[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <e90c75a2e4e3f529c8d42edabcb103885449de3e.camel@gmx.de>
Date: Wed, 03 Jul 2024 15:45:13 +0200
From: Mike Galbraith <efault@....de>
To: Chen Yu <yu.c.chen@...el.com>
Cc: Raghavendra K T <raghavendra.kt@....com>, Peter Zijlstra
<peterz@...radead.org>, Ingo Molnar <mingo@...hat.com>, Juri Lelli
<juri.lelli@...hat.com>, Vincent Guittot <vincent.guittot@...aro.org>, Tim
Chen <tim.c.chen@...el.com>, Yujie Liu <yujie.liu@...el.com>, K Prateek
Nayak <kprateek.nayak@....com>, "Gautham R . Shenoy"
<gautham.shenoy@....com>, Chen Yu <yu.chen.surf@...il.com>,
linux-kernel@...r.kernel.org
Subject: Re: [PATCH 1/2] sched/fair: Record the average duration of a task
On Wed, 2024-07-03 at 21:23 +0800, Chen Yu wrote:
> On 2024-07-03 at 13:57:19 +0200, Mike Galbraith wrote:
> >
> > Oooh dear.. as soon as you mention threads, the question of who's
> > wheelhouse is this in springs to mind, ie should the kernel be
> > overriding userspace by targeting bits of threaded programs for forced
> > serialization?
>
> Do you mean the threads within the same process should run in parallel
> as much as possible, regardless of sharing the same data, because the thread
> is designed to do so? If so then we should probably skip the 1:1 task stacking
> if the waker and the wakee are in the same process.
That's seems like the obviously correct answer.. except for that
leaving virtually the entire issue you're squabbling with lying on the
floor. Grr, how annoying.
-Mike
Powered by blists - more mailing lists