Messages in this thread | | | Date | Wed, 10 Dec 2008 23:08:24 +0900 | From | Daisuke Nishimura <> | Subject | Re: [RFC][RFT] memcg fix cgroup_mutex deadlock when cpuset reclaims memory |
| |
On Wed, 10 Dec 2008 18:36:07 +0530 Balbir Singh <balbir@linux.vnet.ibm.com> wrote:
> * Daisuke Nishimura <d-nishimura@mtf.biglobe.ne.jp> [2008-12-10 20:53:37]: > > > On Wed, 10 Dec 2008 17:18:36 +0900 > > KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> wrote: > > > > > On Wed, 10 Dec 2008 16:41:26 +0900 > > > Daisuke Nishimura <nishimura@mxp.nes.nec.co.jp> wrote: > > > > > > > On Wed, 10 Dec 2008 15:19:48 +0900, Daisuke Nishimura <nishimura@mxp.nes.nec.co.jp> wrote: > > > > > On Wed, 10 Dec 2008 10:49:47 +0530, Balbir Singh <balbir@linux.vnet.ibm.com> wrote: > > > > > > Hi, > > > > > > > > > > > > Here is a proposed fix for the memory controller cgroup_mutex deadlock > > > > > > reported. It is lightly tested and reviewed. I need help with review > > > > > > and test. Is the reported deadlock reproducible after this patch? A > > > > > > careful review of the cpuset impact will also be highly appreciated. > > > > > > > > > > > > From: Balbir Singh <balbir@linux.vnet.ibm.com> > > > > > > > > > > > > cpuset_migrate_mm() holds cgroup_mutex throughout the duration of > > > > > > do_migrate_pages(). The issue with that is that > > > > > > > > > > > > 1. It can lead to deadlock with memcg, as do_migrate_pages() > > > > > > enters reclaim > > > > > > 2. It can lead to long latencies, preventing users from creating/ > > > > > > destroying other cgroups anywhere else > > > > > > > > > > > > The patch holds callback_mutex through the duration of cpuset_migrate_mm() and > > > > > > gives up cgroup_mutex while doing so. > > > > > > > > > > > I agree changing cpuset_migrate_mm not to hold cgroup_mutex to fix the dead lock > > > > > is one choice, and it looks good to me at the first impression. > > > > > > > > > > But I'm not sure it's good to change cpuset(other subsystem) code because of memcg. > > > > > > > > > > Anyway, I'll test this patch and report the result tomorrow. > > > > > (Sorry, I don't have enough time today.) > > > > > > > > > Unfortunately, this patch doesn't seem enough. > > > > > > > > This patch can fix dead lock caused by "circular lock of cgroup_mutex", > > > > but cannot that of caused by "race between page reclaim and cpuset_attach(mpol_rebind_mm)". > > > > > > > > (The dead lock I fixed in memcg-avoid-dead-lock-caused-by-race-between-oom-and-cpuset_attach.patch > > > > was caused by "race between memcg's oom and mpol_rebind_mm, and was independent of hierarchy.) > > > > > > > > I attach logs I got in testing this patch. > > > > > > > Hmm, ok then, what you mention to is this race. > > > -- > > > cgroup_lock() > > > -> cpuset_attach() > > > -> down_write(&mm->mmap_sem); > > > > > > down_read() > > > -> page fault > > > -> reclaim in memcg > > > -> cgroup_lock(). > > > -- > > > What this patch tries to fix is this recursive locks > > > -- > > > cgroup_lock() > > > -> cpuset_attach() > > > -> cpuset_migrate_mm() > > > -> charge to migration > > > -> go to reclaim and meet cgroup_lock. > > > -- > > > > > > > > > Right ? > > > > > Yes. > > Thank you for explaining in detail. > > > > Sorry, I don't understand the context, I am unable to figure out > > 1. How to reproduce the problem that Daisuke-San reported Ah.. sorry.
1) mount memory cgroup and cpuset. (I mount them on different mount points, but I think this can also happen even when mounting on the same hierarchy.) 2) make directories 2-1) memory - make a directory(/cgroup/memory/01) - set memory.limit_in_bytes(no need to set memsw.limit_in_bytes). - enable hierarchy(no need to make a child). 2-2) cpuset - make 2(at least) directories(/cgroup/cpuset/01,02) - set different "mems". - set memory_migrate on. 3) attach shell to /cgroup/*/01 4) run some programs enough to cause swap out/in 5) trigger page migration by cpuset between 01 and 02 repeatedly. I think Documentation/controllers/memcg_test.txt would help.
feel free to ask me if you need additional information.
> 2. Whether the patch is correct or causing more problems or needs more > stuff to completely fix the race. > I should consider more to tell whether it's all right to release cgroup_mutex under attach_task, but some more stuff is needed at least.
Thanks, Daisuke Nishimura.
| |