In 7545aa2dd2 ("gas: improve interaction between read_a_source_file()
and s_linefile()") I didn't pay attention to the dual purpose of the
nul character previously used. This was to a fair degree because of the
open-coding of certain operations. Insert the earlier found line
terminator instead of a hard-coded newline, and do so early in this
special case (bypassing the later general insertion point). Plus
properly use sb_terminate() to mark the end of the string. (Note that
saved_eol_char was misnamed: Without calling sb_terminate() there's
simply random data at that position in the buffer.)
In 162c6aef1f ("gas: fold symbol table entries generated for
.startof.() / .sizeof.()") I screwed up quite badly, inverting the case
sensitive and case insensitive comparison functions.
read_a_source_file() would bump line numbers only when seeing a newline,
whereas is_end_of_line[] indicates further end-of-line characters, in
particular the nul character. s_linefile() attempts to compensate for
the bump, but was too aggressive with this so far: It should only adjust
when a newline ends the line. To facilitate such a check, the check for
nothing else on the line needs to move ahead, which luckily is easily
possible: The relevant two conditions match, and the function can
simply return from the body of that earlier instance of the conditional.
The more strict treatment in s_linefile() then requires an adjustment
to buffer_and_nest()'s invocation of the function: The line terminator
now needs to be a newline, not nul.
Move all the function local static state variables to file scope,
in order to tidy memory on exit and to reinit everything for that
annoying oss-fuzz. Also fix a couple memory leaks.
* read.h (read_begin, read_end): Declare.
* read.c (read_begin): Call stabs_begin.
(read_end): Call stabs_end.
* stabs.c (stabs_begin, stabs_end): New functions.
(in_dot_func_p): Delete, use current_function_label instead.
(cached_sec): Move from s_stab_generic.
(last_asm_file, file_label_count): Move from generate_asm_file.
(line_label_count, prev_lineno, prev_line_file): Move from
stabs_generate_asm_lineno.
(void_emitted_p): Move from stabs_generate_asm_func.
(endfunc_label_count): Move from stabs_generate_asm_endfunc.
(stabs_generate_asm_lineno): Simplify setting of
prev_line_file.
(stabs_generate_asm_func): Don't leak current_function_label.
(stabs_generate_asm_endfunc): Likewise.
There is some inconsistency between the behaviour of objdump -D and
objdump -s, both supposedly operating on all sections by default.
objdump -s ignores bss sections, while objdump -D dissassembles the
zeros. Fix this by making objdump -D ignore bss sections too.
Furthermore, "objdump -s -j .bss" doesn't dump .bss as it should,
since the user is specifically asking to look at all those zeros.
This change does find some tests that used objdump -D with expected
output in bss-style sections. I've updated all the msp430 tests that
just wanted to find a non-empty section to look at section headers
instead, making the tests slightly more stringent. The ppc xcoff and
spu tests are fixed by adding -j options to objdump, which makes the
tests somewhat more lenient.
binutils/
* objdump.c (disassemble_section): Ignore sections without
contents, unless overridden by -j.
(dump_section): Allow -j to override the default of not
displaying sections without contents.
* doc/binutils.texi (objdump options): Update -D, -s and -j
description.
gas/
* testsuite/gas/ppc/xcoff-tls-32.d: Select wanted objdump
sections with -j.
* testsuite/gas/ppc/xcoff-tls-64.d: Likewise.
ld/
* testsuite/ld-msp430-elf/main-bss-lower.d,
* testsuite/ld-msp430-elf/main-bss-upper.d,
* testsuite/ld-msp430-elf/main-const-lower.d,
* testsuite/ld-msp430-elf/main-const-upper.d,
* testsuite/ld-msp430-elf/main-text-lower.d,
* testsuite/ld-msp430-elf/main-text-upper.d,
* testsuite/ld-msp430-elf/main-var-lower.d,
* testsuite/ld-msp430-elf/main-var-upper.d: Expect -wh output.
* testsuite/ld-msp430-elf/msp430-elf.exp: Use objdump -wh
rather than objdump -D or objdump -d with tests checking for
non-empty given sections.
* testsuite/ld-spu/ear.d,
* testsuite/ld-spu/icache1.d,
* testsuite/ld-spu/ovl.d,
* testsuite/ld-spu/ovl2.d: Select wanted objdump sections.
"Local labels are never absolute" says the comment. Except when they
are. Testcase
.offset
0:
a=0b
I don't see any particular reason to disallow local labels inside
struct definitions, so delete the comment and assertions.
* expr.c (integer_constant): Delete local label assertions.
The attribute really specifies that the sum of register and memory
operands is 4. Express it like that in most places, while using the 2nd
(apart from XOP) CPU feature flags (FMA4) in reversed operand matching
logic.
With the use in build_modrm_byte() gone, part of an assertion there
also becomes meaningless - simplify that at the same time.
With all uses of the opcode modifier field gone, also drop that.
The few XOP insns which used it wrongly didn't have VexVVVV specified.
With that added, the only further missing piece to use more generic code
elsewhere is SwapSources - see e.g. the BMI2 insns for similar operand
patterns.
With the only users gone, drop the #define as well as the special case
code.
The VPROT* forms with an immediate operand are entirely standard in the
way their ModR/M bytes are built. There's no reason to invoke special
case code. With that the handling of an immediate there can also be
dropped; it was partially bogus anyway, as in its "no memory operands"
portion it ignores the possibility of an immediate operand (which was
okay only because that case was already handled by more generic code).
This really isn't a "modifier" and rather ought to live next to the base
opcode anyway. Use the bits we presently have available to fit in the
field, renaming it to opcode_space. As an intended side effect this
helps readability at the use sites, by shortening the references quite a
bit.
In generated code arrange for human readable output, by using the
SPACE_* constants there rather than raw numbers. This may aid debugging
down the road.
Fold adjacent comparisons when, by ORing in a certain mask, the same
effect can be achieved by a single one. In load_insn_p() this extends
to further uses of an already available local variable.
Now that we have identifiers for the mnemonic strings we can avoid
opcode based comparisons, for (in many cases) being more expensive and
(in a few cases) being a little fragile and not self-documenting.
Note that the MOV optimization can be engaged by the earlier LEA one,
and hence LEA also needs checking for there.
This is the assertion
know (*input_line_pointer != ' ');
after calling operand.
The usual exit from operand calls SKIP_ALL_WHITESPACE.
* expr.c (operand): Call SKIP_ALL_WHITESPACE after call to expr.
Swapping operands for commutative insns occurs outside of
optimize_encoding() and hence needs explicit checking for a request to
avoid any optimizations.
Dropping a meaningless segment prefix occurs outside of
optimize_encoding() and hence needs explicit checking for a request to
avoid any optimizations.
Insn width granularity being 16 bits, producing byte granular output
isn't very useful. With there being a way to specific otherwise
unknown insns to the assembler, use that same representation (to be
precise: its <length>,<encoding> flavor) for disassembly.
Don't define this. It is defined just before elf-bfd.h is included,
but doesn't have any relevance there. Instead is for aout64.h where
the default is 4 anyway.
Provide a way for config/obj-* to clean up at end of assembly, and do
so for ELF.
* obj.h (struct format_ops): Add "end".
* config/obj-aout.c (aout_format_ops): Init new field.
* config/obj-coff.c (coff_format_ops): Likewise.
* config/obj-ecoff.c (ecoff_format_ops): Likewise.
* config/obj-elf.c (elf_format_ops): Likewise.
(elf_begin): Move later in file. Clear some more variables.
(comment_section): Make file scope.
(free_section_idx): Rewrite.
(elf_adjust_symtab): Expand str_htab_create call and use
free_section_idx as delete function.
(elf_frob_file_after_relocs): Don't clean up groups.indexes here.
(elf_end): New function.
* config/obj-elf.h (obj_end): Define.
* config/obj-multi.h (obj_end): Define.
* output-file.c (output_file_close): Call obj_end.
This function has a gas_assert, ie. possible call to as_abort, which
calls as_report_context, which calls as_info_where.
* messages.c (as_info_where): Don't gas_assert.
Currently, when using -gdwarf-2, gas emits a v3 .debug_line contribution.
Fix this by emitting a v2 .debug_line contribution instead.
gas/ChangeLog:
2023-01-31 Tom de Vries <tdevries@suse.de>
PR 23941
* dwarf2dbg.c (DWARF2_LINE_VERSION): Set to 2 for -gdwarf-2.
(DWARF2_LINE_OPCODE_BASE): Handle DWARF2_LINE_VERSION == 2.
(dwarf2_directive_loc): Bump dwarf_level when encountering
v3 .loc options.
(out_debug_line): Don't output v3 standard opcodes for v2.
* testsuite/gas/i386/debug1.d: Update.
* testsuite/gas/i386/dwarf2-line-1.d: Update.
* testsuite/gas/i386/dwarf2-line-4.d: Update.
I noticed that some of the Power6 DFP instructions were not covered by
the assembler tests. I've added a new test file which I believe
covers all the DFP Power6 instructions.
The existing gas/testsuite/gas/ppc/power6.d test is called:
POWER6 tests (includes DFP and Altivec)
And does cover some of the DFP instructions. But, given the number of
additional instructions I'm adding I opted to add a whole new test
file. I've left the original power6.d unchanged, so there is now some
overlap, but I don't think that should hurt much.
Along with the normal JAL alias, the C-extension one should have been
moved as well by 839189bc93 ("RISC-V: re-arrange opcode table for
consistent alias handling"), for the assembler to actually be able to
use it where/when possible.
Since neither this nor any other compressed branch insn was being tested
so far, take the opportunity and introduce a new testcase covering those.
Ideally we'd do away with this somewhat questionable adjustment (leaving
i.types[] untouched). That's non-trivial though as it looks, so only
- move the logic into process_operands(), putting it closer to related
logic and eliminating a conditional for operand-less insns,
- make it consistent (i.e. also affect %xmm0), eliminating an ugly
special case later in the function.
All (there are just two) SSE2AVX templates with %xmm0 as first operand
also specify VEX3SOURCES. Hence there's no need for an "else" to the
respective if(), and the if() itself can become an assertion.
Now that we have identifiers for the mnemonic strings we can avoid
opcode based comparisons, for (in many cases) being more expensive and
(in a few cases) being a little fragile and not self-documenting.
bfd_cache_close_all is good for closing file descriptors, but doesn't
do the cleanup of bfd memory as in bfd_close_all_done.
PR 13056
* output-file.c (output_file_close): Call bfd_close_all_done,
not bfd_cache_close_all.
This tidies memory allocated for entries in macro_hash. Freeing the
macro name requires a little restructuring of the define_macro
interface due to the name being used in the error message, and exposed
the fact that the name and other fields were not initialised by the
iq2000 backend.
There is also a fix for
.macro .macro
.endm
.macro .macro
.endm
which prior to this patch reported
mac.s:1: Warning: attempt to redefine pseudo-op `.macro' ignored
mac.s:3: Error: Macro `.macro' was already defined
rather than reporting the attempt to redefine twice.
* macro.c (macro_del_f): New function.
(macro_init): Use it when creating macro_hash.
(free_macro): Free macro name too.
(define_macro): Return the macro_entry, remove idx, file, line and
namep params. Call as_where. Report errors here. Delete macro
from macro_hash on attempt to redefined pseudo-op.
(delete_macro): Don't call free_macro.
* macro.h (define_macro): Update prototype.
* read.c (s_macro): Adjust to suit.
* config/tc-iq2000.c (iq2000_add_macro): Init all fields of
macro_entry.
There's a comment a few lines earlier saying that demand_copy_C_string
has already reported an error if it returns NULL. Given the proximity
I decided not to duplicate the comment.
* dwarf2dbg.c (dwarf2_directive_filename): Check return of
demand_copy_C_string for file.
This patch adds support for the .secidx directive and its corresponding
relocation to aarch64-w64-mingw32. As with x86, this is a two-byte LE
integer which gets filled in with the 1-based index of the output
section that a symbol ends up in.
This is needed for PDBs, which represent addresses as a .secrel32,
.secidx pair.
The test is substantially the same as for amd64, but with changes made
for padding and alignment.
Now that we have identifiers for the mnemonic strings we can avoid
strcmp() in a number of places, comparing the offsets into the mnemonic
string table instead. While doing this also
- convert a leftover strncmp() to startswith() (apparently an oversight
when rebasing the original patch introducing the startswith() uses),
- use the new shorthand for current_templates->start also elsewhere in
md_assemble() (valid up to the point where match_template() is
called).
Using full pointers to reference the insn mnemonic strings is not very
efficient. With overall string size presently just slightly over 20k,
even a 16-bit value would suffice. Use "unsigned int" for now, as
there's no good use we could presently make of the otherwise saved 16
bits.
For 64-bit builds this reduces table size by 6.25% (prior to the recent
ISA extension additions it would have been 12.5%), with a similar effect
on cache occupation of table entries accessed. For PIE builds of gas
this also reduces the number of base relocations quite a bit (obviously
independent of bitness).
In preparation for changing the representation of the "name" field
introduce a wrapper function. This keeps the mechanical change separate
from the functional one.
We noticed that a warning message about the use of scalar fp16
instructions being UNPREDICTABLE when conditionalized in an IT
block referenced the specific A-class architecture revision
ARMv8.2-A.
Many of these instructions are now also part of ARMv8.1-M, so
the warning message had become misleading. Here we just change
the message to not specify an architecture revision at all and
update all testing accordingly. This was done with a simple
find-n-replace within the binutils sources. No tests have
regressed for the arm target.
gas/ChangeLog:
* config/tc-arm.c (do_scalar_fp16_v82_encode): Remove
ARMv8.2-A from the warning message.
(do_neon_movhf): Likewise
* testsuite/gas/arm/armv8-2-fp16-scalar-bad.l: Likewise
* testsuite/gas/arm/mve-vaddsub-it-bad.l: Likewise
* testsuite/gas/arm/mve-vcvtne-it-bad.l: Likewise
* testsuite/gas/arm/mve-vcvtne-it.d: Likewise
Previously we had experienced issues with assembling a "VCVTNE" instruction
in the presence of the MVE architecture extension, because it could be
interpreted both as:
* The base instruction VCVT + NE for IT predication when inside an IT block.
* The MVE instruction VCVTN + E in the Else of a VPT block.
Given a C reproducer of:
```
int test_function(float value)
{
int ret_val = 10;
if (value != 0.0)
{
ret_val = (int) value;
}
return ret_val;
}
```
GCC generates a VCVTNE instruction based on the `truncsisf2_vfp`
pattern, which will look like:
`vcvtne.s32.f32 s-reg, s-reg`
This still triggers an error due to being misidentified as "vcvtn+e"
Similar errors were found with other type combinations and instruction
patterns (these have all been added to the testing of this patch).
This class of errors was previously worked around by:
https://sourceware.org/pipermail/binutils/2020-August/112728.html
which addressed this by looking at the operand types, however,
that isn't adequate to cover all the extra cases that have been
found. Instead, we add some special-casing logic earlier when
the instructions are parsed that is conditional on whether we are
in a VPT block or not, when the instruction is parsed.
gas/ChangeLog:
* config/tc-arm.c (opcode_lookup): Add special vcvtn handling.
* testsuite/gas/arm/mve-vcvtne-it-bad.l: Add further testing.
* testsuite/gas/arm/mve-vcvtne-it-bad.s: Likewise.
* testsuite/gas/arm/mve-vcvtne-it.d: Likewise.
* testsuite/gas/arm/mve-vcvtne-it.s: Likewise.
