[PATCH] kmemleak: add unlikely() optimization to kmemleak_late_init

[PATCH] kmemleak: add unlikely() optimization to kmemleak_late_init

[Davidlohr Bueso]

kmemleak: add unlikely() optimization to kmemleak_late_init()

Since there it is not very probable that kmemleak_disable() was called right after setting kmemleak_initialized, we could benefit from predicting it.

Signed-off-by: Davidlohr Bueso <dave@gnu.org>
---
 mm/kmemleak.c |    2 +-
 1 files changed, 1 insertions(+), 1 deletions(-)

diff --git a/mm/kmemleak.c b/mm/kmemleak.c
index bd9bc21..7e2f260 100644
--- a/mm/kmemleak.c
+++ b/mm/kmemleak.c
@@ -1728,7 +1728,7 @@ static int __init kmemleak_late_init(void)
 
 	atomic_set(&kmemleak_initialized, 1);
 
-	if (atomic_read(&kmemleak_error)) {
+	if (unlikely(atomic_read(&kmemleak_error))) {
 		/*
 		 * Some error occured and kmemleak was disabled. There is a
 		 * small chance that kmemleak_disable() was called immediately
-- 
1.7.0.4

Re: [PATCH] kmemleak: add unlikely() optimization to kmemleak_late_init

[Catalin Marinas]

Davidlohr Bueso <dave@gnu.org> wrote:
> kmemleak: add unlikely() optimization to kmemleak_late_init()
>
> Since there it is not very probable that kmemleak_disable() was called right after setting kmemleak_initialized, we could benefit from predicting it.

The patch is correct and you can add my Acked-by but I don't really
think it's worth. We save ~2 cycles during booting, no-one would ever
notice.

-- 
Catalin

Re: [PATCH] kmemleak: add unlikely() optimization to kmemleak_late_init

[Davidlohr Bueso]

On Wed, 2010-09-22 at 11:06 +0100, Catalin Marinas wrote:
> Davidlohr Bueso <dave@gnu.org> wrote:
> > kmemleak: add unlikely() optimization to kmemleak_late_init()
> >
> > Since there it is not very probable that kmemleak_disable() was called right after setting kmemleak_initialized, we could benefit from predicting it.
> 
> The patch is correct and you can add my Acked-by but I don't really
> think it's worth. We save ~2 cycles during booting, no-one would ever
> notice.
> 

Yeah, it's not a big performance booster, but IMHO it wouldn't hurt to
add it. Since you are the maintainer it's up to you if you include it
upstream.

Thanks,
Davidlohr