Files
linux/include/linux
Sean Christopherson c2655835fd mm/mmu_notifiers: ensure range_end() is paired with range_start()
If one or more notifiers fails .invalidate_range_start(), invoke
.invalidate_range_end() for "all" notifiers.  If there are multiple
notifiers, those that did not fail are expecting _start() and _end() to
be paired, e.g.  KVM's mmu_notifier_count would become imbalanced.
Disallow notifiers that can fail _start() from implementing _end() so
that it's unnecessary to either track which notifiers rejected _start(),
or had already succeeded prior to a failed _start().

Note, the existing behavior of calling _start() on all notifiers even
after a previous notifier failed _start() was an unintented "feature".
Make it canon now that the behavior is depended on for correctness.

As of today, the bug is likely benign:

  1. The only caller of the non-blocking notifier is OOM kill.
  2. The only notifiers that can fail _start() are the i915 and Nouveau
     drivers.
  3. The only notifiers that utilize _end() are the SGI UV GRU driver
     and KVM.
  4. The GRU driver will never coincide with the i195/Nouveau drivers.
  5. An imbalanced kvm->mmu_notifier_count only causes soft lockup in the
     _guest_, and the guest is already doomed due to being an OOM victim.

Fix the bug now to play nice with future usage, e.g.  KVM has a
potential use case for blocking memslot updates in KVM while an
invalidation is in-progress, and failure to unblock would result in said
updates being blocked indefinitely and hanging.

Found by inspection.  Verified by adding a second notifier in KVM that
periodically returns -EAGAIN on non-blockable ranges, triggering OOM,
and observing that KVM exits with an elevated notifier count.

Link: https://lkml.kernel.org/r/20210311180057.1582638-1-seanjc@google.com
Fixes: 93065ac753 ("mm, oom: distinguish blockable mode for mmu notifiers")
Signed-off-by: Sean Christopherson <seanjc@google.com>
Suggested-by: Jason Gunthorpe <jgg@ziepe.ca>
Reviewed-by: Jason Gunthorpe <jgg@nvidia.com>
Cc: David Rientjes <rientjes@google.com>
Cc: Ben Gardon <bgardon@google.com>
Cc: Michal Hocko <mhocko@suse.com>
Cc: "Jérôme Glisse" <jglisse@redhat.com>
Cc: Andrea Arcangeli <aarcange@redhat.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: Dimitri Sivanich <dimitri.sivanich@hpe.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2021-03-25 09:22:55 -07:00
..
2021-02-02 00:16:57 +01:00
2020-12-09 19:26:02 -06:00
2021-01-23 14:57:21 +01:00
2021-01-24 14:27:17 +01:00
2021-02-26 09:41:03 -08:00
2021-02-17 14:07:48 +01:00
2020-12-19 11:18:38 -08:00
2021-01-21 14:06:00 -07:00
2021-01-16 15:12:06 -05:00
2021-01-18 14:26:51 +01:00
2021-02-08 12:28:07 +01:00
2020-12-10 12:42:59 -06:00
2021-02-11 13:24:44 -08:00
2021-01-27 12:27:36 +01:00
2020-12-15 16:19:31 +01:00
2021-02-26 09:41:03 -08:00
2020-12-15 16:19:31 +01:00
2021-02-17 14:12:42 +01:00
2021-01-21 16:16:10 +00:00
2021-02-26 09:41:02 -08:00
2021-02-26 09:41:03 -08:00
2021-01-22 15:09:42 +01:00
2021-01-14 11:20:17 +01:00
2021-02-26 09:41:03 -08:00
2021-03-13 11:27:30 -08:00
2021-01-04 11:42:21 -05:00
2021-02-16 16:11:14 -05:00
2021-02-02 10:26:12 +01:00
2021-02-17 13:17:49 -08:00
2021-02-26 09:40:59 -08:00
2021-01-21 14:06:00 -07:00
2021-02-03 19:05:50 +01:00
2021-01-24 14:27:17 +01:00
2021-01-24 14:27:20 +01:00
2021-02-26 09:41:03 -08:00
2020-12-03 11:20:52 +01:00
2021-01-16 23:19:26 +01:00
2020-12-07 17:12:49 -07:00
2020-12-10 16:17:15 +01:00
2021-01-28 00:35:03 +01:00
2020-12-10 10:45:36 +01:00
2021-03-02 17:25:46 -07:00
2021-02-20 10:13:32 -05:00
2021-01-06 16:24:59 -08:00
2021-02-09 12:27:29 -05:00
2020-12-04 16:54:35 +01:00
2021-02-13 17:17:53 +01:00
2021-01-21 16:16:10 +00:00
2021-02-01 13:20:07 -07:00
2021-01-18 10:52:41 +01:00
2021-02-09 12:15:07 +01:00
2021-01-21 14:06:00 -07:00
2021-01-24 14:27:17 +01:00
2021-02-08 22:58:55 +01:00