linux-mm.kvack.org archive mirror
 help / color / mirror / Atom feed
* shrink_slab(), shrinkers and aggresivity
@ 2013-09-10 17:28 Robert Jarzmik
  0 siblings, 0 replies; only message in thread
From: Robert Jarzmik @ 2013-09-10 17:28 UTC (permalink / raw)
  To: linux-mm

Hi,

I was wondering if there was such a notion as "shrinkers aggresivity".
Or, in other words, I wondered why shrink_slab(), in its parameters, doesn't
have the priority, ie. the (struct scan_control)->priority.

The usecase I have in mind would be a shrinker which behaves differently,
depending on this priority :
 - if priority is low, only drop a subset of its objects, the "cold objects"
 - if priority is high, drop each and every possible object

In a GPU cache for example, they are objects that are not used anymore in the
GPU, but some are still mapped into the GPU's MMU table. The GPU MMU
manipulation being costly, such a shrinker would :
 - on low priority, consider only the GPU MMU unmapped objects
 - on high prioriy, consider all GPU objects

Is it in the shrinker definition that no priority should ever be considered, is
it silly to consider having priority in (struct shrink_control) ?

Thanks in advance for your explanation.

-- 
Robert

--
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>

^ permalink raw reply	[flat|nested] only message in thread

only message in thread, other threads:[~2013-09-10 17:28 UTC | newest]

Thread overview: (only message) (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2013-09-10 17:28 shrink_slab(), shrinkers and aggresivity Robert Jarzmik

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox