From: Michal Hocko <email@example.com> To: Tetsuo Handa <firstname.lastname@example.org> Cc: Andrew Morton <email@example.com>, Johannes Weiner <firstname.lastname@example.org>, David Rientjes <email@example.com>, firstname.lastname@example.org, Yong-Taek Lee <email@example.com>, Paul McKenney <firstname.lastname@example.org>, Linus Torvalds <email@example.com> Subject: Re: [PATCH] mm, oom: Tolerate processes sharing mm with different view of oom_score_adj. Date: Wed, 16 Jan 2019 14:41:31 +0100 [thread overview] Message-ID: <20190116134131.GP24149@dhcp22.suse.cz> (raw) In-Reply-To: <firstname.lastname@example.org> On Wed 16-01-19 22:32:50, Tetsuo Handa wrote: > On 2019/01/16 21:19, Michal Hocko wrote: > > On Wed 16-01-19 20:30:25, Tetsuo Handa wrote: > >> On 2019/01/16 20:09, Michal Hocko wrote: > >>> On Wed 16-01-19 19:55:21, Tetsuo Handa wrote: > >>>> This patch reverts both commit 44a70adec910d692 ("mm, oom_adj: make sure > >>>> processes sharing mm have same view of oom_score_adj") and commit > >>>> 97fd49c2355ffded ("mm, oom: kill all tasks sharing the mm") in order to > >>>> close a race and reduce the latency at __set_oom_adj(), and reduces the > >>>> warning at __oom_kill_process() in order to minimize the latency. > >>>> > >>>> Commit 36324a990cf578b5 ("oom: clear TIF_MEMDIE after oom_reaper managed > >>>> to unmap the address space") introduced the worst case mentioned in > >>>> 44a70adec910d692. But since the OOM killer skips mm with MMF_OOM_SKIP set, > >>>> only administrators can trigger the worst case. > >>>> > >>>> Since 44a70adec910d692 did not take latency into account, we can hold RCU > >>>> for minutes and trigger RCU stall warnings by calling printk() on many > >>>> thousands of thread groups. Even without calling printk(), the latency is > >>>> mentioned by Yong-Taek Lee . And I noticed that 44a70adec910d692 is > >>>> racy, and trying to fix the race will require a global lock which is too > >>>> costly for rare events. > >>>> > >>>> If the worst case in 44a70adec910d692 happens, it is an administrator's > >>>> request. Therefore, tolerate the worst case and speed up __set_oom_adj(). > >>> > >>> I really do not think we care about latency. I consider the overal API > >>> sanity much more important. Besides that the original report you are > >>> referring to was never exaplained/shown to represent real world usecase. > >>> oom_score_adj is not really a an interface to be tweaked in hot paths. > >> > >> I do care about the latency. Holding RCU for more than 2 minutes is insane. > > > > Creating 8k threads could be considered insane as well. But more > > seriously. I absolutely do not insist on holding a single RCU section > > for the whole operation. But that doesn't really mean that we want to > > revert these changes. for_each_process is by far not only called from > > this path. > > Unlike check_hung_uninterruptible_tasks() where failing to resume after > breaking RCU section is tolerable, failing to resume after breaking RCU > section for __set_oom_adj() is not tolerable; it leaves the possibility > of different oom_score_adj. Then make sure that no threads are really missed. Really I fail to see what you are actually arguing about. for_each_process is expensive. No question about that. If you can replace it for this specific and odd usecase then go ahead. But there is absolutely zero reason to have a broken oom_score_adj semantic just because somebody might have thousands of threads and want to update the score faster. > Unless it is inevitable (e.g. SysRq-t), I think > that calling printk() on each thread from RCU section is a poor choice. > > What if thousands of threads concurrently called __set_oom_adj() when > each __set_oom_adj() call involves printk() on thousands of threads > which can take more than 2 minutes? How long will it take to complete? I really do not mind removing printk if that is what really bothers users. The primary purpose of this printk was to catch users who wouldn't expect this change. There were exactly zero. -- Michal Hocko SUSE Labs
next prev parent reply other threads:[~2019-01-16 13:41 UTC|newest] Thread overview: 14+ messages / expand[flat|nested] mbox.gz Atom feed top 2019-01-16 10:55 Tetsuo Handa 2019-01-16 11:09 ` Michal Hocko 2019-01-16 11:30 ` Tetsuo Handa 2019-01-16 12:19 ` Michal Hocko 2019-01-16 13:32 ` Tetsuo Handa 2019-01-16 13:41 ` Michal Hocko [this message] 2019-01-17 10:40 ` Tetsuo Handa 2019-01-17 15:51 ` Michal Hocko 2019-01-30 22:49 ` [PATCH v2] " Tetsuo Handa 2019-01-31 7:11 ` Michal Hocko 2019-01-31 20:59 ` Tetsuo Handa 2019-02-01 9:14 ` Michal Hocko 2019-02-02 11:06 ` Tetsuo Handa 2019-02-11 15:07 ` Michal Hocko
Reply instructions: You may reply publicly to this message via plain-text email using any one of the following methods: * Save the following mbox file, import it into your mail client, and reply-to-all from there: mbox Avoid top-posting and favor interleaved quoting: https://en.wikipedia.org/wiki/Posting_style#Interleaved_style * Reply using the --to, --cc, and --in-reply-to switches of git-send-email(1): git send-email \ --in-reply-to=20190116134131.GP24149@dhcp22.suse.cz \ --email@example.com \ --firstname.lastname@example.org \ --email@example.com \ --firstname.lastname@example.org \ --email@example.com \ --firstname.lastname@example.org \ --email@example.com \ --firstname.lastname@example.org \ --email@example.com \ --subject='Re: [PATCH] mm, oom: Tolerate processes sharing mm with different view of oom_score_adj.' \ /path/to/YOUR_REPLY https://kernel.org/pub/software/scm/git/docs/git-send-email.html * If your mail client supports setting the In-Reply-To header via mailto: links, try the mailto: link
This is an external index of several public inboxes, see mirroring instructions on how to clone and mirror all data and code used by this external index.