Re: Kswapd in 3.2.0-rc5 is a CPU hog

[Dave Chinner]

On Tue, Dec 27, 2011 at 01:44:05PM +0900, KAMEZAWA Hiroyuki wrote:
> To me,  it seems kswapd does usual work...reclaim small memory until free
> gets enough. And it seems 'dd' allocates its memory from ZONE_DMA32 because
> of gfp_t fallbacks.
> 
> 
> Memo.
> 
> 1. why shrink_slab() should be called per zone, which is not zone aware.
>    Isn't it enough to call it per priority ?

It is intended that it should be zone aware, but the current
shrinkers only have global LRUs and hence cannot discriminate
between objects from different zones easily. And if only a single
node/zone is being scanned, then we still have to call shirnk_slab()
to try to free objects in that zone/node, despite it's current
global scope.

I have some prototype patches that make the major slab caches and
shrinkers zone/node aware - that is the eventual goal here - but
first all the major slab cache LRUs need to be converted to be node
aware first. Then we can pass a nodemask into shrink_slab() and down
to the shrinkers so that those that have per-node LRUs can scan only
the appropriate nodes for objects to free. This is someting that I'm
working on in my spare time, but I have very little of that at the
moment, unfortunately.

> 2. what spinlock contention that perf showed ?
>    And if shrink_slab() doesn't consume cpu as trace shows, why perf 
>    says shrink_slab() is heavy..

There isn't any spin lock contention - it's just showing how
expensive locking superblocks is when it's being done every few
microseconds for no good reason.

> 3. because 8/9 of memory is in DMA32, calling shrink_slab() frequently
>    at scanning NORMAL seems to be time wasting.

Especially as the shrink_slab() calls are returning zero pages freed
every single time (i.e. the slab caches are empty). kswapd needs to
back off here, I think, or free more memory at a time. Only freeing
100 pages at a time is pretty inefficient, esp. as we have 4 orders
of magnitude more pages on the LRU and that is consuming >90% of
RAM...

Cheers,

Dave.
-- 
Dave Chinner
david@xxxxxxxxxxxxx
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/