Re: mempool - only init waitqueue in slow path

[Andrew Morton <akpm@osdl.org>]

Benjamin LaHaise <bcrl@kvack.org> wrote:
>
> Here's a small patch to improve the performance of mempool_alloc by only 
>  initializing the wait queue when we're about to wait.
> 
>  		-ben
>  -- 
>  "Time is what keeps everything from happening all at once." -- John Wheeler
> 
>  Signed-off-by: Benjamin LaHaise <benjamin.c.lahaise@intel.com>
>  diff -purN v2.6.12-rc4/mm/mempool.c mempool-rc4/mm/mempool.c
>  --- v2.6.12-rc4/mm/mempool.c	2005-05-09 15:47:01.000000000 -0400
>  +++ mempool-rc4/mm/mempool.c	2005-05-13 10:04:54.000000000 -0400
>  @@ -197,7 +197,7 @@ void * mempool_alloc(mempool_t *pool, un
>   {
>   	void *element;
>   	unsigned long flags;
>  -	DEFINE_WAIT(wait);
>  +	wait_queue_t wait;
>   	int gfp_temp;
>   
>   	might_sleep_if(gfp_mask & __GFP_WAIT);
>  @@ -228,6 +228,7 @@ repeat_alloc:
>   
>   	/* Now start performing page reclaim */
>   	gfp_temp = gfp_mask;
>  +	init_wait(&wait);
>   	prepare_to_wait(&pool->wait, &wait, TASK_UNINTERRUPTIBLE);
>   	smp_mb();

hmm, I'd have thought that gcc was smart enough to defer the
initialisations until the thing-which-is-being-initialised is about to be
used.  But with gcc-3.4 this patch changes the generated asm in large ways.

Odd.
--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"aart@kvack.org"> aart@kvack.org </a>