> What is the correct way of interpreting slabinfo? Quick summary from a
> mm guru? What do the two cols mean?
Well, this is part of the draft I am (heh) writing to document /proc
Clarifications, additions, etc, very welcome as this is mostly 'guessing',
since I haven't had a look to how mm works yet.
slabinfo
Information about the cache memory allocated for the various structures
held in the system. Each structure type is called a slab, and each
structure of that type is called an object.
%s cache name
(c_name)
%d active objects
%d total objects. These objects, albeit unused, are kept in the slab cache
so we don't have grow that particular slab in case we need more objects
-Following fields are only available #if SLAB_STATS-
%d active slabs
%d total slabs
%d pages per slab
%d the highest active object in the slab
(c_high_mark)
%d incremented each time the slab allocates space for an object
(c_num_allocations)
%d incremented each time the slab increases its size
(c_grown)
%d incremented each time the slab gets reaped (removed from the slab list)
(c_reaped)
%d incremented each time there's an error allocing or freeing kmem
(c_errors)
> Is this slabinfo gonna be usefull to fixing/diagnosing the current mm
> problems? Perhaps someone would like to collect a bunch of them (off
> list) and summarize?
Dunno, but if SLAB_STATS still works it could be sueful.
____/| Ragnar Hojland (ragnar@lightside.ddns.org) Fingerprint 94C4B
\ o.O| 2F0D27DE025BE2302C
=(_)= "Thou shalt not follow the NULL pointer for 104B78C56 B72F0822
U chaos and madness await thee at its end." hkp://keys.pgp.com
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.altern.org/andrebalsa/doc/lkml-faq.html