Re: [PATCH] slob: poor man's NUMA, take 2.

[Matt Mackall <mpm@selenic.com>]

On Wed, Jun 13, 2007 at 06:21:09PM +0900, Paul Mundt wrote:
> Here's an updated copy with the node variants always defined.
> 
> I've left the nid=-1 case in as the default for the non-node variants, as
> this is the approach also used by SLUB. alloc_pages() is special cased
> for NUMA, and takes the memory policy under advisement when doing the
> allocation, so the page ends up in a reasonable place.
> 

> +void *__kmalloc(size_t size, gfp_t gfp)
> +{
> +	return __kmalloc_node(size, gfp, -1);
> +}
>  EXPORT_SYMBOL(__kmalloc);

> +void *kmem_cache_alloc(struct kmem_cache *c, gfp_t flags)
> +{
> +	return kmem_cache_alloc_node(c, flags, -1);
> +}
>  EXPORT_SYMBOL(kmem_cache_alloc);

Now promote these guys to inlines in slab.h. At which point all the
new NUMA code become a no-op on !NUMA.

-- 
Mathematics is the supreme nostalgia of our time.

--
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:"dont@kvack.org"> email@kvack.org </a>