mbox series

[v8,perf,bpf,00/15] perf annotation of BPF programs

Message ID 20190311195555.889995-1-songliubraving@fb.com
Headers show
Series perf annotation of BPF programs | expand

Message

Song Liu March 11, 2019, 7:55 p.m. UTC
Changes v7 to v8:
1. Address issues suggested by Jiri and Stanislav.

Changes v6 to v7:
1. Fix minor issues suggested by Jiri.

Changes v5 to v6:
1. Improve side band evlist interface;
2. Minor style fixes.

Changes v4 to v5:
1. Rebase to latest bpf-next;
2. Add dependency of 94816add0005 from Arnaldo's tree;
3. More details in change logs;
4. Add perf_env__init() to init bpf related lock and rbtrees;
5. Small clean ups.

Changes v3 to v4:
1. Incorporate feedbacks from Jiri and Namhyung;
2. Fixed compilation error with different feature-disassembler-four-args;
3. Split some patches to smaller patches;
4. Improved error handleing in symbol__disassemble_bpf();
5. Made side band thread more generic;
6. Added comments as suggested.

Changes v2 to v3:
1. Remove unnecessary include in header files;
2. Improved error handling;
3. Better naming of functions, variables, etc.;
4. Enable bpf events by default for perf-top.

Changes v1 to v2:
1. Fix compilation error with different feature-disassembler-four-args;
2. Fix a segfault in perf-record;
3. Split patches 5/9 and 6/9 so that perf_env changes and perf.data changes
   are in separate patches.

This series enables annotation of BPF programs in perf.

perf tool gathers information via sys_bpf and (optionally) stores them in
perf.data as headers.

Patch 1/15 fixes a minor issue in kernel;
Patch 2/15 to 4/15 introduce new helper functions and use them in perf and
     bpftool;
Patch 5/15 to 9/15 saves information of bpf program in perf_env;
Patch 10/15 adds --bpf-event options to perf-top;
Patch 11/15 to 13/15 enables annotation of bpf progs based on information
     gathered in 5/15 to 9/15;
Patch 14/15 introduces side band polling thread that gathers information
     for special kernel events during perf-record or perf-top.
Patch 15/15 handles information of short living BPF program using the new
     side band polling thread.

Commands tested during developments are perf-top, perf-record, perf-report,
and perf-annotate.

===================== Note on patch dependency  ========================
This set has dependency in both bpf-next tree and tip/perf/core. Current
version is developed on bpf-next tree with the following commits
cherry-picked from tip/perf/core (or acme/perf/core):

(from 1/11 to 11/11)
commit 76193a94522f ("perf, bpf: Introduce PERF_RECORD_KSYMBOL")
commit d764ac646491 ("tools headers uapi: Sync tools/include/uapi/linux/perf_event.h")
commit 6ee52e2a3fe4 ("perf, bpf: Introduce PERF_RECORD_BPF_EVENT")
commit df063c83aa2c ("tools headers uapi: Sync tools/include/uapi/linux/perf_event.h")
commit 9aa0bfa370b2 ("perf tools: Handle PERF_RECORD_KSYMBOL")
commit 45178a928a4b ("perf tools: Handle PERF_RECORD_BPF_EVENT")
commit 7b612e291a5a ("perf tools: Synthesize PERF_RECORD_* for loaded BPF programs")
commit a40b95bcd30c ("perf top: Synthesize BPF events for pre-existing loaded BPF programs")
commit 6934058d9fb6 ("bpf: Add module name [bpf] to ksymbols for bpf programs")
commit 811184fb6977 ("perf bpf: Fix synthesized PERF_RECORD_KSYMBOL/BPF_EVENT")
comimt 94816add0005 ("perf tools: Add perf_exe() helper to find perf binary")
========================================================================

This set is also available at:

https://github.com/liu-song-6/linux/tree/bpf-annotation

Thanks!!

Song Liu (15):
  perf, bpf: consider events with attr.bpf_event as side-band events
  bpf: libbpf: introduce bpf_program__get_prog_info_linear()
  bpf: bpftool: use bpf_program__get_prog_info_linear() in
    prog.c:do_dump()
  perf, bpf: synthesize bpf events with
    bpf_program__get_prog_info_linear()
  perf: change prototype of perf_event__synthesize_bpf_events()
  perf, bpf: save bpf_prog_info in a rbtree in perf_env
  perf, bpf: save bpf_prog_info information as headers to perf.data
  perf, bpf: save btf in a rbtree in perf_env
  perf, bpf: save btf information as headers to perf.data
  perf-top: add option --no-bpf-event
  perf: add -lopcodes to feature-libbfd
  perf, bpf: enable annotation of bpf program
  perf, bpf: process PERF_BPF_EVENT_PROG_LOAD for annotation
  perf: introduce side band thread
  perf, bpf: save bpf_prog_info and btf of short living bpf programs

 kernel/events/core.c         |   3 +-
 tools/bpf/bpftool/prog.c     | 266 +++++++---------------------
 tools/build/Makefile.feature |   6 +-
 tools/lib/bpf/libbpf.c       | 251 +++++++++++++++++++++++++++
 tools/lib/bpf/libbpf.h       |  63 +++++++
 tools/lib/bpf/libbpf.map     |   3 +
 tools/perf/Makefile.config   |  10 +-
 tools/perf/builtin-record.c  |   9 +-
 tools/perf/builtin-top.c     |  12 +-
 tools/perf/perf.c            |   1 +
 tools/perf/util/annotate.c   | 150 +++++++++++++++-
 tools/perf/util/bpf-event.c  | 327 ++++++++++++++++++++++++++---------
 tools/perf/util/bpf-event.h  |  33 +++-
 tools/perf/util/dso.c        |   1 +
 tools/perf/util/dso.h        |  33 ++--
 tools/perf/util/env.c        | 151 ++++++++++++++++
 tools/perf/util/env.h        |  24 +++
 tools/perf/util/evlist.c     | 119 +++++++++++++
 tools/perf/util/evlist.h     |  12 ++
 tools/perf/util/evsel.h      |   6 +
 tools/perf/util/header.c     | 241 +++++++++++++++++++++++++-
 tools/perf/util/header.h     |   2 +
 tools/perf/util/session.c    |   1 +
 tools/perf/util/symbol.c     |   1 +
 tools/perf/util/top.h        |   1 +
 25 files changed, 1413 insertions(+), 313 deletions(-)

--
2.17.1

Comments

Arnaldo Carvalho de Melo March 11, 2019, 8:24 p.m. UTC | #1
Em Mon, Mar 11, 2019 at 12:55:40PM -0700, Song Liu escreveu:
> ===================== Note on patch dependency  ========================
> This set has dependency in both bpf-next tree and tip/perf/core. Current
> version is developed on bpf-next tree with the following commits
> cherry-picked from tip/perf/core (or acme/perf/core):
> 
> (from 1/11 to 11/11)
> commit 76193a94522f ("perf, bpf: Introduce PERF_RECORD_KSYMBOL")
> commit d764ac646491 ("tools headers uapi: Sync tools/include/uapi/linux/perf_event.h")
> commit 6ee52e2a3fe4 ("perf, bpf: Introduce PERF_RECORD_BPF_EVENT")
> commit df063c83aa2c ("tools headers uapi: Sync tools/include/uapi/linux/perf_event.h")
> commit 9aa0bfa370b2 ("perf tools: Handle PERF_RECORD_KSYMBOL")
> commit 45178a928a4b ("perf tools: Handle PERF_RECORD_BPF_EVENT")
> commit 7b612e291a5a ("perf tools: Synthesize PERF_RECORD_* for loaded BPF programs")
> commit a40b95bcd30c ("perf top: Synthesize BPF events for pre-existing loaded BPF programs")
> commit 6934058d9fb6 ("bpf: Add module name [bpf] to ksymbols for bpf programs")
> commit 811184fb6977 ("perf bpf: Fix synthesized PERF_RECORD_KSYMBOL/BPF_EVENT")
> comimt 94816add0005 ("perf tools: Add perf_exe() helper to find perf binary")
> ========================================================================

These csets are already upstream, i.e. already hit torvalds/master, i.e.
will be in 5.1, can you check that we still need doing this
cherry-picking?

- Arnaldo
Song Liu March 11, 2019, 9:01 p.m. UTC | #2
> On Mar 11, 2019, at 1:24 PM, Arnaldo Carvalho de Melo <acme@redhat.com> wrote:
> 
> Em Mon, Mar 11, 2019 at 12:55:40PM -0700, Song Liu escreveu:
>> ===================== Note on patch dependency  ========================
>> This set has dependency in both bpf-next tree and tip/perf/core. Current
>> version is developed on bpf-next tree with the following commits
>> cherry-picked from tip/perf/core (or acme/perf/core):
>> 
>> (from 1/11 to 11/11)
>> commit 76193a94522f ("perf, bpf: Introduce PERF_RECORD_KSYMBOL")
>> commit d764ac646491 ("tools headers uapi: Sync tools/include/uapi/linux/perf_event.h")
>> commit 6ee52e2a3fe4 ("perf, bpf: Introduce PERF_RECORD_BPF_EVENT")
>> commit df063c83aa2c ("tools headers uapi: Sync tools/include/uapi/linux/perf_event.h")
>> commit 9aa0bfa370b2 ("perf tools: Handle PERF_RECORD_KSYMBOL")
>> commit 45178a928a4b ("perf tools: Handle PERF_RECORD_BPF_EVENT")
>> commit 7b612e291a5a ("perf tools: Synthesize PERF_RECORD_* for loaded BPF programs")
>> commit a40b95bcd30c ("perf top: Synthesize BPF events for pre-existing loaded BPF programs")
>> commit 6934058d9fb6 ("bpf: Add module name [bpf] to ksymbols for bpf programs")
>> commit 811184fb6977 ("perf bpf: Fix synthesized PERF_RECORD_KSYMBOL/BPF_EVENT")
>> comimt 94816add0005 ("perf tools: Add perf_exe() helper to find perf binary")
>> ========================================================================
> 
> These csets are already upstream, i.e. already hit torvalds/master, i.e.
> will be in 5.1, can you check that we still need doing this
> cherry-picking?
> 
> - Arnaldo

I just checked that 02-15 of the set could apply to torvalds/master (with 
some minor conflicts), and it works as expected. 

Thanks,
Song
Arnaldo Carvalho de Melo March 11, 2019, 9:05 p.m. UTC | #3
Em Mon, Mar 11, 2019 at 09:01:48PM +0000, Song Liu escreveu:
> 
> 
> > On Mar 11, 2019, at 1:24 PM, Arnaldo Carvalho de Melo <acme@redhat.com> wrote:
> > 
> > Em Mon, Mar 11, 2019 at 12:55:40PM -0700, Song Liu escreveu:
> >> ===================== Note on patch dependency  ========================
> >> This set has dependency in both bpf-next tree and tip/perf/core. Current
> >> version is developed on bpf-next tree with the following commits
> >> cherry-picked from tip/perf/core (or acme/perf/core):
> >> 
> >> (from 1/11 to 11/11)
> >> commit 76193a94522f ("perf, bpf: Introduce PERF_RECORD_KSYMBOL")
> >> commit d764ac646491 ("tools headers uapi: Sync tools/include/uapi/linux/perf_event.h")
> >> commit 6ee52e2a3fe4 ("perf, bpf: Introduce PERF_RECORD_BPF_EVENT")
> >> commit df063c83aa2c ("tools headers uapi: Sync tools/include/uapi/linux/perf_event.h")
> >> commit 9aa0bfa370b2 ("perf tools: Handle PERF_RECORD_KSYMBOL")
> >> commit 45178a928a4b ("perf tools: Handle PERF_RECORD_BPF_EVENT")
> >> commit 7b612e291a5a ("perf tools: Synthesize PERF_RECORD_* for loaded BPF programs")
> >> commit a40b95bcd30c ("perf top: Synthesize BPF events for pre-existing loaded BPF programs")
> >> commit 6934058d9fb6 ("bpf: Add module name [bpf] to ksymbols for bpf programs")
> >> commit 811184fb6977 ("perf bpf: Fix synthesized PERF_RECORD_KSYMBOL/BPF_EVENT")
> >> comimt 94816add0005 ("perf tools: Add perf_exe() helper to find perf binary")
> >> ========================================================================
> > 
> > These csets are already upstream, i.e. already hit torvalds/master, i.e.
> > will be in 5.1, can you check that we still need doing this
> > cherry-picking?
> > 
> > - Arnaldo
> 
> I just checked that 02-15 of the set could apply to torvalds/master (with 
> some minor conflicts), and it works as expected. 

So no need to cherrypick, just use torvalds/master as a base, ok.

- Arnaldo
Song Liu March 11, 2019, 9:08 p.m. UTC | #4
> On Mar 11, 2019, at 2:05 PM, Arnaldo Carvalho de Melo <arnaldo.melo@gmail.com> wrote:
> 
> Em Mon, Mar 11, 2019 at 09:01:48PM +0000, Song Liu escreveu:
>> 
>> 
>>> On Mar 11, 2019, at 1:24 PM, Arnaldo Carvalho de Melo <acme@redhat.com> wrote:
>>> 
>>> Em Mon, Mar 11, 2019 at 12:55:40PM -0700, Song Liu escreveu:
>>>> ===================== Note on patch dependency  ========================
>>>> This set has dependency in both bpf-next tree and tip/perf/core. Current
>>>> version is developed on bpf-next tree with the following commits
>>>> cherry-picked from tip/perf/core (or acme/perf/core):
>>>> 
>>>> (from 1/11 to 11/11)
>>>> commit 76193a94522f ("perf, bpf: Introduce PERF_RECORD_KSYMBOL")
>>>> commit d764ac646491 ("tools headers uapi: Sync tools/include/uapi/linux/perf_event.h")
>>>> commit 6ee52e2a3fe4 ("perf, bpf: Introduce PERF_RECORD_BPF_EVENT")
>>>> commit df063c83aa2c ("tools headers uapi: Sync tools/include/uapi/linux/perf_event.h")
>>>> commit 9aa0bfa370b2 ("perf tools: Handle PERF_RECORD_KSYMBOL")
>>>> commit 45178a928a4b ("perf tools: Handle PERF_RECORD_BPF_EVENT")
>>>> commit 7b612e291a5a ("perf tools: Synthesize PERF_RECORD_* for loaded BPF programs")
>>>> commit a40b95bcd30c ("perf top: Synthesize BPF events for pre-existing loaded BPF programs")
>>>> commit 6934058d9fb6 ("bpf: Add module name [bpf] to ksymbols for bpf programs")
>>>> commit 811184fb6977 ("perf bpf: Fix synthesized PERF_RECORD_KSYMBOL/BPF_EVENT")
>>>> comimt 94816add0005 ("perf tools: Add perf_exe() helper to find perf binary")
>>>> ========================================================================
>>> 
>>> These csets are already upstream, i.e. already hit torvalds/master, i.e.
>>> will be in 5.1, can you check that we still need doing this
>>> cherry-picking?
>>> 
>>> - Arnaldo
>> 
>> I just checked that 02-15 of the set could apply to torvalds/master (with 
>> some minor conflicts), and it works as expected. 
> 
> So no need to cherrypick, just use torvalds/master as a base, ok.
> 
> - Arnaldo

Do you want me to resend the set with the conflict resolved? 

Thanks,
Song
Arnaldo Carvalho de Melo March 11, 2019, 9:18 p.m. UTC | #5
Em Mon, Mar 11, 2019 at 09:08:31PM +0000, Song Liu escreveu:
> > On Mar 11, 2019, at 2:05 PM, Arnaldo Carvalho de Melo <arnaldo.melo@gmail.com> wrote:
> > Em Mon, Mar 11, 2019 at 09:01:48PM +0000, Song Liu escreveu:
> >>> On Mar 11, 2019, at 1:24 PM, Arnaldo Carvalho de Melo <acme@redhat.com> wrote:
> >>> These csets are already upstream, i.e. already hit torvalds/master, i.e.
> >>> will be in 5.1, can you check that we still need doing this
> >>> cherry-picking?

> >> I just checked that 02-15 of the set could apply to torvalds/master (with 
> >> some minor conflicts), and it works as expected. 

> > So no need to cherrypick, just use torvalds/master as a base, ok.
 
> Do you want me to resend the set with the conflict resolved? 

No need for today, I think Jiri is looking at that branch I saw you guys
chatting today, so if he has some new comments, you can then address and
repost, fixing the conflicts then.

- Arnaldo
Song Liu March 11, 2019, 9:22 p.m. UTC | #6
> On Mar 11, 2019, at 2:18 PM, Arnaldo Carvalho de Melo <arnaldo.melo@gmail.com> wrote:
> 
> Em Mon, Mar 11, 2019 at 09:08:31PM +0000, Song Liu escreveu:
>>> On Mar 11, 2019, at 2:05 PM, Arnaldo Carvalho de Melo <arnaldo.melo@gmail.com> wrote:
>>> Em Mon, Mar 11, 2019 at 09:01:48PM +0000, Song Liu escreveu:
>>>>> On Mar 11, 2019, at 1:24 PM, Arnaldo Carvalho de Melo <acme@redhat.com> wrote:
>>>>> These csets are already upstream, i.e. already hit torvalds/master, i.e.
>>>>> will be in 5.1, can you check that we still need doing this
>>>>> cherry-picking?
> 
>>>> I just checked that 02-15 of the set could apply to torvalds/master (with 
>>>> some minor conflicts), and it works as expected. 
> 
>>> So no need to cherrypick, just use torvalds/master as a base, ok.
> 
>> Do you want me to resend the set with the conflict resolved? 
> 
> No need for today, I think Jiri is looking at that branch I saw you guys
> chatting today, so if he has some new comments, you can then address and
> repost, fixing the conflicts then.
> 
> - Arnaldo

Got it. I will include the --bpf-event option change and repost later. 

Thanks,
Song