[lkml]   [2012]   [Jun]   [4]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
Messages in this thread
SubjectRe: [PATCH 0/5] Some vmevent fixes...
> KOSAKI, AFAIRC, you are a person who hates android low memory killer.
> Why do you hate it? If it solve problems I mentioned, do you have a concern, still?
> If so, please, list up.

1) it took tasklist_lock. (it was solved already)
2) hacky lowmem_deathpending_timeout
3) No ZONE awareness. global_page_state(), lowmem_minfree[] and shrink_slab interface
don't realize real memory footprint.
4) No memcg, cpuset awareness.
5) strange score calculation. the score is calculated from anon_rss+file_rss,
but oom killer only free anon_rss.
6) strange oom_score_adj overload
7) much duplicate code w/ oom_killer. we can make common helper function, I think.
8) John's fallocate(VOLATILE) is more cleaner.

 \ /
  Last update: 2012-06-04 22:41    [W:0.128 / U:60.528 seconds]
©2003-2018 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site