Thread (6 messages) 6 messages, 2 authors, 2018-08-23

Re: [PATCH] mm, oom: Always call tlb_finish_mmu().

From: Michal Hocko <mhocko@suse.com>
Date: 2018-08-23 12:00:01

On Thu 23-08-18 20:30:48, Tetsuo Handa wrote:
Commit 93065ac753e44438 ("mm, oom: distinguish blockable mode for mmu
notifiers") added "continue;" without calling tlb_finish_mmu(). I don't
know whether tlb_flush_pending imbalance causes problems other than
extra cost, but at least it looks strange.
tlb_flush_pending has mm scope and it would confuse
mm_tlb_flush_pending. At least ptep_clear_flush could get confused and
flush unnecessarily for prot_none entries AFAICS. Other paths shouldn't
trigger for oom victims. Even ptep_clear_flush is unlikely to happen.
So nothing really earth shattering but I do agree that it looks weird
and should be fixed.
More worrisome part in that patch is that I don't know whether using
trylock if blockable == false at entry is really sufficient. For example,
since __gnttab_unmap_refs_async() from gnttab_unmap_refs_async() from
gnttab_unmap_refs_sync() from __unmap_grant_pages() from
unmap_grant_pages() from unmap_if_in_range() from mn_invl_range_start()
involves schedule_delayed_work() which could be blocked on memory
allocation under OOM situation, wait_for_completion() from
gnttab_unmap_refs_sync() might deadlock? I don't know...
Not really sure why this is in the changelog as it is unrelated to the
fix. Anyway let me try to check...

OK, so I've added in_range(map, start, end) check to not go that
direction. But for some reason that check doesn't consider blockable
value. So it looks definitely wrong. I must have screwed up when
rebasing or something. Thanks for catching that up. I will send a fix.
quoted hunk ↗ jump to hunk
Signed-off-by: Tetsuo Handa <penguin-kernel@I-love.SAKURA.ne.jp>
---
 mm/oom_kill.c | 1 +
 1 file changed, 1 insertion(+)
diff --git a/mm/oom_kill.c b/mm/oom_kill.c
index b5b25e4..4f431c1 100644
--- a/mm/oom_kill.c
+++ b/mm/oom_kill.c
@@ -522,6 +522,7 @@ bool __oom_reap_task_mm(struct mm_struct *mm)
 
 			tlb_gather_mmu(&tlb, mm, start, end);
 			if (mmu_notifier_invalidate_range_start_nonblock(mm, start, end)) {
+				tlb_finish_mmu(&tlb, start, end);
 				ret = false;
 				continue;
 			}
-- 
1.8.3.1
-- 
Michal Hocko
SUSE Labs
Keyboard shortcuts
hback out one level
jnext message in thread
kprevious message in thread
ldrill in
Escclose help / fold thread tree
?toggle this help