Re: [patch] slub: change the formula which calculates min_objects based on number of processors

Re: [patch] slub: change the formula which calculates min_objects based on number of processors

[Pekka Enberg]

Zhang, Yanmin wrote:
> Current formula to calculate min_objects based on number of processors is 
> '4 * fls(nr_cpu_ids)', which is not the best optimization on 16-core tigerton.
> If I add 4 to its result, hackbench result is better.
> 
> On 16-core tigerton, by run
> ./hackbench 100 process 2000
> results are:
> 1) 2.6.25-rc6slab: 23.5seconds
> 2) 2.6.25-rc7SLUB+slub_min_objects=20: 31seconds
> 3) 2.6.25-rc7SLUB+slub_min_objects=24: 23.5seconds
> 
> So adding 4 to the output of '4 * fls(nr_cpu_ids)' could get the similar result
> like CONFIG_SLAB=y.

Applied, thanks!

--
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>

Re: [patch 04/12] slub: Add kmem_cache_order_objects struct

[Christoph Lameter]

On Mon, 7 Apr 2008, Pekka Enberg wrote:

> Hi Christoph,
> 
> Christoph Lameter wrote:
> > @@ -1143,7 +1165,7 @@ static struct page *new_slab(struct kmem
> >  	start = page_address(page);
> >   	if (unlikely(s->flags & SLAB_POISON))
> > -		memset(start, POISON_INUSE, PAGE_SIZE << s->order);
> > +		memset(start, POISON_INUSE, PAGE_SIZE << oo_order(oo));
> 
> This should be compound_order(page) as allocate_slab() can fall back to
> smaller page order.

Ack. I distinctly remember fixing this once before. Sigh.



--
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>