mirror of
git://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git
synced 2026-03-22 07:27:12 +08:00
perf script: Fix script_fetch_insn for more than just x86
The script_fetch_insn code was only supported on natively running x86. Implement a crude elf_machine_max_instruction_length function and use to give an instruction length on more than just x86. Use the ELF machine to determine the length to use to support cross-architecture development. Signed-off-by: Ian Rogers <irogers@google.com> Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Andi Kleen <ak@linux.intel.com> Cc: Ingo Molnar <mingo@redhat.com> Cc: James Clark <james.clark@linaro.org> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Leo Yan <leo.yan@arm.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Shimin Guo <shimin.guo@skydio.com> Cc: Yujie Liu <yujie.liu@intel.com> [ Conditionally define EM_CSKY and EM_LOONGARCH for older distros ] Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
This commit is contained in:
committed by
Arnaldo Carvalho de Melo
parent
0a6fb66047
commit
7ce6dfc603
@@ -14,6 +14,5 @@ perf-util-y += iostat.o
|
||||
perf-util-$(CONFIG_LOCAL_LIBUNWIND) += unwind-libunwind.o
|
||||
|
||||
perf-util-y += auxtrace.o
|
||||
perf-util-y += archinsn.o
|
||||
perf-util-y += intel-pt.o
|
||||
perf-util-y += intel-bts.o
|
||||
|
||||
@@ -1,27 +0,0 @@
|
||||
// SPDX-License-Identifier: GPL-2.0
|
||||
#include "archinsn.h"
|
||||
#include "event.h"
|
||||
#include "machine.h"
|
||||
#include "thread.h"
|
||||
#include "symbol.h"
|
||||
#include "../../../../arch/x86/include/asm/insn.h"
|
||||
|
||||
void arch_fetch_insn(struct perf_sample *sample,
|
||||
struct thread *thread,
|
||||
struct machine *machine)
|
||||
{
|
||||
struct insn insn;
|
||||
int len, ret;
|
||||
bool is64bit = false;
|
||||
|
||||
if (!sample->ip)
|
||||
return;
|
||||
len = thread__memcpy(thread, machine, sample->insn, sample->ip, sizeof(sample->insn), &is64bit);
|
||||
if (len <= 0)
|
||||
return;
|
||||
|
||||
ret = insn_decode(&insn, sample->insn, len,
|
||||
is64bit ? INSN_MODE_64 : INSN_MODE_32);
|
||||
if (ret >= 0 && insn.length <= len)
|
||||
sample->insn_len = insn.length;
|
||||
}
|
||||
@@ -37,7 +37,6 @@
|
||||
#include "ui/ui.h"
|
||||
#include "print_binary.h"
|
||||
#include "print_insn.h"
|
||||
#include "archinsn.h"
|
||||
#include <linux/bitmap.h>
|
||||
#include <linux/compiler.h>
|
||||
#include <linux/kernel.h>
|
||||
@@ -90,7 +89,6 @@ static bool print_flags;
|
||||
static const char *cpu_list;
|
||||
static DECLARE_BITMAP(cpu_bitmap, MAX_NR_CPUS);
|
||||
static int max_blocks;
|
||||
static bool native_arch;
|
||||
static struct dlfilter *dlfilter;
|
||||
static int dlargc;
|
||||
static char **dlargv;
|
||||
@@ -1627,7 +1625,7 @@ static int perf_sample__fprintf_insn(struct perf_sample *sample,
|
||||
{
|
||||
int printed = 0;
|
||||
|
||||
script_fetch_insn(sample, thread, machine, native_arch);
|
||||
perf_sample__fetch_insn(sample, thread, machine);
|
||||
|
||||
if (PRINT_FIELD(INSNLEN))
|
||||
printed += fprintf(fp, " ilen: %d", sample->insn_len);
|
||||
@@ -4034,7 +4032,6 @@ int cmd_script(int argc, const char **argv)
|
||||
.set = false,
|
||||
.default_no_sample = true,
|
||||
};
|
||||
struct utsname uts;
|
||||
char *script_path = NULL;
|
||||
const char *dlfilter_file = NULL;
|
||||
const char **__argv;
|
||||
@@ -4456,17 +4453,6 @@ script_found:
|
||||
if (symbol__init(env) < 0)
|
||||
goto out_delete;
|
||||
|
||||
uname(&uts);
|
||||
if (data.is_pipe) { /* Assume pipe_mode indicates native_arch */
|
||||
native_arch = true;
|
||||
} else if (env->arch) {
|
||||
if (!strcmp(uts.machine, env->arch))
|
||||
native_arch = true;
|
||||
else if (!strcmp(uts.machine, "x86_64") &&
|
||||
!strcmp(env->arch, "i386"))
|
||||
native_arch = true;
|
||||
}
|
||||
|
||||
script.session = session;
|
||||
script__setup_sample_type(&script);
|
||||
|
||||
|
||||
@@ -93,7 +93,7 @@ static PyObject *perf_sample_insn(PyObject *obj, PyObject *args)
|
||||
if (c->sample->ip && !c->sample->insn_len && thread__maps(c->al->thread)) {
|
||||
struct machine *machine = maps__machine(thread__maps(c->al->thread));
|
||||
|
||||
script_fetch_insn(c->sample, c->al->thread, machine, /*native_arch=*/true);
|
||||
perf_sample__fetch_insn(c->sample, c->al->thread, machine);
|
||||
}
|
||||
if (!c->sample->insn_len)
|
||||
Py_RETURN_NONE; /* N.B. This is a return statement */
|
||||
|
||||
@@ -30,7 +30,6 @@
|
||||
#include "symbol.h"
|
||||
#include "synthetic-events.h"
|
||||
#include "util.h"
|
||||
#include "archinsn.h"
|
||||
#include "dlfilter.h"
|
||||
#include "tests.h"
|
||||
#include "util/sample.h"
|
||||
|
||||
@@ -1,12 +0,0 @@
|
||||
#ifndef INSN_H
|
||||
#define INSN_H 1
|
||||
|
||||
struct perf_sample;
|
||||
struct machine;
|
||||
struct thread;
|
||||
|
||||
void arch_fetch_insn(struct perf_sample *sample,
|
||||
struct thread *thread,
|
||||
struct machine *machine);
|
||||
|
||||
#endif
|
||||
@@ -234,8 +234,7 @@ static const __u8 *dlfilter__insn(void *ctx, __u32 *len)
|
||||
struct machine *machine = maps__machine(thread__maps(al->thread));
|
||||
|
||||
if (machine)
|
||||
script_fetch_insn(d->sample, al->thread, machine,
|
||||
/*native_arch=*/true);
|
||||
perf_sample__fetch_insn(d->sample, al->thread, machine);
|
||||
}
|
||||
}
|
||||
|
||||
|
||||
@@ -1,9 +1,18 @@
|
||||
/* SPDX-License-Identifier: GPL-2.0 */
|
||||
#include "sample.h"
|
||||
#include "debug.h"
|
||||
#include "thread.h"
|
||||
#include <elf.h>
|
||||
#ifndef EM_CSKY
|
||||
#define EM_CSKY 252
|
||||
#endif
|
||||
#ifndef EM_LOONGARCH
|
||||
#define EM_LOONGARCH 258
|
||||
#endif
|
||||
#include <linux/zalloc.h>
|
||||
#include <stdlib.h>
|
||||
#include <string.h>
|
||||
#include "../../arch/x86/include/asm/insn.h"
|
||||
|
||||
void perf_sample__init(struct perf_sample *sample, bool all)
|
||||
{
|
||||
@@ -41,3 +50,71 @@ struct regs_dump *perf_sample__intr_regs(struct perf_sample *sample)
|
||||
}
|
||||
return sample->intr_regs;
|
||||
}
|
||||
|
||||
static int elf_machine_max_instruction_length(uint16_t e_machine)
|
||||
{
|
||||
switch (e_machine) {
|
||||
/* Fixed 4-byte (32-bit) architectures */
|
||||
case EM_AARCH64:
|
||||
case EM_PPC:
|
||||
case EM_PPC64:
|
||||
case EM_MIPS:
|
||||
case EM_SPARC:
|
||||
case EM_SPARCV9:
|
||||
case EM_ALPHA:
|
||||
case EM_LOONGARCH:
|
||||
case EM_PARISC:
|
||||
case EM_SH:
|
||||
return 4;
|
||||
|
||||
/* Variable length or mixed-mode architectures */
|
||||
case EM_ARM: /* Variable due to Thumb/Thumb-2 */
|
||||
case EM_RISCV: /* Variable due to Compressed (C) extension */
|
||||
case EM_CSKY: /* Variable (16 or 32 bit) */
|
||||
case EM_ARC: /* Variable (ARCompact) */
|
||||
return 4;
|
||||
case EM_S390: /* Variable (2, 4, or 6 bytes) */
|
||||
return 6;
|
||||
case EM_68K:
|
||||
return 10;
|
||||
case EM_386:
|
||||
case EM_X86_64:
|
||||
return 15;
|
||||
case EM_XTENSA: /* Variable (FLIX) */
|
||||
return 16;
|
||||
default:
|
||||
return MAX_INSN;
|
||||
}
|
||||
}
|
||||
|
||||
void perf_sample__fetch_insn(struct perf_sample *sample,
|
||||
struct thread *thread,
|
||||
struct machine *machine)
|
||||
{
|
||||
int ret, len;
|
||||
bool is64bit = false;
|
||||
uint16_t e_machine;
|
||||
|
||||
if (!sample->ip || sample->insn_len != 0)
|
||||
return;
|
||||
|
||||
e_machine = thread__e_machine(thread, machine);
|
||||
len = elf_machine_max_instruction_length(e_machine);
|
||||
len = thread__memcpy(thread, machine, sample->insn,
|
||||
sample->ip, len,
|
||||
&is64bit);
|
||||
if (len <= 0)
|
||||
return;
|
||||
|
||||
sample->insn_len = len;
|
||||
|
||||
if (e_machine == EM_386 || e_machine == EM_X86_64) {
|
||||
/* Refine the x86 instruction length with the decoder. */
|
||||
struct insn insn;
|
||||
|
||||
ret = insn_decode(&insn, sample->insn, len,
|
||||
is64bit ? INSN_MODE_64 : INSN_MODE_32);
|
||||
if (ret >= 0 && insn.length <= len)
|
||||
sample->insn_len = insn.length;
|
||||
}
|
||||
}
|
||||
|
||||
@@ -5,6 +5,9 @@
|
||||
#include <linux/perf_event.h>
|
||||
#include <linux/types.h>
|
||||
|
||||
struct machine;
|
||||
struct thread;
|
||||
|
||||
/* number of register is bound by the number of bits in regs_dump::mask (64) */
|
||||
#define PERF_SAMPLE_REGS_CACHE_SIZE (8 * sizeof(u64))
|
||||
|
||||
@@ -127,6 +130,10 @@ void perf_sample__exit(struct perf_sample *sample);
|
||||
struct regs_dump *perf_sample__user_regs(struct perf_sample *sample);
|
||||
struct regs_dump *perf_sample__intr_regs(struct perf_sample *sample);
|
||||
|
||||
void perf_sample__fetch_insn(struct perf_sample *sample,
|
||||
struct thread *thread,
|
||||
struct machine *machine);
|
||||
|
||||
/*
|
||||
* raw_data is always 4 bytes from an 8-byte boundary, so subtract 4 to get
|
||||
* 8-byte alignment.
|
||||
|
||||
@@ -13,7 +13,6 @@
|
||||
#include <event-parse.h>
|
||||
#endif
|
||||
|
||||
#include "archinsn.h"
|
||||
#include "debug.h"
|
||||
#include "event.h"
|
||||
#include "trace-event.h"
|
||||
@@ -274,21 +273,6 @@ void setup_perl_scripting(void)
|
||||
#endif
|
||||
#endif
|
||||
|
||||
#if !defined(__i386__) && !defined(__x86_64__)
|
||||
void arch_fetch_insn(struct perf_sample *sample __maybe_unused,
|
||||
struct thread *thread __maybe_unused,
|
||||
struct machine *machine __maybe_unused)
|
||||
{
|
||||
}
|
||||
#endif
|
||||
|
||||
void script_fetch_insn(struct perf_sample *sample, struct thread *thread,
|
||||
struct machine *machine, bool native_arch)
|
||||
{
|
||||
if (sample->insn_len == 0 && native_arch)
|
||||
arch_fetch_insn(sample, thread, machine);
|
||||
}
|
||||
|
||||
static const struct {
|
||||
u32 flags;
|
||||
const char *name;
|
||||
|
||||
@@ -116,9 +116,6 @@ extern unsigned int scripting_max_stack;
|
||||
struct scripting_ops *script_spec__lookup(const char *spec);
|
||||
int script_spec__for_each(int (*cb)(struct scripting_ops *ops, const char *spec));
|
||||
|
||||
void script_fetch_insn(struct perf_sample *sample, struct thread *thread,
|
||||
struct machine *machine, bool native_arch);
|
||||
|
||||
void setup_perl_scripting(void);
|
||||
void setup_python_scripting(void);
|
||||
|
||||
|
||||
Reference in New Issue
Block a user