Re: [PATCH 7/7 v3] sched: fix wrong utilization accounting when switching to fair class

2016-09-26 Thread Peter Zijlstra
On Thu, Sep 22, 2016 at 02:25:18PM +0200, Peter Zijlstra wrote: > On Tue, Sep 20, 2016 at 03:06:04PM +0200, Vincent Guittot wrote: > > > Also, even it things magically worked out, its still very icky to mix > > > the normalized vruntime into things. > > > > I agree > > In any case, I pushed out a

Re: [PATCH 7/7 v3] sched: fix wrong utilization accounting when switching to fair class

2016-09-22 Thread bsegall
Peter Zijlstra writes: > On Tue, Sep 20, 2016 at 09:59:08AM -0700, bseg...@google.com wrote: >> Peter Zijlstra writes: >> > Google has this patch-set replacing min_vruntime with an actual global >> > 0-lag, which greatly simplifies things. If only they'd post it sometime >> > :/ /me prods pjt an

Re: [PATCH 7/7 v3] sched: fix wrong utilization accounting when switching to fair class

2016-09-22 Thread Peter Zijlstra
On Tue, Sep 20, 2016 at 03:06:04PM +0200, Vincent Guittot wrote: > > Also, even it things magically worked out, its still very icky to mix > > the normalized vruntime into things. > > I agree In any case, I pushed out a bunch of patches to: git://git.kernel.org/pub/scm/linux/kernel/git/peterz/

Re: [PATCH 7/7 v3] sched: fix wrong utilization accounting when switching to fair class

2016-09-22 Thread Peter Zijlstra
On Tue, Sep 20, 2016 at 09:59:08AM -0700, bseg...@google.com wrote: > Peter Zijlstra writes: > > Google has this patch-set replacing min_vruntime with an actual global > > 0-lag, which greatly simplifies things. If only they'd post it sometime > > :/ /me prods pjt and ben with a sharp stick :-) >

Re: [PATCH 7/7 v3] sched: fix wrong utilization accounting when switching to fair class

2016-09-20 Thread bsegall
Peter Zijlstra writes: > On Fri, Sep 16, 2016 at 04:23:16PM +0200, Vincent Guittot wrote: >> On 16 September 2016 at 14:16, Peter Zijlstra wrote: > >> >> > Also, the normalize comment in dequeue_entity() worries me, 'someone' >> >> > didn't update that when he moved update_min_vruntime() around.

Re: [PATCH 7/7 v3] sched: fix wrong utilization accounting when switching to fair class

2016-09-20 Thread Vincent Guittot
On 20 September 2016 at 13:54, Peter Zijlstra wrote: > On Fri, Sep 16, 2016 at 04:23:16PM +0200, Vincent Guittot wrote: >> On 16 September 2016 at 14:16, Peter Zijlstra wrote: > >> >> > Also, the normalize comment in dequeue_entity() worries me, 'someone' >> >> > didn't update that when he moved

Re: [PATCH 7/7 v3] sched: fix wrong utilization accounting when switching to fair class

2016-09-20 Thread Peter Zijlstra
On Fri, Sep 16, 2016 at 04:23:16PM +0200, Vincent Guittot wrote: > On 16 September 2016 at 14:16, Peter Zijlstra wrote: > >> > Also, the normalize comment in dequeue_entity() worries me, 'someone' > >> > didn't update that when he moved update_min_vruntime() around. > > > > I now worry more, so w

Re: [PATCH 7/7 v3] sched: fix wrong utilization accounting when switching to fair class

2016-09-16 Thread Vincent Guittot
On 16 September 2016 at 14:16, Peter Zijlstra wrote: > On Thu, Sep 15, 2016 at 05:36:58PM +0200, Vincent Guittot wrote: >> On 15 September 2016 at 15:18, Peter Zijlstra wrote: >> > On Mon, Sep 12, 2016 at 09:47:52AM +0200, Vincent Guittot wrote: > >> >> Update the sequence to follow the right one

Re: [PATCH 7/7 v3] sched: fix wrong utilization accounting when switching to fair class

2016-09-16 Thread Vincent Guittot
On 16 September 2016 at 12:51, Peter Zijlstra wrote: > > On Mon, Sep 12, 2016 at 09:47:52AM +0200, Vincent Guittot wrote: > > > -dequeue task > > -put task > > -change the property > > -enqueue task > > -set task as current task > > > diff --git a/kernel/sched/core.c b/kernel/sched/core.c > > inde

Re: [PATCH 7/7 v3] sched: fix wrong utilization accounting when switching to fair class

2016-09-16 Thread Peter Zijlstra
On Thu, Sep 15, 2016 at 05:36:58PM +0200, Vincent Guittot wrote: > On 15 September 2016 at 15:18, Peter Zijlstra wrote: > > On Mon, Sep 12, 2016 at 09:47:52AM +0200, Vincent Guittot wrote: > >> Update the sequence to follow the right one: > >> -dequeue task > >> -put task > >> -change the propert

Re: [PATCH 7/7 v3] sched: fix wrong utilization accounting when switching to fair class

2016-09-16 Thread Peter Zijlstra
On Mon, Sep 12, 2016 at 09:47:52AM +0200, Vincent Guittot wrote: > -dequeue task > -put task > -change the property > -enqueue task > -set task as current task > diff --git a/kernel/sched/core.c b/kernel/sched/core.c > index 3e52d08..7a9c9b9 100644 > --- a/kernel/sched/core.c > +++ b/kernel/sched

Re: [PATCH 7/7 v3] sched: fix wrong utilization accounting when switching to fair class

2016-09-15 Thread Vincent Guittot
On 15 September 2016 at 15:18, Peter Zijlstra wrote: > On Mon, Sep 12, 2016 at 09:47:52AM +0200, Vincent Guittot wrote: >> When a task switches to fair scheduling class, the period between now and >> the last update of its utilization is accounted as running time whatever >> happened during this p

Re: [PATCH 7/7 v3] sched: fix wrong utilization accounting when switching to fair class

2016-09-15 Thread Peter Zijlstra
On Mon, Sep 12, 2016 at 09:47:52AM +0200, Vincent Guittot wrote: > When a task switches to fair scheduling class, the period between now and > the last update of its utilization is accounted as running time whatever > happened during this period. This wrong accounting applies to the task > and also

[PATCH 7/7 v3] sched: fix wrong utilization accounting when switching to fair class

2016-09-12 Thread Vincent Guittot
When a task switches to fair scheduling class, the period between now and the last update of its utilization is accounted as running time whatever happened during this period. This wrong accounting applies to the task and also to the task group branch. When changing the property of a running task