Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]
Groups > linux.kernel > #1522535 > unrolled thread
| Started by | Daniel Bristot de Oliveira <bristot@redhat.com> |
|---|---|
| First post | 2016-11-15 11:40 +0100 |
| Last post | 2016-11-15 18:00 +0100 |
| Articles | 5 — 3 participants |
Back to article view | Back to linux.kernel
[PATCH 0/2] sched/rt: disable RT_RUNTIME_SHARE by default and document it Daniel Bristot de Oliveira <bristot@redhat.com> - 2016-11-15 11:40 +0100
[PATCH 1/2] sched/rt: Disable RT_RUNTIME_SHARE by default Daniel Bristot de Oliveira <bristot@redhat.com> - 2016-11-15 11:50 +0100
Re: [PATCH 0/2] sched/rt: disable RT_RUNTIME_SHARE by default and document it "Luis Claudio R. Goncalves" <lclaudio@uudg.org> - 2016-11-15 14:10 +0100
Re: [PATCH 0/2] sched/rt: disable RT_RUNTIME_SHARE by default and document it Steven Rostedt <rostedt@goodmis.org> - 2016-11-15 16:30 +0100
Re: [PATCH 0/2] sched/rt: disable RT_RUNTIME_SHARE by default and document it Daniel Bristot de Oliveira <bristot@redhat.com> - 2016-11-15 18:00 +0100
| From | Daniel Bristot de Oliveira <bristot@redhat.com> |
|---|---|
| Date | 2016-11-15 11:40 +0100 |
| Subject | [PATCH 0/2] sched/rt: disable RT_RUNTIME_SHARE by default and document it |
| Message-ID | <sDJjH-3JZ-5@gated-at.bofh.it> |
Disable RT_RUNTIME_SHARE by default and document it. Cc: Ingo Molnar <mingo@redhat.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Clark Williams <williams@redhat.com> Cc: Luis Claudio R. Goncalves <lgoncalv@redhat.com> Cc: linux-kernel@vger.kernel.org Daniel Bristot de Oliveira (2): sched/rt: Disable RT_RUNTIME_SHARE by default sched/rt: Document RT_RUNTIME_SHARE sched feature kernel/sched/features.h | 12 +++++++++++- 1 file changed, 11 insertions(+), 1 deletion(-) -- 2.7.4
[toc] | [next] | [standalone]
| From | Daniel Bristot de Oliveira <bristot@redhat.com> |
|---|---|
| Date | 2016-11-15 11:50 +0100 |
| Subject | [PATCH 1/2] sched/rt: Disable RT_RUNTIME_SHARE by default |
| Message-ID | <sDJto-3Nv-39@gated-at.bofh.it> |
| In reply to | #1522535 |
The RT_RUNTIME_SHARE sched feature enables the sharing of rt_runtime between CPUs, allowing a CPU to run a real-time task up to 100% of the time while leaving more space for non-real-time tasks to run on the CPU that lend rt_runtime. The problem is that a CPU can easily borrow enough rt_runtime to allow a spinning rt-task to run forever, starving per-cpu tasks like kworkers, which are non-real-time by design. This patch disables RT_RUNTIME_SHARE by default, avoiding this problem. The feature will still be present for users that want to enable it, though. Signed-off-by: Daniel Bristot de Oliveira <bristot@redhat.com> Cc: Ingo Molnar <mingo@redhat.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Clark Williams <williams@redhat.com> Cc: Luis Claudio R. Goncalves <lgoncalv@redhat.com> Cc: linux-kernel@vger.kernel.org diff --git a/kernel/sched/features.h b/kernel/sched/features.h index 69631fa..5fffebfe 100644 --- a/kernel/sched/features.h +++ b/kernel/sched/features.h @@ -65,7 +65,7 @@ SCHED_FEAT(RT_PUSH_IPI, true) #endif SCHED_FEAT(FORCE_SD_OVERLAP, false) -SCHED_FEAT(RT_RUNTIME_SHARE, true) +SCHED_FEAT(RT_RUNTIME_SHARE, false) SCHED_FEAT(LB_MIN, false) SCHED_FEAT(ATTACH_AGE_LOAD, true) -- 2.7.4
[toc] | [prev] | [next] | [standalone]
| From | "Luis Claudio R. Goncalves" <lclaudio@uudg.org> |
|---|---|
| Date | 2016-11-15 14:10 +0100 |
| Subject | Re: [PATCH 0/2] sched/rt: disable RT_RUNTIME_SHARE by default and document it |
| Message-ID | <sDLET-5lG-73@gated-at.bofh.it> |
| In reply to | #1522535 |
On Tue, Nov 15, 2016 at 11:39:09AM +0100, Daniel Bristot de Oliveira wrote: > Disable RT_RUNTIME_SHARE by default and document it. > > Cc: Ingo Molnar <mingo@redhat.com> > Cc: Peter Zijlstra <peterz@infradead.org> > Cc: Steven Rostedt <rostedt@goodmis.org> > Cc: Clark Williams <williams@redhat.com> > Cc: Luis Claudio R. Goncalves <lgoncalv@redhat.com> > Cc: linux-kernel@vger.kernel.org > > Daniel Bristot de Oliveira (2): > sched/rt: Disable RT_RUNTIME_SHARE by default > sched/rt: Document RT_RUNTIME_SHARE sched feature > > kernel/sched/features.h | 12 +++++++++++- > 1 file changed, 11 insertions(+), 1 deletion(-) Acked-by: Luis Claudio R. Goncalves <lgoncalv@redhat.com> In light of the discussion about "sched/rt: RT_RUNTIME_GREED sched feature", disabling RT_RUNTIME_SHARE by default restores a portion of determinism we gave away by using the clever mechanism behind RT_RUNTIME_SHARE. (I should have prefaced that with a IMHO). Luis
[toc] | [prev] | [next] | [standalone]
| From | Steven Rostedt <rostedt@goodmis.org> |
|---|---|
| Date | 2016-11-15 16:30 +0100 |
| Subject | Re: [PATCH 0/2] sched/rt: disable RT_RUNTIME_SHARE by default and document it |
| Message-ID | <sDNQm-6Go-37@gated-at.bofh.it> |
| In reply to | #1522535 |
On Tue, 15 Nov 2016 11:39:09 +0100 Daniel Bristot de Oliveira <bristot@redhat.com> wrote: > Disable RT_RUNTIME_SHARE by default and document it. > > Cc: Ingo Molnar <mingo@redhat.com> > Cc: Peter Zijlstra <peterz@infradead.org> > Cc: Steven Rostedt <rostedt@goodmis.org> > Cc: Clark Williams <williams@redhat.com> > Cc: Luis Claudio R. Goncalves <lgoncalv@redhat.com> > Cc: linux-kernel@vger.kernel.org > > Daniel Bristot de Oliveira (2): > sched/rt: Disable RT_RUNTIME_SHARE by default > sched/rt: Document RT_RUNTIME_SHARE sched feature > > kernel/sched/features.h | 12 +++++++++++- > 1 file changed, 11 insertions(+), 1 deletion(-) > I don't see why this is two patches. A change and a comment can go together, and probably should, especially since the comment explains the change as well. With a folded patch... Acked-by: Steven Rostedt <rostedt@goodmis.org> -- Steve
[toc] | [prev] | [next] | [standalone]
| From | Daniel Bristot de Oliveira <bristot@redhat.com> |
|---|---|
| Date | 2016-11-15 18:00 +0100 |
| Subject | Re: [PATCH 0/2] sched/rt: disable RT_RUNTIME_SHARE by default and document it |
| Message-ID | <sDPfr-7tr-3@gated-at.bofh.it> |
| In reply to | #1522825 |
On 11/15/2016 04:27 PM, Steven Rostedt wrote: > I don't see why this is two patches. A change and a comment can go > together, and probably should, especially since the comment explains > the change as well. > > With a folded patch... I will cook a v2 in a single patch! Thanks! -- Daniel
[toc] | [prev] | [standalone]
Back to top | Article view | linux.kernel
csiph-web