linux-toolchains.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Namhyung Kim <namhyung@kernel.org>
To: Arnaldo Carvalho de Melo <acme@kernel.org>
Cc: Jiri Olsa <jolsa@kernel.org>,
	Peter Zijlstra <peterz@infradead.org>,
	 Ian Rogers <irogers@google.com>,
	Adrian Hunter <adrian.hunter@intel.com>,
	 Ingo Molnar <mingo@kernel.org>,
	LKML <linux-kernel@vger.kernel.org>,
	 linux-perf-users@vger.kernel.org,
	 Linus Torvalds <torvalds@linux-foundation.org>,
	Stephane Eranian <eranian@google.com>,
	 Masami Hiramatsu <mhiramat@kernel.org>,
	Andi Kleen <ak@linux.intel.com>,
	 linux-trace-devel@vger.kernel.org,
	linux-toolchains@vger.kernel.org
Subject: Re: [PATCH 18/52] perf report: Add 'type' sort key
Date: Wed, 22 Nov 2023 10:49:13 -0800	[thread overview]
Message-ID: <CAM9d7cijprv2fOM_jwYbrV0jbhno2ZQDCT3SExG=Ot=b2mANmw@mail.gmail.com> (raw)
In-Reply-To: <ZVzvFen/214ylf32@kernel.org>

On Tue, Nov 21, 2023 at 9:55 AM Arnaldo Carvalho de Melo
<acme@kernel.org> wrote:
>
> Em Thu, Nov 09, 2023 at 03:59:37PM -0800, Namhyung Kim escreveu:
> > The 'type' sort key is to aggregate hist entries by data type they
> > access.  Add mem_type field to hist_entry struct to save the type.
> > If hist_entry__get_data_type() returns NULL, it'd use the
> > 'unknown_type' instance.
>
> I built up to here and then tried on a random perf.data file:
>
> ⬢[acme@toolbox perf-tools-next]$ perf evlist
> cycles:Pu
> ⬢[acme@toolbox perf-tools-next]$ perf evlist -v
> cycles:Pu: type: 0 (PERF_TYPE_HARDWARE), size: 136, config: 0 (PERF_COUNT_HW_CPU_CYCLES), { sample_period, sample_freq }: 4000, sample_type: IP|TID|TIME|PERIOD, read_format: ID|LOST, disabled: 1, inherit: 1, exclude_kernel: 1, exclude_hv: 1, mmap: 1, comm: 1, freq: 1, enable_on_exec: 1, task: 1, sample_id_all: 1, exclude_guest: 1, mmap2: 1, comm_exec: 1, ksymbol: 1, bpf_event: 1
> ⬢[acme@toolbox perf-tools-next]$
>
> And got:
>
> ⬢[acme@toolbox perf-tools-next]$ perf report -s type
> perf: Segmentation fault
> -------- backtrace --------
> perf[0x69f743]
> /lib64/libc.so.6(+0x3dbb0)[0x7f89b4778bb0]
> perf[0x505af6]
> perf[0x512d47]
> perf[0x512f82]
> perf[0x5b3461]
> perf[0x5b3516]
> perf[0x5b3a3e]
> perf[0x5bbb05]
> perf[0x5bc68f]
> perf[0x5bca7c]
> perf[0x42ead1]
> perf[0x42fa08]
> perf[0x43200d]
> perf[0x504856]
> perf[0x504ac5]
> perf[0x504c14]
> perf[0x504f01]
> /lib64/libc.so.6(+0x27b8a)[0x7f89b4762b8a]
> /lib64/libc.so.6(__libc_start_main+0x8b)[0x7f89b4762c4b]
> perf[0x40ed65]
> ⬢[acme@toolbox perf-tools-next]$

Right, the 'type' sort key was added here but unfortunately
it's not ready for prime time yet.  It also needs the next patch
19/52 ("perf report: Support data type profiling") to fully enable
the feature.  Do you think it's better to squash into here?

Thanks,
Namhyung

>
> Using gdb:
>
> (gdb) run report --stdio -s type
> Starting program: /home/acme/bin/perf report --stdio -s type
>
> This GDB supports auto-downloading debuginfo from the following URLs:
>   <https://debuginfod.fedoraproject.org/>
> Enable debuginfod for this session? (y or [n]) y
> Debuginfod has been enabled.
> To make this setting permanent, add 'set debuginfod enabled on' to .gdbinit.
> [Thread debugging using libthread_db enabled]
> Using host libthread_db library "/lib64/libthread_db.so.1".
> [Detaching after fork from child process 811109]
>
> Program received signal SIGSEGV, Segmentation fault.
> 0x0000000000505af6 in list_empty (head=0x14c20) at /home/acme/git/perf-tools-next/tools/include/linux/list.h:189
> 189             return head->next == head;
> (gdb) bt
> #0  0x0000000000505af6 in list_empty (head=0x14c20) at /home/acme/git/perf-tools-next/tools/include/linux/list.h:189
> #1  0x0000000000512d47 in symbol__ensure_annotate (ms=0xe6f258, evsel=0xe276f0) at util/annotate.c:3640
> #2  0x0000000000512f82 in hist_entry__get_data_type (he=0xe6f1e0) at util/annotate.c:3696
> #3  0x00000000005b3461 in sort__type_init (he=0xe6f1e0) at util/sort.c:2152
> #4  0x00000000005b3516 in sort__type_collapse (left=0xe6ed80, right=0xe6f1e0) at util/sort.c:2169
> #5  0x00000000005b3a3e in __sort__hpp_collapse (fmt=0xe448f0, a=0xe6ed80, b=0xe6f1e0) at util/sort.c:2394
> #6  0x00000000005bbb05 in hist_entry__collapse (left=0xe6ed80, right=0xe6f1e0) at util/hist.c:1306
> #7  0x00000000005bc68f in hists__collapse_insert_entry (hists=0xe27960, root=0xe27998, he=0xe6f1e0) at util/hist.c:1613
> #8  0x00000000005bca7c in hists__collapse_resort (hists=0xe27960, prog=0x7fffffffb820) at util/hist.c:1697
> #9  0x000000000042ead1 in report__collapse_hists (rep=0x7fffffffbac0) at builtin-report.c:723
> #10 0x000000000042fa08 in __cmd_report (rep=0x7fffffffbac0) at builtin-report.c:1042
> #11 0x000000000043200d in cmd_report (argc=0, argv=0x7fffffffe1b0) at builtin-report.c:1733
> #12 0x0000000000504856 in run_builtin (p=0xdf7da0 <commands+288>, argc=4, argv=0x7fffffffe1b0) at perf.c:322
> #13 0x0000000000504ac5 in handle_internal_command (argc=4, argv=0x7fffffffe1b0) at perf.c:375
> #14 0x0000000000504c14 in run_argv (argcp=0x7fffffffdfcc, argv=0x7fffffffdfc0) at perf.c:419
> #15 0x0000000000504f01 in main (argc=4, argv=0x7fffffffe1b0) at perf.c:535
> (gdb)
>
>
> static void symbol__ensure_annotate(struct map_symbol *ms, struct evsel *evsel)
> +{
> +       struct disasm_line *dl, *tmp_dl;
> +       struct annotation *notes;
> +
> +       notes = symbol__annotation(ms->sym);
> +       if (!list_empty(&notes->src->source))
> +               return;
> +
> +       if (symbol__annotate(ms, evsel, notes->options, NULL) < 0)
> +               return;
> +
> +       /* remove non-insn disasm lines for simplicity */
> +       list_for_each_entry_safe(dl, tmp_dl, &notes->src->source, al.node) {
> +               if (dl->al.offset == -1) {
> +                       list_del(&dl->al.node);
> +                       free(dl);
> +               }
> +       }
> +}
>
> Probably annotated_source__new() wasn't called? Yeah, seems so:
>
> (gdb) b annotated_source__new
> Breakpoint 1 at 0x50a894: file util/annotate.c, line 851.
> (gdb) run report --stdio -s type
> Starting program: /home/acme/bin/perf report --stdio -s type
>
> This GDB supports auto-downloading debuginfo from the following URLs:
>   <https://debuginfod.fedoraproject.org/>
> Enable debuginfod for this session? (y or [n]) y
> Debuginfod has been enabled.
> To make this setting permanent, add 'set debuginfod enabled on' to .gdbinit.
> [Thread debugging using libthread_db enabled]
> Using host libthread_db library "/lib64/libthread_db.so.1".
> [Detaching after fork from child process 818292]
>
> Program received signal SIGSEGV, Segmentation fault.
> 0x0000000000505af6 in list_empty (head=0x14c20) at /home/acme/git/perf-tools-next/tools/include/linux/list.h:189
> 189             return head->next == head;
> (gdb)

  reply	other threads:[~2023-11-22 18:49 UTC|newest]

Thread overview: 69+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2023-11-09 23:59 [RFC 00/52] perf tools: Introduce data type profiling (v2) Namhyung Kim
2023-11-09 23:59 ` [PATCH 01/52] perf annotate: Pass "-l" option to objdump conditionally Namhyung Kim
2023-11-09 23:59 ` [PATCH 02/52] perf annotate: Move raw_comment and raw_func_start Namhyung Kim
2023-11-09 23:59 ` [PATCH 03/52] perf tools: Add util/debuginfo.[ch] files Namhyung Kim
2023-11-09 23:59 ` [PATCH 04/52] perf dwarf-aux: Fix die_get_typename() for void * Namhyung Kim
2023-11-09 23:59 ` [PATCH 05/52] perf dwarf-aux: Move #ifdef code to the header file Namhyung Kim
2023-11-09 23:59 ` [PATCH 06/52] perf dwarf-aux: Add die_get_scopes() helper Namhyung Kim
2023-11-09 23:59 ` [PATCH 07/52] perf dwarf-aux: Add die_find_variable_by_reg() helper Namhyung Kim
2023-11-09 23:59 ` [PATCH 08/52] perf build: Add feature check for dwarf_getcfi() Namhyung Kim
2023-11-10 10:26   ` Masami Hiramatsu
2023-11-09 23:59 ` [PATCH 09/52] perf probe: Convert to check dwarf_getcfi feature Namhyung Kim
2023-11-10 10:25   ` Masami Hiramatsu
2023-11-09 23:59 ` [PATCH 10/52] perf dwarf-aux: Factor out die_get_typename_from_type() Namhyung Kim
2023-11-09 23:59 ` [PATCH 11/52] perf dwarf-regs: Add get_dwarf_regnum() Namhyung Kim
2023-11-09 23:59 ` [PATCH 12/52] perf annotate-data: Add find_data_type() Namhyung Kim
     [not found]   ` <CA+JHD90fkWNrQWO5DrHeV8mCmFyKKqJ8fV=KwztRi7TSw+8yDg@mail.gmail.com>
2023-11-20 20:43     ` Namhyung Kim
2023-11-09 23:59 ` [PATCH 13/52] perf annotate-data: Add dso->data_types tree Namhyung Kim
2023-12-21 20:10   ` Arnaldo Carvalho de Melo
2023-12-21 20:13     ` Arnaldo Carvalho de Melo
2023-12-21 20:32       ` Arnaldo Carvalho de Melo
2023-11-09 23:59 ` [PATCH 14/52] perf annotate: Factor out evsel__get_arch() Namhyung Kim
2023-12-23 14:14   ` Arnaldo Carvalho de Melo
2023-11-09 23:59 ` [PATCH 15/52] perf annotate: Check if operand has multiple regs Namhyung Kim
2023-11-09 23:59 ` [PATCH 16/52] perf annotate: Add annotate_get_insn_location() Namhyung Kim
2023-11-09 23:59 ` [PATCH 17/52] perf annotate: Implement hist_entry__get_data_type() Namhyung Kim
2023-11-09 23:59 ` [PATCH 18/52] perf report: Add 'type' sort key Namhyung Kim
2023-11-21 17:55   ` Arnaldo Carvalho de Melo
2023-11-22 18:49     ` Namhyung Kim [this message]
2023-11-22 19:54       ` Arnaldo Carvalho de Melo
2023-11-22 21:13         ` Namhyung Kim
2023-11-23 13:40           ` Arnaldo Carvalho de Melo
2023-11-09 23:59 ` [PATCH 19/52] perf report: Support data type profiling Namhyung Kim
2023-11-09 23:59 ` [PATCH 20/52] perf annotate-data: Add member field in the data type Namhyung Kim
2023-11-09 23:59 ` [PATCH 21/52] perf annotate-data: Update sample histogram for type Namhyung Kim
2023-11-09 23:59 ` [PATCH 22/52] perf report: Add 'typeoff' sort key Namhyung Kim
2023-11-09 23:59 ` [PATCH 23/52] perf report: Add 'symoff' " Namhyung Kim
2023-12-23 14:29   ` Arnaldo Carvalho de Melo
2023-11-09 23:59 ` [PATCH 24/52] perf annotate: Add --data-type option Namhyung Kim
2023-11-09 23:59 ` [PATCH 25/52] perf annotate: Support event group display Namhyung Kim
2023-11-09 23:59 ` [PATCH 26/52] perf annotate: Add --type-stat option for debugging Namhyung Kim
2023-11-09 23:59 ` [PATCH 27/52] perf annotate: Add --insn-stat " Namhyung Kim
2023-11-09 23:59 ` [PATCH 28/52] perf annotate-data: Parse 'lock' prefix from llvm-objdump Namhyung Kim
2023-11-09 23:59 ` [PATCH 29/52] perf annotate-data: Handle macro fusion on x86 Namhyung Kim
2023-11-09 23:59 ` [PATCH 30/52] perf annotate-data: Handle array style accesses Namhyung Kim
2023-11-09 23:59 ` [PATCH 31/52] perf annotate-data: Add stack operation pseudo type Namhyung Kim
2023-11-09 23:59 ` [PATCH 32/52] perf dwarf-aux: Add die_find_variable_by_addr() Namhyung Kim
2023-11-27 22:07   ` Arnaldo Carvalho de Melo
2023-11-09 23:59 ` [PATCH 33/52] perf annotate-data: Handle PC-relative addressing Namhyung Kim
2023-11-09 23:59 ` [PATCH 34/52] perf annotate-data: Support global variables Namhyung Kim
2023-11-09 23:59 ` [PATCH 35/52] perf dwarf-aux: Add die_get_cfa() Namhyung Kim
2023-11-09 23:59 ` [PATCH 36/52] perf annotate-data: Support stack variables Namhyung Kim
2023-11-09 23:59 ` [PATCH 37/52] perf dwarf-aux: Check allowed DWARF Ops Namhyung Kim
2023-11-09 23:59 ` [PATCH 38/52] perf dwarf-aux: Add die_collect_vars() Namhyung Kim
2023-11-09 23:59 ` [PATCH 39/52] perf dwarf-aux: Handle type transfer for memory access Namhyung Kim
2023-11-09 23:59 ` [PATCH 40/52] perf annotate-data: Introduce struct data_loc_info Namhyung Kim
2023-11-10  0:00 ` [PATCH 41/52] perf map: Add map__objdump_2rip() Namhyung Kim
2023-11-10  0:00 ` [PATCH 42/52] perf annotate: Add annotate_get_basic_blocks() Namhyung Kim
2023-11-10  0:00 ` [PATCH 43/52] perf annotate-data: Maintain variable type info Namhyung Kim
2023-11-10  0:00 ` [PATCH 44/52] perf annotate-data: Add update_insn_state() Namhyung Kim
2023-11-10  0:00 ` [PATCH 45/52] perf annotate-data: Handle global variable access Namhyung Kim
2023-11-10  0:00 ` [PATCH 46/52] perf annotate-data: Handle call instructions Namhyung Kim
2023-11-10  0:00 ` [PATCH 47/52] perf annotate-data: Implement instruction tracking Namhyung Kim
2023-11-10  0:00 ` [PATCH 48/52] perf annotate: Parse x86 segment register location Namhyung Kim
2023-11-10  0:00 ` [PATCH 49/52] perf annotate-data: Handle this-cpu variables in kernel Namhyung Kim
2023-11-10  0:00 ` [PATCH 50/52] perf annotate-data: Track instructions with a this-cpu variable Namhyung Kim
2023-11-10  0:00 ` [PATCH 51/52] perf annotate-data: Add stack canary type Namhyung Kim
2023-11-10  0:00 ` [PATCH 52/52] perf annotate-data: Add debug message Namhyung Kim
2023-11-10 12:05 ` [RFC 00/52] perf tools: Introduce data type profiling (v2) Arnaldo Carvalho de Melo
2023-11-11  2:27   ` Namhyung Kim

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to='CAM9d7cijprv2fOM_jwYbrV0jbhno2ZQDCT3SExG=Ot=b2mANmw@mail.gmail.com' \
    --to=namhyung@kernel.org \
    --cc=acme@kernel.org \
    --cc=adrian.hunter@intel.com \
    --cc=ak@linux.intel.com \
    --cc=eranian@google.com \
    --cc=irogers@google.com \
    --cc=jolsa@kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-perf-users@vger.kernel.org \
    --cc=linux-toolchains@vger.kernel.org \
    --cc=linux-trace-devel@vger.kernel.org \
    --cc=mhiramat@kernel.org \
    --cc=mingo@kernel.org \
    --cc=peterz@infradead.org \
    --cc=torvalds@linux-foundation.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).