mm/shmem: fix infinite loop when swap in shmem error at swapoff time

When swap in shmem error at swapoff time, there would be a infinite loop
in the while loop in shmem_unuse_inode().  It's because swapin error is
deliberately ignored now and thus info->swapped will never reach 0.  So we
can't escape the loop in shmem_unuse().

In order to fix the issue, swapin_error entry is stored in the mapping
when swapin error occurs.  So the swapcache page can be freed and the user
won't end up with a permanently mounted swap because a sector is bad.  If
the page is accessed later, the user process will be killed so that
corrupted data is never consumed.  On the other hand, if the page is never
accessed, the user won't even notice it.

Link: https://lkml.kernel.org/r/20220519125030.21486-5-linmiaohe@huawei.com
Signed-off-by: Miaohe Lin <linmiaohe@huawei.com>
Reported-by: Naoya Horiguchi <naoya.horiguchi@nec.com>
Reviewed-by: Naoya Horiguchi <naoya.horiguchi@nec.com>
Cc: Alistair Popple <apopple@nvidia.com>
Cc: David Hildenbrand <david@redhat.com>
Cc: David Howells <dhowells@redhat.com>
Cc: Hugh Dickins <hughd@google.com>
Cc: Matthew Wilcox (Oracle) <willy@infradead.org>
Cc: NeilBrown <neilb@suse.de>
Cc: Peter Xu <peterx@redhat.com>
Cc: Ralph Campbell <rcampbell@nvidia.com>
Cc: Suren Baghdasaryan <surenb@google.com>
Cc: Vlastimil Babka <vbabka@suse.cz>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm/madvise: free hwpoison and swapin error entry in madvise_free_pte_range

Once the MADV_FREE operation has succeeded, callers can expect they might
get zero-fill pages if accessing the memory again.  Therefore it should be
safe to delete the hwpoison entry and swapin error entry.  There is no
reason to kill the process if it has called MADV_FREE on the range.

Link: https://lkml.kernel.org/r/20220519125030.21486-4-linmiaohe@huawei.com
Signed-off-by: Miaohe Lin <linmiaohe@huawei.com>
Suggested-by: Alistair Popple <apopple@nvidia.com>
Acked-by: David Hildenbrand <david@redhat.com>
Reviewed-by: Naoya Horiguchi <naoya.horiguchi@nec.com>
Cc: David Howells <dhowells@redhat.com>
Cc: Hugh Dickins <hughd@google.com>
Cc: Matthew Wilcox (Oracle) <willy@infradead.org>
Cc: NeilBrown <neilb@suse.de>
Cc: Peter Xu <peterx@redhat.com>
Cc: Ralph Campbell <rcampbell@nvidia.com>
Cc: Suren Baghdasaryan <surenb@google.com>
Cc: Vlastimil Babka <vbabka@suse.cz>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm/swapfile: fix lost swap bits in unuse_pte()

This is observed by code review only but not any real report.

When we turn off swapping we could have lost the bits stored in the swap
ptes.  The new rmap-exclusive bit is fine since that turned into a page
flag, but not for soft-dirty and uffd-wp.  Add them.

Link: https://lkml.kernel.org/r/20220519125030.21486-3-linmiaohe@huawei.com
Signed-off-by: Miaohe Lin <linmiaohe@huawei.com>
Suggested-by: Peter Xu <peterx@redhat.com>
Reviewed-by: David Hildenbrand <david@redhat.com>
Cc: Alistair Popple <apopple@nvidia.com>
Cc: David Howells <dhowells@redhat.com>
Cc: Hugh Dickins <hughd@google.com>
Cc: Matthew Wilcox (Oracle) <willy@infradead.org>
Cc: Naoya Horiguchi <naoya.horiguchi@nec.com>
Cc: NeilBrown <neilb@suse.de>
Cc: Ralph Campbell <rcampbell@nvidia.com>
Cc: Suren Baghdasaryan <surenb@google.com>
Cc: Vlastimil Babka <vbabka@suse.cz>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm/swapfile: unuse_pte can map random data if swap read fails

Patch series "A few fixup patches for mm", v4.

This series contains a few patches to avoid mapping random data if swap
read fails and fix lost swap bits in unuse_pte.  Also we free hwpoison and
swapin error entry in madvise_free_pte_range and so on.  More details can
be found in the respective changelogs.

This patch (of 5):

There is a bug in unuse_pte(): when swap page happens to be unreadable,
page filled with random data is mapped into user address space.  In case
of error, a special swap entry indicating swap read fails is set to the
page table.  So the swapcache page can be freed and the user won't end up
with a permanently mounted swap because a sector is bad.  And if the page
is accessed later, the user process will be killed so that corrupted data
is never consumed.  On the other hand, if the page is never accessed, the
user won't even notice it.

Link: https://lkml.kernel.org/r/20220519125030.21486-1-linmiaohe@huawei.com
Link: https://lkml.kernel.org/r/20220519125030.21486-2-linmiaohe@huawei.com
Signed-off-by: Miaohe Lin <linmiaohe@huawei.com>
Acked-by: David Hildenbrand <david@redhat.com>
Cc: Hugh Dickins <hughd@google.com>
Cc: Matthew Wilcox (Oracle) <willy@infradead.org>
Cc: Vlastimil Babka <vbabka@suse.cz>
Cc: David Howells <dhowells@redhat.com>
Cc: NeilBrown <neilb@suse.de>
Cc: Alistair Popple <apopple@nvidia.com>
Cc: Suren Baghdasaryan <surenb@google.com>
Cc: Peter Xu <peterx@redhat.com>
Cc: Ralph Campbell <rcampbell@nvidia.com>
Cc: Naoya Horiguchi <naoya.horiguchi@nec.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

selftests: memcg: factor out common parts of memory.{low,min} tests

The memory protection test setup and runtime is almost equal for
memory.low and memory.min cases.

It makes modification of the common parts prone to mistakes, since the
protections are similar not only in setup but also in principle, factor
the common part out.

Past exceptions between the tests:
- missing memory.min is fine (kept),
- test_memcg_low protected orphaned pagecache (adapted like
  test_memcg_min and we keep the processes of protected memory running).

The evaluation in two tests is different (OOM of allocator vs low events
of protégés), this is kept different.

Link: https://lkml.kernel.org/r/20220518161859.21565-6-mkoutny@suse.com
Signed-off-by: Michal Koutný <mkoutny@suse.com>
Acked-by: Roman Gushchin <roman.gushchin@linux.dev>
CC: Johannes Weiner <hannes@cmpxchg.org>
Cc: Michal Hocko <mhocko@kernel.org>
Cc: Shakeel Butt <shakeelb@google.com>
Cc: Richard Palethorpe <rpalethorpe@suse.de>
Cc: David Vernet <void@manifault.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

selftests: memcg: remove protection from top level memcg

The reclaim is triggered by memory limit in a subtree, therefore the
testcase does not need configured protection against external reclaim.

Also, correct respective comments.

Link: https://lkml.kernel.org/r/20220518161859.21565-5-mkoutny@suse.com
Signed-off-by: Michal Koutný <mkoutny@suse.com>
Acked-by: Roman Gushchin <roman.gushchin@linux.dev>
Cc: David Vernet <void@manifault.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: Michal Hocko <mhocko@kernel.org>
Cc: Richard Palethorpe <rpalethorpe@suse.de>
Cc: Shakeel Butt <shakeelb@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

selftests: memcg: adjust expected reclaim values of protected cgroups

The numbers are not easy to derive in a closed form (certainly mere
protections ratios do not apply), therefore use a simulation to obtain
expected numbers.

Link: https://lkml.kernel.org/r/20220518161859.21565-4-mkoutny@suse.com
Signed-off-by: Michal Koutný <mkoutny@suse.com>
Acked-by: Roman Gushchin <roman.gushchin@linux.dev>
Cc: David Vernet <void@manifault.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: Michal Hocko <mhocko@kernel.org>
Cc: Richard Palethorpe <rpalethorpe@suse.de>
Cc: Shakeel Butt <shakeelb@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

selftests: memcg: expect no low events in unprotected sibling

This is effectively a revert of commit cdc69458a5f3 ("cgroup: account for
memory_recursiveprot in test_memcg_low()").  The case test_memcg_low will
fail with memory_recursiveprot until resolved in reclaim code.

However, this patch preserves the existing helpers and variables for later
uses.

Link: https://lkml.kernel.org/r/20220518161859.21565-3-mkoutny@suse.com
Signed-off-by: Michal Koutný <mkoutny@suse.com>
Reviewed-by: David Vernet <void@manifault.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: Michal Hocko <mhocko@kernel.org>
Cc: Richard Palethorpe <rpalethorpe@suse.de>
Cc: Roman Gushchin <roman.gushchin@linux.dev>
Cc: Shakeel Butt <shakeelb@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

selftests: memcg: fix compilation

Patch series "memcontrol selftests fixups", v2.

Flushing the patches to make memcontrol selftests check the events
behavior we had consensus about (test_memcg_low fails).

(test_memcg_reclaim, test_memcg_swap_max fail for me now but it's present
even before the refactoring.)

The two bigger changes are:
- adjustment of the protected values to make tests succeed with the given
  tolerance,
- both test_memcg_low and test_memcg_min check protection of memory in
  populated cgroups (actually as per Documentation/admin-guide/cgroup-v2.rst
  memory.min should not apply to empty cgroups, which is not the case
  currently. Therefore I unified tests with the populated case in order to to
  bring more broken tests).

This patch (of 5):

This fixes mis-applied changes from commit 72b1e03aa725 ("cgroup: account
for memory_localevents in test_memcg_oom_group_leaf_events()").

Link: https://lkml.kernel.org/r/20220518161859.21565-1-mkoutny@suse.com
Link: https://lkml.kernel.org/r/20220518161859.21565-2-mkoutny@suse.com
Signed-off-by: Michal Koutný <mkoutny@suse.com>
Reviewed-by: David Vernet <void@manifault.com>
Acked-by: Roman Gushchin <roman.gushchin@linux.dev>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: Michal Hocko <mhocko@kernel.org>
Cc: Richard Palethorpe <rpalethorpe@suse.de>
Cc: Shakeel Butt <shakeelb@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm/z3fold: fix z3fold_page_migrate races with z3fold_map

Think about the below scenario:

CPU1 CPU2
 z3fold_page_migrate z3fold_map
  z3fold_page_trylock
  ...
  z3fold_page_unlock
  /* slots still points to old zhdr*/
 get_z3fold_header
  get slots from handle
  get old zhdr from slots
  z3fold_page_trylock
  return *old* zhdr
  encode_handle(new_zhdr, FIRST|LAST|MIDDLE)
  put_page(page) /* zhdr is freed! */
 but zhdr is still used by caller!

z3fold_map can map freed z3fold page and lead to use-after-free bug.  To
fix it, we add PAGE_MIGRATED to indicate z3fold page is migrated and soon
to be released.  So get_z3fold_header won't return such page.

Link: https://lkml.kernel.org/r/20220429064051.61552-10-linmiaohe@huawei.com
Fixes: 1f862989b04a ("mm/z3fold.c: support page migration")
Signed-off-by: Miaohe Lin <linmiaohe@huawei.com>
Reviewed-by: Vitaly Wool <vitaly.wool@konsulko.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm/z3fold: fix z3fold_reclaim_page races with z3fold_free

Think about the below scenario:

CPU1 CPU2
z3fold_reclaim_page z3fold_free
 spin_lock(&pool->lock)  get_z3fold_header -- hold page_lock
 kref_get_unless_zero
 kref_put--zhdr->refcount can be 1 now
 !z3fold_page_trylock
  kref_put -- zhdr->refcount is 0 now
   release_z3fold_page
    WARN_ON(!list_empty(&zhdr->buddy)); -- we're on buddy now!
    spin_lock(&pool->lock); -- deadlock here!

z3fold_reclaim_page might race with z3fold_free and will lead to pool lock
deadlock and zhdr buddy non-empty warning.  To fix this, defer getting the
refcount until page_lock is held just like what __z3fold_alloc does.  Note
this has the side effect that we won't break the reclaim if we meet a soon
to be released z3fold page now.

Link: https://lkml.kernel.org/r/20220429064051.61552-9-linmiaohe@huawei.com
Fixes: dcf5aedb24f8 ("z3fold: stricter locking and more careful reclaim")
Signed-off-by: Miaohe Lin <linmiaohe@huawei.com>
Reviewed-by: Vitaly Wool <vitaly.wool@konsulko.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm/z3fold: always clear PAGE_CLAIMED under z3fold page lock

Think about the below race window:

CPU1 CPU2
z3fold_reclaim_page z3fold_free
 test_and_set_bit PAGE_CLAIMED
 failed to reclaim page
 z3fold_page_lock(zhdr);
 add back to the lru list;
 z3fold_page_unlock(zhdr);
 get_z3fold_header
 page_claimed=test_and_set_bit PAGE_CLAIMED

 clear_bit(PAGE_CLAIMED, &page->private);

 if (!page_claimed) /* it's false true */
  free_handle is not called

free_handle won't be called in this case. So z3fold_buddy_slots will leak.
Fix it by always clear PAGE_CLAIMED under z3fold page lock.

Link: https://lkml.kernel.org/r/20220429064051.61552-8-linmiaohe@huawei.com
Signed-off-by: Miaohe Lin <linmiaohe@huawei.com>
Reviewed-by: Vitaly Wool <vitaly.wool@konsulko.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm/z3fold: put z3fold page back into unbuddied list when reclaim or migration fails

When doing z3fold page reclaim or migration, the page is removed from
unbuddied list.  If reclaim or migration succeeds, it's fine as page is
released.  But in case it fails, the page is not put back into unbuddied
list now.  The page will be leaked until next compaction work, reclaim or
migration is done.

Link: https://lkml.kernel.org/r/20220429064051.61552-7-linmiaohe@huawei.com
Signed-off-by: Miaohe Lin <linmiaohe@huawei.com>
Reviewed-by: Vitaly Wool <vitaly.wool@konsulko.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

revert "mm/z3fold.c: allow __GFP_HIGHMEM in z3fold_alloc"

Revert commit f1549cb5ab2b ("mm/z3fold.c: allow __GFP_HIGHMEM in
z3fold_alloc").

z3fold can't support GFP_HIGHMEM page now.  page_address is used directly
at all places.  Moreover, z3fold_header is on per cpu unbuddied list which
could be accessed anytime.  So we should remove the support of GFP_HIGHMEM
allocation for z3fold.

Link: https://lkml.kernel.org/r/20220429064051.61552-6-linmiaohe@huawei.com
Signed-off-by: Miaohe Lin <linmiaohe@huawei.com>
Cc: Vitaly Wool <vitaly.wool@konsulko.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm/z3fold: throw warning on failure of trylock_page in z3fold_alloc

If trylock_page fails, the page won't be non-lru movable page.  When this
page is freed via free_z3fold_page, it will trigger bug on PageMovable
check in __ClearPageMovable.  Throw warning on failure of trylock_page to
guard against such rare case just as what zsmalloc does.

Link: https://lkml.kernel.org/r/20220429064051.61552-5-linmiaohe@huawei.com
Signed-off-by: Miaohe Lin <linmiaohe@huawei.com>
Cc: Vitaly Wool <vitaly.wool@konsulko.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm/z3fold: remove buggy use of stale list for allocation

Currently if z3fold couldn't find an unbuddied page it would first try to
pull a page off the stale list.  But this approach is problematic.  If
init z3fold page fails later, the page should be freed via
free_z3fold_page to clean up the relevant resource instead of using
__free_page directly.  And if page is successfully reused, it will BUG_ON
later in __SetPageMovable because it's already non-lru movable page, i.e.
PAGE_MAPPING_MOVABLE is already set in page->mapping.  In order to fix all
of these issues, we can simply remove the buggy use of stale list for
allocation because can_sleep should always be false and we never really
hit the reusing code path now.

Link: https://lkml.kernel.org/r/20220429064051.61552-4-linmiaohe@huawei.com
Signed-off-by: Miaohe Lin <linmiaohe@huawei.com>
Reviewed-by: Vitaly Wool <vitaly.wool@konsulko.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm/z3fold: fix possible null pointer dereferencing

alloc_slots could fail to allocate memory under heavy memory pressure.  So
we should check zhdr->slots against NULL to avoid future null pointer
dereferencing.

Link: https://lkml.kernel.org/r/20220429064051.61552-3-linmiaohe@huawei.com
Fixes: fc5488651c7d ("z3fold: simplify freeing slots")
Signed-off-by: Miaohe Lin <linmiaohe@huawei.com>
Reviewed-by: Vitaly Wool <vitaly.wool@konsulko.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm/z3fold: fix sheduling while atomic

Patch series "A few fixup patches for z3fold".

This series contains a few fixup patches to fix sheduling while atomic,
fix possible null pointer dereferencing, fix various race conditions and
so on. More details can be found in the respective changelogs.

This patch (of 9):

z3fold's page_lock is always held when calling alloc_slots.  So gfp should
be GFP_ATOMIC to avoid "scheduling while atomic" bug.

Link: https://lkml.kernel.org/r/20220429064051.61552-1-linmiaohe@huawei.com
Link: https://lkml.kernel.org/r/20220429064051.61552-2-linmiaohe@huawei.com
Fixes: fc5488651c7d ("z3fold: simplify freeing slots")
Signed-off-by: Miaohe Lin <linmiaohe@huawei.com>
Reviewed-by: Vitaly Wool <vitaly.wool@konsulko.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm: split free page with properly free memory accounting and without race

In isolate_single_pageblock(), free pages are checked without holding zone
lock, but they can go away in split_free_page() when zone lock is held.
Check the free page and its order again in split_free_page() when zone lock
is held. Recheck the page if the free page is gone under zone lock.

In addition, in split_free_page(), the free page was deleted from the page
list without changing free page accounting. Add the missing free page
accounting code.

Fix the type of order parameter in split_free_page().

Link: https://lore.kernel.org/lkml/20220525103621.987185e2ca0079f7b97b856d@linux-foundation.org/
Link: https://lkml.kernel.org/r/20220526231531.2404977-2-zi.yan@sent.com
Fixes: b2c9e2fbba32 ("mm: make alloc_contig_range work at pageblock granularity")
Signed-off-by: Zi Yan <ziy@nvidia.com>
Reported-by: Doug Berger <opendmb@gmail.com>
Link: https://lore.kernel.org/linux-mm/c3932a6f-77fe-29f7-0c29-fe6b1c67ab7b@gmail.com/
Cc: David Hildenbrand <david@redhat.com>
Cc: Qian Cai <quic_qiancai@quicinc.com>
Cc: Vlastimil Babka <vbabka@suse.cz>
Cc: Mel Gorman <mgorman@techsingularity.net>
Cc: Eric Ren <renzhengeek@gmail.com>
Cc: Mike Rapoport <rppt@kernel.org>
Cc: Oscar Salvador <osalvador@suse.de>
Cc: Christophe Leroy <christophe.leroy@csgroup.eu>
Cc: Marek Szyprowski <m.szyprowski@samsung.com>
Cc: Michael Walle <michael@walle.cc>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm: page-isolation: skip isolated pageblock in start_isolate_page_range()

start_isolate_page_range() first isolates the first and the last
pageblocks in the range and ensure pages across range boundaries are split
during isolation.  But it missed the case when the range is <= a pageblock
and the first and the last pageblocks are the same one, so the second
isolate_single_pageblock() will always fail.  To fix it, skip the
pageblock isolation in second isolate_single_pageblock().

Link: https://lkml.kernel.org/r/20220526231531.2404977-1-zi.yan@sent.com
Fixes: 88ee134320b8 ("mm: fix a potential infinite loop in start_isolate_page_range()")
Signed-off-by: Zi Yan <ziy@nvidia.com>
Reported-by: Marek Szyprowski <m.szyprowski@samsung.com>
Tested-by: Marek Szyprowski <m.szyprowski@samsung.com>
Link: https://lore.kernel.org/linux-mm/ac65adc0-a7e4-cdfe-a0d8-757195b86293@samsung.com/
Reported-by: Michael Walle <michael@walle.cc>
Tested-by: Michael Walle <michael@walle.cc>
Link: https://lore.kernel.org/linux-mm/8ca048ca8b547e0dd1c95387ee05c23d@walle.cc/
Cc: Christophe Leroy <christophe.leroy@csgroup.eu>
Cc: David Hildenbrand <david@redhat.com>
Cc: Doug Berger <opendmb@gmail.com>
Cc: Eric Ren <renzhengeek@gmail.com>
Cc: Mel Gorman <mgorman@techsingularity.net>
Cc: Mike Rapoport <rppt@kernel.org>
Cc: Oscar Salvador <osalvador@suse.de>
Cc: Qian Cai <quic_qiancai@quicinc.com>
Cc: Vlastimil Babka <vbabka@suse.cz>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm: kfence: use PAGE_ALIGNED helper

Use PAGE_ALIGNED macro instead of IS_ALIGNED and passing PAGE_SIZE.

Link: https://lkml.kernel.org/r/20220520021833.121405-1-wangkefeng.wang@huawei.com
Signed-off-by: Kefeng Wang <wangkefeng.wang@huawei.com>
Acked-by: Muchun Song <songmuchun@bytedance.com>
Cc: Marco Elver <elver@google.com>
Cc: Alexander Potapenko <glider@google.com>
Cc: Dmitry Vyukov <dvyukov@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

selftests: vm: add the "settings" file with timeout variable

The default "timeout" for one kselftest is 45 seconds, while some cases in
run_vmtests.sh require more time.  This will cause testing timeout like:

  not ok 4 selftests: vm: run_vmtests.sh # TIMEOUT 45 seconds

Therefore, add the "settings" file with timeout variable so users can set
the "timeout" value.

Link: https://lkml.kernel.org/r/20220521083825.319654-4-patrick.wang.shcn@gmail.com
Signed-off-by: Patrick Wang <patrick.wang.shcn@gmail.com>
Cc: Shuah Khan <shuah@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

selftests: vm: add "test_hmm.sh" to TEST_FILES

The "test_hmm.sh" file used by run_vmtests.sh dose not be installed into
INSTALL_PATH.  Thus run_vmtests.sh can not call it in INSTALL_PATH:

  ---------------------------
  running ./test_hmm.sh smoke
  ---------------------------
  ./run_vmtests.sh: line 74: ./test_hmm.sh: No such file or directory
  [FAIL]
  -----------------------

Add "test_hmm.sh" to TEST_FILES so that it will be installed.

Link: https://lkml.kernel.org/r/20220521083825.319654-3-patrick.wang.shcn@gmail.com
Signed-off-by: Patrick Wang <patrick.wang.shcn@gmail.com>
Cc: Shuah Khan <shuah@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

selftests: vm: check numa_available() before operating "merge_across_nodes" in ksm_tests

Patch series "selftests: vm: a few fixup patches".

This series contains three fixup patches for vm selftests.  They are
independent.  Please see the patches.

This patch (of 3):

Currently, ksm_tests operates "merge_across_nodes" with NUMA either
enabled or disabled.  In a system with NUMA disabled, these operations
will fail and output a misleading report given "merge_across_nodes" does
not exist in sysfs:

  ----------------------------
  running ./ksm_tests -M -p 10
  ----------------------------
  f /sys/kernel/mm/ksm/merge_across_nodes
  fopen: No such file or directory
  Cannot save default tunables
  [FAIL]
  ----------------------

So check numa_available() before those operations to skip them if NUMA is
disabled.

Link: https://lkml.kernel.org/r/20220521083825.319654-1-patrick.wang.shcn@gmail.com
Link: https://lkml.kernel.org/r/20220521083825.319654-2-patrick.wang.shcn@gmail.com
Signed-off-by: Patrick Wang <patrick.wang.shcn@gmail.com>
Cc: Shuah Khan <shuah@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

selftests: vm: add migration to the .gitignore

Add newly added migration test object to .gitignore file.

Link: https://lkml.kernel.org/r/20220521094313.166505-1-usama.anjum@collabora.com
Fixes: 0c2d08728470 ("mm: add selftests for migration entries")
Signed-off-by: Muhammad Usama Anjum <usama.anjum@collabora.com>
Reviewed-by: Alistair Popple <apopple@nvidia.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

selftests/vm/pkeys: fix typo in comment

Spelling mistake (triple letters) in comment.  Detected with the help of
Coccinelle.

Link: https://lkml.kernel.org/r/20220521111145.81697-80-Julia.Lawall@inria.fr
Signed-off-by: Julia Lawall <Julia.Lawall@inria.fr>
Reviewed-by: Muchun Song <songmuchun@bytedance.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

ksm: fix typo in comment

Spelling mistake (triple letters) in comment.  Detected with the help of
Coccinelle.

Link: https://lkml.kernel.org/r/20220521111145.81697-94-Julia.Lawall@inria.fr
Signed-off-by: Julia Lawall <Julia.Lawall@inria.fr>
Reviewed-by: Muchun Song <songmuchun@bytedance.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

selftests: vm: add process_mrelease tests

Introduce process_mrelease syscall sanity tests which include tests
which expect to fail:

- process_mrelease with invalid pidfd and flags inputs
- process_mrelease on a live process with no pending signals

and valid process_mrelease usage which is expected to succeed.  Because
process_mrelease has to be used against a process with a pending SIGKILL,
it's possible that the process exits before process_mrelease gets called.
In such cases we retry the test with a victim that allocates twice more
memory up to 1GB.  This would require the victim process to spend more
time during exit and process_mrelease has a better chance of catching the
process before it exits and succeeding.

On success the test reports the amount of memory the child had to allocate
for reaping to succeed.  Sample output:

$ mrelease_test
Success reaping a child with 1MB of memory allocations

On failure the test reports the failure. Sample outputs:

$ mrelease_test
All process_mrelease attempts failed!

$ mrelease_test
process_mrelease: Invalid argument

Link: https://lkml.kernel.org/r/20220518204316.13131-1-surenb@google.com
Signed-off-by: Suren Baghdasaryan <surenb@google.com>
Reviewed-by: Shuah Khan <skhan@linuxfoundation.org>
Acked-by: Christian Brauner (Microsoft) <brauner@kernel.org>
Reviewed-by: Muhammad Usama Anjum <usama.anjum@collabora.com>
Cc: Michal Hocko <mhocko@suse.com>
Cc: David Rientjes <rientjes@google.com>
Cc: Matthew Wilcox (Oracle) <willy@infradead.org>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: Roman Gushchin <guro@fb.com>
Cc: Minchan Kim <minchan@kernel.org>
Cc: "Kirill A . Shutemov" <kirill@shutemov.name>
Cc: Andrea Arcangeli <aarcange@redhat.com>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Oleg Nesterov <oleg@redhat.com>
Cc: David Hildenbrand <david@redhat.com>
Cc: Jann Horn <jannh@google.com>
Cc: Shakeel Butt <shakeelb@google.com>
Cc: Peter Xu <peterx@redhat.com>
Cc: John Hubbard <jhubbard@nvidia.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

Revert "mm/vmscan: never demote for memcg reclaim"

This reverts commit 3a235693d3930e1276c8d9cc0ca5807ef292cf0a.

Its premise was that cgroup reclaim cares about freeing memory inside the
cgroup, and demotion just moves them around within the cgroup limit.
Hence, pages from toptier nodes should be reclaimed directly.

However, with NUMA balancing now doing tier promotions, demotion is part
of the page aging process.  Global reclaim demotes the coldest toptier
pages to secondary memory, where their life continues and from which they
have a chance to get promoted back.  Essentially, tiered memory systems
have an LRU order that spans multiple nodes.

When cgroup reclaims pages coming off the toptier directly, there can be
colder pages on lower tier nodes that were demoted by global reclaim.
This is an aging inversion, not unlike if cgroups were to reclaim directly
from the active lists while there are inactive pages.

Proactive reclaim is another factor.  The goal of that it is to offload
colder pages from expensive RAM to cheaper storage.  When lower tier
memory is available as an intermediate layer, we want offloading to take
advantage of it instead of bypassing to storage.

Revert the patch so that cgroups respect the LRU order spanning the memory
hierarchy.

Of note is a specific undercommit scenario, where all cgroup limits in the
system add up to <= available toptier memory.  In that case, shuffling
pages out to lower tiers first to reclaim them from there is inefficient.
This is something could be optimized/short-circuited later on (although
care must be taken not to accidentally recreate the aging inversion).
Let's ensure correctness first.

Link: https://lkml.kernel.org/r/20220518190911.82400-1-hannes@cmpxchg.org
Signed-off-by: Johannes Weiner <hannes@cmpxchg.org>
Acked-by: Dave Hansen <dave.hansen@linux.intel.com>
Reviewed-by: Yang Shi <shy828301@gmail.com>
Acked-by: Roman Gushchin <roman.gushchin@linux.dev>
Reviewed-by: "Huang, Ying" <ying.huang@intel.com>
Reviewed-by: Muchun Song <songmuchun@bytedance.com>
Acked-by: Michal Hocko <mhocko@suse.com>
Acked-by: Shakeel Butt <shakeelb@google.com>
Acked-by: Tim Chen <tim.c.chen@linux.intel.com>
Cc: Zi Yan <ziy@nvidia.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm/kfence: print disabling or re-enabling message

By printing information, we can friendly prompt the status change
information of kfence by dmesg and record by syslog.

Also, set kfence_enabled to false only when needed.

Link: https://lkml.kernel.org/r/20220518073105.3160335-1-liu.yun@linux.dev
Signed-off-by: Jackie Liu <liuyun01@kylinos.cn>
Co-developed-by: Marco Elver <elver@google.com>
Signed-off-by: Marco Elver <elver@google.com>
Reviewed-by: Marco Elver <elver@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

include/trace/events/percpu.h: cleanup for "percpu: improve percpu_alloc_percpu event trace"

Fix sparse warning about incorrect gfp_t cast.

Link: https://lkml.kernel.org/r/001979f3-e978-0998-cbed-61a4a2ac87b8@openvz.org
Fixes: f67bed134a05 ("percpu: improve percpu_alloc_percpu event trace")
Signed-off-by: Vasily Averin <vvs@openvz.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

include/trace/events/mmflags.h: cleanup for "tracing: incorrect gfp_t conversion"

Redefines __def_gfpflag_names array according to akpm@, willy@ and Joe
Perches recommendations.

Link: https://lkml.kernel.org/r/6f811e19-41c6-f3e8-fca6-23a19a62e313@openvz.org
Fixes: fe573327ffb1 ("tracing: incorrect gfp_t conversion")
Signed-off-by: Vasily Averin <vvs@openvz.org>
Cc: Matthew Wilcox <willy@infradead.org>
Cc: Joe Perches <joe@perches.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm: fix a potential infinite loop in start_isolate_page_range()

In isolate_single_pageblock() called by start_isolate_page_range(), there
are some pageblock isolation issues causing a potential infinite loop when
isolating a page range.  This is reported by Qian Cai.

1. the pageblock was isolated by just changing pageblock migratetype
   without checking unmovable pages. Calling set_migratetype_isolate() to
   isolate pageblock properly.
2. an off-by-one error caused migrating pages unnecessarily, since the page
   is not crossing pageblock boundary.
3. migrating a compound page across pageblock boundary then splitting the
   free page later has a small race window that the free page might be
   allocated again, so that the code will try again, causing an potential
   infinite loop. Temporarily set the to-be-migrated page's pageblock to
   MIGRATE_ISOLATE to prevent that and bail out early if no free page is
   found after page migration.

An additional fix to split_free_page() aims to avoid crashing in
__free_one_page().  When the free page is split at the specified
split_pfn_offset, free_page_order should check both the first bit of
free_page_pfn and the last bit of split_pfn_offset and use the smaller
one.  For example, if free_page_pfn=0x10000, split_pfn_offset=0xc000,
free_page_order should first be 0x8000 then 0x4000, instead of 0x4000 then
0x8000, which the original algorithm did.

[akpm@linux-foundation.org: suppress min() warning]
Link: https://lkml.kernel.org/r/20220524194756.1698351-1-zi.yan@sent.com
Fixes: b2c9e2fbba3253 ("mm: make alloc_contig_range work at pageblock granularity")
Signed-off-by: Zi Yan <ziy@nvidia.com>
Reported-by: Qian Cai <quic_qiancai@quicinc.com>
Cc: Christophe Leroy <christophe.leroy@csgroup.eu>
Cc: David Hildenbrand <david@redhat.com>
Cc: Eric Ren <renzhengeek@gmail.com>
Cc: Mel Gorman <mgorman@techsingularity.net>
Cc: Mike Rapoport <rppt@linux.ibm.com>
Cc: Minchan Kim <minchan@kernel.org>
Cc: Oscar Salvador <osalvador@suse.de>
Cc: Vlastimil Babka <vbabka@suse.cz>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

MAINTAINERS: add Muchun as co-maintainer for HugeTLB

I have been focusing on mm for the past two years.  e.g.  developing,
fixing bugs, reviewing related to HugeTLB system.  I would like to help
Mike and other people working on HugeTLB by reviewing their work.

When I first introduced the vmemmmap reduction, I forgot to update
MAINTAINERS file.  Let's update it as well.  And rename "HUGETLB
FILESYSTEM" to "HUGETLB SUBSYSTEM" since some files are not only related
to filesystem but also memory management (the name of FILESYSTEM cannot
cover this area).

Link: https://lkml.kernel.org/r/20220521074103.79468-1-songmuchun@bytedance.com
Signed-off-by: Muchun Song <songmuchun@bytedance.com>
Acked-by: Mike Kravetz <mike.kravetz@oracle.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

zram: fix Kconfig dependency warning

ZSMALLOC depends on MMU so ZRAM should also depend on MMU since 'select'
does not follow any dependency chains.

Fixes this Kconfig warning:

WARNING: unmet direct dependencies detected for ZSMALLOC
  Depends on [n]: MMU [=n]
  Selected by [y]:
  - ZRAM [=y] && BLK_DEV [=y] && BLOCK [=y] && SYSFS [=y] && (CRYPTO_LZO [=y] || CRYPTO_ZSTD [=m] || CRYPTO_LZ4 [=m] || CRYPTO_LZ4HC [=n] || CRYPTO_842 [=n])

Link: https://lkml.kernel.org/r/20220522204027.22964-1-rdunlap@infradead.org
Fixes: b3fbd58fcbb10 ("mm: Kconfig: simplify zswap configuration")
Signed-off-by: Randy Dunlap <rdunlap@infradead.org>
Acked-by: Johannes Weiner <hannes@cmpxchg.org>
Cc: Minchan Kim <minchan@kernel.org>
Cc: Nitin Gupta <ngupta@vflare.org>
Cc: Sergey Senozhatsky <senozhatsky@chromium.org>
Cc: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm/shmem: fix shmem folio swapoff hang

Shmem swapoff makes no progress: the index to indices is not incremented.
But "ret" is no longer a return value, so use folio_batch_count() instead.

Link: https://lkml.kernel.org/r/c32bee8a-f0aa-245-f94e-24dd271924fa@google.com
Fixes: da08e9b79323 ("mm/shmem: convert shmem_swapin_page() to shmem_swapin_folio()")
Signed-off-by: Hugh Dickins <hughd@google.com>
Reviewed-by: Miaohe Lin <linmiaohe@huawei.com>
Tested-by: Miaohe Lin <linmiaohe@huawei.com>
Cc: Matthew Wilcox <willy@infradead.org>
Cc: Christoph Hellwig <hch@lst.de>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

cgroup: fix an error handling path in alloc_pagecache_max_30M()

If the first goto is taken, 'fd' is not opened yet (and is un-initialized).
So a direct return is safer.

Link: https://lkml.kernel.org/r/628312312eb40e0e39463a2c06415fde5295c716.1653229120.git.christophe.jaillet@wanadoo.fr
Fixes: c1a31a2f7a9c ("cgroup: fix racy check in alloc_pagecache_max_30M() helper function")
Signed-off-by: Christophe JAILLET <christophe.jaillet@wanadoo.fr>
Reviewed-by: Andrew Morton <akpm@linux-foundation.org>
Cc: Dan Carpenter <dan.carpenter@oracle.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: Michal Hocko <mhocko@kernel.org>
Cc: Roman Gushchin <roman.gushchin@linux.dev>
Cc: Shakeel Butt <shakeelb@google.com>
Cc: Muchun Song <songmuchun@bytedance.com>
Cc: Tejun Heo <tj@kernel.org>
Cc: Zefan Li <lizefan.x@bytedance.com>
Cc: Shuah Khan <shuah@kernel.org>
Cc: David Vernet <void@manifault.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm: damon: use HPAGE_PMD_SIZE

Use HPAGE_PMD_SIZE instead of open coding.

Link: https://lkml.kernel.org/r/20220517145120.118523-1-wangkefeng.wang@huawei.com
Signed-off-by: Kefeng Wang <wangkefeng.wang@huawei.com>
Reviewed-by: SeongJae Park <sj@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

tracing: incorrect isolate_mote_t cast in mm_vmscan_lru_isolate

Fixes following sparse warnings:

  CHECK   mm/vmscan.c
mm/vmscan.c: note: in included file (through
include/trace/trace_events.h, include/trace/define_trace.h,
include/trace/events/vmscan.h):
./include/trace/events/vmscan.h:281:1: sparse: warning:
 cast to restricted isolate_mode_t
./include/trace/events/vmscan.h:281:1: sparse: warning:
 restricted isolate_mode_t degrades to integer

Link: https://lkml.kernel.org/r/e85d7ff2-fd10-53f8-c24e-ba0458439c1b@openvz.org
Signed-off-by: Vasily Averin <vvs@openvz.org>
Acked-by: Steven Rostedt (Google) <rostedt@goodmis.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

nodemask.h: fix compilation error with GCC12

With gcc version 12.0.1 20220401 (Red Hat 12.0.1-0), building with
defconfig results in the following compilation error:

|   CC      mm/swapfile.o
| mm/swapfile.c: In function `setup_swap_info':
| mm/swapfile.c:2291:47: error: array subscript -1 is below array bounds
|  of `struct plist_node[]' [-Werror=array-bounds]
|  2291 |                                 p->avail_lists[i].prio = 1;
|       |                                 ~~~~~~~~~~~~~~^~~
| In file included from mm/swapfile.c:16:
| ./include/linux/swap.h:292:27: note: while referencing `avail_lists'
|   292 |         struct plist_node avail_lists[]; /*
|       |                           ^~~~~~~~~~~

This is due to the compiler detecting that the mask in
node_states[__state] could theoretically be zero, which would lead to
first_node() returning -1 through find_first_bit.

I believe that the warning/error is legitimate.  I first tried adding a
test to check that the node mask is not emtpy, since a similar test exists
in the case where MAX_NUMNODES == 1.

However, adding the if statement causes other warnings to appear in
for_each_cpu_node_but, because it introduces a dangling else ambiguity.
And unfortunately, GCC is not smart enough to detect that the added test
makes the case where (node) == -1 impossible, so it still complains with
the same message.

This is why I settled on replacing that with a harmless, but relatively
useless (node) >= 0 test.  Based on the warning for the dangling else, I
also decided to fix the case where MAX_NUMNODES == 1 by moving the
condition inside the for loop.  It will still only be tested once.  This
ensures that the meaning of an else following for_each_node_mask or
derivatives would not silently have a different meaning depending on the
configuration.

Link: https://lkml.kernel.org/r/20220414150855.2407137-3-dinechin@redhat.com
Signed-off-by: Christophe de Dinechin <christophe@dinechin.org>
Signed-off-by: Christophe de Dinechin <dinechin@redhat.com>
Reviewed-by: Andrew Morton <akpm@linux-foundation.org>
Cc: Ben Segall <bsegall@google.com>
Cc: "Michael S. Tsirkin" <mst@redhat.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Dietmar Eggemann <dietmar.eggemann@arm.com>
Cc: Vincent Guittot <vincent.guittot@linaro.org>
Cc: Paolo Bonzini <pbonzini@redhat.com>
Cc: Daniel Bristot de Oliveira <bristot@redhat.com>
Cc: Jason Wang <jasowang@redhat.com>
Cc: Zhen Lei <thunder.leizhen@huawei.com>
Cc: Juri Lelli <juri.lelli@redhat.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm: fix missing handler for __GFP_NOWARN

We expect no warnings to be issued when we specify __GFP_NOWARN, but
currently in paths like alloc_pages() and kmalloc(), there are still some
warnings printed, fix it.

But for some warnings that report usage problems, we don't deal with them.
If such warnings are printed, then we should fix the usage problems.
Such as the following case:

WARN_ON_ONCE((gfp_flags & __GFP_NOFAIL) && (order > 1));

[zhengqi.arch@bytedance.com: v2]
Link: https://lkml.kernel.org/r/20220511061951.1114-1-zhengqi.arch@bytedance.com
Link: https://lkml.kernel.org/r/20220510113809.80626-1-zhengqi.arch@bytedance.com
Signed-off-by: Qi Zheng <zhengqi.arch@bytedance.com>
Cc: Akinobu Mita <akinobu.mita@gmail.com>
Cc: Vlastimil Babka <vbabka@suse.cz>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Cc: Jiri Slaby <jirislaby@kernel.org>
Cc: Steven Rostedt (Google) <rostedt@goodmis.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm/page_alloc: fix tracepoint mm_page_alloc_zone_locked()

Currently, trace point mm_page_alloc_zone_locked() doesn't show correct
information.

First, when alloc_flag has ALLOC_HARDER/ALLOC_CMA, page can be allocated
from MIGRATE_HIGHATOMIC/MIGRATE_CMA.  Nevertheless, tracepoint use
requested migration type not MIGRATE_HIGHATOMIC and MIGRATE_CMA.

Second, after commit 44042b4498728 ("mm/page_alloc: allow high-order pages
to be stored on the per-cpu lists") percpu-list can store high order
pages.  But trace point determine whether it is a refiil of percpu-list by
comparing requested order and 0.

To handle these problems, make mm_page_alloc_zone_locked() only be called
by __rmqueue_smallest with correct migration type.  With a new argument
called percpu_refill, it can show roughly whether it is a refill of
percpu-list.

Link: https://lkml.kernel.org/r/20220512025307.57924-1-vvghjk1234@gmail.com
Signed-off-by: Wonhyuk Yang <vvghjk1234@gmail.com>
Acked-by: Mel Gorman <mgorman@suse.de>
Cc: Baik Song An <bsahn@etri.re.kr>
Cc: Hong Yeon Kim <kimhy@etri.re.kr>
Cc: Taeung Song <taeung@reallinux.co.kr>
Cc: <linuxgeek@linuxgeek.io>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Ingo Molnar <mingo@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm/page_owner.c: add missing __initdata attribute

This patch fixes two issues:
1. Add __initdata attribute according to include/linux/init.h:
For initialized data:
You should insert __initdata between the variable name and equal
sign followed by value

2. Fix below error reported by checkpatch.pl:
ERROR: do not initialise statics to false

Special thanks to Muchun Song :)

Link: https://lkml.kernel.org/r/20220516030039.1487005-1-bh1scw@gmail.com
Signed-off-by: Fanjun Kong <bh1scw@gmail.com>
Suggested-by: Muchun Song <songmuchun@bytedance.com>
Reviewed-by: Muchun Song <songmuchun@bytedance.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

tmpfs: fix undefined-behaviour in shmem_reconfigure()

When shmem_reconfigure() calls __percpu_counter_compare(), the second
parameter is unsigned long long.  But in the definition of
__percpu_counter_compare(), the second parameter is s64.  So when
__percpu_counter_compare() executes abs(count - rhs), UBSAN shows the
following warning:

================================================================================
UBSAN: Undefined behaviour in lib/percpu_counter.c:209:6
signed integer overflow:
0 - -9223372036854775808 cannot be represented in type 'long long int'
CPU: 1 PID: 9636 Comm: syz-executor.2 Tainted: G                 ---------r-  - 4.18.0 #2
Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS rel-1.14.0-0-g155821a1990b-prebuilt.qemu.org 04/01/2014
Call Trace:
 __dump_stack home/install/linux-rh-3-10/lib/dump_stack.c:77 [inline]
 dump_stack+0x125/0x1ae home/install/linux-rh-3-10/lib/dump_stack.c:117
 ubsan_epilogue+0xe/0x81 home/install/linux-rh-3-10/lib/ubsan.c:159
 handle_overflow+0x19d/0x1ec home/install/linux-rh-3-10/lib/ubsan.c:190
 __percpu_counter_compare+0x124/0x140 home/install/linux-rh-3-10/lib/percpu_counter.c:209
 percpu_counter_compare home/install/linux-rh-3-10/./include/linux/percpu_counter.h:50 [inline]
 shmem_remount_fs+0x1ce/0x6b0 home/install/linux-rh-3-10/mm/shmem.c:3530
 do_remount_sb+0x11b/0x530 home/install/linux-rh-3-10/fs/super.c:888
 do_remount home/install/linux-rh-3-10/fs/namespace.c:2344 [inline]
 do_mount+0xf8d/0x26b0 home/install/linux-rh-3-10/fs/namespace.c:2844
 ksys_mount+0xad/0x120 home/install/linux-rh-3-10/fs/namespace.c:3075
 __do_sys_mount home/install/linux-rh-3-10/fs/namespace.c:3089 [inline]
 __se_sys_mount home/install/linux-rh-3-10/fs/namespace.c:3086 [inline]
 __x64_sys_mount+0xbf/0x160 home/install/linux-rh-3-10/fs/namespace.c:3086
 do_syscall_64+0xca/0x5c0 home/install/linux-rh-3-10/arch/x86/entry/common.c:298
 entry_SYSCALL_64_after_hwframe+0x6a/0xdf
RIP: 0033:0x46b5e9
Code: 5d db fa ff c3 66 2e 0f 1f 84 00 00 00 00 00 66 90 48 89 f8 48 89 f7 48 89 d6 48 89 ca 4d 89 c2 4d 89 c8 4c 8b 4c 24 08 0f 05 <48> 3d 01 f0 ff ff 0f 83 2b db fa ff c3 66 2e 0f 1f 84 00 00 00 00
RSP: 002b:00007f54d5f22c68 EFLAGS: 00000246 ORIG_RAX: 00000000000000a5
RAX: ffffffffffffffda RBX: 000000000077bf60 RCX: 000000000046b5e9
RDX: 0000000000000000 RSI: 0000000020000000 RDI: 0000000000000000
RBP: 000000000077bf60 R08: 0000000020000140 R09: 0000000000000000
R10: 00000000026740a4 R11: 0000000000000246 R12: 0000000000000000
R13: 00007ffd1fb1592f R14: 00007f54d5f239c0 R15: 000000000077bf6c
================================================================================

[akpm@linux-foundation.org: tweak error message text]
Link: https://lkml.kernel.org/r/20220513025225.2678727-1-luomeng12@huawei.com
Signed-off-by: Luo Meng <luomeng12@huawei.com>
Cc: Hugh Dickins <hughd@google.com>
Cc: Yu Kuai <yukuai3@huawei.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm/mempolicy: fix uninit-value in mpol_rebind_policy()

mpol_set_nodemask()(mm/mempolicy.c) does not set up nodemask when
pol->mode is MPOL_LOCAL.  Check pol->mode before access
pol->w.cpuset_mems_allowed in mpol_rebind_policy()(mm/mempolicy.c).

BUG: KMSAN: uninit-value in mpol_rebind_policy mm/mempolicy.c:352 [inline]
BUG: KMSAN: uninit-value in mpol_rebind_task+0x2ac/0x2c0 mm/mempolicy.c:368
 mpol_rebind_policy mm/mempolicy.c:352 [inline]
 mpol_rebind_task+0x2ac/0x2c0 mm/mempolicy.c:368
 cpuset_change_task_nodemask kernel/cgroup/cpuset.c:1711 [inline]
 cpuset_attach+0x787/0x15e0 kernel/cgroup/cpuset.c:2278
 cgroup_migrate_execute+0x1023/0x1d20 kernel/cgroup/cgroup.c:2515
 cgroup_migrate kernel/cgroup/cgroup.c:2771 [inline]
 cgroup_attach_task+0x540/0x8b0 kernel/cgroup/cgroup.c:2804
 __cgroup1_procs_write+0x5cc/0x7a0 kernel/cgroup/cgroup-v1.c:520
 cgroup1_tasks_write+0x94/0xb0 kernel/cgroup/cgroup-v1.c:539
 cgroup_file_write+0x4c2/0x9e0 kernel/cgroup/cgroup.c:3852
 kernfs_fop_write_iter+0x66a/0x9f0 fs/kernfs/file.c:296
 call_write_iter include/linux/fs.h:2162 [inline]
 new_sync_write fs/read_write.c:503 [inline]
 vfs_write+0x1318/0x2030 fs/read_write.c:590
 ksys_write+0x28b/0x510 fs/read_write.c:643
 __do_sys_write fs/read_write.c:655 [inline]
 __se_sys_write fs/read_write.c:652 [inline]
 __x64_sys_write+0xdb/0x120 fs/read_write.c:652
 do_syscall_x64 arch/x86/entry/common.c:51 [inline]
 do_syscall_64+0x54/0xd0 arch/x86/entry/common.c:82
 entry_SYSCALL_64_after_hwframe+0x44/0xae

Uninit was created at:
 slab_post_alloc_hook mm/slab.h:524 [inline]
 slab_alloc_node mm/slub.c:3251 [inline]
 slab_alloc mm/slub.c:3259 [inline]
 kmem_cache_alloc+0x902/0x11c0 mm/slub.c:3264
 mpol_new mm/mempolicy.c:293 [inline]
 do_set_mempolicy+0x421/0xb70 mm/mempolicy.c:853
 kernel_set_mempolicy mm/mempolicy.c:1504 [inline]
 __do_sys_set_mempolicy mm/mempolicy.c:1510 [inline]
 __se_sys_set_mempolicy+0x44c/0xb60 mm/mempolicy.c:1507
 __x64_sys_set_mempolicy+0xd8/0x110 mm/mempolicy.c:1507
 do_syscall_x64 arch/x86/entry/common.c:51 [inline]
 do_syscall_64+0x54/0xd0 arch/x86/entry/common.c:82
 entry_SYSCALL_64_after_hwframe+0x44/0xae

KMSAN: uninit-value in mpol_rebind_task (2)
https://syzkaller.appspot.com/bug?id=d6eb90f952c2a5de9ea718a1b873c55cb13b59dc

This patch seems to fix below bug too.
KMSAN: uninit-value in mpol_rebind_mm (2)
https://syzkaller.appspot.com/bug?id=f2fecd0d7013f54ec4162f60743a2b28df40926b

The uninit-value is pol->w.cpuset_mems_allowed in mpol_rebind_policy().
When syzkaller reproducer runs to the beginning of mpol_new(),

    mpol_new() mm/mempolicy.c
  do_mbind() mm/mempolicy.c
kernel_mbind() mm/mempolicy.c

`mode` is 1(MPOL_PREFERRED), nodes_empty(*nodes) is `true` and `flags`
is 0. Then

mode = MPOL_LOCAL;
...
policy->mode = mode;
policy->flags = flags;

will be executed. So in mpol_set_nodemask(),

    mpol_set_nodemask() mm/mempolicy.c
  do_mbind()
kernel_mbind()

pol->mode is 4 (MPOL_LOCAL), that `nodemask` in `pol` is not initialized,
which will be accessed in mpol_rebind_policy().

Link: https://lkml.kernel.org/r/20220512123428.fq3wofedp6oiotd4@ppc.localdomain
Signed-off-by: Wang Cheng <wanngchenng@gmail.com>
Reported-by: <syzbot+217f792c92599518a2ab@syzkaller.appspotmail.com>
Tested-by: <syzbot+217f792c92599518a2ab@syzkaller.appspotmail.com>
Cc: David Rientjes <rientjes@google.com>
Cc: Vlastimil Babka <vbabka@suse.cz>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm: don't be stuck to rmap lock on reclaim path

The rmap locks(i_mmap_rwsem and anon_vma->root->rwsem) could be contended
under memory pressure if processes keep working on their vmas(e.g., fork,
mmap, munmap).  It makes reclaim path stuck.  In our real workload traces,
we see kswapd is waiting the lock for 300ms+(worst case, a sec) and it
makes other processes entering direct reclaim, which were also stuck on
the lock.

This patch makes lru aging path try_lock mode like shink_page_list so the
reclaim context will keep working with next lru pages without being stuck.
if it found the rmap lock contended, it rotates the page back to head of
lru in both active/inactive lrus to make them consistent behavior, which
is basic starting point rather than adding more heristic.

Since this patch introduces a new "contended" field as out-param along
with try_lock in-param in rmap_walk_control, it's not immutable any longer
if the try_lock is set so remove const keywords on rmap related functions.
Since rmap walking is already expensive operation, I doubt the const
would help sizable benefit( And we didn't have it until 5.17).

In a heavy app workload in Android, trace shows following statistics.  It
almost removes rmap lock contention from reclaim path.

Martin Liu reported:

Before:

   max_dur(ms)  min_dur(ms)  max-min(dur)ms  avg_dur(ms)  sum_dur(ms)  count blocked_function
         1632            0            1631   151.542173        31672    209  page_lock_anon_vma_read
          601            0             601   145.544681        28817    198  rmap_walk_file

After:

   max_dur(ms)  min_dur(ms)  max-min(dur)ms  avg_dur(ms)  sum_dur(ms)  count blocked_function
          NaN          NaN              NaN          NaN          NaN    0.0             NaN
            0            0                0     0.127645            1     12  rmap_walk_file

[minchan@kernel.org: add comment, per Matthew]
Link: https://lkml.kernel.org/r/YnNqeB5tUf6LZ57b@google.com
Link: https://lkml.kernel.org/r/20220510215423.164547-1-minchan@kernel.org
Signed-off-by: Minchan Kim <minchan@kernel.org>
Acked-by: Johannes Weiner <hannes@cmpxchg.org>
Cc: Suren Baghdasaryan <surenb@google.com>
Cc: Michal Hocko <mhocko@suse.com>
Cc: John Dias <joaodias@google.com>
Cc: Tim Murray <timmurray@google.com>
Cc: Matthew Wilcox <willy@infradead.org>
Cc: Vladimir Davydov <vdavydov.dev@gmail.com>
Cc: Martin Liu <liumartin@google.com>
Cc: Minchan Kim <minchan@kernel.org>
Cc: Matthew Wilcox <willy@infradead.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

zswap: memcg accounting

Applications can currently escape their cgroup memory containment when
zswap is enabled.  This patch adds per-cgroup tracking and limiting of
zswap backend memory to rectify this.

The existing cgroup2 memory.stat file is extended to show zswap statistics
analogous to what's in meminfo and vmstat.  Furthermore, two new control
files, memory.zswap.current and memory.zswap.max, are added to allow
tuning zswap usage on a per-workload basis.  This is important since not
all workloads benefit from zswap equally; some even suffer compared to
disk swap when memory contents don't compress well.  The optimal size of
the zswap pool, and the threshold for writeback, also depends on the size
of the workload's warm set.

The implementation doesn't use a traditional page_counter transaction.
zswap is unconventional as a memory consumer in that we only know the
amount of memory to charge once expensive compression has occurred.  If
zwap is disabled or the limit is already exceeded we obviously don't want
to compress page upon page only to reject them all.  Instead, the limit is
checked against current usage, then we compress and charge.  This allows
some limit overrun, but not enough to matter in practice.

[hannes@cmpxchg.org: fix for CONFIG_SLOB builds]
Link: https://lkml.kernel.org/r/YnwD14zxYjUJPc2w@cmpxchg.org
[hannes@cmpxchg.org: opt out of cgroups v1]
Link: https://lkml.kernel.org/r/Yn6it9mBYFA+/lTb@cmpxchg.org
Link: https://lkml.kernel.org/r/20220510152847.230957-7-hannes@cmpxchg.org
Signed-off-by: Johannes Weiner <hannes@cmpxchg.org>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Roman Gushchin <guro@fb.com>
Cc: Shakeel Butt <shakeelb@google.com>
Cc: Seth Jennings <sjenning@redhat.com>
Cc: Dan Streetman <ddstreet@ieee.org>
Cc: Minchan Kim <minchan@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm: zswap: add basic meminfo and vmstat coverage

Currently it requires poking at debugfs to figure out the size and
population of the zswap cache on a host.  There are no counters for reads
and writes against the cache.  As a result, it's difficult to understand
zswap behavior on production systems.

Print zswap memory consumption and how many pages are zswapped out in
/proc/meminfo.  Count zswapouts and zswapins in /proc/vmstat.

Link: https://lkml.kernel.org/r/20220510152847.230957-6-hannes@cmpxchg.org
Signed-off-by: Johannes Weiner <hannes@cmpxchg.org>
Acked-by: David Hildenbrand <david@redhat.com>
Cc: Dan Streetman <ddstreet@ieee.org>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Minchan Kim <minchan@kernel.org>
Cc: Roman Gushchin <guro@fb.com>
Cc: Seth Jennings <sjenning@redhat.com>
Cc: Shakeel Butt <shakeelb@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm: Kconfig: simplify zswap configuration

- CONFIG_ZRAM: Zram is a user-facing feature, whereas zsmalloc is
  not. Don't make the user chase down a technical dependency like
  that, just select it in automatically when zram is requested. The
  CONFIG_CRYPTO dependency is redundant due to more specific deps.

- CONFIG_ZPOOL: This is not a user-facing feature. Hide the symbol and
  have it selected in as needed.

- CONFIG_ZSWAP: Select CRYPTO instead of depend. Common pattern.

- Make the ZSWAP suboptions and their descriptions (compression,
  allocation backend) a bit more straight-forward for the user.

Link: https://lkml.kernel.org/r/20220510152847.230957-5-hannes@cmpxchg.org
Signed-off-by: Johannes Weiner <hannes@cmpxchg.org>
Cc: Dan Streetman <ddstreet@ieee.org>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Minchan Kim <minchan@kernel.org>
Cc: Roman Gushchin <guro@fb.com>
Cc: Seth Jennings <sjenning@redhat.com>
Cc: Shakeel Butt <shakeelb@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm: Kconfig: group swap, slab, hotplug and thp options into submenus

There are several clusters of related config options spread throughout the
mostly flat MM submenu.  Group them together and put specialization
options into further subdirectories to make the MM submenu a bit more
organized and easier to navigate.

[hannes@cmpxchg.org: fix kbuild warnings]
Link: https://lkml.kernel.org/r/YnvkSVivfnT57Vwh@cmpxchg.org
[hannes@cmpxchg.org: fix more kbuild warnings]
Link: https://lkml.kernel.org/r/Ynz8NusTdEGcCnJN@cmpxchg.org
Link: https://lkml.kernel.org/r/20220510152847.230957-4-hannes@cmpxchg.org
Signed-off-by: Johannes Weiner <hannes@cmpxchg.org>
Cc: Dan Streetman <ddstreet@ieee.org>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Minchan Kim <minchan@kernel.org>
Cc: Roman Gushchin <guro@fb.com>
Cc: Seth Jennings <sjenning@redhat.com>
Cc: Shakeel Butt <shakeelb@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm: Kconfig: move swap and slab config options to the MM section

These are currently under General Setup. MM seems like a better fit.

Link: https://lkml.kernel.org/r/20220510152847.230957-3-hannes@cmpxchg.org
Signed-off-by: Johannes Weiner <hannes@cmpxchg.org>
Cc: Dan Streetman <ddstreet@ieee.org>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Minchan Kim <minchan@kernel.org>
Cc: Roman Gushchin <guro@fb.com>
Cc: Seth Jennings <sjenning@redhat.com>
Cc: Shakeel Butt <shakeelb@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

Documentation: filesystems: proc: update meminfo section

Patch series "zswap: accounting & cgroup control", v2.

Zswap can consume nearly a quarter of RAM in the default configuration,
yet it's neither listed in /proc/meminfo, nor is it accounted and
manageable on a per-cgroup basis.

This makes reasoning about the memory situation on a host in general
rather difficult.  On shared/cgrouped hosts, the consequences are worse.
First, workloads can escape memory containment and cause resource priority
inversions: a lo-pri group can fill the global zswap pool and force a
hi-pri group out to disk.  Second, not all workloads benefit from zswap
equally.  Some even suffer when memory contents compress poorly, and are
better off going to disk swap directly.  On a host with mixed workloads,
it's currently not possible to enable zswap for one workload but not for
the other.

This series implements the missing global accounting as well as cgroup
tracking & control for zswap backing memory:

- Patch 1 refreshes the very out-of-date meminfo documentation in
  Documentation/filesystems/proc.rst.

- Patches 2-4 clean up related and adjacent options in Kconfig. Not
  actual dependencies, just things I noticed during development.

- Patch 5 adds meminfo and vmstat coverage for zswap consumption and
  activity.

- Patch 6 implements per-cgroup tracking & control of zswap memory.

This patch (of 6):

Add new entries.  Minor corrections and cleanups.

[hannes@cmpxchg.org: fix htmldocs warnings]
Link: https://lkml.kernel.org/r/Ynve8dg4zJyhH2gW@cmpxchg.org
[hannes@cmpxchg.org: change `Unevictable' wording, per David]
Link: https://lkml.kernel.org/r/YnwFraZlVWQoCjz3@cmpxchg.org
Link: https://lkml.kernel.org/r/20220510152847.230957-1-hannes@cmpxchg.org
Link: https://lkml.kernel.org/r/20220510152847.230957-2-hannes@cmpxchg.org
Signed-off-by: Johannes Weiner <hannes@cmpxchg.org>
Acked-by: David Hildenbrand <david@redhat.com>
Cc: Dan Streetman <ddstreet@ieee.org>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Minchan Kim <minchan@kernel.org>
Cc: Roman Gushchin <guro@fb.com>
Cc: Seth Jennings <sjenning@redhat.com>
Cc: Shakeel Butt <shakeelb@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm/swap: fix comment about swap extent

Since commit 4efaceb1c5f8 ("mm, swap: use rbtree for swap_extent"), rbtree
is used for swap extent.  Also curr_swap_extent is removed at that time.
Update the corresponding comment.

Link: https://lkml.kernel.org/r/20220509131416.17553-16-linmiaohe@huawei.com
Signed-off-by: Miaohe Lin <linmiaohe@huawei.com>
Cc: Alistair Popple <apopple@nvidia.com>
Cc: David Hildenbrand <david@redhat.com>
Cc: David Howells <dhowells@redhat.com>
Cc: Matthew Wilcox <willy@infradead.org>
Cc: Naoya Horiguchi <naoya.horiguchi@nec.com>
Cc: NeilBrown <neilb@suse.de>
Cc: Peter Xu <peterx@redhat.com>
Cc: Suren Baghdasaryan <surenb@google.com>
Cc: Vlastimil Babka <vbabka@suse.cz>
Cc: Oscar Salvador <osalvador@suse.de>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm/swap: fix the comment of get_kernel_pages

If no pages were pinned, 0 is returned in fact.  Fix the corresponding
comment.

[akpm@linux-foundation.org: s/nr_pages/nr_segs/ also, per David, reflow comment]
Link: https://lkml.kernel.org/r/20220509131416.17553-15-linmiaohe@huawei.com
Signed-off-by: Miaohe Lin <linmiaohe@huawei.com>
Reviewed-by: David Hildenbrand <david@redhat.com>
Cc: Alistair Popple <apopple@nvidia.com>
Cc: David Howells <dhowells@redhat.com>
Cc: Matthew Wilcox <willy@infradead.org>
Cc: Naoya Horiguchi <naoya.horiguchi@nec.com>
Cc: NeilBrown <neilb@suse.de>
Cc: Peter Xu <peterx@redhat.com>
Cc: Suren Baghdasaryan <surenb@google.com>
Cc: Vlastimil Babka <vbabka@suse.cz>
Cc: Oscar Salvador <osalvador@suse.de>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm/swap: clean up the comment of find_next_to_unuse

Since commit 10a9c496789f ("mm: simplify try_to_unuse"), frontswap
parameter is removed.  Update the corresponding comment.

Link: https://lkml.kernel.org/r/20220509131416.17553-14-linmiaohe@huawei.com
Signed-off-by: Miaohe Lin <linmiaohe@huawei.com>
Reviewed-by: David Hildenbrand <david@redhat.com>
Cc: Alistair Popple <apopple@nvidia.com>
Cc: David Howells <dhowells@redhat.com>
Cc: Matthew Wilcox <willy@infradead.org>
Cc: Naoya Horiguchi <naoya.horiguchi@nec.com>
Cc: NeilBrown <neilb@suse.de>
Cc: Peter Xu <peterx@redhat.com>
Cc: Suren Baghdasaryan <surenb@google.com>
Cc: Vlastimil Babka <vbabka@suse.cz>
Cc: Oscar Salvador <osalvador@suse.de>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm/swap: fix the obsolete comment for SWP_TYPE_SHIFT

Since commit 3159f943aafd ("xarray: Replace exceptional entries"), there
is only one bit of 'type' can be shifted up.  Update the corresponding
comment.

Link: https://lkml.kernel.org/r/20220509131416.17553-13-linmiaohe@huawei.com
Signed-off-by: Miaohe Lin <linmiaohe@huawei.com>
Reviewed-by: David Hildenbrand <david@redhat.com>
Cc: Alistair Popple <apopple@nvidia.com>
Cc: David Howells <dhowells@redhat.com>
Cc: Matthew Wilcox <willy@infradead.org>
Cc: Naoya Horiguchi <naoya.horiguchi@nec.com>
Cc: NeilBrown <neilb@suse.de>
Cc: Peter Xu <peterx@redhat.com>
Cc: Suren Baghdasaryan <surenb@google.com>
Cc: Vlastimil Babka <vbabka@suse.cz>
Cc: Oscar Salvador <osalvador@suse.de>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm/swap: add helper swap_offset_available()

Add helper swap_offset_available() to remove some duplicated codes.  Minor
readability improvement.

[akpm@linux-foundation.org: s/swap_offset_available/swap_offset_available_and_locked/, per Neil]
Link: https://lkml.kernel.org/r/20220509131416.17553-12-linmiaohe@huawei.com
Signed-off-by: Miaohe Lin <linmiaohe@huawei.com>
Cc: Alistair Popple <apopple@nvidia.com>
Cc: David Hildenbrand <david@redhat.com>
Cc: David Howells <dhowells@redhat.com>
Cc: Matthew Wilcox <willy@infradead.org>
Cc: Naoya Horiguchi <naoya.horiguchi@nec.com>
Cc: NeilBrown <neilb@suse.de>
Cc: Peter Xu <peterx@redhat.com>
Cc: Suren Baghdasaryan <surenb@google.com>
Cc: Vlastimil Babka <vbabka@suse.cz>
Cc: Oscar Salvador <osalvador@suse.de>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm/swap: avoid calling swp_swap_info when try to check SWP_STABLE_WRITES

Use flags of si directly to check SWP_STABLE_WRITES to avoid possible
READ_ONCE and thus save some cpu cycles.

[akpm@linux-foundation.org: use data_race() on si->flags, per Neil]
Link: https://lkml.kernel.org/r/20220509131416.17553-10-linmiaohe@huawei.com
Signed-off-by: Miaohe Lin <linmiaohe@huawei.com>
Cc: Alistair Popple <apopple@nvidia.com>
Cc: David Hildenbrand <david@redhat.com>
Cc: David Howells <dhowells@redhat.com>
Cc: Matthew Wilcox <willy@infradead.org>
Cc: Naoya Horiguchi <naoya.horiguchi@nec.com>
Cc: NeilBrown <neilb@suse.de>
Cc: Peter Xu <peterx@redhat.com>
Cc: Suren Baghdasaryan <surenb@google.com>
Cc: Vlastimil Babka <vbabka@suse.cz>
Cc: Oscar Salvador <osalvador@suse.de>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm/swap: make page_swapcount and __lru_add_drain_all static

Make page_swapcount and __lru_add_drain_all static.  They are only used
within the file now.

Link: https://lkml.kernel.org/r/20220509131416.17553-9-linmiaohe@huawei.com
Signed-off-by: Miaohe Lin <linmiaohe@huawei.com>
Reviewed-by: David Hildenbrand <david@redhat.com>
Reviewed-by: Oscar Salvador <osalvador@suse.de>
Cc: Alistair Popple <apopple@nvidia.com>
Cc: David Howells <dhowells@redhat.com>
Cc: Matthew Wilcox <willy@infradead.org>
Cc: Naoya Horiguchi <naoya.horiguchi@nec.com>
Cc: NeilBrown <neilb@suse.de>
Cc: Peter Xu <peterx@redhat.com>
Cc: Suren Baghdasaryan <surenb@google.com>
Cc: Vlastimil Babka <vbabka@suse.cz>
Cc: Oscar Salvador <osalvador@suse.de>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm/swap: remove unneeded p != NULL check in __swap_duplicate

If p is NULL, __swap_duplicate will already return -EINVAL.  So if we
reach here, p must be non-NULL.

Link: https://lkml.kernel.org/r/20220509131416.17553-8-linmiaohe@huawei.com
Signed-off-by: Miaohe Lin <linmiaohe@huawei.com>
Reviewed-by: David Hildenbrand <david@redhat.com>
Reviewed-by: Oscar Salvador <osalvador@suse.de>
Cc: Alistair Popple <apopple@nvidia.com>
Cc: David Howells <dhowells@redhat.com>
Cc: Matthew Wilcox <willy@infradead.org>
Cc: Naoya Horiguchi <naoya.horiguchi@nec.com>
Cc: NeilBrown <neilb@suse.de>
Cc: Peter Xu <peterx@redhat.com>
Cc: Suren Baghdasaryan <surenb@google.com>
Cc: Vlastimil Babka <vbabka@suse.cz>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm/swap: remove buggy cache->nr check in refill_swap_slots_cache

refill_swap_slots_cache is always called when cache->nr is 0.  So remove
such buggy and confusing check.

Link: https://lkml.kernel.org/r/20220509131416.17553-7-linmiaohe@huawei.com
Signed-off-by: Miaohe Lin <linmiaohe@huawei.com>
Acked-by: David Hildenbrand <david@redhat.com>
Reviewed-by: Oscar Salvador <osalvador@suse.de>
Cc: Alistair Popple <apopple@nvidia.com>
Cc: David Howells <dhowells@redhat.com>
Cc: Matthew Wilcox <willy@infradead.org>
Cc: Naoya Horiguchi <naoya.horiguchi@nec.com>
Cc: NeilBrown <neilb@suse.de>
Cc: Peter Xu <peterx@redhat.com>
Cc: Suren Baghdasaryan <surenb@google.com>
Cc: Vlastimil Babka <vbabka@suse.cz>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm/swap: print bad swap offset entry in get_swap_device

If offset exceeds the si->max, print bad swap offset entry to help debug
the unexpected case.

Link: https://lkml.kernel.org/r/20220509131416.17553-6-linmiaohe@huawei.com
Signed-off-by: Miaohe Lin <linmiaohe@huawei.com>
Reviewed-by: David Hildenbrand <david@redhat.com>
Reviewed-by: Oscar Salvador <osalvador@suse.de>
Cc: Alistair Popple <apopple@nvidia.com>
Cc: David Howells <dhowells@redhat.com>
Cc: Matthew Wilcox <willy@infradead.org>
Cc: Naoya Horiguchi <naoya.horiguchi@nec.com>
Cc: NeilBrown <neilb@suse.de>
Cc: Peter Xu <peterx@redhat.com>
Cc: Suren Baghdasaryan <surenb@google.com>
Cc: Vlastimil Babka <vbabka@suse.cz>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm/swap: remove unneeded return value of free_swap_slot

The return value of free_swap_slot is always 0 and also ignored now.
Remove it to clean up the code.

Link: https://lkml.kernel.org/r/20220509131416.17553-5-linmiaohe@huawei.com
Signed-off-by: Miaohe Lin <linmiaohe@huawei.com>
Reviewed-by: David Hildenbrand <david@redhat.com>
Reviewed-by: Oscar Salvador <osalvador@suse.de>
Cc: Alistair Popple <apopple@nvidia.com>
Cc: David Howells <dhowells@redhat.com>
Cc: Matthew Wilcox <willy@infradead.org>
Cc: Naoya Horiguchi <naoya.horiguchi@nec.com>
Cc: NeilBrown <neilb@suse.de>
Cc: Peter Xu <peterx@redhat.com>
Cc: Suren Baghdasaryan <surenb@google.com>
Cc: Vlastimil Babka <vbabka@suse.cz>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm/swap: fold __swap_info_get() into its sole caller

Fold __swap_info_get() into its sole caller to make code more clear.
Minor readability improvement.

Link: https://lkml.kernel.org/r/20220509131416.17553-4-linmiaohe@huawei.com
Signed-off-by: Miaohe Lin <linmiaohe@huawei.com>
Reviewed-by: David Hildenbrand <david@redhat.com>
Reviewed-by: Oscar Salvador <osalvador@suse.de>
Cc: Alistair Popple <apopple@nvidia.com>
Cc: David Howells <dhowells@redhat.com>
Cc: Matthew Wilcox <willy@infradead.org>
Cc: Naoya Horiguchi <naoya.horiguchi@nec.com>
Cc: NeilBrown <neilb@suse.de>
Cc: Peter Xu <peterx@redhat.com>
Cc: Suren Baghdasaryan <surenb@google.com>
Cc: Vlastimil Babka <vbabka@suse.cz>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm/swap: use helper macro __ATTR_RW

Use helper macro __ATTR_RW to define vma_ra_enabled_attr to make code more
clear.  Minor readability improvement.

Link: https://lkml.kernel.org/r/20220509131416.17553-3-linmiaohe@huawei.com
Signed-off-by: Miaohe Lin <linmiaohe@huawei.com>
Reviewed-by: David Hildenbrand <david@redhat.com>
Reviewed-by: Oscar Salvador <osalvador@suse.de>
Cc: Alistair Popple <apopple@nvidia.com>
Cc: David Howells <dhowells@redhat.com>
Cc: Matthew Wilcox <willy@infradead.org>
Cc: Naoya Horiguchi <naoya.horiguchi@nec.com>
Cc: NeilBrown <neilb@suse.de>
Cc: Peter Xu <peterx@redhat.com>
Cc: Suren Baghdasaryan <surenb@google.com>
Cc: Vlastimil Babka <vbabka@suse.cz>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm/swap: use helper is_swap_pte() in swap_vma_readahead

Patch series "A few cleanup patches for swap".

This series contains a few patches to fix the comment, remove unneeded
return value, use some helpers and so on.  More details can be found in
the respective changelogs.

This patch (of 14):

Use helper is_swap_pte() to check whether pte is swap entry to make code
more clear.  Minor readability improvement.

Link: https://lkml.kernel.org/r/20220509131416.17553-1-linmiaohe@huawei.com
Link: https://lkml.kernel.org/r/20220509131416.17553-2-linmiaohe@huawei.com
Signed-off-by: Miaohe Lin <linmiaohe@huawei.com>
Reviewed-by: David Hildenbrand <david@redhat.com>
Reviewed-by: Oscar Salvador <osalvador@suse.de>
Cc: Matthew Wilcox <willy@infradead.org>
Cc: Vlastimil Babka <vbabka@suse.cz>
Cc: David Howells <dhowells@redhat.com>
Cc: NeilBrown <neilb@suse.de>
Cc: Alistair Popple <apopple@nvidia.com>
Cc: Suren Baghdasaryan <surenb@google.com>
Cc: Peter Xu <peterx@redhat.com>
Cc: Naoya Horiguchi <naoya.horiguchi@nec.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm: mmap: register suitable readonly file vmas for khugepaged

The readonly FS THP relies on khugepaged to collapse THP for suitable
vmas.  But the behavior is inconsistent for "always" mode
(https://lore.kernel.org/linux-mm/00f195d4-d039-3cf2-d3a1-a2c88de397a0@suse.cz/).

The "always" mode means THP allocation should be tried all the time and
khugepaged should try to collapse THP all the time.  Of course the
allocation and collapse may fail due to other factors and conditions.

Currently file THP may not be collapsed by khugepaged even though all the
conditions are met.  That does break the semantics of "always" mode.

So make sure readonly FS vmas are registered to khugepaged to fix the
break.

Register suitable vmas in common mmap path, that could cover both readonly
FS vmas and shmem vmas, so remove the khugepaged calls in shmem.c.

Still need to keep the khugepaged call in vma_merge() since vma_merge() is
called in a lot of places, for example, madvise, mprotect, etc.

Link: https://lkml.kernel.org/r/20220510203222.24246-9-shy828301@gmail.com
Signed-off-by: Yang Shi <shy828301@gmail.com>
Reported-by: Vlastimil Babka <vbabka@suse.cz>
Acked-by: Vlastmil Babka <vbabka@suse.cz>
Cc: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: Miaohe Lin <linmiaohe@huawei.com>
Cc: Song Liu <songliubraving@fb.com>
Cc: Rik van Riel <riel@surriel.com>
Cc: Matthew Wilcox (Oracle) <willy@infradead.org>
Cc: Zi Yan <ziy@nvidia.com>
Cc: Theodore Ts'o <tytso@mit.edu>
Cc: Song Liu <song@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm: khugepaged: introduce khugepaged_enter_vma() helper

The khugepaged_enter_vma_merge() actually does as the same thing as the
khugepaged_enter() section called by shmem_mmap(), so consolidate them
into one helper and rename it to khugepaged_enter_vma().

Link: https://lkml.kernel.org/r/20220510203222.24246-8-shy828301@gmail.com
Signed-off-by: Yang Shi <shy828301@gmail.com>
Acked-by: Vlastmil Babka <vbabka@suse.cz>
Cc: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: Matthew Wilcox (Oracle) <willy@infradead.org>
Cc: Miaohe Lin <linmiaohe@huawei.com>
Cc: Rik van Riel <riel@surriel.com>
Cc: Song Liu <song@kernel.org>
Cc: Song Liu <songliubraving@fb.com>
Cc: Theodore Ts'o <tytso@mit.edu>
Cc: Zi Yan <ziy@nvidia.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm: khugepaged: make hugepage_vma_check() non-static

The hugepage_vma_check() could be reused by khugepaged_enter() and
khugepaged_enter_vma_merge(), but it is static in khugepaged.c.  Make it
non-static and declare it in khugepaged.h.

Link: https://lkml.kernel.org/r/20220510203222.24246-7-shy828301@gmail.com
Signed-off-by: Yang Shi <shy828301@gmail.com>
Suggested-by: Vlastimil Babka <vbabka@suse.cz>
Cc: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: Matthew Wilcox (Oracle) <willy@infradead.org>
Cc: Miaohe Lin <linmiaohe@huawei.com>
Cc: Rik van Riel <riel@surriel.com>
Cc: Song Liu <song@kernel.org>
Cc: Song Liu <songliubraving@fb.com>
Cc: Theodore Ts'o <tytso@mit.edu>
Cc: Zi Yan <ziy@nvidia.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm: khugepaged: make khugepaged_enter() void function

The most callers of khugepaged_enter() don't care about the return value.
Only dup_mmap(), anonymous THP page fault and MADV_HUGEPAGE handle the
error by returning -ENOMEM.  Actually it is not harmful for them to ignore
the error case either.  It also sounds overkilling to fail fork() and page
fault early due to khugepaged_enter() error, and MADV_HUGEPAGE does set
VM_HUGEPAGE flag regardless of the error.

Link: https://lkml.kernel.org/r/20220510203222.24246-6-shy828301@gmail.com
Signed-off-by: Yang Shi <shy828301@gmail.com>
Acked-by: Song Liu <song@kernel.org>
Acked-by: Vlastmil Babka <vbabka@suse.cz>
Cc: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: Matthew Wilcox (Oracle) <willy@infradead.org>
Cc: Miaohe Lin <linmiaohe@huawei.com>
Cc: Rik van Riel <riel@surriel.com>
Cc: Song Liu <songliubraving@fb.com>
Cc: Theodore Ts'o <tytso@mit.edu>
Cc: Zi Yan <ziy@nvidia.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm: thp: only regular file could be THP eligible

Since commit a4aeaa06d45e ("mm: khugepaged: skip huge page collapse for
special files"), khugepaged just collapses THP for regular file which is
the intended usecase for readonly fs THP.  Only show regular file as THP
eligible accordingly.

And make file_thp_enabled() available for khugepaged too in order to
remove duplicate code.

Link: https://lkml.kernel.org/r/20220510203222.24246-5-shy828301@gmail.com
Signed-off-by: Yang Shi <shy828301@gmail.com>
Acked-by: Song Liu <song@kernel.org>
Acked-by: Vlastmil Babka <vbabka@suse.cz>
Cc: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: Matthew Wilcox (Oracle) <willy@infradead.org>
Cc: Miaohe Lin <linmiaohe@huawei.com>
Cc: Rik van Riel <riel@surriel.com>
Cc: Song Liu <songliubraving@fb.com>
Cc: Theodore Ts'o <tytso@mit.edu>
Cc: Zi Yan <ziy@nvidia.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm: khugepaged: skip DAX vma

The DAX vma may be seen by khugepaged when the mm has other khugepaged
suitable vmas.  So khugepaged may try to collapse THP for DAX vma, but it
will fail due to page sanity check, for example, page is not on LRU.

So it is not harmful, but it is definitely pointless to run khugepaged
against DAX vma, so skip it in early check.

Link: https://lkml.kernel.org/r/20220510203222.24246-4-shy828301@gmail.com
Signed-off-by: Yang Shi <shy828301@gmail.com>
Reviewed-by: Miaohe Lin <linmiaohe@huawei.com>
Acked-by: Song Liu <song@kernel.org>
Acked-by: Vlastmil Babka <vbabka@suse.cz>
Cc: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: Matthew Wilcox (Oracle) <willy@infradead.org>
Cc: Rik van Riel <riel@surriel.com>
Cc: Song Liu <songliubraving@fb.com>
Cc: Theodore Ts'o <tytso@mit.edu>
Cc: Zi Yan <ziy@nvidia.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm: khugepaged: remove redundant check for VM_NO_KHUGEPAGED

The hugepage_vma_check() called by khugepaged_enter_vma_merge() does check
VM_NO_KHUGEPAGED.  Remove the check from caller and move the check in
hugepage_vma_check() up.

More checks may be run for VM_NO_KHUGEPAGED vmas, but MADV_HUGEPAGE is
definitely not a hot path, so cleaner code does outweigh.

Link: https://lkml.kernel.org/r/20220510203222.24246-3-shy828301@gmail.com
Signed-off-by: Yang Shi <shy828301@gmail.com>
Reviewed-by: Miaohe Lin <linmiaohe@huawei.com>
Acked-by: Song Liu <song@kernel.org>
Acked-by: Vlastmil Babka <vbabka@suse.cz>
Cc: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: Matthew Wilcox (Oracle) <willy@infradead.org>
Cc: Rik van Riel <riel@surriel.com>
Cc: Song Liu <songliubraving@fb.com>
Cc: Theodore Ts'o <tytso@mit.edu>
Cc: Zi Yan <ziy@nvidia.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

sched: coredump.h: clarify the use of MMF_VM_HUGEPAGE

Patch series "Make khugepaged collapse readonly FS THP more consistent", v4.

The readonly FS THP relies on khugepaged to collapse THP for suitable
vmas.  But the behavior is inconsistent for "always" mode
(https://lore.kernel.org/linux-mm/00f195d4-d039-3cf2-d3a1-a2c88de397a0@suse.cz/).

The "always" mode means THP allocation should be tried all the time and
khugepaged should try to collapse THP all the time.  Of course the
allocation and collapse may fail due to other factors and conditions.

Currently file THP may not be collapsed by khugepaged even though all the
conditions are met.  That does break the semantics of "always" mode.

So make sure readonly FS vmas are registered to khugepaged to fix the
break.

Register suitable vmas in common mmap path, that could cover both readonly
FS vmas and shmem vmas, so remove the khugepaged calls in shmem.c.

The patch 1-7 are minor bug fixes, clean up and preparation patches.
Patch 8 is the real meat.

Tested with khugepaged test in selftests and the testcase provided by
Vlastimil Babka in
https://lore.kernel.org/lkml/df3b5d1c-a36b-2c73-3e27-99e74983de3a@suse.cz/
by commenting out MADV_HUGEPAGE call.

This patch (of 8):

MMF_VM_HUGEPAGE is set as long as the mm is available for khugepaged by
khugepaged_enter(), not only when VM_HUGEPAGE is set on vma.  Correct the
comment to avoid confusion.

Link: https://lkml.kernel.org/r/20220510203222.24246-1-shy828301@gmail.com
Link: https://lkml.kernel.org/r/20220510203222.24246-2-shy828301@gmail.com
Signed-off-by: Yang Shi <shy828301@gmail.com>
Reviewed-by: Miaohe Lin <linmiaohe@huawei.com>
Acked-by: Song Liu <song@kernel.org>
Acked-by: Vlastmil Babka <vbabka@suse.cz>
Cc: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: Matthew Wilcox (Oracle) <willy@infradead.org>
Cc: Rik van Riel <riel@surriel.com>
Cc: Song Liu <songliubraving@fb.com>
Cc: Theodore Ts'o <tytso@mit.edu>
Cc: Zi Yan <ziy@nvidia.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

arm64/mm: fix page table check compile error for CONFIG_PGTABLE_LEVELS=2

If CONFIG_PGTABLE_LEVELS=2 and CONFIG_ARCH_SUPPORTS_PAGE_TABLE_CHECK=y,
then we trigger a compile error:

  error: implicit declaration of function 'pte_user_accessible_page'

Move the definition of page table check helper out of branch
CONFIG_PGTABLE_LEVELS > 2

Link: https://lkml.kernel.org/r/20220517074548.2227779-3-tongtiangen@huawei.com
Fixes: daf214c14dbe ("arm64/mm: enable ARCH_SUPPORTS_PAGE_TABLE_CHECK")
Signed-off-by: Tong Tiangen <tongtiangen@huawei.com>
Acked-by: Catalin Marinas <catalin.marinas@arm.com>
Cc: Anshuman Khandual <anshuman.khandual@arm.com>
Cc: Pasha Tatashin <pasha.tatashin@soleen.com>
Cc: Anshuman Khandual <anshuman.khandual@arm.com>
Cc: Will Deacon <will@kernel.org>
Cc: Paul Walmsley <paul.walmsley@sifive.com>
Cc: Palmer Dabbelt <palmer@dabbelt.com>
Cc: Albert Ou <aou@eecs.berkeley.edu>
Cc: Arnd Bergmann <arnd@arndb.de>
Cc: Guohanjun <guohanjun@huawei.com>
Cc: Xie XiuQi <xiexiuqi@huawei.com>
Cc: kernel test robot <lkp@intel.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

riscv/mm: fix two page table check related issues

Two page table check related issues have been fixed here.

1. Open CONFIG_PAGE_TABLE_CHECK in riscv32, we got a compile error[1]:

   error: implicit declaration of function 'pud_leaf'

   Add pud_leaf() definition to incluce/asm-generic/pgtable-nopmd.h to fix
   this issue.

2. Keep consistent with other pud_xxx() helpers, move pud_user() to
   pgtable-64.h and add pud_user() to pgtable-nopmd.h.

[1]https://lore.kernel.org/linux-mm/202205161811.2nLxmN2O-lkp@intel.com/T/

Link: https://lkml.kernel.org/r/20220517074548.2227779-2-tongtiangen@huawei.com
Fixes: 856eed79f8d3 ("riscv/mm: enable ARCH_SUPPORTS_PAGE_TABLE_CHECK")
Signed-off-by: Tong Tiangen <tongtiangen@huawei.com>
Reported-by: kernel test robot <lkp@intel.com>
Cc: Anshuman Khandual <anshuman.khandual@arm.com>
Cc: Pasha Tatashin <pasha.tatashin@soleen.com>
Cc: Anshuman Khandual <anshuman.khandual@arm.com>
Cc: Albert Ou <aou@eecs.berkeley.edu>
Cc: Arnd Bergmann <arnd@arndb.de>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Guohanjun <guohanjun@huawei.com>
Cc: Palmer Dabbelt <palmer@dabbelt.com>
Cc: Paul Walmsley <paul.walmsley@sifive.com>
Cc: Will Deacon <will@kernel.org>
Cc: Xie XiuQi <xiexiuqi@huawei.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm, compaction: fast_find_migrateblock() should return pfn in the target zone

At present, pages not in the target zone are added to cc->migratepages
list in isolate_migratepages_block().  As a result, pages may migrate
between nodes unintentionally.

This would be a serious problem for older kernels without commit
a984226f457f849e ("mm: memcontrol: remove the pgdata parameter of
mem_cgroup_page_lruvec"), because it can corrupt the lru list by
handling pages in list without holding proper lru_lock.

Avoid returning a pfn outside the target zone in the case that it is
not aligned with a pageblock boundary.  Otherwise
isolate_migratepages_block() will handle pages not in the target zone.

Link: https://lkml.kernel.org/r/20220511044300.4069-1-yamamoto.rei@jp.fujitsu.com
Fixes: 70b44595eafe ("mm, compaction: use free lists to quickly locate a migration source")
Signed-off-by: Rei Yamamoto <yamamoto.rei@jp.fujitsu.com>
Reviewed-by: Miaohe Lin <linmiaohe@huawei.com>
Acked-by: Mel Gorman <mgorman@techsingularity.net>
Reviewed-by: Oscar Salvador <osalvador@suse.de>
Cc: Don Dutile <ddutile@redhat.com>
Cc: Wonhyuk Yang <vvghjk1234@gmail.com>
Cc: Rei Yamamoto <yamamoto.rei@jp.fujitsu.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm/damon: add documentation for Enum value

Fix the warning - "Enum value 'NR_DAMON_OPS' not described in enum
'damon_ops_id'" generated by the command "make pdfdocs"

Link: https://lkml.kernel.org/r/20220508073316.141401-1-gautammenghani201@gmail.com
Signed-off-by: Gautam Menghani <gautammenghani201@gmail.com>
Reviewed-by: SeongJae Park <sj@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm/memcontrol: export memcg->watermark via sysfs for v2 memcg

We run a lot of automated tests when building our software and run into
OOM scenarios when the tests run unbounded.  v1 memcg exports
memcg->watermark as "memory.max_usage_in_bytes" in sysfs.  We use this
metric to heuristically limit the number of tests that can run in parallel
based on per test historical data.

This metric is currently not exported for v2 memcg and there is no other
easy way of getting this information.  getrusage() syscall returns
"ru_maxrss" which can be used as an approximation but that's the max RSS
of a single child process across all children instead of the aggregated
max for all child processes.  The only work around is to periodically poll
"memory.current" but that's not practical for short-lived one-off cgroups.

Hence, expose memcg->watermark as "memory.peak" for v2 memcg.

Link: https://lkml.kernel.org/r/20220507050916.GA13577@us192.sjc.aristanetworks.com
Signed-off-by: Ganesan Rajagopal <rganesan@arista.com>
Acked-by: Shakeel Butt <shakeelb@google.com>
Acked-by: Johannes Weiner <hannes@cmpxchg.org>
Acked-by: Roman Gushchin <roman.gushchin@linux.dev>
Acked-by: Michal Hocko <mhocko@suse.com>
Cc: Roman Gushchin <roman.gushchin@linux.dev>
Reviewed-by: Michal Koutný <mkoutny@suse.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm: hugetlb_vmemmap: add hugetlb_optimize_vmemmap sysctl

We must add hugetlb_free_vmemmap=on (or "off") to the boot cmdline and
reboot the server to enable or disable the feature of optimizing vmemmap
pages associated with HugeTLB pages.  However, rebooting usually takes a
long time.  So add a sysctl to enable or disable the feature at runtime
without rebooting.  Why we need this?  There are 3 use cases.

1) The feature of minimizing overhead of struct page associated with
   each HugeTLB is disabled by default without passing
   "hugetlb_free_vmemmap=on" to the boot cmdline.  When we (ByteDance)
   deliver the servers to the users who want to enable this feature, they
   have to configure the grub (change boot cmdline) and reboot the
   servers, whereas rebooting usually takes a long time (we have thousands
   of servers).  It's a very bad experience for the users.  So we need a
   approach to enable this feature after rebooting.  This is a use case in
   our practical environment.

2) Some use cases are that HugeTLB pages are allocated 'on the fly'
   instead of being pulled from the HugeTLB pool, those workloads would be
   affected with this feature enabled.  Those workloads could be
   identified by the characteristics of they never explicitly allocating
   huge pages with 'nr_hugepages' but only set 'nr_overcommit_hugepages'
   and then let the pages be allocated from the buddy allocator at fault
   time.  We can confirm it is a real use case from the commit
   099730d67417.  For those workloads, the page fault time could be ~2x
   slower than before.  We suspect those users want to disable this
   feature if the system has enabled this before and they don't think the
   memory savings benefit is enough to make up for the performance drop.

3) If the workload which wants vmemmap pages to be optimized and the
   workload which wants to set 'nr_overcommit_hugepages' and does not want
   the extera overhead at fault time when the overcommitted pages be
   allocated from the buddy allocator are deployed in the same server.
   The user could enable this feature and set 'nr_hugepages' and
   'nr_overcommit_hugepages', then disable the feature.  In this case, the
   overcommited HugeTLB pages will not encounter the extra overhead at
   fault time.

Link: https://lkml.kernel.org/r/20220512041142.39501-5-songmuchun@bytedance.com
Signed-off-by: Muchun Song <songmuchun@bytedance.com>
Reviewed-by: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: Luis Chamberlain <mcgrof@kernel.org>
Cc: Kees Cook <keescook@chromium.org>
Cc: Iurii Zaikin <yzaikin@google.com>
Cc: Oscar Salvador <osalvador@suse.de>
Cc: David Hildenbrand <david@redhat.com>
Cc: Masahiro Yamada <masahiroy@kernel.org>
Cc: Xiongchun Duan <duanxiongchun@bytedance.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm: hugetlb_vmemmap: use kstrtobool for hugetlb_vmemmap param parsing

Use kstrtobool rather than open coding "on" and "off" parsing in
mm/hugetlb_vmemmap.c, which is more powerful to handle all kinds of
parameters like 'Yy1Nn0' or [oO][NnFf] for "on" and "off".

Link: https://lkml.kernel.org/r/20220512041142.39501-4-songmuchun@bytedance.com
Signed-off-by: Muchun Song <songmuchun@bytedance.com>
Reviewed-by: Mike Kravetz <mike.kravetz@oracle.com>
Cc: David Hildenbrand <david@redhat.com>
Cc: Iurii Zaikin <yzaikin@google.com>
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: Kees Cook <keescook@chromium.org>
Cc: Luis Chamberlain <mcgrof@kernel.org>
Cc: Masahiro Yamada <masahiroy@kernel.org>
Cc: Oscar Salvador <osalvador@suse.de>
Cc: Xiongchun Duan <duanxiongchun@bytedance.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm: memory_hotplug: override memmap_on_memory when hugetlb_free_vmemmap=on

Optimizing HugeTLB vmemmap pages is not compatible with allocating memmap
on hot added memory.  If "hugetlb_free_vmemmap=on" and
memory_hotplug.memmap_on_memory" are both passed on the kernel command
line, optimizing hugetlb pages takes precedence.  However, the global
variable memmap_on_memory will still be set to 1, even though we will not
try to allocate memmap on hot added memory.

Also introduce mhp_memmap_on_memory() helper to move the definition of
"memmap_on_memory" to the scope of CONFIG_MHP_MEMMAP_ON_MEMORY.  In the
next patch, mhp_memmap_on_memory() will also be exported to be used in
hugetlb_vmemmap.c.

Link: https://lkml.kernel.org/r/20220512041142.39501-3-songmuchun@bytedance.com
Signed-off-by: Muchun Song <songmuchun@bytedance.com>
Acked-by: Mike Kravetz <mike.kravetz@oracle.com>
Cc: David Hildenbrand <david@redhat.com>
Cc: Iurii Zaikin <yzaikin@google.com>
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: Kees Cook <keescook@chromium.org>
Cc: Luis Chamberlain <mcgrof@kernel.org>
Cc: Masahiro Yamada <masahiroy@kernel.org>
Cc: Oscar Salvador <osalvador@suse.de>
Cc: Xiongchun Duan <duanxiongchun@bytedance.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm: hugetlb_vmemmap: disable hugetlb_optimize_vmemmap when struct page crosses page boundaries

Patch series "add hugetlb_optimize_vmemmap sysctl", v11.

This series aims to add hugetlb_optimize_vmemmap sysctl to enable or
disable the feature of optimizing vmemmap pages associated with HugeTLB
pages.

This patch (of 4):

If the size of "struct page" is not the power of two but with the feature
of minimizing overhead of struct page associated with each HugeTLB is
enabled, then the vmemmap pages of HugeTLB will be corrupted after
remapping (panic is about to happen in theory).  But this only exists when
!CONFIG_MEMCG && !CONFIG_SLUB on x86_64.  However, it is not a
conventional configuration nowadays.  So it is not a real word issue, just
the result of a code review.

But we cannot prevent anyone from configuring that combined configure.
This hugetlb_optimize_vmemmap should be disable in this case to fix this
issue.

Link: https://lkml.kernel.org/r/20220512041142.39501-1-songmuchun@bytedance.com
Link: https://lkml.kernel.org/r/20220512041142.39501-2-songmuchun@bytedance.com
Signed-off-by: Muchun Song <songmuchun@bytedance.com>
Reviewed-by: Mike Kravetz <mike.kravetz@oracle.com>
Cc: David Hildenbrand <david@redhat.com>
Cc: Iurii Zaikin <yzaikin@google.com>
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: Kees Cook <keescook@chromium.org>
Cc: Luis Chamberlain <mcgrof@kernel.org>
Cc: Masahiro Yamada <masahiroy@kernel.org>
Cc: Oscar Salvador <osalvador@suse.de>
Cc: Xiongchun Duan <duanxiongchun@bytedance.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm: rmap: fix CONT-PTE/PMD size hugetlb issue when unmapping

On some architectures (like ARM64), it can support CONT-PTE/PMD size
hugetlb, which means it can support not only PMD/PUD size hugetlb: 2M and
1G, but also CONT-PTE/PMD size: 64K and 32M if a 4K page size specified.

When unmapping a hugetlb page, we will get the relevant page table entry
by huge_pte_offset() only once to nuke it.  This is correct for PMD or PUD
size hugetlb, since they always contain only one pmd entry or pud entry in
the page table.

However this is incorrect for CONT-PTE and CONT-PMD size hugetlb, since
they can contain several continuous pte or pmd entry with same page table
attributes, so we will nuke only one pte or pmd entry for this
CONT-PTE/PMD size hugetlb page.

And now try_to_unmap() is only passed a hugetlb page in the case where the
hugetlb page is poisoned.  Which means now we will unmap only one pte
entry for a CONT-PTE or CONT-PMD size poisoned hugetlb page, and we can
still access other subpages of a CONT-PTE or CONT-PMD size poisoned
hugetlb page, which will cause serious issues possibly.

So we should change to use huge_ptep_clear_flush() to nuke the hugetlb
page table to fix this issue, which already considered CONT-PTE and
CONT-PMD size hugetlb.

We've already used set_huge_swap_pte_at() to set a poisoned swap entry for
a poisoned hugetlb page.  Meanwhile adding a VM_BUG_ON() to make sure the
passed hugetlb page is poisoned in try_to_unmap().

Link: https://lkml.kernel.org/r/0a2e547238cad5bc153a85c3e9658cb9d55f9cac.1652270205.git.baolin.wang@linux.alibaba.com
Link: https://lkml.kernel.org/r/730ea4b6d292f32fb10b7a4e87dad49b0eb30474.1652147571.git.baolin.wang@linux.alibaba.com
Signed-off-by: Baolin Wang <baolin.wang@linux.alibaba.com>
Reviewed-by: Muchun Song <songmuchun@bytedance.com>
Reviewed-by: Mike Kravetz <mike.kravetz@oracle.com>
Acked-by: David Hildenbrand <david@redhat.com>
Cc: Alexander Gordeev <agordeev@linux.ibm.com>
Cc: Arnd Bergmann <arnd@arndb.de>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Christian Borntraeger <borntraeger@linux.ibm.com>
Cc: David S. Miller <davem@davemloft.net>
Cc: Gerald Schaefer <gerald.schaefer@linux.ibm.com>
Cc: Heiko Carstens <hca@linux.ibm.com>
Cc: Helge Deller <deller@gmx.de>
Cc: James Bottomley <James.Bottomley@HansenPartnership.com>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Rich Felker <dalias@libc.org>
Cc: Sven Schnelle <svens@linux.ibm.com>
Cc: Thomas Bogendoerfer <tsbogend@alpha.franken.de>
Cc: Vasily Gorbik <gor@linux.ibm.com>
Cc: Will Deacon <will@kernel.org>
Cc: Yoshinori Sato <ysato@users.osdn.me>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm: rmap: fix CONT-PTE/PMD size hugetlb issue when migration

On some architectures (like ARM64), it can support CONT-PTE/PMD size
hugetlb, which means it can support not only PMD/PUD size hugetlb: 2M and
1G, but also CONT-PTE/PMD size: 64K and 32M if a 4K page size specified.

When migrating a hugetlb page, we will get the relevant page table entry
by huge_pte_offset() only once to nuke it and remap it with a migration
pte entry.  This is correct for PMD or PUD size hugetlb, since they always
contain only one pmd entry or pud entry in the page table.

However this is incorrect for CONT-PTE and CONT-PMD size hugetlb, since
they can contain several continuous pte or pmd entry with same page table
attributes.  So we will nuke or remap only one pte or pmd entry for this
CONT-PTE/PMD size hugetlb page, which is not expected for hugetlb
migration.  The problem is we can still continue to modify the subpages'
data of a hugetlb page during migrating a hugetlb page, which can cause a
serious data consistent issue, since we did not nuke the page table entry
and set a migration pte for the subpages of a hugetlb page.

To fix this issue, we should change to use huge_ptep_clear_flush() to nuke
a hugetlb page table, and remap it with set_huge_pte_at() and
set_huge_swap_pte_at() when migrating a hugetlb page, which already
considered the CONT-PTE or CONT-PMD size hugetlb.

[akpm@linux-foundation.org: fix nommu build]
[baolin.wang@linux.alibaba.com: fix build errors for !CONFIG_MMU]
Link: https://lkml.kernel.org/r/a4baca670aca637e7198d9ae4543b8873cb224dc.1652270205.git.baolin.wang@linux.alibaba.com
Link: https://lkml.kernel.org/r/ea5abf529f0997b5430961012bfda6166c1efc8c.1652147571.git.baolin.wang@linux.alibaba.com
Signed-off-by: Baolin Wang <baolin.wang@linux.alibaba.com>
Reviewed-by: Muchun Song <songmuchun@bytedance.com>
Reviewed-by: Mike Kravetz <mike.kravetz@oracle.com>
Acked-by: David Hildenbrand <david@redhat.com>
Cc: Alexander Gordeev <agordeev@linux.ibm.com>
Cc: Arnd Bergmann <arnd@arndb.de>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Christian Borntraeger <borntraeger@linux.ibm.com>
Cc: David S. Miller <davem@davemloft.net>
Cc: Gerald Schaefer <gerald.schaefer@linux.ibm.com>
Cc: Heiko Carstens <hca@linux.ibm.com>
Cc: Helge Deller <deller@gmx.de>
Cc: James Bottomley <James.Bottomley@HansenPartnership.com>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Rich Felker <dalias@libc.org>
Cc: Sven Schnelle <svens@linux.ibm.com>
Cc: Thomas Bogendoerfer <tsbogend@alpha.franken.de>
Cc: Vasily Gorbik <gor@linux.ibm.com>
Cc: Will Deacon <will@kernel.org>
Cc: Yoshinori Sato <ysato@users.osdn.me>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm: change huge_ptep_clear_flush() to return the original pte

Patch series "Fix CONT-PTE/PMD size hugetlb issue when unmapping or migrating", v4.

presently, migrating a hugetlb page or unmapping a poisoned hugetlb page,
we'll use ptep_clear_flush() and set_pte_at() to nuke the page table entry
and remap it, and this is incorrect for CONT-PTE or CONT-PMD size hugetlb
page, which will cause potential data consistent issue.  This patch set
will change to use hugetlb related APIs to fix this issue.

Note: Mike pointed out the huge_ptep_get() will only return the one
specific value, and it would not take into account the dirty or young bits
of CONT-PTE/PMDs like the huge_ptep_get_and_clear() [1].  This
inconsistent issue is not introduced by this patch set, and this issue
will be addressed in another thread [2].  Meanwhile the uffd for hugetlb
case [3] pointed out by Gerald also needs another patch to address.

[1] https://lore.kernel.org/linux-mm/85bd80b4-b4fd-0d3f-a2e5-149559f2f387@oracle.com/
[2] https://lore.kernel.org/all/cover.1651998586.git.baolin.wang@linux.alibaba.com/
[3] https://lore.kernel.org/linux-mm/20220503120343.6264e126@thinkpad/

This patch (of 3):

It is incorrect to use ptep_clear_flush() to nuke a hugetlb page table
when unmapping or migrating a hugetlb page, and will change to use
huge_ptep_clear_flush() instead in the following patches.

So this is a preparation patch, which changes the huge_ptep_clear_flush()
to return the original pte to help to nuke a hugetlb page table.

[baolin.wang@linux.alibaba.com: fix build in several more architectures]
Link: https://lkml.kernel.org/r/0009a4cd-2826-e8be-e671-f050d4f18d5d@linux.alibaba.com
[sfr@canb.auug.org.au: fixup]
Link: https://lkml.kernel.org/r/20220511181531.7f27a5c1@canb.auug.org.au
Link: https://lkml.kernel.org/r/cover.1652270205.git.baolin.wang@linux.alibaba.com
Link: https://lkml.kernel.org/r/20f77ddab90baa249bd24504c413189b82acde69.1652270205.git.baolin.wang@linux.alibaba.com
Link: https://lkml.kernel.org/r/cover.1652147571.git.baolin.wang@linux.alibaba.com
Link: https://lkml.kernel.org/r/dcf065868cce35bceaf138613ad27f17bb7c0c19.1652147571.git.baolin.wang@linux.alibaba.com
Signed-off-by: Baolin Wang <baolin.wang@linux.alibaba.com>
Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au>
Acked-by: Mike Kravetz <mike.kravetz@oracle.com>
Reviewed-by: Muchun Song <songmuchun@bytedance.com>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Will Deacon <will@kernel.org>
Cc: Thomas Bogendoerfer <tsbogend@alpha.franken.de>
Cc: James Bottomley <James.Bottomley@HansenPartnership.com>
Cc: Helge Deller <deller@gmx.de>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Heiko Carstens <hca@linux.ibm.com>
Cc: Vasily Gorbik <gor@linux.ibm.com>
Cc: Alexander Gordeev <agordeev@linux.ibm.com>
Cc: Christian Borntraeger <borntraeger@linux.ibm.com>
Cc: Sven Schnelle <svens@linux.ibm.com>
Cc: Yoshinori Sato <ysato@users.osdn.me>
Cc: Rich Felker <dalias@libc.org>
Cc: David S. Miller <davem@davemloft.net>
Cc: Arnd Bergmann <arnd@arndb.de>
Cc: Gerald Schaefer <gerald.schaefer@linux.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

Documentation/vm: rework "Temporary Virtual Mappings" section

Extend and rework the "Temporary Virtual Mappings" section of the
highmem.rst documentation.

Despite the local kmaps were introduced by Thomas Gleixner in October
2020, documentation was still missing information about them.  These
additions rely largely on Gleixner's patches, Jonathan Corbet's LWN
articles, comments by Ira Weiny and Matthew Wilcox, and in-code comments
from ./include/linux/highmem.h.

1) Add a paragraph to document kmap_local_page().
2) Reorder the list of functions by decreasing order of preference
   of use.
3) Rework part of the kmap() entry in list.

Link: https://lkml.kernel.org/r/20220428212455.892-5-fmdefrancesco@gmail.com
Signed-off-by: Fabio M. De Francesco <fmdefrancesco@gmail.com>
Suggested-by: Ira Weiny <ira.weiny@intel.com>
Reviewed-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Reviewed-by: Ira Weiny <ira.weiny@intel.com>
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: Matthew Wilcox <willy@infradead.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Mike Rapoport <rppt@linux.ibm.com>
Cc: Peter Collingbourne <pcc@google.com>
Cc: Vlastimil Babka <vbabka@suse.cz>
Cc: Will Deacon <will@kernel.org>
Cc: Jonathan Corbet <corbet@lwn.net>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

Documentation/vm: move "Using kmap-atomic" to highmem.h

The use of kmap_atomic() is new code is being deprecated in favor of
kmap_local_page().  For this reason the "Using kmap_atomic" section in
highmem.rst is obsolete and unnecessary, but it can still help developers
if it were moved to kdocs in highmem.h.

Therefore, move the relevant parts of this section from highmem.rst and
merge them with the kdocs in highmem.h.

Link: https://lkml.kernel.org/r/20220428212455.892-4-fmdefrancesco@gmail.com
Signed-off-by: Fabio M. De Francesco <fmdefrancesco@gmail.com>
Suggested-by: Ira Weiny <ira.weiny@intel.com>
Reviewed-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Reviewed-by: Ira Weiny <ira.weiny@intel.com>
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: Matthew Wilcox <willy@infradead.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Mike Rapoport <rppt@linux.ibm.com>
Cc: Peter Collingbourne <pcc@google.com>
Cc: Vlastimil Babka <vbabka@suse.cz>
Cc: Will Deacon <will@kernel.org>
Cc: Jonathan Corbet <corbet@lwn.net>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

Documentation/vm: include kdocs from highmem*.h into highmem.rst

kernel-docs that are in include/linux/highmem.h and in
include/linux/highmem-internal.h should be included in highmem.rst.

Use kdocs directives to include the above-mentioned comments into
highmem.rst.

Link: https://lkml.kernel.org/r/20220428212455.892-3-fmdefrancesco@gmail.com
Signed-off-by: Fabio M. De Francesco <fmdefrancesco@gmail.com>
Acked-by: Mike Rapoport <rppt@linux.ibm.com>
Reviewed-by: Ira Weiny <ira.weiny@intel.com>
Suggested-by: Ira Weiny <ira.weiny@intel.com>
Reviewed-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: Matthew Wilcox <willy@infradead.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Peter Collingbourne <pcc@google.com>
Cc: Vlastimil Babka <vbabka@suse.cz>
Cc: Will Deacon <will@kernel.org>
Cc: Jonathan Corbet <corbet@lwn.net>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm/highmem: fix kernel-doc warnings in highmem*.h

Patch series "Extend and reorganize Highmem's documentation", v4.

This series has the purpose to extend and reorganize Highmem's
documentation.

This is a work in progress because some information should still be moved
from highmem.rst to highmem.h and highmem-internal.h.  Specifically I'm
talking about moving the "how to" information to the relevant headers, as
it as been suggested by Ira Weiny (Intel).

Also, this is a work in progress because some kdocs in highmem.h and
highmem-internal.h should be improved.

This patch (of 4):

`scripts/kernel-doc -v -none include/linux/highmem*` reports the following
warnings:

include/linux/highmem.h:160: warning: expecting prototype for kunmap_atomic(). Prototype was for nr_free_highpages() instead
include/linux/highmem.h:204: warning: No description found for return value of 'alloc_zeroed_user_highpage_movable'
include/linux/highmem-internal.h:256: warning: Function parameter or member '__addr' not described in 'kunmap_atomic'
include/linux/highmem-internal.h:256: warning: Excess function parameter 'addr' description in 'kunmap_atomic'

Fix these warnings by (1) moving the kernel-doc comments from highmem.h to
highmem-internal.h (which is the file were the kunmap_atomic() macro is
actually defined), (2) extending and merging it with the comment which was
already in highmem-internal.h, and (3) using correct parameter names (4)
correcting a few technical inaccuracies in comments, and (5) adding a
deprecation notice in kunmap_atomic() for consistency with kmap_atomic().

Link: https://lkml.kernel.org/r/20220428212455.892-1-fmdefrancesco@gmail.com
Link: https://lkml.kernel.org/r/20220428212455.892-2-fmdefrancesco@gmail.com
Signed-off-by: Fabio M. De Francesco <fmdefrancesco@gmail.com>
Reviewed-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Reviewed-by: Ira Weiny <ira.weiny@intel.com>
Cc: Matthew Wilcox <willy@infradead.org>
Cc: Mike Rapoport <rppt@linux.ibm.com>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Will Deacon <will@kernel.org>
Cc: Peter Collingbourne <pcc@google.com>
Cc: Vlastimil Babka <vbabka@suse.cz>
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Jonathan Corbet <corbet@lwn.net>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm/memory-failure.c: simplify num_poisoned_pages_inc/dec

Originally, do num_poisoned_pages_inc() in memory failure routine, use
num_poisoned_pages_dec() to rollback the number if filtered/ cancelled.

Suggested by Naoya, do num_poisoned_pages_inc() only in action_result(),
this make this clear and simple.

Link: https://lkml.kernel.org/r/20220509105641.491313-6-pizhenwei@bytedance.com
Signed-off-by: zhenwei pi <pizhenwei@bytedance.com>
Acked-by: Naoya Horiguchi <naoya.horiguchi@nec.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm/hwpoison: disable hwpoison filter during removing

hwpoison filter is enabled by hwpoison-inject module, after removing this
module, hwpoison filter still works.  What is worse, user can not find the
debugfs entries to know this.

Disable the hwpoison filter during removing hwpoison-inject module.

Link: https://lkml.kernel.org/r/20220509105641.491313-5-pizhenwei@bytedance.com
Signed-off-by: zhenwei pi <pizhenwei@bytedance.com>
Acked-by: Naoya Horiguchi <naoya.horiguchi@nec.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm/memory-failure.c: add hwpoison_filter for soft offline

hwpoison_filter is missing in the soft offline path, this leads an issue:
after enabling the corrupt filter, the user process still has a chance to
inject hwpoison fault by madvise(addr, len, MADV_SOFT_OFFLINE) at PFN
which is expected to reject.

Also do a minor change in comment of memory_failure().

Link: https://lkml.kernel.org/r/20220509105641.491313-4-pizhenwei@bytedance.com
Signed-off-by: zhenwei pi <pizhenwei@bytedance.com>
Acked-by: Naoya Horiguchi <naoya.horiguchi@nec.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm/memory-failure.c: simplify num_poisoned_pages_dec

Don't decrease the number of poisoned pages in page_alloc.c, let the
memory-failure.c do inc/dec poisoned pages only.

Also simplify unpoison_memory(), only decrease the number of
poisoned pages when:
 - TestClearPageHWPoison() succeed
 - put_page_back_buddy succeed

After decreasing, print necessary log.

Finally, remove clear_page_hwpoison() and unpoison_taken_off_page().

Link: https://lkml.kernel.org/r/20220509105641.491313-3-pizhenwei@bytedance.com
Signed-off-by: zhenwei pi <pizhenwei@bytedance.com>
Acked-by: Naoya Horiguchi <naoya.horiguchi@nec.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm/memory-failure.c: move clear_hwpoisoned_pages

Patch series "memory-failure: fix hwpoison_filter", v2.

As well known, the memory failure mechanism handles memory corrupted
event, and try to send SIGBUS to the user process which uses this
corrupted page.

For the virtualization case, QEMU catches SIGBUS and tries to inject MCE
into the guest, and the guest handles memory failure again.  Thus the
guest gets the minimal effect from hardware memory corruption.

The further step I'm working on:

1, try to modify code to decrease poisoned pages in a single place
   (mm/memofy-failure.c: simplify num_poisoned_pages_dec in this series).

2, try to use page_handle_poison() to handle SetPageHWPoison() and
   num_poisoned_pages_inc() together.  It would be best to call
   num_poisoned_pages_inc() in a single place too.

3, introduce memory failure notifier list in memory-failure.c: notify
   the corrupted PFN to someone who registers this list.  If I can
   complete [1] and [2] part, [3] will be quite easy(just call notifier
   list after increasing poisoned page).

4, introduce memory recover VQ for memory balloon device, and registers
   memory failure notifier list.  During the guest kernel handles memory
   failure, balloon device gets notified by memory failure notifier list,
   and tells the host to recover the corrupted PFN(GPA) by the new VQ.

5, host side remaps the corrupted page(HVA), and tells the guest side
   to unpoison the PFN(GPA).  Then the guest fixes the corrupted page(GPA)
   dynamically.

This patch (of 5):

clear_hwpoisoned_pages() clears HWPoison flag and decreases the number of
poisoned pages, this actually works as part of memory failure.

Move this function from sparse.c to memory-failure.c, finally there is no
CONFIG_MEMORY_FAILURE in sparse.c.

Link: https://lkml.kernel.org/r/20220509105641.491313-1-pizhenwei@bytedance.com
Link: https://lkml.kernel.org/r/20220509105641.491313-2-pizhenwei@bytedance.com
Signed-off-by: zhenwei pi <pizhenwei@bytedance.com>
Acked-by: Naoya Horiguchi <naoya.horiguchi@nec.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

mm/page_owner: use strscpy() instead of strlcpy()

current->comm[] is not a string (no guarantee for a zero byte in it).

strlcpy(s1, s2, l) is calling strlen(s2), potentially
causing out-of-bound access, as reported by syzbot:

detected buffer overflow in __fortify_strlen
------------[ cut here ]------------
kernel BUG at lib/string_helpers.c:980!
invalid opcode: 0000 [#1] PREEMPT SMP KASAN
CPU: 0 PID: 4087 Comm: dhcpcd-run-hooks Not tainted 5.18.0-rc3-syzkaller-01537-g20b87e7c29df #0
Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011
RIP: 0010:fortify_panic+0x18/0x1a lib/string_helpers.c:980
Code: 8c e8 c5 ba e1 fa e9 23 0f bf fa e8 0b 5d 8c f8 eb db 55 48 89 fd e8 e0 49 40 f8 48 89 ee 48 c7 c7 80 f5 26 8a e8 99 09 f1 ff <0f> 0b e8 ca 49 40 f8 48 8b 54 24 18 4c 89 f1 48 c7 c7 00 00 27 8a
RSP: 0018:ffffc900000074a8 EFLAGS: 00010286

RAX: 000000000000002c RBX: ffff88801226b728 RCX: 0000000000000000
RDX: ffff8880198e0000 RSI: ffffffff81600458 RDI: fffff52000000e87
RBP: ffffffff89da2aa0 R08: 000000000000002c R09: 0000000000000000
R10: ffffffff815fae2e R11: 0000000000000000 R12: ffff88801226b700
R13: ffff8880198e0830 R14: 0000000000000000 R15: 0000000000000000
FS:  0000000000000000(0000) GS:ffff8880b9c00000(0000) knlGS:0000000000000000
CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: 00007f5876ad6ff8 CR3: 000000001a48c000 CR4: 00000000003506f0
DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000600
Call Trace:
 <IRQ>
 __fortify_strlen include/linux/fortify-string.h:128 [inline]
 strlcpy include/linux/fortify-string.h:143 [inline]
 __set_page_owner_handle+0x2b1/0x3e0 mm/page_owner.c:171
 __set_page_owner+0x3e/0x50 mm/page_owner.c:190
 prep_new_page mm/page_alloc.c:2441 [inline]
 get_page_from_freelist+0xba2/0x3e00 mm/page_alloc.c:4182
 __alloc_pages+0x1b2/0x500 mm/page_alloc.c:5408
 alloc_pages+0x1aa/0x310 mm/mempolicy.c:2272
 alloc_slab_page mm/slub.c:1799 [inline]
 allocate_slab+0x26c/0x3c0 mm/slub.c:1944
 new_slab mm/slub.c:2004 [inline]
 ___slab_alloc+0x8df/0xf20 mm/slub.c:3005
 __slab_alloc.constprop.0+0x4d/0xa0 mm/slub.c:3092
 slab_alloc_node mm/slub.c:3183 [inline]
 slab_alloc mm/slub.c:3225 [inline]
 __kmem_cache_alloc_lru mm/slub.c:3232 [inline]
 kmem_cache_alloc+0x360/0x3b0 mm/slub.c:3242
 dst_alloc+0x146/0x1f0 net/core/dst.c:92

Link: https://lkml.kernel.org/r/20220509145949.265184-1-eric.dumazet@gmail.com
Fixes: 865ed6a32786 ("mm/page_owner: record task command name")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reported-by: syzbot <syzkaller@googlegroups.com>
Acked-by: Waiman Long <longman@redhat.com>
Acked-by: Shakeel Butt <shakeelb@google.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

kasan: clean-up kconfig options descriptions

Various readability clean-ups of KASAN Kconfig options.

No functional changes.

Link: https://lkml.kernel.org/r/c160840dd9e4b1ad5529ecfdb0bba35d9a14d826.1652203271.git.andreyknvl@google.com
Link: https://lkml.kernel.org/r/47afaecec29221347bee49f58c258ac1ced3b429.1652123204.git.andreyknvl@google.com
Signed-off-by: Andrey Konovalov <andreyknvl@google.com>
Reviewed-by: Marco Elver <elver@google.com>
Cc: Alexander Potapenko <glider@google.com>
Cc: Dmitry Vyukov <dvyukov@google.com>
Cc: Andrey Ryabinin <ryabinin.a.a@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

kasan: move boot parameters section in documentation

Move the "Boot parameters" section in KASAN documentation next to the
section that describes KASAN build options.

No content changes.

Link: https://lkml.kernel.org/r/870628e1293b4f44edf7cbcb92374ff9eb7503d7.1652203271.git.andreyknvl@google.com
Link: https://lkml.kernel.org/r/ec9c923f35e7c5312836c4624a7f317dc1ee2c1c.1652123204.git.andreyknvl@google.com
Signed-off-by: Andrey Konovalov <andreyknvl@google.com>
Reviewed-by: Marco Elver <elver@google.com>
Cc: Alexander Potapenko <glider@google.com>
Cc: Andrey Ryabinin <ryabinin.a.a@gmail.com>
Cc: Dmitry Vyukov <dvyukov@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

kasan: update documentation

Do assorted clean-ups and improvements to KASAN documentation, including:

- Describe each mode in a dedicated paragraph.
- Split out a Support section that describes in details which compilers,
  architectures and memory types each mode requires/supports.
- Capitalize the first letter in the names of each KASAN mode.

[andreyknvl@google.com: rewording, per Marco]
Link: https://lkml.kernel.org/r/896b2d914d6b50d677fd7b38f76967cc705c01ba.1652203271.git.andreyknvl@google.com
Link: https://lkml.kernel.org/r/5bd58ebebf066593ce0e1d265d60278b5f5a1874.1652123204.git.andreyknvl@google.com
Signed-off-by: Andrey Konovalov <andreyknvl@google.com>
Reviewed-by: Marco Elver <elver@google.com>
Cc: Alexander Potapenko <glider@google.com>
Cc: Andrey Ryabinin <ryabinin.a.a@gmail.com>
Cc: Dmitry Vyukov <dvyukov@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

kasan: give better names to shadow values

Rename KASAN_KMALLOC_* shadow values to KASAN_SLAB_*, as they are used for
all slab allocations, not only for kmalloc.

Also rename KASAN_FREE_PAGE to KASAN_PAGE_FREE to be consistent with
KASAN_PAGE_REDZONE and KASAN_SLAB_FREE.

Link: https://lkml.kernel.org/r/bebcaf4eafdb0cabae0401a69c0af956aa87fcaa.1652111464.git.andreyknvl@google.com
Signed-off-by: Andrey Konovalov <andreyknvl@google.com>
Reviewed-by: Alexander Potapenko <glider@google.com>
Cc: Andrey Ryabinin <ryabinin.a.a@gmail.com>
Cc: Dmitry Vyukov <dvyukov@google.com>
Cc: Marco Elver <elver@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>