linux/tools/objtool/check.c

4105 lines
95 KiB
C
Raw Normal View History

treewide: Replace GPLv2 boilerplate/reference with SPDX - rule 13 Based on 2 normalized pattern(s): this program 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 2 of the license or at your option any later version this program 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 if not see http www gnu org licenses this program 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 2 of the license or at your option any later version this program 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 [based] [from] [clk] [highbank] [c] you should have received a copy of the gnu general public license along with this program if not see http www gnu org licenses extracted by the scancode license scanner the SPDX license identifier GPL-2.0-or-later has been chosen to replace the boilerplate/reference in 355 file(s). Signed-off-by: Thomas Gleixner <tglx@linutronix.de> Reviewed-by: Kate Stewart <kstewart@linuxfoundation.org> Reviewed-by: Jilayne Lovejoy <opensource@jilayne.com> Reviewed-by: Steve Winslow <swinslow@gmail.com> Reviewed-by: Allison Randal <allison@lohutok.net> Cc: linux-spdx@vger.kernel.org Link: https://lkml.kernel.org/r/20190519154041.837383322@linutronix.de Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2019-05-19 21:51:43 +08:00
// SPDX-License-Identifier: GPL-2.0-or-later
/*
* Copyright (C) 2015-2017 Josh Poimboeuf <jpoimboe@redhat.com>
*/
#include <string.h>
#include <stdlib.h>
#include <inttypes.h>
#include <sys/mman.h>
objtool: Rework header include paths Currently objtool headers are being included either by their base name or included via ../ from a parent directory. In case of a base name usage: #include "warn.h" #include "arch_elf.h" it does not make it apparent from which directory the file comes from. To make it slightly better, and actually to avoid name clashes some arch specific files have "arch_" suffix. And files from an arch folder have to revert to including via ../ e.g: #include "../../elf.h" With additional architectures support and the code base growth there is a need for clearer headers naming scheme for multiple reasons: 1. to make it instantly obvious where these files come from (objtool itself / objtool arch|generic folders / some other external files), 2. to avoid name clashes of objtool arch specific headers, potential obtool arch generic headers and the system header files (there is /usr/include/elf.h already), 3. to avoid ../ includes and improve code readability. 4. to give a warm fuzzy feeling to developers who are mostly kernel developers and are accustomed to linux kernel headers arranging scheme. Doesn't this make it instantly obvious where are these files come from? #include <objtool/warn.h> #include <arch/elf.h> And doesn't it look nicer to avoid ugly ../ includes? Which also guarantees this is elf.h from the objtool and not /usr/include/elf.h. #include <objtool/elf.h> This patch defines and implements new objtool headers arranging scheme. Which is: - all generic headers go to include/objtool (similar to include/linux) - all arch headers go to arch/$(SRCARCH)/include/arch (to get arch prefix). This is similar to linux arch specific "asm/*" headers but we are not abusing "asm" name and calling it what it is. This also helps to prevent name clashes (arch is not used in system headers or kernel exports). To bring objtool to this state the following things are done: 1. current top level tools/objtool/ headers are moved into include/objtool/ subdirectory, 2. arch specific headers, currently only arch/x86/include/ are moved into arch/x86/include/arch/ and were stripped of "arch_" suffix, 3. new -I$(srctree)/tools/objtool/include include path to make includes like <objtool/warn.h> possible, 4. rewriting file includes, 5. make git not to ignore include/objtool/ subdirectory. Signed-off-by: Vasily Gorbik <gor@linux.ibm.com> Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org> Acked-by: Masami Hiramatsu <mhiramat@kernel.org> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
2020-11-13 07:03:32 +08:00
#include <arch/elf.h>
#include <objtool/builtin.h>
#include <objtool/cfi.h>
#include <objtool/arch.h>
#include <objtool/check.h>
#include <objtool/special.h>
#include <objtool/warn.h>
#include <objtool/endianness.h>
#include <linux/objtool.h>
#include <linux/hashtable.h>
#include <linux/kernel.h>
#include <linux/static_call_types.h>
struct alternative {
struct list_head list;
struct instruction *insn;
bool skip_orig;
};
static unsigned long nr_cfi, nr_cfi_reused, nr_cfi_cache;
static struct cfi_init_state initial_func_cfi;
static struct cfi_state init_cfi;
static struct cfi_state func_cfi;
struct instruction *find_insn(struct objtool_file *file,
struct section *sec, unsigned long offset)
{
struct instruction *insn;
hash_for_each_possible(file->insn_hash, insn, hash, sec_offset_hash(sec, offset)) {
if (insn->sec == sec && insn->offset == offset)
return insn;
}
return NULL;
}
static struct instruction *next_insn_same_sec(struct objtool_file *file,
struct instruction *insn)
{
struct instruction *next = list_next_entry(insn, list);
if (!next || &next->list == &file->insn_list || next->sec != insn->sec)
return NULL;
return next;
}
static struct instruction *next_insn_same_func(struct objtool_file *file,
struct instruction *insn)
{
struct instruction *next = list_next_entry(insn, list);
struct symbol *func = insn->func;
if (!func)
return NULL;
if (&next->list != &file->insn_list && next->func == func)
return next;
/* Check if we're already in the subfunction: */
if (func == func->cfunc)
return NULL;
/* Move to the subfunction: */
return find_insn(file, func->cfunc->sec, func->cfunc->offset);
}
static struct instruction *prev_insn_same_sym(struct objtool_file *file,
struct instruction *insn)
{
struct instruction *prev = list_prev_entry(insn, list);
if (&prev->list != &file->insn_list && prev->func == insn->func)
return prev;
return NULL;
}
#define func_for_each_insn(file, func, insn) \
for (insn = find_insn(file, func->sec, func->offset); \
insn; \
insn = next_insn_same_func(file, insn))
#define sym_for_each_insn(file, sym, insn) \
for (insn = find_insn(file, sym->sec, sym->offset); \
insn && &insn->list != &file->insn_list && \
insn->sec == sym->sec && \
insn->offset < sym->offset + sym->len; \
insn = list_next_entry(insn, list))
#define sym_for_each_insn_continue_reverse(file, sym, insn) \
for (insn = list_prev_entry(insn, list); \
&insn->list != &file->insn_list && \
insn->sec == sym->sec && insn->offset >= sym->offset; \
insn = list_prev_entry(insn, list))
#define sec_for_each_insn_from(file, insn) \
for (; insn; insn = next_insn_same_sec(file, insn))
#define sec_for_each_insn_continue(file, insn) \
for (insn = next_insn_same_sec(file, insn); insn; \
insn = next_insn_same_sec(file, insn))
static bool is_jump_table_jump(struct instruction *insn)
{
struct alt_group *alt_group = insn->alt_group;
if (insn->jump_table)
return true;
/* Retpoline alternative for a jump table? */
return alt_group && alt_group->orig_group &&
alt_group->orig_group->first_insn->jump_table;
}
static bool is_sibling_call(struct instruction *insn)
{
/*
* Assume only ELF functions can make sibling calls. This ensures
* sibling call detection consistency between vmlinux.o and individual
* objects.
*/
if (!insn->func)
return false;
/* An indirect jump is either a sibling call or a jump to a table. */
if (insn->type == INSN_JUMP_DYNAMIC)
return !is_jump_table_jump(insn);
/* add_jump_destinations() sets insn->call_dest for sibling calls. */
return (is_static_jump(insn) && insn->call_dest);
}
/*
* This checks to see if the given function is a "noreturn" function.
*
* For global functions which are outside the scope of this object file, we
* have to keep a manual list of them.
*
* For local functions, we have to detect them manually by simply looking for
* the lack of a return instruction.
*/
static bool __dead_end_function(struct objtool_file *file, struct symbol *func,
int recursion)
{
int i;
struct instruction *insn;
bool empty = true;
/*
* Unfortunately these have to be hard coded because the noreturn
* attribute isn't provided in ELF data.
*/
static const char * const global_noreturns[] = {
"__stack_chk_fail",
"panic",
"do_exit",
"do_task_dead",
"kthread_exit",
"make_task_dead",
"__module_put_and_kthread_exit",
"kthread_complete_and_exit",
"__reiserfs_panic",
"lbug_with_loc",
"fortify_panic",
"usercopy_abort",
"machine_real_restart",
"rewind_stack_and_make_dead",
"kunit_try_catch_throw",
"xen_start_kernel",
"cpu_bringup_and_idle",
"do_group_exit",
"stop_this_cpu",
"__invalid_creds",
objtool: Mark __ubsan_handle_builtin_unreachable() as noreturn fs/ntfs3/ntfs3.prelink.o: warning: objtool: ni_read_frame() falls through to next function ni_readpage_cmpr.cold() That is in fact: 000000000000124a <ni_read_frame.cold>: 124a: 44 89 e0 mov %r12d,%eax 124d: 0f b6 55 98 movzbl -0x68(%rbp),%edx 1251: 48 c7 c7 00 00 00 00 mov $0x0,%rdi 1254: R_X86_64_32S .data+0x1380 1258: 48 89 c6 mov %rax,%rsi 125b: e8 00 00 00 00 call 1260 <ni_read_frame.cold+0x16> 125c: R_X86_64_PLT32 __ubsan_handle_shift_out_of_bounds-0x4 1260: 48 8d 7d cc lea -0x34(%rbp),%rdi 1264: e8 00 00 00 00 call 1269 <ni_read_frame.cold+0x1f> 1265: R_X86_64_PLT32 __tsan_read4-0x4 1269: 8b 45 cc mov -0x34(%rbp),%eax 126c: e9 00 00 00 00 jmp 1271 <ni_read_frame.cold+0x27> 126d: R_X86_64_PC32 .text+0x19109 1271: 48 8b 75 a0 mov -0x60(%rbp),%rsi 1275: 48 63 d0 movslq %eax,%rdx 1278: 48 c7 c7 00 00 00 00 mov $0x0,%rdi 127b: R_X86_64_32S .data+0x13a0 127f: 89 45 88 mov %eax,-0x78(%rbp) 1282: e8 00 00 00 00 call 1287 <ni_read_frame.cold+0x3d> 1283: R_X86_64_PLT32 __ubsan_handle_shift_out_of_bounds-0x4 1287: 8b 45 88 mov -0x78(%rbp),%eax 128a: e9 00 00 00 00 jmp 128f <ni_read_frame.cold+0x45> 128b: R_X86_64_PC32 .text+0x19098 128f: 48 c7 c7 00 00 00 00 mov $0x0,%rdi 1292: R_X86_64_32S .data+0x11f0 1296: e8 00 00 00 00 call 129b <ni_readpage_cmpr.cold> 1297: R_X86_64_PLT32 __ubsan_handle_builtin_unreachable-0x4 000000000000129b <ni_readpage_cmpr.cold>: Tell objtool that __ubsan_handle_builtin_unreachable() is a noreturn. Reported-by: kernel test robot <lkp@intel.com> Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Link: https://lkml.kernel.org/r/20220502091514.GB479834@worktop.programming.kicks-ass.net
2022-05-02 17:15:14 +08:00
"cpu_startup_entry",
"__ubsan_handle_builtin_unreachable",
"ex_handler_msr_mce",
};
if (!func)
return false;
if (func->bind == STB_WEAK)
return false;
if (func->bind == STB_GLOBAL)
for (i = 0; i < ARRAY_SIZE(global_noreturns); i++)
if (!strcmp(func->name, global_noreturns[i]))
return true;
if (!func->len)
return false;
insn = find_insn(file, func->sec, func->offset);
if (!insn->func)
return false;
func_for_each_insn(file, func, insn) {
empty = false;
if (insn->type == INSN_RETURN)
return false;
}
if (empty)
return false;
/*
* A function can have a sibling call instead of a return. In that
* case, the function's dead-end status depends on whether the target
* of the sibling call returns.
*/
func_for_each_insn(file, func, insn) {
if (is_sibling_call(insn)) {
struct instruction *dest = insn->jump_dest;
if (!dest)
/* sibling call to another file */
return false;
/* local sibling call */
if (recursion == 5) {
/*
* Infinite recursion: two functions have
* sibling calls to each other. This is a very
* rare case. It means they aren't dead ends.
*/
return false;
}
return __dead_end_function(file, dest->func, recursion+1);
}
}
return true;
}
static bool dead_end_function(struct objtool_file *file, struct symbol *func)
{
return __dead_end_function(file, func, 0);
}
static void init_cfi_state(struct cfi_state *cfi)
{
int i;
objtool: Handle GCC stack pointer adjustment bug Arnd Bergmann reported the following warning with GCC 7.1.1: fs/fs_pin.o: warning: objtool: pin_kill()+0x139: stack state mismatch: cfa1=7+88 cfa2=7+96 And the kbuild robot reported the following warnings with GCC 5.4.1: fs/fs_pin.o: warning: objtool: pin_kill()+0x182: return with modified stack frame fs/quota/dquot.o: warning: objtool: dquot_alloc_inode()+0x140: stack state mismatch: cfa1=7+120 cfa2=7+128 fs/quota/dquot.o: warning: objtool: dquot_free_inode()+0x11a: stack state mismatch: cfa1=7+112 cfa2=7+120 Those warnings are caused by an unusual GCC non-optimization where it uses an intermediate register to adjust the stack pointer. It does: lea 0x8(%rsp), %rcx ... mov %rcx, %rsp Instead of the obvious: add $0x8, %rsp It makes no sense to use an intermediate register, so I opened a GCC bug to track it: https://gcc.gnu.org/bugzilla/show_bug.cgi?id=81813 But it's not exactly a high-priority bug and it looks like we'll be stuck with this issue for a while. So for now we have to track register values when they're loaded with stack pointer offsets. This is kind of a big workaround for a tiny problem, but c'est la vie. I hope to eventually create a GCC plugin to implement a big chunk of objtool's functionality. Hopefully at that point we'll be able to remove of a lot of these GCC-isms from the objtool code. Reported-by: Arnd Bergmann <arnd@arndb.de> Reported-by: kbuild test robot <fengguang.wu@intel.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Thomas Gleixner <tglx@linutronix.de> Link: http://lkml.kernel.org/r/6a41a96884c725e7f05413bb7df40cfe824b2444.1504028945.git.jpoimboe@redhat.com Signed-off-by: Ingo Molnar <mingo@kernel.org>
2017-08-30 01:51:03 +08:00
for (i = 0; i < CFI_NUM_REGS; i++) {
cfi->regs[i].base = CFI_UNDEFINED;
cfi->vals[i].base = CFI_UNDEFINED;
objtool: Handle GCC stack pointer adjustment bug Arnd Bergmann reported the following warning with GCC 7.1.1: fs/fs_pin.o: warning: objtool: pin_kill()+0x139: stack state mismatch: cfa1=7+88 cfa2=7+96 And the kbuild robot reported the following warnings with GCC 5.4.1: fs/fs_pin.o: warning: objtool: pin_kill()+0x182: return with modified stack frame fs/quota/dquot.o: warning: objtool: dquot_alloc_inode()+0x140: stack state mismatch: cfa1=7+120 cfa2=7+128 fs/quota/dquot.o: warning: objtool: dquot_free_inode()+0x11a: stack state mismatch: cfa1=7+112 cfa2=7+120 Those warnings are caused by an unusual GCC non-optimization where it uses an intermediate register to adjust the stack pointer. It does: lea 0x8(%rsp), %rcx ... mov %rcx, %rsp Instead of the obvious: add $0x8, %rsp It makes no sense to use an intermediate register, so I opened a GCC bug to track it: https://gcc.gnu.org/bugzilla/show_bug.cgi?id=81813 But it's not exactly a high-priority bug and it looks like we'll be stuck with this issue for a while. So for now we have to track register values when they're loaded with stack pointer offsets. This is kind of a big workaround for a tiny problem, but c'est la vie. I hope to eventually create a GCC plugin to implement a big chunk of objtool's functionality. Hopefully at that point we'll be able to remove of a lot of these GCC-isms from the objtool code. Reported-by: Arnd Bergmann <arnd@arndb.de> Reported-by: kbuild test robot <fengguang.wu@intel.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Thomas Gleixner <tglx@linutronix.de> Link: http://lkml.kernel.org/r/6a41a96884c725e7f05413bb7df40cfe824b2444.1504028945.git.jpoimboe@redhat.com Signed-off-by: Ingo Molnar <mingo@kernel.org>
2017-08-30 01:51:03 +08:00
}
cfi->cfa.base = CFI_UNDEFINED;
cfi->drap_reg = CFI_UNDEFINED;
cfi->drap_offset = -1;
}
static void init_insn_state(struct objtool_file *file, struct insn_state *state,
struct section *sec)
{
memset(state, 0, sizeof(*state));
init_cfi_state(&state->cfi);
/*
* We need the full vmlinux for noinstr validation, otherwise we can
* not correctly determine insn->call_dest->sec (external symbols do
* not have a section).
*/
if (opts.link && opts.noinstr && sec)
state->noinstr = sec->noinstr;
}
static struct cfi_state *cfi_alloc(void)
{
struct cfi_state *cfi = calloc(sizeof(struct cfi_state), 1);
if (!cfi) {
WARN("calloc failed");
exit(1);
}
nr_cfi++;
return cfi;
}
static int cfi_bits;
static struct hlist_head *cfi_hash;
static inline bool cficmp(struct cfi_state *cfi1, struct cfi_state *cfi2)
{
return memcmp((void *)cfi1 + sizeof(cfi1->hash),
(void *)cfi2 + sizeof(cfi2->hash),
sizeof(struct cfi_state) - sizeof(struct hlist_node));
}
static inline u32 cfi_key(struct cfi_state *cfi)
{
return jhash((void *)cfi + sizeof(cfi->hash),
sizeof(*cfi) - sizeof(cfi->hash), 0);
}
static struct cfi_state *cfi_hash_find_or_add(struct cfi_state *cfi)
{
struct hlist_head *head = &cfi_hash[hash_min(cfi_key(cfi), cfi_bits)];
struct cfi_state *obj;
hlist_for_each_entry(obj, head, hash) {
if (!cficmp(cfi, obj)) {
nr_cfi_cache++;
return obj;
}
}
obj = cfi_alloc();
*obj = *cfi;
hlist_add_head(&obj->hash, head);
return obj;
}
static void cfi_hash_add(struct cfi_state *cfi)
{
struct hlist_head *head = &cfi_hash[hash_min(cfi_key(cfi), cfi_bits)];
hlist_add_head(&cfi->hash, head);
}
static void *cfi_hash_alloc(unsigned long size)
{
cfi_bits = max(10, ilog2(size));
cfi_hash = mmap(NULL, sizeof(struct hlist_head) << cfi_bits,
PROT_READ|PROT_WRITE,
MAP_PRIVATE|MAP_ANON, -1, 0);
if (cfi_hash == (void *)-1L) {
WARN("mmap fail cfi_hash");
cfi_hash = NULL;
} else if (opts.stats) {
printf("cfi_bits: %d\n", cfi_bits);
}
return cfi_hash;
}
static unsigned long nr_insns;
static unsigned long nr_insns_visited;
/*
* Call the arch-specific instruction decoder for all the instructions and add
* them to the global instruction list.
*/
static int decode_instructions(struct objtool_file *file)
{
struct section *sec;
struct symbol *func;
unsigned long offset;
struct instruction *insn;
int ret;
for_each_sec(file, sec) {
if (!(sec->sh.sh_flags & SHF_EXECINSTR))
continue;
if (strcmp(sec->name, ".altinstr_replacement") &&
strcmp(sec->name, ".altinstr_aux") &&
strncmp(sec->name, ".discard.", 9))
sec->text = true;
if (!strcmp(sec->name, ".noinstr.text") ||
!strcmp(sec->name, ".entry.text") ||
!strncmp(sec->name, ".text.__x86.", 12))
sec->noinstr = true;
for (offset = 0; offset < sec->sh.sh_size; offset += insn->len) {
insn = malloc(sizeof(*insn));
if (!insn) {
WARN("malloc failed");
return -1;
}
memset(insn, 0, sizeof(*insn));
INIT_LIST_HEAD(&insn->alts);
INIT_LIST_HEAD(&insn->stack_ops);
INIT_LIST_HEAD(&insn->call_node);
insn->sec = sec;
insn->offset = offset;
ret = arch_decode_instruction(file, sec, offset,
sec->sh.sh_size - offset,
&insn->len, &insn->type,
&insn->immediate,
&insn->stack_ops);
if (ret)
goto err;
/*
* By default, "ud2" is a dead end unless otherwise
* annotated, because GCC 7 inserts it for certain
* divide-by-zero cases.
*/
if (insn->type == INSN_BUG)
insn->dead_end = true;
hash_add(file->insn_hash, &insn->hash, sec_offset_hash(sec, insn->offset));
list_add_tail(&insn->list, &file->insn_list);
nr_insns++;
}
list_for_each_entry(func, &sec->symbol_list, list) {
if (func->type != STT_FUNC || func->alias != func)
continue;
if (!find_insn(file, sec, func->offset)) {
WARN("%s(): can't find starting instruction",
func->name);
return -1;
}
sym_for_each_insn(file, func, insn) {
insn->func = func;
if (insn->type == INSN_ENDBR && list_empty(&insn->call_node)) {
if (insn->offset == insn->func->offset) {
list_add_tail(&insn->call_node, &file->endbr_list);
file->nr_endbr++;
} else {
file->nr_endbr_int++;
}
}
}
}
}
if (opts.stats)
printf("nr_insns: %lu\n", nr_insns);
return 0;
err:
free(insn);
return ret;
}
/*
* Read the pv_ops[] .data table to find the static initialized values.
*/
static int add_pv_ops(struct objtool_file *file, const char *symname)
{
struct symbol *sym, *func;
unsigned long off, end;
struct reloc *rel;
int idx;
sym = find_symbol_by_name(file->elf, symname);
if (!sym)
return 0;
off = sym->offset;
end = off + sym->len;
for (;;) {
rel = find_reloc_by_dest_range(file->elf, sym->sec, off, end - off);
if (!rel)
break;
func = rel->sym;
if (func->type == STT_SECTION)
func = find_symbol_by_offset(rel->sym->sec, rel->addend);
idx = (rel->offset - sym->offset) / sizeof(unsigned long);
objtool_pv_add(file, idx, func);
off = rel->offset + 1;
if (off > end)
break;
}
return 0;
}
/*
* Allocate and initialize file->pv_ops[].
*/
static int init_pv_ops(struct objtool_file *file)
{
static const char *pv_ops_tables[] = {
"pv_ops",
"xen_cpu_ops",
"xen_irq_ops",
"xen_mmu_ops",
NULL,
};
const char *pv_ops;
struct symbol *sym;
int idx, nr;
if (!opts.noinstr)
return 0;
file->pv_ops = NULL;
sym = find_symbol_by_name(file->elf, "pv_ops");
if (!sym)
return 0;
nr = sym->len / sizeof(unsigned long);
file->pv_ops = calloc(sizeof(struct pv_state), nr);
if (!file->pv_ops)
return -1;
for (idx = 0; idx < nr; idx++)
INIT_LIST_HEAD(&file->pv_ops[idx].targets);
for (idx = 0; (pv_ops = pv_ops_tables[idx]); idx++)
add_pv_ops(file, pv_ops);
return 0;
}
static struct instruction *find_last_insn(struct objtool_file *file,
struct section *sec)
{
struct instruction *insn = NULL;
unsigned int offset;
unsigned int end = (sec->sh.sh_size > 10) ? sec->sh.sh_size - 10 : 0;
for (offset = sec->sh.sh_size - 1; offset >= end && !insn; offset--)
insn = find_insn(file, sec, offset);
return insn;
}
/*
objtool: Assume unannotated UD2 instructions are dead ends Arnd reported some false positive warnings with GCC 7: drivers/hid/wacom_wac.o: warning: objtool: wacom_bpt3_touch()+0x2a5: stack state mismatch: cfa1=7+8 cfa2=6+16 drivers/iio/adc/vf610_adc.o: warning: objtool: vf610_adc_calculate_rates() falls through to next function vf610_adc_sample_set() drivers/pwm/pwm-hibvt.o: warning: objtool: hibvt_pwm_get_state() falls through to next function hibvt_pwm_remove() drivers/pwm/pwm-mediatek.o: warning: objtool: mtk_pwm_config() falls through to next function mtk_pwm_enable() drivers/spi/spi-bcm2835.o: warning: objtool: .text: unexpected end of section drivers/spi/spi-bcm2835aux.o: warning: objtool: .text: unexpected end of section drivers/watchdog/digicolor_wdt.o: warning: objtool: dc_wdt_get_timeleft() falls through to next function dc_wdt_restart() When GCC 7 detects a potential divide-by-zero condition, it sometimes inserts a UD2 instruction for the case where the divisor is zero, instead of letting the hardware trap on the divide instruction. Objtool doesn't consider UD2 to be fatal unless it's annotated with unreachable(). So it considers the GCC-generated UD2 to be non-fatal, and it tries to follow the control flow past the UD2 and gets confused. Previously, objtool *did* assume UD2 was always a dead end. That changed with the following commit: d1091c7fa3d5 ("objtool: Improve detection of BUG() and other dead ends") The motivation behind that change was that Peter was planning on using UD2 for __WARN(), which is *not* a dead end. However, it turns out that some emulators rely on UD2 being fatal, so he ended up using 'ud0' instead: 9a93848fe787 ("x86/debug: Implement __WARN() using UD0") For GCC 4.5+, it should be safe to go back to the previous assumption that UD2 is fatal, even when it's not annotated with unreachable(). But for pre-4.5 versions of GCC, the unreachable() macro isn't supported, so such cases of UD2 need to be explicitly annotated as reachable. Reported-by: Arnd Bergmann <arnd@arndb.de> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Thomas Gleixner <tglx@linutronix.de> Fixes: d1091c7fa3d5 ("objtool: Improve detection of BUG() and other dead ends") Link: http://lkml.kernel.org/r/e57fa9dfede25f79487da8126ee9cdf7b856db65.1501188854.git.jpoimboe@redhat.com Signed-off-by: Ingo Molnar <mingo@kernel.org>
2017-07-28 04:56:53 +08:00
* Mark "ud2" instructions and manually annotated dead ends.
*/
static int add_dead_ends(struct objtool_file *file)
{
struct section *sec;
objtool: Rename rela to reloc Before supporting additional relocation types rename the relevant types and functions from "rela" to "reloc". This work be done with the following regex: sed -e 's/struct rela/struct reloc/g' \ -e 's/\([_\*]\)rela\(s\{0,1\}\)/\1reloc\2/g' \ -e 's/tmprela\(s\{0,1\}\)/tmpreloc\1/g' \ -e 's/relasec/relocsec/g' \ -e 's/rela_list/reloc_list/g' \ -e 's/rela_hash/reloc_hash/g' \ -e 's/add_rela/add_reloc/g' \ -e 's/rela->/reloc->/g' \ -e '/rela[,\.]/{ s/\([^\.>]\)rela\([\.,]\)/\1reloc\2/g ; }' \ -e 's/rela =/reloc =/g' \ -e 's/relas =/relocs =/g' \ -e 's/relas\[/relocs[/g' \ -e 's/relaname =/relocname =/g' \ -e 's/= rela\;/= reloc\;/g' \ -e 's/= relas\;/= relocs\;/g' \ -e 's/= relaname\;/= relocname\;/g' \ -e 's/, rela)/, reloc)/g' \ -e 's/\([ @]\)rela\([ "]\)/\1reloc\2/g' \ -e 's/ rela$/ reloc/g' \ -e 's/, relaname/, relocname/g' \ -e 's/sec->rela/sec->reloc/g' \ -e 's/(\(!\{0,1\}\)rela/(\1reloc/g' \ -i \ arch.h \ arch/x86/decode.c \ check.c \ check.h \ elf.c \ elf.h \ orc_gen.c \ special.c Notable exceptions which complicate the regex include gelf_* library calls and standard/expected section names which still use "rela" because they encode the type of relocation expected. Also, keep "rela" in the struct because it encodes a specific type of relocation we currently expect. It will eventually turn into a member of an anonymous union when a susequent patch adds implicit addend, or "rel", relocation support. Signed-off-by: Matt Helsley <mhelsley@vmware.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
2020-05-30 05:01:13 +08:00
struct reloc *reloc;
struct instruction *insn;
objtool: Assume unannotated UD2 instructions are dead ends Arnd reported some false positive warnings with GCC 7: drivers/hid/wacom_wac.o: warning: objtool: wacom_bpt3_touch()+0x2a5: stack state mismatch: cfa1=7+8 cfa2=6+16 drivers/iio/adc/vf610_adc.o: warning: objtool: vf610_adc_calculate_rates() falls through to next function vf610_adc_sample_set() drivers/pwm/pwm-hibvt.o: warning: objtool: hibvt_pwm_get_state() falls through to next function hibvt_pwm_remove() drivers/pwm/pwm-mediatek.o: warning: objtool: mtk_pwm_config() falls through to next function mtk_pwm_enable() drivers/spi/spi-bcm2835.o: warning: objtool: .text: unexpected end of section drivers/spi/spi-bcm2835aux.o: warning: objtool: .text: unexpected end of section drivers/watchdog/digicolor_wdt.o: warning: objtool: dc_wdt_get_timeleft() falls through to next function dc_wdt_restart() When GCC 7 detects a potential divide-by-zero condition, it sometimes inserts a UD2 instruction for the case where the divisor is zero, instead of letting the hardware trap on the divide instruction. Objtool doesn't consider UD2 to be fatal unless it's annotated with unreachable(). So it considers the GCC-generated UD2 to be non-fatal, and it tries to follow the control flow past the UD2 and gets confused. Previously, objtool *did* assume UD2 was always a dead end. That changed with the following commit: d1091c7fa3d5 ("objtool: Improve detection of BUG() and other dead ends") The motivation behind that change was that Peter was planning on using UD2 for __WARN(), which is *not* a dead end. However, it turns out that some emulators rely on UD2 being fatal, so he ended up using 'ud0' instead: 9a93848fe787 ("x86/debug: Implement __WARN() using UD0") For GCC 4.5+, it should be safe to go back to the previous assumption that UD2 is fatal, even when it's not annotated with unreachable(). But for pre-4.5 versions of GCC, the unreachable() macro isn't supported, so such cases of UD2 need to be explicitly annotated as reachable. Reported-by: Arnd Bergmann <arnd@arndb.de> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Thomas Gleixner <tglx@linutronix.de> Fixes: d1091c7fa3d5 ("objtool: Improve detection of BUG() and other dead ends") Link: http://lkml.kernel.org/r/e57fa9dfede25f79487da8126ee9cdf7b856db65.1501188854.git.jpoimboe@redhat.com Signed-off-by: Ingo Molnar <mingo@kernel.org>
2017-07-28 04:56:53 +08:00
/*
* Check for manually annotated dead ends.
*/
sec = find_section_by_name(file->elf, ".rela.discard.unreachable");
if (!sec)
objtool: Assume unannotated UD2 instructions are dead ends Arnd reported some false positive warnings with GCC 7: drivers/hid/wacom_wac.o: warning: objtool: wacom_bpt3_touch()+0x2a5: stack state mismatch: cfa1=7+8 cfa2=6+16 drivers/iio/adc/vf610_adc.o: warning: objtool: vf610_adc_calculate_rates() falls through to next function vf610_adc_sample_set() drivers/pwm/pwm-hibvt.o: warning: objtool: hibvt_pwm_get_state() falls through to next function hibvt_pwm_remove() drivers/pwm/pwm-mediatek.o: warning: objtool: mtk_pwm_config() falls through to next function mtk_pwm_enable() drivers/spi/spi-bcm2835.o: warning: objtool: .text: unexpected end of section drivers/spi/spi-bcm2835aux.o: warning: objtool: .text: unexpected end of section drivers/watchdog/digicolor_wdt.o: warning: objtool: dc_wdt_get_timeleft() falls through to next function dc_wdt_restart() When GCC 7 detects a potential divide-by-zero condition, it sometimes inserts a UD2 instruction for the case where the divisor is zero, instead of letting the hardware trap on the divide instruction. Objtool doesn't consider UD2 to be fatal unless it's annotated with unreachable(). So it considers the GCC-generated UD2 to be non-fatal, and it tries to follow the control flow past the UD2 and gets confused. Previously, objtool *did* assume UD2 was always a dead end. That changed with the following commit: d1091c7fa3d5 ("objtool: Improve detection of BUG() and other dead ends") The motivation behind that change was that Peter was planning on using UD2 for __WARN(), which is *not* a dead end. However, it turns out that some emulators rely on UD2 being fatal, so he ended up using 'ud0' instead: 9a93848fe787 ("x86/debug: Implement __WARN() using UD0") For GCC 4.5+, it should be safe to go back to the previous assumption that UD2 is fatal, even when it's not annotated with unreachable(). But for pre-4.5 versions of GCC, the unreachable() macro isn't supported, so such cases of UD2 need to be explicitly annotated as reachable. Reported-by: Arnd Bergmann <arnd@arndb.de> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Thomas Gleixner <tglx@linutronix.de> Fixes: d1091c7fa3d5 ("objtool: Improve detection of BUG() and other dead ends") Link: http://lkml.kernel.org/r/e57fa9dfede25f79487da8126ee9cdf7b856db65.1501188854.git.jpoimboe@redhat.com Signed-off-by: Ingo Molnar <mingo@kernel.org>
2017-07-28 04:56:53 +08:00
goto reachable;
objtool: Rename rela to reloc Before supporting additional relocation types rename the relevant types and functions from "rela" to "reloc". This work be done with the following regex: sed -e 's/struct rela/struct reloc/g' \ -e 's/\([_\*]\)rela\(s\{0,1\}\)/\1reloc\2/g' \ -e 's/tmprela\(s\{0,1\}\)/tmpreloc\1/g' \ -e 's/relasec/relocsec/g' \ -e 's/rela_list/reloc_list/g' \ -e 's/rela_hash/reloc_hash/g' \ -e 's/add_rela/add_reloc/g' \ -e 's/rela->/reloc->/g' \ -e '/rela[,\.]/{ s/\([^\.>]\)rela\([\.,]\)/\1reloc\2/g ; }' \ -e 's/rela =/reloc =/g' \ -e 's/relas =/relocs =/g' \ -e 's/relas\[/relocs[/g' \ -e 's/relaname =/relocname =/g' \ -e 's/= rela\;/= reloc\;/g' \ -e 's/= relas\;/= relocs\;/g' \ -e 's/= relaname\;/= relocname\;/g' \ -e 's/, rela)/, reloc)/g' \ -e 's/\([ @]\)rela\([ "]\)/\1reloc\2/g' \ -e 's/ rela$/ reloc/g' \ -e 's/, relaname/, relocname/g' \ -e 's/sec->rela/sec->reloc/g' \ -e 's/(\(!\{0,1\}\)rela/(\1reloc/g' \ -i \ arch.h \ arch/x86/decode.c \ check.c \ check.h \ elf.c \ elf.h \ orc_gen.c \ special.c Notable exceptions which complicate the regex include gelf_* library calls and standard/expected section names which still use "rela" because they encode the type of relocation expected. Also, keep "rela" in the struct because it encodes a specific type of relocation we currently expect. It will eventually turn into a member of an anonymous union when a susequent patch adds implicit addend, or "rel", relocation support. Signed-off-by: Matt Helsley <mhelsley@vmware.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
2020-05-30 05:01:13 +08:00
list_for_each_entry(reloc, &sec->reloc_list, list) {
if (reloc->sym->type != STT_SECTION) {
WARN("unexpected relocation symbol type in %s", sec->name);
return -1;
}
objtool: Rename rela to reloc Before supporting additional relocation types rename the relevant types and functions from "rela" to "reloc". This work be done with the following regex: sed -e 's/struct rela/struct reloc/g' \ -e 's/\([_\*]\)rela\(s\{0,1\}\)/\1reloc\2/g' \ -e 's/tmprela\(s\{0,1\}\)/tmpreloc\1/g' \ -e 's/relasec/relocsec/g' \ -e 's/rela_list/reloc_list/g' \ -e 's/rela_hash/reloc_hash/g' \ -e 's/add_rela/add_reloc/g' \ -e 's/rela->/reloc->/g' \ -e '/rela[,\.]/{ s/\([^\.>]\)rela\([\.,]\)/\1reloc\2/g ; }' \ -e 's/rela =/reloc =/g' \ -e 's/relas =/relocs =/g' \ -e 's/relas\[/relocs[/g' \ -e 's/relaname =/relocname =/g' \ -e 's/= rela\;/= reloc\;/g' \ -e 's/= relas\;/= relocs\;/g' \ -e 's/= relaname\;/= relocname\;/g' \ -e 's/, rela)/, reloc)/g' \ -e 's/\([ @]\)rela\([ "]\)/\1reloc\2/g' \ -e 's/ rela$/ reloc/g' \ -e 's/, relaname/, relocname/g' \ -e 's/sec->rela/sec->reloc/g' \ -e 's/(\(!\{0,1\}\)rela/(\1reloc/g' \ -i \ arch.h \ arch/x86/decode.c \ check.c \ check.h \ elf.c \ elf.h \ orc_gen.c \ special.c Notable exceptions which complicate the regex include gelf_* library calls and standard/expected section names which still use "rela" because they encode the type of relocation expected. Also, keep "rela" in the struct because it encodes a specific type of relocation we currently expect. It will eventually turn into a member of an anonymous union when a susequent patch adds implicit addend, or "rel", relocation support. Signed-off-by: Matt Helsley <mhelsley@vmware.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
2020-05-30 05:01:13 +08:00
insn = find_insn(file, reloc->sym->sec, reloc->addend);
if (insn)
insn = list_prev_entry(insn, list);
else if (reloc->addend == reloc->sym->sec->sh.sh_size) {
objtool: Rename rela to reloc Before supporting additional relocation types rename the relevant types and functions from "rela" to "reloc". This work be done with the following regex: sed -e 's/struct rela/struct reloc/g' \ -e 's/\([_\*]\)rela\(s\{0,1\}\)/\1reloc\2/g' \ -e 's/tmprela\(s\{0,1\}\)/tmpreloc\1/g' \ -e 's/relasec/relocsec/g' \ -e 's/rela_list/reloc_list/g' \ -e 's/rela_hash/reloc_hash/g' \ -e 's/add_rela/add_reloc/g' \ -e 's/rela->/reloc->/g' \ -e '/rela[,\.]/{ s/\([^\.>]\)rela\([\.,]\)/\1reloc\2/g ; }' \ -e 's/rela =/reloc =/g' \ -e 's/relas =/relocs =/g' \ -e 's/relas\[/relocs[/g' \ -e 's/relaname =/relocname =/g' \ -e 's/= rela\;/= reloc\;/g' \ -e 's/= relas\;/= relocs\;/g' \ -e 's/= relaname\;/= relocname\;/g' \ -e 's/, rela)/, reloc)/g' \ -e 's/\([ @]\)rela\([ "]\)/\1reloc\2/g' \ -e 's/ rela$/ reloc/g' \ -e 's/, relaname/, relocname/g' \ -e 's/sec->rela/sec->reloc/g' \ -e 's/(\(!\{0,1\}\)rela/(\1reloc/g' \ -i \ arch.h \ arch/x86/decode.c \ check.c \ check.h \ elf.c \ elf.h \ orc_gen.c \ special.c Notable exceptions which complicate the regex include gelf_* library calls and standard/expected section names which still use "rela" because they encode the type of relocation expected. Also, keep "rela" in the struct because it encodes a specific type of relocation we currently expect. It will eventually turn into a member of an anonymous union when a susequent patch adds implicit addend, or "rel", relocation support. Signed-off-by: Matt Helsley <mhelsley@vmware.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
2020-05-30 05:01:13 +08:00
insn = find_last_insn(file, reloc->sym->sec);
if (!insn) {
WARN("can't find unreachable insn at %s+0x%" PRIx64,
objtool: Rename rela to reloc Before supporting additional relocation types rename the relevant types and functions from "rela" to "reloc". This work be done with the following regex: sed -e 's/struct rela/struct reloc/g' \ -e 's/\([_\*]\)rela\(s\{0,1\}\)/\1reloc\2/g' \ -e 's/tmprela\(s\{0,1\}\)/tmpreloc\1/g' \ -e 's/relasec/relocsec/g' \ -e 's/rela_list/reloc_list/g' \ -e 's/rela_hash/reloc_hash/g' \ -e 's/add_rela/add_reloc/g' \ -e 's/rela->/reloc->/g' \ -e '/rela[,\.]/{ s/\([^\.>]\)rela\([\.,]\)/\1reloc\2/g ; }' \ -e 's/rela =/reloc =/g' \ -e 's/relas =/relocs =/g' \ -e 's/relas\[/relocs[/g' \ -e 's/relaname =/relocname =/g' \ -e 's/= rela\;/= reloc\;/g' \ -e 's/= relas\;/= relocs\;/g' \ -e 's/= relaname\;/= relocname\;/g' \ -e 's/, rela)/, reloc)/g' \ -e 's/\([ @]\)rela\([ "]\)/\1reloc\2/g' \ -e 's/ rela$/ reloc/g' \ -e 's/, relaname/, relocname/g' \ -e 's/sec->rela/sec->reloc/g' \ -e 's/(\(!\{0,1\}\)rela/(\1reloc/g' \ -i \ arch.h \ arch/x86/decode.c \ check.c \ check.h \ elf.c \ elf.h \ orc_gen.c \ special.c Notable exceptions which complicate the regex include gelf_* library calls and standard/expected section names which still use "rela" because they encode the type of relocation expected. Also, keep "rela" in the struct because it encodes a specific type of relocation we currently expect. It will eventually turn into a member of an anonymous union when a susequent patch adds implicit addend, or "rel", relocation support. Signed-off-by: Matt Helsley <mhelsley@vmware.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
2020-05-30 05:01:13 +08:00
reloc->sym->sec->name, reloc->addend);
return -1;
}
} else {
WARN("can't find unreachable insn at %s+0x%" PRIx64,
objtool: Rename rela to reloc Before supporting additional relocation types rename the relevant types and functions from "rela" to "reloc". This work be done with the following regex: sed -e 's/struct rela/struct reloc/g' \ -e 's/\([_\*]\)rela\(s\{0,1\}\)/\1reloc\2/g' \ -e 's/tmprela\(s\{0,1\}\)/tmpreloc\1/g' \ -e 's/relasec/relocsec/g' \ -e 's/rela_list/reloc_list/g' \ -e 's/rela_hash/reloc_hash/g' \ -e 's/add_rela/add_reloc/g' \ -e 's/rela->/reloc->/g' \ -e '/rela[,\.]/{ s/\([^\.>]\)rela\([\.,]\)/\1reloc\2/g ; }' \ -e 's/rela =/reloc =/g' \ -e 's/relas =/relocs =/g' \ -e 's/relas\[/relocs[/g' \ -e 's/relaname =/relocname =/g' \ -e 's/= rela\;/= reloc\;/g' \ -e 's/= relas\;/= relocs\;/g' \ -e 's/= relaname\;/= relocname\;/g' \ -e 's/, rela)/, reloc)/g' \ -e 's/\([ @]\)rela\([ "]\)/\1reloc\2/g' \ -e 's/ rela$/ reloc/g' \ -e 's/, relaname/, relocname/g' \ -e 's/sec->rela/sec->reloc/g' \ -e 's/(\(!\{0,1\}\)rela/(\1reloc/g' \ -i \ arch.h \ arch/x86/decode.c \ check.c \ check.h \ elf.c \ elf.h \ orc_gen.c \ special.c Notable exceptions which complicate the regex include gelf_* library calls and standard/expected section names which still use "rela" because they encode the type of relocation expected. Also, keep "rela" in the struct because it encodes a specific type of relocation we currently expect. It will eventually turn into a member of an anonymous union when a susequent patch adds implicit addend, or "rel", relocation support. Signed-off-by: Matt Helsley <mhelsley@vmware.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
2020-05-30 05:01:13 +08:00
reloc->sym->sec->name, reloc->addend);
return -1;
}
insn->dead_end = true;
}
objtool: Assume unannotated UD2 instructions are dead ends Arnd reported some false positive warnings with GCC 7: drivers/hid/wacom_wac.o: warning: objtool: wacom_bpt3_touch()+0x2a5: stack state mismatch: cfa1=7+8 cfa2=6+16 drivers/iio/adc/vf610_adc.o: warning: objtool: vf610_adc_calculate_rates() falls through to next function vf610_adc_sample_set() drivers/pwm/pwm-hibvt.o: warning: objtool: hibvt_pwm_get_state() falls through to next function hibvt_pwm_remove() drivers/pwm/pwm-mediatek.o: warning: objtool: mtk_pwm_config() falls through to next function mtk_pwm_enable() drivers/spi/spi-bcm2835.o: warning: objtool: .text: unexpected end of section drivers/spi/spi-bcm2835aux.o: warning: objtool: .text: unexpected end of section drivers/watchdog/digicolor_wdt.o: warning: objtool: dc_wdt_get_timeleft() falls through to next function dc_wdt_restart() When GCC 7 detects a potential divide-by-zero condition, it sometimes inserts a UD2 instruction for the case where the divisor is zero, instead of letting the hardware trap on the divide instruction. Objtool doesn't consider UD2 to be fatal unless it's annotated with unreachable(). So it considers the GCC-generated UD2 to be non-fatal, and it tries to follow the control flow past the UD2 and gets confused. Previously, objtool *did* assume UD2 was always a dead end. That changed with the following commit: d1091c7fa3d5 ("objtool: Improve detection of BUG() and other dead ends") The motivation behind that change was that Peter was planning on using UD2 for __WARN(), which is *not* a dead end. However, it turns out that some emulators rely on UD2 being fatal, so he ended up using 'ud0' instead: 9a93848fe787 ("x86/debug: Implement __WARN() using UD0") For GCC 4.5+, it should be safe to go back to the previous assumption that UD2 is fatal, even when it's not annotated with unreachable(). But for pre-4.5 versions of GCC, the unreachable() macro isn't supported, so such cases of UD2 need to be explicitly annotated as reachable. Reported-by: Arnd Bergmann <arnd@arndb.de> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Thomas Gleixner <tglx@linutronix.de> Fixes: d1091c7fa3d5 ("objtool: Improve detection of BUG() and other dead ends") Link: http://lkml.kernel.org/r/e57fa9dfede25f79487da8126ee9cdf7b856db65.1501188854.git.jpoimboe@redhat.com Signed-off-by: Ingo Molnar <mingo@kernel.org>
2017-07-28 04:56:53 +08:00
reachable:
/*
* These manually annotated reachable checks are needed for GCC 4.4,
* where the Linux unreachable() macro isn't supported. In that case
* GCC doesn't know the "ud2" is fatal, so it generates code as if it's
* not a dead end.
*/
sec = find_section_by_name(file->elf, ".rela.discard.reachable");
if (!sec)
return 0;
objtool: Rename rela to reloc Before supporting additional relocation types rename the relevant types and functions from "rela" to "reloc". This work be done with the following regex: sed -e 's/struct rela/struct reloc/g' \ -e 's/\([_\*]\)rela\(s\{0,1\}\)/\1reloc\2/g' \ -e 's/tmprela\(s\{0,1\}\)/tmpreloc\1/g' \ -e 's/relasec/relocsec/g' \ -e 's/rela_list/reloc_list/g' \ -e 's/rela_hash/reloc_hash/g' \ -e 's/add_rela/add_reloc/g' \ -e 's/rela->/reloc->/g' \ -e '/rela[,\.]/{ s/\([^\.>]\)rela\([\.,]\)/\1reloc\2/g ; }' \ -e 's/rela =/reloc =/g' \ -e 's/relas =/relocs =/g' \ -e 's/relas\[/relocs[/g' \ -e 's/relaname =/relocname =/g' \ -e 's/= rela\;/= reloc\;/g' \ -e 's/= relas\;/= relocs\;/g' \ -e 's/= relaname\;/= relocname\;/g' \ -e 's/, rela)/, reloc)/g' \ -e 's/\([ @]\)rela\([ "]\)/\1reloc\2/g' \ -e 's/ rela$/ reloc/g' \ -e 's/, relaname/, relocname/g' \ -e 's/sec->rela/sec->reloc/g' \ -e 's/(\(!\{0,1\}\)rela/(\1reloc/g' \ -i \ arch.h \ arch/x86/decode.c \ check.c \ check.h \ elf.c \ elf.h \ orc_gen.c \ special.c Notable exceptions which complicate the regex include gelf_* library calls and standard/expected section names which still use "rela" because they encode the type of relocation expected. Also, keep "rela" in the struct because it encodes a specific type of relocation we currently expect. It will eventually turn into a member of an anonymous union when a susequent patch adds implicit addend, or "rel", relocation support. Signed-off-by: Matt Helsley <mhelsley@vmware.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
2020-05-30 05:01:13 +08:00
list_for_each_entry(reloc, &sec->reloc_list, list) {
if (reloc->sym->type != STT_SECTION) {
objtool: Assume unannotated UD2 instructions are dead ends Arnd reported some false positive warnings with GCC 7: drivers/hid/wacom_wac.o: warning: objtool: wacom_bpt3_touch()+0x2a5: stack state mismatch: cfa1=7+8 cfa2=6+16 drivers/iio/adc/vf610_adc.o: warning: objtool: vf610_adc_calculate_rates() falls through to next function vf610_adc_sample_set() drivers/pwm/pwm-hibvt.o: warning: objtool: hibvt_pwm_get_state() falls through to next function hibvt_pwm_remove() drivers/pwm/pwm-mediatek.o: warning: objtool: mtk_pwm_config() falls through to next function mtk_pwm_enable() drivers/spi/spi-bcm2835.o: warning: objtool: .text: unexpected end of section drivers/spi/spi-bcm2835aux.o: warning: objtool: .text: unexpected end of section drivers/watchdog/digicolor_wdt.o: warning: objtool: dc_wdt_get_timeleft() falls through to next function dc_wdt_restart() When GCC 7 detects a potential divide-by-zero condition, it sometimes inserts a UD2 instruction for the case where the divisor is zero, instead of letting the hardware trap on the divide instruction. Objtool doesn't consider UD2 to be fatal unless it's annotated with unreachable(). So it considers the GCC-generated UD2 to be non-fatal, and it tries to follow the control flow past the UD2 and gets confused. Previously, objtool *did* assume UD2 was always a dead end. That changed with the following commit: d1091c7fa3d5 ("objtool: Improve detection of BUG() and other dead ends") The motivation behind that change was that Peter was planning on using UD2 for __WARN(), which is *not* a dead end. However, it turns out that some emulators rely on UD2 being fatal, so he ended up using 'ud0' instead: 9a93848fe787 ("x86/debug: Implement __WARN() using UD0") For GCC 4.5+, it should be safe to go back to the previous assumption that UD2 is fatal, even when it's not annotated with unreachable(). But for pre-4.5 versions of GCC, the unreachable() macro isn't supported, so such cases of UD2 need to be explicitly annotated as reachable. Reported-by: Arnd Bergmann <arnd@arndb.de> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Thomas Gleixner <tglx@linutronix.de> Fixes: d1091c7fa3d5 ("objtool: Improve detection of BUG() and other dead ends") Link: http://lkml.kernel.org/r/e57fa9dfede25f79487da8126ee9cdf7b856db65.1501188854.git.jpoimboe@redhat.com Signed-off-by: Ingo Molnar <mingo@kernel.org>
2017-07-28 04:56:53 +08:00
WARN("unexpected relocation symbol type in %s", sec->name);
return -1;
}
objtool: Rename rela to reloc Before supporting additional relocation types rename the relevant types and functions from "rela" to "reloc". This work be done with the following regex: sed -e 's/struct rela/struct reloc/g' \ -e 's/\([_\*]\)rela\(s\{0,1\}\)/\1reloc\2/g' \ -e 's/tmprela\(s\{0,1\}\)/tmpreloc\1/g' \ -e 's/relasec/relocsec/g' \ -e 's/rela_list/reloc_list/g' \ -e 's/rela_hash/reloc_hash/g' \ -e 's/add_rela/add_reloc/g' \ -e 's/rela->/reloc->/g' \ -e '/rela[,\.]/{ s/\([^\.>]\)rela\([\.,]\)/\1reloc\2/g ; }' \ -e 's/rela =/reloc =/g' \ -e 's/relas =/relocs =/g' \ -e 's/relas\[/relocs[/g' \ -e 's/relaname =/relocname =/g' \ -e 's/= rela\;/= reloc\;/g' \ -e 's/= relas\;/= relocs\;/g' \ -e 's/= relaname\;/= relocname\;/g' \ -e 's/, rela)/, reloc)/g' \ -e 's/\([ @]\)rela\([ "]\)/\1reloc\2/g' \ -e 's/ rela$/ reloc/g' \ -e 's/, relaname/, relocname/g' \ -e 's/sec->rela/sec->reloc/g' \ -e 's/(\(!\{0,1\}\)rela/(\1reloc/g' \ -i \ arch.h \ arch/x86/decode.c \ check.c \ check.h \ elf.c \ elf.h \ orc_gen.c \ special.c Notable exceptions which complicate the regex include gelf_* library calls and standard/expected section names which still use "rela" because they encode the type of relocation expected. Also, keep "rela" in the struct because it encodes a specific type of relocation we currently expect. It will eventually turn into a member of an anonymous union when a susequent patch adds implicit addend, or "rel", relocation support. Signed-off-by: Matt Helsley <mhelsley@vmware.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
2020-05-30 05:01:13 +08:00
insn = find_insn(file, reloc->sym->sec, reloc->addend);
objtool: Assume unannotated UD2 instructions are dead ends Arnd reported some false positive warnings with GCC 7: drivers/hid/wacom_wac.o: warning: objtool: wacom_bpt3_touch()+0x2a5: stack state mismatch: cfa1=7+8 cfa2=6+16 drivers/iio/adc/vf610_adc.o: warning: objtool: vf610_adc_calculate_rates() falls through to next function vf610_adc_sample_set() drivers/pwm/pwm-hibvt.o: warning: objtool: hibvt_pwm_get_state() falls through to next function hibvt_pwm_remove() drivers/pwm/pwm-mediatek.o: warning: objtool: mtk_pwm_config() falls through to next function mtk_pwm_enable() drivers/spi/spi-bcm2835.o: warning: objtool: .text: unexpected end of section drivers/spi/spi-bcm2835aux.o: warning: objtool: .text: unexpected end of section drivers/watchdog/digicolor_wdt.o: warning: objtool: dc_wdt_get_timeleft() falls through to next function dc_wdt_restart() When GCC 7 detects a potential divide-by-zero condition, it sometimes inserts a UD2 instruction for the case where the divisor is zero, instead of letting the hardware trap on the divide instruction. Objtool doesn't consider UD2 to be fatal unless it's annotated with unreachable(). So it considers the GCC-generated UD2 to be non-fatal, and it tries to follow the control flow past the UD2 and gets confused. Previously, objtool *did* assume UD2 was always a dead end. That changed with the following commit: d1091c7fa3d5 ("objtool: Improve detection of BUG() and other dead ends") The motivation behind that change was that Peter was planning on using UD2 for __WARN(), which is *not* a dead end. However, it turns out that some emulators rely on UD2 being fatal, so he ended up using 'ud0' instead: 9a93848fe787 ("x86/debug: Implement __WARN() using UD0") For GCC 4.5+, it should be safe to go back to the previous assumption that UD2 is fatal, even when it's not annotated with unreachable(). But for pre-4.5 versions of GCC, the unreachable() macro isn't supported, so such cases of UD2 need to be explicitly annotated as reachable. Reported-by: Arnd Bergmann <arnd@arndb.de> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Thomas Gleixner <tglx@linutronix.de> Fixes: d1091c7fa3d5 ("objtool: Improve detection of BUG() and other dead ends") Link: http://lkml.kernel.org/r/e57fa9dfede25f79487da8126ee9cdf7b856db65.1501188854.git.jpoimboe@redhat.com Signed-off-by: Ingo Molnar <mingo@kernel.org>
2017-07-28 04:56:53 +08:00
if (insn)
insn = list_prev_entry(insn, list);
else if (reloc->addend == reloc->sym->sec->sh.sh_size) {
objtool: Rename rela to reloc Before supporting additional relocation types rename the relevant types and functions from "rela" to "reloc". This work be done with the following regex: sed -e 's/struct rela/struct reloc/g' \ -e 's/\([_\*]\)rela\(s\{0,1\}\)/\1reloc\2/g' \ -e 's/tmprela\(s\{0,1\}\)/tmpreloc\1/g' \ -e 's/relasec/relocsec/g' \ -e 's/rela_list/reloc_list/g' \ -e 's/rela_hash/reloc_hash/g' \ -e 's/add_rela/add_reloc/g' \ -e 's/rela->/reloc->/g' \ -e '/rela[,\.]/{ s/\([^\.>]\)rela\([\.,]\)/\1reloc\2/g ; }' \ -e 's/rela =/reloc =/g' \ -e 's/relas =/relocs =/g' \ -e 's/relas\[/relocs[/g' \ -e 's/relaname =/relocname =/g' \ -e 's/= rela\;/= reloc\;/g' \ -e 's/= relas\;/= relocs\;/g' \ -e 's/= relaname\;/= relocname\;/g' \ -e 's/, rela)/, reloc)/g' \ -e 's/\([ @]\)rela\([ "]\)/\1reloc\2/g' \ -e 's/ rela$/ reloc/g' \ -e 's/, relaname/, relocname/g' \ -e 's/sec->rela/sec->reloc/g' \ -e 's/(\(!\{0,1\}\)rela/(\1reloc/g' \ -i \ arch.h \ arch/x86/decode.c \ check.c \ check.h \ elf.c \ elf.h \ orc_gen.c \ special.c Notable exceptions which complicate the regex include gelf_* library calls and standard/expected section names which still use "rela" because they encode the type of relocation expected. Also, keep "rela" in the struct because it encodes a specific type of relocation we currently expect. It will eventually turn into a member of an anonymous union when a susequent patch adds implicit addend, or "rel", relocation support. Signed-off-by: Matt Helsley <mhelsley@vmware.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
2020-05-30 05:01:13 +08:00
insn = find_last_insn(file, reloc->sym->sec);
if (!insn) {
WARN("can't find reachable insn at %s+0x%" PRIx64,
objtool: Rename rela to reloc Before supporting additional relocation types rename the relevant types and functions from "rela" to "reloc". This work be done with the following regex: sed -e 's/struct rela/struct reloc/g' \ -e 's/\([_\*]\)rela\(s\{0,1\}\)/\1reloc\2/g' \ -e 's/tmprela\(s\{0,1\}\)/tmpreloc\1/g' \ -e 's/relasec/relocsec/g' \ -e 's/rela_list/reloc_list/g' \ -e 's/rela_hash/reloc_hash/g' \ -e 's/add_rela/add_reloc/g' \ -e 's/rela->/reloc->/g' \ -e '/rela[,\.]/{ s/\([^\.>]\)rela\([\.,]\)/\1reloc\2/g ; }' \ -e 's/rela =/reloc =/g' \ -e 's/relas =/relocs =/g' \ -e 's/relas\[/relocs[/g' \ -e 's/relaname =/relocname =/g' \ -e 's/= rela\;/= reloc\;/g' \ -e 's/= relas\;/= relocs\;/g' \ -e 's/= relaname\;/= relocname\;/g' \ -e 's/, rela)/, reloc)/g' \ -e 's/\([ @]\)rela\([ "]\)/\1reloc\2/g' \ -e 's/ rela$/ reloc/g' \ -e 's/, relaname/, relocname/g' \ -e 's/sec->rela/sec->reloc/g' \ -e 's/(\(!\{0,1\}\)rela/(\1reloc/g' \ -i \ arch.h \ arch/x86/decode.c \ check.c \ check.h \ elf.c \ elf.h \ orc_gen.c \ special.c Notable exceptions which complicate the regex include gelf_* library calls and standard/expected section names which still use "rela" because they encode the type of relocation expected. Also, keep "rela" in the struct because it encodes a specific type of relocation we currently expect. It will eventually turn into a member of an anonymous union when a susequent patch adds implicit addend, or "rel", relocation support. Signed-off-by: Matt Helsley <mhelsley@vmware.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
2020-05-30 05:01:13 +08:00
reloc->sym->sec->name, reloc->addend);
objtool: Assume unannotated UD2 instructions are dead ends Arnd reported some false positive warnings with GCC 7: drivers/hid/wacom_wac.o: warning: objtool: wacom_bpt3_touch()+0x2a5: stack state mismatch: cfa1=7+8 cfa2=6+16 drivers/iio/adc/vf610_adc.o: warning: objtool: vf610_adc_calculate_rates() falls through to next function vf610_adc_sample_set() drivers/pwm/pwm-hibvt.o: warning: objtool: hibvt_pwm_get_state() falls through to next function hibvt_pwm_remove() drivers/pwm/pwm-mediatek.o: warning: objtool: mtk_pwm_config() falls through to next function mtk_pwm_enable() drivers/spi/spi-bcm2835.o: warning: objtool: .text: unexpected end of section drivers/spi/spi-bcm2835aux.o: warning: objtool: .text: unexpected end of section drivers/watchdog/digicolor_wdt.o: warning: objtool: dc_wdt_get_timeleft() falls through to next function dc_wdt_restart() When GCC 7 detects a potential divide-by-zero condition, it sometimes inserts a UD2 instruction for the case where the divisor is zero, instead of letting the hardware trap on the divide instruction. Objtool doesn't consider UD2 to be fatal unless it's annotated with unreachable(). So it considers the GCC-generated UD2 to be non-fatal, and it tries to follow the control flow past the UD2 and gets confused. Previously, objtool *did* assume UD2 was always a dead end. That changed with the following commit: d1091c7fa3d5 ("objtool: Improve detection of BUG() and other dead ends") The motivation behind that change was that Peter was planning on using UD2 for __WARN(), which is *not* a dead end. However, it turns out that some emulators rely on UD2 being fatal, so he ended up using 'ud0' instead: 9a93848fe787 ("x86/debug: Implement __WARN() using UD0") For GCC 4.5+, it should be safe to go back to the previous assumption that UD2 is fatal, even when it's not annotated with unreachable(). But for pre-4.5 versions of GCC, the unreachable() macro isn't supported, so such cases of UD2 need to be explicitly annotated as reachable. Reported-by: Arnd Bergmann <arnd@arndb.de> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Thomas Gleixner <tglx@linutronix.de> Fixes: d1091c7fa3d5 ("objtool: Improve detection of BUG() and other dead ends") Link: http://lkml.kernel.org/r/e57fa9dfede25f79487da8126ee9cdf7b856db65.1501188854.git.jpoimboe@redhat.com Signed-off-by: Ingo Molnar <mingo@kernel.org>
2017-07-28 04:56:53 +08:00
return -1;
}
} else {
WARN("can't find reachable insn at %s+0x%" PRIx64,
objtool: Rename rela to reloc Before supporting additional relocation types rename the relevant types and functions from "rela" to "reloc". This work be done with the following regex: sed -e 's/struct rela/struct reloc/g' \ -e 's/\([_\*]\)rela\(s\{0,1\}\)/\1reloc\2/g' \ -e 's/tmprela\(s\{0,1\}\)/tmpreloc\1/g' \ -e 's/relasec/relocsec/g' \ -e 's/rela_list/reloc_list/g' \ -e 's/rela_hash/reloc_hash/g' \ -e 's/add_rela/add_reloc/g' \ -e 's/rela->/reloc->/g' \ -e '/rela[,\.]/{ s/\([^\.>]\)rela\([\.,]\)/\1reloc\2/g ; }' \ -e 's/rela =/reloc =/g' \ -e 's/relas =/relocs =/g' \ -e 's/relas\[/relocs[/g' \ -e 's/relaname =/relocname =/g' \ -e 's/= rela\;/= reloc\;/g' \ -e 's/= relas\;/= relocs\;/g' \ -e 's/= relaname\;/= relocname\;/g' \ -e 's/, rela)/, reloc)/g' \ -e 's/\([ @]\)rela\([ "]\)/\1reloc\2/g' \ -e 's/ rela$/ reloc/g' \ -e 's/, relaname/, relocname/g' \ -e 's/sec->rela/sec->reloc/g' \ -e 's/(\(!\{0,1\}\)rela/(\1reloc/g' \ -i \ arch.h \ arch/x86/decode.c \ check.c \ check.h \ elf.c \ elf.h \ orc_gen.c \ special.c Notable exceptions which complicate the regex include gelf_* library calls and standard/expected section names which still use "rela" because they encode the type of relocation expected. Also, keep "rela" in the struct because it encodes a specific type of relocation we currently expect. It will eventually turn into a member of an anonymous union when a susequent patch adds implicit addend, or "rel", relocation support. Signed-off-by: Matt Helsley <mhelsley@vmware.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
2020-05-30 05:01:13 +08:00
reloc->sym->sec->name, reloc->addend);
objtool: Assume unannotated UD2 instructions are dead ends Arnd reported some false positive warnings with GCC 7: drivers/hid/wacom_wac.o: warning: objtool: wacom_bpt3_touch()+0x2a5: stack state mismatch: cfa1=7+8 cfa2=6+16 drivers/iio/adc/vf610_adc.o: warning: objtool: vf610_adc_calculate_rates() falls through to next function vf610_adc_sample_set() drivers/pwm/pwm-hibvt.o: warning: objtool: hibvt_pwm_get_state() falls through to next function hibvt_pwm_remove() drivers/pwm/pwm-mediatek.o: warning: objtool: mtk_pwm_config() falls through to next function mtk_pwm_enable() drivers/spi/spi-bcm2835.o: warning: objtool: .text: unexpected end of section drivers/spi/spi-bcm2835aux.o: warning: objtool: .text: unexpected end of section drivers/watchdog/digicolor_wdt.o: warning: objtool: dc_wdt_get_timeleft() falls through to next function dc_wdt_restart() When GCC 7 detects a potential divide-by-zero condition, it sometimes inserts a UD2 instruction for the case where the divisor is zero, instead of letting the hardware trap on the divide instruction. Objtool doesn't consider UD2 to be fatal unless it's annotated with unreachable(). So it considers the GCC-generated UD2 to be non-fatal, and it tries to follow the control flow past the UD2 and gets confused. Previously, objtool *did* assume UD2 was always a dead end. That changed with the following commit: d1091c7fa3d5 ("objtool: Improve detection of BUG() and other dead ends") The motivation behind that change was that Peter was planning on using UD2 for __WARN(), which is *not* a dead end. However, it turns out that some emulators rely on UD2 being fatal, so he ended up using 'ud0' instead: 9a93848fe787 ("x86/debug: Implement __WARN() using UD0") For GCC 4.5+, it should be safe to go back to the previous assumption that UD2 is fatal, even when it's not annotated with unreachable(). But for pre-4.5 versions of GCC, the unreachable() macro isn't supported, so such cases of UD2 need to be explicitly annotated as reachable. Reported-by: Arnd Bergmann <arnd@arndb.de> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Thomas Gleixner <tglx@linutronix.de> Fixes: d1091c7fa3d5 ("objtool: Improve detection of BUG() and other dead ends") Link: http://lkml.kernel.org/r/e57fa9dfede25f79487da8126ee9cdf7b856db65.1501188854.git.jpoimboe@redhat.com Signed-off-by: Ingo Molnar <mingo@kernel.org>
2017-07-28 04:56:53 +08:00
return -1;
}
insn->dead_end = false;
}
return 0;
}
static int create_static_call_sections(struct objtool_file *file)
{
struct section *sec;
struct static_call_site *site;
struct instruction *insn;
struct symbol *key_sym;
char *key_name, *tmp;
int idx;
sec = find_section_by_name(file->elf, ".static_call_sites");
if (sec) {
INIT_LIST_HEAD(&file->static_call_list);
WARN("file already has .static_call_sites section, skipping");
return 0;
}
if (list_empty(&file->static_call_list))
return 0;
idx = 0;
list_for_each_entry(insn, &file->static_call_list, call_node)
idx++;
sec = elf_create_section(file->elf, ".static_call_sites", SHF_WRITE,
sizeof(struct static_call_site), idx);
if (!sec)
return -1;
idx = 0;
list_for_each_entry(insn, &file->static_call_list, call_node) {
site = (struct static_call_site *)sec->data->d_buf + idx;
memset(site, 0, sizeof(struct static_call_site));
/* populate reloc for 'addr' */
if (elf_add_reloc_to_insn(file->elf, sec,
idx * sizeof(struct static_call_site),
R_X86_64_PC32,
insn->sec, insn->offset))
return -1;
/* find key symbol */
key_name = strdup(insn->call_dest->name);
if (!key_name) {
perror("strdup");
return -1;
}
if (strncmp(key_name, STATIC_CALL_TRAMP_PREFIX_STR,
STATIC_CALL_TRAMP_PREFIX_LEN)) {
WARN("static_call: trampoline name malformed: %s", key_name);
return -1;
}
tmp = key_name + STATIC_CALL_TRAMP_PREFIX_LEN - STATIC_CALL_KEY_PREFIX_LEN;
memcpy(tmp, STATIC_CALL_KEY_PREFIX_STR, STATIC_CALL_KEY_PREFIX_LEN);
key_sym = find_symbol_by_name(file->elf, tmp);
if (!key_sym) {
if (!opts.module) {
WARN("static_call: can't find static_call_key symbol: %s", tmp);
return -1;
}
/*
* For modules(), the key might not be exported, which
* means the module can make static calls but isn't
* allowed to change them.
*
* In that case we temporarily set the key to be the
* trampoline address. This is fixed up in
* static_call_add_module().
*/
key_sym = insn->call_dest;
}
free(key_name);
/* populate reloc for 'key' */
if (elf_add_reloc(file->elf, sec,
idx * sizeof(struct static_call_site) + 4,
R_X86_64_PC32, key_sym,
is_sibling_call(insn) * STATIC_CALL_SITE_TAIL))
return -1;
idx++;
}
return 0;
}
static int create_retpoline_sites_sections(struct objtool_file *file)
{
struct instruction *insn;
struct section *sec;
int idx;
sec = find_section_by_name(file->elf, ".retpoline_sites");
if (sec) {
WARN("file already has .retpoline_sites, skipping");
return 0;
}
idx = 0;
list_for_each_entry(insn, &file->retpoline_call_list, call_node)
idx++;
if (!idx)
return 0;
sec = elf_create_section(file->elf, ".retpoline_sites", 0,
sizeof(int), idx);
if (!sec) {
WARN("elf_create_section: .retpoline_sites");
return -1;
}
idx = 0;
list_for_each_entry(insn, &file->retpoline_call_list, call_node) {
int *site = (int *)sec->data->d_buf + idx;
*site = 0;
if (elf_add_reloc_to_insn(file->elf, sec,
idx * sizeof(int),
R_X86_64_PC32,
insn->sec, insn->offset)) {
WARN("elf_add_reloc_to_insn: .retpoline_sites");
return -1;
}
idx++;
}
return 0;
}
static int create_return_sites_sections(struct objtool_file *file)
{
struct instruction *insn;
struct section *sec;
int idx;
sec = find_section_by_name(file->elf, ".return_sites");
if (sec) {
WARN("file already has .return_sites, skipping");
return 0;
}
idx = 0;
list_for_each_entry(insn, &file->return_thunk_list, call_node)
idx++;
if (!idx)
return 0;
sec = elf_create_section(file->elf, ".return_sites", 0,
sizeof(int), idx);
if (!sec) {
WARN("elf_create_section: .return_sites");
return -1;
}
idx = 0;
list_for_each_entry(insn, &file->return_thunk_list, call_node) {
int *site = (int *)sec->data->d_buf + idx;
*site = 0;
if (elf_add_reloc_to_insn(file->elf, sec,
idx * sizeof(int),
R_X86_64_PC32,
insn->sec, insn->offset)) {
WARN("elf_add_reloc_to_insn: .return_sites");
return -1;
}
idx++;
}
return 0;
}
static int create_ibt_endbr_seal_sections(struct objtool_file *file)
{
struct instruction *insn;
struct section *sec;
int idx;
sec = find_section_by_name(file->elf, ".ibt_endbr_seal");
if (sec) {
WARN("file already has .ibt_endbr_seal, skipping");
return 0;
}
idx = 0;
list_for_each_entry(insn, &file->endbr_list, call_node)
idx++;
if (opts.stats) {
printf("ibt: ENDBR at function start: %d\n", file->nr_endbr);
printf("ibt: ENDBR inside functions: %d\n", file->nr_endbr_int);
printf("ibt: superfluous ENDBR: %d\n", idx);
}
if (!idx)
return 0;
sec = elf_create_section(file->elf, ".ibt_endbr_seal", 0,
sizeof(int), idx);
if (!sec) {
WARN("elf_create_section: .ibt_endbr_seal");
return -1;
}
idx = 0;
list_for_each_entry(insn, &file->endbr_list, call_node) {
int *site = (int *)sec->data->d_buf + idx;
*site = 0;
if (elf_add_reloc_to_insn(file->elf, sec,
idx * sizeof(int),
R_X86_64_PC32,
insn->sec, insn->offset)) {
WARN("elf_add_reloc_to_insn: .ibt_endbr_seal");
return -1;
}
idx++;
}
return 0;
}
static int create_mcount_loc_sections(struct objtool_file *file)
{
struct section *sec;
unsigned long *loc;
struct instruction *insn;
int idx;
sec = find_section_by_name(file->elf, "__mcount_loc");
if (sec) {
INIT_LIST_HEAD(&file->mcount_loc_list);
WARN("file already has __mcount_loc section, skipping");
return 0;
}
if (list_empty(&file->mcount_loc_list))
return 0;
idx = 0;
list_for_each_entry(insn, &file->mcount_loc_list, call_node)
idx++;
sec = elf_create_section(file->elf, "__mcount_loc", 0, sizeof(unsigned long), idx);
if (!sec)
return -1;
idx = 0;
list_for_each_entry(insn, &file->mcount_loc_list, call_node) {
loc = (unsigned long *)sec->data->d_buf + idx;
memset(loc, 0, sizeof(unsigned long));
if (elf_add_reloc_to_insn(file->elf, sec,
idx * sizeof(unsigned long),
R_X86_64_64,
insn->sec, insn->offset))
return -1;
idx++;
}
return 0;
}
/*
* Warnings shouldn't be reported for ignored functions.
*/
static void add_ignores(struct objtool_file *file)
{
struct instruction *insn;
struct section *sec;
struct symbol *func;
objtool: Rename rela to reloc Before supporting additional relocation types rename the relevant types and functions from "rela" to "reloc". This work be done with the following regex: sed -e 's/struct rela/struct reloc/g' \ -e 's/\([_\*]\)rela\(s\{0,1\}\)/\1reloc\2/g' \ -e 's/tmprela\(s\{0,1\}\)/tmpreloc\1/g' \ -e 's/relasec/relocsec/g' \ -e 's/rela_list/reloc_list/g' \ -e 's/rela_hash/reloc_hash/g' \ -e 's/add_rela/add_reloc/g' \ -e 's/rela->/reloc->/g' \ -e '/rela[,\.]/{ s/\([^\.>]\)rela\([\.,]\)/\1reloc\2/g ; }' \ -e 's/rela =/reloc =/g' \ -e 's/relas =/relocs =/g' \ -e 's/relas\[/relocs[/g' \ -e 's/relaname =/relocname =/g' \ -e 's/= rela\;/= reloc\;/g' \ -e 's/= relas\;/= relocs\;/g' \ -e 's/= relaname\;/= relocname\;/g' \ -e 's/, rela)/, reloc)/g' \ -e 's/\([ @]\)rela\([ "]\)/\1reloc\2/g' \ -e 's/ rela$/ reloc/g' \ -e 's/, relaname/, relocname/g' \ -e 's/sec->rela/sec->reloc/g' \ -e 's/(\(!\{0,1\}\)rela/(\1reloc/g' \ -i \ arch.h \ arch/x86/decode.c \ check.c \ check.h \ elf.c \ elf.h \ orc_gen.c \ special.c Notable exceptions which complicate the regex include gelf_* library calls and standard/expected section names which still use "rela" because they encode the type of relocation expected. Also, keep "rela" in the struct because it encodes a specific type of relocation we currently expect. It will eventually turn into a member of an anonymous union when a susequent patch adds implicit addend, or "rel", relocation support. Signed-off-by: Matt Helsley <mhelsley@vmware.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
2020-05-30 05:01:13 +08:00
struct reloc *reloc;
sec = find_section_by_name(file->elf, ".rela.discard.func_stack_frame_non_standard");
if (!sec)
return;
objtool: Rename rela to reloc Before supporting additional relocation types rename the relevant types and functions from "rela" to "reloc". This work be done with the following regex: sed -e 's/struct rela/struct reloc/g' \ -e 's/\([_\*]\)rela\(s\{0,1\}\)/\1reloc\2/g' \ -e 's/tmprela\(s\{0,1\}\)/tmpreloc\1/g' \ -e 's/relasec/relocsec/g' \ -e 's/rela_list/reloc_list/g' \ -e 's/rela_hash/reloc_hash/g' \ -e 's/add_rela/add_reloc/g' \ -e 's/rela->/reloc->/g' \ -e '/rela[,\.]/{ s/\([^\.>]\)rela\([\.,]\)/\1reloc\2/g ; }' \ -e 's/rela =/reloc =/g' \ -e 's/relas =/relocs =/g' \ -e 's/relas\[/relocs[/g' \ -e 's/relaname =/relocname =/g' \ -e 's/= rela\;/= reloc\;/g' \ -e 's/= relas\;/= relocs\;/g' \ -e 's/= relaname\;/= relocname\;/g' \ -e 's/, rela)/, reloc)/g' \ -e 's/\([ @]\)rela\([ "]\)/\1reloc\2/g' \ -e 's/ rela$/ reloc/g' \ -e 's/, relaname/, relocname/g' \ -e 's/sec->rela/sec->reloc/g' \ -e 's/(\(!\{0,1\}\)rela/(\1reloc/g' \ -i \ arch.h \ arch/x86/decode.c \ check.c \ check.h \ elf.c \ elf.h \ orc_gen.c \ special.c Notable exceptions which complicate the regex include gelf_* library calls and standard/expected section names which still use "rela" because they encode the type of relocation expected. Also, keep "rela" in the struct because it encodes a specific type of relocation we currently expect. It will eventually turn into a member of an anonymous union when a susequent patch adds implicit addend, or "rel", relocation support. Signed-off-by: Matt Helsley <mhelsley@vmware.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
2020-05-30 05:01:13 +08:00
list_for_each_entry(reloc, &sec->reloc_list, list) {
switch (reloc->sym->type) {
case STT_FUNC:
objtool: Rename rela to reloc Before supporting additional relocation types rename the relevant types and functions from "rela" to "reloc". This work be done with the following regex: sed -e 's/struct rela/struct reloc/g' \ -e 's/\([_\*]\)rela\(s\{0,1\}\)/\1reloc\2/g' \ -e 's/tmprela\(s\{0,1\}\)/tmpreloc\1/g' \ -e 's/relasec/relocsec/g' \ -e 's/rela_list/reloc_list/g' \ -e 's/rela_hash/reloc_hash/g' \ -e 's/add_rela/add_reloc/g' \ -e 's/rela->/reloc->/g' \ -e '/rela[,\.]/{ s/\([^\.>]\)rela\([\.,]\)/\1reloc\2/g ; }' \ -e 's/rela =/reloc =/g' \ -e 's/relas =/relocs =/g' \ -e 's/relas\[/relocs[/g' \ -e 's/relaname =/relocname =/g' \ -e 's/= rela\;/= reloc\;/g' \ -e 's/= relas\;/= relocs\;/g' \ -e 's/= relaname\;/= relocname\;/g' \ -e 's/, rela)/, reloc)/g' \ -e 's/\([ @]\)rela\([ "]\)/\1reloc\2/g' \ -e 's/ rela$/ reloc/g' \ -e 's/, relaname/, relocname/g' \ -e 's/sec->rela/sec->reloc/g' \ -e 's/(\(!\{0,1\}\)rela/(\1reloc/g' \ -i \ arch.h \ arch/x86/decode.c \ check.c \ check.h \ elf.c \ elf.h \ orc_gen.c \ special.c Notable exceptions which complicate the regex include gelf_* library calls and standard/expected section names which still use "rela" because they encode the type of relocation expected. Also, keep "rela" in the struct because it encodes a specific type of relocation we currently expect. It will eventually turn into a member of an anonymous union when a susequent patch adds implicit addend, or "rel", relocation support. Signed-off-by: Matt Helsley <mhelsley@vmware.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
2020-05-30 05:01:13 +08:00
func = reloc->sym;
break;
case STT_SECTION:
objtool: Rename rela to reloc Before supporting additional relocation types rename the relevant types and functions from "rela" to "reloc". This work be done with the following regex: sed -e 's/struct rela/struct reloc/g' \ -e 's/\([_\*]\)rela\(s\{0,1\}\)/\1reloc\2/g' \ -e 's/tmprela\(s\{0,1\}\)/tmpreloc\1/g' \ -e 's/relasec/relocsec/g' \ -e 's/rela_list/reloc_list/g' \ -e 's/rela_hash/reloc_hash/g' \ -e 's/add_rela/add_reloc/g' \ -e 's/rela->/reloc->/g' \ -e '/rela[,\.]/{ s/\([^\.>]\)rela\([\.,]\)/\1reloc\2/g ; }' \ -e 's/rela =/reloc =/g' \ -e 's/relas =/relocs =/g' \ -e 's/relas\[/relocs[/g' \ -e 's/relaname =/relocname =/g' \ -e 's/= rela\;/= reloc\;/g' \ -e 's/= relas\;/= relocs\;/g' \ -e 's/= relaname\;/= relocname\;/g' \ -e 's/, rela)/, reloc)/g' \ -e 's/\([ @]\)rela\([ "]\)/\1reloc\2/g' \ -e 's/ rela$/ reloc/g' \ -e 's/, relaname/, relocname/g' \ -e 's/sec->rela/sec->reloc/g' \ -e 's/(\(!\{0,1\}\)rela/(\1reloc/g' \ -i \ arch.h \ arch/x86/decode.c \ check.c \ check.h \ elf.c \ elf.h \ orc_gen.c \ special.c Notable exceptions which complicate the regex include gelf_* library calls and standard/expected section names which still use "rela" because they encode the type of relocation expected. Also, keep "rela" in the struct because it encodes a specific type of relocation we currently expect. It will eventually turn into a member of an anonymous union when a susequent patch adds implicit addend, or "rel", relocation support. Signed-off-by: Matt Helsley <mhelsley@vmware.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
2020-05-30 05:01:13 +08:00
func = find_func_by_offset(reloc->sym->sec, reloc->addend);
if (!func)
continue;
break;
default:
objtool: Rename rela to reloc Before supporting additional relocation types rename the relevant types and functions from "rela" to "reloc". This work be done with the following regex: sed -e 's/struct rela/struct reloc/g' \ -e 's/\([_\*]\)rela\(s\{0,1\}\)/\1reloc\2/g' \ -e 's/tmprela\(s\{0,1\}\)/tmpreloc\1/g' \ -e 's/relasec/relocsec/g' \ -e 's/rela_list/reloc_list/g' \ -e 's/rela_hash/reloc_hash/g' \ -e 's/add_rela/add_reloc/g' \ -e 's/rela->/reloc->/g' \ -e '/rela[,\.]/{ s/\([^\.>]\)rela\([\.,]\)/\1reloc\2/g ; }' \ -e 's/rela =/reloc =/g' \ -e 's/relas =/relocs =/g' \ -e 's/relas\[/relocs[/g' \ -e 's/relaname =/relocname =/g' \ -e 's/= rela\;/= reloc\;/g' \ -e 's/= relas\;/= relocs\;/g' \ -e 's/= relaname\;/= relocname\;/g' \ -e 's/, rela)/, reloc)/g' \ -e 's/\([ @]\)rela\([ "]\)/\1reloc\2/g' \ -e 's/ rela$/ reloc/g' \ -e 's/, relaname/, relocname/g' \ -e 's/sec->rela/sec->reloc/g' \ -e 's/(\(!\{0,1\}\)rela/(\1reloc/g' \ -i \ arch.h \ arch/x86/decode.c \ check.c \ check.h \ elf.c \ elf.h \ orc_gen.c \ special.c Notable exceptions which complicate the regex include gelf_* library calls and standard/expected section names which still use "rela" because they encode the type of relocation expected. Also, keep "rela" in the struct because it encodes a specific type of relocation we currently expect. It will eventually turn into a member of an anonymous union when a susequent patch adds implicit addend, or "rel", relocation support. Signed-off-by: Matt Helsley <mhelsley@vmware.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
2020-05-30 05:01:13 +08:00
WARN("unexpected relocation symbol type in %s: %d", sec->name, reloc->sym->type);
continue;
}
func_for_each_insn(file, func, insn)
insn->ignore = true;
}
}
/*
* This is a whitelist of functions that is allowed to be called with AC set.
* The list is meant to be minimal and only contains compiler instrumentation
* ABI and a few functions used to implement *_{to,from}_user() functions.
*
* These functions must not directly change AC, but may PUSHF/POPF.
*/
static const char *uaccess_safe_builtin[] = {
/* KASAN */
"kasan_report",
kasan: prefix global functions with kasan_ Patch series "kasan: HW_TAGS tests support and fixes", v4. This patchset adds support for running KASAN-KUnit tests with the hardware tag-based mode and also contains a few fixes. This patch (of 15): There's a number of internal KASAN functions that are used across multiple source code files and therefore aren't marked as static inline. To avoid littering the kernel function names list with generic function names, prefix all such KASAN functions with kasan_. As a part of this change: - Rename internal (un)poison_range() to kasan_(un)poison() (no _range) to avoid name collision with a public kasan_unpoison_range(). - Rename check_memory_region() to kasan_check_range(), as it's a more fitting name. Link: https://lkml.kernel.org/r/cover.1610733117.git.andreyknvl@google.com Link: https://linux-review.googlesource.com/id/I719cc93483d4ba288a634dba80ee6b7f2809cd26 Link: https://lkml.kernel.org/r/13777aedf8d3ebbf35891136e1f2287e2f34aaba.1610733117.git.andreyknvl@google.com Signed-off-by: Andrey Konovalov <andreyknvl@google.com> Suggested-by: Marco Elver <elver@google.com> Reviewed-by: Marco Elver <elver@google.com> Reviewed-by: Alexander Potapenko <glider@google.com> Cc: Catalin Marinas <catalin.marinas@arm.com> Cc: Vincenzo Frascino <vincenzo.frascino@arm.com> Cc: Dmitry Vyukov <dvyukov@google.com> Cc: Will Deacon <will.deacon@arm.com> Cc: Andrey Ryabinin <aryabinin@virtuozzo.com> Cc: Peter Collingbourne <pcc@google.com> Cc: Evgenii Stepanov <eugenis@google.com> Cc: Branislav Rankov <Branislav.Rankov@arm.com> Cc: Kevin Brodsky <kevin.brodsky@arm.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2021-02-25 04:05:05 +08:00
"kasan_check_range",
/* KASAN out-of-line */
"__asan_loadN_noabort",
"__asan_load1_noabort",
"__asan_load2_noabort",
"__asan_load4_noabort",
"__asan_load8_noabort",
"__asan_load16_noabort",
"__asan_storeN_noabort",
"__asan_store1_noabort",
"__asan_store2_noabort",
"__asan_store4_noabort",
"__asan_store8_noabort",
"__asan_store16_noabort",
objtool: Permit __kasan_check_{read,write} under UACCESS Building linux-next with JUMP_LABEL=n and KASAN=y, I got this objtool warning: arch/x86/lib/copy_mc.o: warning: objtool: copy_mc_to_user()+0x22: call to __kasan_check_read() with UACCESS enabled What happens here is that copy_mc_to_user() branches on a static key in a UACCESS region:         __uaccess_begin();         if (static_branch_unlikely(&copy_mc_fragile_key))                 ret = copy_mc_fragile(to, from, len);         ret = copy_mc_generic(to, from, len);         __uaccess_end(); and the !CONFIG_JUMP_LABEL version of static_branch_unlikely() uses static_key_enabled(), which uses static_key_count(), which uses atomic_read(), which calls instrument_atomic_read(), which uses kasan_check_read(), which is __kasan_check_read(). Let's permit these KASAN helpers in UACCESS regions - static keys should probably work under UACCESS, I think. PeterZ adds: It's not a matter of permitting, it's a matter of being safe and correct. In this case it is, because it's a thin wrapper around check_memory_region() which was already marked safe. check_memory_region() is correct because the only thing it ends up calling is kasa_report() and that is also marked safe because that is annotated with user_access_save/restore() before it does anything else. On top of that, all of KASAN is noinstr, so nothing in here will end up in tracing and/or call schedule() before the user_access_save(). Signed-off-by: Jann Horn <jannh@google.com> Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
2020-09-29 06:49:16 +08:00
"__kasan_check_read",
"__kasan_check_write",
/* KASAN in-line */
"__asan_report_load_n_noabort",
"__asan_report_load1_noabort",
"__asan_report_load2_noabort",
"__asan_report_load4_noabort",
"__asan_report_load8_noabort",
"__asan_report_load16_noabort",
"__asan_report_store_n_noabort",
"__asan_report_store1_noabort",
"__asan_report_store2_noabort",
"__asan_report_store4_noabort",
"__asan_report_store8_noabort",
"__asan_report_store16_noabort",
/* KCSAN */
"__kcsan_check_access",
"__kcsan_mb",
"__kcsan_wmb",
"__kcsan_rmb",
"__kcsan_release",
"kcsan_found_watchpoint",
"kcsan_setup_watchpoint",
"kcsan_check_scoped_accesses",
"kcsan_disable_current",
"kcsan_enable_current_nowarn",
/* KCSAN/TSAN */
"__tsan_func_entry",
"__tsan_func_exit",
"__tsan_read_range",
"__tsan_write_range",
"__tsan_read1",
"__tsan_read2",
"__tsan_read4",
"__tsan_read8",
"__tsan_read16",
"__tsan_write1",
"__tsan_write2",
"__tsan_write4",
"__tsan_write8",
"__tsan_write16",
"__tsan_read_write1",
"__tsan_read_write2",
"__tsan_read_write4",
"__tsan_read_write8",
"__tsan_read_write16",
"__tsan_atomic8_load",
"__tsan_atomic16_load",
"__tsan_atomic32_load",
"__tsan_atomic64_load",
"__tsan_atomic8_store",
"__tsan_atomic16_store",
"__tsan_atomic32_store",
"__tsan_atomic64_store",
"__tsan_atomic8_exchange",
"__tsan_atomic16_exchange",
"__tsan_atomic32_exchange",
"__tsan_atomic64_exchange",
"__tsan_atomic8_fetch_add",
"__tsan_atomic16_fetch_add",
"__tsan_atomic32_fetch_add",
"__tsan_atomic64_fetch_add",
"__tsan_atomic8_fetch_sub",
"__tsan_atomic16_fetch_sub",
"__tsan_atomic32_fetch_sub",
"__tsan_atomic64_fetch_sub",
"__tsan_atomic8_fetch_and",
"__tsan_atomic16_fetch_and",
"__tsan_atomic32_fetch_and",
"__tsan_atomic64_fetch_and",
"__tsan_atomic8_fetch_or",
"__tsan_atomic16_fetch_or",
"__tsan_atomic32_fetch_or",
"__tsan_atomic64_fetch_or",
"__tsan_atomic8_fetch_xor",
"__tsan_atomic16_fetch_xor",
"__tsan_atomic32_fetch_xor",
"__tsan_atomic64_fetch_xor",
"__tsan_atomic8_fetch_nand",
"__tsan_atomic16_fetch_nand",
"__tsan_atomic32_fetch_nand",
"__tsan_atomic64_fetch_nand",
"__tsan_atomic8_compare_exchange_strong",
"__tsan_atomic16_compare_exchange_strong",
"__tsan_atomic32_compare_exchange_strong",
"__tsan_atomic64_compare_exchange_strong",
"__tsan_atomic8_compare_exchange_weak",
"__tsan_atomic16_compare_exchange_weak",
"__tsan_atomic32_compare_exchange_weak",
"__tsan_atomic64_compare_exchange_weak",
"__tsan_atomic8_compare_exchange_val",
"__tsan_atomic16_compare_exchange_val",
"__tsan_atomic32_compare_exchange_val",
"__tsan_atomic64_compare_exchange_val",
"__tsan_atomic_thread_fence",
"__tsan_atomic_signal_fence",
/* KCOV */
"write_comp_data",
"check_kcov_mode",
"__sanitizer_cov_trace_pc",
"__sanitizer_cov_trace_const_cmp1",
"__sanitizer_cov_trace_const_cmp2",
"__sanitizer_cov_trace_const_cmp4",
"__sanitizer_cov_trace_const_cmp8",
"__sanitizer_cov_trace_cmp1",
"__sanitizer_cov_trace_cmp2",
"__sanitizer_cov_trace_cmp4",
"__sanitizer_cov_trace_cmp8",
"__sanitizer_cov_trace_switch",
/* UBSAN */
"ubsan_type_mismatch_common",
"__ubsan_handle_type_mismatch",
"__ubsan_handle_type_mismatch_v1",
"__ubsan_handle_shift_out_of_bounds",
/* misc */
"csum_partial_copy_generic",
x86, powerpc: Rename memcpy_mcsafe() to copy_mc_to_{user, kernel}() In reaction to a proposal to introduce a memcpy_mcsafe_fast() implementation Linus points out that memcpy_mcsafe() is poorly named relative to communicating the scope of the interface. Specifically what addresses are valid to pass as source, destination, and what faults / exceptions are handled. Of particular concern is that even though x86 might be able to handle the semantics of copy_mc_to_user() with its common copy_user_generic() implementation other archs likely need / want an explicit path for this case: On Fri, May 1, 2020 at 11:28 AM Linus Torvalds <torvalds@linux-foundation.org> wrote: > > On Thu, Apr 30, 2020 at 6:21 PM Dan Williams <dan.j.williams@intel.com> wrote: > > > > However now I see that copy_user_generic() works for the wrong reason. > > It works because the exception on the source address due to poison > > looks no different than a write fault on the user address to the > > caller, it's still just a short copy. So it makes copy_to_user() work > > for the wrong reason relative to the name. > > Right. > > And it won't work that way on other architectures. On x86, we have a > generic function that can take faults on either side, and we use it > for both cases (and for the "in_user" case too), but that's an > artifact of the architecture oddity. > > In fact, it's probably wrong even on x86 - because it can hide bugs - > but writing those things is painful enough that everybody prefers > having just one function. Replace a single top-level memcpy_mcsafe() with either copy_mc_to_user(), or copy_mc_to_kernel(). Introduce an x86 copy_mc_fragile() name as the rename for the low-level x86 implementation formerly named memcpy_mcsafe(). It is used as the slow / careful backend that is supplanted by a fast copy_mc_generic() in a follow-on patch. One side-effect of this reorganization is that separating copy_mc_64.S to its own file means that perf no longer needs to track dependencies for its memcpy_64.S benchmarks. [ bp: Massage a bit. ] Signed-off-by: Dan Williams <dan.j.williams@intel.com> Signed-off-by: Borislav Petkov <bp@suse.de> Reviewed-by: Tony Luck <tony.luck@intel.com> Acked-by: Michael Ellerman <mpe@ellerman.id.au> Cc: <stable@vger.kernel.org> Link: http://lore.kernel.org/r/CAHk-=wjSqtXAqfUJxFtWNwmguFASTgB0dz1dT3V-78Quiezqbg@mail.gmail.com Link: https://lkml.kernel.org/r/160195561680.2163339.11574962055305783722.stgit@dwillia2-desk3.amr.corp.intel.com
2020-10-06 11:40:16 +08:00
"copy_mc_fragile",
"copy_mc_fragile_handle_tail",
x86/copy_mc: Introduce copy_mc_enhanced_fast_string() The motivations to go rework memcpy_mcsafe() are that the benefit of doing slow and careful copies is obviated on newer CPUs, and that the current opt-in list of CPUs to instrument recovery is broken relative to those CPUs. There is no need to keep an opt-in list up to date on an ongoing basis if pmem/dax operations are instrumented for recovery by default. With recovery enabled by default the old "mcsafe_key" opt-in to careful copying can be made a "fragile" opt-out. Where the "fragile" list takes steps to not consume poison across cachelines. The discussion with Linus made clear that the current "_mcsafe" suffix was imprecise to a fault. The operations that are needed by pmem/dax are to copy from a source address that might throw #MC to a destination that may write-fault, if it is a user page. So copy_to_user_mcsafe() becomes copy_mc_to_user() to indicate the separate precautions taken on source and destination. copy_mc_to_kernel() is introduced as a non-SMAP version that does not expect write-faults on the destination, but is still prepared to abort with an error code upon taking #MC. The original copy_mc_fragile() implementation had negative performance implications since it did not use the fast-string instruction sequence to perform copies. For this reason copy_mc_to_kernel() fell back to plain memcpy() to preserve performance on platforms that did not indicate the capability to recover from machine check exceptions. However, that capability detection was not architectural and now that some platforms can recover from fast-string consumption of memory errors the memcpy() fallback now causes these more capable platforms to fail. Introduce copy_mc_enhanced_fast_string() as the fast default implementation of copy_mc_to_kernel() and finalize the transition of copy_mc_fragile() to be a platform quirk to indicate 'copy-carefully'. With this in place, copy_mc_to_kernel() is fast and recovery-ready by default regardless of hardware capability. Thanks to Vivek for identifying that copy_user_generic() is not suitable as the copy_mc_to_user() backend since the #MC handler explicitly checks ex_has_fault_handler(). Thanks to the 0day robot for catching a performance bug in the x86/copy_mc_to_user implementation. [ bp: Add the "why" for this change from the 0/2th message, massage. ] Fixes: 92b0729c34ca ("x86/mm, x86/mce: Add memcpy_mcsafe()") Reported-by: Erwin Tsaur <erwin.tsaur@intel.com> Reported-by: 0day robot <lkp@intel.com> Signed-off-by: Dan Williams <dan.j.williams@intel.com> Signed-off-by: Borislav Petkov <bp@suse.de> Reviewed-by: Tony Luck <tony.luck@intel.com> Tested-by: Erwin Tsaur <erwin.tsaur@intel.com> Cc: <stable@vger.kernel.org> Link: https://lkml.kernel.org/r/160195562556.2163339.18063423034951948973.stgit@dwillia2-desk3.amr.corp.intel.com
2020-10-06 11:40:25 +08:00
"copy_mc_enhanced_fast_string",
"ftrace_likely_update", /* CONFIG_TRACE_BRANCH_PROFILING */
NULL
};
static void add_uaccess_safe(struct objtool_file *file)
{
struct symbol *func;
const char **name;
if (!opts.uaccess)
return;
for (name = uaccess_safe_builtin; *name; name++) {
func = find_symbol_by_name(file->elf, *name);
if (!func)
continue;
func->uaccess_safe = true;
}
}
/*
* FIXME: For now, just ignore any alternatives which add retpolines. This is
* a temporary hack, as it doesn't allow ORC to unwind from inside a retpoline.
* But it at least allows objtool to understand the control flow *around* the
* retpoline.
*/
static int add_ignore_alternatives(struct objtool_file *file)
{
struct section *sec;
objtool: Rename rela to reloc Before supporting additional relocation types rename the relevant types and functions from "rela" to "reloc". This work be done with the following regex: sed -e 's/struct rela/struct reloc/g' \ -e 's/\([_\*]\)rela\(s\{0,1\}\)/\1reloc\2/g' \ -e 's/tmprela\(s\{0,1\}\)/tmpreloc\1/g' \ -e 's/relasec/relocsec/g' \ -e 's/rela_list/reloc_list/g' \ -e 's/rela_hash/reloc_hash/g' \ -e 's/add_rela/add_reloc/g' \ -e 's/rela->/reloc->/g' \ -e '/rela[,\.]/{ s/\([^\.>]\)rela\([\.,]\)/\1reloc\2/g ; }' \ -e 's/rela =/reloc =/g' \ -e 's/relas =/relocs =/g' \ -e 's/relas\[/relocs[/g' \ -e 's/relaname =/relocname =/g' \ -e 's/= rela\;/= reloc\;/g' \ -e 's/= relas\;/= relocs\;/g' \ -e 's/= relaname\;/= relocname\;/g' \ -e 's/, rela)/, reloc)/g' \ -e 's/\([ @]\)rela\([ "]\)/\1reloc\2/g' \ -e 's/ rela$/ reloc/g' \ -e 's/, relaname/, relocname/g' \ -e 's/sec->rela/sec->reloc/g' \ -e 's/(\(!\{0,1\}\)rela/(\1reloc/g' \ -i \ arch.h \ arch/x86/decode.c \ check.c \ check.h \ elf.c \ elf.h \ orc_gen.c \ special.c Notable exceptions which complicate the regex include gelf_* library calls and standard/expected section names which still use "rela" because they encode the type of relocation expected. Also, keep "rela" in the struct because it encodes a specific type of relocation we currently expect. It will eventually turn into a member of an anonymous union when a susequent patch adds implicit addend, or "rel", relocation support. Signed-off-by: Matt Helsley <mhelsley@vmware.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
2020-05-30 05:01:13 +08:00
struct reloc *reloc;
struct instruction *insn;
sec = find_section_by_name(file->elf, ".rela.discard.ignore_alts");
if (!sec)
return 0;
objtool: Rename rela to reloc Before supporting additional relocation types rename the relevant types and functions from "rela" to "reloc". This work be done with the following regex: sed -e 's/struct rela/struct reloc/g' \ -e 's/\([_\*]\)rela\(s\{0,1\}\)/\1reloc\2/g' \ -e 's/tmprela\(s\{0,1\}\)/tmpreloc\1/g' \ -e 's/relasec/relocsec/g' \ -e 's/rela_list/reloc_list/g' \ -e 's/rela_hash/reloc_hash/g' \ -e 's/add_rela/add_reloc/g' \ -e 's/rela->/reloc->/g' \ -e '/rela[,\.]/{ s/\([^\.>]\)rela\([\.,]\)/\1reloc\2/g ; }' \ -e 's/rela =/reloc =/g' \ -e 's/relas =/relocs =/g' \ -e 's/relas\[/relocs[/g' \ -e 's/relaname =/relocname =/g' \ -e 's/= rela\;/= reloc\;/g' \ -e 's/= relas\;/= relocs\;/g' \ -e 's/= relaname\;/= relocname\;/g' \ -e 's/, rela)/, reloc)/g' \ -e 's/\([ @]\)rela\([ "]\)/\1reloc\2/g' \ -e 's/ rela$/ reloc/g' \ -e 's/, relaname/, relocname/g' \ -e 's/sec->rela/sec->reloc/g' \ -e 's/(\(!\{0,1\}\)rela/(\1reloc/g' \ -i \ arch.h \ arch/x86/decode.c \ check.c \ check.h \ elf.c \ elf.h \ orc_gen.c \ special.c Notable exceptions which complicate the regex include gelf_* library calls and standard/expected section names which still use "rela" because they encode the type of relocation expected. Also, keep "rela" in the struct because it encodes a specific type of relocation we currently expect. It will eventually turn into a member of an anonymous union when a susequent patch adds implicit addend, or "rel", relocation support. Signed-off-by: Matt Helsley <mhelsley@vmware.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
2020-05-30 05:01:13 +08:00
list_for_each_entry(reloc, &sec->reloc_list, list) {
if (reloc->sym->type != STT_SECTION) {
WARN("unexpected relocation symbol type in %s", sec->name);
return -1;
}
objtool: Rename rela to reloc Before supporting additional relocation types rename the relevant types and functions from "rela" to "reloc". This work be done with the following regex: sed -e 's/struct rela/struct reloc/g' \ -e 's/\([_\*]\)rela\(s\{0,1\}\)/\1reloc\2/g' \ -e 's/tmprela\(s\{0,1\}\)/tmpreloc\1/g' \ -e 's/relasec/relocsec/g' \ -e 's/rela_list/reloc_list/g' \ -e 's/rela_hash/reloc_hash/g' \ -e 's/add_rela/add_reloc/g' \ -e 's/rela->/reloc->/g' \ -e '/rela[,\.]/{ s/\([^\.>]\)rela\([\.,]\)/\1reloc\2/g ; }' \ -e 's/rela =/reloc =/g' \ -e 's/relas =/relocs =/g' \ -e 's/relas\[/relocs[/g' \ -e 's/relaname =/relocname =/g' \ -e 's/= rela\;/= reloc\;/g' \ -e 's/= relas\;/= relocs\;/g' \ -e 's/= relaname\;/= relocname\;/g' \ -e 's/, rela)/, reloc)/g' \ -e 's/\([ @]\)rela\([ "]\)/\1reloc\2/g' \ -e 's/ rela$/ reloc/g' \ -e 's/, relaname/, relocname/g' \ -e 's/sec->rela/sec->reloc/g' \ -e 's/(\(!\{0,1\}\)rela/(\1reloc/g' \ -i \ arch.h \ arch/x86/decode.c \ check.c \ check.h \ elf.c \ elf.h \ orc_gen.c \ special.c Notable exceptions which complicate the regex include gelf_* library calls and standard/expected section names which still use "rela" because they encode the type of relocation expected. Also, keep "rela" in the struct because it encodes a specific type of relocation we currently expect. It will eventually turn into a member of an anonymous union when a susequent patch adds implicit addend, or "rel", relocation support. Signed-off-by: Matt Helsley <mhelsley@vmware.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
2020-05-30 05:01:13 +08:00
insn = find_insn(file, reloc->sym->sec, reloc->addend);
if (!insn) {
WARN("bad .discard.ignore_alts entry");
return -1;
}
insn->ignore_alts = true;
}
return 0;
}
__weak bool arch_is_retpoline(struct symbol *sym)
{
return false;
}
__weak bool arch_is_rethunk(struct symbol *sym)
{
return false;
}
#define NEGATIVE_RELOC ((void *)-1L)
static struct reloc *insn_reloc(struct objtool_file *file, struct instruction *insn)
{
if (insn->reloc == NEGATIVE_RELOC)
return NULL;
if (!insn->reloc) {
if (!file)
return NULL;
insn->reloc = find_reloc_by_dest_range(file->elf, insn->sec,
insn->offset, insn->len);
if (!insn->reloc) {
insn->reloc = NEGATIVE_RELOC;
return NULL;
}
}
return insn->reloc;
}
static void remove_insn_ops(struct instruction *insn)
{
struct stack_op *op, *tmp;
list_for_each_entry_safe(op, tmp, &insn->stack_ops, list) {
list_del(&op->list);
free(op);
}
}
static void annotate_call_site(struct objtool_file *file,
struct instruction *insn, bool sibling)
{
struct reloc *reloc = insn_reloc(file, insn);
struct symbol *sym = insn->call_dest;
if (!sym)
sym = reloc->sym;
/*
* Alternative replacement code is just template code which is
* sometimes copied to the original instruction. For now, don't
* annotate it. (In the future we might consider annotating the
* original instruction if/when it ever makes sense to do so.)
*/
if (!strcmp(insn->sec->name, ".altinstr_replacement"))
return;
if (sym->static_call_tramp) {
list_add_tail(&insn->call_node, &file->static_call_list);
return;
}
if (sym->retpoline_thunk) {
list_add_tail(&insn->call_node, &file->retpoline_call_list);
return;
}
/*
* Many compilers cannot disable KCOV or sanitizer calls with a function
* attribute so they need a little help, NOP out any such calls from
* noinstr text.
*/
if (opts.hack_noinstr && insn->sec->noinstr && sym->profiling_func) {
if (reloc) {
reloc->type = R_NONE;
elf_write_reloc(file->elf, reloc);
}
elf_write_insn(file->elf, insn->sec,
insn->offset, insn->len,
sibling ? arch_ret_insn(insn->len)
: arch_nop_insn(insn->len));
insn->type = sibling ? INSN_RETURN : INSN_NOP;
if (sibling) {
/*
* We've replaced the tail-call JMP insn by two new
* insn: RET; INT3, except we only have a single struct
* insn here. Mark it retpoline_safe to avoid the SLS
* warning, instead of adding another insn.
*/
insn->retpoline_safe = true;
}
return;
}
if (opts.mcount && sym->fentry) {
if (sibling)
WARN_FUNC("Tail call to __fentry__ !?!?", insn->sec, insn->offset);
if (reloc) {
reloc->type = R_NONE;
elf_write_reloc(file->elf, reloc);
}
elf_write_insn(file->elf, insn->sec,
insn->offset, insn->len,
arch_nop_insn(insn->len));
insn->type = INSN_NOP;
list_add_tail(&insn->call_node, &file->mcount_loc_list);
return;
}
if (!sibling && dead_end_function(file, sym))
insn->dead_end = true;
}
static void add_call_dest(struct objtool_file *file, struct instruction *insn,
struct symbol *dest, bool sibling)
{
insn->call_dest = dest;
if (!dest)
return;
/*
* Whatever stack impact regular CALLs have, should be undone
* by the RETURN of the called function.
*
* Annotated intra-function calls retain the stack_ops but
* are converted to JUMP, see read_intra_function_calls().
*/
remove_insn_ops(insn);
annotate_call_site(file, insn, sibling);
}
static void add_retpoline_call(struct objtool_file *file, struct instruction *insn)
{
/*
* Retpoline calls/jumps are really dynamic calls/jumps in disguise,
* so convert them accordingly.
*/
switch (insn->type) {
case INSN_CALL:
insn->type = INSN_CALL_DYNAMIC;
break;
case INSN_JUMP_UNCONDITIONAL:
insn->type = INSN_JUMP_DYNAMIC;
break;
case INSN_JUMP_CONDITIONAL:
insn->type = INSN_JUMP_DYNAMIC_CONDITIONAL;
break;
default:
return;
}
insn->retpoline_safe = true;
/*
* Whatever stack impact regular CALLs have, should be undone
* by the RETURN of the called function.
*
* Annotated intra-function calls retain the stack_ops but
* are converted to JUMP, see read_intra_function_calls().
*/
remove_insn_ops(insn);
annotate_call_site(file, insn, false);
}
static void add_return_call(struct objtool_file *file, struct instruction *insn, bool add)
{
/*
* Return thunk tail calls are really just returns in disguise,
* so convert them accordingly.
*/
insn->type = INSN_RETURN;
insn->retpoline_safe = true;
if (add)
list_add_tail(&insn->call_node, &file->return_thunk_list);
}
static bool same_function(struct instruction *insn1, struct instruction *insn2)
{
return insn1->func->pfunc == insn2->func->pfunc;
}
objtool: Fix IBT tail-call detection Objtool reports: arch/x86/crypto/poly1305-x86_64.o: warning: objtool: poly1305_blocks_avx() falls through to next function poly1305_blocks_x86_64() arch/x86/crypto/poly1305-x86_64.o: warning: objtool: poly1305_emit_avx() falls through to next function poly1305_emit_x86_64() arch/x86/crypto/poly1305-x86_64.o: warning: objtool: poly1305_blocks_avx2() falls through to next function poly1305_blocks_x86_64() Which reads like: 0000000000000040 <poly1305_blocks_x86_64>: 40: f3 0f 1e fa endbr64 ... 0000000000000400 <poly1305_blocks_avx>: 400: f3 0f 1e fa endbr64 404: 44 8b 47 14 mov 0x14(%rdi),%r8d 408: 48 81 fa 80 00 00 00 cmp $0x80,%rdx 40f: 73 09 jae 41a <poly1305_blocks_avx+0x1a> 411: 45 85 c0 test %r8d,%r8d 414: 0f 84 2a fc ff ff je 44 <poly1305_blocks_x86_64+0x4> ... These are simple conditional tail-calls and *should* be recognised as such by objtool, however due to a mistake in commit 08f87a93c8ec ("objtool: Validate IBT assumptions") this is failing. Specifically, the jump_dest is +4, this means the instruction pointed at will not be ENDBR and as such it will fail the second clause of is_first_func_insn() that was supposed to capture this exact case. Instead, have is_first_func_insn() look at the previous instruction. Fixes: 08f87a93c8ec ("objtool: Validate IBT assumptions") Reported-by: Stephen Rothwell <sfr@canb.auug.org.au> Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Link: https://lkml.kernel.org/r/20220322115125.811582125@infradead.org
2022-03-22 19:33:31 +08:00
static bool is_first_func_insn(struct objtool_file *file, struct instruction *insn)
{
objtool: Fix IBT tail-call detection Objtool reports: arch/x86/crypto/poly1305-x86_64.o: warning: objtool: poly1305_blocks_avx() falls through to next function poly1305_blocks_x86_64() arch/x86/crypto/poly1305-x86_64.o: warning: objtool: poly1305_emit_avx() falls through to next function poly1305_emit_x86_64() arch/x86/crypto/poly1305-x86_64.o: warning: objtool: poly1305_blocks_avx2() falls through to next function poly1305_blocks_x86_64() Which reads like: 0000000000000040 <poly1305_blocks_x86_64>: 40: f3 0f 1e fa endbr64 ... 0000000000000400 <poly1305_blocks_avx>: 400: f3 0f 1e fa endbr64 404: 44 8b 47 14 mov 0x14(%rdi),%r8d 408: 48 81 fa 80 00 00 00 cmp $0x80,%rdx 40f: 73 09 jae 41a <poly1305_blocks_avx+0x1a> 411: 45 85 c0 test %r8d,%r8d 414: 0f 84 2a fc ff ff je 44 <poly1305_blocks_x86_64+0x4> ... These are simple conditional tail-calls and *should* be recognised as such by objtool, however due to a mistake in commit 08f87a93c8ec ("objtool: Validate IBT assumptions") this is failing. Specifically, the jump_dest is +4, this means the instruction pointed at will not be ENDBR and as such it will fail the second clause of is_first_func_insn() that was supposed to capture this exact case. Instead, have is_first_func_insn() look at the previous instruction. Fixes: 08f87a93c8ec ("objtool: Validate IBT assumptions") Reported-by: Stephen Rothwell <sfr@canb.auug.org.au> Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Link: https://lkml.kernel.org/r/20220322115125.811582125@infradead.org
2022-03-22 19:33:31 +08:00
if (insn->offset == insn->func->offset)
return true;
if (opts.ibt) {
objtool: Fix IBT tail-call detection Objtool reports: arch/x86/crypto/poly1305-x86_64.o: warning: objtool: poly1305_blocks_avx() falls through to next function poly1305_blocks_x86_64() arch/x86/crypto/poly1305-x86_64.o: warning: objtool: poly1305_emit_avx() falls through to next function poly1305_emit_x86_64() arch/x86/crypto/poly1305-x86_64.o: warning: objtool: poly1305_blocks_avx2() falls through to next function poly1305_blocks_x86_64() Which reads like: 0000000000000040 <poly1305_blocks_x86_64>: 40: f3 0f 1e fa endbr64 ... 0000000000000400 <poly1305_blocks_avx>: 400: f3 0f 1e fa endbr64 404: 44 8b 47 14 mov 0x14(%rdi),%r8d 408: 48 81 fa 80 00 00 00 cmp $0x80,%rdx 40f: 73 09 jae 41a <poly1305_blocks_avx+0x1a> 411: 45 85 c0 test %r8d,%r8d 414: 0f 84 2a fc ff ff je 44 <poly1305_blocks_x86_64+0x4> ... These are simple conditional tail-calls and *should* be recognised as such by objtool, however due to a mistake in commit 08f87a93c8ec ("objtool: Validate IBT assumptions") this is failing. Specifically, the jump_dest is +4, this means the instruction pointed at will not be ENDBR and as such it will fail the second clause of is_first_func_insn() that was supposed to capture this exact case. Instead, have is_first_func_insn() look at the previous instruction. Fixes: 08f87a93c8ec ("objtool: Validate IBT assumptions") Reported-by: Stephen Rothwell <sfr@canb.auug.org.au> Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Link: https://lkml.kernel.org/r/20220322115125.811582125@infradead.org
2022-03-22 19:33:31 +08:00
struct instruction *prev = prev_insn_same_sym(file, insn);
if (prev && prev->type == INSN_ENDBR &&
insn->offset == insn->func->offset + prev->len)
return true;
}
return false;
}
/*
* Find the destination instructions for all jumps.
*/
static int add_jump_destinations(struct objtool_file *file)
{
struct instruction *insn, *jump_dest;
objtool: Rename rela to reloc Before supporting additional relocation types rename the relevant types and functions from "rela" to "reloc". This work be done with the following regex: sed -e 's/struct rela/struct reloc/g' \ -e 's/\([_\*]\)rela\(s\{0,1\}\)/\1reloc\2/g' \ -e 's/tmprela\(s\{0,1\}\)/tmpreloc\1/g' \ -e 's/relasec/relocsec/g' \ -e 's/rela_list/reloc_list/g' \ -e 's/rela_hash/reloc_hash/g' \ -e 's/add_rela/add_reloc/g' \ -e 's/rela->/reloc->/g' \ -e '/rela[,\.]/{ s/\([^\.>]\)rela\([\.,]\)/\1reloc\2/g ; }' \ -e 's/rela =/reloc =/g' \ -e 's/relas =/relocs =/g' \ -e 's/relas\[/relocs[/g' \ -e 's/relaname =/relocname =/g' \ -e 's/= rela\;/= reloc\;/g' \ -e 's/= relas\;/= relocs\;/g' \ -e 's/= relaname\;/= relocname\;/g' \ -e 's/, rela)/, reloc)/g' \ -e 's/\([ @]\)rela\([ "]\)/\1reloc\2/g' \ -e 's/ rela$/ reloc/g' \ -e 's/, relaname/, relocname/g' \ -e 's/sec->rela/sec->reloc/g' \ -e 's/(\(!\{0,1\}\)rela/(\1reloc/g' \ -i \ arch.h \ arch/x86/decode.c \ check.c \ check.h \ elf.c \ elf.h \ orc_gen.c \ special.c Notable exceptions which complicate the regex include gelf_* library calls and standard/expected section names which still use "rela" because they encode the type of relocation expected. Also, keep "rela" in the struct because it encodes a specific type of relocation we currently expect. It will eventually turn into a member of an anonymous union when a susequent patch adds implicit addend, or "rel", relocation support. Signed-off-by: Matt Helsley <mhelsley@vmware.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
2020-05-30 05:01:13 +08:00
struct reloc *reloc;
struct section *dest_sec;
unsigned long dest_off;
for_each_insn(file, insn) {
if (insn->jump_dest) {
/*
* handle_group_alt() may have previously set
* 'jump_dest' for some alternatives.
*/
continue;
}
if (!is_static_jump(insn))
continue;
reloc = insn_reloc(file, insn);
objtool: Rename rela to reloc Before supporting additional relocation types rename the relevant types and functions from "rela" to "reloc". This work be done with the following regex: sed -e 's/struct rela/struct reloc/g' \ -e 's/\([_\*]\)rela\(s\{0,1\}\)/\1reloc\2/g' \ -e 's/tmprela\(s\{0,1\}\)/tmpreloc\1/g' \ -e 's/relasec/relocsec/g' \ -e 's/rela_list/reloc_list/g' \ -e 's/rela_hash/reloc_hash/g' \ -e 's/add_rela/add_reloc/g' \ -e 's/rela->/reloc->/g' \ -e '/rela[,\.]/{ s/\([^\.>]\)rela\([\.,]\)/\1reloc\2/g ; }' \ -e 's/rela =/reloc =/g' \ -e 's/relas =/relocs =/g' \ -e 's/relas\[/relocs[/g' \ -e 's/relaname =/relocname =/g' \ -e 's/= rela\;/= reloc\;/g' \ -e 's/= relas\;/= relocs\;/g' \ -e 's/= relaname\;/= relocname\;/g' \ -e 's/, rela)/, reloc)/g' \ -e 's/\([ @]\)rela\([ "]\)/\1reloc\2/g' \ -e 's/ rela$/ reloc/g' \ -e 's/, relaname/, relocname/g' \ -e 's/sec->rela/sec->reloc/g' \ -e 's/(\(!\{0,1\}\)rela/(\1reloc/g' \ -i \ arch.h \ arch/x86/decode.c \ check.c \ check.h \ elf.c \ elf.h \ orc_gen.c \ special.c Notable exceptions which complicate the regex include gelf_* library calls and standard/expected section names which still use "rela" because they encode the type of relocation expected. Also, keep "rela" in the struct because it encodes a specific type of relocation we currently expect. It will eventually turn into a member of an anonymous union when a susequent patch adds implicit addend, or "rel", relocation support. Signed-off-by: Matt Helsley <mhelsley@vmware.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
2020-05-30 05:01:13 +08:00
if (!reloc) {
dest_sec = insn->sec;
dest_off = arch_jump_destination(insn);
objtool: Rename rela to reloc Before supporting additional relocation types rename the relevant types and functions from "rela" to "reloc". This work be done with the following regex: sed -e 's/struct rela/struct reloc/g' \ -e 's/\([_\*]\)rela\(s\{0,1\}\)/\1reloc\2/g' \ -e 's/tmprela\(s\{0,1\}\)/tmpreloc\1/g' \ -e 's/relasec/relocsec/g' \ -e 's/rela_list/reloc_list/g' \ -e 's/rela_hash/reloc_hash/g' \ -e 's/add_rela/add_reloc/g' \ -e 's/rela->/reloc->/g' \ -e '/rela[,\.]/{ s/\([^\.>]\)rela\([\.,]\)/\1reloc\2/g ; }' \ -e 's/rela =/reloc =/g' \ -e 's/relas =/relocs =/g' \ -e 's/relas\[/relocs[/g' \ -e 's/relaname =/relocname =/g' \ -e 's/= rela\;/= reloc\;/g' \ -e 's/= relas\;/= relocs\;/g' \ -e 's/= relaname\;/= relocname\;/g' \ -e 's/, rela)/, reloc)/g' \ -e 's/\([ @]\)rela\([ "]\)/\1reloc\2/g' \ -e 's/ rela$/ reloc/g' \ -e 's/, relaname/, relocname/g' \ -e 's/sec->rela/sec->reloc/g' \ -e 's/(\(!\{0,1\}\)rela/(\1reloc/g' \ -i \ arch.h \ arch/x86/decode.c \ check.c \ check.h \ elf.c \ elf.h \ orc_gen.c \ special.c Notable exceptions which complicate the regex include gelf_* library calls and standard/expected section names which still use "rela" because they encode the type of relocation expected. Also, keep "rela" in the struct because it encodes a specific type of relocation we currently expect. It will eventually turn into a member of an anonymous union when a susequent patch adds implicit addend, or "rel", relocation support. Signed-off-by: Matt Helsley <mhelsley@vmware.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
2020-05-30 05:01:13 +08:00
} else if (reloc->sym->type == STT_SECTION) {
dest_sec = reloc->sym->sec;
dest_off = arch_dest_reloc_offset(reloc->addend);
} else if (reloc->sym->retpoline_thunk) {
add_retpoline_call(file, insn);
continue;
} else if (reloc->sym->return_thunk) {
add_return_call(file, insn, true);
continue;
} else if (insn->func) {
/*
* External sibling call or internal sibling call with
* STT_FUNC reloc.
*/
add_call_dest(file, insn, reloc->sym, true);
continue;
} else if (reloc->sym->sec->idx) {
dest_sec = reloc->sym->sec;
dest_off = reloc->sym->sym.st_value +
arch_dest_reloc_offset(reloc->addend);
} else {
/* non-func asm code jumping to another file */
continue;
}
jump_dest = find_insn(file, dest_sec, dest_off);
if (!jump_dest) {
struct symbol *sym = find_symbol_by_offset(dest_sec, dest_off);
/*
* This is a special case for zen_untrain_ret().
* It jumps to __x86_return_thunk(), but objtool
* can't find the thunk's starting RET
* instruction, because the RET is also in the
* middle of another instruction. Objtool only
* knows about the outer instruction.
*/
if (sym && sym->return_thunk) {
add_return_call(file, insn, false);
continue;
}
WARN_FUNC("can't find jump dest instruction at %s+0x%lx",
insn->sec, insn->offset, dest_sec->name,
dest_off);
return -1;
}
objtool: Fix GCC 8 cold subfunction detection for aliased functions The kbuild test robot reported the following issue: kernel/time/posix-stubs.o: warning: objtool: sys_ni_posix_timers.cold.1()+0x0: unreachable instruction This file creates symbol aliases for the sys_ni_posix_timers() function. So there are multiple ELF function symbols for the same function: 23: 0000000000000150 26 FUNC GLOBAL DEFAULT 1 __x64_sys_timer_create 24: 0000000000000150 26 FUNC GLOBAL DEFAULT 1 sys_ni_posix_timers 25: 0000000000000150 26 FUNC GLOBAL DEFAULT 1 __ia32_sys_timer_create 26: 0000000000000150 26 FUNC GLOBAL DEFAULT 1 __x64_sys_timer_gettime Here's the corresponding cold subfunction: 11: 0000000000000000 45 FUNC LOCAL DEFAULT 6 sys_ni_posix_timers.cold.1 When analyzing overlapping functions, objtool only looks at the first one in the symbol list. The rest of the functions are basically ignored because they point to instructions which have already been analyzed. So in this case it analyzes the __x64_sys_timer_create() function, but then it fails to recognize that its cold subfunction is sys_ni_posix_timers.cold.1(), because the names are different. Make the subfunction detection a little smarter by associating each subfunction with the first function which jumps to it, since that's the one which will be analyzed. Unfortunately we still have to leave the original subfunction detection code in place, thanks to GCC switch tables. (See the comment for more details.) Fixes: 13810435b9a7 ("objtool: Support GCC 8's cold subfunctions") Reported-by: kbuild test robot <lkp@intel.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com> Signed-off-by: Thomas Gleixner <tglx@linutronix.de> Cc: Peter Zijlstra <peterz@infradead.org> Link: https://lkml.kernel.org/r/d3ba52662cbc8e3a64a3b64d44b4efc5674fd9ab.1527855808.git.jpoimboe@redhat.com
2018-06-01 20:23:51 +08:00
/*
* Cross-function jump.
objtool: Fix GCC 8 cold subfunction detection for aliased functions The kbuild test robot reported the following issue: kernel/time/posix-stubs.o: warning: objtool: sys_ni_posix_timers.cold.1()+0x0: unreachable instruction This file creates symbol aliases for the sys_ni_posix_timers() function. So there are multiple ELF function symbols for the same function: 23: 0000000000000150 26 FUNC GLOBAL DEFAULT 1 __x64_sys_timer_create 24: 0000000000000150 26 FUNC GLOBAL DEFAULT 1 sys_ni_posix_timers 25: 0000000000000150 26 FUNC GLOBAL DEFAULT 1 __ia32_sys_timer_create 26: 0000000000000150 26 FUNC GLOBAL DEFAULT 1 __x64_sys_timer_gettime Here's the corresponding cold subfunction: 11: 0000000000000000 45 FUNC LOCAL DEFAULT 6 sys_ni_posix_timers.cold.1 When analyzing overlapping functions, objtool only looks at the first one in the symbol list. The rest of the functions are basically ignored because they point to instructions which have already been analyzed. So in this case it analyzes the __x64_sys_timer_create() function, but then it fails to recognize that its cold subfunction is sys_ni_posix_timers.cold.1(), because the names are different. Make the subfunction detection a little smarter by associating each subfunction with the first function which jumps to it, since that's the one which will be analyzed. Unfortunately we still have to leave the original subfunction detection code in place, thanks to GCC switch tables. (See the comment for more details.) Fixes: 13810435b9a7 ("objtool: Support GCC 8's cold subfunctions") Reported-by: kbuild test robot <lkp@intel.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com> Signed-off-by: Thomas Gleixner <tglx@linutronix.de> Cc: Peter Zijlstra <peterz@infradead.org> Link: https://lkml.kernel.org/r/d3ba52662cbc8e3a64a3b64d44b4efc5674fd9ab.1527855808.git.jpoimboe@redhat.com
2018-06-01 20:23:51 +08:00
*/
if (insn->func && jump_dest->func &&
insn->func != jump_dest->func) {
/*
* For GCC 8+, create parent/child links for any cold
* subfunctions. This is _mostly_ redundant with a
* similar initialization in read_symbols().
*
* If a function has aliases, we want the *first* such
* function in the symbol table to be the subfunction's
* parent. In that case we overwrite the
* initialization done in read_symbols().
*
* However this code can't completely replace the
* read_symbols() code because this doesn't detect the
* case where the parent function's only reference to a
* subfunction is through a jump table.
*/
if (!strstr(insn->func->name, ".cold") &&
strstr(jump_dest->func->name, ".cold")) {
insn->func->cfunc = jump_dest->func;
jump_dest->func->pfunc = insn->func;
} else if (!same_function(insn, jump_dest) &&
is_first_func_insn(file, jump_dest)) {
/*
* Internal sibling call without reloc or with
* STT_SECTION reloc.
*/
add_call_dest(file, insn, jump_dest->func, true);
continue;
}
objtool: Fix GCC 8 cold subfunction detection for aliased functions The kbuild test robot reported the following issue: kernel/time/posix-stubs.o: warning: objtool: sys_ni_posix_timers.cold.1()+0x0: unreachable instruction This file creates symbol aliases for the sys_ni_posix_timers() function. So there are multiple ELF function symbols for the same function: 23: 0000000000000150 26 FUNC GLOBAL DEFAULT 1 __x64_sys_timer_create 24: 0000000000000150 26 FUNC GLOBAL DEFAULT 1 sys_ni_posix_timers 25: 0000000000000150 26 FUNC GLOBAL DEFAULT 1 __ia32_sys_timer_create 26: 0000000000000150 26 FUNC GLOBAL DEFAULT 1 __x64_sys_timer_gettime Here's the corresponding cold subfunction: 11: 0000000000000000 45 FUNC LOCAL DEFAULT 6 sys_ni_posix_timers.cold.1 When analyzing overlapping functions, objtool only looks at the first one in the symbol list. The rest of the functions are basically ignored because they point to instructions which have already been analyzed. So in this case it analyzes the __x64_sys_timer_create() function, but then it fails to recognize that its cold subfunction is sys_ni_posix_timers.cold.1(), because the names are different. Make the subfunction detection a little smarter by associating each subfunction with the first function which jumps to it, since that's the one which will be analyzed. Unfortunately we still have to leave the original subfunction detection code in place, thanks to GCC switch tables. (See the comment for more details.) Fixes: 13810435b9a7 ("objtool: Support GCC 8's cold subfunctions") Reported-by: kbuild test robot <lkp@intel.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com> Signed-off-by: Thomas Gleixner <tglx@linutronix.de> Cc: Peter Zijlstra <peterz@infradead.org> Link: https://lkml.kernel.org/r/d3ba52662cbc8e3a64a3b64d44b4efc5674fd9ab.1527855808.git.jpoimboe@redhat.com
2018-06-01 20:23:51 +08:00
}
insn->jump_dest = jump_dest;
}
return 0;
}
static struct symbol *find_call_destination(struct section *sec, unsigned long offset)
{
struct symbol *call_dest;
call_dest = find_func_by_offset(sec, offset);
if (!call_dest)
call_dest = find_symbol_by_offset(sec, offset);
return call_dest;
}
/*
* Find the destination instructions for all calls.
*/
static int add_call_destinations(struct objtool_file *file)
{
struct instruction *insn;
unsigned long dest_off;
struct symbol *dest;
objtool: Rename rela to reloc Before supporting additional relocation types rename the relevant types and functions from "rela" to "reloc". This work be done with the following regex: sed -e 's/struct rela/struct reloc/g' \ -e 's/\([_\*]\)rela\(s\{0,1\}\)/\1reloc\2/g' \ -e 's/tmprela\(s\{0,1\}\)/tmpreloc\1/g' \ -e 's/relasec/relocsec/g' \ -e 's/rela_list/reloc_list/g' \ -e 's/rela_hash/reloc_hash/g' \ -e 's/add_rela/add_reloc/g' \ -e 's/rela->/reloc->/g' \ -e '/rela[,\.]/{ s/\([^\.>]\)rela\([\.,]\)/\1reloc\2/g ; }' \ -e 's/rela =/reloc =/g' \ -e 's/relas =/relocs =/g' \ -e 's/relas\[/relocs[/g' \ -e 's/relaname =/relocname =/g' \ -e 's/= rela\;/= reloc\;/g' \ -e 's/= relas\;/= relocs\;/g' \ -e 's/= relaname\;/= relocname\;/g' \ -e 's/, rela)/, reloc)/g' \ -e 's/\([ @]\)rela\([ "]\)/\1reloc\2/g' \ -e 's/ rela$/ reloc/g' \ -e 's/, relaname/, relocname/g' \ -e 's/sec->rela/sec->reloc/g' \ -e 's/(\(!\{0,1\}\)rela/(\1reloc/g' \ -i \ arch.h \ arch/x86/decode.c \ check.c \ check.h \ elf.c \ elf.h \ orc_gen.c \ special.c Notable exceptions which complicate the regex include gelf_* library calls and standard/expected section names which still use "rela" because they encode the type of relocation expected. Also, keep "rela" in the struct because it encodes a specific type of relocation we currently expect. It will eventually turn into a member of an anonymous union when a susequent patch adds implicit addend, or "rel", relocation support. Signed-off-by: Matt Helsley <mhelsley@vmware.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
2020-05-30 05:01:13 +08:00
struct reloc *reloc;
for_each_insn(file, insn) {
if (insn->type != INSN_CALL)
continue;
reloc = insn_reloc(file, insn);
objtool: Rename rela to reloc Before supporting additional relocation types rename the relevant types and functions from "rela" to "reloc". This work be done with the following regex: sed -e 's/struct rela/struct reloc/g' \ -e 's/\([_\*]\)rela\(s\{0,1\}\)/\1reloc\2/g' \ -e 's/tmprela\(s\{0,1\}\)/tmpreloc\1/g' \ -e 's/relasec/relocsec/g' \ -e 's/rela_list/reloc_list/g' \ -e 's/rela_hash/reloc_hash/g' \ -e 's/add_rela/add_reloc/g' \ -e 's/rela->/reloc->/g' \ -e '/rela[,\.]/{ s/\([^\.>]\)rela\([\.,]\)/\1reloc\2/g ; }' \ -e 's/rela =/reloc =/g' \ -e 's/relas =/relocs =/g' \ -e 's/relas\[/relocs[/g' \ -e 's/relaname =/relocname =/g' \ -e 's/= rela\;/= reloc\;/g' \ -e 's/= relas\;/= relocs\;/g' \ -e 's/= relaname\;/= relocname\;/g' \ -e 's/, rela)/, reloc)/g' \ -e 's/\([ @]\)rela\([ "]\)/\1reloc\2/g' \ -e 's/ rela$/ reloc/g' \ -e 's/, relaname/, relocname/g' \ -e 's/sec->rela/sec->reloc/g' \ -e 's/(\(!\{0,1\}\)rela/(\1reloc/g' \ -i \ arch.h \ arch/x86/decode.c \ check.c \ check.h \ elf.c \ elf.h \ orc_gen.c \ special.c Notable exceptions which complicate the regex include gelf_* library calls and standard/expected section names which still use "rela" because they encode the type of relocation expected. Also, keep "rela" in the struct because it encodes a specific type of relocation we currently expect. It will eventually turn into a member of an anonymous union when a susequent patch adds implicit addend, or "rel", relocation support. Signed-off-by: Matt Helsley <mhelsley@vmware.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
2020-05-30 05:01:13 +08:00
if (!reloc) {
dest_off = arch_jump_destination(insn);
dest = find_call_destination(insn->sec, dest_off);
add_call_dest(file, insn, dest, false);
if (insn->ignore)
continue;
if (!insn->call_dest) {
WARN_FUNC("unannotated intra-function call", insn->sec, insn->offset);
return -1;
}
if (insn->func && insn->call_dest->type != STT_FUNC) {
WARN_FUNC("unsupported call to non-function",
insn->sec, insn->offset);
return -1;
}
objtool: Rename rela to reloc Before supporting additional relocation types rename the relevant types and functions from "rela" to "reloc". This work be done with the following regex: sed -e 's/struct rela/struct reloc/g' \ -e 's/\([_\*]\)rela\(s\{0,1\}\)/\1reloc\2/g' \ -e 's/tmprela\(s\{0,1\}\)/tmpreloc\1/g' \ -e 's/relasec/relocsec/g' \ -e 's/rela_list/reloc_list/g' \ -e 's/rela_hash/reloc_hash/g' \ -e 's/add_rela/add_reloc/g' \ -e 's/rela->/reloc->/g' \ -e '/rela[,\.]/{ s/\([^\.>]\)rela\([\.,]\)/\1reloc\2/g ; }' \ -e 's/rela =/reloc =/g' \ -e 's/relas =/relocs =/g' \ -e 's/relas\[/relocs[/g' \ -e 's/relaname =/relocname =/g' \ -e 's/= rela\;/= reloc\;/g' \ -e 's/= relas\;/= relocs\;/g' \ -e 's/= relaname\;/= relocname\;/g' \ -e 's/, rela)/, reloc)/g' \ -e 's/\([ @]\)rela\([ "]\)/\1reloc\2/g' \ -e 's/ rela$/ reloc/g' \ -e 's/, relaname/, relocname/g' \ -e 's/sec->rela/sec->reloc/g' \ -e 's/(\(!\{0,1\}\)rela/(\1reloc/g' \ -i \ arch.h \ arch/x86/decode.c \ check.c \ check.h \ elf.c \ elf.h \ orc_gen.c \ special.c Notable exceptions which complicate the regex include gelf_* library calls and standard/expected section names which still use "rela" because they encode the type of relocation expected. Also, keep "rela" in the struct because it encodes a specific type of relocation we currently expect. It will eventually turn into a member of an anonymous union when a susequent patch adds implicit addend, or "rel", relocation support. Signed-off-by: Matt Helsley <mhelsley@vmware.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
2020-05-30 05:01:13 +08:00
} else if (reloc->sym->type == STT_SECTION) {
dest_off = arch_dest_reloc_offset(reloc->addend);
dest = find_call_destination(reloc->sym->sec, dest_off);
if (!dest) {
WARN_FUNC("can't find call dest symbol at %s+0x%lx",
insn->sec, insn->offset,
objtool: Rename rela to reloc Before supporting additional relocation types rename the relevant types and functions from "rela" to "reloc". This work be done with the following regex: sed -e 's/struct rela/struct reloc/g' \ -e 's/\([_\*]\)rela\(s\{0,1\}\)/\1reloc\2/g' \ -e 's/tmprela\(s\{0,1\}\)/tmpreloc\1/g' \ -e 's/relasec/relocsec/g' \ -e 's/rela_list/reloc_list/g' \ -e 's/rela_hash/reloc_hash/g' \ -e 's/add_rela/add_reloc/g' \ -e 's/rela->/reloc->/g' \ -e '/rela[,\.]/{ s/\([^\.>]\)rela\([\.,]\)/\1reloc\2/g ; }' \ -e 's/rela =/reloc =/g' \ -e 's/relas =/relocs =/g' \ -e 's/relas\[/relocs[/g' \ -e 's/relaname =/relocname =/g' \ -e 's/= rela\;/= reloc\;/g' \ -e 's/= relas\;/= relocs\;/g' \ -e 's/= relaname\;/= relocname\;/g' \ -e 's/, rela)/, reloc)/g' \ -e 's/\([ @]\)rela\([ "]\)/\1reloc\2/g' \ -e 's/ rela$/ reloc/g' \ -e 's/, relaname/, relocname/g' \ -e 's/sec->rela/sec->reloc/g' \ -e 's/(\(!\{0,1\}\)rela/(\1reloc/g' \ -i \ arch.h \ arch/x86/decode.c \ check.c \ check.h \ elf.c \ elf.h \ orc_gen.c \ special.c Notable exceptions which complicate the regex include gelf_* library calls and standard/expected section names which still use "rela" because they encode the type of relocation expected. Also, keep "rela" in the struct because it encodes a specific type of relocation we currently expect. It will eventually turn into a member of an anonymous union when a susequent patch adds implicit addend, or "rel", relocation support. Signed-off-by: Matt Helsley <mhelsley@vmware.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
2020-05-30 05:01:13 +08:00
reloc->sym->sec->name,
dest_off);
return -1;
}
add_call_dest(file, insn, dest, false);
} else if (reloc->sym->retpoline_thunk) {
add_retpoline_call(file, insn);
} else
add_call_dest(file, insn, reloc->sym, false);
}
return 0;
}
/*
objtool: Support stack layout changes in alternatives The ORC unwinder showed a warning [1] which revealed the stack layout didn't match what was expected. The problem was that paravirt patching had replaced "CALL *pv_ops.irq.save_fl" with "PUSHF;POP". That changed the stack layout between the PUSHF and the POP, so unwinding from an interrupt which occurred between those two instructions would fail. Part of the agreed upon solution was to rework the custom paravirt patching code to use alternatives instead, since objtool already knows how to read alternatives (and converging runtime patching infrastructure is always a good thing anyway). But the main problem still remains, which is that runtime patching can change the stack layout. Making stack layout changes in alternatives was disallowed with commit 7117f16bf460 ("objtool: Fix ORC vs alternatives"), but now that paravirt is going to be doing it, it needs to be supported. One way to do so would be to modify the ORC table when the code gets patched. But ORC is simple -- a good thing! -- and it's best to leave it alone. Instead, support stack layout changes by "flattening" all possible stack states (CFI) from parallel alternative code streams into a single set of linear states. The only necessary limitation is that CFI conflicts are disallowed at all possible instruction boundaries. For example, this scenario is allowed: Alt1 Alt2 Alt3 0x00 CALL *pv_ops.save_fl CALL xen_save_fl PUSHF 0x01 POP %RAX 0x02 NOP ... 0x05 NOP ... 0x07 <insn> The unwind information for offset-0x00 is identical for all 3 alternatives. Similarly offset-0x05 and higher also are identical (and the same as 0x00). However offset-0x01 has deviating CFI, but that is only relevant for Alt3, neither of the other alternative instruction streams will ever hit that offset. This scenario is NOT allowed: Alt1 Alt2 0x00 CALL *pv_ops.save_fl PUSHF 0x01 NOP6 ... 0x07 NOP POP %RAX The problem here is that offset-0x7, which is an instruction boundary in both possible instruction patch streams, has two conflicting stack layouts. [ The above examples were stolen from Peter Zijlstra. ] The new flattened CFI array is used both for the detection of conflicts (like the second example above) and the generation of linear ORC entries. BTW, another benefit of these changes is that, thanks to some related cleanups (new fake nops and alt_group struct) objtool can finally be rid of fake jumps, which were a constant source of headaches. [1] https://lkml.kernel.org/r/20201111170536.arx2zbn4ngvjoov7@treble Cc: Shinichiro Kawasaki <shinichiro.kawasaki@wdc.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
2020-12-19 04:26:21 +08:00
* The .alternatives section requires some extra special care over and above
* other special sections because alternatives are patched in place.
*/
static int handle_group_alt(struct objtool_file *file,
struct special_alt *special_alt,
struct instruction *orig_insn,
struct instruction **new_insn)
{
objtool: Support stack layout changes in alternatives The ORC unwinder showed a warning [1] which revealed the stack layout didn't match what was expected. The problem was that paravirt patching had replaced "CALL *pv_ops.irq.save_fl" with "PUSHF;POP". That changed the stack layout between the PUSHF and the POP, so unwinding from an interrupt which occurred between those two instructions would fail. Part of the agreed upon solution was to rework the custom paravirt patching code to use alternatives instead, since objtool already knows how to read alternatives (and converging runtime patching infrastructure is always a good thing anyway). But the main problem still remains, which is that runtime patching can change the stack layout. Making stack layout changes in alternatives was disallowed with commit 7117f16bf460 ("objtool: Fix ORC vs alternatives"), but now that paravirt is going to be doing it, it needs to be supported. One way to do so would be to modify the ORC table when the code gets patched. But ORC is simple -- a good thing! -- and it's best to leave it alone. Instead, support stack layout changes by "flattening" all possible stack states (CFI) from parallel alternative code streams into a single set of linear states. The only necessary limitation is that CFI conflicts are disallowed at all possible instruction boundaries. For example, this scenario is allowed: Alt1 Alt2 Alt3 0x00 CALL *pv_ops.save_fl CALL xen_save_fl PUSHF 0x01 POP %RAX 0x02 NOP ... 0x05 NOP ... 0x07 <insn> The unwind information for offset-0x00 is identical for all 3 alternatives. Similarly offset-0x05 and higher also are identical (and the same as 0x00). However offset-0x01 has deviating CFI, but that is only relevant for Alt3, neither of the other alternative instruction streams will ever hit that offset. This scenario is NOT allowed: Alt1 Alt2 0x00 CALL *pv_ops.save_fl PUSHF 0x01 NOP6 ... 0x07 NOP POP %RAX The problem here is that offset-0x7, which is an instruction boundary in both possible instruction patch streams, has two conflicting stack layouts. [ The above examples were stolen from Peter Zijlstra. ] The new flattened CFI array is used both for the detection of conflicts (like the second example above) and the generation of linear ORC entries. BTW, another benefit of these changes is that, thanks to some related cleanups (new fake nops and alt_group struct) objtool can finally be rid of fake jumps, which were a constant source of headaches. [1] https://lkml.kernel.org/r/20201111170536.arx2zbn4ngvjoov7@treble Cc: Shinichiro Kawasaki <shinichiro.kawasaki@wdc.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
2020-12-19 04:26:21 +08:00
struct instruction *last_orig_insn, *last_new_insn = NULL, *insn, *nop = NULL;
struct alt_group *orig_alt_group, *new_alt_group;
unsigned long dest_off;
orig_alt_group = malloc(sizeof(*orig_alt_group));
if (!orig_alt_group) {
WARN("malloc failed");
return -1;
}
objtool: Support stack layout changes in alternatives The ORC unwinder showed a warning [1] which revealed the stack layout didn't match what was expected. The problem was that paravirt patching had replaced "CALL *pv_ops.irq.save_fl" with "PUSHF;POP". That changed the stack layout between the PUSHF and the POP, so unwinding from an interrupt which occurred between those two instructions would fail. Part of the agreed upon solution was to rework the custom paravirt patching code to use alternatives instead, since objtool already knows how to read alternatives (and converging runtime patching infrastructure is always a good thing anyway). But the main problem still remains, which is that runtime patching can change the stack layout. Making stack layout changes in alternatives was disallowed with commit 7117f16bf460 ("objtool: Fix ORC vs alternatives"), but now that paravirt is going to be doing it, it needs to be supported. One way to do so would be to modify the ORC table when the code gets patched. But ORC is simple -- a good thing! -- and it's best to leave it alone. Instead, support stack layout changes by "flattening" all possible stack states (CFI) from parallel alternative code streams into a single set of linear states. The only necessary limitation is that CFI conflicts are disallowed at all possible instruction boundaries. For example, this scenario is allowed: Alt1 Alt2 Alt3 0x00 CALL *pv_ops.save_fl CALL xen_save_fl PUSHF 0x01 POP %RAX 0x02 NOP ... 0x05 NOP ... 0x07 <insn> The unwind information for offset-0x00 is identical for all 3 alternatives. Similarly offset-0x05 and higher also are identical (and the same as 0x00). However offset-0x01 has deviating CFI, but that is only relevant for Alt3, neither of the other alternative instruction streams will ever hit that offset. This scenario is NOT allowed: Alt1 Alt2 0x00 CALL *pv_ops.save_fl PUSHF 0x01 NOP6 ... 0x07 NOP POP %RAX The problem here is that offset-0x7, which is an instruction boundary in both possible instruction patch streams, has two conflicting stack layouts. [ The above examples were stolen from Peter Zijlstra. ] The new flattened CFI array is used both for the detection of conflicts (like the second example above) and the generation of linear ORC entries. BTW, another benefit of these changes is that, thanks to some related cleanups (new fake nops and alt_group struct) objtool can finally be rid of fake jumps, which were a constant source of headaches. [1] https://lkml.kernel.org/r/20201111170536.arx2zbn4ngvjoov7@treble Cc: Shinichiro Kawasaki <shinichiro.kawasaki@wdc.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
2020-12-19 04:26:21 +08:00
orig_alt_group->cfi = calloc(special_alt->orig_len,
sizeof(struct cfi_state *));
if (!orig_alt_group->cfi) {
WARN("calloc failed");
return -1;
}
last_orig_insn = NULL;
insn = orig_insn;
sec_for_each_insn_from(file, insn) {
if (insn->offset >= special_alt->orig_off + special_alt->orig_len)
break;
insn->alt_group = orig_alt_group;
last_orig_insn = insn;
}
orig_alt_group->orig_group = NULL;
orig_alt_group->first_insn = orig_insn;
orig_alt_group->last_insn = last_orig_insn;
objtool: Support stack layout changes in alternatives The ORC unwinder showed a warning [1] which revealed the stack layout didn't match what was expected. The problem was that paravirt patching had replaced "CALL *pv_ops.irq.save_fl" with "PUSHF;POP". That changed the stack layout between the PUSHF and the POP, so unwinding from an interrupt which occurred between those two instructions would fail. Part of the agreed upon solution was to rework the custom paravirt patching code to use alternatives instead, since objtool already knows how to read alternatives (and converging runtime patching infrastructure is always a good thing anyway). But the main problem still remains, which is that runtime patching can change the stack layout. Making stack layout changes in alternatives was disallowed with commit 7117f16bf460 ("objtool: Fix ORC vs alternatives"), but now that paravirt is going to be doing it, it needs to be supported. One way to do so would be to modify the ORC table when the code gets patched. But ORC is simple -- a good thing! -- and it's best to leave it alone. Instead, support stack layout changes by "flattening" all possible stack states (CFI) from parallel alternative code streams into a single set of linear states. The only necessary limitation is that CFI conflicts are disallowed at all possible instruction boundaries. For example, this scenario is allowed: Alt1 Alt2 Alt3 0x00 CALL *pv_ops.save_fl CALL xen_save_fl PUSHF 0x01 POP %RAX 0x02 NOP ... 0x05 NOP ... 0x07 <insn> The unwind information for offset-0x00 is identical for all 3 alternatives. Similarly offset-0x05 and higher also are identical (and the same as 0x00). However offset-0x01 has deviating CFI, but that is only relevant for Alt3, neither of the other alternative instruction streams will ever hit that offset. This scenario is NOT allowed: Alt1 Alt2 0x00 CALL *pv_ops.save_fl PUSHF 0x01 NOP6 ... 0x07 NOP POP %RAX The problem here is that offset-0x7, which is an instruction boundary in both possible instruction patch streams, has two conflicting stack layouts. [ The above examples were stolen from Peter Zijlstra. ] The new flattened CFI array is used both for the detection of conflicts (like the second example above) and the generation of linear ORC entries. BTW, another benefit of these changes is that, thanks to some related cleanups (new fake nops and alt_group struct) objtool can finally be rid of fake jumps, which were a constant source of headaches. [1] https://lkml.kernel.org/r/20201111170536.arx2zbn4ngvjoov7@treble Cc: Shinichiro Kawasaki <shinichiro.kawasaki@wdc.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
2020-12-19 04:26:21 +08:00
new_alt_group = malloc(sizeof(*new_alt_group));
if (!new_alt_group) {
WARN("malloc failed");
return -1;
}
objtool: Support stack layout changes in alternatives The ORC unwinder showed a warning [1] which revealed the stack layout didn't match what was expected. The problem was that paravirt patching had replaced "CALL *pv_ops.irq.save_fl" with "PUSHF;POP". That changed the stack layout between the PUSHF and the POP, so unwinding from an interrupt which occurred between those two instructions would fail. Part of the agreed upon solution was to rework the custom paravirt patching code to use alternatives instead, since objtool already knows how to read alternatives (and converging runtime patching infrastructure is always a good thing anyway). But the main problem still remains, which is that runtime patching can change the stack layout. Making stack layout changes in alternatives was disallowed with commit 7117f16bf460 ("objtool: Fix ORC vs alternatives"), but now that paravirt is going to be doing it, it needs to be supported. One way to do so would be to modify the ORC table when the code gets patched. But ORC is simple -- a good thing! -- and it's best to leave it alone. Instead, support stack layout changes by "flattening" all possible stack states (CFI) from parallel alternative code streams into a single set of linear states. The only necessary limitation is that CFI conflicts are disallowed at all possible instruction boundaries. For example, this scenario is allowed: Alt1 Alt2 Alt3 0x00 CALL *pv_ops.save_fl CALL xen_save_fl PUSHF 0x01 POP %RAX 0x02 NOP ... 0x05 NOP ... 0x07 <insn> The unwind information for offset-0x00 is identical for all 3 alternatives. Similarly offset-0x05 and higher also are identical (and the same as 0x00). However offset-0x01 has deviating CFI, but that is only relevant for Alt3, neither of the other alternative instruction streams will ever hit that offset. This scenario is NOT allowed: Alt1 Alt2 0x00 CALL *pv_ops.save_fl PUSHF 0x01 NOP6 ... 0x07 NOP POP %RAX The problem here is that offset-0x7, which is an instruction boundary in both possible instruction patch streams, has two conflicting stack layouts. [ The above examples were stolen from Peter Zijlstra. ] The new flattened CFI array is used both for the detection of conflicts (like the second example above) and the generation of linear ORC entries. BTW, another benefit of these changes is that, thanks to some related cleanups (new fake nops and alt_group struct) objtool can finally be rid of fake jumps, which were a constant source of headaches. [1] https://lkml.kernel.org/r/20201111170536.arx2zbn4ngvjoov7@treble Cc: Shinichiro Kawasaki <shinichiro.kawasaki@wdc.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
2020-12-19 04:26:21 +08:00
if (special_alt->new_len < special_alt->orig_len) {
/*
* Insert a fake nop at the end to make the replacement
* alt_group the same size as the original. This is needed to
* allow propagate_alt_cfi() to do its magic. When the last
* instruction affects the stack, the instruction after it (the
* nop) will propagate the new state to the shared CFI array.
*/
nop = malloc(sizeof(*nop));
if (!nop) {
WARN("malloc failed");
return -1;
}
objtool: Support stack layout changes in alternatives The ORC unwinder showed a warning [1] which revealed the stack layout didn't match what was expected. The problem was that paravirt patching had replaced "CALL *pv_ops.irq.save_fl" with "PUSHF;POP". That changed the stack layout between the PUSHF and the POP, so unwinding from an interrupt which occurred between those two instructions would fail. Part of the agreed upon solution was to rework the custom paravirt patching code to use alternatives instead, since objtool already knows how to read alternatives (and converging runtime patching infrastructure is always a good thing anyway). But the main problem still remains, which is that runtime patching can change the stack layout. Making stack layout changes in alternatives was disallowed with commit 7117f16bf460 ("objtool: Fix ORC vs alternatives"), but now that paravirt is going to be doing it, it needs to be supported. One way to do so would be to modify the ORC table when the code gets patched. But ORC is simple -- a good thing! -- and it's best to leave it alone. Instead, support stack layout changes by "flattening" all possible stack states (CFI) from parallel alternative code streams into a single set of linear states. The only necessary limitation is that CFI conflicts are disallowed at all possible instruction boundaries. For example, this scenario is allowed: Alt1 Alt2 Alt3 0x00 CALL *pv_ops.save_fl CALL xen_save_fl PUSHF 0x01 POP %RAX 0x02 NOP ... 0x05 NOP ... 0x07 <insn> The unwind information for offset-0x00 is identical for all 3 alternatives. Similarly offset-0x05 and higher also are identical (and the same as 0x00). However offset-0x01 has deviating CFI, but that is only relevant for Alt3, neither of the other alternative instruction streams will ever hit that offset. This scenario is NOT allowed: Alt1 Alt2 0x00 CALL *pv_ops.save_fl PUSHF 0x01 NOP6 ... 0x07 NOP POP %RAX The problem here is that offset-0x7, which is an instruction boundary in both possible instruction patch streams, has two conflicting stack layouts. [ The above examples were stolen from Peter Zijlstra. ] The new flattened CFI array is used both for the detection of conflicts (like the second example above) and the generation of linear ORC entries. BTW, another benefit of these changes is that, thanks to some related cleanups (new fake nops and alt_group struct) objtool can finally be rid of fake jumps, which were a constant source of headaches. [1] https://lkml.kernel.org/r/20201111170536.arx2zbn4ngvjoov7@treble Cc: Shinichiro Kawasaki <shinichiro.kawasaki@wdc.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
2020-12-19 04:26:21 +08:00
memset(nop, 0, sizeof(*nop));
INIT_LIST_HEAD(&nop->alts);
INIT_LIST_HEAD(&nop->stack_ops);
nop->sec = special_alt->new_sec;
nop->offset = special_alt->new_off + special_alt->new_len;
nop->len = special_alt->orig_len - special_alt->new_len;
nop->type = INSN_NOP;
nop->func = orig_insn->func;
nop->alt_group = new_alt_group;
nop->ignore = orig_insn->ignore_alts;
}
objtool: Support stack layout changes in alternatives The ORC unwinder showed a warning [1] which revealed the stack layout didn't match what was expected. The problem was that paravirt patching had replaced "CALL *pv_ops.irq.save_fl" with "PUSHF;POP". That changed the stack layout between the PUSHF and the POP, so unwinding from an interrupt which occurred between those two instructions would fail. Part of the agreed upon solution was to rework the custom paravirt patching code to use alternatives instead, since objtool already knows how to read alternatives (and converging runtime patching infrastructure is always a good thing anyway). But the main problem still remains, which is that runtime patching can change the stack layout. Making stack layout changes in alternatives was disallowed with commit 7117f16bf460 ("objtool: Fix ORC vs alternatives"), but now that paravirt is going to be doing it, it needs to be supported. One way to do so would be to modify the ORC table when the code gets patched. But ORC is simple -- a good thing! -- and it's best to leave it alone. Instead, support stack layout changes by "flattening" all possible stack states (CFI) from parallel alternative code streams into a single set of linear states. The only necessary limitation is that CFI conflicts are disallowed at all possible instruction boundaries. For example, this scenario is allowed: Alt1 Alt2 Alt3 0x00 CALL *pv_ops.save_fl CALL xen_save_fl PUSHF 0x01 POP %RAX 0x02 NOP ... 0x05 NOP ... 0x07 <insn> The unwind information for offset-0x00 is identical for all 3 alternatives. Similarly offset-0x05 and higher also are identical (and the same as 0x00). However offset-0x01 has deviating CFI, but that is only relevant for Alt3, neither of the other alternative instruction streams will ever hit that offset. This scenario is NOT allowed: Alt1 Alt2 0x00 CALL *pv_ops.save_fl PUSHF 0x01 NOP6 ... 0x07 NOP POP %RAX The problem here is that offset-0x7, which is an instruction boundary in both possible instruction patch streams, has two conflicting stack layouts. [ The above examples were stolen from Peter Zijlstra. ] The new flattened CFI array is used both for the detection of conflicts (like the second example above) and the generation of linear ORC entries. BTW, another benefit of these changes is that, thanks to some related cleanups (new fake nops and alt_group struct) objtool can finally be rid of fake jumps, which were a constant source of headaches. [1] https://lkml.kernel.org/r/20201111170536.arx2zbn4ngvjoov7@treble Cc: Shinichiro Kawasaki <shinichiro.kawasaki@wdc.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
2020-12-19 04:26:21 +08:00
if (!special_alt->new_len) {
*new_insn = nop;
goto end;
}
insn = *new_insn;
sec_for_each_insn_from(file, insn) {
struct reloc *alt_reloc;
if (insn->offset >= special_alt->new_off + special_alt->new_len)
break;
last_new_insn = insn;
insn->ignore = orig_insn->ignore_alts;
insn->func = orig_insn->func;
insn->alt_group = new_alt_group;
/*
* Since alternative replacement code is copy/pasted by the
* kernel after applying relocations, generally such code can't
* have relative-address relocation references to outside the
* .altinstr_replacement section, unless the arch's
* alternatives code can adjust the relative offsets
* accordingly.
*/
alt_reloc = insn_reloc(file, insn);
if (alt_reloc &&
!arch_support_alt_relocation(special_alt, insn, alt_reloc)) {
WARN_FUNC("unsupported relocation in alternatives section",
insn->sec, insn->offset);
return -1;
}
if (!is_static_jump(insn))
continue;
if (!insn->immediate)
continue;
dest_off = arch_jump_destination(insn);
if (dest_off == special_alt->new_off + special_alt->new_len) {
objtool: Support stack layout changes in alternatives The ORC unwinder showed a warning [1] which revealed the stack layout didn't match what was expected. The problem was that paravirt patching had replaced "CALL *pv_ops.irq.save_fl" with "PUSHF;POP". That changed the stack layout between the PUSHF and the POP, so unwinding from an interrupt which occurred between those two instructions would fail. Part of the agreed upon solution was to rework the custom paravirt patching code to use alternatives instead, since objtool already knows how to read alternatives (and converging runtime patching infrastructure is always a good thing anyway). But the main problem still remains, which is that runtime patching can change the stack layout. Making stack layout changes in alternatives was disallowed with commit 7117f16bf460 ("objtool: Fix ORC vs alternatives"), but now that paravirt is going to be doing it, it needs to be supported. One way to do so would be to modify the ORC table when the code gets patched. But ORC is simple -- a good thing! -- and it's best to leave it alone. Instead, support stack layout changes by "flattening" all possible stack states (CFI) from parallel alternative code streams into a single set of linear states. The only necessary limitation is that CFI conflicts are disallowed at all possible instruction boundaries. For example, this scenario is allowed: Alt1 Alt2 Alt3 0x00 CALL *pv_ops.save_fl CALL xen_save_fl PUSHF 0x01 POP %RAX 0x02 NOP ... 0x05 NOP ... 0x07 <insn> The unwind information for offset-0x00 is identical for all 3 alternatives. Similarly offset-0x05 and higher also are identical (and the same as 0x00). However offset-0x01 has deviating CFI, but that is only relevant for Alt3, neither of the other alternative instruction streams will ever hit that offset. This scenario is NOT allowed: Alt1 Alt2 0x00 CALL *pv_ops.save_fl PUSHF 0x01 NOP6 ... 0x07 NOP POP %RAX The problem here is that offset-0x7, which is an instruction boundary in both possible instruction patch streams, has two conflicting stack layouts. [ The above examples were stolen from Peter Zijlstra. ] The new flattened CFI array is used both for the detection of conflicts (like the second example above) and the generation of linear ORC entries. BTW, another benefit of these changes is that, thanks to some related cleanups (new fake nops and alt_group struct) objtool can finally be rid of fake jumps, which were a constant source of headaches. [1] https://lkml.kernel.org/r/20201111170536.arx2zbn4ngvjoov7@treble Cc: Shinichiro Kawasaki <shinichiro.kawasaki@wdc.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
2020-12-19 04:26:21 +08:00
insn->jump_dest = next_insn_same_sec(file, last_orig_insn);
if (!insn->jump_dest) {
WARN_FUNC("can't find alternative jump destination",
insn->sec, insn->offset);
return -1;
}
}
}
if (!last_new_insn) {
WARN_FUNC("can't find last new alternative instruction",
special_alt->new_sec, special_alt->new_off);
return -1;
}
objtool: Support stack layout changes in alternatives The ORC unwinder showed a warning [1] which revealed the stack layout didn't match what was expected. The problem was that paravirt patching had replaced "CALL *pv_ops.irq.save_fl" with "PUSHF;POP". That changed the stack layout between the PUSHF and the POP, so unwinding from an interrupt which occurred between those two instructions would fail. Part of the agreed upon solution was to rework the custom paravirt patching code to use alternatives instead, since objtool already knows how to read alternatives (and converging runtime patching infrastructure is always a good thing anyway). But the main problem still remains, which is that runtime patching can change the stack layout. Making stack layout changes in alternatives was disallowed with commit 7117f16bf460 ("objtool: Fix ORC vs alternatives"), but now that paravirt is going to be doing it, it needs to be supported. One way to do so would be to modify the ORC table when the code gets patched. But ORC is simple -- a good thing! -- and it's best to leave it alone. Instead, support stack layout changes by "flattening" all possible stack states (CFI) from parallel alternative code streams into a single set of linear states. The only necessary limitation is that CFI conflicts are disallowed at all possible instruction boundaries. For example, this scenario is allowed: Alt1 Alt2 Alt3 0x00 CALL *pv_ops.save_fl CALL xen_save_fl PUSHF 0x01 POP %RAX 0x02 NOP ... 0x05 NOP ... 0x07 <insn> The unwind information for offset-0x00 is identical for all 3 alternatives. Similarly offset-0x05 and higher also are identical (and the same as 0x00). However offset-0x01 has deviating CFI, but that is only relevant for Alt3, neither of the other alternative instruction streams will ever hit that offset. This scenario is NOT allowed: Alt1 Alt2 0x00 CALL *pv_ops.save_fl PUSHF 0x01 NOP6 ... 0x07 NOP POP %RAX The problem here is that offset-0x7, which is an instruction boundary in both possible instruction patch streams, has two conflicting stack layouts. [ The above examples were stolen from Peter Zijlstra. ] The new flattened CFI array is used both for the detection of conflicts (like the second example above) and the generation of linear ORC entries. BTW, another benefit of these changes is that, thanks to some related cleanups (new fake nops and alt_group struct) objtool can finally be rid of fake jumps, which were a constant source of headaches. [1] https://lkml.kernel.org/r/20201111170536.arx2zbn4ngvjoov7@treble Cc: Shinichiro Kawasaki <shinichiro.kawasaki@wdc.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
2020-12-19 04:26:21 +08:00
if (nop)
list_add(&nop->list, &last_new_insn->list);
end:
new_alt_group->orig_group = orig_alt_group;
new_alt_group->first_insn = *new_insn;
objtool: Support stack layout changes in alternatives The ORC unwinder showed a warning [1] which revealed the stack layout didn't match what was expected. The problem was that paravirt patching had replaced "CALL *pv_ops.irq.save_fl" with "PUSHF;POP". That changed the stack layout between the PUSHF and the POP, so unwinding from an interrupt which occurred between those two instructions would fail. Part of the agreed upon solution was to rework the custom paravirt patching code to use alternatives instead, since objtool already knows how to read alternatives (and converging runtime patching infrastructure is always a good thing anyway). But the main problem still remains, which is that runtime patching can change the stack layout. Making stack layout changes in alternatives was disallowed with commit 7117f16bf460 ("objtool: Fix ORC vs alternatives"), but now that paravirt is going to be doing it, it needs to be supported. One way to do so would be to modify the ORC table when the code gets patched. But ORC is simple -- a good thing! -- and it's best to leave it alone. Instead, support stack layout changes by "flattening" all possible stack states (CFI) from parallel alternative code streams into a single set of linear states. The only necessary limitation is that CFI conflicts are disallowed at all possible instruction boundaries. For example, this scenario is allowed: Alt1 Alt2 Alt3 0x00 CALL *pv_ops.save_fl CALL xen_save_fl PUSHF 0x01 POP %RAX 0x02 NOP ... 0x05 NOP ... 0x07 <insn> The unwind information for offset-0x00 is identical for all 3 alternatives. Similarly offset-0x05 and higher also are identical (and the same as 0x00). However offset-0x01 has deviating CFI, but that is only relevant for Alt3, neither of the other alternative instruction streams will ever hit that offset. This scenario is NOT allowed: Alt1 Alt2 0x00 CALL *pv_ops.save_fl PUSHF 0x01 NOP6 ... 0x07 NOP POP %RAX The problem here is that offset-0x7, which is an instruction boundary in both possible instruction patch streams, has two conflicting stack layouts. [ The above examples were stolen from Peter Zijlstra. ] The new flattened CFI array is used both for the detection of conflicts (like the second example above) and the generation of linear ORC entries. BTW, another benefit of these changes is that, thanks to some related cleanups (new fake nops and alt_group struct) objtool can finally be rid of fake jumps, which were a constant source of headaches. [1] https://lkml.kernel.org/r/20201111170536.arx2zbn4ngvjoov7@treble Cc: Shinichiro Kawasaki <shinichiro.kawasaki@wdc.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
2020-12-19 04:26:21 +08:00
new_alt_group->last_insn = nop ? : last_new_insn;
new_alt_group->cfi = orig_alt_group->cfi;
return 0;
}
/*
* A jump table entry can either convert a nop to a jump or a jump to a nop.
* If the original instruction is a jump, make the alt entry an effective nop
* by just skipping the original instruction.
*/
static int handle_jump_alt(struct objtool_file *file,
struct special_alt *special_alt,
struct instruction *orig_insn,
struct instruction **new_insn)
{
if (orig_insn->type != INSN_JUMP_UNCONDITIONAL &&
orig_insn->type != INSN_NOP) {
WARN_FUNC("unsupported instruction at jump label",
orig_insn->sec, orig_insn->offset);
return -1;
}
if (opts.hack_jump_label && special_alt->key_addend & 2) {
struct reloc *reloc = insn_reloc(file, orig_insn);
if (reloc) {
reloc->type = R_NONE;
elf_write_reloc(file->elf, reloc);
}
elf_write_insn(file->elf, orig_insn->sec,
orig_insn->offset, orig_insn->len,
arch_nop_insn(orig_insn->len));
orig_insn->type = INSN_NOP;
}
if (orig_insn->type == INSN_NOP) {
if (orig_insn->len == 2)
file->jl_nop_short++;
else
file->jl_nop_long++;
return 0;
}
if (orig_insn->len == 2)
file->jl_short++;
else
file->jl_long++;
*new_insn = list_next_entry(orig_insn, list);
return 0;
}
/*
* Read all the special sections which have alternate instructions which can be
* patched in or redirected to at runtime. Each instruction having alternate
* instruction(s) has them added to its insn->alts list, which will be
* traversed in validate_branch().
*/
static int add_special_section_alts(struct objtool_file *file)
{
struct list_head special_alts;
struct instruction *orig_insn, *new_insn;
struct special_alt *special_alt, *tmp;
struct alternative *alt;
int ret;
ret = special_get_alts(file->elf, &special_alts);
if (ret)
return ret;
list_for_each_entry_safe(special_alt, tmp, &special_alts, list) {
orig_insn = find_insn(file, special_alt->orig_sec,
special_alt->orig_off);
if (!orig_insn) {
WARN_FUNC("special: can't find orig instruction",
special_alt->orig_sec, special_alt->orig_off);
ret = -1;
goto out;
}
new_insn = NULL;
if (!special_alt->group || special_alt->new_len) {
new_insn = find_insn(file, special_alt->new_sec,
special_alt->new_off);
if (!new_insn) {
WARN_FUNC("special: can't find new instruction",
special_alt->new_sec,
special_alt->new_off);
ret = -1;
goto out;
}
}
if (special_alt->group) {
if (!special_alt->orig_len) {
WARN_FUNC("empty alternative entry",
orig_insn->sec, orig_insn->offset);
continue;
}
ret = handle_group_alt(file, special_alt, orig_insn,
&new_insn);
if (ret)
goto out;
} else if (special_alt->jump_or_nop) {
ret = handle_jump_alt(file, special_alt, orig_insn,
&new_insn);
if (ret)
goto out;
}
alt = malloc(sizeof(*alt));
if (!alt) {
WARN("malloc failed");
ret = -1;
goto out;
}
alt->insn = new_insn;
alt->skip_orig = special_alt->skip_orig;
orig_insn->ignore_alts |= special_alt->skip_alt;
list_add_tail(&alt->list, &orig_insn->alts);
list_del(&special_alt->list);
free(special_alt);
}
if (opts.stats) {
printf("jl\\\tNOP\tJMP\n");
printf("short:\t%ld\t%ld\n", file->jl_nop_short, file->jl_short);
printf("long:\t%ld\t%ld\n", file->jl_nop_long, file->jl_long);
}
out:
return ret;
}
static int add_jump_table(struct objtool_file *file, struct instruction *insn,
objtool: Rename rela to reloc Before supporting additional relocation types rename the relevant types and functions from "rela" to "reloc". This work be done with the following regex: sed -e 's/struct rela/struct reloc/g' \ -e 's/\([_\*]\)rela\(s\{0,1\}\)/\1reloc\2/g' \ -e 's/tmprela\(s\{0,1\}\)/tmpreloc\1/g' \ -e 's/relasec/relocsec/g' \ -e 's/rela_list/reloc_list/g' \ -e 's/rela_hash/reloc_hash/g' \ -e 's/add_rela/add_reloc/g' \ -e 's/rela->/reloc->/g' \ -e '/rela[,\.]/{ s/\([^\.>]\)rela\([\.,]\)/\1reloc\2/g ; }' \ -e 's/rela =/reloc =/g' \ -e 's/relas =/relocs =/g' \ -e 's/relas\[/relocs[/g' \ -e 's/relaname =/relocname =/g' \ -e 's/= rela\;/= reloc\;/g' \ -e 's/= relas\;/= relocs\;/g' \ -e 's/= relaname\;/= relocname\;/g' \ -e 's/, rela)/, reloc)/g' \ -e 's/\([ @]\)rela\([ "]\)/\1reloc\2/g' \ -e 's/ rela$/ reloc/g' \ -e 's/, relaname/, relocname/g' \ -e 's/sec->rela/sec->reloc/g' \ -e 's/(\(!\{0,1\}\)rela/(\1reloc/g' \ -i \ arch.h \ arch/x86/decode.c \ check.c \ check.h \ elf.c \ elf.h \ orc_gen.c \ special.c Notable exceptions which complicate the regex include gelf_* library calls and standard/expected section names which still use "rela" because they encode the type of relocation expected. Also, keep "rela" in the struct because it encodes a specific type of relocation we currently expect. It will eventually turn into a member of an anonymous union when a susequent patch adds implicit addend, or "rel", relocation support. Signed-off-by: Matt Helsley <mhelsley@vmware.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
2020-05-30 05:01:13 +08:00
struct reloc *table)
{
objtool: Rename rela to reloc Before supporting additional relocation types rename the relevant types and functions from "rela" to "reloc". This work be done with the following regex: sed -e 's/struct rela/struct reloc/g' \ -e 's/\([_\*]\)rela\(s\{0,1\}\)/\1reloc\2/g' \ -e 's/tmprela\(s\{0,1\}\)/tmpreloc\1/g' \ -e 's/relasec/relocsec/g' \ -e 's/rela_list/reloc_list/g' \ -e 's/rela_hash/reloc_hash/g' \ -e 's/add_rela/add_reloc/g' \ -e 's/rela->/reloc->/g' \ -e '/rela[,\.]/{ s/\([^\.>]\)rela\([\.,]\)/\1reloc\2/g ; }' \ -e 's/rela =/reloc =/g' \ -e 's/relas =/relocs =/g' \ -e 's/relas\[/relocs[/g' \ -e 's/relaname =/relocname =/g' \ -e 's/= rela\;/= reloc\;/g' \ -e 's/= relas\;/= relocs\;/g' \ -e 's/= relaname\;/= relocname\;/g' \ -e 's/, rela)/, reloc)/g' \ -e 's/\([ @]\)rela\([ "]\)/\1reloc\2/g' \ -e 's/ rela$/ reloc/g' \ -e 's/, relaname/, relocname/g' \ -e 's/sec->rela/sec->reloc/g' \ -e 's/(\(!\{0,1\}\)rela/(\1reloc/g' \ -i \ arch.h \ arch/x86/decode.c \ check.c \ check.h \ elf.c \ elf.h \ orc_gen.c \ special.c Notable exceptions which complicate the regex include gelf_* library calls and standard/expected section names which still use "rela" because they encode the type of relocation expected. Also, keep "rela" in the struct because it encodes a specific type of relocation we currently expect. It will eventually turn into a member of an anonymous union when a susequent patch adds implicit addend, or "rel", relocation support. Signed-off-by: Matt Helsley <mhelsley@vmware.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
2020-05-30 05:01:13 +08:00
struct reloc *reloc = table;
struct instruction *dest_insn;
struct alternative *alt;
objtool: Support GCC 8 switch tables With GCC 8, some issues were found with the objtool switch table detection. 1) In the .rodata section, immediately after the switch table, there can be another object which contains a pointer to the function which had the switch statement. In this case objtool wrongly considers the function pointer to be part of the switch table. Fix it by: a) making sure there are no pointers to the beginning of the function; and b) making sure there are no gaps in the switch table. Only the former was needed, the latter adds additional protection for future optimizations. 2) In find_switch_table(), case 1 and case 2 are missing the check to ensure that the .rodata switch table data is anonymous, i.e. that it isn't already associated with an ELF symbol. Fix it by adding the same find_symbol_containing() check which is used for case 3. This fixes the following warnings with GCC 8: drivers/block/virtio_blk.o: warning: objtool: virtio_queue_rq()+0x0: stack state mismatch: cfa1=7+8 cfa2=7+72 net/ipv6/icmp.o: warning: objtool: icmpv6_rcv()+0x0: stack state mismatch: cfa1=7+8 cfa2=7+64 drivers/usb/core/quirks.o: warning: objtool: quirks_param_set()+0x0: stack state mismatch: cfa1=7+8 cfa2=7+48 drivers/mtd/nand/raw/nand_hynix.o: warning: objtool: hynix_nand_decode_id()+0x0: stack state mismatch: cfa1=7+8 cfa2=7+24 drivers/mtd/nand/raw/nand_samsung.o: warning: objtool: samsung_nand_decode_id()+0x0: stack state mismatch: cfa1=7+8 cfa2=7+32 drivers/gpu/drm/nouveau/nvkm/subdev/top/gk104.o: warning: objtool: gk104_top_oneinit()+0x0: stack state mismatch: cfa1=7+8 cfa2=7+64 Reported-by: Arnd Bergmann <arnd@arndb.de> Reported-by: kbuild test robot <lkp@intel.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com> Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org> Cc: David Laight <David.Laight@ACULAB.COM> Cc: Greg KH <gregkh@linuxfoundation.org> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Randy Dunlap <rdunlap@infradead.org> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: damian <damian.tometzki@icloud.com> Link: http://lkml.kernel.org/r/20180510224849.xwi34d6tzheb5wgw@treble Signed-off-by: Ingo Molnar <mingo@kernel.org>
2018-05-11 06:48:49 +08:00
struct symbol *pfunc = insn->func->pfunc;
unsigned int prev_offset = 0;
/*
objtool: Rename rela to reloc Before supporting additional relocation types rename the relevant types and functions from "rela" to "reloc". This work be done with the following regex: sed -e 's/struct rela/struct reloc/g' \ -e 's/\([_\*]\)rela\(s\{0,1\}\)/\1reloc\2/g' \ -e 's/tmprela\(s\{0,1\}\)/tmpreloc\1/g' \ -e 's/relasec/relocsec/g' \ -e 's/rela_list/reloc_list/g' \ -e 's/rela_hash/reloc_hash/g' \ -e 's/add_rela/add_reloc/g' \ -e 's/rela->/reloc->/g' \ -e '/rela[,\.]/{ s/\([^\.>]\)rela\([\.,]\)/\1reloc\2/g ; }' \ -e 's/rela =/reloc =/g' \ -e 's/relas =/relocs =/g' \ -e 's/relas\[/relocs[/g' \ -e 's/relaname =/relocname =/g' \ -e 's/= rela\;/= reloc\;/g' \ -e 's/= relas\;/= relocs\;/g' \ -e 's/= relaname\;/= relocname\;/g' \ -e 's/, rela)/, reloc)/g' \ -e 's/\([ @]\)rela\([ "]\)/\1reloc\2/g' \ -e 's/ rela$/ reloc/g' \ -e 's/, relaname/, relocname/g' \ -e 's/sec->rela/sec->reloc/g' \ -e 's/(\(!\{0,1\}\)rela/(\1reloc/g' \ -i \ arch.h \ arch/x86/decode.c \ check.c \ check.h \ elf.c \ elf.h \ orc_gen.c \ special.c Notable exceptions which complicate the regex include gelf_* library calls and standard/expected section names which still use "rela" because they encode the type of relocation expected. Also, keep "rela" in the struct because it encodes a specific type of relocation we currently expect. It will eventually turn into a member of an anonymous union when a susequent patch adds implicit addend, or "rel", relocation support. Signed-off-by: Matt Helsley <mhelsley@vmware.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
2020-05-30 05:01:13 +08:00
* Each @reloc is a switch table relocation which points to the target
* instruction.
*/
objtool: Rename rela to reloc Before supporting additional relocation types rename the relevant types and functions from "rela" to "reloc". This work be done with the following regex: sed -e 's/struct rela/struct reloc/g' \ -e 's/\([_\*]\)rela\(s\{0,1\}\)/\1reloc\2/g' \ -e 's/tmprela\(s\{0,1\}\)/tmpreloc\1/g' \ -e 's/relasec/relocsec/g' \ -e 's/rela_list/reloc_list/g' \ -e 's/rela_hash/reloc_hash/g' \ -e 's/add_rela/add_reloc/g' \ -e 's/rela->/reloc->/g' \ -e '/rela[,\.]/{ s/\([^\.>]\)rela\([\.,]\)/\1reloc\2/g ; }' \ -e 's/rela =/reloc =/g' \ -e 's/relas =/relocs =/g' \ -e 's/relas\[/relocs[/g' \ -e 's/relaname =/relocname =/g' \ -e 's/= rela\;/= reloc\;/g' \ -e 's/= relas\;/= relocs\;/g' \ -e 's/= relaname\;/= relocname\;/g' \ -e 's/, rela)/, reloc)/g' \ -e 's/\([ @]\)rela\([ "]\)/\1reloc\2/g' \ -e 's/ rela$/ reloc/g' \ -e 's/, relaname/, relocname/g' \ -e 's/sec->rela/sec->reloc/g' \ -e 's/(\(!\{0,1\}\)rela/(\1reloc/g' \ -i \ arch.h \ arch/x86/decode.c \ check.c \ check.h \ elf.c \ elf.h \ orc_gen.c \ special.c Notable exceptions which complicate the regex include gelf_* library calls and standard/expected section names which still use "rela" because they encode the type of relocation expected. Also, keep "rela" in the struct because it encodes a specific type of relocation we currently expect. It will eventually turn into a member of an anonymous union when a susequent patch adds implicit addend, or "rel", relocation support. Signed-off-by: Matt Helsley <mhelsley@vmware.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
2020-05-30 05:01:13 +08:00
list_for_each_entry_from(reloc, &table->sec->reloc_list, list) {
objtool: Support repeated uses of the same C jump table This fixes objtool for both a GCC issue and a Clang issue: 1) GCC issue: kernel/bpf/core.o: warning: objtool: ___bpf_prog_run()+0x8d5: sibling call from callable instruction with modified stack frame With CONFIG_RETPOLINE=n, GCC is doing the following optimization in ___bpf_prog_run(). Before: select_insn: jmp *jumptable(,%rax,8) ... ALU64_ADD_X: ... jmp select_insn ALU_ADD_X: ... jmp select_insn After: select_insn: jmp *jumptable(, %rax, 8) ... ALU64_ADD_X: ... jmp *jumptable(, %rax, 8) ALU_ADD_X: ... jmp *jumptable(, %rax, 8) This confuses objtool. It has never seen multiple indirect jump sites which use the same jump table. For GCC switch tables, the only way of detecting the size of a table is by continuing to scan for more tables. The size of the previous table can only be determined after another switch table is found, or when the scan reaches the end of the function. That logic was reused for C jump tables, and was based on the assumption that each jump table only has a single jump site. The above optimization breaks that assumption. 2) Clang issue: drivers/usb/misc/sisusbvga/sisusb.o: warning: objtool: sisusb_write_mem_bulk()+0x588: can't find switch jump table With clang 9, code can be generated where a function contains two indirect jump instructions which use the same switch table. The fix is the same for both issues: split the jump table parsing into two passes. In the first pass, locate the heads of all switch tables for the function and mark their locations. In the second pass, parse the switch tables and add them. Fixes: e55a73251da3 ("bpf: Fix ORC unwinding in non-JIT BPF code") Reported-by: Randy Dunlap <rdunlap@infradead.org> Reported-by: Arnd Bergmann <arnd@arndb.de> Signed-off-by: Jann Horn <jannh@google.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com> Signed-off-by: Thomas Gleixner <tglx@linutronix.de> Tested-by: Nick Desaulniers <ndesaulniers@google.com> Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org> Link: https://lkml.kernel.org/r/e995befaada9d4d8b2cf788ff3f566ba900d2b4d.1563413318.git.jpoimboe@redhat.com Co-developed-by: Josh Poimboeuf <jpoimboe@redhat.com>
2019-07-18 09:36:54 +08:00
/* Check for the end of the table: */
objtool: Rename rela to reloc Before supporting additional relocation types rename the relevant types and functions from "rela" to "reloc". This work be done with the following regex: sed -e 's/struct rela/struct reloc/g' \ -e 's/\([_\*]\)rela\(s\{0,1\}\)/\1reloc\2/g' \ -e 's/tmprela\(s\{0,1\}\)/tmpreloc\1/g' \ -e 's/relasec/relocsec/g' \ -e 's/rela_list/reloc_list/g' \ -e 's/rela_hash/reloc_hash/g' \ -e 's/add_rela/add_reloc/g' \ -e 's/rela->/reloc->/g' \ -e '/rela[,\.]/{ s/\([^\.>]\)rela\([\.,]\)/\1reloc\2/g ; }' \ -e 's/rela =/reloc =/g' \ -e 's/relas =/relocs =/g' \ -e 's/relas\[/relocs[/g' \ -e 's/relaname =/relocname =/g' \ -e 's/= rela\;/= reloc\;/g' \ -e 's/= relas\;/= relocs\;/g' \ -e 's/= relaname\;/= relocname\;/g' \ -e 's/, rela)/, reloc)/g' \ -e 's/\([ @]\)rela\([ "]\)/\1reloc\2/g' \ -e 's/ rela$/ reloc/g' \ -e 's/, relaname/, relocname/g' \ -e 's/sec->rela/sec->reloc/g' \ -e 's/(\(!\{0,1\}\)rela/(\1reloc/g' \ -i \ arch.h \ arch/x86/decode.c \ check.c \ check.h \ elf.c \ elf.h \ orc_gen.c \ special.c Notable exceptions which complicate the regex include gelf_* library calls and standard/expected section names which still use "rela" because they encode the type of relocation expected. Also, keep "rela" in the struct because it encodes a specific type of relocation we currently expect. It will eventually turn into a member of an anonymous union when a susequent patch adds implicit addend, or "rel", relocation support. Signed-off-by: Matt Helsley <mhelsley@vmware.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
2020-05-30 05:01:13 +08:00
if (reloc != table && reloc->jump_table_start)
break;
/* Make sure the table entries are consecutive: */
objtool: Rename rela to reloc Before supporting additional relocation types rename the relevant types and functions from "rela" to "reloc". This work be done with the following regex: sed -e 's/struct rela/struct reloc/g' \ -e 's/\([_\*]\)rela\(s\{0,1\}\)/\1reloc\2/g' \ -e 's/tmprela\(s\{0,1\}\)/tmpreloc\1/g' \ -e 's/relasec/relocsec/g' \ -e 's/rela_list/reloc_list/g' \ -e 's/rela_hash/reloc_hash/g' \ -e 's/add_rela/add_reloc/g' \ -e 's/rela->/reloc->/g' \ -e '/rela[,\.]/{ s/\([^\.>]\)rela\([\.,]\)/\1reloc\2/g ; }' \ -e 's/rela =/reloc =/g' \ -e 's/relas =/relocs =/g' \ -e 's/relas\[/relocs[/g' \ -e 's/relaname =/relocname =/g' \ -e 's/= rela\;/= reloc\;/g' \ -e 's/= relas\;/= relocs\;/g' \ -e 's/= relaname\;/= relocname\;/g' \ -e 's/, rela)/, reloc)/g' \ -e 's/\([ @]\)rela\([ "]\)/\1reloc\2/g' \ -e 's/ rela$/ reloc/g' \ -e 's/, relaname/, relocname/g' \ -e 's/sec->rela/sec->reloc/g' \ -e 's/(\(!\{0,1\}\)rela/(\1reloc/g' \ -i \ arch.h \ arch/x86/decode.c \ check.c \ check.h \ elf.c \ elf.h \ orc_gen.c \ special.c Notable exceptions which complicate the regex include gelf_* library calls and standard/expected section names which still use "rela" because they encode the type of relocation expected. Also, keep "rela" in the struct because it encodes a specific type of relocation we currently expect. It will eventually turn into a member of an anonymous union when a susequent patch adds implicit addend, or "rel", relocation support. Signed-off-by: Matt Helsley <mhelsley@vmware.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
2020-05-30 05:01:13 +08:00
if (prev_offset && reloc->offset != prev_offset + 8)
objtool: Support GCC 8 switch tables With GCC 8, some issues were found with the objtool switch table detection. 1) In the .rodata section, immediately after the switch table, there can be another object which contains a pointer to the function which had the switch statement. In this case objtool wrongly considers the function pointer to be part of the switch table. Fix it by: a) making sure there are no pointers to the beginning of the function; and b) making sure there are no gaps in the switch table. Only the former was needed, the latter adds additional protection for future optimizations. 2) In find_switch_table(), case 1 and case 2 are missing the check to ensure that the .rodata switch table data is anonymous, i.e. that it isn't already associated with an ELF symbol. Fix it by adding the same find_symbol_containing() check which is used for case 3. This fixes the following warnings with GCC 8: drivers/block/virtio_blk.o: warning: objtool: virtio_queue_rq()+0x0: stack state mismatch: cfa1=7+8 cfa2=7+72 net/ipv6/icmp.o: warning: objtool: icmpv6_rcv()+0x0: stack state mismatch: cfa1=7+8 cfa2=7+64 drivers/usb/core/quirks.o: warning: objtool: quirks_param_set()+0x0: stack state mismatch: cfa1=7+8 cfa2=7+48 drivers/mtd/nand/raw/nand_hynix.o: warning: objtool: hynix_nand_decode_id()+0x0: stack state mismatch: cfa1=7+8 cfa2=7+24 drivers/mtd/nand/raw/nand_samsung.o: warning: objtool: samsung_nand_decode_id()+0x0: stack state mismatch: cfa1=7+8 cfa2=7+32 drivers/gpu/drm/nouveau/nvkm/subdev/top/gk104.o: warning: objtool: gk104_top_oneinit()+0x0: stack state mismatch: cfa1=7+8 cfa2=7+64 Reported-by: Arnd Bergmann <arnd@arndb.de> Reported-by: kbuild test robot <lkp@intel.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com> Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org> Cc: David Laight <David.Laight@ACULAB.COM> Cc: Greg KH <gregkh@linuxfoundation.org> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Randy Dunlap <rdunlap@infradead.org> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: damian <damian.tometzki@icloud.com> Link: http://lkml.kernel.org/r/20180510224849.xwi34d6tzheb5wgw@treble Signed-off-by: Ingo Molnar <mingo@kernel.org>
2018-05-11 06:48:49 +08:00
break;
/* Detect function pointers from contiguous objects: */
objtool: Rename rela to reloc Before supporting additional relocation types rename the relevant types and functions from "rela" to "reloc". This work be done with the following regex: sed -e 's/struct rela/struct reloc/g' \ -e 's/\([_\*]\)rela\(s\{0,1\}\)/\1reloc\2/g' \ -e 's/tmprela\(s\{0,1\}\)/tmpreloc\1/g' \ -e 's/relasec/relocsec/g' \ -e 's/rela_list/reloc_list/g' \ -e 's/rela_hash/reloc_hash/g' \ -e 's/add_rela/add_reloc/g' \ -e 's/rela->/reloc->/g' \ -e '/rela[,\.]/{ s/\([^\.>]\)rela\([\.,]\)/\1reloc\2/g ; }' \ -e 's/rela =/reloc =/g' \ -e 's/relas =/relocs =/g' \ -e 's/relas\[/relocs[/g' \ -e 's/relaname =/relocname =/g' \ -e 's/= rela\;/= reloc\;/g' \ -e 's/= relas\;/= relocs\;/g' \ -e 's/= relaname\;/= relocname\;/g' \ -e 's/, rela)/, reloc)/g' \ -e 's/\([ @]\)rela\([ "]\)/\1reloc\2/g' \ -e 's/ rela$/ reloc/g' \ -e 's/, relaname/, relocname/g' \ -e 's/sec->rela/sec->reloc/g' \ -e 's/(\(!\{0,1\}\)rela/(\1reloc/g' \ -i \ arch.h \ arch/x86/decode.c \ check.c \ check.h \ elf.c \ elf.h \ orc_gen.c \ special.c Notable exceptions which complicate the regex include gelf_* library calls and standard/expected section names which still use "rela" because they encode the type of relocation expected. Also, keep "rela" in the struct because it encodes a specific type of relocation we currently expect. It will eventually turn into a member of an anonymous union when a susequent patch adds implicit addend, or "rel", relocation support. Signed-off-by: Matt Helsley <mhelsley@vmware.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
2020-05-30 05:01:13 +08:00
if (reloc->sym->sec == pfunc->sec &&
reloc->addend == pfunc->offset)
objtool: Support GCC 8 switch tables With GCC 8, some issues were found with the objtool switch table detection. 1) In the .rodata section, immediately after the switch table, there can be another object which contains a pointer to the function which had the switch statement. In this case objtool wrongly considers the function pointer to be part of the switch table. Fix it by: a) making sure there are no pointers to the beginning of the function; and b) making sure there are no gaps in the switch table. Only the former was needed, the latter adds additional protection for future optimizations. 2) In find_switch_table(), case 1 and case 2 are missing the check to ensure that the .rodata switch table data is anonymous, i.e. that it isn't already associated with an ELF symbol. Fix it by adding the same find_symbol_containing() check which is used for case 3. This fixes the following warnings with GCC 8: drivers/block/virtio_blk.o: warning: objtool: virtio_queue_rq()+0x0: stack state mismatch: cfa1=7+8 cfa2=7+72 net/ipv6/icmp.o: warning: objtool: icmpv6_rcv()+0x0: stack state mismatch: cfa1=7+8 cfa2=7+64 drivers/usb/core/quirks.o: warning: objtool: quirks_param_set()+0x0: stack state mismatch: cfa1=7+8 cfa2=7+48 drivers/mtd/nand/raw/nand_hynix.o: warning: objtool: hynix_nand_decode_id()+0x0: stack state mismatch: cfa1=7+8 cfa2=7+24 drivers/mtd/nand/raw/nand_samsung.o: warning: objtool: samsung_nand_decode_id()+0x0: stack state mismatch: cfa1=7+8 cfa2=7+32 drivers/gpu/drm/nouveau/nvkm/subdev/top/gk104.o: warning: objtool: gk104_top_oneinit()+0x0: stack state mismatch: cfa1=7+8 cfa2=7+64 Reported-by: Arnd Bergmann <arnd@arndb.de> Reported-by: kbuild test robot <lkp@intel.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com> Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org> Cc: David Laight <David.Laight@ACULAB.COM> Cc: Greg KH <gregkh@linuxfoundation.org> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Randy Dunlap <rdunlap@infradead.org> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: damian <damian.tometzki@icloud.com> Link: http://lkml.kernel.org/r/20180510224849.xwi34d6tzheb5wgw@treble Signed-off-by: Ingo Molnar <mingo@kernel.org>
2018-05-11 06:48:49 +08:00
break;
objtool: Rename rela to reloc Before supporting additional relocation types rename the relevant types and functions from "rela" to "reloc". This work be done with the following regex: sed -e 's/struct rela/struct reloc/g' \ -e 's/\([_\*]\)rela\(s\{0,1\}\)/\1reloc\2/g' \ -e 's/tmprela\(s\{0,1\}\)/tmpreloc\1/g' \ -e 's/relasec/relocsec/g' \ -e 's/rela_list/reloc_list/g' \ -e 's/rela_hash/reloc_hash/g' \ -e 's/add_rela/add_reloc/g' \ -e 's/rela->/reloc->/g' \ -e '/rela[,\.]/{ s/\([^\.>]\)rela\([\.,]\)/\1reloc\2/g ; }' \ -e 's/rela =/reloc =/g' \ -e 's/relas =/relocs =/g' \ -e 's/relas\[/relocs[/g' \ -e 's/relaname =/relocname =/g' \ -e 's/= rela\;/= reloc\;/g' \ -e 's/= relas\;/= relocs\;/g' \ -e 's/= relaname\;/= relocname\;/g' \ -e 's/, rela)/, reloc)/g' \ -e 's/\([ @]\)rela\([ "]\)/\1reloc\2/g' \ -e 's/ rela$/ reloc/g' \ -e 's/, relaname/, relocname/g' \ -e 's/sec->rela/sec->reloc/g' \ -e 's/(\(!\{0,1\}\)rela/(\1reloc/g' \ -i \ arch.h \ arch/x86/decode.c \ check.c \ check.h \ elf.c \ elf.h \ orc_gen.c \ special.c Notable exceptions which complicate the regex include gelf_* library calls and standard/expected section names which still use "rela" because they encode the type of relocation expected. Also, keep "rela" in the struct because it encodes a specific type of relocation we currently expect. It will eventually turn into a member of an anonymous union when a susequent patch adds implicit addend, or "rel", relocation support. Signed-off-by: Matt Helsley <mhelsley@vmware.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
2020-05-30 05:01:13 +08:00
dest_insn = find_insn(file, reloc->sym->sec, reloc->addend);
if (!dest_insn)
break;
/* Make sure the destination is in the same function: */
objtool: Fix seg fault on bad switch table entry In one rare case, Clang generated the following code: 5ca: 83 e0 21 and $0x21,%eax 5cd: b9 04 00 00 00 mov $0x4,%ecx 5d2: ff 24 c5 00 00 00 00 jmpq *0x0(,%rax,8) 5d5: R_X86_64_32S .rodata+0x38 which uses the corresponding jump table relocations: 000000000038 000200000001 R_X86_64_64 0000000000000000 .text + 834 000000000040 000200000001 R_X86_64_64 0000000000000000 .text + 5d9 000000000048 000200000001 R_X86_64_64 0000000000000000 .text + b96 000000000050 000200000001 R_X86_64_64 0000000000000000 .text + b96 000000000058 000200000001 R_X86_64_64 0000000000000000 .text + b96 000000000060 000200000001 R_X86_64_64 0000000000000000 .text + b96 000000000068 000200000001 R_X86_64_64 0000000000000000 .text + b96 000000000070 000200000001 R_X86_64_64 0000000000000000 .text + b96 000000000078 000200000001 R_X86_64_64 0000000000000000 .text + b96 000000000080 000200000001 R_X86_64_64 0000000000000000 .text + b96 000000000088 000200000001 R_X86_64_64 0000000000000000 .text + b96 000000000090 000200000001 R_X86_64_64 0000000000000000 .text + b96 000000000098 000200000001 R_X86_64_64 0000000000000000 .text + b96 0000000000a0 000200000001 R_X86_64_64 0000000000000000 .text + b96 0000000000a8 000200000001 R_X86_64_64 0000000000000000 .text + b96 0000000000b0 000200000001 R_X86_64_64 0000000000000000 .text + b96 0000000000b8 000200000001 R_X86_64_64 0000000000000000 .text + b96 0000000000c0 000200000001 R_X86_64_64 0000000000000000 .text + b96 0000000000c8 000200000001 R_X86_64_64 0000000000000000 .text + b96 0000000000d0 000200000001 R_X86_64_64 0000000000000000 .text + b96 0000000000d8 000200000001 R_X86_64_64 0000000000000000 .text + b96 0000000000e0 000200000001 R_X86_64_64 0000000000000000 .text + b96 0000000000e8 000200000001 R_X86_64_64 0000000000000000 .text + b96 0000000000f0 000200000001 R_X86_64_64 0000000000000000 .text + b96 0000000000f8 000200000001 R_X86_64_64 0000000000000000 .text + b96 000000000100 000200000001 R_X86_64_64 0000000000000000 .text + b96 000000000108 000200000001 R_X86_64_64 0000000000000000 .text + b96 000000000110 000200000001 R_X86_64_64 0000000000000000 .text + b96 000000000118 000200000001 R_X86_64_64 0000000000000000 .text + b96 000000000120 000200000001 R_X86_64_64 0000000000000000 .text + b96 000000000128 000200000001 R_X86_64_64 0000000000000000 .text + b96 000000000130 000200000001 R_X86_64_64 0000000000000000 .text + b96 000000000138 000200000001 R_X86_64_64 0000000000000000 .text + 82f 000000000140 000200000001 R_X86_64_64 0000000000000000 .text + 828 Since %eax was masked with 0x21, only the first two and the last two entries are possible. Objtool doesn't actually emulate all the code, so it isn't smart enough to know that all the middle entries aren't reachable. They point to the NOP padding area after the end of the function, so objtool seg faulted when it tried to dereference a NULL insn->func. After this fix, objtool still gives an "unreachable" error because it stops reading the jump table when it encounters the bad addresses: /home/jpoimboe/objtool-tests/adm1275.o: warning: objtool: adm1275_probe()+0x828: unreachable instruction While the above code is technically correct, it's very wasteful of memory -- it uses 34 jump table entries when only 4 are needed. It's also not possible for objtool to validate this type of switch table because the unused entries point outside the function and objtool has no way of determining if that's intentional. Hopefully the Clang folks can fix it. Reported-by: Arnd Bergmann <arnd@arndb.de> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com> Signed-off-by: Thomas Gleixner <tglx@linutronix.de> Tested-by: Nick Desaulniers <ndesaulniers@google.com> Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org> Link: https://lkml.kernel.org/r/a9db88eec4f1ca089e040989846961748238b6d8.1563413318.git.jpoimboe@redhat.com
2019-07-18 09:36:55 +08:00
if (!dest_insn->func || dest_insn->func->pfunc != pfunc)
break;
alt = malloc(sizeof(*alt));
if (!alt) {
WARN("malloc failed");
return -1;
}
alt->insn = dest_insn;
list_add_tail(&alt->list, &insn->alts);
objtool: Rename rela to reloc Before supporting additional relocation types rename the relevant types and functions from "rela" to "reloc". This work be done with the following regex: sed -e 's/struct rela/struct reloc/g' \ -e 's/\([_\*]\)rela\(s\{0,1\}\)/\1reloc\2/g' \ -e 's/tmprela\(s\{0,1\}\)/tmpreloc\1/g' \ -e 's/relasec/relocsec/g' \ -e 's/rela_list/reloc_list/g' \ -e 's/rela_hash/reloc_hash/g' \ -e 's/add_rela/add_reloc/g' \ -e 's/rela->/reloc->/g' \ -e '/rela[,\.]/{ s/\([^\.>]\)rela\([\.,]\)/\1reloc\2/g ; }' \ -e 's/rela =/reloc =/g' \ -e 's/relas =/relocs =/g' \ -e 's/relas\[/relocs[/g' \ -e 's/relaname =/relocname =/g' \ -e 's/= rela\;/= reloc\;/g' \ -e 's/= relas\;/= relocs\;/g' \ -e 's/= relaname\;/= relocname\;/g' \ -e 's/, rela)/, reloc)/g' \ -e 's/\([ @]\)rela\([ "]\)/\1reloc\2/g' \ -e 's/ rela$/ reloc/g' \ -e 's/, relaname/, relocname/g' \ -e 's/sec->rela/sec->reloc/g' \ -e 's/(\(!\{0,1\}\)rela/(\1reloc/g' \ -i \ arch.h \ arch/x86/decode.c \ check.c \ check.h \ elf.c \ elf.h \ orc_gen.c \ special.c Notable exceptions which complicate the regex include gelf_* library calls and standard/expected section names which still use "rela" because they encode the type of relocation expected. Also, keep "rela" in the struct because it encodes a specific type of relocation we currently expect. It will eventually turn into a member of an anonymous union when a susequent patch adds implicit addend, or "rel", relocation support. Signed-off-by: Matt Helsley <mhelsley@vmware.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
2020-05-30 05:01:13 +08:00
prev_offset = reloc->offset;
objtool: Support GCC 8 switch tables With GCC 8, some issues were found with the objtool switch table detection. 1) In the .rodata section, immediately after the switch table, there can be another object which contains a pointer to the function which had the switch statement. In this case objtool wrongly considers the function pointer to be part of the switch table. Fix it by: a) making sure there are no pointers to the beginning of the function; and b) making sure there are no gaps in the switch table. Only the former was needed, the latter adds additional protection for future optimizations. 2) In find_switch_table(), case 1 and case 2 are missing the check to ensure that the .rodata switch table data is anonymous, i.e. that it isn't already associated with an ELF symbol. Fix it by adding the same find_symbol_containing() check which is used for case 3. This fixes the following warnings with GCC 8: drivers/block/virtio_blk.o: warning: objtool: virtio_queue_rq()+0x0: stack state mismatch: cfa1=7+8 cfa2=7+72 net/ipv6/icmp.o: warning: objtool: icmpv6_rcv()+0x0: stack state mismatch: cfa1=7+8 cfa2=7+64 drivers/usb/core/quirks.o: warning: objtool: quirks_param_set()+0x0: stack state mismatch: cfa1=7+8 cfa2=7+48 drivers/mtd/nand/raw/nand_hynix.o: warning: objtool: hynix_nand_decode_id()+0x0: stack state mismatch: cfa1=7+8 cfa2=7+24 drivers/mtd/nand/raw/nand_samsung.o: warning: objtool: samsung_nand_decode_id()+0x0: stack state mismatch: cfa1=7+8 cfa2=7+32 drivers/gpu/drm/nouveau/nvkm/subdev/top/gk104.o: warning: objtool: gk104_top_oneinit()+0x0: stack state mismatch: cfa1=7+8 cfa2=7+64 Reported-by: Arnd Bergmann <arnd@arndb.de> Reported-by: kbuild test robot <lkp@intel.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com> Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org> Cc: David Laight <David.Laight@ACULAB.COM> Cc: Greg KH <gregkh@linuxfoundation.org> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Randy Dunlap <rdunlap@infradead.org> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: damian <damian.tometzki@icloud.com> Link: http://lkml.kernel.org/r/20180510224849.xwi34d6tzheb5wgw@treble Signed-off-by: Ingo Molnar <mingo@kernel.org>
2018-05-11 06:48:49 +08:00
}
if (!prev_offset) {
WARN_FUNC("can't find switch jump table",
insn->sec, insn->offset);
return -1;
}
return 0;
}
/*
* find_jump_table() - Given a dynamic jump, find the switch jump table
* associated with it.
*/
objtool: Rename rela to reloc Before supporting additional relocation types rename the relevant types and functions from "rela" to "reloc". This work be done with the following regex: sed -e 's/struct rela/struct reloc/g' \ -e 's/\([_\*]\)rela\(s\{0,1\}\)/\1reloc\2/g' \ -e 's/tmprela\(s\{0,1\}\)/tmpreloc\1/g' \ -e 's/relasec/relocsec/g' \ -e 's/rela_list/reloc_list/g' \ -e 's/rela_hash/reloc_hash/g' \ -e 's/add_rela/add_reloc/g' \ -e 's/rela->/reloc->/g' \ -e '/rela[,\.]/{ s/\([^\.>]\)rela\([\.,]\)/\1reloc\2/g ; }' \ -e 's/rela =/reloc =/g' \ -e 's/relas =/relocs =/g' \ -e 's/relas\[/relocs[/g' \ -e 's/relaname =/relocname =/g' \ -e 's/= rela\;/= reloc\;/g' \ -e 's/= relas\;/= relocs\;/g' \ -e 's/= relaname\;/= relocname\;/g' \ -e 's/, rela)/, reloc)/g' \ -e 's/\([ @]\)rela\([ "]\)/\1reloc\2/g' \ -e 's/ rela$/ reloc/g' \ -e 's/, relaname/, relocname/g' \ -e 's/sec->rela/sec->reloc/g' \ -e 's/(\(!\{0,1\}\)rela/(\1reloc/g' \ -i \ arch.h \ arch/x86/decode.c \ check.c \ check.h \ elf.c \ elf.h \ orc_gen.c \ special.c Notable exceptions which complicate the regex include gelf_* library calls and standard/expected section names which still use "rela" because they encode the type of relocation expected. Also, keep "rela" in the struct because it encodes a specific type of relocation we currently expect. It will eventually turn into a member of an anonymous union when a susequent patch adds implicit addend, or "rel", relocation support. Signed-off-by: Matt Helsley <mhelsley@vmware.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
2020-05-30 05:01:13 +08:00
static struct reloc *find_jump_table(struct objtool_file *file,
struct symbol *func,
struct instruction *insn)
{
struct reloc *table_reloc;
struct instruction *dest_insn, *orig_insn = insn;
/*
* Backward search using the @first_jump_src links, these help avoid
* much of the 'in between' code. Which avoids us getting confused by
* it.
*/
objtool: Detect RIP-relative switch table references, part 2 With the following commit: fd35c88b7417 ("objtool: Support GCC 8 switch tables") I added a "can't find switch jump table" warning, to stop covering up silent failures if add_switch_table() can't find anything. That warning found yet another bug in the objtool switch table detection logic. For cases 1 and 2 (as described in the comments of find_switch_table()), the find_symbol_containing() check doesn't adjust the offset for RIP-relative switch jumps. Incidentally, this bug was already fixed for case 3 with: 6f5ec2993b1f ("objtool: Detect RIP-relative switch table references") However, that commit missed the fix for cases 1 and 2. The different cases are now starting to look more and more alike. So fix the bug by consolidating them into a single case, by checking the original dynamic jump instruction in the case 3 loop. This also simplifies the code and makes it more robust against future switch table detection issues -- of which I'm sure there will be many... Switch table detection has been the most fragile area of objtool, by far. I long for the day when we'll have a GCC plugin for annotating switch tables. Linus asked me to delay such a plugin due to the flakiness of the plugin infrastructure in older versions of GCC, so this rickety code is what we're stuck with for now. At least the code is now a little simpler than it was. Reported-by: kbuild test robot <lkp@intel.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Thomas Gleixner <tglx@linutronix.de> Link: http://lkml.kernel.org/r/f400541613d45689086329432f3095119ffbc328.1526674218.git.jpoimboe@redhat.com Signed-off-by: Ingo Molnar <mingo@kernel.org>
2018-05-19 04:10:34 +08:00
for (;
insn && insn->func && insn->func->pfunc == func;
insn = insn->first_jump_src ?: prev_insn_same_sym(file, insn)) {
objtool: Detect RIP-relative switch table references, part 2 With the following commit: fd35c88b7417 ("objtool: Support GCC 8 switch tables") I added a "can't find switch jump table" warning, to stop covering up silent failures if add_switch_table() can't find anything. That warning found yet another bug in the objtool switch table detection logic. For cases 1 and 2 (as described in the comments of find_switch_table()), the find_symbol_containing() check doesn't adjust the offset for RIP-relative switch jumps. Incidentally, this bug was already fixed for case 3 with: 6f5ec2993b1f ("objtool: Detect RIP-relative switch table references") However, that commit missed the fix for cases 1 and 2. The different cases are now starting to look more and more alike. So fix the bug by consolidating them into a single case, by checking the original dynamic jump instruction in the case 3 loop. This also simplifies the code and makes it more robust against future switch table detection issues -- of which I'm sure there will be many... Switch table detection has been the most fragile area of objtool, by far. I long for the day when we'll have a GCC plugin for annotating switch tables. Linus asked me to delay such a plugin due to the flakiness of the plugin infrastructure in older versions of GCC, so this rickety code is what we're stuck with for now. At least the code is now a little simpler than it was. Reported-by: kbuild test robot <lkp@intel.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Thomas Gleixner <tglx@linutronix.de> Link: http://lkml.kernel.org/r/f400541613d45689086329432f3095119ffbc328.1526674218.git.jpoimboe@redhat.com Signed-off-by: Ingo Molnar <mingo@kernel.org>
2018-05-19 04:10:34 +08:00
if (insn != orig_insn && insn->type == INSN_JUMP_DYNAMIC)
break;
/* allow small jumps within the range */
if (insn->type == INSN_JUMP_UNCONDITIONAL &&
insn->jump_dest &&
(insn->jump_dest->offset <= insn->offset ||
insn->jump_dest->offset > orig_insn->offset))
break;
table_reloc = arch_find_switch_table(file, insn);
objtool: Rename rela to reloc Before supporting additional relocation types rename the relevant types and functions from "rela" to "reloc". This work be done with the following regex: sed -e 's/struct rela/struct reloc/g' \ -e 's/\([_\*]\)rela\(s\{0,1\}\)/\1reloc\2/g' \ -e 's/tmprela\(s\{0,1\}\)/tmpreloc\1/g' \ -e 's/relasec/relocsec/g' \ -e 's/rela_list/reloc_list/g' \ -e 's/rela_hash/reloc_hash/g' \ -e 's/add_rela/add_reloc/g' \ -e 's/rela->/reloc->/g' \ -e '/rela[,\.]/{ s/\([^\.>]\)rela\([\.,]\)/\1reloc\2/g ; }' \ -e 's/rela =/reloc =/g' \ -e 's/relas =/relocs =/g' \ -e 's/relas\[/relocs[/g' \ -e 's/relaname =/relocname =/g' \ -e 's/= rela\;/= reloc\;/g' \ -e 's/= relas\;/= relocs\;/g' \ -e 's/= relaname\;/= relocname\;/g' \ -e 's/, rela)/, reloc)/g' \ -e 's/\([ @]\)rela\([ "]\)/\1reloc\2/g' \ -e 's/ rela$/ reloc/g' \ -e 's/, relaname/, relocname/g' \ -e 's/sec->rela/sec->reloc/g' \ -e 's/(\(!\{0,1\}\)rela/(\1reloc/g' \ -i \ arch.h \ arch/x86/decode.c \ check.c \ check.h \ elf.c \ elf.h \ orc_gen.c \ special.c Notable exceptions which complicate the regex include gelf_* library calls and standard/expected section names which still use "rela" because they encode the type of relocation expected. Also, keep "rela" in the struct because it encodes a specific type of relocation we currently expect. It will eventually turn into a member of an anonymous union when a susequent patch adds implicit addend, or "rel", relocation support. Signed-off-by: Matt Helsley <mhelsley@vmware.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
2020-05-30 05:01:13 +08:00
if (!table_reloc)
continue;
objtool: Rename rela to reloc Before supporting additional relocation types rename the relevant types and functions from "rela" to "reloc". This work be done with the following regex: sed -e 's/struct rela/struct reloc/g' \ -e 's/\([_\*]\)rela\(s\{0,1\}\)/\1reloc\2/g' \ -e 's/tmprela\(s\{0,1\}\)/tmpreloc\1/g' \ -e 's/relasec/relocsec/g' \ -e 's/rela_list/reloc_list/g' \ -e 's/rela_hash/reloc_hash/g' \ -e 's/add_rela/add_reloc/g' \ -e 's/rela->/reloc->/g' \ -e '/rela[,\.]/{ s/\([^\.>]\)rela\([\.,]\)/\1reloc\2/g ; }' \ -e 's/rela =/reloc =/g' \ -e 's/relas =/relocs =/g' \ -e 's/relas\[/relocs[/g' \ -e 's/relaname =/relocname =/g' \ -e 's/= rela\;/= reloc\;/g' \ -e 's/= relas\;/= relocs\;/g' \ -e 's/= relaname\;/= relocname\;/g' \ -e 's/, rela)/, reloc)/g' \ -e 's/\([ @]\)rela\([ "]\)/\1reloc\2/g' \ -e 's/ rela$/ reloc/g' \ -e 's/, relaname/, relocname/g' \ -e 's/sec->rela/sec->reloc/g' \ -e 's/(\(!\{0,1\}\)rela/(\1reloc/g' \ -i \ arch.h \ arch/x86/decode.c \ check.c \ check.h \ elf.c \ elf.h \ orc_gen.c \ special.c Notable exceptions which complicate the regex include gelf_* library calls and standard/expected section names which still use "rela" because they encode the type of relocation expected. Also, keep "rela" in the struct because it encodes a specific type of relocation we currently expect. It will eventually turn into a member of an anonymous union when a susequent patch adds implicit addend, or "rel", relocation support. Signed-off-by: Matt Helsley <mhelsley@vmware.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
2020-05-30 05:01:13 +08:00
dest_insn = find_insn(file, table_reloc->sym->sec, table_reloc->addend);
if (!dest_insn || !dest_insn->func || dest_insn->func->pfunc != func)
continue;
objtool: Detect RIP-relative switch table references, part 2 With the following commit: fd35c88b7417 ("objtool: Support GCC 8 switch tables") I added a "can't find switch jump table" warning, to stop covering up silent failures if add_switch_table() can't find anything. That warning found yet another bug in the objtool switch table detection logic. For cases 1 and 2 (as described in the comments of find_switch_table()), the find_symbol_containing() check doesn't adjust the offset for RIP-relative switch jumps. Incidentally, this bug was already fixed for case 3 with: 6f5ec2993b1f ("objtool: Detect RIP-relative switch table references") However, that commit missed the fix for cases 1 and 2. The different cases are now starting to look more and more alike. So fix the bug by consolidating them into a single case, by checking the original dynamic jump instruction in the case 3 loop. This also simplifies the code and makes it more robust against future switch table detection issues -- of which I'm sure there will be many... Switch table detection has been the most fragile area of objtool, by far. I long for the day when we'll have a GCC plugin for annotating switch tables. Linus asked me to delay such a plugin due to the flakiness of the plugin infrastructure in older versions of GCC, so this rickety code is what we're stuck with for now. At least the code is now a little simpler than it was. Reported-by: kbuild test robot <lkp@intel.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Thomas Gleixner <tglx@linutronix.de> Link: http://lkml.kernel.org/r/f400541613d45689086329432f3095119ffbc328.1526674218.git.jpoimboe@redhat.com Signed-off-by: Ingo Molnar <mingo@kernel.org>
2018-05-19 04:10:34 +08:00
objtool: Rename rela to reloc Before supporting additional relocation types rename the relevant types and functions from "rela" to "reloc". This work be done with the following regex: sed -e 's/struct rela/struct reloc/g' \ -e 's/\([_\*]\)rela\(s\{0,1\}\)/\1reloc\2/g' \ -e 's/tmprela\(s\{0,1\}\)/tmpreloc\1/g' \ -e 's/relasec/relocsec/g' \ -e 's/rela_list/reloc_list/g' \ -e 's/rela_hash/reloc_hash/g' \ -e 's/add_rela/add_reloc/g' \ -e 's/rela->/reloc->/g' \ -e '/rela[,\.]/{ s/\([^\.>]\)rela\([\.,]\)/\1reloc\2/g ; }' \ -e 's/rela =/reloc =/g' \ -e 's/relas =/relocs =/g' \ -e 's/relas\[/relocs[/g' \ -e 's/relaname =/relocname =/g' \ -e 's/= rela\;/= reloc\;/g' \ -e 's/= relas\;/= relocs\;/g' \ -e 's/= relaname\;/= relocname\;/g' \ -e 's/, rela)/, reloc)/g' \ -e 's/\([ @]\)rela\([ "]\)/\1reloc\2/g' \ -e 's/ rela$/ reloc/g' \ -e 's/, relaname/, relocname/g' \ -e 's/sec->rela/sec->reloc/g' \ -e 's/(\(!\{0,1\}\)rela/(\1reloc/g' \ -i \ arch.h \ arch/x86/decode.c \ check.c \ check.h \ elf.c \ elf.h \ orc_gen.c \ special.c Notable exceptions which complicate the regex include gelf_* library calls and standard/expected section names which still use "rela" because they encode the type of relocation expected. Also, keep "rela" in the struct because it encodes a specific type of relocation we currently expect. It will eventually turn into a member of an anonymous union when a susequent patch adds implicit addend, or "rel", relocation support. Signed-off-by: Matt Helsley <mhelsley@vmware.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
2020-05-30 05:01:13 +08:00
return table_reloc;
}
return NULL;
}
objtool: Support repeated uses of the same C jump table This fixes objtool for both a GCC issue and a Clang issue: 1) GCC issue: kernel/bpf/core.o: warning: objtool: ___bpf_prog_run()+0x8d5: sibling call from callable instruction with modified stack frame With CONFIG_RETPOLINE=n, GCC is doing the following optimization in ___bpf_prog_run(). Before: select_insn: jmp *jumptable(,%rax,8) ... ALU64_ADD_X: ... jmp select_insn ALU_ADD_X: ... jmp select_insn After: select_insn: jmp *jumptable(, %rax, 8) ... ALU64_ADD_X: ... jmp *jumptable(, %rax, 8) ALU_ADD_X: ... jmp *jumptable(, %rax, 8) This confuses objtool. It has never seen multiple indirect jump sites which use the same jump table. For GCC switch tables, the only way of detecting the size of a table is by continuing to scan for more tables. The size of the previous table can only be determined after another switch table is found, or when the scan reaches the end of the function. That logic was reused for C jump tables, and was based on the assumption that each jump table only has a single jump site. The above optimization breaks that assumption. 2) Clang issue: drivers/usb/misc/sisusbvga/sisusb.o: warning: objtool: sisusb_write_mem_bulk()+0x588: can't find switch jump table With clang 9, code can be generated where a function contains two indirect jump instructions which use the same switch table. The fix is the same for both issues: split the jump table parsing into two passes. In the first pass, locate the heads of all switch tables for the function and mark their locations. In the second pass, parse the switch tables and add them. Fixes: e55a73251da3 ("bpf: Fix ORC unwinding in non-JIT BPF code") Reported-by: Randy Dunlap <rdunlap@infradead.org> Reported-by: Arnd Bergmann <arnd@arndb.de> Signed-off-by: Jann Horn <jannh@google.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com> Signed-off-by: Thomas Gleixner <tglx@linutronix.de> Tested-by: Nick Desaulniers <ndesaulniers@google.com> Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org> Link: https://lkml.kernel.org/r/e995befaada9d4d8b2cf788ff3f566ba900d2b4d.1563413318.git.jpoimboe@redhat.com Co-developed-by: Josh Poimboeuf <jpoimboe@redhat.com>
2019-07-18 09:36:54 +08:00
/*
* First pass: Mark the head of each jump table so that in the next pass,
* we know when a given jump table ends and the next one starts.
*/
static void mark_func_jump_tables(struct objtool_file *file,
struct symbol *func)
{
objtool: Support repeated uses of the same C jump table This fixes objtool for both a GCC issue and a Clang issue: 1) GCC issue: kernel/bpf/core.o: warning: objtool: ___bpf_prog_run()+0x8d5: sibling call from callable instruction with modified stack frame With CONFIG_RETPOLINE=n, GCC is doing the following optimization in ___bpf_prog_run(). Before: select_insn: jmp *jumptable(,%rax,8) ... ALU64_ADD_X: ... jmp select_insn ALU_ADD_X: ... jmp select_insn After: select_insn: jmp *jumptable(, %rax, 8) ... ALU64_ADD_X: ... jmp *jumptable(, %rax, 8) ALU_ADD_X: ... jmp *jumptable(, %rax, 8) This confuses objtool. It has never seen multiple indirect jump sites which use the same jump table. For GCC switch tables, the only way of detecting the size of a table is by continuing to scan for more tables. The size of the previous table can only be determined after another switch table is found, or when the scan reaches the end of the function. That logic was reused for C jump tables, and was based on the assumption that each jump table only has a single jump site. The above optimization breaks that assumption. 2) Clang issue: drivers/usb/misc/sisusbvga/sisusb.o: warning: objtool: sisusb_write_mem_bulk()+0x588: can't find switch jump table With clang 9, code can be generated where a function contains two indirect jump instructions which use the same switch table. The fix is the same for both issues: split the jump table parsing into two passes. In the first pass, locate the heads of all switch tables for the function and mark their locations. In the second pass, parse the switch tables and add them. Fixes: e55a73251da3 ("bpf: Fix ORC unwinding in non-JIT BPF code") Reported-by: Randy Dunlap <rdunlap@infradead.org> Reported-by: Arnd Bergmann <arnd@arndb.de> Signed-off-by: Jann Horn <jannh@google.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com> Signed-off-by: Thomas Gleixner <tglx@linutronix.de> Tested-by: Nick Desaulniers <ndesaulniers@google.com> Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org> Link: https://lkml.kernel.org/r/e995befaada9d4d8b2cf788ff3f566ba900d2b4d.1563413318.git.jpoimboe@redhat.com Co-developed-by: Josh Poimboeuf <jpoimboe@redhat.com>
2019-07-18 09:36:54 +08:00
struct instruction *insn, *last = NULL;
objtool: Rename rela to reloc Before supporting additional relocation types rename the relevant types and functions from "rela" to "reloc". This work be done with the following regex: sed -e 's/struct rela/struct reloc/g' \ -e 's/\([_\*]\)rela\(s\{0,1\}\)/\1reloc\2/g' \ -e 's/tmprela\(s\{0,1\}\)/tmpreloc\1/g' \ -e 's/relasec/relocsec/g' \ -e 's/rela_list/reloc_list/g' \ -e 's/rela_hash/reloc_hash/g' \ -e 's/add_rela/add_reloc/g' \ -e 's/rela->/reloc->/g' \ -e '/rela[,\.]/{ s/\([^\.>]\)rela\([\.,]\)/\1reloc\2/g ; }' \ -e 's/rela =/reloc =/g' \ -e 's/relas =/relocs =/g' \ -e 's/relas\[/relocs[/g' \ -e 's/relaname =/relocname =/g' \ -e 's/= rela\;/= reloc\;/g' \ -e 's/= relas\;/= relocs\;/g' \ -e 's/= relaname\;/= relocname\;/g' \ -e 's/, rela)/, reloc)/g' \ -e 's/\([ @]\)rela\([ "]\)/\1reloc\2/g' \ -e 's/ rela$/ reloc/g' \ -e 's/, relaname/, relocname/g' \ -e 's/sec->rela/sec->reloc/g' \ -e 's/(\(!\{0,1\}\)rela/(\1reloc/g' \ -i \ arch.h \ arch/x86/decode.c \ check.c \ check.h \ elf.c \ elf.h \ orc_gen.c \ special.c Notable exceptions which complicate the regex include gelf_* library calls and standard/expected section names which still use "rela" because they encode the type of relocation expected. Also, keep "rela" in the struct because it encodes a specific type of relocation we currently expect. It will eventually turn into a member of an anonymous union when a susequent patch adds implicit addend, or "rel", relocation support. Signed-off-by: Matt Helsley <mhelsley@vmware.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
2020-05-30 05:01:13 +08:00
struct reloc *reloc;
func_for_each_insn(file, func, insn) {
if (!last)
last = insn;
/*
* Store back-pointers for unconditional forward jumps such
* that find_jump_table() can back-track using those and
* avoid some potentially confusing code.
*/
if (insn->type == INSN_JUMP_UNCONDITIONAL && insn->jump_dest &&
insn->offset > last->offset &&
insn->jump_dest->offset > insn->offset &&
!insn->jump_dest->first_jump_src) {
insn->jump_dest->first_jump_src = insn;
last = insn->jump_dest;
}
if (insn->type != INSN_JUMP_DYNAMIC)
continue;
objtool: Rename rela to reloc Before supporting additional relocation types rename the relevant types and functions from "rela" to "reloc". This work be done with the following regex: sed -e 's/struct rela/struct reloc/g' \ -e 's/\([_\*]\)rela\(s\{0,1\}\)/\1reloc\2/g' \ -e 's/tmprela\(s\{0,1\}\)/tmpreloc\1/g' \ -e 's/relasec/relocsec/g' \ -e 's/rela_list/reloc_list/g' \ -e 's/rela_hash/reloc_hash/g' \ -e 's/add_rela/add_reloc/g' \ -e 's/rela->/reloc->/g' \ -e '/rela[,\.]/{ s/\([^\.>]\)rela\([\.,]\)/\1reloc\2/g ; }' \ -e 's/rela =/reloc =/g' \ -e 's/relas =/relocs =/g' \ -e 's/relas\[/relocs[/g' \ -e 's/relaname =/relocname =/g' \ -e 's/= rela\;/= reloc\;/g' \ -e 's/= relas\;/= relocs\;/g' \ -e 's/= relaname\;/= relocname\;/g' \ -e 's/, rela)/, reloc)/g' \ -e 's/\([ @]\)rela\([ "]\)/\1reloc\2/g' \ -e 's/ rela$/ reloc/g' \ -e 's/, relaname/, relocname/g' \ -e 's/sec->rela/sec->reloc/g' \ -e 's/(\(!\{0,1\}\)rela/(\1reloc/g' \ -i \ arch.h \ arch/x86/decode.c \ check.c \ check.h \ elf.c \ elf.h \ orc_gen.c \ special.c Notable exceptions which complicate the regex include gelf_* library calls and standard/expected section names which still use "rela" because they encode the type of relocation expected. Also, keep "rela" in the struct because it encodes a specific type of relocation we currently expect. It will eventually turn into a member of an anonymous union when a susequent patch adds implicit addend, or "rel", relocation support. Signed-off-by: Matt Helsley <mhelsley@vmware.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
2020-05-30 05:01:13 +08:00
reloc = find_jump_table(file, func, insn);
if (reloc) {
reloc->jump_table_start = true;
insn->jump_table = reloc;
}
}
objtool: Support repeated uses of the same C jump table This fixes objtool for both a GCC issue and a Clang issue: 1) GCC issue: kernel/bpf/core.o: warning: objtool: ___bpf_prog_run()+0x8d5: sibling call from callable instruction with modified stack frame With CONFIG_RETPOLINE=n, GCC is doing the following optimization in ___bpf_prog_run(). Before: select_insn: jmp *jumptable(,%rax,8) ... ALU64_ADD_X: ... jmp select_insn ALU_ADD_X: ... jmp select_insn After: select_insn: jmp *jumptable(, %rax, 8) ... ALU64_ADD_X: ... jmp *jumptable(, %rax, 8) ALU_ADD_X: ... jmp *jumptable(, %rax, 8) This confuses objtool. It has never seen multiple indirect jump sites which use the same jump table. For GCC switch tables, the only way of detecting the size of a table is by continuing to scan for more tables. The size of the previous table can only be determined after another switch table is found, or when the scan reaches the end of the function. That logic was reused for C jump tables, and was based on the assumption that each jump table only has a single jump site. The above optimization breaks that assumption. 2) Clang issue: drivers/usb/misc/sisusbvga/sisusb.o: warning: objtool: sisusb_write_mem_bulk()+0x588: can't find switch jump table With clang 9, code can be generated where a function contains two indirect jump instructions which use the same switch table. The fix is the same for both issues: split the jump table parsing into two passes. In the first pass, locate the heads of all switch tables for the function and mark their locations. In the second pass, parse the switch tables and add them. Fixes: e55a73251da3 ("bpf: Fix ORC unwinding in non-JIT BPF code") Reported-by: Randy Dunlap <rdunlap@infradead.org> Reported-by: Arnd Bergmann <arnd@arndb.de> Signed-off-by: Jann Horn <jannh@google.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com> Signed-off-by: Thomas Gleixner <tglx@linutronix.de> Tested-by: Nick Desaulniers <ndesaulniers@google.com> Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org> Link: https://lkml.kernel.org/r/e995befaada9d4d8b2cf788ff3f566ba900d2b4d.1563413318.git.jpoimboe@redhat.com Co-developed-by: Josh Poimboeuf <jpoimboe@redhat.com>
2019-07-18 09:36:54 +08:00
}
static int add_func_jump_tables(struct objtool_file *file,
struct symbol *func)
{
struct instruction *insn;
int ret;
func_for_each_insn(file, func, insn) {
objtool: Support repeated uses of the same C jump table This fixes objtool for both a GCC issue and a Clang issue: 1) GCC issue: kernel/bpf/core.o: warning: objtool: ___bpf_prog_run()+0x8d5: sibling call from callable instruction with modified stack frame With CONFIG_RETPOLINE=n, GCC is doing the following optimization in ___bpf_prog_run(). Before: select_insn: jmp *jumptable(,%rax,8) ... ALU64_ADD_X: ... jmp select_insn ALU_ADD_X: ... jmp select_insn After: select_insn: jmp *jumptable(, %rax, 8) ... ALU64_ADD_X: ... jmp *jumptable(, %rax, 8) ALU_ADD_X: ... jmp *jumptable(, %rax, 8) This confuses objtool. It has never seen multiple indirect jump sites which use the same jump table. For GCC switch tables, the only way of detecting the size of a table is by continuing to scan for more tables. The size of the previous table can only be determined after another switch table is found, or when the scan reaches the end of the function. That logic was reused for C jump tables, and was based on the assumption that each jump table only has a single jump site. The above optimization breaks that assumption. 2) Clang issue: drivers/usb/misc/sisusbvga/sisusb.o: warning: objtool: sisusb_write_mem_bulk()+0x588: can't find switch jump table With clang 9, code can be generated where a function contains two indirect jump instructions which use the same switch table. The fix is the same for both issues: split the jump table parsing into two passes. In the first pass, locate the heads of all switch tables for the function and mark their locations. In the second pass, parse the switch tables and add them. Fixes: e55a73251da3 ("bpf: Fix ORC unwinding in non-JIT BPF code") Reported-by: Randy Dunlap <rdunlap@infradead.org> Reported-by: Arnd Bergmann <arnd@arndb.de> Signed-off-by: Jann Horn <jannh@google.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com> Signed-off-by: Thomas Gleixner <tglx@linutronix.de> Tested-by: Nick Desaulniers <ndesaulniers@google.com> Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org> Link: https://lkml.kernel.org/r/e995befaada9d4d8b2cf788ff3f566ba900d2b4d.1563413318.git.jpoimboe@redhat.com Co-developed-by: Josh Poimboeuf <jpoimboe@redhat.com>
2019-07-18 09:36:54 +08:00
if (!insn->jump_table)
continue;
objtool: Support repeated uses of the same C jump table This fixes objtool for both a GCC issue and a Clang issue: 1) GCC issue: kernel/bpf/core.o: warning: objtool: ___bpf_prog_run()+0x8d5: sibling call from callable instruction with modified stack frame With CONFIG_RETPOLINE=n, GCC is doing the following optimization in ___bpf_prog_run(). Before: select_insn: jmp *jumptable(,%rax,8) ... ALU64_ADD_X: ... jmp select_insn ALU_ADD_X: ... jmp select_insn After: select_insn: jmp *jumptable(, %rax, 8) ... ALU64_ADD_X: ... jmp *jumptable(, %rax, 8) ALU_ADD_X: ... jmp *jumptable(, %rax, 8) This confuses objtool. It has never seen multiple indirect jump sites which use the same jump table. For GCC switch tables, the only way of detecting the size of a table is by continuing to scan for more tables. The size of the previous table can only be determined after another switch table is found, or when the scan reaches the end of the function. That logic was reused for C jump tables, and was based on the assumption that each jump table only has a single jump site. The above optimization breaks that assumption. 2) Clang issue: drivers/usb/misc/sisusbvga/sisusb.o: warning: objtool: sisusb_write_mem_bulk()+0x588: can't find switch jump table With clang 9, code can be generated where a function contains two indirect jump instructions which use the same switch table. The fix is the same for both issues: split the jump table parsing into two passes. In the first pass, locate the heads of all switch tables for the function and mark their locations. In the second pass, parse the switch tables and add them. Fixes: e55a73251da3 ("bpf: Fix ORC unwinding in non-JIT BPF code") Reported-by: Randy Dunlap <rdunlap@infradead.org> Reported-by: Arnd Bergmann <arnd@arndb.de> Signed-off-by: Jann Horn <jannh@google.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com> Signed-off-by: Thomas Gleixner <tglx@linutronix.de> Tested-by: Nick Desaulniers <ndesaulniers@google.com> Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org> Link: https://lkml.kernel.org/r/e995befaada9d4d8b2cf788ff3f566ba900d2b4d.1563413318.git.jpoimboe@redhat.com Co-developed-by: Josh Poimboeuf <jpoimboe@redhat.com>
2019-07-18 09:36:54 +08:00
ret = add_jump_table(file, insn, insn->jump_table);
if (ret)
return ret;
}
return 0;
}
/*
* For some switch statements, gcc generates a jump table in the .rodata
* section which contains a list of addresses within the function to jump to.
* This finds these jump tables and adds them to the insn->alts lists.
*/
static int add_jump_table_alts(struct objtool_file *file)
{
struct section *sec;
struct symbol *func;
int ret;
objtool: Support per-function rodata sections Add support for processing switch jump tables in objects with multiple .rodata sections, such as those created by '-ffunction-sections' and '-fdata-sections'. Currently, objtool always looks in .rodata for jump table information, which results in many "sibling call from callable instruction with modified stack frame" warnings with objects compiled using those flags. The fix is comprised of three parts: 1. Flagging all .rodata sections when importing ELF information for easier checking later. 2. Keeping a reference to the section each relocation is from in order to get the list_head for the other relocations in that section. 3. Finding jump tables by following relocations to .rodata sections, rather than always referencing a single global .rodata section. The patch has been tested without data sections enabled and no differences in the resulting orc unwind information were seen. Note that as objtool adds terminators to end of each .text section the unwind information generated between a function+data sections build and a normal build aren't directly comparable. Manual inspection suggests that objtool is now generating the correct information, or at least making more of an effort to do so than it did previously. Signed-off-by: Allan Xavier <allan.x.xavier@oracle.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com> Signed-off-by: Thomas Gleixner <tglx@linutronix.de> Link: https://lkml.kernel.org/r/099bdc375195c490dda04db777ee0b95d566ded1.1536325914.git.jpoimboe@redhat.com
2018-09-07 21:12:01 +08:00
if (!file->rodata)
return 0;
for_each_sec(file, sec) {
list_for_each_entry(func, &sec->symbol_list, list) {
if (func->type != STT_FUNC)
continue;
objtool: Support repeated uses of the same C jump table This fixes objtool for both a GCC issue and a Clang issue: 1) GCC issue: kernel/bpf/core.o: warning: objtool: ___bpf_prog_run()+0x8d5: sibling call from callable instruction with modified stack frame With CONFIG_RETPOLINE=n, GCC is doing the following optimization in ___bpf_prog_run(). Before: select_insn: jmp *jumptable(,%rax,8) ... ALU64_ADD_X: ... jmp select_insn ALU_ADD_X: ... jmp select_insn After: select_insn: jmp *jumptable(, %rax, 8) ... ALU64_ADD_X: ... jmp *jumptable(, %rax, 8) ALU_ADD_X: ... jmp *jumptable(, %rax, 8) This confuses objtool. It has never seen multiple indirect jump sites which use the same jump table. For GCC switch tables, the only way of detecting the size of a table is by continuing to scan for more tables. The size of the previous table can only be determined after another switch table is found, or when the scan reaches the end of the function. That logic was reused for C jump tables, and was based on the assumption that each jump table only has a single jump site. The above optimization breaks that assumption. 2) Clang issue: drivers/usb/misc/sisusbvga/sisusb.o: warning: objtool: sisusb_write_mem_bulk()+0x588: can't find switch jump table With clang 9, code can be generated where a function contains two indirect jump instructions which use the same switch table. The fix is the same for both issues: split the jump table parsing into two passes. In the first pass, locate the heads of all switch tables for the function and mark their locations. In the second pass, parse the switch tables and add them. Fixes: e55a73251da3 ("bpf: Fix ORC unwinding in non-JIT BPF code") Reported-by: Randy Dunlap <rdunlap@infradead.org> Reported-by: Arnd Bergmann <arnd@arndb.de> Signed-off-by: Jann Horn <jannh@google.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com> Signed-off-by: Thomas Gleixner <tglx@linutronix.de> Tested-by: Nick Desaulniers <ndesaulniers@google.com> Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org> Link: https://lkml.kernel.org/r/e995befaada9d4d8b2cf788ff3f566ba900d2b4d.1563413318.git.jpoimboe@redhat.com Co-developed-by: Josh Poimboeuf <jpoimboe@redhat.com>
2019-07-18 09:36:54 +08:00
mark_func_jump_tables(file, func);
ret = add_func_jump_tables(file, func);
if (ret)
return ret;
}
}
return 0;
}
static void set_func_state(struct cfi_state *state)
{
state->cfa = initial_func_cfi.cfa;
memcpy(&state->regs, &initial_func_cfi.regs,
CFI_NUM_REGS * sizeof(struct cfi_reg));
state->stack_size = initial_func_cfi.cfa.offset;
}
static int read_unwind_hints(struct objtool_file *file)
{
struct cfi_state cfi = init_cfi;
objtool: Rename rela to reloc Before supporting additional relocation types rename the relevant types and functions from "rela" to "reloc". This work be done with the following regex: sed -e 's/struct rela/struct reloc/g' \ -e 's/\([_\*]\)rela\(s\{0,1\}\)/\1reloc\2/g' \ -e 's/tmprela\(s\{0,1\}\)/tmpreloc\1/g' \ -e 's/relasec/relocsec/g' \ -e 's/rela_list/reloc_list/g' \ -e 's/rela_hash/reloc_hash/g' \ -e 's/add_rela/add_reloc/g' \ -e 's/rela->/reloc->/g' \ -e '/rela[,\.]/{ s/\([^\.>]\)rela\([\.,]\)/\1reloc\2/g ; }' \ -e 's/rela =/reloc =/g' \ -e 's/relas =/relocs =/g' \ -e 's/relas\[/relocs[/g' \ -e 's/relaname =/relocname =/g' \ -e 's/= rela\;/= reloc\;/g' \ -e 's/= relas\;/= relocs\;/g' \ -e 's/= relaname\;/= relocname\;/g' \ -e 's/, rela)/, reloc)/g' \ -e 's/\([ @]\)rela\([ "]\)/\1reloc\2/g' \ -e 's/ rela$/ reloc/g' \ -e 's/, relaname/, relocname/g' \ -e 's/sec->rela/sec->reloc/g' \ -e 's/(\(!\{0,1\}\)rela/(\1reloc/g' \ -i \ arch.h \ arch/x86/decode.c \ check.c \ check.h \ elf.c \ elf.h \ orc_gen.c \ special.c Notable exceptions which complicate the regex include gelf_* library calls and standard/expected section names which still use "rela" because they encode the type of relocation expected. Also, keep "rela" in the struct because it encodes a specific type of relocation we currently expect. It will eventually turn into a member of an anonymous union when a susequent patch adds implicit addend, or "rel", relocation support. Signed-off-by: Matt Helsley <mhelsley@vmware.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
2020-05-30 05:01:13 +08:00
struct section *sec, *relocsec;
struct unwind_hint *hint;
struct instruction *insn;
struct reloc *reloc;
int i;
sec = find_section_by_name(file->elf, ".discard.unwind_hints");
if (!sec)
return 0;
objtool: Rename rela to reloc Before supporting additional relocation types rename the relevant types and functions from "rela" to "reloc". This work be done with the following regex: sed -e 's/struct rela/struct reloc/g' \ -e 's/\([_\*]\)rela\(s\{0,1\}\)/\1reloc\2/g' \ -e 's/tmprela\(s\{0,1\}\)/tmpreloc\1/g' \ -e 's/relasec/relocsec/g' \ -e 's/rela_list/reloc_list/g' \ -e 's/rela_hash/reloc_hash/g' \ -e 's/add_rela/add_reloc/g' \ -e 's/rela->/reloc->/g' \ -e '/rela[,\.]/{ s/\([^\.>]\)rela\([\.,]\)/\1reloc\2/g ; }' \ -e 's/rela =/reloc =/g' \ -e 's/relas =/relocs =/g' \ -e 's/relas\[/relocs[/g' \ -e 's/relaname =/relocname =/g' \ -e 's/= rela\;/= reloc\;/g' \ -e 's/= relas\;/= relocs\;/g' \ -e 's/= relaname\;/= relocname\;/g' \ -e 's/, rela)/, reloc)/g' \ -e 's/\([ @]\)rela\([ "]\)/\1reloc\2/g' \ -e 's/ rela$/ reloc/g' \ -e 's/, relaname/, relocname/g' \ -e 's/sec->rela/sec->reloc/g' \ -e 's/(\(!\{0,1\}\)rela/(\1reloc/g' \ -i \ arch.h \ arch/x86/decode.c \ check.c \ check.h \ elf.c \ elf.h \ orc_gen.c \ special.c Notable exceptions which complicate the regex include gelf_* library calls and standard/expected section names which still use "rela" because they encode the type of relocation expected. Also, keep "rela" in the struct because it encodes a specific type of relocation we currently expect. It will eventually turn into a member of an anonymous union when a susequent patch adds implicit addend, or "rel", relocation support. Signed-off-by: Matt Helsley <mhelsley@vmware.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
2020-05-30 05:01:13 +08:00
relocsec = sec->reloc;
if (!relocsec) {
WARN("missing .rela.discard.unwind_hints section");
return -1;
}
if (sec->sh.sh_size % sizeof(struct unwind_hint)) {
WARN("struct unwind_hint size mismatch");
return -1;
}
file->hints = true;
for (i = 0; i < sec->sh.sh_size / sizeof(struct unwind_hint); i++) {
hint = (struct unwind_hint *)sec->data->d_buf + i;
objtool: Rename rela to reloc Before supporting additional relocation types rename the relevant types and functions from "rela" to "reloc". This work be done with the following regex: sed -e 's/struct rela/struct reloc/g' \ -e 's/\([_\*]\)rela\(s\{0,1\}\)/\1reloc\2/g' \ -e 's/tmprela\(s\{0,1\}\)/tmpreloc\1/g' \ -e 's/relasec/relocsec/g' \ -e 's/rela_list/reloc_list/g' \ -e 's/rela_hash/reloc_hash/g' \ -e 's/add_rela/add_reloc/g' \ -e 's/rela->/reloc->/g' \ -e '/rela[,\.]/{ s/\([^\.>]\)rela\([\.,]\)/\1reloc\2/g ; }' \ -e 's/rela =/reloc =/g' \ -e 's/relas =/relocs =/g' \ -e 's/relas\[/relocs[/g' \ -e 's/relaname =/relocname =/g' \ -e 's/= rela\;/= reloc\;/g' \ -e 's/= relas\;/= relocs\;/g' \ -e 's/= relaname\;/= relocname\;/g' \ -e 's/, rela)/, reloc)/g' \ -e 's/\([ @]\)rela\([ "]\)/\1reloc\2/g' \ -e 's/ rela$/ reloc/g' \ -e 's/, relaname/, relocname/g' \ -e 's/sec->rela/sec->reloc/g' \ -e 's/(\(!\{0,1\}\)rela/(\1reloc/g' \ -i \ arch.h \ arch/x86/decode.c \ check.c \ check.h \ elf.c \ elf.h \ orc_gen.c \ special.c Notable exceptions which complicate the regex include gelf_* library calls and standard/expected section names which still use "rela" because they encode the type of relocation expected. Also, keep "rela" in the struct because it encodes a specific type of relocation we currently expect. It will eventually turn into a member of an anonymous union when a susequent patch adds implicit addend, or "rel", relocation support. Signed-off-by: Matt Helsley <mhelsley@vmware.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
2020-05-30 05:01:13 +08:00
reloc = find_reloc_by_dest(file->elf, sec, i * sizeof(*hint));
if (!reloc) {
WARN("can't find reloc for unwind_hints[%d]", i);
return -1;
}
objtool: Rename rela to reloc Before supporting additional relocation types rename the relevant types and functions from "rela" to "reloc". This work be done with the following regex: sed -e 's/struct rela/struct reloc/g' \ -e 's/\([_\*]\)rela\(s\{0,1\}\)/\1reloc\2/g' \ -e 's/tmprela\(s\{0,1\}\)/tmpreloc\1/g' \ -e 's/relasec/relocsec/g' \ -e 's/rela_list/reloc_list/g' \ -e 's/rela_hash/reloc_hash/g' \ -e 's/add_rela/add_reloc/g' \ -e 's/rela->/reloc->/g' \ -e '/rela[,\.]/{ s/\([^\.>]\)rela\([\.,]\)/\1reloc\2/g ; }' \ -e 's/rela =/reloc =/g' \ -e 's/relas =/relocs =/g' \ -e 's/relas\[/relocs[/g' \ -e 's/relaname =/relocname =/g' \ -e 's/= rela\;/= reloc\;/g' \ -e 's/= relas\;/= relocs\;/g' \ -e 's/= relaname\;/= relocname\;/g' \ -e 's/, rela)/, reloc)/g' \ -e 's/\([ @]\)rela\([ "]\)/\1reloc\2/g' \ -e 's/ rela$/ reloc/g' \ -e 's/, relaname/, relocname/g' \ -e 's/sec->rela/sec->reloc/g' \ -e 's/(\(!\{0,1\}\)rela/(\1reloc/g' \ -i \ arch.h \ arch/x86/decode.c \ check.c \ check.h \ elf.c \ elf.h \ orc_gen.c \ special.c Notable exceptions which complicate the regex include gelf_* library calls and standard/expected section names which still use "rela" because they encode the type of relocation expected. Also, keep "rela" in the struct because it encodes a specific type of relocation we currently expect. It will eventually turn into a member of an anonymous union when a susequent patch adds implicit addend, or "rel", relocation support. Signed-off-by: Matt Helsley <mhelsley@vmware.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
2020-05-30 05:01:13 +08:00
insn = find_insn(file, reloc->sym->sec, reloc->addend);
if (!insn) {
WARN("can't find insn for unwind_hints[%d]", i);
return -1;
}
insn->hint = true;
if (opts.ibt && hint->type == UNWIND_HINT_TYPE_REGS_PARTIAL) {
struct symbol *sym = find_symbol_by_offset(insn->sec, insn->offset);
if (sym && sym->bind == STB_GLOBAL &&
insn->type != INSN_ENDBR && !insn->noendbr) {
WARN_FUNC("UNWIND_HINT_IRET_REGS without ENDBR",
insn->sec, insn->offset);
}
}
if (hint->type == UNWIND_HINT_TYPE_FUNC) {
insn->cfi = &func_cfi;
continue;
}
if (insn->cfi)
cfi = *(insn->cfi);
if (arch_decode_hint_reg(hint->sp_reg, &cfi.cfa.base)) {
WARN_FUNC("unsupported unwind_hint sp base reg %d",
insn->sec, insn->offset, hint->sp_reg);
return -1;
}
cfi.cfa.offset = bswap_if_needed(hint->sp_offset);
cfi.type = hint->type;
cfi.end = hint->end;
insn->cfi = cfi_hash_find_or_add(&cfi);
}
return 0;
}
static int read_noendbr_hints(struct objtool_file *file)
{
struct section *sec;
struct instruction *insn;
struct reloc *reloc;
sec = find_section_by_name(file->elf, ".rela.discard.noendbr");
if (!sec)
return 0;
list_for_each_entry(reloc, &sec->reloc_list, list) {
insn = find_insn(file, reloc->sym->sec, reloc->sym->offset + reloc->addend);
if (!insn) {
WARN("bad .discard.noendbr entry");
return -1;
}
if (insn->type == INSN_ENDBR)
WARN_FUNC("ANNOTATE_NOENDBR on ENDBR", insn->sec, insn->offset);
insn->noendbr = 1;
}
return 0;
}
static int read_retpoline_hints(struct objtool_file *file)
{
struct section *sec;
struct instruction *insn;
objtool: Rename rela to reloc Before supporting additional relocation types rename the relevant types and functions from "rela" to "reloc". This work be done with the following regex: sed -e 's/struct rela/struct reloc/g' \ -e 's/\([_\*]\)rela\(s\{0,1\}\)/\1reloc\2/g' \ -e 's/tmprela\(s\{0,1\}\)/tmpreloc\1/g' \ -e 's/relasec/relocsec/g' \ -e 's/rela_list/reloc_list/g' \ -e 's/rela_hash/reloc_hash/g' \ -e 's/add_rela/add_reloc/g' \ -e 's/rela->/reloc->/g' \ -e '/rela[,\.]/{ s/\([^\.>]\)rela\([\.,]\)/\1reloc\2/g ; }' \ -e 's/rela =/reloc =/g' \ -e 's/relas =/relocs =/g' \ -e 's/relas\[/relocs[/g' \ -e 's/relaname =/relocname =/g' \ -e 's/= rela\;/= reloc\;/g' \ -e 's/= relas\;/= relocs\;/g' \ -e 's/= relaname\;/= relocname\;/g' \ -e 's/, rela)/, reloc)/g' \ -e 's/\([ @]\)rela\([ "]\)/\1reloc\2/g' \ -e 's/ rela$/ reloc/g' \ -e 's/, relaname/, relocname/g' \ -e 's/sec->rela/sec->reloc/g' \ -e 's/(\(!\{0,1\}\)rela/(\1reloc/g' \ -i \ arch.h \ arch/x86/decode.c \ check.c \ check.h \ elf.c \ elf.h \ orc_gen.c \ special.c Notable exceptions which complicate the regex include gelf_* library calls and standard/expected section names which still use "rela" because they encode the type of relocation expected. Also, keep "rela" in the struct because it encodes a specific type of relocation we currently expect. It will eventually turn into a member of an anonymous union when a susequent patch adds implicit addend, or "rel", relocation support. Signed-off-by: Matt Helsley <mhelsley@vmware.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
2020-05-30 05:01:13 +08:00
struct reloc *reloc;
sec = find_section_by_name(file->elf, ".rela.discard.retpoline_safe");
if (!sec)
return 0;
objtool: Rename rela to reloc Before supporting additional relocation types rename the relevant types and functions from "rela" to "reloc". This work be done with the following regex: sed -e 's/struct rela/struct reloc/g' \ -e 's/\([_\*]\)rela\(s\{0,1\}\)/\1reloc\2/g' \ -e 's/tmprela\(s\{0,1\}\)/tmpreloc\1/g' \ -e 's/relasec/relocsec/g' \ -e 's/rela_list/reloc_list/g' \ -e 's/rela_hash/reloc_hash/g' \ -e 's/add_rela/add_reloc/g' \ -e 's/rela->/reloc->/g' \ -e '/rela[,\.]/{ s/\([^\.>]\)rela\([\.,]\)/\1reloc\2/g ; }' \ -e 's/rela =/reloc =/g' \ -e 's/relas =/relocs =/g' \ -e 's/relas\[/relocs[/g' \ -e 's/relaname =/relocname =/g' \ -e 's/= rela\;/= reloc\;/g' \ -e 's/= relas\;/= relocs\;/g' \ -e 's/= relaname\;/= relocname\;/g' \ -e 's/, rela)/, reloc)/g' \ -e 's/\([ @]\)rela\([ "]\)/\1reloc\2/g' \ -e 's/ rela$/ reloc/g' \ -e 's/, relaname/, relocname/g' \ -e 's/sec->rela/sec->reloc/g' \ -e 's/(\(!\{0,1\}\)rela/(\1reloc/g' \ -i \ arch.h \ arch/x86/decode.c \ check.c \ check.h \ elf.c \ elf.h \ orc_gen.c \ special.c Notable exceptions which complicate the regex include gelf_* library calls and standard/expected section names which still use "rela" because they encode the type of relocation expected. Also, keep "rela" in the struct because it encodes a specific type of relocation we currently expect. It will eventually turn into a member of an anonymous union when a susequent patch adds implicit addend, or "rel", relocation support. Signed-off-by: Matt Helsley <mhelsley@vmware.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
2020-05-30 05:01:13 +08:00
list_for_each_entry(reloc, &sec->reloc_list, list) {
if (reloc->sym->type != STT_SECTION) {
WARN("unexpected relocation symbol type in %s", sec->name);
return -1;
}
objtool: Rename rela to reloc Before supporting additional relocation types rename the relevant types and functions from "rela" to "reloc". This work be done with the following regex: sed -e 's/struct rela/struct reloc/g' \ -e 's/\([_\*]\)rela\(s\{0,1\}\)/\1reloc\2/g' \ -e 's/tmprela\(s\{0,1\}\)/tmpreloc\1/g' \ -e 's/relasec/relocsec/g' \ -e 's/rela_list/reloc_list/g' \ -e 's/rela_hash/reloc_hash/g' \ -e 's/add_rela/add_reloc/g' \ -e 's/rela->/reloc->/g' \ -e '/rela[,\.]/{ s/\([^\.>]\)rela\([\.,]\)/\1reloc\2/g ; }' \ -e 's/rela =/reloc =/g' \ -e 's/relas =/relocs =/g' \ -e 's/relas\[/relocs[/g' \ -e 's/relaname =/relocname =/g' \ -e 's/= rela\;/= reloc\;/g' \ -e 's/= relas\;/= relocs\;/g' \ -e 's/= relaname\;/= relocname\;/g' \ -e 's/, rela)/, reloc)/g' \ -e 's/\([ @]\)rela\([ "]\)/\1reloc\2/g' \ -e 's/ rela$/ reloc/g' \ -e 's/, relaname/, relocname/g' \ -e 's/sec->rela/sec->reloc/g' \ -e 's/(\(!\{0,1\}\)rela/(\1reloc/g' \ -i \ arch.h \ arch/x86/decode.c \ check.c \ check.h \ elf.c \ elf.h \ orc_gen.c \ special.c Notable exceptions which complicate the regex include gelf_* library calls and standard/expected section names which still use "rela" because they encode the type of relocation expected. Also, keep "rela" in the struct because it encodes a specific type of relocation we currently expect. It will eventually turn into a member of an anonymous union when a susequent patch adds implicit addend, or "rel", relocation support. Signed-off-by: Matt Helsley <mhelsley@vmware.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
2020-05-30 05:01:13 +08:00
insn = find_insn(file, reloc->sym->sec, reloc->addend);
if (!insn) {
WARN("bad .discard.retpoline_safe entry");
return -1;
}
if (insn->type != INSN_JUMP_DYNAMIC &&
insn->type != INSN_CALL_DYNAMIC) {
WARN_FUNC("retpoline_safe hint not an indirect jump/call",
insn->sec, insn->offset);
return -1;
}
insn->retpoline_safe = true;
}
return 0;
}
static int read_instr_hints(struct objtool_file *file)
{
struct section *sec;
struct instruction *insn;
objtool: Rename rela to reloc Before supporting additional relocation types rename the relevant types and functions from "rela" to "reloc". This work be done with the following regex: sed -e 's/struct rela/struct reloc/g' \ -e 's/\([_\*]\)rela\(s\{0,1\}\)/\1reloc\2/g' \ -e 's/tmprela\(s\{0,1\}\)/tmpreloc\1/g' \ -e 's/relasec/relocsec/g' \ -e 's/rela_list/reloc_list/g' \ -e 's/rela_hash/reloc_hash/g' \ -e 's/add_rela/add_reloc/g' \ -e 's/rela->/reloc->/g' \ -e '/rela[,\.]/{ s/\([^\.>]\)rela\([\.,]\)/\1reloc\2/g ; }' \ -e 's/rela =/reloc =/g' \ -e 's/relas =/relocs =/g' \ -e 's/relas\[/relocs[/g' \ -e 's/relaname =/relocname =/g' \ -e 's/= rela\;/= reloc\;/g' \ -e 's/= relas\;/= relocs\;/g' \ -e 's/= relaname\;/= relocname\;/g' \ -e 's/, rela)/, reloc)/g' \ -e 's/\([ @]\)rela\([ "]\)/\1reloc\2/g' \ -e 's/ rela$/ reloc/g' \ -e 's/, relaname/, relocname/g' \ -e 's/sec->rela/sec->reloc/g' \ -e 's/(\(!\{0,1\}\)rela/(\1reloc/g' \ -i \ arch.h \ arch/x86/decode.c \ check.c \ check.h \ elf.c \ elf.h \ orc_gen.c \ special.c Notable exceptions which complicate the regex include gelf_* library calls and standard/expected section names which still use "rela" because they encode the type of relocation expected. Also, keep "rela" in the struct because it encodes a specific type of relocation we currently expect. It will eventually turn into a member of an anonymous union when a susequent patch adds implicit addend, or "rel", relocation support. Signed-off-by: Matt Helsley <mhelsley@vmware.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
2020-05-30 05:01:13 +08:00
struct reloc *reloc;
sec = find_section_by_name(file->elf, ".rela.discard.instr_end");
if (!sec)
return 0;
objtool: Rename rela to reloc Before supporting additional relocation types rename the relevant types and functions from "rela" to "reloc". This work be done with the following regex: sed -e 's/struct rela/struct reloc/g' \ -e 's/\([_\*]\)rela\(s\{0,1\}\)/\1reloc\2/g' \ -e 's/tmprela\(s\{0,1\}\)/tmpreloc\1/g' \ -e 's/relasec/relocsec/g' \ -e 's/rela_list/reloc_list/g' \ -e 's/rela_hash/reloc_hash/g' \ -e 's/add_rela/add_reloc/g' \ -e 's/rela->/reloc->/g' \ -e '/rela[,\.]/{ s/\([^\.>]\)rela\([\.,]\)/\1reloc\2/g ; }' \ -e 's/rela =/reloc =/g' \ -e 's/relas =/relocs =/g' \ -e 's/relas\[/relocs[/g' \ -e 's/relaname =/relocname =/g' \ -e 's/= rela\;/= reloc\;/g' \ -e 's/= relas\;/= relocs\;/g' \ -e 's/= relaname\;/= relocname\;/g' \ -e 's/, rela)/, reloc)/g' \ -e 's/\([ @]\)rela\([ "]\)/\1reloc\2/g' \ -e 's/ rela$/ reloc/g' \ -e 's/, relaname/, relocname/g' \ -e 's/sec->rela/sec->reloc/g' \ -e 's/(\(!\{0,1\}\)rela/(\1reloc/g' \ -i \ arch.h \ arch/x86/decode.c \ check.c \ check.h \ elf.c \ elf.h \ orc_gen.c \ special.c Notable exceptions which complicate the regex include gelf_* library calls and standard/expected section names which still use "rela" because they encode the type of relocation expected. Also, keep "rela" in the struct because it encodes a specific type of relocation we currently expect. It will eventually turn into a member of an anonymous union when a susequent patch adds implicit addend, or "rel", relocation support. Signed-off-by: Matt Helsley <mhelsley@vmware.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
2020-05-30 05:01:13 +08:00
list_for_each_entry(reloc, &sec->reloc_list, list) {
if (reloc->sym->type != STT_SECTION) {
WARN("unexpected relocation symbol type in %s", sec->name);
return -1;
}
objtool: Rename rela to reloc Before supporting additional relocation types rename the relevant types and functions from "rela" to "reloc". This work be done with the following regex: sed -e 's/struct rela/struct reloc/g' \ -e 's/\([_\*]\)rela\(s\{0,1\}\)/\1reloc\2/g' \ -e 's/tmprela\(s\{0,1\}\)/tmpreloc\1/g' \ -e 's/relasec/relocsec/g' \ -e 's/rela_list/reloc_list/g' \ -e 's/rela_hash/reloc_hash/g' \ -e 's/add_rela/add_reloc/g' \ -e 's/rela->/reloc->/g' \ -e '/rela[,\.]/{ s/\([^\.>]\)rela\([\.,]\)/\1reloc\2/g ; }' \ -e 's/rela =/reloc =/g' \ -e 's/relas =/relocs =/g' \ -e 's/relas\[/relocs[/g' \ -e 's/relaname =/relocname =/g' \ -e 's/= rela\;/= reloc\;/g' \ -e 's/= relas\;/= relocs\;/g' \ -e 's/= relaname\;/= relocname\;/g' \ -e 's/, rela)/, reloc)/g' \ -e 's/\([ @]\)rela\([ "]\)/\1reloc\2/g' \ -e 's/ rela$/ reloc/g' \ -e 's/, relaname/, relocname/g' \ -e 's/sec->rela/sec->reloc/g' \ -e 's/(\(!\{0,1\}\)rela/(\1reloc/g' \ -i \ arch.h \ arch/x86/decode.c \ check.c \ check.h \ elf.c \ elf.h \ orc_gen.c \ special.c Notable exceptions which complicate the regex include gelf_* library calls and standard/expected section names which still use "rela" because they encode the type of relocation expected. Also, keep "rela" in the struct because it encodes a specific type of relocation we currently expect. It will eventually turn into a member of an anonymous union when a susequent patch adds implicit addend, or "rel", relocation support. Signed-off-by: Matt Helsley <mhelsley@vmware.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
2020-05-30 05:01:13 +08:00
insn = find_insn(file, reloc->sym->sec, reloc->addend);
if (!insn) {
WARN("bad .discard.instr_end entry");
return -1;
}
insn->instr--;
}
sec = find_section_by_name(file->elf, ".rela.discard.instr_begin");
if (!sec)
return 0;
objtool: Rename rela to reloc Before supporting additional relocation types rename the relevant types and functions from "rela" to "reloc". This work be done with the following regex: sed -e 's/struct rela/struct reloc/g' \ -e 's/\([_\*]\)rela\(s\{0,1\}\)/\1reloc\2/g' \ -e 's/tmprela\(s\{0,1\}\)/tmpreloc\1/g' \ -e 's/relasec/relocsec/g' \ -e 's/rela_list/reloc_list/g' \ -e 's/rela_hash/reloc_hash/g' \ -e 's/add_rela/add_reloc/g' \ -e 's/rela->/reloc->/g' \ -e '/rela[,\.]/{ s/\([^\.>]\)rela\([\.,]\)/\1reloc\2/g ; }' \ -e 's/rela =/reloc =/g' \ -e 's/relas =/relocs =/g' \ -e 's/relas\[/relocs[/g' \ -e 's/relaname =/relocname =/g' \ -e 's/= rela\;/= reloc\;/g' \ -e 's/= relas\;/= relocs\;/g' \ -e 's/= relaname\;/= relocname\;/g' \ -e 's/, rela)/, reloc)/g' \ -e 's/\([ @]\)rela\([ "]\)/\1reloc\2/g' \ -e 's/ rela$/ reloc/g' \ -e 's/, relaname/, relocname/g' \ -e 's/sec->rela/sec->reloc/g' \ -e 's/(\(!\{0,1\}\)rela/(\1reloc/g' \ -i \ arch.h \ arch/x86/decode.c \ check.c \ check.h \ elf.c \ elf.h \ orc_gen.c \ special.c Notable exceptions which complicate the regex include gelf_* library calls and standard/expected section names which still use "rela" because they encode the type of relocation expected. Also, keep "rela" in the struct because it encodes a specific type of relocation we currently expect. It will eventually turn into a member of an anonymous union when a susequent patch adds implicit addend, or "rel", relocation support. Signed-off-by: Matt Helsley <mhelsley@vmware.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
2020-05-30 05:01:13 +08:00
list_for_each_entry(reloc, &sec->reloc_list, list) {
if (reloc->sym->type != STT_SECTION) {
WARN("unexpected relocation symbol type in %s", sec->name);
return -1;
}
objtool: Rename rela to reloc Before supporting additional relocation types rename the relevant types and functions from "rela" to "reloc". This work be done with the following regex: sed -e 's/struct rela/struct reloc/g' \ -e 's/\([_\*]\)rela\(s\{0,1\}\)/\1reloc\2/g' \ -e 's/tmprela\(s\{0,1\}\)/tmpreloc\1/g' \ -e 's/relasec/relocsec/g' \ -e 's/rela_list/reloc_list/g' \ -e 's/rela_hash/reloc_hash/g' \ -e 's/add_rela/add_reloc/g' \ -e 's/rela->/reloc->/g' \ -e '/rela[,\.]/{ s/\([^\.>]\)rela\([\.,]\)/\1reloc\2/g ; }' \ -e 's/rela =/reloc =/g' \ -e 's/relas =/relocs =/g' \ -e 's/relas\[/relocs[/g' \ -e 's/relaname =/relocname =/g' \ -e 's/= rela\;/= reloc\;/g' \ -e 's/= relas\;/= relocs\;/g' \ -e 's/= relaname\;/= relocname\;/g' \ -e 's/, rela)/, reloc)/g' \ -e 's/\([ @]\)rela\([ "]\)/\1reloc\2/g' \ -e 's/ rela$/ reloc/g' \ -e 's/, relaname/, relocname/g' \ -e 's/sec->rela/sec->reloc/g' \ -e 's/(\(!\{0,1\}\)rela/(\1reloc/g' \ -i \ arch.h \ arch/x86/decode.c \ check.c \ check.h \ elf.c \ elf.h \ orc_gen.c \ special.c Notable exceptions which complicate the regex include gelf_* library calls and standard/expected section names which still use "rela" because they encode the type of relocation expected. Also, keep "rela" in the struct because it encodes a specific type of relocation we currently expect. It will eventually turn into a member of an anonymous union when a susequent patch adds implicit addend, or "rel", relocation support. Signed-off-by: Matt Helsley <mhelsley@vmware.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
2020-05-30 05:01:13 +08:00
insn = find_insn(file, reloc->sym->sec, reloc->addend);
if (!insn) {
WARN("bad .discard.instr_begin entry");
return -1;
}
insn->instr++;
}
return 0;
}
static int read_intra_function_calls(struct objtool_file *file)
{
struct instruction *insn;
struct section *sec;
objtool: Rename rela to reloc Before supporting additional relocation types rename the relevant types and functions from "rela" to "reloc". This work be done with the following regex: sed -e 's/struct rela/struct reloc/g' \ -e 's/\([_\*]\)rela\(s\{0,1\}\)/\1reloc\2/g' \ -e 's/tmprela\(s\{0,1\}\)/tmpreloc\1/g' \ -e 's/relasec/relocsec/g' \ -e 's/rela_list/reloc_list/g' \ -e 's/rela_hash/reloc_hash/g' \ -e 's/add_rela/add_reloc/g' \ -e 's/rela->/reloc->/g' \ -e '/rela[,\.]/{ s/\([^\.>]\)rela\([\.,]\)/\1reloc\2/g ; }' \ -e 's/rela =/reloc =/g' \ -e 's/relas =/relocs =/g' \ -e 's/relas\[/relocs[/g' \ -e 's/relaname =/relocname =/g' \ -e 's/= rela\;/= reloc\;/g' \ -e 's/= relas\;/= relocs\;/g' \ -e 's/= relaname\;/= relocname\;/g' \ -e 's/, rela)/, reloc)/g' \ -e 's/\([ @]\)rela\([ "]\)/\1reloc\2/g' \ -e 's/ rela$/ reloc/g' \ -e 's/, relaname/, relocname/g' \ -e 's/sec->rela/sec->reloc/g' \ -e 's/(\(!\{0,1\}\)rela/(\1reloc/g' \ -i \ arch.h \ arch/x86/decode.c \ check.c \ check.h \ elf.c \ elf.h \ orc_gen.c \ special.c Notable exceptions which complicate the regex include gelf_* library calls and standard/expected section names which still use "rela" because they encode the type of relocation expected. Also, keep "rela" in the struct because it encodes a specific type of relocation we currently expect. It will eventually turn into a member of an anonymous union when a susequent patch adds implicit addend, or "rel", relocation support. Signed-off-by: Matt Helsley <mhelsley@vmware.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
2020-05-30 05:01:13 +08:00
struct reloc *reloc;
sec = find_section_by_name(file->elf, ".rela.discard.intra_function_calls");
if (!sec)
return 0;
objtool: Rename rela to reloc Before supporting additional relocation types rename the relevant types and functions from "rela" to "reloc". This work be done with the following regex: sed -e 's/struct rela/struct reloc/g' \ -e 's/\([_\*]\)rela\(s\{0,1\}\)/\1reloc\2/g' \ -e 's/tmprela\(s\{0,1\}\)/tmpreloc\1/g' \ -e 's/relasec/relocsec/g' \ -e 's/rela_list/reloc_list/g' \ -e 's/rela_hash/reloc_hash/g' \ -e 's/add_rela/add_reloc/g' \ -e 's/rela->/reloc->/g' \ -e '/rela[,\.]/{ s/\([^\.>]\)rela\([\.,]\)/\1reloc\2/g ; }' \ -e 's/rela =/reloc =/g' \ -e 's/relas =/relocs =/g' \ -e 's/relas\[/relocs[/g' \ -e 's/relaname =/relocname =/g' \ -e 's/= rela\;/= reloc\;/g' \ -e 's/= relas\;/= relocs\;/g' \ -e 's/= relaname\;/= relocname\;/g' \ -e 's/, rela)/, reloc)/g' \ -e 's/\([ @]\)rela\([ "]\)/\1reloc\2/g' \ -e 's/ rela$/ reloc/g' \ -e 's/, relaname/, relocname/g' \ -e 's/sec->rela/sec->reloc/g' \ -e 's/(\(!\{0,1\}\)rela/(\1reloc/g' \ -i \ arch.h \ arch/x86/decode.c \ check.c \ check.h \ elf.c \ elf.h \ orc_gen.c \ special.c Notable exceptions which complicate the regex include gelf_* library calls and standard/expected section names which still use "rela" because they encode the type of relocation expected. Also, keep "rela" in the struct because it encodes a specific type of relocation we currently expect. It will eventually turn into a member of an anonymous union when a susequent patch adds implicit addend, or "rel", relocation support. Signed-off-by: Matt Helsley <mhelsley@vmware.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
2020-05-30 05:01:13 +08:00
list_for_each_entry(reloc, &sec->reloc_list, list) {
unsigned long dest_off;
objtool: Rename rela to reloc Before supporting additional relocation types rename the relevant types and functions from "rela" to "reloc". This work be done with the following regex: sed -e 's/struct rela/struct reloc/g' \ -e 's/\([_\*]\)rela\(s\{0,1\}\)/\1reloc\2/g' \ -e 's/tmprela\(s\{0,1\}\)/tmpreloc\1/g' \ -e 's/relasec/relocsec/g' \ -e 's/rela_list/reloc_list/g' \ -e 's/rela_hash/reloc_hash/g' \ -e 's/add_rela/add_reloc/g' \ -e 's/rela->/reloc->/g' \ -e '/rela[,\.]/{ s/\([^\.>]\)rela\([\.,]\)/\1reloc\2/g ; }' \ -e 's/rela =/reloc =/g' \ -e 's/relas =/relocs =/g' \ -e 's/relas\[/relocs[/g' \ -e 's/relaname =/relocname =/g' \ -e 's/= rela\;/= reloc\;/g' \ -e 's/= relas\;/= relocs\;/g' \ -e 's/= relaname\;/= relocname\;/g' \ -e 's/, rela)/, reloc)/g' \ -e 's/\([ @]\)rela\([ "]\)/\1reloc\2/g' \ -e 's/ rela$/ reloc/g' \ -e 's/, relaname/, relocname/g' \ -e 's/sec->rela/sec->reloc/g' \ -e 's/(\(!\{0,1\}\)rela/(\1reloc/g' \ -i \ arch.h \ arch/x86/decode.c \ check.c \ check.h \ elf.c \ elf.h \ orc_gen.c \ special.c Notable exceptions which complicate the regex include gelf_* library calls and standard/expected section names which still use "rela" because they encode the type of relocation expected. Also, keep "rela" in the struct because it encodes a specific type of relocation we currently expect. It will eventually turn into a member of an anonymous union when a susequent patch adds implicit addend, or "rel", relocation support. Signed-off-by: Matt Helsley <mhelsley@vmware.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
2020-05-30 05:01:13 +08:00
if (reloc->sym->type != STT_SECTION) {
WARN("unexpected relocation symbol type in %s",
sec->name);
return -1;
}
objtool: Rename rela to reloc Before supporting additional relocation types rename the relevant types and functions from "rela" to "reloc". This work be done with the following regex: sed -e 's/struct rela/struct reloc/g' \ -e 's/\([_\*]\)rela\(s\{0,1\}\)/\1reloc\2/g' \ -e 's/tmprela\(s\{0,1\}\)/tmpreloc\1/g' \ -e 's/relasec/relocsec/g' \ -e 's/rela_list/reloc_list/g' \ -e 's/rela_hash/reloc_hash/g' \ -e 's/add_rela/add_reloc/g' \ -e 's/rela->/reloc->/g' \ -e '/rela[,\.]/{ s/\([^\.>]\)rela\([\.,]\)/\1reloc\2/g ; }' \ -e 's/rela =/reloc =/g' \ -e 's/relas =/relocs =/g' \ -e 's/relas\[/relocs[/g' \ -e 's/relaname =/relocname =/g' \ -e 's/= rela\;/= reloc\;/g' \ -e 's/= relas\;/= relocs\;/g' \ -e 's/= relaname\;/= relocname\;/g' \ -e 's/, rela)/, reloc)/g' \ -e 's/\([ @]\)rela\([ "]\)/\1reloc\2/g' \ -e 's/ rela$/ reloc/g' \ -e 's/, relaname/, relocname/g' \ -e 's/sec->rela/sec->reloc/g' \ -e 's/(\(!\{0,1\}\)rela/(\1reloc/g' \ -i \ arch.h \ arch/x86/decode.c \ check.c \ check.h \ elf.c \ elf.h \ orc_gen.c \ special.c Notable exceptions which complicate the regex include gelf_* library calls and standard/expected section names which still use "rela" because they encode the type of relocation expected. Also, keep "rela" in the struct because it encodes a specific type of relocation we currently expect. It will eventually turn into a member of an anonymous union when a susequent patch adds implicit addend, or "rel", relocation support. Signed-off-by: Matt Helsley <mhelsley@vmware.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
2020-05-30 05:01:13 +08:00
insn = find_insn(file, reloc->sym->sec, reloc->addend);
if (!insn) {
WARN("bad .discard.intra_function_call entry");
return -1;
}
if (insn->type != INSN_CALL) {
WARN_FUNC("intra_function_call not a direct call",
insn->sec, insn->offset);
return -1;
}
/*
* Treat intra-function CALLs as JMPs, but with a stack_op.
* See add_call_destinations(), which strips stack_ops from
* normal CALLs.
*/
insn->type = INSN_JUMP_UNCONDITIONAL;
dest_off = insn->offset + insn->len + insn->immediate;
insn->jump_dest = find_insn(file, insn->sec, dest_off);
if (!insn->jump_dest) {
WARN_FUNC("can't find call dest at %s+0x%lx",
insn->sec, insn->offset,
insn->sec->name, dest_off);
return -1;
}
}
return 0;
}
/*
* Return true if name matches an instrumentation function, where calls to that
* function from noinstr code can safely be removed, but compilers won't do so.
*/
static bool is_profiling_func(const char *name)
{
/*
* Many compilers cannot disable KCOV with a function attribute.
*/
if (!strncmp(name, "__sanitizer_cov_", 16))
return true;
/*
* Some compilers currently do not remove __tsan_func_entry/exit nor
* __tsan_atomic_signal_fence (used for barrier instrumentation) with
* the __no_sanitize_thread attribute, remove them. Once the kernel's
* minimum Clang version is 14.0, this can be removed.
*/
if (!strncmp(name, "__tsan_func_", 12) ||
!strcmp(name, "__tsan_atomic_signal_fence"))
return true;
return false;
}
static int classify_symbols(struct objtool_file *file)
{
struct section *sec;
struct symbol *func;
for_each_sec(file, sec) {
list_for_each_entry(func, &sec->symbol_list, list) {
if (func->bind != STB_GLOBAL)
continue;
if (!strncmp(func->name, STATIC_CALL_TRAMP_PREFIX_STR,
strlen(STATIC_CALL_TRAMP_PREFIX_STR)))
func->static_call_tramp = true;
if (arch_is_retpoline(func))
func->retpoline_thunk = true;
if (arch_is_rethunk(func))
func->return_thunk = true;
if (!strcmp(func->name, "__fentry__"))
func->fentry = true;
if (is_profiling_func(func->name))
func->profiling_func = true;
}
}
return 0;
}
objtool: Support per-function rodata sections Add support for processing switch jump tables in objects with multiple .rodata sections, such as those created by '-ffunction-sections' and '-fdata-sections'. Currently, objtool always looks in .rodata for jump table information, which results in many "sibling call from callable instruction with modified stack frame" warnings with objects compiled using those flags. The fix is comprised of three parts: 1. Flagging all .rodata sections when importing ELF information for easier checking later. 2. Keeping a reference to the section each relocation is from in order to get the list_head for the other relocations in that section. 3. Finding jump tables by following relocations to .rodata sections, rather than always referencing a single global .rodata section. The patch has been tested without data sections enabled and no differences in the resulting orc unwind information were seen. Note that as objtool adds terminators to end of each .text section the unwind information generated between a function+data sections build and a normal build aren't directly comparable. Manual inspection suggests that objtool is now generating the correct information, or at least making more of an effort to do so than it did previously. Signed-off-by: Allan Xavier <allan.x.xavier@oracle.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com> Signed-off-by: Thomas Gleixner <tglx@linutronix.de> Link: https://lkml.kernel.org/r/099bdc375195c490dda04db777ee0b95d566ded1.1536325914.git.jpoimboe@redhat.com
2018-09-07 21:12:01 +08:00
static void mark_rodata(struct objtool_file *file)
{
struct section *sec;
bool found = false;
/*
* Search for the following rodata sections, each of which can
* potentially contain jump tables:
*
* - .rodata: can contain GCC switch tables
* - .rodata.<func>: same, if -fdata-sections is being used
* - .rodata..c_jump_table: contains C annotated jump tables
*
* .rodata.str1.* sections are ignored; they don't contain jump tables.
objtool: Support per-function rodata sections Add support for processing switch jump tables in objects with multiple .rodata sections, such as those created by '-ffunction-sections' and '-fdata-sections'. Currently, objtool always looks in .rodata for jump table information, which results in many "sibling call from callable instruction with modified stack frame" warnings with objects compiled using those flags. The fix is comprised of three parts: 1. Flagging all .rodata sections when importing ELF information for easier checking later. 2. Keeping a reference to the section each relocation is from in order to get the list_head for the other relocations in that section. 3. Finding jump tables by following relocations to .rodata sections, rather than always referencing a single global .rodata section. The patch has been tested without data sections enabled and no differences in the resulting orc unwind information were seen. Note that as objtool adds terminators to end of each .text section the unwind information generated between a function+data sections build and a normal build aren't directly comparable. Manual inspection suggests that objtool is now generating the correct information, or at least making more of an effort to do so than it did previously. Signed-off-by: Allan Xavier <allan.x.xavier@oracle.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com> Signed-off-by: Thomas Gleixner <tglx@linutronix.de> Link: https://lkml.kernel.org/r/099bdc375195c490dda04db777ee0b95d566ded1.1536325914.git.jpoimboe@redhat.com
2018-09-07 21:12:01 +08:00
*/
for_each_sec(file, sec) {
if (!strncmp(sec->name, ".rodata", 7) &&
!strstr(sec->name, ".str1.")) {
objtool: Support per-function rodata sections Add support for processing switch jump tables in objects with multiple .rodata sections, such as those created by '-ffunction-sections' and '-fdata-sections'. Currently, objtool always looks in .rodata for jump table information, which results in many "sibling call from callable instruction with modified stack frame" warnings with objects compiled using those flags. The fix is comprised of three parts: 1. Flagging all .rodata sections when importing ELF information for easier checking later. 2. Keeping a reference to the section each relocation is from in order to get the list_head for the other relocations in that section. 3. Finding jump tables by following relocations to .rodata sections, rather than always referencing a single global .rodata section. The patch has been tested without data sections enabled and no differences in the resulting orc unwind information were seen. Note that as objtool adds terminators to end of each .text section the unwind information generated between a function+data sections build and a normal build aren't directly comparable. Manual inspection suggests that objtool is now generating the correct information, or at least making more of an effort to do so than it did previously. Signed-off-by: Allan Xavier <allan.x.xavier@oracle.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com> Signed-off-by: Thomas Gleixner <tglx@linutronix.de> Link: https://lkml.kernel.org/r/099bdc375195c490dda04db777ee0b95d566ded1.1536325914.git.jpoimboe@redhat.com
2018-09-07 21:12:01 +08:00
sec->rodata = true;
found = true;
}
}
file->rodata = found;
}
static int decode_sections(struct objtool_file *file)
{
int ret;
objtool: Support per-function rodata sections Add support for processing switch jump tables in objects with multiple .rodata sections, such as those created by '-ffunction-sections' and '-fdata-sections'. Currently, objtool always looks in .rodata for jump table information, which results in many "sibling call from callable instruction with modified stack frame" warnings with objects compiled using those flags. The fix is comprised of three parts: 1. Flagging all .rodata sections when importing ELF information for easier checking later. 2. Keeping a reference to the section each relocation is from in order to get the list_head for the other relocations in that section. 3. Finding jump tables by following relocations to .rodata sections, rather than always referencing a single global .rodata section. The patch has been tested without data sections enabled and no differences in the resulting orc unwind information were seen. Note that as objtool adds terminators to end of each .text section the unwind information generated between a function+data sections build and a normal build aren't directly comparable. Manual inspection suggests that objtool is now generating the correct information, or at least making more of an effort to do so than it did previously. Signed-off-by: Allan Xavier <allan.x.xavier@oracle.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com> Signed-off-by: Thomas Gleixner <tglx@linutronix.de> Link: https://lkml.kernel.org/r/099bdc375195c490dda04db777ee0b95d566ded1.1536325914.git.jpoimboe@redhat.com
2018-09-07 21:12:01 +08:00
mark_rodata(file);
ret = init_pv_ops(file);
if (ret)
return ret;
ret = decode_instructions(file);
if (ret)
return ret;
add_ignores(file);
add_uaccess_safe(file);
ret = add_ignore_alternatives(file);
if (ret)
return ret;
/*
* Must be before read_unwind_hints() since that needs insn->noendbr.
*/
ret = read_noendbr_hints(file);
if (ret)
return ret;
/*
* Must be before add_{jump_call}_destination.
*/
ret = classify_symbols(file);
if (ret)
return ret;
/*
* Must be before add_jump_destinations(), which depends on 'func'
* being set for alternatives, to enable proper sibling call detection.
*/
ret = add_special_section_alts(file);
if (ret)
return ret;
ret = add_jump_destinations(file);
if (ret)
return ret;
/*
* Must be before add_call_destination(); it changes INSN_CALL to
* INSN_JUMP.
*/
ret = read_intra_function_calls(file);
if (ret)
return ret;
ret = add_call_destinations(file);
if (ret)
return ret;
/*
* Must be after add_call_destinations() such that it can override
* dead_end_function() marks.
*/
ret = add_dead_ends(file);
if (ret)
return ret;
ret = add_jump_table_alts(file);
if (ret)
return ret;
ret = read_unwind_hints(file);
if (ret)
return ret;
ret = read_retpoline_hints(file);
if (ret)
return ret;
ret = read_instr_hints(file);
if (ret)
return ret;
return 0;
}
static bool is_fentry_call(struct instruction *insn)
{
if (insn->type == INSN_CALL &&
insn->call_dest &&
insn->call_dest->fentry)
return true;
return false;
}
static bool has_modified_stack_frame(struct instruction *insn, struct insn_state *state)
{
struct cfi_state *cfi = &state->cfi;
int i;
if (cfi->cfa.base != initial_func_cfi.cfa.base || cfi->drap)
return true;
if (cfi->cfa.offset != initial_func_cfi.cfa.offset)
return true;
if (cfi->stack_size != initial_func_cfi.cfa.offset)
return true;
for (i = 0; i < CFI_NUM_REGS; i++) {
if (cfi->regs[i].base != initial_func_cfi.regs[i].base ||
cfi->regs[i].offset != initial_func_cfi.regs[i].offset)
return true;
}
return false;
}
static bool check_reg_frame_pos(const struct cfi_reg *reg,
int expected_offset)
{
return reg->base == CFI_CFA &&
reg->offset == expected_offset;
}
static bool has_valid_stack_frame(struct insn_state *state)
{
struct cfi_state *cfi = &state->cfi;
if (cfi->cfa.base == CFI_BP &&
check_reg_frame_pos(&cfi->regs[CFI_BP], -cfi->cfa.offset) &&
check_reg_frame_pos(&cfi->regs[CFI_RA], -cfi->cfa.offset + 8))
return true;
if (cfi->drap && cfi->regs[CFI_BP].base == CFI_BP)
return true;
return false;
}
static int update_cfi_state_regs(struct instruction *insn,
struct cfi_state *cfi,
struct stack_op *op)
{
struct cfi_reg *cfa = &cfi->cfa;
if (cfa->base != CFI_SP && cfa->base != CFI_SP_INDIRECT)
return 0;
/* push */
if (op->dest.type == OP_DEST_PUSH || op->dest.type == OP_DEST_PUSHF)
cfa->offset += 8;
/* pop */
if (op->src.type == OP_SRC_POP || op->src.type == OP_SRC_POPF)
cfa->offset -= 8;
/* add immediate to sp */
if (op->dest.type == OP_DEST_REG && op->src.type == OP_SRC_ADD &&
op->dest.reg == CFI_SP && op->src.reg == CFI_SP)
cfa->offset -= op->src.offset;
return 0;
}
static void save_reg(struct cfi_state *cfi, unsigned char reg, int base, int offset)
{
if (arch_callee_saved_reg(reg) &&
cfi->regs[reg].base == CFI_UNDEFINED) {
cfi->regs[reg].base = base;
cfi->regs[reg].offset = offset;
}
}
static void restore_reg(struct cfi_state *cfi, unsigned char reg)
{
cfi->regs[reg].base = initial_func_cfi.regs[reg].base;
cfi->regs[reg].offset = initial_func_cfi.regs[reg].offset;
}
/*
* A note about DRAP stack alignment:
*
* GCC has the concept of a DRAP register, which is used to help keep track of
* the stack pointer when aligning the stack. r10 or r13 is used as the DRAP
* register. The typical DRAP pattern is:
*
* 4c 8d 54 24 08 lea 0x8(%rsp),%r10
* 48 83 e4 c0 and $0xffffffffffffffc0,%rsp
* 41 ff 72 f8 pushq -0x8(%r10)
* 55 push %rbp
* 48 89 e5 mov %rsp,%rbp
* (more pushes)
* 41 52 push %r10
* ...
* 41 5a pop %r10
* (more pops)
* 5d pop %rbp
* 49 8d 62 f8 lea -0x8(%r10),%rsp
* c3 retq
*
* There are some variations in the epilogues, like:
*
* 5b pop %rbx
* 41 5a pop %r10
* 41 5c pop %r12
* 41 5d pop %r13
* 41 5e pop %r14
* c9 leaveq
* 49 8d 62 f8 lea -0x8(%r10),%rsp
* c3 retq
*
* and:
*
* 4c 8b 55 e8 mov -0x18(%rbp),%r10
* 48 8b 5d e0 mov -0x20(%rbp),%rbx
* 4c 8b 65 f0 mov -0x10(%rbp),%r12
* 4c 8b 6d f8 mov -0x8(%rbp),%r13
* c9 leaveq
* 49 8d 62 f8 lea -0x8(%r10),%rsp
* c3 retq
*
* Sometimes r13 is used as the DRAP register, in which case it's saved and
* restored beforehand:
*
* 41 55 push %r13
* 4c 8d 6c 24 10 lea 0x10(%rsp),%r13
* 48 83 e4 f0 and $0xfffffffffffffff0,%rsp
* ...
* 49 8d 65 f0 lea -0x10(%r13),%rsp
* 41 5d pop %r13
* c3 retq
*/
static int update_cfi_state(struct instruction *insn,
struct instruction *next_insn,
struct cfi_state *cfi, struct stack_op *op)
{
struct cfi_reg *cfa = &cfi->cfa;
struct cfi_reg *regs = cfi->regs;
/* stack operations don't make sense with an undefined CFA */
if (cfa->base == CFI_UNDEFINED) {
if (insn->func) {
WARN_FUNC("undefined stack state", insn->sec, insn->offset);
return -1;
}
return 0;
}
if (cfi->type == UNWIND_HINT_TYPE_REGS ||
cfi->type == UNWIND_HINT_TYPE_REGS_PARTIAL)
return update_cfi_state_regs(insn, cfi, op);
switch (op->dest.type) {
case OP_DEST_REG:
switch (op->src.type) {
case OP_SRC_REG:
objtool: Handle another GCC stack pointer adjustment bug The kbuild bot reported the following warning with GCC 4.4 and a randconfig: net/socket.o: warning: objtool: compat_sock_ioctl()+0x1083: stack state mismatch: cfa1=7+160 cfa2=-1+0 This is caused by another GCC non-optimization, where it backs up and restores the stack pointer for no apparent reason: 2f91: 48 89 e0 mov %rsp,%rax 2f94: 4c 89 e7 mov %r12,%rdi 2f97: 4c 89 f6 mov %r14,%rsi 2f9a: ba 20 00 00 00 mov $0x20,%edx 2f9f: 48 89 c4 mov %rax,%rsp This issue would have been happily ignored before the following commit: dd88a0a0c861 ("objtool: Handle GCC stack pointer adjustment bug") But now that objtool is paying attention to such stack pointer writes to/from a register, it needs to understand them properly. In this case that means recognizing that the "mov %rsp, %rax" instruction is potentially a backup of the stack pointer. Reported-by: kbuild test robot <fengguang.wu@intel.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com> Cc: Alexander Potapenko <glider@google.com> Cc: Andrey Ryabinin <aryabinin@virtuozzo.com> Cc: Andy Lutomirski <luto@kernel.org> Cc: Arnd Bergmann <arnd@arndb.de> Cc: Dmitriy Vyukov <dvyukov@google.com> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Matthias Kaehlcke <mka@chromium.org> Cc: Miguel Bernal Marin <miguel.bernal.marin@linux.intel.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Thomas Gleixner <tglx@linutronix.de> Fixes: dd88a0a0c861 ("objtool: Handle GCC stack pointer adjustment bug") Link: http://lkml.kernel.org/r/8c7aa8e9a36fbbb6655d9d8e7cea58958c912da8.1505942196.git.jpoimboe@redhat.com Signed-off-by: Ingo Molnar <mingo@kernel.org>
2017-09-21 05:24:32 +08:00
if (op->src.reg == CFI_SP && op->dest.reg == CFI_BP &&
cfa->base == CFI_SP &&
check_reg_frame_pos(&regs[CFI_BP], -cfa->offset)) {
objtool: Handle another GCC stack pointer adjustment bug The kbuild bot reported the following warning with GCC 4.4 and a randconfig: net/socket.o: warning: objtool: compat_sock_ioctl()+0x1083: stack state mismatch: cfa1=7+160 cfa2=-1+0 This is caused by another GCC non-optimization, where it backs up and restores the stack pointer for no apparent reason: 2f91: 48 89 e0 mov %rsp,%rax 2f94: 4c 89 e7 mov %r12,%rdi 2f97: 4c 89 f6 mov %r14,%rsi 2f9a: ba 20 00 00 00 mov $0x20,%edx 2f9f: 48 89 c4 mov %rax,%rsp This issue would have been happily ignored before the following commit: dd88a0a0c861 ("objtool: Handle GCC stack pointer adjustment bug") But now that objtool is paying attention to such stack pointer writes to/from a register, it needs to understand them properly. In this case that means recognizing that the "mov %rsp, %rax" instruction is potentially a backup of the stack pointer. Reported-by: kbuild test robot <fengguang.wu@intel.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com> Cc: Alexander Potapenko <glider@google.com> Cc: Andrey Ryabinin <aryabinin@virtuozzo.com> Cc: Andy Lutomirski <luto@kernel.org> Cc: Arnd Bergmann <arnd@arndb.de> Cc: Dmitriy Vyukov <dvyukov@google.com> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Matthias Kaehlcke <mka@chromium.org> Cc: Miguel Bernal Marin <miguel.bernal.marin@linux.intel.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Thomas Gleixner <tglx@linutronix.de> Fixes: dd88a0a0c861 ("objtool: Handle GCC stack pointer adjustment bug") Link: http://lkml.kernel.org/r/8c7aa8e9a36fbbb6655d9d8e7cea58958c912da8.1505942196.git.jpoimboe@redhat.com Signed-off-by: Ingo Molnar <mingo@kernel.org>
2017-09-21 05:24:32 +08:00
/* mov %rsp, %rbp */
cfa->base = op->dest.reg;
cfi->bp_scratch = false;
objtool: Handle another GCC stack pointer adjustment bug The kbuild bot reported the following warning with GCC 4.4 and a randconfig: net/socket.o: warning: objtool: compat_sock_ioctl()+0x1083: stack state mismatch: cfa1=7+160 cfa2=-1+0 This is caused by another GCC non-optimization, where it backs up and restores the stack pointer for no apparent reason: 2f91: 48 89 e0 mov %rsp,%rax 2f94: 4c 89 e7 mov %r12,%rdi 2f97: 4c 89 f6 mov %r14,%rsi 2f9a: ba 20 00 00 00 mov $0x20,%edx 2f9f: 48 89 c4 mov %rax,%rsp This issue would have been happily ignored before the following commit: dd88a0a0c861 ("objtool: Handle GCC stack pointer adjustment bug") But now that objtool is paying attention to such stack pointer writes to/from a register, it needs to understand them properly. In this case that means recognizing that the "mov %rsp, %rax" instruction is potentially a backup of the stack pointer. Reported-by: kbuild test robot <fengguang.wu@intel.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com> Cc: Alexander Potapenko <glider@google.com> Cc: Andrey Ryabinin <aryabinin@virtuozzo.com> Cc: Andy Lutomirski <luto@kernel.org> Cc: Arnd Bergmann <arnd@arndb.de> Cc: Dmitriy Vyukov <dvyukov@google.com> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Matthias Kaehlcke <mka@chromium.org> Cc: Miguel Bernal Marin <miguel.bernal.marin@linux.intel.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Thomas Gleixner <tglx@linutronix.de> Fixes: dd88a0a0c861 ("objtool: Handle GCC stack pointer adjustment bug") Link: http://lkml.kernel.org/r/8c7aa8e9a36fbbb6655d9d8e7cea58958c912da8.1505942196.git.jpoimboe@redhat.com Signed-off-by: Ingo Molnar <mingo@kernel.org>
2017-09-21 05:24:32 +08:00
}
objtool: Handle GCC stack pointer adjustment bug Arnd Bergmann reported the following warning with GCC 7.1.1: fs/fs_pin.o: warning: objtool: pin_kill()+0x139: stack state mismatch: cfa1=7+88 cfa2=7+96 And the kbuild robot reported the following warnings with GCC 5.4.1: fs/fs_pin.o: warning: objtool: pin_kill()+0x182: return with modified stack frame fs/quota/dquot.o: warning: objtool: dquot_alloc_inode()+0x140: stack state mismatch: cfa1=7+120 cfa2=7+128 fs/quota/dquot.o: warning: objtool: dquot_free_inode()+0x11a: stack state mismatch: cfa1=7+112 cfa2=7+120 Those warnings are caused by an unusual GCC non-optimization where it uses an intermediate register to adjust the stack pointer. It does: lea 0x8(%rsp), %rcx ... mov %rcx, %rsp Instead of the obvious: add $0x8, %rsp It makes no sense to use an intermediate register, so I opened a GCC bug to track it: https://gcc.gnu.org/bugzilla/show_bug.cgi?id=81813 But it's not exactly a high-priority bug and it looks like we'll be stuck with this issue for a while. So for now we have to track register values when they're loaded with stack pointer offsets. This is kind of a big workaround for a tiny problem, but c'est la vie. I hope to eventually create a GCC plugin to implement a big chunk of objtool's functionality. Hopefully at that point we'll be able to remove of a lot of these GCC-isms from the objtool code. Reported-by: Arnd Bergmann <arnd@arndb.de> Reported-by: kbuild test robot <fengguang.wu@intel.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Thomas Gleixner <tglx@linutronix.de> Link: http://lkml.kernel.org/r/6a41a96884c725e7f05413bb7df40cfe824b2444.1504028945.git.jpoimboe@redhat.com Signed-off-by: Ingo Molnar <mingo@kernel.org>
2017-08-30 01:51:03 +08:00
objtool: Handle another GCC stack pointer adjustment bug The kbuild bot reported the following warning with GCC 4.4 and a randconfig: net/socket.o: warning: objtool: compat_sock_ioctl()+0x1083: stack state mismatch: cfa1=7+160 cfa2=-1+0 This is caused by another GCC non-optimization, where it backs up and restores the stack pointer for no apparent reason: 2f91: 48 89 e0 mov %rsp,%rax 2f94: 4c 89 e7 mov %r12,%rdi 2f97: 4c 89 f6 mov %r14,%rsi 2f9a: ba 20 00 00 00 mov $0x20,%edx 2f9f: 48 89 c4 mov %rax,%rsp This issue would have been happily ignored before the following commit: dd88a0a0c861 ("objtool: Handle GCC stack pointer adjustment bug") But now that objtool is paying attention to such stack pointer writes to/from a register, it needs to understand them properly. In this case that means recognizing that the "mov %rsp, %rax" instruction is potentially a backup of the stack pointer. Reported-by: kbuild test robot <fengguang.wu@intel.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com> Cc: Alexander Potapenko <glider@google.com> Cc: Andrey Ryabinin <aryabinin@virtuozzo.com> Cc: Andy Lutomirski <luto@kernel.org> Cc: Arnd Bergmann <arnd@arndb.de> Cc: Dmitriy Vyukov <dvyukov@google.com> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Matthias Kaehlcke <mka@chromium.org> Cc: Miguel Bernal Marin <miguel.bernal.marin@linux.intel.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Thomas Gleixner <tglx@linutronix.de> Fixes: dd88a0a0c861 ("objtool: Handle GCC stack pointer adjustment bug") Link: http://lkml.kernel.org/r/8c7aa8e9a36fbbb6655d9d8e7cea58958c912da8.1505942196.git.jpoimboe@redhat.com Signed-off-by: Ingo Molnar <mingo@kernel.org>
2017-09-21 05:24:32 +08:00
else if (op->src.reg == CFI_SP &&
op->dest.reg == CFI_BP && cfi->drap) {
objtool: Handle GCC stack pointer adjustment bug Arnd Bergmann reported the following warning with GCC 7.1.1: fs/fs_pin.o: warning: objtool: pin_kill()+0x139: stack state mismatch: cfa1=7+88 cfa2=7+96 And the kbuild robot reported the following warnings with GCC 5.4.1: fs/fs_pin.o: warning: objtool: pin_kill()+0x182: return with modified stack frame fs/quota/dquot.o: warning: objtool: dquot_alloc_inode()+0x140: stack state mismatch: cfa1=7+120 cfa2=7+128 fs/quota/dquot.o: warning: objtool: dquot_free_inode()+0x11a: stack state mismatch: cfa1=7+112 cfa2=7+120 Those warnings are caused by an unusual GCC non-optimization where it uses an intermediate register to adjust the stack pointer. It does: lea 0x8(%rsp), %rcx ... mov %rcx, %rsp Instead of the obvious: add $0x8, %rsp It makes no sense to use an intermediate register, so I opened a GCC bug to track it: https://gcc.gnu.org/bugzilla/show_bug.cgi?id=81813 But it's not exactly a high-priority bug and it looks like we'll be stuck with this issue for a while. So for now we have to track register values when they're loaded with stack pointer offsets. This is kind of a big workaround for a tiny problem, but c'est la vie. I hope to eventually create a GCC plugin to implement a big chunk of objtool's functionality. Hopefully at that point we'll be able to remove of a lot of these GCC-isms from the objtool code. Reported-by: Arnd Bergmann <arnd@arndb.de> Reported-by: kbuild test robot <fengguang.wu@intel.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Thomas Gleixner <tglx@linutronix.de> Link: http://lkml.kernel.org/r/6a41a96884c725e7f05413bb7df40cfe824b2444.1504028945.git.jpoimboe@redhat.com Signed-off-by: Ingo Molnar <mingo@kernel.org>
2017-08-30 01:51:03 +08:00
objtool: Handle another GCC stack pointer adjustment bug The kbuild bot reported the following warning with GCC 4.4 and a randconfig: net/socket.o: warning: objtool: compat_sock_ioctl()+0x1083: stack state mismatch: cfa1=7+160 cfa2=-1+0 This is caused by another GCC non-optimization, where it backs up and restores the stack pointer for no apparent reason: 2f91: 48 89 e0 mov %rsp,%rax 2f94: 4c 89 e7 mov %r12,%rdi 2f97: 4c 89 f6 mov %r14,%rsi 2f9a: ba 20 00 00 00 mov $0x20,%edx 2f9f: 48 89 c4 mov %rax,%rsp This issue would have been happily ignored before the following commit: dd88a0a0c861 ("objtool: Handle GCC stack pointer adjustment bug") But now that objtool is paying attention to such stack pointer writes to/from a register, it needs to understand them properly. In this case that means recognizing that the "mov %rsp, %rax" instruction is potentially a backup of the stack pointer. Reported-by: kbuild test robot <fengguang.wu@intel.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com> Cc: Alexander Potapenko <glider@google.com> Cc: Andrey Ryabinin <aryabinin@virtuozzo.com> Cc: Andy Lutomirski <luto@kernel.org> Cc: Arnd Bergmann <arnd@arndb.de> Cc: Dmitriy Vyukov <dvyukov@google.com> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Matthias Kaehlcke <mka@chromium.org> Cc: Miguel Bernal Marin <miguel.bernal.marin@linux.intel.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Thomas Gleixner <tglx@linutronix.de> Fixes: dd88a0a0c861 ("objtool: Handle GCC stack pointer adjustment bug") Link: http://lkml.kernel.org/r/8c7aa8e9a36fbbb6655d9d8e7cea58958c912da8.1505942196.git.jpoimboe@redhat.com Signed-off-by: Ingo Molnar <mingo@kernel.org>
2017-09-21 05:24:32 +08:00
/* drap: mov %rsp, %rbp */
regs[CFI_BP].base = CFI_BP;
regs[CFI_BP].offset = -cfi->stack_size;
cfi->bp_scratch = false;
objtool: Handle another GCC stack pointer adjustment bug The kbuild bot reported the following warning with GCC 4.4 and a randconfig: net/socket.o: warning: objtool: compat_sock_ioctl()+0x1083: stack state mismatch: cfa1=7+160 cfa2=-1+0 This is caused by another GCC non-optimization, where it backs up and restores the stack pointer for no apparent reason: 2f91: 48 89 e0 mov %rsp,%rax 2f94: 4c 89 e7 mov %r12,%rdi 2f97: 4c 89 f6 mov %r14,%rsi 2f9a: ba 20 00 00 00 mov $0x20,%edx 2f9f: 48 89 c4 mov %rax,%rsp This issue would have been happily ignored before the following commit: dd88a0a0c861 ("objtool: Handle GCC stack pointer adjustment bug") But now that objtool is paying attention to such stack pointer writes to/from a register, it needs to understand them properly. In this case that means recognizing that the "mov %rsp, %rax" instruction is potentially a backup of the stack pointer. Reported-by: kbuild test robot <fengguang.wu@intel.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com> Cc: Alexander Potapenko <glider@google.com> Cc: Andrey Ryabinin <aryabinin@virtuozzo.com> Cc: Andy Lutomirski <luto@kernel.org> Cc: Arnd Bergmann <arnd@arndb.de> Cc: Dmitriy Vyukov <dvyukov@google.com> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Matthias Kaehlcke <mka@chromium.org> Cc: Miguel Bernal Marin <miguel.bernal.marin@linux.intel.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Thomas Gleixner <tglx@linutronix.de> Fixes: dd88a0a0c861 ("objtool: Handle GCC stack pointer adjustment bug") Link: http://lkml.kernel.org/r/8c7aa8e9a36fbbb6655d9d8e7cea58958c912da8.1505942196.git.jpoimboe@redhat.com Signed-off-by: Ingo Molnar <mingo@kernel.org>
2017-09-21 05:24:32 +08:00
}
objtool: Handle GCC stack pointer adjustment bug Arnd Bergmann reported the following warning with GCC 7.1.1: fs/fs_pin.o: warning: objtool: pin_kill()+0x139: stack state mismatch: cfa1=7+88 cfa2=7+96 And the kbuild robot reported the following warnings with GCC 5.4.1: fs/fs_pin.o: warning: objtool: pin_kill()+0x182: return with modified stack frame fs/quota/dquot.o: warning: objtool: dquot_alloc_inode()+0x140: stack state mismatch: cfa1=7+120 cfa2=7+128 fs/quota/dquot.o: warning: objtool: dquot_free_inode()+0x11a: stack state mismatch: cfa1=7+112 cfa2=7+120 Those warnings are caused by an unusual GCC non-optimization where it uses an intermediate register to adjust the stack pointer. It does: lea 0x8(%rsp), %rcx ... mov %rcx, %rsp Instead of the obvious: add $0x8, %rsp It makes no sense to use an intermediate register, so I opened a GCC bug to track it: https://gcc.gnu.org/bugzilla/show_bug.cgi?id=81813 But it's not exactly a high-priority bug and it looks like we'll be stuck with this issue for a while. So for now we have to track register values when they're loaded with stack pointer offsets. This is kind of a big workaround for a tiny problem, but c'est la vie. I hope to eventually create a GCC plugin to implement a big chunk of objtool's functionality. Hopefully at that point we'll be able to remove of a lot of these GCC-isms from the objtool code. Reported-by: Arnd Bergmann <arnd@arndb.de> Reported-by: kbuild test robot <fengguang.wu@intel.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Thomas Gleixner <tglx@linutronix.de> Link: http://lkml.kernel.org/r/6a41a96884c725e7f05413bb7df40cfe824b2444.1504028945.git.jpoimboe@redhat.com Signed-off-by: Ingo Molnar <mingo@kernel.org>
2017-08-30 01:51:03 +08:00
objtool: Handle another GCC stack pointer adjustment bug The kbuild bot reported the following warning with GCC 4.4 and a randconfig: net/socket.o: warning: objtool: compat_sock_ioctl()+0x1083: stack state mismatch: cfa1=7+160 cfa2=-1+0 This is caused by another GCC non-optimization, where it backs up and restores the stack pointer for no apparent reason: 2f91: 48 89 e0 mov %rsp,%rax 2f94: 4c 89 e7 mov %r12,%rdi 2f97: 4c 89 f6 mov %r14,%rsi 2f9a: ba 20 00 00 00 mov $0x20,%edx 2f9f: 48 89 c4 mov %rax,%rsp This issue would have been happily ignored before the following commit: dd88a0a0c861 ("objtool: Handle GCC stack pointer adjustment bug") But now that objtool is paying attention to such stack pointer writes to/from a register, it needs to understand them properly. In this case that means recognizing that the "mov %rsp, %rax" instruction is potentially a backup of the stack pointer. Reported-by: kbuild test robot <fengguang.wu@intel.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com> Cc: Alexander Potapenko <glider@google.com> Cc: Andrey Ryabinin <aryabinin@virtuozzo.com> Cc: Andy Lutomirski <luto@kernel.org> Cc: Arnd Bergmann <arnd@arndb.de> Cc: Dmitriy Vyukov <dvyukov@google.com> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Matthias Kaehlcke <mka@chromium.org> Cc: Miguel Bernal Marin <miguel.bernal.marin@linux.intel.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Thomas Gleixner <tglx@linutronix.de> Fixes: dd88a0a0c861 ("objtool: Handle GCC stack pointer adjustment bug") Link: http://lkml.kernel.org/r/8c7aa8e9a36fbbb6655d9d8e7cea58958c912da8.1505942196.git.jpoimboe@redhat.com Signed-off-by: Ingo Molnar <mingo@kernel.org>
2017-09-21 05:24:32 +08:00
else if (op->src.reg == CFI_SP && cfa->base == CFI_SP) {
/*
* mov %rsp, %reg
*
* This is needed for the rare case where GCC
* does:
*
* mov %rsp, %rax
* ...
* mov %rax, %rsp
*/
cfi->vals[op->dest.reg].base = CFI_CFA;
cfi->vals[op->dest.reg].offset = -cfi->stack_size;
objtool: Handle GCC stack pointer adjustment bug Arnd Bergmann reported the following warning with GCC 7.1.1: fs/fs_pin.o: warning: objtool: pin_kill()+0x139: stack state mismatch: cfa1=7+88 cfa2=7+96 And the kbuild robot reported the following warnings with GCC 5.4.1: fs/fs_pin.o: warning: objtool: pin_kill()+0x182: return with modified stack frame fs/quota/dquot.o: warning: objtool: dquot_alloc_inode()+0x140: stack state mismatch: cfa1=7+120 cfa2=7+128 fs/quota/dquot.o: warning: objtool: dquot_free_inode()+0x11a: stack state mismatch: cfa1=7+112 cfa2=7+120 Those warnings are caused by an unusual GCC non-optimization where it uses an intermediate register to adjust the stack pointer. It does: lea 0x8(%rsp), %rcx ... mov %rcx, %rsp Instead of the obvious: add $0x8, %rsp It makes no sense to use an intermediate register, so I opened a GCC bug to track it: https://gcc.gnu.org/bugzilla/show_bug.cgi?id=81813 But it's not exactly a high-priority bug and it looks like we'll be stuck with this issue for a while. So for now we have to track register values when they're loaded with stack pointer offsets. This is kind of a big workaround for a tiny problem, but c'est la vie. I hope to eventually create a GCC plugin to implement a big chunk of objtool's functionality. Hopefully at that point we'll be able to remove of a lot of these GCC-isms from the objtool code. Reported-by: Arnd Bergmann <arnd@arndb.de> Reported-by: kbuild test robot <fengguang.wu@intel.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Thomas Gleixner <tglx@linutronix.de> Link: http://lkml.kernel.org/r/6a41a96884c725e7f05413bb7df40cfe824b2444.1504028945.git.jpoimboe@redhat.com Signed-off-by: Ingo Molnar <mingo@kernel.org>
2017-08-30 01:51:03 +08:00
}
else if (op->src.reg == CFI_BP && op->dest.reg == CFI_SP &&
(cfa->base == CFI_BP || cfa->base == cfi->drap_reg)) {
/*
* mov %rbp, %rsp
*
* Restore the original stack pointer (Clang).
*/
cfi->stack_size = -cfi->regs[CFI_BP].offset;
}
objtool: Handle GCC stack pointer adjustment bug Arnd Bergmann reported the following warning with GCC 7.1.1: fs/fs_pin.o: warning: objtool: pin_kill()+0x139: stack state mismatch: cfa1=7+88 cfa2=7+96 And the kbuild robot reported the following warnings with GCC 5.4.1: fs/fs_pin.o: warning: objtool: pin_kill()+0x182: return with modified stack frame fs/quota/dquot.o: warning: objtool: dquot_alloc_inode()+0x140: stack state mismatch: cfa1=7+120 cfa2=7+128 fs/quota/dquot.o: warning: objtool: dquot_free_inode()+0x11a: stack state mismatch: cfa1=7+112 cfa2=7+120 Those warnings are caused by an unusual GCC non-optimization where it uses an intermediate register to adjust the stack pointer. It does: lea 0x8(%rsp), %rcx ... mov %rcx, %rsp Instead of the obvious: add $0x8, %rsp It makes no sense to use an intermediate register, so I opened a GCC bug to track it: https://gcc.gnu.org/bugzilla/show_bug.cgi?id=81813 But it's not exactly a high-priority bug and it looks like we'll be stuck with this issue for a while. So for now we have to track register values when they're loaded with stack pointer offsets. This is kind of a big workaround for a tiny problem, but c'est la vie. I hope to eventually create a GCC plugin to implement a big chunk of objtool's functionality. Hopefully at that point we'll be able to remove of a lot of these GCC-isms from the objtool code. Reported-by: Arnd Bergmann <arnd@arndb.de> Reported-by: kbuild test robot <fengguang.wu@intel.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Thomas Gleixner <tglx@linutronix.de> Link: http://lkml.kernel.org/r/6a41a96884c725e7f05413bb7df40cfe824b2444.1504028945.git.jpoimboe@redhat.com Signed-off-by: Ingo Molnar <mingo@kernel.org>
2017-08-30 01:51:03 +08:00
else if (op->dest.reg == cfa->base) {
/* mov %reg, %rsp */
if (cfa->base == CFI_SP &&
cfi->vals[op->src.reg].base == CFI_CFA) {
objtool: Handle GCC stack pointer adjustment bug Arnd Bergmann reported the following warning with GCC 7.1.1: fs/fs_pin.o: warning: objtool: pin_kill()+0x139: stack state mismatch: cfa1=7+88 cfa2=7+96 And the kbuild robot reported the following warnings with GCC 5.4.1: fs/fs_pin.o: warning: objtool: pin_kill()+0x182: return with modified stack frame fs/quota/dquot.o: warning: objtool: dquot_alloc_inode()+0x140: stack state mismatch: cfa1=7+120 cfa2=7+128 fs/quota/dquot.o: warning: objtool: dquot_free_inode()+0x11a: stack state mismatch: cfa1=7+112 cfa2=7+120 Those warnings are caused by an unusual GCC non-optimization where it uses an intermediate register to adjust the stack pointer. It does: lea 0x8(%rsp), %rcx ... mov %rcx, %rsp Instead of the obvious: add $0x8, %rsp It makes no sense to use an intermediate register, so I opened a GCC bug to track it: https://gcc.gnu.org/bugzilla/show_bug.cgi?id=81813 But it's not exactly a high-priority bug and it looks like we'll be stuck with this issue for a while. So for now we have to track register values when they're loaded with stack pointer offsets. This is kind of a big workaround for a tiny problem, but c'est la vie. I hope to eventually create a GCC plugin to implement a big chunk of objtool's functionality. Hopefully at that point we'll be able to remove of a lot of these GCC-isms from the objtool code. Reported-by: Arnd Bergmann <arnd@arndb.de> Reported-by: kbuild test robot <fengguang.wu@intel.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Thomas Gleixner <tglx@linutronix.de> Link: http://lkml.kernel.org/r/6a41a96884c725e7f05413bb7df40cfe824b2444.1504028945.git.jpoimboe@redhat.com Signed-off-by: Ingo Molnar <mingo@kernel.org>
2017-08-30 01:51:03 +08:00
/*
* This is needed for the rare case
* where GCC does something dumb like:
*
* lea 0x8(%rsp), %rcx
* ...
* mov %rcx, %rsp
*/
cfa->offset = -cfi->vals[op->src.reg].offset;
cfi->stack_size = cfa->offset;
objtool: Handle GCC stack pointer adjustment bug Arnd Bergmann reported the following warning with GCC 7.1.1: fs/fs_pin.o: warning: objtool: pin_kill()+0x139: stack state mismatch: cfa1=7+88 cfa2=7+96 And the kbuild robot reported the following warnings with GCC 5.4.1: fs/fs_pin.o: warning: objtool: pin_kill()+0x182: return with modified stack frame fs/quota/dquot.o: warning: objtool: dquot_alloc_inode()+0x140: stack state mismatch: cfa1=7+120 cfa2=7+128 fs/quota/dquot.o: warning: objtool: dquot_free_inode()+0x11a: stack state mismatch: cfa1=7+112 cfa2=7+120 Those warnings are caused by an unusual GCC non-optimization where it uses an intermediate register to adjust the stack pointer. It does: lea 0x8(%rsp), %rcx ... mov %rcx, %rsp Instead of the obvious: add $0x8, %rsp It makes no sense to use an intermediate register, so I opened a GCC bug to track it: https://gcc.gnu.org/bugzilla/show_bug.cgi?id=81813 But it's not exactly a high-priority bug and it looks like we'll be stuck with this issue for a while. So for now we have to track register values when they're loaded with stack pointer offsets. This is kind of a big workaround for a tiny problem, but c'est la vie. I hope to eventually create a GCC plugin to implement a big chunk of objtool's functionality. Hopefully at that point we'll be able to remove of a lot of these GCC-isms from the objtool code. Reported-by: Arnd Bergmann <arnd@arndb.de> Reported-by: kbuild test robot <fengguang.wu@intel.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Thomas Gleixner <tglx@linutronix.de> Link: http://lkml.kernel.org/r/6a41a96884c725e7f05413bb7df40cfe824b2444.1504028945.git.jpoimboe@redhat.com Signed-off-by: Ingo Molnar <mingo@kernel.org>
2017-08-30 01:51:03 +08:00
} else if (cfa->base == CFI_SP &&
cfi->vals[op->src.reg].base == CFI_SP_INDIRECT &&
cfi->vals[op->src.reg].offset == cfa->offset) {
/*
* Stack swizzle:
*
* 1: mov %rsp, (%[tos])
* 2: mov %[tos], %rsp
* ...
* 3: pop %rsp
*
* Where:
*
* 1 - places a pointer to the previous
* stack at the Top-of-Stack of the
* new stack.
*
* 2 - switches to the new stack.
*
* 3 - pops the Top-of-Stack to restore
* the original stack.
*
* Note: we set base to SP_INDIRECT
* here and preserve offset. Therefore
* when the unwinder reaches ToS it
* will dereference SP and then add the
* offset to find the next frame, IOW:
* (%rsp) + offset.
*/
cfa->base = CFI_SP_INDIRECT;
objtool: Handle GCC stack pointer adjustment bug Arnd Bergmann reported the following warning with GCC 7.1.1: fs/fs_pin.o: warning: objtool: pin_kill()+0x139: stack state mismatch: cfa1=7+88 cfa2=7+96 And the kbuild robot reported the following warnings with GCC 5.4.1: fs/fs_pin.o: warning: objtool: pin_kill()+0x182: return with modified stack frame fs/quota/dquot.o: warning: objtool: dquot_alloc_inode()+0x140: stack state mismatch: cfa1=7+120 cfa2=7+128 fs/quota/dquot.o: warning: objtool: dquot_free_inode()+0x11a: stack state mismatch: cfa1=7+112 cfa2=7+120 Those warnings are caused by an unusual GCC non-optimization where it uses an intermediate register to adjust the stack pointer. It does: lea 0x8(%rsp), %rcx ... mov %rcx, %rsp Instead of the obvious: add $0x8, %rsp It makes no sense to use an intermediate register, so I opened a GCC bug to track it: https://gcc.gnu.org/bugzilla/show_bug.cgi?id=81813 But it's not exactly a high-priority bug and it looks like we'll be stuck with this issue for a while. So for now we have to track register values when they're loaded with stack pointer offsets. This is kind of a big workaround for a tiny problem, but c'est la vie. I hope to eventually create a GCC plugin to implement a big chunk of objtool's functionality. Hopefully at that point we'll be able to remove of a lot of these GCC-isms from the objtool code. Reported-by: Arnd Bergmann <arnd@arndb.de> Reported-by: kbuild test robot <fengguang.wu@intel.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Thomas Gleixner <tglx@linutronix.de> Link: http://lkml.kernel.org/r/6a41a96884c725e7f05413bb7df40cfe824b2444.1504028945.git.jpoimboe@redhat.com Signed-off-by: Ingo Molnar <mingo@kernel.org>
2017-08-30 01:51:03 +08:00
} else {
cfa->base = CFI_UNDEFINED;
cfa->offset = 0;
}
}
else if (op->dest.reg == CFI_SP &&
cfi->vals[op->src.reg].base == CFI_SP_INDIRECT &&
cfi->vals[op->src.reg].offset == cfa->offset) {
/*
* The same stack swizzle case 2) as above. But
* because we can't change cfa->base, case 3)
* will become a regular POP. Pretend we're a
* PUSH so things don't go unbalanced.
*/
cfi->stack_size += 8;
}
break;
case OP_SRC_ADD:
if (op->dest.reg == CFI_SP && op->src.reg == CFI_SP) {
/* add imm, %rsp */
cfi->stack_size -= op->src.offset;
if (cfa->base == CFI_SP)
cfa->offset -= op->src.offset;
break;
}
if (op->dest.reg == CFI_SP && op->src.reg == CFI_BP) {
/* lea disp(%rbp), %rsp */
cfi->stack_size = -(op->src.offset + regs[CFI_BP].offset);
break;
}
if (!cfi->drap && op->src.reg == CFI_SP &&
op->dest.reg == CFI_BP && cfa->base == CFI_SP &&
check_reg_frame_pos(&regs[CFI_BP], -cfa->offset + op->src.offset)) {
/* lea disp(%rsp), %rbp */
cfa->base = CFI_BP;
cfa->offset -= op->src.offset;
cfi->bp_scratch = false;
break;
}
objtool: Handle GCC stack pointer adjustment bug Arnd Bergmann reported the following warning with GCC 7.1.1: fs/fs_pin.o: warning: objtool: pin_kill()+0x139: stack state mismatch: cfa1=7+88 cfa2=7+96 And the kbuild robot reported the following warnings with GCC 5.4.1: fs/fs_pin.o: warning: objtool: pin_kill()+0x182: return with modified stack frame fs/quota/dquot.o: warning: objtool: dquot_alloc_inode()+0x140: stack state mismatch: cfa1=7+120 cfa2=7+128 fs/quota/dquot.o: warning: objtool: dquot_free_inode()+0x11a: stack state mismatch: cfa1=7+112 cfa2=7+120 Those warnings are caused by an unusual GCC non-optimization where it uses an intermediate register to adjust the stack pointer. It does: lea 0x8(%rsp), %rcx ... mov %rcx, %rsp Instead of the obvious: add $0x8, %rsp It makes no sense to use an intermediate register, so I opened a GCC bug to track it: https://gcc.gnu.org/bugzilla/show_bug.cgi?id=81813 But it's not exactly a high-priority bug and it looks like we'll be stuck with this issue for a while. So for now we have to track register values when they're loaded with stack pointer offsets. This is kind of a big workaround for a tiny problem, but c'est la vie. I hope to eventually create a GCC plugin to implement a big chunk of objtool's functionality. Hopefully at that point we'll be able to remove of a lot of these GCC-isms from the objtool code. Reported-by: Arnd Bergmann <arnd@arndb.de> Reported-by: kbuild test robot <fengguang.wu@intel.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Thomas Gleixner <tglx@linutronix.de> Link: http://lkml.kernel.org/r/6a41a96884c725e7f05413bb7df40cfe824b2444.1504028945.git.jpoimboe@redhat.com Signed-off-by: Ingo Molnar <mingo@kernel.org>
2017-08-30 01:51:03 +08:00
if (op->src.reg == CFI_SP && cfa->base == CFI_SP) {
/* drap: lea disp(%rsp), %drap */
cfi->drap_reg = op->dest.reg;
objtool: Handle GCC stack pointer adjustment bug Arnd Bergmann reported the following warning with GCC 7.1.1: fs/fs_pin.o: warning: objtool: pin_kill()+0x139: stack state mismatch: cfa1=7+88 cfa2=7+96 And the kbuild robot reported the following warnings with GCC 5.4.1: fs/fs_pin.o: warning: objtool: pin_kill()+0x182: return with modified stack frame fs/quota/dquot.o: warning: objtool: dquot_alloc_inode()+0x140: stack state mismatch: cfa1=7+120 cfa2=7+128 fs/quota/dquot.o: warning: objtool: dquot_free_inode()+0x11a: stack state mismatch: cfa1=7+112 cfa2=7+120 Those warnings are caused by an unusual GCC non-optimization where it uses an intermediate register to adjust the stack pointer. It does: lea 0x8(%rsp), %rcx ... mov %rcx, %rsp Instead of the obvious: add $0x8, %rsp It makes no sense to use an intermediate register, so I opened a GCC bug to track it: https://gcc.gnu.org/bugzilla/show_bug.cgi?id=81813 But it's not exactly a high-priority bug and it looks like we'll be stuck with this issue for a while. So for now we have to track register values when they're loaded with stack pointer offsets. This is kind of a big workaround for a tiny problem, but c'est la vie. I hope to eventually create a GCC plugin to implement a big chunk of objtool's functionality. Hopefully at that point we'll be able to remove of a lot of these GCC-isms from the objtool code. Reported-by: Arnd Bergmann <arnd@arndb.de> Reported-by: kbuild test robot <fengguang.wu@intel.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Thomas Gleixner <tglx@linutronix.de> Link: http://lkml.kernel.org/r/6a41a96884c725e7f05413bb7df40cfe824b2444.1504028945.git.jpoimboe@redhat.com Signed-off-by: Ingo Molnar <mingo@kernel.org>
2017-08-30 01:51:03 +08:00
/*
* lea disp(%rsp), %reg
*
* This is needed for the rare case where GCC
* does something dumb like:
*
* lea 0x8(%rsp), %rcx
* ...
* mov %rcx, %rsp
*/
cfi->vals[op->dest.reg].base = CFI_CFA;
cfi->vals[op->dest.reg].offset = \
-cfi->stack_size + op->src.offset;
objtool: Handle GCC stack pointer adjustment bug Arnd Bergmann reported the following warning with GCC 7.1.1: fs/fs_pin.o: warning: objtool: pin_kill()+0x139: stack state mismatch: cfa1=7+88 cfa2=7+96 And the kbuild robot reported the following warnings with GCC 5.4.1: fs/fs_pin.o: warning: objtool: pin_kill()+0x182: return with modified stack frame fs/quota/dquot.o: warning: objtool: dquot_alloc_inode()+0x140: stack state mismatch: cfa1=7+120 cfa2=7+128 fs/quota/dquot.o: warning: objtool: dquot_free_inode()+0x11a: stack state mismatch: cfa1=7+112 cfa2=7+120 Those warnings are caused by an unusual GCC non-optimization where it uses an intermediate register to adjust the stack pointer. It does: lea 0x8(%rsp), %rcx ... mov %rcx, %rsp Instead of the obvious: add $0x8, %rsp It makes no sense to use an intermediate register, so I opened a GCC bug to track it: https://gcc.gnu.org/bugzilla/show_bug.cgi?id=81813 But it's not exactly a high-priority bug and it looks like we'll be stuck with this issue for a while. So for now we have to track register values when they're loaded with stack pointer offsets. This is kind of a big workaround for a tiny problem, but c'est la vie. I hope to eventually create a GCC plugin to implement a big chunk of objtool's functionality. Hopefully at that point we'll be able to remove of a lot of these GCC-isms from the objtool code. Reported-by: Arnd Bergmann <arnd@arndb.de> Reported-by: kbuild test robot <fengguang.wu@intel.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Thomas Gleixner <tglx@linutronix.de> Link: http://lkml.kernel.org/r/6a41a96884c725e7f05413bb7df40cfe824b2444.1504028945.git.jpoimboe@redhat.com Signed-off-by: Ingo Molnar <mingo@kernel.org>
2017-08-30 01:51:03 +08:00
break;
}
if (cfi->drap && op->dest.reg == CFI_SP &&
op->src.reg == cfi->drap_reg) {
/* drap: lea disp(%drap), %rsp */
cfa->base = CFI_SP;
cfa->offset = cfi->stack_size = -op->src.offset;
cfi->drap_reg = CFI_UNDEFINED;
cfi->drap = false;
break;
}
if (op->dest.reg == cfi->cfa.base && !(next_insn && next_insn->hint)) {
WARN_FUNC("unsupported stack register modification",
insn->sec, insn->offset);
return -1;
}
break;
case OP_SRC_AND:
if (op->dest.reg != CFI_SP ||
(cfi->drap_reg != CFI_UNDEFINED && cfa->base != CFI_SP) ||
(cfi->drap_reg == CFI_UNDEFINED && cfa->base != CFI_BP)) {
WARN_FUNC("unsupported stack pointer realignment",
insn->sec, insn->offset);
return -1;
}
if (cfi->drap_reg != CFI_UNDEFINED) {
/* drap: and imm, %rsp */
cfa->base = cfi->drap_reg;
cfa->offset = cfi->stack_size = 0;
cfi->drap = true;
}
/*
* Older versions of GCC (4.8ish) realign the stack
* without DRAP, with a frame pointer.
*/
break;
case OP_SRC_POP:
case OP_SRC_POPF:
if (op->dest.reg == CFI_SP && cfa->base == CFI_SP_INDIRECT) {
/* pop %rsp; # restore from a stack swizzle */
cfa->base = CFI_SP;
break;
}
if (!cfi->drap && op->dest.reg == cfa->base) {
/* pop %rbp */
cfa->base = CFI_SP;
}
if (cfi->drap && cfa->base == CFI_BP_INDIRECT &&
op->dest.reg == cfi->drap_reg &&
cfi->drap_offset == -cfi->stack_size) {
/* drap: pop %drap */
cfa->base = cfi->drap_reg;
cfa->offset = 0;
cfi->drap_offset = -1;
} else if (cfi->stack_size == -regs[op->dest.reg].offset) {
/* pop %reg */
restore_reg(cfi, op->dest.reg);
}
cfi->stack_size -= 8;
if (cfa->base == CFI_SP)
cfa->offset -= 8;
break;
case OP_SRC_REG_INDIRECT:
if (!cfi->drap && op->dest.reg == cfa->base &&
op->dest.reg == CFI_BP) {
/* mov disp(%rsp), %rbp */
cfa->base = CFI_SP;
cfa->offset = cfi->stack_size;
}
if (cfi->drap && op->src.reg == CFI_BP &&
op->src.offset == cfi->drap_offset) {
/* drap: mov disp(%rbp), %drap */
cfa->base = cfi->drap_reg;
cfa->offset = 0;
cfi->drap_offset = -1;
}
if (cfi->drap && op->src.reg == CFI_BP &&
op->src.offset == regs[op->dest.reg].offset) {
/* drap: mov disp(%rbp), %reg */
restore_reg(cfi, op->dest.reg);
} else if (op->src.reg == cfa->base &&
op->src.offset == regs[op->dest.reg].offset + cfa->offset) {
/* mov disp(%rbp), %reg */
/* mov disp(%rsp), %reg */
restore_reg(cfi, op->dest.reg);
} else if (op->src.reg == CFI_SP &&
op->src.offset == regs[op->dest.reg].offset + cfi->stack_size) {
/* mov disp(%rsp), %reg */
restore_reg(cfi, op->dest.reg);
}
break;
default:
WARN_FUNC("unknown stack-related instruction",
insn->sec, insn->offset);
return -1;
}
break;
case OP_DEST_PUSH:
case OP_DEST_PUSHF:
cfi->stack_size += 8;
if (cfa->base == CFI_SP)
cfa->offset += 8;
if (op->src.type != OP_SRC_REG)
break;
if (cfi->drap) {
if (op->src.reg == cfa->base && op->src.reg == cfi->drap_reg) {
/* drap: push %drap */
cfa->base = CFI_BP_INDIRECT;
cfa->offset = -cfi->stack_size;
/* save drap so we know when to restore it */
cfi->drap_offset = -cfi->stack_size;
} else if (op->src.reg == CFI_BP && cfa->base == cfi->drap_reg) {
/* drap: push %rbp */
cfi->stack_size = 0;
} else {
/* drap: push %reg */
save_reg(cfi, op->src.reg, CFI_BP, -cfi->stack_size);
}
} else {
/* push %reg */
save_reg(cfi, op->src.reg, CFI_CFA, -cfi->stack_size);
}
/* detect when asm code uses rbp as a scratch register */
if (opts.stackval && insn->func && op->src.reg == CFI_BP &&
cfa->base != CFI_BP)
cfi->bp_scratch = true;
break;
case OP_DEST_REG_INDIRECT:
if (cfi->drap) {
if (op->src.reg == cfa->base && op->src.reg == cfi->drap_reg) {
/* drap: mov %drap, disp(%rbp) */
cfa->base = CFI_BP_INDIRECT;
cfa->offset = op->dest.offset;
/* save drap offset so we know when to restore it */
cfi->drap_offset = op->dest.offset;
} else {
/* drap: mov reg, disp(%rbp) */
save_reg(cfi, op->src.reg, CFI_BP, op->dest.offset);
}
} else if (op->dest.reg == cfa->base) {
/* mov reg, disp(%rbp) */
/* mov reg, disp(%rsp) */
save_reg(cfi, op->src.reg, CFI_CFA,
op->dest.offset - cfi->cfa.offset);
} else if (op->dest.reg == CFI_SP) {
/* mov reg, disp(%rsp) */
save_reg(cfi, op->src.reg, CFI_CFA,
op->dest.offset - cfi->stack_size);
} else if (op->src.reg == CFI_SP && op->dest.offset == 0) {
/* mov %rsp, (%reg); # setup a stack swizzle. */
cfi->vals[op->dest.reg].base = CFI_SP_INDIRECT;
cfi->vals[op->dest.reg].offset = cfa->offset;
}
break;
case OP_DEST_MEM:
if (op->src.type != OP_SRC_POP && op->src.type != OP_SRC_POPF) {
WARN_FUNC("unknown stack-related memory operation",
insn->sec, insn->offset);
return -1;
}
/* pop mem */
cfi->stack_size -= 8;
if (cfa->base == CFI_SP)
cfa->offset -= 8;
break;
default:
WARN_FUNC("unknown stack-related instruction",
insn->sec, insn->offset);
return -1;
}
return 0;
}
objtool: Support stack layout changes in alternatives The ORC unwinder showed a warning [1] which revealed the stack layout didn't match what was expected. The problem was that paravirt patching had replaced "CALL *pv_ops.irq.save_fl" with "PUSHF;POP". That changed the stack layout between the PUSHF and the POP, so unwinding from an interrupt which occurred between those two instructions would fail. Part of the agreed upon solution was to rework the custom paravirt patching code to use alternatives instead, since objtool already knows how to read alternatives (and converging runtime patching infrastructure is always a good thing anyway). But the main problem still remains, which is that runtime patching can change the stack layout. Making stack layout changes in alternatives was disallowed with commit 7117f16bf460 ("objtool: Fix ORC vs alternatives"), but now that paravirt is going to be doing it, it needs to be supported. One way to do so would be to modify the ORC table when the code gets patched. But ORC is simple -- a good thing! -- and it's best to leave it alone. Instead, support stack layout changes by "flattening" all possible stack states (CFI) from parallel alternative code streams into a single set of linear states. The only necessary limitation is that CFI conflicts are disallowed at all possible instruction boundaries. For example, this scenario is allowed: Alt1 Alt2 Alt3 0x00 CALL *pv_ops.save_fl CALL xen_save_fl PUSHF 0x01 POP %RAX 0x02 NOP ... 0x05 NOP ... 0x07 <insn> The unwind information for offset-0x00 is identical for all 3 alternatives. Similarly offset-0x05 and higher also are identical (and the same as 0x00). However offset-0x01 has deviating CFI, but that is only relevant for Alt3, neither of the other alternative instruction streams will ever hit that offset. This scenario is NOT allowed: Alt1 Alt2 0x00 CALL *pv_ops.save_fl PUSHF 0x01 NOP6 ... 0x07 NOP POP %RAX The problem here is that offset-0x7, which is an instruction boundary in both possible instruction patch streams, has two conflicting stack layouts. [ The above examples were stolen from Peter Zijlstra. ] The new flattened CFI array is used both for the detection of conflicts (like the second example above) and the generation of linear ORC entries. BTW, another benefit of these changes is that, thanks to some related cleanups (new fake nops and alt_group struct) objtool can finally be rid of fake jumps, which were a constant source of headaches. [1] https://lkml.kernel.org/r/20201111170536.arx2zbn4ngvjoov7@treble Cc: Shinichiro Kawasaki <shinichiro.kawasaki@wdc.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
2020-12-19 04:26:21 +08:00
/*
* The stack layouts of alternatives instructions can sometimes diverge when
* they have stack modifications. That's fine as long as the potential stack
* layouts don't conflict at any given potential instruction boundary.
*
* Flatten the CFIs of the different alternative code streams (both original
* and replacement) into a single shared CFI array which can be used to detect
* conflicts and nicely feed a linear array of ORC entries to the unwinder.
*/
static int propagate_alt_cfi(struct objtool_file *file, struct instruction *insn)
{
objtool: Support stack layout changes in alternatives The ORC unwinder showed a warning [1] which revealed the stack layout didn't match what was expected. The problem was that paravirt patching had replaced "CALL *pv_ops.irq.save_fl" with "PUSHF;POP". That changed the stack layout between the PUSHF and the POP, so unwinding from an interrupt which occurred between those two instructions would fail. Part of the agreed upon solution was to rework the custom paravirt patching code to use alternatives instead, since objtool already knows how to read alternatives (and converging runtime patching infrastructure is always a good thing anyway). But the main problem still remains, which is that runtime patching can change the stack layout. Making stack layout changes in alternatives was disallowed with commit 7117f16bf460 ("objtool: Fix ORC vs alternatives"), but now that paravirt is going to be doing it, it needs to be supported. One way to do so would be to modify the ORC table when the code gets patched. But ORC is simple -- a good thing! -- and it's best to leave it alone. Instead, support stack layout changes by "flattening" all possible stack states (CFI) from parallel alternative code streams into a single set of linear states. The only necessary limitation is that CFI conflicts are disallowed at all possible instruction boundaries. For example, this scenario is allowed: Alt1 Alt2 Alt3 0x00 CALL *pv_ops.save_fl CALL xen_save_fl PUSHF 0x01 POP %RAX 0x02 NOP ... 0x05 NOP ... 0x07 <insn> The unwind information for offset-0x00 is identical for all 3 alternatives. Similarly offset-0x05 and higher also are identical (and the same as 0x00). However offset-0x01 has deviating CFI, but that is only relevant for Alt3, neither of the other alternative instruction streams will ever hit that offset. This scenario is NOT allowed: Alt1 Alt2 0x00 CALL *pv_ops.save_fl PUSHF 0x01 NOP6 ... 0x07 NOP POP %RAX The problem here is that offset-0x7, which is an instruction boundary in both possible instruction patch streams, has two conflicting stack layouts. [ The above examples were stolen from Peter Zijlstra. ] The new flattened CFI array is used both for the detection of conflicts (like the second example above) and the generation of linear ORC entries. BTW, another benefit of these changes is that, thanks to some related cleanups (new fake nops and alt_group struct) objtool can finally be rid of fake jumps, which were a constant source of headaches. [1] https://lkml.kernel.org/r/20201111170536.arx2zbn4ngvjoov7@treble Cc: Shinichiro Kawasaki <shinichiro.kawasaki@wdc.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
2020-12-19 04:26:21 +08:00
struct cfi_state **alt_cfi;
int group_off;
objtool: Support stack layout changes in alternatives The ORC unwinder showed a warning [1] which revealed the stack layout didn't match what was expected. The problem was that paravirt patching had replaced "CALL *pv_ops.irq.save_fl" with "PUSHF;POP". That changed the stack layout between the PUSHF and the POP, so unwinding from an interrupt which occurred between those two instructions would fail. Part of the agreed upon solution was to rework the custom paravirt patching code to use alternatives instead, since objtool already knows how to read alternatives (and converging runtime patching infrastructure is always a good thing anyway). But the main problem still remains, which is that runtime patching can change the stack layout. Making stack layout changes in alternatives was disallowed with commit 7117f16bf460 ("objtool: Fix ORC vs alternatives"), but now that paravirt is going to be doing it, it needs to be supported. One way to do so would be to modify the ORC table when the code gets patched. But ORC is simple -- a good thing! -- and it's best to leave it alone. Instead, support stack layout changes by "flattening" all possible stack states (CFI) from parallel alternative code streams into a single set of linear states. The only necessary limitation is that CFI conflicts are disallowed at all possible instruction boundaries. For example, this scenario is allowed: Alt1 Alt2 Alt3 0x00 CALL *pv_ops.save_fl CALL xen_save_fl PUSHF 0x01 POP %RAX 0x02 NOP ... 0x05 NOP ... 0x07 <insn> The unwind information for offset-0x00 is identical for all 3 alternatives. Similarly offset-0x05 and higher also are identical (and the same as 0x00). However offset-0x01 has deviating CFI, but that is only relevant for Alt3, neither of the other alternative instruction streams will ever hit that offset. This scenario is NOT allowed: Alt1 Alt2 0x00 CALL *pv_ops.save_fl PUSHF 0x01 NOP6 ... 0x07 NOP POP %RAX The problem here is that offset-0x7, which is an instruction boundary in both possible instruction patch streams, has two conflicting stack layouts. [ The above examples were stolen from Peter Zijlstra. ] The new flattened CFI array is used both for the detection of conflicts (like the second example above) and the generation of linear ORC entries. BTW, another benefit of these changes is that, thanks to some related cleanups (new fake nops and alt_group struct) objtool can finally be rid of fake jumps, which were a constant source of headaches. [1] https://lkml.kernel.org/r/20201111170536.arx2zbn4ngvjoov7@treble Cc: Shinichiro Kawasaki <shinichiro.kawasaki@wdc.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
2020-12-19 04:26:21 +08:00
if (!insn->alt_group)
return 0;
if (!insn->cfi) {
WARN("CFI missing");
return -1;
}
objtool: Support stack layout changes in alternatives The ORC unwinder showed a warning [1] which revealed the stack layout didn't match what was expected. The problem was that paravirt patching had replaced "CALL *pv_ops.irq.save_fl" with "PUSHF;POP". That changed the stack layout between the PUSHF and the POP, so unwinding from an interrupt which occurred between those two instructions would fail. Part of the agreed upon solution was to rework the custom paravirt patching code to use alternatives instead, since objtool already knows how to read alternatives (and converging runtime patching infrastructure is always a good thing anyway). But the main problem still remains, which is that runtime patching can change the stack layout. Making stack layout changes in alternatives was disallowed with commit 7117f16bf460 ("objtool: Fix ORC vs alternatives"), but now that paravirt is going to be doing it, it needs to be supported. One way to do so would be to modify the ORC table when the code gets patched. But ORC is simple -- a good thing! -- and it's best to leave it alone. Instead, support stack layout changes by "flattening" all possible stack states (CFI) from parallel alternative code streams into a single set of linear states. The only necessary limitation is that CFI conflicts are disallowed at all possible instruction boundaries. For example, this scenario is allowed: Alt1 Alt2 Alt3 0x00 CALL *pv_ops.save_fl CALL xen_save_fl PUSHF 0x01 POP %RAX 0x02 NOP ... 0x05 NOP ... 0x07 <insn> The unwind information for offset-0x00 is identical for all 3 alternatives. Similarly offset-0x05 and higher also are identical (and the same as 0x00). However offset-0x01 has deviating CFI, but that is only relevant for Alt3, neither of the other alternative instruction streams will ever hit that offset. This scenario is NOT allowed: Alt1 Alt2 0x00 CALL *pv_ops.save_fl PUSHF 0x01 NOP6 ... 0x07 NOP POP %RAX The problem here is that offset-0x7, which is an instruction boundary in both possible instruction patch streams, has two conflicting stack layouts. [ The above examples were stolen from Peter Zijlstra. ] The new flattened CFI array is used both for the detection of conflicts (like the second example above) and the generation of linear ORC entries. BTW, another benefit of these changes is that, thanks to some related cleanups (new fake nops and alt_group struct) objtool can finally be rid of fake jumps, which were a constant source of headaches. [1] https://lkml.kernel.org/r/20201111170536.arx2zbn4ngvjoov7@treble Cc: Shinichiro Kawasaki <shinichiro.kawasaki@wdc.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
2020-12-19 04:26:21 +08:00
alt_cfi = insn->alt_group->cfi;
group_off = insn->offset - insn->alt_group->first_insn->offset;
objtool: Support stack layout changes in alternatives The ORC unwinder showed a warning [1] which revealed the stack layout didn't match what was expected. The problem was that paravirt patching had replaced "CALL *pv_ops.irq.save_fl" with "PUSHF;POP". That changed the stack layout between the PUSHF and the POP, so unwinding from an interrupt which occurred between those two instructions would fail. Part of the agreed upon solution was to rework the custom paravirt patching code to use alternatives instead, since objtool already knows how to read alternatives (and converging runtime patching infrastructure is always a good thing anyway). But the main problem still remains, which is that runtime patching can change the stack layout. Making stack layout changes in alternatives was disallowed with commit 7117f16bf460 ("objtool: Fix ORC vs alternatives"), but now that paravirt is going to be doing it, it needs to be supported. One way to do so would be to modify the ORC table when the code gets patched. But ORC is simple -- a good thing! -- and it's best to leave it alone. Instead, support stack layout changes by "flattening" all possible stack states (CFI) from parallel alternative code streams into a single set of linear states. The only necessary limitation is that CFI conflicts are disallowed at all possible instruction boundaries. For example, this scenario is allowed: Alt1 Alt2 Alt3 0x00 CALL *pv_ops.save_fl CALL xen_save_fl PUSHF 0x01 POP %RAX 0x02 NOP ... 0x05 NOP ... 0x07 <insn> The unwind information for offset-0x00 is identical for all 3 alternatives. Similarly offset-0x05 and higher also are identical (and the same as 0x00). However offset-0x01 has deviating CFI, but that is only relevant for Alt3, neither of the other alternative instruction streams will ever hit that offset. This scenario is NOT allowed: Alt1 Alt2 0x00 CALL *pv_ops.save_fl PUSHF 0x01 NOP6 ... 0x07 NOP POP %RAX The problem here is that offset-0x7, which is an instruction boundary in both possible instruction patch streams, has two conflicting stack layouts. [ The above examples were stolen from Peter Zijlstra. ] The new flattened CFI array is used both for the detection of conflicts (like the second example above) and the generation of linear ORC entries. BTW, another benefit of these changes is that, thanks to some related cleanups (new fake nops and alt_group struct) objtool can finally be rid of fake jumps, which were a constant source of headaches. [1] https://lkml.kernel.org/r/20201111170536.arx2zbn4ngvjoov7@treble Cc: Shinichiro Kawasaki <shinichiro.kawasaki@wdc.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
2020-12-19 04:26:21 +08:00
if (!alt_cfi[group_off]) {
alt_cfi[group_off] = insn->cfi;
objtool: Support stack layout changes in alternatives The ORC unwinder showed a warning [1] which revealed the stack layout didn't match what was expected. The problem was that paravirt patching had replaced "CALL *pv_ops.irq.save_fl" with "PUSHF;POP". That changed the stack layout between the PUSHF and the POP, so unwinding from an interrupt which occurred between those two instructions would fail. Part of the agreed upon solution was to rework the custom paravirt patching code to use alternatives instead, since objtool already knows how to read alternatives (and converging runtime patching infrastructure is always a good thing anyway). But the main problem still remains, which is that runtime patching can change the stack layout. Making stack layout changes in alternatives was disallowed with commit 7117f16bf460 ("objtool: Fix ORC vs alternatives"), but now that paravirt is going to be doing it, it needs to be supported. One way to do so would be to modify the ORC table when the code gets patched. But ORC is simple -- a good thing! -- and it's best to leave it alone. Instead, support stack layout changes by "flattening" all possible stack states (CFI) from parallel alternative code streams into a single set of linear states. The only necessary limitation is that CFI conflicts are disallowed at all possible instruction boundaries. For example, this scenario is allowed: Alt1 Alt2 Alt3 0x00 CALL *pv_ops.save_fl CALL xen_save_fl PUSHF 0x01 POP %RAX 0x02 NOP ... 0x05 NOP ... 0x07 <insn> The unwind information for offset-0x00 is identical for all 3 alternatives. Similarly offset-0x05 and higher also are identical (and the same as 0x00). However offset-0x01 has deviating CFI, but that is only relevant for Alt3, neither of the other alternative instruction streams will ever hit that offset. This scenario is NOT allowed: Alt1 Alt2 0x00 CALL *pv_ops.save_fl PUSHF 0x01 NOP6 ... 0x07 NOP POP %RAX The problem here is that offset-0x7, which is an instruction boundary in both possible instruction patch streams, has two conflicting stack layouts. [ The above examples were stolen from Peter Zijlstra. ] The new flattened CFI array is used both for the detection of conflicts (like the second example above) and the generation of linear ORC entries. BTW, another benefit of these changes is that, thanks to some related cleanups (new fake nops and alt_group struct) objtool can finally be rid of fake jumps, which were a constant source of headaches. [1] https://lkml.kernel.org/r/20201111170536.arx2zbn4ngvjoov7@treble Cc: Shinichiro Kawasaki <shinichiro.kawasaki@wdc.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
2020-12-19 04:26:21 +08:00
} else {
if (cficmp(alt_cfi[group_off], insn->cfi)) {
objtool: Support stack layout changes in alternatives The ORC unwinder showed a warning [1] which revealed the stack layout didn't match what was expected. The problem was that paravirt patching had replaced "CALL *pv_ops.irq.save_fl" with "PUSHF;POP". That changed the stack layout between the PUSHF and the POP, so unwinding from an interrupt which occurred between those two instructions would fail. Part of the agreed upon solution was to rework the custom paravirt patching code to use alternatives instead, since objtool already knows how to read alternatives (and converging runtime patching infrastructure is always a good thing anyway). But the main problem still remains, which is that runtime patching can change the stack layout. Making stack layout changes in alternatives was disallowed with commit 7117f16bf460 ("objtool: Fix ORC vs alternatives"), but now that paravirt is going to be doing it, it needs to be supported. One way to do so would be to modify the ORC table when the code gets patched. But ORC is simple -- a good thing! -- and it's best to leave it alone. Instead, support stack layout changes by "flattening" all possible stack states (CFI) from parallel alternative code streams into a single set of linear states. The only necessary limitation is that CFI conflicts are disallowed at all possible instruction boundaries. For example, this scenario is allowed: Alt1 Alt2 Alt3 0x00 CALL *pv_ops.save_fl CALL xen_save_fl PUSHF 0x01 POP %RAX 0x02 NOP ... 0x05 NOP ... 0x07 <insn> The unwind information for offset-0x00 is identical for all 3 alternatives. Similarly offset-0x05 and higher also are identical (and the same as 0x00). However offset-0x01 has deviating CFI, but that is only relevant for Alt3, neither of the other alternative instruction streams will ever hit that offset. This scenario is NOT allowed: Alt1 Alt2 0x00 CALL *pv_ops.save_fl PUSHF 0x01 NOP6 ... 0x07 NOP POP %RAX The problem here is that offset-0x7, which is an instruction boundary in both possible instruction patch streams, has two conflicting stack layouts. [ The above examples were stolen from Peter Zijlstra. ] The new flattened CFI array is used both for the detection of conflicts (like the second example above) and the generation of linear ORC entries. BTW, another benefit of these changes is that, thanks to some related cleanups (new fake nops and alt_group struct) objtool can finally be rid of fake jumps, which were a constant source of headaches. [1] https://lkml.kernel.org/r/20201111170536.arx2zbn4ngvjoov7@treble Cc: Shinichiro Kawasaki <shinichiro.kawasaki@wdc.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
2020-12-19 04:26:21 +08:00
WARN_FUNC("stack layout conflict in alternatives",
insn->sec, insn->offset);
return -1;
}
objtool: Support stack layout changes in alternatives The ORC unwinder showed a warning [1] which revealed the stack layout didn't match what was expected. The problem was that paravirt patching had replaced "CALL *pv_ops.irq.save_fl" with "PUSHF;POP". That changed the stack layout between the PUSHF and the POP, so unwinding from an interrupt which occurred between those two instructions would fail. Part of the agreed upon solution was to rework the custom paravirt patching code to use alternatives instead, since objtool already knows how to read alternatives (and converging runtime patching infrastructure is always a good thing anyway). But the main problem still remains, which is that runtime patching can change the stack layout. Making stack layout changes in alternatives was disallowed with commit 7117f16bf460 ("objtool: Fix ORC vs alternatives"), but now that paravirt is going to be doing it, it needs to be supported. One way to do so would be to modify the ORC table when the code gets patched. But ORC is simple -- a good thing! -- and it's best to leave it alone. Instead, support stack layout changes by "flattening" all possible stack states (CFI) from parallel alternative code streams into a single set of linear states. The only necessary limitation is that CFI conflicts are disallowed at all possible instruction boundaries. For example, this scenario is allowed: Alt1 Alt2 Alt3 0x00 CALL *pv_ops.save_fl CALL xen_save_fl PUSHF 0x01 POP %RAX 0x02 NOP ... 0x05 NOP ... 0x07 <insn> The unwind information for offset-0x00 is identical for all 3 alternatives. Similarly offset-0x05 and higher also are identical (and the same as 0x00). However offset-0x01 has deviating CFI, but that is only relevant for Alt3, neither of the other alternative instruction streams will ever hit that offset. This scenario is NOT allowed: Alt1 Alt2 0x00 CALL *pv_ops.save_fl PUSHF 0x01 NOP6 ... 0x07 NOP POP %RAX The problem here is that offset-0x7, which is an instruction boundary in both possible instruction patch streams, has two conflicting stack layouts. [ The above examples were stolen from Peter Zijlstra. ] The new flattened CFI array is used both for the detection of conflicts (like the second example above) and the generation of linear ORC entries. BTW, another benefit of these changes is that, thanks to some related cleanups (new fake nops and alt_group struct) objtool can finally be rid of fake jumps, which were a constant source of headaches. [1] https://lkml.kernel.org/r/20201111170536.arx2zbn4ngvjoov7@treble Cc: Shinichiro Kawasaki <shinichiro.kawasaki@wdc.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
2020-12-19 04:26:21 +08:00
}
return 0;
}
static int handle_insn_ops(struct instruction *insn,
struct instruction *next_insn,
struct insn_state *state)
objtool: Support stack layout changes in alternatives The ORC unwinder showed a warning [1] which revealed the stack layout didn't match what was expected. The problem was that paravirt patching had replaced "CALL *pv_ops.irq.save_fl" with "PUSHF;POP". That changed the stack layout between the PUSHF and the POP, so unwinding from an interrupt which occurred between those two instructions would fail. Part of the agreed upon solution was to rework the custom paravirt patching code to use alternatives instead, since objtool already knows how to read alternatives (and converging runtime patching infrastructure is always a good thing anyway). But the main problem still remains, which is that runtime patching can change the stack layout. Making stack layout changes in alternatives was disallowed with commit 7117f16bf460 ("objtool: Fix ORC vs alternatives"), but now that paravirt is going to be doing it, it needs to be supported. One way to do so would be to modify the ORC table when the code gets patched. But ORC is simple -- a good thing! -- and it's best to leave it alone. Instead, support stack layout changes by "flattening" all possible stack states (CFI) from parallel alternative code streams into a single set of linear states. The only necessary limitation is that CFI conflicts are disallowed at all possible instruction boundaries. For example, this scenario is allowed: Alt1 Alt2 Alt3 0x00 CALL *pv_ops.save_fl CALL xen_save_fl PUSHF 0x01 POP %RAX 0x02 NOP ... 0x05 NOP ... 0x07 <insn> The unwind information for offset-0x00 is identical for all 3 alternatives. Similarly offset-0x05 and higher also are identical (and the same as 0x00). However offset-0x01 has deviating CFI, but that is only relevant for Alt3, neither of the other alternative instruction streams will ever hit that offset. This scenario is NOT allowed: Alt1 Alt2 0x00 CALL *pv_ops.save_fl PUSHF 0x01 NOP6 ... 0x07 NOP POP %RAX The problem here is that offset-0x7, which is an instruction boundary in both possible instruction patch streams, has two conflicting stack layouts. [ The above examples were stolen from Peter Zijlstra. ] The new flattened CFI array is used both for the detection of conflicts (like the second example above) and the generation of linear ORC entries. BTW, another benefit of these changes is that, thanks to some related cleanups (new fake nops and alt_group struct) objtool can finally be rid of fake jumps, which were a constant source of headaches. [1] https://lkml.kernel.org/r/20201111170536.arx2zbn4ngvjoov7@treble Cc: Shinichiro Kawasaki <shinichiro.kawasaki@wdc.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
2020-12-19 04:26:21 +08:00
{
struct stack_op *op;
list_for_each_entry(op, &insn->stack_ops, list) {
if (update_cfi_state(insn, next_insn, &state->cfi, op))
objtool: Support stack layout changes in alternatives The ORC unwinder showed a warning [1] which revealed the stack layout didn't match what was expected. The problem was that paravirt patching had replaced "CALL *pv_ops.irq.save_fl" with "PUSHF;POP". That changed the stack layout between the PUSHF and the POP, so unwinding from an interrupt which occurred between those two instructions would fail. Part of the agreed upon solution was to rework the custom paravirt patching code to use alternatives instead, since objtool already knows how to read alternatives (and converging runtime patching infrastructure is always a good thing anyway). But the main problem still remains, which is that runtime patching can change the stack layout. Making stack layout changes in alternatives was disallowed with commit 7117f16bf460 ("objtool: Fix ORC vs alternatives"), but now that paravirt is going to be doing it, it needs to be supported. One way to do so would be to modify the ORC table when the code gets patched. But ORC is simple -- a good thing! -- and it's best to leave it alone. Instead, support stack layout changes by "flattening" all possible stack states (CFI) from parallel alternative code streams into a single set of linear states. The only necessary limitation is that CFI conflicts are disallowed at all possible instruction boundaries. For example, this scenario is allowed: Alt1 Alt2 Alt3 0x00 CALL *pv_ops.save_fl CALL xen_save_fl PUSHF 0x01 POP %RAX 0x02 NOP ... 0x05 NOP ... 0x07 <insn> The unwind information for offset-0x00 is identical for all 3 alternatives. Similarly offset-0x05 and higher also are identical (and the same as 0x00). However offset-0x01 has deviating CFI, but that is only relevant for Alt3, neither of the other alternative instruction streams will ever hit that offset. This scenario is NOT allowed: Alt1 Alt2 0x00 CALL *pv_ops.save_fl PUSHF 0x01 NOP6 ... 0x07 NOP POP %RAX The problem here is that offset-0x7, which is an instruction boundary in both possible instruction patch streams, has two conflicting stack layouts. [ The above examples were stolen from Peter Zijlstra. ] The new flattened CFI array is used both for the detection of conflicts (like the second example above) and the generation of linear ORC entries. BTW, another benefit of these changes is that, thanks to some related cleanups (new fake nops and alt_group struct) objtool can finally be rid of fake jumps, which were a constant source of headaches. [1] https://lkml.kernel.org/r/20201111170536.arx2zbn4ngvjoov7@treble Cc: Shinichiro Kawasaki <shinichiro.kawasaki@wdc.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
2020-12-19 04:26:21 +08:00
return 1;
if (!insn->alt_group)
continue;
if (op->dest.type == OP_DEST_PUSHF) {
if (!state->uaccess_stack) {
state->uaccess_stack = 1;
} else if (state->uaccess_stack >> 31) {
WARN_FUNC("PUSHF stack exhausted",
insn->sec, insn->offset);
return 1;
}
state->uaccess_stack <<= 1;
state->uaccess_stack |= state->uaccess;
}
if (op->src.type == OP_SRC_POPF) {
if (state->uaccess_stack) {
state->uaccess = state->uaccess_stack & 1;
state->uaccess_stack >>= 1;
if (state->uaccess_stack == 1)
state->uaccess_stack = 0;
}
}
}
return 0;
}
static bool insn_cfi_match(struct instruction *insn, struct cfi_state *cfi2)
{
struct cfi_state *cfi1 = insn->cfi;
int i;
if (!cfi1) {
WARN("CFI missing");
return false;
}
if (memcmp(&cfi1->cfa, &cfi2->cfa, sizeof(cfi1->cfa))) {
WARN_FUNC("stack state mismatch: cfa1=%d%+d cfa2=%d%+d",
insn->sec, insn->offset,
cfi1->cfa.base, cfi1->cfa.offset,
cfi2->cfa.base, cfi2->cfa.offset);
} else if (memcmp(&cfi1->regs, &cfi2->regs, sizeof(cfi1->regs))) {
for (i = 0; i < CFI_NUM_REGS; i++) {
if (!memcmp(&cfi1->regs[i], &cfi2->regs[i],
sizeof(struct cfi_reg)))
continue;
WARN_FUNC("stack state mismatch: reg1[%d]=%d%+d reg2[%d]=%d%+d",
insn->sec, insn->offset,
i, cfi1->regs[i].base, cfi1->regs[i].offset,
i, cfi2->regs[i].base, cfi2->regs[i].offset);
break;
}
} else if (cfi1->type != cfi2->type) {
WARN_FUNC("stack state mismatch: type1=%d type2=%d",
insn->sec, insn->offset, cfi1->type, cfi2->type);
} else if (cfi1->drap != cfi2->drap ||
(cfi1->drap && cfi1->drap_reg != cfi2->drap_reg) ||
(cfi1->drap && cfi1->drap_offset != cfi2->drap_offset)) {
WARN_FUNC("stack state mismatch: drap1=%d(%d,%d) drap2=%d(%d,%d)",
insn->sec, insn->offset,
cfi1->drap, cfi1->drap_reg, cfi1->drap_offset,
cfi2->drap, cfi2->drap_reg, cfi2->drap_offset);
} else
return true;
return false;
}
static inline bool func_uaccess_safe(struct symbol *func)
{
if (func)
return func->uaccess_safe;
return false;
}
static inline const char *call_dest_name(struct instruction *insn)
{
static char pvname[19];
struct reloc *rel;
int idx;
if (insn->call_dest)
return insn->call_dest->name;
rel = insn_reloc(NULL, insn);
if (rel && !strcmp(rel->sym->name, "pv_ops")) {
idx = (rel->addend / sizeof(void *));
snprintf(pvname, sizeof(pvname), "pv_ops[%d]", idx);
return pvname;
}
return "{dynamic}";
}
static bool pv_call_dest(struct objtool_file *file, struct instruction *insn)
{
struct symbol *target;
struct reloc *rel;
int idx;
rel = insn_reloc(file, insn);
if (!rel || strcmp(rel->sym->name, "pv_ops"))
return false;
idx = (arch_dest_reloc_offset(rel->addend) / sizeof(void *));
if (file->pv_ops[idx].clean)
return true;
file->pv_ops[idx].clean = true;
list_for_each_entry(target, &file->pv_ops[idx].targets, pv_target) {
if (!target->sec->noinstr) {
WARN("pv_ops[%d]: %s", idx, target->name);
file->pv_ops[idx].clean = false;
}
}
return file->pv_ops[idx].clean;
}
static inline bool noinstr_call_dest(struct objtool_file *file,
struct instruction *insn,
struct symbol *func)
{
/*
* We can't deal with indirect function calls at present;
* assume they're instrumented.
*/
if (!func) {
if (file->pv_ops)
return pv_call_dest(file, insn);
return false;
}
/*
* If the symbol is from a noinstr section; we good.
*/
if (func->sec->noinstr)
return true;
/*
* The __ubsan_handle_*() calls are like WARN(), they only happen when
* something 'BAD' happened. At the risk of taking the machine down,
* let them proceed to get the message out.
*/
if (!strncmp(func->name, "__ubsan_handle_", 15))
return true;
return false;
}
static int validate_call(struct objtool_file *file,
struct instruction *insn,
struct insn_state *state)
{
if (state->noinstr && state->instr <= 0 &&
!noinstr_call_dest(file, insn, insn->call_dest)) {
WARN_FUNC("call to %s() leaves .noinstr.text section",
insn->sec, insn->offset, call_dest_name(insn));
return 1;
}
if (state->uaccess && !func_uaccess_safe(insn->call_dest)) {
WARN_FUNC("call to %s() with UACCESS enabled",
insn->sec, insn->offset, call_dest_name(insn));
return 1;
}
if (state->df) {
WARN_FUNC("call to %s() with DF set",
insn->sec, insn->offset, call_dest_name(insn));
return 1;
}
return 0;
}
static int validate_sibling_call(struct objtool_file *file,
struct instruction *insn,
struct insn_state *state)
{
if (has_modified_stack_frame(insn, state)) {
WARN_FUNC("sibling call from callable instruction with modified stack frame",
insn->sec, insn->offset);
return 1;
}
return validate_call(file, insn, state);
}
static int validate_return(struct symbol *func, struct instruction *insn, struct insn_state *state)
{
if (state->noinstr && state->instr > 0) {
WARN_FUNC("return with instrumentation enabled",
insn->sec, insn->offset);
return 1;
}
if (state->uaccess && !func_uaccess_safe(func)) {
WARN_FUNC("return with UACCESS enabled",
insn->sec, insn->offset);
return 1;
}
if (!state->uaccess && func_uaccess_safe(func)) {
WARN_FUNC("return with UACCESS disabled from a UACCESS-safe function",
insn->sec, insn->offset);
return 1;
}
if (state->df) {
WARN_FUNC("return with DF set",
insn->sec, insn->offset);
return 1;
}
if (func && has_modified_stack_frame(insn, state)) {
WARN_FUNC("return with modified stack frame",
insn->sec, insn->offset);
return 1;
}
if (state->cfi.bp_scratch) {
WARN_FUNC("BP used as a scratch register",
insn->sec, insn->offset);
return 1;
}
return 0;
}
objtool: Support stack layout changes in alternatives The ORC unwinder showed a warning [1] which revealed the stack layout didn't match what was expected. The problem was that paravirt patching had replaced "CALL *pv_ops.irq.save_fl" with "PUSHF;POP". That changed the stack layout between the PUSHF and the POP, so unwinding from an interrupt which occurred between those two instructions would fail. Part of the agreed upon solution was to rework the custom paravirt patching code to use alternatives instead, since objtool already knows how to read alternatives (and converging runtime patching infrastructure is always a good thing anyway). But the main problem still remains, which is that runtime patching can change the stack layout. Making stack layout changes in alternatives was disallowed with commit 7117f16bf460 ("objtool: Fix ORC vs alternatives"), but now that paravirt is going to be doing it, it needs to be supported. One way to do so would be to modify the ORC table when the code gets patched. But ORC is simple -- a good thing! -- and it's best to leave it alone. Instead, support stack layout changes by "flattening" all possible stack states (CFI) from parallel alternative code streams into a single set of linear states. The only necessary limitation is that CFI conflicts are disallowed at all possible instruction boundaries. For example, this scenario is allowed: Alt1 Alt2 Alt3 0x00 CALL *pv_ops.save_fl CALL xen_save_fl PUSHF 0x01 POP %RAX 0x02 NOP ... 0x05 NOP ... 0x07 <insn> The unwind information for offset-0x00 is identical for all 3 alternatives. Similarly offset-0x05 and higher also are identical (and the same as 0x00). However offset-0x01 has deviating CFI, but that is only relevant for Alt3, neither of the other alternative instruction streams will ever hit that offset. This scenario is NOT allowed: Alt1 Alt2 0x00 CALL *pv_ops.save_fl PUSHF 0x01 NOP6 ... 0x07 NOP POP %RAX The problem here is that offset-0x7, which is an instruction boundary in both possible instruction patch streams, has two conflicting stack layouts. [ The above examples were stolen from Peter Zijlstra. ] The new flattened CFI array is used both for the detection of conflicts (like the second example above) and the generation of linear ORC entries. BTW, another benefit of these changes is that, thanks to some related cleanups (new fake nops and alt_group struct) objtool can finally be rid of fake jumps, which were a constant source of headaches. [1] https://lkml.kernel.org/r/20201111170536.arx2zbn4ngvjoov7@treble Cc: Shinichiro Kawasaki <shinichiro.kawasaki@wdc.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
2020-12-19 04:26:21 +08:00
static struct instruction *next_insn_to_validate(struct objtool_file *file,
struct instruction *insn)
objtool: Fix ORC vs alternatives Jann reported that (for instance) entry_64.o:general_protection has very odd ORC data: 0000000000000f40 <general_protection>: #######sp:sp+8 bp:(und) type:iret end:0 f40: 90 nop #######sp:(und) bp:(und) type:call end:0 f41: 90 nop f42: 90 nop #######sp:sp+8 bp:(und) type:iret end:0 f43: e8 a8 01 00 00 callq 10f0 <error_entry> #######sp:sp+0 bp:(und) type:regs end:0 f48: f6 84 24 88 00 00 00 testb $0x3,0x88(%rsp) f4f: 03 f50: 74 00 je f52 <general_protection+0x12> f52: 48 89 e7 mov %rsp,%rdi f55: 48 8b 74 24 78 mov 0x78(%rsp),%rsi f5a: 48 c7 44 24 78 ff ff movq $0xffffffffffffffff,0x78(%rsp) f61: ff ff f63: e8 00 00 00 00 callq f68 <general_protection+0x28> f68: e9 73 02 00 00 jmpq 11e0 <error_exit> #######sp:(und) bp:(und) type:call end:0 f6d: 0f 1f 00 nopl (%rax) Note the entry at 0xf41. Josh found this was the result of commit: 764eef4b109a ("objtool: Rewrite alt->skip_orig") Due to the early return in validate_branch() we no longer set insn->cfi of the original instruction stream (the NOPs at 0xf41 and 0xf42) and we'll end up with the above weirdness. In other discussions we realized alternatives should be ORC invariant; that is, due to there being only a single ORC table, it must be valid for all alternatives. The easiest way to ensure this is to not allow any stack modifications in alternatives. When we enforce this latter observation, we get the property that the whole alternative must have the same CFI, which we can employ to fix the former report. Fixes: 764eef4b109a ("objtool: Rewrite alt->skip_orig") Reported-by: Jann Horn <jannh@google.com> Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Reviewed-by: Miroslav Benes <mbenes@suse.cz> Acked-by: Josh Poimboeuf <jpoimboe@redhat.com> Link: https://lkml.kernel.org/r/20200428191659.499074346@infradead.org
2020-04-29 01:37:01 +08:00
{
struct alt_group *alt_group = insn->alt_group;
objtool: Fix ORC vs alternatives Jann reported that (for instance) entry_64.o:general_protection has very odd ORC data: 0000000000000f40 <general_protection>: #######sp:sp+8 bp:(und) type:iret end:0 f40: 90 nop #######sp:(und) bp:(und) type:call end:0 f41: 90 nop f42: 90 nop #######sp:sp+8 bp:(und) type:iret end:0 f43: e8 a8 01 00 00 callq 10f0 <error_entry> #######sp:sp+0 bp:(und) type:regs end:0 f48: f6 84 24 88 00 00 00 testb $0x3,0x88(%rsp) f4f: 03 f50: 74 00 je f52 <general_protection+0x12> f52: 48 89 e7 mov %rsp,%rdi f55: 48 8b 74 24 78 mov 0x78(%rsp),%rsi f5a: 48 c7 44 24 78 ff ff movq $0xffffffffffffffff,0x78(%rsp) f61: ff ff f63: e8 00 00 00 00 callq f68 <general_protection+0x28> f68: e9 73 02 00 00 jmpq 11e0 <error_exit> #######sp:(und) bp:(und) type:call end:0 f6d: 0f 1f 00 nopl (%rax) Note the entry at 0xf41. Josh found this was the result of commit: 764eef4b109a ("objtool: Rewrite alt->skip_orig") Due to the early return in validate_branch() we no longer set insn->cfi of the original instruction stream (the NOPs at 0xf41 and 0xf42) and we'll end up with the above weirdness. In other discussions we realized alternatives should be ORC invariant; that is, due to there being only a single ORC table, it must be valid for all alternatives. The easiest way to ensure this is to not allow any stack modifications in alternatives. When we enforce this latter observation, we get the property that the whole alternative must have the same CFI, which we can employ to fix the former report. Fixes: 764eef4b109a ("objtool: Rewrite alt->skip_orig") Reported-by: Jann Horn <jannh@google.com> Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Reviewed-by: Miroslav Benes <mbenes@suse.cz> Acked-by: Josh Poimboeuf <jpoimboe@redhat.com> Link: https://lkml.kernel.org/r/20200428191659.499074346@infradead.org
2020-04-29 01:37:01 +08:00
objtool: Support stack layout changes in alternatives The ORC unwinder showed a warning [1] which revealed the stack layout didn't match what was expected. The problem was that paravirt patching had replaced "CALL *pv_ops.irq.save_fl" with "PUSHF;POP". That changed the stack layout between the PUSHF and the POP, so unwinding from an interrupt which occurred between those two instructions would fail. Part of the agreed upon solution was to rework the custom paravirt patching code to use alternatives instead, since objtool already knows how to read alternatives (and converging runtime patching infrastructure is always a good thing anyway). But the main problem still remains, which is that runtime patching can change the stack layout. Making stack layout changes in alternatives was disallowed with commit 7117f16bf460 ("objtool: Fix ORC vs alternatives"), but now that paravirt is going to be doing it, it needs to be supported. One way to do so would be to modify the ORC table when the code gets patched. But ORC is simple -- a good thing! -- and it's best to leave it alone. Instead, support stack layout changes by "flattening" all possible stack states (CFI) from parallel alternative code streams into a single set of linear states. The only necessary limitation is that CFI conflicts are disallowed at all possible instruction boundaries. For example, this scenario is allowed: Alt1 Alt2 Alt3 0x00 CALL *pv_ops.save_fl CALL xen_save_fl PUSHF 0x01 POP %RAX 0x02 NOP ... 0x05 NOP ... 0x07 <insn> The unwind information for offset-0x00 is identical for all 3 alternatives. Similarly offset-0x05 and higher also are identical (and the same as 0x00). However offset-0x01 has deviating CFI, but that is only relevant for Alt3, neither of the other alternative instruction streams will ever hit that offset. This scenario is NOT allowed: Alt1 Alt2 0x00 CALL *pv_ops.save_fl PUSHF 0x01 NOP6 ... 0x07 NOP POP %RAX The problem here is that offset-0x7, which is an instruction boundary in both possible instruction patch streams, has two conflicting stack layouts. [ The above examples were stolen from Peter Zijlstra. ] The new flattened CFI array is used both for the detection of conflicts (like the second example above) and the generation of linear ORC entries. BTW, another benefit of these changes is that, thanks to some related cleanups (new fake nops and alt_group struct) objtool can finally be rid of fake jumps, which were a constant source of headaches. [1] https://lkml.kernel.org/r/20201111170536.arx2zbn4ngvjoov7@treble Cc: Shinichiro Kawasaki <shinichiro.kawasaki@wdc.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
2020-12-19 04:26:21 +08:00
/*
* Simulate the fact that alternatives are patched in-place. When the
* end of a replacement alt_group is reached, redirect objtool flow to
* the end of the original alt_group.
*/
if (alt_group && insn == alt_group->last_insn && alt_group->orig_group)
return next_insn_same_sec(file, alt_group->orig_group->last_insn);
return next_insn_same_sec(file, insn);
objtool: Fix ORC vs alternatives Jann reported that (for instance) entry_64.o:general_protection has very odd ORC data: 0000000000000f40 <general_protection>: #######sp:sp+8 bp:(und) type:iret end:0 f40: 90 nop #######sp:(und) bp:(und) type:call end:0 f41: 90 nop f42: 90 nop #######sp:sp+8 bp:(und) type:iret end:0 f43: e8 a8 01 00 00 callq 10f0 <error_entry> #######sp:sp+0 bp:(und) type:regs end:0 f48: f6 84 24 88 00 00 00 testb $0x3,0x88(%rsp) f4f: 03 f50: 74 00 je f52 <general_protection+0x12> f52: 48 89 e7 mov %rsp,%rdi f55: 48 8b 74 24 78 mov 0x78(%rsp),%rsi f5a: 48 c7 44 24 78 ff ff movq $0xffffffffffffffff,0x78(%rsp) f61: ff ff f63: e8 00 00 00 00 callq f68 <general_protection+0x28> f68: e9 73 02 00 00 jmpq 11e0 <error_exit> #######sp:(und) bp:(und) type:call end:0 f6d: 0f 1f 00 nopl (%rax) Note the entry at 0xf41. Josh found this was the result of commit: 764eef4b109a ("objtool: Rewrite alt->skip_orig") Due to the early return in validate_branch() we no longer set insn->cfi of the original instruction stream (the NOPs at 0xf41 and 0xf42) and we'll end up with the above weirdness. In other discussions we realized alternatives should be ORC invariant; that is, due to there being only a single ORC table, it must be valid for all alternatives. The easiest way to ensure this is to not allow any stack modifications in alternatives. When we enforce this latter observation, we get the property that the whole alternative must have the same CFI, which we can employ to fix the former report. Fixes: 764eef4b109a ("objtool: Rewrite alt->skip_orig") Reported-by: Jann Horn <jannh@google.com> Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Reviewed-by: Miroslav Benes <mbenes@suse.cz> Acked-by: Josh Poimboeuf <jpoimboe@redhat.com> Link: https://lkml.kernel.org/r/20200428191659.499074346@infradead.org
2020-04-29 01:37:01 +08:00
}
/*
* Follow the branch starting at the given instruction, and recursively follow
* any other branches (jumps). Meanwhile, track the frame pointer state at
* each instruction and validate all the rules described in
* tools/objtool/Documentation/stack-validation.txt.
*/
static int validate_branch(struct objtool_file *file, struct symbol *func,
struct instruction *insn, struct insn_state state)
{
struct alternative *alt;
struct instruction *next_insn, *prev_insn = NULL;
struct section *sec;
objtool: Improve UACCESS coverage A clang build reported an (obvious) double CLAC while a GCC build did not; it turns out that objtool only re-visits instructions if the first visit was with AC=0. If OTOH the first visit was with AC=1, it completely ignores any subsequent visit, even when it has AC=0. Fix this by using a visited mask instead of a boolean, and (explicitly) mark the AC state. $ ./objtool check -b --no-fp --retpoline --uaccess drivers/gpu/drm/i915/gem/i915_gem_execbuffer.o drivers/gpu/drm/i915/gem/i915_gem_execbuffer.o: warning: objtool: .altinstr_replacement+0x22: redundant UACCESS disable drivers/gpu/drm/i915/gem/i915_gem_execbuffer.o: warning: objtool: eb_copy_relocations.isra.34()+0xea: (alt) drivers/gpu/drm/i915/gem/i915_gem_execbuffer.o: warning: objtool: .altinstr_replacement+0xffffffffffffffff: (branch) drivers/gpu/drm/i915/gem/i915_gem_execbuffer.o: warning: objtool: eb_copy_relocations.isra.34()+0xd9: (alt) drivers/gpu/drm/i915/gem/i915_gem_execbuffer.o: warning: objtool: eb_copy_relocations.isra.34()+0xb2: (branch) drivers/gpu/drm/i915/gem/i915_gem_execbuffer.o: warning: objtool: eb_copy_relocations.isra.34()+0x39: (branch) drivers/gpu/drm/i915/gem/i915_gem_execbuffer.o: warning: objtool: eb_copy_relocations.isra.34()+0x0: <=== (func) Reported-by: Josh Poimboeuf <jpoimboe@redhat.com> Reported-by: Thomas Gleixner <tglx@linutronix.de> Reported-by: Sedat Dilek <sedat.dilek@gmail.com> Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com> Signed-off-by: Thomas Gleixner <tglx@linutronix.de> Tested-by: Nathan Chancellor <natechancellor@gmail.com> Tested-by: Nick Desaulniers <ndesaulniers@google.com> Tested-by: Sedat Dilek <sedat.dilek@gmail.com> Link: https://github.com/ClangBuiltLinux/linux/issues/617 Link: https://lkml.kernel.org/r/5359166aad2d53f3145cd442d83d0e5115e0cd17.1564007838.git.jpoimboe@redhat.com
2019-07-25 06:47:26 +08:00
u8 visited;
int ret;
sec = insn->sec;
while (1) {
objtool: Support stack layout changes in alternatives The ORC unwinder showed a warning [1] which revealed the stack layout didn't match what was expected. The problem was that paravirt patching had replaced "CALL *pv_ops.irq.save_fl" with "PUSHF;POP". That changed the stack layout between the PUSHF and the POP, so unwinding from an interrupt which occurred between those two instructions would fail. Part of the agreed upon solution was to rework the custom paravirt patching code to use alternatives instead, since objtool already knows how to read alternatives (and converging runtime patching infrastructure is always a good thing anyway). But the main problem still remains, which is that runtime patching can change the stack layout. Making stack layout changes in alternatives was disallowed with commit 7117f16bf460 ("objtool: Fix ORC vs alternatives"), but now that paravirt is going to be doing it, it needs to be supported. One way to do so would be to modify the ORC table when the code gets patched. But ORC is simple -- a good thing! -- and it's best to leave it alone. Instead, support stack layout changes by "flattening" all possible stack states (CFI) from parallel alternative code streams into a single set of linear states. The only necessary limitation is that CFI conflicts are disallowed at all possible instruction boundaries. For example, this scenario is allowed: Alt1 Alt2 Alt3 0x00 CALL *pv_ops.save_fl CALL xen_save_fl PUSHF 0x01 POP %RAX 0x02 NOP ... 0x05 NOP ... 0x07 <insn> The unwind information for offset-0x00 is identical for all 3 alternatives. Similarly offset-0x05 and higher also are identical (and the same as 0x00). However offset-0x01 has deviating CFI, but that is only relevant for Alt3, neither of the other alternative instruction streams will ever hit that offset. This scenario is NOT allowed: Alt1 Alt2 0x00 CALL *pv_ops.save_fl PUSHF 0x01 NOP6 ... 0x07 NOP POP %RAX The problem here is that offset-0x7, which is an instruction boundary in both possible instruction patch streams, has two conflicting stack layouts. [ The above examples were stolen from Peter Zijlstra. ] The new flattened CFI array is used both for the detection of conflicts (like the second example above) and the generation of linear ORC entries. BTW, another benefit of these changes is that, thanks to some related cleanups (new fake nops and alt_group struct) objtool can finally be rid of fake jumps, which were a constant source of headaches. [1] https://lkml.kernel.org/r/20201111170536.arx2zbn4ngvjoov7@treble Cc: Shinichiro Kawasaki <shinichiro.kawasaki@wdc.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
2020-12-19 04:26:21 +08:00
next_insn = next_insn_to_validate(file, insn);
if (func && insn->func && func != insn->func->pfunc) {
WARN("%s() falls through to next function %s()",
func->name, insn->func->name);
return 1;
}
if (func && insn->ignore) {
WARN_FUNC("BUG: why am I validating an ignored function?",
sec, insn->offset);
return 1;
}
objtool: Improve UACCESS coverage A clang build reported an (obvious) double CLAC while a GCC build did not; it turns out that objtool only re-visits instructions if the first visit was with AC=0. If OTOH the first visit was with AC=1, it completely ignores any subsequent visit, even when it has AC=0. Fix this by using a visited mask instead of a boolean, and (explicitly) mark the AC state. $ ./objtool check -b --no-fp --retpoline --uaccess drivers/gpu/drm/i915/gem/i915_gem_execbuffer.o drivers/gpu/drm/i915/gem/i915_gem_execbuffer.o: warning: objtool: .altinstr_replacement+0x22: redundant UACCESS disable drivers/gpu/drm/i915/gem/i915_gem_execbuffer.o: warning: objtool: eb_copy_relocations.isra.34()+0xea: (alt) drivers/gpu/drm/i915/gem/i915_gem_execbuffer.o: warning: objtool: .altinstr_replacement+0xffffffffffffffff: (branch) drivers/gpu/drm/i915/gem/i915_gem_execbuffer.o: warning: objtool: eb_copy_relocations.isra.34()+0xd9: (alt) drivers/gpu/drm/i915/gem/i915_gem_execbuffer.o: warning: objtool: eb_copy_relocations.isra.34()+0xb2: (branch) drivers/gpu/drm/i915/gem/i915_gem_execbuffer.o: warning: objtool: eb_copy_relocations.isra.34()+0x39: (branch) drivers/gpu/drm/i915/gem/i915_gem_execbuffer.o: warning: objtool: eb_copy_relocations.isra.34()+0x0: <=== (func) Reported-by: Josh Poimboeuf <jpoimboe@redhat.com> Reported-by: Thomas Gleixner <tglx@linutronix.de> Reported-by: Sedat Dilek <sedat.dilek@gmail.com> Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com> Signed-off-by: Thomas Gleixner <tglx@linutronix.de> Tested-by: Nathan Chancellor <natechancellor@gmail.com> Tested-by: Nick Desaulniers <ndesaulniers@google.com> Tested-by: Sedat Dilek <sedat.dilek@gmail.com> Link: https://github.com/ClangBuiltLinux/linux/issues/617 Link: https://lkml.kernel.org/r/5359166aad2d53f3145cd442d83d0e5115e0cd17.1564007838.git.jpoimboe@redhat.com
2019-07-25 06:47:26 +08:00
visited = 1 << state.uaccess;
if (insn->visited) {
if (!insn->hint && !insn_cfi_match(insn, &state.cfi))
return 1;
objtool: Improve UACCESS coverage A clang build reported an (obvious) double CLAC while a GCC build did not; it turns out that objtool only re-visits instructions if the first visit was with AC=0. If OTOH the first visit was with AC=1, it completely ignores any subsequent visit, even when it has AC=0. Fix this by using a visited mask instead of a boolean, and (explicitly) mark the AC state. $ ./objtool check -b --no-fp --retpoline --uaccess drivers/gpu/drm/i915/gem/i915_gem_execbuffer.o drivers/gpu/drm/i915/gem/i915_gem_execbuffer.o: warning: objtool: .altinstr_replacement+0x22: redundant UACCESS disable drivers/gpu/drm/i915/gem/i915_gem_execbuffer.o: warning: objtool: eb_copy_relocations.isra.34()+0xea: (alt) drivers/gpu/drm/i915/gem/i915_gem_execbuffer.o: warning: objtool: .altinstr_replacement+0xffffffffffffffff: (branch) drivers/gpu/drm/i915/gem/i915_gem_execbuffer.o: warning: objtool: eb_copy_relocations.isra.34()+0xd9: (alt) drivers/gpu/drm/i915/gem/i915_gem_execbuffer.o: warning: objtool: eb_copy_relocations.isra.34()+0xb2: (branch) drivers/gpu/drm/i915/gem/i915_gem_execbuffer.o: warning: objtool: eb_copy_relocations.isra.34()+0x39: (branch) drivers/gpu/drm/i915/gem/i915_gem_execbuffer.o: warning: objtool: eb_copy_relocations.isra.34()+0x0: <=== (func) Reported-by: Josh Poimboeuf <jpoimboe@redhat.com> Reported-by: Thomas Gleixner <tglx@linutronix.de> Reported-by: Sedat Dilek <sedat.dilek@gmail.com> Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com> Signed-off-by: Thomas Gleixner <tglx@linutronix.de> Tested-by: Nathan Chancellor <natechancellor@gmail.com> Tested-by: Nick Desaulniers <ndesaulniers@google.com> Tested-by: Sedat Dilek <sedat.dilek@gmail.com> Link: https://github.com/ClangBuiltLinux/linux/issues/617 Link: https://lkml.kernel.org/r/5359166aad2d53f3145cd442d83d0e5115e0cd17.1564007838.git.jpoimboe@redhat.com
2019-07-25 06:47:26 +08:00
if (insn->visited & visited)
return 0;
} else {
nr_insns_visited++;
}
if (state.noinstr)
state.instr += insn->instr;
if (insn->hint) {
state.cfi = *insn->cfi;
} else {
/* XXX track if we actually changed state.cfi */
if (prev_insn && !cficmp(prev_insn->cfi, &state.cfi)) {
insn->cfi = prev_insn->cfi;
nr_cfi_reused++;
} else {
insn->cfi = cfi_hash_find_or_add(&state.cfi);
}
}
objtool: Improve UACCESS coverage A clang build reported an (obvious) double CLAC while a GCC build did not; it turns out that objtool only re-visits instructions if the first visit was with AC=0. If OTOH the first visit was with AC=1, it completely ignores any subsequent visit, even when it has AC=0. Fix this by using a visited mask instead of a boolean, and (explicitly) mark the AC state. $ ./objtool check -b --no-fp --retpoline --uaccess drivers/gpu/drm/i915/gem/i915_gem_execbuffer.o drivers/gpu/drm/i915/gem/i915_gem_execbuffer.o: warning: objtool: .altinstr_replacement+0x22: redundant UACCESS disable drivers/gpu/drm/i915/gem/i915_gem_execbuffer.o: warning: objtool: eb_copy_relocations.isra.34()+0xea: (alt) drivers/gpu/drm/i915/gem/i915_gem_execbuffer.o: warning: objtool: .altinstr_replacement+0xffffffffffffffff: (branch) drivers/gpu/drm/i915/gem/i915_gem_execbuffer.o: warning: objtool: eb_copy_relocations.isra.34()+0xd9: (alt) drivers/gpu/drm/i915/gem/i915_gem_execbuffer.o: warning: objtool: eb_copy_relocations.isra.34()+0xb2: (branch) drivers/gpu/drm/i915/gem/i915_gem_execbuffer.o: warning: objtool: eb_copy_relocations.isra.34()+0x39: (branch) drivers/gpu/drm/i915/gem/i915_gem_execbuffer.o: warning: objtool: eb_copy_relocations.isra.34()+0x0: <=== (func) Reported-by: Josh Poimboeuf <jpoimboe@redhat.com> Reported-by: Thomas Gleixner <tglx@linutronix.de> Reported-by: Sedat Dilek <sedat.dilek@gmail.com> Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com> Signed-off-by: Thomas Gleixner <tglx@linutronix.de> Tested-by: Nathan Chancellor <natechancellor@gmail.com> Tested-by: Nick Desaulniers <ndesaulniers@google.com> Tested-by: Sedat Dilek <sedat.dilek@gmail.com> Link: https://github.com/ClangBuiltLinux/linux/issues/617 Link: https://lkml.kernel.org/r/5359166aad2d53f3145cd442d83d0e5115e0cd17.1564007838.git.jpoimboe@redhat.com
2019-07-25 06:47:26 +08:00
insn->visited |= visited;
objtool: Support stack layout changes in alternatives The ORC unwinder showed a warning [1] which revealed the stack layout didn't match what was expected. The problem was that paravirt patching had replaced "CALL *pv_ops.irq.save_fl" with "PUSHF;POP". That changed the stack layout between the PUSHF and the POP, so unwinding from an interrupt which occurred between those two instructions would fail. Part of the agreed upon solution was to rework the custom paravirt patching code to use alternatives instead, since objtool already knows how to read alternatives (and converging runtime patching infrastructure is always a good thing anyway). But the main problem still remains, which is that runtime patching can change the stack layout. Making stack layout changes in alternatives was disallowed with commit 7117f16bf460 ("objtool: Fix ORC vs alternatives"), but now that paravirt is going to be doing it, it needs to be supported. One way to do so would be to modify the ORC table when the code gets patched. But ORC is simple -- a good thing! -- and it's best to leave it alone. Instead, support stack layout changes by "flattening" all possible stack states (CFI) from parallel alternative code streams into a single set of linear states. The only necessary limitation is that CFI conflicts are disallowed at all possible instruction boundaries. For example, this scenario is allowed: Alt1 Alt2 Alt3 0x00 CALL *pv_ops.save_fl CALL xen_save_fl PUSHF 0x01 POP %RAX 0x02 NOP ... 0x05 NOP ... 0x07 <insn> The unwind information for offset-0x00 is identical for all 3 alternatives. Similarly offset-0x05 and higher also are identical (and the same as 0x00). However offset-0x01 has deviating CFI, but that is only relevant for Alt3, neither of the other alternative instruction streams will ever hit that offset. This scenario is NOT allowed: Alt1 Alt2 0x00 CALL *pv_ops.save_fl PUSHF 0x01 NOP6 ... 0x07 NOP POP %RAX The problem here is that offset-0x7, which is an instruction boundary in both possible instruction patch streams, has two conflicting stack layouts. [ The above examples were stolen from Peter Zijlstra. ] The new flattened CFI array is used both for the detection of conflicts (like the second example above) and the generation of linear ORC entries. BTW, another benefit of these changes is that, thanks to some related cleanups (new fake nops and alt_group struct) objtool can finally be rid of fake jumps, which were a constant source of headaches. [1] https://lkml.kernel.org/r/20201111170536.arx2zbn4ngvjoov7@treble Cc: Shinichiro Kawasaki <shinichiro.kawasaki@wdc.com> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
2020-12-19 04:26:21 +08:00
if (propagate_alt_cfi(file, insn))
return 1;
objtool: Fix ORC vs alternatives Jann reported that (for instance) entry_64.o:general_protection has very odd ORC data: 0000000000000f40 <general_protection>: #######sp:sp+8 bp:(und) type:iret end:0 f40: 90 nop #######sp:(und) bp:(und) type:call end:0 f41: 90 nop f42: 90 nop #######sp:sp+8 bp:(und) type:iret end:0 f43: e8 a8 01 00 00 callq 10f0 <error_entry> #######sp:sp+0 bp:(und) type:regs end:0 f48: f6 84 24 88 00 00 00 testb $0x3,0x88(%rsp) f4f: 03 f50: 74 00 je f52 <general_protection+0x12> f52: 48 89 e7 mov %rsp,%rdi f55: 48 8b 74 24 78 mov 0x78(%rsp),%rsi f5a: 48 c7 44 24 78 ff ff movq $0xffffffffffffffff,0x78(%rsp) f61: ff ff f63: e8 00 00 00 00 callq f68 <general_protection+0x28> f68: e9 73 02 00 00 jmpq 11e0 <error_exit> #######sp:(und) bp:(und) type:call end:0 f6d: 0f 1f 00 nopl (%rax) Note the entry at 0xf41. Josh found this was the result of commit: 764eef4b109a ("objtool: Rewrite alt->skip_orig") Due to the early return in validate_branch() we no longer set insn->cfi of the original instruction stream (the NOPs at 0xf41 and 0xf42) and we'll end up with the above weirdness. In other discussions we realized alternatives should be ORC invariant; that is, due to there being only a single ORC table, it must be valid for all alternatives. The easiest way to ensure this is to not allow any stack modifications in alternatives. When we enforce this latter observation, we get the property that the whole alternative must have the same CFI, which we can employ to fix the former report. Fixes: 764eef4b109a ("objtool: Rewrite alt->skip_orig") Reported-by: Jann Horn <jannh@google.com> Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Reviewed-by: Miroslav Benes <mbenes@suse.cz> Acked-by: Josh Poimboeuf <jpoimboe@redhat.com> Link: https://lkml.kernel.org/r/20200428191659.499074346@infradead.org
2020-04-29 01:37:01 +08:00
if (!insn->ignore_alts && !list_empty(&insn->alts)) {
bool skip_orig = false;
list_for_each_entry(alt, &insn->alts, list) {
if (alt->skip_orig)
skip_orig = true;
ret = validate_branch(file, func, alt->insn, state);
objtool: Add --backtrace support For when you want to know the path that reached your fail state: $ ./objtool check --no-fp --backtrace arch/x86/lib/usercopy_64.o arch/x86/lib/usercopy_64.o: warning: objtool: .altinstr_replacement+0x3: UACCESS disable without MEMOPs: __clear_user() arch/x86/lib/usercopy_64.o: warning: objtool: __clear_user()+0x3a: (alt) arch/x86/lib/usercopy_64.o: warning: objtool: __clear_user()+0x2e: (branch) arch/x86/lib/usercopy_64.o: warning: objtool: __clear_user()+0x18: (branch) arch/x86/lib/usercopy_64.o: warning: objtool: .altinstr_replacement+0xffffffffffffffff: (branch) arch/x86/lib/usercopy_64.o: warning: objtool: __clear_user()+0x5: (alt) arch/x86/lib/usercopy_64.o: warning: objtool: __clear_user()+0x0: <=== (func) 0000000000000000 <__clear_user>: 0: e8 00 00 00 00 callq 5 <__clear_user+0x5> 1: R_X86_64_PLT32 __fentry__-0x4 5: 90 nop 6: 90 nop 7: 90 nop 8: 48 89 f0 mov %rsi,%rax b: 48 c1 ee 03 shr $0x3,%rsi f: 83 e0 07 and $0x7,%eax 12: 48 89 f1 mov %rsi,%rcx 15: 48 85 c9 test %rcx,%rcx 18: 74 0f je 29 <__clear_user+0x29> 1a: 48 c7 07 00 00 00 00 movq $0x0,(%rdi) 21: 48 83 c7 08 add $0x8,%rdi 25: ff c9 dec %ecx 27: 75 f1 jne 1a <__clear_user+0x1a> 29: 48 89 c1 mov %rax,%rcx 2c: 85 c9 test %ecx,%ecx 2e: 74 0a je 3a <__clear_user+0x3a> 30: c6 07 00 movb $0x0,(%rdi) 33: 48 ff c7 inc %rdi 36: ff c9 dec %ecx 38: 75 f6 jne 30 <__clear_user+0x30> 3a: 90 nop 3b: 90 nop 3c: 90 nop 3d: 48 89 c8 mov %rcx,%rax 40: c3 retq Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Acked-by: Josh Poimboeuf <jpoimboe@redhat.com> Cc: Borislav Petkov <bp@alien8.de> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Thomas Gleixner <tglx@linutronix.de> Signed-off-by: Ingo Molnar <mingo@kernel.org>
2019-03-01 18:15:49 +08:00
if (ret) {
if (opts.backtrace)
objtool: Add --backtrace support For when you want to know the path that reached your fail state: $ ./objtool check --no-fp --backtrace arch/x86/lib/usercopy_64.o arch/x86/lib/usercopy_64.o: warning: objtool: .altinstr_replacement+0x3: UACCESS disable without MEMOPs: __clear_user() arch/x86/lib/usercopy_64.o: warning: objtool: __clear_user()+0x3a: (alt) arch/x86/lib/usercopy_64.o: warning: objtool: __clear_user()+0x2e: (branch) arch/x86/lib/usercopy_64.o: warning: objtool: __clear_user()+0x18: (branch) arch/x86/lib/usercopy_64.o: warning: objtool: .altinstr_replacement+0xffffffffffffffff: (branch) arch/x86/lib/usercopy_64.o: warning: objtool: __clear_user()+0x5: (alt) arch/x86/lib/usercopy_64.o: warning: objtool: __clear_user()+0x0: <=== (func) 0000000000000000 <__clear_user>: 0: e8 00 00 00 00 callq 5 <__clear_user+0x5> 1: R_X86_64_PLT32 __fentry__-0x4 5: 90 nop 6: 90 nop 7: 90 nop 8: 48 89 f0 mov %rsi,%rax b: 48 c1 ee 03 shr $0x3,%rsi f: 83 e0 07 and $0x7,%eax 12: 48 89 f1 mov %rsi,%rcx 15: 48 85 c9 test %rcx,%rcx 18: 74 0f je 29 <__clear_user+0x29> 1a: 48 c7 07 00 00 00 00 movq $0x0,(%rdi) 21: 48 83 c7 08 add $0x8,%rdi 25: ff c9 dec %ecx 27: 75 f1 jne 1a <__clear_user+0x1a> 29: 48 89 c1 mov %rax,%rcx 2c: 85 c9 test %ecx,%ecx 2e: 74 0a je 3a <__clear_user+0x3a> 30: c6 07 00 movb $0x0,(%rdi) 33: 48 ff c7 inc %rdi 36: ff c9 dec %ecx 38: 75 f6 jne 30 <__clear_user+0x30> 3a: 90 nop 3b: 90 nop 3c: 90 nop 3d: 48 89 c8 mov %rcx,%rax 40: c3 retq Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Acked-by: Josh Poimboeuf <jpoimboe@redhat.com> Cc: Borislav Petkov <bp@alien8.de> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Thomas Gleixner <tglx@linutronix.de> Signed-off-by: Ingo Molnar <mingo@kernel.org>
2019-03-01 18:15:49 +08:00
BT_FUNC("(alt)", insn);
return ret;
}
}
if (skip_orig)
return 0;
}
if (handle_insn_ops(insn, next_insn, &state))
return 1;
switch (insn->type) {
case INSN_RETURN:
return validate_return(func, insn, &state);
case INSN_CALL:
case INSN_CALL_DYNAMIC:
ret = validate_call(file, insn, &state);
if (ret)
return ret;
if (opts.stackval && func && !is_fentry_call(insn) &&
!has_valid_stack_frame(&state)) {
WARN_FUNC("call without frame pointer save/setup",
sec, insn->offset);
return 1;
}
if (insn->dead_end)
return 0;
break;
case INSN_JUMP_CONDITIONAL:
case INSN_JUMP_UNCONDITIONAL:
if (is_sibling_call(insn)) {
ret = validate_sibling_call(file, insn, &state);
if (ret)
return ret;
} else if (insn->jump_dest) {
ret = validate_branch(file, func,
insn->jump_dest, state);
objtool: Add --backtrace support For when you want to know the path that reached your fail state: $ ./objtool check --no-fp --backtrace arch/x86/lib/usercopy_64.o arch/x86/lib/usercopy_64.o: warning: objtool: .altinstr_replacement+0x3: UACCESS disable without MEMOPs: __clear_user() arch/x86/lib/usercopy_64.o: warning: objtool: __clear_user()+0x3a: (alt) arch/x86/lib/usercopy_64.o: warning: objtool: __clear_user()+0x2e: (branch) arch/x86/lib/usercopy_64.o: warning: objtool: __clear_user()+0x18: (branch) arch/x86/lib/usercopy_64.o: warning: objtool: .altinstr_replacement+0xffffffffffffffff: (branch) arch/x86/lib/usercopy_64.o: warning: objtool: __clear_user()+0x5: (alt) arch/x86/lib/usercopy_64.o: warning: objtool: __clear_user()+0x0: <=== (func) 0000000000000000 <__clear_user>: 0: e8 00 00 00 00 callq 5 <__clear_user+0x5> 1: R_X86_64_PLT32 __fentry__-0x4 5: 90 nop 6: 90 nop 7: 90 nop 8: 48 89 f0 mov %rsi,%rax b: 48 c1 ee 03 shr $0x3,%rsi f: 83 e0 07 and $0x7,%eax 12: 48 89 f1 mov %rsi,%rcx 15: 48 85 c9 test %rcx,%rcx 18: 74 0f je 29 <__clear_user+0x29> 1a: 48 c7 07 00 00 00 00 movq $0x0,(%rdi) 21: 48 83 c7 08 add $0x8,%rdi 25: ff c9 dec %ecx 27: 75 f1 jne 1a <__clear_user+0x1a> 29: 48 89 c1 mov %rax,%rcx 2c: 85 c9 test %ecx,%ecx 2e: 74 0a je 3a <__clear_user+0x3a> 30: c6 07 00 movb $0x0,(%rdi) 33: 48 ff c7 inc %rdi 36: ff c9 dec %ecx 38: 75 f6 jne 30 <__clear_user+0x30> 3a: 90 nop 3b: 90 nop 3c: 90 nop 3d: 48 89 c8 mov %rcx,%rax 40: c3 retq Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Acked-by: Josh Poimboeuf <jpoimboe@redhat.com> Cc: Borislav Petkov <bp@alien8.de> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Thomas Gleixner <tglx@linutronix.de> Signed-off-by: Ingo Molnar <mingo@kernel.org>
2019-03-01 18:15:49 +08:00
if (ret) {
if (opts.backtrace)
objtool: Add --backtrace support For when you want to know the path that reached your fail state: $ ./objtool check --no-fp --backtrace arch/x86/lib/usercopy_64.o arch/x86/lib/usercopy_64.o: warning: objtool: .altinstr_replacement+0x3: UACCESS disable without MEMOPs: __clear_user() arch/x86/lib/usercopy_64.o: warning: objtool: __clear_user()+0x3a: (alt) arch/x86/lib/usercopy_64.o: warning: objtool: __clear_user()+0x2e: (branch) arch/x86/lib/usercopy_64.o: warning: objtool: __clear_user()+0x18: (branch) arch/x86/lib/usercopy_64.o: warning: objtool: .altinstr_replacement+0xffffffffffffffff: (branch) arch/x86/lib/usercopy_64.o: warning: objtool: __clear_user()+0x5: (alt) arch/x86/lib/usercopy_64.o: warning: objtool: __clear_user()+0x0: <=== (func) 0000000000000000 <__clear_user>: 0: e8 00 00 00 00 callq 5 <__clear_user+0x5> 1: R_X86_64_PLT32 __fentry__-0x4 5: 90 nop 6: 90 nop 7: 90 nop 8: 48 89 f0 mov %rsi,%rax b: 48 c1 ee 03 shr $0x3,%rsi f: 83 e0 07 and $0x7,%eax 12: 48 89 f1 mov %rsi,%rcx 15: 48 85 c9 test %rcx,%rcx 18: 74 0f je 29 <__clear_user+0x29> 1a: 48 c7 07 00 00 00 00 movq $0x0,(%rdi) 21: 48 83 c7 08 add $0x8,%rdi 25: ff c9 dec %ecx 27: 75 f1 jne 1a <__clear_user+0x1a> 29: 48 89 c1 mov %rax,%rcx 2c: 85 c9 test %ecx,%ecx 2e: 74 0a je 3a <__clear_user+0x3a> 30: c6 07 00 movb $0x0,(%rdi) 33: 48 ff c7 inc %rdi 36: ff c9 dec %ecx 38: 75 f6 jne 30 <__clear_user+0x30> 3a: 90 nop 3b: 90 nop 3c: 90 nop 3d: 48 89 c8 mov %rcx,%rax 40: c3 retq Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Acked-by: Josh Poimboeuf <jpoimboe@redhat.com> Cc: Borislav Petkov <bp@alien8.de> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Thomas Gleixner <tglx@linutronix.de> Signed-off-by: Ingo Molnar <mingo@kernel.org>
2019-03-01 18:15:49 +08:00
BT_FUNC("(branch)", insn);
return ret;
}
}
if (insn->type == INSN_JUMP_UNCONDITIONAL)
return 0;
break;
case INSN_JUMP_DYNAMIC:
case INSN_JUMP_DYNAMIC_CONDITIONAL:
if (is_sibling_call(insn)) {
ret = validate_sibling_call(file, insn, &state);
if (ret)
return ret;
}
if (insn->type == INSN_JUMP_DYNAMIC)
return 0;
break;
case INSN_CONTEXT_SWITCH:
if (func && (!next_insn || !next_insn->hint)) {
WARN_FUNC("unsupported instruction in callable function",
sec, insn->offset);
return 1;
}
return 0;
case INSN_STAC:
if (state.uaccess) {
WARN_FUNC("recursive UACCESS enable", sec, insn->offset);
return 1;
}
state.uaccess = true;
break;
case INSN_CLAC:
if (!state.uaccess && func) {
WARN_FUNC("redundant UACCESS disable", sec, insn->offset);
return 1;
}
if (func_uaccess_safe(func) && !state.uaccess_stack) {
WARN_FUNC("UACCESS-safe disables UACCESS", sec, insn->offset);
return 1;
}
state.uaccess = false;
break;
case INSN_STD:
if (state.df) {
WARN_FUNC("recursive STD", sec, insn->offset);
return 1;
}
state.df = true;
break;
case INSN_CLD:
if (!state.df && func) {
WARN_FUNC("redundant CLD", sec, insn->offset);
return 1;
}
state.df = false;
break;
default:
break;
}
if (insn->dead_end)
return 0;
if (!next_insn) {
if (state.cfi.cfa.base == CFI_UNDEFINED)
return 0;
WARN("%s: unexpected end of section", sec->name);
return 1;
}
prev_insn = insn;
insn = next_insn;
}
return 0;
}
static int validate_unwind_hints(struct objtool_file *file, struct section *sec)
{
struct instruction *insn;
struct insn_state state;
int ret, warnings = 0;
if (!file->hints)
return 0;
init_insn_state(file, &state, sec);
if (sec) {
insn = find_insn(file, sec, 0);
if (!insn)
return 0;
} else {
insn = list_first_entry(&file->insn_list, typeof(*insn), list);
}
while (&insn->list != &file->insn_list && (!sec || insn->sec == sec)) {
if (insn->hint && !insn->visited && !insn->ignore) {
ret = validate_branch(file, insn->func, insn, state);
if (ret && opts.backtrace)
objtool: Add --backtrace support For when you want to know the path that reached your fail state: $ ./objtool check --no-fp --backtrace arch/x86/lib/usercopy_64.o arch/x86/lib/usercopy_64.o: warning: objtool: .altinstr_replacement+0x3: UACCESS disable without MEMOPs: __clear_user() arch/x86/lib/usercopy_64.o: warning: objtool: __clear_user()+0x3a: (alt) arch/x86/lib/usercopy_64.o: warning: objtool: __clear_user()+0x2e: (branch) arch/x86/lib/usercopy_64.o: warning: objtool: __clear_user()+0x18: (branch) arch/x86/lib/usercopy_64.o: warning: objtool: .altinstr_replacement+0xffffffffffffffff: (branch) arch/x86/lib/usercopy_64.o: warning: objtool: __clear_user()+0x5: (alt) arch/x86/lib/usercopy_64.o: warning: objtool: __clear_user()+0x0: <=== (func) 0000000000000000 <__clear_user>: 0: e8 00 00 00 00 callq 5 <__clear_user+0x5> 1: R_X86_64_PLT32 __fentry__-0x4 5: 90 nop 6: 90 nop 7: 90 nop 8: 48 89 f0 mov %rsi,%rax b: 48 c1 ee 03 shr $0x3,%rsi f: 83 e0 07 and $0x7,%eax 12: 48 89 f1 mov %rsi,%rcx 15: 48 85 c9 test %rcx,%rcx 18: 74 0f je 29 <__clear_user+0x29> 1a: 48 c7 07 00 00 00 00 movq $0x0,(%rdi) 21: 48 83 c7 08 add $0x8,%rdi 25: ff c9 dec %ecx 27: 75 f1 jne 1a <__clear_user+0x1a> 29: 48 89 c1 mov %rax,%rcx 2c: 85 c9 test %ecx,%ecx 2e: 74 0a je 3a <__clear_user+0x3a> 30: c6 07 00 movb $0x0,(%rdi) 33: 48 ff c7 inc %rdi 36: ff c9 dec %ecx 38: 75 f6 jne 30 <__clear_user+0x30> 3a: 90 nop 3b: 90 nop 3c: 90 nop 3d: 48 89 c8 mov %rcx,%rax 40: c3 retq Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Acked-by: Josh Poimboeuf <jpoimboe@redhat.com> Cc: Borislav Petkov <bp@alien8.de> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Thomas Gleixner <tglx@linutronix.de> Signed-off-by: Ingo Molnar <mingo@kernel.org>
2019-03-01 18:15:49 +08:00
BT_FUNC("<=== (hint)", insn);
warnings += ret;
}
insn = list_next_entry(insn, list);
}
return warnings;
}
static int validate_retpoline(struct objtool_file *file)
{
struct instruction *insn;
int warnings = 0;
for_each_insn(file, insn) {
if (insn->type != INSN_JUMP_DYNAMIC &&
insn->type != INSN_CALL_DYNAMIC)
continue;
if (insn->retpoline_safe)
continue;
/*
* .init.text code is ran before userspace and thus doesn't
* strictly need retpolines, except for modules which are
* loaded late, they very much do need retpoline in their
* .init.text
*/
if (!strcmp(insn->sec->name, ".init.text") && !opts.module)
continue;
WARN_FUNC("indirect %s found in RETPOLINE build",
insn->sec, insn->offset,
insn->type == INSN_JUMP_DYNAMIC ? "jump" : "call");
warnings++;
}
return warnings;
}
static bool is_kasan_insn(struct instruction *insn)
{
return (insn->type == INSN_CALL &&
!strcmp(insn->call_dest->name, "__asan_handle_no_return"));
}
static bool is_ubsan_insn(struct instruction *insn)
{
return (insn->type == INSN_CALL &&
!strcmp(insn->call_dest->name,
"__ubsan_handle_builtin_unreachable"));
}
static bool ignore_unreachable_insn(struct objtool_file *file, struct instruction *insn)
{
int i;
struct instruction *prev_insn;
if (insn->ignore || insn->type == INSN_NOP || insn->type == INSN_TRAP)
return true;
/*
* Ignore alternative replacement instructions. This can happen
* when a whitelisted function uses one of the ALTERNATIVE macros.
*/
if (!strcmp(insn->sec->name, ".altinstr_replacement") ||
!strcmp(insn->sec->name, ".altinstr_aux"))
return true;
/*
* Whole archive runs might encounter dead code from weak symbols.
* This is where the linker will have dropped the weak symbol in
* favour of a regular symbol, but leaves the code in place.
*
* In this case we'll find a piece of code (whole function) that is not
* covered by a !section symbol. Ignore them.
*/
if (opts.link && !insn->func) {
int size = find_symbol_hole_containing(insn->sec, insn->offset);
unsigned long end = insn->offset + size;
if (!size) /* not a hole */
return false;
if (size < 0) /* hole until the end */
return true;
sec_for_each_insn_continue(file, insn) {
/*
* If we reach a visited instruction at or before the
* end of the hole, ignore the unreachable.
*/
if (insn->visited)
return true;
if (insn->offset >= end)
break;
/*
* If this hole jumps to a .cold function, mark it ignore too.
*/
if (insn->jump_dest && insn->jump_dest->func &&
strstr(insn->jump_dest->func->name, ".cold")) {
struct instruction *dest = insn->jump_dest;
func_for_each_insn(file, dest->func, dest)
dest->ignore = true;
}
}
return false;
}
if (!insn->func)
return false;
if (insn->func->static_call_tramp)
return true;
/*
* CONFIG_UBSAN_TRAP inserts a UD2 when it sees
* __builtin_unreachable(). The BUG() macro has an unreachable() after
* the UD2, which causes GCC's undefined trap logic to emit another UD2
* (or occasionally a JMP to UD2).
*
* It may also insert a UD2 after calling a __noreturn function.
*/
prev_insn = list_prev_entry(insn, list);
if ((prev_insn->dead_end || dead_end_function(file, prev_insn->call_dest)) &&
(insn->type == INSN_BUG ||
(insn->type == INSN_JUMP_UNCONDITIONAL &&
insn->jump_dest && insn->jump_dest->type == INSN_BUG)))
return true;
/*
* Check if this (or a subsequent) instruction is related to
* CONFIG_UBSAN or CONFIG_KASAN.
*
* End the search at 5 instructions to avoid going into the weeds.
*/
for (i = 0; i < 5; i++) {
if (is_kasan_insn(insn) || is_ubsan_insn(insn))
return true;
if (insn->type == INSN_JUMP_UNCONDITIONAL) {
if (insn->jump_dest &&
insn->jump_dest->func == insn->func) {
insn = insn->jump_dest;
continue;
}
break;
}
if (insn->offset + insn->len >= insn->func->offset + insn->func->len)
break;
insn = list_next_entry(insn, list);
}
return false;
}
static int validate_symbol(struct objtool_file *file, struct section *sec,
struct symbol *sym, struct insn_state *state)
{
struct instruction *insn;
int ret;
if (!sym->len) {
WARN("%s() is missing an ELF size annotation", sym->name);
return 1;
}
if (sym->pfunc != sym || sym->alias != sym)
return 0;
insn = find_insn(file, sec, sym->offset);
if (!insn || insn->ignore || insn->visited)
return 0;
state->uaccess = sym->uaccess_safe;
ret = validate_branch(file, insn->func, insn, *state);
if (ret && opts.backtrace)
BT_FUNC("<=== (sym)", insn);
return ret;
}
static int validate_section(struct objtool_file *file, struct section *sec)
{
struct insn_state state;
struct symbol *func;
int warnings = 0;
list_for_each_entry(func, &sec->symbol_list, list) {
if (func->type != STT_FUNC)
continue;
init_insn_state(file, &state, sec);
set_func_state(&state.cfi);
warnings += validate_symbol(file, sec, func, &state);
}
return warnings;
}
static int validate_noinstr_sections(struct objtool_file *file)
{
struct section *sec;
int warnings = 0;
sec = find_section_by_name(file->elf, ".noinstr.text");
if (sec) {
warnings += validate_section(file, sec);
warnings += validate_unwind_hints(file, sec);
}
sec = find_section_by_name(file->elf, ".entry.text");
if (sec) {
warnings += validate_section(file, sec);
warnings += validate_unwind_hints(file, sec);
}
return warnings;
}
static int validate_functions(struct objtool_file *file)
{
struct section *sec;
int warnings = 0;
for_each_sec(file, sec) {
if (!(sec->sh.sh_flags & SHF_EXECINSTR))
continue;
warnings += validate_section(file, sec);
}
return warnings;
}
static void mark_endbr_used(struct instruction *insn)
{
if (!list_empty(&insn->call_node))
list_del_init(&insn->call_node);
}
static int validate_ibt_insn(struct objtool_file *file, struct instruction *insn)
{
struct instruction *dest;
struct reloc *reloc;
unsigned long off;
int warnings = 0;
/*
* Looking for function pointer load relocations. Ignore
* direct/indirect branches:
*/
switch (insn->type) {
case INSN_CALL:
case INSN_CALL_DYNAMIC:
case INSN_JUMP_CONDITIONAL:
case INSN_JUMP_UNCONDITIONAL:
case INSN_JUMP_DYNAMIC:
case INSN_JUMP_DYNAMIC_CONDITIONAL:
case INSN_RETURN:
case INSN_NOP:
return 0;
default:
break;
}
for (reloc = insn_reloc(file, insn);
reloc;
reloc = find_reloc_by_dest_range(file->elf, insn->sec,
reloc->offset + 1,
(insn->offset + insn->len) - (reloc->offset + 1))) {
/*
* static_call_update() references the trampoline, which
* doesn't have (or need) ENDBR. Skip warning in that case.
*/
if (reloc->sym->static_call_tramp)
continue;
off = reloc->sym->offset;
if (reloc->type == R_X86_64_PC32 || reloc->type == R_X86_64_PLT32)
off += arch_dest_reloc_offset(reloc->addend);
else
off += reloc->addend;
dest = find_insn(file, reloc->sym->sec, off);
if (!dest)
continue;
if (dest->type == INSN_ENDBR) {
mark_endbr_used(dest);
continue;
}
if (dest->func && dest->func == insn->func) {
/*
* Anything from->to self is either _THIS_IP_ or
* IRET-to-self.
*
* There is no sane way to annotate _THIS_IP_ since the
* compiler treats the relocation as a constant and is
* happy to fold in offsets, skewing any annotation we
* do, leading to vast amounts of false-positives.
*
* There's also compiler generated _THIS_IP_ through
* KCOV and such which we have no hope of annotating.
*
* As such, blanket accept self-references without
* issue.
*/
continue;
}
if (dest->noendbr)
continue;
WARN_FUNC("relocation to !ENDBR: %s",
insn->sec, insn->offset,
offstr(dest->sec, dest->offset));
warnings++;
}
return warnings;
}
static int validate_ibt_data_reloc(struct objtool_file *file,
struct reloc *reloc)
{
struct instruction *dest;
dest = find_insn(file, reloc->sym->sec,
reloc->sym->offset + reloc->addend);
if (!dest)
return 0;
if (dest->type == INSN_ENDBR) {
mark_endbr_used(dest);
return 0;
}
if (dest->noendbr)
return 0;
WARN_FUNC("data relocation to !ENDBR: %s",
reloc->sec->base, reloc->offset,
offstr(dest->sec, dest->offset));
return 1;
}
/*
* Validate IBT rules and remove used ENDBR instructions from the seal list.
* Unused ENDBR instructions will be annotated for sealing (i.e., replaced with
* NOPs) later, in create_ibt_endbr_seal_sections().
*/
static int validate_ibt(struct objtool_file *file)
{
struct section *sec;
struct reloc *reloc;
struct instruction *insn;
int warnings = 0;
for_each_insn(file, insn)
warnings += validate_ibt_insn(file, insn);
for_each_sec(file, sec) {
/* Already done by validate_ibt_insn() */
if (sec->sh.sh_flags & SHF_EXECINSTR)
continue;
if (!sec->reloc)
continue;
/*
* These sections can reference text addresses, but not with
* the intent to indirect branch to them.
*/
if (!strncmp(sec->name, ".discard", 8) ||
!strncmp(sec->name, ".debug", 6) ||
!strcmp(sec->name, ".altinstructions") ||
!strcmp(sec->name, ".ibt_endbr_seal") ||
!strcmp(sec->name, ".orc_unwind_ip") ||
!strcmp(sec->name, ".parainstructions") ||
!strcmp(sec->name, ".retpoline_sites") ||
!strcmp(sec->name, ".smp_locks") ||
!strcmp(sec->name, ".static_call_sites") ||
!strcmp(sec->name, "_error_injection_whitelist") ||
!strcmp(sec->name, "_kprobe_blacklist") ||
!strcmp(sec->name, "__bug_table") ||
!strcmp(sec->name, "__ex_table") ||
!strcmp(sec->name, "__jump_table") ||
!strcmp(sec->name, "__mcount_loc") ||
!strcmp(sec->name, "__tracepoints"))
continue;
list_for_each_entry(reloc, &sec->reloc->reloc_list, list)
warnings += validate_ibt_data_reloc(file, reloc);
}
return warnings;
}
static int validate_sls(struct objtool_file *file)
{
struct instruction *insn, *next_insn;
int warnings = 0;
for_each_insn(file, insn) {
next_insn = next_insn_same_sec(file, insn);
if (insn->retpoline_safe)
continue;
switch (insn->type) {
case INSN_RETURN:
if (!next_insn || next_insn->type != INSN_TRAP) {
WARN_FUNC("missing int3 after ret",
insn->sec, insn->offset);
warnings++;
}
break;
case INSN_JUMP_DYNAMIC:
if (!next_insn || next_insn->type != INSN_TRAP) {
WARN_FUNC("missing int3 after indirect jump",
insn->sec, insn->offset);
warnings++;
}
break;
default:
break;
}
}
return warnings;
}
static int validate_reachable_instructions(struct objtool_file *file)
{
struct instruction *insn;
if (file->ignore_unreachables)
return 0;
for_each_insn(file, insn) {
if (insn->visited || ignore_unreachable_insn(file, insn))
continue;
WARN_FUNC("unreachable instruction", insn->sec, insn->offset);
return 1;
}
return 0;
}
int check(struct objtool_file *file)
{
int ret, warnings = 0;
arch_initial_func_cfi_state(&initial_func_cfi);
init_cfi_state(&init_cfi);
init_cfi_state(&func_cfi);
set_func_state(&func_cfi);
if (!cfi_hash_alloc(1UL << (file->elf->symbol_bits - 3)))
goto out;
cfi_hash_add(&init_cfi);
cfi_hash_add(&func_cfi);
ret = decode_sections(file);
if (ret < 0)
goto out;
warnings += ret;
if (list_empty(&file->insn_list))
goto out;
if (opts.retpoline) {
ret = validate_retpoline(file);
if (ret < 0)
return ret;
warnings += ret;
}
if (opts.stackval || opts.orc || opts.uaccess) {
ret = validate_functions(file);
if (ret < 0)
goto out;
warnings += ret;
ret = validate_unwind_hints(file, NULL);
if (ret < 0)
goto out;
warnings += ret;
if (!warnings) {
ret = validate_reachable_instructions(file);
if (ret < 0)
goto out;
warnings += ret;
}
} else if (opts.noinstr) {
ret = validate_noinstr_sections(file);
if (ret < 0)
goto out;
warnings += ret;
}
if (opts.ibt) {
ret = validate_ibt(file);
if (ret < 0)
goto out;
warnings += ret;
}
if (opts.sls) {
ret = validate_sls(file);
if (ret < 0)
goto out;
warnings += ret;
}
if (opts.static_call) {
ret = create_static_call_sections(file);
if (ret < 0)
goto out;
warnings += ret;
}
if (opts.retpoline) {
ret = create_retpoline_sites_sections(file);
if (ret < 0)
goto out;
warnings += ret;
ret = create_return_sites_sections(file);
if (ret < 0)
goto out;
warnings += ret;
}
if (opts.mcount) {
ret = create_mcount_loc_sections(file);
if (ret < 0)
goto out;
warnings += ret;
}
if (opts.ibt) {
ret = create_ibt_endbr_seal_sections(file);
if (ret < 0)
goto out;
warnings += ret;
}
if (opts.orc && !list_empty(&file->insn_list)) {
ret = orc_create(file);
if (ret < 0)
goto out;
warnings += ret;
}
if (opts.stats) {
printf("nr_insns_visited: %ld\n", nr_insns_visited);
printf("nr_cfi: %ld\n", nr_cfi);
printf("nr_cfi_reused: %ld\n", nr_cfi_reused);
printf("nr_cfi_cache: %ld\n", nr_cfi_cache);
}
out:
/*
* For now, don't fail the kernel build on fatal warnings. These
* errors are still fairly common due to the growing matrix of
* supported toolchains and their recent pace of change.
*/
return 0;
}