[linux-mm-cc] [PATCH 00/12] Avoid OOM-Killer with Ccache

IKEDA Munehiro m-ikeda at ds.jp.nec.com
Fri Jul 20 07:12:00 EDT 2007


> Below are memorandum about remained work related these patches, 
> in my mind.
> 
> Simply freeing behavior should be applied code path of process 
> exiting same as memory reclaiming path.
> 
> Current chunk_head locking mechanism is based on spinlock.
> Page locking mechanism in mainline kernel is based on wait-on-bit-
> lock.
> Should we change the mechanism like mainline kernel?
> Probably we should evaluate them and decide which is preferable 
> because it may affects performance, IMHO.
> (I have a trial implementation already in fact.)

I forgot to mention what I think very important.

(1)
These patches can avoid OOM-Killer caused by that Ccache is full
of fs_backed chunks at least on my system, which is UP x86.
More tests are needed on SMP or architectures other than x86.

(2)
If you "cat /proc/ccache_stats" reading many files, you can
probably see so many pages are freed when Ccache is shrunk,
even though only 1 or 2 pages are needed at the time.
This is because of the fragmentation of chunks, which means
compressed data of a page is stored in many divided chunks
which are located in many pages.
For more efficiency and higher performance, avoiding fragmen-
tation like it is next important target of development, IMHO.


Best regards,
IKEDA, Munehiro


More information about the linux-mm-cc mailing list