Re: [PATCH] sched/fair: Cleanup definition of NOHZ blocked load functions

2019-06-03 Thread Peter Zijlstra
On Mon, Jun 03, 2019 at 11:22:40AM +0100, Valentin Schneider wrote: > On 03/06/2019 10:38, Peter Zijlstra wrote: > [...] > > > > I'm thinking the below can go on top to further clean up? > > > > Yep, that's even better indeed! Want me to resend with that extra diff? Yes please, I didn't even

Re: [PATCH] sched/fair: Cleanup definition of NOHZ blocked load functions

2019-06-03 Thread Valentin Schneider
On 03/06/2019 10:38, Peter Zijlstra wrote: [...] > > I'm thinking the below can go on top to further clean up? > Yep, that's even better indeed! Want me to resend with that extra diff? [...]

Re: [PATCH] sched/fair: Cleanup definition of NOHZ blocked load functions

2019-06-03 Thread Vincent Guittot
On Mon, 3 Jun 2019 at 11:38, Peter Zijlstra wrote: > > On Sun, Jun 02, 2019 at 05:41:10PM +0100, Valentin Schneider wrote: > > cfs_rq_has_blocked() and others_have_blocked() are only used within > > update_blocked_averages(). The !CONFIG_FAIR_GROUP_SCHED version of the > > latter calls them

Re: [PATCH] sched/fair: Cleanup definition of NOHZ blocked load functions

2019-06-03 Thread Peter Zijlstra
On Sun, Jun 02, 2019 at 05:41:10PM +0100, Valentin Schneider wrote: > cfs_rq_has_blocked() and others_have_blocked() are only used within > update_blocked_averages(). The !CONFIG_FAIR_GROUP_SCHED version of the > latter calls them within a #define CONFIG_NO_HZ_COMMON block, whereas > the

[PATCH] sched/fair: Cleanup definition of NOHZ blocked load functions

2019-06-02 Thread Valentin Schneider
cfs_rq_has_blocked() and others_have_blocked() are only used within update_blocked_averages(). The !CONFIG_FAIR_GROUP_SCHED version of the latter calls them within a #define CONFIG_NO_HZ_COMMON block, whereas the CONFIG_FAIR_GROUP_SCHED one calls them unconditionnally. As reported by Qian, the