kernel/git/stable/linux.git/tools/testing/selftests/mm, branch master

mm/secretmem: properly account locked pages

2026-09-03T17:56:10+00:00

secretmem accounts folios by treating memory as if it were mlock()'d and thus limited by the RLIMIT_MEMLOCK limit. However the folios are unevictable and remain so until the inode is evicted, eliminating usual mlock() semantics - mapping folios then unmapping them does not clear their unevictable state, since it depends on AS_UNEVICTABLE, not PG_mlocked. A user can therefore easily work around the RLIMIT_MEMLOCK limit - simply map then unmap and VmLck no longer counts the secretmem range. Worse, folios are not accounted in the process's RSS, meaning the OOM killer won't know to kill the process. Repeatedly mapping/unmapping (or forking) can then result in the consumption of all available system memory with unevictable folios and cause system instability. A secretmem fd can be passed between processes and over fork so a per-process limit simply does not make sense, so follow the precedent set by io_uring, perf, skbuff, iommufd and xdp by tracking the number of locked pages in user_struct->locked_vm. Since the scope tracked is actually inode lifetime, the RLIMIT_MEMLOCK applies per-user not per-process, so it doesn't make sense to bypass for users with CAP_IPC_LOCK, therefore remove this bypass. There is simply no reason to carry on marking the mapping as mlock()'d since it's misleading and the lifecycle is now correctly handled, so remove this too. Note that secretmem does not support any form of truncation (including hole punching) and the folios are unreclaimable, so the folios need only be accounted on fault and unaccounted on inode destruction. __secretmem_account_pages() is more or less a duplicate of the code that io_uring etc. use, but since this is a bug fix that needs backporting, defer any de-duplication efforts to a follow-up. test_mlock_limit() asserts mlock_future_ok() on mmap(), however this has been removed, so remove the test altogether for the fix. A new test will be sent separately for upstream. Link: https://lore.kernel.org/20260826-secretmem-accounting-v3-1-94cb04399510@kernel.org Fixes: 1507f51255c9 ("mm: introduce memfd_secret system call to create "secret" memory areas") Signed-off-by: Lorenzo Stoakes (ARM) Reported-by: Daehyeon Ko <4ncienth@gmail.com> Closes: https://lore.kernel.org/linux-mm/20260813225328.2010303-1-4ncienth@gmail.com/ Reviewed-by: Mike Rapoport (Microsoft) Acked-by: David Hildenbrand (Arm) Tested-by: Daehyeon Ko <4ncienth@gmail.com> Cc: Alexei Starovoitov Cc: David Hildenbrand Cc: David S. Miller Cc: Hagen Paul Pfeifer Cc: Jakub Kacinski Cc: James Bottomley Cc: Jesper Dangaard Brouer Cc: John Fastabend Cc: Liam R. Howlett Cc: Michal Hocko Cc: Stanislav Fomichev Cc: Suren Baghdasaryan Cc: Vlastimil Babka Cc: Signed-off-by: Andrew Morton

selftests/mm: check stat() return value in khugepaged get_finfo()

2026-08-25T01:43:27+00:00

get_finfo() calls stat() to get metadata about the target directory, but never checks the return value. On failure, stat() returns -1 and leaves path_stat unmodified, so path_stat.st_mode may contain uninitialized stack data. The code then checks S_ISDIR(path_stat.st_mode) against this potentially garbage value. This can produce a misleading "Not a directory" error when the real problem is a nonexistent or inaccessible path, or, in the worst case, the check could pass by chance on garbage data and let the function continue using an invalid path_stat for the rest of its logic. Check the return value and fail with a clear error message if stat() fails, matching the error-handling style already used for statfs() and read_file() later in the same function. Link: https://lore.kernel.org/20260819121426.49500-1-anshumantewari123@gmail.com Signed-off-by: Anshuman Reviewed-by: Andrew Morton Reviewed-by: SJ Park Reviewed-by: Sarthak Sharma Acked-by: David Hildenbrand (Arm) Cc: Lorenzo Stoakes Cc: Shuah Khan Signed-off-by: Andrew Morton

selftests/mm: fix unchecked ftruncate return value in soft-dirty test

2026-08-25T01:43:27+00:00

test_mprotect() calls ftruncate() to resize the backing file before mmap()'ing it, but never checks the return value. If ftruncate() fails, the file may remain shorter than the requested mapping size. The subsequent mmap() with MAP_SHARED can still succeed in this case, but the very next line writes directly into the mapped memory (*map = 1), which can trigger SIGBUS if the mapping extends beyond the actual file size. Check the return value and fail cleanly with ksft_exit_fail_msg() if ftruncate() fails, matching the error-handling style already used for the mmap() call immediately below it. Link: https://lore.kernel.org/20260818133206.39503-1-anshumantewari123@gmail.com Signed-off-by: Anshuman Reviewed-by: Andrew Morton Reviewed-by: Sarthak Sharma Cc: David Hildenbrand Cc: Shuah Khan Signed-off-by: Andrew Morton

selftests/mm: drop redundant open() in mprotect_tests()

2026-08-25T01:43:26+00:00

Remove duplicate open() for local pagemap_fd in mprotect_tests() that shadows the global pagemap_fd already opened in main(). The local fd is never used in the function. Link: https://lore.kernel.org/20260817080616.52946-1-hongfu.li@linux.dev Signed-off-by: Hongfu Li Reviewed-by: Lorenzo Stoakes (ARM) Reviewed-by: Muhammad Usama Anjum Reviewed-by: SJ Park Acked-by: David Hildenbrand (Arm) Reviewed-by: Anshuman Khandual Cc: Liam R. Howlett Cc: Michal Hocko Cc: Mike Rapoport Cc: Shuah Khan Cc: Suren Baghdasaryan Cc: Vlastimil Babka Signed-off-by: Andrew Morton

selftests/mm: kmemleak: drop stale min_unref_scans default from comments

2026-08-25T01:43:18+00:00

The test writes min_unref_scans explicitly for every case, so its comments describing 1 as the default are both unnecessary and, since the default is now conditional, wrong. Refer to the threshold values directly. No functional change. Link: https://lore.kernel.org/20260731-kmemleak_hardened-v2-3-7b9689ac77cb@debian.org Signed-off-by: Breno Leitao Acked-by: Catalin Marinas Cc: David Hildenbrand Cc: Jonathan Corbet Cc: Liam R. Howlett Cc: Lorenzo Stoakes Cc: Michal Hocko Cc: Mike Rapoport Cc: Shuah Khan Cc: Suren Baghdasaryan Cc: Vlastimil Babka Signed-off-by: Andrew Morton

selftests: mm: test kmemleak's N-consecutive-scan leak confirmation

2026-08-25T01:43:18+00:00

Add a functional test for the min_unref_scans kmemleak module parameter. Using samples/kmemleak's helper module it checks that min_unref_scans=1 reports an orphan on the first scan, min_unref_scans=2 reports nothing on the first scan but does on the second, and that the parameter reads back what was written. It counts only the helper module's own orphans (matched by their [kmemleak_test] backtrace, with the module kept loaded so the symbols resolve) so unrelated leaks already present on the system do not perturb the result. The test skips when run as non-root, without CONFIG_DEBUG_KMEMLEAK / CONFIG_SAMPLE_KMEMLEAK, on a kernel without the parameter, or when the helper yields no detectable orphan. Link: https://lore.kernel.org/20260713-catalin_pto-v1-4-5b93b1131089@debian.org Signed-off-by: Breno Leitao Reviewed-by: Catalin Marinas Cc: David Hildenbrand Cc: Geert Uytterhoeven Cc: Jonathan Corbet Cc: Liam R. Howlett Cc: Lorenzo Stoakes Cc: Michal Hocko Cc: Mike Rapoport Cc: Shuah Khan Cc: Suren Baghdasaryan Cc: Vlastimil Babka Signed-off-by: Andrew Morton

selftests/mm: fix read_file() return value check

2026-08-25T01:43:11+00:00

read_file() returns 0 on open/read failures and never returns negative values. Existing < 0 error checks never trigger, so read failures are silently ignored. Check for zero return to detect read_file() failures. Also fix misleading error message in get_finfo(). The error string incorrectly references read_num when reading uevent files. Link: https://lore.kernel.org/20260807013555.36525-1-hongfu.li@linux.dev Fixes: e0c13f9761df ("khugepaged: add self test") Signed-off-by: Hongfu Li Acked-by: David Hildenbrand (Arm) Cc: Baolin Wang Cc: Barry Song Cc: Dev Jain Cc: Lance Yang Cc: Liam R. Howlett Cc: Lorenzo Stoakes Cc: Michal Hocko Cc: Mike Rapoport Cc: Ryan Roberts Cc: Shuah Khan Cc: Suren Baghdasaryan Cc: Vlastimil Babka Cc: Zi Yan Signed-off-by: Andrew Morton

selftests/mm: retry migration failures for the full runtime

2026-08-25T01:43:10+00:00

move_pages() is best effort and can temporarily fail when concurrent faults race with page unmapping. A busy shared-anon workload can exhaust the current 100 retries long before the intended 20-second runtime and produce a false failure. Use the full runtime as the retry window. Since the initial page location is unknown, require it to reach both alternating NUMA targets to confirm that cross-node migration made progress despite transient contention. Link: https://lore.kernel.org/20260727095225.372655-6-usama.anjum@arm.com Signed-off-by: Muhammad Usama Anjum Cc: Baolin Wang Cc: Barry Song Cc: David Hildenbrand Cc: Dev Jain Cc: Lance Yang Cc: Liam R. Howlett Cc: Lorenzo Stoakes Cc: Miaohe Lin Cc: Michal Hocko Cc: Mike Rapoport Cc: Naoya Horiguchi Cc: Nico Pache Cc: Ryan Roberts Cc: Sarthak Sharma Cc: Shuah Khan Cc: Suren Baghdasaryan Cc: Usama Arif Cc: Vlastimil Babka Cc: Zi Yan Signed-off-by: Andrew Morton

selftests/mm: skip hard dirty page-cache test on NFS

2026-08-25T01:43:09+00:00

The hard dirty_pagecache variant uses MADV_HWPOISON to exercise recovery of a dirty file-backed page. The recovery path records -EIO in the address_space mapping, which NFS later reports when the test closes the file. This makes the test fail after the hwpoison checks have completed. Skip this variant when the test file is on NFS. Keep the hard clean-page and both soft-offline variants enabled because they use folio removal, invalidation, or migration rather than recording a delayed writeback error. The unsupported-filesystem path in clean_pagecache() also returns without closing the opened test file. Close the descriptor before skipping there and in dirty_pagecache(). Link: https://lore.kernel.org/20260727095225.372655-5-usama.anjum@arm.com Signed-off-by: Muhammad Usama Anjum Reviewed-by: Miaohe Lin Cc: Baolin Wang Cc: Barry Song Cc: David Hildenbrand Cc: Dev Jain Cc: Lance Yang Cc: Liam R. Howlett Cc: Lorenzo Stoakes Cc: Michal Hocko Cc: Mike Rapoport Cc: Naoya Horiguchi Cc: Nico Pache Cc: Ryan Roberts Cc: Sarthak Sharma Cc: Shuah Khan Cc: Suren Baghdasaryan Cc: Usama Arif Cc: Vlastimil Babka Cc: Zi Yan Signed-off-by: Andrew Morton

selftests/mm: skip guard hole-punch test if MADV_REMOVE is unsupported

2026-08-25T01:43:09+00:00

The hole_punch case verifies that guard regions survive MADV_REMOVE and that the backing range is punched out. MADV_REMOVE delegates the hole punch to the backing filesystem, which may reject the operation with EOPNOTSUPP. That result means the test cannot establish the state whose guard semantics it intends to validate. Treating the missing filesystem capability as a guard-region failure creates a false regression. Unmap the range and skip only when MADV_REMOVE fails with EOPNOTSUPP. Preserve the assertion for all other errors so failures on supported configurations remain visible. Link: https://lore.kernel.org/20260727095225.372655-3-usama.anjum@arm.com Signed-off-by: Muhammad Usama Anjum Tested-by: Sarthak Sharma Acked-by: Usama Arif Acked-by: David Hildenbrand (Arm) Cc: Baolin Wang Cc: Barry Song Cc: Dev Jain Cc: Lance Yang Cc: Liam R. Howlett Cc: Lorenzo Stoakes Cc: Miaohe Lin Cc: Michal Hocko Cc: Mike Rapoport Cc: Naoya Horiguchi Cc: Nico Pache Cc: Ryan Roberts Cc: Shuah Khan Cc: Suren Baghdasaryan Cc: Vlastimil Babka Cc: Zi Yan Signed-off-by: Andrew Morton