From: Arun Sharma <asharma@fb.com>
To: avagin@openvz.org
Cc: Arnaldo Carvalho de Melo <acme@ghostprotocols.net>,
linux-kernel@vger.kernel.org,
Peter Zijlstra <a.p.zijlstra@chello.nl>,
linux-perf-users@vger.kernel.org,
Paul Mackerras <paulus@samba.org>,
devel@openvz.org, David Ahern <dsahern@gmail.com>,
Ingo Molnar <mingo@elte.hu>
Subject: Re: [Devel] [PATCH 0/7] Profiling sleep times (v3)
Date: Wed, 7 Dec 2011 12:33:48 -0800 [thread overview]
Message-ID: <4EDFCDAC.3060600@fb.com> (raw)
In-Reply-To: <4EDDC0FB.6080005@openvz.org>
On 12/5/11 11:15 PM, Andrey Vagin wrote:
> Arun Sharma said, that the second versions of patches works ok for him.
> (Arun is the first user of this functionality after me.)
Yes - Andrey's patches (v2) have been functional for me when used via:
perf record -agP -e sched:sched_switch --filter "prev_state == 1 ||
prev_state == 2" -e sched:sched_stat_sleep,sched:sched_stat_iowait --
sleep 3
mv perf.data{,.old}; perf inject -s -i perf.data.old -o perf.data
perf report --stdio -g graph --sort pid -C command-name
There are two major issues though:
* The above command lines cause us to collect way too much data and perf
can't keep up on a busy server
Eg:
Warning:
Processed 55182 events and lost 13 chunks!
Check IO/CPU overload!
* Requires root access
I suspect there is a way to stash the delay information available at
enqueue_sleeper() into the task_struct (or some place else) and make it
available at the sched:sched_switch tracepoint. This would solve both of
the problems above, but I don't have a patch yet.
-Arun
WARNING: multiple messages have this Message-ID (diff)
From: Arun Sharma <asharma@fb.com>
To: <avagin@openvz.org>
Cc: Arnaldo Carvalho de Melo <acme@ghostprotocols.net>,
<linux-kernel@vger.kernel.org>,
Peter Zijlstra <a.p.zijlstra@chello.nl>,
<linux-perf-users@vger.kernel.org>,
Paul Mackerras <paulus@samba.org>, <devel@openvz.org>,
David Ahern <dsahern@gmail.com>, Ingo Molnar <mingo@elte.hu>
Subject: Re: [Devel] [PATCH 0/7] Profiling sleep times (v3)
Date: Wed, 7 Dec 2011 12:33:48 -0800 [thread overview]
Message-ID: <4EDFCDAC.3060600@fb.com> (raw)
In-Reply-To: <4EDDC0FB.6080005@openvz.org>
On 12/5/11 11:15 PM, Andrey Vagin wrote:
> Arun Sharma said, that the second versions of patches works ok for him.
> (Arun is the first user of this functionality after me.)
Yes - Andrey's patches (v2) have been functional for me when used via:
perf record -agP -e sched:sched_switch --filter "prev_state == 1 ||
prev_state == 2" -e sched:sched_stat_sleep,sched:sched_stat_iowait --
sleep 3
mv perf.data{,.old}; perf inject -s -i perf.data.old -o perf.data
perf report --stdio -g graph --sort pid -C command-name
There are two major issues though:
* The above command lines cause us to collect way too much data and perf
can't keep up on a busy server
Eg:
Warning:
Processed 55182 events and lost 13 chunks!
Check IO/CPU overload!
* Requires root access
I suspect there is a way to stash the delay information available at
enqueue_sleeper() into the task_struct (or some place else) and make it
available at the sched:sched_switch tracepoint. This would solve both of
the problems above, but I don't have a patch yet.
-Arun
next prev parent reply other threads:[~2011-12-07 20:33 UTC|newest]
Thread overview: 26+ messages / expand[flat|nested] mbox.gz Atom feed top
2011-11-28 9:03 [PATCH 0/7] Profiling sleep times (v3) Andrew Vagin
2011-11-28 9:03 ` [PATCH 1/7] perf: use event_name() to get an event name Andrew Vagin
2011-12-06 14:02 ` Arnaldo Carvalho de Melo
2011-12-08 4:34 ` [tip:perf/urgent] perf header: Use " tip-bot for Andrew Vagin
2011-11-28 9:03 ` [PATCH 2/7] perf: add ability to record event period Andrew Vagin
2011-12-06 14:02 ` Arnaldo Carvalho de Melo
2011-11-28 9:03 ` [PATCH 3/7] perf: add ability to change event according to sample (v2) Andrew Vagin
2011-12-06 14:19 ` Arnaldo Carvalho de Melo
2011-12-06 14:24 ` Arnaldo Carvalho de Melo
2011-12-06 14:57 ` David Ahern
2011-12-06 15:06 ` Arnaldo Carvalho de Melo
2011-11-28 9:03 ` [PATCH 4/7] perf: teach "perf inject" to work with files Andrew Vagin
2011-11-28 9:03 ` [PATCH 5/7] perf: teach perf inject to merge sched_stat_* and sched_switch events Andrew Vagin
2011-11-28 9:03 ` [PATCH 6/7] perf: add scripts for profiling sleep times (v2) Andrew Vagin
2011-11-28 9:03 ` [PATCH 7/7] event: add tracepoint for accounting block time Andrew Vagin
2011-11-28 11:42 ` Peter Zijlstra
2011-11-28 14:02 ` Arjan van de Ven
2011-11-28 14:31 ` Arnaldo Carvalho de Melo
2011-11-28 14:43 ` Arjan van de Ven
2011-11-28 14:59 ` Arnaldo Carvalho de Melo
2011-12-06 9:48 ` [tip:sched/core] events, sched: Add tracepoint for accounting blocked time tip-bot for Andrew Vagin
2011-12-06 7:15 ` [Devel] [PATCH 0/7] Profiling sleep times (v3) Andrey Vagin
2011-12-06 8:30 ` Ingo Molnar
2011-12-06 13:45 ` Arnaldo Carvalho de Melo
2011-12-07 20:33 ` Arun Sharma [this message]
2011-12-07 20:33 ` Arun Sharma
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=4EDFCDAC.3060600@fb.com \
--to=asharma@fb.com \
--cc=a.p.zijlstra@chello.nl \
--cc=acme@ghostprotocols.net \
--cc=avagin@openvz.org \
--cc=devel@openvz.org \
--cc=dsahern@gmail.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-perf-users@vger.kernel.org \
--cc=mingo@elte.hu \
--cc=paulus@samba.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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.