linux-perf-users.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Michael Petlan <mpetlan@redhat.com>
To: Namhyung Kim <namhyung@gmail.com>
Cc: linux-perf-users@vger.kernel.org, acme@kernel.org,
	acme@redhat.com, qzhao@redhat.com, cjense@google.com
Subject: Re: [PATCH 2/2] perf test: Fix JSON format linter test checks
Date: Tue, 24 Jan 2023 17:49:07 +0100 (CET)	[thread overview]
Message-ID: <alpine.LRH.2.20.2301241747420.5281@Diego> (raw)
In-Reply-To: <CAM9d7ch-if6Ov9dT+zHtYzVOdOjGOe=YsQvmb_5AOdbK9NDmZQ@mail.gmail.com>

On Sun, 22 Jan 2023, Namhyung Kim wrote:
> On Fri, Jan 20, 2023 at 6:00 AM Michael Petlan <mpetlan@redhat.com> wrote:
> >
> > The test fails on CPUs with topdown metrics, where it is common to print
> > two metrics per line. Since these are included in default event set for
> > `perf stat -- something`, it causes the "no args" subtest to fail due to
> > unexpected member count. We need to accept 7 or 9 members in that case.
> >
> > Coming to that, counting commas in the JSON line and consider it a number
> > of elements is incorrect and misleading. There should be +1 element than
> > the count of commas, while also, commas can potentially appear in the
> > event names. Count " : " substrings rather, since these better fit to
> > what should be actually counted.
> 
> I'm also curious if ":" could appear in the event/metric names.
> Probably more chances to have ":" than "," in the name, no?

Yes, ':' can appear in the event names, but not " : " (with the spaces
around).

Michael

> 
> Thanks,
> Namhyung
> 
> 
> >
> > Fixes: 0c343af2a2f8 ("perf test: JSON format checking")
> >
> > Signed-off-by: Michael Petlan <mpetlan@redhat.com>
> > ---
> >  .../tests/shell/lib/perf_json_output_lint.py     | 16 ++++++++--------
> >  1 file changed, 8 insertions(+), 8 deletions(-)
> >
> > diff --git a/tools/perf/tests/shell/lib/perf_json_output_lint.py b/tools/perf/tests/shell/lib/perf_json_output_lint.py
> > index d90f8d102eb9..4f1bbb3f07ec 100644
> > --- a/tools/perf/tests/shell/lib/perf_json_output_lint.py
> > +++ b/tools/perf/tests/shell/lib/perf_json_output_lint.py
> > @@ -40,17 +40,17 @@ def is_counter_value(num):
> >    return isfloat(num) or num == '<not counted>' or num == '<not supported>'
> >
> >  def check_json_output(expected_items):
> > -  if expected_items != -1:
> > +  if expected_items:
> >      for line in Lines:
> >        if 'failed' not in line:
> >          count = 0
> > -        count = line.count(',')
> > -        if count != expected_items and count >= 1 and count <= 3 and 'metric-value' in line:
> > +        count = line.count(' : ')
> > +        if count not in expected_items and count >= 2 and count <= 4 and 'metric-value' in line:
> >            # Events that generate >1 metric may have isolated metric
> >            # values and possibly other prefixes like interval, core and
> >            # aggregate-number.
> >            continue
> > -        if count != expected_items:
> > +        if count not in expected_items:
> >            raise RuntimeError(f'wrong number of fields. counted {count} expected {expected_items}'
> >                               f' in \'{line}\'')
> >    checks = {
> > @@ -82,14 +82,14 @@ def check_json_output(expected_items):
> >
> >  try:
> >    if args.no_args or args.system_wide or args.event:
> > -    expected_items = 6
> > +    expected_items = [7, 9]
> >    elif args.interval or args.per_thread or args.system_wide_no_aggr:
> > -    expected_items = 7
> > +    expected_items = [8, 10]
> >    elif args.per_core or args.per_socket or args.per_node or args.per_die:
> > -    expected_items = 8
> > +    expected_items = [9, 11]
> >    else:
> >      # If no option is specified, don't check the number of items.
> > -    expected_items = -1
> > +    expected_items = []
> >    check_json_output(expected_items)
> >  except:
> >    print('Test failed for input:\n' + '\n'.join(Lines))
> > --
> > 2.18.4
> >
> 
> 


  reply	other threads:[~2023-01-24 16:50 UTC|newest]

Thread overview: 18+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2023-01-20 13:40 [PATCH 0/2] perf test: Fix JSON linter Michael Petlan
2023-01-20 13:40 ` [PATCH 1/2] perf stat: Fix JSON metric printout for multiple metrics per line Michael Petlan
2023-01-23  6:31   ` Namhyung Kim
2023-05-22 12:11   ` Michael Petlan
2023-06-06 11:16     ` Michael Petlan
2023-01-20 13:40 ` [PATCH 2/2] perf test: Fix JSON format linter test checks Michael Petlan
2023-01-23  6:36   ` Namhyung Kim
2023-01-24 16:49     ` Michael Petlan [this message]
2023-01-24 17:26       ` Namhyung Kim
2023-01-27 12:26   ` Arnaldo Carvalho de Melo
2023-01-27 12:30     ` Arnaldo Carvalho de Melo
2023-01-31 17:14       ` Michael Petlan
2023-02-02  1:18         ` Arnaldo Carvalho de Melo
2023-01-20 17:37 ` [PATCH 0/2] perf test: Fix JSON linter Ian Rogers
2023-01-23  6:48   ` Namhyung Kim
2023-01-23 13:38   ` Arnaldo Carvalho de Melo
2023-01-24 17:39   ` Michael Petlan
2023-01-25  0:37     ` Ian Rogers

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=alpine.LRH.2.20.2301241747420.5281@Diego \
    --to=mpetlan@redhat.com \
    --cc=acme@kernel.org \
    --cc=acme@redhat.com \
    --cc=cjense@google.com \
    --cc=linux-perf-users@vger.kernel.org \
    --cc=namhyung@gmail.com \
    --cc=qzhao@redhat.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
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).