From: Tejun Heo <tj@kernel.org>
To: Avi Kivity <avi@redhat.com>
Cc: Peter Zijlstra <peterz@infradead.org>,
mingo@elte.hu, efault@gmx.de, paulus@samba.org, acme@redhat.com,
linux-kernel@vger.kernel.org
Subject: Re: [RFC PATCHSET] sched,perf: unify tracers in sched and move perf on top of TP
Date: Wed, 05 May 2010 15:09:14 +0200 [thread overview]
Message-ID: <4BE16DFA.3070204@kernel.org> (raw)
In-Reply-To: <4BE165B0.1070300@redhat.com>
Hello,
On 05/05/2010 02:33 PM, Avi Kivity wrote:
> On 05/05/2010 02:38 PM, Peter Zijlstra wrote:
>> As it stands I'd argue to simply drop this whole idea. The SCHED_EVENT()
>> thing doesn't look like its worth the obfuscation and I'm very much
>> opposed to making perf and sched_notifiers rely on tracepoints.
>
> We can make perf rely on sched_notifiers, I think it's exactly the same
> case as kvm - extra cpu state that needs switching at context switch
> time. We could also make fpu switching use sched_notifiers, but that's
> pushing it a bit.
Oh, that was my first attempt too. The difference between TPs and
sched_notifiers is the enable condition. TP is system-wide while
sched_notifier is per-task. The thing is that for perf hooks,
system-wide enable condition works much better. Also, w/
sched_notifiers, there's the problem that no one else but self can
enable/disable it (which can be changed but doing it without causing
call time overhead requires a bit of effort).
Thanks.
--
tejun
next prev parent reply other threads:[~2010-05-05 13:10 UTC|newest]
Thread overview: 55+ messages / expand[flat|nested] mbox.gz Atom feed top
2010-05-04 12:38 Tejun Heo
2010-05-04 12:38 ` [PATCH 01/12] sched: drop @cpu argument from sched_in preempt notifier Tejun Heo
2010-05-04 17:11 ` Peter Zijlstra
2010-05-04 12:38 ` [PATCH 02/12] sched: rename preempt_notifiers to sched_notifiers and refactor implementation Tejun Heo
2010-05-04 12:38 ` [PATCH 03/12] perf: add perf_event_task_migrate() Tejun Heo
2010-05-05 5:08 ` Frederic Weisbecker
2010-05-05 5:16 ` Tejun Heo
2010-05-05 9:11 ` Peter Zijlstra
2010-05-05 9:37 ` Tejun Heo
2010-05-05 9:50 ` Peter Zijlstra
2010-05-05 9:56 ` Tejun Heo
2010-05-04 12:38 ` [PATCH 04/12] perf: add @rq to perf_event_task_sched_out() Tejun Heo
2010-05-04 17:11 ` Peter Zijlstra
2010-05-04 17:22 ` Tejun Heo
2010-05-04 17:42 ` Peter Zijlstra
2010-05-04 18:22 ` Steven Rostedt
2010-05-04 18:26 ` Peter Zijlstra
2010-05-04 18:32 ` Steven Rostedt
2010-05-05 4:48 ` Tejun Heo
2010-05-05 9:58 ` Tejun Heo
2010-05-07 18:41 ` [tip:sched/core] sched: Remove rq argument to the tracepoints tip-bot for Peter Zijlstra
2010-05-04 12:38 ` [PATCH 05/12] perf: move perf_event_task_sched_in() next to fire_sched_notifiers_in() Tejun Heo
2010-05-04 12:38 ` [PATCH 06/12] sched: relocate fire_sched_notifiers_out() and trace_sched_switch() Tejun Heo
2010-05-04 12:38 ` [PATCH 07/12] sched: coalesce event notifiers Tejun Heo
2010-05-04 12:38 ` [PATCH 08/12] sched: add switch_in and tick tracepoints Tejun Heo
2010-05-04 12:38 ` [PATCH 09/12] perf: factor out perf_event_switch_clones() Tejun Heo
2010-05-04 12:38 ` [PATCH 10/12] perf: make nr_events an int and add perf_online_mutex to protect it Tejun Heo
2010-05-04 12:38 ` [PATCH 11/12] perf: prepare to move sched perf functions on top of tracepoints Tejun Heo
2010-05-04 12:38 ` [PATCH 12/12] perf: " Tejun Heo
2010-05-04 17:29 ` [RFC PATCHSET] sched,perf: unify tracers in sched and move perf on top of TP Peter Zijlstra
2010-05-05 5:00 ` Tejun Heo
2010-05-05 9:06 ` Peter Zijlstra
2010-05-05 9:32 ` Tejun Heo
2010-05-05 9:51 ` Peter Zijlstra
2010-05-05 9:54 ` Tejun Heo
2010-05-05 11:38 ` Peter Zijlstra
2010-05-05 12:28 ` Tejun Heo
2010-05-05 16:55 ` Ingo Molnar
2010-05-05 18:12 ` Peter Zijlstra
2010-05-05 18:16 ` Peter Zijlstra
2010-05-05 18:30 ` Frederic Weisbecker
2010-05-06 6:28 ` Ingo Molnar
2010-05-06 7:11 ` Tejun Heo
2010-05-06 8:27 ` Ingo Molnar
2010-05-06 8:41 ` Tejun Heo
2010-05-06 8:18 ` Avi Kivity
2010-05-06 6:31 ` Ingo Molnar
2010-05-06 7:04 ` Peter Zijlstra
2010-05-06 7:11 ` Ingo Molnar
2010-05-06 7:29 ` Tejun Heo
2010-05-06 7:33 ` Tejun Heo
2010-05-05 12:33 ` Avi Kivity
2010-05-05 13:09 ` Tejun Heo [this message]
2010-05-10 5:20 ` Paul Mackerras
2010-05-10 5:48 ` Tejun Heo
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=4BE16DFA.3070204@kernel.org \
--to=tj@kernel.org \
--cc=acme@redhat.com \
--cc=avi@redhat.com \
--cc=efault@gmx.de \
--cc=linux-kernel@vger.kernel.org \
--cc=mingo@elte.hu \
--cc=paulus@samba.org \
--cc=peterz@infradead.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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox
all inboxes | Powered by JetHome®