binutils-gdb/opcodes/aarch64-opc.c

5741 lines
179 KiB
C
Raw Normal View History

/* aarch64-opc.c -- AArch64 opcode support.
Copyright (C) 2009-2024 Free Software Foundation, Inc.
Contributed by ARM Ltd.
This file is part of the GNU opcodes library.
This library is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 3, or (at your option)
any later version.
It is distributed in the hope that it will be useful, but WITHOUT
ANY WARRANTY; without even the implied warranty of MERCHANTABILITY
or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public
License for more details.
You should have received a copy of the GNU General Public License
along with this program; see the file COPYING3. If not,
see <http://www.gnu.org/licenses/>. */
#include "sysdep.h"
#include <assert.h>
#include <stdlib.h>
#include <stdio.h>
Remove bfd_stdint.h If we require C99 for binutils then stdint.h is available. bfd/ * .gitignore: Delete bfd_stdint.h entry. * Makefile.am (bfdinclude_HEADERS): Delete bfd_stdint.h. (BUILD_HFILES, LOCAL_H_DEPS): Likewise. * bfd-in.h: Include stdint.h in place of bfd_stdint.h. * configure.ac: Don't invoke GCC_HEADER_STDINT. * configure.com: Don't create bfd_stdint.h. * Makefile.in: Regenerate. * aclocal.m4: Regenerate. * bfd-in2.h: Regenerate. * config.in: Regenerate. * configure: Regenerate. * doc/Makefile.in: Regenerate. * po/BLD-POTFILES.in: Regenerate. binutils/ * coffdump.c: Include stdint.h in place of bfd_stdint.h. * dwarf.c: Likewise. gas/ * config/tc-aarch64.c: Include stdint.h in place of bfd_stdint.h. * config/tc-crx.c: Likewise. * config/tc-nds32.h: Likewise. include/ * cgen/basic-modes.h: Include stdint.h in place of bfd_stdint.h. * elf/nfp.h: Likewise. * opcode/aarch64.h: Likewise. * opcode/cgen.h: Likewise. * opcode/nfp.h: Likewise. * opcode/ppc.h: Likewise. ld/ * elf-hints-local.h: Include stdint.h in place of bfd_stdint.h. * emultempl/nds32elf.em: Likewise. * testsuite/ld-elf/mbind2b.c: Likewise. * testsuite/ld-elf/pr18718.c: Likewise. * testsuite/ld-elf/pr18720a.c: Likewise. * testsuite/ld-elf/pr25749-1.c: Likewise. * testsuite/ld-elf/pr25749-1a.c: Likewise. * testsuite/ld-elf/pr25749-1b.c: Likewise. * testsuite/ld-elf/pr25749-1c.c: Likewise. * testsuite/ld-elf/pr25749-1d.c: Likewise. * testsuite/ld-elf/pr25749-2.c: Likewise. * testsuite/ld-elf/pr25754-1a.c: Likewise. * testsuite/ld-elf/pr25754-2a.c: Likewise. * testsuite/ld-elf/pr25754-3a.c: Likewise. * testsuite/ld-elf/pr25754-4a.c: Likewise. * testsuite/ld-elf/pr25754-5a.c: Likewise. * testsuite/ld-elf/pr25754-6a.c: Likewise. opcodes/ * aarch64-dis.c: Include stdint.h in place of bfd_stdint.h. * aarch64-dis.h: Likewise. * aarch64-opc.c: Likewise. * avr-dis.c: Likewise. * csky-dis.c: Likewise. * nds32-asm.c: Likewise. * nds32-dis.c: Likewise. * nfp-dis.c: Likewise. * riscv-dis.c: Likewise. * s12z-dis.c: Likewise. * wasm32-dis.c: Likewise.
2021-03-31 07:37:02 +08:00
#include <stdint.h>
#include <stdarg.h>
#include <inttypes.h>
#include "opintl.h"
[AArch64][SVE 23/32] Add SVE pattern and prfop operands The SVE instructions have two enumerated operands: one to select a vector pattern and another to select a prefetch operation. The latter is a cut-down version of the base AArch64 prefetch operation. Both types of operand can also be specified as raw enum values such as #31. Reserved values can only be specified this way. If it hadn't been for the pattern operand, I would have been tempted to use the existing parsing for prefetch operations and add extra checks for SVE. However, since the patterns needed new enum parsing code anyway, it seeemed cleaner to reuse it for the prefetches too. Because of the small number of enum values, I don't think we'd gain anything by using hash tables. include/ * opcode/aarch64.h (AARCH64_OPND_SVE_PATTERN): New aarch64_opnd. (AARCH64_OPND_SVE_PRFOP): Likewise. (aarch64_sve_pattern_array): Declare. (aarch64_sve_prfop_array): Likewise. opcodes/ * aarch64-tbl.h (AARCH64_OPERANDS): Add entries for AARCH64_OPND_SVE_PATTERN and AARCH64_OPND_SVE_PRFOP. * aarch64-opc.h (FLD_SVE_pattern): New aarch64_field_kind. (FLD_SVE_prfop): Likewise. * aarch64-opc.c: Include libiberty.h. (aarch64_sve_pattern_array): New variable. (aarch64_sve_prfop_array): Likewise. (fields): Add entries for FLD_SVE_pattern and FLD_SVE_prfop. (aarch64_print_operand): Handle AARCH64_OPND_SVE_PATTERN and AARCH64_OPND_SVE_PRFOP. * aarch64-asm-2.c: Regenerate. * aarch64-dis-2.c: Likewise. * aarch64-opc-2.c: Likewise. gas/ * config/tc-aarch64.c (parse_enum_string): New function. (po_enum_or_fail): New macro. (parse_operands): Handle AARCH64_OPND_SVE_PATTERN and AARCH64_OPND_SVE_PRFOP.
2016-09-21 23:54:53 +08:00
#include "libiberty.h"
#include "aarch64-opc.h"
#ifdef DEBUG_AARCH64
int debug_dump = false;
#endif /* DEBUG_AARCH64 */
[AArch64][SVE 23/32] Add SVE pattern and prfop operands The SVE instructions have two enumerated operands: one to select a vector pattern and another to select a prefetch operation. The latter is a cut-down version of the base AArch64 prefetch operation. Both types of operand can also be specified as raw enum values such as #31. Reserved values can only be specified this way. If it hadn't been for the pattern operand, I would have been tempted to use the existing parsing for prefetch operations and add extra checks for SVE. However, since the patterns needed new enum parsing code anyway, it seeemed cleaner to reuse it for the prefetches too. Because of the small number of enum values, I don't think we'd gain anything by using hash tables. include/ * opcode/aarch64.h (AARCH64_OPND_SVE_PATTERN): New aarch64_opnd. (AARCH64_OPND_SVE_PRFOP): Likewise. (aarch64_sve_pattern_array): Declare. (aarch64_sve_prfop_array): Likewise. opcodes/ * aarch64-tbl.h (AARCH64_OPERANDS): Add entries for AARCH64_OPND_SVE_PATTERN and AARCH64_OPND_SVE_PRFOP. * aarch64-opc.h (FLD_SVE_pattern): New aarch64_field_kind. (FLD_SVE_prfop): Likewise. * aarch64-opc.c: Include libiberty.h. (aarch64_sve_pattern_array): New variable. (aarch64_sve_prfop_array): Likewise. (fields): Add entries for FLD_SVE_pattern and FLD_SVE_prfop. (aarch64_print_operand): Handle AARCH64_OPND_SVE_PATTERN and AARCH64_OPND_SVE_PRFOP. * aarch64-asm-2.c: Regenerate. * aarch64-dis-2.c: Likewise. * aarch64-opc-2.c: Likewise. gas/ * config/tc-aarch64.c (parse_enum_string): New function. (po_enum_or_fail): New macro. (parse_operands): Handle AARCH64_OPND_SVE_PATTERN and AARCH64_OPND_SVE_PRFOP.
2016-09-21 23:54:53 +08:00
/* The enumeration strings associated with each value of a 5-bit SVE
pattern operand. A null entry indicates a reserved meaning. */
const char *const aarch64_sve_pattern_array[32] = {
/* 0-7. */
"pow2",
"vl1",
"vl2",
"vl3",
"vl4",
"vl5",
"vl6",
"vl7",
/* 8-15. */
"vl8",
"vl16",
"vl32",
"vl64",
"vl128",
"vl256",
0,
0,
/* 16-23. */
0,
0,
0,
0,
0,
0,
0,
0,
/* 24-31. */
0,
0,
0,
0,
0,
"mul4",
"mul3",
"all"
};
/* The enumeration strings associated with each value of a 4-bit SVE
prefetch operand. A null entry indicates a reserved meaning. */
const char *const aarch64_sve_prfop_array[16] = {
/* 0-7. */
"pldl1keep",
"pldl1strm",
"pldl2keep",
"pldl2strm",
"pldl3keep",
"pldl3strm",
0,
0,
/* 8-15. */
"pstl1keep",
"pstl1strm",
"pstl2keep",
"pstl2strm",
"pstl3keep",
"pstl3strm",
0,
0
};
/* The enumeration strings associated with each value of a 6-bit RPRFM
operation. */
const char *const aarch64_rprfmop_array[64] = {
"pldkeep",
"pstkeep",
0,
0,
"pldstrm",
"pststrm"
};
/* Vector length multiples for a predicate-as-counter operand. Used in things
like AARCH64_OPND_SME_VLxN_10. */
const char *const aarch64_sme_vlxn_array[2] = {
"vlx2",
"vlx4"
};
/* Helper functions to determine which operand to be used to encode/decode
the size:Q fields for AdvSIMD instructions. */
static inline bool
vector_qualifier_p (enum aarch64_opnd_qualifier qualifier)
{
TRUE/FALSE simplification There is really no need to write code like "foo != 0 ? TRUE : FALSE" unless we had stupidly defined FALSE as something other than 0 or TRUE as something other than 1. The simpler "foo != 0" does just as well. Similarly "(condition == TRUE)" or "(condition == FALSE) can be simplified to "(condition)" and "(!condition)" respectively. I'll note that there is reason to use "integer_expression != 0" when assigning a bfd_boolean rather than the simpler "integer_expression", if you expect the variable to have 0 or 1 value. It's probably even a good idea to not rely on implicit conversion if bfd_boolean were _Bool. bfd/ * aoutx.h (aout_link_write_symbols): Don't cast boolean expression to bfd_boolean. * elf32-or1k.c (or1k_set_got_and_rela_sizes): Dont compare booleans against FALSE. * elf32-arc.c (name_for_global_symbol): Don't compare boolean to TRUE. (is_reloc_PC_relative): Don't use "boolean_condition ? TRUE : FALSE". (is_reloc_SDA_relative, is_reloc_for_GOT): Likewise. (is_reloc_for_PLT, is_reloc_for_TLS): Likewise. * elf32-arm.c (stm32l4xx_need_create_replacing_stub): Likewise. * elf32-nds32.c (insert_nds32_elf_blank): Likewise. * elf32-rx.c (rx_set_section_contents): Likewise. * elfnn-aarch64.c (elfNN_aarch64_final_link_relocate): Likewise. * elfxx-mips.c (_bfd_mips_elf_ignore_undef_symbol): Likewise. * mach-o.c (bfd_mach_o_read_command): Likewise. * targets.c (bfd_get_target_info): Likewise. binutils/ * dlltool.c (main): Don't use "boolean_condition ? TRUE : FALSE". * dwarf.c (read_and_display_attr_value): Likewise. (display_debug_str_offsets): Likewise. * objdump.c (dump_bfd): Likewise. * readelf.c (dump_section_as_strings): Likewise. (dump_section_as_bytes): Likewise. gas/ * atof-generic.c (FALSE, TRUE): Don't define. * config/obj-elf.h (FALSE, TRUE): Don't define. * config/obj-som.h (FALSE, TRUE): Don't define. * config/tc-hppa.h (FALSE, TRUE): Don't define. * config/tc-pdp11.c (FALSE, TRUE): Don't define. * config/tc-iq2000.h (obj_fix_adjustable): Delete. * config/tc-m32r.h (TC_FIX_ADJUSTABLE): Delete. * config/tc-mt.h (obj_fix_adjustable): Delete. * config/tc-nds32.h (TC_FIX_ADJUSTABLE): Delete. * config/tc-arc.c (parse_opcode_flags): Simplify boolean expression. (relaxable_flag, relaxable_operand, assemble_insn): Likewise. (tokenize_extregister): Likewise. * config/tc-csky.c (parse_opcode, get_operand_value): Likewise. (parse_operands_op, parse_operands, md_assemble): Likewise. * config/tc-d10v.c (build_insn): Likewise. * config/tc-score.c (s3_gen_insn_frag): Likewise. * config/tc-score7.c (s7_gen_insn_frag, s7_relax_frag): Likewise. * config/tc-tic6x.c (tic6x_update_features, md_assemble): Likewise. * config/tc-z80.c (emit_byte): Likewise. include/ * opcode/aarch64.h (alias_opcode_p): Simplify boolean expression. (opcode_has_alias, pseudo_opcode_p, optional_operand_p): Likewise. (opcode_has_special_coder): Likewise. ld/ * emultempl/aix.em (gld${EMULATION_NAME}_before_allocation): Simplify boolean expression. * lexsup.c (parse_args): Likewise. * pe-dll.c (pe_dll_id_target): Likewise. opcodes/ * aarch64-opc.c (vector_qualifier_p): Simplify boolean expression. (fp_qualifier_p, get_data_pattern): Likewise. (aarch64_get_operand_modifier_from_value): Likewise. (aarch64_extend_operator_p, aarch64_shift_operator_p): Likewise. (operand_variant_qualifier_p): Likewise. (qualifier_value_in_range_constraint_p): Likewise. (aarch64_get_qualifier_esize): Likewise. (aarch64_get_qualifier_nelem): Likewise. (aarch64_get_qualifier_standard_value): Likewise. (get_lower_bound, get_upper_bound): Likewise. (aarch64_find_best_match, match_operands_qualifier): Likewise. (aarch64_print_operand): Likewise. * aarch64-opc.h (operand_has_inserter, operand_has_extractor): Likewise. (operand_need_sign_extension, operand_need_shift_by_two): Likewise. (operand_need_shift_by_four, operand_maybe_stack_pointer): Likewise. * arm-dis.c (print_insn_mve, print_insn_thumb32): Likewise. * tic6x-dis.c (tic6x_check_fetch_packet_header): Likewise. (print_insn_tic6x): Likewise.
2021-03-29 07:22:56 +08:00
return (qualifier >= AARCH64_OPND_QLF_V_8B
&& qualifier <= AARCH64_OPND_QLF_V_1Q);
}
static inline bool
fp_qualifier_p (enum aarch64_opnd_qualifier qualifier)
{
TRUE/FALSE simplification There is really no need to write code like "foo != 0 ? TRUE : FALSE" unless we had stupidly defined FALSE as something other than 0 or TRUE as something other than 1. The simpler "foo != 0" does just as well. Similarly "(condition == TRUE)" or "(condition == FALSE) can be simplified to "(condition)" and "(!condition)" respectively. I'll note that there is reason to use "integer_expression != 0" when assigning a bfd_boolean rather than the simpler "integer_expression", if you expect the variable to have 0 or 1 value. It's probably even a good idea to not rely on implicit conversion if bfd_boolean were _Bool. bfd/ * aoutx.h (aout_link_write_symbols): Don't cast boolean expression to bfd_boolean. * elf32-or1k.c (or1k_set_got_and_rela_sizes): Dont compare booleans against FALSE. * elf32-arc.c (name_for_global_symbol): Don't compare boolean to TRUE. (is_reloc_PC_relative): Don't use "boolean_condition ? TRUE : FALSE". (is_reloc_SDA_relative, is_reloc_for_GOT): Likewise. (is_reloc_for_PLT, is_reloc_for_TLS): Likewise. * elf32-arm.c (stm32l4xx_need_create_replacing_stub): Likewise. * elf32-nds32.c (insert_nds32_elf_blank): Likewise. * elf32-rx.c (rx_set_section_contents): Likewise. * elfnn-aarch64.c (elfNN_aarch64_final_link_relocate): Likewise. * elfxx-mips.c (_bfd_mips_elf_ignore_undef_symbol): Likewise. * mach-o.c (bfd_mach_o_read_command): Likewise. * targets.c (bfd_get_target_info): Likewise. binutils/ * dlltool.c (main): Don't use "boolean_condition ? TRUE : FALSE". * dwarf.c (read_and_display_attr_value): Likewise. (display_debug_str_offsets): Likewise. * objdump.c (dump_bfd): Likewise. * readelf.c (dump_section_as_strings): Likewise. (dump_section_as_bytes): Likewise. gas/ * atof-generic.c (FALSE, TRUE): Don't define. * config/obj-elf.h (FALSE, TRUE): Don't define. * config/obj-som.h (FALSE, TRUE): Don't define. * config/tc-hppa.h (FALSE, TRUE): Don't define. * config/tc-pdp11.c (FALSE, TRUE): Don't define. * config/tc-iq2000.h (obj_fix_adjustable): Delete. * config/tc-m32r.h (TC_FIX_ADJUSTABLE): Delete. * config/tc-mt.h (obj_fix_adjustable): Delete. * config/tc-nds32.h (TC_FIX_ADJUSTABLE): Delete. * config/tc-arc.c (parse_opcode_flags): Simplify boolean expression. (relaxable_flag, relaxable_operand, assemble_insn): Likewise. (tokenize_extregister): Likewise. * config/tc-csky.c (parse_opcode, get_operand_value): Likewise. (parse_operands_op, parse_operands, md_assemble): Likewise. * config/tc-d10v.c (build_insn): Likewise. * config/tc-score.c (s3_gen_insn_frag): Likewise. * config/tc-score7.c (s7_gen_insn_frag, s7_relax_frag): Likewise. * config/tc-tic6x.c (tic6x_update_features, md_assemble): Likewise. * config/tc-z80.c (emit_byte): Likewise. include/ * opcode/aarch64.h (alias_opcode_p): Simplify boolean expression. (opcode_has_alias, pseudo_opcode_p, optional_operand_p): Likewise. (opcode_has_special_coder): Likewise. ld/ * emultempl/aix.em (gld${EMULATION_NAME}_before_allocation): Simplify boolean expression. * lexsup.c (parse_args): Likewise. * pe-dll.c (pe_dll_id_target): Likewise. opcodes/ * aarch64-opc.c (vector_qualifier_p): Simplify boolean expression. (fp_qualifier_p, get_data_pattern): Likewise. (aarch64_get_operand_modifier_from_value): Likewise. (aarch64_extend_operator_p, aarch64_shift_operator_p): Likewise. (operand_variant_qualifier_p): Likewise. (qualifier_value_in_range_constraint_p): Likewise. (aarch64_get_qualifier_esize): Likewise. (aarch64_get_qualifier_nelem): Likewise. (aarch64_get_qualifier_standard_value): Likewise. (get_lower_bound, get_upper_bound): Likewise. (aarch64_find_best_match, match_operands_qualifier): Likewise. (aarch64_print_operand): Likewise. * aarch64-opc.h (operand_has_inserter, operand_has_extractor): Likewise. (operand_need_sign_extension, operand_need_shift_by_two): Likewise. (operand_need_shift_by_four, operand_maybe_stack_pointer): Likewise. * arm-dis.c (print_insn_mve, print_insn_thumb32): Likewise. * tic6x-dis.c (tic6x_check_fetch_packet_header): Likewise. (print_insn_tic6x): Likewise.
2021-03-29 07:22:56 +08:00
return (qualifier >= AARCH64_OPND_QLF_S_B
&& qualifier <= AARCH64_OPND_QLF_S_Q);
}
enum data_pattern
{
DP_UNKNOWN,
DP_VECTOR_3SAME,
DP_VECTOR_LONG,
DP_VECTOR_WIDE,
DP_VECTOR_ACROSS_LANES,
};
static const char significant_operand_index [] =
{
0, /* DP_UNKNOWN, by default using operand 0. */
0, /* DP_VECTOR_3SAME */
1, /* DP_VECTOR_LONG */
2, /* DP_VECTOR_WIDE */
1, /* DP_VECTOR_ACROSS_LANES */
};
/* Given a sequence of qualifiers in QUALIFIERS, determine and return
the data pattern.
N.B. QUALIFIERS is a possible sequence of qualifiers each of which
corresponds to one of a sequence of operands. */
static enum data_pattern
get_data_pattern (const aarch64_opnd_qualifier_seq_t qualifiers)
{
TRUE/FALSE simplification There is really no need to write code like "foo != 0 ? TRUE : FALSE" unless we had stupidly defined FALSE as something other than 0 or TRUE as something other than 1. The simpler "foo != 0" does just as well. Similarly "(condition == TRUE)" or "(condition == FALSE) can be simplified to "(condition)" and "(!condition)" respectively. I'll note that there is reason to use "integer_expression != 0" when assigning a bfd_boolean rather than the simpler "integer_expression", if you expect the variable to have 0 or 1 value. It's probably even a good idea to not rely on implicit conversion if bfd_boolean were _Bool. bfd/ * aoutx.h (aout_link_write_symbols): Don't cast boolean expression to bfd_boolean. * elf32-or1k.c (or1k_set_got_and_rela_sizes): Dont compare booleans against FALSE. * elf32-arc.c (name_for_global_symbol): Don't compare boolean to TRUE. (is_reloc_PC_relative): Don't use "boolean_condition ? TRUE : FALSE". (is_reloc_SDA_relative, is_reloc_for_GOT): Likewise. (is_reloc_for_PLT, is_reloc_for_TLS): Likewise. * elf32-arm.c (stm32l4xx_need_create_replacing_stub): Likewise. * elf32-nds32.c (insert_nds32_elf_blank): Likewise. * elf32-rx.c (rx_set_section_contents): Likewise. * elfnn-aarch64.c (elfNN_aarch64_final_link_relocate): Likewise. * elfxx-mips.c (_bfd_mips_elf_ignore_undef_symbol): Likewise. * mach-o.c (bfd_mach_o_read_command): Likewise. * targets.c (bfd_get_target_info): Likewise. binutils/ * dlltool.c (main): Don't use "boolean_condition ? TRUE : FALSE". * dwarf.c (read_and_display_attr_value): Likewise. (display_debug_str_offsets): Likewise. * objdump.c (dump_bfd): Likewise. * readelf.c (dump_section_as_strings): Likewise. (dump_section_as_bytes): Likewise. gas/ * atof-generic.c (FALSE, TRUE): Don't define. * config/obj-elf.h (FALSE, TRUE): Don't define. * config/obj-som.h (FALSE, TRUE): Don't define. * config/tc-hppa.h (FALSE, TRUE): Don't define. * config/tc-pdp11.c (FALSE, TRUE): Don't define. * config/tc-iq2000.h (obj_fix_adjustable): Delete. * config/tc-m32r.h (TC_FIX_ADJUSTABLE): Delete. * config/tc-mt.h (obj_fix_adjustable): Delete. * config/tc-nds32.h (TC_FIX_ADJUSTABLE): Delete. * config/tc-arc.c (parse_opcode_flags): Simplify boolean expression. (relaxable_flag, relaxable_operand, assemble_insn): Likewise. (tokenize_extregister): Likewise. * config/tc-csky.c (parse_opcode, get_operand_value): Likewise. (parse_operands_op, parse_operands, md_assemble): Likewise. * config/tc-d10v.c (build_insn): Likewise. * config/tc-score.c (s3_gen_insn_frag): Likewise. * config/tc-score7.c (s7_gen_insn_frag, s7_relax_frag): Likewise. * config/tc-tic6x.c (tic6x_update_features, md_assemble): Likewise. * config/tc-z80.c (emit_byte): Likewise. include/ * opcode/aarch64.h (alias_opcode_p): Simplify boolean expression. (opcode_has_alias, pseudo_opcode_p, optional_operand_p): Likewise. (opcode_has_special_coder): Likewise. ld/ * emultempl/aix.em (gld${EMULATION_NAME}_before_allocation): Simplify boolean expression. * lexsup.c (parse_args): Likewise. * pe-dll.c (pe_dll_id_target): Likewise. opcodes/ * aarch64-opc.c (vector_qualifier_p): Simplify boolean expression. (fp_qualifier_p, get_data_pattern): Likewise. (aarch64_get_operand_modifier_from_value): Likewise. (aarch64_extend_operator_p, aarch64_shift_operator_p): Likewise. (operand_variant_qualifier_p): Likewise. (qualifier_value_in_range_constraint_p): Likewise. (aarch64_get_qualifier_esize): Likewise. (aarch64_get_qualifier_nelem): Likewise. (aarch64_get_qualifier_standard_value): Likewise. (get_lower_bound, get_upper_bound): Likewise. (aarch64_find_best_match, match_operands_qualifier): Likewise. (aarch64_print_operand): Likewise. * aarch64-opc.h (operand_has_inserter, operand_has_extractor): Likewise. (operand_need_sign_extension, operand_need_shift_by_two): Likewise. (operand_need_shift_by_four, operand_maybe_stack_pointer): Likewise. * arm-dis.c (print_insn_mve, print_insn_thumb32): Likewise. * tic6x-dis.c (tic6x_check_fetch_packet_header): Likewise. (print_insn_tic6x): Likewise.
2021-03-29 07:22:56 +08:00
if (vector_qualifier_p (qualifiers[0]))
{
/* e.g. v.4s, v.4s, v.4s
or v.4h, v.4h, v.h[3]. */
if (qualifiers[0] == qualifiers[1]
TRUE/FALSE simplification There is really no need to write code like "foo != 0 ? TRUE : FALSE" unless we had stupidly defined FALSE as something other than 0 or TRUE as something other than 1. The simpler "foo != 0" does just as well. Similarly "(condition == TRUE)" or "(condition == FALSE) can be simplified to "(condition)" and "(!condition)" respectively. I'll note that there is reason to use "integer_expression != 0" when assigning a bfd_boolean rather than the simpler "integer_expression", if you expect the variable to have 0 or 1 value. It's probably even a good idea to not rely on implicit conversion if bfd_boolean were _Bool. bfd/ * aoutx.h (aout_link_write_symbols): Don't cast boolean expression to bfd_boolean. * elf32-or1k.c (or1k_set_got_and_rela_sizes): Dont compare booleans against FALSE. * elf32-arc.c (name_for_global_symbol): Don't compare boolean to TRUE. (is_reloc_PC_relative): Don't use "boolean_condition ? TRUE : FALSE". (is_reloc_SDA_relative, is_reloc_for_GOT): Likewise. (is_reloc_for_PLT, is_reloc_for_TLS): Likewise. * elf32-arm.c (stm32l4xx_need_create_replacing_stub): Likewise. * elf32-nds32.c (insert_nds32_elf_blank): Likewise. * elf32-rx.c (rx_set_section_contents): Likewise. * elfnn-aarch64.c (elfNN_aarch64_final_link_relocate): Likewise. * elfxx-mips.c (_bfd_mips_elf_ignore_undef_symbol): Likewise. * mach-o.c (bfd_mach_o_read_command): Likewise. * targets.c (bfd_get_target_info): Likewise. binutils/ * dlltool.c (main): Don't use "boolean_condition ? TRUE : FALSE". * dwarf.c (read_and_display_attr_value): Likewise. (display_debug_str_offsets): Likewise. * objdump.c (dump_bfd): Likewise. * readelf.c (dump_section_as_strings): Likewise. (dump_section_as_bytes): Likewise. gas/ * atof-generic.c (FALSE, TRUE): Don't define. * config/obj-elf.h (FALSE, TRUE): Don't define. * config/obj-som.h (FALSE, TRUE): Don't define. * config/tc-hppa.h (FALSE, TRUE): Don't define. * config/tc-pdp11.c (FALSE, TRUE): Don't define. * config/tc-iq2000.h (obj_fix_adjustable): Delete. * config/tc-m32r.h (TC_FIX_ADJUSTABLE): Delete. * config/tc-mt.h (obj_fix_adjustable): Delete. * config/tc-nds32.h (TC_FIX_ADJUSTABLE): Delete. * config/tc-arc.c (parse_opcode_flags): Simplify boolean expression. (relaxable_flag, relaxable_operand, assemble_insn): Likewise. (tokenize_extregister): Likewise. * config/tc-csky.c (parse_opcode, get_operand_value): Likewise. (parse_operands_op, parse_operands, md_assemble): Likewise. * config/tc-d10v.c (build_insn): Likewise. * config/tc-score.c (s3_gen_insn_frag): Likewise. * config/tc-score7.c (s7_gen_insn_frag, s7_relax_frag): Likewise. * config/tc-tic6x.c (tic6x_update_features, md_assemble): Likewise. * config/tc-z80.c (emit_byte): Likewise. include/ * opcode/aarch64.h (alias_opcode_p): Simplify boolean expression. (opcode_has_alias, pseudo_opcode_p, optional_operand_p): Likewise. (opcode_has_special_coder): Likewise. ld/ * emultempl/aix.em (gld${EMULATION_NAME}_before_allocation): Simplify boolean expression. * lexsup.c (parse_args): Likewise. * pe-dll.c (pe_dll_id_target): Likewise. opcodes/ * aarch64-opc.c (vector_qualifier_p): Simplify boolean expression. (fp_qualifier_p, get_data_pattern): Likewise. (aarch64_get_operand_modifier_from_value): Likewise. (aarch64_extend_operator_p, aarch64_shift_operator_p): Likewise. (operand_variant_qualifier_p): Likewise. (qualifier_value_in_range_constraint_p): Likewise. (aarch64_get_qualifier_esize): Likewise. (aarch64_get_qualifier_nelem): Likewise. (aarch64_get_qualifier_standard_value): Likewise. (get_lower_bound, get_upper_bound): Likewise. (aarch64_find_best_match, match_operands_qualifier): Likewise. (aarch64_print_operand): Likewise. * aarch64-opc.h (operand_has_inserter, operand_has_extractor): Likewise. (operand_need_sign_extension, operand_need_shift_by_two): Likewise. (operand_need_shift_by_four, operand_maybe_stack_pointer): Likewise. * arm-dis.c (print_insn_mve, print_insn_thumb32): Likewise. * tic6x-dis.c (tic6x_check_fetch_packet_header): Likewise. (print_insn_tic6x): Likewise.
2021-03-29 07:22:56 +08:00
&& vector_qualifier_p (qualifiers[2])
&& (aarch64_get_qualifier_esize (qualifiers[0])
== aarch64_get_qualifier_esize (qualifiers[1]))
&& (aarch64_get_qualifier_esize (qualifiers[0])
== aarch64_get_qualifier_esize (qualifiers[2])))
return DP_VECTOR_3SAME;
/* e.g. v.8h, v.8b, v.8b.
or v.4s, v.4h, v.h[2].
or v.8h, v.16b. */
TRUE/FALSE simplification There is really no need to write code like "foo != 0 ? TRUE : FALSE" unless we had stupidly defined FALSE as something other than 0 or TRUE as something other than 1. The simpler "foo != 0" does just as well. Similarly "(condition == TRUE)" or "(condition == FALSE) can be simplified to "(condition)" and "(!condition)" respectively. I'll note that there is reason to use "integer_expression != 0" when assigning a bfd_boolean rather than the simpler "integer_expression", if you expect the variable to have 0 or 1 value. It's probably even a good idea to not rely on implicit conversion if bfd_boolean were _Bool. bfd/ * aoutx.h (aout_link_write_symbols): Don't cast boolean expression to bfd_boolean. * elf32-or1k.c (or1k_set_got_and_rela_sizes): Dont compare booleans against FALSE. * elf32-arc.c (name_for_global_symbol): Don't compare boolean to TRUE. (is_reloc_PC_relative): Don't use "boolean_condition ? TRUE : FALSE". (is_reloc_SDA_relative, is_reloc_for_GOT): Likewise. (is_reloc_for_PLT, is_reloc_for_TLS): Likewise. * elf32-arm.c (stm32l4xx_need_create_replacing_stub): Likewise. * elf32-nds32.c (insert_nds32_elf_blank): Likewise. * elf32-rx.c (rx_set_section_contents): Likewise. * elfnn-aarch64.c (elfNN_aarch64_final_link_relocate): Likewise. * elfxx-mips.c (_bfd_mips_elf_ignore_undef_symbol): Likewise. * mach-o.c (bfd_mach_o_read_command): Likewise. * targets.c (bfd_get_target_info): Likewise. binutils/ * dlltool.c (main): Don't use "boolean_condition ? TRUE : FALSE". * dwarf.c (read_and_display_attr_value): Likewise. (display_debug_str_offsets): Likewise. * objdump.c (dump_bfd): Likewise. * readelf.c (dump_section_as_strings): Likewise. (dump_section_as_bytes): Likewise. gas/ * atof-generic.c (FALSE, TRUE): Don't define. * config/obj-elf.h (FALSE, TRUE): Don't define. * config/obj-som.h (FALSE, TRUE): Don't define. * config/tc-hppa.h (FALSE, TRUE): Don't define. * config/tc-pdp11.c (FALSE, TRUE): Don't define. * config/tc-iq2000.h (obj_fix_adjustable): Delete. * config/tc-m32r.h (TC_FIX_ADJUSTABLE): Delete. * config/tc-mt.h (obj_fix_adjustable): Delete. * config/tc-nds32.h (TC_FIX_ADJUSTABLE): Delete. * config/tc-arc.c (parse_opcode_flags): Simplify boolean expression. (relaxable_flag, relaxable_operand, assemble_insn): Likewise. (tokenize_extregister): Likewise. * config/tc-csky.c (parse_opcode, get_operand_value): Likewise. (parse_operands_op, parse_operands, md_assemble): Likewise. * config/tc-d10v.c (build_insn): Likewise. * config/tc-score.c (s3_gen_insn_frag): Likewise. * config/tc-score7.c (s7_gen_insn_frag, s7_relax_frag): Likewise. * config/tc-tic6x.c (tic6x_update_features, md_assemble): Likewise. * config/tc-z80.c (emit_byte): Likewise. include/ * opcode/aarch64.h (alias_opcode_p): Simplify boolean expression. (opcode_has_alias, pseudo_opcode_p, optional_operand_p): Likewise. (opcode_has_special_coder): Likewise. ld/ * emultempl/aix.em (gld${EMULATION_NAME}_before_allocation): Simplify boolean expression. * lexsup.c (parse_args): Likewise. * pe-dll.c (pe_dll_id_target): Likewise. opcodes/ * aarch64-opc.c (vector_qualifier_p): Simplify boolean expression. (fp_qualifier_p, get_data_pattern): Likewise. (aarch64_get_operand_modifier_from_value): Likewise. (aarch64_extend_operator_p, aarch64_shift_operator_p): Likewise. (operand_variant_qualifier_p): Likewise. (qualifier_value_in_range_constraint_p): Likewise. (aarch64_get_qualifier_esize): Likewise. (aarch64_get_qualifier_nelem): Likewise. (aarch64_get_qualifier_standard_value): Likewise. (get_lower_bound, get_upper_bound): Likewise. (aarch64_find_best_match, match_operands_qualifier): Likewise. (aarch64_print_operand): Likewise. * aarch64-opc.h (operand_has_inserter, operand_has_extractor): Likewise. (operand_need_sign_extension, operand_need_shift_by_two): Likewise. (operand_need_shift_by_four, operand_maybe_stack_pointer): Likewise. * arm-dis.c (print_insn_mve, print_insn_thumb32): Likewise. * tic6x-dis.c (tic6x_check_fetch_packet_header): Likewise. (print_insn_tic6x): Likewise.
2021-03-29 07:22:56 +08:00
if (vector_qualifier_p (qualifiers[1])
&& aarch64_get_qualifier_esize (qualifiers[0]) != 0
&& (aarch64_get_qualifier_esize (qualifiers[0])
== aarch64_get_qualifier_esize (qualifiers[1]) << 1))
return DP_VECTOR_LONG;
/* e.g. v.8h, v.8h, v.8b. */
if (qualifiers[0] == qualifiers[1]
TRUE/FALSE simplification There is really no need to write code like "foo != 0 ? TRUE : FALSE" unless we had stupidly defined FALSE as something other than 0 or TRUE as something other than 1. The simpler "foo != 0" does just as well. Similarly "(condition == TRUE)" or "(condition == FALSE) can be simplified to "(condition)" and "(!condition)" respectively. I'll note that there is reason to use "integer_expression != 0" when assigning a bfd_boolean rather than the simpler "integer_expression", if you expect the variable to have 0 or 1 value. It's probably even a good idea to not rely on implicit conversion if bfd_boolean were _Bool. bfd/ * aoutx.h (aout_link_write_symbols): Don't cast boolean expression to bfd_boolean. * elf32-or1k.c (or1k_set_got_and_rela_sizes): Dont compare booleans against FALSE. * elf32-arc.c (name_for_global_symbol): Don't compare boolean to TRUE. (is_reloc_PC_relative): Don't use "boolean_condition ? TRUE : FALSE". (is_reloc_SDA_relative, is_reloc_for_GOT): Likewise. (is_reloc_for_PLT, is_reloc_for_TLS): Likewise. * elf32-arm.c (stm32l4xx_need_create_replacing_stub): Likewise. * elf32-nds32.c (insert_nds32_elf_blank): Likewise. * elf32-rx.c (rx_set_section_contents): Likewise. * elfnn-aarch64.c (elfNN_aarch64_final_link_relocate): Likewise. * elfxx-mips.c (_bfd_mips_elf_ignore_undef_symbol): Likewise. * mach-o.c (bfd_mach_o_read_command): Likewise. * targets.c (bfd_get_target_info): Likewise. binutils/ * dlltool.c (main): Don't use "boolean_condition ? TRUE : FALSE". * dwarf.c (read_and_display_attr_value): Likewise. (display_debug_str_offsets): Likewise. * objdump.c (dump_bfd): Likewise. * readelf.c (dump_section_as_strings): Likewise. (dump_section_as_bytes): Likewise. gas/ * atof-generic.c (FALSE, TRUE): Don't define. * config/obj-elf.h (FALSE, TRUE): Don't define. * config/obj-som.h (FALSE, TRUE): Don't define. * config/tc-hppa.h (FALSE, TRUE): Don't define. * config/tc-pdp11.c (FALSE, TRUE): Don't define. * config/tc-iq2000.h (obj_fix_adjustable): Delete. * config/tc-m32r.h (TC_FIX_ADJUSTABLE): Delete. * config/tc-mt.h (obj_fix_adjustable): Delete. * config/tc-nds32.h (TC_FIX_ADJUSTABLE): Delete. * config/tc-arc.c (parse_opcode_flags): Simplify boolean expression. (relaxable_flag, relaxable_operand, assemble_insn): Likewise. (tokenize_extregister): Likewise. * config/tc-csky.c (parse_opcode, get_operand_value): Likewise. (parse_operands_op, parse_operands, md_assemble): Likewise. * config/tc-d10v.c (build_insn): Likewise. * config/tc-score.c (s3_gen_insn_frag): Likewise. * config/tc-score7.c (s7_gen_insn_frag, s7_relax_frag): Likewise. * config/tc-tic6x.c (tic6x_update_features, md_assemble): Likewise. * config/tc-z80.c (emit_byte): Likewise. include/ * opcode/aarch64.h (alias_opcode_p): Simplify boolean expression. (opcode_has_alias, pseudo_opcode_p, optional_operand_p): Likewise. (opcode_has_special_coder): Likewise. ld/ * emultempl/aix.em (gld${EMULATION_NAME}_before_allocation): Simplify boolean expression. * lexsup.c (parse_args): Likewise. * pe-dll.c (pe_dll_id_target): Likewise. opcodes/ * aarch64-opc.c (vector_qualifier_p): Simplify boolean expression. (fp_qualifier_p, get_data_pattern): Likewise. (aarch64_get_operand_modifier_from_value): Likewise. (aarch64_extend_operator_p, aarch64_shift_operator_p): Likewise. (operand_variant_qualifier_p): Likewise. (qualifier_value_in_range_constraint_p): Likewise. (aarch64_get_qualifier_esize): Likewise. (aarch64_get_qualifier_nelem): Likewise. (aarch64_get_qualifier_standard_value): Likewise. (get_lower_bound, get_upper_bound): Likewise. (aarch64_find_best_match, match_operands_qualifier): Likewise. (aarch64_print_operand): Likewise. * aarch64-opc.h (operand_has_inserter, operand_has_extractor): Likewise. (operand_need_sign_extension, operand_need_shift_by_two): Likewise. (operand_need_shift_by_four, operand_maybe_stack_pointer): Likewise. * arm-dis.c (print_insn_mve, print_insn_thumb32): Likewise. * tic6x-dis.c (tic6x_check_fetch_packet_header): Likewise. (print_insn_tic6x): Likewise.
2021-03-29 07:22:56 +08:00
&& vector_qualifier_p (qualifiers[2])
&& aarch64_get_qualifier_esize (qualifiers[0]) != 0
&& (aarch64_get_qualifier_esize (qualifiers[0])
== aarch64_get_qualifier_esize (qualifiers[2]) << 1)
&& (aarch64_get_qualifier_esize (qualifiers[0])
== aarch64_get_qualifier_esize (qualifiers[1])))
return DP_VECTOR_WIDE;
}
TRUE/FALSE simplification There is really no need to write code like "foo != 0 ? TRUE : FALSE" unless we had stupidly defined FALSE as something other than 0 or TRUE as something other than 1. The simpler "foo != 0" does just as well. Similarly "(condition == TRUE)" or "(condition == FALSE) can be simplified to "(condition)" and "(!condition)" respectively. I'll note that there is reason to use "integer_expression != 0" when assigning a bfd_boolean rather than the simpler "integer_expression", if you expect the variable to have 0 or 1 value. It's probably even a good idea to not rely on implicit conversion if bfd_boolean were _Bool. bfd/ * aoutx.h (aout_link_write_symbols): Don't cast boolean expression to bfd_boolean. * elf32-or1k.c (or1k_set_got_and_rela_sizes): Dont compare booleans against FALSE. * elf32-arc.c (name_for_global_symbol): Don't compare boolean to TRUE. (is_reloc_PC_relative): Don't use "boolean_condition ? TRUE : FALSE". (is_reloc_SDA_relative, is_reloc_for_GOT): Likewise. (is_reloc_for_PLT, is_reloc_for_TLS): Likewise. * elf32-arm.c (stm32l4xx_need_create_replacing_stub): Likewise. * elf32-nds32.c (insert_nds32_elf_blank): Likewise. * elf32-rx.c (rx_set_section_contents): Likewise. * elfnn-aarch64.c (elfNN_aarch64_final_link_relocate): Likewise. * elfxx-mips.c (_bfd_mips_elf_ignore_undef_symbol): Likewise. * mach-o.c (bfd_mach_o_read_command): Likewise. * targets.c (bfd_get_target_info): Likewise. binutils/ * dlltool.c (main): Don't use "boolean_condition ? TRUE : FALSE". * dwarf.c (read_and_display_attr_value): Likewise. (display_debug_str_offsets): Likewise. * objdump.c (dump_bfd): Likewise. * readelf.c (dump_section_as_strings): Likewise. (dump_section_as_bytes): Likewise. gas/ * atof-generic.c (FALSE, TRUE): Don't define. * config/obj-elf.h (FALSE, TRUE): Don't define. * config/obj-som.h (FALSE, TRUE): Don't define. * config/tc-hppa.h (FALSE, TRUE): Don't define. * config/tc-pdp11.c (FALSE, TRUE): Don't define. * config/tc-iq2000.h (obj_fix_adjustable): Delete. * config/tc-m32r.h (TC_FIX_ADJUSTABLE): Delete. * config/tc-mt.h (obj_fix_adjustable): Delete. * config/tc-nds32.h (TC_FIX_ADJUSTABLE): Delete. * config/tc-arc.c (parse_opcode_flags): Simplify boolean expression. (relaxable_flag, relaxable_operand, assemble_insn): Likewise. (tokenize_extregister): Likewise. * config/tc-csky.c (parse_opcode, get_operand_value): Likewise. (parse_operands_op, parse_operands, md_assemble): Likewise. * config/tc-d10v.c (build_insn): Likewise. * config/tc-score.c (s3_gen_insn_frag): Likewise. * config/tc-score7.c (s7_gen_insn_frag, s7_relax_frag): Likewise. * config/tc-tic6x.c (tic6x_update_features, md_assemble): Likewise. * config/tc-z80.c (emit_byte): Likewise. include/ * opcode/aarch64.h (alias_opcode_p): Simplify boolean expression. (opcode_has_alias, pseudo_opcode_p, optional_operand_p): Likewise. (opcode_has_special_coder): Likewise. ld/ * emultempl/aix.em (gld${EMULATION_NAME}_before_allocation): Simplify boolean expression. * lexsup.c (parse_args): Likewise. * pe-dll.c (pe_dll_id_target): Likewise. opcodes/ * aarch64-opc.c (vector_qualifier_p): Simplify boolean expression. (fp_qualifier_p, get_data_pattern): Likewise. (aarch64_get_operand_modifier_from_value): Likewise. (aarch64_extend_operator_p, aarch64_shift_operator_p): Likewise. (operand_variant_qualifier_p): Likewise. (qualifier_value_in_range_constraint_p): Likewise. (aarch64_get_qualifier_esize): Likewise. (aarch64_get_qualifier_nelem): Likewise. (aarch64_get_qualifier_standard_value): Likewise. (get_lower_bound, get_upper_bound): Likewise. (aarch64_find_best_match, match_operands_qualifier): Likewise. (aarch64_print_operand): Likewise. * aarch64-opc.h (operand_has_inserter, operand_has_extractor): Likewise. (operand_need_sign_extension, operand_need_shift_by_two): Likewise. (operand_need_shift_by_four, operand_maybe_stack_pointer): Likewise. * arm-dis.c (print_insn_mve, print_insn_thumb32): Likewise. * tic6x-dis.c (tic6x_check_fetch_packet_header): Likewise. (print_insn_tic6x): Likewise.
2021-03-29 07:22:56 +08:00
else if (fp_qualifier_p (qualifiers[0]))
{
/* e.g. SADDLV <V><d>, <Vn>.<T>. */
TRUE/FALSE simplification There is really no need to write code like "foo != 0 ? TRUE : FALSE" unless we had stupidly defined FALSE as something other than 0 or TRUE as something other than 1. The simpler "foo != 0" does just as well. Similarly "(condition == TRUE)" or "(condition == FALSE) can be simplified to "(condition)" and "(!condition)" respectively. I'll note that there is reason to use "integer_expression != 0" when assigning a bfd_boolean rather than the simpler "integer_expression", if you expect the variable to have 0 or 1 value. It's probably even a good idea to not rely on implicit conversion if bfd_boolean were _Bool. bfd/ * aoutx.h (aout_link_write_symbols): Don't cast boolean expression to bfd_boolean. * elf32-or1k.c (or1k_set_got_and_rela_sizes): Dont compare booleans against FALSE. * elf32-arc.c (name_for_global_symbol): Don't compare boolean to TRUE. (is_reloc_PC_relative): Don't use "boolean_condition ? TRUE : FALSE". (is_reloc_SDA_relative, is_reloc_for_GOT): Likewise. (is_reloc_for_PLT, is_reloc_for_TLS): Likewise. * elf32-arm.c (stm32l4xx_need_create_replacing_stub): Likewise. * elf32-nds32.c (insert_nds32_elf_blank): Likewise. * elf32-rx.c (rx_set_section_contents): Likewise. * elfnn-aarch64.c (elfNN_aarch64_final_link_relocate): Likewise. * elfxx-mips.c (_bfd_mips_elf_ignore_undef_symbol): Likewise. * mach-o.c (bfd_mach_o_read_command): Likewise. * targets.c (bfd_get_target_info): Likewise. binutils/ * dlltool.c (main): Don't use "boolean_condition ? TRUE : FALSE". * dwarf.c (read_and_display_attr_value): Likewise. (display_debug_str_offsets): Likewise. * objdump.c (dump_bfd): Likewise. * readelf.c (dump_section_as_strings): Likewise. (dump_section_as_bytes): Likewise. gas/ * atof-generic.c (FALSE, TRUE): Don't define. * config/obj-elf.h (FALSE, TRUE): Don't define. * config/obj-som.h (FALSE, TRUE): Don't define. * config/tc-hppa.h (FALSE, TRUE): Don't define. * config/tc-pdp11.c (FALSE, TRUE): Don't define. * config/tc-iq2000.h (obj_fix_adjustable): Delete. * config/tc-m32r.h (TC_FIX_ADJUSTABLE): Delete. * config/tc-mt.h (obj_fix_adjustable): Delete. * config/tc-nds32.h (TC_FIX_ADJUSTABLE): Delete. * config/tc-arc.c (parse_opcode_flags): Simplify boolean expression. (relaxable_flag, relaxable_operand, assemble_insn): Likewise. (tokenize_extregister): Likewise. * config/tc-csky.c (parse_opcode, get_operand_value): Likewise. (parse_operands_op, parse_operands, md_assemble): Likewise. * config/tc-d10v.c (build_insn): Likewise. * config/tc-score.c (s3_gen_insn_frag): Likewise. * config/tc-score7.c (s7_gen_insn_frag, s7_relax_frag): Likewise. * config/tc-tic6x.c (tic6x_update_features, md_assemble): Likewise. * config/tc-z80.c (emit_byte): Likewise. include/ * opcode/aarch64.h (alias_opcode_p): Simplify boolean expression. (opcode_has_alias, pseudo_opcode_p, optional_operand_p): Likewise. (opcode_has_special_coder): Likewise. ld/ * emultempl/aix.em (gld${EMULATION_NAME}_before_allocation): Simplify boolean expression. * lexsup.c (parse_args): Likewise. * pe-dll.c (pe_dll_id_target): Likewise. opcodes/ * aarch64-opc.c (vector_qualifier_p): Simplify boolean expression. (fp_qualifier_p, get_data_pattern): Likewise. (aarch64_get_operand_modifier_from_value): Likewise. (aarch64_extend_operator_p, aarch64_shift_operator_p): Likewise. (operand_variant_qualifier_p): Likewise. (qualifier_value_in_range_constraint_p): Likewise. (aarch64_get_qualifier_esize): Likewise. (aarch64_get_qualifier_nelem): Likewise. (aarch64_get_qualifier_standard_value): Likewise. (get_lower_bound, get_upper_bound): Likewise. (aarch64_find_best_match, match_operands_qualifier): Likewise. (aarch64_print_operand): Likewise. * aarch64-opc.h (operand_has_inserter, operand_has_extractor): Likewise. (operand_need_sign_extension, operand_need_shift_by_two): Likewise. (operand_need_shift_by_four, operand_maybe_stack_pointer): Likewise. * arm-dis.c (print_insn_mve, print_insn_thumb32): Likewise. * tic6x-dis.c (tic6x_check_fetch_packet_header): Likewise. (print_insn_tic6x): Likewise.
2021-03-29 07:22:56 +08:00
if (vector_qualifier_p (qualifiers[1])
&& qualifiers[2] == AARCH64_OPND_QLF_NIL)
return DP_VECTOR_ACROSS_LANES;
}
return DP_UNKNOWN;
}
/* Select the operand to do the encoding/decoding of the 'size:Q' fields in
the AdvSIMD instructions. */
/* N.B. it is possible to do some optimization that doesn't call
get_data_pattern each time when we need to select an operand. We can
either buffer the caculated the result or statically generate the data,
however, it is not obvious that the optimization will bring significant
benefit. */
int
aarch64_select_operand_for_sizeq_field_coding (const aarch64_opcode *opcode)
{
return
significant_operand_index [get_data_pattern (opcode->qualifiers_list[0])];
}
aarch64: [SME] Add LD1x, ST1x, LDR and STR instructions This patch is adding new loads and stores defined by SME instructions. gas/ChangeLog: * config/tc-aarch64.c (parse_sme_address): New parser. (parse_sme_za_hv_tiles_operand_with_braces): New parser. (parse_sme_za_array): New parser. (output_operand_error_record): Print error details if present. (parse_operands): Support new operands. * testsuite/gas/aarch64/sme-5-illegal.d: New test. * testsuite/gas/aarch64/sme-5-illegal.l: New test. * testsuite/gas/aarch64/sme-5-illegal.s: New test. * testsuite/gas/aarch64/sme-5.d: New test. * testsuite/gas/aarch64/sme-5.s: New test. * testsuite/gas/aarch64/sme-6-illegal.d: New test. * testsuite/gas/aarch64/sme-6-illegal.l: New test. * testsuite/gas/aarch64/sme-6-illegal.s: New test. * testsuite/gas/aarch64/sme-6.d: New test. * testsuite/gas/aarch64/sme-6.s: New test. * testsuite/gas/aarch64/sme-7-illegal.d: New test. * testsuite/gas/aarch64/sme-7-illegal.l: New test. * testsuite/gas/aarch64/sme-7-illegal.s: New test. * testsuite/gas/aarch64/sme-7.d: New test. * testsuite/gas/aarch64/sme-7.s: New test. include/ChangeLog: * opcode/aarch64.h (enum aarch64_opnd): New operands. (enum aarch64_insn_class): Added sme_ldr and sme_str. (AARCH64_OPDE_UNTIED_IMMS): New operand error kind. opcodes/ChangeLog: * aarch64-asm.c (aarch64_ins_sme_za_hv_tiles): New inserter. (aarch64_ins_sme_za_list): New inserter. (aarch64_ins_sme_za_array): New inserter. (aarch64_ins_sme_addr_ri_u4xvl): New inserter. * aarch64-asm.h (AARCH64_DECL_OPD_INSERTER): Added ins_sme_za_list, ins_sme_za_array and ins_sme_addr_ri_u4xvl. * aarch64-dis.c (aarch64_ext_sme_za_hv_tiles): New extractor. (aarch64_ext_sme_za_list): New extractor. (aarch64_ext_sme_za_array): New extractor. (aarch64_ext_sme_addr_ri_u4xvl): New extractor. * aarch64-dis.h (AARCH64_DECL_OPD_EXTRACTOR): Added ext_sme_za_list, ext_sme_za_array and ext_sme_addr_ri_u4xvl. * aarch64-opc.c (operand_general_constraint_met_p): (aarch64_match_operands_constraint): Handle sme_ldr, sme_str and sme_misc. (aarch64_print_operand): New operands supported. * aarch64-tbl.h (OP_SVE_QUU): New qualifier. (OP_SVE_QZU): New qualifier. aarch64-asm-2.c: Regenerate. aarch64-dis-2.c: Regenerate. aarch64-opc-2.c: Regenerate.
2021-11-18 04:02:06 +08:00
/* Instruction bit-fields.
+ Keep synced with 'enum aarch64_field_kind'. */
const aarch64_field fields[] =
{
{ 0, 0 }, /* NIL. */
{ 8, 4 }, /* CRm: in the system instructions. */
{ 10, 2 }, /* CRm_dsb_nxs: 2-bit imm. encoded in CRm<3:2>. */
{ 12, 4 }, /* CRn: in the system instructions. */
{ 10, 8 }, /* CSSC_imm8. */
{ 11, 1 }, /* H: in advsimd scalar x indexed element instructions. */
{ 21, 1 }, /* L: in advsimd scalar x indexed element instructions. */
{ 0, 5 }, /* LSE128_Rt: Shared input+output operand register. */
{ 16, 5 }, /* LSE128_Rt2: Shared input+output operand register 2. */
{ 20, 1 }, /* M: in advsimd scalar x indexed element instructions. */
{ 22, 1 }, /* N: in logical (immediate) instructions. */
{ 30, 1 }, /* Q: in most AdvSIMD instructions. */
{ 10, 5 }, /* Ra: in fp instructions. */
{ 0, 5 }, /* Rd: in many integer instructions. */
{ 16, 5 }, /* Rm: in ld/st reg offset and some integer inst. */
{ 5, 5 }, /* Rn: in many integer instructions. */
{ 16, 5 }, /* Rs: in load/store exclusive instructions. */
{ 0, 5 }, /* Rt: in load/store instructions. */
{ 10, 5 }, /* Rt2: in load/store pair instructions. */
{ 12, 1 }, /* S: in load/store reg offset instructions. */
{ 12, 2 }, /* SM3_imm2: Indexed element SM3 2 bits index immediate. */
{ 1, 3 }, /* SME_Pdx2: predicate register, multiple of 2, [3:1]. */
{ 13, 3 }, /* SME_Pm: second source scalable predicate register P0-P7. */
{ 0, 3 }, /* SME_PNd3: PN0-PN7, bits [2:0]. */
{ 5, 3 }, /* SME_PNn3: PN0-PN7, bits [7:5]. */
{ 16, 1 }, /* SME_Q: Q class bit, bit 16. */
{ 16, 2 }, /* SME_Rm: index base register W12-W15 [17:16]. */
{ 13, 2 }, /* SME_Rv: vector select register W12-W15, bits [14:13]. */
{ 15, 1 }, /* SME_V: (horizontal / vertical tiles), bit 15. */
{ 10, 1 }, /* SME_VL_10: VLx2 or VLx4, bit [10]. */
{ 13, 1 }, /* SME_VL_13: VLx2 or VLx4, bit [13]. */
{ 0, 2 }, /* SME_ZAda_2b: tile ZA0-ZA3. */
{ 0, 3 }, /* SME_ZAda_3b: tile ZA0-ZA7. */
{ 1, 4 }, /* SME_Zdn2: Z0-Z31, multiple of 2, bits [4:1]. */
{ 2, 3 }, /* SME_Zdn4: Z0-Z31, multiple of 4, bits [4:2]. */
{ 16, 4 }, /* SME_Zm: Z0-Z15, bits [19:16]. */
{ 17, 4 }, /* SME_Zm2: Z0-Z31, multiple of 2, bits [20:17]. */
{ 18, 3 }, /* SME_Zm4: Z0-Z31, multiple of 4, bits [20:18]. */
{ 6, 4 }, /* SME_Zn2: Z0-Z31, multiple of 2, bits [9:6]. */
{ 7, 3 }, /* SME_Zn4: Z0-Z31, multiple of 4, bits [9:7]. */
{ 4, 1 }, /* SME_ZtT: upper bit of Zt, bit [4]. */
{ 0, 3 }, /* SME_Zt3: lower 3 bits of Zt, bits [2:0]. */
{ 0, 2 }, /* SME_Zt2: lower 2 bits of Zt, bits [1:0]. */
{ 23, 1 }, /* SME_i1: immediate field, bit 23. */
{ 12, 2 }, /* SME_size_12: bits [13:12]. */
{ 22, 2 }, /* SME_size_22: size<1>, size<0> class field, [23:22]. */
{ 23, 1 }, /* SME_sz_23: bit [23]. */
{ 22, 1 }, /* SME_tszh: immediate and qualifier field, bit 22. */
{ 18, 3 }, /* SME_tszl: immediate and qualifier field, bits [20:18]. */
{ 0, 8 }, /* SME_zero_mask: list of up to 8 tile names separated by commas [7:0]. */
{ 4, 1 }, /* SVE_M_4: Merge/zero select, bit 4. */
{ 14, 1 }, /* SVE_M_14: Merge/zero select, bit 14. */
{ 16, 1 }, /* SVE_M_16: Merge/zero select, bit 16. */
[AArch64][SVE 27/32] Add SVE integer immediate operands This patch adds the new SVE integer immediate operands. There are three kinds: - simple signed and unsigned ranges, but with new widths and positions. - 13-bit logical immediates. These have the same form as in base AArch64, but at a different bit position. In the case of the "MOV Zn.<T>, #<limm>" alias of DUPM, the logical immediate <limm> is not allowed to be a valid DUP immediate, since DUP is preferred over DUPM for constants that both instructions can handle. - a new 9-bit arithmetic immediate, of the form "<imm8>{, LSL #8}". In some contexts the operand is signed and in others it's unsigned. As an extension, we allow shifted immediates to be written as a single integer, e.g. "#256" is equivalent to "#1, LSL #8". We also use the shiftless form as the preferred disassembly, except for the special case of "#0, LSL #8" (a redundant encoding of 0). include/ * opcode/aarch64.h (AARCH64_OPND_SIMM5): New aarch64_opnd. (AARCH64_OPND_SVE_AIMM, AARCH64_OPND_SVE_ASIMM) (AARCH64_OPND_SVE_INV_LIMM, AARCH64_OPND_SVE_LIMM) (AARCH64_OPND_SVE_LIMM_MOV, AARCH64_OPND_SVE_SHLIMM_PRED) (AARCH64_OPND_SVE_SHLIMM_UNPRED, AARCH64_OPND_SVE_SHRIMM_PRED) (AARCH64_OPND_SVE_SHRIMM_UNPRED, AARCH64_OPND_SVE_SIMM5) (AARCH64_OPND_SVE_SIMM5B, AARCH64_OPND_SVE_SIMM6) (AARCH64_OPND_SVE_SIMM8, AARCH64_OPND_SVE_UIMM3) (AARCH64_OPND_SVE_UIMM7, AARCH64_OPND_SVE_UIMM8) (AARCH64_OPND_SVE_UIMM8_53): Likewise. (aarch64_sve_dupm_mov_immediate_p): Declare. opcodes/ * aarch64-tbl.h (AARCH64_OPERANDS): Add entries for the new SVE integer immediate operands. * aarch64-opc.h (FLD_SVE_immN, FLD_SVE_imm3, FLD_SVE_imm5) (FLD_SVE_imm5b, FLD_SVE_imm7, FLD_SVE_imm8, FLD_SVE_imm9) (FLD_SVE_immr, FLD_SVE_imms, FLD_SVE_tszh): New aarch64_field_kinds. * aarch64-opc.c (fields): Add corresponding entries. (operand_general_constraint_met_p): Handle the new SVE integer immediate operands. (aarch64_print_operand): Likewise. (aarch64_sve_dupm_mov_immediate_p): New function. * aarch64-opc-2.c: Regenerate. * aarch64-asm.h (ins_inv_limm, ins_sve_aimm, ins_sve_asimm) (ins_sve_limm_mov, ins_sve_shlimm, ins_sve_shrimm): New inserters. * aarch64-asm.c (aarch64_ins_limm_1): New function, split out from... (aarch64_ins_limm): ...here. (aarch64_ins_inv_limm): New function. (aarch64_ins_sve_aimm): Likewise. (aarch64_ins_sve_asimm): Likewise. (aarch64_ins_sve_limm_mov): Likewise. (aarch64_ins_sve_shlimm): Likewise. (aarch64_ins_sve_shrimm): Likewise. * aarch64-asm-2.c: Regenerate. * aarch64-dis.h (ext_inv_limm, ext_sve_aimm, ext_sve_asimm) (ext_sve_limm_mov, ext_sve_shlimm, ext_sve_shrimm): New extractors. * aarch64-dis.c (decode_limm): New function, split out from... (aarch64_ext_limm): ...here. (aarch64_ext_inv_limm): New function. (decode_sve_aimm): Likewise. (aarch64_ext_sve_aimm): Likewise. (aarch64_ext_sve_asimm): Likewise. (aarch64_ext_sve_limm_mov): Likewise. (aarch64_top_bit): Likewise. (aarch64_ext_sve_shlimm): Likewise. (aarch64_ext_sve_shrimm): Likewise. * aarch64-dis-2.c: Regenerate. gas/ * config/tc-aarch64.c (parse_operands): Handle the new SVE integer immediate operands.
2016-09-21 23:56:57 +08:00
{ 17, 1 }, /* SVE_N: SVE equivalent of N. */
[AArch64][SVE 21/32] Add Zn and Pn registers This patch adds the Zn and Pn registers, and associated fields and operands. include/ * opcode/aarch64.h (AARCH64_OPND_CLASS_SVE_REG): New aarch64_operand_class. (AARCH64_OPND_CLASS_PRED_REG): Likewise. (AARCH64_OPND_SVE_Pd, AARCH64_OPND_SVE_Pg3, AARCH64_OPND_SVE_Pg4_5) (AARCH64_OPND_SVE_Pg4_10, AARCH64_OPND_SVE_Pg4_16) (AARCH64_OPND_SVE_Pm, AARCH64_OPND_SVE_Pn, AARCH64_OPND_SVE_Pt) (AARCH64_OPND_SVE_Za_5, AARCH64_OPND_SVE_Za_16, AARCH64_OPND_SVE_Zd) (AARCH64_OPND_SVE_Zm_5, AARCH64_OPND_SVE_Zm_16, AARCH64_OPND_SVE_Zn) (AARCH64_OPND_SVE_Zn_INDEX, AARCH64_OPND_SVE_ZnxN) (AARCH64_OPND_SVE_Zt, AARCH64_OPND_SVE_ZtxN): New aarch64_opnds. opcodes/ * aarch64-tbl.h (AARCH64_OPERANDS): Add entries for new SVE operands. * aarch64-opc.h (FLD_SVE_Pd, FLD_SVE_Pg3, FLD_SVE_Pg4_5) (FLD_SVE_Pg4_10, FLD_SVE_Pg4_16, FLD_SVE_Pm, FLD_SVE_Pn, FLD_SVE_Pt) (FLD_SVE_Za_5, FLD_SVE_Za_16, FLD_SVE_Zd, FLD_SVE_Zm_5, FLD_SVE_Zm_16) (FLD_SVE_Zn, FLD_SVE_Zt, FLD_SVE_tzsh): New aarch64_field_kinds. * aarch64-opc.c (fields): Add corresponding entries here. (operand_general_constraint_met_p): Check that SVE register lists have the correct length. Check the ranges of SVE index registers. Check for cases where p8-p15 are used in 3-bit predicate fields. (aarch64_print_operand): Handle the new SVE operands. * aarch64-opc-2.c: Regenerate. * aarch64-asm.h (ins_sve_index, ins_sve_reglist): New inserters. * aarch64-asm.c (aarch64_ins_sve_index): New function. (aarch64_ins_sve_reglist): Likewise. * aarch64-asm-2.c: Regenerate. * aarch64-dis.h (ext_sve_index, ext_sve_reglist): New extractors. * aarch64-dis.c (aarch64_ext_sve_index): New function. (aarch64_ext_sve_reglist): Likewise. * aarch64-dis-2.c: Regenerate. gas/ * config/tc-aarch64.c (NTA_HASVARWIDTH): New macro. (AARCH64_REG_TYPES): Add ZN and PN. (get_reg_expected_msg): Handle them. (parse_vector_type_for_operand): Add a reg_type parameter. Skip the width for Zn and Pn registers. (parse_typed_reg): Extend vector handling to Zn and Pn. Update the call to parse_vector_type_for_operand. Set HASVARTYPE for Zn and Pn, expecting the width to be 0. (parse_vector_reg_list): Restrict error about [BHSD]nn operands to REG_TYPE_VN. (vectype_to_qualifier): Use S_[BHSD] qualifiers for NTA_HASVARWIDTH. (parse_operands): Handle the new Zn and Pn operands. (REGSET16): New macro, split out from... (REGSET31): ...here. (reg_names): Add Zn and Pn entries.
2016-09-21 23:53:54 +08:00
{ 0, 4 }, /* SVE_Pd: p0-p15, bits [3,0]. */
{ 10, 3 }, /* SVE_Pg3: p0-p7, bits [12,10]. */
{ 5, 4 }, /* SVE_Pg4_5: p0-p15, bits [8,5]. */
{ 10, 4 }, /* SVE_Pg4_10: p0-p15, bits [13,10]. */
{ 16, 4 }, /* SVE_Pg4_16: p0-p15, bits [19,16]. */
{ 16, 4 }, /* SVE_Pm: p0-p15, bits [19,16]. */
{ 5, 4 }, /* SVE_Pn: p0-p15, bits [8,5]. */
{ 0, 4 }, /* SVE_Pt: p0-p15, bits [3,0]. */
{ 5, 5 }, /* SVE_Rm: SVE alternative position for Rm. */
{ 16, 5 }, /* SVE_Rn: SVE alternative position for Rn. */
{ 0, 5 }, /* SVE_Vd: Scalar SIMD&FP register, bits [4,0]. */
{ 5, 5 }, /* SVE_Vm: Scalar SIMD&FP register, bits [9,5]. */
{ 5, 5 }, /* SVE_Vn: Scalar SIMD&FP register, bits [9,5]. */
[AArch64][SVE 21/32] Add Zn and Pn registers This patch adds the Zn and Pn registers, and associated fields and operands. include/ * opcode/aarch64.h (AARCH64_OPND_CLASS_SVE_REG): New aarch64_operand_class. (AARCH64_OPND_CLASS_PRED_REG): Likewise. (AARCH64_OPND_SVE_Pd, AARCH64_OPND_SVE_Pg3, AARCH64_OPND_SVE_Pg4_5) (AARCH64_OPND_SVE_Pg4_10, AARCH64_OPND_SVE_Pg4_16) (AARCH64_OPND_SVE_Pm, AARCH64_OPND_SVE_Pn, AARCH64_OPND_SVE_Pt) (AARCH64_OPND_SVE_Za_5, AARCH64_OPND_SVE_Za_16, AARCH64_OPND_SVE_Zd) (AARCH64_OPND_SVE_Zm_5, AARCH64_OPND_SVE_Zm_16, AARCH64_OPND_SVE_Zn) (AARCH64_OPND_SVE_Zn_INDEX, AARCH64_OPND_SVE_ZnxN) (AARCH64_OPND_SVE_Zt, AARCH64_OPND_SVE_ZtxN): New aarch64_opnds. opcodes/ * aarch64-tbl.h (AARCH64_OPERANDS): Add entries for new SVE operands. * aarch64-opc.h (FLD_SVE_Pd, FLD_SVE_Pg3, FLD_SVE_Pg4_5) (FLD_SVE_Pg4_10, FLD_SVE_Pg4_16, FLD_SVE_Pm, FLD_SVE_Pn, FLD_SVE_Pt) (FLD_SVE_Za_5, FLD_SVE_Za_16, FLD_SVE_Zd, FLD_SVE_Zm_5, FLD_SVE_Zm_16) (FLD_SVE_Zn, FLD_SVE_Zt, FLD_SVE_tzsh): New aarch64_field_kinds. * aarch64-opc.c (fields): Add corresponding entries here. (operand_general_constraint_met_p): Check that SVE register lists have the correct length. Check the ranges of SVE index registers. Check for cases where p8-p15 are used in 3-bit predicate fields. (aarch64_print_operand): Handle the new SVE operands. * aarch64-opc-2.c: Regenerate. * aarch64-asm.h (ins_sve_index, ins_sve_reglist): New inserters. * aarch64-asm.c (aarch64_ins_sve_index): New function. (aarch64_ins_sve_reglist): Likewise. * aarch64-asm-2.c: Regenerate. * aarch64-dis.h (ext_sve_index, ext_sve_reglist): New extractors. * aarch64-dis.c (aarch64_ext_sve_index): New function. (aarch64_ext_sve_reglist): Likewise. * aarch64-dis-2.c: Regenerate. gas/ * config/tc-aarch64.c (NTA_HASVARWIDTH): New macro. (AARCH64_REG_TYPES): Add ZN and PN. (get_reg_expected_msg): Handle them. (parse_vector_type_for_operand): Add a reg_type parameter. Skip the width for Zn and Pn registers. (parse_typed_reg): Extend vector handling to Zn and Pn. Update the call to parse_vector_type_for_operand. Set HASVARTYPE for Zn and Pn, expecting the width to be 0. (parse_vector_reg_list): Restrict error about [BHSD]nn operands to REG_TYPE_VN. (vectype_to_qualifier): Use S_[BHSD] qualifiers for NTA_HASVARWIDTH. (parse_operands): Handle the new Zn and Pn operands. (REGSET16): New macro, split out from... (REGSET31): ...here. (reg_names): Add Zn and Pn entries.
2016-09-21 23:53:54 +08:00
{ 5, 5 }, /* SVE_Za_5: SVE vector register, bits [9,5]. */
{ 16, 5 }, /* SVE_Za_16: SVE vector register, bits [20,16]. */
{ 0, 5 }, /* SVE_Zd: SVE vector register. bits [4,0]. */
{ 5, 5 }, /* SVE_Zm_5: SVE vector register, bits [9,5]. */
{ 16, 5 }, /* SVE_Zm_16: SVE vector register, bits [20,16]. */
{ 5, 5 }, /* SVE_Zn: SVE vector register, bits [9,5]. */
{ 0, 5 }, /* SVE_Zt: SVE vector register, bits [4,0]. */
[AArch64][SVE 28/32] Add SVE FP immediate operands This patch adds support for the new SVE floating-point immediate operands. One operand uses the same 8-bit encoding as base AArch64, but in a different position. The others use a single bit to select between two values. One of the single-bit operands is a choice between 0 and 1, where 0 is not a valid 8-bit encoding. I think the cleanest way of handling these single-bit immediates is therefore to use the IEEE float encoding itself as the immediate value and select between the two possible values when encoding and decoding. As described in the covering note for the patch that added F_STRICT, we get better error messages by accepting unsuffixed vector registers and leaving the qualifier matching code to report an error. This means that we carry on parsing the other operands, and so can try to parse FP immediates for invalid instructions like: fcpy z0, #2.5 In this case there is no suffix to tell us whether the immediate should be treated as single or double precision. Again, we get better error messages by picking one (arbitrary) immediate size and reporting an error for the missing suffix later. include/ * opcode/aarch64.h (AARCH64_OPND_SVE_FPIMM8): New aarch64_opnd. (AARCH64_OPND_SVE_I1_HALF_ONE, AARCH64_OPND_SVE_I1_HALF_TWO) (AARCH64_OPND_SVE_I1_ZERO_ONE): Likewise. opcodes/ * aarch64-tbl.h (AARCH64_OPERANDS): Add entries for the new SVE FP immediate operands. * aarch64-opc.h (FLD_SVE_i1): New aarch64_field_kind. * aarch64-opc.c (fields): Add corresponding entry. (operand_general_constraint_met_p): Handle the new SVE FP immediate operands. (aarch64_print_operand): Likewise. * aarch64-opc-2.c: Regenerate. * aarch64-asm.h (ins_sve_float_half_one, ins_sve_float_half_two) (ins_sve_float_zero_one): New inserters. * aarch64-asm.c (aarch64_ins_sve_float_half_one): New function. (aarch64_ins_sve_float_half_two): Likewise. (aarch64_ins_sve_float_zero_one): Likewise. * aarch64-asm-2.c: Regenerate. * aarch64-dis.h (ext_sve_float_half_one, ext_sve_float_half_two) (ext_sve_float_zero_one): New extractors. * aarch64-dis.c (aarch64_ext_sve_float_half_one): New function. (aarch64_ext_sve_float_half_two): Likewise. (aarch64_ext_sve_float_zero_one): Likewise. * aarch64-dis-2.c: Regenerate. gas/ * config/tc-aarch64.c (double_precision_operand_p): New function. (parse_operands): Use it to calculate the dp_p input to parse_aarch64_imm_float. Handle the new SVE FP immediate operands.
2016-09-21 23:57:22 +08:00
{ 5, 1 }, /* SVE_i1: single-bit immediate. */
{ 20, 1 }, /* SVE_i2h: high bit of 2bit immediate, bits. */
[AArch64] Additional SVE instructions This patch supports some additions to the SVE architecture prior to its public release. include/ * opcode/aarch64.h (AARCH64_OPND_SVE_ADDR_RI_S4x16) (AARCH64_OPND_SVE_IMM_ROT1, AARCH64_OPND_SVE_IMM_ROT2) (AARCH64_OPND_SVE_Zm3_INDEX, AARCH64_OPND_SVE_Zm3_22_INDEX) (AARCH64_OPND_SVE_Zm4_INDEX): New aarch64_opnds. opcodes/ * aarch64-tbl.h (OP_SVE_HMH, OP_SVE_VMU_HSD, OP_SVE_VMVU_HSD) (OP_SVE_VMVV_HSD, OP_SVE_VMVVU_HSD, OP_SVE_VM_HSD, OP_SVE_VUVV_HSD) (OP_SVE_VUV_HSD, OP_SVE_VU_HSD, OP_SVE_VVVU_H, OP_SVE_VVVU_S) (OP_SVE_VVVU_HSD, OP_SVE_VVV_D, OP_SVE_VVV_D_H, OP_SVE_VVV_H) (OP_SVE_VVV_HSD, OP_SVE_VVV_S, OP_SVE_VVV_S_B, OP_SVE_VVV_SD_BH) (OP_SVE_VV_BHSDQ, OP_SVE_VV_HSD, OP_SVE_VZVV_HSD, OP_SVE_VZV_HSD) (OP_SVE_V_HSD): New macros. (OP_SVE_VMU_SD, OP_SVE_VMVU_SD, OP_SVE_VM_SD, OP_SVE_VUVV_SD) (OP_SVE_VU_SD, OP_SVE_VVVU_SD, OP_SVE_VVV_SD, OP_SVE_VZVV_SD) (OP_SVE_VZV_SD, OP_SVE_V_SD): Delete. (aarch64_opcode_table): Add new SVE instructions. (aarch64_opcode_table): Use imm_rotate{1,2} instead of imm_rotate for rotation operands. Add new SVE operands. * aarch64-asm.h (ins_sve_addr_ri_s4): New inserter. (ins_sve_quad_index): Likewise. (ins_imm_rotate): Split into... (ins_imm_rotate1, ins_imm_rotate2): ...these two inserters. * aarch64-asm.c (aarch64_ins_imm_rotate): Split into... (aarch64_ins_imm_rotate1, aarch64_ins_imm_rotate2): ...these two functions. (aarch64_ins_sve_addr_ri_s4): New function. (aarch64_ins_sve_quad_index): Likewise. (do_misc_encoding): Handle "MOV Zn.Q, Qm". * aarch64-asm-2.c: Regenerate. * aarch64-dis.h (ext_sve_addr_ri_s4): New extractor. (ext_sve_quad_index): Likewise. (ext_imm_rotate): Split into... (ext_imm_rotate1, ext_imm_rotate2): ...these two extractors. * aarch64-dis.c (aarch64_ext_imm_rotate): Split into... (aarch64_ext_imm_rotate1, aarch64_ext_imm_rotate2): ...these two functions. (aarch64_ext_sve_addr_ri_s4): New function. (aarch64_ext_sve_quad_index): Likewise. (aarch64_ext_sve_index): Allow quad indices. (do_misc_decoding): Likewise. * aarch64-dis-2.c: Regenerate. * aarch64-opc.h (FLD_SVE_i3h, FLD_SVE_rot1, FLD_SVE_rot2): New aarch64_field_kinds. (OPD_F_OD_MASK): Widen by one bit. (OPD_F_NO_ZR): Bump accordingly. (get_operand_field_width): New function. * aarch64-opc.c (fields): Add new SVE fields. (operand_general_constraint_met_p): Handle new SVE operands. (aarch64_print_operand): Likewise. * aarch64-opc-2.c: Regenerate. gas/ * doc/c-aarch64.texi: Document that sve implies fp16, simd and compnum. * config/tc-aarch64.c (parse_vector_type_for_operand): Allow .q to be used with SVE registers. (parse_operands): Handle new SVE operands. (aarch64_features): Make "sve" require F16 rather than FP. Also require COMPNUM. * testsuite/gas/aarch64/sve.s: Add tests for new instructions. Include compnum tests. * testsuite/gas/aarch64/sve.d: Update accordingly. * testsuite/gas/aarch64/sve-invalid.s: Add tests for new instructions. * testsuite/gas/aarch64/sve-invalid.l: Update accordingly. Also update expected output for new FMOV and MOV alternatives.
2017-02-25 02:29:00 +08:00
{ 22, 1 }, /* SVE_i3h: high bit of 3-bit immediate. */
{ 19, 2 }, /* SVE_i3h2: two high bits of 3bit immediate, bits [20,19]. */
{ 11, 1 }, /* SVE_i3l: low bit of 3-bit immediate. */
[AArch64][SVE 27/32] Add SVE integer immediate operands This patch adds the new SVE integer immediate operands. There are three kinds: - simple signed and unsigned ranges, but with new widths and positions. - 13-bit logical immediates. These have the same form as in base AArch64, but at a different bit position. In the case of the "MOV Zn.<T>, #<limm>" alias of DUPM, the logical immediate <limm> is not allowed to be a valid DUP immediate, since DUP is preferred over DUPM for constants that both instructions can handle. - a new 9-bit arithmetic immediate, of the form "<imm8>{, LSL #8}". In some contexts the operand is signed and in others it's unsigned. As an extension, we allow shifted immediates to be written as a single integer, e.g. "#256" is equivalent to "#1, LSL #8". We also use the shiftless form as the preferred disassembly, except for the special case of "#0, LSL #8" (a redundant encoding of 0). include/ * opcode/aarch64.h (AARCH64_OPND_SIMM5): New aarch64_opnd. (AARCH64_OPND_SVE_AIMM, AARCH64_OPND_SVE_ASIMM) (AARCH64_OPND_SVE_INV_LIMM, AARCH64_OPND_SVE_LIMM) (AARCH64_OPND_SVE_LIMM_MOV, AARCH64_OPND_SVE_SHLIMM_PRED) (AARCH64_OPND_SVE_SHLIMM_UNPRED, AARCH64_OPND_SVE_SHRIMM_PRED) (AARCH64_OPND_SVE_SHRIMM_UNPRED, AARCH64_OPND_SVE_SIMM5) (AARCH64_OPND_SVE_SIMM5B, AARCH64_OPND_SVE_SIMM6) (AARCH64_OPND_SVE_SIMM8, AARCH64_OPND_SVE_UIMM3) (AARCH64_OPND_SVE_UIMM7, AARCH64_OPND_SVE_UIMM8) (AARCH64_OPND_SVE_UIMM8_53): Likewise. (aarch64_sve_dupm_mov_immediate_p): Declare. opcodes/ * aarch64-tbl.h (AARCH64_OPERANDS): Add entries for the new SVE integer immediate operands. * aarch64-opc.h (FLD_SVE_immN, FLD_SVE_imm3, FLD_SVE_imm5) (FLD_SVE_imm5b, FLD_SVE_imm7, FLD_SVE_imm8, FLD_SVE_imm9) (FLD_SVE_immr, FLD_SVE_imms, FLD_SVE_tszh): New aarch64_field_kinds. * aarch64-opc.c (fields): Add corresponding entries. (operand_general_constraint_met_p): Handle the new SVE integer immediate operands. (aarch64_print_operand): Likewise. (aarch64_sve_dupm_mov_immediate_p): New function. * aarch64-opc-2.c: Regenerate. * aarch64-asm.h (ins_inv_limm, ins_sve_aimm, ins_sve_asimm) (ins_sve_limm_mov, ins_sve_shlimm, ins_sve_shrimm): New inserters. * aarch64-asm.c (aarch64_ins_limm_1): New function, split out from... (aarch64_ins_limm): ...here. (aarch64_ins_inv_limm): New function. (aarch64_ins_sve_aimm): Likewise. (aarch64_ins_sve_asimm): Likewise. (aarch64_ins_sve_limm_mov): Likewise. (aarch64_ins_sve_shlimm): Likewise. (aarch64_ins_sve_shrimm): Likewise. * aarch64-asm-2.c: Regenerate. * aarch64-dis.h (ext_inv_limm, ext_sve_aimm, ext_sve_asimm) (ext_sve_limm_mov, ext_sve_shlimm, ext_sve_shrimm): New extractors. * aarch64-dis.c (decode_limm): New function, split out from... (aarch64_ext_limm): ...here. (aarch64_ext_inv_limm): New function. (decode_sve_aimm): Likewise. (aarch64_ext_sve_aimm): Likewise. (aarch64_ext_sve_asimm): Likewise. (aarch64_ext_sve_limm_mov): Likewise. (aarch64_top_bit): Likewise. (aarch64_ext_sve_shlimm): Likewise. (aarch64_ext_sve_shrimm): Likewise. * aarch64-dis-2.c: Regenerate. gas/ * config/tc-aarch64.c (parse_operands): Handle the new SVE integer immediate operands.
2016-09-21 23:56:57 +08:00
{ 16, 3 }, /* SVE_imm3: 3-bit immediate field. */
[AArch64][SVE 24/32] Add AARCH64_OPND_SVE_PATTERN_SCALED Some SVE instructions count the number of elements in a given vector pattern and allow a scale factor of [1, 16] to be applied to the result. This scale factor is written ", MUL #n", where "MUL" is a new operator. E.g.: UQINCD X0, POW2, MUL #2 This patch adds support for this kind of operand. All existing operators were shifts of some kind, so there was a natural range of [0, 63] regardless of context. This was then narrowered further by later checks (e.g. to [0, 31] when used for 32-bit values). In contrast, MUL doesn't really have a natural context-independent range. Rather than pick one arbitrarily, it seemed better to make the "shift" amount a full 64-bit value and leave the range test to the usual operand-checking code. I've rearranged the fields of aarch64_opnd_info so that this doesn't increase the size of the structure (although I don't think its size is critical anyway). include/ * opcode/aarch64.h (AARCH64_OPND_SVE_PATTERN_SCALED): New aarch64_opnd. (AARCH64_MOD_MUL): New aarch64_modifier_kind. (aarch64_opnd_info): Make shifter.amount an int64_t and rearrange the fields. opcodes/ * aarch64-tbl.h (AARCH64_OPERANDS): Add an entry for AARCH64_OPND_SVE_PATTERN_SCALED. * aarch64-opc.h (FLD_SVE_imm4): New aarch64_field_kind. * aarch64-opc.c (fields): Add a corresponding entry. (set_multiplier_out_of_range_error): New function. (aarch64_operand_modifiers): Add entry for AARCH64_MOD_MUL. (operand_general_constraint_met_p): Handle AARCH64_OPND_SVE_PATTERN_SCALED. (print_register_offset_address): Use PRIi64 to print the shift amount. (aarch64_print_operand): Likewise. Handle AARCH64_OPND_SVE_PATTERN_SCALED. * aarch64-opc-2.c: Regenerate. * aarch64-asm.h (ins_sve_scale): New inserter. * aarch64-asm.c (aarch64_ins_sve_scale): New function. * aarch64-asm-2.c: Regenerate. * aarch64-dis.h (ext_sve_scale): New inserter. * aarch64-dis.c (aarch64_ext_sve_scale): New function. * aarch64-dis-2.c: Regenerate. gas/ * config/tc-aarch64.c (SHIFTED_MUL): New parse_shift_mode. (parse_shift): Handle it. Reject AARCH64_MOD_MUL for all other shift modes. Skip range tests for AARCH64_MOD_MUL. (process_omitted_operand): Handle AARCH64_OPND_SVE_PATTERN_SCALED. (parse_operands): Likewise.
2016-09-21 23:55:22 +08:00
{ 16, 4 }, /* SVE_imm4: 4-bit immediate field. */
[AArch64][SVE 27/32] Add SVE integer immediate operands This patch adds the new SVE integer immediate operands. There are three kinds: - simple signed and unsigned ranges, but with new widths and positions. - 13-bit logical immediates. These have the same form as in base AArch64, but at a different bit position. In the case of the "MOV Zn.<T>, #<limm>" alias of DUPM, the logical immediate <limm> is not allowed to be a valid DUP immediate, since DUP is preferred over DUPM for constants that both instructions can handle. - a new 9-bit arithmetic immediate, of the form "<imm8>{, LSL #8}". In some contexts the operand is signed and in others it's unsigned. As an extension, we allow shifted immediates to be written as a single integer, e.g. "#256" is equivalent to "#1, LSL #8". We also use the shiftless form as the preferred disassembly, except for the special case of "#0, LSL #8" (a redundant encoding of 0). include/ * opcode/aarch64.h (AARCH64_OPND_SIMM5): New aarch64_opnd. (AARCH64_OPND_SVE_AIMM, AARCH64_OPND_SVE_ASIMM) (AARCH64_OPND_SVE_INV_LIMM, AARCH64_OPND_SVE_LIMM) (AARCH64_OPND_SVE_LIMM_MOV, AARCH64_OPND_SVE_SHLIMM_PRED) (AARCH64_OPND_SVE_SHLIMM_UNPRED, AARCH64_OPND_SVE_SHRIMM_PRED) (AARCH64_OPND_SVE_SHRIMM_UNPRED, AARCH64_OPND_SVE_SIMM5) (AARCH64_OPND_SVE_SIMM5B, AARCH64_OPND_SVE_SIMM6) (AARCH64_OPND_SVE_SIMM8, AARCH64_OPND_SVE_UIMM3) (AARCH64_OPND_SVE_UIMM7, AARCH64_OPND_SVE_UIMM8) (AARCH64_OPND_SVE_UIMM8_53): Likewise. (aarch64_sve_dupm_mov_immediate_p): Declare. opcodes/ * aarch64-tbl.h (AARCH64_OPERANDS): Add entries for the new SVE integer immediate operands. * aarch64-opc.h (FLD_SVE_immN, FLD_SVE_imm3, FLD_SVE_imm5) (FLD_SVE_imm5b, FLD_SVE_imm7, FLD_SVE_imm8, FLD_SVE_imm9) (FLD_SVE_immr, FLD_SVE_imms, FLD_SVE_tszh): New aarch64_field_kinds. * aarch64-opc.c (fields): Add corresponding entries. (operand_general_constraint_met_p): Handle the new SVE integer immediate operands. (aarch64_print_operand): Likewise. (aarch64_sve_dupm_mov_immediate_p): New function. * aarch64-opc-2.c: Regenerate. * aarch64-asm.h (ins_inv_limm, ins_sve_aimm, ins_sve_asimm) (ins_sve_limm_mov, ins_sve_shlimm, ins_sve_shrimm): New inserters. * aarch64-asm.c (aarch64_ins_limm_1): New function, split out from... (aarch64_ins_limm): ...here. (aarch64_ins_inv_limm): New function. (aarch64_ins_sve_aimm): Likewise. (aarch64_ins_sve_asimm): Likewise. (aarch64_ins_sve_limm_mov): Likewise. (aarch64_ins_sve_shlimm): Likewise. (aarch64_ins_sve_shrimm): Likewise. * aarch64-asm-2.c: Regenerate. * aarch64-dis.h (ext_inv_limm, ext_sve_aimm, ext_sve_asimm) (ext_sve_limm_mov, ext_sve_shlimm, ext_sve_shrimm): New extractors. * aarch64-dis.c (decode_limm): New function, split out from... (aarch64_ext_limm): ...here. (aarch64_ext_inv_limm): New function. (decode_sve_aimm): Likewise. (aarch64_ext_sve_aimm): Likewise. (aarch64_ext_sve_asimm): Likewise. (aarch64_ext_sve_limm_mov): Likewise. (aarch64_top_bit): Likewise. (aarch64_ext_sve_shlimm): Likewise. (aarch64_ext_sve_shrimm): Likewise. * aarch64-dis-2.c: Regenerate. gas/ * config/tc-aarch64.c (parse_operands): Handle the new SVE integer immediate operands.
2016-09-21 23:56:57 +08:00
{ 5, 5 }, /* SVE_imm5: 5-bit immediate field. */
{ 16, 5 }, /* SVE_imm5b: secondary 5-bit immediate field. */
[AArch64][SVE 25/32] Add support for SVE addressing modes This patch adds most of the new SVE addressing modes and associated operands. A follow-on patch adds MUL VL, since handling it separately makes the changes easier to read. The patch also introduces a new "operand-dependent data" field to the operand flags, based closely on the existing one for opcode flags. For SVE this new field needs only 2 bits, but it could be widened in future if necessary. include/ * opcode/aarch64.h (AARCH64_OPND_SVE_ADDR_RI_U6): New aarch64_opnd. (AARCH64_OPND_SVE_ADDR_RI_U6x2, AARCH64_OPND_SVE_ADDR_RI_U6x4) (AARCH64_OPND_SVE_ADDR_RI_U6x8, AARCH64_OPND_SVE_ADDR_RR) (AARCH64_OPND_SVE_ADDR_RR_LSL1, AARCH64_OPND_SVE_ADDR_RR_LSL2) (AARCH64_OPND_SVE_ADDR_RR_LSL3, AARCH64_OPND_SVE_ADDR_RX) (AARCH64_OPND_SVE_ADDR_RX_LSL1, AARCH64_OPND_SVE_ADDR_RX_LSL2) (AARCH64_OPND_SVE_ADDR_RX_LSL3, AARCH64_OPND_SVE_ADDR_RZ) (AARCH64_OPND_SVE_ADDR_RZ_LSL1, AARCH64_OPND_SVE_ADDR_RZ_LSL2) (AARCH64_OPND_SVE_ADDR_RZ_LSL3, AARCH64_OPND_SVE_ADDR_RZ_XTW_14) (AARCH64_OPND_SVE_ADDR_RZ_XTW_22, AARCH64_OPND_SVE_ADDR_RZ_XTW1_14) (AARCH64_OPND_SVE_ADDR_RZ_XTW1_22, AARCH64_OPND_SVE_ADDR_RZ_XTW2_14) (AARCH64_OPND_SVE_ADDR_RZ_XTW2_22, AARCH64_OPND_SVE_ADDR_RZ_XTW3_14) (AARCH64_OPND_SVE_ADDR_RZ_XTW3_22, AARCH64_OPND_SVE_ADDR_ZI_U5) (AARCH64_OPND_SVE_ADDR_ZI_U5x2, AARCH64_OPND_SVE_ADDR_ZI_U5x4) (AARCH64_OPND_SVE_ADDR_ZI_U5x8, AARCH64_OPND_SVE_ADDR_ZZ_LSL) (AARCH64_OPND_SVE_ADDR_ZZ_SXTW, AARCH64_OPND_SVE_ADDR_ZZ_UXTW): Likewise. opcodes/ * aarch64-tbl.h (AARCH64_OPERANDS): Add entries for the new SVE address operands. * aarch64-opc.h (FLD_SVE_imm6, FLD_SVE_msz, FLD_SVE_xs_14) (FLD_SVE_xs_22): New aarch64_field_kinds. (OPD_F_OD_MASK, OPD_F_OD_LSB, OPD_F_NO_ZR): New flags. (get_operand_specific_data): New function. * aarch64-opc.c (fields): Add entries for FLD_SVE_imm6, FLD_SVE_msz, FLD_SVE_xs_14 and FLD_SVE_xs_22. (operand_general_constraint_met_p): Handle the new SVE address operands. (sve_reg): New array. (get_addr_sve_reg_name): New function. (aarch64_print_operand): Handle the new SVE address operands. * aarch64-opc-2.c: Regenerate. * aarch64-asm.h (ins_sve_addr_ri_u6, ins_sve_addr_rr_lsl) (ins_sve_addr_rz_xtw, ins_sve_addr_zi_u5, ins_sve_addr_zz_lsl) (ins_sve_addr_zz_sxtw, ins_sve_addr_zz_uxtw): New inserters. * aarch64-asm.c (aarch64_ins_sve_addr_ri_u6): New function. (aarch64_ins_sve_addr_rr_lsl): Likewise. (aarch64_ins_sve_addr_rz_xtw): Likewise. (aarch64_ins_sve_addr_zi_u5): Likewise. (aarch64_ins_sve_addr_zz): Likewise. (aarch64_ins_sve_addr_zz_lsl): Likewise. (aarch64_ins_sve_addr_zz_sxtw): Likewise. (aarch64_ins_sve_addr_zz_uxtw): Likewise. * aarch64-asm-2.c: Regenerate. * aarch64-dis.h (ext_sve_addr_ri_u6, ext_sve_addr_rr_lsl) (ext_sve_addr_rz_xtw, ext_sve_addr_zi_u5, ext_sve_addr_zz_lsl) (ext_sve_addr_zz_sxtw, ext_sve_addr_zz_uxtw): New extractors. * aarch64-dis.c (aarch64_ext_sve_add_reg_imm): New function. (aarch64_ext_sve_addr_ri_u6): Likewise. (aarch64_ext_sve_addr_rr_lsl): Likewise. (aarch64_ext_sve_addr_rz_xtw): Likewise. (aarch64_ext_sve_addr_zi_u5): Likewise. (aarch64_ext_sve_addr_zz): Likewise. (aarch64_ext_sve_addr_zz_lsl): Likewise. (aarch64_ext_sve_addr_zz_sxtw): Likewise. (aarch64_ext_sve_addr_zz_uxtw): Likewise. * aarch64-dis-2.c: Regenerate. gas/ * config/tc-aarch64.c (REG_TYPE_SVE_BASE, REG_TYPE_SVE_OFFSET): New register types. (get_reg_expected_msg): Handle them. (aarch64_addr_reg_parse): New function, split out from aarch64_reg_parse_32_64. Handle Z registers too. (aarch64_reg_parse_32_64): Call it. (parse_address_main): Add base_qualifier, offset_qualifier, base_type and offset_type parameters. Handle SVE base and offset registers. (parse_address): Update call to parse_address_main. (parse_sve_address): New function. (parse_operands): Parse the new SVE address operands.
2016-09-21 23:55:49 +08:00
{ 16, 6 }, /* SVE_imm6: 6-bit immediate field. */
[AArch64][SVE 27/32] Add SVE integer immediate operands This patch adds the new SVE integer immediate operands. There are three kinds: - simple signed and unsigned ranges, but with new widths and positions. - 13-bit logical immediates. These have the same form as in base AArch64, but at a different bit position. In the case of the "MOV Zn.<T>, #<limm>" alias of DUPM, the logical immediate <limm> is not allowed to be a valid DUP immediate, since DUP is preferred over DUPM for constants that both instructions can handle. - a new 9-bit arithmetic immediate, of the form "<imm8>{, LSL #8}". In some contexts the operand is signed and in others it's unsigned. As an extension, we allow shifted immediates to be written as a single integer, e.g. "#256" is equivalent to "#1, LSL #8". We also use the shiftless form as the preferred disassembly, except for the special case of "#0, LSL #8" (a redundant encoding of 0). include/ * opcode/aarch64.h (AARCH64_OPND_SIMM5): New aarch64_opnd. (AARCH64_OPND_SVE_AIMM, AARCH64_OPND_SVE_ASIMM) (AARCH64_OPND_SVE_INV_LIMM, AARCH64_OPND_SVE_LIMM) (AARCH64_OPND_SVE_LIMM_MOV, AARCH64_OPND_SVE_SHLIMM_PRED) (AARCH64_OPND_SVE_SHLIMM_UNPRED, AARCH64_OPND_SVE_SHRIMM_PRED) (AARCH64_OPND_SVE_SHRIMM_UNPRED, AARCH64_OPND_SVE_SIMM5) (AARCH64_OPND_SVE_SIMM5B, AARCH64_OPND_SVE_SIMM6) (AARCH64_OPND_SVE_SIMM8, AARCH64_OPND_SVE_UIMM3) (AARCH64_OPND_SVE_UIMM7, AARCH64_OPND_SVE_UIMM8) (AARCH64_OPND_SVE_UIMM8_53): Likewise. (aarch64_sve_dupm_mov_immediate_p): Declare. opcodes/ * aarch64-tbl.h (AARCH64_OPERANDS): Add entries for the new SVE integer immediate operands. * aarch64-opc.h (FLD_SVE_immN, FLD_SVE_imm3, FLD_SVE_imm5) (FLD_SVE_imm5b, FLD_SVE_imm7, FLD_SVE_imm8, FLD_SVE_imm9) (FLD_SVE_immr, FLD_SVE_imms, FLD_SVE_tszh): New aarch64_field_kinds. * aarch64-opc.c (fields): Add corresponding entries. (operand_general_constraint_met_p): Handle the new SVE integer immediate operands. (aarch64_print_operand): Likewise. (aarch64_sve_dupm_mov_immediate_p): New function. * aarch64-opc-2.c: Regenerate. * aarch64-asm.h (ins_inv_limm, ins_sve_aimm, ins_sve_asimm) (ins_sve_limm_mov, ins_sve_shlimm, ins_sve_shrimm): New inserters. * aarch64-asm.c (aarch64_ins_limm_1): New function, split out from... (aarch64_ins_limm): ...here. (aarch64_ins_inv_limm): New function. (aarch64_ins_sve_aimm): Likewise. (aarch64_ins_sve_asimm): Likewise. (aarch64_ins_sve_limm_mov): Likewise. (aarch64_ins_sve_shlimm): Likewise. (aarch64_ins_sve_shrimm): Likewise. * aarch64-asm-2.c: Regenerate. * aarch64-dis.h (ext_inv_limm, ext_sve_aimm, ext_sve_asimm) (ext_sve_limm_mov, ext_sve_shlimm, ext_sve_shrimm): New extractors. * aarch64-dis.c (decode_limm): New function, split out from... (aarch64_ext_limm): ...here. (aarch64_ext_inv_limm): New function. (decode_sve_aimm): Likewise. (aarch64_ext_sve_aimm): Likewise. (aarch64_ext_sve_asimm): Likewise. (aarch64_ext_sve_limm_mov): Likewise. (aarch64_top_bit): Likewise. (aarch64_ext_sve_shlimm): Likewise. (aarch64_ext_sve_shrimm): Likewise. * aarch64-dis-2.c: Regenerate. gas/ * config/tc-aarch64.c (parse_operands): Handle the new SVE integer immediate operands.
2016-09-21 23:56:57 +08:00
{ 14, 7 }, /* SVE_imm7: 7-bit immediate field. */
{ 5, 8 }, /* SVE_imm8: 8-bit immediate field. */
{ 5, 9 }, /* SVE_imm9: 9-bit immediate field. */
{ 11, 6 }, /* SVE_immr: SVE equivalent of immr. */
{ 5, 6 }, /* SVE_imms: SVE equivalent of imms. */
[AArch64][SVE 25/32] Add support for SVE addressing modes This patch adds most of the new SVE addressing modes and associated operands. A follow-on patch adds MUL VL, since handling it separately makes the changes easier to read. The patch also introduces a new "operand-dependent data" field to the operand flags, based closely on the existing one for opcode flags. For SVE this new field needs only 2 bits, but it could be widened in future if necessary. include/ * opcode/aarch64.h (AARCH64_OPND_SVE_ADDR_RI_U6): New aarch64_opnd. (AARCH64_OPND_SVE_ADDR_RI_U6x2, AARCH64_OPND_SVE_ADDR_RI_U6x4) (AARCH64_OPND_SVE_ADDR_RI_U6x8, AARCH64_OPND_SVE_ADDR_RR) (AARCH64_OPND_SVE_ADDR_RR_LSL1, AARCH64_OPND_SVE_ADDR_RR_LSL2) (AARCH64_OPND_SVE_ADDR_RR_LSL3, AARCH64_OPND_SVE_ADDR_RX) (AARCH64_OPND_SVE_ADDR_RX_LSL1, AARCH64_OPND_SVE_ADDR_RX_LSL2) (AARCH64_OPND_SVE_ADDR_RX_LSL3, AARCH64_OPND_SVE_ADDR_RZ) (AARCH64_OPND_SVE_ADDR_RZ_LSL1, AARCH64_OPND_SVE_ADDR_RZ_LSL2) (AARCH64_OPND_SVE_ADDR_RZ_LSL3, AARCH64_OPND_SVE_ADDR_RZ_XTW_14) (AARCH64_OPND_SVE_ADDR_RZ_XTW_22, AARCH64_OPND_SVE_ADDR_RZ_XTW1_14) (AARCH64_OPND_SVE_ADDR_RZ_XTW1_22, AARCH64_OPND_SVE_ADDR_RZ_XTW2_14) (AARCH64_OPND_SVE_ADDR_RZ_XTW2_22, AARCH64_OPND_SVE_ADDR_RZ_XTW3_14) (AARCH64_OPND_SVE_ADDR_RZ_XTW3_22, AARCH64_OPND_SVE_ADDR_ZI_U5) (AARCH64_OPND_SVE_ADDR_ZI_U5x2, AARCH64_OPND_SVE_ADDR_ZI_U5x4) (AARCH64_OPND_SVE_ADDR_ZI_U5x8, AARCH64_OPND_SVE_ADDR_ZZ_LSL) (AARCH64_OPND_SVE_ADDR_ZZ_SXTW, AARCH64_OPND_SVE_ADDR_ZZ_UXTW): Likewise. opcodes/ * aarch64-tbl.h (AARCH64_OPERANDS): Add entries for the new SVE address operands. * aarch64-opc.h (FLD_SVE_imm6, FLD_SVE_msz, FLD_SVE_xs_14) (FLD_SVE_xs_22): New aarch64_field_kinds. (OPD_F_OD_MASK, OPD_F_OD_LSB, OPD_F_NO_ZR): New flags. (get_operand_specific_data): New function. * aarch64-opc.c (fields): Add entries for FLD_SVE_imm6, FLD_SVE_msz, FLD_SVE_xs_14 and FLD_SVE_xs_22. (operand_general_constraint_met_p): Handle the new SVE address operands. (sve_reg): New array. (get_addr_sve_reg_name): New function. (aarch64_print_operand): Handle the new SVE address operands. * aarch64-opc-2.c: Regenerate. * aarch64-asm.h (ins_sve_addr_ri_u6, ins_sve_addr_rr_lsl) (ins_sve_addr_rz_xtw, ins_sve_addr_zi_u5, ins_sve_addr_zz_lsl) (ins_sve_addr_zz_sxtw, ins_sve_addr_zz_uxtw): New inserters. * aarch64-asm.c (aarch64_ins_sve_addr_ri_u6): New function. (aarch64_ins_sve_addr_rr_lsl): Likewise. (aarch64_ins_sve_addr_rz_xtw): Likewise. (aarch64_ins_sve_addr_zi_u5): Likewise. (aarch64_ins_sve_addr_zz): Likewise. (aarch64_ins_sve_addr_zz_lsl): Likewise. (aarch64_ins_sve_addr_zz_sxtw): Likewise. (aarch64_ins_sve_addr_zz_uxtw): Likewise. * aarch64-asm-2.c: Regenerate. * aarch64-dis.h (ext_sve_addr_ri_u6, ext_sve_addr_rr_lsl) (ext_sve_addr_rz_xtw, ext_sve_addr_zi_u5, ext_sve_addr_zz_lsl) (ext_sve_addr_zz_sxtw, ext_sve_addr_zz_uxtw): New extractors. * aarch64-dis.c (aarch64_ext_sve_add_reg_imm): New function. (aarch64_ext_sve_addr_ri_u6): Likewise. (aarch64_ext_sve_addr_rr_lsl): Likewise. (aarch64_ext_sve_addr_rz_xtw): Likewise. (aarch64_ext_sve_addr_zi_u5): Likewise. (aarch64_ext_sve_addr_zz): Likewise. (aarch64_ext_sve_addr_zz_lsl): Likewise. (aarch64_ext_sve_addr_zz_sxtw): Likewise. (aarch64_ext_sve_addr_zz_uxtw): Likewise. * aarch64-dis-2.c: Regenerate. gas/ * config/tc-aarch64.c (REG_TYPE_SVE_BASE, REG_TYPE_SVE_OFFSET): New register types. (get_reg_expected_msg): Handle them. (aarch64_addr_reg_parse): New function, split out from aarch64_reg_parse_32_64. Handle Z registers too. (aarch64_reg_parse_32_64): Call it. (parse_address_main): Add base_qualifier, offset_qualifier, base_type and offset_type parameters. Handle SVE base and offset registers. (parse_address): Update call to parse_address_main. (parse_sve_address): New function. (parse_operands): Parse the new SVE address operands.
2016-09-21 23:55:49 +08:00
{ 10, 2 }, /* SVE_msz: 2-bit shift amount for ADR. */
[AArch64][SVE 23/32] Add SVE pattern and prfop operands The SVE instructions have two enumerated operands: one to select a vector pattern and another to select a prefetch operation. The latter is a cut-down version of the base AArch64 prefetch operation. Both types of operand can also be specified as raw enum values such as #31. Reserved values can only be specified this way. If it hadn't been for the pattern operand, I would have been tempted to use the existing parsing for prefetch operations and add extra checks for SVE. However, since the patterns needed new enum parsing code anyway, it seeemed cleaner to reuse it for the prefetches too. Because of the small number of enum values, I don't think we'd gain anything by using hash tables. include/ * opcode/aarch64.h (AARCH64_OPND_SVE_PATTERN): New aarch64_opnd. (AARCH64_OPND_SVE_PRFOP): Likewise. (aarch64_sve_pattern_array): Declare. (aarch64_sve_prfop_array): Likewise. opcodes/ * aarch64-tbl.h (AARCH64_OPERANDS): Add entries for AARCH64_OPND_SVE_PATTERN and AARCH64_OPND_SVE_PRFOP. * aarch64-opc.h (FLD_SVE_pattern): New aarch64_field_kind. (FLD_SVE_prfop): Likewise. * aarch64-opc.c: Include libiberty.h. (aarch64_sve_pattern_array): New variable. (aarch64_sve_prfop_array): Likewise. (fields): Add entries for FLD_SVE_pattern and FLD_SVE_prfop. (aarch64_print_operand): Handle AARCH64_OPND_SVE_PATTERN and AARCH64_OPND_SVE_PRFOP. * aarch64-asm-2.c: Regenerate. * aarch64-dis-2.c: Likewise. * aarch64-opc-2.c: Likewise. gas/ * config/tc-aarch64.c (parse_enum_string): New function. (po_enum_or_fail): New macro. (parse_operands): Handle AARCH64_OPND_SVE_PATTERN and AARCH64_OPND_SVE_PRFOP.
2016-09-21 23:54:53 +08:00
{ 5, 5 }, /* SVE_pattern: vector pattern enumeration. */
{ 0, 4 }, /* SVE_prfop: prefetch operation for SVE PRF[BHWD]. */
[AArch64] Additional SVE instructions This patch supports some additions to the SVE architecture prior to its public release. include/ * opcode/aarch64.h (AARCH64_OPND_SVE_ADDR_RI_S4x16) (AARCH64_OPND_SVE_IMM_ROT1, AARCH64_OPND_SVE_IMM_ROT2) (AARCH64_OPND_SVE_Zm3_INDEX, AARCH64_OPND_SVE_Zm3_22_INDEX) (AARCH64_OPND_SVE_Zm4_INDEX): New aarch64_opnds. opcodes/ * aarch64-tbl.h (OP_SVE_HMH, OP_SVE_VMU_HSD, OP_SVE_VMVU_HSD) (OP_SVE_VMVV_HSD, OP_SVE_VMVVU_HSD, OP_SVE_VM_HSD, OP_SVE_VUVV_HSD) (OP_SVE_VUV_HSD, OP_SVE_VU_HSD, OP_SVE_VVVU_H, OP_SVE_VVVU_S) (OP_SVE_VVVU_HSD, OP_SVE_VVV_D, OP_SVE_VVV_D_H, OP_SVE_VVV_H) (OP_SVE_VVV_HSD, OP_SVE_VVV_S, OP_SVE_VVV_S_B, OP_SVE_VVV_SD_BH) (OP_SVE_VV_BHSDQ, OP_SVE_VV_HSD, OP_SVE_VZVV_HSD, OP_SVE_VZV_HSD) (OP_SVE_V_HSD): New macros. (OP_SVE_VMU_SD, OP_SVE_VMVU_SD, OP_SVE_VM_SD, OP_SVE_VUVV_SD) (OP_SVE_VU_SD, OP_SVE_VVVU_SD, OP_SVE_VVV_SD, OP_SVE_VZVV_SD) (OP_SVE_VZV_SD, OP_SVE_V_SD): Delete. (aarch64_opcode_table): Add new SVE instructions. (aarch64_opcode_table): Use imm_rotate{1,2} instead of imm_rotate for rotation operands. Add new SVE operands. * aarch64-asm.h (ins_sve_addr_ri_s4): New inserter. (ins_sve_quad_index): Likewise. (ins_imm_rotate): Split into... (ins_imm_rotate1, ins_imm_rotate2): ...these two inserters. * aarch64-asm.c (aarch64_ins_imm_rotate): Split into... (aarch64_ins_imm_rotate1, aarch64_ins_imm_rotate2): ...these two functions. (aarch64_ins_sve_addr_ri_s4): New function. (aarch64_ins_sve_quad_index): Likewise. (do_misc_encoding): Handle "MOV Zn.Q, Qm". * aarch64-asm-2.c: Regenerate. * aarch64-dis.h (ext_sve_addr_ri_s4): New extractor. (ext_sve_quad_index): Likewise. (ext_imm_rotate): Split into... (ext_imm_rotate1, ext_imm_rotate2): ...these two extractors. * aarch64-dis.c (aarch64_ext_imm_rotate): Split into... (aarch64_ext_imm_rotate1, aarch64_ext_imm_rotate2): ...these two functions. (aarch64_ext_sve_addr_ri_s4): New function. (aarch64_ext_sve_quad_index): Likewise. (aarch64_ext_sve_index): Allow quad indices. (do_misc_decoding): Likewise. * aarch64-dis-2.c: Regenerate. * aarch64-opc.h (FLD_SVE_i3h, FLD_SVE_rot1, FLD_SVE_rot2): New aarch64_field_kinds. (OPD_F_OD_MASK): Widen by one bit. (OPD_F_NO_ZR): Bump accordingly. (get_operand_field_width): New function. * aarch64-opc.c (fields): Add new SVE fields. (operand_general_constraint_met_p): Handle new SVE operands. (aarch64_print_operand): Likewise. * aarch64-opc-2.c: Regenerate. gas/ * doc/c-aarch64.texi: Document that sve implies fp16, simd and compnum. * config/tc-aarch64.c (parse_vector_type_for_operand): Allow .q to be used with SVE registers. (parse_operands): Handle new SVE operands. (aarch64_features): Make "sve" require F16 rather than FP. Also require COMPNUM. * testsuite/gas/aarch64/sve.s: Add tests for new instructions. Include compnum tests. * testsuite/gas/aarch64/sve.d: Update accordingly. * testsuite/gas/aarch64/sve-invalid.s: Add tests for new instructions. * testsuite/gas/aarch64/sve-invalid.l: Update accordingly. Also update expected output for new FMOV and MOV alternatives.
2017-02-25 02:29:00 +08:00
{ 16, 1 }, /* SVE_rot1: 1-bit rotation amount. */
{ 10, 2 }, /* SVE_rot2: 2-bit rotation amount. */
{ 10, 1 }, /* SVE_rot3: 1-bit rotation amount at bit 10. */
{ 17, 2 }, /* SVE_size: 2-bit element size, bits [18,17]. */
{ 22, 1 }, /* SVE_sz: 1-bit element size select. */
{ 30, 1 }, /* SVE_sz2: 1-bit element size select. */
{ 16, 4 }, /* SVE_tsz: triangular size select. */
[AArch64][SVE 21/32] Add Zn and Pn registers This patch adds the Zn and Pn registers, and associated fields and operands. include/ * opcode/aarch64.h (AARCH64_OPND_CLASS_SVE_REG): New aarch64_operand_class. (AARCH64_OPND_CLASS_PRED_REG): Likewise. (AARCH64_OPND_SVE_Pd, AARCH64_OPND_SVE_Pg3, AARCH64_OPND_SVE_Pg4_5) (AARCH64_OPND_SVE_Pg4_10, AARCH64_OPND_SVE_Pg4_16) (AARCH64_OPND_SVE_Pm, AARCH64_OPND_SVE_Pn, AARCH64_OPND_SVE_Pt) (AARCH64_OPND_SVE_Za_5, AARCH64_OPND_SVE_Za_16, AARCH64_OPND_SVE_Zd) (AARCH64_OPND_SVE_Zm_5, AARCH64_OPND_SVE_Zm_16, AARCH64_OPND_SVE_Zn) (AARCH64_OPND_SVE_Zn_INDEX, AARCH64_OPND_SVE_ZnxN) (AARCH64_OPND_SVE_Zt, AARCH64_OPND_SVE_ZtxN): New aarch64_opnds. opcodes/ * aarch64-tbl.h (AARCH64_OPERANDS): Add entries for new SVE operands. * aarch64-opc.h (FLD_SVE_Pd, FLD_SVE_Pg3, FLD_SVE_Pg4_5) (FLD_SVE_Pg4_10, FLD_SVE_Pg4_16, FLD_SVE_Pm, FLD_SVE_Pn, FLD_SVE_Pt) (FLD_SVE_Za_5, FLD_SVE_Za_16, FLD_SVE_Zd, FLD_SVE_Zm_5, FLD_SVE_Zm_16) (FLD_SVE_Zn, FLD_SVE_Zt, FLD_SVE_tzsh): New aarch64_field_kinds. * aarch64-opc.c (fields): Add corresponding entries here. (operand_general_constraint_met_p): Check that SVE register lists have the correct length. Check the ranges of SVE index registers. Check for cases where p8-p15 are used in 3-bit predicate fields. (aarch64_print_operand): Handle the new SVE operands. * aarch64-opc-2.c: Regenerate. * aarch64-asm.h (ins_sve_index, ins_sve_reglist): New inserters. * aarch64-asm.c (aarch64_ins_sve_index): New function. (aarch64_ins_sve_reglist): Likewise. * aarch64-asm-2.c: Regenerate. * aarch64-dis.h (ext_sve_index, ext_sve_reglist): New extractors. * aarch64-dis.c (aarch64_ext_sve_index): New function. (aarch64_ext_sve_reglist): Likewise. * aarch64-dis-2.c: Regenerate. gas/ * config/tc-aarch64.c (NTA_HASVARWIDTH): New macro. (AARCH64_REG_TYPES): Add ZN and PN. (get_reg_expected_msg): Handle them. (parse_vector_type_for_operand): Add a reg_type parameter. Skip the width for Zn and Pn registers. (parse_typed_reg): Extend vector handling to Zn and Pn. Update the call to parse_vector_type_for_operand. Set HASVARTYPE for Zn and Pn, expecting the width to be 0. (parse_vector_reg_list): Restrict error about [BHSD]nn operands to REG_TYPE_VN. (vectype_to_qualifier): Use S_[BHSD] qualifiers for NTA_HASVARWIDTH. (parse_operands): Handle the new Zn and Pn operands. (REGSET16): New macro, split out from... (REGSET31): ...here. (reg_names): Add Zn and Pn entries.
2016-09-21 23:53:54 +08:00
{ 22, 2 }, /* SVE_tszh: triangular size select high, bits [23,22]. */
{ 8, 2 }, /* SVE_tszl_8: triangular size select low, bits [9,8]. */
{ 19, 2 }, /* SVE_tszl_19: triangular size select low, bits [20,19]. */
[AArch64][SVE 25/32] Add support for SVE addressing modes This patch adds most of the new SVE addressing modes and associated operands. A follow-on patch adds MUL VL, since handling it separately makes the changes easier to read. The patch also introduces a new "operand-dependent data" field to the operand flags, based closely on the existing one for opcode flags. For SVE this new field needs only 2 bits, but it could be widened in future if necessary. include/ * opcode/aarch64.h (AARCH64_OPND_SVE_ADDR_RI_U6): New aarch64_opnd. (AARCH64_OPND_SVE_ADDR_RI_U6x2, AARCH64_OPND_SVE_ADDR_RI_U6x4) (AARCH64_OPND_SVE_ADDR_RI_U6x8, AARCH64_OPND_SVE_ADDR_RR) (AARCH64_OPND_SVE_ADDR_RR_LSL1, AARCH64_OPND_SVE_ADDR_RR_LSL2) (AARCH64_OPND_SVE_ADDR_RR_LSL3, AARCH64_OPND_SVE_ADDR_RX) (AARCH64_OPND_SVE_ADDR_RX_LSL1, AARCH64_OPND_SVE_ADDR_RX_LSL2) (AARCH64_OPND_SVE_ADDR_RX_LSL3, AARCH64_OPND_SVE_ADDR_RZ) (AARCH64_OPND_SVE_ADDR_RZ_LSL1, AARCH64_OPND_SVE_ADDR_RZ_LSL2) (AARCH64_OPND_SVE_ADDR_RZ_LSL3, AARCH64_OPND_SVE_ADDR_RZ_XTW_14) (AARCH64_OPND_SVE_ADDR_RZ_XTW_22, AARCH64_OPND_SVE_ADDR_RZ_XTW1_14) (AARCH64_OPND_SVE_ADDR_RZ_XTW1_22, AARCH64_OPND_SVE_ADDR_RZ_XTW2_14) (AARCH64_OPND_SVE_ADDR_RZ_XTW2_22, AARCH64_OPND_SVE_ADDR_RZ_XTW3_14) (AARCH64_OPND_SVE_ADDR_RZ_XTW3_22, AARCH64_OPND_SVE_ADDR_ZI_U5) (AARCH64_OPND_SVE_ADDR_ZI_U5x2, AARCH64_OPND_SVE_ADDR_ZI_U5x4) (AARCH64_OPND_SVE_ADDR_ZI_U5x8, AARCH64_OPND_SVE_ADDR_ZZ_LSL) (AARCH64_OPND_SVE_ADDR_ZZ_SXTW, AARCH64_OPND_SVE_ADDR_ZZ_UXTW): Likewise. opcodes/ * aarch64-tbl.h (AARCH64_OPERANDS): Add entries for the new SVE address operands. * aarch64-opc.h (FLD_SVE_imm6, FLD_SVE_msz, FLD_SVE_xs_14) (FLD_SVE_xs_22): New aarch64_field_kinds. (OPD_F_OD_MASK, OPD_F_OD_LSB, OPD_F_NO_ZR): New flags. (get_operand_specific_data): New function. * aarch64-opc.c (fields): Add entries for FLD_SVE_imm6, FLD_SVE_msz, FLD_SVE_xs_14 and FLD_SVE_xs_22. (operand_general_constraint_met_p): Handle the new SVE address operands. (sve_reg): New array. (get_addr_sve_reg_name): New function. (aarch64_print_operand): Handle the new SVE address operands. * aarch64-opc-2.c: Regenerate. * aarch64-asm.h (ins_sve_addr_ri_u6, ins_sve_addr_rr_lsl) (ins_sve_addr_rz_xtw, ins_sve_addr_zi_u5, ins_sve_addr_zz_lsl) (ins_sve_addr_zz_sxtw, ins_sve_addr_zz_uxtw): New inserters. * aarch64-asm.c (aarch64_ins_sve_addr_ri_u6): New function. (aarch64_ins_sve_addr_rr_lsl): Likewise. (aarch64_ins_sve_addr_rz_xtw): Likewise. (aarch64_ins_sve_addr_zi_u5): Likewise. (aarch64_ins_sve_addr_zz): Likewise. (aarch64_ins_sve_addr_zz_lsl): Likewise. (aarch64_ins_sve_addr_zz_sxtw): Likewise. (aarch64_ins_sve_addr_zz_uxtw): Likewise. * aarch64-asm-2.c: Regenerate. * aarch64-dis.h (ext_sve_addr_ri_u6, ext_sve_addr_rr_lsl) (ext_sve_addr_rz_xtw, ext_sve_addr_zi_u5, ext_sve_addr_zz_lsl) (ext_sve_addr_zz_sxtw, ext_sve_addr_zz_uxtw): New extractors. * aarch64-dis.c (aarch64_ext_sve_add_reg_imm): New function. (aarch64_ext_sve_addr_ri_u6): Likewise. (aarch64_ext_sve_addr_rr_lsl): Likewise. (aarch64_ext_sve_addr_rz_xtw): Likewise. (aarch64_ext_sve_addr_zi_u5): Likewise. (aarch64_ext_sve_addr_zz): Likewise. (aarch64_ext_sve_addr_zz_lsl): Likewise. (aarch64_ext_sve_addr_zz_sxtw): Likewise. (aarch64_ext_sve_addr_zz_uxtw): Likewise. * aarch64-dis-2.c: Regenerate. gas/ * config/tc-aarch64.c (REG_TYPE_SVE_BASE, REG_TYPE_SVE_OFFSET): New register types. (get_reg_expected_msg): Handle them. (aarch64_addr_reg_parse): New function, split out from aarch64_reg_parse_32_64. Handle Z registers too. (aarch64_reg_parse_32_64): Call it. (parse_address_main): Add base_qualifier, offset_qualifier, base_type and offset_type parameters. Handle SVE base and offset registers. (parse_address): Update call to parse_address_main. (parse_sve_address): New function. (parse_operands): Parse the new SVE address operands.
2016-09-21 23:55:49 +08:00
{ 14, 1 }, /* SVE_xs_14: UXTW/SXTW select (bit 14). */
[AArch64] Add ARMv8.3 FCMLA and FCADD instructions Add support for FCMLA and FCADD complex arithmetic SIMD instructions. FCMLA has an indexed element variant where the index range has to be treated specially because a complex number takes two elements and the indexed vector size depends on the other operands. These complex number SIMD instructions are part of ARMv8.3 https://community.arm.com/groups/processors/blog/2016/10/27/armv8-a-architecture-2016-additions include/ 2016-11-18 Szabolcs Nagy <szabolcs.nagy@arm.com> * opcode/aarch64.h (enum aarch64_opnd): Add AARCH64_OPND_IMM_ROT1, AARCH64_OPND_IMM_ROT2, AARCH64_OPND_IMM_ROT3. (enum aarch64_op): Add OP_FCMLA_ELEM. opcodes/ 2016-11-18 Szabolcs Nagy <szabolcs.nagy@arm.com> * aarch64-tbl.h (QL_V3SAMEHSD_ROT, QL_ELEMENT_ROT): Define. (aarch64_feature_simd_v8_3, SIMD_V8_3): Define. (aarch64_opcode_table): Add fcmla and fcadd. (AARCH64_OPERANDS): Add IMM_ROT{1,2,3}. * aarch64-asm.h (aarch64_ins_imm_rotate): Declare. * aarch64-asm.c (aarch64_ins_imm_rotate): Define. * aarch64-dis.h (aarch64_ext_imm_rotate): Declare. * aarch64-dis.c (aarch64_ext_imm_rotate): Define. * aarch64-opc.h (enum aarch64_field_kind): Add FLD_rotate{1,2,3}. * aarch64-opc.c (fields): Add FLD_rotate{1,2,3}. (operand_general_constraint_met_p): Rotate and index range check. (aarch64_print_operand): Handle rotate operand. * aarch64-asm-2.c: Regenerate. * aarch64-dis-2.c: Likewise. * aarch64-opc-2.c: Likewise. gas/ 2016-11-18 Szabolcs Nagy <szabolcs.nagy@arm.com> * config/tc-aarch64.c (parse_operands): Handle AARCH64_OPND_IMM_ROT*. * testsuite/gas/aarch64/advsimd-armv8_3.d: New. * testsuite/gas/aarch64/advsimd-armv8_3.s: New. * testsuite/gas/aarch64/illegal-fcmla.s: New. * testsuite/gas/aarch64/illegal-fcmla.l: New. * testsuite/gas/aarch64/illegal-fcmla.d: New.
2016-11-18 18:02:16 +08:00
{ 22, 1 }, /* SVE_xs_22: UXTW/SXTW select (bit 22). */
{ 22, 1 }, /* S_imm10: in LDRAA and LDRAB instructions. */
{ 16, 3 }, /* abc: a:b:c bits in AdvSIMD modified immediate. */
{ 13, 3 }, /* asisdlso_opcode: opcode in advsimd ld/st single element. */
{ 19, 5 }, /* b40: in the test bit and branch instructions. */
{ 31, 1 }, /* b5: in the test bit and branch instructions. */
{ 12, 4 }, /* cmode: in advsimd modified immediate instructions. */
{ 12, 4 }, /* cond: condition flags as a source operand. */
{ 0, 4 }, /* cond2: condition in truly conditional-executed inst. */
{ 5, 5 }, /* defgh: d:e:f:g:h bits in AdvSIMD modified immediate. */
{ 21, 2 }, /* hw: in move wide constant instructions. */
{ 0, 1 }, /* imm1_0: general immediate in bits [0]. */
{ 2, 1 }, /* imm1_2: general immediate in bits [2]. */
{ 8, 1 }, /* imm1_8: general immediate in bits [8]. */
{ 10, 1 }, /* imm1_10: general immediate in bits [10]. */
{ 15, 1 }, /* imm1_15: general immediate in bits [15]. */
{ 16, 1 }, /* imm1_16: general immediate in bits [16]. */
{ 0, 2 }, /* imm2_0: general immediate in bits [1:0]. */
{ 1, 2 }, /* imm2_1: general immediate in bits [2:1]. */
{ 8, 2 }, /* imm2_8: general immediate in bits [9:8]. */
{ 10, 2 }, /* imm2_10: 2-bit immediate, bits [11:10] */
{ 12, 2 }, /* imm2_12: 2-bit immediate, bits [13:12] */
{ 15, 2 }, /* imm2_15: 2-bit immediate, bits [16:15] */
{ 16, 2 }, /* imm2_16: 2-bit immediate, bits [17:16] */
{ 19, 2 }, /* imm2_19: 2-bit immediate, bits [20:19] */
{ 0, 3 }, /* imm3_0: general immediate in bits [2:0]. */
{ 5, 3 }, /* imm3_5: general immediate in bits [7:5]. */
{ 10, 3 }, /* imm3_10: in add/sub extended reg instructions. */
{ 12, 3 }, /* imm3_12: general immediate in bits [14:12]. */
{ 14, 3 }, /* imm3_14: general immediate in bits [16:14]. */
{ 15, 3 }, /* imm3_15: general immediate in bits [17:15]. */
{ 0, 4 }, /* imm4_0: in rmif instructions. */
{ 5, 4 }, /* imm4_5: in SME instructions. */
{ 10, 4 }, /* imm4_10: in adddg/subg instructions. */
{ 11, 4 }, /* imm4_11: in advsimd ext and advsimd ins instructions. */
{ 14, 4 }, /* imm4_14: general immediate in bits [17:14]. */
{ 16, 5 }, /* imm5: in conditional compare (immediate) instructions. */
{ 10, 6 }, /* imm6_10: in add/sub reg shifted instructions. */
{ 15, 6 }, /* imm6_15: in rmif instructions. */
{ 15, 7 }, /* imm7: in load/store pair pre/post index instructions. */
{ 13, 8 }, /* imm8: in floating-point scalar move immediate inst. */
{ 12, 9 }, /* imm9: in load/store pre/post index instructions. */
{ 10, 12 }, /* imm12: in ld/st unsigned imm or add/sub shifted inst. */
{ 5, 14 }, /* imm14: in test bit and branch instructions. */
{ 0, 16 }, /* imm16_0: in udf instruction. */
{ 5, 16 }, /* imm16_5: in exception instructions. */
{ 5, 19 }, /* imm19: e.g. in CBZ. */
{ 0, 26 }, /* imm26: in unconditional branch instructions. */
{ 16, 3 }, /* immb: in advsimd shift by immediate instructions. */
{ 19, 4 }, /* immh: in advsimd shift by immediate instructions. */
{ 5, 19 }, /* immhi: e.g. in ADRP. */
{ 29, 2 }, /* immlo: e.g. in ADRP. */
{ 16, 6 }, /* immr: in bitfield and logical immediate instructions. */
{ 10, 6 }, /* imms: in bitfield and logical immediate instructions. */
{ 11, 1 }, /* index: in ld/st inst deciding the pre/post-index. */
{ 24, 1 }, /* index2: in ld/st pair inst deciding the pre/post-index. */
{ 30, 2 }, /* ldst_size: size field in ld/st reg offset inst. */
{ 13, 2 }, /* len: in advsimd tbl/tbx instructions. */
{ 30, 1 }, /* lse_sz: in LSE extension atomic instructions. */
{ 0, 4 }, /* nzcv: flag bit specifier, encoded in the "nzcv" field. */
{ 29, 1 }, /* op: in AdvSIMD modified immediate instructions. */
{ 19, 2 }, /* op0: in the system instructions. */
{ 16, 3 }, /* op1: in the system instructions. */
{ 5, 3 }, /* op2: in the system instructions. */
{ 22, 2 }, /* opc: in load/store reg offset instructions. */
{ 23, 1 }, /* opc1: in load/store reg offset instructions. */
{ 12, 4 }, /* opcode: in advsimd load/store instructions. */
{ 13, 3 }, /* option: in ld/st reg offset + add/sub extended reg inst. */
[AArch64] Add ARMv8.3 FCMLA and FCADD instructions Add support for FCMLA and FCADD complex arithmetic SIMD instructions. FCMLA has an indexed element variant where the index range has to be treated specially because a complex number takes two elements and the indexed vector size depends on the other operands. These complex number SIMD instructions are part of ARMv8.3 https://community.arm.com/groups/processors/blog/2016/10/27/armv8-a-architecture-2016-additions include/ 2016-11-18 Szabolcs Nagy <szabolcs.nagy@arm.com> * opcode/aarch64.h (enum aarch64_opnd): Add AARCH64_OPND_IMM_ROT1, AARCH64_OPND_IMM_ROT2, AARCH64_OPND_IMM_ROT3. (enum aarch64_op): Add OP_FCMLA_ELEM. opcodes/ 2016-11-18 Szabolcs Nagy <szabolcs.nagy@arm.com> * aarch64-tbl.h (QL_V3SAMEHSD_ROT, QL_ELEMENT_ROT): Define. (aarch64_feature_simd_v8_3, SIMD_V8_3): Define. (aarch64_opcode_table): Add fcmla and fcadd. (AARCH64_OPERANDS): Add IMM_ROT{1,2,3}. * aarch64-asm.h (aarch64_ins_imm_rotate): Declare. * aarch64-asm.c (aarch64_ins_imm_rotate): Define. * aarch64-dis.h (aarch64_ext_imm_rotate): Declare. * aarch64-dis.c (aarch64_ext_imm_rotate): Define. * aarch64-opc.h (enum aarch64_field_kind): Add FLD_rotate{1,2,3}. * aarch64-opc.c (fields): Add FLD_rotate{1,2,3}. (operand_general_constraint_met_p): Rotate and index range check. (aarch64_print_operand): Handle rotate operand. * aarch64-asm-2.c: Regenerate. * aarch64-dis-2.c: Likewise. * aarch64-opc-2.c: Likewise. gas/ 2016-11-18 Szabolcs Nagy <szabolcs.nagy@arm.com> * config/tc-aarch64.c (parse_operands): Handle AARCH64_OPND_IMM_ROT*. * testsuite/gas/aarch64/advsimd-armv8_3.d: New. * testsuite/gas/aarch64/advsimd-armv8_3.s: New. * testsuite/gas/aarch64/illegal-fcmla.s: New. * testsuite/gas/aarch64/illegal-fcmla.l: New. * testsuite/gas/aarch64/illegal-fcmla.d: New.
2016-11-18 18:02:16 +08:00
{ 11, 2 }, /* rotate1: FCMLA immediate rotate. */
{ 13, 2 }, /* rotate2: Indexed element FCMLA immediate rotate. */
{ 12, 1 }, /* rotate3: FCADD immediate rotate. */
{ 10, 6 }, /* scale: in the fixed-point scalar to fp converting inst. */
{ 31, 1 }, /* sf: in integer data processing instructions. */
{ 22, 2 }, /* shift: in add/sub reg/imm shifted instructions. */
{ 22, 2 }, /* size: in most AdvSIMD and floating-point instructions. */
{ 22, 1 }, /* sz: 1-bit element size select. */
{ 22, 2 }, /* type: floating point type field in fp data inst. */
{ 10, 2 }, /* vldst_size: size field in the AdvSIMD load/store inst. */
{ 5, 3 }, /* off3: immediate offset used to calculate slice number in a
ZA tile. */
{ 5, 2 }, /* off2: immediate offset used to calculate slice number in
a ZA tile. */
{ 7, 1 }, /* ZAn_1: name of the 1bit encoded ZA tile. */
{ 5, 1 }, /* ol: immediate offset used to calculate slice number in a ZA
tile. */
{ 6, 2 }, /* ZAn_2: name of the 2bit encoded ZA tile. */
{ 5, 3 }, /* ZAn_3: name of the 3bit encoded ZA tile. */
{ 6, 1 }, /* ZAn: name of the bit encoded ZA tile. */
{ 12, 4 }, /* opc2: in rcpc3 ld/st inst deciding the pre/post-index. */
{ 30, 2 }, /* rcpc3_size: in rcpc3 ld/st, field controls Rt/Rt2 width. */
};
enum aarch64_operand_class
aarch64_get_operand_class (enum aarch64_opnd type)
{
return aarch64_operands[type].op_class;
}
const char *
aarch64_get_operand_name (enum aarch64_opnd type)
{
return aarch64_operands[type].name;
}
/* Get operand description string.
This is usually for the diagnosis purpose. */
const char *
aarch64_get_operand_desc (enum aarch64_opnd type)
{
return aarch64_operands[type].desc;
}
/* Table of all conditional affixes. */
const aarch64_cond aarch64_conds[16] =
{
[AArch64] Add SVE condition codes SVE defines new names for existing NZCV conditions, to reflect the result of instructions like PTEST. This patch adds support for these names. The patch also adds comments to the disassembly output to show the alternative names of a condition code. For example: cinv x0, x1, cc becomes: cinv x0, x1, cc // cc = lo, ul, last and: b.cc f0 <...> becomes: b.cc f0 <...> // b.lo, b.ul, b.last Doing this for the SVE names follows the practice recommended by the SVE specification and is definitely useful when reading SVE code. If the feeling is that it's too distracting elsewhere, we could add an option to turn it off. include/ * opcode/aarch64.h (aarch64_cond): Bump array size to 4. opcodes/ * aarch64-dis.c (remove_dot_suffix): New function, split out from... (print_mnemonic_name): ...here. (print_comment): New function. (print_aarch64_insn): Call it. * aarch64-opc.c (aarch64_conds): Add SVE names. (aarch64_print_operand): Print alternative condition names in a comment. gas/ * config/tc-aarch64.c (opcode_lookup): Search for the end of a condition name, rather than assuming that it will have exactly 2 characters. (parse_operands): Likewise. * testsuite/gas/aarch64/alias.d: Add new condition-code comments to the expected output. * testsuite/gas/aarch64/beq_1.d: Likewise. * testsuite/gas/aarch64/float-fp16.d: Likewise. * testsuite/gas/aarch64/int-insns.d: Likewise. * testsuite/gas/aarch64/no-aliases.d: Likewise. * testsuite/gas/aarch64/programmer-friendly.d: Likewise. * testsuite/gas/aarch64/reloc-insn.d: Likewise. * testsuite/gas/aarch64/b_c_1.d, testsuite/gas/aarch64/b_c_1.s: New test. ld/ * testsuite/ld-aarch64/emit-relocs-280.d: Match branch comments. * testsuite/ld-aarch64/weak-undefined.d: Likewise.
2016-09-22 00:09:59 +08:00
{{"eq", "none"}, 0x0},
{{"ne", "any"}, 0x1},
{{"cs", "hs", "nlast"}, 0x2},
{{"cc", "lo", "ul", "last"}, 0x3},
{{"mi", "first"}, 0x4},
{{"pl", "nfrst"}, 0x5},
{{"vs"}, 0x6},
{{"vc"}, 0x7},
[AArch64] Add SVE condition codes SVE defines new names for existing NZCV conditions, to reflect the result of instructions like PTEST. This patch adds support for these names. The patch also adds comments to the disassembly output to show the alternative names of a condition code. For example: cinv x0, x1, cc becomes: cinv x0, x1, cc // cc = lo, ul, last and: b.cc f0 <...> becomes: b.cc f0 <...> // b.lo, b.ul, b.last Doing this for the SVE names follows the practice recommended by the SVE specification and is definitely useful when reading SVE code. If the feeling is that it's too distracting elsewhere, we could add an option to turn it off. include/ * opcode/aarch64.h (aarch64_cond): Bump array size to 4. opcodes/ * aarch64-dis.c (remove_dot_suffix): New function, split out from... (print_mnemonic_name): ...here. (print_comment): New function. (print_aarch64_insn): Call it. * aarch64-opc.c (aarch64_conds): Add SVE names. (aarch64_print_operand): Print alternative condition names in a comment. gas/ * config/tc-aarch64.c (opcode_lookup): Search for the end of a condition name, rather than assuming that it will have exactly 2 characters. (parse_operands): Likewise. * testsuite/gas/aarch64/alias.d: Add new condition-code comments to the expected output. * testsuite/gas/aarch64/beq_1.d: Likewise. * testsuite/gas/aarch64/float-fp16.d: Likewise. * testsuite/gas/aarch64/int-insns.d: Likewise. * testsuite/gas/aarch64/no-aliases.d: Likewise. * testsuite/gas/aarch64/programmer-friendly.d: Likewise. * testsuite/gas/aarch64/reloc-insn.d: Likewise. * testsuite/gas/aarch64/b_c_1.d, testsuite/gas/aarch64/b_c_1.s: New test. ld/ * testsuite/ld-aarch64/emit-relocs-280.d: Match branch comments. * testsuite/ld-aarch64/weak-undefined.d: Likewise.
2016-09-22 00:09:59 +08:00
{{"hi", "pmore"}, 0x8},
{{"ls", "plast"}, 0x9},
{{"ge", "tcont"}, 0xa},
{{"lt", "tstop"}, 0xb},
{{"gt"}, 0xc},
{{"le"}, 0xd},
{{"al"}, 0xe},
{{"nv"}, 0xf},
};
const aarch64_cond *
get_cond_from_value (aarch64_insn value)
{
assert (value < 16);
return &aarch64_conds[(unsigned int) value];
}
const aarch64_cond *
get_inverted_cond (const aarch64_cond *cond)
{
return &aarch64_conds[cond->value ^ 0x1];
}
/* Table describing the operand extension/shifting operators; indexed by
enum aarch64_modifier_kind.
The value column provides the most common values for encoding modifiers,
which enables table-driven encoding/decoding for the modifiers. */
const struct aarch64_name_value_pair aarch64_operand_modifiers [] =
{
{"none", 0x0},
{"msl", 0x0},
{"ror", 0x3},
{"asr", 0x2},
{"lsr", 0x1},
{"lsl", 0x0},
{"uxtb", 0x0},
{"uxth", 0x1},
{"uxtw", 0x2},
{"uxtx", 0x3},
{"sxtb", 0x4},
{"sxth", 0x5},
{"sxtw", 0x6},
{"sxtx", 0x7},
[AArch64][SVE 24/32] Add AARCH64_OPND_SVE_PATTERN_SCALED Some SVE instructions count the number of elements in a given vector pattern and allow a scale factor of [1, 16] to be applied to the result. This scale factor is written ", MUL #n", where "MUL" is a new operator. E.g.: UQINCD X0, POW2, MUL #2 This patch adds support for this kind of operand. All existing operators were shifts of some kind, so there was a natural range of [0, 63] regardless of context. This was then narrowered further by later checks (e.g. to [0, 31] when used for 32-bit values). In contrast, MUL doesn't really have a natural context-independent range. Rather than pick one arbitrarily, it seemed better to make the "shift" amount a full 64-bit value and leave the range test to the usual operand-checking code. I've rearranged the fields of aarch64_opnd_info so that this doesn't increase the size of the structure (although I don't think its size is critical anyway). include/ * opcode/aarch64.h (AARCH64_OPND_SVE_PATTERN_SCALED): New aarch64_opnd. (AARCH64_MOD_MUL): New aarch64_modifier_kind. (aarch64_opnd_info): Make shifter.amount an int64_t and rearrange the fields. opcodes/ * aarch64-tbl.h (AARCH64_OPERANDS): Add an entry for AARCH64_OPND_SVE_PATTERN_SCALED. * aarch64-opc.h (FLD_SVE_imm4): New aarch64_field_kind. * aarch64-opc.c (fields): Add a corresponding entry. (set_multiplier_out_of_range_error): New function. (aarch64_operand_modifiers): Add entry for AARCH64_MOD_MUL. (operand_general_constraint_met_p): Handle AARCH64_OPND_SVE_PATTERN_SCALED. (print_register_offset_address): Use PRIi64 to print the shift amount. (aarch64_print_operand): Likewise. Handle AARCH64_OPND_SVE_PATTERN_SCALED. * aarch64-opc-2.c: Regenerate. * aarch64-asm.h (ins_sve_scale): New inserter. * aarch64-asm.c (aarch64_ins_sve_scale): New function. * aarch64-asm-2.c: Regenerate. * aarch64-dis.h (ext_sve_scale): New inserter. * aarch64-dis.c (aarch64_ext_sve_scale): New function. * aarch64-dis-2.c: Regenerate. gas/ * config/tc-aarch64.c (SHIFTED_MUL): New parse_shift_mode. (parse_shift): Handle it. Reject AARCH64_MOD_MUL for all other shift modes. Skip range tests for AARCH64_MOD_MUL. (process_omitted_operand): Handle AARCH64_OPND_SVE_PATTERN_SCALED. (parse_operands): Likewise.
2016-09-21 23:55:22 +08:00
{"mul", 0x0},
[AArch64][SVE 26/32] Add SVE MUL VL addressing modes This patch adds support for addresses of the form: [<base>, #<offset>, MUL VL] This involves adding a new AARCH64_MOD_MUL_VL modifier, which is why I split it out from the other addressing modes. For LD2, LD3 and LD4, the offset must be a multiple of the structure size, so for LD3 the possible values are 0, 3, 6, .... The patch therefore extends value_aligned_p to handle non-power-of-2 alignments. include/ * opcode/aarch64.h (AARCH64_OPND_SVE_ADDR_RI_S4xVL): New aarch64_opnd. (AARCH64_OPND_SVE_ADDR_RI_S4x2xVL, AARCH64_OPND_SVE_ADDR_RI_S4x3xVL) (AARCH64_OPND_SVE_ADDR_RI_S4x4xVL, AARCH64_OPND_SVE_ADDR_RI_S6xVL) (AARCH64_OPND_SVE_ADDR_RI_S9xVL): Likewise. (AARCH64_MOD_MUL_VL): New aarch64_modifier_kind. opcodes/ * aarch64-tbl.h (AARCH64_OPERANDS): Add entries for new MUL VL operands. * aarch64-opc.c (aarch64_operand_modifiers): Initialize the AARCH64_MOD_MUL_VL entry. (value_aligned_p): Cope with non-power-of-two alignments. (operand_general_constraint_met_p): Handle the new MUL VL addresses. (print_immediate_offset_address): Likewise. (aarch64_print_operand): Likewise. * aarch64-opc-2.c: Regenerate. * aarch64-asm.h (ins_sve_addr_ri_s4xvl, ins_sve_addr_ri_s6xvl) (ins_sve_addr_ri_s9xvl): New inserters. * aarch64-asm.c (aarch64_ins_sve_addr_ri_s4xvl): New function. (aarch64_ins_sve_addr_ri_s6xvl): Likewise. (aarch64_ins_sve_addr_ri_s9xvl): Likewise. * aarch64-asm-2.c: Regenerate. * aarch64-dis.h (ext_sve_addr_ri_s4xvl, ext_sve_addr_ri_s6xvl) (ext_sve_addr_ri_s9xvl): New extractors. * aarch64-dis.c (aarch64_ext_sve_addr_reg_mul_vl): New function. (aarch64_ext_sve_addr_ri_s4xvl): Likewise. (aarch64_ext_sve_addr_ri_s6xvl): Likewise. (aarch64_ext_sve_addr_ri_s9xvl): Likewise. * aarch64-dis-2.c: Regenerate. gas/ * config/tc-aarch64.c (SHIFTED_NONE, SHIFTED_MUL_VL): New parse_shift_modes. (parse_shift): Handle SHIFTED_MUL_VL. (parse_address_main): Add an imm_shift_mode parameter. (parse_address, parse_sve_address): Update accordingly. (parse_operands): Handle MUL VL addressing modes.
2016-09-21 23:56:15 +08:00
{"mul vl", 0x0},
{NULL, 0},
};
enum aarch64_modifier_kind
aarch64_get_operand_modifier (const struct aarch64_name_value_pair *desc)
{
return desc - aarch64_operand_modifiers;
}
aarch64_insn
aarch64_get_operand_modifier_value (enum aarch64_modifier_kind kind)
{
return aarch64_operand_modifiers[kind].value;
}
enum aarch64_modifier_kind
aarch64_get_operand_modifier_from_value (aarch64_insn value,
bool extend_p)
{
TRUE/FALSE simplification There is really no need to write code like "foo != 0 ? TRUE : FALSE" unless we had stupidly defined FALSE as something other than 0 or TRUE as something other than 1. The simpler "foo != 0" does just as well. Similarly "(condition == TRUE)" or "(condition == FALSE) can be simplified to "(condition)" and "(!condition)" respectively. I'll note that there is reason to use "integer_expression != 0" when assigning a bfd_boolean rather than the simpler "integer_expression", if you expect the variable to have 0 or 1 value. It's probably even a good idea to not rely on implicit conversion if bfd_boolean were _Bool. bfd/ * aoutx.h (aout_link_write_symbols): Don't cast boolean expression to bfd_boolean. * elf32-or1k.c (or1k_set_got_and_rela_sizes): Dont compare booleans against FALSE. * elf32-arc.c (name_for_global_symbol): Don't compare boolean to TRUE. (is_reloc_PC_relative): Don't use "boolean_condition ? TRUE : FALSE". (is_reloc_SDA_relative, is_reloc_for_GOT): Likewise. (is_reloc_for_PLT, is_reloc_for_TLS): Likewise. * elf32-arm.c (stm32l4xx_need_create_replacing_stub): Likewise. * elf32-nds32.c (insert_nds32_elf_blank): Likewise. * elf32-rx.c (rx_set_section_contents): Likewise. * elfnn-aarch64.c (elfNN_aarch64_final_link_relocate): Likewise. * elfxx-mips.c (_bfd_mips_elf_ignore_undef_symbol): Likewise. * mach-o.c (bfd_mach_o_read_command): Likewise. * targets.c (bfd_get_target_info): Likewise. binutils/ * dlltool.c (main): Don't use "boolean_condition ? TRUE : FALSE". * dwarf.c (read_and_display_attr_value): Likewise. (display_debug_str_offsets): Likewise. * objdump.c (dump_bfd): Likewise. * readelf.c (dump_section_as_strings): Likewise. (dump_section_as_bytes): Likewise. gas/ * atof-generic.c (FALSE, TRUE): Don't define. * config/obj-elf.h (FALSE, TRUE): Don't define. * config/obj-som.h (FALSE, TRUE): Don't define. * config/tc-hppa.h (FALSE, TRUE): Don't define. * config/tc-pdp11.c (FALSE, TRUE): Don't define. * config/tc-iq2000.h (obj_fix_adjustable): Delete. * config/tc-m32r.h (TC_FIX_ADJUSTABLE): Delete. * config/tc-mt.h (obj_fix_adjustable): Delete. * config/tc-nds32.h (TC_FIX_ADJUSTABLE): Delete. * config/tc-arc.c (parse_opcode_flags): Simplify boolean expression. (relaxable_flag, relaxable_operand, assemble_insn): Likewise. (tokenize_extregister): Likewise. * config/tc-csky.c (parse_opcode, get_operand_value): Likewise. (parse_operands_op, parse_operands, md_assemble): Likewise. * config/tc-d10v.c (build_insn): Likewise. * config/tc-score.c (s3_gen_insn_frag): Likewise. * config/tc-score7.c (s7_gen_insn_frag, s7_relax_frag): Likewise. * config/tc-tic6x.c (tic6x_update_features, md_assemble): Likewise. * config/tc-z80.c (emit_byte): Likewise. include/ * opcode/aarch64.h (alias_opcode_p): Simplify boolean expression. (opcode_has_alias, pseudo_opcode_p, optional_operand_p): Likewise. (opcode_has_special_coder): Likewise. ld/ * emultempl/aix.em (gld${EMULATION_NAME}_before_allocation): Simplify boolean expression. * lexsup.c (parse_args): Likewise. * pe-dll.c (pe_dll_id_target): Likewise. opcodes/ * aarch64-opc.c (vector_qualifier_p): Simplify boolean expression. (fp_qualifier_p, get_data_pattern): Likewise. (aarch64_get_operand_modifier_from_value): Likewise. (aarch64_extend_operator_p, aarch64_shift_operator_p): Likewise. (operand_variant_qualifier_p): Likewise. (qualifier_value_in_range_constraint_p): Likewise. (aarch64_get_qualifier_esize): Likewise. (aarch64_get_qualifier_nelem): Likewise. (aarch64_get_qualifier_standard_value): Likewise. (get_lower_bound, get_upper_bound): Likewise. (aarch64_find_best_match, match_operands_qualifier): Likewise. (aarch64_print_operand): Likewise. * aarch64-opc.h (operand_has_inserter, operand_has_extractor): Likewise. (operand_need_sign_extension, operand_need_shift_by_two): Likewise. (operand_need_shift_by_four, operand_maybe_stack_pointer): Likewise. * arm-dis.c (print_insn_mve, print_insn_thumb32): Likewise. * tic6x-dis.c (tic6x_check_fetch_packet_header): Likewise. (print_insn_tic6x): Likewise.
2021-03-29 07:22:56 +08:00
if (extend_p)
return AARCH64_MOD_UXTB + value;
else
return AARCH64_MOD_LSL - value;
}
bool
aarch64_extend_operator_p (enum aarch64_modifier_kind kind)
{
TRUE/FALSE simplification There is really no need to write code like "foo != 0 ? TRUE : FALSE" unless we had stupidly defined FALSE as something other than 0 or TRUE as something other than 1. The simpler "foo != 0" does just as well. Similarly "(condition == TRUE)" or "(condition == FALSE) can be simplified to "(condition)" and "(!condition)" respectively. I'll note that there is reason to use "integer_expression != 0" when assigning a bfd_boolean rather than the simpler "integer_expression", if you expect the variable to have 0 or 1 value. It's probably even a good idea to not rely on implicit conversion if bfd_boolean were _Bool. bfd/ * aoutx.h (aout_link_write_symbols): Don't cast boolean expression to bfd_boolean. * elf32-or1k.c (or1k_set_got_and_rela_sizes): Dont compare booleans against FALSE. * elf32-arc.c (name_for_global_symbol): Don't compare boolean to TRUE. (is_reloc_PC_relative): Don't use "boolean_condition ? TRUE : FALSE". (is_reloc_SDA_relative, is_reloc_for_GOT): Likewise. (is_reloc_for_PLT, is_reloc_for_TLS): Likewise. * elf32-arm.c (stm32l4xx_need_create_replacing_stub): Likewise. * elf32-nds32.c (insert_nds32_elf_blank): Likewise. * elf32-rx.c (rx_set_section_contents): Likewise. * elfnn-aarch64.c (elfNN_aarch64_final_link_relocate): Likewise. * elfxx-mips.c (_bfd_mips_elf_ignore_undef_symbol): Likewise. * mach-o.c (bfd_mach_o_read_command): Likewise. * targets.c (bfd_get_target_info): Likewise. binutils/ * dlltool.c (main): Don't use "boolean_condition ? TRUE : FALSE". * dwarf.c (read_and_display_attr_value): Likewise. (display_debug_str_offsets): Likewise. * objdump.c (dump_bfd): Likewise. * readelf.c (dump_section_as_strings): Likewise. (dump_section_as_bytes): Likewise. gas/ * atof-generic.c (FALSE, TRUE): Don't define. * config/obj-elf.h (FALSE, TRUE): Don't define. * config/obj-som.h (FALSE, TRUE): Don't define. * config/tc-hppa.h (FALSE, TRUE): Don't define. * config/tc-pdp11.c (FALSE, TRUE): Don't define. * config/tc-iq2000.h (obj_fix_adjustable): Delete. * config/tc-m32r.h (TC_FIX_ADJUSTABLE): Delete. * config/tc-mt.h (obj_fix_adjustable): Delete. * config/tc-nds32.h (TC_FIX_ADJUSTABLE): Delete. * config/tc-arc.c (parse_opcode_flags): Simplify boolean expression. (relaxable_flag, relaxable_operand, assemble_insn): Likewise. (tokenize_extregister): Likewise. * config/tc-csky.c (parse_opcode, get_operand_value): Likewise. (parse_operands_op, parse_operands, md_assemble): Likewise. * config/tc-d10v.c (build_insn): Likewise. * config/tc-score.c (s3_gen_insn_frag): Likewise. * config/tc-score7.c (s7_gen_insn_frag, s7_relax_frag): Likewise. * config/tc-tic6x.c (tic6x_update_features, md_assemble): Likewise. * config/tc-z80.c (emit_byte): Likewise. include/ * opcode/aarch64.h (alias_opcode_p): Simplify boolean expression. (opcode_has_alias, pseudo_opcode_p, optional_operand_p): Likewise. (opcode_has_special_coder): Likewise. ld/ * emultempl/aix.em (gld${EMULATION_NAME}_before_allocation): Simplify boolean expression. * lexsup.c (parse_args): Likewise. * pe-dll.c (pe_dll_id_target): Likewise. opcodes/ * aarch64-opc.c (vector_qualifier_p): Simplify boolean expression. (fp_qualifier_p, get_data_pattern): Likewise. (aarch64_get_operand_modifier_from_value): Likewise. (aarch64_extend_operator_p, aarch64_shift_operator_p): Likewise. (operand_variant_qualifier_p): Likewise. (qualifier_value_in_range_constraint_p): Likewise. (aarch64_get_qualifier_esize): Likewise. (aarch64_get_qualifier_nelem): Likewise. (aarch64_get_qualifier_standard_value): Likewise. (get_lower_bound, get_upper_bound): Likewise. (aarch64_find_best_match, match_operands_qualifier): Likewise. (aarch64_print_operand): Likewise. * aarch64-opc.h (operand_has_inserter, operand_has_extractor): Likewise. (operand_need_sign_extension, operand_need_shift_by_two): Likewise. (operand_need_shift_by_four, operand_maybe_stack_pointer): Likewise. * arm-dis.c (print_insn_mve, print_insn_thumb32): Likewise. * tic6x-dis.c (tic6x_check_fetch_packet_header): Likewise. (print_insn_tic6x): Likewise.
2021-03-29 07:22:56 +08:00
return kind > AARCH64_MOD_LSL && kind <= AARCH64_MOD_SXTX;
}
static inline bool
aarch64_shift_operator_p (enum aarch64_modifier_kind kind)
{
TRUE/FALSE simplification There is really no need to write code like "foo != 0 ? TRUE : FALSE" unless we had stupidly defined FALSE as something other than 0 or TRUE as something other than 1. The simpler "foo != 0" does just as well. Similarly "(condition == TRUE)" or "(condition == FALSE) can be simplified to "(condition)" and "(!condition)" respectively. I'll note that there is reason to use "integer_expression != 0" when assigning a bfd_boolean rather than the simpler "integer_expression", if you expect the variable to have 0 or 1 value. It's probably even a good idea to not rely on implicit conversion if bfd_boolean were _Bool. bfd/ * aoutx.h (aout_link_write_symbols): Don't cast boolean expression to bfd_boolean. * elf32-or1k.c (or1k_set_got_and_rela_sizes): Dont compare booleans against FALSE. * elf32-arc.c (name_for_global_symbol): Don't compare boolean to TRUE. (is_reloc_PC_relative): Don't use "boolean_condition ? TRUE : FALSE". (is_reloc_SDA_relative, is_reloc_for_GOT): Likewise. (is_reloc_for_PLT, is_reloc_for_TLS): Likewise. * elf32-arm.c (stm32l4xx_need_create_replacing_stub): Likewise. * elf32-nds32.c (insert_nds32_elf_blank): Likewise. * elf32-rx.c (rx_set_section_contents): Likewise. * elfnn-aarch64.c (elfNN_aarch64_final_link_relocate): Likewise. * elfxx-mips.c (_bfd_mips_elf_ignore_undef_symbol): Likewise. * mach-o.c (bfd_mach_o_read_command): Likewise. * targets.c (bfd_get_target_info): Likewise. binutils/ * dlltool.c (main): Don't use "boolean_condition ? TRUE : FALSE". * dwarf.c (read_and_display_attr_value): Likewise. (display_debug_str_offsets): Likewise. * objdump.c (dump_bfd): Likewise. * readelf.c (dump_section_as_strings): Likewise. (dump_section_as_bytes): Likewise. gas/ * atof-generic.c (FALSE, TRUE): Don't define. * config/obj-elf.h (FALSE, TRUE): Don't define. * config/obj-som.h (FALSE, TRUE): Don't define. * config/tc-hppa.h (FALSE, TRUE): Don't define. * config/tc-pdp11.c (FALSE, TRUE): Don't define. * config/tc-iq2000.h (obj_fix_adjustable): Delete. * config/tc-m32r.h (TC_FIX_ADJUSTABLE): Delete. * config/tc-mt.h (obj_fix_adjustable): Delete. * config/tc-nds32.h (TC_FIX_ADJUSTABLE): Delete. * config/tc-arc.c (parse_opcode_flags): Simplify boolean expression. (relaxable_flag, relaxable_operand, assemble_insn): Likewise. (tokenize_extregister): Likewise. * config/tc-csky.c (parse_opcode, get_operand_value): Likewise. (parse_operands_op, parse_operands, md_assemble): Likewise. * config/tc-d10v.c (build_insn): Likewise. * config/tc-score.c (s3_gen_insn_frag): Likewise. * config/tc-score7.c (s7_gen_insn_frag, s7_relax_frag): Likewise. * config/tc-tic6x.c (tic6x_update_features, md_assemble): Likewise. * config/tc-z80.c (emit_byte): Likewise. include/ * opcode/aarch64.h (alias_opcode_p): Simplify boolean expression. (opcode_has_alias, pseudo_opcode_p, optional_operand_p): Likewise. (opcode_has_special_coder): Likewise. ld/ * emultempl/aix.em (gld${EMULATION_NAME}_before_allocation): Simplify boolean expression. * lexsup.c (parse_args): Likewise. * pe-dll.c (pe_dll_id_target): Likewise. opcodes/ * aarch64-opc.c (vector_qualifier_p): Simplify boolean expression. (fp_qualifier_p, get_data_pattern): Likewise. (aarch64_get_operand_modifier_from_value): Likewise. (aarch64_extend_operator_p, aarch64_shift_operator_p): Likewise. (operand_variant_qualifier_p): Likewise. (qualifier_value_in_range_constraint_p): Likewise. (aarch64_get_qualifier_esize): Likewise. (aarch64_get_qualifier_nelem): Likewise. (aarch64_get_qualifier_standard_value): Likewise. (get_lower_bound, get_upper_bound): Likewise. (aarch64_find_best_match, match_operands_qualifier): Likewise. (aarch64_print_operand): Likewise. * aarch64-opc.h (operand_has_inserter, operand_has_extractor): Likewise. (operand_need_sign_extension, operand_need_shift_by_two): Likewise. (operand_need_shift_by_four, operand_maybe_stack_pointer): Likewise. * arm-dis.c (print_insn_mve, print_insn_thumb32): Likewise. * tic6x-dis.c (tic6x_check_fetch_packet_header): Likewise. (print_insn_tic6x): Likewise.
2021-03-29 07:22:56 +08:00
return kind >= AARCH64_MOD_ROR && kind <= AARCH64_MOD_LSL;
}
const struct aarch64_name_value_pair aarch64_barrier_options[16] =
{
{ "#0x00", 0x0 },
{ "oshld", 0x1 },
{ "oshst", 0x2 },
{ "osh", 0x3 },
{ "#0x04", 0x4 },
{ "nshld", 0x5 },
{ "nshst", 0x6 },
{ "nsh", 0x7 },
{ "#0x08", 0x8 },
{ "ishld", 0x9 },
{ "ishst", 0xa },
{ "ish", 0xb },
{ "#0x0c", 0xc },
{ "ld", 0xd },
{ "st", 0xe },
{ "sy", 0xf },
};
aarch64: Add DSB instruction Armv8.7-a variant This patch adds new variant (nXS) of DSB memory barrier instruction available in Armv8.7-a. New nXS variant has different encoding in comparison with pre Armv8.7-a DSB memory barrier variant thus new instruction and new operand was added. DSB memory nXS barrier variant specifies the limitation on the barrier operation. Allowed values are: DSB SYnXS|#28 DSB ISHnXS|#24 DSB NSHnXS|#20 DSB OSHnXS|#16 Please note that till now, for barriers, barrier operation was encoded in 4-bit unsigned immediate CRm field (in the range 0 to 15). For DSB memory nXS barrier variant, barrier operation is a 5-bit unsigned assembly instruction immediate, encoded in instruction in two bits CRm<3:2>: CRm<3:2> #imm 00 16 01 20 10 24 11 28 This patch extends current AArch64 barrier instructions with above mapping. Notable patch changes include: + New DSB memory barrier variant encoding for Armv8.7-a. + New operand BARRIER_DSB_NXS for above instruction in order to distinguish between existing and new DSB instruction flavour. + New set of DSB nXS barrier options. + New instruction inserter and extractor map between instruction immediate 5-bit value and 2-bit CRm field of the instruction itself (see FLD_CRm_dsb_nxs). + Regeneration of aarch64-[asm|dis|opc]-2.c files. + Test cases to cover new instruction assembling and disassembling. For more details regarding DSB memory barrier instruction and its Armv8.7-a flavour please refer to Arm A64 Instruction set documentation for Armv8-A architecture profile, see document pages 132-133 of [0]. [0]: https://developer.arm.com/docs/ddi0596/i gas/ChangeLog: 2020-10-23 Przemyslaw Wirkus <przemyslaw.wirkus@arm.com> * NEWS: Docs update. * config/tc-aarch64.c (parse_operands): Add AARCH64_OPND_BARRIER_DSB_NXS handler. (md_begin): Add content of aarch64_barrier_dsb_nxs_options to aarch64_barrier_opt_hsh hash. * testsuite/gas/aarch64/system-4-invalid.d: New test. * testsuite/gas/aarch64/system-4-invalid.l: New test. * testsuite/gas/aarch64/system-4-invalid.s: New test. * testsuite/gas/aarch64/system-4.d: New test. * testsuite/gas/aarch64/system-4.s: New test. include/ChangeLog: 2020-10-23 Przemyslaw Wirkus <przemyslaw.wirkus@arm.com> * opcode/aarch64.h (enum aarch64_opnd): New operand AARCH64_OPND_BARRIER_DSB_NXS. (aarch64_barrier_dsb_nxs_options): Declare DSB nXS options. opcodes/ChangeLog: 2020-10-23 Przemyslaw Wirkus <przemyslaw.wirkus@arm.com> * aarch64-asm.c (aarch64_ins_barrier_dsb_nxs): New inserter. * aarch64-asm.h (AARCH64_DECL_OPD_INSERTER): New inserter ins_barrier_dsb_nx. * aarch64-dis.c (aarch64_ext_barrier_dsb_nxs): New extractor. * aarch64-dis.h (AARCH64_DECL_OPD_EXTRACTOR): New extractor ext_barrier_dsb_nx. * aarch64-opc.c (aarch64_print_operand): New options table aarch64_barrier_dsb_nxs_options. * aarch64-opc.h (enum aarch64_field_kind): New field name FLD_CRm_dsb_nxs. * aarch64-tbl.h (struct aarch64_opcode): Define DSB nXS barrier Armv8.7-a instruction. * aarch64-asm-2.c: Regenerated. * aarch64-dis-2.c: Regenerated. * aarch64-opc-2.c: Regenerated.
2020-10-28 22:01:36 +08:00
const struct aarch64_name_value_pair aarch64_barrier_dsb_nxs_options[4] =
{ /* CRm<3:2> #imm */
{ "oshnxs", 16 }, /* 00 16 */
{ "nshnxs", 20 }, /* 01 20 */
{ "ishnxs", 24 }, /* 10 24 */
{ "synxs", 28 }, /* 11 28 */
};
/* Table describing the operands supported by the aliases of the HINT
instruction.
The name column is the operand that is accepted for the alias. The value
column is the hint number of the alias. The list of operands is terminated
by NULL in the name column. */
const struct aarch64_name_value_pair aarch64_hint_options[] =
{
[PATCH, BINUTILS, AARCH64, 7/9] Add BTI instruction This patch is part of the patch series to add support for ARMv8.5-A extensions. (https://developer.arm.com/products/architecture/cpu-architecture/a-profile/docs/ddi0596/a/a64-base-instructions-alphabetic-order/bti-branch-target-identification) The Branch Target Identification instructions (BTI) are allocated to existing HINT space, using HINT numbers 32, 34, 36, 38, such that bits[7:6] of the instruction identify the compatibility of the BTI instruction to different branches. BTI {<targets>} where <targets> one of the following, specifying which type of indirection is allowed: j : Can be a target of any BR Xn isntruction. c : Can be a target of any BLR Xn and BR {X16|X17}. jc: Can be a target of any free branch. A BTI instruction without any <targets> is the strictest of all and can not be a target of nay free branch. *** include/ChangeLog *** 2018-10-09 Sudakshina Das <sudi.das@arm.com> * opcode/aarch64.h (AARCH64_FEATURE_BTI): New. (AARCH64_ARCH_V8_5): Add AARCH64_FEATURE_BTI by default. (aarch64_opnd): Add AARCH64_OPND_BTI_TARGET. (HINT_OPD_CSYNC, HINT_OPD_C, HINT_OPD_J): New macros to define HINT #imm values. (HINT_OPD_JC, HINT_OPD_NULL): Likewise. *** opcodes/ChangeLog *** 2018-10-09 Sudakshina Das <sudi.das@arm.com> * aarch64-opc.h (HINT_OPD_NOPRINT, HINT_ENCODE): New. (HINT_FLAG, HINT_VALUE): New macros to encode NO_PRINT flag with the hint immediate. * aarch64-opc.c (aarch64_hint_options): New entries for c, j, jc and default (with HINT_OPD_F_NOPRINT flag) for BTI. (aarch64_print_operand): Add case for AARCH64_OPND_BTI_TARGET while checking for HINT_OPD_F_NOPRINT flag. * aarch64-dis.c (aarch64_ext_hint): Use new HINT_VALUE to extract value. * aarch64-tbl.h (aarch64_feature_bti, BTI, BTI_INSN): New. (aarch64_opcode_table): Add entry for BTI. (AARCH64_OPERANDS): Add new description for BTI targets. * aarch64-asm-2.c: Regenerate. * aarch64-dis-2.c: Regenerate. * aarch64-opc-2.c: Regenerate. *** gas/ChangeLog *** 2018-10-09 Sudakshina Das <sudi.das@arm.com> * config/tc-aarch64.c (parse_bti_operand): New. (process_omitted_operand): Add case for AARCH64_OPND_BTI_TARGET. (parse_operands): Likewise. * testsuite/gas/aarch64/system.d: Update for BTI. * testsuite/gas/aarch64/bti.s: New. * testsuite/gas/aarch64/bti.d: New. * testsuite/gas/aarch64/illegal-bti.d: New. * testsuite/gas/aarch64/illegal-bti.l: New.
2018-09-26 18:00:49 +08:00
/* BTI. This is also the F_DEFAULT entry for AARCH64_OPND_BTI_TARGET. */
{ " ", HINT_ENCODE (HINT_OPD_F_NOPRINT, 0x20) },
{ "csync", HINT_OPD_CSYNC }, /* PSB CSYNC. */
{ "dsync", HINT_OPD_DSYNC }, /* GCSB DSYNC. */
[PATCH, BINUTILS, AARCH64, 7/9] Add BTI instruction This patch is part of the patch series to add support for ARMv8.5-A extensions. (https://developer.arm.com/products/architecture/cpu-architecture/a-profile/docs/ddi0596/a/a64-base-instructions-alphabetic-order/bti-branch-target-identification) The Branch Target Identification instructions (BTI) are allocated to existing HINT space, using HINT numbers 32, 34, 36, 38, such that bits[7:6] of the instruction identify the compatibility of the BTI instruction to different branches. BTI {<targets>} where <targets> one of the following, specifying which type of indirection is allowed: j : Can be a target of any BR Xn isntruction. c : Can be a target of any BLR Xn and BR {X16|X17}. jc: Can be a target of any free branch. A BTI instruction without any <targets> is the strictest of all and can not be a target of nay free branch. *** include/ChangeLog *** 2018-10-09 Sudakshina Das <sudi.das@arm.com> * opcode/aarch64.h (AARCH64_FEATURE_BTI): New. (AARCH64_ARCH_V8_5): Add AARCH64_FEATURE_BTI by default. (aarch64_opnd): Add AARCH64_OPND_BTI_TARGET. (HINT_OPD_CSYNC, HINT_OPD_C, HINT_OPD_J): New macros to define HINT #imm values. (HINT_OPD_JC, HINT_OPD_NULL): Likewise. *** opcodes/ChangeLog *** 2018-10-09 Sudakshina Das <sudi.das@arm.com> * aarch64-opc.h (HINT_OPD_NOPRINT, HINT_ENCODE): New. (HINT_FLAG, HINT_VALUE): New macros to encode NO_PRINT flag with the hint immediate. * aarch64-opc.c (aarch64_hint_options): New entries for c, j, jc and default (with HINT_OPD_F_NOPRINT flag) for BTI. (aarch64_print_operand): Add case for AARCH64_OPND_BTI_TARGET while checking for HINT_OPD_F_NOPRINT flag. * aarch64-dis.c (aarch64_ext_hint): Use new HINT_VALUE to extract value. * aarch64-tbl.h (aarch64_feature_bti, BTI, BTI_INSN): New. (aarch64_opcode_table): Add entry for BTI. (AARCH64_OPERANDS): Add new description for BTI targets. * aarch64-asm-2.c: Regenerate. * aarch64-dis-2.c: Regenerate. * aarch64-opc-2.c: Regenerate. *** gas/ChangeLog *** 2018-10-09 Sudakshina Das <sudi.das@arm.com> * config/tc-aarch64.c (parse_bti_operand): New. (process_omitted_operand): Add case for AARCH64_OPND_BTI_TARGET. (parse_operands): Likewise. * testsuite/gas/aarch64/system.d: Update for BTI. * testsuite/gas/aarch64/bti.s: New. * testsuite/gas/aarch64/bti.d: New. * testsuite/gas/aarch64/illegal-bti.d: New. * testsuite/gas/aarch64/illegal-bti.l: New.
2018-09-26 18:00:49 +08:00
{ "c", HINT_OPD_C }, /* BTI C. */
{ "j", HINT_OPD_J }, /* BTI J. */
{ "jc", HINT_OPD_JC }, /* BTI JC. */
{ NULL, HINT_OPD_NULL },
};
/* op -> op: load = 0 instruction = 1 store = 2
l -> level: 1-3
t -> temporal: temporal (retained) = 0 non-temporal (streaming) = 1 */
#define B(op,l,t) (((op) << 3) | (((l) - 1) << 1) | (t))
const struct aarch64_name_value_pair aarch64_prfops[32] =
{
{ "pldl1keep", B(0, 1, 0) },
{ "pldl1strm", B(0, 1, 1) },
{ "pldl2keep", B(0, 2, 0) },
{ "pldl2strm", B(0, 2, 1) },
{ "pldl3keep", B(0, 3, 0) },
{ "pldl3strm", B(0, 3, 1) },
{ "pldslckeep", B(0, 4, 0) },
{ "pldslcstrm", B(0, 4, 1) },
{ "plil1keep", B(1, 1, 0) },
{ "plil1strm", B(1, 1, 1) },
{ "plil2keep", B(1, 2, 0) },
{ "plil2strm", B(1, 2, 1) },
{ "plil3keep", B(1, 3, 0) },
{ "plil3strm", B(1, 3, 1) },
{ "plislckeep", B(1, 4, 0) },
{ "plislcstrm", B(1, 4, 1) },
{ "pstl1keep", B(2, 1, 0) },
{ "pstl1strm", B(2, 1, 1) },
{ "pstl2keep", B(2, 2, 0) },
{ "pstl2strm", B(2, 2, 1) },
{ "pstl3keep", B(2, 3, 0) },
{ "pstl3strm", B(2, 3, 1) },
{ "pstslckeep", B(2, 4, 0) },
{ "pstslcstrm", B(2, 4, 1) },
{ NULL, 0x18 },
{ NULL, 0x19 },
{ NULL, 0x1a },
{ NULL, 0x1b },
{ NULL, 0x1c },
{ NULL, 0x1d },
{ NULL, 0x1e },
{ NULL, 0x1f },
};
#undef B
/* Utilities on value constraint. */
static inline int
value_in_range_p (int64_t value, int low, int high)
{
return (value >= low && value <= high) ? 1 : 0;
}
[AArch64][SVE 26/32] Add SVE MUL VL addressing modes This patch adds support for addresses of the form: [<base>, #<offset>, MUL VL] This involves adding a new AARCH64_MOD_MUL_VL modifier, which is why I split it out from the other addressing modes. For LD2, LD3 and LD4, the offset must be a multiple of the structure size, so for LD3 the possible values are 0, 3, 6, .... The patch therefore extends value_aligned_p to handle non-power-of-2 alignments. include/ * opcode/aarch64.h (AARCH64_OPND_SVE_ADDR_RI_S4xVL): New aarch64_opnd. (AARCH64_OPND_SVE_ADDR_RI_S4x2xVL, AARCH64_OPND_SVE_ADDR_RI_S4x3xVL) (AARCH64_OPND_SVE_ADDR_RI_S4x4xVL, AARCH64_OPND_SVE_ADDR_RI_S6xVL) (AARCH64_OPND_SVE_ADDR_RI_S9xVL): Likewise. (AARCH64_MOD_MUL_VL): New aarch64_modifier_kind. opcodes/ * aarch64-tbl.h (AARCH64_OPERANDS): Add entries for new MUL VL operands. * aarch64-opc.c (aarch64_operand_modifiers): Initialize the AARCH64_MOD_MUL_VL entry. (value_aligned_p): Cope with non-power-of-two alignments. (operand_general_constraint_met_p): Handle the new MUL VL addresses. (print_immediate_offset_address): Likewise. (aarch64_print_operand): Likewise. * aarch64-opc-2.c: Regenerate. * aarch64-asm.h (ins_sve_addr_ri_s4xvl, ins_sve_addr_ri_s6xvl) (ins_sve_addr_ri_s9xvl): New inserters. * aarch64-asm.c (aarch64_ins_sve_addr_ri_s4xvl): New function. (aarch64_ins_sve_addr_ri_s6xvl): Likewise. (aarch64_ins_sve_addr_ri_s9xvl): Likewise. * aarch64-asm-2.c: Regenerate. * aarch64-dis.h (ext_sve_addr_ri_s4xvl, ext_sve_addr_ri_s6xvl) (ext_sve_addr_ri_s9xvl): New extractors. * aarch64-dis.c (aarch64_ext_sve_addr_reg_mul_vl): New function. (aarch64_ext_sve_addr_ri_s4xvl): Likewise. (aarch64_ext_sve_addr_ri_s6xvl): Likewise. (aarch64_ext_sve_addr_ri_s9xvl): Likewise. * aarch64-dis-2.c: Regenerate. gas/ * config/tc-aarch64.c (SHIFTED_NONE, SHIFTED_MUL_VL): New parse_shift_modes. (parse_shift): Handle SHIFTED_MUL_VL. (parse_address_main): Add an imm_shift_mode parameter. (parse_address, parse_sve_address): Update accordingly. (parse_operands): Handle MUL VL addressing modes.
2016-09-21 23:56:15 +08:00
/* Return true if VALUE is a multiple of ALIGN. */
static inline int
value_aligned_p (int64_t value, int align)
{
[AArch64][SVE 26/32] Add SVE MUL VL addressing modes This patch adds support for addresses of the form: [<base>, #<offset>, MUL VL] This involves adding a new AARCH64_MOD_MUL_VL modifier, which is why I split it out from the other addressing modes. For LD2, LD3 and LD4, the offset must be a multiple of the structure size, so for LD3 the possible values are 0, 3, 6, .... The patch therefore extends value_aligned_p to handle non-power-of-2 alignments. include/ * opcode/aarch64.h (AARCH64_OPND_SVE_ADDR_RI_S4xVL): New aarch64_opnd. (AARCH64_OPND_SVE_ADDR_RI_S4x2xVL, AARCH64_OPND_SVE_ADDR_RI_S4x3xVL) (AARCH64_OPND_SVE_ADDR_RI_S4x4xVL, AARCH64_OPND_SVE_ADDR_RI_S6xVL) (AARCH64_OPND_SVE_ADDR_RI_S9xVL): Likewise. (AARCH64_MOD_MUL_VL): New aarch64_modifier_kind. opcodes/ * aarch64-tbl.h (AARCH64_OPERANDS): Add entries for new MUL VL operands. * aarch64-opc.c (aarch64_operand_modifiers): Initialize the AARCH64_MOD_MUL_VL entry. (value_aligned_p): Cope with non-power-of-two alignments. (operand_general_constraint_met_p): Handle the new MUL VL addresses. (print_immediate_offset_address): Likewise. (aarch64_print_operand): Likewise. * aarch64-opc-2.c: Regenerate. * aarch64-asm.h (ins_sve_addr_ri_s4xvl, ins_sve_addr_ri_s6xvl) (ins_sve_addr_ri_s9xvl): New inserters. * aarch64-asm.c (aarch64_ins_sve_addr_ri_s4xvl): New function. (aarch64_ins_sve_addr_ri_s6xvl): Likewise. (aarch64_ins_sve_addr_ri_s9xvl): Likewise. * aarch64-asm-2.c: Regenerate. * aarch64-dis.h (ext_sve_addr_ri_s4xvl, ext_sve_addr_ri_s6xvl) (ext_sve_addr_ri_s9xvl): New extractors. * aarch64-dis.c (aarch64_ext_sve_addr_reg_mul_vl): New function. (aarch64_ext_sve_addr_ri_s4xvl): Likewise. (aarch64_ext_sve_addr_ri_s6xvl): Likewise. (aarch64_ext_sve_addr_ri_s9xvl): Likewise. * aarch64-dis-2.c: Regenerate. gas/ * config/tc-aarch64.c (SHIFTED_NONE, SHIFTED_MUL_VL): New parse_shift_modes. (parse_shift): Handle SHIFTED_MUL_VL. (parse_address_main): Add an imm_shift_mode parameter. (parse_address, parse_sve_address): Update accordingly. (parse_operands): Handle MUL VL addressing modes.
2016-09-21 23:56:15 +08:00
return (value % align) == 0;
}
/* A signed value fits in a field. */
static inline int
value_fit_signed_field_p (int64_t value, unsigned width)
{
assert (width < 32);
if (width < sizeof (value) * 8)
{
int64_t lim = (uint64_t) 1 << (width - 1);
if (value >= -lim && value < lim)
return 1;
}
return 0;
}
/* An unsigned value fits in a field. */
static inline int
value_fit_unsigned_field_p (int64_t value, unsigned width)
{
assert (width < 32);
if (width < sizeof (value) * 8)
{
int64_t lim = (uint64_t) 1 << width;
if (value >= 0 && value < lim)
return 1;
}
return 0;
}
/* Return 1 if OPERAND is SP or WSP. */
int
aarch64_stack_pointer_p (const aarch64_opnd_info *operand)
{
return ((aarch64_get_operand_class (operand->type)
== AARCH64_OPND_CLASS_INT_REG)
&& operand_maybe_stack_pointer (aarch64_operands + operand->type)
&& operand->reg.regno == 31);
}
/* Return 1 if OPERAND is XZR or WZP. */
int
aarch64_zero_register_p (const aarch64_opnd_info *operand)
{
return ((aarch64_get_operand_class (operand->type)
== AARCH64_OPND_CLASS_INT_REG)
&& !operand_maybe_stack_pointer (aarch64_operands + operand->type)
&& operand->reg.regno == 31);
}
/* Return true if the operand *OPERAND that has the operand code
OPERAND->TYPE and been qualified by OPERAND->QUALIFIER can be also
qualified by the qualifier TARGET. */
static inline int
operand_also_qualified_p (const struct aarch64_opnd_info *operand,
aarch64_opnd_qualifier_t target)
{
switch (operand->qualifier)
{
case AARCH64_OPND_QLF_W:
if (target == AARCH64_OPND_QLF_WSP && aarch64_stack_pointer_p (operand))
return 1;
break;
case AARCH64_OPND_QLF_X:
if (target == AARCH64_OPND_QLF_SP && aarch64_stack_pointer_p (operand))
return 1;
break;
case AARCH64_OPND_QLF_WSP:
if (target == AARCH64_OPND_QLF_W
&& operand_maybe_stack_pointer (aarch64_operands + operand->type))
return 1;
break;
case AARCH64_OPND_QLF_SP:
if (target == AARCH64_OPND_QLF_X
&& operand_maybe_stack_pointer (aarch64_operands + operand->type))
return 1;
break;
default:
break;
}
return 0;
}
/* Given qualifier sequence list QSEQ_LIST and the known qualifier KNOWN_QLF
for operand KNOWN_IDX, return the expected qualifier for operand IDX.
Return NIL if more than one expected qualifiers are found. */
aarch64_opnd_qualifier_t
aarch64_get_expected_qualifier (const aarch64_opnd_qualifier_seq_t *qseq_list,
int idx,
const aarch64_opnd_qualifier_t known_qlf,
int known_idx)
{
int i, saved_i;
/* Special case.
When the known qualifier is NIL, we have to assume that there is only
one qualifier sequence in the *QSEQ_LIST and return the corresponding
qualifier directly. One scenario is that for instruction
PRFM <prfop>, [<Xn|SP>, #:lo12:<symbol>]
which has only one possible valid qualifier sequence
NIL, S_D
the caller may pass NIL in KNOWN_QLF to obtain S_D so that it can
determine the correct relocation type (i.e. LDST64_LO12) for PRFM.
Because the qualifier NIL has dual roles in the qualifier sequence:
it can mean no qualifier for the operand, or the qualifer sequence is
not in use (when all qualifiers in the sequence are NILs), we have to
handle this special case here. */
if (known_qlf == AARCH64_OPND_NIL)
{
assert (qseq_list[0][known_idx] == AARCH64_OPND_NIL);
return qseq_list[0][idx];
}
for (i = 0, saved_i = -1; i < AARCH64_MAX_QLF_SEQ_NUM; ++i)
{
if (qseq_list[i][known_idx] == known_qlf)
{
if (saved_i != -1)
/* More than one sequences are found to have KNOWN_QLF at
KNOWN_IDX. */
return AARCH64_OPND_NIL;
saved_i = i;
}
}
return qseq_list[saved_i][idx];
}
enum operand_qualifier_kind
{
OQK_NIL,
OQK_OPD_VARIANT,
OQK_VALUE_IN_RANGE,
OQK_MISC,
};
/* Operand qualifier description. */
struct operand_qualifier_data
{
/* The usage of the three data fields depends on the qualifier kind. */
int data0;
int data1;
int data2;
/* Description. */
const char *desc;
/* Kind. */
enum operand_qualifier_kind kind;
};
/* Indexed by the operand qualifier enumerators. */
struct operand_qualifier_data aarch64_opnd_qualifiers[] =
{
{0, 0, 0, "NIL", OQK_NIL},
/* Operand variant qualifiers.
First 3 fields:
element size, number of elements and common value for encoding. */
{4, 1, 0x0, "w", OQK_OPD_VARIANT},
{8, 1, 0x1, "x", OQK_OPD_VARIANT},
{4, 1, 0x0, "wsp", OQK_OPD_VARIANT},
{8, 1, 0x1, "sp", OQK_OPD_VARIANT},
{1, 1, 0x0, "b", OQK_OPD_VARIANT},
{2, 1, 0x1, "h", OQK_OPD_VARIANT},
{4, 1, 0x2, "s", OQK_OPD_VARIANT},
{8, 1, 0x3, "d", OQK_OPD_VARIANT},
{16, 1, 0x4, "q", OQK_OPD_VARIANT},
AArch64: Fix error checking for SIMD udot (by element) Committed on behalf of Matthew Malcomson: The SIMD UDOT instruction assembly has an unusual operand that selects a single 32 bit element with the mnemonic 4B. This unusual mnemonic is handled by a special operand qualifier and associated qualifier data in `aarch64_opnd_qualifiers`. The current qualifier data describes 4 1-byte elements with the structure {1, 4, 0x0, "4b", OQK_OPD_VARIANT} This makes sense, as the instruction does work on 4 1-byte elements, however some logic in the `operand_general_constraint_met_p` makes assumptions about the range of index allowed when selecting a SIMD_ELEMENT depending on element size. That function reasons that e.g. in order to select a byte-sized element in a 16 byte V register an index must allow selection of one of the 16 elements and hence its range will be in [0,15]. This reasoning breaks with the above description of a 4 part selection of 1 byte elements and allows an index outside the valid [0,3] range, triggering an assert later on in the program in `aarch64_ins_reglane`. vshcmd: > echo 'udot v0.2s, v1.8b, v2.4b[4]' | ../src/binutils-build/gas/as-new -march=armv8.4-a as-new: ../../binutils-gdb/opcodes/aarch64-asm.c:134: aarch64_ins_reglane: Assertion `reglane_index < 4' failed. {standard input}: Assembler messages: {standard input}:1: Internal error (Aborted). Please report this bug. This patch changes the operand qualifier data so that it describes a single 32 bit element. {4, 1, 0x0, "4b", OQK_OPD_VARIANT} Hence the calculation in `operand_general_constraint_met_p` provides the correct answer and the usual error checking machinery is used. vshcmd: > echo 'udot v0.2s, v1.8b, v2.4b[4]' | ../src/binutils-build/gas/as-new -march=armv8.4-a {standard input}: Assembler messages: {standard input}:1: Error: register element index out of range 0 to 3 at operand 3 -- `udot v0.2s,v1.8b,v2.4b[4]'
2018-10-17 01:49:36 +08:00
{4, 1, 0x0, "4b", OQK_OPD_VARIANT},
[binutils][aarch64] Bfloat16 enablement [2/X] Hi, This patch is part of a series that adds support for Armv8.6-A (Matrix Multiply and BFloat16 extensions) to binutils. This patch introduces the following BFloat16 instructions to the aarch64 backend: bfdot, bfmmla, bfcvt, bfcvtnt, bfmlal[t/b], bfcvtn2. Committed on behalf of Mihail Ionescu. gas/ChangeLog: 2019-11-07 Mihail Ionescu <mihail.ionescu@arm.com> 2019-11-07 Matthew Malcomson <matthew.malcomson@arm.com> * config/tc-aarch64.c (vectype_to_qualifier): Special case the S_2H operand qualifier. * doc/c-aarch64.texi: Document bf16 and bf16mmla4 extensions. * testsuite/gas/aarch64/bfloat16.d: New test. * testsuite/gas/aarch64/bfloat16.s: New test. * testsuite/gas/aarch64/illegal-bfloat16.d: New test. * testsuite/gas/aarch64/illegal-bfloat16.l: New test. * testsuite/gas/aarch64/illegal-bfloat16.s: New test. * testsuite/gas/aarch64/sve-bfloat-movprfx.s: New test. * testsuite/gas/aarch64/sve-bfloat-movprfx.d: New test. include/ChangeLog: 2019-11-07 Mihail Ionescu <mihail.ionescu@arm.com> 2019-11-07 Matthew Malcomson <matthew.malcomson@arm.com> * opcode/aarch64.h (AARCH64_FEATURE_BFLOAT16): New feature macros. (AARCH64_ARCH_V8_6): Include BFloat16 feature macros. (enum aarch64_opnd_qualifier): Introduce new operand qualifier AARCH64_OPND_QLF_S_2H. (enum aarch64_insn_class): Introduce new class "bfloat16". (BFLOAT16_SVE_INSNC): New feature set for bfloat16 instructions to support the movprfx constraint. opcodes/ChangeLog: 2019-11-07 Mihail Ionescu <mihail.ionescu@arm.com> 2019-11-07 Matthew Malcomson <matthew.malcomson@arm.com> * aarch64-asm.c (aarch64_ins_reglane): Use AARCH64_OPND_QLF_S_2H in reglane special case. * aarch64-dis-2.c (aarch64_opcode_lookup_1, aarch64_find_next_opcode): Account for new instructions. * aarch64-dis.c (aarch64_ext_reglane): Use AARCH64_OPND_QLF_S_2H in reglane special case. * aarch64-opc.c (struct operand_qualifier_data): Add data for new AARCH64_OPND_QLF_S_2H qualifier. * aarch64-tbl.h (QL_BFDOT QL_BFDOT64, QL_BFDOT64I, QL_BFMMLA2, QL_BFCVT64, QL_BFCVTN64, QL_BFCVTN2_64): New qualifiers. (aarch64_feature_bfloat16, aarch64_feature_bfloat16_sve, aarch64_feature_bfloat16_bfmmla4): New feature sets. (BFLOAT_SVE, BFLOAT): New feature set macros. (BFLOAT_SVE_INSN, BFLOAT_BFMMLA4_INSN, BFLOAT_INSN): New macros to define BFloat16 instructions. (aarch64_opcode_table): Define new instructions bfdot, bfmmla, bfcvt, bfcvtnt, bfdot, bfdot, bfcvtn, bfmlal[b/t] bfcvtn2, bfcvt. Regression tested on aarch64-elf. Is it ok for trunk? Regards, Mihail
2019-11-08 00:38:59 +08:00
{4, 1, 0x0, "2h", OQK_OPD_VARIANT},
{1, 4, 0x0, "4b", OQK_OPD_VARIANT},
{1, 8, 0x0, "8b", OQK_OPD_VARIANT},
{1, 16, 0x1, "16b", OQK_OPD_VARIANT},
{2, 2, 0x0, "2h", OQK_OPD_VARIANT},
{2, 4, 0x2, "4h", OQK_OPD_VARIANT},
{2, 8, 0x3, "8h", OQK_OPD_VARIANT},
{4, 2, 0x4, "2s", OQK_OPD_VARIANT},
{4, 4, 0x5, "4s", OQK_OPD_VARIANT},
{8, 1, 0x6, "1d", OQK_OPD_VARIANT},
{8, 2, 0x7, "2d", OQK_OPD_VARIANT},
{16, 1, 0x8, "1q", OQK_OPD_VARIANT},
{0, 0, 0, "z", OQK_OPD_VARIANT},
{0, 0, 0, "m", OQK_OPD_VARIANT},
[BINUTILS, AARCH64, 4/8] Add Tag setting instructions in Memory Tagging Extension This patch is part of the patch series to add support for ARMv8.5-A Memory Tagging Extensions which is an optional extension to ARMv8.5-A and is enabled using the +memtag command line option. This patch add support to the Tag setting instructions from MTE which consists of the following instructions: - STG [<Xn|SP>, #<simm>] - STG [<Xn|SP>, #<simm>]! - STG [<Xn|SP>], #<simm> - STZG [<Xn|SP>, #<simm>] - STZG [<Xn|SP>, #<simm>]! - STZG [<Xn|SP>], #<simm> - ST2G [<Xn|SP>, #<simm>] - ST2G [<Xn|SP>, #<simm>]! - ST2G [<Xn|SP>], #<simm> - STZ2G [<Xn|SP>, #<simm>] - STZ2G [<Xn|SP>, #<simm>]! - STZ2G [<Xn|SP>], #<simm> - STGP <Xt>, <Xt2>, [<Xn|SP>, #<imm>] - STGP <Xt>, <Xt2>, [<Xn|SP>, #<imm>]! - STGP <Xt>, <Xt2>, [<Xn|SP>], #<imm> where <Xn|SP> : Is the 64-bit GPR or Stack pointer. <simm> : Is the optional signed immediate offset, a multiple of 16 in the range -4096 to 4080, defaulting to 0. *** include/ChangeLog *** 2018-11-12 Sudakshina Das <sudi.das@arm.com> * opcode/aarch64.h (aarch64_opnd): Add AARCH64_OPND_ADDR_SIMM11 and AARCH64_OPND_ADDR_SIMM13. (aarch64_opnd_qualifier): Add new AARCH64_OPND_QLF_imm_tag. *** opcodes/ChangeLog *** 2018-11-12 Sudakshina Das <sudi.das@arm.com> * aarch64-opc.c (aarch64_opnd_qualifiers): Add new data for AARCH64_OPND_QLF_imm_tag. (operand_general_constraint_met_p): Add case for AARCH64_OPND_ADDR_SIMM11 and AARCH64_OPND_ADDR_SIMM13. (aarch64_print_operand): Likewise. * aarch64-tbl.h (QL_LDST_AT, QL_STGP): New. (aarch64_opcode_table): Add stg, stzg, st2g, stz2g and stgp for both offset and pre/post indexed versions. (AARCH64_OPERANDS): Define ADDR_SIMM11 and ADDR_SIMM13. * aarch64-asm-2.c: Regenerated. * aarch64-dis-2.c: Regenerated. * aarch64-opc-2.c: Regenerated. *** gas/ChangeLog *** 2018-11-12 Sudakshina Das <sudi.das@arm.com> * config/tc-aarch64.c (parse_operands): Add switch case for AARCH64_OPND_ADDR_SIMM11 and AARCH64_OPND_ADDR_SIMM13. (fix_insn): Likewise. (warn_unpredictable_ldst): Exempt STGP. * testsuite/gas/aarch64/armv8_5-a-memtag.s: Add tests for stg, st2g, stzg, stz2g and stgp. * testsuite/gas/aarch64/armv8_5-a-memtag.d: Likewise. * testsuite/gas/aarch64/illegal-memtag.s: Likewise. * testsuite/gas/aarch64/illegal-memtag.l: Likewise.
2018-11-12 21:09:55 +08:00
/* Qualifier for scaled immediate for Tag granule (stg,st2g,etc). */
{16, 0, 0, "tag", OQK_OPD_VARIANT},
/* Qualifiers constraining the value range.
First 3 fields:
Lower bound, higher bound, unused. */
{0, 15, 0, "CR", OQK_VALUE_IN_RANGE},
{0, 7, 0, "imm_0_7" , OQK_VALUE_IN_RANGE},
{0, 15, 0, "imm_0_15", OQK_VALUE_IN_RANGE},
{0, 31, 0, "imm_0_31", OQK_VALUE_IN_RANGE},
{0, 63, 0, "imm_0_63", OQK_VALUE_IN_RANGE},
{1, 32, 0, "imm_1_32", OQK_VALUE_IN_RANGE},
{1, 64, 0, "imm_1_64", OQK_VALUE_IN_RANGE},
/* Qualifiers for miscellaneous purpose.
First 3 fields:
unused, unused and unused. */
{0, 0, 0, "lsl", 0},
{0, 0, 0, "msl", 0},
{0, 0, 0, "retrieving", 0},
};
static inline bool
operand_variant_qualifier_p (aarch64_opnd_qualifier_t qualifier)
{
TRUE/FALSE simplification There is really no need to write code like "foo != 0 ? TRUE : FALSE" unless we had stupidly defined FALSE as something other than 0 or TRUE as something other than 1. The simpler "foo != 0" does just as well. Similarly "(condition == TRUE)" or "(condition == FALSE) can be simplified to "(condition)" and "(!condition)" respectively. I'll note that there is reason to use "integer_expression != 0" when assigning a bfd_boolean rather than the simpler "integer_expression", if you expect the variable to have 0 or 1 value. It's probably even a good idea to not rely on implicit conversion if bfd_boolean were _Bool. bfd/ * aoutx.h (aout_link_write_symbols): Don't cast boolean expression to bfd_boolean. * elf32-or1k.c (or1k_set_got_and_rela_sizes): Dont compare booleans against FALSE. * elf32-arc.c (name_for_global_symbol): Don't compare boolean to TRUE. (is_reloc_PC_relative): Don't use "boolean_condition ? TRUE : FALSE". (is_reloc_SDA_relative, is_reloc_for_GOT): Likewise. (is_reloc_for_PLT, is_reloc_for_TLS): Likewise. * elf32-arm.c (stm32l4xx_need_create_replacing_stub): Likewise. * elf32-nds32.c (insert_nds32_elf_blank): Likewise. * elf32-rx.c (rx_set_section_contents): Likewise. * elfnn-aarch64.c (elfNN_aarch64_final_link_relocate): Likewise. * elfxx-mips.c (_bfd_mips_elf_ignore_undef_symbol): Likewise. * mach-o.c (bfd_mach_o_read_command): Likewise. * targets.c (bfd_get_target_info): Likewise. binutils/ * dlltool.c (main): Don't use "boolean_condition ? TRUE : FALSE". * dwarf.c (read_and_display_attr_value): Likewise. (display_debug_str_offsets): Likewise. * objdump.c (dump_bfd): Likewise. * readelf.c (dump_section_as_strings): Likewise. (dump_section_as_bytes): Likewise. gas/ * atof-generic.c (FALSE, TRUE): Don't define. * config/obj-elf.h (FALSE, TRUE): Don't define. * config/obj-som.h (FALSE, TRUE): Don't define. * config/tc-hppa.h (FALSE, TRUE): Don't define. * config/tc-pdp11.c (FALSE, TRUE): Don't define. * config/tc-iq2000.h (obj_fix_adjustable): Delete. * config/tc-m32r.h (TC_FIX_ADJUSTABLE): Delete. * config/tc-mt.h (obj_fix_adjustable): Delete. * config/tc-nds32.h (TC_FIX_ADJUSTABLE): Delete. * config/tc-arc.c (parse_opcode_flags): Simplify boolean expression. (relaxable_flag, relaxable_operand, assemble_insn): Likewise. (tokenize_extregister): Likewise. * config/tc-csky.c (parse_opcode, get_operand_value): Likewise. (parse_operands_op, parse_operands, md_assemble): Likewise. * config/tc-d10v.c (build_insn): Likewise. * config/tc-score.c (s3_gen_insn_frag): Likewise. * config/tc-score7.c (s7_gen_insn_frag, s7_relax_frag): Likewise. * config/tc-tic6x.c (tic6x_update_features, md_assemble): Likewise. * config/tc-z80.c (emit_byte): Likewise. include/ * opcode/aarch64.h (alias_opcode_p): Simplify boolean expression. (opcode_has_alias, pseudo_opcode_p, optional_operand_p): Likewise. (opcode_has_special_coder): Likewise. ld/ * emultempl/aix.em (gld${EMULATION_NAME}_before_allocation): Simplify boolean expression. * lexsup.c (parse_args): Likewise. * pe-dll.c (pe_dll_id_target): Likewise. opcodes/ * aarch64-opc.c (vector_qualifier_p): Simplify boolean expression. (fp_qualifier_p, get_data_pattern): Likewise. (aarch64_get_operand_modifier_from_value): Likewise. (aarch64_extend_operator_p, aarch64_shift_operator_p): Likewise. (operand_variant_qualifier_p): Likewise. (qualifier_value_in_range_constraint_p): Likewise. (aarch64_get_qualifier_esize): Likewise. (aarch64_get_qualifier_nelem): Likewise. (aarch64_get_qualifier_standard_value): Likewise. (get_lower_bound, get_upper_bound): Likewise. (aarch64_find_best_match, match_operands_qualifier): Likewise. (aarch64_print_operand): Likewise. * aarch64-opc.h (operand_has_inserter, operand_has_extractor): Likewise. (operand_need_sign_extension, operand_need_shift_by_two): Likewise. (operand_need_shift_by_four, operand_maybe_stack_pointer): Likewise. * arm-dis.c (print_insn_mve, print_insn_thumb32): Likewise. * tic6x-dis.c (tic6x_check_fetch_packet_header): Likewise. (print_insn_tic6x): Likewise.
2021-03-29 07:22:56 +08:00
return aarch64_opnd_qualifiers[qualifier].kind == OQK_OPD_VARIANT;
}
static inline bool
qualifier_value_in_range_constraint_p (aarch64_opnd_qualifier_t qualifier)
{
TRUE/FALSE simplification There is really no need to write code like "foo != 0 ? TRUE : FALSE" unless we had stupidly defined FALSE as something other than 0 or TRUE as something other than 1. The simpler "foo != 0" does just as well. Similarly "(condition == TRUE)" or "(condition == FALSE) can be simplified to "(condition)" and "(!condition)" respectively. I'll note that there is reason to use "integer_expression != 0" when assigning a bfd_boolean rather than the simpler "integer_expression", if you expect the variable to have 0 or 1 value. It's probably even a good idea to not rely on implicit conversion if bfd_boolean were _Bool. bfd/ * aoutx.h (aout_link_write_symbols): Don't cast boolean expression to bfd_boolean. * elf32-or1k.c (or1k_set_got_and_rela_sizes): Dont compare booleans against FALSE. * elf32-arc.c (name_for_global_symbol): Don't compare boolean to TRUE. (is_reloc_PC_relative): Don't use "boolean_condition ? TRUE : FALSE". (is_reloc_SDA_relative, is_reloc_for_GOT): Likewise. (is_reloc_for_PLT, is_reloc_for_TLS): Likewise. * elf32-arm.c (stm32l4xx_need_create_replacing_stub): Likewise. * elf32-nds32.c (insert_nds32_elf_blank): Likewise. * elf32-rx.c (rx_set_section_contents): Likewise. * elfnn-aarch64.c (elfNN_aarch64_final_link_relocate): Likewise. * elfxx-mips.c (_bfd_mips_elf_ignore_undef_symbol): Likewise. * mach-o.c (bfd_mach_o_read_command): Likewise. * targets.c (bfd_get_target_info): Likewise. binutils/ * dlltool.c (main): Don't use "boolean_condition ? TRUE : FALSE". * dwarf.c (read_and_display_attr_value): Likewise. (display_debug_str_offsets): Likewise. * objdump.c (dump_bfd): Likewise. * readelf.c (dump_section_as_strings): Likewise. (dump_section_as_bytes): Likewise. gas/ * atof-generic.c (FALSE, TRUE): Don't define. * config/obj-elf.h (FALSE, TRUE): Don't define. * config/obj-som.h (FALSE, TRUE): Don't define. * config/tc-hppa.h (FALSE, TRUE): Don't define. * config/tc-pdp11.c (FALSE, TRUE): Don't define. * config/tc-iq2000.h (obj_fix_adjustable): Delete. * config/tc-m32r.h (TC_FIX_ADJUSTABLE): Delete. * config/tc-mt.h (obj_fix_adjustable): Delete. * config/tc-nds32.h (TC_FIX_ADJUSTABLE): Delete. * config/tc-arc.c (parse_opcode_flags): Simplify boolean expression. (relaxable_flag, relaxable_operand, assemble_insn): Likewise. (tokenize_extregister): Likewise. * config/tc-csky.c (parse_opcode, get_operand_value): Likewise. (parse_operands_op, parse_operands, md_assemble): Likewise. * config/tc-d10v.c (build_insn): Likewise. * config/tc-score.c (s3_gen_insn_frag): Likewise. * config/tc-score7.c (s7_gen_insn_frag, s7_relax_frag): Likewise. * config/tc-tic6x.c (tic6x_update_features, md_assemble): Likewise. * config/tc-z80.c (emit_byte): Likewise. include/ * opcode/aarch64.h (alias_opcode_p): Simplify boolean expression. (opcode_has_alias, pseudo_opcode_p, optional_operand_p): Likewise. (opcode_has_special_coder): Likewise. ld/ * emultempl/aix.em (gld${EMULATION_NAME}_before_allocation): Simplify boolean expression. * lexsup.c (parse_args): Likewise. * pe-dll.c (pe_dll_id_target): Likewise. opcodes/ * aarch64-opc.c (vector_qualifier_p): Simplify boolean expression. (fp_qualifier_p, get_data_pattern): Likewise. (aarch64_get_operand_modifier_from_value): Likewise. (aarch64_extend_operator_p, aarch64_shift_operator_p): Likewise. (operand_variant_qualifier_p): Likewise. (qualifier_value_in_range_constraint_p): Likewise. (aarch64_get_qualifier_esize): Likewise. (aarch64_get_qualifier_nelem): Likewise. (aarch64_get_qualifier_standard_value): Likewise. (get_lower_bound, get_upper_bound): Likewise. (aarch64_find_best_match, match_operands_qualifier): Likewise. (aarch64_print_operand): Likewise. * aarch64-opc.h (operand_has_inserter, operand_has_extractor): Likewise. (operand_need_sign_extension, operand_need_shift_by_two): Likewise. (operand_need_shift_by_four, operand_maybe_stack_pointer): Likewise. * arm-dis.c (print_insn_mve, print_insn_thumb32): Likewise. * tic6x-dis.c (tic6x_check_fetch_packet_header): Likewise. (print_insn_tic6x): Likewise.
2021-03-29 07:22:56 +08:00
return aarch64_opnd_qualifiers[qualifier].kind == OQK_VALUE_IN_RANGE;
}
const char*
aarch64_get_qualifier_name (aarch64_opnd_qualifier_t qualifier)
{
return aarch64_opnd_qualifiers[qualifier].desc;
}
/* Given an operand qualifier, return the expected data element size
of a qualified operand. */
unsigned char
aarch64_get_qualifier_esize (aarch64_opnd_qualifier_t qualifier)
{
TRUE/FALSE simplification There is really no need to write code like "foo != 0 ? TRUE : FALSE" unless we had stupidly defined FALSE as something other than 0 or TRUE as something other than 1. The simpler "foo != 0" does just as well. Similarly "(condition == TRUE)" or "(condition == FALSE) can be simplified to "(condition)" and "(!condition)" respectively. I'll note that there is reason to use "integer_expression != 0" when assigning a bfd_boolean rather than the simpler "integer_expression", if you expect the variable to have 0 or 1 value. It's probably even a good idea to not rely on implicit conversion if bfd_boolean were _Bool. bfd/ * aoutx.h (aout_link_write_symbols): Don't cast boolean expression to bfd_boolean. * elf32-or1k.c (or1k_set_got_and_rela_sizes): Dont compare booleans against FALSE. * elf32-arc.c (name_for_global_symbol): Don't compare boolean to TRUE. (is_reloc_PC_relative): Don't use "boolean_condition ? TRUE : FALSE". (is_reloc_SDA_relative, is_reloc_for_GOT): Likewise. (is_reloc_for_PLT, is_reloc_for_TLS): Likewise. * elf32-arm.c (stm32l4xx_need_create_replacing_stub): Likewise. * elf32-nds32.c (insert_nds32_elf_blank): Likewise. * elf32-rx.c (rx_set_section_contents): Likewise. * elfnn-aarch64.c (elfNN_aarch64_final_link_relocate): Likewise. * elfxx-mips.c (_bfd_mips_elf_ignore_undef_symbol): Likewise. * mach-o.c (bfd_mach_o_read_command): Likewise. * targets.c (bfd_get_target_info): Likewise. binutils/ * dlltool.c (main): Don't use "boolean_condition ? TRUE : FALSE". * dwarf.c (read_and_display_attr_value): Likewise. (display_debug_str_offsets): Likewise. * objdump.c (dump_bfd): Likewise. * readelf.c (dump_section_as_strings): Likewise. (dump_section_as_bytes): Likewise. gas/ * atof-generic.c (FALSE, TRUE): Don't define. * config/obj-elf.h (FALSE, TRUE): Don't define. * config/obj-som.h (FALSE, TRUE): Don't define. * config/tc-hppa.h (FALSE, TRUE): Don't define. * config/tc-pdp11.c (FALSE, TRUE): Don't define. * config/tc-iq2000.h (obj_fix_adjustable): Delete. * config/tc-m32r.h (TC_FIX_ADJUSTABLE): Delete. * config/tc-mt.h (obj_fix_adjustable): Delete. * config/tc-nds32.h (TC_FIX_ADJUSTABLE): Delete. * config/tc-arc.c (parse_opcode_flags): Simplify boolean expression. (relaxable_flag, relaxable_operand, assemble_insn): Likewise. (tokenize_extregister): Likewise. * config/tc-csky.c (parse_opcode, get_operand_value): Likewise. (parse_operands_op, parse_operands, md_assemble): Likewise. * config/tc-d10v.c (build_insn): Likewise. * config/tc-score.c (s3_gen_insn_frag): Likewise. * config/tc-score7.c (s7_gen_insn_frag, s7_relax_frag): Likewise. * config/tc-tic6x.c (tic6x_update_features, md_assemble): Likewise. * config/tc-z80.c (emit_byte): Likewise. include/ * opcode/aarch64.h (alias_opcode_p): Simplify boolean expression. (opcode_has_alias, pseudo_opcode_p, optional_operand_p): Likewise. (opcode_has_special_coder): Likewise. ld/ * emultempl/aix.em (gld${EMULATION_NAME}_before_allocation): Simplify boolean expression. * lexsup.c (parse_args): Likewise. * pe-dll.c (pe_dll_id_target): Likewise. opcodes/ * aarch64-opc.c (vector_qualifier_p): Simplify boolean expression. (fp_qualifier_p, get_data_pattern): Likewise. (aarch64_get_operand_modifier_from_value): Likewise. (aarch64_extend_operator_p, aarch64_shift_operator_p): Likewise. (operand_variant_qualifier_p): Likewise. (qualifier_value_in_range_constraint_p): Likewise. (aarch64_get_qualifier_esize): Likewise. (aarch64_get_qualifier_nelem): Likewise. (aarch64_get_qualifier_standard_value): Likewise. (get_lower_bound, get_upper_bound): Likewise. (aarch64_find_best_match, match_operands_qualifier): Likewise. (aarch64_print_operand): Likewise. * aarch64-opc.h (operand_has_inserter, operand_has_extractor): Likewise. (operand_need_sign_extension, operand_need_shift_by_two): Likewise. (operand_need_shift_by_four, operand_maybe_stack_pointer): Likewise. * arm-dis.c (print_insn_mve, print_insn_thumb32): Likewise. * tic6x-dis.c (tic6x_check_fetch_packet_header): Likewise. (print_insn_tic6x): Likewise.
2021-03-29 07:22:56 +08:00
assert (operand_variant_qualifier_p (qualifier));
return aarch64_opnd_qualifiers[qualifier].data0;
}
unsigned char
aarch64_get_qualifier_nelem (aarch64_opnd_qualifier_t qualifier)
{
TRUE/FALSE simplification There is really no need to write code like "foo != 0 ? TRUE : FALSE" unless we had stupidly defined FALSE as something other than 0 or TRUE as something other than 1. The simpler "foo != 0" does just as well. Similarly "(condition == TRUE)" or "(condition == FALSE) can be simplified to "(condition)" and "(!condition)" respectively. I'll note that there is reason to use "integer_expression != 0" when assigning a bfd_boolean rather than the simpler "integer_expression", if you expect the variable to have 0 or 1 value. It's probably even a good idea to not rely on implicit conversion if bfd_boolean were _Bool. bfd/ * aoutx.h (aout_link_write_symbols): Don't cast boolean expression to bfd_boolean. * elf32-or1k.c (or1k_set_got_and_rela_sizes): Dont compare booleans against FALSE. * elf32-arc.c (name_for_global_symbol): Don't compare boolean to TRUE. (is_reloc_PC_relative): Don't use "boolean_condition ? TRUE : FALSE". (is_reloc_SDA_relative, is_reloc_for_GOT): Likewise. (is_reloc_for_PLT, is_reloc_for_TLS): Likewise. * elf32-arm.c (stm32l4xx_need_create_replacing_stub): Likewise. * elf32-nds32.c (insert_nds32_elf_blank): Likewise. * elf32-rx.c (rx_set_section_contents): Likewise. * elfnn-aarch64.c (elfNN_aarch64_final_link_relocate): Likewise. * elfxx-mips.c (_bfd_mips_elf_ignore_undef_symbol): Likewise. * mach-o.c (bfd_mach_o_read_command): Likewise. * targets.c (bfd_get_target_info): Likewise. binutils/ * dlltool.c (main): Don't use "boolean_condition ? TRUE : FALSE". * dwarf.c (read_and_display_attr_value): Likewise. (display_debug_str_offsets): Likewise. * objdump.c (dump_bfd): Likewise. * readelf.c (dump_section_as_strings): Likewise. (dump_section_as_bytes): Likewise. gas/ * atof-generic.c (FALSE, TRUE): Don't define. * config/obj-elf.h (FALSE, TRUE): Don't define. * config/obj-som.h (FALSE, TRUE): Don't define. * config/tc-hppa.h (FALSE, TRUE): Don't define. * config/tc-pdp11.c (FALSE, TRUE): Don't define. * config/tc-iq2000.h (obj_fix_adjustable): Delete. * config/tc-m32r.h (TC_FIX_ADJUSTABLE): Delete. * config/tc-mt.h (obj_fix_adjustable): Delete. * config/tc-nds32.h (TC_FIX_ADJUSTABLE): Delete. * config/tc-arc.c (parse_opcode_flags): Simplify boolean expression. (relaxable_flag, relaxable_operand, assemble_insn): Likewise. (tokenize_extregister): Likewise. * config/tc-csky.c (parse_opcode, get_operand_value): Likewise. (parse_operands_op, parse_operands, md_assemble): Likewise. * config/tc-d10v.c (build_insn): Likewise. * config/tc-score.c (s3_gen_insn_frag): Likewise. * config/tc-score7.c (s7_gen_insn_frag, s7_relax_frag): Likewise. * config/tc-tic6x.c (tic6x_update_features, md_assemble): Likewise. * config/tc-z80.c (emit_byte): Likewise. include/ * opcode/aarch64.h (alias_opcode_p): Simplify boolean expression. (opcode_has_alias, pseudo_opcode_p, optional_operand_p): Likewise. (opcode_has_special_coder): Likewise. ld/ * emultempl/aix.em (gld${EMULATION_NAME}_before_allocation): Simplify boolean expression. * lexsup.c (parse_args): Likewise. * pe-dll.c (pe_dll_id_target): Likewise. opcodes/ * aarch64-opc.c (vector_qualifier_p): Simplify boolean expression. (fp_qualifier_p, get_data_pattern): Likewise. (aarch64_get_operand_modifier_from_value): Likewise. (aarch64_extend_operator_p, aarch64_shift_operator_p): Likewise. (operand_variant_qualifier_p): Likewise. (qualifier_value_in_range_constraint_p): Likewise. (aarch64_get_qualifier_esize): Likewise. (aarch64_get_qualifier_nelem): Likewise. (aarch64_get_qualifier_standard_value): Likewise. (get_lower_bound, get_upper_bound): Likewise. (aarch64_find_best_match, match_operands_qualifier): Likewise. (aarch64_print_operand): Likewise. * aarch64-opc.h (operand_has_inserter, operand_has_extractor): Likewise. (operand_need_sign_extension, operand_need_shift_by_two): Likewise. (operand_need_shift_by_four, operand_maybe_stack_pointer): Likewise. * arm-dis.c (print_insn_mve, print_insn_thumb32): Likewise. * tic6x-dis.c (tic6x_check_fetch_packet_header): Likewise. (print_insn_tic6x): Likewise.
2021-03-29 07:22:56 +08:00
assert (operand_variant_qualifier_p (qualifier));
return aarch64_opnd_qualifiers[qualifier].data1;
}
aarch64_insn
aarch64_get_qualifier_standard_value (aarch64_opnd_qualifier_t qualifier)
{
TRUE/FALSE simplification There is really no need to write code like "foo != 0 ? TRUE : FALSE" unless we had stupidly defined FALSE as something other than 0 or TRUE as something other than 1. The simpler "foo != 0" does just as well. Similarly "(condition == TRUE)" or "(condition == FALSE) can be simplified to "(condition)" and "(!condition)" respectively. I'll note that there is reason to use "integer_expression != 0" when assigning a bfd_boolean rather than the simpler "integer_expression", if you expect the variable to have 0 or 1 value. It's probably even a good idea to not rely on implicit conversion if bfd_boolean were _Bool. bfd/ * aoutx.h (aout_link_write_symbols): Don't cast boolean expression to bfd_boolean. * elf32-or1k.c (or1k_set_got_and_rela_sizes): Dont compare booleans against FALSE. * elf32-arc.c (name_for_global_symbol): Don't compare boolean to TRUE. (is_reloc_PC_relative): Don't use "boolean_condition ? TRUE : FALSE". (is_reloc_SDA_relative, is_reloc_for_GOT): Likewise. (is_reloc_for_PLT, is_reloc_for_TLS): Likewise. * elf32-arm.c (stm32l4xx_need_create_replacing_stub): Likewise. * elf32-nds32.c (insert_nds32_elf_blank): Likewise. * elf32-rx.c (rx_set_section_contents): Likewise. * elfnn-aarch64.c (elfNN_aarch64_final_link_relocate): Likewise. * elfxx-mips.c (_bfd_mips_elf_ignore_undef_symbol): Likewise. * mach-o.c (bfd_mach_o_read_command): Likewise. * targets.c (bfd_get_target_info): Likewise. binutils/ * dlltool.c (main): Don't use "boolean_condition ? TRUE : FALSE". * dwarf.c (read_and_display_attr_value): Likewise. (display_debug_str_offsets): Likewise. * objdump.c (dump_bfd): Likewise. * readelf.c (dump_section_as_strings): Likewise. (dump_section_as_bytes): Likewise. gas/ * atof-generic.c (FALSE, TRUE): Don't define. * config/obj-elf.h (FALSE, TRUE): Don't define. * config/obj-som.h (FALSE, TRUE): Don't define. * config/tc-hppa.h (FALSE, TRUE): Don't define. * config/tc-pdp11.c (FALSE, TRUE): Don't define. * config/tc-iq2000.h (obj_fix_adjustable): Delete. * config/tc-m32r.h (TC_FIX_ADJUSTABLE): Delete. * config/tc-mt.h (obj_fix_adjustable): Delete. * config/tc-nds32.h (TC_FIX_ADJUSTABLE): Delete. * config/tc-arc.c (parse_opcode_flags): Simplify boolean expression. (relaxable_flag, relaxable_operand, assemble_insn): Likewise. (tokenize_extregister): Likewise. * config/tc-csky.c (parse_opcode, get_operand_value): Likewise. (parse_operands_op, parse_operands, md_assemble): Likewise. * config/tc-d10v.c (build_insn): Likewise. * config/tc-score.c (s3_gen_insn_frag): Likewise. * config/tc-score7.c (s7_gen_insn_frag, s7_relax_frag): Likewise. * config/tc-tic6x.c (tic6x_update_features, md_assemble): Likewise. * config/tc-z80.c (emit_byte): Likewise. include/ * opcode/aarch64.h (alias_opcode_p): Simplify boolean expression. (opcode_has_alias, pseudo_opcode_p, optional_operand_p): Likewise. (opcode_has_special_coder): Likewise. ld/ * emultempl/aix.em (gld${EMULATION_NAME}_before_allocation): Simplify boolean expression. * lexsup.c (parse_args): Likewise. * pe-dll.c (pe_dll_id_target): Likewise. opcodes/ * aarch64-opc.c (vector_qualifier_p): Simplify boolean expression. (fp_qualifier_p, get_data_pattern): Likewise. (aarch64_get_operand_modifier_from_value): Likewise. (aarch64_extend_operator_p, aarch64_shift_operator_p): Likewise. (operand_variant_qualifier_p): Likewise. (qualifier_value_in_range_constraint_p): Likewise. (aarch64_get_qualifier_esize): Likewise. (aarch64_get_qualifier_nelem): Likewise. (aarch64_get_qualifier_standard_value): Likewise. (get_lower_bound, get_upper_bound): Likewise. (aarch64_find_best_match, match_operands_qualifier): Likewise. (aarch64_print_operand): Likewise. * aarch64-opc.h (operand_has_inserter, operand_has_extractor): Likewise. (operand_need_sign_extension, operand_need_shift_by_two): Likewise. (operand_need_shift_by_four, operand_maybe_stack_pointer): Likewise. * arm-dis.c (print_insn_mve, print_insn_thumb32): Likewise. * tic6x-dis.c (tic6x_check_fetch_packet_header): Likewise. (print_insn_tic6x): Likewise.
2021-03-29 07:22:56 +08:00
assert (operand_variant_qualifier_p (qualifier));
return aarch64_opnd_qualifiers[qualifier].data2;
}
static int
get_lower_bound (aarch64_opnd_qualifier_t qualifier)
{
TRUE/FALSE simplification There is really no need to write code like "foo != 0 ? TRUE : FALSE" unless we had stupidly defined FALSE as something other than 0 or TRUE as something other than 1. The simpler "foo != 0" does just as well. Similarly "(condition == TRUE)" or "(condition == FALSE) can be simplified to "(condition)" and "(!condition)" respectively. I'll note that there is reason to use "integer_expression != 0" when assigning a bfd_boolean rather than the simpler "integer_expression", if you expect the variable to have 0 or 1 value. It's probably even a good idea to not rely on implicit conversion if bfd_boolean were _Bool. bfd/ * aoutx.h (aout_link_write_symbols): Don't cast boolean expression to bfd_boolean. * elf32-or1k.c (or1k_set_got_and_rela_sizes): Dont compare booleans against FALSE. * elf32-arc.c (name_for_global_symbol): Don't compare boolean to TRUE. (is_reloc_PC_relative): Don't use "boolean_condition ? TRUE : FALSE". (is_reloc_SDA_relative, is_reloc_for_GOT): Likewise. (is_reloc_for_PLT, is_reloc_for_TLS): Likewise. * elf32-arm.c (stm32l4xx_need_create_replacing_stub): Likewise. * elf32-nds32.c (insert_nds32_elf_blank): Likewise. * elf32-rx.c (rx_set_section_contents): Likewise. * elfnn-aarch64.c (elfNN_aarch64_final_link_relocate): Likewise. * elfxx-mips.c (_bfd_mips_elf_ignore_undef_symbol): Likewise. * mach-o.c (bfd_mach_o_read_command): Likewise. * targets.c (bfd_get_target_info): Likewise. binutils/ * dlltool.c (main): Don't use "boolean_condition ? TRUE : FALSE". * dwarf.c (read_and_display_attr_value): Likewise. (display_debug_str_offsets): Likewise. * objdump.c (dump_bfd): Likewise. * readelf.c (dump_section_as_strings): Likewise. (dump_section_as_bytes): Likewise. gas/ * atof-generic.c (FALSE, TRUE): Don't define. * config/obj-elf.h (FALSE, TRUE): Don't define. * config/obj-som.h (FALSE, TRUE): Don't define. * config/tc-hppa.h (FALSE, TRUE): Don't define. * config/tc-pdp11.c (FALSE, TRUE): Don't define. * config/tc-iq2000.h (obj_fix_adjustable): Delete. * config/tc-m32r.h (TC_FIX_ADJUSTABLE): Delete. * config/tc-mt.h (obj_fix_adjustable): Delete. * config/tc-nds32.h (TC_FIX_ADJUSTABLE): Delete. * config/tc-arc.c (parse_opcode_flags): Simplify boolean expression. (relaxable_flag, relaxable_operand, assemble_insn): Likewise. (tokenize_extregister): Likewise. * config/tc-csky.c (parse_opcode, get_operand_value): Likewise. (parse_operands_op, parse_operands, md_assemble): Likewise. * config/tc-d10v.c (build_insn): Likewise. * config/tc-score.c (s3_gen_insn_frag): Likewise. * config/tc-score7.c (s7_gen_insn_frag, s7_relax_frag): Likewise. * config/tc-tic6x.c (tic6x_update_features, md_assemble): Likewise. * config/tc-z80.c (emit_byte): Likewise. include/ * opcode/aarch64.h (alias_opcode_p): Simplify boolean expression. (opcode_has_alias, pseudo_opcode_p, optional_operand_p): Likewise. (opcode_has_special_coder): Likewise. ld/ * emultempl/aix.em (gld${EMULATION_NAME}_before_allocation): Simplify boolean expression. * lexsup.c (parse_args): Likewise. * pe-dll.c (pe_dll_id_target): Likewise. opcodes/ * aarch64-opc.c (vector_qualifier_p): Simplify boolean expression. (fp_qualifier_p, get_data_pattern): Likewise. (aarch64_get_operand_modifier_from_value): Likewise. (aarch64_extend_operator_p, aarch64_shift_operator_p): Likewise. (operand_variant_qualifier_p): Likewise. (qualifier_value_in_range_constraint_p): Likewise. (aarch64_get_qualifier_esize): Likewise. (aarch64_get_qualifier_nelem): Likewise. (aarch64_get_qualifier_standard_value): Likewise. (get_lower_bound, get_upper_bound): Likewise. (aarch64_find_best_match, match_operands_qualifier): Likewise. (aarch64_print_operand): Likewise. * aarch64-opc.h (operand_has_inserter, operand_has_extractor): Likewise. (operand_need_sign_extension, operand_need_shift_by_two): Likewise. (operand_need_shift_by_four, operand_maybe_stack_pointer): Likewise. * arm-dis.c (print_insn_mve, print_insn_thumb32): Likewise. * tic6x-dis.c (tic6x_check_fetch_packet_header): Likewise. (print_insn_tic6x): Likewise.
2021-03-29 07:22:56 +08:00
assert (qualifier_value_in_range_constraint_p (qualifier));
return aarch64_opnd_qualifiers[qualifier].data0;
}
static int
get_upper_bound (aarch64_opnd_qualifier_t qualifier)
{
TRUE/FALSE simplification There is really no need to write code like "foo != 0 ? TRUE : FALSE" unless we had stupidly defined FALSE as something other than 0 or TRUE as something other than 1. The simpler "foo != 0" does just as well. Similarly "(condition == TRUE)" or "(condition == FALSE) can be simplified to "(condition)" and "(!condition)" respectively. I'll note that there is reason to use "integer_expression != 0" when assigning a bfd_boolean rather than the simpler "integer_expression", if you expect the variable to have 0 or 1 value. It's probably even a good idea to not rely on implicit conversion if bfd_boolean were _Bool. bfd/ * aoutx.h (aout_link_write_symbols): Don't cast boolean expression to bfd_boolean. * elf32-or1k.c (or1k_set_got_and_rela_sizes): Dont compare booleans against FALSE. * elf32-arc.c (name_for_global_symbol): Don't compare boolean to TRUE. (is_reloc_PC_relative): Don't use "boolean_condition ? TRUE : FALSE". (is_reloc_SDA_relative, is_reloc_for_GOT): Likewise. (is_reloc_for_PLT, is_reloc_for_TLS): Likewise. * elf32-arm.c (stm32l4xx_need_create_replacing_stub): Likewise. * elf32-nds32.c (insert_nds32_elf_blank): Likewise. * elf32-rx.c (rx_set_section_contents): Likewise. * elfnn-aarch64.c (elfNN_aarch64_final_link_relocate): Likewise. * elfxx-mips.c (_bfd_mips_elf_ignore_undef_symbol): Likewise. * mach-o.c (bfd_mach_o_read_command): Likewise. * targets.c (bfd_get_target_info): Likewise. binutils/ * dlltool.c (main): Don't use "boolean_condition ? TRUE : FALSE". * dwarf.c (read_and_display_attr_value): Likewise. (display_debug_str_offsets): Likewise. * objdump.c (dump_bfd): Likewise. * readelf.c (dump_section_as_strings): Likewise. (dump_section_as_bytes): Likewise. gas/ * atof-generic.c (FALSE, TRUE): Don't define. * config/obj-elf.h (FALSE, TRUE): Don't define. * config/obj-som.h (FALSE, TRUE): Don't define. * config/tc-hppa.h (FALSE, TRUE): Don't define. * config/tc-pdp11.c (FALSE, TRUE): Don't define. * config/tc-iq2000.h (obj_fix_adjustable): Delete. * config/tc-m32r.h (TC_FIX_ADJUSTABLE): Delete. * config/tc-mt.h (obj_fix_adjustable): Delete. * config/tc-nds32.h (TC_FIX_ADJUSTABLE): Delete. * config/tc-arc.c (parse_opcode_flags): Simplify boolean expression. (relaxable_flag, relaxable_operand, assemble_insn): Likewise. (tokenize_extregister): Likewise. * config/tc-csky.c (parse_opcode, get_operand_value): Likewise. (parse_operands_op, parse_operands, md_assemble): Likewise. * config/tc-d10v.c (build_insn): Likewise. * config/tc-score.c (s3_gen_insn_frag): Likewise. * config/tc-score7.c (s7_gen_insn_frag, s7_relax_frag): Likewise. * config/tc-tic6x.c (tic6x_update_features, md_assemble): Likewise. * config/tc-z80.c (emit_byte): Likewise. include/ * opcode/aarch64.h (alias_opcode_p): Simplify boolean expression. (opcode_has_alias, pseudo_opcode_p, optional_operand_p): Likewise. (opcode_has_special_coder): Likewise. ld/ * emultempl/aix.em (gld${EMULATION_NAME}_before_allocation): Simplify boolean expression. * lexsup.c (parse_args): Likewise. * pe-dll.c (pe_dll_id_target): Likewise. opcodes/ * aarch64-opc.c (vector_qualifier_p): Simplify boolean expression. (fp_qualifier_p, get_data_pattern): Likewise. (aarch64_get_operand_modifier_from_value): Likewise. (aarch64_extend_operator_p, aarch64_shift_operator_p): Likewise. (operand_variant_qualifier_p): Likewise. (qualifier_value_in_range_constraint_p): Likewise. (aarch64_get_qualifier_esize): Likewise. (aarch64_get_qualifier_nelem): Likewise. (aarch64_get_qualifier_standard_value): Likewise. (get_lower_bound, get_upper_bound): Likewise. (aarch64_find_best_match, match_operands_qualifier): Likewise. (aarch64_print_operand): Likewise. * aarch64-opc.h (operand_has_inserter, operand_has_extractor): Likewise. (operand_need_sign_extension, operand_need_shift_by_two): Likewise. (operand_need_shift_by_four, operand_maybe_stack_pointer): Likewise. * arm-dis.c (print_insn_mve, print_insn_thumb32): Likewise. * tic6x-dis.c (tic6x_check_fetch_packet_header): Likewise. (print_insn_tic6x): Likewise.
2021-03-29 07:22:56 +08:00
assert (qualifier_value_in_range_constraint_p (qualifier));
return aarch64_opnd_qualifiers[qualifier].data1;
}
#ifdef DEBUG_AARCH64
void
aarch64_verbose (const char *str, ...)
{
va_list ap;
va_start (ap, str);
printf ("#### ");
vprintf (str, ap);
printf ("\n");
va_end (ap);
}
static inline void
dump_qualifier_sequence (const aarch64_opnd_qualifier_t *qualifier)
{
int i;
printf ("#### \t");
for (i = 0; i < AARCH64_MAX_OPND_NUM; ++i, ++qualifier)
printf ("%s,", aarch64_get_qualifier_name (*qualifier));
printf ("\n");
}
static void
dump_match_qualifiers (const struct aarch64_opnd_info *opnd,
const aarch64_opnd_qualifier_t *qualifier)
{
int i;
aarch64_opnd_qualifier_t curr[AARCH64_MAX_OPND_NUM];
aarch64_verbose ("dump_match_qualifiers:");
for (i = 0; i < AARCH64_MAX_OPND_NUM; ++i)
curr[i] = opnd[i].qualifier;
dump_qualifier_sequence (curr);
aarch64_verbose ("against");
dump_qualifier_sequence (qualifier);
}
#endif /* DEBUG_AARCH64 */
AArch64: Add SVE constraints verifier. This patch adds the verification rules for move prefix constraints. The Arm SVE instruction MOVPRFX introduces[1] constraints on the instruction at PC+4. Particularly the following constraints are handled by this patch * MOVPRFX must be followed by an instruction. * MOVPRFX can only be followed by non-layout altering directives. * MOVPRFX destination register MUST be used as the destination register in the instruction at PC+4, and is not allowed to be used in any other position other than destructive input. This includes registers that architecturally overlap. e.g. x1 should be treated as z1. * MOVPRFX must be followed by a restricted set of SVE instructions. * The size of the destination register of MOVPRFX must be equal to that of the operation at PC+4. * The predicate register and operation of MOVPRFX must match that of the instruction at PC+4 * The predicated instruction at PC+4 must use the merging predicate. * Architectural aliases and pseudo-instructions need to be supported as well. * MOVPRFX cannot be the last instruction in a sequence Any failure to adhere to any of these constrains will emit an assembly warning and a disassembly note. [1] https://developer.arm.com/docs/ddi0584/latest/arm-architecture-reference-manual-supplement-the-scalable-vector-extension-sve-for-armv8-a include/ * opcode/aarch64.h (aarch64_inst): Remove. (enum err_type): Add ERR_VFI. (aarch64_is_destructive_by_operands): New. (init_insn_sequence): New. (aarch64_decode_insn): Remove param name. opcodes/ * aarch64-opc.c (init_insn_block): New. (verify_constraints, aarch64_is_destructive_by_operands): New. * aarch64-opc.h (verify_constraints): New. gas/ * config/tc-aarch64.c (output_operand_error_report): Order warnings.
2018-10-04 01:38:42 +08:00
/* This function checks if the given instruction INSN is a destructive
instruction based on the usage of the registers. It does not recognize
unary destructive instructions. */
bool
AArch64: Add SVE constraints verifier. This patch adds the verification rules for move prefix constraints. The Arm SVE instruction MOVPRFX introduces[1] constraints on the instruction at PC+4. Particularly the following constraints are handled by this patch * MOVPRFX must be followed by an instruction. * MOVPRFX can only be followed by non-layout altering directives. * MOVPRFX destination register MUST be used as the destination register in the instruction at PC+4, and is not allowed to be used in any other position other than destructive input. This includes registers that architecturally overlap. e.g. x1 should be treated as z1. * MOVPRFX must be followed by a restricted set of SVE instructions. * The size of the destination register of MOVPRFX must be equal to that of the operation at PC+4. * The predicate register and operation of MOVPRFX must match that of the instruction at PC+4 * The predicated instruction at PC+4 must use the merging predicate. * Architectural aliases and pseudo-instructions need to be supported as well. * MOVPRFX cannot be the last instruction in a sequence Any failure to adhere to any of these constrains will emit an assembly warning and a disassembly note. [1] https://developer.arm.com/docs/ddi0584/latest/arm-architecture-reference-manual-supplement-the-scalable-vector-extension-sve-for-armv8-a include/ * opcode/aarch64.h (aarch64_inst): Remove. (enum err_type): Add ERR_VFI. (aarch64_is_destructive_by_operands): New. (init_insn_sequence): New. (aarch64_decode_insn): Remove param name. opcodes/ * aarch64-opc.c (init_insn_block): New. (verify_constraints, aarch64_is_destructive_by_operands): New. * aarch64-opc.h (verify_constraints): New. gas/ * config/tc-aarch64.c (output_operand_error_report): Order warnings.
2018-10-04 01:38:42 +08:00
aarch64_is_destructive_by_operands (const aarch64_opcode *opcode)
{
int i = 0;
const enum aarch64_opnd *opnds = opcode->operands;
if (opnds[0] == AARCH64_OPND_NIL)
return false;
AArch64: Add SVE constraints verifier. This patch adds the verification rules for move prefix constraints. The Arm SVE instruction MOVPRFX introduces[1] constraints on the instruction at PC+4. Particularly the following constraints are handled by this patch * MOVPRFX must be followed by an instruction. * MOVPRFX can only be followed by non-layout altering directives. * MOVPRFX destination register MUST be used as the destination register in the instruction at PC+4, and is not allowed to be used in any other position other than destructive input. This includes registers that architecturally overlap. e.g. x1 should be treated as z1. * MOVPRFX must be followed by a restricted set of SVE instructions. * The size of the destination register of MOVPRFX must be equal to that of the operation at PC+4. * The predicate register and operation of MOVPRFX must match that of the instruction at PC+4 * The predicated instruction at PC+4 must use the merging predicate. * Architectural aliases and pseudo-instructions need to be supported as well. * MOVPRFX cannot be the last instruction in a sequence Any failure to adhere to any of these constrains will emit an assembly warning and a disassembly note. [1] https://developer.arm.com/docs/ddi0584/latest/arm-architecture-reference-manual-supplement-the-scalable-vector-extension-sve-for-armv8-a include/ * opcode/aarch64.h (aarch64_inst): Remove. (enum err_type): Add ERR_VFI. (aarch64_is_destructive_by_operands): New. (init_insn_sequence): New. (aarch64_decode_insn): Remove param name. opcodes/ * aarch64-opc.c (init_insn_block): New. (verify_constraints, aarch64_is_destructive_by_operands): New. * aarch64-opc.h (verify_constraints): New. gas/ * config/tc-aarch64.c (output_operand_error_report): Order warnings.
2018-10-04 01:38:42 +08:00
while (opnds[++i] != AARCH64_OPND_NIL)
if (opnds[i] == opnds[0])
return true;
AArch64: Add SVE constraints verifier. This patch adds the verification rules for move prefix constraints. The Arm SVE instruction MOVPRFX introduces[1] constraints on the instruction at PC+4. Particularly the following constraints are handled by this patch * MOVPRFX must be followed by an instruction. * MOVPRFX can only be followed by non-layout altering directives. * MOVPRFX destination register MUST be used as the destination register in the instruction at PC+4, and is not allowed to be used in any other position other than destructive input. This includes registers that architecturally overlap. e.g. x1 should be treated as z1. * MOVPRFX must be followed by a restricted set of SVE instructions. * The size of the destination register of MOVPRFX must be equal to that of the operation at PC+4. * The predicate register and operation of MOVPRFX must match that of the instruction at PC+4 * The predicated instruction at PC+4 must use the merging predicate. * Architectural aliases and pseudo-instructions need to be supported as well. * MOVPRFX cannot be the last instruction in a sequence Any failure to adhere to any of these constrains will emit an assembly warning and a disassembly note. [1] https://developer.arm.com/docs/ddi0584/latest/arm-architecture-reference-manual-supplement-the-scalable-vector-extension-sve-for-armv8-a include/ * opcode/aarch64.h (aarch64_inst): Remove. (enum err_type): Add ERR_VFI. (aarch64_is_destructive_by_operands): New. (init_insn_sequence): New. (aarch64_decode_insn): Remove param name. opcodes/ * aarch64-opc.c (init_insn_block): New. (verify_constraints, aarch64_is_destructive_by_operands): New. * aarch64-opc.h (verify_constraints): New. gas/ * config/tc-aarch64.c (output_operand_error_report): Order warnings.
2018-10-04 01:38:42 +08:00
return false;
AArch64: Add SVE constraints verifier. This patch adds the verification rules for move prefix constraints. The Arm SVE instruction MOVPRFX introduces[1] constraints on the instruction at PC+4. Particularly the following constraints are handled by this patch * MOVPRFX must be followed by an instruction. * MOVPRFX can only be followed by non-layout altering directives. * MOVPRFX destination register MUST be used as the destination register in the instruction at PC+4, and is not allowed to be used in any other position other than destructive input. This includes registers that architecturally overlap. e.g. x1 should be treated as z1. * MOVPRFX must be followed by a restricted set of SVE instructions. * The size of the destination register of MOVPRFX must be equal to that of the operation at PC+4. * The predicate register and operation of MOVPRFX must match that of the instruction at PC+4 * The predicated instruction at PC+4 must use the merging predicate. * Architectural aliases and pseudo-instructions need to be supported as well. * MOVPRFX cannot be the last instruction in a sequence Any failure to adhere to any of these constrains will emit an assembly warning and a disassembly note. [1] https://developer.arm.com/docs/ddi0584/latest/arm-architecture-reference-manual-supplement-the-scalable-vector-extension-sve-for-armv8-a include/ * opcode/aarch64.h (aarch64_inst): Remove. (enum err_type): Add ERR_VFI. (aarch64_is_destructive_by_operands): New. (init_insn_sequence): New. (aarch64_decode_insn): Remove param name. opcodes/ * aarch64-opc.c (init_insn_block): New. (verify_constraints, aarch64_is_destructive_by_operands): New. * aarch64-opc.h (verify_constraints): New. gas/ * config/tc-aarch64.c (output_operand_error_report): Order warnings.
2018-10-04 01:38:42 +08:00
}
/* TODO improve this, we can have an extra field at the runtime to
store the number of operands rather than calculating it every time. */
int
aarch64_num_of_operands (const aarch64_opcode *opcode)
{
int i = 0;
const enum aarch64_opnd *opnds = opcode->operands;
while (opnds[i++] != AARCH64_OPND_NIL)
;
--i;
assert (i >= 0 && i <= AARCH64_MAX_OPND_NUM);
return i;
}
/* Find the best matched qualifier sequence in *QUALIFIERS_LIST for INST.
If succeeds, fill the found sequence in *RET, return 1; otherwise return 0.
Store the smallest number of non-matching qualifiers in *INVALID_COUNT.
This is always 0 if the function succeeds.
N.B. on the entry, it is very likely that only some operands in *INST
have had their qualifiers been established.
If STOP_AT is not -1, the function will only try to match
the qualifier sequence for operands before and including the operand
of index STOP_AT; and on success *RET will only be filled with the first
(STOP_AT+1) qualifiers.
A couple examples of the matching algorithm:
X,W,NIL should match
X,W,NIL
NIL,NIL should match
X ,NIL
Apart from serving the main encoding routine, this can also be called
during or after the operand decoding. */
int
aarch64_find_best_match (const aarch64_inst *inst,
const aarch64_opnd_qualifier_seq_t *qualifiers_list,
int stop_at, aarch64_opnd_qualifier_t *ret,
int *invalid_count)
{
int i, num_opnds, invalid, min_invalid;
const aarch64_opnd_qualifier_t *qualifiers;
num_opnds = aarch64_num_of_operands (inst->opcode);
if (num_opnds == 0)
{
DEBUG_TRACE ("SUCCEED: no operand");
*invalid_count = 0;
return 1;
}
if (stop_at < 0 || stop_at >= num_opnds)
stop_at = num_opnds - 1;
/* For each pattern. */
min_invalid = num_opnds;
for (i = 0; i < AARCH64_MAX_QLF_SEQ_NUM; ++i, ++qualifiers_list)
{
int j;
qualifiers = *qualifiers_list;
/* Start as positive. */
invalid = 0;
DEBUG_TRACE ("%d", i);
#ifdef DEBUG_AARCH64
if (debug_dump)
dump_match_qualifiers (inst->operands, qualifiers);
#endif
/* The first entry should be taken literally, even if it's an empty
qualifier sequence. (This matters for strict testing.) In other
positions an empty sequence acts as a terminator. */
if (i > 0 && empty_qualifier_sequence_p (qualifiers))
break;
for (j = 0; j < num_opnds && j <= stop_at; ++j, ++qualifiers)
{
if (inst->operands[j].qualifier == AARCH64_OPND_QLF_NIL
&& !(inst->opcode->flags & F_STRICT))
{
/* Either the operand does not have qualifier, or the qualifier
for the operand needs to be deduced from the qualifier
sequence.
In the latter case, any constraint checking related with
the obtained qualifier should be done later in
operand_general_constraint_met_p. */
continue;
}
else if (*qualifiers != inst->operands[j].qualifier)
{
/* Unless the target qualifier can also qualify the operand
(which has already had a non-nil qualifier), non-equal
qualifiers are generally un-matched. */
if (operand_also_qualified_p (inst->operands + j, *qualifiers))
continue;
else
invalid += 1;
}
else
continue; /* Equal qualifiers are certainly matched. */
}
if (min_invalid > invalid)
min_invalid = invalid;
/* Qualifiers established. */
if (min_invalid == 0)
break;
}
*invalid_count = min_invalid;
if (min_invalid == 0)
{
/* Fill the result in *RET. */
int j;
qualifiers = *qualifiers_list;
DEBUG_TRACE ("complete qualifiers using list %d", i);
#ifdef DEBUG_AARCH64
if (debug_dump)
dump_qualifier_sequence (qualifiers);
#endif
for (j = 0; j <= stop_at; ++j, ++qualifiers)
ret[j] = *qualifiers;
for (; j < AARCH64_MAX_OPND_NUM; ++j)
ret[j] = AARCH64_OPND_QLF_NIL;
DEBUG_TRACE ("SUCCESS");
return 1;
}
DEBUG_TRACE ("FAIL");
return 0;
}
/* Operand qualifier matching and resolving.
Return 1 if the operand qualifier(s) in *INST match one of the qualifier
sequences in INST->OPCODE->qualifiers_list; otherwise return 0.
Store the smallest number of non-matching qualifiers in *INVALID_COUNT.
This is always 0 if the function succeeds.
TRUE/FALSE simplification There is really no need to write code like "foo != 0 ? TRUE : FALSE" unless we had stupidly defined FALSE as something other than 0 or TRUE as something other than 1. The simpler "foo != 0" does just as well. Similarly "(condition == TRUE)" or "(condition == FALSE) can be simplified to "(condition)" and "(!condition)" respectively. I'll note that there is reason to use "integer_expression != 0" when assigning a bfd_boolean rather than the simpler "integer_expression", if you expect the variable to have 0 or 1 value. It's probably even a good idea to not rely on implicit conversion if bfd_boolean were _Bool. bfd/ * aoutx.h (aout_link_write_symbols): Don't cast boolean expression to bfd_boolean. * elf32-or1k.c (or1k_set_got_and_rela_sizes): Dont compare booleans against FALSE. * elf32-arc.c (name_for_global_symbol): Don't compare boolean to TRUE. (is_reloc_PC_relative): Don't use "boolean_condition ? TRUE : FALSE". (is_reloc_SDA_relative, is_reloc_for_GOT): Likewise. (is_reloc_for_PLT, is_reloc_for_TLS): Likewise. * elf32-arm.c (stm32l4xx_need_create_replacing_stub): Likewise. * elf32-nds32.c (insert_nds32_elf_blank): Likewise. * elf32-rx.c (rx_set_section_contents): Likewise. * elfnn-aarch64.c (elfNN_aarch64_final_link_relocate): Likewise. * elfxx-mips.c (_bfd_mips_elf_ignore_undef_symbol): Likewise. * mach-o.c (bfd_mach_o_read_command): Likewise. * targets.c (bfd_get_target_info): Likewise. binutils/ * dlltool.c (main): Don't use "boolean_condition ? TRUE : FALSE". * dwarf.c (read_and_display_attr_value): Likewise. (display_debug_str_offsets): Likewise. * objdump.c (dump_bfd): Likewise. * readelf.c (dump_section_as_strings): Likewise. (dump_section_as_bytes): Likewise. gas/ * atof-generic.c (FALSE, TRUE): Don't define. * config/obj-elf.h (FALSE, TRUE): Don't define. * config/obj-som.h (FALSE, TRUE): Don't define. * config/tc-hppa.h (FALSE, TRUE): Don't define. * config/tc-pdp11.c (FALSE, TRUE): Don't define. * config/tc-iq2000.h (obj_fix_adjustable): Delete. * config/tc-m32r.h (TC_FIX_ADJUSTABLE): Delete. * config/tc-mt.h (obj_fix_adjustable): Delete. * config/tc-nds32.h (TC_FIX_ADJUSTABLE): Delete. * config/tc-arc.c (parse_opcode_flags): Simplify boolean expression. (relaxable_flag, relaxable_operand, assemble_insn): Likewise. (tokenize_extregister): Likewise. * config/tc-csky.c (parse_opcode, get_operand_value): Likewise. (parse_operands_op, parse_operands, md_assemble): Likewise. * config/tc-d10v.c (build_insn): Likewise. * config/tc-score.c (s3_gen_insn_frag): Likewise. * config/tc-score7.c (s7_gen_insn_frag, s7_relax_frag): Likewise. * config/tc-tic6x.c (tic6x_update_features, md_assemble): Likewise. * config/tc-z80.c (emit_byte): Likewise. include/ * opcode/aarch64.h (alias_opcode_p): Simplify boolean expression. (opcode_has_alias, pseudo_opcode_p, optional_operand_p): Likewise. (opcode_has_special_coder): Likewise. ld/ * emultempl/aix.em (gld${EMULATION_NAME}_before_allocation): Simplify boolean expression. * lexsup.c (parse_args): Likewise. * pe-dll.c (pe_dll_id_target): Likewise. opcodes/ * aarch64-opc.c (vector_qualifier_p): Simplify boolean expression. (fp_qualifier_p, get_data_pattern): Likewise. (aarch64_get_operand_modifier_from_value): Likewise. (aarch64_extend_operator_p, aarch64_shift_operator_p): Likewise. (operand_variant_qualifier_p): Likewise. (qualifier_value_in_range_constraint_p): Likewise. (aarch64_get_qualifier_esize): Likewise. (aarch64_get_qualifier_nelem): Likewise. (aarch64_get_qualifier_standard_value): Likewise. (get_lower_bound, get_upper_bound): Likewise. (aarch64_find_best_match, match_operands_qualifier): Likewise. (aarch64_print_operand): Likewise. * aarch64-opc.h (operand_has_inserter, operand_has_extractor): Likewise. (operand_need_sign_extension, operand_need_shift_by_two): Likewise. (operand_need_shift_by_four, operand_maybe_stack_pointer): Likewise. * arm-dis.c (print_insn_mve, print_insn_thumb32): Likewise. * tic6x-dis.c (tic6x_check_fetch_packet_header): Likewise. (print_insn_tic6x): Likewise.
2021-03-29 07:22:56 +08:00
if UPDATE_P, update the qualifier(s) in *INST after the matching
succeeds. */
static int
match_operands_qualifier (aarch64_inst *inst, bool update_p,
int *invalid_count)
{
int i;
aarch64_opnd_qualifier_seq_t qualifiers;
if (!aarch64_find_best_match (inst, inst->opcode->qualifiers_list, -1,
qualifiers, invalid_count))
{
DEBUG_TRACE ("matching FAIL");
return 0;
}
/* Update the qualifiers. */
TRUE/FALSE simplification There is really no need to write code like "foo != 0 ? TRUE : FALSE" unless we had stupidly defined FALSE as something other than 0 or TRUE as something other than 1. The simpler "foo != 0" does just as well. Similarly "(condition == TRUE)" or "(condition == FALSE) can be simplified to "(condition)" and "(!condition)" respectively. I'll note that there is reason to use "integer_expression != 0" when assigning a bfd_boolean rather than the simpler "integer_expression", if you expect the variable to have 0 or 1 value. It's probably even a good idea to not rely on implicit conversion if bfd_boolean were _Bool. bfd/ * aoutx.h (aout_link_write_symbols): Don't cast boolean expression to bfd_boolean. * elf32-or1k.c (or1k_set_got_and_rela_sizes): Dont compare booleans against FALSE. * elf32-arc.c (name_for_global_symbol): Don't compare boolean to TRUE. (is_reloc_PC_relative): Don't use "boolean_condition ? TRUE : FALSE". (is_reloc_SDA_relative, is_reloc_for_GOT): Likewise. (is_reloc_for_PLT, is_reloc_for_TLS): Likewise. * elf32-arm.c (stm32l4xx_need_create_replacing_stub): Likewise. * elf32-nds32.c (insert_nds32_elf_blank): Likewise. * elf32-rx.c (rx_set_section_contents): Likewise. * elfnn-aarch64.c (elfNN_aarch64_final_link_relocate): Likewise. * elfxx-mips.c (_bfd_mips_elf_ignore_undef_symbol): Likewise. * mach-o.c (bfd_mach_o_read_command): Likewise. * targets.c (bfd_get_target_info): Likewise. binutils/ * dlltool.c (main): Don't use "boolean_condition ? TRUE : FALSE". * dwarf.c (read_and_display_attr_value): Likewise. (display_debug_str_offsets): Likewise. * objdump.c (dump_bfd): Likewise. * readelf.c (dump_section_as_strings): Likewise. (dump_section_as_bytes): Likewise. gas/ * atof-generic.c (FALSE, TRUE): Don't define. * config/obj-elf.h (FALSE, TRUE): Don't define. * config/obj-som.h (FALSE, TRUE): Don't define. * config/tc-hppa.h (FALSE, TRUE): Don't define. * config/tc-pdp11.c (FALSE, TRUE): Don't define. * config/tc-iq2000.h (obj_fix_adjustable): Delete. * config/tc-m32r.h (TC_FIX_ADJUSTABLE): Delete. * config/tc-mt.h (obj_fix_adjustable): Delete. * config/tc-nds32.h (TC_FIX_ADJUSTABLE): Delete. * config/tc-arc.c (parse_opcode_flags): Simplify boolean expression. (relaxable_flag, relaxable_operand, assemble_insn): Likewise. (tokenize_extregister): Likewise. * config/tc-csky.c (parse_opcode, get_operand_value): Likewise. (parse_operands_op, parse_operands, md_assemble): Likewise. * config/tc-d10v.c (build_insn): Likewise. * config/tc-score.c (s3_gen_insn_frag): Likewise. * config/tc-score7.c (s7_gen_insn_frag, s7_relax_frag): Likewise. * config/tc-tic6x.c (tic6x_update_features, md_assemble): Likewise. * config/tc-z80.c (emit_byte): Likewise. include/ * opcode/aarch64.h (alias_opcode_p): Simplify boolean expression. (opcode_has_alias, pseudo_opcode_p, optional_operand_p): Likewise. (opcode_has_special_coder): Likewise. ld/ * emultempl/aix.em (gld${EMULATION_NAME}_before_allocation): Simplify boolean expression. * lexsup.c (parse_args): Likewise. * pe-dll.c (pe_dll_id_target): Likewise. opcodes/ * aarch64-opc.c (vector_qualifier_p): Simplify boolean expression. (fp_qualifier_p, get_data_pattern): Likewise. (aarch64_get_operand_modifier_from_value): Likewise. (aarch64_extend_operator_p, aarch64_shift_operator_p): Likewise. (operand_variant_qualifier_p): Likewise. (qualifier_value_in_range_constraint_p): Likewise. (aarch64_get_qualifier_esize): Likewise. (aarch64_get_qualifier_nelem): Likewise. (aarch64_get_qualifier_standard_value): Likewise. (get_lower_bound, get_upper_bound): Likewise. (aarch64_find_best_match, match_operands_qualifier): Likewise. (aarch64_print_operand): Likewise. * aarch64-opc.h (operand_has_inserter, operand_has_extractor): Likewise. (operand_need_sign_extension, operand_need_shift_by_two): Likewise. (operand_need_shift_by_four, operand_maybe_stack_pointer): Likewise. * arm-dis.c (print_insn_mve, print_insn_thumb32): Likewise. * tic6x-dis.c (tic6x_check_fetch_packet_header): Likewise. (print_insn_tic6x): Likewise.
2021-03-29 07:22:56 +08:00
if (update_p)
for (i = 0; i < AARCH64_MAX_OPND_NUM; ++i)
{
if (inst->opcode->operands[i] == AARCH64_OPND_NIL)
break;
DEBUG_TRACE_IF (inst->operands[i].qualifier != qualifiers[i],
"update %s with %s for operand %d",
aarch64_get_qualifier_name (inst->operands[i].qualifier),
aarch64_get_qualifier_name (qualifiers[i]), i);
inst->operands[i].qualifier = qualifiers[i];
}
DEBUG_TRACE ("matching SUCCESS");
return 1;
}
/* Return TRUE if VALUE is a wide constant that can be moved into a general
register by MOVZ.
IS32 indicates whether value is a 32-bit immediate or not.
If SHIFT_AMOUNT is not NULL, on the return of TRUE, the logical left shift
amount will be returned in *SHIFT_AMOUNT. */
bool
aarch64_wide_constant_p (uint64_t value, int is32, unsigned int *shift_amount)
{
int amount;
DEBUG_TRACE ("enter with 0x%" PRIx64 "(%" PRIi64 ")", value, value);
if (is32)
{
/* Allow all zeros or all ones in top 32-bits, so that
32-bit constant expressions like ~0x80000000 are
permitted. */
if (value >> 32 != 0 && value >> 32 != 0xffffffff)
/* Immediate out of range. */
return false;
value &= 0xffffffff;
}
/* first, try movz then movn */
amount = -1;
if ((value & ((uint64_t) 0xffff << 0)) == value)
amount = 0;
else if ((value & ((uint64_t) 0xffff << 16)) == value)
amount = 16;
else if (!is32 && (value & ((uint64_t) 0xffff << 32)) == value)
amount = 32;
else if (!is32 && (value & ((uint64_t) 0xffff << 48)) == value)
amount = 48;
if (amount == -1)
{
DEBUG_TRACE ("exit false with 0x%" PRIx64 "(%" PRIi64 ")", value, value);
return false;
}
if (shift_amount != NULL)
*shift_amount = amount;
DEBUG_TRACE ("exit true with amount %d", amount);
return true;
}
/* Build the accepted values for immediate logical SIMD instructions.
The standard encodings of the immediate value are:
N imms immr SIMD size R S
1 ssssss rrrrrr 64 UInt(rrrrrr) UInt(ssssss)
0 0sssss 0rrrrr 32 UInt(rrrrr) UInt(sssss)
0 10ssss 00rrrr 16 UInt(rrrr) UInt(ssss)
0 110sss 000rrr 8 UInt(rrr) UInt(sss)
0 1110ss 0000rr 4 UInt(rr) UInt(ss)
0 11110s 00000r 2 UInt(r) UInt(s)
where all-ones value of S is reserved.
Let's call E the SIMD size.
The immediate value is: S+1 bits '1' rotated to the right by R.
The total of valid encodings is 64*63 + 32*31 + ... + 2*1 = 5334
(remember S != E - 1). */
#define TOTAL_IMM_NB 5334
typedef struct
{
uint64_t imm;
aarch64_insn encoding;
} simd_imm_encoding;
static simd_imm_encoding simd_immediates[TOTAL_IMM_NB];
static int
simd_imm_encoding_cmp(const void *i1, const void *i2)
{
const simd_imm_encoding *imm1 = (const simd_imm_encoding *)i1;
const simd_imm_encoding *imm2 = (const simd_imm_encoding *)i2;
if (imm1->imm < imm2->imm)
return -1;
if (imm1->imm > imm2->imm)
return +1;
return 0;
}
/* immediate bitfield standard encoding
imm13<12> imm13<5:0> imm13<11:6> SIMD size R S
1 ssssss rrrrrr 64 rrrrrr ssssss
0 0sssss 0rrrrr 32 rrrrr sssss
0 10ssss 00rrrr 16 rrrr ssss
0 110sss 000rrr 8 rrr sss
0 1110ss 0000rr 4 rr ss
0 11110s 00000r 2 r s */
static inline int
encode_immediate_bitfield (int is64, uint32_t s, uint32_t r)
{
return (is64 << 12) | (r << 6) | s;
}
static void
build_immediate_table (void)
{
uint32_t log_e, e, s, r, s_mask;
uint64_t mask, imm;
int nb_imms;
int is64;
nb_imms = 0;
for (log_e = 1; log_e <= 6; log_e++)
{
/* Get element size. */
e = 1u << log_e;
if (log_e == 6)
{
is64 = 1;
mask = 0xffffffffffffffffull;
s_mask = 0;
}
else
{
is64 = 0;
mask = (1ull << e) - 1;
/* log_e s_mask
1 ((1 << 4) - 1) << 2 = 111100
2 ((1 << 3) - 1) << 3 = 111000
3 ((1 << 2) - 1) << 4 = 110000
4 ((1 << 1) - 1) << 5 = 100000
5 ((1 << 0) - 1) << 6 = 000000 */
s_mask = ((1u << (5 - log_e)) - 1) << (log_e + 1);
}
for (s = 0; s < e - 1; s++)
for (r = 0; r < e; r++)
{
/* s+1 consecutive bits to 1 (s < 63) */
imm = (1ull << (s + 1)) - 1;
/* rotate right by r */
if (r != 0)
imm = (imm >> r) | ((imm << (e - r)) & mask);
/* replicate the constant depending on SIMD size */
switch (log_e)
{
case 1: imm = (imm << 2) | imm;
-Wimplicit-fallthrough warning fixes Comment changes. bfd/ * coff-h8300.c: Spell fall through comments consistently. * coffgen.c: Likewise. * elf32-hppa.c: Likewise. * elf32-ppc.c: Likewise. * elf32-score.c: Likewise. * elf32-score7.c: Likewise. * elf64-ppc.c: Likewise. * elfxx-aarch64.c: Likewise. * elfxx-mips.c: Likewise. * cpu-ns32k.c: Add missing fall through comments. * elf-m10300.c: Likewise. * elf32-arm.c: Likewise. * elf32-avr.c: Likewise. * elf32-bfin.c: Likewise. * elf32-frv.c: Likewise. * elf32-i386.c: Likewise. * elf32-microblaze.c: Likewise. * elf32-nds32.c: Likewise. * elf32-ppc.c: Likewise. * elf32-rl78.c: Likewise. * elf32-rx.c: Likewise. * elf32-s390.c: Likewise. * elf32-sh.c: Likewise. * elf32-tic6x.c: Likewise. * elf64-ia64-vms.c: Likewise. * elf64-ppc.c: Likewise. * elf64-s390.c: Likewise. * elf64-x86-64.c: Likewise. * elflink.c: Likewise. * elfnn-aarch64.c: Likewise. * elfnn-ia64.c: Likewise. * ieee.c: Likewise. * oasys.c: Likewise. * pdp11.c: Likewise. * srec.c: Likewise. * versados.c: Likewise. opcodes/ * aarch64-opc.c: Spell fall through comments consistently. * i386-dis.c: Likewise. * aarch64-dis.c: Add missing fall through comments. * aarch64-opc.c: Likewise. * arc-dis.c: Likewise. * arm-dis.c: Likewise. * i386-dis.c: Likewise. * m68k-dis.c: Likewise. * mep-asm.c: Likewise. * ns32k-dis.c: Likewise. * sh-dis.c: Likewise. * tic4x-dis.c: Likewise. * tic6x-dis.c: Likewise. * vax-dis.c: Likewise. binutils/ * dlltool.c: Spell fall through comments consistently. * objcopy.c: Likewise. * readelf.c: Likewise. * dwarf.c: Add missing fall through comments. * elfcomm.c: Likewise. * sysinfo.y: Likewise. * readelf.c: Likewise. Also remove extraneous comments. gas/ * app.c: Add missing fall through comments. * dw2gencfi.c: Likewise. * expr.c: Likewise. * config/tc-alpha.c: Likewise. * config/tc-arc.c: Likewise. * config/tc-arm.c: Likewise. * config/tc-cr16.c: Likewise. * config/tc-crx.c: Likewise. * config/tc-dlx.c: Likewise. * config/tc-h8300.c: Likewise. * config/tc-hppa.c: Likewise. * config/tc-i370.c: Likewise. * config/tc-i386.c: Likewise. * config/tc-i960.c: Likewise. * config/tc-ia64.c: Likewise. * config/tc-m68hc11.c: Likewise. * config/tc-m68k.c: Likewise. * config/tc-mep.c: Likewise. * config/tc-metag.c: Likewise. * config/tc-microblaze.c: Likewise. * config/tc-mips.c: Likewise. * config/tc-ns32k.c: Likewise. * config/tc-rx.c: Likewise. * config/tc-score.c: Likewise. * config/tc-score7.c: Likewise. * config/tc-sh.c: Likewise. * config/tc-tic4x.c: Likewise. * config/tc-vax.c: Likewise. * config/tc-xstormy16.c: Likewise. * config/tc-z80.c: Likewise. * config/tc-z8k.c: Likewise. * config/obj-elf.c: Likewise. * config/tc-i386.c: Likewise. * depend.c: Spell fall through comments consistently. * config/tc-arm.c: Likewise. * config/tc-d10v.c: Likewise. * config/tc-i960.c: Likewise. * config/tc-ia64.c: Likewise. * config/tc-m68k.c: Likewise. * config/tc-mcore.c: Likewise. * config/tc-mep.c: Likewise. * config/tc-ns32k.c: Likewise. * config/tc-visium.c: Likewise. * config/tc-xstormy16.c: Likewise. * config/tc-z8k.c: Likewise. gprof/ * gprof.c: Add missing fall through comments. ld/ * lexsup.c: Spell fall through comments consistently and add missing fall through comments.
2016-10-05 15:47:02 +08:00
/* Fall through. */
case 2: imm = (imm << 4) | imm;
-Wimplicit-fallthrough warning fixes Comment changes. bfd/ * coff-h8300.c: Spell fall through comments consistently. * coffgen.c: Likewise. * elf32-hppa.c: Likewise. * elf32-ppc.c: Likewise. * elf32-score.c: Likewise. * elf32-score7.c: Likewise. * elf64-ppc.c: Likewise. * elfxx-aarch64.c: Likewise. * elfxx-mips.c: Likewise. * cpu-ns32k.c: Add missing fall through comments. * elf-m10300.c: Likewise. * elf32-arm.c: Likewise. * elf32-avr.c: Likewise. * elf32-bfin.c: Likewise. * elf32-frv.c: Likewise. * elf32-i386.c: Likewise. * elf32-microblaze.c: Likewise. * elf32-nds32.c: Likewise. * elf32-ppc.c: Likewise. * elf32-rl78.c: Likewise. * elf32-rx.c: Likewise. * elf32-s390.c: Likewise. * elf32-sh.c: Likewise. * elf32-tic6x.c: Likewise. * elf64-ia64-vms.c: Likewise. * elf64-ppc.c: Likewise. * elf64-s390.c: Likewise. * elf64-x86-64.c: Likewise. * elflink.c: Likewise. * elfnn-aarch64.c: Likewise. * elfnn-ia64.c: Likewise. * ieee.c: Likewise. * oasys.c: Likewise. * pdp11.c: Likewise. * srec.c: Likewise. * versados.c: Likewise. opcodes/ * aarch64-opc.c: Spell fall through comments consistently. * i386-dis.c: Likewise. * aarch64-dis.c: Add missing fall through comments. * aarch64-opc.c: Likewise. * arc-dis.c: Likewise. * arm-dis.c: Likewise. * i386-dis.c: Likewise. * m68k-dis.c: Likewise. * mep-asm.c: Likewise. * ns32k-dis.c: Likewise. * sh-dis.c: Likewise. * tic4x-dis.c: Likewise. * tic6x-dis.c: Likewise. * vax-dis.c: Likewise. binutils/ * dlltool.c: Spell fall through comments consistently. * objcopy.c: Likewise. * readelf.c: Likewise. * dwarf.c: Add missing fall through comments. * elfcomm.c: Likewise. * sysinfo.y: Likewise. * readelf.c: Likewise. Also remove extraneous comments. gas/ * app.c: Add missing fall through comments. * dw2gencfi.c: Likewise. * expr.c: Likewise. * config/tc-alpha.c: Likewise. * config/tc-arc.c: Likewise. * config/tc-arm.c: Likewise. * config/tc-cr16.c: Likewise. * config/tc-crx.c: Likewise. * config/tc-dlx.c: Likewise. * config/tc-h8300.c: Likewise. * config/tc-hppa.c: Likewise. * config/tc-i370.c: Likewise. * config/tc-i386.c: Likewise. * config/tc-i960.c: Likewise. * config/tc-ia64.c: Likewise. * config/tc-m68hc11.c: Likewise. * config/tc-m68k.c: Likewise. * config/tc-mep.c: Likewise. * config/tc-metag.c: Likewise. * config/tc-microblaze.c: Likewise. * config/tc-mips.c: Likewise. * config/tc-ns32k.c: Likewise. * config/tc-rx.c: Likewise. * config/tc-score.c: Likewise. * config/tc-score7.c: Likewise. * config/tc-sh.c: Likewise. * config/tc-tic4x.c: Likewise. * config/tc-vax.c: Likewise. * config/tc-xstormy16.c: Likewise. * config/tc-z80.c: Likewise. * config/tc-z8k.c: Likewise. * config/obj-elf.c: Likewise. * config/tc-i386.c: Likewise. * depend.c: Spell fall through comments consistently. * config/tc-arm.c: Likewise. * config/tc-d10v.c: Likewise. * config/tc-i960.c: Likewise. * config/tc-ia64.c: Likewise. * config/tc-m68k.c: Likewise. * config/tc-mcore.c: Likewise. * config/tc-mep.c: Likewise. * config/tc-ns32k.c: Likewise. * config/tc-visium.c: Likewise. * config/tc-xstormy16.c: Likewise. * config/tc-z8k.c: Likewise. gprof/ * gprof.c: Add missing fall through comments. ld/ * lexsup.c: Spell fall through comments consistently and add missing fall through comments.
2016-10-05 15:47:02 +08:00
/* Fall through. */
case 3: imm = (imm << 8) | imm;
-Wimplicit-fallthrough warning fixes Comment changes. bfd/ * coff-h8300.c: Spell fall through comments consistently. * coffgen.c: Likewise. * elf32-hppa.c: Likewise. * elf32-ppc.c: Likewise. * elf32-score.c: Likewise. * elf32-score7.c: Likewise. * elf64-ppc.c: Likewise. * elfxx-aarch64.c: Likewise. * elfxx-mips.c: Likewise. * cpu-ns32k.c: Add missing fall through comments. * elf-m10300.c: Likewise. * elf32-arm.c: Likewise. * elf32-avr.c: Likewise. * elf32-bfin.c: Likewise. * elf32-frv.c: Likewise. * elf32-i386.c: Likewise. * elf32-microblaze.c: Likewise. * elf32-nds32.c: Likewise. * elf32-ppc.c: Likewise. * elf32-rl78.c: Likewise. * elf32-rx.c: Likewise. * elf32-s390.c: Likewise. * elf32-sh.c: Likewise. * elf32-tic6x.c: Likewise. * elf64-ia64-vms.c: Likewise. * elf64-ppc.c: Likewise. * elf64-s390.c: Likewise. * elf64-x86-64.c: Likewise. * elflink.c: Likewise. * elfnn-aarch64.c: Likewise. * elfnn-ia64.c: Likewise. * ieee.c: Likewise. * oasys.c: Likewise. * pdp11.c: Likewise. * srec.c: Likewise. * versados.c: Likewise. opcodes/ * aarch64-opc.c: Spell fall through comments consistently. * i386-dis.c: Likewise. * aarch64-dis.c: Add missing fall through comments. * aarch64-opc.c: Likewise. * arc-dis.c: Likewise. * arm-dis.c: Likewise. * i386-dis.c: Likewise. * m68k-dis.c: Likewise. * mep-asm.c: Likewise. * ns32k-dis.c: Likewise. * sh-dis.c: Likewise. * tic4x-dis.c: Likewise. * tic6x-dis.c: Likewise. * vax-dis.c: Likewise. binutils/ * dlltool.c: Spell fall through comments consistently. * objcopy.c: Likewise. * readelf.c: Likewise. * dwarf.c: Add missing fall through comments. * elfcomm.c: Likewise. * sysinfo.y: Likewise. * readelf.c: Likewise. Also remove extraneous comments. gas/ * app.c: Add missing fall through comments. * dw2gencfi.c: Likewise. * expr.c: Likewise. * config/tc-alpha.c: Likewise. * config/tc-arc.c: Likewise. * config/tc-arm.c: Likewise. * config/tc-cr16.c: Likewise. * config/tc-crx.c: Likewise. * config/tc-dlx.c: Likewise. * config/tc-h8300.c: Likewise. * config/tc-hppa.c: Likewise. * config/tc-i370.c: Likewise. * config/tc-i386.c: Likewise. * config/tc-i960.c: Likewise. * config/tc-ia64.c: Likewise. * config/tc-m68hc11.c: Likewise. * config/tc-m68k.c: Likewise. * config/tc-mep.c: Likewise. * config/tc-metag.c: Likewise. * config/tc-microblaze.c: Likewise. * config/tc-mips.c: Likewise. * config/tc-ns32k.c: Likewise. * config/tc-rx.c: Likewise. * config/tc-score.c: Likewise. * config/tc-score7.c: Likewise. * config/tc-sh.c: Likewise. * config/tc-tic4x.c: Likewise. * config/tc-vax.c: Likewise. * config/tc-xstormy16.c: Likewise. * config/tc-z80.c: Likewise. * config/tc-z8k.c: Likewise. * config/obj-elf.c: Likewise. * config/tc-i386.c: Likewise. * depend.c: Spell fall through comments consistently. * config/tc-arm.c: Likewise. * config/tc-d10v.c: Likewise. * config/tc-i960.c: Likewise. * config/tc-ia64.c: Likewise. * config/tc-m68k.c: Likewise. * config/tc-mcore.c: Likewise. * config/tc-mep.c: Likewise. * config/tc-ns32k.c: Likewise. * config/tc-visium.c: Likewise. * config/tc-xstormy16.c: Likewise. * config/tc-z8k.c: Likewise. gprof/ * gprof.c: Add missing fall through comments. ld/ * lexsup.c: Spell fall through comments consistently and add missing fall through comments.
2016-10-05 15:47:02 +08:00
/* Fall through. */
case 4: imm = (imm << 16) | imm;
-Wimplicit-fallthrough warning fixes Comment changes. bfd/ * coff-h8300.c: Spell fall through comments consistently. * coffgen.c: Likewise. * elf32-hppa.c: Likewise. * elf32-ppc.c: Likewise. * elf32-score.c: Likewise. * elf32-score7.c: Likewise. * elf64-ppc.c: Likewise. * elfxx-aarch64.c: Likewise. * elfxx-mips.c: Likewise. * cpu-ns32k.c: Add missing fall through comments. * elf-m10300.c: Likewise. * elf32-arm.c: Likewise. * elf32-avr.c: Likewise. * elf32-bfin.c: Likewise. * elf32-frv.c: Likewise. * elf32-i386.c: Likewise. * elf32-microblaze.c: Likewise. * elf32-nds32.c: Likewise. * elf32-ppc.c: Likewise. * elf32-rl78.c: Likewise. * elf32-rx.c: Likewise. * elf32-s390.c: Likewise. * elf32-sh.c: Likewise. * elf32-tic6x.c: Likewise. * elf64-ia64-vms.c: Likewise. * elf64-ppc.c: Likewise. * elf64-s390.c: Likewise. * elf64-x86-64.c: Likewise. * elflink.c: Likewise. * elfnn-aarch64.c: Likewise. * elfnn-ia64.c: Likewise. * ieee.c: Likewise. * oasys.c: Likewise. * pdp11.c: Likewise. * srec.c: Likewise. * versados.c: Likewise. opcodes/ * aarch64-opc.c: Spell fall through comments consistently. * i386-dis.c: Likewise. * aarch64-dis.c: Add missing fall through comments. * aarch64-opc.c: Likewise. * arc-dis.c: Likewise. * arm-dis.c: Likewise. * i386-dis.c: Likewise. * m68k-dis.c: Likewise. * mep-asm.c: Likewise. * ns32k-dis.c: Likewise. * sh-dis.c: Likewise. * tic4x-dis.c: Likewise. * tic6x-dis.c: Likewise. * vax-dis.c: Likewise. binutils/ * dlltool.c: Spell fall through comments consistently. * objcopy.c: Likewise. * readelf.c: Likewise. * dwarf.c: Add missing fall through comments. * elfcomm.c: Likewise. * sysinfo.y: Likewise. * readelf.c: Likewise. Also remove extraneous comments. gas/ * app.c: Add missing fall through comments. * dw2gencfi.c: Likewise. * expr.c: Likewise. * config/tc-alpha.c: Likewise. * config/tc-arc.c: Likewise. * config/tc-arm.c: Likewise. * config/tc-cr16.c: Likewise. * config/tc-crx.c: Likewise. * config/tc-dlx.c: Likewise. * config/tc-h8300.c: Likewise. * config/tc-hppa.c: Likewise. * config/tc-i370.c: Likewise. * config/tc-i386.c: Likewise. * config/tc-i960.c: Likewise. * config/tc-ia64.c: Likewise. * config/tc-m68hc11.c: Likewise. * config/tc-m68k.c: Likewise. * config/tc-mep.c: Likewise. * config/tc-metag.c: Likewise. * config/tc-microblaze.c: Likewise. * config/tc-mips.c: Likewise. * config/tc-ns32k.c: Likewise. * config/tc-rx.c: Likewise. * config/tc-score.c: Likewise. * config/tc-score7.c: Likewise. * config/tc-sh.c: Likewise. * config/tc-tic4x.c: Likewise. * config/tc-vax.c: Likewise. * config/tc-xstormy16.c: Likewise. * config/tc-z80.c: Likewise. * config/tc-z8k.c: Likewise. * config/obj-elf.c: Likewise. * config/tc-i386.c: Likewise. * depend.c: Spell fall through comments consistently. * config/tc-arm.c: Likewise. * config/tc-d10v.c: Likewise. * config/tc-i960.c: Likewise. * config/tc-ia64.c: Likewise. * config/tc-m68k.c: Likewise. * config/tc-mcore.c: Likewise. * config/tc-mep.c: Likewise. * config/tc-ns32k.c: Likewise. * config/tc-visium.c: Likewise. * config/tc-xstormy16.c: Likewise. * config/tc-z8k.c: Likewise. gprof/ * gprof.c: Add missing fall through comments. ld/ * lexsup.c: Spell fall through comments consistently and add missing fall through comments.
2016-10-05 15:47:02 +08:00
/* Fall through. */
case 5: imm = (imm << 32) | imm;
-Wimplicit-fallthrough warning fixes Comment changes. bfd/ * coff-h8300.c: Spell fall through comments consistently. * coffgen.c: Likewise. * elf32-hppa.c: Likewise. * elf32-ppc.c: Likewise. * elf32-score.c: Likewise. * elf32-score7.c: Likewise. * elf64-ppc.c: Likewise. * elfxx-aarch64.c: Likewise. * elfxx-mips.c: Likewise. * cpu-ns32k.c: Add missing fall through comments. * elf-m10300.c: Likewise. * elf32-arm.c: Likewise. * elf32-avr.c: Likewise. * elf32-bfin.c: Likewise. * elf32-frv.c: Likewise. * elf32-i386.c: Likewise. * elf32-microblaze.c: Likewise. * elf32-nds32.c: Likewise. * elf32-ppc.c: Likewise. * elf32-rl78.c: Likewise. * elf32-rx.c: Likewise. * elf32-s390.c: Likewise. * elf32-sh.c: Likewise. * elf32-tic6x.c: Likewise. * elf64-ia64-vms.c: Likewise. * elf64-ppc.c: Likewise. * elf64-s390.c: Likewise. * elf64-x86-64.c: Likewise. * elflink.c: Likewise. * elfnn-aarch64.c: Likewise. * elfnn-ia64.c: Likewise. * ieee.c: Likewise. * oasys.c: Likewise. * pdp11.c: Likewise. * srec.c: Likewise. * versados.c: Likewise. opcodes/ * aarch64-opc.c: Spell fall through comments consistently. * i386-dis.c: Likewise. * aarch64-dis.c: Add missing fall through comments. * aarch64-opc.c: Likewise. * arc-dis.c: Likewise. * arm-dis.c: Likewise. * i386-dis.c: Likewise. * m68k-dis.c: Likewise. * mep-asm.c: Likewise. * ns32k-dis.c: Likewise. * sh-dis.c: Likewise. * tic4x-dis.c: Likewise. * tic6x-dis.c: Likewise. * vax-dis.c: Likewise. binutils/ * dlltool.c: Spell fall through comments consistently. * objcopy.c: Likewise. * readelf.c: Likewise. * dwarf.c: Add missing fall through comments. * elfcomm.c: Likewise. * sysinfo.y: Likewise. * readelf.c: Likewise. Also remove extraneous comments. gas/ * app.c: Add missing fall through comments. * dw2gencfi.c: Likewise. * expr.c: Likewise. * config/tc-alpha.c: Likewise. * config/tc-arc.c: Likewise. * config/tc-arm.c: Likewise. * config/tc-cr16.c: Likewise. * config/tc-crx.c: Likewise. * config/tc-dlx.c: Likewise. * config/tc-h8300.c: Likewise. * config/tc-hppa.c: Likewise. * config/tc-i370.c: Likewise. * config/tc-i386.c: Likewise. * config/tc-i960.c: Likewise. * config/tc-ia64.c: Likewise. * config/tc-m68hc11.c: Likewise. * config/tc-m68k.c: Likewise. * config/tc-mep.c: Likewise. * config/tc-metag.c: Likewise. * config/tc-microblaze.c: Likewise. * config/tc-mips.c: Likewise. * config/tc-ns32k.c: Likewise. * config/tc-rx.c: Likewise. * config/tc-score.c: Likewise. * config/tc-score7.c: Likewise. * config/tc-sh.c: Likewise. * config/tc-tic4x.c: Likewise. * config/tc-vax.c: Likewise. * config/tc-xstormy16.c: Likewise. * config/tc-z80.c: Likewise. * config/tc-z8k.c: Likewise. * config/obj-elf.c: Likewise. * config/tc-i386.c: Likewise. * depend.c: Spell fall through comments consistently. * config/tc-arm.c: Likewise. * config/tc-d10v.c: Likewise. * config/tc-i960.c: Likewise. * config/tc-ia64.c: Likewise. * config/tc-m68k.c: Likewise. * config/tc-mcore.c: Likewise. * config/tc-mep.c: Likewise. * config/tc-ns32k.c: Likewise. * config/tc-visium.c: Likewise. * config/tc-xstormy16.c: Likewise. * config/tc-z8k.c: Likewise. gprof/ * gprof.c: Add missing fall through comments. ld/ * lexsup.c: Spell fall through comments consistently and add missing fall through comments.
2016-10-05 15:47:02 +08:00
/* Fall through. */
case 6: break;
default: abort ();
}
simd_immediates[nb_imms].imm = imm;
simd_immediates[nb_imms].encoding =
encode_immediate_bitfield(is64, s | s_mask, r);
nb_imms++;
}
}
assert (nb_imms == TOTAL_IMM_NB);
qsort(simd_immediates, nb_imms,
sizeof(simd_immediates[0]), simd_imm_encoding_cmp);
}
/* Return TRUE if VALUE is a valid logical immediate, i.e. bitmask, that can
be accepted by logical (immediate) instructions
e.g. ORR <Xd|SP>, <Xn>, #<imm>.
ESIZE is the number of bytes in the decoded immediate value.
If ENCODING is not NULL, on the return of TRUE, the standard encoding for
VALUE will be returned in *ENCODING. */
bool
aarch64_logical_immediate_p (uint64_t value, int esize, aarch64_insn *encoding)
{
simd_imm_encoding imm_enc;
const simd_imm_encoding *imm_encoding;
static bool initialized = false;
uint64_t upper;
int i;
DEBUG_TRACE ("enter with 0x%" PRIx64 "(%" PRIi64 "), esize: %d", value,
value, esize);
if (!initialized)
{
build_immediate_table ();
initialized = true;
}
/* Allow all zeros or all ones in top bits, so that
constant expressions like ~1 are permitted. */
upper = (uint64_t) -1 << (esize * 4) << (esize * 4);
if ((value & ~upper) != value && (value | upper) != value)
return false;
/* Replicate to a full 64-bit value. */
value &= ~upper;
for (i = esize * 8; i < 64; i *= 2)
value |= (value << i);
imm_enc.imm = value;
imm_encoding = (const simd_imm_encoding *)
bsearch(&imm_enc, simd_immediates, TOTAL_IMM_NB,
sizeof(simd_immediates[0]), simd_imm_encoding_cmp);
if (imm_encoding == NULL)
{
DEBUG_TRACE ("exit with false");
return false;
}
if (encoding != NULL)
*encoding = imm_encoding->encoding;
DEBUG_TRACE ("exit with true");
return true;
}
/* If 64-bit immediate IMM is in the format of
"aaaaaaaabbbbbbbbccccccccddddddddeeeeeeeeffffffffgggggggghhhhhhhh",
where a, b, c, d, e, f, g and h are independently 0 or 1, return an integer
of value "abcdefgh". Otherwise return -1. */
int
aarch64_shrink_expanded_imm8 (uint64_t imm)
{
int i, ret;
uint32_t byte;
ret = 0;
for (i = 0; i < 8; i++)
{
byte = (imm >> (8 * i)) & 0xff;
if (byte == 0xff)
ret |= 1 << i;
else if (byte != 0x00)
return -1;
}
return ret;
}
/* Utility inline functions for operand_general_constraint_met_p. */
static inline void
set_error (aarch64_operand_error *mismatch_detail,
enum aarch64_operand_error_kind kind, int idx,
const char* error)
{
if (mismatch_detail == NULL)
return;
mismatch_detail->kind = kind;
mismatch_detail->index = idx;
mismatch_detail->error = error;
}
static inline void
set_syntax_error (aarch64_operand_error *mismatch_detail, int idx,
const char* error)
{
if (mismatch_detail == NULL)
return;
set_error (mismatch_detail, AARCH64_OPDE_SYNTAX_ERROR, idx, error);
}
static inline void
set_invalid_regno_error (aarch64_operand_error *mismatch_detail, int idx,
const char *prefix, int lower_bound, int upper_bound)
{
if (mismatch_detail == NULL)
return;
set_error (mismatch_detail, AARCH64_OPDE_INVALID_REGNO, idx, NULL);
mismatch_detail->data[0].s = prefix;
mismatch_detail->data[1].i = lower_bound;
mismatch_detail->data[2].i = upper_bound;
}
static inline void
set_out_of_range_error (aarch64_operand_error *mismatch_detail,
int idx, int lower_bound, int upper_bound,
const char* error)
{
if (mismatch_detail == NULL)
return;
set_error (mismatch_detail, AARCH64_OPDE_OUT_OF_RANGE, idx, error);
aarch64: Enforce P/M/E order for MOPS instructions The MOPS instructions should be used as a triple, such as: cpyfp [x0]!, [x1]!, x2! cpyfm [x0]!, [x1]!, x2! cpyfe [x0]!, [x1]!, x2! The registers should also be the same for each writeback operand. This patch adds a warning for code that doesn't follow this rule, along similar lines to the warning that we already emit for invalid uses of MOVPRFX. include/ * opcode/aarch64.h (C_SCAN_MOPS_P, C_SCAN_MOPS_M, C_SCAN_MOPS_E) (C_SCAN_MOPS_PME): New macros. (AARCH64_OPDE_A_SHOULD_FOLLOW_B): New aarch64_operand_error_kind. (AARCH64_OPDE_EXPECTED_A_AFTER_B): Likewise. (aarch64_operand_error): Make each data value a union between an int and a string. opcodes/ * aarch64-tbl.h (MOPS_CPY_OP1_OP2_INSN): Add scan flags. (MOPS_SET_OP1_OP2_INSN): Likewise. * aarch64-opc.c (set_out_of_range_error): Update after change to aarch64_operand_error. (set_unaligned_error, set_reg_list_error): Likewise. (init_insn_sequence): Use a 3-instruction sequence for MOPS P instructions. (verify_mops_pme_sequence): New function. (verify_constraints): Call it. * aarch64-dis.c (print_verifier_notes): Handle AARCH64_OPDE_A_SHOULD_FOLLOW_B and AARCH64_OPDE_EXPECTED_A_AFTER_B. gas/ * config/tc-aarch64.c (operand_mismatch_kind_names): Add entries for AARCH64_OPDE_A_SHOULD_FOLLOW_B and AARCH64_OPDE_EXPECTED_A_AFTER_B. (operand_error_higher_severity_p): Check that AARCH64_OPDE_A_SHOULD_FOLLOW_B and AARCH64_OPDE_EXPECTED_A_AFTER_B come between AARCH64_OPDE_RECOVERABLE and AARCH64_OPDE_SYNTAX_ERROR; their relative order is not significant. (record_operand_error_with_data): Update after change to aarch64_operand_error. (output_operand_error_record): Likewise. Handle AARCH64_OPDE_A_SHOULD_FOLLOW_B and AARCH64_OPDE_EXPECTED_A_AFTER_B. * testsuite/gas/aarch64/mops_invalid_2.s, testsuite/gas/aarch64/mops_invalid_2.d, testsuite/gas/aarch64/mops_invalid_2.l: New test.
2021-12-02 23:00:57 +08:00
mismatch_detail->data[0].i = lower_bound;
mismatch_detail->data[1].i = upper_bound;
}
static inline void
set_imm_out_of_range_error (aarch64_operand_error *mismatch_detail,
int idx, int lower_bound, int upper_bound)
{
if (mismatch_detail == NULL)
return;
set_out_of_range_error (mismatch_detail, idx, lower_bound, upper_bound,
_("immediate value"));
}
static inline void
set_offset_out_of_range_error (aarch64_operand_error *mismatch_detail,
int idx, int lower_bound, int upper_bound)
{
if (mismatch_detail == NULL)
return;
set_out_of_range_error (mismatch_detail, idx, lower_bound, upper_bound,
_("immediate offset"));
}
static inline void
set_regno_out_of_range_error (aarch64_operand_error *mismatch_detail,
int idx, int lower_bound, int upper_bound)
{
if (mismatch_detail == NULL)
return;
set_out_of_range_error (mismatch_detail, idx, lower_bound, upper_bound,
_("register number"));
}
static inline void
set_elem_idx_out_of_range_error (aarch64_operand_error *mismatch_detail,
int idx, int lower_bound, int upper_bound)
{
if (mismatch_detail == NULL)
return;
set_out_of_range_error (mismatch_detail, idx, lower_bound, upper_bound,
_("register element index"));
}
static inline void
set_sft_amount_out_of_range_error (aarch64_operand_error *mismatch_detail,
int idx, int lower_bound, int upper_bound)
{
if (mismatch_detail == NULL)
return;
set_out_of_range_error (mismatch_detail, idx, lower_bound, upper_bound,
_("shift amount"));
}
[AArch64][SVE 24/32] Add AARCH64_OPND_SVE_PATTERN_SCALED Some SVE instructions count the number of elements in a given vector pattern and allow a scale factor of [1, 16] to be applied to the result. This scale factor is written ", MUL #n", where "MUL" is a new operator. E.g.: UQINCD X0, POW2, MUL #2 This patch adds support for this kind of operand. All existing operators were shifts of some kind, so there was a natural range of [0, 63] regardless of context. This was then narrowered further by later checks (e.g. to [0, 31] when used for 32-bit values). In contrast, MUL doesn't really have a natural context-independent range. Rather than pick one arbitrarily, it seemed better to make the "shift" amount a full 64-bit value and leave the range test to the usual operand-checking code. I've rearranged the fields of aarch64_opnd_info so that this doesn't increase the size of the structure (although I don't think its size is critical anyway). include/ * opcode/aarch64.h (AARCH64_OPND_SVE_PATTERN_SCALED): New aarch64_opnd. (AARCH64_MOD_MUL): New aarch64_modifier_kind. (aarch64_opnd_info): Make shifter.amount an int64_t and rearrange the fields. opcodes/ * aarch64-tbl.h (AARCH64_OPERANDS): Add an entry for AARCH64_OPND_SVE_PATTERN_SCALED. * aarch64-opc.h (FLD_SVE_imm4): New aarch64_field_kind. * aarch64-opc.c (fields): Add a corresponding entry. (set_multiplier_out_of_range_error): New function. (aarch64_operand_modifiers): Add entry for AARCH64_MOD_MUL. (operand_general_constraint_met_p): Handle AARCH64_OPND_SVE_PATTERN_SCALED. (print_register_offset_address): Use PRIi64 to print the shift amount. (aarch64_print_operand): Likewise. Handle AARCH64_OPND_SVE_PATTERN_SCALED. * aarch64-opc-2.c: Regenerate. * aarch64-asm.h (ins_sve_scale): New inserter. * aarch64-asm.c (aarch64_ins_sve_scale): New function. * aarch64-asm-2.c: Regenerate. * aarch64-dis.h (ext_sve_scale): New inserter. * aarch64-dis.c (aarch64_ext_sve_scale): New function. * aarch64-dis-2.c: Regenerate. gas/ * config/tc-aarch64.c (SHIFTED_MUL): New parse_shift_mode. (parse_shift): Handle it. Reject AARCH64_MOD_MUL for all other shift modes. Skip range tests for AARCH64_MOD_MUL. (process_omitted_operand): Handle AARCH64_OPND_SVE_PATTERN_SCALED. (parse_operands): Likewise.
2016-09-21 23:55:22 +08:00
/* Report that the MUL modifier in operand IDX should be in the range
[LOWER_BOUND, UPPER_BOUND]. */
static inline void
set_multiplier_out_of_range_error (aarch64_operand_error *mismatch_detail,
int idx, int lower_bound, int upper_bound)
{
if (mismatch_detail == NULL)
return;
set_out_of_range_error (mismatch_detail, idx, lower_bound, upper_bound,
_("multiplier"));
}
static inline void
set_unaligned_error (aarch64_operand_error *mismatch_detail, int idx,
int alignment)
{
if (mismatch_detail == NULL)
return;
set_error (mismatch_detail, AARCH64_OPDE_UNALIGNED, idx, NULL);
aarch64: Enforce P/M/E order for MOPS instructions The MOPS instructions should be used as a triple, such as: cpyfp [x0]!, [x1]!, x2! cpyfm [x0]!, [x1]!, x2! cpyfe [x0]!, [x1]!, x2! The registers should also be the same for each writeback operand. This patch adds a warning for code that doesn't follow this rule, along similar lines to the warning that we already emit for invalid uses of MOVPRFX. include/ * opcode/aarch64.h (C_SCAN_MOPS_P, C_SCAN_MOPS_M, C_SCAN_MOPS_E) (C_SCAN_MOPS_PME): New macros. (AARCH64_OPDE_A_SHOULD_FOLLOW_B): New aarch64_operand_error_kind. (AARCH64_OPDE_EXPECTED_A_AFTER_B): Likewise. (aarch64_operand_error): Make each data value a union between an int and a string. opcodes/ * aarch64-tbl.h (MOPS_CPY_OP1_OP2_INSN): Add scan flags. (MOPS_SET_OP1_OP2_INSN): Likewise. * aarch64-opc.c (set_out_of_range_error): Update after change to aarch64_operand_error. (set_unaligned_error, set_reg_list_error): Likewise. (init_insn_sequence): Use a 3-instruction sequence for MOPS P instructions. (verify_mops_pme_sequence): New function. (verify_constraints): Call it. * aarch64-dis.c (print_verifier_notes): Handle AARCH64_OPDE_A_SHOULD_FOLLOW_B and AARCH64_OPDE_EXPECTED_A_AFTER_B. gas/ * config/tc-aarch64.c (operand_mismatch_kind_names): Add entries for AARCH64_OPDE_A_SHOULD_FOLLOW_B and AARCH64_OPDE_EXPECTED_A_AFTER_B. (operand_error_higher_severity_p): Check that AARCH64_OPDE_A_SHOULD_FOLLOW_B and AARCH64_OPDE_EXPECTED_A_AFTER_B come between AARCH64_OPDE_RECOVERABLE and AARCH64_OPDE_SYNTAX_ERROR; their relative order is not significant. (record_operand_error_with_data): Update after change to aarch64_operand_error. (output_operand_error_record): Likewise. Handle AARCH64_OPDE_A_SHOULD_FOLLOW_B and AARCH64_OPDE_EXPECTED_A_AFTER_B. * testsuite/gas/aarch64/mops_invalid_2.s, testsuite/gas/aarch64/mops_invalid_2.d, testsuite/gas/aarch64/mops_invalid_2.l: New test.
2021-12-02 23:00:57 +08:00
mismatch_detail->data[0].i = alignment;
}
static inline void
set_reg_list_length_error (aarch64_operand_error *mismatch_detail, int idx,
int expected_num)
{
if (mismatch_detail == NULL)
return;
set_error (mismatch_detail, AARCH64_OPDE_REG_LIST_LENGTH, idx, NULL);
mismatch_detail->data[0].i = 1 << expected_num;
}
static inline void
set_reg_list_stride_error (aarch64_operand_error *mismatch_detail, int idx,
int expected_num)
{
if (mismatch_detail == NULL)
return;
set_error (mismatch_detail, AARCH64_OPDE_REG_LIST_STRIDE, idx, NULL);
mismatch_detail->data[0].i = 1 << expected_num;
}
static inline void
set_invalid_vg_size (aarch64_operand_error *mismatch_detail,
int idx, int expected)
{
if (mismatch_detail == NULL)
return;
set_error (mismatch_detail, AARCH64_OPDE_INVALID_VG_SIZE, idx, NULL);
mismatch_detail->data[0].i = expected;
}
static inline void
set_other_error (aarch64_operand_error *mismatch_detail, int idx,
const char* error)
{
if (mismatch_detail == NULL)
return;
set_error (mismatch_detail, AARCH64_OPDE_OTHER_ERROR, idx, error);
}
/* Check that indexed register operand OPND has a register in the range
[MIN_REGNO, MAX_REGNO] and an index in the range [MIN_INDEX, MAX_INDEX].
PREFIX is the register prefix, such as "z" for SVE vector registers. */
static bool
check_reglane (const aarch64_opnd_info *opnd,
aarch64_operand_error *mismatch_detail, int idx,
const char *prefix, int min_regno, int max_regno,
int min_index, int max_index)
{
if (!value_in_range_p (opnd->reglane.regno, min_regno, max_regno))
{
set_invalid_regno_error (mismatch_detail, idx, prefix, min_regno,
max_regno);
return false;
}
if (!value_in_range_p (opnd->reglane.index, min_index, max_index))
{
set_elem_idx_out_of_range_error (mismatch_detail, idx, min_index,
max_index);
return false;
}
return true;
}
/* Check that register list operand OPND has NUM_REGS registers and a
register stride of STRIDE. */
static bool
check_reglist (const aarch64_opnd_info *opnd,
aarch64_operand_error *mismatch_detail, int idx,
int num_regs, int stride)
{
if (opnd->reglist.num_regs != num_regs)
{
set_reg_list_length_error (mismatch_detail, idx, num_regs);
return false;
}
if (opnd->reglist.stride != stride)
{
set_reg_list_stride_error (mismatch_detail, idx, stride);
return false;
}
return true;
}
/* Check that indexed ZA operand OPND has:
- a selection register in the range [MIN_WREG, MIN_WREG + 3]
- RANGE_SIZE consecutive immediate offsets.
- an initial immediate offset that is a multiple of RANGE_SIZE
in the range [0, MAX_VALUE * RANGE_SIZE]
- a vector group size of GROUP_SIZE. */
static bool
check_za_access (const aarch64_opnd_info *opnd,
aarch64_operand_error *mismatch_detail, int idx,
int min_wreg, int max_value, unsigned int range_size,
int group_size)
{
if (!value_in_range_p (opnd->indexed_za.index.regno, min_wreg, min_wreg + 3))
{
if (min_wreg == 12)
set_other_error (mismatch_detail, idx,
_("expected a selection register in the"
" range w12-w15"));
else if (min_wreg == 8)
set_other_error (mismatch_detail, idx,
_("expected a selection register in the"
" range w8-w11"));
else
abort ();
return false;
}
int max_index = max_value * range_size;
if (!value_in_range_p (opnd->indexed_za.index.imm, 0, max_index))
{
set_offset_out_of_range_error (mismatch_detail, idx, 0, max_index);
return false;
}
if ((opnd->indexed_za.index.imm % range_size) != 0)
{
assert (range_size == 2 || range_size == 4);
set_other_error (mismatch_detail, idx,
range_size == 2
? _("starting offset is not a multiple of 2")
: _("starting offset is not a multiple of 4"));
return false;
}
if (opnd->indexed_za.index.countm1 != range_size - 1)
{
if (range_size == 1)
set_other_error (mismatch_detail, idx,
_("expected a single offset rather than"
" a range"));
else if (range_size == 2)
set_other_error (mismatch_detail, idx,
_("expected a range of two offsets"));
else if (range_size == 4)
set_other_error (mismatch_detail, idx,
_("expected a range of four offsets"));
else
abort ();
return false;
}
/* The vector group specifier is optional in assembly code. */
if (opnd->indexed_za.group_size != 0
&& opnd->indexed_za.group_size != group_size)
{
set_invalid_vg_size (mismatch_detail, idx, group_size);
return false;
}
return true;
}
/* Given a load/store operation, calculate the size of transferred data via a
cumulative sum of qualifier sizes preceding the address operand in the
OPNDS operand list argument. */
int
calc_ldst_datasize (const aarch64_opnd_info *opnds)
{
unsigned num_bytes = 0; /* total number of bytes transferred. */
enum aarch64_operand_class opnd_class;
enum aarch64_opnd type;
for (int i = 0; i < AARCH64_MAX_OPND_NUM; i++)
{
type = opnds[i].type;
opnd_class = aarch64_operands[type].op_class;
if (opnd_class == AARCH64_OPND_CLASS_ADDRESS)
break;
num_bytes += aarch64_get_qualifier_esize (opnds[i].qualifier);
}
return num_bytes;
}
/* General constraint checking based on operand code.
Return 1 if OPNDS[IDX] meets the general constraint of operand code TYPE
as the IDXth operand of opcode OPCODE. Otherwise return 0.
This function has to be called after the qualifiers for all operands
have been resolved.
Mismatching error message is returned in *MISMATCH_DETAIL upon request,
i.e. when MISMATCH_DETAIL is non-NULL. This avoids the generation
of error message during the disassembling where error message is not
wanted. We avoid the dynamic construction of strings of error messages
here (i.e. in libopcodes), as it is costly and complicated; instead, we
use a combination of error code, static string and some integer data to
represent an error. */
static int
operand_general_constraint_met_p (const aarch64_opnd_info *opnds, int idx,
enum aarch64_opnd type,
const aarch64_opcode *opcode,
aarch64_operand_error *mismatch_detail)
{
[AArch64][SVE 27/32] Add SVE integer immediate operands This patch adds the new SVE integer immediate operands. There are three kinds: - simple signed and unsigned ranges, but with new widths and positions. - 13-bit logical immediates. These have the same form as in base AArch64, but at a different bit position. In the case of the "MOV Zn.<T>, #<limm>" alias of DUPM, the logical immediate <limm> is not allowed to be a valid DUP immediate, since DUP is preferred over DUPM for constants that both instructions can handle. - a new 9-bit arithmetic immediate, of the form "<imm8>{, LSL #8}". In some contexts the operand is signed and in others it's unsigned. As an extension, we allow shifted immediates to be written as a single integer, e.g. "#256" is equivalent to "#1, LSL #8". We also use the shiftless form as the preferred disassembly, except for the special case of "#0, LSL #8" (a redundant encoding of 0). include/ * opcode/aarch64.h (AARCH64_OPND_SIMM5): New aarch64_opnd. (AARCH64_OPND_SVE_AIMM, AARCH64_OPND_SVE_ASIMM) (AARCH64_OPND_SVE_INV_LIMM, AARCH64_OPND_SVE_LIMM) (AARCH64_OPND_SVE_LIMM_MOV, AARCH64_OPND_SVE_SHLIMM_PRED) (AARCH64_OPND_SVE_SHLIMM_UNPRED, AARCH64_OPND_SVE_SHRIMM_PRED) (AARCH64_OPND_SVE_SHRIMM_UNPRED, AARCH64_OPND_SVE_SIMM5) (AARCH64_OPND_SVE_SIMM5B, AARCH64_OPND_SVE_SIMM6) (AARCH64_OPND_SVE_SIMM8, AARCH64_OPND_SVE_UIMM3) (AARCH64_OPND_SVE_UIMM7, AARCH64_OPND_SVE_UIMM8) (AARCH64_OPND_SVE_UIMM8_53): Likewise. (aarch64_sve_dupm_mov_immediate_p): Declare. opcodes/ * aarch64-tbl.h (AARCH64_OPERANDS): Add entries for the new SVE integer immediate operands. * aarch64-opc.h (FLD_SVE_immN, FLD_SVE_imm3, FLD_SVE_imm5) (FLD_SVE_imm5b, FLD_SVE_imm7, FLD_SVE_imm8, FLD_SVE_imm9) (FLD_SVE_immr, FLD_SVE_imms, FLD_SVE_tszh): New aarch64_field_kinds. * aarch64-opc.c (fields): Add corresponding entries. (operand_general_constraint_met_p): Handle the new SVE integer immediate operands. (aarch64_print_operand): Likewise. (aarch64_sve_dupm_mov_immediate_p): New function. * aarch64-opc-2.c: Regenerate. * aarch64-asm.h (ins_inv_limm, ins_sve_aimm, ins_sve_asimm) (ins_sve_limm_mov, ins_sve_shlimm, ins_sve_shrimm): New inserters. * aarch64-asm.c (aarch64_ins_limm_1): New function, split out from... (aarch64_ins_limm): ...here. (aarch64_ins_inv_limm): New function. (aarch64_ins_sve_aimm): Likewise. (aarch64_ins_sve_asimm): Likewise. (aarch64_ins_sve_limm_mov): Likewise. (aarch64_ins_sve_shlimm): Likewise. (aarch64_ins_sve_shrimm): Likewise. * aarch64-asm-2.c: Regenerate. * aarch64-dis.h (ext_inv_limm, ext_sve_aimm, ext_sve_asimm) (ext_sve_limm_mov, ext_sve_shlimm, ext_sve_shrimm): New extractors. * aarch64-dis.c (decode_limm): New function, split out from... (aarch64_ext_limm): ...here. (aarch64_ext_inv_limm): New function. (decode_sve_aimm): Likewise. (aarch64_ext_sve_aimm): Likewise. (aarch64_ext_sve_asimm): Likewise. (aarch64_ext_sve_limm_mov): Likewise. (aarch64_top_bit): Likewise. (aarch64_ext_sve_shlimm): Likewise. (aarch64_ext_sve_shrimm): Likewise. * aarch64-dis-2.c: Regenerate. gas/ * config/tc-aarch64.c (parse_operands): Handle the new SVE integer immediate operands.
2016-09-21 23:56:57 +08:00
unsigned num, modifiers, shift;
unsigned char size;
[AArch64][SVE 25/32] Add support for SVE addressing modes This patch adds most of the new SVE addressing modes and associated operands. A follow-on patch adds MUL VL, since handling it separately makes the changes easier to read. The patch also introduces a new "operand-dependent data" field to the operand flags, based closely on the existing one for opcode flags. For SVE this new field needs only 2 bits, but it could be widened in future if necessary. include/ * opcode/aarch64.h (AARCH64_OPND_SVE_ADDR_RI_U6): New aarch64_opnd. (AARCH64_OPND_SVE_ADDR_RI_U6x2, AARCH64_OPND_SVE_ADDR_RI_U6x4) (AARCH64_OPND_SVE_ADDR_RI_U6x8, AARCH64_OPND_SVE_ADDR_RR) (AARCH64_OPND_SVE_ADDR_RR_LSL1, AARCH64_OPND_SVE_ADDR_RR_LSL2) (AARCH64_OPND_SVE_ADDR_RR_LSL3, AARCH64_OPND_SVE_ADDR_RX) (AARCH64_OPND_SVE_ADDR_RX_LSL1, AARCH64_OPND_SVE_ADDR_RX_LSL2) (AARCH64_OPND_SVE_ADDR_RX_LSL3, AARCH64_OPND_SVE_ADDR_RZ) (AARCH64_OPND_SVE_ADDR_RZ_LSL1, AARCH64_OPND_SVE_ADDR_RZ_LSL2) (AARCH64_OPND_SVE_ADDR_RZ_LSL3, AARCH64_OPND_SVE_ADDR_RZ_XTW_14) (AARCH64_OPND_SVE_ADDR_RZ_XTW_22, AARCH64_OPND_SVE_ADDR_RZ_XTW1_14) (AARCH64_OPND_SVE_ADDR_RZ_XTW1_22, AARCH64_OPND_SVE_ADDR_RZ_XTW2_14) (AARCH64_OPND_SVE_ADDR_RZ_XTW2_22, AARCH64_OPND_SVE_ADDR_RZ_XTW3_14) (AARCH64_OPND_SVE_ADDR_RZ_XTW3_22, AARCH64_OPND_SVE_ADDR_ZI_U5) (AARCH64_OPND_SVE_ADDR_ZI_U5x2, AARCH64_OPND_SVE_ADDR_ZI_U5x4) (AARCH64_OPND_SVE_ADDR_ZI_U5x8, AARCH64_OPND_SVE_ADDR_ZZ_LSL) (AARCH64_OPND_SVE_ADDR_ZZ_SXTW, AARCH64_OPND_SVE_ADDR_ZZ_UXTW): Likewise. opcodes/ * aarch64-tbl.h (AARCH64_OPERANDS): Add entries for the new SVE address operands. * aarch64-opc.h (FLD_SVE_imm6, FLD_SVE_msz, FLD_SVE_xs_14) (FLD_SVE_xs_22): New aarch64_field_kinds. (OPD_F_OD_MASK, OPD_F_OD_LSB, OPD_F_NO_ZR): New flags. (get_operand_specific_data): New function. * aarch64-opc.c (fields): Add entries for FLD_SVE_imm6, FLD_SVE_msz, FLD_SVE_xs_14 and FLD_SVE_xs_22. (operand_general_constraint_met_p): Handle the new SVE address operands. (sve_reg): New array. (get_addr_sve_reg_name): New function. (aarch64_print_operand): Handle the new SVE address operands. * aarch64-opc-2.c: Regenerate. * aarch64-asm.h (ins_sve_addr_ri_u6, ins_sve_addr_rr_lsl) (ins_sve_addr_rz_xtw, ins_sve_addr_zi_u5, ins_sve_addr_zz_lsl) (ins_sve_addr_zz_sxtw, ins_sve_addr_zz_uxtw): New inserters. * aarch64-asm.c (aarch64_ins_sve_addr_ri_u6): New function. (aarch64_ins_sve_addr_rr_lsl): Likewise. (aarch64_ins_sve_addr_rz_xtw): Likewise. (aarch64_ins_sve_addr_zi_u5): Likewise. (aarch64_ins_sve_addr_zz): Likewise. (aarch64_ins_sve_addr_zz_lsl): Likewise. (aarch64_ins_sve_addr_zz_sxtw): Likewise. (aarch64_ins_sve_addr_zz_uxtw): Likewise. * aarch64-asm-2.c: Regenerate. * aarch64-dis.h (ext_sve_addr_ri_u6, ext_sve_addr_rr_lsl) (ext_sve_addr_rz_xtw, ext_sve_addr_zi_u5, ext_sve_addr_zz_lsl) (ext_sve_addr_zz_sxtw, ext_sve_addr_zz_uxtw): New extractors. * aarch64-dis.c (aarch64_ext_sve_add_reg_imm): New function. (aarch64_ext_sve_addr_ri_u6): Likewise. (aarch64_ext_sve_addr_rr_lsl): Likewise. (aarch64_ext_sve_addr_rz_xtw): Likewise. (aarch64_ext_sve_addr_zi_u5): Likewise. (aarch64_ext_sve_addr_zz): Likewise. (aarch64_ext_sve_addr_zz_lsl): Likewise. (aarch64_ext_sve_addr_zz_sxtw): Likewise. (aarch64_ext_sve_addr_zz_uxtw): Likewise. * aarch64-dis-2.c: Regenerate. gas/ * config/tc-aarch64.c (REG_TYPE_SVE_BASE, REG_TYPE_SVE_OFFSET): New register types. (get_reg_expected_msg): Handle them. (aarch64_addr_reg_parse): New function, split out from aarch64_reg_parse_32_64. Handle Z registers too. (aarch64_reg_parse_32_64): Call it. (parse_address_main): Add base_qualifier, offset_qualifier, base_type and offset_type parameters. Handle SVE base and offset registers. (parse_address): Update call to parse_address_main. (parse_sve_address): New function. (parse_operands): Parse the new SVE address operands.
2016-09-21 23:55:49 +08:00
int64_t imm, min_value, max_value;
[AArch64][SVE 27/32] Add SVE integer immediate operands This patch adds the new SVE integer immediate operands. There are three kinds: - simple signed and unsigned ranges, but with new widths and positions. - 13-bit logical immediates. These have the same form as in base AArch64, but at a different bit position. In the case of the "MOV Zn.<T>, #<limm>" alias of DUPM, the logical immediate <limm> is not allowed to be a valid DUP immediate, since DUP is preferred over DUPM for constants that both instructions can handle. - a new 9-bit arithmetic immediate, of the form "<imm8>{, LSL #8}". In some contexts the operand is signed and in others it's unsigned. As an extension, we allow shifted immediates to be written as a single integer, e.g. "#256" is equivalent to "#1, LSL #8". We also use the shiftless form as the preferred disassembly, except for the special case of "#0, LSL #8" (a redundant encoding of 0). include/ * opcode/aarch64.h (AARCH64_OPND_SIMM5): New aarch64_opnd. (AARCH64_OPND_SVE_AIMM, AARCH64_OPND_SVE_ASIMM) (AARCH64_OPND_SVE_INV_LIMM, AARCH64_OPND_SVE_LIMM) (AARCH64_OPND_SVE_LIMM_MOV, AARCH64_OPND_SVE_SHLIMM_PRED) (AARCH64_OPND_SVE_SHLIMM_UNPRED, AARCH64_OPND_SVE_SHRIMM_PRED) (AARCH64_OPND_SVE_SHRIMM_UNPRED, AARCH64_OPND_SVE_SIMM5) (AARCH64_OPND_SVE_SIMM5B, AARCH64_OPND_SVE_SIMM6) (AARCH64_OPND_SVE_SIMM8, AARCH64_OPND_SVE_UIMM3) (AARCH64_OPND_SVE_UIMM7, AARCH64_OPND_SVE_UIMM8) (AARCH64_OPND_SVE_UIMM8_53): Likewise. (aarch64_sve_dupm_mov_immediate_p): Declare. opcodes/ * aarch64-tbl.h (AARCH64_OPERANDS): Add entries for the new SVE integer immediate operands. * aarch64-opc.h (FLD_SVE_immN, FLD_SVE_imm3, FLD_SVE_imm5) (FLD_SVE_imm5b, FLD_SVE_imm7, FLD_SVE_imm8, FLD_SVE_imm9) (FLD_SVE_immr, FLD_SVE_imms, FLD_SVE_tszh): New aarch64_field_kinds. * aarch64-opc.c (fields): Add corresponding entries. (operand_general_constraint_met_p): Handle the new SVE integer immediate operands. (aarch64_print_operand): Likewise. (aarch64_sve_dupm_mov_immediate_p): New function. * aarch64-opc-2.c: Regenerate. * aarch64-asm.h (ins_inv_limm, ins_sve_aimm, ins_sve_asimm) (ins_sve_limm_mov, ins_sve_shlimm, ins_sve_shrimm): New inserters. * aarch64-asm.c (aarch64_ins_limm_1): New function, split out from... (aarch64_ins_limm): ...here. (aarch64_ins_inv_limm): New function. (aarch64_ins_sve_aimm): Likewise. (aarch64_ins_sve_asimm): Likewise. (aarch64_ins_sve_limm_mov): Likewise. (aarch64_ins_sve_shlimm): Likewise. (aarch64_ins_sve_shrimm): Likewise. * aarch64-asm-2.c: Regenerate. * aarch64-dis.h (ext_inv_limm, ext_sve_aimm, ext_sve_asimm) (ext_sve_limm_mov, ext_sve_shlimm, ext_sve_shrimm): New extractors. * aarch64-dis.c (decode_limm): New function, split out from... (aarch64_ext_limm): ...here. (aarch64_ext_inv_limm): New function. (decode_sve_aimm): Likewise. (aarch64_ext_sve_aimm): Likewise. (aarch64_ext_sve_asimm): Likewise. (aarch64_ext_sve_limm_mov): Likewise. (aarch64_top_bit): Likewise. (aarch64_ext_sve_shlimm): Likewise. (aarch64_ext_sve_shrimm): Likewise. * aarch64-dis-2.c: Regenerate. gas/ * config/tc-aarch64.c (parse_operands): Handle the new SVE integer immediate operands.
2016-09-21 23:56:57 +08:00
uint64_t uvalue, mask;
const aarch64_opnd_info *opnd = opnds + idx;
aarch64_opnd_qualifier_t qualifier = opnd->qualifier;
aarch64: Add maximum immediate value to aarch64_sys_reg The immediate form of MSR has a 4-bit immediate field (in CRm). However, many forms of MSR require a smaller immediate. These cases are identified by value in operand_general_constraint_met_p, but they're now the common case rather than the exception. This patch therefore adds the maximum value to the sys_reg description and gets the range from there. It also enforces the minimum of 0, which avoids a situation in which: msr dit, #2 would give the expected: Error: immediate value out of range 0 to 1 whereas: msr dit, #-1 would give: Error: immediate value out of range 0 to 15 (from the later UIMM4 checking). Also: - we were reporting the first error above against the wrong operand - TCO takes a single-bit immediate, but we previously allowed all 16 values. [https://developer.arm.com/documentation/ddi0596/2021-09/Base-Instructions/MSR--immediate---Move-immediate-value-to-Special-Register-?lang=en] opcodes/ * aarch64-opc.h (F_REG_MAX_VALUE, F_GET_REG_MAX_VALUE): New macros. * aarch64-opc.c (operand_general_constraint_met_p): Read the maximum MSR immediate value from aarch64_pstatefields. (aarch64_pstatefields): Add the maximum immediate value for each register. gas/ * testsuite/gas/aarch64/sysreg-4.s: Use an immediate value of 1 rather than 8 for the TCO test. * testsuite/gas/aarch64/sysreg-4.d: Update accordingly. * testsuite/gas/aarch64/armv8_2-a-illegal.l: Fix operand number in MSR immediate error messages. * testsuite/gas/aarch64/diagnostic.l: Likewise. * testsuite/gas/aarch64/pan-illegal.l: Likewise. * testsuite/gas/aarch64/ssbs-illegal1.l: Likewise. * testsuite/gas/aarch64/illegal-sysreg-4b.s, * testsuite/gas/aarch64/illegal-sysreg-4b.d, * testsuite/gas/aarch64/illegal-sysreg-4b.l: New test.
2021-12-02 23:00:56 +08:00
int i;
assert (opcode->operands[idx] == opnd->type && opnd->type == type);
switch (aarch64_operands[type].op_class)
{
case AARCH64_OPND_CLASS_INT_REG:
/* Check for pair of xzr registers. */
if (type == AARCH64_OPND_PAIRREG_OR_XZR
&& opnds[idx - 1].reg.regno == 0x1f)
{
if (opnds[idx].reg.regno != 0x1f)
{
set_syntax_error (mismatch_detail, idx - 1,
_("second reg in pair should be xzr if first is"
" xzr"));
return 0;
}
}
/* Check pair reg constraints for instructions taking a pair of
consecutively-numbered general-purpose registers. */
else if (type == AARCH64_OPND_PAIRREG
|| type == AARCH64_OPND_PAIRREG_OR_XZR)
{
assert (idx == 1 || idx == 2 || idx == 3 || idx == 5);
if (opnds[idx - 1].reg.regno % 2 != 0)
{
set_syntax_error (mismatch_detail, idx - 1,
_("reg pair must start from even reg"));
return 0;
}
if (opnds[idx].reg.regno != opnds[idx - 1].reg.regno + 1)
{
set_syntax_error (mismatch_detail, idx,
_("reg pair must be contiguous"));
return 0;
}
break;
}
/* <Xt> may be optional in some IC and TLBI instructions. */
if (type == AARCH64_OPND_Rt_SYS)
{
assert (idx == 1 && (aarch64_get_operand_class (opnds[0].type)
== AARCH64_OPND_CLASS_SYSTEM));
if (opnds[1].present
&& !aarch64_sys_ins_reg_has_xt (opnds[0].sysins_op))
{
set_other_error (mismatch_detail, idx, _("extraneous register"));
return 0;
}
if (!opnds[1].present
&& aarch64_sys_ins_reg_has_xt (opnds[0].sysins_op))
{
set_other_error (mismatch_detail, idx, _("missing register"));
return 0;
}
}
switch (qualifier)
{
case AARCH64_OPND_QLF_WSP:
case AARCH64_OPND_QLF_SP:
if (!aarch64_stack_pointer_p (opnd))
{
set_other_error (mismatch_detail, idx,
aarch64: [SME] Add SME mode selection and state access instructions This patch is adding new SME mode selection and state access instructions: * Add SMSTART and SMSTOP instructions. * Add SVCR system register. gas/ChangeLog: * config/tc-aarch64.c (parse_sme_sm_za): New parser. (parse_operands): New parser. * testsuite/gas/aarch64/sme-8-illegal.d: New test. * testsuite/gas/aarch64/sme-8-illegal.l: New test. * testsuite/gas/aarch64/sme-8-illegal.s: New test. * testsuite/gas/aarch64/sme-8.d: New test. * testsuite/gas/aarch64/sme-8.s: New test. include/ChangeLog: * opcode/aarch64.h (enum aarch64_opnd): New operand AARCH64_OPND_SME_SM_ZA. (enum aarch64_insn_class): New instruction classes sme_start and sme_stop. opcodes/ChangeLog: * aarch64-asm.c (aarch64_ins_pstatefield): New inserter. (aarch64_ins_sme_sm_za): New inserter. * aarch64-dis.c (aarch64_ext_imm): New extractor. (aarch64_ext_pstatefield): New extractor. (aarch64_ext_sme_sm_za): New extractor. * aarch64-opc.c (operand_general_constraint_met_p): New pstatefield value for SME instructions. (aarch64_print_operand): Printout for OPND_SME_SM_ZA. (SR_SME): New register SVCR. * aarch64-opc.h (F_REG_IN_CRM): New register endcoding. * aarch64-opc.h (F_IMM_IN_CRM): New immediate endcoding. (PSTATE_ENCODE_CRM): Encode CRm field. (PSTATE_DECODE_CRM): Decode CRm field. (PSTATE_ENCODE_CRM_IMM): Encode CRm immediate field. (PSTATE_DECODE_CRM_IMM): Decode CRm immediate field. (PSTATE_ENCODE_CRM_AND_IMM): Encode CRm and immediate field. * aarch64-tbl.h (struct aarch64_opcode): New SMSTART and SMSTOP instructions. aarch64-asm-2.c: Regenerate. aarch64-dis-2.c: Regenerate. aarch64-opc-2.c: Regenerate.
2021-11-18 04:15:13 +08:00
_("stack pointer register expected"));
return 0;
}
break;
default:
break;
}
break;
[AArch64][SVE 21/32] Add Zn and Pn registers This patch adds the Zn and Pn registers, and associated fields and operands. include/ * opcode/aarch64.h (AARCH64_OPND_CLASS_SVE_REG): New aarch64_operand_class. (AARCH64_OPND_CLASS_PRED_REG): Likewise. (AARCH64_OPND_SVE_Pd, AARCH64_OPND_SVE_Pg3, AARCH64_OPND_SVE_Pg4_5) (AARCH64_OPND_SVE_Pg4_10, AARCH64_OPND_SVE_Pg4_16) (AARCH64_OPND_SVE_Pm, AARCH64_OPND_SVE_Pn, AARCH64_OPND_SVE_Pt) (AARCH64_OPND_SVE_Za_5, AARCH64_OPND_SVE_Za_16, AARCH64_OPND_SVE_Zd) (AARCH64_OPND_SVE_Zm_5, AARCH64_OPND_SVE_Zm_16, AARCH64_OPND_SVE_Zn) (AARCH64_OPND_SVE_Zn_INDEX, AARCH64_OPND_SVE_ZnxN) (AARCH64_OPND_SVE_Zt, AARCH64_OPND_SVE_ZtxN): New aarch64_opnds. opcodes/ * aarch64-tbl.h (AARCH64_OPERANDS): Add entries for new SVE operands. * aarch64-opc.h (FLD_SVE_Pd, FLD_SVE_Pg3, FLD_SVE_Pg4_5) (FLD_SVE_Pg4_10, FLD_SVE_Pg4_16, FLD_SVE_Pm, FLD_SVE_Pn, FLD_SVE_Pt) (FLD_SVE_Za_5, FLD_SVE_Za_16, FLD_SVE_Zd, FLD_SVE_Zm_5, FLD_SVE_Zm_16) (FLD_SVE_Zn, FLD_SVE_Zt, FLD_SVE_tzsh): New aarch64_field_kinds. * aarch64-opc.c (fields): Add corresponding entries here. (operand_general_constraint_met_p): Check that SVE register lists have the correct length. Check the ranges of SVE index registers. Check for cases where p8-p15 are used in 3-bit predicate fields. (aarch64_print_operand): Handle the new SVE operands. * aarch64-opc-2.c: Regenerate. * aarch64-asm.h (ins_sve_index, ins_sve_reglist): New inserters. * aarch64-asm.c (aarch64_ins_sve_index): New function. (aarch64_ins_sve_reglist): Likewise. * aarch64-asm-2.c: Regenerate. * aarch64-dis.h (ext_sve_index, ext_sve_reglist): New extractors. * aarch64-dis.c (aarch64_ext_sve_index): New function. (aarch64_ext_sve_reglist): Likewise. * aarch64-dis-2.c: Regenerate. gas/ * config/tc-aarch64.c (NTA_HASVARWIDTH): New macro. (AARCH64_REG_TYPES): Add ZN and PN. (get_reg_expected_msg): Handle them. (parse_vector_type_for_operand): Add a reg_type parameter. Skip the width for Zn and Pn registers. (parse_typed_reg): Extend vector handling to Zn and Pn. Update the call to parse_vector_type_for_operand. Set HASVARTYPE for Zn and Pn, expecting the width to be 0. (parse_vector_reg_list): Restrict error about [BHSD]nn operands to REG_TYPE_VN. (vectype_to_qualifier): Use S_[BHSD] qualifiers for NTA_HASVARWIDTH. (parse_operands): Handle the new Zn and Pn operands. (REGSET16): New macro, split out from... (REGSET31): ...here. (reg_names): Add Zn and Pn entries.
2016-09-21 23:53:54 +08:00
case AARCH64_OPND_CLASS_SVE_REG:
switch (type)
{
[AArch64] Additional SVE instructions This patch supports some additions to the SVE architecture prior to its public release. include/ * opcode/aarch64.h (AARCH64_OPND_SVE_ADDR_RI_S4x16) (AARCH64_OPND_SVE_IMM_ROT1, AARCH64_OPND_SVE_IMM_ROT2) (AARCH64_OPND_SVE_Zm3_INDEX, AARCH64_OPND_SVE_Zm3_22_INDEX) (AARCH64_OPND_SVE_Zm4_INDEX): New aarch64_opnds. opcodes/ * aarch64-tbl.h (OP_SVE_HMH, OP_SVE_VMU_HSD, OP_SVE_VMVU_HSD) (OP_SVE_VMVV_HSD, OP_SVE_VMVVU_HSD, OP_SVE_VM_HSD, OP_SVE_VUVV_HSD) (OP_SVE_VUV_HSD, OP_SVE_VU_HSD, OP_SVE_VVVU_H, OP_SVE_VVVU_S) (OP_SVE_VVVU_HSD, OP_SVE_VVV_D, OP_SVE_VVV_D_H, OP_SVE_VVV_H) (OP_SVE_VVV_HSD, OP_SVE_VVV_S, OP_SVE_VVV_S_B, OP_SVE_VVV_SD_BH) (OP_SVE_VV_BHSDQ, OP_SVE_VV_HSD, OP_SVE_VZVV_HSD, OP_SVE_VZV_HSD) (OP_SVE_V_HSD): New macros. (OP_SVE_VMU_SD, OP_SVE_VMVU_SD, OP_SVE_VM_SD, OP_SVE_VUVV_SD) (OP_SVE_VU_SD, OP_SVE_VVVU_SD, OP_SVE_VVV_SD, OP_SVE_VZVV_SD) (OP_SVE_VZV_SD, OP_SVE_V_SD): Delete. (aarch64_opcode_table): Add new SVE instructions. (aarch64_opcode_table): Use imm_rotate{1,2} instead of imm_rotate for rotation operands. Add new SVE operands. * aarch64-asm.h (ins_sve_addr_ri_s4): New inserter. (ins_sve_quad_index): Likewise. (ins_imm_rotate): Split into... (ins_imm_rotate1, ins_imm_rotate2): ...these two inserters. * aarch64-asm.c (aarch64_ins_imm_rotate): Split into... (aarch64_ins_imm_rotate1, aarch64_ins_imm_rotate2): ...these two functions. (aarch64_ins_sve_addr_ri_s4): New function. (aarch64_ins_sve_quad_index): Likewise. (do_misc_encoding): Handle "MOV Zn.Q, Qm". * aarch64-asm-2.c: Regenerate. * aarch64-dis.h (ext_sve_addr_ri_s4): New extractor. (ext_sve_quad_index): Likewise. (ext_imm_rotate): Split into... (ext_imm_rotate1, ext_imm_rotate2): ...these two extractors. * aarch64-dis.c (aarch64_ext_imm_rotate): Split into... (aarch64_ext_imm_rotate1, aarch64_ext_imm_rotate2): ...these two functions. (aarch64_ext_sve_addr_ri_s4): New function. (aarch64_ext_sve_quad_index): Likewise. (aarch64_ext_sve_index): Allow quad indices. (do_misc_decoding): Likewise. * aarch64-dis-2.c: Regenerate. * aarch64-opc.h (FLD_SVE_i3h, FLD_SVE_rot1, FLD_SVE_rot2): New aarch64_field_kinds. (OPD_F_OD_MASK): Widen by one bit. (OPD_F_NO_ZR): Bump accordingly. (get_operand_field_width): New function. * aarch64-opc.c (fields): Add new SVE fields. (operand_general_constraint_met_p): Handle new SVE operands. (aarch64_print_operand): Likewise. * aarch64-opc-2.c: Regenerate. gas/ * doc/c-aarch64.texi: Document that sve implies fp16, simd and compnum. * config/tc-aarch64.c (parse_vector_type_for_operand): Allow .q to be used with SVE registers. (parse_operands): Handle new SVE operands. (aarch64_features): Make "sve" require F16 rather than FP. Also require COMPNUM. * testsuite/gas/aarch64/sve.s: Add tests for new instructions. Include compnum tests. * testsuite/gas/aarch64/sve.d: Update accordingly. * testsuite/gas/aarch64/sve-invalid.s: Add tests for new instructions. * testsuite/gas/aarch64/sve-invalid.l: Update accordingly. Also update expected output for new FMOV and MOV alternatives.
2017-02-25 02:29:00 +08:00
case AARCH64_OPND_SVE_Zm3_INDEX:
case AARCH64_OPND_SVE_Zm3_22_INDEX:
case AARCH64_OPND_SVE_Zm3_19_INDEX:
case AARCH64_OPND_SVE_Zm3_11_INDEX:
case AARCH64_OPND_SVE_Zm4_11_INDEX:
[AArch64] Additional SVE instructions This patch supports some additions to the SVE architecture prior to its public release. include/ * opcode/aarch64.h (AARCH64_OPND_SVE_ADDR_RI_S4x16) (AARCH64_OPND_SVE_IMM_ROT1, AARCH64_OPND_SVE_IMM_ROT2) (AARCH64_OPND_SVE_Zm3_INDEX, AARCH64_OPND_SVE_Zm3_22_INDEX) (AARCH64_OPND_SVE_Zm4_INDEX): New aarch64_opnds. opcodes/ * aarch64-tbl.h (OP_SVE_HMH, OP_SVE_VMU_HSD, OP_SVE_VMVU_HSD) (OP_SVE_VMVV_HSD, OP_SVE_VMVVU_HSD, OP_SVE_VM_HSD, OP_SVE_VUVV_HSD) (OP_SVE_VUV_HSD, OP_SVE_VU_HSD, OP_SVE_VVVU_H, OP_SVE_VVVU_S) (OP_SVE_VVVU_HSD, OP_SVE_VVV_D, OP_SVE_VVV_D_H, OP_SVE_VVV_H) (OP_SVE_VVV_HSD, OP_SVE_VVV_S, OP_SVE_VVV_S_B, OP_SVE_VVV_SD_BH) (OP_SVE_VV_BHSDQ, OP_SVE_VV_HSD, OP_SVE_VZVV_HSD, OP_SVE_VZV_HSD) (OP_SVE_V_HSD): New macros. (OP_SVE_VMU_SD, OP_SVE_VMVU_SD, OP_SVE_VM_SD, OP_SVE_VUVV_SD) (OP_SVE_VU_SD, OP_SVE_VVVU_SD, OP_SVE_VVV_SD, OP_SVE_VZVV_SD) (OP_SVE_VZV_SD, OP_SVE_V_SD): Delete. (aarch64_opcode_table): Add new SVE instructions. (aarch64_opcode_table): Use imm_rotate{1,2} instead of imm_rotate for rotation operands. Add new SVE operands. * aarch64-asm.h (ins_sve_addr_ri_s4): New inserter. (ins_sve_quad_index): Likewise. (ins_imm_rotate): Split into... (ins_imm_rotate1, ins_imm_rotate2): ...these two inserters. * aarch64-asm.c (aarch64_ins_imm_rotate): Split into... (aarch64_ins_imm_rotate1, aarch64_ins_imm_rotate2): ...these two functions. (aarch64_ins_sve_addr_ri_s4): New function. (aarch64_ins_sve_quad_index): Likewise. (do_misc_encoding): Handle "MOV Zn.Q, Qm". * aarch64-asm-2.c: Regenerate. * aarch64-dis.h (ext_sve_addr_ri_s4): New extractor. (ext_sve_quad_index): Likewise. (ext_imm_rotate): Split into... (ext_imm_rotate1, ext_imm_rotate2): ...these two extractors. * aarch64-dis.c (aarch64_ext_imm_rotate): Split into... (aarch64_ext_imm_rotate1, aarch64_ext_imm_rotate2): ...these two functions. (aarch64_ext_sve_addr_ri_s4): New function. (aarch64_ext_sve_quad_index): Likewise. (aarch64_ext_sve_index): Allow quad indices. (do_misc_decoding): Likewise. * aarch64-dis-2.c: Regenerate. * aarch64-opc.h (FLD_SVE_i3h, FLD_SVE_rot1, FLD_SVE_rot2): New aarch64_field_kinds. (OPD_F_OD_MASK): Widen by one bit. (OPD_F_NO_ZR): Bump accordingly. (get_operand_field_width): New function. * aarch64-opc.c (fields): Add new SVE fields. (operand_general_constraint_met_p): Handle new SVE operands. (aarch64_print_operand): Likewise. * aarch64-opc-2.c: Regenerate. gas/ * doc/c-aarch64.texi: Document that sve implies fp16, simd and compnum. * config/tc-aarch64.c (parse_vector_type_for_operand): Allow .q to be used with SVE registers. (parse_operands): Handle new SVE operands. (aarch64_features): Make "sve" require F16 rather than FP. Also require COMPNUM. * testsuite/gas/aarch64/sve.s: Add tests for new instructions. Include compnum tests. * testsuite/gas/aarch64/sve.d: Update accordingly. * testsuite/gas/aarch64/sve-invalid.s: Add tests for new instructions. * testsuite/gas/aarch64/sve-invalid.l: Update accordingly. Also update expected output for new FMOV and MOV alternatives.
2017-02-25 02:29:00 +08:00
case AARCH64_OPND_SVE_Zm4_INDEX:
size = get_operand_fields_width (get_operand_from_code (type));
shift = get_operand_specific_data (&aarch64_operands[type]);
if (!check_reglane (opnd, mismatch_detail, idx,
"z", 0, (1 << shift) - 1,
0, (1u << (size - shift)) - 1))
return 0;
[AArch64] Additional SVE instructions This patch supports some additions to the SVE architecture prior to its public release. include/ * opcode/aarch64.h (AARCH64_OPND_SVE_ADDR_RI_S4x16) (AARCH64_OPND_SVE_IMM_ROT1, AARCH64_OPND_SVE_IMM_ROT2) (AARCH64_OPND_SVE_Zm3_INDEX, AARCH64_OPND_SVE_Zm3_22_INDEX) (AARCH64_OPND_SVE_Zm4_INDEX): New aarch64_opnds. opcodes/ * aarch64-tbl.h (OP_SVE_HMH, OP_SVE_VMU_HSD, OP_SVE_VMVU_HSD) (OP_SVE_VMVV_HSD, OP_SVE_VMVVU_HSD, OP_SVE_VM_HSD, OP_SVE_VUVV_HSD) (OP_SVE_VUV_HSD, OP_SVE_VU_HSD, OP_SVE_VVVU_H, OP_SVE_VVVU_S) (OP_SVE_VVVU_HSD, OP_SVE_VVV_D, OP_SVE_VVV_D_H, OP_SVE_VVV_H) (OP_SVE_VVV_HSD, OP_SVE_VVV_S, OP_SVE_VVV_S_B, OP_SVE_VVV_SD_BH) (OP_SVE_VV_BHSDQ, OP_SVE_VV_HSD, OP_SVE_VZVV_HSD, OP_SVE_VZV_HSD) (OP_SVE_V_HSD): New macros. (OP_SVE_VMU_SD, OP_SVE_VMVU_SD, OP_SVE_VM_SD, OP_SVE_VUVV_SD) (OP_SVE_VU_SD, OP_SVE_VVVU_SD, OP_SVE_VVV_SD, OP_SVE_VZVV_SD) (OP_SVE_VZV_SD, OP_SVE_V_SD): Delete. (aarch64_opcode_table): Add new SVE instructions. (aarch64_opcode_table): Use imm_rotate{1,2} instead of imm_rotate for rotation operands. Add new SVE operands. * aarch64-asm.h (ins_sve_addr_ri_s4): New inserter. (ins_sve_quad_index): Likewise. (ins_imm_rotate): Split into... (ins_imm_rotate1, ins_imm_rotate2): ...these two inserters. * aarch64-asm.c (aarch64_ins_imm_rotate): Split into... (aarch64_ins_imm_rotate1, aarch64_ins_imm_rotate2): ...these two functions. (aarch64_ins_sve_addr_ri_s4): New function. (aarch64_ins_sve_quad_index): Likewise. (do_misc_encoding): Handle "MOV Zn.Q, Qm". * aarch64-asm-2.c: Regenerate. * aarch64-dis.h (ext_sve_addr_ri_s4): New extractor. (ext_sve_quad_index): Likewise. (ext_imm_rotate): Split into... (ext_imm_rotate1, ext_imm_rotate2): ...these two extractors. * aarch64-dis.c (aarch64_ext_imm_rotate): Split into... (aarch64_ext_imm_rotate1, aarch64_ext_imm_rotate2): ...these two functions. (aarch64_ext_sve_addr_ri_s4): New function. (aarch64_ext_sve_quad_index): Likewise. (aarch64_ext_sve_index): Allow quad indices. (do_misc_decoding): Likewise. * aarch64-dis-2.c: Regenerate. * aarch64-opc.h (FLD_SVE_i3h, FLD_SVE_rot1, FLD_SVE_rot2): New aarch64_field_kinds. (OPD_F_OD_MASK): Widen by one bit. (OPD_F_NO_ZR): Bump accordingly. (get_operand_field_width): New function. * aarch64-opc.c (fields): Add new SVE fields. (operand_general_constraint_met_p): Handle new SVE operands. (aarch64_print_operand): Likewise. * aarch64-opc-2.c: Regenerate. gas/ * doc/c-aarch64.texi: Document that sve implies fp16, simd and compnum. * config/tc-aarch64.c (parse_vector_type_for_operand): Allow .q to be used with SVE registers. (parse_operands): Handle new SVE operands. (aarch64_features): Make "sve" require F16 rather than FP. Also require COMPNUM. * testsuite/gas/aarch64/sve.s: Add tests for new instructions. Include compnum tests. * testsuite/gas/aarch64/sve.d: Update accordingly. * testsuite/gas/aarch64/sve-invalid.s: Add tests for new instructions. * testsuite/gas/aarch64/sve-invalid.l: Update accordingly. Also update expected output for new FMOV and MOV alternatives.
2017-02-25 02:29:00 +08:00
break;
[AArch64][SVE 21/32] Add Zn and Pn registers This patch adds the Zn and Pn registers, and associated fields and operands. include/ * opcode/aarch64.h (AARCH64_OPND_CLASS_SVE_REG): New aarch64_operand_class. (AARCH64_OPND_CLASS_PRED_REG): Likewise. (AARCH64_OPND_SVE_Pd, AARCH64_OPND_SVE_Pg3, AARCH64_OPND_SVE_Pg4_5) (AARCH64_OPND_SVE_Pg4_10, AARCH64_OPND_SVE_Pg4_16) (AARCH64_OPND_SVE_Pm, AARCH64_OPND_SVE_Pn, AARCH64_OPND_SVE_Pt) (AARCH64_OPND_SVE_Za_5, AARCH64_OPND_SVE_Za_16, AARCH64_OPND_SVE_Zd) (AARCH64_OPND_SVE_Zm_5, AARCH64_OPND_SVE_Zm_16, AARCH64_OPND_SVE_Zn) (AARCH64_OPND_SVE_Zn_INDEX, AARCH64_OPND_SVE_ZnxN) (AARCH64_OPND_SVE_Zt, AARCH64_OPND_SVE_ZtxN): New aarch64_opnds. opcodes/ * aarch64-tbl.h (AARCH64_OPERANDS): Add entries for new SVE operands. * aarch64-opc.h (FLD_SVE_Pd, FLD_SVE_Pg3, FLD_SVE_Pg4_5) (FLD_SVE_Pg4_10, FLD_SVE_Pg4_16, FLD_SVE_Pm, FLD_SVE_Pn, FLD_SVE_Pt) (FLD_SVE_Za_5, FLD_SVE_Za_16, FLD_SVE_Zd, FLD_SVE_Zm_5, FLD_SVE_Zm_16) (FLD_SVE_Zn, FLD_SVE_Zt, FLD_SVE_tzsh): New aarch64_field_kinds. * aarch64-opc.c (fields): Add corresponding entries here. (operand_general_constraint_met_p): Check that SVE register lists have the correct length. Check the ranges of SVE index registers. Check for cases where p8-p15 are used in 3-bit predicate fields. (aarch64_print_operand): Handle the new SVE operands. * aarch64-opc-2.c: Regenerate. * aarch64-asm.h (ins_sve_index, ins_sve_reglist): New inserters. * aarch64-asm.c (aarch64_ins_sve_index): New function. (aarch64_ins_sve_reglist): Likewise. * aarch64-asm-2.c: Regenerate. * aarch64-dis.h (ext_sve_index, ext_sve_reglist): New extractors. * aarch64-dis.c (aarch64_ext_sve_index): New function. (aarch64_ext_sve_reglist): Likewise. * aarch64-dis-2.c: Regenerate. gas/ * config/tc-aarch64.c (NTA_HASVARWIDTH): New macro. (AARCH64_REG_TYPES): Add ZN and PN. (get_reg_expected_msg): Handle them. (parse_vector_type_for_operand): Add a reg_type parameter. Skip the width for Zn and Pn registers. (parse_typed_reg): Extend vector handling to Zn and Pn. Update the call to parse_vector_type_for_operand. Set HASVARTYPE for Zn and Pn, expecting the width to be 0. (parse_vector_reg_list): Restrict error about [BHSD]nn operands to REG_TYPE_VN. (vectype_to_qualifier): Use S_[BHSD] qualifiers for NTA_HASVARWIDTH. (parse_operands): Handle the new Zn and Pn operands. (REGSET16): New macro, split out from... (REGSET31): ...here. (reg_names): Add Zn and Pn entries.
2016-09-21 23:53:54 +08:00
case AARCH64_OPND_SVE_Zn_INDEX:
size = aarch64_get_qualifier_esize (opnd->qualifier);
if (!check_reglane (opnd, mismatch_detail, idx, "z", 0, 31,
0, 64 / size - 1))
return 0;
[AArch64][SVE 21/32] Add Zn and Pn registers This patch adds the Zn and Pn registers, and associated fields and operands. include/ * opcode/aarch64.h (AARCH64_OPND_CLASS_SVE_REG): New aarch64_operand_class. (AARCH64_OPND_CLASS_PRED_REG): Likewise. (AARCH64_OPND_SVE_Pd, AARCH64_OPND_SVE_Pg3, AARCH64_OPND_SVE_Pg4_5) (AARCH64_OPND_SVE_Pg4_10, AARCH64_OPND_SVE_Pg4_16) (AARCH64_OPND_SVE_Pm, AARCH64_OPND_SVE_Pn, AARCH64_OPND_SVE_Pt) (AARCH64_OPND_SVE_Za_5, AARCH64_OPND_SVE_Za_16, AARCH64_OPND_SVE_Zd) (AARCH64_OPND_SVE_Zm_5, AARCH64_OPND_SVE_Zm_16, AARCH64_OPND_SVE_Zn) (AARCH64_OPND_SVE_Zn_INDEX, AARCH64_OPND_SVE_ZnxN) (AARCH64_OPND_SVE_Zt, AARCH64_OPND_SVE_ZtxN): New aarch64_opnds. opcodes/ * aarch64-tbl.h (AARCH64_OPERANDS): Add entries for new SVE operands. * aarch64-opc.h (FLD_SVE_Pd, FLD_SVE_Pg3, FLD_SVE_Pg4_5) (FLD_SVE_Pg4_10, FLD_SVE_Pg4_16, FLD_SVE_Pm, FLD_SVE_Pn, FLD_SVE_Pt) (FLD_SVE_Za_5, FLD_SVE_Za_16, FLD_SVE_Zd, FLD_SVE_Zm_5, FLD_SVE_Zm_16) (FLD_SVE_Zn, FLD_SVE_Zt, FLD_SVE_tzsh): New aarch64_field_kinds. * aarch64-opc.c (fields): Add corresponding entries here. (operand_general_constraint_met_p): Check that SVE register lists have the correct length. Check the ranges of SVE index registers. Check for cases where p8-p15 are used in 3-bit predicate fields. (aarch64_print_operand): Handle the new SVE operands. * aarch64-opc-2.c: Regenerate. * aarch64-asm.h (ins_sve_index, ins_sve_reglist): New inserters. * aarch64-asm.c (aarch64_ins_sve_index): New function. (aarch64_ins_sve_reglist): Likewise. * aarch64-asm-2.c: Regenerate. * aarch64-dis.h (ext_sve_index, ext_sve_reglist): New extractors. * aarch64-dis.c (aarch64_ext_sve_index): New function. (aarch64_ext_sve_reglist): Likewise. * aarch64-dis-2.c: Regenerate. gas/ * config/tc-aarch64.c (NTA_HASVARWIDTH): New macro. (AARCH64_REG_TYPES): Add ZN and PN. (get_reg_expected_msg): Handle them. (parse_vector_type_for_operand): Add a reg_type parameter. Skip the width for Zn and Pn registers. (parse_typed_reg): Extend vector handling to Zn and Pn. Update the call to parse_vector_type_for_operand. Set HASVARTYPE for Zn and Pn, expecting the width to be 0. (parse_vector_reg_list): Restrict error about [BHSD]nn operands to REG_TYPE_VN. (vectype_to_qualifier): Use S_[BHSD] qualifiers for NTA_HASVARWIDTH. (parse_operands): Handle the new Zn and Pn operands. (REGSET16): New macro, split out from... (REGSET31): ...here. (reg_names): Add Zn and Pn entries.
2016-09-21 23:53:54 +08:00
break;
case AARCH64_OPND_SVE_Zm_imm4:
if (!check_reglane (opnd, mismatch_detail, idx, "z", 0, 31, 0, 15))
return 0;
break;
case AARCH64_OPND_SVE_Zn_5_INDEX:
size = aarch64_get_qualifier_esize (opnd->qualifier);
if (!check_reglane (opnd, mismatch_detail, idx, "z", 0, 31,
0, 16 / size - 1))
return 0;
break;
case AARCH64_OPND_SME_PNn3_INDEX1:
case AARCH64_OPND_SME_PNn3_INDEX2:
size = get_operand_field_width (get_operand_from_code (type), 1);
if (!check_reglane (opnd, mismatch_detail, idx, "pn", 8, 15,
0, (1 << size) - 1))
return 0;
break;
case AARCH64_OPND_SME_Zn_INDEX1_16:
case AARCH64_OPND_SME_Zn_INDEX2_15:
case AARCH64_OPND_SME_Zn_INDEX2_16:
case AARCH64_OPND_SME_Zn_INDEX3_14:
case AARCH64_OPND_SME_Zn_INDEX3_15:
case AARCH64_OPND_SME_Zn_INDEX4_14:
size = get_operand_fields_width (get_operand_from_code (type)) - 5;
if (!check_reglane (opnd, mismatch_detail, idx, "z", 0, 31,
0, (1 << size) - 1))
return 0;
break;
case AARCH64_OPND_SME_Zm_INDEX1:
case AARCH64_OPND_SME_Zm_INDEX2:
case AARCH64_OPND_SME_Zm_INDEX3_1:
case AARCH64_OPND_SME_Zm_INDEX3_2:
case AARCH64_OPND_SME_Zm_INDEX3_10:
case AARCH64_OPND_SME_Zm_INDEX4_1:
case AARCH64_OPND_SME_Zm_INDEX4_10:
size = get_operand_fields_width (get_operand_from_code (type)) - 4;
if (!check_reglane (opnd, mismatch_detail, idx, "z", 0, 15,
0, (1 << size) - 1))
return 0;
break;
case AARCH64_OPND_SME_Zm:
if (opnd->reg.regno > 15)
{
set_invalid_regno_error (mismatch_detail, idx, "z", 0, 15);
return 0;
}
break;
case AARCH64_OPND_SME_PnT_Wm_imm:
size = aarch64_get_qualifier_esize (opnd->qualifier);
max_value = 16 / size - 1;
if (!check_za_access (opnd, mismatch_detail, idx,
12, max_value, 1, 0))
return 0;
break;
[AArch64][SVE 21/32] Add Zn and Pn registers This patch adds the Zn and Pn registers, and associated fields and operands. include/ * opcode/aarch64.h (AARCH64_OPND_CLASS_SVE_REG): New aarch64_operand_class. (AARCH64_OPND_CLASS_PRED_REG): Likewise. (AARCH64_OPND_SVE_Pd, AARCH64_OPND_SVE_Pg3, AARCH64_OPND_SVE_Pg4_5) (AARCH64_OPND_SVE_Pg4_10, AARCH64_OPND_SVE_Pg4_16) (AARCH64_OPND_SVE_Pm, AARCH64_OPND_SVE_Pn, AARCH64_OPND_SVE_Pt) (AARCH64_OPND_SVE_Za_5, AARCH64_OPND_SVE_Za_16, AARCH64_OPND_SVE_Zd) (AARCH64_OPND_SVE_Zm_5, AARCH64_OPND_SVE_Zm_16, AARCH64_OPND_SVE_Zn) (AARCH64_OPND_SVE_Zn_INDEX, AARCH64_OPND_SVE_ZnxN) (AARCH64_OPND_SVE_Zt, AARCH64_OPND_SVE_ZtxN): New aarch64_opnds. opcodes/ * aarch64-tbl.h (AARCH64_OPERANDS): Add entries for new SVE operands. * aarch64-opc.h (FLD_SVE_Pd, FLD_SVE_Pg3, FLD_SVE_Pg4_5) (FLD_SVE_Pg4_10, FLD_SVE_Pg4_16, FLD_SVE_Pm, FLD_SVE_Pn, FLD_SVE_Pt) (FLD_SVE_Za_5, FLD_SVE_Za_16, FLD_SVE_Zd, FLD_SVE_Zm_5, FLD_SVE_Zm_16) (FLD_SVE_Zn, FLD_SVE_Zt, FLD_SVE_tzsh): New aarch64_field_kinds. * aarch64-opc.c (fields): Add corresponding entries here. (operand_general_constraint_met_p): Check that SVE register lists have the correct length. Check the ranges of SVE index registers. Check for cases where p8-p15 are used in 3-bit predicate fields. (aarch64_print_operand): Handle the new SVE operands. * aarch64-opc-2.c: Regenerate. * aarch64-asm.h (ins_sve_index, ins_sve_reglist): New inserters. * aarch64-asm.c (aarch64_ins_sve_index): New function. (aarch64_ins_sve_reglist): Likewise. * aarch64-asm-2.c: Regenerate. * aarch64-dis.h (ext_sve_index, ext_sve_reglist): New extractors. * aarch64-dis.c (aarch64_ext_sve_index): New function. (aarch64_ext_sve_reglist): Likewise. * aarch64-dis-2.c: Regenerate. gas/ * config/tc-aarch64.c (NTA_HASVARWIDTH): New macro. (AARCH64_REG_TYPES): Add ZN and PN. (get_reg_expected_msg): Handle them. (parse_vector_type_for_operand): Add a reg_type parameter. Skip the width for Zn and Pn registers. (parse_typed_reg): Extend vector handling to Zn and Pn. Update the call to parse_vector_type_for_operand. Set HASVARTYPE for Zn and Pn, expecting the width to be 0. (parse_vector_reg_list): Restrict error about [BHSD]nn operands to REG_TYPE_VN. (vectype_to_qualifier): Use S_[BHSD] qualifiers for NTA_HASVARWIDTH. (parse_operands): Handle the new Zn and Pn operands. (REGSET16): New macro, split out from... (REGSET31): ...here. (reg_names): Add Zn and Pn entries.
2016-09-21 23:53:54 +08:00
default:
break;
}
break;
case AARCH64_OPND_CLASS_SVE_REGLIST:
switch (type)
{
case AARCH64_OPND_SME_Pdx2:
case AARCH64_OPND_SME_Zdnx2:
case AARCH64_OPND_SME_Zdnx4:
case AARCH64_OPND_SME_Zmx2:
case AARCH64_OPND_SME_Zmx4:
case AARCH64_OPND_SME_Znx2:
case AARCH64_OPND_SME_Znx4:
case AARCH64_OPND_SME_Zt2:
case AARCH64_OPND_SME_Zt3:
case AARCH64_OPND_SME_Zt4:
num = get_operand_specific_data (&aarch64_operands[type]);
if (!check_reglist (opnd, mismatch_detail, idx, num, 1))
return 0;
if ((opnd->reglist.first_regno % num) != 0)
{
set_other_error (mismatch_detail, idx,
_("start register out of range"));
return 0;
}
break;
case AARCH64_OPND_SME_Ztx2_STRIDED:
case AARCH64_OPND_SME_Ztx4_STRIDED:
/* 2-register lists have a stride of 8 and 4-register lists
have a stride of 4. */
num = get_operand_specific_data (&aarch64_operands[type]);
if (!check_reglist (opnd, mismatch_detail, idx, num, 16 / num))
return 0;
num = 16 | (opnd->reglist.stride - 1);
if ((opnd->reglist.first_regno & ~num) != 0)
{
set_other_error (mismatch_detail, idx,
_("start register out of range"));
return 0;
}
break;
case AARCH64_OPND_SME_PdxN:
case AARCH64_OPND_SVE_ZnxN:
case AARCH64_OPND_SVE_ZtxN:
num = get_opcode_dependent_value (opcode);
if (!check_reglist (opnd, mismatch_detail, idx, num, 1))
return 0;
break;
default:
abort ();
}
break;
case AARCH64_OPND_CLASS_ZA_ACCESS:
switch (type)
{
case AARCH64_OPND_SME_ZA_HV_idx_src:
case AARCH64_OPND_SME_ZA_HV_idx_dest:
case AARCH64_OPND_SME_ZA_HV_idx_ldstr:
size = aarch64_get_qualifier_esize (opnd->qualifier);
max_value = 16 / size - 1;
if (!check_za_access (opnd, mismatch_detail, idx, 12, max_value, 1,
get_opcode_dependent_value (opcode)))
return 0;
break;
case AARCH64_OPND_SME_ZA_array_off4:
if (!check_za_access (opnd, mismatch_detail, idx, 12, 15, 1,
get_opcode_dependent_value (opcode)))
return 0;
break;
case AARCH64_OPND_SME_ZA_array_off3_0:
case AARCH64_OPND_SME_ZA_array_off3_5:
if (!check_za_access (opnd, mismatch_detail, idx, 8, 7, 1,
get_opcode_dependent_value (opcode)))
return 0;
break;
case AARCH64_OPND_SME_ZA_array_off1x4:
if (!check_za_access (opnd, mismatch_detail, idx, 8, 1, 4,
get_opcode_dependent_value (opcode)))
return 0;
break;
case AARCH64_OPND_SME_ZA_array_off2x2:
if (!check_za_access (opnd, mismatch_detail, idx, 8, 3, 2,
get_opcode_dependent_value (opcode)))
return 0;
break;
case AARCH64_OPND_SME_ZA_array_off2x4:
if (!check_za_access (opnd, mismatch_detail, idx, 8, 3, 4,
get_opcode_dependent_value (opcode)))
return 0;
break;
case AARCH64_OPND_SME_ZA_array_off3x2:
if (!check_za_access (opnd, mismatch_detail, idx, 8, 7, 2,
get_opcode_dependent_value (opcode)))
return 0;
break;
case AARCH64_OPND_SME_ZA_array_vrsb_1:
if (!check_za_access (opnd, mismatch_detail, idx, 12, 7, 2,
get_opcode_dependent_value (opcode)))
return 0;
break;
case AARCH64_OPND_SME_ZA_array_vrsh_1:
if (!check_za_access (opnd, mismatch_detail, idx, 12, 3, 2,
get_opcode_dependent_value (opcode)))
return 0;
break;
case AARCH64_OPND_SME_ZA_array_vrss_1:
if (!check_za_access (opnd, mismatch_detail, idx, 12, 1, 2,
get_opcode_dependent_value (opcode)))
return 0;
break;
case AARCH64_OPND_SME_ZA_array_vrsd_1:
if (!check_za_access (opnd, mismatch_detail, idx, 12, 0, 2,
get_opcode_dependent_value (opcode)))
return 0;
break;
case AARCH64_OPND_SME_ZA_array_vrsb_2:
if (!check_za_access (opnd, mismatch_detail, idx, 12, 3, 4,
get_opcode_dependent_value (opcode)))
return 0;
break;
case AARCH64_OPND_SME_ZA_array_vrsh_2:
if (!check_za_access (opnd, mismatch_detail, idx, 12, 1, 4,
get_opcode_dependent_value (opcode)))
return 0;
break;
case AARCH64_OPND_SME_ZA_array_vrss_2:
case AARCH64_OPND_SME_ZA_array_vrsd_2:
if (!check_za_access (opnd, mismatch_detail, idx, 12, 0, 4,
get_opcode_dependent_value (opcode)))
return 0;
break;
case AARCH64_OPND_SME_ZA_HV_idx_srcxN:
case AARCH64_OPND_SME_ZA_HV_idx_destxN:
size = aarch64_get_qualifier_esize (opnd->qualifier);
num = get_opcode_dependent_value (opcode);
max_value = 16 / num / size;
if (max_value > 0)
max_value -= 1;
if (!check_za_access (opnd, mismatch_detail, idx,
12, max_value, num, 0))
return 0;
break;
default:
abort ();
}
break;
[AArch64][SVE 21/32] Add Zn and Pn registers This patch adds the Zn and Pn registers, and associated fields and operands. include/ * opcode/aarch64.h (AARCH64_OPND_CLASS_SVE_REG): New aarch64_operand_class. (AARCH64_OPND_CLASS_PRED_REG): Likewise. (AARCH64_OPND_SVE_Pd, AARCH64_OPND_SVE_Pg3, AARCH64_OPND_SVE_Pg4_5) (AARCH64_OPND_SVE_Pg4_10, AARCH64_OPND_SVE_Pg4_16) (AARCH64_OPND_SVE_Pm, AARCH64_OPND_SVE_Pn, AARCH64_OPND_SVE_Pt) (AARCH64_OPND_SVE_Za_5, AARCH64_OPND_SVE_Za_16, AARCH64_OPND_SVE_Zd) (AARCH64_OPND_SVE_Zm_5, AARCH64_OPND_SVE_Zm_16, AARCH64_OPND_SVE_Zn) (AARCH64_OPND_SVE_Zn_INDEX, AARCH64_OPND_SVE_ZnxN) (AARCH64_OPND_SVE_Zt, AARCH64_OPND_SVE_ZtxN): New aarch64_opnds. opcodes/ * aarch64-tbl.h (AARCH64_OPERANDS): Add entries for new SVE operands. * aarch64-opc.h (FLD_SVE_Pd, FLD_SVE_Pg3, FLD_SVE_Pg4_5) (FLD_SVE_Pg4_10, FLD_SVE_Pg4_16, FLD_SVE_Pm, FLD_SVE_Pn, FLD_SVE_Pt) (FLD_SVE_Za_5, FLD_SVE_Za_16, FLD_SVE_Zd, FLD_SVE_Zm_5, FLD_SVE_Zm_16) (FLD_SVE_Zn, FLD_SVE_Zt, FLD_SVE_tzsh): New aarch64_field_kinds. * aarch64-opc.c (fields): Add corresponding entries here. (operand_general_constraint_met_p): Check that SVE register lists have the correct length. Check the ranges of SVE index registers. Check for cases where p8-p15 are used in 3-bit predicate fields. (aarch64_print_operand): Handle the new SVE operands. * aarch64-opc-2.c: Regenerate. * aarch64-asm.h (ins_sve_index, ins_sve_reglist): New inserters. * aarch64-asm.c (aarch64_ins_sve_index): New function. (aarch64_ins_sve_reglist): Likewise. * aarch64-asm-2.c: Regenerate. * aarch64-dis.h (ext_sve_index, ext_sve_reglist): New extractors. * aarch64-dis.c (aarch64_ext_sve_index): New function. (aarch64_ext_sve_reglist): Likewise. * aarch64-dis-2.c: Regenerate. gas/ * config/tc-aarch64.c (NTA_HASVARWIDTH): New macro. (AARCH64_REG_TYPES): Add ZN and PN. (get_reg_expected_msg): Handle them. (parse_vector_type_for_operand): Add a reg_type parameter. Skip the width for Zn and Pn registers. (parse_typed_reg): Extend vector handling to Zn and Pn. Update the call to parse_vector_type_for_operand. Set HASVARTYPE for Zn and Pn, expecting the width to be 0. (parse_vector_reg_list): Restrict error about [BHSD]nn operands to REG_TYPE_VN. (vectype_to_qualifier): Use S_[BHSD] qualifiers for NTA_HASVARWIDTH. (parse_operands): Handle the new Zn and Pn operands. (REGSET16): New macro, split out from... (REGSET31): ...here. (reg_names): Add Zn and Pn entries.
2016-09-21 23:53:54 +08:00
case AARCH64_OPND_CLASS_PRED_REG:
switch (type)
[AArch64][SVE 21/32] Add Zn and Pn registers This patch adds the Zn and Pn registers, and associated fields and operands. include/ * opcode/aarch64.h (AARCH64_OPND_CLASS_SVE_REG): New aarch64_operand_class. (AARCH64_OPND_CLASS_PRED_REG): Likewise. (AARCH64_OPND_SVE_Pd, AARCH64_OPND_SVE_Pg3, AARCH64_OPND_SVE_Pg4_5) (AARCH64_OPND_SVE_Pg4_10, AARCH64_OPND_SVE_Pg4_16) (AARCH64_OPND_SVE_Pm, AARCH64_OPND_SVE_Pn, AARCH64_OPND_SVE_Pt) (AARCH64_OPND_SVE_Za_5, AARCH64_OPND_SVE_Za_16, AARCH64_OPND_SVE_Zd) (AARCH64_OPND_SVE_Zm_5, AARCH64_OPND_SVE_Zm_16, AARCH64_OPND_SVE_Zn) (AARCH64_OPND_SVE_Zn_INDEX, AARCH64_OPND_SVE_ZnxN) (AARCH64_OPND_SVE_Zt, AARCH64_OPND_SVE_ZtxN): New aarch64_opnds. opcodes/ * aarch64-tbl.h (AARCH64_OPERANDS): Add entries for new SVE operands. * aarch64-opc.h (FLD_SVE_Pd, FLD_SVE_Pg3, FLD_SVE_Pg4_5) (FLD_SVE_Pg4_10, FLD_SVE_Pg4_16, FLD_SVE_Pm, FLD_SVE_Pn, FLD_SVE_Pt) (FLD_SVE_Za_5, FLD_SVE_Za_16, FLD_SVE_Zd, FLD_SVE_Zm_5, FLD_SVE_Zm_16) (FLD_SVE_Zn, FLD_SVE_Zt, FLD_SVE_tzsh): New aarch64_field_kinds. * aarch64-opc.c (fields): Add corresponding entries here. (operand_general_constraint_met_p): Check that SVE register lists have the correct length. Check the ranges of SVE index registers. Check for cases where p8-p15 are used in 3-bit predicate fields. (aarch64_print_operand): Handle the new SVE operands. * aarch64-opc-2.c: Regenerate. * aarch64-asm.h (ins_sve_index, ins_sve_reglist): New inserters. * aarch64-asm.c (aarch64_ins_sve_index): New function. (aarch64_ins_sve_reglist): Likewise. * aarch64-asm-2.c: Regenerate. * aarch64-dis.h (ext_sve_index, ext_sve_reglist): New extractors. * aarch64-dis.c (aarch64_ext_sve_index): New function. (aarch64_ext_sve_reglist): Likewise. * aarch64-dis-2.c: Regenerate. gas/ * config/tc-aarch64.c (NTA_HASVARWIDTH): New macro. (AARCH64_REG_TYPES): Add ZN and PN. (get_reg_expected_msg): Handle them. (parse_vector_type_for_operand): Add a reg_type parameter. Skip the width for Zn and Pn registers. (parse_typed_reg): Extend vector handling to Zn and Pn. Update the call to parse_vector_type_for_operand. Set HASVARTYPE for Zn and Pn, expecting the width to be 0. (parse_vector_reg_list): Restrict error about [BHSD]nn operands to REG_TYPE_VN. (vectype_to_qualifier): Use S_[BHSD] qualifiers for NTA_HASVARWIDTH. (parse_operands): Handle the new Zn and Pn operands. (REGSET16): New macro, split out from... (REGSET31): ...here. (reg_names): Add Zn and Pn entries.
2016-09-21 23:53:54 +08:00
{
case AARCH64_OPND_SME_PNd3:
case AARCH64_OPND_SME_PNg3:
if (opnd->reg.regno < 8)
{
set_invalid_regno_error (mismatch_detail, idx, "pn", 8, 15);
return 0;
}
break;
default:
if (opnd->reg.regno >= 8
&& get_operand_fields_width (get_operand_from_code (type)) == 3)
{
set_invalid_regno_error (mismatch_detail, idx, "p", 0, 7);
return 0;
}
break;
[AArch64][SVE 21/32] Add Zn and Pn registers This patch adds the Zn and Pn registers, and associated fields and operands. include/ * opcode/aarch64.h (AARCH64_OPND_CLASS_SVE_REG): New aarch64_operand_class. (AARCH64_OPND_CLASS_PRED_REG): Likewise. (AARCH64_OPND_SVE_Pd, AARCH64_OPND_SVE_Pg3, AARCH64_OPND_SVE_Pg4_5) (AARCH64_OPND_SVE_Pg4_10, AARCH64_OPND_SVE_Pg4_16) (AARCH64_OPND_SVE_Pm, AARCH64_OPND_SVE_Pn, AARCH64_OPND_SVE_Pt) (AARCH64_OPND_SVE_Za_5, AARCH64_OPND_SVE_Za_16, AARCH64_OPND_SVE_Zd) (AARCH64_OPND_SVE_Zm_5, AARCH64_OPND_SVE_Zm_16, AARCH64_OPND_SVE_Zn) (AARCH64_OPND_SVE_Zn_INDEX, AARCH64_OPND_SVE_ZnxN) (AARCH64_OPND_SVE_Zt, AARCH64_OPND_SVE_ZtxN): New aarch64_opnds. opcodes/ * aarch64-tbl.h (AARCH64_OPERANDS): Add entries for new SVE operands. * aarch64-opc.h (FLD_SVE_Pd, FLD_SVE_Pg3, FLD_SVE_Pg4_5) (FLD_SVE_Pg4_10, FLD_SVE_Pg4_16, FLD_SVE_Pm, FLD_SVE_Pn, FLD_SVE_Pt) (FLD_SVE_Za_5, FLD_SVE_Za_16, FLD_SVE_Zd, FLD_SVE_Zm_5, FLD_SVE_Zm_16) (FLD_SVE_Zn, FLD_SVE_Zt, FLD_SVE_tzsh): New aarch64_field_kinds. * aarch64-opc.c (fields): Add corresponding entries here. (operand_general_constraint_met_p): Check that SVE register lists have the correct length. Check the ranges of SVE index registers. Check for cases where p8-p15 are used in 3-bit predicate fields. (aarch64_print_operand): Handle the new SVE operands. * aarch64-opc-2.c: Regenerate. * aarch64-asm.h (ins_sve_index, ins_sve_reglist): New inserters. * aarch64-asm.c (aarch64_ins_sve_index): New function. (aarch64_ins_sve_reglist): Likewise. * aarch64-asm-2.c: Regenerate. * aarch64-dis.h (ext_sve_index, ext_sve_reglist): New extractors. * aarch64-dis.c (aarch64_ext_sve_index): New function. (aarch64_ext_sve_reglist): Likewise. * aarch64-dis-2.c: Regenerate. gas/ * config/tc-aarch64.c (NTA_HASVARWIDTH): New macro. (AARCH64_REG_TYPES): Add ZN and PN. (get_reg_expected_msg): Handle them. (parse_vector_type_for_operand): Add a reg_type parameter. Skip the width for Zn and Pn registers. (parse_typed_reg): Extend vector handling to Zn and Pn. Update the call to parse_vector_type_for_operand. Set HASVARTYPE for Zn and Pn, expecting the width to be 0. (parse_vector_reg_list): Restrict error about [BHSD]nn operands to REG_TYPE_VN. (vectype_to_qualifier): Use S_[BHSD] qualifiers for NTA_HASVARWIDTH. (parse_operands): Handle the new Zn and Pn operands. (REGSET16): New macro, split out from... (REGSET31): ...here. (reg_names): Add Zn and Pn entries.
2016-09-21 23:53:54 +08:00
}
break;
case AARCH64_OPND_CLASS_COND:
if (type == AARCH64_OPND_COND1
&& (opnds[idx].cond->value & 0xe) == 0xe)
{
/* Not allow AL or NV. */
set_syntax_error (mismatch_detail, idx, NULL);
}
break;
case AARCH64_OPND_CLASS_ADDRESS:
/* Check writeback. */
switch (opcode->iclass)
{
case ldst_pos:
case ldst_unscaled:
case ldstnapair_offs:
case ldstpair_off:
case ldst_unpriv:
if (opnd->addr.writeback == 1)
{
set_syntax_error (mismatch_detail, idx,
_("unexpected address writeback"));
return 0;
}
break;
[AArch64] Add ARMv8.3 combined pointer authentication load instructions Add support for ARMv8.3 LDRAA and LDRAB combined pointer authentication and load instructions. These instructions authenticate the base register and load 8 byte from it plus a scaled 10-bit offset with optional writeback to update the base register. A new instruction class (ldst_imm10) and operand type (AARCH64_OPND_ADDR_SIMM10) were introduced to handle the special addressing form. include/ 2016-11-18 Szabolcs Nagy <szabolcs.nagy@arm.com> * opcode/aarch64.h (enum aarch64_opnd): Add AARCH64_OPND_ADDR_SIMM10. (enum aarch64_insn_class): Add ldst_imm10. opcodes/ 2016-11-18 Szabolcs Nagy <szabolcs.nagy@arm.com> * aarch64-tbl.h (QL_X1NIL): New. (arch64_opcode_table): Add ldraa, ldrab. (AARCH64_OPERANDS): Add "ADDR_SIMM10". * aarch64-asm.h (aarch64_ins_addr_simm10): Declare. * aarch64-asm.c (aarch64_ins_addr_simm10): Define. * aarch64-dis.h (aarch64_ext_addr_simm10): Declare. * aarch64-dis.c (aarch64_ext_addr_simm10): Define. * aarch64-opc.h (enum aarch64_field_kind): Add FLD_S_simm10. * aarch64-opc.c (fields): Add data for FLD_S_simm10. (operand_general_constraint_met_p): Handle AARCH64_OPND_ADDR_SIMM10. (aarch64_print_operand): Likewise. * aarch64-asm-2.c: Regenerate. * aarch64-dis-2.c: Regenerate. * aarch64-opc-2.c: Regenerate. gas/ 2016-11-18 Szabolcs Nagy <szabolcs.nagy@arm.com> * config/tc-aarch64.c (parse_operands): Handle AARCH64_OPND_ADDR_SIMM10. (fix_insn): Likewise. (warn_unpredictable_ldst): Handle ldst_imm10. * testsuite/gas/aarch64/pac.s: Add ldraa and ldrab tests. * testsuite/gas/aarch64/pac.d: Likewise. * testsuite/gas/aarch64/illegal-ldraa.s: New. * testsuite/gas/aarch64/illegal-ldraa.l: New. * testsuite/gas/aarch64/illegal-ldraa.d: New.
2016-11-18 17:49:06 +08:00
case ldst_imm10:
if (opnd->addr.writeback == 1 && opnd->addr.preind != 1)
{
set_syntax_error (mismatch_detail, idx,
_("unexpected address writeback"));
return 0;
}
break;
case ldst_imm9:
case ldstpair_indexed:
case asisdlsep:
case asisdlsop:
if (opnd->addr.writeback == 0)
{
set_syntax_error (mismatch_detail, idx,
_("address writeback expected"));
return 0;
}
break;
case rcpc3:
if (opnd->addr.writeback)
if ((type == AARCH64_OPND_RCPC3_ADDR_PREIND_WB
&& !opnd->addr.preind)
|| (type == AARCH64_OPND_RCPC3_ADDR_POSTIND
&& !opnd->addr.postind))
{
set_syntax_error (mismatch_detail, idx,
_("unexpected address writeback"));
return 0;
}
break;
default:
assert (opnd->addr.writeback == 0);
break;
}
switch (type)
{
case AARCH64_OPND_ADDR_SIMM7:
/* Scaled signed 7 bits immediate offset. */
/* Get the size of the data element that is accessed, which may be
different from that of the source register size,
e.g. in strb/ldrb. */
size = aarch64_get_qualifier_esize (opnd->qualifier);
if (!value_in_range_p (opnd->addr.offset.imm, -64 * size, 63 * size))
{
set_offset_out_of_range_error (mismatch_detail, idx,
-64 * size, 63 * size);
return 0;
}
if (!value_aligned_p (opnd->addr.offset.imm, size))
{
set_unaligned_error (mismatch_detail, idx, size);
return 0;
}
break;
case AARCH64_OPND_ADDR_OFFSET:
case AARCH64_OPND_ADDR_SIMM9:
/* Unscaled signed 9 bits immediate offset. */
if (!value_in_range_p (opnd->addr.offset.imm, -256, 255))
{
set_offset_out_of_range_error (mismatch_detail, idx, -256, 255);
return 0;
}
break;
case AARCH64_OPND_ADDR_SIMM9_2:
/* Unscaled signed 9 bits immediate offset, which has to be negative
or unaligned. */
size = aarch64_get_qualifier_esize (qualifier);
if ((value_in_range_p (opnd->addr.offset.imm, 0, 255)
&& !value_aligned_p (opnd->addr.offset.imm, size))
|| value_in_range_p (opnd->addr.offset.imm, -256, -1))
return 1;
set_other_error (mismatch_detail, idx,
_("negative or unaligned offset expected"));
return 0;
[AArch64] Add ARMv8.3 combined pointer authentication load instructions Add support for ARMv8.3 LDRAA and LDRAB combined pointer authentication and load instructions. These instructions authenticate the base register and load 8 byte from it plus a scaled 10-bit offset with optional writeback to update the base register. A new instruction class (ldst_imm10) and operand type (AARCH64_OPND_ADDR_SIMM10) were introduced to handle the special addressing form. include/ 2016-11-18 Szabolcs Nagy <szabolcs.nagy@arm.com> * opcode/aarch64.h (enum aarch64_opnd): Add AARCH64_OPND_ADDR_SIMM10. (enum aarch64_insn_class): Add ldst_imm10. opcodes/ 2016-11-18 Szabolcs Nagy <szabolcs.nagy@arm.com> * aarch64-tbl.h (QL_X1NIL): New. (arch64_opcode_table): Add ldraa, ldrab. (AARCH64_OPERANDS): Add "ADDR_SIMM10". * aarch64-asm.h (aarch64_ins_addr_simm10): Declare. * aarch64-asm.c (aarch64_ins_addr_simm10): Define. * aarch64-dis.h (aarch64_ext_addr_simm10): Declare. * aarch64-dis.c (aarch64_ext_addr_simm10): Define. * aarch64-opc.h (enum aarch64_field_kind): Add FLD_S_simm10. * aarch64-opc.c (fields): Add data for FLD_S_simm10. (operand_general_constraint_met_p): Handle AARCH64_OPND_ADDR_SIMM10. (aarch64_print_operand): Likewise. * aarch64-asm-2.c: Regenerate. * aarch64-dis-2.c: Regenerate. * aarch64-opc-2.c: Regenerate. gas/ 2016-11-18 Szabolcs Nagy <szabolcs.nagy@arm.com> * config/tc-aarch64.c (parse_operands): Handle AARCH64_OPND_ADDR_SIMM10. (fix_insn): Likewise. (warn_unpredictable_ldst): Handle ldst_imm10. * testsuite/gas/aarch64/pac.s: Add ldraa and ldrab tests. * testsuite/gas/aarch64/pac.d: Likewise. * testsuite/gas/aarch64/illegal-ldraa.s: New. * testsuite/gas/aarch64/illegal-ldraa.l: New. * testsuite/gas/aarch64/illegal-ldraa.d: New.
2016-11-18 17:49:06 +08:00
case AARCH64_OPND_ADDR_SIMM10:
/* Scaled signed 10 bits immediate offset. */
if (!value_in_range_p (opnd->addr.offset.imm, -4096, 4088))
{
set_offset_out_of_range_error (mismatch_detail, idx, -4096, 4088);
return 0;
}
if (!value_aligned_p (opnd->addr.offset.imm, 8))
{
set_unaligned_error (mismatch_detail, idx, 8);
return 0;
}
break;
[BINUTILS, AARCH64, 4/8] Add Tag setting instructions in Memory Tagging Extension This patch is part of the patch series to add support for ARMv8.5-A Memory Tagging Extensions which is an optional extension to ARMv8.5-A and is enabled using the +memtag command line option. This patch add support to the Tag setting instructions from MTE which consists of the following instructions: - STG [<Xn|SP>, #<simm>] - STG [<Xn|SP>, #<simm>]! - STG [<Xn|SP>], #<simm> - STZG [<Xn|SP>, #<simm>] - STZG [<Xn|SP>, #<simm>]! - STZG [<Xn|SP>], #<simm> - ST2G [<Xn|SP>, #<simm>] - ST2G [<Xn|SP>, #<simm>]! - ST2G [<Xn|SP>], #<simm> - STZ2G [<Xn|SP>, #<simm>] - STZ2G [<Xn|SP>, #<simm>]! - STZ2G [<Xn|SP>], #<simm> - STGP <Xt>, <Xt2>, [<Xn|SP>, #<imm>] - STGP <Xt>, <Xt2>, [<Xn|SP>, #<imm>]! - STGP <Xt>, <Xt2>, [<Xn|SP>], #<imm> where <Xn|SP> : Is the 64-bit GPR or Stack pointer. <simm> : Is the optional signed immediate offset, a multiple of 16 in the range -4096 to 4080, defaulting to 0. *** include/ChangeLog *** 2018-11-12 Sudakshina Das <sudi.das@arm.com> * opcode/aarch64.h (aarch64_opnd): Add AARCH64_OPND_ADDR_SIMM11 and AARCH64_OPND_ADDR_SIMM13. (aarch64_opnd_qualifier): Add new AARCH64_OPND_QLF_imm_tag. *** opcodes/ChangeLog *** 2018-11-12 Sudakshina Das <sudi.das@arm.com> * aarch64-opc.c (aarch64_opnd_qualifiers): Add new data for AARCH64_OPND_QLF_imm_tag. (operand_general_constraint_met_p): Add case for AARCH64_OPND_ADDR_SIMM11 and AARCH64_OPND_ADDR_SIMM13. (aarch64_print_operand): Likewise. * aarch64-tbl.h (QL_LDST_AT, QL_STGP): New. (aarch64_opcode_table): Add stg, stzg, st2g, stz2g and stgp for both offset and pre/post indexed versions. (AARCH64_OPERANDS): Define ADDR_SIMM11 and ADDR_SIMM13. * aarch64-asm-2.c: Regenerated. * aarch64-dis-2.c: Regenerated. * aarch64-opc-2.c: Regenerated. *** gas/ChangeLog *** 2018-11-12 Sudakshina Das <sudi.das@arm.com> * config/tc-aarch64.c (parse_operands): Add switch case for AARCH64_OPND_ADDR_SIMM11 and AARCH64_OPND_ADDR_SIMM13. (fix_insn): Likewise. (warn_unpredictable_ldst): Exempt STGP. * testsuite/gas/aarch64/armv8_5-a-memtag.s: Add tests for stg, st2g, stzg, stz2g and stgp. * testsuite/gas/aarch64/armv8_5-a-memtag.d: Likewise. * testsuite/gas/aarch64/illegal-memtag.s: Likewise. * testsuite/gas/aarch64/illegal-memtag.l: Likewise.
2018-11-12 21:09:55 +08:00
case AARCH64_OPND_ADDR_SIMM11:
/* Signed 11 bits immediate offset (multiple of 16). */
if (!value_in_range_p (opnd->addr.offset.imm, -1024, 1008))
{
set_offset_out_of_range_error (mismatch_detail, idx, -1024, 1008);
return 0;
}
if (!value_aligned_p (opnd->addr.offset.imm, 16))
{
set_unaligned_error (mismatch_detail, idx, 16);
return 0;
}
break;
case AARCH64_OPND_ADDR_SIMM13:
/* Signed 13 bits immediate offset (multiple of 16). */
if (!value_in_range_p (opnd->addr.offset.imm, -4096, 4080))
{
set_offset_out_of_range_error (mismatch_detail, idx, -4096, 4080);
return 0;
}
if (!value_aligned_p (opnd->addr.offset.imm, 16))
{
set_unaligned_error (mismatch_detail, idx, 16);
return 0;
}
break;
case AARCH64_OPND_SIMD_ADDR_POST:
/* AdvSIMD load/store multiple structures, post-index. */
assert (idx == 1);
if (opnd->addr.offset.is_reg)
{
if (value_in_range_p (opnd->addr.offset.regno, 0, 30))
return 1;
else
{
set_other_error (mismatch_detail, idx,
_("invalid register offset"));
return 0;
}
}
else
{
const aarch64_opnd_info *prev = &opnds[idx-1];
unsigned num_bytes; /* total number of bytes transferred. */
/* The opcode dependent area stores the number of elements in
each structure to be loaded/stored. */
int is_ld1r = get_opcode_dependent_value (opcode) == 1;
if (opcode->operands[0] == AARCH64_OPND_LVt_AL)
/* Special handling of loading single structure to all lane. */
num_bytes = (is_ld1r ? 1 : prev->reglist.num_regs)
* aarch64_get_qualifier_esize (prev->qualifier);
else
num_bytes = prev->reglist.num_regs
* aarch64_get_qualifier_esize (prev->qualifier)
* aarch64_get_qualifier_nelem (prev->qualifier);
if ((int) num_bytes != opnd->addr.offset.imm)
{
set_other_error (mismatch_detail, idx,
_("invalid post-increment amount"));
return 0;
}
}
break;
case AARCH64_OPND_ADDR_REGOFF:
/* Get the size of the data element that is accessed, which may be
different from that of the source register size,
e.g. in strb/ldrb. */
size = aarch64_get_qualifier_esize (opnd->qualifier);
/* It is either no shift or shift by the binary logarithm of SIZE. */
if (opnd->shifter.amount != 0
&& opnd->shifter.amount != (int)get_logsz (size))
{
set_other_error (mismatch_detail, idx,
_("invalid shift amount"));
return 0;
}
/* Only UXTW, LSL, SXTW and SXTX are the accepted extending
operators. */
switch (opnd->shifter.kind)
{
case AARCH64_MOD_UXTW:
case AARCH64_MOD_LSL:
case AARCH64_MOD_SXTW:
case AARCH64_MOD_SXTX: break;
default:
set_other_error (mismatch_detail, idx,
_("invalid extend/shift operator"));
return 0;
}
break;
case AARCH64_OPND_ADDR_UIMM12:
imm = opnd->addr.offset.imm;
/* Get the size of the data element that is accessed, which may be
different from that of the source register size,
e.g. in strb/ldrb. */
size = aarch64_get_qualifier_esize (qualifier);
if (!value_in_range_p (opnd->addr.offset.imm, 0, 4095 * size))
{
set_offset_out_of_range_error (mismatch_detail, idx,
0, 4095 * size);
return 0;
}
if (!value_aligned_p (opnd->addr.offset.imm, size))
{
set_unaligned_error (mismatch_detail, idx, size);
return 0;
}
break;
case AARCH64_OPND_ADDR_PCREL14:
case AARCH64_OPND_ADDR_PCREL19:
case AARCH64_OPND_ADDR_PCREL21:
case AARCH64_OPND_ADDR_PCREL26:
imm = opnd->imm.value;
if (operand_need_shift_by_two (get_operand_from_code (type)))
{
/* The offset value in a PC-relative branch instruction is alway
4-byte aligned and is encoded without the lowest 2 bits. */
if (!value_aligned_p (imm, 4))
{
set_unaligned_error (mismatch_detail, idx, 4);
return 0;
}
/* Right shift by 2 so that we can carry out the following check
canonically. */
imm >>= 2;
}
size = get_operand_fields_width (get_operand_from_code (type));
if (!value_fit_signed_field_p (imm, size))
{
set_other_error (mismatch_detail, idx,
_("immediate out of range"));
return 0;
}
break;
aarch64: [SME] Add LD1x, ST1x, LDR and STR instructions This patch is adding new loads and stores defined by SME instructions. gas/ChangeLog: * config/tc-aarch64.c (parse_sme_address): New parser. (parse_sme_za_hv_tiles_operand_with_braces): New parser. (parse_sme_za_array): New parser. (output_operand_error_record): Print error details if present. (parse_operands): Support new operands. * testsuite/gas/aarch64/sme-5-illegal.d: New test. * testsuite/gas/aarch64/sme-5-illegal.l: New test. * testsuite/gas/aarch64/sme-5-illegal.s: New test. * testsuite/gas/aarch64/sme-5.d: New test. * testsuite/gas/aarch64/sme-5.s: New test. * testsuite/gas/aarch64/sme-6-illegal.d: New test. * testsuite/gas/aarch64/sme-6-illegal.l: New test. * testsuite/gas/aarch64/sme-6-illegal.s: New test. * testsuite/gas/aarch64/sme-6.d: New test. * testsuite/gas/aarch64/sme-6.s: New test. * testsuite/gas/aarch64/sme-7-illegal.d: New test. * testsuite/gas/aarch64/sme-7-illegal.l: New test. * testsuite/gas/aarch64/sme-7-illegal.s: New test. * testsuite/gas/aarch64/sme-7.d: New test. * testsuite/gas/aarch64/sme-7.s: New test. include/ChangeLog: * opcode/aarch64.h (enum aarch64_opnd): New operands. (enum aarch64_insn_class): Added sme_ldr and sme_str. (AARCH64_OPDE_UNTIED_IMMS): New operand error kind. opcodes/ChangeLog: * aarch64-asm.c (aarch64_ins_sme_za_hv_tiles): New inserter. (aarch64_ins_sme_za_list): New inserter. (aarch64_ins_sme_za_array): New inserter. (aarch64_ins_sme_addr_ri_u4xvl): New inserter. * aarch64-asm.h (AARCH64_DECL_OPD_INSERTER): Added ins_sme_za_list, ins_sme_za_array and ins_sme_addr_ri_u4xvl. * aarch64-dis.c (aarch64_ext_sme_za_hv_tiles): New extractor. (aarch64_ext_sme_za_list): New extractor. (aarch64_ext_sme_za_array): New extractor. (aarch64_ext_sme_addr_ri_u4xvl): New extractor. * aarch64-dis.h (AARCH64_DECL_OPD_EXTRACTOR): Added ext_sme_za_list, ext_sme_za_array and ext_sme_addr_ri_u4xvl. * aarch64-opc.c (operand_general_constraint_met_p): (aarch64_match_operands_constraint): Handle sme_ldr, sme_str and sme_misc. (aarch64_print_operand): New operands supported. * aarch64-tbl.h (OP_SVE_QUU): New qualifier. (OP_SVE_QZU): New qualifier. aarch64-asm-2.c: Regenerate. aarch64-dis-2.c: Regenerate. aarch64-opc-2.c: Regenerate.
2021-11-18 04:02:06 +08:00
case AARCH64_OPND_SME_ADDR_RI_U4xVL:
if (!value_in_range_p (opnd->addr.offset.imm, 0, 15))
{
set_offset_out_of_range_error (mismatch_detail, idx, 0, 15);
return 0;
}
break;
[AArch64][SVE 26/32] Add SVE MUL VL addressing modes This patch adds support for addresses of the form: [<base>, #<offset>, MUL VL] This involves adding a new AARCH64_MOD_MUL_VL modifier, which is why I split it out from the other addressing modes. For LD2, LD3 and LD4, the offset must be a multiple of the structure size, so for LD3 the possible values are 0, 3, 6, .... The patch therefore extends value_aligned_p to handle non-power-of-2 alignments. include/ * opcode/aarch64.h (AARCH64_OPND_SVE_ADDR_RI_S4xVL): New aarch64_opnd. (AARCH64_OPND_SVE_ADDR_RI_S4x2xVL, AARCH64_OPND_SVE_ADDR_RI_S4x3xVL) (AARCH64_OPND_SVE_ADDR_RI_S4x4xVL, AARCH64_OPND_SVE_ADDR_RI_S6xVL) (AARCH64_OPND_SVE_ADDR_RI_S9xVL): Likewise. (AARCH64_MOD_MUL_VL): New aarch64_modifier_kind. opcodes/ * aarch64-tbl.h (AARCH64_OPERANDS): Add entries for new MUL VL operands. * aarch64-opc.c (aarch64_operand_modifiers): Initialize the AARCH64_MOD_MUL_VL entry. (value_aligned_p): Cope with non-power-of-two alignments. (operand_general_constraint_met_p): Handle the new MUL VL addresses. (print_immediate_offset_address): Likewise. (aarch64_print_operand): Likewise. * aarch64-opc-2.c: Regenerate. * aarch64-asm.h (ins_sve_addr_ri_s4xvl, ins_sve_addr_ri_s6xvl) (ins_sve_addr_ri_s9xvl): New inserters. * aarch64-asm.c (aarch64_ins_sve_addr_ri_s4xvl): New function. (aarch64_ins_sve_addr_ri_s6xvl): Likewise. (aarch64_ins_sve_addr_ri_s9xvl): Likewise. * aarch64-asm-2.c: Regenerate. * aarch64-dis.h (ext_sve_addr_ri_s4xvl, ext_sve_addr_ri_s6xvl) (ext_sve_addr_ri_s9xvl): New extractors. * aarch64-dis.c (aarch64_ext_sve_addr_reg_mul_vl): New function. (aarch64_ext_sve_addr_ri_s4xvl): Likewise. (aarch64_ext_sve_addr_ri_s6xvl): Likewise. (aarch64_ext_sve_addr_ri_s9xvl): Likewise. * aarch64-dis-2.c: Regenerate. gas/ * config/tc-aarch64.c (SHIFTED_NONE, SHIFTED_MUL_VL): New parse_shift_modes. (parse_shift): Handle SHIFTED_MUL_VL. (parse_address_main): Add an imm_shift_mode parameter. (parse_address, parse_sve_address): Update accordingly. (parse_operands): Handle MUL VL addressing modes.
2016-09-21 23:56:15 +08:00
case AARCH64_OPND_SVE_ADDR_RI_S4xVL:
case AARCH64_OPND_SVE_ADDR_RI_S4x2xVL:
case AARCH64_OPND_SVE_ADDR_RI_S4x3xVL:
case AARCH64_OPND_SVE_ADDR_RI_S4x4xVL:
min_value = -8;
max_value = 7;
sve_imm_offset_vl:
assert (!opnd->addr.offset.is_reg);
assert (opnd->addr.preind);
num = 1 + get_operand_specific_data (&aarch64_operands[type]);
min_value *= num;
max_value *= num;
if ((opnd->addr.offset.imm != 0 && !opnd->shifter.operator_present)
|| (opnd->shifter.operator_present
&& opnd->shifter.kind != AARCH64_MOD_MUL_VL))
{
set_other_error (mismatch_detail, idx,
_("invalid addressing mode"));
return 0;
}
if (!value_in_range_p (opnd->addr.offset.imm, min_value, max_value))
{
set_offset_out_of_range_error (mismatch_detail, idx,
min_value, max_value);
return 0;
}
if (!value_aligned_p (opnd->addr.offset.imm, num))
{
set_unaligned_error (mismatch_detail, idx, num);
return 0;
}
break;
case AARCH64_OPND_SVE_ADDR_RI_S6xVL:
min_value = -32;
max_value = 31;
goto sve_imm_offset_vl;
case AARCH64_OPND_SVE_ADDR_RI_S9xVL:
min_value = -256;
max_value = 255;
goto sve_imm_offset_vl;
[AArch64][SVE 25/32] Add support for SVE addressing modes This patch adds most of the new SVE addressing modes and associated operands. A follow-on patch adds MUL VL, since handling it separately makes the changes easier to read. The patch also introduces a new "operand-dependent data" field to the operand flags, based closely on the existing one for opcode flags. For SVE this new field needs only 2 bits, but it could be widened in future if necessary. include/ * opcode/aarch64.h (AARCH64_OPND_SVE_ADDR_RI_U6): New aarch64_opnd. (AARCH64_OPND_SVE_ADDR_RI_U6x2, AARCH64_OPND_SVE_ADDR_RI_U6x4) (AARCH64_OPND_SVE_ADDR_RI_U6x8, AARCH64_OPND_SVE_ADDR_RR) (AARCH64_OPND_SVE_ADDR_RR_LSL1, AARCH64_OPND_SVE_ADDR_RR_LSL2) (AARCH64_OPND_SVE_ADDR_RR_LSL3, AARCH64_OPND_SVE_ADDR_RX) (AARCH64_OPND_SVE_ADDR_RX_LSL1, AARCH64_OPND_SVE_ADDR_RX_LSL2) (AARCH64_OPND_SVE_ADDR_RX_LSL3, AARCH64_OPND_SVE_ADDR_RZ) (AARCH64_OPND_SVE_ADDR_RZ_LSL1, AARCH64_OPND_SVE_ADDR_RZ_LSL2) (AARCH64_OPND_SVE_ADDR_RZ_LSL3, AARCH64_OPND_SVE_ADDR_RZ_XTW_14) (AARCH64_OPND_SVE_ADDR_RZ_XTW_22, AARCH64_OPND_SVE_ADDR_RZ_XTW1_14) (AARCH64_OPND_SVE_ADDR_RZ_XTW1_22, AARCH64_OPND_SVE_ADDR_RZ_XTW2_14) (AARCH64_OPND_SVE_ADDR_RZ_XTW2_22, AARCH64_OPND_SVE_ADDR_RZ_XTW3_14) (AARCH64_OPND_SVE_ADDR_RZ_XTW3_22, AARCH64_OPND_SVE_ADDR_ZI_U5) (AARCH64_OPND_SVE_ADDR_ZI_U5x2, AARCH64_OPND_SVE_ADDR_ZI_U5x4) (AARCH64_OPND_SVE_ADDR_ZI_U5x8, AARCH64_OPND_SVE_ADDR_ZZ_LSL) (AARCH64_OPND_SVE_ADDR_ZZ_SXTW, AARCH64_OPND_SVE_ADDR_ZZ_UXTW): Likewise. opcodes/ * aarch64-tbl.h (AARCH64_OPERANDS): Add entries for the new SVE address operands. * aarch64-opc.h (FLD_SVE_imm6, FLD_SVE_msz, FLD_SVE_xs_14) (FLD_SVE_xs_22): New aarch64_field_kinds. (OPD_F_OD_MASK, OPD_F_OD_LSB, OPD_F_NO_ZR): New flags. (get_operand_specific_data): New function. * aarch64-opc.c (fields): Add entries for FLD_SVE_imm6, FLD_SVE_msz, FLD_SVE_xs_14 and FLD_SVE_xs_22. (operand_general_constraint_met_p): Handle the new SVE address operands. (sve_reg): New array. (get_addr_sve_reg_name): New function. (aarch64_print_operand): Handle the new SVE address operands. * aarch64-opc-2.c: Regenerate. * aarch64-asm.h (ins_sve_addr_ri_u6, ins_sve_addr_rr_lsl) (ins_sve_addr_rz_xtw, ins_sve_addr_zi_u5, ins_sve_addr_zz_lsl) (ins_sve_addr_zz_sxtw, ins_sve_addr_zz_uxtw): New inserters. * aarch64-asm.c (aarch64_ins_sve_addr_ri_u6): New function. (aarch64_ins_sve_addr_rr_lsl): Likewise. (aarch64_ins_sve_addr_rz_xtw): Likewise. (aarch64_ins_sve_addr_zi_u5): Likewise. (aarch64_ins_sve_addr_zz): Likewise. (aarch64_ins_sve_addr_zz_lsl): Likewise. (aarch64_ins_sve_addr_zz_sxtw): Likewise. (aarch64_ins_sve_addr_zz_uxtw): Likewise. * aarch64-asm-2.c: Regenerate. * aarch64-dis.h (ext_sve_addr_ri_u6, ext_sve_addr_rr_lsl) (ext_sve_addr_rz_xtw, ext_sve_addr_zi_u5, ext_sve_addr_zz_lsl) (ext_sve_addr_zz_sxtw, ext_sve_addr_zz_uxtw): New extractors. * aarch64-dis.c (aarch64_ext_sve_add_reg_imm): New function. (aarch64_ext_sve_addr_ri_u6): Likewise. (aarch64_ext_sve_addr_rr_lsl): Likewise. (aarch64_ext_sve_addr_rz_xtw): Likewise. (aarch64_ext_sve_addr_zi_u5): Likewise. (aarch64_ext_sve_addr_zz): Likewise. (aarch64_ext_sve_addr_zz_lsl): Likewise. (aarch64_ext_sve_addr_zz_sxtw): Likewise. (aarch64_ext_sve_addr_zz_uxtw): Likewise. * aarch64-dis-2.c: Regenerate. gas/ * config/tc-aarch64.c (REG_TYPE_SVE_BASE, REG_TYPE_SVE_OFFSET): New register types. (get_reg_expected_msg): Handle them. (aarch64_addr_reg_parse): New function, split out from aarch64_reg_parse_32_64. Handle Z registers too. (aarch64_reg_parse_32_64): Call it. (parse_address_main): Add base_qualifier, offset_qualifier, base_type and offset_type parameters. Handle SVE base and offset registers. (parse_address): Update call to parse_address_main. (parse_sve_address): New function. (parse_operands): Parse the new SVE address operands.
2016-09-21 23:55:49 +08:00
case AARCH64_OPND_SVE_ADDR_RI_U6:
case AARCH64_OPND_SVE_ADDR_RI_U6x2:
case AARCH64_OPND_SVE_ADDR_RI_U6x4:
case AARCH64_OPND_SVE_ADDR_RI_U6x8:
min_value = 0;
max_value = 63;
sve_imm_offset:
assert (!opnd->addr.offset.is_reg);
assert (opnd->addr.preind);
num = 1 << get_operand_specific_data (&aarch64_operands[type]);
min_value *= num;
max_value *= num;
if (opnd->shifter.operator_present
|| opnd->shifter.amount_present)
{
set_other_error (mismatch_detail, idx,
_("invalid addressing mode"));
return 0;
}
if (!value_in_range_p (opnd->addr.offset.imm, min_value, max_value))
{
set_offset_out_of_range_error (mismatch_detail, idx,
min_value, max_value);
return 0;
}
if (!value_aligned_p (opnd->addr.offset.imm, num))
{
set_unaligned_error (mismatch_detail, idx, num);
return 0;
}
break;
[AArch64] Additional SVE instructions This patch supports some additions to the SVE architecture prior to its public release. include/ * opcode/aarch64.h (AARCH64_OPND_SVE_ADDR_RI_S4x16) (AARCH64_OPND_SVE_IMM_ROT1, AARCH64_OPND_SVE_IMM_ROT2) (AARCH64_OPND_SVE_Zm3_INDEX, AARCH64_OPND_SVE_Zm3_22_INDEX) (AARCH64_OPND_SVE_Zm4_INDEX): New aarch64_opnds. opcodes/ * aarch64-tbl.h (OP_SVE_HMH, OP_SVE_VMU_HSD, OP_SVE_VMVU_HSD) (OP_SVE_VMVV_HSD, OP_SVE_VMVVU_HSD, OP_SVE_VM_HSD, OP_SVE_VUVV_HSD) (OP_SVE_VUV_HSD, OP_SVE_VU_HSD, OP_SVE_VVVU_H, OP_SVE_VVVU_S) (OP_SVE_VVVU_HSD, OP_SVE_VVV_D, OP_SVE_VVV_D_H, OP_SVE_VVV_H) (OP_SVE_VVV_HSD, OP_SVE_VVV_S, OP_SVE_VVV_S_B, OP_SVE_VVV_SD_BH) (OP_SVE_VV_BHSDQ, OP_SVE_VV_HSD, OP_SVE_VZVV_HSD, OP_SVE_VZV_HSD) (OP_SVE_V_HSD): New macros. (OP_SVE_VMU_SD, OP_SVE_VMVU_SD, OP_SVE_VM_SD, OP_SVE_VUVV_SD) (OP_SVE_VU_SD, OP_SVE_VVVU_SD, OP_SVE_VVV_SD, OP_SVE_VZVV_SD) (OP_SVE_VZV_SD, OP_SVE_V_SD): Delete. (aarch64_opcode_table): Add new SVE instructions. (aarch64_opcode_table): Use imm_rotate{1,2} instead of imm_rotate for rotation operands. Add new SVE operands. * aarch64-asm.h (ins_sve_addr_ri_s4): New inserter. (ins_sve_quad_index): Likewise. (ins_imm_rotate): Split into... (ins_imm_rotate1, ins_imm_rotate2): ...these two inserters. * aarch64-asm.c (aarch64_ins_imm_rotate): Split into... (aarch64_ins_imm_rotate1, aarch64_ins_imm_rotate2): ...these two functions. (aarch64_ins_sve_addr_ri_s4): New function. (aarch64_ins_sve_quad_index): Likewise. (do_misc_encoding): Handle "MOV Zn.Q, Qm". * aarch64-asm-2.c: Regenerate. * aarch64-dis.h (ext_sve_addr_ri_s4): New extractor. (ext_sve_quad_index): Likewise. (ext_imm_rotate): Split into... (ext_imm_rotate1, ext_imm_rotate2): ...these two extractors. * aarch64-dis.c (aarch64_ext_imm_rotate): Split into... (aarch64_ext_imm_rotate1, aarch64_ext_imm_rotate2): ...these two functions. (aarch64_ext_sve_addr_ri_s4): New function. (aarch64_ext_sve_quad_index): Likewise. (aarch64_ext_sve_index): Allow quad indices. (do_misc_decoding): Likewise. * aarch64-dis-2.c: Regenerate. * aarch64-opc.h (FLD_SVE_i3h, FLD_SVE_rot1, FLD_SVE_rot2): New aarch64_field_kinds. (OPD_F_OD_MASK): Widen by one bit. (OPD_F_NO_ZR): Bump accordingly. (get_operand_field_width): New function. * aarch64-opc.c (fields): Add new SVE fields. (operand_general_constraint_met_p): Handle new SVE operands. (aarch64_print_operand): Likewise. * aarch64-opc-2.c: Regenerate. gas/ * doc/c-aarch64.texi: Document that sve implies fp16, simd and compnum. * config/tc-aarch64.c (parse_vector_type_for_operand): Allow .q to be used with SVE registers. (parse_operands): Handle new SVE operands. (aarch64_features): Make "sve" require F16 rather than FP. Also require COMPNUM. * testsuite/gas/aarch64/sve.s: Add tests for new instructions. Include compnum tests. * testsuite/gas/aarch64/sve.d: Update accordingly. * testsuite/gas/aarch64/sve-invalid.s: Add tests for new instructions. * testsuite/gas/aarch64/sve-invalid.l: Update accordingly. Also update expected output for new FMOV and MOV alternatives.
2017-02-25 02:29:00 +08:00
case AARCH64_OPND_SVE_ADDR_RI_S4x16:
[binutils][aarch64] Matrix Multiply extension enablement [8/X] Hi, This patch is part of a series that adds support for Armv8.6-A (Matrix Multiply and BFloat16 extensions) to binutils. This patch introduces the Matrix Multiply (Int8, F32, F64) extensions to the aarch64 backend. The following instructions are added: {s/u}mmla, usmmla, {us/su}dot, fmmla, ld1rob, ld1roh, d1row, ld1rod, uzip{1/2}, trn{1/2}. Committed on behalf of Mihail Ionescu. gas/ChangeLog: 2019-11-07 Mihail Ionescu <mihail.ionescu@arm.com> * config/tc-aarch64.c: Add new arch fetures to suppport the mm extension. (parse_operands): Add new operand. * testsuite/gas/aarch64/i8mm.s: New test. * testsuite/gas/aarch64/i8mm.d: New test. * testsuite/gas/aarch64/f32mm.s: New test. * testsuite/gas/aarch64/f32mm.d: New test. * testsuite/gas/aarch64/f64mm.s: New test. * testsuite/gas/aarch64/f64mm.d: New test. * testsuite/gas/aarch64/sve-movprfx-mm.s: New test. * testsuite/gas/aarch64/sve-movprfx-mm.d: New test. include/ChangeLog: 2019-11-07 Mihail Ionescu <mihail.ionescu@arm.com> * opcode/aarch64.h (AARCH64_FEATURE_I8MM): New. (AARCH64_FEATURE_F32MM): New. (AARCH64_FEATURE_F64MM): New. (AARCH64_OPND_SVE_ADDR_RI_S4x32): New. (enum aarch64_insn_class): Add new instruction class "aarch64_misc" for instructions that do not require special handling. opcodes/ChangeLog: 2019-11-07 Mihail Ionescu <mihail.ionescu@arm.com> * aarch64-tbl.h (aarch64_feature_i8mm_sve, aarch64_feature_f32mm_sve, aarch64_feature_f64mm_sve, aarch64_feature_i8mm, aarch64_feature_f32mm, aarch64_feature_f64mm): New feature sets. (INT8MATMUL_INSN, F64MATMUL_SVE_INSN, F64MATMUL_INSN, F32MATMUL_SVE_INSN, F32MATMUL_INSN): New macros to define matrix multiply instructions. (I8MM_SVE, F32MM_SVE, F64MM_SVE, I8MM, F32MM, F64MM): New feature set macros. (QL_MMLA64, OP_SVE_SBB): New qualifiers. (OP_SVE_QQQ): New qualifier. (INT8MATMUL_SVE_INSNC, F64MATMUL_SVE_INSNC, F32MATMUL_SVE_INSNC): New feature set for bfloat16 instructions to support the movprfx constraint. (aarch64_opcode_table): Support for SVE_ADDR_RI_S4x32. (aarch64_opcode_table): Define new instructions smmla, ummla, usmmla, usdot, sudot, fmmla, ld1rob, ld1roh, ld1row, ld1rod uzip{1/2}, trn{1/2}. * aarch64-opc.c (operand_general_constraint_met_p): Handle AARCH64_OPND_SVE_ADDR_RI_S4x32. (aarch64_print_operand): Handle AARCH64_OPND_SVE_ADDR_RI_S4x32. * aarch64-dis-2.c (aarch64_opcode_lookup_1, aarch64_find_next_opcode): Account for new instructions. * opcodes/aarch64-asm-2.c (aarch64_insert_operand): Support the new S4x32 operand. * aarch64-opc-2.c (aarch64_operands): Support the new S4x32 operand. Regression tested on arm-none-eabi. Is it ok for trunk? Regards, Mihail
2019-11-08 01:10:01 +08:00
case AARCH64_OPND_SVE_ADDR_RI_S4x32:
[AArch64] Additional SVE instructions This patch supports some additions to the SVE architecture prior to its public release. include/ * opcode/aarch64.h (AARCH64_OPND_SVE_ADDR_RI_S4x16) (AARCH64_OPND_SVE_IMM_ROT1, AARCH64_OPND_SVE_IMM_ROT2) (AARCH64_OPND_SVE_Zm3_INDEX, AARCH64_OPND_SVE_Zm3_22_INDEX) (AARCH64_OPND_SVE_Zm4_INDEX): New aarch64_opnds. opcodes/ * aarch64-tbl.h (OP_SVE_HMH, OP_SVE_VMU_HSD, OP_SVE_VMVU_HSD) (OP_SVE_VMVV_HSD, OP_SVE_VMVVU_HSD, OP_SVE_VM_HSD, OP_SVE_VUVV_HSD) (OP_SVE_VUV_HSD, OP_SVE_VU_HSD, OP_SVE_VVVU_H, OP_SVE_VVVU_S) (OP_SVE_VVVU_HSD, OP_SVE_VVV_D, OP_SVE_VVV_D_H, OP_SVE_VVV_H) (OP_SVE_VVV_HSD, OP_SVE_VVV_S, OP_SVE_VVV_S_B, OP_SVE_VVV_SD_BH) (OP_SVE_VV_BHSDQ, OP_SVE_VV_HSD, OP_SVE_VZVV_HSD, OP_SVE_VZV_HSD) (OP_SVE_V_HSD): New macros. (OP_SVE_VMU_SD, OP_SVE_VMVU_SD, OP_SVE_VM_SD, OP_SVE_VUVV_SD) (OP_SVE_VU_SD, OP_SVE_VVVU_SD, OP_SVE_VVV_SD, OP_SVE_VZVV_SD) (OP_SVE_VZV_SD, OP_SVE_V_SD): Delete. (aarch64_opcode_table): Add new SVE instructions. (aarch64_opcode_table): Use imm_rotate{1,2} instead of imm_rotate for rotation operands. Add new SVE operands. * aarch64-asm.h (ins_sve_addr_ri_s4): New inserter. (ins_sve_quad_index): Likewise. (ins_imm_rotate): Split into... (ins_imm_rotate1, ins_imm_rotate2): ...these two inserters. * aarch64-asm.c (aarch64_ins_imm_rotate): Split into... (aarch64_ins_imm_rotate1, aarch64_ins_imm_rotate2): ...these two functions. (aarch64_ins_sve_addr_ri_s4): New function. (aarch64_ins_sve_quad_index): Likewise. (do_misc_encoding): Handle "MOV Zn.Q, Qm". * aarch64-asm-2.c: Regenerate. * aarch64-dis.h (ext_sve_addr_ri_s4): New extractor. (ext_sve_quad_index): Likewise. (ext_imm_rotate): Split into... (ext_imm_rotate1, ext_imm_rotate2): ...these two extractors. * aarch64-dis.c (aarch64_ext_imm_rotate): Split into... (aarch64_ext_imm_rotate1, aarch64_ext_imm_rotate2): ...these two functions. (aarch64_ext_sve_addr_ri_s4): New function. (aarch64_ext_sve_quad_index): Likewise. (aarch64_ext_sve_index): Allow quad indices. (do_misc_decoding): Likewise. * aarch64-dis-2.c: Regenerate. * aarch64-opc.h (FLD_SVE_i3h, FLD_SVE_rot1, FLD_SVE_rot2): New aarch64_field_kinds. (OPD_F_OD_MASK): Widen by one bit. (OPD_F_NO_ZR): Bump accordingly. (get_operand_field_width): New function. * aarch64-opc.c (fields): Add new SVE fields. (operand_general_constraint_met_p): Handle new SVE operands. (aarch64_print_operand): Likewise. * aarch64-opc-2.c: Regenerate. gas/ * doc/c-aarch64.texi: Document that sve implies fp16, simd and compnum. * config/tc-aarch64.c (parse_vector_type_for_operand): Allow .q to be used with SVE registers. (parse_operands): Handle new SVE operands. (aarch64_features): Make "sve" require F16 rather than FP. Also require COMPNUM. * testsuite/gas/aarch64/sve.s: Add tests for new instructions. Include compnum tests. * testsuite/gas/aarch64/sve.d: Update accordingly. * testsuite/gas/aarch64/sve-invalid.s: Add tests for new instructions. * testsuite/gas/aarch64/sve-invalid.l: Update accordingly. Also update expected output for new FMOV and MOV alternatives.
2017-02-25 02:29:00 +08:00
min_value = -8;
max_value = 7;
goto sve_imm_offset;
case AARCH64_OPND_SVE_ADDR_ZX:
/* Everything is already ensured by parse_operands or
aarch64_ext_sve_addr_rr_lsl (because this is a very specific
argument type). */
assert (opnd->addr.offset.is_reg);
assert (opnd->addr.preind);
assert ((aarch64_operands[type].flags & OPD_F_NO_ZR) == 0);
assert (opnd->shifter.kind == AARCH64_MOD_LSL);
assert (opnd->shifter.operator_present == 0);
break;
case AARCH64_OPND_SVE_ADDR_R:
[AArch64][SVE 25/32] Add support for SVE addressing modes This patch adds most of the new SVE addressing modes and associated operands. A follow-on patch adds MUL VL, since handling it separately makes the changes easier to read. The patch also introduces a new "operand-dependent data" field to the operand flags, based closely on the existing one for opcode flags. For SVE this new field needs only 2 bits, but it could be widened in future if necessary. include/ * opcode/aarch64.h (AARCH64_OPND_SVE_ADDR_RI_U6): New aarch64_opnd. (AARCH64_OPND_SVE_ADDR_RI_U6x2, AARCH64_OPND_SVE_ADDR_RI_U6x4) (AARCH64_OPND_SVE_ADDR_RI_U6x8, AARCH64_OPND_SVE_ADDR_RR) (AARCH64_OPND_SVE_ADDR_RR_LSL1, AARCH64_OPND_SVE_ADDR_RR_LSL2) (AARCH64_OPND_SVE_ADDR_RR_LSL3, AARCH64_OPND_SVE_ADDR_RX) (AARCH64_OPND_SVE_ADDR_RX_LSL1, AARCH64_OPND_SVE_ADDR_RX_LSL2) (AARCH64_OPND_SVE_ADDR_RX_LSL3, AARCH64_OPND_SVE_ADDR_RZ) (AARCH64_OPND_SVE_ADDR_RZ_LSL1, AARCH64_OPND_SVE_ADDR_RZ_LSL2) (AARCH64_OPND_SVE_ADDR_RZ_LSL3, AARCH64_OPND_SVE_ADDR_RZ_XTW_14) (AARCH64_OPND_SVE_ADDR_RZ_XTW_22, AARCH64_OPND_SVE_ADDR_RZ_XTW1_14) (AARCH64_OPND_SVE_ADDR_RZ_XTW1_22, AARCH64_OPND_SVE_ADDR_RZ_XTW2_14) (AARCH64_OPND_SVE_ADDR_RZ_XTW2_22, AARCH64_OPND_SVE_ADDR_RZ_XTW3_14) (AARCH64_OPND_SVE_ADDR_RZ_XTW3_22, AARCH64_OPND_SVE_ADDR_ZI_U5) (AARCH64_OPND_SVE_ADDR_ZI_U5x2, AARCH64_OPND_SVE_ADDR_ZI_U5x4) (AARCH64_OPND_SVE_ADDR_ZI_U5x8, AARCH64_OPND_SVE_ADDR_ZZ_LSL) (AARCH64_OPND_SVE_ADDR_ZZ_SXTW, AARCH64_OPND_SVE_ADDR_ZZ_UXTW): Likewise. opcodes/ * aarch64-tbl.h (AARCH64_OPERANDS): Add entries for the new SVE address operands. * aarch64-opc.h (FLD_SVE_imm6, FLD_SVE_msz, FLD_SVE_xs_14) (FLD_SVE_xs_22): New aarch64_field_kinds. (OPD_F_OD_MASK, OPD_F_OD_LSB, OPD_F_NO_ZR): New flags. (get_operand_specific_data): New function. * aarch64-opc.c (fields): Add entries for FLD_SVE_imm6, FLD_SVE_msz, FLD_SVE_xs_14 and FLD_SVE_xs_22. (operand_general_constraint_met_p): Handle the new SVE address operands. (sve_reg): New array. (get_addr_sve_reg_name): New function. (aarch64_print_operand): Handle the new SVE address operands. * aarch64-opc-2.c: Regenerate. * aarch64-asm.h (ins_sve_addr_ri_u6, ins_sve_addr_rr_lsl) (ins_sve_addr_rz_xtw, ins_sve_addr_zi_u5, ins_sve_addr_zz_lsl) (ins_sve_addr_zz_sxtw, ins_sve_addr_zz_uxtw): New inserters. * aarch64-asm.c (aarch64_ins_sve_addr_ri_u6): New function. (aarch64_ins_sve_addr_rr_lsl): Likewise. (aarch64_ins_sve_addr_rz_xtw): Likewise. (aarch64_ins_sve_addr_zi_u5): Likewise. (aarch64_ins_sve_addr_zz): Likewise. (aarch64_ins_sve_addr_zz_lsl): Likewise. (aarch64_ins_sve_addr_zz_sxtw): Likewise. (aarch64_ins_sve_addr_zz_uxtw): Likewise. * aarch64-asm-2.c: Regenerate. * aarch64-dis.h (ext_sve_addr_ri_u6, ext_sve_addr_rr_lsl) (ext_sve_addr_rz_xtw, ext_sve_addr_zi_u5, ext_sve_addr_zz_lsl) (ext_sve_addr_zz_sxtw, ext_sve_addr_zz_uxtw): New extractors. * aarch64-dis.c (aarch64_ext_sve_add_reg_imm): New function. (aarch64_ext_sve_addr_ri_u6): Likewise. (aarch64_ext_sve_addr_rr_lsl): Likewise. (aarch64_ext_sve_addr_rz_xtw): Likewise. (aarch64_ext_sve_addr_zi_u5): Likewise. (aarch64_ext_sve_addr_zz): Likewise. (aarch64_ext_sve_addr_zz_lsl): Likewise. (aarch64_ext_sve_addr_zz_sxtw): Likewise. (aarch64_ext_sve_addr_zz_uxtw): Likewise. * aarch64-dis-2.c: Regenerate. gas/ * config/tc-aarch64.c (REG_TYPE_SVE_BASE, REG_TYPE_SVE_OFFSET): New register types. (get_reg_expected_msg): Handle them. (aarch64_addr_reg_parse): New function, split out from aarch64_reg_parse_32_64. Handle Z registers too. (aarch64_reg_parse_32_64): Call it. (parse_address_main): Add base_qualifier, offset_qualifier, base_type and offset_type parameters. Handle SVE base and offset registers. (parse_address): Update call to parse_address_main. (parse_sve_address): New function. (parse_operands): Parse the new SVE address operands.
2016-09-21 23:55:49 +08:00
case AARCH64_OPND_SVE_ADDR_RR:
case AARCH64_OPND_SVE_ADDR_RR_LSL1:
case AARCH64_OPND_SVE_ADDR_RR_LSL2:
case AARCH64_OPND_SVE_ADDR_RR_LSL3:
aarch64: [SME] Add LD1x, ST1x, LDR and STR instructions This patch is adding new loads and stores defined by SME instructions. gas/ChangeLog: * config/tc-aarch64.c (parse_sme_address): New parser. (parse_sme_za_hv_tiles_operand_with_braces): New parser. (parse_sme_za_array): New parser. (output_operand_error_record): Print error details if present. (parse_operands): Support new operands. * testsuite/gas/aarch64/sme-5-illegal.d: New test. * testsuite/gas/aarch64/sme-5-illegal.l: New test. * testsuite/gas/aarch64/sme-5-illegal.s: New test. * testsuite/gas/aarch64/sme-5.d: New test. * testsuite/gas/aarch64/sme-5.s: New test. * testsuite/gas/aarch64/sme-6-illegal.d: New test. * testsuite/gas/aarch64/sme-6-illegal.l: New test. * testsuite/gas/aarch64/sme-6-illegal.s: New test. * testsuite/gas/aarch64/sme-6.d: New test. * testsuite/gas/aarch64/sme-6.s: New test. * testsuite/gas/aarch64/sme-7-illegal.d: New test. * testsuite/gas/aarch64/sme-7-illegal.l: New test. * testsuite/gas/aarch64/sme-7-illegal.s: New test. * testsuite/gas/aarch64/sme-7.d: New test. * testsuite/gas/aarch64/sme-7.s: New test. include/ChangeLog: * opcode/aarch64.h (enum aarch64_opnd): New operands. (enum aarch64_insn_class): Added sme_ldr and sme_str. (AARCH64_OPDE_UNTIED_IMMS): New operand error kind. opcodes/ChangeLog: * aarch64-asm.c (aarch64_ins_sme_za_hv_tiles): New inserter. (aarch64_ins_sme_za_list): New inserter. (aarch64_ins_sme_za_array): New inserter. (aarch64_ins_sme_addr_ri_u4xvl): New inserter. * aarch64-asm.h (AARCH64_DECL_OPD_INSERTER): Added ins_sme_za_list, ins_sme_za_array and ins_sme_addr_ri_u4xvl. * aarch64-dis.c (aarch64_ext_sme_za_hv_tiles): New extractor. (aarch64_ext_sme_za_list): New extractor. (aarch64_ext_sme_za_array): New extractor. (aarch64_ext_sme_addr_ri_u4xvl): New extractor. * aarch64-dis.h (AARCH64_DECL_OPD_EXTRACTOR): Added ext_sme_za_list, ext_sme_za_array and ext_sme_addr_ri_u4xvl. * aarch64-opc.c (operand_general_constraint_met_p): (aarch64_match_operands_constraint): Handle sme_ldr, sme_str and sme_misc. (aarch64_print_operand): New operands supported. * aarch64-tbl.h (OP_SVE_QUU): New qualifier. (OP_SVE_QZU): New qualifier. aarch64-asm-2.c: Regenerate. aarch64-dis-2.c: Regenerate. aarch64-opc-2.c: Regenerate.
2021-11-18 04:02:06 +08:00
case AARCH64_OPND_SVE_ADDR_RR_LSL4:
[AArch64][SVE 25/32] Add support for SVE addressing modes This patch adds most of the new SVE addressing modes and associated operands. A follow-on patch adds MUL VL, since handling it separately makes the changes easier to read. The patch also introduces a new "operand-dependent data" field to the operand flags, based closely on the existing one for opcode flags. For SVE this new field needs only 2 bits, but it could be widened in future if necessary. include/ * opcode/aarch64.h (AARCH64_OPND_SVE_ADDR_RI_U6): New aarch64_opnd. (AARCH64_OPND_SVE_ADDR_RI_U6x2, AARCH64_OPND_SVE_ADDR_RI_U6x4) (AARCH64_OPND_SVE_ADDR_RI_U6x8, AARCH64_OPND_SVE_ADDR_RR) (AARCH64_OPND_SVE_ADDR_RR_LSL1, AARCH64_OPND_SVE_ADDR_RR_LSL2) (AARCH64_OPND_SVE_ADDR_RR_LSL3, AARCH64_OPND_SVE_ADDR_RX) (AARCH64_OPND_SVE_ADDR_RX_LSL1, AARCH64_OPND_SVE_ADDR_RX_LSL2) (AARCH64_OPND_SVE_ADDR_RX_LSL3, AARCH64_OPND_SVE_ADDR_RZ) (AARCH64_OPND_SVE_ADDR_RZ_LSL1, AARCH64_OPND_SVE_ADDR_RZ_LSL2) (AARCH64_OPND_SVE_ADDR_RZ_LSL3, AARCH64_OPND_SVE_ADDR_RZ_XTW_14) (AARCH64_OPND_SVE_ADDR_RZ_XTW_22, AARCH64_OPND_SVE_ADDR_RZ_XTW1_14) (AARCH64_OPND_SVE_ADDR_RZ_XTW1_22, AARCH64_OPND_SVE_ADDR_RZ_XTW2_14) (AARCH64_OPND_SVE_ADDR_RZ_XTW2_22, AARCH64_OPND_SVE_ADDR_RZ_XTW3_14) (AARCH64_OPND_SVE_ADDR_RZ_XTW3_22, AARCH64_OPND_SVE_ADDR_ZI_U5) (AARCH64_OPND_SVE_ADDR_ZI_U5x2, AARCH64_OPND_SVE_ADDR_ZI_U5x4) (AARCH64_OPND_SVE_ADDR_ZI_U5x8, AARCH64_OPND_SVE_ADDR_ZZ_LSL) (AARCH64_OPND_SVE_ADDR_ZZ_SXTW, AARCH64_OPND_SVE_ADDR_ZZ_UXTW): Likewise. opcodes/ * aarch64-tbl.h (AARCH64_OPERANDS): Add entries for the new SVE address operands. * aarch64-opc.h (FLD_SVE_imm6, FLD_SVE_msz, FLD_SVE_xs_14) (FLD_SVE_xs_22): New aarch64_field_kinds. (OPD_F_OD_MASK, OPD_F_OD_LSB, OPD_F_NO_ZR): New flags. (get_operand_specific_data): New function. * aarch64-opc.c (fields): Add entries for FLD_SVE_imm6, FLD_SVE_msz, FLD_SVE_xs_14 and FLD_SVE_xs_22. (operand_general_constraint_met_p): Handle the new SVE address operands. (sve_reg): New array. (get_addr_sve_reg_name): New function. (aarch64_print_operand): Handle the new SVE address operands. * aarch64-opc-2.c: Regenerate. * aarch64-asm.h (ins_sve_addr_ri_u6, ins_sve_addr_rr_lsl) (ins_sve_addr_rz_xtw, ins_sve_addr_zi_u5, ins_sve_addr_zz_lsl) (ins_sve_addr_zz_sxtw, ins_sve_addr_zz_uxtw): New inserters. * aarch64-asm.c (aarch64_ins_sve_addr_ri_u6): New function. (aarch64_ins_sve_addr_rr_lsl): Likewise. (aarch64_ins_sve_addr_rz_xtw): Likewise. (aarch64_ins_sve_addr_zi_u5): Likewise. (aarch64_ins_sve_addr_zz): Likewise. (aarch64_ins_sve_addr_zz_lsl): Likewise. (aarch64_ins_sve_addr_zz_sxtw): Likewise. (aarch64_ins_sve_addr_zz_uxtw): Likewise. * aarch64-asm-2.c: Regenerate. * aarch64-dis.h (ext_sve_addr_ri_u6, ext_sve_addr_rr_lsl) (ext_sve_addr_rz_xtw, ext_sve_addr_zi_u5, ext_sve_addr_zz_lsl) (ext_sve_addr_zz_sxtw, ext_sve_addr_zz_uxtw): New extractors. * aarch64-dis.c (aarch64_ext_sve_add_reg_imm): New function. (aarch64_ext_sve_addr_ri_u6): Likewise. (aarch64_ext_sve_addr_rr_lsl): Likewise. (aarch64_ext_sve_addr_rz_xtw): Likewise. (aarch64_ext_sve_addr_zi_u5): Likewise. (aarch64_ext_sve_addr_zz): Likewise. (aarch64_ext_sve_addr_zz_lsl): Likewise. (aarch64_ext_sve_addr_zz_sxtw): Likewise. (aarch64_ext_sve_addr_zz_uxtw): Likewise. * aarch64-dis-2.c: Regenerate. gas/ * config/tc-aarch64.c (REG_TYPE_SVE_BASE, REG_TYPE_SVE_OFFSET): New register types. (get_reg_expected_msg): Handle them. (aarch64_addr_reg_parse): New function, split out from aarch64_reg_parse_32_64. Handle Z registers too. (aarch64_reg_parse_32_64): Call it. (parse_address_main): Add base_qualifier, offset_qualifier, base_type and offset_type parameters. Handle SVE base and offset registers. (parse_address): Update call to parse_address_main. (parse_sve_address): New function. (parse_operands): Parse the new SVE address operands.
2016-09-21 23:55:49 +08:00
case AARCH64_OPND_SVE_ADDR_RX:
case AARCH64_OPND_SVE_ADDR_RX_LSL1:
case AARCH64_OPND_SVE_ADDR_RX_LSL2:
case AARCH64_OPND_SVE_ADDR_RX_LSL3:
case AARCH64_OPND_SVE_ADDR_RZ:
case AARCH64_OPND_SVE_ADDR_RZ_LSL1:
case AARCH64_OPND_SVE_ADDR_RZ_LSL2:
case AARCH64_OPND_SVE_ADDR_RZ_LSL3:
modifiers = 1 << AARCH64_MOD_LSL;
sve_rr_operand:
assert (opnd->addr.offset.is_reg);
assert (opnd->addr.preind);
if ((aarch64_operands[type].flags & OPD_F_NO_ZR) != 0
&& opnd->addr.offset.regno == 31)
{
set_other_error (mismatch_detail, idx,
_("index register xzr is not allowed"));
return 0;
}
if (((1 << opnd->shifter.kind) & modifiers) == 0
|| (opnd->shifter.amount
!= get_operand_specific_data (&aarch64_operands[type])))
{
set_other_error (mismatch_detail, idx,
_("invalid addressing mode"));
return 0;
}
break;
case AARCH64_OPND_SVE_ADDR_RZ_XTW_14:
case AARCH64_OPND_SVE_ADDR_RZ_XTW_22:
case AARCH64_OPND_SVE_ADDR_RZ_XTW1_14:
case AARCH64_OPND_SVE_ADDR_RZ_XTW1_22:
case AARCH64_OPND_SVE_ADDR_RZ_XTW2_14:
case AARCH64_OPND_SVE_ADDR_RZ_XTW2_22:
case AARCH64_OPND_SVE_ADDR_RZ_XTW3_14:
case AARCH64_OPND_SVE_ADDR_RZ_XTW3_22:
modifiers = (1 << AARCH64_MOD_SXTW) | (1 << AARCH64_MOD_UXTW);
goto sve_rr_operand;
case AARCH64_OPND_SVE_ADDR_ZI_U5:
case AARCH64_OPND_SVE_ADDR_ZI_U5x2:
case AARCH64_OPND_SVE_ADDR_ZI_U5x4:
case AARCH64_OPND_SVE_ADDR_ZI_U5x8:
min_value = 0;
max_value = 31;
goto sve_imm_offset;
case AARCH64_OPND_SVE_ADDR_ZZ_LSL:
modifiers = 1 << AARCH64_MOD_LSL;
sve_zz_operand:
assert (opnd->addr.offset.is_reg);
assert (opnd->addr.preind);
if (((1 << opnd->shifter.kind) & modifiers) == 0
|| opnd->shifter.amount < 0
|| opnd->shifter.amount > 3)
{
set_other_error (mismatch_detail, idx,
_("invalid addressing mode"));
return 0;
}
break;
case AARCH64_OPND_SVE_ADDR_ZZ_SXTW:
modifiers = (1 << AARCH64_MOD_SXTW);
goto sve_zz_operand;
case AARCH64_OPND_SVE_ADDR_ZZ_UXTW:
modifiers = 1 << AARCH64_MOD_UXTW;
goto sve_zz_operand;
case AARCH64_OPND_RCPC3_ADDR_OPT_PREIND_WB:
case AARCH64_OPND_RCPC3_ADDR_OPT_POSTIND:
case AARCH64_OPND_RCPC3_ADDR_PREIND_WB:
case AARCH64_OPND_RCPC3_ADDR_POSTIND:
{
int num_bytes = calc_ldst_datasize (opnds);
int abs_offset = (type == AARCH64_OPND_RCPC3_ADDR_OPT_PREIND_WB
|| type == AARCH64_OPND_RCPC3_ADDR_PREIND_WB)
? opnd->addr.offset.imm * -1
: opnd->addr.offset.imm;
if ((int) num_bytes != abs_offset
&& opnd->addr.offset.imm != 0)
{
set_other_error (mismatch_detail, idx,
_("invalid increment amount"));
return 0;
}
}
break;
case AARCH64_OPND_RCPC3_ADDR_OFFSET:
if (!value_in_range_p (opnd->addr.offset.imm, -256, 255))
{
set_imm_out_of_range_error (mismatch_detail, idx, -256, 255);
return 0;
}
default:
break;
}
break;
case AARCH64_OPND_CLASS_SIMD_REGLIST:
if (type == AARCH64_OPND_LEt)
{
/* Get the upper bound for the element index. */
num = 16 / aarch64_get_qualifier_esize (qualifier) - 1;
if (!value_in_range_p (opnd->reglist.index, 0, num))
{
set_elem_idx_out_of_range_error (mismatch_detail, idx, 0, num);
return 0;
}
}
/* The opcode dependent area stores the number of elements in
each structure to be loaded/stored. */
num = get_opcode_dependent_value (opcode);
switch (type)
{
case AARCH64_OPND_LVt:
assert (num >= 1 && num <= 4);
/* Unless LD1/ST1, the number of registers should be equal to that
of the structure elements. */
if (num != 1 && !check_reglist (opnd, mismatch_detail, idx, num, 1))
return 0;
break;
case AARCH64_OPND_LVt_AL:
case AARCH64_OPND_LEt:
assert (num >= 1 && num <= 4);
/* The number of registers should be equal to that of the structure
elements. */
if (!check_reglist (opnd, mismatch_detail, idx, num, 1))
return 0;
break;
default:
break;
}
if (opnd->reglist.stride != 1)
{
set_reg_list_stride_error (mismatch_detail, idx, 1);
return 0;
}
break;
case AARCH64_OPND_CLASS_IMMEDIATE:
/* Constraint check on immediate operand. */
imm = opnd->imm.value;
/* E.g. imm_0_31 constrains value to be 0..31. */
if (qualifier_value_in_range_constraint_p (qualifier)
&& !value_in_range_p (imm, get_lower_bound (qualifier),
get_upper_bound (qualifier)))
{
set_imm_out_of_range_error (mismatch_detail, idx,
get_lower_bound (qualifier),
get_upper_bound (qualifier));
return 0;
}
switch (type)
{
case AARCH64_OPND_AIMM:
if (opnd->shifter.kind != AARCH64_MOD_LSL)
{
set_other_error (mismatch_detail, idx,
_("invalid shift operator"));
return 0;
}
if (opnd->shifter.amount != 0 && opnd->shifter.amount != 12)
{
set_other_error (mismatch_detail, idx,
_("shift amount must be 0 or 12"));
return 0;
}
if (!value_fit_unsigned_field_p (opnd->imm.value, 12))
{
set_other_error (mismatch_detail, idx,
_("immediate out of range"));
return 0;
}
break;
case AARCH64_OPND_HALF:
assert (idx == 1 && opnds[0].type == AARCH64_OPND_Rd);
if (opnd->shifter.kind != AARCH64_MOD_LSL)
{
set_other_error (mismatch_detail, idx,
_("invalid shift operator"));
return 0;
}
size = aarch64_get_qualifier_esize (opnds[0].qualifier);
if (!value_aligned_p (opnd->shifter.amount, 16))
{
set_other_error (mismatch_detail, idx,
_("shift amount must be a multiple of 16"));
return 0;
}
if (!value_in_range_p (opnd->shifter.amount, 0, size * 8 - 16))
{
set_sft_amount_out_of_range_error (mismatch_detail, idx,
0, size * 8 - 16);
return 0;
}
if (opnd->imm.value < 0)
{
set_other_error (mismatch_detail, idx,
_("negative immediate value not allowed"));
return 0;
}
if (!value_fit_unsigned_field_p (opnd->imm.value, 16))
{
set_other_error (mismatch_detail, idx,
_("immediate out of range"));
return 0;
}
break;
case AARCH64_OPND_IMM_MOV:
{
int esize = aarch64_get_qualifier_esize (opnds[0].qualifier);
imm = opnd->imm.value;
assert (idx == 1);
switch (opcode->op)
{
case OP_MOV_IMM_WIDEN:
imm = ~imm;
-Wimplicit-fallthrough warning fixes Comment changes. bfd/ * coff-h8300.c: Spell fall through comments consistently. * coffgen.c: Likewise. * elf32-hppa.c: Likewise. * elf32-ppc.c: Likewise. * elf32-score.c: Likewise. * elf32-score7.c: Likewise. * elf64-ppc.c: Likewise. * elfxx-aarch64.c: Likewise. * elfxx-mips.c: Likewise. * cpu-ns32k.c: Add missing fall through comments. * elf-m10300.c: Likewise. * elf32-arm.c: Likewise. * elf32-avr.c: Likewise. * elf32-bfin.c: Likewise. * elf32-frv.c: Likewise. * elf32-i386.c: Likewise. * elf32-microblaze.c: Likewise. * elf32-nds32.c: Likewise. * elf32-ppc.c: Likewise. * elf32-rl78.c: Likewise. * elf32-rx.c: Likewise. * elf32-s390.c: Likewise. * elf32-sh.c: Likewise. * elf32-tic6x.c: Likewise. * elf64-ia64-vms.c: Likewise. * elf64-ppc.c: Likewise. * elf64-s390.c: Likewise. * elf64-x86-64.c: Likewise. * elflink.c: Likewise. * elfnn-aarch64.c: Likewise. * elfnn-ia64.c: Likewise. * ieee.c: Likewise. * oasys.c: Likewise. * pdp11.c: Likewise. * srec.c: Likewise. * versados.c: Likewise. opcodes/ * aarch64-opc.c: Spell fall through comments consistently. * i386-dis.c: Likewise. * aarch64-dis.c: Add missing fall through comments. * aarch64-opc.c: Likewise. * arc-dis.c: Likewise. * arm-dis.c: Likewise. * i386-dis.c: Likewise. * m68k-dis.c: Likewise. * mep-asm.c: Likewise. * ns32k-dis.c: Likewise. * sh-dis.c: Likewise. * tic4x-dis.c: Likewise. * tic6x-dis.c: Likewise. * vax-dis.c: Likewise. binutils/ * dlltool.c: Spell fall through comments consistently. * objcopy.c: Likewise. * readelf.c: Likewise. * dwarf.c: Add missing fall through comments. * elfcomm.c: Likewise. * sysinfo.y: Likewise. * readelf.c: Likewise. Also remove extraneous comments. gas/ * app.c: Add missing fall through comments. * dw2gencfi.c: Likewise. * expr.c: Likewise. * config/tc-alpha.c: Likewise. * config/tc-arc.c: Likewise. * config/tc-arm.c: Likewise. * config/tc-cr16.c: Likewise. * config/tc-crx.c: Likewise. * config/tc-dlx.c: Likewise. * config/tc-h8300.c: Likewise. * config/tc-hppa.c: Likewise. * config/tc-i370.c: Likewise. * config/tc-i386.c: Likewise. * config/tc-i960.c: Likewise. * config/tc-ia64.c: Likewise. * config/tc-m68hc11.c: Likewise. * config/tc-m68k.c: Likewise. * config/tc-mep.c: Likewise. * config/tc-metag.c: Likewise. * config/tc-microblaze.c: Likewise. * config/tc-mips.c: Likewise. * config/tc-ns32k.c: Likewise. * config/tc-rx.c: Likewise. * config/tc-score.c: Likewise. * config/tc-score7.c: Likewise. * config/tc-sh.c: Likewise. * config/tc-tic4x.c: Likewise. * config/tc-vax.c: Likewise. * config/tc-xstormy16.c: Likewise. * config/tc-z80.c: Likewise. * config/tc-z8k.c: Likewise. * config/obj-elf.c: Likewise. * config/tc-i386.c: Likewise. * depend.c: Spell fall through comments consistently. * config/tc-arm.c: Likewise. * config/tc-d10v.c: Likewise. * config/tc-i960.c: Likewise. * config/tc-ia64.c: Likewise. * config/tc-m68k.c: Likewise. * config/tc-mcore.c: Likewise. * config/tc-mep.c: Likewise. * config/tc-ns32k.c: Likewise. * config/tc-visium.c: Likewise. * config/tc-xstormy16.c: Likewise. * config/tc-z8k.c: Likewise. gprof/ * gprof.c: Add missing fall through comments. ld/ * lexsup.c: Spell fall through comments consistently and add missing fall through comments.
2016-10-05 15:47:02 +08:00
/* Fall through. */
case OP_MOV_IMM_WIDE:
if (!aarch64_wide_constant_p (imm, esize == 4, NULL))
{
set_other_error (mismatch_detail, idx,
_("immediate out of range"));
return 0;
}
break;
case OP_MOV_IMM_LOG:
if (!aarch64_logical_immediate_p (imm, esize, NULL))
{
set_other_error (mismatch_detail, idx,
_("immediate out of range"));
return 0;
}
break;
default:
assert (0);
return 0;
}
}
break;
case AARCH64_OPND_NZCV:
case AARCH64_OPND_CCMP_IMM:
case AARCH64_OPND_EXCEPTION:
case AARCH64_OPND_UNDEFINED:
[BINUTILS, AArch64] Enable Transactional Memory Extension This patch enables the new Transactional Memory Extension added recently as part of Arm's new architecture technologies. We introduce a new optional extension "tme" to enable this. The following instructions are part of the extension: * tstart <Xt> * ttest <Xt> * tcommit * tcancel #<imm> The ISA for the above can be found here: https://developer.arm.com/docs/ddi0602/latest/base-instructions-alphabetic-order *** gas/ChangeLog *** 2019-05-01 Sudakshina Das <sudi.das@arm.com> * config/tc-aarch64.c (parse_operands): Add case for AARCH64_OPND_TME_UIMM16. (aarch64_features): Add "tme". * doc/c-aarch64.texi: Document the same. * testsuite/gas/aarch64/tme-invalid.d: New test. * testsuite/gas/aarch64/tme-invalid.l: New test. * testsuite/gas/aarch64/tme-invalid.s: New test. * testsuite/gas/aarch64/tme.d: New test. * testsuite/gas/aarch64/tme.s: New test. *** include/ChangeLog *** 2019-05-01 Sudakshina Das <sudi.das@arm.com> * opcode/aarch64.h (AARCH64_FEATURE_TME): New. (enum aarch64_opnd): Add AARCH64_OPND_TME_UIMM16. *** opcodes/ChangeLog *** 2019-05-01 Sudakshina Das <sudi.das@arm.com> * aarch64-asm-2.c: Regenerated. * aarch64-dis-2.c: Regenerated. * aarch64-opc-2.c: Regenerated. * aarch64-opc.c (operand_general_constraint_met_p): Add case for AARCH64_OPND_TME_UIMM16. (aarch64_print_operand): Likewise. * aarch64-tbl.h (QL_IMM_NIL): New. (TME): New. (_TME_INSN): New. (struct aarch64_opcode): Add tstart, tcommit, ttest and tcancel.
2019-05-02 00:14:01 +08:00
case AARCH64_OPND_TME_UIMM16:
case AARCH64_OPND_UIMM4:
[BINUTILS, AARCH64, 2/8] Add Tag generation instructions in Memory Tagging Extension This patch is part of the patch series to add support for ARMv8.5-A Memory Tagging Extensions which is an optional extension to ARMv8.5-A and is enabled using the +memtag command line option. This patch add support to the Tag generation instructions from MTE. These are the following instructions added in this patch: - IRG <Xd|SP>, <Xn|SP>{, Xm} - ADDG <Xd|SP>, <Xn|SP>, #<uimm1>. #<uimm2> - SUBG <Xd|SP>, <Xn|SP>, #<uimm1>. #<uimm2> - GMI <Xd>, <Xn|SP>, <Xm> where <Xd|SP> : Is the 64-bit destination GPR or Stack pointer. <Xn|SP> : Is the 64-bit source GPR or Stack pointer. <uimm6> : Is the unsigned immediate, a multiple of 16 in the range 0 to 1008. <uimm4> : Is the unsigned immediate, in the range 0 to 15. *** include/ChangeLog *** 2018-11-12 Sudakshina Das <sudi.das@arm.com> * opcode/aarch64.h (aarch64_opnd): Add AARCH64_OPND_UIMM4_ADDG and AARCH64_OPND_UIMM10 as new enums. *** opcodes/ChangeLog *** 2018-11-12 Sudakshina Das <sudi.das@arm.com> * aarch64-opc.h (aarch64_field_kind): New FLD_imm4_3. (OPD_F_SHIFT_BY_4, operand_need_shift_by_four): New. * aarch64-opc.c (fields): Add entry for imm4_3. (operand_general_constraint_met_p): Add cases for AARCH64_OPND_UIMM4_ADDG and AARCH64_OPND_UIMM10. (aarch64_print_operand): Likewise. * aarch64-tbl.h (QL_ADDG): New. (aarch64_opcode_table): Add addg, subg, irg and gmi. (AARCH64_OPERANDS): Define UIMM4_ADDG and UIMM10. * aarch64-asm.c (aarch64_ins_imm): Add case for operand_need_shift_by_four. * aarch64-asm-2.c: Regenerated. * aarch64-dis-2.c: Regenerated. * aarch64-opc-2.c: Regenerated. *** gas/ChangeLog *** 2018-11-12 Sudakshina Das <sudi.das@arm.com> * config/tc-aarch64.c (parse_operands): Add switch case for AARCH64_OPND_UIMM4_ADDG and AARCH64_OPND_UIMM10. * testsuite/gas/aarch64/armv8_5-a-memtag.s: New. * testsuite/gas/aarch64/armv8_5-a-memtag.d: Likewise. * testsuite/gas/aarch64/illegal-memtag.s: Likewise. * testsuite/gas/aarch64/illegal-memtag.l: Likewise. * testsuite/gas/aarch64/illegal-memtag.d: Likewise.
2018-11-12 20:52:55 +08:00
case AARCH64_OPND_UIMM4_ADDG:
case AARCH64_OPND_UIMM7:
case AARCH64_OPND_UIMM3_OP1:
case AARCH64_OPND_UIMM3_OP2:
[AArch64][SVE 27/32] Add SVE integer immediate operands This patch adds the new SVE integer immediate operands. There are three kinds: - simple signed and unsigned ranges, but with new widths and positions. - 13-bit logical immediates. These have the same form as in base AArch64, but at a different bit position. In the case of the "MOV Zn.<T>, #<limm>" alias of DUPM, the logical immediate <limm> is not allowed to be a valid DUP immediate, since DUP is preferred over DUPM for constants that both instructions can handle. - a new 9-bit arithmetic immediate, of the form "<imm8>{, LSL #8}". In some contexts the operand is signed and in others it's unsigned. As an extension, we allow shifted immediates to be written as a single integer, e.g. "#256" is equivalent to "#1, LSL #8". We also use the shiftless form as the preferred disassembly, except for the special case of "#0, LSL #8" (a redundant encoding of 0). include/ * opcode/aarch64.h (AARCH64_OPND_SIMM5): New aarch64_opnd. (AARCH64_OPND_SVE_AIMM, AARCH64_OPND_SVE_ASIMM) (AARCH64_OPND_SVE_INV_LIMM, AARCH64_OPND_SVE_LIMM) (AARCH64_OPND_SVE_LIMM_MOV, AARCH64_OPND_SVE_SHLIMM_PRED) (AARCH64_OPND_SVE_SHLIMM_UNPRED, AARCH64_OPND_SVE_SHRIMM_PRED) (AARCH64_OPND_SVE_SHRIMM_UNPRED, AARCH64_OPND_SVE_SIMM5) (AARCH64_OPND_SVE_SIMM5B, AARCH64_OPND_SVE_SIMM6) (AARCH64_OPND_SVE_SIMM8, AARCH64_OPND_SVE_UIMM3) (AARCH64_OPND_SVE_UIMM7, AARCH64_OPND_SVE_UIMM8) (AARCH64_OPND_SVE_UIMM8_53): Likewise. (aarch64_sve_dupm_mov_immediate_p): Declare. opcodes/ * aarch64-tbl.h (AARCH64_OPERANDS): Add entries for the new SVE integer immediate operands. * aarch64-opc.h (FLD_SVE_immN, FLD_SVE_imm3, FLD_SVE_imm5) (FLD_SVE_imm5b, FLD_SVE_imm7, FLD_SVE_imm8, FLD_SVE_imm9) (FLD_SVE_immr, FLD_SVE_imms, FLD_SVE_tszh): New aarch64_field_kinds. * aarch64-opc.c (fields): Add corresponding entries. (operand_general_constraint_met_p): Handle the new SVE integer immediate operands. (aarch64_print_operand): Likewise. (aarch64_sve_dupm_mov_immediate_p): New function. * aarch64-opc-2.c: Regenerate. * aarch64-asm.h (ins_inv_limm, ins_sve_aimm, ins_sve_asimm) (ins_sve_limm_mov, ins_sve_shlimm, ins_sve_shrimm): New inserters. * aarch64-asm.c (aarch64_ins_limm_1): New function, split out from... (aarch64_ins_limm): ...here. (aarch64_ins_inv_limm): New function. (aarch64_ins_sve_aimm): Likewise. (aarch64_ins_sve_asimm): Likewise. (aarch64_ins_sve_limm_mov): Likewise. (aarch64_ins_sve_shlimm): Likewise. (aarch64_ins_sve_shrimm): Likewise. * aarch64-asm-2.c: Regenerate. * aarch64-dis.h (ext_inv_limm, ext_sve_aimm, ext_sve_asimm) (ext_sve_limm_mov, ext_sve_shlimm, ext_sve_shrimm): New extractors. * aarch64-dis.c (decode_limm): New function, split out from... (aarch64_ext_limm): ...here. (aarch64_ext_inv_limm): New function. (decode_sve_aimm): Likewise. (aarch64_ext_sve_aimm): Likewise. (aarch64_ext_sve_asimm): Likewise. (aarch64_ext_sve_limm_mov): Likewise. (aarch64_top_bit): Likewise. (aarch64_ext_sve_shlimm): Likewise. (aarch64_ext_sve_shrimm): Likewise. * aarch64-dis-2.c: Regenerate. gas/ * config/tc-aarch64.c (parse_operands): Handle the new SVE integer immediate operands.
2016-09-21 23:56:57 +08:00
case AARCH64_OPND_SVE_UIMM3:
case AARCH64_OPND_SVE_UIMM7:
case AARCH64_OPND_SVE_UIMM8:
case AARCH64_OPND_SVE_UIMM8_53:
case AARCH64_OPND_CSSC_UIMM8:
size = get_operand_fields_width (get_operand_from_code (type));
assert (size < 32);
if (!value_fit_unsigned_field_p (opnd->imm.value, size))
{
set_imm_out_of_range_error (mismatch_detail, idx, 0,
(1u << size) - 1);
return 0;
}
break;
[BINUTILS, AARCH64, 2/8] Add Tag generation instructions in Memory Tagging Extension This patch is part of the patch series to add support for ARMv8.5-A Memory Tagging Extensions which is an optional extension to ARMv8.5-A and is enabled using the +memtag command line option. This patch add support to the Tag generation instructions from MTE. These are the following instructions added in this patch: - IRG <Xd|SP>, <Xn|SP>{, Xm} - ADDG <Xd|SP>, <Xn|SP>, #<uimm1>. #<uimm2> - SUBG <Xd|SP>, <Xn|SP>, #<uimm1>. #<uimm2> - GMI <Xd>, <Xn|SP>, <Xm> where <Xd|SP> : Is the 64-bit destination GPR or Stack pointer. <Xn|SP> : Is the 64-bit source GPR or Stack pointer. <uimm6> : Is the unsigned immediate, a multiple of 16 in the range 0 to 1008. <uimm4> : Is the unsigned immediate, in the range 0 to 15. *** include/ChangeLog *** 2018-11-12 Sudakshina Das <sudi.das@arm.com> * opcode/aarch64.h (aarch64_opnd): Add AARCH64_OPND_UIMM4_ADDG and AARCH64_OPND_UIMM10 as new enums. *** opcodes/ChangeLog *** 2018-11-12 Sudakshina Das <sudi.das@arm.com> * aarch64-opc.h (aarch64_field_kind): New FLD_imm4_3. (OPD_F_SHIFT_BY_4, operand_need_shift_by_four): New. * aarch64-opc.c (fields): Add entry for imm4_3. (operand_general_constraint_met_p): Add cases for AARCH64_OPND_UIMM4_ADDG and AARCH64_OPND_UIMM10. (aarch64_print_operand): Likewise. * aarch64-tbl.h (QL_ADDG): New. (aarch64_opcode_table): Add addg, subg, irg and gmi. (AARCH64_OPERANDS): Define UIMM4_ADDG and UIMM10. * aarch64-asm.c (aarch64_ins_imm): Add case for operand_need_shift_by_four. * aarch64-asm-2.c: Regenerated. * aarch64-dis-2.c: Regenerated. * aarch64-opc-2.c: Regenerated. *** gas/ChangeLog *** 2018-11-12 Sudakshina Das <sudi.das@arm.com> * config/tc-aarch64.c (parse_operands): Add switch case for AARCH64_OPND_UIMM4_ADDG and AARCH64_OPND_UIMM10. * testsuite/gas/aarch64/armv8_5-a-memtag.s: New. * testsuite/gas/aarch64/armv8_5-a-memtag.d: Likewise. * testsuite/gas/aarch64/illegal-memtag.s: Likewise. * testsuite/gas/aarch64/illegal-memtag.l: Likewise. * testsuite/gas/aarch64/illegal-memtag.d: Likewise.
2018-11-12 20:52:55 +08:00
case AARCH64_OPND_UIMM10:
/* Scaled unsigned 10 bits immediate offset. */
if (!value_in_range_p (opnd->imm.value, 0, 1008))
{
set_imm_out_of_range_error (mismatch_detail, idx, 0, 1008);
return 0;
}
if (!value_aligned_p (opnd->imm.value, 16))
{
set_unaligned_error (mismatch_detail, idx, 16);
return 0;
}
break;
[AArch64][SVE 27/32] Add SVE integer immediate operands This patch adds the new SVE integer immediate operands. There are three kinds: - simple signed and unsigned ranges, but with new widths and positions. - 13-bit logical immediates. These have the same form as in base AArch64, but at a different bit position. In the case of the "MOV Zn.<T>, #<limm>" alias of DUPM, the logical immediate <limm> is not allowed to be a valid DUP immediate, since DUP is preferred over DUPM for constants that both instructions can handle. - a new 9-bit arithmetic immediate, of the form "<imm8>{, LSL #8}". In some contexts the operand is signed and in others it's unsigned. As an extension, we allow shifted immediates to be written as a single integer, e.g. "#256" is equivalent to "#1, LSL #8". We also use the shiftless form as the preferred disassembly, except for the special case of "#0, LSL #8" (a redundant encoding of 0). include/ * opcode/aarch64.h (AARCH64_OPND_SIMM5): New aarch64_opnd. (AARCH64_OPND_SVE_AIMM, AARCH64_OPND_SVE_ASIMM) (AARCH64_OPND_SVE_INV_LIMM, AARCH64_OPND_SVE_LIMM) (AARCH64_OPND_SVE_LIMM_MOV, AARCH64_OPND_SVE_SHLIMM_PRED) (AARCH64_OPND_SVE_SHLIMM_UNPRED, AARCH64_OPND_SVE_SHRIMM_PRED) (AARCH64_OPND_SVE_SHRIMM_UNPRED, AARCH64_OPND_SVE_SIMM5) (AARCH64_OPND_SVE_SIMM5B, AARCH64_OPND_SVE_SIMM6) (AARCH64_OPND_SVE_SIMM8, AARCH64_OPND_SVE_UIMM3) (AARCH64_OPND_SVE_UIMM7, AARCH64_OPND_SVE_UIMM8) (AARCH64_OPND_SVE_UIMM8_53): Likewise. (aarch64_sve_dupm_mov_immediate_p): Declare. opcodes/ * aarch64-tbl.h (AARCH64_OPERANDS): Add entries for the new SVE integer immediate operands. * aarch64-opc.h (FLD_SVE_immN, FLD_SVE_imm3, FLD_SVE_imm5) (FLD_SVE_imm5b, FLD_SVE_imm7, FLD_SVE_imm8, FLD_SVE_imm9) (FLD_SVE_immr, FLD_SVE_imms, FLD_SVE_tszh): New aarch64_field_kinds. * aarch64-opc.c (fields): Add corresponding entries. (operand_general_constraint_met_p): Handle the new SVE integer immediate operands. (aarch64_print_operand): Likewise. (aarch64_sve_dupm_mov_immediate_p): New function. * aarch64-opc-2.c: Regenerate. * aarch64-asm.h (ins_inv_limm, ins_sve_aimm, ins_sve_asimm) (ins_sve_limm_mov, ins_sve_shlimm, ins_sve_shrimm): New inserters. * aarch64-asm.c (aarch64_ins_limm_1): New function, split out from... (aarch64_ins_limm): ...here. (aarch64_ins_inv_limm): New function. (aarch64_ins_sve_aimm): Likewise. (aarch64_ins_sve_asimm): Likewise. (aarch64_ins_sve_limm_mov): Likewise. (aarch64_ins_sve_shlimm): Likewise. (aarch64_ins_sve_shrimm): Likewise. * aarch64-asm-2.c: Regenerate. * aarch64-dis.h (ext_inv_limm, ext_sve_aimm, ext_sve_asimm) (ext_sve_limm_mov, ext_sve_shlimm, ext_sve_shrimm): New extractors. * aarch64-dis.c (decode_limm): New function, split out from... (aarch64_ext_limm): ...here. (aarch64_ext_inv_limm): New function. (decode_sve_aimm): Likewise. (aarch64_ext_sve_aimm): Likewise. (aarch64_ext_sve_asimm): Likewise. (aarch64_ext_sve_limm_mov): Likewise. (aarch64_top_bit): Likewise. (aarch64_ext_sve_shlimm): Likewise. (aarch64_ext_sve_shrimm): Likewise. * aarch64-dis-2.c: Regenerate. gas/ * config/tc-aarch64.c (parse_operands): Handle the new SVE integer immediate operands.
2016-09-21 23:56:57 +08:00
case AARCH64_OPND_SIMM5:
case AARCH64_OPND_SVE_SIMM5:
case AARCH64_OPND_SVE_SIMM5B:
case AARCH64_OPND_SVE_SIMM6:
case AARCH64_OPND_SVE_SIMM8:
case AARCH64_OPND_CSSC_SIMM8:
[AArch64][SVE 27/32] Add SVE integer immediate operands This patch adds the new SVE integer immediate operands. There are three kinds: - simple signed and unsigned ranges, but with new widths and positions. - 13-bit logical immediates. These have the same form as in base AArch64, but at a different bit position. In the case of the "MOV Zn.<T>, #<limm>" alias of DUPM, the logical immediate <limm> is not allowed to be a valid DUP immediate, since DUP is preferred over DUPM for constants that both instructions can handle. - a new 9-bit arithmetic immediate, of the form "<imm8>{, LSL #8}". In some contexts the operand is signed and in others it's unsigned. As an extension, we allow shifted immediates to be written as a single integer, e.g. "#256" is equivalent to "#1, LSL #8". We also use the shiftless form as the preferred disassembly, except for the special case of "#0, LSL #8" (a redundant encoding of 0). include/ * opcode/aarch64.h (AARCH64_OPND_SIMM5): New aarch64_opnd. (AARCH64_OPND_SVE_AIMM, AARCH64_OPND_SVE_ASIMM) (AARCH64_OPND_SVE_INV_LIMM, AARCH64_OPND_SVE_LIMM) (AARCH64_OPND_SVE_LIMM_MOV, AARCH64_OPND_SVE_SHLIMM_PRED) (AARCH64_OPND_SVE_SHLIMM_UNPRED, AARCH64_OPND_SVE_SHRIMM_PRED) (AARCH64_OPND_SVE_SHRIMM_UNPRED, AARCH64_OPND_SVE_SIMM5) (AARCH64_OPND_SVE_SIMM5B, AARCH64_OPND_SVE_SIMM6) (AARCH64_OPND_SVE_SIMM8, AARCH64_OPND_SVE_UIMM3) (AARCH64_OPND_SVE_UIMM7, AARCH64_OPND_SVE_UIMM8) (AARCH64_OPND_SVE_UIMM8_53): Likewise. (aarch64_sve_dupm_mov_immediate_p): Declare. opcodes/ * aarch64-tbl.h (AARCH64_OPERANDS): Add entries for the new SVE integer immediate operands. * aarch64-opc.h (FLD_SVE_immN, FLD_SVE_imm3, FLD_SVE_imm5) (FLD_SVE_imm5b, FLD_SVE_imm7, FLD_SVE_imm8, FLD_SVE_imm9) (FLD_SVE_immr, FLD_SVE_imms, FLD_SVE_tszh): New aarch64_field_kinds. * aarch64-opc.c (fields): Add corresponding entries. (operand_general_constraint_met_p): Handle the new SVE integer immediate operands. (aarch64_print_operand): Likewise. (aarch64_sve_dupm_mov_immediate_p): New function. * aarch64-opc-2.c: Regenerate. * aarch64-asm.h (ins_inv_limm, ins_sve_aimm, ins_sve_asimm) (ins_sve_limm_mov, ins_sve_shlimm, ins_sve_shrimm): New inserters. * aarch64-asm.c (aarch64_ins_limm_1): New function, split out from... (aarch64_ins_limm): ...here. (aarch64_ins_inv_limm): New function. (aarch64_ins_sve_aimm): Likewise. (aarch64_ins_sve_asimm): Likewise. (aarch64_ins_sve_limm_mov): Likewise. (aarch64_ins_sve_shlimm): Likewise. (aarch64_ins_sve_shrimm): Likewise. * aarch64-asm-2.c: Regenerate. * aarch64-dis.h (ext_inv_limm, ext_sve_aimm, ext_sve_asimm) (ext_sve_limm_mov, ext_sve_shlimm, ext_sve_shrimm): New extractors. * aarch64-dis.c (decode_limm): New function, split out from... (aarch64_ext_limm): ...here. (aarch64_ext_inv_limm): New function. (decode_sve_aimm): Likewise. (aarch64_ext_sve_aimm): Likewise. (aarch64_ext_sve_asimm): Likewise. (aarch64_ext_sve_limm_mov): Likewise. (aarch64_top_bit): Likewise. (aarch64_ext_sve_shlimm): Likewise. (aarch64_ext_sve_shrimm): Likewise. * aarch64-dis-2.c: Regenerate. gas/ * config/tc-aarch64.c (parse_operands): Handle the new SVE integer immediate operands.
2016-09-21 23:56:57 +08:00
size = get_operand_fields_width (get_operand_from_code (type));
assert (size < 32);
if (!value_fit_signed_field_p (opnd->imm.value, size))
{
set_imm_out_of_range_error (mismatch_detail, idx,
-(1 << (size - 1)),
(1 << (size - 1)) - 1);
return 0;
}
break;
case AARCH64_OPND_WIDTH:
assert (idx > 1 && opnds[idx-1].type == AARCH64_OPND_IMM
&& opnds[0].type == AARCH64_OPND_Rd);
size = get_upper_bound (qualifier);
if (opnd->imm.value + opnds[idx-1].imm.value > size)
/* lsb+width <= reg.size */
{
set_imm_out_of_range_error (mismatch_detail, idx, 1,
size - opnds[idx-1].imm.value);
return 0;
}
break;
case AARCH64_OPND_LIMM:
[AArch64][SVE 27/32] Add SVE integer immediate operands This patch adds the new SVE integer immediate operands. There are three kinds: - simple signed and unsigned ranges, but with new widths and positions. - 13-bit logical immediates. These have the same form as in base AArch64, but at a different bit position. In the case of the "MOV Zn.<T>, #<limm>" alias of DUPM, the logical immediate <limm> is not allowed to be a valid DUP immediate, since DUP is preferred over DUPM for constants that both instructions can handle. - a new 9-bit arithmetic immediate, of the form "<imm8>{, LSL #8}". In some contexts the operand is signed and in others it's unsigned. As an extension, we allow shifted immediates to be written as a single integer, e.g. "#256" is equivalent to "#1, LSL #8". We also use the shiftless form as the preferred disassembly, except for the special case of "#0, LSL #8" (a redundant encoding of 0). include/ * opcode/aarch64.h (AARCH64_OPND_SIMM5): New aarch64_opnd. (AARCH64_OPND_SVE_AIMM, AARCH64_OPND_SVE_ASIMM) (AARCH64_OPND_SVE_INV_LIMM, AARCH64_OPND_SVE_LIMM) (AARCH64_OPND_SVE_LIMM_MOV, AARCH64_OPND_SVE_SHLIMM_PRED) (AARCH64_OPND_SVE_SHLIMM_UNPRED, AARCH64_OPND_SVE_SHRIMM_PRED) (AARCH64_OPND_SVE_SHRIMM_UNPRED, AARCH64_OPND_SVE_SIMM5) (AARCH64_OPND_SVE_SIMM5B, AARCH64_OPND_SVE_SIMM6) (AARCH64_OPND_SVE_SIMM8, AARCH64_OPND_SVE_UIMM3) (AARCH64_OPND_SVE_UIMM7, AARCH64_OPND_SVE_UIMM8) (AARCH64_OPND_SVE_UIMM8_53): Likewise. (aarch64_sve_dupm_mov_immediate_p): Declare. opcodes/ * aarch64-tbl.h (AARCH64_OPERANDS): Add entries for the new SVE integer immediate operands. * aarch64-opc.h (FLD_SVE_immN, FLD_SVE_imm3, FLD_SVE_imm5) (FLD_SVE_imm5b, FLD_SVE_imm7, FLD_SVE_imm8, FLD_SVE_imm9) (FLD_SVE_immr, FLD_SVE_imms, FLD_SVE_tszh): New aarch64_field_kinds. * aarch64-opc.c (fields): Add corresponding entries. (operand_general_constraint_met_p): Handle the new SVE integer immediate operands. (aarch64_print_operand): Likewise. (aarch64_sve_dupm_mov_immediate_p): New function. * aarch64-opc-2.c: Regenerate. * aarch64-asm.h (ins_inv_limm, ins_sve_aimm, ins_sve_asimm) (ins_sve_limm_mov, ins_sve_shlimm, ins_sve_shrimm): New inserters. * aarch64-asm.c (aarch64_ins_limm_1): New function, split out from... (aarch64_ins_limm): ...here. (aarch64_ins_inv_limm): New function. (aarch64_ins_sve_aimm): Likewise. (aarch64_ins_sve_asimm): Likewise. (aarch64_ins_sve_limm_mov): Likewise. (aarch64_ins_sve_shlimm): Likewise. (aarch64_ins_sve_shrimm): Likewise. * aarch64-asm-2.c: Regenerate. * aarch64-dis.h (ext_inv_limm, ext_sve_aimm, ext_sve_asimm) (ext_sve_limm_mov, ext_sve_shlimm, ext_sve_shrimm): New extractors. * aarch64-dis.c (decode_limm): New function, split out from... (aarch64_ext_limm): ...here. (aarch64_ext_inv_limm): New function. (decode_sve_aimm): Likewise. (aarch64_ext_sve_aimm): Likewise. (aarch64_ext_sve_asimm): Likewise. (aarch64_ext_sve_limm_mov): Likewise. (aarch64_top_bit): Likewise. (aarch64_ext_sve_shlimm): Likewise. (aarch64_ext_sve_shrimm): Likewise. * aarch64-dis-2.c: Regenerate. gas/ * config/tc-aarch64.c (parse_operands): Handle the new SVE integer immediate operands.
2016-09-21 23:56:57 +08:00
case AARCH64_OPND_SVE_LIMM:
{
int esize = aarch64_get_qualifier_esize (opnds[0].qualifier);
uint64_t uimm = opnd->imm.value;
if (opcode->op == OP_BIC)
uimm = ~uimm;
if (!aarch64_logical_immediate_p (uimm, esize, NULL))
{
set_other_error (mismatch_detail, idx,
_("immediate out of range"));
return 0;
}
}
break;
case AARCH64_OPND_IMM0:
case AARCH64_OPND_FPIMM0:
if (opnd->imm.value != 0)
{
set_other_error (mismatch_detail, idx,
_("immediate zero expected"));
return 0;
}
break;
[AArch64] Add ARMv8.3 FCMLA and FCADD instructions Add support for FCMLA and FCADD complex arithmetic SIMD instructions. FCMLA has an indexed element variant where the index range has to be treated specially because a complex number takes two elements and the indexed vector size depends on the other operands. These complex number SIMD instructions are part of ARMv8.3 https://community.arm.com/groups/processors/blog/2016/10/27/armv8-a-architecture-2016-additions include/ 2016-11-18 Szabolcs Nagy <szabolcs.nagy@arm.com> * opcode/aarch64.h (enum aarch64_opnd): Add AARCH64_OPND_IMM_ROT1, AARCH64_OPND_IMM_ROT2, AARCH64_OPND_IMM_ROT3. (enum aarch64_op): Add OP_FCMLA_ELEM. opcodes/ 2016-11-18 Szabolcs Nagy <szabolcs.nagy@arm.com> * aarch64-tbl.h (QL_V3SAMEHSD_ROT, QL_ELEMENT_ROT): Define. (aarch64_feature_simd_v8_3, SIMD_V8_3): Define. (aarch64_opcode_table): Add fcmla and fcadd. (AARCH64_OPERANDS): Add IMM_ROT{1,2,3}. * aarch64-asm.h (aarch64_ins_imm_rotate): Declare. * aarch64-asm.c (aarch64_ins_imm_rotate): Define. * aarch64-dis.h (aarch64_ext_imm_rotate): Declare. * aarch64-dis.c (aarch64_ext_imm_rotate): Define. * aarch64-opc.h (enum aarch64_field_kind): Add FLD_rotate{1,2,3}. * aarch64-opc.c (fields): Add FLD_rotate{1,2,3}. (operand_general_constraint_met_p): Rotate and index range check. (aarch64_print_operand): Handle rotate operand. * aarch64-asm-2.c: Regenerate. * aarch64-dis-2.c: Likewise. * aarch64-opc-2.c: Likewise. gas/ 2016-11-18 Szabolcs Nagy <szabolcs.nagy@arm.com> * config/tc-aarch64.c (parse_operands): Handle AARCH64_OPND_IMM_ROT*. * testsuite/gas/aarch64/advsimd-armv8_3.d: New. * testsuite/gas/aarch64/advsimd-armv8_3.s: New. * testsuite/gas/aarch64/illegal-fcmla.s: New. * testsuite/gas/aarch64/illegal-fcmla.l: New. * testsuite/gas/aarch64/illegal-fcmla.d: New.
2016-11-18 18:02:16 +08:00
case AARCH64_OPND_IMM_ROT1:
case AARCH64_OPND_IMM_ROT2:
[AArch64] Additional SVE instructions This patch supports some additions to the SVE architecture prior to its public release. include/ * opcode/aarch64.h (AARCH64_OPND_SVE_ADDR_RI_S4x16) (AARCH64_OPND_SVE_IMM_ROT1, AARCH64_OPND_SVE_IMM_ROT2) (AARCH64_OPND_SVE_Zm3_INDEX, AARCH64_OPND_SVE_Zm3_22_INDEX) (AARCH64_OPND_SVE_Zm4_INDEX): New aarch64_opnds. opcodes/ * aarch64-tbl.h (OP_SVE_HMH, OP_SVE_VMU_HSD, OP_SVE_VMVU_HSD) (OP_SVE_VMVV_HSD, OP_SVE_VMVVU_HSD, OP_SVE_VM_HSD, OP_SVE_VUVV_HSD) (OP_SVE_VUV_HSD, OP_SVE_VU_HSD, OP_SVE_VVVU_H, OP_SVE_VVVU_S) (OP_SVE_VVVU_HSD, OP_SVE_VVV_D, OP_SVE_VVV_D_H, OP_SVE_VVV_H) (OP_SVE_VVV_HSD, OP_SVE_VVV_S, OP_SVE_VVV_S_B, OP_SVE_VVV_SD_BH) (OP_SVE_VV_BHSDQ, OP_SVE_VV_HSD, OP_SVE_VZVV_HSD, OP_SVE_VZV_HSD) (OP_SVE_V_HSD): New macros. (OP_SVE_VMU_SD, OP_SVE_VMVU_SD, OP_SVE_VM_SD, OP_SVE_VUVV_SD) (OP_SVE_VU_SD, OP_SVE_VVVU_SD, OP_SVE_VVV_SD, OP_SVE_VZVV_SD) (OP_SVE_VZV_SD, OP_SVE_V_SD): Delete. (aarch64_opcode_table): Add new SVE instructions. (aarch64_opcode_table): Use imm_rotate{1,2} instead of imm_rotate for rotation operands. Add new SVE operands. * aarch64-asm.h (ins_sve_addr_ri_s4): New inserter. (ins_sve_quad_index): Likewise. (ins_imm_rotate): Split into... (ins_imm_rotate1, ins_imm_rotate2): ...these two inserters. * aarch64-asm.c (aarch64_ins_imm_rotate): Split into... (aarch64_ins_imm_rotate1, aarch64_ins_imm_rotate2): ...these two functions. (aarch64_ins_sve_addr_ri_s4): New function. (aarch64_ins_sve_quad_index): Likewise. (do_misc_encoding): Handle "MOV Zn.Q, Qm". * aarch64-asm-2.c: Regenerate. * aarch64-dis.h (ext_sve_addr_ri_s4): New extractor. (ext_sve_quad_index): Likewise. (ext_imm_rotate): Split into... (ext_imm_rotate1, ext_imm_rotate2): ...these two extractors. * aarch64-dis.c (aarch64_ext_imm_rotate): Split into... (aarch64_ext_imm_rotate1, aarch64_ext_imm_rotate2): ...these two functions. (aarch64_ext_sve_addr_ri_s4): New function. (aarch64_ext_sve_quad_index): Likewise. (aarch64_ext_sve_index): Allow quad indices. (do_misc_decoding): Likewise. * aarch64-dis-2.c: Regenerate. * aarch64-opc.h (FLD_SVE_i3h, FLD_SVE_rot1, FLD_SVE_rot2): New aarch64_field_kinds. (OPD_F_OD_MASK): Widen by one bit. (OPD_F_NO_ZR): Bump accordingly. (get_operand_field_width): New function. * aarch64-opc.c (fields): Add new SVE fields. (operand_general_constraint_met_p): Handle new SVE operands. (aarch64_print_operand): Likewise. * aarch64-opc-2.c: Regenerate. gas/ * doc/c-aarch64.texi: Document that sve implies fp16, simd and compnum. * config/tc-aarch64.c (parse_vector_type_for_operand): Allow .q to be used with SVE registers. (parse_operands): Handle new SVE operands. (aarch64_features): Make "sve" require F16 rather than FP. Also require COMPNUM. * testsuite/gas/aarch64/sve.s: Add tests for new instructions. Include compnum tests. * testsuite/gas/aarch64/sve.d: Update accordingly. * testsuite/gas/aarch64/sve-invalid.s: Add tests for new instructions. * testsuite/gas/aarch64/sve-invalid.l: Update accordingly. Also update expected output for new FMOV and MOV alternatives.
2017-02-25 02:29:00 +08:00
case AARCH64_OPND_SVE_IMM_ROT2:
[AArch64] Add ARMv8.3 FCMLA and FCADD instructions Add support for FCMLA and FCADD complex arithmetic SIMD instructions. FCMLA has an indexed element variant where the index range has to be treated specially because a complex number takes two elements and the indexed vector size depends on the other operands. These complex number SIMD instructions are part of ARMv8.3 https://community.arm.com/groups/processors/blog/2016/10/27/armv8-a-architecture-2016-additions include/ 2016-11-18 Szabolcs Nagy <szabolcs.nagy@arm.com> * opcode/aarch64.h (enum aarch64_opnd): Add AARCH64_OPND_IMM_ROT1, AARCH64_OPND_IMM_ROT2, AARCH64_OPND_IMM_ROT3. (enum aarch64_op): Add OP_FCMLA_ELEM. opcodes/ 2016-11-18 Szabolcs Nagy <szabolcs.nagy@arm.com> * aarch64-tbl.h (QL_V3SAMEHSD_ROT, QL_ELEMENT_ROT): Define. (aarch64_feature_simd_v8_3, SIMD_V8_3): Define. (aarch64_opcode_table): Add fcmla and fcadd. (AARCH64_OPERANDS): Add IMM_ROT{1,2,3}. * aarch64-asm.h (aarch64_ins_imm_rotate): Declare. * aarch64-asm.c (aarch64_ins_imm_rotate): Define. * aarch64-dis.h (aarch64_ext_imm_rotate): Declare. * aarch64-dis.c (aarch64_ext_imm_rotate): Define. * aarch64-opc.h (enum aarch64_field_kind): Add FLD_rotate{1,2,3}. * aarch64-opc.c (fields): Add FLD_rotate{1,2,3}. (operand_general_constraint_met_p): Rotate and index range check. (aarch64_print_operand): Handle rotate operand. * aarch64-asm-2.c: Regenerate. * aarch64-dis-2.c: Likewise. * aarch64-opc-2.c: Likewise. gas/ 2016-11-18 Szabolcs Nagy <szabolcs.nagy@arm.com> * config/tc-aarch64.c (parse_operands): Handle AARCH64_OPND_IMM_ROT*. * testsuite/gas/aarch64/advsimd-armv8_3.d: New. * testsuite/gas/aarch64/advsimd-armv8_3.s: New. * testsuite/gas/aarch64/illegal-fcmla.s: New. * testsuite/gas/aarch64/illegal-fcmla.l: New. * testsuite/gas/aarch64/illegal-fcmla.d: New.
2016-11-18 18:02:16 +08:00
if (opnd->imm.value != 0
&& opnd->imm.value != 90
&& opnd->imm.value != 180
&& opnd->imm.value != 270)
{
set_other_error (mismatch_detail, idx,
_("rotate expected to be 0, 90, 180 or 270"));
return 0;
}
break;
case AARCH64_OPND_IMM_ROT3:
[AArch64] Additional SVE instructions This patch supports some additions to the SVE architecture prior to its public release. include/ * opcode/aarch64.h (AARCH64_OPND_SVE_ADDR_RI_S4x16) (AARCH64_OPND_SVE_IMM_ROT1, AARCH64_OPND_SVE_IMM_ROT2) (AARCH64_OPND_SVE_Zm3_INDEX, AARCH64_OPND_SVE_Zm3_22_INDEX) (AARCH64_OPND_SVE_Zm4_INDEX): New aarch64_opnds. opcodes/ * aarch64-tbl.h (OP_SVE_HMH, OP_SVE_VMU_HSD, OP_SVE_VMVU_HSD) (OP_SVE_VMVV_HSD, OP_SVE_VMVVU_HSD, OP_SVE_VM_HSD, OP_SVE_VUVV_HSD) (OP_SVE_VUV_HSD, OP_SVE_VU_HSD, OP_SVE_VVVU_H, OP_SVE_VVVU_S) (OP_SVE_VVVU_HSD, OP_SVE_VVV_D, OP_SVE_VVV_D_H, OP_SVE_VVV_H) (OP_SVE_VVV_HSD, OP_SVE_VVV_S, OP_SVE_VVV_S_B, OP_SVE_VVV_SD_BH) (OP_SVE_VV_BHSDQ, OP_SVE_VV_HSD, OP_SVE_VZVV_HSD, OP_SVE_VZV_HSD) (OP_SVE_V_HSD): New macros. (OP_SVE_VMU_SD, OP_SVE_VMVU_SD, OP_SVE_VM_SD, OP_SVE_VUVV_SD) (OP_SVE_VU_SD, OP_SVE_VVVU_SD, OP_SVE_VVV_SD, OP_SVE_VZVV_SD) (OP_SVE_VZV_SD, OP_SVE_V_SD): Delete. (aarch64_opcode_table): Add new SVE instructions. (aarch64_opcode_table): Use imm_rotate{1,2} instead of imm_rotate for rotation operands. Add new SVE operands. * aarch64-asm.h (ins_sve_addr_ri_s4): New inserter. (ins_sve_quad_index): Likewise. (ins_imm_rotate): Split into... (ins_imm_rotate1, ins_imm_rotate2): ...these two inserters. * aarch64-asm.c (aarch64_ins_imm_rotate): Split into... (aarch64_ins_imm_rotate1, aarch64_ins_imm_rotate2): ...these two functions. (aarch64_ins_sve_addr_ri_s4): New function. (aarch64_ins_sve_quad_index): Likewise. (do_misc_encoding): Handle "MOV Zn.Q, Qm". * aarch64-asm-2.c: Regenerate. * aarch64-dis.h (ext_sve_addr_ri_s4): New extractor. (ext_sve_quad_index): Likewise. (ext_imm_rotate): Split into... (ext_imm_rotate1, ext_imm_rotate2): ...these two extractors. * aarch64-dis.c (aarch64_ext_imm_rotate): Split into... (aarch64_ext_imm_rotate1, aarch64_ext_imm_rotate2): ...these two functions. (aarch64_ext_sve_addr_ri_s4): New function. (aarch64_ext_sve_quad_index): Likewise. (aarch64_ext_sve_index): Allow quad indices. (do_misc_decoding): Likewise. * aarch64-dis-2.c: Regenerate. * aarch64-opc.h (FLD_SVE_i3h, FLD_SVE_rot1, FLD_SVE_rot2): New aarch64_field_kinds. (OPD_F_OD_MASK): Widen by one bit. (OPD_F_NO_ZR): Bump accordingly. (get_operand_field_width): New function. * aarch64-opc.c (fields): Add new SVE fields. (operand_general_constraint_met_p): Handle new SVE operands. (aarch64_print_operand): Likewise. * aarch64-opc-2.c: Regenerate. gas/ * doc/c-aarch64.texi: Document that sve implies fp16, simd and compnum. * config/tc-aarch64.c (parse_vector_type_for_operand): Allow .q to be used with SVE registers. (parse_operands): Handle new SVE operands. (aarch64_features): Make "sve" require F16 rather than FP. Also require COMPNUM. * testsuite/gas/aarch64/sve.s: Add tests for new instructions. Include compnum tests. * testsuite/gas/aarch64/sve.d: Update accordingly. * testsuite/gas/aarch64/sve-invalid.s: Add tests for new instructions. * testsuite/gas/aarch64/sve-invalid.l: Update accordingly. Also update expected output for new FMOV and MOV alternatives.
2017-02-25 02:29:00 +08:00
case AARCH64_OPND_SVE_IMM_ROT1:
case AARCH64_OPND_SVE_IMM_ROT3:
[AArch64] Add ARMv8.3 FCMLA and FCADD instructions Add support for FCMLA and FCADD complex arithmetic SIMD instructions. FCMLA has an indexed element variant where the index range has to be treated specially because a complex number takes two elements and the indexed vector size depends on the other operands. These complex number SIMD instructions are part of ARMv8.3 https://community.arm.com/groups/processors/blog/2016/10/27/armv8-a-architecture-2016-additions include/ 2016-11-18 Szabolcs Nagy <szabolcs.nagy@arm.com> * opcode/aarch64.h (enum aarch64_opnd): Add AARCH64_OPND_IMM_ROT1, AARCH64_OPND_IMM_ROT2, AARCH64_OPND_IMM_ROT3. (enum aarch64_op): Add OP_FCMLA_ELEM. opcodes/ 2016-11-18 Szabolcs Nagy <szabolcs.nagy@arm.com> * aarch64-tbl.h (QL_V3SAMEHSD_ROT, QL_ELEMENT_ROT): Define. (aarch64_feature_simd_v8_3, SIMD_V8_3): Define. (aarch64_opcode_table): Add fcmla and fcadd. (AARCH64_OPERANDS): Add IMM_ROT{1,2,3}. * aarch64-asm.h (aarch64_ins_imm_rotate): Declare. * aarch64-asm.c (aarch64_ins_imm_rotate): Define. * aarch64-dis.h (aarch64_ext_imm_rotate): Declare. * aarch64-dis.c (aarch64_ext_imm_rotate): Define. * aarch64-opc.h (enum aarch64_field_kind): Add FLD_rotate{1,2,3}. * aarch64-opc.c (fields): Add FLD_rotate{1,2,3}. (operand_general_constraint_met_p): Rotate and index range check. (aarch64_print_operand): Handle rotate operand. * aarch64-asm-2.c: Regenerate. * aarch64-dis-2.c: Likewise. * aarch64-opc-2.c: Likewise. gas/ 2016-11-18 Szabolcs Nagy <szabolcs.nagy@arm.com> * config/tc-aarch64.c (parse_operands): Handle AARCH64_OPND_IMM_ROT*. * testsuite/gas/aarch64/advsimd-armv8_3.d: New. * testsuite/gas/aarch64/advsimd-armv8_3.s: New. * testsuite/gas/aarch64/illegal-fcmla.s: New. * testsuite/gas/aarch64/illegal-fcmla.l: New. * testsuite/gas/aarch64/illegal-fcmla.d: New.
2016-11-18 18:02:16 +08:00
if (opnd->imm.value != 90 && opnd->imm.value != 270)
{
set_other_error (mismatch_detail, idx,
_("rotate expected to be 90 or 270"));
return 0;
}
break;
case AARCH64_OPND_SHLL_IMM:
assert (idx == 2);
size = 8 * aarch64_get_qualifier_esize (opnds[idx - 1].qualifier);
if (opnd->imm.value != size)
{
set_other_error (mismatch_detail, idx,
_("invalid shift amount"));
return 0;
}
break;
case AARCH64_OPND_IMM_VLSL:
size = aarch64_get_qualifier_esize (qualifier);
if (!value_in_range_p (opnd->imm.value, 0, size * 8 - 1))
{
set_imm_out_of_range_error (mismatch_detail, idx, 0,
size * 8 - 1);
return 0;
}
break;
case AARCH64_OPND_IMM_VLSR:
size = aarch64_get_qualifier_esize (qualifier);
if (!value_in_range_p (opnd->imm.value, 1, size * 8))
{
set_imm_out_of_range_error (mismatch_detail, idx, 1, size * 8);
return 0;
}
break;
case AARCH64_OPND_SIMD_IMM:
case AARCH64_OPND_SIMD_IMM_SFT:
/* Qualifier check. */
switch (qualifier)
{
case AARCH64_OPND_QLF_LSL:
if (opnd->shifter.kind != AARCH64_MOD_LSL)
{
set_other_error (mismatch_detail, idx,
_("invalid shift operator"));
return 0;
}
break;
case AARCH64_OPND_QLF_MSL:
if (opnd->shifter.kind != AARCH64_MOD_MSL)
{
set_other_error (mismatch_detail, idx,
_("invalid shift operator"));
return 0;
}
break;
case AARCH64_OPND_QLF_NIL:
if (opnd->shifter.kind != AARCH64_MOD_NONE)
{
set_other_error (mismatch_detail, idx,
_("shift is not permitted"));
return 0;
}
break;
default:
assert (0);
return 0;
}
/* Is the immediate valid? */
assert (idx == 1);
if (aarch64_get_qualifier_esize (opnds[0].qualifier) != 8)
{
/* uimm8 or simm8 */
if (!value_in_range_p (opnd->imm.value, -128, 255))
{
set_imm_out_of_range_error (mismatch_detail, idx, -128, 255);
return 0;
}
}
else if (aarch64_shrink_expanded_imm8 (opnd->imm.value) < 0)
{
/* uimm64 is not
'aaaaaaaabbbbbbbbccccccccddddddddeeeeeeee
ffffffffgggggggghhhhhhhh'. */
set_other_error (mismatch_detail, idx,
_("invalid value for immediate"));
return 0;
}
/* Is the shift amount valid? */
switch (opnd->shifter.kind)
{
case AARCH64_MOD_LSL:
size = aarch64_get_qualifier_esize (opnds[0].qualifier);
if (!value_in_range_p (opnd->shifter.amount, 0, (size - 1) * 8))
{
set_sft_amount_out_of_range_error (mismatch_detail, idx, 0,
(size - 1) * 8);
return 0;
}
if (!value_aligned_p (opnd->shifter.amount, 8))
{
set_unaligned_error (mismatch_detail, idx, 8);
return 0;
}
break;
case AARCH64_MOD_MSL:
/* Only 8 and 16 are valid shift amount. */
if (opnd->shifter.amount != 8 && opnd->shifter.amount != 16)
{
set_other_error (mismatch_detail, idx,
_("shift amount must be 0 or 16"));
return 0;
}
break;
default:
if (opnd->shifter.kind != AARCH64_MOD_NONE)
{
set_other_error (mismatch_detail, idx,
_("invalid shift operator"));
return 0;
}
break;
}
break;
case AARCH64_OPND_FPIMM:
case AARCH64_OPND_SIMD_FPIMM:
[AArch64][SVE 28/32] Add SVE FP immediate operands This patch adds support for the new SVE floating-point immediate operands. One operand uses the same 8-bit encoding as base AArch64, but in a different position. The others use a single bit to select between two values. One of the single-bit operands is a choice between 0 and 1, where 0 is not a valid 8-bit encoding. I think the cleanest way of handling these single-bit immediates is therefore to use the IEEE float encoding itself as the immediate value and select between the two possible values when encoding and decoding. As described in the covering note for the patch that added F_STRICT, we get better error messages by accepting unsuffixed vector registers and leaving the qualifier matching code to report an error. This means that we carry on parsing the other operands, and so can try to parse FP immediates for invalid instructions like: fcpy z0, #2.5 In this case there is no suffix to tell us whether the immediate should be treated as single or double precision. Again, we get better error messages by picking one (arbitrary) immediate size and reporting an error for the missing suffix later. include/ * opcode/aarch64.h (AARCH64_OPND_SVE_FPIMM8): New aarch64_opnd. (AARCH64_OPND_SVE_I1_HALF_ONE, AARCH64_OPND_SVE_I1_HALF_TWO) (AARCH64_OPND_SVE_I1_ZERO_ONE): Likewise. opcodes/ * aarch64-tbl.h (AARCH64_OPERANDS): Add entries for the new SVE FP immediate operands. * aarch64-opc.h (FLD_SVE_i1): New aarch64_field_kind. * aarch64-opc.c (fields): Add corresponding entry. (operand_general_constraint_met_p): Handle the new SVE FP immediate operands. (aarch64_print_operand): Likewise. * aarch64-opc-2.c: Regenerate. * aarch64-asm.h (ins_sve_float_half_one, ins_sve_float_half_two) (ins_sve_float_zero_one): New inserters. * aarch64-asm.c (aarch64_ins_sve_float_half_one): New function. (aarch64_ins_sve_float_half_two): Likewise. (aarch64_ins_sve_float_zero_one): Likewise. * aarch64-asm-2.c: Regenerate. * aarch64-dis.h (ext_sve_float_half_one, ext_sve_float_half_two) (ext_sve_float_zero_one): New extractors. * aarch64-dis.c (aarch64_ext_sve_float_half_one): New function. (aarch64_ext_sve_float_half_two): Likewise. (aarch64_ext_sve_float_zero_one): Likewise. * aarch64-dis-2.c: Regenerate. gas/ * config/tc-aarch64.c (double_precision_operand_p): New function. (parse_operands): Use it to calculate the dp_p input to parse_aarch64_imm_float. Handle the new SVE FP immediate operands.
2016-09-21 23:57:22 +08:00
case AARCH64_OPND_SVE_FPIMM8:
if (opnd->imm.is_fp == 0)
{
set_other_error (mismatch_detail, idx,
_("floating-point immediate expected"));
return 0;
}
/* The value is expected to be an 8-bit floating-point constant with
sign, 3-bit exponent and normalized 4 bits of precision, encoded
in "a:b:c:d:e:f:g:h" or FLD_imm8 (depending on the type of the
instruction). */
if (!value_in_range_p (opnd->imm.value, 0, 255))
{
set_other_error (mismatch_detail, idx,
_("immediate out of range"));
return 0;
}
if (opnd->shifter.kind != AARCH64_MOD_NONE)
{
set_other_error (mismatch_detail, idx,
_("invalid shift operator"));
return 0;
}
break;
[AArch64][SVE 27/32] Add SVE integer immediate operands This patch adds the new SVE integer immediate operands. There are three kinds: - simple signed and unsigned ranges, but with new widths and positions. - 13-bit logical immediates. These have the same form as in base AArch64, but at a different bit position. In the case of the "MOV Zn.<T>, #<limm>" alias of DUPM, the logical immediate <limm> is not allowed to be a valid DUP immediate, since DUP is preferred over DUPM for constants that both instructions can handle. - a new 9-bit arithmetic immediate, of the form "<imm8>{, LSL #8}". In some contexts the operand is signed and in others it's unsigned. As an extension, we allow shifted immediates to be written as a single integer, e.g. "#256" is equivalent to "#1, LSL #8". We also use the shiftless form as the preferred disassembly, except for the special case of "#0, LSL #8" (a redundant encoding of 0). include/ * opcode/aarch64.h (AARCH64_OPND_SIMM5): New aarch64_opnd. (AARCH64_OPND_SVE_AIMM, AARCH64_OPND_SVE_ASIMM) (AARCH64_OPND_SVE_INV_LIMM, AARCH64_OPND_SVE_LIMM) (AARCH64_OPND_SVE_LIMM_MOV, AARCH64_OPND_SVE_SHLIMM_PRED) (AARCH64_OPND_SVE_SHLIMM_UNPRED, AARCH64_OPND_SVE_SHRIMM_PRED) (AARCH64_OPND_SVE_SHRIMM_UNPRED, AARCH64_OPND_SVE_SIMM5) (AARCH64_OPND_SVE_SIMM5B, AARCH64_OPND_SVE_SIMM6) (AARCH64_OPND_SVE_SIMM8, AARCH64_OPND_SVE_UIMM3) (AARCH64_OPND_SVE_UIMM7, AARCH64_OPND_SVE_UIMM8) (AARCH64_OPND_SVE_UIMM8_53): Likewise. (aarch64_sve_dupm_mov_immediate_p): Declare. opcodes/ * aarch64-tbl.h (AARCH64_OPERANDS): Add entries for the new SVE integer immediate operands. * aarch64-opc.h (FLD_SVE_immN, FLD_SVE_imm3, FLD_SVE_imm5) (FLD_SVE_imm5b, FLD_SVE_imm7, FLD_SVE_imm8, FLD_SVE_imm9) (FLD_SVE_immr, FLD_SVE_imms, FLD_SVE_tszh): New aarch64_field_kinds. * aarch64-opc.c (fields): Add corresponding entries. (operand_general_constraint_met_p): Handle the new SVE integer immediate operands. (aarch64_print_operand): Likewise. (aarch64_sve_dupm_mov_immediate_p): New function. * aarch64-opc-2.c: Regenerate. * aarch64-asm.h (ins_inv_limm, ins_sve_aimm, ins_sve_asimm) (ins_sve_limm_mov, ins_sve_shlimm, ins_sve_shrimm): New inserters. * aarch64-asm.c (aarch64_ins_limm_1): New function, split out from... (aarch64_ins_limm): ...here. (aarch64_ins_inv_limm): New function. (aarch64_ins_sve_aimm): Likewise. (aarch64_ins_sve_asimm): Likewise. (aarch64_ins_sve_limm_mov): Likewise. (aarch64_ins_sve_shlimm): Likewise. (aarch64_ins_sve_shrimm): Likewise. * aarch64-asm-2.c: Regenerate. * aarch64-dis.h (ext_inv_limm, ext_sve_aimm, ext_sve_asimm) (ext_sve_limm_mov, ext_sve_shlimm, ext_sve_shrimm): New extractors. * aarch64-dis.c (decode_limm): New function, split out from... (aarch64_ext_limm): ...here. (aarch64_ext_inv_limm): New function. (decode_sve_aimm): Likewise. (aarch64_ext_sve_aimm): Likewise. (aarch64_ext_sve_asimm): Likewise. (aarch64_ext_sve_limm_mov): Likewise. (aarch64_top_bit): Likewise. (aarch64_ext_sve_shlimm): Likewise. (aarch64_ext_sve_shrimm): Likewise. * aarch64-dis-2.c: Regenerate. gas/ * config/tc-aarch64.c (parse_operands): Handle the new SVE integer immediate operands.
2016-09-21 23:56:57 +08:00
case AARCH64_OPND_SVE_AIMM:
min_value = 0;
sve_aimm:
assert (opnd->shifter.kind == AARCH64_MOD_LSL);
size = aarch64_get_qualifier_esize (opnds[0].qualifier);
mask = ~((uint64_t) -1 << (size * 4) << (size * 4));
uvalue = opnd->imm.value;
shift = opnd->shifter.amount;
if (size == 1)
{
if (shift != 0)
{
set_other_error (mismatch_detail, idx,
_("no shift amount allowed for"
" 8-bit constants"));
return 0;
}
}
else
{
if (shift != 0 && shift != 8)
{
set_other_error (mismatch_detail, idx,
_("shift amount must be 0 or 8"));
return 0;
}
if (shift == 0 && (uvalue & 0xff) == 0)
{
shift = 8;
uvalue = (int64_t) uvalue / 256;
}
}
mask >>= shift;
if ((uvalue & mask) != uvalue && (uvalue | ~mask) != uvalue)
{
set_other_error (mismatch_detail, idx,
_("immediate too big for element size"));
return 0;
}
uvalue = (uvalue - min_value) & mask;
if (uvalue > 0xff)
{
set_other_error (mismatch_detail, idx,
_("invalid arithmetic immediate"));
return 0;
}
break;
case AARCH64_OPND_SVE_ASIMM:
min_value = -128;
goto sve_aimm;
[AArch64][SVE 28/32] Add SVE FP immediate operands This patch adds support for the new SVE floating-point immediate operands. One operand uses the same 8-bit encoding as base AArch64, but in a different position. The others use a single bit to select between two values. One of the single-bit operands is a choice between 0 and 1, where 0 is not a valid 8-bit encoding. I think the cleanest way of handling these single-bit immediates is therefore to use the IEEE float encoding itself as the immediate value and select between the two possible values when encoding and decoding. As described in the covering note for the patch that added F_STRICT, we get better error messages by accepting unsuffixed vector registers and leaving the qualifier matching code to report an error. This means that we carry on parsing the other operands, and so can try to parse FP immediates for invalid instructions like: fcpy z0, #2.5 In this case there is no suffix to tell us whether the immediate should be treated as single or double precision. Again, we get better error messages by picking one (arbitrary) immediate size and reporting an error for the missing suffix later. include/ * opcode/aarch64.h (AARCH64_OPND_SVE_FPIMM8): New aarch64_opnd. (AARCH64_OPND_SVE_I1_HALF_ONE, AARCH64_OPND_SVE_I1_HALF_TWO) (AARCH64_OPND_SVE_I1_ZERO_ONE): Likewise. opcodes/ * aarch64-tbl.h (AARCH64_OPERANDS): Add entries for the new SVE FP immediate operands. * aarch64-opc.h (FLD_SVE_i1): New aarch64_field_kind. * aarch64-opc.c (fields): Add corresponding entry. (operand_general_constraint_met_p): Handle the new SVE FP immediate operands. (aarch64_print_operand): Likewise. * aarch64-opc-2.c: Regenerate. * aarch64-asm.h (ins_sve_float_half_one, ins_sve_float_half_two) (ins_sve_float_zero_one): New inserters. * aarch64-asm.c (aarch64_ins_sve_float_half_one): New function. (aarch64_ins_sve_float_half_two): Likewise. (aarch64_ins_sve_float_zero_one): Likewise. * aarch64-asm-2.c: Regenerate. * aarch64-dis.h (ext_sve_float_half_one, ext_sve_float_half_two) (ext_sve_float_zero_one): New extractors. * aarch64-dis.c (aarch64_ext_sve_float_half_one): New function. (aarch64_ext_sve_float_half_two): Likewise. (aarch64_ext_sve_float_zero_one): Likewise. * aarch64-dis-2.c: Regenerate. gas/ * config/tc-aarch64.c (double_precision_operand_p): New function. (parse_operands): Use it to calculate the dp_p input to parse_aarch64_imm_float. Handle the new SVE FP immediate operands.
2016-09-21 23:57:22 +08:00
case AARCH64_OPND_SVE_I1_HALF_ONE:
assert (opnd->imm.is_fp);
if (opnd->imm.value != 0x3f000000 && opnd->imm.value != 0x3f800000)
{
set_other_error (mismatch_detail, idx,
_("floating-point value must be 0.5 or 1.0"));
return 0;
}
break;
case AARCH64_OPND_SVE_I1_HALF_TWO:
assert (opnd->imm.is_fp);
if (opnd->imm.value != 0x3f000000 && opnd->imm.value != 0x40000000)
{
set_other_error (mismatch_detail, idx,
_("floating-point value must be 0.5 or 2.0"));
return 0;
}
break;
case AARCH64_OPND_SVE_I1_ZERO_ONE:
assert (opnd->imm.is_fp);
if (opnd->imm.value != 0 && opnd->imm.value != 0x3f800000)
{
set_other_error (mismatch_detail, idx,
_("floating-point value must be 0.0 or 1.0"));
return 0;
}
break;
[AArch64][SVE 27/32] Add SVE integer immediate operands This patch adds the new SVE integer immediate operands. There are three kinds: - simple signed and unsigned ranges, but with new widths and positions. - 13-bit logical immediates. These have the same form as in base AArch64, but at a different bit position. In the case of the "MOV Zn.<T>, #<limm>" alias of DUPM, the logical immediate <limm> is not allowed to be a valid DUP immediate, since DUP is preferred over DUPM for constants that both instructions can handle. - a new 9-bit arithmetic immediate, of the form "<imm8>{, LSL #8}". In some contexts the operand is signed and in others it's unsigned. As an extension, we allow shifted immediates to be written as a single integer, e.g. "#256" is equivalent to "#1, LSL #8". We also use the shiftless form as the preferred disassembly, except for the special case of "#0, LSL #8" (a redundant encoding of 0). include/ * opcode/aarch64.h (AARCH64_OPND_SIMM5): New aarch64_opnd. (AARCH64_OPND_SVE_AIMM, AARCH64_OPND_SVE_ASIMM) (AARCH64_OPND_SVE_INV_LIMM, AARCH64_OPND_SVE_LIMM) (AARCH64_OPND_SVE_LIMM_MOV, AARCH64_OPND_SVE_SHLIMM_PRED) (AARCH64_OPND_SVE_SHLIMM_UNPRED, AARCH64_OPND_SVE_SHRIMM_PRED) (AARCH64_OPND_SVE_SHRIMM_UNPRED, AARCH64_OPND_SVE_SIMM5) (AARCH64_OPND_SVE_SIMM5B, AARCH64_OPND_SVE_SIMM6) (AARCH64_OPND_SVE_SIMM8, AARCH64_OPND_SVE_UIMM3) (AARCH64_OPND_SVE_UIMM7, AARCH64_OPND_SVE_UIMM8) (AARCH64_OPND_SVE_UIMM8_53): Likewise. (aarch64_sve_dupm_mov_immediate_p): Declare. opcodes/ * aarch64-tbl.h (AARCH64_OPERANDS): Add entries for the new SVE integer immediate operands. * aarch64-opc.h (FLD_SVE_immN, FLD_SVE_imm3, FLD_SVE_imm5) (FLD_SVE_imm5b, FLD_SVE_imm7, FLD_SVE_imm8, FLD_SVE_imm9) (FLD_SVE_immr, FLD_SVE_imms, FLD_SVE_tszh): New aarch64_field_kinds. * aarch64-opc.c (fields): Add corresponding entries. (operand_general_constraint_met_p): Handle the new SVE integer immediate operands. (aarch64_print_operand): Likewise. (aarch64_sve_dupm_mov_immediate_p): New function. * aarch64-opc-2.c: Regenerate. * aarch64-asm.h (ins_inv_limm, ins_sve_aimm, ins_sve_asimm) (ins_sve_limm_mov, ins_sve_shlimm, ins_sve_shrimm): New inserters. * aarch64-asm.c (aarch64_ins_limm_1): New function, split out from... (aarch64_ins_limm): ...here. (aarch64_ins_inv_limm): New function. (aarch64_ins_sve_aimm): Likewise. (aarch64_ins_sve_asimm): Likewise. (aarch64_ins_sve_limm_mov): Likewise. (aarch64_ins_sve_shlimm): Likewise. (aarch64_ins_sve_shrimm): Likewise. * aarch64-asm-2.c: Regenerate. * aarch64-dis.h (ext_inv_limm, ext_sve_aimm, ext_sve_asimm) (ext_sve_limm_mov, ext_sve_shlimm, ext_sve_shrimm): New extractors. * aarch64-dis.c (decode_limm): New function, split out from... (aarch64_ext_limm): ...here. (aarch64_ext_inv_limm): New function. (decode_sve_aimm): Likewise. (aarch64_ext_sve_aimm): Likewise. (aarch64_ext_sve_asimm): Likewise. (aarch64_ext_sve_limm_mov): Likewise. (aarch64_top_bit): Likewise. (aarch64_ext_sve_shlimm): Likewise. (aarch64_ext_sve_shrimm): Likewise. * aarch64-dis-2.c: Regenerate. gas/ * config/tc-aarch64.c (parse_operands): Handle the new SVE integer immediate operands.
2016-09-21 23:56:57 +08:00
case AARCH64_OPND_SVE_INV_LIMM:
{
int esize = aarch64_get_qualifier_esize (opnds[0].qualifier);
uint64_t uimm = ~opnd->imm.value;
if (!aarch64_logical_immediate_p (uimm, esize, NULL))
{
set_other_error (mismatch_detail, idx,
_("immediate out of range"));
return 0;
}
}
break;
case AARCH64_OPND_SVE_LIMM_MOV:
{
int esize = aarch64_get_qualifier_esize (opnds[0].qualifier);
uint64_t uimm = opnd->imm.value;
if (!aarch64_logical_immediate_p (uimm, esize, NULL))
{
set_other_error (mismatch_detail, idx,
_("immediate out of range"));
return 0;
}
if (!aarch64_sve_dupm_mov_immediate_p (uimm, esize))
{
set_other_error (mismatch_detail, idx,
_("invalid replicated MOV immediate"));
return 0;
}
}
break;
[AArch64][SVE 24/32] Add AARCH64_OPND_SVE_PATTERN_SCALED Some SVE instructions count the number of elements in a given vector pattern and allow a scale factor of [1, 16] to be applied to the result. This scale factor is written ", MUL #n", where "MUL" is a new operator. E.g.: UQINCD X0, POW2, MUL #2 This patch adds support for this kind of operand. All existing operators were shifts of some kind, so there was a natural range of [0, 63] regardless of context. This was then narrowered further by later checks (e.g. to [0, 31] when used for 32-bit values). In contrast, MUL doesn't really have a natural context-independent range. Rather than pick one arbitrarily, it seemed better to make the "shift" amount a full 64-bit value and leave the range test to the usual operand-checking code. I've rearranged the fields of aarch64_opnd_info so that this doesn't increase the size of the structure (although I don't think its size is critical anyway). include/ * opcode/aarch64.h (AARCH64_OPND_SVE_PATTERN_SCALED): New aarch64_opnd. (AARCH64_MOD_MUL): New aarch64_modifier_kind. (aarch64_opnd_info): Make shifter.amount an int64_t and rearrange the fields. opcodes/ * aarch64-tbl.h (AARCH64_OPERANDS): Add an entry for AARCH64_OPND_SVE_PATTERN_SCALED. * aarch64-opc.h (FLD_SVE_imm4): New aarch64_field_kind. * aarch64-opc.c (fields): Add a corresponding entry. (set_multiplier_out_of_range_error): New function. (aarch64_operand_modifiers): Add entry for AARCH64_MOD_MUL. (operand_general_constraint_met_p): Handle AARCH64_OPND_SVE_PATTERN_SCALED. (print_register_offset_address): Use PRIi64 to print the shift amount. (aarch64_print_operand): Likewise. Handle AARCH64_OPND_SVE_PATTERN_SCALED. * aarch64-opc-2.c: Regenerate. * aarch64-asm.h (ins_sve_scale): New inserter. * aarch64-asm.c (aarch64_ins_sve_scale): New function. * aarch64-asm-2.c: Regenerate. * aarch64-dis.h (ext_sve_scale): New inserter. * aarch64-dis.c (aarch64_ext_sve_scale): New function. * aarch64-dis-2.c: Regenerate. gas/ * config/tc-aarch64.c (SHIFTED_MUL): New parse_shift_mode. (parse_shift): Handle it. Reject AARCH64_MOD_MUL for all other shift modes. Skip range tests for AARCH64_MOD_MUL. (process_omitted_operand): Handle AARCH64_OPND_SVE_PATTERN_SCALED. (parse_operands): Likewise.
2016-09-21 23:55:22 +08:00
case AARCH64_OPND_SVE_PATTERN_SCALED:
assert (opnd->shifter.kind == AARCH64_MOD_MUL);
if (!value_in_range_p (opnd->shifter.amount, 1, 16))
{
set_multiplier_out_of_range_error (mismatch_detail, idx, 1, 16);
return 0;
}
break;
[AArch64][SVE 27/32] Add SVE integer immediate operands This patch adds the new SVE integer immediate operands. There are three kinds: - simple signed and unsigned ranges, but with new widths and positions. - 13-bit logical immediates. These have the same form as in base AArch64, but at a different bit position. In the case of the "MOV Zn.<T>, #<limm>" alias of DUPM, the logical immediate <limm> is not allowed to be a valid DUP immediate, since DUP is preferred over DUPM for constants that both instructions can handle. - a new 9-bit arithmetic immediate, of the form "<imm8>{, LSL #8}". In some contexts the operand is signed and in others it's unsigned. As an extension, we allow shifted immediates to be written as a single integer, e.g. "#256" is equivalent to "#1, LSL #8". We also use the shiftless form as the preferred disassembly, except for the special case of "#0, LSL #8" (a redundant encoding of 0). include/ * opcode/aarch64.h (AARCH64_OPND_SIMM5): New aarch64_opnd. (AARCH64_OPND_SVE_AIMM, AARCH64_OPND_SVE_ASIMM) (AARCH64_OPND_SVE_INV_LIMM, AARCH64_OPND_SVE_LIMM) (AARCH64_OPND_SVE_LIMM_MOV, AARCH64_OPND_SVE_SHLIMM_PRED) (AARCH64_OPND_SVE_SHLIMM_UNPRED, AARCH64_OPND_SVE_SHRIMM_PRED) (AARCH64_OPND_SVE_SHRIMM_UNPRED, AARCH64_OPND_SVE_SIMM5) (AARCH64_OPND_SVE_SIMM5B, AARCH64_OPND_SVE_SIMM6) (AARCH64_OPND_SVE_SIMM8, AARCH64_OPND_SVE_UIMM3) (AARCH64_OPND_SVE_UIMM7, AARCH64_OPND_SVE_UIMM8) (AARCH64_OPND_SVE_UIMM8_53): Likewise. (aarch64_sve_dupm_mov_immediate_p): Declare. opcodes/ * aarch64-tbl.h (AARCH64_OPERANDS): Add entries for the new SVE integer immediate operands. * aarch64-opc.h (FLD_SVE_immN, FLD_SVE_imm3, FLD_SVE_imm5) (FLD_SVE_imm5b, FLD_SVE_imm7, FLD_SVE_imm8, FLD_SVE_imm9) (FLD_SVE_immr, FLD_SVE_imms, FLD_SVE_tszh): New aarch64_field_kinds. * aarch64-opc.c (fields): Add corresponding entries. (operand_general_constraint_met_p): Handle the new SVE integer immediate operands. (aarch64_print_operand): Likewise. (aarch64_sve_dupm_mov_immediate_p): New function. * aarch64-opc-2.c: Regenerate. * aarch64-asm.h (ins_inv_limm, ins_sve_aimm, ins_sve_asimm) (ins_sve_limm_mov, ins_sve_shlimm, ins_sve_shrimm): New inserters. * aarch64-asm.c (aarch64_ins_limm_1): New function, split out from... (aarch64_ins_limm): ...here. (aarch64_ins_inv_limm): New function. (aarch64_ins_sve_aimm): Likewise. (aarch64_ins_sve_asimm): Likewise. (aarch64_ins_sve_limm_mov): Likewise. (aarch64_ins_sve_shlimm): Likewise. (aarch64_ins_sve_shrimm): Likewise. * aarch64-asm-2.c: Regenerate. * aarch64-dis.h (ext_inv_limm, ext_sve_aimm, ext_sve_asimm) (ext_sve_limm_mov, ext_sve_shlimm, ext_sve_shrimm): New extractors. * aarch64-dis.c (decode_limm): New function, split out from... (aarch64_ext_limm): ...here. (aarch64_ext_inv_limm): New function. (decode_sve_aimm): Likewise. (aarch64_ext_sve_aimm): Likewise. (aarch64_ext_sve_asimm): Likewise. (aarch64_ext_sve_limm_mov): Likewise. (aarch64_top_bit): Likewise. (aarch64_ext_sve_shlimm): Likewise. (aarch64_ext_sve_shrimm): Likewise. * aarch64-dis-2.c: Regenerate. gas/ * config/tc-aarch64.c (parse_operands): Handle the new SVE integer immediate operands.
2016-09-21 23:56:57 +08:00
case AARCH64_OPND_SVE_SHLIMM_PRED:
case AARCH64_OPND_SVE_SHLIMM_UNPRED:
case AARCH64_OPND_SVE_SHLIMM_UNPRED_22:
[AArch64][SVE 27/32] Add SVE integer immediate operands This patch adds the new SVE integer immediate operands. There are three kinds: - simple signed and unsigned ranges, but with new widths and positions. - 13-bit logical immediates. These have the same form as in base AArch64, but at a different bit position. In the case of the "MOV Zn.<T>, #<limm>" alias of DUPM, the logical immediate <limm> is not allowed to be a valid DUP immediate, since DUP is preferred over DUPM for constants that both instructions can handle. - a new 9-bit arithmetic immediate, of the form "<imm8>{, LSL #8}". In some contexts the operand is signed and in others it's unsigned. As an extension, we allow shifted immediates to be written as a single integer, e.g. "#256" is equivalent to "#1, LSL #8". We also use the shiftless form as the preferred disassembly, except for the special case of "#0, LSL #8" (a redundant encoding of 0). include/ * opcode/aarch64.h (AARCH64_OPND_SIMM5): New aarch64_opnd. (AARCH64_OPND_SVE_AIMM, AARCH64_OPND_SVE_ASIMM) (AARCH64_OPND_SVE_INV_LIMM, AARCH64_OPND_SVE_LIMM) (AARCH64_OPND_SVE_LIMM_MOV, AARCH64_OPND_SVE_SHLIMM_PRED) (AARCH64_OPND_SVE_SHLIMM_UNPRED, AARCH64_OPND_SVE_SHRIMM_PRED) (AARCH64_OPND_SVE_SHRIMM_UNPRED, AARCH64_OPND_SVE_SIMM5) (AARCH64_OPND_SVE_SIMM5B, AARCH64_OPND_SVE_SIMM6) (AARCH64_OPND_SVE_SIMM8, AARCH64_OPND_SVE_UIMM3) (AARCH64_OPND_SVE_UIMM7, AARCH64_OPND_SVE_UIMM8) (AARCH64_OPND_SVE_UIMM8_53): Likewise. (aarch64_sve_dupm_mov_immediate_p): Declare. opcodes/ * aarch64-tbl.h (AARCH64_OPERANDS): Add entries for the new SVE integer immediate operands. * aarch64-opc.h (FLD_SVE_immN, FLD_SVE_imm3, FLD_SVE_imm5) (FLD_SVE_imm5b, FLD_SVE_imm7, FLD_SVE_imm8, FLD_SVE_imm9) (FLD_SVE_immr, FLD_SVE_imms, FLD_SVE_tszh): New aarch64_field_kinds. * aarch64-opc.c (fields): Add corresponding entries. (operand_general_constraint_met_p): Handle the new SVE integer immediate operands. (aarch64_print_operand): Likewise. (aarch64_sve_dupm_mov_immediate_p): New function. * aarch64-opc-2.c: Regenerate. * aarch64-asm.h (ins_inv_limm, ins_sve_aimm, ins_sve_asimm) (ins_sve_limm_mov, ins_sve_shlimm, ins_sve_shrimm): New inserters. * aarch64-asm.c (aarch64_ins_limm_1): New function, split out from... (aarch64_ins_limm): ...here. (aarch64_ins_inv_limm): New function. (aarch64_ins_sve_aimm): Likewise. (aarch64_ins_sve_asimm): Likewise. (aarch64_ins_sve_limm_mov): Likewise. (aarch64_ins_sve_shlimm): Likewise. (aarch64_ins_sve_shrimm): Likewise. * aarch64-asm-2.c: Regenerate. * aarch64-dis.h (ext_inv_limm, ext_sve_aimm, ext_sve_asimm) (ext_sve_limm_mov, ext_sve_shlimm, ext_sve_shrimm): New extractors. * aarch64-dis.c (decode_limm): New function, split out from... (aarch64_ext_limm): ...here. (aarch64_ext_inv_limm): New function. (decode_sve_aimm): Likewise. (aarch64_ext_sve_aimm): Likewise. (aarch64_ext_sve_asimm): Likewise. (aarch64_ext_sve_limm_mov): Likewise. (aarch64_top_bit): Likewise. (aarch64_ext_sve_shlimm): Likewise. (aarch64_ext_sve_shrimm): Likewise. * aarch64-dis-2.c: Regenerate. gas/ * config/tc-aarch64.c (parse_operands): Handle the new SVE integer immediate operands.
2016-09-21 23:56:57 +08:00
size = aarch64_get_qualifier_esize (opnds[idx - 1].qualifier);
if (!value_in_range_p (opnd->imm.value, 0, 8 * size - 1))
{
set_imm_out_of_range_error (mismatch_detail, idx,
0, 8 * size - 1);
return 0;
}
break;
case AARCH64_OPND_SME_SHRIMM4:
size = 1 << get_operand_fields_width (get_operand_from_code (type));
if (!value_in_range_p (opnd->imm.value, 1, size))
{
set_imm_out_of_range_error (mismatch_detail, idx, 1, size);
return 0;
}
break;
case AARCH64_OPND_SME_SHRIMM5:
[AArch64][SVE 27/32] Add SVE integer immediate operands This patch adds the new SVE integer immediate operands. There are three kinds: - simple signed and unsigned ranges, but with new widths and positions. - 13-bit logical immediates. These have the same form as in base AArch64, but at a different bit position. In the case of the "MOV Zn.<T>, #<limm>" alias of DUPM, the logical immediate <limm> is not allowed to be a valid DUP immediate, since DUP is preferred over DUPM for constants that both instructions can handle. - a new 9-bit arithmetic immediate, of the form "<imm8>{, LSL #8}". In some contexts the operand is signed and in others it's unsigned. As an extension, we allow shifted immediates to be written as a single integer, e.g. "#256" is equivalent to "#1, LSL #8". We also use the shiftless form as the preferred disassembly, except for the special case of "#0, LSL #8" (a redundant encoding of 0). include/ * opcode/aarch64.h (AARCH64_OPND_SIMM5): New aarch64_opnd. (AARCH64_OPND_SVE_AIMM, AARCH64_OPND_SVE_ASIMM) (AARCH64_OPND_SVE_INV_LIMM, AARCH64_OPND_SVE_LIMM) (AARCH64_OPND_SVE_LIMM_MOV, AARCH64_OPND_SVE_SHLIMM_PRED) (AARCH64_OPND_SVE_SHLIMM_UNPRED, AARCH64_OPND_SVE_SHRIMM_PRED) (AARCH64_OPND_SVE_SHRIMM_UNPRED, AARCH64_OPND_SVE_SIMM5) (AARCH64_OPND_SVE_SIMM5B, AARCH64_OPND_SVE_SIMM6) (AARCH64_OPND_SVE_SIMM8, AARCH64_OPND_SVE_UIMM3) (AARCH64_OPND_SVE_UIMM7, AARCH64_OPND_SVE_UIMM8) (AARCH64_OPND_SVE_UIMM8_53): Likewise. (aarch64_sve_dupm_mov_immediate_p): Declare. opcodes/ * aarch64-tbl.h (AARCH64_OPERANDS): Add entries for the new SVE integer immediate operands. * aarch64-opc.h (FLD_SVE_immN, FLD_SVE_imm3, FLD_SVE_imm5) (FLD_SVE_imm5b, FLD_SVE_imm7, FLD_SVE_imm8, FLD_SVE_imm9) (FLD_SVE_immr, FLD_SVE_imms, FLD_SVE_tszh): New aarch64_field_kinds. * aarch64-opc.c (fields): Add corresponding entries. (operand_general_constraint_met_p): Handle the new SVE integer immediate operands. (aarch64_print_operand): Likewise. (aarch64_sve_dupm_mov_immediate_p): New function. * aarch64-opc-2.c: Regenerate. * aarch64-asm.h (ins_inv_limm, ins_sve_aimm, ins_sve_asimm) (ins_sve_limm_mov, ins_sve_shlimm, ins_sve_shrimm): New inserters. * aarch64-asm.c (aarch64_ins_limm_1): New function, split out from... (aarch64_ins_limm): ...here. (aarch64_ins_inv_limm): New function. (aarch64_ins_sve_aimm): Likewise. (aarch64_ins_sve_asimm): Likewise. (aarch64_ins_sve_limm_mov): Likewise. (aarch64_ins_sve_shlimm): Likewise. (aarch64_ins_sve_shrimm): Likewise. * aarch64-asm-2.c: Regenerate. * aarch64-dis.h (ext_inv_limm, ext_sve_aimm, ext_sve_asimm) (ext_sve_limm_mov, ext_sve_shlimm, ext_sve_shrimm): New extractors. * aarch64-dis.c (decode_limm): New function, split out from... (aarch64_ext_limm): ...here. (aarch64_ext_inv_limm): New function. (decode_sve_aimm): Likewise. (aarch64_ext_sve_aimm): Likewise. (aarch64_ext_sve_asimm): Likewise. (aarch64_ext_sve_limm_mov): Likewise. (aarch64_top_bit): Likewise. (aarch64_ext_sve_shlimm): Likewise. (aarch64_ext_sve_shrimm): Likewise. * aarch64-dis-2.c: Regenerate. gas/ * config/tc-aarch64.c (parse_operands): Handle the new SVE integer immediate operands.
2016-09-21 23:56:57 +08:00
case AARCH64_OPND_SVE_SHRIMM_PRED:
case AARCH64_OPND_SVE_SHRIMM_UNPRED:
[binutils][aarch64] New SVE_SHRIMM_UNPRED_22 operand. Include a new iclass to extract the variant from the most significant 3 bits of this operand. Instructions such as rshrnb include a constant shift amount as an operand, where the most significant three bits of this operand determine what size elements the instruction is operating on. The new SVE_SHRIMM_UNPRED_22 operand denotes this constant encoded in bits 22:20-19:18-16 while the new sve_shift_tsz_hsd iclass denotes that the SVE qualifier is encoded in bits 22:20-19. gas/ChangeLog: 2019-05-09 Matthew Malcomson <matthew.malcomson@arm.com> * config/tc-aarch64.c (parse_operands): Handle new SVE_SHRIMM_UNPRED_22 operand. include/ChangeLog: 2019-05-09 Matthew Malcomson <matthew.malcomson@arm.com> * opcode/aarch64.h (enum aarch64_opnd): New SVE_SHRIMM_UNPRED_22 operand. (enum aarch64_insn_class): Add sve_shift_tsz_hsd iclass. opcodes/ChangeLog: 2019-05-09 Matthew Malcomson <matthew.malcomson@arm.com> * aarch64-asm-2.c: Regenerated. * aarch64-dis-2.c: Regenerated. * aarch64-opc-2.c: Regenerated. * aarch64-asm.c (aarch64_ins_sve_shrimm): (aarch64_encode_variant_using_iclass): Handle sve_shift_tsz_hsd iclass encode. * aarch64-dis.c (aarch64_decode_variant_using_iclass): Handle sve_shift_tsz_hsd iclass decode. * aarch64-opc.c (operand_general_constraint_met_p): Constraint checking for SVE_SHRIMM_UNPRED_22. (aarch64_print_operand): Add printing for SVE_SHRIMM_UNPRED_22. * aarch64-tbl.h (AARCH64_OPERANDS): Use new SVE_SHRIMM_UNPRED_22 operand.
2019-05-09 17:29:22 +08:00
case AARCH64_OPND_SVE_SHRIMM_UNPRED_22:
num = (type == AARCH64_OPND_SVE_SHRIMM_UNPRED_22) ? 2 : 1;
size = aarch64_get_qualifier_esize (opnds[idx - num].qualifier);
if (!value_in_range_p (opnd->imm.value, 1, 8 * size))
[AArch64][SVE 27/32] Add SVE integer immediate operands This patch adds the new SVE integer immediate operands. There are three kinds: - simple signed and unsigned ranges, but with new widths and positions. - 13-bit logical immediates. These have the same form as in base AArch64, but at a different bit position. In the case of the "MOV Zn.<T>, #<limm>" alias of DUPM, the logical immediate <limm> is not allowed to be a valid DUP immediate, since DUP is preferred over DUPM for constants that both instructions can handle. - a new 9-bit arithmetic immediate, of the form "<imm8>{, LSL #8}". In some contexts the operand is signed and in others it's unsigned. As an extension, we allow shifted immediates to be written as a single integer, e.g. "#256" is equivalent to "#1, LSL #8". We also use the shiftless form as the preferred disassembly, except for the special case of "#0, LSL #8" (a redundant encoding of 0). include/ * opcode/aarch64.h (AARCH64_OPND_SIMM5): New aarch64_opnd. (AARCH64_OPND_SVE_AIMM, AARCH64_OPND_SVE_ASIMM) (AARCH64_OPND_SVE_INV_LIMM, AARCH64_OPND_SVE_LIMM) (AARCH64_OPND_SVE_LIMM_MOV, AARCH64_OPND_SVE_SHLIMM_PRED) (AARCH64_OPND_SVE_SHLIMM_UNPRED, AARCH64_OPND_SVE_SHRIMM_PRED) (AARCH64_OPND_SVE_SHRIMM_UNPRED, AARCH64_OPND_SVE_SIMM5) (AARCH64_OPND_SVE_SIMM5B, AARCH64_OPND_SVE_SIMM6) (AARCH64_OPND_SVE_SIMM8, AARCH64_OPND_SVE_UIMM3) (AARCH64_OPND_SVE_UIMM7, AARCH64_OPND_SVE_UIMM8) (AARCH64_OPND_SVE_UIMM8_53): Likewise. (aarch64_sve_dupm_mov_immediate_p): Declare. opcodes/ * aarch64-tbl.h (AARCH64_OPERANDS): Add entries for the new SVE integer immediate operands. * aarch64-opc.h (FLD_SVE_immN, FLD_SVE_imm3, FLD_SVE_imm5) (FLD_SVE_imm5b, FLD_SVE_imm7, FLD_SVE_imm8, FLD_SVE_imm9) (FLD_SVE_immr, FLD_SVE_imms, FLD_SVE_tszh): New aarch64_field_kinds. * aarch64-opc.c (fields): Add corresponding entries. (operand_general_constraint_met_p): Handle the new SVE integer immediate operands. (aarch64_print_operand): Likewise. (aarch64_sve_dupm_mov_immediate_p): New function. * aarch64-opc-2.c: Regenerate. * aarch64-asm.h (ins_inv_limm, ins_sve_aimm, ins_sve_asimm) (ins_sve_limm_mov, ins_sve_shlimm, ins_sve_shrimm): New inserters. * aarch64-asm.c (aarch64_ins_limm_1): New function, split out from... (aarch64_ins_limm): ...here. (aarch64_ins_inv_limm): New function. (aarch64_ins_sve_aimm): Likewise. (aarch64_ins_sve_asimm): Likewise. (aarch64_ins_sve_limm_mov): Likewise. (aarch64_ins_sve_shlimm): Likewise. (aarch64_ins_sve_shrimm): Likewise. * aarch64-asm-2.c: Regenerate. * aarch64-dis.h (ext_inv_limm, ext_sve_aimm, ext_sve_asimm) (ext_sve_limm_mov, ext_sve_shlimm, ext_sve_shrimm): New extractors. * aarch64-dis.c (decode_limm): New function, split out from... (aarch64_ext_limm): ...here. (aarch64_ext_inv_limm): New function. (decode_sve_aimm): Likewise. (aarch64_ext_sve_aimm): Likewise. (aarch64_ext_sve_asimm): Likewise. (aarch64_ext_sve_limm_mov): Likewise. (aarch64_top_bit): Likewise. (aarch64_ext_sve_shlimm): Likewise. (aarch64_ext_sve_shrimm): Likewise. * aarch64-dis-2.c: Regenerate. gas/ * config/tc-aarch64.c (parse_operands): Handle the new SVE integer immediate operands.
2016-09-21 23:56:57 +08:00
{
set_imm_out_of_range_error (mismatch_detail, idx, 1, 8*size);
return 0;
}
break;
[AArch64][SVE 27/32] Add SVE integer immediate operands This patch adds the new SVE integer immediate operands. There are three kinds: - simple signed and unsigned ranges, but with new widths and positions. - 13-bit logical immediates. These have the same form as in base AArch64, but at a different bit position. In the case of the "MOV Zn.<T>, #<limm>" alias of DUPM, the logical immediate <limm> is not allowed to be a valid DUP immediate, since DUP is preferred over DUPM for constants that both instructions can handle. - a new 9-bit arithmetic immediate, of the form "<imm8>{, LSL #8}". In some contexts the operand is signed and in others it's unsigned. As an extension, we allow shifted immediates to be written as a single integer, e.g. "#256" is equivalent to "#1, LSL #8". We also use the shiftless form as the preferred disassembly, except for the special case of "#0, LSL #8" (a redundant encoding of 0). include/ * opcode/aarch64.h (AARCH64_OPND_SIMM5): New aarch64_opnd. (AARCH64_OPND_SVE_AIMM, AARCH64_OPND_SVE_ASIMM) (AARCH64_OPND_SVE_INV_LIMM, AARCH64_OPND_SVE_LIMM) (AARCH64_OPND_SVE_LIMM_MOV, AARCH64_OPND_SVE_SHLIMM_PRED) (AARCH64_OPND_SVE_SHLIMM_UNPRED, AARCH64_OPND_SVE_SHRIMM_PRED) (AARCH64_OPND_SVE_SHRIMM_UNPRED, AARCH64_OPND_SVE_SIMM5) (AARCH64_OPND_SVE_SIMM5B, AARCH64_OPND_SVE_SIMM6) (AARCH64_OPND_SVE_SIMM8, AARCH64_OPND_SVE_UIMM3) (AARCH64_OPND_SVE_UIMM7, AARCH64_OPND_SVE_UIMM8) (AARCH64_OPND_SVE_UIMM8_53): Likewise. (aarch64_sve_dupm_mov_immediate_p): Declare. opcodes/ * aarch64-tbl.h (AARCH64_OPERANDS): Add entries for the new SVE integer immediate operands. * aarch64-opc.h (FLD_SVE_immN, FLD_SVE_imm3, FLD_SVE_imm5) (FLD_SVE_imm5b, FLD_SVE_imm7, FLD_SVE_imm8, FLD_SVE_imm9) (FLD_SVE_immr, FLD_SVE_imms, FLD_SVE_tszh): New aarch64_field_kinds. * aarch64-opc.c (fields): Add corresponding entries. (operand_general_constraint_met_p): Handle the new SVE integer immediate operands. (aarch64_print_operand): Likewise. (aarch64_sve_dupm_mov_immediate_p): New function. * aarch64-opc-2.c: Regenerate. * aarch64-asm.h (ins_inv_limm, ins_sve_aimm, ins_sve_asimm) (ins_sve_limm_mov, ins_sve_shlimm, ins_sve_shrimm): New inserters. * aarch64-asm.c (aarch64_ins_limm_1): New function, split out from... (aarch64_ins_limm): ...here. (aarch64_ins_inv_limm): New function. (aarch64_ins_sve_aimm): Likewise. (aarch64_ins_sve_asimm): Likewise. (aarch64_ins_sve_limm_mov): Likewise. (aarch64_ins_sve_shlimm): Likewise. (aarch64_ins_sve_shrimm): Likewise. * aarch64-asm-2.c: Regenerate. * aarch64-dis.h (ext_inv_limm, ext_sve_aimm, ext_sve_asimm) (ext_sve_limm_mov, ext_sve_shlimm, ext_sve_shrimm): New extractors. * aarch64-dis.c (decode_limm): New function, split out from... (aarch64_ext_limm): ...here. (aarch64_ext_inv_limm): New function. (decode_sve_aimm): Likewise. (aarch64_ext_sve_aimm): Likewise. (aarch64_ext_sve_asimm): Likewise. (aarch64_ext_sve_limm_mov): Likewise. (aarch64_top_bit): Likewise. (aarch64_ext_sve_shlimm): Likewise. (aarch64_ext_sve_shrimm): Likewise. * aarch64-dis-2.c: Regenerate. gas/ * config/tc-aarch64.c (parse_operands): Handle the new SVE integer immediate operands.
2016-09-21 23:56:57 +08:00
case AARCH64_OPND_SME_ZT0_INDEX:
if (!value_in_range_p (opnd->imm.value, 0, 56))
{
set_elem_idx_out_of_range_error (mismatch_detail, idx, 0, 56);
return 0;
}
if (opnd->imm.value % 8 != 0)
{
set_other_error (mismatch_detail, idx,
_("byte index must be a multiple of 8"));
return 0;
}
break;
default:
break;
}
break;
case AARCH64_OPND_CLASS_SYSTEM:
switch (type)
{
case AARCH64_OPND_PSTATEFIELD:
aarch64: Add maximum immediate value to aarch64_sys_reg The immediate form of MSR has a 4-bit immediate field (in CRm). However, many forms of MSR require a smaller immediate. These cases are identified by value in operand_general_constraint_met_p, but they're now the common case rather than the exception. This patch therefore adds the maximum value to the sys_reg description and gets the range from there. It also enforces the minimum of 0, which avoids a situation in which: msr dit, #2 would give the expected: Error: immediate value out of range 0 to 1 whereas: msr dit, #-1 would give: Error: immediate value out of range 0 to 15 (from the later UIMM4 checking). Also: - we were reporting the first error above against the wrong operand - TCO takes a single-bit immediate, but we previously allowed all 16 values. [https://developer.arm.com/documentation/ddi0596/2021-09/Base-Instructions/MSR--immediate---Move-immediate-value-to-Special-Register-?lang=en] opcodes/ * aarch64-opc.h (F_REG_MAX_VALUE, F_GET_REG_MAX_VALUE): New macros. * aarch64-opc.c (operand_general_constraint_met_p): Read the maximum MSR immediate value from aarch64_pstatefields. (aarch64_pstatefields): Add the maximum immediate value for each register. gas/ * testsuite/gas/aarch64/sysreg-4.s: Use an immediate value of 1 rather than 8 for the TCO test. * testsuite/gas/aarch64/sysreg-4.d: Update accordingly. * testsuite/gas/aarch64/armv8_2-a-illegal.l: Fix operand number in MSR immediate error messages. * testsuite/gas/aarch64/diagnostic.l: Likewise. * testsuite/gas/aarch64/pan-illegal.l: Likewise. * testsuite/gas/aarch64/ssbs-illegal1.l: Likewise. * testsuite/gas/aarch64/illegal-sysreg-4b.s, * testsuite/gas/aarch64/illegal-sysreg-4b.d, * testsuite/gas/aarch64/illegal-sysreg-4b.l: New test.
2021-12-02 23:00:56 +08:00
for (i = 0; aarch64_pstatefields[i].name; ++i)
if (aarch64_pstatefields[i].value == opnd->pstatefield)
break;
assert (aarch64_pstatefields[i].name);
assert (idx == 0 && opnds[1].type == AARCH64_OPND_UIMM4);
aarch64: Add maximum immediate value to aarch64_sys_reg The immediate form of MSR has a 4-bit immediate field (in CRm). However, many forms of MSR require a smaller immediate. These cases are identified by value in operand_general_constraint_met_p, but they're now the common case rather than the exception. This patch therefore adds the maximum value to the sys_reg description and gets the range from there. It also enforces the minimum of 0, which avoids a situation in which: msr dit, #2 would give the expected: Error: immediate value out of range 0 to 1 whereas: msr dit, #-1 would give: Error: immediate value out of range 0 to 15 (from the later UIMM4 checking). Also: - we were reporting the first error above against the wrong operand - TCO takes a single-bit immediate, but we previously allowed all 16 values. [https://developer.arm.com/documentation/ddi0596/2021-09/Base-Instructions/MSR--immediate---Move-immediate-value-to-Special-Register-?lang=en] opcodes/ * aarch64-opc.h (F_REG_MAX_VALUE, F_GET_REG_MAX_VALUE): New macros. * aarch64-opc.c (operand_general_constraint_met_p): Read the maximum MSR immediate value from aarch64_pstatefields. (aarch64_pstatefields): Add the maximum immediate value for each register. gas/ * testsuite/gas/aarch64/sysreg-4.s: Use an immediate value of 1 rather than 8 for the TCO test. * testsuite/gas/aarch64/sysreg-4.d: Update accordingly. * testsuite/gas/aarch64/armv8_2-a-illegal.l: Fix operand number in MSR immediate error messages. * testsuite/gas/aarch64/diagnostic.l: Likewise. * testsuite/gas/aarch64/pan-illegal.l: Likewise. * testsuite/gas/aarch64/ssbs-illegal1.l: Likewise. * testsuite/gas/aarch64/illegal-sysreg-4b.s, * testsuite/gas/aarch64/illegal-sysreg-4b.d, * testsuite/gas/aarch64/illegal-sysreg-4b.l: New test.
2021-12-02 23:00:56 +08:00
max_value = F_GET_REG_MAX_VALUE (aarch64_pstatefields[i].flags);
if (opnds[1].imm.value < 0 || opnds[1].imm.value > max_value)
{
aarch64: Add maximum immediate value to aarch64_sys_reg The immediate form of MSR has a 4-bit immediate field (in CRm). However, many forms of MSR require a smaller immediate. These cases are identified by value in operand_general_constraint_met_p, but they're now the common case rather than the exception. This patch therefore adds the maximum value to the sys_reg description and gets the range from there. It also enforces the minimum of 0, which avoids a situation in which: msr dit, #2 would give the expected: Error: immediate value out of range 0 to 1 whereas: msr dit, #-1 would give: Error: immediate value out of range 0 to 15 (from the later UIMM4 checking). Also: - we were reporting the first error above against the wrong operand - TCO takes a single-bit immediate, but we previously allowed all 16 values. [https://developer.arm.com/documentation/ddi0596/2021-09/Base-Instructions/MSR--immediate---Move-immediate-value-to-Special-Register-?lang=en] opcodes/ * aarch64-opc.h (F_REG_MAX_VALUE, F_GET_REG_MAX_VALUE): New macros. * aarch64-opc.c (operand_general_constraint_met_p): Read the maximum MSR immediate value from aarch64_pstatefields. (aarch64_pstatefields): Add the maximum immediate value for each register. gas/ * testsuite/gas/aarch64/sysreg-4.s: Use an immediate value of 1 rather than 8 for the TCO test. * testsuite/gas/aarch64/sysreg-4.d: Update accordingly. * testsuite/gas/aarch64/armv8_2-a-illegal.l: Fix operand number in MSR immediate error messages. * testsuite/gas/aarch64/diagnostic.l: Likewise. * testsuite/gas/aarch64/pan-illegal.l: Likewise. * testsuite/gas/aarch64/ssbs-illegal1.l: Likewise. * testsuite/gas/aarch64/illegal-sysreg-4b.s, * testsuite/gas/aarch64/illegal-sysreg-4b.d, * testsuite/gas/aarch64/illegal-sysreg-4b.l: New test.
2021-12-02 23:00:56 +08:00
set_imm_out_of_range_error (mismatch_detail, 1, 0, max_value);
return 0;
}
break;
case AARCH64_OPND_PRFOP:
if (opcode->iclass == ldst_regoff && opnd->prfop->value >= 24)
{
set_other_error (mismatch_detail, idx,
_("the register-index form of PRFM does"
" not accept opcodes in the range 24-31"));
return 0;
}
break;
default:
break;
}
break;
case AARCH64_OPND_CLASS_SIMD_ELEMENT:
/* Get the upper bound for the element index. */
[AArch64] Add ARMv8.3 FCMLA and FCADD instructions Add support for FCMLA and FCADD complex arithmetic SIMD instructions. FCMLA has an indexed element variant where the index range has to be treated specially because a complex number takes two elements and the indexed vector size depends on the other operands. These complex number SIMD instructions are part of ARMv8.3 https://community.arm.com/groups/processors/blog/2016/10/27/armv8-a-architecture-2016-additions include/ 2016-11-18 Szabolcs Nagy <szabolcs.nagy@arm.com> * opcode/aarch64.h (enum aarch64_opnd): Add AARCH64_OPND_IMM_ROT1, AARCH64_OPND_IMM_ROT2, AARCH64_OPND_IMM_ROT3. (enum aarch64_op): Add OP_FCMLA_ELEM. opcodes/ 2016-11-18 Szabolcs Nagy <szabolcs.nagy@arm.com> * aarch64-tbl.h (QL_V3SAMEHSD_ROT, QL_ELEMENT_ROT): Define. (aarch64_feature_simd_v8_3, SIMD_V8_3): Define. (aarch64_opcode_table): Add fcmla and fcadd. (AARCH64_OPERANDS): Add IMM_ROT{1,2,3}. * aarch64-asm.h (aarch64_ins_imm_rotate): Declare. * aarch64-asm.c (aarch64_ins_imm_rotate): Define. * aarch64-dis.h (aarch64_ext_imm_rotate): Declare. * aarch64-dis.c (aarch64_ext_imm_rotate): Define. * aarch64-opc.h (enum aarch64_field_kind): Add FLD_rotate{1,2,3}. * aarch64-opc.c (fields): Add FLD_rotate{1,2,3}. (operand_general_constraint_met_p): Rotate and index range check. (aarch64_print_operand): Handle rotate operand. * aarch64-asm-2.c: Regenerate. * aarch64-dis-2.c: Likewise. * aarch64-opc-2.c: Likewise. gas/ 2016-11-18 Szabolcs Nagy <szabolcs.nagy@arm.com> * config/tc-aarch64.c (parse_operands): Handle AARCH64_OPND_IMM_ROT*. * testsuite/gas/aarch64/advsimd-armv8_3.d: New. * testsuite/gas/aarch64/advsimd-armv8_3.s: New. * testsuite/gas/aarch64/illegal-fcmla.s: New. * testsuite/gas/aarch64/illegal-fcmla.l: New. * testsuite/gas/aarch64/illegal-fcmla.d: New.
2016-11-18 18:02:16 +08:00
if (opcode->op == OP_FCMLA_ELEM)
/* FCMLA index range depends on the vector size of other operands
and is halfed because complex numbers take two elements. */
num = aarch64_get_qualifier_nelem (opnds[0].qualifier)
* aarch64_get_qualifier_esize (opnds[0].qualifier) / 2;
else
num = 16;
num = num / aarch64_get_qualifier_esize (qualifier) - 1;
AArch64: Fix error checking for SIMD udot (by element) Committed on behalf of Matthew Malcomson: The SIMD UDOT instruction assembly has an unusual operand that selects a single 32 bit element with the mnemonic 4B. This unusual mnemonic is handled by a special operand qualifier and associated qualifier data in `aarch64_opnd_qualifiers`. The current qualifier data describes 4 1-byte elements with the structure {1, 4, 0x0, "4b", OQK_OPD_VARIANT} This makes sense, as the instruction does work on 4 1-byte elements, however some logic in the `operand_general_constraint_met_p` makes assumptions about the range of index allowed when selecting a SIMD_ELEMENT depending on element size. That function reasons that e.g. in order to select a byte-sized element in a 16 byte V register an index must allow selection of one of the 16 elements and hence its range will be in [0,15]. This reasoning breaks with the above description of a 4 part selection of 1 byte elements and allows an index outside the valid [0,3] range, triggering an assert later on in the program in `aarch64_ins_reglane`. vshcmd: > echo 'udot v0.2s, v1.8b, v2.4b[4]' | ../src/binutils-build/gas/as-new -march=armv8.4-a as-new: ../../binutils-gdb/opcodes/aarch64-asm.c:134: aarch64_ins_reglane: Assertion `reglane_index < 4' failed. {standard input}: Assembler messages: {standard input}:1: Internal error (Aborted). Please report this bug. This patch changes the operand qualifier data so that it describes a single 32 bit element. {4, 1, 0x0, "4b", OQK_OPD_VARIANT} Hence the calculation in `operand_general_constraint_met_p` provides the correct answer and the usual error checking machinery is used. vshcmd: > echo 'udot v0.2s, v1.8b, v2.4b[4]' | ../src/binutils-build/gas/as-new -march=armv8.4-a {standard input}: Assembler messages: {standard input}:1: Error: register element index out of range 0 to 3 at operand 3 -- `udot v0.2s,v1.8b,v2.4b[4]'
2018-10-17 01:49:36 +08:00
assert (aarch64_get_qualifier_nelem (qualifier) == 1);
[AArch64] Add ARMv8.3 FCMLA and FCADD instructions Add support for FCMLA and FCADD complex arithmetic SIMD instructions. FCMLA has an indexed element variant where the index range has to be treated specially because a complex number takes two elements and the indexed vector size depends on the other operands. These complex number SIMD instructions are part of ARMv8.3 https://community.arm.com/groups/processors/blog/2016/10/27/armv8-a-architecture-2016-additions include/ 2016-11-18 Szabolcs Nagy <szabolcs.nagy@arm.com> * opcode/aarch64.h (enum aarch64_opnd): Add AARCH64_OPND_IMM_ROT1, AARCH64_OPND_IMM_ROT2, AARCH64_OPND_IMM_ROT3. (enum aarch64_op): Add OP_FCMLA_ELEM. opcodes/ 2016-11-18 Szabolcs Nagy <szabolcs.nagy@arm.com> * aarch64-tbl.h (QL_V3SAMEHSD_ROT, QL_ELEMENT_ROT): Define. (aarch64_feature_simd_v8_3, SIMD_V8_3): Define. (aarch64_opcode_table): Add fcmla and fcadd. (AARCH64_OPERANDS): Add IMM_ROT{1,2,3}. * aarch64-asm.h (aarch64_ins_imm_rotate): Declare. * aarch64-asm.c (aarch64_ins_imm_rotate): Define. * aarch64-dis.h (aarch64_ext_imm_rotate): Declare. * aarch64-dis.c (aarch64_ext_imm_rotate): Define. * aarch64-opc.h (enum aarch64_field_kind): Add FLD_rotate{1,2,3}. * aarch64-opc.c (fields): Add FLD_rotate{1,2,3}. (operand_general_constraint_met_p): Rotate and index range check. (aarch64_print_operand): Handle rotate operand. * aarch64-asm-2.c: Regenerate. * aarch64-dis-2.c: Likewise. * aarch64-opc-2.c: Likewise. gas/ 2016-11-18 Szabolcs Nagy <szabolcs.nagy@arm.com> * config/tc-aarch64.c (parse_operands): Handle AARCH64_OPND_IMM_ROT*. * testsuite/gas/aarch64/advsimd-armv8_3.d: New. * testsuite/gas/aarch64/advsimd-armv8_3.s: New. * testsuite/gas/aarch64/illegal-fcmla.s: New. * testsuite/gas/aarch64/illegal-fcmla.l: New. * testsuite/gas/aarch64/illegal-fcmla.d: New.
2016-11-18 18:02:16 +08:00
/* Index out-of-range. */
if (!value_in_range_p (opnd->reglane.index, 0, num))
{
set_elem_idx_out_of_range_error (mismatch_detail, idx, 0, num);
return 0;
}
/* SMLAL<Q> <Vd>.<Ta>, <Vn>.<Tb>, <Vm>.<Ts>[<index>].
<Vm> Is the vector register (V0-V31) or (V0-V15), whose
number is encoded in "size:M:Rm":
size <Vm>
00 RESERVED
01 0:Rm
10 M:Rm
11 RESERVED */
Fix AArch64 encodings for by element instructions. Some instructions in Armv8-a place a limitation on FP16 registers that can be used as the register from which to select an element from. e.g. fmla restricts Rm to 4 bits when using an FP16 register. This restriction does not apply for all instructions, e.g. fcmla does not have this restriction as it gets an extra bit from the M field. Unfortunately, this restriction to S_H was added for all _Em operands before, meaning for a large number of instructions you couldn't use the full register file. This fixes the issue by introducing a new operand _Em16 which applies this restriction only when paired with S_H and leaves the _Em and the other qualifiers for _Em16 unbounded (i.e. using the full 5 bit range). Also the patch updates all instructions that should be affected by this. opcodes/ PR binutils/23192 * aarch64-asm-2.c: Regenerate. * aarch64-dis-2.c: Likewise. * aarch64-opc-2.c: Likewise. * aarch64-dis.c (aarch64_ext_reglane): Add AARCH64_OPND_Em16 constraint. * aarch64-opc.c (operand_general_constraint_met_p, aarch64_print_operand): Likewise. * aarch64-tbl.h (aarch64_opcode_table): Change Em to Em16 for smlal, smlal2, fmla, fmls, fmul, fmulx, sqrdmlah, sqrdlsh, fmlal, fmlsl, fmlal2, fmlsl2. (AARCH64_OPERANDS): Add Em2. gas/ PR binutils/23192 * config/tc-aarch64.c (process_omitted_operand, parse_operands): Add AARCH64_OPND_Em16 * testsuite/gas/aarch64/advsimd-armv8_3.s: Expand tests to cover upper 16 registers. * testsuite/gas/aarch64/advsimd-armv8_3.d: Likewise. * testsuite/gas/aarch64/advsimd-compnum.s: Likewise. * testsuite/gas/aarch64/advsimd-compnum.d: Likewise. * testsuite/gas/aarch64/sve.d: Likewise. include/ PR binutils/23192 *opcode/aarch64.h (aarch64_opnd): Add AARCH64_OPND_Em16.
2018-06-29 19:12:27 +08:00
if (type == AARCH64_OPND_Em16 && qualifier == AARCH64_OPND_QLF_S_H
&& !value_in_range_p (opnd->reglane.regno, 0, 15))
{
set_regno_out_of_range_error (mismatch_detail, idx, 0, 15);
return 0;
}
break;
case AARCH64_OPND_CLASS_MODIFIED_REG:
assert (idx == 1 || idx == 2);
switch (type)
{
case AARCH64_OPND_Rm_EXT:
if (!aarch64_extend_operator_p (opnd->shifter.kind)
&& opnd->shifter.kind != AARCH64_MOD_LSL)
{
set_other_error (mismatch_detail, idx,
_("extend operator expected"));
return 0;
}
/* It is not optional unless at least one of "Rd" or "Rn" is '11111'
(i.e. SP), in which case it defaults to LSL. The LSL alias is
only valid when "Rd" or "Rn" is '11111', and is preferred in that
case. */
if (!aarch64_stack_pointer_p (opnds + 0)
&& (idx != 2 || !aarch64_stack_pointer_p (opnds + 1)))
{
if (!opnd->shifter.operator_present)
{
set_other_error (mismatch_detail, idx,
_("missing extend operator"));
return 0;
}
else if (opnd->shifter.kind == AARCH64_MOD_LSL)
{
set_other_error (mismatch_detail, idx,
_("'LSL' operator not allowed"));
return 0;
}
}
assert (opnd->shifter.operator_present /* Default to LSL. */
|| opnd->shifter.kind == AARCH64_MOD_LSL);
if (!value_in_range_p (opnd->shifter.amount, 0, 4))
{
set_sft_amount_out_of_range_error (mismatch_detail, idx, 0, 4);
return 0;
}
/* In the 64-bit form, the final register operand is written as Wm
for all but the (possibly omitted) UXTX/LSL and SXTX
operators.
N.B. GAS allows X register to be used with any operator as a
programming convenience. */
if (qualifier == AARCH64_OPND_QLF_X
&& opnd->shifter.kind != AARCH64_MOD_LSL
&& opnd->shifter.kind != AARCH64_MOD_UXTX
&& opnd->shifter.kind != AARCH64_MOD_SXTX)
{
set_other_error (mismatch_detail, idx, _("W register expected"));
return 0;
}
break;
case AARCH64_OPND_Rm_SFT:
/* ROR is not available to the shifted register operand in
arithmetic instructions. */
if (!aarch64_shift_operator_p (opnd->shifter.kind))
{
set_other_error (mismatch_detail, idx,
_("shift operator expected"));
return 0;
}
if (opnd->shifter.kind == AARCH64_MOD_ROR
&& opcode->iclass != log_shift)
{
set_other_error (mismatch_detail, idx,
_("'ROR' operator not allowed"));
return 0;
}
num = qualifier == AARCH64_OPND_QLF_W ? 31 : 63;
if (!value_in_range_p (opnd->shifter.amount, 0, num))
{
set_sft_amount_out_of_range_error (mismatch_detail, idx, 0, num);
return 0;
}
break;
case AARCH64_OPND_Rm_LSL:
/* We expect here that opnd->shifter.kind != AARCH64_MOD_LSL
because the parser already restricts the type of shift to LSL only,
so another check of shift kind would be redundant. */
if (!value_in_range_p (opnd->shifter.amount, 0, 7))
{
set_sft_amount_out_of_range_error (mismatch_detail, idx, 0, 7);
return 0;
}
break;
default:
break;
}
break;
default:
break;
}
return 1;
}
/* Main entrypoint for the operand constraint checking.
Return 1 if operands of *INST meet the constraint applied by the operand
codes and operand qualifiers; otherwise return 0 and if MISMATCH_DETAIL is
not NULL, return the detail of the error in *MISMATCH_DETAIL. N.B. when
adding more constraint checking, make sure MISMATCH_DETAIL->KIND is set
with a proper error kind rather than AARCH64_OPDE_NIL (GAS asserts non-NIL
error kind when it is notified that an instruction does not pass the check).
Un-determined operand qualifiers may get established during the process. */
int
aarch64_match_operands_constraint (aarch64_inst *inst,
aarch64_operand_error *mismatch_detail)
{
int i;
DEBUG_TRACE ("enter");
i = inst->opcode->tied_operand;
aarch64: [SME] Add LD1x, ST1x, LDR and STR instructions This patch is adding new loads and stores defined by SME instructions. gas/ChangeLog: * config/tc-aarch64.c (parse_sme_address): New parser. (parse_sme_za_hv_tiles_operand_with_braces): New parser. (parse_sme_za_array): New parser. (output_operand_error_record): Print error details if present. (parse_operands): Support new operands. * testsuite/gas/aarch64/sme-5-illegal.d: New test. * testsuite/gas/aarch64/sme-5-illegal.l: New test. * testsuite/gas/aarch64/sme-5-illegal.s: New test. * testsuite/gas/aarch64/sme-5.d: New test. * testsuite/gas/aarch64/sme-5.s: New test. * testsuite/gas/aarch64/sme-6-illegal.d: New test. * testsuite/gas/aarch64/sme-6-illegal.l: New test. * testsuite/gas/aarch64/sme-6-illegal.s: New test. * testsuite/gas/aarch64/sme-6.d: New test. * testsuite/gas/aarch64/sme-6.s: New test. * testsuite/gas/aarch64/sme-7-illegal.d: New test. * testsuite/gas/aarch64/sme-7-illegal.l: New test. * testsuite/gas/aarch64/sme-7-illegal.s: New test. * testsuite/gas/aarch64/sme-7.d: New test. * testsuite/gas/aarch64/sme-7.s: New test. include/ChangeLog: * opcode/aarch64.h (enum aarch64_opnd): New operands. (enum aarch64_insn_class): Added sme_ldr and sme_str. (AARCH64_OPDE_UNTIED_IMMS): New operand error kind. opcodes/ChangeLog: * aarch64-asm.c (aarch64_ins_sme_za_hv_tiles): New inserter. (aarch64_ins_sme_za_list): New inserter. (aarch64_ins_sme_za_array): New inserter. (aarch64_ins_sme_addr_ri_u4xvl): New inserter. * aarch64-asm.h (AARCH64_DECL_OPD_INSERTER): Added ins_sme_za_list, ins_sme_za_array and ins_sme_addr_ri_u4xvl. * aarch64-dis.c (aarch64_ext_sme_za_hv_tiles): New extractor. (aarch64_ext_sme_za_list): New extractor. (aarch64_ext_sme_za_array): New extractor. (aarch64_ext_sme_addr_ri_u4xvl): New extractor. * aarch64-dis.h (AARCH64_DECL_OPD_EXTRACTOR): Added ext_sme_za_list, ext_sme_za_array and ext_sme_addr_ri_u4xvl. * aarch64-opc.c (operand_general_constraint_met_p): (aarch64_match_operands_constraint): Handle sme_ldr, sme_str and sme_misc. (aarch64_print_operand): New operands supported. * aarch64-tbl.h (OP_SVE_QUU): New qualifier. (OP_SVE_QZU): New qualifier. aarch64-asm-2.c: Regenerate. aarch64-dis-2.c: Regenerate. aarch64-opc-2.c: Regenerate.
2021-11-18 04:02:06 +08:00
if (i > 0)
{
aarch64: [SME] Add LD1x, ST1x, LDR and STR instructions This patch is adding new loads and stores defined by SME instructions. gas/ChangeLog: * config/tc-aarch64.c (parse_sme_address): New parser. (parse_sme_za_hv_tiles_operand_with_braces): New parser. (parse_sme_za_array): New parser. (output_operand_error_record): Print error details if present. (parse_operands): Support new operands. * testsuite/gas/aarch64/sme-5-illegal.d: New test. * testsuite/gas/aarch64/sme-5-illegal.l: New test. * testsuite/gas/aarch64/sme-5-illegal.s: New test. * testsuite/gas/aarch64/sme-5.d: New test. * testsuite/gas/aarch64/sme-5.s: New test. * testsuite/gas/aarch64/sme-6-illegal.d: New test. * testsuite/gas/aarch64/sme-6-illegal.l: New test. * testsuite/gas/aarch64/sme-6-illegal.s: New test. * testsuite/gas/aarch64/sme-6.d: New test. * testsuite/gas/aarch64/sme-6.s: New test. * testsuite/gas/aarch64/sme-7-illegal.d: New test. * testsuite/gas/aarch64/sme-7-illegal.l: New test. * testsuite/gas/aarch64/sme-7-illegal.s: New test. * testsuite/gas/aarch64/sme-7.d: New test. * testsuite/gas/aarch64/sme-7.s: New test. include/ChangeLog: * opcode/aarch64.h (enum aarch64_opnd): New operands. (enum aarch64_insn_class): Added sme_ldr and sme_str. (AARCH64_OPDE_UNTIED_IMMS): New operand error kind. opcodes/ChangeLog: * aarch64-asm.c (aarch64_ins_sme_za_hv_tiles): New inserter. (aarch64_ins_sme_za_list): New inserter. (aarch64_ins_sme_za_array): New inserter. (aarch64_ins_sme_addr_ri_u4xvl): New inserter. * aarch64-asm.h (AARCH64_DECL_OPD_INSERTER): Added ins_sme_za_list, ins_sme_za_array and ins_sme_addr_ri_u4xvl. * aarch64-dis.c (aarch64_ext_sme_za_hv_tiles): New extractor. (aarch64_ext_sme_za_list): New extractor. (aarch64_ext_sme_za_array): New extractor. (aarch64_ext_sme_addr_ri_u4xvl): New extractor. * aarch64-dis.h (AARCH64_DECL_OPD_EXTRACTOR): Added ext_sme_za_list, ext_sme_za_array and ext_sme_addr_ri_u4xvl. * aarch64-opc.c (operand_general_constraint_met_p): (aarch64_match_operands_constraint): Handle sme_ldr, sme_str and sme_misc. (aarch64_print_operand): New operands supported. * aarch64-tbl.h (OP_SVE_QUU): New qualifier. (OP_SVE_QZU): New qualifier. aarch64-asm-2.c: Regenerate. aarch64-dis-2.c: Regenerate. aarch64-opc-2.c: Regenerate.
2021-11-18 04:02:06 +08:00
/* Check for tied_operands with specific opcode iclass. */
switch (inst->opcode->iclass)
{
/* For SME LDR and STR instructions #imm must have the same numerical
value for both operands.
*/
case sme_ldr:
case sme_str:
assert (inst->operands[0].type == AARCH64_OPND_SME_ZA_array_off4);
aarch64: [SME] Add LD1x, ST1x, LDR and STR instructions This patch is adding new loads and stores defined by SME instructions. gas/ChangeLog: * config/tc-aarch64.c (parse_sme_address): New parser. (parse_sme_za_hv_tiles_operand_with_braces): New parser. (parse_sme_za_array): New parser. (output_operand_error_record): Print error details if present. (parse_operands): Support new operands. * testsuite/gas/aarch64/sme-5-illegal.d: New test. * testsuite/gas/aarch64/sme-5-illegal.l: New test. * testsuite/gas/aarch64/sme-5-illegal.s: New test. * testsuite/gas/aarch64/sme-5.d: New test. * testsuite/gas/aarch64/sme-5.s: New test. * testsuite/gas/aarch64/sme-6-illegal.d: New test. * testsuite/gas/aarch64/sme-6-illegal.l: New test. * testsuite/gas/aarch64/sme-6-illegal.s: New test. * testsuite/gas/aarch64/sme-6.d: New test. * testsuite/gas/aarch64/sme-6.s: New test. * testsuite/gas/aarch64/sme-7-illegal.d: New test. * testsuite/gas/aarch64/sme-7-illegal.l: New test. * testsuite/gas/aarch64/sme-7-illegal.s: New test. * testsuite/gas/aarch64/sme-7.d: New test. * testsuite/gas/aarch64/sme-7.s: New test. include/ChangeLog: * opcode/aarch64.h (enum aarch64_opnd): New operands. (enum aarch64_insn_class): Added sme_ldr and sme_str. (AARCH64_OPDE_UNTIED_IMMS): New operand error kind. opcodes/ChangeLog: * aarch64-asm.c (aarch64_ins_sme_za_hv_tiles): New inserter. (aarch64_ins_sme_za_list): New inserter. (aarch64_ins_sme_za_array): New inserter. (aarch64_ins_sme_addr_ri_u4xvl): New inserter. * aarch64-asm.h (AARCH64_DECL_OPD_INSERTER): Added ins_sme_za_list, ins_sme_za_array and ins_sme_addr_ri_u4xvl. * aarch64-dis.c (aarch64_ext_sme_za_hv_tiles): New extractor. (aarch64_ext_sme_za_list): New extractor. (aarch64_ext_sme_za_array): New extractor. (aarch64_ext_sme_addr_ri_u4xvl): New extractor. * aarch64-dis.h (AARCH64_DECL_OPD_EXTRACTOR): Added ext_sme_za_list, ext_sme_za_array and ext_sme_addr_ri_u4xvl. * aarch64-opc.c (operand_general_constraint_met_p): (aarch64_match_operands_constraint): Handle sme_ldr, sme_str and sme_misc. (aarch64_print_operand): New operands supported. * aarch64-tbl.h (OP_SVE_QUU): New qualifier. (OP_SVE_QZU): New qualifier. aarch64-asm-2.c: Regenerate. aarch64-dis-2.c: Regenerate. aarch64-opc-2.c: Regenerate.
2021-11-18 04:02:06 +08:00
assert (inst->operands[1].type == AARCH64_OPND_SME_ADDR_RI_U4xVL);
if (inst->operands[0].indexed_za.index.imm
aarch64: [SME] Add LD1x, ST1x, LDR and STR instructions This patch is adding new loads and stores defined by SME instructions. gas/ChangeLog: * config/tc-aarch64.c (parse_sme_address): New parser. (parse_sme_za_hv_tiles_operand_with_braces): New parser. (parse_sme_za_array): New parser. (output_operand_error_record): Print error details if present. (parse_operands): Support new operands. * testsuite/gas/aarch64/sme-5-illegal.d: New test. * testsuite/gas/aarch64/sme-5-illegal.l: New test. * testsuite/gas/aarch64/sme-5-illegal.s: New test. * testsuite/gas/aarch64/sme-5.d: New test. * testsuite/gas/aarch64/sme-5.s: New test. * testsuite/gas/aarch64/sme-6-illegal.d: New test. * testsuite/gas/aarch64/sme-6-illegal.l: New test. * testsuite/gas/aarch64/sme-6-illegal.s: New test. * testsuite/gas/aarch64/sme-6.d: New test. * testsuite/gas/aarch64/sme-6.s: New test. * testsuite/gas/aarch64/sme-7-illegal.d: New test. * testsuite/gas/aarch64/sme-7-illegal.l: New test. * testsuite/gas/aarch64/sme-7-illegal.s: New test. * testsuite/gas/aarch64/sme-7.d: New test. * testsuite/gas/aarch64/sme-7.s: New test. include/ChangeLog: * opcode/aarch64.h (enum aarch64_opnd): New operands. (enum aarch64_insn_class): Added sme_ldr and sme_str. (AARCH64_OPDE_UNTIED_IMMS): New operand error kind. opcodes/ChangeLog: * aarch64-asm.c (aarch64_ins_sme_za_hv_tiles): New inserter. (aarch64_ins_sme_za_list): New inserter. (aarch64_ins_sme_za_array): New inserter. (aarch64_ins_sme_addr_ri_u4xvl): New inserter. * aarch64-asm.h (AARCH64_DECL_OPD_INSERTER): Added ins_sme_za_list, ins_sme_za_array and ins_sme_addr_ri_u4xvl. * aarch64-dis.c (aarch64_ext_sme_za_hv_tiles): New extractor. (aarch64_ext_sme_za_list): New extractor. (aarch64_ext_sme_za_array): New extractor. (aarch64_ext_sme_addr_ri_u4xvl): New extractor. * aarch64-dis.h (AARCH64_DECL_OPD_EXTRACTOR): Added ext_sme_za_list, ext_sme_za_array and ext_sme_addr_ri_u4xvl. * aarch64-opc.c (operand_general_constraint_met_p): (aarch64_match_operands_constraint): Handle sme_ldr, sme_str and sme_misc. (aarch64_print_operand): New operands supported. * aarch64-tbl.h (OP_SVE_QUU): New qualifier. (OP_SVE_QZU): New qualifier. aarch64-asm-2.c: Regenerate. aarch64-dis-2.c: Regenerate. aarch64-opc-2.c: Regenerate.
2021-11-18 04:02:06 +08:00
!= inst->operands[1].addr.offset.imm)
{
if (mismatch_detail)
{
mismatch_detail->kind = AARCH64_OPDE_UNTIED_IMMS;
mismatch_detail->index = i;
}
return 0;
}
break;
default:
{
/* Check for cases where a source register needs to be the
same as the destination register. Do this before
matching qualifiers since if an instruction has both
invalid tying and invalid qualifiers, the error about
qualifiers would suggest several alternative instructions
that also have invalid tying. */
enum aarch64_operand_class op_class
= aarch64_get_operand_class (inst->operands[0].type);
assert (aarch64_get_operand_class (inst->operands[i].type)
== op_class);
if (op_class == AARCH64_OPND_CLASS_SVE_REGLIST
? ((inst->operands[0].reglist.first_regno
!= inst->operands[i].reglist.first_regno)
|| (inst->operands[0].reglist.num_regs
!= inst->operands[i].reglist.num_regs)
|| (inst->operands[0].reglist.stride
!= inst->operands[i].reglist.stride))
: (inst->operands[0].reg.regno
!= inst->operands[i].reg.regno))
{
if (mismatch_detail)
{
mismatch_detail->kind = AARCH64_OPDE_UNTIED_OPERAND;
mismatch_detail->index = i;
mismatch_detail->error = NULL;
}
return 0;
}
break;
}
aarch64: [SME] Add LD1x, ST1x, LDR and STR instructions This patch is adding new loads and stores defined by SME instructions. gas/ChangeLog: * config/tc-aarch64.c (parse_sme_address): New parser. (parse_sme_za_hv_tiles_operand_with_braces): New parser. (parse_sme_za_array): New parser. (output_operand_error_record): Print error details if present. (parse_operands): Support new operands. * testsuite/gas/aarch64/sme-5-illegal.d: New test. * testsuite/gas/aarch64/sme-5-illegal.l: New test. * testsuite/gas/aarch64/sme-5-illegal.s: New test. * testsuite/gas/aarch64/sme-5.d: New test. * testsuite/gas/aarch64/sme-5.s: New test. * testsuite/gas/aarch64/sme-6-illegal.d: New test. * testsuite/gas/aarch64/sme-6-illegal.l: New test. * testsuite/gas/aarch64/sme-6-illegal.s: New test. * testsuite/gas/aarch64/sme-6.d: New test. * testsuite/gas/aarch64/sme-6.s: New test. * testsuite/gas/aarch64/sme-7-illegal.d: New test. * testsuite/gas/aarch64/sme-7-illegal.l: New test. * testsuite/gas/aarch64/sme-7-illegal.s: New test. * testsuite/gas/aarch64/sme-7.d: New test. * testsuite/gas/aarch64/sme-7.s: New test. include/ChangeLog: * opcode/aarch64.h (enum aarch64_opnd): New operands. (enum aarch64_insn_class): Added sme_ldr and sme_str. (AARCH64_OPDE_UNTIED_IMMS): New operand error kind. opcodes/ChangeLog: * aarch64-asm.c (aarch64_ins_sme_za_hv_tiles): New inserter. (aarch64_ins_sme_za_list): New inserter. (aarch64_ins_sme_za_array): New inserter. (aarch64_ins_sme_addr_ri_u4xvl): New inserter. * aarch64-asm.h (AARCH64_DECL_OPD_INSERTER): Added ins_sme_za_list, ins_sme_za_array and ins_sme_addr_ri_u4xvl. * aarch64-dis.c (aarch64_ext_sme_za_hv_tiles): New extractor. (aarch64_ext_sme_za_list): New extractor. (aarch64_ext_sme_za_array): New extractor. (aarch64_ext_sme_addr_ri_u4xvl): New extractor. * aarch64-dis.h (AARCH64_DECL_OPD_EXTRACTOR): Added ext_sme_za_list, ext_sme_za_array and ext_sme_addr_ri_u4xvl. * aarch64-opc.c (operand_general_constraint_met_p): (aarch64_match_operands_constraint): Handle sme_ldr, sme_str and sme_misc. (aarch64_print_operand): New operands supported. * aarch64-tbl.h (OP_SVE_QUU): New qualifier. (OP_SVE_QZU): New qualifier. aarch64-asm-2.c: Regenerate. aarch64-dis-2.c: Regenerate. aarch64-opc-2.c: Regenerate.
2021-11-18 04:02:06 +08:00
}
}
/* Match operands' qualifier.
*INST has already had qualifier establish for some, if not all, of
its operands; we need to find out whether these established
qualifiers match one of the qualifier sequence in
INST->OPCODE->QUALIFIERS_LIST. If yes, we will assign each operand
with the corresponding qualifier in such a sequence.
Only basic operand constraint checking is done here; the more thorough
constraint checking will carried out by operand_general_constraint_met_p,
which has be to called after this in order to get all of the operands'
qualifiers established. */
int invalid_count;
if (match_operands_qualifier (inst, true /* update_p */,
&invalid_count) == 0)
{
DEBUG_TRACE ("FAIL on operand qualifier matching");
if (mismatch_detail)
{
/* Return an error type to indicate that it is the qualifier
matching failure; we don't care about which operand as there
are enough information in the opcode table to reproduce it. */
mismatch_detail->kind = AARCH64_OPDE_INVALID_VARIANT;
mismatch_detail->index = -1;
mismatch_detail->error = NULL;
mismatch_detail->data[0].i = invalid_count;
}
return 0;
}
/* Match operands' constraint. */
for (i = 0; i < AARCH64_MAX_OPND_NUM; ++i)
{
enum aarch64_opnd type = inst->opcode->operands[i];
if (type == AARCH64_OPND_NIL)
break;
if (inst->operands[i].skip)
{
DEBUG_TRACE ("skip the incomplete operand %d", i);
continue;
}
if (operand_general_constraint_met_p (inst->operands, i, type,
inst->opcode, mismatch_detail) == 0)
{
DEBUG_TRACE ("FAIL on operand %d", i);
return 0;
}
}
DEBUG_TRACE ("PASS");
return 1;
}
/* Replace INST->OPCODE with OPCODE and return the replaced OPCODE.
Also updates the TYPE of each INST->OPERANDS with the corresponding
value of OPCODE->OPERANDS.
Note that some operand qualifiers may need to be manually cleared by
the caller before it further calls the aarch64_opcode_encode; by
doing this, it helps the qualifier matching facilities work
properly. */
const aarch64_opcode*
aarch64_replace_opcode (aarch64_inst *inst, const aarch64_opcode *opcode)
{
int i;
const aarch64_opcode *old = inst->opcode;
inst->opcode = opcode;
/* Update the operand types. */
for (i = 0; i < AARCH64_MAX_OPND_NUM; ++i)
{
inst->operands[i].type = opcode->operands[i];
if (opcode->operands[i] == AARCH64_OPND_NIL)
break;
}
DEBUG_TRACE ("replace %s with %s", old->name, opcode->name);
return old;
}
int
aarch64_operand_index (const enum aarch64_opnd *operands, enum aarch64_opnd operand)
{
int i;
for (i = 0; i < AARCH64_MAX_OPND_NUM; ++i)
if (operands[i] == operand)
return i;
else if (operands[i] == AARCH64_OPND_NIL)
break;
return -1;
}
/* R0...R30, followed by FOR31. */
#define BANK(R, FOR31) \
{ R (0), R (1), R (2), R (3), R (4), R (5), R (6), R (7), \
R (8), R (9), R (10), R (11), R (12), R (13), R (14), R (15), \
R (16), R (17), R (18), R (19), R (20), R (21), R (22), R (23), \
R (24), R (25), R (26), R (27), R (28), R (29), R (30), FOR31 }
/* [0][0] 32-bit integer regs with sp Wn
[0][1] 64-bit integer regs with sp Xn sf=1
[1][0] 32-bit integer regs with #0 Wn
[1][1] 64-bit integer regs with #0 Xn sf=1 */
static const char *int_reg[2][2][32] = {
#define R32(X) "w" #X
#define R64(X) "x" #X
{ BANK (R32, "wsp"), BANK (R64, "sp") },
{ BANK (R32, "wzr"), BANK (R64, "xzr") }
#undef R64
#undef R32
};
[AArch64][SVE 25/32] Add support for SVE addressing modes This patch adds most of the new SVE addressing modes and associated operands. A follow-on patch adds MUL VL, since handling it separately makes the changes easier to read. The patch also introduces a new "operand-dependent data" field to the operand flags, based closely on the existing one for opcode flags. For SVE this new field needs only 2 bits, but it could be widened in future if necessary. include/ * opcode/aarch64.h (AARCH64_OPND_SVE_ADDR_RI_U6): New aarch64_opnd. (AARCH64_OPND_SVE_ADDR_RI_U6x2, AARCH64_OPND_SVE_ADDR_RI_U6x4) (AARCH64_OPND_SVE_ADDR_RI_U6x8, AARCH64_OPND_SVE_ADDR_RR) (AARCH64_OPND_SVE_ADDR_RR_LSL1, AARCH64_OPND_SVE_ADDR_RR_LSL2) (AARCH64_OPND_SVE_ADDR_RR_LSL3, AARCH64_OPND_SVE_ADDR_RX) (AARCH64_OPND_SVE_ADDR_RX_LSL1, AARCH64_OPND_SVE_ADDR_RX_LSL2) (AARCH64_OPND_SVE_ADDR_RX_LSL3, AARCH64_OPND_SVE_ADDR_RZ) (AARCH64_OPND_SVE_ADDR_RZ_LSL1, AARCH64_OPND_SVE_ADDR_RZ_LSL2) (AARCH64_OPND_SVE_ADDR_RZ_LSL3, AARCH64_OPND_SVE_ADDR_RZ_XTW_14) (AARCH64_OPND_SVE_ADDR_RZ_XTW_22, AARCH64_OPND_SVE_ADDR_RZ_XTW1_14) (AARCH64_OPND_SVE_ADDR_RZ_XTW1_22, AARCH64_OPND_SVE_ADDR_RZ_XTW2_14) (AARCH64_OPND_SVE_ADDR_RZ_XTW2_22, AARCH64_OPND_SVE_ADDR_RZ_XTW3_14) (AARCH64_OPND_SVE_ADDR_RZ_XTW3_22, AARCH64_OPND_SVE_ADDR_ZI_U5) (AARCH64_OPND_SVE_ADDR_ZI_U5x2, AARCH64_OPND_SVE_ADDR_ZI_U5x4) (AARCH64_OPND_SVE_ADDR_ZI_U5x8, AARCH64_OPND_SVE_ADDR_ZZ_LSL) (AARCH64_OPND_SVE_ADDR_ZZ_SXTW, AARCH64_OPND_SVE_ADDR_ZZ_UXTW): Likewise. opcodes/ * aarch64-tbl.h (AARCH64_OPERANDS): Add entries for the new SVE address operands. * aarch64-opc.h (FLD_SVE_imm6, FLD_SVE_msz, FLD_SVE_xs_14) (FLD_SVE_xs_22): New aarch64_field_kinds. (OPD_F_OD_MASK, OPD_F_OD_LSB, OPD_F_NO_ZR): New flags. (get_operand_specific_data): New function. * aarch64-opc.c (fields): Add entries for FLD_SVE_imm6, FLD_SVE_msz, FLD_SVE_xs_14 and FLD_SVE_xs_22. (operand_general_constraint_met_p): Handle the new SVE address operands. (sve_reg): New array. (get_addr_sve_reg_name): New function. (aarch64_print_operand): Handle the new SVE address operands. * aarch64-opc-2.c: Regenerate. * aarch64-asm.h (ins_sve_addr_ri_u6, ins_sve_addr_rr_lsl) (ins_sve_addr_rz_xtw, ins_sve_addr_zi_u5, ins_sve_addr_zz_lsl) (ins_sve_addr_zz_sxtw, ins_sve_addr_zz_uxtw): New inserters. * aarch64-asm.c (aarch64_ins_sve_addr_ri_u6): New function. (aarch64_ins_sve_addr_rr_lsl): Likewise. (aarch64_ins_sve_addr_rz_xtw): Likewise. (aarch64_ins_sve_addr_zi_u5): Likewise. (aarch64_ins_sve_addr_zz): Likewise. (aarch64_ins_sve_addr_zz_lsl): Likewise. (aarch64_ins_sve_addr_zz_sxtw): Likewise. (aarch64_ins_sve_addr_zz_uxtw): Likewise. * aarch64-asm-2.c: Regenerate. * aarch64-dis.h (ext_sve_addr_ri_u6, ext_sve_addr_rr_lsl) (ext_sve_addr_rz_xtw, ext_sve_addr_zi_u5, ext_sve_addr_zz_lsl) (ext_sve_addr_zz_sxtw, ext_sve_addr_zz_uxtw): New extractors. * aarch64-dis.c (aarch64_ext_sve_add_reg_imm): New function. (aarch64_ext_sve_addr_ri_u6): Likewise. (aarch64_ext_sve_addr_rr_lsl): Likewise. (aarch64_ext_sve_addr_rz_xtw): Likewise. (aarch64_ext_sve_addr_zi_u5): Likewise. (aarch64_ext_sve_addr_zz): Likewise. (aarch64_ext_sve_addr_zz_lsl): Likewise. (aarch64_ext_sve_addr_zz_sxtw): Likewise. (aarch64_ext_sve_addr_zz_uxtw): Likewise. * aarch64-dis-2.c: Regenerate. gas/ * config/tc-aarch64.c (REG_TYPE_SVE_BASE, REG_TYPE_SVE_OFFSET): New register types. (get_reg_expected_msg): Handle them. (aarch64_addr_reg_parse): New function, split out from aarch64_reg_parse_32_64. Handle Z registers too. (aarch64_reg_parse_32_64): Call it. (parse_address_main): Add base_qualifier, offset_qualifier, base_type and offset_type parameters. Handle SVE base and offset registers. (parse_address): Update call to parse_address_main. (parse_sve_address): New function. (parse_operands): Parse the new SVE address operands.
2016-09-21 23:55:49 +08:00
/* Names of the SVE vector registers, first with .S suffixes,
then with .D suffixes. */
static const char *sve_reg[2][32] = {
#define ZS(X) "z" #X ".s"
#define ZD(X) "z" #X ".d"
BANK (ZS, ZS (31)), BANK (ZD, ZD (31))
#undef ZD
#undef ZS
};
#undef BANK
/* Return the integer register name.
if SP_REG_P is not 0, R31 is an SP reg, other R31 is the zero reg. */
static inline const char *
get_int_reg_name (int regno, aarch64_opnd_qualifier_t qualifier, int sp_reg_p)
{
const int has_zr = sp_reg_p ? 0 : 1;
const int is_64 = aarch64_get_qualifier_esize (qualifier) == 4 ? 0 : 1;
return int_reg[has_zr][is_64][regno];
}
/* Like get_int_reg_name, but IS_64 is always 1. */
static inline const char *
get_64bit_int_reg_name (int regno, int sp_reg_p)
{
const int has_zr = sp_reg_p ? 0 : 1;
return int_reg[has_zr][1][regno];
}
/* Get the name of the integer offset register in OPND, using the shift type
to decide whether it's a word or doubleword. */
static inline const char *
get_offset_int_reg_name (const aarch64_opnd_info *opnd)
{
switch (opnd->shifter.kind)
{
case AARCH64_MOD_UXTW:
case AARCH64_MOD_SXTW:
return get_int_reg_name (opnd->addr.offset.regno, AARCH64_OPND_QLF_W, 0);
case AARCH64_MOD_LSL:
case AARCH64_MOD_SXTX:
return get_int_reg_name (opnd->addr.offset.regno, AARCH64_OPND_QLF_X, 0);
default:
abort ();
}
}
[AArch64][SVE 25/32] Add support for SVE addressing modes This patch adds most of the new SVE addressing modes and associated operands. A follow-on patch adds MUL VL, since handling it separately makes the changes easier to read. The patch also introduces a new "operand-dependent data" field to the operand flags, based closely on the existing one for opcode flags. For SVE this new field needs only 2 bits, but it could be widened in future if necessary. include/ * opcode/aarch64.h (AARCH64_OPND_SVE_ADDR_RI_U6): New aarch64_opnd. (AARCH64_OPND_SVE_ADDR_RI_U6x2, AARCH64_OPND_SVE_ADDR_RI_U6x4) (AARCH64_OPND_SVE_ADDR_RI_U6x8, AARCH64_OPND_SVE_ADDR_RR) (AARCH64_OPND_SVE_ADDR_RR_LSL1, AARCH64_OPND_SVE_ADDR_RR_LSL2) (AARCH64_OPND_SVE_ADDR_RR_LSL3, AARCH64_OPND_SVE_ADDR_RX) (AARCH64_OPND_SVE_ADDR_RX_LSL1, AARCH64_OPND_SVE_ADDR_RX_LSL2) (AARCH64_OPND_SVE_ADDR_RX_LSL3, AARCH64_OPND_SVE_ADDR_RZ) (AARCH64_OPND_SVE_ADDR_RZ_LSL1, AARCH64_OPND_SVE_ADDR_RZ_LSL2) (AARCH64_OPND_SVE_ADDR_RZ_LSL3, AARCH64_OPND_SVE_ADDR_RZ_XTW_14) (AARCH64_OPND_SVE_ADDR_RZ_XTW_22, AARCH64_OPND_SVE_ADDR_RZ_XTW1_14) (AARCH64_OPND_SVE_ADDR_RZ_XTW1_22, AARCH64_OPND_SVE_ADDR_RZ_XTW2_14) (AARCH64_OPND_SVE_ADDR_RZ_XTW2_22, AARCH64_OPND_SVE_ADDR_RZ_XTW3_14) (AARCH64_OPND_SVE_ADDR_RZ_XTW3_22, AARCH64_OPND_SVE_ADDR_ZI_U5) (AARCH64_OPND_SVE_ADDR_ZI_U5x2, AARCH64_OPND_SVE_ADDR_ZI_U5x4) (AARCH64_OPND_SVE_ADDR_ZI_U5x8, AARCH64_OPND_SVE_ADDR_ZZ_LSL) (AARCH64_OPND_SVE_ADDR_ZZ_SXTW, AARCH64_OPND_SVE_ADDR_ZZ_UXTW): Likewise. opcodes/ * aarch64-tbl.h (AARCH64_OPERANDS): Add entries for the new SVE address operands. * aarch64-opc.h (FLD_SVE_imm6, FLD_SVE_msz, FLD_SVE_xs_14) (FLD_SVE_xs_22): New aarch64_field_kinds. (OPD_F_OD_MASK, OPD_F_OD_LSB, OPD_F_NO_ZR): New flags. (get_operand_specific_data): New function. * aarch64-opc.c (fields): Add entries for FLD_SVE_imm6, FLD_SVE_msz, FLD_SVE_xs_14 and FLD_SVE_xs_22. (operand_general_constraint_met_p): Handle the new SVE address operands. (sve_reg): New array. (get_addr_sve_reg_name): New function. (aarch64_print_operand): Handle the new SVE address operands. * aarch64-opc-2.c: Regenerate. * aarch64-asm.h (ins_sve_addr_ri_u6, ins_sve_addr_rr_lsl) (ins_sve_addr_rz_xtw, ins_sve_addr_zi_u5, ins_sve_addr_zz_lsl) (ins_sve_addr_zz_sxtw, ins_sve_addr_zz_uxtw): New inserters. * aarch64-asm.c (aarch64_ins_sve_addr_ri_u6): New function. (aarch64_ins_sve_addr_rr_lsl): Likewise. (aarch64_ins_sve_addr_rz_xtw): Likewise. (aarch64_ins_sve_addr_zi_u5): Likewise. (aarch64_ins_sve_addr_zz): Likewise. (aarch64_ins_sve_addr_zz_lsl): Likewise. (aarch64_ins_sve_addr_zz_sxtw): Likewise. (aarch64_ins_sve_addr_zz_uxtw): Likewise. * aarch64-asm-2.c: Regenerate. * aarch64-dis.h (ext_sve_addr_ri_u6, ext_sve_addr_rr_lsl) (ext_sve_addr_rz_xtw, ext_sve_addr_zi_u5, ext_sve_addr_zz_lsl) (ext_sve_addr_zz_sxtw, ext_sve_addr_zz_uxtw): New extractors. * aarch64-dis.c (aarch64_ext_sve_add_reg_imm): New function. (aarch64_ext_sve_addr_ri_u6): Likewise. (aarch64_ext_sve_addr_rr_lsl): Likewise. (aarch64_ext_sve_addr_rz_xtw): Likewise. (aarch64_ext_sve_addr_zi_u5): Likewise. (aarch64_ext_sve_addr_zz): Likewise. (aarch64_ext_sve_addr_zz_lsl): Likewise. (aarch64_ext_sve_addr_zz_sxtw): Likewise. (aarch64_ext_sve_addr_zz_uxtw): Likewise. * aarch64-dis-2.c: Regenerate. gas/ * config/tc-aarch64.c (REG_TYPE_SVE_BASE, REG_TYPE_SVE_OFFSET): New register types. (get_reg_expected_msg): Handle them. (aarch64_addr_reg_parse): New function, split out from aarch64_reg_parse_32_64. Handle Z registers too. (aarch64_reg_parse_32_64): Call it. (parse_address_main): Add base_qualifier, offset_qualifier, base_type and offset_type parameters. Handle SVE base and offset registers. (parse_address): Update call to parse_address_main. (parse_sve_address): New function. (parse_operands): Parse the new SVE address operands.
2016-09-21 23:55:49 +08:00
/* Get the name of the SVE vector offset register in OPND, using the operand
qualifier to decide whether the suffix should be .S or .D. */
static inline const char *
get_addr_sve_reg_name (int regno, aarch64_opnd_qualifier_t qualifier)
{
assert (qualifier == AARCH64_OPND_QLF_S_S
|| qualifier == AARCH64_OPND_QLF_S_D);
return sve_reg[qualifier == AARCH64_OPND_QLF_S_D][regno];
}
/* Types for expanding an encoded 8-bit value to a floating-point value. */
typedef union
{
uint64_t i;
double d;
} double_conv_t;
typedef union
{
uint32_t i;
float f;
} single_conv_t;
typedef union
{
uint32_t i;
float f;
} half_conv_t;
/* IMM8 is an 8-bit floating-point constant with sign, 3-bit exponent and
normalized 4 bits of precision, encoded in "a:b:c:d:e:f:g:h" or FLD_imm8
(depending on the type of the instruction). IMM8 will be expanded to a
single-precision floating-point value (SIZE == 4) or a double-precision
floating-point value (SIZE == 8). A half-precision floating-point value
(SIZE == 2) is expanded to a single-precision floating-point value. The
expanded value is returned. */
static uint64_t
expand_fp_imm (int size, uint32_t imm8)
{
uint64_t imm = 0;
uint32_t imm8_7, imm8_6_0, imm8_6, imm8_6_repl4;
imm8_7 = (imm8 >> 7) & 0x01; /* imm8<7> */
imm8_6_0 = imm8 & 0x7f; /* imm8<6:0> */
imm8_6 = imm8_6_0 >> 6; /* imm8<6> */
imm8_6_repl4 = (imm8_6 << 3) | (imm8_6 << 2)
| (imm8_6 << 1) | imm8_6; /* Replicate(imm8<6>,4) */
if (size == 8)
{
imm = (imm8_7 << (63-32)) /* imm8<7> */
| ((imm8_6 ^ 1) << (62-32)) /* NOT(imm8<6) */
| (imm8_6_repl4 << (58-32)) | (imm8_6 << (57-32))
| (imm8_6 << (56-32)) | (imm8_6 << (55-32)) /* Replicate(imm8<6>,7) */
| (imm8_6_0 << (48-32)); /* imm8<6>:imm8<5:0> */
imm <<= 32;
}
else if (size == 4 || size == 2)
{
imm = (imm8_7 << 31) /* imm8<7> */
| ((imm8_6 ^ 1) << 30) /* NOT(imm8<6>) */
| (imm8_6_repl4 << 26) /* Replicate(imm8<6>,4) */
| (imm8_6_0 << 19); /* imm8<6>:imm8<5:0> */
}
else
{
/* An unsupported size. */
assert (0);
}
return imm;
}
libopcodes/aarch64: add support for disassembler styling This commit enables disassembler styling for AArch64. After this commit it is possible to have objdump style AArch64 disassembler output (using --disassembler-color option). Once the required GDB patches are merged, GDB will also style the disassembler output. The changes to support styling are mostly split between two files opcodes/aarch64-dis.c and opcodes/aarch64-opc.c. The entry point for the AArch64 disassembler can be found in aarch64-dis.c, this file handles printing the instruction mnemonics, and assembler directives (e.g. '.byte', '.word', etc). Some operands, mostly relating to assembler directives are also printed from this file. This commit changes all of this to pass through suitable styling information. However, for most "normal" instructions, the instruction operands are printed using a two step process. From aarch64-dis.c, in the print_operands function, the function aarch64_print_operand is called, this function is in aarch64-opc.c, and converts an instruction operand into a string. Then, back in print_operands (aarch64-dis.c), the operand string is printed. Unfortunately, the string returned by aarch64_print_operand can be quite complex, it will include syntax elements, like '[' and ']', in addition to register names and immediate values. In some cases, a single operand will expand into what will appear (to the user) as multiple operands separated with a ','. This makes the task of styling more complex, all these different components need to by styled differently, so we need to get the styling information out of aarch64_print_operand in some way. The solution that I propose here is similar to the solution that I used for the i386 disassembler. Currently, aarch64_print_operand uses snprintf to write the operand text into a buffer provided by the caller. What I propose is that we pass an extra argument to the aarch64_print_operand function, this argument will be a structure, the structure contains a callback function and some state. When aarch64_print_operand needs to format part of its output this can be done by using the callback function within the new structure, this callback returns a string with special embedded markers that indicate which mode should be used for each piece of text. Back in aarch64-dis.c we can spot these special style markers and use this to split the disassembler output up and apply the correct style to each piece. To make aarch64-opc.c clearer a series of new static functions have been added, e.g. 'style_reg', 'style_imm', etc. Each of these functions formats a piece of text in a different style, 'register' and 'immediate' in this case. Here's an example taken from aarch64-opc.c of the new functions in use: snprintf (buf, size, "[%s, %s]!", style_reg (styler, base), style_imm (styler, "#%d", opnd->addr.offset.imm)); The aarch64_print_operand function is also called from the assembler to aid in printing diagnostic messages. Right now I have no plans to add styling to the assembler output, and so, the callback function used in the assembler ignores the styling information and just returns an plain string. I've used the source files in gas/testsuite/gas/aarch64/ for testing, and have manually gone through and checked that the styling looks reasonable, however, I'm not an AArch64 expert, so it is possible that the odd piece is styled incorrectly. Please point out any mistakes I've made. With objdump disassembler color turned off, there should be no change in the output after this commit.
2022-04-28 20:31:07 +08:00
/* Return a string based on FMT with the register style applied. */
static const char *
style_reg (struct aarch64_styler *styler, const char *fmt, ...)
{
const char *txt;
va_list ap;
va_start (ap, fmt);
txt = styler->apply_style (styler, dis_style_register, fmt, ap);
va_end (ap);
return txt;
}
/* Return a string based on FMT with the immediate style applied. */
static const char *
style_imm (struct aarch64_styler *styler, const char *fmt, ...)
{
const char *txt;
va_list ap;
va_start (ap, fmt);
txt = styler->apply_style (styler, dis_style_immediate, fmt, ap);
va_end (ap);
return txt;
}
/* Return a string based on FMT with the sub-mnemonic style applied. */
static const char *
style_sub_mnem (struct aarch64_styler *styler, const char *fmt, ...)
{
const char *txt;
va_list ap;
va_start (ap, fmt);
txt = styler->apply_style (styler, dis_style_sub_mnemonic, fmt, ap);
va_end (ap);
return txt;
}
/* Return a string based on FMT with the address style applied. */
static const char *
style_addr (struct aarch64_styler *styler, const char *fmt, ...)
{
const char *txt;
va_list ap;
va_start (ap, fmt);
txt = styler->apply_style (styler, dis_style_address, fmt, ap);
va_end (ap);
return txt;
}
/* Produce the string representation of the register list operand *OPND
in the buffer pointed by BUF of size SIZE. PREFIX is the part of
the register name that comes before the register number, such as "v". */
static void
print_register_list (char *buf, size_t size, const aarch64_opnd_info *opnd,
libopcodes/aarch64: add support for disassembler styling This commit enables disassembler styling for AArch64. After this commit it is possible to have objdump style AArch64 disassembler output (using --disassembler-color option). Once the required GDB patches are merged, GDB will also style the disassembler output. The changes to support styling are mostly split between two files opcodes/aarch64-dis.c and opcodes/aarch64-opc.c. The entry point for the AArch64 disassembler can be found in aarch64-dis.c, this file handles printing the instruction mnemonics, and assembler directives (e.g. '.byte', '.word', etc). Some operands, mostly relating to assembler directives are also printed from this file. This commit changes all of this to pass through suitable styling information. However, for most "normal" instructions, the instruction operands are printed using a two step process. From aarch64-dis.c, in the print_operands function, the function aarch64_print_operand is called, this function is in aarch64-opc.c, and converts an instruction operand into a string. Then, back in print_operands (aarch64-dis.c), the operand string is printed. Unfortunately, the string returned by aarch64_print_operand can be quite complex, it will include syntax elements, like '[' and ']', in addition to register names and immediate values. In some cases, a single operand will expand into what will appear (to the user) as multiple operands separated with a ','. This makes the task of styling more complex, all these different components need to by styled differently, so we need to get the styling information out of aarch64_print_operand in some way. The solution that I propose here is similar to the solution that I used for the i386 disassembler. Currently, aarch64_print_operand uses snprintf to write the operand text into a buffer provided by the caller. What I propose is that we pass an extra argument to the aarch64_print_operand function, this argument will be a structure, the structure contains a callback function and some state. When aarch64_print_operand needs to format part of its output this can be done by using the callback function within the new structure, this callback returns a string with special embedded markers that indicate which mode should be used for each piece of text. Back in aarch64-dis.c we can spot these special style markers and use this to split the disassembler output up and apply the correct style to each piece. To make aarch64-opc.c clearer a series of new static functions have been added, e.g. 'style_reg', 'style_imm', etc. Each of these functions formats a piece of text in a different style, 'register' and 'immediate' in this case. Here's an example taken from aarch64-opc.c of the new functions in use: snprintf (buf, size, "[%s, %s]!", style_reg (styler, base), style_imm (styler, "#%d", opnd->addr.offset.imm)); The aarch64_print_operand function is also called from the assembler to aid in printing diagnostic messages. Right now I have no plans to add styling to the assembler output, and so, the callback function used in the assembler ignores the styling information and just returns an plain string. I've used the source files in gas/testsuite/gas/aarch64/ for testing, and have manually gone through and checked that the styling looks reasonable, however, I'm not an AArch64 expert, so it is possible that the odd piece is styled incorrectly. Please point out any mistakes I've made. With objdump disassembler color turned off, there should be no change in the output after this commit.
2022-04-28 20:31:07 +08:00
const char *prefix, struct aarch64_styler *styler)
{
const int mask = (prefix[0] == 'p' ? 15 : 31);
const int num_regs = opnd->reglist.num_regs;
const int stride = opnd->reglist.stride;
const int first_reg = opnd->reglist.first_regno;
const int last_reg = (first_reg + (num_regs - 1) * stride) & mask;
const char *qlf_name = aarch64_get_qualifier_name (opnd->qualifier);
libopcodes/aarch64: add support for disassembler styling This commit enables disassembler styling for AArch64. After this commit it is possible to have objdump style AArch64 disassembler output (using --disassembler-color option). Once the required GDB patches are merged, GDB will also style the disassembler output. The changes to support styling are mostly split between two files opcodes/aarch64-dis.c and opcodes/aarch64-opc.c. The entry point for the AArch64 disassembler can be found in aarch64-dis.c, this file handles printing the instruction mnemonics, and assembler directives (e.g. '.byte', '.word', etc). Some operands, mostly relating to assembler directives are also printed from this file. This commit changes all of this to pass through suitable styling information. However, for most "normal" instructions, the instruction operands are printed using a two step process. From aarch64-dis.c, in the print_operands function, the function aarch64_print_operand is called, this function is in aarch64-opc.c, and converts an instruction operand into a string. Then, back in print_operands (aarch64-dis.c), the operand string is printed. Unfortunately, the string returned by aarch64_print_operand can be quite complex, it will include syntax elements, like '[' and ']', in addition to register names and immediate values. In some cases, a single operand will expand into what will appear (to the user) as multiple operands separated with a ','. This makes the task of styling more complex, all these different components need to by styled differently, so we need to get the styling information out of aarch64_print_operand in some way. The solution that I propose here is similar to the solution that I used for the i386 disassembler. Currently, aarch64_print_operand uses snprintf to write the operand text into a buffer provided by the caller. What I propose is that we pass an extra argument to the aarch64_print_operand function, this argument will be a structure, the structure contains a callback function and some state. When aarch64_print_operand needs to format part of its output this can be done by using the callback function within the new structure, this callback returns a string with special embedded markers that indicate which mode should be used for each piece of text. Back in aarch64-dis.c we can spot these special style markers and use this to split the disassembler output up and apply the correct style to each piece. To make aarch64-opc.c clearer a series of new static functions have been added, e.g. 'style_reg', 'style_imm', etc. Each of these functions formats a piece of text in a different style, 'register' and 'immediate' in this case. Here's an example taken from aarch64-opc.c of the new functions in use: snprintf (buf, size, "[%s, %s]!", style_reg (styler, base), style_imm (styler, "#%d", opnd->addr.offset.imm)); The aarch64_print_operand function is also called from the assembler to aid in printing diagnostic messages. Right now I have no plans to add styling to the assembler output, and so, the callback function used in the assembler ignores the styling information and just returns an plain string. I've used the source files in gas/testsuite/gas/aarch64/ for testing, and have manually gone through and checked that the styling looks reasonable, however, I'm not an AArch64 expert, so it is possible that the odd piece is styled incorrectly. Please point out any mistakes I've made. With objdump disassembler color turned off, there should be no change in the output after this commit.
2022-04-28 20:31:07 +08:00
char tb[16]; /* Temporary buffer. */
assert (opnd->type != AARCH64_OPND_LEt || opnd->reglist.has_index);
assert (num_regs >= 1 && num_regs <= 4);
/* Prepare the index if any. */
if (opnd->reglist.has_index)
/* PR 21096: The %100 is to silence a warning about possible truncation. */
libopcodes/aarch64: add support for disassembler styling This commit enables disassembler styling for AArch64. After this commit it is possible to have objdump style AArch64 disassembler output (using --disassembler-color option). Once the required GDB patches are merged, GDB will also style the disassembler output. The changes to support styling are mostly split between two files opcodes/aarch64-dis.c and opcodes/aarch64-opc.c. The entry point for the AArch64 disassembler can be found in aarch64-dis.c, this file handles printing the instruction mnemonics, and assembler directives (e.g. '.byte', '.word', etc). Some operands, mostly relating to assembler directives are also printed from this file. This commit changes all of this to pass through suitable styling information. However, for most "normal" instructions, the instruction operands are printed using a two step process. From aarch64-dis.c, in the print_operands function, the function aarch64_print_operand is called, this function is in aarch64-opc.c, and converts an instruction operand into a string. Then, back in print_operands (aarch64-dis.c), the operand string is printed. Unfortunately, the string returned by aarch64_print_operand can be quite complex, it will include syntax elements, like '[' and ']', in addition to register names and immediate values. In some cases, a single operand will expand into what will appear (to the user) as multiple operands separated with a ','. This makes the task of styling more complex, all these different components need to by styled differently, so we need to get the styling information out of aarch64_print_operand in some way. The solution that I propose here is similar to the solution that I used for the i386 disassembler. Currently, aarch64_print_operand uses snprintf to write the operand text into a buffer provided by the caller. What I propose is that we pass an extra argument to the aarch64_print_operand function, this argument will be a structure, the structure contains a callback function and some state. When aarch64_print_operand needs to format part of its output this can be done by using the callback function within the new structure, this callback returns a string with special embedded markers that indicate which mode should be used for each piece of text. Back in aarch64-dis.c we can spot these special style markers and use this to split the disassembler output up and apply the correct style to each piece. To make aarch64-opc.c clearer a series of new static functions have been added, e.g. 'style_reg', 'style_imm', etc. Each of these functions formats a piece of text in a different style, 'register' and 'immediate' in this case. Here's an example taken from aarch64-opc.c of the new functions in use: snprintf (buf, size, "[%s, %s]!", style_reg (styler, base), style_imm (styler, "#%d", opnd->addr.offset.imm)); The aarch64_print_operand function is also called from the assembler to aid in printing diagnostic messages. Right now I have no plans to add styling to the assembler output, and so, the callback function used in the assembler ignores the styling information and just returns an plain string. I've used the source files in gas/testsuite/gas/aarch64/ for testing, and have manually gone through and checked that the styling looks reasonable, however, I'm not an AArch64 expert, so it is possible that the odd piece is styled incorrectly. Please point out any mistakes I've made. With objdump disassembler color turned off, there should be no change in the output after this commit.
2022-04-28 20:31:07 +08:00
snprintf (tb, sizeof (tb), "[%s]",
style_imm (styler, "%" PRIi64, (opnd->reglist.index % 100)));
else
tb[0] = '\0';
/* The hyphenated form is preferred for disassembly if there is
more than one register in the list, and the register numbers
are monotonically increasing in increments of one. */
if (stride == 1 && num_regs > 1
&& ((opnd->type != AARCH64_OPND_SME_Zt2)
&& (opnd->type != AARCH64_OPND_SME_Zt3)
&& (opnd->type != AARCH64_OPND_SME_Zt4)))
libopcodes/aarch64: add support for disassembler styling This commit enables disassembler styling for AArch64. After this commit it is possible to have objdump style AArch64 disassembler output (using --disassembler-color option). Once the required GDB patches are merged, GDB will also style the disassembler output. The changes to support styling are mostly split between two files opcodes/aarch64-dis.c and opcodes/aarch64-opc.c. The entry point for the AArch64 disassembler can be found in aarch64-dis.c, this file handles printing the instruction mnemonics, and assembler directives (e.g. '.byte', '.word', etc). Some operands, mostly relating to assembler directives are also printed from this file. This commit changes all of this to pass through suitable styling information. However, for most "normal" instructions, the instruction operands are printed using a two step process. From aarch64-dis.c, in the print_operands function, the function aarch64_print_operand is called, this function is in aarch64-opc.c, and converts an instruction operand into a string. Then, back in print_operands (aarch64-dis.c), the operand string is printed. Unfortunately, the string returned by aarch64_print_operand can be quite complex, it will include syntax elements, like '[' and ']', in addition to register names and immediate values. In some cases, a single operand will expand into what will appear (to the user) as multiple operands separated with a ','. This makes the task of styling more complex, all these different components need to by styled differently, so we need to get the styling information out of aarch64_print_operand in some way. The solution that I propose here is similar to the solution that I used for the i386 disassembler. Currently, aarch64_print_operand uses snprintf to write the operand text into a buffer provided by the caller. What I propose is that we pass an extra argument to the aarch64_print_operand function, this argument will be a structure, the structure contains a callback function and some state. When aarch64_print_operand needs to format part of its output this can be done by using the callback function within the new structure, this callback returns a string with special embedded markers that indicate which mode should be used for each piece of text. Back in aarch64-dis.c we can spot these special style markers and use this to split the disassembler output up and apply the correct style to each piece. To make aarch64-opc.c clearer a series of new static functions have been added, e.g. 'style_reg', 'style_imm', etc. Each of these functions formats a piece of text in a different style, 'register' and 'immediate' in this case. Here's an example taken from aarch64-opc.c of the new functions in use: snprintf (buf, size, "[%s, %s]!", style_reg (styler, base), style_imm (styler, "#%d", opnd->addr.offset.imm)); The aarch64_print_operand function is also called from the assembler to aid in printing diagnostic messages. Right now I have no plans to add styling to the assembler output, and so, the callback function used in the assembler ignores the styling information and just returns an plain string. I've used the source files in gas/testsuite/gas/aarch64/ for testing, and have manually gone through and checked that the styling looks reasonable, however, I'm not an AArch64 expert, so it is possible that the odd piece is styled incorrectly. Please point out any mistakes I've made. With objdump disassembler color turned off, there should be no change in the output after this commit.
2022-04-28 20:31:07 +08:00
snprintf (buf, size, "{%s-%s}%s",
style_reg (styler, "%s%d.%s", prefix, first_reg, qlf_name),
style_reg (styler, "%s%d.%s", prefix, last_reg, qlf_name), tb);
else
{
const int reg0 = first_reg;
const int reg1 = (first_reg + stride) & mask;
const int reg2 = (first_reg + stride * 2) & mask;
const int reg3 = (first_reg + stride * 3) & mask;
switch (num_regs)
{
case 1:
libopcodes/aarch64: add support for disassembler styling This commit enables disassembler styling for AArch64. After this commit it is possible to have objdump style AArch64 disassembler output (using --disassembler-color option). Once the required GDB patches are merged, GDB will also style the disassembler output. The changes to support styling are mostly split between two files opcodes/aarch64-dis.c and opcodes/aarch64-opc.c. The entry point for the AArch64 disassembler can be found in aarch64-dis.c, this file handles printing the instruction mnemonics, and assembler directives (e.g. '.byte', '.word', etc). Some operands, mostly relating to assembler directives are also printed from this file. This commit changes all of this to pass through suitable styling information. However, for most "normal" instructions, the instruction operands are printed using a two step process. From aarch64-dis.c, in the print_operands function, the function aarch64_print_operand is called, this function is in aarch64-opc.c, and converts an instruction operand into a string. Then, back in print_operands (aarch64-dis.c), the operand string is printed. Unfortunately, the string returned by aarch64_print_operand can be quite complex, it will include syntax elements, like '[' and ']', in addition to register names and immediate values. In some cases, a single operand will expand into what will appear (to the user) as multiple operands separated with a ','. This makes the task of styling more complex, all these different components need to by styled differently, so we need to get the styling information out of aarch64_print_operand in some way. The solution that I propose here is similar to the solution that I used for the i386 disassembler. Currently, aarch64_print_operand uses snprintf to write the operand text into a buffer provided by the caller. What I propose is that we pass an extra argument to the aarch64_print_operand function, this argument will be a structure, the structure contains a callback function and some state. When aarch64_print_operand needs to format part of its output this can be done by using the callback function within the new structure, this callback returns a string with special embedded markers that indicate which mode should be used for each piece of text. Back in aarch64-dis.c we can spot these special style markers and use this to split the disassembler output up and apply the correct style to each piece. To make aarch64-opc.c clearer a series of new static functions have been added, e.g. 'style_reg', 'style_imm', etc. Each of these functions formats a piece of text in a different style, 'register' and 'immediate' in this case. Here's an example taken from aarch64-opc.c of the new functions in use: snprintf (buf, size, "[%s, %s]!", style_reg (styler, base), style_imm (styler, "#%d", opnd->addr.offset.imm)); The aarch64_print_operand function is also called from the assembler to aid in printing diagnostic messages. Right now I have no plans to add styling to the assembler output, and so, the callback function used in the assembler ignores the styling information and just returns an plain string. I've used the source files in gas/testsuite/gas/aarch64/ for testing, and have manually gone through and checked that the styling looks reasonable, however, I'm not an AArch64 expert, so it is possible that the odd piece is styled incorrectly. Please point out any mistakes I've made. With objdump disassembler color turned off, there should be no change in the output after this commit.
2022-04-28 20:31:07 +08:00
snprintf (buf, size, "{%s}%s",
style_reg (styler, "%s%d.%s", prefix, reg0, qlf_name),
tb);
break;
case 2:
libopcodes/aarch64: add support for disassembler styling This commit enables disassembler styling for AArch64. After this commit it is possible to have objdump style AArch64 disassembler output (using --disassembler-color option). Once the required GDB patches are merged, GDB will also style the disassembler output. The changes to support styling are mostly split between two files opcodes/aarch64-dis.c and opcodes/aarch64-opc.c. The entry point for the AArch64 disassembler can be found in aarch64-dis.c, this file handles printing the instruction mnemonics, and assembler directives (e.g. '.byte', '.word', etc). Some operands, mostly relating to assembler directives are also printed from this file. This commit changes all of this to pass through suitable styling information. However, for most "normal" instructions, the instruction operands are printed using a two step process. From aarch64-dis.c, in the print_operands function, the function aarch64_print_operand is called, this function is in aarch64-opc.c, and converts an instruction operand into a string. Then, back in print_operands (aarch64-dis.c), the operand string is printed. Unfortunately, the string returned by aarch64_print_operand can be quite complex, it will include syntax elements, like '[' and ']', in addition to register names and immediate values. In some cases, a single operand will expand into what will appear (to the user) as multiple operands separated with a ','. This makes the task of styling more complex, all these different components need to by styled differently, so we need to get the styling information out of aarch64_print_operand in some way. The solution that I propose here is similar to the solution that I used for the i386 disassembler. Currently, aarch64_print_operand uses snprintf to write the operand text into a buffer provided by the caller. What I propose is that we pass an extra argument to the aarch64_print_operand function, this argument will be a structure, the structure contains a callback function and some state. When aarch64_print_operand needs to format part of its output this can be done by using the callback function within the new structure, this callback returns a string with special embedded markers that indicate which mode should be used for each piece of text. Back in aarch64-dis.c we can spot these special style markers and use this to split the disassembler output up and apply the correct style to each piece. To make aarch64-opc.c clearer a series of new static functions have been added, e.g. 'style_reg', 'style_imm', etc. Each of these functions formats a piece of text in a different style, 'register' and 'immediate' in this case. Here's an example taken from aarch64-opc.c of the new functions in use: snprintf (buf, size, "[%s, %s]!", style_reg (styler, base), style_imm (styler, "#%d", opnd->addr.offset.imm)); The aarch64_print_operand function is also called from the assembler to aid in printing diagnostic messages. Right now I have no plans to add styling to the assembler output, and so, the callback function used in the assembler ignores the styling information and just returns an plain string. I've used the source files in gas/testsuite/gas/aarch64/ for testing, and have manually gone through and checked that the styling looks reasonable, however, I'm not an AArch64 expert, so it is possible that the odd piece is styled incorrectly. Please point out any mistakes I've made. With objdump disassembler color turned off, there should be no change in the output after this commit.
2022-04-28 20:31:07 +08:00
snprintf (buf, size, "{%s, %s}%s",
style_reg (styler, "%s%d.%s", prefix, reg0, qlf_name),
style_reg (styler, "%s%d.%s", prefix, reg1, qlf_name),
tb);
break;
case 3:
libopcodes/aarch64: add support for disassembler styling This commit enables disassembler styling for AArch64. After this commit it is possible to have objdump style AArch64 disassembler output (using --disassembler-color option). Once the required GDB patches are merged, GDB will also style the disassembler output. The changes to support styling are mostly split between two files opcodes/aarch64-dis.c and opcodes/aarch64-opc.c. The entry point for the AArch64 disassembler can be found in aarch64-dis.c, this file handles printing the instruction mnemonics, and assembler directives (e.g. '.byte', '.word', etc). Some operands, mostly relating to assembler directives are also printed from this file. This commit changes all of this to pass through suitable styling information. However, for most "normal" instructions, the instruction operands are printed using a two step process. From aarch64-dis.c, in the print_operands function, the function aarch64_print_operand is called, this function is in aarch64-opc.c, and converts an instruction operand into a string. Then, back in print_operands (aarch64-dis.c), the operand string is printed. Unfortunately, the string returned by aarch64_print_operand can be quite complex, it will include syntax elements, like '[' and ']', in addition to register names and immediate values. In some cases, a single operand will expand into what will appear (to the user) as multiple operands separated with a ','. This makes the task of styling more complex, all these different components need to by styled differently, so we need to get the styling information out of aarch64_print_operand in some way. The solution that I propose here is similar to the solution that I used for the i386 disassembler. Currently, aarch64_print_operand uses snprintf to write the operand text into a buffer provided by the caller. What I propose is that we pass an extra argument to the aarch64_print_operand function, this argument will be a structure, the structure contains a callback function and some state. When aarch64_print_operand needs to format part of its output this can be done by using the callback function within the new structure, this callback returns a string with special embedded markers that indicate which mode should be used for each piece of text. Back in aarch64-dis.c we can spot these special style markers and use this to split the disassembler output up and apply the correct style to each piece. To make aarch64-opc.c clearer a series of new static functions have been added, e.g. 'style_reg', 'style_imm', etc. Each of these functions formats a piece of text in a different style, 'register' and 'immediate' in this case. Here's an example taken from aarch64-opc.c of the new functions in use: snprintf (buf, size, "[%s, %s]!", style_reg (styler, base), style_imm (styler, "#%d", opnd->addr.offset.imm)); The aarch64_print_operand function is also called from the assembler to aid in printing diagnostic messages. Right now I have no plans to add styling to the assembler output, and so, the callback function used in the assembler ignores the styling information and just returns an plain string. I've used the source files in gas/testsuite/gas/aarch64/ for testing, and have manually gone through and checked that the styling looks reasonable, however, I'm not an AArch64 expert, so it is possible that the odd piece is styled incorrectly. Please point out any mistakes I've made. With objdump disassembler color turned off, there should be no change in the output after this commit.
2022-04-28 20:31:07 +08:00
snprintf (buf, size, "{%s, %s, %s}%s",
style_reg (styler, "%s%d.%s", prefix, reg0, qlf_name),
style_reg (styler, "%s%d.%s", prefix, reg1, qlf_name),
style_reg (styler, "%s%d.%s", prefix, reg2, qlf_name),
tb);
break;
case 4:
libopcodes/aarch64: add support for disassembler styling This commit enables disassembler styling for AArch64. After this commit it is possible to have objdump style AArch64 disassembler output (using --disassembler-color option). Once the required GDB patches are merged, GDB will also style the disassembler output. The changes to support styling are mostly split between two files opcodes/aarch64-dis.c and opcodes/aarch64-opc.c. The entry point for the AArch64 disassembler can be found in aarch64-dis.c, this file handles printing the instruction mnemonics, and assembler directives (e.g. '.byte', '.word', etc). Some operands, mostly relating to assembler directives are also printed from this file. This commit changes all of this to pass through suitable styling information. However, for most "normal" instructions, the instruction operands are printed using a two step process. From aarch64-dis.c, in the print_operands function, the function aarch64_print_operand is called, this function is in aarch64-opc.c, and converts an instruction operand into a string. Then, back in print_operands (aarch64-dis.c), the operand string is printed. Unfortunately, the string returned by aarch64_print_operand can be quite complex, it will include syntax elements, like '[' and ']', in addition to register names and immediate values. In some cases, a single operand will expand into what will appear (to the user) as multiple operands separated with a ','. This makes the task of styling more complex, all these different components need to by styled differently, so we need to get the styling information out of aarch64_print_operand in some way. The solution that I propose here is similar to the solution that I used for the i386 disassembler. Currently, aarch64_print_operand uses snprintf to write the operand text into a buffer provided by the caller. What I propose is that we pass an extra argument to the aarch64_print_operand function, this argument will be a structure, the structure contains a callback function and some state. When aarch64_print_operand needs to format part of its output this can be done by using the callback function within the new structure, this callback returns a string with special embedded markers that indicate which mode should be used for each piece of text. Back in aarch64-dis.c we can spot these special style markers and use this to split the disassembler output up and apply the correct style to each piece. To make aarch64-opc.c clearer a series of new static functions have been added, e.g. 'style_reg', 'style_imm', etc. Each of these functions formats a piece of text in a different style, 'register' and 'immediate' in this case. Here's an example taken from aarch64-opc.c of the new functions in use: snprintf (buf, size, "[%s, %s]!", style_reg (styler, base), style_imm (styler, "#%d", opnd->addr.offset.imm)); The aarch64_print_operand function is also called from the assembler to aid in printing diagnostic messages. Right now I have no plans to add styling to the assembler output, and so, the callback function used in the assembler ignores the styling information and just returns an plain string. I've used the source files in gas/testsuite/gas/aarch64/ for testing, and have manually gone through and checked that the styling looks reasonable, however, I'm not an AArch64 expert, so it is possible that the odd piece is styled incorrectly. Please point out any mistakes I've made. With objdump disassembler color turned off, there should be no change in the output after this commit.
2022-04-28 20:31:07 +08:00
snprintf (buf, size, "{%s, %s, %s, %s}%s",
style_reg (styler, "%s%d.%s", prefix, reg0, qlf_name),
style_reg (styler, "%s%d.%s", prefix, reg1, qlf_name),
style_reg (styler, "%s%d.%s", prefix, reg2, qlf_name),
style_reg (styler, "%s%d.%s", prefix, reg3, qlf_name),
tb);
break;
}
}
}
/* Print the register+immediate address in OPND to BUF, which has SIZE
characters. BASE is the name of the base register. */
static void
print_immediate_offset_address (char *buf, size_t size,
const aarch64_opnd_info *opnd,
libopcodes/aarch64: add support for disassembler styling This commit enables disassembler styling for AArch64. After this commit it is possible to have objdump style AArch64 disassembler output (using --disassembler-color option). Once the required GDB patches are merged, GDB will also style the disassembler output. The changes to support styling are mostly split between two files opcodes/aarch64-dis.c and opcodes/aarch64-opc.c. The entry point for the AArch64 disassembler can be found in aarch64-dis.c, this file handles printing the instruction mnemonics, and assembler directives (e.g. '.byte', '.word', etc). Some operands, mostly relating to assembler directives are also printed from this file. This commit changes all of this to pass through suitable styling information. However, for most "normal" instructions, the instruction operands are printed using a two step process. From aarch64-dis.c, in the print_operands function, the function aarch64_print_operand is called, this function is in aarch64-opc.c, and converts an instruction operand into a string. Then, back in print_operands (aarch64-dis.c), the operand string is printed. Unfortunately, the string returned by aarch64_print_operand can be quite complex, it will include syntax elements, like '[' and ']', in addition to register names and immediate values. In some cases, a single operand will expand into what will appear (to the user) as multiple operands separated with a ','. This makes the task of styling more complex, all these different components need to by styled differently, so we need to get the styling information out of aarch64_print_operand in some way. The solution that I propose here is similar to the solution that I used for the i386 disassembler. Currently, aarch64_print_operand uses snprintf to write the operand text into a buffer provided by the caller. What I propose is that we pass an extra argument to the aarch64_print_operand function, this argument will be a structure, the structure contains a callback function and some state. When aarch64_print_operand needs to format part of its output this can be done by using the callback function within the new structure, this callback returns a string with special embedded markers that indicate which mode should be used for each piece of text. Back in aarch64-dis.c we can spot these special style markers and use this to split the disassembler output up and apply the correct style to each piece. To make aarch64-opc.c clearer a series of new static functions have been added, e.g. 'style_reg', 'style_imm', etc. Each of these functions formats a piece of text in a different style, 'register' and 'immediate' in this case. Here's an example taken from aarch64-opc.c of the new functions in use: snprintf (buf, size, "[%s, %s]!", style_reg (styler, base), style_imm (styler, "#%d", opnd->addr.offset.imm)); The aarch64_print_operand function is also called from the assembler to aid in printing diagnostic messages. Right now I have no plans to add styling to the assembler output, and so, the callback function used in the assembler ignores the styling information and just returns an plain string. I've used the source files in gas/testsuite/gas/aarch64/ for testing, and have manually gone through and checked that the styling looks reasonable, however, I'm not an AArch64 expert, so it is possible that the odd piece is styled incorrectly. Please point out any mistakes I've made. With objdump disassembler color turned off, there should be no change in the output after this commit.
2022-04-28 20:31:07 +08:00
const char *base,
struct aarch64_styler *styler)
{
if (opnd->addr.writeback)
{
if (opnd->addr.preind)
{
if (opnd->type == AARCH64_OPND_ADDR_SIMM10 && !opnd->addr.offset.imm)
libopcodes/aarch64: add support for disassembler styling This commit enables disassembler styling for AArch64. After this commit it is possible to have objdump style AArch64 disassembler output (using --disassembler-color option). Once the required GDB patches are merged, GDB will also style the disassembler output. The changes to support styling are mostly split between two files opcodes/aarch64-dis.c and opcodes/aarch64-opc.c. The entry point for the AArch64 disassembler can be found in aarch64-dis.c, this file handles printing the instruction mnemonics, and assembler directives (e.g. '.byte', '.word', etc). Some operands, mostly relating to assembler directives are also printed from this file. This commit changes all of this to pass through suitable styling information. However, for most "normal" instructions, the instruction operands are printed using a two step process. From aarch64-dis.c, in the print_operands function, the function aarch64_print_operand is called, this function is in aarch64-opc.c, and converts an instruction operand into a string. Then, back in print_operands (aarch64-dis.c), the operand string is printed. Unfortunately, the string returned by aarch64_print_operand can be quite complex, it will include syntax elements, like '[' and ']', in addition to register names and immediate values. In some cases, a single operand will expand into what will appear (to the user) as multiple operands separated with a ','. This makes the task of styling more complex, all these different components need to by styled differently, so we need to get the styling information out of aarch64_print_operand in some way. The solution that I propose here is similar to the solution that I used for the i386 disassembler. Currently, aarch64_print_operand uses snprintf to write the operand text into a buffer provided by the caller. What I propose is that we pass an extra argument to the aarch64_print_operand function, this argument will be a structure, the structure contains a callback function and some state. When aarch64_print_operand needs to format part of its output this can be done by using the callback function within the new structure, this callback returns a string with special embedded markers that indicate which mode should be used for each piece of text. Back in aarch64-dis.c we can spot these special style markers and use this to split the disassembler output up and apply the correct style to each piece. To make aarch64-opc.c clearer a series of new static functions have been added, e.g. 'style_reg', 'style_imm', etc. Each of these functions formats a piece of text in a different style, 'register' and 'immediate' in this case. Here's an example taken from aarch64-opc.c of the new functions in use: snprintf (buf, size, "[%s, %s]!", style_reg (styler, base), style_imm (styler, "#%d", opnd->addr.offset.imm)); The aarch64_print_operand function is also called from the assembler to aid in printing diagnostic messages. Right now I have no plans to add styling to the assembler output, and so, the callback function used in the assembler ignores the styling information and just returns an plain string. I've used the source files in gas/testsuite/gas/aarch64/ for testing, and have manually gone through and checked that the styling looks reasonable, however, I'm not an AArch64 expert, so it is possible that the odd piece is styled incorrectly. Please point out any mistakes I've made. With objdump disassembler color turned off, there should be no change in the output after this commit.
2022-04-28 20:31:07 +08:00
snprintf (buf, size, "[%s]!", style_reg (styler, base));
else
libopcodes/aarch64: add support for disassembler styling This commit enables disassembler styling for AArch64. After this commit it is possible to have objdump style AArch64 disassembler output (using --disassembler-color option). Once the required GDB patches are merged, GDB will also style the disassembler output. The changes to support styling are mostly split between two files opcodes/aarch64-dis.c and opcodes/aarch64-opc.c. The entry point for the AArch64 disassembler can be found in aarch64-dis.c, this file handles printing the instruction mnemonics, and assembler directives (e.g. '.byte', '.word', etc). Some operands, mostly relating to assembler directives are also printed from this file. This commit changes all of this to pass through suitable styling information. However, for most "normal" instructions, the instruction operands are printed using a two step process. From aarch64-dis.c, in the print_operands function, the function aarch64_print_operand is called, this function is in aarch64-opc.c, and converts an instruction operand into a string. Then, back in print_operands (aarch64-dis.c), the operand string is printed. Unfortunately, the string returned by aarch64_print_operand can be quite complex, it will include syntax elements, like '[' and ']', in addition to register names and immediate values. In some cases, a single operand will expand into what will appear (to the user) as multiple operands separated with a ','. This makes the task of styling more complex, all these different components need to by styled differently, so we need to get the styling information out of aarch64_print_operand in some way. The solution that I propose here is similar to the solution that I used for the i386 disassembler. Currently, aarch64_print_operand uses snprintf to write the operand text into a buffer provided by the caller. What I propose is that we pass an extra argument to the aarch64_print_operand function, this argument will be a structure, the structure contains a callback function and some state. When aarch64_print_operand needs to format part of its output this can be done by using the callback function within the new structure, this callback returns a string with special embedded markers that indicate which mode should be used for each piece of text. Back in aarch64-dis.c we can spot these special style markers and use this to split the disassembler output up and apply the correct style to each piece. To make aarch64-opc.c clearer a series of new static functions have been added, e.g. 'style_reg', 'style_imm', etc. Each of these functions formats a piece of text in a different style, 'register' and 'immediate' in this case. Here's an example taken from aarch64-opc.c of the new functions in use: snprintf (buf, size, "[%s, %s]!", style_reg (styler, base), style_imm (styler, "#%d", opnd->addr.offset.imm)); The aarch64_print_operand function is also called from the assembler to aid in printing diagnostic messages. Right now I have no plans to add styling to the assembler output, and so, the callback function used in the assembler ignores the styling information and just returns an plain string. I've used the source files in gas/testsuite/gas/aarch64/ for testing, and have manually gone through and checked that the styling looks reasonable, however, I'm not an AArch64 expert, so it is possible that the odd piece is styled incorrectly. Please point out any mistakes I've made. With objdump disassembler color turned off, there should be no change in the output after this commit.
2022-04-28 20:31:07 +08:00
snprintf (buf, size, "[%s, %s]!",
style_reg (styler, base),
style_imm (styler, "#%d", opnd->addr.offset.imm));
}
else
libopcodes/aarch64: add support for disassembler styling This commit enables disassembler styling for AArch64. After this commit it is possible to have objdump style AArch64 disassembler output (using --disassembler-color option). Once the required GDB patches are merged, GDB will also style the disassembler output. The changes to support styling are mostly split between two files opcodes/aarch64-dis.c and opcodes/aarch64-opc.c. The entry point for the AArch64 disassembler can be found in aarch64-dis.c, this file handles printing the instruction mnemonics, and assembler directives (e.g. '.byte', '.word', etc). Some operands, mostly relating to assembler directives are also printed from this file. This commit changes all of this to pass through suitable styling information. However, for most "normal" instructions, the instruction operands are printed using a two step process. From aarch64-dis.c, in the print_operands function, the function aarch64_print_operand is called, this function is in aarch64-opc.c, and converts an instruction operand into a string. Then, back in print_operands (aarch64-dis.c), the operand string is printed. Unfortunately, the string returned by aarch64_print_operand can be quite complex, it will include syntax elements, like '[' and ']', in addition to register names and immediate values. In some cases, a single operand will expand into what will appear (to the user) as multiple operands separated with a ','. This makes the task of styling more complex, all these different components need to by styled differently, so we need to get the styling information out of aarch64_print_operand in some way. The solution that I propose here is similar to the solution that I used for the i386 disassembler. Currently, aarch64_print_operand uses snprintf to write the operand text into a buffer provided by the caller. What I propose is that we pass an extra argument to the aarch64_print_operand function, this argument will be a structure, the structure contains a callback function and some state. When aarch64_print_operand needs to format part of its output this can be done by using the callback function within the new structure, this callback returns a string with special embedded markers that indicate which mode should be used for each piece of text. Back in aarch64-dis.c we can spot these special style markers and use this to split the disassembler output up and apply the correct style to each piece. To make aarch64-opc.c clearer a series of new static functions have been added, e.g. 'style_reg', 'style_imm', etc. Each of these functions formats a piece of text in a different style, 'register' and 'immediate' in this case. Here's an example taken from aarch64-opc.c of the new functions in use: snprintf (buf, size, "[%s, %s]!", style_reg (styler, base), style_imm (styler, "#%d", opnd->addr.offset.imm)); The aarch64_print_operand function is also called from the assembler to aid in printing diagnostic messages. Right now I have no plans to add styling to the assembler output, and so, the callback function used in the assembler ignores the styling information and just returns an plain string. I've used the source files in gas/testsuite/gas/aarch64/ for testing, and have manually gone through and checked that the styling looks reasonable, however, I'm not an AArch64 expert, so it is possible that the odd piece is styled incorrectly. Please point out any mistakes I've made. With objdump disassembler color turned off, there should be no change in the output after this commit.
2022-04-28 20:31:07 +08:00
snprintf (buf, size, "[%s], %s",
style_reg (styler, base),
style_imm (styler, "#%d", opnd->addr.offset.imm));
}
else
{
[AArch64][SVE 26/32] Add SVE MUL VL addressing modes This patch adds support for addresses of the form: [<base>, #<offset>, MUL VL] This involves adding a new AARCH64_MOD_MUL_VL modifier, which is why I split it out from the other addressing modes. For LD2, LD3 and LD4, the offset must be a multiple of the structure size, so for LD3 the possible values are 0, 3, 6, .... The patch therefore extends value_aligned_p to handle non-power-of-2 alignments. include/ * opcode/aarch64.h (AARCH64_OPND_SVE_ADDR_RI_S4xVL): New aarch64_opnd. (AARCH64_OPND_SVE_ADDR_RI_S4x2xVL, AARCH64_OPND_SVE_ADDR_RI_S4x3xVL) (AARCH64_OPND_SVE_ADDR_RI_S4x4xVL, AARCH64_OPND_SVE_ADDR_RI_S6xVL) (AARCH64_OPND_SVE_ADDR_RI_S9xVL): Likewise. (AARCH64_MOD_MUL_VL): New aarch64_modifier_kind. opcodes/ * aarch64-tbl.h (AARCH64_OPERANDS): Add entries for new MUL VL operands. * aarch64-opc.c (aarch64_operand_modifiers): Initialize the AARCH64_MOD_MUL_VL entry. (value_aligned_p): Cope with non-power-of-two alignments. (operand_general_constraint_met_p): Handle the new MUL VL addresses. (print_immediate_offset_address): Likewise. (aarch64_print_operand): Likewise. * aarch64-opc-2.c: Regenerate. * aarch64-asm.h (ins_sve_addr_ri_s4xvl, ins_sve_addr_ri_s6xvl) (ins_sve_addr_ri_s9xvl): New inserters. * aarch64-asm.c (aarch64_ins_sve_addr_ri_s4xvl): New function. (aarch64_ins_sve_addr_ri_s6xvl): Likewise. (aarch64_ins_sve_addr_ri_s9xvl): Likewise. * aarch64-asm-2.c: Regenerate. * aarch64-dis.h (ext_sve_addr_ri_s4xvl, ext_sve_addr_ri_s6xvl) (ext_sve_addr_ri_s9xvl): New extractors. * aarch64-dis.c (aarch64_ext_sve_addr_reg_mul_vl): New function. (aarch64_ext_sve_addr_ri_s4xvl): Likewise. (aarch64_ext_sve_addr_ri_s6xvl): Likewise. (aarch64_ext_sve_addr_ri_s9xvl): Likewise. * aarch64-dis-2.c: Regenerate. gas/ * config/tc-aarch64.c (SHIFTED_NONE, SHIFTED_MUL_VL): New parse_shift_modes. (parse_shift): Handle SHIFTED_MUL_VL. (parse_address_main): Add an imm_shift_mode parameter. (parse_address, parse_sve_address): Update accordingly. (parse_operands): Handle MUL VL addressing modes.
2016-09-21 23:56:15 +08:00
if (opnd->shifter.operator_present)
{
assert (opnd->shifter.kind == AARCH64_MOD_MUL_VL);
libopcodes/aarch64: add support for disassembler styling This commit enables disassembler styling for AArch64. After this commit it is possible to have objdump style AArch64 disassembler output (using --disassembler-color option). Once the required GDB patches are merged, GDB will also style the disassembler output. The changes to support styling are mostly split between two files opcodes/aarch64-dis.c and opcodes/aarch64-opc.c. The entry point for the AArch64 disassembler can be found in aarch64-dis.c, this file handles printing the instruction mnemonics, and assembler directives (e.g. '.byte', '.word', etc). Some operands, mostly relating to assembler directives are also printed from this file. This commit changes all of this to pass through suitable styling information. However, for most "normal" instructions, the instruction operands are printed using a two step process. From aarch64-dis.c, in the print_operands function, the function aarch64_print_operand is called, this function is in aarch64-opc.c, and converts an instruction operand into a string. Then, back in print_operands (aarch64-dis.c), the operand string is printed. Unfortunately, the string returned by aarch64_print_operand can be quite complex, it will include syntax elements, like '[' and ']', in addition to register names and immediate values. In some cases, a single operand will expand into what will appear (to the user) as multiple operands separated with a ','. This makes the task of styling more complex, all these different components need to by styled differently, so we need to get the styling information out of aarch64_print_operand in some way. The solution that I propose here is similar to the solution that I used for the i386 disassembler. Currently, aarch64_print_operand uses snprintf to write the operand text into a buffer provided by the caller. What I propose is that we pass an extra argument to the aarch64_print_operand function, this argument will be a structure, the structure contains a callback function and some state. When aarch64_print_operand needs to format part of its output this can be done by using the callback function within the new structure, this callback returns a string with special embedded markers that indicate which mode should be used for each piece of text. Back in aarch64-dis.c we can spot these special style markers and use this to split the disassembler output up and apply the correct style to each piece. To make aarch64-opc.c clearer a series of new static functions have been added, e.g. 'style_reg', 'style_imm', etc. Each of these functions formats a piece of text in a different style, 'register' and 'immediate' in this case. Here's an example taken from aarch64-opc.c of the new functions in use: snprintf (buf, size, "[%s, %s]!", style_reg (styler, base), style_imm (styler, "#%d", opnd->addr.offset.imm)); The aarch64_print_operand function is also called from the assembler to aid in printing diagnostic messages. Right now I have no plans to add styling to the assembler output, and so, the callback function used in the assembler ignores the styling information and just returns an plain string. I've used the source files in gas/testsuite/gas/aarch64/ for testing, and have manually gone through and checked that the styling looks reasonable, however, I'm not an AArch64 expert, so it is possible that the odd piece is styled incorrectly. Please point out any mistakes I've made. With objdump disassembler color turned off, there should be no change in the output after this commit.
2022-04-28 20:31:07 +08:00
snprintf (buf, size, "[%s, %s, %s]",
style_reg (styler, base),
style_imm (styler, "#%d", opnd->addr.offset.imm),
style_sub_mnem (styler, "mul vl"));
[AArch64][SVE 26/32] Add SVE MUL VL addressing modes This patch adds support for addresses of the form: [<base>, #<offset>, MUL VL] This involves adding a new AARCH64_MOD_MUL_VL modifier, which is why I split it out from the other addressing modes. For LD2, LD3 and LD4, the offset must be a multiple of the structure size, so for LD3 the possible values are 0, 3, 6, .... The patch therefore extends value_aligned_p to handle non-power-of-2 alignments. include/ * opcode/aarch64.h (AARCH64_OPND_SVE_ADDR_RI_S4xVL): New aarch64_opnd. (AARCH64_OPND_SVE_ADDR_RI_S4x2xVL, AARCH64_OPND_SVE_ADDR_RI_S4x3xVL) (AARCH64_OPND_SVE_ADDR_RI_S4x4xVL, AARCH64_OPND_SVE_ADDR_RI_S6xVL) (AARCH64_OPND_SVE_ADDR_RI_S9xVL): Likewise. (AARCH64_MOD_MUL_VL): New aarch64_modifier_kind. opcodes/ * aarch64-tbl.h (AARCH64_OPERANDS): Add entries for new MUL VL operands. * aarch64-opc.c (aarch64_operand_modifiers): Initialize the AARCH64_MOD_MUL_VL entry. (value_aligned_p): Cope with non-power-of-two alignments. (operand_general_constraint_met_p): Handle the new MUL VL addresses. (print_immediate_offset_address): Likewise. (aarch64_print_operand): Likewise. * aarch64-opc-2.c: Regenerate. * aarch64-asm.h (ins_sve_addr_ri_s4xvl, ins_sve_addr_ri_s6xvl) (ins_sve_addr_ri_s9xvl): New inserters. * aarch64-asm.c (aarch64_ins_sve_addr_ri_s4xvl): New function. (aarch64_ins_sve_addr_ri_s6xvl): Likewise. (aarch64_ins_sve_addr_ri_s9xvl): Likewise. * aarch64-asm-2.c: Regenerate. * aarch64-dis.h (ext_sve_addr_ri_s4xvl, ext_sve_addr_ri_s6xvl) (ext_sve_addr_ri_s9xvl): New extractors. * aarch64-dis.c (aarch64_ext_sve_addr_reg_mul_vl): New function. (aarch64_ext_sve_addr_ri_s4xvl): Likewise. (aarch64_ext_sve_addr_ri_s6xvl): Likewise. (aarch64_ext_sve_addr_ri_s9xvl): Likewise. * aarch64-dis-2.c: Regenerate. gas/ * config/tc-aarch64.c (SHIFTED_NONE, SHIFTED_MUL_VL): New parse_shift_modes. (parse_shift): Handle SHIFTED_MUL_VL. (parse_address_main): Add an imm_shift_mode parameter. (parse_address, parse_sve_address): Update accordingly. (parse_operands): Handle MUL VL addressing modes.
2016-09-21 23:56:15 +08:00
}
else if (opnd->addr.offset.imm)
libopcodes/aarch64: add support for disassembler styling This commit enables disassembler styling for AArch64. After this commit it is possible to have objdump style AArch64 disassembler output (using --disassembler-color option). Once the required GDB patches are merged, GDB will also style the disassembler output. The changes to support styling are mostly split between two files opcodes/aarch64-dis.c and opcodes/aarch64-opc.c. The entry point for the AArch64 disassembler can be found in aarch64-dis.c, this file handles printing the instruction mnemonics, and assembler directives (e.g. '.byte', '.word', etc). Some operands, mostly relating to assembler directives are also printed from this file. This commit changes all of this to pass through suitable styling information. However, for most "normal" instructions, the instruction operands are printed using a two step process. From aarch64-dis.c, in the print_operands function, the function aarch64_print_operand is called, this function is in aarch64-opc.c, and converts an instruction operand into a string. Then, back in print_operands (aarch64-dis.c), the operand string is printed. Unfortunately, the string returned by aarch64_print_operand can be quite complex, it will include syntax elements, like '[' and ']', in addition to register names and immediate values. In some cases, a single operand will expand into what will appear (to the user) as multiple operands separated with a ','. This makes the task of styling more complex, all these different components need to by styled differently, so we need to get the styling information out of aarch64_print_operand in some way. The solution that I propose here is similar to the solution that I used for the i386 disassembler. Currently, aarch64_print_operand uses snprintf to write the operand text into a buffer provided by the caller. What I propose is that we pass an extra argument to the aarch64_print_operand function, this argument will be a structure, the structure contains a callback function and some state. When aarch64_print_operand needs to format part of its output this can be done by using the callback function within the new structure, this callback returns a string with special embedded markers that indicate which mode should be used for each piece of text. Back in aarch64-dis.c we can spot these special style markers and use this to split the disassembler output up and apply the correct style to each piece. To make aarch64-opc.c clearer a series of new static functions have been added, e.g. 'style_reg', 'style_imm', etc. Each of these functions formats a piece of text in a different style, 'register' and 'immediate' in this case. Here's an example taken from aarch64-opc.c of the new functions in use: snprintf (buf, size, "[%s, %s]!", style_reg (styler, base), style_imm (styler, "#%d", opnd->addr.offset.imm)); The aarch64_print_operand function is also called from the assembler to aid in printing diagnostic messages. Right now I have no plans to add styling to the assembler output, and so, the callback function used in the assembler ignores the styling information and just returns an plain string. I've used the source files in gas/testsuite/gas/aarch64/ for testing, and have manually gone through and checked that the styling looks reasonable, however, I'm not an AArch64 expert, so it is possible that the odd piece is styled incorrectly. Please point out any mistakes I've made. With objdump disassembler color turned off, there should be no change in the output after this commit.
2022-04-28 20:31:07 +08:00
snprintf (buf, size, "[%s, %s]",
style_reg (styler, base),
style_imm (styler, "#%d", opnd->addr.offset.imm));
else
libopcodes/aarch64: add support for disassembler styling This commit enables disassembler styling for AArch64. After this commit it is possible to have objdump style AArch64 disassembler output (using --disassembler-color option). Once the required GDB patches are merged, GDB will also style the disassembler output. The changes to support styling are mostly split between two files opcodes/aarch64-dis.c and opcodes/aarch64-opc.c. The entry point for the AArch64 disassembler can be found in aarch64-dis.c, this file handles printing the instruction mnemonics, and assembler directives (e.g. '.byte', '.word', etc). Some operands, mostly relating to assembler directives are also printed from this file. This commit changes all of this to pass through suitable styling information. However, for most "normal" instructions, the instruction operands are printed using a two step process. From aarch64-dis.c, in the print_operands function, the function aarch64_print_operand is called, this function is in aarch64-opc.c, and converts an instruction operand into a string. Then, back in print_operands (aarch64-dis.c), the operand string is printed. Unfortunately, the string returned by aarch64_print_operand can be quite complex, it will include syntax elements, like '[' and ']', in addition to register names and immediate values. In some cases, a single operand will expand into what will appear (to the user) as multiple operands separated with a ','. This makes the task of styling more complex, all these different components need to by styled differently, so we need to get the styling information out of aarch64_print_operand in some way. The solution that I propose here is similar to the solution that I used for the i386 disassembler. Currently, aarch64_print_operand uses snprintf to write the operand text into a buffer provided by the caller. What I propose is that we pass an extra argument to the aarch64_print_operand function, this argument will be a structure, the structure contains a callback function and some state. When aarch64_print_operand needs to format part of its output this can be done by using the callback function within the new structure, this callback returns a string with special embedded markers that indicate which mode should be used for each piece of text. Back in aarch64-dis.c we can spot these special style markers and use this to split the disassembler output up and apply the correct style to each piece. To make aarch64-opc.c clearer a series of new static functions have been added, e.g. 'style_reg', 'style_imm', etc. Each of these functions formats a piece of text in a different style, 'register' and 'immediate' in this case. Here's an example taken from aarch64-opc.c of the new functions in use: snprintf (buf, size, "[%s, %s]!", style_reg (styler, base), style_imm (styler, "#%d", opnd->addr.offset.imm)); The aarch64_print_operand function is also called from the assembler to aid in printing diagnostic messages. Right now I have no plans to add styling to the assembler output, and so, the callback function used in the assembler ignores the styling information and just returns an plain string. I've used the source files in gas/testsuite/gas/aarch64/ for testing, and have manually gone through and checked that the styling looks reasonable, however, I'm not an AArch64 expert, so it is possible that the odd piece is styled incorrectly. Please point out any mistakes I've made. With objdump disassembler color turned off, there should be no change in the output after this commit.
2022-04-28 20:31:07 +08:00
snprintf (buf, size, "[%s]", style_reg (styler, base));
}
}
/* Produce the string representation of the register offset address operand
*OPND in the buffer pointed by BUF of size SIZE. BASE and OFFSET are
the names of the base and offset registers. */
static void
print_register_offset_address (char *buf, size_t size,
const aarch64_opnd_info *opnd,
libopcodes/aarch64: add support for disassembler styling This commit enables disassembler styling for AArch64. After this commit it is possible to have objdump style AArch64 disassembler output (using --disassembler-color option). Once the required GDB patches are merged, GDB will also style the disassembler output. The changes to support styling are mostly split between two files opcodes/aarch64-dis.c and opcodes/aarch64-opc.c. The entry point for the AArch64 disassembler can be found in aarch64-dis.c, this file handles printing the instruction mnemonics, and assembler directives (e.g. '.byte', '.word', etc). Some operands, mostly relating to assembler directives are also printed from this file. This commit changes all of this to pass through suitable styling information. However, for most "normal" instructions, the instruction operands are printed using a two step process. From aarch64-dis.c, in the print_operands function, the function aarch64_print_operand is called, this function is in aarch64-opc.c, and converts an instruction operand into a string. Then, back in print_operands (aarch64-dis.c), the operand string is printed. Unfortunately, the string returned by aarch64_print_operand can be quite complex, it will include syntax elements, like '[' and ']', in addition to register names and immediate values. In some cases, a single operand will expand into what will appear (to the user) as multiple operands separated with a ','. This makes the task of styling more complex, all these different components need to by styled differently, so we need to get the styling information out of aarch64_print_operand in some way. The solution that I propose here is similar to the solution that I used for the i386 disassembler. Currently, aarch64_print_operand uses snprintf to write the operand text into a buffer provided by the caller. What I propose is that we pass an extra argument to the aarch64_print_operand function, this argument will be a structure, the structure contains a callback function and some state. When aarch64_print_operand needs to format part of its output this can be done by using the callback function within the new structure, this callback returns a string with special embedded markers that indicate which mode should be used for each piece of text. Back in aarch64-dis.c we can spot these special style markers and use this to split the disassembler output up and apply the correct style to each piece. To make aarch64-opc.c clearer a series of new static functions have been added, e.g. 'style_reg', 'style_imm', etc. Each of these functions formats a piece of text in a different style, 'register' and 'immediate' in this case. Here's an example taken from aarch64-opc.c of the new functions in use: snprintf (buf, size, "[%s, %s]!", style_reg (styler, base), style_imm (styler, "#%d", opnd->addr.offset.imm)); The aarch64_print_operand function is also called from the assembler to aid in printing diagnostic messages. Right now I have no plans to add styling to the assembler output, and so, the callback function used in the assembler ignores the styling information and just returns an plain string. I've used the source files in gas/testsuite/gas/aarch64/ for testing, and have manually gone through and checked that the styling looks reasonable, however, I'm not an AArch64 expert, so it is possible that the odd piece is styled incorrectly. Please point out any mistakes I've made. With objdump disassembler color turned off, there should be no change in the output after this commit.
2022-04-28 20:31:07 +08:00
const char *base, const char *offset,
struct aarch64_styler *styler)
{
libopcodes/aarch64: add support for disassembler styling This commit enables disassembler styling for AArch64. After this commit it is possible to have objdump style AArch64 disassembler output (using --disassembler-color option). Once the required GDB patches are merged, GDB will also style the disassembler output. The changes to support styling are mostly split between two files opcodes/aarch64-dis.c and opcodes/aarch64-opc.c. The entry point for the AArch64 disassembler can be found in aarch64-dis.c, this file handles printing the instruction mnemonics, and assembler directives (e.g. '.byte', '.word', etc). Some operands, mostly relating to assembler directives are also printed from this file. This commit changes all of this to pass through suitable styling information. However, for most "normal" instructions, the instruction operands are printed using a two step process. From aarch64-dis.c, in the print_operands function, the function aarch64_print_operand is called, this function is in aarch64-opc.c, and converts an instruction operand into a string. Then, back in print_operands (aarch64-dis.c), the operand string is printed. Unfortunately, the string returned by aarch64_print_operand can be quite complex, it will include syntax elements, like '[' and ']', in addition to register names and immediate values. In some cases, a single operand will expand into what will appear (to the user) as multiple operands separated with a ','. This makes the task of styling more complex, all these different components need to by styled differently, so we need to get the styling information out of aarch64_print_operand in some way. The solution that I propose here is similar to the solution that I used for the i386 disassembler. Currently, aarch64_print_operand uses snprintf to write the operand text into a buffer provided by the caller. What I propose is that we pass an extra argument to the aarch64_print_operand function, this argument will be a structure, the structure contains a callback function and some state. When aarch64_print_operand needs to format part of its output this can be done by using the callback function within the new structure, this callback returns a string with special embedded markers that indicate which mode should be used for each piece of text. Back in aarch64-dis.c we can spot these special style markers and use this to split the disassembler output up and apply the correct style to each piece. To make aarch64-opc.c clearer a series of new static functions have been added, e.g. 'style_reg', 'style_imm', etc. Each of these functions formats a piece of text in a different style, 'register' and 'immediate' in this case. Here's an example taken from aarch64-opc.c of the new functions in use: snprintf (buf, size, "[%s, %s]!", style_reg (styler, base), style_imm (styler, "#%d", opnd->addr.offset.imm)); The aarch64_print_operand function is also called from the assembler to aid in printing diagnostic messages. Right now I have no plans to add styling to the assembler output, and so, the callback function used in the assembler ignores the styling information and just returns an plain string. I've used the source files in gas/testsuite/gas/aarch64/ for testing, and have manually gone through and checked that the styling looks reasonable, however, I'm not an AArch64 expert, so it is possible that the odd piece is styled incorrectly. Please point out any mistakes I've made. With objdump disassembler color turned off, there should be no change in the output after this commit.
2022-04-28 20:31:07 +08:00
char tb[32]; /* Temporary buffer. */
bool print_extend_p = true;
bool print_amount_p = true;
const char *shift_name = aarch64_operand_modifiers[opnd->shifter.kind].name;
if (!opnd->shifter.amount && (opnd->qualifier != AARCH64_OPND_QLF_S_B
|| !opnd->shifter.amount_present))
{
/* Not print the shift/extend amount when the amount is zero and
when it is not the special case of 8-bit load/store instruction. */
print_amount_p = false;
/* Likewise, no need to print the shift operator LSL in such a
situation. */
if (opnd->shifter.kind == AARCH64_MOD_LSL)
print_extend_p = false;
}
/* Prepare for the extend/shift. */
if (print_extend_p)
{
if (print_amount_p)
libopcodes/aarch64: add support for disassembler styling This commit enables disassembler styling for AArch64. After this commit it is possible to have objdump style AArch64 disassembler output (using --disassembler-color option). Once the required GDB patches are merged, GDB will also style the disassembler output. The changes to support styling are mostly split between two files opcodes/aarch64-dis.c and opcodes/aarch64-opc.c. The entry point for the AArch64 disassembler can be found in aarch64-dis.c, this file handles printing the instruction mnemonics, and assembler directives (e.g. '.byte', '.word', etc). Some operands, mostly relating to assembler directives are also printed from this file. This commit changes all of this to pass through suitable styling information. However, for most "normal" instructions, the instruction operands are printed using a two step process. From aarch64-dis.c, in the print_operands function, the function aarch64_print_operand is called, this function is in aarch64-opc.c, and converts an instruction operand into a string. Then, back in print_operands (aarch64-dis.c), the operand string is printed. Unfortunately, the string returned by aarch64_print_operand can be quite complex, it will include syntax elements, like '[' and ']', in addition to register names and immediate values. In some cases, a single operand will expand into what will appear (to the user) as multiple operands separated with a ','. This makes the task of styling more complex, all these different components need to by styled differently, so we need to get the styling information out of aarch64_print_operand in some way. The solution that I propose here is similar to the solution that I used for the i386 disassembler. Currently, aarch64_print_operand uses snprintf to write the operand text into a buffer provided by the caller. What I propose is that we pass an extra argument to the aarch64_print_operand function, this argument will be a structure, the structure contains a callback function and some state. When aarch64_print_operand needs to format part of its output this can be done by using the callback function within the new structure, this callback returns a string with special embedded markers that indicate which mode should be used for each piece of text. Back in aarch64-dis.c we can spot these special style markers and use this to split the disassembler output up and apply the correct style to each piece. To make aarch64-opc.c clearer a series of new static functions have been added, e.g. 'style_reg', 'style_imm', etc. Each of these functions formats a piece of text in a different style, 'register' and 'immediate' in this case. Here's an example taken from aarch64-opc.c of the new functions in use: snprintf (buf, size, "[%s, %s]!", style_reg (styler, base), style_imm (styler, "#%d", opnd->addr.offset.imm)); The aarch64_print_operand function is also called from the assembler to aid in printing diagnostic messages. Right now I have no plans to add styling to the assembler output, and so, the callback function used in the assembler ignores the styling information and just returns an plain string. I've used the source files in gas/testsuite/gas/aarch64/ for testing, and have manually gone through and checked that the styling looks reasonable, however, I'm not an AArch64 expert, so it is possible that the odd piece is styled incorrectly. Please point out any mistakes I've made. With objdump disassembler color turned off, there should be no change in the output after this commit.
2022-04-28 20:31:07 +08:00
snprintf (tb, sizeof (tb), ", %s %s",
style_sub_mnem (styler, shift_name),
style_imm (styler, "#%" PRIi64,
/* PR 21096: The %100 is to silence a warning about possible truncation. */
libopcodes/aarch64: add support for disassembler styling This commit enables disassembler styling for AArch64. After this commit it is possible to have objdump style AArch64 disassembler output (using --disassembler-color option). Once the required GDB patches are merged, GDB will also style the disassembler output. The changes to support styling are mostly split between two files opcodes/aarch64-dis.c and opcodes/aarch64-opc.c. The entry point for the AArch64 disassembler can be found in aarch64-dis.c, this file handles printing the instruction mnemonics, and assembler directives (e.g. '.byte', '.word', etc). Some operands, mostly relating to assembler directives are also printed from this file. This commit changes all of this to pass through suitable styling information. However, for most "normal" instructions, the instruction operands are printed using a two step process. From aarch64-dis.c, in the print_operands function, the function aarch64_print_operand is called, this function is in aarch64-opc.c, and converts an instruction operand into a string. Then, back in print_operands (aarch64-dis.c), the operand string is printed. Unfortunately, the string returned by aarch64_print_operand can be quite complex, it will include syntax elements, like '[' and ']', in addition to register names and immediate values. In some cases, a single operand will expand into what will appear (to the user) as multiple operands separated with a ','. This makes the task of styling more complex, all these different components need to by styled differently, so we need to get the styling information out of aarch64_print_operand in some way. The solution that I propose here is similar to the solution that I used for the i386 disassembler. Currently, aarch64_print_operand uses snprintf to write the operand text into a buffer provided by the caller. What I propose is that we pass an extra argument to the aarch64_print_operand function, this argument will be a structure, the structure contains a callback function and some state. When aarch64_print_operand needs to format part of its output this can be done by using the callback function within the new structure, this callback returns a string with special embedded markers that indicate which mode should be used for each piece of text. Back in aarch64-dis.c we can spot these special style markers and use this to split the disassembler output up and apply the correct style to each piece. To make aarch64-opc.c clearer a series of new static functions have been added, e.g. 'style_reg', 'style_imm', etc. Each of these functions formats a piece of text in a different style, 'register' and 'immediate' in this case. Here's an example taken from aarch64-opc.c of the new functions in use: snprintf (buf, size, "[%s, %s]!", style_reg (styler, base), style_imm (styler, "#%d", opnd->addr.offset.imm)); The aarch64_print_operand function is also called from the assembler to aid in printing diagnostic messages. Right now I have no plans to add styling to the assembler output, and so, the callback function used in the assembler ignores the styling information and just returns an plain string. I've used the source files in gas/testsuite/gas/aarch64/ for testing, and have manually gone through and checked that the styling looks reasonable, however, I'm not an AArch64 expert, so it is possible that the odd piece is styled incorrectly. Please point out any mistakes I've made. With objdump disassembler color turned off, there should be no change in the output after this commit.
2022-04-28 20:31:07 +08:00
(opnd->shifter.amount % 100)));
else
libopcodes/aarch64: add support for disassembler styling This commit enables disassembler styling for AArch64. After this commit it is possible to have objdump style AArch64 disassembler output (using --disassembler-color option). Once the required GDB patches are merged, GDB will also style the disassembler output. The changes to support styling are mostly split between two files opcodes/aarch64-dis.c and opcodes/aarch64-opc.c. The entry point for the AArch64 disassembler can be found in aarch64-dis.c, this file handles printing the instruction mnemonics, and assembler directives (e.g. '.byte', '.word', etc). Some operands, mostly relating to assembler directives are also printed from this file. This commit changes all of this to pass through suitable styling information. However, for most "normal" instructions, the instruction operands are printed using a two step process. From aarch64-dis.c, in the print_operands function, the function aarch64_print_operand is called, this function is in aarch64-opc.c, and converts an instruction operand into a string. Then, back in print_operands (aarch64-dis.c), the operand string is printed. Unfortunately, the string returned by aarch64_print_operand can be quite complex, it will include syntax elements, like '[' and ']', in addition to register names and immediate values. In some cases, a single operand will expand into what will appear (to the user) as multiple operands separated with a ','. This makes the task of styling more complex, all these different components need to by styled differently, so we need to get the styling information out of aarch64_print_operand in some way. The solution that I propose here is similar to the solution that I used for the i386 disassembler. Currently, aarch64_print_operand uses snprintf to write the operand text into a buffer provided by the caller. What I propose is that we pass an extra argument to the aarch64_print_operand function, this argument will be a structure, the structure contains a callback function and some state. When aarch64_print_operand needs to format part of its output this can be done by using the callback function within the new structure, this callback returns a string with special embedded markers that indicate which mode should be used for each piece of text. Back in aarch64-dis.c we can spot these special style markers and use this to split the disassembler output up and apply the correct style to each piece. To make aarch64-opc.c clearer a series of new static functions have been added, e.g. 'style_reg', 'style_imm', etc. Each of these functions formats a piece of text in a different style, 'register' and 'immediate' in this case. Here's an example taken from aarch64-opc.c of the new functions in use: snprintf (buf, size, "[%s, %s]!", style_reg (styler, base), style_imm (styler, "#%d", opnd->addr.offset.imm)); The aarch64_print_operand function is also called from the assembler to aid in printing diagnostic messages. Right now I have no plans to add styling to the assembler output, and so, the callback function used in the assembler ignores the styling information and just returns an plain string. I've used the source files in gas/testsuite/gas/aarch64/ for testing, and have manually gone through and checked that the styling looks reasonable, however, I'm not an AArch64 expert, so it is possible that the odd piece is styled incorrectly. Please point out any mistakes I've made. With objdump disassembler color turned off, there should be no change in the output after this commit.
2022-04-28 20:31:07 +08:00
snprintf (tb, sizeof (tb), ", %s",
style_sub_mnem (styler, shift_name));
}
else
tb[0] = '\0';
libopcodes/aarch64: add support for disassembler styling This commit enables disassembler styling for AArch64. After this commit it is possible to have objdump style AArch64 disassembler output (using --disassembler-color option). Once the required GDB patches are merged, GDB will also style the disassembler output. The changes to support styling are mostly split between two files opcodes/aarch64-dis.c and opcodes/aarch64-opc.c. The entry point for the AArch64 disassembler can be found in aarch64-dis.c, this file handles printing the instruction mnemonics, and assembler directives (e.g. '.byte', '.word', etc). Some operands, mostly relating to assembler directives are also printed from this file. This commit changes all of this to pass through suitable styling information. However, for most "normal" instructions, the instruction operands are printed using a two step process. From aarch64-dis.c, in the print_operands function, the function aarch64_print_operand is called, this function is in aarch64-opc.c, and converts an instruction operand into a string. Then, back in print_operands (aarch64-dis.c), the operand string is printed. Unfortunately, the string returned by aarch64_print_operand can be quite complex, it will include syntax elements, like '[' and ']', in addition to register names and immediate values. In some cases, a single operand will expand into what will appear (to the user) as multiple operands separated with a ','. This makes the task of styling more complex, all these different components need to by styled differently, so we need to get the styling information out of aarch64_print_operand in some way. The solution that I propose here is similar to the solution that I used for the i386 disassembler. Currently, aarch64_print_operand uses snprintf to write the operand text into a buffer provided by the caller. What I propose is that we pass an extra argument to the aarch64_print_operand function, this argument will be a structure, the structure contains a callback function and some state. When aarch64_print_operand needs to format part of its output this can be done by using the callback function within the new structure, this callback returns a string with special embedded markers that indicate which mode should be used for each piece of text. Back in aarch64-dis.c we can spot these special style markers and use this to split the disassembler output up and apply the correct style to each piece. To make aarch64-opc.c clearer a series of new static functions have been added, e.g. 'style_reg', 'style_imm', etc. Each of these functions formats a piece of text in a different style, 'register' and 'immediate' in this case. Here's an example taken from aarch64-opc.c of the new functions in use: snprintf (buf, size, "[%s, %s]!", style_reg (styler, base), style_imm (styler, "#%d", opnd->addr.offset.imm)); The aarch64_print_operand function is also called from the assembler to aid in printing diagnostic messages. Right now I have no plans to add styling to the assembler output, and so, the callback function used in the assembler ignores the styling information and just returns an plain string. I've used the source files in gas/testsuite/gas/aarch64/ for testing, and have manually gone through and checked that the styling looks reasonable, however, I'm not an AArch64 expert, so it is possible that the odd piece is styled incorrectly. Please point out any mistakes I've made. With objdump disassembler color turned off, there should be no change in the output after this commit.
2022-04-28 20:31:07 +08:00
snprintf (buf, size, "[%s, %s%s]", style_reg (styler, base),
style_reg (styler, offset), tb);
}
/* Print ZA tiles from imm8 in ZERO instruction.
The preferred disassembly of this instruction uses the shortest list of tile
names that represent the encoded immediate mask.
For example:
* An all-ones immediate is disassembled as {ZA}.
* An all-zeros immediate is disassembled as an empty list { }.
*/
static void
libopcodes/aarch64: add support for disassembler styling This commit enables disassembler styling for AArch64. After this commit it is possible to have objdump style AArch64 disassembler output (using --disassembler-color option). Once the required GDB patches are merged, GDB will also style the disassembler output. The changes to support styling are mostly split between two files opcodes/aarch64-dis.c and opcodes/aarch64-opc.c. The entry point for the AArch64 disassembler can be found in aarch64-dis.c, this file handles printing the instruction mnemonics, and assembler directives (e.g. '.byte', '.word', etc). Some operands, mostly relating to assembler directives are also printed from this file. This commit changes all of this to pass through suitable styling information. However, for most "normal" instructions, the instruction operands are printed using a two step process. From aarch64-dis.c, in the print_operands function, the function aarch64_print_operand is called, this function is in aarch64-opc.c, and converts an instruction operand into a string. Then, back in print_operands (aarch64-dis.c), the operand string is printed. Unfortunately, the string returned by aarch64_print_operand can be quite complex, it will include syntax elements, like '[' and ']', in addition to register names and immediate values. In some cases, a single operand will expand into what will appear (to the user) as multiple operands separated with a ','. This makes the task of styling more complex, all these different components need to by styled differently, so we need to get the styling information out of aarch64_print_operand in some way. The solution that I propose here is similar to the solution that I used for the i386 disassembler. Currently, aarch64_print_operand uses snprintf to write the operand text into a buffer provided by the caller. What I propose is that we pass an extra argument to the aarch64_print_operand function, this argument will be a structure, the structure contains a callback function and some state. When aarch64_print_operand needs to format part of its output this can be done by using the callback function within the new structure, this callback returns a string with special embedded markers that indicate which mode should be used for each piece of text. Back in aarch64-dis.c we can spot these special style markers and use this to split the disassembler output up and apply the correct style to each piece. To make aarch64-opc.c clearer a series of new static functions have been added, e.g. 'style_reg', 'style_imm', etc. Each of these functions formats a piece of text in a different style, 'register' and 'immediate' in this case. Here's an example taken from aarch64-opc.c of the new functions in use: snprintf (buf, size, "[%s, %s]!", style_reg (styler, base), style_imm (styler, "#%d", opnd->addr.offset.imm)); The aarch64_print_operand function is also called from the assembler to aid in printing diagnostic messages. Right now I have no plans to add styling to the assembler output, and so, the callback function used in the assembler ignores the styling information and just returns an plain string. I've used the source files in gas/testsuite/gas/aarch64/ for testing, and have manually gone through and checked that the styling looks reasonable, however, I'm not an AArch64 expert, so it is possible that the odd piece is styled incorrectly. Please point out any mistakes I've made. With objdump disassembler color turned off, there should be no change in the output after this commit.
2022-04-28 20:31:07 +08:00
print_sme_za_list (char *buf, size_t size, int mask,
struct aarch64_styler *styler)
{
const char* zan[] = { "za", "za0.h", "za1.h", "za0.s",
"za1.s", "za2.s", "za3.s", "za0.d",
"za1.d", "za2.d", "za3.d", "za4.d",
"za5.d", "za6.d", "za7.d", " " };
const int zan_v[] = { 0xff, 0x55, 0xaa, 0x11,
0x22, 0x44, 0x88, 0x01,
0x02, 0x04, 0x08, 0x10,
0x20, 0x40, 0x80, 0x00 };
int i, k;
const int ZAN_SIZE = sizeof(zan) / sizeof(zan[0]);
k = snprintf (buf, size, "{");
for (i = 0; i < ZAN_SIZE; i++)
{
if ((mask & zan_v[i]) == zan_v[i])
{
mask &= ~zan_v[i];
if (k > 1)
libopcodes/aarch64: add support for disassembler styling This commit enables disassembler styling for AArch64. After this commit it is possible to have objdump style AArch64 disassembler output (using --disassembler-color option). Once the required GDB patches are merged, GDB will also style the disassembler output. The changes to support styling are mostly split between two files opcodes/aarch64-dis.c and opcodes/aarch64-opc.c. The entry point for the AArch64 disassembler can be found in aarch64-dis.c, this file handles printing the instruction mnemonics, and assembler directives (e.g. '.byte', '.word', etc). Some operands, mostly relating to assembler directives are also printed from this file. This commit changes all of this to pass through suitable styling information. However, for most "normal" instructions, the instruction operands are printed using a two step process. From aarch64-dis.c, in the print_operands function, the function aarch64_print_operand is called, this function is in aarch64-opc.c, and converts an instruction operand into a string. Then, back in print_operands (aarch64-dis.c), the operand string is printed. Unfortunately, the string returned by aarch64_print_operand can be quite complex, it will include syntax elements, like '[' and ']', in addition to register names and immediate values. In some cases, a single operand will expand into what will appear (to the user) as multiple operands separated with a ','. This makes the task of styling more complex, all these different components need to by styled differently, so we need to get the styling information out of aarch64_print_operand in some way. The solution that I propose here is similar to the solution that I used for the i386 disassembler. Currently, aarch64_print_operand uses snprintf to write the operand text into a buffer provided by the caller. What I propose is that we pass an extra argument to the aarch64_print_operand function, this argument will be a structure, the structure contains a callback function and some state. When aarch64_print_operand needs to format part of its output this can be done by using the callback function within the new structure, this callback returns a string with special embedded markers that indicate which mode should be used for each piece of text. Back in aarch64-dis.c we can spot these special style markers and use this to split the disassembler output up and apply the correct style to each piece. To make aarch64-opc.c clearer a series of new static functions have been added, e.g. 'style_reg', 'style_imm', etc. Each of these functions formats a piece of text in a different style, 'register' and 'immediate' in this case. Here's an example taken from aarch64-opc.c of the new functions in use: snprintf (buf, size, "[%s, %s]!", style_reg (styler, base), style_imm (styler, "#%d", opnd->addr.offset.imm)); The aarch64_print_operand function is also called from the assembler to aid in printing diagnostic messages. Right now I have no plans to add styling to the assembler output, and so, the callback function used in the assembler ignores the styling information and just returns an plain string. I've used the source files in gas/testsuite/gas/aarch64/ for testing, and have manually gone through and checked that the styling looks reasonable, however, I'm not an AArch64 expert, so it is possible that the odd piece is styled incorrectly. Please point out any mistakes I've made. With objdump disassembler color turned off, there should be no change in the output after this commit.
2022-04-28 20:31:07 +08:00
k += snprintf (buf + k, size - k, ", ");
k += snprintf (buf + k, size - k, "%s", style_reg (styler, zan[i]));
}
if (mask == 0)
break;
}
snprintf (buf + k, size - k, "}");
}
/* Generate the string representation of the operand OPNDS[IDX] for OPCODE
in *BUF. The caller should pass in the maximum size of *BUF in SIZE.
PC, PCREL_P and ADDRESS are used to pass in and return information about
the PC-relative address calculation, where the PC value is passed in
PC. If the operand is pc-relative related, *PCREL_P (if PCREL_P non-NULL)
will return 1 and *ADDRESS (if ADDRESS non-NULL) will return the
calculated address; otherwise, *PCREL_P (if PCREL_P non-NULL) returns 0.
The function serves both the disassembler and the assembler diagnostics
issuer, which is the reason why it lives in this file. */
void
aarch64_print_operand (char *buf, size_t size, bfd_vma pc,
const aarch64_opcode *opcode,
const aarch64_opnd_info *opnds, int idx, int *pcrel_p,
aarch64: Add support for Armv8-R system registers This patch adds support for the system registers introduced in Armv8-R AArch64. gas/ChangeLog: 2020-09-08 Alex Coplan <alex.coplan@arm.com> * config/tc-aarch64.c (parse_sys_reg): Also pass sysreg name to validation function. (parse_sys_ins_reg): Likewise. (print_operands): Pass CPU features to aarch64_print_operand(). * testsuite/gas/aarch64/v8-r-bad-sysregs.d: New test. * testsuite/gas/aarch64/v8-r-bad-sysregs.l: Error output. * testsuite/gas/aarch64/v8-r-bad-sysregs.s: Input. * testsuite/gas/aarch64/v8-r-sysregs-need-arch.d: New test. * testsuite/gas/aarch64/v8-r-sysregs-need-arch.l: Error output. * testsuite/gas/aarch64/v8-r-sysregs.d: New test. * testsuite/gas/aarch64/v8-r-sysregs.s: Input for previous two tests. include/ChangeLog: 2020-09-08 Alex Coplan <alex.coplan@arm.com> * opcode/aarch64.h (aarch64_sys_ins_reg_supported_p): Also take system register name in order to simplify validation for v8-R. (aarch64_print_operand): Also take CPU feature set, as disassembly for system registers now depends on arch variant. opcodes/ChangeLog: 2020-09-08 Alex Coplan <alex.coplan@arm.com> * aarch64-dis.c (print_operands): Pass CPU features to aarch64_print_operand(). * aarch64-opc.c (aarch64_print_operand): Use CPU features to determine preferred disassembly of system registers. (SR_RNG): Refactor to use new SR_FEAT2 macro. (SR_FEAT2): New. (SR_V8_1_A): New. (SR_V8_4_A): New. (SR_V8_A): New. (SR_V8_R): New. (SR_EXPAND_ELx): New. (SR_EXPAND_EL12): New. (aarch64_sys_regs): Specify which registers are only on A-profile, add R-profile system registers. (ENC_BARLAR): New. (PRBARn_ELx): New. (PRLARn_ELx): New. (aarch64_sys_ins_reg_supported_p): Reject EL3 registers for Armv8-R AArch64.
2020-09-08 21:21:44 +08:00
bfd_vma *address, char** notes,
opcodes/aarch64: split off creation of comment text in disassembler The function aarch64_print_operand (aarch64-opc.c) is responsible for converting an instruction operand into the textual representation of that operand. In some cases, a comment is included in the operand representation, though this (currently) only happens for the last operand of the instruction. In a future commit I would like to enable the new libopcodes styling for AArch64, this will allow objdump and GDB[1] to syntax highlight the disassembler output, however, having operands and comments combined in a single string like this makes such styling harder. In this commit, I propose to extend aarch64_print_operand to take a second buffer. Any comments for the instruction are written into this extra buffer. The two callers of aarch64_print_operand are then updated to pass an extra buffer, and print any resulting comment. In this commit no styling is added, that will come later. However, I have adjusted the output slightly. Before this commit some comments would be separated from the instruction operands with a tab character, while in other cases the comment was separated with two single spaces. After this commit I use a single tab character in all cases. This means a few test cases needed updated. If people would prefer me to move everyone to use the two spaces, then just let me know. Or maybe there was a good reason why we used a mix of styles, I could probably figure out a way to maintain the old output exactly if that is critical. Other than that, there should be no user visible changes after this commit. [1] GDB patches have not been merged yet, but have been posted to the GDB mailing list: https://sourceware.org/pipermail/gdb-patches/2022-June/190142.html
2022-06-16 20:46:41 +08:00
char *comment, size_t comment_size,
libopcodes/aarch64: add support for disassembler styling This commit enables disassembler styling for AArch64. After this commit it is possible to have objdump style AArch64 disassembler output (using --disassembler-color option). Once the required GDB patches are merged, GDB will also style the disassembler output. The changes to support styling are mostly split between two files opcodes/aarch64-dis.c and opcodes/aarch64-opc.c. The entry point for the AArch64 disassembler can be found in aarch64-dis.c, this file handles printing the instruction mnemonics, and assembler directives (e.g. '.byte', '.word', etc). Some operands, mostly relating to assembler directives are also printed from this file. This commit changes all of this to pass through suitable styling information. However, for most "normal" instructions, the instruction operands are printed using a two step process. From aarch64-dis.c, in the print_operands function, the function aarch64_print_operand is called, this function is in aarch64-opc.c, and converts an instruction operand into a string. Then, back in print_operands (aarch64-dis.c), the operand string is printed. Unfortunately, the string returned by aarch64_print_operand can be quite complex, it will include syntax elements, like '[' and ']', in addition to register names and immediate values. In some cases, a single operand will expand into what will appear (to the user) as multiple operands separated with a ','. This makes the task of styling more complex, all these different components need to by styled differently, so we need to get the styling information out of aarch64_print_operand in some way. The solution that I propose here is similar to the solution that I used for the i386 disassembler. Currently, aarch64_print_operand uses snprintf to write the operand text into a buffer provided by the caller. What I propose is that we pass an extra argument to the aarch64_print_operand function, this argument will be a structure, the structure contains a callback function and some state. When aarch64_print_operand needs to format part of its output this can be done by using the callback function within the new structure, this callback returns a string with special embedded markers that indicate which mode should be used for each piece of text. Back in aarch64-dis.c we can spot these special style markers and use this to split the disassembler output up and apply the correct style to each piece. To make aarch64-opc.c clearer a series of new static functions have been added, e.g. 'style_reg', 'style_imm', etc. Each of these functions formats a piece of text in a different style, 'register' and 'immediate' in this case. Here's an example taken from aarch64-opc.c of the new functions in use: snprintf (buf, size, "[%s, %s]!", style_reg (styler, base), style_imm (styler, "#%d", opnd->addr.offset.imm)); The aarch64_print_operand function is also called from the assembler to aid in printing diagnostic messages. Right now I have no plans to add styling to the assembler output, and so, the callback function used in the assembler ignores the styling information and just returns an plain string. I've used the source files in gas/testsuite/gas/aarch64/ for testing, and have manually gone through and checked that the styling looks reasonable, however, I'm not an AArch64 expert, so it is possible that the odd piece is styled incorrectly. Please point out any mistakes I've made. With objdump disassembler color turned off, there should be no change in the output after this commit.
2022-04-28 20:31:07 +08:00
aarch64_feature_set features,
struct aarch64_styler *styler)
{
[AArch64] Add SVE condition codes SVE defines new names for existing NZCV conditions, to reflect the result of instructions like PTEST. This patch adds support for these names. The patch also adds comments to the disassembly output to show the alternative names of a condition code. For example: cinv x0, x1, cc becomes: cinv x0, x1, cc // cc = lo, ul, last and: b.cc f0 <...> becomes: b.cc f0 <...> // b.lo, b.ul, b.last Doing this for the SVE names follows the practice recommended by the SVE specification and is definitely useful when reading SVE code. If the feeling is that it's too distracting elsewhere, we could add an option to turn it off. include/ * opcode/aarch64.h (aarch64_cond): Bump array size to 4. opcodes/ * aarch64-dis.c (remove_dot_suffix): New function, split out from... (print_mnemonic_name): ...here. (print_comment): New function. (print_aarch64_insn): Call it. * aarch64-opc.c (aarch64_conds): Add SVE names. (aarch64_print_operand): Print alternative condition names in a comment. gas/ * config/tc-aarch64.c (opcode_lookup): Search for the end of a condition name, rather than assuming that it will have exactly 2 characters. (parse_operands): Likewise. * testsuite/gas/aarch64/alias.d: Add new condition-code comments to the expected output. * testsuite/gas/aarch64/beq_1.d: Likewise. * testsuite/gas/aarch64/float-fp16.d: Likewise. * testsuite/gas/aarch64/int-insns.d: Likewise. * testsuite/gas/aarch64/no-aliases.d: Likewise. * testsuite/gas/aarch64/programmer-friendly.d: Likewise. * testsuite/gas/aarch64/reloc-insn.d: Likewise. * testsuite/gas/aarch64/b_c_1.d, testsuite/gas/aarch64/b_c_1.s: New test. ld/ * testsuite/ld-aarch64/emit-relocs-280.d: Match branch comments. * testsuite/ld-aarch64/weak-undefined.d: Likewise.
2016-09-22 00:09:59 +08:00
unsigned int i, num_conds;
const char *name = NULL;
const aarch64_opnd_info *opnd = opnds + idx;
enum aarch64_modifier_kind kind;
[AArch64][SVE 23/32] Add SVE pattern and prfop operands The SVE instructions have two enumerated operands: one to select a vector pattern and another to select a prefetch operation. The latter is a cut-down version of the base AArch64 prefetch operation. Both types of operand can also be specified as raw enum values such as #31. Reserved values can only be specified this way. If it hadn't been for the pattern operand, I would have been tempted to use the existing parsing for prefetch operations and add extra checks for SVE. However, since the patterns needed new enum parsing code anyway, it seeemed cleaner to reuse it for the prefetches too. Because of the small number of enum values, I don't think we'd gain anything by using hash tables. include/ * opcode/aarch64.h (AARCH64_OPND_SVE_PATTERN): New aarch64_opnd. (AARCH64_OPND_SVE_PRFOP): Likewise. (aarch64_sve_pattern_array): Declare. (aarch64_sve_prfop_array): Likewise. opcodes/ * aarch64-tbl.h (AARCH64_OPERANDS): Add entries for AARCH64_OPND_SVE_PATTERN and AARCH64_OPND_SVE_PRFOP. * aarch64-opc.h (FLD_SVE_pattern): New aarch64_field_kind. (FLD_SVE_prfop): Likewise. * aarch64-opc.c: Include libiberty.h. (aarch64_sve_pattern_array): New variable. (aarch64_sve_prfop_array): Likewise. (fields): Add entries for FLD_SVE_pattern and FLD_SVE_prfop. (aarch64_print_operand): Handle AARCH64_OPND_SVE_PATTERN and AARCH64_OPND_SVE_PRFOP. * aarch64-asm-2.c: Regenerate. * aarch64-dis-2.c: Likewise. * aarch64-opc-2.c: Likewise. gas/ * config/tc-aarch64.c (parse_enum_string): New function. (po_enum_or_fail): New macro. (parse_operands): Handle AARCH64_OPND_SVE_PATTERN and AARCH64_OPND_SVE_PRFOP.
2016-09-21 23:54:53 +08:00
uint64_t addr, enum_value;
opcodes/aarch64: split off creation of comment text in disassembler The function aarch64_print_operand (aarch64-opc.c) is responsible for converting an instruction operand into the textual representation of that operand. In some cases, a comment is included in the operand representation, though this (currently) only happens for the last operand of the instruction. In a future commit I would like to enable the new libopcodes styling for AArch64, this will allow objdump and GDB[1] to syntax highlight the disassembler output, however, having operands and comments combined in a single string like this makes such styling harder. In this commit, I propose to extend aarch64_print_operand to take a second buffer. Any comments for the instruction are written into this extra buffer. The two callers of aarch64_print_operand are then updated to pass an extra buffer, and print any resulting comment. In this commit no styling is added, that will come later. However, I have adjusted the output slightly. Before this commit some comments would be separated from the instruction operands with a tab character, while in other cases the comment was separated with two single spaces. After this commit I use a single tab character in all cases. This means a few test cases needed updated. If people would prefer me to move everyone to use the two spaces, then just let me know. Or maybe there was a good reason why we used a mix of styles, I could probably figure out a way to maintain the old output exactly if that is critical. Other than that, there should be no user visible changes after this commit. [1] GDB patches have not been merged yet, but have been posted to the GDB mailing list: https://sourceware.org/pipermail/gdb-patches/2022-June/190142.html
2022-06-16 20:46:41 +08:00
if (comment != NULL)
{
assert (comment_size > 0);
comment[0] = '\0';
}
else
assert (comment_size == 0);
buf[0] = '\0';
if (pcrel_p)
*pcrel_p = 0;
switch (opnd->type)
{
case AARCH64_OPND_Rd:
case AARCH64_OPND_Rn:
case AARCH64_OPND_Rm:
case AARCH64_OPND_Rt:
case AARCH64_OPND_Rt2:
case AARCH64_OPND_Rs:
case AARCH64_OPND_Ra:
case AARCH64_OPND_Rt_LS64:
case AARCH64_OPND_Rt_SYS:
case AARCH64_OPND_PAIRREG:
case AARCH64_OPND_PAIRREG_OR_XZR:
case AARCH64_OPND_SVE_Rm:
case AARCH64_OPND_LSE128_Rt:
case AARCH64_OPND_LSE128_Rt2:
/* The optional-ness of <Xt> in e.g. IC <ic_op>{, <Xt>} is determined by
2017-07-18 23:58:14 +08:00
the <ic_op>, therefore we use opnd->present to override the
generic optional-ness information. */
if (opnd->type == AARCH64_OPND_Rt_SYS)
{
if (!opnd->present)
break;
}
/* Omit the operand, e.g. RET. */
else if (optional_operand_p (opcode, idx)
&& (opnd->reg.regno
== get_optional_operand_default_value (opcode)))
break;
assert (opnd->qualifier == AARCH64_OPND_QLF_W
|| opnd->qualifier == AARCH64_OPND_QLF_X);
snprintf (buf, size, "%s",
libopcodes/aarch64: add support for disassembler styling This commit enables disassembler styling for AArch64. After this commit it is possible to have objdump style AArch64 disassembler output (using --disassembler-color option). Once the required GDB patches are merged, GDB will also style the disassembler output. The changes to support styling are mostly split between two files opcodes/aarch64-dis.c and opcodes/aarch64-opc.c. The entry point for the AArch64 disassembler can be found in aarch64-dis.c, this file handles printing the instruction mnemonics, and assembler directives (e.g. '.byte', '.word', etc). Some operands, mostly relating to assembler directives are also printed from this file. This commit changes all of this to pass through suitable styling information. However, for most "normal" instructions, the instruction operands are printed using a two step process. From aarch64-dis.c, in the print_operands function, the function aarch64_print_operand is called, this function is in aarch64-opc.c, and converts an instruction operand into a string. Then, back in print_operands (aarch64-dis.c), the operand string is printed. Unfortunately, the string returned by aarch64_print_operand can be quite complex, it will include syntax elements, like '[' and ']', in addition to register names and immediate values. In some cases, a single operand will expand into what will appear (to the user) as multiple operands separated with a ','. This makes the task of styling more complex, all these different components need to by styled differently, so we need to get the styling information out of aarch64_print_operand in some way. The solution that I propose here is similar to the solution that I used for the i386 disassembler. Currently, aarch64_print_operand uses snprintf to write the operand text into a buffer provided by the caller. What I propose is that we pass an extra argument to the aarch64_print_operand function, this argument will be a structure, the structure contains a callback function and some state. When aarch64_print_operand needs to format part of its output this can be done by using the callback function within the new structure, this callback returns a string with special embedded markers that indicate which mode should be used for each piece of text. Back in aarch64-dis.c we can spot these special style markers and use this to split the disassembler output up and apply the correct style to each piece. To make aarch64-opc.c clearer a series of new static functions have been added, e.g. 'style_reg', 'style_imm', etc. Each of these functions formats a piece of text in a different style, 'register' and 'immediate' in this case. Here's an example taken from aarch64-opc.c of the new functions in use: snprintf (buf, size, "[%s, %s]!", style_reg (styler, base), style_imm (styler, "#%d", opnd->addr.offset.imm)); The aarch64_print_operand function is also called from the assembler to aid in printing diagnostic messages. Right now I have no plans to add styling to the assembler output, and so, the callback function used in the assembler ignores the styling information and just returns an plain string. I've used the source files in gas/testsuite/gas/aarch64/ for testing, and have manually gone through and checked that the styling looks reasonable, however, I'm not an AArch64 expert, so it is possible that the odd piece is styled incorrectly. Please point out any mistakes I've made. With objdump disassembler color turned off, there should be no change in the output after this commit.
2022-04-28 20:31:07 +08:00
style_reg (styler, get_int_reg_name (opnd->reg.regno,
opnd->qualifier, 0)));
break;
case AARCH64_OPND_Rd_SP:
case AARCH64_OPND_Rn_SP:
[BINUTILS, AArch64, 2/2] Update Store Allocation Tag instructions This patch updates the Store allocation tags instructions in Armv8.5-A Memory Tagging Extension. This is part of the changes that have been introduced recently in the 00bet10 release All of these instructions have an updated register operand (Xt -> <Xt|SP>) - STG <Xt|SP>, [<Xn|SP>, #<simm>] - STG <Xt|SP>, [<Xn|SP>, #<simm>]! - STG <Xt|SP>, [<Xn|SP>], #<simm> - STZG <Xt|SP>, [<Xn|SP>, #<simm>] - STZG <Xt|SP>, [<Xn|SP>, #<simm>]! - STZG <Xt|SP>, [<Xn|SP>], #<simm> - ST2G <Xt|SP>, [<Xn|SP>, #<simm>] - ST2G <Xt|SP>, [<Xn|SP>, #<simm>]! - ST2G <Xt|SP>, [<Xn|SP>], #<simm> - STZ2G <Xt|SP>, [<Xn|SP>, #<simm>] - STZ2G <Xt|SP>, [<Xn|SP>, #<simm>]! - STZ2G <Xt|SP>, [<Xn|SP>], #<simm> In order to accept <Rt|SP> a new operand type Rt_SP is introduced which has the same field as FLD_Rt but follows other semantics of Rn_SP. *** gas/ChangeLog *** 2019-04-11 Sudakshina Das <sudi.das@arm.com> * config/tc-aarch64.c (process_omitted_operand): Add case for AARCH64_OPND_Rt_SP. (parse_operands): Likewise. * testsuite/gas/aarch64/armv8_5-a-memtag.d: Update tests. * testsuite/gas/aarch64/armv8_5-a-memtag.s: Likewise. * testsuite/gas/aarch64/illegal-memtag.l: Likewise. * testsuite/gas/aarch64/illegal-memtag.s: Likewise. *** include/ChangeLog *** 2019-04-11 Sudakshina Das <sudi.das@arm.com> * opcode/aarch64.h (enum aarch64_opnd): Add AARCH64_OPND_Rt_SP. *** opcodes/ChangeLog *** 2019-04-11 Sudakshina Das <sudi.das@arm.com> * aarch64-opc.c (aarch64_print_operand): Add case for AARCH64_OPND_Rt_SP. (verify_constraints): Likewise. * aarch64-tbl.h (QL_LDST_AT): Update to add SP qualifier. (struct aarch64_opcode): Update stg, stzg, st2g, stz2g instructions to accept Rt|SP as first operand. (AARCH64_OPERANDS): Add new Rt_SP. * aarch64-asm-2.c: Regenerated. * aarch64-dis-2.c: Regenerated. * aarch64-opc-2.c: Regenerated.
2019-04-11 17:19:37 +08:00
case AARCH64_OPND_Rt_SP:
case AARCH64_OPND_SVE_Rn_SP:
case AARCH64_OPND_Rm_SP:
assert (opnd->qualifier == AARCH64_OPND_QLF_W
|| opnd->qualifier == AARCH64_OPND_QLF_WSP
|| opnd->qualifier == AARCH64_OPND_QLF_X
|| opnd->qualifier == AARCH64_OPND_QLF_SP);
snprintf (buf, size, "%s",
libopcodes/aarch64: add support for disassembler styling This commit enables disassembler styling for AArch64. After this commit it is possible to have objdump style AArch64 disassembler output (using --disassembler-color option). Once the required GDB patches are merged, GDB will also style the disassembler output. The changes to support styling are mostly split between two files opcodes/aarch64-dis.c and opcodes/aarch64-opc.c. The entry point for the AArch64 disassembler can be found in aarch64-dis.c, this file handles printing the instruction mnemonics, and assembler directives (e.g. '.byte', '.word', etc). Some operands, mostly relating to assembler directives are also printed from this file. This commit changes all of this to pass through suitable styling information. However, for most "normal" instructions, the instruction operands are printed using a two step process. From aarch64-dis.c, in the print_operands function, the function aarch64_print_operand is called, this function is in aarch64-opc.c, and converts an instruction operand into a string. Then, back in print_operands (aarch64-dis.c), the operand string is printed. Unfortunately, the string returned by aarch64_print_operand can be quite complex, it will include syntax elements, like '[' and ']', in addition to register names and immediate values. In some cases, a single operand will expand into what will appear (to the user) as multiple operands separated with a ','. This makes the task of styling more complex, all these different components need to by styled differently, so we need to get the styling information out of aarch64_print_operand in some way. The solution that I propose here is similar to the solution that I used for the i386 disassembler. Currently, aarch64_print_operand uses snprintf to write the operand text into a buffer provided by the caller. What I propose is that we pass an extra argument to the aarch64_print_operand function, this argument will be a structure, the structure contains a callback function and some state. When aarch64_print_operand needs to format part of its output this can be done by using the callback function within the new structure, this callback returns a string with special embedded markers that indicate which mode should be used for each piece of text. Back in aarch64-dis.c we can spot these special style markers and use this to split the disassembler output up and apply the correct style to each piece. To make aarch64-opc.c clearer a series of new static functions have been added, e.g. 'style_reg', 'style_imm', etc. Each of these functions formats a piece of text in a different style, 'register' and 'immediate' in this case. Here's an example taken from aarch64-opc.c of the new functions in use: snprintf (buf, size, "[%s, %s]!", style_reg (styler, base), style_imm (styler, "#%d", opnd->addr.offset.imm)); The aarch64_print_operand function is also called from the assembler to aid in printing diagnostic messages. Right now I have no plans to add styling to the assembler output, and so, the callback function used in the assembler ignores the styling information and just returns an plain string. I've used the source files in gas/testsuite/gas/aarch64/ for testing, and have manually gone through and checked that the styling looks reasonable, however, I'm not an AArch64 expert, so it is possible that the odd piece is styled incorrectly. Please point out any mistakes I've made. With objdump disassembler color turned off, there should be no change in the output after this commit.
2022-04-28 20:31:07 +08:00
style_reg (styler, get_int_reg_name (opnd->reg.regno,
opnd->qualifier, 1)));
break;
case AARCH64_OPND_Rm_EXT:
kind = opnd->shifter.kind;
assert (idx == 1 || idx == 2);
if ((aarch64_stack_pointer_p (opnds)
|| (idx == 2 && aarch64_stack_pointer_p (opnds + 1)))
&& ((opnd->qualifier == AARCH64_OPND_QLF_W
&& opnds[0].qualifier == AARCH64_OPND_QLF_W
&& kind == AARCH64_MOD_UXTW)
|| (opnd->qualifier == AARCH64_OPND_QLF_X
&& kind == AARCH64_MOD_UXTX)))
{
/* 'LSL' is the preferred form in this case. */
kind = AARCH64_MOD_LSL;
if (opnd->shifter.amount == 0)
{
/* Shifter omitted. */
snprintf (buf, size, "%s",
libopcodes/aarch64: add support for disassembler styling This commit enables disassembler styling for AArch64. After this commit it is possible to have objdump style AArch64 disassembler output (using --disassembler-color option). Once the required GDB patches are merged, GDB will also style the disassembler output. The changes to support styling are mostly split between two files opcodes/aarch64-dis.c and opcodes/aarch64-opc.c. The entry point for the AArch64 disassembler can be found in aarch64-dis.c, this file handles printing the instruction mnemonics, and assembler directives (e.g. '.byte', '.word', etc). Some operands, mostly relating to assembler directives are also printed from this file. This commit changes all of this to pass through suitable styling information. However, for most "normal" instructions, the instruction operands are printed using a two step process. From aarch64-dis.c, in the print_operands function, the function aarch64_print_operand is called, this function is in aarch64-opc.c, and converts an instruction operand into a string. Then, back in print_operands (aarch64-dis.c), the operand string is printed. Unfortunately, the string returned by aarch64_print_operand can be quite complex, it will include syntax elements, like '[' and ']', in addition to register names and immediate values. In some cases, a single operand will expand into what will appear (to the user) as multiple operands separated with a ','. This makes the task of styling more complex, all these different components need to by styled differently, so we need to get the styling information out of aarch64_print_operand in some way. The solution that I propose here is similar to the solution that I used for the i386 disassembler. Currently, aarch64_print_operand uses snprintf to write the operand text into a buffer provided by the caller. What I propose is that we pass an extra argument to the aarch64_print_operand function, this argument will be a structure, the structure contains a callback function and some state. When aarch64_print_operand needs to format part of its output this can be done by using the callback function within the new structure, this callback returns a string with special embedded markers that indicate which mode should be used for each piece of text. Back in aarch64-dis.c we can spot these special style markers and use this to split the disassembler output up and apply the correct style to each piece. To make aarch64-opc.c clearer a series of new static functions have been added, e.g. 'style_reg', 'style_imm', etc. Each of these functions formats a piece of text in a different style, 'register' and 'immediate' in this case. Here's an example taken from aarch64-opc.c of the new functions in use: snprintf (buf, size, "[%s, %s]!", style_reg (styler, base), style_imm (styler, "#%d", opnd->addr.offset.imm)); The aarch64_print_operand function is also called from the assembler to aid in printing diagnostic messages. Right now I have no plans to add styling to the assembler output, and so, the callback function used in the assembler ignores the styling information and just returns an plain string. I've used the source files in gas/testsuite/gas/aarch64/ for testing, and have manually gone through and checked that the styling looks reasonable, however, I'm not an AArch64 expert, so it is possible that the odd piece is styled incorrectly. Please point out any mistakes I've made. With objdump disassembler color turned off, there should be no change in the output after this commit.
2022-04-28 20:31:07 +08:00
style_reg (styler,
get_int_reg_name (opnd->reg.regno,
opnd->qualifier, 0)));
break;
}
}
if (opnd->shifter.amount)
libopcodes/aarch64: add support for disassembler styling This commit enables disassembler styling for AArch64. After this commit it is possible to have objdump style AArch64 disassembler output (using --disassembler-color option). Once the required GDB patches are merged, GDB will also style the disassembler output. The changes to support styling are mostly split between two files opcodes/aarch64-dis.c and opcodes/aarch64-opc.c. The entry point for the AArch64 disassembler can be found in aarch64-dis.c, this file handles printing the instruction mnemonics, and assembler directives (e.g. '.byte', '.word', etc). Some operands, mostly relating to assembler directives are also printed from this file. This commit changes all of this to pass through suitable styling information. However, for most "normal" instructions, the instruction operands are printed using a two step process. From aarch64-dis.c, in the print_operands function, the function aarch64_print_operand is called, this function is in aarch64-opc.c, and converts an instruction operand into a string. Then, back in print_operands (aarch64-dis.c), the operand string is printed. Unfortunately, the string returned by aarch64_print_operand can be quite complex, it will include syntax elements, like '[' and ']', in addition to register names and immediate values. In some cases, a single operand will expand into what will appear (to the user) as multiple operands separated with a ','. This makes the task of styling more complex, all these different components need to by styled differently, so we need to get the styling information out of aarch64_print_operand in some way. The solution that I propose here is similar to the solution that I used for the i386 disassembler. Currently, aarch64_print_operand uses snprintf to write the operand text into a buffer provided by the caller. What I propose is that we pass an extra argument to the aarch64_print_operand function, this argument will be a structure, the structure contains a callback function and some state. When aarch64_print_operand needs to format part of its output this can be done by using the callback function within the new structure, this callback returns a string with special embedded markers that indicate which mode should be used for each piece of text. Back in aarch64-dis.c we can spot these special style markers and use this to split the disassembler output up and apply the correct style to each piece. To make aarch64-opc.c clearer a series of new static functions have been added, e.g. 'style_reg', 'style_imm', etc. Each of these functions formats a piece of text in a different style, 'register' and 'immediate' in this case. Here's an example taken from aarch64-opc.c of the new functions in use: snprintf (buf, size, "[%s, %s]!", style_reg (styler, base), style_imm (styler, "#%d", opnd->addr.offset.imm)); The aarch64_print_operand function is also called from the assembler to aid in printing diagnostic messages. Right now I have no plans to add styling to the assembler output, and so, the callback function used in the assembler ignores the styling information and just returns an plain string. I've used the source files in gas/testsuite/gas/aarch64/ for testing, and have manually gone through and checked that the styling looks reasonable, however, I'm not an AArch64 expert, so it is possible that the odd piece is styled incorrectly. Please point out any mistakes I've made. With objdump disassembler color turned off, there should be no change in the output after this commit.
2022-04-28 20:31:07 +08:00
snprintf (buf, size, "%s, %s %s",
style_reg (styler, get_int_reg_name (opnd->reg.regno, opnd->qualifier, 0)),
style_sub_mnem (styler, aarch64_operand_modifiers[kind].name),
style_imm (styler, "#%" PRIi64, opnd->shifter.amount));
else
snprintf (buf, size, "%s, %s",
libopcodes/aarch64: add support for disassembler styling This commit enables disassembler styling for AArch64. After this commit it is possible to have objdump style AArch64 disassembler output (using --disassembler-color option). Once the required GDB patches are merged, GDB will also style the disassembler output. The changes to support styling are mostly split between two files opcodes/aarch64-dis.c and opcodes/aarch64-opc.c. The entry point for the AArch64 disassembler can be found in aarch64-dis.c, this file handles printing the instruction mnemonics, and assembler directives (e.g. '.byte', '.word', etc). Some operands, mostly relating to assembler directives are also printed from this file. This commit changes all of this to pass through suitable styling information. However, for most "normal" instructions, the instruction operands are printed using a two step process. From aarch64-dis.c, in the print_operands function, the function aarch64_print_operand is called, this function is in aarch64-opc.c, and converts an instruction operand into a string. Then, back in print_operands (aarch64-dis.c), the operand string is printed. Unfortunately, the string returned by aarch64_print_operand can be quite complex, it will include syntax elements, like '[' and ']', in addition to register names and immediate values. In some cases, a single operand will expand into what will appear (to the user) as multiple operands separated with a ','. This makes the task of styling more complex, all these different components need to by styled differently, so we need to get the styling information out of aarch64_print_operand in some way. The solution that I propose here is similar to the solution that I used for the i386 disassembler. Currently, aarch64_print_operand uses snprintf to write the operand text into a buffer provided by the caller. What I propose is that we pass an extra argument to the aarch64_print_operand function, this argument will be a structure, the structure contains a callback function and some state. When aarch64_print_operand needs to format part of its output this can be done by using the callback function within the new structure, this callback returns a string with special embedded markers that indicate which mode should be used for each piece of text. Back in aarch64-dis.c we can spot these special style markers and use this to split the disassembler output up and apply the correct style to each piece. To make aarch64-opc.c clearer a series of new static functions have been added, e.g. 'style_reg', 'style_imm', etc. Each of these functions formats a piece of text in a different style, 'register' and 'immediate' in this case. Here's an example taken from aarch64-opc.c of the new functions in use: snprintf (buf, size, "[%s, %s]!", style_reg (styler, base), style_imm (styler, "#%d", opnd->addr.offset.imm)); The aarch64_print_operand function is also called from the assembler to aid in printing diagnostic messages. Right now I have no plans to add styling to the assembler output, and so, the callback function used in the assembler ignores the styling information and just returns an plain string. I've used the source files in gas/testsuite/gas/aarch64/ for testing, and have manually gone through and checked that the styling looks reasonable, however, I'm not an AArch64 expert, so it is possible that the odd piece is styled incorrectly. Please point out any mistakes I've made. With objdump disassembler color turned off, there should be no change in the output after this commit.
2022-04-28 20:31:07 +08:00
style_reg (styler, get_int_reg_name (opnd->reg.regno, opnd->qualifier, 0)),
style_sub_mnem (styler, aarch64_operand_modifiers[kind].name));
break;
case AARCH64_OPND_Rm_SFT:
assert (opnd->qualifier == AARCH64_OPND_QLF_W
|| opnd->qualifier == AARCH64_OPND_QLF_X);
if (opnd->shifter.amount == 0 && opnd->shifter.kind == AARCH64_MOD_LSL)
snprintf (buf, size, "%s",
style_reg (styler, get_int_reg_name (opnd->reg.regno,
opnd->qualifier, 0)));
else
snprintf (buf, size, "%s, %s %s",
style_reg (styler, get_int_reg_name (opnd->reg.regno, opnd->qualifier, 0)),
style_sub_mnem (styler, aarch64_operand_modifiers[opnd->shifter.kind].name),
style_imm (styler, "#%" PRIi64, opnd->shifter.amount));
break;
case AARCH64_OPND_Rm_LSL:
assert (opnd->qualifier == AARCH64_OPND_QLF_X);
assert (opnd->shifter.kind == AARCH64_MOD_LSL);
if (opnd->shifter.amount == 0)
snprintf (buf, size, "%s",
libopcodes/aarch64: add support for disassembler styling This commit enables disassembler styling for AArch64. After this commit it is possible to have objdump style AArch64 disassembler output (using --disassembler-color option). Once the required GDB patches are merged, GDB will also style the disassembler output. The changes to support styling are mostly split between two files opcodes/aarch64-dis.c and opcodes/aarch64-opc.c. The entry point for the AArch64 disassembler can be found in aarch64-dis.c, this file handles printing the instruction mnemonics, and assembler directives (e.g. '.byte', '.word', etc). Some operands, mostly relating to assembler directives are also printed from this file. This commit changes all of this to pass through suitable styling information. However, for most "normal" instructions, the instruction operands are printed using a two step process. From aarch64-dis.c, in the print_operands function, the function aarch64_print_operand is called, this function is in aarch64-opc.c, and converts an instruction operand into a string. Then, back in print_operands (aarch64-dis.c), the operand string is printed. Unfortunately, the string returned by aarch64_print_operand can be quite complex, it will include syntax elements, like '[' and ']', in addition to register names and immediate values. In some cases, a single operand will expand into what will appear (to the user) as multiple operands separated with a ','. This makes the task of styling more complex, all these different components need to by styled differently, so we need to get the styling information out of aarch64_print_operand in some way. The solution that I propose here is similar to the solution that I used for the i386 disassembler. Currently, aarch64_print_operand uses snprintf to write the operand text into a buffer provided by the caller. What I propose is that we pass an extra argument to the aarch64_print_operand function, this argument will be a structure, the structure contains a callback function and some state. When aarch64_print_operand needs to format part of its output this can be done by using the callback function within the new structure, this callback returns a string with special embedded markers that indicate which mode should be used for each piece of text. Back in aarch64-dis.c we can spot these special style markers and use this to split the disassembler output up and apply the correct style to each piece. To make aarch64-opc.c clearer a series of new static functions have been added, e.g. 'style_reg', 'style_imm', etc. Each of these functions formats a piece of text in a different style, 'register' and 'immediate' in this case. Here's an example taken from aarch64-opc.c of the new functions in use: snprintf (buf, size, "[%s, %s]!", style_reg (styler, base), style_imm (styler, "#%d", opnd->addr.offset.imm)); The aarch64_print_operand function is also called from the assembler to aid in printing diagnostic messages. Right now I have no plans to add styling to the assembler output, and so, the callback function used in the assembler ignores the styling information and just returns an plain string. I've used the source files in gas/testsuite/gas/aarch64/ for testing, and have manually gone through and checked that the styling looks reasonable, however, I'm not an AArch64 expert, so it is possible that the odd piece is styled incorrectly. Please point out any mistakes I've made. With objdump disassembler color turned off, there should be no change in the output after this commit.
2022-04-28 20:31:07 +08:00
style_reg (styler, get_int_reg_name (opnd->reg.regno,
opnd->qualifier, 0)));
else
libopcodes/aarch64: add support for disassembler styling This commit enables disassembler styling for AArch64. After this commit it is possible to have objdump style AArch64 disassembler output (using --disassembler-color option). Once the required GDB patches are merged, GDB will also style the disassembler output. The changes to support styling are mostly split between two files opcodes/aarch64-dis.c and opcodes/aarch64-opc.c. The entry point for the AArch64 disassembler can be found in aarch64-dis.c, this file handles printing the instruction mnemonics, and assembler directives (e.g. '.byte', '.word', etc). Some operands, mostly relating to assembler directives are also printed from this file. This commit changes all of this to pass through suitable styling information. However, for most "normal" instructions, the instruction operands are printed using a two step process. From aarch64-dis.c, in the print_operands function, the function aarch64_print_operand is called, this function is in aarch64-opc.c, and converts an instruction operand into a string. Then, back in print_operands (aarch64-dis.c), the operand string is printed. Unfortunately, the string returned by aarch64_print_operand can be quite complex, it will include syntax elements, like '[' and ']', in addition to register names and immediate values. In some cases, a single operand will expand into what will appear (to the user) as multiple operands separated with a ','. This makes the task of styling more complex, all these different components need to by styled differently, so we need to get the styling information out of aarch64_print_operand in some way. The solution that I propose here is similar to the solution that I used for the i386 disassembler. Currently, aarch64_print_operand uses snprintf to write the operand text into a buffer provided by the caller. What I propose is that we pass an extra argument to the aarch64_print_operand function, this argument will be a structure, the structure contains a callback function and some state. When aarch64_print_operand needs to format part of its output this can be done by using the callback function within the new structure, this callback returns a string with special embedded markers that indicate which mode should be used for each piece of text. Back in aarch64-dis.c we can spot these special style markers and use this to split the disassembler output up and apply the correct style to each piece. To make aarch64-opc.c clearer a series of new static functions have been added, e.g. 'style_reg', 'style_imm', etc. Each of these functions formats a piece of text in a different style, 'register' and 'immediate' in this case. Here's an example taken from aarch64-opc.c of the new functions in use: snprintf (buf, size, "[%s, %s]!", style_reg (styler, base), style_imm (styler, "#%d", opnd->addr.offset.imm)); The aarch64_print_operand function is also called from the assembler to aid in printing diagnostic messages. Right now I have no plans to add styling to the assembler output, and so, the callback function used in the assembler ignores the styling information and just returns an plain string. I've used the source files in gas/testsuite/gas/aarch64/ for testing, and have manually gone through and checked that the styling looks reasonable, however, I'm not an AArch64 expert, so it is possible that the odd piece is styled incorrectly. Please point out any mistakes I've made. With objdump disassembler color turned off, there should be no change in the output after this commit.
2022-04-28 20:31:07 +08:00
snprintf (buf, size, "%s, %s %s",
style_reg (styler, get_int_reg_name (opnd->reg.regno, opnd->qualifier, 0)),
style_sub_mnem (styler, aarch64_operand_modifiers[opnd->shifter.kind].name),
style_imm (styler, "#%" PRIi64, opnd->shifter.amount));
break;
case AARCH64_OPND_Fd:
case AARCH64_OPND_Fn:
case AARCH64_OPND_Fm:
case AARCH64_OPND_Fa:
case AARCH64_OPND_Ft:
case AARCH64_OPND_Ft2:
case AARCH64_OPND_Sd:
case AARCH64_OPND_Sn:
case AARCH64_OPND_Sm:
case AARCH64_OPND_SVE_VZn:
case AARCH64_OPND_SVE_Vd:
case AARCH64_OPND_SVE_Vm:
case AARCH64_OPND_SVE_Vn:
libopcodes/aarch64: add support for disassembler styling This commit enables disassembler styling for AArch64. After this commit it is possible to have objdump style AArch64 disassembler output (using --disassembler-color option). Once the required GDB patches are merged, GDB will also style the disassembler output. The changes to support styling are mostly split between two files opcodes/aarch64-dis.c and opcodes/aarch64-opc.c. The entry point for the AArch64 disassembler can be found in aarch64-dis.c, this file handles printing the instruction mnemonics, and assembler directives (e.g. '.byte', '.word', etc). Some operands, mostly relating to assembler directives are also printed from this file. This commit changes all of this to pass through suitable styling information. However, for most "normal" instructions, the instruction operands are printed using a two step process. From aarch64-dis.c, in the print_operands function, the function aarch64_print_operand is called, this function is in aarch64-opc.c, and converts an instruction operand into a string. Then, back in print_operands (aarch64-dis.c), the operand string is printed. Unfortunately, the string returned by aarch64_print_operand can be quite complex, it will include syntax elements, like '[' and ']', in addition to register names and immediate values. In some cases, a single operand will expand into what will appear (to the user) as multiple operands separated with a ','. This makes the task of styling more complex, all these different components need to by styled differently, so we need to get the styling information out of aarch64_print_operand in some way. The solution that I propose here is similar to the solution that I used for the i386 disassembler. Currently, aarch64_print_operand uses snprintf to write the operand text into a buffer provided by the caller. What I propose is that we pass an extra argument to the aarch64_print_operand function, this argument will be a structure, the structure contains a callback function and some state. When aarch64_print_operand needs to format part of its output this can be done by using the callback function within the new structure, this callback returns a string with special embedded markers that indicate which mode should be used for each piece of text. Back in aarch64-dis.c we can spot these special style markers and use this to split the disassembler output up and apply the correct style to each piece. To make aarch64-opc.c clearer a series of new static functions have been added, e.g. 'style_reg', 'style_imm', etc. Each of these functions formats a piece of text in a different style, 'register' and 'immediate' in this case. Here's an example taken from aarch64-opc.c of the new functions in use: snprintf (buf, size, "[%s, %s]!", style_reg (styler, base), style_imm (styler, "#%d", opnd->addr.offset.imm)); The aarch64_print_operand function is also called from the assembler to aid in printing diagnostic messages. Right now I have no plans to add styling to the assembler output, and so, the callback function used in the assembler ignores the styling information and just returns an plain string. I've used the source files in gas/testsuite/gas/aarch64/ for testing, and have manually gone through and checked that the styling looks reasonable, however, I'm not an AArch64 expert, so it is possible that the odd piece is styled incorrectly. Please point out any mistakes I've made. With objdump disassembler color turned off, there should be no change in the output after this commit.
2022-04-28 20:31:07 +08:00
snprintf (buf, size, "%s",
style_reg (styler, "%s%d",
aarch64_get_qualifier_name (opnd->qualifier),
opnd->reg.regno));
break;
case AARCH64_OPND_Va:
case AARCH64_OPND_Vd:
case AARCH64_OPND_Vn:
case AARCH64_OPND_Vm:
libopcodes/aarch64: add support for disassembler styling This commit enables disassembler styling for AArch64. After this commit it is possible to have objdump style AArch64 disassembler output (using --disassembler-color option). Once the required GDB patches are merged, GDB will also style the disassembler output. The changes to support styling are mostly split between two files opcodes/aarch64-dis.c and opcodes/aarch64-opc.c. The entry point for the AArch64 disassembler can be found in aarch64-dis.c, this file handles printing the instruction mnemonics, and assembler directives (e.g. '.byte', '.word', etc). Some operands, mostly relating to assembler directives are also printed from this file. This commit changes all of this to pass through suitable styling information. However, for most "normal" instructions, the instruction operands are printed using a two step process. From aarch64-dis.c, in the print_operands function, the function aarch64_print_operand is called, this function is in aarch64-opc.c, and converts an instruction operand into a string. Then, back in print_operands (aarch64-dis.c), the operand string is printed. Unfortunately, the string returned by aarch64_print_operand can be quite complex, it will include syntax elements, like '[' and ']', in addition to register names and immediate values. In some cases, a single operand will expand into what will appear (to the user) as multiple operands separated with a ','. This makes the task of styling more complex, all these different components need to by styled differently, so we need to get the styling information out of aarch64_print_operand in some way. The solution that I propose here is similar to the solution that I used for the i386 disassembler. Currently, aarch64_print_operand uses snprintf to write the operand text into a buffer provided by the caller. What I propose is that we pass an extra argument to the aarch64_print_operand function, this argument will be a structure, the structure contains a callback function and some state. When aarch64_print_operand needs to format part of its output this can be done by using the callback function within the new structure, this callback returns a string with special embedded markers that indicate which mode should be used for each piece of text. Back in aarch64-dis.c we can spot these special style markers and use this to split the disassembler output up and apply the correct style to each piece. To make aarch64-opc.c clearer a series of new static functions have been added, e.g. 'style_reg', 'style_imm', etc. Each of these functions formats a piece of text in a different style, 'register' and 'immediate' in this case. Here's an example taken from aarch64-opc.c of the new functions in use: snprintf (buf, size, "[%s, %s]!", style_reg (styler, base), style_imm (styler, "#%d", opnd->addr.offset.imm)); The aarch64_print_operand function is also called from the assembler to aid in printing diagnostic messages. Right now I have no plans to add styling to the assembler output, and so, the callback function used in the assembler ignores the styling information and just returns an plain string. I've used the source files in gas/testsuite/gas/aarch64/ for testing, and have manually gone through and checked that the styling looks reasonable, however, I'm not an AArch64 expert, so it is possible that the odd piece is styled incorrectly. Please point out any mistakes I've made. With objdump disassembler color turned off, there should be no change in the output after this commit.
2022-04-28 20:31:07 +08:00
snprintf (buf, size, "%s",
style_reg (styler, "v%d.%s", opnd->reg.regno,
aarch64_get_qualifier_name (opnd->qualifier)));
break;
case AARCH64_OPND_Ed:
case AARCH64_OPND_En:
case AARCH64_OPND_Em:
Fix AArch64 encodings for by element instructions. Some instructions in Armv8-a place a limitation on FP16 registers that can be used as the register from which to select an element from. e.g. fmla restricts Rm to 4 bits when using an FP16 register. This restriction does not apply for all instructions, e.g. fcmla does not have this restriction as it gets an extra bit from the M field. Unfortunately, this restriction to S_H was added for all _Em operands before, meaning for a large number of instructions you couldn't use the full register file. This fixes the issue by introducing a new operand _Em16 which applies this restriction only when paired with S_H and leaves the _Em and the other qualifiers for _Em16 unbounded (i.e. using the full 5 bit range). Also the patch updates all instructions that should be affected by this. opcodes/ PR binutils/23192 * aarch64-asm-2.c: Regenerate. * aarch64-dis-2.c: Likewise. * aarch64-opc-2.c: Likewise. * aarch64-dis.c (aarch64_ext_reglane): Add AARCH64_OPND_Em16 constraint. * aarch64-opc.c (operand_general_constraint_met_p, aarch64_print_operand): Likewise. * aarch64-tbl.h (aarch64_opcode_table): Change Em to Em16 for smlal, smlal2, fmla, fmls, fmul, fmulx, sqrdmlah, sqrdlsh, fmlal, fmlsl, fmlal2, fmlsl2. (AARCH64_OPERANDS): Add Em2. gas/ PR binutils/23192 * config/tc-aarch64.c (process_omitted_operand, parse_operands): Add AARCH64_OPND_Em16 * testsuite/gas/aarch64/advsimd-armv8_3.s: Expand tests to cover upper 16 registers. * testsuite/gas/aarch64/advsimd-armv8_3.d: Likewise. * testsuite/gas/aarch64/advsimd-compnum.s: Likewise. * testsuite/gas/aarch64/advsimd-compnum.d: Likewise. * testsuite/gas/aarch64/sve.d: Likewise. include/ PR binutils/23192 *opcode/aarch64.h (aarch64_opnd): Add AARCH64_OPND_Em16.
2018-06-29 19:12:27 +08:00
case AARCH64_OPND_Em16:
case AARCH64_OPND_SM3_IMM2:
libopcodes/aarch64: add support for disassembler styling This commit enables disassembler styling for AArch64. After this commit it is possible to have objdump style AArch64 disassembler output (using --disassembler-color option). Once the required GDB patches are merged, GDB will also style the disassembler output. The changes to support styling are mostly split between two files opcodes/aarch64-dis.c and opcodes/aarch64-opc.c. The entry point for the AArch64 disassembler can be found in aarch64-dis.c, this file handles printing the instruction mnemonics, and assembler directives (e.g. '.byte', '.word', etc). Some operands, mostly relating to assembler directives are also printed from this file. This commit changes all of this to pass through suitable styling information. However, for most "normal" instructions, the instruction operands are printed using a two step process. From aarch64-dis.c, in the print_operands function, the function aarch64_print_operand is called, this function is in aarch64-opc.c, and converts an instruction operand into a string. Then, back in print_operands (aarch64-dis.c), the operand string is printed. Unfortunately, the string returned by aarch64_print_operand can be quite complex, it will include syntax elements, like '[' and ']', in addition to register names and immediate values. In some cases, a single operand will expand into what will appear (to the user) as multiple operands separated with a ','. This makes the task of styling more complex, all these different components need to by styled differently, so we need to get the styling information out of aarch64_print_operand in some way. The solution that I propose here is similar to the solution that I used for the i386 disassembler. Currently, aarch64_print_operand uses snprintf to write the operand text into a buffer provided by the caller. What I propose is that we pass an extra argument to the aarch64_print_operand function, this argument will be a structure, the structure contains a callback function and some state. When aarch64_print_operand needs to format part of its output this can be done by using the callback function within the new structure, this callback returns a string with special embedded markers that indicate which mode should be used for each piece of text. Back in aarch64-dis.c we can spot these special style markers and use this to split the disassembler output up and apply the correct style to each piece. To make aarch64-opc.c clearer a series of new static functions have been added, e.g. 'style_reg', 'style_imm', etc. Each of these functions formats a piece of text in a different style, 'register' and 'immediate' in this case. Here's an example taken from aarch64-opc.c of the new functions in use: snprintf (buf, size, "[%s, %s]!", style_reg (styler, base), style_imm (styler, "#%d", opnd->addr.offset.imm)); The aarch64_print_operand function is also called from the assembler to aid in printing diagnostic messages. Right now I have no plans to add styling to the assembler output, and so, the callback function used in the assembler ignores the styling information and just returns an plain string. I've used the source files in gas/testsuite/gas/aarch64/ for testing, and have manually gone through and checked that the styling looks reasonable, however, I'm not an AArch64 expert, so it is possible that the odd piece is styled incorrectly. Please point out any mistakes I've made. With objdump disassembler color turned off, there should be no change in the output after this commit.
2022-04-28 20:31:07 +08:00
snprintf (buf, size, "%s[%s]",
style_reg (styler, "v%d.%s", opnd->reglane.regno,
aarch64_get_qualifier_name (opnd->qualifier)),
style_imm (styler, "%" PRIi64, opnd->reglane.index));
break;
case AARCH64_OPND_VdD1:
case AARCH64_OPND_VnD1:
libopcodes/aarch64: add support for disassembler styling This commit enables disassembler styling for AArch64. After this commit it is possible to have objdump style AArch64 disassembler output (using --disassembler-color option). Once the required GDB patches are merged, GDB will also style the disassembler output. The changes to support styling are mostly split between two files opcodes/aarch64-dis.c and opcodes/aarch64-opc.c. The entry point for the AArch64 disassembler can be found in aarch64-dis.c, this file handles printing the instruction mnemonics, and assembler directives (e.g. '.byte', '.word', etc). Some operands, mostly relating to assembler directives are also printed from this file. This commit changes all of this to pass through suitable styling information. However, for most "normal" instructions, the instruction operands are printed using a two step process. From aarch64-dis.c, in the print_operands function, the function aarch64_print_operand is called, this function is in aarch64-opc.c, and converts an instruction operand into a string. Then, back in print_operands (aarch64-dis.c), the operand string is printed. Unfortunately, the string returned by aarch64_print_operand can be quite complex, it will include syntax elements, like '[' and ']', in addition to register names and immediate values. In some cases, a single operand will expand into what will appear (to the user) as multiple operands separated with a ','. This makes the task of styling more complex, all these different components need to by styled differently, so we need to get the styling information out of aarch64_print_operand in some way. The solution that I propose here is similar to the solution that I used for the i386 disassembler. Currently, aarch64_print_operand uses snprintf to write the operand text into a buffer provided by the caller. What I propose is that we pass an extra argument to the aarch64_print_operand function, this argument will be a structure, the structure contains a callback function and some state. When aarch64_print_operand needs to format part of its output this can be done by using the callback function within the new structure, this callback returns a string with special embedded markers that indicate which mode should be used for each piece of text. Back in aarch64-dis.c we can spot these special style markers and use this to split the disassembler output up and apply the correct style to each piece. To make aarch64-opc.c clearer a series of new static functions have been added, e.g. 'style_reg', 'style_imm', etc. Each of these functions formats a piece of text in a different style, 'register' and 'immediate' in this case. Here's an example taken from aarch64-opc.c of the new functions in use: snprintf (buf, size, "[%s, %s]!", style_reg (styler, base), style_imm (styler, "#%d", opnd->addr.offset.imm)); The aarch64_print_operand function is also called from the assembler to aid in printing diagnostic messages. Right now I have no plans to add styling to the assembler output, and so, the callback function used in the assembler ignores the styling information and just returns an plain string. I've used the source files in gas/testsuite/gas/aarch64/ for testing, and have manually gone through and checked that the styling looks reasonable, however, I'm not an AArch64 expert, so it is possible that the odd piece is styled incorrectly. Please point out any mistakes I've made. With objdump disassembler color turned off, there should be no change in the output after this commit.
2022-04-28 20:31:07 +08:00
snprintf (buf, size, "%s[%s]",
style_reg (styler, "v%d.d", opnd->reg.regno),
style_imm (styler, "1"));
break;
case AARCH64_OPND_LVn:
case AARCH64_OPND_LVt:
case AARCH64_OPND_LVt_AL:
case AARCH64_OPND_LEt:
libopcodes/aarch64: add support for disassembler styling This commit enables disassembler styling for AArch64. After this commit it is possible to have objdump style AArch64 disassembler output (using --disassembler-color option). Once the required GDB patches are merged, GDB will also style the disassembler output. The changes to support styling are mostly split between two files opcodes/aarch64-dis.c and opcodes/aarch64-opc.c. The entry point for the AArch64 disassembler can be found in aarch64-dis.c, this file handles printing the instruction mnemonics, and assembler directives (e.g. '.byte', '.word', etc). Some operands, mostly relating to assembler directives are also printed from this file. This commit changes all of this to pass through suitable styling information. However, for most "normal" instructions, the instruction operands are printed using a two step process. From aarch64-dis.c, in the print_operands function, the function aarch64_print_operand is called, this function is in aarch64-opc.c, and converts an instruction operand into a string. Then, back in print_operands (aarch64-dis.c), the operand string is printed. Unfortunately, the string returned by aarch64_print_operand can be quite complex, it will include syntax elements, like '[' and ']', in addition to register names and immediate values. In some cases, a single operand will expand into what will appear (to the user) as multiple operands separated with a ','. This makes the task of styling more complex, all these different components need to by styled differently, so we need to get the styling information out of aarch64_print_operand in some way. The solution that I propose here is similar to the solution that I used for the i386 disassembler. Currently, aarch64_print_operand uses snprintf to write the operand text into a buffer provided by the caller. What I propose is that we pass an extra argument to the aarch64_print_operand function, this argument will be a structure, the structure contains a callback function and some state. When aarch64_print_operand needs to format part of its output this can be done by using the callback function within the new structure, this callback returns a string with special embedded markers that indicate which mode should be used for each piece of text. Back in aarch64-dis.c we can spot these special style markers and use this to split the disassembler output up and apply the correct style to each piece. To make aarch64-opc.c clearer a series of new static functions have been added, e.g. 'style_reg', 'style_imm', etc. Each of these functions formats a piece of text in a different style, 'register' and 'immediate' in this case. Here's an example taken from aarch64-opc.c of the new functions in use: snprintf (buf, size, "[%s, %s]!", style_reg (styler, base), style_imm (styler, "#%d", opnd->addr.offset.imm)); The aarch64_print_operand function is also called from the assembler to aid in printing diagnostic messages. Right now I have no plans to add styling to the assembler output, and so, the callback function used in the assembler ignores the styling information and just returns an plain string. I've used the source files in gas/testsuite/gas/aarch64/ for testing, and have manually gone through and checked that the styling looks reasonable, however, I'm not an AArch64 expert, so it is possible that the odd piece is styled incorrectly. Please point out any mistakes I've made. With objdump disassembler color turned off, there should be no change in the output after this commit.
2022-04-28 20:31:07 +08:00
print_register_list (buf, size, opnd, "v", styler);
break;
[AArch64][SVE 21/32] Add Zn and Pn registers This patch adds the Zn and Pn registers, and associated fields and operands. include/ * opcode/aarch64.h (AARCH64_OPND_CLASS_SVE_REG): New aarch64_operand_class. (AARCH64_OPND_CLASS_PRED_REG): Likewise. (AARCH64_OPND_SVE_Pd, AARCH64_OPND_SVE_Pg3, AARCH64_OPND_SVE_Pg4_5) (AARCH64_OPND_SVE_Pg4_10, AARCH64_OPND_SVE_Pg4_16) (AARCH64_OPND_SVE_Pm, AARCH64_OPND_SVE_Pn, AARCH64_OPND_SVE_Pt) (AARCH64_OPND_SVE_Za_5, AARCH64_OPND_SVE_Za_16, AARCH64_OPND_SVE_Zd) (AARCH64_OPND_SVE_Zm_5, AARCH64_OPND_SVE_Zm_16, AARCH64_OPND_SVE_Zn) (AARCH64_OPND_SVE_Zn_INDEX, AARCH64_OPND_SVE_ZnxN) (AARCH64_OPND_SVE_Zt, AARCH64_OPND_SVE_ZtxN): New aarch64_opnds. opcodes/ * aarch64-tbl.h (AARCH64_OPERANDS): Add entries for new SVE operands. * aarch64-opc.h (FLD_SVE_Pd, FLD_SVE_Pg3, FLD_SVE_Pg4_5) (FLD_SVE_Pg4_10, FLD_SVE_Pg4_16, FLD_SVE_Pm, FLD_SVE_Pn, FLD_SVE_Pt) (FLD_SVE_Za_5, FLD_SVE_Za_16, FLD_SVE_Zd, FLD_SVE_Zm_5, FLD_SVE_Zm_16) (FLD_SVE_Zn, FLD_SVE_Zt, FLD_SVE_tzsh): New aarch64_field_kinds. * aarch64-opc.c (fields): Add corresponding entries here. (operand_general_constraint_met_p): Check that SVE register lists have the correct length. Check the ranges of SVE index registers. Check for cases where p8-p15 are used in 3-bit predicate fields. (aarch64_print_operand): Handle the new SVE operands. * aarch64-opc-2.c: Regenerate. * aarch64-asm.h (ins_sve_index, ins_sve_reglist): New inserters. * aarch64-asm.c (aarch64_ins_sve_index): New function. (aarch64_ins_sve_reglist): Likewise. * aarch64-asm-2.c: Regenerate. * aarch64-dis.h (ext_sve_index, ext_sve_reglist): New extractors. * aarch64-dis.c (aarch64_ext_sve_index): New function. (aarch64_ext_sve_reglist): Likewise. * aarch64-dis-2.c: Regenerate. gas/ * config/tc-aarch64.c (NTA_HASVARWIDTH): New macro. (AARCH64_REG_TYPES): Add ZN and PN. (get_reg_expected_msg): Handle them. (parse_vector_type_for_operand): Add a reg_type parameter. Skip the width for Zn and Pn registers. (parse_typed_reg): Extend vector handling to Zn and Pn. Update the call to parse_vector_type_for_operand. Set HASVARTYPE for Zn and Pn, expecting the width to be 0. (parse_vector_reg_list): Restrict error about [BHSD]nn operands to REG_TYPE_VN. (vectype_to_qualifier): Use S_[BHSD] qualifiers for NTA_HASVARWIDTH. (parse_operands): Handle the new Zn and Pn operands. (REGSET16): New macro, split out from... (REGSET31): ...here. (reg_names): Add Zn and Pn entries.
2016-09-21 23:53:54 +08:00
case AARCH64_OPND_SVE_Pd:
case AARCH64_OPND_SVE_Pg3:
case AARCH64_OPND_SVE_Pg4_5:
case AARCH64_OPND_SVE_Pg4_10:
case AARCH64_OPND_SVE_Pg4_16:
case AARCH64_OPND_SVE_Pm:
case AARCH64_OPND_SVE_Pn:
case AARCH64_OPND_SVE_Pt:
aarch64: [SME] Add SME instructions Patch is adding new SME matrix instructions. Please note additional instructions will be added in following patches. gas/ChangeLog: * config/tc-aarch64.c (parse_sme_zada_operand): New parser. * config/tc-aarch64.c (parse_reg_with_qual): New reg parser. * config/tc-aarch64.c (R_ZA): New egister type. (parse_operands): New parser. * testsuite/gas/aarch64/sme-illegal.d: New test. * testsuite/gas/aarch64/sme-illegal.l: New test. * testsuite/gas/aarch64/sme-illegal.s: New test. * testsuite/gas/aarch64/sme.d: New test. * testsuite/gas/aarch64/sme.s: New test. * testsuite/gas/aarch64/sme-f64.d: New test. * testsuite/gas/aarch64/sme-f64.s: New test. * testsuite/gas/aarch64/sme-i64.d: New test. * testsuite/gas/aarch64/sme-i64.s: New test. include/ChangeLog: * opcode/aarch64.h (enum aarch64_opnd): New operands AARCH64_OPND_SME_ZAda_2b, AARCH64_OPND_SME_ZAda_3b and AARCH64_OPND_SME_Pm. (enum aarch64_insn_class): New instruction class sme_misc. opcodes/ChangeLog: * aarch64-opc.c (aarch64_print_operand): Print OPND_SME_ZAda_2b and OPND_SME_ZAda_3b operands. (verify_constraints): Handle OPND_SME_Pm. * aarch64-opc.h (enum aarch64_field_kind): New bit fields FLD_SME_ZAda_2b, FLD_SME_ZAda_3b and FLD_SME_Pm. * aarch64-tbl.h (OP_SME_ZADA_PN_PM_ZN_S): New qualifier set. (OP_SME_ZADA_PN_PM_ZN_D): New qualifier. (OP_SME_ZADA_PN_PM_ZN_ZM): New qualifier. (OP_SME_ZADA_S_PM_PM_S_S): New qualifier. (OP_SME_ZADA_D_PM_PM_D_D): New qualifier. (OP_SME_ZADA_S_PM_PM_H_H): New qualifier. (OP_SME_ZADA_S_PM_PM_B_B): New qualifier. (OP_SME_ZADA_D_PM_PM_H_H): New qualifier. (SME_INSN): New instruction macro. (SME_F64_INSN): New instruction macro. (SME_I64_INSN): New instruction macro. (SME_INSNC): New instruction macro. (struct aarch64_opcode): New SME instructions. aarch64-asm-2.c: Regenerate. aarch64-dis-2.c: Regenerate. aarch64-opc-2.c: Regenerate.
2021-11-18 03:21:33 +08:00
case AARCH64_OPND_SME_Pm:
[AArch64][SVE 21/32] Add Zn and Pn registers This patch adds the Zn and Pn registers, and associated fields and operands. include/ * opcode/aarch64.h (AARCH64_OPND_CLASS_SVE_REG): New aarch64_operand_class. (AARCH64_OPND_CLASS_PRED_REG): Likewise. (AARCH64_OPND_SVE_Pd, AARCH64_OPND_SVE_Pg3, AARCH64_OPND_SVE_Pg4_5) (AARCH64_OPND_SVE_Pg4_10, AARCH64_OPND_SVE_Pg4_16) (AARCH64_OPND_SVE_Pm, AARCH64_OPND_SVE_Pn, AARCH64_OPND_SVE_Pt) (AARCH64_OPND_SVE_Za_5, AARCH64_OPND_SVE_Za_16, AARCH64_OPND_SVE_Zd) (AARCH64_OPND_SVE_Zm_5, AARCH64_OPND_SVE_Zm_16, AARCH64_OPND_SVE_Zn) (AARCH64_OPND_SVE_Zn_INDEX, AARCH64_OPND_SVE_ZnxN) (AARCH64_OPND_SVE_Zt, AARCH64_OPND_SVE_ZtxN): New aarch64_opnds. opcodes/ * aarch64-tbl.h (AARCH64_OPERANDS): Add entries for new SVE operands. * aarch64-opc.h (FLD_SVE_Pd, FLD_SVE_Pg3, FLD_SVE_Pg4_5) (FLD_SVE_Pg4_10, FLD_SVE_Pg4_16, FLD_SVE_Pm, FLD_SVE_Pn, FLD_SVE_Pt) (FLD_SVE_Za_5, FLD_SVE_Za_16, FLD_SVE_Zd, FLD_SVE_Zm_5, FLD_SVE_Zm_16) (FLD_SVE_Zn, FLD_SVE_Zt, FLD_SVE_tzsh): New aarch64_field_kinds. * aarch64-opc.c (fields): Add corresponding entries here. (operand_general_constraint_met_p): Check that SVE register lists have the correct length. Check the ranges of SVE index registers. Check for cases where p8-p15 are used in 3-bit predicate fields. (aarch64_print_operand): Handle the new SVE operands. * aarch64-opc-2.c: Regenerate. * aarch64-asm.h (ins_sve_index, ins_sve_reglist): New inserters. * aarch64-asm.c (aarch64_ins_sve_index): New function. (aarch64_ins_sve_reglist): Likewise. * aarch64-asm-2.c: Regenerate. * aarch64-dis.h (ext_sve_index, ext_sve_reglist): New extractors. * aarch64-dis.c (aarch64_ext_sve_index): New function. (aarch64_ext_sve_reglist): Likewise. * aarch64-dis-2.c: Regenerate. gas/ * config/tc-aarch64.c (NTA_HASVARWIDTH): New macro. (AARCH64_REG_TYPES): Add ZN and PN. (get_reg_expected_msg): Handle them. (parse_vector_type_for_operand): Add a reg_type parameter. Skip the width for Zn and Pn registers. (parse_typed_reg): Extend vector handling to Zn and Pn. Update the call to parse_vector_type_for_operand. Set HASVARTYPE for Zn and Pn, expecting the width to be 0. (parse_vector_reg_list): Restrict error about [BHSD]nn operands to REG_TYPE_VN. (vectype_to_qualifier): Use S_[BHSD] qualifiers for NTA_HASVARWIDTH. (parse_operands): Handle the new Zn and Pn operands. (REGSET16): New macro, split out from... (REGSET31): ...here. (reg_names): Add Zn and Pn entries.
2016-09-21 23:53:54 +08:00
if (opnd->qualifier == AARCH64_OPND_QLF_NIL)
libopcodes/aarch64: add support for disassembler styling This commit enables disassembler styling for AArch64. After this commit it is possible to have objdump style AArch64 disassembler output (using --disassembler-color option). Once the required GDB patches are merged, GDB will also style the disassembler output. The changes to support styling are mostly split between two files opcodes/aarch64-dis.c and opcodes/aarch64-opc.c. The entry point for the AArch64 disassembler can be found in aarch64-dis.c, this file handles printing the instruction mnemonics, and assembler directives (e.g. '.byte', '.word', etc). Some operands, mostly relating to assembler directives are also printed from this file. This commit changes all of this to pass through suitable styling information. However, for most "normal" instructions, the instruction operands are printed using a two step process. From aarch64-dis.c, in the print_operands function, the function aarch64_print_operand is called, this function is in aarch64-opc.c, and converts an instruction operand into a string. Then, back in print_operands (aarch64-dis.c), the operand string is printed. Unfortunately, the string returned by aarch64_print_operand can be quite complex, it will include syntax elements, like '[' and ']', in addition to register names and immediate values. In some cases, a single operand will expand into what will appear (to the user) as multiple operands separated with a ','. This makes the task of styling more complex, all these different components need to by styled differently, so we need to get the styling information out of aarch64_print_operand in some way. The solution that I propose here is similar to the solution that I used for the i386 disassembler. Currently, aarch64_print_operand uses snprintf to write the operand text into a buffer provided by the caller. What I propose is that we pass an extra argument to the aarch64_print_operand function, this argument will be a structure, the structure contains a callback function and some state. When aarch64_print_operand needs to format part of its output this can be done by using the callback function within the new structure, this callback returns a string with special embedded markers that indicate which mode should be used for each piece of text. Back in aarch64-dis.c we can spot these special style markers and use this to split the disassembler output up and apply the correct style to each piece. To make aarch64-opc.c clearer a series of new static functions have been added, e.g. 'style_reg', 'style_imm', etc. Each of these functions formats a piece of text in a different style, 'register' and 'immediate' in this case. Here's an example taken from aarch64-opc.c of the new functions in use: snprintf (buf, size, "[%s, %s]!", style_reg (styler, base), style_imm (styler, "#%d", opnd->addr.offset.imm)); The aarch64_print_operand function is also called from the assembler to aid in printing diagnostic messages. Right now I have no plans to add styling to the assembler output, and so, the callback function used in the assembler ignores the styling information and just returns an plain string. I've used the source files in gas/testsuite/gas/aarch64/ for testing, and have manually gone through and checked that the styling looks reasonable, however, I'm not an AArch64 expert, so it is possible that the odd piece is styled incorrectly. Please point out any mistakes I've made. With objdump disassembler color turned off, there should be no change in the output after this commit.
2022-04-28 20:31:07 +08:00
snprintf (buf, size, "%s",
style_reg (styler, "p%d", opnd->reg.regno));
else if (opnd->qualifier == AARCH64_OPND_QLF_P_Z
|| opnd->qualifier == AARCH64_OPND_QLF_P_M)
libopcodes/aarch64: add support for disassembler styling This commit enables disassembler styling for AArch64. After this commit it is possible to have objdump style AArch64 disassembler output (using --disassembler-color option). Once the required GDB patches are merged, GDB will also style the disassembler output. The changes to support styling are mostly split between two files opcodes/aarch64-dis.c and opcodes/aarch64-opc.c. The entry point for the AArch64 disassembler can be found in aarch64-dis.c, this file handles printing the instruction mnemonics, and assembler directives (e.g. '.byte', '.word', etc). Some operands, mostly relating to assembler directives are also printed from this file. This commit changes all of this to pass through suitable styling information. However, for most "normal" instructions, the instruction operands are printed using a two step process. From aarch64-dis.c, in the print_operands function, the function aarch64_print_operand is called, this function is in aarch64-opc.c, and converts an instruction operand into a string. Then, back in print_operands (aarch64-dis.c), the operand string is printed. Unfortunately, the string returned by aarch64_print_operand can be quite complex, it will include syntax elements, like '[' and ']', in addition to register names and immediate values. In some cases, a single operand will expand into what will appear (to the user) as multiple operands separated with a ','. This makes the task of styling more complex, all these different components need to by styled differently, so we need to get the styling information out of aarch64_print_operand in some way. The solution that I propose here is similar to the solution that I used for the i386 disassembler. Currently, aarch64_print_operand uses snprintf to write the operand text into a buffer provided by the caller. What I propose is that we pass an extra argument to the aarch64_print_operand function, this argument will be a structure, the structure contains a callback function and some state. When aarch64_print_operand needs to format part of its output this can be done by using the callback function within the new structure, this callback returns a string with special embedded markers that indicate which mode should be used for each piece of text. Back in aarch64-dis.c we can spot these special style markers and use this to split the disassembler output up and apply the correct style to each piece. To make aarch64-opc.c clearer a series of new static functions have been added, e.g. 'style_reg', 'style_imm', etc. Each of these functions formats a piece of text in a different style, 'register' and 'immediate' in this case. Here's an example taken from aarch64-opc.c of the new functions in use: snprintf (buf, size, "[%s, %s]!", style_reg (styler, base), style_imm (styler, "#%d", opnd->addr.offset.imm)); The aarch64_print_operand function is also called from the assembler to aid in printing diagnostic messages. Right now I have no plans to add styling to the assembler output, and so, the callback function used in the assembler ignores the styling information and just returns an plain string. I've used the source files in gas/testsuite/gas/aarch64/ for testing, and have manually gone through and checked that the styling looks reasonable, however, I'm not an AArch64 expert, so it is possible that the odd piece is styled incorrectly. Please point out any mistakes I've made. With objdump disassembler color turned off, there should be no change in the output after this commit.
2022-04-28 20:31:07 +08:00
snprintf (buf, size, "%s",
style_reg (styler, "p%d/%s", opnd->reg.regno,
aarch64_get_qualifier_name (opnd->qualifier)));
[AArch64][SVE 21/32] Add Zn and Pn registers This patch adds the Zn and Pn registers, and associated fields and operands. include/ * opcode/aarch64.h (AARCH64_OPND_CLASS_SVE_REG): New aarch64_operand_class. (AARCH64_OPND_CLASS_PRED_REG): Likewise. (AARCH64_OPND_SVE_Pd, AARCH64_OPND_SVE_Pg3, AARCH64_OPND_SVE_Pg4_5) (AARCH64_OPND_SVE_Pg4_10, AARCH64_OPND_SVE_Pg4_16) (AARCH64_OPND_SVE_Pm, AARCH64_OPND_SVE_Pn, AARCH64_OPND_SVE_Pt) (AARCH64_OPND_SVE_Za_5, AARCH64_OPND_SVE_Za_16, AARCH64_OPND_SVE_Zd) (AARCH64_OPND_SVE_Zm_5, AARCH64_OPND_SVE_Zm_16, AARCH64_OPND_SVE_Zn) (AARCH64_OPND_SVE_Zn_INDEX, AARCH64_OPND_SVE_ZnxN) (AARCH64_OPND_SVE_Zt, AARCH64_OPND_SVE_ZtxN): New aarch64_opnds. opcodes/ * aarch64-tbl.h (AARCH64_OPERANDS): Add entries for new SVE operands. * aarch64-opc.h (FLD_SVE_Pd, FLD_SVE_Pg3, FLD_SVE_Pg4_5) (FLD_SVE_Pg4_10, FLD_SVE_Pg4_16, FLD_SVE_Pm, FLD_SVE_Pn, FLD_SVE_Pt) (FLD_SVE_Za_5, FLD_SVE_Za_16, FLD_SVE_Zd, FLD_SVE_Zm_5, FLD_SVE_Zm_16) (FLD_SVE_Zn, FLD_SVE_Zt, FLD_SVE_tzsh): New aarch64_field_kinds. * aarch64-opc.c (fields): Add corresponding entries here. (operand_general_constraint_met_p): Check that SVE register lists have the correct length. Check the ranges of SVE index registers. Check for cases where p8-p15 are used in 3-bit predicate fields. (aarch64_print_operand): Handle the new SVE operands. * aarch64-opc-2.c: Regenerate. * aarch64-asm.h (ins_sve_index, ins_sve_reglist): New inserters. * aarch64-asm.c (aarch64_ins_sve_index): New function. (aarch64_ins_sve_reglist): Likewise. * aarch64-asm-2.c: Regenerate. * aarch64-dis.h (ext_sve_index, ext_sve_reglist): New extractors. * aarch64-dis.c (aarch64_ext_sve_index): New function. (aarch64_ext_sve_reglist): Likewise. * aarch64-dis-2.c: Regenerate. gas/ * config/tc-aarch64.c (NTA_HASVARWIDTH): New macro. (AARCH64_REG_TYPES): Add ZN and PN. (get_reg_expected_msg): Handle them. (parse_vector_type_for_operand): Add a reg_type parameter. Skip the width for Zn and Pn registers. (parse_typed_reg): Extend vector handling to Zn and Pn. Update the call to parse_vector_type_for_operand. Set HASVARTYPE for Zn and Pn, expecting the width to be 0. (parse_vector_reg_list): Restrict error about [BHSD]nn operands to REG_TYPE_VN. (vectype_to_qualifier): Use S_[BHSD] qualifiers for NTA_HASVARWIDTH. (parse_operands): Handle the new Zn and Pn operands. (REGSET16): New macro, split out from... (REGSET31): ...here. (reg_names): Add Zn and Pn entries.
2016-09-21 23:53:54 +08:00
else
libopcodes/aarch64: add support for disassembler styling This commit enables disassembler styling for AArch64. After this commit it is possible to have objdump style AArch64 disassembler output (using --disassembler-color option). Once the required GDB patches are merged, GDB will also style the disassembler output. The changes to support styling are mostly split between two files opcodes/aarch64-dis.c and opcodes/aarch64-opc.c. The entry point for the AArch64 disassembler can be found in aarch64-dis.c, this file handles printing the instruction mnemonics, and assembler directives (e.g. '.byte', '.word', etc). Some operands, mostly relating to assembler directives are also printed from this file. This commit changes all of this to pass through suitable styling information. However, for most "normal" instructions, the instruction operands are printed using a two step process. From aarch64-dis.c, in the print_operands function, the function aarch64_print_operand is called, this function is in aarch64-opc.c, and converts an instruction operand into a string. Then, back in print_operands (aarch64-dis.c), the operand string is printed. Unfortunately, the string returned by aarch64_print_operand can be quite complex, it will include syntax elements, like '[' and ']', in addition to register names and immediate values. In some cases, a single operand will expand into what will appear (to the user) as multiple operands separated with a ','. This makes the task of styling more complex, all these different components need to by styled differently, so we need to get the styling information out of aarch64_print_operand in some way. The solution that I propose here is similar to the solution that I used for the i386 disassembler. Currently, aarch64_print_operand uses snprintf to write the operand text into a buffer provided by the caller. What I propose is that we pass an extra argument to the aarch64_print_operand function, this argument will be a structure, the structure contains a callback function and some state. When aarch64_print_operand needs to format part of its output this can be done by using the callback function within the new structure, this callback returns a string with special embedded markers that indicate which mode should be used for each piece of text. Back in aarch64-dis.c we can spot these special style markers and use this to split the disassembler output up and apply the correct style to each piece. To make aarch64-opc.c clearer a series of new static functions have been added, e.g. 'style_reg', 'style_imm', etc. Each of these functions formats a piece of text in a different style, 'register' and 'immediate' in this case. Here's an example taken from aarch64-opc.c of the new functions in use: snprintf (buf, size, "[%s, %s]!", style_reg (styler, base), style_imm (styler, "#%d", opnd->addr.offset.imm)); The aarch64_print_operand function is also called from the assembler to aid in printing diagnostic messages. Right now I have no plans to add styling to the assembler output, and so, the callback function used in the assembler ignores the styling information and just returns an plain string. I've used the source files in gas/testsuite/gas/aarch64/ for testing, and have manually gone through and checked that the styling looks reasonable, however, I'm not an AArch64 expert, so it is possible that the odd piece is styled incorrectly. Please point out any mistakes I've made. With objdump disassembler color turned off, there should be no change in the output after this commit.
2022-04-28 20:31:07 +08:00
snprintf (buf, size, "%s",
style_reg (styler, "p%d.%s", opnd->reg.regno,
aarch64_get_qualifier_name (opnd->qualifier)));
[AArch64][SVE 21/32] Add Zn and Pn registers This patch adds the Zn and Pn registers, and associated fields and operands. include/ * opcode/aarch64.h (AARCH64_OPND_CLASS_SVE_REG): New aarch64_operand_class. (AARCH64_OPND_CLASS_PRED_REG): Likewise. (AARCH64_OPND_SVE_Pd, AARCH64_OPND_SVE_Pg3, AARCH64_OPND_SVE_Pg4_5) (AARCH64_OPND_SVE_Pg4_10, AARCH64_OPND_SVE_Pg4_16) (AARCH64_OPND_SVE_Pm, AARCH64_OPND_SVE_Pn, AARCH64_OPND_SVE_Pt) (AARCH64_OPND_SVE_Za_5, AARCH64_OPND_SVE_Za_16, AARCH64_OPND_SVE_Zd) (AARCH64_OPND_SVE_Zm_5, AARCH64_OPND_SVE_Zm_16, AARCH64_OPND_SVE_Zn) (AARCH64_OPND_SVE_Zn_INDEX, AARCH64_OPND_SVE_ZnxN) (AARCH64_OPND_SVE_Zt, AARCH64_OPND_SVE_ZtxN): New aarch64_opnds. opcodes/ * aarch64-tbl.h (AARCH64_OPERANDS): Add entries for new SVE operands. * aarch64-opc.h (FLD_SVE_Pd, FLD_SVE_Pg3, FLD_SVE_Pg4_5) (FLD_SVE_Pg4_10, FLD_SVE_Pg4_16, FLD_SVE_Pm, FLD_SVE_Pn, FLD_SVE_Pt) (FLD_SVE_Za_5, FLD_SVE_Za_16, FLD_SVE_Zd, FLD_SVE_Zm_5, FLD_SVE_Zm_16) (FLD_SVE_Zn, FLD_SVE_Zt, FLD_SVE_tzsh): New aarch64_field_kinds. * aarch64-opc.c (fields): Add corresponding entries here. (operand_general_constraint_met_p): Check that SVE register lists have the correct length. Check the ranges of SVE index registers. Check for cases where p8-p15 are used in 3-bit predicate fields. (aarch64_print_operand): Handle the new SVE operands. * aarch64-opc-2.c: Regenerate. * aarch64-asm.h (ins_sve_index, ins_sve_reglist): New inserters. * aarch64-asm.c (aarch64_ins_sve_index): New function. (aarch64_ins_sve_reglist): Likewise. * aarch64-asm-2.c: Regenerate. * aarch64-dis.h (ext_sve_index, ext_sve_reglist): New extractors. * aarch64-dis.c (aarch64_ext_sve_index): New function. (aarch64_ext_sve_reglist): Likewise. * aarch64-dis-2.c: Regenerate. gas/ * config/tc-aarch64.c (NTA_HASVARWIDTH): New macro. (AARCH64_REG_TYPES): Add ZN and PN. (get_reg_expected_msg): Handle them. (parse_vector_type_for_operand): Add a reg_type parameter. Skip the width for Zn and Pn registers. (parse_typed_reg): Extend vector handling to Zn and Pn. Update the call to parse_vector_type_for_operand. Set HASVARTYPE for Zn and Pn, expecting the width to be 0. (parse_vector_reg_list): Restrict error about [BHSD]nn operands to REG_TYPE_VN. (vectype_to_qualifier): Use S_[BHSD] qualifiers for NTA_HASVARWIDTH. (parse_operands): Handle the new Zn and Pn operands. (REGSET16): New macro, split out from... (REGSET31): ...here. (reg_names): Add Zn and Pn entries.
2016-09-21 23:53:54 +08:00
break;
case AARCH64_OPND_SVE_PNd:
case AARCH64_OPND_SVE_PNg4_10:
case AARCH64_OPND_SVE_PNn:
case AARCH64_OPND_SVE_PNt:
case AARCH64_OPND_SME_PNd3:
case AARCH64_OPND_SME_PNg3:
case AARCH64_OPND_SME_PNn:
if (opnd->qualifier == AARCH64_OPND_QLF_NIL)
snprintf (buf, size, "%s",
style_reg (styler, "pn%d", opnd->reg.regno));
else if (opnd->qualifier == AARCH64_OPND_QLF_P_Z
|| opnd->qualifier == AARCH64_OPND_QLF_P_M)
snprintf (buf, size, "%s",
style_reg (styler, "pn%d/%s", opnd->reg.regno,
aarch64_get_qualifier_name (opnd->qualifier)));
else
snprintf (buf, size, "%s",
style_reg (styler, "pn%d.%s", opnd->reg.regno,
aarch64_get_qualifier_name (opnd->qualifier)));
break;
case AARCH64_OPND_SME_Pdx2:
case AARCH64_OPND_SME_PdxN:
print_register_list (buf, size, opnd, "p", styler);
break;
case AARCH64_OPND_SME_PNn3_INDEX1:
case AARCH64_OPND_SME_PNn3_INDEX2:
snprintf (buf, size, "%s[%s]",
style_reg (styler, "pn%d", opnd->reglane.regno),
style_imm (styler, "%" PRIi64, opnd->reglane.index));
break;
[AArch64][SVE 21/32] Add Zn and Pn registers This patch adds the Zn and Pn registers, and associated fields and operands. include/ * opcode/aarch64.h (AARCH64_OPND_CLASS_SVE_REG): New aarch64_operand_class. (AARCH64_OPND_CLASS_PRED_REG): Likewise. (AARCH64_OPND_SVE_Pd, AARCH64_OPND_SVE_Pg3, AARCH64_OPND_SVE_Pg4_5) (AARCH64_OPND_SVE_Pg4_10, AARCH64_OPND_SVE_Pg4_16) (AARCH64_OPND_SVE_Pm, AARCH64_OPND_SVE_Pn, AARCH64_OPND_SVE_Pt) (AARCH64_OPND_SVE_Za_5, AARCH64_OPND_SVE_Za_16, AARCH64_OPND_SVE_Zd) (AARCH64_OPND_SVE_Zm_5, AARCH64_OPND_SVE_Zm_16, AARCH64_OPND_SVE_Zn) (AARCH64_OPND_SVE_Zn_INDEX, AARCH64_OPND_SVE_ZnxN) (AARCH64_OPND_SVE_Zt, AARCH64_OPND_SVE_ZtxN): New aarch64_opnds. opcodes/ * aarch64-tbl.h (AARCH64_OPERANDS): Add entries for new SVE operands. * aarch64-opc.h (FLD_SVE_Pd, FLD_SVE_Pg3, FLD_SVE_Pg4_5) (FLD_SVE_Pg4_10, FLD_SVE_Pg4_16, FLD_SVE_Pm, FLD_SVE_Pn, FLD_SVE_Pt) (FLD_SVE_Za_5, FLD_SVE_Za_16, FLD_SVE_Zd, FLD_SVE_Zm_5, FLD_SVE_Zm_16) (FLD_SVE_Zn, FLD_SVE_Zt, FLD_SVE_tzsh): New aarch64_field_kinds. * aarch64-opc.c (fields): Add corresponding entries here. (operand_general_constraint_met_p): Check that SVE register lists have the correct length. Check the ranges of SVE index registers. Check for cases where p8-p15 are used in 3-bit predicate fields. (aarch64_print_operand): Handle the new SVE operands. * aarch64-opc-2.c: Regenerate. * aarch64-asm.h (ins_sve_index, ins_sve_reglist): New inserters. * aarch64-asm.c (aarch64_ins_sve_index): New function. (aarch64_ins_sve_reglist): Likewise. * aarch64-asm-2.c: Regenerate. * aarch64-dis.h (ext_sve_index, ext_sve_reglist): New extractors. * aarch64-dis.c (aarch64_ext_sve_index): New function. (aarch64_ext_sve_reglist): Likewise. * aarch64-dis-2.c: Regenerate. gas/ * config/tc-aarch64.c (NTA_HASVARWIDTH): New macro. (AARCH64_REG_TYPES): Add ZN and PN. (get_reg_expected_msg): Handle them. (parse_vector_type_for_operand): Add a reg_type parameter. Skip the width for Zn and Pn registers. (parse_typed_reg): Extend vector handling to Zn and Pn. Update the call to parse_vector_type_for_operand. Set HASVARTYPE for Zn and Pn, expecting the width to be 0. (parse_vector_reg_list): Restrict error about [BHSD]nn operands to REG_TYPE_VN. (vectype_to_qualifier): Use S_[BHSD] qualifiers for NTA_HASVARWIDTH. (parse_operands): Handle the new Zn and Pn operands. (REGSET16): New macro, split out from... (REGSET31): ...here. (reg_names): Add Zn and Pn entries.
2016-09-21 23:53:54 +08:00
case AARCH64_OPND_SVE_Za_5:
case AARCH64_OPND_SVE_Za_16:
case AARCH64_OPND_SVE_Zd:
case AARCH64_OPND_SVE_Zm_5:
case AARCH64_OPND_SVE_Zm_16:
case AARCH64_OPND_SVE_Zn:
case AARCH64_OPND_SVE_Zt:
case AARCH64_OPND_SME_Zm:
[AArch64][SVE 21/32] Add Zn and Pn registers This patch adds the Zn and Pn registers, and associated fields and operands. include/ * opcode/aarch64.h (AARCH64_OPND_CLASS_SVE_REG): New aarch64_operand_class. (AARCH64_OPND_CLASS_PRED_REG): Likewise. (AARCH64_OPND_SVE_Pd, AARCH64_OPND_SVE_Pg3, AARCH64_OPND_SVE_Pg4_5) (AARCH64_OPND_SVE_Pg4_10, AARCH64_OPND_SVE_Pg4_16) (AARCH64_OPND_SVE_Pm, AARCH64_OPND_SVE_Pn, AARCH64_OPND_SVE_Pt) (AARCH64_OPND_SVE_Za_5, AARCH64_OPND_SVE_Za_16, AARCH64_OPND_SVE_Zd) (AARCH64_OPND_SVE_Zm_5, AARCH64_OPND_SVE_Zm_16, AARCH64_OPND_SVE_Zn) (AARCH64_OPND_SVE_Zn_INDEX, AARCH64_OPND_SVE_ZnxN) (AARCH64_OPND_SVE_Zt, AARCH64_OPND_SVE_ZtxN): New aarch64_opnds. opcodes/ * aarch64-tbl.h (AARCH64_OPERANDS): Add entries for new SVE operands. * aarch64-opc.h (FLD_SVE_Pd, FLD_SVE_Pg3, FLD_SVE_Pg4_5) (FLD_SVE_Pg4_10, FLD_SVE_Pg4_16, FLD_SVE_Pm, FLD_SVE_Pn, FLD_SVE_Pt) (FLD_SVE_Za_5, FLD_SVE_Za_16, FLD_SVE_Zd, FLD_SVE_Zm_5, FLD_SVE_Zm_16) (FLD_SVE_Zn, FLD_SVE_Zt, FLD_SVE_tzsh): New aarch64_field_kinds. * aarch64-opc.c (fields): Add corresponding entries here. (operand_general_constraint_met_p): Check that SVE register lists have the correct length. Check the ranges of SVE index registers. Check for cases where p8-p15 are used in 3-bit predicate fields. (aarch64_print_operand): Handle the new SVE operands. * aarch64-opc-2.c: Regenerate. * aarch64-asm.h (ins_sve_index, ins_sve_reglist): New inserters. * aarch64-asm.c (aarch64_ins_sve_index): New function. (aarch64_ins_sve_reglist): Likewise. * aarch64-asm-2.c: Regenerate. * aarch64-dis.h (ext_sve_index, ext_sve_reglist): New extractors. * aarch64-dis.c (aarch64_ext_sve_index): New function. (aarch64_ext_sve_reglist): Likewise. * aarch64-dis-2.c: Regenerate. gas/ * config/tc-aarch64.c (NTA_HASVARWIDTH): New macro. (AARCH64_REG_TYPES): Add ZN and PN. (get_reg_expected_msg): Handle them. (parse_vector_type_for_operand): Add a reg_type parameter. Skip the width for Zn and Pn registers. (parse_typed_reg): Extend vector handling to Zn and Pn. Update the call to parse_vector_type_for_operand. Set HASVARTYPE for Zn and Pn, expecting the width to be 0. (parse_vector_reg_list): Restrict error about [BHSD]nn operands to REG_TYPE_VN. (vectype_to_qualifier): Use S_[BHSD] qualifiers for NTA_HASVARWIDTH. (parse_operands): Handle the new Zn and Pn operands. (REGSET16): New macro, split out from... (REGSET31): ...here. (reg_names): Add Zn and Pn entries.
2016-09-21 23:53:54 +08:00
if (opnd->qualifier == AARCH64_OPND_QLF_NIL)
libopcodes/aarch64: add support for disassembler styling This commit enables disassembler styling for AArch64. After this commit it is possible to have objdump style AArch64 disassembler output (using --disassembler-color option). Once the required GDB patches are merged, GDB will also style the disassembler output. The changes to support styling are mostly split between two files opcodes/aarch64-dis.c and opcodes/aarch64-opc.c. The entry point for the AArch64 disassembler can be found in aarch64-dis.c, this file handles printing the instruction mnemonics, and assembler directives (e.g. '.byte', '.word', etc). Some operands, mostly relating to assembler directives are also printed from this file. This commit changes all of this to pass through suitable styling information. However, for most "normal" instructions, the instruction operands are printed using a two step process. From aarch64-dis.c, in the print_operands function, the function aarch64_print_operand is called, this function is in aarch64-opc.c, and converts an instruction operand into a string. Then, back in print_operands (aarch64-dis.c), the operand string is printed. Unfortunately, the string returned by aarch64_print_operand can be quite complex, it will include syntax elements, like '[' and ']', in addition to register names and immediate values. In some cases, a single operand will expand into what will appear (to the user) as multiple operands separated with a ','. This makes the task of styling more complex, all these different components need to by styled differently, so we need to get the styling information out of aarch64_print_operand in some way. The solution that I propose here is similar to the solution that I used for the i386 disassembler. Currently, aarch64_print_operand uses snprintf to write the operand text into a buffer provided by the caller. What I propose is that we pass an extra argument to the aarch64_print_operand function, this argument will be a structure, the structure contains a callback function and some state. When aarch64_print_operand needs to format part of its output this can be done by using the callback function within the new structure, this callback returns a string with special embedded markers that indicate which mode should be used for each piece of text. Back in aarch64-dis.c we can spot these special style markers and use this to split the disassembler output up and apply the correct style to each piece. To make aarch64-opc.c clearer a series of new static functions have been added, e.g. 'style_reg', 'style_imm', etc. Each of these functions formats a piece of text in a different style, 'register' and 'immediate' in this case. Here's an example taken from aarch64-opc.c of the new functions in use: snprintf (buf, size, "[%s, %s]!", style_reg (styler, base), style_imm (styler, "#%d", opnd->addr.offset.imm)); The aarch64_print_operand function is also called from the assembler to aid in printing diagnostic messages. Right now I have no plans to add styling to the assembler output, and so, the callback function used in the assembler ignores the styling information and just returns an plain string. I've used the source files in gas/testsuite/gas/aarch64/ for testing, and have manually gone through and checked that the styling looks reasonable, however, I'm not an AArch64 expert, so it is possible that the odd piece is styled incorrectly. Please point out any mistakes I've made. With objdump disassembler color turned off, there should be no change in the output after this commit.
2022-04-28 20:31:07 +08:00
snprintf (buf, size, "%s", style_reg (styler, "z%d", opnd->reg.regno));
[AArch64][SVE 21/32] Add Zn and Pn registers This patch adds the Zn and Pn registers, and associated fields and operands. include/ * opcode/aarch64.h (AARCH64_OPND_CLASS_SVE_REG): New aarch64_operand_class. (AARCH64_OPND_CLASS_PRED_REG): Likewise. (AARCH64_OPND_SVE_Pd, AARCH64_OPND_SVE_Pg3, AARCH64_OPND_SVE_Pg4_5) (AARCH64_OPND_SVE_Pg4_10, AARCH64_OPND_SVE_Pg4_16) (AARCH64_OPND_SVE_Pm, AARCH64_OPND_SVE_Pn, AARCH64_OPND_SVE_Pt) (AARCH64_OPND_SVE_Za_5, AARCH64_OPND_SVE_Za_16, AARCH64_OPND_SVE_Zd) (AARCH64_OPND_SVE_Zm_5, AARCH64_OPND_SVE_Zm_16, AARCH64_OPND_SVE_Zn) (AARCH64_OPND_SVE_Zn_INDEX, AARCH64_OPND_SVE_ZnxN) (AARCH64_OPND_SVE_Zt, AARCH64_OPND_SVE_ZtxN): New aarch64_opnds. opcodes/ * aarch64-tbl.h (AARCH64_OPERANDS): Add entries for new SVE operands. * aarch64-opc.h (FLD_SVE_Pd, FLD_SVE_Pg3, FLD_SVE_Pg4_5) (FLD_SVE_Pg4_10, FLD_SVE_Pg4_16, FLD_SVE_Pm, FLD_SVE_Pn, FLD_SVE_Pt) (FLD_SVE_Za_5, FLD_SVE_Za_16, FLD_SVE_Zd, FLD_SVE_Zm_5, FLD_SVE_Zm_16) (FLD_SVE_Zn, FLD_SVE_Zt, FLD_SVE_tzsh): New aarch64_field_kinds. * aarch64-opc.c (fields): Add corresponding entries here. (operand_general_constraint_met_p): Check that SVE register lists have the correct length. Check the ranges of SVE index registers. Check for cases where p8-p15 are used in 3-bit predicate fields. (aarch64_print_operand): Handle the new SVE operands. * aarch64-opc-2.c: Regenerate. * aarch64-asm.h (ins_sve_index, ins_sve_reglist): New inserters. * aarch64-asm.c (aarch64_ins_sve_index): New function. (aarch64_ins_sve_reglist): Likewise. * aarch64-asm-2.c: Regenerate. * aarch64-dis.h (ext_sve_index, ext_sve_reglist): New extractors. * aarch64-dis.c (aarch64_ext_sve_index): New function. (aarch64_ext_sve_reglist): Likewise. * aarch64-dis-2.c: Regenerate. gas/ * config/tc-aarch64.c (NTA_HASVARWIDTH): New macro. (AARCH64_REG_TYPES): Add ZN and PN. (get_reg_expected_msg): Handle them. (parse_vector_type_for_operand): Add a reg_type parameter. Skip the width for Zn and Pn registers. (parse_typed_reg): Extend vector handling to Zn and Pn. Update the call to parse_vector_type_for_operand. Set HASVARTYPE for Zn and Pn, expecting the width to be 0. (parse_vector_reg_list): Restrict error about [BHSD]nn operands to REG_TYPE_VN. (vectype_to_qualifier): Use S_[BHSD] qualifiers for NTA_HASVARWIDTH. (parse_operands): Handle the new Zn and Pn operands. (REGSET16): New macro, split out from... (REGSET31): ...here. (reg_names): Add Zn and Pn entries.
2016-09-21 23:53:54 +08:00
else
libopcodes/aarch64: add support for disassembler styling This commit enables disassembler styling for AArch64. After this commit it is possible to have objdump style AArch64 disassembler output (using --disassembler-color option). Once the required GDB patches are merged, GDB will also style the disassembler output. The changes to support styling are mostly split between two files opcodes/aarch64-dis.c and opcodes/aarch64-opc.c. The entry point for the AArch64 disassembler can be found in aarch64-dis.c, this file handles printing the instruction mnemonics, and assembler directives (e.g. '.byte', '.word', etc). Some operands, mostly relating to assembler directives are also printed from this file. This commit changes all of this to pass through suitable styling information. However, for most "normal" instructions, the instruction operands are printed using a two step process. From aarch64-dis.c, in the print_operands function, the function aarch64_print_operand is called, this function is in aarch64-opc.c, and converts an instruction operand into a string. Then, back in print_operands (aarch64-dis.c), the operand string is printed. Unfortunately, the string returned by aarch64_print_operand can be quite complex, it will include syntax elements, like '[' and ']', in addition to register names and immediate values. In some cases, a single operand will expand into what will appear (to the user) as multiple operands separated with a ','. This makes the task of styling more complex, all these different components need to by styled differently, so we need to get the styling information out of aarch64_print_operand in some way. The solution that I propose here is similar to the solution that I used for the i386 disassembler. Currently, aarch64_print_operand uses snprintf to write the operand text into a buffer provided by the caller. What I propose is that we pass an extra argument to the aarch64_print_operand function, this argument will be a structure, the structure contains a callback function and some state. When aarch64_print_operand needs to format part of its output this can be done by using the callback function within the new structure, this callback returns a string with special embedded markers that indicate which mode should be used for each piece of text. Back in aarch64-dis.c we can spot these special style markers and use this to split the disassembler output up and apply the correct style to each piece. To make aarch64-opc.c clearer a series of new static functions have been added, e.g. 'style_reg', 'style_imm', etc. Each of these functions formats a piece of text in a different style, 'register' and 'immediate' in this case. Here's an example taken from aarch64-opc.c of the new functions in use: snprintf (buf, size, "[%s, %s]!", style_reg (styler, base), style_imm (styler, "#%d", opnd->addr.offset.imm)); The aarch64_print_operand function is also called from the assembler to aid in printing diagnostic messages. Right now I have no plans to add styling to the assembler output, and so, the callback function used in the assembler ignores the styling information and just returns an plain string. I've used the source files in gas/testsuite/gas/aarch64/ for testing, and have manually gone through and checked that the styling looks reasonable, however, I'm not an AArch64 expert, so it is possible that the odd piece is styled incorrectly. Please point out any mistakes I've made. With objdump disassembler color turned off, there should be no change in the output after this commit.
2022-04-28 20:31:07 +08:00
snprintf (buf, size, "%s",
style_reg (styler, "z%d.%s", opnd->reg.regno,
aarch64_get_qualifier_name (opnd->qualifier)));
[AArch64][SVE 21/32] Add Zn and Pn registers This patch adds the Zn and Pn registers, and associated fields and operands. include/ * opcode/aarch64.h (AARCH64_OPND_CLASS_SVE_REG): New aarch64_operand_class. (AARCH64_OPND_CLASS_PRED_REG): Likewise. (AARCH64_OPND_SVE_Pd, AARCH64_OPND_SVE_Pg3, AARCH64_OPND_SVE_Pg4_5) (AARCH64_OPND_SVE_Pg4_10, AARCH64_OPND_SVE_Pg4_16) (AARCH64_OPND_SVE_Pm, AARCH64_OPND_SVE_Pn, AARCH64_OPND_SVE_Pt) (AARCH64_OPND_SVE_Za_5, AARCH64_OPND_SVE_Za_16, AARCH64_OPND_SVE_Zd) (AARCH64_OPND_SVE_Zm_5, AARCH64_OPND_SVE_Zm_16, AARCH64_OPND_SVE_Zn) (AARCH64_OPND_SVE_Zn_INDEX, AARCH64_OPND_SVE_ZnxN) (AARCH64_OPND_SVE_Zt, AARCH64_OPND_SVE_ZtxN): New aarch64_opnds. opcodes/ * aarch64-tbl.h (AARCH64_OPERANDS): Add entries for new SVE operands. * aarch64-opc.h (FLD_SVE_Pd, FLD_SVE_Pg3, FLD_SVE_Pg4_5) (FLD_SVE_Pg4_10, FLD_SVE_Pg4_16, FLD_SVE_Pm, FLD_SVE_Pn, FLD_SVE_Pt) (FLD_SVE_Za_5, FLD_SVE_Za_16, FLD_SVE_Zd, FLD_SVE_Zm_5, FLD_SVE_Zm_16) (FLD_SVE_Zn, FLD_SVE_Zt, FLD_SVE_tzsh): New aarch64_field_kinds. * aarch64-opc.c (fields): Add corresponding entries here. (operand_general_constraint_met_p): Check that SVE register lists have the correct length. Check the ranges of SVE index registers. Check for cases where p8-p15 are used in 3-bit predicate fields. (aarch64_print_operand): Handle the new SVE operands. * aarch64-opc-2.c: Regenerate. * aarch64-asm.h (ins_sve_index, ins_sve_reglist): New inserters. * aarch64-asm.c (aarch64_ins_sve_index): New function. (aarch64_ins_sve_reglist): Likewise. * aarch64-asm-2.c: Regenerate. * aarch64-dis.h (ext_sve_index, ext_sve_reglist): New extractors. * aarch64-dis.c (aarch64_ext_sve_index): New function. (aarch64_ext_sve_reglist): Likewise. * aarch64-dis-2.c: Regenerate. gas/ * config/tc-aarch64.c (NTA_HASVARWIDTH): New macro. (AARCH64_REG_TYPES): Add ZN and PN. (get_reg_expected_msg): Handle them. (parse_vector_type_for_operand): Add a reg_type parameter. Skip the width for Zn and Pn registers. (parse_typed_reg): Extend vector handling to Zn and Pn. Update the call to parse_vector_type_for_operand. Set HASVARTYPE for Zn and Pn, expecting the width to be 0. (parse_vector_reg_list): Restrict error about [BHSD]nn operands to REG_TYPE_VN. (vectype_to_qualifier): Use S_[BHSD] qualifiers for NTA_HASVARWIDTH. (parse_operands): Handle the new Zn and Pn operands. (REGSET16): New macro, split out from... (REGSET31): ...here. (reg_names): Add Zn and Pn entries.
2016-09-21 23:53:54 +08:00
break;
case AARCH64_OPND_SVE_ZnxN:
case AARCH64_OPND_SVE_ZtxN:
case AARCH64_OPND_SME_Zdnx2:
case AARCH64_OPND_SME_Zdnx4:
case AARCH64_OPND_SME_Zmx2:
case AARCH64_OPND_SME_Zmx4:
case AARCH64_OPND_SME_Znx2:
case AARCH64_OPND_SME_Znx4:
case AARCH64_OPND_SME_Ztx2_STRIDED:
case AARCH64_OPND_SME_Ztx4_STRIDED:
case AARCH64_OPND_SME_Zt2:
case AARCH64_OPND_SME_Zt3:
case AARCH64_OPND_SME_Zt4:
libopcodes/aarch64: add support for disassembler styling This commit enables disassembler styling for AArch64. After this commit it is possible to have objdump style AArch64 disassembler output (using --disassembler-color option). Once the required GDB patches are merged, GDB will also style the disassembler output. The changes to support styling are mostly split between two files opcodes/aarch64-dis.c and opcodes/aarch64-opc.c. The entry point for the AArch64 disassembler can be found in aarch64-dis.c, this file handles printing the instruction mnemonics, and assembler directives (e.g. '.byte', '.word', etc). Some operands, mostly relating to assembler directives are also printed from this file. This commit changes all of this to pass through suitable styling information. However, for most "normal" instructions, the instruction operands are printed using a two step process. From aarch64-dis.c, in the print_operands function, the function aarch64_print_operand is called, this function is in aarch64-opc.c, and converts an instruction operand into a string. Then, back in print_operands (aarch64-dis.c), the operand string is printed. Unfortunately, the string returned by aarch64_print_operand can be quite complex, it will include syntax elements, like '[' and ']', in addition to register names and immediate values. In some cases, a single operand will expand into what will appear (to the user) as multiple operands separated with a ','. This makes the task of styling more complex, all these different components need to by styled differently, so we need to get the styling information out of aarch64_print_operand in some way. The solution that I propose here is similar to the solution that I used for the i386 disassembler. Currently, aarch64_print_operand uses snprintf to write the operand text into a buffer provided by the caller. What I propose is that we pass an extra argument to the aarch64_print_operand function, this argument will be a structure, the structure contains a callback function and some state. When aarch64_print_operand needs to format part of its output this can be done by using the callback function within the new structure, this callback returns a string with special embedded markers that indicate which mode should be used for each piece of text. Back in aarch64-dis.c we can spot these special style markers and use this to split the disassembler output up and apply the correct style to each piece. To make aarch64-opc.c clearer a series of new static functions have been added, e.g. 'style_reg', 'style_imm', etc. Each of these functions formats a piece of text in a different style, 'register' and 'immediate' in this case. Here's an example taken from aarch64-opc.c of the new functions in use: snprintf (buf, size, "[%s, %s]!", style_reg (styler, base), style_imm (styler, "#%d", opnd->addr.offset.imm)); The aarch64_print_operand function is also called from the assembler to aid in printing diagnostic messages. Right now I have no plans to add styling to the assembler output, and so, the callback function used in the assembler ignores the styling information and just returns an plain string. I've used the source files in gas/testsuite/gas/aarch64/ for testing, and have manually gone through and checked that the styling looks reasonable, however, I'm not an AArch64 expert, so it is possible that the odd piece is styled incorrectly. Please point out any mistakes I've made. With objdump disassembler color turned off, there should be no change in the output after this commit.
2022-04-28 20:31:07 +08:00
print_register_list (buf, size, opnd, "z", styler);
[AArch64][SVE 21/32] Add Zn and Pn registers This patch adds the Zn and Pn registers, and associated fields and operands. include/ * opcode/aarch64.h (AARCH64_OPND_CLASS_SVE_REG): New aarch64_operand_class. (AARCH64_OPND_CLASS_PRED_REG): Likewise. (AARCH64_OPND_SVE_Pd, AARCH64_OPND_SVE_Pg3, AARCH64_OPND_SVE_Pg4_5) (AARCH64_OPND_SVE_Pg4_10, AARCH64_OPND_SVE_Pg4_16) (AARCH64_OPND_SVE_Pm, AARCH64_OPND_SVE_Pn, AARCH64_OPND_SVE_Pt) (AARCH64_OPND_SVE_Za_5, AARCH64_OPND_SVE_Za_16, AARCH64_OPND_SVE_Zd) (AARCH64_OPND_SVE_Zm_5, AARCH64_OPND_SVE_Zm_16, AARCH64_OPND_SVE_Zn) (AARCH64_OPND_SVE_Zn_INDEX, AARCH64_OPND_SVE_ZnxN) (AARCH64_OPND_SVE_Zt, AARCH64_OPND_SVE_ZtxN): New aarch64_opnds. opcodes/ * aarch64-tbl.h (AARCH64_OPERANDS): Add entries for new SVE operands. * aarch64-opc.h (FLD_SVE_Pd, FLD_SVE_Pg3, FLD_SVE_Pg4_5) (FLD_SVE_Pg4_10, FLD_SVE_Pg4_16, FLD_SVE_Pm, FLD_SVE_Pn, FLD_SVE_Pt) (FLD_SVE_Za_5, FLD_SVE_Za_16, FLD_SVE_Zd, FLD_SVE_Zm_5, FLD_SVE_Zm_16) (FLD_SVE_Zn, FLD_SVE_Zt, FLD_SVE_tzsh): New aarch64_field_kinds. * aarch64-opc.c (fields): Add corresponding entries here. (operand_general_constraint_met_p): Check that SVE register lists have the correct length. Check the ranges of SVE index registers. Check for cases where p8-p15 are used in 3-bit predicate fields. (aarch64_print_operand): Handle the new SVE operands. * aarch64-opc-2.c: Regenerate. * aarch64-asm.h (ins_sve_index, ins_sve_reglist): New inserters. * aarch64-asm.c (aarch64_ins_sve_index): New function. (aarch64_ins_sve_reglist): Likewise. * aarch64-asm-2.c: Regenerate. * aarch64-dis.h (ext_sve_index, ext_sve_reglist): New extractors. * aarch64-dis.c (aarch64_ext_sve_index): New function. (aarch64_ext_sve_reglist): Likewise. * aarch64-dis-2.c: Regenerate. gas/ * config/tc-aarch64.c (NTA_HASVARWIDTH): New macro. (AARCH64_REG_TYPES): Add ZN and PN. (get_reg_expected_msg): Handle them. (parse_vector_type_for_operand): Add a reg_type parameter. Skip the width for Zn and Pn registers. (parse_typed_reg): Extend vector handling to Zn and Pn. Update the call to parse_vector_type_for_operand. Set HASVARTYPE for Zn and Pn, expecting the width to be 0. (parse_vector_reg_list): Restrict error about [BHSD]nn operands to REG_TYPE_VN. (vectype_to_qualifier): Use S_[BHSD] qualifiers for NTA_HASVARWIDTH. (parse_operands): Handle the new Zn and Pn operands. (REGSET16): New macro, split out from... (REGSET31): ...here. (reg_names): Add Zn and Pn entries.
2016-09-21 23:53:54 +08:00
break;
[AArch64] Additional SVE instructions This patch supports some additions to the SVE architecture prior to its public release. include/ * opcode/aarch64.h (AARCH64_OPND_SVE_ADDR_RI_S4x16) (AARCH64_OPND_SVE_IMM_ROT1, AARCH64_OPND_SVE_IMM_ROT2) (AARCH64_OPND_SVE_Zm3_INDEX, AARCH64_OPND_SVE_Zm3_22_INDEX) (AARCH64_OPND_SVE_Zm4_INDEX): New aarch64_opnds. opcodes/ * aarch64-tbl.h (OP_SVE_HMH, OP_SVE_VMU_HSD, OP_SVE_VMVU_HSD) (OP_SVE_VMVV_HSD, OP_SVE_VMVVU_HSD, OP_SVE_VM_HSD, OP_SVE_VUVV_HSD) (OP_SVE_VUV_HSD, OP_SVE_VU_HSD, OP_SVE_VVVU_H, OP_SVE_VVVU_S) (OP_SVE_VVVU_HSD, OP_SVE_VVV_D, OP_SVE_VVV_D_H, OP_SVE_VVV_H) (OP_SVE_VVV_HSD, OP_SVE_VVV_S, OP_SVE_VVV_S_B, OP_SVE_VVV_SD_BH) (OP_SVE_VV_BHSDQ, OP_SVE_VV_HSD, OP_SVE_VZVV_HSD, OP_SVE_VZV_HSD) (OP_SVE_V_HSD): New macros. (OP_SVE_VMU_SD, OP_SVE_VMVU_SD, OP_SVE_VM_SD, OP_SVE_VUVV_SD) (OP_SVE_VU_SD, OP_SVE_VVVU_SD, OP_SVE_VVV_SD, OP_SVE_VZVV_SD) (OP_SVE_VZV_SD, OP_SVE_V_SD): Delete. (aarch64_opcode_table): Add new SVE instructions. (aarch64_opcode_table): Use imm_rotate{1,2} instead of imm_rotate for rotation operands. Add new SVE operands. * aarch64-asm.h (ins_sve_addr_ri_s4): New inserter. (ins_sve_quad_index): Likewise. (ins_imm_rotate): Split into... (ins_imm_rotate1, ins_imm_rotate2): ...these two inserters. * aarch64-asm.c (aarch64_ins_imm_rotate): Split into... (aarch64_ins_imm_rotate1, aarch64_ins_imm_rotate2): ...these two functions. (aarch64_ins_sve_addr_ri_s4): New function. (aarch64_ins_sve_quad_index): Likewise. (do_misc_encoding): Handle "MOV Zn.Q, Qm". * aarch64-asm-2.c: Regenerate. * aarch64-dis.h (ext_sve_addr_ri_s4): New extractor. (ext_sve_quad_index): Likewise. (ext_imm_rotate): Split into... (ext_imm_rotate1, ext_imm_rotate2): ...these two extractors. * aarch64-dis.c (aarch64_ext_imm_rotate): Split into... (aarch64_ext_imm_rotate1, aarch64_ext_imm_rotate2): ...these two functions. (aarch64_ext_sve_addr_ri_s4): New function. (aarch64_ext_sve_quad_index): Likewise. (aarch64_ext_sve_index): Allow quad indices. (do_misc_decoding): Likewise. * aarch64-dis-2.c: Regenerate. * aarch64-opc.h (FLD_SVE_i3h, FLD_SVE_rot1, FLD_SVE_rot2): New aarch64_field_kinds. (OPD_F_OD_MASK): Widen by one bit. (OPD_F_NO_ZR): Bump accordingly. (get_operand_field_width): New function. * aarch64-opc.c (fields): Add new SVE fields. (operand_general_constraint_met_p): Handle new SVE operands. (aarch64_print_operand): Likewise. * aarch64-opc-2.c: Regenerate. gas/ * doc/c-aarch64.texi: Document that sve implies fp16, simd and compnum. * config/tc-aarch64.c (parse_vector_type_for_operand): Allow .q to be used with SVE registers. (parse_operands): Handle new SVE operands. (aarch64_features): Make "sve" require F16 rather than FP. Also require COMPNUM. * testsuite/gas/aarch64/sve.s: Add tests for new instructions. Include compnum tests. * testsuite/gas/aarch64/sve.d: Update accordingly. * testsuite/gas/aarch64/sve-invalid.s: Add tests for new instructions. * testsuite/gas/aarch64/sve-invalid.l: Update accordingly. Also update expected output for new FMOV and MOV alternatives.
2017-02-25 02:29:00 +08:00
case AARCH64_OPND_SVE_Zm3_INDEX:
case AARCH64_OPND_SVE_Zm3_22_INDEX:
case AARCH64_OPND_SVE_Zm3_19_INDEX:
case AARCH64_OPND_SVE_Zm3_11_INDEX:
case AARCH64_OPND_SVE_Zm4_11_INDEX:
[AArch64] Additional SVE instructions This patch supports some additions to the SVE architecture prior to its public release. include/ * opcode/aarch64.h (AARCH64_OPND_SVE_ADDR_RI_S4x16) (AARCH64_OPND_SVE_IMM_ROT1, AARCH64_OPND_SVE_IMM_ROT2) (AARCH64_OPND_SVE_Zm3_INDEX, AARCH64_OPND_SVE_Zm3_22_INDEX) (AARCH64_OPND_SVE_Zm4_INDEX): New aarch64_opnds. opcodes/ * aarch64-tbl.h (OP_SVE_HMH, OP_SVE_VMU_HSD, OP_SVE_VMVU_HSD) (OP_SVE_VMVV_HSD, OP_SVE_VMVVU_HSD, OP_SVE_VM_HSD, OP_SVE_VUVV_HSD) (OP_SVE_VUV_HSD, OP_SVE_VU_HSD, OP_SVE_VVVU_H, OP_SVE_VVVU_S) (OP_SVE_VVVU_HSD, OP_SVE_VVV_D, OP_SVE_VVV_D_H, OP_SVE_VVV_H) (OP_SVE_VVV_HSD, OP_SVE_VVV_S, OP_SVE_VVV_S_B, OP_SVE_VVV_SD_BH) (OP_SVE_VV_BHSDQ, OP_SVE_VV_HSD, OP_SVE_VZVV_HSD, OP_SVE_VZV_HSD) (OP_SVE_V_HSD): New macros. (OP_SVE_VMU_SD, OP_SVE_VMVU_SD, OP_SVE_VM_SD, OP_SVE_VUVV_SD) (OP_SVE_VU_SD, OP_SVE_VVVU_SD, OP_SVE_VVV_SD, OP_SVE_VZVV_SD) (OP_SVE_VZV_SD, OP_SVE_V_SD): Delete. (aarch64_opcode_table): Add new SVE instructions. (aarch64_opcode_table): Use imm_rotate{1,2} instead of imm_rotate for rotation operands. Add new SVE operands. * aarch64-asm.h (ins_sve_addr_ri_s4): New inserter. (ins_sve_quad_index): Likewise. (ins_imm_rotate): Split into... (ins_imm_rotate1, ins_imm_rotate2): ...these two inserters. * aarch64-asm.c (aarch64_ins_imm_rotate): Split into... (aarch64_ins_imm_rotate1, aarch64_ins_imm_rotate2): ...these two functions. (aarch64_ins_sve_addr_ri_s4): New function. (aarch64_ins_sve_quad_index): Likewise. (do_misc_encoding): Handle "MOV Zn.Q, Qm". * aarch64-asm-2.c: Regenerate. * aarch64-dis.h (ext_sve_addr_ri_s4): New extractor. (ext_sve_quad_index): Likewise. (ext_imm_rotate): Split into... (ext_imm_rotate1, ext_imm_rotate2): ...these two extractors. * aarch64-dis.c (aarch64_ext_imm_rotate): Split into... (aarch64_ext_imm_rotate1, aarch64_ext_imm_rotate2): ...these two functions. (aarch64_ext_sve_addr_ri_s4): New function. (aarch64_ext_sve_quad_index): Likewise. (aarch64_ext_sve_index): Allow quad indices. (do_misc_decoding): Likewise. * aarch64-dis-2.c: Regenerate. * aarch64-opc.h (FLD_SVE_i3h, FLD_SVE_rot1, FLD_SVE_rot2): New aarch64_field_kinds. (OPD_F_OD_MASK): Widen by one bit. (OPD_F_NO_ZR): Bump accordingly. (get_operand_field_width): New function. * aarch64-opc.c (fields): Add new SVE fields. (operand_general_constraint_met_p): Handle new SVE operands. (aarch64_print_operand): Likewise. * aarch64-opc-2.c: Regenerate. gas/ * doc/c-aarch64.texi: Document that sve implies fp16, simd and compnum. * config/tc-aarch64.c (parse_vector_type_for_operand): Allow .q to be used with SVE registers. (parse_operands): Handle new SVE operands. (aarch64_features): Make "sve" require F16 rather than FP. Also require COMPNUM. * testsuite/gas/aarch64/sve.s: Add tests for new instructions. Include compnum tests. * testsuite/gas/aarch64/sve.d: Update accordingly. * testsuite/gas/aarch64/sve-invalid.s: Add tests for new instructions. * testsuite/gas/aarch64/sve-invalid.l: Update accordingly. Also update expected output for new FMOV and MOV alternatives.
2017-02-25 02:29:00 +08:00
case AARCH64_OPND_SVE_Zm4_INDEX:
[AArch64][SVE 21/32] Add Zn and Pn registers This patch adds the Zn and Pn registers, and associated fields and operands. include/ * opcode/aarch64.h (AARCH64_OPND_CLASS_SVE_REG): New aarch64_operand_class. (AARCH64_OPND_CLASS_PRED_REG): Likewise. (AARCH64_OPND_SVE_Pd, AARCH64_OPND_SVE_Pg3, AARCH64_OPND_SVE_Pg4_5) (AARCH64_OPND_SVE_Pg4_10, AARCH64_OPND_SVE_Pg4_16) (AARCH64_OPND_SVE_Pm, AARCH64_OPND_SVE_Pn, AARCH64_OPND_SVE_Pt) (AARCH64_OPND_SVE_Za_5, AARCH64_OPND_SVE_Za_16, AARCH64_OPND_SVE_Zd) (AARCH64_OPND_SVE_Zm_5, AARCH64_OPND_SVE_Zm_16, AARCH64_OPND_SVE_Zn) (AARCH64_OPND_SVE_Zn_INDEX, AARCH64_OPND_SVE_ZnxN) (AARCH64_OPND_SVE_Zt, AARCH64_OPND_SVE_ZtxN): New aarch64_opnds. opcodes/ * aarch64-tbl.h (AARCH64_OPERANDS): Add entries for new SVE operands. * aarch64-opc.h (FLD_SVE_Pd, FLD_SVE_Pg3, FLD_SVE_Pg4_5) (FLD_SVE_Pg4_10, FLD_SVE_Pg4_16, FLD_SVE_Pm, FLD_SVE_Pn, FLD_SVE_Pt) (FLD_SVE_Za_5, FLD_SVE_Za_16, FLD_SVE_Zd, FLD_SVE_Zm_5, FLD_SVE_Zm_16) (FLD_SVE_Zn, FLD_SVE_Zt, FLD_SVE_tzsh): New aarch64_field_kinds. * aarch64-opc.c (fields): Add corresponding entries here. (operand_general_constraint_met_p): Check that SVE register lists have the correct length. Check the ranges of SVE index registers. Check for cases where p8-p15 are used in 3-bit predicate fields. (aarch64_print_operand): Handle the new SVE operands. * aarch64-opc-2.c: Regenerate. * aarch64-asm.h (ins_sve_index, ins_sve_reglist): New inserters. * aarch64-asm.c (aarch64_ins_sve_index): New function. (aarch64_ins_sve_reglist): Likewise. * aarch64-asm-2.c: Regenerate. * aarch64-dis.h (ext_sve_index, ext_sve_reglist): New extractors. * aarch64-dis.c (aarch64_ext_sve_index): New function. (aarch64_ext_sve_reglist): Likewise. * aarch64-dis-2.c: Regenerate. gas/ * config/tc-aarch64.c (NTA_HASVARWIDTH): New macro. (AARCH64_REG_TYPES): Add ZN and PN. (get_reg_expected_msg): Handle them. (parse_vector_type_for_operand): Add a reg_type parameter. Skip the width for Zn and Pn registers. (parse_typed_reg): Extend vector handling to Zn and Pn. Update the call to parse_vector_type_for_operand. Set HASVARTYPE for Zn and Pn, expecting the width to be 0. (parse_vector_reg_list): Restrict error about [BHSD]nn operands to REG_TYPE_VN. (vectype_to_qualifier): Use S_[BHSD] qualifiers for NTA_HASVARWIDTH. (parse_operands): Handle the new Zn and Pn operands. (REGSET16): New macro, split out from... (REGSET31): ...here. (reg_names): Add Zn and Pn entries.
2016-09-21 23:53:54 +08:00
case AARCH64_OPND_SVE_Zn_INDEX:
case AARCH64_OPND_SME_Zm_INDEX1:
case AARCH64_OPND_SME_Zm_INDEX2:
case AARCH64_OPND_SME_Zm_INDEX3_1:
case AARCH64_OPND_SME_Zm_INDEX3_2:
case AARCH64_OPND_SME_Zm_INDEX3_10:
case AARCH64_OPND_SVE_Zn_5_INDEX:
case AARCH64_OPND_SME_Zm_INDEX4_1:
case AARCH64_OPND_SME_Zm_INDEX4_10:
case AARCH64_OPND_SME_Zn_INDEX1_16:
case AARCH64_OPND_SME_Zn_INDEX2_15:
case AARCH64_OPND_SME_Zn_INDEX2_16:
case AARCH64_OPND_SME_Zn_INDEX3_14:
case AARCH64_OPND_SME_Zn_INDEX3_15:
case AARCH64_OPND_SME_Zn_INDEX4_14:
case AARCH64_OPND_SVE_Zm_imm4:
libopcodes/aarch64: add support for disassembler styling This commit enables disassembler styling for AArch64. After this commit it is possible to have objdump style AArch64 disassembler output (using --disassembler-color option). Once the required GDB patches are merged, GDB will also style the disassembler output. The changes to support styling are mostly split between two files opcodes/aarch64-dis.c and opcodes/aarch64-opc.c. The entry point for the AArch64 disassembler can be found in aarch64-dis.c, this file handles printing the instruction mnemonics, and assembler directives (e.g. '.byte', '.word', etc). Some operands, mostly relating to assembler directives are also printed from this file. This commit changes all of this to pass through suitable styling information. However, for most "normal" instructions, the instruction operands are printed using a two step process. From aarch64-dis.c, in the print_operands function, the function aarch64_print_operand is called, this function is in aarch64-opc.c, and converts an instruction operand into a string. Then, back in print_operands (aarch64-dis.c), the operand string is printed. Unfortunately, the string returned by aarch64_print_operand can be quite complex, it will include syntax elements, like '[' and ']', in addition to register names and immediate values. In some cases, a single operand will expand into what will appear (to the user) as multiple operands separated with a ','. This makes the task of styling more complex, all these different components need to by styled differently, so we need to get the styling information out of aarch64_print_operand in some way. The solution that I propose here is similar to the solution that I used for the i386 disassembler. Currently, aarch64_print_operand uses snprintf to write the operand text into a buffer provided by the caller. What I propose is that we pass an extra argument to the aarch64_print_operand function, this argument will be a structure, the structure contains a callback function and some state. When aarch64_print_operand needs to format part of its output this can be done by using the callback function within the new structure, this callback returns a string with special embedded markers that indicate which mode should be used for each piece of text. Back in aarch64-dis.c we can spot these special style markers and use this to split the disassembler output up and apply the correct style to each piece. To make aarch64-opc.c clearer a series of new static functions have been added, e.g. 'style_reg', 'style_imm', etc. Each of these functions formats a piece of text in a different style, 'register' and 'immediate' in this case. Here's an example taken from aarch64-opc.c of the new functions in use: snprintf (buf, size, "[%s, %s]!", style_reg (styler, base), style_imm (styler, "#%d", opnd->addr.offset.imm)); The aarch64_print_operand function is also called from the assembler to aid in printing diagnostic messages. Right now I have no plans to add styling to the assembler output, and so, the callback function used in the assembler ignores the styling information and just returns an plain string. I've used the source files in gas/testsuite/gas/aarch64/ for testing, and have manually gone through and checked that the styling looks reasonable, however, I'm not an AArch64 expert, so it is possible that the odd piece is styled incorrectly. Please point out any mistakes I've made. With objdump disassembler color turned off, there should be no change in the output after this commit.
2022-04-28 20:31:07 +08:00
snprintf (buf, size, "%s[%s]",
(opnd->qualifier == AARCH64_OPND_QLF_NIL
? style_reg (styler, "z%d", opnd->reglane.regno)
: style_reg (styler, "z%d.%s", opnd->reglane.regno,
aarch64_get_qualifier_name (opnd->qualifier))),
libopcodes/aarch64: add support for disassembler styling This commit enables disassembler styling for AArch64. After this commit it is possible to have objdump style AArch64 disassembler output (using --disassembler-color option). Once the required GDB patches are merged, GDB will also style the disassembler output. The changes to support styling are mostly split between two files opcodes/aarch64-dis.c and opcodes/aarch64-opc.c. The entry point for the AArch64 disassembler can be found in aarch64-dis.c, this file handles printing the instruction mnemonics, and assembler directives (e.g. '.byte', '.word', etc). Some operands, mostly relating to assembler directives are also printed from this file. This commit changes all of this to pass through suitable styling information. However, for most "normal" instructions, the instruction operands are printed using a two step process. From aarch64-dis.c, in the print_operands function, the function aarch64_print_operand is called, this function is in aarch64-opc.c, and converts an instruction operand into a string. Then, back in print_operands (aarch64-dis.c), the operand string is printed. Unfortunately, the string returned by aarch64_print_operand can be quite complex, it will include syntax elements, like '[' and ']', in addition to register names and immediate values. In some cases, a single operand will expand into what will appear (to the user) as multiple operands separated with a ','. This makes the task of styling more complex, all these different components need to by styled differently, so we need to get the styling information out of aarch64_print_operand in some way. The solution that I propose here is similar to the solution that I used for the i386 disassembler. Currently, aarch64_print_operand uses snprintf to write the operand text into a buffer provided by the caller. What I propose is that we pass an extra argument to the aarch64_print_operand function, this argument will be a structure, the structure contains a callback function and some state. When aarch64_print_operand needs to format part of its output this can be done by using the callback function within the new structure, this callback returns a string with special embedded markers that indicate which mode should be used for each piece of text. Back in aarch64-dis.c we can spot these special style markers and use this to split the disassembler output up and apply the correct style to each piece. To make aarch64-opc.c clearer a series of new static functions have been added, e.g. 'style_reg', 'style_imm', etc. Each of these functions formats a piece of text in a different style, 'register' and 'immediate' in this case. Here's an example taken from aarch64-opc.c of the new functions in use: snprintf (buf, size, "[%s, %s]!", style_reg (styler, base), style_imm (styler, "#%d", opnd->addr.offset.imm)); The aarch64_print_operand function is also called from the assembler to aid in printing diagnostic messages. Right now I have no plans to add styling to the assembler output, and so, the callback function used in the assembler ignores the styling information and just returns an plain string. I've used the source files in gas/testsuite/gas/aarch64/ for testing, and have manually gone through and checked that the styling looks reasonable, however, I'm not an AArch64 expert, so it is possible that the odd piece is styled incorrectly. Please point out any mistakes I've made. With objdump disassembler color turned off, there should be no change in the output after this commit.
2022-04-28 20:31:07 +08:00
style_imm (styler, "%" PRIi64, opnd->reglane.index));
[AArch64][SVE 21/32] Add Zn and Pn registers This patch adds the Zn and Pn registers, and associated fields and operands. include/ * opcode/aarch64.h (AARCH64_OPND_CLASS_SVE_REG): New aarch64_operand_class. (AARCH64_OPND_CLASS_PRED_REG): Likewise. (AARCH64_OPND_SVE_Pd, AARCH64_OPND_SVE_Pg3, AARCH64_OPND_SVE_Pg4_5) (AARCH64_OPND_SVE_Pg4_10, AARCH64_OPND_SVE_Pg4_16) (AARCH64_OPND_SVE_Pm, AARCH64_OPND_SVE_Pn, AARCH64_OPND_SVE_Pt) (AARCH64_OPND_SVE_Za_5, AARCH64_OPND_SVE_Za_16, AARCH64_OPND_SVE_Zd) (AARCH64_OPND_SVE_Zm_5, AARCH64_OPND_SVE_Zm_16, AARCH64_OPND_SVE_Zn) (AARCH64_OPND_SVE_Zn_INDEX, AARCH64_OPND_SVE_ZnxN) (AARCH64_OPND_SVE_Zt, AARCH64_OPND_SVE_ZtxN): New aarch64_opnds. opcodes/ * aarch64-tbl.h (AARCH64_OPERANDS): Add entries for new SVE operands. * aarch64-opc.h (FLD_SVE_Pd, FLD_SVE_Pg3, FLD_SVE_Pg4_5) (FLD_SVE_Pg4_10, FLD_SVE_Pg4_16, FLD_SVE_Pm, FLD_SVE_Pn, FLD_SVE_Pt) (FLD_SVE_Za_5, FLD_SVE_Za_16, FLD_SVE_Zd, FLD_SVE_Zm_5, FLD_SVE_Zm_16) (FLD_SVE_Zn, FLD_SVE_Zt, FLD_SVE_tzsh): New aarch64_field_kinds. * aarch64-opc.c (fields): Add corresponding entries here. (operand_general_constraint_met_p): Check that SVE register lists have the correct length. Check the ranges of SVE index registers. Check for cases where p8-p15 are used in 3-bit predicate fields. (aarch64_print_operand): Handle the new SVE operands. * aarch64-opc-2.c: Regenerate. * aarch64-asm.h (ins_sve_index, ins_sve_reglist): New inserters. * aarch64-asm.c (aarch64_ins_sve_index): New function. (aarch64_ins_sve_reglist): Likewise. * aarch64-asm-2.c: Regenerate. * aarch64-dis.h (ext_sve_index, ext_sve_reglist): New extractors. * aarch64-dis.c (aarch64_ext_sve_index): New function. (aarch64_ext_sve_reglist): Likewise. * aarch64-dis-2.c: Regenerate. gas/ * config/tc-aarch64.c (NTA_HASVARWIDTH): New macro. (AARCH64_REG_TYPES): Add ZN and PN. (get_reg_expected_msg): Handle them. (parse_vector_type_for_operand): Add a reg_type parameter. Skip the width for Zn and Pn registers. (parse_typed_reg): Extend vector handling to Zn and Pn. Update the call to parse_vector_type_for_operand. Set HASVARTYPE for Zn and Pn, expecting the width to be 0. (parse_vector_reg_list): Restrict error about [BHSD]nn operands to REG_TYPE_VN. (vectype_to_qualifier): Use S_[BHSD] qualifiers for NTA_HASVARWIDTH. (parse_operands): Handle the new Zn and Pn operands. (REGSET16): New macro, split out from... (REGSET31): ...here. (reg_names): Add Zn and Pn entries.
2016-09-21 23:53:54 +08:00
break;
aarch64: [SME] Add SME instructions Patch is adding new SME matrix instructions. Please note additional instructions will be added in following patches. gas/ChangeLog: * config/tc-aarch64.c (parse_sme_zada_operand): New parser. * config/tc-aarch64.c (parse_reg_with_qual): New reg parser. * config/tc-aarch64.c (R_ZA): New egister type. (parse_operands): New parser. * testsuite/gas/aarch64/sme-illegal.d: New test. * testsuite/gas/aarch64/sme-illegal.l: New test. * testsuite/gas/aarch64/sme-illegal.s: New test. * testsuite/gas/aarch64/sme.d: New test. * testsuite/gas/aarch64/sme.s: New test. * testsuite/gas/aarch64/sme-f64.d: New test. * testsuite/gas/aarch64/sme-f64.s: New test. * testsuite/gas/aarch64/sme-i64.d: New test. * testsuite/gas/aarch64/sme-i64.s: New test. include/ChangeLog: * opcode/aarch64.h (enum aarch64_opnd): New operands AARCH64_OPND_SME_ZAda_2b, AARCH64_OPND_SME_ZAda_3b and AARCH64_OPND_SME_Pm. (enum aarch64_insn_class): New instruction class sme_misc. opcodes/ChangeLog: * aarch64-opc.c (aarch64_print_operand): Print OPND_SME_ZAda_2b and OPND_SME_ZAda_3b operands. (verify_constraints): Handle OPND_SME_Pm. * aarch64-opc.h (enum aarch64_field_kind): New bit fields FLD_SME_ZAda_2b, FLD_SME_ZAda_3b and FLD_SME_Pm. * aarch64-tbl.h (OP_SME_ZADA_PN_PM_ZN_S): New qualifier set. (OP_SME_ZADA_PN_PM_ZN_D): New qualifier. (OP_SME_ZADA_PN_PM_ZN_ZM): New qualifier. (OP_SME_ZADA_S_PM_PM_S_S): New qualifier. (OP_SME_ZADA_D_PM_PM_D_D): New qualifier. (OP_SME_ZADA_S_PM_PM_H_H): New qualifier. (OP_SME_ZADA_S_PM_PM_B_B): New qualifier. (OP_SME_ZADA_D_PM_PM_H_H): New qualifier. (SME_INSN): New instruction macro. (SME_F64_INSN): New instruction macro. (SME_I64_INSN): New instruction macro. (SME_INSNC): New instruction macro. (struct aarch64_opcode): New SME instructions. aarch64-asm-2.c: Regenerate. aarch64-dis-2.c: Regenerate. aarch64-opc-2.c: Regenerate.
2021-11-18 03:21:33 +08:00
case AARCH64_OPND_SME_ZAda_2b:
case AARCH64_OPND_SME_ZAda_3b:
libopcodes/aarch64: add support for disassembler styling This commit enables disassembler styling for AArch64. After this commit it is possible to have objdump style AArch64 disassembler output (using --disassembler-color option). Once the required GDB patches are merged, GDB will also style the disassembler output. The changes to support styling are mostly split between two files opcodes/aarch64-dis.c and opcodes/aarch64-opc.c. The entry point for the AArch64 disassembler can be found in aarch64-dis.c, this file handles printing the instruction mnemonics, and assembler directives (e.g. '.byte', '.word', etc). Some operands, mostly relating to assembler directives are also printed from this file. This commit changes all of this to pass through suitable styling information. However, for most "normal" instructions, the instruction operands are printed using a two step process. From aarch64-dis.c, in the print_operands function, the function aarch64_print_operand is called, this function is in aarch64-opc.c, and converts an instruction operand into a string. Then, back in print_operands (aarch64-dis.c), the operand string is printed. Unfortunately, the string returned by aarch64_print_operand can be quite complex, it will include syntax elements, like '[' and ']', in addition to register names and immediate values. In some cases, a single operand will expand into what will appear (to the user) as multiple operands separated with a ','. This makes the task of styling more complex, all these different components need to by styled differently, so we need to get the styling information out of aarch64_print_operand in some way. The solution that I propose here is similar to the solution that I used for the i386 disassembler. Currently, aarch64_print_operand uses snprintf to write the operand text into a buffer provided by the caller. What I propose is that we pass an extra argument to the aarch64_print_operand function, this argument will be a structure, the structure contains a callback function and some state. When aarch64_print_operand needs to format part of its output this can be done by using the callback function within the new structure, this callback returns a string with special embedded markers that indicate which mode should be used for each piece of text. Back in aarch64-dis.c we can spot these special style markers and use this to split the disassembler output up and apply the correct style to each piece. To make aarch64-opc.c clearer a series of new static functions have been added, e.g. 'style_reg', 'style_imm', etc. Each of these functions formats a piece of text in a different style, 'register' and 'immediate' in this case. Here's an example taken from aarch64-opc.c of the new functions in use: snprintf (buf, size, "[%s, %s]!", style_reg (styler, base), style_imm (styler, "#%d", opnd->addr.offset.imm)); The aarch64_print_operand function is also called from the assembler to aid in printing diagnostic messages. Right now I have no plans to add styling to the assembler output, and so, the callback function used in the assembler ignores the styling information and just returns an plain string. I've used the source files in gas/testsuite/gas/aarch64/ for testing, and have manually gone through and checked that the styling looks reasonable, however, I'm not an AArch64 expert, so it is possible that the odd piece is styled incorrectly. Please point out any mistakes I've made. With objdump disassembler color turned off, there should be no change in the output after this commit.
2022-04-28 20:31:07 +08:00
snprintf (buf, size, "%s",
style_reg (styler, "za%d.%s", opnd->reg.regno,
aarch64_get_qualifier_name (opnd->qualifier)));
aarch64: [SME] Add SME instructions Patch is adding new SME matrix instructions. Please note additional instructions will be added in following patches. gas/ChangeLog: * config/tc-aarch64.c (parse_sme_zada_operand): New parser. * config/tc-aarch64.c (parse_reg_with_qual): New reg parser. * config/tc-aarch64.c (R_ZA): New egister type. (parse_operands): New parser. * testsuite/gas/aarch64/sme-illegal.d: New test. * testsuite/gas/aarch64/sme-illegal.l: New test. * testsuite/gas/aarch64/sme-illegal.s: New test. * testsuite/gas/aarch64/sme.d: New test. * testsuite/gas/aarch64/sme.s: New test. * testsuite/gas/aarch64/sme-f64.d: New test. * testsuite/gas/aarch64/sme-f64.s: New test. * testsuite/gas/aarch64/sme-i64.d: New test. * testsuite/gas/aarch64/sme-i64.s: New test. include/ChangeLog: * opcode/aarch64.h (enum aarch64_opnd): New operands AARCH64_OPND_SME_ZAda_2b, AARCH64_OPND_SME_ZAda_3b and AARCH64_OPND_SME_Pm. (enum aarch64_insn_class): New instruction class sme_misc. opcodes/ChangeLog: * aarch64-opc.c (aarch64_print_operand): Print OPND_SME_ZAda_2b and OPND_SME_ZAda_3b operands. (verify_constraints): Handle OPND_SME_Pm. * aarch64-opc.h (enum aarch64_field_kind): New bit fields FLD_SME_ZAda_2b, FLD_SME_ZAda_3b and FLD_SME_Pm. * aarch64-tbl.h (OP_SME_ZADA_PN_PM_ZN_S): New qualifier set. (OP_SME_ZADA_PN_PM_ZN_D): New qualifier. (OP_SME_ZADA_PN_PM_ZN_ZM): New qualifier. (OP_SME_ZADA_S_PM_PM_S_S): New qualifier. (OP_SME_ZADA_D_PM_PM_D_D): New qualifier. (OP_SME_ZADA_S_PM_PM_H_H): New qualifier. (OP_SME_ZADA_S_PM_PM_B_B): New qualifier. (OP_SME_ZADA_D_PM_PM_H_H): New qualifier. (SME_INSN): New instruction macro. (SME_F64_INSN): New instruction macro. (SME_I64_INSN): New instruction macro. (SME_INSNC): New instruction macro. (struct aarch64_opcode): New SME instructions. aarch64-asm-2.c: Regenerate. aarch64-dis-2.c: Regenerate. aarch64-opc-2.c: Regenerate.
2021-11-18 03:21:33 +08:00
break;
aarch64: [SME] Add MOV and MOVA instructions This patch is adding new MOV (alias) and MOVA SME instruction. gas/ChangeLog: * config/tc-aarch64.c (enum sme_hv_slice): new enum. (struct reloc_entry): Added ZAH and ZAV registers. (parse_sme_immediate): Immediate parser. (parse_sme_za_hv_tiles_operand): ZA tile parser. (parse_sme_za_hv_tiles_operand_index): Index parser. (parse_operands): Added ZA tile parser calls. (REGNUMS): New macro. Regs with suffix. (REGSET16S): New macro. 16 regs with suffix. * testsuite/gas/aarch64/sme-2-illegal.d: New test. * testsuite/gas/aarch64/sme-2-illegal.l: New test. * testsuite/gas/aarch64/sme-2-illegal.s: New test. * testsuite/gas/aarch64/sme-2.d: New test. * testsuite/gas/aarch64/sme-2.s: New test. * testsuite/gas/aarch64/sme-2a.d: New test. * testsuite/gas/aarch64/sme-2a.s: New test. * testsuite/gas/aarch64/sme-3-illegal.d: New test. * testsuite/gas/aarch64/sme-3-illegal.l: New test. * testsuite/gas/aarch64/sme-3-illegal.s: New test. * testsuite/gas/aarch64/sme-3.d: New test. * testsuite/gas/aarch64/sme-3.s: New test. * testsuite/gas/aarch64/sme-3a.d: New test. * testsuite/gas/aarch64/sme-3a.s: New test. include/ChangeLog: * opcode/aarch64.h (enum aarch64_opnd): New enums AARCH64_OPND_SME_ZA_HV_idx_src and AARCH64_OPND_SME_ZA_HV_idx_dest. (struct aarch64_opnd_info): New ZA tile vector struct. opcodes/ChangeLog: * aarch64-asm.c (aarch64_ins_sme_za_hv_tiles): New inserter. * aarch64-asm.h (AARCH64_DECL_OPD_INSERTER): New inserter ins_sme_za_hv_tiles. * aarch64-dis.c (aarch64_ext_sme_za_hv_tiles): New extractor. * aarch64-dis.h (AARCH64_DECL_OPD_EXTRACTOR): New extractor ext_sme_za_hv_tiles. * aarch64-opc.c (aarch64_print_operand): Handle SME_ZA_HV_idx_src and SME_ZA_HV_idx_dest. * aarch64-opc.h (enum aarch64_field_kind): New enums FLD_SME_size_10, FLD_SME_Q, FLD_SME_V and FLD_SME_Rv. (struct aarch64_operand): Increase fields size to 5. * aarch64-tbl.h (OP_SME_BHSDQ_PM_BHSDQ): New qualifiers aarch64-asm-2.c: Regenerate. aarch64-dis-2.c: Regenerate. aarch64-opc-2.c: Regenerate.
2021-11-18 03:31:25 +08:00
case AARCH64_OPND_SME_ZA_HV_idx_src:
case AARCH64_OPND_SME_ZA_HV_idx_srcxN:
aarch64: [SME] Add MOV and MOVA instructions This patch is adding new MOV (alias) and MOVA SME instruction. gas/ChangeLog: * config/tc-aarch64.c (enum sme_hv_slice): new enum. (struct reloc_entry): Added ZAH and ZAV registers. (parse_sme_immediate): Immediate parser. (parse_sme_za_hv_tiles_operand): ZA tile parser. (parse_sme_za_hv_tiles_operand_index): Index parser. (parse_operands): Added ZA tile parser calls. (REGNUMS): New macro. Regs with suffix. (REGSET16S): New macro. 16 regs with suffix. * testsuite/gas/aarch64/sme-2-illegal.d: New test. * testsuite/gas/aarch64/sme-2-illegal.l: New test. * testsuite/gas/aarch64/sme-2-illegal.s: New test. * testsuite/gas/aarch64/sme-2.d: New test. * testsuite/gas/aarch64/sme-2.s: New test. * testsuite/gas/aarch64/sme-2a.d: New test. * testsuite/gas/aarch64/sme-2a.s: New test. * testsuite/gas/aarch64/sme-3-illegal.d: New test. * testsuite/gas/aarch64/sme-3-illegal.l: New test. * testsuite/gas/aarch64/sme-3-illegal.s: New test. * testsuite/gas/aarch64/sme-3.d: New test. * testsuite/gas/aarch64/sme-3.s: New test. * testsuite/gas/aarch64/sme-3a.d: New test. * testsuite/gas/aarch64/sme-3a.s: New test. include/ChangeLog: * opcode/aarch64.h (enum aarch64_opnd): New enums AARCH64_OPND_SME_ZA_HV_idx_src and AARCH64_OPND_SME_ZA_HV_idx_dest. (struct aarch64_opnd_info): New ZA tile vector struct. opcodes/ChangeLog: * aarch64-asm.c (aarch64_ins_sme_za_hv_tiles): New inserter. * aarch64-asm.h (AARCH64_DECL_OPD_INSERTER): New inserter ins_sme_za_hv_tiles. * aarch64-dis.c (aarch64_ext_sme_za_hv_tiles): New extractor. * aarch64-dis.h (AARCH64_DECL_OPD_EXTRACTOR): New extractor ext_sme_za_hv_tiles. * aarch64-opc.c (aarch64_print_operand): Handle SME_ZA_HV_idx_src and SME_ZA_HV_idx_dest. * aarch64-opc.h (enum aarch64_field_kind): New enums FLD_SME_size_10, FLD_SME_Q, FLD_SME_V and FLD_SME_Rv. (struct aarch64_operand): Increase fields size to 5. * aarch64-tbl.h (OP_SME_BHSDQ_PM_BHSDQ): New qualifiers aarch64-asm-2.c: Regenerate. aarch64-dis-2.c: Regenerate. aarch64-opc-2.c: Regenerate.
2021-11-18 03:31:25 +08:00
case AARCH64_OPND_SME_ZA_HV_idx_dest:
case AARCH64_OPND_SME_ZA_HV_idx_destxN:
aarch64: [SME] Add LD1x, ST1x, LDR and STR instructions This patch is adding new loads and stores defined by SME instructions. gas/ChangeLog: * config/tc-aarch64.c (parse_sme_address): New parser. (parse_sme_za_hv_tiles_operand_with_braces): New parser. (parse_sme_za_array): New parser. (output_operand_error_record): Print error details if present. (parse_operands): Support new operands. * testsuite/gas/aarch64/sme-5-illegal.d: New test. * testsuite/gas/aarch64/sme-5-illegal.l: New test. * testsuite/gas/aarch64/sme-5-illegal.s: New test. * testsuite/gas/aarch64/sme-5.d: New test. * testsuite/gas/aarch64/sme-5.s: New test. * testsuite/gas/aarch64/sme-6-illegal.d: New test. * testsuite/gas/aarch64/sme-6-illegal.l: New test. * testsuite/gas/aarch64/sme-6-illegal.s: New test. * testsuite/gas/aarch64/sme-6.d: New test. * testsuite/gas/aarch64/sme-6.s: New test. * testsuite/gas/aarch64/sme-7-illegal.d: New test. * testsuite/gas/aarch64/sme-7-illegal.l: New test. * testsuite/gas/aarch64/sme-7-illegal.s: New test. * testsuite/gas/aarch64/sme-7.d: New test. * testsuite/gas/aarch64/sme-7.s: New test. include/ChangeLog: * opcode/aarch64.h (enum aarch64_opnd): New operands. (enum aarch64_insn_class): Added sme_ldr and sme_str. (AARCH64_OPDE_UNTIED_IMMS): New operand error kind. opcodes/ChangeLog: * aarch64-asm.c (aarch64_ins_sme_za_hv_tiles): New inserter. (aarch64_ins_sme_za_list): New inserter. (aarch64_ins_sme_za_array): New inserter. (aarch64_ins_sme_addr_ri_u4xvl): New inserter. * aarch64-asm.h (AARCH64_DECL_OPD_INSERTER): Added ins_sme_za_list, ins_sme_za_array and ins_sme_addr_ri_u4xvl. * aarch64-dis.c (aarch64_ext_sme_za_hv_tiles): New extractor. (aarch64_ext_sme_za_list): New extractor. (aarch64_ext_sme_za_array): New extractor. (aarch64_ext_sme_addr_ri_u4xvl): New extractor. * aarch64-dis.h (AARCH64_DECL_OPD_EXTRACTOR): Added ext_sme_za_list, ext_sme_za_array and ext_sme_addr_ri_u4xvl. * aarch64-opc.c (operand_general_constraint_met_p): (aarch64_match_operands_constraint): Handle sme_ldr, sme_str and sme_misc. (aarch64_print_operand): New operands supported. * aarch64-tbl.h (OP_SVE_QUU): New qualifier. (OP_SVE_QZU): New qualifier. aarch64-asm-2.c: Regenerate. aarch64-dis-2.c: Regenerate. aarch64-opc-2.c: Regenerate.
2021-11-18 04:02:06 +08:00
case AARCH64_OPND_SME_ZA_HV_idx_ldstr:
snprintf (buf, size, "%s%s[%s, %s%s%s%s%s]%s",
libopcodes/aarch64: add support for disassembler styling This commit enables disassembler styling for AArch64. After this commit it is possible to have objdump style AArch64 disassembler output (using --disassembler-color option). Once the required GDB patches are merged, GDB will also style the disassembler output. The changes to support styling are mostly split between two files opcodes/aarch64-dis.c and opcodes/aarch64-opc.c. The entry point for the AArch64 disassembler can be found in aarch64-dis.c, this file handles printing the instruction mnemonics, and assembler directives (e.g. '.byte', '.word', etc). Some operands, mostly relating to assembler directives are also printed from this file. This commit changes all of this to pass through suitable styling information. However, for most "normal" instructions, the instruction operands are printed using a two step process. From aarch64-dis.c, in the print_operands function, the function aarch64_print_operand is called, this function is in aarch64-opc.c, and converts an instruction operand into a string. Then, back in print_operands (aarch64-dis.c), the operand string is printed. Unfortunately, the string returned by aarch64_print_operand can be quite complex, it will include syntax elements, like '[' and ']', in addition to register names and immediate values. In some cases, a single operand will expand into what will appear (to the user) as multiple operands separated with a ','. This makes the task of styling more complex, all these different components need to by styled differently, so we need to get the styling information out of aarch64_print_operand in some way. The solution that I propose here is similar to the solution that I used for the i386 disassembler. Currently, aarch64_print_operand uses snprintf to write the operand text into a buffer provided by the caller. What I propose is that we pass an extra argument to the aarch64_print_operand function, this argument will be a structure, the structure contains a callback function and some state. When aarch64_print_operand needs to format part of its output this can be done by using the callback function within the new structure, this callback returns a string with special embedded markers that indicate which mode should be used for each piece of text. Back in aarch64-dis.c we can spot these special style markers and use this to split the disassembler output up and apply the correct style to each piece. To make aarch64-opc.c clearer a series of new static functions have been added, e.g. 'style_reg', 'style_imm', etc. Each of these functions formats a piece of text in a different style, 'register' and 'immediate' in this case. Here's an example taken from aarch64-opc.c of the new functions in use: snprintf (buf, size, "[%s, %s]!", style_reg (styler, base), style_imm (styler, "#%d", opnd->addr.offset.imm)); The aarch64_print_operand function is also called from the assembler to aid in printing diagnostic messages. Right now I have no plans to add styling to the assembler output, and so, the callback function used in the assembler ignores the styling information and just returns an plain string. I've used the source files in gas/testsuite/gas/aarch64/ for testing, and have manually gone through and checked that the styling looks reasonable, however, I'm not an AArch64 expert, so it is possible that the odd piece is styled incorrectly. Please point out any mistakes I've made. With objdump disassembler color turned off, there should be no change in the output after this commit.
2022-04-28 20:31:07 +08:00
opnd->type == AARCH64_OPND_SME_ZA_HV_idx_ldstr ? "{" : "",
style_reg (styler, "za%d%c.%s",
opnd->indexed_za.regno,
opnd->indexed_za.v == 1 ? 'v' : 'h',
libopcodes/aarch64: add support for disassembler styling This commit enables disassembler styling for AArch64. After this commit it is possible to have objdump style AArch64 disassembler output (using --disassembler-color option). Once the required GDB patches are merged, GDB will also style the disassembler output. The changes to support styling are mostly split between two files opcodes/aarch64-dis.c and opcodes/aarch64-opc.c. The entry point for the AArch64 disassembler can be found in aarch64-dis.c, this file handles printing the instruction mnemonics, and assembler directives (e.g. '.byte', '.word', etc). Some operands, mostly relating to assembler directives are also printed from this file. This commit changes all of this to pass through suitable styling information. However, for most "normal" instructions, the instruction operands are printed using a two step process. From aarch64-dis.c, in the print_operands function, the function aarch64_print_operand is called, this function is in aarch64-opc.c, and converts an instruction operand into a string. Then, back in print_operands (aarch64-dis.c), the operand string is printed. Unfortunately, the string returned by aarch64_print_operand can be quite complex, it will include syntax elements, like '[' and ']', in addition to register names and immediate values. In some cases, a single operand will expand into what will appear (to the user) as multiple operands separated with a ','. This makes the task of styling more complex, all these different components need to by styled differently, so we need to get the styling information out of aarch64_print_operand in some way. The solution that I propose here is similar to the solution that I used for the i386 disassembler. Currently, aarch64_print_operand uses snprintf to write the operand text into a buffer provided by the caller. What I propose is that we pass an extra argument to the aarch64_print_operand function, this argument will be a structure, the structure contains a callback function and some state. When aarch64_print_operand needs to format part of its output this can be done by using the callback function within the new structure, this callback returns a string with special embedded markers that indicate which mode should be used for each piece of text. Back in aarch64-dis.c we can spot these special style markers and use this to split the disassembler output up and apply the correct style to each piece. To make aarch64-opc.c clearer a series of new static functions have been added, e.g. 'style_reg', 'style_imm', etc. Each of these functions formats a piece of text in a different style, 'register' and 'immediate' in this case. Here's an example taken from aarch64-opc.c of the new functions in use: snprintf (buf, size, "[%s, %s]!", style_reg (styler, base), style_imm (styler, "#%d", opnd->addr.offset.imm)); The aarch64_print_operand function is also called from the assembler to aid in printing diagnostic messages. Right now I have no plans to add styling to the assembler output, and so, the callback function used in the assembler ignores the styling information and just returns an plain string. I've used the source files in gas/testsuite/gas/aarch64/ for testing, and have manually gone through and checked that the styling looks reasonable, however, I'm not an AArch64 expert, so it is possible that the odd piece is styled incorrectly. Please point out any mistakes I've made. With objdump disassembler color turned off, there should be no change in the output after this commit.
2022-04-28 20:31:07 +08:00
aarch64_get_qualifier_name (opnd->qualifier)),
style_reg (styler, "w%d", opnd->indexed_za.index.regno),
style_imm (styler, "%" PRIi64, opnd->indexed_za.index.imm),
opnd->indexed_za.index.countm1 ? ":" : "",
(opnd->indexed_za.index.countm1
? style_imm (styler, "%d",
opnd->indexed_za.index.imm
+ opnd->indexed_za.index.countm1)
: ""),
opnd->indexed_za.group_size ? ", " : "",
opnd->indexed_za.group_size == 2
? style_sub_mnem (styler, "vgx2")
: opnd->indexed_za.group_size == 4
? style_sub_mnem (styler, "vgx4") : "",
libopcodes/aarch64: add support for disassembler styling This commit enables disassembler styling for AArch64. After this commit it is possible to have objdump style AArch64 disassembler output (using --disassembler-color option). Once the required GDB patches are merged, GDB will also style the disassembler output. The changes to support styling are mostly split between two files opcodes/aarch64-dis.c and opcodes/aarch64-opc.c. The entry point for the AArch64 disassembler can be found in aarch64-dis.c, this file handles printing the instruction mnemonics, and assembler directives (e.g. '.byte', '.word', etc). Some operands, mostly relating to assembler directives are also printed from this file. This commit changes all of this to pass through suitable styling information. However, for most "normal" instructions, the instruction operands are printed using a two step process. From aarch64-dis.c, in the print_operands function, the function aarch64_print_operand is called, this function is in aarch64-opc.c, and converts an instruction operand into a string. Then, back in print_operands (aarch64-dis.c), the operand string is printed. Unfortunately, the string returned by aarch64_print_operand can be quite complex, it will include syntax elements, like '[' and ']', in addition to register names and immediate values. In some cases, a single operand will expand into what will appear (to the user) as multiple operands separated with a ','. This makes the task of styling more complex, all these different components need to by styled differently, so we need to get the styling information out of aarch64_print_operand in some way. The solution that I propose here is similar to the solution that I used for the i386 disassembler. Currently, aarch64_print_operand uses snprintf to write the operand text into a buffer provided by the caller. What I propose is that we pass an extra argument to the aarch64_print_operand function, this argument will be a structure, the structure contains a callback function and some state. When aarch64_print_operand needs to format part of its output this can be done by using the callback function within the new structure, this callback returns a string with special embedded markers that indicate which mode should be used for each piece of text. Back in aarch64-dis.c we can spot these special style markers and use this to split the disassembler output up and apply the correct style to each piece. To make aarch64-opc.c clearer a series of new static functions have been added, e.g. 'style_reg', 'style_imm', etc. Each of these functions formats a piece of text in a different style, 'register' and 'immediate' in this case. Here's an example taken from aarch64-opc.c of the new functions in use: snprintf (buf, size, "[%s, %s]!", style_reg (styler, base), style_imm (styler, "#%d", opnd->addr.offset.imm)); The aarch64_print_operand function is also called from the assembler to aid in printing diagnostic messages. Right now I have no plans to add styling to the assembler output, and so, the callback function used in the assembler ignores the styling information and just returns an plain string. I've used the source files in gas/testsuite/gas/aarch64/ for testing, and have manually gone through and checked that the styling looks reasonable, however, I'm not an AArch64 expert, so it is possible that the odd piece is styled incorrectly. Please point out any mistakes I've made. With objdump disassembler color turned off, there should be no change in the output after this commit.
2022-04-28 20:31:07 +08:00
opnd->type == AARCH64_OPND_SME_ZA_HV_idx_ldstr ? "}" : "");
aarch64: [SME] Add MOV and MOVA instructions This patch is adding new MOV (alias) and MOVA SME instruction. gas/ChangeLog: * config/tc-aarch64.c (enum sme_hv_slice): new enum. (struct reloc_entry): Added ZAH and ZAV registers. (parse_sme_immediate): Immediate parser. (parse_sme_za_hv_tiles_operand): ZA tile parser. (parse_sme_za_hv_tiles_operand_index): Index parser. (parse_operands): Added ZA tile parser calls. (REGNUMS): New macro. Regs with suffix. (REGSET16S): New macro. 16 regs with suffix. * testsuite/gas/aarch64/sme-2-illegal.d: New test. * testsuite/gas/aarch64/sme-2-illegal.l: New test. * testsuite/gas/aarch64/sme-2-illegal.s: New test. * testsuite/gas/aarch64/sme-2.d: New test. * testsuite/gas/aarch64/sme-2.s: New test. * testsuite/gas/aarch64/sme-2a.d: New test. * testsuite/gas/aarch64/sme-2a.s: New test. * testsuite/gas/aarch64/sme-3-illegal.d: New test. * testsuite/gas/aarch64/sme-3-illegal.l: New test. * testsuite/gas/aarch64/sme-3-illegal.s: New test. * testsuite/gas/aarch64/sme-3.d: New test. * testsuite/gas/aarch64/sme-3.s: New test. * testsuite/gas/aarch64/sme-3a.d: New test. * testsuite/gas/aarch64/sme-3a.s: New test. include/ChangeLog: * opcode/aarch64.h (enum aarch64_opnd): New enums AARCH64_OPND_SME_ZA_HV_idx_src and AARCH64_OPND_SME_ZA_HV_idx_dest. (struct aarch64_opnd_info): New ZA tile vector struct. opcodes/ChangeLog: * aarch64-asm.c (aarch64_ins_sme_za_hv_tiles): New inserter. * aarch64-asm.h (AARCH64_DECL_OPD_INSERTER): New inserter ins_sme_za_hv_tiles. * aarch64-dis.c (aarch64_ext_sme_za_hv_tiles): New extractor. * aarch64-dis.h (AARCH64_DECL_OPD_EXTRACTOR): New extractor ext_sme_za_hv_tiles. * aarch64-opc.c (aarch64_print_operand): Handle SME_ZA_HV_idx_src and SME_ZA_HV_idx_dest. * aarch64-opc.h (enum aarch64_field_kind): New enums FLD_SME_size_10, FLD_SME_Q, FLD_SME_V and FLD_SME_Rv. (struct aarch64_operand): Increase fields size to 5. * aarch64-tbl.h (OP_SME_BHSDQ_PM_BHSDQ): New qualifiers aarch64-asm-2.c: Regenerate. aarch64-dis-2.c: Regenerate. aarch64-opc-2.c: Regenerate.
2021-11-18 03:31:25 +08:00
break;
case AARCH64_OPND_SME_list_of_64bit_tiles:
print_sme_za_list (buf, size, opnd->imm.value, styler);
break;
case AARCH64_OPND_SME_ZA_array_off1x4:
case AARCH64_OPND_SME_ZA_array_off2x2:
case AARCH64_OPND_SME_ZA_array_off2x4:
case AARCH64_OPND_SME_ZA_array_off3_0:
case AARCH64_OPND_SME_ZA_array_off3_5:
case AARCH64_OPND_SME_ZA_array_off3x2:
case AARCH64_OPND_SME_ZA_array_off4:
snprintf (buf, size, "%s[%s, %s%s%s%s%s]",
style_reg (styler, "za%s%s",
opnd->qualifier == AARCH64_OPND_QLF_NIL ? "" : ".",
(opnd->qualifier == AARCH64_OPND_QLF_NIL
? ""
: aarch64_get_qualifier_name (opnd->qualifier))),
style_reg (styler, "w%d", opnd->indexed_za.index.regno),
style_imm (styler, "%" PRIi64, opnd->indexed_za.index.imm),
opnd->indexed_za.index.countm1 ? ":" : "",
(opnd->indexed_za.index.countm1
? style_imm (styler, "%d",
opnd->indexed_za.index.imm
+ opnd->indexed_za.index.countm1)
: ""),
opnd->indexed_za.group_size ? ", " : "",
opnd->indexed_za.group_size == 2
? style_sub_mnem (styler, "vgx2")
: opnd->indexed_za.group_size == 4
? style_sub_mnem (styler, "vgx4") : "");
aarch64: [SME] Add LD1x, ST1x, LDR and STR instructions This patch is adding new loads and stores defined by SME instructions. gas/ChangeLog: * config/tc-aarch64.c (parse_sme_address): New parser. (parse_sme_za_hv_tiles_operand_with_braces): New parser. (parse_sme_za_array): New parser. (output_operand_error_record): Print error details if present. (parse_operands): Support new operands. * testsuite/gas/aarch64/sme-5-illegal.d: New test. * testsuite/gas/aarch64/sme-5-illegal.l: New test. * testsuite/gas/aarch64/sme-5-illegal.s: New test. * testsuite/gas/aarch64/sme-5.d: New test. * testsuite/gas/aarch64/sme-5.s: New test. * testsuite/gas/aarch64/sme-6-illegal.d: New test. * testsuite/gas/aarch64/sme-6-illegal.l: New test. * testsuite/gas/aarch64/sme-6-illegal.s: New test. * testsuite/gas/aarch64/sme-6.d: New test. * testsuite/gas/aarch64/sme-6.s: New test. * testsuite/gas/aarch64/sme-7-illegal.d: New test. * testsuite/gas/aarch64/sme-7-illegal.l: New test. * testsuite/gas/aarch64/sme-7-illegal.s: New test. * testsuite/gas/aarch64/sme-7.d: New test. * testsuite/gas/aarch64/sme-7.s: New test. include/ChangeLog: * opcode/aarch64.h (enum aarch64_opnd): New operands. (enum aarch64_insn_class): Added sme_ldr and sme_str. (AARCH64_OPDE_UNTIED_IMMS): New operand error kind. opcodes/ChangeLog: * aarch64-asm.c (aarch64_ins_sme_za_hv_tiles): New inserter. (aarch64_ins_sme_za_list): New inserter. (aarch64_ins_sme_za_array): New inserter. (aarch64_ins_sme_addr_ri_u4xvl): New inserter. * aarch64-asm.h (AARCH64_DECL_OPD_INSERTER): Added ins_sme_za_list, ins_sme_za_array and ins_sme_addr_ri_u4xvl. * aarch64-dis.c (aarch64_ext_sme_za_hv_tiles): New extractor. (aarch64_ext_sme_za_list): New extractor. (aarch64_ext_sme_za_array): New extractor. (aarch64_ext_sme_addr_ri_u4xvl): New extractor. * aarch64-dis.h (AARCH64_DECL_OPD_EXTRACTOR): Added ext_sme_za_list, ext_sme_za_array and ext_sme_addr_ri_u4xvl. * aarch64-opc.c (operand_general_constraint_met_p): (aarch64_match_operands_constraint): Handle sme_ldr, sme_str and sme_misc. (aarch64_print_operand): New operands supported. * aarch64-tbl.h (OP_SVE_QUU): New qualifier. (OP_SVE_QZU): New qualifier. aarch64-asm-2.c: Regenerate. aarch64-dis-2.c: Regenerate. aarch64-opc-2.c: Regenerate.
2021-11-18 04:02:06 +08:00
break;
case AARCH64_OPND_SME_ZA_array_vrsb_1:
case AARCH64_OPND_SME_ZA_array_vrsh_1:
case AARCH64_OPND_SME_ZA_array_vrss_1:
case AARCH64_OPND_SME_ZA_array_vrsd_1:
case AARCH64_OPND_SME_ZA_array_vrsb_2:
case AARCH64_OPND_SME_ZA_array_vrsh_2:
case AARCH64_OPND_SME_ZA_array_vrss_2:
case AARCH64_OPND_SME_ZA_array_vrsd_2:
snprintf (buf, size, "%s [%s, %s%s%s]",
style_reg (styler, "za%d%c%s%s",
opnd->indexed_za.regno,
opnd->indexed_za.v ? 'v': 'h',
opnd->qualifier == AARCH64_OPND_QLF_NIL ? "" : ".",
(opnd->qualifier == AARCH64_OPND_QLF_NIL
? ""
: aarch64_get_qualifier_name (opnd->qualifier))),
style_reg (styler, "w%d", opnd->indexed_za.index.regno),
style_imm (styler, "%" PRIi64, opnd->indexed_za.index.imm),
opnd->indexed_za.index.countm1 ? ":" : "",
opnd->indexed_za.index.countm1 ? style_imm (styler, "%d",
opnd->indexed_za.index.imm
+ opnd->indexed_za.index.countm1):"");
break;
aarch64: [SME] Add SME mode selection and state access instructions This patch is adding new SME mode selection and state access instructions: * Add SMSTART and SMSTOP instructions. * Add SVCR system register. gas/ChangeLog: * config/tc-aarch64.c (parse_sme_sm_za): New parser. (parse_operands): New parser. * testsuite/gas/aarch64/sme-8-illegal.d: New test. * testsuite/gas/aarch64/sme-8-illegal.l: New test. * testsuite/gas/aarch64/sme-8-illegal.s: New test. * testsuite/gas/aarch64/sme-8.d: New test. * testsuite/gas/aarch64/sme-8.s: New test. include/ChangeLog: * opcode/aarch64.h (enum aarch64_opnd): New operand AARCH64_OPND_SME_SM_ZA. (enum aarch64_insn_class): New instruction classes sme_start and sme_stop. opcodes/ChangeLog: * aarch64-asm.c (aarch64_ins_pstatefield): New inserter. (aarch64_ins_sme_sm_za): New inserter. * aarch64-dis.c (aarch64_ext_imm): New extractor. (aarch64_ext_pstatefield): New extractor. (aarch64_ext_sme_sm_za): New extractor. * aarch64-opc.c (operand_general_constraint_met_p): New pstatefield value for SME instructions. (aarch64_print_operand): Printout for OPND_SME_SM_ZA. (SR_SME): New register SVCR. * aarch64-opc.h (F_REG_IN_CRM): New register endcoding. * aarch64-opc.h (F_IMM_IN_CRM): New immediate endcoding. (PSTATE_ENCODE_CRM): Encode CRm field. (PSTATE_DECODE_CRM): Decode CRm field. (PSTATE_ENCODE_CRM_IMM): Encode CRm immediate field. (PSTATE_DECODE_CRM_IMM): Decode CRm immediate field. (PSTATE_ENCODE_CRM_AND_IMM): Encode CRm and immediate field. * aarch64-tbl.h (struct aarch64_opcode): New SMSTART and SMSTOP instructions. aarch64-asm-2.c: Regenerate. aarch64-dis-2.c: Regenerate. aarch64-opc-2.c: Regenerate.
2021-11-18 04:15:13 +08:00
case AARCH64_OPND_SME_SM_ZA:
libopcodes/aarch64: add support for disassembler styling This commit enables disassembler styling for AArch64. After this commit it is possible to have objdump style AArch64 disassembler output (using --disassembler-color option). Once the required GDB patches are merged, GDB will also style the disassembler output. The changes to support styling are mostly split between two files opcodes/aarch64-dis.c and opcodes/aarch64-opc.c. The entry point for the AArch64 disassembler can be found in aarch64-dis.c, this file handles printing the instruction mnemonics, and assembler directives (e.g. '.byte', '.word', etc). Some operands, mostly relating to assembler directives are also printed from this file. This commit changes all of this to pass through suitable styling information. However, for most "normal" instructions, the instruction operands are printed using a two step process. From aarch64-dis.c, in the print_operands function, the function aarch64_print_operand is called, this function is in aarch64-opc.c, and converts an instruction operand into a string. Then, back in print_operands (aarch64-dis.c), the operand string is printed. Unfortunately, the string returned by aarch64_print_operand can be quite complex, it will include syntax elements, like '[' and ']', in addition to register names and immediate values. In some cases, a single operand will expand into what will appear (to the user) as multiple operands separated with a ','. This makes the task of styling more complex, all these different components need to by styled differently, so we need to get the styling information out of aarch64_print_operand in some way. The solution that I propose here is similar to the solution that I used for the i386 disassembler. Currently, aarch64_print_operand uses snprintf to write the operand text into a buffer provided by the caller. What I propose is that we pass an extra argument to the aarch64_print_operand function, this argument will be a structure, the structure contains a callback function and some state. When aarch64_print_operand needs to format part of its output this can be done by using the callback function within the new structure, this callback returns a string with special embedded markers that indicate which mode should be used for each piece of text. Back in aarch64-dis.c we can spot these special style markers and use this to split the disassembler output up and apply the correct style to each piece. To make aarch64-opc.c clearer a series of new static functions have been added, e.g. 'style_reg', 'style_imm', etc. Each of these functions formats a piece of text in a different style, 'register' and 'immediate' in this case. Here's an example taken from aarch64-opc.c of the new functions in use: snprintf (buf, size, "[%s, %s]!", style_reg (styler, base), style_imm (styler, "#%d", opnd->addr.offset.imm)); The aarch64_print_operand function is also called from the assembler to aid in printing diagnostic messages. Right now I have no plans to add styling to the assembler output, and so, the callback function used in the assembler ignores the styling information and just returns an plain string. I've used the source files in gas/testsuite/gas/aarch64/ for testing, and have manually gone through and checked that the styling looks reasonable, however, I'm not an AArch64 expert, so it is possible that the odd piece is styled incorrectly. Please point out any mistakes I've made. With objdump disassembler color turned off, there should be no change in the output after this commit.
2022-04-28 20:31:07 +08:00
snprintf (buf, size, "%s",
style_reg (styler, opnd->reg.regno == 's' ? "sm" : "za"));
aarch64: [SME] Add SME mode selection and state access instructions This patch is adding new SME mode selection and state access instructions: * Add SMSTART and SMSTOP instructions. * Add SVCR system register. gas/ChangeLog: * config/tc-aarch64.c (parse_sme_sm_za): New parser. (parse_operands): New parser. * testsuite/gas/aarch64/sme-8-illegal.d: New test. * testsuite/gas/aarch64/sme-8-illegal.l: New test. * testsuite/gas/aarch64/sme-8-illegal.s: New test. * testsuite/gas/aarch64/sme-8.d: New test. * testsuite/gas/aarch64/sme-8.s: New test. include/ChangeLog: * opcode/aarch64.h (enum aarch64_opnd): New operand AARCH64_OPND_SME_SM_ZA. (enum aarch64_insn_class): New instruction classes sme_start and sme_stop. opcodes/ChangeLog: * aarch64-asm.c (aarch64_ins_pstatefield): New inserter. (aarch64_ins_sme_sm_za): New inserter. * aarch64-dis.c (aarch64_ext_imm): New extractor. (aarch64_ext_pstatefield): New extractor. (aarch64_ext_sme_sm_za): New extractor. * aarch64-opc.c (operand_general_constraint_met_p): New pstatefield value for SME instructions. (aarch64_print_operand): Printout for OPND_SME_SM_ZA. (SR_SME): New register SVCR. * aarch64-opc.h (F_REG_IN_CRM): New register endcoding. * aarch64-opc.h (F_IMM_IN_CRM): New immediate endcoding. (PSTATE_ENCODE_CRM): Encode CRm field. (PSTATE_DECODE_CRM): Decode CRm field. (PSTATE_ENCODE_CRM_IMM): Encode CRm immediate field. (PSTATE_DECODE_CRM_IMM): Decode CRm immediate field. (PSTATE_ENCODE_CRM_AND_IMM): Encode CRm and immediate field. * aarch64-tbl.h (struct aarch64_opcode): New SMSTART and SMSTOP instructions. aarch64-asm-2.c: Regenerate. aarch64-dis-2.c: Regenerate. aarch64-opc-2.c: Regenerate.
2021-11-18 04:15:13 +08:00
break;
case AARCH64_OPND_SME_PnT_Wm_imm:
libopcodes/aarch64: add support for disassembler styling This commit enables disassembler styling for AArch64. After this commit it is possible to have objdump style AArch64 disassembler output (using --disassembler-color option). Once the required GDB patches are merged, GDB will also style the disassembler output. The changes to support styling are mostly split between two files opcodes/aarch64-dis.c and opcodes/aarch64-opc.c. The entry point for the AArch64 disassembler can be found in aarch64-dis.c, this file handles printing the instruction mnemonics, and assembler directives (e.g. '.byte', '.word', etc). Some operands, mostly relating to assembler directives are also printed from this file. This commit changes all of this to pass through suitable styling information. However, for most "normal" instructions, the instruction operands are printed using a two step process. From aarch64-dis.c, in the print_operands function, the function aarch64_print_operand is called, this function is in aarch64-opc.c, and converts an instruction operand into a string. Then, back in print_operands (aarch64-dis.c), the operand string is printed. Unfortunately, the string returned by aarch64_print_operand can be quite complex, it will include syntax elements, like '[' and ']', in addition to register names and immediate values. In some cases, a single operand will expand into what will appear (to the user) as multiple operands separated with a ','. This makes the task of styling more complex, all these different components need to by styled differently, so we need to get the styling information out of aarch64_print_operand in some way. The solution that I propose here is similar to the solution that I used for the i386 disassembler. Currently, aarch64_print_operand uses snprintf to write the operand text into a buffer provided by the caller. What I propose is that we pass an extra argument to the aarch64_print_operand function, this argument will be a structure, the structure contains a callback function and some state. When aarch64_print_operand needs to format part of its output this can be done by using the callback function within the new structure, this callback returns a string with special embedded markers that indicate which mode should be used for each piece of text. Back in aarch64-dis.c we can spot these special style markers and use this to split the disassembler output up and apply the correct style to each piece. To make aarch64-opc.c clearer a series of new static functions have been added, e.g. 'style_reg', 'style_imm', etc. Each of these functions formats a piece of text in a different style, 'register' and 'immediate' in this case. Here's an example taken from aarch64-opc.c of the new functions in use: snprintf (buf, size, "[%s, %s]!", style_reg (styler, base), style_imm (styler, "#%d", opnd->addr.offset.imm)); The aarch64_print_operand function is also called from the assembler to aid in printing diagnostic messages. Right now I have no plans to add styling to the assembler output, and so, the callback function used in the assembler ignores the styling information and just returns an plain string. I've used the source files in gas/testsuite/gas/aarch64/ for testing, and have manually gone through and checked that the styling looks reasonable, however, I'm not an AArch64 expert, so it is possible that the odd piece is styled incorrectly. Please point out any mistakes I've made. With objdump disassembler color turned off, there should be no change in the output after this commit.
2022-04-28 20:31:07 +08:00
snprintf (buf, size, "%s[%s, %s]",
style_reg (styler, "p%d.%s", opnd->indexed_za.regno,
libopcodes/aarch64: add support for disassembler styling This commit enables disassembler styling for AArch64. After this commit it is possible to have objdump style AArch64 disassembler output (using --disassembler-color option). Once the required GDB patches are merged, GDB will also style the disassembler output. The changes to support styling are mostly split between two files opcodes/aarch64-dis.c and opcodes/aarch64-opc.c. The entry point for the AArch64 disassembler can be found in aarch64-dis.c, this file handles printing the instruction mnemonics, and assembler directives (e.g. '.byte', '.word', etc). Some operands, mostly relating to assembler directives are also printed from this file. This commit changes all of this to pass through suitable styling information. However, for most "normal" instructions, the instruction operands are printed using a two step process. From aarch64-dis.c, in the print_operands function, the function aarch64_print_operand is called, this function is in aarch64-opc.c, and converts an instruction operand into a string. Then, back in print_operands (aarch64-dis.c), the operand string is printed. Unfortunately, the string returned by aarch64_print_operand can be quite complex, it will include syntax elements, like '[' and ']', in addition to register names and immediate values. In some cases, a single operand will expand into what will appear (to the user) as multiple operands separated with a ','. This makes the task of styling more complex, all these different components need to by styled differently, so we need to get the styling information out of aarch64_print_operand in some way. The solution that I propose here is similar to the solution that I used for the i386 disassembler. Currently, aarch64_print_operand uses snprintf to write the operand text into a buffer provided by the caller. What I propose is that we pass an extra argument to the aarch64_print_operand function, this argument will be a structure, the structure contains a callback function and some state. When aarch64_print_operand needs to format part of its output this can be done by using the callback function within the new structure, this callback returns a string with special embedded markers that indicate which mode should be used for each piece of text. Back in aarch64-dis.c we can spot these special style markers and use this to split the disassembler output up and apply the correct style to each piece. To make aarch64-opc.c clearer a series of new static functions have been added, e.g. 'style_reg', 'style_imm', etc. Each of these functions formats a piece of text in a different style, 'register' and 'immediate' in this case. Here's an example taken from aarch64-opc.c of the new functions in use: snprintf (buf, size, "[%s, %s]!", style_reg (styler, base), style_imm (styler, "#%d", opnd->addr.offset.imm)); The aarch64_print_operand function is also called from the assembler to aid in printing diagnostic messages. Right now I have no plans to add styling to the assembler output, and so, the callback function used in the assembler ignores the styling information and just returns an plain string. I've used the source files in gas/testsuite/gas/aarch64/ for testing, and have manually gone through and checked that the styling looks reasonable, however, I'm not an AArch64 expert, so it is possible that the odd piece is styled incorrectly. Please point out any mistakes I've made. With objdump disassembler color turned off, there should be no change in the output after this commit.
2022-04-28 20:31:07 +08:00
aarch64_get_qualifier_name (opnd->qualifier)),
style_reg (styler, "w%d", opnd->indexed_za.index.regno),
style_imm (styler, "%" PRIi64, opnd->indexed_za.index.imm));
break;
case AARCH64_OPND_SME_VLxN_10:
case AARCH64_OPND_SME_VLxN_13:
enum_value = opnd->imm.value;
assert (enum_value < ARRAY_SIZE (aarch64_sme_vlxn_array));
snprintf (buf, size, "%s",
style_sub_mnem (styler, aarch64_sme_vlxn_array[enum_value]));
break;
case AARCH64_OPND_CRn:
case AARCH64_OPND_CRm:
libopcodes/aarch64: add support for disassembler styling This commit enables disassembler styling for AArch64. After this commit it is possible to have objdump style AArch64 disassembler output (using --disassembler-color option). Once the required GDB patches are merged, GDB will also style the disassembler output. The changes to support styling are mostly split between two files opcodes/aarch64-dis.c and opcodes/aarch64-opc.c. The entry point for the AArch64 disassembler can be found in aarch64-dis.c, this file handles printing the instruction mnemonics, and assembler directives (e.g. '.byte', '.word', etc). Some operands, mostly relating to assembler directives are also printed from this file. This commit changes all of this to pass through suitable styling information. However, for most "normal" instructions, the instruction operands are printed using a two step process. From aarch64-dis.c, in the print_operands function, the function aarch64_print_operand is called, this function is in aarch64-opc.c, and converts an instruction operand into a string. Then, back in print_operands (aarch64-dis.c), the operand string is printed. Unfortunately, the string returned by aarch64_print_operand can be quite complex, it will include syntax elements, like '[' and ']', in addition to register names and immediate values. In some cases, a single operand will expand into what will appear (to the user) as multiple operands separated with a ','. This makes the task of styling more complex, all these different components need to by styled differently, so we need to get the styling information out of aarch64_print_operand in some way. The solution that I propose here is similar to the solution that I used for the i386 disassembler. Currently, aarch64_print_operand uses snprintf to write the operand text into a buffer provided by the caller. What I propose is that we pass an extra argument to the aarch64_print_operand function, this argument will be a structure, the structure contains a callback function and some state. When aarch64_print_operand needs to format part of its output this can be done by using the callback function within the new structure, this callback returns a string with special embedded markers that indicate which mode should be used for each piece of text. Back in aarch64-dis.c we can spot these special style markers and use this to split the disassembler output up and apply the correct style to each piece. To make aarch64-opc.c clearer a series of new static functions have been added, e.g. 'style_reg', 'style_imm', etc. Each of these functions formats a piece of text in a different style, 'register' and 'immediate' in this case. Here's an example taken from aarch64-opc.c of the new functions in use: snprintf (buf, size, "[%s, %s]!", style_reg (styler, base), style_imm (styler, "#%d", opnd->addr.offset.imm)); The aarch64_print_operand function is also called from the assembler to aid in printing diagnostic messages. Right now I have no plans to add styling to the assembler output, and so, the callback function used in the assembler ignores the styling information and just returns an plain string. I've used the source files in gas/testsuite/gas/aarch64/ for testing, and have manually gone through and checked that the styling looks reasonable, however, I'm not an AArch64 expert, so it is possible that the odd piece is styled incorrectly. Please point out any mistakes I've made. With objdump disassembler color turned off, there should be no change in the output after this commit.
2022-04-28 20:31:07 +08:00
snprintf (buf, size, "%s",
style_reg (styler, "C%" PRIi64, opnd->imm.value));
break;
case AARCH64_OPND_IDX:
case AARCH64_OPND_MASK:
case AARCH64_OPND_IMM:
case AARCH64_OPND_IMM_2:
case AARCH64_OPND_WIDTH:
case AARCH64_OPND_UIMM3_OP1:
case AARCH64_OPND_UIMM3_OP2:
case AARCH64_OPND_BIT_NUM:
case AARCH64_OPND_IMM_VLSL:
case AARCH64_OPND_IMM_VLSR:
case AARCH64_OPND_SHLL_IMM:
case AARCH64_OPND_IMM0:
case AARCH64_OPND_IMMR:
case AARCH64_OPND_IMMS:
case AARCH64_OPND_UNDEFINED:
case AARCH64_OPND_FBITS:
[BINUTILS, AArch64] Enable Transactional Memory Extension This patch enables the new Transactional Memory Extension added recently as part of Arm's new architecture technologies. We introduce a new optional extension "tme" to enable this. The following instructions are part of the extension: * tstart <Xt> * ttest <Xt> * tcommit * tcancel #<imm> The ISA for the above can be found here: https://developer.arm.com/docs/ddi0602/latest/base-instructions-alphabetic-order *** gas/ChangeLog *** 2019-05-01 Sudakshina Das <sudi.das@arm.com> * config/tc-aarch64.c (parse_operands): Add case for AARCH64_OPND_TME_UIMM16. (aarch64_features): Add "tme". * doc/c-aarch64.texi: Document the same. * testsuite/gas/aarch64/tme-invalid.d: New test. * testsuite/gas/aarch64/tme-invalid.l: New test. * testsuite/gas/aarch64/tme-invalid.s: New test. * testsuite/gas/aarch64/tme.d: New test. * testsuite/gas/aarch64/tme.s: New test. *** include/ChangeLog *** 2019-05-01 Sudakshina Das <sudi.das@arm.com> * opcode/aarch64.h (AARCH64_FEATURE_TME): New. (enum aarch64_opnd): Add AARCH64_OPND_TME_UIMM16. *** opcodes/ChangeLog *** 2019-05-01 Sudakshina Das <sudi.das@arm.com> * aarch64-asm-2.c: Regenerated. * aarch64-dis-2.c: Regenerated. * aarch64-opc-2.c: Regenerated. * aarch64-opc.c (operand_general_constraint_met_p): Add case for AARCH64_OPND_TME_UIMM16. (aarch64_print_operand): Likewise. * aarch64-tbl.h (QL_IMM_NIL): New. (TME): New. (_TME_INSN): New. (struct aarch64_opcode): Add tstart, tcommit, ttest and tcancel.
2019-05-02 00:14:01 +08:00
case AARCH64_OPND_TME_UIMM16:
[AArch64][SVE 27/32] Add SVE integer immediate operands This patch adds the new SVE integer immediate operands. There are three kinds: - simple signed and unsigned ranges, but with new widths and positions. - 13-bit logical immediates. These have the same form as in base AArch64, but at a different bit position. In the case of the "MOV Zn.<T>, #<limm>" alias of DUPM, the logical immediate <limm> is not allowed to be a valid DUP immediate, since DUP is preferred over DUPM for constants that both instructions can handle. - a new 9-bit arithmetic immediate, of the form "<imm8>{, LSL #8}". In some contexts the operand is signed and in others it's unsigned. As an extension, we allow shifted immediates to be written as a single integer, e.g. "#256" is equivalent to "#1, LSL #8". We also use the shiftless form as the preferred disassembly, except for the special case of "#0, LSL #8" (a redundant encoding of 0). include/ * opcode/aarch64.h (AARCH64_OPND_SIMM5): New aarch64_opnd. (AARCH64_OPND_SVE_AIMM, AARCH64_OPND_SVE_ASIMM) (AARCH64_OPND_SVE_INV_LIMM, AARCH64_OPND_SVE_LIMM) (AARCH64_OPND_SVE_LIMM_MOV, AARCH64_OPND_SVE_SHLIMM_PRED) (AARCH64_OPND_SVE_SHLIMM_UNPRED, AARCH64_OPND_SVE_SHRIMM_PRED) (AARCH64_OPND_SVE_SHRIMM_UNPRED, AARCH64_OPND_SVE_SIMM5) (AARCH64_OPND_SVE_SIMM5B, AARCH64_OPND_SVE_SIMM6) (AARCH64_OPND_SVE_SIMM8, AARCH64_OPND_SVE_UIMM3) (AARCH64_OPND_SVE_UIMM7, AARCH64_OPND_SVE_UIMM8) (AARCH64_OPND_SVE_UIMM8_53): Likewise. (aarch64_sve_dupm_mov_immediate_p): Declare. opcodes/ * aarch64-tbl.h (AARCH64_OPERANDS): Add entries for the new SVE integer immediate operands. * aarch64-opc.h (FLD_SVE_immN, FLD_SVE_imm3, FLD_SVE_imm5) (FLD_SVE_imm5b, FLD_SVE_imm7, FLD_SVE_imm8, FLD_SVE_imm9) (FLD_SVE_immr, FLD_SVE_imms, FLD_SVE_tszh): New aarch64_field_kinds. * aarch64-opc.c (fields): Add corresponding entries. (operand_general_constraint_met_p): Handle the new SVE integer immediate operands. (aarch64_print_operand): Likewise. (aarch64_sve_dupm_mov_immediate_p): New function. * aarch64-opc-2.c: Regenerate. * aarch64-asm.h (ins_inv_limm, ins_sve_aimm, ins_sve_asimm) (ins_sve_limm_mov, ins_sve_shlimm, ins_sve_shrimm): New inserters. * aarch64-asm.c (aarch64_ins_limm_1): New function, split out from... (aarch64_ins_limm): ...here. (aarch64_ins_inv_limm): New function. (aarch64_ins_sve_aimm): Likewise. (aarch64_ins_sve_asimm): Likewise. (aarch64_ins_sve_limm_mov): Likewise. (aarch64_ins_sve_shlimm): Likewise. (aarch64_ins_sve_shrimm): Likewise. * aarch64-asm-2.c: Regenerate. * aarch64-dis.h (ext_inv_limm, ext_sve_aimm, ext_sve_asimm) (ext_sve_limm_mov, ext_sve_shlimm, ext_sve_shrimm): New extractors. * aarch64-dis.c (decode_limm): New function, split out from... (aarch64_ext_limm): ...here. (aarch64_ext_inv_limm): New function. (decode_sve_aimm): Likewise. (aarch64_ext_sve_aimm): Likewise. (aarch64_ext_sve_asimm): Likewise. (aarch64_ext_sve_limm_mov): Likewise. (aarch64_top_bit): Likewise. (aarch64_ext_sve_shlimm): Likewise. (aarch64_ext_sve_shrimm): Likewise. * aarch64-dis-2.c: Regenerate. gas/ * config/tc-aarch64.c (parse_operands): Handle the new SVE integer immediate operands.
2016-09-21 23:56:57 +08:00
case AARCH64_OPND_SIMM5:
case AARCH64_OPND_SME_SHRIMM4:
case AARCH64_OPND_SME_SHRIMM5:
[AArch64][SVE 27/32] Add SVE integer immediate operands This patch adds the new SVE integer immediate operands. There are three kinds: - simple signed and unsigned ranges, but with new widths and positions. - 13-bit logical immediates. These have the same form as in base AArch64, but at a different bit position. In the case of the "MOV Zn.<T>, #<limm>" alias of DUPM, the logical immediate <limm> is not allowed to be a valid DUP immediate, since DUP is preferred over DUPM for constants that both instructions can handle. - a new 9-bit arithmetic immediate, of the form "<imm8>{, LSL #8}". In some contexts the operand is signed and in others it's unsigned. As an extension, we allow shifted immediates to be written as a single integer, e.g. "#256" is equivalent to "#1, LSL #8". We also use the shiftless form as the preferred disassembly, except for the special case of "#0, LSL #8" (a redundant encoding of 0). include/ * opcode/aarch64.h (AARCH64_OPND_SIMM5): New aarch64_opnd. (AARCH64_OPND_SVE_AIMM, AARCH64_OPND_SVE_ASIMM) (AARCH64_OPND_SVE_INV_LIMM, AARCH64_OPND_SVE_LIMM) (AARCH64_OPND_SVE_LIMM_MOV, AARCH64_OPND_SVE_SHLIMM_PRED) (AARCH64_OPND_SVE_SHLIMM_UNPRED, AARCH64_OPND_SVE_SHRIMM_PRED) (AARCH64_OPND_SVE_SHRIMM_UNPRED, AARCH64_OPND_SVE_SIMM5) (AARCH64_OPND_SVE_SIMM5B, AARCH64_OPND_SVE_SIMM6) (AARCH64_OPND_SVE_SIMM8, AARCH64_OPND_SVE_UIMM3) (AARCH64_OPND_SVE_UIMM7, AARCH64_OPND_SVE_UIMM8) (AARCH64_OPND_SVE_UIMM8_53): Likewise. (aarch64_sve_dupm_mov_immediate_p): Declare. opcodes/ * aarch64-tbl.h (AARCH64_OPERANDS): Add entries for the new SVE integer immediate operands. * aarch64-opc.h (FLD_SVE_immN, FLD_SVE_imm3, FLD_SVE_imm5) (FLD_SVE_imm5b, FLD_SVE_imm7, FLD_SVE_imm8, FLD_SVE_imm9) (FLD_SVE_immr, FLD_SVE_imms, FLD_SVE_tszh): New aarch64_field_kinds. * aarch64-opc.c (fields): Add corresponding entries. (operand_general_constraint_met_p): Handle the new SVE integer immediate operands. (aarch64_print_operand): Likewise. (aarch64_sve_dupm_mov_immediate_p): New function. * aarch64-opc-2.c: Regenerate. * aarch64-asm.h (ins_inv_limm, ins_sve_aimm, ins_sve_asimm) (ins_sve_limm_mov, ins_sve_shlimm, ins_sve_shrimm): New inserters. * aarch64-asm.c (aarch64_ins_limm_1): New function, split out from... (aarch64_ins_limm): ...here. (aarch64_ins_inv_limm): New function. (aarch64_ins_sve_aimm): Likewise. (aarch64_ins_sve_asimm): Likewise. (aarch64_ins_sve_limm_mov): Likewise. (aarch64_ins_sve_shlimm): Likewise. (aarch64_ins_sve_shrimm): Likewise. * aarch64-asm-2.c: Regenerate. * aarch64-dis.h (ext_inv_limm, ext_sve_aimm, ext_sve_asimm) (ext_sve_limm_mov, ext_sve_shlimm, ext_sve_shrimm): New extractors. * aarch64-dis.c (decode_limm): New function, split out from... (aarch64_ext_limm): ...here. (aarch64_ext_inv_limm): New function. (decode_sve_aimm): Likewise. (aarch64_ext_sve_aimm): Likewise. (aarch64_ext_sve_asimm): Likewise. (aarch64_ext_sve_limm_mov): Likewise. (aarch64_top_bit): Likewise. (aarch64_ext_sve_shlimm): Likewise. (aarch64_ext_sve_shrimm): Likewise. * aarch64-dis-2.c: Regenerate. gas/ * config/tc-aarch64.c (parse_operands): Handle the new SVE integer immediate operands.
2016-09-21 23:56:57 +08:00
case AARCH64_OPND_SVE_SHLIMM_PRED:
case AARCH64_OPND_SVE_SHLIMM_UNPRED:
case AARCH64_OPND_SVE_SHLIMM_UNPRED_22:
[AArch64][SVE 27/32] Add SVE integer immediate operands This patch adds the new SVE integer immediate operands. There are three kinds: - simple signed and unsigned ranges, but with new widths and positions. - 13-bit logical immediates. These have the same form as in base AArch64, but at a different bit position. In the case of the "MOV Zn.<T>, #<limm>" alias of DUPM, the logical immediate <limm> is not allowed to be a valid DUP immediate, since DUP is preferred over DUPM for constants that both instructions can handle. - a new 9-bit arithmetic immediate, of the form "<imm8>{, LSL #8}". In some contexts the operand is signed and in others it's unsigned. As an extension, we allow shifted immediates to be written as a single integer, e.g. "#256" is equivalent to "#1, LSL #8". We also use the shiftless form as the preferred disassembly, except for the special case of "#0, LSL #8" (a redundant encoding of 0). include/ * opcode/aarch64.h (AARCH64_OPND_SIMM5): New aarch64_opnd. (AARCH64_OPND_SVE_AIMM, AARCH64_OPND_SVE_ASIMM) (AARCH64_OPND_SVE_INV_LIMM, AARCH64_OPND_SVE_LIMM) (AARCH64_OPND_SVE_LIMM_MOV, AARCH64_OPND_SVE_SHLIMM_PRED) (AARCH64_OPND_SVE_SHLIMM_UNPRED, AARCH64_OPND_SVE_SHRIMM_PRED) (AARCH64_OPND_SVE_SHRIMM_UNPRED, AARCH64_OPND_SVE_SIMM5) (AARCH64_OPND_SVE_SIMM5B, AARCH64_OPND_SVE_SIMM6) (AARCH64_OPND_SVE_SIMM8, AARCH64_OPND_SVE_UIMM3) (AARCH64_OPND_SVE_UIMM7, AARCH64_OPND_SVE_UIMM8) (AARCH64_OPND_SVE_UIMM8_53): Likewise. (aarch64_sve_dupm_mov_immediate_p): Declare. opcodes/ * aarch64-tbl.h (AARCH64_OPERANDS): Add entries for the new SVE integer immediate operands. * aarch64-opc.h (FLD_SVE_immN, FLD_SVE_imm3, FLD_SVE_imm5) (FLD_SVE_imm5b, FLD_SVE_imm7, FLD_SVE_imm8, FLD_SVE_imm9) (FLD_SVE_immr, FLD_SVE_imms, FLD_SVE_tszh): New aarch64_field_kinds. * aarch64-opc.c (fields): Add corresponding entries. (operand_general_constraint_met_p): Handle the new SVE integer immediate operands. (aarch64_print_operand): Likewise. (aarch64_sve_dupm_mov_immediate_p): New function. * aarch64-opc-2.c: Regenerate. * aarch64-asm.h (ins_inv_limm, ins_sve_aimm, ins_sve_asimm) (ins_sve_limm_mov, ins_sve_shlimm, ins_sve_shrimm): New inserters. * aarch64-asm.c (aarch64_ins_limm_1): New function, split out from... (aarch64_ins_limm): ...here. (aarch64_ins_inv_limm): New function. (aarch64_ins_sve_aimm): Likewise. (aarch64_ins_sve_asimm): Likewise. (aarch64_ins_sve_limm_mov): Likewise. (aarch64_ins_sve_shlimm): Likewise. (aarch64_ins_sve_shrimm): Likewise. * aarch64-asm-2.c: Regenerate. * aarch64-dis.h (ext_inv_limm, ext_sve_aimm, ext_sve_asimm) (ext_sve_limm_mov, ext_sve_shlimm, ext_sve_shrimm): New extractors. * aarch64-dis.c (decode_limm): New function, split out from... (aarch64_ext_limm): ...here. (aarch64_ext_inv_limm): New function. (decode_sve_aimm): Likewise. (aarch64_ext_sve_aimm): Likewise. (aarch64_ext_sve_asimm): Likewise. (aarch64_ext_sve_limm_mov): Likewise. (aarch64_top_bit): Likewise. (aarch64_ext_sve_shlimm): Likewise. (aarch64_ext_sve_shrimm): Likewise. * aarch64-dis-2.c: Regenerate. gas/ * config/tc-aarch64.c (parse_operands): Handle the new SVE integer immediate operands.
2016-09-21 23:56:57 +08:00
case AARCH64_OPND_SVE_SHRIMM_PRED:
case AARCH64_OPND_SVE_SHRIMM_UNPRED:
[binutils][aarch64] New SVE_SHRIMM_UNPRED_22 operand. Include a new iclass to extract the variant from the most significant 3 bits of this operand. Instructions such as rshrnb include a constant shift amount as an operand, where the most significant three bits of this operand determine what size elements the instruction is operating on. The new SVE_SHRIMM_UNPRED_22 operand denotes this constant encoded in bits 22:20-19:18-16 while the new sve_shift_tsz_hsd iclass denotes that the SVE qualifier is encoded in bits 22:20-19. gas/ChangeLog: 2019-05-09 Matthew Malcomson <matthew.malcomson@arm.com> * config/tc-aarch64.c (parse_operands): Handle new SVE_SHRIMM_UNPRED_22 operand. include/ChangeLog: 2019-05-09 Matthew Malcomson <matthew.malcomson@arm.com> * opcode/aarch64.h (enum aarch64_opnd): New SVE_SHRIMM_UNPRED_22 operand. (enum aarch64_insn_class): Add sve_shift_tsz_hsd iclass. opcodes/ChangeLog: 2019-05-09 Matthew Malcomson <matthew.malcomson@arm.com> * aarch64-asm-2.c: Regenerated. * aarch64-dis-2.c: Regenerated. * aarch64-opc-2.c: Regenerated. * aarch64-asm.c (aarch64_ins_sve_shrimm): (aarch64_encode_variant_using_iclass): Handle sve_shift_tsz_hsd iclass encode. * aarch64-dis.c (aarch64_decode_variant_using_iclass): Handle sve_shift_tsz_hsd iclass decode. * aarch64-opc.c (operand_general_constraint_met_p): Constraint checking for SVE_SHRIMM_UNPRED_22. (aarch64_print_operand): Add printing for SVE_SHRIMM_UNPRED_22. * aarch64-tbl.h (AARCH64_OPERANDS): Use new SVE_SHRIMM_UNPRED_22 operand.
2019-05-09 17:29:22 +08:00
case AARCH64_OPND_SVE_SHRIMM_UNPRED_22:
[AArch64][SVE 27/32] Add SVE integer immediate operands This patch adds the new SVE integer immediate operands. There are three kinds: - simple signed and unsigned ranges, but with new widths and positions. - 13-bit logical immediates. These have the same form as in base AArch64, but at a different bit position. In the case of the "MOV Zn.<T>, #<limm>" alias of DUPM, the logical immediate <limm> is not allowed to be a valid DUP immediate, since DUP is preferred over DUPM for constants that both instructions can handle. - a new 9-bit arithmetic immediate, of the form "<imm8>{, LSL #8}". In some contexts the operand is signed and in others it's unsigned. As an extension, we allow shifted immediates to be written as a single integer, e.g. "#256" is equivalent to "#1, LSL #8". We also use the shiftless form as the preferred disassembly, except for the special case of "#0, LSL #8" (a redundant encoding of 0). include/ * opcode/aarch64.h (AARCH64_OPND_SIMM5): New aarch64_opnd. (AARCH64_OPND_SVE_AIMM, AARCH64_OPND_SVE_ASIMM) (AARCH64_OPND_SVE_INV_LIMM, AARCH64_OPND_SVE_LIMM) (AARCH64_OPND_SVE_LIMM_MOV, AARCH64_OPND_SVE_SHLIMM_PRED) (AARCH64_OPND_SVE_SHLIMM_UNPRED, AARCH64_OPND_SVE_SHRIMM_PRED) (AARCH64_OPND_SVE_SHRIMM_UNPRED, AARCH64_OPND_SVE_SIMM5) (AARCH64_OPND_SVE_SIMM5B, AARCH64_OPND_SVE_SIMM6) (AARCH64_OPND_SVE_SIMM8, AARCH64_OPND_SVE_UIMM3) (AARCH64_OPND_SVE_UIMM7, AARCH64_OPND_SVE_UIMM8) (AARCH64_OPND_SVE_UIMM8_53): Likewise. (aarch64_sve_dupm_mov_immediate_p): Declare. opcodes/ * aarch64-tbl.h (AARCH64_OPERANDS): Add entries for the new SVE integer immediate operands. * aarch64-opc.h (FLD_SVE_immN, FLD_SVE_imm3, FLD_SVE_imm5) (FLD_SVE_imm5b, FLD_SVE_imm7, FLD_SVE_imm8, FLD_SVE_imm9) (FLD_SVE_immr, FLD_SVE_imms, FLD_SVE_tszh): New aarch64_field_kinds. * aarch64-opc.c (fields): Add corresponding entries. (operand_general_constraint_met_p): Handle the new SVE integer immediate operands. (aarch64_print_operand): Likewise. (aarch64_sve_dupm_mov_immediate_p): New function. * aarch64-opc-2.c: Regenerate. * aarch64-asm.h (ins_inv_limm, ins_sve_aimm, ins_sve_asimm) (ins_sve_limm_mov, ins_sve_shlimm, ins_sve_shrimm): New inserters. * aarch64-asm.c (aarch64_ins_limm_1): New function, split out from... (aarch64_ins_limm): ...here. (aarch64_ins_inv_limm): New function. (aarch64_ins_sve_aimm): Likewise. (aarch64_ins_sve_asimm): Likewise. (aarch64_ins_sve_limm_mov): Likewise. (aarch64_ins_sve_shlimm): Likewise. (aarch64_ins_sve_shrimm): Likewise. * aarch64-asm-2.c: Regenerate. * aarch64-dis.h (ext_inv_limm, ext_sve_aimm, ext_sve_asimm) (ext_sve_limm_mov, ext_sve_shlimm, ext_sve_shrimm): New extractors. * aarch64-dis.c (decode_limm): New function, split out from... (aarch64_ext_limm): ...here. (aarch64_ext_inv_limm): New function. (decode_sve_aimm): Likewise. (aarch64_ext_sve_aimm): Likewise. (aarch64_ext_sve_asimm): Likewise. (aarch64_ext_sve_limm_mov): Likewise. (aarch64_top_bit): Likewise. (aarch64_ext_sve_shlimm): Likewise. (aarch64_ext_sve_shrimm): Likewise. * aarch64-dis-2.c: Regenerate. gas/ * config/tc-aarch64.c (parse_operands): Handle the new SVE integer immediate operands.
2016-09-21 23:56:57 +08:00
case AARCH64_OPND_SVE_SIMM5:
case AARCH64_OPND_SVE_SIMM5B:
case AARCH64_OPND_SVE_SIMM6:
case AARCH64_OPND_SVE_SIMM8:
case AARCH64_OPND_SVE_UIMM3:
case AARCH64_OPND_SVE_UIMM7:
case AARCH64_OPND_SVE_UIMM8:
case AARCH64_OPND_SVE_UIMM8_53:
[AArch64] Add ARMv8.3 FCMLA and FCADD instructions Add support for FCMLA and FCADD complex arithmetic SIMD instructions. FCMLA has an indexed element variant where the index range has to be treated specially because a complex number takes two elements and the indexed vector size depends on the other operands. These complex number SIMD instructions are part of ARMv8.3 https://community.arm.com/groups/processors/blog/2016/10/27/armv8-a-architecture-2016-additions include/ 2016-11-18 Szabolcs Nagy <szabolcs.nagy@arm.com> * opcode/aarch64.h (enum aarch64_opnd): Add AARCH64_OPND_IMM_ROT1, AARCH64_OPND_IMM_ROT2, AARCH64_OPND_IMM_ROT3. (enum aarch64_op): Add OP_FCMLA_ELEM. opcodes/ 2016-11-18 Szabolcs Nagy <szabolcs.nagy@arm.com> * aarch64-tbl.h (QL_V3SAMEHSD_ROT, QL_ELEMENT_ROT): Define. (aarch64_feature_simd_v8_3, SIMD_V8_3): Define. (aarch64_opcode_table): Add fcmla and fcadd. (AARCH64_OPERANDS): Add IMM_ROT{1,2,3}. * aarch64-asm.h (aarch64_ins_imm_rotate): Declare. * aarch64-asm.c (aarch64_ins_imm_rotate): Define. * aarch64-dis.h (aarch64_ext_imm_rotate): Declare. * aarch64-dis.c (aarch64_ext_imm_rotate): Define. * aarch64-opc.h (enum aarch64_field_kind): Add FLD_rotate{1,2,3}. * aarch64-opc.c (fields): Add FLD_rotate{1,2,3}. (operand_general_constraint_met_p): Rotate and index range check. (aarch64_print_operand): Handle rotate operand. * aarch64-asm-2.c: Regenerate. * aarch64-dis-2.c: Likewise. * aarch64-opc-2.c: Likewise. gas/ 2016-11-18 Szabolcs Nagy <szabolcs.nagy@arm.com> * config/tc-aarch64.c (parse_operands): Handle AARCH64_OPND_IMM_ROT*. * testsuite/gas/aarch64/advsimd-armv8_3.d: New. * testsuite/gas/aarch64/advsimd-armv8_3.s: New. * testsuite/gas/aarch64/illegal-fcmla.s: New. * testsuite/gas/aarch64/illegal-fcmla.l: New. * testsuite/gas/aarch64/illegal-fcmla.d: New.
2016-11-18 18:02:16 +08:00
case AARCH64_OPND_IMM_ROT1:
case AARCH64_OPND_IMM_ROT2:
case AARCH64_OPND_IMM_ROT3:
[AArch64] Additional SVE instructions This patch supports some additions to the SVE architecture prior to its public release. include/ * opcode/aarch64.h (AARCH64_OPND_SVE_ADDR_RI_S4x16) (AARCH64_OPND_SVE_IMM_ROT1, AARCH64_OPND_SVE_IMM_ROT2) (AARCH64_OPND_SVE_Zm3_INDEX, AARCH64_OPND_SVE_Zm3_22_INDEX) (AARCH64_OPND_SVE_Zm4_INDEX): New aarch64_opnds. opcodes/ * aarch64-tbl.h (OP_SVE_HMH, OP_SVE_VMU_HSD, OP_SVE_VMVU_HSD) (OP_SVE_VMVV_HSD, OP_SVE_VMVVU_HSD, OP_SVE_VM_HSD, OP_SVE_VUVV_HSD) (OP_SVE_VUV_HSD, OP_SVE_VU_HSD, OP_SVE_VVVU_H, OP_SVE_VVVU_S) (OP_SVE_VVVU_HSD, OP_SVE_VVV_D, OP_SVE_VVV_D_H, OP_SVE_VVV_H) (OP_SVE_VVV_HSD, OP_SVE_VVV_S, OP_SVE_VVV_S_B, OP_SVE_VVV_SD_BH) (OP_SVE_VV_BHSDQ, OP_SVE_VV_HSD, OP_SVE_VZVV_HSD, OP_SVE_VZV_HSD) (OP_SVE_V_HSD): New macros. (OP_SVE_VMU_SD, OP_SVE_VMVU_SD, OP_SVE_VM_SD, OP_SVE_VUVV_SD) (OP_SVE_VU_SD, OP_SVE_VVVU_SD, OP_SVE_VVV_SD, OP_SVE_VZVV_SD) (OP_SVE_VZV_SD, OP_SVE_V_SD): Delete. (aarch64_opcode_table): Add new SVE instructions. (aarch64_opcode_table): Use imm_rotate{1,2} instead of imm_rotate for rotation operands. Add new SVE operands. * aarch64-asm.h (ins_sve_addr_ri_s4): New inserter. (ins_sve_quad_index): Likewise. (ins_imm_rotate): Split into... (ins_imm_rotate1, ins_imm_rotate2): ...these two inserters. * aarch64-asm.c (aarch64_ins_imm_rotate): Split into... (aarch64_ins_imm_rotate1, aarch64_ins_imm_rotate2): ...these two functions. (aarch64_ins_sve_addr_ri_s4): New function. (aarch64_ins_sve_quad_index): Likewise. (do_misc_encoding): Handle "MOV Zn.Q, Qm". * aarch64-asm-2.c: Regenerate. * aarch64-dis.h (ext_sve_addr_ri_s4): New extractor. (ext_sve_quad_index): Likewise. (ext_imm_rotate): Split into... (ext_imm_rotate1, ext_imm_rotate2): ...these two extractors. * aarch64-dis.c (aarch64_ext_imm_rotate): Split into... (aarch64_ext_imm_rotate1, aarch64_ext_imm_rotate2): ...these two functions. (aarch64_ext_sve_addr_ri_s4): New function. (aarch64_ext_sve_quad_index): Likewise. (aarch64_ext_sve_index): Allow quad indices. (do_misc_decoding): Likewise. * aarch64-dis-2.c: Regenerate. * aarch64-opc.h (FLD_SVE_i3h, FLD_SVE_rot1, FLD_SVE_rot2): New aarch64_field_kinds. (OPD_F_OD_MASK): Widen by one bit. (OPD_F_NO_ZR): Bump accordingly. (get_operand_field_width): New function. * aarch64-opc.c (fields): Add new SVE fields. (operand_general_constraint_met_p): Handle new SVE operands. (aarch64_print_operand): Likewise. * aarch64-opc-2.c: Regenerate. gas/ * doc/c-aarch64.texi: Document that sve implies fp16, simd and compnum. * config/tc-aarch64.c (parse_vector_type_for_operand): Allow .q to be used with SVE registers. (parse_operands): Handle new SVE operands. (aarch64_features): Make "sve" require F16 rather than FP. Also require COMPNUM. * testsuite/gas/aarch64/sve.s: Add tests for new instructions. Include compnum tests. * testsuite/gas/aarch64/sve.d: Update accordingly. * testsuite/gas/aarch64/sve-invalid.s: Add tests for new instructions. * testsuite/gas/aarch64/sve-invalid.l: Update accordingly. Also update expected output for new FMOV and MOV alternatives.
2017-02-25 02:29:00 +08:00
case AARCH64_OPND_SVE_IMM_ROT1:
case AARCH64_OPND_SVE_IMM_ROT2:
case AARCH64_OPND_SVE_IMM_ROT3:
case AARCH64_OPND_CSSC_SIMM8:
case AARCH64_OPND_CSSC_UIMM8:
libopcodes/aarch64: add support for disassembler styling This commit enables disassembler styling for AArch64. After this commit it is possible to have objdump style AArch64 disassembler output (using --disassembler-color option). Once the required GDB patches are merged, GDB will also style the disassembler output. The changes to support styling are mostly split between two files opcodes/aarch64-dis.c and opcodes/aarch64-opc.c. The entry point for the AArch64 disassembler can be found in aarch64-dis.c, this file handles printing the instruction mnemonics, and assembler directives (e.g. '.byte', '.word', etc). Some operands, mostly relating to assembler directives are also printed from this file. This commit changes all of this to pass through suitable styling information. However, for most "normal" instructions, the instruction operands are printed using a two step process. From aarch64-dis.c, in the print_operands function, the function aarch64_print_operand is called, this function is in aarch64-opc.c, and converts an instruction operand into a string. Then, back in print_operands (aarch64-dis.c), the operand string is printed. Unfortunately, the string returned by aarch64_print_operand can be quite complex, it will include syntax elements, like '[' and ']', in addition to register names and immediate values. In some cases, a single operand will expand into what will appear (to the user) as multiple operands separated with a ','. This makes the task of styling more complex, all these different components need to by styled differently, so we need to get the styling information out of aarch64_print_operand in some way. The solution that I propose here is similar to the solution that I used for the i386 disassembler. Currently, aarch64_print_operand uses snprintf to write the operand text into a buffer provided by the caller. What I propose is that we pass an extra argument to the aarch64_print_operand function, this argument will be a structure, the structure contains a callback function and some state. When aarch64_print_operand needs to format part of its output this can be done by using the callback function within the new structure, this callback returns a string with special embedded markers that indicate which mode should be used for each piece of text. Back in aarch64-dis.c we can spot these special style markers and use this to split the disassembler output up and apply the correct style to each piece. To make aarch64-opc.c clearer a series of new static functions have been added, e.g. 'style_reg', 'style_imm', etc. Each of these functions formats a piece of text in a different style, 'register' and 'immediate' in this case. Here's an example taken from aarch64-opc.c of the new functions in use: snprintf (buf, size, "[%s, %s]!", style_reg (styler, base), style_imm (styler, "#%d", opnd->addr.offset.imm)); The aarch64_print_operand function is also called from the assembler to aid in printing diagnostic messages. Right now I have no plans to add styling to the assembler output, and so, the callback function used in the assembler ignores the styling information and just returns an plain string. I've used the source files in gas/testsuite/gas/aarch64/ for testing, and have manually gone through and checked that the styling looks reasonable, however, I'm not an AArch64 expert, so it is possible that the odd piece is styled incorrectly. Please point out any mistakes I've made. With objdump disassembler color turned off, there should be no change in the output after this commit.
2022-04-28 20:31:07 +08:00
snprintf (buf, size, "%s",
style_imm (styler, "#%" PRIi64, opnd->imm.value));
break;
[AArch64][SVE 28/32] Add SVE FP immediate operands This patch adds support for the new SVE floating-point immediate operands. One operand uses the same 8-bit encoding as base AArch64, but in a different position. The others use a single bit to select between two values. One of the single-bit operands is a choice between 0 and 1, where 0 is not a valid 8-bit encoding. I think the cleanest way of handling these single-bit immediates is therefore to use the IEEE float encoding itself as the immediate value and select between the two possible values when encoding and decoding. As described in the covering note for the patch that added F_STRICT, we get better error messages by accepting unsuffixed vector registers and leaving the qualifier matching code to report an error. This means that we carry on parsing the other operands, and so can try to parse FP immediates for invalid instructions like: fcpy z0, #2.5 In this case there is no suffix to tell us whether the immediate should be treated as single or double precision. Again, we get better error messages by picking one (arbitrary) immediate size and reporting an error for the missing suffix later. include/ * opcode/aarch64.h (AARCH64_OPND_SVE_FPIMM8): New aarch64_opnd. (AARCH64_OPND_SVE_I1_HALF_ONE, AARCH64_OPND_SVE_I1_HALF_TWO) (AARCH64_OPND_SVE_I1_ZERO_ONE): Likewise. opcodes/ * aarch64-tbl.h (AARCH64_OPERANDS): Add entries for the new SVE FP immediate operands. * aarch64-opc.h (FLD_SVE_i1): New aarch64_field_kind. * aarch64-opc.c (fields): Add corresponding entry. (operand_general_constraint_met_p): Handle the new SVE FP immediate operands. (aarch64_print_operand): Likewise. * aarch64-opc-2.c: Regenerate. * aarch64-asm.h (ins_sve_float_half_one, ins_sve_float_half_two) (ins_sve_float_zero_one): New inserters. * aarch64-asm.c (aarch64_ins_sve_float_half_one): New function. (aarch64_ins_sve_float_half_two): Likewise. (aarch64_ins_sve_float_zero_one): Likewise. * aarch64-asm-2.c: Regenerate. * aarch64-dis.h (ext_sve_float_half_one, ext_sve_float_half_two) (ext_sve_float_zero_one): New extractors. * aarch64-dis.c (aarch64_ext_sve_float_half_one): New function. (aarch64_ext_sve_float_half_two): Likewise. (aarch64_ext_sve_float_zero_one): Likewise. * aarch64-dis-2.c: Regenerate. gas/ * config/tc-aarch64.c (double_precision_operand_p): New function. (parse_operands): Use it to calculate the dp_p input to parse_aarch64_imm_float. Handle the new SVE FP immediate operands.
2016-09-21 23:57:22 +08:00
case AARCH64_OPND_SVE_I1_HALF_ONE:
case AARCH64_OPND_SVE_I1_HALF_TWO:
case AARCH64_OPND_SVE_I1_ZERO_ONE:
{
single_conv_t c;
c.i = opnd->imm.value;
libopcodes/aarch64: add support for disassembler styling This commit enables disassembler styling for AArch64. After this commit it is possible to have objdump style AArch64 disassembler output (using --disassembler-color option). Once the required GDB patches are merged, GDB will also style the disassembler output. The changes to support styling are mostly split between two files opcodes/aarch64-dis.c and opcodes/aarch64-opc.c. The entry point for the AArch64 disassembler can be found in aarch64-dis.c, this file handles printing the instruction mnemonics, and assembler directives (e.g. '.byte', '.word', etc). Some operands, mostly relating to assembler directives are also printed from this file. This commit changes all of this to pass through suitable styling information. However, for most "normal" instructions, the instruction operands are printed using a two step process. From aarch64-dis.c, in the print_operands function, the function aarch64_print_operand is called, this function is in aarch64-opc.c, and converts an instruction operand into a string. Then, back in print_operands (aarch64-dis.c), the operand string is printed. Unfortunately, the string returned by aarch64_print_operand can be quite complex, it will include syntax elements, like '[' and ']', in addition to register names and immediate values. In some cases, a single operand will expand into what will appear (to the user) as multiple operands separated with a ','. This makes the task of styling more complex, all these different components need to by styled differently, so we need to get the styling information out of aarch64_print_operand in some way. The solution that I propose here is similar to the solution that I used for the i386 disassembler. Currently, aarch64_print_operand uses snprintf to write the operand text into a buffer provided by the caller. What I propose is that we pass an extra argument to the aarch64_print_operand function, this argument will be a structure, the structure contains a callback function and some state. When aarch64_print_operand needs to format part of its output this can be done by using the callback function within the new structure, this callback returns a string with special embedded markers that indicate which mode should be used for each piece of text. Back in aarch64-dis.c we can spot these special style markers and use this to split the disassembler output up and apply the correct style to each piece. To make aarch64-opc.c clearer a series of new static functions have been added, e.g. 'style_reg', 'style_imm', etc. Each of these functions formats a piece of text in a different style, 'register' and 'immediate' in this case. Here's an example taken from aarch64-opc.c of the new functions in use: snprintf (buf, size, "[%s, %s]!", style_reg (styler, base), style_imm (styler, "#%d", opnd->addr.offset.imm)); The aarch64_print_operand function is also called from the assembler to aid in printing diagnostic messages. Right now I have no plans to add styling to the assembler output, and so, the callback function used in the assembler ignores the styling information and just returns an plain string. I've used the source files in gas/testsuite/gas/aarch64/ for testing, and have manually gone through and checked that the styling looks reasonable, however, I'm not an AArch64 expert, so it is possible that the odd piece is styled incorrectly. Please point out any mistakes I've made. With objdump disassembler color turned off, there should be no change in the output after this commit.
2022-04-28 20:31:07 +08:00
snprintf (buf, size, "%s", style_imm (styler, "#%.1f", c.f));
[AArch64][SVE 28/32] Add SVE FP immediate operands This patch adds support for the new SVE floating-point immediate operands. One operand uses the same 8-bit encoding as base AArch64, but in a different position. The others use a single bit to select between two values. One of the single-bit operands is a choice between 0 and 1, where 0 is not a valid 8-bit encoding. I think the cleanest way of handling these single-bit immediates is therefore to use the IEEE float encoding itself as the immediate value and select between the two possible values when encoding and decoding. As described in the covering note for the patch that added F_STRICT, we get better error messages by accepting unsuffixed vector registers and leaving the qualifier matching code to report an error. This means that we carry on parsing the other operands, and so can try to parse FP immediates for invalid instructions like: fcpy z0, #2.5 In this case there is no suffix to tell us whether the immediate should be treated as single or double precision. Again, we get better error messages by picking one (arbitrary) immediate size and reporting an error for the missing suffix later. include/ * opcode/aarch64.h (AARCH64_OPND_SVE_FPIMM8): New aarch64_opnd. (AARCH64_OPND_SVE_I1_HALF_ONE, AARCH64_OPND_SVE_I1_HALF_TWO) (AARCH64_OPND_SVE_I1_ZERO_ONE): Likewise. opcodes/ * aarch64-tbl.h (AARCH64_OPERANDS): Add entries for the new SVE FP immediate operands. * aarch64-opc.h (FLD_SVE_i1): New aarch64_field_kind. * aarch64-opc.c (fields): Add corresponding entry. (operand_general_constraint_met_p): Handle the new SVE FP immediate operands. (aarch64_print_operand): Likewise. * aarch64-opc-2.c: Regenerate. * aarch64-asm.h (ins_sve_float_half_one, ins_sve_float_half_two) (ins_sve_float_zero_one): New inserters. * aarch64-asm.c (aarch64_ins_sve_float_half_one): New function. (aarch64_ins_sve_float_half_two): Likewise. (aarch64_ins_sve_float_zero_one): Likewise. * aarch64-asm-2.c: Regenerate. * aarch64-dis.h (ext_sve_float_half_one, ext_sve_float_half_two) (ext_sve_float_zero_one): New extractors. * aarch64-dis.c (aarch64_ext_sve_float_half_one): New function. (aarch64_ext_sve_float_half_two): Likewise. (aarch64_ext_sve_float_zero_one): Likewise. * aarch64-dis-2.c: Regenerate. gas/ * config/tc-aarch64.c (double_precision_operand_p): New function. (parse_operands): Use it to calculate the dp_p input to parse_aarch64_imm_float. Handle the new SVE FP immediate operands.
2016-09-21 23:57:22 +08:00
break;
}
[AArch64][SVE 23/32] Add SVE pattern and prfop operands The SVE instructions have two enumerated operands: one to select a vector pattern and another to select a prefetch operation. The latter is a cut-down version of the base AArch64 prefetch operation. Both types of operand can also be specified as raw enum values such as #31. Reserved values can only be specified this way. If it hadn't been for the pattern operand, I would have been tempted to use the existing parsing for prefetch operations and add extra checks for SVE. However, since the patterns needed new enum parsing code anyway, it seeemed cleaner to reuse it for the prefetches too. Because of the small number of enum values, I don't think we'd gain anything by using hash tables. include/ * opcode/aarch64.h (AARCH64_OPND_SVE_PATTERN): New aarch64_opnd. (AARCH64_OPND_SVE_PRFOP): Likewise. (aarch64_sve_pattern_array): Declare. (aarch64_sve_prfop_array): Likewise. opcodes/ * aarch64-tbl.h (AARCH64_OPERANDS): Add entries for AARCH64_OPND_SVE_PATTERN and AARCH64_OPND_SVE_PRFOP. * aarch64-opc.h (FLD_SVE_pattern): New aarch64_field_kind. (FLD_SVE_prfop): Likewise. * aarch64-opc.c: Include libiberty.h. (aarch64_sve_pattern_array): New variable. (aarch64_sve_prfop_array): Likewise. (fields): Add entries for FLD_SVE_pattern and FLD_SVE_prfop. (aarch64_print_operand): Handle AARCH64_OPND_SVE_PATTERN and AARCH64_OPND_SVE_PRFOP. * aarch64-asm-2.c: Regenerate. * aarch64-dis-2.c: Likewise. * aarch64-opc-2.c: Likewise. gas/ * config/tc-aarch64.c (parse_enum_string): New function. (po_enum_or_fail): New macro. (parse_operands): Handle AARCH64_OPND_SVE_PATTERN and AARCH64_OPND_SVE_PRFOP.
2016-09-21 23:54:53 +08:00
case AARCH64_OPND_SVE_PATTERN:
if (optional_operand_p (opcode, idx)
&& opnd->imm.value == get_optional_operand_default_value (opcode))
break;
enum_value = opnd->imm.value;
assert (enum_value < ARRAY_SIZE (aarch64_sve_pattern_array));
if (aarch64_sve_pattern_array[enum_value])
libopcodes/aarch64: add support for disassembler styling This commit enables disassembler styling for AArch64. After this commit it is possible to have objdump style AArch64 disassembler output (using --disassembler-color option). Once the required GDB patches are merged, GDB will also style the disassembler output. The changes to support styling are mostly split between two files opcodes/aarch64-dis.c and opcodes/aarch64-opc.c. The entry point for the AArch64 disassembler can be found in aarch64-dis.c, this file handles printing the instruction mnemonics, and assembler directives (e.g. '.byte', '.word', etc). Some operands, mostly relating to assembler directives are also printed from this file. This commit changes all of this to pass through suitable styling information. However, for most "normal" instructions, the instruction operands are printed using a two step process. From aarch64-dis.c, in the print_operands function, the function aarch64_print_operand is called, this function is in aarch64-opc.c, and converts an instruction operand into a string. Then, back in print_operands (aarch64-dis.c), the operand string is printed. Unfortunately, the string returned by aarch64_print_operand can be quite complex, it will include syntax elements, like '[' and ']', in addition to register names and immediate values. In some cases, a single operand will expand into what will appear (to the user) as multiple operands separated with a ','. This makes the task of styling more complex, all these different components need to by styled differently, so we need to get the styling information out of aarch64_print_operand in some way. The solution that I propose here is similar to the solution that I used for the i386 disassembler. Currently, aarch64_print_operand uses snprintf to write the operand text into a buffer provided by the caller. What I propose is that we pass an extra argument to the aarch64_print_operand function, this argument will be a structure, the structure contains a callback function and some state. When aarch64_print_operand needs to format part of its output this can be done by using the callback function within the new structure, this callback returns a string with special embedded markers that indicate which mode should be used for each piece of text. Back in aarch64-dis.c we can spot these special style markers and use this to split the disassembler output up and apply the correct style to each piece. To make aarch64-opc.c clearer a series of new static functions have been added, e.g. 'style_reg', 'style_imm', etc. Each of these functions formats a piece of text in a different style, 'register' and 'immediate' in this case. Here's an example taken from aarch64-opc.c of the new functions in use: snprintf (buf, size, "[%s, %s]!", style_reg (styler, base), style_imm (styler, "#%d", opnd->addr.offset.imm)); The aarch64_print_operand function is also called from the assembler to aid in printing diagnostic messages. Right now I have no plans to add styling to the assembler output, and so, the callback function used in the assembler ignores the styling information and just returns an plain string. I've used the source files in gas/testsuite/gas/aarch64/ for testing, and have manually gone through and checked that the styling looks reasonable, however, I'm not an AArch64 expert, so it is possible that the odd piece is styled incorrectly. Please point out any mistakes I've made. With objdump disassembler color turned off, there should be no change in the output after this commit.
2022-04-28 20:31:07 +08:00
snprintf (buf, size, "%s",
style_reg (styler, aarch64_sve_pattern_array[enum_value]));
[AArch64][SVE 23/32] Add SVE pattern and prfop operands The SVE instructions have two enumerated operands: one to select a vector pattern and another to select a prefetch operation. The latter is a cut-down version of the base AArch64 prefetch operation. Both types of operand can also be specified as raw enum values such as #31. Reserved values can only be specified this way. If it hadn't been for the pattern operand, I would have been tempted to use the existing parsing for prefetch operations and add extra checks for SVE. However, since the patterns needed new enum parsing code anyway, it seeemed cleaner to reuse it for the prefetches too. Because of the small number of enum values, I don't think we'd gain anything by using hash tables. include/ * opcode/aarch64.h (AARCH64_OPND_SVE_PATTERN): New aarch64_opnd. (AARCH64_OPND_SVE_PRFOP): Likewise. (aarch64_sve_pattern_array): Declare. (aarch64_sve_prfop_array): Likewise. opcodes/ * aarch64-tbl.h (AARCH64_OPERANDS): Add entries for AARCH64_OPND_SVE_PATTERN and AARCH64_OPND_SVE_PRFOP. * aarch64-opc.h (FLD_SVE_pattern): New aarch64_field_kind. (FLD_SVE_prfop): Likewise. * aarch64-opc.c: Include libiberty.h. (aarch64_sve_pattern_array): New variable. (aarch64_sve_prfop_array): Likewise. (fields): Add entries for FLD_SVE_pattern and FLD_SVE_prfop. (aarch64_print_operand): Handle AARCH64_OPND_SVE_PATTERN and AARCH64_OPND_SVE_PRFOP. * aarch64-asm-2.c: Regenerate. * aarch64-dis-2.c: Likewise. * aarch64-opc-2.c: Likewise. gas/ * config/tc-aarch64.c (parse_enum_string): New function. (po_enum_or_fail): New macro. (parse_operands): Handle AARCH64_OPND_SVE_PATTERN and AARCH64_OPND_SVE_PRFOP.
2016-09-21 23:54:53 +08:00
else
libopcodes/aarch64: add support for disassembler styling This commit enables disassembler styling for AArch64. After this commit it is possible to have objdump style AArch64 disassembler output (using --disassembler-color option). Once the required GDB patches are merged, GDB will also style the disassembler output. The changes to support styling are mostly split between two files opcodes/aarch64-dis.c and opcodes/aarch64-opc.c. The entry point for the AArch64 disassembler can be found in aarch64-dis.c, this file handles printing the instruction mnemonics, and assembler directives (e.g. '.byte', '.word', etc). Some operands, mostly relating to assembler directives are also printed from this file. This commit changes all of this to pass through suitable styling information. However, for most "normal" instructions, the instruction operands are printed using a two step process. From aarch64-dis.c, in the print_operands function, the function aarch64_print_operand is called, this function is in aarch64-opc.c, and converts an instruction operand into a string. Then, back in print_operands (aarch64-dis.c), the operand string is printed. Unfortunately, the string returned by aarch64_print_operand can be quite complex, it will include syntax elements, like '[' and ']', in addition to register names and immediate values. In some cases, a single operand will expand into what will appear (to the user) as multiple operands separated with a ','. This makes the task of styling more complex, all these different components need to by styled differently, so we need to get the styling information out of aarch64_print_operand in some way. The solution that I propose here is similar to the solution that I used for the i386 disassembler. Currently, aarch64_print_operand uses snprintf to write the operand text into a buffer provided by the caller. What I propose is that we pass an extra argument to the aarch64_print_operand function, this argument will be a structure, the structure contains a callback function and some state. When aarch64_print_operand needs to format part of its output this can be done by using the callback function within the new structure, this callback returns a string with special embedded markers that indicate which mode should be used for each piece of text. Back in aarch64-dis.c we can spot these special style markers and use this to split the disassembler output up and apply the correct style to each piece. To make aarch64-opc.c clearer a series of new static functions have been added, e.g. 'style_reg', 'style_imm', etc. Each of these functions formats a piece of text in a different style, 'register' and 'immediate' in this case. Here's an example taken from aarch64-opc.c of the new functions in use: snprintf (buf, size, "[%s, %s]!", style_reg (styler, base), style_imm (styler, "#%d", opnd->addr.offset.imm)); The aarch64_print_operand function is also called from the assembler to aid in printing diagnostic messages. Right now I have no plans to add styling to the assembler output, and so, the callback function used in the assembler ignores the styling information and just returns an plain string. I've used the source files in gas/testsuite/gas/aarch64/ for testing, and have manually gone through and checked that the styling looks reasonable, however, I'm not an AArch64 expert, so it is possible that the odd piece is styled incorrectly. Please point out any mistakes I've made. With objdump disassembler color turned off, there should be no change in the output after this commit.
2022-04-28 20:31:07 +08:00
snprintf (buf, size, "%s",
style_imm (styler, "#%" PRIi64, opnd->imm.value));
[AArch64][SVE 23/32] Add SVE pattern and prfop operands The SVE instructions have two enumerated operands: one to select a vector pattern and another to select a prefetch operation. The latter is a cut-down version of the base AArch64 prefetch operation. Both types of operand can also be specified as raw enum values such as #31. Reserved values can only be specified this way. If it hadn't been for the pattern operand, I would have been tempted to use the existing parsing for prefetch operations and add extra checks for SVE. However, since the patterns needed new enum parsing code anyway, it seeemed cleaner to reuse it for the prefetches too. Because of the small number of enum values, I don't think we'd gain anything by using hash tables. include/ * opcode/aarch64.h (AARCH64_OPND_SVE_PATTERN): New aarch64_opnd. (AARCH64_OPND_SVE_PRFOP): Likewise. (aarch64_sve_pattern_array): Declare. (aarch64_sve_prfop_array): Likewise. opcodes/ * aarch64-tbl.h (AARCH64_OPERANDS): Add entries for AARCH64_OPND_SVE_PATTERN and AARCH64_OPND_SVE_PRFOP. * aarch64-opc.h (FLD_SVE_pattern): New aarch64_field_kind. (FLD_SVE_prfop): Likewise. * aarch64-opc.c: Include libiberty.h. (aarch64_sve_pattern_array): New variable. (aarch64_sve_prfop_array): Likewise. (fields): Add entries for FLD_SVE_pattern and FLD_SVE_prfop. (aarch64_print_operand): Handle AARCH64_OPND_SVE_PATTERN and AARCH64_OPND_SVE_PRFOP. * aarch64-asm-2.c: Regenerate. * aarch64-dis-2.c: Likewise. * aarch64-opc-2.c: Likewise. gas/ * config/tc-aarch64.c (parse_enum_string): New function. (po_enum_or_fail): New macro. (parse_operands): Handle AARCH64_OPND_SVE_PATTERN and AARCH64_OPND_SVE_PRFOP.
2016-09-21 23:54:53 +08:00
break;
[AArch64][SVE 24/32] Add AARCH64_OPND_SVE_PATTERN_SCALED Some SVE instructions count the number of elements in a given vector pattern and allow a scale factor of [1, 16] to be applied to the result. This scale factor is written ", MUL #n", where "MUL" is a new operator. E.g.: UQINCD X0, POW2, MUL #2 This patch adds support for this kind of operand. All existing operators were shifts of some kind, so there was a natural range of [0, 63] regardless of context. This was then narrowered further by later checks (e.g. to [0, 31] when used for 32-bit values). In contrast, MUL doesn't really have a natural context-independent range. Rather than pick one arbitrarily, it seemed better to make the "shift" amount a full 64-bit value and leave the range test to the usual operand-checking code. I've rearranged the fields of aarch64_opnd_info so that this doesn't increase the size of the structure (although I don't think its size is critical anyway). include/ * opcode/aarch64.h (AARCH64_OPND_SVE_PATTERN_SCALED): New aarch64_opnd. (AARCH64_MOD_MUL): New aarch64_modifier_kind. (aarch64_opnd_info): Make shifter.amount an int64_t and rearrange the fields. opcodes/ * aarch64-tbl.h (AARCH64_OPERANDS): Add an entry for AARCH64_OPND_SVE_PATTERN_SCALED. * aarch64-opc.h (FLD_SVE_imm4): New aarch64_field_kind. * aarch64-opc.c (fields): Add a corresponding entry. (set_multiplier_out_of_range_error): New function. (aarch64_operand_modifiers): Add entry for AARCH64_MOD_MUL. (operand_general_constraint_met_p): Handle AARCH64_OPND_SVE_PATTERN_SCALED. (print_register_offset_address): Use PRIi64 to print the shift amount. (aarch64_print_operand): Likewise. Handle AARCH64_OPND_SVE_PATTERN_SCALED. * aarch64-opc-2.c: Regenerate. * aarch64-asm.h (ins_sve_scale): New inserter. * aarch64-asm.c (aarch64_ins_sve_scale): New function. * aarch64-asm-2.c: Regenerate. * aarch64-dis.h (ext_sve_scale): New inserter. * aarch64-dis.c (aarch64_ext_sve_scale): New function. * aarch64-dis-2.c: Regenerate. gas/ * config/tc-aarch64.c (SHIFTED_MUL): New parse_shift_mode. (parse_shift): Handle it. Reject AARCH64_MOD_MUL for all other shift modes. Skip range tests for AARCH64_MOD_MUL. (process_omitted_operand): Handle AARCH64_OPND_SVE_PATTERN_SCALED. (parse_operands): Likewise.
2016-09-21 23:55:22 +08:00
case AARCH64_OPND_SVE_PATTERN_SCALED:
if (optional_operand_p (opcode, idx)
&& !opnd->shifter.operator_present
&& opnd->imm.value == get_optional_operand_default_value (opcode))
break;
enum_value = opnd->imm.value;
assert (enum_value < ARRAY_SIZE (aarch64_sve_pattern_array));
if (aarch64_sve_pattern_array[opnd->imm.value])
libopcodes/aarch64: add support for disassembler styling This commit enables disassembler styling for AArch64. After this commit it is possible to have objdump style AArch64 disassembler output (using --disassembler-color option). Once the required GDB patches are merged, GDB will also style the disassembler output. The changes to support styling are mostly split between two files opcodes/aarch64-dis.c and opcodes/aarch64-opc.c. The entry point for the AArch64 disassembler can be found in aarch64-dis.c, this file handles printing the instruction mnemonics, and assembler directives (e.g. '.byte', '.word', etc). Some operands, mostly relating to assembler directives are also printed from this file. This commit changes all of this to pass through suitable styling information. However, for most "normal" instructions, the instruction operands are printed using a two step process. From aarch64-dis.c, in the print_operands function, the function aarch64_print_operand is called, this function is in aarch64-opc.c, and converts an instruction operand into a string. Then, back in print_operands (aarch64-dis.c), the operand string is printed. Unfortunately, the string returned by aarch64_print_operand can be quite complex, it will include syntax elements, like '[' and ']', in addition to register names and immediate values. In some cases, a single operand will expand into what will appear (to the user) as multiple operands separated with a ','. This makes the task of styling more complex, all these different components need to by styled differently, so we need to get the styling information out of aarch64_print_operand in some way. The solution that I propose here is similar to the solution that I used for the i386 disassembler. Currently, aarch64_print_operand uses snprintf to write the operand text into a buffer provided by the caller. What I propose is that we pass an extra argument to the aarch64_print_operand function, this argument will be a structure, the structure contains a callback function and some state. When aarch64_print_operand needs to format part of its output this can be done by using the callback function within the new structure, this callback returns a string with special embedded markers that indicate which mode should be used for each piece of text. Back in aarch64-dis.c we can spot these special style markers and use this to split the disassembler output up and apply the correct style to each piece. To make aarch64-opc.c clearer a series of new static functions have been added, e.g. 'style_reg', 'style_imm', etc. Each of these functions formats a piece of text in a different style, 'register' and 'immediate' in this case. Here's an example taken from aarch64-opc.c of the new functions in use: snprintf (buf, size, "[%s, %s]!", style_reg (styler, base), style_imm (styler, "#%d", opnd->addr.offset.imm)); The aarch64_print_operand function is also called from the assembler to aid in printing diagnostic messages. Right now I have no plans to add styling to the assembler output, and so, the callback function used in the assembler ignores the styling information and just returns an plain string. I've used the source files in gas/testsuite/gas/aarch64/ for testing, and have manually gone through and checked that the styling looks reasonable, however, I'm not an AArch64 expert, so it is possible that the odd piece is styled incorrectly. Please point out any mistakes I've made. With objdump disassembler color turned off, there should be no change in the output after this commit.
2022-04-28 20:31:07 +08:00
snprintf (buf, size, "%s",
style_reg (styler,
aarch64_sve_pattern_array[opnd->imm.value]));
[AArch64][SVE 24/32] Add AARCH64_OPND_SVE_PATTERN_SCALED Some SVE instructions count the number of elements in a given vector pattern and allow a scale factor of [1, 16] to be applied to the result. This scale factor is written ", MUL #n", where "MUL" is a new operator. E.g.: UQINCD X0, POW2, MUL #2 This patch adds support for this kind of operand. All existing operators were shifts of some kind, so there was a natural range of [0, 63] regardless of context. This was then narrowered further by later checks (e.g. to [0, 31] when used for 32-bit values). In contrast, MUL doesn't really have a natural context-independent range. Rather than pick one arbitrarily, it seemed better to make the "shift" amount a full 64-bit value and leave the range test to the usual operand-checking code. I've rearranged the fields of aarch64_opnd_info so that this doesn't increase the size of the structure (although I don't think its size is critical anyway). include/ * opcode/aarch64.h (AARCH64_OPND_SVE_PATTERN_SCALED): New aarch64_opnd. (AARCH64_MOD_MUL): New aarch64_modifier_kind. (aarch64_opnd_info): Make shifter.amount an int64_t and rearrange the fields. opcodes/ * aarch64-tbl.h (AARCH64_OPERANDS): Add an entry for AARCH64_OPND_SVE_PATTERN_SCALED. * aarch64-opc.h (FLD_SVE_imm4): New aarch64_field_kind. * aarch64-opc.c (fields): Add a corresponding entry. (set_multiplier_out_of_range_error): New function. (aarch64_operand_modifiers): Add entry for AARCH64_MOD_MUL. (operand_general_constraint_met_p): Handle AARCH64_OPND_SVE_PATTERN_SCALED. (print_register_offset_address): Use PRIi64 to print the shift amount. (aarch64_print_operand): Likewise. Handle AARCH64_OPND_SVE_PATTERN_SCALED. * aarch64-opc-2.c: Regenerate. * aarch64-asm.h (ins_sve_scale): New inserter. * aarch64-asm.c (aarch64_ins_sve_scale): New function. * aarch64-asm-2.c: Regenerate. * aarch64-dis.h (ext_sve_scale): New inserter. * aarch64-dis.c (aarch64_ext_sve_scale): New function. * aarch64-dis-2.c: Regenerate. gas/ * config/tc-aarch64.c (SHIFTED_MUL): New parse_shift_mode. (parse_shift): Handle it. Reject AARCH64_MOD_MUL for all other shift modes. Skip range tests for AARCH64_MOD_MUL. (process_omitted_operand): Handle AARCH64_OPND_SVE_PATTERN_SCALED. (parse_operands): Likewise.
2016-09-21 23:55:22 +08:00
else
libopcodes/aarch64: add support for disassembler styling This commit enables disassembler styling for AArch64. After this commit it is possible to have objdump style AArch64 disassembler output (using --disassembler-color option). Once the required GDB patches are merged, GDB will also style the disassembler output. The changes to support styling are mostly split between two files opcodes/aarch64-dis.c and opcodes/aarch64-opc.c. The entry point for the AArch64 disassembler can be found in aarch64-dis.c, this file handles printing the instruction mnemonics, and assembler directives (e.g. '.byte', '.word', etc). Some operands, mostly relating to assembler directives are also printed from this file. This commit changes all of this to pass through suitable styling information. However, for most "normal" instructions, the instruction operands are printed using a two step process. From aarch64-dis.c, in the print_operands function, the function aarch64_print_operand is called, this function is in aarch64-opc.c, and converts an instruction operand into a string. Then, back in print_operands (aarch64-dis.c), the operand string is printed. Unfortunately, the string returned by aarch64_print_operand can be quite complex, it will include syntax elements, like '[' and ']', in addition to register names and immediate values. In some cases, a single operand will expand into what will appear (to the user) as multiple operands separated with a ','. This makes the task of styling more complex, all these different components need to by styled differently, so we need to get the styling information out of aarch64_print_operand in some way. The solution that I propose here is similar to the solution that I used for the i386 disassembler. Currently, aarch64_print_operand uses snprintf to write the operand text into a buffer provided by the caller. What I propose is that we pass an extra argument to the aarch64_print_operand function, this argument will be a structure, the structure contains a callback function and some state. When aarch64_print_operand needs to format part of its output this can be done by using the callback function within the new structure, this callback returns a string with special embedded markers that indicate which mode should be used for each piece of text. Back in aarch64-dis.c we can spot these special style markers and use this to split the disassembler output up and apply the correct style to each piece. To make aarch64-opc.c clearer a series of new static functions have been added, e.g. 'style_reg', 'style_imm', etc. Each of these functions formats a piece of text in a different style, 'register' and 'immediate' in this case. Here's an example taken from aarch64-opc.c of the new functions in use: snprintf (buf, size, "[%s, %s]!", style_reg (styler, base), style_imm (styler, "#%d", opnd->addr.offset.imm)); The aarch64_print_operand function is also called from the assembler to aid in printing diagnostic messages. Right now I have no plans to add styling to the assembler output, and so, the callback function used in the assembler ignores the styling information and just returns an plain string. I've used the source files in gas/testsuite/gas/aarch64/ for testing, and have manually gone through and checked that the styling looks reasonable, however, I'm not an AArch64 expert, so it is possible that the odd piece is styled incorrectly. Please point out any mistakes I've made. With objdump disassembler color turned off, there should be no change in the output after this commit.
2022-04-28 20:31:07 +08:00
snprintf (buf, size, "%s",
style_imm (styler, "#%" PRIi64, opnd->imm.value));
[AArch64][SVE 24/32] Add AARCH64_OPND_SVE_PATTERN_SCALED Some SVE instructions count the number of elements in a given vector pattern and allow a scale factor of [1, 16] to be applied to the result. This scale factor is written ", MUL #n", where "MUL" is a new operator. E.g.: UQINCD X0, POW2, MUL #2 This patch adds support for this kind of operand. All existing operators were shifts of some kind, so there was a natural range of [0, 63] regardless of context. This was then narrowered further by later checks (e.g. to [0, 31] when used for 32-bit values). In contrast, MUL doesn't really have a natural context-independent range. Rather than pick one arbitrarily, it seemed better to make the "shift" amount a full 64-bit value and leave the range test to the usual operand-checking code. I've rearranged the fields of aarch64_opnd_info so that this doesn't increase the size of the structure (although I don't think its size is critical anyway). include/ * opcode/aarch64.h (AARCH64_OPND_SVE_PATTERN_SCALED): New aarch64_opnd. (AARCH64_MOD_MUL): New aarch64_modifier_kind. (aarch64_opnd_info): Make shifter.amount an int64_t and rearrange the fields. opcodes/ * aarch64-tbl.h (AARCH64_OPERANDS): Add an entry for AARCH64_OPND_SVE_PATTERN_SCALED. * aarch64-opc.h (FLD_SVE_imm4): New aarch64_field_kind. * aarch64-opc.c (fields): Add a corresponding entry. (set_multiplier_out_of_range_error): New function. (aarch64_operand_modifiers): Add entry for AARCH64_MOD_MUL. (operand_general_constraint_met_p): Handle AARCH64_OPND_SVE_PATTERN_SCALED. (print_register_offset_address): Use PRIi64 to print the shift amount. (aarch64_print_operand): Likewise. Handle AARCH64_OPND_SVE_PATTERN_SCALED. * aarch64-opc-2.c: Regenerate. * aarch64-asm.h (ins_sve_scale): New inserter. * aarch64-asm.c (aarch64_ins_sve_scale): New function. * aarch64-asm-2.c: Regenerate. * aarch64-dis.h (ext_sve_scale): New inserter. * aarch64-dis.c (aarch64_ext_sve_scale): New function. * aarch64-dis-2.c: Regenerate. gas/ * config/tc-aarch64.c (SHIFTED_MUL): New parse_shift_mode. (parse_shift): Handle it. Reject AARCH64_MOD_MUL for all other shift modes. Skip range tests for AARCH64_MOD_MUL. (process_omitted_operand): Handle AARCH64_OPND_SVE_PATTERN_SCALED. (parse_operands): Likewise.
2016-09-21 23:55:22 +08:00
if (opnd->shifter.operator_present)
{
size_t len = strlen (buf);
libopcodes/aarch64: add support for disassembler styling This commit enables disassembler styling for AArch64. After this commit it is possible to have objdump style AArch64 disassembler output (using --disassembler-color option). Once the required GDB patches are merged, GDB will also style the disassembler output. The changes to support styling are mostly split between two files opcodes/aarch64-dis.c and opcodes/aarch64-opc.c. The entry point for the AArch64 disassembler can be found in aarch64-dis.c, this file handles printing the instruction mnemonics, and assembler directives (e.g. '.byte', '.word', etc). Some operands, mostly relating to assembler directives are also printed from this file. This commit changes all of this to pass through suitable styling information. However, for most "normal" instructions, the instruction operands are printed using a two step process. From aarch64-dis.c, in the print_operands function, the function aarch64_print_operand is called, this function is in aarch64-opc.c, and converts an instruction operand into a string. Then, back in print_operands (aarch64-dis.c), the operand string is printed. Unfortunately, the string returned by aarch64_print_operand can be quite complex, it will include syntax elements, like '[' and ']', in addition to register names and immediate values. In some cases, a single operand will expand into what will appear (to the user) as multiple operands separated with a ','. This makes the task of styling more complex, all these different components need to by styled differently, so we need to get the styling information out of aarch64_print_operand in some way. The solution that I propose here is similar to the solution that I used for the i386 disassembler. Currently, aarch64_print_operand uses snprintf to write the operand text into a buffer provided by the caller. What I propose is that we pass an extra argument to the aarch64_print_operand function, this argument will be a structure, the structure contains a callback function and some state. When aarch64_print_operand needs to format part of its output this can be done by using the callback function within the new structure, this callback returns a string with special embedded markers that indicate which mode should be used for each piece of text. Back in aarch64-dis.c we can spot these special style markers and use this to split the disassembler output up and apply the correct style to each piece. To make aarch64-opc.c clearer a series of new static functions have been added, e.g. 'style_reg', 'style_imm', etc. Each of these functions formats a piece of text in a different style, 'register' and 'immediate' in this case. Here's an example taken from aarch64-opc.c of the new functions in use: snprintf (buf, size, "[%s, %s]!", style_reg (styler, base), style_imm (styler, "#%d", opnd->addr.offset.imm)); The aarch64_print_operand function is also called from the assembler to aid in printing diagnostic messages. Right now I have no plans to add styling to the assembler output, and so, the callback function used in the assembler ignores the styling information and just returns an plain string. I've used the source files in gas/testsuite/gas/aarch64/ for testing, and have manually gone through and checked that the styling looks reasonable, however, I'm not an AArch64 expert, so it is possible that the odd piece is styled incorrectly. Please point out any mistakes I've made. With objdump disassembler color turned off, there should be no change in the output after this commit.
2022-04-28 20:31:07 +08:00
const char *shift_name
= aarch64_operand_modifiers[opnd->shifter.kind].name;
snprintf (buf + len, size - len, ", %s %s",
style_sub_mnem (styler, shift_name),
style_imm (styler, "#%" PRIi64, opnd->shifter.amount));
[AArch64][SVE 24/32] Add AARCH64_OPND_SVE_PATTERN_SCALED Some SVE instructions count the number of elements in a given vector pattern and allow a scale factor of [1, 16] to be applied to the result. This scale factor is written ", MUL #n", where "MUL" is a new operator. E.g.: UQINCD X0, POW2, MUL #2 This patch adds support for this kind of operand. All existing operators were shifts of some kind, so there was a natural range of [0, 63] regardless of context. This was then narrowered further by later checks (e.g. to [0, 31] when used for 32-bit values). In contrast, MUL doesn't really have a natural context-independent range. Rather than pick one arbitrarily, it seemed better to make the "shift" amount a full 64-bit value and leave the range test to the usual operand-checking code. I've rearranged the fields of aarch64_opnd_info so that this doesn't increase the size of the structure (although I don't think its size is critical anyway). include/ * opcode/aarch64.h (AARCH64_OPND_SVE_PATTERN_SCALED): New aarch64_opnd. (AARCH64_MOD_MUL): New aarch64_modifier_kind. (aarch64_opnd_info): Make shifter.amount an int64_t and rearrange the fields. opcodes/ * aarch64-tbl.h (AARCH64_OPERANDS): Add an entry for AARCH64_OPND_SVE_PATTERN_SCALED. * aarch64-opc.h (FLD_SVE_imm4): New aarch64_field_kind. * aarch64-opc.c (fields): Add a corresponding entry. (set_multiplier_out_of_range_error): New function. (aarch64_operand_modifiers): Add entry for AARCH64_MOD_MUL. (operand_general_constraint_met_p): Handle AARCH64_OPND_SVE_PATTERN_SCALED. (print_register_offset_address): Use PRIi64 to print the shift amount. (aarch64_print_operand): Likewise. Handle AARCH64_OPND_SVE_PATTERN_SCALED. * aarch64-opc-2.c: Regenerate. * aarch64-asm.h (ins_sve_scale): New inserter. * aarch64-asm.c (aarch64_ins_sve_scale): New function. * aarch64-asm-2.c: Regenerate. * aarch64-dis.h (ext_sve_scale): New inserter. * aarch64-dis.c (aarch64_ext_sve_scale): New function. * aarch64-dis-2.c: Regenerate. gas/ * config/tc-aarch64.c (SHIFTED_MUL): New parse_shift_mode. (parse_shift): Handle it. Reject AARCH64_MOD_MUL for all other shift modes. Skip range tests for AARCH64_MOD_MUL. (process_omitted_operand): Handle AARCH64_OPND_SVE_PATTERN_SCALED. (parse_operands): Likewise.
2016-09-21 23:55:22 +08:00
}
break;
[AArch64][SVE 23/32] Add SVE pattern and prfop operands The SVE instructions have two enumerated operands: one to select a vector pattern and another to select a prefetch operation. The latter is a cut-down version of the base AArch64 prefetch operation. Both types of operand can also be specified as raw enum values such as #31. Reserved values can only be specified this way. If it hadn't been for the pattern operand, I would have been tempted to use the existing parsing for prefetch operations and add extra checks for SVE. However, since the patterns needed new enum parsing code anyway, it seeemed cleaner to reuse it for the prefetches too. Because of the small number of enum values, I don't think we'd gain anything by using hash tables. include/ * opcode/aarch64.h (AARCH64_OPND_SVE_PATTERN): New aarch64_opnd. (AARCH64_OPND_SVE_PRFOP): Likewise. (aarch64_sve_pattern_array): Declare. (aarch64_sve_prfop_array): Likewise. opcodes/ * aarch64-tbl.h (AARCH64_OPERANDS): Add entries for AARCH64_OPND_SVE_PATTERN and AARCH64_OPND_SVE_PRFOP. * aarch64-opc.h (FLD_SVE_pattern): New aarch64_field_kind. (FLD_SVE_prfop): Likewise. * aarch64-opc.c: Include libiberty.h. (aarch64_sve_pattern_array): New variable. (aarch64_sve_prfop_array): Likewise. (fields): Add entries for FLD_SVE_pattern and FLD_SVE_prfop. (aarch64_print_operand): Handle AARCH64_OPND_SVE_PATTERN and AARCH64_OPND_SVE_PRFOP. * aarch64-asm-2.c: Regenerate. * aarch64-dis-2.c: Likewise. * aarch64-opc-2.c: Likewise. gas/ * config/tc-aarch64.c (parse_enum_string): New function. (po_enum_or_fail): New macro. (parse_operands): Handle AARCH64_OPND_SVE_PATTERN and AARCH64_OPND_SVE_PRFOP.
2016-09-21 23:54:53 +08:00
case AARCH64_OPND_SVE_PRFOP:
enum_value = opnd->imm.value;
assert (enum_value < ARRAY_SIZE (aarch64_sve_prfop_array));
if (aarch64_sve_prfop_array[enum_value])
libopcodes/aarch64: add support for disassembler styling This commit enables disassembler styling for AArch64. After this commit it is possible to have objdump style AArch64 disassembler output (using --disassembler-color option). Once the required GDB patches are merged, GDB will also style the disassembler output. The changes to support styling are mostly split between two files opcodes/aarch64-dis.c and opcodes/aarch64-opc.c. The entry point for the AArch64 disassembler can be found in aarch64-dis.c, this file handles printing the instruction mnemonics, and assembler directives (e.g. '.byte', '.word', etc). Some operands, mostly relating to assembler directives are also printed from this file. This commit changes all of this to pass through suitable styling information. However, for most "normal" instructions, the instruction operands are printed using a two step process. From aarch64-dis.c, in the print_operands function, the function aarch64_print_operand is called, this function is in aarch64-opc.c, and converts an instruction operand into a string. Then, back in print_operands (aarch64-dis.c), the operand string is printed. Unfortunately, the string returned by aarch64_print_operand can be quite complex, it will include syntax elements, like '[' and ']', in addition to register names and immediate values. In some cases, a single operand will expand into what will appear (to the user) as multiple operands separated with a ','. This makes the task of styling more complex, all these different components need to by styled differently, so we need to get the styling information out of aarch64_print_operand in some way. The solution that I propose here is similar to the solution that I used for the i386 disassembler. Currently, aarch64_print_operand uses snprintf to write the operand text into a buffer provided by the caller. What I propose is that we pass an extra argument to the aarch64_print_operand function, this argument will be a structure, the structure contains a callback function and some state. When aarch64_print_operand needs to format part of its output this can be done by using the callback function within the new structure, this callback returns a string with special embedded markers that indicate which mode should be used for each piece of text. Back in aarch64-dis.c we can spot these special style markers and use this to split the disassembler output up and apply the correct style to each piece. To make aarch64-opc.c clearer a series of new static functions have been added, e.g. 'style_reg', 'style_imm', etc. Each of these functions formats a piece of text in a different style, 'register' and 'immediate' in this case. Here's an example taken from aarch64-opc.c of the new functions in use: snprintf (buf, size, "[%s, %s]!", style_reg (styler, base), style_imm (styler, "#%d", opnd->addr.offset.imm)); The aarch64_print_operand function is also called from the assembler to aid in printing diagnostic messages. Right now I have no plans to add styling to the assembler output, and so, the callback function used in the assembler ignores the styling information and just returns an plain string. I've used the source files in gas/testsuite/gas/aarch64/ for testing, and have manually gone through and checked that the styling looks reasonable, however, I'm not an AArch64 expert, so it is possible that the odd piece is styled incorrectly. Please point out any mistakes I've made. With objdump disassembler color turned off, there should be no change in the output after this commit.
2022-04-28 20:31:07 +08:00
snprintf (buf, size, "%s",
style_reg (styler, aarch64_sve_prfop_array[enum_value]));
[AArch64][SVE 23/32] Add SVE pattern and prfop operands The SVE instructions have two enumerated operands: one to select a vector pattern and another to select a prefetch operation. The latter is a cut-down version of the base AArch64 prefetch operation. Both types of operand can also be specified as raw enum values such as #31. Reserved values can only be specified this way. If it hadn't been for the pattern operand, I would have been tempted to use the existing parsing for prefetch operations and add extra checks for SVE. However, since the patterns needed new enum parsing code anyway, it seeemed cleaner to reuse it for the prefetches too. Because of the small number of enum values, I don't think we'd gain anything by using hash tables. include/ * opcode/aarch64.h (AARCH64_OPND_SVE_PATTERN): New aarch64_opnd. (AARCH64_OPND_SVE_PRFOP): Likewise. (aarch64_sve_pattern_array): Declare. (aarch64_sve_prfop_array): Likewise. opcodes/ * aarch64-tbl.h (AARCH64_OPERANDS): Add entries for AARCH64_OPND_SVE_PATTERN and AARCH64_OPND_SVE_PRFOP. * aarch64-opc.h (FLD_SVE_pattern): New aarch64_field_kind. (FLD_SVE_prfop): Likewise. * aarch64-opc.c: Include libiberty.h. (aarch64_sve_pattern_array): New variable. (aarch64_sve_prfop_array): Likewise. (fields): Add entries for FLD_SVE_pattern and FLD_SVE_prfop. (aarch64_print_operand): Handle AARCH64_OPND_SVE_PATTERN and AARCH64_OPND_SVE_PRFOP. * aarch64-asm-2.c: Regenerate. * aarch64-dis-2.c: Likewise. * aarch64-opc-2.c: Likewise. gas/ * config/tc-aarch64.c (parse_enum_string): New function. (po_enum_or_fail): New macro. (parse_operands): Handle AARCH64_OPND_SVE_PATTERN and AARCH64_OPND_SVE_PRFOP.
2016-09-21 23:54:53 +08:00
else
libopcodes/aarch64: add support for disassembler styling This commit enables disassembler styling for AArch64. After this commit it is possible to have objdump style AArch64 disassembler output (using --disassembler-color option). Once the required GDB patches are merged, GDB will also style the disassembler output. The changes to support styling are mostly split between two files opcodes/aarch64-dis.c and opcodes/aarch64-opc.c. The entry point for the AArch64 disassembler can be found in aarch64-dis.c, this file handles printing the instruction mnemonics, and assembler directives (e.g. '.byte', '.word', etc). Some operands, mostly relating to assembler directives are also printed from this file. This commit changes all of this to pass through suitable styling information. However, for most "normal" instructions, the instruction operands are printed using a two step process. From aarch64-dis.c, in the print_operands function, the function aarch64_print_operand is called, this function is in aarch64-opc.c, and converts an instruction operand into a string. Then, back in print_operands (aarch64-dis.c), the operand string is printed. Unfortunately, the string returned by aarch64_print_operand can be quite complex, it will include syntax elements, like '[' and ']', in addition to register names and immediate values. In some cases, a single operand will expand into what will appear (to the user) as multiple operands separated with a ','. This makes the task of styling more complex, all these different components need to by styled differently, so we need to get the styling information out of aarch64_print_operand in some way. The solution that I propose here is similar to the solution that I used for the i386 disassembler. Currently, aarch64_print_operand uses snprintf to write the operand text into a buffer provided by the caller. What I propose is that we pass an extra argument to the aarch64_print_operand function, this argument will be a structure, the structure contains a callback function and some state. When aarch64_print_operand needs to format part of its output this can be done by using the callback function within the new structure, this callback returns a string with special embedded markers that indicate which mode should be used for each piece of text. Back in aarch64-dis.c we can spot these special style markers and use this to split the disassembler output up and apply the correct style to each piece. To make aarch64-opc.c clearer a series of new static functions have been added, e.g. 'style_reg', 'style_imm', etc. Each of these functions formats a piece of text in a different style, 'register' and 'immediate' in this case. Here's an example taken from aarch64-opc.c of the new functions in use: snprintf (buf, size, "[%s, %s]!", style_reg (styler, base), style_imm (styler, "#%d", opnd->addr.offset.imm)); The aarch64_print_operand function is also called from the assembler to aid in printing diagnostic messages. Right now I have no plans to add styling to the assembler output, and so, the callback function used in the assembler ignores the styling information and just returns an plain string. I've used the source files in gas/testsuite/gas/aarch64/ for testing, and have manually gone through and checked that the styling looks reasonable, however, I'm not an AArch64 expert, so it is possible that the odd piece is styled incorrectly. Please point out any mistakes I've made. With objdump disassembler color turned off, there should be no change in the output after this commit.
2022-04-28 20:31:07 +08:00
snprintf (buf, size, "%s",
style_imm (styler, "#%" PRIi64, opnd->imm.value));
[AArch64][SVE 23/32] Add SVE pattern and prfop operands The SVE instructions have two enumerated operands: one to select a vector pattern and another to select a prefetch operation. The latter is a cut-down version of the base AArch64 prefetch operation. Both types of operand can also be specified as raw enum values such as #31. Reserved values can only be specified this way. If it hadn't been for the pattern operand, I would have been tempted to use the existing parsing for prefetch operations and add extra checks for SVE. However, since the patterns needed new enum parsing code anyway, it seeemed cleaner to reuse it for the prefetches too. Because of the small number of enum values, I don't think we'd gain anything by using hash tables. include/ * opcode/aarch64.h (AARCH64_OPND_SVE_PATTERN): New aarch64_opnd. (AARCH64_OPND_SVE_PRFOP): Likewise. (aarch64_sve_pattern_array): Declare. (aarch64_sve_prfop_array): Likewise. opcodes/ * aarch64-tbl.h (AARCH64_OPERANDS): Add entries for AARCH64_OPND_SVE_PATTERN and AARCH64_OPND_SVE_PRFOP. * aarch64-opc.h (FLD_SVE_pattern): New aarch64_field_kind. (FLD_SVE_prfop): Likewise. * aarch64-opc.c: Include libiberty.h. (aarch64_sve_pattern_array): New variable. (aarch64_sve_prfop_array): Likewise. (fields): Add entries for FLD_SVE_pattern and FLD_SVE_prfop. (aarch64_print_operand): Handle AARCH64_OPND_SVE_PATTERN and AARCH64_OPND_SVE_PRFOP. * aarch64-asm-2.c: Regenerate. * aarch64-dis-2.c: Likewise. * aarch64-opc-2.c: Likewise. gas/ * config/tc-aarch64.c (parse_enum_string): New function. (po_enum_or_fail): New macro. (parse_operands): Handle AARCH64_OPND_SVE_PATTERN and AARCH64_OPND_SVE_PRFOP.
2016-09-21 23:54:53 +08:00
break;
case AARCH64_OPND_IMM_MOV:
switch (aarch64_get_qualifier_esize (opnds[0].qualifier))
{
case 4: /* e.g. MOV Wd, #<imm32>. */
{
int imm32 = opnd->imm.value;
libopcodes/aarch64: add support for disassembler styling This commit enables disassembler styling for AArch64. After this commit it is possible to have objdump style AArch64 disassembler output (using --disassembler-color option). Once the required GDB patches are merged, GDB will also style the disassembler output. The changes to support styling are mostly split between two files opcodes/aarch64-dis.c and opcodes/aarch64-opc.c. The entry point for the AArch64 disassembler can be found in aarch64-dis.c, this file handles printing the instruction mnemonics, and assembler directives (e.g. '.byte', '.word', etc). Some operands, mostly relating to assembler directives are also printed from this file. This commit changes all of this to pass through suitable styling information. However, for most "normal" instructions, the instruction operands are printed using a two step process. From aarch64-dis.c, in the print_operands function, the function aarch64_print_operand is called, this function is in aarch64-opc.c, and converts an instruction operand into a string. Then, back in print_operands (aarch64-dis.c), the operand string is printed. Unfortunately, the string returned by aarch64_print_operand can be quite complex, it will include syntax elements, like '[' and ']', in addition to register names and immediate values. In some cases, a single operand will expand into what will appear (to the user) as multiple operands separated with a ','. This makes the task of styling more complex, all these different components need to by styled differently, so we need to get the styling information out of aarch64_print_operand in some way. The solution that I propose here is similar to the solution that I used for the i386 disassembler. Currently, aarch64_print_operand uses snprintf to write the operand text into a buffer provided by the caller. What I propose is that we pass an extra argument to the aarch64_print_operand function, this argument will be a structure, the structure contains a callback function and some state. When aarch64_print_operand needs to format part of its output this can be done by using the callback function within the new structure, this callback returns a string with special embedded markers that indicate which mode should be used for each piece of text. Back in aarch64-dis.c we can spot these special style markers and use this to split the disassembler output up and apply the correct style to each piece. To make aarch64-opc.c clearer a series of new static functions have been added, e.g. 'style_reg', 'style_imm', etc. Each of these functions formats a piece of text in a different style, 'register' and 'immediate' in this case. Here's an example taken from aarch64-opc.c of the new functions in use: snprintf (buf, size, "[%s, %s]!", style_reg (styler, base), style_imm (styler, "#%d", opnd->addr.offset.imm)); The aarch64_print_operand function is also called from the assembler to aid in printing diagnostic messages. Right now I have no plans to add styling to the assembler output, and so, the callback function used in the assembler ignores the styling information and just returns an plain string. I've used the source files in gas/testsuite/gas/aarch64/ for testing, and have manually gone through and checked that the styling looks reasonable, however, I'm not an AArch64 expert, so it is possible that the odd piece is styled incorrectly. Please point out any mistakes I've made. With objdump disassembler color turned off, there should be no change in the output after this commit.
2022-04-28 20:31:07 +08:00
snprintf (buf, size, "%s",
style_imm (styler, "#0x%-20x", imm32));
opcodes/aarch64: split off creation of comment text in disassembler The function aarch64_print_operand (aarch64-opc.c) is responsible for converting an instruction operand into the textual representation of that operand. In some cases, a comment is included in the operand representation, though this (currently) only happens for the last operand of the instruction. In a future commit I would like to enable the new libopcodes styling for AArch64, this will allow objdump and GDB[1] to syntax highlight the disassembler output, however, having operands and comments combined in a single string like this makes such styling harder. In this commit, I propose to extend aarch64_print_operand to take a second buffer. Any comments for the instruction are written into this extra buffer. The two callers of aarch64_print_operand are then updated to pass an extra buffer, and print any resulting comment. In this commit no styling is added, that will come later. However, I have adjusted the output slightly. Before this commit some comments would be separated from the instruction operands with a tab character, while in other cases the comment was separated with two single spaces. After this commit I use a single tab character in all cases. This means a few test cases needed updated. If people would prefer me to move everyone to use the two spaces, then just let me know. Or maybe there was a good reason why we used a mix of styles, I could probably figure out a way to maintain the old output exactly if that is critical. Other than that, there should be no user visible changes after this commit. [1] GDB patches have not been merged yet, but have been posted to the GDB mailing list: https://sourceware.org/pipermail/gdb-patches/2022-June/190142.html
2022-06-16 20:46:41 +08:00
snprintf (comment, comment_size, "#%d", imm32);
}
break;
case 8: /* e.g. MOV Xd, #<imm64>. */
libopcodes/aarch64: add support for disassembler styling This commit enables disassembler styling for AArch64. After this commit it is possible to have objdump style AArch64 disassembler output (using --disassembler-color option). Once the required GDB patches are merged, GDB will also style the disassembler output. The changes to support styling are mostly split between two files opcodes/aarch64-dis.c and opcodes/aarch64-opc.c. The entry point for the AArch64 disassembler can be found in aarch64-dis.c, this file handles printing the instruction mnemonics, and assembler directives (e.g. '.byte', '.word', etc). Some operands, mostly relating to assembler directives are also printed from this file. This commit changes all of this to pass through suitable styling information. However, for most "normal" instructions, the instruction operands are printed using a two step process. From aarch64-dis.c, in the print_operands function, the function aarch64_print_operand is called, this function is in aarch64-opc.c, and converts an instruction operand into a string. Then, back in print_operands (aarch64-dis.c), the operand string is printed. Unfortunately, the string returned by aarch64_print_operand can be quite complex, it will include syntax elements, like '[' and ']', in addition to register names and immediate values. In some cases, a single operand will expand into what will appear (to the user) as multiple operands separated with a ','. This makes the task of styling more complex, all these different components need to by styled differently, so we need to get the styling information out of aarch64_print_operand in some way. The solution that I propose here is similar to the solution that I used for the i386 disassembler. Currently, aarch64_print_operand uses snprintf to write the operand text into a buffer provided by the caller. What I propose is that we pass an extra argument to the aarch64_print_operand function, this argument will be a structure, the structure contains a callback function and some state. When aarch64_print_operand needs to format part of its output this can be done by using the callback function within the new structure, this callback returns a string with special embedded markers that indicate which mode should be used for each piece of text. Back in aarch64-dis.c we can spot these special style markers and use this to split the disassembler output up and apply the correct style to each piece. To make aarch64-opc.c clearer a series of new static functions have been added, e.g. 'style_reg', 'style_imm', etc. Each of these functions formats a piece of text in a different style, 'register' and 'immediate' in this case. Here's an example taken from aarch64-opc.c of the new functions in use: snprintf (buf, size, "[%s, %s]!", style_reg (styler, base), style_imm (styler, "#%d", opnd->addr.offset.imm)); The aarch64_print_operand function is also called from the assembler to aid in printing diagnostic messages. Right now I have no plans to add styling to the assembler output, and so, the callback function used in the assembler ignores the styling information and just returns an plain string. I've used the source files in gas/testsuite/gas/aarch64/ for testing, and have manually gone through and checked that the styling looks reasonable, however, I'm not an AArch64 expert, so it is possible that the odd piece is styled incorrectly. Please point out any mistakes I've made. With objdump disassembler color turned off, there should be no change in the output after this commit.
2022-04-28 20:31:07 +08:00
snprintf (buf, size, "%s", style_imm (styler, "#0x%-20" PRIx64,
opnd->imm.value));
opcodes/aarch64: split off creation of comment text in disassembler The function aarch64_print_operand (aarch64-opc.c) is responsible for converting an instruction operand into the textual representation of that operand. In some cases, a comment is included in the operand representation, though this (currently) only happens for the last operand of the instruction. In a future commit I would like to enable the new libopcodes styling for AArch64, this will allow objdump and GDB[1] to syntax highlight the disassembler output, however, having operands and comments combined in a single string like this makes such styling harder. In this commit, I propose to extend aarch64_print_operand to take a second buffer. Any comments for the instruction are written into this extra buffer. The two callers of aarch64_print_operand are then updated to pass an extra buffer, and print any resulting comment. In this commit no styling is added, that will come later. However, I have adjusted the output slightly. Before this commit some comments would be separated from the instruction operands with a tab character, while in other cases the comment was separated with two single spaces. After this commit I use a single tab character in all cases. This means a few test cases needed updated. If people would prefer me to move everyone to use the two spaces, then just let me know. Or maybe there was a good reason why we used a mix of styles, I could probably figure out a way to maintain the old output exactly if that is critical. Other than that, there should be no user visible changes after this commit. [1] GDB patches have not been merged yet, but have been posted to the GDB mailing list: https://sourceware.org/pipermail/gdb-patches/2022-June/190142.html
2022-06-16 20:46:41 +08:00
snprintf (comment, comment_size, "#%" PRIi64, opnd->imm.value);
break;
default:
snprintf (buf, size, "<invalid>");
break;
}
break;
case AARCH64_OPND_FPIMM0:
libopcodes/aarch64: add support for disassembler styling This commit enables disassembler styling for AArch64. After this commit it is possible to have objdump style AArch64 disassembler output (using --disassembler-color option). Once the required GDB patches are merged, GDB will also style the disassembler output. The changes to support styling are mostly split between two files opcodes/aarch64-dis.c and opcodes/aarch64-opc.c. The entry point for the AArch64 disassembler can be found in aarch64-dis.c, this file handles printing the instruction mnemonics, and assembler directives (e.g. '.byte', '.word', etc). Some operands, mostly relating to assembler directives are also printed from this file. This commit changes all of this to pass through suitable styling information. However, for most "normal" instructions, the instruction operands are printed using a two step process. From aarch64-dis.c, in the print_operands function, the function aarch64_print_operand is called, this function is in aarch64-opc.c, and converts an instruction operand into a string. Then, back in print_operands (aarch64-dis.c), the operand string is printed. Unfortunately, the string returned by aarch64_print_operand can be quite complex, it will include syntax elements, like '[' and ']', in addition to register names and immediate values. In some cases, a single operand will expand into what will appear (to the user) as multiple operands separated with a ','. This makes the task of styling more complex, all these different components need to by styled differently, so we need to get the styling information out of aarch64_print_operand in some way. The solution that I propose here is similar to the solution that I used for the i386 disassembler. Currently, aarch64_print_operand uses snprintf to write the operand text into a buffer provided by the caller. What I propose is that we pass an extra argument to the aarch64_print_operand function, this argument will be a structure, the structure contains a callback function and some state. When aarch64_print_operand needs to format part of its output this can be done by using the callback function within the new structure, this callback returns a string with special embedded markers that indicate which mode should be used for each piece of text. Back in aarch64-dis.c we can spot these special style markers and use this to split the disassembler output up and apply the correct style to each piece. To make aarch64-opc.c clearer a series of new static functions have been added, e.g. 'style_reg', 'style_imm', etc. Each of these functions formats a piece of text in a different style, 'register' and 'immediate' in this case. Here's an example taken from aarch64-opc.c of the new functions in use: snprintf (buf, size, "[%s, %s]!", style_reg (styler, base), style_imm (styler, "#%d", opnd->addr.offset.imm)); The aarch64_print_operand function is also called from the assembler to aid in printing diagnostic messages. Right now I have no plans to add styling to the assembler output, and so, the callback function used in the assembler ignores the styling information and just returns an plain string. I've used the source files in gas/testsuite/gas/aarch64/ for testing, and have manually gone through and checked that the styling looks reasonable, however, I'm not an AArch64 expert, so it is possible that the odd piece is styled incorrectly. Please point out any mistakes I've made. With objdump disassembler color turned off, there should be no change in the output after this commit.
2022-04-28 20:31:07 +08:00
snprintf (buf, size, "%s", style_imm (styler, "#0.0"));
break;
case AARCH64_OPND_LIMM:
case AARCH64_OPND_AIMM:
case AARCH64_OPND_HALF:
[AArch64][SVE 27/32] Add SVE integer immediate operands This patch adds the new SVE integer immediate operands. There are three kinds: - simple signed and unsigned ranges, but with new widths and positions. - 13-bit logical immediates. These have the same form as in base AArch64, but at a different bit position. In the case of the "MOV Zn.<T>, #<limm>" alias of DUPM, the logical immediate <limm> is not allowed to be a valid DUP immediate, since DUP is preferred over DUPM for constants that both instructions can handle. - a new 9-bit arithmetic immediate, of the form "<imm8>{, LSL #8}". In some contexts the operand is signed and in others it's unsigned. As an extension, we allow shifted immediates to be written as a single integer, e.g. "#256" is equivalent to "#1, LSL #8". We also use the shiftless form as the preferred disassembly, except for the special case of "#0, LSL #8" (a redundant encoding of 0). include/ * opcode/aarch64.h (AARCH64_OPND_SIMM5): New aarch64_opnd. (AARCH64_OPND_SVE_AIMM, AARCH64_OPND_SVE_ASIMM) (AARCH64_OPND_SVE_INV_LIMM, AARCH64_OPND_SVE_LIMM) (AARCH64_OPND_SVE_LIMM_MOV, AARCH64_OPND_SVE_SHLIMM_PRED) (AARCH64_OPND_SVE_SHLIMM_UNPRED, AARCH64_OPND_SVE_SHRIMM_PRED) (AARCH64_OPND_SVE_SHRIMM_UNPRED, AARCH64_OPND_SVE_SIMM5) (AARCH64_OPND_SVE_SIMM5B, AARCH64_OPND_SVE_SIMM6) (AARCH64_OPND_SVE_SIMM8, AARCH64_OPND_SVE_UIMM3) (AARCH64_OPND_SVE_UIMM7, AARCH64_OPND_SVE_UIMM8) (AARCH64_OPND_SVE_UIMM8_53): Likewise. (aarch64_sve_dupm_mov_immediate_p): Declare. opcodes/ * aarch64-tbl.h (AARCH64_OPERANDS): Add entries for the new SVE integer immediate operands. * aarch64-opc.h (FLD_SVE_immN, FLD_SVE_imm3, FLD_SVE_imm5) (FLD_SVE_imm5b, FLD_SVE_imm7, FLD_SVE_imm8, FLD_SVE_imm9) (FLD_SVE_immr, FLD_SVE_imms, FLD_SVE_tszh): New aarch64_field_kinds. * aarch64-opc.c (fields): Add corresponding entries. (operand_general_constraint_met_p): Handle the new SVE integer immediate operands. (aarch64_print_operand): Likewise. (aarch64_sve_dupm_mov_immediate_p): New function. * aarch64-opc-2.c: Regenerate. * aarch64-asm.h (ins_inv_limm, ins_sve_aimm, ins_sve_asimm) (ins_sve_limm_mov, ins_sve_shlimm, ins_sve_shrimm): New inserters. * aarch64-asm.c (aarch64_ins_limm_1): New function, split out from... (aarch64_ins_limm): ...here. (aarch64_ins_inv_limm): New function. (aarch64_ins_sve_aimm): Likewise. (aarch64_ins_sve_asimm): Likewise. (aarch64_ins_sve_limm_mov): Likewise. (aarch64_ins_sve_shlimm): Likewise. (aarch64_ins_sve_shrimm): Likewise. * aarch64-asm-2.c: Regenerate. * aarch64-dis.h (ext_inv_limm, ext_sve_aimm, ext_sve_asimm) (ext_sve_limm_mov, ext_sve_shlimm, ext_sve_shrimm): New extractors. * aarch64-dis.c (decode_limm): New function, split out from... (aarch64_ext_limm): ...here. (aarch64_ext_inv_limm): New function. (decode_sve_aimm): Likewise. (aarch64_ext_sve_aimm): Likewise. (aarch64_ext_sve_asimm): Likewise. (aarch64_ext_sve_limm_mov): Likewise. (aarch64_top_bit): Likewise. (aarch64_ext_sve_shlimm): Likewise. (aarch64_ext_sve_shrimm): Likewise. * aarch64-dis-2.c: Regenerate. gas/ * config/tc-aarch64.c (parse_operands): Handle the new SVE integer immediate operands.
2016-09-21 23:56:57 +08:00
case AARCH64_OPND_SVE_INV_LIMM:
case AARCH64_OPND_SVE_LIMM:
case AARCH64_OPND_SVE_LIMM_MOV:
if (opnd->shifter.amount)
libopcodes/aarch64: add support for disassembler styling This commit enables disassembler styling for AArch64. After this commit it is possible to have objdump style AArch64 disassembler output (using --disassembler-color option). Once the required GDB patches are merged, GDB will also style the disassembler output. The changes to support styling are mostly split between two files opcodes/aarch64-dis.c and opcodes/aarch64-opc.c. The entry point for the AArch64 disassembler can be found in aarch64-dis.c, this file handles printing the instruction mnemonics, and assembler directives (e.g. '.byte', '.word', etc). Some operands, mostly relating to assembler directives are also printed from this file. This commit changes all of this to pass through suitable styling information. However, for most "normal" instructions, the instruction operands are printed using a two step process. From aarch64-dis.c, in the print_operands function, the function aarch64_print_operand is called, this function is in aarch64-opc.c, and converts an instruction operand into a string. Then, back in print_operands (aarch64-dis.c), the operand string is printed. Unfortunately, the string returned by aarch64_print_operand can be quite complex, it will include syntax elements, like '[' and ']', in addition to register names and immediate values. In some cases, a single operand will expand into what will appear (to the user) as multiple operands separated with a ','. This makes the task of styling more complex, all these different components need to by styled differently, so we need to get the styling information out of aarch64_print_operand in some way. The solution that I propose here is similar to the solution that I used for the i386 disassembler. Currently, aarch64_print_operand uses snprintf to write the operand text into a buffer provided by the caller. What I propose is that we pass an extra argument to the aarch64_print_operand function, this argument will be a structure, the structure contains a callback function and some state. When aarch64_print_operand needs to format part of its output this can be done by using the callback function within the new structure, this callback returns a string with special embedded markers that indicate which mode should be used for each piece of text. Back in aarch64-dis.c we can spot these special style markers and use this to split the disassembler output up and apply the correct style to each piece. To make aarch64-opc.c clearer a series of new static functions have been added, e.g. 'style_reg', 'style_imm', etc. Each of these functions formats a piece of text in a different style, 'register' and 'immediate' in this case. Here's an example taken from aarch64-opc.c of the new functions in use: snprintf (buf, size, "[%s, %s]!", style_reg (styler, base), style_imm (styler, "#%d", opnd->addr.offset.imm)); The aarch64_print_operand function is also called from the assembler to aid in printing diagnostic messages. Right now I have no plans to add styling to the assembler output, and so, the callback function used in the assembler ignores the styling information and just returns an plain string. I've used the source files in gas/testsuite/gas/aarch64/ for testing, and have manually gone through and checked that the styling looks reasonable, however, I'm not an AArch64 expert, so it is possible that the odd piece is styled incorrectly. Please point out any mistakes I've made. With objdump disassembler color turned off, there should be no change in the output after this commit.
2022-04-28 20:31:07 +08:00
snprintf (buf, size, "%s, %s %s",
style_imm (styler, "#0x%" PRIx64, opnd->imm.value),
style_sub_mnem (styler, "lsl"),
style_imm (styler, "#%" PRIi64, opnd->shifter.amount));
else
libopcodes/aarch64: add support for disassembler styling This commit enables disassembler styling for AArch64. After this commit it is possible to have objdump style AArch64 disassembler output (using --disassembler-color option). Once the required GDB patches are merged, GDB will also style the disassembler output. The changes to support styling are mostly split between two files opcodes/aarch64-dis.c and opcodes/aarch64-opc.c. The entry point for the AArch64 disassembler can be found in aarch64-dis.c, this file handles printing the instruction mnemonics, and assembler directives (e.g. '.byte', '.word', etc). Some operands, mostly relating to assembler directives are also printed from this file. This commit changes all of this to pass through suitable styling information. However, for most "normal" instructions, the instruction operands are printed using a two step process. From aarch64-dis.c, in the print_operands function, the function aarch64_print_operand is called, this function is in aarch64-opc.c, and converts an instruction operand into a string. Then, back in print_operands (aarch64-dis.c), the operand string is printed. Unfortunately, the string returned by aarch64_print_operand can be quite complex, it will include syntax elements, like '[' and ']', in addition to register names and immediate values. In some cases, a single operand will expand into what will appear (to the user) as multiple operands separated with a ','. This makes the task of styling more complex, all these different components need to by styled differently, so we need to get the styling information out of aarch64_print_operand in some way. The solution that I propose here is similar to the solution that I used for the i386 disassembler. Currently, aarch64_print_operand uses snprintf to write the operand text into a buffer provided by the caller. What I propose is that we pass an extra argument to the aarch64_print_operand function, this argument will be a structure, the structure contains a callback function and some state. When aarch64_print_operand needs to format part of its output this can be done by using the callback function within the new structure, this callback returns a string with special embedded markers that indicate which mode should be used for each piece of text. Back in aarch64-dis.c we can spot these special style markers and use this to split the disassembler output up and apply the correct style to each piece. To make aarch64-opc.c clearer a series of new static functions have been added, e.g. 'style_reg', 'style_imm', etc. Each of these functions formats a piece of text in a different style, 'register' and 'immediate' in this case. Here's an example taken from aarch64-opc.c of the new functions in use: snprintf (buf, size, "[%s, %s]!", style_reg (styler, base), style_imm (styler, "#%d", opnd->addr.offset.imm)); The aarch64_print_operand function is also called from the assembler to aid in printing diagnostic messages. Right now I have no plans to add styling to the assembler output, and so, the callback function used in the assembler ignores the styling information and just returns an plain string. I've used the source files in gas/testsuite/gas/aarch64/ for testing, and have manually gone through and checked that the styling looks reasonable, however, I'm not an AArch64 expert, so it is possible that the odd piece is styled incorrectly. Please point out any mistakes I've made. With objdump disassembler color turned off, there should be no change in the output after this commit.
2022-04-28 20:31:07 +08:00
snprintf (buf, size, "%s",
style_imm (styler, "#0x%" PRIx64, opnd->imm.value));
break;
case AARCH64_OPND_SIMD_IMM:
case AARCH64_OPND_SIMD_IMM_SFT:
if ((! opnd->shifter.amount && opnd->shifter.kind == AARCH64_MOD_LSL)
|| opnd->shifter.kind == AARCH64_MOD_NONE)
libopcodes/aarch64: add support for disassembler styling This commit enables disassembler styling for AArch64. After this commit it is possible to have objdump style AArch64 disassembler output (using --disassembler-color option). Once the required GDB patches are merged, GDB will also style the disassembler output. The changes to support styling are mostly split between two files opcodes/aarch64-dis.c and opcodes/aarch64-opc.c. The entry point for the AArch64 disassembler can be found in aarch64-dis.c, this file handles printing the instruction mnemonics, and assembler directives (e.g. '.byte', '.word', etc). Some operands, mostly relating to assembler directives are also printed from this file. This commit changes all of this to pass through suitable styling information. However, for most "normal" instructions, the instruction operands are printed using a two step process. From aarch64-dis.c, in the print_operands function, the function aarch64_print_operand is called, this function is in aarch64-opc.c, and converts an instruction operand into a string. Then, back in print_operands (aarch64-dis.c), the operand string is printed. Unfortunately, the string returned by aarch64_print_operand can be quite complex, it will include syntax elements, like '[' and ']', in addition to register names and immediate values. In some cases, a single operand will expand into what will appear (to the user) as multiple operands separated with a ','. This makes the task of styling more complex, all these different components need to by styled differently, so we need to get the styling information out of aarch64_print_operand in some way. The solution that I propose here is similar to the solution that I used for the i386 disassembler. Currently, aarch64_print_operand uses snprintf to write the operand text into a buffer provided by the caller. What I propose is that we pass an extra argument to the aarch64_print_operand function, this argument will be a structure, the structure contains a callback function and some state. When aarch64_print_operand needs to format part of its output this can be done by using the callback function within the new structure, this callback returns a string with special embedded markers that indicate which mode should be used for each piece of text. Back in aarch64-dis.c we can spot these special style markers and use this to split the disassembler output up and apply the correct style to each piece. To make aarch64-opc.c clearer a series of new static functions have been added, e.g. 'style_reg', 'style_imm', etc. Each of these functions formats a piece of text in a different style, 'register' and 'immediate' in this case. Here's an example taken from aarch64-opc.c of the new functions in use: snprintf (buf, size, "[%s, %s]!", style_reg (styler, base), style_imm (styler, "#%d", opnd->addr.offset.imm)); The aarch64_print_operand function is also called from the assembler to aid in printing diagnostic messages. Right now I have no plans to add styling to the assembler output, and so, the callback function used in the assembler ignores the styling information and just returns an plain string. I've used the source files in gas/testsuite/gas/aarch64/ for testing, and have manually gone through and checked that the styling looks reasonable, however, I'm not an AArch64 expert, so it is possible that the odd piece is styled incorrectly. Please point out any mistakes I've made. With objdump disassembler color turned off, there should be no change in the output after this commit.
2022-04-28 20:31:07 +08:00
snprintf (buf, size, "%s",
style_imm (styler, "#0x%" PRIx64, opnd->imm.value));
else
libopcodes/aarch64: add support for disassembler styling This commit enables disassembler styling for AArch64. After this commit it is possible to have objdump style AArch64 disassembler output (using --disassembler-color option). Once the required GDB patches are merged, GDB will also style the disassembler output. The changes to support styling are mostly split between two files opcodes/aarch64-dis.c and opcodes/aarch64-opc.c. The entry point for the AArch64 disassembler can be found in aarch64-dis.c, this file handles printing the instruction mnemonics, and assembler directives (e.g. '.byte', '.word', etc). Some operands, mostly relating to assembler directives are also printed from this file. This commit changes all of this to pass through suitable styling information. However, for most "normal" instructions, the instruction operands are printed using a two step process. From aarch64-dis.c, in the print_operands function, the function aarch64_print_operand is called, this function is in aarch64-opc.c, and converts an instruction operand into a string. Then, back in print_operands (aarch64-dis.c), the operand string is printed. Unfortunately, the string returned by aarch64_print_operand can be quite complex, it will include syntax elements, like '[' and ']', in addition to register names and immediate values. In some cases, a single operand will expand into what will appear (to the user) as multiple operands separated with a ','. This makes the task of styling more complex, all these different components need to by styled differently, so we need to get the styling information out of aarch64_print_operand in some way. The solution that I propose here is similar to the solution that I used for the i386 disassembler. Currently, aarch64_print_operand uses snprintf to write the operand text into a buffer provided by the caller. What I propose is that we pass an extra argument to the aarch64_print_operand function, this argument will be a structure, the structure contains a callback function and some state. When aarch64_print_operand needs to format part of its output this can be done by using the callback function within the new structure, this callback returns a string with special embedded markers that indicate which mode should be used for each piece of text. Back in aarch64-dis.c we can spot these special style markers and use this to split the disassembler output up and apply the correct style to each piece. To make aarch64-opc.c clearer a series of new static functions have been added, e.g. 'style_reg', 'style_imm', etc. Each of these functions formats a piece of text in a different style, 'register' and 'immediate' in this case. Here's an example taken from aarch64-opc.c of the new functions in use: snprintf (buf, size, "[%s, %s]!", style_reg (styler, base), style_imm (styler, "#%d", opnd->addr.offset.imm)); The aarch64_print_operand function is also called from the assembler to aid in printing diagnostic messages. Right now I have no plans to add styling to the assembler output, and so, the callback function used in the assembler ignores the styling information and just returns an plain string. I've used the source files in gas/testsuite/gas/aarch64/ for testing, and have manually gone through and checked that the styling looks reasonable, however, I'm not an AArch64 expert, so it is possible that the odd piece is styled incorrectly. Please point out any mistakes I've made. With objdump disassembler color turned off, there should be no change in the output after this commit.
2022-04-28 20:31:07 +08:00
snprintf (buf, size, "%s, %s %s",
style_imm (styler, "#0x%" PRIx64, opnd->imm.value),
style_sub_mnem (styler, aarch64_operand_modifiers[opnd->shifter.kind].name),
style_imm (styler, "#%" PRIi64, opnd->shifter.amount));
break;
[AArch64][SVE 27/32] Add SVE integer immediate operands This patch adds the new SVE integer immediate operands. There are three kinds: - simple signed and unsigned ranges, but with new widths and positions. - 13-bit logical immediates. These have the same form as in base AArch64, but at a different bit position. In the case of the "MOV Zn.<T>, #<limm>" alias of DUPM, the logical immediate <limm> is not allowed to be a valid DUP immediate, since DUP is preferred over DUPM for constants that both instructions can handle. - a new 9-bit arithmetic immediate, of the form "<imm8>{, LSL #8}". In some contexts the operand is signed and in others it's unsigned. As an extension, we allow shifted immediates to be written as a single integer, e.g. "#256" is equivalent to "#1, LSL #8". We also use the shiftless form as the preferred disassembly, except for the special case of "#0, LSL #8" (a redundant encoding of 0). include/ * opcode/aarch64.h (AARCH64_OPND_SIMM5): New aarch64_opnd. (AARCH64_OPND_SVE_AIMM, AARCH64_OPND_SVE_ASIMM) (AARCH64_OPND_SVE_INV_LIMM, AARCH64_OPND_SVE_LIMM) (AARCH64_OPND_SVE_LIMM_MOV, AARCH64_OPND_SVE_SHLIMM_PRED) (AARCH64_OPND_SVE_SHLIMM_UNPRED, AARCH64_OPND_SVE_SHRIMM_PRED) (AARCH64_OPND_SVE_SHRIMM_UNPRED, AARCH64_OPND_SVE_SIMM5) (AARCH64_OPND_SVE_SIMM5B, AARCH64_OPND_SVE_SIMM6) (AARCH64_OPND_SVE_SIMM8, AARCH64_OPND_SVE_UIMM3) (AARCH64_OPND_SVE_UIMM7, AARCH64_OPND_SVE_UIMM8) (AARCH64_OPND_SVE_UIMM8_53): Likewise. (aarch64_sve_dupm_mov_immediate_p): Declare. opcodes/ * aarch64-tbl.h (AARCH64_OPERANDS): Add entries for the new SVE integer immediate operands. * aarch64-opc.h (FLD_SVE_immN, FLD_SVE_imm3, FLD_SVE_imm5) (FLD_SVE_imm5b, FLD_SVE_imm7, FLD_SVE_imm8, FLD_SVE_imm9) (FLD_SVE_immr, FLD_SVE_imms, FLD_SVE_tszh): New aarch64_field_kinds. * aarch64-opc.c (fields): Add corresponding entries. (operand_general_constraint_met_p): Handle the new SVE integer immediate operands. (aarch64_print_operand): Likewise. (aarch64_sve_dupm_mov_immediate_p): New function. * aarch64-opc-2.c: Regenerate. * aarch64-asm.h (ins_inv_limm, ins_sve_aimm, ins_sve_asimm) (ins_sve_limm_mov, ins_sve_shlimm, ins_sve_shrimm): New inserters. * aarch64-asm.c (aarch64_ins_limm_1): New function, split out from... (aarch64_ins_limm): ...here. (aarch64_ins_inv_limm): New function. (aarch64_ins_sve_aimm): Likewise. (aarch64_ins_sve_asimm): Likewise. (aarch64_ins_sve_limm_mov): Likewise. (aarch64_ins_sve_shlimm): Likewise. (aarch64_ins_sve_shrimm): Likewise. * aarch64-asm-2.c: Regenerate. * aarch64-dis.h (ext_inv_limm, ext_sve_aimm, ext_sve_asimm) (ext_sve_limm_mov, ext_sve_shlimm, ext_sve_shrimm): New extractors. * aarch64-dis.c (decode_limm): New function, split out from... (aarch64_ext_limm): ...here. (aarch64_ext_inv_limm): New function. (decode_sve_aimm): Likewise. (aarch64_ext_sve_aimm): Likewise. (aarch64_ext_sve_asimm): Likewise. (aarch64_ext_sve_limm_mov): Likewise. (aarch64_top_bit): Likewise. (aarch64_ext_sve_shlimm): Likewise. (aarch64_ext_sve_shrimm): Likewise. * aarch64-dis-2.c: Regenerate. gas/ * config/tc-aarch64.c (parse_operands): Handle the new SVE integer immediate operands.
2016-09-21 23:56:57 +08:00
case AARCH64_OPND_SVE_AIMM:
case AARCH64_OPND_SVE_ASIMM:
if (opnd->shifter.amount)
libopcodes/aarch64: add support for disassembler styling This commit enables disassembler styling for AArch64. After this commit it is possible to have objdump style AArch64 disassembler output (using --disassembler-color option). Once the required GDB patches are merged, GDB will also style the disassembler output. The changes to support styling are mostly split between two files opcodes/aarch64-dis.c and opcodes/aarch64-opc.c. The entry point for the AArch64 disassembler can be found in aarch64-dis.c, this file handles printing the instruction mnemonics, and assembler directives (e.g. '.byte', '.word', etc). Some operands, mostly relating to assembler directives are also printed from this file. This commit changes all of this to pass through suitable styling information. However, for most "normal" instructions, the instruction operands are printed using a two step process. From aarch64-dis.c, in the print_operands function, the function aarch64_print_operand is called, this function is in aarch64-opc.c, and converts an instruction operand into a string. Then, back in print_operands (aarch64-dis.c), the operand string is printed. Unfortunately, the string returned by aarch64_print_operand can be quite complex, it will include syntax elements, like '[' and ']', in addition to register names and immediate values. In some cases, a single operand will expand into what will appear (to the user) as multiple operands separated with a ','. This makes the task of styling more complex, all these different components need to by styled differently, so we need to get the styling information out of aarch64_print_operand in some way. The solution that I propose here is similar to the solution that I used for the i386 disassembler. Currently, aarch64_print_operand uses snprintf to write the operand text into a buffer provided by the caller. What I propose is that we pass an extra argument to the aarch64_print_operand function, this argument will be a structure, the structure contains a callback function and some state. When aarch64_print_operand needs to format part of its output this can be done by using the callback function within the new structure, this callback returns a string with special embedded markers that indicate which mode should be used for each piece of text. Back in aarch64-dis.c we can spot these special style markers and use this to split the disassembler output up and apply the correct style to each piece. To make aarch64-opc.c clearer a series of new static functions have been added, e.g. 'style_reg', 'style_imm', etc. Each of these functions formats a piece of text in a different style, 'register' and 'immediate' in this case. Here's an example taken from aarch64-opc.c of the new functions in use: snprintf (buf, size, "[%s, %s]!", style_reg (styler, base), style_imm (styler, "#%d", opnd->addr.offset.imm)); The aarch64_print_operand function is also called from the assembler to aid in printing diagnostic messages. Right now I have no plans to add styling to the assembler output, and so, the callback function used in the assembler ignores the styling information and just returns an plain string. I've used the source files in gas/testsuite/gas/aarch64/ for testing, and have manually gone through and checked that the styling looks reasonable, however, I'm not an AArch64 expert, so it is possible that the odd piece is styled incorrectly. Please point out any mistakes I've made. With objdump disassembler color turned off, there should be no change in the output after this commit.
2022-04-28 20:31:07 +08:00
snprintf (buf, size, "%s, %s %s",
style_imm (styler, "#%" PRIi64, opnd->imm.value),
style_sub_mnem (styler, "lsl"),
style_imm (styler, "#%" PRIi64, opnd->shifter.amount));
[AArch64][SVE 27/32] Add SVE integer immediate operands This patch adds the new SVE integer immediate operands. There are three kinds: - simple signed and unsigned ranges, but with new widths and positions. - 13-bit logical immediates. These have the same form as in base AArch64, but at a different bit position. In the case of the "MOV Zn.<T>, #<limm>" alias of DUPM, the logical immediate <limm> is not allowed to be a valid DUP immediate, since DUP is preferred over DUPM for constants that both instructions can handle. - a new 9-bit arithmetic immediate, of the form "<imm8>{, LSL #8}". In some contexts the operand is signed and in others it's unsigned. As an extension, we allow shifted immediates to be written as a single integer, e.g. "#256" is equivalent to "#1, LSL #8". We also use the shiftless form as the preferred disassembly, except for the special case of "#0, LSL #8" (a redundant encoding of 0). include/ * opcode/aarch64.h (AARCH64_OPND_SIMM5): New aarch64_opnd. (AARCH64_OPND_SVE_AIMM, AARCH64_OPND_SVE_ASIMM) (AARCH64_OPND_SVE_INV_LIMM, AARCH64_OPND_SVE_LIMM) (AARCH64_OPND_SVE_LIMM_MOV, AARCH64_OPND_SVE_SHLIMM_PRED) (AARCH64_OPND_SVE_SHLIMM_UNPRED, AARCH64_OPND_SVE_SHRIMM_PRED) (AARCH64_OPND_SVE_SHRIMM_UNPRED, AARCH64_OPND_SVE_SIMM5) (AARCH64_OPND_SVE_SIMM5B, AARCH64_OPND_SVE_SIMM6) (AARCH64_OPND_SVE_SIMM8, AARCH64_OPND_SVE_UIMM3) (AARCH64_OPND_SVE_UIMM7, AARCH64_OPND_SVE_UIMM8) (AARCH64_OPND_SVE_UIMM8_53): Likewise. (aarch64_sve_dupm_mov_immediate_p): Declare. opcodes/ * aarch64-tbl.h (AARCH64_OPERANDS): Add entries for the new SVE integer immediate operands. * aarch64-opc.h (FLD_SVE_immN, FLD_SVE_imm3, FLD_SVE_imm5) (FLD_SVE_imm5b, FLD_SVE_imm7, FLD_SVE_imm8, FLD_SVE_imm9) (FLD_SVE_immr, FLD_SVE_imms, FLD_SVE_tszh): New aarch64_field_kinds. * aarch64-opc.c (fields): Add corresponding entries. (operand_general_constraint_met_p): Handle the new SVE integer immediate operands. (aarch64_print_operand): Likewise. (aarch64_sve_dupm_mov_immediate_p): New function. * aarch64-opc-2.c: Regenerate. * aarch64-asm.h (ins_inv_limm, ins_sve_aimm, ins_sve_asimm) (ins_sve_limm_mov, ins_sve_shlimm, ins_sve_shrimm): New inserters. * aarch64-asm.c (aarch64_ins_limm_1): New function, split out from... (aarch64_ins_limm): ...here. (aarch64_ins_inv_limm): New function. (aarch64_ins_sve_aimm): Likewise. (aarch64_ins_sve_asimm): Likewise. (aarch64_ins_sve_limm_mov): Likewise. (aarch64_ins_sve_shlimm): Likewise. (aarch64_ins_sve_shrimm): Likewise. * aarch64-asm-2.c: Regenerate. * aarch64-dis.h (ext_inv_limm, ext_sve_aimm, ext_sve_asimm) (ext_sve_limm_mov, ext_sve_shlimm, ext_sve_shrimm): New extractors. * aarch64-dis.c (decode_limm): New function, split out from... (aarch64_ext_limm): ...here. (aarch64_ext_inv_limm): New function. (decode_sve_aimm): Likewise. (aarch64_ext_sve_aimm): Likewise. (aarch64_ext_sve_asimm): Likewise. (aarch64_ext_sve_limm_mov): Likewise. (aarch64_top_bit): Likewise. (aarch64_ext_sve_shlimm): Likewise. (aarch64_ext_sve_shrimm): Likewise. * aarch64-dis-2.c: Regenerate. gas/ * config/tc-aarch64.c (parse_operands): Handle the new SVE integer immediate operands.
2016-09-21 23:56:57 +08:00
else
libopcodes/aarch64: add support for disassembler styling This commit enables disassembler styling for AArch64. After this commit it is possible to have objdump style AArch64 disassembler output (using --disassembler-color option). Once the required GDB patches are merged, GDB will also style the disassembler output. The changes to support styling are mostly split between two files opcodes/aarch64-dis.c and opcodes/aarch64-opc.c. The entry point for the AArch64 disassembler can be found in aarch64-dis.c, this file handles printing the instruction mnemonics, and assembler directives (e.g. '.byte', '.word', etc). Some operands, mostly relating to assembler directives are also printed from this file. This commit changes all of this to pass through suitable styling information. However, for most "normal" instructions, the instruction operands are printed using a two step process. From aarch64-dis.c, in the print_operands function, the function aarch64_print_operand is called, this function is in aarch64-opc.c, and converts an instruction operand into a string. Then, back in print_operands (aarch64-dis.c), the operand string is printed. Unfortunately, the string returned by aarch64_print_operand can be quite complex, it will include syntax elements, like '[' and ']', in addition to register names and immediate values. In some cases, a single operand will expand into what will appear (to the user) as multiple operands separated with a ','. This makes the task of styling more complex, all these different components need to by styled differently, so we need to get the styling information out of aarch64_print_operand in some way. The solution that I propose here is similar to the solution that I used for the i386 disassembler. Currently, aarch64_print_operand uses snprintf to write the operand text into a buffer provided by the caller. What I propose is that we pass an extra argument to the aarch64_print_operand function, this argument will be a structure, the structure contains a callback function and some state. When aarch64_print_operand needs to format part of its output this can be done by using the callback function within the new structure, this callback returns a string with special embedded markers that indicate which mode should be used for each piece of text. Back in aarch64-dis.c we can spot these special style markers and use this to split the disassembler output up and apply the correct style to each piece. To make aarch64-opc.c clearer a series of new static functions have been added, e.g. 'style_reg', 'style_imm', etc. Each of these functions formats a piece of text in a different style, 'register' and 'immediate' in this case. Here's an example taken from aarch64-opc.c of the new functions in use: snprintf (buf, size, "[%s, %s]!", style_reg (styler, base), style_imm (styler, "#%d", opnd->addr.offset.imm)); The aarch64_print_operand function is also called from the assembler to aid in printing diagnostic messages. Right now I have no plans to add styling to the assembler output, and so, the callback function used in the assembler ignores the styling information and just returns an plain string. I've used the source files in gas/testsuite/gas/aarch64/ for testing, and have manually gone through and checked that the styling looks reasonable, however, I'm not an AArch64 expert, so it is possible that the odd piece is styled incorrectly. Please point out any mistakes I've made. With objdump disassembler color turned off, there should be no change in the output after this commit.
2022-04-28 20:31:07 +08:00
snprintf (buf, size, "%s",
style_imm (styler, "#%" PRIi64, opnd->imm.value));
[AArch64][SVE 27/32] Add SVE integer immediate operands This patch adds the new SVE integer immediate operands. There are three kinds: - simple signed and unsigned ranges, but with new widths and positions. - 13-bit logical immediates. These have the same form as in base AArch64, but at a different bit position. In the case of the "MOV Zn.<T>, #<limm>" alias of DUPM, the logical immediate <limm> is not allowed to be a valid DUP immediate, since DUP is preferred over DUPM for constants that both instructions can handle. - a new 9-bit arithmetic immediate, of the form "<imm8>{, LSL #8}". In some contexts the operand is signed and in others it's unsigned. As an extension, we allow shifted immediates to be written as a single integer, e.g. "#256" is equivalent to "#1, LSL #8". We also use the shiftless form as the preferred disassembly, except for the special case of "#0, LSL #8" (a redundant encoding of 0). include/ * opcode/aarch64.h (AARCH64_OPND_SIMM5): New aarch64_opnd. (AARCH64_OPND_SVE_AIMM, AARCH64_OPND_SVE_ASIMM) (AARCH64_OPND_SVE_INV_LIMM, AARCH64_OPND_SVE_LIMM) (AARCH64_OPND_SVE_LIMM_MOV, AARCH64_OPND_SVE_SHLIMM_PRED) (AARCH64_OPND_SVE_SHLIMM_UNPRED, AARCH64_OPND_SVE_SHRIMM_PRED) (AARCH64_OPND_SVE_SHRIMM_UNPRED, AARCH64_OPND_SVE_SIMM5) (AARCH64_OPND_SVE_SIMM5B, AARCH64_OPND_SVE_SIMM6) (AARCH64_OPND_SVE_SIMM8, AARCH64_OPND_SVE_UIMM3) (AARCH64_OPND_SVE_UIMM7, AARCH64_OPND_SVE_UIMM8) (AARCH64_OPND_SVE_UIMM8_53): Likewise. (aarch64_sve_dupm_mov_immediate_p): Declare. opcodes/ * aarch64-tbl.h (AARCH64_OPERANDS): Add entries for the new SVE integer immediate operands. * aarch64-opc.h (FLD_SVE_immN, FLD_SVE_imm3, FLD_SVE_imm5) (FLD_SVE_imm5b, FLD_SVE_imm7, FLD_SVE_imm8, FLD_SVE_imm9) (FLD_SVE_immr, FLD_SVE_imms, FLD_SVE_tszh): New aarch64_field_kinds. * aarch64-opc.c (fields): Add corresponding entries. (operand_general_constraint_met_p): Handle the new SVE integer immediate operands. (aarch64_print_operand): Likewise. (aarch64_sve_dupm_mov_immediate_p): New function. * aarch64-opc-2.c: Regenerate. * aarch64-asm.h (ins_inv_limm, ins_sve_aimm, ins_sve_asimm) (ins_sve_limm_mov, ins_sve_shlimm, ins_sve_shrimm): New inserters. * aarch64-asm.c (aarch64_ins_limm_1): New function, split out from... (aarch64_ins_limm): ...here. (aarch64_ins_inv_limm): New function. (aarch64_ins_sve_aimm): Likewise. (aarch64_ins_sve_asimm): Likewise. (aarch64_ins_sve_limm_mov): Likewise. (aarch64_ins_sve_shlimm): Likewise. (aarch64_ins_sve_shrimm): Likewise. * aarch64-asm-2.c: Regenerate. * aarch64-dis.h (ext_inv_limm, ext_sve_aimm, ext_sve_asimm) (ext_sve_limm_mov, ext_sve_shlimm, ext_sve_shrimm): New extractors. * aarch64-dis.c (decode_limm): New function, split out from... (aarch64_ext_limm): ...here. (aarch64_ext_inv_limm): New function. (decode_sve_aimm): Likewise. (aarch64_ext_sve_aimm): Likewise. (aarch64_ext_sve_asimm): Likewise. (aarch64_ext_sve_limm_mov): Likewise. (aarch64_top_bit): Likewise. (aarch64_ext_sve_shlimm): Likewise. (aarch64_ext_sve_shrimm): Likewise. * aarch64-dis-2.c: Regenerate. gas/ * config/tc-aarch64.c (parse_operands): Handle the new SVE integer immediate operands.
2016-09-21 23:56:57 +08:00
break;
case AARCH64_OPND_FPIMM:
case AARCH64_OPND_SIMD_FPIMM:
[AArch64][SVE 28/32] Add SVE FP immediate operands This patch adds support for the new SVE floating-point immediate operands. One operand uses the same 8-bit encoding as base AArch64, but in a different position. The others use a single bit to select between two values. One of the single-bit operands is a choice between 0 and 1, where 0 is not a valid 8-bit encoding. I think the cleanest way of handling these single-bit immediates is therefore to use the IEEE float encoding itself as the immediate value and select between the two possible values when encoding and decoding. As described in the covering note for the patch that added F_STRICT, we get better error messages by accepting unsuffixed vector registers and leaving the qualifier matching code to report an error. This means that we carry on parsing the other operands, and so can try to parse FP immediates for invalid instructions like: fcpy z0, #2.5 In this case there is no suffix to tell us whether the immediate should be treated as single or double precision. Again, we get better error messages by picking one (arbitrary) immediate size and reporting an error for the missing suffix later. include/ * opcode/aarch64.h (AARCH64_OPND_SVE_FPIMM8): New aarch64_opnd. (AARCH64_OPND_SVE_I1_HALF_ONE, AARCH64_OPND_SVE_I1_HALF_TWO) (AARCH64_OPND_SVE_I1_ZERO_ONE): Likewise. opcodes/ * aarch64-tbl.h (AARCH64_OPERANDS): Add entries for the new SVE FP immediate operands. * aarch64-opc.h (FLD_SVE_i1): New aarch64_field_kind. * aarch64-opc.c (fields): Add corresponding entry. (operand_general_constraint_met_p): Handle the new SVE FP immediate operands. (aarch64_print_operand): Likewise. * aarch64-opc-2.c: Regenerate. * aarch64-asm.h (ins_sve_float_half_one, ins_sve_float_half_two) (ins_sve_float_zero_one): New inserters. * aarch64-asm.c (aarch64_ins_sve_float_half_one): New function. (aarch64_ins_sve_float_half_two): Likewise. (aarch64_ins_sve_float_zero_one): Likewise. * aarch64-asm-2.c: Regenerate. * aarch64-dis.h (ext_sve_float_half_one, ext_sve_float_half_two) (ext_sve_float_zero_one): New extractors. * aarch64-dis.c (aarch64_ext_sve_float_half_one): New function. (aarch64_ext_sve_float_half_two): Likewise. (aarch64_ext_sve_float_zero_one): Likewise. * aarch64-dis-2.c: Regenerate. gas/ * config/tc-aarch64.c (double_precision_operand_p): New function. (parse_operands): Use it to calculate the dp_p input to parse_aarch64_imm_float. Handle the new SVE FP immediate operands.
2016-09-21 23:57:22 +08:00
case AARCH64_OPND_SVE_FPIMM8:
switch (aarch64_get_qualifier_esize (opnds[0].qualifier))
{
case 2: /* e.g. FMOV <Hd>, #<imm>. */
{
half_conv_t c;
c.i = expand_fp_imm (2, opnd->imm.value);
libopcodes/aarch64: add support for disassembler styling This commit enables disassembler styling for AArch64. After this commit it is possible to have objdump style AArch64 disassembler output (using --disassembler-color option). Once the required GDB patches are merged, GDB will also style the disassembler output. The changes to support styling are mostly split between two files opcodes/aarch64-dis.c and opcodes/aarch64-opc.c. The entry point for the AArch64 disassembler can be found in aarch64-dis.c, this file handles printing the instruction mnemonics, and assembler directives (e.g. '.byte', '.word', etc). Some operands, mostly relating to assembler directives are also printed from this file. This commit changes all of this to pass through suitable styling information. However, for most "normal" instructions, the instruction operands are printed using a two step process. From aarch64-dis.c, in the print_operands function, the function aarch64_print_operand is called, this function is in aarch64-opc.c, and converts an instruction operand into a string. Then, back in print_operands (aarch64-dis.c), the operand string is printed. Unfortunately, the string returned by aarch64_print_operand can be quite complex, it will include syntax elements, like '[' and ']', in addition to register names and immediate values. In some cases, a single operand will expand into what will appear (to the user) as multiple operands separated with a ','. This makes the task of styling more complex, all these different components need to by styled differently, so we need to get the styling information out of aarch64_print_operand in some way. The solution that I propose here is similar to the solution that I used for the i386 disassembler. Currently, aarch64_print_operand uses snprintf to write the operand text into a buffer provided by the caller. What I propose is that we pass an extra argument to the aarch64_print_operand function, this argument will be a structure, the structure contains a callback function and some state. When aarch64_print_operand needs to format part of its output this can be done by using the callback function within the new structure, this callback returns a string with special embedded markers that indicate which mode should be used for each piece of text. Back in aarch64-dis.c we can spot these special style markers and use this to split the disassembler output up and apply the correct style to each piece. To make aarch64-opc.c clearer a series of new static functions have been added, e.g. 'style_reg', 'style_imm', etc. Each of these functions formats a piece of text in a different style, 'register' and 'immediate' in this case. Here's an example taken from aarch64-opc.c of the new functions in use: snprintf (buf, size, "[%s, %s]!", style_reg (styler, base), style_imm (styler, "#%d", opnd->addr.offset.imm)); The aarch64_print_operand function is also called from the assembler to aid in printing diagnostic messages. Right now I have no plans to add styling to the assembler output, and so, the callback function used in the assembler ignores the styling information and just returns an plain string. I've used the source files in gas/testsuite/gas/aarch64/ for testing, and have manually gone through and checked that the styling looks reasonable, however, I'm not an AArch64 expert, so it is possible that the odd piece is styled incorrectly. Please point out any mistakes I've made. With objdump disassembler color turned off, there should be no change in the output after this commit.
2022-04-28 20:31:07 +08:00
snprintf (buf, size, "%s", style_imm (styler, "#%.18e", c.f));
}
break;
case 4: /* e.g. FMOV <Vd>.4S, #<imm>. */
{
single_conv_t c;
c.i = expand_fp_imm (4, opnd->imm.value);
libopcodes/aarch64: add support for disassembler styling This commit enables disassembler styling for AArch64. After this commit it is possible to have objdump style AArch64 disassembler output (using --disassembler-color option). Once the required GDB patches are merged, GDB will also style the disassembler output. The changes to support styling are mostly split between two files opcodes/aarch64-dis.c and opcodes/aarch64-opc.c. The entry point for the AArch64 disassembler can be found in aarch64-dis.c, this file handles printing the instruction mnemonics, and assembler directives (e.g. '.byte', '.word', etc). Some operands, mostly relating to assembler directives are also printed from this file. This commit changes all of this to pass through suitable styling information. However, for most "normal" instructions, the instruction operands are printed using a two step process. From aarch64-dis.c, in the print_operands function, the function aarch64_print_operand is called, this function is in aarch64-opc.c, and converts an instruction operand into a string. Then, back in print_operands (aarch64-dis.c), the operand string is printed. Unfortunately, the string returned by aarch64_print_operand can be quite complex, it will include syntax elements, like '[' and ']', in addition to register names and immediate values. In some cases, a single operand will expand into what will appear (to the user) as multiple operands separated with a ','. This makes the task of styling more complex, all these different components need to by styled differently, so we need to get the styling information out of aarch64_print_operand in some way. The solution that I propose here is similar to the solution that I used for the i386 disassembler. Currently, aarch64_print_operand uses snprintf to write the operand text into a buffer provided by the caller. What I propose is that we pass an extra argument to the aarch64_print_operand function, this argument will be a structure, the structure contains a callback function and some state. When aarch64_print_operand needs to format part of its output this can be done by using the callback function within the new structure, this callback returns a string with special embedded markers that indicate which mode should be used for each piece of text. Back in aarch64-dis.c we can spot these special style markers and use this to split the disassembler output up and apply the correct style to each piece. To make aarch64-opc.c clearer a series of new static functions have been added, e.g. 'style_reg', 'style_imm', etc. Each of these functions formats a piece of text in a different style, 'register' and 'immediate' in this case. Here's an example taken from aarch64-opc.c of the new functions in use: snprintf (buf, size, "[%s, %s]!", style_reg (styler, base), style_imm (styler, "#%d", opnd->addr.offset.imm)); The aarch64_print_operand function is also called from the assembler to aid in printing diagnostic messages. Right now I have no plans to add styling to the assembler output, and so, the callback function used in the assembler ignores the styling information and just returns an plain string. I've used the source files in gas/testsuite/gas/aarch64/ for testing, and have manually gone through and checked that the styling looks reasonable, however, I'm not an AArch64 expert, so it is possible that the odd piece is styled incorrectly. Please point out any mistakes I've made. With objdump disassembler color turned off, there should be no change in the output after this commit.
2022-04-28 20:31:07 +08:00
snprintf (buf, size, "%s", style_imm (styler, "#%.18e", c.f));
}
break;
case 8: /* e.g. FMOV <Sd>, #<imm>. */
{
double_conv_t c;
c.i = expand_fp_imm (8, opnd->imm.value);
libopcodes/aarch64: add support for disassembler styling This commit enables disassembler styling for AArch64. After this commit it is possible to have objdump style AArch64 disassembler output (using --disassembler-color option). Once the required GDB patches are merged, GDB will also style the disassembler output. The changes to support styling are mostly split between two files opcodes/aarch64-dis.c and opcodes/aarch64-opc.c. The entry point for the AArch64 disassembler can be found in aarch64-dis.c, this file handles printing the instruction mnemonics, and assembler directives (e.g. '.byte', '.word', etc). Some operands, mostly relating to assembler directives are also printed from this file. This commit changes all of this to pass through suitable styling information. However, for most "normal" instructions, the instruction operands are printed using a two step process. From aarch64-dis.c, in the print_operands function, the function aarch64_print_operand is called, this function is in aarch64-opc.c, and converts an instruction operand into a string. Then, back in print_operands (aarch64-dis.c), the operand string is printed. Unfortunately, the string returned by aarch64_print_operand can be quite complex, it will include syntax elements, like '[' and ']', in addition to register names and immediate values. In some cases, a single operand will expand into what will appear (to the user) as multiple operands separated with a ','. This makes the task of styling more complex, all these different components need to by styled differently, so we need to get the styling information out of aarch64_print_operand in some way. The solution that I propose here is similar to the solution that I used for the i386 disassembler. Currently, aarch64_print_operand uses snprintf to write the operand text into a buffer provided by the caller. What I propose is that we pass an extra argument to the aarch64_print_operand function, this argument will be a structure, the structure contains a callback function and some state. When aarch64_print_operand needs to format part of its output this can be done by using the callback function within the new structure, this callback returns a string with special embedded markers that indicate which mode should be used for each piece of text. Back in aarch64-dis.c we can spot these special style markers and use this to split the disassembler output up and apply the correct style to each piece. To make aarch64-opc.c clearer a series of new static functions have been added, e.g. 'style_reg', 'style_imm', etc. Each of these functions formats a piece of text in a different style, 'register' and 'immediate' in this case. Here's an example taken from aarch64-opc.c of the new functions in use: snprintf (buf, size, "[%s, %s]!", style_reg (styler, base), style_imm (styler, "#%d", opnd->addr.offset.imm)); The aarch64_print_operand function is also called from the assembler to aid in printing diagnostic messages. Right now I have no plans to add styling to the assembler output, and so, the callback function used in the assembler ignores the styling information and just returns an plain string. I've used the source files in gas/testsuite/gas/aarch64/ for testing, and have manually gone through and checked that the styling looks reasonable, however, I'm not an AArch64 expert, so it is possible that the odd piece is styled incorrectly. Please point out any mistakes I've made. With objdump disassembler color turned off, there should be no change in the output after this commit.
2022-04-28 20:31:07 +08:00
snprintf (buf, size, "%s", style_imm (styler, "#%.18e", c.d));
}
break;
default:
snprintf (buf, size, "<invalid>");
break;
}
break;
case AARCH64_OPND_CCMP_IMM:
case AARCH64_OPND_NZCV:
case AARCH64_OPND_EXCEPTION:
case AARCH64_OPND_UIMM4:
[BINUTILS, AARCH64, 2/8] Add Tag generation instructions in Memory Tagging Extension This patch is part of the patch series to add support for ARMv8.5-A Memory Tagging Extensions which is an optional extension to ARMv8.5-A and is enabled using the +memtag command line option. This patch add support to the Tag generation instructions from MTE. These are the following instructions added in this patch: - IRG <Xd|SP>, <Xn|SP>{, Xm} - ADDG <Xd|SP>, <Xn|SP>, #<uimm1>. #<uimm2> - SUBG <Xd|SP>, <Xn|SP>, #<uimm1>. #<uimm2> - GMI <Xd>, <Xn|SP>, <Xm> where <Xd|SP> : Is the 64-bit destination GPR or Stack pointer. <Xn|SP> : Is the 64-bit source GPR or Stack pointer. <uimm6> : Is the unsigned immediate, a multiple of 16 in the range 0 to 1008. <uimm4> : Is the unsigned immediate, in the range 0 to 15. *** include/ChangeLog *** 2018-11-12 Sudakshina Das <sudi.das@arm.com> * opcode/aarch64.h (aarch64_opnd): Add AARCH64_OPND_UIMM4_ADDG and AARCH64_OPND_UIMM10 as new enums. *** opcodes/ChangeLog *** 2018-11-12 Sudakshina Das <sudi.das@arm.com> * aarch64-opc.h (aarch64_field_kind): New FLD_imm4_3. (OPD_F_SHIFT_BY_4, operand_need_shift_by_four): New. * aarch64-opc.c (fields): Add entry for imm4_3. (operand_general_constraint_met_p): Add cases for AARCH64_OPND_UIMM4_ADDG and AARCH64_OPND_UIMM10. (aarch64_print_operand): Likewise. * aarch64-tbl.h (QL_ADDG): New. (aarch64_opcode_table): Add addg, subg, irg and gmi. (AARCH64_OPERANDS): Define UIMM4_ADDG and UIMM10. * aarch64-asm.c (aarch64_ins_imm): Add case for operand_need_shift_by_four. * aarch64-asm-2.c: Regenerated. * aarch64-dis-2.c: Regenerated. * aarch64-opc-2.c: Regenerated. *** gas/ChangeLog *** 2018-11-12 Sudakshina Das <sudi.das@arm.com> * config/tc-aarch64.c (parse_operands): Add switch case for AARCH64_OPND_UIMM4_ADDG and AARCH64_OPND_UIMM10. * testsuite/gas/aarch64/armv8_5-a-memtag.s: New. * testsuite/gas/aarch64/armv8_5-a-memtag.d: Likewise. * testsuite/gas/aarch64/illegal-memtag.s: Likewise. * testsuite/gas/aarch64/illegal-memtag.l: Likewise. * testsuite/gas/aarch64/illegal-memtag.d: Likewise.
2018-11-12 20:52:55 +08:00
case AARCH64_OPND_UIMM4_ADDG:
case AARCH64_OPND_UIMM7:
[BINUTILS, AARCH64, 2/8] Add Tag generation instructions in Memory Tagging Extension This patch is part of the patch series to add support for ARMv8.5-A Memory Tagging Extensions which is an optional extension to ARMv8.5-A and is enabled using the +memtag command line option. This patch add support to the Tag generation instructions from MTE. These are the following instructions added in this patch: - IRG <Xd|SP>, <Xn|SP>{, Xm} - ADDG <Xd|SP>, <Xn|SP>, #<uimm1>. #<uimm2> - SUBG <Xd|SP>, <Xn|SP>, #<uimm1>. #<uimm2> - GMI <Xd>, <Xn|SP>, <Xm> where <Xd|SP> : Is the 64-bit destination GPR or Stack pointer. <Xn|SP> : Is the 64-bit source GPR or Stack pointer. <uimm6> : Is the unsigned immediate, a multiple of 16 in the range 0 to 1008. <uimm4> : Is the unsigned immediate, in the range 0 to 15. *** include/ChangeLog *** 2018-11-12 Sudakshina Das <sudi.das@arm.com> * opcode/aarch64.h (aarch64_opnd): Add AARCH64_OPND_UIMM4_ADDG and AARCH64_OPND_UIMM10 as new enums. *** opcodes/ChangeLog *** 2018-11-12 Sudakshina Das <sudi.das@arm.com> * aarch64-opc.h (aarch64_field_kind): New FLD_imm4_3. (OPD_F_SHIFT_BY_4, operand_need_shift_by_four): New. * aarch64-opc.c (fields): Add entry for imm4_3. (operand_general_constraint_met_p): Add cases for AARCH64_OPND_UIMM4_ADDG and AARCH64_OPND_UIMM10. (aarch64_print_operand): Likewise. * aarch64-tbl.h (QL_ADDG): New. (aarch64_opcode_table): Add addg, subg, irg and gmi. (AARCH64_OPERANDS): Define UIMM4_ADDG and UIMM10. * aarch64-asm.c (aarch64_ins_imm): Add case for operand_need_shift_by_four. * aarch64-asm-2.c: Regenerated. * aarch64-dis-2.c: Regenerated. * aarch64-opc-2.c: Regenerated. *** gas/ChangeLog *** 2018-11-12 Sudakshina Das <sudi.das@arm.com> * config/tc-aarch64.c (parse_operands): Add switch case for AARCH64_OPND_UIMM4_ADDG and AARCH64_OPND_UIMM10. * testsuite/gas/aarch64/armv8_5-a-memtag.s: New. * testsuite/gas/aarch64/armv8_5-a-memtag.d: Likewise. * testsuite/gas/aarch64/illegal-memtag.s: Likewise. * testsuite/gas/aarch64/illegal-memtag.l: Likewise. * testsuite/gas/aarch64/illegal-memtag.d: Likewise.
2018-11-12 20:52:55 +08:00
case AARCH64_OPND_UIMM10:
TRUE/FALSE simplification There is really no need to write code like "foo != 0 ? TRUE : FALSE" unless we had stupidly defined FALSE as something other than 0 or TRUE as something other than 1. The simpler "foo != 0" does just as well. Similarly "(condition == TRUE)" or "(condition == FALSE) can be simplified to "(condition)" and "(!condition)" respectively. I'll note that there is reason to use "integer_expression != 0" when assigning a bfd_boolean rather than the simpler "integer_expression", if you expect the variable to have 0 or 1 value. It's probably even a good idea to not rely on implicit conversion if bfd_boolean were _Bool. bfd/ * aoutx.h (aout_link_write_symbols): Don't cast boolean expression to bfd_boolean. * elf32-or1k.c (or1k_set_got_and_rela_sizes): Dont compare booleans against FALSE. * elf32-arc.c (name_for_global_symbol): Don't compare boolean to TRUE. (is_reloc_PC_relative): Don't use "boolean_condition ? TRUE : FALSE". (is_reloc_SDA_relative, is_reloc_for_GOT): Likewise. (is_reloc_for_PLT, is_reloc_for_TLS): Likewise. * elf32-arm.c (stm32l4xx_need_create_replacing_stub): Likewise. * elf32-nds32.c (insert_nds32_elf_blank): Likewise. * elf32-rx.c (rx_set_section_contents): Likewise. * elfnn-aarch64.c (elfNN_aarch64_final_link_relocate): Likewise. * elfxx-mips.c (_bfd_mips_elf_ignore_undef_symbol): Likewise. * mach-o.c (bfd_mach_o_read_command): Likewise. * targets.c (bfd_get_target_info): Likewise. binutils/ * dlltool.c (main): Don't use "boolean_condition ? TRUE : FALSE". * dwarf.c (read_and_display_attr_value): Likewise. (display_debug_str_offsets): Likewise. * objdump.c (dump_bfd): Likewise. * readelf.c (dump_section_as_strings): Likewise. (dump_section_as_bytes): Likewise. gas/ * atof-generic.c (FALSE, TRUE): Don't define. * config/obj-elf.h (FALSE, TRUE): Don't define. * config/obj-som.h (FALSE, TRUE): Don't define. * config/tc-hppa.h (FALSE, TRUE): Don't define. * config/tc-pdp11.c (FALSE, TRUE): Don't define. * config/tc-iq2000.h (obj_fix_adjustable): Delete. * config/tc-m32r.h (TC_FIX_ADJUSTABLE): Delete. * config/tc-mt.h (obj_fix_adjustable): Delete. * config/tc-nds32.h (TC_FIX_ADJUSTABLE): Delete. * config/tc-arc.c (parse_opcode_flags): Simplify boolean expression. (relaxable_flag, relaxable_operand, assemble_insn): Likewise. (tokenize_extregister): Likewise. * config/tc-csky.c (parse_opcode, get_operand_value): Likewise. (parse_operands_op, parse_operands, md_assemble): Likewise. * config/tc-d10v.c (build_insn): Likewise. * config/tc-score.c (s3_gen_insn_frag): Likewise. * config/tc-score7.c (s7_gen_insn_frag, s7_relax_frag): Likewise. * config/tc-tic6x.c (tic6x_update_features, md_assemble): Likewise. * config/tc-z80.c (emit_byte): Likewise. include/ * opcode/aarch64.h (alias_opcode_p): Simplify boolean expression. (opcode_has_alias, pseudo_opcode_p, optional_operand_p): Likewise. (opcode_has_special_coder): Likewise. ld/ * emultempl/aix.em (gld${EMULATION_NAME}_before_allocation): Simplify boolean expression. * lexsup.c (parse_args): Likewise. * pe-dll.c (pe_dll_id_target): Likewise. opcodes/ * aarch64-opc.c (vector_qualifier_p): Simplify boolean expression. (fp_qualifier_p, get_data_pattern): Likewise. (aarch64_get_operand_modifier_from_value): Likewise. (aarch64_extend_operator_p, aarch64_shift_operator_p): Likewise. (operand_variant_qualifier_p): Likewise. (qualifier_value_in_range_constraint_p): Likewise. (aarch64_get_qualifier_esize): Likewise. (aarch64_get_qualifier_nelem): Likewise. (aarch64_get_qualifier_standard_value): Likewise. (get_lower_bound, get_upper_bound): Likewise. (aarch64_find_best_match, match_operands_qualifier): Likewise. (aarch64_print_operand): Likewise. * aarch64-opc.h (operand_has_inserter, operand_has_extractor): Likewise. (operand_need_sign_extension, operand_need_shift_by_two): Likewise. (operand_need_shift_by_four, operand_maybe_stack_pointer): Likewise. * arm-dis.c (print_insn_mve, print_insn_thumb32): Likewise. * tic6x-dis.c (tic6x_check_fetch_packet_header): Likewise. (print_insn_tic6x): Likewise.
2021-03-29 07:22:56 +08:00
if (optional_operand_p (opcode, idx)
&& (opnd->imm.value ==
(int64_t) get_optional_operand_default_value (opcode)))
/* Omit the operand, e.g. DCPS1. */
break;
libopcodes/aarch64: add support for disassembler styling This commit enables disassembler styling for AArch64. After this commit it is possible to have objdump style AArch64 disassembler output (using --disassembler-color option). Once the required GDB patches are merged, GDB will also style the disassembler output. The changes to support styling are mostly split between two files opcodes/aarch64-dis.c and opcodes/aarch64-opc.c. The entry point for the AArch64 disassembler can be found in aarch64-dis.c, this file handles printing the instruction mnemonics, and assembler directives (e.g. '.byte', '.word', etc). Some operands, mostly relating to assembler directives are also printed from this file. This commit changes all of this to pass through suitable styling information. However, for most "normal" instructions, the instruction operands are printed using a two step process. From aarch64-dis.c, in the print_operands function, the function aarch64_print_operand is called, this function is in aarch64-opc.c, and converts an instruction operand into a string. Then, back in print_operands (aarch64-dis.c), the operand string is printed. Unfortunately, the string returned by aarch64_print_operand can be quite complex, it will include syntax elements, like '[' and ']', in addition to register names and immediate values. In some cases, a single operand will expand into what will appear (to the user) as multiple operands separated with a ','. This makes the task of styling more complex, all these different components need to by styled differently, so we need to get the styling information out of aarch64_print_operand in some way. The solution that I propose here is similar to the solution that I used for the i386 disassembler. Currently, aarch64_print_operand uses snprintf to write the operand text into a buffer provided by the caller. What I propose is that we pass an extra argument to the aarch64_print_operand function, this argument will be a structure, the structure contains a callback function and some state. When aarch64_print_operand needs to format part of its output this can be done by using the callback function within the new structure, this callback returns a string with special embedded markers that indicate which mode should be used for each piece of text. Back in aarch64-dis.c we can spot these special style markers and use this to split the disassembler output up and apply the correct style to each piece. To make aarch64-opc.c clearer a series of new static functions have been added, e.g. 'style_reg', 'style_imm', etc. Each of these functions formats a piece of text in a different style, 'register' and 'immediate' in this case. Here's an example taken from aarch64-opc.c of the new functions in use: snprintf (buf, size, "[%s, %s]!", style_reg (styler, base), style_imm (styler, "#%d", opnd->addr.offset.imm)); The aarch64_print_operand function is also called from the assembler to aid in printing diagnostic messages. Right now I have no plans to add styling to the assembler output, and so, the callback function used in the assembler ignores the styling information and just returns an plain string. I've used the source files in gas/testsuite/gas/aarch64/ for testing, and have manually gone through and checked that the styling looks reasonable, however, I'm not an AArch64 expert, so it is possible that the odd piece is styled incorrectly. Please point out any mistakes I've made. With objdump disassembler color turned off, there should be no change in the output after this commit.
2022-04-28 20:31:07 +08:00
snprintf (buf, size, "%s",
style_imm (styler, "#0x%x", (unsigned int) opnd->imm.value));
break;
case AARCH64_OPND_COND:
case AARCH64_OPND_COND1:
libopcodes/aarch64: add support for disassembler styling This commit enables disassembler styling for AArch64. After this commit it is possible to have objdump style AArch64 disassembler output (using --disassembler-color option). Once the required GDB patches are merged, GDB will also style the disassembler output. The changes to support styling are mostly split between two files opcodes/aarch64-dis.c and opcodes/aarch64-opc.c. The entry point for the AArch64 disassembler can be found in aarch64-dis.c, this file handles printing the instruction mnemonics, and assembler directives (e.g. '.byte', '.word', etc). Some operands, mostly relating to assembler directives are also printed from this file. This commit changes all of this to pass through suitable styling information. However, for most "normal" instructions, the instruction operands are printed using a two step process. From aarch64-dis.c, in the print_operands function, the function aarch64_print_operand is called, this function is in aarch64-opc.c, and converts an instruction operand into a string. Then, back in print_operands (aarch64-dis.c), the operand string is printed. Unfortunately, the string returned by aarch64_print_operand can be quite complex, it will include syntax elements, like '[' and ']', in addition to register names and immediate values. In some cases, a single operand will expand into what will appear (to the user) as multiple operands separated with a ','. This makes the task of styling more complex, all these different components need to by styled differently, so we need to get the styling information out of aarch64_print_operand in some way. The solution that I propose here is similar to the solution that I used for the i386 disassembler. Currently, aarch64_print_operand uses snprintf to write the operand text into a buffer provided by the caller. What I propose is that we pass an extra argument to the aarch64_print_operand function, this argument will be a structure, the structure contains a callback function and some state. When aarch64_print_operand needs to format part of its output this can be done by using the callback function within the new structure, this callback returns a string with special embedded markers that indicate which mode should be used for each piece of text. Back in aarch64-dis.c we can spot these special style markers and use this to split the disassembler output up and apply the correct style to each piece. To make aarch64-opc.c clearer a series of new static functions have been added, e.g. 'style_reg', 'style_imm', etc. Each of these functions formats a piece of text in a different style, 'register' and 'immediate' in this case. Here's an example taken from aarch64-opc.c of the new functions in use: snprintf (buf, size, "[%s, %s]!", style_reg (styler, base), style_imm (styler, "#%d", opnd->addr.offset.imm)); The aarch64_print_operand function is also called from the assembler to aid in printing diagnostic messages. Right now I have no plans to add styling to the assembler output, and so, the callback function used in the assembler ignores the styling information and just returns an plain string. I've used the source files in gas/testsuite/gas/aarch64/ for testing, and have manually gone through and checked that the styling looks reasonable, however, I'm not an AArch64 expert, so it is possible that the odd piece is styled incorrectly. Please point out any mistakes I've made. With objdump disassembler color turned off, there should be no change in the output after this commit.
2022-04-28 20:31:07 +08:00
snprintf (buf, size, "%s",
style_sub_mnem (styler, opnd->cond->names[0]));
[AArch64] Add SVE condition codes SVE defines new names for existing NZCV conditions, to reflect the result of instructions like PTEST. This patch adds support for these names. The patch also adds comments to the disassembly output to show the alternative names of a condition code. For example: cinv x0, x1, cc becomes: cinv x0, x1, cc // cc = lo, ul, last and: b.cc f0 <...> becomes: b.cc f0 <...> // b.lo, b.ul, b.last Doing this for the SVE names follows the practice recommended by the SVE specification and is definitely useful when reading SVE code. If the feeling is that it's too distracting elsewhere, we could add an option to turn it off. include/ * opcode/aarch64.h (aarch64_cond): Bump array size to 4. opcodes/ * aarch64-dis.c (remove_dot_suffix): New function, split out from... (print_mnemonic_name): ...here. (print_comment): New function. (print_aarch64_insn): Call it. * aarch64-opc.c (aarch64_conds): Add SVE names. (aarch64_print_operand): Print alternative condition names in a comment. gas/ * config/tc-aarch64.c (opcode_lookup): Search for the end of a condition name, rather than assuming that it will have exactly 2 characters. (parse_operands): Likewise. * testsuite/gas/aarch64/alias.d: Add new condition-code comments to the expected output. * testsuite/gas/aarch64/beq_1.d: Likewise. * testsuite/gas/aarch64/float-fp16.d: Likewise. * testsuite/gas/aarch64/int-insns.d: Likewise. * testsuite/gas/aarch64/no-aliases.d: Likewise. * testsuite/gas/aarch64/programmer-friendly.d: Likewise. * testsuite/gas/aarch64/reloc-insn.d: Likewise. * testsuite/gas/aarch64/b_c_1.d, testsuite/gas/aarch64/b_c_1.s: New test. ld/ * testsuite/ld-aarch64/emit-relocs-280.d: Match branch comments. * testsuite/ld-aarch64/weak-undefined.d: Likewise.
2016-09-22 00:09:59 +08:00
num_conds = ARRAY_SIZE (opnd->cond->names);
for (i = 1; i < num_conds && opnd->cond->names[i]; ++i)
{
opcodes/aarch64: split off creation of comment text in disassembler The function aarch64_print_operand (aarch64-opc.c) is responsible for converting an instruction operand into the textual representation of that operand. In some cases, a comment is included in the operand representation, though this (currently) only happens for the last operand of the instruction. In a future commit I would like to enable the new libopcodes styling for AArch64, this will allow objdump and GDB[1] to syntax highlight the disassembler output, however, having operands and comments combined in a single string like this makes such styling harder. In this commit, I propose to extend aarch64_print_operand to take a second buffer. Any comments for the instruction are written into this extra buffer. The two callers of aarch64_print_operand are then updated to pass an extra buffer, and print any resulting comment. In this commit no styling is added, that will come later. However, I have adjusted the output slightly. Before this commit some comments would be separated from the instruction operands with a tab character, while in other cases the comment was separated with two single spaces. After this commit I use a single tab character in all cases. This means a few test cases needed updated. If people would prefer me to move everyone to use the two spaces, then just let me know. Or maybe there was a good reason why we used a mix of styles, I could probably figure out a way to maintain the old output exactly if that is critical. Other than that, there should be no user visible changes after this commit. [1] GDB patches have not been merged yet, but have been posted to the GDB mailing list: https://sourceware.org/pipermail/gdb-patches/2022-June/190142.html
2022-06-16 20:46:41 +08:00
size_t len = comment != NULL ? strlen (comment) : 0;
[AArch64] Add SVE condition codes SVE defines new names for existing NZCV conditions, to reflect the result of instructions like PTEST. This patch adds support for these names. The patch also adds comments to the disassembly output to show the alternative names of a condition code. For example: cinv x0, x1, cc becomes: cinv x0, x1, cc // cc = lo, ul, last and: b.cc f0 <...> becomes: b.cc f0 <...> // b.lo, b.ul, b.last Doing this for the SVE names follows the practice recommended by the SVE specification and is definitely useful when reading SVE code. If the feeling is that it's too distracting elsewhere, we could add an option to turn it off. include/ * opcode/aarch64.h (aarch64_cond): Bump array size to 4. opcodes/ * aarch64-dis.c (remove_dot_suffix): New function, split out from... (print_mnemonic_name): ...here. (print_comment): New function. (print_aarch64_insn): Call it. * aarch64-opc.c (aarch64_conds): Add SVE names. (aarch64_print_operand): Print alternative condition names in a comment. gas/ * config/tc-aarch64.c (opcode_lookup): Search for the end of a condition name, rather than assuming that it will have exactly 2 characters. (parse_operands): Likewise. * testsuite/gas/aarch64/alias.d: Add new condition-code comments to the expected output. * testsuite/gas/aarch64/beq_1.d: Likewise. * testsuite/gas/aarch64/float-fp16.d: Likewise. * testsuite/gas/aarch64/int-insns.d: Likewise. * testsuite/gas/aarch64/no-aliases.d: Likewise. * testsuite/gas/aarch64/programmer-friendly.d: Likewise. * testsuite/gas/aarch64/reloc-insn.d: Likewise. * testsuite/gas/aarch64/b_c_1.d, testsuite/gas/aarch64/b_c_1.s: New test. ld/ * testsuite/ld-aarch64/emit-relocs-280.d: Match branch comments. * testsuite/ld-aarch64/weak-undefined.d: Likewise.
2016-09-22 00:09:59 +08:00
if (i == 1)
opcodes/aarch64: split off creation of comment text in disassembler The function aarch64_print_operand (aarch64-opc.c) is responsible for converting an instruction operand into the textual representation of that operand. In some cases, a comment is included in the operand representation, though this (currently) only happens for the last operand of the instruction. In a future commit I would like to enable the new libopcodes styling for AArch64, this will allow objdump and GDB[1] to syntax highlight the disassembler output, however, having operands and comments combined in a single string like this makes such styling harder. In this commit, I propose to extend aarch64_print_operand to take a second buffer. Any comments for the instruction are written into this extra buffer. The two callers of aarch64_print_operand are then updated to pass an extra buffer, and print any resulting comment. In this commit no styling is added, that will come later. However, I have adjusted the output slightly. Before this commit some comments would be separated from the instruction operands with a tab character, while in other cases the comment was separated with two single spaces. After this commit I use a single tab character in all cases. This means a few test cases needed updated. If people would prefer me to move everyone to use the two spaces, then just let me know. Or maybe there was a good reason why we used a mix of styles, I could probably figure out a way to maintain the old output exactly if that is critical. Other than that, there should be no user visible changes after this commit. [1] GDB patches have not been merged yet, but have been posted to the GDB mailing list: https://sourceware.org/pipermail/gdb-patches/2022-June/190142.html
2022-06-16 20:46:41 +08:00
snprintf (comment + len, comment_size - len, "%s = %s",
[AArch64] Add SVE condition codes SVE defines new names for existing NZCV conditions, to reflect the result of instructions like PTEST. This patch adds support for these names. The patch also adds comments to the disassembly output to show the alternative names of a condition code. For example: cinv x0, x1, cc becomes: cinv x0, x1, cc // cc = lo, ul, last and: b.cc f0 <...> becomes: b.cc f0 <...> // b.lo, b.ul, b.last Doing this for the SVE names follows the practice recommended by the SVE specification and is definitely useful when reading SVE code. If the feeling is that it's too distracting elsewhere, we could add an option to turn it off. include/ * opcode/aarch64.h (aarch64_cond): Bump array size to 4. opcodes/ * aarch64-dis.c (remove_dot_suffix): New function, split out from... (print_mnemonic_name): ...here. (print_comment): New function. (print_aarch64_insn): Call it. * aarch64-opc.c (aarch64_conds): Add SVE names. (aarch64_print_operand): Print alternative condition names in a comment. gas/ * config/tc-aarch64.c (opcode_lookup): Search for the end of a condition name, rather than assuming that it will have exactly 2 characters. (parse_operands): Likewise. * testsuite/gas/aarch64/alias.d: Add new condition-code comments to the expected output. * testsuite/gas/aarch64/beq_1.d: Likewise. * testsuite/gas/aarch64/float-fp16.d: Likewise. * testsuite/gas/aarch64/int-insns.d: Likewise. * testsuite/gas/aarch64/no-aliases.d: Likewise. * testsuite/gas/aarch64/programmer-friendly.d: Likewise. * testsuite/gas/aarch64/reloc-insn.d: Likewise. * testsuite/gas/aarch64/b_c_1.d, testsuite/gas/aarch64/b_c_1.s: New test. ld/ * testsuite/ld-aarch64/emit-relocs-280.d: Match branch comments. * testsuite/ld-aarch64/weak-undefined.d: Likewise.
2016-09-22 00:09:59 +08:00
opnd->cond->names[0], opnd->cond->names[i]);
else
opcodes/aarch64: split off creation of comment text in disassembler The function aarch64_print_operand (aarch64-opc.c) is responsible for converting an instruction operand into the textual representation of that operand. In some cases, a comment is included in the operand representation, though this (currently) only happens for the last operand of the instruction. In a future commit I would like to enable the new libopcodes styling for AArch64, this will allow objdump and GDB[1] to syntax highlight the disassembler output, however, having operands and comments combined in a single string like this makes such styling harder. In this commit, I propose to extend aarch64_print_operand to take a second buffer. Any comments for the instruction are written into this extra buffer. The two callers of aarch64_print_operand are then updated to pass an extra buffer, and print any resulting comment. In this commit no styling is added, that will come later. However, I have adjusted the output slightly. Before this commit some comments would be separated from the instruction operands with a tab character, while in other cases the comment was separated with two single spaces. After this commit I use a single tab character in all cases. This means a few test cases needed updated. If people would prefer me to move everyone to use the two spaces, then just let me know. Or maybe there was a good reason why we used a mix of styles, I could probably figure out a way to maintain the old output exactly if that is critical. Other than that, there should be no user visible changes after this commit. [1] GDB patches have not been merged yet, but have been posted to the GDB mailing list: https://sourceware.org/pipermail/gdb-patches/2022-June/190142.html
2022-06-16 20:46:41 +08:00
snprintf (comment + len, comment_size - len, ", %s",
[AArch64] Add SVE condition codes SVE defines new names for existing NZCV conditions, to reflect the result of instructions like PTEST. This patch adds support for these names. The patch also adds comments to the disassembly output to show the alternative names of a condition code. For example: cinv x0, x1, cc becomes: cinv x0, x1, cc // cc = lo, ul, last and: b.cc f0 <...> becomes: b.cc f0 <...> // b.lo, b.ul, b.last Doing this for the SVE names follows the practice recommended by the SVE specification and is definitely useful when reading SVE code. If the feeling is that it's too distracting elsewhere, we could add an option to turn it off. include/ * opcode/aarch64.h (aarch64_cond): Bump array size to 4. opcodes/ * aarch64-dis.c (remove_dot_suffix): New function, split out from... (print_mnemonic_name): ...here. (print_comment): New function. (print_aarch64_insn): Call it. * aarch64-opc.c (aarch64_conds): Add SVE names. (aarch64_print_operand): Print alternative condition names in a comment. gas/ * config/tc-aarch64.c (opcode_lookup): Search for the end of a condition name, rather than assuming that it will have exactly 2 characters. (parse_operands): Likewise. * testsuite/gas/aarch64/alias.d: Add new condition-code comments to the expected output. * testsuite/gas/aarch64/beq_1.d: Likewise. * testsuite/gas/aarch64/float-fp16.d: Likewise. * testsuite/gas/aarch64/int-insns.d: Likewise. * testsuite/gas/aarch64/no-aliases.d: Likewise. * testsuite/gas/aarch64/programmer-friendly.d: Likewise. * testsuite/gas/aarch64/reloc-insn.d: Likewise. * testsuite/gas/aarch64/b_c_1.d, testsuite/gas/aarch64/b_c_1.s: New test. ld/ * testsuite/ld-aarch64/emit-relocs-280.d: Match branch comments. * testsuite/ld-aarch64/weak-undefined.d: Likewise.
2016-09-22 00:09:59 +08:00
opnd->cond->names[i]);
}
break;
case AARCH64_OPND_ADDR_ADRP:
addr = ((pc + AARCH64_PCREL_OFFSET) & ~(uint64_t)0xfff)
+ opnd->imm.value;
if (pcrel_p)
*pcrel_p = 1;
if (address)
*address = addr;
/* This is not necessary during the disassembling, as print_address_func
in the disassemble_info will take care of the printing. But some
other callers may be still interested in getting the string in *STR,
so here we do snprintf regardless. */
libopcodes/aarch64: add support for disassembler styling This commit enables disassembler styling for AArch64. After this commit it is possible to have objdump style AArch64 disassembler output (using --disassembler-color option). Once the required GDB patches are merged, GDB will also style the disassembler output. The changes to support styling are mostly split between two files opcodes/aarch64-dis.c and opcodes/aarch64-opc.c. The entry point for the AArch64 disassembler can be found in aarch64-dis.c, this file handles printing the instruction mnemonics, and assembler directives (e.g. '.byte', '.word', etc). Some operands, mostly relating to assembler directives are also printed from this file. This commit changes all of this to pass through suitable styling information. However, for most "normal" instructions, the instruction operands are printed using a two step process. From aarch64-dis.c, in the print_operands function, the function aarch64_print_operand is called, this function is in aarch64-opc.c, and converts an instruction operand into a string. Then, back in print_operands (aarch64-dis.c), the operand string is printed. Unfortunately, the string returned by aarch64_print_operand can be quite complex, it will include syntax elements, like '[' and ']', in addition to register names and immediate values. In some cases, a single operand will expand into what will appear (to the user) as multiple operands separated with a ','. This makes the task of styling more complex, all these different components need to by styled differently, so we need to get the styling information out of aarch64_print_operand in some way. The solution that I propose here is similar to the solution that I used for the i386 disassembler. Currently, aarch64_print_operand uses snprintf to write the operand text into a buffer provided by the caller. What I propose is that we pass an extra argument to the aarch64_print_operand function, this argument will be a structure, the structure contains a callback function and some state. When aarch64_print_operand needs to format part of its output this can be done by using the callback function within the new structure, this callback returns a string with special embedded markers that indicate which mode should be used for each piece of text. Back in aarch64-dis.c we can spot these special style markers and use this to split the disassembler output up and apply the correct style to each piece. To make aarch64-opc.c clearer a series of new static functions have been added, e.g. 'style_reg', 'style_imm', etc. Each of these functions formats a piece of text in a different style, 'register' and 'immediate' in this case. Here's an example taken from aarch64-opc.c of the new functions in use: snprintf (buf, size, "[%s, %s]!", style_reg (styler, base), style_imm (styler, "#%d", opnd->addr.offset.imm)); The aarch64_print_operand function is also called from the assembler to aid in printing diagnostic messages. Right now I have no plans to add styling to the assembler output, and so, the callback function used in the assembler ignores the styling information and just returns an plain string. I've used the source files in gas/testsuite/gas/aarch64/ for testing, and have manually gone through and checked that the styling looks reasonable, however, I'm not an AArch64 expert, so it is possible that the odd piece is styled incorrectly. Please point out any mistakes I've made. With objdump disassembler color turned off, there should be no change in the output after this commit.
2022-04-28 20:31:07 +08:00
snprintf (buf, size, "%s", style_addr (styler, "#0x%" PRIx64 , addr));
break;
case AARCH64_OPND_ADDR_PCREL14:
case AARCH64_OPND_ADDR_PCREL19:
case AARCH64_OPND_ADDR_PCREL21:
case AARCH64_OPND_ADDR_PCREL26:
addr = pc + AARCH64_PCREL_OFFSET + opnd->imm.value;
if (pcrel_p)
*pcrel_p = 1;
if (address)
*address = addr;
/* This is not necessary during the disassembling, as print_address_func
in the disassemble_info will take care of the printing. But some
other callers may be still interested in getting the string in *STR,
so here we do snprintf regardless. */
libopcodes/aarch64: add support for disassembler styling This commit enables disassembler styling for AArch64. After this commit it is possible to have objdump style AArch64 disassembler output (using --disassembler-color option). Once the required GDB patches are merged, GDB will also style the disassembler output. The changes to support styling are mostly split between two files opcodes/aarch64-dis.c and opcodes/aarch64-opc.c. The entry point for the AArch64 disassembler can be found in aarch64-dis.c, this file handles printing the instruction mnemonics, and assembler directives (e.g. '.byte', '.word', etc). Some operands, mostly relating to assembler directives are also printed from this file. This commit changes all of this to pass through suitable styling information. However, for most "normal" instructions, the instruction operands are printed using a two step process. From aarch64-dis.c, in the print_operands function, the function aarch64_print_operand is called, this function is in aarch64-opc.c, and converts an instruction operand into a string. Then, back in print_operands (aarch64-dis.c), the operand string is printed. Unfortunately, the string returned by aarch64_print_operand can be quite complex, it will include syntax elements, like '[' and ']', in addition to register names and immediate values. In some cases, a single operand will expand into what will appear (to the user) as multiple operands separated with a ','. This makes the task of styling more complex, all these different components need to by styled differently, so we need to get the styling information out of aarch64_print_operand in some way. The solution that I propose here is similar to the solution that I used for the i386 disassembler. Currently, aarch64_print_operand uses snprintf to write the operand text into a buffer provided by the caller. What I propose is that we pass an extra argument to the aarch64_print_operand function, this argument will be a structure, the structure contains a callback function and some state. When aarch64_print_operand needs to format part of its output this can be done by using the callback function within the new structure, this callback returns a string with special embedded markers that indicate which mode should be used for each piece of text. Back in aarch64-dis.c we can spot these special style markers and use this to split the disassembler output up and apply the correct style to each piece. To make aarch64-opc.c clearer a series of new static functions have been added, e.g. 'style_reg', 'style_imm', etc. Each of these functions formats a piece of text in a different style, 'register' and 'immediate' in this case. Here's an example taken from aarch64-opc.c of the new functions in use: snprintf (buf, size, "[%s, %s]!", style_reg (styler, base), style_imm (styler, "#%d", opnd->addr.offset.imm)); The aarch64_print_operand function is also called from the assembler to aid in printing diagnostic messages. Right now I have no plans to add styling to the assembler output, and so, the callback function used in the assembler ignores the styling information and just returns an plain string. I've used the source files in gas/testsuite/gas/aarch64/ for testing, and have manually gone through and checked that the styling looks reasonable, however, I'm not an AArch64 expert, so it is possible that the odd piece is styled incorrectly. Please point out any mistakes I've made. With objdump disassembler color turned off, there should be no change in the output after this commit.
2022-04-28 20:31:07 +08:00
snprintf (buf, size, "%s", style_addr (styler, "#0x%" PRIx64, addr));
break;
case AARCH64_OPND_ADDR_SIMPLE:
case AARCH64_OPND_SIMD_ADDR_SIMPLE:
case AARCH64_OPND_SIMD_ADDR_POST:
name = get_64bit_int_reg_name (opnd->addr.base_regno, 1);
if (opnd->type == AARCH64_OPND_SIMD_ADDR_POST)
{
if (opnd->addr.offset.is_reg)
libopcodes/aarch64: add support for disassembler styling This commit enables disassembler styling for AArch64. After this commit it is possible to have objdump style AArch64 disassembler output (using --disassembler-color option). Once the required GDB patches are merged, GDB will also style the disassembler output. The changes to support styling are mostly split between two files opcodes/aarch64-dis.c and opcodes/aarch64-opc.c. The entry point for the AArch64 disassembler can be found in aarch64-dis.c, this file handles printing the instruction mnemonics, and assembler directives (e.g. '.byte', '.word', etc). Some operands, mostly relating to assembler directives are also printed from this file. This commit changes all of this to pass through suitable styling information. However, for most "normal" instructions, the instruction operands are printed using a two step process. From aarch64-dis.c, in the print_operands function, the function aarch64_print_operand is called, this function is in aarch64-opc.c, and converts an instruction operand into a string. Then, back in print_operands (aarch64-dis.c), the operand string is printed. Unfortunately, the string returned by aarch64_print_operand can be quite complex, it will include syntax elements, like '[' and ']', in addition to register names and immediate values. In some cases, a single operand will expand into what will appear (to the user) as multiple operands separated with a ','. This makes the task of styling more complex, all these different components need to by styled differently, so we need to get the styling information out of aarch64_print_operand in some way. The solution that I propose here is similar to the solution that I used for the i386 disassembler. Currently, aarch64_print_operand uses snprintf to write the operand text into a buffer provided by the caller. What I propose is that we pass an extra argument to the aarch64_print_operand function, this argument will be a structure, the structure contains a callback function and some state. When aarch64_print_operand needs to format part of its output this can be done by using the callback function within the new structure, this callback returns a string with special embedded markers that indicate which mode should be used for each piece of text. Back in aarch64-dis.c we can spot these special style markers and use this to split the disassembler output up and apply the correct style to each piece. To make aarch64-opc.c clearer a series of new static functions have been added, e.g. 'style_reg', 'style_imm', etc. Each of these functions formats a piece of text in a different style, 'register' and 'immediate' in this case. Here's an example taken from aarch64-opc.c of the new functions in use: snprintf (buf, size, "[%s, %s]!", style_reg (styler, base), style_imm (styler, "#%d", opnd->addr.offset.imm)); The aarch64_print_operand function is also called from the assembler to aid in printing diagnostic messages. Right now I have no plans to add styling to the assembler output, and so, the callback function used in the assembler ignores the styling information and just returns an plain string. I've used the source files in gas/testsuite/gas/aarch64/ for testing, and have manually gone through and checked that the styling looks reasonable, however, I'm not an AArch64 expert, so it is possible that the odd piece is styled incorrectly. Please point out any mistakes I've made. With objdump disassembler color turned off, there should be no change in the output after this commit.
2022-04-28 20:31:07 +08:00
snprintf (buf, size, "[%s], %s",
style_reg (styler, name),
style_reg (styler, "x%d", opnd->addr.offset.regno));
else
libopcodes/aarch64: add support for disassembler styling This commit enables disassembler styling for AArch64. After this commit it is possible to have objdump style AArch64 disassembler output (using --disassembler-color option). Once the required GDB patches are merged, GDB will also style the disassembler output. The changes to support styling are mostly split between two files opcodes/aarch64-dis.c and opcodes/aarch64-opc.c. The entry point for the AArch64 disassembler can be found in aarch64-dis.c, this file handles printing the instruction mnemonics, and assembler directives (e.g. '.byte', '.word', etc). Some operands, mostly relating to assembler directives are also printed from this file. This commit changes all of this to pass through suitable styling information. However, for most "normal" instructions, the instruction operands are printed using a two step process. From aarch64-dis.c, in the print_operands function, the function aarch64_print_operand is called, this function is in aarch64-opc.c, and converts an instruction operand into a string. Then, back in print_operands (aarch64-dis.c), the operand string is printed. Unfortunately, the string returned by aarch64_print_operand can be quite complex, it will include syntax elements, like '[' and ']', in addition to register names and immediate values. In some cases, a single operand will expand into what will appear (to the user) as multiple operands separated with a ','. This makes the task of styling more complex, all these different components need to by styled differently, so we need to get the styling information out of aarch64_print_operand in some way. The solution that I propose here is similar to the solution that I used for the i386 disassembler. Currently, aarch64_print_operand uses snprintf to write the operand text into a buffer provided by the caller. What I propose is that we pass an extra argument to the aarch64_print_operand function, this argument will be a structure, the structure contains a callback function and some state. When aarch64_print_operand needs to format part of its output this can be done by using the callback function within the new structure, this callback returns a string with special embedded markers that indicate which mode should be used for each piece of text. Back in aarch64-dis.c we can spot these special style markers and use this to split the disassembler output up and apply the correct style to each piece. To make aarch64-opc.c clearer a series of new static functions have been added, e.g. 'style_reg', 'style_imm', etc. Each of these functions formats a piece of text in a different style, 'register' and 'immediate' in this case. Here's an example taken from aarch64-opc.c of the new functions in use: snprintf (buf, size, "[%s, %s]!", style_reg (styler, base), style_imm (styler, "#%d", opnd->addr.offset.imm)); The aarch64_print_operand function is also called from the assembler to aid in printing diagnostic messages. Right now I have no plans to add styling to the assembler output, and so, the callback function used in the assembler ignores the styling information and just returns an plain string. I've used the source files in gas/testsuite/gas/aarch64/ for testing, and have manually gone through and checked that the styling looks reasonable, however, I'm not an AArch64 expert, so it is possible that the odd piece is styled incorrectly. Please point out any mistakes I've made. With objdump disassembler color turned off, there should be no change in the output after this commit.
2022-04-28 20:31:07 +08:00
snprintf (buf, size, "[%s], %s",
style_reg (styler, name),
style_imm (styler, "#%d", opnd->addr.offset.imm));
}
else
libopcodes/aarch64: add support for disassembler styling This commit enables disassembler styling for AArch64. After this commit it is possible to have objdump style AArch64 disassembler output (using --disassembler-color option). Once the required GDB patches are merged, GDB will also style the disassembler output. The changes to support styling are mostly split between two files opcodes/aarch64-dis.c and opcodes/aarch64-opc.c. The entry point for the AArch64 disassembler can be found in aarch64-dis.c, this file handles printing the instruction mnemonics, and assembler directives (e.g. '.byte', '.word', etc). Some operands, mostly relating to assembler directives are also printed from this file. This commit changes all of this to pass through suitable styling information. However, for most "normal" instructions, the instruction operands are printed using a two step process. From aarch64-dis.c, in the print_operands function, the function aarch64_print_operand is called, this function is in aarch64-opc.c, and converts an instruction operand into a string. Then, back in print_operands (aarch64-dis.c), the operand string is printed. Unfortunately, the string returned by aarch64_print_operand can be quite complex, it will include syntax elements, like '[' and ']', in addition to register names and immediate values. In some cases, a single operand will expand into what will appear (to the user) as multiple operands separated with a ','. This makes the task of styling more complex, all these different components need to by styled differently, so we need to get the styling information out of aarch64_print_operand in some way. The solution that I propose here is similar to the solution that I used for the i386 disassembler. Currently, aarch64_print_operand uses snprintf to write the operand text into a buffer provided by the caller. What I propose is that we pass an extra argument to the aarch64_print_operand function, this argument will be a structure, the structure contains a callback function and some state. When aarch64_print_operand needs to format part of its output this can be done by using the callback function within the new structure, this callback returns a string with special embedded markers that indicate which mode should be used for each piece of text. Back in aarch64-dis.c we can spot these special style markers and use this to split the disassembler output up and apply the correct style to each piece. To make aarch64-opc.c clearer a series of new static functions have been added, e.g. 'style_reg', 'style_imm', etc. Each of these functions formats a piece of text in a different style, 'register' and 'immediate' in this case. Here's an example taken from aarch64-opc.c of the new functions in use: snprintf (buf, size, "[%s, %s]!", style_reg (styler, base), style_imm (styler, "#%d", opnd->addr.offset.imm)); The aarch64_print_operand function is also called from the assembler to aid in printing diagnostic messages. Right now I have no plans to add styling to the assembler output, and so, the callback function used in the assembler ignores the styling information and just returns an plain string. I've used the source files in gas/testsuite/gas/aarch64/ for testing, and have manually gone through and checked that the styling looks reasonable, however, I'm not an AArch64 expert, so it is possible that the odd piece is styled incorrectly. Please point out any mistakes I've made. With objdump disassembler color turned off, there should be no change in the output after this commit.
2022-04-28 20:31:07 +08:00
snprintf (buf, size, "[%s]", style_reg (styler, name));
break;
case AARCH64_OPND_ADDR_REGOFF:
case AARCH64_OPND_SVE_ADDR_R:
[AArch64][SVE 25/32] Add support for SVE addressing modes This patch adds most of the new SVE addressing modes and associated operands. A follow-on patch adds MUL VL, since handling it separately makes the changes easier to read. The patch also introduces a new "operand-dependent data" field to the operand flags, based closely on the existing one for opcode flags. For SVE this new field needs only 2 bits, but it could be widened in future if necessary. include/ * opcode/aarch64.h (AARCH64_OPND_SVE_ADDR_RI_U6): New aarch64_opnd. (AARCH64_OPND_SVE_ADDR_RI_U6x2, AARCH64_OPND_SVE_ADDR_RI_U6x4) (AARCH64_OPND_SVE_ADDR_RI_U6x8, AARCH64_OPND_SVE_ADDR_RR) (AARCH64_OPND_SVE_ADDR_RR_LSL1, AARCH64_OPND_SVE_ADDR_RR_LSL2) (AARCH64_OPND_SVE_ADDR_RR_LSL3, AARCH64_OPND_SVE_ADDR_RX) (AARCH64_OPND_SVE_ADDR_RX_LSL1, AARCH64_OPND_SVE_ADDR_RX_LSL2) (AARCH64_OPND_SVE_ADDR_RX_LSL3, AARCH64_OPND_SVE_ADDR_RZ) (AARCH64_OPND_SVE_ADDR_RZ_LSL1, AARCH64_OPND_SVE_ADDR_RZ_LSL2) (AARCH64_OPND_SVE_ADDR_RZ_LSL3, AARCH64_OPND_SVE_ADDR_RZ_XTW_14) (AARCH64_OPND_SVE_ADDR_RZ_XTW_22, AARCH64_OPND_SVE_ADDR_RZ_XTW1_14) (AARCH64_OPND_SVE_ADDR_RZ_XTW1_22, AARCH64_OPND_SVE_ADDR_RZ_XTW2_14) (AARCH64_OPND_SVE_ADDR_RZ_XTW2_22, AARCH64_OPND_SVE_ADDR_RZ_XTW3_14) (AARCH64_OPND_SVE_ADDR_RZ_XTW3_22, AARCH64_OPND_SVE_ADDR_ZI_U5) (AARCH64_OPND_SVE_ADDR_ZI_U5x2, AARCH64_OPND_SVE_ADDR_ZI_U5x4) (AARCH64_OPND_SVE_ADDR_ZI_U5x8, AARCH64_OPND_SVE_ADDR_ZZ_LSL) (AARCH64_OPND_SVE_ADDR_ZZ_SXTW, AARCH64_OPND_SVE_ADDR_ZZ_UXTW): Likewise. opcodes/ * aarch64-tbl.h (AARCH64_OPERANDS): Add entries for the new SVE address operands. * aarch64-opc.h (FLD_SVE_imm6, FLD_SVE_msz, FLD_SVE_xs_14) (FLD_SVE_xs_22): New aarch64_field_kinds. (OPD_F_OD_MASK, OPD_F_OD_LSB, OPD_F_NO_ZR): New flags. (get_operand_specific_data): New function. * aarch64-opc.c (fields): Add entries for FLD_SVE_imm6, FLD_SVE_msz, FLD_SVE_xs_14 and FLD_SVE_xs_22. (operand_general_constraint_met_p): Handle the new SVE address operands. (sve_reg): New array. (get_addr_sve_reg_name): New function. (aarch64_print_operand): Handle the new SVE address operands. * aarch64-opc-2.c: Regenerate. * aarch64-asm.h (ins_sve_addr_ri_u6, ins_sve_addr_rr_lsl) (ins_sve_addr_rz_xtw, ins_sve_addr_zi_u5, ins_sve_addr_zz_lsl) (ins_sve_addr_zz_sxtw, ins_sve_addr_zz_uxtw): New inserters. * aarch64-asm.c (aarch64_ins_sve_addr_ri_u6): New function. (aarch64_ins_sve_addr_rr_lsl): Likewise. (aarch64_ins_sve_addr_rz_xtw): Likewise. (aarch64_ins_sve_addr_zi_u5): Likewise. (aarch64_ins_sve_addr_zz): Likewise. (aarch64_ins_sve_addr_zz_lsl): Likewise. (aarch64_ins_sve_addr_zz_sxtw): Likewise. (aarch64_ins_sve_addr_zz_uxtw): Likewise. * aarch64-asm-2.c: Regenerate. * aarch64-dis.h (ext_sve_addr_ri_u6, ext_sve_addr_rr_lsl) (ext_sve_addr_rz_xtw, ext_sve_addr_zi_u5, ext_sve_addr_zz_lsl) (ext_sve_addr_zz_sxtw, ext_sve_addr_zz_uxtw): New extractors. * aarch64-dis.c (aarch64_ext_sve_add_reg_imm): New function. (aarch64_ext_sve_addr_ri_u6): Likewise. (aarch64_ext_sve_addr_rr_lsl): Likewise. (aarch64_ext_sve_addr_rz_xtw): Likewise. (aarch64_ext_sve_addr_zi_u5): Likewise. (aarch64_ext_sve_addr_zz): Likewise. (aarch64_ext_sve_addr_zz_lsl): Likewise. (aarch64_ext_sve_addr_zz_sxtw): Likewise. (aarch64_ext_sve_addr_zz_uxtw): Likewise. * aarch64-dis-2.c: Regenerate. gas/ * config/tc-aarch64.c (REG_TYPE_SVE_BASE, REG_TYPE_SVE_OFFSET): New register types. (get_reg_expected_msg): Handle them. (aarch64_addr_reg_parse): New function, split out from aarch64_reg_parse_32_64. Handle Z registers too. (aarch64_reg_parse_32_64): Call it. (parse_address_main): Add base_qualifier, offset_qualifier, base_type and offset_type parameters. Handle SVE base and offset registers. (parse_address): Update call to parse_address_main. (parse_sve_address): New function. (parse_operands): Parse the new SVE address operands.
2016-09-21 23:55:49 +08:00
case AARCH64_OPND_SVE_ADDR_RR:
case AARCH64_OPND_SVE_ADDR_RR_LSL1:
case AARCH64_OPND_SVE_ADDR_RR_LSL2:
case AARCH64_OPND_SVE_ADDR_RR_LSL3:
aarch64: [SME] Add LD1x, ST1x, LDR and STR instructions This patch is adding new loads and stores defined by SME instructions. gas/ChangeLog: * config/tc-aarch64.c (parse_sme_address): New parser. (parse_sme_za_hv_tiles_operand_with_braces): New parser. (parse_sme_za_array): New parser. (output_operand_error_record): Print error details if present. (parse_operands): Support new operands. * testsuite/gas/aarch64/sme-5-illegal.d: New test. * testsuite/gas/aarch64/sme-5-illegal.l: New test. * testsuite/gas/aarch64/sme-5-illegal.s: New test. * testsuite/gas/aarch64/sme-5.d: New test. * testsuite/gas/aarch64/sme-5.s: New test. * testsuite/gas/aarch64/sme-6-illegal.d: New test. * testsuite/gas/aarch64/sme-6-illegal.l: New test. * testsuite/gas/aarch64/sme-6-illegal.s: New test. * testsuite/gas/aarch64/sme-6.d: New test. * testsuite/gas/aarch64/sme-6.s: New test. * testsuite/gas/aarch64/sme-7-illegal.d: New test. * testsuite/gas/aarch64/sme-7-illegal.l: New test. * testsuite/gas/aarch64/sme-7-illegal.s: New test. * testsuite/gas/aarch64/sme-7.d: New test. * testsuite/gas/aarch64/sme-7.s: New test. include/ChangeLog: * opcode/aarch64.h (enum aarch64_opnd): New operands. (enum aarch64_insn_class): Added sme_ldr and sme_str. (AARCH64_OPDE_UNTIED_IMMS): New operand error kind. opcodes/ChangeLog: * aarch64-asm.c (aarch64_ins_sme_za_hv_tiles): New inserter. (aarch64_ins_sme_za_list): New inserter. (aarch64_ins_sme_za_array): New inserter. (aarch64_ins_sme_addr_ri_u4xvl): New inserter. * aarch64-asm.h (AARCH64_DECL_OPD_INSERTER): Added ins_sme_za_list, ins_sme_za_array and ins_sme_addr_ri_u4xvl. * aarch64-dis.c (aarch64_ext_sme_za_hv_tiles): New extractor. (aarch64_ext_sme_za_list): New extractor. (aarch64_ext_sme_za_array): New extractor. (aarch64_ext_sme_addr_ri_u4xvl): New extractor. * aarch64-dis.h (AARCH64_DECL_OPD_EXTRACTOR): Added ext_sme_za_list, ext_sme_za_array and ext_sme_addr_ri_u4xvl. * aarch64-opc.c (operand_general_constraint_met_p): (aarch64_match_operands_constraint): Handle sme_ldr, sme_str and sme_misc. (aarch64_print_operand): New operands supported. * aarch64-tbl.h (OP_SVE_QUU): New qualifier. (OP_SVE_QZU): New qualifier. aarch64-asm-2.c: Regenerate. aarch64-dis-2.c: Regenerate. aarch64-opc-2.c: Regenerate.
2021-11-18 04:02:06 +08:00
case AARCH64_OPND_SVE_ADDR_RR_LSL4:
[AArch64][SVE 25/32] Add support for SVE addressing modes This patch adds most of the new SVE addressing modes and associated operands. A follow-on patch adds MUL VL, since handling it separately makes the changes easier to read. The patch also introduces a new "operand-dependent data" field to the operand flags, based closely on the existing one for opcode flags. For SVE this new field needs only 2 bits, but it could be widened in future if necessary. include/ * opcode/aarch64.h (AARCH64_OPND_SVE_ADDR_RI_U6): New aarch64_opnd. (AARCH64_OPND_SVE_ADDR_RI_U6x2, AARCH64_OPND_SVE_ADDR_RI_U6x4) (AARCH64_OPND_SVE_ADDR_RI_U6x8, AARCH64_OPND_SVE_ADDR_RR) (AARCH64_OPND_SVE_ADDR_RR_LSL1, AARCH64_OPND_SVE_ADDR_RR_LSL2) (AARCH64_OPND_SVE_ADDR_RR_LSL3, AARCH64_OPND_SVE_ADDR_RX) (AARCH64_OPND_SVE_ADDR_RX_LSL1, AARCH64_OPND_SVE_ADDR_RX_LSL2) (AARCH64_OPND_SVE_ADDR_RX_LSL3, AARCH64_OPND_SVE_ADDR_RZ) (AARCH64_OPND_SVE_ADDR_RZ_LSL1, AARCH64_OPND_SVE_ADDR_RZ_LSL2) (AARCH64_OPND_SVE_ADDR_RZ_LSL3, AARCH64_OPND_SVE_ADDR_RZ_XTW_14) (AARCH64_OPND_SVE_ADDR_RZ_XTW_22, AARCH64_OPND_SVE_ADDR_RZ_XTW1_14) (AARCH64_OPND_SVE_ADDR_RZ_XTW1_22, AARCH64_OPND_SVE_ADDR_RZ_XTW2_14) (AARCH64_OPND_SVE_ADDR_RZ_XTW2_22, AARCH64_OPND_SVE_ADDR_RZ_XTW3_14) (AARCH64_OPND_SVE_ADDR_RZ_XTW3_22, AARCH64_OPND_SVE_ADDR_ZI_U5) (AARCH64_OPND_SVE_ADDR_ZI_U5x2, AARCH64_OPND_SVE_ADDR_ZI_U5x4) (AARCH64_OPND_SVE_ADDR_ZI_U5x8, AARCH64_OPND_SVE_ADDR_ZZ_LSL) (AARCH64_OPND_SVE_ADDR_ZZ_SXTW, AARCH64_OPND_SVE_ADDR_ZZ_UXTW): Likewise. opcodes/ * aarch64-tbl.h (AARCH64_OPERANDS): Add entries for the new SVE address operands. * aarch64-opc.h (FLD_SVE_imm6, FLD_SVE_msz, FLD_SVE_xs_14) (FLD_SVE_xs_22): New aarch64_field_kinds. (OPD_F_OD_MASK, OPD_F_OD_LSB, OPD_F_NO_ZR): New flags. (get_operand_specific_data): New function. * aarch64-opc.c (fields): Add entries for FLD_SVE_imm6, FLD_SVE_msz, FLD_SVE_xs_14 and FLD_SVE_xs_22. (operand_general_constraint_met_p): Handle the new SVE address operands. (sve_reg): New array. (get_addr_sve_reg_name): New function. (aarch64_print_operand): Handle the new SVE address operands. * aarch64-opc-2.c: Regenerate. * aarch64-asm.h (ins_sve_addr_ri_u6, ins_sve_addr_rr_lsl) (ins_sve_addr_rz_xtw, ins_sve_addr_zi_u5, ins_sve_addr_zz_lsl) (ins_sve_addr_zz_sxtw, ins_sve_addr_zz_uxtw): New inserters. * aarch64-asm.c (aarch64_ins_sve_addr_ri_u6): New function. (aarch64_ins_sve_addr_rr_lsl): Likewise. (aarch64_ins_sve_addr_rz_xtw): Likewise. (aarch64_ins_sve_addr_zi_u5): Likewise. (aarch64_ins_sve_addr_zz): Likewise. (aarch64_ins_sve_addr_zz_lsl): Likewise. (aarch64_ins_sve_addr_zz_sxtw): Likewise. (aarch64_ins_sve_addr_zz_uxtw): Likewise. * aarch64-asm-2.c: Regenerate. * aarch64-dis.h (ext_sve_addr_ri_u6, ext_sve_addr_rr_lsl) (ext_sve_addr_rz_xtw, ext_sve_addr_zi_u5, ext_sve_addr_zz_lsl) (ext_sve_addr_zz_sxtw, ext_sve_addr_zz_uxtw): New extractors. * aarch64-dis.c (aarch64_ext_sve_add_reg_imm): New function. (aarch64_ext_sve_addr_ri_u6): Likewise. (aarch64_ext_sve_addr_rr_lsl): Likewise. (aarch64_ext_sve_addr_rz_xtw): Likewise. (aarch64_ext_sve_addr_zi_u5): Likewise. (aarch64_ext_sve_addr_zz): Likewise. (aarch64_ext_sve_addr_zz_lsl): Likewise. (aarch64_ext_sve_addr_zz_sxtw): Likewise. (aarch64_ext_sve_addr_zz_uxtw): Likewise. * aarch64-dis-2.c: Regenerate. gas/ * config/tc-aarch64.c (REG_TYPE_SVE_BASE, REG_TYPE_SVE_OFFSET): New register types. (get_reg_expected_msg): Handle them. (aarch64_addr_reg_parse): New function, split out from aarch64_reg_parse_32_64. Handle Z registers too. (aarch64_reg_parse_32_64): Call it. (parse_address_main): Add base_qualifier, offset_qualifier, base_type and offset_type parameters. Handle SVE base and offset registers. (parse_address): Update call to parse_address_main. (parse_sve_address): New function. (parse_operands): Parse the new SVE address operands.
2016-09-21 23:55:49 +08:00
case AARCH64_OPND_SVE_ADDR_RX:
case AARCH64_OPND_SVE_ADDR_RX_LSL1:
case AARCH64_OPND_SVE_ADDR_RX_LSL2:
case AARCH64_OPND_SVE_ADDR_RX_LSL3:
print_register_offset_address
(buf, size, opnd, get_64bit_int_reg_name (opnd->addr.base_regno, 1),
libopcodes/aarch64: add support for disassembler styling This commit enables disassembler styling for AArch64. After this commit it is possible to have objdump style AArch64 disassembler output (using --disassembler-color option). Once the required GDB patches are merged, GDB will also style the disassembler output. The changes to support styling are mostly split between two files opcodes/aarch64-dis.c and opcodes/aarch64-opc.c. The entry point for the AArch64 disassembler can be found in aarch64-dis.c, this file handles printing the instruction mnemonics, and assembler directives (e.g. '.byte', '.word', etc). Some operands, mostly relating to assembler directives are also printed from this file. This commit changes all of this to pass through suitable styling information. However, for most "normal" instructions, the instruction operands are printed using a two step process. From aarch64-dis.c, in the print_operands function, the function aarch64_print_operand is called, this function is in aarch64-opc.c, and converts an instruction operand into a string. Then, back in print_operands (aarch64-dis.c), the operand string is printed. Unfortunately, the string returned by aarch64_print_operand can be quite complex, it will include syntax elements, like '[' and ']', in addition to register names and immediate values. In some cases, a single operand will expand into what will appear (to the user) as multiple operands separated with a ','. This makes the task of styling more complex, all these different components need to by styled differently, so we need to get the styling information out of aarch64_print_operand in some way. The solution that I propose here is similar to the solution that I used for the i386 disassembler. Currently, aarch64_print_operand uses snprintf to write the operand text into a buffer provided by the caller. What I propose is that we pass an extra argument to the aarch64_print_operand function, this argument will be a structure, the structure contains a callback function and some state. When aarch64_print_operand needs to format part of its output this can be done by using the callback function within the new structure, this callback returns a string with special embedded markers that indicate which mode should be used for each piece of text. Back in aarch64-dis.c we can spot these special style markers and use this to split the disassembler output up and apply the correct style to each piece. To make aarch64-opc.c clearer a series of new static functions have been added, e.g. 'style_reg', 'style_imm', etc. Each of these functions formats a piece of text in a different style, 'register' and 'immediate' in this case. Here's an example taken from aarch64-opc.c of the new functions in use: snprintf (buf, size, "[%s, %s]!", style_reg (styler, base), style_imm (styler, "#%d", opnd->addr.offset.imm)); The aarch64_print_operand function is also called from the assembler to aid in printing diagnostic messages. Right now I have no plans to add styling to the assembler output, and so, the callback function used in the assembler ignores the styling information and just returns an plain string. I've used the source files in gas/testsuite/gas/aarch64/ for testing, and have manually gone through and checked that the styling looks reasonable, however, I'm not an AArch64 expert, so it is possible that the odd piece is styled incorrectly. Please point out any mistakes I've made. With objdump disassembler color turned off, there should be no change in the output after this commit.
2022-04-28 20:31:07 +08:00
get_offset_int_reg_name (opnd), styler);
break;
case AARCH64_OPND_SVE_ADDR_ZX:
print_register_offset_address
(buf, size, opnd,
get_addr_sve_reg_name (opnd->addr.base_regno, opnd->qualifier),
libopcodes/aarch64: add support for disassembler styling This commit enables disassembler styling for AArch64. After this commit it is possible to have objdump style AArch64 disassembler output (using --disassembler-color option). Once the required GDB patches are merged, GDB will also style the disassembler output. The changes to support styling are mostly split between two files opcodes/aarch64-dis.c and opcodes/aarch64-opc.c. The entry point for the AArch64 disassembler can be found in aarch64-dis.c, this file handles printing the instruction mnemonics, and assembler directives (e.g. '.byte', '.word', etc). Some operands, mostly relating to assembler directives are also printed from this file. This commit changes all of this to pass through suitable styling information. However, for most "normal" instructions, the instruction operands are printed using a two step process. From aarch64-dis.c, in the print_operands function, the function aarch64_print_operand is called, this function is in aarch64-opc.c, and converts an instruction operand into a string. Then, back in print_operands (aarch64-dis.c), the operand string is printed. Unfortunately, the string returned by aarch64_print_operand can be quite complex, it will include syntax elements, like '[' and ']', in addition to register names and immediate values. In some cases, a single operand will expand into what will appear (to the user) as multiple operands separated with a ','. This makes the task of styling more complex, all these different components need to by styled differently, so we need to get the styling information out of aarch64_print_operand in some way. The solution that I propose here is similar to the solution that I used for the i386 disassembler. Currently, aarch64_print_operand uses snprintf to write the operand text into a buffer provided by the caller. What I propose is that we pass an extra argument to the aarch64_print_operand function, this argument will be a structure, the structure contains a callback function and some state. When aarch64_print_operand needs to format part of its output this can be done by using the callback function within the new structure, this callback returns a string with special embedded markers that indicate which mode should be used for each piece of text. Back in aarch64-dis.c we can spot these special style markers and use this to split the disassembler output up and apply the correct style to each piece. To make aarch64-opc.c clearer a series of new static functions have been added, e.g. 'style_reg', 'style_imm', etc. Each of these functions formats a piece of text in a different style, 'register' and 'immediate' in this case. Here's an example taken from aarch64-opc.c of the new functions in use: snprintf (buf, size, "[%s, %s]!", style_reg (styler, base), style_imm (styler, "#%d", opnd->addr.offset.imm)); The aarch64_print_operand function is also called from the assembler to aid in printing diagnostic messages. Right now I have no plans to add styling to the assembler output, and so, the callback function used in the assembler ignores the styling information and just returns an plain string. I've used the source files in gas/testsuite/gas/aarch64/ for testing, and have manually gone through and checked that the styling looks reasonable, however, I'm not an AArch64 expert, so it is possible that the odd piece is styled incorrectly. Please point out any mistakes I've made. With objdump disassembler color turned off, there should be no change in the output after this commit.
2022-04-28 20:31:07 +08:00
get_64bit_int_reg_name (opnd->addr.offset.regno, 0), styler);
break;
[AArch64][SVE 25/32] Add support for SVE addressing modes This patch adds most of the new SVE addressing modes and associated operands. A follow-on patch adds MUL VL, since handling it separately makes the changes easier to read. The patch also introduces a new "operand-dependent data" field to the operand flags, based closely on the existing one for opcode flags. For SVE this new field needs only 2 bits, but it could be widened in future if necessary. include/ * opcode/aarch64.h (AARCH64_OPND_SVE_ADDR_RI_U6): New aarch64_opnd. (AARCH64_OPND_SVE_ADDR_RI_U6x2, AARCH64_OPND_SVE_ADDR_RI_U6x4) (AARCH64_OPND_SVE_ADDR_RI_U6x8, AARCH64_OPND_SVE_ADDR_RR) (AARCH64_OPND_SVE_ADDR_RR_LSL1, AARCH64_OPND_SVE_ADDR_RR_LSL2) (AARCH64_OPND_SVE_ADDR_RR_LSL3, AARCH64_OPND_SVE_ADDR_RX) (AARCH64_OPND_SVE_ADDR_RX_LSL1, AARCH64_OPND_SVE_ADDR_RX_LSL2) (AARCH64_OPND_SVE_ADDR_RX_LSL3, AARCH64_OPND_SVE_ADDR_RZ) (AARCH64_OPND_SVE_ADDR_RZ_LSL1, AARCH64_OPND_SVE_ADDR_RZ_LSL2) (AARCH64_OPND_SVE_ADDR_RZ_LSL3, AARCH64_OPND_SVE_ADDR_RZ_XTW_14) (AARCH64_OPND_SVE_ADDR_RZ_XTW_22, AARCH64_OPND_SVE_ADDR_RZ_XTW1_14) (AARCH64_OPND_SVE_ADDR_RZ_XTW1_22, AARCH64_OPND_SVE_ADDR_RZ_XTW2_14) (AARCH64_OPND_SVE_ADDR_RZ_XTW2_22, AARCH64_OPND_SVE_ADDR_RZ_XTW3_14) (AARCH64_OPND_SVE_ADDR_RZ_XTW3_22, AARCH64_OPND_SVE_ADDR_ZI_U5) (AARCH64_OPND_SVE_ADDR_ZI_U5x2, AARCH64_OPND_SVE_ADDR_ZI_U5x4) (AARCH64_OPND_SVE_ADDR_ZI_U5x8, AARCH64_OPND_SVE_ADDR_ZZ_LSL) (AARCH64_OPND_SVE_ADDR_ZZ_SXTW, AARCH64_OPND_SVE_ADDR_ZZ_UXTW): Likewise. opcodes/ * aarch64-tbl.h (AARCH64_OPERANDS): Add entries for the new SVE address operands. * aarch64-opc.h (FLD_SVE_imm6, FLD_SVE_msz, FLD_SVE_xs_14) (FLD_SVE_xs_22): New aarch64_field_kinds. (OPD_F_OD_MASK, OPD_F_OD_LSB, OPD_F_NO_ZR): New flags. (get_operand_specific_data): New function. * aarch64-opc.c (fields): Add entries for FLD_SVE_imm6, FLD_SVE_msz, FLD_SVE_xs_14 and FLD_SVE_xs_22. (operand_general_constraint_met_p): Handle the new SVE address operands. (sve_reg): New array. (get_addr_sve_reg_name): New function. (aarch64_print_operand): Handle the new SVE address operands. * aarch64-opc-2.c: Regenerate. * aarch64-asm.h (ins_sve_addr_ri_u6, ins_sve_addr_rr_lsl) (ins_sve_addr_rz_xtw, ins_sve_addr_zi_u5, ins_sve_addr_zz_lsl) (ins_sve_addr_zz_sxtw, ins_sve_addr_zz_uxtw): New inserters. * aarch64-asm.c (aarch64_ins_sve_addr_ri_u6): New function. (aarch64_ins_sve_addr_rr_lsl): Likewise. (aarch64_ins_sve_addr_rz_xtw): Likewise. (aarch64_ins_sve_addr_zi_u5): Likewise. (aarch64_ins_sve_addr_zz): Likewise. (aarch64_ins_sve_addr_zz_lsl): Likewise. (aarch64_ins_sve_addr_zz_sxtw): Likewise. (aarch64_ins_sve_addr_zz_uxtw): Likewise. * aarch64-asm-2.c: Regenerate. * aarch64-dis.h (ext_sve_addr_ri_u6, ext_sve_addr_rr_lsl) (ext_sve_addr_rz_xtw, ext_sve_addr_zi_u5, ext_sve_addr_zz_lsl) (ext_sve_addr_zz_sxtw, ext_sve_addr_zz_uxtw): New extractors. * aarch64-dis.c (aarch64_ext_sve_add_reg_imm): New function. (aarch64_ext_sve_addr_ri_u6): Likewise. (aarch64_ext_sve_addr_rr_lsl): Likewise. (aarch64_ext_sve_addr_rz_xtw): Likewise. (aarch64_ext_sve_addr_zi_u5): Likewise. (aarch64_ext_sve_addr_zz): Likewise. (aarch64_ext_sve_addr_zz_lsl): Likewise. (aarch64_ext_sve_addr_zz_sxtw): Likewise. (aarch64_ext_sve_addr_zz_uxtw): Likewise. * aarch64-dis-2.c: Regenerate. gas/ * config/tc-aarch64.c (REG_TYPE_SVE_BASE, REG_TYPE_SVE_OFFSET): New register types. (get_reg_expected_msg): Handle them. (aarch64_addr_reg_parse): New function, split out from aarch64_reg_parse_32_64. Handle Z registers too. (aarch64_reg_parse_32_64): Call it. (parse_address_main): Add base_qualifier, offset_qualifier, base_type and offset_type parameters. Handle SVE base and offset registers. (parse_address): Update call to parse_address_main. (parse_sve_address): New function. (parse_operands): Parse the new SVE address operands.
2016-09-21 23:55:49 +08:00
case AARCH64_OPND_SVE_ADDR_RZ:
case AARCH64_OPND_SVE_ADDR_RZ_LSL1:
case AARCH64_OPND_SVE_ADDR_RZ_LSL2:
case AARCH64_OPND_SVE_ADDR_RZ_LSL3:
case AARCH64_OPND_SVE_ADDR_RZ_XTW_14:
case AARCH64_OPND_SVE_ADDR_RZ_XTW_22:
case AARCH64_OPND_SVE_ADDR_RZ_XTW1_14:
case AARCH64_OPND_SVE_ADDR_RZ_XTW1_22:
case AARCH64_OPND_SVE_ADDR_RZ_XTW2_14:
case AARCH64_OPND_SVE_ADDR_RZ_XTW2_22:
case AARCH64_OPND_SVE_ADDR_RZ_XTW3_14:
case AARCH64_OPND_SVE_ADDR_RZ_XTW3_22:
print_register_offset_address
(buf, size, opnd, get_64bit_int_reg_name (opnd->addr.base_regno, 1),
libopcodes/aarch64: add support for disassembler styling This commit enables disassembler styling for AArch64. After this commit it is possible to have objdump style AArch64 disassembler output (using --disassembler-color option). Once the required GDB patches are merged, GDB will also style the disassembler output. The changes to support styling are mostly split between two files opcodes/aarch64-dis.c and opcodes/aarch64-opc.c. The entry point for the AArch64 disassembler can be found in aarch64-dis.c, this file handles printing the instruction mnemonics, and assembler directives (e.g. '.byte', '.word', etc). Some operands, mostly relating to assembler directives are also printed from this file. This commit changes all of this to pass through suitable styling information. However, for most "normal" instructions, the instruction operands are printed using a two step process. From aarch64-dis.c, in the print_operands function, the function aarch64_print_operand is called, this function is in aarch64-opc.c, and converts an instruction operand into a string. Then, back in print_operands (aarch64-dis.c), the operand string is printed. Unfortunately, the string returned by aarch64_print_operand can be quite complex, it will include syntax elements, like '[' and ']', in addition to register names and immediate values. In some cases, a single operand will expand into what will appear (to the user) as multiple operands separated with a ','. This makes the task of styling more complex, all these different components need to by styled differently, so we need to get the styling information out of aarch64_print_operand in some way. The solution that I propose here is similar to the solution that I used for the i386 disassembler. Currently, aarch64_print_operand uses snprintf to write the operand text into a buffer provided by the caller. What I propose is that we pass an extra argument to the aarch64_print_operand function, this argument will be a structure, the structure contains a callback function and some state. When aarch64_print_operand needs to format part of its output this can be done by using the callback function within the new structure, this callback returns a string with special embedded markers that indicate which mode should be used for each piece of text. Back in aarch64-dis.c we can spot these special style markers and use this to split the disassembler output up and apply the correct style to each piece. To make aarch64-opc.c clearer a series of new static functions have been added, e.g. 'style_reg', 'style_imm', etc. Each of these functions formats a piece of text in a different style, 'register' and 'immediate' in this case. Here's an example taken from aarch64-opc.c of the new functions in use: snprintf (buf, size, "[%s, %s]!", style_reg (styler, base), style_imm (styler, "#%d", opnd->addr.offset.imm)); The aarch64_print_operand function is also called from the assembler to aid in printing diagnostic messages. Right now I have no plans to add styling to the assembler output, and so, the callback function used in the assembler ignores the styling information and just returns an plain string. I've used the source files in gas/testsuite/gas/aarch64/ for testing, and have manually gone through and checked that the styling looks reasonable, however, I'm not an AArch64 expert, so it is possible that the odd piece is styled incorrectly. Please point out any mistakes I've made. With objdump disassembler color turned off, there should be no change in the output after this commit.
2022-04-28 20:31:07 +08:00
get_addr_sve_reg_name (opnd->addr.offset.regno, opnd->qualifier),
styler);
[AArch64][SVE 25/32] Add support for SVE addressing modes This patch adds most of the new SVE addressing modes and associated operands. A follow-on patch adds MUL VL, since handling it separately makes the changes easier to read. The patch also introduces a new "operand-dependent data" field to the operand flags, based closely on the existing one for opcode flags. For SVE this new field needs only 2 bits, but it could be widened in future if necessary. include/ * opcode/aarch64.h (AARCH64_OPND_SVE_ADDR_RI_U6): New aarch64_opnd. (AARCH64_OPND_SVE_ADDR_RI_U6x2, AARCH64_OPND_SVE_ADDR_RI_U6x4) (AARCH64_OPND_SVE_ADDR_RI_U6x8, AARCH64_OPND_SVE_ADDR_RR) (AARCH64_OPND_SVE_ADDR_RR_LSL1, AARCH64_OPND_SVE_ADDR_RR_LSL2) (AARCH64_OPND_SVE_ADDR_RR_LSL3, AARCH64_OPND_SVE_ADDR_RX) (AARCH64_OPND_SVE_ADDR_RX_LSL1, AARCH64_OPND_SVE_ADDR_RX_LSL2) (AARCH64_OPND_SVE_ADDR_RX_LSL3, AARCH64_OPND_SVE_ADDR_RZ) (AARCH64_OPND_SVE_ADDR_RZ_LSL1, AARCH64_OPND_SVE_ADDR_RZ_LSL2) (AARCH64_OPND_SVE_ADDR_RZ_LSL3, AARCH64_OPND_SVE_ADDR_RZ_XTW_14) (AARCH64_OPND_SVE_ADDR_RZ_XTW_22, AARCH64_OPND_SVE_ADDR_RZ_XTW1_14) (AARCH64_OPND_SVE_ADDR_RZ_XTW1_22, AARCH64_OPND_SVE_ADDR_RZ_XTW2_14) (AARCH64_OPND_SVE_ADDR_RZ_XTW2_22, AARCH64_OPND_SVE_ADDR_RZ_XTW3_14) (AARCH64_OPND_SVE_ADDR_RZ_XTW3_22, AARCH64_OPND_SVE_ADDR_ZI_U5) (AARCH64_OPND_SVE_ADDR_ZI_U5x2, AARCH64_OPND_SVE_ADDR_ZI_U5x4) (AARCH64_OPND_SVE_ADDR_ZI_U5x8, AARCH64_OPND_SVE_ADDR_ZZ_LSL) (AARCH64_OPND_SVE_ADDR_ZZ_SXTW, AARCH64_OPND_SVE_ADDR_ZZ_UXTW): Likewise. opcodes/ * aarch64-tbl.h (AARCH64_OPERANDS): Add entries for the new SVE address operands. * aarch64-opc.h (FLD_SVE_imm6, FLD_SVE_msz, FLD_SVE_xs_14) (FLD_SVE_xs_22): New aarch64_field_kinds. (OPD_F_OD_MASK, OPD_F_OD_LSB, OPD_F_NO_ZR): New flags. (get_operand_specific_data): New function. * aarch64-opc.c (fields): Add entries for FLD_SVE_imm6, FLD_SVE_msz, FLD_SVE_xs_14 and FLD_SVE_xs_22. (operand_general_constraint_met_p): Handle the new SVE address operands. (sve_reg): New array. (get_addr_sve_reg_name): New function. (aarch64_print_operand): Handle the new SVE address operands. * aarch64-opc-2.c: Regenerate. * aarch64-asm.h (ins_sve_addr_ri_u6, ins_sve_addr_rr_lsl) (ins_sve_addr_rz_xtw, ins_sve_addr_zi_u5, ins_sve_addr_zz_lsl) (ins_sve_addr_zz_sxtw, ins_sve_addr_zz_uxtw): New inserters. * aarch64-asm.c (aarch64_ins_sve_addr_ri_u6): New function. (aarch64_ins_sve_addr_rr_lsl): Likewise. (aarch64_ins_sve_addr_rz_xtw): Likewise. (aarch64_ins_sve_addr_zi_u5): Likewise. (aarch64_ins_sve_addr_zz): Likewise. (aarch64_ins_sve_addr_zz_lsl): Likewise. (aarch64_ins_sve_addr_zz_sxtw): Likewise. (aarch64_ins_sve_addr_zz_uxtw): Likewise. * aarch64-asm-2.c: Regenerate. * aarch64-dis.h (ext_sve_addr_ri_u6, ext_sve_addr_rr_lsl) (ext_sve_addr_rz_xtw, ext_sve_addr_zi_u5, ext_sve_addr_zz_lsl) (ext_sve_addr_zz_sxtw, ext_sve_addr_zz_uxtw): New extractors. * aarch64-dis.c (aarch64_ext_sve_add_reg_imm): New function. (aarch64_ext_sve_addr_ri_u6): Likewise. (aarch64_ext_sve_addr_rr_lsl): Likewise. (aarch64_ext_sve_addr_rz_xtw): Likewise. (aarch64_ext_sve_addr_zi_u5): Likewise. (aarch64_ext_sve_addr_zz): Likewise. (aarch64_ext_sve_addr_zz_lsl): Likewise. (aarch64_ext_sve_addr_zz_sxtw): Likewise. (aarch64_ext_sve_addr_zz_uxtw): Likewise. * aarch64-dis-2.c: Regenerate. gas/ * config/tc-aarch64.c (REG_TYPE_SVE_BASE, REG_TYPE_SVE_OFFSET): New register types. (get_reg_expected_msg): Handle them. (aarch64_addr_reg_parse): New function, split out from aarch64_reg_parse_32_64. Handle Z registers too. (aarch64_reg_parse_32_64): Call it. (parse_address_main): Add base_qualifier, offset_qualifier, base_type and offset_type parameters. Handle SVE base and offset registers. (parse_address): Update call to parse_address_main. (parse_sve_address): New function. (parse_operands): Parse the new SVE address operands.
2016-09-21 23:55:49 +08:00
break;
case AARCH64_OPND_ADDR_SIMM7:
case AARCH64_OPND_ADDR_SIMM9:
case AARCH64_OPND_ADDR_SIMM9_2:
[AArch64] Add ARMv8.3 combined pointer authentication load instructions Add support for ARMv8.3 LDRAA and LDRAB combined pointer authentication and load instructions. These instructions authenticate the base register and load 8 byte from it plus a scaled 10-bit offset with optional writeback to update the base register. A new instruction class (ldst_imm10) and operand type (AARCH64_OPND_ADDR_SIMM10) were introduced to handle the special addressing form. include/ 2016-11-18 Szabolcs Nagy <szabolcs.nagy@arm.com> * opcode/aarch64.h (enum aarch64_opnd): Add AARCH64_OPND_ADDR_SIMM10. (enum aarch64_insn_class): Add ldst_imm10. opcodes/ 2016-11-18 Szabolcs Nagy <szabolcs.nagy@arm.com> * aarch64-tbl.h (QL_X1NIL): New. (arch64_opcode_table): Add ldraa, ldrab. (AARCH64_OPERANDS): Add "ADDR_SIMM10". * aarch64-asm.h (aarch64_ins_addr_simm10): Declare. * aarch64-asm.c (aarch64_ins_addr_simm10): Define. * aarch64-dis.h (aarch64_ext_addr_simm10): Declare. * aarch64-dis.c (aarch64_ext_addr_simm10): Define. * aarch64-opc.h (enum aarch64_field_kind): Add FLD_S_simm10. * aarch64-opc.c (fields): Add data for FLD_S_simm10. (operand_general_constraint_met_p): Handle AARCH64_OPND_ADDR_SIMM10. (aarch64_print_operand): Likewise. * aarch64-asm-2.c: Regenerate. * aarch64-dis-2.c: Regenerate. * aarch64-opc-2.c: Regenerate. gas/ 2016-11-18 Szabolcs Nagy <szabolcs.nagy@arm.com> * config/tc-aarch64.c (parse_operands): Handle AARCH64_OPND_ADDR_SIMM10. (fix_insn): Likewise. (warn_unpredictable_ldst): Handle ldst_imm10. * testsuite/gas/aarch64/pac.s: Add ldraa and ldrab tests. * testsuite/gas/aarch64/pac.d: Likewise. * testsuite/gas/aarch64/illegal-ldraa.s: New. * testsuite/gas/aarch64/illegal-ldraa.l: New. * testsuite/gas/aarch64/illegal-ldraa.d: New.
2016-11-18 17:49:06 +08:00
case AARCH64_OPND_ADDR_SIMM10:
[BINUTILS, AARCH64, 4/8] Add Tag setting instructions in Memory Tagging Extension This patch is part of the patch series to add support for ARMv8.5-A Memory Tagging Extensions which is an optional extension to ARMv8.5-A and is enabled using the +memtag command line option. This patch add support to the Tag setting instructions from MTE which consists of the following instructions: - STG [<Xn|SP>, #<simm>] - STG [<Xn|SP>, #<simm>]! - STG [<Xn|SP>], #<simm> - STZG [<Xn|SP>, #<simm>] - STZG [<Xn|SP>, #<simm>]! - STZG [<Xn|SP>], #<simm> - ST2G [<Xn|SP>, #<simm>] - ST2G [<Xn|SP>, #<simm>]! - ST2G [<Xn|SP>], #<simm> - STZ2G [<Xn|SP>, #<simm>] - STZ2G [<Xn|SP>, #<simm>]! - STZ2G [<Xn|SP>], #<simm> - STGP <Xt>, <Xt2>, [<Xn|SP>, #<imm>] - STGP <Xt>, <Xt2>, [<Xn|SP>, #<imm>]! - STGP <Xt>, <Xt2>, [<Xn|SP>], #<imm> where <Xn|SP> : Is the 64-bit GPR or Stack pointer. <simm> : Is the optional signed immediate offset, a multiple of 16 in the range -4096 to 4080, defaulting to 0. *** include/ChangeLog *** 2018-11-12 Sudakshina Das <sudi.das@arm.com> * opcode/aarch64.h (aarch64_opnd): Add AARCH64_OPND_ADDR_SIMM11 and AARCH64_OPND_ADDR_SIMM13. (aarch64_opnd_qualifier): Add new AARCH64_OPND_QLF_imm_tag. *** opcodes/ChangeLog *** 2018-11-12 Sudakshina Das <sudi.das@arm.com> * aarch64-opc.c (aarch64_opnd_qualifiers): Add new data for AARCH64_OPND_QLF_imm_tag. (operand_general_constraint_met_p): Add case for AARCH64_OPND_ADDR_SIMM11 and AARCH64_OPND_ADDR_SIMM13. (aarch64_print_operand): Likewise. * aarch64-tbl.h (QL_LDST_AT, QL_STGP): New. (aarch64_opcode_table): Add stg, stzg, st2g, stz2g and stgp for both offset and pre/post indexed versions. (AARCH64_OPERANDS): Define ADDR_SIMM11 and ADDR_SIMM13. * aarch64-asm-2.c: Regenerated. * aarch64-dis-2.c: Regenerated. * aarch64-opc-2.c: Regenerated. *** gas/ChangeLog *** 2018-11-12 Sudakshina Das <sudi.das@arm.com> * config/tc-aarch64.c (parse_operands): Add switch case for AARCH64_OPND_ADDR_SIMM11 and AARCH64_OPND_ADDR_SIMM13. (fix_insn): Likewise. (warn_unpredictable_ldst): Exempt STGP. * testsuite/gas/aarch64/armv8_5-a-memtag.s: Add tests for stg, st2g, stzg, stz2g and stgp. * testsuite/gas/aarch64/armv8_5-a-memtag.d: Likewise. * testsuite/gas/aarch64/illegal-memtag.s: Likewise. * testsuite/gas/aarch64/illegal-memtag.l: Likewise.
2018-11-12 21:09:55 +08:00
case AARCH64_OPND_ADDR_SIMM11:
case AARCH64_OPND_ADDR_SIMM13:
case AARCH64_OPND_RCPC3_ADDR_OFFSET:
case AARCH64_OPND_ADDR_OFFSET:
case AARCH64_OPND_RCPC3_ADDR_OPT_POSTIND:
case AARCH64_OPND_RCPC3_ADDR_OPT_PREIND_WB:
case AARCH64_OPND_RCPC3_ADDR_POSTIND:
case AARCH64_OPND_RCPC3_ADDR_PREIND_WB:
aarch64: [SME] Add LD1x, ST1x, LDR and STR instructions This patch is adding new loads and stores defined by SME instructions. gas/ChangeLog: * config/tc-aarch64.c (parse_sme_address): New parser. (parse_sme_za_hv_tiles_operand_with_braces): New parser. (parse_sme_za_array): New parser. (output_operand_error_record): Print error details if present. (parse_operands): Support new operands. * testsuite/gas/aarch64/sme-5-illegal.d: New test. * testsuite/gas/aarch64/sme-5-illegal.l: New test. * testsuite/gas/aarch64/sme-5-illegal.s: New test. * testsuite/gas/aarch64/sme-5.d: New test. * testsuite/gas/aarch64/sme-5.s: New test. * testsuite/gas/aarch64/sme-6-illegal.d: New test. * testsuite/gas/aarch64/sme-6-illegal.l: New test. * testsuite/gas/aarch64/sme-6-illegal.s: New test. * testsuite/gas/aarch64/sme-6.d: New test. * testsuite/gas/aarch64/sme-6.s: New test. * testsuite/gas/aarch64/sme-7-illegal.d: New test. * testsuite/gas/aarch64/sme-7-illegal.l: New test. * testsuite/gas/aarch64/sme-7-illegal.s: New test. * testsuite/gas/aarch64/sme-7.d: New test. * testsuite/gas/aarch64/sme-7.s: New test. include/ChangeLog: * opcode/aarch64.h (enum aarch64_opnd): New operands. (enum aarch64_insn_class): Added sme_ldr and sme_str. (AARCH64_OPDE_UNTIED_IMMS): New operand error kind. opcodes/ChangeLog: * aarch64-asm.c (aarch64_ins_sme_za_hv_tiles): New inserter. (aarch64_ins_sme_za_list): New inserter. (aarch64_ins_sme_za_array): New inserter. (aarch64_ins_sme_addr_ri_u4xvl): New inserter. * aarch64-asm.h (AARCH64_DECL_OPD_INSERTER): Added ins_sme_za_list, ins_sme_za_array and ins_sme_addr_ri_u4xvl. * aarch64-dis.c (aarch64_ext_sme_za_hv_tiles): New extractor. (aarch64_ext_sme_za_list): New extractor. (aarch64_ext_sme_za_array): New extractor. (aarch64_ext_sme_addr_ri_u4xvl): New extractor. * aarch64-dis.h (AARCH64_DECL_OPD_EXTRACTOR): Added ext_sme_za_list, ext_sme_za_array and ext_sme_addr_ri_u4xvl. * aarch64-opc.c (operand_general_constraint_met_p): (aarch64_match_operands_constraint): Handle sme_ldr, sme_str and sme_misc. (aarch64_print_operand): New operands supported. * aarch64-tbl.h (OP_SVE_QUU): New qualifier. (OP_SVE_QZU): New qualifier. aarch64-asm-2.c: Regenerate. aarch64-dis-2.c: Regenerate. aarch64-opc-2.c: Regenerate.
2021-11-18 04:02:06 +08:00
case AARCH64_OPND_SME_ADDR_RI_U4xVL:
[AArch64] Additional SVE instructions This patch supports some additions to the SVE architecture prior to its public release. include/ * opcode/aarch64.h (AARCH64_OPND_SVE_ADDR_RI_S4x16) (AARCH64_OPND_SVE_IMM_ROT1, AARCH64_OPND_SVE_IMM_ROT2) (AARCH64_OPND_SVE_Zm3_INDEX, AARCH64_OPND_SVE_Zm3_22_INDEX) (AARCH64_OPND_SVE_Zm4_INDEX): New aarch64_opnds. opcodes/ * aarch64-tbl.h (OP_SVE_HMH, OP_SVE_VMU_HSD, OP_SVE_VMVU_HSD) (OP_SVE_VMVV_HSD, OP_SVE_VMVVU_HSD, OP_SVE_VM_HSD, OP_SVE_VUVV_HSD) (OP_SVE_VUV_HSD, OP_SVE_VU_HSD, OP_SVE_VVVU_H, OP_SVE_VVVU_S) (OP_SVE_VVVU_HSD, OP_SVE_VVV_D, OP_SVE_VVV_D_H, OP_SVE_VVV_H) (OP_SVE_VVV_HSD, OP_SVE_VVV_S, OP_SVE_VVV_S_B, OP_SVE_VVV_SD_BH) (OP_SVE_VV_BHSDQ, OP_SVE_VV_HSD, OP_SVE_VZVV_HSD, OP_SVE_VZV_HSD) (OP_SVE_V_HSD): New macros. (OP_SVE_VMU_SD, OP_SVE_VMVU_SD, OP_SVE_VM_SD, OP_SVE_VUVV_SD) (OP_SVE_VU_SD, OP_SVE_VVVU_SD, OP_SVE_VVV_SD, OP_SVE_VZVV_SD) (OP_SVE_VZV_SD, OP_SVE_V_SD): Delete. (aarch64_opcode_table): Add new SVE instructions. (aarch64_opcode_table): Use imm_rotate{1,2} instead of imm_rotate for rotation operands. Add new SVE operands. * aarch64-asm.h (ins_sve_addr_ri_s4): New inserter. (ins_sve_quad_index): Likewise. (ins_imm_rotate): Split into... (ins_imm_rotate1, ins_imm_rotate2): ...these two inserters. * aarch64-asm.c (aarch64_ins_imm_rotate): Split into... (aarch64_ins_imm_rotate1, aarch64_ins_imm_rotate2): ...these two functions. (aarch64_ins_sve_addr_ri_s4): New function. (aarch64_ins_sve_quad_index): Likewise. (do_misc_encoding): Handle "MOV Zn.Q, Qm". * aarch64-asm-2.c: Regenerate. * aarch64-dis.h (ext_sve_addr_ri_s4): New extractor. (ext_sve_quad_index): Likewise. (ext_imm_rotate): Split into... (ext_imm_rotate1, ext_imm_rotate2): ...these two extractors. * aarch64-dis.c (aarch64_ext_imm_rotate): Split into... (aarch64_ext_imm_rotate1, aarch64_ext_imm_rotate2): ...these two functions. (aarch64_ext_sve_addr_ri_s4): New function. (aarch64_ext_sve_quad_index): Likewise. (aarch64_ext_sve_index): Allow quad indices. (do_misc_decoding): Likewise. * aarch64-dis-2.c: Regenerate. * aarch64-opc.h (FLD_SVE_i3h, FLD_SVE_rot1, FLD_SVE_rot2): New aarch64_field_kinds. (OPD_F_OD_MASK): Widen by one bit. (OPD_F_NO_ZR): Bump accordingly. (get_operand_field_width): New function. * aarch64-opc.c (fields): Add new SVE fields. (operand_general_constraint_met_p): Handle new SVE operands. (aarch64_print_operand): Likewise. * aarch64-opc-2.c: Regenerate. gas/ * doc/c-aarch64.texi: Document that sve implies fp16, simd and compnum. * config/tc-aarch64.c (parse_vector_type_for_operand): Allow .q to be used with SVE registers. (parse_operands): Handle new SVE operands. (aarch64_features): Make "sve" require F16 rather than FP. Also require COMPNUM. * testsuite/gas/aarch64/sve.s: Add tests for new instructions. Include compnum tests. * testsuite/gas/aarch64/sve.d: Update accordingly. * testsuite/gas/aarch64/sve-invalid.s: Add tests for new instructions. * testsuite/gas/aarch64/sve-invalid.l: Update accordingly. Also update expected output for new FMOV and MOV alternatives.
2017-02-25 02:29:00 +08:00
case AARCH64_OPND_SVE_ADDR_RI_S4x16:
[binutils][aarch64] Matrix Multiply extension enablement [8/X] Hi, This patch is part of a series that adds support for Armv8.6-A (Matrix Multiply and BFloat16 extensions) to binutils. This patch introduces the Matrix Multiply (Int8, F32, F64) extensions to the aarch64 backend. The following instructions are added: {s/u}mmla, usmmla, {us/su}dot, fmmla, ld1rob, ld1roh, d1row, ld1rod, uzip{1/2}, trn{1/2}. Committed on behalf of Mihail Ionescu. gas/ChangeLog: 2019-11-07 Mihail Ionescu <mihail.ionescu@arm.com> * config/tc-aarch64.c: Add new arch fetures to suppport the mm extension. (parse_operands): Add new operand. * testsuite/gas/aarch64/i8mm.s: New test. * testsuite/gas/aarch64/i8mm.d: New test. * testsuite/gas/aarch64/f32mm.s: New test. * testsuite/gas/aarch64/f32mm.d: New test. * testsuite/gas/aarch64/f64mm.s: New test. * testsuite/gas/aarch64/f64mm.d: New test. * testsuite/gas/aarch64/sve-movprfx-mm.s: New test. * testsuite/gas/aarch64/sve-movprfx-mm.d: New test. include/ChangeLog: 2019-11-07 Mihail Ionescu <mihail.ionescu@arm.com> * opcode/aarch64.h (AARCH64_FEATURE_I8MM): New. (AARCH64_FEATURE_F32MM): New. (AARCH64_FEATURE_F64MM): New. (AARCH64_OPND_SVE_ADDR_RI_S4x32): New. (enum aarch64_insn_class): Add new instruction class "aarch64_misc" for instructions that do not require special handling. opcodes/ChangeLog: 2019-11-07 Mihail Ionescu <mihail.ionescu@arm.com> * aarch64-tbl.h (aarch64_feature_i8mm_sve, aarch64_feature_f32mm_sve, aarch64_feature_f64mm_sve, aarch64_feature_i8mm, aarch64_feature_f32mm, aarch64_feature_f64mm): New feature sets. (INT8MATMUL_INSN, F64MATMUL_SVE_INSN, F64MATMUL_INSN, F32MATMUL_SVE_INSN, F32MATMUL_INSN): New macros to define matrix multiply instructions. (I8MM_SVE, F32MM_SVE, F64MM_SVE, I8MM, F32MM, F64MM): New feature set macros. (QL_MMLA64, OP_SVE_SBB): New qualifiers. (OP_SVE_QQQ): New qualifier. (INT8MATMUL_SVE_INSNC, F64MATMUL_SVE_INSNC, F32MATMUL_SVE_INSNC): New feature set for bfloat16 instructions to support the movprfx constraint. (aarch64_opcode_table): Support for SVE_ADDR_RI_S4x32. (aarch64_opcode_table): Define new instructions smmla, ummla, usmmla, usdot, sudot, fmmla, ld1rob, ld1roh, ld1row, ld1rod uzip{1/2}, trn{1/2}. * aarch64-opc.c (operand_general_constraint_met_p): Handle AARCH64_OPND_SVE_ADDR_RI_S4x32. (aarch64_print_operand): Handle AARCH64_OPND_SVE_ADDR_RI_S4x32. * aarch64-dis-2.c (aarch64_opcode_lookup_1, aarch64_find_next_opcode): Account for new instructions. * opcodes/aarch64-asm-2.c (aarch64_insert_operand): Support the new S4x32 operand. * aarch64-opc-2.c (aarch64_operands): Support the new S4x32 operand. Regression tested on arm-none-eabi. Is it ok for trunk? Regards, Mihail
2019-11-08 01:10:01 +08:00
case AARCH64_OPND_SVE_ADDR_RI_S4x32:
[AArch64][SVE 26/32] Add SVE MUL VL addressing modes This patch adds support for addresses of the form: [<base>, #<offset>, MUL VL] This involves adding a new AARCH64_MOD_MUL_VL modifier, which is why I split it out from the other addressing modes. For LD2, LD3 and LD4, the offset must be a multiple of the structure size, so for LD3 the possible values are 0, 3, 6, .... The patch therefore extends value_aligned_p to handle non-power-of-2 alignments. include/ * opcode/aarch64.h (AARCH64_OPND_SVE_ADDR_RI_S4xVL): New aarch64_opnd. (AARCH64_OPND_SVE_ADDR_RI_S4x2xVL, AARCH64_OPND_SVE_ADDR_RI_S4x3xVL) (AARCH64_OPND_SVE_ADDR_RI_S4x4xVL, AARCH64_OPND_SVE_ADDR_RI_S6xVL) (AARCH64_OPND_SVE_ADDR_RI_S9xVL): Likewise. (AARCH64_MOD_MUL_VL): New aarch64_modifier_kind. opcodes/ * aarch64-tbl.h (AARCH64_OPERANDS): Add entries for new MUL VL operands. * aarch64-opc.c (aarch64_operand_modifiers): Initialize the AARCH64_MOD_MUL_VL entry. (value_aligned_p): Cope with non-power-of-two alignments. (operand_general_constraint_met_p): Handle the new MUL VL addresses. (print_immediate_offset_address): Likewise. (aarch64_print_operand): Likewise. * aarch64-opc-2.c: Regenerate. * aarch64-asm.h (ins_sve_addr_ri_s4xvl, ins_sve_addr_ri_s6xvl) (ins_sve_addr_ri_s9xvl): New inserters. * aarch64-asm.c (aarch64_ins_sve_addr_ri_s4xvl): New function. (aarch64_ins_sve_addr_ri_s6xvl): Likewise. (aarch64_ins_sve_addr_ri_s9xvl): Likewise. * aarch64-asm-2.c: Regenerate. * aarch64-dis.h (ext_sve_addr_ri_s4xvl, ext_sve_addr_ri_s6xvl) (ext_sve_addr_ri_s9xvl): New extractors. * aarch64-dis.c (aarch64_ext_sve_addr_reg_mul_vl): New function. (aarch64_ext_sve_addr_ri_s4xvl): Likewise. (aarch64_ext_sve_addr_ri_s6xvl): Likewise. (aarch64_ext_sve_addr_ri_s9xvl): Likewise. * aarch64-dis-2.c: Regenerate. gas/ * config/tc-aarch64.c (SHIFTED_NONE, SHIFTED_MUL_VL): New parse_shift_modes. (parse_shift): Handle SHIFTED_MUL_VL. (parse_address_main): Add an imm_shift_mode parameter. (parse_address, parse_sve_address): Update accordingly. (parse_operands): Handle MUL VL addressing modes.
2016-09-21 23:56:15 +08:00
case AARCH64_OPND_SVE_ADDR_RI_S4xVL:
case AARCH64_OPND_SVE_ADDR_RI_S4x2xVL:
case AARCH64_OPND_SVE_ADDR_RI_S4x3xVL:
case AARCH64_OPND_SVE_ADDR_RI_S4x4xVL:
case AARCH64_OPND_SVE_ADDR_RI_S6xVL:
case AARCH64_OPND_SVE_ADDR_RI_S9xVL:
[AArch64][SVE 25/32] Add support for SVE addressing modes This patch adds most of the new SVE addressing modes and associated operands. A follow-on patch adds MUL VL, since handling it separately makes the changes easier to read. The patch also introduces a new "operand-dependent data" field to the operand flags, based closely on the existing one for opcode flags. For SVE this new field needs only 2 bits, but it could be widened in future if necessary. include/ * opcode/aarch64.h (AARCH64_OPND_SVE_ADDR_RI_U6): New aarch64_opnd. (AARCH64_OPND_SVE_ADDR_RI_U6x2, AARCH64_OPND_SVE_ADDR_RI_U6x4) (AARCH64_OPND_SVE_ADDR_RI_U6x8, AARCH64_OPND_SVE_ADDR_RR) (AARCH64_OPND_SVE_ADDR_RR_LSL1, AARCH64_OPND_SVE_ADDR_RR_LSL2) (AARCH64_OPND_SVE_ADDR_RR_LSL3, AARCH64_OPND_SVE_ADDR_RX) (AARCH64_OPND_SVE_ADDR_RX_LSL1, AARCH64_OPND_SVE_ADDR_RX_LSL2) (AARCH64_OPND_SVE_ADDR_RX_LSL3, AARCH64_OPND_SVE_ADDR_RZ) (AARCH64_OPND_SVE_ADDR_RZ_LSL1, AARCH64_OPND_SVE_ADDR_RZ_LSL2) (AARCH64_OPND_SVE_ADDR_RZ_LSL3, AARCH64_OPND_SVE_ADDR_RZ_XTW_14) (AARCH64_OPND_SVE_ADDR_RZ_XTW_22, AARCH64_OPND_SVE_ADDR_RZ_XTW1_14) (AARCH64_OPND_SVE_ADDR_RZ_XTW1_22, AARCH64_OPND_SVE_ADDR_RZ_XTW2_14) (AARCH64_OPND_SVE_ADDR_RZ_XTW2_22, AARCH64_OPND_SVE_ADDR_RZ_XTW3_14) (AARCH64_OPND_SVE_ADDR_RZ_XTW3_22, AARCH64_OPND_SVE_ADDR_ZI_U5) (AARCH64_OPND_SVE_ADDR_ZI_U5x2, AARCH64_OPND_SVE_ADDR_ZI_U5x4) (AARCH64_OPND_SVE_ADDR_ZI_U5x8, AARCH64_OPND_SVE_ADDR_ZZ_LSL) (AARCH64_OPND_SVE_ADDR_ZZ_SXTW, AARCH64_OPND_SVE_ADDR_ZZ_UXTW): Likewise. opcodes/ * aarch64-tbl.h (AARCH64_OPERANDS): Add entries for the new SVE address operands. * aarch64-opc.h (FLD_SVE_imm6, FLD_SVE_msz, FLD_SVE_xs_14) (FLD_SVE_xs_22): New aarch64_field_kinds. (OPD_F_OD_MASK, OPD_F_OD_LSB, OPD_F_NO_ZR): New flags. (get_operand_specific_data): New function. * aarch64-opc.c (fields): Add entries for FLD_SVE_imm6, FLD_SVE_msz, FLD_SVE_xs_14 and FLD_SVE_xs_22. (operand_general_constraint_met_p): Handle the new SVE address operands. (sve_reg): New array. (get_addr_sve_reg_name): New function. (aarch64_print_operand): Handle the new SVE address operands. * aarch64-opc-2.c: Regenerate. * aarch64-asm.h (ins_sve_addr_ri_u6, ins_sve_addr_rr_lsl) (ins_sve_addr_rz_xtw, ins_sve_addr_zi_u5, ins_sve_addr_zz_lsl) (ins_sve_addr_zz_sxtw, ins_sve_addr_zz_uxtw): New inserters. * aarch64-asm.c (aarch64_ins_sve_addr_ri_u6): New function. (aarch64_ins_sve_addr_rr_lsl): Likewise. (aarch64_ins_sve_addr_rz_xtw): Likewise. (aarch64_ins_sve_addr_zi_u5): Likewise. (aarch64_ins_sve_addr_zz): Likewise. (aarch64_ins_sve_addr_zz_lsl): Likewise. (aarch64_ins_sve_addr_zz_sxtw): Likewise. (aarch64_ins_sve_addr_zz_uxtw): Likewise. * aarch64-asm-2.c: Regenerate. * aarch64-dis.h (ext_sve_addr_ri_u6, ext_sve_addr_rr_lsl) (ext_sve_addr_rz_xtw, ext_sve_addr_zi_u5, ext_sve_addr_zz_lsl) (ext_sve_addr_zz_sxtw, ext_sve_addr_zz_uxtw): New extractors. * aarch64-dis.c (aarch64_ext_sve_add_reg_imm): New function. (aarch64_ext_sve_addr_ri_u6): Likewise. (aarch64_ext_sve_addr_rr_lsl): Likewise. (aarch64_ext_sve_addr_rz_xtw): Likewise. (aarch64_ext_sve_addr_zi_u5): Likewise. (aarch64_ext_sve_addr_zz): Likewise. (aarch64_ext_sve_addr_zz_lsl): Likewise. (aarch64_ext_sve_addr_zz_sxtw): Likewise. (aarch64_ext_sve_addr_zz_uxtw): Likewise. * aarch64-dis-2.c: Regenerate. gas/ * config/tc-aarch64.c (REG_TYPE_SVE_BASE, REG_TYPE_SVE_OFFSET): New register types. (get_reg_expected_msg): Handle them. (aarch64_addr_reg_parse): New function, split out from aarch64_reg_parse_32_64. Handle Z registers too. (aarch64_reg_parse_32_64): Call it. (parse_address_main): Add base_qualifier, offset_qualifier, base_type and offset_type parameters. Handle SVE base and offset registers. (parse_address): Update call to parse_address_main. (parse_sve_address): New function. (parse_operands): Parse the new SVE address operands.
2016-09-21 23:55:49 +08:00
case AARCH64_OPND_SVE_ADDR_RI_U6:
case AARCH64_OPND_SVE_ADDR_RI_U6x2:
case AARCH64_OPND_SVE_ADDR_RI_U6x4:
case AARCH64_OPND_SVE_ADDR_RI_U6x8:
print_immediate_offset_address
libopcodes/aarch64: add support for disassembler styling This commit enables disassembler styling for AArch64. After this commit it is possible to have objdump style AArch64 disassembler output (using --disassembler-color option). Once the required GDB patches are merged, GDB will also style the disassembler output. The changes to support styling are mostly split between two files opcodes/aarch64-dis.c and opcodes/aarch64-opc.c. The entry point for the AArch64 disassembler can be found in aarch64-dis.c, this file handles printing the instruction mnemonics, and assembler directives (e.g. '.byte', '.word', etc). Some operands, mostly relating to assembler directives are also printed from this file. This commit changes all of this to pass through suitable styling information. However, for most "normal" instructions, the instruction operands are printed using a two step process. From aarch64-dis.c, in the print_operands function, the function aarch64_print_operand is called, this function is in aarch64-opc.c, and converts an instruction operand into a string. Then, back in print_operands (aarch64-dis.c), the operand string is printed. Unfortunately, the string returned by aarch64_print_operand can be quite complex, it will include syntax elements, like '[' and ']', in addition to register names and immediate values. In some cases, a single operand will expand into what will appear (to the user) as multiple operands separated with a ','. This makes the task of styling more complex, all these different components need to by styled differently, so we need to get the styling information out of aarch64_print_operand in some way. The solution that I propose here is similar to the solution that I used for the i386 disassembler. Currently, aarch64_print_operand uses snprintf to write the operand text into a buffer provided by the caller. What I propose is that we pass an extra argument to the aarch64_print_operand function, this argument will be a structure, the structure contains a callback function and some state. When aarch64_print_operand needs to format part of its output this can be done by using the callback function within the new structure, this callback returns a string with special embedded markers that indicate which mode should be used for each piece of text. Back in aarch64-dis.c we can spot these special style markers and use this to split the disassembler output up and apply the correct style to each piece. To make aarch64-opc.c clearer a series of new static functions have been added, e.g. 'style_reg', 'style_imm', etc. Each of these functions formats a piece of text in a different style, 'register' and 'immediate' in this case. Here's an example taken from aarch64-opc.c of the new functions in use: snprintf (buf, size, "[%s, %s]!", style_reg (styler, base), style_imm (styler, "#%d", opnd->addr.offset.imm)); The aarch64_print_operand function is also called from the assembler to aid in printing diagnostic messages. Right now I have no plans to add styling to the assembler output, and so, the callback function used in the assembler ignores the styling information and just returns an plain string. I've used the source files in gas/testsuite/gas/aarch64/ for testing, and have manually gone through and checked that the styling looks reasonable, however, I'm not an AArch64 expert, so it is possible that the odd piece is styled incorrectly. Please point out any mistakes I've made. With objdump disassembler color turned off, there should be no change in the output after this commit.
2022-04-28 20:31:07 +08:00
(buf, size, opnd, get_64bit_int_reg_name (opnd->addr.base_regno, 1),
styler);
break;
[AArch64][SVE 25/32] Add support for SVE addressing modes This patch adds most of the new SVE addressing modes and associated operands. A follow-on patch adds MUL VL, since handling it separately makes the changes easier to read. The patch also introduces a new "operand-dependent data" field to the operand flags, based closely on the existing one for opcode flags. For SVE this new field needs only 2 bits, but it could be widened in future if necessary. include/ * opcode/aarch64.h (AARCH64_OPND_SVE_ADDR_RI_U6): New aarch64_opnd. (AARCH64_OPND_SVE_ADDR_RI_U6x2, AARCH64_OPND_SVE_ADDR_RI_U6x4) (AARCH64_OPND_SVE_ADDR_RI_U6x8, AARCH64_OPND_SVE_ADDR_RR) (AARCH64_OPND_SVE_ADDR_RR_LSL1, AARCH64_OPND_SVE_ADDR_RR_LSL2) (AARCH64_OPND_SVE_ADDR_RR_LSL3, AARCH64_OPND_SVE_ADDR_RX) (AARCH64_OPND_SVE_ADDR_RX_LSL1, AARCH64_OPND_SVE_ADDR_RX_LSL2) (AARCH64_OPND_SVE_ADDR_RX_LSL3, AARCH64_OPND_SVE_ADDR_RZ) (AARCH64_OPND_SVE_ADDR_RZ_LSL1, AARCH64_OPND_SVE_ADDR_RZ_LSL2) (AARCH64_OPND_SVE_ADDR_RZ_LSL3, AARCH64_OPND_SVE_ADDR_RZ_XTW_14) (AARCH64_OPND_SVE_ADDR_RZ_XTW_22, AARCH64_OPND_SVE_ADDR_RZ_XTW1_14) (AARCH64_OPND_SVE_ADDR_RZ_XTW1_22, AARCH64_OPND_SVE_ADDR_RZ_XTW2_14) (AARCH64_OPND_SVE_ADDR_RZ_XTW2_22, AARCH64_OPND_SVE_ADDR_RZ_XTW3_14) (AARCH64_OPND_SVE_ADDR_RZ_XTW3_22, AARCH64_OPND_SVE_ADDR_ZI_U5) (AARCH64_OPND_SVE_ADDR_ZI_U5x2, AARCH64_OPND_SVE_ADDR_ZI_U5x4) (AARCH64_OPND_SVE_ADDR_ZI_U5x8, AARCH64_OPND_SVE_ADDR_ZZ_LSL) (AARCH64_OPND_SVE_ADDR_ZZ_SXTW, AARCH64_OPND_SVE_ADDR_ZZ_UXTW): Likewise. opcodes/ * aarch64-tbl.h (AARCH64_OPERANDS): Add entries for the new SVE address operands. * aarch64-opc.h (FLD_SVE_imm6, FLD_SVE_msz, FLD_SVE_xs_14) (FLD_SVE_xs_22): New aarch64_field_kinds. (OPD_F_OD_MASK, OPD_F_OD_LSB, OPD_F_NO_ZR): New flags. (get_operand_specific_data): New function. * aarch64-opc.c (fields): Add entries for FLD_SVE_imm6, FLD_SVE_msz, FLD_SVE_xs_14 and FLD_SVE_xs_22. (operand_general_constraint_met_p): Handle the new SVE address operands. (sve_reg): New array. (get_addr_sve_reg_name): New function. (aarch64_print_operand): Handle the new SVE address operands. * aarch64-opc-2.c: Regenerate. * aarch64-asm.h (ins_sve_addr_ri_u6, ins_sve_addr_rr_lsl) (ins_sve_addr_rz_xtw, ins_sve_addr_zi_u5, ins_sve_addr_zz_lsl) (ins_sve_addr_zz_sxtw, ins_sve_addr_zz_uxtw): New inserters. * aarch64-asm.c (aarch64_ins_sve_addr_ri_u6): New function. (aarch64_ins_sve_addr_rr_lsl): Likewise. (aarch64_ins_sve_addr_rz_xtw): Likewise. (aarch64_ins_sve_addr_zi_u5): Likewise. (aarch64_ins_sve_addr_zz): Likewise. (aarch64_ins_sve_addr_zz_lsl): Likewise. (aarch64_ins_sve_addr_zz_sxtw): Likewise. (aarch64_ins_sve_addr_zz_uxtw): Likewise. * aarch64-asm-2.c: Regenerate. * aarch64-dis.h (ext_sve_addr_ri_u6, ext_sve_addr_rr_lsl) (ext_sve_addr_rz_xtw, ext_sve_addr_zi_u5, ext_sve_addr_zz_lsl) (ext_sve_addr_zz_sxtw, ext_sve_addr_zz_uxtw): New extractors. * aarch64-dis.c (aarch64_ext_sve_add_reg_imm): New function. (aarch64_ext_sve_addr_ri_u6): Likewise. (aarch64_ext_sve_addr_rr_lsl): Likewise. (aarch64_ext_sve_addr_rz_xtw): Likewise. (aarch64_ext_sve_addr_zi_u5): Likewise. (aarch64_ext_sve_addr_zz): Likewise. (aarch64_ext_sve_addr_zz_lsl): Likewise. (aarch64_ext_sve_addr_zz_sxtw): Likewise. (aarch64_ext_sve_addr_zz_uxtw): Likewise. * aarch64-dis-2.c: Regenerate. gas/ * config/tc-aarch64.c (REG_TYPE_SVE_BASE, REG_TYPE_SVE_OFFSET): New register types. (get_reg_expected_msg): Handle them. (aarch64_addr_reg_parse): New function, split out from aarch64_reg_parse_32_64. Handle Z registers too. (aarch64_reg_parse_32_64): Call it. (parse_address_main): Add base_qualifier, offset_qualifier, base_type and offset_type parameters. Handle SVE base and offset registers. (parse_address): Update call to parse_address_main. (parse_sve_address): New function. (parse_operands): Parse the new SVE address operands.
2016-09-21 23:55:49 +08:00
case AARCH64_OPND_SVE_ADDR_ZI_U5:
case AARCH64_OPND_SVE_ADDR_ZI_U5x2:
case AARCH64_OPND_SVE_ADDR_ZI_U5x4:
case AARCH64_OPND_SVE_ADDR_ZI_U5x8:
print_immediate_offset_address
(buf, size, opnd,
libopcodes/aarch64: add support for disassembler styling This commit enables disassembler styling for AArch64. After this commit it is possible to have objdump style AArch64 disassembler output (using --disassembler-color option). Once the required GDB patches are merged, GDB will also style the disassembler output. The changes to support styling are mostly split between two files opcodes/aarch64-dis.c and opcodes/aarch64-opc.c. The entry point for the AArch64 disassembler can be found in aarch64-dis.c, this file handles printing the instruction mnemonics, and assembler directives (e.g. '.byte', '.word', etc). Some operands, mostly relating to assembler directives are also printed from this file. This commit changes all of this to pass through suitable styling information. However, for most "normal" instructions, the instruction operands are printed using a two step process. From aarch64-dis.c, in the print_operands function, the function aarch64_print_operand is called, this function is in aarch64-opc.c, and converts an instruction operand into a string. Then, back in print_operands (aarch64-dis.c), the operand string is printed. Unfortunately, the string returned by aarch64_print_operand can be quite complex, it will include syntax elements, like '[' and ']', in addition to register names and immediate values. In some cases, a single operand will expand into what will appear (to the user) as multiple operands separated with a ','. This makes the task of styling more complex, all these different components need to by styled differently, so we need to get the styling information out of aarch64_print_operand in some way. The solution that I propose here is similar to the solution that I used for the i386 disassembler. Currently, aarch64_print_operand uses snprintf to write the operand text into a buffer provided by the caller. What I propose is that we pass an extra argument to the aarch64_print_operand function, this argument will be a structure, the structure contains a callback function and some state. When aarch64_print_operand needs to format part of its output this can be done by using the callback function within the new structure, this callback returns a string with special embedded markers that indicate which mode should be used for each piece of text. Back in aarch64-dis.c we can spot these special style markers and use this to split the disassembler output up and apply the correct style to each piece. To make aarch64-opc.c clearer a series of new static functions have been added, e.g. 'style_reg', 'style_imm', etc. Each of these functions formats a piece of text in a different style, 'register' and 'immediate' in this case. Here's an example taken from aarch64-opc.c of the new functions in use: snprintf (buf, size, "[%s, %s]!", style_reg (styler, base), style_imm (styler, "#%d", opnd->addr.offset.imm)); The aarch64_print_operand function is also called from the assembler to aid in printing diagnostic messages. Right now I have no plans to add styling to the assembler output, and so, the callback function used in the assembler ignores the styling information and just returns an plain string. I've used the source files in gas/testsuite/gas/aarch64/ for testing, and have manually gone through and checked that the styling looks reasonable, however, I'm not an AArch64 expert, so it is possible that the odd piece is styled incorrectly. Please point out any mistakes I've made. With objdump disassembler color turned off, there should be no change in the output after this commit.
2022-04-28 20:31:07 +08:00
get_addr_sve_reg_name (opnd->addr.base_regno, opnd->qualifier),
styler);
[AArch64][SVE 25/32] Add support for SVE addressing modes This patch adds most of the new SVE addressing modes and associated operands. A follow-on patch adds MUL VL, since handling it separately makes the changes easier to read. The patch also introduces a new "operand-dependent data" field to the operand flags, based closely on the existing one for opcode flags. For SVE this new field needs only 2 bits, but it could be widened in future if necessary. include/ * opcode/aarch64.h (AARCH64_OPND_SVE_ADDR_RI_U6): New aarch64_opnd. (AARCH64_OPND_SVE_ADDR_RI_U6x2, AARCH64_OPND_SVE_ADDR_RI_U6x4) (AARCH64_OPND_SVE_ADDR_RI_U6x8, AARCH64_OPND_SVE_ADDR_RR) (AARCH64_OPND_SVE_ADDR_RR_LSL1, AARCH64_OPND_SVE_ADDR_RR_LSL2) (AARCH64_OPND_SVE_ADDR_RR_LSL3, AARCH64_OPND_SVE_ADDR_RX) (AARCH64_OPND_SVE_ADDR_RX_LSL1, AARCH64_OPND_SVE_ADDR_RX_LSL2) (AARCH64_OPND_SVE_ADDR_RX_LSL3, AARCH64_OPND_SVE_ADDR_RZ) (AARCH64_OPND_SVE_ADDR_RZ_LSL1, AARCH64_OPND_SVE_ADDR_RZ_LSL2) (AARCH64_OPND_SVE_ADDR_RZ_LSL3, AARCH64_OPND_SVE_ADDR_RZ_XTW_14) (AARCH64_OPND_SVE_ADDR_RZ_XTW_22, AARCH64_OPND_SVE_ADDR_RZ_XTW1_14) (AARCH64_OPND_SVE_ADDR_RZ_XTW1_22, AARCH64_OPND_SVE_ADDR_RZ_XTW2_14) (AARCH64_OPND_SVE_ADDR_RZ_XTW2_22, AARCH64_OPND_SVE_ADDR_RZ_XTW3_14) (AARCH64_OPND_SVE_ADDR_RZ_XTW3_22, AARCH64_OPND_SVE_ADDR_ZI_U5) (AARCH64_OPND_SVE_ADDR_ZI_U5x2, AARCH64_OPND_SVE_ADDR_ZI_U5x4) (AARCH64_OPND_SVE_ADDR_ZI_U5x8, AARCH64_OPND_SVE_ADDR_ZZ_LSL) (AARCH64_OPND_SVE_ADDR_ZZ_SXTW, AARCH64_OPND_SVE_ADDR_ZZ_UXTW): Likewise. opcodes/ * aarch64-tbl.h (AARCH64_OPERANDS): Add entries for the new SVE address operands. * aarch64-opc.h (FLD_SVE_imm6, FLD_SVE_msz, FLD_SVE_xs_14) (FLD_SVE_xs_22): New aarch64_field_kinds. (OPD_F_OD_MASK, OPD_F_OD_LSB, OPD_F_NO_ZR): New flags. (get_operand_specific_data): New function. * aarch64-opc.c (fields): Add entries for FLD_SVE_imm6, FLD_SVE_msz, FLD_SVE_xs_14 and FLD_SVE_xs_22. (operand_general_constraint_met_p): Handle the new SVE address operands. (sve_reg): New array. (get_addr_sve_reg_name): New function. (aarch64_print_operand): Handle the new SVE address operands. * aarch64-opc-2.c: Regenerate. * aarch64-asm.h (ins_sve_addr_ri_u6, ins_sve_addr_rr_lsl) (ins_sve_addr_rz_xtw, ins_sve_addr_zi_u5, ins_sve_addr_zz_lsl) (ins_sve_addr_zz_sxtw, ins_sve_addr_zz_uxtw): New inserters. * aarch64-asm.c (aarch64_ins_sve_addr_ri_u6): New function. (aarch64_ins_sve_addr_rr_lsl): Likewise. (aarch64_ins_sve_addr_rz_xtw): Likewise. (aarch64_ins_sve_addr_zi_u5): Likewise. (aarch64_ins_sve_addr_zz): Likewise. (aarch64_ins_sve_addr_zz_lsl): Likewise. (aarch64_ins_sve_addr_zz_sxtw): Likewise. (aarch64_ins_sve_addr_zz_uxtw): Likewise. * aarch64-asm-2.c: Regenerate. * aarch64-dis.h (ext_sve_addr_ri_u6, ext_sve_addr_rr_lsl) (ext_sve_addr_rz_xtw, ext_sve_addr_zi_u5, ext_sve_addr_zz_lsl) (ext_sve_addr_zz_sxtw, ext_sve_addr_zz_uxtw): New extractors. * aarch64-dis.c (aarch64_ext_sve_add_reg_imm): New function. (aarch64_ext_sve_addr_ri_u6): Likewise. (aarch64_ext_sve_addr_rr_lsl): Likewise. (aarch64_ext_sve_addr_rz_xtw): Likewise. (aarch64_ext_sve_addr_zi_u5): Likewise. (aarch64_ext_sve_addr_zz): Likewise. (aarch64_ext_sve_addr_zz_lsl): Likewise. (aarch64_ext_sve_addr_zz_sxtw): Likewise. (aarch64_ext_sve_addr_zz_uxtw): Likewise. * aarch64-dis-2.c: Regenerate. gas/ * config/tc-aarch64.c (REG_TYPE_SVE_BASE, REG_TYPE_SVE_OFFSET): New register types. (get_reg_expected_msg): Handle them. (aarch64_addr_reg_parse): New function, split out from aarch64_reg_parse_32_64. Handle Z registers too. (aarch64_reg_parse_32_64): Call it. (parse_address_main): Add base_qualifier, offset_qualifier, base_type and offset_type parameters. Handle SVE base and offset registers. (parse_address): Update call to parse_address_main. (parse_sve_address): New function. (parse_operands): Parse the new SVE address operands.
2016-09-21 23:55:49 +08:00
break;
case AARCH64_OPND_SVE_ADDR_ZZ_LSL:
case AARCH64_OPND_SVE_ADDR_ZZ_SXTW:
case AARCH64_OPND_SVE_ADDR_ZZ_UXTW:
print_register_offset_address
(buf, size, opnd,
get_addr_sve_reg_name (opnd->addr.base_regno, opnd->qualifier),
libopcodes/aarch64: add support for disassembler styling This commit enables disassembler styling for AArch64. After this commit it is possible to have objdump style AArch64 disassembler output (using --disassembler-color option). Once the required GDB patches are merged, GDB will also style the disassembler output. The changes to support styling are mostly split between two files opcodes/aarch64-dis.c and opcodes/aarch64-opc.c. The entry point for the AArch64 disassembler can be found in aarch64-dis.c, this file handles printing the instruction mnemonics, and assembler directives (e.g. '.byte', '.word', etc). Some operands, mostly relating to assembler directives are also printed from this file. This commit changes all of this to pass through suitable styling information. However, for most "normal" instructions, the instruction operands are printed using a two step process. From aarch64-dis.c, in the print_operands function, the function aarch64_print_operand is called, this function is in aarch64-opc.c, and converts an instruction operand into a string. Then, back in print_operands (aarch64-dis.c), the operand string is printed. Unfortunately, the string returned by aarch64_print_operand can be quite complex, it will include syntax elements, like '[' and ']', in addition to register names and immediate values. In some cases, a single operand will expand into what will appear (to the user) as multiple operands separated with a ','. This makes the task of styling more complex, all these different components need to by styled differently, so we need to get the styling information out of aarch64_print_operand in some way. The solution that I propose here is similar to the solution that I used for the i386 disassembler. Currently, aarch64_print_operand uses snprintf to write the operand text into a buffer provided by the caller. What I propose is that we pass an extra argument to the aarch64_print_operand function, this argument will be a structure, the structure contains a callback function and some state. When aarch64_print_operand needs to format part of its output this can be done by using the callback function within the new structure, this callback returns a string with special embedded markers that indicate which mode should be used for each piece of text. Back in aarch64-dis.c we can spot these special style markers and use this to split the disassembler output up and apply the correct style to each piece. To make aarch64-opc.c clearer a series of new static functions have been added, e.g. 'style_reg', 'style_imm', etc. Each of these functions formats a piece of text in a different style, 'register' and 'immediate' in this case. Here's an example taken from aarch64-opc.c of the new functions in use: snprintf (buf, size, "[%s, %s]!", style_reg (styler, base), style_imm (styler, "#%d", opnd->addr.offset.imm)); The aarch64_print_operand function is also called from the assembler to aid in printing diagnostic messages. Right now I have no plans to add styling to the assembler output, and so, the callback function used in the assembler ignores the styling information and just returns an plain string. I've used the source files in gas/testsuite/gas/aarch64/ for testing, and have manually gone through and checked that the styling looks reasonable, however, I'm not an AArch64 expert, so it is possible that the odd piece is styled incorrectly. Please point out any mistakes I've made. With objdump disassembler color turned off, there should be no change in the output after this commit.
2022-04-28 20:31:07 +08:00
get_addr_sve_reg_name (opnd->addr.offset.regno, opnd->qualifier),
styler);
[AArch64][SVE 25/32] Add support for SVE addressing modes This patch adds most of the new SVE addressing modes and associated operands. A follow-on patch adds MUL VL, since handling it separately makes the changes easier to read. The patch also introduces a new "operand-dependent data" field to the operand flags, based closely on the existing one for opcode flags. For SVE this new field needs only 2 bits, but it could be widened in future if necessary. include/ * opcode/aarch64.h (AARCH64_OPND_SVE_ADDR_RI_U6): New aarch64_opnd. (AARCH64_OPND_SVE_ADDR_RI_U6x2, AARCH64_OPND_SVE_ADDR_RI_U6x4) (AARCH64_OPND_SVE_ADDR_RI_U6x8, AARCH64_OPND_SVE_ADDR_RR) (AARCH64_OPND_SVE_ADDR_RR_LSL1, AARCH64_OPND_SVE_ADDR_RR_LSL2) (AARCH64_OPND_SVE_ADDR_RR_LSL3, AARCH64_OPND_SVE_ADDR_RX) (AARCH64_OPND_SVE_ADDR_RX_LSL1, AARCH64_OPND_SVE_ADDR_RX_LSL2) (AARCH64_OPND_SVE_ADDR_RX_LSL3, AARCH64_OPND_SVE_ADDR_RZ) (AARCH64_OPND_SVE_ADDR_RZ_LSL1, AARCH64_OPND_SVE_ADDR_RZ_LSL2) (AARCH64_OPND_SVE_ADDR_RZ_LSL3, AARCH64_OPND_SVE_ADDR_RZ_XTW_14) (AARCH64_OPND_SVE_ADDR_RZ_XTW_22, AARCH64_OPND_SVE_ADDR_RZ_XTW1_14) (AARCH64_OPND_SVE_ADDR_RZ_XTW1_22, AARCH64_OPND_SVE_ADDR_RZ_XTW2_14) (AARCH64_OPND_SVE_ADDR_RZ_XTW2_22, AARCH64_OPND_SVE_ADDR_RZ_XTW3_14) (AARCH64_OPND_SVE_ADDR_RZ_XTW3_22, AARCH64_OPND_SVE_ADDR_ZI_U5) (AARCH64_OPND_SVE_ADDR_ZI_U5x2, AARCH64_OPND_SVE_ADDR_ZI_U5x4) (AARCH64_OPND_SVE_ADDR_ZI_U5x8, AARCH64_OPND_SVE_ADDR_ZZ_LSL) (AARCH64_OPND_SVE_ADDR_ZZ_SXTW, AARCH64_OPND_SVE_ADDR_ZZ_UXTW): Likewise. opcodes/ * aarch64-tbl.h (AARCH64_OPERANDS): Add entries for the new SVE address operands. * aarch64-opc.h (FLD_SVE_imm6, FLD_SVE_msz, FLD_SVE_xs_14) (FLD_SVE_xs_22): New aarch64_field_kinds. (OPD_F_OD_MASK, OPD_F_OD_LSB, OPD_F_NO_ZR): New flags. (get_operand_specific_data): New function. * aarch64-opc.c (fields): Add entries for FLD_SVE_imm6, FLD_SVE_msz, FLD_SVE_xs_14 and FLD_SVE_xs_22. (operand_general_constraint_met_p): Handle the new SVE address operands. (sve_reg): New array. (get_addr_sve_reg_name): New function. (aarch64_print_operand): Handle the new SVE address operands. * aarch64-opc-2.c: Regenerate. * aarch64-asm.h (ins_sve_addr_ri_u6, ins_sve_addr_rr_lsl) (ins_sve_addr_rz_xtw, ins_sve_addr_zi_u5, ins_sve_addr_zz_lsl) (ins_sve_addr_zz_sxtw, ins_sve_addr_zz_uxtw): New inserters. * aarch64-asm.c (aarch64_ins_sve_addr_ri_u6): New function. (aarch64_ins_sve_addr_rr_lsl): Likewise. (aarch64_ins_sve_addr_rz_xtw): Likewise. (aarch64_ins_sve_addr_zi_u5): Likewise. (aarch64_ins_sve_addr_zz): Likewise. (aarch64_ins_sve_addr_zz_lsl): Likewise. (aarch64_ins_sve_addr_zz_sxtw): Likewise. (aarch64_ins_sve_addr_zz_uxtw): Likewise. * aarch64-asm-2.c: Regenerate. * aarch64-dis.h (ext_sve_addr_ri_u6, ext_sve_addr_rr_lsl) (ext_sve_addr_rz_xtw, ext_sve_addr_zi_u5, ext_sve_addr_zz_lsl) (ext_sve_addr_zz_sxtw, ext_sve_addr_zz_uxtw): New extractors. * aarch64-dis.c (aarch64_ext_sve_add_reg_imm): New function. (aarch64_ext_sve_addr_ri_u6): Likewise. (aarch64_ext_sve_addr_rr_lsl): Likewise. (aarch64_ext_sve_addr_rz_xtw): Likewise. (aarch64_ext_sve_addr_zi_u5): Likewise. (aarch64_ext_sve_addr_zz): Likewise. (aarch64_ext_sve_addr_zz_lsl): Likewise. (aarch64_ext_sve_addr_zz_sxtw): Likewise. (aarch64_ext_sve_addr_zz_uxtw): Likewise. * aarch64-dis-2.c: Regenerate. gas/ * config/tc-aarch64.c (REG_TYPE_SVE_BASE, REG_TYPE_SVE_OFFSET): New register types. (get_reg_expected_msg): Handle them. (aarch64_addr_reg_parse): New function, split out from aarch64_reg_parse_32_64. Handle Z registers too. (aarch64_reg_parse_32_64): Call it. (parse_address_main): Add base_qualifier, offset_qualifier, base_type and offset_type parameters. Handle SVE base and offset registers. (parse_address): Update call to parse_address_main. (parse_sve_address): New function. (parse_operands): Parse the new SVE address operands.
2016-09-21 23:55:49 +08:00
break;
case AARCH64_OPND_ADDR_UIMM12:
name = get_64bit_int_reg_name (opnd->addr.base_regno, 1);
if (opnd->addr.offset.imm)
libopcodes/aarch64: add support for disassembler styling This commit enables disassembler styling for AArch64. After this commit it is possible to have objdump style AArch64 disassembler output (using --disassembler-color option). Once the required GDB patches are merged, GDB will also style the disassembler output. The changes to support styling are mostly split between two files opcodes/aarch64-dis.c and opcodes/aarch64-opc.c. The entry point for the AArch64 disassembler can be found in aarch64-dis.c, this file handles printing the instruction mnemonics, and assembler directives (e.g. '.byte', '.word', etc). Some operands, mostly relating to assembler directives are also printed from this file. This commit changes all of this to pass through suitable styling information. However, for most "normal" instructions, the instruction operands are printed using a two step process. From aarch64-dis.c, in the print_operands function, the function aarch64_print_operand is called, this function is in aarch64-opc.c, and converts an instruction operand into a string. Then, back in print_operands (aarch64-dis.c), the operand string is printed. Unfortunately, the string returned by aarch64_print_operand can be quite complex, it will include syntax elements, like '[' and ']', in addition to register names and immediate values. In some cases, a single operand will expand into what will appear (to the user) as multiple operands separated with a ','. This makes the task of styling more complex, all these different components need to by styled differently, so we need to get the styling information out of aarch64_print_operand in some way. The solution that I propose here is similar to the solution that I used for the i386 disassembler. Currently, aarch64_print_operand uses snprintf to write the operand text into a buffer provided by the caller. What I propose is that we pass an extra argument to the aarch64_print_operand function, this argument will be a structure, the structure contains a callback function and some state. When aarch64_print_operand needs to format part of its output this can be done by using the callback function within the new structure, this callback returns a string with special embedded markers that indicate which mode should be used for each piece of text. Back in aarch64-dis.c we can spot these special style markers and use this to split the disassembler output up and apply the correct style to each piece. To make aarch64-opc.c clearer a series of new static functions have been added, e.g. 'style_reg', 'style_imm', etc. Each of these functions formats a piece of text in a different style, 'register' and 'immediate' in this case. Here's an example taken from aarch64-opc.c of the new functions in use: snprintf (buf, size, "[%s, %s]!", style_reg (styler, base), style_imm (styler, "#%d", opnd->addr.offset.imm)); The aarch64_print_operand function is also called from the assembler to aid in printing diagnostic messages. Right now I have no plans to add styling to the assembler output, and so, the callback function used in the assembler ignores the styling information and just returns an plain string. I've used the source files in gas/testsuite/gas/aarch64/ for testing, and have manually gone through and checked that the styling looks reasonable, however, I'm not an AArch64 expert, so it is possible that the odd piece is styled incorrectly. Please point out any mistakes I've made. With objdump disassembler color turned off, there should be no change in the output after this commit.
2022-04-28 20:31:07 +08:00
snprintf (buf, size, "[%s, %s]",
style_reg (styler, name),
style_imm (styler, "#%d", opnd->addr.offset.imm));
else
libopcodes/aarch64: add support for disassembler styling This commit enables disassembler styling for AArch64. After this commit it is possible to have objdump style AArch64 disassembler output (using --disassembler-color option). Once the required GDB patches are merged, GDB will also style the disassembler output. The changes to support styling are mostly split between two files opcodes/aarch64-dis.c and opcodes/aarch64-opc.c. The entry point for the AArch64 disassembler can be found in aarch64-dis.c, this file handles printing the instruction mnemonics, and assembler directives (e.g. '.byte', '.word', etc). Some operands, mostly relating to assembler directives are also printed from this file. This commit changes all of this to pass through suitable styling information. However, for most "normal" instructions, the instruction operands are printed using a two step process. From aarch64-dis.c, in the print_operands function, the function aarch64_print_operand is called, this function is in aarch64-opc.c, and converts an instruction operand into a string. Then, back in print_operands (aarch64-dis.c), the operand string is printed. Unfortunately, the string returned by aarch64_print_operand can be quite complex, it will include syntax elements, like '[' and ']', in addition to register names and immediate values. In some cases, a single operand will expand into what will appear (to the user) as multiple operands separated with a ','. This makes the task of styling more complex, all these different components need to by styled differently, so we need to get the styling information out of aarch64_print_operand in some way. The solution that I propose here is similar to the solution that I used for the i386 disassembler. Currently, aarch64_print_operand uses snprintf to write the operand text into a buffer provided by the caller. What I propose is that we pass an extra argument to the aarch64_print_operand function, this argument will be a structure, the structure contains a callback function and some state. When aarch64_print_operand needs to format part of its output this can be done by using the callback function within the new structure, this callback returns a string with special embedded markers that indicate which mode should be used for each piece of text. Back in aarch64-dis.c we can spot these special style markers and use this to split the disassembler output up and apply the correct style to each piece. To make aarch64-opc.c clearer a series of new static functions have been added, e.g. 'style_reg', 'style_imm', etc. Each of these functions formats a piece of text in a different style, 'register' and 'immediate' in this case. Here's an example taken from aarch64-opc.c of the new functions in use: snprintf (buf, size, "[%s, %s]!", style_reg (styler, base), style_imm (styler, "#%d", opnd->addr.offset.imm)); The aarch64_print_operand function is also called from the assembler to aid in printing diagnostic messages. Right now I have no plans to add styling to the assembler output, and so, the callback function used in the assembler ignores the styling information and just returns an plain string. I've used the source files in gas/testsuite/gas/aarch64/ for testing, and have manually gone through and checked that the styling looks reasonable, however, I'm not an AArch64 expert, so it is possible that the odd piece is styled incorrectly. Please point out any mistakes I've made. With objdump disassembler color turned off, there should be no change in the output after this commit.
2022-04-28 20:31:07 +08:00
snprintf (buf, size, "[%s]", style_reg (styler, name));
break;
case AARCH64_OPND_SYSREG:
case AARCH64_OPND_SYSREG128:
for (i = 0; aarch64_sys_regs[i].name; ++i)
Implement Read/Write constraints on system registers on AArch64 This patch adds constraints for read and write only system registers with the msr and mrs instructions. The code will treat having both flags set and none set as the same. These flags add constraints that must be matched up. e.g. a system register with a READ only flag set, can only be used with mrs. If The constraint fails a warning is emitted. Examples of the warnings generated: test.s: Assembler messages: test.s:5: Warning: specified register cannot be written to at operand 1 -- `msr dbgdtrrx_el0,x3' test.s:7: Warning: specified register cannot be read from at operand 2 -- `mrs x3,dbgdtrtx_el0' test.s:8: Warning: specified register cannot be written to at operand 1 -- `msr midr_el1,x3' and disassembly notes: 0000000000000000 <main>: 0: d5130503 msr dbgdtrtx_el0, x3 4: d5130503 msr dbgdtrtx_el0, x3 8: d5330503 mrs x3, dbgdtrrx_el0 c: d5330503 mrs x3, dbgdtrrx_el0 10: d5180003 msr midr_el1, x3 ; note: writing to a read-only register. Note that because dbgdtrrx_el0 and dbgdtrtx_el0 have the same encoding, during disassembly the constraints are use to disambiguate between the two. An exact constraint match is always prefered over partial ones if available. As always the warnings can be suppressed with -w and also be made errors using warnings as errors. binutils/ PR binutils/21446 * doc/binutils.texi (-M): Document AArch64 options. gas/ PR binutils/21446 * testsuite/gas/aarch64/illegal-sysreg-2.s: Fix pmbidr_el1 test. * testsuite/gas/aarch64/illegal-sysreg-2.l: Likewise. * testsuite/gas/aarch64/illegal-sysreg-2.d: Likewise. * testsuite/gas/aarch64/sysreg-diagnostic.s: New. * testsuite/gas/aarch64/sysreg-diagnostic.l: New. * testsuite/gas/aarch64/sysreg-diagnostic.d: New. include/ PR binutils/21446 * opcode/aarch64.h (F_SYS_READ, F_SYS_WRITE): New. opcodes/ PR binutils/21446 * aarch64-asm.c (opintl.h): Include. (aarch64_ins_sysreg): Enforce read/write constraints. * aarch64-dis.c (aarch64_ext_sysreg): Likewise. * aarch64-opc.h (F_DEPRECATED, F_ARCHEXT, F_HASXT): Moved here. (F_REG_READ, F_REG_WRITE): New. * aarch64-opc.c (aarch64_print_operand): Generate notes for AARCH64_OPND_SYSREG. (F_DEPRECATED, F_ARCHEXT, F_HASXT): Move to aarch64-opc.h. (aarch64_sys_regs): Add constraints to currentel, midr_el1, ctr_el0, mpidr_el1, revidr_el1, aidr_el1, dczid_el0, id_dfr0_el1, id_pfr0_el1, id_pfr1_el1, id_afr0_el1, id_mmfr0_el1, id_mmfr1_el1, id_mmfr2_el1, id_mmfr3_el1, id_mmfr4_el1, id_isar0_el1, id_isar1_el1, id_isar2_el1, id_isar3_el1, id_isar4_el1, id_isar5_el1, mvfr0_el1, mvfr1_el1, mvfr2_el1, ccsidr_el1, id_aa64pfr0_el1, id_aa64pfr1_el1, id_aa64dfr0_el1, id_aa64dfr1_el1, id_aa64isar0_el1, id_aa64isar1_el1, id_aa64mmfr0_el1, id_aa64mmfr1_el1, id_aa64mmfr2_el1, id_aa64afr0_el1, id_aa64afr0_el1, id_aa64afr1_el1, id_aa64zfr0_el1, clidr_el1, csselr_el1, vsesr_el2, erridr_el1, erxfr_el1, rvbar_el1, rvbar_el2, rvbar_el3, isr_el1, tpidrro_el0, cntfrq_el0, cntpct_el0, cntvct_el0, mdccsr_el0, dbgdtrrx_el0, dbgdtrtx_el0, osdtrrx_el1, osdtrtx_el1, mdrar_el1, oslar_el1, oslsr_el1, dbgauthstatus_el1, pmbidr_el1, pmsidr_el1, pmswinc_el0, pmceid0_el0, pmceid1_el0. * aarch64-tbl.h (aarch64_opcode_table): Add constraints to msr (F_SYS_WRITE), mrs (F_SYS_READ).
2018-05-15 23:37:20 +08:00
{
aarch64: Add support for Armv8-R system registers This patch adds support for the system registers introduced in Armv8-R AArch64. gas/ChangeLog: 2020-09-08 Alex Coplan <alex.coplan@arm.com> * config/tc-aarch64.c (parse_sys_reg): Also pass sysreg name to validation function. (parse_sys_ins_reg): Likewise. (print_operands): Pass CPU features to aarch64_print_operand(). * testsuite/gas/aarch64/v8-r-bad-sysregs.d: New test. * testsuite/gas/aarch64/v8-r-bad-sysregs.l: Error output. * testsuite/gas/aarch64/v8-r-bad-sysregs.s: Input. * testsuite/gas/aarch64/v8-r-sysregs-need-arch.d: New test. * testsuite/gas/aarch64/v8-r-sysregs-need-arch.l: Error output. * testsuite/gas/aarch64/v8-r-sysregs.d: New test. * testsuite/gas/aarch64/v8-r-sysregs.s: Input for previous two tests. include/ChangeLog: 2020-09-08 Alex Coplan <alex.coplan@arm.com> * opcode/aarch64.h (aarch64_sys_ins_reg_supported_p): Also take system register name in order to simplify validation for v8-R. (aarch64_print_operand): Also take CPU feature set, as disassembly for system registers now depends on arch variant. opcodes/ChangeLog: 2020-09-08 Alex Coplan <alex.coplan@arm.com> * aarch64-dis.c (print_operands): Pass CPU features to aarch64_print_operand(). * aarch64-opc.c (aarch64_print_operand): Use CPU features to determine preferred disassembly of system registers. (SR_RNG): Refactor to use new SR_FEAT2 macro. (SR_FEAT2): New. (SR_V8_1_A): New. (SR_V8_4_A): New. (SR_V8_A): New. (SR_V8_R): New. (SR_EXPAND_ELx): New. (SR_EXPAND_EL12): New. (aarch64_sys_regs): Specify which registers are only on A-profile, add R-profile system registers. (ENC_BARLAR): New. (PRBARn_ELx): New. (PRLARn_ELx): New. (aarch64_sys_ins_reg_supported_p): Reject EL3 registers for Armv8-R AArch64.
2020-09-08 21:21:44 +08:00
const aarch64_sys_reg *sr = aarch64_sys_regs + i;
bool exact_match
aarch64: Add support for Armv8-R system registers This patch adds support for the system registers introduced in Armv8-R AArch64. gas/ChangeLog: 2020-09-08 Alex Coplan <alex.coplan@arm.com> * config/tc-aarch64.c (parse_sys_reg): Also pass sysreg name to validation function. (parse_sys_ins_reg): Likewise. (print_operands): Pass CPU features to aarch64_print_operand(). * testsuite/gas/aarch64/v8-r-bad-sysregs.d: New test. * testsuite/gas/aarch64/v8-r-bad-sysregs.l: Error output. * testsuite/gas/aarch64/v8-r-bad-sysregs.s: Input. * testsuite/gas/aarch64/v8-r-sysregs-need-arch.d: New test. * testsuite/gas/aarch64/v8-r-sysregs-need-arch.l: Error output. * testsuite/gas/aarch64/v8-r-sysregs.d: New test. * testsuite/gas/aarch64/v8-r-sysregs.s: Input for previous two tests. include/ChangeLog: 2020-09-08 Alex Coplan <alex.coplan@arm.com> * opcode/aarch64.h (aarch64_sys_ins_reg_supported_p): Also take system register name in order to simplify validation for v8-R. (aarch64_print_operand): Also take CPU feature set, as disassembly for system registers now depends on arch variant. opcodes/ChangeLog: 2020-09-08 Alex Coplan <alex.coplan@arm.com> * aarch64-dis.c (print_operands): Pass CPU features to aarch64_print_operand(). * aarch64-opc.c (aarch64_print_operand): Use CPU features to determine preferred disassembly of system registers. (SR_RNG): Refactor to use new SR_FEAT2 macro. (SR_FEAT2): New. (SR_V8_1_A): New. (SR_V8_4_A): New. (SR_V8_A): New. (SR_V8_R): New. (SR_EXPAND_ELx): New. (SR_EXPAND_EL12): New. (aarch64_sys_regs): Specify which registers are only on A-profile, add R-profile system registers. (ENC_BARLAR): New. (PRBARn_ELx): New. (PRLARn_ELx): New. (aarch64_sys_ins_reg_supported_p): Reject EL3 registers for Armv8-R AArch64.
2020-09-08 21:21:44 +08:00
= (!(sr->flags & (F_REG_READ | F_REG_WRITE))
|| (sr->flags & opnd->sysreg.flags) == opnd->sysreg.flags)
&& AARCH64_CPU_HAS_ALL_FEATURES (features, sr->features);
Implement Read/Write constraints on system registers on AArch64 This patch adds constraints for read and write only system registers with the msr and mrs instructions. The code will treat having both flags set and none set as the same. These flags add constraints that must be matched up. e.g. a system register with a READ only flag set, can only be used with mrs. If The constraint fails a warning is emitted. Examples of the warnings generated: test.s: Assembler messages: test.s:5: Warning: specified register cannot be written to at operand 1 -- `msr dbgdtrrx_el0,x3' test.s:7: Warning: specified register cannot be read from at operand 2 -- `mrs x3,dbgdtrtx_el0' test.s:8: Warning: specified register cannot be written to at operand 1 -- `msr midr_el1,x3' and disassembly notes: 0000000000000000 <main>: 0: d5130503 msr dbgdtrtx_el0, x3 4: d5130503 msr dbgdtrtx_el0, x3 8: d5330503 mrs x3, dbgdtrrx_el0 c: d5330503 mrs x3, dbgdtrrx_el0 10: d5180003 msr midr_el1, x3 ; note: writing to a read-only register. Note that because dbgdtrrx_el0 and dbgdtrtx_el0 have the same encoding, during disassembly the constraints are use to disambiguate between the two. An exact constraint match is always prefered over partial ones if available. As always the warnings can be suppressed with -w and also be made errors using warnings as errors. binutils/ PR binutils/21446 * doc/binutils.texi (-M): Document AArch64 options. gas/ PR binutils/21446 * testsuite/gas/aarch64/illegal-sysreg-2.s: Fix pmbidr_el1 test. * testsuite/gas/aarch64/illegal-sysreg-2.l: Likewise. * testsuite/gas/aarch64/illegal-sysreg-2.d: Likewise. * testsuite/gas/aarch64/sysreg-diagnostic.s: New. * testsuite/gas/aarch64/sysreg-diagnostic.l: New. * testsuite/gas/aarch64/sysreg-diagnostic.d: New. include/ PR binutils/21446 * opcode/aarch64.h (F_SYS_READ, F_SYS_WRITE): New. opcodes/ PR binutils/21446 * aarch64-asm.c (opintl.h): Include. (aarch64_ins_sysreg): Enforce read/write constraints. * aarch64-dis.c (aarch64_ext_sysreg): Likewise. * aarch64-opc.h (F_DEPRECATED, F_ARCHEXT, F_HASXT): Moved here. (F_REG_READ, F_REG_WRITE): New. * aarch64-opc.c (aarch64_print_operand): Generate notes for AARCH64_OPND_SYSREG. (F_DEPRECATED, F_ARCHEXT, F_HASXT): Move to aarch64-opc.h. (aarch64_sys_regs): Add constraints to currentel, midr_el1, ctr_el0, mpidr_el1, revidr_el1, aidr_el1, dczid_el0, id_dfr0_el1, id_pfr0_el1, id_pfr1_el1, id_afr0_el1, id_mmfr0_el1, id_mmfr1_el1, id_mmfr2_el1, id_mmfr3_el1, id_mmfr4_el1, id_isar0_el1, id_isar1_el1, id_isar2_el1, id_isar3_el1, id_isar4_el1, id_isar5_el1, mvfr0_el1, mvfr1_el1, mvfr2_el1, ccsidr_el1, id_aa64pfr0_el1, id_aa64pfr1_el1, id_aa64dfr0_el1, id_aa64dfr1_el1, id_aa64isar0_el1, id_aa64isar1_el1, id_aa64mmfr0_el1, id_aa64mmfr1_el1, id_aa64mmfr2_el1, id_aa64afr0_el1, id_aa64afr0_el1, id_aa64afr1_el1, id_aa64zfr0_el1, clidr_el1, csselr_el1, vsesr_el2, erridr_el1, erxfr_el1, rvbar_el1, rvbar_el2, rvbar_el3, isr_el1, tpidrro_el0, cntfrq_el0, cntpct_el0, cntvct_el0, mdccsr_el0, dbgdtrrx_el0, dbgdtrtx_el0, osdtrrx_el1, osdtrtx_el1, mdrar_el1, oslar_el1, oslsr_el1, dbgauthstatus_el1, pmbidr_el1, pmsidr_el1, pmswinc_el0, pmceid0_el0, pmceid1_el0. * aarch64-tbl.h (aarch64_opcode_table): Add constraints to msr (F_SYS_WRITE), mrs (F_SYS_READ).
2018-05-15 23:37:20 +08:00
/* Try and find an exact match, But if that fails, return the first
partial match that was found. */
if (aarch64_sys_regs[i].value == opnd->sysreg.value
&& ! aarch64_sys_reg_deprecated_p (aarch64_sys_regs[i].flags)
&& ! aarch64_sys_reg_alias_p (aarch64_sys_regs[i].flags)
Implement Read/Write constraints on system registers on AArch64 This patch adds constraints for read and write only system registers with the msr and mrs instructions. The code will treat having both flags set and none set as the same. These flags add constraints that must be matched up. e.g. a system register with a READ only flag set, can only be used with mrs. If The constraint fails a warning is emitted. Examples of the warnings generated: test.s: Assembler messages: test.s:5: Warning: specified register cannot be written to at operand 1 -- `msr dbgdtrrx_el0,x3' test.s:7: Warning: specified register cannot be read from at operand 2 -- `mrs x3,dbgdtrtx_el0' test.s:8: Warning: specified register cannot be written to at operand 1 -- `msr midr_el1,x3' and disassembly notes: 0000000000000000 <main>: 0: d5130503 msr dbgdtrtx_el0, x3 4: d5130503 msr dbgdtrtx_el0, x3 8: d5330503 mrs x3, dbgdtrrx_el0 c: d5330503 mrs x3, dbgdtrrx_el0 10: d5180003 msr midr_el1, x3 ; note: writing to a read-only register. Note that because dbgdtrrx_el0 and dbgdtrtx_el0 have the same encoding, during disassembly the constraints are use to disambiguate between the two. An exact constraint match is always prefered over partial ones if available. As always the warnings can be suppressed with -w and also be made errors using warnings as errors. binutils/ PR binutils/21446 * doc/binutils.texi (-M): Document AArch64 options. gas/ PR binutils/21446 * testsuite/gas/aarch64/illegal-sysreg-2.s: Fix pmbidr_el1 test. * testsuite/gas/aarch64/illegal-sysreg-2.l: Likewise. * testsuite/gas/aarch64/illegal-sysreg-2.d: Likewise. * testsuite/gas/aarch64/sysreg-diagnostic.s: New. * testsuite/gas/aarch64/sysreg-diagnostic.l: New. * testsuite/gas/aarch64/sysreg-diagnostic.d: New. include/ PR binutils/21446 * opcode/aarch64.h (F_SYS_READ, F_SYS_WRITE): New. opcodes/ PR binutils/21446 * aarch64-asm.c (opintl.h): Include. (aarch64_ins_sysreg): Enforce read/write constraints. * aarch64-dis.c (aarch64_ext_sysreg): Likewise. * aarch64-opc.h (F_DEPRECATED, F_ARCHEXT, F_HASXT): Moved here. (F_REG_READ, F_REG_WRITE): New. * aarch64-opc.c (aarch64_print_operand): Generate notes for AARCH64_OPND_SYSREG. (F_DEPRECATED, F_ARCHEXT, F_HASXT): Move to aarch64-opc.h. (aarch64_sys_regs): Add constraints to currentel, midr_el1, ctr_el0, mpidr_el1, revidr_el1, aidr_el1, dczid_el0, id_dfr0_el1, id_pfr0_el1, id_pfr1_el1, id_afr0_el1, id_mmfr0_el1, id_mmfr1_el1, id_mmfr2_el1, id_mmfr3_el1, id_mmfr4_el1, id_isar0_el1, id_isar1_el1, id_isar2_el1, id_isar3_el1, id_isar4_el1, id_isar5_el1, mvfr0_el1, mvfr1_el1, mvfr2_el1, ccsidr_el1, id_aa64pfr0_el1, id_aa64pfr1_el1, id_aa64dfr0_el1, id_aa64dfr1_el1, id_aa64isar0_el1, id_aa64isar1_el1, id_aa64mmfr0_el1, id_aa64mmfr1_el1, id_aa64mmfr2_el1, id_aa64afr0_el1, id_aa64afr0_el1, id_aa64afr1_el1, id_aa64zfr0_el1, clidr_el1, csselr_el1, vsesr_el2, erridr_el1, erxfr_el1, rvbar_el1, rvbar_el2, rvbar_el3, isr_el1, tpidrro_el0, cntfrq_el0, cntpct_el0, cntvct_el0, mdccsr_el0, dbgdtrrx_el0, dbgdtrtx_el0, osdtrrx_el1, osdtrtx_el1, mdrar_el1, oslar_el1, oslsr_el1, dbgauthstatus_el1, pmbidr_el1, pmsidr_el1, pmswinc_el0, pmceid0_el0, pmceid1_el0. * aarch64-tbl.h (aarch64_opcode_table): Add constraints to msr (F_SYS_WRITE), mrs (F_SYS_READ).
2018-05-15 23:37:20 +08:00
&& (name == NULL || exact_match))
{
name = aarch64_sys_regs[i].name;
if (exact_match)
{
if (notes)
*notes = NULL;
break;
}
/* If we didn't match exactly, that means the presense of a flag
indicates what we didn't want for this instruction. e.g. If
F_REG_READ is there, that means we were looking for a write
register. See aarch64_ext_sysreg. */
if (aarch64_sys_regs[i].flags & F_REG_WRITE)
*notes = _("reading from a write-only register");
Implement Read/Write constraints on system registers on AArch64 This patch adds constraints for read and write only system registers with the msr and mrs instructions. The code will treat having both flags set and none set as the same. These flags add constraints that must be matched up. e.g. a system register with a READ only flag set, can only be used with mrs. If The constraint fails a warning is emitted. Examples of the warnings generated: test.s: Assembler messages: test.s:5: Warning: specified register cannot be written to at operand 1 -- `msr dbgdtrrx_el0,x3' test.s:7: Warning: specified register cannot be read from at operand 2 -- `mrs x3,dbgdtrtx_el0' test.s:8: Warning: specified register cannot be written to at operand 1 -- `msr midr_el1,x3' and disassembly notes: 0000000000000000 <main>: 0: d5130503 msr dbgdtrtx_el0, x3 4: d5130503 msr dbgdtrtx_el0, x3 8: d5330503 mrs x3, dbgdtrrx_el0 c: d5330503 mrs x3, dbgdtrrx_el0 10: d5180003 msr midr_el1, x3 ; note: writing to a read-only register. Note that because dbgdtrrx_el0 and dbgdtrtx_el0 have the same encoding, during disassembly the constraints are use to disambiguate between the two. An exact constraint match is always prefered over partial ones if available. As always the warnings can be suppressed with -w and also be made errors using warnings as errors. binutils/ PR binutils/21446 * doc/binutils.texi (-M): Document AArch64 options. gas/ PR binutils/21446 * testsuite/gas/aarch64/illegal-sysreg-2.s: Fix pmbidr_el1 test. * testsuite/gas/aarch64/illegal-sysreg-2.l: Likewise. * testsuite/gas/aarch64/illegal-sysreg-2.d: Likewise. * testsuite/gas/aarch64/sysreg-diagnostic.s: New. * testsuite/gas/aarch64/sysreg-diagnostic.l: New. * testsuite/gas/aarch64/sysreg-diagnostic.d: New. include/ PR binutils/21446 * opcode/aarch64.h (F_SYS_READ, F_SYS_WRITE): New. opcodes/ PR binutils/21446 * aarch64-asm.c (opintl.h): Include. (aarch64_ins_sysreg): Enforce read/write constraints. * aarch64-dis.c (aarch64_ext_sysreg): Likewise. * aarch64-opc.h (F_DEPRECATED, F_ARCHEXT, F_HASXT): Moved here. (F_REG_READ, F_REG_WRITE): New. * aarch64-opc.c (aarch64_print_operand): Generate notes for AARCH64_OPND_SYSREG. (F_DEPRECATED, F_ARCHEXT, F_HASXT): Move to aarch64-opc.h. (aarch64_sys_regs): Add constraints to currentel, midr_el1, ctr_el0, mpidr_el1, revidr_el1, aidr_el1, dczid_el0, id_dfr0_el1, id_pfr0_el1, id_pfr1_el1, id_afr0_el1, id_mmfr0_el1, id_mmfr1_el1, id_mmfr2_el1, id_mmfr3_el1, id_mmfr4_el1, id_isar0_el1, id_isar1_el1, id_isar2_el1, id_isar3_el1, id_isar4_el1, id_isar5_el1, mvfr0_el1, mvfr1_el1, mvfr2_el1, ccsidr_el1, id_aa64pfr0_el1, id_aa64pfr1_el1, id_aa64dfr0_el1, id_aa64dfr1_el1, id_aa64isar0_el1, id_aa64isar1_el1, id_aa64mmfr0_el1, id_aa64mmfr1_el1, id_aa64mmfr2_el1, id_aa64afr0_el1, id_aa64afr0_el1, id_aa64afr1_el1, id_aa64zfr0_el1, clidr_el1, csselr_el1, vsesr_el2, erridr_el1, erxfr_el1, rvbar_el1, rvbar_el2, rvbar_el3, isr_el1, tpidrro_el0, cntfrq_el0, cntpct_el0, cntvct_el0, mdccsr_el0, dbgdtrrx_el0, dbgdtrtx_el0, osdtrrx_el1, osdtrtx_el1, mdrar_el1, oslar_el1, oslsr_el1, dbgauthstatus_el1, pmbidr_el1, pmsidr_el1, pmswinc_el0, pmceid0_el0, pmceid1_el0. * aarch64-tbl.h (aarch64_opcode_table): Add constraints to msr (F_SYS_WRITE), mrs (F_SYS_READ).
2018-05-15 23:37:20 +08:00
else if (aarch64_sys_regs[i].flags & F_REG_READ)
*notes = _("writing to a read-only register");
Implement Read/Write constraints on system registers on AArch64 This patch adds constraints for read and write only system registers with the msr and mrs instructions. The code will treat having both flags set and none set as the same. These flags add constraints that must be matched up. e.g. a system register with a READ only flag set, can only be used with mrs. If The constraint fails a warning is emitted. Examples of the warnings generated: test.s: Assembler messages: test.s:5: Warning: specified register cannot be written to at operand 1 -- `msr dbgdtrrx_el0,x3' test.s:7: Warning: specified register cannot be read from at operand 2 -- `mrs x3,dbgdtrtx_el0' test.s:8: Warning: specified register cannot be written to at operand 1 -- `msr midr_el1,x3' and disassembly notes: 0000000000000000 <main>: 0: d5130503 msr dbgdtrtx_el0, x3 4: d5130503 msr dbgdtrtx_el0, x3 8: d5330503 mrs x3, dbgdtrrx_el0 c: d5330503 mrs x3, dbgdtrrx_el0 10: d5180003 msr midr_el1, x3 ; note: writing to a read-only register. Note that because dbgdtrrx_el0 and dbgdtrtx_el0 have the same encoding, during disassembly the constraints are use to disambiguate between the two. An exact constraint match is always prefered over partial ones if available. As always the warnings can be suppressed with -w and also be made errors using warnings as errors. binutils/ PR binutils/21446 * doc/binutils.texi (-M): Document AArch64 options. gas/ PR binutils/21446 * testsuite/gas/aarch64/illegal-sysreg-2.s: Fix pmbidr_el1 test. * testsuite/gas/aarch64/illegal-sysreg-2.l: Likewise. * testsuite/gas/aarch64/illegal-sysreg-2.d: Likewise. * testsuite/gas/aarch64/sysreg-diagnostic.s: New. * testsuite/gas/aarch64/sysreg-diagnostic.l: New. * testsuite/gas/aarch64/sysreg-diagnostic.d: New. include/ PR binutils/21446 * opcode/aarch64.h (F_SYS_READ, F_SYS_WRITE): New. opcodes/ PR binutils/21446 * aarch64-asm.c (opintl.h): Include. (aarch64_ins_sysreg): Enforce read/write constraints. * aarch64-dis.c (aarch64_ext_sysreg): Likewise. * aarch64-opc.h (F_DEPRECATED, F_ARCHEXT, F_HASXT): Moved here. (F_REG_READ, F_REG_WRITE): New. * aarch64-opc.c (aarch64_print_operand): Generate notes for AARCH64_OPND_SYSREG. (F_DEPRECATED, F_ARCHEXT, F_HASXT): Move to aarch64-opc.h. (aarch64_sys_regs): Add constraints to currentel, midr_el1, ctr_el0, mpidr_el1, revidr_el1, aidr_el1, dczid_el0, id_dfr0_el1, id_pfr0_el1, id_pfr1_el1, id_afr0_el1, id_mmfr0_el1, id_mmfr1_el1, id_mmfr2_el1, id_mmfr3_el1, id_mmfr4_el1, id_isar0_el1, id_isar1_el1, id_isar2_el1, id_isar3_el1, id_isar4_el1, id_isar5_el1, mvfr0_el1, mvfr1_el1, mvfr2_el1, ccsidr_el1, id_aa64pfr0_el1, id_aa64pfr1_el1, id_aa64dfr0_el1, id_aa64dfr1_el1, id_aa64isar0_el1, id_aa64isar1_el1, id_aa64mmfr0_el1, id_aa64mmfr1_el1, id_aa64mmfr2_el1, id_aa64afr0_el1, id_aa64afr0_el1, id_aa64afr1_el1, id_aa64zfr0_el1, clidr_el1, csselr_el1, vsesr_el2, erridr_el1, erxfr_el1, rvbar_el1, rvbar_el2, rvbar_el3, isr_el1, tpidrro_el0, cntfrq_el0, cntpct_el0, cntvct_el0, mdccsr_el0, dbgdtrrx_el0, dbgdtrtx_el0, osdtrrx_el1, osdtrtx_el1, mdrar_el1, oslar_el1, oslsr_el1, dbgauthstatus_el1, pmbidr_el1, pmsidr_el1, pmswinc_el0, pmceid0_el0, pmceid1_el0. * aarch64-tbl.h (aarch64_opcode_table): Add constraints to msr (F_SYS_WRITE), mrs (F_SYS_READ).
2018-05-15 23:37:20 +08:00
}
}
if (name)
libopcodes/aarch64: add support for disassembler styling This commit enables disassembler styling for AArch64. After this commit it is possible to have objdump style AArch64 disassembler output (using --disassembler-color option). Once the required GDB patches are merged, GDB will also style the disassembler output. The changes to support styling are mostly split between two files opcodes/aarch64-dis.c and opcodes/aarch64-opc.c. The entry point for the AArch64 disassembler can be found in aarch64-dis.c, this file handles printing the instruction mnemonics, and assembler directives (e.g. '.byte', '.word', etc). Some operands, mostly relating to assembler directives are also printed from this file. This commit changes all of this to pass through suitable styling information. However, for most "normal" instructions, the instruction operands are printed using a two step process. From aarch64-dis.c, in the print_operands function, the function aarch64_print_operand is called, this function is in aarch64-opc.c, and converts an instruction operand into a string. Then, back in print_operands (aarch64-dis.c), the operand string is printed. Unfortunately, the string returned by aarch64_print_operand can be quite complex, it will include syntax elements, like '[' and ']', in addition to register names and immediate values. In some cases, a single operand will expand into what will appear (to the user) as multiple operands separated with a ','. This makes the task of styling more complex, all these different components need to by styled differently, so we need to get the styling information out of aarch64_print_operand in some way. The solution that I propose here is similar to the solution that I used for the i386 disassembler. Currently, aarch64_print_operand uses snprintf to write the operand text into a buffer provided by the caller. What I propose is that we pass an extra argument to the aarch64_print_operand function, this argument will be a structure, the structure contains a callback function and some state. When aarch64_print_operand needs to format part of its output this can be done by using the callback function within the new structure, this callback returns a string with special embedded markers that indicate which mode should be used for each piece of text. Back in aarch64-dis.c we can spot these special style markers and use this to split the disassembler output up and apply the correct style to each piece. To make aarch64-opc.c clearer a series of new static functions have been added, e.g. 'style_reg', 'style_imm', etc. Each of these functions formats a piece of text in a different style, 'register' and 'immediate' in this case. Here's an example taken from aarch64-opc.c of the new functions in use: snprintf (buf, size, "[%s, %s]!", style_reg (styler, base), style_imm (styler, "#%d", opnd->addr.offset.imm)); The aarch64_print_operand function is also called from the assembler to aid in printing diagnostic messages. Right now I have no plans to add styling to the assembler output, and so, the callback function used in the assembler ignores the styling information and just returns an plain string. I've used the source files in gas/testsuite/gas/aarch64/ for testing, and have manually gone through and checked that the styling looks reasonable, however, I'm not an AArch64 expert, so it is possible that the odd piece is styled incorrectly. Please point out any mistakes I've made. With objdump disassembler color turned off, there should be no change in the output after this commit.
2022-04-28 20:31:07 +08:00
snprintf (buf, size, "%s", style_reg (styler, name));
else
{
/* Implementation defined system register. */
Modify AArch64 Assembly and disassembly functions to be able to fail and report why. This patch if the first patch in a series to add the ability to add constraints to system registers that an instruction must adhere to in order for the register to be usable with that instruction. These constraints can also be used to disambiguate between registers with the same encoding during disassembly. This patch adds a new flags entry in the sysreg structures and ensures it is filled in and read out during assembly/disassembly. It also adds the ability for the assemble and disassemble functions to be able to gracefully fail and re-use the existing error reporting infrastructure. The return type of these functions are changed to a boolean to denote success or failure and the error structure is passed around to them. This requires aarch64-gen changes so a lot of the changes here are just mechanical. gas/ PR binutils/21446 * config/tc-aarch64.c (parse_sys_reg): Return register flags. (parse_operands): Fill in register flags. gdb/ PR binutils/21446 * aarch64-tdep.c (aarch64_analyze_prologue, aarch64_software_single_step, aarch64_displaced_step_copy_insn): Indicate not interested in errors. include/ PR binutils/21446 * opcode/aarch64.h (aarch64_opnd_info): Change sysreg to struct. (aarch64_decode_insn): Accept error struct. opcodes/ PR binutils/21446 * aarch64-asm.h (aarch64_insert_operand, aarch64_##x): Return boolean and take error struct. * aarch64-asm.c (aarch64_ext_regno, aarch64_ins_reglane, aarch64_ins_reglist, aarch64_ins_ldst_reglist, aarch64_ins_ldst_reglist_r, aarch64_ins_ldst_elemlist, aarch64_ins_advsimd_imm_shift, aarch64_ins_imm, aarch64_ins_imm_half, aarch64_ins_advsimd_imm_modified, aarch64_ins_fpimm, aarch64_ins_imm_rotate1, aarch64_ins_imm_rotate2, aarch64_ins_fbits, aarch64_ins_aimm, aarch64_ins_limm_1, aarch64_ins_limm, aarch64_ins_inv_limm, aarch64_ins_ft, aarch64_ins_addr_simple, aarch64_ins_addr_regoff, aarch64_ins_addr_offset, aarch64_ins_addr_simm, aarch64_ins_addr_simm10, aarch64_ins_addr_uimm12, aarch64_ins_simd_addr_post, aarch64_ins_cond, aarch64_ins_sysreg, aarch64_ins_pstatefield, aarch64_ins_sysins_op, aarch64_ins_barrier, aarch64_ins_prfop, aarch64_ins_hint, aarch64_ins_reg_extended, aarch64_ins_reg_shifted, aarch64_ins_sve_addr_ri_s4xvl, aarch64_ins_sve_addr_ri_s6xvl, aarch64_ins_sve_addr_ri_s9xvl, aarch64_ins_sve_addr_ri_s4, aarch64_ins_sve_addr_ri_u6, aarch64_ins_sve_addr_rr_lsl, aarch64_ins_sve_addr_rz_xtw, aarch64_ins_sve_addr_zi_u5, aarch64_ext_sve_addr_zz, aarch64_ins_sve_addr_zz_lsl, aarch64_ins_sve_addr_zz_sxtw, aarch64_ins_sve_addr_zz_uxtw, aarch64_ins_sve_aimm, aarch64_ins_sve_asimm, aarch64_ins_sve_index, aarch64_ins_sve_limm_mov, aarch64_ins_sve_quad_index, aarch64_ins_sve_reglist, aarch64_ins_sve_scale, aarch64_ins_sve_shlimm, aarch64_ins_sve_shrimm, aarch64_ins_sve_float_half_one, aarch64_ins_sve_float_half_two, aarch64_ins_sve_float_zero_one, aarch64_opcode_encode): Likewise. * aarch64-dis.h (aarch64_extract_operand, aarch64_##x): Likewise. * aarch64-dis.c (aarch64_ext_regno, aarch64_ext_reglane, aarch64_ext_reglist, aarch64_ext_ldst_reglist, aarch64_ext_ldst_reglist_r, aarch64_ext_ldst_elemlist, aarch64_ext_advsimd_imm_shift, aarch64_ext_imm, aarch64_ext_imm_half, aarch64_ext_advsimd_imm_modified, aarch64_ext_fpimm, aarch64_ext_imm_rotate1, aarch64_ext_imm_rotate2, aarch64_ext_fbits, aarch64_ext_aimm, aarch64_ext_limm_1, aarch64_ext_limm, decode_limm, aarch64_ext_inv_limm, aarch64_ext_ft, aarch64_ext_addr_simple, aarch64_ext_addr_regoff, aarch64_ext_addr_offset, aarch64_ext_addr_simm, aarch64_ext_addr_simm10, aarch64_ext_addr_uimm12, aarch64_ext_simd_addr_post, aarch64_ext_cond, aarch64_ext_sysreg, aarch64_ext_pstatefield, aarch64_ext_sysins_op, aarch64_ext_barrier, aarch64_ext_prfop, aarch64_ext_hint, aarch64_ext_reg_extended, aarch64_ext_reg_shifted, aarch64_ext_sve_addr_ri_s4xvl, aarch64_ext_sve_addr_ri_s6xvl, aarch64_ext_sve_addr_ri_s9xvl, aarch64_ext_sve_addr_ri_s4, aarch64_ext_sve_addr_ri_u6, aarch64_ext_sve_addr_rr_lsl, aarch64_ext_sve_addr_rz_xtw, aarch64_ext_sve_addr_zi_u5, aarch64_ext_sve_addr_zz, aarch64_ext_sve_addr_zz_lsl, aarch64_ext_sve_addr_zz_sxtw, aarch64_ext_sve_addr_zz_uxtw, aarch64_ext_sve_aimm, aarch64_ext_sve_asimm, aarch64_ext_sve_index, aarch64_ext_sve_limm_mov, aarch64_ext_sve_quad_index, aarch64_ext_sve_reglist, aarch64_ext_sve_scale, aarch64_ext_sve_shlimm, aarch64_ext_sve_shrimm, aarch64_ext_sve_float_half_one, aarch64_ext_sve_float_half_two, aarch64_ext_sve_float_zero_one, aarch64_opcode_decode): Likewise. (determine_disassembling_preference, aarch64_decode_insn, print_insn_aarch64_word, print_insn_data): Take errors struct. (print_insn_aarch64): Use errors. * aarch64-asm-2.c: Regenerate. * aarch64-dis-2.c: Regenerate. * aarch64-gen.c (print_operand_inserter): Use errors and change type to boolean in aarch64_insert_operan. (print_operand_extractor): Likewise. * aarch64-opc.c (aarch64_print_operand): Use sysreg struct.
2018-05-15 23:11:42 +08:00
unsigned int value = opnd->sysreg.value;
libopcodes/aarch64: add support for disassembler styling This commit enables disassembler styling for AArch64. After this commit it is possible to have objdump style AArch64 disassembler output (using --disassembler-color option). Once the required GDB patches are merged, GDB will also style the disassembler output. The changes to support styling are mostly split between two files opcodes/aarch64-dis.c and opcodes/aarch64-opc.c. The entry point for the AArch64 disassembler can be found in aarch64-dis.c, this file handles printing the instruction mnemonics, and assembler directives (e.g. '.byte', '.word', etc). Some operands, mostly relating to assembler directives are also printed from this file. This commit changes all of this to pass through suitable styling information. However, for most "normal" instructions, the instruction operands are printed using a two step process. From aarch64-dis.c, in the print_operands function, the function aarch64_print_operand is called, this function is in aarch64-opc.c, and converts an instruction operand into a string. Then, back in print_operands (aarch64-dis.c), the operand string is printed. Unfortunately, the string returned by aarch64_print_operand can be quite complex, it will include syntax elements, like '[' and ']', in addition to register names and immediate values. In some cases, a single operand will expand into what will appear (to the user) as multiple operands separated with a ','. This makes the task of styling more complex, all these different components need to by styled differently, so we need to get the styling information out of aarch64_print_operand in some way. The solution that I propose here is similar to the solution that I used for the i386 disassembler. Currently, aarch64_print_operand uses snprintf to write the operand text into a buffer provided by the caller. What I propose is that we pass an extra argument to the aarch64_print_operand function, this argument will be a structure, the structure contains a callback function and some state. When aarch64_print_operand needs to format part of its output this can be done by using the callback function within the new structure, this callback returns a string with special embedded markers that indicate which mode should be used for each piece of text. Back in aarch64-dis.c we can spot these special style markers and use this to split the disassembler output up and apply the correct style to each piece. To make aarch64-opc.c clearer a series of new static functions have been added, e.g. 'style_reg', 'style_imm', etc. Each of these functions formats a piece of text in a different style, 'register' and 'immediate' in this case. Here's an example taken from aarch64-opc.c of the new functions in use: snprintf (buf, size, "[%s, %s]!", style_reg (styler, base), style_imm (styler, "#%d", opnd->addr.offset.imm)); The aarch64_print_operand function is also called from the assembler to aid in printing diagnostic messages. Right now I have no plans to add styling to the assembler output, and so, the callback function used in the assembler ignores the styling information and just returns an plain string. I've used the source files in gas/testsuite/gas/aarch64/ for testing, and have manually gone through and checked that the styling looks reasonable, however, I'm not an AArch64 expert, so it is possible that the odd piece is styled incorrectly. Please point out any mistakes I've made. With objdump disassembler color turned off, there should be no change in the output after this commit.
2022-04-28 20:31:07 +08:00
snprintf (buf, size, "%s",
style_reg (styler, "s%u_%u_c%u_c%u_%u",
(value >> 14) & 0x3, (value >> 11) & 0x7,
(value >> 7) & 0xf, (value >> 3) & 0xf,
value & 0x7));
}
break;
case AARCH64_OPND_PSTATEFIELD:
for (i = 0; aarch64_pstatefields[i].name; ++i)
aarch64: [SME] Add SME mode selection and state access instructions This patch is adding new SME mode selection and state access instructions: * Add SMSTART and SMSTOP instructions. * Add SVCR system register. gas/ChangeLog: * config/tc-aarch64.c (parse_sme_sm_za): New parser. (parse_operands): New parser. * testsuite/gas/aarch64/sme-8-illegal.d: New test. * testsuite/gas/aarch64/sme-8-illegal.l: New test. * testsuite/gas/aarch64/sme-8-illegal.s: New test. * testsuite/gas/aarch64/sme-8.d: New test. * testsuite/gas/aarch64/sme-8.s: New test. include/ChangeLog: * opcode/aarch64.h (enum aarch64_opnd): New operand AARCH64_OPND_SME_SM_ZA. (enum aarch64_insn_class): New instruction classes sme_start and sme_stop. opcodes/ChangeLog: * aarch64-asm.c (aarch64_ins_pstatefield): New inserter. (aarch64_ins_sme_sm_za): New inserter. * aarch64-dis.c (aarch64_ext_imm): New extractor. (aarch64_ext_pstatefield): New extractor. (aarch64_ext_sme_sm_za): New extractor. * aarch64-opc.c (operand_general_constraint_met_p): New pstatefield value for SME instructions. (aarch64_print_operand): Printout for OPND_SME_SM_ZA. (SR_SME): New register SVCR. * aarch64-opc.h (F_REG_IN_CRM): New register endcoding. * aarch64-opc.h (F_IMM_IN_CRM): New immediate endcoding. (PSTATE_ENCODE_CRM): Encode CRm field. (PSTATE_DECODE_CRM): Decode CRm field. (PSTATE_ENCODE_CRM_IMM): Encode CRm immediate field. (PSTATE_DECODE_CRM_IMM): Decode CRm immediate field. (PSTATE_ENCODE_CRM_AND_IMM): Encode CRm and immediate field. * aarch64-tbl.h (struct aarch64_opcode): New SMSTART and SMSTOP instructions. aarch64-asm-2.c: Regenerate. aarch64-dis-2.c: Regenerate. aarch64-opc-2.c: Regenerate.
2021-11-18 04:15:13 +08:00
if (aarch64_pstatefields[i].value == opnd->pstatefield)
{
/* PSTATEFIELD name is encoded partially in CRm[3:1] for SVCRSM,
SVCRZA and SVCRSMZA. */
uint32_t flags = aarch64_pstatefields[i].flags;
if (flags & F_REG_IN_CRM
&& (PSTATE_DECODE_CRM (opnd->sysreg.flags)
!= PSTATE_DECODE_CRM (flags)))
continue;
break;
}
assert (aarch64_pstatefields[i].name);
libopcodes/aarch64: add support for disassembler styling This commit enables disassembler styling for AArch64. After this commit it is possible to have objdump style AArch64 disassembler output (using --disassembler-color option). Once the required GDB patches are merged, GDB will also style the disassembler output. The changes to support styling are mostly split between two files opcodes/aarch64-dis.c and opcodes/aarch64-opc.c. The entry point for the AArch64 disassembler can be found in aarch64-dis.c, this file handles printing the instruction mnemonics, and assembler directives (e.g. '.byte', '.word', etc). Some operands, mostly relating to assembler directives are also printed from this file. This commit changes all of this to pass through suitable styling information. However, for most "normal" instructions, the instruction operands are printed using a two step process. From aarch64-dis.c, in the print_operands function, the function aarch64_print_operand is called, this function is in aarch64-opc.c, and converts an instruction operand into a string. Then, back in print_operands (aarch64-dis.c), the operand string is printed. Unfortunately, the string returned by aarch64_print_operand can be quite complex, it will include syntax elements, like '[' and ']', in addition to register names and immediate values. In some cases, a single operand will expand into what will appear (to the user) as multiple operands separated with a ','. This makes the task of styling more complex, all these different components need to by styled differently, so we need to get the styling information out of aarch64_print_operand in some way. The solution that I propose here is similar to the solution that I used for the i386 disassembler. Currently, aarch64_print_operand uses snprintf to write the operand text into a buffer provided by the caller. What I propose is that we pass an extra argument to the aarch64_print_operand function, this argument will be a structure, the structure contains a callback function and some state. When aarch64_print_operand needs to format part of its output this can be done by using the callback function within the new structure, this callback returns a string with special embedded markers that indicate which mode should be used for each piece of text. Back in aarch64-dis.c we can spot these special style markers and use this to split the disassembler output up and apply the correct style to each piece. To make aarch64-opc.c clearer a series of new static functions have been added, e.g. 'style_reg', 'style_imm', etc. Each of these functions formats a piece of text in a different style, 'register' and 'immediate' in this case. Here's an example taken from aarch64-opc.c of the new functions in use: snprintf (buf, size, "[%s, %s]!", style_reg (styler, base), style_imm (styler, "#%d", opnd->addr.offset.imm)); The aarch64_print_operand function is also called from the assembler to aid in printing diagnostic messages. Right now I have no plans to add styling to the assembler output, and so, the callback function used in the assembler ignores the styling information and just returns an plain string. I've used the source files in gas/testsuite/gas/aarch64/ for testing, and have manually gone through and checked that the styling looks reasonable, however, I'm not an AArch64 expert, so it is possible that the odd piece is styled incorrectly. Please point out any mistakes I've made. With objdump disassembler color turned off, there should be no change in the output after this commit.
2022-04-28 20:31:07 +08:00
snprintf (buf, size, "%s",
style_reg (styler, aarch64_pstatefields[i].name));
break;
case AARCH64_OPND_SYSREG_AT:
case AARCH64_OPND_SYSREG_DC:
case AARCH64_OPND_SYSREG_IC:
case AARCH64_OPND_SYSREG_TLBI:
case AARCH64_OPND_SYSREG_TLBIP:
[PATCH, BINUTILS, AARCH64, 4/9] Add Execution and Data Restriction instructions This patch is part of the patch series to add support for ARMv8.5-A extensions. (https://developer.arm.com/products/architecture/cpu-architecture/a-profile/docs/ddi0596/a/a64-base-instructions-alphabetic-order) This patch adds the prediction restriction instructions (that is, cfp, dvp, cpp). These instructions are retrospectively made optional for all versions of the architecture from ARMv8.0 to ARMv8.4 and is mandatory from ARMv8.5. Hence adding a new +predres which can be used by the older architectures. *** include/ChangeLog *** 2018-10-09 Sudakshina Das <sudi.das@arm.com> * opcode/aarch64.h (AARCH64_FEATURE_PREDRES): New. (AARCH64_ARCH_V8_5): Add AARCH64_FEATURE_PREDRES by default. (aarch64_opnd): Add AARCH64_OPND_SYSREG_SR. (aarch64_sys_regs_sr): Declare new table. *** opcodes/ChangeLog *** 2018-10-09 Sudakshina Das <sudi.das@arm.com> * aarch64-dis.c (aarch64_ext_sysins_op): Add case for AARCH64_OPND_SYSREG_SR. * aarch64-opc.c (aarch64_print_operand): Likewise. (aarch64_sys_regs_sr): Define table. (aarch64_sys_ins_reg_supported_p): Check for RCTX with AARCH64_FEATURE_PREDRES. * aarch64-tbl.h (aarch64_feature_predres): New. (PREDRES, PREDRES_INSN): New. (aarch64_opcode_table): Add entries for cfp, dvp and cpp. (AARCH64_OPERANDS): Add new description for SYSREG_SR. * aarch64-asm-2.c: Regenerate. * aarch64-dis-2.c: Regenerate. * aarch64-opc-2.c: Regenerate. *** gas/ChangeLog *** 2018-10-09 Sudakshina Das <sudi.das@arm.com> * config/tc-aarch64.c (aarch64_sys_regs_sr_hsh): New. (parse_operands): Add entry for AARCH64_OPND_SYSREG_SR. (md_begin): Allocate and initialize aarch64_sys_regs_sr_hsh with aarch64_sys_regs_sr. (aarch64_features): Add new "predres" option for older architectures. * doc/c-aarch64.texi: Document the same. * testsuite/gas/aarch64/sysreg-4.s: New. * testsuite/gas/aarch64/sysreg-4.d: New. * testsuite/gas/aarch64/illegal-sysreg-4.d: New. * testsuite/gas/aarch64/illegal-sysreg-4.l: New. * testsuite/gas/aarch64/predres.s: New. * testsuite/gas/aarch64/predres.d: New.
2018-09-26 17:52:51 +08:00
case AARCH64_OPND_SYSREG_SR:
libopcodes/aarch64: add support for disassembler styling This commit enables disassembler styling for AArch64. After this commit it is possible to have objdump style AArch64 disassembler output (using --disassembler-color option). Once the required GDB patches are merged, GDB will also style the disassembler output. The changes to support styling are mostly split between two files opcodes/aarch64-dis.c and opcodes/aarch64-opc.c. The entry point for the AArch64 disassembler can be found in aarch64-dis.c, this file handles printing the instruction mnemonics, and assembler directives (e.g. '.byte', '.word', etc). Some operands, mostly relating to assembler directives are also printed from this file. This commit changes all of this to pass through suitable styling information. However, for most "normal" instructions, the instruction operands are printed using a two step process. From aarch64-dis.c, in the print_operands function, the function aarch64_print_operand is called, this function is in aarch64-opc.c, and converts an instruction operand into a string. Then, back in print_operands (aarch64-dis.c), the operand string is printed. Unfortunately, the string returned by aarch64_print_operand can be quite complex, it will include syntax elements, like '[' and ']', in addition to register names and immediate values. In some cases, a single operand will expand into what will appear (to the user) as multiple operands separated with a ','. This makes the task of styling more complex, all these different components need to by styled differently, so we need to get the styling information out of aarch64_print_operand in some way. The solution that I propose here is similar to the solution that I used for the i386 disassembler. Currently, aarch64_print_operand uses snprintf to write the operand text into a buffer provided by the caller. What I propose is that we pass an extra argument to the aarch64_print_operand function, this argument will be a structure, the structure contains a callback function and some state. When aarch64_print_operand needs to format part of its output this can be done by using the callback function within the new structure, this callback returns a string with special embedded markers that indicate which mode should be used for each piece of text. Back in aarch64-dis.c we can spot these special style markers and use this to split the disassembler output up and apply the correct style to each piece. To make aarch64-opc.c clearer a series of new static functions have been added, e.g. 'style_reg', 'style_imm', etc. Each of these functions formats a piece of text in a different style, 'register' and 'immediate' in this case. Here's an example taken from aarch64-opc.c of the new functions in use: snprintf (buf, size, "[%s, %s]!", style_reg (styler, base), style_imm (styler, "#%d", opnd->addr.offset.imm)); The aarch64_print_operand function is also called from the assembler to aid in printing diagnostic messages. Right now I have no plans to add styling to the assembler output, and so, the callback function used in the assembler ignores the styling information and just returns an plain string. I've used the source files in gas/testsuite/gas/aarch64/ for testing, and have manually gone through and checked that the styling looks reasonable, however, I'm not an AArch64 expert, so it is possible that the odd piece is styled incorrectly. Please point out any mistakes I've made. With objdump disassembler color turned off, there should be no change in the output after this commit.
2022-04-28 20:31:07 +08:00
snprintf (buf, size, "%s", style_reg (styler, opnd->sysins_op->name));
break;
case AARCH64_OPND_BARRIER:
aarch64: Add DSB instruction Armv8.7-a variant This patch adds new variant (nXS) of DSB memory barrier instruction available in Armv8.7-a. New nXS variant has different encoding in comparison with pre Armv8.7-a DSB memory barrier variant thus new instruction and new operand was added. DSB memory nXS barrier variant specifies the limitation on the barrier operation. Allowed values are: DSB SYnXS|#28 DSB ISHnXS|#24 DSB NSHnXS|#20 DSB OSHnXS|#16 Please note that till now, for barriers, barrier operation was encoded in 4-bit unsigned immediate CRm field (in the range 0 to 15). For DSB memory nXS barrier variant, barrier operation is a 5-bit unsigned assembly instruction immediate, encoded in instruction in two bits CRm<3:2>: CRm<3:2> #imm 00 16 01 20 10 24 11 28 This patch extends current AArch64 barrier instructions with above mapping. Notable patch changes include: + New DSB memory barrier variant encoding for Armv8.7-a. + New operand BARRIER_DSB_NXS for above instruction in order to distinguish between existing and new DSB instruction flavour. + New set of DSB nXS barrier options. + New instruction inserter and extractor map between instruction immediate 5-bit value and 2-bit CRm field of the instruction itself (see FLD_CRm_dsb_nxs). + Regeneration of aarch64-[asm|dis|opc]-2.c files. + Test cases to cover new instruction assembling and disassembling. For more details regarding DSB memory barrier instruction and its Armv8.7-a flavour please refer to Arm A64 Instruction set documentation for Armv8-A architecture profile, see document pages 132-133 of [0]. [0]: https://developer.arm.com/docs/ddi0596/i gas/ChangeLog: 2020-10-23 Przemyslaw Wirkus <przemyslaw.wirkus@arm.com> * NEWS: Docs update. * config/tc-aarch64.c (parse_operands): Add AARCH64_OPND_BARRIER_DSB_NXS handler. (md_begin): Add content of aarch64_barrier_dsb_nxs_options to aarch64_barrier_opt_hsh hash. * testsuite/gas/aarch64/system-4-invalid.d: New test. * testsuite/gas/aarch64/system-4-invalid.l: New test. * testsuite/gas/aarch64/system-4-invalid.s: New test. * testsuite/gas/aarch64/system-4.d: New test. * testsuite/gas/aarch64/system-4.s: New test. include/ChangeLog: 2020-10-23 Przemyslaw Wirkus <przemyslaw.wirkus@arm.com> * opcode/aarch64.h (enum aarch64_opnd): New operand AARCH64_OPND_BARRIER_DSB_NXS. (aarch64_barrier_dsb_nxs_options): Declare DSB nXS options. opcodes/ChangeLog: 2020-10-23 Przemyslaw Wirkus <przemyslaw.wirkus@arm.com> * aarch64-asm.c (aarch64_ins_barrier_dsb_nxs): New inserter. * aarch64-asm.h (AARCH64_DECL_OPD_INSERTER): New inserter ins_barrier_dsb_nx. * aarch64-dis.c (aarch64_ext_barrier_dsb_nxs): New extractor. * aarch64-dis.h (AARCH64_DECL_OPD_EXTRACTOR): New extractor ext_barrier_dsb_nx. * aarch64-opc.c (aarch64_print_operand): New options table aarch64_barrier_dsb_nxs_options. * aarch64-opc.h (enum aarch64_field_kind): New field name FLD_CRm_dsb_nxs. * aarch64-tbl.h (struct aarch64_opcode): Define DSB nXS barrier Armv8.7-a instruction. * aarch64-asm-2.c: Regenerated. * aarch64-dis-2.c: Regenerated. * aarch64-opc-2.c: Regenerated.
2020-10-28 22:01:36 +08:00
case AARCH64_OPND_BARRIER_DSB_NXS:
libopcodes/aarch64: add support for disassembler styling This commit enables disassembler styling for AArch64. After this commit it is possible to have objdump style AArch64 disassembler output (using --disassembler-color option). Once the required GDB patches are merged, GDB will also style the disassembler output. The changes to support styling are mostly split between two files opcodes/aarch64-dis.c and opcodes/aarch64-opc.c. The entry point for the AArch64 disassembler can be found in aarch64-dis.c, this file handles printing the instruction mnemonics, and assembler directives (e.g. '.byte', '.word', etc). Some operands, mostly relating to assembler directives are also printed from this file. This commit changes all of this to pass through suitable styling information. However, for most "normal" instructions, the instruction operands are printed using a two step process. From aarch64-dis.c, in the print_operands function, the function aarch64_print_operand is called, this function is in aarch64-opc.c, and converts an instruction operand into a string. Then, back in print_operands (aarch64-dis.c), the operand string is printed. Unfortunately, the string returned by aarch64_print_operand can be quite complex, it will include syntax elements, like '[' and ']', in addition to register names and immediate values. In some cases, a single operand will expand into what will appear (to the user) as multiple operands separated with a ','. This makes the task of styling more complex, all these different components need to by styled differently, so we need to get the styling information out of aarch64_print_operand in some way. The solution that I propose here is similar to the solution that I used for the i386 disassembler. Currently, aarch64_print_operand uses snprintf to write the operand text into a buffer provided by the caller. What I propose is that we pass an extra argument to the aarch64_print_operand function, this argument will be a structure, the structure contains a callback function and some state. When aarch64_print_operand needs to format part of its output this can be done by using the callback function within the new structure, this callback returns a string with special embedded markers that indicate which mode should be used for each piece of text. Back in aarch64-dis.c we can spot these special style markers and use this to split the disassembler output up and apply the correct style to each piece. To make aarch64-opc.c clearer a series of new static functions have been added, e.g. 'style_reg', 'style_imm', etc. Each of these functions formats a piece of text in a different style, 'register' and 'immediate' in this case. Here's an example taken from aarch64-opc.c of the new functions in use: snprintf (buf, size, "[%s, %s]!", style_reg (styler, base), style_imm (styler, "#%d", opnd->addr.offset.imm)); The aarch64_print_operand function is also called from the assembler to aid in printing diagnostic messages. Right now I have no plans to add styling to the assembler output, and so, the callback function used in the assembler ignores the styling information and just returns an plain string. I've used the source files in gas/testsuite/gas/aarch64/ for testing, and have manually gone through and checked that the styling looks reasonable, however, I'm not an AArch64 expert, so it is possible that the odd piece is styled incorrectly. Please point out any mistakes I've made. With objdump disassembler color turned off, there should be no change in the output after this commit.
2022-04-28 20:31:07 +08:00
{
if (opnd->barrier->name[0] == '#')
snprintf (buf, size, "%s", style_imm (styler, opnd->barrier->name));
else
snprintf (buf, size, "%s",
style_sub_mnem (styler, opnd->barrier->name));
}
break;
case AARCH64_OPND_BARRIER_ISB:
/* Operand can be omitted, e.g. in DCPS1. */
if (! optional_operand_p (opcode, idx)
|| (opnd->barrier->value
!= get_optional_operand_default_value (opcode)))
libopcodes/aarch64: add support for disassembler styling This commit enables disassembler styling for AArch64. After this commit it is possible to have objdump style AArch64 disassembler output (using --disassembler-color option). Once the required GDB patches are merged, GDB will also style the disassembler output. The changes to support styling are mostly split between two files opcodes/aarch64-dis.c and opcodes/aarch64-opc.c. The entry point for the AArch64 disassembler can be found in aarch64-dis.c, this file handles printing the instruction mnemonics, and assembler directives (e.g. '.byte', '.word', etc). Some operands, mostly relating to assembler directives are also printed from this file. This commit changes all of this to pass through suitable styling information. However, for most "normal" instructions, the instruction operands are printed using a two step process. From aarch64-dis.c, in the print_operands function, the function aarch64_print_operand is called, this function is in aarch64-opc.c, and converts an instruction operand into a string. Then, back in print_operands (aarch64-dis.c), the operand string is printed. Unfortunately, the string returned by aarch64_print_operand can be quite complex, it will include syntax elements, like '[' and ']', in addition to register names and immediate values. In some cases, a single operand will expand into what will appear (to the user) as multiple operands separated with a ','. This makes the task of styling more complex, all these different components need to by styled differently, so we need to get the styling information out of aarch64_print_operand in some way. The solution that I propose here is similar to the solution that I used for the i386 disassembler. Currently, aarch64_print_operand uses snprintf to write the operand text into a buffer provided by the caller. What I propose is that we pass an extra argument to the aarch64_print_operand function, this argument will be a structure, the structure contains a callback function and some state. When aarch64_print_operand needs to format part of its output this can be done by using the callback function within the new structure, this callback returns a string with special embedded markers that indicate which mode should be used for each piece of text. Back in aarch64-dis.c we can spot these special style markers and use this to split the disassembler output up and apply the correct style to each piece. To make aarch64-opc.c clearer a series of new static functions have been added, e.g. 'style_reg', 'style_imm', etc. Each of these functions formats a piece of text in a different style, 'register' and 'immediate' in this case. Here's an example taken from aarch64-opc.c of the new functions in use: snprintf (buf, size, "[%s, %s]!", style_reg (styler, base), style_imm (styler, "#%d", opnd->addr.offset.imm)); The aarch64_print_operand function is also called from the assembler to aid in printing diagnostic messages. Right now I have no plans to add styling to the assembler output, and so, the callback function used in the assembler ignores the styling information and just returns an plain string. I've used the source files in gas/testsuite/gas/aarch64/ for testing, and have manually gone through and checked that the styling looks reasonable, however, I'm not an AArch64 expert, so it is possible that the odd piece is styled incorrectly. Please point out any mistakes I've made. With objdump disassembler color turned off, there should be no change in the output after this commit.
2022-04-28 20:31:07 +08:00
snprintf (buf, size, "%s",
style_imm (styler, "#0x%x", opnd->barrier->value));
break;
case AARCH64_OPND_PRFOP:
if (opnd->prfop->name != NULL)
libopcodes/aarch64: add support for disassembler styling This commit enables disassembler styling for AArch64. After this commit it is possible to have objdump style AArch64 disassembler output (using --disassembler-color option). Once the required GDB patches are merged, GDB will also style the disassembler output. The changes to support styling are mostly split between two files opcodes/aarch64-dis.c and opcodes/aarch64-opc.c. The entry point for the AArch64 disassembler can be found in aarch64-dis.c, this file handles printing the instruction mnemonics, and assembler directives (e.g. '.byte', '.word', etc). Some operands, mostly relating to assembler directives are also printed from this file. This commit changes all of this to pass through suitable styling information. However, for most "normal" instructions, the instruction operands are printed using a two step process. From aarch64-dis.c, in the print_operands function, the function aarch64_print_operand is called, this function is in aarch64-opc.c, and converts an instruction operand into a string. Then, back in print_operands (aarch64-dis.c), the operand string is printed. Unfortunately, the string returned by aarch64_print_operand can be quite complex, it will include syntax elements, like '[' and ']', in addition to register names and immediate values. In some cases, a single operand will expand into what will appear (to the user) as multiple operands separated with a ','. This makes the task of styling more complex, all these different components need to by styled differently, so we need to get the styling information out of aarch64_print_operand in some way. The solution that I propose here is similar to the solution that I used for the i386 disassembler. Currently, aarch64_print_operand uses snprintf to write the operand text into a buffer provided by the caller. What I propose is that we pass an extra argument to the aarch64_print_operand function, this argument will be a structure, the structure contains a callback function and some state. When aarch64_print_operand needs to format part of its output this can be done by using the callback function within the new structure, this callback returns a string with special embedded markers that indicate which mode should be used for each piece of text. Back in aarch64-dis.c we can spot these special style markers and use this to split the disassembler output up and apply the correct style to each piece. To make aarch64-opc.c clearer a series of new static functions have been added, e.g. 'style_reg', 'style_imm', etc. Each of these functions formats a piece of text in a different style, 'register' and 'immediate' in this case. Here's an example taken from aarch64-opc.c of the new functions in use: snprintf (buf, size, "[%s, %s]!", style_reg (styler, base), style_imm (styler, "#%d", opnd->addr.offset.imm)); The aarch64_print_operand function is also called from the assembler to aid in printing diagnostic messages. Right now I have no plans to add styling to the assembler output, and so, the callback function used in the assembler ignores the styling information and just returns an plain string. I've used the source files in gas/testsuite/gas/aarch64/ for testing, and have manually gone through and checked that the styling looks reasonable, however, I'm not an AArch64 expert, so it is possible that the odd piece is styled incorrectly. Please point out any mistakes I've made. With objdump disassembler color turned off, there should be no change in the output after this commit.
2022-04-28 20:31:07 +08:00
snprintf (buf, size, "%s", style_sub_mnem (styler, opnd->prfop->name));
else
libopcodes/aarch64: add support for disassembler styling This commit enables disassembler styling for AArch64. After this commit it is possible to have objdump style AArch64 disassembler output (using --disassembler-color option). Once the required GDB patches are merged, GDB will also style the disassembler output. The changes to support styling are mostly split between two files opcodes/aarch64-dis.c and opcodes/aarch64-opc.c. The entry point for the AArch64 disassembler can be found in aarch64-dis.c, this file handles printing the instruction mnemonics, and assembler directives (e.g. '.byte', '.word', etc). Some operands, mostly relating to assembler directives are also printed from this file. This commit changes all of this to pass through suitable styling information. However, for most "normal" instructions, the instruction operands are printed using a two step process. From aarch64-dis.c, in the print_operands function, the function aarch64_print_operand is called, this function is in aarch64-opc.c, and converts an instruction operand into a string. Then, back in print_operands (aarch64-dis.c), the operand string is printed. Unfortunately, the string returned by aarch64_print_operand can be quite complex, it will include syntax elements, like '[' and ']', in addition to register names and immediate values. In some cases, a single operand will expand into what will appear (to the user) as multiple operands separated with a ','. This makes the task of styling more complex, all these different components need to by styled differently, so we need to get the styling information out of aarch64_print_operand in some way. The solution that I propose here is similar to the solution that I used for the i386 disassembler. Currently, aarch64_print_operand uses snprintf to write the operand text into a buffer provided by the caller. What I propose is that we pass an extra argument to the aarch64_print_operand function, this argument will be a structure, the structure contains a callback function and some state. When aarch64_print_operand needs to format part of its output this can be done by using the callback function within the new structure, this callback returns a string with special embedded markers that indicate which mode should be used for each piece of text. Back in aarch64-dis.c we can spot these special style markers and use this to split the disassembler output up and apply the correct style to each piece. To make aarch64-opc.c clearer a series of new static functions have been added, e.g. 'style_reg', 'style_imm', etc. Each of these functions formats a piece of text in a different style, 'register' and 'immediate' in this case. Here's an example taken from aarch64-opc.c of the new functions in use: snprintf (buf, size, "[%s, %s]!", style_reg (styler, base), style_imm (styler, "#%d", opnd->addr.offset.imm)); The aarch64_print_operand function is also called from the assembler to aid in printing diagnostic messages. Right now I have no plans to add styling to the assembler output, and so, the callback function used in the assembler ignores the styling information and just returns an plain string. I've used the source files in gas/testsuite/gas/aarch64/ for testing, and have manually gone through and checked that the styling looks reasonable, however, I'm not an AArch64 expert, so it is possible that the odd piece is styled incorrectly. Please point out any mistakes I've made. With objdump disassembler color turned off, there should be no change in the output after this commit.
2022-04-28 20:31:07 +08:00
snprintf (buf, size, "%s", style_imm (styler, "#0x%02x",
opnd->prfop->value));
break;
case AARCH64_OPND_RPRFMOP:
enum_value = opnd->imm.value;
if (enum_value < ARRAY_SIZE (aarch64_rprfmop_array)
&& aarch64_rprfmop_array[enum_value])
snprintf (buf, size, "%s",
style_reg (styler, aarch64_rprfmop_array[enum_value]));
else
snprintf (buf, size, "%s",
style_imm (styler, "#%" PRIi64, opnd->imm.value));
break;
[AArch64][Patch 5/5] Add instruction PSB CSYNC The Statistical Profile Extension adds the instruction PSB CSYNC as an alias for the HINT #17 instruction. This patch adds the instruction to binutils as a HINT alias that takes an operand. A new operand type, AARCH64_OPND_BARRIER_PSB, is added to represent the operand to PSB. A parser for the operand type is added to the assembler and a printer to the disassembler. The operand name "csync" is added to the list of HINT options with HINT number #17. Encoding and decoding of the operand is handled by the ins_hint/ext_hint functions added in the preceding patches. gas/ 2015-12-11 Matthew Wahab <matthew.wahab@arm.com> * config/tc-aarch64.c (aarch64_hint_opt_hsh): New. (parse_barrier_psb): New. (parse_operands): Add case for AARCH64_OPND_BARRIER_PSB. (md_begin): Set up aarch64_hint_opt_hsh. gas/testsuite/ 2015-12-11 Matthew Wahab <matthew.wahab@arm.com> * gas/aarch64/system-2.d: Enable the statistical profiling extension. Update the expected output. * gas/aarch64/system-2.s: Add tests for PSB CSYNC. * gas/aarch64/system.d: Update the expected output. include/opcode/ 2015-12-11 Matthew Wahab <matthew.wahab@arm.com> * aarch64.h (aarch64_opnd): Add AARCH64_OPND_BARRIER_PSB. * aarch64-asm-2.c: Regenerate. * aarch64-dis-2.c: Regenerate. * aarch64-opc-2.c: Regenerate. * aarch64-opc.c (aarch64_hint_options): Add "csync". (aarch64_print_operands): Handle AARCH64_OPND_BARRIER_PSB. * aarch64-tbl.h (aarch64_feature_stat_profile): New. (STAT_PROFILE): New. (aarch64_opcode_table): Add "psb". (AARCH64_OPERANDS): Add "BARRIER_PSB". Change-Id: I5ffb672d26a8b15b48785478d359350a9b70ca09
2015-12-11 18:22:40 +08:00
case AARCH64_OPND_BARRIER_PSB:
libopcodes/aarch64: add support for disassembler styling This commit enables disassembler styling for AArch64. After this commit it is possible to have objdump style AArch64 disassembler output (using --disassembler-color option). Once the required GDB patches are merged, GDB will also style the disassembler output. The changes to support styling are mostly split between two files opcodes/aarch64-dis.c and opcodes/aarch64-opc.c. The entry point for the AArch64 disassembler can be found in aarch64-dis.c, this file handles printing the instruction mnemonics, and assembler directives (e.g. '.byte', '.word', etc). Some operands, mostly relating to assembler directives are also printed from this file. This commit changes all of this to pass through suitable styling information. However, for most "normal" instructions, the instruction operands are printed using a two step process. From aarch64-dis.c, in the print_operands function, the function aarch64_print_operand is called, this function is in aarch64-opc.c, and converts an instruction operand into a string. Then, back in print_operands (aarch64-dis.c), the operand string is printed. Unfortunately, the string returned by aarch64_print_operand can be quite complex, it will include syntax elements, like '[' and ']', in addition to register names and immediate values. In some cases, a single operand will expand into what will appear (to the user) as multiple operands separated with a ','. This makes the task of styling more complex, all these different components need to by styled differently, so we need to get the styling information out of aarch64_print_operand in some way. The solution that I propose here is similar to the solution that I used for the i386 disassembler. Currently, aarch64_print_operand uses snprintf to write the operand text into a buffer provided by the caller. What I propose is that we pass an extra argument to the aarch64_print_operand function, this argument will be a structure, the structure contains a callback function and some state. When aarch64_print_operand needs to format part of its output this can be done by using the callback function within the new structure, this callback returns a string with special embedded markers that indicate which mode should be used for each piece of text. Back in aarch64-dis.c we can spot these special style markers and use this to split the disassembler output up and apply the correct style to each piece. To make aarch64-opc.c clearer a series of new static functions have been added, e.g. 'style_reg', 'style_imm', etc. Each of these functions formats a piece of text in a different style, 'register' and 'immediate' in this case. Here's an example taken from aarch64-opc.c of the new functions in use: snprintf (buf, size, "[%s, %s]!", style_reg (styler, base), style_imm (styler, "#%d", opnd->addr.offset.imm)); The aarch64_print_operand function is also called from the assembler to aid in printing diagnostic messages. Right now I have no plans to add styling to the assembler output, and so, the callback function used in the assembler ignores the styling information and just returns an plain string. I've used the source files in gas/testsuite/gas/aarch64/ for testing, and have manually gone through and checked that the styling looks reasonable, however, I'm not an AArch64 expert, so it is possible that the odd piece is styled incorrectly. Please point out any mistakes I've made. With objdump disassembler color turned off, there should be no change in the output after this commit.
2022-04-28 20:31:07 +08:00
snprintf (buf, size, "%s", style_sub_mnem (styler, "csync"));
[AArch64, Binutils] Add missing TSB instruction This patch implements the TSB instructions: https://developer.arm.com/docs/ddi0596/f/base-instructions-alphabetic-order/ tsb-csync-trace-synchronization-barrier Since TSB and PSB both use the same (and only) argument "CSYNC", this patch reuses it for TSB. However, the same argument would imply different value for CRm:Op2 which are anyway fixed values, so I have diverted the inserter/extracter function to dummy versions instead of the "hint" version. The operand checker part still uses the existing infratructure for AARCH64_OPND_BARRIER_PSB to make sure the operand is parsed correctly. gas/ChangeLog: 2020-04-20 Sudakshina Das <sudi.das@arm.com> * config/tc-aarch64.c (parse_barrier_psb): Update error messages to include TSB. * testsuite/gas/aarch64/system-2.d: Update -march and new tsb tests. * testsuite/gas/aarch64/system-2.s: Add new tsb tests. * testsuite/gas/aarch64/system.d: Update. opcodes/ChangeLog: 2020-04-20 Sudakshina Das <sudi.das@arm.com> * aarch64-asm.c (aarch64_ins_none): New. * aarch64-asm.h (ins_none): New declaration. * aarch64-dis.c (aarch64_ext_none): New. * aarch64-dis.h (ext_none): New declaration. * aarch64-opc.c (aarch64_print_operand): Update case for AARCH64_OPND_BARRIER_PSB. * aarch64-tbl.h (aarch64_opcode_table): Add tsb. (AARCH64_OPERANDS): Update inserter/extracter for AARCH64_OPND_BARRIER_PSB to use new dummy functions. * aarch64-asm-2.c: Regenerated. * aarch64-dis-2.c: Regenerated. * aarch64-opc-2.c: Regenerated.
2020-04-20 17:58:16 +08:00
break;
case AARCH64_OPND_X16:
snprintf (buf, size, "%s", style_reg (styler, "x16"));
break;
case AARCH64_OPND_SME_ZT0:
snprintf (buf, size, "%s", style_reg (styler, "zt0"));
break;
case AARCH64_OPND_SME_ZT0_INDEX:
snprintf (buf, size, "%s[%s]", style_reg (styler, "zt0"),
style_imm (styler, "%d", (int) opnd->imm.value));
break;
case AARCH64_OPND_SME_ZT0_LIST:
snprintf (buf, size, "{%s}", style_reg (styler, "zt0"));
break;
case AARCH64_OPND_BARRIER_GCSB:
snprintf (buf, size, "%s", style_sub_mnem (styler, "dsync"));
break;
[PATCH, BINUTILS, AARCH64, 7/9] Add BTI instruction This patch is part of the patch series to add support for ARMv8.5-A extensions. (https://developer.arm.com/products/architecture/cpu-architecture/a-profile/docs/ddi0596/a/a64-base-instructions-alphabetic-order/bti-branch-target-identification) The Branch Target Identification instructions (BTI) are allocated to existing HINT space, using HINT numbers 32, 34, 36, 38, such that bits[7:6] of the instruction identify the compatibility of the BTI instruction to different branches. BTI {<targets>} where <targets> one of the following, specifying which type of indirection is allowed: j : Can be a target of any BR Xn isntruction. c : Can be a target of any BLR Xn and BR {X16|X17}. jc: Can be a target of any free branch. A BTI instruction without any <targets> is the strictest of all and can not be a target of nay free branch. *** include/ChangeLog *** 2018-10-09 Sudakshina Das <sudi.das@arm.com> * opcode/aarch64.h (AARCH64_FEATURE_BTI): New. (AARCH64_ARCH_V8_5): Add AARCH64_FEATURE_BTI by default. (aarch64_opnd): Add AARCH64_OPND_BTI_TARGET. (HINT_OPD_CSYNC, HINT_OPD_C, HINT_OPD_J): New macros to define HINT #imm values. (HINT_OPD_JC, HINT_OPD_NULL): Likewise. *** opcodes/ChangeLog *** 2018-10-09 Sudakshina Das <sudi.das@arm.com> * aarch64-opc.h (HINT_OPD_NOPRINT, HINT_ENCODE): New. (HINT_FLAG, HINT_VALUE): New macros to encode NO_PRINT flag with the hint immediate. * aarch64-opc.c (aarch64_hint_options): New entries for c, j, jc and default (with HINT_OPD_F_NOPRINT flag) for BTI. (aarch64_print_operand): Add case for AARCH64_OPND_BTI_TARGET while checking for HINT_OPD_F_NOPRINT flag. * aarch64-dis.c (aarch64_ext_hint): Use new HINT_VALUE to extract value. * aarch64-tbl.h (aarch64_feature_bti, BTI, BTI_INSN): New. (aarch64_opcode_table): Add entry for BTI. (AARCH64_OPERANDS): Add new description for BTI targets. * aarch64-asm-2.c: Regenerate. * aarch64-dis-2.c: Regenerate. * aarch64-opc-2.c: Regenerate. *** gas/ChangeLog *** 2018-10-09 Sudakshina Das <sudi.das@arm.com> * config/tc-aarch64.c (parse_bti_operand): New. (process_omitted_operand): Add case for AARCH64_OPND_BTI_TARGET. (parse_operands): Likewise. * testsuite/gas/aarch64/system.d: Update for BTI. * testsuite/gas/aarch64/bti.s: New. * testsuite/gas/aarch64/bti.d: New. * testsuite/gas/aarch64/illegal-bti.d: New. * testsuite/gas/aarch64/illegal-bti.l: New.
2018-09-26 18:00:49 +08:00
case AARCH64_OPND_BTI_TARGET:
if ((HINT_FLAG (opnd->hint_option->value) & HINT_OPD_F_NOPRINT) == 0)
libopcodes/aarch64: add support for disassembler styling This commit enables disassembler styling for AArch64. After this commit it is possible to have objdump style AArch64 disassembler output (using --disassembler-color option). Once the required GDB patches are merged, GDB will also style the disassembler output. The changes to support styling are mostly split between two files opcodes/aarch64-dis.c and opcodes/aarch64-opc.c. The entry point for the AArch64 disassembler can be found in aarch64-dis.c, this file handles printing the instruction mnemonics, and assembler directives (e.g. '.byte', '.word', etc). Some operands, mostly relating to assembler directives are also printed from this file. This commit changes all of this to pass through suitable styling information. However, for most "normal" instructions, the instruction operands are printed using a two step process. From aarch64-dis.c, in the print_operands function, the function aarch64_print_operand is called, this function is in aarch64-opc.c, and converts an instruction operand into a string. Then, back in print_operands (aarch64-dis.c), the operand string is printed. Unfortunately, the string returned by aarch64_print_operand can be quite complex, it will include syntax elements, like '[' and ']', in addition to register names and immediate values. In some cases, a single operand will expand into what will appear (to the user) as multiple operands separated with a ','. This makes the task of styling more complex, all these different components need to by styled differently, so we need to get the styling information out of aarch64_print_operand in some way. The solution that I propose here is similar to the solution that I used for the i386 disassembler. Currently, aarch64_print_operand uses snprintf to write the operand text into a buffer provided by the caller. What I propose is that we pass an extra argument to the aarch64_print_operand function, this argument will be a structure, the structure contains a callback function and some state. When aarch64_print_operand needs to format part of its output this can be done by using the callback function within the new structure, this callback returns a string with special embedded markers that indicate which mode should be used for each piece of text. Back in aarch64-dis.c we can spot these special style markers and use this to split the disassembler output up and apply the correct style to each piece. To make aarch64-opc.c clearer a series of new static functions have been added, e.g. 'style_reg', 'style_imm', etc. Each of these functions formats a piece of text in a different style, 'register' and 'immediate' in this case. Here's an example taken from aarch64-opc.c of the new functions in use: snprintf (buf, size, "[%s, %s]!", style_reg (styler, base), style_imm (styler, "#%d", opnd->addr.offset.imm)); The aarch64_print_operand function is also called from the assembler to aid in printing diagnostic messages. Right now I have no plans to add styling to the assembler output, and so, the callback function used in the assembler ignores the styling information and just returns an plain string. I've used the source files in gas/testsuite/gas/aarch64/ for testing, and have manually gone through and checked that the styling looks reasonable, however, I'm not an AArch64 expert, so it is possible that the odd piece is styled incorrectly. Please point out any mistakes I've made. With objdump disassembler color turned off, there should be no change in the output after this commit.
2022-04-28 20:31:07 +08:00
snprintf (buf, size, "%s",
style_sub_mnem (styler, opnd->hint_option->name));
[AArch64][Patch 5/5] Add instruction PSB CSYNC The Statistical Profile Extension adds the instruction PSB CSYNC as an alias for the HINT #17 instruction. This patch adds the instruction to binutils as a HINT alias that takes an operand. A new operand type, AARCH64_OPND_BARRIER_PSB, is added to represent the operand to PSB. A parser for the operand type is added to the assembler and a printer to the disassembler. The operand name "csync" is added to the list of HINT options with HINT number #17. Encoding and decoding of the operand is handled by the ins_hint/ext_hint functions added in the preceding patches. gas/ 2015-12-11 Matthew Wahab <matthew.wahab@arm.com> * config/tc-aarch64.c (aarch64_hint_opt_hsh): New. (parse_barrier_psb): New. (parse_operands): Add case for AARCH64_OPND_BARRIER_PSB. (md_begin): Set up aarch64_hint_opt_hsh. gas/testsuite/ 2015-12-11 Matthew Wahab <matthew.wahab@arm.com> * gas/aarch64/system-2.d: Enable the statistical profiling extension. Update the expected output. * gas/aarch64/system-2.s: Add tests for PSB CSYNC. * gas/aarch64/system.d: Update the expected output. include/opcode/ 2015-12-11 Matthew Wahab <matthew.wahab@arm.com> * aarch64.h (aarch64_opnd): Add AARCH64_OPND_BARRIER_PSB. * aarch64-asm-2.c: Regenerate. * aarch64-dis-2.c: Regenerate. * aarch64-opc-2.c: Regenerate. * aarch64-opc.c (aarch64_hint_options): Add "csync". (aarch64_print_operands): Handle AARCH64_OPND_BARRIER_PSB. * aarch64-tbl.h (aarch64_feature_stat_profile): New. (STAT_PROFILE): New. (aarch64_opcode_table): Add "psb". (AARCH64_OPERANDS): Add "BARRIER_PSB". Change-Id: I5ffb672d26a8b15b48785478d359350a9b70ca09
2015-12-11 18:22:40 +08:00
break;
aarch64: Add support for +mops This patch adds support for FEAT_MOPS, an Armv8.8-A extension that provides memcpy and memset acceleration instructions. I took the perhaps controversial decision to generate the individual instruction forms using macros rather than list them out individually. This becomes useful with a follow-on patch to check that code follows the correct P/M/E sequence. [https://developer.arm.com/documentation/ddi0596/2021-09/Base-Instructions?lang=en] include/ * opcode/aarch64.h (AARCH64_FEATURE_MOPS): New macro. (AARCH64_ARCH_V8_8): Make armv8.8-a imply AARCH64_FEATURE_MOPS. (AARCH64_OPND_MOPS_ADDR_Rd): New aarch64_opnd. (AARCH64_OPND_MOPS_ADDR_Rs): Likewise. (AARCH64_OPND_MOPS_WB_Rn): Likewise. opcodes/ * aarch64-asm.h (ins_x0_to_x30): New inserter. * aarch64-asm.c (aarch64_ins_x0_to_x30): New function. * aarch64-dis.h (ext_x0_to_x30): New extractor. * aarch64-dis.c (aarch64_ext_x0_to_x30): New function. * aarch64-tbl.h (aarch64_feature_mops): New feature set. (aarch64_feature_mops_memtag): Likewise. (MOPS, MOPS_MEMTAG, MOPS_INSN, MOPS_MEMTAG_INSN) (MOPS_CPY_OP1_OP2_PME_INSN, MOPS_CPY_OP1_OP2_INSN, MOPS_CPY_OP1_INSN) (MOPS_CPY_INSN, MOPS_SET_OP1_OP2_PME_INSN, MOPS_SET_OP1_OP2_INSN) (MOPS_SET_INSN): New macros. (aarch64_opcode_table): Add MOPS instructions. (aarch64_opcode_table): Add entries for AARCH64_OPND_MOPS_ADDR_Rd, AARCH64_OPND_MOPS_ADDR_Rs and AARCH64_OPND_MOPS_WB_Rn. * aarch64-opc.c (aarch64_print_operand): Handle AARCH64_OPND_MOPS_ADDR_Rd, AARCH64_OPND_MOPS_ADDR_Rs and AARCH64_OPND_MOPS_WB_Rn. (verify_three_different_regs): New function. * aarch64-asm-2.c: Regenerate. * aarch64-dis-2.c: Likewise. * aarch64-opc-2.c: Likewise. gas/ * doc/c-aarch64.texi: Document +mops. * config/tc-aarch64.c (parse_x0_to_x30): New function. (parse_operands): Handle AARCH64_OPND_MOPS_ADDR_Rd, AARCH64_OPND_MOPS_ADDR_Rs and AARCH64_OPND_MOPS_WB_Rn. (aarch64_features): Add "mops". * testsuite/gas/aarch64/mops.s, testsuite/gas/aarch64/mops.d: New test. * testsuite/gas/aarch64/mops_invalid.s, * testsuite/gas/aarch64/mops_invalid.d, * testsuite/gas/aarch64/mops_invalid.l: Likewise.
2021-12-02 23:00:57 +08:00
case AARCH64_OPND_MOPS_ADDR_Rd:
case AARCH64_OPND_MOPS_ADDR_Rs:
snprintf (buf, size, "[%s]!",
libopcodes/aarch64: add support for disassembler styling This commit enables disassembler styling for AArch64. After this commit it is possible to have objdump style AArch64 disassembler output (using --disassembler-color option). Once the required GDB patches are merged, GDB will also style the disassembler output. The changes to support styling are mostly split between two files opcodes/aarch64-dis.c and opcodes/aarch64-opc.c. The entry point for the AArch64 disassembler can be found in aarch64-dis.c, this file handles printing the instruction mnemonics, and assembler directives (e.g. '.byte', '.word', etc). Some operands, mostly relating to assembler directives are also printed from this file. This commit changes all of this to pass through suitable styling information. However, for most "normal" instructions, the instruction operands are printed using a two step process. From aarch64-dis.c, in the print_operands function, the function aarch64_print_operand is called, this function is in aarch64-opc.c, and converts an instruction operand into a string. Then, back in print_operands (aarch64-dis.c), the operand string is printed. Unfortunately, the string returned by aarch64_print_operand can be quite complex, it will include syntax elements, like '[' and ']', in addition to register names and immediate values. In some cases, a single operand will expand into what will appear (to the user) as multiple operands separated with a ','. This makes the task of styling more complex, all these different components need to by styled differently, so we need to get the styling information out of aarch64_print_operand in some way. The solution that I propose here is similar to the solution that I used for the i386 disassembler. Currently, aarch64_print_operand uses snprintf to write the operand text into a buffer provided by the caller. What I propose is that we pass an extra argument to the aarch64_print_operand function, this argument will be a structure, the structure contains a callback function and some state. When aarch64_print_operand needs to format part of its output this can be done by using the callback function within the new structure, this callback returns a string with special embedded markers that indicate which mode should be used for each piece of text. Back in aarch64-dis.c we can spot these special style markers and use this to split the disassembler output up and apply the correct style to each piece. To make aarch64-opc.c clearer a series of new static functions have been added, e.g. 'style_reg', 'style_imm', etc. Each of these functions formats a piece of text in a different style, 'register' and 'immediate' in this case. Here's an example taken from aarch64-opc.c of the new functions in use: snprintf (buf, size, "[%s, %s]!", style_reg (styler, base), style_imm (styler, "#%d", opnd->addr.offset.imm)); The aarch64_print_operand function is also called from the assembler to aid in printing diagnostic messages. Right now I have no plans to add styling to the assembler output, and so, the callback function used in the assembler ignores the styling information and just returns an plain string. I've used the source files in gas/testsuite/gas/aarch64/ for testing, and have manually gone through and checked that the styling looks reasonable, however, I'm not an AArch64 expert, so it is possible that the odd piece is styled incorrectly. Please point out any mistakes I've made. With objdump disassembler color turned off, there should be no change in the output after this commit.
2022-04-28 20:31:07 +08:00
style_reg (styler,
get_int_reg_name (opnd->reg.regno,
AARCH64_OPND_QLF_X, 0)));
aarch64: Add support for +mops This patch adds support for FEAT_MOPS, an Armv8.8-A extension that provides memcpy and memset acceleration instructions. I took the perhaps controversial decision to generate the individual instruction forms using macros rather than list them out individually. This becomes useful with a follow-on patch to check that code follows the correct P/M/E sequence. [https://developer.arm.com/documentation/ddi0596/2021-09/Base-Instructions?lang=en] include/ * opcode/aarch64.h (AARCH64_FEATURE_MOPS): New macro. (AARCH64_ARCH_V8_8): Make armv8.8-a imply AARCH64_FEATURE_MOPS. (AARCH64_OPND_MOPS_ADDR_Rd): New aarch64_opnd. (AARCH64_OPND_MOPS_ADDR_Rs): Likewise. (AARCH64_OPND_MOPS_WB_Rn): Likewise. opcodes/ * aarch64-asm.h (ins_x0_to_x30): New inserter. * aarch64-asm.c (aarch64_ins_x0_to_x30): New function. * aarch64-dis.h (ext_x0_to_x30): New extractor. * aarch64-dis.c (aarch64_ext_x0_to_x30): New function. * aarch64-tbl.h (aarch64_feature_mops): New feature set. (aarch64_feature_mops_memtag): Likewise. (MOPS, MOPS_MEMTAG, MOPS_INSN, MOPS_MEMTAG_INSN) (MOPS_CPY_OP1_OP2_PME_INSN, MOPS_CPY_OP1_OP2_INSN, MOPS_CPY_OP1_INSN) (MOPS_CPY_INSN, MOPS_SET_OP1_OP2_PME_INSN, MOPS_SET_OP1_OP2_INSN) (MOPS_SET_INSN): New macros. (aarch64_opcode_table): Add MOPS instructions. (aarch64_opcode_table): Add entries for AARCH64_OPND_MOPS_ADDR_Rd, AARCH64_OPND_MOPS_ADDR_Rs and AARCH64_OPND_MOPS_WB_Rn. * aarch64-opc.c (aarch64_print_operand): Handle AARCH64_OPND_MOPS_ADDR_Rd, AARCH64_OPND_MOPS_ADDR_Rs and AARCH64_OPND_MOPS_WB_Rn. (verify_three_different_regs): New function. * aarch64-asm-2.c: Regenerate. * aarch64-dis-2.c: Likewise. * aarch64-opc-2.c: Likewise. gas/ * doc/c-aarch64.texi: Document +mops. * config/tc-aarch64.c (parse_x0_to_x30): New function. (parse_operands): Handle AARCH64_OPND_MOPS_ADDR_Rd, AARCH64_OPND_MOPS_ADDR_Rs and AARCH64_OPND_MOPS_WB_Rn. (aarch64_features): Add "mops". * testsuite/gas/aarch64/mops.s, testsuite/gas/aarch64/mops.d: New test. * testsuite/gas/aarch64/mops_invalid.s, * testsuite/gas/aarch64/mops_invalid.d, * testsuite/gas/aarch64/mops_invalid.l: Likewise.
2021-12-02 23:00:57 +08:00
break;
case AARCH64_OPND_MOPS_WB_Rn:
snprintf (buf, size, "%s!",
libopcodes/aarch64: add support for disassembler styling This commit enables disassembler styling for AArch64. After this commit it is possible to have objdump style AArch64 disassembler output (using --disassembler-color option). Once the required GDB patches are merged, GDB will also style the disassembler output. The changes to support styling are mostly split between two files opcodes/aarch64-dis.c and opcodes/aarch64-opc.c. The entry point for the AArch64 disassembler can be found in aarch64-dis.c, this file handles printing the instruction mnemonics, and assembler directives (e.g. '.byte', '.word', etc). Some operands, mostly relating to assembler directives are also printed from this file. This commit changes all of this to pass through suitable styling information. However, for most "normal" instructions, the instruction operands are printed using a two step process. From aarch64-dis.c, in the print_operands function, the function aarch64_print_operand is called, this function is in aarch64-opc.c, and converts an instruction operand into a string. Then, back in print_operands (aarch64-dis.c), the operand string is printed. Unfortunately, the string returned by aarch64_print_operand can be quite complex, it will include syntax elements, like '[' and ']', in addition to register names and immediate values. In some cases, a single operand will expand into what will appear (to the user) as multiple operands separated with a ','. This makes the task of styling more complex, all these different components need to by styled differently, so we need to get the styling information out of aarch64_print_operand in some way. The solution that I propose here is similar to the solution that I used for the i386 disassembler. Currently, aarch64_print_operand uses snprintf to write the operand text into a buffer provided by the caller. What I propose is that we pass an extra argument to the aarch64_print_operand function, this argument will be a structure, the structure contains a callback function and some state. When aarch64_print_operand needs to format part of its output this can be done by using the callback function within the new structure, this callback returns a string with special embedded markers that indicate which mode should be used for each piece of text. Back in aarch64-dis.c we can spot these special style markers and use this to split the disassembler output up and apply the correct style to each piece. To make aarch64-opc.c clearer a series of new static functions have been added, e.g. 'style_reg', 'style_imm', etc. Each of these functions formats a piece of text in a different style, 'register' and 'immediate' in this case. Here's an example taken from aarch64-opc.c of the new functions in use: snprintf (buf, size, "[%s, %s]!", style_reg (styler, base), style_imm (styler, "#%d", opnd->addr.offset.imm)); The aarch64_print_operand function is also called from the assembler to aid in printing diagnostic messages. Right now I have no plans to add styling to the assembler output, and so, the callback function used in the assembler ignores the styling information and just returns an plain string. I've used the source files in gas/testsuite/gas/aarch64/ for testing, and have manually gone through and checked that the styling looks reasonable, however, I'm not an AArch64 expert, so it is possible that the odd piece is styled incorrectly. Please point out any mistakes I've made. With objdump disassembler color turned off, there should be no change in the output after this commit.
2022-04-28 20:31:07 +08:00
style_reg (styler, get_int_reg_name (opnd->reg.regno,
AARCH64_OPND_QLF_X, 0)));
aarch64: Add support for +mops This patch adds support for FEAT_MOPS, an Armv8.8-A extension that provides memcpy and memset acceleration instructions. I took the perhaps controversial decision to generate the individual instruction forms using macros rather than list them out individually. This becomes useful with a follow-on patch to check that code follows the correct P/M/E sequence. [https://developer.arm.com/documentation/ddi0596/2021-09/Base-Instructions?lang=en] include/ * opcode/aarch64.h (AARCH64_FEATURE_MOPS): New macro. (AARCH64_ARCH_V8_8): Make armv8.8-a imply AARCH64_FEATURE_MOPS. (AARCH64_OPND_MOPS_ADDR_Rd): New aarch64_opnd. (AARCH64_OPND_MOPS_ADDR_Rs): Likewise. (AARCH64_OPND_MOPS_WB_Rn): Likewise. opcodes/ * aarch64-asm.h (ins_x0_to_x30): New inserter. * aarch64-asm.c (aarch64_ins_x0_to_x30): New function. * aarch64-dis.h (ext_x0_to_x30): New extractor. * aarch64-dis.c (aarch64_ext_x0_to_x30): New function. * aarch64-tbl.h (aarch64_feature_mops): New feature set. (aarch64_feature_mops_memtag): Likewise. (MOPS, MOPS_MEMTAG, MOPS_INSN, MOPS_MEMTAG_INSN) (MOPS_CPY_OP1_OP2_PME_INSN, MOPS_CPY_OP1_OP2_INSN, MOPS_CPY_OP1_INSN) (MOPS_CPY_INSN, MOPS_SET_OP1_OP2_PME_INSN, MOPS_SET_OP1_OP2_INSN) (MOPS_SET_INSN): New macros. (aarch64_opcode_table): Add MOPS instructions. (aarch64_opcode_table): Add entries for AARCH64_OPND_MOPS_ADDR_Rd, AARCH64_OPND_MOPS_ADDR_Rs and AARCH64_OPND_MOPS_WB_Rn. * aarch64-opc.c (aarch64_print_operand): Handle AARCH64_OPND_MOPS_ADDR_Rd, AARCH64_OPND_MOPS_ADDR_Rs and AARCH64_OPND_MOPS_WB_Rn. (verify_three_different_regs): New function. * aarch64-asm-2.c: Regenerate. * aarch64-dis-2.c: Likewise. * aarch64-opc-2.c: Likewise. gas/ * doc/c-aarch64.texi: Document +mops. * config/tc-aarch64.c (parse_x0_to_x30): New function. (parse_operands): Handle AARCH64_OPND_MOPS_ADDR_Rd, AARCH64_OPND_MOPS_ADDR_Rs and AARCH64_OPND_MOPS_WB_Rn. (aarch64_features): Add "mops". * testsuite/gas/aarch64/mops.s, testsuite/gas/aarch64/mops.d: New test. * testsuite/gas/aarch64/mops_invalid.s, * testsuite/gas/aarch64/mops_invalid.d, * testsuite/gas/aarch64/mops_invalid.l: Likewise.
2021-12-02 23:00:57 +08:00
break;
default:
snprintf (buf, size, "<invalid>");
break;
}
}
#define CPENC(op0,op1,crn,crm,op2) \
((((op0) << 19) | ((op1) << 16) | ((crn) << 12) | ((crm) << 8) | ((op2) << 5)) >> 5)
/* for 3.9.3 Instructions for Accessing Special Purpose Registers */
#define CPEN_(op1,crm,op2) CPENC(3,(op1),4,(crm),(op2))
/* for 3.9.10 System Instructions */
#define CPENS(op1,crn,crm,op2) CPENC(1,(op1),(crn),(crm),(op2))
#define C0 0
#define C1 1
#define C2 2
#define C3 3
#define C4 4
#define C5 5
#define C6 6
#define C7 7
#define C8 8
#define C9 9
#define C10 10
#define C11 11
#define C12 12
#define C13 13
#define C14 14
#define C15 15
Implement Read/Write constraints on system registers on AArch64 This patch adds constraints for read and write only system registers with the msr and mrs instructions. The code will treat having both flags set and none set as the same. These flags add constraints that must be matched up. e.g. a system register with a READ only flag set, can only be used with mrs. If The constraint fails a warning is emitted. Examples of the warnings generated: test.s: Assembler messages: test.s:5: Warning: specified register cannot be written to at operand 1 -- `msr dbgdtrrx_el0,x3' test.s:7: Warning: specified register cannot be read from at operand 2 -- `mrs x3,dbgdtrtx_el0' test.s:8: Warning: specified register cannot be written to at operand 1 -- `msr midr_el1,x3' and disassembly notes: 0000000000000000 <main>: 0: d5130503 msr dbgdtrtx_el0, x3 4: d5130503 msr dbgdtrtx_el0, x3 8: d5330503 mrs x3, dbgdtrrx_el0 c: d5330503 mrs x3, dbgdtrrx_el0 10: d5180003 msr midr_el1, x3 ; note: writing to a read-only register. Note that because dbgdtrrx_el0 and dbgdtrtx_el0 have the same encoding, during disassembly the constraints are use to disambiguate between the two. An exact constraint match is always prefered over partial ones if available. As always the warnings can be suppressed with -w and also be made errors using warnings as errors. binutils/ PR binutils/21446 * doc/binutils.texi (-M): Document AArch64 options. gas/ PR binutils/21446 * testsuite/gas/aarch64/illegal-sysreg-2.s: Fix pmbidr_el1 test. * testsuite/gas/aarch64/illegal-sysreg-2.l: Likewise. * testsuite/gas/aarch64/illegal-sysreg-2.d: Likewise. * testsuite/gas/aarch64/sysreg-diagnostic.s: New. * testsuite/gas/aarch64/sysreg-diagnostic.l: New. * testsuite/gas/aarch64/sysreg-diagnostic.d: New. include/ PR binutils/21446 * opcode/aarch64.h (F_SYS_READ, F_SYS_WRITE): New. opcodes/ PR binutils/21446 * aarch64-asm.c (opintl.h): Include. (aarch64_ins_sysreg): Enforce read/write constraints. * aarch64-dis.c (aarch64_ext_sysreg): Likewise. * aarch64-opc.h (F_DEPRECATED, F_ARCHEXT, F_HASXT): Moved here. (F_REG_READ, F_REG_WRITE): New. * aarch64-opc.c (aarch64_print_operand): Generate notes for AARCH64_OPND_SYSREG. (F_DEPRECATED, F_ARCHEXT, F_HASXT): Move to aarch64-opc.h. (aarch64_sys_regs): Add constraints to currentel, midr_el1, ctr_el0, mpidr_el1, revidr_el1, aidr_el1, dczid_el0, id_dfr0_el1, id_pfr0_el1, id_pfr1_el1, id_afr0_el1, id_mmfr0_el1, id_mmfr1_el1, id_mmfr2_el1, id_mmfr3_el1, id_mmfr4_el1, id_isar0_el1, id_isar1_el1, id_isar2_el1, id_isar3_el1, id_isar4_el1, id_isar5_el1, mvfr0_el1, mvfr1_el1, mvfr2_el1, ccsidr_el1, id_aa64pfr0_el1, id_aa64pfr1_el1, id_aa64dfr0_el1, id_aa64dfr1_el1, id_aa64isar0_el1, id_aa64isar1_el1, id_aa64mmfr0_el1, id_aa64mmfr1_el1, id_aa64mmfr2_el1, id_aa64afr0_el1, id_aa64afr0_el1, id_aa64afr1_el1, id_aa64zfr0_el1, clidr_el1, csselr_el1, vsesr_el2, erridr_el1, erxfr_el1, rvbar_el1, rvbar_el2, rvbar_el3, isr_el1, tpidrro_el0, cntfrq_el0, cntpct_el0, cntvct_el0, mdccsr_el0, dbgdtrrx_el0, dbgdtrtx_el0, osdtrrx_el1, osdtrtx_el1, mdrar_el1, oslar_el1, oslsr_el1, dbgauthstatus_el1, pmbidr_el1, pmsidr_el1, pmswinc_el0, pmceid0_el0, pmceid1_el0. * aarch64-tbl.h (aarch64_opcode_table): Add constraints to msr (F_SYS_WRITE), mrs (F_SYS_READ).
2018-05-15 23:37:20 +08:00
/* TODO there is one more issues need to be resolved
[PATCH]: aarch64: Refactor representation of system registers Prior to this patch, the information describing the AArch64 system registers was separate from the information describing which system registers are available depending on the CPU feature set. Indeed, the latter was implemented as a separate function from the main table with the system register information. This patch remedies this situation and puts the feature information into the system register table itself. This has several advantages: * Having all the information described in one place is easier to maintain. * The logic to check whether a system register is supported now becomes trivial (and much more efficient). Since this patch ended up touching every line of the system register table, I took the opportunity to make the formatting more consistent and remove some redundant comments. Note that there is still more refactoring that could be done along the same lines here (e.g. with the TLB instructions) but this seemed like a reasonable first pass. Testing: * Regression tested an x64 -> aarch64-none-elf cross binutils. * Built aarch64-none-elf cross toolchain, checked newlib startup code still works. * Bootstrapped binutils on aarch64-linux-gnu, regression tested. * Built aarch64 kernel using new binutils with allyesconfig. OK for master? If so, I'll need a maintainer to commit on my behalf since I don't have write access. Thanks, Alex --- include/ChangeLog: 2020-06-11 Alex Coplan <alex.coplan@arm.com> * opcode/aarch64.h (aarch64_sys_reg): Add required features to struct describing system registers. opcodes/ChangeLog: 2020-06-11 Alex Coplan <alex.coplan@arm.com> * aarch64-opc.c (SYSREG): New macro for describing system registers. (SR_CORE): Likewise. (SR_FEAT): Likewise. (SR_RNG): Likewise. (SR_V8_1): Likewise. (SR_V8_2): Likewise. (SR_V8_3): Likewise. (SR_V8_4): Likewise. (SR_PAN): Likewise. (SR_RAS): Likewise. (SR_SSBS): Likewise. (SR_SVE): Likewise. (SR_ID_PFR2): Likewise. (SR_PROFILE): Likewise. (SR_MEMTAG): Likewise. (SR_SCXTNUM): Likewise. (aarch64_sys_regs): Refactor to store feature information in the table. (aarch64_sys_reg_supported_p): Collapse logic for system registers that now describe their own features. (aarch64_pstatefield_supported_p): Likewise.
2020-06-11 19:34:37 +08:00
1. handle cpu-implementation-defined system registers.
Note that the F_REG_{READ,WRITE} flags mean read-only and write-only
respectively. If neither of these are set then the register is read-write. */
const aarch64_sys_reg aarch64_sys_regs [] =
{
#define SYSREG(name, encoding, flags, features) \
{ name, encoding, flags, features },
#include "aarch64-sys-regs.def"
{ 0, CPENC (0,0,0,0,0), 0, AARCH64_NO_FEATURES }
#undef SYSREG
};
bool
aarch64_sys_reg_deprecated_p (const uint32_t reg_flags)
{
return (reg_flags & F_DEPRECATED) != 0;
}
bool
aarch64_sys_reg_128bit_p (const uint32_t reg_flags)
{
return (reg_flags & F_REG_128) != 0;
}
bool
aarch64_sys_reg_alias_p (const uint32_t reg_flags)
{
return (reg_flags & F_REG_ALIAS) != 0;
}
/* The CPENC below is fairly misleading, the fields
here are not in CPENC form. They are in op2op1 form. The fields are encoded
by ins_pstatefield, which just shifts the value by the width of the fields
in a loop. So if you CPENC them only the first value will be set, the rest
are masked out to 0. As an example. op2 = 3, op1=2. CPENC would produce a
value of 0b110000000001000000 (0x30040) while what you want is
0b011010 (0x1a). */
const aarch64_sys_reg aarch64_pstatefields [] =
{
{ "spsel", 0x05, F_REG_MAX_VALUE (1), AARCH64_NO_FEATURES },
{ "daifset", 0x1e, F_REG_MAX_VALUE (15), AARCH64_NO_FEATURES },
{ "daifclr", 0x1f, F_REG_MAX_VALUE (15), AARCH64_NO_FEATURES },
{ "pan", 0x04, F_REG_MAX_VALUE (1) | F_ARCHEXT, AARCH64_FEATURE (PAN) },
{ "uao", 0x03, F_REG_MAX_VALUE (1) | F_ARCHEXT, AARCH64_FEATURE (V8_2A) },
{ "ssbs", 0x19, F_REG_MAX_VALUE (1) | F_ARCHEXT, AARCH64_FEATURE (SSBS) },
{ "dit", 0x1a, F_REG_MAX_VALUE (1) | F_ARCHEXT, AARCH64_FEATURE (V8_4A) },
{ "tco", 0x1c, F_REG_MAX_VALUE (1) | F_ARCHEXT, AARCH64_FEATURE (MEMTAG) },
{ "svcrsm", 0x1b, PSTATE_ENCODE_CRM_AND_IMM (0x2,0x1) | F_REG_MAX_VALUE (1)
| F_ARCHEXT, AARCH64_FEATURE (SME) },
{ "svcrza", 0x1b, PSTATE_ENCODE_CRM_AND_IMM (0x4,0x1) | F_REG_MAX_VALUE (1)
| F_ARCHEXT, AARCH64_FEATURE (SME) },
{ "svcrsmza", 0x1b, PSTATE_ENCODE_CRM_AND_IMM (0x6,0x1) | F_REG_MAX_VALUE (1)
| F_ARCHEXT, AARCH64_FEATURE (SME) },
{ "allint", 0x08, F_REG_MAX_VALUE (1) | F_ARCHEXT, AARCH64_FEATURE (V8_8A) },
{ 0, CPENC (0,0,0,0,0), 0, AARCH64_NO_FEATURES },
};
bool
aarch64_pstatefield_supported_p (const aarch64_feature_set features,
const aarch64_sys_reg *reg)
{
if (!(reg->flags & F_ARCHEXT))
return true;
[PATCH]: aarch64: Refactor representation of system registers Prior to this patch, the information describing the AArch64 system registers was separate from the information describing which system registers are available depending on the CPU feature set. Indeed, the latter was implemented as a separate function from the main table with the system register information. This patch remedies this situation and puts the feature information into the system register table itself. This has several advantages: * Having all the information described in one place is easier to maintain. * The logic to check whether a system register is supported now becomes trivial (and much more efficient). Since this patch ended up touching every line of the system register table, I took the opportunity to make the formatting more consistent and remove some redundant comments. Note that there is still more refactoring that could be done along the same lines here (e.g. with the TLB instructions) but this seemed like a reasonable first pass. Testing: * Regression tested an x64 -> aarch64-none-elf cross binutils. * Built aarch64-none-elf cross toolchain, checked newlib startup code still works. * Bootstrapped binutils on aarch64-linux-gnu, regression tested. * Built aarch64 kernel using new binutils with allyesconfig. OK for master? If so, I'll need a maintainer to commit on my behalf since I don't have write access. Thanks, Alex --- include/ChangeLog: 2020-06-11 Alex Coplan <alex.coplan@arm.com> * opcode/aarch64.h (aarch64_sys_reg): Add required features to struct describing system registers. opcodes/ChangeLog: 2020-06-11 Alex Coplan <alex.coplan@arm.com> * aarch64-opc.c (SYSREG): New macro for describing system registers. (SR_CORE): Likewise. (SR_FEAT): Likewise. (SR_RNG): Likewise. (SR_V8_1): Likewise. (SR_V8_2): Likewise. (SR_V8_3): Likewise. (SR_V8_4): Likewise. (SR_PAN): Likewise. (SR_RAS): Likewise. (SR_SSBS): Likewise. (SR_SVE): Likewise. (SR_ID_PFR2): Likewise. (SR_PROFILE): Likewise. (SR_MEMTAG): Likewise. (SR_SCXTNUM): Likewise. (aarch64_sys_regs): Refactor to store feature information in the table. (aarch64_sys_reg_supported_p): Collapse logic for system registers that now describe their own features. (aarch64_pstatefield_supported_p): Likewise.
2020-06-11 19:34:37 +08:00
return AARCH64_CPU_HAS_ALL_FEATURES (features, reg->features);
}
const aarch64_sys_ins_reg aarch64_sys_regs_ic[] =
{
{ "ialluis", CPENS(0,C7,C1,0), 0, AARCH64_NO_FEATURES },
{ "iallu", CPENS(0,C7,C5,0), 0, AARCH64_NO_FEATURES },
{ "ivau", CPENS (3, C7, C5, 1), F_HASXT, AARCH64_NO_FEATURES },
{ 0, CPENS(0,0,0,0), 0, AARCH64_NO_FEATURES }
};
const aarch64_sys_ins_reg aarch64_sys_regs_dc[] =
{
{ "zva", CPENS (3, C7, C4, 1), F_HASXT, AARCH64_NO_FEATURES },
{ "gva", CPENS (3, C7, C4, 3), F_HASXT | F_ARCHEXT, AARCH64_FEATURE (MEMTAG) },
{ "gzva", CPENS (3, C7, C4, 4), F_HASXT | F_ARCHEXT, AARCH64_FEATURE (MEMTAG) },
{ "ivac", CPENS (0, C7, C6, 1), F_HASXT, AARCH64_NO_FEATURES },
{ "igvac", CPENS (0, C7, C6, 3), F_HASXT | F_ARCHEXT, AARCH64_FEATURE (MEMTAG) },
{ "igsw", CPENS (0, C7, C6, 4), F_HASXT | F_ARCHEXT, AARCH64_FEATURE (MEMTAG) },
{ "isw", CPENS (0, C7, C6, 2), F_HASXT, AARCH64_NO_FEATURES },
{ "igdvac", CPENS (0, C7, C6, 5), F_HASXT | F_ARCHEXT, AARCH64_FEATURE (MEMTAG) },
{ "igdsw", CPENS (0, C7, C6, 6), F_HASXT | F_ARCHEXT, AARCH64_FEATURE (MEMTAG) },
{ "cvac", CPENS (3, C7, C10, 1), F_HASXT, AARCH64_NO_FEATURES },
{ "cgvac", CPENS (3, C7, C10, 3), F_HASXT | F_ARCHEXT, AARCH64_FEATURE (MEMTAG) },
{ "cgdvac", CPENS (3, C7, C10, 5), F_HASXT | F_ARCHEXT, AARCH64_FEATURE (MEMTAG) },
{ "csw", CPENS (0, C7, C10, 2), F_HASXT, AARCH64_NO_FEATURES },
{ "cgsw", CPENS (0, C7, C10, 4), F_HASXT | F_ARCHEXT, AARCH64_FEATURE (MEMTAG) },
{ "cgdsw", CPENS (0, C7, C10, 6), F_HASXT | F_ARCHEXT, AARCH64_FEATURE (MEMTAG) },
{ "cvau", CPENS (3, C7, C11, 1), F_HASXT, AARCH64_NO_FEATURES },
{ "cvap", CPENS (3, C7, C12, 1), F_HASXT | F_ARCHEXT, AARCH64_FEATURE (V8_2A) },
{ "cgvap", CPENS (3, C7, C12, 3), F_HASXT | F_ARCHEXT, AARCH64_FEATURE (MEMTAG) },
{ "cgdvap", CPENS (3, C7, C12, 5), F_HASXT | F_ARCHEXT, AARCH64_FEATURE (MEMTAG) },
{ "cvadp", CPENS (3, C7, C13, 1), F_HASXT | F_ARCHEXT, AARCH64_FEATURE (CVADP) },
{ "cgvadp", CPENS (3, C7, C13, 3), F_HASXT | F_ARCHEXT, AARCH64_FEATURE (MEMTAG) },
{ "cgdvadp", CPENS (3, C7, C13, 5), F_HASXT | F_ARCHEXT, AARCH64_FEATURE (MEMTAG) },
{ "civac", CPENS (3, C7, C14, 1), F_HASXT, AARCH64_NO_FEATURES },
{ "cigvac", CPENS (3, C7, C14, 3), F_HASXT | F_ARCHEXT, AARCH64_FEATURE (MEMTAG) },
{ "cigdvac", CPENS (3, C7, C14, 5), F_HASXT | F_ARCHEXT, AARCH64_FEATURE (MEMTAG) },
{ "cisw", CPENS (0, C7, C14, 2), F_HASXT, AARCH64_NO_FEATURES },
{ "cigsw", CPENS (0, C7, C14, 4), F_HASXT | F_ARCHEXT, AARCH64_FEATURE (MEMTAG) },
{ "cigdsw", CPENS (0, C7, C14, 6), F_HASXT | F_ARCHEXT, AARCH64_FEATURE (MEMTAG) },
{ "cipapa", CPENS (6, C7, C14, 1), F_HASXT, AARCH64_NO_FEATURES },
{ "cigdpapa", CPENS (6, C7, C14, 5), F_HASXT, AARCH64_NO_FEATURES },
{ 0, CPENS(0,0,0,0), 0, AARCH64_NO_FEATURES }
};
const aarch64_sys_ins_reg aarch64_sys_regs_at[] =
{
{ "s1e1r", CPENS (0, C7, C8, 0), F_HASXT, AARCH64_NO_FEATURES },
{ "s1e1w", CPENS (0, C7, C8, 1), F_HASXT, AARCH64_NO_FEATURES },
{ "s1e0r", CPENS (0, C7, C8, 2), F_HASXT, AARCH64_NO_FEATURES },
{ "s1e0w", CPENS (0, C7, C8, 3), F_HASXT, AARCH64_NO_FEATURES },
{ "s12e1r", CPENS (4, C7, C8, 4), F_HASXT, AARCH64_NO_FEATURES },
{ "s12e1w", CPENS (4, C7, C8, 5), F_HASXT, AARCH64_NO_FEATURES },
{ "s12e0r", CPENS (4, C7, C8, 6), F_HASXT, AARCH64_NO_FEATURES },
{ "s12e0w", CPENS (4, C7, C8, 7), F_HASXT, AARCH64_NO_FEATURES },
{ "s1e2r", CPENS (4, C7, C8, 0), F_HASXT, AARCH64_NO_FEATURES },
{ "s1e2w", CPENS (4, C7, C8, 1), F_HASXT, AARCH64_NO_FEATURES },
{ "s1e3r", CPENS (6, C7, C8, 0), F_HASXT, AARCH64_NO_FEATURES },
{ "s1e3w", CPENS (6, C7, C8, 1), F_HASXT, AARCH64_NO_FEATURES },
{ "s1e1rp", CPENS (0, C7, C9, 0), F_HASXT | F_ARCHEXT, AARCH64_FEATURE (V8_2A) },
{ "s1e1wp", CPENS (0, C7, C9, 1), F_HASXT | F_ARCHEXT, AARCH64_FEATURE (V8_2A) },
{ "s1e1a", CPENS (0, C7, C9, 2), F_HASXT | F_ARCHEXT, AARCH64_FEATURE (ATS1A) },
{ "s1e2a", CPENS (4, C7, C9, 2), F_HASXT | F_ARCHEXT, AARCH64_FEATURE (ATS1A) },
{ "s1e3a", CPENS (6, C7, C9, 2), F_HASXT | F_ARCHEXT, AARCH64_FEATURE (ATS1A) },
{ 0, CPENS(0,0,0,0), 0, AARCH64_NO_FEATURES }
};
const aarch64_sys_ins_reg aarch64_sys_regs_tlbi[] =
{
{ "rpaos", CPENS (6, C8, C4, 3), F_HASXT, AARCH64_NO_FEATURES },
{ "rpalos", CPENS (6, C8, C4, 7), F_HASXT, AARCH64_NO_FEATURES },
{ "paallos", CPENS (6, C8, C1, 4), 0, AARCH64_NO_FEATURES },
{ "paall", CPENS (6, C8, C7, 4), 0, AARCH64_NO_FEATURES },
#define TLBI_XS_OP(OP, CODE, FLAGS) \
{ OP, CODE, FLAGS, AARCH64_NO_FEATURES }, \
{ OP "nxs", CODE | CPENS (0, C9, 0, 0), FLAGS | F_ARCHEXT, AARCH64_FEATURE (XS) },
TLBI_XS_OP ( "vmalle1", CPENS (0, C8, C7, 0), 0)
TLBI_XS_OP ( "vae1", CPENS (0, C8, C7, 1), F_HASXT | F_REG_128)
TLBI_XS_OP ( "aside1", CPENS (0, C8, C7, 2), F_HASXT )
TLBI_XS_OP ( "vaae1", CPENS (0, C8, C7, 3), F_HASXT | F_REG_128)
TLBI_XS_OP ( "vmalle1is", CPENS (0, C8, C3, 0), 0)
TLBI_XS_OP ( "vae1is", CPENS (0, C8, C3, 1), F_HASXT | F_REG_128)
TLBI_XS_OP ( "aside1is", CPENS (0, C8, C3, 2), F_HASXT )
TLBI_XS_OP ( "vaae1is", CPENS (0, C8, C3, 3), F_HASXT | F_REG_128)
TLBI_XS_OP ( "ipas2e1is", CPENS (4, C8, C0, 1), F_HASXT | F_REG_128)
TLBI_XS_OP ( "ipas2le1is",CPENS (4, C8, C0, 5), F_HASXT | F_REG_128)
TLBI_XS_OP ( "ipas2e1", CPENS (4, C8, C4, 1), F_HASXT | F_REG_128)
TLBI_XS_OP ( "ipas2le1", CPENS (4, C8, C4, 5), F_HASXT | F_REG_128)
TLBI_XS_OP ( "vae2", CPENS (4, C8, C7, 1), F_HASXT | F_REG_128)
TLBI_XS_OP ( "vae2is", CPENS (4, C8, C3, 1), F_HASXT | F_REG_128)
TLBI_XS_OP ( "vmalls12e1",CPENS (4, C8, C7, 6), 0)
TLBI_XS_OP ( "vmalls12e1is",CPENS(4,C8, C3, 6), 0)
TLBI_XS_OP ( "vae3", CPENS (6, C8, C7, 1), F_HASXT | F_REG_128)
TLBI_XS_OP ( "vae3is", CPENS (6, C8, C3, 1), F_HASXT | F_REG_128)
TLBI_XS_OP ( "alle2", CPENS (4, C8, C7, 0), 0)
TLBI_XS_OP ( "alle2is", CPENS (4, C8, C3, 0), 0)
TLBI_XS_OP ( "alle1", CPENS (4, C8, C7, 4), 0)
TLBI_XS_OP ( "alle1is", CPENS (4, C8, C3, 4), 0)
TLBI_XS_OP ( "alle3", CPENS (6, C8, C7, 0), 0)
TLBI_XS_OP ( "alle3is", CPENS (6, C8, C3, 0), 0)
TLBI_XS_OP ( "vale1is", CPENS (0, C8, C3, 5), F_HASXT | F_REG_128)
TLBI_XS_OP ( "vale2is", CPENS (4, C8, C3, 5), F_HASXT | F_REG_128)
TLBI_XS_OP ( "vale3is", CPENS (6, C8, C3, 5), F_HASXT | F_REG_128)
TLBI_XS_OP ( "vaale1is", CPENS (0, C8, C3, 7), F_HASXT | F_REG_128)
TLBI_XS_OP ( "vale1", CPENS (0, C8, C7, 5), F_HASXT | F_REG_128)
TLBI_XS_OP ( "vale2", CPENS (4, C8, C7, 5), F_HASXT | F_REG_128)
TLBI_XS_OP ( "vale3", CPENS (6, C8, C7, 5), F_HASXT | F_REG_128)
TLBI_XS_OP ( "vaale1", CPENS (0, C8, C7, 7), F_HASXT | F_REG_128)
#undef TLBI_XS_OP
#define TLBI_XS_OP(OP, CODE, FLAGS) \
{ OP, CODE, FLAGS | F_ARCHEXT, AARCH64_FEATURE (V8_4A) }, \
{ OP "nxs", CODE | CPENS (0, C9, 0, 0), FLAGS | F_ARCHEXT, AARCH64_FEATURE (XS) },
TLBI_XS_OP ( "vmalle1os", CPENS (0, C8, C1, 0), 0 )
TLBI_XS_OP ( "vae1os", CPENS (0, C8, C1, 1), F_HASXT | F_REG_128 )
TLBI_XS_OP ( "aside1os", CPENS (0, C8, C1, 2), F_HASXT )
TLBI_XS_OP ( "vaae1os", CPENS (0, C8, C1, 3), F_HASXT | F_REG_128 )
TLBI_XS_OP ( "vale1os", CPENS (0, C8, C1, 5), F_HASXT | F_REG_128 )
TLBI_XS_OP ( "vaale1os", CPENS (0, C8, C1, 7), F_HASXT | F_REG_128 )
TLBI_XS_OP ( "ipas2e1os", CPENS (4, C8, C4, 0), F_HASXT | F_REG_128 )
TLBI_XS_OP ( "ipas2le1os", CPENS (4, C8, C4, 4), F_HASXT | F_REG_128 )
TLBI_XS_OP ( "vae2os", CPENS (4, C8, C1, 1), F_HASXT | F_REG_128 )
TLBI_XS_OP ( "vale2os", CPENS (4, C8, C1, 5), F_HASXT | F_REG_128 )
TLBI_XS_OP ( "vmalls12e1os", CPENS (4, C8, C1, 6), 0 )
TLBI_XS_OP ( "vae3os", CPENS (6, C8, C1, 1), F_HASXT | F_REG_128 )
TLBI_XS_OP ( "vale3os", CPENS (6, C8, C1, 5), F_HASXT | F_REG_128 )
TLBI_XS_OP ( "alle2os", CPENS (4, C8, C1, 0), 0 )
TLBI_XS_OP ( "alle1os", CPENS (4, C8, C1, 4), 0 )
TLBI_XS_OP ( "alle3os", CPENS (6, C8, C1, 0), 0 )
TLBI_XS_OP ( "rvae1", CPENS (0, C8, C6, 1), F_HASXT | F_REG_128 )
TLBI_XS_OP ( "rvaae1", CPENS (0, C8, C6, 3), F_HASXT | F_REG_128 )
TLBI_XS_OP ( "rvale1", CPENS (0, C8, C6, 5), F_HASXT | F_REG_128 )
TLBI_XS_OP ( "rvaale1", CPENS (0, C8, C6, 7), F_HASXT | F_REG_128 )
TLBI_XS_OP ( "rvae1is", CPENS (0, C8, C2, 1), F_HASXT | F_REG_128 )
TLBI_XS_OP ( "rvaae1is", CPENS (0, C8, C2, 3), F_HASXT | F_REG_128 )
TLBI_XS_OP ( "rvale1is", CPENS (0, C8, C2, 5), F_HASXT | F_REG_128 )
TLBI_XS_OP ( "rvaale1is", CPENS (0, C8, C2, 7), F_HASXT | F_REG_128 )
TLBI_XS_OP ( "rvae1os", CPENS (0, C8, C5, 1), F_HASXT | F_REG_128 )
TLBI_XS_OP ( "rvaae1os", CPENS (0, C8, C5, 3), F_HASXT | F_REG_128 )
TLBI_XS_OP ( "rvale1os", CPENS (0, C8, C5, 5), F_HASXT | F_REG_128 )
TLBI_XS_OP ( "rvaale1os", CPENS (0, C8, C5, 7), F_HASXT | F_REG_128 )
TLBI_XS_OP ( "ripas2e1is", CPENS (4, C8, C0, 2), F_HASXT | F_REG_128 )
TLBI_XS_OP ( "ripas2le1is",CPENS (4, C8, C0, 6), F_HASXT | F_REG_128 )
TLBI_XS_OP ( "ripas2e1", CPENS (4, C8, C4, 2), F_HASXT | F_REG_128 )
TLBI_XS_OP ( "ripas2le1", CPENS (4, C8, C4, 6), F_HASXT | F_REG_128 )
TLBI_XS_OP ( "ripas2e1os", CPENS (4, C8, C4, 3), F_HASXT | F_REG_128 )
TLBI_XS_OP ( "ripas2le1os",CPENS (4, C8, C4, 7), F_HASXT | F_REG_128 )
TLBI_XS_OP ( "rvae2", CPENS (4, C8, C6, 1), F_HASXT | F_REG_128 )
TLBI_XS_OP ( "rvale2", CPENS (4, C8, C6, 5), F_HASXT | F_REG_128 )
TLBI_XS_OP ( "rvae2is", CPENS (4, C8, C2, 1), F_HASXT | F_REG_128 )
TLBI_XS_OP ( "rvale2is", CPENS (4, C8, C2, 5), F_HASXT | F_REG_128 )
TLBI_XS_OP ( "rvae2os", CPENS (4, C8, C5, 1), F_HASXT | F_REG_128 )
TLBI_XS_OP ( "rvale2os", CPENS (4, C8, C5, 5), F_HASXT | F_REG_128 )
TLBI_XS_OP ( "rvae3", CPENS (6, C8, C6, 1), F_HASXT | F_REG_128 )
TLBI_XS_OP ( "rvale3", CPENS (6, C8, C6, 5), F_HASXT | F_REG_128 )
TLBI_XS_OP ( "rvae3is", CPENS (6, C8, C2, 1), F_HASXT | F_REG_128 )
TLBI_XS_OP ( "rvale3is", CPENS (6, C8, C2, 5), F_HASXT | F_REG_128 )
TLBI_XS_OP ( "rvae3os", CPENS (6, C8, C5, 1), F_HASXT | F_REG_128 )
TLBI_XS_OP ( "rvale3os", CPENS (6, C8, C5, 5), F_HASXT | F_REG_128 )
#undef TLBI_XS_OP
{ 0, CPENS(0,0,0,0), 0, AARCH64_NO_FEATURES }
};
[PATCH, BINUTILS, AARCH64, 4/9] Add Execution and Data Restriction instructions This patch is part of the patch series to add support for ARMv8.5-A extensions. (https://developer.arm.com/products/architecture/cpu-architecture/a-profile/docs/ddi0596/a/a64-base-instructions-alphabetic-order) This patch adds the prediction restriction instructions (that is, cfp, dvp, cpp). These instructions are retrospectively made optional for all versions of the architecture from ARMv8.0 to ARMv8.4 and is mandatory from ARMv8.5. Hence adding a new +predres which can be used by the older architectures. *** include/ChangeLog *** 2018-10-09 Sudakshina Das <sudi.das@arm.com> * opcode/aarch64.h (AARCH64_FEATURE_PREDRES): New. (AARCH64_ARCH_V8_5): Add AARCH64_FEATURE_PREDRES by default. (aarch64_opnd): Add AARCH64_OPND_SYSREG_SR. (aarch64_sys_regs_sr): Declare new table. *** opcodes/ChangeLog *** 2018-10-09 Sudakshina Das <sudi.das@arm.com> * aarch64-dis.c (aarch64_ext_sysins_op): Add case for AARCH64_OPND_SYSREG_SR. * aarch64-opc.c (aarch64_print_operand): Likewise. (aarch64_sys_regs_sr): Define table. (aarch64_sys_ins_reg_supported_p): Check for RCTX with AARCH64_FEATURE_PREDRES. * aarch64-tbl.h (aarch64_feature_predres): New. (PREDRES, PREDRES_INSN): New. (aarch64_opcode_table): Add entries for cfp, dvp and cpp. (AARCH64_OPERANDS): Add new description for SYSREG_SR. * aarch64-asm-2.c: Regenerate. * aarch64-dis-2.c: Regenerate. * aarch64-opc-2.c: Regenerate. *** gas/ChangeLog *** 2018-10-09 Sudakshina Das <sudi.das@arm.com> * config/tc-aarch64.c (aarch64_sys_regs_sr_hsh): New. (parse_operands): Add entry for AARCH64_OPND_SYSREG_SR. (md_begin): Allocate and initialize aarch64_sys_regs_sr_hsh with aarch64_sys_regs_sr. (aarch64_features): Add new "predres" option for older architectures. * doc/c-aarch64.texi: Document the same. * testsuite/gas/aarch64/sysreg-4.s: New. * testsuite/gas/aarch64/sysreg-4.d: New. * testsuite/gas/aarch64/illegal-sysreg-4.d: New. * testsuite/gas/aarch64/illegal-sysreg-4.l: New. * testsuite/gas/aarch64/predres.s: New. * testsuite/gas/aarch64/predres.d: New.
2018-09-26 17:52:51 +08:00
const aarch64_sys_ins_reg aarch64_sys_regs_sr[] =
{
/* RCTX is somewhat unique in a way that it has different values
(op2) based on the instruction in which it is used (cfp/dvp/cpp).
Thus op2 is masked out and instead encoded directly in the
aarch64_opcode_table entries for the respective instructions. */
{ "rctx", CPENS(3,C7,C3,0), F_HASXT | F_ARCHEXT | F_REG_WRITE, AARCH64_FEATURE (PREDRES) }, /* WO */
{ 0, CPENS(0,0,0,0), 0, AARCH64_NO_FEATURES }
[PATCH, BINUTILS, AARCH64, 4/9] Add Execution and Data Restriction instructions This patch is part of the patch series to add support for ARMv8.5-A extensions. (https://developer.arm.com/products/architecture/cpu-architecture/a-profile/docs/ddi0596/a/a64-base-instructions-alphabetic-order) This patch adds the prediction restriction instructions (that is, cfp, dvp, cpp). These instructions are retrospectively made optional for all versions of the architecture from ARMv8.0 to ARMv8.4 and is mandatory from ARMv8.5. Hence adding a new +predres which can be used by the older architectures. *** include/ChangeLog *** 2018-10-09 Sudakshina Das <sudi.das@arm.com> * opcode/aarch64.h (AARCH64_FEATURE_PREDRES): New. (AARCH64_ARCH_V8_5): Add AARCH64_FEATURE_PREDRES by default. (aarch64_opnd): Add AARCH64_OPND_SYSREG_SR. (aarch64_sys_regs_sr): Declare new table. *** opcodes/ChangeLog *** 2018-10-09 Sudakshina Das <sudi.das@arm.com> * aarch64-dis.c (aarch64_ext_sysins_op): Add case for AARCH64_OPND_SYSREG_SR. * aarch64-opc.c (aarch64_print_operand): Likewise. (aarch64_sys_regs_sr): Define table. (aarch64_sys_ins_reg_supported_p): Check for RCTX with AARCH64_FEATURE_PREDRES. * aarch64-tbl.h (aarch64_feature_predres): New. (PREDRES, PREDRES_INSN): New. (aarch64_opcode_table): Add entries for cfp, dvp and cpp. (AARCH64_OPERANDS): Add new description for SYSREG_SR. * aarch64-asm-2.c: Regenerate. * aarch64-dis-2.c: Regenerate. * aarch64-opc-2.c: Regenerate. *** gas/ChangeLog *** 2018-10-09 Sudakshina Das <sudi.das@arm.com> * config/tc-aarch64.c (aarch64_sys_regs_sr_hsh): New. (parse_operands): Add entry for AARCH64_OPND_SYSREG_SR. (md_begin): Allocate and initialize aarch64_sys_regs_sr_hsh with aarch64_sys_regs_sr. (aarch64_features): Add new "predres" option for older architectures. * doc/c-aarch64.texi: Document the same. * testsuite/gas/aarch64/sysreg-4.s: New. * testsuite/gas/aarch64/sysreg-4.d: New. * testsuite/gas/aarch64/illegal-sysreg-4.d: New. * testsuite/gas/aarch64/illegal-sysreg-4.l: New. * testsuite/gas/aarch64/predres.s: New. * testsuite/gas/aarch64/predres.d: New.
2018-09-26 17:52:51 +08:00
};
bool
aarch64_sys_ins_reg_has_xt (const aarch64_sys_ins_reg *sys_ins_reg)
{
return (sys_ins_reg->flags & F_HASXT) != 0;
}
extern bool
aarch64_sys_ins_reg_supported_p (const aarch64_feature_set features,
const char *reg_name,
uint32_t reg_flags,
const aarch64_feature_set *reg_features)
{
aarch64: Add support for Armv8-R system registers This patch adds support for the system registers introduced in Armv8-R AArch64. gas/ChangeLog: 2020-09-08 Alex Coplan <alex.coplan@arm.com> * config/tc-aarch64.c (parse_sys_reg): Also pass sysreg name to validation function. (parse_sys_ins_reg): Likewise. (print_operands): Pass CPU features to aarch64_print_operand(). * testsuite/gas/aarch64/v8-r-bad-sysregs.d: New test. * testsuite/gas/aarch64/v8-r-bad-sysregs.l: Error output. * testsuite/gas/aarch64/v8-r-bad-sysregs.s: Input. * testsuite/gas/aarch64/v8-r-sysregs-need-arch.d: New test. * testsuite/gas/aarch64/v8-r-sysregs-need-arch.l: Error output. * testsuite/gas/aarch64/v8-r-sysregs.d: New test. * testsuite/gas/aarch64/v8-r-sysregs.s: Input for previous two tests. include/ChangeLog: 2020-09-08 Alex Coplan <alex.coplan@arm.com> * opcode/aarch64.h (aarch64_sys_ins_reg_supported_p): Also take system register name in order to simplify validation for v8-R. (aarch64_print_operand): Also take CPU feature set, as disassembly for system registers now depends on arch variant. opcodes/ChangeLog: 2020-09-08 Alex Coplan <alex.coplan@arm.com> * aarch64-dis.c (print_operands): Pass CPU features to aarch64_print_operand(). * aarch64-opc.c (aarch64_print_operand): Use CPU features to determine preferred disassembly of system registers. (SR_RNG): Refactor to use new SR_FEAT2 macro. (SR_FEAT2): New. (SR_V8_1_A): New. (SR_V8_4_A): New. (SR_V8_A): New. (SR_V8_R): New. (SR_EXPAND_ELx): New. (SR_EXPAND_EL12): New. (aarch64_sys_regs): Specify which registers are only on A-profile, add R-profile system registers. (ENC_BARLAR): New. (PRBARn_ELx): New. (PRLARn_ELx): New. (aarch64_sys_ins_reg_supported_p): Reject EL3 registers for Armv8-R AArch64.
2020-09-08 21:21:44 +08:00
/* Armv8-R has no EL3. */
if (AARCH64_CPU_HAS_FEATURE (features, V8R))
aarch64: Add support for Armv8-R system registers This patch adds support for the system registers introduced in Armv8-R AArch64. gas/ChangeLog: 2020-09-08 Alex Coplan <alex.coplan@arm.com> * config/tc-aarch64.c (parse_sys_reg): Also pass sysreg name to validation function. (parse_sys_ins_reg): Likewise. (print_operands): Pass CPU features to aarch64_print_operand(). * testsuite/gas/aarch64/v8-r-bad-sysregs.d: New test. * testsuite/gas/aarch64/v8-r-bad-sysregs.l: Error output. * testsuite/gas/aarch64/v8-r-bad-sysregs.s: Input. * testsuite/gas/aarch64/v8-r-sysregs-need-arch.d: New test. * testsuite/gas/aarch64/v8-r-sysregs-need-arch.l: Error output. * testsuite/gas/aarch64/v8-r-sysregs.d: New test. * testsuite/gas/aarch64/v8-r-sysregs.s: Input for previous two tests. include/ChangeLog: 2020-09-08 Alex Coplan <alex.coplan@arm.com> * opcode/aarch64.h (aarch64_sys_ins_reg_supported_p): Also take system register name in order to simplify validation for v8-R. (aarch64_print_operand): Also take CPU feature set, as disassembly for system registers now depends on arch variant. opcodes/ChangeLog: 2020-09-08 Alex Coplan <alex.coplan@arm.com> * aarch64-dis.c (print_operands): Pass CPU features to aarch64_print_operand(). * aarch64-opc.c (aarch64_print_operand): Use CPU features to determine preferred disassembly of system registers. (SR_RNG): Refactor to use new SR_FEAT2 macro. (SR_FEAT2): New. (SR_V8_1_A): New. (SR_V8_4_A): New. (SR_V8_A): New. (SR_V8_R): New. (SR_EXPAND_ELx): New. (SR_EXPAND_EL12): New. (aarch64_sys_regs): Specify which registers are only on A-profile, add R-profile system registers. (ENC_BARLAR): New. (PRBARn_ELx): New. (PRLARn_ELx): New. (aarch64_sys_ins_reg_supported_p): Reject EL3 registers for Armv8-R AArch64.
2020-09-08 21:21:44 +08:00
{
const char *suffix = strrchr (reg_name, '_');
if (suffix && !strcmp (suffix, "_el3"))
return false;
aarch64: Add support for Armv8-R system registers This patch adds support for the system registers introduced in Armv8-R AArch64. gas/ChangeLog: 2020-09-08 Alex Coplan <alex.coplan@arm.com> * config/tc-aarch64.c (parse_sys_reg): Also pass sysreg name to validation function. (parse_sys_ins_reg): Likewise. (print_operands): Pass CPU features to aarch64_print_operand(). * testsuite/gas/aarch64/v8-r-bad-sysregs.d: New test. * testsuite/gas/aarch64/v8-r-bad-sysregs.l: Error output. * testsuite/gas/aarch64/v8-r-bad-sysregs.s: Input. * testsuite/gas/aarch64/v8-r-sysregs-need-arch.d: New test. * testsuite/gas/aarch64/v8-r-sysregs-need-arch.l: Error output. * testsuite/gas/aarch64/v8-r-sysregs.d: New test. * testsuite/gas/aarch64/v8-r-sysregs.s: Input for previous two tests. include/ChangeLog: 2020-09-08 Alex Coplan <alex.coplan@arm.com> * opcode/aarch64.h (aarch64_sys_ins_reg_supported_p): Also take system register name in order to simplify validation for v8-R. (aarch64_print_operand): Also take CPU feature set, as disassembly for system registers now depends on arch variant. opcodes/ChangeLog: 2020-09-08 Alex Coplan <alex.coplan@arm.com> * aarch64-dis.c (print_operands): Pass CPU features to aarch64_print_operand(). * aarch64-opc.c (aarch64_print_operand): Use CPU features to determine preferred disassembly of system registers. (SR_RNG): Refactor to use new SR_FEAT2 macro. (SR_FEAT2): New. (SR_V8_1_A): New. (SR_V8_4_A): New. (SR_V8_A): New. (SR_V8_R): New. (SR_EXPAND_ELx): New. (SR_EXPAND_EL12): New. (aarch64_sys_regs): Specify which registers are only on A-profile, add R-profile system registers. (ENC_BARLAR): New. (PRBARn_ELx): New. (PRLARn_ELx): New. (aarch64_sys_ins_reg_supported_p): Reject EL3 registers for Armv8-R AArch64.
2020-09-08 21:21:44 +08:00
}
if (!(reg_flags & F_ARCHEXT))
return true;
return AARCH64_CPU_HAS_ALL_FEATURES (features, *reg_features);
}
#undef C0
#undef C1
#undef C2
#undef C3
#undef C4
#undef C5
#undef C6
#undef C7
#undef C8
#undef C9
#undef C10
#undef C11
#undef C12
#undef C13
#undef C14
#undef C15
#define BIT(INSN,BT) (((INSN) >> (BT)) & 1)
#define BITS(INSN,HI,LO) (((INSN) >> (LO)) & ((1 << (((HI) - (LO)) + 1)) - 1))
static enum err_type
verify_ldpsw (const struct aarch64_inst *inst ATTRIBUTE_UNUSED,
const aarch64_insn insn, bfd_vma pc ATTRIBUTE_UNUSED,
bool encoding ATTRIBUTE_UNUSED,
aarch64_operand_error *mismatch_detail ATTRIBUTE_UNUSED,
AArch64: Add SVE constraints verifier. This patch adds the verification rules for move prefix constraints. The Arm SVE instruction MOVPRFX introduces[1] constraints on the instruction at PC+4. Particularly the following constraints are handled by this patch * MOVPRFX must be followed by an instruction. * MOVPRFX can only be followed by non-layout altering directives. * MOVPRFX destination register MUST be used as the destination register in the instruction at PC+4, and is not allowed to be used in any other position other than destructive input. This includes registers that architecturally overlap. e.g. x1 should be treated as z1. * MOVPRFX must be followed by a restricted set of SVE instructions. * The size of the destination register of MOVPRFX must be equal to that of the operation at PC+4. * The predicate register and operation of MOVPRFX must match that of the instruction at PC+4 * The predicated instruction at PC+4 must use the merging predicate. * Architectural aliases and pseudo-instructions need to be supported as well. * MOVPRFX cannot be the last instruction in a sequence Any failure to adhere to any of these constrains will emit an assembly warning and a disassembly note. [1] https://developer.arm.com/docs/ddi0584/latest/arm-architecture-reference-manual-supplement-the-scalable-vector-extension-sve-for-armv8-a include/ * opcode/aarch64.h (aarch64_inst): Remove. (enum err_type): Add ERR_VFI. (aarch64_is_destructive_by_operands): New. (init_insn_sequence): New. (aarch64_decode_insn): Remove param name. opcodes/ * aarch64-opc.c (init_insn_block): New. (verify_constraints, aarch64_is_destructive_by_operands): New. * aarch64-opc.h (verify_constraints): New. gas/ * config/tc-aarch64.c (output_operand_error_report): Order warnings.
2018-10-04 01:38:42 +08:00
aarch64_instr_sequence *insn_sequence ATTRIBUTE_UNUSED)
{
int t = BITS (insn, 4, 0);
int n = BITS (insn, 9, 5);
int t2 = BITS (insn, 14, 10);
if (BIT (insn, 23))
{
/* Write back enabled. */
if ((t == n || t2 == n) && n != 31)
return ERR_UND;
}
if (BIT (insn, 22))
{
/* Load */
if (t == t2)
return ERR_UND;
}
return ERR_OK;
}
/* Verifier for vector by element 3 operands functions where the
conditions `if sz:L == 11 then UNDEFINED` holds. */
static enum err_type
verify_elem_sd (const struct aarch64_inst *inst, const aarch64_insn insn,
bfd_vma pc ATTRIBUTE_UNUSED, bool encoding,
aarch64_operand_error *mismatch_detail ATTRIBUTE_UNUSED,
aarch64_instr_sequence *insn_sequence ATTRIBUTE_UNUSED)
{
const aarch64_insn undef_pattern = 0x3;
aarch64_insn value;
assert (inst->opcode);
assert (inst->opcode->operands[2] == AARCH64_OPND_Em);
value = encoding ? inst->value : insn;
assert (value);
if (undef_pattern == extract_fields (value, 0, 2, FLD_sz, FLD_L))
return ERR_UND;
return ERR_OK;
}
aarch64: Add support for +mops This patch adds support for FEAT_MOPS, an Armv8.8-A extension that provides memcpy and memset acceleration instructions. I took the perhaps controversial decision to generate the individual instruction forms using macros rather than list them out individually. This becomes useful with a follow-on patch to check that code follows the correct P/M/E sequence. [https://developer.arm.com/documentation/ddi0596/2021-09/Base-Instructions?lang=en] include/ * opcode/aarch64.h (AARCH64_FEATURE_MOPS): New macro. (AARCH64_ARCH_V8_8): Make armv8.8-a imply AARCH64_FEATURE_MOPS. (AARCH64_OPND_MOPS_ADDR_Rd): New aarch64_opnd. (AARCH64_OPND_MOPS_ADDR_Rs): Likewise. (AARCH64_OPND_MOPS_WB_Rn): Likewise. opcodes/ * aarch64-asm.h (ins_x0_to_x30): New inserter. * aarch64-asm.c (aarch64_ins_x0_to_x30): New function. * aarch64-dis.h (ext_x0_to_x30): New extractor. * aarch64-dis.c (aarch64_ext_x0_to_x30): New function. * aarch64-tbl.h (aarch64_feature_mops): New feature set. (aarch64_feature_mops_memtag): Likewise. (MOPS, MOPS_MEMTAG, MOPS_INSN, MOPS_MEMTAG_INSN) (MOPS_CPY_OP1_OP2_PME_INSN, MOPS_CPY_OP1_OP2_INSN, MOPS_CPY_OP1_INSN) (MOPS_CPY_INSN, MOPS_SET_OP1_OP2_PME_INSN, MOPS_SET_OP1_OP2_INSN) (MOPS_SET_INSN): New macros. (aarch64_opcode_table): Add MOPS instructions. (aarch64_opcode_table): Add entries for AARCH64_OPND_MOPS_ADDR_Rd, AARCH64_OPND_MOPS_ADDR_Rs and AARCH64_OPND_MOPS_WB_Rn. * aarch64-opc.c (aarch64_print_operand): Handle AARCH64_OPND_MOPS_ADDR_Rd, AARCH64_OPND_MOPS_ADDR_Rs and AARCH64_OPND_MOPS_WB_Rn. (verify_three_different_regs): New function. * aarch64-asm-2.c: Regenerate. * aarch64-dis-2.c: Likewise. * aarch64-opc-2.c: Likewise. gas/ * doc/c-aarch64.texi: Document +mops. * config/tc-aarch64.c (parse_x0_to_x30): New function. (parse_operands): Handle AARCH64_OPND_MOPS_ADDR_Rd, AARCH64_OPND_MOPS_ADDR_Rs and AARCH64_OPND_MOPS_WB_Rn. (aarch64_features): Add "mops". * testsuite/gas/aarch64/mops.s, testsuite/gas/aarch64/mops.d: New test. * testsuite/gas/aarch64/mops_invalid.s, * testsuite/gas/aarch64/mops_invalid.d, * testsuite/gas/aarch64/mops_invalid.l: Likewise.
2021-12-02 23:00:57 +08:00
/* Check an instruction that takes three register operands and that
requires the register numbers to be distinct from one another. */
static enum err_type
verify_three_different_regs (const struct aarch64_inst *inst,
const aarch64_insn insn ATTRIBUTE_UNUSED,
bfd_vma pc ATTRIBUTE_UNUSED,
bool encoding ATTRIBUTE_UNUSED,
aarch64_operand_error *mismatch_detail
ATTRIBUTE_UNUSED,
aarch64_instr_sequence *insn_sequence
ATTRIBUTE_UNUSED)
{
int rd, rs, rn;
rd = inst->operands[0].reg.regno;
rs = inst->operands[1].reg.regno;
rn = inst->operands[2].reg.regno;
if (rd == rs || rd == rn || rs == rn)
{
mismatch_detail->kind = AARCH64_OPDE_SYNTAX_ERROR;
mismatch_detail->error
= _("the three register operands must be distinct from one another");
mismatch_detail->index = -1;
return ERR_UND;
}
return ERR_OK;
}
/* Add INST to the end of INSN_SEQUENCE. */
static void
add_insn_to_sequence (const struct aarch64_inst *inst,
aarch64_instr_sequence *insn_sequence)
{
insn_sequence->instr[insn_sequence->num_added_insns++] = *inst;
}
AArch64: Add SVE constraints verifier. This patch adds the verification rules for move prefix constraints. The Arm SVE instruction MOVPRFX introduces[1] constraints on the instruction at PC+4. Particularly the following constraints are handled by this patch * MOVPRFX must be followed by an instruction. * MOVPRFX can only be followed by non-layout altering directives. * MOVPRFX destination register MUST be used as the destination register in the instruction at PC+4, and is not allowed to be used in any other position other than destructive input. This includes registers that architecturally overlap. e.g. x1 should be treated as z1. * MOVPRFX must be followed by a restricted set of SVE instructions. * The size of the destination register of MOVPRFX must be equal to that of the operation at PC+4. * The predicate register and operation of MOVPRFX must match that of the instruction at PC+4 * The predicated instruction at PC+4 must use the merging predicate. * Architectural aliases and pseudo-instructions need to be supported as well. * MOVPRFX cannot be the last instruction in a sequence Any failure to adhere to any of these constrains will emit an assembly warning and a disassembly note. [1] https://developer.arm.com/docs/ddi0584/latest/arm-architecture-reference-manual-supplement-the-scalable-vector-extension-sve-for-armv8-a include/ * opcode/aarch64.h (aarch64_inst): Remove. (enum err_type): Add ERR_VFI. (aarch64_is_destructive_by_operands): New. (init_insn_sequence): New. (aarch64_decode_insn): Remove param name. opcodes/ * aarch64-opc.c (init_insn_block): New. (verify_constraints, aarch64_is_destructive_by_operands): New. * aarch64-opc.h (verify_constraints): New. gas/ * config/tc-aarch64.c (output_operand_error_report): Order warnings.
2018-10-04 01:38:42 +08:00
/* Initialize an instruction sequence insn_sequence with the instruction INST.
If INST is NULL the given insn_sequence is cleared and the sequence is left
uninitialized. */
void
init_insn_sequence (const struct aarch64_inst *inst,
aarch64_instr_sequence *insn_sequence)
{
int num_req_entries = 0;
if (insn_sequence->instr)
AArch64: Add SVE constraints verifier. This patch adds the verification rules for move prefix constraints. The Arm SVE instruction MOVPRFX introduces[1] constraints on the instruction at PC+4. Particularly the following constraints are handled by this patch * MOVPRFX must be followed by an instruction. * MOVPRFX can only be followed by non-layout altering directives. * MOVPRFX destination register MUST be used as the destination register in the instruction at PC+4, and is not allowed to be used in any other position other than destructive input. This includes registers that architecturally overlap. e.g. x1 should be treated as z1. * MOVPRFX must be followed by a restricted set of SVE instructions. * The size of the destination register of MOVPRFX must be equal to that of the operation at PC+4. * The predicate register and operation of MOVPRFX must match that of the instruction at PC+4 * The predicated instruction at PC+4 must use the merging predicate. * Architectural aliases and pseudo-instructions need to be supported as well. * MOVPRFX cannot be the last instruction in a sequence Any failure to adhere to any of these constrains will emit an assembly warning and a disassembly note. [1] https://developer.arm.com/docs/ddi0584/latest/arm-architecture-reference-manual-supplement-the-scalable-vector-extension-sve-for-armv8-a include/ * opcode/aarch64.h (aarch64_inst): Remove. (enum err_type): Add ERR_VFI. (aarch64_is_destructive_by_operands): New. (init_insn_sequence): New. (aarch64_decode_insn): Remove param name. opcodes/ * aarch64-opc.c (init_insn_block): New. (verify_constraints, aarch64_is_destructive_by_operands): New. * aarch64-opc.h (verify_constraints): New. gas/ * config/tc-aarch64.c (output_operand_error_report): Order warnings.
2018-10-04 01:38:42 +08:00
{
XDELETE (insn_sequence->instr);
insn_sequence->instr = NULL;
AArch64: Add SVE constraints verifier. This patch adds the verification rules for move prefix constraints. The Arm SVE instruction MOVPRFX introduces[1] constraints on the instruction at PC+4. Particularly the following constraints are handled by this patch * MOVPRFX must be followed by an instruction. * MOVPRFX can only be followed by non-layout altering directives. * MOVPRFX destination register MUST be used as the destination register in the instruction at PC+4, and is not allowed to be used in any other position other than destructive input. This includes registers that architecturally overlap. e.g. x1 should be treated as z1. * MOVPRFX must be followed by a restricted set of SVE instructions. * The size of the destination register of MOVPRFX must be equal to that of the operation at PC+4. * The predicate register and operation of MOVPRFX must match that of the instruction at PC+4 * The predicated instruction at PC+4 must use the merging predicate. * Architectural aliases and pseudo-instructions need to be supported as well. * MOVPRFX cannot be the last instruction in a sequence Any failure to adhere to any of these constrains will emit an assembly warning and a disassembly note. [1] https://developer.arm.com/docs/ddi0584/latest/arm-architecture-reference-manual-supplement-the-scalable-vector-extension-sve-for-armv8-a include/ * opcode/aarch64.h (aarch64_inst): Remove. (enum err_type): Add ERR_VFI. (aarch64_is_destructive_by_operands): New. (init_insn_sequence): New. (aarch64_decode_insn): Remove param name. opcodes/ * aarch64-opc.c (init_insn_block): New. (verify_constraints, aarch64_is_destructive_by_operands): New. * aarch64-opc.h (verify_constraints): New. gas/ * config/tc-aarch64.c (output_operand_error_report): Order warnings.
2018-10-04 01:38:42 +08:00
}
/* Handle all the cases here. May need to think of something smarter than
a giant if/else chain if this grows. At that time, a lookup table may be
best. */
if (inst && inst->opcode->constraints & C_SCAN_MOVPRFX)
num_req_entries = 1;
aarch64: Enforce P/M/E order for MOPS instructions The MOPS instructions should be used as a triple, such as: cpyfp [x0]!, [x1]!, x2! cpyfm [x0]!, [x1]!, x2! cpyfe [x0]!, [x1]!, x2! The registers should also be the same for each writeback operand. This patch adds a warning for code that doesn't follow this rule, along similar lines to the warning that we already emit for invalid uses of MOVPRFX. include/ * opcode/aarch64.h (C_SCAN_MOPS_P, C_SCAN_MOPS_M, C_SCAN_MOPS_E) (C_SCAN_MOPS_PME): New macros. (AARCH64_OPDE_A_SHOULD_FOLLOW_B): New aarch64_operand_error_kind. (AARCH64_OPDE_EXPECTED_A_AFTER_B): Likewise. (aarch64_operand_error): Make each data value a union between an int and a string. opcodes/ * aarch64-tbl.h (MOPS_CPY_OP1_OP2_INSN): Add scan flags. (MOPS_SET_OP1_OP2_INSN): Likewise. * aarch64-opc.c (set_out_of_range_error): Update after change to aarch64_operand_error. (set_unaligned_error, set_reg_list_error): Likewise. (init_insn_sequence): Use a 3-instruction sequence for MOPS P instructions. (verify_mops_pme_sequence): New function. (verify_constraints): Call it. * aarch64-dis.c (print_verifier_notes): Handle AARCH64_OPDE_A_SHOULD_FOLLOW_B and AARCH64_OPDE_EXPECTED_A_AFTER_B. gas/ * config/tc-aarch64.c (operand_mismatch_kind_names): Add entries for AARCH64_OPDE_A_SHOULD_FOLLOW_B and AARCH64_OPDE_EXPECTED_A_AFTER_B. (operand_error_higher_severity_p): Check that AARCH64_OPDE_A_SHOULD_FOLLOW_B and AARCH64_OPDE_EXPECTED_A_AFTER_B come between AARCH64_OPDE_RECOVERABLE and AARCH64_OPDE_SYNTAX_ERROR; their relative order is not significant. (record_operand_error_with_data): Update after change to aarch64_operand_error. (output_operand_error_record): Likewise. Handle AARCH64_OPDE_A_SHOULD_FOLLOW_B and AARCH64_OPDE_EXPECTED_A_AFTER_B. * testsuite/gas/aarch64/mops_invalid_2.s, testsuite/gas/aarch64/mops_invalid_2.d, testsuite/gas/aarch64/mops_invalid_2.l: New test.
2021-12-02 23:00:57 +08:00
if (inst && (inst->opcode->constraints & C_SCAN_MOPS_PME) == C_SCAN_MOPS_P)
num_req_entries = 2;
AArch64: Add SVE constraints verifier. This patch adds the verification rules for move prefix constraints. The Arm SVE instruction MOVPRFX introduces[1] constraints on the instruction at PC+4. Particularly the following constraints are handled by this patch * MOVPRFX must be followed by an instruction. * MOVPRFX can only be followed by non-layout altering directives. * MOVPRFX destination register MUST be used as the destination register in the instruction at PC+4, and is not allowed to be used in any other position other than destructive input. This includes registers that architecturally overlap. e.g. x1 should be treated as z1. * MOVPRFX must be followed by a restricted set of SVE instructions. * The size of the destination register of MOVPRFX must be equal to that of the operation at PC+4. * The predicate register and operation of MOVPRFX must match that of the instruction at PC+4 * The predicated instruction at PC+4 must use the merging predicate. * Architectural aliases and pseudo-instructions need to be supported as well. * MOVPRFX cannot be the last instruction in a sequence Any failure to adhere to any of these constrains will emit an assembly warning and a disassembly note. [1] https://developer.arm.com/docs/ddi0584/latest/arm-architecture-reference-manual-supplement-the-scalable-vector-extension-sve-for-armv8-a include/ * opcode/aarch64.h (aarch64_inst): Remove. (enum err_type): Add ERR_VFI. (aarch64_is_destructive_by_operands): New. (init_insn_sequence): New. (aarch64_decode_insn): Remove param name. opcodes/ * aarch64-opc.c (init_insn_block): New. (verify_constraints, aarch64_is_destructive_by_operands): New. * aarch64-opc.h (verify_constraints): New. gas/ * config/tc-aarch64.c (output_operand_error_report): Order warnings.
2018-10-04 01:38:42 +08:00
insn_sequence->num_added_insns = 0;
insn_sequence->num_allocated_insns = num_req_entries;
AArch64: Add SVE constraints verifier. This patch adds the verification rules for move prefix constraints. The Arm SVE instruction MOVPRFX introduces[1] constraints on the instruction at PC+4. Particularly the following constraints are handled by this patch * MOVPRFX must be followed by an instruction. * MOVPRFX can only be followed by non-layout altering directives. * MOVPRFX destination register MUST be used as the destination register in the instruction at PC+4, and is not allowed to be used in any other position other than destructive input. This includes registers that architecturally overlap. e.g. x1 should be treated as z1. * MOVPRFX must be followed by a restricted set of SVE instructions. * The size of the destination register of MOVPRFX must be equal to that of the operation at PC+4. * The predicate register and operation of MOVPRFX must match that of the instruction at PC+4 * The predicated instruction at PC+4 must use the merging predicate. * Architectural aliases and pseudo-instructions need to be supported as well. * MOVPRFX cannot be the last instruction in a sequence Any failure to adhere to any of these constrains will emit an assembly warning and a disassembly note. [1] https://developer.arm.com/docs/ddi0584/latest/arm-architecture-reference-manual-supplement-the-scalable-vector-extension-sve-for-armv8-a include/ * opcode/aarch64.h (aarch64_inst): Remove. (enum err_type): Add ERR_VFI. (aarch64_is_destructive_by_operands): New. (init_insn_sequence): New. (aarch64_decode_insn): Remove param name. opcodes/ * aarch64-opc.c (init_insn_block): New. (verify_constraints, aarch64_is_destructive_by_operands): New. * aarch64-opc.h (verify_constraints): New. gas/ * config/tc-aarch64.c (output_operand_error_report): Order warnings.
2018-10-04 01:38:42 +08:00
if (num_req_entries != 0)
{
insn_sequence->instr = XCNEWVEC (aarch64_inst, num_req_entries);
add_insn_to_sequence (inst, insn_sequence);
AArch64: Add SVE constraints verifier. This patch adds the verification rules for move prefix constraints. The Arm SVE instruction MOVPRFX introduces[1] constraints on the instruction at PC+4. Particularly the following constraints are handled by this patch * MOVPRFX must be followed by an instruction. * MOVPRFX can only be followed by non-layout altering directives. * MOVPRFX destination register MUST be used as the destination register in the instruction at PC+4, and is not allowed to be used in any other position other than destructive input. This includes registers that architecturally overlap. e.g. x1 should be treated as z1. * MOVPRFX must be followed by a restricted set of SVE instructions. * The size of the destination register of MOVPRFX must be equal to that of the operation at PC+4. * The predicate register and operation of MOVPRFX must match that of the instruction at PC+4 * The predicated instruction at PC+4 must use the merging predicate. * Architectural aliases and pseudo-instructions need to be supported as well. * MOVPRFX cannot be the last instruction in a sequence Any failure to adhere to any of these constrains will emit an assembly warning and a disassembly note. [1] https://developer.arm.com/docs/ddi0584/latest/arm-architecture-reference-manual-supplement-the-scalable-vector-extension-sve-for-armv8-a include/ * opcode/aarch64.h (aarch64_inst): Remove. (enum err_type): Add ERR_VFI. (aarch64_is_destructive_by_operands): New. (init_insn_sequence): New. (aarch64_decode_insn): Remove param name. opcodes/ * aarch64-opc.c (init_insn_block): New. (verify_constraints, aarch64_is_destructive_by_operands): New. * aarch64-opc.h (verify_constraints): New. gas/ * config/tc-aarch64.c (output_operand_error_report): Order warnings.
2018-10-04 01:38:42 +08:00
}
}
aarch64: Enforce P/M/E order for MOPS instructions The MOPS instructions should be used as a triple, such as: cpyfp [x0]!, [x1]!, x2! cpyfm [x0]!, [x1]!, x2! cpyfe [x0]!, [x1]!, x2! The registers should also be the same for each writeback operand. This patch adds a warning for code that doesn't follow this rule, along similar lines to the warning that we already emit for invalid uses of MOVPRFX. include/ * opcode/aarch64.h (C_SCAN_MOPS_P, C_SCAN_MOPS_M, C_SCAN_MOPS_E) (C_SCAN_MOPS_PME): New macros. (AARCH64_OPDE_A_SHOULD_FOLLOW_B): New aarch64_operand_error_kind. (AARCH64_OPDE_EXPECTED_A_AFTER_B): Likewise. (aarch64_operand_error): Make each data value a union between an int and a string. opcodes/ * aarch64-tbl.h (MOPS_CPY_OP1_OP2_INSN): Add scan flags. (MOPS_SET_OP1_OP2_INSN): Likewise. * aarch64-opc.c (set_out_of_range_error): Update after change to aarch64_operand_error. (set_unaligned_error, set_reg_list_error): Likewise. (init_insn_sequence): Use a 3-instruction sequence for MOPS P instructions. (verify_mops_pme_sequence): New function. (verify_constraints): Call it. * aarch64-dis.c (print_verifier_notes): Handle AARCH64_OPDE_A_SHOULD_FOLLOW_B and AARCH64_OPDE_EXPECTED_A_AFTER_B. gas/ * config/tc-aarch64.c (operand_mismatch_kind_names): Add entries for AARCH64_OPDE_A_SHOULD_FOLLOW_B and AARCH64_OPDE_EXPECTED_A_AFTER_B. (operand_error_higher_severity_p): Check that AARCH64_OPDE_A_SHOULD_FOLLOW_B and AARCH64_OPDE_EXPECTED_A_AFTER_B come between AARCH64_OPDE_RECOVERABLE and AARCH64_OPDE_SYNTAX_ERROR; their relative order is not significant. (record_operand_error_with_data): Update after change to aarch64_operand_error. (output_operand_error_record): Likewise. Handle AARCH64_OPDE_A_SHOULD_FOLLOW_B and AARCH64_OPDE_EXPECTED_A_AFTER_B. * testsuite/gas/aarch64/mops_invalid_2.s, testsuite/gas/aarch64/mops_invalid_2.d, testsuite/gas/aarch64/mops_invalid_2.l: New test.
2021-12-02 23:00:57 +08:00
/* Subroutine of verify_constraints. Check whether the instruction
is part of a MOPS P/M/E sequence and, if so, whether sequencing
expectations are met. Return true if the check passes, otherwise
describe the problem in MISMATCH_DETAIL.
IS_NEW_SECTION is true if INST is assumed to start a new section.
The other arguments are as for verify_constraints. */
static bool
verify_mops_pme_sequence (const struct aarch64_inst *inst,
bool is_new_section,
aarch64_operand_error *mismatch_detail,
aarch64_instr_sequence *insn_sequence)
{
const struct aarch64_opcode *opcode;
const struct aarch64_inst *prev_insn;
int i;
opcode = inst->opcode;
if (insn_sequence->instr)
prev_insn = insn_sequence->instr + (insn_sequence->num_added_insns - 1);
else
prev_insn = NULL;
if (prev_insn
&& (prev_insn->opcode->constraints & C_SCAN_MOPS_PME)
&& prev_insn->opcode != opcode - 1)
{
mismatch_detail->kind = AARCH64_OPDE_EXPECTED_A_AFTER_B;
mismatch_detail->error = NULL;
aarch64: Enforce P/M/E order for MOPS instructions The MOPS instructions should be used as a triple, such as: cpyfp [x0]!, [x1]!, x2! cpyfm [x0]!, [x1]!, x2! cpyfe [x0]!, [x1]!, x2! The registers should also be the same for each writeback operand. This patch adds a warning for code that doesn't follow this rule, along similar lines to the warning that we already emit for invalid uses of MOVPRFX. include/ * opcode/aarch64.h (C_SCAN_MOPS_P, C_SCAN_MOPS_M, C_SCAN_MOPS_E) (C_SCAN_MOPS_PME): New macros. (AARCH64_OPDE_A_SHOULD_FOLLOW_B): New aarch64_operand_error_kind. (AARCH64_OPDE_EXPECTED_A_AFTER_B): Likewise. (aarch64_operand_error): Make each data value a union between an int and a string. opcodes/ * aarch64-tbl.h (MOPS_CPY_OP1_OP2_INSN): Add scan flags. (MOPS_SET_OP1_OP2_INSN): Likewise. * aarch64-opc.c (set_out_of_range_error): Update after change to aarch64_operand_error. (set_unaligned_error, set_reg_list_error): Likewise. (init_insn_sequence): Use a 3-instruction sequence for MOPS P instructions. (verify_mops_pme_sequence): New function. (verify_constraints): Call it. * aarch64-dis.c (print_verifier_notes): Handle AARCH64_OPDE_A_SHOULD_FOLLOW_B and AARCH64_OPDE_EXPECTED_A_AFTER_B. gas/ * config/tc-aarch64.c (operand_mismatch_kind_names): Add entries for AARCH64_OPDE_A_SHOULD_FOLLOW_B and AARCH64_OPDE_EXPECTED_A_AFTER_B. (operand_error_higher_severity_p): Check that AARCH64_OPDE_A_SHOULD_FOLLOW_B and AARCH64_OPDE_EXPECTED_A_AFTER_B come between AARCH64_OPDE_RECOVERABLE and AARCH64_OPDE_SYNTAX_ERROR; their relative order is not significant. (record_operand_error_with_data): Update after change to aarch64_operand_error. (output_operand_error_record): Likewise. Handle AARCH64_OPDE_A_SHOULD_FOLLOW_B and AARCH64_OPDE_EXPECTED_A_AFTER_B. * testsuite/gas/aarch64/mops_invalid_2.s, testsuite/gas/aarch64/mops_invalid_2.d, testsuite/gas/aarch64/mops_invalid_2.l: New test.
2021-12-02 23:00:57 +08:00
mismatch_detail->index = -1;
mismatch_detail->data[0].s = prev_insn->opcode[1].name;
mismatch_detail->data[1].s = prev_insn->opcode->name;
mismatch_detail->non_fatal = true;
return false;
}
if (opcode->constraints & C_SCAN_MOPS_PME)
{
if (is_new_section || !prev_insn || prev_insn->opcode != opcode - 1)
{
mismatch_detail->kind = AARCH64_OPDE_A_SHOULD_FOLLOW_B;
mismatch_detail->error = NULL;
aarch64: Enforce P/M/E order for MOPS instructions The MOPS instructions should be used as a triple, such as: cpyfp [x0]!, [x1]!, x2! cpyfm [x0]!, [x1]!, x2! cpyfe [x0]!, [x1]!, x2! The registers should also be the same for each writeback operand. This patch adds a warning for code that doesn't follow this rule, along similar lines to the warning that we already emit for invalid uses of MOVPRFX. include/ * opcode/aarch64.h (C_SCAN_MOPS_P, C_SCAN_MOPS_M, C_SCAN_MOPS_E) (C_SCAN_MOPS_PME): New macros. (AARCH64_OPDE_A_SHOULD_FOLLOW_B): New aarch64_operand_error_kind. (AARCH64_OPDE_EXPECTED_A_AFTER_B): Likewise. (aarch64_operand_error): Make each data value a union between an int and a string. opcodes/ * aarch64-tbl.h (MOPS_CPY_OP1_OP2_INSN): Add scan flags. (MOPS_SET_OP1_OP2_INSN): Likewise. * aarch64-opc.c (set_out_of_range_error): Update after change to aarch64_operand_error. (set_unaligned_error, set_reg_list_error): Likewise. (init_insn_sequence): Use a 3-instruction sequence for MOPS P instructions. (verify_mops_pme_sequence): New function. (verify_constraints): Call it. * aarch64-dis.c (print_verifier_notes): Handle AARCH64_OPDE_A_SHOULD_FOLLOW_B and AARCH64_OPDE_EXPECTED_A_AFTER_B. gas/ * config/tc-aarch64.c (operand_mismatch_kind_names): Add entries for AARCH64_OPDE_A_SHOULD_FOLLOW_B and AARCH64_OPDE_EXPECTED_A_AFTER_B. (operand_error_higher_severity_p): Check that AARCH64_OPDE_A_SHOULD_FOLLOW_B and AARCH64_OPDE_EXPECTED_A_AFTER_B come between AARCH64_OPDE_RECOVERABLE and AARCH64_OPDE_SYNTAX_ERROR; their relative order is not significant. (record_operand_error_with_data): Update after change to aarch64_operand_error. (output_operand_error_record): Likewise. Handle AARCH64_OPDE_A_SHOULD_FOLLOW_B and AARCH64_OPDE_EXPECTED_A_AFTER_B. * testsuite/gas/aarch64/mops_invalid_2.s, testsuite/gas/aarch64/mops_invalid_2.d, testsuite/gas/aarch64/mops_invalid_2.l: New test.
2021-12-02 23:00:57 +08:00
mismatch_detail->index = -1;
mismatch_detail->data[0].s = opcode->name;
mismatch_detail->data[1].s = opcode[-1].name;
mismatch_detail->non_fatal = true;
return false;
}
for (i = 0; i < 3; ++i)
/* There's no specific requirement for the data register to be
the same between consecutive SET* instructions. */
if ((opcode->operands[i] == AARCH64_OPND_MOPS_ADDR_Rd
|| opcode->operands[i] == AARCH64_OPND_MOPS_ADDR_Rs
|| opcode->operands[i] == AARCH64_OPND_MOPS_WB_Rn)
&& prev_insn->operands[i].reg.regno != inst->operands[i].reg.regno)
{
mismatch_detail->kind = AARCH64_OPDE_SYNTAX_ERROR;
if (opcode->operands[i] == AARCH64_OPND_MOPS_ADDR_Rd)
mismatch_detail->error = _("destination register differs from "
"preceding instruction");
else if (opcode->operands[i] == AARCH64_OPND_MOPS_ADDR_Rs)
mismatch_detail->error = _("source register differs from "
"preceding instruction");
else
mismatch_detail->error = _("size register differs from "
"preceding instruction");
mismatch_detail->index = i;
mismatch_detail->non_fatal = true;
return false;
}
}
return true;
}
AArch64: Add SVE constraints verifier. This patch adds the verification rules for move prefix constraints. The Arm SVE instruction MOVPRFX introduces[1] constraints on the instruction at PC+4. Particularly the following constraints are handled by this patch * MOVPRFX must be followed by an instruction. * MOVPRFX can only be followed by non-layout altering directives. * MOVPRFX destination register MUST be used as the destination register in the instruction at PC+4, and is not allowed to be used in any other position other than destructive input. This includes registers that architecturally overlap. e.g. x1 should be treated as z1. * MOVPRFX must be followed by a restricted set of SVE instructions. * The size of the destination register of MOVPRFX must be equal to that of the operation at PC+4. * The predicate register and operation of MOVPRFX must match that of the instruction at PC+4 * The predicated instruction at PC+4 must use the merging predicate. * Architectural aliases and pseudo-instructions need to be supported as well. * MOVPRFX cannot be the last instruction in a sequence Any failure to adhere to any of these constrains will emit an assembly warning and a disassembly note. [1] https://developer.arm.com/docs/ddi0584/latest/arm-architecture-reference-manual-supplement-the-scalable-vector-extension-sve-for-armv8-a include/ * opcode/aarch64.h (aarch64_inst): Remove. (enum err_type): Add ERR_VFI. (aarch64_is_destructive_by_operands): New. (init_insn_sequence): New. (aarch64_decode_insn): Remove param name. opcodes/ * aarch64-opc.c (init_insn_block): New. (verify_constraints, aarch64_is_destructive_by_operands): New. * aarch64-opc.h (verify_constraints): New. gas/ * config/tc-aarch64.c (output_operand_error_report): Order warnings.
2018-10-04 01:38:42 +08:00
/* This function verifies that the instruction INST adheres to its specified
constraints. If it does then ERR_OK is returned, if not then ERR_VFI is
returned and MISMATCH_DETAIL contains the reason why verification failed.
The function is called both during assembly and disassembly. If assembling
then ENCODING will be TRUE, else FALSE. If dissassembling PC will be set
and will contain the PC of the current instruction w.r.t to the section.
If ENCODING and PC=0 then you are at a start of a section. The constraints
are verified against the given state insn_sequence which is updated as it
transitions through the verification. */
enum err_type
verify_constraints (const struct aarch64_inst *inst,
const aarch64_insn insn ATTRIBUTE_UNUSED,
bfd_vma pc,
bool encoding,
AArch64: Add SVE constraints verifier. This patch adds the verification rules for move prefix constraints. The Arm SVE instruction MOVPRFX introduces[1] constraints on the instruction at PC+4. Particularly the following constraints are handled by this patch * MOVPRFX must be followed by an instruction. * MOVPRFX can only be followed by non-layout altering directives. * MOVPRFX destination register MUST be used as the destination register in the instruction at PC+4, and is not allowed to be used in any other position other than destructive input. This includes registers that architecturally overlap. e.g. x1 should be treated as z1. * MOVPRFX must be followed by a restricted set of SVE instructions. * The size of the destination register of MOVPRFX must be equal to that of the operation at PC+4. * The predicate register and operation of MOVPRFX must match that of the instruction at PC+4 * The predicated instruction at PC+4 must use the merging predicate. * Architectural aliases and pseudo-instructions need to be supported as well. * MOVPRFX cannot be the last instruction in a sequence Any failure to adhere to any of these constrains will emit an assembly warning and a disassembly note. [1] https://developer.arm.com/docs/ddi0584/latest/arm-architecture-reference-manual-supplement-the-scalable-vector-extension-sve-for-armv8-a include/ * opcode/aarch64.h (aarch64_inst): Remove. (enum err_type): Add ERR_VFI. (aarch64_is_destructive_by_operands): New. (init_insn_sequence): New. (aarch64_decode_insn): Remove param name. opcodes/ * aarch64-opc.c (init_insn_block): New. (verify_constraints, aarch64_is_destructive_by_operands): New. * aarch64-opc.h (verify_constraints): New. gas/ * config/tc-aarch64.c (output_operand_error_report): Order warnings.
2018-10-04 01:38:42 +08:00
aarch64_operand_error *mismatch_detail,
aarch64_instr_sequence *insn_sequence)
{
assert (inst);
assert (inst->opcode);
const struct aarch64_opcode *opcode = inst->opcode;
if (!opcode->constraints && !insn_sequence->instr)
return ERR_OK;
assert (insn_sequence);
enum err_type res = ERR_OK;
/* This instruction puts a constraint on the insn_sequence. */
if (opcode->flags & F_SCAN)
{
if (insn_sequence->instr)
{
mismatch_detail->kind = AARCH64_OPDE_SYNTAX_ERROR;
mismatch_detail->error = _("instruction opens new dependency "
"sequence without ending previous one");
mismatch_detail->index = -1;
mismatch_detail->non_fatal = true;
AArch64: Add SVE constraints verifier. This patch adds the verification rules for move prefix constraints. The Arm SVE instruction MOVPRFX introduces[1] constraints on the instruction at PC+4. Particularly the following constraints are handled by this patch * MOVPRFX must be followed by an instruction. * MOVPRFX can only be followed by non-layout altering directives. * MOVPRFX destination register MUST be used as the destination register in the instruction at PC+4, and is not allowed to be used in any other position other than destructive input. This includes registers that architecturally overlap. e.g. x1 should be treated as z1. * MOVPRFX must be followed by a restricted set of SVE instructions. * The size of the destination register of MOVPRFX must be equal to that of the operation at PC+4. * The predicate register and operation of MOVPRFX must match that of the instruction at PC+4 * The predicated instruction at PC+4 must use the merging predicate. * Architectural aliases and pseudo-instructions need to be supported as well. * MOVPRFX cannot be the last instruction in a sequence Any failure to adhere to any of these constrains will emit an assembly warning and a disassembly note. [1] https://developer.arm.com/docs/ddi0584/latest/arm-architecture-reference-manual-supplement-the-scalable-vector-extension-sve-for-armv8-a include/ * opcode/aarch64.h (aarch64_inst): Remove. (enum err_type): Add ERR_VFI. (aarch64_is_destructive_by_operands): New. (init_insn_sequence): New. (aarch64_decode_insn): Remove param name. opcodes/ * aarch64-opc.c (init_insn_block): New. (verify_constraints, aarch64_is_destructive_by_operands): New. * aarch64-opc.h (verify_constraints): New. gas/ * config/tc-aarch64.c (output_operand_error_report): Order warnings.
2018-10-04 01:38:42 +08:00
res = ERR_VFI;
}
init_insn_sequence (inst, insn_sequence);
return res;
}
aarch64: Enforce P/M/E order for MOPS instructions The MOPS instructions should be used as a triple, such as: cpyfp [x0]!, [x1]!, x2! cpyfm [x0]!, [x1]!, x2! cpyfe [x0]!, [x1]!, x2! The registers should also be the same for each writeback operand. This patch adds a warning for code that doesn't follow this rule, along similar lines to the warning that we already emit for invalid uses of MOVPRFX. include/ * opcode/aarch64.h (C_SCAN_MOPS_P, C_SCAN_MOPS_M, C_SCAN_MOPS_E) (C_SCAN_MOPS_PME): New macros. (AARCH64_OPDE_A_SHOULD_FOLLOW_B): New aarch64_operand_error_kind. (AARCH64_OPDE_EXPECTED_A_AFTER_B): Likewise. (aarch64_operand_error): Make each data value a union between an int and a string. opcodes/ * aarch64-tbl.h (MOPS_CPY_OP1_OP2_INSN): Add scan flags. (MOPS_SET_OP1_OP2_INSN): Likewise. * aarch64-opc.c (set_out_of_range_error): Update after change to aarch64_operand_error. (set_unaligned_error, set_reg_list_error): Likewise. (init_insn_sequence): Use a 3-instruction sequence for MOPS P instructions. (verify_mops_pme_sequence): New function. (verify_constraints): Call it. * aarch64-dis.c (print_verifier_notes): Handle AARCH64_OPDE_A_SHOULD_FOLLOW_B and AARCH64_OPDE_EXPECTED_A_AFTER_B. gas/ * config/tc-aarch64.c (operand_mismatch_kind_names): Add entries for AARCH64_OPDE_A_SHOULD_FOLLOW_B and AARCH64_OPDE_EXPECTED_A_AFTER_B. (operand_error_higher_severity_p): Check that AARCH64_OPDE_A_SHOULD_FOLLOW_B and AARCH64_OPDE_EXPECTED_A_AFTER_B come between AARCH64_OPDE_RECOVERABLE and AARCH64_OPDE_SYNTAX_ERROR; their relative order is not significant. (record_operand_error_with_data): Update after change to aarch64_operand_error. (output_operand_error_record): Likewise. Handle AARCH64_OPDE_A_SHOULD_FOLLOW_B and AARCH64_OPDE_EXPECTED_A_AFTER_B. * testsuite/gas/aarch64/mops_invalid_2.s, testsuite/gas/aarch64/mops_invalid_2.d, testsuite/gas/aarch64/mops_invalid_2.l: New test.
2021-12-02 23:00:57 +08:00
bool is_new_section = (!encoding && pc == 0);
if (!verify_mops_pme_sequence (inst, is_new_section, mismatch_detail,
insn_sequence))
{
res = ERR_VFI;
if ((opcode->constraints & C_SCAN_MOPS_PME) != C_SCAN_MOPS_M)
init_insn_sequence (NULL, insn_sequence);
}
AArch64: Add SVE constraints verifier. This patch adds the verification rules for move prefix constraints. The Arm SVE instruction MOVPRFX introduces[1] constraints on the instruction at PC+4. Particularly the following constraints are handled by this patch * MOVPRFX must be followed by an instruction. * MOVPRFX can only be followed by non-layout altering directives. * MOVPRFX destination register MUST be used as the destination register in the instruction at PC+4, and is not allowed to be used in any other position other than destructive input. This includes registers that architecturally overlap. e.g. x1 should be treated as z1. * MOVPRFX must be followed by a restricted set of SVE instructions. * The size of the destination register of MOVPRFX must be equal to that of the operation at PC+4. * The predicate register and operation of MOVPRFX must match that of the instruction at PC+4 * The predicated instruction at PC+4 must use the merging predicate. * Architectural aliases and pseudo-instructions need to be supported as well. * MOVPRFX cannot be the last instruction in a sequence Any failure to adhere to any of these constrains will emit an assembly warning and a disassembly note. [1] https://developer.arm.com/docs/ddi0584/latest/arm-architecture-reference-manual-supplement-the-scalable-vector-extension-sve-for-armv8-a include/ * opcode/aarch64.h (aarch64_inst): Remove. (enum err_type): Add ERR_VFI. (aarch64_is_destructive_by_operands): New. (init_insn_sequence): New. (aarch64_decode_insn): Remove param name. opcodes/ * aarch64-opc.c (init_insn_block): New. (verify_constraints, aarch64_is_destructive_by_operands): New. * aarch64-opc.h (verify_constraints): New. gas/ * config/tc-aarch64.c (output_operand_error_report): Order warnings.
2018-10-04 01:38:42 +08:00
/* Verify constraints on an existing sequence. */
if (insn_sequence->instr)
{
const struct aarch64_opcode* inst_opcode = insn_sequence->instr->opcode;
/* If we're decoding and we hit PC=0 with an open sequence then we haven't
closed a previous one that we should have. */
aarch64: Enforce P/M/E order for MOPS instructions The MOPS instructions should be used as a triple, such as: cpyfp [x0]!, [x1]!, x2! cpyfm [x0]!, [x1]!, x2! cpyfe [x0]!, [x1]!, x2! The registers should also be the same for each writeback operand. This patch adds a warning for code that doesn't follow this rule, along similar lines to the warning that we already emit for invalid uses of MOVPRFX. include/ * opcode/aarch64.h (C_SCAN_MOPS_P, C_SCAN_MOPS_M, C_SCAN_MOPS_E) (C_SCAN_MOPS_PME): New macros. (AARCH64_OPDE_A_SHOULD_FOLLOW_B): New aarch64_operand_error_kind. (AARCH64_OPDE_EXPECTED_A_AFTER_B): Likewise. (aarch64_operand_error): Make each data value a union between an int and a string. opcodes/ * aarch64-tbl.h (MOPS_CPY_OP1_OP2_INSN): Add scan flags. (MOPS_SET_OP1_OP2_INSN): Likewise. * aarch64-opc.c (set_out_of_range_error): Update after change to aarch64_operand_error. (set_unaligned_error, set_reg_list_error): Likewise. (init_insn_sequence): Use a 3-instruction sequence for MOPS P instructions. (verify_mops_pme_sequence): New function. (verify_constraints): Call it. * aarch64-dis.c (print_verifier_notes): Handle AARCH64_OPDE_A_SHOULD_FOLLOW_B and AARCH64_OPDE_EXPECTED_A_AFTER_B. gas/ * config/tc-aarch64.c (operand_mismatch_kind_names): Add entries for AARCH64_OPDE_A_SHOULD_FOLLOW_B and AARCH64_OPDE_EXPECTED_A_AFTER_B. (operand_error_higher_severity_p): Check that AARCH64_OPDE_A_SHOULD_FOLLOW_B and AARCH64_OPDE_EXPECTED_A_AFTER_B come between AARCH64_OPDE_RECOVERABLE and AARCH64_OPDE_SYNTAX_ERROR; their relative order is not significant. (record_operand_error_with_data): Update after change to aarch64_operand_error. (output_operand_error_record): Likewise. Handle AARCH64_OPDE_A_SHOULD_FOLLOW_B and AARCH64_OPDE_EXPECTED_A_AFTER_B. * testsuite/gas/aarch64/mops_invalid_2.s, testsuite/gas/aarch64/mops_invalid_2.d, testsuite/gas/aarch64/mops_invalid_2.l: New test.
2021-12-02 23:00:57 +08:00
if (is_new_section && res == ERR_OK)
AArch64: Add SVE constraints verifier. This patch adds the verification rules for move prefix constraints. The Arm SVE instruction MOVPRFX introduces[1] constraints on the instruction at PC+4. Particularly the following constraints are handled by this patch * MOVPRFX must be followed by an instruction. * MOVPRFX can only be followed by non-layout altering directives. * MOVPRFX destination register MUST be used as the destination register in the instruction at PC+4, and is not allowed to be used in any other position other than destructive input. This includes registers that architecturally overlap. e.g. x1 should be treated as z1. * MOVPRFX must be followed by a restricted set of SVE instructions. * The size of the destination register of MOVPRFX must be equal to that of the operation at PC+4. * The predicate register and operation of MOVPRFX must match that of the instruction at PC+4 * The predicated instruction at PC+4 must use the merging predicate. * Architectural aliases and pseudo-instructions need to be supported as well. * MOVPRFX cannot be the last instruction in a sequence Any failure to adhere to any of these constrains will emit an assembly warning and a disassembly note. [1] https://developer.arm.com/docs/ddi0584/latest/arm-architecture-reference-manual-supplement-the-scalable-vector-extension-sve-for-armv8-a include/ * opcode/aarch64.h (aarch64_inst): Remove. (enum err_type): Add ERR_VFI. (aarch64_is_destructive_by_operands): New. (init_insn_sequence): New. (aarch64_decode_insn): Remove param name. opcodes/ * aarch64-opc.c (init_insn_block): New. (verify_constraints, aarch64_is_destructive_by_operands): New. * aarch64-opc.h (verify_constraints): New. gas/ * config/tc-aarch64.c (output_operand_error_report): Order warnings.
2018-10-04 01:38:42 +08:00
{
mismatch_detail->kind = AARCH64_OPDE_SYNTAX_ERROR;
mismatch_detail->error = _("previous `movprfx' sequence not closed");
mismatch_detail->index = -1;
mismatch_detail->non_fatal = true;
AArch64: Add SVE constraints verifier. This patch adds the verification rules for move prefix constraints. The Arm SVE instruction MOVPRFX introduces[1] constraints on the instruction at PC+4. Particularly the following constraints are handled by this patch * MOVPRFX must be followed by an instruction. * MOVPRFX can only be followed by non-layout altering directives. * MOVPRFX destination register MUST be used as the destination register in the instruction at PC+4, and is not allowed to be used in any other position other than destructive input. This includes registers that architecturally overlap. e.g. x1 should be treated as z1. * MOVPRFX must be followed by a restricted set of SVE instructions. * The size of the destination register of MOVPRFX must be equal to that of the operation at PC+4. * The predicate register and operation of MOVPRFX must match that of the instruction at PC+4 * The predicated instruction at PC+4 must use the merging predicate. * Architectural aliases and pseudo-instructions need to be supported as well. * MOVPRFX cannot be the last instruction in a sequence Any failure to adhere to any of these constrains will emit an assembly warning and a disassembly note. [1] https://developer.arm.com/docs/ddi0584/latest/arm-architecture-reference-manual-supplement-the-scalable-vector-extension-sve-for-armv8-a include/ * opcode/aarch64.h (aarch64_inst): Remove. (enum err_type): Add ERR_VFI. (aarch64_is_destructive_by_operands): New. (init_insn_sequence): New. (aarch64_decode_insn): Remove param name. opcodes/ * aarch64-opc.c (init_insn_block): New. (verify_constraints, aarch64_is_destructive_by_operands): New. * aarch64-opc.h (verify_constraints): New. gas/ * config/tc-aarch64.c (output_operand_error_report): Order warnings.
2018-10-04 01:38:42 +08:00
res = ERR_VFI;
/* Reset the sequence. */
init_insn_sequence (NULL, insn_sequence);
return res;
}
/* Validate C_SCAN_MOVPRFX constraints. Move this to a lookup table. */
if (inst_opcode->constraints & C_SCAN_MOVPRFX)
{
/* Check to see if the MOVPRFX SVE instruction is followed by an SVE
instruction for better error messages. */
if (!opcode->avariant
|| (!AARCH64_CPU_HAS_FEATURE (*opcode->avariant, SVE)
&& !AARCH64_CPU_HAS_FEATURE (*opcode->avariant, SVE2)))
AArch64: Add SVE constraints verifier. This patch adds the verification rules for move prefix constraints. The Arm SVE instruction MOVPRFX introduces[1] constraints on the instruction at PC+4. Particularly the following constraints are handled by this patch * MOVPRFX must be followed by an instruction. * MOVPRFX can only be followed by non-layout altering directives. * MOVPRFX destination register MUST be used as the destination register in the instruction at PC+4, and is not allowed to be used in any other position other than destructive input. This includes registers that architecturally overlap. e.g. x1 should be treated as z1. * MOVPRFX must be followed by a restricted set of SVE instructions. * The size of the destination register of MOVPRFX must be equal to that of the operation at PC+4. * The predicate register and operation of MOVPRFX must match that of the instruction at PC+4 * The predicated instruction at PC+4 must use the merging predicate. * Architectural aliases and pseudo-instructions need to be supported as well. * MOVPRFX cannot be the last instruction in a sequence Any failure to adhere to any of these constrains will emit an assembly warning and a disassembly note. [1] https://developer.arm.com/docs/ddi0584/latest/arm-architecture-reference-manual-supplement-the-scalable-vector-extension-sve-for-armv8-a include/ * opcode/aarch64.h (aarch64_inst): Remove. (enum err_type): Add ERR_VFI. (aarch64_is_destructive_by_operands): New. (init_insn_sequence): New. (aarch64_decode_insn): Remove param name. opcodes/ * aarch64-opc.c (init_insn_block): New. (verify_constraints, aarch64_is_destructive_by_operands): New. * aarch64-opc.h (verify_constraints): New. gas/ * config/tc-aarch64.c (output_operand_error_report): Order warnings.
2018-10-04 01:38:42 +08:00
{
mismatch_detail->kind = AARCH64_OPDE_SYNTAX_ERROR;
mismatch_detail->error = _("SVE instruction expected after "
"`movprfx'");
mismatch_detail->index = -1;
mismatch_detail->non_fatal = true;
AArch64: Add SVE constraints verifier. This patch adds the verification rules for move prefix constraints. The Arm SVE instruction MOVPRFX introduces[1] constraints on the instruction at PC+4. Particularly the following constraints are handled by this patch * MOVPRFX must be followed by an instruction. * MOVPRFX can only be followed by non-layout altering directives. * MOVPRFX destination register MUST be used as the destination register in the instruction at PC+4, and is not allowed to be used in any other position other than destructive input. This includes registers that architecturally overlap. e.g. x1 should be treated as z1. * MOVPRFX must be followed by a restricted set of SVE instructions. * The size of the destination register of MOVPRFX must be equal to that of the operation at PC+4. * The predicate register and operation of MOVPRFX must match that of the instruction at PC+4 * The predicated instruction at PC+4 must use the merging predicate. * Architectural aliases and pseudo-instructions need to be supported as well. * MOVPRFX cannot be the last instruction in a sequence Any failure to adhere to any of these constrains will emit an assembly warning and a disassembly note. [1] https://developer.arm.com/docs/ddi0584/latest/arm-architecture-reference-manual-supplement-the-scalable-vector-extension-sve-for-armv8-a include/ * opcode/aarch64.h (aarch64_inst): Remove. (enum err_type): Add ERR_VFI. (aarch64_is_destructive_by_operands): New. (init_insn_sequence): New. (aarch64_decode_insn): Remove param name. opcodes/ * aarch64-opc.c (init_insn_block): New. (verify_constraints, aarch64_is_destructive_by_operands): New. * aarch64-opc.h (verify_constraints): New. gas/ * config/tc-aarch64.c (output_operand_error_report): Order warnings.
2018-10-04 01:38:42 +08:00
res = ERR_VFI;
goto done;
}
/* Check to see if the MOVPRFX SVE instruction is followed by an SVE
instruction that is allowed to be used with a MOVPRFX. */
if (!(opcode->constraints & C_SCAN_MOVPRFX))
{
mismatch_detail->kind = AARCH64_OPDE_SYNTAX_ERROR;
mismatch_detail->error = _("SVE `movprfx' compatible instruction "
"expected");
mismatch_detail->index = -1;
mismatch_detail->non_fatal = true;
AArch64: Add SVE constraints verifier. This patch adds the verification rules for move prefix constraints. The Arm SVE instruction MOVPRFX introduces[1] constraints on the instruction at PC+4. Particularly the following constraints are handled by this patch * MOVPRFX must be followed by an instruction. * MOVPRFX can only be followed by non-layout altering directives. * MOVPRFX destination register MUST be used as the destination register in the instruction at PC+4, and is not allowed to be used in any other position other than destructive input. This includes registers that architecturally overlap. e.g. x1 should be treated as z1. * MOVPRFX must be followed by a restricted set of SVE instructions. * The size of the destination register of MOVPRFX must be equal to that of the operation at PC+4. * The predicate register and operation of MOVPRFX must match that of the instruction at PC+4 * The predicated instruction at PC+4 must use the merging predicate. * Architectural aliases and pseudo-instructions need to be supported as well. * MOVPRFX cannot be the last instruction in a sequence Any failure to adhere to any of these constrains will emit an assembly warning and a disassembly note. [1] https://developer.arm.com/docs/ddi0584/latest/arm-architecture-reference-manual-supplement-the-scalable-vector-extension-sve-for-armv8-a include/ * opcode/aarch64.h (aarch64_inst): Remove. (enum err_type): Add ERR_VFI. (aarch64_is_destructive_by_operands): New. (init_insn_sequence): New. (aarch64_decode_insn): Remove param name. opcodes/ * aarch64-opc.c (init_insn_block): New. (verify_constraints, aarch64_is_destructive_by_operands): New. * aarch64-opc.h (verify_constraints): New. gas/ * config/tc-aarch64.c (output_operand_error_report): Order warnings.
2018-10-04 01:38:42 +08:00
res = ERR_VFI;
goto done;
}
/* Next check for usage of the predicate register. */
aarch64_opnd_info blk_dest = insn_sequence->instr->operands[0];
aarch64_opnd_info blk_pred, inst_pred;
memset (&blk_pred, 0, sizeof (aarch64_opnd_info));
memset (&inst_pred, 0, sizeof (aarch64_opnd_info));
bool predicated = false;
AArch64: Add SVE constraints verifier. This patch adds the verification rules for move prefix constraints. The Arm SVE instruction MOVPRFX introduces[1] constraints on the instruction at PC+4. Particularly the following constraints are handled by this patch * MOVPRFX must be followed by an instruction. * MOVPRFX can only be followed by non-layout altering directives. * MOVPRFX destination register MUST be used as the destination register in the instruction at PC+4, and is not allowed to be used in any other position other than destructive input. This includes registers that architecturally overlap. e.g. x1 should be treated as z1. * MOVPRFX must be followed by a restricted set of SVE instructions. * The size of the destination register of MOVPRFX must be equal to that of the operation at PC+4. * The predicate register and operation of MOVPRFX must match that of the instruction at PC+4 * The predicated instruction at PC+4 must use the merging predicate. * Architectural aliases and pseudo-instructions need to be supported as well. * MOVPRFX cannot be the last instruction in a sequence Any failure to adhere to any of these constrains will emit an assembly warning and a disassembly note. [1] https://developer.arm.com/docs/ddi0584/latest/arm-architecture-reference-manual-supplement-the-scalable-vector-extension-sve-for-armv8-a include/ * opcode/aarch64.h (aarch64_inst): Remove. (enum err_type): Add ERR_VFI. (aarch64_is_destructive_by_operands): New. (init_insn_sequence): New. (aarch64_decode_insn): Remove param name. opcodes/ * aarch64-opc.c (init_insn_block): New. (verify_constraints, aarch64_is_destructive_by_operands): New. * aarch64-opc.h (verify_constraints): New. gas/ * config/tc-aarch64.c (output_operand_error_report): Order warnings.
2018-10-04 01:38:42 +08:00
assert (blk_dest.type == AARCH64_OPND_SVE_Zd);
/* Determine if the movprfx instruction used is predicated or not. */
if (insn_sequence->instr->operands[1].type == AARCH64_OPND_SVE_Pg3)
{
predicated = true;
AArch64: Add SVE constraints verifier. This patch adds the verification rules for move prefix constraints. The Arm SVE instruction MOVPRFX introduces[1] constraints on the instruction at PC+4. Particularly the following constraints are handled by this patch * MOVPRFX must be followed by an instruction. * MOVPRFX can only be followed by non-layout altering directives. * MOVPRFX destination register MUST be used as the destination register in the instruction at PC+4, and is not allowed to be used in any other position other than destructive input. This includes registers that architecturally overlap. e.g. x1 should be treated as z1. * MOVPRFX must be followed by a restricted set of SVE instructions. * The size of the destination register of MOVPRFX must be equal to that of the operation at PC+4. * The predicate register and operation of MOVPRFX must match that of the instruction at PC+4 * The predicated instruction at PC+4 must use the merging predicate. * Architectural aliases and pseudo-instructions need to be supported as well. * MOVPRFX cannot be the last instruction in a sequence Any failure to adhere to any of these constrains will emit an assembly warning and a disassembly note. [1] https://developer.arm.com/docs/ddi0584/latest/arm-architecture-reference-manual-supplement-the-scalable-vector-extension-sve-for-armv8-a include/ * opcode/aarch64.h (aarch64_inst): Remove. (enum err_type): Add ERR_VFI. (aarch64_is_destructive_by_operands): New. (init_insn_sequence): New. (aarch64_decode_insn): Remove param name. opcodes/ * aarch64-opc.c (init_insn_block): New. (verify_constraints, aarch64_is_destructive_by_operands): New. * aarch64-opc.h (verify_constraints): New. gas/ * config/tc-aarch64.c (output_operand_error_report): Order warnings.
2018-10-04 01:38:42 +08:00
blk_pred = insn_sequence->instr->operands[1];
}
unsigned char max_elem_size = 0;
unsigned char current_elem_size;
int num_op_used = 0, last_op_usage = 0;
int i, inst_pred_idx = -1;
int num_ops = aarch64_num_of_operands (opcode);
for (i = 0; i < num_ops; i++)
{
aarch64_opnd_info inst_op = inst->operands[i];
switch (inst_op.type)
{
case AARCH64_OPND_SVE_Zd:
case AARCH64_OPND_SVE_Zm_5:
case AARCH64_OPND_SVE_Zm_16:
case AARCH64_OPND_SVE_Zn:
case AARCH64_OPND_SVE_Zt:
case AARCH64_OPND_SVE_Vm:
case AARCH64_OPND_SVE_Vn:
case AARCH64_OPND_Va:
case AARCH64_OPND_Vn:
case AARCH64_OPND_Vm:
case AARCH64_OPND_Sn:
case AARCH64_OPND_Sm:
if (inst_op.reg.regno == blk_dest.reg.regno)
{
num_op_used++;
last_op_usage = i;
}
current_elem_size
= aarch64_get_qualifier_esize (inst_op.qualifier);
if (current_elem_size > max_elem_size)
max_elem_size = current_elem_size;
break;
case AARCH64_OPND_SVE_Pd:
case AARCH64_OPND_SVE_Pg3:
case AARCH64_OPND_SVE_Pg4_5:
case AARCH64_OPND_SVE_Pg4_10:
case AARCH64_OPND_SVE_Pg4_16:
case AARCH64_OPND_SVE_Pm:
case AARCH64_OPND_SVE_Pn:
case AARCH64_OPND_SVE_Pt:
aarch64: [SME] Add SME instructions Patch is adding new SME matrix instructions. Please note additional instructions will be added in following patches. gas/ChangeLog: * config/tc-aarch64.c (parse_sme_zada_operand): New parser. * config/tc-aarch64.c (parse_reg_with_qual): New reg parser. * config/tc-aarch64.c (R_ZA): New egister type. (parse_operands): New parser. * testsuite/gas/aarch64/sme-illegal.d: New test. * testsuite/gas/aarch64/sme-illegal.l: New test. * testsuite/gas/aarch64/sme-illegal.s: New test. * testsuite/gas/aarch64/sme.d: New test. * testsuite/gas/aarch64/sme.s: New test. * testsuite/gas/aarch64/sme-f64.d: New test. * testsuite/gas/aarch64/sme-f64.s: New test. * testsuite/gas/aarch64/sme-i64.d: New test. * testsuite/gas/aarch64/sme-i64.s: New test. include/ChangeLog: * opcode/aarch64.h (enum aarch64_opnd): New operands AARCH64_OPND_SME_ZAda_2b, AARCH64_OPND_SME_ZAda_3b and AARCH64_OPND_SME_Pm. (enum aarch64_insn_class): New instruction class sme_misc. opcodes/ChangeLog: * aarch64-opc.c (aarch64_print_operand): Print OPND_SME_ZAda_2b and OPND_SME_ZAda_3b operands. (verify_constraints): Handle OPND_SME_Pm. * aarch64-opc.h (enum aarch64_field_kind): New bit fields FLD_SME_ZAda_2b, FLD_SME_ZAda_3b and FLD_SME_Pm. * aarch64-tbl.h (OP_SME_ZADA_PN_PM_ZN_S): New qualifier set. (OP_SME_ZADA_PN_PM_ZN_D): New qualifier. (OP_SME_ZADA_PN_PM_ZN_ZM): New qualifier. (OP_SME_ZADA_S_PM_PM_S_S): New qualifier. (OP_SME_ZADA_D_PM_PM_D_D): New qualifier. (OP_SME_ZADA_S_PM_PM_H_H): New qualifier. (OP_SME_ZADA_S_PM_PM_B_B): New qualifier. (OP_SME_ZADA_D_PM_PM_H_H): New qualifier. (SME_INSN): New instruction macro. (SME_F64_INSN): New instruction macro. (SME_I64_INSN): New instruction macro. (SME_INSNC): New instruction macro. (struct aarch64_opcode): New SME instructions. aarch64-asm-2.c: Regenerate. aarch64-dis-2.c: Regenerate. aarch64-opc-2.c: Regenerate.
2021-11-18 03:21:33 +08:00
case AARCH64_OPND_SME_Pm:
AArch64: Add SVE constraints verifier. This patch adds the verification rules for move prefix constraints. The Arm SVE instruction MOVPRFX introduces[1] constraints on the instruction at PC+4. Particularly the following constraints are handled by this patch * MOVPRFX must be followed by an instruction. * MOVPRFX can only be followed by non-layout altering directives. * MOVPRFX destination register MUST be used as the destination register in the instruction at PC+4, and is not allowed to be used in any other position other than destructive input. This includes registers that architecturally overlap. e.g. x1 should be treated as z1. * MOVPRFX must be followed by a restricted set of SVE instructions. * The size of the destination register of MOVPRFX must be equal to that of the operation at PC+4. * The predicate register and operation of MOVPRFX must match that of the instruction at PC+4 * The predicated instruction at PC+4 must use the merging predicate. * Architectural aliases and pseudo-instructions need to be supported as well. * MOVPRFX cannot be the last instruction in a sequence Any failure to adhere to any of these constrains will emit an assembly warning and a disassembly note. [1] https://developer.arm.com/docs/ddi0584/latest/arm-architecture-reference-manual-supplement-the-scalable-vector-extension-sve-for-armv8-a include/ * opcode/aarch64.h (aarch64_inst): Remove. (enum err_type): Add ERR_VFI. (aarch64_is_destructive_by_operands): New. (init_insn_sequence): New. (aarch64_decode_insn): Remove param name. opcodes/ * aarch64-opc.c (init_insn_block): New. (verify_constraints, aarch64_is_destructive_by_operands): New. * aarch64-opc.h (verify_constraints): New. gas/ * config/tc-aarch64.c (output_operand_error_report): Order warnings.
2018-10-04 01:38:42 +08:00
inst_pred = inst_op;
inst_pred_idx = i;
break;
default:
break;
}
}
assert (max_elem_size != 0);
aarch64_opnd_info inst_dest = inst->operands[0];
/* Determine the size that should be used to compare against the
movprfx size. */
current_elem_size
= opcode->constraints & C_MAX_ELEM
? max_elem_size
: aarch64_get_qualifier_esize (inst_dest.qualifier);
/* If movprfx is predicated do some extra checks. */
if (predicated)
{
/* The instruction must be predicated. */
if (inst_pred_idx < 0)
{
mismatch_detail->kind = AARCH64_OPDE_SYNTAX_ERROR;
mismatch_detail->error = _("predicated instruction expected "
"after `movprfx'");
mismatch_detail->index = -1;
mismatch_detail->non_fatal = true;
AArch64: Add SVE constraints verifier. This patch adds the verification rules for move prefix constraints. The Arm SVE instruction MOVPRFX introduces[1] constraints on the instruction at PC+4. Particularly the following constraints are handled by this patch * MOVPRFX must be followed by an instruction. * MOVPRFX can only be followed by non-layout altering directives. * MOVPRFX destination register MUST be used as the destination register in the instruction at PC+4, and is not allowed to be used in any other position other than destructive input. This includes registers that architecturally overlap. e.g. x1 should be treated as z1. * MOVPRFX must be followed by a restricted set of SVE instructions. * The size of the destination register of MOVPRFX must be equal to that of the operation at PC+4. * The predicate register and operation of MOVPRFX must match that of the instruction at PC+4 * The predicated instruction at PC+4 must use the merging predicate. * Architectural aliases and pseudo-instructions need to be supported as well. * MOVPRFX cannot be the last instruction in a sequence Any failure to adhere to any of these constrains will emit an assembly warning and a disassembly note. [1] https://developer.arm.com/docs/ddi0584/latest/arm-architecture-reference-manual-supplement-the-scalable-vector-extension-sve-for-armv8-a include/ * opcode/aarch64.h (aarch64_inst): Remove. (enum err_type): Add ERR_VFI. (aarch64_is_destructive_by_operands): New. (init_insn_sequence): New. (aarch64_decode_insn): Remove param name. opcodes/ * aarch64-opc.c (init_insn_block): New. (verify_constraints, aarch64_is_destructive_by_operands): New. * aarch64-opc.h (verify_constraints): New. gas/ * config/tc-aarch64.c (output_operand_error_report): Order warnings.
2018-10-04 01:38:42 +08:00
res = ERR_VFI;
goto done;
}
/* The instruction must have a merging predicate. */
if (inst_pred.qualifier != AARCH64_OPND_QLF_P_M)
{
mismatch_detail->kind = AARCH64_OPDE_SYNTAX_ERROR;
mismatch_detail->error = _("merging predicate expected due "
"to preceding `movprfx'");
mismatch_detail->index = inst_pred_idx;
mismatch_detail->non_fatal = true;
AArch64: Add SVE constraints verifier. This patch adds the verification rules for move prefix constraints. The Arm SVE instruction MOVPRFX introduces[1] constraints on the instruction at PC+4. Particularly the following constraints are handled by this patch * MOVPRFX must be followed by an instruction. * MOVPRFX can only be followed by non-layout altering directives. * MOVPRFX destination register MUST be used as the destination register in the instruction at PC+4, and is not allowed to be used in any other position other than destructive input. This includes registers that architecturally overlap. e.g. x1 should be treated as z1. * MOVPRFX must be followed by a restricted set of SVE instructions. * The size of the destination register of MOVPRFX must be equal to that of the operation at PC+4. * The predicate register and operation of MOVPRFX must match that of the instruction at PC+4 * The predicated instruction at PC+4 must use the merging predicate. * Architectural aliases and pseudo-instructions need to be supported as well. * MOVPRFX cannot be the last instruction in a sequence Any failure to adhere to any of these constrains will emit an assembly warning and a disassembly note. [1] https://developer.arm.com/docs/ddi0584/latest/arm-architecture-reference-manual-supplement-the-scalable-vector-extension-sve-for-armv8-a include/ * opcode/aarch64.h (aarch64_inst): Remove. (enum err_type): Add ERR_VFI. (aarch64_is_destructive_by_operands): New. (init_insn_sequence): New. (aarch64_decode_insn): Remove param name. opcodes/ * aarch64-opc.c (init_insn_block): New. (verify_constraints, aarch64_is_destructive_by_operands): New. * aarch64-opc.h (verify_constraints): New. gas/ * config/tc-aarch64.c (output_operand_error_report): Order warnings.
2018-10-04 01:38:42 +08:00
res = ERR_VFI;
goto done;
}
/* The same register must be used in instruction. */
if (blk_pred.reg.regno != inst_pred.reg.regno)
{
mismatch_detail->kind = AARCH64_OPDE_SYNTAX_ERROR;
mismatch_detail->error = _("predicate register differs "
"from that in preceding "
"`movprfx'");
mismatch_detail->index = inst_pred_idx;
mismatch_detail->non_fatal = true;
AArch64: Add SVE constraints verifier. This patch adds the verification rules for move prefix constraints. The Arm SVE instruction MOVPRFX introduces[1] constraints on the instruction at PC+4. Particularly the following constraints are handled by this patch * MOVPRFX must be followed by an instruction. * MOVPRFX can only be followed by non-layout altering directives. * MOVPRFX destination register MUST be used as the destination register in the instruction at PC+4, and is not allowed to be used in any other position other than destructive input. This includes registers that architecturally overlap. e.g. x1 should be treated as z1. * MOVPRFX must be followed by a restricted set of SVE instructions. * The size of the destination register of MOVPRFX must be equal to that of the operation at PC+4. * The predicate register and operation of MOVPRFX must match that of the instruction at PC+4 * The predicated instruction at PC+4 must use the merging predicate. * Architectural aliases and pseudo-instructions need to be supported as well. * MOVPRFX cannot be the last instruction in a sequence Any failure to adhere to any of these constrains will emit an assembly warning and a disassembly note. [1] https://developer.arm.com/docs/ddi0584/latest/arm-architecture-reference-manual-supplement-the-scalable-vector-extension-sve-for-armv8-a include/ * opcode/aarch64.h (aarch64_inst): Remove. (enum err_type): Add ERR_VFI. (aarch64_is_destructive_by_operands): New. (init_insn_sequence): New. (aarch64_decode_insn): Remove param name. opcodes/ * aarch64-opc.c (init_insn_block): New. (verify_constraints, aarch64_is_destructive_by_operands): New. * aarch64-opc.h (verify_constraints): New. gas/ * config/tc-aarch64.c (output_operand_error_report): Order warnings.
2018-10-04 01:38:42 +08:00
res = ERR_VFI;
goto done;
}
}
/* Destructive operations by definition must allow one usage of the
same register. */
int allowed_usage
= aarch64_is_destructive_by_operands (opcode) ? 2 : 1;
/* Operand is not used at all. */
if (num_op_used == 0)
{
mismatch_detail->kind = AARCH64_OPDE_SYNTAX_ERROR;
mismatch_detail->error = _("output register of preceding "
"`movprfx' not used in current "
"instruction");
mismatch_detail->index = 0;
mismatch_detail->non_fatal = true;
AArch64: Add SVE constraints verifier. This patch adds the verification rules for move prefix constraints. The Arm SVE instruction MOVPRFX introduces[1] constraints on the instruction at PC+4. Particularly the following constraints are handled by this patch * MOVPRFX must be followed by an instruction. * MOVPRFX can only be followed by non-layout altering directives. * MOVPRFX destination register MUST be used as the destination register in the instruction at PC+4, and is not allowed to be used in any other position other than destructive input. This includes registers that architecturally overlap. e.g. x1 should be treated as z1. * MOVPRFX must be followed by a restricted set of SVE instructions. * The size of the destination register of MOVPRFX must be equal to that of the operation at PC+4. * The predicate register and operation of MOVPRFX must match that of the instruction at PC+4 * The predicated instruction at PC+4 must use the merging predicate. * Architectural aliases and pseudo-instructions need to be supported as well. * MOVPRFX cannot be the last instruction in a sequence Any failure to adhere to any of these constrains will emit an assembly warning and a disassembly note. [1] https://developer.arm.com/docs/ddi0584/latest/arm-architecture-reference-manual-supplement-the-scalable-vector-extension-sve-for-armv8-a include/ * opcode/aarch64.h (aarch64_inst): Remove. (enum err_type): Add ERR_VFI. (aarch64_is_destructive_by_operands): New. (init_insn_sequence): New. (aarch64_decode_insn): Remove param name. opcodes/ * aarch64-opc.c (init_insn_block): New. (verify_constraints, aarch64_is_destructive_by_operands): New. * aarch64-opc.h (verify_constraints): New. gas/ * config/tc-aarch64.c (output_operand_error_report): Order warnings.
2018-10-04 01:38:42 +08:00
res = ERR_VFI;
goto done;
}
/* We now know it's used, now determine exactly where it's used. */
if (blk_dest.reg.regno != inst_dest.reg.regno)
{
mismatch_detail->kind = AARCH64_OPDE_SYNTAX_ERROR;
mismatch_detail->error = _("output register of preceding "
"`movprfx' expected as output");
mismatch_detail->index = 0;
mismatch_detail->non_fatal = true;
AArch64: Add SVE constraints verifier. This patch adds the verification rules for move prefix constraints. The Arm SVE instruction MOVPRFX introduces[1] constraints on the instruction at PC+4. Particularly the following constraints are handled by this patch * MOVPRFX must be followed by an instruction. * MOVPRFX can only be followed by non-layout altering directives. * MOVPRFX destination register MUST be used as the destination register in the instruction at PC+4, and is not allowed to be used in any other position other than destructive input. This includes registers that architecturally overlap. e.g. x1 should be treated as z1. * MOVPRFX must be followed by a restricted set of SVE instructions. * The size of the destination register of MOVPRFX must be equal to that of the operation at PC+4. * The predicate register and operation of MOVPRFX must match that of the instruction at PC+4 * The predicated instruction at PC+4 must use the merging predicate. * Architectural aliases and pseudo-instructions need to be supported as well. * MOVPRFX cannot be the last instruction in a sequence Any failure to adhere to any of these constrains will emit an assembly warning and a disassembly note. [1] https://developer.arm.com/docs/ddi0584/latest/arm-architecture-reference-manual-supplement-the-scalable-vector-extension-sve-for-armv8-a include/ * opcode/aarch64.h (aarch64_inst): Remove. (enum err_type): Add ERR_VFI. (aarch64_is_destructive_by_operands): New. (init_insn_sequence): New. (aarch64_decode_insn): Remove param name. opcodes/ * aarch64-opc.c (init_insn_block): New. (verify_constraints, aarch64_is_destructive_by_operands): New. * aarch64-opc.h (verify_constraints): New. gas/ * config/tc-aarch64.c (output_operand_error_report): Order warnings.
2018-10-04 01:38:42 +08:00
res = ERR_VFI;
goto done;
}
/* Operand used more than allowed for the specific opcode type. */
if (num_op_used > allowed_usage)
{
mismatch_detail->kind = AARCH64_OPDE_SYNTAX_ERROR;
mismatch_detail->error = _("output register of preceding "
"`movprfx' used as input");
mismatch_detail->index = last_op_usage;
mismatch_detail->non_fatal = true;
AArch64: Add SVE constraints verifier. This patch adds the verification rules for move prefix constraints. The Arm SVE instruction MOVPRFX introduces[1] constraints on the instruction at PC+4. Particularly the following constraints are handled by this patch * MOVPRFX must be followed by an instruction. * MOVPRFX can only be followed by non-layout altering directives. * MOVPRFX destination register MUST be used as the destination register in the instruction at PC+4, and is not allowed to be used in any other position other than destructive input. This includes registers that architecturally overlap. e.g. x1 should be treated as z1. * MOVPRFX must be followed by a restricted set of SVE instructions. * The size of the destination register of MOVPRFX must be equal to that of the operation at PC+4. * The predicate register and operation of MOVPRFX must match that of the instruction at PC+4 * The predicated instruction at PC+4 must use the merging predicate. * Architectural aliases and pseudo-instructions need to be supported as well. * MOVPRFX cannot be the last instruction in a sequence Any failure to adhere to any of these constrains will emit an assembly warning and a disassembly note. [1] https://developer.arm.com/docs/ddi0584/latest/arm-architecture-reference-manual-supplement-the-scalable-vector-extension-sve-for-armv8-a include/ * opcode/aarch64.h (aarch64_inst): Remove. (enum err_type): Add ERR_VFI. (aarch64_is_destructive_by_operands): New. (init_insn_sequence): New. (aarch64_decode_insn): Remove param name. opcodes/ * aarch64-opc.c (init_insn_block): New. (verify_constraints, aarch64_is_destructive_by_operands): New. * aarch64-opc.h (verify_constraints): New. gas/ * config/tc-aarch64.c (output_operand_error_report): Order warnings.
2018-10-04 01:38:42 +08:00
res = ERR_VFI;
goto done;
}
/* Now the only thing left is the qualifiers checks. The register
must have the same maximum element size. */
if (inst_dest.qualifier
&& blk_dest.qualifier
&& current_elem_size
!= aarch64_get_qualifier_esize (blk_dest.qualifier))
{
mismatch_detail->kind = AARCH64_OPDE_SYNTAX_ERROR;
mismatch_detail->error = _("register size not compatible with "
"previous `movprfx'");
mismatch_detail->index = 0;
mismatch_detail->non_fatal = true;
AArch64: Add SVE constraints verifier. This patch adds the verification rules for move prefix constraints. The Arm SVE instruction MOVPRFX introduces[1] constraints on the instruction at PC+4. Particularly the following constraints are handled by this patch * MOVPRFX must be followed by an instruction. * MOVPRFX can only be followed by non-layout altering directives. * MOVPRFX destination register MUST be used as the destination register in the instruction at PC+4, and is not allowed to be used in any other position other than destructive input. This includes registers that architecturally overlap. e.g. x1 should be treated as z1. * MOVPRFX must be followed by a restricted set of SVE instructions. * The size of the destination register of MOVPRFX must be equal to that of the operation at PC+4. * The predicate register and operation of MOVPRFX must match that of the instruction at PC+4 * The predicated instruction at PC+4 must use the merging predicate. * Architectural aliases and pseudo-instructions need to be supported as well. * MOVPRFX cannot be the last instruction in a sequence Any failure to adhere to any of these constrains will emit an assembly warning and a disassembly note. [1] https://developer.arm.com/docs/ddi0584/latest/arm-architecture-reference-manual-supplement-the-scalable-vector-extension-sve-for-armv8-a include/ * opcode/aarch64.h (aarch64_inst): Remove. (enum err_type): Add ERR_VFI. (aarch64_is_destructive_by_operands): New. (init_insn_sequence): New. (aarch64_decode_insn): Remove param name. opcodes/ * aarch64-opc.c (init_insn_block): New. (verify_constraints, aarch64_is_destructive_by_operands): New. * aarch64-opc.h (verify_constraints): New. gas/ * config/tc-aarch64.c (output_operand_error_report): Order warnings.
2018-10-04 01:38:42 +08:00
res = ERR_VFI;
goto done;
}
}
Indent labels Labels don't go in the first column according to standard emacs C indent rules, and I got annoyed enough at seeing diff -p show a label rather than the function name to fix this. bfd/ * aoutx.h: Indent labels correctly. Format error strings. * archive.c: Likewise. * archive64.c: Likewise. * coff-arm.c: Likewise. * coff-rs6000.c: Likewise. * coff-stgo32.c: Likewise. * cpu-arm.c: Likewise. * dwarf2.c: Likewise. * elf-ifunc.c: Likewise. * elf-properties.c: Likewise. * elf-s390-common.c: Likewise. * elf-strtab.c: Likewise. * elf.c: Likewise. * elf32-arm.c: Likewise. * elf32-bfin.c: Likewise. * elf32-cr16.c: Likewise. * elf32-csky.c: Likewise. * elf32-i386.c: Likewise. * elf32-m68k.c: Likewise. * elf32-msp430.c: Likewise. * elf32-nds32.c: Likewise. * elf32-nios2.c: Likewise. * elf32-pru.c: Likewise. * elf32-xtensa.c: Likewise. * elf64-ia64-vms.c: Likewise. * elf64-x86-64.c: Likewise. * elfcode.h: Likewise. * elfcore.h: Likewise. * elflink.c: Likewise. * elfnn-aarch64.c: Likewise. * elfnn-ia64.c: Likewise. * elfnn-riscv.c: Likewise. * elfxx-mips.c: Likewise. * elfxx-sparc.c: Likewise. * elfxx-x86.c: Likewise. * i386lynx.c: Likewise. * merge.c: Likewise. * pdp11.c: Likewise. * plugin.c: Likewise. * reloc.c: Likewise. binutils/ * elfedit.c: Indent labels correctly. * readelf.c: Likewise. * resres.c: Likewise. gas/ * config/obj-elf.c: Indent labels correctly. * config/obj-macho.c: Likewise. * config/tc-aarch64.c: Likewise. * config/tc-alpha.c: Likewise. * config/tc-arm.c: Likewise. * config/tc-cr16.c: Likewise. * config/tc-crx.c: Likewise. * config/tc-frv.c: Likewise. * config/tc-i386-intel.c: Likewise. * config/tc-i386.c: Likewise. * config/tc-ia64.c: Likewise. * config/tc-mn10200.c: Likewise. * config/tc-mn10300.c: Likewise. * config/tc-nds32.c: Likewise. * config/tc-riscv.c: Likewise. * config/tc-s12z.c: Likewise. * config/tc-xtensa.c: Likewise. * config/tc-z80.c: Likewise. * read.c: Likewise. * symbols.c: Likewise. * write.c: Likewise. ld/ * emultempl/cskyelf.em: Indent labels correctly. * ldfile.c: Likewise. * ldlang.c: Likewise. * plugin.c: Likewise. opcodes/ * aarch64-asm.c: Indent labels correctly. * aarch64-dis.c: Likewise. * aarch64-gen.c: Likewise. * aarch64-opc.c: Likewise. * alpha-dis.c: Likewise. * i386-dis.c: Likewise. * nds32-asm.c: Likewise. * nfp-dis.c: Likewise. * visium-dis.c: Likewise.
2020-02-25 13:04:46 +08:00
done:
if (insn_sequence->num_added_insns == insn_sequence->num_allocated_insns)
/* We've checked the last instruction in the sequence and so
don't need the sequence any more. */
init_insn_sequence (NULL, insn_sequence);
else
add_insn_to_sequence (inst, insn_sequence);
AArch64: Add SVE constraints verifier. This patch adds the verification rules for move prefix constraints. The Arm SVE instruction MOVPRFX introduces[1] constraints on the instruction at PC+4. Particularly the following constraints are handled by this patch * MOVPRFX must be followed by an instruction. * MOVPRFX can only be followed by non-layout altering directives. * MOVPRFX destination register MUST be used as the destination register in the instruction at PC+4, and is not allowed to be used in any other position other than destructive input. This includes registers that architecturally overlap. e.g. x1 should be treated as z1. * MOVPRFX must be followed by a restricted set of SVE instructions. * The size of the destination register of MOVPRFX must be equal to that of the operation at PC+4. * The predicate register and operation of MOVPRFX must match that of the instruction at PC+4 * The predicated instruction at PC+4 must use the merging predicate. * Architectural aliases and pseudo-instructions need to be supported as well. * MOVPRFX cannot be the last instruction in a sequence Any failure to adhere to any of these constrains will emit an assembly warning and a disassembly note. [1] https://developer.arm.com/docs/ddi0584/latest/arm-architecture-reference-manual-supplement-the-scalable-vector-extension-sve-for-armv8-a include/ * opcode/aarch64.h (aarch64_inst): Remove. (enum err_type): Add ERR_VFI. (aarch64_is_destructive_by_operands): New. (init_insn_sequence): New. (aarch64_decode_insn): Remove param name. opcodes/ * aarch64-opc.c (init_insn_block): New. (verify_constraints, aarch64_is_destructive_by_operands): New. * aarch64-opc.h (verify_constraints): New. gas/ * config/tc-aarch64.c (output_operand_error_report): Order warnings.
2018-10-04 01:38:42 +08:00
}
return res;
}
[AArch64][SVE 27/32] Add SVE integer immediate operands This patch adds the new SVE integer immediate operands. There are three kinds: - simple signed and unsigned ranges, but with new widths and positions. - 13-bit logical immediates. These have the same form as in base AArch64, but at a different bit position. In the case of the "MOV Zn.<T>, #<limm>" alias of DUPM, the logical immediate <limm> is not allowed to be a valid DUP immediate, since DUP is preferred over DUPM for constants that both instructions can handle. - a new 9-bit arithmetic immediate, of the form "<imm8>{, LSL #8}". In some contexts the operand is signed and in others it's unsigned. As an extension, we allow shifted immediates to be written as a single integer, e.g. "#256" is equivalent to "#1, LSL #8". We also use the shiftless form as the preferred disassembly, except for the special case of "#0, LSL #8" (a redundant encoding of 0). include/ * opcode/aarch64.h (AARCH64_OPND_SIMM5): New aarch64_opnd. (AARCH64_OPND_SVE_AIMM, AARCH64_OPND_SVE_ASIMM) (AARCH64_OPND_SVE_INV_LIMM, AARCH64_OPND_SVE_LIMM) (AARCH64_OPND_SVE_LIMM_MOV, AARCH64_OPND_SVE_SHLIMM_PRED) (AARCH64_OPND_SVE_SHLIMM_UNPRED, AARCH64_OPND_SVE_SHRIMM_PRED) (AARCH64_OPND_SVE_SHRIMM_UNPRED, AARCH64_OPND_SVE_SIMM5) (AARCH64_OPND_SVE_SIMM5B, AARCH64_OPND_SVE_SIMM6) (AARCH64_OPND_SVE_SIMM8, AARCH64_OPND_SVE_UIMM3) (AARCH64_OPND_SVE_UIMM7, AARCH64_OPND_SVE_UIMM8) (AARCH64_OPND_SVE_UIMM8_53): Likewise. (aarch64_sve_dupm_mov_immediate_p): Declare. opcodes/ * aarch64-tbl.h (AARCH64_OPERANDS): Add entries for the new SVE integer immediate operands. * aarch64-opc.h (FLD_SVE_immN, FLD_SVE_imm3, FLD_SVE_imm5) (FLD_SVE_imm5b, FLD_SVE_imm7, FLD_SVE_imm8, FLD_SVE_imm9) (FLD_SVE_immr, FLD_SVE_imms, FLD_SVE_tszh): New aarch64_field_kinds. * aarch64-opc.c (fields): Add corresponding entries. (operand_general_constraint_met_p): Handle the new SVE integer immediate operands. (aarch64_print_operand): Likewise. (aarch64_sve_dupm_mov_immediate_p): New function. * aarch64-opc-2.c: Regenerate. * aarch64-asm.h (ins_inv_limm, ins_sve_aimm, ins_sve_asimm) (ins_sve_limm_mov, ins_sve_shlimm, ins_sve_shrimm): New inserters. * aarch64-asm.c (aarch64_ins_limm_1): New function, split out from... (aarch64_ins_limm): ...here. (aarch64_ins_inv_limm): New function. (aarch64_ins_sve_aimm): Likewise. (aarch64_ins_sve_asimm): Likewise. (aarch64_ins_sve_limm_mov): Likewise. (aarch64_ins_sve_shlimm): Likewise. (aarch64_ins_sve_shrimm): Likewise. * aarch64-asm-2.c: Regenerate. * aarch64-dis.h (ext_inv_limm, ext_sve_aimm, ext_sve_asimm) (ext_sve_limm_mov, ext_sve_shlimm, ext_sve_shrimm): New extractors. * aarch64-dis.c (decode_limm): New function, split out from... (aarch64_ext_limm): ...here. (aarch64_ext_inv_limm): New function. (decode_sve_aimm): Likewise. (aarch64_ext_sve_aimm): Likewise. (aarch64_ext_sve_asimm): Likewise. (aarch64_ext_sve_limm_mov): Likewise. (aarch64_top_bit): Likewise. (aarch64_ext_sve_shlimm): Likewise. (aarch64_ext_sve_shrimm): Likewise. * aarch64-dis-2.c: Regenerate. gas/ * config/tc-aarch64.c (parse_operands): Handle the new SVE integer immediate operands.
2016-09-21 23:56:57 +08:00
/* Return true if VALUE cannot be moved into an SVE register using DUP
(with any element size, not just ESIZE) and if using DUPM would
therefore be OK. ESIZE is the number of bytes in the immediate. */
bool
[AArch64][SVE 27/32] Add SVE integer immediate operands This patch adds the new SVE integer immediate operands. There are three kinds: - simple signed and unsigned ranges, but with new widths and positions. - 13-bit logical immediates. These have the same form as in base AArch64, but at a different bit position. In the case of the "MOV Zn.<T>, #<limm>" alias of DUPM, the logical immediate <limm> is not allowed to be a valid DUP immediate, since DUP is preferred over DUPM for constants that both instructions can handle. - a new 9-bit arithmetic immediate, of the form "<imm8>{, LSL #8}". In some contexts the operand is signed and in others it's unsigned. As an extension, we allow shifted immediates to be written as a single integer, e.g. "#256" is equivalent to "#1, LSL #8". We also use the shiftless form as the preferred disassembly, except for the special case of "#0, LSL #8" (a redundant encoding of 0). include/ * opcode/aarch64.h (AARCH64_OPND_SIMM5): New aarch64_opnd. (AARCH64_OPND_SVE_AIMM, AARCH64_OPND_SVE_ASIMM) (AARCH64_OPND_SVE_INV_LIMM, AARCH64_OPND_SVE_LIMM) (AARCH64_OPND_SVE_LIMM_MOV, AARCH64_OPND_SVE_SHLIMM_PRED) (AARCH64_OPND_SVE_SHLIMM_UNPRED, AARCH64_OPND_SVE_SHRIMM_PRED) (AARCH64_OPND_SVE_SHRIMM_UNPRED, AARCH64_OPND_SVE_SIMM5) (AARCH64_OPND_SVE_SIMM5B, AARCH64_OPND_SVE_SIMM6) (AARCH64_OPND_SVE_SIMM8, AARCH64_OPND_SVE_UIMM3) (AARCH64_OPND_SVE_UIMM7, AARCH64_OPND_SVE_UIMM8) (AARCH64_OPND_SVE_UIMM8_53): Likewise. (aarch64_sve_dupm_mov_immediate_p): Declare. opcodes/ * aarch64-tbl.h (AARCH64_OPERANDS): Add entries for the new SVE integer immediate operands. * aarch64-opc.h (FLD_SVE_immN, FLD_SVE_imm3, FLD_SVE_imm5) (FLD_SVE_imm5b, FLD_SVE_imm7, FLD_SVE_imm8, FLD_SVE_imm9) (FLD_SVE_immr, FLD_SVE_imms, FLD_SVE_tszh): New aarch64_field_kinds. * aarch64-opc.c (fields): Add corresponding entries. (operand_general_constraint_met_p): Handle the new SVE integer immediate operands. (aarch64_print_operand): Likewise. (aarch64_sve_dupm_mov_immediate_p): New function. * aarch64-opc-2.c: Regenerate. * aarch64-asm.h (ins_inv_limm, ins_sve_aimm, ins_sve_asimm) (ins_sve_limm_mov, ins_sve_shlimm, ins_sve_shrimm): New inserters. * aarch64-asm.c (aarch64_ins_limm_1): New function, split out from... (aarch64_ins_limm): ...here. (aarch64_ins_inv_limm): New function. (aarch64_ins_sve_aimm): Likewise. (aarch64_ins_sve_asimm): Likewise. (aarch64_ins_sve_limm_mov): Likewise. (aarch64_ins_sve_shlimm): Likewise. (aarch64_ins_sve_shrimm): Likewise. * aarch64-asm-2.c: Regenerate. * aarch64-dis.h (ext_inv_limm, ext_sve_aimm, ext_sve_asimm) (ext_sve_limm_mov, ext_sve_shlimm, ext_sve_shrimm): New extractors. * aarch64-dis.c (decode_limm): New function, split out from... (aarch64_ext_limm): ...here. (aarch64_ext_inv_limm): New function. (decode_sve_aimm): Likewise. (aarch64_ext_sve_aimm): Likewise. (aarch64_ext_sve_asimm): Likewise. (aarch64_ext_sve_limm_mov): Likewise. (aarch64_top_bit): Likewise. (aarch64_ext_sve_shlimm): Likewise. (aarch64_ext_sve_shrimm): Likewise. * aarch64-dis-2.c: Regenerate. gas/ * config/tc-aarch64.c (parse_operands): Handle the new SVE integer immediate operands.
2016-09-21 23:56:57 +08:00
aarch64_sve_dupm_mov_immediate_p (uint64_t uvalue, int esize)
{
int64_t svalue = uvalue;
uint64_t upper = (uint64_t) -1 << (esize * 4) << (esize * 4);
if ((uvalue & ~upper) != uvalue && (uvalue | upper) != uvalue)
return false;
[AArch64][SVE 27/32] Add SVE integer immediate operands This patch adds the new SVE integer immediate operands. There are three kinds: - simple signed and unsigned ranges, but with new widths and positions. - 13-bit logical immediates. These have the same form as in base AArch64, but at a different bit position. In the case of the "MOV Zn.<T>, #<limm>" alias of DUPM, the logical immediate <limm> is not allowed to be a valid DUP immediate, since DUP is preferred over DUPM for constants that both instructions can handle. - a new 9-bit arithmetic immediate, of the form "<imm8>{, LSL #8}". In some contexts the operand is signed and in others it's unsigned. As an extension, we allow shifted immediates to be written as a single integer, e.g. "#256" is equivalent to "#1, LSL #8". We also use the shiftless form as the preferred disassembly, except for the special case of "#0, LSL #8" (a redundant encoding of 0). include/ * opcode/aarch64.h (AARCH64_OPND_SIMM5): New aarch64_opnd. (AARCH64_OPND_SVE_AIMM, AARCH64_OPND_SVE_ASIMM) (AARCH64_OPND_SVE_INV_LIMM, AARCH64_OPND_SVE_LIMM) (AARCH64_OPND_SVE_LIMM_MOV, AARCH64_OPND_SVE_SHLIMM_PRED) (AARCH64_OPND_SVE_SHLIMM_UNPRED, AARCH64_OPND_SVE_SHRIMM_PRED) (AARCH64_OPND_SVE_SHRIMM_UNPRED, AARCH64_OPND_SVE_SIMM5) (AARCH64_OPND_SVE_SIMM5B, AARCH64_OPND_SVE_SIMM6) (AARCH64_OPND_SVE_SIMM8, AARCH64_OPND_SVE_UIMM3) (AARCH64_OPND_SVE_UIMM7, AARCH64_OPND_SVE_UIMM8) (AARCH64_OPND_SVE_UIMM8_53): Likewise. (aarch64_sve_dupm_mov_immediate_p): Declare. opcodes/ * aarch64-tbl.h (AARCH64_OPERANDS): Add entries for the new SVE integer immediate operands. * aarch64-opc.h (FLD_SVE_immN, FLD_SVE_imm3, FLD_SVE_imm5) (FLD_SVE_imm5b, FLD_SVE_imm7, FLD_SVE_imm8, FLD_SVE_imm9) (FLD_SVE_immr, FLD_SVE_imms, FLD_SVE_tszh): New aarch64_field_kinds. * aarch64-opc.c (fields): Add corresponding entries. (operand_general_constraint_met_p): Handle the new SVE integer immediate operands. (aarch64_print_operand): Likewise. (aarch64_sve_dupm_mov_immediate_p): New function. * aarch64-opc-2.c: Regenerate. * aarch64-asm.h (ins_inv_limm, ins_sve_aimm, ins_sve_asimm) (ins_sve_limm_mov, ins_sve_shlimm, ins_sve_shrimm): New inserters. * aarch64-asm.c (aarch64_ins_limm_1): New function, split out from... (aarch64_ins_limm): ...here. (aarch64_ins_inv_limm): New function. (aarch64_ins_sve_aimm): Likewise. (aarch64_ins_sve_asimm): Likewise. (aarch64_ins_sve_limm_mov): Likewise. (aarch64_ins_sve_shlimm): Likewise. (aarch64_ins_sve_shrimm): Likewise. * aarch64-asm-2.c: Regenerate. * aarch64-dis.h (ext_inv_limm, ext_sve_aimm, ext_sve_asimm) (ext_sve_limm_mov, ext_sve_shlimm, ext_sve_shrimm): New extractors. * aarch64-dis.c (decode_limm): New function, split out from... (aarch64_ext_limm): ...here. (aarch64_ext_inv_limm): New function. (decode_sve_aimm): Likewise. (aarch64_ext_sve_aimm): Likewise. (aarch64_ext_sve_asimm): Likewise. (aarch64_ext_sve_limm_mov): Likewise. (aarch64_top_bit): Likewise. (aarch64_ext_sve_shlimm): Likewise. (aarch64_ext_sve_shrimm): Likewise. * aarch64-dis-2.c: Regenerate. gas/ * config/tc-aarch64.c (parse_operands): Handle the new SVE integer immediate operands.
2016-09-21 23:56:57 +08:00
if (esize <= 4 || (uint32_t) uvalue == (uint32_t) (uvalue >> 32))
{
svalue = (int32_t) uvalue;
if (esize <= 2 || (uint16_t) uvalue == (uint16_t) (uvalue >> 16))
{
svalue = (int16_t) uvalue;
if (esize == 1 || (uint8_t) uvalue == (uint8_t) (uvalue >> 8))
return false;
[AArch64][SVE 27/32] Add SVE integer immediate operands This patch adds the new SVE integer immediate operands. There are three kinds: - simple signed and unsigned ranges, but with new widths and positions. - 13-bit logical immediates. These have the same form as in base AArch64, but at a different bit position. In the case of the "MOV Zn.<T>, #<limm>" alias of DUPM, the logical immediate <limm> is not allowed to be a valid DUP immediate, since DUP is preferred over DUPM for constants that both instructions can handle. - a new 9-bit arithmetic immediate, of the form "<imm8>{, LSL #8}". In some contexts the operand is signed and in others it's unsigned. As an extension, we allow shifted immediates to be written as a single integer, e.g. "#256" is equivalent to "#1, LSL #8". We also use the shiftless form as the preferred disassembly, except for the special case of "#0, LSL #8" (a redundant encoding of 0). include/ * opcode/aarch64.h (AARCH64_OPND_SIMM5): New aarch64_opnd. (AARCH64_OPND_SVE_AIMM, AARCH64_OPND_SVE_ASIMM) (AARCH64_OPND_SVE_INV_LIMM, AARCH64_OPND_SVE_LIMM) (AARCH64_OPND_SVE_LIMM_MOV, AARCH64_OPND_SVE_SHLIMM_PRED) (AARCH64_OPND_SVE_SHLIMM_UNPRED, AARCH64_OPND_SVE_SHRIMM_PRED) (AARCH64_OPND_SVE_SHRIMM_UNPRED, AARCH64_OPND_SVE_SIMM5) (AARCH64_OPND_SVE_SIMM5B, AARCH64_OPND_SVE_SIMM6) (AARCH64_OPND_SVE_SIMM8, AARCH64_OPND_SVE_UIMM3) (AARCH64_OPND_SVE_UIMM7, AARCH64_OPND_SVE_UIMM8) (AARCH64_OPND_SVE_UIMM8_53): Likewise. (aarch64_sve_dupm_mov_immediate_p): Declare. opcodes/ * aarch64-tbl.h (AARCH64_OPERANDS): Add entries for the new SVE integer immediate operands. * aarch64-opc.h (FLD_SVE_immN, FLD_SVE_imm3, FLD_SVE_imm5) (FLD_SVE_imm5b, FLD_SVE_imm7, FLD_SVE_imm8, FLD_SVE_imm9) (FLD_SVE_immr, FLD_SVE_imms, FLD_SVE_tszh): New aarch64_field_kinds. * aarch64-opc.c (fields): Add corresponding entries. (operand_general_constraint_met_p): Handle the new SVE integer immediate operands. (aarch64_print_operand): Likewise. (aarch64_sve_dupm_mov_immediate_p): New function. * aarch64-opc-2.c: Regenerate. * aarch64-asm.h (ins_inv_limm, ins_sve_aimm, ins_sve_asimm) (ins_sve_limm_mov, ins_sve_shlimm, ins_sve_shrimm): New inserters. * aarch64-asm.c (aarch64_ins_limm_1): New function, split out from... (aarch64_ins_limm): ...here. (aarch64_ins_inv_limm): New function. (aarch64_ins_sve_aimm): Likewise. (aarch64_ins_sve_asimm): Likewise. (aarch64_ins_sve_limm_mov): Likewise. (aarch64_ins_sve_shlimm): Likewise. (aarch64_ins_sve_shrimm): Likewise. * aarch64-asm-2.c: Regenerate. * aarch64-dis.h (ext_inv_limm, ext_sve_aimm, ext_sve_asimm) (ext_sve_limm_mov, ext_sve_shlimm, ext_sve_shrimm): New extractors. * aarch64-dis.c (decode_limm): New function, split out from... (aarch64_ext_limm): ...here. (aarch64_ext_inv_limm): New function. (decode_sve_aimm): Likewise. (aarch64_ext_sve_aimm): Likewise. (aarch64_ext_sve_asimm): Likewise. (aarch64_ext_sve_limm_mov): Likewise. (aarch64_top_bit): Likewise. (aarch64_ext_sve_shlimm): Likewise. (aarch64_ext_sve_shrimm): Likewise. * aarch64-dis-2.c: Regenerate. gas/ * config/tc-aarch64.c (parse_operands): Handle the new SVE integer immediate operands.
2016-09-21 23:56:57 +08:00
}
}
if ((svalue & 0xff) == 0)
svalue /= 256;
return svalue < -128 || svalue >= 128;
}
/* Return true if a CPU with the AARCH64_FEATURE_* bits in CPU_VARIANT
supports the instruction described by INST. */
bool
aarch64_cpu_supports_inst_p (aarch64_feature_set cpu_variant,
aarch64_inst *inst)
{
if (!inst->opcode->avariant
|| !AARCH64_CPU_HAS_ALL_FEATURES (cpu_variant, *inst->opcode->avariant))
return false;
if (inst->opcode->iclass == sme_fp_sd
&& inst->operands[0].qualifier == AARCH64_OPND_QLF_S_D
&& !AARCH64_CPU_HAS_FEATURE (cpu_variant, SME_F64F64))
return false;
if (inst->opcode->iclass == sme_int_sd
&& inst->operands[0].qualifier == AARCH64_OPND_QLF_S_D
&& !AARCH64_CPU_HAS_FEATURE (cpu_variant, SME_I16I64))
return false;
return true;
}
/* Include the opcode description table as well as the operand description
table. */
#define VERIFIER(x) verify_##x
#include "aarch64-tbl.h"