On Thu 17-08-17 19:12:40, Andrea Arcangeli wrote:
> On Wed, Aug 16, 2017 at 03:23:29PM +0200, Michal Hocko wrote:
> > Reviewed-by: Michal Hocko <mhocko@xxxxxxxx>
>
> Thanks for the review!
>
> There's this further possible microoptimization that can be folded on top.
>
> >From 76bf017f923581d15fe01249af92b0d757752a9f Mon Sep 17 00:00:00 2001
> From: Andrea Arcangeli <aarcange@xxxxxxxxxx>
> Date: Thu, 17 Aug 2017 18:39:46 +0200
> Subject: [PATCH 1/1] mm: oom: let oom_reap_task and exit_mmap run
> concurrently, add unlikely
>
> Microoptimization to fold before merging.
>
> Signed-off-by: Andrea Arcangeli <aarcange@xxxxxxxxxx>
> ---
> mm/mmap.c | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/mm/mmap.c b/mm/mmap.c
> index 013170e5c8a4..ab0026a8acc4 100644
> --- a/mm/mmap.c
> +++ b/mm/mmap.c
> @@ -3003,7 +3003,7 @@ void exit_mmap(struct mm_struct *mm)
> unmap_vmas(&tlb, vma, 0, -1);
>
> set_bit(MMF_OOM_SKIP, &mm->flags);
> - if (tsk_is_oom_victim(current)) {
> + if (unlikely(tsk_is_oom_victim(current))) {
I dunno. This doesn't make any difference in the generated code for
me (with gcc 6.4). If anything we might wan't to putt unlikely inside
tsk_is_oom_victim. Or even go further and use a jump label to get any
conditional paths out of way.
> /*
> * Wait for oom_reap_task() to stop working on this
> * mm. Because MMF_OOM_SKIP is already set before
>
--
Michal Hocko
SUSE Labs
--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@xxxxxxxxx. For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@xxxxxxxxx";> email@xxxxxxxxx </a>
