linux-kernel - Re: [PATCH] perf evsel: Correct clock unit to nanosecond

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <20181130152521.GA10748@kernel.org>
Date:   Fri, 30 Nov 2018 12:25:21 -0300
From:   Arnaldo Carvalho de Melo <acme@...nel.org>
To:     leo.yan@...aro.org
Cc:     Jiri Olsa <jolsa@...hat.com>,
        Alexander Shishkin <alexander.shishkin@...ux.intel.com>,
        Andi Kleen <ak@...ux.intel.com>, linux-kernel@...r.kernel.org,
        Daniel Thompson <daniel.thompson@...aro.org>
Subject: Re: [PATCH] perf evsel: Correct clock unit to nanosecond

Em Fri, Nov 30, 2018 at 10:21:40PM +0800, leo.yan@...aro.org escreveu:
> On Fri, Nov 30, 2018 at 11:21:54AM +0100, Jiri Olsa wrote:
> > On Fri, Nov 30, 2018 at 06:06:05PM +0800, Leo Yan wrote:
> > > Since commit 0aa802a79469 ("perf stat: Get rid of extra clock display
> > > function"), the cpu and task clock unit has been changed from
> > > nanosecond value to millisecond value.  This introduces confusion for
> > > CPU run time statistics, we can see in below flow the clock value is
> > > scaled from nanosecond value to millisecond value; but this is
> > > contradiction with statistics type 'STAT_NSECS', which always takes
> > > clock as nanosecond unit.
> 
> [...]
> 
> > there was fix for this recently, could you please check
> > if it's working for you:
> >   6e269c85dcea perf stat: Fix shadow stats for clock events
> 
> Ah, I missed this patch ...
> 
> Yeah, I tested this patch and can confirm this patch fixes
> this issue.

My fault, new notebook update process got in the way causing
acme/perf/core to pile up, a perf/core pull req is about to be sent to
Ingo now tho :-)

- Arnaldo