linux/tools/perf/util/annotate.c

3263 lines
80 KiB
C
Raw Normal View History

// SPDX-License-Identifier: GPL-2.0-only
/*
* Copyright (C) 2011, Red Hat Inc, Arnaldo Carvalho de Melo <acme@redhat.com>
*
* Parts came from builtin-annotate.c, see those files for further
* copyright notes.
*/
#include <errno.h>
#include <inttypes.h>
#include <libgen.h>
#include <stdlib.h>
#include "util.h" // hex_width()
#include "ui/ui.h"
#include "sort.h"
#include "build-id.h"
#include "color.h"
#include "config.h"
#include "dso.h"
#include "env.h"
#include "map.h"
#include "maps.h"
#include "symbol.h"
#include "srcline.h"
#include "units.h"
#include "debug.h"
#include "annotate.h"
#include "evsel.h"
#include "evlist.h"
perf annotate: Enable annotation of BPF programs In symbol__disassemble(), DSO_BINARY_TYPE__BPF_PROG_INFO dso calls into a new function symbol__disassemble_bpf(), where annotation line information is filled based on the bpf_prog_info and btf data saved in given perf_env. symbol__disassemble_bpf() uses binutils's libopcodes to disassemble bpf programs. Committer testing: After fixing this: - u64 *addrs = (u64 *)(info_linear->info.jited_ksyms); + u64 *addrs = (u64 *)(uintptr_t)(info_linear->info.jited_ksyms); Detected when crossbuilding to a 32-bit arch. And making all this dependent on HAVE_LIBBFD_SUPPORT and HAVE_LIBBPF_SUPPORT: 1) Have a BPF program running, one that has BTF info, etc, I used the tools/perf/examples/bpf/augmented_raw_syscalls.c put in place by 'perf trace'. # grep -B1 augmented_raw ~/.perfconfig [trace] add_events = /home/acme/git/perf/tools/perf/examples/bpf/augmented_raw_syscalls.c # # perf trace -e *mmsg dnf/6245 sendmmsg(20, 0x7f5485a88030, 2, MSG_NOSIGNAL) = 2 NetworkManager/10055 sendmmsg(22<socket:[1056822]>, 0x7f8126ad1bb0, 2, MSG_NOSIGNAL) = 2 2) Then do a 'perf record' system wide for a while: # perf record -a ^C[ perf record: Woken up 68 times to write data ] [ perf record: Captured and wrote 19.427 MB perf.data (366891 samples) ] # 3) Check that we captured BPF and BTF info in the perf.data file: # perf report --header-only | grep 'b[pt]f' # event : name = cycles:ppp, , id = { 294789, 294790, 294791, 294792, 294793, 294794, 294795, 294796 }, size = 112, { sample_period, sample_freq } = 4000, sample_type = IP|TID|TIME|CPU|PERIOD, read_format = ID, disabled = 1, inherit = 1, mmap = 1, comm = 1, freq = 1, task = 1, precise_ip = 3, sample_id_all = 1, exclude_guest = 1, mmap2 = 1, comm_exec = 1, ksymbol = 1, bpf_event = 1 # bpf_prog_info of id 13 # bpf_prog_info of id 14 # bpf_prog_info of id 15 # bpf_prog_info of id 16 # bpf_prog_info of id 17 # bpf_prog_info of id 18 # bpf_prog_info of id 21 # bpf_prog_info of id 22 # bpf_prog_info of id 41 # bpf_prog_info of id 42 # btf info of id 2 # 4) Check which programs got recorded: # perf report | grep bpf_prog | head 0.16% exe bpf_prog_819967866022f1e1_sys_enter [k] bpf_prog_819967866022f1e1_sys_enter 0.14% exe bpf_prog_c1bd85c092d6e4aa_sys_exit [k] bpf_prog_c1bd85c092d6e4aa_sys_exit 0.08% fuse-overlayfs bpf_prog_819967866022f1e1_sys_enter [k] bpf_prog_819967866022f1e1_sys_enter 0.07% fuse-overlayfs bpf_prog_c1bd85c092d6e4aa_sys_exit [k] bpf_prog_c1bd85c092d6e4aa_sys_exit 0.01% clang-4.0 bpf_prog_c1bd85c092d6e4aa_sys_exit [k] bpf_prog_c1bd85c092d6e4aa_sys_exit 0.01% clang-4.0 bpf_prog_819967866022f1e1_sys_enter [k] bpf_prog_819967866022f1e1_sys_enter 0.00% clang bpf_prog_c1bd85c092d6e4aa_sys_exit [k] bpf_prog_c1bd85c092d6e4aa_sys_exit 0.00% runc bpf_prog_819967866022f1e1_sys_enter [k] bpf_prog_819967866022f1e1_sys_enter 0.00% clang bpf_prog_819967866022f1e1_sys_enter [k] bpf_prog_819967866022f1e1_sys_enter 0.00% sh bpf_prog_c1bd85c092d6e4aa_sys_exit [k] bpf_prog_c1bd85c092d6e4aa_sys_exit # This was with the default --sort order for 'perf report', which is: --sort comm,dso,symbol If we just look for the symbol, for instance: # perf report --sort symbol | grep bpf_prog | head 0.26% [k] bpf_prog_819967866022f1e1_sys_enter - - 0.24% [k] bpf_prog_c1bd85c092d6e4aa_sys_exit - - # or the DSO: # perf report --sort dso | grep bpf_prog | head 0.26% bpf_prog_819967866022f1e1_sys_enter 0.24% bpf_prog_c1bd85c092d6e4aa_sys_exit # We'll see the two BPF programs that augmented_raw_syscalls.o puts in place, one attached to the raw_syscalls:sys_enter and another to the raw_syscalls:sys_exit tracepoints, as expected. Now we can finally do, from the command line, annotation for one of those two symbols, with the original BPF program source coude intermixed with the disassembled JITed code: # perf annotate --stdio2 bpf_prog_819967866022f1e1_sys_enter Samples: 950 of event 'cycles:ppp', 4000 Hz, Event count (approx.): 553756947, [percent: local period] bpf_prog_819967866022f1e1_sys_enter() bpf_prog_819967866022f1e1_sys_enter Percent int sys_enter(struct syscall_enter_args *args) 53.41 push %rbp 0.63 mov %rsp,%rbp 0.31 sub $0x170,%rsp 1.93 sub $0x28,%rbp 7.02 mov %rbx,0x0(%rbp) 3.20 mov %r13,0x8(%rbp) 1.07 mov %r14,0x10(%rbp) 0.61 mov %r15,0x18(%rbp) 0.11 xor %eax,%eax 1.29 mov %rax,0x20(%rbp) 0.11 mov %rdi,%rbx return bpf_get_current_pid_tgid(); 2.02 → callq *ffffffffda6776d9 2.76 mov %eax,-0x148(%rbp) mov %rbp,%rsi int sys_enter(struct syscall_enter_args *args) add $0xfffffffffffffeb8,%rsi return bpf_map_lookup_elem(pids, &pid) != NULL; movabs $0xffff975ac2607800,%rdi 1.26 → callq *ffffffffda6789e9 cmp $0x0,%rax 2.43 → je 0 add $0x38,%rax 0.21 xor %r13d,%r13d if (pid_filter__has(&pids_filtered, getpid())) 0.81 cmp $0x0,%rax → jne 0 mov %rbp,%rdi probe_read(&augmented_args.args, sizeof(augmented_args.args), args); 2.22 add $0xfffffffffffffeb8,%rdi 0.11 mov $0x40,%esi 0.32 mov %rbx,%rdx 2.74 → callq *ffffffffda658409 syscall = bpf_map_lookup_elem(&syscalls, &augmented_args.args.syscall_nr); 0.22 mov %rbp,%rsi 1.69 add $0xfffffffffffffec0,%rsi syscall = bpf_map_lookup_elem(&syscalls, &augmented_args.args.syscall_nr); movabs $0xffff975bfcd36000,%rdi add $0xd0,%rdi 0.21 mov 0x0(%rsi),%eax 0.93 cmp $0x200,%rax → jae 0 0.10 shl $0x3,%rax 0.11 add %rdi,%rax 0.11 → jmp 0 xor %eax,%eax if (syscall == NULL || !syscall->enabled) 1.07 cmp $0x0,%rax → je 0 if (syscall == NULL || !syscall->enabled) 6.57 movzbq 0x0(%rax),%rdi if (syscall == NULL || !syscall->enabled) cmp $0x0,%rdi 0.95 → je 0 mov $0x40,%r8d switch (augmented_args.args.syscall_nr) { mov -0x140(%rbp),%rdi switch (augmented_args.args.syscall_nr) { cmp $0x2,%rdi → je 0 cmp $0x101,%rdi → je 0 cmp $0x15,%rdi → jne 0 case SYS_OPEN: filename_arg = (const void *)args->args[0]; mov 0x10(%rbx),%rdx → jmp 0 case SYS_OPENAT: filename_arg = (const void *)args->args[1]; mov 0x18(%rbx),%rdx if (filename_arg != NULL) { cmp $0x0,%rdx → je 0 xor %edi,%edi augmented_args.filename.reserved = 0; mov %edi,-0x104(%rbp) augmented_args.filename.size = probe_read_str(&augmented_args.filename.value, mov %rbp,%rdi add $0xffffffffffffff00,%rdi augmented_args.filename.size = probe_read_str(&augmented_args.filename.value, mov $0x100,%esi → callq *ffffffffda658499 mov $0x148,%r8d augmented_args.filename.size = probe_read_str(&augmented_args.filename.value, mov %eax,-0x108(%rbp) augmented_args.filename.size = probe_read_str(&augmented_args.filename.value, mov %rax,%rdi shl $0x20,%rdi shr $0x20,%rdi if (augmented_args.filename.size < sizeof(augmented_args.filename.value)) { cmp $0xff,%rdi → ja 0 len -= sizeof(augmented_args.filename.value) - augmented_args.filename.size; add $0x48,%rax len &= sizeof(augmented_args.filename.value) - 1; and $0xff,%rax mov %rax,%r8 mov %rbp,%rcx return perf_event_output(args, &__augmented_syscalls__, BPF_F_CURRENT_CPU, &augmented_args, len); add $0xfffffffffffffeb8,%rcx mov %rbx,%rdi movabs $0xffff975fbd72d800,%rsi mov $0xffffffff,%edx → callq *ffffffffda658ad9 mov %rax,%r13 } mov %r13,%rax 0.72 mov 0x0(%rbp),%rbx mov 0x8(%rbp),%r13 1.16 mov 0x10(%rbp),%r14 0.10 mov 0x18(%rbp),%r15 0.42 add $0x28,%rbp 0.54 leaveq 0.54 ← retq # Please see 'man perf-config' to see how to control what should be seen, via ~/.perfconfig [annotate] section, for instance, one can suppress the source code and see just the disassembly, etc. Alternatively, use the TUI bu just using 'perf annotate', press '/bpf_prog' to see the bpf symbols, press enter and do the interactive annotation, which allows for dumping to a file after selecting the the various output tunables, for instance, the above without source code intermixed, plus showing all the instruction offsets: # perf annotate bpf_prog_819967866022f1e1_sys_enter Then press: 's' to hide the source code + 'O' twice to show all instruction offsets, then 'P' to print to the bpf_prog_819967866022f1e1_sys_enter.annotation file, which will have: # cat bpf_prog_819967866022f1e1_sys_enter.annotation bpf_prog_819967866022f1e1_sys_enter() bpf_prog_819967866022f1e1_sys_enter Event: cycles:ppp 53.41 0: push %rbp 0.63 1: mov %rsp,%rbp 0.31 4: sub $0x170,%rsp 1.93 b: sub $0x28,%rbp 7.02 f: mov %rbx,0x0(%rbp) 3.20 13: mov %r13,0x8(%rbp) 1.07 17: mov %r14,0x10(%rbp) 0.61 1b: mov %r15,0x18(%rbp) 0.11 1f: xor %eax,%eax 1.29 21: mov %rax,0x20(%rbp) 0.11 25: mov %rdi,%rbx 2.02 28: → callq *ffffffffda6776d9 2.76 2d: mov %eax,-0x148(%rbp) 33: mov %rbp,%rsi 36: add $0xfffffffffffffeb8,%rsi 3d: movabs $0xffff975ac2607800,%rdi 1.26 47: → callq *ffffffffda6789e9 4c: cmp $0x0,%rax 2.43 50: → je 0 52: add $0x38,%rax 0.21 56: xor %r13d,%r13d 0.81 59: cmp $0x0,%rax 5d: → jne 0 63: mov %rbp,%rdi 2.22 66: add $0xfffffffffffffeb8,%rdi 0.11 6d: mov $0x40,%esi 0.32 72: mov %rbx,%rdx 2.74 75: → callq *ffffffffda658409 0.22 7a: mov %rbp,%rsi 1.69 7d: add $0xfffffffffffffec0,%rsi 84: movabs $0xffff975bfcd36000,%rdi 8e: add $0xd0,%rdi 0.21 95: mov 0x0(%rsi),%eax 0.93 98: cmp $0x200,%rax 9f: → jae 0 0.10 a1: shl $0x3,%rax 0.11 a5: add %rdi,%rax 0.11 a8: → jmp 0 aa: xor %eax,%eax 1.07 ac: cmp $0x0,%rax b0: → je 0 6.57 b6: movzbq 0x0(%rax),%rdi bb: cmp $0x0,%rdi 0.95 bf: → je 0 c5: mov $0x40,%r8d cb: mov -0x140(%rbp),%rdi d2: cmp $0x2,%rdi d6: → je 0 d8: cmp $0x101,%rdi df: → je 0 e1: cmp $0x15,%rdi e5: → jne 0 e7: mov 0x10(%rbx),%rdx eb: → jmp 0 ed: mov 0x18(%rbx),%rdx f1: cmp $0x0,%rdx f5: → je 0 f7: xor %edi,%edi f9: mov %edi,-0x104(%rbp) ff: mov %rbp,%rdi 102: add $0xffffffffffffff00,%rdi 109: mov $0x100,%esi 10e: → callq *ffffffffda658499 113: mov $0x148,%r8d 119: mov %eax,-0x108(%rbp) 11f: mov %rax,%rdi 122: shl $0x20,%rdi 126: shr $0x20,%rdi 12a: cmp $0xff,%rdi 131: → ja 0 133: add $0x48,%rax 137: and $0xff,%rax 13d: mov %rax,%r8 140: mov %rbp,%rcx 143: add $0xfffffffffffffeb8,%rcx 14a: mov %rbx,%rdi 14d: movabs $0xffff975fbd72d800,%rsi 157: mov $0xffffffff,%edx 15c: → callq *ffffffffda658ad9 161: mov %rax,%r13 164: mov %r13,%rax 0.72 167: mov 0x0(%rbp),%rbx 16b: mov 0x8(%rbp),%r13 1.16 16f: mov 0x10(%rbp),%r14 0.10 173: mov 0x18(%rbp),%r15 0.42 177: add $0x28,%rbp 0.54 17b: leaveq 0.54 17c: ← retq Another cool way to test all this is to symple use 'perf top' look for those symbols, go there and press enter, annotate it live :-) Signed-off-by: Song Liu <songliubraving@fb.com> Reviewed-by: Jiri Olsa <jolsa@kernel.org> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Alexei Starovoitov <ast@kernel.org> Cc: Daniel Borkmann <daniel@iogearbox.net> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Stanislav Fomichev <sdf@google.com> Link: http://lkml.kernel.org/r/20190312053051.2690567-13-songliubraving@fb.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2019-03-12 08:30:48 +03:00
#include "bpf-event.h"
#include "bpf-utils.h"
perf annotate: Add branch stack / basic block I wanted to know the hottest path through a function and figured the branch-stack (LBR) information should be able to help out with that. The below uses the branch-stack to create basic blocks and generate statistics from them. from to branch_i * ----> * | | block v * ----> * from to branch_i+1 The blocks are broken down into non-overlapping ranges, while tracking if the start of each range is an entry point and/or the end of a range is a branch. Each block iterates all ranges it covers (while splitting where required to exactly match the block) and increments the 'coverage' count. For the range including the branch we increment the taken counter, as well as the pred counter if flags.predicted. Using these number we can find if an instruction: - had coverage; given by: br->coverage / br->sym->max_coverage This metric ensures each symbol has a 100% spot, which reflects the observation that each symbol must have a most covered/hottest block. - is a branch target: br->is_target && br->start == add - for targets, how much of a branch's coverages comes from it: target->entry / branch->coverage - is a branch: br->is_branch && br->end == addr - for branches, how often it was taken: br->taken / br->coverage after all, all execution that didn't take the branch would have incremented the coverage and continued onward to a later branch. - for branches, how often it was predicted: br->pred / br->taken The coverage percentage is used to color the address and asm sections; for low (<1%) coverage we use NORMAL (uncolored), indicating that these instructions are not 'important'. For high coverage (>75%) we color the address RED. For each branch, we add an asm comment after the instruction with information on how often it was taken and predicted. Output looks like (sans color, which does loose a lot of the information :/) $ perf record --branch-filter u,any -e cycles:p ./branches 27 $ perf annotate branches Percent | Source code & Disassembly of branches for cycles:pu (217 samples) --------------------------------------------------------------------------------- : branches(): 0.00 : 40057a: push %rbp 0.00 : 40057b: mov %rsp,%rbp 0.00 : 40057e: sub $0x20,%rsp 0.00 : 400582: mov %rdi,-0x18(%rbp) 0.00 : 400586: mov %rsi,-0x20(%rbp) 0.00 : 40058a: mov -0x18(%rbp),%rax 0.00 : 40058e: mov %rax,-0x10(%rbp) 0.00 : 400592: movq $0x0,-0x8(%rbp) 0.00 : 40059a: jmpq 400656 <branches+0xdc> 1.84 : 40059f: mov -0x10(%rbp),%rax # +100.00% 3.23 : 4005a3: and $0x1,%eax 1.84 : 4005a6: test %rax,%rax 0.00 : 4005a9: je 4005bf <branches+0x45> # -54.50% (p:42.00%) 0.46 : 4005ab: mov 0x200bbe(%rip),%rax # 601170 <acc> 12.90 : 4005b2: add $0x1,%rax 2.30 : 4005b6: mov %rax,0x200bb3(%rip) # 601170 <acc> 0.46 : 4005bd: jmp 4005d1 <branches+0x57> # -100.00% (p:100.00%) 0.92 : 4005bf: mov 0x200baa(%rip),%rax # 601170 <acc> # +49.54% 13.82 : 4005c6: sub $0x1,%rax 0.46 : 4005ca: mov %rax,0x200b9f(%rip) # 601170 <acc> 2.30 : 4005d1: mov -0x10(%rbp),%rax # +50.46% 0.46 : 4005d5: mov %rax,%rdi 0.46 : 4005d8: callq 400526 <lfsr> # -100.00% (p:100.00%) 0.00 : 4005dd: mov %rax,-0x10(%rbp) # +100.00% 0.92 : 4005e1: mov -0x18(%rbp),%rax 0.00 : 4005e5: and $0x1,%eax 0.00 : 4005e8: test %rax,%rax 0.00 : 4005eb: je 4005ff <branches+0x85> # -100.00% (p:100.00%) 0.00 : 4005ed: mov 0x200b7c(%rip),%rax # 601170 <acc> 0.00 : 4005f4: shr $0x2,%rax 0.00 : 4005f8: mov %rax,0x200b71(%rip) # 601170 <acc> 0.00 : 4005ff: mov -0x10(%rbp),%rax # +100.00% 7.37 : 400603: and $0x1,%eax 3.69 : 400606: test %rax,%rax 0.00 : 400609: jne 400612 <branches+0x98> # -59.25% (p:42.99%) 1.84 : 40060b: mov $0x1,%eax 14.29 : 400610: jmp 400617 <branches+0x9d> # -100.00% (p:100.00%) 1.38 : 400612: mov $0x0,%eax # +57.65% 10.14 : 400617: test %al,%al # +42.35% 0.00 : 400619: je 40062f <branches+0xb5> # -57.65% (p:100.00%) 0.46 : 40061b: mov 0x200b4e(%rip),%rax # 601170 <acc> 2.76 : 400622: sub $0x1,%rax 0.00 : 400626: mov %rax,0x200b43(%rip) # 601170 <acc> 0.46 : 40062d: jmp 400641 <branches+0xc7> # -100.00% (p:100.00%) 0.92 : 40062f: mov 0x200b3a(%rip),%rax # 601170 <acc> # +56.13% 2.30 : 400636: add $0x1,%rax 0.92 : 40063a: mov %rax,0x200b2f(%rip) # 601170 <acc> 0.92 : 400641: mov -0x10(%rbp),%rax # +43.87% 2.30 : 400645: mov %rax,%rdi 0.00 : 400648: callq 400526 <lfsr> # -100.00% (p:100.00%) 0.00 : 40064d: mov %rax,-0x10(%rbp) # +100.00% 1.84 : 400651: addq $0x1,-0x8(%rbp) 0.92 : 400656: mov -0x8(%rbp),%rax 5.07 : 40065a: cmp -0x20(%rbp),%rax 0.00 : 40065e: jb 40059f <branches+0x25> # -100.00% (p:100.00%) 0.00 : 400664: nop 0.00 : 400665: leaveq 0.00 : 400666: retq (Note: the --branch-filter u,any was used to avoid spurious target and branch points due to interrupts/faults, they show up as very small -/+ annotations on 'weird' locations) Committer note: Please take a look at: http://vger.kernel.org/~acme/perf/annotate_basic_blocks.png To see the colors. Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Andi Kleen <andi@firstfloor.org> Cc: Anshuman Khandual <khandual@linux.vnet.ibm.com> Cc: David Carrillo-Cisneros <davidcc@google.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Kan Liang <kan.liang@intel.com> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Stephane Eranian <eranian@google.com> [ Moved sym->max_coverage to 'struct annotate', aka symbol__annotate(sym) ] Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-09-05 22:08:12 +03:00
#include "block-range.h"
#include "string2.h"
#include "util/event.h"
#include "arch/common.h"
#include <regex.h>
#include <pthread.h>
#include <linux/bitops.h>
#include <linux/kernel.h>
#include <linux/string.h>
#include <subcmd/parse-options.h>
#include <subcmd/run-command.h>
/* FIXME: For the HE_COLORSET */
#include "ui/browser.h"
/*
* FIXME: Using the same values as slang.h,
* but that header may not be available everywhere
*/
#define LARROW_CHAR ((unsigned char)',')
#define RARROW_CHAR ((unsigned char)'+')
#define DARROW_CHAR ((unsigned char)'.')
#define UARROW_CHAR ((unsigned char)'-')
#include <linux/ctype.h>
struct annotation_options annotation__default_options = {
.use_offset = true,
.jump_arrows = true,
.annotate_src = true,
.offset_level = ANNOTATION__OFFSET_JUMP_TARGETS,
.percent_type = PERCENT_PERIOD_LOCAL,
};
static regex_t file_lineno;
perf annotate: Remove duplicate 'name' field from disasm_line The disasm_line::name field is always equal to ins::name, being used just to locate the instruction's ins_ops from the per-arch instructions table. Eliminate this duplication, nuking that field and instead make ins__find() return an ins_ops, store it in disasm_line::ins.ops, and keep just in disasm_line::ins.name what was in disasm_line::name, this way we end up not keeping a reference to entries in the per-arch instructions table. This in turn will help supporting multiple ways to manage the per-arch instructions table, allowing resorting that array, for instance, when the entries will move after references to its addresses were made. The same problem is avoided when one grows the array with realloc. So architectures simply keeping a constant array will work as well as architectures building the table using regular expressions or other logic that involves resorting the table. Reviewed-by: Ravi Bangoria <ravi.bangoria@linux.vnet.ibm.com> Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Chris Riyder <chris.ryder@arm.com> Cc: David Ahern <dsahern@gmail.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Kim Phillips <kim.phillips@arm.com> Cc: Markus Trippelsdorf <markus@trippelsdorf.de> Cc: Masami Hiramatsu <mhiramat@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Naveen N. Rao <naveen.n.rao@linux.vnet.ibm.com> Cc: Pawel Moll <pawel.moll@arm.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Russell King <rmk+kernel@arm.linux.org.uk> Cc: Taeung Song <treeze.taeung@gmail.com> Cc: Wang Nan <wangnan0@huawei.com> Link: http://lkml.kernel.org/n/tip-vr899azvabnw9gtuepuqfd9t@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-11-24 17:16:06 +03:00
static struct ins_ops *ins__find(struct arch *arch, const char *name);
perf annotate: Introduce alternative method of keeping instructions table Some arches may want to dynamically populate the table using regular expressions on the instruction names to associate them with a set of parsing/formatting/etc functions (struct ins_ops), so provide a fallback for when the ins__find() method fails. That fall back will be able to resize the arch->instructions, setting arch->nr_instructions appropriately, helper functions to associate an ins_ops to an instruction name, growing the arch->instructions if needed and resorting it are provided, all the arch specific callback needs to do is to decide if the missing instruction should be added to arch->instructions with a ins_ops association. Reviewed-by: Ravi Bangoria <ravi.bangoria@linux.vnet.ibm.com> Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Chris Riyder <chris.ryder@arm.com> Cc: David Ahern <dsahern@gmail.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Kim Phillips <kim.phillips@arm.com> Cc: Markus Trippelsdorf <markus@trippelsdorf.de> Cc: Masami Hiramatsu <mhiramat@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Naveen N. Rao <naveen.n.rao@linux.vnet.ibm.com> Cc: Pawel Moll <pawel.moll@arm.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Russell King <rmk+kernel@arm.linux.org.uk> Cc: Taeung Song <treeze.taeung@gmail.com> Cc: Wang Nan <wangnan0@huawei.com> Link: http://lkml.kernel.org/n/tip-auu13yradxf7g5dgtpnzt97a@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-11-24 17:37:08 +03:00
static void ins__sort(struct arch *arch);
perf annotate: Remove duplicate 'name' field from disasm_line The disasm_line::name field is always equal to ins::name, being used just to locate the instruction's ins_ops from the per-arch instructions table. Eliminate this duplication, nuking that field and instead make ins__find() return an ins_ops, store it in disasm_line::ins.ops, and keep just in disasm_line::ins.name what was in disasm_line::name, this way we end up not keeping a reference to entries in the per-arch instructions table. This in turn will help supporting multiple ways to manage the per-arch instructions table, allowing resorting that array, for instance, when the entries will move after references to its addresses were made. The same problem is avoided when one grows the array with realloc. So architectures simply keeping a constant array will work as well as architectures building the table using regular expressions or other logic that involves resorting the table. Reviewed-by: Ravi Bangoria <ravi.bangoria@linux.vnet.ibm.com> Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Chris Riyder <chris.ryder@arm.com> Cc: David Ahern <dsahern@gmail.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Kim Phillips <kim.phillips@arm.com> Cc: Markus Trippelsdorf <markus@trippelsdorf.de> Cc: Masami Hiramatsu <mhiramat@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Naveen N. Rao <naveen.n.rao@linux.vnet.ibm.com> Cc: Pawel Moll <pawel.moll@arm.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Russell King <rmk+kernel@arm.linux.org.uk> Cc: Taeung Song <treeze.taeung@gmail.com> Cc: Wang Nan <wangnan0@huawei.com> Link: http://lkml.kernel.org/n/tip-vr899azvabnw9gtuepuqfd9t@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-11-24 17:16:06 +03:00
static int disasm_line__parse(char *line, const char **namep, char **rawp);
struct arch {
const char *name;
struct ins *instructions;
size_t nr_instructions;
perf annotate: Introduce alternative method of keeping instructions table Some arches may want to dynamically populate the table using regular expressions on the instruction names to associate them with a set of parsing/formatting/etc functions (struct ins_ops), so provide a fallback for when the ins__find() method fails. That fall back will be able to resize the arch->instructions, setting arch->nr_instructions appropriately, helper functions to associate an ins_ops to an instruction name, growing the arch->instructions if needed and resorting it are provided, all the arch specific callback needs to do is to decide if the missing instruction should be added to arch->instructions with a ins_ops association. Reviewed-by: Ravi Bangoria <ravi.bangoria@linux.vnet.ibm.com> Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Chris Riyder <chris.ryder@arm.com> Cc: David Ahern <dsahern@gmail.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Kim Phillips <kim.phillips@arm.com> Cc: Markus Trippelsdorf <markus@trippelsdorf.de> Cc: Masami Hiramatsu <mhiramat@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Naveen N. Rao <naveen.n.rao@linux.vnet.ibm.com> Cc: Pawel Moll <pawel.moll@arm.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Russell King <rmk+kernel@arm.linux.org.uk> Cc: Taeung Song <treeze.taeung@gmail.com> Cc: Wang Nan <wangnan0@huawei.com> Link: http://lkml.kernel.org/n/tip-auu13yradxf7g5dgtpnzt97a@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-11-24 17:37:08 +03:00
size_t nr_instructions_allocated;
struct ins_ops *(*associate_instruction_ops)(struct arch *arch, const char *name);
bool sorted_instructions;
bool initialized;
void *priv;
perf annotate: Check for fused instructions Macro fusion merges two instructions to a single micro-op. Intel core platform performs this hardware optimization under limited circumstances. For example, CMP + JCC can be "fused" and executed /retired together. While with sampling this can result in the sample sometimes being on the JCC and sometimes on the CMP. So for the fused instruction pair, they could be considered together. On Nehalem, fused instruction pairs: cmp/test + jcc. On other new CPU: cmp/test/add/sub/and/inc/dec + jcc. This patch adds an x86-specific function which checks if 2 instructions are in a "fused" pair. For non-x86 arch, the function is just NULL. Changelog: v4: Move the CPU model checking to symbol__disassemble and save the CPU family/model in arch structure. It avoids checking every time when jump arrow printed. v3: Add checking for Nehalem (CMP, TEST). For other newer Intel CPUs just check it by default (CMP, TEST, ADD, SUB, AND, INC, DEC). v2: Remove the original weak function. Arnaldo points out that doing it as a weak function that will be overridden by the host arch doesn't work. So now it's implemented as an arch-specific function. Committer fix: Do not access evsel->evlist->env->cpuid, ->env can be null, introduce perf_evsel__env_cpuid(), just like perf_evsel__env_arch(), also used in this function call. The original patch was segfaulting 'perf top' + annotation. But this essentially disables this fused instructions augmentation in 'perf top', the right thing is to get the cpuid from the running kernel, left for a later patch tho. Signed-off-by: Yao Jin <yao.jin@linux.intel.com> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Andi Kleen <ak@linux.intel.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Kan Liang <kan.liang@intel.com> Cc: Peter Zijlstra <peterz@infradead.org> Link: http://lkml.kernel.org/r/1499403995-19857-2-git-send-email-yao.jin@linux.intel.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2017-07-07 08:06:34 +03:00
unsigned int model;
unsigned int family;
int (*init)(struct arch *arch, char *cpuid);
perf annotate: Check for fused instructions Macro fusion merges two instructions to a single micro-op. Intel core platform performs this hardware optimization under limited circumstances. For example, CMP + JCC can be "fused" and executed /retired together. While with sampling this can result in the sample sometimes being on the JCC and sometimes on the CMP. So for the fused instruction pair, they could be considered together. On Nehalem, fused instruction pairs: cmp/test + jcc. On other new CPU: cmp/test/add/sub/and/inc/dec + jcc. This patch adds an x86-specific function which checks if 2 instructions are in a "fused" pair. For non-x86 arch, the function is just NULL. Changelog: v4: Move the CPU model checking to symbol__disassemble and save the CPU family/model in arch structure. It avoids checking every time when jump arrow printed. v3: Add checking for Nehalem (CMP, TEST). For other newer Intel CPUs just check it by default (CMP, TEST, ADD, SUB, AND, INC, DEC). v2: Remove the original weak function. Arnaldo points out that doing it as a weak function that will be overridden by the host arch doesn't work. So now it's implemented as an arch-specific function. Committer fix: Do not access evsel->evlist->env->cpuid, ->env can be null, introduce perf_evsel__env_cpuid(), just like perf_evsel__env_arch(), also used in this function call. The original patch was segfaulting 'perf top' + annotation. But this essentially disables this fused instructions augmentation in 'perf top', the right thing is to get the cpuid from the running kernel, left for a later patch tho. Signed-off-by: Yao Jin <yao.jin@linux.intel.com> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Andi Kleen <ak@linux.intel.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Kan Liang <kan.liang@intel.com> Cc: Peter Zijlstra <peterz@infradead.org> Link: http://lkml.kernel.org/r/1499403995-19857-2-git-send-email-yao.jin@linux.intel.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2017-07-07 08:06:34 +03:00
bool (*ins_is_fused)(struct arch *arch, const char *ins1,
const char *ins2);
struct {
char comment_char;
char skip_functions_char;
} objdump;
};
static struct ins_ops call_ops;
static struct ins_ops dec_ops;
static struct ins_ops jump_ops;
static struct ins_ops mov_ops;
static struct ins_ops nop_ops;
static struct ins_ops lock_ops;
static struct ins_ops ret_ops;
perf annotate: Introduce alternative method of keeping instructions table Some arches may want to dynamically populate the table using regular expressions on the instruction names to associate them with a set of parsing/formatting/etc functions (struct ins_ops), so provide a fallback for when the ins__find() method fails. That fall back will be able to resize the arch->instructions, setting arch->nr_instructions appropriately, helper functions to associate an ins_ops to an instruction name, growing the arch->instructions if needed and resorting it are provided, all the arch specific callback needs to do is to decide if the missing instruction should be added to arch->instructions with a ins_ops association. Reviewed-by: Ravi Bangoria <ravi.bangoria@linux.vnet.ibm.com> Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Chris Riyder <chris.ryder@arm.com> Cc: David Ahern <dsahern@gmail.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Kim Phillips <kim.phillips@arm.com> Cc: Markus Trippelsdorf <markus@trippelsdorf.de> Cc: Masami Hiramatsu <mhiramat@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Naveen N. Rao <naveen.n.rao@linux.vnet.ibm.com> Cc: Pawel Moll <pawel.moll@arm.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Russell King <rmk+kernel@arm.linux.org.uk> Cc: Taeung Song <treeze.taeung@gmail.com> Cc: Wang Nan <wangnan0@huawei.com> Link: http://lkml.kernel.org/n/tip-auu13yradxf7g5dgtpnzt97a@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-11-24 17:37:08 +03:00
static int arch__grow_instructions(struct arch *arch)
{
struct ins *new_instructions;
size_t new_nr_allocated;
if (arch->nr_instructions_allocated == 0 && arch->instructions)
goto grow_from_non_allocated_table;
new_nr_allocated = arch->nr_instructions_allocated + 128;
new_instructions = realloc(arch->instructions, new_nr_allocated * sizeof(struct ins));
if (new_instructions == NULL)
return -1;
out_update_instructions:
arch->instructions = new_instructions;
arch->nr_instructions_allocated = new_nr_allocated;
return 0;
grow_from_non_allocated_table:
new_nr_allocated = arch->nr_instructions + 128;
new_instructions = calloc(new_nr_allocated, sizeof(struct ins));
if (new_instructions == NULL)
return -1;
memcpy(new_instructions, arch->instructions, arch->nr_instructions);
goto out_update_instructions;
}
static int arch__associate_ins_ops(struct arch* arch, const char *name, struct ins_ops *ops)
perf annotate: Introduce alternative method of keeping instructions table Some arches may want to dynamically populate the table using regular expressions on the instruction names to associate them with a set of parsing/formatting/etc functions (struct ins_ops), so provide a fallback for when the ins__find() method fails. That fall back will be able to resize the arch->instructions, setting arch->nr_instructions appropriately, helper functions to associate an ins_ops to an instruction name, growing the arch->instructions if needed and resorting it are provided, all the arch specific callback needs to do is to decide if the missing instruction should be added to arch->instructions with a ins_ops association. Reviewed-by: Ravi Bangoria <ravi.bangoria@linux.vnet.ibm.com> Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Chris Riyder <chris.ryder@arm.com> Cc: David Ahern <dsahern@gmail.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Kim Phillips <kim.phillips@arm.com> Cc: Markus Trippelsdorf <markus@trippelsdorf.de> Cc: Masami Hiramatsu <mhiramat@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Naveen N. Rao <naveen.n.rao@linux.vnet.ibm.com> Cc: Pawel Moll <pawel.moll@arm.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Russell King <rmk+kernel@arm.linux.org.uk> Cc: Taeung Song <treeze.taeung@gmail.com> Cc: Wang Nan <wangnan0@huawei.com> Link: http://lkml.kernel.org/n/tip-auu13yradxf7g5dgtpnzt97a@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-11-24 17:37:08 +03:00
{
struct ins *ins;
if (arch->nr_instructions == arch->nr_instructions_allocated &&
arch__grow_instructions(arch))
return -1;
ins = &arch->instructions[arch->nr_instructions];
ins->name = strdup(name);
if (!ins->name)
return -1;
ins->ops = ops;
arch->nr_instructions++;
ins__sort(arch);
return 0;
}
#include "arch/arc/annotate/instructions.c"
#include "arch/arm/annotate/instructions.c"
perf annotate: AArch64 support This is a regex converted version from the original: https://lkml.org/lkml/2016/5/19/461 Add basic support to recognise AArch64 assembly. This allows perf to identify AArch64 instructions that branch to other parts within the same function, thereby properly annotating them. Rebased onto new cross-arch annotation bits: https://lkml.org/lkml/2016/11/25/546 Sample output: security_file_permission vmlinux 5.80 │ ← ret ▒ │70: ldr w0, [x21,#68] ▒ 4.44 │ ↓ tbnz d0 ▒ │ mov w0, #0x24 // #36 ▒ 1.37 │ ands w0, w22, w0 ▒ │ ↑ b.eq 60 ▒ 1.37 │ ↓ tbnz e4 ▒ │ mov w19, #0x20000 // #131072 ▒ 1.02 │ ↓ tbz ec ▒ │90:┌─→ldr x3, [x21,#24] ▒ 1.37 │ │ add x21, x21, #0x10 ▒ │ │ mov w2, w19 ▒ 1.02 │ │ mov x0, x21 ▒ │ │ mov x1, x3 ▒ 1.71 │ │ ldr x20, [x3,#48] ▒ │ │→ bl __fsnotify_parent ▒ 0.68 │ │↑ cbnz 60 ▒ │ │ mov x2, x21 ▒ 1.37 │ │ mov w1, w19 ▒ │ │ mov x0, x20 ▒ 0.68 │ │ mov w5, #0x0 // #0 ▒ │ │ mov x4, #0x0 // #0 ▒ 1.71 │ │ mov w3, #0x1 // #1 ▒ │ │→ bl fsnotify ▒ 1.37 │ │↑ b 60 ▒ │d0:│ mov w0, #0x0 // #0 ▒ │ │ ldp x19, x20, [sp,#16] ▒ │ │ ldp x21, x22, [sp,#32] ▒ │ │ ldp x29, x30, [sp],#48 ▒ │ │← ret ▒ │e4:│ mov w19, #0x10000 // #65536 ▒ │ └──b 90 ◆ │ec: brk #0x800 ▒ Press 'h' for help on key bindings Signed-off-by: Kim Phillips <kim.phillips@arm.com> Signed-off-by: Chris Ryder <chris.ryder@arm.com> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Mark Rutland <mark.rutland@arm.com> Cc: Pawel Moll <pawel.moll@arm.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Will Deacon <will.deacon@arm.com> Link: http://lkml.kernel.org/r/20161130092344.012e18e3e623bea395162f95@arm.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-11-30 18:23:44 +03:00
#include "arch/arm64/annotate/instructions.c"
#include "arch/csky/annotate/instructions.c"
perf annotate mips: Add perf arch instructions annotate handlers Support the MIPS architecture using the ins_ops association method. With this patch, perf-annotate can work well on MIPS. Testing it with a perf.data file collected on a mips machine: $./perf annotate -i perf.data : Disassembly of section .text: : : 00000000000be6a0 <get_next_seq>: : get_next_seq(): 0.00 : be6a0: lw v0,0(a0) 0.00 : be6a4: daddiu sp,sp,-128 0.00 : be6a8: ld a7,72(a0) 0.00 : be6ac: gssq s5,s4,80(sp) 0.00 : be6b0: gssq s1,s0,48(sp) 0.00 : be6b4: gssq s8,gp,112(sp) 0.00 : be6b8: gssq s7,s6,96(sp) 0.00 : be6bc: gssq s3,s2,64(sp) 0.00 : be6c0: sd a3,0(sp) 0.00 : be6c4: move s0,a0 0.00 : be6c8: sd v0,32(sp) 0.00 : be6cc: sd a5,8(sp) 0.00 : be6d0: sd zero,8(a0) 0.00 : be6d4: sd a6,16(sp) 0.00 : be6d8: ld s2,48(a0) 8.53 : be6dc: ld s1,40(a0) 9.42 : be6e0: ld v1,32(a0) 0.00 : be6e4: nop 0.00 : be6e8: ld s4,24(a0) 0.00 : be6ec: ld s5,16(a0) 0.00 : be6f0: sd a7,40(sp) 10.11 : be6f4: ld s6,64(a0) ... The original patch link: https://lore.kernel.org/patchwork/patch/1180480/ Signed-off-by: Dengcheng Zhu <dzhu@wavecomp.com> Cc: Dengcheng Zhu <dzhu@wavecomp.com> Cc: Jiaxun Yang <jiaxun.yang@flygoat.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Xuefeng Li <lixuefeng@loongson.cn> Cc: linux-mips@vger.kernel.org [ fanpeng@loongson.cn: Add missing "bgtzl", "bltzl", "bgezl", "blezl", "beql" and "bnel" for pre-R6processors ] Signed-off-by: Peng Fan <fanpeng@loongson.cn> Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2020-10-19 10:21:24 +03:00
#include "arch/mips/annotate/instructions.c"
#include "arch/x86/annotate/instructions.c"
perf annotate: Initial PowerPC support Support the PowerPC architecture using the ins_ops association method. Committer notes: Testing it with a perf.data file collected on a PowerPC machine and cross-annotated on a x86_64 workstation, using the associated vmlinux file: $ perf report -i perf.data.f22vm.powerdev --vmlinux vmlinux.powerpc .ktime_get vmlinux.powerpc │ clrldi r9,r28,63 8.57 │ ┌──bne e0 <- TUI cursor positioned here │54:│ lwsync 2.86 │ │ std r2,40(r1) │ │ ld r9,144(r31) │ │ ld r3,136(r31) │ │ ld r30,184(r31) │ │ ld r10,0(r9) │ │ mtctr r10 │ │ ld r2,8(r9) 8.57 │ │→ bctrl │ │ ld r2,40(r1) │ │ ld r10,160(r31) │ │ ld r5,152(r31) │ │ lwz r7,168(r31) │ │ ld r9,176(r31) 8.57 │ │ lwz r6,172(r31) │ │ lwsync 2.86 │ │ lwz r8,128(r31) │ │ cmpw cr7,r8,r28 2.86 │ │↑ bne 48 │ │ subf r10,r10,r3 │ │ mr r3,r29 │ │ and r10,r10,r5 2.86 │ │ mulld r10,r10,r7 │ │ add r9,r10,r9 │ │ srd r9,r9,r6 │ │ add r9,r9,r30 │ │ std r9,0(r29) │ │ addi r1,r1,144 │ │ ld r0,16(r1) │ │ ld r28,-32(r1) │ │ ld r29,-24(r1) │ │ ld r30,-16(r1) │ │ mtlr r0 │ │ ld r31,-8(r1) │ │← blr 5.71 │e0:└─→mr r1,r1 11.43 │ mr r2,r2 11.43 │ lwz r28,128(r31) Press 'h' for help on key bindings $ perf report -i perf.data.f22vm.powerdev --header-only # ======== # captured on: Thu Nov 24 12:40:38 2016 # hostname : pdev-f22-qemu # os release : 4.4.10-200.fc22.ppc64 # perf version : 4.9.rc1.g6298ce # arch : ppc64 # nrcpus online : 48 # nrcpus avail : 48 # cpudesc : POWER7 (architected), altivec supported # cpuid : 74,513 # total memory : 4158976 kB # cmdline : /home/ravi/Workspace/linux/tools/perf/perf record -a # event : name = cycles:ppp, , size = 112, { sample_period, sample_freq } = 4000, sample_type = IP|TID|TIME|CPU|PERIOD, disabled = 1, inherit = 1, mmap = 1, comm = 1, freq = 1, task = 1, precise_ip = 3, sample_id_all = 1, exclude_guest = 1, mmap2 = 1, comm_exec = 1 # HEADER_CPU_TOPOLOGY info available, use -I to display # HEADER_NUMA_TOPOLOGY info available, use -I to display # pmu mappings: cpu = 4, software = 1, tracepoint = 2, breakpoint = 5 # missing features: HEADER_TRACING_DATA HEADER_BRANCH_STACK HEADER_GROUP_DESC HEADER_AUXTRACE HEADER_STAT HEADER_CACHE # ======== # $ Signed-off-by: Ravi Bangoria <ravi.bangoria@linux.vnet.ibm.com> Signed-off-by: Naveen N. Rao <naveen.n.rao@linux.vnet.ibm.com> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Kim Phillips <kim.phillips@arm.com> Link: http://lkml.kernel.org/n/tip-tbjnp40ddoxxl474uvhwi6g4@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-11-23 19:03:46 +03:00
#include "arch/powerpc/annotate/instructions.c"
#include "arch/riscv64/annotate/instructions.c"
#include "arch/s390/annotate/instructions.c"
#include "arch/sparc/annotate/instructions.c"
static struct arch architectures[] = {
{
.name = "arc",
.init = arc__annotate_init,
},
{
.name = "arm",
.init = arm__annotate_init,
},
perf annotate: AArch64 support This is a regex converted version from the original: https://lkml.org/lkml/2016/5/19/461 Add basic support to recognise AArch64 assembly. This allows perf to identify AArch64 instructions that branch to other parts within the same function, thereby properly annotating them. Rebased onto new cross-arch annotation bits: https://lkml.org/lkml/2016/11/25/546 Sample output: security_file_permission vmlinux 5.80 │ ← ret ▒ │70: ldr w0, [x21,#68] ▒ 4.44 │ ↓ tbnz d0 ▒ │ mov w0, #0x24 // #36 ▒ 1.37 │ ands w0, w22, w0 ▒ │ ↑ b.eq 60 ▒ 1.37 │ ↓ tbnz e4 ▒ │ mov w19, #0x20000 // #131072 ▒ 1.02 │ ↓ tbz ec ▒ │90:┌─→ldr x3, [x21,#24] ▒ 1.37 │ │ add x21, x21, #0x10 ▒ │ │ mov w2, w19 ▒ 1.02 │ │ mov x0, x21 ▒ │ │ mov x1, x3 ▒ 1.71 │ │ ldr x20, [x3,#48] ▒ │ │→ bl __fsnotify_parent ▒ 0.68 │ │↑ cbnz 60 ▒ │ │ mov x2, x21 ▒ 1.37 │ │ mov w1, w19 ▒ │ │ mov x0, x20 ▒ 0.68 │ │ mov w5, #0x0 // #0 ▒ │ │ mov x4, #0x0 // #0 ▒ 1.71 │ │ mov w3, #0x1 // #1 ▒ │ │→ bl fsnotify ▒ 1.37 │ │↑ b 60 ▒ │d0:│ mov w0, #0x0 // #0 ▒ │ │ ldp x19, x20, [sp,#16] ▒ │ │ ldp x21, x22, [sp,#32] ▒ │ │ ldp x29, x30, [sp],#48 ▒ │ │← ret ▒ │e4:│ mov w19, #0x10000 // #65536 ▒ │ └──b 90 ◆ │ec: brk #0x800 ▒ Press 'h' for help on key bindings Signed-off-by: Kim Phillips <kim.phillips@arm.com> Signed-off-by: Chris Ryder <chris.ryder@arm.com> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Mark Rutland <mark.rutland@arm.com> Cc: Pawel Moll <pawel.moll@arm.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Will Deacon <will.deacon@arm.com> Link: http://lkml.kernel.org/r/20161130092344.012e18e3e623bea395162f95@arm.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-11-30 18:23:44 +03:00
{
.name = "arm64",
.init = arm64__annotate_init,
},
{
.name = "csky",
.init = csky__annotate_init,
},
perf annotate mips: Add perf arch instructions annotate handlers Support the MIPS architecture using the ins_ops association method. With this patch, perf-annotate can work well on MIPS. Testing it with a perf.data file collected on a mips machine: $./perf annotate -i perf.data : Disassembly of section .text: : : 00000000000be6a0 <get_next_seq>: : get_next_seq(): 0.00 : be6a0: lw v0,0(a0) 0.00 : be6a4: daddiu sp,sp,-128 0.00 : be6a8: ld a7,72(a0) 0.00 : be6ac: gssq s5,s4,80(sp) 0.00 : be6b0: gssq s1,s0,48(sp) 0.00 : be6b4: gssq s8,gp,112(sp) 0.00 : be6b8: gssq s7,s6,96(sp) 0.00 : be6bc: gssq s3,s2,64(sp) 0.00 : be6c0: sd a3,0(sp) 0.00 : be6c4: move s0,a0 0.00 : be6c8: sd v0,32(sp) 0.00 : be6cc: sd a5,8(sp) 0.00 : be6d0: sd zero,8(a0) 0.00 : be6d4: sd a6,16(sp) 0.00 : be6d8: ld s2,48(a0) 8.53 : be6dc: ld s1,40(a0) 9.42 : be6e0: ld v1,32(a0) 0.00 : be6e4: nop 0.00 : be6e8: ld s4,24(a0) 0.00 : be6ec: ld s5,16(a0) 0.00 : be6f0: sd a7,40(sp) 10.11 : be6f4: ld s6,64(a0) ... The original patch link: https://lore.kernel.org/patchwork/patch/1180480/ Signed-off-by: Dengcheng Zhu <dzhu@wavecomp.com> Cc: Dengcheng Zhu <dzhu@wavecomp.com> Cc: Jiaxun Yang <jiaxun.yang@flygoat.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Xuefeng Li <lixuefeng@loongson.cn> Cc: linux-mips@vger.kernel.org [ fanpeng@loongson.cn: Add missing "bgtzl", "bltzl", "bgezl", "blezl", "beql" and "bnel" for pre-R6processors ] Signed-off-by: Peng Fan <fanpeng@loongson.cn> Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2020-10-19 10:21:24 +03:00
{
.name = "mips",
.init = mips__annotate_init,
.objdump = {
.comment_char = '#',
},
},
{
.name = "x86",
.init = x86__annotate_init,
.instructions = x86__instructions,
.nr_instructions = ARRAY_SIZE(x86__instructions),
.objdump = {
.comment_char = '#',
},
},
perf annotate: Initial PowerPC support Support the PowerPC architecture using the ins_ops association method. Committer notes: Testing it with a perf.data file collected on a PowerPC machine and cross-annotated on a x86_64 workstation, using the associated vmlinux file: $ perf report -i perf.data.f22vm.powerdev --vmlinux vmlinux.powerpc .ktime_get vmlinux.powerpc │ clrldi r9,r28,63 8.57 │ ┌──bne e0 <- TUI cursor positioned here │54:│ lwsync 2.86 │ │ std r2,40(r1) │ │ ld r9,144(r31) │ │ ld r3,136(r31) │ │ ld r30,184(r31) │ │ ld r10,0(r9) │ │ mtctr r10 │ │ ld r2,8(r9) 8.57 │ │→ bctrl │ │ ld r2,40(r1) │ │ ld r10,160(r31) │ │ ld r5,152(r31) │ │ lwz r7,168(r31) │ │ ld r9,176(r31) 8.57 │ │ lwz r6,172(r31) │ │ lwsync 2.86 │ │ lwz r8,128(r31) │ │ cmpw cr7,r8,r28 2.86 │ │↑ bne 48 │ │ subf r10,r10,r3 │ │ mr r3,r29 │ │ and r10,r10,r5 2.86 │ │ mulld r10,r10,r7 │ │ add r9,r10,r9 │ │ srd r9,r9,r6 │ │ add r9,r9,r30 │ │ std r9,0(r29) │ │ addi r1,r1,144 │ │ ld r0,16(r1) │ │ ld r28,-32(r1) │ │ ld r29,-24(r1) │ │ ld r30,-16(r1) │ │ mtlr r0 │ │ ld r31,-8(r1) │ │← blr 5.71 │e0:└─→mr r1,r1 11.43 │ mr r2,r2 11.43 │ lwz r28,128(r31) Press 'h' for help on key bindings $ perf report -i perf.data.f22vm.powerdev --header-only # ======== # captured on: Thu Nov 24 12:40:38 2016 # hostname : pdev-f22-qemu # os release : 4.4.10-200.fc22.ppc64 # perf version : 4.9.rc1.g6298ce # arch : ppc64 # nrcpus online : 48 # nrcpus avail : 48 # cpudesc : POWER7 (architected), altivec supported # cpuid : 74,513 # total memory : 4158976 kB # cmdline : /home/ravi/Workspace/linux/tools/perf/perf record -a # event : name = cycles:ppp, , size = 112, { sample_period, sample_freq } = 4000, sample_type = IP|TID|TIME|CPU|PERIOD, disabled = 1, inherit = 1, mmap = 1, comm = 1, freq = 1, task = 1, precise_ip = 3, sample_id_all = 1, exclude_guest = 1, mmap2 = 1, comm_exec = 1 # HEADER_CPU_TOPOLOGY info available, use -I to display # HEADER_NUMA_TOPOLOGY info available, use -I to display # pmu mappings: cpu = 4, software = 1, tracepoint = 2, breakpoint = 5 # missing features: HEADER_TRACING_DATA HEADER_BRANCH_STACK HEADER_GROUP_DESC HEADER_AUXTRACE HEADER_STAT HEADER_CACHE # ======== # $ Signed-off-by: Ravi Bangoria <ravi.bangoria@linux.vnet.ibm.com> Signed-off-by: Naveen N. Rao <naveen.n.rao@linux.vnet.ibm.com> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Kim Phillips <kim.phillips@arm.com> Link: http://lkml.kernel.org/n/tip-tbjnp40ddoxxl474uvhwi6g4@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-11-23 19:03:46 +03:00
{
.name = "powerpc",
.init = powerpc__annotate_init,
},
{
.name = "riscv64",
.init = riscv64__annotate_init,
},
{
.name = "s390",
.init = s390__annotate_init,
.objdump = {
.comment_char = '#',
},
},
{
.name = "sparc",
.init = sparc__annotate_init,
.objdump = {
.comment_char = '#',
},
},
};
static void ins__delete(struct ins_operands *ops)
{
2015-03-05 21:27:28 +03:00
if (ops == NULL)
return;
zfree(&ops->source.raw);
zfree(&ops->source.name);
zfree(&ops->target.raw);
zfree(&ops->target.name);
}
static int ins__raw_scnprintf(struct ins *ins, char *bf, size_t size,
perf annotate: Calculate the max instruction name, align column to that We were hardcoding '6' as the max instruction name, and we have lots that are longer than that, see the diff from two 'P' printed TUI annotations for a libc function that uses instructions with long names, such as 'vpmovmskb' with its 9 chars: --- __strcmp_avx2.annotation.before 2019-03-06 16:31:39.368020425 -0300 +++ __strcmp_avx2.annotation 2019-03-06 16:32:12.079450508 -0300 @@ -2,284 +2,284 @@ Event: cycles:ppp Percent endbr64 - 0.10 mov %edi,%eax + 0.10 mov %edi,%eax - xor %edx,%edx + xor %edx,%edx - 3.54 vpxor %ymm7,%ymm7,%ymm7 + 3.54 vpxor %ymm7,%ymm7,%ymm7 - or %esi,%eax + or %esi,%eax - and $0xfff,%eax + and $0xfff,%eax - cmp $0xf80,%eax + cmp $0xf80,%eax - ↓ jg 370 + ↓ jg 370 - 27.07 vmovdqu (%rdi),%ymm1 + 27.07 vmovdqu (%rdi),%ymm1 - 7.97 vpcmpeqb (%rsi),%ymm1,%ymm0 + 7.97 vpcmpeqb (%rsi),%ymm1,%ymm0 - 2.15 vpminub %ymm1,%ymm0,%ymm0 + 2.15 vpminub %ymm1,%ymm0,%ymm0 - 4.09 vpcmpeqb %ymm7,%ymm0,%ymm0 + 4.09 vpcmpeqb %ymm7,%ymm0,%ymm0 - 0.43 vpmovmskb %ymm0,%ecx + 0.43 vpmovmskb %ymm0,%ecx - 1.53 test %ecx,%ecx + 1.53 test %ecx,%ecx - ↓ je b0 + ↓ je b0 - 5.26 tzcnt %ecx,%edx + 5.26 tzcnt %ecx,%edx - 18.40 movzbl (%rdi,%rdx,1),%eax + 18.40 movzbl (%rdi,%rdx,1),%eax - 7.09 movzbl (%rsi,%rdx,1),%edx + 7.09 movzbl (%rsi,%rdx,1),%edx - 3.34 sub %edx,%eax + 3.34 sub %edx,%eax 2.37 vzeroupper ← retq nop - 50: tzcnt %ecx,%edx + 50: tzcnt %ecx,%edx - movzbl 0x20(%rdi,%rdx,1),%eax + movzbl 0x20(%rdi,%rdx,1),%eax - movzbl 0x20(%rsi,%rdx,1),%edx + movzbl 0x20(%rsi,%rdx,1),%edx - sub %edx,%eax + sub %edx,%eax vzeroupper ← retq - data16 nopw %cs:0x0(%rax,%rax,1) + data16 nopw %cs:0x0(%rax,%rax,1) Reported-by: Travis Downs <travis.downs@gmail.com> LPU-Reference: CAOBGo4z1KfmWeOm6Et0cnX5Z6DWsG2PQbAvRn1MhVPJmXHrc5g@mail.gmail.com Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Link: https://lkml.kernel.org/n/tip-89wsdd9h9g6bvq52sgp6d0u4@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2019-03-06 22:40:15 +03:00
struct ins_operands *ops, int max_ins_name)
{
perf annotate: Calculate the max instruction name, align column to that We were hardcoding '6' as the max instruction name, and we have lots that are longer than that, see the diff from two 'P' printed TUI annotations for a libc function that uses instructions with long names, such as 'vpmovmskb' with its 9 chars: --- __strcmp_avx2.annotation.before 2019-03-06 16:31:39.368020425 -0300 +++ __strcmp_avx2.annotation 2019-03-06 16:32:12.079450508 -0300 @@ -2,284 +2,284 @@ Event: cycles:ppp Percent endbr64 - 0.10 mov %edi,%eax + 0.10 mov %edi,%eax - xor %edx,%edx + xor %edx,%edx - 3.54 vpxor %ymm7,%ymm7,%ymm7 + 3.54 vpxor %ymm7,%ymm7,%ymm7 - or %esi,%eax + or %esi,%eax - and $0xfff,%eax + and $0xfff,%eax - cmp $0xf80,%eax + cmp $0xf80,%eax - ↓ jg 370 + ↓ jg 370 - 27.07 vmovdqu (%rdi),%ymm1 + 27.07 vmovdqu (%rdi),%ymm1 - 7.97 vpcmpeqb (%rsi),%ymm1,%ymm0 + 7.97 vpcmpeqb (%rsi),%ymm1,%ymm0 - 2.15 vpminub %ymm1,%ymm0,%ymm0 + 2.15 vpminub %ymm1,%ymm0,%ymm0 - 4.09 vpcmpeqb %ymm7,%ymm0,%ymm0 + 4.09 vpcmpeqb %ymm7,%ymm0,%ymm0 - 0.43 vpmovmskb %ymm0,%ecx + 0.43 vpmovmskb %ymm0,%ecx - 1.53 test %ecx,%ecx + 1.53 test %ecx,%ecx - ↓ je b0 + ↓ je b0 - 5.26 tzcnt %ecx,%edx + 5.26 tzcnt %ecx,%edx - 18.40 movzbl (%rdi,%rdx,1),%eax + 18.40 movzbl (%rdi,%rdx,1),%eax - 7.09 movzbl (%rsi,%rdx,1),%edx + 7.09 movzbl (%rsi,%rdx,1),%edx - 3.34 sub %edx,%eax + 3.34 sub %edx,%eax 2.37 vzeroupper ← retq nop - 50: tzcnt %ecx,%edx + 50: tzcnt %ecx,%edx - movzbl 0x20(%rdi,%rdx,1),%eax + movzbl 0x20(%rdi,%rdx,1),%eax - movzbl 0x20(%rsi,%rdx,1),%edx + movzbl 0x20(%rsi,%rdx,1),%edx - sub %edx,%eax + sub %edx,%eax vzeroupper ← retq - data16 nopw %cs:0x0(%rax,%rax,1) + data16 nopw %cs:0x0(%rax,%rax,1) Reported-by: Travis Downs <travis.downs@gmail.com> LPU-Reference: CAOBGo4z1KfmWeOm6Et0cnX5Z6DWsG2PQbAvRn1MhVPJmXHrc5g@mail.gmail.com Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Link: https://lkml.kernel.org/n/tip-89wsdd9h9g6bvq52sgp6d0u4@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2019-03-06 22:40:15 +03:00
return scnprintf(bf, size, "%-*s %s", max_ins_name, ins->name, ops->raw);
}
int ins__scnprintf(struct ins *ins, char *bf, size_t size,
perf annotate: Calculate the max instruction name, align column to that We were hardcoding '6' as the max instruction name, and we have lots that are longer than that, see the diff from two 'P' printed TUI annotations for a libc function that uses instructions with long names, such as 'vpmovmskb' with its 9 chars: --- __strcmp_avx2.annotation.before 2019-03-06 16:31:39.368020425 -0300 +++ __strcmp_avx2.annotation 2019-03-06 16:32:12.079450508 -0300 @@ -2,284 +2,284 @@ Event: cycles:ppp Percent endbr64 - 0.10 mov %edi,%eax + 0.10 mov %edi,%eax - xor %edx,%edx + xor %edx,%edx - 3.54 vpxor %ymm7,%ymm7,%ymm7 + 3.54 vpxor %ymm7,%ymm7,%ymm7 - or %esi,%eax + or %esi,%eax - and $0xfff,%eax + and $0xfff,%eax - cmp $0xf80,%eax + cmp $0xf80,%eax - ↓ jg 370 + ↓ jg 370 - 27.07 vmovdqu (%rdi),%ymm1 + 27.07 vmovdqu (%rdi),%ymm1 - 7.97 vpcmpeqb (%rsi),%ymm1,%ymm0 + 7.97 vpcmpeqb (%rsi),%ymm1,%ymm0 - 2.15 vpminub %ymm1,%ymm0,%ymm0 + 2.15 vpminub %ymm1,%ymm0,%ymm0 - 4.09 vpcmpeqb %ymm7,%ymm0,%ymm0 + 4.09 vpcmpeqb %ymm7,%ymm0,%ymm0 - 0.43 vpmovmskb %ymm0,%ecx + 0.43 vpmovmskb %ymm0,%ecx - 1.53 test %ecx,%ecx + 1.53 test %ecx,%ecx - ↓ je b0 + ↓ je b0 - 5.26 tzcnt %ecx,%edx + 5.26 tzcnt %ecx,%edx - 18.40 movzbl (%rdi,%rdx,1),%eax + 18.40 movzbl (%rdi,%rdx,1),%eax - 7.09 movzbl (%rsi,%rdx,1),%edx + 7.09 movzbl (%rsi,%rdx,1),%edx - 3.34 sub %edx,%eax + 3.34 sub %edx,%eax 2.37 vzeroupper ← retq nop - 50: tzcnt %ecx,%edx + 50: tzcnt %ecx,%edx - movzbl 0x20(%rdi,%rdx,1),%eax + movzbl 0x20(%rdi,%rdx,1),%eax - movzbl 0x20(%rsi,%rdx,1),%edx + movzbl 0x20(%rsi,%rdx,1),%edx - sub %edx,%eax + sub %edx,%eax vzeroupper ← retq - data16 nopw %cs:0x0(%rax,%rax,1) + data16 nopw %cs:0x0(%rax,%rax,1) Reported-by: Travis Downs <travis.downs@gmail.com> LPU-Reference: CAOBGo4z1KfmWeOm6Et0cnX5Z6DWsG2PQbAvRn1MhVPJmXHrc5g@mail.gmail.com Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Link: https://lkml.kernel.org/n/tip-89wsdd9h9g6bvq52sgp6d0u4@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2019-03-06 22:40:15 +03:00
struct ins_operands *ops, int max_ins_name)
{
if (ins->ops->scnprintf)
perf annotate: Calculate the max instruction name, align column to that We were hardcoding '6' as the max instruction name, and we have lots that are longer than that, see the diff from two 'P' printed TUI annotations for a libc function that uses instructions with long names, such as 'vpmovmskb' with its 9 chars: --- __strcmp_avx2.annotation.before 2019-03-06 16:31:39.368020425 -0300 +++ __strcmp_avx2.annotation 2019-03-06 16:32:12.079450508 -0300 @@ -2,284 +2,284 @@ Event: cycles:ppp Percent endbr64 - 0.10 mov %edi,%eax + 0.10 mov %edi,%eax - xor %edx,%edx + xor %edx,%edx - 3.54 vpxor %ymm7,%ymm7,%ymm7 + 3.54 vpxor %ymm7,%ymm7,%ymm7 - or %esi,%eax + or %esi,%eax - and $0xfff,%eax + and $0xfff,%eax - cmp $0xf80,%eax + cmp $0xf80,%eax - ↓ jg 370 + ↓ jg 370 - 27.07 vmovdqu (%rdi),%ymm1 + 27.07 vmovdqu (%rdi),%ymm1 - 7.97 vpcmpeqb (%rsi),%ymm1,%ymm0 + 7.97 vpcmpeqb (%rsi),%ymm1,%ymm0 - 2.15 vpminub %ymm1,%ymm0,%ymm0 + 2.15 vpminub %ymm1,%ymm0,%ymm0 - 4.09 vpcmpeqb %ymm7,%ymm0,%ymm0 + 4.09 vpcmpeqb %ymm7,%ymm0,%ymm0 - 0.43 vpmovmskb %ymm0,%ecx + 0.43 vpmovmskb %ymm0,%ecx - 1.53 test %ecx,%ecx + 1.53 test %ecx,%ecx - ↓ je b0 + ↓ je b0 - 5.26 tzcnt %ecx,%edx + 5.26 tzcnt %ecx,%edx - 18.40 movzbl (%rdi,%rdx,1),%eax + 18.40 movzbl (%rdi,%rdx,1),%eax - 7.09 movzbl (%rsi,%rdx,1),%edx + 7.09 movzbl (%rsi,%rdx,1),%edx - 3.34 sub %edx,%eax + 3.34 sub %edx,%eax 2.37 vzeroupper ← retq nop - 50: tzcnt %ecx,%edx + 50: tzcnt %ecx,%edx - movzbl 0x20(%rdi,%rdx,1),%eax + movzbl 0x20(%rdi,%rdx,1),%eax - movzbl 0x20(%rsi,%rdx,1),%edx + movzbl 0x20(%rsi,%rdx,1),%edx - sub %edx,%eax + sub %edx,%eax vzeroupper ← retq - data16 nopw %cs:0x0(%rax,%rax,1) + data16 nopw %cs:0x0(%rax,%rax,1) Reported-by: Travis Downs <travis.downs@gmail.com> LPU-Reference: CAOBGo4z1KfmWeOm6Et0cnX5Z6DWsG2PQbAvRn1MhVPJmXHrc5g@mail.gmail.com Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Link: https://lkml.kernel.org/n/tip-89wsdd9h9g6bvq52sgp6d0u4@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2019-03-06 22:40:15 +03:00
return ins->ops->scnprintf(ins, bf, size, ops, max_ins_name);
perf annotate: Calculate the max instruction name, align column to that We were hardcoding '6' as the max instruction name, and we have lots that are longer than that, see the diff from two 'P' printed TUI annotations for a libc function that uses instructions with long names, such as 'vpmovmskb' with its 9 chars: --- __strcmp_avx2.annotation.before 2019-03-06 16:31:39.368020425 -0300 +++ __strcmp_avx2.annotation 2019-03-06 16:32:12.079450508 -0300 @@ -2,284 +2,284 @@ Event: cycles:ppp Percent endbr64 - 0.10 mov %edi,%eax + 0.10 mov %edi,%eax - xor %edx,%edx + xor %edx,%edx - 3.54 vpxor %ymm7,%ymm7,%ymm7 + 3.54 vpxor %ymm7,%ymm7,%ymm7 - or %esi,%eax + or %esi,%eax - and $0xfff,%eax + and $0xfff,%eax - cmp $0xf80,%eax + cmp $0xf80,%eax - ↓ jg 370 + ↓ jg 370 - 27.07 vmovdqu (%rdi),%ymm1 + 27.07 vmovdqu (%rdi),%ymm1 - 7.97 vpcmpeqb (%rsi),%ymm1,%ymm0 + 7.97 vpcmpeqb (%rsi),%ymm1,%ymm0 - 2.15 vpminub %ymm1,%ymm0,%ymm0 + 2.15 vpminub %ymm1,%ymm0,%ymm0 - 4.09 vpcmpeqb %ymm7,%ymm0,%ymm0 + 4.09 vpcmpeqb %ymm7,%ymm0,%ymm0 - 0.43 vpmovmskb %ymm0,%ecx + 0.43 vpmovmskb %ymm0,%ecx - 1.53 test %ecx,%ecx + 1.53 test %ecx,%ecx - ↓ je b0 + ↓ je b0 - 5.26 tzcnt %ecx,%edx + 5.26 tzcnt %ecx,%edx - 18.40 movzbl (%rdi,%rdx,1),%eax + 18.40 movzbl (%rdi,%rdx,1),%eax - 7.09 movzbl (%rsi,%rdx,1),%edx + 7.09 movzbl (%rsi,%rdx,1),%edx - 3.34 sub %edx,%eax + 3.34 sub %edx,%eax 2.37 vzeroupper ← retq nop - 50: tzcnt %ecx,%edx + 50: tzcnt %ecx,%edx - movzbl 0x20(%rdi,%rdx,1),%eax + movzbl 0x20(%rdi,%rdx,1),%eax - movzbl 0x20(%rsi,%rdx,1),%edx + movzbl 0x20(%rsi,%rdx,1),%edx - sub %edx,%eax + sub %edx,%eax vzeroupper ← retq - data16 nopw %cs:0x0(%rax,%rax,1) + data16 nopw %cs:0x0(%rax,%rax,1) Reported-by: Travis Downs <travis.downs@gmail.com> LPU-Reference: CAOBGo4z1KfmWeOm6Et0cnX5Z6DWsG2PQbAvRn1MhVPJmXHrc5g@mail.gmail.com Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Link: https://lkml.kernel.org/n/tip-89wsdd9h9g6bvq52sgp6d0u4@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2019-03-06 22:40:15 +03:00
return ins__raw_scnprintf(ins, bf, size, ops, max_ins_name);
}
perf annotate: Check for fused instructions Macro fusion merges two instructions to a single micro-op. Intel core platform performs this hardware optimization under limited circumstances. For example, CMP + JCC can be "fused" and executed /retired together. While with sampling this can result in the sample sometimes being on the JCC and sometimes on the CMP. So for the fused instruction pair, they could be considered together. On Nehalem, fused instruction pairs: cmp/test + jcc. On other new CPU: cmp/test/add/sub/and/inc/dec + jcc. This patch adds an x86-specific function which checks if 2 instructions are in a "fused" pair. For non-x86 arch, the function is just NULL. Changelog: v4: Move the CPU model checking to symbol__disassemble and save the CPU family/model in arch structure. It avoids checking every time when jump arrow printed. v3: Add checking for Nehalem (CMP, TEST). For other newer Intel CPUs just check it by default (CMP, TEST, ADD, SUB, AND, INC, DEC). v2: Remove the original weak function. Arnaldo points out that doing it as a weak function that will be overridden by the host arch doesn't work. So now it's implemented as an arch-specific function. Committer fix: Do not access evsel->evlist->env->cpuid, ->env can be null, introduce perf_evsel__env_cpuid(), just like perf_evsel__env_arch(), also used in this function call. The original patch was segfaulting 'perf top' + annotation. But this essentially disables this fused instructions augmentation in 'perf top', the right thing is to get the cpuid from the running kernel, left for a later patch tho. Signed-off-by: Yao Jin <yao.jin@linux.intel.com> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Andi Kleen <ak@linux.intel.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Kan Liang <kan.liang@intel.com> Cc: Peter Zijlstra <peterz@infradead.org> Link: http://lkml.kernel.org/r/1499403995-19857-2-git-send-email-yao.jin@linux.intel.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2017-07-07 08:06:34 +03:00
bool ins__is_fused(struct arch *arch, const char *ins1, const char *ins2)
{
if (!arch || !arch->ins_is_fused)
return false;
return arch->ins_is_fused(arch, ins1, ins2);
}
static int call__parse(struct arch *arch, struct ins_operands *ops, struct map_symbol *ms)
{
char *endptr, *tok, *name;
struct map *map = ms->map;
struct addr_map_symbol target = {
.ms = { .map = map, },
};
ops->target.addr = strtoull(ops->raw, &endptr, 16);
name = strchr(endptr, '<');
if (name == NULL)
goto indirect_call;
name++;
if (arch->objdump.skip_functions_char &&
strchr(name, arch->objdump.skip_functions_char))
return -1;
tok = strchr(name, '>');
if (tok == NULL)
return -1;
*tok = '\0';
ops->target.name = strdup(name);
*tok = '>';
if (ops->target.name == NULL)
return -1;
find_target:
target.addr = map__objdump_2mem(map, ops->target.addr);
if (maps__find_ams(ms->maps, &target) == 0 &&
map__rip_2objdump(target.ms.map, map->map_ip(target.ms.map, target.addr)) == ops->target.addr)
ops->target.sym = target.ms.sym;
return 0;
indirect_call:
tok = strchr(endptr, '*');
if (tok != NULL) {
endptr++;
/* Indirect call can use a non-rip register and offset: callq *0x8(%rbx).
* Do not parse such instruction. */
if (strstr(endptr, "(%r") == NULL)
ops->target.addr = strtoull(endptr, NULL, 16);
}
goto find_target;
}
static int call__scnprintf(struct ins *ins, char *bf, size_t size,
perf annotate: Calculate the max instruction name, align column to that We were hardcoding '6' as the max instruction name, and we have lots that are longer than that, see the diff from two 'P' printed TUI annotations for a libc function that uses instructions with long names, such as 'vpmovmskb' with its 9 chars: --- __strcmp_avx2.annotation.before 2019-03-06 16:31:39.368020425 -0300 +++ __strcmp_avx2.annotation 2019-03-06 16:32:12.079450508 -0300 @@ -2,284 +2,284 @@ Event: cycles:ppp Percent endbr64 - 0.10 mov %edi,%eax + 0.10 mov %edi,%eax - xor %edx,%edx + xor %edx,%edx - 3.54 vpxor %ymm7,%ymm7,%ymm7 + 3.54 vpxor %ymm7,%ymm7,%ymm7 - or %esi,%eax + or %esi,%eax - and $0xfff,%eax + and $0xfff,%eax - cmp $0xf80,%eax + cmp $0xf80,%eax - ↓ jg 370 + ↓ jg 370 - 27.07 vmovdqu (%rdi),%ymm1 + 27.07 vmovdqu (%rdi),%ymm1 - 7.97 vpcmpeqb (%rsi),%ymm1,%ymm0 + 7.97 vpcmpeqb (%rsi),%ymm1,%ymm0 - 2.15 vpminub %ymm1,%ymm0,%ymm0 + 2.15 vpminub %ymm1,%ymm0,%ymm0 - 4.09 vpcmpeqb %ymm7,%ymm0,%ymm0 + 4.09 vpcmpeqb %ymm7,%ymm0,%ymm0 - 0.43 vpmovmskb %ymm0,%ecx + 0.43 vpmovmskb %ymm0,%ecx - 1.53 test %ecx,%ecx + 1.53 test %ecx,%ecx - ↓ je b0 + ↓ je b0 - 5.26 tzcnt %ecx,%edx + 5.26 tzcnt %ecx,%edx - 18.40 movzbl (%rdi,%rdx,1),%eax + 18.40 movzbl (%rdi,%rdx,1),%eax - 7.09 movzbl (%rsi,%rdx,1),%edx + 7.09 movzbl (%rsi,%rdx,1),%edx - 3.34 sub %edx,%eax + 3.34 sub %edx,%eax 2.37 vzeroupper ← retq nop - 50: tzcnt %ecx,%edx + 50: tzcnt %ecx,%edx - movzbl 0x20(%rdi,%rdx,1),%eax + movzbl 0x20(%rdi,%rdx,1),%eax - movzbl 0x20(%rsi,%rdx,1),%edx + movzbl 0x20(%rsi,%rdx,1),%edx - sub %edx,%eax + sub %edx,%eax vzeroupper ← retq - data16 nopw %cs:0x0(%rax,%rax,1) + data16 nopw %cs:0x0(%rax,%rax,1) Reported-by: Travis Downs <travis.downs@gmail.com> LPU-Reference: CAOBGo4z1KfmWeOm6Et0cnX5Z6DWsG2PQbAvRn1MhVPJmXHrc5g@mail.gmail.com Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Link: https://lkml.kernel.org/n/tip-89wsdd9h9g6bvq52sgp6d0u4@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2019-03-06 22:40:15 +03:00
struct ins_operands *ops, int max_ins_name)
{
if (ops->target.sym)
perf annotate: Calculate the max instruction name, align column to that We were hardcoding '6' as the max instruction name, and we have lots that are longer than that, see the diff from two 'P' printed TUI annotations for a libc function that uses instructions with long names, such as 'vpmovmskb' with its 9 chars: --- __strcmp_avx2.annotation.before 2019-03-06 16:31:39.368020425 -0300 +++ __strcmp_avx2.annotation 2019-03-06 16:32:12.079450508 -0300 @@ -2,284 +2,284 @@ Event: cycles:ppp Percent endbr64 - 0.10 mov %edi,%eax + 0.10 mov %edi,%eax - xor %edx,%edx + xor %edx,%edx - 3.54 vpxor %ymm7,%ymm7,%ymm7 + 3.54 vpxor %ymm7,%ymm7,%ymm7 - or %esi,%eax + or %esi,%eax - and $0xfff,%eax + and $0xfff,%eax - cmp $0xf80,%eax + cmp $0xf80,%eax - ↓ jg 370 + ↓ jg 370 - 27.07 vmovdqu (%rdi),%ymm1 + 27.07 vmovdqu (%rdi),%ymm1 - 7.97 vpcmpeqb (%rsi),%ymm1,%ymm0 + 7.97 vpcmpeqb (%rsi),%ymm1,%ymm0 - 2.15 vpminub %ymm1,%ymm0,%ymm0 + 2.15 vpminub %ymm1,%ymm0,%ymm0 - 4.09 vpcmpeqb %ymm7,%ymm0,%ymm0 + 4.09 vpcmpeqb %ymm7,%ymm0,%ymm0 - 0.43 vpmovmskb %ymm0,%ecx + 0.43 vpmovmskb %ymm0,%ecx - 1.53 test %ecx,%ecx + 1.53 test %ecx,%ecx - ↓ je b0 + ↓ je b0 - 5.26 tzcnt %ecx,%edx + 5.26 tzcnt %ecx,%edx - 18.40 movzbl (%rdi,%rdx,1),%eax + 18.40 movzbl (%rdi,%rdx,1),%eax - 7.09 movzbl (%rsi,%rdx,1),%edx + 7.09 movzbl (%rsi,%rdx,1),%edx - 3.34 sub %edx,%eax + 3.34 sub %edx,%eax 2.37 vzeroupper ← retq nop - 50: tzcnt %ecx,%edx + 50: tzcnt %ecx,%edx - movzbl 0x20(%rdi,%rdx,1),%eax + movzbl 0x20(%rdi,%rdx,1),%eax - movzbl 0x20(%rsi,%rdx,1),%edx + movzbl 0x20(%rsi,%rdx,1),%edx - sub %edx,%eax + sub %edx,%eax vzeroupper ← retq - data16 nopw %cs:0x0(%rax,%rax,1) + data16 nopw %cs:0x0(%rax,%rax,1) Reported-by: Travis Downs <travis.downs@gmail.com> LPU-Reference: CAOBGo4z1KfmWeOm6Et0cnX5Z6DWsG2PQbAvRn1MhVPJmXHrc5g@mail.gmail.com Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Link: https://lkml.kernel.org/n/tip-89wsdd9h9g6bvq52sgp6d0u4@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2019-03-06 22:40:15 +03:00
return scnprintf(bf, size, "%-*s %s", max_ins_name, ins->name, ops->target.sym->name);
if (ops->target.addr == 0)
perf annotate: Calculate the max instruction name, align column to that We were hardcoding '6' as the max instruction name, and we have lots that are longer than that, see the diff from two 'P' printed TUI annotations for a libc function that uses instructions with long names, such as 'vpmovmskb' with its 9 chars: --- __strcmp_avx2.annotation.before 2019-03-06 16:31:39.368020425 -0300 +++ __strcmp_avx2.annotation 2019-03-06 16:32:12.079450508 -0300 @@ -2,284 +2,284 @@ Event: cycles:ppp Percent endbr64 - 0.10 mov %edi,%eax + 0.10 mov %edi,%eax - xor %edx,%edx + xor %edx,%edx - 3.54 vpxor %ymm7,%ymm7,%ymm7 + 3.54 vpxor %ymm7,%ymm7,%ymm7 - or %esi,%eax + or %esi,%eax - and $0xfff,%eax + and $0xfff,%eax - cmp $0xf80,%eax + cmp $0xf80,%eax - ↓ jg 370 + ↓ jg 370 - 27.07 vmovdqu (%rdi),%ymm1 + 27.07 vmovdqu (%rdi),%ymm1 - 7.97 vpcmpeqb (%rsi),%ymm1,%ymm0 + 7.97 vpcmpeqb (%rsi),%ymm1,%ymm0 - 2.15 vpminub %ymm1,%ymm0,%ymm0 + 2.15 vpminub %ymm1,%ymm0,%ymm0 - 4.09 vpcmpeqb %ymm7,%ymm0,%ymm0 + 4.09 vpcmpeqb %ymm7,%ymm0,%ymm0 - 0.43 vpmovmskb %ymm0,%ecx + 0.43 vpmovmskb %ymm0,%ecx - 1.53 test %ecx,%ecx + 1.53 test %ecx,%ecx - ↓ je b0 + ↓ je b0 - 5.26 tzcnt %ecx,%edx + 5.26 tzcnt %ecx,%edx - 18.40 movzbl (%rdi,%rdx,1),%eax + 18.40 movzbl (%rdi,%rdx,1),%eax - 7.09 movzbl (%rsi,%rdx,1),%edx + 7.09 movzbl (%rsi,%rdx,1),%edx - 3.34 sub %edx,%eax + 3.34 sub %edx,%eax 2.37 vzeroupper ← retq nop - 50: tzcnt %ecx,%edx + 50: tzcnt %ecx,%edx - movzbl 0x20(%rdi,%rdx,1),%eax + movzbl 0x20(%rdi,%rdx,1),%eax - movzbl 0x20(%rsi,%rdx,1),%edx + movzbl 0x20(%rsi,%rdx,1),%edx - sub %edx,%eax + sub %edx,%eax vzeroupper ← retq - data16 nopw %cs:0x0(%rax,%rax,1) + data16 nopw %cs:0x0(%rax,%rax,1) Reported-by: Travis Downs <travis.downs@gmail.com> LPU-Reference: CAOBGo4z1KfmWeOm6Et0cnX5Z6DWsG2PQbAvRn1MhVPJmXHrc5g@mail.gmail.com Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Link: https://lkml.kernel.org/n/tip-89wsdd9h9g6bvq52sgp6d0u4@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2019-03-06 22:40:15 +03:00
return ins__raw_scnprintf(ins, bf, size, ops, max_ins_name);
perf annotate: Use ops->target.name when available for unresolved call targets There is a bug where when using 'perf annotate timerqueue_add' the target for its only routine called with the 'callq' instruction, 'rb_insert_color', doesn't get resolved from its address when parsing that 'callq' instruction. That symbol resolution works when using 'perf report --tui' and then doing annotation for 'timerqueue_add' from there, the vmlinux dso->symbols rb_tree somehow gets in a state that we can't find that address, that is a bug that has to be further investigated. But since the objdump output has the function name, i.e. the raw objdump disassembled line looks like: So, before: # perf annotate timerqueue_add │ mov %rbx,%rdi │ mov %rbx,(%rdx) │ → callq *ffffffff8184dc80 │ mov 0x8(%rbp),%rdx │ test %rdx,%rdx │ ↓ je 67 # perf report │ mov %rbx,%rdi │ mov %rbx,(%rdx) │ → callq rb_insert_color │ mov 0x8(%rbp),%rdx │ test %rdx,%rdx │ ↓ je 67 And after both look the same: # perf annotate timerqueue_add │ mov %rbx,%rdi │ mov %rbx,(%rdx) │ → callq rb_insert_color │ mov 0x8(%rbp),%rdx │ test %rdx,%rdx │ ↓ je 67 From 'perf report' one can annotate and navigate to that 'rb_insert_color' function, but not directly from 'perf annotate timerqueue_add', that remains to be investigated and fixed. Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Andi Kleen <ak@linux.intel.com> Cc: David Ahern <dsahern@gmail.com> Cc: Jin Yao <yao.jin@linux.intel.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Wang Nan <wangnan0@huawei.com> Link: https://lkml.kernel.org/n/tip-nkktz6355rhqtq7o8atr8f8r@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2018-03-16 19:28:09 +03:00
if (ops->target.name)
perf annotate: Calculate the max instruction name, align column to that We were hardcoding '6' as the max instruction name, and we have lots that are longer than that, see the diff from two 'P' printed TUI annotations for a libc function that uses instructions with long names, such as 'vpmovmskb' with its 9 chars: --- __strcmp_avx2.annotation.before 2019-03-06 16:31:39.368020425 -0300 +++ __strcmp_avx2.annotation 2019-03-06 16:32:12.079450508 -0300 @@ -2,284 +2,284 @@ Event: cycles:ppp Percent endbr64 - 0.10 mov %edi,%eax + 0.10 mov %edi,%eax - xor %edx,%edx + xor %edx,%edx - 3.54 vpxor %ymm7,%ymm7,%ymm7 + 3.54 vpxor %ymm7,%ymm7,%ymm7 - or %esi,%eax + or %esi,%eax - and $0xfff,%eax + and $0xfff,%eax - cmp $0xf80,%eax + cmp $0xf80,%eax - ↓ jg 370 + ↓ jg 370 - 27.07 vmovdqu (%rdi),%ymm1 + 27.07 vmovdqu (%rdi),%ymm1 - 7.97 vpcmpeqb (%rsi),%ymm1,%ymm0 + 7.97 vpcmpeqb (%rsi),%ymm1,%ymm0 - 2.15 vpminub %ymm1,%ymm0,%ymm0 + 2.15 vpminub %ymm1,%ymm0,%ymm0 - 4.09 vpcmpeqb %ymm7,%ymm0,%ymm0 + 4.09 vpcmpeqb %ymm7,%ymm0,%ymm0 - 0.43 vpmovmskb %ymm0,%ecx + 0.43 vpmovmskb %ymm0,%ecx - 1.53 test %ecx,%ecx + 1.53 test %ecx,%ecx - ↓ je b0 + ↓ je b0 - 5.26 tzcnt %ecx,%edx + 5.26 tzcnt %ecx,%edx - 18.40 movzbl (%rdi,%rdx,1),%eax + 18.40 movzbl (%rdi,%rdx,1),%eax - 7.09 movzbl (%rsi,%rdx,1),%edx + 7.09 movzbl (%rsi,%rdx,1),%edx - 3.34 sub %edx,%eax + 3.34 sub %edx,%eax 2.37 vzeroupper ← retq nop - 50: tzcnt %ecx,%edx + 50: tzcnt %ecx,%edx - movzbl 0x20(%rdi,%rdx,1),%eax + movzbl 0x20(%rdi,%rdx,1),%eax - movzbl 0x20(%rsi,%rdx,1),%edx + movzbl 0x20(%rsi,%rdx,1),%edx - sub %edx,%eax + sub %edx,%eax vzeroupper ← retq - data16 nopw %cs:0x0(%rax,%rax,1) + data16 nopw %cs:0x0(%rax,%rax,1) Reported-by: Travis Downs <travis.downs@gmail.com> LPU-Reference: CAOBGo4z1KfmWeOm6Et0cnX5Z6DWsG2PQbAvRn1MhVPJmXHrc5g@mail.gmail.com Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Link: https://lkml.kernel.org/n/tip-89wsdd9h9g6bvq52sgp6d0u4@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2019-03-06 22:40:15 +03:00
return scnprintf(bf, size, "%-*s %s", max_ins_name, ins->name, ops->target.name);
perf annotate: Use ops->target.name when available for unresolved call targets There is a bug where when using 'perf annotate timerqueue_add' the target for its only routine called with the 'callq' instruction, 'rb_insert_color', doesn't get resolved from its address when parsing that 'callq' instruction. That symbol resolution works when using 'perf report --tui' and then doing annotation for 'timerqueue_add' from there, the vmlinux dso->symbols rb_tree somehow gets in a state that we can't find that address, that is a bug that has to be further investigated. But since the objdump output has the function name, i.e. the raw objdump disassembled line looks like: So, before: # perf annotate timerqueue_add │ mov %rbx,%rdi │ mov %rbx,(%rdx) │ → callq *ffffffff8184dc80 │ mov 0x8(%rbp),%rdx │ test %rdx,%rdx │ ↓ je 67 # perf report │ mov %rbx,%rdi │ mov %rbx,(%rdx) │ → callq rb_insert_color │ mov 0x8(%rbp),%rdx │ test %rdx,%rdx │ ↓ je 67 And after both look the same: # perf annotate timerqueue_add │ mov %rbx,%rdi │ mov %rbx,(%rdx) │ → callq rb_insert_color │ mov 0x8(%rbp),%rdx │ test %rdx,%rdx │ ↓ je 67 From 'perf report' one can annotate and navigate to that 'rb_insert_color' function, but not directly from 'perf annotate timerqueue_add', that remains to be investigated and fixed. Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Andi Kleen <ak@linux.intel.com> Cc: David Ahern <dsahern@gmail.com> Cc: Jin Yao <yao.jin@linux.intel.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Wang Nan <wangnan0@huawei.com> Link: https://lkml.kernel.org/n/tip-nkktz6355rhqtq7o8atr8f8r@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2018-03-16 19:28:09 +03:00
perf annotate: Calculate the max instruction name, align column to that We were hardcoding '6' as the max instruction name, and we have lots that are longer than that, see the diff from two 'P' printed TUI annotations for a libc function that uses instructions with long names, such as 'vpmovmskb' with its 9 chars: --- __strcmp_avx2.annotation.before 2019-03-06 16:31:39.368020425 -0300 +++ __strcmp_avx2.annotation 2019-03-06 16:32:12.079450508 -0300 @@ -2,284 +2,284 @@ Event: cycles:ppp Percent endbr64 - 0.10 mov %edi,%eax + 0.10 mov %edi,%eax - xor %edx,%edx + xor %edx,%edx - 3.54 vpxor %ymm7,%ymm7,%ymm7 + 3.54 vpxor %ymm7,%ymm7,%ymm7 - or %esi,%eax + or %esi,%eax - and $0xfff,%eax + and $0xfff,%eax - cmp $0xf80,%eax + cmp $0xf80,%eax - ↓ jg 370 + ↓ jg 370 - 27.07 vmovdqu (%rdi),%ymm1 + 27.07 vmovdqu (%rdi),%ymm1 - 7.97 vpcmpeqb (%rsi),%ymm1,%ymm0 + 7.97 vpcmpeqb (%rsi),%ymm1,%ymm0 - 2.15 vpminub %ymm1,%ymm0,%ymm0 + 2.15 vpminub %ymm1,%ymm0,%ymm0 - 4.09 vpcmpeqb %ymm7,%ymm0,%ymm0 + 4.09 vpcmpeqb %ymm7,%ymm0,%ymm0 - 0.43 vpmovmskb %ymm0,%ecx + 0.43 vpmovmskb %ymm0,%ecx - 1.53 test %ecx,%ecx + 1.53 test %ecx,%ecx - ↓ je b0 + ↓ je b0 - 5.26 tzcnt %ecx,%edx + 5.26 tzcnt %ecx,%edx - 18.40 movzbl (%rdi,%rdx,1),%eax + 18.40 movzbl (%rdi,%rdx,1),%eax - 7.09 movzbl (%rsi,%rdx,1),%edx + 7.09 movzbl (%rsi,%rdx,1),%edx - 3.34 sub %edx,%eax + 3.34 sub %edx,%eax 2.37 vzeroupper ← retq nop - 50: tzcnt %ecx,%edx + 50: tzcnt %ecx,%edx - movzbl 0x20(%rdi,%rdx,1),%eax + movzbl 0x20(%rdi,%rdx,1),%eax - movzbl 0x20(%rsi,%rdx,1),%edx + movzbl 0x20(%rsi,%rdx,1),%edx - sub %edx,%eax + sub %edx,%eax vzeroupper ← retq - data16 nopw %cs:0x0(%rax,%rax,1) + data16 nopw %cs:0x0(%rax,%rax,1) Reported-by: Travis Downs <travis.downs@gmail.com> LPU-Reference: CAOBGo4z1KfmWeOm6Et0cnX5Z6DWsG2PQbAvRn1MhVPJmXHrc5g@mail.gmail.com Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Link: https://lkml.kernel.org/n/tip-89wsdd9h9g6bvq52sgp6d0u4@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2019-03-06 22:40:15 +03:00
return scnprintf(bf, size, "%-*s *%" PRIx64, max_ins_name, ins->name, ops->target.addr);
}
static struct ins_ops call_ops = {
.parse = call__parse,
.scnprintf = call__scnprintf,
};
bool ins__is_call(const struct ins *ins)
{
return ins->ops == &call_ops || ins->ops == &s390_call_ops;
}
perf annotate: Fix parsing aarch64 branch instructions after objdump update Starting with binutils 2.28, aarch64 objdump adds comments to the disassembly output to show the alternative names of a condition code [1]. It is assumed that commas in objdump comments could occur in other arches now or in the future, so this fix is arch-independent. The fix could have been done with arm64 specific jump__parse and jump__scnprintf functions, but the jump__scnprintf instruction would have to have its comment character be a literal, since the scnprintf functions cannot receive a struct arch easily. This inconvenience also applies to the generic jump__scnprintf, which is why we add a raw_comment pointer to struct ins_operands, so the __parse function assigns it to be re-used by its corresponding __scnprintf function. Example differences in 'perf annotate --stdio2' output on an aarch64 perf.data file: BEFORE: → b.cs ffff200008133d1c <unwind_frame+0x18c> // b.hs, dffff7ecc47b AFTER : ↓ b.cs 18c BEFORE: → b.cc ffff200008d8d9cc <get_alloc_profile+0x31c> // b.lo, b.ul, dffff727295b AFTER : ↓ b.cc 31c The branch target labels 18c and 31c also now appear in the output: BEFORE: add x26, x29, #0x80 AFTER : 18c: add x26, x29, #0x80 BEFORE: add x21, x21, #0x8 AFTER : 31c: add x21, x21, #0x8 The Fixes: tag below is added so stable branches will get the update; it doesn't necessarily mean that commit was broken at the time, rather it didn't withstand the aarch64 objdump update. Tested no difference in output for sample x86_64, power arch perf.data files. [1] https://sourceware.org/git/gitweb.cgi?p=binutils-gdb.git;a=commit;h=bb7eff5206e4795ac79c177a80fe9f4630aaf730 Signed-off-by: Kim Phillips <kim.phillips@arm.com> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Anton Blanchard <anton@samba.org> Cc: Christian Borntraeger <borntraeger@de.ibm.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Mark Rutland <mark.rutland@arm.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Ravi Bangoria <ravi.bangoria@linux.vnet.ibm.com> Cc: Robin Murphy <robin.murphy@arm.com> Cc: Taeung Song <treeze.taeung@gmail.com> Cc: linux-arm-kernel@lists.infradead.org Fixes: b13bbeee5ee6 ("perf annotate: Fix branch instruction with multiple operands") Link: http://lkml.kernel.org/r/20180827125340.a2f7e291901d17cea05daba4@arm.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2018-08-27 20:53:40 +03:00
/*
* Prevents from matching commas in the comment section, e.g.:
* ffff200008446e70: b.cs ffff2000084470f4 <generic_exec_single+0x314> // b.hs, b.nlast
perf annotate: Fix jump parsing for C++ code. Considering the following testcase: int foo(int a, int b) { for (unsigned i = 0; i < 1000000000; i++) a += b; return a; } int main() { foo (3, 4); return 0; } 'perf annotate' displays: 86.52 │40055e: → ja 40056c <foo(int, int)+0x26> 13.37 │400560: mov -0x18(%rbp),%eax │400563: add %eax,-0x14(%rbp) │400566: addl $0x1,-0x4(%rbp) 0.11 │40056a: → jmp 400557 <foo(int, int)+0x11> │40056c: mov -0x14(%rbp),%eax │40056f: pop %rbp and the 'ja 40056c' does not link to the location in the function. It's caused by fact that comma is wrongly parsed, it's part of function signature. With my patch I see: 86.52 │ ┌──ja 26 13.37 │ │ mov -0x18(%rbp),%eax │ │ add %eax,-0x14(%rbp) │ │ addl $0x1,-0x4(%rbp) 0.11 │ │↑ jmp 11 │26:└─→mov -0x14(%rbp),%eax and 'o' output prints: 86.52 │4005┌── ↓ ja 40056c <foo(int, int)+0x26> 13.37 │4005│0: mov -0x18(%rbp),%eax │4005│3: add %eax,-0x14(%rbp) │4005│6: addl $0x1,-0x4(%rbp) 0.11 │4005│a: ↑ jmp 400557 <foo(int, int)+0x11> │4005└─→ mov -0x14(%rbp),%eax On the contrary, compiling the very same file with gcc -x c, the parsing is fine because function arguments are not displayed: jmp 400543 <foo+0x1d> Committer testing: Before: $ cat cpp_args_annotate.c int foo(int a, int b) { for (unsigned i = 0; i < 1000000000; i++) a += b; return a; } int main() { foo (3, 4); return 0; } $ gcc --version |& head -1 gcc (GCC) 10.2.1 20201125 (Red Hat 10.2.1-9) $ gcc -g cpp_args_annotate.c -o cpp_args_annotate $ perf record ./cpp_args_annotate [ perf record: Woken up 2 times to write data ] [ perf record: Captured and wrote 0.275 MB perf.data (7188 samples) ] $ perf annotate --stdio2 foo Samples: 7K of event 'cycles:u', 4000 Hz, Event count (approx.): 7468429289, [percent: local period] foo() /home/acme/c/cpp_args_annotate Percent 0000000000401106 <foo>: foo(): int foo(int a, int b) { push %rbp mov %rsp,%rbp mov %edi,-0x14(%rbp) mov %esi,-0x18(%rbp) for (unsigned i = 0; i < 1000000000; i++) movl $0x0,-0x4(%rbp) ↓ jmp 1d a += b; 13.45 13: mov -0x18(%rbp),%eax add %eax,-0x14(%rbp) for (unsigned i = 0; i < 1000000000; i++) addl $0x1,-0x4(%rbp) 0.09 1d: cmpl $0x3b9ac9ff,-0x4(%rbp) 86.46 ↑ jbe 13 return a; mov -0x14(%rbp),%eax } pop %rbp ← retq $ I.e. works for C, now lets switch to C++: $ g++ -g cpp_args_annotate.c -o cpp_args_annotate $ perf record ./cpp_args_annotate [ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 0.268 MB perf.data (6976 samples) ] $ perf annotate --stdio2 foo Samples: 6K of event 'cycles:u', 4000 Hz, Event count (approx.): 7380681761, [percent: local period] foo() /home/acme/c/cpp_args_annotate Percent 0000000000401106 <foo(int, int)>: foo(int, int): int foo(int a, int b) { push %rbp mov %rsp,%rbp mov %edi,-0x14(%rbp) mov %esi,-0x18(%rbp) for (unsigned i = 0; i < 1000000000; i++) movl $0x0,-0x4(%rbp) cmpl $0x3b9ac9ff,-0x4(%rbp) 86.53 → ja 40112c <foo(int, int)+0x26> a += b; 13.32 mov -0x18(%rbp),%eax 0.00 add %eax,-0x14(%rbp) for (unsigned i = 0; i < 1000000000; i++) addl $0x1,-0x4(%rbp) 0.15 → jmp 401117 <foo(int, int)+0x11> return a; mov -0x14(%rbp),%eax } pop %rbp ← retq $ Reproduced. Now with this patch: Reusing the C++ built binary, as we can see here: $ readelf -wi cpp_args_annotate | grep producer <c> DW_AT_producer : (indirect string, offset: 0x2e): GNU C++14 10.2.1 20201125 (Red Hat 10.2.1-9) -mtune=generic -march=x86-64 -g $ And furthermore: $ file cpp_args_annotate cpp_args_annotate: ELF 64-bit LSB executable, x86-64, version 1 (SYSV), dynamically linked, interpreter /lib64/ld-linux-x86-64.so.2, BuildID[sha1]=4fe3cab260204765605ec630d0dc7a7e93c361a9, for GNU/Linux 3.2.0, with debug_info, not stripped $ perf buildid-list -i cpp_args_annotate 4fe3cab260204765605ec630d0dc7a7e93c361a9 $ perf buildid-list | grep cpp_args_annotate 4fe3cab260204765605ec630d0dc7a7e93c361a9 /home/acme/c/cpp_args_annotate $ It now works: $ perf annotate --stdio2 foo Samples: 6K of event 'cycles:u', 4000 Hz, Event count (approx.): 7380681761, [percent: local period] foo() /home/acme/c/cpp_args_annotate Percent 0000000000401106 <foo(int, int)>: foo(int, int): int foo(int a, int b) { push %rbp mov %rsp,%rbp mov %edi,-0x14(%rbp) mov %esi,-0x18(%rbp) for (unsigned i = 0; i < 1000000000; i++) movl $0x0,-0x4(%rbp) 11: cmpl $0x3b9ac9ff,-0x4(%rbp) 86.53 ↓ ja 26 a += b; 13.32 mov -0x18(%rbp),%eax 0.00 add %eax,-0x14(%rbp) for (unsigned i = 0; i < 1000000000; i++) addl $0x1,-0x4(%rbp) 0.15 ↑ jmp 11 return a; 26: mov -0x14(%rbp),%eax } pop %rbp ← retq $ Signed-off-by: Martin Liška <mliska@suse.cz> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Jiri Slaby <jslaby@suse.cz> Link: http://lore.kernel.org/lkml/13e1a405-edf9-e4c2-4327-a9b454353730@suse.cz Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2021-02-11 15:37:55 +03:00
*
* and skip comma as part of function arguments, e.g.:
* 1d8b4ac <linemap_lookup(line_maps const*, unsigned int)+0xcc>
perf annotate: Fix parsing aarch64 branch instructions after objdump update Starting with binutils 2.28, aarch64 objdump adds comments to the disassembly output to show the alternative names of a condition code [1]. It is assumed that commas in objdump comments could occur in other arches now or in the future, so this fix is arch-independent. The fix could have been done with arm64 specific jump__parse and jump__scnprintf functions, but the jump__scnprintf instruction would have to have its comment character be a literal, since the scnprintf functions cannot receive a struct arch easily. This inconvenience also applies to the generic jump__scnprintf, which is why we add a raw_comment pointer to struct ins_operands, so the __parse function assigns it to be re-used by its corresponding __scnprintf function. Example differences in 'perf annotate --stdio2' output on an aarch64 perf.data file: BEFORE: → b.cs ffff200008133d1c <unwind_frame+0x18c> // b.hs, dffff7ecc47b AFTER : ↓ b.cs 18c BEFORE: → b.cc ffff200008d8d9cc <get_alloc_profile+0x31c> // b.lo, b.ul, dffff727295b AFTER : ↓ b.cc 31c The branch target labels 18c and 31c also now appear in the output: BEFORE: add x26, x29, #0x80 AFTER : 18c: add x26, x29, #0x80 BEFORE: add x21, x21, #0x8 AFTER : 31c: add x21, x21, #0x8 The Fixes: tag below is added so stable branches will get the update; it doesn't necessarily mean that commit was broken at the time, rather it didn't withstand the aarch64 objdump update. Tested no difference in output for sample x86_64, power arch perf.data files. [1] https://sourceware.org/git/gitweb.cgi?p=binutils-gdb.git;a=commit;h=bb7eff5206e4795ac79c177a80fe9f4630aaf730 Signed-off-by: Kim Phillips <kim.phillips@arm.com> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Anton Blanchard <anton@samba.org> Cc: Christian Borntraeger <borntraeger@de.ibm.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Mark Rutland <mark.rutland@arm.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Ravi Bangoria <ravi.bangoria@linux.vnet.ibm.com> Cc: Robin Murphy <robin.murphy@arm.com> Cc: Taeung Song <treeze.taeung@gmail.com> Cc: linux-arm-kernel@lists.infradead.org Fixes: b13bbeee5ee6 ("perf annotate: Fix branch instruction with multiple operands") Link: http://lkml.kernel.org/r/20180827125340.a2f7e291901d17cea05daba4@arm.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2018-08-27 20:53:40 +03:00
*/
static inline const char *validate_comma(const char *c, struct ins_operands *ops)
{
if (ops->raw_comment && c > ops->raw_comment)
return NULL;
perf annotate: Fix jump parsing for C++ code. Considering the following testcase: int foo(int a, int b) { for (unsigned i = 0; i < 1000000000; i++) a += b; return a; } int main() { foo (3, 4); return 0; } 'perf annotate' displays: 86.52 │40055e: → ja 40056c <foo(int, int)+0x26> 13.37 │400560: mov -0x18(%rbp),%eax │400563: add %eax,-0x14(%rbp) │400566: addl $0x1,-0x4(%rbp) 0.11 │40056a: → jmp 400557 <foo(int, int)+0x11> │40056c: mov -0x14(%rbp),%eax │40056f: pop %rbp and the 'ja 40056c' does not link to the location in the function. It's caused by fact that comma is wrongly parsed, it's part of function signature. With my patch I see: 86.52 │ ┌──ja 26 13.37 │ │ mov -0x18(%rbp),%eax │ │ add %eax,-0x14(%rbp) │ │ addl $0x1,-0x4(%rbp) 0.11 │ │↑ jmp 11 │26:└─→mov -0x14(%rbp),%eax and 'o' output prints: 86.52 │4005┌── ↓ ja 40056c <foo(int, int)+0x26> 13.37 │4005│0: mov -0x18(%rbp),%eax │4005│3: add %eax,-0x14(%rbp) │4005│6: addl $0x1,-0x4(%rbp) 0.11 │4005│a: ↑ jmp 400557 <foo(int, int)+0x11> │4005└─→ mov -0x14(%rbp),%eax On the contrary, compiling the very same file with gcc -x c, the parsing is fine because function arguments are not displayed: jmp 400543 <foo+0x1d> Committer testing: Before: $ cat cpp_args_annotate.c int foo(int a, int b) { for (unsigned i = 0; i < 1000000000; i++) a += b; return a; } int main() { foo (3, 4); return 0; } $ gcc --version |& head -1 gcc (GCC) 10.2.1 20201125 (Red Hat 10.2.1-9) $ gcc -g cpp_args_annotate.c -o cpp_args_annotate $ perf record ./cpp_args_annotate [ perf record: Woken up 2 times to write data ] [ perf record: Captured and wrote 0.275 MB perf.data (7188 samples) ] $ perf annotate --stdio2 foo Samples: 7K of event 'cycles:u', 4000 Hz, Event count (approx.): 7468429289, [percent: local period] foo() /home/acme/c/cpp_args_annotate Percent 0000000000401106 <foo>: foo(): int foo(int a, int b) { push %rbp mov %rsp,%rbp mov %edi,-0x14(%rbp) mov %esi,-0x18(%rbp) for (unsigned i = 0; i < 1000000000; i++) movl $0x0,-0x4(%rbp) ↓ jmp 1d a += b; 13.45 13: mov -0x18(%rbp),%eax add %eax,-0x14(%rbp) for (unsigned i = 0; i < 1000000000; i++) addl $0x1,-0x4(%rbp) 0.09 1d: cmpl $0x3b9ac9ff,-0x4(%rbp) 86.46 ↑ jbe 13 return a; mov -0x14(%rbp),%eax } pop %rbp ← retq $ I.e. works for C, now lets switch to C++: $ g++ -g cpp_args_annotate.c -o cpp_args_annotate $ perf record ./cpp_args_annotate [ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 0.268 MB perf.data (6976 samples) ] $ perf annotate --stdio2 foo Samples: 6K of event 'cycles:u', 4000 Hz, Event count (approx.): 7380681761, [percent: local period] foo() /home/acme/c/cpp_args_annotate Percent 0000000000401106 <foo(int, int)>: foo(int, int): int foo(int a, int b) { push %rbp mov %rsp,%rbp mov %edi,-0x14(%rbp) mov %esi,-0x18(%rbp) for (unsigned i = 0; i < 1000000000; i++) movl $0x0,-0x4(%rbp) cmpl $0x3b9ac9ff,-0x4(%rbp) 86.53 → ja 40112c <foo(int, int)+0x26> a += b; 13.32 mov -0x18(%rbp),%eax 0.00 add %eax,-0x14(%rbp) for (unsigned i = 0; i < 1000000000; i++) addl $0x1,-0x4(%rbp) 0.15 → jmp 401117 <foo(int, int)+0x11> return a; mov -0x14(%rbp),%eax } pop %rbp ← retq $ Reproduced. Now with this patch: Reusing the C++ built binary, as we can see here: $ readelf -wi cpp_args_annotate | grep producer <c> DW_AT_producer : (indirect string, offset: 0x2e): GNU C++14 10.2.1 20201125 (Red Hat 10.2.1-9) -mtune=generic -march=x86-64 -g $ And furthermore: $ file cpp_args_annotate cpp_args_annotate: ELF 64-bit LSB executable, x86-64, version 1 (SYSV), dynamically linked, interpreter /lib64/ld-linux-x86-64.so.2, BuildID[sha1]=4fe3cab260204765605ec630d0dc7a7e93c361a9, for GNU/Linux 3.2.0, with debug_info, not stripped $ perf buildid-list -i cpp_args_annotate 4fe3cab260204765605ec630d0dc7a7e93c361a9 $ perf buildid-list | grep cpp_args_annotate 4fe3cab260204765605ec630d0dc7a7e93c361a9 /home/acme/c/cpp_args_annotate $ It now works: $ perf annotate --stdio2 foo Samples: 6K of event 'cycles:u', 4000 Hz, Event count (approx.): 7380681761, [percent: local period] foo() /home/acme/c/cpp_args_annotate Percent 0000000000401106 <foo(int, int)>: foo(int, int): int foo(int a, int b) { push %rbp mov %rsp,%rbp mov %edi,-0x14(%rbp) mov %esi,-0x18(%rbp) for (unsigned i = 0; i < 1000000000; i++) movl $0x0,-0x4(%rbp) 11: cmpl $0x3b9ac9ff,-0x4(%rbp) 86.53 ↓ ja 26 a += b; 13.32 mov -0x18(%rbp),%eax 0.00 add %eax,-0x14(%rbp) for (unsigned i = 0; i < 1000000000; i++) addl $0x1,-0x4(%rbp) 0.15 ↑ jmp 11 return a; 26: mov -0x14(%rbp),%eax } pop %rbp ← retq $ Signed-off-by: Martin Liška <mliska@suse.cz> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Jiri Slaby <jslaby@suse.cz> Link: http://lore.kernel.org/lkml/13e1a405-edf9-e4c2-4327-a9b454353730@suse.cz Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2021-02-11 15:37:55 +03:00
if (ops->raw_func_start && c > ops->raw_func_start)
return NULL;
perf annotate: Fix parsing aarch64 branch instructions after objdump update Starting with binutils 2.28, aarch64 objdump adds comments to the disassembly output to show the alternative names of a condition code [1]. It is assumed that commas in objdump comments could occur in other arches now or in the future, so this fix is arch-independent. The fix could have been done with arm64 specific jump__parse and jump__scnprintf functions, but the jump__scnprintf instruction would have to have its comment character be a literal, since the scnprintf functions cannot receive a struct arch easily. This inconvenience also applies to the generic jump__scnprintf, which is why we add a raw_comment pointer to struct ins_operands, so the __parse function assigns it to be re-used by its corresponding __scnprintf function. Example differences in 'perf annotate --stdio2' output on an aarch64 perf.data file: BEFORE: → b.cs ffff200008133d1c <unwind_frame+0x18c> // b.hs, dffff7ecc47b AFTER : ↓ b.cs 18c BEFORE: → b.cc ffff200008d8d9cc <get_alloc_profile+0x31c> // b.lo, b.ul, dffff727295b AFTER : ↓ b.cc 31c The branch target labels 18c and 31c also now appear in the output: BEFORE: add x26, x29, #0x80 AFTER : 18c: add x26, x29, #0x80 BEFORE: add x21, x21, #0x8 AFTER : 31c: add x21, x21, #0x8 The Fixes: tag below is added so stable branches will get the update; it doesn't necessarily mean that commit was broken at the time, rather it didn't withstand the aarch64 objdump update. Tested no difference in output for sample x86_64, power arch perf.data files. [1] https://sourceware.org/git/gitweb.cgi?p=binutils-gdb.git;a=commit;h=bb7eff5206e4795ac79c177a80fe9f4630aaf730 Signed-off-by: Kim Phillips <kim.phillips@arm.com> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Anton Blanchard <anton@samba.org> Cc: Christian Borntraeger <borntraeger@de.ibm.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Mark Rutland <mark.rutland@arm.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Ravi Bangoria <ravi.bangoria@linux.vnet.ibm.com> Cc: Robin Murphy <robin.murphy@arm.com> Cc: Taeung Song <treeze.taeung@gmail.com> Cc: linux-arm-kernel@lists.infradead.org Fixes: b13bbeee5ee6 ("perf annotate: Fix branch instruction with multiple operands") Link: http://lkml.kernel.org/r/20180827125340.a2f7e291901d17cea05daba4@arm.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2018-08-27 20:53:40 +03:00
return c;
}
static int jump__parse(struct arch *arch, struct ins_operands *ops, struct map_symbol *ms)
{
struct map *map = ms->map;
struct symbol *sym = ms->sym;
struct addr_map_symbol target = {
.ms = { .map = map, },
};
const char *c = strchr(ops->raw, ',');
u64 start, end;
perf annotate: Fix parsing aarch64 branch instructions after objdump update Starting with binutils 2.28, aarch64 objdump adds comments to the disassembly output to show the alternative names of a condition code [1]. It is assumed that commas in objdump comments could occur in other arches now or in the future, so this fix is arch-independent. The fix could have been done with arm64 specific jump__parse and jump__scnprintf functions, but the jump__scnprintf instruction would have to have its comment character be a literal, since the scnprintf functions cannot receive a struct arch easily. This inconvenience also applies to the generic jump__scnprintf, which is why we add a raw_comment pointer to struct ins_operands, so the __parse function assigns it to be re-used by its corresponding __scnprintf function. Example differences in 'perf annotate --stdio2' output on an aarch64 perf.data file: BEFORE: → b.cs ffff200008133d1c <unwind_frame+0x18c> // b.hs, dffff7ecc47b AFTER : ↓ b.cs 18c BEFORE: → b.cc ffff200008d8d9cc <get_alloc_profile+0x31c> // b.lo, b.ul, dffff727295b AFTER : ↓ b.cc 31c The branch target labels 18c and 31c also now appear in the output: BEFORE: add x26, x29, #0x80 AFTER : 18c: add x26, x29, #0x80 BEFORE: add x21, x21, #0x8 AFTER : 31c: add x21, x21, #0x8 The Fixes: tag below is added so stable branches will get the update; it doesn't necessarily mean that commit was broken at the time, rather it didn't withstand the aarch64 objdump update. Tested no difference in output for sample x86_64, power arch perf.data files. [1] https://sourceware.org/git/gitweb.cgi?p=binutils-gdb.git;a=commit;h=bb7eff5206e4795ac79c177a80fe9f4630aaf730 Signed-off-by: Kim Phillips <kim.phillips@arm.com> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Anton Blanchard <anton@samba.org> Cc: Christian Borntraeger <borntraeger@de.ibm.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Mark Rutland <mark.rutland@arm.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Ravi Bangoria <ravi.bangoria@linux.vnet.ibm.com> Cc: Robin Murphy <robin.murphy@arm.com> Cc: Taeung Song <treeze.taeung@gmail.com> Cc: linux-arm-kernel@lists.infradead.org Fixes: b13bbeee5ee6 ("perf annotate: Fix branch instruction with multiple operands") Link: http://lkml.kernel.org/r/20180827125340.a2f7e291901d17cea05daba4@arm.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2018-08-27 20:53:40 +03:00
ops->raw_comment = strchr(ops->raw, arch->objdump.comment_char);
perf annotate: Fix jump parsing for C++ code. Considering the following testcase: int foo(int a, int b) { for (unsigned i = 0; i < 1000000000; i++) a += b; return a; } int main() { foo (3, 4); return 0; } 'perf annotate' displays: 86.52 │40055e: → ja 40056c <foo(int, int)+0x26> 13.37 │400560: mov -0x18(%rbp),%eax │400563: add %eax,-0x14(%rbp) │400566: addl $0x1,-0x4(%rbp) 0.11 │40056a: → jmp 400557 <foo(int, int)+0x11> │40056c: mov -0x14(%rbp),%eax │40056f: pop %rbp and the 'ja 40056c' does not link to the location in the function. It's caused by fact that comma is wrongly parsed, it's part of function signature. With my patch I see: 86.52 │ ┌──ja 26 13.37 │ │ mov -0x18(%rbp),%eax │ │ add %eax,-0x14(%rbp) │ │ addl $0x1,-0x4(%rbp) 0.11 │ │↑ jmp 11 │26:└─→mov -0x14(%rbp),%eax and 'o' output prints: 86.52 │4005┌── ↓ ja 40056c <foo(int, int)+0x26> 13.37 │4005│0: mov -0x18(%rbp),%eax │4005│3: add %eax,-0x14(%rbp) │4005│6: addl $0x1,-0x4(%rbp) 0.11 │4005│a: ↑ jmp 400557 <foo(int, int)+0x11> │4005└─→ mov -0x14(%rbp),%eax On the contrary, compiling the very same file with gcc -x c, the parsing is fine because function arguments are not displayed: jmp 400543 <foo+0x1d> Committer testing: Before: $ cat cpp_args_annotate.c int foo(int a, int b) { for (unsigned i = 0; i < 1000000000; i++) a += b; return a; } int main() { foo (3, 4); return 0; } $ gcc --version |& head -1 gcc (GCC) 10.2.1 20201125 (Red Hat 10.2.1-9) $ gcc -g cpp_args_annotate.c -o cpp_args_annotate $ perf record ./cpp_args_annotate [ perf record: Woken up 2 times to write data ] [ perf record: Captured and wrote 0.275 MB perf.data (7188 samples) ] $ perf annotate --stdio2 foo Samples: 7K of event 'cycles:u', 4000 Hz, Event count (approx.): 7468429289, [percent: local period] foo() /home/acme/c/cpp_args_annotate Percent 0000000000401106 <foo>: foo(): int foo(int a, int b) { push %rbp mov %rsp,%rbp mov %edi,-0x14(%rbp) mov %esi,-0x18(%rbp) for (unsigned i = 0; i < 1000000000; i++) movl $0x0,-0x4(%rbp) ↓ jmp 1d a += b; 13.45 13: mov -0x18(%rbp),%eax add %eax,-0x14(%rbp) for (unsigned i = 0; i < 1000000000; i++) addl $0x1,-0x4(%rbp) 0.09 1d: cmpl $0x3b9ac9ff,-0x4(%rbp) 86.46 ↑ jbe 13 return a; mov -0x14(%rbp),%eax } pop %rbp ← retq $ I.e. works for C, now lets switch to C++: $ g++ -g cpp_args_annotate.c -o cpp_args_annotate $ perf record ./cpp_args_annotate [ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 0.268 MB perf.data (6976 samples) ] $ perf annotate --stdio2 foo Samples: 6K of event 'cycles:u', 4000 Hz, Event count (approx.): 7380681761, [percent: local period] foo() /home/acme/c/cpp_args_annotate Percent 0000000000401106 <foo(int, int)>: foo(int, int): int foo(int a, int b) { push %rbp mov %rsp,%rbp mov %edi,-0x14(%rbp) mov %esi,-0x18(%rbp) for (unsigned i = 0; i < 1000000000; i++) movl $0x0,-0x4(%rbp) cmpl $0x3b9ac9ff,-0x4(%rbp) 86.53 → ja 40112c <foo(int, int)+0x26> a += b; 13.32 mov -0x18(%rbp),%eax 0.00 add %eax,-0x14(%rbp) for (unsigned i = 0; i < 1000000000; i++) addl $0x1,-0x4(%rbp) 0.15 → jmp 401117 <foo(int, int)+0x11> return a; mov -0x14(%rbp),%eax } pop %rbp ← retq $ Reproduced. Now with this patch: Reusing the C++ built binary, as we can see here: $ readelf -wi cpp_args_annotate | grep producer <c> DW_AT_producer : (indirect string, offset: 0x2e): GNU C++14 10.2.1 20201125 (Red Hat 10.2.1-9) -mtune=generic -march=x86-64 -g $ And furthermore: $ file cpp_args_annotate cpp_args_annotate: ELF 64-bit LSB executable, x86-64, version 1 (SYSV), dynamically linked, interpreter /lib64/ld-linux-x86-64.so.2, BuildID[sha1]=4fe3cab260204765605ec630d0dc7a7e93c361a9, for GNU/Linux 3.2.0, with debug_info, not stripped $ perf buildid-list -i cpp_args_annotate 4fe3cab260204765605ec630d0dc7a7e93c361a9 $ perf buildid-list | grep cpp_args_annotate 4fe3cab260204765605ec630d0dc7a7e93c361a9 /home/acme/c/cpp_args_annotate $ It now works: $ perf annotate --stdio2 foo Samples: 6K of event 'cycles:u', 4000 Hz, Event count (approx.): 7380681761, [percent: local period] foo() /home/acme/c/cpp_args_annotate Percent 0000000000401106 <foo(int, int)>: foo(int, int): int foo(int a, int b) { push %rbp mov %rsp,%rbp mov %edi,-0x14(%rbp) mov %esi,-0x18(%rbp) for (unsigned i = 0; i < 1000000000; i++) movl $0x0,-0x4(%rbp) 11: cmpl $0x3b9ac9ff,-0x4(%rbp) 86.53 ↓ ja 26 a += b; 13.32 mov -0x18(%rbp),%eax 0.00 add %eax,-0x14(%rbp) for (unsigned i = 0; i < 1000000000; i++) addl $0x1,-0x4(%rbp) 0.15 ↑ jmp 11 return a; 26: mov -0x14(%rbp),%eax } pop %rbp ← retq $ Signed-off-by: Martin Liška <mliska@suse.cz> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Jiri Slaby <jslaby@suse.cz> Link: http://lore.kernel.org/lkml/13e1a405-edf9-e4c2-4327-a9b454353730@suse.cz Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2021-02-11 15:37:55 +03:00
ops->raw_func_start = strchr(ops->raw, '<');
perf annotate: Fix parsing aarch64 branch instructions after objdump update Starting with binutils 2.28, aarch64 objdump adds comments to the disassembly output to show the alternative names of a condition code [1]. It is assumed that commas in objdump comments could occur in other arches now or in the future, so this fix is arch-independent. The fix could have been done with arm64 specific jump__parse and jump__scnprintf functions, but the jump__scnprintf instruction would have to have its comment character be a literal, since the scnprintf functions cannot receive a struct arch easily. This inconvenience also applies to the generic jump__scnprintf, which is why we add a raw_comment pointer to struct ins_operands, so the __parse function assigns it to be re-used by its corresponding __scnprintf function. Example differences in 'perf annotate --stdio2' output on an aarch64 perf.data file: BEFORE: → b.cs ffff200008133d1c <unwind_frame+0x18c> // b.hs, dffff7ecc47b AFTER : ↓ b.cs 18c BEFORE: → b.cc ffff200008d8d9cc <get_alloc_profile+0x31c> // b.lo, b.ul, dffff727295b AFTER : ↓ b.cc 31c The branch target labels 18c and 31c also now appear in the output: BEFORE: add x26, x29, #0x80 AFTER : 18c: add x26, x29, #0x80 BEFORE: add x21, x21, #0x8 AFTER : 31c: add x21, x21, #0x8 The Fixes: tag below is added so stable branches will get the update; it doesn't necessarily mean that commit was broken at the time, rather it didn't withstand the aarch64 objdump update. Tested no difference in output for sample x86_64, power arch perf.data files. [1] https://sourceware.org/git/gitweb.cgi?p=binutils-gdb.git;a=commit;h=bb7eff5206e4795ac79c177a80fe9f4630aaf730 Signed-off-by: Kim Phillips <kim.phillips@arm.com> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Anton Blanchard <anton@samba.org> Cc: Christian Borntraeger <borntraeger@de.ibm.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Mark Rutland <mark.rutland@arm.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Ravi Bangoria <ravi.bangoria@linux.vnet.ibm.com> Cc: Robin Murphy <robin.murphy@arm.com> Cc: Taeung Song <treeze.taeung@gmail.com> Cc: linux-arm-kernel@lists.infradead.org Fixes: b13bbeee5ee6 ("perf annotate: Fix branch instruction with multiple operands") Link: http://lkml.kernel.org/r/20180827125340.a2f7e291901d17cea05daba4@arm.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2018-08-27 20:53:40 +03:00
c = validate_comma(c, ops);
/*
* Examples of lines to parse for the _cpp_lex_token@@Base
* function:
*
* 1159e6c: jne 115aa32 <_cpp_lex_token@@Base+0xf92>
* 1159e8b: jne c469be <cpp_named_operator2name@@Base+0xa72>
*
* The first is a jump to an offset inside the same function,
* the second is to another function, i.e. that 0xa72 is an
* offset in the cpp_named_operator2name@@base function.
*/
perf annotate: Fix branch instruction with multiple operands 'perf annotate' is dropping the cr* fields from branch instructions. Fix it by adding support to display branch instructions having multiple operands. Power Arch objdump of int_sqrt: 20.36 | c0000000004d2694: subf r10,r10,r3 | c0000000004d2698: v bgt cr6,c0000000004d26a0 <int_sqrt+0x40> 1.82 | c0000000004d269c: mr r3,r10 29.18 | c0000000004d26a0: mr r10,r8 | c0000000004d26a4: v bgt cr7,c0000000004d26ac <int_sqrt+0x4c> | c0000000004d26a8: mr r10,r7 Power Arch Before Patch: 20.36 | subf r10,r10,r3 | v bgt 40 1.82 | mr r3,r10 29.18 | 40: mr r10,r8 | v bgt 4c | mr r10,r7 Power Arch After patch: 20.36 | subf r10,r10,r3 | v bgt cr6,40 1.82 | mr r3,r10 29.18 | 40: mr r10,r8 | v bgt cr7,4c | mr r10,r7 Also support AArch64 conditional branch instructions, which can have up to three operands: Aarch64 Non-simplified (raw objdump) view: │ffff0000083cd11c: ↑ cbz w0, ffff0000083cd100 <security_fil▒ ... 4.44 │ffff000│083cd134: ↓ tbnz w0, #26, ffff0000083cd190 <securit▒ ... 1.37 │ffff000│083cd144: ↓ tbnz w22, #5, ffff0000083cd1a4 <securit▒ │ffff000│083cd148: mov w19, #0x20000 //▒ 1.02 │ffff000│083cd14c: ↓ tbz w22, #2, ffff0000083cd1ac <securit▒ ... 0.68 │ffff000└──3cd16c: ↑ cbnz w0, ffff0000083cd120 <security_fil▒ Aarch64 Simplified, before this patch: │ ↑ cbz 40 ... 4.44 │ │↓ tbnz w0, #26, ffff0000083cd190 <security_file_permiss▒ ... 1.37 │ │↓ tbnz w22, #5, ffff0000083cd1a4 <security_file_permiss▒ │ │ mov w19, #0x20000 // #131072 1.02 │ │↓ tbz w22, #2, ffff0000083cd1ac <security_file_permiss▒ ... 0.68 │ └──cbnz 60 the cbz operand is missing, and the tbz doesn't get simplified processing at all because the parsing function failed to match an address. Aarch64 Simplified, After this patch applied: │ ↑ cbz w0, 40 ... 4.44 │ │↓ tbnz w0, #26, d0 ... 1.37 │ │↓ tbnz w22, #5, e4 │ │ mov w19, #0x20000 // #131072 1.02 │ │↓ tbz w22, #2, ec ... 0.68 │ └──cbnz w0, 60 Originally-by: Ravi Bangoria <ravi.bangoria@linux.vnet.ibm.com> Tested-by: Ravi Bangoria <ravi.bangoria@linux.vnet.ibm.com> Reported-by: Anton Blanchard <anton@samba.org> Reported-by: Robin Murphy <robin.murphy@arm.com> Signed-off-by: Kim Phillips <kim.phillips@arm.com> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Christian Borntraeger <borntraeger@de.ibm.com> Cc: Mark Rutland <mark.rutland@arm.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Taeung Song <treeze.taeung@gmail.com> Link: http://lkml.kernel.org/r/20170601092959.f60d98912e8a1b66fd1e4c0e@arm.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2017-06-01 17:29:59 +03:00
/*
* skip over possible up to 2 operands to get to address, e.g.:
* tbnz w0, #26, ffff0000083cd190 <security_file_permission+0xd0>
*/
if (c++ != NULL) {
ops->target.addr = strtoull(c, NULL, 16);
perf annotate: Fix branch instruction with multiple operands 'perf annotate' is dropping the cr* fields from branch instructions. Fix it by adding support to display branch instructions having multiple operands. Power Arch objdump of int_sqrt: 20.36 | c0000000004d2694: subf r10,r10,r3 | c0000000004d2698: v bgt cr6,c0000000004d26a0 <int_sqrt+0x40> 1.82 | c0000000004d269c: mr r3,r10 29.18 | c0000000004d26a0: mr r10,r8 | c0000000004d26a4: v bgt cr7,c0000000004d26ac <int_sqrt+0x4c> | c0000000004d26a8: mr r10,r7 Power Arch Before Patch: 20.36 | subf r10,r10,r3 | v bgt 40 1.82 | mr r3,r10 29.18 | 40: mr r10,r8 | v bgt 4c | mr r10,r7 Power Arch After patch: 20.36 | subf r10,r10,r3 | v bgt cr6,40 1.82 | mr r3,r10 29.18 | 40: mr r10,r8 | v bgt cr7,4c | mr r10,r7 Also support AArch64 conditional branch instructions, which can have up to three operands: Aarch64 Non-simplified (raw objdump) view: │ffff0000083cd11c: ↑ cbz w0, ffff0000083cd100 <security_fil▒ ... 4.44 │ffff000│083cd134: ↓ tbnz w0, #26, ffff0000083cd190 <securit▒ ... 1.37 │ffff000│083cd144: ↓ tbnz w22, #5, ffff0000083cd1a4 <securit▒ │ffff000│083cd148: mov w19, #0x20000 //▒ 1.02 │ffff000│083cd14c: ↓ tbz w22, #2, ffff0000083cd1ac <securit▒ ... 0.68 │ffff000└──3cd16c: ↑ cbnz w0, ffff0000083cd120 <security_fil▒ Aarch64 Simplified, before this patch: │ ↑ cbz 40 ... 4.44 │ │↓ tbnz w0, #26, ffff0000083cd190 <security_file_permiss▒ ... 1.37 │ │↓ tbnz w22, #5, ffff0000083cd1a4 <security_file_permiss▒ │ │ mov w19, #0x20000 // #131072 1.02 │ │↓ tbz w22, #2, ffff0000083cd1ac <security_file_permiss▒ ... 0.68 │ └──cbnz 60 the cbz operand is missing, and the tbz doesn't get simplified processing at all because the parsing function failed to match an address. Aarch64 Simplified, After this patch applied: │ ↑ cbz w0, 40 ... 4.44 │ │↓ tbnz w0, #26, d0 ... 1.37 │ │↓ tbnz w22, #5, e4 │ │ mov w19, #0x20000 // #131072 1.02 │ │↓ tbz w22, #2, ec ... 0.68 │ └──cbnz w0, 60 Originally-by: Ravi Bangoria <ravi.bangoria@linux.vnet.ibm.com> Tested-by: Ravi Bangoria <ravi.bangoria@linux.vnet.ibm.com> Reported-by: Anton Blanchard <anton@samba.org> Reported-by: Robin Murphy <robin.murphy@arm.com> Signed-off-by: Kim Phillips <kim.phillips@arm.com> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Christian Borntraeger <borntraeger@de.ibm.com> Cc: Mark Rutland <mark.rutland@arm.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Taeung Song <treeze.taeung@gmail.com> Link: http://lkml.kernel.org/r/20170601092959.f60d98912e8a1b66fd1e4c0e@arm.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2017-06-01 17:29:59 +03:00
if (!ops->target.addr) {
c = strchr(c, ',');
perf annotate: Fix parsing aarch64 branch instructions after objdump update Starting with binutils 2.28, aarch64 objdump adds comments to the disassembly output to show the alternative names of a condition code [1]. It is assumed that commas in objdump comments could occur in other arches now or in the future, so this fix is arch-independent. The fix could have been done with arm64 specific jump__parse and jump__scnprintf functions, but the jump__scnprintf instruction would have to have its comment character be a literal, since the scnprintf functions cannot receive a struct arch easily. This inconvenience also applies to the generic jump__scnprintf, which is why we add a raw_comment pointer to struct ins_operands, so the __parse function assigns it to be re-used by its corresponding __scnprintf function. Example differences in 'perf annotate --stdio2' output on an aarch64 perf.data file: BEFORE: → b.cs ffff200008133d1c <unwind_frame+0x18c> // b.hs, dffff7ecc47b AFTER : ↓ b.cs 18c BEFORE: → b.cc ffff200008d8d9cc <get_alloc_profile+0x31c> // b.lo, b.ul, dffff727295b AFTER : ↓ b.cc 31c The branch target labels 18c and 31c also now appear in the output: BEFORE: add x26, x29, #0x80 AFTER : 18c: add x26, x29, #0x80 BEFORE: add x21, x21, #0x8 AFTER : 31c: add x21, x21, #0x8 The Fixes: tag below is added so stable branches will get the update; it doesn't necessarily mean that commit was broken at the time, rather it didn't withstand the aarch64 objdump update. Tested no difference in output for sample x86_64, power arch perf.data files. [1] https://sourceware.org/git/gitweb.cgi?p=binutils-gdb.git;a=commit;h=bb7eff5206e4795ac79c177a80fe9f4630aaf730 Signed-off-by: Kim Phillips <kim.phillips@arm.com> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Anton Blanchard <anton@samba.org> Cc: Christian Borntraeger <borntraeger@de.ibm.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Mark Rutland <mark.rutland@arm.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Ravi Bangoria <ravi.bangoria@linux.vnet.ibm.com> Cc: Robin Murphy <robin.murphy@arm.com> Cc: Taeung Song <treeze.taeung@gmail.com> Cc: linux-arm-kernel@lists.infradead.org Fixes: b13bbeee5ee6 ("perf annotate: Fix branch instruction with multiple operands") Link: http://lkml.kernel.org/r/20180827125340.a2f7e291901d17cea05daba4@arm.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2018-08-27 20:53:40 +03:00
c = validate_comma(c, ops);
perf annotate: Fix branch instruction with multiple operands 'perf annotate' is dropping the cr* fields from branch instructions. Fix it by adding support to display branch instructions having multiple operands. Power Arch objdump of int_sqrt: 20.36 | c0000000004d2694: subf r10,r10,r3 | c0000000004d2698: v bgt cr6,c0000000004d26a0 <int_sqrt+0x40> 1.82 | c0000000004d269c: mr r3,r10 29.18 | c0000000004d26a0: mr r10,r8 | c0000000004d26a4: v bgt cr7,c0000000004d26ac <int_sqrt+0x4c> | c0000000004d26a8: mr r10,r7 Power Arch Before Patch: 20.36 | subf r10,r10,r3 | v bgt 40 1.82 | mr r3,r10 29.18 | 40: mr r10,r8 | v bgt 4c | mr r10,r7 Power Arch After patch: 20.36 | subf r10,r10,r3 | v bgt cr6,40 1.82 | mr r3,r10 29.18 | 40: mr r10,r8 | v bgt cr7,4c | mr r10,r7 Also support AArch64 conditional branch instructions, which can have up to three operands: Aarch64 Non-simplified (raw objdump) view: │ffff0000083cd11c: ↑ cbz w0, ffff0000083cd100 <security_fil▒ ... 4.44 │ffff000│083cd134: ↓ tbnz w0, #26, ffff0000083cd190 <securit▒ ... 1.37 │ffff000│083cd144: ↓ tbnz w22, #5, ffff0000083cd1a4 <securit▒ │ffff000│083cd148: mov w19, #0x20000 //▒ 1.02 │ffff000│083cd14c: ↓ tbz w22, #2, ffff0000083cd1ac <securit▒ ... 0.68 │ffff000└──3cd16c: ↑ cbnz w0, ffff0000083cd120 <security_fil▒ Aarch64 Simplified, before this patch: │ ↑ cbz 40 ... 4.44 │ │↓ tbnz w0, #26, ffff0000083cd190 <security_file_permiss▒ ... 1.37 │ │↓ tbnz w22, #5, ffff0000083cd1a4 <security_file_permiss▒ │ │ mov w19, #0x20000 // #131072 1.02 │ │↓ tbz w22, #2, ffff0000083cd1ac <security_file_permiss▒ ... 0.68 │ └──cbnz 60 the cbz operand is missing, and the tbz doesn't get simplified processing at all because the parsing function failed to match an address. Aarch64 Simplified, After this patch applied: │ ↑ cbz w0, 40 ... 4.44 │ │↓ tbnz w0, #26, d0 ... 1.37 │ │↓ tbnz w22, #5, e4 │ │ mov w19, #0x20000 // #131072 1.02 │ │↓ tbz w22, #2, ec ... 0.68 │ └──cbnz w0, 60 Originally-by: Ravi Bangoria <ravi.bangoria@linux.vnet.ibm.com> Tested-by: Ravi Bangoria <ravi.bangoria@linux.vnet.ibm.com> Reported-by: Anton Blanchard <anton@samba.org> Reported-by: Robin Murphy <robin.murphy@arm.com> Signed-off-by: Kim Phillips <kim.phillips@arm.com> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Christian Borntraeger <borntraeger@de.ibm.com> Cc: Mark Rutland <mark.rutland@arm.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Taeung Song <treeze.taeung@gmail.com> Link: http://lkml.kernel.org/r/20170601092959.f60d98912e8a1b66fd1e4c0e@arm.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2017-06-01 17:29:59 +03:00
if (c++ != NULL)
ops->target.addr = strtoull(c, NULL, 16);
}
} else {
ops->target.addr = strtoull(ops->raw, NULL, 16);
perf annotate: Fix branch instruction with multiple operands 'perf annotate' is dropping the cr* fields from branch instructions. Fix it by adding support to display branch instructions having multiple operands. Power Arch objdump of int_sqrt: 20.36 | c0000000004d2694: subf r10,r10,r3 | c0000000004d2698: v bgt cr6,c0000000004d26a0 <int_sqrt+0x40> 1.82 | c0000000004d269c: mr r3,r10 29.18 | c0000000004d26a0: mr r10,r8 | c0000000004d26a4: v bgt cr7,c0000000004d26ac <int_sqrt+0x4c> | c0000000004d26a8: mr r10,r7 Power Arch Before Patch: 20.36 | subf r10,r10,r3 | v bgt 40 1.82 | mr r3,r10 29.18 | 40: mr r10,r8 | v bgt 4c | mr r10,r7 Power Arch After patch: 20.36 | subf r10,r10,r3 | v bgt cr6,40 1.82 | mr r3,r10 29.18 | 40: mr r10,r8 | v bgt cr7,4c | mr r10,r7 Also support AArch64 conditional branch instructions, which can have up to three operands: Aarch64 Non-simplified (raw objdump) view: │ffff0000083cd11c: ↑ cbz w0, ffff0000083cd100 <security_fil▒ ... 4.44 │ffff000│083cd134: ↓ tbnz w0, #26, ffff0000083cd190 <securit▒ ... 1.37 │ffff000│083cd144: ↓ tbnz w22, #5, ffff0000083cd1a4 <securit▒ │ffff000│083cd148: mov w19, #0x20000 //▒ 1.02 │ffff000│083cd14c: ↓ tbz w22, #2, ffff0000083cd1ac <securit▒ ... 0.68 │ffff000└──3cd16c: ↑ cbnz w0, ffff0000083cd120 <security_fil▒ Aarch64 Simplified, before this patch: │ ↑ cbz 40 ... 4.44 │ │↓ tbnz w0, #26, ffff0000083cd190 <security_file_permiss▒ ... 1.37 │ │↓ tbnz w22, #5, ffff0000083cd1a4 <security_file_permiss▒ │ │ mov w19, #0x20000 // #131072 1.02 │ │↓ tbz w22, #2, ffff0000083cd1ac <security_file_permiss▒ ... 0.68 │ └──cbnz 60 the cbz operand is missing, and the tbz doesn't get simplified processing at all because the parsing function failed to match an address. Aarch64 Simplified, After this patch applied: │ ↑ cbz w0, 40 ... 4.44 │ │↓ tbnz w0, #26, d0 ... 1.37 │ │↓ tbnz w22, #5, e4 │ │ mov w19, #0x20000 // #131072 1.02 │ │↓ tbz w22, #2, ec ... 0.68 │ └──cbnz w0, 60 Originally-by: Ravi Bangoria <ravi.bangoria@linux.vnet.ibm.com> Tested-by: Ravi Bangoria <ravi.bangoria@linux.vnet.ibm.com> Reported-by: Anton Blanchard <anton@samba.org> Reported-by: Robin Murphy <robin.murphy@arm.com> Signed-off-by: Kim Phillips <kim.phillips@arm.com> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Christian Borntraeger <borntraeger@de.ibm.com> Cc: Mark Rutland <mark.rutland@arm.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Taeung Song <treeze.taeung@gmail.com> Link: http://lkml.kernel.org/r/20170601092959.f60d98912e8a1b66fd1e4c0e@arm.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2017-06-01 17:29:59 +03:00
}
target.addr = map__objdump_2mem(map, ops->target.addr);
start = map->unmap_ip(map, sym->start),
end = map->unmap_ip(map, sym->end);
ops->target.outside = target.addr < start || target.addr > end;
/*
* FIXME: things like this in _cpp_lex_token (gcc's cc1 program):
cpp_named_operator2name@@Base+0xa72
* Point to a place that is after the cpp_named_operator2name
* boundaries, i.e. in the ELF symbol table for cc1
* cpp_named_operator2name is marked as being 32-bytes long, but it in
* fact is much larger than that, so we seem to need a symbols__find()
* routine that looks for >= current->start and < next_symbol->start,
* possibly just for C++ objects?
*
* For now lets just make some progress by marking jumps to outside the
* current function as call like.
*
* Actual navigation will come next, with further understanding of how
* the symbol searching and disassembly should be done.
perf annotate: Support jumping from one function to another For instance: entry_SYSCALL_64 /lib/modules/4.16.0-rc5-00086-gdf09348f78dc/build/vmlinux 5.50 │ → callq do_syscall_64 14.56 │ mov 0x58(%rsp),%rcx 7.44 │ mov 0x80(%rsp),%r11 0.32 │ cmp %rcx,%r11 │ → jne swapgs_restore_regs_and_return_to_usermode 0.32 │ shl $0x10,%rcx 0.32 │ sar $0x10,%rcx 3.24 │ cmp %rcx,%r11 │ → jne swapgs_restore_regs_and_return_to_usermode 2.27 │ cmpq $0x33,0x88(%rsp) 1.29 │ → jne swapgs_restore_regs_and_return_to_usermode │ mov 0x30(%rsp),%r11 8.74 │ cmp %r11,0x90(%rsp) │ → jne swapgs_restore_regs_and_return_to_usermode 0.32 │ test $0x10100,%r11 │ → jne swapgs_restore_regs_and_return_to_usermode 0.32 │ cmpq $0x2b,0xa0(%rsp) 0.65 │ → jne swapgs_restore_regs_and_return_to_usermode It'll behave just like a "call" instruction, i.e. press enter or right arrow over one such line and the browser will navigate to the annotated disassembly of that function, which when exited, via left arrow or esc, will come back to the calling function. Now to support jump to an offset on a different function... Reported-by: Linus Torvalds <torvalds@linux-foundation.org> Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Andi Kleen <ak@linux.intel.com> Cc: David Ahern <dsahern@gmail.com> Cc: Jin Yao <yao.jin@linux.intel.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Wang Nan <wangnan0@huawei.com> Link: https://lkml.kernel.org/n/tip-78o508mqvr8inhj63ddtw7mo@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2018-03-23 16:50:35 +03:00
*/
if (maps__find_ams(ms->maps, &target) == 0 &&
map__rip_2objdump(target.ms.map, map->map_ip(target.ms.map, target.addr)) == ops->target.addr)
ops->target.sym = target.ms.sym;
perf annotate: Use absolute addresses to calculate jump target offsets These types of jumps were confusing the annotate browser: entry_SYSCALL_64 /lib/modules/4.16.0-rc5-00086-gdf09348f78dc/build/vmlinux entry_SYSCALL_64 /lib/modules/4.16.0-rc5-00086-gdf09348f78dc/build/vmlinux Percent│ffffffff81a00020: swapgs <SNIP> │ffffffff81a00128: ↓ jae ffffffff81a00139 <syscall_return_via_sysret+0x53> <SNIP> │ffffffff81a00155: → jmpq *0x825d2d(%rip) # ffffffff82225e88 <pv_cpu_ops+0xe8> I.e. the syscall_return_via_sysret function is actually "inside" the entry_SYSCALL_64 function, and the offsets in jumps like these (+0x53) are relative to syscall_return_via_sysret, not to syscall_return_via_sysret. Or this may be some artifact in how the assembler marks the start and end of a function and how this ends up in the ELF symtab for vmlinux, i.e. syscall_return_via_sysret() isn't "inside" entry_SYSCALL_64, but just right after it. From readelf -sw vmlinux: 80267: ffffffff81a00020 315 NOTYPE GLOBAL DEFAULT 1 entry_SYSCALL_64 316: ffffffff81a000e6 0 NOTYPE LOCAL DEFAULT 1 syscall_return_via_sysret 0xffffffff81a00020 + 315 > 0xffffffff81a000e6 So instead of looking for offsets after that last '+' sign, calculate offsets for jump target addresses that are inside the function being disassembled from the absolute address, 0xffffffff81a00139 in this case, subtracting from it the objdump address for the start of the function being disassembled, entry_SYSCALL_64() in this case. So, before this patch: entry_SYSCALL_64 /lib/modules/4.16.0-rc5-00086-gdf09348f78dc/build/vmlinux Percent│ pop %r10 │ pop %r9 │ pop %r8 │ pop %rax │ pop %rsi │ pop %rdx │ pop %rsi │ mov %rsp,%rdi │ mov %gs:0x5004,%rsp │ pushq 0x28(%rdi) │ pushq (%rdi) │ push %rax │ ↑ jmp 6c │ mov %cr3,%rdi │ ↑ jmp 62 │ mov %rdi,%rax │ and $0x7ff,%rdi │ bt %rdi,%gs:0x2219a │ ↑ jae 53 │ btr %rdi,%gs:0x2219a │ mov %rax,%rdi │ ↑ jmp 5b After: entry_SYSCALL_64 /lib/modules/4.16.0-rc5-00086-gdf09348f78dc/build/vmlinux 0.65 │ → jne swapgs_restore_regs_and_return_to_usermode │ pop %r10 │ pop %r9 │ pop %r8 │ pop %rax │ pop %rsi │ pop %rdx │ pop %rsi │ mov %rsp,%rdi │ mov %gs:0x5004,%rsp │ pushq 0x28(%rdi) │ pushq (%rdi) │ push %rax │ ↓ jmp 132 │ mov %cr3,%rdi │ ┌──jmp 128 │ │ mov %rdi,%rax │ │ and $0x7ff,%rdi │ │ bt %rdi,%gs:0x2219a │ │↓ jae 119 │ │ btr %rdi,%gs:0x2219a │ │ mov %rax,%rdi │ │↓ jmp 121 │119:│ mov %rax,%rdi │ │ bts $0x3f,%rdi │121:│ or $0x800,%rdi │128:└─→or $0x1000,%rdi │ mov %rdi,%cr3 │132: pop %rax │ pop %rdi │ pop %rsp │ → jmpq *0x825d2d(%rip) # ffffffff82225e88 <pv_cpu_ops+0xe8> With those at least navigating to the right destination, an improvement for these cases seems to be to be to somehow mark those inner functions, which in this case could be: entry_SYSCALL_64 /lib/modules/4.16.0-rc5-00086-gdf09348f78dc/build/vmlinux │syscall_return_via_sysret: │ pop %r15 │ pop %r14 │ pop %r13 │ pop %r12 │ pop %rbp │ pop %rbx │ pop %rsi │ pop %r10 │ pop %r9 │ pop %r8 │ pop %rax │ pop %rsi │ pop %rdx │ pop %rsi │ mov %rsp,%rdi │ mov %gs:0x5004,%rsp │ pushq 0x28(%rdi) │ pushq (%rdi) │ push %rax │ ↓ jmp 132 │ mov %cr3,%rdi │ ┌──jmp 128 │ │ mov %rdi,%rax │ │ and $0x7ff,%rdi │ │ bt %rdi,%gs:0x2219a │ │↓ jae 119 │ │ btr %rdi,%gs:0x2219a │ │ mov %rax,%rdi │ │↓ jmp 121 │119:│ mov %rax,%rdi │ │ bts $0x3f,%rdi │121:│ or $0x800,%rdi │128:└─→or $0x1000,%rdi │ mov %rdi,%cr3 │132: pop %rax │ pop %rdi │ pop %rsp │ → jmpq *0x825d2d(%rip) # ffffffff82225e88 <pv_cpu_ops+0xe8> This all gets much better viewed if one uses 'perf report --ignore-vmlinux' forcing the usage of /proc/kcore + /proc/kallsyms, when the above actually gets down to: # perf report --ignore-vmlinux ## do '/64', will show the function names containing '64', ## navigate to /entry_SYSCALL_64_after_hwframe.annotation, ## press 'A' to annotate, then 'P' to print that annotation ## to a file ## From another xterm (or see on screen, this 'P' thing is for ## getting rid of those right side scroll bars/spaces): # cat /entry_SYSCALL_64_after_hwframe.annotation entry_SYSCALL_64_after_hwframe() /proc/kcore Event: cycles:ppp Percent Disassembly of section load0: ffffffff9aa00044 <load0>: 11.97 push %rax 4.85 push %rdi push %rsi 2.59 push %rdx 2.27 push %rcx 0.32 pushq $0xffffffffffffffda 1.29 push %r8 xor %r8d,%r8d 1.62 push %r9 0.65 xor %r9d,%r9d 1.62 push %r10 xor %r10d,%r10d 5.50 push %r11 xor %r11d,%r11d 3.56 push %rbx xor %ebx,%ebx 4.21 push %rbp xor %ebp,%ebp 2.59 push %r12 0.97 xor %r12d,%r12d 3.24 push %r13 xor %r13d,%r13d 2.27 push %r14 xor %r14d,%r14d 4.21 push %r15 xor %r15d,%r15d 0.97 mov %rsp,%rdi 5.50 → callq do_syscall_64 14.56 mov 0x58(%rsp),%rcx 7.44 mov 0x80(%rsp),%r11 0.32 cmp %rcx,%r11 → jne swapgs_restore_regs_and_return_to_usermode 0.32 shl $0x10,%rcx 0.32 sar $0x10,%rcx 3.24 cmp %rcx,%r11 → jne swapgs_restore_regs_and_return_to_usermode 2.27 cmpq $0x33,0x88(%rsp) 1.29 → jne swapgs_restore_regs_and_return_to_usermode mov 0x30(%rsp),%r11 8.74 cmp %r11,0x90(%rsp) → jne swapgs_restore_regs_and_return_to_usermode 0.32 test $0x10100,%r11 → jne swapgs_restore_regs_and_return_to_usermode 0.32 cmpq $0x2b,0xa0(%rsp) 0.65 → jne swapgs_restore_regs_and_return_to_usermode I.e. using kallsyms makes the function start/end be done differently than using what is in the vmlinux ELF symtab and actually the hits goes to entry_SYSCALL_64_after_hwframe, which is a GLOBAL() after the start of entry_SYSCALL_64: ENTRY(entry_SYSCALL_64) UNWIND_HINT_EMPTY <SNIP> pushq $__USER_CS /* pt_regs->cs */ pushq %rcx /* pt_regs->ip */ GLOBAL(entry_SYSCALL_64_after_hwframe) pushq %rax /* pt_regs->orig_ax */ PUSH_AND_CLEAR_REGS rax=$-ENOSYS And it goes and ends at: cmpq $__USER_DS, SS(%rsp) /* SS must match SYSRET */ jne swapgs_restore_regs_and_return_to_usermode /* * We win! This label is here just for ease of understanding * perf profiles. Nothing jumps here. */ syscall_return_via_sysret: /* rcx and r11 are already restored (see code above) */ UNWIND_HINT_EMPTY POP_REGS pop_rdi=0 skip_r11rcx=1 So perhaps some people should really just play with '--ignore-vmlinux' to force /proc/kcore + kallsyms. One idea is to do both, i.e. have a vmlinux annotation and a kcore+kallsyms one, when possible, and even show the patched location, etc. Reported-by: Linus Torvalds <torvalds@linux-foundation.org> Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Andi Kleen <ak@linux.intel.com> Cc: David Ahern <dsahern@gmail.com> Cc: Jin Yao <yao.jin@linux.intel.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Wang Nan <wangnan0@huawei.com> Link: https://lkml.kernel.org/n/tip-r11knxv8voesav31xokjiuo6@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2018-03-23 18:26:39 +03:00
if (!ops->target.outside) {
ops->target.offset = target.addr - start;
perf annotate: Fix jump target outside of function address range If jump target is outside of function range, perf is not handling it correctly. Especially when target address is lesser than function start address, target offset will be negative. But, target address declared to be unsigned, converts negative number into 2's complement. See below example. Here target of 'jumpq' instruction at 34cf8 is 34ac0 which is lesser than function start address(34cf0). 34ac0 - 34cf0 = -0x230 = 0xfffffffffffffdd0 Objdump output: 0000000000034cf0 <__sigaction>: __GI___sigaction(): 34cf0: lea -0x20(%rdi),%eax 34cf3: cmp -bashx1,%eax 34cf6: jbe 34d00 <__sigaction+0x10> 34cf8: jmpq 34ac0 <__GI___libc_sigaction> 34cfd: nopl (%rax) 34d00: mov 0x386161(%rip),%rax # 3bae68 <_DYNAMIC+0x2e8> 34d07: movl -bashx16,%fs:(%rax) 34d0e: mov -bashxffffffff,%eax 34d13: retq perf annotate before applying patch: __GI___sigaction /usr/lib64/libc-2.22.so lea -0x20(%rdi),%eax cmp -bashx1,%eax v jbe 10 v jmpq fffffffffffffdd0 nop 10: mov _DYNAMIC+0x2e8,%rax movl -bashx16,%fs:(%rax) mov -bashxffffffff,%eax retq perf annotate after applying patch: __GI___sigaction /usr/lib64/libc-2.22.so lea -0x20(%rdi),%eax cmp -bashx1,%eax v jbe 10 ^ jmpq 34ac0 <__GI___libc_sigaction> nop 10: mov _DYNAMIC+0x2e8,%rax movl -bashx16,%fs:(%rax) mov -bashxffffffff,%eax retq Signed-off-by: Ravi Bangoria <ravi.bangoria@linux.vnet.ibm.com> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Chris Riyder <chris.ryder@arm.com> Cc: Kim Phillips <kim.phillips@arm.com> Cc: Markus Trippelsdorf <markus@trippelsdorf.de> Cc: Masami Hiramatsu <mhiramat@kernel.org> Cc: Naveen N. Rao <naveen.n.rao@linux.vnet.ibm.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Taeung Song <treeze.taeung@gmail.com> Cc: linuxppc-dev@lists.ozlabs.org Link: http://lkml.kernel.org/r/1480953407-7605-3-git-send-email-ravi.bangoria@linux.vnet.ibm.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-12-05 18:56:47 +03:00
ops->target.offset_avail = true;
} else {
ops->target.offset_avail = false;
}
return 0;
}
static int jump__scnprintf(struct ins *ins, char *bf, size_t size,
perf annotate: Calculate the max instruction name, align column to that We were hardcoding '6' as the max instruction name, and we have lots that are longer than that, see the diff from two 'P' printed TUI annotations for a libc function that uses instructions with long names, such as 'vpmovmskb' with its 9 chars: --- __strcmp_avx2.annotation.before 2019-03-06 16:31:39.368020425 -0300 +++ __strcmp_avx2.annotation 2019-03-06 16:32:12.079450508 -0300 @@ -2,284 +2,284 @@ Event: cycles:ppp Percent endbr64 - 0.10 mov %edi,%eax + 0.10 mov %edi,%eax - xor %edx,%edx + xor %edx,%edx - 3.54 vpxor %ymm7,%ymm7,%ymm7 + 3.54 vpxor %ymm7,%ymm7,%ymm7 - or %esi,%eax + or %esi,%eax - and $0xfff,%eax + and $0xfff,%eax - cmp $0xf80,%eax + cmp $0xf80,%eax - ↓ jg 370 + ↓ jg 370 - 27.07 vmovdqu (%rdi),%ymm1 + 27.07 vmovdqu (%rdi),%ymm1 - 7.97 vpcmpeqb (%rsi),%ymm1,%ymm0 + 7.97 vpcmpeqb (%rsi),%ymm1,%ymm0 - 2.15 vpminub %ymm1,%ymm0,%ymm0 + 2.15 vpminub %ymm1,%ymm0,%ymm0 - 4.09 vpcmpeqb %ymm7,%ymm0,%ymm0 + 4.09 vpcmpeqb %ymm7,%ymm0,%ymm0 - 0.43 vpmovmskb %ymm0,%ecx + 0.43 vpmovmskb %ymm0,%ecx - 1.53 test %ecx,%ecx + 1.53 test %ecx,%ecx - ↓ je b0 + ↓ je b0 - 5.26 tzcnt %ecx,%edx + 5.26 tzcnt %ecx,%edx - 18.40 movzbl (%rdi,%rdx,1),%eax + 18.40 movzbl (%rdi,%rdx,1),%eax - 7.09 movzbl (%rsi,%rdx,1),%edx + 7.09 movzbl (%rsi,%rdx,1),%edx - 3.34 sub %edx,%eax + 3.34 sub %edx,%eax 2.37 vzeroupper ← retq nop - 50: tzcnt %ecx,%edx + 50: tzcnt %ecx,%edx - movzbl 0x20(%rdi,%rdx,1),%eax + movzbl 0x20(%rdi,%rdx,1),%eax - movzbl 0x20(%rsi,%rdx,1),%edx + movzbl 0x20(%rsi,%rdx,1),%edx - sub %edx,%eax + sub %edx,%eax vzeroupper ← retq - data16 nopw %cs:0x0(%rax,%rax,1) + data16 nopw %cs:0x0(%rax,%rax,1) Reported-by: Travis Downs <travis.downs@gmail.com> LPU-Reference: CAOBGo4z1KfmWeOm6Et0cnX5Z6DWsG2PQbAvRn1MhVPJmXHrc5g@mail.gmail.com Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Link: https://lkml.kernel.org/n/tip-89wsdd9h9g6bvq52sgp6d0u4@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2019-03-06 22:40:15 +03:00
struct ins_operands *ops, int max_ins_name)
{
const char *c;
perf annotate: Fix branch instruction with multiple operands 'perf annotate' is dropping the cr* fields from branch instructions. Fix it by adding support to display branch instructions having multiple operands. Power Arch objdump of int_sqrt: 20.36 | c0000000004d2694: subf r10,r10,r3 | c0000000004d2698: v bgt cr6,c0000000004d26a0 <int_sqrt+0x40> 1.82 | c0000000004d269c: mr r3,r10 29.18 | c0000000004d26a0: mr r10,r8 | c0000000004d26a4: v bgt cr7,c0000000004d26ac <int_sqrt+0x4c> | c0000000004d26a8: mr r10,r7 Power Arch Before Patch: 20.36 | subf r10,r10,r3 | v bgt 40 1.82 | mr r3,r10 29.18 | 40: mr r10,r8 | v bgt 4c | mr r10,r7 Power Arch After patch: 20.36 | subf r10,r10,r3 | v bgt cr6,40 1.82 | mr r3,r10 29.18 | 40: mr r10,r8 | v bgt cr7,4c | mr r10,r7 Also support AArch64 conditional branch instructions, which can have up to three operands: Aarch64 Non-simplified (raw objdump) view: │ffff0000083cd11c: ↑ cbz w0, ffff0000083cd100 <security_fil▒ ... 4.44 │ffff000│083cd134: ↓ tbnz w0, #26, ffff0000083cd190 <securit▒ ... 1.37 │ffff000│083cd144: ↓ tbnz w22, #5, ffff0000083cd1a4 <securit▒ │ffff000│083cd148: mov w19, #0x20000 //▒ 1.02 │ffff000│083cd14c: ↓ tbz w22, #2, ffff0000083cd1ac <securit▒ ... 0.68 │ffff000└──3cd16c: ↑ cbnz w0, ffff0000083cd120 <security_fil▒ Aarch64 Simplified, before this patch: │ ↑ cbz 40 ... 4.44 │ │↓ tbnz w0, #26, ffff0000083cd190 <security_file_permiss▒ ... 1.37 │ │↓ tbnz w22, #5, ffff0000083cd1a4 <security_file_permiss▒ │ │ mov w19, #0x20000 // #131072 1.02 │ │↓ tbz w22, #2, ffff0000083cd1ac <security_file_permiss▒ ... 0.68 │ └──cbnz 60 the cbz operand is missing, and the tbz doesn't get simplified processing at all because the parsing function failed to match an address. Aarch64 Simplified, After this patch applied: │ ↑ cbz w0, 40 ... 4.44 │ │↓ tbnz w0, #26, d0 ... 1.37 │ │↓ tbnz w22, #5, e4 │ │ mov w19, #0x20000 // #131072 1.02 │ │↓ tbz w22, #2, ec ... 0.68 │ └──cbnz w0, 60 Originally-by: Ravi Bangoria <ravi.bangoria@linux.vnet.ibm.com> Tested-by: Ravi Bangoria <ravi.bangoria@linux.vnet.ibm.com> Reported-by: Anton Blanchard <anton@samba.org> Reported-by: Robin Murphy <robin.murphy@arm.com> Signed-off-by: Kim Phillips <kim.phillips@arm.com> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Christian Borntraeger <borntraeger@de.ibm.com> Cc: Mark Rutland <mark.rutland@arm.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Taeung Song <treeze.taeung@gmail.com> Link: http://lkml.kernel.org/r/20170601092959.f60d98912e8a1b66fd1e4c0e@arm.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2017-06-01 17:29:59 +03:00
perf annotate: Fix jump target outside of function address range If jump target is outside of function range, perf is not handling it correctly. Especially when target address is lesser than function start address, target offset will be negative. But, target address declared to be unsigned, converts negative number into 2's complement. See below example. Here target of 'jumpq' instruction at 34cf8 is 34ac0 which is lesser than function start address(34cf0). 34ac0 - 34cf0 = -0x230 = 0xfffffffffffffdd0 Objdump output: 0000000000034cf0 <__sigaction>: __GI___sigaction(): 34cf0: lea -0x20(%rdi),%eax 34cf3: cmp -bashx1,%eax 34cf6: jbe 34d00 <__sigaction+0x10> 34cf8: jmpq 34ac0 <__GI___libc_sigaction> 34cfd: nopl (%rax) 34d00: mov 0x386161(%rip),%rax # 3bae68 <_DYNAMIC+0x2e8> 34d07: movl -bashx16,%fs:(%rax) 34d0e: mov -bashxffffffff,%eax 34d13: retq perf annotate before applying patch: __GI___sigaction /usr/lib64/libc-2.22.so lea -0x20(%rdi),%eax cmp -bashx1,%eax v jbe 10 v jmpq fffffffffffffdd0 nop 10: mov _DYNAMIC+0x2e8,%rax movl -bashx16,%fs:(%rax) mov -bashxffffffff,%eax retq perf annotate after applying patch: __GI___sigaction /usr/lib64/libc-2.22.so lea -0x20(%rdi),%eax cmp -bashx1,%eax v jbe 10 ^ jmpq 34ac0 <__GI___libc_sigaction> nop 10: mov _DYNAMIC+0x2e8,%rax movl -bashx16,%fs:(%rax) mov -bashxffffffff,%eax retq Signed-off-by: Ravi Bangoria <ravi.bangoria@linux.vnet.ibm.com> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Chris Riyder <chris.ryder@arm.com> Cc: Kim Phillips <kim.phillips@arm.com> Cc: Markus Trippelsdorf <markus@trippelsdorf.de> Cc: Masami Hiramatsu <mhiramat@kernel.org> Cc: Naveen N. Rao <naveen.n.rao@linux.vnet.ibm.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Taeung Song <treeze.taeung@gmail.com> Cc: linuxppc-dev@lists.ozlabs.org Link: http://lkml.kernel.org/r/1480953407-7605-3-git-send-email-ravi.bangoria@linux.vnet.ibm.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-12-05 18:56:47 +03:00
if (!ops->target.addr || ops->target.offset < 0)
perf annotate: Calculate the max instruction name, align column to that We were hardcoding '6' as the max instruction name, and we have lots that are longer than that, see the diff from two 'P' printed TUI annotations for a libc function that uses instructions with long names, such as 'vpmovmskb' with its 9 chars: --- __strcmp_avx2.annotation.before 2019-03-06 16:31:39.368020425 -0300 +++ __strcmp_avx2.annotation 2019-03-06 16:32:12.079450508 -0300 @@ -2,284 +2,284 @@ Event: cycles:ppp Percent endbr64 - 0.10 mov %edi,%eax + 0.10 mov %edi,%eax - xor %edx,%edx + xor %edx,%edx - 3.54 vpxor %ymm7,%ymm7,%ymm7 + 3.54 vpxor %ymm7,%ymm7,%ymm7 - or %esi,%eax + or %esi,%eax - and $0xfff,%eax + and $0xfff,%eax - cmp $0xf80,%eax + cmp $0xf80,%eax - ↓ jg 370 + ↓ jg 370 - 27.07 vmovdqu (%rdi),%ymm1 + 27.07 vmovdqu (%rdi),%ymm1 - 7.97 vpcmpeqb (%rsi),%ymm1,%ymm0 + 7.97 vpcmpeqb (%rsi),%ymm1,%ymm0 - 2.15 vpminub %ymm1,%ymm0,%ymm0 + 2.15 vpminub %ymm1,%ymm0,%ymm0 - 4.09 vpcmpeqb %ymm7,%ymm0,%ymm0 + 4.09 vpcmpeqb %ymm7,%ymm0,%ymm0 - 0.43 vpmovmskb %ymm0,%ecx + 0.43 vpmovmskb %ymm0,%ecx - 1.53 test %ecx,%ecx + 1.53 test %ecx,%ecx - ↓ je b0 + ↓ je b0 - 5.26 tzcnt %ecx,%edx + 5.26 tzcnt %ecx,%edx - 18.40 movzbl (%rdi,%rdx,1),%eax + 18.40 movzbl (%rdi,%rdx,1),%eax - 7.09 movzbl (%rsi,%rdx,1),%edx + 7.09 movzbl (%rsi,%rdx,1),%edx - 3.34 sub %edx,%eax + 3.34 sub %edx,%eax 2.37 vzeroupper ← retq nop - 50: tzcnt %ecx,%edx + 50: tzcnt %ecx,%edx - movzbl 0x20(%rdi,%rdx,1),%eax + movzbl 0x20(%rdi,%rdx,1),%eax - movzbl 0x20(%rsi,%rdx,1),%edx + movzbl 0x20(%rsi,%rdx,1),%edx - sub %edx,%eax + sub %edx,%eax vzeroupper ← retq - data16 nopw %cs:0x0(%rax,%rax,1) + data16 nopw %cs:0x0(%rax,%rax,1) Reported-by: Travis Downs <travis.downs@gmail.com> LPU-Reference: CAOBGo4z1KfmWeOm6Et0cnX5Z6DWsG2PQbAvRn1MhVPJmXHrc5g@mail.gmail.com Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Link: https://lkml.kernel.org/n/tip-89wsdd9h9g6bvq52sgp6d0u4@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2019-03-06 22:40:15 +03:00
return ins__raw_scnprintf(ins, bf, size, ops, max_ins_name);
perf annotate: Show raw form for jump instruction with indirect target For jump instructions that does not include target address as direct operand, show the original disassembled line for them. This is needed for certain powerpc jump instructions that use target address in a register (such as bctr, btar, ...). Before: ld r12,32088(r12) mtctr r12 v bctr ffffffffffffca2c std r2,24(r1) addis r12,r2,-1 After: ld r12,32088(r12) mtctr r12 v bctr std r2,24(r1) addis r12,r2,-1 Committer notes: Testing it using a perf.data file and vmlinux for powerpc64, cross-annotating it on a x86_64 workstation: Before: .__bpf_prog_run vmlinux.powerpc │ std r10,512(r9) ▒ │ lbz r9,0(r31) ▒ │ rldicr r9,r9,3,60 ▒ │ ldx r9,r30,r9 ▒ │ mtctr r9 ▒ 100.00 │ ↓ bctr 3fffffffffe01510 ▒ │ lwa r10,4(r31) ▒ │ lwz r9,0(r31) ▒ <SNIP> Invalid jump offset: 3fffffffffe01510 After: .__bpf_prog_run vmlinux.powerpc │ std r10,512(r9) ▒ │ lbz r9,0(r31) ▒ │ rldicr r9,r9,3,60 ▒ │ ldx r9,r30,r9 ▒ │ mtctr r9 ▒ 100.00 │ ↓ bctr ▒ │ lwa r10,4(r31) ▒ │ lwz r9,0(r31) ▒ <SNIP> Invalid jump offset: 3fffffffffe01510 This, in turn, uncovers another problem with jumps without operands, the ENTER/-> operation, to jump to the target, still continues using the bogus target :-) BTW, this was the file used for the above tests: [acme@jouet ravi_bangoria]$ perf report --header-only -i perf.data.f22vm.powerdev # ======== # captured on: Thu Nov 24 12:40:38 2016 # hostname : pdev-f22-qemu # os release : 4.4.10-200.fc22.ppc64 # perf version : 4.9.rc1.g6298ce # arch : ppc64 # nrcpus online : 48 # nrcpus avail : 48 # cpudesc : POWER7 (architected), altivec supported # cpuid : 74,513 # total memory : 4158976 kB # cmdline : /home/ravi/Workspace/linux/tools/perf/perf record -a # event : name = cycles:ppp, , size = 112, { sample_period, sample_freq } = 4000, sample_type = IP|TID|TIME|CPU|PERIOD, disabled = 1, inherit = 1, mmap = 1, c # HEADER_CPU_TOPOLOGY info available, use -I to display # HEADER_NUMA_TOPOLOGY info available, use -I to display # pmu mappings: cpu = 4, software = 1, tracepoint = 2, breakpoint = 5 # missing features: HEADER_TRACING_DATA HEADER_BRANCH_STACK HEADER_GROUP_DESC HEADER_AUXTRACE HEADER_STAT HEADER_CACHE # ======== # [acme@jouet ravi_bangoria]$ Suggested-by: Michael Ellerman <mpe@ellerman.id.au> Signed-off-by: Ravi Bangoria <ravi.bangoria@linux.vnet.ibm.com> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Chris Riyder <chris.ryder@arm.com> Cc: Kim Phillips <kim.phillips@arm.com> Cc: Markus Trippelsdorf <markus@trippelsdorf.de> Cc: Masami Hiramatsu <mhiramat@kernel.org> Cc: Naveen N. Rao <naveen.n.rao@linux.vnet.ibm.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Taeung Song <treeze.taeung@gmail.com> Cc: linuxppc-dev@lists.ozlabs.org Link: http://lkml.kernel.org/r/1480953407-7605-1-git-send-email-ravi.bangoria@linux.vnet.ibm.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-12-05 18:56:45 +03:00
perf annotate: Support jumping from one function to another For instance: entry_SYSCALL_64 /lib/modules/4.16.0-rc5-00086-gdf09348f78dc/build/vmlinux 5.50 │ → callq do_syscall_64 14.56 │ mov 0x58(%rsp),%rcx 7.44 │ mov 0x80(%rsp),%r11 0.32 │ cmp %rcx,%r11 │ → jne swapgs_restore_regs_and_return_to_usermode 0.32 │ shl $0x10,%rcx 0.32 │ sar $0x10,%rcx 3.24 │ cmp %rcx,%r11 │ → jne swapgs_restore_regs_and_return_to_usermode 2.27 │ cmpq $0x33,0x88(%rsp) 1.29 │ → jne swapgs_restore_regs_and_return_to_usermode │ mov 0x30(%rsp),%r11 8.74 │ cmp %r11,0x90(%rsp) │ → jne swapgs_restore_regs_and_return_to_usermode 0.32 │ test $0x10100,%r11 │ → jne swapgs_restore_regs_and_return_to_usermode 0.32 │ cmpq $0x2b,0xa0(%rsp) 0.65 │ → jne swapgs_restore_regs_and_return_to_usermode It'll behave just like a "call" instruction, i.e. press enter or right arrow over one such line and the browser will navigate to the annotated disassembly of that function, which when exited, via left arrow or esc, will come back to the calling function. Now to support jump to an offset on a different function... Reported-by: Linus Torvalds <torvalds@linux-foundation.org> Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Andi Kleen <ak@linux.intel.com> Cc: David Ahern <dsahern@gmail.com> Cc: Jin Yao <yao.jin@linux.intel.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Wang Nan <wangnan0@huawei.com> Link: https://lkml.kernel.org/n/tip-78o508mqvr8inhj63ddtw7mo@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2018-03-23 16:50:35 +03:00
if (ops->target.outside && ops->target.sym != NULL)
perf annotate: Calculate the max instruction name, align column to that We were hardcoding '6' as the max instruction name, and we have lots that are longer than that, see the diff from two 'P' printed TUI annotations for a libc function that uses instructions with long names, such as 'vpmovmskb' with its 9 chars: --- __strcmp_avx2.annotation.before 2019-03-06 16:31:39.368020425 -0300 +++ __strcmp_avx2.annotation 2019-03-06 16:32:12.079450508 -0300 @@ -2,284 +2,284 @@ Event: cycles:ppp Percent endbr64 - 0.10 mov %edi,%eax + 0.10 mov %edi,%eax - xor %edx,%edx + xor %edx,%edx - 3.54 vpxor %ymm7,%ymm7,%ymm7 + 3.54 vpxor %ymm7,%ymm7,%ymm7 - or %esi,%eax + or %esi,%eax - and $0xfff,%eax + and $0xfff,%eax - cmp $0xf80,%eax + cmp $0xf80,%eax - ↓ jg 370 + ↓ jg 370 - 27.07 vmovdqu (%rdi),%ymm1 + 27.07 vmovdqu (%rdi),%ymm1 - 7.97 vpcmpeqb (%rsi),%ymm1,%ymm0 + 7.97 vpcmpeqb (%rsi),%ymm1,%ymm0 - 2.15 vpminub %ymm1,%ymm0,%ymm0 + 2.15 vpminub %ymm1,%ymm0,%ymm0 - 4.09 vpcmpeqb %ymm7,%ymm0,%ymm0 + 4.09 vpcmpeqb %ymm7,%ymm0,%ymm0 - 0.43 vpmovmskb %ymm0,%ecx + 0.43 vpmovmskb %ymm0,%ecx - 1.53 test %ecx,%ecx + 1.53 test %ecx,%ecx - ↓ je b0 + ↓ je b0 - 5.26 tzcnt %ecx,%edx + 5.26 tzcnt %ecx,%edx - 18.40 movzbl (%rdi,%rdx,1),%eax + 18.40 movzbl (%rdi,%rdx,1),%eax - 7.09 movzbl (%rsi,%rdx,1),%edx + 7.09 movzbl (%rsi,%rdx,1),%edx - 3.34 sub %edx,%eax + 3.34 sub %edx,%eax 2.37 vzeroupper ← retq nop - 50: tzcnt %ecx,%edx + 50: tzcnt %ecx,%edx - movzbl 0x20(%rdi,%rdx,1),%eax + movzbl 0x20(%rdi,%rdx,1),%eax - movzbl 0x20(%rsi,%rdx,1),%edx + movzbl 0x20(%rsi,%rdx,1),%edx - sub %edx,%eax + sub %edx,%eax vzeroupper ← retq - data16 nopw %cs:0x0(%rax,%rax,1) + data16 nopw %cs:0x0(%rax,%rax,1) Reported-by: Travis Downs <travis.downs@gmail.com> LPU-Reference: CAOBGo4z1KfmWeOm6Et0cnX5Z6DWsG2PQbAvRn1MhVPJmXHrc5g@mail.gmail.com Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Link: https://lkml.kernel.org/n/tip-89wsdd9h9g6bvq52sgp6d0u4@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2019-03-06 22:40:15 +03:00
return scnprintf(bf, size, "%-*s %s", max_ins_name, ins->name, ops->target.sym->name);
perf annotate: Support jumping from one function to another For instance: entry_SYSCALL_64 /lib/modules/4.16.0-rc5-00086-gdf09348f78dc/build/vmlinux 5.50 │ → callq do_syscall_64 14.56 │ mov 0x58(%rsp),%rcx 7.44 │ mov 0x80(%rsp),%r11 0.32 │ cmp %rcx,%r11 │ → jne swapgs_restore_regs_and_return_to_usermode 0.32 │ shl $0x10,%rcx 0.32 │ sar $0x10,%rcx 3.24 │ cmp %rcx,%r11 │ → jne swapgs_restore_regs_and_return_to_usermode 2.27 │ cmpq $0x33,0x88(%rsp) 1.29 │ → jne swapgs_restore_regs_and_return_to_usermode │ mov 0x30(%rsp),%r11 8.74 │ cmp %r11,0x90(%rsp) │ → jne swapgs_restore_regs_and_return_to_usermode 0.32 │ test $0x10100,%r11 │ → jne swapgs_restore_regs_and_return_to_usermode 0.32 │ cmpq $0x2b,0xa0(%rsp) 0.65 │ → jne swapgs_restore_regs_and_return_to_usermode It'll behave just like a "call" instruction, i.e. press enter or right arrow over one such line and the browser will navigate to the annotated disassembly of that function, which when exited, via left arrow or esc, will come back to the calling function. Now to support jump to an offset on a different function... Reported-by: Linus Torvalds <torvalds@linux-foundation.org> Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Andi Kleen <ak@linux.intel.com> Cc: David Ahern <dsahern@gmail.com> Cc: Jin Yao <yao.jin@linux.intel.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Wang Nan <wangnan0@huawei.com> Link: https://lkml.kernel.org/n/tip-78o508mqvr8inhj63ddtw7mo@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2018-03-23 16:50:35 +03:00
c = strchr(ops->raw, ',');
perf annotate: Fix parsing aarch64 branch instructions after objdump update Starting with binutils 2.28, aarch64 objdump adds comments to the disassembly output to show the alternative names of a condition code [1]. It is assumed that commas in objdump comments could occur in other arches now or in the future, so this fix is arch-independent. The fix could have been done with arm64 specific jump__parse and jump__scnprintf functions, but the jump__scnprintf instruction would have to have its comment character be a literal, since the scnprintf functions cannot receive a struct arch easily. This inconvenience also applies to the generic jump__scnprintf, which is why we add a raw_comment pointer to struct ins_operands, so the __parse function assigns it to be re-used by its corresponding __scnprintf function. Example differences in 'perf annotate --stdio2' output on an aarch64 perf.data file: BEFORE: → b.cs ffff200008133d1c <unwind_frame+0x18c> // b.hs, dffff7ecc47b AFTER : ↓ b.cs 18c BEFORE: → b.cc ffff200008d8d9cc <get_alloc_profile+0x31c> // b.lo, b.ul, dffff727295b AFTER : ↓ b.cc 31c The branch target labels 18c and 31c also now appear in the output: BEFORE: add x26, x29, #0x80 AFTER : 18c: add x26, x29, #0x80 BEFORE: add x21, x21, #0x8 AFTER : 31c: add x21, x21, #0x8 The Fixes: tag below is added so stable branches will get the update; it doesn't necessarily mean that commit was broken at the time, rather it didn't withstand the aarch64 objdump update. Tested no difference in output for sample x86_64, power arch perf.data files. [1] https://sourceware.org/git/gitweb.cgi?p=binutils-gdb.git;a=commit;h=bb7eff5206e4795ac79c177a80fe9f4630aaf730 Signed-off-by: Kim Phillips <kim.phillips@arm.com> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Anton Blanchard <anton@samba.org> Cc: Christian Borntraeger <borntraeger@de.ibm.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Mark Rutland <mark.rutland@arm.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Ravi Bangoria <ravi.bangoria@linux.vnet.ibm.com> Cc: Robin Murphy <robin.murphy@arm.com> Cc: Taeung Song <treeze.taeung@gmail.com> Cc: linux-arm-kernel@lists.infradead.org Fixes: b13bbeee5ee6 ("perf annotate: Fix branch instruction with multiple operands") Link: http://lkml.kernel.org/r/20180827125340.a2f7e291901d17cea05daba4@arm.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2018-08-27 20:53:40 +03:00
c = validate_comma(c, ops);
perf annotate: Fix branch instruction with multiple operands 'perf annotate' is dropping the cr* fields from branch instructions. Fix it by adding support to display branch instructions having multiple operands. Power Arch objdump of int_sqrt: 20.36 | c0000000004d2694: subf r10,r10,r3 | c0000000004d2698: v bgt cr6,c0000000004d26a0 <int_sqrt+0x40> 1.82 | c0000000004d269c: mr r3,r10 29.18 | c0000000004d26a0: mr r10,r8 | c0000000004d26a4: v bgt cr7,c0000000004d26ac <int_sqrt+0x4c> | c0000000004d26a8: mr r10,r7 Power Arch Before Patch: 20.36 | subf r10,r10,r3 | v bgt 40 1.82 | mr r3,r10 29.18 | 40: mr r10,r8 | v bgt 4c | mr r10,r7 Power Arch After patch: 20.36 | subf r10,r10,r3 | v bgt cr6,40 1.82 | mr r3,r10 29.18 | 40: mr r10,r8 | v bgt cr7,4c | mr r10,r7 Also support AArch64 conditional branch instructions, which can have up to three operands: Aarch64 Non-simplified (raw objdump) view: │ffff0000083cd11c: ↑ cbz w0, ffff0000083cd100 <security_fil▒ ... 4.44 │ffff000│083cd134: ↓ tbnz w0, #26, ffff0000083cd190 <securit▒ ... 1.37 │ffff000│083cd144: ↓ tbnz w22, #5, ffff0000083cd1a4 <securit▒ │ffff000│083cd148: mov w19, #0x20000 //▒ 1.02 │ffff000│083cd14c: ↓ tbz w22, #2, ffff0000083cd1ac <securit▒ ... 0.68 │ffff000└──3cd16c: ↑ cbnz w0, ffff0000083cd120 <security_fil▒ Aarch64 Simplified, before this patch: │ ↑ cbz 40 ... 4.44 │ │↓ tbnz w0, #26, ffff0000083cd190 <security_file_permiss▒ ... 1.37 │ │↓ tbnz w22, #5, ffff0000083cd1a4 <security_file_permiss▒ │ │ mov w19, #0x20000 // #131072 1.02 │ │↓ tbz w22, #2, ffff0000083cd1ac <security_file_permiss▒ ... 0.68 │ └──cbnz 60 the cbz operand is missing, and the tbz doesn't get simplified processing at all because the parsing function failed to match an address. Aarch64 Simplified, After this patch applied: │ ↑ cbz w0, 40 ... 4.44 │ │↓ tbnz w0, #26, d0 ... 1.37 │ │↓ tbnz w22, #5, e4 │ │ mov w19, #0x20000 // #131072 1.02 │ │↓ tbz w22, #2, ec ... 0.68 │ └──cbnz w0, 60 Originally-by: Ravi Bangoria <ravi.bangoria@linux.vnet.ibm.com> Tested-by: Ravi Bangoria <ravi.bangoria@linux.vnet.ibm.com> Reported-by: Anton Blanchard <anton@samba.org> Reported-by: Robin Murphy <robin.murphy@arm.com> Signed-off-by: Kim Phillips <kim.phillips@arm.com> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Christian Borntraeger <borntraeger@de.ibm.com> Cc: Mark Rutland <mark.rutland@arm.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Taeung Song <treeze.taeung@gmail.com> Link: http://lkml.kernel.org/r/20170601092959.f60d98912e8a1b66fd1e4c0e@arm.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2017-06-01 17:29:59 +03:00
if (c != NULL) {
const char *c2 = strchr(c + 1, ',');
perf annotate: Fix parsing aarch64 branch instructions after objdump update Starting with binutils 2.28, aarch64 objdump adds comments to the disassembly output to show the alternative names of a condition code [1]. It is assumed that commas in objdump comments could occur in other arches now or in the future, so this fix is arch-independent. The fix could have been done with arm64 specific jump__parse and jump__scnprintf functions, but the jump__scnprintf instruction would have to have its comment character be a literal, since the scnprintf functions cannot receive a struct arch easily. This inconvenience also applies to the generic jump__scnprintf, which is why we add a raw_comment pointer to struct ins_operands, so the __parse function assigns it to be re-used by its corresponding __scnprintf function. Example differences in 'perf annotate --stdio2' output on an aarch64 perf.data file: BEFORE: → b.cs ffff200008133d1c <unwind_frame+0x18c> // b.hs, dffff7ecc47b AFTER : ↓ b.cs 18c BEFORE: → b.cc ffff200008d8d9cc <get_alloc_profile+0x31c> // b.lo, b.ul, dffff727295b AFTER : ↓ b.cc 31c The branch target labels 18c and 31c also now appear in the output: BEFORE: add x26, x29, #0x80 AFTER : 18c: add x26, x29, #0x80 BEFORE: add x21, x21, #0x8 AFTER : 31c: add x21, x21, #0x8 The Fixes: tag below is added so stable branches will get the update; it doesn't necessarily mean that commit was broken at the time, rather it didn't withstand the aarch64 objdump update. Tested no difference in output for sample x86_64, power arch perf.data files. [1] https://sourceware.org/git/gitweb.cgi?p=binutils-gdb.git;a=commit;h=bb7eff5206e4795ac79c177a80fe9f4630aaf730 Signed-off-by: Kim Phillips <kim.phillips@arm.com> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Anton Blanchard <anton@samba.org> Cc: Christian Borntraeger <borntraeger@de.ibm.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Mark Rutland <mark.rutland@arm.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Ravi Bangoria <ravi.bangoria@linux.vnet.ibm.com> Cc: Robin Murphy <robin.murphy@arm.com> Cc: Taeung Song <treeze.taeung@gmail.com> Cc: linux-arm-kernel@lists.infradead.org Fixes: b13bbeee5ee6 ("perf annotate: Fix branch instruction with multiple operands") Link: http://lkml.kernel.org/r/20180827125340.a2f7e291901d17cea05daba4@arm.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2018-08-27 20:53:40 +03:00
c2 = validate_comma(c2, ops);
perf annotate: Fix branch instruction with multiple operands 'perf annotate' is dropping the cr* fields from branch instructions. Fix it by adding support to display branch instructions having multiple operands. Power Arch objdump of int_sqrt: 20.36 | c0000000004d2694: subf r10,r10,r3 | c0000000004d2698: v bgt cr6,c0000000004d26a0 <int_sqrt+0x40> 1.82 | c0000000004d269c: mr r3,r10 29.18 | c0000000004d26a0: mr r10,r8 | c0000000004d26a4: v bgt cr7,c0000000004d26ac <int_sqrt+0x4c> | c0000000004d26a8: mr r10,r7 Power Arch Before Patch: 20.36 | subf r10,r10,r3 | v bgt 40 1.82 | mr r3,r10 29.18 | 40: mr r10,r8 | v bgt 4c | mr r10,r7 Power Arch After patch: 20.36 | subf r10,r10,r3 | v bgt cr6,40 1.82 | mr r3,r10 29.18 | 40: mr r10,r8 | v bgt cr7,4c | mr r10,r7 Also support AArch64 conditional branch instructions, which can have up to three operands: Aarch64 Non-simplified (raw objdump) view: │ffff0000083cd11c: ↑ cbz w0, ffff0000083cd100 <security_fil▒ ... 4.44 │ffff000│083cd134: ↓ tbnz w0, #26, ffff0000083cd190 <securit▒ ... 1.37 │ffff000│083cd144: ↓ tbnz w22, #5, ffff0000083cd1a4 <securit▒ │ffff000│083cd148: mov w19, #0x20000 //▒ 1.02 │ffff000│083cd14c: ↓ tbz w22, #2, ffff0000083cd1ac <securit▒ ... 0.68 │ffff000└──3cd16c: ↑ cbnz w0, ffff0000083cd120 <security_fil▒ Aarch64 Simplified, before this patch: │ ↑ cbz 40 ... 4.44 │ │↓ tbnz w0, #26, ffff0000083cd190 <security_file_permiss▒ ... 1.37 │ │↓ tbnz w22, #5, ffff0000083cd1a4 <security_file_permiss▒ │ │ mov w19, #0x20000 // #131072 1.02 │ │↓ tbz w22, #2, ffff0000083cd1ac <security_file_permiss▒ ... 0.68 │ └──cbnz 60 the cbz operand is missing, and the tbz doesn't get simplified processing at all because the parsing function failed to match an address. Aarch64 Simplified, After this patch applied: │ ↑ cbz w0, 40 ... 4.44 │ │↓ tbnz w0, #26, d0 ... 1.37 │ │↓ tbnz w22, #5, e4 │ │ mov w19, #0x20000 // #131072 1.02 │ │↓ tbz w22, #2, ec ... 0.68 │ └──cbnz w0, 60 Originally-by: Ravi Bangoria <ravi.bangoria@linux.vnet.ibm.com> Tested-by: Ravi Bangoria <ravi.bangoria@linux.vnet.ibm.com> Reported-by: Anton Blanchard <anton@samba.org> Reported-by: Robin Murphy <robin.murphy@arm.com> Signed-off-by: Kim Phillips <kim.phillips@arm.com> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Christian Borntraeger <borntraeger@de.ibm.com> Cc: Mark Rutland <mark.rutland@arm.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Taeung Song <treeze.taeung@gmail.com> Link: http://lkml.kernel.org/r/20170601092959.f60d98912e8a1b66fd1e4c0e@arm.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2017-06-01 17:29:59 +03:00
/* check for 3-op insn */
if (c2 != NULL)
c = c2;
c++;
/* mirror arch objdump's space-after-comma style */
if (*c == ' ')
c++;
}
perf annotate: Calculate the max instruction name, align column to that We were hardcoding '6' as the max instruction name, and we have lots that are longer than that, see the diff from two 'P' printed TUI annotations for a libc function that uses instructions with long names, such as 'vpmovmskb' with its 9 chars: --- __strcmp_avx2.annotation.before 2019-03-06 16:31:39.368020425 -0300 +++ __strcmp_avx2.annotation 2019-03-06 16:32:12.079450508 -0300 @@ -2,284 +2,284 @@ Event: cycles:ppp Percent endbr64 - 0.10 mov %edi,%eax + 0.10 mov %edi,%eax - xor %edx,%edx + xor %edx,%edx - 3.54 vpxor %ymm7,%ymm7,%ymm7 + 3.54 vpxor %ymm7,%ymm7,%ymm7 - or %esi,%eax + or %esi,%eax - and $0xfff,%eax + and $0xfff,%eax - cmp $0xf80,%eax + cmp $0xf80,%eax - ↓ jg 370 + ↓ jg 370 - 27.07 vmovdqu (%rdi),%ymm1 + 27.07 vmovdqu (%rdi),%ymm1 - 7.97 vpcmpeqb (%rsi),%ymm1,%ymm0 + 7.97 vpcmpeqb (%rsi),%ymm1,%ymm0 - 2.15 vpminub %ymm1,%ymm0,%ymm0 + 2.15 vpminub %ymm1,%ymm0,%ymm0 - 4.09 vpcmpeqb %ymm7,%ymm0,%ymm0 + 4.09 vpcmpeqb %ymm7,%ymm0,%ymm0 - 0.43 vpmovmskb %ymm0,%ecx + 0.43 vpmovmskb %ymm0,%ecx - 1.53 test %ecx,%ecx + 1.53 test %ecx,%ecx - ↓ je b0 + ↓ je b0 - 5.26 tzcnt %ecx,%edx + 5.26 tzcnt %ecx,%edx - 18.40 movzbl (%rdi,%rdx,1),%eax + 18.40 movzbl (%rdi,%rdx,1),%eax - 7.09 movzbl (%rsi,%rdx,1),%edx + 7.09 movzbl (%rsi,%rdx,1),%edx - 3.34 sub %edx,%eax + 3.34 sub %edx,%eax 2.37 vzeroupper ← retq nop - 50: tzcnt %ecx,%edx + 50: tzcnt %ecx,%edx - movzbl 0x20(%rdi,%rdx,1),%eax + movzbl 0x20(%rdi,%rdx,1),%eax - movzbl 0x20(%rsi,%rdx,1),%edx + movzbl 0x20(%rsi,%rdx,1),%edx - sub %edx,%eax + sub %edx,%eax vzeroupper ← retq - data16 nopw %cs:0x0(%rax,%rax,1) + data16 nopw %cs:0x0(%rax,%rax,1) Reported-by: Travis Downs <travis.downs@gmail.com> LPU-Reference: CAOBGo4z1KfmWeOm6Et0cnX5Z6DWsG2PQbAvRn1MhVPJmXHrc5g@mail.gmail.com Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Link: https://lkml.kernel.org/n/tip-89wsdd9h9g6bvq52sgp6d0u4@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2019-03-06 22:40:15 +03:00
return scnprintf(bf, size, "%-*s %.*s%" PRIx64, max_ins_name,
perf annotate: Fix branch instruction with multiple operands 'perf annotate' is dropping the cr* fields from branch instructions. Fix it by adding support to display branch instructions having multiple operands. Power Arch objdump of int_sqrt: 20.36 | c0000000004d2694: subf r10,r10,r3 | c0000000004d2698: v bgt cr6,c0000000004d26a0 <int_sqrt+0x40> 1.82 | c0000000004d269c: mr r3,r10 29.18 | c0000000004d26a0: mr r10,r8 | c0000000004d26a4: v bgt cr7,c0000000004d26ac <int_sqrt+0x4c> | c0000000004d26a8: mr r10,r7 Power Arch Before Patch: 20.36 | subf r10,r10,r3 | v bgt 40 1.82 | mr r3,r10 29.18 | 40: mr r10,r8 | v bgt 4c | mr r10,r7 Power Arch After patch: 20.36 | subf r10,r10,r3 | v bgt cr6,40 1.82 | mr r3,r10 29.18 | 40: mr r10,r8 | v bgt cr7,4c | mr r10,r7 Also support AArch64 conditional branch instructions, which can have up to three operands: Aarch64 Non-simplified (raw objdump) view: │ffff0000083cd11c: ↑ cbz w0, ffff0000083cd100 <security_fil▒ ... 4.44 │ffff000│083cd134: ↓ tbnz w0, #26, ffff0000083cd190 <securit▒ ... 1.37 │ffff000│083cd144: ↓ tbnz w22, #5, ffff0000083cd1a4 <securit▒ │ffff000│083cd148: mov w19, #0x20000 //▒ 1.02 │ffff000│083cd14c: ↓ tbz w22, #2, ffff0000083cd1ac <securit▒ ... 0.68 │ffff000└──3cd16c: ↑ cbnz w0, ffff0000083cd120 <security_fil▒ Aarch64 Simplified, before this patch: │ ↑ cbz 40 ... 4.44 │ │↓ tbnz w0, #26, ffff0000083cd190 <security_file_permiss▒ ... 1.37 │ │↓ tbnz w22, #5, ffff0000083cd1a4 <security_file_permiss▒ │ │ mov w19, #0x20000 // #131072 1.02 │ │↓ tbz w22, #2, ffff0000083cd1ac <security_file_permiss▒ ... 0.68 │ └──cbnz 60 the cbz operand is missing, and the tbz doesn't get simplified processing at all because the parsing function failed to match an address. Aarch64 Simplified, After this patch applied: │ ↑ cbz w0, 40 ... 4.44 │ │↓ tbnz w0, #26, d0 ... 1.37 │ │↓ tbnz w22, #5, e4 │ │ mov w19, #0x20000 // #131072 1.02 │ │↓ tbz w22, #2, ec ... 0.68 │ └──cbnz w0, 60 Originally-by: Ravi Bangoria <ravi.bangoria@linux.vnet.ibm.com> Tested-by: Ravi Bangoria <ravi.bangoria@linux.vnet.ibm.com> Reported-by: Anton Blanchard <anton@samba.org> Reported-by: Robin Murphy <robin.murphy@arm.com> Signed-off-by: Kim Phillips <kim.phillips@arm.com> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Christian Borntraeger <borntraeger@de.ibm.com> Cc: Mark Rutland <mark.rutland@arm.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Taeung Song <treeze.taeung@gmail.com> Link: http://lkml.kernel.org/r/20170601092959.f60d98912e8a1b66fd1e4c0e@arm.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2017-06-01 17:29:59 +03:00
ins->name, c ? c - ops->raw : 0, ops->raw,
ops->target.offset);
}
static struct ins_ops jump_ops = {
.parse = jump__parse,
.scnprintf = jump__scnprintf,
};
bool ins__is_jump(const struct ins *ins)
{
return ins->ops == &jump_ops;
}
static int comment__symbol(char *raw, char *comment, u64 *addrp, char **namep)
{
char *endptr, *name, *t;
if (strstr(raw, "(%rip)") == NULL)
return 0;
*addrp = strtoull(comment, &endptr, 16);
perf annotate: Fix objdump comment parsing for Intel mov dissassembly The command 'perf annotate' parses the output of objdump and also investigates the comments produced by objdump. For example the output of objdump produces (on x86): 23eee: 4c 8b 3d 13 01 21 00 mov 0x210113(%rip),%r15 # 234008 <stderr@@GLIBC_2.2.5+0x9a8> and the function mov__parse() is called to investigate the complete line. Mov__parse() breaks this line into several parts and finally calls function comment__symbol() to parse the data after the comment character '#'. Comment__symbol() expects a hexadecimal address followed by a symbol in '<' and '>' brackets. However the 2nd parameter given to function comment__symbol() always points to the comment character '#'. The address parsing always returns 0 because the character '#' is not a digit and strtoull() fails without being noticed. Fix this by advancing the second parameter to function comment__symbol() by one byte before invocation and add an error check after strtoull() has been called. Signed-off-by: Thomas Richter <tmricht@linux.vnet.ibm.com> Reviewed-by: Hendrik Brueckner <brueckner@linux.vnet.ibm.com> Acked-by: Ravi Bangoria <ravi.bangoria@linux.vnet.ibm.com> Cc: Heiko Carstens <heiko.carstens@de.ibm.com> Cc: Martin Schwidefsky <schwidefsky@de.ibm.com> Fixes: 6de783b6f50f ("perf annotate: Resolve symbols using objdump comment") Link: http://lkml.kernel.org/r/20171128075632.72182-1-tmricht@linux.vnet.ibm.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2017-11-28 10:56:32 +03:00
if (endptr == comment)
return 0;
name = strchr(endptr, '<');
if (name == NULL)
return -1;
name++;
t = strchr(name, '>');
if (t == NULL)
return 0;
*t = '\0';
*namep = strdup(name);
*t = '>';
return 0;
}
static int lock__parse(struct arch *arch, struct ins_operands *ops, struct map_symbol *ms)
{
ops->locked.ops = zalloc(sizeof(*ops->locked.ops));
if (ops->locked.ops == NULL)
return 0;
perf annotate: Remove duplicate 'name' field from disasm_line The disasm_line::name field is always equal to ins::name, being used just to locate the instruction's ins_ops from the per-arch instructions table. Eliminate this duplication, nuking that field and instead make ins__find() return an ins_ops, store it in disasm_line::ins.ops, and keep just in disasm_line::ins.name what was in disasm_line::name, this way we end up not keeping a reference to entries in the per-arch instructions table. This in turn will help supporting multiple ways to manage the per-arch instructions table, allowing resorting that array, for instance, when the entries will move after references to its addresses were made. The same problem is avoided when one grows the array with realloc. So architectures simply keeping a constant array will work as well as architectures building the table using regular expressions or other logic that involves resorting the table. Reviewed-by: Ravi Bangoria <ravi.bangoria@linux.vnet.ibm.com> Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Chris Riyder <chris.ryder@arm.com> Cc: David Ahern <dsahern@gmail.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Kim Phillips <kim.phillips@arm.com> Cc: Markus Trippelsdorf <markus@trippelsdorf.de> Cc: Masami Hiramatsu <mhiramat@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Naveen N. Rao <naveen.n.rao@linux.vnet.ibm.com> Cc: Pawel Moll <pawel.moll@arm.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Russell King <rmk+kernel@arm.linux.org.uk> Cc: Taeung Song <treeze.taeung@gmail.com> Cc: Wang Nan <wangnan0@huawei.com> Link: http://lkml.kernel.org/n/tip-vr899azvabnw9gtuepuqfd9t@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-11-24 17:16:06 +03:00
if (disasm_line__parse(ops->raw, &ops->locked.ins.name, &ops->locked.ops->raw) < 0)
goto out_free_ops;
perf annotate: Remove duplicate 'name' field from disasm_line The disasm_line::name field is always equal to ins::name, being used just to locate the instruction's ins_ops from the per-arch instructions table. Eliminate this duplication, nuking that field and instead make ins__find() return an ins_ops, store it in disasm_line::ins.ops, and keep just in disasm_line::ins.name what was in disasm_line::name, this way we end up not keeping a reference to entries in the per-arch instructions table. This in turn will help supporting multiple ways to manage the per-arch instructions table, allowing resorting that array, for instance, when the entries will move after references to its addresses were made. The same problem is avoided when one grows the array with realloc. So architectures simply keeping a constant array will work as well as architectures building the table using regular expressions or other logic that involves resorting the table. Reviewed-by: Ravi Bangoria <ravi.bangoria@linux.vnet.ibm.com> Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Chris Riyder <chris.ryder@arm.com> Cc: David Ahern <dsahern@gmail.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Kim Phillips <kim.phillips@arm.com> Cc: Markus Trippelsdorf <markus@trippelsdorf.de> Cc: Masami Hiramatsu <mhiramat@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Naveen N. Rao <naveen.n.rao@linux.vnet.ibm.com> Cc: Pawel Moll <pawel.moll@arm.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Russell King <rmk+kernel@arm.linux.org.uk> Cc: Taeung Song <treeze.taeung@gmail.com> Cc: Wang Nan <wangnan0@huawei.com> Link: http://lkml.kernel.org/n/tip-vr899azvabnw9gtuepuqfd9t@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-11-24 17:16:06 +03:00
ops->locked.ins.ops = ins__find(arch, ops->locked.ins.name);
perf annotate: Remove duplicate 'name' field from disasm_line The disasm_line::name field is always equal to ins::name, being used just to locate the instruction's ins_ops from the per-arch instructions table. Eliminate this duplication, nuking that field and instead make ins__find() return an ins_ops, store it in disasm_line::ins.ops, and keep just in disasm_line::ins.name what was in disasm_line::name, this way we end up not keeping a reference to entries in the per-arch instructions table. This in turn will help supporting multiple ways to manage the per-arch instructions table, allowing resorting that array, for instance, when the entries will move after references to its addresses were made. The same problem is avoided when one grows the array with realloc. So architectures simply keeping a constant array will work as well as architectures building the table using regular expressions or other logic that involves resorting the table. Reviewed-by: Ravi Bangoria <ravi.bangoria@linux.vnet.ibm.com> Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Chris Riyder <chris.ryder@arm.com> Cc: David Ahern <dsahern@gmail.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Kim Phillips <kim.phillips@arm.com> Cc: Markus Trippelsdorf <markus@trippelsdorf.de> Cc: Masami Hiramatsu <mhiramat@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Naveen N. Rao <naveen.n.rao@linux.vnet.ibm.com> Cc: Pawel Moll <pawel.moll@arm.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Russell King <rmk+kernel@arm.linux.org.uk> Cc: Taeung Song <treeze.taeung@gmail.com> Cc: Wang Nan <wangnan0@huawei.com> Link: http://lkml.kernel.org/n/tip-vr899azvabnw9gtuepuqfd9t@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-11-24 17:16:06 +03:00
if (ops->locked.ins.ops == NULL)
goto out_free_ops;
perf annotate: Remove duplicate 'name' field from disasm_line The disasm_line::name field is always equal to ins::name, being used just to locate the instruction's ins_ops from the per-arch instructions table. Eliminate this duplication, nuking that field and instead make ins__find() return an ins_ops, store it in disasm_line::ins.ops, and keep just in disasm_line::ins.name what was in disasm_line::name, this way we end up not keeping a reference to entries in the per-arch instructions table. This in turn will help supporting multiple ways to manage the per-arch instructions table, allowing resorting that array, for instance, when the entries will move after references to its addresses were made. The same problem is avoided when one grows the array with realloc. So architectures simply keeping a constant array will work as well as architectures building the table using regular expressions or other logic that involves resorting the table. Reviewed-by: Ravi Bangoria <ravi.bangoria@linux.vnet.ibm.com> Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Chris Riyder <chris.ryder@arm.com> Cc: David Ahern <dsahern@gmail.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Kim Phillips <kim.phillips@arm.com> Cc: Markus Trippelsdorf <markus@trippelsdorf.de> Cc: Masami Hiramatsu <mhiramat@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Naveen N. Rao <naveen.n.rao@linux.vnet.ibm.com> Cc: Pawel Moll <pawel.moll@arm.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Russell King <rmk+kernel@arm.linux.org.uk> Cc: Taeung Song <treeze.taeung@gmail.com> Cc: Wang Nan <wangnan0@huawei.com> Link: http://lkml.kernel.org/n/tip-vr899azvabnw9gtuepuqfd9t@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-11-24 17:16:06 +03:00
if (ops->locked.ins.ops->parse &&
ops->locked.ins.ops->parse(arch, ops->locked.ops, ms) < 0)
goto out_free_ops;
return 0;
out_free_ops:
zfree(&ops->locked.ops);
return 0;
}
static int lock__scnprintf(struct ins *ins, char *bf, size_t size,
perf annotate: Calculate the max instruction name, align column to that We were hardcoding '6' as the max instruction name, and we have lots that are longer than that, see the diff from two 'P' printed TUI annotations for a libc function that uses instructions with long names, such as 'vpmovmskb' with its 9 chars: --- __strcmp_avx2.annotation.before 2019-03-06 16:31:39.368020425 -0300 +++ __strcmp_avx2.annotation 2019-03-06 16:32:12.079450508 -0300 @@ -2,284 +2,284 @@ Event: cycles:ppp Percent endbr64 - 0.10 mov %edi,%eax + 0.10 mov %edi,%eax - xor %edx,%edx + xor %edx,%edx - 3.54 vpxor %ymm7,%ymm7,%ymm7 + 3.54 vpxor %ymm7,%ymm7,%ymm7 - or %esi,%eax + or %esi,%eax - and $0xfff,%eax + and $0xfff,%eax - cmp $0xf80,%eax + cmp $0xf80,%eax - ↓ jg 370 + ↓ jg 370 - 27.07 vmovdqu (%rdi),%ymm1 + 27.07 vmovdqu (%rdi),%ymm1 - 7.97 vpcmpeqb (%rsi),%ymm1,%ymm0 + 7.97 vpcmpeqb (%rsi),%ymm1,%ymm0 - 2.15 vpminub %ymm1,%ymm0,%ymm0 + 2.15 vpminub %ymm1,%ymm0,%ymm0 - 4.09 vpcmpeqb %ymm7,%ymm0,%ymm0 + 4.09 vpcmpeqb %ymm7,%ymm0,%ymm0 - 0.43 vpmovmskb %ymm0,%ecx + 0.43 vpmovmskb %ymm0,%ecx - 1.53 test %ecx,%ecx + 1.53 test %ecx,%ecx - ↓ je b0 + ↓ je b0 - 5.26 tzcnt %ecx,%edx + 5.26 tzcnt %ecx,%edx - 18.40 movzbl (%rdi,%rdx,1),%eax + 18.40 movzbl (%rdi,%rdx,1),%eax - 7.09 movzbl (%rsi,%rdx,1),%edx + 7.09 movzbl (%rsi,%rdx,1),%edx - 3.34 sub %edx,%eax + 3.34 sub %edx,%eax 2.37 vzeroupper ← retq nop - 50: tzcnt %ecx,%edx + 50: tzcnt %ecx,%edx - movzbl 0x20(%rdi,%rdx,1),%eax + movzbl 0x20(%rdi,%rdx,1),%eax - movzbl 0x20(%rsi,%rdx,1),%edx + movzbl 0x20(%rsi,%rdx,1),%edx - sub %edx,%eax + sub %edx,%eax vzeroupper ← retq - data16 nopw %cs:0x0(%rax,%rax,1) + data16 nopw %cs:0x0(%rax,%rax,1) Reported-by: Travis Downs <travis.downs@gmail.com> LPU-Reference: CAOBGo4z1KfmWeOm6Et0cnX5Z6DWsG2PQbAvRn1MhVPJmXHrc5g@mail.gmail.com Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Link: https://lkml.kernel.org/n/tip-89wsdd9h9g6bvq52sgp6d0u4@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2019-03-06 22:40:15 +03:00
struct ins_operands *ops, int max_ins_name)
{
int printed;
perf annotate: Remove duplicate 'name' field from disasm_line The disasm_line::name field is always equal to ins::name, being used just to locate the instruction's ins_ops from the per-arch instructions table. Eliminate this duplication, nuking that field and instead make ins__find() return an ins_ops, store it in disasm_line::ins.ops, and keep just in disasm_line::ins.name what was in disasm_line::name, this way we end up not keeping a reference to entries in the per-arch instructions table. This in turn will help supporting multiple ways to manage the per-arch instructions table, allowing resorting that array, for instance, when the entries will move after references to its addresses were made. The same problem is avoided when one grows the array with realloc. So architectures simply keeping a constant array will work as well as architectures building the table using regular expressions or other logic that involves resorting the table. Reviewed-by: Ravi Bangoria <ravi.bangoria@linux.vnet.ibm.com> Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Chris Riyder <chris.ryder@arm.com> Cc: David Ahern <dsahern@gmail.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Kim Phillips <kim.phillips@arm.com> Cc: Markus Trippelsdorf <markus@trippelsdorf.de> Cc: Masami Hiramatsu <mhiramat@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Naveen N. Rao <naveen.n.rao@linux.vnet.ibm.com> Cc: Pawel Moll <pawel.moll@arm.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Russell King <rmk+kernel@arm.linux.org.uk> Cc: Taeung Song <treeze.taeung@gmail.com> Cc: Wang Nan <wangnan0@huawei.com> Link: http://lkml.kernel.org/n/tip-vr899azvabnw9gtuepuqfd9t@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-11-24 17:16:06 +03:00
if (ops->locked.ins.ops == NULL)
perf annotate: Calculate the max instruction name, align column to that We were hardcoding '6' as the max instruction name, and we have lots that are longer than that, see the diff from two 'P' printed TUI annotations for a libc function that uses instructions with long names, such as 'vpmovmskb' with its 9 chars: --- __strcmp_avx2.annotation.before 2019-03-06 16:31:39.368020425 -0300 +++ __strcmp_avx2.annotation 2019-03-06 16:32:12.079450508 -0300 @@ -2,284 +2,284 @@ Event: cycles:ppp Percent endbr64 - 0.10 mov %edi,%eax + 0.10 mov %edi,%eax - xor %edx,%edx + xor %edx,%edx - 3.54 vpxor %ymm7,%ymm7,%ymm7 + 3.54 vpxor %ymm7,%ymm7,%ymm7 - or %esi,%eax + or %esi,%eax - and $0xfff,%eax + and $0xfff,%eax - cmp $0xf80,%eax + cmp $0xf80,%eax - ↓ jg 370 + ↓ jg 370 - 27.07 vmovdqu (%rdi),%ymm1 + 27.07 vmovdqu (%rdi),%ymm1 - 7.97 vpcmpeqb (%rsi),%ymm1,%ymm0 + 7.97 vpcmpeqb (%rsi),%ymm1,%ymm0 - 2.15 vpminub %ymm1,%ymm0,%ymm0 + 2.15 vpminub %ymm1,%ymm0,%ymm0 - 4.09 vpcmpeqb %ymm7,%ymm0,%ymm0 + 4.09 vpcmpeqb %ymm7,%ymm0,%ymm0 - 0.43 vpmovmskb %ymm0,%ecx + 0.43 vpmovmskb %ymm0,%ecx - 1.53 test %ecx,%ecx + 1.53 test %ecx,%ecx - ↓ je b0 + ↓ je b0 - 5.26 tzcnt %ecx,%edx + 5.26 tzcnt %ecx,%edx - 18.40 movzbl (%rdi,%rdx,1),%eax + 18.40 movzbl (%rdi,%rdx,1),%eax - 7.09 movzbl (%rsi,%rdx,1),%edx + 7.09 movzbl (%rsi,%rdx,1),%edx - 3.34 sub %edx,%eax + 3.34 sub %edx,%eax 2.37 vzeroupper ← retq nop - 50: tzcnt %ecx,%edx + 50: tzcnt %ecx,%edx - movzbl 0x20(%rdi,%rdx,1),%eax + movzbl 0x20(%rdi,%rdx,1),%eax - movzbl 0x20(%rsi,%rdx,1),%edx + movzbl 0x20(%rsi,%rdx,1),%edx - sub %edx,%eax + sub %edx,%eax vzeroupper ← retq - data16 nopw %cs:0x0(%rax,%rax,1) + data16 nopw %cs:0x0(%rax,%rax,1) Reported-by: Travis Downs <travis.downs@gmail.com> LPU-Reference: CAOBGo4z1KfmWeOm6Et0cnX5Z6DWsG2PQbAvRn1MhVPJmXHrc5g@mail.gmail.com Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Link: https://lkml.kernel.org/n/tip-89wsdd9h9g6bvq52sgp6d0u4@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2019-03-06 22:40:15 +03:00
return ins__raw_scnprintf(ins, bf, size, ops, max_ins_name);
perf annotate: Calculate the max instruction name, align column to that We were hardcoding '6' as the max instruction name, and we have lots that are longer than that, see the diff from two 'P' printed TUI annotations for a libc function that uses instructions with long names, such as 'vpmovmskb' with its 9 chars: --- __strcmp_avx2.annotation.before 2019-03-06 16:31:39.368020425 -0300 +++ __strcmp_avx2.annotation 2019-03-06 16:32:12.079450508 -0300 @@ -2,284 +2,284 @@ Event: cycles:ppp Percent endbr64 - 0.10 mov %edi,%eax + 0.10 mov %edi,%eax - xor %edx,%edx + xor %edx,%edx - 3.54 vpxor %ymm7,%ymm7,%ymm7 + 3.54 vpxor %ymm7,%ymm7,%ymm7 - or %esi,%eax + or %esi,%eax - and $0xfff,%eax + and $0xfff,%eax - cmp $0xf80,%eax + cmp $0xf80,%eax - ↓ jg 370 + ↓ jg 370 - 27.07 vmovdqu (%rdi),%ymm1 + 27.07 vmovdqu (%rdi),%ymm1 - 7.97 vpcmpeqb (%rsi),%ymm1,%ymm0 + 7.97 vpcmpeqb (%rsi),%ymm1,%ymm0 - 2.15 vpminub %ymm1,%ymm0,%ymm0 + 2.15 vpminub %ymm1,%ymm0,%ymm0 - 4.09 vpcmpeqb %ymm7,%ymm0,%ymm0 + 4.09 vpcmpeqb %ymm7,%ymm0,%ymm0 - 0.43 vpmovmskb %ymm0,%ecx + 0.43 vpmovmskb %ymm0,%ecx - 1.53 test %ecx,%ecx + 1.53 test %ecx,%ecx - ↓ je b0 + ↓ je b0 - 5.26 tzcnt %ecx,%edx + 5.26 tzcnt %ecx,%edx - 18.40 movzbl (%rdi,%rdx,1),%eax + 18.40 movzbl (%rdi,%rdx,1),%eax - 7.09 movzbl (%rsi,%rdx,1),%edx + 7.09 movzbl (%rsi,%rdx,1),%edx - 3.34 sub %edx,%eax + 3.34 sub %edx,%eax 2.37 vzeroupper ← retq nop - 50: tzcnt %ecx,%edx + 50: tzcnt %ecx,%edx - movzbl 0x20(%rdi,%rdx,1),%eax + movzbl 0x20(%rdi,%rdx,1),%eax - movzbl 0x20(%rsi,%rdx,1),%edx + movzbl 0x20(%rsi,%rdx,1),%edx - sub %edx,%eax + sub %edx,%eax vzeroupper ← retq - data16 nopw %cs:0x0(%rax,%rax,1) + data16 nopw %cs:0x0(%rax,%rax,1) Reported-by: Travis Downs <travis.downs@gmail.com> LPU-Reference: CAOBGo4z1KfmWeOm6Et0cnX5Z6DWsG2PQbAvRn1MhVPJmXHrc5g@mail.gmail.com Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Link: https://lkml.kernel.org/n/tip-89wsdd9h9g6bvq52sgp6d0u4@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2019-03-06 22:40:15 +03:00
printed = scnprintf(bf, size, "%-*s ", max_ins_name, ins->name);
perf annotate: Remove duplicate 'name' field from disasm_line The disasm_line::name field is always equal to ins::name, being used just to locate the instruction's ins_ops from the per-arch instructions table. Eliminate this duplication, nuking that field and instead make ins__find() return an ins_ops, store it in disasm_line::ins.ops, and keep just in disasm_line::ins.name what was in disasm_line::name, this way we end up not keeping a reference to entries in the per-arch instructions table. This in turn will help supporting multiple ways to manage the per-arch instructions table, allowing resorting that array, for instance, when the entries will move after references to its addresses were made. The same problem is avoided when one grows the array with realloc. So architectures simply keeping a constant array will work as well as architectures building the table using regular expressions or other logic that involves resorting the table. Reviewed-by: Ravi Bangoria <ravi.bangoria@linux.vnet.ibm.com> Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Chris Riyder <chris.ryder@arm.com> Cc: David Ahern <dsahern@gmail.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Kim Phillips <kim.phillips@arm.com> Cc: Markus Trippelsdorf <markus@trippelsdorf.de> Cc: Masami Hiramatsu <mhiramat@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Naveen N. Rao <naveen.n.rao@linux.vnet.ibm.com> Cc: Pawel Moll <pawel.moll@arm.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Russell King <rmk+kernel@arm.linux.org.uk> Cc: Taeung Song <treeze.taeung@gmail.com> Cc: Wang Nan <wangnan0@huawei.com> Link: http://lkml.kernel.org/n/tip-vr899azvabnw9gtuepuqfd9t@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-11-24 17:16:06 +03:00
return printed + ins__scnprintf(&ops->locked.ins, bf + printed,
perf annotate: Calculate the max instruction name, align column to that We were hardcoding '6' as the max instruction name, and we have lots that are longer than that, see the diff from two 'P' printed TUI annotations for a libc function that uses instructions with long names, such as 'vpmovmskb' with its 9 chars: --- __strcmp_avx2.annotation.before 2019-03-06 16:31:39.368020425 -0300 +++ __strcmp_avx2.annotation 2019-03-06 16:32:12.079450508 -0300 @@ -2,284 +2,284 @@ Event: cycles:ppp Percent endbr64 - 0.10 mov %edi,%eax + 0.10 mov %edi,%eax - xor %edx,%edx + xor %edx,%edx - 3.54 vpxor %ymm7,%ymm7,%ymm7 + 3.54 vpxor %ymm7,%ymm7,%ymm7 - or %esi,%eax + or %esi,%eax - and $0xfff,%eax + and $0xfff,%eax - cmp $0xf80,%eax + cmp $0xf80,%eax - ↓ jg 370 + ↓ jg 370 - 27.07 vmovdqu (%rdi),%ymm1 + 27.07 vmovdqu (%rdi),%ymm1 - 7.97 vpcmpeqb (%rsi),%ymm1,%ymm0 + 7.97 vpcmpeqb (%rsi),%ymm1,%ymm0 - 2.15 vpminub %ymm1,%ymm0,%ymm0 + 2.15 vpminub %ymm1,%ymm0,%ymm0 - 4.09 vpcmpeqb %ymm7,%ymm0,%ymm0 + 4.09 vpcmpeqb %ymm7,%ymm0,%ymm0 - 0.43 vpmovmskb %ymm0,%ecx + 0.43 vpmovmskb %ymm0,%ecx - 1.53 test %ecx,%ecx + 1.53 test %ecx,%ecx - ↓ je b0 + ↓ je b0 - 5.26 tzcnt %ecx,%edx + 5.26 tzcnt %ecx,%edx - 18.40 movzbl (%rdi,%rdx,1),%eax + 18.40 movzbl (%rdi,%rdx,1),%eax - 7.09 movzbl (%rsi,%rdx,1),%edx + 7.09 movzbl (%rsi,%rdx,1),%edx - 3.34 sub %edx,%eax + 3.34 sub %edx,%eax 2.37 vzeroupper ← retq nop - 50: tzcnt %ecx,%edx + 50: tzcnt %ecx,%edx - movzbl 0x20(%rdi,%rdx,1),%eax + movzbl 0x20(%rdi,%rdx,1),%eax - movzbl 0x20(%rsi,%rdx,1),%edx + movzbl 0x20(%rsi,%rdx,1),%edx - sub %edx,%eax + sub %edx,%eax vzeroupper ← retq - data16 nopw %cs:0x0(%rax,%rax,1) + data16 nopw %cs:0x0(%rax,%rax,1) Reported-by: Travis Downs <travis.downs@gmail.com> LPU-Reference: CAOBGo4z1KfmWeOm6Et0cnX5Z6DWsG2PQbAvRn1MhVPJmXHrc5g@mail.gmail.com Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Link: https://lkml.kernel.org/n/tip-89wsdd9h9g6bvq52sgp6d0u4@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2019-03-06 22:40:15 +03:00
size - printed, ops->locked.ops, max_ins_name);
}
static void lock__delete(struct ins_operands *ops)
{
perf annotate: Remove duplicate 'name' field from disasm_line The disasm_line::name field is always equal to ins::name, being used just to locate the instruction's ins_ops from the per-arch instructions table. Eliminate this duplication, nuking that field and instead make ins__find() return an ins_ops, store it in disasm_line::ins.ops, and keep just in disasm_line::ins.name what was in disasm_line::name, this way we end up not keeping a reference to entries in the per-arch instructions table. This in turn will help supporting multiple ways to manage the per-arch instructions table, allowing resorting that array, for instance, when the entries will move after references to its addresses were made. The same problem is avoided when one grows the array with realloc. So architectures simply keeping a constant array will work as well as architectures building the table using regular expressions or other logic that involves resorting the table. Reviewed-by: Ravi Bangoria <ravi.bangoria@linux.vnet.ibm.com> Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Chris Riyder <chris.ryder@arm.com> Cc: David Ahern <dsahern@gmail.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Kim Phillips <kim.phillips@arm.com> Cc: Markus Trippelsdorf <markus@trippelsdorf.de> Cc: Masami Hiramatsu <mhiramat@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Naveen N. Rao <naveen.n.rao@linux.vnet.ibm.com> Cc: Pawel Moll <pawel.moll@arm.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Russell King <rmk+kernel@arm.linux.org.uk> Cc: Taeung Song <treeze.taeung@gmail.com> Cc: Wang Nan <wangnan0@huawei.com> Link: http://lkml.kernel.org/n/tip-vr899azvabnw9gtuepuqfd9t@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-11-24 17:16:06 +03:00
struct ins *ins = &ops->locked.ins;
perf annotate: Remove duplicate 'name' field from disasm_line The disasm_line::name field is always equal to ins::name, being used just to locate the instruction's ins_ops from the per-arch instructions table. Eliminate this duplication, nuking that field and instead make ins__find() return an ins_ops, store it in disasm_line::ins.ops, and keep just in disasm_line::ins.name what was in disasm_line::name, this way we end up not keeping a reference to entries in the per-arch instructions table. This in turn will help supporting multiple ways to manage the per-arch instructions table, allowing resorting that array, for instance, when the entries will move after references to its addresses were made. The same problem is avoided when one grows the array with realloc. So architectures simply keeping a constant array will work as well as architectures building the table using regular expressions or other logic that involves resorting the table. Reviewed-by: Ravi Bangoria <ravi.bangoria@linux.vnet.ibm.com> Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Chris Riyder <chris.ryder@arm.com> Cc: David Ahern <dsahern@gmail.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Kim Phillips <kim.phillips@arm.com> Cc: Markus Trippelsdorf <markus@trippelsdorf.de> Cc: Masami Hiramatsu <mhiramat@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Naveen N. Rao <naveen.n.rao@linux.vnet.ibm.com> Cc: Pawel Moll <pawel.moll@arm.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Russell King <rmk+kernel@arm.linux.org.uk> Cc: Taeung Song <treeze.taeung@gmail.com> Cc: Wang Nan <wangnan0@huawei.com> Link: http://lkml.kernel.org/n/tip-vr899azvabnw9gtuepuqfd9t@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-11-24 17:16:06 +03:00
if (ins->ops && ins->ops->free)
ins->ops->free(ops->locked.ops);
else
ins__delete(ops->locked.ops);
zfree(&ops->locked.ops);
zfree(&ops->target.raw);
zfree(&ops->target.name);
}
static struct ins_ops lock_ops = {
.free = lock__delete,
.parse = lock__parse,
.scnprintf = lock__scnprintf,
};
static int mov__parse(struct arch *arch, struct ins_operands *ops, struct map_symbol *ms __maybe_unused)
{
char *s = strchr(ops->raw, ','), *target, *comment, prev;
if (s == NULL)
return -1;
*s = '\0';
ops->source.raw = strdup(ops->raw);
*s = ',';
if (ops->source.raw == NULL)
return -1;
target = ++s;
comment = strchr(s, arch->objdump.comment_char);
if (comment != NULL)
s = comment - 1;
else
s = strchr(s, '\0') - 1;
while (s > target && isspace(s[0]))
--s;
s++;
prev = *s;
*s = '\0';
ops->target.raw = strdup(target);
*s = prev;
if (ops->target.raw == NULL)
goto out_free_source;
if (comment == NULL)
return 0;
comment = skip_spaces(comment);
perf annotate: Fix objdump comment parsing for Intel mov dissassembly The command 'perf annotate' parses the output of objdump and also investigates the comments produced by objdump. For example the output of objdump produces (on x86): 23eee: 4c 8b 3d 13 01 21 00 mov 0x210113(%rip),%r15 # 234008 <stderr@@GLIBC_2.2.5+0x9a8> and the function mov__parse() is called to investigate the complete line. Mov__parse() breaks this line into several parts and finally calls function comment__symbol() to parse the data after the comment character '#'. Comment__symbol() expects a hexadecimal address followed by a symbol in '<' and '>' brackets. However the 2nd parameter given to function comment__symbol() always points to the comment character '#'. The address parsing always returns 0 because the character '#' is not a digit and strtoull() fails without being noticed. Fix this by advancing the second parameter to function comment__symbol() by one byte before invocation and add an error check after strtoull() has been called. Signed-off-by: Thomas Richter <tmricht@linux.vnet.ibm.com> Reviewed-by: Hendrik Brueckner <brueckner@linux.vnet.ibm.com> Acked-by: Ravi Bangoria <ravi.bangoria@linux.vnet.ibm.com> Cc: Heiko Carstens <heiko.carstens@de.ibm.com> Cc: Martin Schwidefsky <schwidefsky@de.ibm.com> Fixes: 6de783b6f50f ("perf annotate: Resolve symbols using objdump comment") Link: http://lkml.kernel.org/r/20171128075632.72182-1-tmricht@linux.vnet.ibm.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2017-11-28 10:56:32 +03:00
comment__symbol(ops->source.raw, comment + 1, &ops->source.addr, &ops->source.name);
comment__symbol(ops->target.raw, comment + 1, &ops->target.addr, &ops->target.name);
return 0;
out_free_source:
zfree(&ops->source.raw);
return -1;
}
static int mov__scnprintf(struct ins *ins, char *bf, size_t size,
perf annotate: Calculate the max instruction name, align column to that We were hardcoding '6' as the max instruction name, and we have lots that are longer than that, see the diff from two 'P' printed TUI annotations for a libc function that uses instructions with long names, such as 'vpmovmskb' with its 9 chars: --- __strcmp_avx2.annotation.before 2019-03-06 16:31:39.368020425 -0300 +++ __strcmp_avx2.annotation 2019-03-06 16:32:12.079450508 -0300 @@ -2,284 +2,284 @@ Event: cycles:ppp Percent endbr64 - 0.10 mov %edi,%eax + 0.10 mov %edi,%eax - xor %edx,%edx + xor %edx,%edx - 3.54 vpxor %ymm7,%ymm7,%ymm7 + 3.54 vpxor %ymm7,%ymm7,%ymm7 - or %esi,%eax + or %esi,%eax - and $0xfff,%eax + and $0xfff,%eax - cmp $0xf80,%eax + cmp $0xf80,%eax - ↓ jg 370 + ↓ jg 370 - 27.07 vmovdqu (%rdi),%ymm1 + 27.07 vmovdqu (%rdi),%ymm1 - 7.97 vpcmpeqb (%rsi),%ymm1,%ymm0 + 7.97 vpcmpeqb (%rsi),%ymm1,%ymm0 - 2.15 vpminub %ymm1,%ymm0,%ymm0 + 2.15 vpminub %ymm1,%ymm0,%ymm0 - 4.09 vpcmpeqb %ymm7,%ymm0,%ymm0 + 4.09 vpcmpeqb %ymm7,%ymm0,%ymm0 - 0.43 vpmovmskb %ymm0,%ecx + 0.43 vpmovmskb %ymm0,%ecx - 1.53 test %ecx,%ecx + 1.53 test %ecx,%ecx - ↓ je b0 + ↓ je b0 - 5.26 tzcnt %ecx,%edx + 5.26 tzcnt %ecx,%edx - 18.40 movzbl (%rdi,%rdx,1),%eax + 18.40 movzbl (%rdi,%rdx,1),%eax - 7.09 movzbl (%rsi,%rdx,1),%edx + 7.09 movzbl (%rsi,%rdx,1),%edx - 3.34 sub %edx,%eax + 3.34 sub %edx,%eax 2.37 vzeroupper ← retq nop - 50: tzcnt %ecx,%edx + 50: tzcnt %ecx,%edx - movzbl 0x20(%rdi,%rdx,1),%eax + movzbl 0x20(%rdi,%rdx,1),%eax - movzbl 0x20(%rsi,%rdx,1),%edx + movzbl 0x20(%rsi,%rdx,1),%edx - sub %edx,%eax + sub %edx,%eax vzeroupper ← retq - data16 nopw %cs:0x0(%rax,%rax,1) + data16 nopw %cs:0x0(%rax,%rax,1) Reported-by: Travis Downs <travis.downs@gmail.com> LPU-Reference: CAOBGo4z1KfmWeOm6Et0cnX5Z6DWsG2PQbAvRn1MhVPJmXHrc5g@mail.gmail.com Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Link: https://lkml.kernel.org/n/tip-89wsdd9h9g6bvq52sgp6d0u4@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2019-03-06 22:40:15 +03:00
struct ins_operands *ops, int max_ins_name)
{
perf annotate: Calculate the max instruction name, align column to that We were hardcoding '6' as the max instruction name, and we have lots that are longer than that, see the diff from two 'P' printed TUI annotations for a libc function that uses instructions with long names, such as 'vpmovmskb' with its 9 chars: --- __strcmp_avx2.annotation.before 2019-03-06 16:31:39.368020425 -0300 +++ __strcmp_avx2.annotation 2019-03-06 16:32:12.079450508 -0300 @@ -2,284 +2,284 @@ Event: cycles:ppp Percent endbr64 - 0.10 mov %edi,%eax + 0.10 mov %edi,%eax - xor %edx,%edx + xor %edx,%edx - 3.54 vpxor %ymm7,%ymm7,%ymm7 + 3.54 vpxor %ymm7,%ymm7,%ymm7 - or %esi,%eax + or %esi,%eax - and $0xfff,%eax + and $0xfff,%eax - cmp $0xf80,%eax + cmp $0xf80,%eax - ↓ jg 370 + ↓ jg 370 - 27.07 vmovdqu (%rdi),%ymm1 + 27.07 vmovdqu (%rdi),%ymm1 - 7.97 vpcmpeqb (%rsi),%ymm1,%ymm0 + 7.97 vpcmpeqb (%rsi),%ymm1,%ymm0 - 2.15 vpminub %ymm1,%ymm0,%ymm0 + 2.15 vpminub %ymm1,%ymm0,%ymm0 - 4.09 vpcmpeqb %ymm7,%ymm0,%ymm0 + 4.09 vpcmpeqb %ymm7,%ymm0,%ymm0 - 0.43 vpmovmskb %ymm0,%ecx + 0.43 vpmovmskb %ymm0,%ecx - 1.53 test %ecx,%ecx + 1.53 test %ecx,%ecx - ↓ je b0 + ↓ je b0 - 5.26 tzcnt %ecx,%edx + 5.26 tzcnt %ecx,%edx - 18.40 movzbl (%rdi,%rdx,1),%eax + 18.40 movzbl (%rdi,%rdx,1),%eax - 7.09 movzbl (%rsi,%rdx,1),%edx + 7.09 movzbl (%rsi,%rdx,1),%edx - 3.34 sub %edx,%eax + 3.34 sub %edx,%eax 2.37 vzeroupper ← retq nop - 50: tzcnt %ecx,%edx + 50: tzcnt %ecx,%edx - movzbl 0x20(%rdi,%rdx,1),%eax + movzbl 0x20(%rdi,%rdx,1),%eax - movzbl 0x20(%rsi,%rdx,1),%edx + movzbl 0x20(%rsi,%rdx,1),%edx - sub %edx,%eax + sub %edx,%eax vzeroupper ← retq - data16 nopw %cs:0x0(%rax,%rax,1) + data16 nopw %cs:0x0(%rax,%rax,1) Reported-by: Travis Downs <travis.downs@gmail.com> LPU-Reference: CAOBGo4z1KfmWeOm6Et0cnX5Z6DWsG2PQbAvRn1MhVPJmXHrc5g@mail.gmail.com Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Link: https://lkml.kernel.org/n/tip-89wsdd9h9g6bvq52sgp6d0u4@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2019-03-06 22:40:15 +03:00
return scnprintf(bf, size, "%-*s %s,%s", max_ins_name, ins->name,
ops->source.name ?: ops->source.raw,
ops->target.name ?: ops->target.raw);
}
static struct ins_ops mov_ops = {
.parse = mov__parse,
.scnprintf = mov__scnprintf,
};
static int dec__parse(struct arch *arch __maybe_unused, struct ins_operands *ops, struct map_symbol *ms __maybe_unused)
{
char *target, *comment, *s, prev;
target = s = ops->raw;
while (s[0] != '\0' && !isspace(s[0]))
++s;
prev = *s;
*s = '\0';
ops->target.raw = strdup(target);
*s = prev;
if (ops->target.raw == NULL)
return -1;
comment = strchr(s, arch->objdump.comment_char);
if (comment == NULL)
return 0;
comment = skip_spaces(comment);
perf annotate: Fix objdump comment parsing for Intel mov dissassembly The command 'perf annotate' parses the output of objdump and also investigates the comments produced by objdump. For example the output of objdump produces (on x86): 23eee: 4c 8b 3d 13 01 21 00 mov 0x210113(%rip),%r15 # 234008 <stderr@@GLIBC_2.2.5+0x9a8> and the function mov__parse() is called to investigate the complete line. Mov__parse() breaks this line into several parts and finally calls function comment__symbol() to parse the data after the comment character '#'. Comment__symbol() expects a hexadecimal address followed by a symbol in '<' and '>' brackets. However the 2nd parameter given to function comment__symbol() always points to the comment character '#'. The address parsing always returns 0 because the character '#' is not a digit and strtoull() fails without being noticed. Fix this by advancing the second parameter to function comment__symbol() by one byte before invocation and add an error check after strtoull() has been called. Signed-off-by: Thomas Richter <tmricht@linux.vnet.ibm.com> Reviewed-by: Hendrik Brueckner <brueckner@linux.vnet.ibm.com> Acked-by: Ravi Bangoria <ravi.bangoria@linux.vnet.ibm.com> Cc: Heiko Carstens <heiko.carstens@de.ibm.com> Cc: Martin Schwidefsky <schwidefsky@de.ibm.com> Fixes: 6de783b6f50f ("perf annotate: Resolve symbols using objdump comment") Link: http://lkml.kernel.org/r/20171128075632.72182-1-tmricht@linux.vnet.ibm.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2017-11-28 10:56:32 +03:00
comment__symbol(ops->target.raw, comment + 1, &ops->target.addr, &ops->target.name);
return 0;
}
static int dec__scnprintf(struct ins *ins, char *bf, size_t size,
perf annotate: Calculate the max instruction name, align column to that We were hardcoding '6' as the max instruction name, and we have lots that are longer than that, see the diff from two 'P' printed TUI annotations for a libc function that uses instructions with long names, such as 'vpmovmskb' with its 9 chars: --- __strcmp_avx2.annotation.before 2019-03-06 16:31:39.368020425 -0300 +++ __strcmp_avx2.annotation 2019-03-06 16:32:12.079450508 -0300 @@ -2,284 +2,284 @@ Event: cycles:ppp Percent endbr64 - 0.10 mov %edi,%eax + 0.10 mov %edi,%eax - xor %edx,%edx + xor %edx,%edx - 3.54 vpxor %ymm7,%ymm7,%ymm7 + 3.54 vpxor %ymm7,%ymm7,%ymm7 - or %esi,%eax + or %esi,%eax - and $0xfff,%eax + and $0xfff,%eax - cmp $0xf80,%eax + cmp $0xf80,%eax - ↓ jg 370 + ↓ jg 370 - 27.07 vmovdqu (%rdi),%ymm1 + 27.07 vmovdqu (%rdi),%ymm1 - 7.97 vpcmpeqb (%rsi),%ymm1,%ymm0 + 7.97 vpcmpeqb (%rsi),%ymm1,%ymm0 - 2.15 vpminub %ymm1,%ymm0,%ymm0 + 2.15 vpminub %ymm1,%ymm0,%ymm0 - 4.09 vpcmpeqb %ymm7,%ymm0,%ymm0 + 4.09 vpcmpeqb %ymm7,%ymm0,%ymm0 - 0.43 vpmovmskb %ymm0,%ecx + 0.43 vpmovmskb %ymm0,%ecx - 1.53 test %ecx,%ecx + 1.53 test %ecx,%ecx - ↓ je b0 + ↓ je b0 - 5.26 tzcnt %ecx,%edx + 5.26 tzcnt %ecx,%edx - 18.40 movzbl (%rdi,%rdx,1),%eax + 18.40 movzbl (%rdi,%rdx,1),%eax - 7.09 movzbl (%rsi,%rdx,1),%edx + 7.09 movzbl (%rsi,%rdx,1),%edx - 3.34 sub %edx,%eax + 3.34 sub %edx,%eax 2.37 vzeroupper ← retq nop - 50: tzcnt %ecx,%edx + 50: tzcnt %ecx,%edx - movzbl 0x20(%rdi,%rdx,1),%eax + movzbl 0x20(%rdi,%rdx,1),%eax - movzbl 0x20(%rsi,%rdx,1),%edx + movzbl 0x20(%rsi,%rdx,1),%edx - sub %edx,%eax + sub %edx,%eax vzeroupper ← retq - data16 nopw %cs:0x0(%rax,%rax,1) + data16 nopw %cs:0x0(%rax,%rax,1) Reported-by: Travis Downs <travis.downs@gmail.com> LPU-Reference: CAOBGo4z1KfmWeOm6Et0cnX5Z6DWsG2PQbAvRn1MhVPJmXHrc5g@mail.gmail.com Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Link: https://lkml.kernel.org/n/tip-89wsdd9h9g6bvq52sgp6d0u4@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2019-03-06 22:40:15 +03:00
struct ins_operands *ops, int max_ins_name)
{
perf annotate: Calculate the max instruction name, align column to that We were hardcoding '6' as the max instruction name, and we have lots that are longer than that, see the diff from two 'P' printed TUI annotations for a libc function that uses instructions with long names, such as 'vpmovmskb' with its 9 chars: --- __strcmp_avx2.annotation.before 2019-03-06 16:31:39.368020425 -0300 +++ __strcmp_avx2.annotation 2019-03-06 16:32:12.079450508 -0300 @@ -2,284 +2,284 @@ Event: cycles:ppp Percent endbr64 - 0.10 mov %edi,%eax + 0.10 mov %edi,%eax - xor %edx,%edx + xor %edx,%edx - 3.54 vpxor %ymm7,%ymm7,%ymm7 + 3.54 vpxor %ymm7,%ymm7,%ymm7 - or %esi,%eax + or %esi,%eax - and $0xfff,%eax + and $0xfff,%eax - cmp $0xf80,%eax + cmp $0xf80,%eax - ↓ jg 370 + ↓ jg 370 - 27.07 vmovdqu (%rdi),%ymm1 + 27.07 vmovdqu (%rdi),%ymm1 - 7.97 vpcmpeqb (%rsi),%ymm1,%ymm0 + 7.97 vpcmpeqb (%rsi),%ymm1,%ymm0 - 2.15 vpminub %ymm1,%ymm0,%ymm0 + 2.15 vpminub %ymm1,%ymm0,%ymm0 - 4.09 vpcmpeqb %ymm7,%ymm0,%ymm0 + 4.09 vpcmpeqb %ymm7,%ymm0,%ymm0 - 0.43 vpmovmskb %ymm0,%ecx + 0.43 vpmovmskb %ymm0,%ecx - 1.53 test %ecx,%ecx + 1.53 test %ecx,%ecx - ↓ je b0 + ↓ je b0 - 5.26 tzcnt %ecx,%edx + 5.26 tzcnt %ecx,%edx - 18.40 movzbl (%rdi,%rdx,1),%eax + 18.40 movzbl (%rdi,%rdx,1),%eax - 7.09 movzbl (%rsi,%rdx,1),%edx + 7.09 movzbl (%rsi,%rdx,1),%edx - 3.34 sub %edx,%eax + 3.34 sub %edx,%eax 2.37 vzeroupper ← retq nop - 50: tzcnt %ecx,%edx + 50: tzcnt %ecx,%edx - movzbl 0x20(%rdi,%rdx,1),%eax + movzbl 0x20(%rdi,%rdx,1),%eax - movzbl 0x20(%rsi,%rdx,1),%edx + movzbl 0x20(%rsi,%rdx,1),%edx - sub %edx,%eax + sub %edx,%eax vzeroupper ← retq - data16 nopw %cs:0x0(%rax,%rax,1) + data16 nopw %cs:0x0(%rax,%rax,1) Reported-by: Travis Downs <travis.downs@gmail.com> LPU-Reference: CAOBGo4z1KfmWeOm6Et0cnX5Z6DWsG2PQbAvRn1MhVPJmXHrc5g@mail.gmail.com Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Link: https://lkml.kernel.org/n/tip-89wsdd9h9g6bvq52sgp6d0u4@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2019-03-06 22:40:15 +03:00
return scnprintf(bf, size, "%-*s %s", max_ins_name, ins->name,
ops->target.name ?: ops->target.raw);
}
static struct ins_ops dec_ops = {
.parse = dec__parse,
.scnprintf = dec__scnprintf,
};
perf tools: Use __maybe_used for unused variables perf defines both __used and __unused variables to use for marking unused variables. The variable __used is defined to __attribute__((__unused__)), which contradicts the kernel definition to __attribute__((__used__)) for new gcc versions. On Android, __used is also defined in system headers and this leads to warnings like: warning: '__used__' attribute ignored __unused is not defined in the kernel and is not a standard definition. If __unused is included everywhere instead of __used, this leads to conflicts with glibc headers, since glibc has a variables with this name in its headers. The best approach is to use __maybe_unused, the definition used in the kernel for __attribute__((unused)). In this way there is only one definition in perf sources (instead of 2 definitions that point to the same thing: __used and __unused) and it works on both Linux and Android. This patch simply replaces all instances of __used and __unused with __maybe_unused. Signed-off-by: Irina Tirdea <irina.tirdea@intel.com> Acked-by: Pekka Enberg <penberg@kernel.org> Cc: David Ahern <dsahern@gmail.com> Cc: Ingo Molnar <mingo@redhat.com> Cc: Namhyung Kim <namhyung.kim@lge.com> Cc: Paul Mackerras <paulus@samba.org> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Steven Rostedt <rostedt@goodmis.org> Link: http://lkml.kernel.org/r/1347315303-29906-7-git-send-email-irina.tirdea@intel.com [ committer note: fixed up conflict with a116e05 in builtin-sched.c ] Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2012-09-11 02:15:03 +04:00
static int nop__scnprintf(struct ins *ins __maybe_unused, char *bf, size_t size,
perf annotate: Calculate the max instruction name, align column to that We were hardcoding '6' as the max instruction name, and we have lots that are longer than that, see the diff from two 'P' printed TUI annotations for a libc function that uses instructions with long names, such as 'vpmovmskb' with its 9 chars: --- __strcmp_avx2.annotation.before 2019-03-06 16:31:39.368020425 -0300 +++ __strcmp_avx2.annotation 2019-03-06 16:32:12.079450508 -0300 @@ -2,284 +2,284 @@ Event: cycles:ppp Percent endbr64 - 0.10 mov %edi,%eax + 0.10 mov %edi,%eax - xor %edx,%edx + xor %edx,%edx - 3.54 vpxor %ymm7,%ymm7,%ymm7 + 3.54 vpxor %ymm7,%ymm7,%ymm7 - or %esi,%eax + or %esi,%eax - and $0xfff,%eax + and $0xfff,%eax - cmp $0xf80,%eax + cmp $0xf80,%eax - ↓ jg 370 + ↓ jg 370 - 27.07 vmovdqu (%rdi),%ymm1 + 27.07 vmovdqu (%rdi),%ymm1 - 7.97 vpcmpeqb (%rsi),%ymm1,%ymm0 + 7.97 vpcmpeqb (%rsi),%ymm1,%ymm0 - 2.15 vpminub %ymm1,%ymm0,%ymm0 + 2.15 vpminub %ymm1,%ymm0,%ymm0 - 4.09 vpcmpeqb %ymm7,%ymm0,%ymm0 + 4.09 vpcmpeqb %ymm7,%ymm0,%ymm0 - 0.43 vpmovmskb %ymm0,%ecx + 0.43 vpmovmskb %ymm0,%ecx - 1.53 test %ecx,%ecx + 1.53 test %ecx,%ecx - ↓ je b0 + ↓ je b0 - 5.26 tzcnt %ecx,%edx + 5.26 tzcnt %ecx,%edx - 18.40 movzbl (%rdi,%rdx,1),%eax + 18.40 movzbl (%rdi,%rdx,1),%eax - 7.09 movzbl (%rsi,%rdx,1),%edx + 7.09 movzbl (%rsi,%rdx,1),%edx - 3.34 sub %edx,%eax + 3.34 sub %edx,%eax 2.37 vzeroupper ← retq nop - 50: tzcnt %ecx,%edx + 50: tzcnt %ecx,%edx - movzbl 0x20(%rdi,%rdx,1),%eax + movzbl 0x20(%rdi,%rdx,1),%eax - movzbl 0x20(%rsi,%rdx,1),%edx + movzbl 0x20(%rsi,%rdx,1),%edx - sub %edx,%eax + sub %edx,%eax vzeroupper ← retq - data16 nopw %cs:0x0(%rax,%rax,1) + data16 nopw %cs:0x0(%rax,%rax,1) Reported-by: Travis Downs <travis.downs@gmail.com> LPU-Reference: CAOBGo4z1KfmWeOm6Et0cnX5Z6DWsG2PQbAvRn1MhVPJmXHrc5g@mail.gmail.com Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Link: https://lkml.kernel.org/n/tip-89wsdd9h9g6bvq52sgp6d0u4@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2019-03-06 22:40:15 +03:00
struct ins_operands *ops __maybe_unused, int max_ins_name)
{
perf annotate: Calculate the max instruction name, align column to that We were hardcoding '6' as the max instruction name, and we have lots that are longer than that, see the diff from two 'P' printed TUI annotations for a libc function that uses instructions with long names, such as 'vpmovmskb' with its 9 chars: --- __strcmp_avx2.annotation.before 2019-03-06 16:31:39.368020425 -0300 +++ __strcmp_avx2.annotation 2019-03-06 16:32:12.079450508 -0300 @@ -2,284 +2,284 @@ Event: cycles:ppp Percent endbr64 - 0.10 mov %edi,%eax + 0.10 mov %edi,%eax - xor %edx,%edx + xor %edx,%edx - 3.54 vpxor %ymm7,%ymm7,%ymm7 + 3.54 vpxor %ymm7,%ymm7,%ymm7 - or %esi,%eax + or %esi,%eax - and $0xfff,%eax + and $0xfff,%eax - cmp $0xf80,%eax + cmp $0xf80,%eax - ↓ jg 370 + ↓ jg 370 - 27.07 vmovdqu (%rdi),%ymm1 + 27.07 vmovdqu (%rdi),%ymm1 - 7.97 vpcmpeqb (%rsi),%ymm1,%ymm0 + 7.97 vpcmpeqb (%rsi),%ymm1,%ymm0 - 2.15 vpminub %ymm1,%ymm0,%ymm0 + 2.15 vpminub %ymm1,%ymm0,%ymm0 - 4.09 vpcmpeqb %ymm7,%ymm0,%ymm0 + 4.09 vpcmpeqb %ymm7,%ymm0,%ymm0 - 0.43 vpmovmskb %ymm0,%ecx + 0.43 vpmovmskb %ymm0,%ecx - 1.53 test %ecx,%ecx + 1.53 test %ecx,%ecx - ↓ je b0 + ↓ je b0 - 5.26 tzcnt %ecx,%edx + 5.26 tzcnt %ecx,%edx - 18.40 movzbl (%rdi,%rdx,1),%eax + 18.40 movzbl (%rdi,%rdx,1),%eax - 7.09 movzbl (%rsi,%rdx,1),%edx + 7.09 movzbl (%rsi,%rdx,1),%edx - 3.34 sub %edx,%eax + 3.34 sub %edx,%eax 2.37 vzeroupper ← retq nop - 50: tzcnt %ecx,%edx + 50: tzcnt %ecx,%edx - movzbl 0x20(%rdi,%rdx,1),%eax + movzbl 0x20(%rdi,%rdx,1),%eax - movzbl 0x20(%rsi,%rdx,1),%edx + movzbl 0x20(%rsi,%rdx,1),%edx - sub %edx,%eax + sub %edx,%eax vzeroupper ← retq - data16 nopw %cs:0x0(%rax,%rax,1) + data16 nopw %cs:0x0(%rax,%rax,1) Reported-by: Travis Downs <travis.downs@gmail.com> LPU-Reference: CAOBGo4z1KfmWeOm6Et0cnX5Z6DWsG2PQbAvRn1MhVPJmXHrc5g@mail.gmail.com Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Link: https://lkml.kernel.org/n/tip-89wsdd9h9g6bvq52sgp6d0u4@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2019-03-06 22:40:15 +03:00
return scnprintf(bf, size, "%-*s", max_ins_name, "nop");
}
static struct ins_ops nop_ops = {
.scnprintf = nop__scnprintf,
};
static struct ins_ops ret_ops = {
.scnprintf = ins__raw_scnprintf,
};
bool ins__is_ret(const struct ins *ins)
{
return ins->ops == &ret_ops;
}
bool ins__is_lock(const struct ins *ins)
{
return ins->ops == &lock_ops;
}
static int ins__key_cmp(const void *name, const void *insp)
{
const struct ins *ins = insp;
return strcmp(name, ins->name);
}
static int ins__cmp(const void *a, const void *b)
{
const struct ins *ia = a;
const struct ins *ib = b;
return strcmp(ia->name, ib->name);
}
static void ins__sort(struct arch *arch)
{
const int nmemb = arch->nr_instructions;
qsort(arch->instructions, nmemb, sizeof(struct ins), ins__cmp);
}
perf annotate: Introduce alternative method of keeping instructions table Some arches may want to dynamically populate the table using regular expressions on the instruction names to associate them with a set of parsing/formatting/etc functions (struct ins_ops), so provide a fallback for when the ins__find() method fails. That fall back will be able to resize the arch->instructions, setting arch->nr_instructions appropriately, helper functions to associate an ins_ops to an instruction name, growing the arch->instructions if needed and resorting it are provided, all the arch specific callback needs to do is to decide if the missing instruction should be added to arch->instructions with a ins_ops association. Reviewed-by: Ravi Bangoria <ravi.bangoria@linux.vnet.ibm.com> Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Chris Riyder <chris.ryder@arm.com> Cc: David Ahern <dsahern@gmail.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Kim Phillips <kim.phillips@arm.com> Cc: Markus Trippelsdorf <markus@trippelsdorf.de> Cc: Masami Hiramatsu <mhiramat@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Naveen N. Rao <naveen.n.rao@linux.vnet.ibm.com> Cc: Pawel Moll <pawel.moll@arm.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Russell King <rmk+kernel@arm.linux.org.uk> Cc: Taeung Song <treeze.taeung@gmail.com> Cc: Wang Nan <wangnan0@huawei.com> Link: http://lkml.kernel.org/n/tip-auu13yradxf7g5dgtpnzt97a@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-11-24 17:37:08 +03:00
static struct ins_ops *__ins__find(struct arch *arch, const char *name)
{
perf annotate: Remove duplicate 'name' field from disasm_line The disasm_line::name field is always equal to ins::name, being used just to locate the instruction's ins_ops from the per-arch instructions table. Eliminate this duplication, nuking that field and instead make ins__find() return an ins_ops, store it in disasm_line::ins.ops, and keep just in disasm_line::ins.name what was in disasm_line::name, this way we end up not keeping a reference to entries in the per-arch instructions table. This in turn will help supporting multiple ways to manage the per-arch instructions table, allowing resorting that array, for instance, when the entries will move after references to its addresses were made. The same problem is avoided when one grows the array with realloc. So architectures simply keeping a constant array will work as well as architectures building the table using regular expressions or other logic that involves resorting the table. Reviewed-by: Ravi Bangoria <ravi.bangoria@linux.vnet.ibm.com> Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Chris Riyder <chris.ryder@arm.com> Cc: David Ahern <dsahern@gmail.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Kim Phillips <kim.phillips@arm.com> Cc: Markus Trippelsdorf <markus@trippelsdorf.de> Cc: Masami Hiramatsu <mhiramat@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Naveen N. Rao <naveen.n.rao@linux.vnet.ibm.com> Cc: Pawel Moll <pawel.moll@arm.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Russell King <rmk+kernel@arm.linux.org.uk> Cc: Taeung Song <treeze.taeung@gmail.com> Cc: Wang Nan <wangnan0@huawei.com> Link: http://lkml.kernel.org/n/tip-vr899azvabnw9gtuepuqfd9t@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-11-24 17:16:06 +03:00
struct ins *ins;
const int nmemb = arch->nr_instructions;
if (!arch->sorted_instructions) {
ins__sort(arch);
arch->sorted_instructions = true;
}
perf annotate: Remove duplicate 'name' field from disasm_line The disasm_line::name field is always equal to ins::name, being used just to locate the instruction's ins_ops from the per-arch instructions table. Eliminate this duplication, nuking that field and instead make ins__find() return an ins_ops, store it in disasm_line::ins.ops, and keep just in disasm_line::ins.name what was in disasm_line::name, this way we end up not keeping a reference to entries in the per-arch instructions table. This in turn will help supporting multiple ways to manage the per-arch instructions table, allowing resorting that array, for instance, when the entries will move after references to its addresses were made. The same problem is avoided when one grows the array with realloc. So architectures simply keeping a constant array will work as well as architectures building the table using regular expressions or other logic that involves resorting the table. Reviewed-by: Ravi Bangoria <ravi.bangoria@linux.vnet.ibm.com> Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Chris Riyder <chris.ryder@arm.com> Cc: David Ahern <dsahern@gmail.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Kim Phillips <kim.phillips@arm.com> Cc: Markus Trippelsdorf <markus@trippelsdorf.de> Cc: Masami Hiramatsu <mhiramat@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Naveen N. Rao <naveen.n.rao@linux.vnet.ibm.com> Cc: Pawel Moll <pawel.moll@arm.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Russell King <rmk+kernel@arm.linux.org.uk> Cc: Taeung Song <treeze.taeung@gmail.com> Cc: Wang Nan <wangnan0@huawei.com> Link: http://lkml.kernel.org/n/tip-vr899azvabnw9gtuepuqfd9t@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-11-24 17:16:06 +03:00
ins = bsearch(name, arch->instructions, nmemb, sizeof(struct ins), ins__key_cmp);
return ins ? ins->ops : NULL;
}
perf annotate: Introduce alternative method of keeping instructions table Some arches may want to dynamically populate the table using regular expressions on the instruction names to associate them with a set of parsing/formatting/etc functions (struct ins_ops), so provide a fallback for when the ins__find() method fails. That fall back will be able to resize the arch->instructions, setting arch->nr_instructions appropriately, helper functions to associate an ins_ops to an instruction name, growing the arch->instructions if needed and resorting it are provided, all the arch specific callback needs to do is to decide if the missing instruction should be added to arch->instructions with a ins_ops association. Reviewed-by: Ravi Bangoria <ravi.bangoria@linux.vnet.ibm.com> Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Chris Riyder <chris.ryder@arm.com> Cc: David Ahern <dsahern@gmail.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Kim Phillips <kim.phillips@arm.com> Cc: Markus Trippelsdorf <markus@trippelsdorf.de> Cc: Masami Hiramatsu <mhiramat@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Naveen N. Rao <naveen.n.rao@linux.vnet.ibm.com> Cc: Pawel Moll <pawel.moll@arm.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Russell King <rmk+kernel@arm.linux.org.uk> Cc: Taeung Song <treeze.taeung@gmail.com> Cc: Wang Nan <wangnan0@huawei.com> Link: http://lkml.kernel.org/n/tip-auu13yradxf7g5dgtpnzt97a@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-11-24 17:37:08 +03:00
static struct ins_ops *ins__find(struct arch *arch, const char *name)
{
struct ins_ops *ops = __ins__find(arch, name);
if (!ops && arch->associate_instruction_ops)
ops = arch->associate_instruction_ops(arch, name);
return ops;
}
static int arch__key_cmp(const void *name, const void *archp)
{
const struct arch *arch = archp;
return strcmp(name, arch->name);
}
static int arch__cmp(const void *a, const void *b)
{
const struct arch *aa = a;
const struct arch *ab = b;
return strcmp(aa->name, ab->name);
}
static void arch__sort(void)
{
const int nmemb = ARRAY_SIZE(architectures);
qsort(architectures, nmemb, sizeof(struct arch), arch__cmp);
}
static struct arch *arch__find(const char *name)
{
const int nmemb = ARRAY_SIZE(architectures);
static bool sorted;
if (!sorted) {
arch__sort();
sorted = true;
}
return bsearch(name, architectures, nmemb, sizeof(struct arch), arch__key_cmp);
}
static struct annotated_source *annotated_source__new(void)
{
struct annotated_source *src = zalloc(sizeof(*src));
if (src != NULL)
INIT_LIST_HEAD(&src->source);
return src;
}
static __maybe_unused void annotated_source__delete(struct annotated_source *src)
{
if (src == NULL)
return;
zfree(&src->histograms);
zfree(&src->cycles_hist);
free(src);
}
static int annotated_source__alloc_histograms(struct annotated_source *src,
size_t size, int nr_hists)
{
size_t sizeof_sym_hist;
perf symbols: Fix memory corruption because of zero length symbols Perf top is often crashing at very random locations on powerpc. After investigating, I found the crash only happens when sample is of zero length symbol. Powerpc kernel has many such symbols which does not contain length details in vmlinux binary and thus start and end addresses of such symbols are same. Structure struct sym_hist { u64 nr_samples; u64 period; struct sym_hist_entry addr[0]; }; has last member 'addr[]' of size zero. 'addr[]' is an array of addresses that belongs to one symbol (function). If function consist of 100 instructions, 'addr' points to an array of 100 'struct sym_hist_entry' elements. For zero length symbol, it points to the *empty* array, i.e. no members in the array and thus offset 0 is also invalid for such array. static int __symbol__inc_addr_samples(...) { ... offset = addr - sym->start; h = annotation__histogram(notes, evidx); h->nr_samples++; h->addr[offset].nr_samples++; h->period += sample->period; h->addr[offset].period += sample->period; ... } Here, when 'addr' is same as 'sym->start', 'offset' becomes 0, which is valid for normal symbols but *invalid* for zero length symbols and thus updating h->addr[offset] causes memory corruption. Fix this by adding one dummy element for zero length symbols. Link: https://lkml.org/lkml/2016/10/10/148 Fixes: edee44be5919 ("perf annotate: Don't throw error for zero length symbols") Signed-off-by: Ravi Bangoria <ravi.bangoria@linux.vnet.ibm.com> Acked-by: Jiri Olsa <jolsa@kernel.org> Acked-by: Namhyung Kim <namhyung@kernel.org> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Jin Yao <yao.jin@linux.intel.com> Cc: Kim Phillips <kim.phillips@arm.com> Cc: Naveen N. Rao <naveen.n.rao@linux.vnet.ibm.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Taeung Song <treeze.taeung@gmail.com> Link: http://lkml.kernel.org/r/1508854806-10542-1-git-send-email-ravi.bangoria@linux.vnet.ibm.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2017-10-24 17:20:06 +03:00
/*
* Add buffer of one element for zero length symbol.
* When sample is taken from first instruction of
* zero length symbol, perf still resolves it and
* shows symbol name in perf report and allows to
* annotate it.
*/
if (size == 0)
size = 1;
/* Check for overflow when calculating sizeof_sym_hist */
if (size > (SIZE_MAX - sizeof(struct sym_hist)) / sizeof(struct sym_hist_entry))
return -1;
sizeof_sym_hist = (sizeof(struct sym_hist) + size * sizeof(struct sym_hist_entry));
/* Check for overflow in zalloc argument */
if (sizeof_sym_hist > SIZE_MAX / nr_hists)
return -1;
src->sizeof_sym_hist = sizeof_sym_hist;
src->nr_histograms = nr_hists;
src->histograms = calloc(nr_hists, sizeof_sym_hist) ;
return src->histograms ? 0 : -1;
}
/* The cycles histogram is lazily allocated. */
static int symbol__alloc_hist_cycles(struct symbol *sym)
{
struct annotation *notes = symbol__annotation(sym);
const size_t size = symbol__size(sym);
notes->src->cycles_hist = calloc(size, sizeof(struct cyc_hist));
if (notes->src->cycles_hist == NULL)
return -1;
return 0;
}
void symbol__annotate_zero_histograms(struct symbol *sym)
{
struct annotation *notes = symbol__annotation(sym);
pthread_mutex_lock(&notes->lock);
if (notes->src != NULL) {
memset(notes->src->histograms, 0,
notes->src->nr_histograms * notes->src->sizeof_sym_hist);
if (notes->src->cycles_hist)
memset(notes->src->cycles_hist, 0,
symbol__size(sym) * sizeof(struct cyc_hist));
}
pthread_mutex_unlock(&notes->lock);
}
static int __symbol__account_cycles(struct cyc_hist *ch,
u64 start,
unsigned offset, unsigned cycles,
unsigned have_start)
{
/*
* For now we can only account one basic block per
* final jump. But multiple could be overlapping.
* Always account the longest one. So when
* a shorter one has been already seen throw it away.
*
* We separately always account the full cycles.
*/
ch[offset].num_aggr++;
ch[offset].cycles_aggr += cycles;
if (cycles > ch[offset].cycles_max)
ch[offset].cycles_max = cycles;
if (ch[offset].cycles_min) {
if (cycles && cycles < ch[offset].cycles_min)
ch[offset].cycles_min = cycles;
} else
ch[offset].cycles_min = cycles;
if (!have_start && ch[offset].have_start)
return 0;
if (ch[offset].num) {
if (have_start && (!ch[offset].have_start ||
ch[offset].start > start)) {
ch[offset].have_start = 0;
ch[offset].cycles = 0;
ch[offset].num = 0;
if (ch[offset].reset < 0xffff)
ch[offset].reset++;
} else if (have_start &&
ch[offset].start < start)
return 0;
}
perf diff: Report noisy for cycles diff This patch prints the stddev and hist for the cycles diff of program block. It can help us to understand if the cycles is noisy or not. This patch is inspired by Andi Kleen's patch: https://lwn.net/Articles/600471/ We create new option '--cycles-hist'. Example: perf record -b ./div perf record -b ./div perf diff -c cycles # Baseline [Program Block Range] Cycles Diff Shared Object Symbol # ........ .......................................................... .... ................. ............................ # 46.72% [div.c:40 -> div.c:40] 0 div [.] main 46.72% [div.c:42 -> div.c:44] 0 div [.] main 46.72% [div.c:42 -> div.c:39] 0 div [.] main 20.54% [random_r.c:357 -> random_r.c:394] 1 libc-2.27.so [.] __random_r 20.54% [random_r.c:357 -> random_r.c:380] 0 libc-2.27.so [.] __random_r 20.54% [random_r.c:388 -> random_r.c:388] 0 libc-2.27.so [.] __random_r 20.54% [random_r.c:388 -> random_r.c:391] 0 libc-2.27.so [.] __random_r 17.04% [random.c:288 -> random.c:291] 0 libc-2.27.so [.] __random 17.04% [random.c:291 -> random.c:291] 0 libc-2.27.so [.] __random 17.04% [random.c:293 -> random.c:293] 0 libc-2.27.so [.] __random 17.04% [random.c:295 -> random.c:295] 0 libc-2.27.so [.] __random 17.04% [random.c:295 -> random.c:295] 0 libc-2.27.so [.] __random 17.04% [random.c:298 -> random.c:298] 0 libc-2.27.so [.] __random 8.40% [div.c:22 -> div.c:25] 0 div [.] compute_flag 8.40% [div.c:27 -> div.c:28] 0 div [.] compute_flag 5.14% [rand.c:26 -> rand.c:27] 0 libc-2.27.so [.] rand 5.14% [rand.c:28 -> rand.c:28] 0 libc-2.27.so [.] rand 2.15% [rand@plt+0 -> rand@plt+0] 0 div [.] rand@plt 0.00% [kernel.kallsyms] [k] __x86_indirect_thunk_rax 0.00% [do_mmap+714 -> do_mmap+732] -10 [kernel.kallsyms] [k] do_mmap 0.00% [do_mmap+737 -> do_mmap+765] 1 [kernel.kallsyms] [k] do_mmap 0.00% [do_mmap+262 -> do_mmap+299] 0 [kernel.kallsyms] [k] do_mmap 0.00% [__x86_indirect_thunk_r15+0 -> __x86_indirect_thunk_r15+0] 7 [kernel.kallsyms] [k] __x86_indirect_thunk_r15 0.00% [native_sched_clock+0 -> native_sched_clock+119] -1 [kernel.kallsyms] [k] native_sched_clock 0.00% [native_write_msr+0 -> native_write_msr+16] -13 [kernel.kallsyms] [k] native_write_msr When we enable the option '--cycles-hist', the output is perf diff -c cycles --cycles-hist # Baseline [Program Block Range] Cycles Diff stddev/Hist Shared Object Symbol # ........ .......................................................... .... ................. ................. ............................ # 46.72% [div.c:40 -> div.c:40] 0 ± 37.8% ▁█▁▁██▁█ div [.] main 46.72% [div.c:42 -> div.c:44] 0 ± 49.4% ▁▁▂█▂▂▂▂ div [.] main 46.72% [div.c:42 -> div.c:39] 0 ± 24.1% ▃█▂▄▁▃▂▁ div [.] main 20.54% [random_r.c:357 -> random_r.c:394] 1 ± 33.5% ▅▂▁█▃▁▂▁ libc-2.27.so [.] __random_r 20.54% [random_r.c:357 -> random_r.c:380] 0 ± 39.4% ▁▁█▁██▅▁ libc-2.27.so [.] __random_r 20.54% [random_r.c:388 -> random_r.c:388] 0 libc-2.27.so [.] __random_r 20.54% [random_r.c:388 -> random_r.c:391] 0 ± 41.2% ▁▃▁▂█▄▃▁ libc-2.27.so [.] __random_r 17.04% [random.c:288 -> random.c:291] 0 ± 48.8% ▁▁▁▁███▁ libc-2.27.so [.] __random 17.04% [random.c:291 -> random.c:291] 0 ±100.0% ▁█▁▁▁▁▁▁ libc-2.27.so [.] __random 17.04% [random.c:293 -> random.c:293] 0 ±100.0% ▁█▁▁▁▁▁▁ libc-2.27.so [.] __random 17.04% [random.c:295 -> random.c:295] 0 ±100.0% ▁█▁▁▁▁▁▁ libc-2.27.so [.] __random 17.04% [random.c:295 -> random.c:295] 0 libc-2.27.so [.] __random 17.04% [random.c:298 -> random.c:298] 0 ± 75.6% ▃█▁▁▁▁▁▁ libc-2.27.so [.] __random 8.40% [div.c:22 -> div.c:25] 0 ± 42.1% ▁▃▁▁███▁ div [.] compute_flag 8.40% [div.c:27 -> div.c:28] 0 ± 41.8% ██▁▁▄▁▁▄ div [.] compute_flag 5.14% [rand.c:26 -> rand.c:27] 0 ± 37.8% ▁▁▁████▁ libc-2.27.so [.] rand 5.14% [rand.c:28 -> rand.c:28] 0 libc-2.27.so [.] rand 2.15% [rand@plt+0 -> rand@plt+0] 0 div [.] rand@plt 0.00% [kernel.kallsyms] [k] __x86_indirect_thunk_rax 0.00% [do_mmap+714 -> do_mmap+732] -10 [kernel.kallsyms] [k] do_mmap 0.00% [do_mmap+737 -> do_mmap+765] 1 [kernel.kallsyms] [k] do_mmap 0.00% [do_mmap+262 -> do_mmap+299] 0 [kernel.kallsyms] [k] do_mmap 0.00% [__x86_indirect_thunk_r15+0 -> __x86_indirect_thunk_r15+0] 7 [kernel.kallsyms] [k] __x86_indirect_thunk_r15 0.00% [native_sched_clock+0 -> native_sched_clock+119] -1 ± 38.5% ▄█▁ [kernel.kallsyms] [k] native_sched_clock 0.00% [native_write_msr+0 -> native_write_msr+16] -13 ± 47.1% ▁█▇▃▁▁ [kernel.kallsyms] [k] native_write_msr v8: --- Rebase to perf/core branch v7: --- 1. v6 got Jiri's ACK. 2. Rebase to latest perf/core branch. v6: --- 1. Jiri provides better code for using data__hpp_register() in ui_init(). Use this code in v6. v5: --- 1. Refine the use of data__hpp_register() in ui_init() according to Jiri's suggestion. v4: --- 1. Rename the new option from '--noisy' to '--cycles-hist' 2. Remove the option '-n'. 3. Only update the spark value and stats when '--cycles-hist' is enabled. 4. Remove the code of printing '..'. v3: --- 1. Move the histogram to a separate column 2. Move the svals[] out of struct stats v2: --- Jiri got a compile error, CC builtin-diff.o builtin-diff.c: In function ‘compute_cycles_diff’: builtin-diff.c:712:10: error: taking the absolute value of unsigned type ‘u64’ {aka ‘long unsigned int’} has no effect [-Werror=absolute-value] 712 | labs(pair->block_info->cycles_spark[i] - | ^~~~ Because the result of u64 - u64 is still u64. Now we change the type of cycles_spark[] to s64. Signed-off-by: Jin Yao <yao.jin@linux.intel.com> Acked-by: Jiri Olsa <jolsa@kernel.org> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Andi Kleen <ak@linux.intel.com> Cc: Kan Liang <kan.liang@linux.intel.com> Cc: Peter Zijlstra <peterz@infradead.org> Link: http://lore.kernel.org/lkml/20190925011446.30678-1-yao.jin@linux.intel.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2019-09-25 04:14:46 +03:00
if (ch[offset].num < NUM_SPARKS)
ch[offset].cycles_spark[ch[offset].num] = cycles;
ch[offset].have_start = have_start;
ch[offset].start = start;
ch[offset].cycles += cycles;
ch[offset].num++;
return 0;
}
static int __symbol__inc_addr_samples(struct map_symbol *ms,
struct annotated_source *src, int evidx, u64 addr,
struct perf_sample *sample)
{
struct symbol *sym = ms->sym;
unsigned offset;
struct sym_hist *h;
pr_debug3("%s: addr=%#" PRIx64 "\n", __func__, ms->map->unmap_ip(ms->map, addr));
if ((addr < sym->start || addr >= sym->end) &&
(addr != sym->end || sym->start != sym->end)) {
pr_debug("%s(%d): ERANGE! sym->name=%s, start=%#" PRIx64 ", addr=%#" PRIx64 ", end=%#" PRIx64 "\n",
__func__, __LINE__, sym->name, sym->start, addr, sym->end);
return -ERANGE;
}
offset = addr - sym->start;
h = annotated_source__histogram(src, evidx);
if (h == NULL) {
pr_debug("%s(%d): ENOMEM! sym->name=%s, start=%#" PRIx64 ", addr=%#" PRIx64 ", end=%#" PRIx64 ", func: %d\n",
__func__, __LINE__, sym->name, sym->start, addr, sym->end, sym->type == STT_FUNC);
return -ENOMEM;
}
h->nr_samples++;
h->addr[offset].nr_samples++;
h->period += sample->period;
h->addr[offset].period += sample->period;
pr_debug3("%#" PRIx64 " %s: period++ [addr: %#" PRIx64 ", %#" PRIx64
", evidx=%d] => nr_samples: %" PRIu64 ", period: %" PRIu64 "\n",
sym->start, sym->name, addr, addr - sym->start, evidx,
h->addr[offset].nr_samples, h->addr[offset].period);
return 0;
}
static struct cyc_hist *symbol__cycles_hist(struct symbol *sym)
{
struct annotation *notes = symbol__annotation(sym);
if (notes->src == NULL) {
notes->src = annotated_source__new();
if (notes->src == NULL)
return NULL;
goto alloc_cycles_hist;
}
if (!notes->src->cycles_hist) {
alloc_cycles_hist:
symbol__alloc_hist_cycles(sym);
}
return notes->src->cycles_hist;
}
struct annotated_source *symbol__hists(struct symbol *sym, int nr_hists)
{
struct annotation *notes = symbol__annotation(sym);
if (notes->src == NULL) {
notes->src = annotated_source__new();
if (notes->src == NULL)
return NULL;
goto alloc_histograms;
}
if (notes->src->histograms == NULL) {
alloc_histograms:
annotated_source__alloc_histograms(notes->src, symbol__size(sym),
nr_hists);
}
return notes->src;
}
static int symbol__inc_addr_samples(struct map_symbol *ms,
struct evsel *evsel, u64 addr,
struct perf_sample *sample)
{
struct symbol *sym = ms->sym;
struct annotated_source *src;
if (sym == NULL)
return 0;
src = symbol__hists(sym, evsel->evlist->core.nr_entries);
return src ? __symbol__inc_addr_samples(ms, src, evsel->core.idx, addr, sample) : 0;
}
static int symbol__account_cycles(u64 addr, u64 start,
struct symbol *sym, unsigned cycles)
{
struct cyc_hist *cycles_hist;
unsigned offset;
if (sym == NULL)
return 0;
cycles_hist = symbol__cycles_hist(sym);
if (cycles_hist == NULL)
return -ENOMEM;
if (addr < sym->start || addr >= sym->end)
return -ERANGE;
if (start) {
if (start < sym->start || start >= sym->end)
return -ERANGE;
if (start >= addr)
start = 0;
}
offset = addr - sym->start;
return __symbol__account_cycles(cycles_hist,
start ? start - sym->start : 0,
offset, cycles,
!!start);
}
int addr_map_symbol__account_cycles(struct addr_map_symbol *ams,
struct addr_map_symbol *start,
unsigned cycles)
{
u64 saddr = 0;
int err;
if (!cycles)
return 0;
/*
* Only set start when IPC can be computed. We can only
* compute it when the basic block is completely in a single
* function.
* Special case the case when the jump is elsewhere, but
* it starts on the function start.
*/
if (start &&
(start->ms.sym == ams->ms.sym ||
(ams->ms.sym &&
start->addr == ams->ms.sym->start + ams->ms.map->start)))
saddr = start->al_addr;
if (saddr == 0)
pr_debug2("BB with bad start: addr %"PRIx64" start %"PRIx64" sym %"PRIx64" saddr %"PRIx64"\n",
ams->addr,
start ? start->addr : 0,
ams->ms.sym ? ams->ms.sym->start + ams->ms.map->start : 0,
saddr);
err = symbol__account_cycles(ams->al_addr, saddr, ams->ms.sym, cycles);
if (err)
pr_debug2("account_cycles failed %d\n", err);
return err;
}
static unsigned annotation__count_insn(struct annotation *notes, u64 start, u64 end)
{
unsigned n_insn = 0;
u64 offset;
for (offset = start; offset <= end; offset++) {
if (notes->offsets[offset])
n_insn++;
}
return n_insn;
}
static void annotation__count_and_fill(struct annotation *notes, u64 start, u64 end, struct cyc_hist *ch)
{
unsigned n_insn;
perf annotate: Compute average IPC and IPC coverage per symbol Add support to 'perf report' annotate view or 'perf annotate --stdio2' to aggregate the IPC derived from timed LBRs per symbol. We compute the average IPC and the IPC coverage percentage. For example: $ perf annotate --stdio2 Percent IPC Cycle (Average IPC: 2.30, IPC Coverage: 54.8%) Disassembly of section .text: 000000000003aac0 <random@@GLIBC_2.2.5>: 8.32 3.28 sub $0x18,%rsp 3.28 mov $0x1,%esi 3.28 xor %eax,%eax 3.28 cmpl $0x0,argp_program_version_hook@@GLIBC_2.2.5+0x1e0 11.57 3.28 1 ↓ je 20 lock cmpxchg %esi,__abort_msg@@GLIBC_PRIVATE+0x8a0 ↓ jne 29 ↓ jmp 43 11.57 1.10 20: cmpxchg %esi,__abort_msg@@GLIBC_PRIVATE+0x8a0 0.00 1.10 1 ↓ je 43 29: lea __abort_msg@@GLIBC_PRIVATE+0x8a0,%rdi sub $0x80,%rsp → callq __lll_lock_wait_private add $0x80,%rsp 0.00 3.00 43: lea __ctype_b@GLIBC_2.2.5+0x38,%rdi 3.00 lea 0xc(%rsp),%rsi 8.49 3.00 1 → callq __random_r 7.91 1.94 cmpl $0x0,argp_program_version_hook@@GLIBC_2.2.5+0x1e0 0.00 1.94 1 ↓ je 68 lock decl __abort_msg@@GLIBC_PRIVATE+0x8a0 ↓ jne 70 ↓ jmp 8a 0.00 2.00 68: decl __abort_msg@@GLIBC_PRIVATE+0x8a0 21.56 2.00 1 ↓ je 8a 70: lea __abort_msg@@GLIBC_PRIVATE+0x8a0,%rdi sub $0x80,%rsp → callq __lll_unlock_wake_private add $0x80,%rsp 21.56 2.90 8a: movslq 0xc(%rsp),%rax 2.90 add $0x18,%rsp 9.03 2.90 1 ← retq It shows for this symbol the average IPC is 2.30 and the IPC coverage is 54.8%. Signed-off-by: Jin Yao <yao.jin@linux.intel.com> Reviewed-by: Ingo Molnar <mingo@kernel.org> Reviewed-by: Jiri Olsa <jolsa@kernel.org> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Andi Kleen <ak@linux.intel.com> Cc: Kan Liang <kan.liang@linux.intel.com> Cc: Peter Zijlstra <peterz@infradead.org> Link: http://lkml.kernel.org/r/1543586097-27632-2-git-send-email-yao.jin@linux.intel.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2018-11-30 16:54:54 +03:00
unsigned int cover_insn = 0;
u64 offset;
n_insn = annotation__count_insn(notes, start, end);
if (n_insn && ch->num && ch->cycles) {
float ipc = n_insn / ((double)ch->cycles / (double)ch->num);
/* Hide data when there are too many overlaps. */
perf annotate: Remove hist__account_cycles() from callback The hist__account_cycles() function is executed when the hist_iter__branch_callback() is called. But it looks it's not necessary. In hist__account_cycles, it already walks on all branch entries. This patch moves the hist__account_cycles out of callback, now the data processing is much faster than before. Previous code has an issue that the ch[offset].num++ (in __symbol__account_cycles) is executed repeatedly since hist__account_cycles is called in each hist_iter__branch_callback, so the counting of ch[offset].num is not correct (too big). With this patch, the issue is fixed. And we don't need the code of "ch->reset >= ch->num / 2" to check if there are too many overlaps (in annotation__count_and_fill), otherwise some data would be hidden. Now, we can try, for example: perf record -b ... perf annotate or perf report -s symbol The before/after output should be no change. v3: --- Fix the crash in stdio mode. Like previous code, it needs the checking of ui__has_annotation() before hist__account_cycles() v2: --- 1. Cover the similar perf report 2. Remove the checking code "ch->reset >= ch->num / 2" Signed-off-by: Jin Yao <yao.jin@linux.intel.com> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Andi Kleen <ak@linux.intel.com> Cc: Jin Yao <yao.jin@intel.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Kan Liang <kan.liang@linux.intel.com> Cc: Peter Zijlstra <peterz@infradead.org> Link: http://lkml.kernel.org/r/1552684577-29041-1-git-send-email-yao.jin@linux.intel.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2019-03-16 00:16:17 +03:00
if (ch->reset >= 0x7fff)
return;
for (offset = start; offset <= end; offset++) {
struct annotation_line *al = notes->offsets[offset];
perf annotate: Compute average IPC and IPC coverage per symbol Add support to 'perf report' annotate view or 'perf annotate --stdio2' to aggregate the IPC derived from timed LBRs per symbol. We compute the average IPC and the IPC coverage percentage. For example: $ perf annotate --stdio2 Percent IPC Cycle (Average IPC: 2.30, IPC Coverage: 54.8%) Disassembly of section .text: 000000000003aac0 <random@@GLIBC_2.2.5>: 8.32 3.28 sub $0x18,%rsp 3.28 mov $0x1,%esi 3.28 xor %eax,%eax 3.28 cmpl $0x0,argp_program_version_hook@@GLIBC_2.2.5+0x1e0 11.57 3.28 1 ↓ je 20 lock cmpxchg %esi,__abort_msg@@GLIBC_PRIVATE+0x8a0 ↓ jne 29 ↓ jmp 43 11.57 1.10 20: cmpxchg %esi,__abort_msg@@GLIBC_PRIVATE+0x8a0 0.00 1.10 1 ↓ je 43 29: lea __abort_msg@@GLIBC_PRIVATE+0x8a0,%rdi sub $0x80,%rsp → callq __lll_lock_wait_private add $0x80,%rsp 0.00 3.00 43: lea __ctype_b@GLIBC_2.2.5+0x38,%rdi 3.00 lea 0xc(%rsp),%rsi 8.49 3.00 1 → callq __random_r 7.91 1.94 cmpl $0x0,argp_program_version_hook@@GLIBC_2.2.5+0x1e0 0.00 1.94 1 ↓ je 68 lock decl __abort_msg@@GLIBC_PRIVATE+0x8a0 ↓ jne 70 ↓ jmp 8a 0.00 2.00 68: decl __abort_msg@@GLIBC_PRIVATE+0x8a0 21.56 2.00 1 ↓ je 8a 70: lea __abort_msg@@GLIBC_PRIVATE+0x8a0,%rdi sub $0x80,%rsp → callq __lll_unlock_wake_private add $0x80,%rsp 21.56 2.90 8a: movslq 0xc(%rsp),%rax 2.90 add $0x18,%rsp 9.03 2.90 1 ← retq It shows for this symbol the average IPC is 2.30 and the IPC coverage is 54.8%. Signed-off-by: Jin Yao <yao.jin@linux.intel.com> Reviewed-by: Ingo Molnar <mingo@kernel.org> Reviewed-by: Jiri Olsa <jolsa@kernel.org> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Andi Kleen <ak@linux.intel.com> Cc: Kan Liang <kan.liang@linux.intel.com> Cc: Peter Zijlstra <peterz@infradead.org> Link: http://lkml.kernel.org/r/1543586097-27632-2-git-send-email-yao.jin@linux.intel.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2018-11-30 16:54:54 +03:00
if (al && al->ipc == 0.0) {
al->ipc = ipc;
perf annotate: Compute average IPC and IPC coverage per symbol Add support to 'perf report' annotate view or 'perf annotate --stdio2' to aggregate the IPC derived from timed LBRs per symbol. We compute the average IPC and the IPC coverage percentage. For example: $ perf annotate --stdio2 Percent IPC Cycle (Average IPC: 2.30, IPC Coverage: 54.8%) Disassembly of section .text: 000000000003aac0 <random@@GLIBC_2.2.5>: 8.32 3.28 sub $0x18,%rsp 3.28 mov $0x1,%esi 3.28 xor %eax,%eax 3.28 cmpl $0x0,argp_program_version_hook@@GLIBC_2.2.5+0x1e0 11.57 3.28 1 ↓ je 20 lock cmpxchg %esi,__abort_msg@@GLIBC_PRIVATE+0x8a0 ↓ jne 29 ↓ jmp 43 11.57 1.10 20: cmpxchg %esi,__abort_msg@@GLIBC_PRIVATE+0x8a0 0.00 1.10 1 ↓ je 43 29: lea __abort_msg@@GLIBC_PRIVATE+0x8a0,%rdi sub $0x80,%rsp → callq __lll_lock_wait_private add $0x80,%rsp 0.00 3.00 43: lea __ctype_b@GLIBC_2.2.5+0x38,%rdi 3.00 lea 0xc(%rsp),%rsi 8.49 3.00 1 → callq __random_r 7.91 1.94 cmpl $0x0,argp_program_version_hook@@GLIBC_2.2.5+0x1e0 0.00 1.94 1 ↓ je 68 lock decl __abort_msg@@GLIBC_PRIVATE+0x8a0 ↓ jne 70 ↓ jmp 8a 0.00 2.00 68: decl __abort_msg@@GLIBC_PRIVATE+0x8a0 21.56 2.00 1 ↓ je 8a 70: lea __abort_msg@@GLIBC_PRIVATE+0x8a0,%rdi sub $0x80,%rsp → callq __lll_unlock_wake_private add $0x80,%rsp 21.56 2.90 8a: movslq 0xc(%rsp),%rax 2.90 add $0x18,%rsp 9.03 2.90 1 ← retq It shows for this symbol the average IPC is 2.30 and the IPC coverage is 54.8%. Signed-off-by: Jin Yao <yao.jin@linux.intel.com> Reviewed-by: Ingo Molnar <mingo@kernel.org> Reviewed-by: Jiri Olsa <jolsa@kernel.org> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Andi Kleen <ak@linux.intel.com> Cc: Kan Liang <kan.liang@linux.intel.com> Cc: Peter Zijlstra <peterz@infradead.org> Link: http://lkml.kernel.org/r/1543586097-27632-2-git-send-email-yao.jin@linux.intel.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2018-11-30 16:54:54 +03:00
cover_insn++;
}
}
if (cover_insn) {
notes->hit_cycles += ch->cycles;
notes->hit_insn += n_insn * ch->num;
notes->cover_insn += cover_insn;
}
}
}
void annotation__compute_ipc(struct annotation *notes, size_t size)
{
perf annotate: Compute average IPC and IPC coverage per symbol Add support to 'perf report' annotate view or 'perf annotate --stdio2' to aggregate the IPC derived from timed LBRs per symbol. We compute the average IPC and the IPC coverage percentage. For example: $ perf annotate --stdio2 Percent IPC Cycle (Average IPC: 2.30, IPC Coverage: 54.8%) Disassembly of section .text: 000000000003aac0 <random@@GLIBC_2.2.5>: 8.32 3.28 sub $0x18,%rsp 3.28 mov $0x1,%esi 3.28 xor %eax,%eax 3.28 cmpl $0x0,argp_program_version_hook@@GLIBC_2.2.5+0x1e0 11.57 3.28 1 ↓ je 20 lock cmpxchg %esi,__abort_msg@@GLIBC_PRIVATE+0x8a0 ↓ jne 29 ↓ jmp 43 11.57 1.10 20: cmpxchg %esi,__abort_msg@@GLIBC_PRIVATE+0x8a0 0.00 1.10 1 ↓ je 43 29: lea __abort_msg@@GLIBC_PRIVATE+0x8a0,%rdi sub $0x80,%rsp → callq __lll_lock_wait_private add $0x80,%rsp 0.00 3.00 43: lea __ctype_b@GLIBC_2.2.5+0x38,%rdi 3.00 lea 0xc(%rsp),%rsi 8.49 3.00 1 → callq __random_r 7.91 1.94 cmpl $0x0,argp_program_version_hook@@GLIBC_2.2.5+0x1e0 0.00 1.94 1 ↓ je 68 lock decl __abort_msg@@GLIBC_PRIVATE+0x8a0 ↓ jne 70 ↓ jmp 8a 0.00 2.00 68: decl __abort_msg@@GLIBC_PRIVATE+0x8a0 21.56 2.00 1 ↓ je 8a 70: lea __abort_msg@@GLIBC_PRIVATE+0x8a0,%rdi sub $0x80,%rsp → callq __lll_unlock_wake_private add $0x80,%rsp 21.56 2.90 8a: movslq 0xc(%rsp),%rax 2.90 add $0x18,%rsp 9.03 2.90 1 ← retq It shows for this symbol the average IPC is 2.30 and the IPC coverage is 54.8%. Signed-off-by: Jin Yao <yao.jin@linux.intel.com> Reviewed-by: Ingo Molnar <mingo@kernel.org> Reviewed-by: Jiri Olsa <jolsa@kernel.org> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Andi Kleen <ak@linux.intel.com> Cc: Kan Liang <kan.liang@linux.intel.com> Cc: Peter Zijlstra <peterz@infradead.org> Link: http://lkml.kernel.org/r/1543586097-27632-2-git-send-email-yao.jin@linux.intel.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2018-11-30 16:54:54 +03:00
s64 offset;
if (!notes->src || !notes->src->cycles_hist)
return;
perf annotate: Compute average IPC and IPC coverage per symbol Add support to 'perf report' annotate view or 'perf annotate --stdio2' to aggregate the IPC derived from timed LBRs per symbol. We compute the average IPC and the IPC coverage percentage. For example: $ perf annotate --stdio2 Percent IPC Cycle (Average IPC: 2.30, IPC Coverage: 54.8%) Disassembly of section .text: 000000000003aac0 <random@@GLIBC_2.2.5>: 8.32 3.28 sub $0x18,%rsp 3.28 mov $0x1,%esi 3.28 xor %eax,%eax 3.28 cmpl $0x0,argp_program_version_hook@@GLIBC_2.2.5+0x1e0 11.57 3.28 1 ↓ je 20 lock cmpxchg %esi,__abort_msg@@GLIBC_PRIVATE+0x8a0 ↓ jne 29 ↓ jmp 43 11.57 1.10 20: cmpxchg %esi,__abort_msg@@GLIBC_PRIVATE+0x8a0 0.00 1.10 1 ↓ je 43 29: lea __abort_msg@@GLIBC_PRIVATE+0x8a0,%rdi sub $0x80,%rsp → callq __lll_lock_wait_private add $0x80,%rsp 0.00 3.00 43: lea __ctype_b@GLIBC_2.2.5+0x38,%rdi 3.00 lea 0xc(%rsp),%rsi 8.49 3.00 1 → callq __random_r 7.91 1.94 cmpl $0x0,argp_program_version_hook@@GLIBC_2.2.5+0x1e0 0.00 1.94 1 ↓ je 68 lock decl __abort_msg@@GLIBC_PRIVATE+0x8a0 ↓ jne 70 ↓ jmp 8a 0.00 2.00 68: decl __abort_msg@@GLIBC_PRIVATE+0x8a0 21.56 2.00 1 ↓ je 8a 70: lea __abort_msg@@GLIBC_PRIVATE+0x8a0,%rdi sub $0x80,%rsp → callq __lll_unlock_wake_private add $0x80,%rsp 21.56 2.90 8a: movslq 0xc(%rsp),%rax 2.90 add $0x18,%rsp 9.03 2.90 1 ← retq It shows for this symbol the average IPC is 2.30 and the IPC coverage is 54.8%. Signed-off-by: Jin Yao <yao.jin@linux.intel.com> Reviewed-by: Ingo Molnar <mingo@kernel.org> Reviewed-by: Jiri Olsa <jolsa@kernel.org> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Andi Kleen <ak@linux.intel.com> Cc: Kan Liang <kan.liang@linux.intel.com> Cc: Peter Zijlstra <peterz@infradead.org> Link: http://lkml.kernel.org/r/1543586097-27632-2-git-send-email-yao.jin@linux.intel.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2018-11-30 16:54:54 +03:00
notes->total_insn = annotation__count_insn(notes, 0, size - 1);
notes->hit_cycles = 0;
notes->hit_insn = 0;
notes->cover_insn = 0;
pthread_mutex_lock(&notes->lock);
perf annotate: Compute average IPC and IPC coverage per symbol Add support to 'perf report' annotate view or 'perf annotate --stdio2' to aggregate the IPC derived from timed LBRs per symbol. We compute the average IPC and the IPC coverage percentage. For example: $ perf annotate --stdio2 Percent IPC Cycle (Average IPC: 2.30, IPC Coverage: 54.8%) Disassembly of section .text: 000000000003aac0 <random@@GLIBC_2.2.5>: 8.32 3.28 sub $0x18,%rsp 3.28 mov $0x1,%esi 3.28 xor %eax,%eax 3.28 cmpl $0x0,argp_program_version_hook@@GLIBC_2.2.5+0x1e0 11.57 3.28 1 ↓ je 20 lock cmpxchg %esi,__abort_msg@@GLIBC_PRIVATE+0x8a0 ↓ jne 29 ↓ jmp 43 11.57 1.10 20: cmpxchg %esi,__abort_msg@@GLIBC_PRIVATE+0x8a0 0.00 1.10 1 ↓ je 43 29: lea __abort_msg@@GLIBC_PRIVATE+0x8a0,%rdi sub $0x80,%rsp → callq __lll_lock_wait_private add $0x80,%rsp 0.00 3.00 43: lea __ctype_b@GLIBC_2.2.5+0x38,%rdi 3.00 lea 0xc(%rsp),%rsi 8.49 3.00 1 → callq __random_r 7.91 1.94 cmpl $0x0,argp_program_version_hook@@GLIBC_2.2.5+0x1e0 0.00 1.94 1 ↓ je 68 lock decl __abort_msg@@GLIBC_PRIVATE+0x8a0 ↓ jne 70 ↓ jmp 8a 0.00 2.00 68: decl __abort_msg@@GLIBC_PRIVATE+0x8a0 21.56 2.00 1 ↓ je 8a 70: lea __abort_msg@@GLIBC_PRIVATE+0x8a0,%rdi sub $0x80,%rsp → callq __lll_unlock_wake_private add $0x80,%rsp 21.56 2.90 8a: movslq 0xc(%rsp),%rax 2.90 add $0x18,%rsp 9.03 2.90 1 ← retq It shows for this symbol the average IPC is 2.30 and the IPC coverage is 54.8%. Signed-off-by: Jin Yao <yao.jin@linux.intel.com> Reviewed-by: Ingo Molnar <mingo@kernel.org> Reviewed-by: Jiri Olsa <jolsa@kernel.org> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Andi Kleen <ak@linux.intel.com> Cc: Kan Liang <kan.liang@linux.intel.com> Cc: Peter Zijlstra <peterz@infradead.org> Link: http://lkml.kernel.org/r/1543586097-27632-2-git-send-email-yao.jin@linux.intel.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2018-11-30 16:54:54 +03:00
for (offset = size - 1; offset >= 0; --offset) {
struct cyc_hist *ch;
ch = &notes->src->cycles_hist[offset];
if (ch && ch->cycles) {
struct annotation_line *al;
if (ch->have_start)
annotation__count_and_fill(notes, ch->start, offset, ch);
al = notes->offsets[offset];
if (al && ch->num_aggr) {
al->cycles = ch->cycles_aggr / ch->num_aggr;
al->cycles_max = ch->cycles_max;
al->cycles_min = ch->cycles_min;
}
notes->have_cycles = true;
}
}
pthread_mutex_unlock(&notes->lock);
}
int addr_map_symbol__inc_samples(struct addr_map_symbol *ams, struct perf_sample *sample,
struct evsel *evsel)
{
return symbol__inc_addr_samples(&ams->ms, evsel, ams->al_addr, sample);
}
int hist_entry__inc_addr_samples(struct hist_entry *he, struct perf_sample *sample,
struct evsel *evsel, u64 ip)
{
return symbol__inc_addr_samples(&he->ms, evsel, ip, sample);
}
static void disasm_line__init_ins(struct disasm_line *dl, struct arch *arch, struct map_symbol *ms)
{
perf annotate: Remove duplicate 'name' field from disasm_line The disasm_line::name field is always equal to ins::name, being used just to locate the instruction's ins_ops from the per-arch instructions table. Eliminate this duplication, nuking that field and instead make ins__find() return an ins_ops, store it in disasm_line::ins.ops, and keep just in disasm_line::ins.name what was in disasm_line::name, this way we end up not keeping a reference to entries in the per-arch instructions table. This in turn will help supporting multiple ways to manage the per-arch instructions table, allowing resorting that array, for instance, when the entries will move after references to its addresses were made. The same problem is avoided when one grows the array with realloc. So architectures simply keeping a constant array will work as well as architectures building the table using regular expressions or other logic that involves resorting the table. Reviewed-by: Ravi Bangoria <ravi.bangoria@linux.vnet.ibm.com> Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Chris Riyder <chris.ryder@arm.com> Cc: David Ahern <dsahern@gmail.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Kim Phillips <kim.phillips@arm.com> Cc: Markus Trippelsdorf <markus@trippelsdorf.de> Cc: Masami Hiramatsu <mhiramat@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Naveen N. Rao <naveen.n.rao@linux.vnet.ibm.com> Cc: Pawel Moll <pawel.moll@arm.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Russell King <rmk+kernel@arm.linux.org.uk> Cc: Taeung Song <treeze.taeung@gmail.com> Cc: Wang Nan <wangnan0@huawei.com> Link: http://lkml.kernel.org/n/tip-vr899azvabnw9gtuepuqfd9t@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-11-24 17:16:06 +03:00
dl->ins.ops = ins__find(arch, dl->ins.name);
perf annotate: Remove duplicate 'name' field from disasm_line The disasm_line::name field is always equal to ins::name, being used just to locate the instruction's ins_ops from the per-arch instructions table. Eliminate this duplication, nuking that field and instead make ins__find() return an ins_ops, store it in disasm_line::ins.ops, and keep just in disasm_line::ins.name what was in disasm_line::name, this way we end up not keeping a reference to entries in the per-arch instructions table. This in turn will help supporting multiple ways to manage the per-arch instructions table, allowing resorting that array, for instance, when the entries will move after references to its addresses were made. The same problem is avoided when one grows the array with realloc. So architectures simply keeping a constant array will work as well as architectures building the table using regular expressions or other logic that involves resorting the table. Reviewed-by: Ravi Bangoria <ravi.bangoria@linux.vnet.ibm.com> Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Chris Riyder <chris.ryder@arm.com> Cc: David Ahern <dsahern@gmail.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Kim Phillips <kim.phillips@arm.com> Cc: Markus Trippelsdorf <markus@trippelsdorf.de> Cc: Masami Hiramatsu <mhiramat@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Naveen N. Rao <naveen.n.rao@linux.vnet.ibm.com> Cc: Pawel Moll <pawel.moll@arm.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Russell King <rmk+kernel@arm.linux.org.uk> Cc: Taeung Song <treeze.taeung@gmail.com> Cc: Wang Nan <wangnan0@huawei.com> Link: http://lkml.kernel.org/n/tip-vr899azvabnw9gtuepuqfd9t@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-11-24 17:16:06 +03:00
if (!dl->ins.ops)
return;
if (dl->ins.ops->parse && dl->ins.ops->parse(arch, &dl->ops, ms) < 0)
perf annotate: Remove duplicate 'name' field from disasm_line The disasm_line::name field is always equal to ins::name, being used just to locate the instruction's ins_ops from the per-arch instructions table. Eliminate this duplication, nuking that field and instead make ins__find() return an ins_ops, store it in disasm_line::ins.ops, and keep just in disasm_line::ins.name what was in disasm_line::name, this way we end up not keeping a reference to entries in the per-arch instructions table. This in turn will help supporting multiple ways to manage the per-arch instructions table, allowing resorting that array, for instance, when the entries will move after references to its addresses were made. The same problem is avoided when one grows the array with realloc. So architectures simply keeping a constant array will work as well as architectures building the table using regular expressions or other logic that involves resorting the table. Reviewed-by: Ravi Bangoria <ravi.bangoria@linux.vnet.ibm.com> Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Chris Riyder <chris.ryder@arm.com> Cc: David Ahern <dsahern@gmail.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Kim Phillips <kim.phillips@arm.com> Cc: Markus Trippelsdorf <markus@trippelsdorf.de> Cc: Masami Hiramatsu <mhiramat@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Naveen N. Rao <naveen.n.rao@linux.vnet.ibm.com> Cc: Pawel Moll <pawel.moll@arm.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Russell King <rmk+kernel@arm.linux.org.uk> Cc: Taeung Song <treeze.taeung@gmail.com> Cc: Wang Nan <wangnan0@huawei.com> Link: http://lkml.kernel.org/n/tip-vr899azvabnw9gtuepuqfd9t@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-11-24 17:16:06 +03:00
dl->ins.ops = NULL;
}
perf annotate: Remove duplicate 'name' field from disasm_line The disasm_line::name field is always equal to ins::name, being used just to locate the instruction's ins_ops from the per-arch instructions table. Eliminate this duplication, nuking that field and instead make ins__find() return an ins_ops, store it in disasm_line::ins.ops, and keep just in disasm_line::ins.name what was in disasm_line::name, this way we end up not keeping a reference to entries in the per-arch instructions table. This in turn will help supporting multiple ways to manage the per-arch instructions table, allowing resorting that array, for instance, when the entries will move after references to its addresses were made. The same problem is avoided when one grows the array with realloc. So architectures simply keeping a constant array will work as well as architectures building the table using regular expressions or other logic that involves resorting the table. Reviewed-by: Ravi Bangoria <ravi.bangoria@linux.vnet.ibm.com> Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Chris Riyder <chris.ryder@arm.com> Cc: David Ahern <dsahern@gmail.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Kim Phillips <kim.phillips@arm.com> Cc: Markus Trippelsdorf <markus@trippelsdorf.de> Cc: Masami Hiramatsu <mhiramat@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Naveen N. Rao <naveen.n.rao@linux.vnet.ibm.com> Cc: Pawel Moll <pawel.moll@arm.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Russell King <rmk+kernel@arm.linux.org.uk> Cc: Taeung Song <treeze.taeung@gmail.com> Cc: Wang Nan <wangnan0@huawei.com> Link: http://lkml.kernel.org/n/tip-vr899azvabnw9gtuepuqfd9t@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-11-24 17:16:06 +03:00
static int disasm_line__parse(char *line, const char **namep, char **rawp)
{
char tmp, *name = skip_spaces(line);
if (name[0] == '\0')
return -1;
*rawp = name + 1;
while ((*rawp)[0] != '\0' && !isspace((*rawp)[0]))
++*rawp;
tmp = (*rawp)[0];
(*rawp)[0] = '\0';
*namep = strdup(name);
if (*namep == NULL)
perf annotate: Fix dereferencing freed memory found by the smatch tool Based on the following report from Smatch, fix the potential dereferencing freed memory check. tools/perf/util/annotate.c:1125 disasm_line__parse() error: dereferencing freed memory 'namep' tools/perf/util/annotate.c 1100 static int disasm_line__parse(char *line, const char **namep, char **rawp) 1101 { 1102 char tmp, *name = ltrim(line); [...] 1114 *namep = strdup(name); 1115 1116 if (*namep == NULL) 1117 goto out_free_name; [...] 1124 out_free_name: 1125 free((void *)namep); ^^^^^ 1126 *namep = NULL; ^^^^^^ 1127 return -1; 1128 } If strdup() fails to allocate memory space for *namep, we don't need to free memory with pointer 'namep', which is resident in data structure disasm_line::ins::name; and *namep is NULL pointer for this failure, so it's pointless to assign NULL to *namep again. Committer note: Freeing namep, which is the address of the first entry of the 'struct ins' that is the first member of struct disasm_line would in fact free that disasm_line instance, if it was allocated via malloc/calloc, which, later, would a dereference of freed memory. Signed-off-by: Leo Yan <leo.yan@linaro.org> Acked-by: Jiri Olsa <jolsa@kernel.org> Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Alexey Budankov <alexey.budankov@linux.intel.com> Cc: Alexios Zavras <alexios.zavras@intel.com> Cc: Andi Kleen <ak@linux.intel.com> Cc: Changbin Du <changbin.du@intel.com> Cc: David S. Miller <davem@davemloft.net> Cc: Davidlohr Bueso <dave@stgolabs.net> Cc: Eric Saint-Etienne <eric.saint.etienne@oracle.com> Cc: Jin Yao <yao.jin@linux.intel.com> Cc: Konstantin Khlebnikov <khlebnikov@yandex-team.ru> Cc: Mathieu Poirier <mathieu.poirier@linaro.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Rasmus Villemoes <linux@rasmusvillemoes.dk> Cc: Song Liu <songliubraving@fb.com> Cc: Suzuki Poulouse <suzuki.poulose@arm.com> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: Thomas Richter <tmricht@linux.ibm.com> Cc: linux-arm-kernel@lists.infradead.org Link: http://lkml.kernel.org/r/20190702103420.27540-5-leo.yan@linaro.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2019-07-02 13:34:13 +03:00
goto out;
(*rawp)[0] = tmp;
perf annotate: Fix printing of unaugmented disassembled instructions from BPF The code to disassemble BPF programs uses binutil's disassembling routines, and those use in turn fprintf to print to a memstream FILE, adding a newline at the end of each line, which ends up confusing the TUI routines called from: annotate_browser__write() annotate_line__write() annotate_browser__printf() ui_browser__vprintf() SLsmg_vprintf() The SLsmg_vprintf() function in the slang library gets confused with the terminating newline, so make the disasm_line__parse() function that parses the lines produced by the BPF specific disassembler (that uses binutil's libopcodes) and the lines produced by the objdump based disassembler used for everything else (and that doesn't adds this terminating newline) trim the end of the line in addition of the beginning. This way when disasm_line->ops.raw, i.e. for instructions without a special scnprintf() method, we'll not have that \n getting in the way of filling the screen right after the instruction with spaces to avoid leaving what was on the screen before and thus garbling the annotation screen, breaking scrolling, etc. Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Song Liu <songliubraving@fb.com> Fixes: 6987561c9e86 ("perf annotate: Enable annotation of BPF programs") Link: https://lkml.kernel.org/n/tip-unbr5a5efakobfr6rhxq99ta@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2019-08-06 17:24:09 +03:00
*rawp = strim(*rawp);
return 0;
perf annotate: Fix dereferencing freed memory found by the smatch tool Based on the following report from Smatch, fix the potential dereferencing freed memory check. tools/perf/util/annotate.c:1125 disasm_line__parse() error: dereferencing freed memory 'namep' tools/perf/util/annotate.c 1100 static int disasm_line__parse(char *line, const char **namep, char **rawp) 1101 { 1102 char tmp, *name = ltrim(line); [...] 1114 *namep = strdup(name); 1115 1116 if (*namep == NULL) 1117 goto out_free_name; [...] 1124 out_free_name: 1125 free((void *)namep); ^^^^^ 1126 *namep = NULL; ^^^^^^ 1127 return -1; 1128 } If strdup() fails to allocate memory space for *namep, we don't need to free memory with pointer 'namep', which is resident in data structure disasm_line::ins::name; and *namep is NULL pointer for this failure, so it's pointless to assign NULL to *namep again. Committer note: Freeing namep, which is the address of the first entry of the 'struct ins' that is the first member of struct disasm_line would in fact free that disasm_line instance, if it was allocated via malloc/calloc, which, later, would a dereference of freed memory. Signed-off-by: Leo Yan <leo.yan@linaro.org> Acked-by: Jiri Olsa <jolsa@kernel.org> Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Alexey Budankov <alexey.budankov@linux.intel.com> Cc: Alexios Zavras <alexios.zavras@intel.com> Cc: Andi Kleen <ak@linux.intel.com> Cc: Changbin Du <changbin.du@intel.com> Cc: David S. Miller <davem@davemloft.net> Cc: Davidlohr Bueso <dave@stgolabs.net> Cc: Eric Saint-Etienne <eric.saint.etienne@oracle.com> Cc: Jin Yao <yao.jin@linux.intel.com> Cc: Konstantin Khlebnikov <khlebnikov@yandex-team.ru> Cc: Mathieu Poirier <mathieu.poirier@linaro.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Rasmus Villemoes <linux@rasmusvillemoes.dk> Cc: Song Liu <songliubraving@fb.com> Cc: Suzuki Poulouse <suzuki.poulose@arm.com> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: Thomas Richter <tmricht@linux.ibm.com> Cc: linux-arm-kernel@lists.infradead.org Link: http://lkml.kernel.org/r/20190702103420.27540-5-leo.yan@linaro.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2019-07-02 13:34:13 +03:00
out:
return -1;
}
struct annotate_args {
struct arch *arch;
struct map_symbol ms;
struct evsel *evsel;
struct annotation_options *options;
s64 offset;
char *line;
int line_nr;
char *fileloc;
};
static void annotation_line__init(struct annotation_line *al,
struct annotate_args *args,
int nr)
{
al->offset = args->offset;
al->line = strdup(args->line);
al->line_nr = args->line_nr;
al->fileloc = args->fileloc;
al->data_nr = nr;
}
static void annotation_line__exit(struct annotation_line *al)
{
free_srcline(al->path);
zfree(&al->line);
}
static size_t disasm_line_size(int nr)
{
struct annotation_line *al;
return (sizeof(struct disasm_line) + (sizeof(al->data[0]) * nr));
}
/*
* Allocating the disasm annotation line data with
* following structure:
*
* -------------------------------------------
* struct disasm_line | struct annotation_line
* -------------------------------------------
*
* We have 'struct annotation_line' member as last member
* of 'struct disasm_line' to have an easy access.
*/
static struct disasm_line *disasm_line__new(struct annotate_args *args)
{
struct disasm_line *dl = NULL;
int nr = 1;
if (evsel__is_group_event(args->evsel))
nr = args->evsel->core.nr_members;
dl = zalloc(disasm_line_size(nr));
if (!dl)
return NULL;
annotation_line__init(&dl->al, args, nr);
if (dl->al.line == NULL)
goto out_delete;
if (args->offset != -1) {
if (disasm_line__parse(dl->al.line, &dl->ins.name, &dl->ops.raw) < 0)
goto out_free_line;
disasm_line__init_ins(dl, args->arch, &args->ms);
}
return dl;
out_free_line:
zfree(&dl->al.line);
out_delete:
free(dl);
return NULL;
}
void disasm_line__free(struct disasm_line *dl)
{
perf annotate: Remove duplicate 'name' field from disasm_line The disasm_line::name field is always equal to ins::name, being used just to locate the instruction's ins_ops from the per-arch instructions table. Eliminate this duplication, nuking that field and instead make ins__find() return an ins_ops, store it in disasm_line::ins.ops, and keep just in disasm_line::ins.name what was in disasm_line::name, this way we end up not keeping a reference to entries in the per-arch instructions table. This in turn will help supporting multiple ways to manage the per-arch instructions table, allowing resorting that array, for instance, when the entries will move after references to its addresses were made. The same problem is avoided when one grows the array with realloc. So architectures simply keeping a constant array will work as well as architectures building the table using regular expressions or other logic that involves resorting the table. Reviewed-by: Ravi Bangoria <ravi.bangoria@linux.vnet.ibm.com> Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Chris Riyder <chris.ryder@arm.com> Cc: David Ahern <dsahern@gmail.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Kim Phillips <kim.phillips@arm.com> Cc: Markus Trippelsdorf <markus@trippelsdorf.de> Cc: Masami Hiramatsu <mhiramat@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Naveen N. Rao <naveen.n.rao@linux.vnet.ibm.com> Cc: Pawel Moll <pawel.moll@arm.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Russell King <rmk+kernel@arm.linux.org.uk> Cc: Taeung Song <treeze.taeung@gmail.com> Cc: Wang Nan <wangnan0@huawei.com> Link: http://lkml.kernel.org/n/tip-vr899azvabnw9gtuepuqfd9t@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-11-24 17:16:06 +03:00
if (dl->ins.ops && dl->ins.ops->free)
dl->ins.ops->free(&dl->ops);
else
ins__delete(&dl->ops);
zfree(&dl->ins.name);
annotation_line__exit(&dl->al);
free(dl);
}
perf annotate: Calculate the max instruction name, align column to that We were hardcoding '6' as the max instruction name, and we have lots that are longer than that, see the diff from two 'P' printed TUI annotations for a libc function that uses instructions with long names, such as 'vpmovmskb' with its 9 chars: --- __strcmp_avx2.annotation.before 2019-03-06 16:31:39.368020425 -0300 +++ __strcmp_avx2.annotation 2019-03-06 16:32:12.079450508 -0300 @@ -2,284 +2,284 @@ Event: cycles:ppp Percent endbr64 - 0.10 mov %edi,%eax + 0.10 mov %edi,%eax - xor %edx,%edx + xor %edx,%edx - 3.54 vpxor %ymm7,%ymm7,%ymm7 + 3.54 vpxor %ymm7,%ymm7,%ymm7 - or %esi,%eax + or %esi,%eax - and $0xfff,%eax + and $0xfff,%eax - cmp $0xf80,%eax + cmp $0xf80,%eax - ↓ jg 370 + ↓ jg 370 - 27.07 vmovdqu (%rdi),%ymm1 + 27.07 vmovdqu (%rdi),%ymm1 - 7.97 vpcmpeqb (%rsi),%ymm1,%ymm0 + 7.97 vpcmpeqb (%rsi),%ymm1,%ymm0 - 2.15 vpminub %ymm1,%ymm0,%ymm0 + 2.15 vpminub %ymm1,%ymm0,%ymm0 - 4.09 vpcmpeqb %ymm7,%ymm0,%ymm0 + 4.09 vpcmpeqb %ymm7,%ymm0,%ymm0 - 0.43 vpmovmskb %ymm0,%ecx + 0.43 vpmovmskb %ymm0,%ecx - 1.53 test %ecx,%ecx + 1.53 test %ecx,%ecx - ↓ je b0 + ↓ je b0 - 5.26 tzcnt %ecx,%edx + 5.26 tzcnt %ecx,%edx - 18.40 movzbl (%rdi,%rdx,1),%eax + 18.40 movzbl (%rdi,%rdx,1),%eax - 7.09 movzbl (%rsi,%rdx,1),%edx + 7.09 movzbl (%rsi,%rdx,1),%edx - 3.34 sub %edx,%eax + 3.34 sub %edx,%eax 2.37 vzeroupper ← retq nop - 50: tzcnt %ecx,%edx + 50: tzcnt %ecx,%edx - movzbl 0x20(%rdi,%rdx,1),%eax + movzbl 0x20(%rdi,%rdx,1),%eax - movzbl 0x20(%rsi,%rdx,1),%edx + movzbl 0x20(%rsi,%rdx,1),%edx - sub %edx,%eax + sub %edx,%eax vzeroupper ← retq - data16 nopw %cs:0x0(%rax,%rax,1) + data16 nopw %cs:0x0(%rax,%rax,1) Reported-by: Travis Downs <travis.downs@gmail.com> LPU-Reference: CAOBGo4z1KfmWeOm6Et0cnX5Z6DWsG2PQbAvRn1MhVPJmXHrc5g@mail.gmail.com Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Link: https://lkml.kernel.org/n/tip-89wsdd9h9g6bvq52sgp6d0u4@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2019-03-06 22:40:15 +03:00
int disasm_line__scnprintf(struct disasm_line *dl, char *bf, size_t size, bool raw, int max_ins_name)
{
perf annotate: Remove duplicate 'name' field from disasm_line The disasm_line::name field is always equal to ins::name, being used just to locate the instruction's ins_ops from the per-arch instructions table. Eliminate this duplication, nuking that field and instead make ins__find() return an ins_ops, store it in disasm_line::ins.ops, and keep just in disasm_line::ins.name what was in disasm_line::name, this way we end up not keeping a reference to entries in the per-arch instructions table. This in turn will help supporting multiple ways to manage the per-arch instructions table, allowing resorting that array, for instance, when the entries will move after references to its addresses were made. The same problem is avoided when one grows the array with realloc. So architectures simply keeping a constant array will work as well as architectures building the table using regular expressions or other logic that involves resorting the table. Reviewed-by: Ravi Bangoria <ravi.bangoria@linux.vnet.ibm.com> Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Chris Riyder <chris.ryder@arm.com> Cc: David Ahern <dsahern@gmail.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Kim Phillips <kim.phillips@arm.com> Cc: Markus Trippelsdorf <markus@trippelsdorf.de> Cc: Masami Hiramatsu <mhiramat@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Naveen N. Rao <naveen.n.rao@linux.vnet.ibm.com> Cc: Pawel Moll <pawel.moll@arm.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Russell King <rmk+kernel@arm.linux.org.uk> Cc: Taeung Song <treeze.taeung@gmail.com> Cc: Wang Nan <wangnan0@huawei.com> Link: http://lkml.kernel.org/n/tip-vr899azvabnw9gtuepuqfd9t@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-11-24 17:16:06 +03:00
if (raw || !dl->ins.ops)
perf annotate: Calculate the max instruction name, align column to that We were hardcoding '6' as the max instruction name, and we have lots that are longer than that, see the diff from two 'P' printed TUI annotations for a libc function that uses instructions with long names, such as 'vpmovmskb' with its 9 chars: --- __strcmp_avx2.annotation.before 2019-03-06 16:31:39.368020425 -0300 +++ __strcmp_avx2.annotation 2019-03-06 16:32:12.079450508 -0300 @@ -2,284 +2,284 @@ Event: cycles:ppp Percent endbr64 - 0.10 mov %edi,%eax + 0.10 mov %edi,%eax - xor %edx,%edx + xor %edx,%edx - 3.54 vpxor %ymm7,%ymm7,%ymm7 + 3.54 vpxor %ymm7,%ymm7,%ymm7 - or %esi,%eax + or %esi,%eax - and $0xfff,%eax + and $0xfff,%eax - cmp $0xf80,%eax + cmp $0xf80,%eax - ↓ jg 370 + ↓ jg 370 - 27.07 vmovdqu (%rdi),%ymm1 + 27.07 vmovdqu (%rdi),%ymm1 - 7.97 vpcmpeqb (%rsi),%ymm1,%ymm0 + 7.97 vpcmpeqb (%rsi),%ymm1,%ymm0 - 2.15 vpminub %ymm1,%ymm0,%ymm0 + 2.15 vpminub %ymm1,%ymm0,%ymm0 - 4.09 vpcmpeqb %ymm7,%ymm0,%ymm0 + 4.09 vpcmpeqb %ymm7,%ymm0,%ymm0 - 0.43 vpmovmskb %ymm0,%ecx + 0.43 vpmovmskb %ymm0,%ecx - 1.53 test %ecx,%ecx + 1.53 test %ecx,%ecx - ↓ je b0 + ↓ je b0 - 5.26 tzcnt %ecx,%edx + 5.26 tzcnt %ecx,%edx - 18.40 movzbl (%rdi,%rdx,1),%eax + 18.40 movzbl (%rdi,%rdx,1),%eax - 7.09 movzbl (%rsi,%rdx,1),%edx + 7.09 movzbl (%rsi,%rdx,1),%edx - 3.34 sub %edx,%eax + 3.34 sub %edx,%eax 2.37 vzeroupper ← retq nop - 50: tzcnt %ecx,%edx + 50: tzcnt %ecx,%edx - movzbl 0x20(%rdi,%rdx,1),%eax + movzbl 0x20(%rdi,%rdx,1),%eax - movzbl 0x20(%rsi,%rdx,1),%edx + movzbl 0x20(%rsi,%rdx,1),%edx - sub %edx,%eax + sub %edx,%eax vzeroupper ← retq - data16 nopw %cs:0x0(%rax,%rax,1) + data16 nopw %cs:0x0(%rax,%rax,1) Reported-by: Travis Downs <travis.downs@gmail.com> LPU-Reference: CAOBGo4z1KfmWeOm6Et0cnX5Z6DWsG2PQbAvRn1MhVPJmXHrc5g@mail.gmail.com Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Link: https://lkml.kernel.org/n/tip-89wsdd9h9g6bvq52sgp6d0u4@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2019-03-06 22:40:15 +03:00
return scnprintf(bf, size, "%-*s %s", max_ins_name, dl->ins.name, dl->ops.raw);
perf annotate: Calculate the max instruction name, align column to that We were hardcoding '6' as the max instruction name, and we have lots that are longer than that, see the diff from two 'P' printed TUI annotations for a libc function that uses instructions with long names, such as 'vpmovmskb' with its 9 chars: --- __strcmp_avx2.annotation.before 2019-03-06 16:31:39.368020425 -0300 +++ __strcmp_avx2.annotation 2019-03-06 16:32:12.079450508 -0300 @@ -2,284 +2,284 @@ Event: cycles:ppp Percent endbr64 - 0.10 mov %edi,%eax + 0.10 mov %edi,%eax - xor %edx,%edx + xor %edx,%edx - 3.54 vpxor %ymm7,%ymm7,%ymm7 + 3.54 vpxor %ymm7,%ymm7,%ymm7 - or %esi,%eax + or %esi,%eax - and $0xfff,%eax + and $0xfff,%eax - cmp $0xf80,%eax + cmp $0xf80,%eax - ↓ jg 370 + ↓ jg 370 - 27.07 vmovdqu (%rdi),%ymm1 + 27.07 vmovdqu (%rdi),%ymm1 - 7.97 vpcmpeqb (%rsi),%ymm1,%ymm0 + 7.97 vpcmpeqb (%rsi),%ymm1,%ymm0 - 2.15 vpminub %ymm1,%ymm0,%ymm0 + 2.15 vpminub %ymm1,%ymm0,%ymm0 - 4.09 vpcmpeqb %ymm7,%ymm0,%ymm0 + 4.09 vpcmpeqb %ymm7,%ymm0,%ymm0 - 0.43 vpmovmskb %ymm0,%ecx + 0.43 vpmovmskb %ymm0,%ecx - 1.53 test %ecx,%ecx + 1.53 test %ecx,%ecx - ↓ je b0 + ↓ je b0 - 5.26 tzcnt %ecx,%edx + 5.26 tzcnt %ecx,%edx - 18.40 movzbl (%rdi,%rdx,1),%eax + 18.40 movzbl (%rdi,%rdx,1),%eax - 7.09 movzbl (%rsi,%rdx,1),%edx + 7.09 movzbl (%rsi,%rdx,1),%edx - 3.34 sub %edx,%eax + 3.34 sub %edx,%eax 2.37 vzeroupper ← retq nop - 50: tzcnt %ecx,%edx + 50: tzcnt %ecx,%edx - movzbl 0x20(%rdi,%rdx,1),%eax + movzbl 0x20(%rdi,%rdx,1),%eax - movzbl 0x20(%rsi,%rdx,1),%edx + movzbl 0x20(%rsi,%rdx,1),%edx - sub %edx,%eax + sub %edx,%eax vzeroupper ← retq - data16 nopw %cs:0x0(%rax,%rax,1) + data16 nopw %cs:0x0(%rax,%rax,1) Reported-by: Travis Downs <travis.downs@gmail.com> LPU-Reference: CAOBGo4z1KfmWeOm6Et0cnX5Z6DWsG2PQbAvRn1MhVPJmXHrc5g@mail.gmail.com Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Link: https://lkml.kernel.org/n/tip-89wsdd9h9g6bvq52sgp6d0u4@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2019-03-06 22:40:15 +03:00
return ins__scnprintf(&dl->ins, bf, size, &dl->ops, max_ins_name);
}
perf symbols: Factor out annotation init/exit The exit function fixes a memory leak with the src field as detected by leak sanitizer. An example of which is: Indirect leak of 25133184 byte(s) in 207 object(s) allocated from: #0 0x7f199ecfe987 in __interceptor_calloc libsanitizer/asan/asan_malloc_linux.cpp:154 #1 0x55defe638224 in annotated_source__alloc_histograms util/annotate.c:803 #2 0x55defe6397e4 in symbol__hists util/annotate.c:952 #3 0x55defe639908 in symbol__inc_addr_samples util/annotate.c:968 #4 0x55defe63aa29 in hist_entry__inc_addr_samples util/annotate.c:1119 #5 0x55defe499a79 in hist_iter__report_callback tools/perf/builtin-report.c:182 #6 0x55defe7a859d in hist_entry_iter__add util/hist.c:1236 #7 0x55defe49aa63 in process_sample_event tools/perf/builtin-report.c:315 #8 0x55defe731bc8 in evlist__deliver_sample util/session.c:1473 #9 0x55defe731e38 in machines__deliver_event util/session.c:1510 #10 0x55defe732a23 in perf_session__deliver_event util/session.c:1590 #11 0x55defe72951e in ordered_events__deliver_event util/session.c:183 #12 0x55defe740082 in do_flush util/ordered-events.c:244 #13 0x55defe7407cb in __ordered_events__flush util/ordered-events.c:323 #14 0x55defe740a61 in ordered_events__flush util/ordered-events.c:341 #15 0x55defe73837f in __perf_session__process_events util/session.c:2390 #16 0x55defe7385ff in perf_session__process_events util/session.c:2420 ... Signed-off-by: Ian Rogers <irogers@google.com> Acked-by: Namhyung Kim <namhyung@kernel.org> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Ingo Molnar <mingo@redhat.com> Cc: James Clark <james.clark@arm.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Kajol Jain <kjain@linux.ibm.com> Cc: Mark Rutland <mark.rutland@arm.com> Cc: Martin Liška <mliska@suse.cz> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Stephane Eranian <eranian@google.com> Link: https://lore.kernel.org/r/20211112035124.94327-3-irogers@google.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2021-11-12 06:51:24 +03:00
void annotation__init(struct annotation *notes)
{
pthread_mutex_init(&notes->lock, NULL);
}
void annotation__exit(struct annotation *notes)
{
annotated_source__delete(notes->src);
pthread_mutex_destroy(&notes->lock);
}
static void annotation_line__add(struct annotation_line *al, struct list_head *head)
{
list_add_tail(&al->node, head);
}
struct annotation_line *
annotation_line__next(struct annotation_line *pos, struct list_head *head)
{
list_for_each_entry_continue(pos, head, node)
if (pos->offset >= 0)
return pos;
return NULL;
}
perf annotate: Add branch stack / basic block I wanted to know the hottest path through a function and figured the branch-stack (LBR) information should be able to help out with that. The below uses the branch-stack to create basic blocks and generate statistics from them. from to branch_i * ----> * | | block v * ----> * from to branch_i+1 The blocks are broken down into non-overlapping ranges, while tracking if the start of each range is an entry point and/or the end of a range is a branch. Each block iterates all ranges it covers (while splitting where required to exactly match the block) and increments the 'coverage' count. For the range including the branch we increment the taken counter, as well as the pred counter if flags.predicted. Using these number we can find if an instruction: - had coverage; given by: br->coverage / br->sym->max_coverage This metric ensures each symbol has a 100% spot, which reflects the observation that each symbol must have a most covered/hottest block. - is a branch target: br->is_target && br->start == add - for targets, how much of a branch's coverages comes from it: target->entry / branch->coverage - is a branch: br->is_branch && br->end == addr - for branches, how often it was taken: br->taken / br->coverage after all, all execution that didn't take the branch would have incremented the coverage and continued onward to a later branch. - for branches, how often it was predicted: br->pred / br->taken The coverage percentage is used to color the address and asm sections; for low (<1%) coverage we use NORMAL (uncolored), indicating that these instructions are not 'important'. For high coverage (>75%) we color the address RED. For each branch, we add an asm comment after the instruction with information on how often it was taken and predicted. Output looks like (sans color, which does loose a lot of the information :/) $ perf record --branch-filter u,any -e cycles:p ./branches 27 $ perf annotate branches Percent | Source code & Disassembly of branches for cycles:pu (217 samples) --------------------------------------------------------------------------------- : branches(): 0.00 : 40057a: push %rbp 0.00 : 40057b: mov %rsp,%rbp 0.00 : 40057e: sub $0x20,%rsp 0.00 : 400582: mov %rdi,-0x18(%rbp) 0.00 : 400586: mov %rsi,-0x20(%rbp) 0.00 : 40058a: mov -0x18(%rbp),%rax 0.00 : 40058e: mov %rax,-0x10(%rbp) 0.00 : 400592: movq $0x0,-0x8(%rbp) 0.00 : 40059a: jmpq 400656 <branches+0xdc> 1.84 : 40059f: mov -0x10(%rbp),%rax # +100.00% 3.23 : 4005a3: and $0x1,%eax 1.84 : 4005a6: test %rax,%rax 0.00 : 4005a9: je 4005bf <branches+0x45> # -54.50% (p:42.00%) 0.46 : 4005ab: mov 0x200bbe(%rip),%rax # 601170 <acc> 12.90 : 4005b2: add $0x1,%rax 2.30 : 4005b6: mov %rax,0x200bb3(%rip) # 601170 <acc> 0.46 : 4005bd: jmp 4005d1 <branches+0x57> # -100.00% (p:100.00%) 0.92 : 4005bf: mov 0x200baa(%rip),%rax # 601170 <acc> # +49.54% 13.82 : 4005c6: sub $0x1,%rax 0.46 : 4005ca: mov %rax,0x200b9f(%rip) # 601170 <acc> 2.30 : 4005d1: mov -0x10(%rbp),%rax # +50.46% 0.46 : 4005d5: mov %rax,%rdi 0.46 : 4005d8: callq 400526 <lfsr> # -100.00% (p:100.00%) 0.00 : 4005dd: mov %rax,-0x10(%rbp) # +100.00% 0.92 : 4005e1: mov -0x18(%rbp),%rax 0.00 : 4005e5: and $0x1,%eax 0.00 : 4005e8: test %rax,%rax 0.00 : 4005eb: je 4005ff <branches+0x85> # -100.00% (p:100.00%) 0.00 : 4005ed: mov 0x200b7c(%rip),%rax # 601170 <acc> 0.00 : 4005f4: shr $0x2,%rax 0.00 : 4005f8: mov %rax,0x200b71(%rip) # 601170 <acc> 0.00 : 4005ff: mov -0x10(%rbp),%rax # +100.00% 7.37 : 400603: and $0x1,%eax 3.69 : 400606: test %rax,%rax 0.00 : 400609: jne 400612 <branches+0x98> # -59.25% (p:42.99%) 1.84 : 40060b: mov $0x1,%eax 14.29 : 400610: jmp 400617 <branches+0x9d> # -100.00% (p:100.00%) 1.38 : 400612: mov $0x0,%eax # +57.65% 10.14 : 400617: test %al,%al # +42.35% 0.00 : 400619: je 40062f <branches+0xb5> # -57.65% (p:100.00%) 0.46 : 40061b: mov 0x200b4e(%rip),%rax # 601170 <acc> 2.76 : 400622: sub $0x1,%rax 0.00 : 400626: mov %rax,0x200b43(%rip) # 601170 <acc> 0.46 : 40062d: jmp 400641 <branches+0xc7> # -100.00% (p:100.00%) 0.92 : 40062f: mov 0x200b3a(%rip),%rax # 601170 <acc> # +56.13% 2.30 : 400636: add $0x1,%rax 0.92 : 40063a: mov %rax,0x200b2f(%rip) # 601170 <acc> 0.92 : 400641: mov -0x10(%rbp),%rax # +43.87% 2.30 : 400645: mov %rax,%rdi 0.00 : 400648: callq 400526 <lfsr> # -100.00% (p:100.00%) 0.00 : 40064d: mov %rax,-0x10(%rbp) # +100.00% 1.84 : 400651: addq $0x1,-0x8(%rbp) 0.92 : 400656: mov -0x8(%rbp),%rax 5.07 : 40065a: cmp -0x20(%rbp),%rax 0.00 : 40065e: jb 40059f <branches+0x25> # -100.00% (p:100.00%) 0.00 : 400664: nop 0.00 : 400665: leaveq 0.00 : 400666: retq (Note: the --branch-filter u,any was used to avoid spurious target and branch points due to interrupts/faults, they show up as very small -/+ annotations on 'weird' locations) Committer note: Please take a look at: http://vger.kernel.org/~acme/perf/annotate_basic_blocks.png To see the colors. Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Andi Kleen <andi@firstfloor.org> Cc: Anshuman Khandual <khandual@linux.vnet.ibm.com> Cc: David Carrillo-Cisneros <davidcc@google.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Kan Liang <kan.liang@intel.com> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Stephane Eranian <eranian@google.com> [ Moved sym->max_coverage to 'struct annotate', aka symbol__annotate(sym) ] Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-09-05 22:08:12 +03:00
static const char *annotate__address_color(struct block_range *br)
{
double cov = block_range__coverage(br);
if (cov >= 0) {
/* mark red for >75% coverage */
if (cov > 0.75)
return PERF_COLOR_RED;
/* mark dull for <1% coverage */
if (cov < 0.01)
return PERF_COLOR_NORMAL;
}
return PERF_COLOR_MAGENTA;
}
static const char *annotate__asm_color(struct block_range *br)
{
double cov = block_range__coverage(br);
if (cov >= 0) {
/* mark dull for <1% coverage */
if (cov < 0.01)
return PERF_COLOR_NORMAL;
}
return PERF_COLOR_BLUE;
}
static void annotate__branch_printf(struct block_range *br, u64 addr)
{
bool emit_comment = true;
if (!br)
return;
#if 1
if (br->is_target && br->start == addr) {
struct block_range *branch = br;
double p;
/*
* Find matching branch to our target.
*/
while (!branch->is_branch)
branch = block_range__next(branch);
p = 100 *(double)br->entry / branch->coverage;
if (p > 0.1) {
if (emit_comment) {
emit_comment = false;
printf("\t#");
}
/*
* The percentage of coverage joined at this target in relation
* to the next branch.
*/
printf(" +%.2f%%", p);
}
}
#endif
if (br->is_branch && br->end == addr) {
double p = 100*(double)br->taken / br->coverage;
if (p > 0.1) {
if (emit_comment) {
emit_comment = false;
printf("\t#");
}
/*
* The percentage of coverage leaving at this branch, and
* its prediction ratio.
*/
printf(" -%.2f%% (p:%.2f%%)", p, 100*(double)br->pred / br->taken);
}
}
}
perf annotate: Align source and offset lines Align source with offset lines, which are more advanced, because of the address column. Before: : static void *worker_thread(void *__tdata) : { 0.00 : 48a971: push %rbp 0.00 : 48a972: mov %rsp,%rbp 0.00 : 48a975: sub $0x30,%rsp 0.00 : 48a979: mov %rdi,-0x28(%rbp) 0.00 : 48a97d: mov %fs:0x28,%rax 0.00 : 48a986: mov %rax,-0x8(%rbp) 0.00 : 48a98a: xor %eax,%eax : struct thread_data *td = __tdata; 0.00 : 48a98c: mov -0x28(%rbp),%rax 0.00 : 48a990: mov %rax,-0x10(%rbp) : int m = 0, i; 0.00 : 48a994: movl $0x0,-0x1c(%rbp) : int ret; : : for (i = 0; i < loops; i++) { 0.00 : 48a99b: movl $0x0,-0x18(%rbp) After: : static void *worker_thread(void *__tdata) : { 0.00 : 48a971: push %rbp 0.00 : 48a972: mov %rsp,%rbp 0.00 : 48a975: sub $0x30,%rsp 0.00 : 48a979: mov %rdi,-0x28(%rbp) 0.00 : 48a97d: mov %fs:0x28,%rax 0.00 : 48a986: mov %rax,-0x8(%rbp) 0.00 : 48a98a: xor %eax,%eax : struct thread_data *td = __tdata; 0.00 : 48a98c: mov -0x28(%rbp),%rax 0.00 : 48a990: mov %rax,-0x10(%rbp) : int m = 0, i; 0.00 : 48a994: movl $0x0,-0x1c(%rbp) : int ret; : : for (i = 0; i < loops; i++) { 0.00 : 48a99b: movl $0x0,-0x18(%rbp) It makes bigger different when displaying script sources, where the comment lines looks oddly shifted from the lines which actually hold code. I'll send script support separately. Committer note: Do not use a fixed column width for the addresses, as kernel ones se more than 10 columns, look at the last offset and get the right width. Signed-off-by: Jiri Olsa <jolsa@kernel.org> Cc: Andi Kleen <andi@firstfloor.org> Cc: David Ahern <dsahern@gmail.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Link: http://lkml.kernel.org/r/20171011150158.11895-36-jolsa@kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2017-10-11 18:01:58 +03:00
static int disasm_line__print(struct disasm_line *dl, u64 start, int addr_fmt_width)
{
s64 offset = dl->al.offset;
const u64 addr = start + offset;
struct block_range *br;
br = block_range__find(addr);
perf annotate: Align source and offset lines Align source with offset lines, which are more advanced, because of the address column. Before: : static void *worker_thread(void *__tdata) : { 0.00 : 48a971: push %rbp 0.00 : 48a972: mov %rsp,%rbp 0.00 : 48a975: sub $0x30,%rsp 0.00 : 48a979: mov %rdi,-0x28(%rbp) 0.00 : 48a97d: mov %fs:0x28,%rax 0.00 : 48a986: mov %rax,-0x8(%rbp) 0.00 : 48a98a: xor %eax,%eax : struct thread_data *td = __tdata; 0.00 : 48a98c: mov -0x28(%rbp),%rax 0.00 : 48a990: mov %rax,-0x10(%rbp) : int m = 0, i; 0.00 : 48a994: movl $0x0,-0x1c(%rbp) : int ret; : : for (i = 0; i < loops; i++) { 0.00 : 48a99b: movl $0x0,-0x18(%rbp) After: : static void *worker_thread(void *__tdata) : { 0.00 : 48a971: push %rbp 0.00 : 48a972: mov %rsp,%rbp 0.00 : 48a975: sub $0x30,%rsp 0.00 : 48a979: mov %rdi,-0x28(%rbp) 0.00 : 48a97d: mov %fs:0x28,%rax 0.00 : 48a986: mov %rax,-0x8(%rbp) 0.00 : 48a98a: xor %eax,%eax : struct thread_data *td = __tdata; 0.00 : 48a98c: mov -0x28(%rbp),%rax 0.00 : 48a990: mov %rax,-0x10(%rbp) : int m = 0, i; 0.00 : 48a994: movl $0x0,-0x1c(%rbp) : int ret; : : for (i = 0; i < loops; i++) { 0.00 : 48a99b: movl $0x0,-0x18(%rbp) It makes bigger different when displaying script sources, where the comment lines looks oddly shifted from the lines which actually hold code. I'll send script support separately. Committer note: Do not use a fixed column width for the addresses, as kernel ones se more than 10 columns, look at the last offset and get the right width. Signed-off-by: Jiri Olsa <jolsa@kernel.org> Cc: Andi Kleen <andi@firstfloor.org> Cc: David Ahern <dsahern@gmail.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Link: http://lkml.kernel.org/r/20171011150158.11895-36-jolsa@kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2017-10-11 18:01:58 +03:00
color_fprintf(stdout, annotate__address_color(br), " %*" PRIx64 ":", addr_fmt_width, addr);
color_fprintf(stdout, annotate__asm_color(br), "%s", dl->al.line);
annotate__branch_printf(br, addr);
return 0;
}
static int
annotation_line__print(struct annotation_line *al, struct symbol *sym, u64 start,
struct evsel *evsel, u64 len, int min_pcnt, int printed,
int max_lines, struct annotation_line *queue, int addr_fmt_width,
int percent_type)
{
struct disasm_line *dl = container_of(al, struct disasm_line, al);
static const char *prev_line;
if (al->offset != -1) {
double max_percent = 0.0;
perf annotate: Add basic support to event group view Add --group option to enable event grouping. When enabled, all the group members information will be shown with the leader so skip non-leader events. It only supports --stdio output currently. Later patches will extend additional features. $ perf annotate --group --stdio ... Percent | Source code & Disassembly of libpthread-2.15.so -------------------------------------------------------------------------------- : : : : Disassembly of section .text: : : 000000387dc0aa50 <__pthread_mutex_unlock_usercnt>: 8.08 2.40 5.29 : 387dc0aa50: mov %rdi,%rdx 0.00 0.00 0.00 : 387dc0aa53: mov 0x10(%rdi),%edi 0.00 0.00 0.00 : 387dc0aa56: mov %edi,%eax 0.00 0.80 0.00 : 387dc0aa58: and $0x7f,%eax 3.03 2.40 3.53 : 387dc0aa5b: test $0x7c,%dil 0.00 0.00 0.00 : 387dc0aa5f: jne 387dc0aaa9 <__pthread_mutex_unlock_use 0.00 0.00 0.00 : 387dc0aa61: test %eax,%eax 0.00 0.00 0.00 : 387dc0aa63: jne 387dc0aa85 <__pthread_mutex_unlock_use 0.00 0.00 0.00 : 387dc0aa65: and $0x80,%edi 0.00 0.00 0.00 : 387dc0aa6b: test %esi,%esi 3.03 5.60 7.06 : 387dc0aa6d: movl $0x0,0x8(%rdx) 0.00 0.00 0.59 : 387dc0aa74: je 387dc0aa7a <__pthread_mutex_unlock_use 0.00 0.00 0.00 : 387dc0aa76: subl $0x1,0xc(%rdx) 2.02 5.60 1.18 : 387dc0aa7a: mov %edi,%esi 0.00 0.00 0.00 : 387dc0aa7c: lock decl (%rdx) 83.84 83.20 82.35 : 387dc0aa7f: jne 387dc0aada <_L_unlock_586> 0.00 0.00 0.00 : 387dc0aa81: nop 0.00 0.00 0.00 : 387dc0aa82: xor %eax,%eax 0.00 0.00 0.00 : 387dc0aa84: retq ... Signed-off-by: Namhyung Kim <namhyung@kernel.org> Cc: Andi Kleen <andi@firstfloor.org> Cc: Ingo Molnar <mingo@kernel.org> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Paul Mackerras <paulus@samba.org> Cc: Pekka Enberg <penberg@kernel.org> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Link: http://lkml.kernel.org/r/1362462812-30885-6-git-send-email-namhyung@kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2013-03-05 09:53:25 +04:00
int i, nr_percent = 1;
const char *color;
struct annotation *notes = symbol__annotation(sym);
for (i = 0; i < al->data_nr; i++) {
double percent;
percent = annotation_data__percent(&al->data[i],
percent_type);
if (percent > max_percent)
max_percent = percent;
perf annotate: Add basic support to event group view Add --group option to enable event grouping. When enabled, all the group members information will be shown with the leader so skip non-leader events. It only supports --stdio output currently. Later patches will extend additional features. $ perf annotate --group --stdio ... Percent | Source code & Disassembly of libpthread-2.15.so -------------------------------------------------------------------------------- : : : : Disassembly of section .text: : : 000000387dc0aa50 <__pthread_mutex_unlock_usercnt>: 8.08 2.40 5.29 : 387dc0aa50: mov %rdi,%rdx 0.00 0.00 0.00 : 387dc0aa53: mov 0x10(%rdi),%edi 0.00 0.00 0.00 : 387dc0aa56: mov %edi,%eax 0.00 0.80 0.00 : 387dc0aa58: and $0x7f,%eax 3.03 2.40 3.53 : 387dc0aa5b: test $0x7c,%dil 0.00 0.00 0.00 : 387dc0aa5f: jne 387dc0aaa9 <__pthread_mutex_unlock_use 0.00 0.00 0.00 : 387dc0aa61: test %eax,%eax 0.00 0.00 0.00 : 387dc0aa63: jne 387dc0aa85 <__pthread_mutex_unlock_use 0.00 0.00 0.00 : 387dc0aa65: and $0x80,%edi 0.00 0.00 0.00 : 387dc0aa6b: test %esi,%esi 3.03 5.60 7.06 : 387dc0aa6d: movl $0x0,0x8(%rdx) 0.00 0.00 0.59 : 387dc0aa74: je 387dc0aa7a <__pthread_mutex_unlock_use 0.00 0.00 0.00 : 387dc0aa76: subl $0x1,0xc(%rdx) 2.02 5.60 1.18 : 387dc0aa7a: mov %edi,%esi 0.00 0.00 0.00 : 387dc0aa7c: lock decl (%rdx) 83.84 83.20 82.35 : 387dc0aa7f: jne 387dc0aada <_L_unlock_586> 0.00 0.00 0.00 : 387dc0aa81: nop 0.00 0.00 0.00 : 387dc0aa82: xor %eax,%eax 0.00 0.00 0.00 : 387dc0aa84: retq ... Signed-off-by: Namhyung Kim <namhyung@kernel.org> Cc: Andi Kleen <andi@firstfloor.org> Cc: Ingo Molnar <mingo@kernel.org> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Paul Mackerras <paulus@samba.org> Cc: Pekka Enberg <penberg@kernel.org> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Link: http://lkml.kernel.org/r/1362462812-30885-6-git-send-email-namhyung@kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2013-03-05 09:53:25 +04:00
}
if (al->data_nr > nr_percent)
nr_percent = al->data_nr;
perf annotate: Display all available events on --stdio When we perform the following command lines: $ perf record -e "{cycles,branches}" ./div $ perf annotate main --stdio The output shows only the first event, "cycles" and the displaying format is not correct. Percent | Source code & Disassembly of div for cycles (44550 samples) ----------------------------------------------------------------------------------- : : : : Disassembly of section .text: : : 00000000004004b0 <main>: : main(): : : return i; : } : : int main(void) : { 0.00 : 4004b0: push %rbx : int i; : int flag; : volatile double x = 1212121212, y = 121212; : : s_randseed = time(0); 0.00 : 4004b1: xor %edi,%edi : srand(s_randseed); 0.00 : 4004b3: mov $0x77359400,%ebx : : return i; : } The issue is that the value of the 'nr_percent' variable is hardcoded to 1. This patch fixes it. With this patch, the output is: Percent | Source code & Disassembly of div for cycles (44550 samples) ----------------------------------------------------------------------------------- : : : : Disassembly of section .text: : : 00000000004004b0 <main>: : main(): : : return i; : } : : int main(void) : { 0.00 0.00 : 4004b0: push %rbx : int i; : int flag; : volatile double x = 1212121212, y = 121212; : : s_randseed = time(0); 0.00 0.00 : 4004b1: xor %edi,%edi : srand(s_randseed); 0.00 0.00 : 4004b3: mov $0x77359400,%ebx : : return i; : } Signed-off-by: Jin Yao <yao.jin@linux.intel.com> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Andi Kleen <ak@linux.intel.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Kan Liang <kan.liang@linux.intel.com> Cc: Peter Zijlstra <peterz@infradead.org> Fixes: f681d593d1ce ("perf annotate: Remove disasm__calc_percent() from disasm_line__print()") Link: http://lkml.kernel.org/r/1525881435-4092-1-git-send-email-yao.jin@linux.intel.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2018-05-09 18:57:15 +03:00
perf annotate: Add basic support to event group view Add --group option to enable event grouping. When enabled, all the group members information will be shown with the leader so skip non-leader events. It only supports --stdio output currently. Later patches will extend additional features. $ perf annotate --group --stdio ... Percent | Source code & Disassembly of libpthread-2.15.so -------------------------------------------------------------------------------- : : : : Disassembly of section .text: : : 000000387dc0aa50 <__pthread_mutex_unlock_usercnt>: 8.08 2.40 5.29 : 387dc0aa50: mov %rdi,%rdx 0.00 0.00 0.00 : 387dc0aa53: mov 0x10(%rdi),%edi 0.00 0.00 0.00 : 387dc0aa56: mov %edi,%eax 0.00 0.80 0.00 : 387dc0aa58: and $0x7f,%eax 3.03 2.40 3.53 : 387dc0aa5b: test $0x7c,%dil 0.00 0.00 0.00 : 387dc0aa5f: jne 387dc0aaa9 <__pthread_mutex_unlock_use 0.00 0.00 0.00 : 387dc0aa61: test %eax,%eax 0.00 0.00 0.00 : 387dc0aa63: jne 387dc0aa85 <__pthread_mutex_unlock_use 0.00 0.00 0.00 : 387dc0aa65: and $0x80,%edi 0.00 0.00 0.00 : 387dc0aa6b: test %esi,%esi 3.03 5.60 7.06 : 387dc0aa6d: movl $0x0,0x8(%rdx) 0.00 0.00 0.59 : 387dc0aa74: je 387dc0aa7a <__pthread_mutex_unlock_use 0.00 0.00 0.00 : 387dc0aa76: subl $0x1,0xc(%rdx) 2.02 5.60 1.18 : 387dc0aa7a: mov %edi,%esi 0.00 0.00 0.00 : 387dc0aa7c: lock decl (%rdx) 83.84 83.20 82.35 : 387dc0aa7f: jne 387dc0aada <_L_unlock_586> 0.00 0.00 0.00 : 387dc0aa81: nop 0.00 0.00 0.00 : 387dc0aa82: xor %eax,%eax 0.00 0.00 0.00 : 387dc0aa84: retq ... Signed-off-by: Namhyung Kim <namhyung@kernel.org> Cc: Andi Kleen <andi@firstfloor.org> Cc: Ingo Molnar <mingo@kernel.org> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Paul Mackerras <paulus@samba.org> Cc: Pekka Enberg <penberg@kernel.org> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Link: http://lkml.kernel.org/r/1362462812-30885-6-git-send-email-namhyung@kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2013-03-05 09:53:25 +04:00
if (max_percent < min_pcnt)
return -1;
if (max_lines && printed >= max_lines)
return 1;
if (queue != NULL) {
list_for_each_entry_from(queue, &notes->src->source, node) {
if (queue == al)
break;
annotation_line__print(queue, sym, start, evsel, len,
0, 0, 1, NULL, addr_fmt_width,
percent_type);
}
}
perf annotate: Add basic support to event group view Add --group option to enable event grouping. When enabled, all the group members information will be shown with the leader so skip non-leader events. It only supports --stdio output currently. Later patches will extend additional features. $ perf annotate --group --stdio ... Percent | Source code & Disassembly of libpthread-2.15.so -------------------------------------------------------------------------------- : : : : Disassembly of section .text: : : 000000387dc0aa50 <__pthread_mutex_unlock_usercnt>: 8.08 2.40 5.29 : 387dc0aa50: mov %rdi,%rdx 0.00 0.00 0.00 : 387dc0aa53: mov 0x10(%rdi),%edi 0.00 0.00 0.00 : 387dc0aa56: mov %edi,%eax 0.00 0.80 0.00 : 387dc0aa58: and $0x7f,%eax 3.03 2.40 3.53 : 387dc0aa5b: test $0x7c,%dil 0.00 0.00 0.00 : 387dc0aa5f: jne 387dc0aaa9 <__pthread_mutex_unlock_use 0.00 0.00 0.00 : 387dc0aa61: test %eax,%eax 0.00 0.00 0.00 : 387dc0aa63: jne 387dc0aa85 <__pthread_mutex_unlock_use 0.00 0.00 0.00 : 387dc0aa65: and $0x80,%edi 0.00 0.00 0.00 : 387dc0aa6b: test %esi,%esi 3.03 5.60 7.06 : 387dc0aa6d: movl $0x0,0x8(%rdx) 0.00 0.00 0.59 : 387dc0aa74: je 387dc0aa7a <__pthread_mutex_unlock_use 0.00 0.00 0.00 : 387dc0aa76: subl $0x1,0xc(%rdx) 2.02 5.60 1.18 : 387dc0aa7a: mov %edi,%esi 0.00 0.00 0.00 : 387dc0aa7c: lock decl (%rdx) 83.84 83.20 82.35 : 387dc0aa7f: jne 387dc0aada <_L_unlock_586> 0.00 0.00 0.00 : 387dc0aa81: nop 0.00 0.00 0.00 : 387dc0aa82: xor %eax,%eax 0.00 0.00 0.00 : 387dc0aa84: retq ... Signed-off-by: Namhyung Kim <namhyung@kernel.org> Cc: Andi Kleen <andi@firstfloor.org> Cc: Ingo Molnar <mingo@kernel.org> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Paul Mackerras <paulus@samba.org> Cc: Pekka Enberg <penberg@kernel.org> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Link: http://lkml.kernel.org/r/1362462812-30885-6-git-send-email-namhyung@kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2013-03-05 09:53:25 +04:00
color = get_percent_color(max_percent);
perf annotate: Add basic support to event group view Add --group option to enable event grouping. When enabled, all the group members information will be shown with the leader so skip non-leader events. It only supports --stdio output currently. Later patches will extend additional features. $ perf annotate --group --stdio ... Percent | Source code & Disassembly of libpthread-2.15.so -------------------------------------------------------------------------------- : : : : Disassembly of section .text: : : 000000387dc0aa50 <__pthread_mutex_unlock_usercnt>: 8.08 2.40 5.29 : 387dc0aa50: mov %rdi,%rdx 0.00 0.00 0.00 : 387dc0aa53: mov 0x10(%rdi),%edi 0.00 0.00 0.00 : 387dc0aa56: mov %edi,%eax 0.00 0.80 0.00 : 387dc0aa58: and $0x7f,%eax 3.03 2.40 3.53 : 387dc0aa5b: test $0x7c,%dil 0.00 0.00 0.00 : 387dc0aa5f: jne 387dc0aaa9 <__pthread_mutex_unlock_use 0.00 0.00 0.00 : 387dc0aa61: test %eax,%eax 0.00 0.00 0.00 : 387dc0aa63: jne 387dc0aa85 <__pthread_mutex_unlock_use 0.00 0.00 0.00 : 387dc0aa65: and $0x80,%edi 0.00 0.00 0.00 : 387dc0aa6b: test %esi,%esi 3.03 5.60 7.06 : 387dc0aa6d: movl $0x0,0x8(%rdx) 0.00 0.00 0.59 : 387dc0aa74: je 387dc0aa7a <__pthread_mutex_unlock_use 0.00 0.00 0.00 : 387dc0aa76: subl $0x1,0xc(%rdx) 2.02 5.60 1.18 : 387dc0aa7a: mov %edi,%esi 0.00 0.00 0.00 : 387dc0aa7c: lock decl (%rdx) 83.84 83.20 82.35 : 387dc0aa7f: jne 387dc0aada <_L_unlock_586> 0.00 0.00 0.00 : 387dc0aa81: nop 0.00 0.00 0.00 : 387dc0aa82: xor %eax,%eax 0.00 0.00 0.00 : 387dc0aa84: retq ... Signed-off-by: Namhyung Kim <namhyung@kernel.org> Cc: Andi Kleen <andi@firstfloor.org> Cc: Ingo Molnar <mingo@kernel.org> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Paul Mackerras <paulus@samba.org> Cc: Pekka Enberg <penberg@kernel.org> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Link: http://lkml.kernel.org/r/1362462812-30885-6-git-send-email-namhyung@kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2013-03-05 09:53:25 +04:00
for (i = 0; i < nr_percent; i++) {
struct annotation_data *data = &al->data[i];
double percent;
percent = annotation_data__percent(data, percent_type);
color = get_percent_color(percent);
if (symbol_conf.show_total_period)
color_fprintf(stdout, color, " %11" PRIu64,
data->he.period);
else if (symbol_conf.show_nr_samples)
color_fprintf(stdout, color, " %7" PRIu64,
data->he.nr_samples);
else
color_fprintf(stdout, color, " %7.2f", percent);
perf annotate: Add basic support to event group view Add --group option to enable event grouping. When enabled, all the group members information will be shown with the leader so skip non-leader events. It only supports --stdio output currently. Later patches will extend additional features. $ perf annotate --group --stdio ... Percent | Source code & Disassembly of libpthread-2.15.so -------------------------------------------------------------------------------- : : : : Disassembly of section .text: : : 000000387dc0aa50 <__pthread_mutex_unlock_usercnt>: 8.08 2.40 5.29 : 387dc0aa50: mov %rdi,%rdx 0.00 0.00 0.00 : 387dc0aa53: mov 0x10(%rdi),%edi 0.00 0.00 0.00 : 387dc0aa56: mov %edi,%eax 0.00 0.80 0.00 : 387dc0aa58: and $0x7f,%eax 3.03 2.40 3.53 : 387dc0aa5b: test $0x7c,%dil 0.00 0.00 0.00 : 387dc0aa5f: jne 387dc0aaa9 <__pthread_mutex_unlock_use 0.00 0.00 0.00 : 387dc0aa61: test %eax,%eax 0.00 0.00 0.00 : 387dc0aa63: jne 387dc0aa85 <__pthread_mutex_unlock_use 0.00 0.00 0.00 : 387dc0aa65: and $0x80,%edi 0.00 0.00 0.00 : 387dc0aa6b: test %esi,%esi 3.03 5.60 7.06 : 387dc0aa6d: movl $0x0,0x8(%rdx) 0.00 0.00 0.59 : 387dc0aa74: je 387dc0aa7a <__pthread_mutex_unlock_use 0.00 0.00 0.00 : 387dc0aa76: subl $0x1,0xc(%rdx) 2.02 5.60 1.18 : 387dc0aa7a: mov %edi,%esi 0.00 0.00 0.00 : 387dc0aa7c: lock decl (%rdx) 83.84 83.20 82.35 : 387dc0aa7f: jne 387dc0aada <_L_unlock_586> 0.00 0.00 0.00 : 387dc0aa81: nop 0.00 0.00 0.00 : 387dc0aa82: xor %eax,%eax 0.00 0.00 0.00 : 387dc0aa84: retq ... Signed-off-by: Namhyung Kim <namhyung@kernel.org> Cc: Andi Kleen <andi@firstfloor.org> Cc: Ingo Molnar <mingo@kernel.org> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Paul Mackerras <paulus@samba.org> Cc: Pekka Enberg <penberg@kernel.org> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Link: http://lkml.kernel.org/r/1362462812-30885-6-git-send-email-namhyung@kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2013-03-05 09:53:25 +04:00
}
perf annotate: Align source and offset lines Align source with offset lines, which are more advanced, because of the address column. Before: : static void *worker_thread(void *__tdata) : { 0.00 : 48a971: push %rbp 0.00 : 48a972: mov %rsp,%rbp 0.00 : 48a975: sub $0x30,%rsp 0.00 : 48a979: mov %rdi,-0x28(%rbp) 0.00 : 48a97d: mov %fs:0x28,%rax 0.00 : 48a986: mov %rax,-0x8(%rbp) 0.00 : 48a98a: xor %eax,%eax : struct thread_data *td = __tdata; 0.00 : 48a98c: mov -0x28(%rbp),%rax 0.00 : 48a990: mov %rax,-0x10(%rbp) : int m = 0, i; 0.00 : 48a994: movl $0x0,-0x1c(%rbp) : int ret; : : for (i = 0; i < loops; i++) { 0.00 : 48a99b: movl $0x0,-0x18(%rbp) After: : static void *worker_thread(void *__tdata) : { 0.00 : 48a971: push %rbp 0.00 : 48a972: mov %rsp,%rbp 0.00 : 48a975: sub $0x30,%rsp 0.00 : 48a979: mov %rdi,-0x28(%rbp) 0.00 : 48a97d: mov %fs:0x28,%rax 0.00 : 48a986: mov %rax,-0x8(%rbp) 0.00 : 48a98a: xor %eax,%eax : struct thread_data *td = __tdata; 0.00 : 48a98c: mov -0x28(%rbp),%rax 0.00 : 48a990: mov %rax,-0x10(%rbp) : int m = 0, i; 0.00 : 48a994: movl $0x0,-0x1c(%rbp) : int ret; : : for (i = 0; i < loops; i++) { 0.00 : 48a99b: movl $0x0,-0x18(%rbp) It makes bigger different when displaying script sources, where the comment lines looks oddly shifted from the lines which actually hold code. I'll send script support separately. Committer note: Do not use a fixed column width for the addresses, as kernel ones se more than 10 columns, look at the last offset and get the right width. Signed-off-by: Jiri Olsa <jolsa@kernel.org> Cc: Andi Kleen <andi@firstfloor.org> Cc: David Ahern <dsahern@gmail.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Link: http://lkml.kernel.org/r/20171011150158.11895-36-jolsa@kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2017-10-11 18:01:58 +03:00
printf(" : ");
perf annotate: Add branch stack / basic block I wanted to know the hottest path through a function and figured the branch-stack (LBR) information should be able to help out with that. The below uses the branch-stack to create basic blocks and generate statistics from them. from to branch_i * ----> * | | block v * ----> * from to branch_i+1 The blocks are broken down into non-overlapping ranges, while tracking if the start of each range is an entry point and/or the end of a range is a branch. Each block iterates all ranges it covers (while splitting where required to exactly match the block) and increments the 'coverage' count. For the range including the branch we increment the taken counter, as well as the pred counter if flags.predicted. Using these number we can find if an instruction: - had coverage; given by: br->coverage / br->sym->max_coverage This metric ensures each symbol has a 100% spot, which reflects the observation that each symbol must have a most covered/hottest block. - is a branch target: br->is_target && br->start == add - for targets, how much of a branch's coverages comes from it: target->entry / branch->coverage - is a branch: br->is_branch && br->end == addr - for branches, how often it was taken: br->taken / br->coverage after all, all execution that didn't take the branch would have incremented the coverage and continued onward to a later branch. - for branches, how often it was predicted: br->pred / br->taken The coverage percentage is used to color the address and asm sections; for low (<1%) coverage we use NORMAL (uncolored), indicating that these instructions are not 'important'. For high coverage (>75%) we color the address RED. For each branch, we add an asm comment after the instruction with information on how often it was taken and predicted. Output looks like (sans color, which does loose a lot of the information :/) $ perf record --branch-filter u,any -e cycles:p ./branches 27 $ perf annotate branches Percent | Source code & Disassembly of branches for cycles:pu (217 samples) --------------------------------------------------------------------------------- : branches(): 0.00 : 40057a: push %rbp 0.00 : 40057b: mov %rsp,%rbp 0.00 : 40057e: sub $0x20,%rsp 0.00 : 400582: mov %rdi,-0x18(%rbp) 0.00 : 400586: mov %rsi,-0x20(%rbp) 0.00 : 40058a: mov -0x18(%rbp),%rax 0.00 : 40058e: mov %rax,-0x10(%rbp) 0.00 : 400592: movq $0x0,-0x8(%rbp) 0.00 : 40059a: jmpq 400656 <branches+0xdc> 1.84 : 40059f: mov -0x10(%rbp),%rax # +100.00% 3.23 : 4005a3: and $0x1,%eax 1.84 : 4005a6: test %rax,%rax 0.00 : 4005a9: je 4005bf <branches+0x45> # -54.50% (p:42.00%) 0.46 : 4005ab: mov 0x200bbe(%rip),%rax # 601170 <acc> 12.90 : 4005b2: add $0x1,%rax 2.30 : 4005b6: mov %rax,0x200bb3(%rip) # 601170 <acc> 0.46 : 4005bd: jmp 4005d1 <branches+0x57> # -100.00% (p:100.00%) 0.92 : 4005bf: mov 0x200baa(%rip),%rax # 601170 <acc> # +49.54% 13.82 : 4005c6: sub $0x1,%rax 0.46 : 4005ca: mov %rax,0x200b9f(%rip) # 601170 <acc> 2.30 : 4005d1: mov -0x10(%rbp),%rax # +50.46% 0.46 : 4005d5: mov %rax,%rdi 0.46 : 4005d8: callq 400526 <lfsr> # -100.00% (p:100.00%) 0.00 : 4005dd: mov %rax,-0x10(%rbp) # +100.00% 0.92 : 4005e1: mov -0x18(%rbp),%rax 0.00 : 4005e5: and $0x1,%eax 0.00 : 4005e8: test %rax,%rax 0.00 : 4005eb: je 4005ff <branches+0x85> # -100.00% (p:100.00%) 0.00 : 4005ed: mov 0x200b7c(%rip),%rax # 601170 <acc> 0.00 : 4005f4: shr $0x2,%rax 0.00 : 4005f8: mov %rax,0x200b71(%rip) # 601170 <acc> 0.00 : 4005ff: mov -0x10(%rbp),%rax # +100.00% 7.37 : 400603: and $0x1,%eax 3.69 : 400606: test %rax,%rax 0.00 : 400609: jne 400612 <branches+0x98> # -59.25% (p:42.99%) 1.84 : 40060b: mov $0x1,%eax 14.29 : 400610: jmp 400617 <branches+0x9d> # -100.00% (p:100.00%) 1.38 : 400612: mov $0x0,%eax # +57.65% 10.14 : 400617: test %al,%al # +42.35% 0.00 : 400619: je 40062f <branches+0xb5> # -57.65% (p:100.00%) 0.46 : 40061b: mov 0x200b4e(%rip),%rax # 601170 <acc> 2.76 : 400622: sub $0x1,%rax 0.00 : 400626: mov %rax,0x200b43(%rip) # 601170 <acc> 0.46 : 40062d: jmp 400641 <branches+0xc7> # -100.00% (p:100.00%) 0.92 : 40062f: mov 0x200b3a(%rip),%rax # 601170 <acc> # +56.13% 2.30 : 400636: add $0x1,%rax 0.92 : 40063a: mov %rax,0x200b2f(%rip) # 601170 <acc> 0.92 : 400641: mov -0x10(%rbp),%rax # +43.87% 2.30 : 400645: mov %rax,%rdi 0.00 : 400648: callq 400526 <lfsr> # -100.00% (p:100.00%) 0.00 : 40064d: mov %rax,-0x10(%rbp) # +100.00% 1.84 : 400651: addq $0x1,-0x8(%rbp) 0.92 : 400656: mov -0x8(%rbp),%rax 5.07 : 40065a: cmp -0x20(%rbp),%rax 0.00 : 40065e: jb 40059f <branches+0x25> # -100.00% (p:100.00%) 0.00 : 400664: nop 0.00 : 400665: leaveq 0.00 : 400666: retq (Note: the --branch-filter u,any was used to avoid spurious target and branch points due to interrupts/faults, they show up as very small -/+ annotations on 'weird' locations) Committer note: Please take a look at: http://vger.kernel.org/~acme/perf/annotate_basic_blocks.png To see the colors. Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Andi Kleen <andi@firstfloor.org> Cc: Anshuman Khandual <khandual@linux.vnet.ibm.com> Cc: David Carrillo-Cisneros <davidcc@google.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Kan Liang <kan.liang@intel.com> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Stephane Eranian <eranian@google.com> [ Moved sym->max_coverage to 'struct annotate', aka symbol__annotate(sym) ] Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-09-05 22:08:12 +03:00
perf annotate: Align source and offset lines Align source with offset lines, which are more advanced, because of the address column. Before: : static void *worker_thread(void *__tdata) : { 0.00 : 48a971: push %rbp 0.00 : 48a972: mov %rsp,%rbp 0.00 : 48a975: sub $0x30,%rsp 0.00 : 48a979: mov %rdi,-0x28(%rbp) 0.00 : 48a97d: mov %fs:0x28,%rax 0.00 : 48a986: mov %rax,-0x8(%rbp) 0.00 : 48a98a: xor %eax,%eax : struct thread_data *td = __tdata; 0.00 : 48a98c: mov -0x28(%rbp),%rax 0.00 : 48a990: mov %rax,-0x10(%rbp) : int m = 0, i; 0.00 : 48a994: movl $0x0,-0x1c(%rbp) : int ret; : : for (i = 0; i < loops; i++) { 0.00 : 48a99b: movl $0x0,-0x18(%rbp) After: : static void *worker_thread(void *__tdata) : { 0.00 : 48a971: push %rbp 0.00 : 48a972: mov %rsp,%rbp 0.00 : 48a975: sub $0x30,%rsp 0.00 : 48a979: mov %rdi,-0x28(%rbp) 0.00 : 48a97d: mov %fs:0x28,%rax 0.00 : 48a986: mov %rax,-0x8(%rbp) 0.00 : 48a98a: xor %eax,%eax : struct thread_data *td = __tdata; 0.00 : 48a98c: mov -0x28(%rbp),%rax 0.00 : 48a990: mov %rax,-0x10(%rbp) : int m = 0, i; 0.00 : 48a994: movl $0x0,-0x1c(%rbp) : int ret; : : for (i = 0; i < loops; i++) { 0.00 : 48a99b: movl $0x0,-0x18(%rbp) It makes bigger different when displaying script sources, where the comment lines looks oddly shifted from the lines which actually hold code. I'll send script support separately. Committer note: Do not use a fixed column width for the addresses, as kernel ones se more than 10 columns, look at the last offset and get the right width. Signed-off-by: Jiri Olsa <jolsa@kernel.org> Cc: Andi Kleen <andi@firstfloor.org> Cc: David Ahern <dsahern@gmail.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Link: http://lkml.kernel.org/r/20171011150158.11895-36-jolsa@kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2017-10-11 18:01:58 +03:00
disasm_line__print(dl, start, addr_fmt_width);
perf annotate: Add line number like in TUI and source location at EOL The patch changes the output format in 2 ways: - line number is displayed for all source lines (matching TUI mode) - source locations for the hottest lines are printed at the line end in order to preserve layout Before: 0.00 : 405ef1: inc %r15 : tmpsd * (TD + tmpsd * TDD))); 0.01 : 405ef4: vfmadd213sd 0x2b9b3(%rip),%xmm0,%xmm3 # 4318b0 <_IO_stdin_used+0x8b0> : tmpsd * (TC + eff.c:1811 0.67 : 405efd: vfmadd213sd 0x2b9b2(%rip),%xmm0,%xmm3 # 4318b8 <_IO_stdin_used+0x8b8> : TA + tmpsd * (TB + 0.35 : 405f06: vfmadd213sd 0x2b9b1(%rip),%xmm0,%xmm3 # 4318c0 <_IO_stdin_used+0x8c0> : dumbo = eff.c:1809 1.41 : 405f0f: vfmadd213sd 0x2b9b0(%rip),%xmm0,%xmm3 # 4318c8 <_IO_stdin_used+0x8c8> : sumi -= sj * tmpsd * dij2i * dumbo; eff.c:1813 2.58 : 405f18: vmulsd %xmm3,%xmm0,%xmm0 2.81 : 405f1c: vfnmadd213sd 0x30(%rsp),%xmm1,%xmm0 3.78 : 405f23: vmovsd %xmm0,0x30(%rsp) : for (k = 0; k < lpears[i] + upears[i]; k++) { eff.c:1761 0.90 : 405f29: cmp %r15d,%r12d After: 0.00 : 405ef1: inc %r15 : 1812 tmpsd * (TD + tmpsd * TDD))); 0.01 : 405ef4: vfmadd213sd 0x2b9b3(%rip),%xmm0,%xmm3 # 4318b0 <_IO_stdin_used+0x8b0> : 1811 tmpsd * (TC + 0.67 : 405efd: vfmadd213sd 0x2b9b2(%rip),%xmm0,%xmm3 # 4318b8 <_IO_stdin_used+0x8b8> // eff.c:1811 : 1810 TA + tmpsd * (TB + 0.35 : 405f06: vfmadd213sd 0x2b9b1(%rip),%xmm0,%xmm3 # 4318c0 <_IO_stdin_used+0x8c0> : 1809 dumbo = 1.41 : 405f0f: vfmadd213sd 0x2b9b0(%rip),%xmm0,%xmm3 # 4318c8 <_IO_stdin_used+0x8c8> // eff.c:1809 : 1813 sumi -= sj * tmpsd * dij2i * dumbo; 2.58 : 405f18: vmulsd %xmm3,%xmm0,%xmm0 // eff.c:1813 2.81 : 405f1c: vfnmadd213sd 0x30(%rsp),%xmm1,%xmm0 3.78 : 405f23: vmovsd %xmm0,0x30(%rsp) : 1761 for (k = 0; k < lpears[i] + upears[i]; k++) { Where e.g. '// eff.c:1811' shares the same color as the percentantage at the line beginning. Signed-off-by: Martin Liška <mliska@suse.cz> Link: http://lore.kernel.org/lkml/a0d53f31-f633-5013-c386-a4452391b081@suse.cz Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2021-02-21 15:46:36 +03:00
/*
* Also color the filename and line if needed, with
* the same color than the percentage. Don't print it
* twice for close colored addr with the same filename:line
*/
if (al->path) {
if (!prev_line || strcmp(prev_line, al->path)) {
color_fprintf(stdout, color, " // %s", al->path);
prev_line = al->path;
}
}
perf annotate: Add branch stack / basic block I wanted to know the hottest path through a function and figured the branch-stack (LBR) information should be able to help out with that. The below uses the branch-stack to create basic blocks and generate statistics from them. from to branch_i * ----> * | | block v * ----> * from to branch_i+1 The blocks are broken down into non-overlapping ranges, while tracking if the start of each range is an entry point and/or the end of a range is a branch. Each block iterates all ranges it covers (while splitting where required to exactly match the block) and increments the 'coverage' count. For the range including the branch we increment the taken counter, as well as the pred counter if flags.predicted. Using these number we can find if an instruction: - had coverage; given by: br->coverage / br->sym->max_coverage This metric ensures each symbol has a 100% spot, which reflects the observation that each symbol must have a most covered/hottest block. - is a branch target: br->is_target && br->start == add - for targets, how much of a branch's coverages comes from it: target->entry / branch->coverage - is a branch: br->is_branch && br->end == addr - for branches, how often it was taken: br->taken / br->coverage after all, all execution that didn't take the branch would have incremented the coverage and continued onward to a later branch. - for branches, how often it was predicted: br->pred / br->taken The coverage percentage is used to color the address and asm sections; for low (<1%) coverage we use NORMAL (uncolored), indicating that these instructions are not 'important'. For high coverage (>75%) we color the address RED. For each branch, we add an asm comment after the instruction with information on how often it was taken and predicted. Output looks like (sans color, which does loose a lot of the information :/) $ perf record --branch-filter u,any -e cycles:p ./branches 27 $ perf annotate branches Percent | Source code & Disassembly of branches for cycles:pu (217 samples) --------------------------------------------------------------------------------- : branches(): 0.00 : 40057a: push %rbp 0.00 : 40057b: mov %rsp,%rbp 0.00 : 40057e: sub $0x20,%rsp 0.00 : 400582: mov %rdi,-0x18(%rbp) 0.00 : 400586: mov %rsi,-0x20(%rbp) 0.00 : 40058a: mov -0x18(%rbp),%rax 0.00 : 40058e: mov %rax,-0x10(%rbp) 0.00 : 400592: movq $0x0,-0x8(%rbp) 0.00 : 40059a: jmpq 400656 <branches+0xdc> 1.84 : 40059f: mov -0x10(%rbp),%rax # +100.00% 3.23 : 4005a3: and $0x1,%eax 1.84 : 4005a6: test %rax,%rax 0.00 : 4005a9: je 4005bf <branches+0x45> # -54.50% (p:42.00%) 0.46 : 4005ab: mov 0x200bbe(%rip),%rax # 601170 <acc> 12.90 : 4005b2: add $0x1,%rax 2.30 : 4005b6: mov %rax,0x200bb3(%rip) # 601170 <acc> 0.46 : 4005bd: jmp 4005d1 <branches+0x57> # -100.00% (p:100.00%) 0.92 : 4005bf: mov 0x200baa(%rip),%rax # 601170 <acc> # +49.54% 13.82 : 4005c6: sub $0x1,%rax 0.46 : 4005ca: mov %rax,0x200b9f(%rip) # 601170 <acc> 2.30 : 4005d1: mov -0x10(%rbp),%rax # +50.46% 0.46 : 4005d5: mov %rax,%rdi 0.46 : 4005d8: callq 400526 <lfsr> # -100.00% (p:100.00%) 0.00 : 4005dd: mov %rax,-0x10(%rbp) # +100.00% 0.92 : 4005e1: mov -0x18(%rbp),%rax 0.00 : 4005e5: and $0x1,%eax 0.00 : 4005e8: test %rax,%rax 0.00 : 4005eb: je 4005ff <branches+0x85> # -100.00% (p:100.00%) 0.00 : 4005ed: mov 0x200b7c(%rip),%rax # 601170 <acc> 0.00 : 4005f4: shr $0x2,%rax 0.00 : 4005f8: mov %rax,0x200b71(%rip) # 601170 <acc> 0.00 : 4005ff: mov -0x10(%rbp),%rax # +100.00% 7.37 : 400603: and $0x1,%eax 3.69 : 400606: test %rax,%rax 0.00 : 400609: jne 400612 <branches+0x98> # -59.25% (p:42.99%) 1.84 : 40060b: mov $0x1,%eax 14.29 : 400610: jmp 400617 <branches+0x9d> # -100.00% (p:100.00%) 1.38 : 400612: mov $0x0,%eax # +57.65% 10.14 : 400617: test %al,%al # +42.35% 0.00 : 400619: je 40062f <branches+0xb5> # -57.65% (p:100.00%) 0.46 : 40061b: mov 0x200b4e(%rip),%rax # 601170 <acc> 2.76 : 400622: sub $0x1,%rax 0.00 : 400626: mov %rax,0x200b43(%rip) # 601170 <acc> 0.46 : 40062d: jmp 400641 <branches+0xc7> # -100.00% (p:100.00%) 0.92 : 40062f: mov 0x200b3a(%rip),%rax # 601170 <acc> # +56.13% 2.30 : 400636: add $0x1,%rax 0.92 : 40063a: mov %rax,0x200b2f(%rip) # 601170 <acc> 0.92 : 400641: mov -0x10(%rbp),%rax # +43.87% 2.30 : 400645: mov %rax,%rdi 0.00 : 400648: callq 400526 <lfsr> # -100.00% (p:100.00%) 0.00 : 40064d: mov %rax,-0x10(%rbp) # +100.00% 1.84 : 400651: addq $0x1,-0x8(%rbp) 0.92 : 400656: mov -0x8(%rbp),%rax 5.07 : 40065a: cmp -0x20(%rbp),%rax 0.00 : 40065e: jb 40059f <branches+0x25> # -100.00% (p:100.00%) 0.00 : 400664: nop 0.00 : 400665: leaveq 0.00 : 400666: retq (Note: the --branch-filter u,any was used to avoid spurious target and branch points due to interrupts/faults, they show up as very small -/+ annotations on 'weird' locations) Committer note: Please take a look at: http://vger.kernel.org/~acme/perf/annotate_basic_blocks.png To see the colors. Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Andi Kleen <andi@firstfloor.org> Cc: Anshuman Khandual <khandual@linux.vnet.ibm.com> Cc: David Carrillo-Cisneros <davidcc@google.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Kan Liang <kan.liang@intel.com> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Stephane Eranian <eranian@google.com> [ Moved sym->max_coverage to 'struct annotate', aka symbol__annotate(sym) ] Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-09-05 22:08:12 +03:00
printf("\n");
} else if (max_lines && printed >= max_lines)
return 1;
else {
int width = symbol_conf.show_total_period ? 12 : 8;
perf annotate: Add basic support to event group view Add --group option to enable event grouping. When enabled, all the group members information will be shown with the leader so skip non-leader events. It only supports --stdio output currently. Later patches will extend additional features. $ perf annotate --group --stdio ... Percent | Source code & Disassembly of libpthread-2.15.so -------------------------------------------------------------------------------- : : : : Disassembly of section .text: : : 000000387dc0aa50 <__pthread_mutex_unlock_usercnt>: 8.08 2.40 5.29 : 387dc0aa50: mov %rdi,%rdx 0.00 0.00 0.00 : 387dc0aa53: mov 0x10(%rdi),%edi 0.00 0.00 0.00 : 387dc0aa56: mov %edi,%eax 0.00 0.80 0.00 : 387dc0aa58: and $0x7f,%eax 3.03 2.40 3.53 : 387dc0aa5b: test $0x7c,%dil 0.00 0.00 0.00 : 387dc0aa5f: jne 387dc0aaa9 <__pthread_mutex_unlock_use 0.00 0.00 0.00 : 387dc0aa61: test %eax,%eax 0.00 0.00 0.00 : 387dc0aa63: jne 387dc0aa85 <__pthread_mutex_unlock_use 0.00 0.00 0.00 : 387dc0aa65: and $0x80,%edi 0.00 0.00 0.00 : 387dc0aa6b: test %esi,%esi 3.03 5.60 7.06 : 387dc0aa6d: movl $0x0,0x8(%rdx) 0.00 0.00 0.59 : 387dc0aa74: je 387dc0aa7a <__pthread_mutex_unlock_use 0.00 0.00 0.00 : 387dc0aa76: subl $0x1,0xc(%rdx) 2.02 5.60 1.18 : 387dc0aa7a: mov %edi,%esi 0.00 0.00 0.00 : 387dc0aa7c: lock decl (%rdx) 83.84 83.20 82.35 : 387dc0aa7f: jne 387dc0aada <_L_unlock_586> 0.00 0.00 0.00 : 387dc0aa81: nop 0.00 0.00 0.00 : 387dc0aa82: xor %eax,%eax 0.00 0.00 0.00 : 387dc0aa84: retq ... Signed-off-by: Namhyung Kim <namhyung@kernel.org> Cc: Andi Kleen <andi@firstfloor.org> Cc: Ingo Molnar <mingo@kernel.org> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Paul Mackerras <paulus@samba.org> Cc: Pekka Enberg <penberg@kernel.org> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Link: http://lkml.kernel.org/r/1362462812-30885-6-git-send-email-namhyung@kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2013-03-05 09:53:25 +04:00
if (queue)
return -1;
if (evsel__is_group_event(evsel))
width *= evsel->core.nr_members;
perf annotate: Add basic support to event group view Add --group option to enable event grouping. When enabled, all the group members information will be shown with the leader so skip non-leader events. It only supports --stdio output currently. Later patches will extend additional features. $ perf annotate --group --stdio ... Percent | Source code & Disassembly of libpthread-2.15.so -------------------------------------------------------------------------------- : : : : Disassembly of section .text: : : 000000387dc0aa50 <__pthread_mutex_unlock_usercnt>: 8.08 2.40 5.29 : 387dc0aa50: mov %rdi,%rdx 0.00 0.00 0.00 : 387dc0aa53: mov 0x10(%rdi),%edi 0.00 0.00 0.00 : 387dc0aa56: mov %edi,%eax 0.00 0.80 0.00 : 387dc0aa58: and $0x7f,%eax 3.03 2.40 3.53 : 387dc0aa5b: test $0x7c,%dil 0.00 0.00 0.00 : 387dc0aa5f: jne 387dc0aaa9 <__pthread_mutex_unlock_use 0.00 0.00 0.00 : 387dc0aa61: test %eax,%eax 0.00 0.00 0.00 : 387dc0aa63: jne 387dc0aa85 <__pthread_mutex_unlock_use 0.00 0.00 0.00 : 387dc0aa65: and $0x80,%edi 0.00 0.00 0.00 : 387dc0aa6b: test %esi,%esi 3.03 5.60 7.06 : 387dc0aa6d: movl $0x0,0x8(%rdx) 0.00 0.00 0.59 : 387dc0aa74: je 387dc0aa7a <__pthread_mutex_unlock_use 0.00 0.00 0.00 : 387dc0aa76: subl $0x1,0xc(%rdx) 2.02 5.60 1.18 : 387dc0aa7a: mov %edi,%esi 0.00 0.00 0.00 : 387dc0aa7c: lock decl (%rdx) 83.84 83.20 82.35 : 387dc0aa7f: jne 387dc0aada <_L_unlock_586> 0.00 0.00 0.00 : 387dc0aa81: nop 0.00 0.00 0.00 : 387dc0aa82: xor %eax,%eax 0.00 0.00 0.00 : 387dc0aa84: retq ... Signed-off-by: Namhyung Kim <namhyung@kernel.org> Cc: Andi Kleen <andi@firstfloor.org> Cc: Ingo Molnar <mingo@kernel.org> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Paul Mackerras <paulus@samba.org> Cc: Pekka Enberg <penberg@kernel.org> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Link: http://lkml.kernel.org/r/1362462812-30885-6-git-send-email-namhyung@kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2013-03-05 09:53:25 +04:00
if (!*al->line)
perf annotate: Add basic support to event group view Add --group option to enable event grouping. When enabled, all the group members information will be shown with the leader so skip non-leader events. It only supports --stdio output currently. Later patches will extend additional features. $ perf annotate --group --stdio ... Percent | Source code & Disassembly of libpthread-2.15.so -------------------------------------------------------------------------------- : : : : Disassembly of section .text: : : 000000387dc0aa50 <__pthread_mutex_unlock_usercnt>: 8.08 2.40 5.29 : 387dc0aa50: mov %rdi,%rdx 0.00 0.00 0.00 : 387dc0aa53: mov 0x10(%rdi),%edi 0.00 0.00 0.00 : 387dc0aa56: mov %edi,%eax 0.00 0.80 0.00 : 387dc0aa58: and $0x7f,%eax 3.03 2.40 3.53 : 387dc0aa5b: test $0x7c,%dil 0.00 0.00 0.00 : 387dc0aa5f: jne 387dc0aaa9 <__pthread_mutex_unlock_use 0.00 0.00 0.00 : 387dc0aa61: test %eax,%eax 0.00 0.00 0.00 : 387dc0aa63: jne 387dc0aa85 <__pthread_mutex_unlock_use 0.00 0.00 0.00 : 387dc0aa65: and $0x80,%edi 0.00 0.00 0.00 : 387dc0aa6b: test %esi,%esi 3.03 5.60 7.06 : 387dc0aa6d: movl $0x0,0x8(%rdx) 0.00 0.00 0.59 : 387dc0aa74: je 387dc0aa7a <__pthread_mutex_unlock_use 0.00 0.00 0.00 : 387dc0aa76: subl $0x1,0xc(%rdx) 2.02 5.60 1.18 : 387dc0aa7a: mov %edi,%esi 0.00 0.00 0.00 : 387dc0aa7c: lock decl (%rdx) 83.84 83.20 82.35 : 387dc0aa7f: jne 387dc0aada <_L_unlock_586> 0.00 0.00 0.00 : 387dc0aa81: nop 0.00 0.00 0.00 : 387dc0aa82: xor %eax,%eax 0.00 0.00 0.00 : 387dc0aa84: retq ... Signed-off-by: Namhyung Kim <namhyung@kernel.org> Cc: Andi Kleen <andi@firstfloor.org> Cc: Ingo Molnar <mingo@kernel.org> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Paul Mackerras <paulus@samba.org> Cc: Pekka Enberg <penberg@kernel.org> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Link: http://lkml.kernel.org/r/1362462812-30885-6-git-send-email-namhyung@kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2013-03-05 09:53:25 +04:00
printf(" %*s:\n", width, " ");
else
perf annotate: Add line number like in TUI and source location at EOL The patch changes the output format in 2 ways: - line number is displayed for all source lines (matching TUI mode) - source locations for the hottest lines are printed at the line end in order to preserve layout Before: 0.00 : 405ef1: inc %r15 : tmpsd * (TD + tmpsd * TDD))); 0.01 : 405ef4: vfmadd213sd 0x2b9b3(%rip),%xmm0,%xmm3 # 4318b0 <_IO_stdin_used+0x8b0> : tmpsd * (TC + eff.c:1811 0.67 : 405efd: vfmadd213sd 0x2b9b2(%rip),%xmm0,%xmm3 # 4318b8 <_IO_stdin_used+0x8b8> : TA + tmpsd * (TB + 0.35 : 405f06: vfmadd213sd 0x2b9b1(%rip),%xmm0,%xmm3 # 4318c0 <_IO_stdin_used+0x8c0> : dumbo = eff.c:1809 1.41 : 405f0f: vfmadd213sd 0x2b9b0(%rip),%xmm0,%xmm3 # 4318c8 <_IO_stdin_used+0x8c8> : sumi -= sj * tmpsd * dij2i * dumbo; eff.c:1813 2.58 : 405f18: vmulsd %xmm3,%xmm0,%xmm0 2.81 : 405f1c: vfnmadd213sd 0x30(%rsp),%xmm1,%xmm0 3.78 : 405f23: vmovsd %xmm0,0x30(%rsp) : for (k = 0; k < lpears[i] + upears[i]; k++) { eff.c:1761 0.90 : 405f29: cmp %r15d,%r12d After: 0.00 : 405ef1: inc %r15 : 1812 tmpsd * (TD + tmpsd * TDD))); 0.01 : 405ef4: vfmadd213sd 0x2b9b3(%rip),%xmm0,%xmm3 # 4318b0 <_IO_stdin_used+0x8b0> : 1811 tmpsd * (TC + 0.67 : 405efd: vfmadd213sd 0x2b9b2(%rip),%xmm0,%xmm3 # 4318b8 <_IO_stdin_used+0x8b8> // eff.c:1811 : 1810 TA + tmpsd * (TB + 0.35 : 405f06: vfmadd213sd 0x2b9b1(%rip),%xmm0,%xmm3 # 4318c0 <_IO_stdin_used+0x8c0> : 1809 dumbo = 1.41 : 405f0f: vfmadd213sd 0x2b9b0(%rip),%xmm0,%xmm3 # 4318c8 <_IO_stdin_used+0x8c8> // eff.c:1809 : 1813 sumi -= sj * tmpsd * dij2i * dumbo; 2.58 : 405f18: vmulsd %xmm3,%xmm0,%xmm0 // eff.c:1813 2.81 : 405f1c: vfnmadd213sd 0x30(%rsp),%xmm1,%xmm0 3.78 : 405f23: vmovsd %xmm0,0x30(%rsp) : 1761 for (k = 0; k < lpears[i] + upears[i]; k++) { Where e.g. '// eff.c:1811' shares the same color as the percentantage at the line beginning. Signed-off-by: Martin Liška <mliska@suse.cz> Link: http://lore.kernel.org/lkml/a0d53f31-f633-5013-c386-a4452391b081@suse.cz Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2021-02-21 15:46:36 +03:00
printf(" %*s: %-*d %s\n", width, " ", addr_fmt_width, al->line_nr, al->line);
}
return 0;
}
/*
* symbol__parse_objdump_line() parses objdump output (with -d --no-show-raw)
* which looks like following
*
* 0000000000415500 <_init>:
* 415500: sub $0x8,%rsp
* 415504: mov 0x2f5ad5(%rip),%rax # 70afe0 <_DYNAMIC+0x2f8>
* 41550b: test %rax,%rax
* 41550e: je 415515 <_init+0x15>
* 415510: callq 416e70 <__gmon_start__@plt>
* 415515: add $0x8,%rsp
* 415519: retq
*
* it will be parsed and saved into struct disasm_line as
* <offset> <name> <ops.raw>
*
* The offset will be a relative offset from the start of the symbol and -1
* means that it's not a disassembly line so should be treated differently.
* The ops.raw part will be parsed further according to type of the instruction.
*/
static int symbol__parse_objdump_line(struct symbol *sym,
struct annotate_args *args,
char *parsed_line, int *line_nr, char **fileloc)
{
struct map *map = args->ms.map;
struct annotation *notes = symbol__annotation(sym);
struct disasm_line *dl;
char *tmp;
s64 line_ip, offset = -1;
regmatch_t match[2];
/* /filename:linenr ? Save line number and ignore. */
if (regexec(&file_lineno, parsed_line, 2, match, 0) == 0) {
*line_nr = atoi(parsed_line + match[1].rm_so);
*fileloc = strdup(parsed_line);
return 0;
}
/* Process hex address followed by ':'. */
line_ip = strtoull(parsed_line, &tmp, 16);
if (parsed_line != tmp && tmp[0] == ':' && tmp[1] != '\0') {
u64 start = map__rip_2objdump(map, sym->start),
end = map__rip_2objdump(map, sym->end);
offset = line_ip - start;
if ((u64)line_ip < start || (u64)line_ip >= end)
offset = -1;
else
parsed_line = tmp + 1;
}
args->offset = offset;
args->line = parsed_line;
args->line_nr = *line_nr;
args->fileloc = *fileloc;
args->ms.sym = sym;
dl = disasm_line__new(args);
(*line_nr)++;
if (dl == NULL)
return -1;
if (!disasm_line__has_local_offset(dl)) {
dl->ops.target.offset = dl->ops.target.addr -
map__rip_2objdump(map, sym->start);
perf annotate: Fix jump target outside of function address range If jump target is outside of function range, perf is not handling it correctly. Especially when target address is lesser than function start address, target offset will be negative. But, target address declared to be unsigned, converts negative number into 2's complement. See below example. Here target of 'jumpq' instruction at 34cf8 is 34ac0 which is lesser than function start address(34cf0). 34ac0 - 34cf0 = -0x230 = 0xfffffffffffffdd0 Objdump output: 0000000000034cf0 <__sigaction>: __GI___sigaction(): 34cf0: lea -0x20(%rdi),%eax 34cf3: cmp -bashx1,%eax 34cf6: jbe 34d00 <__sigaction+0x10> 34cf8: jmpq 34ac0 <__GI___libc_sigaction> 34cfd: nopl (%rax) 34d00: mov 0x386161(%rip),%rax # 3bae68 <_DYNAMIC+0x2e8> 34d07: movl -bashx16,%fs:(%rax) 34d0e: mov -bashxffffffff,%eax 34d13: retq perf annotate before applying patch: __GI___sigaction /usr/lib64/libc-2.22.so lea -0x20(%rdi),%eax cmp -bashx1,%eax v jbe 10 v jmpq fffffffffffffdd0 nop 10: mov _DYNAMIC+0x2e8,%rax movl -bashx16,%fs:(%rax) mov -bashxffffffff,%eax retq perf annotate after applying patch: __GI___sigaction /usr/lib64/libc-2.22.so lea -0x20(%rdi),%eax cmp -bashx1,%eax v jbe 10 ^ jmpq 34ac0 <__GI___libc_sigaction> nop 10: mov _DYNAMIC+0x2e8,%rax movl -bashx16,%fs:(%rax) mov -bashxffffffff,%eax retq Signed-off-by: Ravi Bangoria <ravi.bangoria@linux.vnet.ibm.com> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Chris Riyder <chris.ryder@arm.com> Cc: Kim Phillips <kim.phillips@arm.com> Cc: Markus Trippelsdorf <markus@trippelsdorf.de> Cc: Masami Hiramatsu <mhiramat@kernel.org> Cc: Naveen N. Rao <naveen.n.rao@linux.vnet.ibm.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Taeung Song <treeze.taeung@gmail.com> Cc: linuxppc-dev@lists.ozlabs.org Link: http://lkml.kernel.org/r/1480953407-7605-3-git-send-email-ravi.bangoria@linux.vnet.ibm.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-12-05 18:56:47 +03:00
dl->ops.target.offset_avail = true;
}
/* kcore has no symbols, so add the call target symbol */
if (dl->ins.ops && ins__is_call(&dl->ins) && !dl->ops.target.sym) {
struct addr_map_symbol target = {
.addr = dl->ops.target.addr,
.ms = { .map = map, },
};
if (!maps__find_ams(args->ms.maps, &target) &&
target.ms.sym->start == target.al_addr)
dl->ops.target.sym = target.ms.sym;
}
annotation_line__add(&dl->al, &notes->src->source);
return 0;
}
static __attribute__((constructor)) void symbol__init_regexpr(void)
{
regcomp(&file_lineno, "^/[^:]+:([0-9]+)", REG_EXTENDED);
}
static void delete_last_nop(struct symbol *sym)
{
struct annotation *notes = symbol__annotation(sym);
struct list_head *list = &notes->src->source;
struct disasm_line *dl;
while (!list_empty(list)) {
dl = list_entry(list->prev, struct disasm_line, al.node);
perf annotate: Remove duplicate 'name' field from disasm_line The disasm_line::name field is always equal to ins::name, being used just to locate the instruction's ins_ops from the per-arch instructions table. Eliminate this duplication, nuking that field and instead make ins__find() return an ins_ops, store it in disasm_line::ins.ops, and keep just in disasm_line::ins.name what was in disasm_line::name, this way we end up not keeping a reference to entries in the per-arch instructions table. This in turn will help supporting multiple ways to manage the per-arch instructions table, allowing resorting that array, for instance, when the entries will move after references to its addresses were made. The same problem is avoided when one grows the array with realloc. So architectures simply keeping a constant array will work as well as architectures building the table using regular expressions or other logic that involves resorting the table. Reviewed-by: Ravi Bangoria <ravi.bangoria@linux.vnet.ibm.com> Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Chris Riyder <chris.ryder@arm.com> Cc: David Ahern <dsahern@gmail.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Kim Phillips <kim.phillips@arm.com> Cc: Markus Trippelsdorf <markus@trippelsdorf.de> Cc: Masami Hiramatsu <mhiramat@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Naveen N. Rao <naveen.n.rao@linux.vnet.ibm.com> Cc: Pawel Moll <pawel.moll@arm.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Russell King <rmk+kernel@arm.linux.org.uk> Cc: Taeung Song <treeze.taeung@gmail.com> Cc: Wang Nan <wangnan0@huawei.com> Link: http://lkml.kernel.org/n/tip-vr899azvabnw9gtuepuqfd9t@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2016-11-24 17:16:06 +03:00
if (dl->ins.ops) {
if (dl->ins.ops != &nop_ops)
return;
} else {
if (!strstr(dl->al.line, " nop ") &&
!strstr(dl->al.line, " nopl ") &&
!strstr(dl->al.line, " nopw "))
return;
}
list_del_init(&dl->al.node);
disasm_line__free(dl);
}
}
int symbol__strerror_disassemble(struct map_symbol *ms, int errnum, char *buf, size_t buflen)
{
struct dso *dso = ms->map->dso;
BUG_ON(buflen == 0);
if (errnum >= 0) {
str_error_r(errnum, buf, buflen);
return 0;
}
switch (errnum) {
case SYMBOL_ANNOTATE_ERRNO__NO_VMLINUX: {
char bf[SBUILD_ID_SIZE + 15] = " with build id ";
char *build_id_msg = NULL;
if (dso->has_build_id) {
build_id__sprintf(&dso->bid, bf + 15);
build_id_msg = bf;
}
scnprintf(buf, buflen,
"No vmlinux file%s\nwas found in the path.\n\n"
"Note that annotation using /proc/kcore requires CAP_SYS_RAWIO capability.\n\n"
"Please use:\n\n"
" perf buildid-cache -vu vmlinux\n\n"
"or:\n\n"
" --vmlinux vmlinux\n", build_id_msg ?: "");
}
break;
perf annotate: Enable annotation of BPF programs In symbol__disassemble(), DSO_BINARY_TYPE__BPF_PROG_INFO dso calls into a new function symbol__disassemble_bpf(), where annotation line information is filled based on the bpf_prog_info and btf data saved in given perf_env. symbol__disassemble_bpf() uses binutils's libopcodes to disassemble bpf programs. Committer testing: After fixing this: - u64 *addrs = (u64 *)(info_linear->info.jited_ksyms); + u64 *addrs = (u64 *)(uintptr_t)(info_linear->info.jited_ksyms); Detected when crossbuilding to a 32-bit arch. And making all this dependent on HAVE_LIBBFD_SUPPORT and HAVE_LIBBPF_SUPPORT: 1) Have a BPF program running, one that has BTF info, etc, I used the tools/perf/examples/bpf/augmented_raw_syscalls.c put in place by 'perf trace'. # grep -B1 augmented_raw ~/.perfconfig [trace] add_events = /home/acme/git/perf/tools/perf/examples/bpf/augmented_raw_syscalls.c # # perf trace -e *mmsg dnf/6245 sendmmsg(20, 0x7f5485a88030, 2, MSG_NOSIGNAL) = 2 NetworkManager/10055 sendmmsg(22<socket:[1056822]>, 0x7f8126ad1bb0, 2, MSG_NOSIGNAL) = 2 2) Then do a 'perf record' system wide for a while: # perf record -a ^C[ perf record: Woken up 68 times to write data ] [ perf record: Captured and wrote 19.427 MB perf.data (366891 samples) ] # 3) Check that we captured BPF and BTF info in the perf.data file: # perf report --header-only | grep 'b[pt]f' # event : name = cycles:ppp, , id = { 294789, 294790, 294791, 294792, 294793, 294794, 294795, 294796 }, size = 112, { sample_period, sample_freq } = 4000, sample_type = IP|TID|TIME|CPU|PERIOD, read_format = ID, disabled = 1, inherit = 1, mmap = 1, comm = 1, freq = 1, task = 1, precise_ip = 3, sample_id_all = 1, exclude_guest = 1, mmap2 = 1, comm_exec = 1, ksymbol = 1, bpf_event = 1 # bpf_prog_info of id 13 # bpf_prog_info of id 14 # bpf_prog_info of id 15 # bpf_prog_info of id 16 # bpf_prog_info of id 17 # bpf_prog_info of id 18 # bpf_prog_info of id 21 # bpf_prog_info of id 22 # bpf_prog_info of id 41 # bpf_prog_info of id 42 # btf info of id 2 # 4) Check which programs got recorded: # perf report | grep bpf_prog | head 0.16% exe bpf_prog_819967866022f1e1_sys_enter [k] bpf_prog_819967866022f1e1_sys_enter 0.14% exe bpf_prog_c1bd85c092d6e4aa_sys_exit [k] bpf_prog_c1bd85c092d6e4aa_sys_exit 0.08% fuse-overlayfs bpf_prog_819967866022f1e1_sys_enter [k] bpf_prog_819967866022f1e1_sys_enter 0.07% fuse-overlayfs bpf_prog_c1bd85c092d6e4aa_sys_exit [k] bpf_prog_c1bd85c092d6e4aa_sys_exit 0.01% clang-4.0 bpf_prog_c1bd85c092d6e4aa_sys_exit [k] bpf_prog_c1bd85c092d6e4aa_sys_exit 0.01% clang-4.0 bpf_prog_819967866022f1e1_sys_enter [k] bpf_prog_819967866022f1e1_sys_enter 0.00% clang bpf_prog_c1bd85c092d6e4aa_sys_exit [k] bpf_prog_c1bd85c092d6e4aa_sys_exit 0.00% runc bpf_prog_819967866022f1e1_sys_enter [k] bpf_prog_819967866022f1e1_sys_enter 0.00% clang bpf_prog_819967866022f1e1_sys_enter [k] bpf_prog_819967866022f1e1_sys_enter 0.00% sh bpf_prog_c1bd85c092d6e4aa_sys_exit [k] bpf_prog_c1bd85c092d6e4aa_sys_exit # This was with the default --sort order for 'perf report', which is: --sort comm,dso,symbol If we just look for the symbol, for instance: # perf report --sort symbol | grep bpf_prog | head 0.26% [k] bpf_prog_819967866022f1e1_sys_enter - - 0.24% [k] bpf_prog_c1bd85c092d6e4aa_sys_exit - - # or the DSO: # perf report --sort dso | grep bpf_prog | head 0.26% bpf_prog_819967866022f1e1_sys_enter 0.24% bpf_prog_c1bd85c092d6e4aa_sys_exit # We'll see the two BPF programs that augmented_raw_syscalls.o puts in place, one attached to the raw_syscalls:sys_enter and another to the raw_syscalls:sys_exit tracepoints, as expected. Now we can finally do, from the command line, annotation for one of those two symbols, with the original BPF program source coude intermixed with the disassembled JITed code: # perf annotate --stdio2 bpf_prog_819967866022f1e1_sys_enter Samples: 950 of event 'cycles:ppp', 4000 Hz, Event count (approx.): 553756947, [percent: local period] bpf_prog_819967866022f1e1_sys_enter() bpf_prog_819967866022f1e1_sys_enter Percent int sys_enter(struct syscall_enter_args *args) 53.41 push %rbp 0.63 mov %rsp,%rbp 0.31 sub $0x170,%rsp 1.93 sub $0x28,%rbp 7.02 mov %rbx,0x0(%rbp) 3.20 mov %r13,0x8(%rbp) 1.07 mov %r14,0x10(%rbp) 0.61 mov %r15,0x18(%rbp) 0.11 xor %eax,%eax 1.29 mov %rax,0x20(%rbp) 0.11 mov %rdi,%rbx return bpf_get_current_pid_tgid(); 2.02 → callq *ffffffffda6776d9 2.76 mov %eax,-0x148(%rbp) mov %rbp,%rsi int sys_enter(struct syscall_enter_args *args) add $0xfffffffffffffeb8,%rsi return bpf_map_lookup_elem(pids, &pid) != NULL; movabs $0xffff975ac2607800,%rdi 1.26 → callq *ffffffffda6789e9 cmp $0x0,%rax 2.43 → je 0 add $0x38,%rax 0.21 xor %r13d,%r13d if (pid_filter__has(&pids_filtered, getpid())) 0.81 cmp $0x0,%rax → jne 0 mov %rbp,%rdi probe_read(&augmented_args.args, sizeof(augmented_args.args), args); 2.22 add $0xfffffffffffffeb8,%rdi 0.11 mov $0x40,%esi 0.32 mov %rbx,%rdx 2.74 → callq *ffffffffda658409 syscall = bpf_map_lookup_elem(&syscalls, &augmented_args.args.syscall_nr); 0.22 mov %rbp,%rsi 1.69 add $0xfffffffffffffec0,%rsi syscall = bpf_map_lookup_elem(&syscalls, &augmented_args.args.syscall_nr); movabs $0xffff975bfcd36000,%rdi add $0xd0,%rdi 0.21 mov 0x0(%rsi),%eax 0.93 cmp $0x200,%rax → jae 0 0.10 shl $0x3,%rax 0.11 add %rdi,%rax 0.11 → jmp 0 xor %eax,%eax if (syscall == NULL || !syscall->enabled) 1.07 cmp $0x0,%rax → je 0 if (syscall == NULL || !syscall->enabled) 6.57 movzbq 0x0(%rax),%rdi if (syscall == NULL || !syscall->enabled) cmp $0x0,%rdi 0.95 → je 0 mov $0x40,%r8d switch (augmented_args.args.syscall_nr) { mov -0x140(%rbp),%rdi switch (augmented_args.args.syscall_nr) { cmp $0x2,%rdi → je 0 cmp $0x101,%rdi → je 0 cmp $0x15,%rdi → jne 0 case SYS_OPEN: filename_arg = (const void *)args->args[0]; mov 0x10(%rbx),%rdx → jmp 0 case SYS_OPENAT: filename_arg = (const void *)args->args[1]; mov 0x18(%rbx),%rdx if (filename_arg != NULL) { cmp $0x0,%rdx → je 0 xor %edi,%edi augmented_args.filename.reserved = 0; mov %edi,-0x104(%rbp) augmented_args.filename.size = probe_read_str(&augmented_args.filename.value, mov %rbp,%rdi add $0xffffffffffffff00,%rdi augmented_args.filename.size = probe_read_str(&augmented_args.filename.value, mov $0x100,%esi → callq *ffffffffda658499 mov $0x148,%r8d augmented_args.filename.size = probe_read_str(&augmented_args.filename.value, mov %eax,-0x108(%rbp) augmented_args.filename.size = probe_read_str(&augmented_args.filename.value, mov %rax,%rdi shl $0x20,%rdi shr $0x20,%rdi if (augmented_args.filename.size < sizeof(augmented_args.filename.value)) { cmp $0xff,%rdi → ja 0 len -= sizeof(augmented_args.filename.value) - augmented_args.filename.size; add $0x48,%rax len &= sizeof(augmented_args.filename.value) - 1; and $0xff,%rax mov %rax,%r8 mov %rbp,%rcx return perf_event_output(args, &__augmented_syscalls__, BPF_F_CURRENT_CPU, &augmented_args, len); add $0xfffffffffffffeb8,%rcx mov %rbx,%rdi movabs $0xffff975fbd72d800,%rsi mov $0xffffffff,%edx → callq *ffffffffda658ad9 mov %rax,%r13 } mov %r13,%rax 0.72 mov 0x0(%rbp),%rbx mov 0x8(%rbp),%r13 1.16 mov 0x10(%rbp),%r14 0.10 mov 0x18(%rbp),%r15 0.42 add $0x28,%rbp 0.54 leaveq 0.54 ← retq # Please see 'man perf-config' to see how to control what should be seen, via ~/.perfconfig [annotate] section, for instance, one can suppress the source code and see just the disassembly, etc. Alternatively, use the TUI bu just using 'perf annotate', press '/bpf_prog' to see the bpf symbols, press enter and do the interactive annotation, which allows for dumping to a file after selecting the the various output tunables, for instance, the above without source code intermixed, plus showing all the instruction offsets: # perf annotate bpf_prog_819967866022f1e1_sys_enter Then press: 's' to hide the source code + 'O' twice to show all instruction offsets, then 'P' to print to the bpf_prog_819967866022f1e1_sys_enter.annotation file, which will have: # cat bpf_prog_819967866022f1e1_sys_enter.annotation bpf_prog_819967866022f1e1_sys_enter() bpf_prog_819967866022f1e1_sys_enter Event: cycles:ppp 53.41 0: push %rbp 0.63 1: mov %rsp,%rbp 0.31 4: sub $0x170,%rsp 1.93 b: sub $0x28,%rbp 7.02 f: mov %rbx,0x0(%rbp) 3.20 13: mov %r13,0x8(%rbp) 1.07 17: mov %r14,0x10(%rbp) 0.61 1b: mov %r15,0x18(%rbp) 0.11 1f: xor %eax,%eax 1.29 21: mov %rax,0x20(%rbp) 0.11 25: mov %rdi,%rbx 2.02 28: → callq *ffffffffda6776d9 2.76 2d: mov %eax,-0x148(%rbp) 33: mov %rbp,%rsi 36: add $0xfffffffffffffeb8,%rsi 3d: movabs $0xffff975ac2607800,%rdi 1.26 47: → callq *ffffffffda6789e9 4c: cmp $0x0,%rax 2.43 50: → je 0 52: add $0x38,%rax 0.21 56: xor %r13d,%r13d 0.81 59: cmp $0x0,%rax 5d: → jne 0 63: mov %rbp,%rdi 2.22 66: add $0xfffffffffffffeb8,%rdi 0.11 6d: mov $0x40,%esi 0.32 72: mov %rbx,%rdx 2.74 75: → callq *ffffffffda658409 0.22 7a: mov %rbp,%rsi 1.69 7d: add $0xfffffffffffffec0,%rsi 84: movabs $0xffff975bfcd36000,%rdi 8e: add $0xd0,%rdi 0.21 95: mov 0x0(%rsi),%eax 0.93 98: cmp $0x200,%rax 9f: → jae 0 0.10 a1: shl $0x3,%rax 0.11 a5: add %rdi,%rax 0.11 a8: → jmp 0 aa: xor %eax,%eax 1.07 ac: cmp $0x0,%rax b0: → je 0 6.57 b6: movzbq 0x0(%rax),%rdi bb: cmp $0x0,%rdi 0.95 bf: → je 0 c5: mov $0x40,%r8d cb: mov -0x140(%rbp),%rdi d2: cmp $0x2,%rdi d6: → je 0 d8: cmp $0x101,%rdi df: → je 0 e1: cmp $0x15,%rdi e5: → jne 0 e7: mov 0x10(%rbx),%rdx eb: → jmp 0 ed: mov 0x18(%rbx),%rdx f1: cmp $0x0,%rdx f5: → je 0 f7: xor %edi,%edi f9: mov %edi,-0x104(%rbp) ff: mov %rbp,%rdi 102: add $0xffffffffffffff00,%rdi 109: mov $0x100,%esi 10e: → callq *ffffffffda658499 113: mov $0x148,%r8d 119: mov %eax,-0x108(%rbp) 11f: mov %rax,%rdi 122: shl $0x20,%rdi 126: shr $0x20,%rdi 12a: cmp $0xff,%rdi 131: → ja 0 133: add $0x48,%rax 137: and $0xff,%rax 13d: mov %rax,%r8 140: mov %rbp,%rcx 143: add $0xfffffffffffffeb8,%rcx 14a: mov %rbx,%rdi 14d: movabs $0xffff975fbd72d800,%rsi 157: mov $0xffffffff,%edx 15c: → callq *ffffffffda658ad9 161: mov %rax,%r13 164: mov %r13,%rax 0.72 167: mov 0x0(%rbp),%rbx 16b: mov 0x8(%rbp),%r13 1.16 16f: mov 0x10(%rbp),%r14 0.10 173: mov 0x18(%rbp),%r15 0.42 177: add $0x28,%rbp 0.54 17b: leaveq 0.54 17c: ← retq Another cool way to test all this is to symple use 'perf top' look for those symbols, go there and press enter, annotate it live :-) Signed-off-by: Song Liu <songliubraving@fb.com> Reviewed-by: Jiri Olsa <jolsa@kernel.org> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Alexei Starovoitov <ast@kernel.org> Cc: Daniel Borkmann <daniel@iogearbox.net> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Stanislav Fomichev <sdf@google.com> Link: http://lkml.kernel.org/r/20190312053051.2690567-13-songliubraving@fb.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2019-03-12 08:30:48 +03:00
case SYMBOL_ANNOTATE_ERRNO__NO_LIBOPCODES_FOR_BPF:
scnprintf(buf, buflen, "Please link with binutils's libopcode to enable BPF annotation");
break;
case SYMBOL_ANNOTATE_ERRNO__ARCH_INIT_REGEXP:
scnprintf(buf, buflen, "Problems with arch specific instruction name regular expressions.");
break;
case SYMBOL_ANNOTATE_ERRNO__ARCH_INIT_CPUID_PARSING:
scnprintf(buf, buflen, "Problems while parsing the CPUID in the arch specific initialization.");
break;
case SYMBOL_ANNOTATE_ERRNO__BPF_INVALID_FILE:
scnprintf(buf, buflen, "Invalid BPF file: %s.", dso->long_name);
break;
case SYMBOL_ANNOTATE_ERRNO__BPF_MISSING_BTF:
scnprintf(buf, buflen, "The %s BPF file has no BTF section, compile with -g or use pahole -J.",
dso->long_name);
break;
default:
scnprintf(buf, buflen, "Internal error: Invalid %d error code\n", errnum);
break;
}
return 0;
}
static int dso__disassemble_filename(struct dso *dso, char *filename, size_t filename_size)
{
char linkname[PATH_MAX];
char *build_id_filename;
char *build_id_path = NULL;
char *pos;
perf annotate: Fix non-null terminated buffer returned by readlink() Our local MSAN (Memory Sanitizer) build of perf throws a warning that comes from the "dso__disassemble_filename" function in "tools/perf/util/annotate.c" when running perf record. The warning stems from the call to readlink, in which "build_id_path" was being read into "linkname". Since readlink does not null terminate, an uninitialized memory access would later occur when "linkname" is passed into the strstr function. This is simply fixed by null-terminating "linkname" after the call to readlink. To reproduce this warning, build perf by running: $ make -C tools/perf CLANG=1 CC=clang EXTRA_CFLAGS="-fsanitize=memory -fsanitize-memory-track-origins" (Additionally, llvm might have to be installed and clang might have to be specified as the compiler - export CC=/usr/bin/clang) Then running: tools/perf/perf record -o - ls / | tools/perf/perf --no-pager annotate -i - --stdio Please see the cover letter for why false positive warnings may be generated. Signed-off-by: Numfor Mbiziwo-Tiapo <nums@google.com> Acked-by: Ian Rogers <irogers@google.com> Acked-by: Jiri Olsa <jolsa@kernel.org> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Ian Rogers <irogers@google.com> Cc: Mark Drayton <mbd@fb.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Song Liu <songliubraving@fb.com> Cc: Stephane Eranian <eranian@google.com> Link: http://lore.kernel.org/lkml/20190729205750.193289-1-nums@google.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2019-07-29 23:57:50 +03:00
int len;
if (dso->symtab_type == DSO_BINARY_TYPE__KALLSYMS &&
!dso__is_kcore(dso))
return SYMBOL_ANNOTATE_ERRNO__NO_VMLINUX;
build_id_filename = dso__build_id_filename(dso, NULL, 0, false);
if (build_id_filename) {
__symbol__join_symfs(filename, filename_size, build_id_filename);
free(build_id_filename);
} else {
if (dso->has_build_id)
return ENOMEM;
goto fallback;
}
build_id_path = strdup(filename);
if (!build_id_path)
return ENOMEM;
/*
* old style build-id cache has name of XX/XXXXXXX.. while
* new style has XX/XXXXXXX../{elf,kallsyms,vdso}.
* extract the build-id part of dirname in the new style only.
*/
pos = strrchr(build_id_path, '/');
if (pos && strlen(pos) < SBUILD_ID_SIZE - 2)
dirname(build_id_path);
perf annotate: Fix non-null terminated buffer returned by readlink() Our local MSAN (Memory Sanitizer) build of perf throws a warning that comes from the "dso__disassemble_filename" function in "tools/perf/util/annotate.c" when running perf record. The warning stems from the call to readlink, in which "build_id_path" was being read into "linkname". Since readlink does not null terminate, an uninitialized memory access would later occur when "linkname" is passed into the strstr function. This is simply fixed by null-terminating "linkname" after the call to readlink. To reproduce this warning, build perf by running: $ make -C tools/perf CLANG=1 CC=clang EXTRA_CFLAGS="-fsanitize=memory -fsanitize-memory-track-origins" (Additionally, llvm might have to be installed and clang might have to be specified as the compiler - export CC=/usr/bin/clang) Then running: tools/perf/perf record -o - ls / | tools/perf/perf --no-pager annotate -i - --stdio Please see the cover letter for why false positive warnings may be generated. Signed-off-by: Numfor Mbiziwo-Tiapo <nums@google.com> Acked-by: Ian Rogers <irogers@google.com> Acked-by: Jiri Olsa <jolsa@kernel.org> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Ian Rogers <irogers@google.com> Cc: Mark Drayton <mbd@fb.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Song Liu <songliubraving@fb.com> Cc: Stephane Eranian <eranian@google.com> Link: http://lore.kernel.org/lkml/20190729205750.193289-1-nums@google.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2019-07-29 23:57:50 +03:00
if (dso__is_kcore(dso))
goto fallback;
len = readlink(build_id_path, linkname, sizeof(linkname) - 1);
if (len < 0)
goto fallback;
linkname[len] = '\0';
if (strstr(linkname, DSO__NAME_KALLSYMS) ||
access(filename, R_OK)) {
fallback:
/*
* If we don't have build-ids or the build-id file isn't in the
* cache, or is just a kallsyms file, well, lets hope that this
* DSO is the same as when 'perf record' ran.
*/
__symbol__join_symfs(filename, filename_size, dso->long_name);
}
free(build_id_path);
return 0;
}
perf annotate: Enable annotation of BPF programs In symbol__disassemble(), DSO_BINARY_TYPE__BPF_PROG_INFO dso calls into a new function symbol__disassemble_bpf(), where annotation line information is filled based on the bpf_prog_info and btf data saved in given perf_env. symbol__disassemble_bpf() uses binutils's libopcodes to disassemble bpf programs. Committer testing: After fixing this: - u64 *addrs = (u64 *)(info_linear->info.jited_ksyms); + u64 *addrs = (u64 *)(uintptr_t)(info_linear->info.jited_ksyms); Detected when crossbuilding to a 32-bit arch. And making all this dependent on HAVE_LIBBFD_SUPPORT and HAVE_LIBBPF_SUPPORT: 1) Have a BPF program running, one that has BTF info, etc, I used the tools/perf/examples/bpf/augmented_raw_syscalls.c put in place by 'perf trace'. # grep -B1 augmented_raw ~/.perfconfig [trace] add_events = /home/acme/git/perf/tools/perf/examples/bpf/augmented_raw_syscalls.c # # perf trace -e *mmsg dnf/6245 sendmmsg(20, 0x7f5485a88030, 2, MSG_NOSIGNAL) = 2 NetworkManager/10055 sendmmsg(22<socket:[1056822]>, 0x7f8126ad1bb0, 2, MSG_NOSIGNAL) = 2 2) Then do a 'perf record' system wide for a while: # perf record -a ^C[ perf record: Woken up 68 times to write data ] [ perf record: Captured and wrote 19.427 MB perf.data (366891 samples) ] # 3) Check that we captured BPF and BTF info in the perf.data file: # perf report --header-only | grep 'b[pt]f' # event : name = cycles:ppp, , id = { 294789, 294790, 294791, 294792, 294793, 294794, 294795, 294796 }, size = 112, { sample_period, sample_freq } = 4000, sample_type = IP|TID|TIME|CPU|PERIOD, read_format = ID, disabled = 1, inherit = 1, mmap = 1, comm = 1, freq = 1, task = 1, precise_ip = 3, sample_id_all = 1, exclude_guest = 1, mmap2 = 1, comm_exec = 1, ksymbol = 1, bpf_event = 1 # bpf_prog_info of id 13 # bpf_prog_info of id 14 # bpf_prog_info of id 15 # bpf_prog_info of id 16 # bpf_prog_info of id 17 # bpf_prog_info of id 18 # bpf_prog_info of id 21 # bpf_prog_info of id 22 # bpf_prog_info of id 41 # bpf_prog_info of id 42 # btf info of id 2 # 4) Check which programs got recorded: # perf report | grep bpf_prog | head 0.16% exe bpf_prog_819967866022f1e1_sys_enter [k] bpf_prog_819967866022f1e1_sys_enter 0.14% exe bpf_prog_c1bd85c092d6e4aa_sys_exit [k] bpf_prog_c1bd85c092d6e4aa_sys_exit 0.08% fuse-overlayfs bpf_prog_819967866022f1e1_sys_enter [k] bpf_prog_819967866022f1e1_sys_enter 0.07% fuse-overlayfs bpf_prog_c1bd85c092d6e4aa_sys_exit [k] bpf_prog_c1bd85c092d6e4aa_sys_exit 0.01% clang-4.0 bpf_prog_c1bd85c092d6e4aa_sys_exit [k] bpf_prog_c1bd85c092d6e4aa_sys_exit 0.01% clang-4.0 bpf_prog_819967866022f1e1_sys_enter [k] bpf_prog_819967866022f1e1_sys_enter 0.00% clang bpf_prog_c1bd85c092d6e4aa_sys_exit [k] bpf_prog_c1bd85c092d6e4aa_sys_exit 0.00% runc bpf_prog_819967866022f1e1_sys_enter [k] bpf_prog_819967866022f1e1_sys_enter 0.00% clang bpf_prog_819967866022f1e1_sys_enter [k] bpf_prog_819967866022f1e1_sys_enter 0.00% sh bpf_prog_c1bd85c092d6e4aa_sys_exit [k] bpf_prog_c1bd85c092d6e4aa_sys_exit # This was with the default --sort order for 'perf report', which is: --sort comm,dso,symbol If we just look for the symbol, for instance: # perf report --sort symbol | grep bpf_prog | head 0.26% [k] bpf_prog_819967866022f1e1_sys_enter - - 0.24% [k] bpf_prog_c1bd85c092d6e4aa_sys_exit - - # or the DSO: # perf report --sort dso | grep bpf_prog | head 0.26% bpf_prog_819967866022f1e1_sys_enter 0.24% bpf_prog_c1bd85c092d6e4aa_sys_exit # We'll see the two BPF programs that augmented_raw_syscalls.o puts in place, one attached to the raw_syscalls:sys_enter and another to the raw_syscalls:sys_exit tracepoints, as expected. Now we can finally do, from the command line, annotation for one of those two symbols, with the original BPF program source coude intermixed with the disassembled JITed code: # perf annotate --stdio2 bpf_prog_819967866022f1e1_sys_enter Samples: 950 of event 'cycles:ppp', 4000 Hz, Event count (approx.): 553756947, [percent: local period] bpf_prog_819967866022f1e1_sys_enter() bpf_prog_819967866022f1e1_sys_enter Percent int sys_enter(struct syscall_enter_args *args) 53.41 push %rbp 0.63 mov %rsp,%rbp 0.31 sub $0x170,%rsp 1.93 sub $0x28,%rbp 7.02 mov %rbx,0x0(%rbp) 3.20 mov %r13,0x8(%rbp) 1.07 mov %r14,0x10(%rbp) 0.61 mov %r15,0x18(%rbp) 0.11 xor %eax,%eax 1.29 mov %rax,0x20(%rbp) 0.11 mov %rdi,%rbx return bpf_get_current_pid_tgid(); 2.02 → callq *ffffffffda6776d9 2.76 mov %eax,-0x148(%rbp) mov %rbp,%rsi int sys_enter(struct syscall_enter_args *args) add $0xfffffffffffffeb8,%rsi return bpf_map_lookup_elem(pids, &pid) != NULL; movabs $0xffff975ac2607800,%rdi 1.26 → callq *ffffffffda6789e9 cmp $0x0,%rax 2.43 → je 0 add $0x38,%rax 0.21 xor %r13d,%r13d if (pid_filter__has(&pids_filtered, getpid())) 0.81 cmp $0x0,%rax → jne 0 mov %rbp,%rdi probe_read(&augmented_args.args, sizeof(augmented_args.args), args); 2.22 add $0xfffffffffffffeb8,%rdi 0.11 mov $0x40,%esi 0.32 mov %rbx,%rdx 2.74 → callq *ffffffffda658409 syscall = bpf_map_lookup_elem(&syscalls, &augmented_args.args.syscall_nr); 0.22 mov %rbp,%rsi 1.69 add $0xfffffffffffffec0,%rsi syscall = bpf_map_lookup_elem(&syscalls, &augmented_args.args.syscall_nr); movabs $0xffff975bfcd36000,%rdi add $0xd0,%rdi 0.21 mov 0x0(%rsi),%eax 0.93 cmp $0x200,%rax → jae 0 0.10 shl $0x3,%rax 0.11 add %rdi,%rax 0.11 → jmp 0 xor %eax,%eax if (syscall == NULL || !syscall->enabled) 1.07 cmp $0x0,%rax → je 0 if (syscall == NULL || !syscall->enabled) 6.57 movzbq 0x0(%rax),%rdi if (syscall == NULL || !syscall->enabled) cmp $0x0,%rdi 0.95 → je 0 mov $0x40,%r8d switch (augmented_args.args.syscall_nr) { mov -0x140(%rbp),%rdi switch (augmented_args.args.syscall_nr) { cmp $0x2,%rdi → je 0 cmp $0x101,%rdi → je 0 cmp $0x15,%rdi → jne 0 case SYS_OPEN: filename_arg = (const void *)args->args[0]; mov 0x10(%rbx),%rdx → jmp 0 case SYS_OPENAT: filename_arg = (const void *)args->args[1]; mov 0x18(%rbx),%rdx if (filename_arg != NULL) { cmp $0x0,%rdx → je 0 xor %edi,%edi augmented_args.filename.reserved = 0; mov %edi,-0x104(%rbp) augmented_args.filename.size = probe_read_str(&augmented_args.filename.value, mov %rbp,%rdi add $0xffffffffffffff00,%rdi augmented_args.filename.size = probe_read_str(&augmented_args.filename.value, mov $0x100,%esi → callq *ffffffffda658499 mov $0x148,%r8d augmented_args.filename.size = probe_read_str(&augmented_args.filename.value, mov %eax,-0x108(%rbp) augmented_args.filename.size = probe_read_str(&augmented_args.filename.value, mov %rax,%rdi shl $0x20,%rdi shr $0x20,%rdi if (augmented_args.filename.size < sizeof(augmented_args.filename.value)) { cmp $0xff,%rdi → ja 0 len -= sizeof(augmented_args.filename.value) - augmented_args.filename.size; add $0x48,%rax len &= sizeof(augmented_args.filename.value) - 1; and $0xff,%rax mov %rax,%r8 mov %rbp,%rcx return perf_event_output(args, &__augmented_syscalls__, BPF_F_CURRENT_CPU, &augmented_args, len); add $0xfffffffffffffeb8,%rcx mov %rbx,%rdi movabs $0xffff975fbd72d800,%rsi mov $0xffffffff,%edx → callq *ffffffffda658ad9 mov %rax,%r13 } mov %r13,%rax 0.72 mov 0x0(%rbp),%rbx mov 0x8(%rbp),%r13 1.16 mov 0x10(%rbp),%r14 0.10 mov 0x18(%rbp),%r15 0.42 add $0x28,%rbp 0.54 leaveq 0.54 ← retq # Please see 'man perf-config' to see how to control what should be seen, via ~/.perfconfig [annotate] section, for instance, one can suppress the source code and see just the disassembly, etc. Alternatively, use the TUI bu just using 'perf annotate', press '/bpf_prog' to see the bpf symbols, press enter and do the interactive annotation, which allows for dumping to a file after selecting the the various output tunables, for instance, the above without source code intermixed, plus showing all the instruction offsets: # perf annotate bpf_prog_819967866022f1e1_sys_enter Then press: 's' to hide the source code + 'O' twice to show all instruction offsets, then 'P' to print to the bpf_prog_819967866022f1e1_sys_enter.annotation file, which will have: # cat bpf_prog_819967866022f1e1_sys_enter.annotation bpf_prog_819967866022f1e1_sys_enter() bpf_prog_819967866022f1e1_sys_enter Event: cycles:ppp 53.41 0: push %rbp 0.63 1: mov %rsp,%rbp 0.31 4: sub $0x170,%rsp 1.93 b: sub $0x28,%rbp 7.02 f: mov %rbx,0x0(%rbp) 3.20 13: mov %r13,0x8(%rbp) 1.07 17: mov %r14,0x10(%rbp) 0.61 1b: mov %r15,0x18(%rbp) 0.11 1f: xor %eax,%eax 1.29 21: mov %rax,0x20(%rbp) 0.11 25: mov %rdi,%rbx 2.02 28: → callq *ffffffffda6776d9 2.76 2d: mov %eax,-0x148(%rbp) 33: mov %rbp,%rsi 36: add $0xfffffffffffffeb8,%rsi 3d: movabs $0xffff975ac2607800,%rdi 1.26 47: → callq *ffffffffda6789e9 4c: cmp $0x0,%rax 2.43 50: → je 0 52: add $0x38,%rax 0.21 56: xor %r13d,%r13d 0.81 59: cmp $0x0,%rax 5d: → jne 0 63: mov %rbp,%rdi 2.22 66: add $0xfffffffffffffeb8,%rdi 0.11 6d: mov $0x40,%esi 0.32 72: mov %rbx,%rdx 2.74 75: → callq *ffffffffda658409 0.22 7a: mov %rbp,%rsi 1.69 7d: add $0xfffffffffffffec0,%rsi 84: movabs $0xffff975bfcd36000,%rdi 8e: add $0xd0,%rdi 0.21 95: mov 0x0(%rsi),%eax 0.93 98: cmp $0x200,%rax 9f: → jae 0 0.10 a1: shl $0x3,%rax 0.11 a5: add %rdi,%rax 0.11 a8: → jmp 0 aa: xor %eax,%eax 1.07 ac: cmp $0x0,%rax b0: → je 0 6.57 b6: movzbq 0x0(%rax),%rdi bb: cmp $0x0,%rdi 0.95 bf: → je 0 c5: mov $0x40,%r8d cb: mov -0x140(%rbp),%rdi d2: cmp $0x2,%rdi d6: → je 0 d8: cmp $0x101,%rdi df: → je 0 e1: cmp $0x15,%rdi e5: → jne 0 e7: mov 0x10(%rbx),%rdx eb: → jmp 0 ed: mov 0x18(%rbx),%rdx f1: cmp $0x0,%rdx f5: → je 0 f7: xor %edi,%edi f9: mov %edi,-0x104(%rbp) ff: mov %rbp,%rdi 102: add $0xffffffffffffff00,%rdi 109: mov $0x100,%esi 10e: → callq *ffffffffda658499 113: mov $0x148,%r8d 119: mov %eax,-0x108(%rbp) 11f: mov %rax,%rdi 122: shl $0x20,%rdi 126: shr $0x20,%rdi 12a: cmp $0xff,%rdi 131: → ja 0 133: add $0x48,%rax 137: and $0xff,%rax 13d: mov %rax,%r8 140: mov %rbp,%rcx 143: add $0xfffffffffffffeb8,%rcx 14a: mov %rbx,%rdi 14d: movabs $0xffff975fbd72d800,%rsi 157: mov $0xffffffff,%edx 15c: → callq *ffffffffda658ad9 161: mov %rax,%r13 164: mov %r13,%rax 0.72 167: mov 0x0(%rbp),%rbx 16b: mov 0x8(%rbp),%r13 1.16 16f: mov 0x10(%rbp),%r14 0.10 173: mov 0x18(%rbp),%r15 0.42 177: add $0x28,%rbp 0.54 17b: leaveq 0.54 17c: ← retq Another cool way to test all this is to symple use 'perf top' look for those symbols, go there and press enter, annotate it live :-) Signed-off-by: Song Liu <songliubraving@fb.com> Reviewed-by: Jiri Olsa <jolsa@kernel.org> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Alexei Starovoitov <ast@kernel.org> Cc: Daniel Borkmann <daniel@iogearbox.net> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Stanislav Fomichev <sdf@google.com> Link: http://lkml.kernel.org/r/20190312053051.2690567-13-songliubraving@fb.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2019-03-12 08:30:48 +03:00
#if defined(HAVE_LIBBFD_SUPPORT) && defined(HAVE_LIBBPF_SUPPORT)
#define PACKAGE "perf"
#include <bfd.h>
#include <dis-asm.h>
#include <bpf/bpf.h>
#include <bpf/btf.h>
#include <bpf/libbpf.h>
#include <linux/btf.h>
perf annotate: Enable annotation of BPF programs In symbol__disassemble(), DSO_BINARY_TYPE__BPF_PROG_INFO dso calls into a new function symbol__disassemble_bpf(), where annotation line information is filled based on the bpf_prog_info and btf data saved in given perf_env. symbol__disassemble_bpf() uses binutils's libopcodes to disassemble bpf programs. Committer testing: After fixing this: - u64 *addrs = (u64 *)(info_linear->info.jited_ksyms); + u64 *addrs = (u64 *)(uintptr_t)(info_linear->info.jited_ksyms); Detected when crossbuilding to a 32-bit arch. And making all this dependent on HAVE_LIBBFD_SUPPORT and HAVE_LIBBPF_SUPPORT: 1) Have a BPF program running, one that has BTF info, etc, I used the tools/perf/examples/bpf/augmented_raw_syscalls.c put in place by 'perf trace'. # grep -B1 augmented_raw ~/.perfconfig [trace] add_events = /home/acme/git/perf/tools/perf/examples/bpf/augmented_raw_syscalls.c # # perf trace -e *mmsg dnf/6245 sendmmsg(20, 0x7f5485a88030, 2, MSG_NOSIGNAL) = 2 NetworkManager/10055 sendmmsg(22<socket:[1056822]>, 0x7f8126ad1bb0, 2, MSG_NOSIGNAL) = 2 2) Then do a 'perf record' system wide for a while: # perf record -a ^C[ perf record: Woken up 68 times to write data ] [ perf record: Captured and wrote 19.427 MB perf.data (366891 samples) ] # 3) Check that we captured BPF and BTF info in the perf.data file: # perf report --header-only | grep 'b[pt]f' # event : name = cycles:ppp, , id = { 294789, 294790, 294791, 294792, 294793, 294794, 294795, 294796 }, size = 112, { sample_period, sample_freq } = 4000, sample_type = IP|TID|TIME|CPU|PERIOD, read_format = ID, disabled = 1, inherit = 1, mmap = 1, comm = 1, freq = 1, task = 1, precise_ip = 3, sample_id_all = 1, exclude_guest = 1, mmap2 = 1, comm_exec = 1, ksymbol = 1, bpf_event = 1 # bpf_prog_info of id 13 # bpf_prog_info of id 14 # bpf_prog_info of id 15 # bpf_prog_info of id 16 # bpf_prog_info of id 17 # bpf_prog_info of id 18 # bpf_prog_info of id 21 # bpf_prog_info of id 22 # bpf_prog_info of id 41 # bpf_prog_info of id 42 # btf info of id 2 # 4) Check which programs got recorded: # perf report | grep bpf_prog | head 0.16% exe bpf_prog_819967866022f1e1_sys_enter [k] bpf_prog_819967866022f1e1_sys_enter 0.14% exe bpf_prog_c1bd85c092d6e4aa_sys_exit [k] bpf_prog_c1bd85c092d6e4aa_sys_exit 0.08% fuse-overlayfs bpf_prog_819967866022f1e1_sys_enter [k] bpf_prog_819967866022f1e1_sys_enter 0.07% fuse-overlayfs bpf_prog_c1bd85c092d6e4aa_sys_exit [k] bpf_prog_c1bd85c092d6e4aa_sys_exit 0.01% clang-4.0 bpf_prog_c1bd85c092d6e4aa_sys_exit [k] bpf_prog_c1bd85c092d6e4aa_sys_exit 0.01% clang-4.0 bpf_prog_819967866022f1e1_sys_enter [k] bpf_prog_819967866022f1e1_sys_enter 0.00% clang bpf_prog_c1bd85c092d6e4aa_sys_exit [k] bpf_prog_c1bd85c092d6e4aa_sys_exit 0.00% runc bpf_prog_819967866022f1e1_sys_enter [k] bpf_prog_819967866022f1e1_sys_enter 0.00% clang bpf_prog_819967866022f1e1_sys_enter [k] bpf_prog_819967866022f1e1_sys_enter 0.00% sh bpf_prog_c1bd85c092d6e4aa_sys_exit [k] bpf_prog_c1bd85c092d6e4aa_sys_exit # This was with the default --sort order for 'perf report', which is: --sort comm,dso,symbol If we just look for the symbol, for instance: # perf report --sort symbol | grep bpf_prog | head 0.26% [k] bpf_prog_819967866022f1e1_sys_enter - - 0.24% [k] bpf_prog_c1bd85c092d6e4aa_sys_exit - - # or the DSO: # perf report --sort dso | grep bpf_prog | head 0.26% bpf_prog_819967866022f1e1_sys_enter 0.24% bpf_prog_c1bd85c092d6e4aa_sys_exit # We'll see the two BPF programs that augmented_raw_syscalls.o puts in place, one attached to the raw_syscalls:sys_enter and another to the raw_syscalls:sys_exit tracepoints, as expected. Now we can finally do, from the command line, annotation for one of those two symbols, with the original BPF program source coude intermixed with the disassembled JITed code: # perf annotate --stdio2 bpf_prog_819967866022f1e1_sys_enter Samples: 950 of event 'cycles:ppp', 4000 Hz, Event count (approx.): 553756947, [percent: local period] bpf_prog_819967866022f1e1_sys_enter() bpf_prog_819967866022f1e1_sys_enter Percent int sys_enter(struct syscall_enter_args *args) 53.41 push %rbp 0.63 mov %rsp,%rbp 0.31 sub $0x170,%rsp 1.93 sub $0x28,%rbp 7.02 mov %rbx,0x0(%rbp) 3.20 mov %r13,0x8(%rbp) 1.07 mov %r14,0x10(%rbp) 0.61 mov %r15,0x18(%rbp) 0.11 xor %eax,%eax 1.29 mov %rax,0x20(%rbp) 0.11 mov %rdi,%rbx return bpf_get_current_pid_tgid(); 2.02 → callq *ffffffffda6776d9 2.76 mov %eax,-0x148(%rbp) mov %rbp,%rsi int sys_enter(struct syscall_enter_args *args) add $0xfffffffffffffeb8,%rsi return bpf_map_lookup_elem(pids, &pid) != NULL; movabs $0xffff975ac2607800,%rdi 1.26 → callq *ffffffffda6789e9 cmp $0x0,%rax 2.43 → je 0 add $0x38,%rax 0.21 xor %r13d,%r13d if (pid_filter__has(&pids_filtered, getpid())) 0.81 cmp $0x0,%rax → jne 0 mov %rbp,%rdi probe_read(&augmented_args.args, sizeof(augmented_args.args), args); 2.22 add $0xfffffffffffffeb8,%rdi 0.11 mov $0x40,%esi 0.32 mov %rbx,%rdx 2.74 → callq *ffffffffda658409 syscall = bpf_map_lookup_elem(&syscalls, &augmented_args.args.syscall_nr); 0.22 mov %rbp,%rsi 1.69 add $0xfffffffffffffec0,%rsi syscall = bpf_map_lookup_elem(&syscalls, &augmented_args.args.syscall_nr); movabs $0xffff975bfcd36000,%rdi add $0xd0,%rdi 0.21 mov 0x0(%rsi),%eax 0.93 cmp $0x200,%rax → jae 0 0.10 shl $0x3,%rax 0.11 add %rdi,%rax 0.11 → jmp 0 xor %eax,%eax if (syscall == NULL || !syscall->enabled) 1.07 cmp $0x0,%rax → je 0 if (syscall == NULL || !syscall->enabled) 6.57 movzbq 0x0(%rax),%rdi if (syscall == NULL || !syscall->enabled) cmp $0x0,%rdi 0.95 → je 0 mov $0x40,%r8d switch (augmented_args.args.syscall_nr) { mov -0x140(%rbp),%rdi switch (augmented_args.args.syscall_nr) { cmp $0x2,%rdi → je 0 cmp $0x101,%rdi → je 0 cmp $0x15,%rdi → jne 0 case SYS_OPEN: filename_arg = (const void *)args->args[0]; mov 0x10(%rbx),%rdx → jmp 0 case SYS_OPENAT: filename_arg = (const void *)args->args[1]; mov 0x18(%rbx),%rdx if (filename_arg != NULL) { cmp $0x0,%rdx → je 0 xor %edi,%edi augmented_args.filename.reserved = 0; mov %edi,-0x104(%rbp) augmented_args.filename.size = probe_read_str(&augmented_args.filename.value, mov %rbp,%rdi add $0xffffffffffffff00,%rdi augmented_args.filename.size = probe_read_str(&augmented_args.filename.value, mov $0x100,%esi → callq *ffffffffda658499 mov $0x148,%r8d augmented_args.filename.size = probe_read_str(&augmented_args.filename.value, mov %eax,-0x108(%rbp) augmented_args.filename.size = probe_read_str(&augmented_args.filename.value, mov %rax,%rdi shl $0x20,%rdi shr $0x20,%rdi if (augmented_args.filename.size < sizeof(augmented_args.filename.value)) { cmp $0xff,%rdi → ja 0 len -= sizeof(augmented_args.filename.value) - augmented_args.filename.size; add $0x48,%rax len &= sizeof(augmented_args.filename.value) - 1; and $0xff,%rax mov %rax,%r8 mov %rbp,%rcx return perf_event_output(args, &__augmented_syscalls__, BPF_F_CURRENT_CPU, &augmented_args, len); add $0xfffffffffffffeb8,%rcx mov %rbx,%rdi movabs $0xffff975fbd72d800,%rsi mov $0xffffffff,%edx → callq *ffffffffda658ad9 mov %rax,%r13 } mov %r13,%rax 0.72 mov 0x0(%rbp),%rbx mov 0x8(%rbp),%r13 1.16 mov 0x10(%rbp),%r14 0.10 mov 0x18(%rbp),%r15 0.42 add $0x28,%rbp 0.54 leaveq 0.54 ← retq # Please see 'man perf-config' to see how to control what should be seen, via ~/.perfconfig [annotate] section, for instance, one can suppress the source code and see just the disassembly, etc. Alternatively, use the TUI bu just using 'perf annotate', press '/bpf_prog' to see the bpf symbols, press enter and do the interactive annotation, which allows for dumping to a file after selecting the the various output tunables, for instance, the above without source code intermixed, plus showing all the instruction offsets: # perf annotate bpf_prog_819967866022f1e1_sys_enter Then press: 's' to hide the source code + 'O' twice to show all instruction offsets, then 'P' to print to the bpf_prog_819967866022f1e1_sys_enter.annotation file, which will have: # cat bpf_prog_819967866022f1e1_sys_enter.annotation bpf_prog_819967866022f1e1_sys_enter() bpf_prog_819967866022f1e1_sys_enter Event: cycles:ppp 53.41 0: push %rbp 0.63 1: mov %rsp,%rbp 0.31 4: sub $0x170,%rsp 1.93 b: sub $0x28,%rbp 7.02 f: mov %rbx,0x0(%rbp) 3.20 13: mov %r13,0x8(%rbp) 1.07 17: mov %r14,0x10(%rbp) 0.61 1b: mov %r15,0x18(%rbp) 0.11 1f: xor %eax,%eax 1.29 21: mov %rax,0x20(%rbp) 0.11 25: mov %rdi,%rbx 2.02 28: → callq *ffffffffda6776d9 2.76 2d: mov %eax,-0x148(%rbp) 33: mov %rbp,%rsi 36: add $0xfffffffffffffeb8,%rsi 3d: movabs $0xffff975ac2607800,%rdi 1.26 47: → callq *ffffffffda6789e9 4c: cmp $0x0,%rax 2.43 50: → je 0 52: add $0x38,%rax 0.21 56: xor %r13d,%r13d 0.81 59: cmp $0x0,%rax 5d: → jne 0 63: mov %rbp,%rdi 2.22 66: add $0xfffffffffffffeb8,%rdi 0.11 6d: mov $0x40,%esi 0.32 72: mov %rbx,%rdx 2.74 75: → callq *ffffffffda658409 0.22 7a: mov %rbp,%rsi 1.69 7d: add $0xfffffffffffffec0,%rsi 84: movabs $0xffff975bfcd36000,%rdi 8e: add $0xd0,%rdi 0.21 95: mov 0x0(%rsi),%eax 0.93 98: cmp $0x200,%rax 9f: → jae 0 0.10 a1: shl $0x3,%rax 0.11 a5: add %rdi,%rax 0.11 a8: → jmp 0 aa: xor %eax,%eax 1.07 ac: cmp $0x0,%rax b0: → je 0 6.57 b6: movzbq 0x0(%rax),%rdi bb: cmp $0x0,%rdi 0.95 bf: → je 0 c5: mov $0x40,%r8d cb: mov -0x140(%rbp),%rdi d2: cmp $0x2,%rdi d6: → je 0 d8: cmp $0x101,%rdi df: → je 0 e1: cmp $0x15,%rdi e5: → jne 0 e7: mov 0x10(%rbx),%rdx eb: → jmp 0 ed: mov 0x18(%rbx),%rdx f1: cmp $0x0,%rdx f5: → je 0 f7: xor %edi,%edi f9: mov %edi,-0x104(%rbp) ff: mov %rbp,%rdi 102: add $0xffffffffffffff00,%rdi 109: mov $0x100,%esi 10e: → callq *ffffffffda658499 113: mov $0x148,%r8d 119: mov %eax,-0x108(%rbp) 11f: mov %rax,%rdi 122: shl $0x20,%rdi 126: shr $0x20,%rdi 12a: cmp $0xff,%rdi 131: → ja 0 133: add $0x48,%rax 137: and $0xff,%rax 13d: mov %rax,%r8 140: mov %rbp,%rcx 143: add $0xfffffffffffffeb8,%rcx 14a: mov %rbx,%rdi 14d: movabs $0xffff975fbd72d800,%rsi 157: mov $0xffffffff,%edx 15c: → callq *ffffffffda658ad9 161: mov %rax,%r13 164: mov %r13,%rax 0.72 167: mov 0x0(%rbp),%rbx 16b: mov 0x8(%rbp),%r13 1.16 16f: mov 0x10(%rbp),%r14 0.10 173: mov 0x18(%rbp),%r15 0.42 177: add $0x28,%rbp 0.54 17b: leaveq 0.54 17c: ← retq Another cool way to test all this is to symple use 'perf top' look for those symbols, go there and press enter, annotate it live :-) Signed-off-by: Song Liu <songliubraving@fb.com> Reviewed-by: Jiri Olsa <jolsa@kernel.org> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Alexei Starovoitov <ast@kernel.org> Cc: Daniel Borkmann <daniel@iogearbox.net> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Stanislav Fomichev <sdf@google.com> Link: http://lkml.kernel.org/r/20190312053051.2690567-13-songliubraving@fb.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2019-03-12 08:30:48 +03:00
static int symbol__disassemble_bpf(struct symbol *sym,
struct annotate_args *args)
{
struct annotation *notes = symbol__annotation(sym);
struct annotation_options *opts = args->options;
struct bpf_prog_linfo *prog_linfo = NULL;
struct bpf_prog_info_node *info_node;
int len = sym->end - sym->start;
disassembler_ftype disassemble;
struct map *map = args->ms.map;
struct perf_bpil *info_linear;
perf annotate: Enable annotation of BPF programs In symbol__disassemble(), DSO_BINARY_TYPE__BPF_PROG_INFO dso calls into a new function symbol__disassemble_bpf(), where annotation line information is filled based on the bpf_prog_info and btf data saved in given perf_env. symbol__disassemble_bpf() uses binutils's libopcodes to disassemble bpf programs. Committer testing: After fixing this: - u64 *addrs = (u64 *)(info_linear->info.jited_ksyms); + u64 *addrs = (u64 *)(uintptr_t)(info_linear->info.jited_ksyms); Detected when crossbuilding to a 32-bit arch. And making all this dependent on HAVE_LIBBFD_SUPPORT and HAVE_LIBBPF_SUPPORT: 1) Have a BPF program running, one that has BTF info, etc, I used the tools/perf/examples/bpf/augmented_raw_syscalls.c put in place by 'perf trace'. # grep -B1 augmented_raw ~/.perfconfig [trace] add_events = /home/acme/git/perf/tools/perf/examples/bpf/augmented_raw_syscalls.c # # perf trace -e *mmsg dnf/6245 sendmmsg(20, 0x7f5485a88030, 2, MSG_NOSIGNAL) = 2 NetworkManager/10055 sendmmsg(22<socket:[1056822]>, 0x7f8126ad1bb0, 2, MSG_NOSIGNAL) = 2 2) Then do a 'perf record' system wide for a while: # perf record -a ^C[ perf record: Woken up 68 times to write data ] [ perf record: Captured and wrote 19.427 MB perf.data (366891 samples) ] # 3) Check that we captured BPF and BTF info in the perf.data file: # perf report --header-only | grep 'b[pt]f' # event : name = cycles:ppp, , id = { 294789, 294790, 294791, 294792, 294793, 294794, 294795, 294796 }, size = 112, { sample_period, sample_freq } = 4000, sample_type = IP|TID|TIME|CPU|PERIOD, read_format = ID, disabled = 1, inherit = 1, mmap = 1, comm = 1, freq = 1, task = 1, precise_ip = 3, sample_id_all = 1, exclude_guest = 1, mmap2 = 1, comm_exec = 1, ksymbol = 1, bpf_event = 1 # bpf_prog_info of id 13 # bpf_prog_info of id 14 # bpf_prog_info of id 15 # bpf_prog_info of id 16 # bpf_prog_info of id 17 # bpf_prog_info of id 18 # bpf_prog_info of id 21 # bpf_prog_info of id 22 # bpf_prog_info of id 41 # bpf_prog_info of id 42 # btf info of id 2 # 4) Check which programs got recorded: # perf report | grep bpf_prog | head 0.16% exe bpf_prog_819967866022f1e1_sys_enter [k] bpf_prog_819967866022f1e1_sys_enter 0.14% exe bpf_prog_c1bd85c092d6e4aa_sys_exit [k] bpf_prog_c1bd85c092d6e4aa_sys_exit 0.08% fuse-overlayfs bpf_prog_819967866022f1e1_sys_enter [k] bpf_prog_819967866022f1e1_sys_enter 0.07% fuse-overlayfs bpf_prog_c1bd85c092d6e4aa_sys_exit [k] bpf_prog_c1bd85c092d6e4aa_sys_exit 0.01% clang-4.0 bpf_prog_c1bd85c092d6e4aa_sys_exit [k] bpf_prog_c1bd85c092d6e4aa_sys_exit 0.01% clang-4.0 bpf_prog_819967866022f1e1_sys_enter [k] bpf_prog_819967866022f1e1_sys_enter 0.00% clang bpf_prog_c1bd85c092d6e4aa_sys_exit [k] bpf_prog_c1bd85c092d6e4aa_sys_exit 0.00% runc bpf_prog_819967866022f1e1_sys_enter [k] bpf_prog_819967866022f1e1_sys_enter 0.00% clang bpf_prog_819967866022f1e1_sys_enter [k] bpf_prog_819967866022f1e1_sys_enter 0.00% sh bpf_prog_c1bd85c092d6e4aa_sys_exit [k] bpf_prog_c1bd85c092d6e4aa_sys_exit # This was with the default --sort order for 'perf report', which is: --sort comm,dso,symbol If we just look for the symbol, for instance: # perf report --sort symbol | grep bpf_prog | head 0.26% [k] bpf_prog_819967866022f1e1_sys_enter - - 0.24% [k] bpf_prog_c1bd85c092d6e4aa_sys_exit - - # or the DSO: # perf report --sort dso | grep bpf_prog | head 0.26% bpf_prog_819967866022f1e1_sys_enter 0.24% bpf_prog_c1bd85c092d6e4aa_sys_exit # We'll see the two BPF programs that augmented_raw_syscalls.o puts in place, one attached to the raw_syscalls:sys_enter and another to the raw_syscalls:sys_exit tracepoints, as expected. Now we can finally do, from the command line, annotation for one of those two symbols, with the original BPF program source coude intermixed with the disassembled JITed code: # perf annotate --stdio2 bpf_prog_819967866022f1e1_sys_enter Samples: 950 of event 'cycles:ppp', 4000 Hz, Event count (approx.): 553756947, [percent: local period] bpf_prog_819967866022f1e1_sys_enter() bpf_prog_819967866022f1e1_sys_enter Percent int sys_enter(struct syscall_enter_args *args) 53.41 push %rbp 0.63 mov %rsp,%rbp 0.31 sub $0x170,%rsp 1.93 sub $0x28,%rbp 7.02 mov %rbx,0x0(%rbp) 3.20 mov %r13,0x8(%rbp) 1.07 mov %r14,0x10(%rbp) 0.61 mov %r15,0x18(%rbp) 0.11 xor %eax,%eax 1.29 mov %rax,0x20(%rbp) 0.11 mov %rdi,%rbx return bpf_get_current_pid_tgid(); 2.02 → callq *ffffffffda6776d9 2.76 mov %eax,-0x148(%rbp) mov %rbp,%rsi int sys_enter(struct syscall_enter_args *args) add $0xfffffffffffffeb8,%rsi return bpf_map_lookup_elem(pids, &pid) != NULL; movabs $0xffff975ac2607800,%rdi 1.26 → callq *ffffffffda6789e9 cmp $0x0,%rax 2.43 → je 0 add $0x38,%rax 0.21 xor %r13d,%r13d if (pid_filter__has(&pids_filtered, getpid())) 0.81 cmp $0x0,%rax → jne 0 mov %rbp,%rdi probe_read(&augmented_args.args, sizeof(augmented_args.args), args); 2.22 add $0xfffffffffffffeb8,%rdi 0.11 mov $0x40,%esi 0.32 mov %rbx,%rdx 2.74 → callq *ffffffffda658409 syscall = bpf_map_lookup_elem(&syscalls, &augmented_args.args.syscall_nr); 0.22 mov %rbp,%rsi 1.69 add $0xfffffffffffffec0,%rsi syscall = bpf_map_lookup_elem(&syscalls, &augmented_args.args.syscall_nr); movabs $0xffff975bfcd36000,%rdi add $0xd0,%rdi 0.21 mov 0x0(%rsi),%eax 0.93 cmp $0x200,%rax → jae 0 0.10 shl $0x3,%rax 0.11 add %rdi,%rax 0.11 → jmp 0 xor %eax,%eax if (syscall == NULL || !syscall->enabled) 1.07 cmp $0x0,%rax → je 0 if (syscall == NULL || !syscall->enabled) 6.57 movzbq 0x0(%rax),%rdi if (syscall == NULL || !syscall->enabled) cmp $0x0,%rdi 0.95 → je 0 mov $0x40,%r8d switch (augmented_args.args.syscall_nr) { mov -0x140(%rbp),%rdi switch (augmented_args.args.syscall_nr) { cmp $0x2,%rdi → je 0 cmp $0x101,%rdi → je 0 cmp $0x15,%rdi → jne 0 case SYS_OPEN: filename_arg = (const void *)args->args[0]; mov 0x10(%rbx),%rdx → jmp 0 case SYS_OPENAT: filename_arg = (const void *)args->args[1]; mov 0x18(%rbx),%rdx if (filename_arg != NULL) { cmp $0x0,%rdx → je 0 xor %edi,%edi augmented_args.filename.reserved = 0; mov %edi,-0x104(%rbp) augmented_args.filename.size = probe_read_str(&augmented_args.filename.value, mov %rbp,%rdi add $0xffffffffffffff00,%rdi augmented_args.filename.size = probe_read_str(&augmented_args.filename.value, mov $0x100,%esi → callq *ffffffffda658499 mov $0x148,%r8d augmented_args.filename.size = probe_read_str(&augmented_args.filename.value, mov %eax,-0x108(%rbp) augmented_args.filename.size = probe_read_str(&augmented_args.filename.value, mov %rax,%rdi shl $0x20,%rdi shr $0x20,%rdi if (augmented_args.filename.size < sizeof(augmented_args.filename.value)) { cmp $0xff,%rdi → ja 0 len -= sizeof(augmented_args.filename.value) - augmented_args.filename.size; add $0x48,%rax len &= sizeof(augmented_args.filename.value) - 1; and $0xff,%rax mov %rax,%r8 mov %rbp,%rcx return perf_event_output(args, &__augmented_syscalls__, BPF_F_CURRENT_CPU, &augmented_args, len); add $0xfffffffffffffeb8,%rcx mov %rbx,%rdi movabs $0xffff975fbd72d800,%rsi mov $0xffffffff,%edx → callq *ffffffffda658ad9 mov %rax,%r13 } mov %r13,%rax 0.72 mov 0x0(%rbp),%rbx mov 0x8(%rbp),%r13 1.16 mov 0x10(%rbp),%r14 0.10 mov 0x18(%rbp),%r15 0.42 add $0x28,%rbp 0.54 leaveq 0.54 ← retq # Please see 'man perf-config' to see how to control what should be seen, via ~/.perfconfig [annotate] section, for instance, one can suppress the source code and see just the disassembly, etc. Alternatively, use the TUI bu just using 'perf annotate', press '/bpf_prog' to see the bpf symbols, press enter and do the interactive annotation, which allows for dumping to a file after selecting the the various output tunables, for instance, the above without source code intermixed, plus showing all the instruction offsets: # perf annotate bpf_prog_819967866022f1e1_sys_enter Then press: 's' to hide the source code + 'O' twice to show all instruction offsets, then 'P' to print to the bpf_prog_819967866022f1e1_sys_enter.annotation file, which will have: # cat bpf_prog_819967866022f1e1_sys_enter.annotation bpf_prog_819967866022f1e1_sys_enter() bpf_prog_819967866022f1e1_sys_enter Event: cycles:ppp 53.41 0: push %rbp 0.63 1: mov %rsp,%rbp 0.31 4: sub $0x170,%rsp 1.93 b: sub $0x28,%rbp 7.02 f: mov %rbx,0x0(%rbp) 3.20 13: mov %r13,0x8(%rbp) 1.07 17: mov %r14,0x10(%rbp) 0.61 1b: mov %r15,0x18(%rbp) 0.11 1f: xor %eax,%eax 1.29 21: mov %rax,0x20(%rbp) 0.11 25: mov %rdi,%rbx 2.02 28: → callq *ffffffffda6776d9 2.76 2d: mov %eax,-0x148(%rbp) 33: mov %rbp,%rsi 36: add $0xfffffffffffffeb8,%rsi 3d: movabs $0xffff975ac2607800,%rdi 1.26 47: → callq *ffffffffda6789e9 4c: cmp $0x0,%rax 2.43 50: → je 0 52: add $0x38,%rax 0.21 56: xor %r13d,%r13d 0.81 59: cmp $0x0,%rax 5d: → jne 0 63: mov %rbp,%rdi 2.22 66: add $0xfffffffffffffeb8,%rdi 0.11 6d: mov $0x40,%esi 0.32 72: mov %rbx,%rdx 2.74 75: → callq *ffffffffda658409 0.22 7a: mov %rbp,%rsi 1.69 7d: add $0xfffffffffffffec0,%rsi 84: movabs $0xffff975bfcd36000,%rdi 8e: add $0xd0,%rdi 0.21 95: mov 0x0(%rsi),%eax 0.93 98: cmp $0x200,%rax 9f: → jae 0 0.10 a1: shl $0x3,%rax 0.11 a5: add %rdi,%rax 0.11 a8: → jmp 0 aa: xor %eax,%eax 1.07 ac: cmp $0x0,%rax b0: → je 0 6.57 b6: movzbq 0x0(%rax),%rdi bb: cmp $0x0,%rdi 0.95 bf: → je 0 c5: mov $0x40,%r8d cb: mov -0x140(%rbp),%rdi d2: cmp $0x2,%rdi d6: → je 0 d8: cmp $0x101,%rdi df: → je 0 e1: cmp $0x15,%rdi e5: → jne 0 e7: mov 0x10(%rbx),%rdx eb: → jmp 0 ed: mov 0x18(%rbx),%rdx f1: cmp $0x0,%rdx f5: → je 0 f7: xor %edi,%edi f9: mov %edi,-0x104(%rbp) ff: mov %rbp,%rdi 102: add $0xffffffffffffff00,%rdi 109: mov $0x100,%esi 10e: → callq *ffffffffda658499 113: mov $0x148,%r8d 119: mov %eax,-0x108(%rbp) 11f: mov %rax,%rdi 122: shl $0x20,%rdi 126: shr $0x20,%rdi 12a: cmp $0xff,%rdi 131: → ja 0 133: add $0x48,%rax 137: and $0xff,%rax 13d: mov %rax,%r8 140: mov %rbp,%rcx 143: add $0xfffffffffffffeb8,%rcx 14a: mov %rbx,%rdi 14d: movabs $0xffff975fbd72d800,%rsi 157: mov $0xffffffff,%edx 15c: → callq *ffffffffda658ad9 161: mov %rax,%r13 164: mov %r13,%rax 0.72 167: mov 0x0(%rbp),%rbx 16b: mov 0x8(%rbp),%r13 1.16 16f: mov 0x10(%rbp),%r14 0.10 173: mov 0x18(%rbp),%r15 0.42 177: add $0x28,%rbp 0.54 17b: leaveq 0.54 17c: ← retq Another cool way to test all this is to symple use 'perf top' look for those symbols, go there and press enter, annotate it live :-) Signed-off-by: Song Liu <songliubraving@fb.com> Reviewed-by: Jiri Olsa <jolsa@kernel.org> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Alexei Starovoitov <ast@kernel.org> Cc: Daniel Borkmann <daniel@iogearbox.net> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Stanislav Fomichev <sdf@google.com> Link: http://lkml.kernel.org/r/20190312053051.2690567-13-songliubraving@fb.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2019-03-12 08:30:48 +03:00
struct disassemble_info info;
struct dso *dso = map->dso;
int pc = 0, count, sub_id;
struct btf *btf = NULL;
char tpath[PATH_MAX];
size_t buf_size;
int nr_skip = 0;
char *buf;
bfd *bfdf;
int ret;
perf annotate: Enable annotation of BPF programs In symbol__disassemble(), DSO_BINARY_TYPE__BPF_PROG_INFO dso calls into a new function symbol__disassemble_bpf(), where annotation line information is filled based on the bpf_prog_info and btf data saved in given perf_env. symbol__disassemble_bpf() uses binutils's libopcodes to disassemble bpf programs. Committer testing: After fixing this: - u64 *addrs = (u64 *)(info_linear->info.jited_ksyms); + u64 *addrs = (u64 *)(uintptr_t)(info_linear->info.jited_ksyms); Detected when crossbuilding to a 32-bit arch. And making all this dependent on HAVE_LIBBFD_SUPPORT and HAVE_LIBBPF_SUPPORT: 1) Have a BPF program running, one that has BTF info, etc, I used the tools/perf/examples/bpf/augmented_raw_syscalls.c put in place by 'perf trace'. # grep -B1 augmented_raw ~/.perfconfig [trace] add_events = /home/acme/git/perf/tools/perf/examples/bpf/augmented_raw_syscalls.c # # perf trace -e *mmsg dnf/6245 sendmmsg(20, 0x7f5485a88030, 2, MSG_NOSIGNAL) = 2 NetworkManager/10055 sendmmsg(22<socket:[1056822]>, 0x7f8126ad1bb0, 2, MSG_NOSIGNAL) = 2 2) Then do a 'perf record' system wide for a while: # perf record -a ^C[ perf record: Woken up 68 times to write data ] [ perf record: Captured and wrote 19.427 MB perf.data (366891 samples) ] # 3) Check that we captured BPF and BTF info in the perf.data file: # perf report --header-only | grep 'b[pt]f' # event : name = cycles:ppp, , id = { 294789, 294790, 294791, 294792, 294793, 294794, 294795, 294796 }, size = 112, { sample_period, sample_freq } = 4000, sample_type = IP|TID|TIME|CPU|PERIOD, read_format = ID, disabled = 1, inherit = 1, mmap = 1, comm = 1, freq = 1, task = 1, precise_ip = 3, sample_id_all = 1, exclude_guest = 1, mmap2 = 1, comm_exec = 1, ksymbol = 1, bpf_event = 1 # bpf_prog_info of id 13 # bpf_prog_info of id 14 # bpf_prog_info of id 15 # bpf_prog_info of id 16 # bpf_prog_info of id 17 # bpf_prog_info of id 18 # bpf_prog_info of id 21 # bpf_prog_info of id 22 # bpf_prog_info of id 41 # bpf_prog_info of id 42 # btf info of id 2 # 4) Check which programs got recorded: # perf report | grep bpf_prog | head 0.16% exe bpf_prog_819967866022f1e1_sys_enter [k] bpf_prog_819967866022f1e1_sys_enter 0.14% exe bpf_prog_c1bd85c092d6e4aa_sys_exit [k] bpf_prog_c1bd85c092d6e4aa_sys_exit 0.08% fuse-overlayfs bpf_prog_819967866022f1e1_sys_enter [k] bpf_prog_819967866022f1e1_sys_enter 0.07% fuse-overlayfs bpf_prog_c1bd85c092d6e4aa_sys_exit [k] bpf_prog_c1bd85c092d6e4aa_sys_exit 0.01% clang-4.0 bpf_prog_c1bd85c092d6e4aa_sys_exit [k] bpf_prog_c1bd85c092d6e4aa_sys_exit 0.01% clang-4.0 bpf_prog_819967866022f1e1_sys_enter [k] bpf_prog_819967866022f1e1_sys_enter 0.00% clang bpf_prog_c1bd85c092d6e4aa_sys_exit [k] bpf_prog_c1bd85c092d6e4aa_sys_exit 0.00% runc bpf_prog_819967866022f1e1_sys_enter [k] bpf_prog_819967866022f1e1_sys_enter 0.00% clang bpf_prog_819967866022f1e1_sys_enter [k] bpf_prog_819967866022f1e1_sys_enter 0.00% sh bpf_prog_c1bd85c092d6e4aa_sys_exit [k] bpf_prog_c1bd85c092d6e4aa_sys_exit # This was with the default --sort order for 'perf report', which is: --sort comm,dso,symbol If we just look for the symbol, for instance: # perf report --sort symbol | grep bpf_prog | head 0.26% [k] bpf_prog_819967866022f1e1_sys_enter - - 0.24% [k] bpf_prog_c1bd85c092d6e4aa_sys_exit - - # or the DSO: # perf report --sort dso | grep bpf_prog | head 0.26% bpf_prog_819967866022f1e1_sys_enter 0.24% bpf_prog_c1bd85c092d6e4aa_sys_exit # We'll see the two BPF programs that augmented_raw_syscalls.o puts in place, one attached to the raw_syscalls:sys_enter and another to the raw_syscalls:sys_exit tracepoints, as expected. Now we can finally do, from the command line, annotation for one of those two symbols, with the original BPF program source coude intermixed with the disassembled JITed code: # perf annotate --stdio2 bpf_prog_819967866022f1e1_sys_enter Samples: 950 of event 'cycles:ppp', 4000 Hz, Event count (approx.): 553756947, [percent: local period] bpf_prog_819967866022f1e1_sys_enter() bpf_prog_819967866022f1e1_sys_enter Percent int sys_enter(struct syscall_enter_args *args) 53.41 push %rbp 0.63 mov %rsp,%rbp 0.31 sub $0x170,%rsp 1.93 sub $0x28,%rbp 7.02 mov %rbx,0x0(%rbp) 3.20 mov %r13,0x8(%rbp) 1.07 mov %r14,0x10(%rbp) 0.61 mov %r15,0x18(%rbp) 0.11 xor %eax,%eax 1.29 mov %rax,0x20(%rbp) 0.11 mov %rdi,%rbx return bpf_get_current_pid_tgid(); 2.02 → callq *ffffffffda6776d9 2.76 mov %eax,-0x148(%rbp) mov %rbp,%rsi int sys_enter(struct syscall_enter_args *args) add $0xfffffffffffffeb8,%rsi return bpf_map_lookup_elem(pids, &pid) != NULL; movabs $0xffff975ac2607800,%rdi 1.26 → callq *ffffffffda6789e9 cmp $0x0,%rax 2.43 → je 0 add $0x38,%rax 0.21 xor %r13d,%r13d if (pid_filter__has(&pids_filtered, getpid())) 0.81 cmp $0x0,%rax → jne 0 mov %rbp,%rdi probe_read(&augmented_args.args, sizeof(augmented_args.args), args); 2.22 add $0xfffffffffffffeb8,%rdi 0.11 mov $0x40,%esi 0.32 mov %rbx,%rdx 2.74 → callq *ffffffffda658409 syscall = bpf_map_lookup_elem(&syscalls, &augmented_args.args.syscall_nr); 0.22 mov %rbp,%rsi 1.69 add $0xfffffffffffffec0,%rsi syscall = bpf_map_lookup_elem(&syscalls, &augmented_args.args.syscall_nr); movabs $0xffff975bfcd36000,%rdi add $0xd0,%rdi 0.21 mov 0x0(%rsi),%eax 0.93 cmp $0x200,%rax → jae 0 0.10 shl $0x3,%rax 0.11 add %rdi,%rax 0.11 → jmp 0 xor %eax,%eax if (syscall == NULL || !syscall->enabled) 1.07 cmp $0x0,%rax → je 0 if (syscall == NULL || !syscall->enabled) 6.57 movzbq 0x0(%rax),%rdi if (syscall == NULL || !syscall->enabled) cmp $0x0,%rdi 0.95 → je 0 mov $0x40,%r8d switch (augmented_args.args.syscall_nr) { mov -0x140(%rbp),%rdi switch (augmented_args.args.syscall_nr) { cmp $0x2,%rdi → je 0 cmp $0x101,%rdi → je 0 cmp $0x15,%rdi → jne 0 case SYS_OPEN: filename_arg = (const void *)args->args[0]; mov 0x10(%rbx),%rdx → jmp 0 case SYS_OPENAT: filename_arg = (const void *)args->args[1]; mov 0x18(%rbx),%rdx if (filename_arg != NULL) { cmp $0x0,%rdx → je 0 xor %edi,%edi augmented_args.filename.reserved = 0; mov %edi,-0x104(%rbp) augmented_args.filename.size = probe_read_str(&augmented_args.filename.value, mov %rbp,%rdi add $0xffffffffffffff00,%rdi augmented_args.filename.size = probe_read_str(&augmented_args.filename.value, mov $0x100,%esi → callq *ffffffffda658499 mov $0x148,%r8d augmented_args.filename.size = probe_read_str(&augmented_args.filename.value, mov %eax,-0x108(%rbp) augmented_args.filename.size = probe_read_str(&augmented_args.filename.value, mov %rax,%rdi shl $0x20,%rdi shr $0x20,%rdi if (augmented_args.filename.size < sizeof(augmented_args.filename.value)) { cmp $0xff,%rdi → ja 0 len -= sizeof(augmented_args.filename.value) - augmented_args.filename.size; add $0x48,%rax len &= sizeof(augmented_args.filename.value) - 1; and $0xff,%rax mov %rax,%r8 mov %rbp,%rcx return perf_event_output(args, &__augmented_syscalls__, BPF_F_CURRENT_CPU, &augmented_args, len); add $0xfffffffffffffeb8,%rcx mov %rbx,%rdi movabs $0xffff975fbd72d800,%rsi mov $0xffffffff,%edx → callq *ffffffffda658ad9 mov %rax,%r13 } mov %r13,%rax 0.72 mov 0x0(%rbp),%rbx mov 0x8(%rbp),%r13 1.16 mov 0x10(%rbp),%r14 0.10 mov 0x18(%rbp),%r15 0.42 add $0x28,%rbp 0.54 leaveq 0.54 ← retq # Please see 'man perf-config' to see how to control what should be seen, via ~/.perfconfig [annotate] section, for instance, one can suppress the source code and see just the disassembly, etc. Alternatively, use the TUI bu just using 'perf annotate', press '/bpf_prog' to see the bpf symbols, press enter and do the interactive annotation, which allows for dumping to a file after selecting the the various output tunables, for instance, the above without source code intermixed, plus showing all the instruction offsets: # perf annotate bpf_prog_819967866022f1e1_sys_enter Then press: 's' to hide the source code + 'O' twice to show all instruction offsets, then 'P' to print to the bpf_prog_819967866022f1e1_sys_enter.annotation file, which will have: # cat bpf_prog_819967866022f1e1_sys_enter.annotation bpf_prog_819967866022f1e1_sys_enter() bpf_prog_819967866022f1e1_sys_enter Event: cycles:ppp 53.41 0: push %rbp 0.63 1: mov %rsp,%rbp 0.31 4: sub $0x170,%rsp 1.93 b: sub $0x28,%rbp 7.02 f: mov %rbx,0x0(%rbp) 3.20 13: mov %r13,0x8(%rbp) 1.07 17: mov %r14,0x10(%rbp) 0.61 1b: mov %r15,0x18(%rbp) 0.11 1f: xor %eax,%eax 1.29 21: mov %rax,0x20(%rbp) 0.11 25: mov %rdi,%rbx 2.02 28: → callq *ffffffffda6776d9 2.76 2d: mov %eax,-0x148(%rbp) 33: mov %rbp,%rsi 36: add $0xfffffffffffffeb8,%rsi 3d: movabs $0xffff975ac2607800,%rdi 1.26 47: → callq *ffffffffda6789e9 4c: cmp $0x0,%rax 2.43 50: → je 0 52: add $0x38,%rax 0.21 56: xor %r13d,%r13d 0.81 59: cmp $0x0,%rax 5d: → jne 0 63: mov %rbp,%rdi 2.22 66: add $0xfffffffffffffeb8,%rdi 0.11 6d: mov $0x40,%esi 0.32 72: mov %rbx,%rdx 2.74 75: → callq *ffffffffda658409 0.22 7a: mov %rbp,%rsi 1.69 7d: add $0xfffffffffffffec0,%rsi 84: movabs $0xffff975bfcd36000,%rdi 8e: add $0xd0,%rdi 0.21 95: mov 0x0(%rsi),%eax 0.93 98: cmp $0x200,%rax 9f: → jae 0 0.10 a1: shl $0x3,%rax 0.11 a5: add %rdi,%rax 0.11 a8: → jmp 0 aa: xor %eax,%eax 1.07 ac: cmp $0x0,%rax b0: → je 0 6.57 b6: movzbq 0x0(%rax),%rdi bb: cmp $0x0,%rdi 0.95 bf: → je 0 c5: mov $0x40,%r8d cb: mov -0x140(%rbp),%rdi d2: cmp $0x2,%rdi d6: → je 0 d8: cmp $0x101,%rdi df: → je 0 e1: cmp $0x15,%rdi e5: → jne 0 e7: mov 0x10(%rbx),%rdx eb: → jmp 0 ed: mov 0x18(%rbx),%rdx f1: cmp $0x0,%rdx f5: → je 0 f7: xor %edi,%edi f9: mov %edi,-0x104(%rbp) ff: mov %rbp,%rdi 102: add $0xffffffffffffff00,%rdi 109: mov $0x100,%esi 10e: → callq *ffffffffda658499 113: mov $0x148,%r8d 119: mov %eax,-0x108(%rbp) 11f: mov %rax,%rdi 122: shl $0x20,%rdi 126: shr $0x20,%rdi 12a: cmp $0xff,%rdi 131: → ja 0 133: add $0x48,%rax 137: and $0xff,%rax 13d: mov %rax,%r8 140: mov %rbp,%rcx 143: add $0xfffffffffffffeb8,%rcx 14a: mov %rbx,%rdi 14d: movabs $0xffff975fbd72d800,%rsi 157: mov $0xffffffff,%edx 15c: → callq *ffffffffda658ad9 161: mov %rax,%r13 164: mov %r13,%rax 0.72 167: mov 0x0(%rbp),%rbx 16b: mov 0x8(%rbp),%r13 1.16 16f: mov 0x10(%rbp),%r14 0.10 173: mov 0x18(%rbp),%r15 0.42 177: add $0x28,%rbp 0.54 17b: leaveq 0.54 17c: ← retq Another cool way to test all this is to symple use 'perf top' look for those symbols, go there and press enter, annotate it live :-) Signed-off-by: Song Liu <songliubraving@fb.com> Reviewed-by: Jiri Olsa <jolsa@kernel.org> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Alexei Starovoitov <ast@kernel.org> Cc: Daniel Borkmann <daniel@iogearbox.net> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Stanislav Fomichev <sdf@google.com> Link: http://lkml.kernel.org/r/20190312053051.2690567-13-songliubraving@fb.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2019-03-12 08:30:48 +03:00
FILE *s;
if (dso->binary_type != DSO_BINARY_TYPE__BPF_PROG_INFO)
return SYMBOL_ANNOTATE_ERRNO__BPF_INVALID_FILE;
perf annotate: Enable annotation of BPF programs In symbol__disassemble(), DSO_BINARY_TYPE__BPF_PROG_INFO dso calls into a new function symbol__disassemble_bpf(), where annotation line information is filled based on the bpf_prog_info and btf data saved in given perf_env. symbol__disassemble_bpf() uses binutils's libopcodes to disassemble bpf programs. Committer testing: After fixing this: - u64 *addrs = (u64 *)(info_linear->info.jited_ksyms); + u64 *addrs = (u64 *)(uintptr_t)(info_linear->info.jited_ksyms); Detected when crossbuilding to a 32-bit arch. And making all this dependent on HAVE_LIBBFD_SUPPORT and HAVE_LIBBPF_SUPPORT: 1) Have a BPF program running, one that has BTF info, etc, I used the tools/perf/examples/bpf/augmented_raw_syscalls.c put in place by 'perf trace'. # grep -B1 augmented_raw ~/.perfconfig [trace] add_events = /home/acme/git/perf/tools/perf/examples/bpf/augmented_raw_syscalls.c # # perf trace -e *mmsg dnf/6245 sendmmsg(20, 0x7f5485a88030, 2, MSG_NOSIGNAL) = 2 NetworkManager/10055 sendmmsg(22<socket:[1056822]>, 0x7f8126ad1bb0, 2, MSG_NOSIGNAL) = 2 2) Then do a 'perf record' system wide for a while: # perf record -a ^C[ perf record: Woken up 68 times to write data ] [ perf record: Captured and wrote 19.427 MB perf.data (366891 samples) ] # 3) Check that we captured BPF and BTF info in the perf.data file: # perf report --header-only | grep 'b[pt]f' # event : name = cycles:ppp, , id = { 294789, 294790, 294791, 294792, 294793, 294794, 294795, 294796 }, size = 112, { sample_period, sample_freq } = 4000, sample_type = IP|TID|TIME|CPU|PERIOD, read_format = ID, disabled = 1, inherit = 1, mmap = 1, comm = 1, freq = 1, task = 1, precise_ip = 3, sample_id_all = 1, exclude_guest = 1, mmap2 = 1, comm_exec = 1, ksymbol = 1, bpf_event = 1 # bpf_prog_info of id 13 # bpf_prog_info of id 14 # bpf_prog_info of id 15 # bpf_prog_info of id 16 # bpf_prog_info of id 17 # bpf_prog_info of id 18 # bpf_prog_info of id 21 # bpf_prog_info of id 22 # bpf_prog_info of id 41 # bpf_prog_info of id 42 # btf info of id 2 # 4) Check which programs got recorded: # perf report | grep bpf_prog | head 0.16% exe bpf_prog_819967866022f1e1_sys_enter [k] bpf_prog_819967866022f1e1_sys_enter 0.14% exe bpf_prog_c1bd85c092d6e4aa_sys_exit [k] bpf_prog_c1bd85c092d6e4aa_sys_exit 0.08% fuse-overlayfs bpf_prog_819967866022f1e1_sys_enter [k] bpf_prog_819967866022f1e1_sys_enter 0.07% fuse-overlayfs bpf_prog_c1bd85c092d6e4aa_sys_exit [k] bpf_prog_c1bd85c092d6e4aa_sys_exit 0.01% clang-4.0 bpf_prog_c1bd85c092d6e4aa_sys_exit [k] bpf_prog_c1bd85c092d6e4aa_sys_exit 0.01% clang-4.0 bpf_prog_819967866022f1e1_sys_enter [k] bpf_prog_819967866022f1e1_sys_enter 0.00% clang bpf_prog_c1bd85c092d6e4aa_sys_exit [k] bpf_prog_c1bd85c092d6e4aa_sys_exit 0.00% runc bpf_prog_819967866022f1e1_sys_enter [k] bpf_prog_819967866022f1e1_sys_enter 0.00% clang bpf_prog_819967866022f1e1_sys_enter [k] bpf_prog_819967866022f1e1_sys_enter 0.00% sh bpf_prog_c1bd85c092d6e4aa_sys_exit [k] bpf_prog_c1bd85c092d6e4aa_sys_exit # This was with the default --sort order for 'perf report', which is: --sort comm,dso,symbol If we just look for the symbol, for instance: # perf report --sort symbol | grep bpf_prog | head 0.26% [k] bpf_prog_819967866022f1e1_sys_enter - - 0.24% [k] bpf_prog_c1bd85c092d6e4aa_sys_exit - - # or the DSO: # perf report --sort dso | grep bpf_prog | head 0.26% bpf_prog_819967866022f1e1_sys_enter 0.24% bpf_prog_c1bd85c092d6e4aa_sys_exit # We'll see the two BPF programs that augmented_raw_syscalls.o puts in place, one attached to the raw_syscalls:sys_enter and another to the raw_syscalls:sys_exit tracepoints, as expected. Now we can finally do, from the command line, annotation for one of those two symbols, with the original BPF program source coude intermixed with the disassembled JITed code: # perf annotate --stdio2 bpf_prog_819967866022f1e1_sys_enter Samples: 950 of event 'cycles:ppp', 4000 Hz, Event count (approx.): 553756947, [percent: local period] bpf_prog_819967866022f1e1_sys_enter() bpf_prog_819967866022f1e1_sys_enter Percent int sys_enter(struct syscall_enter_args *args) 53.41 push %rbp 0.63 mov %rsp,%rbp 0.31 sub $0x170,%rsp 1.93 sub $0x28,%rbp 7.02 mov %rbx,0x0(%rbp) 3.20 mov %r13,0x8(%rbp) 1.07 mov %r14,0x10(%rbp) 0.61 mov %r15,0x18(%rbp) 0.11 xor %eax,%eax 1.29 mov %rax,0x20(%rbp) 0.11 mov %rdi,%rbx return bpf_get_current_pid_tgid(); 2.02 → callq *ffffffffda6776d9 2.76 mov %eax,-0x148(%rbp) mov %rbp,%rsi int sys_enter(struct syscall_enter_args *args) add $0xfffffffffffffeb8,%rsi return bpf_map_lookup_elem(pids, &pid) != NULL; movabs $0xffff975ac2607800,%rdi 1.26 → callq *ffffffffda6789e9 cmp $0x0,%rax 2.43 → je 0 add $0x38,%rax 0.21 xor %r13d,%r13d if (pid_filter__has(&pids_filtered, getpid())) 0.81 cmp $0x0,%rax → jne 0 mov %rbp,%rdi probe_read(&augmented_args.args, sizeof(augmented_args.args), args); 2.22 add $0xfffffffffffffeb8,%rdi 0.11 mov $0x40,%esi 0.32 mov %rbx,%rdx 2.74 → callq *ffffffffda658409 syscall = bpf_map_lookup_elem(&syscalls, &augmented_args.args.syscall_nr); 0.22 mov %rbp,%rsi 1.69 add $0xfffffffffffffec0,%rsi syscall = bpf_map_lookup_elem(&syscalls, &augmented_args.args.syscall_nr); movabs $0xffff975bfcd36000,%rdi add $0xd0,%rdi 0.21 mov 0x0(%rsi),%eax 0.93 cmp $0x200,%rax → jae 0 0.10 shl $0x3,%rax 0.11 add %rdi,%rax 0.11 → jmp 0 xor %eax,%eax if (syscall == NULL || !syscall->enabled) 1.07 cmp $0x0,%rax → je 0 if (syscall == NULL || !syscall->enabled) 6.57 movzbq 0x0(%rax),%rdi if (syscall == NULL || !syscall->enabled) cmp $0x0,%rdi 0.95 → je 0 mov $0x40,%r8d switch (augmented_args.args.syscall_nr) { mov -0x140(%rbp),%rdi switch (augmented_args.args.syscall_nr) { cmp $0x2,%rdi → je 0 cmp $0x101,%rdi → je 0 cmp $0x15,%rdi → jne 0 case SYS_OPEN: filename_arg = (const void *)args->args[0]; mov 0x10(%rbx),%rdx → jmp 0 case SYS_OPENAT: filename_arg = (const void *)args->args[1]; mov 0x18(%rbx),%rdx if (filename_arg != NULL) { cmp $0x0,%rdx → je 0 xor %edi,%edi augmented_args.filename.reserved = 0; mov %edi,-0x104(%rbp) augmented_args.filename.size = probe_read_str(&augmented_args.filename.value, mov %rbp,%rdi add $0xffffffffffffff00,%rdi augmented_args.filename.size = probe_read_str(&augmented_args.filename.value, mov $0x100,%esi → callq *ffffffffda658499 mov $0x148,%r8d augmented_args.filename.size = probe_read_str(&augmented_args.filename.value, mov %eax,-0x108(%rbp) augmented_args.filename.size = probe_read_str(&augmented_args.filename.value, mov %rax,%rdi shl $0x20,%rdi shr $0x20,%rdi if (augmented_args.filename.size < sizeof(augmented_args.filename.value)) { cmp $0xff,%rdi → ja 0 len -= sizeof(augmented_args.filename.value) - augmented_args.filename.size; add $0x48,%rax len &= sizeof(augmented_args.filename.value) - 1; and $0xff,%rax mov %rax,%r8 mov %rbp,%rcx return perf_event_output(args, &__augmented_syscalls__, BPF_F_CURRENT_CPU, &augmented_args, len); add $0xfffffffffffffeb8,%rcx mov %rbx,%rdi movabs $0xffff975fbd72d800,%rsi mov $0xffffffff,%edx → callq *ffffffffda658ad9 mov %rax,%r13 } mov %r13,%rax 0.72 mov 0x0(%rbp),%rbx mov 0x8(%rbp),%r13 1.16 mov 0x10(%rbp),%r14 0.10 mov 0x18(%rbp),%r15 0.42 add $0x28,%rbp 0.54 leaveq 0.54 ← retq # Please see 'man perf-config' to see how to control what should be seen, via ~/.perfconfig [annotate] section, for instance, one can suppress the source code and see just the disassembly, etc. Alternatively, use the TUI bu just using 'perf annotate', press '/bpf_prog' to see the bpf symbols, press enter and do the interactive annotation, which allows for dumping to a file after selecting the the various output tunables, for instance, the above without source code intermixed, plus showing all the instruction offsets: # perf annotate bpf_prog_819967866022f1e1_sys_enter Then press: 's' to hide the source code + 'O' twice to show all instruction offsets, then 'P' to print to the bpf_prog_819967866022f1e1_sys_enter.annotation file, which will have: # cat bpf_prog_819967866022f1e1_sys_enter.annotation bpf_prog_819967866022f1e1_sys_enter() bpf_prog_819967866022f1e1_sys_enter Event: cycles:ppp 53.41 0: push %rbp 0.63 1: mov %rsp,%rbp 0.31 4: sub $0x170,%rsp 1.93 b: sub $0x28,%rbp 7.02 f: mov %rbx,0x0(%rbp) 3.20 13: mov %r13,0x8(%rbp) 1.07 17: mov %r14,0x10(%rbp) 0.61 1b: mov %r15,0x18(%rbp) 0.11 1f: xor %eax,%eax 1.29 21: mov %rax,0x20(%rbp) 0.11 25: mov %rdi,%rbx 2.02 28: → callq *ffffffffda6776d9 2.76 2d: mov %eax,-0x148(%rbp) 33: mov %rbp,%rsi 36: add $0xfffffffffffffeb8,%rsi 3d: movabs $0xffff975ac2607800,%rdi 1.26 47: → callq *ffffffffda6789e9 4c: cmp $0x0,%rax 2.43 50: → je 0 52: add $0x38,%rax 0.21 56: xor %r13d,%r13d 0.81 59: cmp $0x0,%rax 5d: → jne 0 63: mov %rbp,%rdi 2.22 66: add $0xfffffffffffffeb8,%rdi 0.11 6d: mov $0x40,%esi 0.32 72: mov %rbx,%rdx 2.74 75: → callq *ffffffffda658409 0.22 7a: mov %rbp,%rsi 1.69 7d: add $0xfffffffffffffec0,%rsi 84: movabs $0xffff975bfcd36000,%rdi 8e: add $0xd0,%rdi 0.21 95: mov 0x0(%rsi),%eax 0.93 98: cmp $0x200,%rax 9f: → jae 0 0.10 a1: shl $0x3,%rax 0.11 a5: add %rdi,%rax 0.11 a8: → jmp 0 aa: xor %eax,%eax 1.07 ac: cmp $0x0,%rax b0: → je 0 6.57 b6: movzbq 0x0(%rax),%rdi bb: cmp $0x0,%rdi 0.95 bf: → je 0 c5: mov $0x40,%r8d cb: mov -0x140(%rbp),%rdi d2: cmp $0x2,%rdi d6: → je 0 d8: cmp $0x101,%rdi df: → je 0 e1: cmp $0x15,%rdi e5: → jne 0 e7: mov 0x10(%rbx),%rdx eb: → jmp 0 ed: mov 0x18(%rbx),%rdx f1: cmp $0x0,%rdx f5: → je 0 f7: xor %edi,%edi f9: mov %edi,-0x104(%rbp) ff: mov %rbp,%rdi 102: add $0xffffffffffffff00,%rdi 109: mov $0x100,%esi 10e: → callq *ffffffffda658499 113: mov $0x148,%r8d 119: mov %eax,-0x108(%rbp) 11f: mov %rax,%rdi 122: shl $0x20,%rdi 126: shr $0x20,%rdi 12a: cmp $0xff,%rdi 131: → ja 0 133: add $0x48,%rax 137: and $0xff,%rax 13d: mov %rax,%r8 140: mov %rbp,%rcx 143: add $0xfffffffffffffeb8,%rcx 14a: mov %rbx,%rdi 14d: movabs $0xffff975fbd72d800,%rsi 157: mov $0xffffffff,%edx 15c: → callq *ffffffffda658ad9 161: mov %rax,%r13 164: mov %r13,%rax 0.72 167: mov 0x0(%rbp),%rbx 16b: mov 0x8(%rbp),%r13 1.16 16f: mov 0x10(%rbp),%r14 0.10 173: mov 0x18(%rbp),%r15 0.42 177: add $0x28,%rbp 0.54 17b: leaveq 0.54 17c: ← retq Another cool way to test all this is to symple use 'perf top' look for those symbols, go there and press enter, annotate it live :-) Signed-off-by: Song Liu <songliubraving@fb.com> Reviewed-by: Jiri Olsa <jolsa@kernel.org> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Alexei Starovoitov <ast@kernel.org> Cc: Daniel Borkmann <daniel@iogearbox.net> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Stanislav Fomichev <sdf@google.com> Link: http://lkml.kernel.org/r/20190312053051.2690567-13-songliubraving@fb.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2019-03-12 08:30:48 +03:00
pr_debug("%s: handling sym %s addr %" PRIx64 " len %" PRIx64 "\n", __func__,
sym->name, sym->start, sym->end - sym->start);
perf annotate: Enable annotation of BPF programs In symbol__disassemble(), DSO_BINARY_TYPE__BPF_PROG_INFO dso calls into a new function symbol__disassemble_bpf(), where annotation line information is filled based on the bpf_prog_info and btf data saved in given perf_env. symbol__disassemble_bpf() uses binutils's libopcodes to disassemble bpf programs. Committer testing: After fixing this: - u64 *addrs = (u64 *)(info_linear->info.jited_ksyms); + u64 *addrs = (u64 *)(uintptr_t)(info_linear->info.jited_ksyms); Detected when crossbuilding to a 32-bit arch. And making all this dependent on HAVE_LIBBFD_SUPPORT and HAVE_LIBBPF_SUPPORT: 1) Have a BPF program running, one that has BTF info, etc, I used the tools/perf/examples/bpf/augmented_raw_syscalls.c put in place by 'perf trace'. # grep -B1 augmented_raw ~/.perfconfig [trace] add_events = /home/acme/git/perf/tools/perf/examples/bpf/augmented_raw_syscalls.c # # perf trace -e *mmsg dnf/6245 sendmmsg(20, 0x7f5485a88030, 2, MSG_NOSIGNAL) = 2 NetworkManager/10055 sendmmsg(22<socket:[1056822]>, 0x7f8126ad1bb0, 2, MSG_NOSIGNAL) = 2 2) Then do a 'perf record' system wide for a while: # perf record -a ^C[ perf record: Woken up 68 times to write data ] [ perf record: Captured and wrote 19.427 MB perf.data (366891 samples) ] # 3) Check that we captured BPF and BTF info in the perf.data file: # perf report --header-only | grep 'b[pt]f' # event : name = cycles:ppp, , id = { 294789, 294790, 294791, 294792, 294793, 294794, 294795, 294796 }, size = 112, { sample_period, sample_freq } = 4000, sample_type = IP|TID|TIME|CPU|PERIOD, read_format = ID, disabled = 1, inherit = 1, mmap = 1, comm = 1, freq = 1, task = 1, precise_ip = 3, sample_id_all = 1, exclude_guest = 1, mmap2 = 1, comm_exec = 1, ksymbol = 1, bpf_event = 1 # bpf_prog_info of id 13 # bpf_prog_info of id 14 # bpf_prog_info of id 15 # bpf_prog_info of id 16 # bpf_prog_info of id 17 # bpf_prog_info of id 18 # bpf_prog_info of id 21 # bpf_prog_info of id 22 # bpf_prog_info of id 41 # bpf_prog_info of id 42 # btf info of id 2 # 4) Check which programs got recorded: # perf report | grep bpf_prog | head 0.16% exe bpf_prog_819967866022f1e1_sys_enter [k] bpf_prog_819967866022f1e1_sys_enter 0.14% exe bpf_prog_c1bd85c092d6e4aa_sys_exit [k] bpf_prog_c1bd85c092d6e4aa_sys_exit 0.08% fuse-overlayfs bpf_prog_819967866022f1e1_sys_enter [k] bpf_prog_819967866022f1e1_sys_enter 0.07% fuse-overlayfs bpf_prog_c1bd85c092d6e4aa_sys_exit [k] bpf_prog_c1bd85c092d6e4aa_sys_exit 0.01% clang-4.0 bpf_prog_c1bd85c092d6e4aa_sys_exit [k] bpf_prog_c1bd85c092d6e4aa_sys_exit 0.01% clang-4.0 bpf_prog_819967866022f1e1_sys_enter [k] bpf_prog_819967866022f1e1_sys_enter 0.00% clang bpf_prog_c1bd85c092d6e4aa_sys_exit [k] bpf_prog_c1bd85c092d6e4aa_sys_exit 0.00% runc bpf_prog_819967866022f1e1_sys_enter [k] bpf_prog_819967866022f1e1_sys_enter 0.00% clang bpf_prog_819967866022f1e1_sys_enter [k] bpf_prog_819967866022f1e1_sys_enter 0.00% sh bpf_prog_c1bd85c092d6e4aa_sys_exit [k] bpf_prog_c1bd85c092d6e4aa_sys_exit # This was with the default --sort order for 'perf report', which is: --sort comm,dso,symbol If we just look for the symbol, for instance: # perf report --sort symbol | grep bpf_prog | head 0.26% [k] bpf_prog_819967866022f1e1_sys_enter - - 0.24% [k] bpf_prog_c1bd85c092d6e4aa_sys_exit - - # or the DSO: # perf report --sort dso | grep bpf_prog | head 0.26% bpf_prog_819967866022f1e1_sys_enter 0.24% bpf_prog_c1bd85c092d6e4aa_sys_exit # We'll see the two BPF programs that augmented_raw_syscalls.o puts in place, one attached to the raw_syscalls:sys_enter and another to the raw_syscalls:sys_exit tracepoints, as expected. Now we can finally do, from the command line, annotation for one of those two symbols, with the original BPF program source coude intermixed with the disassembled JITed code: # perf annotate --stdio2 bpf_prog_819967866022f1e1_sys_enter Samples: 950 of event 'cycles:ppp', 4000 Hz, Event count (approx.): 553756947, [percent: local period] bpf_prog_819967866022f1e1_sys_enter() bpf_prog_819967866022f1e1_sys_enter Percent int sys_enter(struct syscall_enter_args *args) 53.41 push %rbp 0.63 mov %rsp,%rbp 0.31 sub $0x170,%rsp 1.93 sub $0x28,%rbp 7.02 mov %rbx,0x0(%rbp) 3.20 mov %r13,0x8(%rbp) 1.07 mov %r14,0x10(%rbp) 0.61 mov %r15,0x18(%rbp) 0.11 xor %eax,%eax 1.29 mov %rax,0x20(%rbp) 0.11 mov %rdi,%rbx return bpf_get_current_pid_tgid(); 2.02 → callq *ffffffffda6776d9 2.76 mov %eax,-0x148(%rbp) mov %rbp,%rsi int sys_enter(struct syscall_enter_args *args) add $0xfffffffffffffeb8,%rsi return bpf_map_lookup_elem(pids, &pid) != NULL; movabs $0xffff975ac2607800,%rdi 1.26 → callq *ffffffffda6789e9 cmp $0x0,%rax 2.43 → je 0 add $0x38,%rax 0.21 xor %r13d,%r13d if (pid_filter__has(&pids_filtered, getpid())) 0.81 cmp $0x0,%rax → jne 0 mov %rbp,%rdi probe_read(&augmented_args.args, sizeof(augmented_args.args), args); 2.22 add $0xfffffffffffffeb8,%rdi 0.11 mov $0x40,%esi 0.32 mov %rbx,%rdx 2.74 → callq *ffffffffda658409 syscall = bpf_map_lookup_elem(&syscalls, &augmented_args.args.syscall_nr); 0.22 mov %rbp,%rsi 1.69 add $0xfffffffffffffec0,%rsi syscall = bpf_map_lookup_elem(&syscalls, &augmented_args.args.syscall_nr); movabs $0xffff975bfcd36000,%rdi add $0xd0,%rdi 0.21 mov 0x0(%rsi),%eax 0.93 cmp $0x200,%rax → jae 0 0.10 shl $0x3,%rax 0.11 add %rdi,%rax 0.11 → jmp 0 xor %eax,%eax if (syscall == NULL || !syscall->enabled) 1.07 cmp $0x0,%rax → je 0 if (syscall == NULL || !syscall->enabled) 6.57 movzbq 0x0(%rax),%rdi if (syscall == NULL || !syscall->enabled) cmp $0x0,%rdi 0.95 → je 0 mov $0x40,%r8d switch (augmented_args.args.syscall_nr) { mov -0x140(%rbp),%rdi switch (augmented_args.args.syscall_nr) { cmp $0x2,%rdi → je 0 cmp $0x101,%rdi → je 0 cmp $0x15,%rdi → jne 0 case SYS_OPEN: filename_arg = (const void *)args->args[0]; mov 0x10(%rbx),%rdx → jmp 0 case SYS_OPENAT: filename_arg = (const void *)args->args[1]; mov 0x18(%rbx),%rdx if (filename_arg != NULL) { cmp $0x0,%rdx → je 0 xor %edi,%edi augmented_args.filename.reserved = 0; mov %edi,-0x104(%rbp) augmented_args.filename.size = probe_read_str(&augmented_args.filename.value, mov %rbp,%rdi add $0xffffffffffffff00,%rdi augmented_args.filename.size = probe_read_str(&augmented_args.filename.value, mov $0x100,%esi → callq *ffffffffda658499 mov $0x148,%r8d augmented_args.filename.size = probe_read_str(&augmented_args.filename.value, mov %eax,-0x108(%rbp) augmented_args.filename.size = probe_read_str(&augmented_args.filename.value, mov %rax,%rdi shl $0x20,%rdi shr $0x20,%rdi if (augmented_args.filename.size < sizeof(augmented_args.filename.value)) { cmp $0xff,%rdi → ja 0 len -= sizeof(augmented_args.filename.value) - augmented_args.filename.size; add $0x48,%rax len &= sizeof(augmented_args.filename.value) - 1; and $0xff,%rax mov %rax,%r8 mov %rbp,%rcx return perf_event_output(args, &__augmented_syscalls__, BPF_F_CURRENT_CPU, &augmented_args, len); add $0xfffffffffffffeb8,%rcx mov %rbx,%rdi movabs $0xffff975fbd72d800,%rsi mov $0xffffffff,%edx → callq *ffffffffda658ad9 mov %rax,%r13 } mov %r13,%rax 0.72 mov 0x0(%rbp),%rbx mov 0x8(%rbp),%r13 1.16 mov 0x10(%rbp),%r14 0.10 mov 0x18(%rbp),%r15 0.42 add $0x28,%rbp 0.54 leaveq 0.54 ← retq # Please see 'man perf-config' to see how to control what should be seen, via ~/.perfconfig [annotate] section, for instance, one can suppress the source code and see just the disassembly, etc. Alternatively, use the TUI bu just using 'perf annotate', press '/bpf_prog' to see the bpf symbols, press enter and do the interactive annotation, which allows for dumping to a file after selecting the the various output tunables, for instance, the above without source code intermixed, plus showing all the instruction offsets: # perf annotate bpf_prog_819967866022f1e1_sys_enter Then press: 's' to hide the source code + 'O' twice to show all instruction offsets, then 'P' to print to the bpf_prog_819967866022f1e1_sys_enter.annotation file, which will have: # cat bpf_prog_819967866022f1e1_sys_enter.annotation bpf_prog_819967866022f1e1_sys_enter() bpf_prog_819967866022f1e1_sys_enter Event: cycles:ppp 53.41 0: push %rbp 0.63 1: mov %rsp,%rbp 0.31 4: sub $0x170,%rsp 1.93 b: sub $0x28,%rbp 7.02 f: mov %rbx,0x0(%rbp) 3.20 13: mov %r13,0x8(%rbp) 1.07 17: mov %r14,0x10(%rbp) 0.61 1b: mov %r15,0x18(%rbp) 0.11 1f: xor %eax,%eax 1.29 21: mov %rax,0x20(%rbp) 0.11 25: mov %rdi,%rbx 2.02 28: → callq *ffffffffda6776d9 2.76 2d: mov %eax,-0x148(%rbp) 33: mov %rbp,%rsi 36: add $0xfffffffffffffeb8,%rsi 3d: movabs $0xffff975ac2607800,%rdi 1.26 47: → callq *ffffffffda6789e9 4c: cmp $0x0,%rax 2.43 50: → je 0 52: add $0x38,%rax 0.21 56: xor %r13d,%r13d 0.81 59: cmp $0x0,%rax 5d: → jne 0 63: mov %rbp,%rdi 2.22 66: add $0xfffffffffffffeb8,%rdi 0.11 6d: mov $0x40,%esi 0.32 72: mov %rbx,%rdx 2.74 75: → callq *ffffffffda658409 0.22 7a: mov %rbp,%rsi 1.69 7d: add $0xfffffffffffffec0,%rsi 84: movabs $0xffff975bfcd36000,%rdi 8e: add $0xd0,%rdi 0.21 95: mov 0x0(%rsi),%eax 0.93 98: cmp $0x200,%rax 9f: → jae 0 0.10 a1: shl $0x3,%rax 0.11 a5: add %rdi,%rax 0.11 a8: → jmp 0 aa: xor %eax,%eax 1.07 ac: cmp $0x0,%rax b0: → je 0 6.57 b6: movzbq 0x0(%rax),%rdi bb: cmp $0x0,%rdi 0.95 bf: → je 0 c5: mov $0x40,%r8d cb: mov -0x140(%rbp),%rdi d2: cmp $0x2,%rdi d6: → je 0 d8: cmp $0x101,%rdi df: → je 0 e1: cmp $0x15,%rdi e5: → jne 0 e7: mov 0x10(%rbx),%rdx eb: → jmp 0 ed: mov 0x18(%rbx),%rdx f1: cmp $0x0,%rdx f5: → je 0 f7: xor %edi,%edi f9: mov %edi,-0x104(%rbp) ff: mov %rbp,%rdi 102: add $0xffffffffffffff00,%rdi 109: mov $0x100,%esi 10e: → callq *ffffffffda658499 113: mov $0x148,%r8d 119: mov %eax,-0x108(%rbp) 11f: mov %rax,%rdi 122: shl $0x20,%rdi 126: shr $0x20,%rdi 12a: cmp $0xff,%rdi 131: → ja 0 133: add $0x48,%rax 137: and $0xff,%rax 13d: mov %rax,%r8 140: mov %rbp,%rcx 143: add $0xfffffffffffffeb8,%rcx 14a: mov %rbx,%rdi 14d: movabs $0xffff975fbd72d800,%rsi 157: mov $0xffffffff,%edx 15c: → callq *ffffffffda658ad9 161: mov %rax,%r13 164: mov %r13,%rax 0.72 167: mov 0x0(%rbp),%rbx 16b: mov 0x8(%rbp),%r13 1.16 16f: mov 0x10(%rbp),%r14 0.10 173: mov 0x18(%rbp),%r15 0.42 177: add $0x28,%rbp 0.54 17b: leaveq 0.54 17c: ← retq Another cool way to test all this is to symple use 'perf top' look for those symbols, go there and press enter, annotate it live :-) Signed-off-by: Song Liu <songliubraving@fb.com> Reviewed-by: Jiri Olsa <jolsa@kernel.org> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Alexei Starovoitov <ast@kernel.org> Cc: Daniel Borkmann <daniel@iogearbox.net> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Stanislav Fomichev <sdf@google.com> Link: http://lkml.kernel.org/r/20190312053051.2690567-13-songliubraving@fb.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2019-03-12 08:30:48 +03:00
memset(tpath, 0, sizeof(tpath));
perf_exe(tpath, sizeof(tpath));
bfdf = bfd_openr(tpath, NULL);
assert(bfdf);
assert(bfd_check_format(bfdf, bfd_object));
s = open_memstream(&buf, &buf_size);
if (!s) {
ret = errno;
perf annotate: Enable annotation of BPF programs In symbol__disassemble(), DSO_BINARY_TYPE__BPF_PROG_INFO dso calls into a new function symbol__disassemble_bpf(), where annotation line information is filled based on the bpf_prog_info and btf data saved in given perf_env. symbol__disassemble_bpf() uses binutils's libopcodes to disassemble bpf programs. Committer testing: After fixing this: - u64 *addrs = (u64 *)(info_linear->info.jited_ksyms); + u64 *addrs = (u64 *)(uintptr_t)(info_linear->info.jited_ksyms); Detected when crossbuilding to a 32-bit arch. And making all this dependent on HAVE_LIBBFD_SUPPORT and HAVE_LIBBPF_SUPPORT: 1) Have a BPF program running, one that has BTF info, etc, I used the tools/perf/examples/bpf/augmented_raw_syscalls.c put in place by 'perf trace'. # grep -B1 augmented_raw ~/.perfconfig [trace] add_events = /home/acme/git/perf/tools/perf/examples/bpf/augmented_raw_syscalls.c # # perf trace -e *mmsg dnf/6245 sendmmsg(20, 0x7f5485a88030, 2, MSG_NOSIGNAL) = 2 NetworkManager/10055 sendmmsg(22<socket:[1056822]>, 0x7f8126ad1bb0, 2, MSG_NOSIGNAL) = 2 2) Then do a 'perf record' system wide for a while: # perf record -a ^C[ perf record: Woken up 68 times to write data ] [ perf record: Captured and wrote 19.427 MB perf.data (366891 samples) ] # 3) Check that we captured BPF and BTF info in the perf.data file: # perf report --header-only | grep 'b[pt]f' # event : name = cycles:ppp, , id = { 294789, 294790, 294791, 294792, 294793, 294794, 294795, 294796 }, size = 112, { sample_period, sample_freq } = 4000, sample_type = IP|TID|TIME|CPU|PERIOD, read_format = ID, disabled = 1, inherit = 1, mmap = 1, comm = 1, freq = 1, task = 1, precise_ip = 3, sample_id_all = 1, exclude_guest = 1, mmap2 = 1, comm_exec = 1, ksymbol = 1, bpf_event = 1 # bpf_prog_info of id 13 # bpf_prog_info of id 14 # bpf_prog_info of id 15 # bpf_prog_info of id 16 # bpf_prog_info of id 17 # bpf_prog_info of id 18 # bpf_prog_info of id 21 # bpf_prog_info of id 22 # bpf_prog_info of id 41 # bpf_prog_info of id 42 # btf info of id 2 # 4) Check which programs got recorded: # perf report | grep bpf_prog | head 0.16% exe bpf_prog_819967866022f1e1_sys_enter [k] bpf_prog_819967866022f1e1_sys_enter 0.14% exe bpf_prog_c1bd85c092d6e4aa_sys_exit [k] bpf_prog_c1bd85c092d6e4aa_sys_exit 0.08% fuse-overlayfs bpf_prog_819967866022f1e1_sys_enter [k] bpf_prog_819967866022f1e1_sys_enter 0.07% fuse-overlayfs bpf_prog_c1bd85c092d6e4aa_sys_exit [k] bpf_prog_c1bd85c092d6e4aa_sys_exit 0.01% clang-4.0 bpf_prog_c1bd85c092d6e4aa_sys_exit [k] bpf_prog_c1bd85c092d6e4aa_sys_exit 0.01% clang-4.0 bpf_prog_819967866022f1e1_sys_enter [k] bpf_prog_819967866022f1e1_sys_enter 0.00% clang bpf_prog_c1bd85c092d6e4aa_sys_exit [k] bpf_prog_c1bd85c092d6e4aa_sys_exit 0.00% runc bpf_prog_819967866022f1e1_sys_enter [k] bpf_prog_819967866022f1e1_sys_enter 0.00% clang bpf_prog_819967866022f1e1_sys_enter [k] bpf_prog_819967866022f1e1_sys_enter 0.00% sh bpf_prog_c1bd85c092d6e4aa_sys_exit [k] bpf_prog_c1bd85c092d6e4aa_sys_exit # This was with the default --sort order for 'perf report', which is: --sort comm,dso,symbol If we just look for the symbol, for instance: # perf report --sort symbol | grep bpf_prog | head 0.26% [k] bpf_prog_819967866022f1e1_sys_enter - - 0.24% [k] bpf_prog_c1bd85c092d6e4aa_sys_exit - - # or the DSO: # perf report --sort dso | grep bpf_prog | head 0.26% bpf_prog_819967866022f1e1_sys_enter 0.24% bpf_prog_c1bd85c092d6e4aa_sys_exit # We'll see the two BPF programs that augmented_raw_syscalls.o puts in place, one attached to the raw_syscalls:sys_enter and another to the raw_syscalls:sys_exit tracepoints, as expected. Now we can finally do, from the command line, annotation for one of those two symbols, with the original BPF program source coude intermixed with the disassembled JITed code: # perf annotate --stdio2 bpf_prog_819967866022f1e1_sys_enter Samples: 950 of event 'cycles:ppp', 4000 Hz, Event count (approx.): 553756947, [percent: local period] bpf_prog_819967866022f1e1_sys_enter() bpf_prog_819967866022f1e1_sys_enter Percent int sys_enter(struct syscall_enter_args *args) 53.41 push %rbp 0.63 mov %rsp,%rbp 0.31 sub $0x170,%rsp 1.93 sub $0x28,%rbp 7.02 mov %rbx,0x0(%rbp) 3.20 mov %r13,0x8(%rbp) 1.07 mov %r14,0x10(%rbp) 0.61 mov %r15,0x18(%rbp) 0.11 xor %eax,%eax 1.29 mov %rax,0x20(%rbp) 0.11 mov %rdi,%rbx return bpf_get_current_pid_tgid(); 2.02 → callq *ffffffffda6776d9 2.76 mov %eax,-0x148(%rbp) mov %rbp,%rsi int sys_enter(struct syscall_enter_args *args) add $0xfffffffffffffeb8,%rsi return bpf_map_lookup_elem(pids, &pid) != NULL; movabs $0xffff975ac2607800,%rdi 1.26 → callq *ffffffffda6789e9 cmp $0x0,%rax 2.43 → je 0 add $0x38,%rax 0.21 xor %r13d,%r13d if (pid_filter__has(&pids_filtered, getpid())) 0.81 cmp $0x0,%rax → jne 0 mov %rbp,%rdi probe_read(&augmented_args.args, sizeof(augmented_args.args), args); 2.22 add $0xfffffffffffffeb8,%rdi 0.11 mov $0x40,%esi 0.32 mov %rbx,%rdx 2.74 → callq *ffffffffda658409 syscall = bpf_map_lookup_elem(&syscalls, &augmented_args.args.syscall_nr); 0.22 mov %rbp,%rsi 1.69 add $0xfffffffffffffec0,%rsi syscall = bpf_map_lookup_elem(&syscalls, &augmented_args.args.syscall_nr); movabs $0xffff975bfcd36000,%rdi add $0xd0,%rdi 0.21 mov 0x0(%rsi),%eax 0.93 cmp $0x200,%rax → jae 0 0.10 shl $0x3,%rax 0.11 add %rdi,%rax 0.11 → jmp 0 xor %eax,%eax if (syscall == NULL || !syscall->enabled) 1.07 cmp $0x0,%rax → je 0 if (syscall == NULL || !syscall->enabled) 6.57 movzbq 0x0(%rax),%rdi if (syscall == NULL || !syscall->enabled) cmp $0x0,%rdi 0.95 → je 0 mov $0x40,%r8d switch (augmented_args.args.syscall_nr) { mov -0x140(%rbp),%rdi switch (augmented_args.args.syscall_nr) { cmp $0x2,%rdi → je 0 cmp $0x101,%rdi → je 0 cmp $0x15,%rdi → jne 0 case SYS_OPEN: filename_arg = (const void *)args->args[0]; mov 0x10(%rbx),%rdx → jmp 0 case SYS_OPENAT: filename_arg = (const void *)args->args[1]; mov 0x18(%rbx),%rdx if (filename_arg != NULL) { cmp $0x0,%rdx → je 0 xor %edi,%edi augmented_args.filename.reserved = 0; mov %edi,-0x104(%rbp) augmented_args.filename.size = probe_read_str(&augmented_args.filename.value, mov %rbp,%rdi add $0xffffffffffffff00,%rdi augmented_args.filename.size = probe_read_str(&augmented_args.filename.value, mov $0x100,%esi → callq *ffffffffda658499 mov $0x148,%r8d augmented_args.filename.size = probe_read_str(&augmented_args.filename.value, mov %eax,-0x108(%rbp) augmented_args.filename.size = probe_read_str(&augmented_args.filename.value, mov %rax,%rdi shl $0x20,%rdi shr $0x20,%rdi if (augmented_args.filename.size < sizeof(augmented_args.filename.value)) { cmp $0xff,%rdi → ja 0 len -= sizeof(augmented_args.filename.value) - augmented_args.filename.size; add $0x48,%rax len &= sizeof(augmented_args.filename.value) - 1; and $0xff,%rax mov %rax,%r8 mov %rbp,%rcx return perf_event_output(args, &__augmented_syscalls__, BPF_F_CURRENT_CPU, &augmented_args, len); add $0xfffffffffffffeb8,%rcx mov %rbx,%rdi movabs $0xffff975fbd72d800,%rsi mov $0xffffffff,%edx → callq *ffffffffda658ad9 mov %rax,%r13 } mov %r13,%rax 0.72 mov 0x0(%rbp),%rbx mov 0x8(%rbp),%r13 1.16 mov 0x10(%rbp),%r14 0.10 mov 0x18(%rbp),%r15 0.42 add $0x28,%rbp 0.54 leaveq 0.54 ← retq # Please see 'man perf-config' to see how to control what should be seen, via ~/.perfconfig [annotate] section, for instance, one can suppress the source code and see just the disassembly, etc. Alternatively, use the TUI bu just using 'perf annotate', press '/bpf_prog' to see the bpf symbols, press enter and do the interactive annotation, which allows for dumping to a file after selecting the the various output tunables, for instance, the above without source code intermixed, plus showing all the instruction offsets: # perf annotate bpf_prog_819967866022f1e1_sys_enter Then press: 's' to hide the source code + 'O' twice to show all instruction offsets, then 'P' to print to the bpf_prog_819967866022f1e1_sys_enter.annotation file, which will have: # cat bpf_prog_819967866022f1e1_sys_enter.annotation bpf_prog_819967866022f1e1_sys_enter() bpf_prog_819967866022f1e1_sys_enter Event: cycles:ppp 53.41 0: push %rbp 0.63 1: mov %rsp,%rbp 0.31 4: sub $0x170,%rsp 1.93 b: sub $0x28,%rbp 7.02 f: mov %rbx,0x0(%rbp) 3.20 13: mov %r13,0x8(%rbp) 1.07 17: mov %r14,0x10(%rbp) 0.61 1b: mov %r15,0x18(%rbp) 0.11 1f: xor %eax,%eax 1.29 21: mov %rax,0x20(%rbp) 0.11 25: mov %rdi,%rbx 2.02 28: → callq *ffffffffda6776d9 2.76 2d: mov %eax,-0x148(%rbp) 33: mov %rbp,%rsi 36: add $0xfffffffffffffeb8,%rsi 3d: movabs $0xffff975ac2607800,%rdi 1.26 47: → callq *ffffffffda6789e9 4c: cmp $0x0,%rax 2.43 50: → je 0 52: add $0x38,%rax 0.21 56: xor %r13d,%r13d 0.81 59: cmp $0x0,%rax 5d: → jne 0 63: mov %rbp,%rdi 2.22 66: add $0xfffffffffffffeb8,%rdi 0.11 6d: mov $0x40,%esi 0.32 72: mov %rbx,%rdx 2.74 75: → callq *ffffffffda658409 0.22 7a: mov %rbp,%rsi 1.69 7d: add $0xfffffffffffffec0,%rsi 84: movabs $0xffff975bfcd36000,%rdi 8e: add $0xd0,%rdi 0.21 95: mov 0x0(%rsi),%eax 0.93 98: cmp $0x200,%rax 9f: → jae 0 0.10 a1: shl $0x3,%rax 0.11 a5: add %rdi,%rax 0.11 a8: → jmp 0 aa: xor %eax,%eax 1.07 ac: cmp $0x0,%rax b0: → je 0 6.57 b6: movzbq 0x0(%rax),%rdi bb: cmp $0x0,%rdi 0.95 bf: → je 0 c5: mov $0x40,%r8d cb: mov -0x140(%rbp),%rdi d2: cmp $0x2,%rdi d6: → je 0 d8: cmp $0x101,%rdi df: → je 0 e1: cmp $0x15,%rdi e5: → jne 0 e7: mov 0x10(%rbx),%rdx eb: → jmp 0 ed: mov 0x18(%rbx),%rdx f1: cmp $0x0,%rdx f5: → je 0 f7: xor %edi,%edi f9: mov %edi,-0x104(%rbp) ff: mov %rbp,%rdi 102: add $0xffffffffffffff00,%rdi 109: mov $0x100,%esi 10e: → callq *ffffffffda658499 113: mov $0x148,%r8d 119: mov %eax,-0x108(%rbp) 11f: mov %rax,%rdi 122: shl $0x20,%rdi 126: shr $0x20,%rdi 12a: cmp $0xff,%rdi 131: → ja 0 133: add $0x48,%rax 137: and $0xff,%rax 13d: mov %rax,%r8 140: mov %rbp,%rcx 143: add $0xfffffffffffffeb8,%rcx 14a: mov %rbx,%rdi 14d: movabs $0xffff975fbd72d800,%rsi 157: mov $0xffffffff,%edx 15c: → callq *ffffffffda658ad9 161: mov %rax,%r13 164: mov %r13,%rax 0.72 167: mov 0x0(%rbp),%rbx 16b: mov 0x8(%rbp),%r13 1.16 16f: mov 0x10(%rbp),%r14 0.10 173: mov 0x18(%rbp),%r15 0.42 177: add $0x28,%rbp 0.54 17b: leaveq 0.54 17c: ← retq Another cool way to test all this is to symple use 'perf top' look for those symbols, go there and press enter, annotate it live :-) Signed-off-by: Song Liu <songliubraving@fb.com> Reviewed-by: Jiri Olsa <jolsa@kernel.org> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Alexei Starovoitov <ast@kernel.org> Cc: Daniel Borkmann <daniel@iogearbox.net> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Stanislav Fomichev <sdf@google.com> Link: http://lkml.kernel.org/r/20190312053051.2690567-13-songliubraving@fb.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2019-03-12 08:30:48 +03:00
goto out;
}
perf annotate: Enable annotation of BPF programs In symbol__disassemble(), DSO_BINARY_TYPE__BPF_PROG_INFO dso calls into a new function symbol__disassemble_bpf(), where annotation line information is filled based on the bpf_prog_info and btf data saved in given perf_env. symbol__disassemble_bpf() uses binutils's libopcodes to disassemble bpf programs. Committer testing: After fixing this: - u64 *addrs = (u64 *)(info_linear->info.jited_ksyms); + u64 *addrs = (u64 *)(uintptr_t)(info_linear->info.jited_ksyms); Detected when crossbuilding to a 32-bit arch. And making all this dependent on HAVE_LIBBFD_SUPPORT and HAVE_LIBBPF_SUPPORT: 1) Have a BPF program running, one that has BTF info, etc, I used the tools/perf/examples/bpf/augmented_raw_syscalls.c put in place by 'perf trace'. # grep -B1 augmented_raw ~/.perfconfig [trace] add_events = /home/acme/git/perf/tools/perf/examples/bpf/augmented_raw_syscalls.c # # perf trace -e *mmsg dnf/6245 sendmmsg(20, 0x7f5485a88030, 2, MSG_NOSIGNAL) = 2 NetworkManager/10055 sendmmsg(22<socket:[1056822]>, 0x7f8126ad1bb0, 2, MSG_NOSIGNAL) = 2 2) Then do a 'perf record' system wide for a while: # perf record -a ^C[ perf record: Woken up 68 times to write data ] [ perf record: Captured and wrote 19.427 MB perf.data (366891 samples) ] # 3) Check that we captured BPF and BTF info in the perf.data file: # perf report --header-only | grep 'b[pt]f' # event : name = cycles:ppp, , id = { 294789, 294790, 294791, 294792, 294793, 294794, 294795, 294796 }, size = 112, { sample_period, sample_freq } = 4000, sample_type = IP|TID|TIME|CPU|PERIOD, read_format = ID, disabled = 1, inherit = 1, mmap = 1, comm = 1, freq = 1, task = 1, precise_ip = 3, sample_id_all = 1, exclude_guest = 1, mmap2 = 1, comm_exec = 1, ksymbol = 1, bpf_event = 1 # bpf_prog_info of id 13 # bpf_prog_info of id 14 # bpf_prog_info of id 15 # bpf_prog_info of id 16 # bpf_prog_info of id 17 # bpf_prog_info of id 18 # bpf_prog_info of id 21 # bpf_prog_info of id 22 # bpf_prog_info of id 41 # bpf_prog_info of id 42 # btf info of id 2 # 4) Check which programs got recorded: # perf report | grep bpf_prog | head 0.16% exe bpf_prog_819967866022f1e1_sys_enter [k] bpf_prog_819967866022f1e1_sys_enter 0.14% exe bpf_prog_c1bd85c092d6e4aa_sys_exit [k] bpf_prog_c1bd85c092d6e4aa_sys_exit 0.08% fuse-overlayfs bpf_prog_819967866022f1e1_sys_enter [k] bpf_prog_819967866022f1e1_sys_enter 0.07% fuse-overlayfs bpf_prog_c1bd85c092d6e4aa_sys_exit [k] bpf_prog_c1bd85c092d6e4aa_sys_exit 0.01% clang-4.0 bpf_prog_c1bd85c092d6e4aa_sys_exit [k] bpf_prog_c1bd85c092d6e4aa_sys_exit 0.01% clang-4.0 bpf_prog_819967866022f1e1_sys_enter [k] bpf_prog_819967866022f1e1_sys_enter 0.00% clang bpf_prog_c1bd85c092d6e4aa_sys_exit [k] bpf_prog_c1bd85c092d6e4aa_sys_exit 0.00% runc bpf_prog_819967866022f1e1_sys_enter [k] bpf_prog_819967866022f1e1_sys_enter 0.00% clang bpf_prog_819967866022f1e1_sys_enter [k] bpf_prog_819967866022f1e1_sys_enter 0.00% sh bpf_prog_c1bd85c092d6e4aa_sys_exit [k] bpf_prog_c1bd85c092d6e4aa_sys_exit # This was with the default --sort order for 'perf report', which is: --sort comm,dso,symbol If we just look for the symbol, for instance: # perf report --sort symbol | grep bpf_prog | head 0.26% [k] bpf_prog_819967866022f1e1_sys_enter - - 0.24% [k] bpf_prog_c1bd85c092d6e4aa_sys_exit - - # or the DSO: # perf report --sort dso | grep bpf_prog | head 0.26% bpf_prog_819967866022f1e1_sys_enter 0.24% bpf_prog_c1bd85c092d6e4aa_sys_exit # We'll see the two BPF programs that augmented_raw_syscalls.o puts in place, one attached to the raw_syscalls:sys_enter and another to the raw_syscalls:sys_exit tracepoints, as expected. Now we can finally do, from the command line, annotation for one of those two symbols, with the original BPF program source coude intermixed with the disassembled JITed code: # perf annotate --stdio2 bpf_prog_819967866022f1e1_sys_enter Samples: 950 of event 'cycles:ppp', 4000 Hz, Event count (approx.): 553756947, [percent: local period] bpf_prog_819967866022f1e1_sys_enter() bpf_prog_819967866022f1e1_sys_enter Percent int sys_enter(struct syscall_enter_args *args) 53.41 push %rbp 0.63 mov %rsp,%rbp 0.31 sub $0x170,%rsp 1.93 sub $0x28,%rbp 7.02 mov %rbx,0x0(%rbp) 3.20 mov %r13,0x8(%rbp) 1.07 mov %r14,0x10(%rbp) 0.61 mov %r15,0x18(%rbp) 0.11 xor %eax,%eax 1.29 mov %rax,0x20(%rbp) 0.11 mov %rdi,%rbx return bpf_get_current_pid_tgid(); 2.02 → callq *ffffffffda6776d9 2.76 mov %eax,-0x148(%rbp) mov %rbp,%rsi int sys_enter(struct syscall_enter_args *args) add $0xfffffffffffffeb8,%rsi return bpf_map_lookup_elem(pids, &pid) != NULL; movabs $0xffff975ac2607800,%rdi 1.26 → callq *ffffffffda6789e9 cmp $0x0,%rax 2.43 → je 0 add $0x38,%rax 0.21 xor %r13d,%r13d if (pid_filter__has(&pids_filtered, getpid())) 0.81 cmp $0x0,%rax → jne 0 mov %rbp,%rdi probe_read(&augmented_args.args, sizeof(augmented_args.args), args); 2.22 add $0xfffffffffffffeb8,%rdi 0.11 mov $0x40,%esi 0.32 mov %rbx,%rdx 2.74 → callq *ffffffffda658409 syscall = bpf_map_lookup_elem(&syscalls, &augmented_args.args.syscall_nr); 0.22 mov %rbp,%rsi 1.69 add $0xfffffffffffffec0,%rsi syscall = bpf_map_lookup_elem(&syscalls, &augmented_args.args.syscall_nr); movabs $0xffff975bfcd36000,%rdi add $0xd0,%rdi 0.21 mov 0x0(%rsi),%eax 0.93 cmp $0x200,%rax → jae 0 0.10 shl $0x3,%rax 0.11 add %rdi,%rax 0.11 → jmp 0 xor %eax,%eax if (syscall == NULL || !syscall->enabled) 1.07 cmp $0x0,%rax → je 0 if (syscall == NULL || !syscall->enabled) 6.57 movzbq 0x0(%rax),%rdi if (syscall == NULL || !syscall->enabled) cmp $0x0,%rdi 0.95 → je 0 mov $0x40,%r8d switch (augmented_args.args.syscall_nr) { mov -0x140(%rbp),%rdi switch (augmented_args.args.syscall_nr) { cmp $0x2,%rdi → je 0 cmp $0x101,%rdi → je 0 cmp $0x15,%rdi → jne 0 case SYS_OPEN: filename_arg = (const void *)args->args[0]; mov 0x10(%rbx),%rdx → jmp 0 case SYS_OPENAT: filename_arg = (const void *)args->args[1]; mov 0x18(%rbx),%rdx if (filename_arg != NULL) { cmp $0x0,%rdx → je 0 xor %edi,%edi augmented_args.filename.reserved = 0; mov %edi,-0x104(%rbp) augmented_args.filename.size = probe_read_str(&augmented_args.filename.value, mov %rbp,%rdi add $0xffffffffffffff00,%rdi augmented_args.filename.size = probe_read_str(&augmented_args.filename.value, mov $0x100,%esi → callq *ffffffffda658499 mov $0x148,%r8d augmented_args.filename.size = probe_read_str(&augmented_args.filename.value, mov %eax,-0x108(%rbp) augmented_args.filename.size = probe_read_str(&augmented_args.filename.value, mov %rax,%rdi shl $0x20,%rdi shr $0x20,%rdi if (augmented_args.filename.size < sizeof(augmented_args.filename.value)) { cmp $0xff,%rdi → ja 0 len -= sizeof(augmented_args.filename.value) - augmented_args.filename.size; add $0x48,%rax len &= sizeof(augmented_args.filename.value) - 1; and $0xff,%rax mov %rax,%r8 mov %rbp,%rcx return perf_event_output(args, &__augmented_syscalls__, BPF_F_CURRENT_CPU, &augmented_args, len); add $0xfffffffffffffeb8,%rcx mov %rbx,%rdi movabs $0xffff975fbd72d800,%rsi mov $0xffffffff,%edx → callq *ffffffffda658ad9 mov %rax,%r13 } mov %r13,%rax 0.72 mov 0x0(%rbp),%rbx mov 0x8(%rbp),%r13 1.16 mov 0x10(%rbp),%r14 0.10 mov 0x18(%rbp),%r15 0.42 add $0x28,%rbp 0.54 leaveq 0.54 ← retq # Please see 'man perf-config' to see how to control what should be seen, via ~/.perfconfig [annotate] section, for instance, one can suppress the source code and see just the disassembly, etc. Alternatively, use the TUI bu just using 'perf annotate', press '/bpf_prog' to see the bpf symbols, press enter and do the interactive annotation, which allows for dumping to a file after selecting the the various output tunables, for instance, the above without source code intermixed, plus showing all the instruction offsets: # perf annotate bpf_prog_819967866022f1e1_sys_enter Then press: 's' to hide the source code + 'O' twice to show all instruction offsets, then 'P' to print to the bpf_prog_819967866022f1e1_sys_enter.annotation file, which will have: # cat bpf_prog_819967866022f1e1_sys_enter.annotation bpf_prog_819967866022f1e1_sys_enter() bpf_prog_819967866022f1e1_sys_enter Event: cycles:ppp 53.41 0: push %rbp 0.63 1: mov %rsp,%rbp 0.31 4: sub $0x170,%rsp 1.93 b: sub $0x28,%rbp 7.02 f: mov %rbx,0x0(%rbp) 3.20 13: mov %r13,0x8(%rbp) 1.07 17: mov %r14,0x10(%rbp) 0.61 1b: mov %r15,0x18(%rbp) 0.11 1f: xor %eax,%eax 1.29 21: mov %rax,0x20(%rbp) 0.11 25: mov %rdi,%rbx 2.02 28: → callq *ffffffffda6776d9 2.76 2d: mov %eax,-0x148(%rbp) 33: mov %rbp,%rsi 36: add $0xfffffffffffffeb8,%rsi 3d: movabs $0xffff975ac2607800,%rdi 1.26 47: → callq *ffffffffda6789e9 4c: cmp $0x0,%rax 2.43 50: → je 0 52: add $0x38,%rax 0.21 56: xor %r13d,%r13d 0.81 59: cmp $0x0,%rax 5d: → jne 0 63: mov %rbp,%rdi 2.22 66: add $0xfffffffffffffeb8,%rdi 0.11 6d: mov $0x40,%esi 0.32 72: mov %rbx,%rdx 2.74 75: → callq *ffffffffda658409 0.22 7a: mov %rbp,%rsi 1.69 7d: add $0xfffffffffffffec0,%rsi 84: movabs $0xffff975bfcd36000,%rdi 8e: add $0xd0,%rdi 0.21 95: mov 0x0(%rsi),%eax 0.93 98: cmp $0x200,%rax 9f: → jae 0 0.10 a1: shl $0x3,%rax 0.11 a5: add %rdi,%rax 0.11 a8: → jmp 0 aa: xor %eax,%eax 1.07 ac: cmp $0x0,%rax b0: → je 0 6.57 b6: movzbq 0x0(%rax),%rdi bb: cmp $0x0,%rdi 0.95 bf: → je 0 c5: mov $0x40,%r8d cb: mov -0x140(%rbp),%rdi d2: cmp $0x2,%rdi d6: → je 0 d8: cmp $0x101,%rdi df: → je 0 e1: cmp $0x15,%rdi e5: → jne 0 e7: mov 0x10(%rbx),%rdx eb: → jmp 0 ed: mov 0x18(%rbx),%rdx f1: cmp $0x0,%rdx f5: → je 0 f7: xor %edi,%edi f9: mov %edi,-0x104(%rbp) ff: mov %rbp,%rdi 102: add $0xffffffffffffff00,%rdi 109: mov $0x100,%esi 10e: → callq *ffffffffda658499 113: mov $0x148,%r8d 119: mov %eax,-0x108(%rbp) 11f: mov %rax,%rdi 122: shl $0x20,%rdi 126: shr $0x20,%rdi 12a: cmp $0xff,%rdi 131: → ja 0 133: add $0x48,%rax 137: and $0xff,%rax 13d: mov %rax,%r8 140: mov %rbp,%rcx 143: add $0xfffffffffffffeb8,%rcx 14a: mov %rbx,%rdi 14d: movabs $0xffff975fbd72d800,%rsi 157: mov $0xffffffff,%edx 15c: → callq *ffffffffda658ad9 161: mov %rax,%r13 164: mov %r13,%rax 0.72 167: mov 0x0(%rbp),%rbx 16b: mov 0x8(%rbp),%r13 1.16 16f: mov 0x10(%rbp),%r14 0.10 173: mov 0x18(%rbp),%r15 0.42 177: add $0x28,%rbp 0.54 17b: leaveq 0.54 17c: ← retq Another cool way to test all this is to symple use 'perf top' look for those symbols, go there and press enter, annotate it live :-) Signed-off-by: Song Liu <songliubraving@fb.com> Reviewed-by: Jiri Olsa <jolsa@kernel.org> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Alexei Starovoitov <ast@kernel.org> Cc: Daniel Borkmann <daniel@iogearbox.net> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Stanislav Fomichev <sdf@google.com> Link: http://lkml.kernel.org/r/20190312053051.2690567-13-songliubraving@fb.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2019-03-12 08:30:48 +03:00
init_disassemble_info(&info, s,
(fprintf_ftype) fprintf);
info.arch = bfd_get_arch(bfdf);
info.mach = bfd_get_mach(bfdf);
info_node = perf_env__find_bpf_prog_info(dso->bpf_prog.env,
dso->bpf_prog.id);
if (!info_node) {
ret = SYMBOL_ANNOTATE_ERRNO__BPF_MISSING_BTF;
perf annotate: Enable annotation of BPF programs In symbol__disassemble(), DSO_BINARY_TYPE__BPF_PROG_INFO dso calls into a new function symbol__disassemble_bpf(), where annotation line information is filled based on the bpf_prog_info and btf data saved in given perf_env. symbol__disassemble_bpf() uses binutils's libopcodes to disassemble bpf programs. Committer testing: After fixing this: - u64 *addrs = (u64 *)(info_linear->info.jited_ksyms); + u64 *addrs = (u64 *)(uintptr_t)(info_linear->info.jited_ksyms); Detected when crossbuilding to a 32-bit arch. And making all this dependent on HAVE_LIBBFD_SUPPORT and HAVE_LIBBPF_SUPPORT: 1) Have a BPF program running, one that has BTF info, etc, I used the tools/perf/examples/bpf/augmented_raw_syscalls.c put in place by 'perf trace'. # grep -B1 augmented_raw ~/.perfconfig [trace] add_events = /home/acme/git/perf/tools/perf/examples/bpf/augmented_raw_syscalls.c # # perf trace -e *mmsg dnf/6245 sendmmsg(20, 0x7f5485a88030, 2, MSG_NOSIGNAL) = 2 NetworkManager/10055 sendmmsg(22<socket:[1056822]>, 0x7f8126ad1bb0, 2, MSG_NOSIGNAL) = 2 2) Then do a 'perf record' system wide for a while: # perf record -a ^C[ perf record: Woken up 68 times to write data ] [ perf record: Captured and wrote 19.427 MB perf.data (366891 samples) ] # 3) Check that we captured BPF and BTF info in the perf.data file: # perf report --header-only | grep 'b[pt]f' # event : name = cycles:ppp, , id = { 294789, 294790, 294791, 294792, 294793, 294794, 294795, 294796 }, size = 112, { sample_period, sample_freq } = 4000, sample_type = IP|TID|TIME|CPU|PERIOD, read_format = ID, disabled = 1, inherit = 1, mmap = 1, comm = 1, freq = 1, task = 1, precise_ip = 3, sample_id_all = 1, exclude_guest = 1, mmap2 = 1, comm_exec = 1, ksymbol = 1, bpf_event = 1 # bpf_prog_info of id 13 # bpf_prog_info of id 14 # bpf_prog_info of id 15 # bpf_prog_info of id 16 # bpf_prog_info of id 17 # bpf_prog_info of id 18 # bpf_prog_info of id 21 # bpf_prog_info of id 22 # bpf_prog_info of id 41 # bpf_prog_info of id 42 # btf info of id 2 # 4) Check which programs got recorded: # perf report | grep bpf_prog | head 0.16% exe bpf_prog_819967866022f1e1_sys_enter [k] bpf_prog_819967866022f1e1_sys_enter 0.14% exe bpf_prog_c1bd85c092d6e4aa_sys_exit [k] bpf_prog_c1bd85c092d6e4aa_sys_exit 0.08% fuse-overlayfs bpf_prog_819967866022f1e1_sys_enter [k] bpf_prog_819967866022f1e1_sys_enter 0.07% fuse-overlayfs bpf_prog_c1bd85c092d6e4aa_sys_exit [k] bpf_prog_c1bd85c092d6e4aa_sys_exit 0.01% clang-4.0 bpf_prog_c1bd85c092d6e4aa_sys_exit [k] bpf_prog_c1bd85c092d6e4aa_sys_exit 0.01% clang-4.0 bpf_prog_819967866022f1e1_sys_enter [k] bpf_prog_819967866022f1e1_sys_enter 0.00% clang bpf_prog_c1bd85c092d6e4aa_sys_exit [k] bpf_prog_c1bd85c092d6e4aa_sys_exit 0.00% runc bpf_prog_819967866022f1e1_sys_enter [k] bpf_prog_819967866022f1e1_sys_enter 0.00% clang bpf_prog_819967866022f1e1_sys_enter [k] bpf_prog_819967866022f1e1_sys_enter 0.00% sh bpf_prog_c1bd85c092d6e4aa_sys_exit [k] bpf_prog_c1bd85c092d6e4aa_sys_exit # This was with the default --sort order for 'perf report', which is: --sort comm,dso,symbol If we just look for the symbol, for instance: # perf report --sort symbol | grep bpf_prog | head 0.26% [k] bpf_prog_819967866022f1e1_sys_enter - - 0.24% [k] bpf_prog_c1bd85c092d6e4aa_sys_exit - - # or the DSO: # perf report --sort dso | grep bpf_prog | head 0.26% bpf_prog_819967866022f1e1_sys_enter 0.24% bpf_prog_c1bd85c092d6e4aa_sys_exit # We'll see the two BPF programs that augmented_raw_syscalls.o puts in place, one attached to the raw_syscalls:sys_enter and another to the raw_syscalls:sys_exit tracepoints, as expected. Now we can finally do, from the command line, annotation for one of those two symbols, with the original BPF program source coude intermixed with the disassembled JITed code: # perf annotate --stdio2 bpf_prog_819967866022f1e1_sys_enter Samples: 950 of event 'cycles:ppp', 4000 Hz, Event count (approx.): 553756947, [percent: local period] bpf_prog_819967866022f1e1_sys_enter() bpf_prog_819967866022f1e1_sys_enter Percent int sys_enter(struct syscall_enter_args *args) 53.41 push %rbp 0.63 mov %rsp,%rbp 0.31 sub $0x170,%rsp 1.93 sub $0x28,%rbp 7.02 mov %rbx,0x0(%rbp) 3.20 mov %r13,0x8(%rbp) 1.07 mov %r14,0x10(%rbp) 0.61 mov %r15,0x18(%rbp) 0.11 xor %eax,%eax 1.29 mov %rax,0x20(%rbp) 0.11 mov %rdi,%rbx return bpf_get_current_pid_tgid(); 2.02 → callq *ffffffffda6776d9 2.76 mov %eax,-0x148(%rbp) mov %rbp,%rsi int sys_enter(struct syscall_enter_args *args) add $0xfffffffffffffeb8,%rsi return bpf_map_lookup_elem(pids, &pid) != NULL; movabs $0xffff975ac2607800,%rdi 1.26 → callq *ffffffffda6789e9 cmp $0x0,%rax 2.43 → je 0 add $0x38,%rax 0.21 xor %r13d,%r13d if (pid_filter__has(&pids_filtered, getpid())) 0.81 cmp $0x0,%rax → jne 0 mov %rbp,%rdi probe_read(&augmented_args.args, sizeof(augmented_args.args), args); 2.22 add $0xfffffffffffffeb8,%rdi 0.11 mov $0x40,%esi 0.32 mov %rbx,%rdx 2.74 → callq *ffffffffda658409 syscall = bpf_map_lookup_elem(&syscalls, &augmented_args.args.syscall_nr); 0.22 mov %rbp,%rsi 1.69 add $0xfffffffffffffec0,%rsi syscall = bpf_map_lookup_elem(&syscalls, &augmented_args.args.syscall_nr); movabs $0xffff975bfcd36000,%rdi add $0xd0,%rdi 0.21 mov 0x0(%rsi),%eax 0.93 cmp $0x200,%rax → jae 0 0.10 shl $0x3,%rax 0.11 add %rdi,%rax 0.11 → jmp 0 xor %eax,%eax if (syscall == NULL || !syscall->enabled) 1.07 cmp $0x0,%rax → je 0 if (syscall == NULL || !syscall->enabled) 6.57 movzbq 0x0(%rax),%rdi if (syscall == NULL || !syscall->enabled) cmp $0x0,%rdi 0.95 → je 0 mov $0x40,%r8d switch (augmented_args.args.syscall_nr) { mov -0x140(%rbp),%rdi switch (augmented_args.args.syscall_nr) { cmp $0x2,%rdi → je 0 cmp $0x101,%rdi → je 0 cmp $0x15,%rdi → jne 0 case SYS_OPEN: filename_arg = (const void *)args->args[0]; mov 0x10(%rbx),%rdx → jmp 0 case SYS_OPENAT: filename_arg = (const void *)args->args[1]; mov 0x18(%rbx),%rdx if (filename_arg != NULL) { cmp $0x0,%rdx → je 0 xor %edi,%edi augmented_args.filename.reserved = 0; mov %edi,-0x104(%rbp) augmented_args.filename.size = probe_read_str(&augmented_args.filename.value, mov %rbp,%rdi add $0xffffffffffffff00,%rdi augmented_args.filename.size = probe_read_str(&augmented_args.filename.value, mov $0x100,%esi → callq *ffffffffda658499 mov $0x148,%r8d augmented_args.filename.size = probe_read_str(&augmented_args.filename.value, mov %eax,-0x108(%rbp) augmented_args.filename.size = probe_read_str(&augmented_args.filename.value, mov %rax,%rdi shl $0x20,%rdi shr $0x20,%rdi if (augmented_args.filename.size < sizeof(augmented_args.filename.value)) { cmp $0xff,%rdi → ja 0 len -= sizeof(augmented_args.filename.value) - augmented_args.filename.size; add $0x48,%rax len &= sizeof(augmented_args.filename.value) - 1; and $0xff,%rax mov %rax,%r8 mov %rbp,%rcx return perf_event_output(args, &__augmented_syscalls__, BPF_F_CURRENT_CPU, &augmented_args, len); add $0xfffffffffffffeb8,%rcx mov %rbx,%rdi movabs $0xffff975fbd72d800,%rsi mov $0xffffffff,%edx → callq *ffffffffda658ad9 mov %rax,%r13 } mov %r13,%rax 0.72 mov 0x0(%rbp),%rbx mov 0x8(%rbp),%r13 1.16 mov 0x10(%rbp),%r14 0.10 mov 0x18(%rbp),%r15 0.42 add $0x28,%rbp 0.54 leaveq 0.54 ← retq # Please see 'man perf-config' to see how to control what should be seen, via ~/.perfconfig [annotate] section, for instance, one can suppress the source code and see just the disassembly, etc. Alternatively, use the TUI bu just using 'perf annotate', press '/bpf_prog' to see the bpf symbols, press enter and do the interactive annotation, which allows for dumping to a file after selecting the the various output tunables, for instance, the above without source code intermixed, plus showing all the instruction offsets: # perf annotate bpf_prog_819967866022f1e1_sys_enter Then press: 's' to hide the source code + 'O' twice to show all instruction offsets, then 'P' to print to the bpf_prog_819967866022f1e1_sys_enter.annotation file, which will have: # cat bpf_prog_819967866022f1e1_sys_enter.annotation bpf_prog_819967866022f1e1_sys_enter() bpf_prog_819967866022f1e1_sys_enter Event: cycles:ppp 53.41 0: push %rbp 0.63 1: mov %rsp,%rbp 0.31 4: sub $0x170,%rsp 1.93 b: sub $0x28,%rbp 7.02 f: mov %rbx,0x0(%rbp) 3.20 13: mov %r13,0x8(%rbp) 1.07 17: mov %r14,0x10(%rbp) 0.61 1b: mov %r15,0x18(%rbp) 0.11 1f: xor %eax,%eax 1.29 21: mov %rax,0x20(%rbp) 0.11 25: mov %rdi,%rbx 2.02 28: → callq *ffffffffda6776d9 2.76 2d: mov %eax,-0x148(%rbp) 33: mov %rbp,%rsi 36: add $0xfffffffffffffeb8,%rsi 3d: movabs $0xffff975ac2607800,%rdi 1.26 47: → callq *ffffffffda6789e9 4c: cmp $0x0,%rax 2.43 50: → je 0 52: add $0x38,%rax 0.21 56: xor %r13d,%r13d 0.81 59: cmp $0x0,%rax 5d: → jne 0 63: mov %rbp,%rdi 2.22 66: add $0xfffffffffffffeb8,%rdi 0.11 6d: mov $0x40,%esi 0.32 72: mov %rbx,%rdx 2.74 75: → callq *ffffffffda658409 0.22 7a: mov %rbp,%rsi 1.69 7d: add $0xfffffffffffffec0,%rsi 84: movabs $0xffff975bfcd36000,%rdi 8e: add $0xd0,%rdi 0.21 95: mov 0x0(%rsi),%eax 0.93 98: cmp $0x200,%rax 9f: → jae 0 0.10 a1: shl $0x3,%rax 0.11 a5: add %rdi,%rax 0.11 a8: → jmp 0 aa: xor %eax,%eax 1.07 ac: cmp $0x0,%rax b0: → je 0 6.57 b6: movzbq 0x0(%rax),%rdi bb: cmp $0x0,%rdi 0.95 bf: → je 0 c5: mov $0x40,%r8d cb: mov -0x140(%rbp),%rdi d2: cmp $0x2,%rdi d6: → je 0 d8: cmp $0x101,%rdi df: → je 0 e1: cmp $0x15,%rdi e5: → jne 0 e7: mov 0x10(%rbx),%rdx eb: → jmp 0 ed: mov 0x18(%rbx),%rdx f1: cmp $0x0,%rdx f5: → je 0 f7: xor %edi,%edi f9: mov %edi,-0x104(%rbp) ff: mov %rbp,%rdi 102: add $0xffffffffffffff00,%rdi 109: mov $0x100,%esi 10e: → callq *ffffffffda658499 113: mov $0x148,%r8d 119: mov %eax,-0x108(%rbp) 11f: mov %rax,%rdi 122: shl $0x20,%rdi 126: shr $0x20,%rdi 12a: cmp $0xff,%rdi 131: → ja 0 133: add $0x48,%rax 137: and $0xff,%rax 13d: mov %rax,%r8 140: mov %rbp,%rcx 143: add $0xfffffffffffffeb8,%rcx 14a: mov %rbx,%rdi 14d: movabs $0xffff975fbd72d800,%rsi 157: mov $0xffffffff,%edx 15c: → callq *ffffffffda658ad9 161: mov %rax,%r13 164: mov %r13,%rax 0.72 167: mov 0x0(%rbp),%rbx 16b: mov 0x8(%rbp),%r13 1.16 16f: mov 0x10(%rbp),%r14 0.10 173: mov 0x18(%rbp),%r15 0.42 177: add $0x28,%rbp 0.54 17b: leaveq 0.54 17c: ← retq Another cool way to test all this is to symple use 'perf top' look for those symbols, go there and press enter, annotate it live :-) Signed-off-by: Song Liu <songliubraving@fb.com> Reviewed-by: Jiri Olsa <jolsa@kernel.org> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Alexei Starovoitov <ast@kernel.org> Cc: Daniel Borkmann <daniel@iogearbox.net> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Stanislav Fomichev <sdf@google.com> Link: http://lkml.kernel.org/r/20190312053051.2690567-13-songliubraving@fb.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2019-03-12 08:30:48 +03:00
goto out;
}
perf annotate: Enable annotation of BPF programs In symbol__disassemble(), DSO_BINARY_TYPE__BPF_PROG_INFO dso calls into a new function symbol__disassemble_bpf(), where annotation line information is filled based on the bpf_prog_info and btf data saved in given perf_env. symbol__disassemble_bpf() uses binutils's libopcodes to disassemble bpf programs. Committer testing: After fixing this: - u64 *addrs = (u64 *)(info_linear->info.jited_ksyms); + u64 *addrs = (u64 *)(uintptr_t)(info_linear->info.jited_ksyms); Detected when crossbuilding to a 32-bit arch. And making all this dependent on HAVE_LIBBFD_SUPPORT and HAVE_LIBBPF_SUPPORT: 1) Have a BPF program running, one that has BTF info, etc, I used the tools/perf/examples/bpf/augmented_raw_syscalls.c put in place by 'perf trace'. # grep -B1 augmented_raw ~/.perfconfig [trace] add_events = /home/acme/git/perf/tools/perf/examples/bpf/augmented_raw_syscalls.c # # perf trace -e *mmsg dnf/6245 sendmmsg(20, 0x7f5485a88030, 2, MSG_NOSIGNAL) = 2 NetworkManager/10055 sendmmsg(22<socket:[1056822]>, 0x7f8126ad1bb0, 2, MSG_NOSIGNAL) = 2 2) Then do a 'perf record' system wide for a while: # perf record -a ^C[ perf record: Woken up 68 times to write data ] [ perf record: Captured and wrote 19.427 MB perf.data (366891 samples) ] # 3) Check that we captured BPF and BTF info in the perf.data file: # perf report --header-only | grep 'b[pt]f' # event : name = cycles:ppp, , id = { 294789, 294790, 294791, 294792, 294793, 294794, 294795, 294796 }, size = 112, { sample_period, sample_freq } = 4000, sample_type = IP|TID|TIME|CPU|PERIOD, read_format = ID, disabled = 1, inherit = 1, mmap = 1, comm = 1, freq = 1, task = 1, precise_ip = 3, sample_id_all = 1, exclude_guest = 1, mmap2 = 1, comm_exec = 1, ksymbol = 1, bpf_event = 1 # bpf_prog_info of id 13 # bpf_prog_info of id 14 # bpf_prog_info of id 15 # bpf_prog_info of id 16 # bpf_prog_info of id 17 # bpf_prog_info of id 18 # bpf_prog_info of id 21 # bpf_prog_info of id 22 # bpf_prog_info of id 41 # bpf_prog_info of id 42 # btf info of id 2 # 4) Check which programs got recorded: # perf report | grep bpf_prog | head 0.16% exe bpf_prog_819967866022f1e1_sys_enter [k] bpf_prog_819967866022f1e1_sys_enter 0.14% exe bpf_prog_c1bd85c092d6e4aa_sys_exit [k] bpf_prog_c1bd85c092d6e4aa_sys_exit 0.08% fuse-overlayfs bpf_prog_819967866022f1e1_sys_enter [k] bpf_prog_819967866022f1e1_sys_enter 0.07% fuse-overlayfs bpf_prog_c1bd85c092d6e4aa_sys_exit [k] bpf_prog_c1bd85c092d6e4aa_sys_exit 0.01% clang-4.0 bpf_prog_c1bd85c092d6e4aa_sys_exit [k] bpf_prog_c1bd85c092d6e4aa_sys_exit 0.01% clang-4.0 bpf_prog_819967866022f1e1_sys_enter [k] bpf_prog_819967866022f1e1_sys_enter 0.00% clang bpf_prog_c1bd85c092d6e4aa_sys_exit [k] bpf_prog_c1bd85c092d6e4aa_sys_exit 0.00% runc bpf_prog_819967866022f1e1_sys_enter [k] bpf_prog_819967866022f1e1_sys_enter 0.00% clang bpf_prog_819967866022f1e1_sys_enter [k] bpf_prog_819967866022f1e1_sys_enter 0.00% sh bpf_prog_c1bd85c092d6e4aa_sys_exit [k] bpf_prog_c1bd85c092d6e4aa_sys_exit # This was with the default --sort order for 'perf report', which is: --sort comm,dso,symbol If we just look for the symbol, for instance: # perf report --sort symbol | grep bpf_prog | head 0.26% [k] bpf_prog_819967866022f1e1_sys_enter - - 0.24% [k] bpf_prog_c1bd85c092d6e4aa_sys_exit - - # or the DSO: # perf report --sort dso | grep bpf_prog | head 0.26% bpf_prog_819967866022f1e1_sys_enter 0.24% bpf_prog_c1bd85c092d6e4aa_sys_exit # We'll see the two BPF programs that augmented_raw_syscalls.o puts in place, one attached to the raw_syscalls:sys_enter and another to the raw_syscalls:sys_exit tracepoints, as expected. Now we can finally do, from the command line, annotation for one of those two symbols, with the original BPF program source coude intermixed with the disassembled JITed code: # perf annotate --stdio2 bpf_prog_819967866022f1e1_sys_enter Samples: 950 of event 'cycles:ppp', 4000 Hz, Event count (approx.): 553756947, [percent: local period] bpf_prog_819967866022f1e1_sys_enter() bpf_prog_819967866022f1e1_sys_enter Percent int sys_enter(struct syscall_enter_args *args) 53.41 push %rbp 0.63 mov %rsp,%rbp 0.31 sub $0x170,%rsp 1.93 sub $0x28,%rbp 7.02 mov %rbx,0x0(%rbp) 3.20 mov %r13,0x8(%rbp) 1.07 mov %r14,0x10(%rbp) 0.61 mov %r15,0x18(%rbp) 0.11 xor %eax,%eax 1.29 mov %rax,0x20(%rbp) 0.11 mov %rdi,%rbx return bpf_get_current_pid_tgid(); 2.02 → callq *ffffffffda6776d9 2.76 mov %eax,-0x148(%rbp) mov %rbp,%rsi int sys_enter(struct syscall_enter_args *args) add $0xfffffffffffffeb8,%rsi return bpf_map_lookup_elem(pids, &pid) != NULL; movabs $0xffff975ac2607800,%rdi 1.26 → callq *ffffffffda6789e9 cmp $0x0,%rax 2.43 → je 0 add $0x38,%rax 0.21 xor %r13d,%r13d if (pid_filter__has(&pids_filtered, getpid())) 0.81 cmp $0x0,%rax → jne 0 mov %rbp,%rdi probe_read(&augmented_args.args, sizeof(augmented_args.args), args); 2.22 add $0xfffffffffffffeb8,%rdi 0.11 mov $0x40,%esi 0.32 mov %rbx,%rdx 2.74 → callq *ffffffffda658409 syscall = bpf_map_lookup_elem(&syscalls, &augmented_args.args.syscall_nr); 0.22 mov %rbp,%rsi 1.69 add $0xfffffffffffffec0,%rsi syscall = bpf_map_lookup_elem(&syscalls, &augmented_args.args.syscall_nr); movabs $0xffff975bfcd36000,%rdi add $0xd0,%rdi 0.21 mov 0x0(%rsi),%eax 0.93 cmp $0x200,%rax → jae 0 0.10 shl $0x3,%rax 0.11 add %rdi,%rax 0.11 → jmp 0 xor %eax,%eax if (syscall == NULL || !syscall->enabled) 1.07 cmp $0x0,%rax → je 0 if (syscall == NULL || !syscall->enabled) 6.57 movzbq 0x0(%rax),%rdi if (syscall == NULL || !syscall->enabled) cmp $0x0,%rdi 0.95 → je 0 mov $0x40,%r8d switch (augmented_args.args.syscall_nr) { mov -0x140(%rbp),%rdi switch (augmented_args.args.syscall_nr) { cmp $0x2,%rdi → je 0 cmp $0x101,%rdi → je 0 cmp $0x15,%rdi → jne 0 case SYS_OPEN: filename_arg = (const void *)args->args[0]; mov 0x10(%rbx),%rdx → jmp 0 case SYS_OPENAT: filename_arg = (const void *)args->args[1]; mov 0x18(%rbx),%rdx if (filename_arg != NULL) { cmp $0x0,%rdx → je 0 xor %edi,%edi augmented_args.filename.reserved = 0; mov %edi,-0x104(%rbp) augmented_args.filename.size = probe_read_str(&augmented_args.filename.value, mov %rbp,%rdi add $0xffffffffffffff00,%rdi augmented_args.filename.size = probe_read_str(&augmented_args.filename.value, mov $0x100,%esi → callq *ffffffffda658499 mov $0x148,%r8d augmented_args.filename.size = probe_read_str(&augmented_args.filename.value, mov %eax,-0x108(%rbp) augmented_args.filename.size = probe_read_str(&augmented_args.filename.value, mov %rax,%rdi shl $0x20,%rdi shr $0x20,%rdi if (augmented_args.filename.size < sizeof(augmented_args.filename.value)) { cmp $0xff,%rdi → ja 0 len -= sizeof(augmented_args.filename.value) - augmented_args.filename.size; add $0x48,%rax len &= sizeof(augmented_args.filename.value) - 1; and $0xff,%rax mov %rax,%r8 mov %rbp,%rcx return perf_event_output(args, &__augmented_syscalls__, BPF_F_CURRENT_CPU, &augmented_args, len); add $0xfffffffffffffeb8,%rcx mov %rbx,%rdi movabs $0xffff975fbd72d800,%rsi mov $0xffffffff,%edx → callq *ffffffffda658ad9 mov %rax,%r13 } mov %r13,%rax 0.72 mov 0x0(%rbp),%rbx mov 0x8(%rbp),%r13 1.16 mov 0x10(%rbp),%r14 0.10 mov 0x18(%rbp),%r15 0.42 add $0x28,%rbp 0.54 leaveq 0.54 ← retq # Please see 'man perf-config' to see how to control what should be seen, via ~/.perfconfig [annotate] section, for instance, one can suppress the source code and see just the disassembly, etc. Alternatively, use the TUI bu just using 'perf annotate', press '/bpf_prog' to see the bpf symbols, press enter and do the interactive annotation, which allows for dumping to a file after selecting the the various output tunables, for instance, the above without source code intermixed, plus showing all the instruction offsets: # perf annotate bpf_prog_819967866022f1e1_sys_enter Then press: 's' to hide the source code + 'O' twice to show all instruction offsets, then 'P' to print to the bpf_prog_819967866022f1e1_sys_enter.annotation file, which will have: # cat bpf_prog_819967866022f1e1_sys_enter.annotation bpf_prog_819967866022f1e1_sys_enter() bpf_prog_819967866022f1e1_sys_enter Event: cycles:ppp 53.41 0: push %rbp 0.63 1: mov %rsp,%rbp 0.31 4: sub $0x170,%rsp 1.93 b: sub $0x28,%rbp 7.02 f: mov %rbx,0x0(%rbp) 3.20 13: mov %r13,0x8(%rbp) 1.07 17: mov %r14,0x10(%rbp) 0.61 1b: mov %r15,0x18(%rbp) 0.11 1f: xor %eax,%eax 1.29 21: mov %rax,0x20(%rbp) 0.11 25: mov %rdi,%rbx 2.02 28: → callq *ffffffffda6776d9 2.76 2d: mov %eax,-0x148(%rbp) 33: mov %rbp,%rsi 36: add $0xfffffffffffffeb8,%rsi 3d: movabs $0xffff975ac2607800,%rdi 1.26 47: → callq *ffffffffda6789e9 4c: cmp $0x0,%rax 2.43 50: → je 0 52: add $0x38,%rax 0.21 56: xor %r13d,%r13d 0.81 59: cmp $0x0,%rax 5d: → jne 0 63: mov %rbp,%rdi 2.22 66: add $0xfffffffffffffeb8,%rdi 0.11 6d: mov $0x40,%esi 0.32 72: mov %rbx,%rdx 2.74 75: → callq *ffffffffda658409 0.22 7a: mov %rbp,%rsi 1.69 7d: add $0xfffffffffffffec0,%rsi 84: movabs $0xffff975bfcd36000,%rdi 8e: add $0xd0,%rdi 0.21 95: mov 0x0(%rsi),%eax 0.93 98: cmp $0x200,%rax 9f: → jae 0 0.10 a1: shl $0x3,%rax 0.11 a5: add %rdi,%rax 0.11 a8: → jmp 0 aa: xor %eax,%eax 1.07 ac: cmp $0x0,%rax b0: → je 0 6.57 b6: movzbq 0x0(%rax),%rdi bb: cmp $0x0,%rdi 0.95 bf: → je 0 c5: mov $0x40,%r8d cb: mov -0x140(%rbp),%rdi d2: cmp $0x2,%rdi d6: → je 0 d8: cmp $0x101,%rdi df: → je 0 e1: cmp $0x15,%rdi e5: → jne 0 e7: mov 0x10(%rbx),%rdx eb: → jmp 0 ed: mov 0x18(%rbx),%rdx f1: cmp $0x0,%rdx f5: → je 0 f7: xor %edi,%edi f9: mov %edi,-0x104(%rbp) ff: mov %rbp,%rdi 102: add $0xffffffffffffff00,%rdi 109: mov $0x100,%esi 10e: → callq *ffffffffda658499 113: mov $0x148,%r8d 119: mov %eax,-0x108(%rbp) 11f: mov %rax,%rdi 122: shl $0x20,%rdi 126: shr $0x20,%rdi 12a: cmp $0xff,%rdi 131: → ja 0 133: add $0x48,%rax 137: and $0xff,%rax 13d: mov %rax,%r8 140: mov %rbp,%rcx 143: add $0xfffffffffffffeb8,%rcx 14a: mov %rbx,%rdi 14d: movabs $0xffff975fbd72d800,%rsi 157: mov $0xffffffff,%edx 15c: → callq *ffffffffda658ad9 161: mov %rax,%r13 164: mov %r13,%rax 0.72 167: mov 0x0(%rbp),%rbx 16b: mov 0x8(%rbp),%r13 1.16 16f: mov 0x10(%rbp),%r14 0.10 173: mov 0x18(%rbp),%r15 0.42 177: add $0x28,%rbp 0.54 17b: leaveq 0.54 17c: ← retq Another cool way to test all this is to symple use 'perf top' look for those symbols, go there and press enter, annotate it live :-) Signed-off-by: Song Liu <songliubraving@fb.com> Reviewed-by: Jiri Olsa <jolsa@kernel.org> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Alexei Starovoitov <ast@kernel.org> Cc: Daniel Borkmann <daniel@iogearbox.net> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Stanislav Fomichev <sdf@google.com> Link: http://lkml.kernel.org/r/20190312053051.2690567-13-songliubraving@fb.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2019-03-12 08:30:48 +03:00
info_linear = info_node->info_linear;
sub_id = dso->bpf_prog.sub_id;
info.buffer = (void *)(uintptr_t)(info_linear->info.jited_prog_insns);
perf annotate: Enable annotation of BPF programs In symbol__disassemble(), DSO_BINARY_TYPE__BPF_PROG_INFO dso calls into a new function symbol__disassemble_bpf(), where annotation line information is filled based on the bpf_prog_info and btf data saved in given perf_env. symbol__disassemble_bpf() uses binutils's libopcodes to disassemble bpf programs. Committer testing: After fixing this: - u64 *addrs = (u64 *)(info_linear->info.jited_ksyms); + u64 *addrs = (u64 *)(uintptr_t)(info_linear->info.jited_ksyms); Detected when crossbuilding to a 32-bit arch. And making all this dependent on HAVE_LIBBFD_SUPPORT and HAVE_LIBBPF_SUPPORT: 1) Have a BPF program running, one that has BTF info, etc, I used the tools/perf/examples/bpf/augmented_raw_syscalls.c put in place by 'perf trace'. # grep -B1 augmented_raw ~/.perfconfig [trace] add_events = /home/acme/git/perf/tools/perf/examples/bpf/augmented_raw_syscalls.c # # perf trace -e *mmsg dnf/6245 sendmmsg(20, 0x7f5485a88030, 2, MSG_NOSIGNAL) = 2 NetworkManager/10055 sendmmsg(22<socket:[1056822]>, 0x7f8126ad1bb0, 2, MSG_NOSIGNAL) = 2 2) Then do a 'perf record' system wide for a while: # perf record -a ^C[ perf record: Woken up 68 times to write data ] [ perf record: Captured and wrote 19.427 MB perf.data (366891 samples) ] # 3) Check that we captured BPF and BTF info in the perf.data file: # perf report --header-only | grep 'b[pt]f' # event : name = cycles:ppp, , id = { 294789, 294790, 294791, 294792, 294793, 294794, 294795, 294796 }, size = 112, { sample_period, sample_freq } = 4000, sample_type = IP|TID|TIME|CPU|PERIOD, read_format = ID, disabled = 1, inherit = 1, mmap = 1, comm = 1, freq = 1, task = 1, precise_ip = 3, sample_id_all = 1, exclude_guest = 1, mmap2 = 1, comm_exec = 1, ksymbol = 1, bpf_event = 1 # bpf_prog_info of id 13 # bpf_prog_info of id 14 # bpf_prog_info of id 15 # bpf_prog_info of id 16 # bpf_prog_info of id 17 # bpf_prog_info of id 18 # bpf_prog_info of id 21 # bpf_prog_info of id 22 # bpf_prog_info of id 41 # bpf_prog_info of id 42 # btf info of id 2 # 4) Check which programs got recorded: # perf report | grep bpf_prog | head 0.16% exe bpf_prog_819967866022f1e1_sys_enter [k] bpf_prog_819967866022f1e1_sys_enter 0.14% exe bpf_prog_c1bd85c092d6e4aa_sys_exit [k] bpf_prog_c1bd85c092d6e4aa_sys_exit 0.08% fuse-overlayfs bpf_prog_819967866022f1e1_sys_enter [k] bpf_prog_819967866022f1e1_sys_enter 0.07% fuse-overlayfs bpf_prog_c1bd85c092d6e4aa_sys_exit [k] bpf_prog_c1bd85c092d6e4aa_sys_exit 0.01% clang-4.0 bpf_prog_c1bd85c092d6e4aa_sys_exit [k] bpf_prog_c1bd85c092d6e4aa_sys_exit 0.01% clang-4.0 bpf_prog_819967866022f1e1_sys_enter [k] bpf_prog_819967866022f1e1_sys_enter 0.00% clang bpf_prog_c1bd85c092d6e4aa_sys_exit [k] bpf_prog_c1bd85c092d6e4aa_sys_exit 0.00% runc bpf_prog_819967866022f1e1_sys_enter [k] bpf_prog_819967866022f1e1_sys_enter 0.00% clang bpf_prog_819967866022f1e1_sys_enter [k] bpf_prog_819967866022f1e1_sys_enter 0.00% sh bpf_prog_c1bd85c092d6e4aa_sys_exit [k] bpf_prog_c1bd85c092d6e4aa_sys_exit # This was with the default --sort order for 'perf report', which is: --sort comm,dso,symbol If we just look for the symbol, for instance: # perf report --sort symbol | grep bpf_prog | head 0.26% [k] bpf_prog_819967866022f1e1_sys_enter - - 0.24% [k] bpf_prog_c1bd85c092d6e4aa_sys_exit - - # or the DSO: # perf report --sort dso | grep bpf_prog | head 0.26% bpf_prog_819967866022f1e1_sys_enter 0.24% bpf_prog_c1bd85c092d6e4aa_sys_exit # We'll see the two BPF programs that augmented_raw_syscalls.o puts in place, one attached to the raw_syscalls:sys_enter and another to the raw_syscalls:sys_exit tracepoints, as expected. Now we can finally do, from the command line, annotation for one of those two symbols, with the original BPF program source coude intermixed with the disassembled JITed code: # perf annotate --stdio2 bpf_prog_819967866022f1e1_sys_enter Samples: 950 of event 'cycles:ppp', 4000 Hz, Event count (approx.): 553756947, [percent: local period] bpf_prog_819967866022f1e1_sys_enter() bpf_prog_819967866022f1e1_sys_enter Percent int sys_enter(struct syscall_enter_args *args) 53.41 push %rbp 0.63 mov %rsp,%rbp 0.31 sub $0x170,%rsp 1.93 sub $0x28,%rbp 7.02 mov %rbx,0x0(%rbp) 3.20 mov %r13,0x8(%rbp) 1.07 mov %r14,0x10(%rbp) 0.61 mov %r15,0x18(%rbp) 0.11 xor %eax,%eax 1.29 mov %rax,0x20(%rbp) 0.11 mov %rdi,%rbx return bpf_get_current_pid_tgid(); 2.02 → callq *ffffffffda6776d9 2.76 mov %eax,-0x148(%rbp) mov %rbp,%rsi int sys_enter(struct syscall_enter_args *args) add $0xfffffffffffffeb8,%rsi return bpf_map_lookup_elem(pids, &pid) != NULL; movabs $0xffff975ac2607800,%rdi 1.26 → callq *ffffffffda6789e9 cmp $0x0,%rax 2.43 → je 0 add $0x38,%rax 0.21 xor %r13d,%r13d if (pid_filter__has(&pids_filtered, getpid())) 0.81 cmp $0x0,%rax → jne 0 mov %rbp,%rdi probe_read(&augmented_args.args, sizeof(augmented_args.args), args); 2.22 add $0xfffffffffffffeb8,%rdi 0.11 mov $0x40,%esi 0.32 mov %rbx,%rdx 2.74 → callq *ffffffffda658409 syscall = bpf_map_lookup_elem(&syscalls, &augmented_args.args.syscall_nr); 0.22 mov %rbp,%rsi 1.69 add $0xfffffffffffffec0,%rsi syscall = bpf_map_lookup_elem(&syscalls, &augmented_args.args.syscall_nr); movabs $0xffff975bfcd36000,%rdi add $0xd0,%rdi 0.21 mov 0x0(%rsi),%eax 0.93 cmp $0x200,%rax → jae 0 0.10 shl $0x3,%rax 0.11 add %rdi,%rax 0.11 → jmp 0 xor %eax,%eax if (syscall == NULL || !syscall->enabled) 1.07 cmp $0x0,%rax → je 0 if (syscall == NULL || !syscall->enabled) 6.57 movzbq 0x0(%rax),%rdi if (syscall == NULL || !syscall->enabled) cmp $0x0,%rdi 0.95 → je 0 mov $0x40,%r8d switch (augmented_args.args.syscall_nr) { mov -0x140(%rbp),%rdi switch (augmented_args.args.syscall_nr) { cmp $0x2,%rdi → je 0 cmp $0x101,%rdi → je 0 cmp $0x15,%rdi → jne 0 case SYS_OPEN: filename_arg = (const void *)args->args[0]; mov 0x10(%rbx),%rdx → jmp 0 case SYS_OPENAT: filename_arg = (const void *)args->args[1]; mov 0x18(%rbx),%rdx if (filename_arg != NULL) { cmp $0x0,%rdx → je 0 xor %edi,%edi augmented_args.filename.reserved = 0; mov %edi,-0x104(%rbp) augmented_args.filename.size = probe_read_str(&augmented_args.filename.value, mov %rbp,%rdi add $0xffffffffffffff00,%rdi augmented_args.filename.size = probe_read_str(&augmented_args.filename.value, mov $0x100,%esi → callq *ffffffffda658499 mov $0x148,%r8d augmented_args.filename.size = probe_read_str(&augmented_args.filename.value, mov %eax,-0x108(%rbp) augmented_args.filename.size = probe_read_str(&augmented_args.filename.value, mov %rax,%rdi shl $0x20,%rdi shr $0x20,%rdi if (augmented_args.filename.size < sizeof(augmented_args.filename.value)) { cmp $0xff,%rdi → ja 0 len -= sizeof(augmented_args.filename.value) - augmented_args.filename.size; add $0x48,%rax len &= sizeof(augmented_args.filename.value) - 1; and $0xff,%rax mov %rax,%r8 mov %rbp,%rcx return perf_event_output(args, &__augmented_syscalls__, BPF_F_CURRENT_CPU, &augmented_args, len); add $0xfffffffffffffeb8,%rcx mov %rbx,%rdi movabs $0xffff975fbd72d800,%rsi mov $0xffffffff,%edx → callq *ffffffffda658ad9 mov %rax,%r13 } mov %r13,%rax 0.72 mov 0x0(%rbp),%rbx mov 0x8(%rbp),%r13 1.16 mov 0x10(%rbp),%r14 0.10 mov 0x18(%rbp),%r15 0.42 add $0x28,%rbp 0.54 leaveq 0.54 ← retq # Please see 'man perf-config' to see how to control what should be seen, via ~/.perfconfig [annotate] section, for instance, one can suppress the source code and see just the disassembly, etc. Alternatively, use the TUI bu just using 'perf annotate', press '/bpf_prog' to see the bpf symbols, press enter and do the interactive annotation, which allows for dumping to a file after selecting the the various output tunables, for instance, the above without source code intermixed, plus showing all the instruction offsets: # perf annotate bpf_prog_819967866022f1e1_sys_enter Then press: 's' to hide the source code + 'O' twice to show all instruction offsets, then 'P' to print to the bpf_prog_819967866022f1e1_sys_enter.annotation file, which will have: # cat bpf_prog_819967866022f1e1_sys_enter.annotation bpf_prog_819967866022f1e1_sys_enter() bpf_prog_819967866022f1e1_sys_enter Event: cycles:ppp 53.41 0: push %rbp 0.63 1: mov %rsp,%rbp 0.31 4: sub $0x170,%rsp 1.93 b: sub $0x28,%rbp 7.02 f: mov %rbx,0x0(%rbp) 3.20 13: mov %r13,0x8(%rbp) 1.07 17: mov %r14,0x10(%rbp) 0.61 1b: mov %r15,0x18(%rbp) 0.11 1f: xor %eax,%eax 1.29 21: mov %rax,0x20(%rbp) 0.11 25: mov %rdi,%rbx 2.02 28: → callq *ffffffffda6776d9 2.76 2d: mov %eax,-0x148(%rbp) 33: mov %rbp,%rsi 36: add $0xfffffffffffffeb8,%rsi 3d: movabs $0xffff975ac2607800,%rdi 1.26 47: → callq *ffffffffda6789e9 4c: cmp $0x0,%rax 2.43 50: → je 0 52: add $0x38,%rax 0.21 56: xor %r13d,%r13d 0.81 59: cmp $0x0,%rax 5d: → jne 0 63: mov %rbp,%rdi 2.22 66: add $0xfffffffffffffeb8,%rdi 0.11 6d: mov $0x40,%esi 0.32 72: mov %rbx,%rdx 2.74 75: → callq *ffffffffda658409 0.22 7a: mov %rbp,%rsi 1.69 7d: add $0xfffffffffffffec0,%rsi 84: movabs $0xffff975bfcd36000,%rdi 8e: add $0xd0,%rdi 0.21 95: mov 0x0(%rsi),%eax 0.93 98: cmp $0x200,%rax 9f: → jae 0 0.10 a1: shl $0x3,%rax 0.11 a5: add %rdi,%rax 0.11 a8: → jmp 0 aa: xor %eax,%eax 1.07 ac: cmp $0x0,%rax b0: → je 0 6.57 b6: movzbq 0x0(%rax),%rdi bb: cmp $0x0,%rdi 0.95 bf: → je 0 c5: mov $0x40,%r8d cb: mov -0x140(%rbp),%rdi d2: cmp $0x2,%rdi d6: → je 0 d8: cmp $0x101,%rdi df: → je 0 e1: cmp $0x15,%rdi e5: → jne 0 e7: mov 0x10(%rbx),%rdx eb: → jmp 0 ed: mov 0x18(%rbx),%rdx f1: cmp $0x0,%rdx f5: → je 0 f7: xor %edi,%edi f9: mov %edi,-0x104(%rbp) ff: mov %rbp,%rdi 102: add $0xffffffffffffff00,%rdi 109: mov $0x100,%esi 10e: → callq *ffffffffda658499 113: mov $0x148,%r8d 119: mov %eax,-0x108(%rbp) 11f: mov %rax,%rdi 122: shl $0x20,%rdi 126: shr $0x20,%rdi 12a: cmp $0xff,%rdi 131: → ja 0 133: add $0x48,%rax 137: and $0xff,%rax 13d: mov %rax,%r8 140: mov %rbp,%rcx 143: add $0xfffffffffffffeb8,%rcx 14a: mov %rbx,%rdi 14d: movabs $0xffff975fbd72d800,%rsi 157: mov $0xffffffff,%edx 15c: → callq *ffffffffda658ad9 161: mov %rax,%r13 164: mov %r13,%rax 0.72 167: mov 0x0(%rbp),%rbx 16b: mov 0x8(%rbp),%r13 1.16 16f: mov 0x10(%rbp),%r14 0.10 173: mov 0x18(%rbp),%r15 0.42 177: add $0x28,%rbp 0.54 17b: leaveq 0.54 17c: ← retq Another cool way to test all this is to symple use 'perf top' look for those symbols, go there and press enter, annotate it live :-) Signed-off-by: Song Liu <songliubraving@fb.com> Reviewed-by: Jiri Olsa <jolsa@kernel.org> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Alexei Starovoitov <ast@kernel.org> Cc: Daniel Borkmann <daniel@iogearbox.net> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Stanislav Fomichev <sdf@google.com> Link: http://lkml.kernel.org/r/20190312053051.2690567-13-songliubraving@fb.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2019-03-12 08:30:48 +03:00
info.buffer_length = info_linear->info.jited_prog_len;
if (info_linear->info.nr_line_info)
prog_linfo = bpf_prog_linfo__new(&info_linear->info);
if (info_linear->info.btf_id) {
struct btf_node *node;
node = perf_env__find_btf(dso->bpf_prog.env,
info_linear->info.btf_id);
if (node)
btf = btf__new((__u8 *)(node->data),
node->data_size);
}
disassemble_init_for_target(&info);
#ifdef DISASM_FOUR_ARGS_SIGNATURE
disassemble = disassembler(info.arch,
bfd_big_endian(bfdf),
info.mach,
bfdf);
#else
disassemble = disassembler(bfdf);
#endif
assert(disassemble);
fflush(s);
do {
const struct bpf_line_info *linfo = NULL;
struct disasm_line *dl;
size_t prev_buf_size;
const char *srcline;
u64 addr;
addr = pc + ((u64 *)(uintptr_t)(info_linear->info.jited_ksyms))[sub_id];
perf annotate: Enable annotation of BPF programs In symbol__disassemble(), DSO_BINARY_TYPE__BPF_PROG_INFO dso calls into a new function symbol__disassemble_bpf(), where annotation line information is filled based on the bpf_prog_info and btf data saved in given perf_env. symbol__disassemble_bpf() uses binutils's libopcodes to disassemble bpf programs. Committer testing: After fixing this: - u64 *addrs = (u64 *)(info_linear->info.jited_ksyms); + u64 *addrs = (u64 *)(uintptr_t)(info_linear->info.jited_ksyms); Detected when crossbuilding to a 32-bit arch. And making all this dependent on HAVE_LIBBFD_SUPPORT and HAVE_LIBBPF_SUPPORT: 1) Have a BPF program running, one that has BTF info, etc, I used the tools/perf/examples/bpf/augmented_raw_syscalls.c put in place by 'perf trace'. # grep -B1 augmented_raw ~/.perfconfig [trace] add_events = /home/acme/git/perf/tools/perf/examples/bpf/augmented_raw_syscalls.c # # perf trace -e *mmsg dnf/6245 sendmmsg(20, 0x7f5485a88030, 2, MSG_NOSIGNAL) = 2 NetworkManager/10055 sendmmsg(22<socket:[1056822]>, 0x7f8126ad1bb0, 2, MSG_NOSIGNAL) = 2 2) Then do a 'perf record' system wide for a while: # perf record -a ^C[ perf record: Woken up 68 times to write data ] [ perf record: Captured and wrote 19.427 MB perf.data (366891 samples) ] # 3) Check that we captured BPF and BTF info in the perf.data file: # perf report --header-only | grep 'b[pt]f' # event : name = cycles:ppp, , id = { 294789, 294790, 294791, 294792, 294793, 294794, 294795, 294796 }, size = 112, { sample_period, sample_freq } = 4000, sample_type = IP|TID|TIME|CPU|PERIOD, read_format = ID, disabled = 1, inherit = 1, mmap = 1, comm = 1, freq = 1, task = 1, precise_ip = 3, sample_id_all = 1, exclude_guest = 1, mmap2 = 1, comm_exec = 1, ksymbol = 1, bpf_event = 1 # bpf_prog_info of id 13 # bpf_prog_info of id 14 # bpf_prog_info of id 15 # bpf_prog_info of id 16 # bpf_prog_info of id 17 # bpf_prog_info of id 18 # bpf_prog_info of id 21 # bpf_prog_info of id 22 # bpf_prog_info of id 41 # bpf_prog_info of id 42 # btf info of id 2 # 4) Check which programs got recorded: # perf report | grep bpf_prog | head 0.16% exe bpf_prog_819967866022f1e1_sys_enter [k] bpf_prog_819967866022f1e1_sys_enter 0.14% exe bpf_prog_c1bd85c092d6e4aa_sys_exit [k] bpf_prog_c1bd85c092d6e4aa_sys_exit 0.08% fuse-overlayfs bpf_prog_819967866022f1e1_sys_enter [k] bpf_prog_819967866022f1e1_sys_enter 0.07% fuse-overlayfs bpf_prog_c1bd85c092d6e4aa_sys_exit [k] bpf_prog_c1bd85c092d6e4aa_sys_exit 0.01% clang-4.0 bpf_prog_c1bd85c092d6e4aa_sys_exit [k] bpf_prog_c1bd85c092d6e4aa_sys_exit 0.01% clang-4.0 bpf_prog_819967866022f1e1_sys_enter [k] bpf_prog_819967866022f1e1_sys_enter 0.00% clang bpf_prog_c1bd85c092d6e4aa_sys_exit [k] bpf_prog_c1bd85c092d6e4aa_sys_exit 0.00% runc bpf_prog_819967866022f1e1_sys_enter [k] bpf_prog_819967866022f1e1_sys_enter 0.00% clang bpf_prog_819967866022f1e1_sys_enter [k] bpf_prog_819967866022f1e1_sys_enter 0.00% sh bpf_prog_c1bd85c092d6e4aa_sys_exit [k] bpf_prog_c1bd85c092d6e4aa_sys_exit # This was with the default --sort order for 'perf report', which is: --sort comm,dso,symbol If we just look for the symbol, for instance: # perf report --sort symbol | grep bpf_prog | head 0.26% [k] bpf_prog_819967866022f1e1_sys_enter - - 0.24% [k] bpf_prog_c1bd85c092d6e4aa_sys_exit - - # or the DSO: # perf report --sort dso | grep bpf_prog | head 0.26% bpf_prog_819967866022f1e1_sys_enter 0.24% bpf_prog_c1bd85c092d6e4aa_sys_exit # We'll see the two BPF programs that augmented_raw_syscalls.o puts in place, one attached to the raw_syscalls:sys_enter and another to the raw_syscalls:sys_exit tracepoints, as expected. Now we can finally do, from the command line, annotation for one of those two symbols, with the original BPF program source coude intermixed with the disassembled JITed code: # perf annotate --stdio2 bpf_prog_819967866022f1e1_sys_enter Samples: 950 of event 'cycles:ppp', 4000 Hz, Event count (approx.): 553756947, [percent: local period] bpf_prog_819967866022f1e1_sys_enter() bpf_prog_819967866022f1e1_sys_enter Percent int sys_enter(struct syscall_enter_args *args) 53.41 push %rbp 0.63 mov %rsp,%rbp 0.31 sub $0x170,%rsp 1.93 sub $0x28,%rbp 7.02 mov %rbx,0x0(%rbp) 3.20 mov %r13,0x8(%rbp) 1.07 mov %r14,0x10(%rbp) 0.61 mov %r15,0x18(%rbp) 0.11 xor %eax,%eax 1.29 mov %rax,0x20(%rbp) 0.11 mov %rdi,%rbx return bpf_get_current_pid_tgid(); 2.02 → callq *ffffffffda6776d9 2.76 mov %eax,-0x148(%rbp) mov %rbp,%rsi int sys_enter(struct syscall_enter_args *args) add $0xfffffffffffffeb8,%rsi return bpf_map_lookup_elem(pids, &pid) != NULL; movabs $0xffff975ac2607800,%rdi 1.26 → callq *ffffffffda6789e9 cmp $0x0,%rax 2.43 → je 0 add $0x38,%rax 0.21 xor %r13d,%r13d if (pid_filter__has(&pids_filtered, getpid())) 0.81 cmp $0x0,%rax → jne 0 mov %rbp,%rdi probe_read(&augmented_args.args, sizeof(augmented_args.args), args); 2.22 add $0xfffffffffffffeb8,%rdi 0.11 mov $0x40,%esi 0.32 mov %rbx,%rdx 2.74 → callq *ffffffffda658409 syscall = bpf_map_lookup_elem(&syscalls, &augmented_args.args.syscall_nr); 0.22 mov %rbp,%rsi 1.69 add $0xfffffffffffffec0,%rsi syscall = bpf_map_lookup_elem(&syscalls, &augmented_args.args.syscall_nr); movabs $0xffff975bfcd36000,%rdi add $0xd0,%rdi 0.21 mov 0x0(%rsi),%eax 0.93 cmp $0x200,%rax → jae 0 0.10 shl $0x3,%rax 0.11 add %rdi,%rax 0.11 → jmp 0 xor %eax,%eax if (syscall == NULL || !syscall->enabled) 1.07 cmp $0x0,%rax → je 0 if (syscall == NULL || !syscall->enabled) 6.57 movzbq 0x0(%rax),%rdi if (syscall == NULL || !syscall->enabled) cmp $0x0,%rdi 0.95 → je 0 mov $0x40,%r8d switch (augmented_args.args.syscall_nr) { mov -0x140(%rbp),%rdi switch (augmented_args.args.syscall_nr) { cmp $0x2,%rdi → je 0 cmp $0x101,%rdi → je 0 cmp $0x15,%rdi → jne 0 case SYS_OPEN: filename_arg = (const void *)args->args[0]; mov 0x10(%rbx),%rdx → jmp 0 case SYS_OPENAT: filename_arg = (const void *)args->args[1]; mov 0x18(%rbx),%rdx if (filename_arg != NULL) { cmp $0x0,%rdx → je 0 xor %edi,%edi augmented_args.filename.reserved = 0; mov %edi,-0x104(%rbp) augmented_args.filename.size = probe_read_str(&augmented_args.filename.value, mov %rbp,%rdi add $0xffffffffffffff00,%rdi augmented_args.filename.size = probe_read_str(&augmented_args.filename.value, mov $0x100,%esi → callq *ffffffffda658499 mov $0x148,%r8d augmented_args.filename.size = probe_read_str(&augmented_args.filename.value, mov %eax,-0x108(%rbp) augmented_args.filename.size = probe_read_str(&augmented_args.filename.value, mov %rax,%rdi shl $0x20,%rdi shr $0x20,%rdi if (augmented_args.filename.size < sizeof(augmented_args.filename.value)) { cmp $0xff,%rdi → ja 0 len -= sizeof(augmented_args.filename.value) - augmented_args.filename.size; add $0x48,%rax len &= sizeof(augmented_args.filename.value) - 1; and $0xff,%rax mov %rax,%r8 mov %rbp,%rcx return perf_event_output(args, &__augmented_syscalls__, BPF_F_CURRENT_CPU, &augmented_args, len); add $0xfffffffffffffeb8,%rcx mov %rbx,%rdi movabs $0xffff975fbd72d800,%rsi mov $0xffffffff,%edx → callq *ffffffffda658ad9 mov %rax,%r13 } mov %r13,%rax 0.72 mov 0x0(%rbp),%rbx mov 0x8(%rbp),%r13 1.16 mov 0x10(%rbp),%r14 0.10 mov 0x18(%rbp),%r15 0.42 add $0x28,%rbp 0.54 leaveq 0.54 ← retq # Please see 'man perf-config' to see how to control what should be seen, via ~/.perfconfig [annotate] section, for instance, one can suppress the source code and see just the disassembly, etc. Alternatively, use the TUI bu just using 'perf annotate', press '/bpf_prog' to see the bpf symbols, press enter and do the interactive annotation, which allows for dumping to a file after selecting the the various output tunables, for instance, the above without source code intermixed, plus showing all the instruction offsets: # perf annotate bpf_prog_819967866022f1e1_sys_enter Then press: 's' to hide the source code + 'O' twice to show all instruction offsets, then 'P' to print to the bpf_prog_819967866022f1e1_sys_enter.annotation file, which will have: # cat bpf_prog_819967866022f1e1_sys_enter.annotation bpf_prog_819967866022f1e1_sys_enter() bpf_prog_819967866022f1e1_sys_enter Event: cycles:ppp 53.41 0: push %rbp 0.63 1: mov %rsp,%rbp 0.31 4: sub $0x170,%rsp 1.93 b: sub $0x28,%rbp 7.02 f: mov %rbx,0x0(%rbp) 3.20 13: mov %r13,0x8(%rbp) 1.07 17: mov %r14,0x10(%rbp) 0.61 1b: mov %r15,0x18(%rbp) 0.11 1f: xor %eax,%eax 1.29 21: mov %rax,0x20(%rbp) 0.11 25: mov %rdi,%rbx 2.02 28: → callq *ffffffffda6776d9 2.76 2d: mov %eax,-0x148(%rbp) 33: mov %rbp,%rsi 36: add $0xfffffffffffffeb8,%rsi 3d: movabs $0xffff975ac2607800,%rdi 1.26 47: → callq *ffffffffda6789e9 4c: cmp $0x0,%rax 2.43 50: → je 0 52: add $0x38,%rax 0.21 56: xor %r13d,%r13d 0.81 59: cmp $0x0,%rax 5d: → jne 0 63: mov %rbp,%rdi 2.22 66: add $0xfffffffffffffeb8,%rdi 0.11 6d: mov $0x40,%esi 0.32 72: mov %rbx,%rdx 2.74 75: → callq *ffffffffda658409 0.22 7a: mov %rbp,%rsi 1.69 7d: add $0xfffffffffffffec0,%rsi 84: movabs $0xffff975bfcd36000,%rdi 8e: add $0xd0,%rdi 0.21 95: mov 0x0(%rsi),%eax 0.93 98: cmp $0x200,%rax 9f: → jae 0 0.10 a1: shl $0x3,%rax 0.11 a5: add %rdi,%rax 0.11 a8: → jmp 0 aa: xor %eax,%eax 1.07 ac: cmp $0x0,%rax b0: → je 0 6.57 b6: movzbq 0x0(%rax),%rdi bb: cmp $0x0,%rdi 0.95 bf: → je 0 c5: mov $0x40,%r8d cb: mov -0x140(%rbp),%rdi d2: cmp $0x2,%rdi d6: → je 0 d8: cmp $0x101,%rdi df: → je 0 e1: cmp $0x15,%rdi e5: → jne 0 e7: mov 0x10(%rbx),%rdx eb: → jmp 0 ed: mov 0x18(%rbx),%rdx f1: cmp $0x0,%rdx f5: → je 0 f7: xor %edi,%edi f9: mov %edi,-0x104(%rbp) ff: mov %rbp,%rdi 102: add $0xffffffffffffff00,%rdi 109: mov $0x100,%esi 10e: → callq *ffffffffda658499 113: mov $0x148,%r8d 119: mov %eax,-0x108(%rbp) 11f: mov %rax,%rdi 122: shl $0x20,%rdi 126: shr $0x20,%rdi 12a: cmp $0xff,%rdi 131: → ja 0 133: add $0x48,%rax 137: and $0xff,%rax 13d: mov %rax,%r8 140: mov %rbp,%rcx 143: add $0xfffffffffffffeb8,%rcx 14a: mov %rbx,%rdi 14d: movabs $0xffff975fbd72d800,%rsi 157: mov $0xffffffff,%edx 15c: → callq *ffffffffda658ad9 161: mov %rax,%r13 164: mov %r13,%rax 0.72 167: mov 0x0(%rbp),%rbx 16b: mov 0x8(%rbp),%r13 1.16 16f: mov 0x10(%rbp),%r14 0.10 173: mov 0x18(%rbp),%r15 0.42 177: add $0x28,%rbp 0.54 17b: leaveq 0.54 17c: ← retq Another cool way to test all this is to symple use 'perf top' look for those symbols, go there and press enter, annotate it live :-) Signed-off-by: Song Liu <songliubraving@fb.com> Reviewed-by: Jiri Olsa <jolsa@kernel.org> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Alexei Starovoitov <ast@kernel.org> Cc: Daniel Borkmann <daniel@iogearbox.net> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Stanislav Fomichev <sdf@google.com> Link: http://lkml.kernel.org/r/20190312053051.2690567-13-songliubraving@fb.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2019-03-12 08:30:48 +03:00
count = disassemble(pc, &info);
if (prog_linfo)
linfo = bpf_prog_linfo__lfind_addr_func(prog_linfo,
addr, sub_id,
nr_skip);
if (linfo && btf) {
srcline = btf__name_by_offset(btf, linfo->line_off);
nr_skip++;
} else
srcline = NULL;
fprintf(s, "\n");
prev_buf_size = buf_size;
fflush(s);
if (!opts->hide_src_code && srcline) {
args->offset = -1;
args->line = strdup(srcline);
args->line_nr = 0;
args->fileloc = NULL;
perf annotate: Enable annotation of BPF programs In symbol__disassemble(), DSO_BINARY_TYPE__BPF_PROG_INFO dso calls into a new function symbol__disassemble_bpf(), where annotation line information is filled based on the bpf_prog_info and btf data saved in given perf_env. symbol__disassemble_bpf() uses binutils's libopcodes to disassemble bpf programs. Committer testing: After fixing this: - u64 *addrs = (u64 *)(info_linear->info.jited_ksyms); + u64 *addrs = (u64 *)(uintptr_t)(info_linear->info.jited_ksyms); Detected when crossbuilding to a 32-bit arch. And making all this dependent on HAVE_LIBBFD_SUPPORT and HAVE_LIBBPF_SUPPORT: 1) Have a BPF program running, one that has BTF info, etc, I used the tools/perf/examples/bpf/augmented_raw_syscalls.c put in place by 'perf trace'. # grep -B1 augmented_raw ~/.perfconfig [trace] add_events = /home/acme/git/perf/tools/perf/examples/bpf/augmented_raw_syscalls.c # # perf trace -e *mmsg dnf/6245 sendmmsg(20, 0x7f5485a88030, 2, MSG_NOSIGNAL) = 2 NetworkManager/10055 sendmmsg(22<socket:[1056822]>, 0x7f8126ad1bb0, 2, MSG_NOSIGNAL) = 2 2) Then do a 'perf record' system wide for a while: # perf record -a ^C[ perf record: Woken up 68 times to write data ] [ perf record: Captured and wrote 19.427 MB perf.data (366891 samples) ] # 3) Check that we captured BPF and BTF info in the perf.data file: # perf report --header-only | grep 'b[pt]f' # event : name = cycles:ppp, , id = { 294789, 294790, 294791, 294792, 294793, 294794, 294795, 294796 }, size = 112, { sample_period, sample_freq } = 4000, sample_type = IP|TID|TIME|CPU|PERIOD, read_format = ID, disabled = 1, inherit = 1, mmap = 1, comm = 1, freq = 1, task = 1, precise_ip = 3, sample_id_all = 1, exclude_guest = 1, mmap2 = 1, comm_exec = 1, ksymbol = 1, bpf_event = 1 # bpf_prog_info of id 13 # bpf_prog_info of id 14 # bpf_prog_info of id 15 # bpf_prog_info of id 16 # bpf_prog_info of id 17 # bpf_prog_info of id 18 # bpf_prog_info of id 21 # bpf_prog_info of id 22 # bpf_prog_info of id 41 # bpf_prog_info of id 42 # btf info of id 2 # 4) Check which programs got recorded: # perf report | grep bpf_prog | head 0.16% exe bpf_prog_819967866022f1e1_sys_enter [k] bpf_prog_819967866022f1e1_sys_enter 0.14% exe bpf_prog_c1bd85c092d6e4aa_sys_exit [k] bpf_prog_c1bd85c092d6e4aa_sys_exit 0.08% fuse-overlayfs bpf_prog_819967866022f1e1_sys_enter [k] bpf_prog_819967866022f1e1_sys_enter 0.07% fuse-overlayfs bpf_prog_c1bd85c092d6e4aa_sys_exit [k] bpf_prog_c1bd85c092d6e4aa_sys_exit 0.01% clang-4.0 bpf_prog_c1bd85c092d6e4aa_sys_exit [k] bpf_prog_c1bd85c092d6e4aa_sys_exit 0.01% clang-4.0 bpf_prog_819967866022f1e1_sys_enter [k] bpf_prog_819967866022f1e1_sys_enter 0.00% clang bpf_prog_c1bd85c092d6e4aa_sys_exit [k] bpf_prog_c1bd85c092d6e4aa_sys_exit 0.00% runc bpf_prog_819967866022f1e1_sys_enter [k] bpf_prog_819967866022f1e1_sys_enter 0.00% clang bpf_prog_819967866022f1e1_sys_enter [k] bpf_prog_819967866022f1e1_sys_enter 0.00% sh bpf_prog_c1bd85c092d6e4aa_sys_exit [k] bpf_prog_c1bd85c092d6e4aa_sys_exit # This was with the default --sort order for 'perf report', which is: --sort comm,dso,symbol If we just look for the symbol, for instance: # perf report --sort symbol | grep bpf_prog | head 0.26% [k] bpf_prog_819967866022f1e1_sys_enter - - 0.24% [k] bpf_prog_c1bd85c092d6e4aa_sys_exit - - # or the DSO: # perf report --sort dso | grep bpf_prog | head 0.26% bpf_prog_819967866022f1e1_sys_enter 0.24% bpf_prog_c1bd85c092d6e4aa_sys_exit # We'll see the two BPF programs that augmented_raw_syscalls.o puts in place, one attached to the raw_syscalls:sys_enter and another to the raw_syscalls:sys_exit tracepoints, as expected. Now we can finally do, from the command line, annotation for one of those two symbols, with the original BPF program source coude intermixed with the disassembled JITed code: # perf annotate --stdio2 bpf_prog_819967866022f1e1_sys_enter Samples: 950 of event 'cycles:ppp', 4000 Hz, Event count (approx.): 553756947, [percent: local period] bpf_prog_819967866022f1e1_sys_enter() bpf_prog_819967866022f1e1_sys_enter Percent int sys_enter(struct syscall_enter_args *args) 53.41 push %rbp 0.63 mov %rsp,%rbp 0.31 sub $0x170,%rsp 1.93 sub $0x28,%rbp 7.02 mov %rbx,0x0(%rbp) 3.20 mov %r13,0x8(%rbp) 1.07 mov %r14,0x10(%rbp) 0.61 mov %r15,0x18(%rbp) 0.11 xor %eax,%eax 1.29 mov %rax,0x20(%rbp) 0.11 mov %rdi,%rbx return bpf_get_current_pid_tgid(); 2.02 → callq *ffffffffda6776d9 2.76 mov %eax,-0x148(%rbp) mov %rbp,%rsi int sys_enter(struct syscall_enter_args *args) add $0xfffffffffffffeb8,%rsi return bpf_map_lookup_elem(pids, &pid) != NULL; movabs $0xffff975ac2607800,%rdi 1.26 → callq *ffffffffda6789e9 cmp $0x0,%rax 2.43 → je 0 add $0x38,%rax 0.21 xor %r13d,%r13d if (pid_filter__has(&pids_filtered, getpid())) 0.81 cmp $0x0,%rax → jne 0 mov %rbp,%rdi probe_read(&augmented_args.args, sizeof(augmented_args.args), args); 2.22 add $0xfffffffffffffeb8,%rdi 0.11 mov $0x40,%esi 0.32 mov %rbx,%rdx 2.74 → callq *ffffffffda658409 syscall = bpf_map_lookup_elem(&syscalls, &augmented_args.args.syscall_nr); 0.22 mov %rbp,%rsi 1.69 add $0xfffffffffffffec0,%rsi syscall = bpf_map_lookup_elem(&syscalls, &augmented_args.args.syscall_nr); movabs $0xffff975bfcd36000,%rdi add $0xd0,%rdi 0.21 mov 0x0(%rsi),%eax 0.93 cmp $0x200,%rax → jae 0 0.10 shl $0x3,%rax 0.11 add %rdi,%rax 0.11 → jmp 0 xor %eax,%eax if (syscall == NULL || !syscall->enabled) 1.07 cmp $0x0,%rax → je 0 if (syscall == NULL || !syscall->enabled) 6.57 movzbq 0x0(%rax),%rdi if (syscall == NULL || !syscall->enabled) cmp $0x0,%rdi 0.95 → je 0 mov $0x40,%r8d switch (augmented_args.args.syscall_nr) { mov -0x140(%rbp),%rdi switch (augmented_args.args.syscall_nr) { cmp $0x2,%rdi → je 0 cmp $0x101,%rdi → je 0 cmp $0x15,%rdi → jne 0 case SYS_OPEN: filename_arg = (const void *)args->args[0]; mov 0x10(%rbx),%rdx → jmp 0 case SYS_OPENAT: filename_arg = (const void *)args->args[1]; mov 0x18(%rbx),%rdx if (filename_arg != NULL) { cmp $0x0,%rdx → je 0 xor %edi,%edi augmented_args.filename.reserved = 0; mov %edi,-0x104(%rbp) augmented_args.filename.size = probe_read_str(&augmented_args.filename.value, mov %rbp,%rdi add $0xffffffffffffff00,%rdi augmented_args.filename.size = probe_read_str(&augmented_args.filename.value, mov $0x100,%esi → callq *ffffffffda658499 mov $0x148,%r8d augmented_args.filename.size = probe_read_str(&augmented_args.filename.value, mov %eax,-0x108(%rbp) augmented_args.filename.size = probe_read_str(&augmented_args.filename.value, mov %rax,%rdi shl $0x20,%rdi shr $0x20,%rdi if (augmented_args.filename.size < sizeof(augmented_args.filename.value)) { cmp $0xff,%rdi → ja 0 len -= sizeof(augmented_args.filename.value) - augmented_args.filename.size; add $0x48,%rax len &= sizeof(augmented_args.filename.value) - 1; and $0xff,%rax mov %rax,%r8 mov %rbp,%rcx return perf_event_output(args, &__augmented_syscalls__, BPF_F_CURRENT_CPU, &augmented_args, len); add $0xfffffffffffffeb8,%rcx mov %rbx,%rdi movabs $0xffff975fbd72d800,%rsi mov $0xffffffff,%edx → callq *ffffffffda658ad9 mov %rax,%r13 } mov %r13,%rax 0.72 mov 0x0(%rbp),%rbx mov 0x8(%rbp),%r13 1.16 mov 0x10(%rbp),%r14 0.10 mov 0x18(%rbp),%r15 0.42 add $0x28,%rbp 0.54 leaveq 0.54 ← retq # Please see 'man perf-config' to see how to control what should be seen, via ~/.perfconfig [annotate] section, for instance, one can suppress the source code and see just the disassembly, etc. Alternatively, use the TUI bu just using 'perf annotate', press '/bpf_prog' to see the bpf symbols, press enter and do the interactive annotation, which allows for dumping to a file after selecting the the various output tunables, for instance, the above without source code intermixed, plus showing all the instruction offsets: # perf annotate bpf_prog_819967866022f1e1_sys_enter Then press: 's' to hide the source code + 'O' twice to show all instruction offsets, then 'P' to print to the bpf_prog_819967866022f1e1_sys_enter.annotation file, which will have: # cat bpf_prog_819967866022f1e1_sys_enter.annotation bpf_prog_819967866022f1e1_sys_enter() bpf_prog_819967866022f1e1_sys_enter Event: cycles:ppp 53.41 0: push %rbp 0.63 1: mov %rsp,%rbp 0.31 4: sub $0x170,%rsp 1.93 b: sub $0x28,%rbp 7.02 f: mov %rbx,0x0(%rbp) 3.20 13: mov %r13,0x8(%rbp) 1.07 17: mov %r14,0x10(%rbp) 0.61 1b: mov %r15,0x18(%rbp) 0.11 1f: xor %eax,%eax 1.29 21: mov %rax,0x20(%rbp) 0.11 25: mov %rdi,%rbx 2.02 28: → callq *ffffffffda6776d9 2.76 2d: mov %eax,-0x148(%rbp) 33: mov %rbp,%rsi 36: add $0xfffffffffffffeb8,%rsi 3d: movabs $0xffff975ac2607800,%rdi 1.26 47: → callq *ffffffffda6789e9 4c: cmp $0x0,%rax 2.43 50: → je 0 52: add $0x38,%rax 0.21 56: xor %r13d,%r13d 0.81 59: cmp $0x0,%rax 5d: → jne 0 63: mov %rbp,%rdi 2.22 66: add $0xfffffffffffffeb8,%rdi 0.11 6d: mov $0x40,%esi 0.32 72: mov %rbx,%rdx 2.74 75: → callq *ffffffffda658409 0.22 7a: mov %rbp,%rsi 1.69 7d: add $0xfffffffffffffec0,%rsi 84: movabs $0xffff975bfcd36000,%rdi 8e: add $0xd0,%rdi 0.21 95: mov 0x0(%rsi),%eax 0.93 98: cmp $0x200,%rax 9f: → jae 0 0.10 a1: shl $0x3,%rax 0.11 a5: add %rdi,%rax 0.11 a8: → jmp 0 aa: xor %eax,%eax 1.07 ac: cmp $0x0,%rax b0: → je 0 6.57 b6: movzbq 0x0(%rax),%rdi bb: cmp $0x0,%rdi 0.95 bf: → je 0 c5: mov $0x40,%r8d cb: mov -0x140(%rbp),%rdi d2: cmp $0x2,%rdi d6: → je 0 d8: cmp $0x101,%rdi df: → je 0 e1: cmp $0x15,%rdi e5: → jne 0 e7: mov 0x10(%rbx),%rdx eb: → jmp 0 ed: mov 0x18(%rbx),%rdx f1: cmp $0x0,%rdx f5: → je 0 f7: xor %edi,%edi f9: mov %edi,-0x104(%rbp) ff: mov %rbp,%rdi 102: add $0xffffffffffffff00,%rdi 109: mov $0x100,%esi 10e: → callq *ffffffffda658499 113: mov $0x148,%r8d 119: mov %eax,-0x108(%rbp) 11f: mov %rax,%rdi 122: shl $0x20,%rdi 126: shr $0x20,%rdi 12a: cmp $0xff,%rdi 131: → ja 0 133: add $0x48,%rax 137: and $0xff,%rax 13d: mov %rax,%r8 140: mov %rbp,%rcx 143: add $0xfffffffffffffeb8,%rcx 14a: mov %rbx,%rdi 14d: movabs $0xffff975fbd72d800,%rsi 157: mov $0xffffffff,%edx 15c: → callq *ffffffffda658ad9 161: mov %rax,%r13 164: mov %r13,%rax 0.72 167: mov 0x0(%rbp),%rbx 16b: mov 0x8(%rbp),%r13 1.16 16f: mov 0x10(%rbp),%r14 0.10 173: mov 0x18(%rbp),%r15 0.42 177: add $0x28,%rbp 0.54 17b: leaveq 0.54 17c: ← retq Another cool way to test all this is to symple use 'perf top' look for those symbols, go there and press enter, annotate it live :-) Signed-off-by: Song Liu <songliubraving@fb.com> Reviewed-by: Jiri Olsa <jolsa@kernel.org> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Alexei Starovoitov <ast@kernel.org> Cc: Daniel Borkmann <daniel@iogearbox.net> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Stanislav Fomichev <sdf@google.com> Link: http://lkml.kernel.org/r/20190312053051.2690567-13-songliubraving@fb.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2019-03-12 08:30:48 +03:00
args->ms.sym = sym;
dl = disasm_line__new(args);
if (dl) {
annotation_line__add(&dl->al,
&notes->src->source);
}
}
args->offset = pc;
args->line = buf + prev_buf_size;
args->line_nr = 0;
args->fileloc = NULL;
perf annotate: Enable annotation of BPF programs In symbol__disassemble(), DSO_BINARY_TYPE__BPF_PROG_INFO dso calls into a new function symbol__disassemble_bpf(), where annotation line information is filled based on the bpf_prog_info and btf data saved in given perf_env. symbol__disassemble_bpf() uses binutils's libopcodes to disassemble bpf programs. Committer testing: After fixing this: - u64 *addrs = (u64 *)(info_linear->info.jited_ksyms); + u64 *addrs = (u64 *)(uintptr_t)(info_linear->info.jited_ksyms); Detected when crossbuilding to a 32-bit arch. And making all this dependent on HAVE_LIBBFD_SUPPORT and HAVE_LIBBPF_SUPPORT: 1) Have a BPF program running, one that has BTF info, etc, I used the tools/perf/examples/bpf/augmented_raw_syscalls.c put in place by 'perf trace'. # grep -B1 augmented_raw ~/.perfconfig [trace] add_events = /home/acme/git/perf/tools/perf/examples/bpf/augmented_raw_syscalls.c # # perf trace -e *mmsg dnf/6245 sendmmsg(20, 0x7f5485a88030, 2, MSG_NOSIGNAL) = 2 NetworkManager/10055 sendmmsg(22<socket:[1056822]>, 0x7f8126ad1bb0, 2, MSG_NOSIGNAL) = 2 2) Then do a 'perf record' system wide for a while: # perf record -a ^C[ perf record: Woken up 68 times to write data ] [ perf record: Captured and wrote 19.427 MB perf.data (366891 samples) ] # 3) Check that we captured BPF and BTF info in the perf.data file: # perf report --header-only | grep 'b[pt]f' # event : name = cycles:ppp, , id = { 294789, 294790, 294791, 294792, 294793, 294794, 294795, 294796 }, size = 112, { sample_period, sample_freq } = 4000, sample_type = IP|TID|TIME|CPU|PERIOD, read_format = ID, disabled = 1, inherit = 1, mmap = 1, comm = 1, freq = 1, task = 1, precise_ip = 3, sample_id_all = 1, exclude_guest = 1, mmap2 = 1, comm_exec = 1, ksymbol = 1, bpf_event = 1 # bpf_prog_info of id 13 # bpf_prog_info of id 14 # bpf_prog_info of id 15 # bpf_prog_info of id 16 # bpf_prog_info of id 17 # bpf_prog_info of id 18 # bpf_prog_info of id 21 # bpf_prog_info of id 22 # bpf_prog_info of id 41 # bpf_prog_info of id 42 # btf info of id 2 # 4) Check which programs got recorded: # perf report | grep bpf_prog | head 0.16% exe bpf_prog_819967866022f1e1_sys_enter [k] bpf_prog_819967866022f1e1_sys_enter 0.14% exe bpf_prog_c1bd85c092d6e4aa_sys_exit [k] bpf_prog_c1bd85c092d6e4aa_sys_exit 0.08% fuse-overlayfs bpf_prog_819967866022f1e1_sys_enter [k] bpf_prog_819967866022f1e1_sys_enter 0.07% fuse-overlayfs bpf_prog_c1bd85c092d6e4aa_sys_exit [k] bpf_prog_c1bd85c092d6e4aa_sys_exit 0.01% clang-4.0 bpf_prog_c1bd85c092d6e4aa_sys_exit [k] bpf_prog_c1bd85c092d6e4aa_sys_exit 0.01% clang-4.0 bpf_prog_819967866022f1e1_sys_enter [k] bpf_prog_819967866022f1e1_sys_enter 0.00% clang bpf_prog_c1bd85c092d6e4aa_sys_exit [k] bpf_prog_c1bd85c092d6e4aa_sys_exit 0.00% runc bpf_prog_819967866022f1e1_sys_enter [k] bpf_prog_819967866022f1e1_sys_enter 0.00% clang bpf_prog_819967866022f1e1_sys_enter [k] bpf_prog_819967866022f1e1_sys_enter 0.00% sh bpf_prog_c1bd85c092d6e4aa_sys_exit [k] bpf_prog_c1bd85c092d6e4aa_sys_exit # This was with the default --sort order for 'perf report', which is: --sort comm,dso,symbol If we just look for the symbol, for instance: # perf report --sort symbol | grep bpf_prog | head 0.26% [k] bpf_prog_819967866022f1e1_sys_enter - - 0.24% [k] bpf_prog_c1bd85c092d6e4aa_sys_exit - - # or the DSO: # perf report --sort dso | grep bpf_prog | head 0.26% bpf_prog_819967866022f1e1_sys_enter 0.24% bpf_prog_c1bd85c092d6e4aa_sys_exit # We'll see the two BPF programs that augmented_raw_syscalls.o puts in place, one attached to the raw_syscalls:sys_enter and another to the raw_syscalls:sys_exit tracepoints, as expected. Now we can finally do, from the command line, annotation for one of those two symbols, with the original BPF program source coude intermixed with the disassembled JITed code: # perf annotate --stdio2 bpf_prog_819967866022f1e1_sys_enter Samples: 950 of event 'cycles:ppp', 4000 Hz, Event count (approx.): 553756947, [percent: local period] bpf_prog_819967866022f1e1_sys_enter() bpf_prog_819967866022f1e1_sys_enter Percent int sys_enter(struct syscall_enter_args *args) 53.41 push %rbp 0.63 mov %rsp,%rbp 0.31 sub $0x170,%rsp 1.93 sub $0x28,%rbp 7.02 mov %rbx,0x0(%rbp) 3.20 mov %r13,0x8(%rbp) 1.07 mov %r14,0x10(%rbp) 0.61 mov %r15,0x18(%rbp) 0.11 xor %eax,%eax 1.29 mov %rax,0x20(%rbp) 0.11 mov %rdi,%rbx return bpf_get_current_pid_tgid(); 2.02 → callq *ffffffffda6776d9 2.76 mov %eax,-0x148(%rbp) mov %rbp,%rsi int sys_enter(struct syscall_enter_args *args) add $0xfffffffffffffeb8,%rsi return bpf_map_lookup_elem(pids, &pid) != NULL; movabs $0xffff975ac2607800,%rdi 1.26 → callq *ffffffffda6789e9 cmp $0x0,%rax 2.43 → je 0 add $0x38,%rax 0.21 xor %r13d,%r13d if (pid_filter__has(&pids_filtered, getpid())) 0.81 cmp $0x0,%rax → jne 0 mov %rbp,%rdi probe_read(&augmented_args.args, sizeof(augmented_args.args), args); 2.22 add $0xfffffffffffffeb8,%rdi 0.11 mov $0x40,%esi 0.32 mov %rbx,%rdx 2.74 → callq *ffffffffda658409 syscall = bpf_map_lookup_elem(&syscalls, &augmented_args.args.syscall_nr); 0.22 mov %rbp,%rsi 1.69 add $0xfffffffffffffec0,%rsi syscall = bpf_map_lookup_elem(&syscalls, &augmented_args.args.syscall_nr); movabs $0xffff975bfcd36000,%rdi add $0xd0,%rdi 0.21 mov 0x0(%rsi),%eax 0.93 cmp $0x200,%rax → jae 0 0.10 shl $0x3,%rax 0.11 add %rdi,%rax 0.11 → jmp 0 xor %eax,%eax if (syscall == NULL || !syscall->enabled) 1.07 cmp $0x0,%rax → je 0 if (syscall == NULL || !syscall->enabled) 6.57 movzbq 0x0(%rax),%rdi if (syscall == NULL || !syscall->enabled) cmp $0x0,%rdi 0.95 → je 0 mov $0x40,%r8d switch (augmented_args.args.syscall_nr) { mov -0x140(%rbp),%rdi switch (augmented_args.args.syscall_nr) { cmp $0x2,%rdi → je 0 cmp $0x101,%rdi → je 0 cmp $0x15,%rdi → jne 0 case SYS_OPEN: filename_arg = (const void *)args->args[0]; mov 0x10(%rbx),%rdx → jmp 0 case SYS_OPENAT: filename_arg = (const void *)args->args[1]; mov 0x18(%rbx),%rdx if (filename_arg != NULL) { cmp $0x0,%rdx → je 0 xor %edi,%edi augmented_args.filename.reserved = 0; mov %edi,-0x104(%rbp) augmented_args.filename.size = probe_read_str(&augmented_args.filename.value, mov %rbp,%rdi add $0xffffffffffffff00,%rdi augmented_args.filename.size = probe_read_str(&augmented_args.filename.value, mov $0x100,%esi → callq *ffffffffda658499 mov $0x148,%r8d augmented_args.filename.size = probe_read_str(&augmented_args.filename.value, mov %eax,-0x108(%rbp) augmented_args.filename.size = probe_read_str(&augmented_args.filename.value, mov %rax,%rdi shl $0x20,%rdi shr $0x20,%rdi if (augmented_args.filename.size < sizeof(augmented_args.filename.value)) { cmp $0xff,%rdi → ja 0 len -= sizeof(augmented_args.filename.value) - augmented_args.filename.size; add $0x48,%rax len &= sizeof(augmented_args.filename.value) - 1; and $0xff,%rax mov %rax,%r8 mov %rbp,%rcx return perf_event_output(args, &__augmented_syscalls__, BPF_F_CURRENT_CPU, &augmented_args, len); add $0xfffffffffffffeb8,%rcx mov %rbx,%rdi movabs $0xffff975fbd72d800,%rsi mov $0xffffffff,%edx → callq *ffffffffda658ad9 mov %rax,%r13 } mov %r13,%rax 0.72 mov 0x0(%rbp),%rbx mov 0x8(%rbp),%r13 1.16 mov 0x10(%rbp),%r14 0.10 mov 0x18(%rbp),%r15 0.42 add $0x28,%rbp 0.54 leaveq 0.54 ← retq # Please see 'man perf-config' to see how to control what should be seen, via ~/.perfconfig [annotate] section, for instance, one can suppress the source code and see just the disassembly, etc. Alternatively, use the TUI bu just using 'perf annotate', press '/bpf_prog' to see the bpf symbols, press enter and do the interactive annotation, which allows for dumping to a file after selecting the the various output tunables, for instance, the above without source code intermixed, plus showing all the instruction offsets: # perf annotate bpf_prog_819967866022f1e1_sys_enter Then press: 's' to hide the source code + 'O' twice to show all instruction offsets, then 'P' to print to the bpf_prog_819967866022f1e1_sys_enter.annotation file, which will have: # cat bpf_prog_819967866022f1e1_sys_enter.annotation bpf_prog_819967866022f1e1_sys_enter() bpf_prog_819967866022f1e1_sys_enter Event: cycles:ppp 53.41 0: push %rbp 0.63 1: mov %rsp,%rbp 0.31 4: sub $0x170,%rsp 1.93 b: sub $0x28,%rbp 7.02 f: mov %rbx,0x0(%rbp) 3.20 13: mov %r13,0x8(%rbp) 1.07 17: mov %r14,0x10(%rbp) 0.61 1b: mov %r15,0x18(%rbp) 0.11 1f: xor %eax,%eax 1.29 21: mov %rax,0x20(%rbp) 0.11 25: mov %rdi,%rbx 2.02 28: → callq *ffffffffda6776d9 2.76 2d: mov %eax,-0x148(%rbp) 33: mov %rbp,%rsi 36: add $0xfffffffffffffeb8,%rsi 3d: movabs $0xffff975ac2607800,%rdi 1.26 47: → callq *ffffffffda6789e9 4c: cmp $0x0,%rax 2.43 50: → je 0 52: add $0x38,%rax 0.21 56: xor %r13d,%r13d 0.81 59: cmp $0x0,%rax 5d: → jne 0 63: mov %rbp,%rdi 2.22 66: add $0xfffffffffffffeb8,%rdi 0.11 6d: mov $0x40,%esi 0.32 72: mov %rbx,%rdx 2.74 75: → callq *ffffffffda658409 0.22 7a: mov %rbp,%rsi 1.69 7d: add $0xfffffffffffffec0,%rsi 84: movabs $0xffff975bfcd36000,%rdi 8e: add $0xd0,%rdi 0.21 95: mov 0x0(%rsi),%eax 0.93 98: cmp $0x200,%rax 9f: → jae 0 0.10 a1: shl $0x3,%rax 0.11 a5: add %rdi,%rax 0.11 a8: → jmp 0 aa: xor %eax,%eax 1.07 ac: cmp $0x0,%rax b0: → je 0 6.57 b6: movzbq 0x0(%rax),%rdi bb: cmp $0x0,%rdi 0.95 bf: → je 0 c5: mov $0x40,%r8d cb: mov -0x140(%rbp),%rdi d2: cmp $0x2,%rdi d6: → je 0 d8: cmp $0x101,%rdi df: → je 0 e1: cmp $0x15,%rdi e5: → jne 0 e7: mov 0x10(%rbx),%rdx eb: → jmp 0 ed: mov 0x18(%rbx),%rdx f1: cmp $0x0,%rdx f5: → je 0 f7: xor %edi,%edi f9: mov %edi,-0x104(%rbp) ff: mov %rbp,%rdi 102: add $0xffffffffffffff00,%rdi 109: mov $0x100,%esi 10e: → callq *ffffffffda658499 113: mov $0x148,%r8d 119: mov %eax,-0x108(%rbp) 11f: mov %rax,%rdi 122: shl $0x20,%rdi 126: shr $0x20,%rdi 12a: cmp $0xff,%rdi 131: → ja 0 133: add $0x48,%rax 137: and $0xff,%rax 13d: mov %rax,%r8 140: mov %rbp,%rcx 143: add $0xfffffffffffffeb8,%rcx 14a: mov %rbx,%rdi 14d: movabs $0xffff975fbd72d800,%rsi 157: mov $0xffffffff,%edx 15c: → callq *ffffffffda658ad9 161: mov %rax,%r13 164: mov %r13,%rax 0.72 167: mov 0x0(%rbp),%rbx 16b: mov 0x8(%rbp),%r13 1.16 16f: mov 0x10(%rbp),%r14 0.10 173: mov 0x18(%rbp),%r15 0.42 177: add $0x28,%rbp 0.54 17b: leaveq 0.54 17c: ← retq Another cool way to test all this is to symple use 'perf top' look for those symbols, go there and press enter, annotate it live :-) Signed-off-by: Song Liu <songliubraving@fb.com> Reviewed-by: Jiri Olsa <jolsa@kernel.org> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Alexei Starovoitov <ast@kernel.org> Cc: Daniel Borkmann <daniel@iogearbox.net> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Stanislav Fomichev <sdf@google.com> Link: http://lkml.kernel.org/r/20190312053051.2690567-13-songliubraving@fb.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2019-03-12 08:30:48 +03:00
args->ms.sym = sym;
dl = disasm_line__new(args);
if (dl)
annotation_line__add(&dl->al, &notes->src->source);
pc += count;
} while (count > 0 && pc < len);
ret = 0;
out:
free(prog_linfo);
btf__free(btf);
perf annotate: Enable annotation of BPF programs In symbol__disassemble(), DSO_BINARY_TYPE__BPF_PROG_INFO dso calls into a new function symbol__disassemble_bpf(), where annotation line information is filled based on the bpf_prog_info and btf data saved in given perf_env. symbol__disassemble_bpf() uses binutils's libopcodes to disassemble bpf programs. Committer testing: After fixing this: - u64 *addrs = (u64 *)(info_linear->info.jited_ksyms); + u64 *addrs = (u64 *)(uintptr_t)(info_linear->info.jited_ksyms); Detected when crossbuilding to a 32-bit arch. And making all this dependent on HAVE_LIBBFD_SUPPORT and HAVE_LIBBPF_SUPPORT: 1) Have a BPF program running, one that has BTF info, etc, I used the tools/perf/examples/bpf/augmented_raw_syscalls.c put in place by 'perf trace'. # grep -B1 augmented_raw ~/.perfconfig [trace] add_events = /home/acme/git/perf/tools/perf/examples/bpf/augmented_raw_syscalls.c # # perf trace -e *mmsg dnf/6245 sendmmsg(20, 0x7f5485a88030, 2, MSG_NOSIGNAL) = 2 NetworkManager/10055 sendmmsg(22<socket:[1056822]>, 0x7f8126ad1bb0, 2, MSG_NOSIGNAL) = 2 2) Then do a 'perf record' system wide for a while: # perf record -a ^C[ perf record: Woken up 68 times to write data ] [ perf record: Captured and wrote 19.427 MB perf.data (366891 samples) ] # 3) Check that we captured BPF and BTF info in the perf.data file: # perf report --header-only | grep 'b[pt]f' # event : name = cycles:ppp, , id = { 294789, 294790, 294791, 294792, 294793, 294794, 294795, 294796 }, size = 112, { sample_period, sample_freq } = 4000, sample_type = IP|TID|TIME|CPU|PERIOD, read_format = ID, disabled = 1, inherit = 1, mmap = 1, comm = 1, freq = 1, task = 1, precise_ip = 3, sample_id_all = 1, exclude_guest = 1, mmap2 = 1, comm_exec = 1, ksymbol = 1, bpf_event = 1 # bpf_prog_info of id 13 # bpf_prog_info of id 14 # bpf_prog_info of id 15 # bpf_prog_info of id 16 # bpf_prog_info of id 17 # bpf_prog_info of id 18 # bpf_prog_info of id 21 # bpf_prog_info of id 22 # bpf_prog_info of id 41 # bpf_prog_info of id 42 # btf info of id 2 # 4) Check which programs got recorded: # perf report | grep bpf_prog | head 0.16% exe bpf_prog_819967866022f1e1_sys_enter [k] bpf_prog_819967866022f1e1_sys_enter 0.14% exe bpf_prog_c1bd85c092d6e4aa_sys_exit [k] bpf_prog_c1bd85c092d6e4aa_sys_exit 0.08% fuse-overlayfs bpf_prog_819967866022f1e1_sys_enter [k] bpf_prog_819967866022f1e1_sys_enter 0.07% fuse-overlayfs bpf_prog_c1bd85c092d6e4aa_sys_exit [k] bpf_prog_c1bd85c092d6e4aa_sys_exit 0.01% clang-4.0 bpf_prog_c1bd85c092d6e4aa_sys_exit [k] bpf_prog_c1bd85c092d6e4aa_sys_exit 0.01% clang-4.0 bpf_prog_819967866022f1e1_sys_enter [k] bpf_prog_819967866022f1e1_sys_enter 0.00% clang bpf_prog_c1bd85c092d6e4aa_sys_exit [k] bpf_prog_c1bd85c092d6e4aa_sys_exit 0.00% runc bpf_prog_819967866022f1e1_sys_enter [k] bpf_prog_819967866022f1e1_sys_enter 0.00% clang bpf_prog_819967866022f1e1_sys_enter [k] bpf_prog_819967866022f1e1_sys_enter 0.00% sh bpf_prog_c1bd85c092d6e4aa_sys_exit [k] bpf_prog_c1bd85c092d6e4aa_sys_exit # This was with the default --sort order for 'perf report', which is: --sort comm,dso,symbol If we just look for the symbol, for instance: # perf report --sort symbol | grep bpf_prog | head 0.26% [k] bpf_prog_819967866022f1e1_sys_enter - - 0.24% [k] bpf_prog_c1bd85c092d6e4aa_sys_exit - - # or the DSO: # perf report --sort dso | grep bpf_prog | head 0.26% bpf_prog_819967866022f1e1_sys_enter 0.24% bpf_prog_c1bd85c092d6e4aa_sys_exit # We'll see the two BPF programs that augmented_raw_syscalls.o puts in place, one attached to the raw_syscalls:sys_enter and another to the raw_syscalls:sys_exit tracepoints, as expected. Now we can finally do, from the command line, annotation for one of those two symbols, with the original BPF program source coude intermixed with the disassembled JITed code: # perf annotate --stdio2 bpf_prog_819967866022f1e1_sys_enter Samples: 950 of event 'cycles:ppp', 4000 Hz, Event count (approx.): 553756947, [percent: local period] bpf_prog_819967866022f1e1_sys_enter() bpf_prog_819967866022f1e1_sys_enter Percent int sys_enter(struct syscall_enter_args *args) 53.41 push %rbp 0.63 mov %rsp,%rbp 0.31 sub $0x170,%rsp 1.93 sub $0x28,%rbp 7.02 mov %rbx,0x0(%rbp) 3.20 mov %r13,0x8(%rbp) 1.07 mov %r14,0x10(%rbp) 0.61 mov %r15,0x18(%rbp) 0.11 xor %eax,%eax 1.29 mov %rax,0x20(%rbp) 0.11 mov %rdi,%rbx return bpf_get_current_pid_tgid(); 2.02 → callq *ffffffffda6776d9 2.76 mov %eax,-0x148(%rbp) mov %rbp,%rsi int sys_enter(struct syscall_enter_args *args) add $0xfffffffffffffeb8,%rsi return bpf_map_lookup_elem(pids, &pid) != NULL; movabs $0xffff975ac2607800,%rdi 1.26 → callq *ffffffffda6789e9 cmp $0x0,%rax 2.43 → je 0 add $0x38,%rax 0.21 xor %r13d,%r13d if (pid_filter__has(&pids_filtered, getpid())) 0.81 cmp $0x0,%rax → jne 0 mov %rbp,%rdi probe_read(&augmented_args.args, sizeof(augmented_args.args), args); 2.22 add $0xfffffffffffffeb8,%rdi 0.11 mov $0x40,%esi 0.32 mov %rbx,%rdx 2.74 → callq *ffffffffda658409 syscall = bpf_map_lookup_elem(&syscalls, &augmented_args.args.syscall_nr); 0.22 mov %rbp,%rsi 1.69 add $0xfffffffffffffec0,%rsi syscall = bpf_map_lookup_elem(&syscalls, &augmented_args.args.syscall_nr); movabs $0xffff975bfcd36000,%rdi add $0xd0,%rdi 0.21 mov 0x0(%rsi),%eax 0.93 cmp $0x200,%rax → jae 0 0.10 shl $0x3,%rax 0.11 add %rdi,%rax 0.11 → jmp 0 xor %eax,%eax if (syscall == NULL || !syscall->enabled) 1.07 cmp $0x0,%rax → je 0 if (syscall == NULL || !syscall->enabled) 6.57 movzbq 0x0(%rax),%rdi if (syscall == NULL || !syscall->enabled) cmp $0x0,%rdi 0.95 → je 0 mov $0x40,%r8d switch (augmented_args.args.syscall_nr) { mov -0x140(%rbp),%rdi switch (augmented_args.args.syscall_nr) { cmp $0x2,%rdi → je 0 cmp $0x101,%rdi → je 0 cmp $0x15,%rdi → jne 0 case SYS_OPEN: filename_arg = (const void *)args->args[0]; mov 0x10(%rbx),%rdx → jmp 0 case SYS_OPENAT: filename_arg = (const void *)args->args[1]; mov 0x18(%rbx),%rdx if (filename_arg != NULL) { cmp $0x0,%rdx → je 0 xor %edi,%edi augmented_args.filename.reserved = 0; mov %edi,-0x104(%rbp) augmented_args.filename.size = probe_read_str(&augmented_args.filename.value, mov %rbp,%rdi add $0xffffffffffffff00,%rdi augmented_args.filename.size = probe_read_str(&augmented_args.filename.value, mov $0x100,%esi → callq *ffffffffda658499 mov $0x148,%r8d augmented_args.filename.size = probe_read_str(&augmented_args.filename.value, mov %eax,-0x108(%rbp) augmented_args.filename.size = probe_read_str(&augmented_args.filename.value, mov %rax,%rdi shl $0x20,%rdi shr $0x20,%rdi if (augmented_args.filename.size < sizeof(augmented_args.filename.value)) { cmp $0xff,%rdi → ja 0 len -= sizeof(augmented_args.filename.value) - augmented_args.filename.size; add $0x48,%rax len &= sizeof(augmented_args.filename.value) - 1; and $0xff,%rax mov %rax,%r8 mov %rbp,%rcx return perf_event_output(args, &__augmented_syscalls__, BPF_F_CURRENT_CPU, &augmented_args, len); add $0xfffffffffffffeb8,%rcx mov %rbx,%rdi movabs $0xffff975fbd72d800,%rsi mov $0xffffffff,%edx → callq *ffffffffda658ad9 mov %rax,%r13 } mov %r13,%rax 0.72 mov 0x0(%rbp),%rbx mov 0x8(%rbp),%r13 1.16 mov 0x10(%rbp),%r14 0.10 mov 0x18(%rbp),%r15 0.42 add $0x28,%rbp 0.54 leaveq 0.54 ← retq # Please see 'man perf-config' to see how to control what should be seen, via ~/.perfconfig [annotate] section, for instance, one can suppress the source code and see just the disassembly, etc. Alternatively, use the TUI bu just using 'perf annotate', press '/bpf_prog' to see the bpf symbols, press enter and do the interactive annotation, which allows for dumping to a file after selecting the the various output tunables, for instance, the above without source code intermixed, plus showing all the instruction offsets: # perf annotate bpf_prog_819967866022f1e1_sys_enter Then press: 's' to hide the source code + 'O' twice to show all instruction offsets, then 'P' to print to the bpf_prog_819967866022f1e1_sys_enter.annotation file, which will have: # cat bpf_prog_819967866022f1e1_sys_enter.annotation bpf_prog_819967866022f1e1_sys_enter() bpf_prog_819967866022f1e1_sys_enter Event: cycles:ppp 53.41 0: push %rbp 0.63 1: mov %rsp,%rbp 0.31 4: sub $0x170,%rsp 1.93 b: sub $0x28,%rbp 7.02 f: mov %rbx,0x0(%rbp) 3.20 13: mov %r13,0x8(%rbp) 1.07 17: mov %r14,0x10(%rbp) 0.61 1b: mov %r15,0x18(%rbp) 0.11 1f: xor %eax,%eax 1.29 21: mov %rax,0x20(%rbp) 0.11 25: mov %rdi,%rbx 2.02 28: → callq *ffffffffda6776d9 2.76 2d: mov %eax,-0x148(%rbp) 33: mov %rbp,%rsi 36: add $0xfffffffffffffeb8,%rsi 3d: movabs $0xffff975ac2607800,%rdi 1.26 47: → callq *ffffffffda6789e9 4c: cmp $0x0,%rax 2.43 50: → je 0 52: add $0x38,%rax 0.21 56: xor %r13d,%r13d 0.81 59: cmp $0x0,%rax 5d: → jne 0 63: mov %rbp,%rdi 2.22 66: add $0xfffffffffffffeb8,%rdi 0.11 6d: mov $0x40,%esi 0.32 72: mov %rbx,%rdx 2.74 75: → callq *ffffffffda658409 0.22 7a: mov %rbp,%rsi 1.69 7d: add $0xfffffffffffffec0,%rsi 84: movabs $0xffff975bfcd36000,%rdi 8e: add $0xd0,%rdi 0.21 95: mov 0x0(%rsi),%eax 0.93 98: cmp $0x200,%rax 9f: → jae 0 0.10 a1: shl $0x3,%rax 0.11 a5: add %rdi,%rax 0.11 a8: → jmp 0 aa: xor %eax,%eax 1.07 ac: cmp $0x0,%rax b0: → je 0 6.57 b6: movzbq 0x0(%rax),%rdi bb: cmp $0x0,%rdi 0.95 bf: → je 0 c5: mov $0x40,%r8d cb: mov -0x140(%rbp),%rdi d2: cmp $0x2,%rdi d6: → je 0 d8: cmp $0x101,%rdi df: → je 0 e1: cmp $0x15,%rdi e5: → jne 0 e7: mov 0x10(%rbx),%rdx eb: → jmp 0 ed: mov 0x18(%rbx),%rdx f1: cmp $0x0,%rdx f5: → je 0 f7: xor %edi,%edi f9: mov %edi,-0x104(%rbp) ff: mov %rbp,%rdi 102: add $0xffffffffffffff00,%rdi 109: mov $0x100,%esi 10e: → callq *ffffffffda658499 113: mov $0x148,%r8d 119: mov %eax,-0x108(%rbp) 11f: mov %rax,%rdi 122: shl $0x20,%rdi 126: shr $0x20,%rdi 12a: cmp $0xff,%rdi 131: → ja 0 133: add $0x48,%rax 137: and $0xff,%rax 13d: mov %rax,%r8 140: mov %rbp,%rcx 143: add $0xfffffffffffffeb8,%rcx 14a: mov %rbx,%rdi 14d: movabs $0xffff975fbd72d800,%rsi 157: mov $0xffffffff,%edx 15c: → callq *ffffffffda658ad9 161: mov %rax,%r13 164: mov %r13,%rax 0.72 167: mov 0x0(%rbp),%rbx 16b: mov 0x8(%rbp),%r13 1.16 16f: mov 0x10(%rbp),%r14 0.10 173: mov 0x18(%rbp),%r15 0.42 177: add $0x28,%rbp 0.54 17b: leaveq 0.54 17c: ← retq Another cool way to test all this is to symple use 'perf top' look for those symbols, go there and press enter, annotate it live :-) Signed-off-by: Song Liu <songliubraving@fb.com> Reviewed-by: Jiri Olsa <jolsa@kernel.org> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Alexei Starovoitov <ast@kernel.org> Cc: Daniel Borkmann <daniel@iogearbox.net> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Stanislav Fomichev <sdf@google.com> Link: http://lkml.kernel.org/r/20190312053051.2690567-13-songliubraving@fb.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2019-03-12 08:30:48 +03:00
fclose(s);
bfd_close(bfdf);
return ret;
}
#else // defined(HAVE_LIBBFD_SUPPORT) && defined(HAVE_LIBBPF_SUPPORT)
static int symbol__disassemble_bpf(struct symbol *sym __maybe_unused,
struct annotate_args *args __maybe_unused)
{
return SYMBOL_ANNOTATE_ERRNO__NO_LIBOPCODES_FOR_BPF;
}
#endif // defined(HAVE_LIBBFD_SUPPORT) && defined(HAVE_LIBBPF_SUPPORT)
static int
symbol__disassemble_bpf_image(struct symbol *sym,
struct annotate_args *args)
{
struct annotation *notes = symbol__annotation(sym);
struct disasm_line *dl;
args->offset = -1;
args->line = strdup("to be implemented");
args->line_nr = 0;
args->fileloc = NULL;
dl = disasm_line__new(args);
if (dl)
annotation_line__add(&dl->al, &notes->src->source);
free(args->line);
return 0;
}
/*
* Possibly create a new version of line with tabs expanded. Returns the
* existing or new line, storage is updated if a new line is allocated. If
* allocation fails then NULL is returned.
*/
static char *expand_tabs(char *line, char **storage, size_t *storage_len)
{
size_t i, src, dst, len, new_storage_len, num_tabs;
char *new_line;
size_t line_len = strlen(line);
for (num_tabs = 0, i = 0; i < line_len; i++)
if (line[i] == '\t')
num_tabs++;
if (num_tabs == 0)
return line;
/*
* Space for the line and '\0', less the leading and trailing
* spaces. Each tab may introduce 7 additional spaces.
*/
new_storage_len = line_len + 1 + (num_tabs * 7);
new_line = malloc(new_storage_len);
if (new_line == NULL) {
pr_err("Failure allocating memory for tab expansion\n");
return NULL;
}
/*
* Copy regions starting at src and expand tabs. If there are two
* adjacent tabs then 'src == i', the memcpy is of size 0 and the spaces
* are inserted.
*/
for (i = 0, src = 0, dst = 0; i < line_len && num_tabs; i++) {
if (line[i] == '\t') {
len = i - src;
memcpy(&new_line[dst], &line[src], len);
dst += len;
new_line[dst++] = ' ';
while (dst % 8 != 0)
new_line[dst++] = ' ';
src = i + 1;
num_tabs--;
}
}
/* Expand the last region. */
len = line_len - src;
memcpy(&new_line[dst], &line[src], len);
dst += len;
new_line[dst] = '\0';
free(*storage);
*storage = new_line;
*storage_len = new_storage_len;
return new_line;
}
static int symbol__disassemble(struct symbol *sym, struct annotate_args *args)
{
struct annotation_options *opts = args->options;
struct map *map = args->ms.map;
struct dso *dso = map->dso;
perf annotate: Use asprintf when formatting objdump command line We were using a local buffer with an arbitrary size, that would have to get increased to avoid truncation as warned by gcc 8: util/annotate.c: In function 'symbol__disassemble': util/annotate.c:1488:4: error: '%s' directive output may be truncated writing up to 4095 bytes into a region of size between 3966 and 8086 [-Werror=format-truncation=] "%s %s%s --start-address=0x%016" PRIx64 ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ util/annotate.c:1498:20: symfs_filename, symfs_filename); ~~~~~~~~~~~~~~ util/annotate.c:1490:50: note: format string is defined here " -l -d %s %s -C \"%s\" 2>/dev/null|grep -v \"%s:\"|expand", ^~ In file included from /usr/include/stdio.h:861, from util/color.h:5, from util/sort.h:8, from util/annotate.c:14: /usr/include/bits/stdio2.h:67:10: note: '__builtin___snprintf_chk' output 116 or more bytes (assuming 8331) into a destination of size 8192 return __builtin___snprintf_chk (__s, __n, __USE_FORTIFY_LEVEL - 1, ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ __bos (__s), __fmt, __va_arg_pack ()); ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ So switch to asprintf, that will make sure enough space is available. Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: David Ahern <dsahern@gmail.com> Cc: Jin Yao <yao.jin@linux.intel.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Wang Nan <wangnan0@huawei.com> Link: https://lkml.kernel.org/n/tip-qagoy2dmbjpc9gdnaj0r3mml@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2018-03-14 16:34:11 +03:00
char *command;
FILE *file;
char symfs_filename[PATH_MAX];
struct kcore_extract kce;
bool delete_extract = false;
bool decomp = false;
int lineno = 0;
char *fileloc = NULL;
int nline;
char *line;
size_t line_len;
const char *objdump_argv[] = {
"/bin/sh",
"-c",
NULL, /* Will be the objdump command to run. */
"--",
NULL, /* Will be the symfs path. */
NULL,
};
struct child_process objdump_process;
int err = dso__disassemble_filename(dso, symfs_filename, sizeof(symfs_filename));
if (err)
return err;
pr_debug("%s: filename=%s, sym=%s, start=%#" PRIx64 ", end=%#" PRIx64 "\n", __func__,
symfs_filename, sym->name, map->unmap_ip(map, sym->start),
map->unmap_ip(map, sym->end));
pr_debug("annotating [%p] %30s : [%p] %30s\n",
dso, dso->long_name, sym, sym->name);
perf annotate: Enable annotation of BPF programs In symbol__disassemble(), DSO_BINARY_TYPE__BPF_PROG_INFO dso calls into a new function symbol__disassemble_bpf(), where annotation line information is filled based on the bpf_prog_info and btf data saved in given perf_env. symbol__disassemble_bpf() uses binutils's libopcodes to disassemble bpf programs. Committer testing: After fixing this: - u64 *addrs = (u64 *)(info_linear->info.jited_ksyms); + u64 *addrs = (u64 *)(uintptr_t)(info_linear->info.jited_ksyms); Detected when crossbuilding to a 32-bit arch. And making all this dependent on HAVE_LIBBFD_SUPPORT and HAVE_LIBBPF_SUPPORT: 1) Have a BPF program running, one that has BTF info, etc, I used the tools/perf/examples/bpf/augmented_raw_syscalls.c put in place by 'perf trace'. # grep -B1 augmented_raw ~/.perfconfig [trace] add_events = /home/acme/git/perf/tools/perf/examples/bpf/augmented_raw_syscalls.c # # perf trace -e *mmsg dnf/6245 sendmmsg(20, 0x7f5485a88030, 2, MSG_NOSIGNAL) = 2 NetworkManager/10055 sendmmsg(22<socket:[1056822]>, 0x7f8126ad1bb0, 2, MSG_NOSIGNAL) = 2 2) Then do a 'perf record' system wide for a while: # perf record -a ^C[ perf record: Woken up 68 times to write data ] [ perf record: Captured and wrote 19.427 MB perf.data (366891 samples) ] # 3) Check that we captured BPF and BTF info in the perf.data file: # perf report --header-only | grep 'b[pt]f' # event : name = cycles:ppp, , id = { 294789, 294790, 294791, 294792, 294793, 294794, 294795, 294796 }, size = 112, { sample_period, sample_freq } = 4000, sample_type = IP|TID|TIME|CPU|PERIOD, read_format = ID, disabled = 1, inherit = 1, mmap = 1, comm = 1, freq = 1, task = 1, precise_ip = 3, sample_id_all = 1, exclude_guest = 1, mmap2 = 1, comm_exec = 1, ksymbol = 1, bpf_event = 1 # bpf_prog_info of id 13 # bpf_prog_info of id 14 # bpf_prog_info of id 15 # bpf_prog_info of id 16 # bpf_prog_info of id 17 # bpf_prog_info of id 18 # bpf_prog_info of id 21 # bpf_prog_info of id 22 # bpf_prog_info of id 41 # bpf_prog_info of id 42 # btf info of id 2 # 4) Check which programs got recorded: # perf report | grep bpf_prog | head 0.16% exe bpf_prog_819967866022f1e1_sys_enter [k] bpf_prog_819967866022f1e1_sys_enter 0.14% exe bpf_prog_c1bd85c092d6e4aa_sys_exit [k] bpf_prog_c1bd85c092d6e4aa_sys_exit 0.08% fuse-overlayfs bpf_prog_819967866022f1e1_sys_enter [k] bpf_prog_819967866022f1e1_sys_enter 0.07% fuse-overlayfs bpf_prog_c1bd85c092d6e4aa_sys_exit [k] bpf_prog_c1bd85c092d6e4aa_sys_exit 0.01% clang-4.0 bpf_prog_c1bd85c092d6e4aa_sys_exit [k] bpf_prog_c1bd85c092d6e4aa_sys_exit 0.01% clang-4.0 bpf_prog_819967866022f1e1_sys_enter [k] bpf_prog_819967866022f1e1_sys_enter 0.00% clang bpf_prog_c1bd85c092d6e4aa_sys_exit [k] bpf_prog_c1bd85c092d6e4aa_sys_exit 0.00% runc bpf_prog_819967866022f1e1_sys_enter [k] bpf_prog_819967866022f1e1_sys_enter 0.00% clang bpf_prog_819967866022f1e1_sys_enter [k] bpf_prog_819967866022f1e1_sys_enter 0.00% sh bpf_prog_c1bd85c092d6e4aa_sys_exit [k] bpf_prog_c1bd85c092d6e4aa_sys_exit # This was with the default --sort order for 'perf report', which is: --sort comm,dso,symbol If we just look for the symbol, for instance: # perf report --sort symbol | grep bpf_prog | head 0.26% [k] bpf_prog_819967866022f1e1_sys_enter - - 0.24% [k] bpf_prog_c1bd85c092d6e4aa_sys_exit - - # or the DSO: # perf report --sort dso | grep bpf_prog | head 0.26% bpf_prog_819967866022f1e1_sys_enter 0.24% bpf_prog_c1bd85c092d6e4aa_sys_exit # We'll see the two BPF programs that augmented_raw_syscalls.o puts in place, one attached to the raw_syscalls:sys_enter and another to the raw_syscalls:sys_exit tracepoints, as expected. Now we can finally do, from the command line, annotation for one of those two symbols, with the original BPF program source coude intermixed with the disassembled JITed code: # perf annotate --stdio2 bpf_prog_819967866022f1e1_sys_enter Samples: 950 of event 'cycles:ppp', 4000 Hz, Event count (approx.): 553756947, [percent: local period] bpf_prog_819967866022f1e1_sys_enter() bpf_prog_819967866022f1e1_sys_enter Percent int sys_enter(struct syscall_enter_args *args) 53.41 push %rbp 0.63 mov %rsp,%rbp 0.31 sub $0x170,%rsp 1.93 sub $0x28,%rbp 7.02 mov %rbx,0x0(%rbp) 3.20 mov %r13,0x8(%rbp) 1.07 mov %r14,0x10(%rbp) 0.61 mov %r15,0x18(%rbp) 0.11 xor %eax,%eax 1.29 mov %rax,0x20(%rbp) 0.11 mov %rdi,%rbx return bpf_get_current_pid_tgid(); 2.02 → callq *ffffffffda6776d9 2.76 mov %eax,-0x148(%rbp) mov %rbp,%rsi int sys_enter(struct syscall_enter_args *args) add $0xfffffffffffffeb8,%rsi return bpf_map_lookup_elem(pids, &pid) != NULL; movabs $0xffff975ac2607800,%rdi 1.26 → callq *ffffffffda6789e9 cmp $0x0,%rax 2.43 → je 0 add $0x38,%rax 0.21 xor %r13d,%r13d if (pid_filter__has(&pids_filtered, getpid())) 0.81 cmp $0x0,%rax → jne 0 mov %rbp,%rdi probe_read(&augmented_args.args, sizeof(augmented_args.args), args); 2.22 add $0xfffffffffffffeb8,%rdi 0.11 mov $0x40,%esi 0.32 mov %rbx,%rdx 2.74 → callq *ffffffffda658409 syscall = bpf_map_lookup_elem(&syscalls, &augmented_args.args.syscall_nr); 0.22 mov %rbp,%rsi 1.69 add $0xfffffffffffffec0,%rsi syscall = bpf_map_lookup_elem(&syscalls, &augmented_args.args.syscall_nr); movabs $0xffff975bfcd36000,%rdi add $0xd0,%rdi 0.21 mov 0x0(%rsi),%eax 0.93 cmp $0x200,%rax → jae 0 0.10 shl $0x3,%rax 0.11 add %rdi,%rax 0.11 → jmp 0 xor %eax,%eax if (syscall == NULL || !syscall->enabled) 1.07 cmp $0x0,%rax → je 0 if (syscall == NULL || !syscall->enabled) 6.57 movzbq 0x0(%rax),%rdi if (syscall == NULL || !syscall->enabled) cmp $0x0,%rdi 0.95 → je 0 mov $0x40,%r8d switch (augmented_args.args.syscall_nr) { mov -0x140(%rbp),%rdi switch (augmented_args.args.syscall_nr) { cmp $0x2,%rdi → je 0 cmp $0x101,%rdi → je 0 cmp $0x15,%rdi → jne 0 case SYS_OPEN: filename_arg = (const void *)args->args[0]; mov 0x10(%rbx),%rdx → jmp 0 case SYS_OPENAT: filename_arg = (const void *)args->args[1]; mov 0x18(%rbx),%rdx if (filename_arg != NULL) { cmp $0x0,%rdx → je 0 xor %edi,%edi augmented_args.filename.reserved = 0; mov %edi,-0x104(%rbp) augmented_args.filename.size = probe_read_str(&augmented_args.filename.value, mov %rbp,%rdi add $0xffffffffffffff00,%rdi augmented_args.filename.size = probe_read_str(&augmented_args.filename.value, mov $0x100,%esi → callq *ffffffffda658499 mov $0x148,%r8d augmented_args.filename.size = probe_read_str(&augmented_args.filename.value, mov %eax,-0x108(%rbp) augmented_args.filename.size = probe_read_str(&augmented_args.filename.value, mov %rax,%rdi shl $0x20,%rdi shr $0x20,%rdi if (augmented_args.filename.size < sizeof(augmented_args.filename.value)) { cmp $0xff,%rdi → ja 0 len -= sizeof(augmented_args.filename.value) - augmented_args.filename.size; add $0x48,%rax len &= sizeof(augmented_args.filename.value) - 1; and $0xff,%rax mov %rax,%r8 mov %rbp,%rcx return perf_event_output(args, &__augmented_syscalls__, BPF_F_CURRENT_CPU, &augmented_args, len); add $0xfffffffffffffeb8,%rcx mov %rbx,%rdi movabs $0xffff975fbd72d800,%rsi mov $0xffffffff,%edx → callq *ffffffffda658ad9 mov %rax,%r13 } mov %r13,%rax 0.72 mov 0x0(%rbp),%rbx mov 0x8(%rbp),%r13 1.16 mov 0x10(%rbp),%r14 0.10 mov 0x18(%rbp),%r15 0.42 add $0x28,%rbp 0.54 leaveq 0.54 ← retq # Please see 'man perf-config' to see how to control what should be seen, via ~/.perfconfig [annotate] section, for instance, one can suppress the source code and see just the disassembly, etc. Alternatively, use the TUI bu just using 'perf annotate', press '/bpf_prog' to see the bpf symbols, press enter and do the interactive annotation, which allows for dumping to a file after selecting the the various output tunables, for instance, the above without source code intermixed, plus showing all the instruction offsets: # perf annotate bpf_prog_819967866022f1e1_sys_enter Then press: 's' to hide the source code + 'O' twice to show all instruction offsets, then 'P' to print to the bpf_prog_819967866022f1e1_sys_enter.annotation file, which will have: # cat bpf_prog_819967866022f1e1_sys_enter.annotation bpf_prog_819967866022f1e1_sys_enter() bpf_prog_819967866022f1e1_sys_enter Event: cycles:ppp 53.41 0: push %rbp 0.63 1: mov %rsp,%rbp 0.31 4: sub $0x170,%rsp 1.93 b: sub $0x28,%rbp 7.02 f: mov %rbx,0x0(%rbp) 3.20 13: mov %r13,0x8(%rbp) 1.07 17: mov %r14,0x10(%rbp) 0.61 1b: mov %r15,0x18(%rbp) 0.11 1f: xor %eax,%eax 1.29 21: mov %rax,0x20(%rbp) 0.11 25: mov %rdi,%rbx 2.02 28: → callq *ffffffffda6776d9 2.76 2d: mov %eax,-0x148(%rbp) 33: mov %rbp,%rsi 36: add $0xfffffffffffffeb8,%rsi 3d: movabs $0xffff975ac2607800,%rdi 1.26 47: → callq *ffffffffda6789e9 4c: cmp $0x0,%rax 2.43 50: → je 0 52: add $0x38,%rax 0.21 56: xor %r13d,%r13d 0.81 59: cmp $0x0,%rax 5d: → jne 0 63: mov %rbp,%rdi 2.22 66: add $0xfffffffffffffeb8,%rdi 0.11 6d: mov $0x40,%esi 0.32 72: mov %rbx,%rdx 2.74 75: → callq *ffffffffda658409 0.22 7a: mov %rbp,%rsi 1.69 7d: add $0xfffffffffffffec0,%rsi 84: movabs $0xffff975bfcd36000,%rdi 8e: add $0xd0,%rdi 0.21 95: mov 0x0(%rsi),%eax 0.93 98: cmp $0x200,%rax 9f: → jae 0 0.10 a1: shl $0x3,%rax 0.11 a5: add %rdi,%rax 0.11 a8: → jmp 0 aa: xor %eax,%eax 1.07 ac: cmp $0x0,%rax b0: → je 0 6.57 b6: movzbq 0x0(%rax),%rdi bb: cmp $0x0,%rdi 0.95 bf: → je 0 c5: mov $0x40,%r8d cb: mov -0x140(%rbp),%rdi d2: cmp $0x2,%rdi d6: → je 0 d8: cmp $0x101,%rdi df: → je 0 e1: cmp $0x15,%rdi e5: → jne 0 e7: mov 0x10(%rbx),%rdx eb: → jmp 0 ed: mov 0x18(%rbx),%rdx f1: cmp $0x0,%rdx f5: → je 0 f7: xor %edi,%edi f9: mov %edi,-0x104(%rbp) ff: mov %rbp,%rdi 102: add $0xffffffffffffff00,%rdi 109: mov $0x100,%esi 10e: → callq *ffffffffda658499 113: mov $0x148,%r8d 119: mov %eax,-0x108(%rbp) 11f: mov %rax,%rdi 122: shl $0x20,%rdi 126: shr $0x20,%rdi 12a: cmp $0xff,%rdi 131: → ja 0 133: add $0x48,%rax 137: and $0xff,%rax 13d: mov %rax,%r8 140: mov %rbp,%rcx 143: add $0xfffffffffffffeb8,%rcx 14a: mov %rbx,%rdi 14d: movabs $0xffff975fbd72d800,%rsi 157: mov $0xffffffff,%edx 15c: → callq *ffffffffda658ad9 161: mov %rax,%r13 164: mov %r13,%rax 0.72 167: mov 0x0(%rbp),%rbx 16b: mov 0x8(%rbp),%r13 1.16 16f: mov 0x10(%rbp),%r14 0.10 173: mov 0x18(%rbp),%r15 0.42 177: add $0x28,%rbp 0.54 17b: leaveq 0.54 17c: ← retq Another cool way to test all this is to symple use 'perf top' look for those symbols, go there and press enter, annotate it live :-) Signed-off-by: Song Liu <songliubraving@fb.com> Reviewed-by: Jiri Olsa <jolsa@kernel.org> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Alexei Starovoitov <ast@kernel.org> Cc: Daniel Borkmann <daniel@iogearbox.net> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Stanislav Fomichev <sdf@google.com> Link: http://lkml.kernel.org/r/20190312053051.2690567-13-songliubraving@fb.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2019-03-12 08:30:48 +03:00
if (dso->binary_type == DSO_BINARY_TYPE__BPF_PROG_INFO) {
return symbol__disassemble_bpf(sym, args);
} else if (dso->binary_type == DSO_BINARY_TYPE__BPF_IMAGE) {
return symbol__disassemble_bpf_image(sym, args);
perf annotate: Enable annotation of BPF programs In symbol__disassemble(), DSO_BINARY_TYPE__BPF_PROG_INFO dso calls into a new function symbol__disassemble_bpf(), where annotation line information is filled based on the bpf_prog_info and btf data saved in given perf_env. symbol__disassemble_bpf() uses binutils's libopcodes to disassemble bpf programs. Committer testing: After fixing this: - u64 *addrs = (u64 *)(info_linear->info.jited_ksyms); + u64 *addrs = (u64 *)(uintptr_t)(info_linear->info.jited_ksyms); Detected when crossbuilding to a 32-bit arch. And making all this dependent on HAVE_LIBBFD_SUPPORT and HAVE_LIBBPF_SUPPORT: 1) Have a BPF program running, one that has BTF info, etc, I used the tools/perf/examples/bpf/augmented_raw_syscalls.c put in place by 'perf trace'. # grep -B1 augmented_raw ~/.perfconfig [trace] add_events = /home/acme/git/perf/tools/perf/examples/bpf/augmented_raw_syscalls.c # # perf trace -e *mmsg dnf/6245 sendmmsg(20, 0x7f5485a88030, 2, MSG_NOSIGNAL) = 2 NetworkManager/10055 sendmmsg(22<socket:[1056822]>, 0x7f8126ad1bb0, 2, MSG_NOSIGNAL) = 2 2) Then do a 'perf record' system wide for a while: # perf record -a ^C[ perf record: Woken up 68 times to write data ] [ perf record: Captured and wrote 19.427 MB perf.data (366891 samples) ] # 3) Check that we captured BPF and BTF info in the perf.data file: # perf report --header-only | grep 'b[pt]f' # event : name = cycles:ppp, , id = { 294789, 294790, 294791, 294792, 294793, 294794, 294795, 294796 }, size = 112, { sample_period, sample_freq } = 4000, sample_type = IP|TID|TIME|CPU|PERIOD, read_format = ID, disabled = 1, inherit = 1, mmap = 1, comm = 1, freq = 1, task = 1, precise_ip = 3, sample_id_all = 1, exclude_guest = 1, mmap2 = 1, comm_exec = 1, ksymbol = 1, bpf_event = 1 # bpf_prog_info of id 13 # bpf_prog_info of id 14 # bpf_prog_info of id 15 # bpf_prog_info of id 16 # bpf_prog_info of id 17 # bpf_prog_info of id 18 # bpf_prog_info of id 21 # bpf_prog_info of id 22 # bpf_prog_info of id 41 # bpf_prog_info of id 42 # btf info of id 2 # 4) Check which programs got recorded: # perf report | grep bpf_prog | head 0.16% exe bpf_prog_819967866022f1e1_sys_enter [k] bpf_prog_819967866022f1e1_sys_enter 0.14% exe bpf_prog_c1bd85c092d6e4aa_sys_exit [k] bpf_prog_c1bd85c092d6e4aa_sys_exit 0.08% fuse-overlayfs bpf_prog_819967866022f1e1_sys_enter [k] bpf_prog_819967866022f1e1_sys_enter 0.07% fuse-overlayfs bpf_prog_c1bd85c092d6e4aa_sys_exit [k] bpf_prog_c1bd85c092d6e4aa_sys_exit 0.01% clang-4.0 bpf_prog_c1bd85c092d6e4aa_sys_exit [k] bpf_prog_c1bd85c092d6e4aa_sys_exit 0.01% clang-4.0 bpf_prog_819967866022f1e1_sys_enter [k] bpf_prog_819967866022f1e1_sys_enter 0.00% clang bpf_prog_c1bd85c092d6e4aa_sys_exit [k] bpf_prog_c1bd85c092d6e4aa_sys_exit 0.00% runc bpf_prog_819967866022f1e1_sys_enter [k] bpf_prog_819967866022f1e1_sys_enter 0.00% clang bpf_prog_819967866022f1e1_sys_enter [k] bpf_prog_819967866022f1e1_sys_enter 0.00% sh bpf_prog_c1bd85c092d6e4aa_sys_exit [k] bpf_prog_c1bd85c092d6e4aa_sys_exit # This was with the default --sort order for 'perf report', which is: --sort comm,dso,symbol If we just look for the symbol, for instance: # perf report --sort symbol | grep bpf_prog | head 0.26% [k] bpf_prog_819967866022f1e1_sys_enter - - 0.24% [k] bpf_prog_c1bd85c092d6e4aa_sys_exit - - # or the DSO: # perf report --sort dso | grep bpf_prog | head 0.26% bpf_prog_819967866022f1e1_sys_enter 0.24% bpf_prog_c1bd85c092d6e4aa_sys_exit # We'll see the two BPF programs that augmented_raw_syscalls.o puts in place, one attached to the raw_syscalls:sys_enter and another to the raw_syscalls:sys_exit tracepoints, as expected. Now we can finally do, from the command line, annotation for one of those two symbols, with the original BPF program source coude intermixed with the disassembled JITed code: # perf annotate --stdio2 bpf_prog_819967866022f1e1_sys_enter Samples: 950 of event 'cycles:ppp', 4000 Hz, Event count (approx.): 553756947, [percent: local period] bpf_prog_819967866022f1e1_sys_enter() bpf_prog_819967866022f1e1_sys_enter Percent int sys_enter(struct syscall_enter_args *args) 53.41 push %rbp 0.63 mov %rsp,%rbp 0.31 sub $0x170,%rsp 1.93 sub $0x28,%rbp 7.02 mov %rbx,0x0(%rbp) 3.20 mov %r13,0x8(%rbp) 1.07 mov %r14,0x10(%rbp) 0.61 mov %r15,0x18(%rbp) 0.11 xor %eax,%eax 1.29 mov %rax,0x20(%rbp) 0.11 mov %rdi,%rbx return bpf_get_current_pid_tgid(); 2.02 → callq *ffffffffda6776d9 2.76 mov %eax,-0x148(%rbp) mov %rbp,%rsi int sys_enter(struct syscall_enter_args *args) add $0xfffffffffffffeb8,%rsi return bpf_map_lookup_elem(pids, &pid) != NULL; movabs $0xffff975ac2607800,%rdi 1.26 → callq *ffffffffda6789e9 cmp $0x0,%rax 2.43 → je 0 add $0x38,%rax 0.21 xor %r13d,%r13d if (pid_filter__has(&pids_filtered, getpid())) 0.81 cmp $0x0,%rax → jne 0 mov %rbp,%rdi probe_read(&augmented_args.args, sizeof(augmented_args.args), args); 2.22 add $0xfffffffffffffeb8,%rdi 0.11 mov $0x40,%esi 0.32 mov %rbx,%rdx 2.74 → callq *ffffffffda658409 syscall = bpf_map_lookup_elem(&syscalls, &augmented_args.args.syscall_nr); 0.22 mov %rbp,%rsi 1.69 add $0xfffffffffffffec0,%rsi syscall = bpf_map_lookup_elem(&syscalls, &augmented_args.args.syscall_nr); movabs $0xffff975bfcd36000,%rdi add $0xd0,%rdi 0.21 mov 0x0(%rsi),%eax 0.93 cmp $0x200,%rax → jae 0 0.10 shl $0x3,%rax 0.11 add %rdi,%rax 0.11 → jmp 0 xor %eax,%eax if (syscall == NULL || !syscall->enabled) 1.07 cmp $0x0,%rax → je 0 if (syscall == NULL || !syscall->enabled) 6.57 movzbq 0x0(%rax),%rdi if (syscall == NULL || !syscall->enabled) cmp $0x0,%rdi 0.95 → je 0 mov $0x40,%r8d switch (augmented_args.args.syscall_nr) { mov -0x140(%rbp),%rdi switch (augmented_args.args.syscall_nr) { cmp $0x2,%rdi → je 0 cmp $0x101,%rdi → je 0 cmp $0x15,%rdi → jne 0 case SYS_OPEN: filename_arg = (const void *)args->args[0]; mov 0x10(%rbx),%rdx → jmp 0 case SYS_OPENAT: filename_arg = (const void *)args->args[1]; mov 0x18(%rbx),%rdx if (filename_arg != NULL) { cmp $0x0,%rdx → je 0 xor %edi,%edi augmented_args.filename.reserved = 0; mov %edi,-0x104(%rbp) augmented_args.filename.size = probe_read_str(&augmented_args.filename.value, mov %rbp,%rdi add $0xffffffffffffff00,%rdi augmented_args.filename.size = probe_read_str(&augmented_args.filename.value, mov $0x100,%esi → callq *ffffffffda658499 mov $0x148,%r8d augmented_args.filename.size = probe_read_str(&augmented_args.filename.value, mov %eax,-0x108(%rbp) augmented_args.filename.size = probe_read_str(&augmented_args.filename.value, mov %rax,%rdi shl $0x20,%rdi shr $0x20,%rdi if (augmented_args.filename.size < sizeof(augmented_args.filename.value)) { cmp $0xff,%rdi → ja 0 len -= sizeof(augmented_args.filename.value) - augmented_args.filename.size; add $0x48,%rax len &= sizeof(augmented_args.filename.value) - 1; and $0xff,%rax mov %rax,%r8 mov %rbp,%rcx return perf_event_output(args, &__augmented_syscalls__, BPF_F_CURRENT_CPU, &augmented_args, len); add $0xfffffffffffffeb8,%rcx mov %rbx,%rdi movabs $0xffff975fbd72d800,%rsi mov $0xffffffff,%edx → callq *ffffffffda658ad9 mov %rax,%r13 } mov %r13,%rax 0.72 mov 0x0(%rbp),%rbx mov 0x8(%rbp),%r13 1.16 mov 0x10(%rbp),%r14 0.10 mov 0x18(%rbp),%r15 0.42 add $0x28,%rbp 0.54 leaveq 0.54 ← retq # Please see 'man perf-config' to see how to control what should be seen, via ~/.perfconfig [annotate] section, for instance, one can suppress the source code and see just the disassembly, etc. Alternatively, use the TUI bu just using 'perf annotate', press '/bpf_prog' to see the bpf symbols, press enter and do the interactive annotation, which allows for dumping to a file after selecting the the various output tunables, for instance, the above without source code intermixed, plus showing all the instruction offsets: # perf annotate bpf_prog_819967866022f1e1_sys_enter Then press: 's' to hide the source code + 'O' twice to show all instruction offsets, then 'P' to print to the bpf_prog_819967866022f1e1_sys_enter.annotation file, which will have: # cat bpf_prog_819967866022f1e1_sys_enter.annotation bpf_prog_819967866022f1e1_sys_enter() bpf_prog_819967866022f1e1_sys_enter Event: cycles:ppp 53.41 0: push %rbp 0.63 1: mov %rsp,%rbp 0.31 4: sub $0x170,%rsp 1.93 b: sub $0x28,%rbp 7.02 f: mov %rbx,0x0(%rbp) 3.20 13: mov %r13,0x8(%rbp) 1.07 17: mov %r14,0x10(%rbp) 0.61 1b: mov %r15,0x18(%rbp) 0.11 1f: xor %eax,%eax 1.29 21: mov %rax,0x20(%rbp) 0.11 25: mov %rdi,%rbx 2.02 28: → callq *ffffffffda6776d9 2.76 2d: mov %eax,-0x148(%rbp) 33: mov %rbp,%rsi 36: add $0xfffffffffffffeb8,%rsi 3d: movabs $0xffff975ac2607800,%rdi 1.26 47: → callq *ffffffffda6789e9 4c: cmp $0x0,%rax 2.43 50: → je 0 52: add $0x38,%rax 0.21 56: xor %r13d,%r13d 0.81 59: cmp $0x0,%rax 5d: → jne 0 63: mov %rbp,%rdi 2.22 66: add $0xfffffffffffffeb8,%rdi 0.11 6d: mov $0x40,%esi 0.32 72: mov %rbx,%rdx 2.74 75: → callq *ffffffffda658409 0.22 7a: mov %rbp,%rsi 1.69 7d: add $0xfffffffffffffec0,%rsi 84: movabs $0xffff975bfcd36000,%rdi 8e: add $0xd0,%rdi 0.21 95: mov 0x0(%rsi),%eax 0.93 98: cmp $0x200,%rax 9f: → jae 0 0.10 a1: shl $0x3,%rax 0.11 a5: add %rdi,%rax 0.11 a8: → jmp 0 aa: xor %eax,%eax 1.07 ac: cmp $0x0,%rax b0: → je 0 6.57 b6: movzbq 0x0(%rax),%rdi bb: cmp $0x0,%rdi 0.95 bf: → je 0 c5: mov $0x40,%r8d cb: mov -0x140(%rbp),%rdi d2: cmp $0x2,%rdi d6: → je 0 d8: cmp $0x101,%rdi df: → je 0 e1: cmp $0x15,%rdi e5: → jne 0 e7: mov 0x10(%rbx),%rdx eb: → jmp 0 ed: mov 0x18(%rbx),%rdx f1: cmp $0x0,%rdx f5: → je 0 f7: xor %edi,%edi f9: mov %edi,-0x104(%rbp) ff: mov %rbp,%rdi 102: add $0xffffffffffffff00,%rdi 109: mov $0x100,%esi 10e: → callq *ffffffffda658499 113: mov $0x148,%r8d 119: mov %eax,-0x108(%rbp) 11f: mov %rax,%rdi 122: shl $0x20,%rdi 126: shr $0x20,%rdi 12a: cmp $0xff,%rdi 131: → ja 0 133: add $0x48,%rax 137: and $0xff,%rax 13d: mov %rax,%r8 140: mov %rbp,%rcx 143: add $0xfffffffffffffeb8,%rcx 14a: mov %rbx,%rdi 14d: movabs $0xffff975fbd72d800,%rsi 157: mov $0xffffffff,%edx 15c: → callq *ffffffffda658ad9 161: mov %rax,%r13 164: mov %r13,%rax 0.72 167: mov 0x0(%rbp),%rbx 16b: mov 0x8(%rbp),%r13 1.16 16f: mov 0x10(%rbp),%r14 0.10 173: mov 0x18(%rbp),%r15 0.42 177: add $0x28,%rbp 0.54 17b: leaveq 0.54 17c: ← retq Another cool way to test all this is to symple use 'perf top' look for those symbols, go there and press enter, annotate it live :-) Signed-off-by: Song Liu <songliubraving@fb.com> Reviewed-by: Jiri Olsa <jolsa@kernel.org> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Alexei Starovoitov <ast@kernel.org> Cc: Daniel Borkmann <daniel@iogearbox.net> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Stanislav Fomichev <sdf@google.com> Link: http://lkml.kernel.org/r/20190312053051.2690567-13-songliubraving@fb.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2019-03-12 08:30:48 +03:00
} else if (dso__is_kcore(dso)) {
kce.kcore_filename = symfs_filename;
kce.addr = map__rip_2objdump(map, sym->start);
kce.offs = sym->start;
kce.len = sym->end - sym->start;
if (!kcore_extract__create(&kce)) {
delete_extract = true;
strlcpy(symfs_filename, kce.extract_filename,
sizeof(symfs_filename));
}
} else if (dso__needs_decompress(dso)) {
char tmp[KMOD_DECOMP_LEN];
if (dso__decompress_kmodule_path(dso, symfs_filename,
tmp, sizeof(tmp)) < 0)
return -1;
decomp = true;
strcpy(symfs_filename, tmp);
}
perf annotate: Use asprintf when formatting objdump command line We were using a local buffer with an arbitrary size, that would have to get increased to avoid truncation as warned by gcc 8: util/annotate.c: In function 'symbol__disassemble': util/annotate.c:1488:4: error: '%s' directive output may be truncated writing up to 4095 bytes into a region of size between 3966 and 8086 [-Werror=format-truncation=] "%s %s%s --start-address=0x%016" PRIx64 ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ util/annotate.c:1498:20: symfs_filename, symfs_filename); ~~~~~~~~~~~~~~ util/annotate.c:1490:50: note: format string is defined here " -l -d %s %s -C \"%s\" 2>/dev/null|grep -v \"%s:\"|expand", ^~ In file included from /usr/include/stdio.h:861, from util/color.h:5, from util/sort.h:8, from util/annotate.c:14: /usr/include/bits/stdio2.h:67:10: note: '__builtin___snprintf_chk' output 116 or more bytes (assuming 8331) into a destination of size 8192 return __builtin___snprintf_chk (__s, __n, __USE_FORTIFY_LEVEL - 1, ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ __bos (__s), __fmt, __va_arg_pack ()); ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ So switch to asprintf, that will make sure enough space is available. Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: David Ahern <dsahern@gmail.com> Cc: Jin Yao <yao.jin@linux.intel.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Wang Nan <wangnan0@huawei.com> Link: https://lkml.kernel.org/n/tip-qagoy2dmbjpc9gdnaj0r3mml@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2018-03-14 16:34:11 +03:00
err = asprintf(&command,
"%s %s%s --start-address=0x%016" PRIx64
" --stop-address=0x%016" PRIx64
" -l -d %s %s %s %c%s%c %s%s -C \"$1\"",
opts->objdump_path ?: "objdump",
opts->disassembler_style ? "-M " : "",
opts->disassembler_style ?: "",
map__rip_2objdump(map, sym->start),
map__rip_2objdump(map, sym->end),
opts->show_asm_raw ? "" : "--no-show-raw-insn",
opts->annotate_src ? "-S" : "",
opts->prefix ? "--prefix " : "",
opts->prefix ? '"' : ' ',
opts->prefix ?: "",
opts->prefix ? '"' : ' ',
opts->prefix_strip ? "--prefix-strip=" : "",
opts->prefix_strip ?: "");
perf annotate: Use asprintf when formatting objdump command line We were using a local buffer with an arbitrary size, that would have to get increased to avoid truncation as warned by gcc 8: util/annotate.c: In function 'symbol__disassemble': util/annotate.c:1488:4: error: '%s' directive output may be truncated writing up to 4095 bytes into a region of size between 3966 and 8086 [-Werror=format-truncation=] "%s %s%s --start-address=0x%016" PRIx64 ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ util/annotate.c:1498:20: symfs_filename, symfs_filename); ~~~~~~~~~~~~~~ util/annotate.c:1490:50: note: format string is defined here " -l -d %s %s -C \"%s\" 2>/dev/null|grep -v \"%s:\"|expand", ^~ In file included from /usr/include/stdio.h:861, from util/color.h:5, from util/sort.h:8, from util/annotate.c:14: /usr/include/bits/stdio2.h:67:10: note: '__builtin___snprintf_chk' output 116 or more bytes (assuming 8331) into a destination of size 8192 return __builtin___snprintf_chk (__s, __n, __USE_FORTIFY_LEVEL - 1, ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ __bos (__s), __fmt, __va_arg_pack ()); ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ So switch to asprintf, that will make sure enough space is available. Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: David Ahern <dsahern@gmail.com> Cc: Jin Yao <yao.jin@linux.intel.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Wang Nan <wangnan0@huawei.com> Link: https://lkml.kernel.org/n/tip-qagoy2dmbjpc9gdnaj0r3mml@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2018-03-14 16:34:11 +03:00
if (err < 0) {
pr_err("Failure allocating memory for the command to run\n");
goto out_remove_tmp;
}
pr_debug("Executing: %s\n", command);
objdump_argv[2] = command;
objdump_argv[4] = symfs_filename;
/* Create a pipe to read from for stdout */
memset(&objdump_process, 0, sizeof(objdump_process));
objdump_process.argv = objdump_argv;
objdump_process.out = -1;
if (start_command(&objdump_process)) {
pr_err("Failure starting to run %s\n", command);
err = -1;
goto out_free_command;
}
file = fdopen(objdump_process.out, "r");
perf annotate: Inform the user about objdump failures in --stdio When the browser fails to annotate it is difficult for users to find out what went wrong. Add some errors for objdump failures that are displayed in the UI. Note it would be even better to handle these errors smarter, like falling back to the binary when the debug info is somehow corrupted. But for now just giving a better error is an improvement. Committer note: This works for --stdio, where errors just scroll by the screen: # perf annotate --stdio intel_idle Failure running objdump --start-address=0xffffffff81418290 --stop-address=0xffffffff814183ae -l -d --no-show-raw -S -C /root/.debug/.build-id/28/2777c262e6b3c0451375163c9a81c893218ab1 2>/dev/null|grep -v /root/.debug/.build-id/28/2777c262e6b3c0451375163c9a81c893218ab1|expand Percent | Source code & Disassembly of vmlinux for cycles:pp ------------------------------------------------------------------ And with that one can use that command line to try to find out more about what happened instead of getting a blank screen, an improvement. We need tho to improve this further to get it to work with other UIs, like --tui and --gtk, where it continues showing a blank screen, no messages, as the pr_err() used is enough just for --stdio. Signed-off-by: Andi Kleen <ak@linux.intel.com> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Link: http://lkml.kernel.org/r/1446779167-18949-1-git-send-email-andi@firstfloor.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-11-06 06:06:07 +03:00
if (!file) {
pr_err("Failure creating FILE stream for %s\n", command);
perf annotate: Inform the user about objdump failures in --stdio When the browser fails to annotate it is difficult for users to find out what went wrong. Add some errors for objdump failures that are displayed in the UI. Note it would be even better to handle these errors smarter, like falling back to the binary when the debug info is somehow corrupted. But for now just giving a better error is an improvement. Committer note: This works for --stdio, where errors just scroll by the screen: # perf annotate --stdio intel_idle Failure running objdump --start-address=0xffffffff81418290 --stop-address=0xffffffff814183ae -l -d --no-show-raw -S -C /root/.debug/.build-id/28/2777c262e6b3c0451375163c9a81c893218ab1 2>/dev/null|grep -v /root/.debug/.build-id/28/2777c262e6b3c0451375163c9a81c893218ab1|expand Percent | Source code & Disassembly of vmlinux for cycles:pp ------------------------------------------------------------------ And with that one can use that command line to try to find out more about what happened instead of getting a blank screen, an improvement. We need tho to improve this further to get it to work with other UIs, like --tui and --gtk, where it continues showing a blank screen, no messages, as the pr_err() used is enough just for --stdio. Signed-off-by: Andi Kleen <ak@linux.intel.com> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Link: http://lkml.kernel.org/r/1446779167-18949-1-git-send-email-andi@firstfloor.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-11-06 06:06:07 +03:00
/*
* If we were using debug info should retry with
* original binary.
*/
err = -1;
goto out_close_stdout;
perf annotate: Inform the user about objdump failures in --stdio When the browser fails to annotate it is difficult for users to find out what went wrong. Add some errors for objdump failures that are displayed in the UI. Note it would be even better to handle these errors smarter, like falling back to the binary when the debug info is somehow corrupted. But for now just giving a better error is an improvement. Committer note: This works for --stdio, where errors just scroll by the screen: # perf annotate --stdio intel_idle Failure running objdump --start-address=0xffffffff81418290 --stop-address=0xffffffff814183ae -l -d --no-show-raw -S -C /root/.debug/.build-id/28/2777c262e6b3c0451375163c9a81c893218ab1 2>/dev/null|grep -v /root/.debug/.build-id/28/2777c262e6b3c0451375163c9a81c893218ab1|expand Percent | Source code & Disassembly of vmlinux for cycles:pp ------------------------------------------------------------------ And with that one can use that command line to try to find out more about what happened instead of getting a blank screen, an improvement. We need tho to improve this further to get it to work with other UIs, like --tui and --gtk, where it continues showing a blank screen, no messages, as the pr_err() used is enough just for --stdio. Signed-off-by: Andi Kleen <ak@linux.intel.com> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Link: http://lkml.kernel.org/r/1446779167-18949-1-git-send-email-andi@firstfloor.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-11-06 06:06:07 +03:00
}
/* Storage for getline. */
line = NULL;
line_len = 0;
perf annotate: Inform the user about objdump failures in --stdio When the browser fails to annotate it is difficult for users to find out what went wrong. Add some errors for objdump failures that are displayed in the UI. Note it would be even better to handle these errors smarter, like falling back to the binary when the debug info is somehow corrupted. But for now just giving a better error is an improvement. Committer note: This works for --stdio, where errors just scroll by the screen: # perf annotate --stdio intel_idle Failure running objdump --start-address=0xffffffff81418290 --stop-address=0xffffffff814183ae -l -d --no-show-raw -S -C /root/.debug/.build-id/28/2777c262e6b3c0451375163c9a81c893218ab1 2>/dev/null|grep -v /root/.debug/.build-id/28/2777c262e6b3c0451375163c9a81c893218ab1|expand Percent | Source code & Disassembly of vmlinux for cycles:pp ------------------------------------------------------------------ And with that one can use that command line to try to find out more about what happened instead of getting a blank screen, an improvement. We need tho to improve this further to get it to work with other UIs, like --tui and --gtk, where it continues showing a blank screen, no messages, as the pr_err() used is enough just for --stdio. Signed-off-by: Andi Kleen <ak@linux.intel.com> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Link: http://lkml.kernel.org/r/1446779167-18949-1-git-send-email-andi@firstfloor.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-11-06 06:06:07 +03:00
nline = 0;
while (!feof(file)) {
const char *match;
char *expanded_line;
if (getline(&line, &line_len, file) < 0 || !line)
break;
/* Skip lines containing "filename:" */
match = strstr(line, symfs_filename);
if (match && match[strlen(symfs_filename)] == ':')
continue;
expanded_line = strim(line);
expanded_line = expand_tabs(expanded_line, &line, &line_len);
if (!expanded_line)
break;
/*
* The source code line number (lineno) needs to be kept in
* across calls to symbol__parse_objdump_line(), so that it
* can associate it with the instructions till the next one.
* See disasm_line__new() and struct disasm_line::line_nr.
*/
if (symbol__parse_objdump_line(sym, args, expanded_line,
&lineno, &fileloc) < 0)
break;
perf annotate: Inform the user about objdump failures in --stdio When the browser fails to annotate it is difficult for users to find out what went wrong. Add some errors for objdump failures that are displayed in the UI. Note it would be even better to handle these errors smarter, like falling back to the binary when the debug info is somehow corrupted. But for now just giving a better error is an improvement. Committer note: This works for --stdio, where errors just scroll by the screen: # perf annotate --stdio intel_idle Failure running objdump --start-address=0xffffffff81418290 --stop-address=0xffffffff814183ae -l -d --no-show-raw -S -C /root/.debug/.build-id/28/2777c262e6b3c0451375163c9a81c893218ab1 2>/dev/null|grep -v /root/.debug/.build-id/28/2777c262e6b3c0451375163c9a81c893218ab1|expand Percent | Source code & Disassembly of vmlinux for cycles:pp ------------------------------------------------------------------ And with that one can use that command line to try to find out more about what happened instead of getting a blank screen, an improvement. We need tho to improve this further to get it to work with other UIs, like --tui and --gtk, where it continues showing a blank screen, no messages, as the pr_err() used is enough just for --stdio. Signed-off-by: Andi Kleen <ak@linux.intel.com> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Link: http://lkml.kernel.org/r/1446779167-18949-1-git-send-email-andi@firstfloor.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-11-06 06:06:07 +03:00
nline++;
}
free(line);
perf annotate: Inform the user about objdump failures in --stdio When the browser fails to annotate it is difficult for users to find out what went wrong. Add some errors for objdump failures that are displayed in the UI. Note it would be even better to handle these errors smarter, like falling back to the binary when the debug info is somehow corrupted. But for now just giving a better error is an improvement. Committer note: This works for --stdio, where errors just scroll by the screen: # perf annotate --stdio intel_idle Failure running objdump --start-address=0xffffffff81418290 --stop-address=0xffffffff814183ae -l -d --no-show-raw -S -C /root/.debug/.build-id/28/2777c262e6b3c0451375163c9a81c893218ab1 2>/dev/null|grep -v /root/.debug/.build-id/28/2777c262e6b3c0451375163c9a81c893218ab1|expand Percent | Source code & Disassembly of vmlinux for cycles:pp ------------------------------------------------------------------ And with that one can use that command line to try to find out more about what happened instead of getting a blank screen, an improvement. We need tho to improve this further to get it to work with other UIs, like --tui and --gtk, where it continues showing a blank screen, no messages, as the pr_err() used is enough just for --stdio. Signed-off-by: Andi Kleen <ak@linux.intel.com> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Link: http://lkml.kernel.org/r/1446779167-18949-1-git-send-email-andi@firstfloor.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-11-06 06:06:07 +03:00
err = finish_command(&objdump_process);
if (err)
pr_err("Error running %s\n", command);
if (nline == 0) {
err = -1;
perf annotate: Inform the user about objdump failures in --stdio When the browser fails to annotate it is difficult for users to find out what went wrong. Add some errors for objdump failures that are displayed in the UI. Note it would be even better to handle these errors smarter, like falling back to the binary when the debug info is somehow corrupted. But for now just giving a better error is an improvement. Committer note: This works for --stdio, where errors just scroll by the screen: # perf annotate --stdio intel_idle Failure running objdump --start-address=0xffffffff81418290 --stop-address=0xffffffff814183ae -l -d --no-show-raw -S -C /root/.debug/.build-id/28/2777c262e6b3c0451375163c9a81c893218ab1 2>/dev/null|grep -v /root/.debug/.build-id/28/2777c262e6b3c0451375163c9a81c893218ab1|expand Percent | Source code & Disassembly of vmlinux for cycles:pp ------------------------------------------------------------------ And with that one can use that command line to try to find out more about what happened instead of getting a blank screen, an improvement. We need tho to improve this further to get it to work with other UIs, like --tui and --gtk, where it continues showing a blank screen, no messages, as the pr_err() used is enough just for --stdio. Signed-off-by: Andi Kleen <ak@linux.intel.com> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Link: http://lkml.kernel.org/r/1446779167-18949-1-git-send-email-andi@firstfloor.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2015-11-06 06:06:07 +03:00
pr_err("No output from %s\n", command);
}
/*
* kallsyms does not have symbol sizes so there may a nop at the end.
* Remove it.
*/
if (dso__is_kcore(dso))
delete_last_nop(sym);
fclose(file);
out_close_stdout:
close(objdump_process.out);
perf annotate: Use asprintf when formatting objdump command line We were using a local buffer with an arbitrary size, that would have to get increased to avoid truncation as warned by gcc 8: util/annotate.c: In function 'symbol__disassemble': util/annotate.c:1488:4: error: '%s' directive output may be truncated writing up to 4095 bytes into a region of size between 3966 and 8086 [-Werror=format-truncation=] "%s %s%s --start-address=0x%016" PRIx64 ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ util/annotate.c:1498:20: symfs_filename, symfs_filename); ~~~~~~~~~~~~~~ util/annotate.c:1490:50: note: format string is defined here " -l -d %s %s -C \"%s\" 2>/dev/null|grep -v \"%s:\"|expand", ^~ In file included from /usr/include/stdio.h:861, from util/color.h:5, from util/sort.h:8, from util/annotate.c:14: /usr/include/bits/stdio2.h:67:10: note: '__builtin___snprintf_chk' output 116 or more bytes (assuming 8331) into a destination of size 8192 return __builtin___snprintf_chk (__s, __n, __USE_FORTIFY_LEVEL - 1, ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ __bos (__s), __fmt, __va_arg_pack ()); ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ So switch to asprintf, that will make sure enough space is available. Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: David Ahern <dsahern@gmail.com> Cc: Jin Yao <yao.jin@linux.intel.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Wang Nan <wangnan0@huawei.com> Link: https://lkml.kernel.org/n/tip-qagoy2dmbjpc9gdnaj0r3mml@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2018-03-14 16:34:11 +03:00
out_free_command:
free(command);
out_remove_tmp:
if (decomp)
unlink(symfs_filename);
if (delete_extract)
kcore_extract__delete(&kce);
return err;
}
static void calc_percent(struct sym_hist *sym_hist,
struct hists *hists,
struct annotation_data *data,
s64 offset, s64 end)
{
unsigned int hits = 0;
u64 period = 0;
while (offset < end) {
hits += sym_hist->addr[offset].nr_samples;
period += sym_hist->addr[offset].period;
++offset;
}
if (sym_hist->nr_samples) {
data->he.period = period;
data->he.nr_samples = hits;
data->percent[PERCENT_HITS_LOCAL] = 100.0 * hits / sym_hist->nr_samples;
}
if (hists->stats.nr_non_filtered_samples)
data->percent[PERCENT_HITS_GLOBAL] = 100.0 * hits / hists->stats.nr_non_filtered_samples;
if (sym_hist->period)
data->percent[PERCENT_PERIOD_LOCAL] = 100.0 * period / sym_hist->period;
if (hists->stats.total_period)
data->percent[PERCENT_PERIOD_GLOBAL] = 100.0 * period / hists->stats.total_period;
}
static void annotation__calc_percent(struct annotation *notes,
struct evsel *leader, s64 len)
{
struct annotation_line *al, *next;
struct evsel *evsel;
list_for_each_entry(al, &notes->src->source, node) {
s64 end;
int i = 0;
if (al->offset == -1)
continue;
next = annotation_line__next(al, &notes->src->source);
end = next ? next->offset : len;
for_each_group_evsel(evsel, leader) {
struct hists *hists = evsel__hists(evsel);
struct annotation_data *data;
struct sym_hist *sym_hist;
BUG_ON(i >= al->data_nr);
sym_hist = annotation__histogram(notes, evsel->core.idx);
data = &al->data[i++];
calc_percent(sym_hist, hists, data, al->offset, end);
}
}
}
void symbol__calc_percent(struct symbol *sym, struct evsel *evsel)
{
struct annotation *notes = symbol__annotation(sym);
annotation__calc_percent(notes, evsel, symbol__size(sym));
}
int symbol__annotate(struct map_symbol *ms, struct evsel *evsel,
struct annotation_options *options, struct arch **parch)
{
struct symbol *sym = ms->sym;
perf annotate: Fix getting source line failure The output of "perf annotate -l --stdio xxx" changed since commit 425859ff0de33 ("perf annotate: No need to calculate notes->start twice") removed notes->start assignment in symbol__calc_lines(). It will get failed in find_address_in_section() from symbol__tty_annotate() subroutine as the a2l->addr is wrong. So the annotate summary doesn't report the line number of source code correctly. Before fix: liwei@euler:~/main_code/hulk_work/hulk/tools/perf$ cat common_while_1.c void hotspot_1(void) { volatile int i; for (i = 0; i < 0x10000000; i++); for (i = 0; i < 0x10000000; i++); for (i = 0; i < 0x10000000; i++); } int main(void) { hotspot_1(); return 0; } liwei@euler:~/main_code/hulk_work/hulk/tools/perf$ gcc common_while_1.c -g -o common_while_1 liwei@euler:~/main_code/hulk_work/hulk/tools/perf$ sudo ./perf record ./common_while_1 [ perf record: Woken up 2 times to write data ] [ perf record: Captured and wrote 0.488 MB perf.data (12498 samples) ] liwei@euler:~/main_code/hulk_work/hulk/tools/perf$ sudo ./perf annotate -l -s hotspot_1 --stdio Sorted summary for file /home/liwei/main_code/hulk_work/hulk/tools/perf/common_while_1 ---------------------------------------------- 19.30 common_while_1[32] 19.03 common_while_1[4e] 19.01 common_while_1[16] 5.04 common_while_1[13] 4.99 common_while_1[4b] 4.78 common_while_1[2c] 4.77 common_while_1[10] 4.66 common_while_1[2f] 4.59 common_while_1[51] 4.59 common_while_1[35] 4.52 common_while_1[19] 4.20 common_while_1[56] 0.51 common_while_1[48] Percent | Source code & Disassembly of common_while_1 for cycles:ppp (12480 samples, percent: local period) ----------------------------------------------------------------------------------------------------------------- : : : : Disassembly of section .text: : : 00000000000005fa <hotspot_1>: : hotspot_1(): : void hotspot_1(void) : { 0.00 : 5fa: push %rbp 0.00 : 5fb: mov %rsp,%rbp : volatile int i; : : for (i = 0; i < 0x10000000; i++); 0.00 : 5fe: movl $0x0,-0x4(%rbp) 0.00 : 605: jmp 610 <hotspot_1+0x16> 0.00 : 607: mov -0x4(%rbp),%eax common_while_1[10] 4.77 : 60a: add $0x1,%eax common_while_1[13] 5.04 : 60d: mov %eax,-0x4(%rbp) common_while_1[16] 19.01 : 610: mov -0x4(%rbp),%eax common_while_1[19] 4.52 : 613: cmp $0xfffffff,%eax 0.00 : 618: jle 607 <hotspot_1+0xd> : for (i = 0; i < 0x10000000; i++); ... After fix: liwei@euler:~/main_code/hulk_work/hulk/tools/perf$ sudo ./perf record ./common_while_1 [ perf record: Woken up 2 times to write data ] [ perf record: Captured and wrote 0.488 MB perf.data (12500 samples) ] liwei@euler:~/main_code/hulk_work/hulk/tools/perf$ sudo ./perf annotate -l -s hotspot_1 --stdio Sorted summary for file /home/liwei/main_code/hulk_work/hulk/tools/perf/common_while_1 ---------------------------------------------- 33.34 common_while_1.c:5 33.34 common_while_1.c:6 33.32 common_while_1.c:7 Percent | Source code & Disassembly of common_while_1 for cycles:ppp (12482 samples, percent: local period) ----------------------------------------------------------------------------------------------------------------- : : : : Disassembly of section .text: : : 00000000000005fa <hotspot_1>: : hotspot_1(): : void hotspot_1(void) : { 0.00 : 5fa: push %rbp 0.00 : 5fb: mov %rsp,%rbp : volatile int i; : : for (i = 0; i < 0x10000000; i++); 0.00 : 5fe: movl $0x0,-0x4(%rbp) 0.00 : 605: jmp 610 <hotspot_1+0x16> 0.00 : 607: mov -0x4(%rbp),%eax common_while_1.c:5 4.70 : 60a: add $0x1,%eax 4.89 : 60d: mov %eax,-0x4(%rbp) common_while_1.c:5 19.03 : 610: mov -0x4(%rbp),%eax common_while_1.c:5 4.72 : 613: cmp $0xfffffff,%eax 0.00 : 618: jle 607 <hotspot_1+0xd> : for (i = 0; i < 0x10000000; i++); 0.00 : 61a: movl $0x0,-0x4(%rbp) 0.00 : 621: jmp 62c <hotspot_1+0x32> 0.00 : 623: mov -0x4(%rbp),%eax common_while_1.c:6 4.54 : 626: add $0x1,%eax 4.73 : 629: mov %eax,-0x4(%rbp) common_while_1.c:6 19.54 : 62c: mov -0x4(%rbp),%eax common_while_1.c:6 4.54 : 62f: cmp $0xfffffff,%eax ... Signed-off-by: Wei Li <liwei391@huawei.com> Acked-by: Jiri Olsa <jolsa@kernel.org> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Jin Yao <yao.jin@linux.intel.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Fixes: 425859ff0de33 ("perf annotate: No need to calculate notes->start twice") Link: http://lkml.kernel.org/r/20190221095716.39529-1-liwei391@huawei.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2019-02-21 12:57:16 +03:00
struct annotation *notes = symbol__annotation(sym);
struct annotate_args args = {
.evsel = evsel,
.options = options,
};
struct perf_env *env = evsel__env(evsel);
const char *arch_name = perf_env__arch(env);
struct arch *arch;
int err;
if (!arch_name)
return errno;
args.arch = arch = arch__find(arch_name);
if (arch == NULL) {
pr_err("%s: unsupported arch %s\n", __func__, arch_name);
return ENOTSUP;
}
if (parch)
*parch = arch;
if (arch->init) {
err = arch->init(arch, env ? env->cpuid : NULL);
if (err) {
pr_err("%s: failed to initialize %s arch priv area\n", __func__, arch->name);
return err;
}
}
args.ms = *ms;
notes->start = map__rip_2objdump(ms->map, sym->start);
return symbol__disassemble(sym, &args);
}
static void insert_source_line(struct rb_root *root, struct annotation_line *al,
struct annotation_options *opts)
{
struct annotation_line *iter;
struct rb_node **p = &root->rb_node;
struct rb_node *parent = NULL;
int i, ret;
while (*p != NULL) {
parent = *p;
iter = rb_entry(parent, struct annotation_line, rb_node);
ret = strcmp(iter->path, al->path);
if (ret == 0) {
for (i = 0; i < al->data_nr; i++) {
iter->data[i].percent_sum += annotation_data__percent(&al->data[i],
opts->percent_type);
}
return;
}
if (ret < 0)
p = &(*p)->rb_left;
else
p = &(*p)->rb_right;
}
for (i = 0; i < al->data_nr; i++) {
al->data[i].percent_sum = annotation_data__percent(&al->data[i],
opts->percent_type);
}
rb_link_node(&al->rb_node, parent, p);
rb_insert_color(&al->rb_node, root);
}
static int cmp_source_line(struct annotation_line *a, struct annotation_line *b)
{
int i;
for (i = 0; i < a->data_nr; i++) {
if (a->data[i].percent_sum == b->data[i].percent_sum)
continue;
return a->data[i].percent_sum > b->data[i].percent_sum;
}
return 0;
}
static void __resort_source_line(struct rb_root *root, struct annotation_line *al)
{
struct annotation_line *iter;
struct rb_node **p = &root->rb_node;
struct rb_node *parent = NULL;
while (*p != NULL) {
parent = *p;
iter = rb_entry(parent, struct annotation_line, rb_node);
if (cmp_source_line(al, iter))
p = &(*p)->rb_left;
else
p = &(*p)->rb_right;
}
rb_link_node(&al->rb_node, parent, p);
rb_insert_color(&al->rb_node, root);
}
static void resort_source_line(struct rb_root *dest_root, struct rb_root *src_root)
{
struct annotation_line *al;
struct rb_node *node;
node = rb_first(src_root);
while (node) {
struct rb_node *next;
al = rb_entry(node, struct annotation_line, rb_node);
next = rb_next(node);
rb_erase(node, src_root);
__resort_source_line(dest_root, al);
node = next;
}
}
static void print_summary(struct rb_root *root, const char *filename)
{
struct annotation_line *al;
struct rb_node *node;
printf("\nSorted summary for file %s\n", filename);
printf("----------------------------------------------\n\n");
if (RB_EMPTY_ROOT(root)) {
printf(" Nothing higher than %1.1f%%\n", MIN_GREEN);
return;
}
node = rb_first(root);
while (node) {
double percent, percent_max = 0.0;
const char *color;
char *path;
int i;
al = rb_entry(node, struct annotation_line, rb_node);
for (i = 0; i < al->data_nr; i++) {
percent = al->data[i].percent_sum;
color = get_percent_color(percent);
color_fprintf(stdout, color, " %7.2f", percent);
if (percent > percent_max)
percent_max = percent;
}
path = al->path;
color = get_percent_color(percent_max);
color_fprintf(stdout, color, " %s\n", path);
node = rb_next(node);
}
}
static void symbol__annotate_hits(struct symbol *sym, struct evsel *evsel)
{
struct annotation *notes = symbol__annotation(sym);
struct sym_hist *h = annotation__histogram(notes, evsel->core.idx);
u64 len = symbol__size(sym), offset;
for (offset = 0; offset < len; ++offset)
if (h->addr[offset].nr_samples != 0)
printf("%*" PRIx64 ": %" PRIu64 "\n", BITS_PER_LONG / 2,
sym->start + offset, h->addr[offset].nr_samples);
printf("%*s: %" PRIu64 "\n", BITS_PER_LONG / 2, "h->nr_samples", h->nr_samples);
}
perf annotate: Align source and offset lines Align source with offset lines, which are more advanced, because of the address column. Before: : static void *worker_thread(void *__tdata) : { 0.00 : 48a971: push %rbp 0.00 : 48a972: mov %rsp,%rbp 0.00 : 48a975: sub $0x30,%rsp 0.00 : 48a979: mov %rdi,-0x28(%rbp) 0.00 : 48a97d: mov %fs:0x28,%rax 0.00 : 48a986: mov %rax,-0x8(%rbp) 0.00 : 48a98a: xor %eax,%eax : struct thread_data *td = __tdata; 0.00 : 48a98c: mov -0x28(%rbp),%rax 0.00 : 48a990: mov %rax,-0x10(%rbp) : int m = 0, i; 0.00 : 48a994: movl $0x0,-0x1c(%rbp) : int ret; : : for (i = 0; i < loops; i++) { 0.00 : 48a99b: movl $0x0,-0x18(%rbp) After: : static void *worker_thread(void *__tdata) : { 0.00 : 48a971: push %rbp 0.00 : 48a972: mov %rsp,%rbp 0.00 : 48a975: sub $0x30,%rsp 0.00 : 48a979: mov %rdi,-0x28(%rbp) 0.00 : 48a97d: mov %fs:0x28,%rax 0.00 : 48a986: mov %rax,-0x8(%rbp) 0.00 : 48a98a: xor %eax,%eax : struct thread_data *td = __tdata; 0.00 : 48a98c: mov -0x28(%rbp),%rax 0.00 : 48a990: mov %rax,-0x10(%rbp) : int m = 0, i; 0.00 : 48a994: movl $0x0,-0x1c(%rbp) : int ret; : : for (i = 0; i < loops; i++) { 0.00 : 48a99b: movl $0x0,-0x18(%rbp) It makes bigger different when displaying script sources, where the comment lines looks oddly shifted from the lines which actually hold code. I'll send script support separately. Committer note: Do not use a fixed column width for the addresses, as kernel ones se more than 10 columns, look at the last offset and get the right width. Signed-off-by: Jiri Olsa <jolsa@kernel.org> Cc: Andi Kleen <andi@firstfloor.org> Cc: David Ahern <dsahern@gmail.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Link: http://lkml.kernel.org/r/20171011150158.11895-36-jolsa@kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2017-10-11 18:01:58 +03:00
static int annotated_source__addr_fmt_width(struct list_head *lines, u64 start)
{
char bf[32];
struct annotation_line *line;
list_for_each_entry_reverse(line, lines, node) {
if (line->offset != -1)
return scnprintf(bf, sizeof(bf), "%" PRIx64, start + line->offset);
}
return 0;
}
int symbol__annotate_printf(struct map_symbol *ms, struct evsel *evsel,
struct annotation_options *opts)
{
struct map *map = ms->map;
struct symbol *sym = ms->sym;
struct dso *dso = map->dso;
char *filename;
const char *d_filename;
const char *evsel_name = evsel__name(evsel);
struct annotation *notes = symbol__annotation(sym);
struct sym_hist *h = annotation__histogram(notes, evsel->core.idx);
struct annotation_line *pos, *queue = NULL;
u64 start = map__rip_2objdump(map, sym->start);
perf annotate: Align source and offset lines Align source with offset lines, which are more advanced, because of the address column. Before: : static void *worker_thread(void *__tdata) : { 0.00 : 48a971: push %rbp 0.00 : 48a972: mov %rsp,%rbp 0.00 : 48a975: sub $0x30,%rsp 0.00 : 48a979: mov %rdi,-0x28(%rbp) 0.00 : 48a97d: mov %fs:0x28,%rax 0.00 : 48a986: mov %rax,-0x8(%rbp) 0.00 : 48a98a: xor %eax,%eax : struct thread_data *td = __tdata; 0.00 : 48a98c: mov -0x28(%rbp),%rax 0.00 : 48a990: mov %rax,-0x10(%rbp) : int m = 0, i; 0.00 : 48a994: movl $0x0,-0x1c(%rbp) : int ret; : : for (i = 0; i < loops; i++) { 0.00 : 48a99b: movl $0x0,-0x18(%rbp) After: : static void *worker_thread(void *__tdata) : { 0.00 : 48a971: push %rbp 0.00 : 48a972: mov %rsp,%rbp 0.00 : 48a975: sub $0x30,%rsp 0.00 : 48a979: mov %rdi,-0x28(%rbp) 0.00 : 48a97d: mov %fs:0x28,%rax 0.00 : 48a986: mov %rax,-0x8(%rbp) 0.00 : 48a98a: xor %eax,%eax : struct thread_data *td = __tdata; 0.00 : 48a98c: mov -0x28(%rbp),%rax 0.00 : 48a990: mov %rax,-0x10(%rbp) : int m = 0, i; 0.00 : 48a994: movl $0x0,-0x1c(%rbp) : int ret; : : for (i = 0; i < loops; i++) { 0.00 : 48a99b: movl $0x0,-0x18(%rbp) It makes bigger different when displaying script sources, where the comment lines looks oddly shifted from the lines which actually hold code. I'll send script support separately. Committer note: Do not use a fixed column width for the addresses, as kernel ones se more than 10 columns, look at the last offset and get the right width. Signed-off-by: Jiri Olsa <jolsa@kernel.org> Cc: Andi Kleen <andi@firstfloor.org> Cc: David Ahern <dsahern@gmail.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Link: http://lkml.kernel.org/r/20171011150158.11895-36-jolsa@kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2017-10-11 18:01:58 +03:00
int printed = 2, queue_len = 0, addr_fmt_width;
int more = 0;
bool context = opts->context;
u64 len;
int width = symbol_conf.show_total_period ? 12 : 8;
int graph_dotted_len;
char buf[512];
filename = strdup(dso->long_name);
if (!filename)
return -ENOMEM;
if (opts->full_path)
d_filename = filename;
else
d_filename = basename(filename);
len = symbol__size(sym);
perf annotate: Add basic support to event group view Add --group option to enable event grouping. When enabled, all the group members information will be shown with the leader so skip non-leader events. It only supports --stdio output currently. Later patches will extend additional features. $ perf annotate --group --stdio ... Percent | Source code & Disassembly of libpthread-2.15.so -------------------------------------------------------------------------------- : : : : Disassembly of section .text: : : 000000387dc0aa50 <__pthread_mutex_unlock_usercnt>: 8.08 2.40 5.29 : 387dc0aa50: mov %rdi,%rdx 0.00 0.00 0.00 : 387dc0aa53: mov 0x10(%rdi),%edi 0.00 0.00 0.00 : 387dc0aa56: mov %edi,%eax 0.00 0.80 0.00 : 387dc0aa58: and $0x7f,%eax 3.03 2.40 3.53 : 387dc0aa5b: test $0x7c,%dil 0.00 0.00 0.00 : 387dc0aa5f: jne 387dc0aaa9 <__pthread_mutex_unlock_use 0.00 0.00 0.00 : 387dc0aa61: test %eax,%eax 0.00 0.00 0.00 : 387dc0aa63: jne 387dc0aa85 <__pthread_mutex_unlock_use 0.00 0.00 0.00 : 387dc0aa65: and $0x80,%edi 0.00 0.00 0.00 : 387dc0aa6b: test %esi,%esi 3.03 5.60 7.06 : 387dc0aa6d: movl $0x0,0x8(%rdx) 0.00 0.00 0.59 : 387dc0aa74: je 387dc0aa7a <__pthread_mutex_unlock_use 0.00 0.00 0.00 : 387dc0aa76: subl $0x1,0xc(%rdx) 2.02 5.60 1.18 : 387dc0aa7a: mov %edi,%esi 0.00 0.00 0.00 : 387dc0aa7c: lock decl (%rdx) 83.84 83.20 82.35 : 387dc0aa7f: jne 387dc0aada <_L_unlock_586> 0.00 0.00 0.00 : 387dc0aa81: nop 0.00 0.00 0.00 : 387dc0aa82: xor %eax,%eax 0.00 0.00 0.00 : 387dc0aa84: retq ... Signed-off-by: Namhyung Kim <namhyung@kernel.org> Cc: Andi Kleen <andi@firstfloor.org> Cc: Ingo Molnar <mingo@kernel.org> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Paul Mackerras <paulus@samba.org> Cc: Pekka Enberg <penberg@kernel.org> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Link: http://lkml.kernel.org/r/1362462812-30885-6-git-send-email-namhyung@kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2013-03-05 09:53:25 +04:00
if (evsel__is_group_event(evsel)) {
width *= evsel->core.nr_members;
evsel__group_desc(evsel, buf, sizeof(buf));
evsel_name = buf;
}
graph_dotted_len = printf(" %-*.*s| Source code & Disassembly of %s for %s (%" PRIu64 " samples, "
"percent: %s)\n",
width, width, symbol_conf.show_total_period ? "Period" :
symbol_conf.show_nr_samples ? "Samples" : "Percent",
d_filename, evsel_name, h->nr_samples,
percent_type_str(opts->percent_type));
printf("%-*.*s----\n",
graph_dotted_len, graph_dotted_len, graph_dotted_line);
if (verbose > 0)
symbol__annotate_hits(sym, evsel);
perf annotate: Align source and offset lines Align source with offset lines, which are more advanced, because of the address column. Before: : static void *worker_thread(void *__tdata) : { 0.00 : 48a971: push %rbp 0.00 : 48a972: mov %rsp,%rbp 0.00 : 48a975: sub $0x30,%rsp 0.00 : 48a979: mov %rdi,-0x28(%rbp) 0.00 : 48a97d: mov %fs:0x28,%rax 0.00 : 48a986: mov %rax,-0x8(%rbp) 0.00 : 48a98a: xor %eax,%eax : struct thread_data *td = __tdata; 0.00 : 48a98c: mov -0x28(%rbp),%rax 0.00 : 48a990: mov %rax,-0x10(%rbp) : int m = 0, i; 0.00 : 48a994: movl $0x0,-0x1c(%rbp) : int ret; : : for (i = 0; i < loops; i++) { 0.00 : 48a99b: movl $0x0,-0x18(%rbp) After: : static void *worker_thread(void *__tdata) : { 0.00 : 48a971: push %rbp 0.00 : 48a972: mov %rsp,%rbp 0.00 : 48a975: sub $0x30,%rsp 0.00 : 48a979: mov %rdi,-0x28(%rbp) 0.00 : 48a97d: mov %fs:0x28,%rax 0.00 : 48a986: mov %rax,-0x8(%rbp) 0.00 : 48a98a: xor %eax,%eax : struct thread_data *td = __tdata; 0.00 : 48a98c: mov -0x28(%rbp),%rax 0.00 : 48a990: mov %rax,-0x10(%rbp) : int m = 0, i; 0.00 : 48a994: movl $0x0,-0x1c(%rbp) : int ret; : : for (i = 0; i < loops; i++) { 0.00 : 48a99b: movl $0x0,-0x18(%rbp) It makes bigger different when displaying script sources, where the comment lines looks oddly shifted from the lines which actually hold code. I'll send script support separately. Committer note: Do not use a fixed column width for the addresses, as kernel ones se more than 10 columns, look at the last offset and get the right width. Signed-off-by: Jiri Olsa <jolsa@kernel.org> Cc: Andi Kleen <andi@firstfloor.org> Cc: David Ahern <dsahern@gmail.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Link: http://lkml.kernel.org/r/20171011150158.11895-36-jolsa@kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2017-10-11 18:01:58 +03:00
addr_fmt_width = annotated_source__addr_fmt_width(&notes->src->source, start);
list_for_each_entry(pos, &notes->src->source, node) {
int err;
if (context && queue == NULL) {
queue = pos;
queue_len = 0;
}
err = annotation_line__print(pos, sym, start, evsel, len,
opts->min_pcnt, printed, opts->max_lines,
queue, addr_fmt_width, opts->percent_type);
switch (err) {
case 0:
++printed;
if (context) {
printed += queue_len;
queue = NULL;
queue_len = 0;
}
break;
case 1:
/* filtered by max_lines */
++more;
break;
case -1:
default:
/*
* Filtered by min_pcnt or non IP lines when
* context != 0
*/
if (!context)
break;
if (queue_len == context)
queue = list_entry(queue->node.next, typeof(*queue), node);
else
++queue_len;
break;
}
}
free(filename);
return more;
}
static void FILE__set_percent_color(void *fp __maybe_unused,
double percent __maybe_unused,
bool current __maybe_unused)
{
}
static int FILE__set_jumps_percent_color(void *fp __maybe_unused,
int nr __maybe_unused, bool current __maybe_unused)
{
return 0;
}
static int FILE__set_color(void *fp __maybe_unused, int color __maybe_unused)
{
return 0;
}
static void FILE__printf(void *fp, const char *fmt, ...)
{
va_list args;
va_start(args, fmt);
vfprintf(fp, fmt, args);
va_end(args);
}
static void FILE__write_graph(void *fp, int graph)
{
const char *s;
switch (graph) {
case DARROW_CHAR: s = ""; break;
case UARROW_CHAR: s = ""; break;
case LARROW_CHAR: s = ""; break;
case RARROW_CHAR: s = ""; break;
default: s = "?"; break;
}
fputs(s, fp);
}
static int symbol__annotate_fprintf2(struct symbol *sym, FILE *fp,
struct annotation_options *opts)
{
struct annotation *notes = symbol__annotation(sym);
struct annotation_write_ops wops = {
.first_line = true,
.obj = fp,
.set_color = FILE__set_color,
.set_percent_color = FILE__set_percent_color,
.set_jumps_percent_color = FILE__set_jumps_percent_color,
.printf = FILE__printf,
.write_graph = FILE__write_graph,
};
struct annotation_line *al;
list_for_each_entry(al, &notes->src->source, node) {
if (annotation_line__filter(al, notes))
continue;
annotation_line__write(al, notes, &wops, opts);
fputc('\n', fp);
wops.first_line = false;
}
return 0;
}
int map_symbol__annotation_dump(struct map_symbol *ms, struct evsel *evsel,
struct annotation_options *opts)
{
const char *ev_name = evsel__name(evsel);
char buf[1024];
char *filename;
int err = -1;
FILE *fp;
if (asprintf(&filename, "%s.annotation", ms->sym->name) < 0)
return -1;
fp = fopen(filename, "w");
if (fp == NULL)
goto out_free_filename;
if (evsel__is_group_event(evsel)) {
evsel__group_desc(evsel, buf, sizeof(buf));
ev_name = buf;
}
fprintf(fp, "%s() %s\nEvent: %s\n\n",
ms->sym->name, ms->map->dso->long_name, ev_name);
symbol__annotate_fprintf2(ms->sym, fp, opts);
fclose(fp);
err = 0;
out_free_filename:
free(filename);
return err;
}
void symbol__annotate_zero_histogram(struct symbol *sym, int evidx)
{
struct annotation *notes = symbol__annotation(sym);
struct sym_hist *h = annotation__histogram(notes, evidx);
memset(h, 0, notes->src->sizeof_sym_hist);
}
void symbol__annotate_decay_histogram(struct symbol *sym, int evidx)
{
struct annotation *notes = symbol__annotation(sym);
struct sym_hist *h = annotation__histogram(notes, evidx);
int len = symbol__size(sym), offset;
h->nr_samples = 0;
for (offset = 0; offset < len; ++offset) {
h->addr[offset].nr_samples = h->addr[offset].nr_samples * 7 / 8;
h->nr_samples += h->addr[offset].nr_samples;
}
}
void annotated_source__purge(struct annotated_source *as)
{
struct annotation_line *al, *n;
list_for_each_entry_safe(al, n, &as->source, node) {
list_del_init(&al->node);
disasm_line__free(disasm_line(al));
}
}
static size_t disasm_line__fprintf(struct disasm_line *dl, FILE *fp)
{
size_t printed;
if (dl->al.offset == -1)
return fprintf(fp, "%s\n", dl->al.line);
printed = fprintf(fp, "%#" PRIx64 " %s", dl->al.offset, dl->ins.name);
if (dl->ops.raw[0] != '\0') {
printed += fprintf(fp, "%.*s %s\n", 6 - (int)printed, " ",
dl->ops.raw);
}
return printed + fprintf(fp, "\n");
}
size_t disasm__fprintf(struct list_head *head, FILE *fp)
{
struct disasm_line *pos;
size_t printed = 0;
list_for_each_entry(pos, head, al.node)
printed += disasm_line__fprintf(pos, fp);
return printed;
}
bool disasm_line__is_valid_local_jump(struct disasm_line *dl, struct symbol *sym)
{
if (!dl || !dl->ins.ops || !ins__is_jump(&dl->ins) ||
!disasm_line__has_local_offset(dl) || dl->ops.target.offset < 0 ||
dl->ops.target.offset >= (s64)symbol__size(sym))
return false;
return true;
}
void annotation__mark_jump_targets(struct annotation *notes, struct symbol *sym)
{
u64 offset, size = symbol__size(sym);
/* PLT symbols contain external offsets */
if (strstr(sym->name, "@plt"))
return;
for (offset = 0; offset < size; ++offset) {
struct annotation_line *al = notes->offsets[offset];
struct disasm_line *dl;
dl = disasm_line(al);
if (!disasm_line__is_valid_local_jump(dl, sym))
continue;
al = notes->offsets[dl->ops.target.offset];
/*
* FIXME: Oops, no jump target? Buggy disassembler? Or do we
* have to adjust to the previous offset?
*/
if (al == NULL)
continue;
if (++al->jump_sources > notes->max_jump_sources)
notes->max_jump_sources = al->jump_sources;
}
}
void annotation__set_offsets(struct annotation *notes, s64 size)
{
struct annotation_line *al;
notes->max_line_len = 0;
notes->nr_entries = 0;
notes->nr_asm_entries = 0;
list_for_each_entry(al, &notes->src->source, node) {
size_t line_len = strlen(al->line);
if (notes->max_line_len < line_len)
notes->max_line_len = line_len;
al->idx = notes->nr_entries++;
if (al->offset != -1) {
al->idx_asm = notes->nr_asm_entries++;
/*
* FIXME: short term bandaid to cope with assembly
* routines that comes with labels in the same column
* as the address in objdump, sigh.
*
* E.g. copy_user_generic_unrolled
*/
if (al->offset < size)
notes->offsets[al->offset] = al;
} else
al->idx_asm = -1;
}
}
static inline int width_jumps(int n)
{
if (n >= 100)
return 5;
if (n / 10)
return 2;
return 1;
}
perf annotate: Calculate the max instruction name, align column to that We were hardcoding '6' as the max instruction name, and we have lots that are longer than that, see the diff from two 'P' printed TUI annotations for a libc function that uses instructions with long names, such as 'vpmovmskb' with its 9 chars: --- __strcmp_avx2.annotation.before 2019-03-06 16:31:39.368020425 -0300 +++ __strcmp_avx2.annotation 2019-03-06 16:32:12.079450508 -0300 @@ -2,284 +2,284 @@ Event: cycles:ppp Percent endbr64 - 0.10 mov %edi,%eax + 0.10 mov %edi,%eax - xor %edx,%edx + xor %edx,%edx - 3.54 vpxor %ymm7,%ymm7,%ymm7 + 3.54 vpxor %ymm7,%ymm7,%ymm7 - or %esi,%eax + or %esi,%eax - and $0xfff,%eax + and $0xfff,%eax - cmp $0xf80,%eax + cmp $0xf80,%eax - ↓ jg 370 + ↓ jg 370 - 27.07 vmovdqu (%rdi),%ymm1 + 27.07 vmovdqu (%rdi),%ymm1 - 7.97 vpcmpeqb (%rsi),%ymm1,%ymm0 + 7.97 vpcmpeqb (%rsi),%ymm1,%ymm0 - 2.15 vpminub %ymm1,%ymm0,%ymm0 + 2.15 vpminub %ymm1,%ymm0,%ymm0 - 4.09 vpcmpeqb %ymm7,%ymm0,%ymm0 + 4.09 vpcmpeqb %ymm7,%ymm0,%ymm0 - 0.43 vpmovmskb %ymm0,%ecx + 0.43 vpmovmskb %ymm0,%ecx - 1.53 test %ecx,%ecx + 1.53 test %ecx,%ecx - ↓ je b0 + ↓ je b0 - 5.26 tzcnt %ecx,%edx + 5.26 tzcnt %ecx,%edx - 18.40 movzbl (%rdi,%rdx,1),%eax + 18.40 movzbl (%rdi,%rdx,1),%eax - 7.09 movzbl (%rsi,%rdx,1),%edx + 7.09 movzbl (%rsi,%rdx,1),%edx - 3.34 sub %edx,%eax + 3.34 sub %edx,%eax 2.37 vzeroupper ← retq nop - 50: tzcnt %ecx,%edx + 50: tzcnt %ecx,%edx - movzbl 0x20(%rdi,%rdx,1),%eax + movzbl 0x20(%rdi,%rdx,1),%eax - movzbl 0x20(%rsi,%rdx,1),%edx + movzbl 0x20(%rsi,%rdx,1),%edx - sub %edx,%eax + sub %edx,%eax vzeroupper ← retq - data16 nopw %cs:0x0(%rax,%rax,1) + data16 nopw %cs:0x0(%rax,%rax,1) Reported-by: Travis Downs <travis.downs@gmail.com> LPU-Reference: CAOBGo4z1KfmWeOm6Et0cnX5Z6DWsG2PQbAvRn1MhVPJmXHrc5g@mail.gmail.com Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Link: https://lkml.kernel.org/n/tip-89wsdd9h9g6bvq52sgp6d0u4@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2019-03-06 22:40:15 +03:00
static int annotation__max_ins_name(struct annotation *notes)
{
int max_name = 0, len;
struct annotation_line *al;
list_for_each_entry(al, &notes->src->source, node) {
if (al->offset == -1)
continue;
len = strlen(disasm_line(al)->ins.name);
if (max_name < len)
max_name = len;
}
return max_name;
}
void annotation__init_column_widths(struct annotation *notes, struct symbol *sym)
{
notes->widths.addr = notes->widths.target =
notes->widths.min_addr = hex_width(symbol__size(sym));
notes->widths.max_addr = hex_width(sym->end);
notes->widths.jumps = width_jumps(notes->max_jump_sources);
perf annotate: Calculate the max instruction name, align column to that We were hardcoding '6' as the max instruction name, and we have lots that are longer than that, see the diff from two 'P' printed TUI annotations for a libc function that uses instructions with long names, such as 'vpmovmskb' with its 9 chars: --- __strcmp_avx2.annotation.before 2019-03-06 16:31:39.368020425 -0300 +++ __strcmp_avx2.annotation 2019-03-06 16:32:12.079450508 -0300 @@ -2,284 +2,284 @@ Event: cycles:ppp Percent endbr64 - 0.10 mov %edi,%eax + 0.10 mov %edi,%eax - xor %edx,%edx + xor %edx,%edx - 3.54 vpxor %ymm7,%ymm7,%ymm7 + 3.54 vpxor %ymm7,%ymm7,%ymm7 - or %esi,%eax + or %esi,%eax - and $0xfff,%eax + and $0xfff,%eax - cmp $0xf80,%eax + cmp $0xf80,%eax - ↓ jg 370 + ↓ jg 370 - 27.07 vmovdqu (%rdi),%ymm1 + 27.07 vmovdqu (%rdi),%ymm1 - 7.97 vpcmpeqb (%rsi),%ymm1,%ymm0 + 7.97 vpcmpeqb (%rsi),%ymm1,%ymm0 - 2.15 vpminub %ymm1,%ymm0,%ymm0 + 2.15 vpminub %ymm1,%ymm0,%ymm0 - 4.09 vpcmpeqb %ymm7,%ymm0,%ymm0 + 4.09 vpcmpeqb %ymm7,%ymm0,%ymm0 - 0.43 vpmovmskb %ymm0,%ecx + 0.43 vpmovmskb %ymm0,%ecx - 1.53 test %ecx,%ecx + 1.53 test %ecx,%ecx - ↓ je b0 + ↓ je b0 - 5.26 tzcnt %ecx,%edx + 5.26 tzcnt %ecx,%edx - 18.40 movzbl (%rdi,%rdx,1),%eax + 18.40 movzbl (%rdi,%rdx,1),%eax - 7.09 movzbl (%rsi,%rdx,1),%edx + 7.09 movzbl (%rsi,%rdx,1),%edx - 3.34 sub %edx,%eax + 3.34 sub %edx,%eax 2.37 vzeroupper ← retq nop - 50: tzcnt %ecx,%edx + 50: tzcnt %ecx,%edx - movzbl 0x20(%rdi,%rdx,1),%eax + movzbl 0x20(%rdi,%rdx,1),%eax - movzbl 0x20(%rsi,%rdx,1),%edx + movzbl 0x20(%rsi,%rdx,1),%edx - sub %edx,%eax + sub %edx,%eax vzeroupper ← retq - data16 nopw %cs:0x0(%rax,%rax,1) + data16 nopw %cs:0x0(%rax,%rax,1) Reported-by: Travis Downs <travis.downs@gmail.com> LPU-Reference: CAOBGo4z1KfmWeOm6Et0cnX5Z6DWsG2PQbAvRn1MhVPJmXHrc5g@mail.gmail.com Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Link: https://lkml.kernel.org/n/tip-89wsdd9h9g6bvq52sgp6d0u4@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2019-03-06 22:40:15 +03:00
notes->widths.max_ins_name = annotation__max_ins_name(notes);
}
void annotation__update_column_widths(struct annotation *notes)
{
if (notes->options->use_offset)
notes->widths.target = notes->widths.min_addr;
else
notes->widths.target = notes->widths.max_addr;
notes->widths.addr = notes->widths.target;
if (notes->options->show_nr_jumps)
notes->widths.addr += notes->widths.jumps + 1;
}
static void annotation__calc_lines(struct annotation *notes, struct map *map,
struct rb_root *root,
struct annotation_options *opts)
{
struct annotation_line *al;
struct rb_root tmp_root = RB_ROOT;
list_for_each_entry(al, &notes->src->source, node) {
double percent_max = 0.0;
int i;
for (i = 0; i < al->data_nr; i++) {
double percent;
percent = annotation_data__percent(&al->data[i],
opts->percent_type);
if (percent > percent_max)
percent_max = percent;
}
if (percent_max <= 0.5)
continue;
al->path = get_srcline(map->dso, notes->start + al->offset, NULL,
false, true, notes->start + al->offset);
insert_source_line(&tmp_root, al, opts);
}
resort_source_line(root, &tmp_root);
}
static void symbol__calc_lines(struct map_symbol *ms, struct rb_root *root,
struct annotation_options *opts)
{
struct annotation *notes = symbol__annotation(ms->sym);
annotation__calc_lines(notes, ms->map, root, opts);
}
int symbol__tty_annotate2(struct map_symbol *ms, struct evsel *evsel,
struct annotation_options *opts)
{
struct dso *dso = ms->map->dso;
struct symbol *sym = ms->sym;
struct rb_root source_line = RB_ROOT;
struct hists *hists = evsel__hists(evsel);
char buf[1024];
int err;
err = symbol__annotate2(ms, evsel, opts, NULL);
if (err) {
char msg[BUFSIZ];
dso->annotate_warned = true;
symbol__strerror_disassemble(ms, err, msg, sizeof(msg));
ui__error("Couldn't annotate %s:\n%s", sym->name, msg);
return -1;
}
if (opts->print_lines) {
srcline_full_filename = opts->full_path;
symbol__calc_lines(ms, &source_line, opts);
print_summary(&source_line, dso->long_name);
}
hists__scnprintf_title(hists, buf, sizeof(buf));
fprintf(stdout, "%s, [percent: %s]\n%s() %s\n",
buf, percent_type_str(opts->percent_type), sym->name, dso->long_name);
symbol__annotate_fprintf2(sym, stdout, opts);
annotated_source__purge(symbol__annotation(sym)->src);
return 0;
}
int symbol__tty_annotate(struct map_symbol *ms, struct evsel *evsel,
struct annotation_options *opts)
{
struct dso *dso = ms->map->dso;
struct symbol *sym = ms->sym;
struct rb_root source_line = RB_ROOT;
int err;
err = symbol__annotate(ms, evsel, opts, NULL);
if (err) {
char msg[BUFSIZ];
dso->annotate_warned = true;
symbol__strerror_disassemble(ms, err, msg, sizeof(msg));
ui__error("Couldn't annotate %s:\n%s", sym->name, msg);
return -1;
}
symbol__calc_percent(sym, evsel);
if (opts->print_lines) {
srcline_full_filename = opts->full_path;
symbol__calc_lines(ms, &source_line, opts);
print_summary(&source_line, dso->long_name);
}
symbol__annotate_printf(ms, evsel, opts);
annotated_source__purge(symbol__annotation(sym)->src);
return 0;
}
bool ui__has_annotation(void)
{
return use_browser == 1 && perf_hpp_list.sym;
}
static double annotation_line__max_percent(struct annotation_line *al,
struct annotation *notes,
unsigned int percent_type)
{
double percent_max = 0.0;
int i;
for (i = 0; i < notes->nr_events; i++) {
double percent;
percent = annotation_data__percent(&al->data[i],
percent_type);
if (percent > percent_max)
percent_max = percent;
}
return percent_max;
}
static void disasm_line__write(struct disasm_line *dl, struct annotation *notes,
void *obj, char *bf, size_t size,
void (*obj__printf)(void *obj, const char *fmt, ...),
void (*obj__write_graph)(void *obj, int graph))
{
if (dl->ins.ops && dl->ins.ops->scnprintf) {
if (ins__is_jump(&dl->ins)) {
bool fwd;
if (dl->ops.target.outside)
goto call_like;
fwd = dl->ops.target.offset > dl->al.offset;
obj__write_graph(obj, fwd ? DARROW_CHAR : UARROW_CHAR);
obj__printf(obj, " ");
} else if (ins__is_call(&dl->ins)) {
call_like:
obj__write_graph(obj, RARROW_CHAR);
obj__printf(obj, " ");
} else if (ins__is_ret(&dl->ins)) {
obj__write_graph(obj, LARROW_CHAR);
obj__printf(obj, " ");
} else {
obj__printf(obj, " ");
}
} else {
obj__printf(obj, " ");
}
perf annotate: Calculate the max instruction name, align column to that We were hardcoding '6' as the max instruction name, and we have lots that are longer than that, see the diff from two 'P' printed TUI annotations for a libc function that uses instructions with long names, such as 'vpmovmskb' with its 9 chars: --- __strcmp_avx2.annotation.before 2019-03-06 16:31:39.368020425 -0300 +++ __strcmp_avx2.annotation 2019-03-06 16:32:12.079450508 -0300 @@ -2,284 +2,284 @@ Event: cycles:ppp Percent endbr64 - 0.10 mov %edi,%eax + 0.10 mov %edi,%eax - xor %edx,%edx + xor %edx,%edx - 3.54 vpxor %ymm7,%ymm7,%ymm7 + 3.54 vpxor %ymm7,%ymm7,%ymm7 - or %esi,%eax + or %esi,%eax - and $0xfff,%eax + and $0xfff,%eax - cmp $0xf80,%eax + cmp $0xf80,%eax - ↓ jg 370 + ↓ jg 370 - 27.07 vmovdqu (%rdi),%ymm1 + 27.07 vmovdqu (%rdi),%ymm1 - 7.97 vpcmpeqb (%rsi),%ymm1,%ymm0 + 7.97 vpcmpeqb (%rsi),%ymm1,%ymm0 - 2.15 vpminub %ymm1,%ymm0,%ymm0 + 2.15 vpminub %ymm1,%ymm0,%ymm0 - 4.09 vpcmpeqb %ymm7,%ymm0,%ymm0 + 4.09 vpcmpeqb %ymm7,%ymm0,%ymm0 - 0.43 vpmovmskb %ymm0,%ecx + 0.43 vpmovmskb %ymm0,%ecx - 1.53 test %ecx,%ecx + 1.53 test %ecx,%ecx - ↓ je b0 + ↓ je b0 - 5.26 tzcnt %ecx,%edx + 5.26 tzcnt %ecx,%edx - 18.40 movzbl (%rdi,%rdx,1),%eax + 18.40 movzbl (%rdi,%rdx,1),%eax - 7.09 movzbl (%rsi,%rdx,1),%edx + 7.09 movzbl (%rsi,%rdx,1),%edx - 3.34 sub %edx,%eax + 3.34 sub %edx,%eax 2.37 vzeroupper ← retq nop - 50: tzcnt %ecx,%edx + 50: tzcnt %ecx,%edx - movzbl 0x20(%rdi,%rdx,1),%eax + movzbl 0x20(%rdi,%rdx,1),%eax - movzbl 0x20(%rsi,%rdx,1),%edx + movzbl 0x20(%rsi,%rdx,1),%edx - sub %edx,%eax + sub %edx,%eax vzeroupper ← retq - data16 nopw %cs:0x0(%rax,%rax,1) + data16 nopw %cs:0x0(%rax,%rax,1) Reported-by: Travis Downs <travis.downs@gmail.com> LPU-Reference: CAOBGo4z1KfmWeOm6Et0cnX5Z6DWsG2PQbAvRn1MhVPJmXHrc5g@mail.gmail.com Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Link: https://lkml.kernel.org/n/tip-89wsdd9h9g6bvq52sgp6d0u4@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2019-03-06 22:40:15 +03:00
disasm_line__scnprintf(dl, bf, size, !notes->options->use_offset, notes->widths.max_ins_name);
}
perf annotate: Compute average IPC and IPC coverage per symbol Add support to 'perf report' annotate view or 'perf annotate --stdio2' to aggregate the IPC derived from timed LBRs per symbol. We compute the average IPC and the IPC coverage percentage. For example: $ perf annotate --stdio2 Percent IPC Cycle (Average IPC: 2.30, IPC Coverage: 54.8%) Disassembly of section .text: 000000000003aac0 <random@@GLIBC_2.2.5>: 8.32 3.28 sub $0x18,%rsp 3.28 mov $0x1,%esi 3.28 xor %eax,%eax 3.28 cmpl $0x0,argp_program_version_hook@@GLIBC_2.2.5+0x1e0 11.57 3.28 1 ↓ je 20 lock cmpxchg %esi,__abort_msg@@GLIBC_PRIVATE+0x8a0 ↓ jne 29 ↓ jmp 43 11.57 1.10 20: cmpxchg %esi,__abort_msg@@GLIBC_PRIVATE+0x8a0 0.00 1.10 1 ↓ je 43 29: lea __abort_msg@@GLIBC_PRIVATE+0x8a0,%rdi sub $0x80,%rsp → callq __lll_lock_wait_private add $0x80,%rsp 0.00 3.00 43: lea __ctype_b@GLIBC_2.2.5+0x38,%rdi 3.00 lea 0xc(%rsp),%rsi 8.49 3.00 1 → callq __random_r 7.91 1.94 cmpl $0x0,argp_program_version_hook@@GLIBC_2.2.5+0x1e0 0.00 1.94 1 ↓ je 68 lock decl __abort_msg@@GLIBC_PRIVATE+0x8a0 ↓ jne 70 ↓ jmp 8a 0.00 2.00 68: decl __abort_msg@@GLIBC_PRIVATE+0x8a0 21.56 2.00 1 ↓ je 8a 70: lea __abort_msg@@GLIBC_PRIVATE+0x8a0,%rdi sub $0x80,%rsp → callq __lll_unlock_wake_private add $0x80,%rsp 21.56 2.90 8a: movslq 0xc(%rsp),%rax 2.90 add $0x18,%rsp 9.03 2.90 1 ← retq It shows for this symbol the average IPC is 2.30 and the IPC coverage is 54.8%. Signed-off-by: Jin Yao <yao.jin@linux.intel.com> Reviewed-by: Ingo Molnar <mingo@kernel.org> Reviewed-by: Jiri Olsa <jolsa@kernel.org> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Andi Kleen <ak@linux.intel.com> Cc: Kan Liang <kan.liang@linux.intel.com> Cc: Peter Zijlstra <peterz@infradead.org> Link: http://lkml.kernel.org/r/1543586097-27632-2-git-send-email-yao.jin@linux.intel.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2018-11-30 16:54:54 +03:00
static void ipc_coverage_string(char *bf, int size, struct annotation *notes)
{
double ipc = 0.0, coverage = 0.0;
if (notes->hit_cycles)
ipc = notes->hit_insn / ((double)notes->hit_cycles);
if (notes->total_insn) {
coverage = notes->cover_insn * 100.0 /
((double)notes->total_insn);
}
scnprintf(bf, size, "(Average IPC: %.2f, IPC Coverage: %.1f%%)",
ipc, coverage);
}
static void __annotation_line__write(struct annotation_line *al, struct annotation *notes,
bool first_line, bool current_entry, bool change_color, int width,
void *obj, unsigned int percent_type,
int (*obj__set_color)(void *obj, int color),
void (*obj__set_percent_color)(void *obj, double percent, bool current),
int (*obj__set_jumps_percent_color)(void *obj, int nr, bool current),
void (*obj__printf)(void *obj, const char *fmt, ...),
void (*obj__write_graph)(void *obj, int graph))
{
double percent_max = annotation_line__max_percent(al, notes, percent_type);
int pcnt_width = annotation__pcnt_width(notes),
cycles_width = annotation__cycles_width(notes);
bool show_title = false;
char bf[256];
int printed;
if (first_line && (al->offset == -1 || percent_max == 0.0)) {
if (notes->have_cycles) {
if (al->ipc == 0.0 && al->cycles == 0)
show_title = true;
} else
show_title = true;
}
if (al->offset != -1 && percent_max != 0.0) {
int i;
for (i = 0; i < notes->nr_events; i++) {
double percent;
percent = annotation_data__percent(&al->data[i], percent_type);
obj__set_percent_color(obj, percent, current_entry);
perf annotate: Fix --show-total-period for tui/stdio2 perf annotate --show-total-period does not really show total period. The reason is we have two separate variables for the same purpose. One is in symbol_conf.show_total_period and another is annotation_options.show_total_period. We save command line option in symbol_conf.show_total_period but uses annotation_option.show_total_period while rendering tui/stdio2 browser. Though, we copy symbol_conf.show_total_period to annotation__default_options.show_total_period but that is not really effective as we don't use annotation__default_options once we copy default options to dynamic variable annotate.opts in cmd_annotate(). Instead of all these complication, keep only one variable and use it all over. symbol_conf.show_total_period is used by perf report/top as well. So let's kill annotation_options.show_total_period. On a side note, I've kept annotation_options.show_total_period definition because it's still used by perf-config code. Follow up patch to fix perf-config for annotate will remove annotation_options.show_total_period. Signed-off-by: Ravi Bangoria <ravi.bangoria@linux.ibm.com> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Alexey Budankov <alexey.budankov@linux.intel.com> Cc: Changbin Du <changbin.du@intel.com> Cc: Ian Rogers <irogers@google.com> Cc: Jin Yao <yao.jin@linux.intel.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Leo Yan <leo.yan@linaro.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Song Liu <songliubraving@fb.com> Cc: Taeung Song <treeze.taeung@gmail.com> Cc: Thomas Richter <tmricht@linux.ibm.com> Cc: Yisheng Xie <xieyisheng1@huawei.com> Link: http://lore.kernel.org/lkml/20200213064306.160480-3-ravi.bangoria@linux.ibm.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2020-02-13 09:43:00 +03:00
if (symbol_conf.show_total_period) {
obj__printf(obj, "%11" PRIu64 " ", al->data[i].he.period);
perf annotate: Fix --show-nr-samples for tui/stdio2 perf annotate --show-nr-samples does not really show number of samples. The reason is we have two separate variables for the same purpose. One is in symbol_conf.show_nr_samples and another is annotation_options.show_nr_samples. We save command line option in symbol_conf.show_nr_samples but uses annotation_option.show_nr_samples while rendering tui/stdio2 browser. Though, we copy symbol_conf.show_nr_samples to annotation__default_options.show_nr_samples but that is not really effective as we don't use annotation__default_options once we copy default options to dynamic variable annotate.opts in cmd_annotate(). Instead of all these complication, keep only one variable and use it all over. symbol_conf.show_nr_samples is used by perf report/top as well. So let's kill annotation_options.show_nr_samples. On a side note, I've kept annotation_options.show_nr_samples definition because it's still used by perf-config code. Follow up patch to fix perf-config for annotate will remove annotation_options.show_nr_samples. Signed-off-by: Ravi Bangoria <ravi.bangoria@linux.ibm.com> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Alexey Budankov <alexey.budankov@linux.intel.com> Cc: Changbin Du <changbin.du@intel.com> Cc: Ian Rogers <irogers@google.com> Cc: Jin Yao <yao.jin@linux.intel.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Leo Yan <leo.yan@linaro.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Song Liu <songliubraving@fb.com> Cc: Taeung Song <treeze.taeung@gmail.com> Cc: Thomas Richter <tmricht@linux.ibm.com> Cc: Yisheng Xie <xieyisheng1@huawei.com> Link: http://lore.kernel.org/lkml/20200213064306.160480-4-ravi.bangoria@linux.ibm.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2020-02-13 09:43:01 +03:00
} else if (symbol_conf.show_nr_samples) {
obj__printf(obj, "%6" PRIu64 " ",
al->data[i].he.nr_samples);
} else {
obj__printf(obj, "%6.2f ", percent);
}
}
} else {
obj__set_percent_color(obj, 0, current_entry);
if (!show_title)
obj__printf(obj, "%-*s", pcnt_width, " ");
else {
obj__printf(obj, "%-*s", pcnt_width,
perf annotate: Fix --show-total-period for tui/stdio2 perf annotate --show-total-period does not really show total period. The reason is we have two separate variables for the same purpose. One is in symbol_conf.show_total_period and another is annotation_options.show_total_period. We save command line option in symbol_conf.show_total_period but uses annotation_option.show_total_period while rendering tui/stdio2 browser. Though, we copy symbol_conf.show_total_period to annotation__default_options.show_total_period but that is not really effective as we don't use annotation__default_options once we copy default options to dynamic variable annotate.opts in cmd_annotate(). Instead of all these complication, keep only one variable and use it all over. symbol_conf.show_total_period is used by perf report/top as well. So let's kill annotation_options.show_total_period. On a side note, I've kept annotation_options.show_total_period definition because it's still used by perf-config code. Follow up patch to fix perf-config for annotate will remove annotation_options.show_total_period. Signed-off-by: Ravi Bangoria <ravi.bangoria@linux.ibm.com> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Alexey Budankov <alexey.budankov@linux.intel.com> Cc: Changbin Du <changbin.du@intel.com> Cc: Ian Rogers <irogers@google.com> Cc: Jin Yao <yao.jin@linux.intel.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Leo Yan <leo.yan@linaro.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Song Liu <songliubraving@fb.com> Cc: Taeung Song <treeze.taeung@gmail.com> Cc: Thomas Richter <tmricht@linux.ibm.com> Cc: Yisheng Xie <xieyisheng1@huawei.com> Link: http://lore.kernel.org/lkml/20200213064306.160480-3-ravi.bangoria@linux.ibm.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2020-02-13 09:43:00 +03:00
symbol_conf.show_total_period ? "Period" :
perf annotate: Fix --show-nr-samples for tui/stdio2 perf annotate --show-nr-samples does not really show number of samples. The reason is we have two separate variables for the same purpose. One is in symbol_conf.show_nr_samples and another is annotation_options.show_nr_samples. We save command line option in symbol_conf.show_nr_samples but uses annotation_option.show_nr_samples while rendering tui/stdio2 browser. Though, we copy symbol_conf.show_nr_samples to annotation__default_options.show_nr_samples but that is not really effective as we don't use annotation__default_options once we copy default options to dynamic variable annotate.opts in cmd_annotate(). Instead of all these complication, keep only one variable and use it all over. symbol_conf.show_nr_samples is used by perf report/top as well. So let's kill annotation_options.show_nr_samples. On a side note, I've kept annotation_options.show_nr_samples definition because it's still used by perf-config code. Follow up patch to fix perf-config for annotate will remove annotation_options.show_nr_samples. Signed-off-by: Ravi Bangoria <ravi.bangoria@linux.ibm.com> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Alexey Budankov <alexey.budankov@linux.intel.com> Cc: Changbin Du <changbin.du@intel.com> Cc: Ian Rogers <irogers@google.com> Cc: Jin Yao <yao.jin@linux.intel.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Leo Yan <leo.yan@linaro.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Song Liu <songliubraving@fb.com> Cc: Taeung Song <treeze.taeung@gmail.com> Cc: Thomas Richter <tmricht@linux.ibm.com> Cc: Yisheng Xie <xieyisheng1@huawei.com> Link: http://lore.kernel.org/lkml/20200213064306.160480-4-ravi.bangoria@linux.ibm.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2020-02-13 09:43:01 +03:00
symbol_conf.show_nr_samples ? "Samples" : "Percent");
}
}
if (notes->have_cycles) {
if (al->ipc)
obj__printf(obj, "%*.2f ", ANNOTATION__IPC_WIDTH - 1, al->ipc);
else if (!show_title)
obj__printf(obj, "%*s", ANNOTATION__IPC_WIDTH, " ");
else
obj__printf(obj, "%*s ", ANNOTATION__IPC_WIDTH - 1, "IPC");
perf annotate: Create hotkey 'c' to show min/max cycles In the 'perf annotate' view, a new hotkey 'c' is created for showing the min/max cycles. For example, when press 'c', the annotate view is: Percent│ IPC Cycle(min/max) │ │ │ Disassembly of section .text: │ │ 000000000003aab0 <random@@GLIBC_2.2.5>: 8.22 │3.92 sub $0x18,%rsp │3.92 mov $0x1,%esi │3.92 xor %eax,%eax │3.92 cmpl $0x0,argp_program_version_hook@@G │3.92 1(2/1) ↓ je 20 │ lock cmpxchg %esi,__abort_msg@@GLIBC_P │ ↓ jne 29 │ ↓ jmp 43 │1.10 20: cmpxchg %esi,__abort_msg@@GLIBC_PRIVATE+ 8.93 │1.10 1(5/1) ↓ je 43 When press 'c' again, the annotate view is switched back: Percent│ IPC Cycle │ │ │ Disassembly of section .text: │ │ 000000000003aab0 <random@@GLIBC_2.2.5>: 8.22 │3.92 sub $0x18,%rsp │3.92 mov $0x1,%esi │3.92 xor %eax,%eax │3.92 cmpl $0x0,argp_program_version_hook@@GLIBC_2.2.5+0x │3.92 1 ↓ je 20 │ lock cmpxchg %esi,__abort_msg@@GLIBC_PRIVATE+0x8a0 │ ↓ jne 29 │ ↓ jmp 43 │1.10 20: cmpxchg %esi,__abort_msg@@GLIBC_PRIVATE+0x8a0 8.93 │1.10 1 ↓ je 43 Signed-off-by: Jin Yao <yao.jin@linux.intel.com> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Andi Kleen <ak@linux.intel.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Kan Liang <kan.liang@linux.intel.com> Cc: Peter Zijlstra <peterz@infradead.org> Link: http://lkml.kernel.org/r/1526569118-14217-3-git-send-email-yao.jin@linux.intel.com [ Rename all maxmin to minmax ] Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2018-05-17 17:58:38 +03:00
if (!notes->options->show_minmax_cycle) {
if (al->cycles)
obj__printf(obj, "%*" PRIu64 " ",
ANNOTATION__CYCLES_WIDTH - 1, al->cycles);
perf annotate: Create hotkey 'c' to show min/max cycles In the 'perf annotate' view, a new hotkey 'c' is created for showing the min/max cycles. For example, when press 'c', the annotate view is: Percent│ IPC Cycle(min/max) │ │ │ Disassembly of section .text: │ │ 000000000003aab0 <random@@GLIBC_2.2.5>: 8.22 │3.92 sub $0x18,%rsp │3.92 mov $0x1,%esi │3.92 xor %eax,%eax │3.92 cmpl $0x0,argp_program_version_hook@@G │3.92 1(2/1) ↓ je 20 │ lock cmpxchg %esi,__abort_msg@@GLIBC_P │ ↓ jne 29 │ ↓ jmp 43 │1.10 20: cmpxchg %esi,__abort_msg@@GLIBC_PRIVATE+ 8.93 │1.10 1(5/1) ↓ je 43 When press 'c' again, the annotate view is switched back: Percent│ IPC Cycle │ │ │ Disassembly of section .text: │ │ 000000000003aab0 <random@@GLIBC_2.2.5>: 8.22 │3.92 sub $0x18,%rsp │3.92 mov $0x1,%esi │3.92 xor %eax,%eax │3.92 cmpl $0x0,argp_program_version_hook@@GLIBC_2.2.5+0x │3.92 1 ↓ je 20 │ lock cmpxchg %esi,__abort_msg@@GLIBC_PRIVATE+0x8a0 │ ↓ jne 29 │ ↓ jmp 43 │1.10 20: cmpxchg %esi,__abort_msg@@GLIBC_PRIVATE+0x8a0 8.93 │1.10 1 ↓ je 43 Signed-off-by: Jin Yao <yao.jin@linux.intel.com> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Andi Kleen <ak@linux.intel.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Kan Liang <kan.liang@linux.intel.com> Cc: Peter Zijlstra <peterz@infradead.org> Link: http://lkml.kernel.org/r/1526569118-14217-3-git-send-email-yao.jin@linux.intel.com [ Rename all maxmin to minmax ] Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2018-05-17 17:58:38 +03:00
else if (!show_title)
obj__printf(obj, "%*s",
ANNOTATION__CYCLES_WIDTH, " ");
else
obj__printf(obj, "%*s ",
ANNOTATION__CYCLES_WIDTH - 1,
"Cycle");
} else {
if (al->cycles) {
char str[32];
scnprintf(str, sizeof(str),
"%" PRIu64 "(%" PRIu64 "/%" PRIu64 ")",
al->cycles, al->cycles_min,
al->cycles_max);
obj__printf(obj, "%*s ",
ANNOTATION__MINMAX_CYCLES_WIDTH - 1,
str);
} else if (!show_title)
obj__printf(obj, "%*s",
ANNOTATION__MINMAX_CYCLES_WIDTH,
" ");
else
obj__printf(obj, "%*s ",
ANNOTATION__MINMAX_CYCLES_WIDTH - 1,
"Cycle(min/max)");
}
perf annotate: Compute average IPC and IPC coverage per symbol Add support to 'perf report' annotate view or 'perf annotate --stdio2' to aggregate the IPC derived from timed LBRs per symbol. We compute the average IPC and the IPC coverage percentage. For example: $ perf annotate --stdio2 Percent IPC Cycle (Average IPC: 2.30, IPC Coverage: 54.8%) Disassembly of section .text: 000000000003aac0 <random@@GLIBC_2.2.5>: 8.32 3.28 sub $0x18,%rsp 3.28 mov $0x1,%esi 3.28 xor %eax,%eax 3.28 cmpl $0x0,argp_program_version_hook@@GLIBC_2.2.5+0x1e0 11.57 3.28 1 ↓ je 20 lock cmpxchg %esi,__abort_msg@@GLIBC_PRIVATE+0x8a0 ↓ jne 29 ↓ jmp 43 11.57 1.10 20: cmpxchg %esi,__abort_msg@@GLIBC_PRIVATE+0x8a0 0.00 1.10 1 ↓ je 43 29: lea __abort_msg@@GLIBC_PRIVATE+0x8a0,%rdi sub $0x80,%rsp → callq __lll_lock_wait_private add $0x80,%rsp 0.00 3.00 43: lea __ctype_b@GLIBC_2.2.5+0x38,%rdi 3.00 lea 0xc(%rsp),%rsi 8.49 3.00 1 → callq __random_r 7.91 1.94 cmpl $0x0,argp_program_version_hook@@GLIBC_2.2.5+0x1e0 0.00 1.94 1 ↓ je 68 lock decl __abort_msg@@GLIBC_PRIVATE+0x8a0 ↓ jne 70 ↓ jmp 8a 0.00 2.00 68: decl __abort_msg@@GLIBC_PRIVATE+0x8a0 21.56 2.00 1 ↓ je 8a 70: lea __abort_msg@@GLIBC_PRIVATE+0x8a0,%rdi sub $0x80,%rsp → callq __lll_unlock_wake_private add $0x80,%rsp 21.56 2.90 8a: movslq 0xc(%rsp),%rax 2.90 add $0x18,%rsp 9.03 2.90 1 ← retq It shows for this symbol the average IPC is 2.30 and the IPC coverage is 54.8%. Signed-off-by: Jin Yao <yao.jin@linux.intel.com> Reviewed-by: Ingo Molnar <mingo@kernel.org> Reviewed-by: Jiri Olsa <jolsa@kernel.org> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Andi Kleen <ak@linux.intel.com> Cc: Kan Liang <kan.liang@linux.intel.com> Cc: Peter Zijlstra <peterz@infradead.org> Link: http://lkml.kernel.org/r/1543586097-27632-2-git-send-email-yao.jin@linux.intel.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2018-11-30 16:54:54 +03:00
if (show_title && !*al->line) {
ipc_coverage_string(bf, sizeof(bf), notes);
obj__printf(obj, "%*s", ANNOTATION__AVG_IPC_WIDTH, bf);
}
}
obj__printf(obj, " ");
if (!*al->line)
obj__printf(obj, "%-*s", width - pcnt_width - cycles_width, " ");
else if (al->offset == -1) {
if (al->line_nr && notes->options->show_linenr)
printed = scnprintf(bf, sizeof(bf), "%-*d ", notes->widths.addr + 1, al->line_nr);
else
printed = scnprintf(bf, sizeof(bf), "%-*s ", notes->widths.addr, " ");
obj__printf(obj, bf);
obj__printf(obj, "%-*s", width - printed - pcnt_width - cycles_width + 1, al->line);
} else {
u64 addr = al->offset;
int color = -1;
if (!notes->options->use_offset)
addr += notes->start;
if (!notes->options->use_offset) {
printed = scnprintf(bf, sizeof(bf), "%" PRIx64 ": ", addr);
} else {
if (al->jump_sources &&
notes->options->offset_level >= ANNOTATION__OFFSET_JUMP_TARGETS) {
if (notes->options->show_nr_jumps) {
int prev;
printed = scnprintf(bf, sizeof(bf), "%*d ",
notes->widths.jumps,
al->jump_sources);
prev = obj__set_jumps_percent_color(obj, al->jump_sources,
current_entry);
obj__printf(obj, bf);
obj__set_color(obj, prev);
}
print_addr:
printed = scnprintf(bf, sizeof(bf), "%*" PRIx64 ": ",
notes->widths.target, addr);
} else if (ins__is_call(&disasm_line(al)->ins) &&
notes->options->offset_level >= ANNOTATION__OFFSET_CALL) {
goto print_addr;
} else if (notes->options->offset_level == ANNOTATION__MAX_OFFSET_LEVEL) {
goto print_addr;
} else {
printed = scnprintf(bf, sizeof(bf), "%-*s ",
notes->widths.addr, " ");
}
}
if (change_color)
color = obj__set_color(obj, HE_COLORSET_ADDR);
obj__printf(obj, bf);
if (change_color)
obj__set_color(obj, color);
disasm_line__write(disasm_line(al), notes, obj, bf, sizeof(bf), obj__printf, obj__write_graph);
obj__printf(obj, "%-*s", width - pcnt_width - cycles_width - 3 - printed, bf);
}
}
void annotation_line__write(struct annotation_line *al, struct annotation *notes,
struct annotation_write_ops *wops,
struct annotation_options *opts)
{
__annotation_line__write(al, notes, wops->first_line, wops->current_entry,
wops->change_color, wops->width, wops->obj,
opts->percent_type,
wops->set_color, wops->set_percent_color,
wops->set_jumps_percent_color, wops->printf,
wops->write_graph);
}
int symbol__annotate2(struct map_symbol *ms, struct evsel *evsel,
struct annotation_options *options, struct arch **parch)
{
struct symbol *sym = ms->sym;
struct annotation *notes = symbol__annotation(sym);
size_t size = symbol__size(sym);
int nr_pcnt = 1, err;
notes->offsets = zalloc(size * sizeof(struct annotation_line *));
if (notes->offsets == NULL)
return ENOMEM;
if (evsel__is_group_event(evsel))
nr_pcnt = evsel->core.nr_members;
err = symbol__annotate(ms, evsel, options, parch);
if (err)
goto out_free_offsets;
notes->options = options;
symbol__calc_percent(sym, evsel);
annotation__set_offsets(notes, size);
annotation__mark_jump_targets(notes, sym);
annotation__compute_ipc(notes, size);
annotation__init_column_widths(notes, sym);
notes->nr_events = nr_pcnt;
annotation__update_column_widths(notes);
sym->annotate2 = 1;
return 0;
out_free_offsets:
zfree(&notes->offsets);
return err;
}
perf annotate: Make perf config effective perf default config set by user in [annotate] section is totally ignored by annotate code. Fix it. Before: $ ./perf config annotate.hide_src_code=true annotate.show_nr_jumps=true annotate.show_nr_samples=true $ ./perf annotate shash │ unsigned h = 0; │ movl $0x0,-0xc(%rbp) │ while (*s) │ ↓ jmp 44 │ h = 65599 * h + *s++; 11.33 │24: mov -0xc(%rbp),%eax 43.50 │ imul $0x1003f,%eax,%ecx │ mov -0x18(%rbp),%rax After: │ movl $0x0,-0xc(%rbp) │ ↓ jmp 44 1 │1 24: mov -0xc(%rbp),%eax 4 │ imul $0x1003f,%eax,%ecx │ mov -0x18(%rbp),%rax Note that we have removed show_nr_samples and show_total_period from annotation_options because they are not used. Instead of them we use symbol_conf.show_nr_samples and symbol_conf.show_total_period. Committer testing: Using 'perf annotate --stdio2' to use the TUI rendering but emitting the output to stdio: # perf config # # perf config annotate.hide_src_code=true # perf config annotate.hide_src_code=true # # perf config annotate.show_nr_jumps=true # perf config annotate.show_nr_samples=true # perf config annotate.hide_src_code=true annotate.show_nr_jumps=true annotate.show_nr_samples=true # # Before: # perf annotate --stdio2 ObjectInstance::weak_pointer_was_finalized Samples: 1 of event 'cycles', 4000 Hz, Event count (approx.): 830873, [percent: local period] ObjectInstance::weak_pointer_was_finalized() /usr/lib64/libgjs.so.0.0.0 Percent 00000000000609f0 <ObjectInstance::weak_pointer_was_finalized()@@Base>: endbr64 cmpq $0x0,0x20(%rdi) ↓ je 10 xor %eax,%eax ← retq xchg %ax,%ax 100.00 10: push %rbp cmpq $0x0,0x18(%rdi) mov %rdi,%rbp ↓ jne 20 1b: xor %eax,%eax pop %rbp ← retq nop 20: lea 0x18(%rdi),%rdi → callq JS_UpdateWeakPointerAfterGC(JS::Heap<JSObject* cmpq $0x0,0x18(%rbp) ↑ jne 1b mov %rbp,%rdi → callq ObjectBase::jsobj_addr() const@plt mov $0x1,%eax pop %rbp ← retq # After: # perf annotate --stdio2 ObjectInstance::weak_pointer_was_finalized 2> /dev/null Samples: 1 of event 'cycles', 4000 Hz, Event count (approx.): 830873, [percent: local period] ObjectInstance::weak_pointer_was_finalized() /usr/lib64/libgjs.so.0.0.0 Samples endbr64 cmpq $0x0,0x20(%rdi) ↓ je 10 xor %eax,%eax ← retq xchg %ax,%ax 1 1 10: push %rbp cmpq $0x0,0x18(%rdi) mov %rdi,%rbp ↓ jne 20 1 1b: xor %eax,%eax pop %rbp ← retq nop 1 20: lea 0x18(%rdi),%rdi → callq JS_UpdateWeakPointerAfterGC(JS::Heap<JSObject* cmpq $0x0,0x18(%rbp) ↑ jne 1b mov %rbp,%rdi → callq ObjectBase::jsobj_addr() const@plt mov $0x1,%eax pop %rbp ← retq # # perf config annotate.show_nr_jumps annotate.show_nr_jumps=true # perf config annotate.show_nr_jumps=false # perf config annotate.show_nr_jumps annotate.show_nr_jumps=false # # perf annotate --stdio2 ObjectInstance::weak_pointer_was_finalized 2> /dev/null Samples: 1 of event 'cycles', 4000 Hz, Event count (approx.): 830873, [percent: local period] ObjectInstance::weak_pointer_was_finalized() /usr/lib64/libgjs.so.0.0.0 Samples endbr64 cmpq $0x0,0x20(%rdi) ↓ je 10 xor %eax,%eax ← retq xchg %ax,%ax 1 10: push %rbp cmpq $0x0,0x18(%rdi) mov %rdi,%rbp ↓ jne 20 1b: xor %eax,%eax pop %rbp ← retq nop 20: lea 0x18(%rdi),%rdi → callq JS_UpdateWeakPointerAfterGC(JS::Heap<JSObject* cmpq $0x0,0x18(%rbp) ↑ jne 1b mov %rbp,%rdi → callq ObjectBase::jsobj_addr() const@plt mov $0x1,%eax pop %rbp ← retq # Signed-off-by: Ravi Bangoria <ravi.bangoria@linux.ibm.com> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Alexey Budankov <alexey.budankov@linux.intel.com> Cc: Changbin Du <changbin.du@intel.com> Cc: Ian Rogers <irogers@google.com> Cc: Jin Yao <yao.jin@linux.intel.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Leo Yan <leo.yan@linaro.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Song Liu <songliubraving@fb.com> Cc: Taeung Song <treeze.taeung@gmail.com> Cc: Thomas Richter <tmricht@linux.ibm.com> Cc: Yisheng Xie <xieyisheng1@huawei.com> Link: http://lore.kernel.org/lkml/20200213064306.160480-6-ravi.bangoria@linux.ibm.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2020-02-13 09:43:03 +03:00
static int annotation__config(const char *var, const char *value, void *data)
{
perf annotate: Make perf config effective perf default config set by user in [annotate] section is totally ignored by annotate code. Fix it. Before: $ ./perf config annotate.hide_src_code=true annotate.show_nr_jumps=true annotate.show_nr_samples=true $ ./perf annotate shash │ unsigned h = 0; │ movl $0x0,-0xc(%rbp) │ while (*s) │ ↓ jmp 44 │ h = 65599 * h + *s++; 11.33 │24: mov -0xc(%rbp),%eax 43.50 │ imul $0x1003f,%eax,%ecx │ mov -0x18(%rbp),%rax After: │ movl $0x0,-0xc(%rbp) │ ↓ jmp 44 1 │1 24: mov -0xc(%rbp),%eax 4 │ imul $0x1003f,%eax,%ecx │ mov -0x18(%rbp),%rax Note that we have removed show_nr_samples and show_total_period from annotation_options because they are not used. Instead of them we use symbol_conf.show_nr_samples and symbol_conf.show_total_period. Committer testing: Using 'perf annotate --stdio2' to use the TUI rendering but emitting the output to stdio: # perf config # # perf config annotate.hide_src_code=true # perf config annotate.hide_src_code=true # # perf config annotate.show_nr_jumps=true # perf config annotate.show_nr_samples=true # perf config annotate.hide_src_code=true annotate.show_nr_jumps=true annotate.show_nr_samples=true # # Before: # perf annotate --stdio2 ObjectInstance::weak_pointer_was_finalized Samples: 1 of event 'cycles', 4000 Hz, Event count (approx.): 830873, [percent: local period] ObjectInstance::weak_pointer_was_finalized() /usr/lib64/libgjs.so.0.0.0 Percent 00000000000609f0 <ObjectInstance::weak_pointer_was_finalized()@@Base>: endbr64 cmpq $0x0,0x20(%rdi) ↓ je 10 xor %eax,%eax ← retq xchg %ax,%ax 100.00 10: push %rbp cmpq $0x0,0x18(%rdi) mov %rdi,%rbp ↓ jne 20 1b: xor %eax,%eax pop %rbp ← retq nop 20: lea 0x18(%rdi),%rdi → callq JS_UpdateWeakPointerAfterGC(JS::Heap<JSObject* cmpq $0x0,0x18(%rbp) ↑ jne 1b mov %rbp,%rdi → callq ObjectBase::jsobj_addr() const@plt mov $0x1,%eax pop %rbp ← retq # After: # perf annotate --stdio2 ObjectInstance::weak_pointer_was_finalized 2> /dev/null Samples: 1 of event 'cycles', 4000 Hz, Event count (approx.): 830873, [percent: local period] ObjectInstance::weak_pointer_was_finalized() /usr/lib64/libgjs.so.0.0.0 Samples endbr64 cmpq $0x0,0x20(%rdi) ↓ je 10 xor %eax,%eax ← retq xchg %ax,%ax 1 1 10: push %rbp cmpq $0x0,0x18(%rdi) mov %rdi,%rbp ↓ jne 20 1 1b: xor %eax,%eax pop %rbp ← retq nop 1 20: lea 0x18(%rdi),%rdi → callq JS_UpdateWeakPointerAfterGC(JS::Heap<JSObject* cmpq $0x0,0x18(%rbp) ↑ jne 1b mov %rbp,%rdi → callq ObjectBase::jsobj_addr() const@plt mov $0x1,%eax pop %rbp ← retq # # perf config annotate.show_nr_jumps annotate.show_nr_jumps=true # perf config annotate.show_nr_jumps=false # perf config annotate.show_nr_jumps annotate.show_nr_jumps=false # # perf annotate --stdio2 ObjectInstance::weak_pointer_was_finalized 2> /dev/null Samples: 1 of event 'cycles', 4000 Hz, Event count (approx.): 830873, [percent: local period] ObjectInstance::weak_pointer_was_finalized() /usr/lib64/libgjs.so.0.0.0 Samples endbr64 cmpq $0x0,0x20(%rdi) ↓ je 10 xor %eax,%eax ← retq xchg %ax,%ax 1 10: push %rbp cmpq $0x0,0x18(%rdi) mov %rdi,%rbp ↓ jne 20 1b: xor %eax,%eax pop %rbp ← retq nop 20: lea 0x18(%rdi),%rdi → callq JS_UpdateWeakPointerAfterGC(JS::Heap<JSObject* cmpq $0x0,0x18(%rbp) ↑ jne 1b mov %rbp,%rdi → callq ObjectBase::jsobj_addr() const@plt mov $0x1,%eax pop %rbp ← retq # Signed-off-by: Ravi Bangoria <ravi.bangoria@linux.ibm.com> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Alexey Budankov <alexey.budankov@linux.intel.com> Cc: Changbin Du <changbin.du@intel.com> Cc: Ian Rogers <irogers@google.com> Cc: Jin Yao <yao.jin@linux.intel.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Leo Yan <leo.yan@linaro.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Song Liu <songliubraving@fb.com> Cc: Taeung Song <treeze.taeung@gmail.com> Cc: Thomas Richter <tmricht@linux.ibm.com> Cc: Yisheng Xie <xieyisheng1@huawei.com> Link: http://lore.kernel.org/lkml/20200213064306.160480-6-ravi.bangoria@linux.ibm.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2020-02-13 09:43:03 +03:00
struct annotation_options *opt = data;
if (!strstarts(var, "annotate."))
return 0;
perf annotate: Make perf config effective perf default config set by user in [annotate] section is totally ignored by annotate code. Fix it. Before: $ ./perf config annotate.hide_src_code=true annotate.show_nr_jumps=true annotate.show_nr_samples=true $ ./perf annotate shash │ unsigned h = 0; │ movl $0x0,-0xc(%rbp) │ while (*s) │ ↓ jmp 44 │ h = 65599 * h + *s++; 11.33 │24: mov -0xc(%rbp),%eax 43.50 │ imul $0x1003f,%eax,%ecx │ mov -0x18(%rbp),%rax After: │ movl $0x0,-0xc(%rbp) │ ↓ jmp 44 1 │1 24: mov -0xc(%rbp),%eax 4 │ imul $0x1003f,%eax,%ecx │ mov -0x18(%rbp),%rax Note that we have removed show_nr_samples and show_total_period from annotation_options because they are not used. Instead of them we use symbol_conf.show_nr_samples and symbol_conf.show_total_period. Committer testing: Using 'perf annotate --stdio2' to use the TUI rendering but emitting the output to stdio: # perf config # # perf config annotate.hide_src_code=true # perf config annotate.hide_src_code=true # # perf config annotate.show_nr_jumps=true # perf config annotate.show_nr_samples=true # perf config annotate.hide_src_code=true annotate.show_nr_jumps=true annotate.show_nr_samples=true # # Before: # perf annotate --stdio2 ObjectInstance::weak_pointer_was_finalized Samples: 1 of event 'cycles', 4000 Hz, Event count (approx.): 830873, [percent: local period] ObjectInstance::weak_pointer_was_finalized() /usr/lib64/libgjs.so.0.0.0 Percent 00000000000609f0 <ObjectInstance::weak_pointer_was_finalized()@@Base>: endbr64 cmpq $0x0,0x20(%rdi) ↓ je 10 xor %eax,%eax ← retq xchg %ax,%ax 100.00 10: push %rbp cmpq $0x0,0x18(%rdi) mov %rdi,%rbp ↓ jne 20 1b: xor %eax,%eax pop %rbp ← retq nop 20: lea 0x18(%rdi),%rdi → callq JS_UpdateWeakPointerAfterGC(JS::Heap<JSObject* cmpq $0x0,0x18(%rbp) ↑ jne 1b mov %rbp,%rdi → callq ObjectBase::jsobj_addr() const@plt mov $0x1,%eax pop %rbp ← retq # After: # perf annotate --stdio2 ObjectInstance::weak_pointer_was_finalized 2> /dev/null Samples: 1 of event 'cycles', 4000 Hz, Event count (approx.): 830873, [percent: local period] ObjectInstance::weak_pointer_was_finalized() /usr/lib64/libgjs.so.0.0.0 Samples endbr64 cmpq $0x0,0x20(%rdi) ↓ je 10 xor %eax,%eax ← retq xchg %ax,%ax 1 1 10: push %rbp cmpq $0x0,0x18(%rdi) mov %rdi,%rbp ↓ jne 20 1 1b: xor %eax,%eax pop %rbp ← retq nop 1 20: lea 0x18(%rdi),%rdi → callq JS_UpdateWeakPointerAfterGC(JS::Heap<JSObject* cmpq $0x0,0x18(%rbp) ↑ jne 1b mov %rbp,%rdi → callq ObjectBase::jsobj_addr() const@plt mov $0x1,%eax pop %rbp ← retq # # perf config annotate.show_nr_jumps annotate.show_nr_jumps=true # perf config annotate.show_nr_jumps=false # perf config annotate.show_nr_jumps annotate.show_nr_jumps=false # # perf annotate --stdio2 ObjectInstance::weak_pointer_was_finalized 2> /dev/null Samples: 1 of event 'cycles', 4000 Hz, Event count (approx.): 830873, [percent: local period] ObjectInstance::weak_pointer_was_finalized() /usr/lib64/libgjs.so.0.0.0 Samples endbr64 cmpq $0x0,0x20(%rdi) ↓ je 10 xor %eax,%eax ← retq xchg %ax,%ax 1 10: push %rbp cmpq $0x0,0x18(%rdi) mov %rdi,%rbp ↓ jne 20 1b: xor %eax,%eax pop %rbp ← retq nop 20: lea 0x18(%rdi),%rdi → callq JS_UpdateWeakPointerAfterGC(JS::Heap<JSObject* cmpq $0x0,0x18(%rbp) ↑ jne 1b mov %rbp,%rdi → callq ObjectBase::jsobj_addr() const@plt mov $0x1,%eax pop %rbp ← retq # Signed-off-by: Ravi Bangoria <ravi.bangoria@linux.ibm.com> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Alexey Budankov <alexey.budankov@linux.intel.com> Cc: Changbin Du <changbin.du@intel.com> Cc: Ian Rogers <irogers@google.com> Cc: Jin Yao <yao.jin@linux.intel.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Leo Yan <leo.yan@linaro.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Song Liu <songliubraving@fb.com> Cc: Taeung Song <treeze.taeung@gmail.com> Cc: Thomas Richter <tmricht@linux.ibm.com> Cc: Yisheng Xie <xieyisheng1@huawei.com> Link: http://lore.kernel.org/lkml/20200213064306.160480-6-ravi.bangoria@linux.ibm.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2020-02-13 09:43:03 +03:00
if (!strcmp(var, "annotate.offset_level")) {
perf_config_u8(&opt->offset_level, "offset_level", value);
if (opt->offset_level > ANNOTATION__MAX_OFFSET_LEVEL)
opt->offset_level = ANNOTATION__MAX_OFFSET_LEVEL;
else if (opt->offset_level < ANNOTATION__MIN_OFFSET_LEVEL)
opt->offset_level = ANNOTATION__MIN_OFFSET_LEVEL;
} else if (!strcmp(var, "annotate.hide_src_code")) {
opt->hide_src_code = perf_config_bool("hide_src_code", value);
} else if (!strcmp(var, "annotate.jump_arrows")) {
opt->jump_arrows = perf_config_bool("jump_arrows", value);
} else if (!strcmp(var, "annotate.show_linenr")) {
opt->show_linenr = perf_config_bool("show_linenr", value);
} else if (!strcmp(var, "annotate.show_nr_jumps")) {
opt->show_nr_jumps = perf_config_bool("show_nr_jumps", value);
} else if (!strcmp(var, "annotate.show_nr_samples")) {
symbol_conf.show_nr_samples = perf_config_bool("show_nr_samples",
value);
} else if (!strcmp(var, "annotate.show_total_period")) {
symbol_conf.show_total_period = perf_config_bool("show_total_period",
value);
} else if (!strcmp(var, "annotate.use_offset")) {
opt->use_offset = perf_config_bool("use_offset", value);
perf annotate: Allow configuring the 'disassembler_style' knob via 'perf config' # perf annotate --stdio2 acpi_processor_ffh_cstate_enter > default # perf config annotate.disassembler_style=intel # perf config annotate.disassembler_style annotate.disassembler_style=intel # perf annotate --stdio2 acpi_processor_ffh_cstate_enter > intel # diff -u default intel --- default 2020-09-04 13:09:26.019205732 -0300 +++ intel 2020-09-04 13:09:52.823795081 -0300 @@ -1,42 +1,42 @@ Samples: 1K of event 'cycles', 4000 Hz, Event count (approx.): 990065316, [percent: local period] acpi_processor_ffh_cstate_enter() /lib/modules/5.9.0-rc3/build/vmlinux -Percent → callq __fentry__ - mov cpu_number,%edx - mov %edx,%edx - mov cpu_cstate_entry,%rax - add -0x7dbe9700(,%rdx,8),%rax - movzbl 0x9(%rdi),%edx - mov 0x4(%rax,%rdx,8),%edi - mov (%rax,%rdx,8),%esi - → jmpq 137ccc6 - 2d: → jmpq 137ccd8 +Percent → call __fentry__ + mov edx,DWORD PTR gs:[rip+0x7e541d74] + mov edx,edx + mov rax,QWORD PTR [rip+0x152b8fb] + add rax,QWORD PTR [rdx*8-0x7dbe9700] + movzx edx,BYTE PTR [rdi+0x9] + mov edi,DWORD PTR [rax+rdx*8+0x4] + mov esi,DWORD PTR [rax+rdx*8] + → jmp 137ccc6 + 2d: → jmp 137ccd8 mfence - mov %gs:0x17bc0,%rax - clflush (%rax) + mov rax,QWORD PTR gs:0x17bc0 + clflush BYTE PTR [rax] mfence - xor %edx,%edx - mov %rdx,%rcx - mov %gs:0x17bc0,%rax - 0.00 monitor %rax,%ecx,%edx - mov (%rax),%rax - test $0x8,%al + xor edx,edx + mov rcx,rdx + mov rax,QWORD PTR gs:0x17bc0 + 0.00 monitor + mov rax,QWORD PTR [rax] + test al,0x8 ↓ jne 71 - ↓ jmpq 68 - verw 0x538b08(%rip) # ffffffff82008150 <ds.0> - 68: mov %rsi,%rax - mov %rdi,%rcx -100.00 mwait %eax,%ecx - 71: mov %gs:0x17bc0,%rax - lock andb $0xdf,0x2(%rax) - lock addl $0x0,-0x4(%rsp) - mov (%rax),%rax - test $0x8,%al + ↓ jmp 68 + verw WORD PTR [rip+0x538b08] # ffffffff82008150 <ds.0> + 68: mov rax,rsi + mov rcx,rdi +100.00 mwait + 71: mov rax,QWORD PTR gs:0x17bc0 + lock and BYTE PTR [rax+0x2],0xdf + lock add DWORD PTR [rsp-0x4],0x0 + mov rax,QWORD PTR [rax] + test al,0x8 ↓ je 97 - andl $0x7fffffff,__preempt_count - 97: ← retq - mov %gs:0x17bc0,%rax - lock orb $0x20,0x2(%rax) - mov (%rax),%rax - test $0x8,%al + and DWORD PTR gs:[rip+0x7e548509],0x7fffffff + 97: ret + mov rax,QWORD PTR gs:0x17bc0 + lock or BYTE PTR [rax+0x2],0x20 + mov rax,QWORD PTR [rax] + test al,0x8 ↑ jne 71 - ↑ jmpq 2d + ↑ jmp 2d # Requested-by: Matt P. Dziubinski <matdzb@gmail.com> Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Ian Rogers <irogers@google.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2020-09-04 19:10:43 +03:00
} else if (!strcmp(var, "annotate.disassembler_style")) {
opt->disassembler_style = value;
} else if (!strcmp(var, "annotate.demangle")) {
symbol_conf.demangle = perf_config_bool("demangle", value);
} else if (!strcmp(var, "annotate.demangle_kernel")) {
symbol_conf.demangle_kernel = perf_config_bool("demangle_kernel", value);
perf annotate: Allow setting the offset level in .perfconfig The default is 1 (jump_target): # perf annotate --ignore-vmlinux --stdio2 _raw_spin_lock_irqsave Samples: 3K of event 'cycles:ppp', 3000 Hz, Event count (approx.): 2766398574 _raw_spin_lock_irqsave() /proc/kcore 0.26 nop 4.61 push %rbx 19.33 pushfq 7.97 pop %rax 0.32 nop 0.06 mov %rax,%rbx 14.63 cli 0.06 nop xor %eax,%eax mov $0x1,%edx 49.94 lock cmpxchg %edx,(%rdi) 0.16 test %eax,%eax ↓ jne 2b 2.66 mov %rbx,%rax pop %rbx ← retq 2b: mov %eax,%esi → callq *ffffffffb30eaed0 mov %rbx,%rax pop %rbx ← retq # But one can ask for showing offsets for call instructions by setting this: # perf annotate --ignore-vmlinux --stdio2 _raw_spin_lock_irqsave Samples: 3K of event 'cycles:ppp', 3000 Hz, Event count (approx.): 2766398574 _raw_spin_lock_irqsave() /proc/kcore 0.26 nop 4.61 push %rbx 19.33 pushfq 7.97 pop %rax 0.32 nop 0.06 mov %rax,%rbx 14.63 cli 0.06 nop xor %eax,%eax mov $0x1,%edx 49.94 lock cmpxchg %edx,(%rdi) 0.16 test %eax,%eax ↓ jne 2b 2.66 mov %rbx,%rax pop %rbx ← retq 2b: mov %eax,%esi 2d: → callq *ffffffffb30eaed0 mov %rbx,%rax pop %rbx ← retq # Or using a big value to ask for all offsets to be shown: # cat ~/.perfconfig [annotate] offset_level = 100 hide_src_code = true # perf annotate --ignore-vmlinux --stdio2 _raw_spin_lock_irqsave Samples: 3K of event 'cycles:ppp', 3000 Hz, Event count (approx.): 2766398574 _raw_spin_lock_irqsave() /proc/kcore 0.26 0: nop 4.61 5: push %rbx 19.33 6: pushfq 7.97 7: pop %rax 0.32 8: nop 0.06 d: mov %rax,%rbx 14.63 10: cli 0.06 11: nop 17: xor %eax,%eax 19: mov $0x1,%edx 49.94 1e: lock cmpxchg %edx,(%rdi) 0.16 22: test %eax,%eax 24: ↓ jne 2b 2.66 26: mov %rbx,%rax 29: pop %rbx 2a: ← retq 2b: mov %eax,%esi 2d: → callq *ffffffffb30eaed0 32: mov %rbx,%rax 35: pop %rbx 36: ← retq # This also affects the TUI, i.e. the default 'perf annotate' and 'perf top/report' -> A hotkey -> annotate interfaces, when slang-devel is present in the build, i.e.: # perf version --build-options | grep slang libslang: [ on ] # HAVE_SLANG_SUPPORT # Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Andi Kleen <ak@linux.intel.com> Cc: David Ahern <dsahern@gmail.com> Cc: Jin Yao <yao.jin@linux.intel.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Martin Liška <mliska@suse.cz> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Ravi Bangoria <ravi.bangoria@linux.vnet.ibm.com> Cc: Thomas Richter <tmricht@linux.vnet.ibm.com> Cc: Wang Nan <wangnan0@huawei.com> Link: https://lkml.kernel.org/n/tip-venm6x5zrt40eu8hxdsmqxz6@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2018-04-12 21:23:02 +03:00
} else {
perf annotate: Make perf config effective perf default config set by user in [annotate] section is totally ignored by annotate code. Fix it. Before: $ ./perf config annotate.hide_src_code=true annotate.show_nr_jumps=true annotate.show_nr_samples=true $ ./perf annotate shash │ unsigned h = 0; │ movl $0x0,-0xc(%rbp) │ while (*s) │ ↓ jmp 44 │ h = 65599 * h + *s++; 11.33 │24: mov -0xc(%rbp),%eax 43.50 │ imul $0x1003f,%eax,%ecx │ mov -0x18(%rbp),%rax After: │ movl $0x0,-0xc(%rbp) │ ↓ jmp 44 1 │1 24: mov -0xc(%rbp),%eax 4 │ imul $0x1003f,%eax,%ecx │ mov -0x18(%rbp),%rax Note that we have removed show_nr_samples and show_total_period from annotation_options because they are not used. Instead of them we use symbol_conf.show_nr_samples and symbol_conf.show_total_period. Committer testing: Using 'perf annotate --stdio2' to use the TUI rendering but emitting the output to stdio: # perf config # # perf config annotate.hide_src_code=true # perf config annotate.hide_src_code=true # # perf config annotate.show_nr_jumps=true # perf config annotate.show_nr_samples=true # perf config annotate.hide_src_code=true annotate.show_nr_jumps=true annotate.show_nr_samples=true # # Before: # perf annotate --stdio2 ObjectInstance::weak_pointer_was_finalized Samples: 1 of event 'cycles', 4000 Hz, Event count (approx.): 830873, [percent: local period] ObjectInstance::weak_pointer_was_finalized() /usr/lib64/libgjs.so.0.0.0 Percent 00000000000609f0 <ObjectInstance::weak_pointer_was_finalized()@@Base>: endbr64 cmpq $0x0,0x20(%rdi) ↓ je 10 xor %eax,%eax ← retq xchg %ax,%ax 100.00 10: push %rbp cmpq $0x0,0x18(%rdi) mov %rdi,%rbp ↓ jne 20 1b: xor %eax,%eax pop %rbp ← retq nop 20: lea 0x18(%rdi),%rdi → callq JS_UpdateWeakPointerAfterGC(JS::Heap<JSObject* cmpq $0x0,0x18(%rbp) ↑ jne 1b mov %rbp,%rdi → callq ObjectBase::jsobj_addr() const@plt mov $0x1,%eax pop %rbp ← retq # After: # perf annotate --stdio2 ObjectInstance::weak_pointer_was_finalized 2> /dev/null Samples: 1 of event 'cycles', 4000 Hz, Event count (approx.): 830873, [percent: local period] ObjectInstance::weak_pointer_was_finalized() /usr/lib64/libgjs.so.0.0.0 Samples endbr64 cmpq $0x0,0x20(%rdi) ↓ je 10 xor %eax,%eax ← retq xchg %ax,%ax 1 1 10: push %rbp cmpq $0x0,0x18(%rdi) mov %rdi,%rbp ↓ jne 20 1 1b: xor %eax,%eax pop %rbp ← retq nop 1 20: lea 0x18(%rdi),%rdi → callq JS_UpdateWeakPointerAfterGC(JS::Heap<JSObject* cmpq $0x0,0x18(%rbp) ↑ jne 1b mov %rbp,%rdi → callq ObjectBase::jsobj_addr() const@plt mov $0x1,%eax pop %rbp ← retq # # perf config annotate.show_nr_jumps annotate.show_nr_jumps=true # perf config annotate.show_nr_jumps=false # perf config annotate.show_nr_jumps annotate.show_nr_jumps=false # # perf annotate --stdio2 ObjectInstance::weak_pointer_was_finalized 2> /dev/null Samples: 1 of event 'cycles', 4000 Hz, Event count (approx.): 830873, [percent: local period] ObjectInstance::weak_pointer_was_finalized() /usr/lib64/libgjs.so.0.0.0 Samples endbr64 cmpq $0x0,0x20(%rdi) ↓ je 10 xor %eax,%eax ← retq xchg %ax,%ax 1 10: push %rbp cmpq $0x0,0x18(%rdi) mov %rdi,%rbp ↓ jne 20 1b: xor %eax,%eax pop %rbp ← retq nop 20: lea 0x18(%rdi),%rdi → callq JS_UpdateWeakPointerAfterGC(JS::Heap<JSObject* cmpq $0x0,0x18(%rbp) ↑ jne 1b mov %rbp,%rdi → callq ObjectBase::jsobj_addr() const@plt mov $0x1,%eax pop %rbp ← retq # Signed-off-by: Ravi Bangoria <ravi.bangoria@linux.ibm.com> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Alexey Budankov <alexey.budankov@linux.intel.com> Cc: Changbin Du <changbin.du@intel.com> Cc: Ian Rogers <irogers@google.com> Cc: Jin Yao <yao.jin@linux.intel.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Leo Yan <leo.yan@linaro.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Song Liu <songliubraving@fb.com> Cc: Taeung Song <treeze.taeung@gmail.com> Cc: Thomas Richter <tmricht@linux.ibm.com> Cc: Yisheng Xie <xieyisheng1@huawei.com> Link: http://lore.kernel.org/lkml/20200213064306.160480-6-ravi.bangoria@linux.ibm.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2020-02-13 09:43:03 +03:00
pr_debug("%s variable unknown, ignoring...", var);
perf annotate: Allow setting the offset level in .perfconfig The default is 1 (jump_target): # perf annotate --ignore-vmlinux --stdio2 _raw_spin_lock_irqsave Samples: 3K of event 'cycles:ppp', 3000 Hz, Event count (approx.): 2766398574 _raw_spin_lock_irqsave() /proc/kcore 0.26 nop 4.61 push %rbx 19.33 pushfq 7.97 pop %rax 0.32 nop 0.06 mov %rax,%rbx 14.63 cli 0.06 nop xor %eax,%eax mov $0x1,%edx 49.94 lock cmpxchg %edx,(%rdi) 0.16 test %eax,%eax ↓ jne 2b 2.66 mov %rbx,%rax pop %rbx ← retq 2b: mov %eax,%esi → callq *ffffffffb30eaed0 mov %rbx,%rax pop %rbx ← retq # But one can ask for showing offsets for call instructions by setting this: # perf annotate --ignore-vmlinux --stdio2 _raw_spin_lock_irqsave Samples: 3K of event 'cycles:ppp', 3000 Hz, Event count (approx.): 2766398574 _raw_spin_lock_irqsave() /proc/kcore 0.26 nop 4.61 push %rbx 19.33 pushfq 7.97 pop %rax 0.32 nop 0.06 mov %rax,%rbx 14.63 cli 0.06 nop xor %eax,%eax mov $0x1,%edx 49.94 lock cmpxchg %edx,(%rdi) 0.16 test %eax,%eax ↓ jne 2b 2.66 mov %rbx,%rax pop %rbx ← retq 2b: mov %eax,%esi 2d: → callq *ffffffffb30eaed0 mov %rbx,%rax pop %rbx ← retq # Or using a big value to ask for all offsets to be shown: # cat ~/.perfconfig [annotate] offset_level = 100 hide_src_code = true # perf annotate --ignore-vmlinux --stdio2 _raw_spin_lock_irqsave Samples: 3K of event 'cycles:ppp', 3000 Hz, Event count (approx.): 2766398574 _raw_spin_lock_irqsave() /proc/kcore 0.26 0: nop 4.61 5: push %rbx 19.33 6: pushfq 7.97 7: pop %rax 0.32 8: nop 0.06 d: mov %rax,%rbx 14.63 10: cli 0.06 11: nop 17: xor %eax,%eax 19: mov $0x1,%edx 49.94 1e: lock cmpxchg %edx,(%rdi) 0.16 22: test %eax,%eax 24: ↓ jne 2b 2.66 26: mov %rbx,%rax 29: pop %rbx 2a: ← retq 2b: mov %eax,%esi 2d: → callq *ffffffffb30eaed0 32: mov %rbx,%rax 35: pop %rbx 36: ← retq # This also affects the TUI, i.e. the default 'perf annotate' and 'perf top/report' -> A hotkey -> annotate interfaces, when slang-devel is present in the build, i.e.: # perf version --build-options | grep slang libslang: [ on ] # HAVE_SLANG_SUPPORT # Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Andi Kleen <ak@linux.intel.com> Cc: David Ahern <dsahern@gmail.com> Cc: Jin Yao <yao.jin@linux.intel.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Martin Liška <mliska@suse.cz> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Ravi Bangoria <ravi.bangoria@linux.vnet.ibm.com> Cc: Thomas Richter <tmricht@linux.vnet.ibm.com> Cc: Wang Nan <wangnan0@huawei.com> Link: https://lkml.kernel.org/n/tip-venm6x5zrt40eu8hxdsmqxz6@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2018-04-12 21:23:02 +03:00
}
perf annotate: Make perf config effective perf default config set by user in [annotate] section is totally ignored by annotate code. Fix it. Before: $ ./perf config annotate.hide_src_code=true annotate.show_nr_jumps=true annotate.show_nr_samples=true $ ./perf annotate shash │ unsigned h = 0; │ movl $0x0,-0xc(%rbp) │ while (*s) │ ↓ jmp 44 │ h = 65599 * h + *s++; 11.33 │24: mov -0xc(%rbp),%eax 43.50 │ imul $0x1003f,%eax,%ecx │ mov -0x18(%rbp),%rax After: │ movl $0x0,-0xc(%rbp) │ ↓ jmp 44 1 │1 24: mov -0xc(%rbp),%eax 4 │ imul $0x1003f,%eax,%ecx │ mov -0x18(%rbp),%rax Note that we have removed show_nr_samples and show_total_period from annotation_options because they are not used. Instead of them we use symbol_conf.show_nr_samples and symbol_conf.show_total_period. Committer testing: Using 'perf annotate --stdio2' to use the TUI rendering but emitting the output to stdio: # perf config # # perf config annotate.hide_src_code=true # perf config annotate.hide_src_code=true # # perf config annotate.show_nr_jumps=true # perf config annotate.show_nr_samples=true # perf config annotate.hide_src_code=true annotate.show_nr_jumps=true annotate.show_nr_samples=true # # Before: # perf annotate --stdio2 ObjectInstance::weak_pointer_was_finalized Samples: 1 of event 'cycles', 4000 Hz, Event count (approx.): 830873, [percent: local period] ObjectInstance::weak_pointer_was_finalized() /usr/lib64/libgjs.so.0.0.0 Percent 00000000000609f0 <ObjectInstance::weak_pointer_was_finalized()@@Base>: endbr64 cmpq $0x0,0x20(%rdi) ↓ je 10 xor %eax,%eax ← retq xchg %ax,%ax 100.00 10: push %rbp cmpq $0x0,0x18(%rdi) mov %rdi,%rbp ↓ jne 20 1b: xor %eax,%eax pop %rbp ← retq nop 20: lea 0x18(%rdi),%rdi → callq JS_UpdateWeakPointerAfterGC(JS::Heap<JSObject* cmpq $0x0,0x18(%rbp) ↑ jne 1b mov %rbp,%rdi → callq ObjectBase::jsobj_addr() const@plt mov $0x1,%eax pop %rbp ← retq # After: # perf annotate --stdio2 ObjectInstance::weak_pointer_was_finalized 2> /dev/null Samples: 1 of event 'cycles', 4000 Hz, Event count (approx.): 830873, [percent: local period] ObjectInstance::weak_pointer_was_finalized() /usr/lib64/libgjs.so.0.0.0 Samples endbr64 cmpq $0x0,0x20(%rdi) ↓ je 10 xor %eax,%eax ← retq xchg %ax,%ax 1 1 10: push %rbp cmpq $0x0,0x18(%rdi) mov %rdi,%rbp ↓ jne 20 1 1b: xor %eax,%eax pop %rbp ← retq nop 1 20: lea 0x18(%rdi),%rdi → callq JS_UpdateWeakPointerAfterGC(JS::Heap<JSObject* cmpq $0x0,0x18(%rbp) ↑ jne 1b mov %rbp,%rdi → callq ObjectBase::jsobj_addr() const@plt mov $0x1,%eax pop %rbp ← retq # # perf config annotate.show_nr_jumps annotate.show_nr_jumps=true # perf config annotate.show_nr_jumps=false # perf config annotate.show_nr_jumps annotate.show_nr_jumps=false # # perf annotate --stdio2 ObjectInstance::weak_pointer_was_finalized 2> /dev/null Samples: 1 of event 'cycles', 4000 Hz, Event count (approx.): 830873, [percent: local period] ObjectInstance::weak_pointer_was_finalized() /usr/lib64/libgjs.so.0.0.0 Samples endbr64 cmpq $0x0,0x20(%rdi) ↓ je 10 xor %eax,%eax ← retq xchg %ax,%ax 1 10: push %rbp cmpq $0x0,0x18(%rdi) mov %rdi,%rbp ↓ jne 20 1b: xor %eax,%eax pop %rbp ← retq nop 20: lea 0x18(%rdi),%rdi → callq JS_UpdateWeakPointerAfterGC(JS::Heap<JSObject* cmpq $0x0,0x18(%rbp) ↑ jne 1b mov %rbp,%rdi → callq ObjectBase::jsobj_addr() const@plt mov $0x1,%eax pop %rbp ← retq # Signed-off-by: Ravi Bangoria <ravi.bangoria@linux.ibm.com> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Alexey Budankov <alexey.budankov@linux.intel.com> Cc: Changbin Du <changbin.du@intel.com> Cc: Ian Rogers <irogers@google.com> Cc: Jin Yao <yao.jin@linux.intel.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Leo Yan <leo.yan@linaro.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Song Liu <songliubraving@fb.com> Cc: Taeung Song <treeze.taeung@gmail.com> Cc: Thomas Richter <tmricht@linux.ibm.com> Cc: Yisheng Xie <xieyisheng1@huawei.com> Link: http://lore.kernel.org/lkml/20200213064306.160480-6-ravi.bangoria@linux.ibm.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2020-02-13 09:43:03 +03:00
return 0;
}
perf annotate: Make perf config effective perf default config set by user in [annotate] section is totally ignored by annotate code. Fix it. Before: $ ./perf config annotate.hide_src_code=true annotate.show_nr_jumps=true annotate.show_nr_samples=true $ ./perf annotate shash │ unsigned h = 0; │ movl $0x0,-0xc(%rbp) │ while (*s) │ ↓ jmp 44 │ h = 65599 * h + *s++; 11.33 │24: mov -0xc(%rbp),%eax 43.50 │ imul $0x1003f,%eax,%ecx │ mov -0x18(%rbp),%rax After: │ movl $0x0,-0xc(%rbp) │ ↓ jmp 44 1 │1 24: mov -0xc(%rbp),%eax 4 │ imul $0x1003f,%eax,%ecx │ mov -0x18(%rbp),%rax Note that we have removed show_nr_samples and show_total_period from annotation_options because they are not used. Instead of them we use symbol_conf.show_nr_samples and symbol_conf.show_total_period. Committer testing: Using 'perf annotate --stdio2' to use the TUI rendering but emitting the output to stdio: # perf config # # perf config annotate.hide_src_code=true # perf config annotate.hide_src_code=true # # perf config annotate.show_nr_jumps=true # perf config annotate.show_nr_samples=true # perf config annotate.hide_src_code=true annotate.show_nr_jumps=true annotate.show_nr_samples=true # # Before: # perf annotate --stdio2 ObjectInstance::weak_pointer_was_finalized Samples: 1 of event 'cycles', 4000 Hz, Event count (approx.): 830873, [percent: local period] ObjectInstance::weak_pointer_was_finalized() /usr/lib64/libgjs.so.0.0.0 Percent 00000000000609f0 <ObjectInstance::weak_pointer_was_finalized()@@Base>: endbr64 cmpq $0x0,0x20(%rdi) ↓ je 10 xor %eax,%eax ← retq xchg %ax,%ax 100.00 10: push %rbp cmpq $0x0,0x18(%rdi) mov %rdi,%rbp ↓ jne 20 1b: xor %eax,%eax pop %rbp ← retq nop 20: lea 0x18(%rdi),%rdi → callq JS_UpdateWeakPointerAfterGC(JS::Heap<JSObject* cmpq $0x0,0x18(%rbp) ↑ jne 1b mov %rbp,%rdi → callq ObjectBase::jsobj_addr() const@plt mov $0x1,%eax pop %rbp ← retq # After: # perf annotate --stdio2 ObjectInstance::weak_pointer_was_finalized 2> /dev/null Samples: 1 of event 'cycles', 4000 Hz, Event count (approx.): 830873, [percent: local period] ObjectInstance::weak_pointer_was_finalized() /usr/lib64/libgjs.so.0.0.0 Samples endbr64 cmpq $0x0,0x20(%rdi) ↓ je 10 xor %eax,%eax ← retq xchg %ax,%ax 1 1 10: push %rbp cmpq $0x0,0x18(%rdi) mov %rdi,%rbp ↓ jne 20 1 1b: xor %eax,%eax pop %rbp ← retq nop 1 20: lea 0x18(%rdi),%rdi → callq JS_UpdateWeakPointerAfterGC(JS::Heap<JSObject* cmpq $0x0,0x18(%rbp) ↑ jne 1b mov %rbp,%rdi → callq ObjectBase::jsobj_addr() const@plt mov $0x1,%eax pop %rbp ← retq # # perf config annotate.show_nr_jumps annotate.show_nr_jumps=true # perf config annotate.show_nr_jumps=false # perf config annotate.show_nr_jumps annotate.show_nr_jumps=false # # perf annotate --stdio2 ObjectInstance::weak_pointer_was_finalized 2> /dev/null Samples: 1 of event 'cycles', 4000 Hz, Event count (approx.): 830873, [percent: local period] ObjectInstance::weak_pointer_was_finalized() /usr/lib64/libgjs.so.0.0.0 Samples endbr64 cmpq $0x0,0x20(%rdi) ↓ je 10 xor %eax,%eax ← retq xchg %ax,%ax 1 10: push %rbp cmpq $0x0,0x18(%rdi) mov %rdi,%rbp ↓ jne 20 1b: xor %eax,%eax pop %rbp ← retq nop 20: lea 0x18(%rdi),%rdi → callq JS_UpdateWeakPointerAfterGC(JS::Heap<JSObject* cmpq $0x0,0x18(%rbp) ↑ jne 1b mov %rbp,%rdi → callq ObjectBase::jsobj_addr() const@plt mov $0x1,%eax pop %rbp ← retq # Signed-off-by: Ravi Bangoria <ravi.bangoria@linux.ibm.com> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Alexey Budankov <alexey.budankov@linux.intel.com> Cc: Changbin Du <changbin.du@intel.com> Cc: Ian Rogers <irogers@google.com> Cc: Jin Yao <yao.jin@linux.intel.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Leo Yan <leo.yan@linaro.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Song Liu <songliubraving@fb.com> Cc: Taeung Song <treeze.taeung@gmail.com> Cc: Thomas Richter <tmricht@linux.ibm.com> Cc: Yisheng Xie <xieyisheng1@huawei.com> Link: http://lore.kernel.org/lkml/20200213064306.160480-6-ravi.bangoria@linux.ibm.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2020-02-13 09:43:03 +03:00
void annotation_config__init(struct annotation_options *opt)
{
perf annotate: Make perf config effective perf default config set by user in [annotate] section is totally ignored by annotate code. Fix it. Before: $ ./perf config annotate.hide_src_code=true annotate.show_nr_jumps=true annotate.show_nr_samples=true $ ./perf annotate shash │ unsigned h = 0; │ movl $0x0,-0xc(%rbp) │ while (*s) │ ↓ jmp 44 │ h = 65599 * h + *s++; 11.33 │24: mov -0xc(%rbp),%eax 43.50 │ imul $0x1003f,%eax,%ecx │ mov -0x18(%rbp),%rax After: │ movl $0x0,-0xc(%rbp) │ ↓ jmp 44 1 │1 24: mov -0xc(%rbp),%eax 4 │ imul $0x1003f,%eax,%ecx │ mov -0x18(%rbp),%rax Note that we have removed show_nr_samples and show_total_period from annotation_options because they are not used. Instead of them we use symbol_conf.show_nr_samples and symbol_conf.show_total_period. Committer testing: Using 'perf annotate --stdio2' to use the TUI rendering but emitting the output to stdio: # perf config # # perf config annotate.hide_src_code=true # perf config annotate.hide_src_code=true # # perf config annotate.show_nr_jumps=true # perf config annotate.show_nr_samples=true # perf config annotate.hide_src_code=true annotate.show_nr_jumps=true annotate.show_nr_samples=true # # Before: # perf annotate --stdio2 ObjectInstance::weak_pointer_was_finalized Samples: 1 of event 'cycles', 4000 Hz, Event count (approx.): 830873, [percent: local period] ObjectInstance::weak_pointer_was_finalized() /usr/lib64/libgjs.so.0.0.0 Percent 00000000000609f0 <ObjectInstance::weak_pointer_was_finalized()@@Base>: endbr64 cmpq $0x0,0x20(%rdi) ↓ je 10 xor %eax,%eax ← retq xchg %ax,%ax 100.00 10: push %rbp cmpq $0x0,0x18(%rdi) mov %rdi,%rbp ↓ jne 20 1b: xor %eax,%eax pop %rbp ← retq nop 20: lea 0x18(%rdi),%rdi → callq JS_UpdateWeakPointerAfterGC(JS::Heap<JSObject* cmpq $0x0,0x18(%rbp) ↑ jne 1b mov %rbp,%rdi → callq ObjectBase::jsobj_addr() const@plt mov $0x1,%eax pop %rbp ← retq # After: # perf annotate --stdio2 ObjectInstance::weak_pointer_was_finalized 2> /dev/null Samples: 1 of event 'cycles', 4000 Hz, Event count (approx.): 830873, [percent: local period] ObjectInstance::weak_pointer_was_finalized() /usr/lib64/libgjs.so.0.0.0 Samples endbr64 cmpq $0x0,0x20(%rdi) ↓ je 10 xor %eax,%eax ← retq xchg %ax,%ax 1 1 10: push %rbp cmpq $0x0,0x18(%rdi) mov %rdi,%rbp ↓ jne 20 1 1b: xor %eax,%eax pop %rbp ← retq nop 1 20: lea 0x18(%rdi),%rdi → callq JS_UpdateWeakPointerAfterGC(JS::Heap<JSObject* cmpq $0x0,0x18(%rbp) ↑ jne 1b mov %rbp,%rdi → callq ObjectBase::jsobj_addr() const@plt mov $0x1,%eax pop %rbp ← retq # # perf config annotate.show_nr_jumps annotate.show_nr_jumps=true # perf config annotate.show_nr_jumps=false # perf config annotate.show_nr_jumps annotate.show_nr_jumps=false # # perf annotate --stdio2 ObjectInstance::weak_pointer_was_finalized 2> /dev/null Samples: 1 of event 'cycles', 4000 Hz, Event count (approx.): 830873, [percent: local period] ObjectInstance::weak_pointer_was_finalized() /usr/lib64/libgjs.so.0.0.0 Samples endbr64 cmpq $0x0,0x20(%rdi) ↓ je 10 xor %eax,%eax ← retq xchg %ax,%ax 1 10: push %rbp cmpq $0x0,0x18(%rdi) mov %rdi,%rbp ↓ jne 20 1b: xor %eax,%eax pop %rbp ← retq nop 20: lea 0x18(%rdi),%rdi → callq JS_UpdateWeakPointerAfterGC(JS::Heap<JSObject* cmpq $0x0,0x18(%rbp) ↑ jne 1b mov %rbp,%rdi → callq ObjectBase::jsobj_addr() const@plt mov $0x1,%eax pop %rbp ← retq # Signed-off-by: Ravi Bangoria <ravi.bangoria@linux.ibm.com> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Alexey Budankov <alexey.budankov@linux.intel.com> Cc: Changbin Du <changbin.du@intel.com> Cc: Ian Rogers <irogers@google.com> Cc: Jin Yao <yao.jin@linux.intel.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Leo Yan <leo.yan@linaro.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Song Liu <songliubraving@fb.com> Cc: Taeung Song <treeze.taeung@gmail.com> Cc: Thomas Richter <tmricht@linux.ibm.com> Cc: Yisheng Xie <xieyisheng1@huawei.com> Link: http://lore.kernel.org/lkml/20200213064306.160480-6-ravi.bangoria@linux.ibm.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
2020-02-13 09:43:03 +03:00
perf_config(annotation__config, opt);
}
static unsigned int parse_percent_type(char *str1, char *str2)
{
unsigned int type = (unsigned int) -1;
if (!strcmp("period", str1)) {
if (!strcmp("local", str2))
type = PERCENT_PERIOD_LOCAL;
else if (!strcmp("global", str2))
type = PERCENT_PERIOD_GLOBAL;
}
if (!strcmp("hits", str1)) {
if (!strcmp("local", str2))
type = PERCENT_HITS_LOCAL;
else if (!strcmp("global", str2))
type = PERCENT_HITS_GLOBAL;
}
return type;
}
int annotate_parse_percent_type(const struct option *opt, const char *_str,
int unset __maybe_unused)
{
struct annotation_options *opts = opt->value;
unsigned int type;
char *str1, *str2;
int err = -1;
str1 = strdup(_str);
if (!str1)
return -ENOMEM;
str2 = strchr(str1, '-');
if (!str2)
goto out;
*str2++ = 0;
type = parse_percent_type(str1, str2);
if (type == (unsigned int) -1)
type = parse_percent_type(str2, str1);
if (type != (unsigned int) -1) {
opts->percent_type = type;
err = 0;
}
out:
free(str1);
return err;
}
int annotate_check_args(struct annotation_options *args)
{
if (args->prefix_strip && !args->prefix) {
pr_err("--prefix-strip requires --prefix\n");
return -1;
}
return 0;
}