Re: [PATCH] [RFC] CMA: clear buffer-head lru before page migration

[Gioh Kim <gioh.kim@lge.com>]

2014-07-15 오전 5:37, Andrew Morton 쓴 글:
> On Mon, 14 Jul 2014 16:02:25 +0900 Joonsoo Kim <iamjoonsoo.kim@lge.com> wrote:
>
>> On Tue, Jul 08, 2014 at 06:46:31PM +0200, Michal Nazarewicz wrote:
>>> On Mon, Jul 07 2014, Andrew Morton <akpm@linux-foundation.org> wrote:
>>>> What I proposed is that CMA call invalidate_bh_lrus() right at the
>>>> outset.  Something along the lines of
>>>>
>>>> --- a/mm/page_alloc.c~a
>>>> +++ a/mm/page_alloc.c
>>>> @@ -6329,6 +6329,14 @@ int alloc_contig_range(unsigned long sta
>>>>   	};
>>>>   	INIT_LIST_HEAD(&cc.migratepages);
>>>>
>>>> +#ifdef CONFIG_CMA
>>>> +	/*
>>>> +	 * Comment goes here
>>>> +	 */
>>>> +	if (migratetype == MIGRATE_CMA)
>>>> +		invalidate_bh_lrus();
>>>> +#endif
>>>> +
>>>
>>> This seems reasonable, except I think it should go after
>>> start_isolate_page_range call because otherwise there's no guarantee
>>> that someone won't grab those pages back.
>>>
>>> Also to avoid the #ifdef perhaps we want this as well:
>>
>> I think that we just want to remove ifdef CONFIG_CMA on above code
>> snippet, because invalidate_bh_lrus() would also help user of
>> alloc_contig_range() with MIGRATE_MOVABLE.
>
> That's what I believe also.  I pinged Mel and Johannes off-list and Mel
> said "I hit it, the invalidation cost wasn't worth it for a THP alloc".
>
> So hm.  I do think it's worth additional investigation but some careful
> testing would be needed to demonstrate that it's worthwhile.  If the
> invalidation cost is hurting then perhaps additional logic will be
> needed to perform the invalidation only as a last-resort thing.
>
>
>
Adding invalidate_bh_lrus() between start_isolate_page_range and
__alloc_contig_migrate_range is working well on my platform.

But I'd like to test performance before sending patch.
Would anybody recommend me a benchmark tool?



--
To unsubscribe from this list: send the line "unsubscribe linux-fsdevel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html