* Steven Rostedt <rost...@goodmis.org> wrote:
>
> Ingo,
>
> The first of the patches is a minor fix to when a woken RT task is about
> to preempt a pinned RT task, push_rt_task() is called to try to
> migrate the woken task if possible (to avoid preempting a pinned RT
> task that may be the second highest priority task in the system).
>
> But the issue is that push_rt_task() won't push it if the woken task
> is higher priority even if the task to be preempted is pinned.
>
> The second two patches is more of a clean up to remove the idle
> hooks in the scheduler proper, and to use the pre/post schedule
> methods instead.
>
> This allows interrupts to be enabled in the idle balance, which slightly
> helps latencies, especially for the -rt kernel. Other patches can be
> added on top. Maybe in the future preemption could be enabled during
> the idle balance as well. But that remains to be seen.
>
> Please pull the latest tip/sched/core tree, which can be found at:
>
> git://git.kernel.org/pub/scm/linux/kernel/git/rostedt/linux-trace.git
> tip/sched/core
>
> Head SHA1: 1db13ecf89054d39922e7b3323d124c7e2921560
>
>
> Steven Rostedt (Red Hat) (3):
> sched/rt: Fix push_rt_task() to have the same checks as the caller did
> sched: Move idle_balance() to post_schedule
> sched: Enable interrupts in idle_balance()
>
> ----
> kernel/sched/core.c | 3 ---
> kernel/sched/fair.c | 8 +++++---
> kernel/sched/idle_task.c | 10 ++++++++++
> kernel/sched/rt.c | 15 ++++++++++-----
> 4 files changed, 25 insertions(+), 11 deletions(-)
Pulled, thanks a lot Steve!
Ingo
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majord...@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
