BPF Archive on lore.kernel.org
 help / color / Atom feed
From: Ian Rogers <irogers@google.com>
To: Arnaldo Carvalho de Melo <acme@kernel.org>
Cc: Jiri Olsa <jolsa@redhat.com>,
	Peter Zijlstra <peterz@infradead.org>,
	Ingo Molnar <mingo@redhat.com>,
	Mark Rutland <mark.rutland@arm.com>,
	Alexander Shishkin <alexander.shishkin@linux.intel.com>,
	Namhyung Kim <namhyung@kernel.org>,
	Song Liu <songliubraving@fb.com>,
	Andrii Nakryiko <andriin@fb.com>,
	Kajol Jain <kjain@linux.ibm.com>, Andi Kleen <ak@linux.intel.com>,
	John Garry <john.garry@huawei.com>,
	Jin Yao <yao.jin@linux.intel.com>,
	Kan Liang <kan.liang@linux.intel.com>,
	Cong Wang <xiyou.wangcong@gmail.com>,
	Kim Phillips <kim.phillips@amd.com>, Paul Clarke <pc@us.ibm.com>,
	Srikar Dronamraju <srikar@linux.vnet.ibm.com>,
	LKML <linux-kernel@vger.kernel.org>,
	Networking <netdev@vger.kernel.org>, bpf <bpf@vger.kernel.org>,
	linux-perf-users <linux-perf-users@vger.kernel.org>,
	Vince Weaver <vincent.weaver@maine.edu>,
	Stephane Eranian <eranian@google.com>
Subject: Re: [PATCH v2 0/7] Share events between metrics
Date: Fri, 22 May 2020 10:56:59 -0700
Message-ID: <CAP-5=fWZYJ2RXeXGGmFXAW9CNnb2S6cGYKc_M=hUQyCng7KJBQ@mail.gmail.com> (raw)
In-Reply-To: <CAP-5=fUaaNpi3RZd9-Q-uCaudop0tU5NN8HFek5e2XLoBZqt6w@mail.gmail.com>

On Fri, May 22, 2020 at 7:59 AM Ian Rogers <irogers@google.com> wrote:
>
>
>
> On Fri, May 22, 2020, 7:49 AM Arnaldo Carvalho de Melo <acme@kernel.org> wrote:
>>
>> Em Fri, May 22, 2020 at 12:13:11PM +0200, Jiri Olsa escreveu:
>> > On Thu, May 21, 2020 at 02:22:35PM -0300, Arnaldo Carvalho de Melo wrote:
>> > > Em Thu, May 21, 2020 at 01:43:25PM +0200, Jiri Olsa escreveu:
>> > > > On Wed, May 20, 2020 at 11:20:04AM -0700, Ian Rogers wrote:
>> > > >
>> > > > SNIP
>> > > >
>> > > > > There are 5 out of 12 metric groups where no events are shared, such
>> > > > > as Power, however, disabling grouping of events always reduces the
>> > > > > number of events.
>> > > > >
>> > > > > The result for Memory_BW needs explanation:
>> > > > >
>> > > > > Metric group: Memory_BW
>> > > > >  - No merging (old default, now --metric-no-merge): 9
>> > > > >  - Merging over metrics (new default)             : 5
>> > > > >  - No event groups and merging (--metric-no-group): 11
>> > > > >
>> > > > > Both with and without merging the groups fail to be set up and so the
>> > > > > event counts here are for broken metrics. The --metric-no-group number
>> > > > > is accurate as all the events are scheduled. Ideally a constraint
>> > > > > would be added for these metrics in the json code to avoid grouping.
>> > > > >
>> > > > > v2. rebases on kernel/git/acme/linux.git branch tmp.perf/core, fixes a
>> > > > > missing comma with metric lists (reported-by Jiri Olsa
>> > > > > <jolsa@redhat.com>) and adds early returns to metricgroup__add_metric
>> > > > > (suggested-by Jiri Olsa).
>> > > >
>> > > > Acked-by: Jiri Olsa <jolsa@redhat.com>
>> > >
>> > > Applied and pushed to tmp.perf/core, will move to perf/core as soon as
>> > > testing finishes,
>> >
>> > I checked tmp.perf/core and I'm getting segfault for 'perf test expr'
>>
>> Right, reproduced here and...
>>
>> >        7: Simple expression parser                              :
>> >       Program received signal SIGSEGV, Segmentation fault.
>> >       0x000000000067841e in hashmap_find_entry (map=0x7fffffffd0c0, key=0xc83b30, hash=9893851511679796638, pprev=0x0, entry=0x7fffffffc658) at hashmap.c:131
>> >       131             for (prev_ptr = &map->buckets[hash], cur = *prev_ptr;
>> >       (gdb) bt
>> >       #0  0x000000000067841e in hashmap_find_entry (map=0x7fffffffd0c0, key=0xc83b30, hash=9893851511679796638, pprev=0x0, entry=0x7fffffffc658) at hashmap.c:131
>> >       #1  0x000000000067853a in hashmap__insert (map=0x7fffffffd0c0, key=0xc83b30, value=0x0, strategy=HASHMAP_SET, old_key=0x7fffffffc718,
>> >           old_value=0x7fffffffc710) at hashmap.c:160
>> >       #2  0x00000000005d3209 in hashmap__set (map=0x7fffffffd0c0, key=0xc83b30, value=0x0, old_key=0x7fffffffc718, old_value=0x7fffffffc710)
>> >           at /home/jolsa/kernel/linux-perf/tools/perf/util/hashmap.h:107
>> >       #3  0x00000000005d3386 in expr__add_id (ctx=0x7fffffffd0c0, name=0xc83b30 "FOO", val=0) at util/expr.c:45
>> >       #4  0x00000000005d27ee in expr_parse (final_val=0x0, ctx=0x7fffffffd0c0, scanner=0xc87990) at util/expr.y:63
>> >       #5  0x00000000005d35b7 in __expr__parse (val=0x0, ctx=0x7fffffffd0c0, expr=0x75a84b "FOO + BAR + BAZ + BOZO", start=259, runtime=1) at util/expr.c:102
>> >       #6  0x00000000005d36c6 in expr__find_other (expr=0x75a84b "FOO + BAR + BAZ + BOZO", one=0x75a791 "FOO", ctx=0x7fffffffd0c0, runtime=1) at util/expr.c:121
>> >       #7  0x00000000004e3aaf in test__expr (t=0xa7bd40 <generic_tests+384>, subtest=-1) at tests/expr.c:55
>> >       #8  0x00000000004b5651 in run_test (test=0xa7bd40 <generic_tests+384>, subtest=-1) at tests/builtin-test.c:393
>> >       #9  0x00000000004b5787 in test_and_print (t=0xa7bd40 <generic_tests+384>, force_skip=false, subtest=-1) at tests/builtin-test.c:423
>> >       #10 0x00000000004b61c4 in __cmd_test (argc=1, argv=0x7fffffffd7f0, skiplist=0x0) at tests/builtin-test.c:628
>> >       #11 0x00000000004b6911 in cmd_test (argc=1, argv=0x7fffffffd7f0) at tests/builtin-test.c:772
>> >       #12 0x00000000004e977b in run_builtin (p=0xa7eee8 <commands+552>, argc=3, argv=0x7fffffffd7f0) at perf.c:312
>> >       #13 0x00000000004e99e8 in handle_internal_command (argc=3, argv=0x7fffffffd7f0) at perf.c:364
>> >       #14 0x00000000004e9b2f in run_argv (argcp=0x7fffffffd64c, argv=0x7fffffffd640) at perf.c:408
>> >       #15 0x00000000004e9efb in main (argc=3, argv=0x7fffffffd7f0) at perf.c:538
>> >
>> > attached patch fixes it for me, but I'm not sure this
>> > should be necessary
>>
>> ... applying the patch below makes the segfault go away. Ian, Ack? I can
>> fold it into the patch introducing the problem.
>
>
> I suspect this patch is a memory leak. The underlying issue is likely the outstanding hashmap_clear fix in libbpf. Let me check.
>
> Thanks,
> Ian

Tested:
$ git checkout -b testing acme/tmp.perf/core
$ make ...
$ perf test 7
7: Simple expression parser                              : FAILED!
$ git cherry-pick 6bca339175bf
[acme-perf-expr-testing 4614bd252003] libbpf: Fix memory leak and
possible double-free in hashmap__c
lear
Author: Andrii Nakryiko <andriin@fb.com>
Date: Tue Apr 28 18:21:04 2020 -0700
1 file changed, 7 insertions(+)
$ make ...
$ perf test 7
7: Simple expression parser                              : Ok

I'd prefer we took the libbpf fix as initializing over the top of the
hashmap will leak. This fix is in the tools/perf/util/hashmap.c.

Thanks,
Ian

>> - Arnaldo
>>
>> > jirka
>> >
>> >
>> > ---
>> > diff --git a/tools/perf/tests/expr.c b/tools/perf/tests/expr.c
>> > index 1cb02ca2b15f..21693fe516c1 100644
>> > --- a/tools/perf/tests/expr.c
>> > +++ b/tools/perf/tests/expr.c
>> > @@ -52,6 +52,7 @@ int test__expr(struct test *t __maybe_unused, int subtest __maybe_unused)
>> >       TEST_ASSERT_VAL("missing operand", ret == -1);
>> >
>> >       expr__ctx_clear(&ctx);
>> > +     expr__ctx_init(&ctx);
>> >       TEST_ASSERT_VAL("find other",
>> >                       expr__find_other("FOO + BAR + BAZ + BOZO", "FOO",
>> >                                        &ctx, 1) == 0);
>> > @@ -64,6 +65,7 @@ int test__expr(struct test *t __maybe_unused, int subtest __maybe_unused)
>> >                                                   (void **)&val_ptr));
>> >
>> >       expr__ctx_clear(&ctx);
>> > +     expr__ctx_init(&ctx);
>> >       TEST_ASSERT_VAL("find other",
>> >                       expr__find_other("EVENT1\\,param\\=?@ + EVENT2\\,param\\=?@",
>> >                                        NULL, &ctx, 3) == 0);
>> >
>> >
>>
>> --
>>
>> - Arnaldo

  parent reply index

Thread overview: 17+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-05-20 18:20 Ian Rogers
2020-05-20 18:20 ` [PATCH v2 1/7] perf metricgroup: Always place duration_time last Ian Rogers
2020-05-20 18:20 ` [PATCH v2 2/7] perf metricgroup: Use early return in add_metric Ian Rogers
2020-05-20 18:20 ` [PATCH v2 3/7] perf metricgroup: Delay events string creation Ian Rogers
2020-05-20 18:20 ` [PATCH v2 4/7] perf metricgroup: Order event groups by size Ian Rogers
2020-05-20 18:20 ` [PATCH v2 5/7] perf metricgroup: Remove duped metric group events Ian Rogers
2020-05-20 18:20 ` [PATCH v2 6/7] perf metricgroup: Add options to not group or merge Ian Rogers
2020-05-20 18:20 ` [PATCH v2 7/7] perf metricgroup: Remove unnecessary ',' from events Ian Rogers
2020-05-21 11:43 ` [PATCH v2 0/7] Share events between metrics Jiri Olsa
2020-05-21 17:22   ` Arnaldo Carvalho de Melo
2020-05-22 10:13     ` Jiri Olsa
2020-05-22 14:49       ` Arnaldo Carvalho de Melo
     [not found]         ` <CAP-5=fUaaNpi3RZd9-Q-uCaudop0tU5NN8HFek5e2XLoBZqt6w@mail.gmail.com>
2020-05-22 17:56           ` Ian Rogers [this message]
2020-05-23 22:19             ` Jiri Olsa
2020-05-25 13:34               ` Arnaldo Carvalho de Melo
2020-05-22  9:25 ` kajoljain
2020-05-22 14:31   ` Arnaldo Carvalho de Melo

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='CAP-5=fWZYJ2RXeXGGmFXAW9CNnb2S6cGYKc_M=hUQyCng7KJBQ@mail.gmail.com' \
    --to=irogers@google.com \
    --cc=acme@kernel.org \
    --cc=ak@linux.intel.com \
    --cc=alexander.shishkin@linux.intel.com \
    --cc=andriin@fb.com \
    --cc=bpf@vger.kernel.org \
    --cc=eranian@google.com \
    --cc=john.garry@huawei.com \
    --cc=jolsa@redhat.com \
    --cc=kan.liang@linux.intel.com \
    --cc=kim.phillips@amd.com \
    --cc=kjain@linux.ibm.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-perf-users@vger.kernel.org \
    --cc=mark.rutland@arm.com \
    --cc=mingo@redhat.com \
    --cc=namhyung@kernel.org \
    --cc=netdev@vger.kernel.org \
    --cc=pc@us.ibm.com \
    --cc=peterz@infradead.org \
    --cc=songliubraving@fb.com \
    --cc=srikar@linux.vnet.ibm.com \
    --cc=vincent.weaver@maine.edu \
    --cc=xiyou.wangcong@gmail.com \
    --cc=yao.jin@linux.intel.com \
    /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

BPF Archive on lore.kernel.org

Archives are clonable:
	git clone --mirror https://lore.kernel.org/bpf/0 bpf/git/0.git

	# If you have public-inbox 1.1+ installed, you may
	# initialize and index your mirror using the following commands:
	public-inbox-init -V2 bpf bpf/ https://lore.kernel.org/bpf \
		bpf@vger.kernel.org
	public-inbox-index bpf

Example config snippet for mirrors

Newsgroup available over NNTP:
	nntp://nntp.lore.kernel.org/org.kernel.vger.bpf


AGPL code for this site: git clone https://public-inbox.org/public-inbox.git