)]}' { "log": [ { "commit": "a370dcbed43563f0462801e889e0eceb93c7cfad", "tree": "b390913ae4b718a3de7fd0847faa513c0e103353", "parents": [ "54422f9a4c094ebdf1f3c56bdfd21e9a9ad71744" ], "author": { "name": "Jan Kara", "email": "jack@suse.cz", "time": "Thu Jul 30 15:18:17 2026" }, "committer": { "name": "Zorro Lang", "email": "zlang@kernel.org", "time": "Wed Aug 12 18:26:53 2026" }, "message": "generic/347: Fix sporadic test failures\n\ngeneric/347 was occasionally failing on ext4 in our QA due to ext4\naborting its journal when the filesystem on thinp device was overfilled.\nI have tracked the problem down to journal checkpointing failing to\nwrite a metadata block to its final location due to ENOSPC failure from\nthe thinp device. Modify the test to first preallocate blocks for the\nfiles and remount the filesystem which practically makes sure all\ninvolved metadata blocks were written and so their further modifications\nwill not fail.\n\nSigned-off-by: Jan Kara \u003cjack@suse.cz\u003e\nSigned-off-by: Zorro Lang \u003czlang@kernel.org\u003e\n" }, { "commit": "54422f9a4c094ebdf1f3c56bdfd21e9a9ad71744", "tree": "710418be0a33175325af675458ff4b1254ab7b7a", "parents": [ "7536b334c4927cd09a1d0ea3d74ae420ada504d4" ], "author": { "name": "Rochan Avlur", "email": "rochan.avlur@gmail.com", "time": "Thu Jul 23 04:11:55 2026" }, "committer": { "name": "Zorro Lang", "email": "zlang@kernel.org", "time": "Wed Aug 12 18:07:58 2026" }, "message": "exfat/001: test that rename and move preserve benign secondary entries\n\nVerify that unrecognized benign secondary directory entries (such as\nvendor extensions) survive same-directory rename, cross-directory move,\nin-place rename to a shorter name, rename overwriting an existing file,\nand operations on files with multiple benign entries.\n\nexFAT spec section 8.2 [1] requires implementations to preserve benign\nsecondary entries they do not recognize. This is a regression test for\nkernel commit 8258ef28001a (\"exfat: handle unreconized benign secondary\nentries\") which incorrectly freed benign entry clusters in the rename\nand move paths.\n\n[1]: https://learn.microsoft.com/en-us/windows/win32/fileio/exfat-specification#8-implementation-notes\n\nSigned-off-by: Rochan Avlur \u003crochan.avlur@gmail.com\u003e\nReviewed-by: Zorro Lang \u003czlang@kernel.org\u003e\nSigned-off-by: Zorro Lang \u003czlang@kernel.org\u003e\n" }, { "commit": "7536b334c4927cd09a1d0ea3d74ae420ada504d4", "tree": "2691d8bda1d6c833a06053fe70c85ff19a725eb4", "parents": [ "8954caba9fc432cf3b0f334e62a6111d795b8f63" ], "author": { "name": "Zorro Lang", "email": "zlang@kernel.org", "time": "Wed Aug 12 16:10:00 2026" }, "committer": { "name": "Zorro Lang", "email": "zlang@kernel.org", "time": "Wed Aug 12 16:10:00 2026" }, "message": "Revert \"fstests: vfstest: skip rejected tmpfile creation\"\n\nThis reverts commit 1cca2bc0d7a9462bb1b59edb108bac7c9e7c424d.\n\nThis commit can be better replaced by 1cca2bc0d7a9 (\"fstests:\nvfstest: skip rejected tmpfile creation\"), so let\u0027s revert\nthis duplicate fix now.\n" }, { "commit": "8954caba9fc432cf3b0f334e62a6111d795b8f63", "tree": "f0027257c171049fd3734630a1b58e5678e3a526", "parents": [ "9b0f378ac991b0257765f203c273f48796db0a0f" ], "author": { "name": "Christian Brauner", "email": "brauner@kernel.org", "time": "Mon Jun 15 15:33:29 2026" }, "committer": { "name": "Zorro Lang", "email": "zlang@kernel.org", "time": "Wed Aug 12 16:09:22 2026" }, "message": "src/vfs: probe O_TMPFILE support on the base mount in the idmapped tests\n\nopenat_tmpfile_supported() in the idmapped mount tests probes O_TMPFILE\non the idmapped mount (open_tree_fd) as the test\u0027s fsuid -- but that\nfsuid is not mapped by the idmapped mount under test, so it really asks\n\"can this unmapped caller create a tmpfile through this idmap?\" instead\nof \"does the filesystem support O_TMPFILE?\".\n\nOn overlayfs this is a false negative: the backing tmpfile ends up owned\nby an unmapped id, overlayfs opens it with O_NOATIME and may_open()\nreturns -EPERM, so the tmpfile sub-tests are wrongly skipped.\n\nO_TMPFILE support is a property of the filesystem, not of the idmapped\nmount or the caller\u0027s mapping. Probe info-\u003et_dir1_fd, the base mount\nthat the caller owns, exactly as the non-idmapped setgid tests already\ndo. The probe then succeeds on every filesystem and the tmpfile\nsub-tests run as the mapped fsuid.\n\nSigned-off-by: Christian Brauner (Amutable) \u003cbrauner@kernel.org\u003e\nReviewed-by: Qu Wenruo \u003cwqu@suse.com\u003e\nReviewed-by: Christoph Hellwig \u003chch@lst.de\u003e\nReviewed-by: Amir Goldstein \u003camir73il@gmail.com\u003e\nSigned-off-by: Zorro Lang \u003czlang@kernel.org\u003e\n" }, { "commit": "9b0f378ac991b0257765f203c273f48796db0a0f", "tree": "caf8e7f49e9dea4829782c6150732c427943902b", "parents": [ "be0d986ddfdff531ffc4b0f51cd65fe4bef213fe" ], "author": { "name": "Pankaj Raghav", "email": "p.raghav@samsung.com", "time": "Fri Jul 17 06:42:16 2026" }, "committer": { "name": "Zorro Lang", "email": "zlang@kernel.org", "time": "Wed Aug 12 16:09:22 2026" }, "message": "generic: add unaligned boundary test cases for WRITE_ZEROES\n\nDuring the review of WRITE_ZEROES support to XFS, Zhang Yi pointed out\nsome important semantics for the boundary blocks when WRITE_ZEROES\ncommand is used.[1]\n\nThe test cases check the boundary blocks when they are in different\nstates and WRITE_ZEROES are issued that straddle the boundary. Along with that,\ntest cases have been added when a EoF straddles an allocation unit.\n\n[1] https://lore.kernel.org/linux-xfs/0b7f1a4f-da1c-4297-8099-98d738070ab7@huaweicloud.com/\n\nSigned-off-by: Pankaj Raghav \u003cp.raghav@samsung.com\u003e\nReviewed-by: Zhang Yi \u003cyi.zhang@huawei.com\u003e\nReviewed-by: Christoph Hellwig \u003chch@lst.de\u003e\nSigned-off-by: Zorro Lang \u003czlang@kernel.org\u003e\n" }, { "commit": "be0d986ddfdff531ffc4b0f51cd65fe4bef213fe", "tree": "65e679f53388e2b5641ce197048d62e6aeeb47af", "parents": [ "b69358f4d8adb1852adf527b1f083eb0fffef13c" ], "author": { "name": "Anthony Iliopoulos", "email": "ailiop@suse.com", "time": "Mon Aug 03 11:58:19 2026" }, "committer": { "name": "Zorro Lang", "email": "zlang@kernel.org", "time": "Mon Aug 10 21:33:18 2026" }, "message": "xfs: use -w instead of --path for man program portability\n\nSome implementations of man(1) do not support the long option --path to\ndisplay the actual location of a man page, which can cause these tests\nto fail on systems using alternative man packages, such as mandoc.\n\nSwitch to using the more standard -w flag across these tests to improve\nportability while maintaining the same behavior, given that the --path\noption is just an alias to -w in the standard man-db utility.\n\nSigned-off-by: Anthony Iliopoulos \u003cailiop@suse.com\u003e\nReviewed-by: Zorro Lang \u003czlang@kernel.org\u003e\nReviewed-by: David Disseldorp \u003cddiss@suse.de\u003e\nSigned-off-by: Zorro Lang \u003czlang@kernel.org\u003e\n" }, { "commit": "b69358f4d8adb1852adf527b1f083eb0fffef13c", "tree": "01e54cdbe880f463893ca0faa668a1071f6a5440", "parents": [ "65ac068039ea058d27b5538a2ec1d6252474536c" ], "author": { "name": "Qu Wenruo", "email": "wqu@suse.com", "time": "Mon Aug 03 23:55:14 2026" }, "committer": { "name": "Zorro Lang", "email": "zlang@kernel.org", "time": "Mon Aug 10 21:21:32 2026" }, "message": "fstests: add a dio-read-into-mmap and sync race test case\n\nThere is a report that on btrfs, if the following workload are running,\nbtrfs can fail:\n\n- A dio read into a mmaped range\n Only the mmap range needs to be on btrfs.\n The dio read source makes no difference.\n\n- Sync_range on the mapped range\n\nThe btrfs errors include:\n\n- Hang during data writeback\n- Filesystem flips RO\n\nThe mmap range is dirtied but written back by the sync_range process,\nthen dio read finished and found that the folios are no longer dirty,\nso dio endio will mark those folios dirty again so that the fs can write\nthem back again.\n\nHowever for non-experimental btrfs with 4K block size and 4K page size,\nthere is a regression in v7.2 that such case is no longer handled\nproperly, due to the enablement of large folios and removal of cow\nfixup.\nAnd btrfs can never handle it for bs \u003c ps from day 1.\n\nAdd a regression test for it.\n\nReported-by: Christian Borntraeger \u003cborntraeger@linux.ibm.com\u003e\nLink: https://lore.kernel.org/linux-btrfs/20260721191152.101118-1-borntraeger@linux.ibm.com/\nSigned-off-by: Qu Wenruo \u003cwqu@suse.com\u003e\nReviewed-by: Zorro Lang \u003czlang@kernel.org\u003e\nSigned-off-by: Zorro Lang \u003czlang@kernel.org\u003e\n" }, { "commit": "65ac068039ea058d27b5538a2ec1d6252474536c", "tree": "5c4d0a13622ea22849cb23c1a9686f5c2c6732cc", "parents": [ "5249d98ebb3bcfe79e6fd87f0d469d5e0e3a4f78" ], "author": { "name": "Donald Douwsma", "email": "ddouwsma@redhat.com", "time": "Thu Jul 23 06:26:17 2026" }, "committer": { "name": "Zorro Lang", "email": "zlang@kernel.org", "time": "Mon Aug 10 20:16:01 2026" }, "message": "xfs: test xfsdump subtree restores\n\nRegression test for cumulative restores where a directory has been\nrenamed outside of the subtree being restored triggering the assert:\n\n xfsrestore: tree.c:1421: noref_elim_recurse: Assertion \u0027isrealpr\u0027 failed\n\nSigned-off-by: Donald Douwsma \u003cddouwsma@redhat.com\u003e\nReviewed-by: Zorro Lang \u003czlang@kernel.org\u003e\nReviewed-by: Christoph Hellwig \u003chch@lst.de\u003e\nSigned-off-by: Zorro Lang \u003czlang@kernel.org\u003e\n" }, { "commit": "5249d98ebb3bcfe79e6fd87f0d469d5e0e3a4f78", "tree": "e955349ab0f049677d06090910d5b4b901d71118", "parents": [ "3fec7b1d80ad2e2d58fe396cac8cc60b6aa1fc43" ], "author": { "name": "ChenXiaoSong", "email": "chenxiaosong@kylinos.cn", "time": "Thu Jul 30 02:12:38 2026" }, "committer": { "name": "Zorro Lang", "email": "zlang@kernel.org", "time": "Mon Aug 10 20:13:24 2026" }, "message": "generic: new test to check nlink returned by fstat()\n\nAdd a test to verify that fstat(2) returns the expected\nst_nlink value as hardlinks are created and removed.\n\nRegression test for kernel commit:\n9dd1964ac59d (\"smb/client: fix incorrect nlink returned by fstat()\")\n\nSuggested-by: Zorro Lang \u003czlang@kernel.org\u003e\nReviewed-by: Christoph Hellwig \u003chch@lst.de\u003e\nSigned-off-by: ChenXiaoSong \u003cchenxiaosong@kylinos.cn\u003e\nReviewed-by: Zorro Lang \u003czlang@kernel.org\u003e\nSigned-off-by: Zorro Lang \u003czlang@kernel.org\u003e\n" }, { "commit": "3fec7b1d80ad2e2d58fe396cac8cc60b6aa1fc43", "tree": "7eb26543fee79d2c44442710d90fa0c7a6e80624", "parents": [ "83760720b81e9b7db686f29f787229fe8cb59159" ], "author": { "name": "ChenXiaoSong", "email": "chenxiaosong@kylinos.cn", "time": "Thu Jul 30 02:12:37 2026" }, "committer": { "name": "Zorro Lang", "email": "zlang@kernel.org", "time": "Mon Aug 10 20:13:24 2026" }, "message": "generic/002: add test to hardlink group\n\nThe test creates and removes hard links, so add it to the hardlink\ngroup.\n\nSuggested-by: Zorro Lang \u003czlang@kernel.org\u003e\nReviewed-by: Christoph Hellwig \u003chch@lst.de\u003e\nReviewed-by: Zorro Lang \u003czlang@kernel.org\u003e\nSigned-off-by: ChenXiaoSong \u003cchenxiaosong@kylinos.cn\u003e\nSigned-off-by: Zorro Lang \u003czlang@kernel.org\u003e\n" }, { "commit": "83760720b81e9b7db686f29f787229fe8cb59159", "tree": "dbf18714184e801d2f4eb5a15411305d1993f4d4", "parents": [ "29d735f251c345a2d41c65bb0e87e115c0917b1b" ], "author": { "name": "Shin\u0027ichiro Kawasaki", "email": "shinichiro.kawasaki@wdc.com", "time": "Sat Jul 25 07:04:12 2026" }, "committer": { "name": "Zorro Lang", "email": "zlang@kernel.org", "time": "Mon Aug 10 20:13:24 2026" }, "message": "generic/563: limit loop device size\n\nThe test case generic/563 sets up a loop device on top of SCRATCH_DEV.\nCurrently, the loop device has the same size as SCRATCH_DEV. When\nSCRATCH_DEV has huge size, mkfs for the loop device can take very long\ntime. Especially, when SCRATCH_DEV is a NVMe device that takes long time\nfor discard operation, mkfs for huge loop device can take hours.\n\nTo avoid the long mkfs time, limit the size of the loop device. Choose a\nsize four times the test\u0027s I/O size, then round it up to the\nfilesystem\u0027s minimum supported size by calling _small_fs_size_mb().\n\nSigned-off-by: Shin\u0027ichiro Kawasaki \u003cshinichiro.kawasaki@wdc.com\u003e\nReviewed-by: Christoph Hellwig \u003chch@lst.de\u003e\nSigned-off-by: Zorro Lang \u003czlang@kernel.org\u003e\n" }, { "commit": "29d735f251c345a2d41c65bb0e87e115c0917b1b", "tree": "832764db832c662ea4a55cad41a173641a1642d3", "parents": [ "4a5ccb5102737aac18f61617f6a6fe4d29d3c4e2" ], "author": { "name": "Shin\u0027ichiro Kawasaki", "email": "shinichiro.kawasaki@wdc.com", "time": "Sat Jul 25 07:04:11 2026" }, "committer": { "name": "Zorro Lang", "email": "zlang@kernel.org", "time": "Mon Aug 10 20:13:24 2026" }, "message": "common/rc: add argument to _create_loop_device() to specify device size\n\nCurrently, _create_loop_device() does not specify the --sizelimit option\nto the losetup command. As a result, the loop device has the same size\nas the backing file or device. To allow controlling the loop device\nsize, add a third optional argument to _create_loop_device(). Also,\nmodify _create_loop_device_like_bdev() to pass its third argument to\n_create_loop_device().\n\nSigned-off-by: Shin\u0027ichiro Kawasaki \u003cshinichiro.kawasaki@wdc.com\u003e\nReviewed-by: Christoph Hellwig \u003chch@lst.de\u003e\nSigned-off-by: Zorro Lang \u003czlang@kernel.org\u003e\n" }, { "commit": "4a5ccb5102737aac18f61617f6a6fe4d29d3c4e2", "tree": "e67ec913fc236fb2d5919d75b438d3f08b2080f6", "parents": [ "e846df8f70211537bb87bcce184efed0a7956640" ], "author": { "name": "Lin Jiapeng", "email": "ljp1205831794@gmail.com", "time": "Mon Aug 03 02:56:13 2026" }, "committer": { "name": "Zorro Lang", "email": "zlang@kernel.org", "time": "Mon Aug 10 20:13:24 2026" }, "message": "xfs: verify CoW after exchangerange with FILE1_WRITTEN on shared extents\n\nWhen a full-file exchangerange is requested with FILE1_WRITTEN for a\nfully sparse donor file, every mapping pair is skipped, so no extents\nactually move. However, the kernel decides to swap the reflink inode\nflag based on the request geometry alone, so the target file used to\nlose its reflink flag to the donor while still owning shared extents.\nA subsequent write then took the non-reflink write path and modified\nthe shared physical blocks in place, silently corrupting the other\nfile sharing them:\n\n --- tests/xfs/842.out\n +++ results/xfs/842.out.bad\n @@ -1,3 +1,4 @@\n QA output created by 842\n +vX.reflink \u003d 0\n vX.reflink \u003d 1\n -vX.reflink \u003d 1\n +orig changed: md5 e6065c4aa2ab1603008fc18410f579d4 -\u003e 32c5189478e6bafa1cc76423f06c88f2 (write hit shared blocks in place)\n\nThis test clones a file so that both share extents, swaps the clone\nagainst a sparse donor with FILE1_WRITTEN, and then checks the defect\ntwice. First it inspects the inodes directly with xfs_db: the clone\nmust still be flagged reflink after the exchange (the donor carrying\nthe flag too is the conservative outcome of the fix and is harmless).\nThen it writes to the clone and verifies after a mount cycle that the\noriginal file\u0027s data is intact, i.e. the write was redirected through\nCoW.\n\nThe kernel fix \"xfs: fix exchange-range reflink flag clearing issue\nwith INO1_WRITTEN\" refuses the reflink flag exchange whenever\nFILE1_WRITTEN is requested, which makes this test pass.\n\nReported-by: Lin Jiapeng(TencentOS Red Team) \u003cjiapenglin@tencent.com\u003e\nLink: https://lore.kernel.org/r/amgekiNKKoAdACnc@infradead.org\nSuggested-by: Zorro Lang \u003czlang@kernel.org\u003e\nSuggested-by: Christoph Hellwig \u003chch@lst.de\u003e\nSuggested-by: \"Darrick J. Wong\" \u003cdjwong@kernel.org\u003e\nSigned-off-by: Lin Jiapeng \u003cjiapenglin@tencent.com\u003e\nReviewed-by: \"Darrick J. Wong\" \u003cdjwong@kernel.org\u003e\nReviewed-by: Zorro Lang \u003czlang@kernel.org\u003e\nSigned-off-by: Zorro Lang \u003czlang@kernel.org\u003e\n" }, { "commit": "e846df8f70211537bb87bcce184efed0a7956640", "tree": "56950f46556dc4be1ddf64d8f4d33540bc3627a2", "parents": [ "7a6f7338d28489762c0e98a73c5d61b07544dbf7" ], "author": { "name": "Ojaswin Mujoo", "email": "ojaswin@linux.ibm.com", "time": "Thu Jul 30 17:22:07 2026" }, "committer": { "name": "Zorro Lang", "email": "zlang@kernel.org", "time": "Mon Aug 10 20:13:24 2026" }, "message": "generic/645: notrun if FS doesn\u0027t have enough free inodes/blocks\n\nWhen running generic/645 with ext4 using 64k block size + bigalloc, the\ntest fails with ENOSPC because the filesystem runs out of inodes before\nthe test completes.\n\nThe test creates approximately 10,001 files, however, in this particular\nconfiguration a standard 5G FS only has around ~5100 inodes resulting in\nthe ENOSPC failure.\n\nParse vfstest output to notrun the test incase we encounter this ENOSPC\ninstead of reporting it as a failure.\n\nReported-by: Disha Goel \u003cdisgoel@linux.ibm.com\u003e\nSigned-off-by: Ojaswin Mujoo \u003cojaswin@linux.ibm.com\u003e\nTested-by: Disha Goel \u003cdisgoel@linux.ibm.com\u003e\nReviewed-by: \"Darrick J. Wong\" \u003cdjwong@kernel.org\u003e\nSigned-off-by: Zorro Lang \u003czlang@kernel.org\u003e\n" }, { "commit": "7a6f7338d28489762c0e98a73c5d61b07544dbf7", "tree": "b5dabf8a7fd01ff263370480e95c1e38191a0080", "parents": [ "140058abb86fb3b5e1c2f0d3ca9638f5294b6eef" ], "author": { "name": "Chao Yu", "email": "chao@kernel.org", "time": "Wed Jul 29 09:49:09 2026" }, "committer": { "name": "Zorro Lang", "email": "zlang@kernel.org", "time": "Mon Aug 10 20:13:24 2026" }, "message": "common/rc: support f2fs in _require_fanotify_ioerrors()\n\nf2fs has supported to report fserror, this patch adds support for f2fs\nin _require_fanotify_ioerrors(), so that we can run generic/791 on f2fs.\n\nCc: Darrick J. Wong \u003cdjwong@kernel.org\u003e\nSigned-off-by: Chao Yu \u003cchao@kernel.org\u003e\nReviewed-by: Zorro Lang \u003czlang@kernel.org\u003e\nReviewed-by: \"Darrick J. Wong\" \u003cdjwong@kernel.org\u003e\nSigned-off-by: Zorro Lang \u003czlang@kernel.org\u003e\n" }, { "commit": "140058abb86fb3b5e1c2f0d3ca9638f5294b6eef", "tree": "bf89a0e0d17cf24224fe1526247b682ff83670d1", "parents": [ "92711dbe73aa390360e0b1738689cef84875da84" ], "author": { "name": "Lukas Herbolt", "email": "lukas@herbolt.com", "time": "Wed Jul 22 05:22:58 2026" }, "committer": { "name": "Zorro Lang", "email": "zlang@kernel.org", "time": "Mon Aug 10 20:13:24 2026" }, "message": "fstests: fix build errors on kernels before v5.1\n\nThe FAN_REPORT_FID and fanotify_event_info_fid were introduced\nin v5.1. Currently there is no check if the kernel supports it\nmaking xfstests build to fail on older kernels.\n\nFixes: 29d120e41c3d (\"generic: test fsnotify filesystem error reporting\")\n\nSigned-off-by: Lukas Herbolt \u003clukas@herbolt.com\u003e\nReviewed-by: Donald Douwsma \u003cddouwsma@redhat.com\u003e\nReviewed-by: Andrey Albershteyn \u003caalbersh@kernel.org\u003e\nSigned-off-by: Zorro Lang \u003czlang@kernel.org\u003e\n" }, { "commit": "92711dbe73aa390360e0b1738689cef84875da84", "tree": "1af2f32ead202f794880c4dc9bd2244c150e3c81", "parents": [ "feb1bf12484df3772dfd44564745bc7288770ef6" ], "author": { "name": "Sun YangKai", "email": "sunyangkai@fygo.io", "time": "Wed Jul 22 07:52:47 2026" }, "committer": { "name": "Zorro Lang", "email": "zlang@kernel.org", "time": "Mon Aug 10 20:13:24 2026" }, "message": "btrfs: test POSIX ACL changes for RO btrfs property\n\nTest creation, modification and deletion of POSIX ACLs on a btrfs\nfilesystem that has the read-only property set to true.\n\nThis exercises a bug fixed by the kernel patch with subject:\n\"btrfs: check if root is readonly when setting posix acl\"\n\nSigned-off-by: Sun YangKai \u003csunyangkai@fygo.io\u003e\nReviewed-by: Filipe Manana \u003cfdmanana@suse.com\u003e\nReviewed-by: Zorro Lang \u003czlang@kernel.org\u003e\nSigned-off-by: Zorro Lang \u003czlang@kernel.org\u003e\n" }, { "commit": "feb1bf12484df3772dfd44564745bc7288770ef6", "tree": "5048f6bf478ac858fa794e509e833b36a8c573cb", "parents": [ "361b5e5d5bf9c2a8d2a93af6bbb6b26bbc727731" ], "author": { "name": "Christoph Hellwig", "email": "hch@lst.de", "time": "Fri Jul 24 12:51:06 2026" }, "committer": { "name": "Zorro Lang", "email": "zlang@kernel.org", "time": "Mon Jul 27 21:05:51 2026" }, "message": "generic/761: add to the rw group\n\nThis tests the read-write path, so add it to the proper group.\n\nSigned-off-by: Christoph Hellwig \u003chch@lst.de\u003e\nReviewed-by: Qu Wenruo \u003cwqu@suse.com\u003e\nSigned-off-by: Zorro Lang \u003czlang@kernel.org\u003e\n" }, { "commit": "361b5e5d5bf9c2a8d2a93af6bbb6b26bbc727731", "tree": "5a15da0c4bb240c6074b9086e985d63648771e84", "parents": [ "bbb3a73297abd4c5381c4ece9a9c9967c8ee4509" ], "author": { "name": "Filipe Manana", "email": "fdmanana@suse.com", "time": "Wed Jul 22 12:49:14 2026" }, "committer": { "name": "Zorro Lang", "email": "zlang@kernel.org", "time": "Mon Jul 27 20:34:20 2026" }, "message": "btrfs: add missing commit IDs to kernel fixes of some tests\n\nSome kernel fixes were already merged into Linus\u0027 tree, so add their IDs\nand in the case of btrfs/351, update the subject of the kernel fix, since\nthe original patch was dropped due to another fix.\n\nSigned-off-by: Filipe Manana \u003cfdmanana@suse.com\u003e\nReviewed-by: Johannes Thumshirn \u003cjohannes.thumshrin@wdc.com\u003e\nSigned-off-by: Zorro Lang \u003czlang@kernel.org\u003e\n" }, { "commit": "bbb3a73297abd4c5381c4ece9a9c9967c8ee4509", "tree": "b72383d58ad667e76c8f888b7a55074a67af2666", "parents": [ "56c410ad0f69da5b13c5807bc47b4876dcfa02b2" ], "author": { "name": "Jan Prusakowski", "email": "jprusakowski@google.com", "time": "Mon Jun 22 07:04:38 2026" }, "committer": { "name": "Zorro Lang", "email": "zlang@kernel.org", "time": "Mon Jul 27 20:22:35 2026" }, "message": "generic/064: allow 50 extents on F2FS after fcollapse\n\nOn F2FS, generic/064 fails with \"extents mismatched before \u003d 1 after \u003d\n50\" following multiple fcollapse (collapse range) operations.\n\nTo ensure crash consistency and checkpoint integrity, F2FS forbids\nin-place SSR (Summary Standalone Replacement) overwrites on valid\ncheckpointed blocks. When collapse range shifts blocks, F2FS allocates\nnew data pages in LFS mode (out-of-place log writes). As a result,\nsequential collapse range calls rewrite shifted blocks at new log\nlocations, intentionally leaving the file with 50 extents.\n\nAdjust the extent verification in generic/064 to expect exactly 50\nextents on F2FS, while preserving the strict 1-extent requirement for\nall other filesystems. Data integrity continues to be verified via byte\ncomparison.\n\nSigned-off-by: Jaegeuk Kim \u003cjaegeuk@kernel.org\u003e\nSigned-off-by: Jan Prusakowski \u003cjprusakowski@google.com\u003e\nReviewed-by: Chao Yu \u003cchao@kernel.org\u003e\nSigned-off-by: Zorro Lang \u003czlang@kernel.org\u003e\n" }, { "commit": "56c410ad0f69da5b13c5807bc47b4876dcfa02b2", "tree": "2791d82415ee27cf7fe6c2c3f0090cc4dfe3ac22", "parents": [ "154d755aa1df69884f2451e51b175a4d66329b4d" ], "author": { "name": "Christoph Hellwig", "email": "hch@lst.de", "time": "Fri Jul 10 05:27:35 2026" }, "committer": { "name": "Zorro Lang", "email": "zlang@kernel.org", "time": "Sun Jul 19 19:35:36 2026" }, "message": "fstests: formalize and fix disabling the RT subvolume\n\nVarious test unset SCRATCH_RTDEV to avoid dealing with metadata\nallocations for the RT subvolume in the main device or differences\nin RT subvolume behavior. Besides being rather ad-hoc, this unset\nalso fails to disable the internal RT zoned subvolume.\n\nAdd a _disable_scratch_realtime helper that not only unsets\nSCRATCH_RTDEV, but also disables the zoned allocator and with\nthat the internal RT subvolume.\n\nSigned-off-by: Christoph Hellwig \u003chch@lst.de\u003e\nReviewed-by: \"Darrick J. Wong\" \u003cdjwong@kernel.org\u003e\nSigned-off-by: Zorro Lang \u003czlang@kernel.org\u003e\n" }, { "commit": "154d755aa1df69884f2451e51b175a4d66329b4d", "tree": "c2fb3199489300a7c96cde60d03bd02dd1b2b565", "parents": [ "1e30cedac5abe8c12e47819ec41ebdcd32cb3c6d" ], "author": { "name": "Darrick J. Wong", "email": "djwong@kernel.org", "time": "Fri Jul 10 05:27:34 2026" }, "committer": { "name": "Zorro Lang", "email": "zlang@kernel.org", "time": "Sun Jul 19 19:35:20 2026" }, "message": "xfs/65[56]: don\u0027t unset SCRATCH_RTDEV here\n\nBoth of these tests claim that they have to unset SCRATCH_RTDEV to avoid\nformatting failures due to the rt bitmap consuming all the space in the\n\"small\" data device. However, the format doesn\u0027t constrain the size of\nthe data device, so the justification doesn\u0027t apply. Remove the whole\nthing.\n\nSigned-off-by: \"Darrick J. Wong\" \u003cdjwong@kernel.org\u003e\nSigned-off-by: Christoph Hellwig \u003chch@lst.de\u003e\nReviewed-by: Zorro Lang \u003czlang@kernel.org\u003e\nSigned-off-by: Zorro Lang \u003czlang@kernel.org\u003e\n" }, { "commit": "1e30cedac5abe8c12e47819ec41ebdcd32cb3c6d", "tree": "747595d2efce79fad89b819253080e5b93576e6d", "parents": [ "1cca2bc0d7a9462bb1b59edb108bac7c9e7c424d" ], "author": { "name": "Darrick J. Wong", "email": "djwong@kernel.org", "time": "Fri Jul 10 05:27:33 2026" }, "committer": { "name": "Zorro Lang", "email": "zlang@kernel.org", "time": "Sun Jul 19 19:34:50 2026" }, "message": "xfs/656: rearrange directio pread/pwrite for zoned filesystems\n\nOn zoned filesystems, a write failure shuts down the filesystem, which\nmakes this test fail with:\n\n --- /run/fstests/bin/tests/xfs/656.out\t2026-03-13 16:19:08.152939212 -0700\n +++ /run/fstests/logs/xfs/656.out.bad\t2026-06-30 18:16:43.847849604 -0700\n @@ -1,10 +1,9 @@\n QA output created by 656\n Format and mount\n pwrite: Input/output error\n -pread: Input/output error\n +stat: Input/output error\n pread: Input/output error\n fsync: Input/output error\n VICTIM pos NUM len NUM: directio_write: Input/output error\n -VICTIM pos NUM len NUM: directio_read: Input/output error\n VICTIM pos NUM len NUM: buffered_read: Input/output error\n VICTIM pos NUM len NUM: buffered_write: Input/output error\n\nReorder the directio pread and pwrite to avoid this unnecessary failure.\nNote that we cycle the mount between the direct and buffered io tests,\nwhich is why we only need to fix the directio test.\n\nSigned-off-by: \"Darrick J. Wong\" \u003cdjwong@kernel.org\u003e\nSigned-off-by: Christoph Hellwig \u003chch@lst.de\u003e\nReviewed-by: Zorro Lang \u003czlang@kernel.org\u003e\nSigned-off-by: Zorro Lang \u003czlang@kernel.org\u003e\n" }, { "commit": "1cca2bc0d7a9462bb1b59edb108bac7c9e7c424d", "tree": "37e9435b6cd0a1fbd0ece3ece328ced507714cea", "parents": [ "748bab28c5936fe8c4ce0ac53fa84fb209f7b2de" ], "author": { "name": "Qu Wenruo", "email": "wqu@suse.com", "time": "Fri Jul 17 07:07:17 2026" }, "committer": { "name": "Zorro Lang", "email": "zlang@kernel.org", "time": "Sun Jul 19 19:22:57 2026" }, "message": "fstests: vfstest: skip rejected tmpfile creation\n\nSince kernel commit 539dce114465 (\"fs: refuse O_TMPFILE creation with an\nunmapped fsuid or fsgid\"), tmpfile creation can be rejected with\nEOVERFLOW, so also skip that error number to avoid false alerts like the\nfollowing:\n\n FSTYP -- ext4\n PLATFORM -- Linux/x86_64 btrfs-vm 7.2.0-rc3-custom+ #411 SMP PREEMPT_DYNAMIC Fri Jul 17 16:24:34 ACST 2026\n MKFS_OPTIONS -- -F /dev/mapper/test-scratch1\n MOUNT_OPTIONS -- -o acl,user_xattr /dev/mapper/test-scratch1 /mnt/scratch\n\n generic/633 1s ... - output mismatch (see /home/adam/xfstests/results//generic/633.out.bad)\n --- tests/generic/633.out\t2022-05-11 11:25:30.816666664 +0930\n +++ /home/adam/xfstests/results//generic/633.out.bad\t2026-07-17 16:25:45.033967541 +0930\n @@ -1,2 +1,4 @@\n QA output created by 633\n Silence is golden\n +utils.c: 948: openat_tmpfile_supported - Value too large for defined data type - failure: create\n +utils.c: 948: openat_tmpfile_supported - Value too large for defined data type - failure: create\n ...\n (Run \u0027diff -u /home/adam/xfstests/tests/generic/633.out /home/adam/xfstests/results//generic/633.out.bad\u0027 to see the entire diff)\n Ran: generic/633\n Failures: generic/633\n Failed 1 of 1 tests\n\nSigned-off-by: Qu Wenruo \u003cwqu@suse.com\u003e\nReviewed-by: Zorro Lang \u003czlang@kernel.org\u003e\nReviewed-by: Boris Burkov \u003cboris@bur.io\u003e\nSigned-off-by: Zorro Lang \u003czlang@kernel.org\u003e\n" }, { "commit": "748bab28c5936fe8c4ce0ac53fa84fb209f7b2de", "tree": "87c1c4291b0ab13574e27054d4755cb10c29959b", "parents": [ "1cd5dcb0545d27f46afe3c816817d06cc240ec87" ], "author": { "name": "Shin\u0027ichiro Kawasaki", "email": "shinichiro.kawasaki@wdc.com", "time": "Thu Jul 16 01:37:47 2026" }, "committer": { "name": "Zorro Lang", "email": "zlang@kernel.org", "time": "Sun Jul 19 19:22:57 2026" }, "message": "xfs/333: require external scratch device\n\nThe test case xfs/333 uses the external scratch device SCRATCH_RTDEV.\nWhen SCRATCH_RTDEV is not specified, the test case fails at\n\"_require_block_device $SCRATCH_RTDEV\". To confirm the requirement, the\ntest case calls the helper function _require_realtime(). However, it\ndoes not check SCRATCH_RTDEV when the zone type of SCRATCH_DEV is\n\u0027none\u0027.\n\nTo avoid the failure, ensure the SCRATCH_RTDEV check by calling\n_require_external_realtime_device() instead of _require_realtime(). With\nthis, the test case is skipped when SCRATCH_RTDEV is not specified\nregardless of the zone type of SCRATCH_DEV.\n\nSigned-off-by: Shin\u0027ichiro Kawasaki \u003cshinichiro.kawasaki@wdc.com\u003e\nReviewed-by: Christoph Hellwig \u003chch@lst.de\u003e\nReviewed-by: Yao Sang \u003csangyao@kylinos.cn\u003e\nSigned-off-by: Zorro Lang \u003czlang@kernel.org\u003e\n" }, { "commit": "1cd5dcb0545d27f46afe3c816817d06cc240ec87", "tree": "23101b07dc7b468bb31c6a62d36f41490af0f699", "parents": [ "b3ee9a117b00250c2ec6233d7a47aaf847be474a" ], "author": { "name": "Andrey Albershteyn", "email": "aalbersh@kernel.org", "time": "Thu Jul 09 11:18:47 2026" }, "committer": { "name": "Zorro Lang", "email": "zlang@kernel.org", "time": "Sun Jul 19 19:22:57 2026" }, "message": "generic/270: kill fsstress by exact name\n\nRunning ./check with fsstress in the arguments (for example -x\nfsstress_scrub) causes pkill to kill parent bash process executing the\ntests. The test suite execution just stops with:\n\nxfstests-start[4099970]: Terminated _run_seq 2\u003e\u00261 | tee $tmp.out\nxfstests-start[4099970]: Ran: generic/269 generic/270\nxfstests-start[4102535]: Interrupted!\nxfstests-start[4102538]: Interrupted!\n\nFix by using exact name, we can not use _pkill_fsstress here as this\ntest copies fsstress binary under custom name.\n\nSigned-off-by: Andrey Albershteyn \u003caalbersh@kernel.org\u003e\nReviewed-by: Christoph Hellwig \u003chch@lst.de\u003e\nSigned-off-by: Zorro Lang \u003czlang@kernel.org\u003e\n" }, { "commit": "b3ee9a117b00250c2ec6233d7a47aaf847be474a", "tree": "b3dfa4741381705e32d01e8cbd1b8a903a7bc767", "parents": [ "8c2fa9b73b15a35ce6d8822b686fc547091ee79c" ], "author": { "name": "Qu Wenruo", "email": "wqu@suse.com", "time": "Tue Jul 07 11:40:30 2026" }, "committer": { "name": "Zorro Lang", "email": "zlang@kernel.org", "time": "Sun Jul 19 19:22:57 2026" }, "message": "fstests: btrfs: new test case for leaking BTRFS_FS_STATE_REMOUNTING flag\n\nThis is a regression test for the patch \"btrfs: fix leaking\nBTRFS_FS_STATE_REMOUNTING flag\", where conflicting mount options can\ncause btrfs leave BTRFS_FS_STATE_REMOUNTING set.\n\nSuch set BTRFS_FS_STATE_REMOUNTING will interrupt several features, and\nthe test case will use \"btrfs qgroup rescan\" to catch the\nBTRFS_FS_STATE_REMOUNTING flag.\n\nSigned-off-by: Qu Wenruo \u003cwqu@suse.com\u003e\nReviewed-by: Anand Jain \u003casj@kernel.org\u003e\nSigned-off-by: Zorro Lang \u003czlang@kernel.org\u003e\n" }, { "commit": "8c2fa9b73b15a35ce6d8822b686fc547091ee79c", "tree": "5f73ad31560955324802745718ab1a90bbaae29b", "parents": [ "eb4a22ad79b9f5505f6a9da62b47ebfba1feb6e5" ], "author": { "name": "Qu Wenruo", "email": "wqu@suse.com", "time": "Wed Jul 15 22:37:16 2026" }, "committer": { "name": "Zorro Lang", "email": "zlang@kernel.org", "time": "Sun Jul 19 19:22:57 2026" }, "message": "fstests: generic: add a basic cachestat test case\n\nThe test case is inspired by LTP, where there is a regression on 64K\npage size systems with btrfs, that after a fsync, cachestat() still\nreport dirty pages.\n\nThe test case itself is pretty simple, fill the file with a buffered write that is\n1/2/4/8/16 page sized, call cachestat() to make sure the cached/dirtied\nnumber match the page number.\n\nThen do a fsync(), and make sure the dirty page number reduced to 0\nmeanwhile cached is still the same.\n\nLink: https://bugzilla.suse.com/show_bug.cgi?id\u003d1270397\nSigned-off-by: Qu Wenruo \u003cwqu@suse.com\u003e\nReviewed-by: Anand Jain \u003casj@kernel.org\u003e\nReviewed-by: Filipe Manana \u003cfdmanana@suse.com\u003e\nReviewed-by: Brian Foster \u003cbfoster@redhat.com\u003e\nSigned-off-by: Zorro Lang \u003czlang@kernel.org\u003e\n" }, { "commit": "eb4a22ad79b9f5505f6a9da62b47ebfba1feb6e5", "tree": "7d981ecfe5a478627cafe7db7c56c90cba8397ae", "parents": [ "b2e97cf30043334a72efaca4f860bb4768b95722" ], "author": { "name": "Avinesh Kumar", "email": "avinesh.kumar@suse.com", "time": "Mon Jul 06 19:13:26 2026" }, "committer": { "name": "Zorro Lang", "email": "zlang@kernel.org", "time": "Sun Jul 19 19:22:57 2026" }, "message": "generic/062: filter setfattr --restore symlink-safety warning\n\nattr 2.6.0 (CVE-2026-54371 fix)[0] makes \"setfattr --restore\" warn when\nused without -P and the dump contains a symlink, which generic/062 does.\nThe warning leaks into the compared output via the setfattr() wrapper\nand fails the test. Adding -P isn\u0027t portable (older attr doesn\u0027t have it),\nso filter the warning in the wrapper instead.\n\n[0] https://cgit.git.savannah.nongnu.org/cgit/attr.git/commit/?id\u003d3fb06b9ba314d37035d0877e6de313de754f1ac8\n\nSigned-off-by: Avinesh Kumar \u003cavinesh.kumar@suse.com\u003e\nReviewed-by: Zorro Lang \u003czlang@kernel.org\u003e\nSigned-off-by: Zorro Lang \u003czlang@kernel.org\u003e\n" }, { "commit": "b2e97cf30043334a72efaca4f860bb4768b95722", "tree": "02296b70aed21cb42b4c6d6276ce1fb835a1886e", "parents": [ "921e3466e0f7e168b4ce4376f5eb2dabc8cd7868" ], "author": { "name": "An Long", "email": "lan@suse.com", "time": "Wed Jul 01 06:14:07 2026" }, "committer": { "name": "Zorro Lang", "email": "zlang@kernel.org", "time": "Fri Jul 17 08:40:41 2026" }, "message": "common/rc: fix mount options quoting in _mount\n\nIn commit 078f320 (\"treewide: convert all $MOUNT_PROG to _mount\"),\ndirect calls to $MOUNT_PROG were converted to call the helper function\n_mount.\nHowever, the _mount helper was implemented using `$*` instead of `\"$@\"`.\nThis unquoted argument expansion triggers word splitting on any arguments\ncontaining spaces (such as the overlay mount options containing spaces in\ntests `overlay/083` and `overlay/086`), causing the mount to fail with\n\"mount: bad usage\".\nFix this regression by using `\"$@\"` in the _mount helper, which correctly\npreserves argument boundaries exactly as they were provided by the caller.\n\nSigned-off-by: An Long \u003clan@suse.com\u003e\nReviewed-by: \"Darrick J. Wong\" \u003cdjwong@kernel.org\u003e\nSigned-off-by: Zorro Lang \u003czlang@kernel.org\u003e\n" }, { "commit": "921e3466e0f7e168b4ce4376f5eb2dabc8cd7868", "tree": "4519f8894febf4da60471646e5532d4df10a08fc", "parents": [ "2c13d0a4a5546ece9046984816b5281e5da3e11e" ], "author": { "name": "Christoph Hellwig", "email": "hch@lst.de", "time": "Fri Jul 10 05:22:08 2026" }, "committer": { "name": "Zorro Lang", "email": "zlang@kernel.org", "time": "Fri Jul 17 08:39:32 2026" }, "message": "common: log mkfs output in _test_streams\n\nLog the mkfs output to allow debugging the mkfs paramters,\nespecially when mkfs and thus the test fails.\n\nSigned-off-by: Christoph Hellwig \u003chch@lst.de\u003e\nReviewed-by: \"Darrick J. Wong\" \u003cdjwong@kernel.org\u003e\nSigned-off-by: Zorro Lang \u003czlang@kernel.org\u003e\n" }, { "commit": "2c13d0a4a5546ece9046984816b5281e5da3e11e", "tree": "b534ef58fe0f62b2992b46fd232cc13a5fdec847", "parents": [ "bd4a8258575a85d26b39e9aeebf8186e5f7f6b91" ], "author": { "name": "Christoph Hellwig", "email": "hch@lst.de", "time": "Thu Jun 18 13:45:28 2026" }, "committer": { "name": "Zorro Lang", "email": "zlang@kernel.org", "time": "Fri Jul 17 08:37:55 2026" }, "message": "xfs: remove the v2log group\n\nv2log have been the default for more than a decade. Merge all xfs log\ntests into the log group instead of splitting them up (with a lot of\noverlap).\n\nSigned-off-by: Christoph Hellwig \u003chch@lst.de\u003e\nReviewed-by: Andrey Albershteyn \u003caalbersh@kernel.org\u003e\nReviewed-by: Zorro Lang \u003czlang@kernel.org\u003e\nSigned-off-by: Zorro Lang \u003czlang@kernel.org\u003e\n" }, { "commit": "bd4a8258575a85d26b39e9aeebf8186e5f7f6b91", "tree": "ea11b800c1258cfc17812a98eb4e12f341c5363a", "parents": [ "60fa62e5e80870040ea89d7dd77df1fcc0bb7092" ], "author": { "name": "Darrick J. Wong", "email": "djwong@kernel.org", "time": "Thu Jun 25 16:51:17 2026" }, "committer": { "name": "Zorro Lang", "email": "zlang@kernel.org", "time": "Mon Jun 29 18:18:56 2026" }, "message": "xfs/21[67]: fix mkfs log concurrency detection\n\nBoth of these tests use _scratch_mkfs_xfs_supports_concurrency to detect\nif they should be adding -lconcurrency\u003d0 to the mkfs options that are\nused to format a loop device.\n\nHowever, this doesn\u0027t work in practice because of three factors: 1) the\nscratch device could have an external log set up; 2) the loop device\ndoes /not/ have an external log device; and 3) -lconcurrency isn\u0027t\ncompatible with -llogdev.\n\nFix this by creating a more general mkfs option detection helper and use\nthat once we have the loop device set up.\n\nCc: fstests@vger.kernel.org # v2026.05.17\nFixes: ffc8bad17e5b2f (\"xfs/21{6,7} Use default -l concurrency\u003d0 on mkfs.xfs that supports it\")\nSigned-off-by: \"Darrick J. Wong\" \u003cdjwong@kernel.org\u003e\nReviewed-by: Christoph Hellwig \u003chch@lst.de\u003e\nSigned-off-by: Zorro Lang \u003czlang@kernel.org\u003e\n" }, { "commit": "60fa62e5e80870040ea89d7dd77df1fcc0bb7092", "tree": "2031b92d990301340d16abf3c7dec1b18f75aa34", "parents": [ "86b976d0deae50f790c119073115d4590bbbae90" ], "author": { "name": "Darrick J. Wong", "email": "djwong@kernel.org", "time": "Thu Jun 25 16:51:02 2026" }, "committer": { "name": "Zorro Lang", "email": "zlang@kernel.org", "time": "Mon Jun 29 18:18:56 2026" }, "message": "common/xfs: fix _scratch_mkfs_xfs_supports_concurrency\n\ngeneric/347 regresses when external logs are in the configuration:\n\n --- a/generic/347.out\t\t2025-07-15 14:45:15.044714644 -0700\n +++ b/generic/347.out.bad\t2026-05-18 23:17:01.687750781 -0700\n @@ -1,2 +1,36 @@\n QA output created by 347\n -\u003d\u003d\u003d completed\n +Cannot specify both -l logdev and -l concurrency\n +Usage: mkfs.xfs\n\nSince this actually tests running mkfs.xfs with $SCRATCH_OPTIONS, the\nhelper ought to have detected that you can\u0027t give both. Unfortunately,\nthere\u0027s a weird switch that turns -l into -d even though they\u0027re not\nquite the same option. Delete that, and fix the indentation (tabs, not\nspaces) in the helper.\n\nCc: fstests@vger.kernel.org # v2026.05.17\nFixes: 7f162f5bcf50fc (\"common/xfs: helper function to check if -l/-d/-r concurrecy flags.\")\nSigned-off-by: \"Darrick J. Wong\" \u003cdjwong@kernel.org\u003e\nReviewed-by: Christoph Hellwig \u003chch@lst.de\u003e\nSigned-off-by: Zorro Lang \u003czlang@kernel.org\u003e\n" }, { "commit": "86b976d0deae50f790c119073115d4590bbbae90", "tree": "7c06decdd67f1620bbedcff36713da11bd0de5a6", "parents": [ "325e2e4daa406d98c4d8841e9cdf9d346db02f81" ], "author": { "name": "Christoph Hellwig", "email": "hch@lst.de", "time": "Fri Jun 19 05:09:28 2026" }, "committer": { "name": "Zorro Lang", "email": "zlang@kernel.org", "time": "Mon Jun 29 18:17:43 2026" }, "message": "xfs/078: disable all concurrency scaling\n\nDisable concurrency scaling to avoid mismatching output on systems with\na large CPU count.\n\nReported-by: Shin\u0027ichiro Kawasaki \u003cshinichiro.kawasaki@wdc.com\u003e\nSigned-off-by: Christoph Hellwig \u003chch@lst.de\u003e\nTested-by: Shin\u0027ichiro Kawasaki \u003cshinichiro.kawasaki@wdc.com\u003e\nReviewed-by: Eric Sandeen \u003csandeen@redhat.com\u003e\nSigned-off-by: Zorro Lang \u003czlang@kernel.org\u003e\n" }, { "commit": "325e2e4daa406d98c4d8841e9cdf9d346db02f81", "tree": "bf366ff6c8484e6b833ef1bb4580093d5ebb127c", "parents": [ "eb4cd02a3da5cd7c3c2cd320586d435180a871bb" ], "author": { "name": "Filipe Manana", "email": "fdmanana@suse.com", "time": "Thu Jun 25 19:20:26 2026" }, "committer": { "name": "Zorro Lang", "email": "zlang@kernel.org", "time": "Mon Jun 29 18:17:43 2026" }, "message": "btrfs: skip tests that fail when autodefrag is enabled\n\nA couple tests, btrfs/004 and btrfs/308, fail if we pass \"-o autodefrag\"\nin MOUNT_OPTIONS, due to changes in extent layouts. So make sure the\ntests are skipped if autodefrag is enabled.\n\nSigned-off-by: Filipe Manana \u003cfdmanana@suse.com\u003e\nReviewed-by: Anand Jain \u003casj@kernel.org\u003e\nSigned-off-by: Zorro Lang \u003czlang@kernel.org\u003e\n" }, { "commit": "eb4cd02a3da5cd7c3c2cd320586d435180a871bb", "tree": "2a2d9cc5dd1d4fd3b5fc2f75ebe4525a0e532b99", "parents": [ "ce0dff4b555b3d0fd6521a62fee1b8dfdec5b656" ], "author": { "name": "Yao Sang", "email": "sangyao@kylinos.cn", "time": "Tue Jun 23 06:26:16 2026" }, "committer": { "name": "Zorro Lang", "email": "zlang@kernel.org", "time": "Mon Jun 29 18:17:43 2026" }, "message": "common: skip data write EIO survival tests on fatal configs\n\ngeneric/441 and xfs/656 inject data write errors after mounting the scratch\nfilesystem on dm-error. These tests do not verify whether a filesystem may\nshut down after media errors. They require the filesystem to survive long\nenough to report the error through the path under test: generic/441 checks\nfsync(2) writeback error reporting, and xfs/656 checks healthmon per-I/O\nerror events.\n\nThat expectation is not valid for filesystem configurations where a data\nwrite error is fatal. Zoned XFS is one such case because a failed data write\ncan leave the in-core zone allocation state untrustworthy, so shutting down\nthe mount is expected behavior.\n\nAdd a common helper for tests that require data write EIO to be reported\nwithout shutting down the mounted filesystem, and use it after mounting the\ndm-error device in both tests. The helper currently detects zoned XFS by\nchecking the XFS per-mount zoned/max_open_zones sysfs attribute for the\nmounted device, so the check follows the filesystem instance rather than the\nunderlying block queue alone.\n\nFatal-on-data-write-error configurations are reported as notrun instead of\ntest failures.\n\nSigned-off-by: Yao Sang \u003csangyao@kylinos.cn\u003e\nReviewed-by: Christoph Hellwig \u003chch@lst.de\u003e\nSigned-off-by: Zorro Lang \u003czlang@kernel.org\u003e\n" }, { "commit": "ce0dff4b555b3d0fd6521a62fee1b8dfdec5b656", "tree": "388805e075a687e92960451d310fb6562cb4778e", "parents": [ "78b2382e6ccfa64f29fad91ad39e22a3cb411863" ], "author": { "name": "Chao Yu", "email": "chao@kernel.org", "time": "Mon Jun 22 02:05:16 2026" }, "committer": { "name": "Zorro Lang", "email": "zlang@kernel.org", "time": "Mon Jun 29 18:17:04 2026" }, "message": "common/f2fs: introduce _require_f2fs_io_command\n\nUpdate existing f2fs testcases to utilize the new\n_require_f2fs_io_command() helper for verifying required f2fs_io\nsubcommands instead of basic executable checks or open coding.\n\nSuggested-by: Zorro Lang \u003czlang@kernel.org\u003e\nSigned-off-by: Chao Yu \u003cchao@kernel.org\u003e\nReviewed-by: Zorro Lang \u003czlang@kernel.org\u003e\nSigned-off-by: Zorro Lang \u003czlang@kernel.org\u003e\n" }, { "commit": "78b2382e6ccfa64f29fad91ad39e22a3cb411863", "tree": "f9173d35f0b06c4e6f9fa4729a0849e330fe5a9f", "parents": [ "b1a8621f6b481824951a61f1eb94de10fab1cba3" ], "author": { "name": "Qu Wenruo", "email": "wqu@suse.com", "time": "Thu Jun 18 21:25:53 2026" }, "committer": { "name": "Zorro Lang", "email": "zlang@kernel.org", "time": "Mon Jun 29 18:17:01 2026" }, "message": "fstests: add a test case for btrfs get_subvol_info ioctl\n\nThere is a regression caused by upstream kernel commit 538e5bdbc899\n(\"btrfs: add 32-bit compat ioctl for BTRFS_IOC_GET_SUBVOL_INFO\"), where\nthe unprivileged get_subvol_info ioctl is returning incorrect results.\n\nUnfortunately that bug is not covered by fstests, and only exposed by\nbtrfs-progs libbtrfsutil python tests, which is not a short run.\n\nSo add a new test case for such btrfs ioctl:\n\n- Introduce a helper program, btrfs_ioctl\n Which will call the dedicated ioctl, print all of its structure\n members.\n\n For now only get_subvol_info is supported, but it can be easily\n appended for new ones.\n\n- Mkfs and create a subvolume on btrfs\n\n- Use \"_btrfs_get_subvolid()\" to grab the subvolid\n That function is utilizing \"btrfs subvolume list\" to grab the info,\n which is not utilizing the unprivileged ioctl.\n\n- Use \"btrfs_ioctl get_subvol_info\" to grab the subvolid\n\n- Compare the two subvolids\n\nThis should be enough to detect the regression.\n\nSigned-off-by: Qu Wenruo \u003cwqu@suse.com\u003e\nReviewed-by: Anand Jain \u003casj@kernel.org\u003e\nSigned-off-by: Zorro Lang \u003czlang@kernel.org\u003e\n" }, { "commit": "b1a8621f6b481824951a61f1eb94de10fab1cba3", "tree": "e7ff0d63d6ce0a5709b83cd43408873e4ca4b1b6", "parents": [ "d2286dbce7f6e5fff3a4b7827a5cee19580e75fc" ], "author": { "name": "German Maglione", "email": "gmaglione@redhat.com", "time": "Thu Jun 18 10:11:54 2026" }, "committer": { "name": "Zorro Lang", "email": "zlang@kernel.org", "time": "Mon Jun 29 18:17:01 2026" }, "message": "generic/633: add _require_acls for tcore_acls() ACL dependency\n\ngeneric/633 runs --test-core which includes tcore_acls(). Without\n_require_acls, the test fails instead of being skipped on filesystems\nthat lack ACL support.\n\nSigned-off-by: German Maglione \u003cgmaglione@redhat.com\u003e\nReviewed-by: Zorro Lang \u003czlang@kernel.org\u003e\nReviewed-by: Christian Brauner (Amutable) \u003cbrauner@kernel.org\u003e\nSigned-off-by: Zorro Lang \u003czlang@kernel.org\u003e\n" }, { "commit": "d2286dbce7f6e5fff3a4b7827a5cee19580e75fc", "tree": "6085122ac188da005812f53dab2d115f3ef52fe7", "parents": [ "acb6d4cb84205a8e3f19ca470cfcf7bf6d93a509" ], "author": { "name": "German Maglione", "email": "gmaglione@redhat.com", "time": "Thu Jun 18 10:11:53 2026" }, "committer": { "name": "Zorro Lang", "email": "zlang@kernel.org", "time": "Mon Jun 29 18:17:01 2026" }, "message": "vfstest: move sys_umount2() to cleanup path in tcore_acls()\n\nsys_umount2() in tcore_acls() only runs on the success path. If setfacl\nfails (e.g. no ACL support), the mount on DIR2 is left behind and\ntest_cleanup() can\u0027t remove T_DIR1. This causes every subsequent test\nto fail with EEXIST on mkdirat(T_DIR1).\n\nMove sys_umount2() into the out: block so it always runs.\n\nSigned-off-by: German Maglione \u003cgmaglione@redhat.com\u003e\nReviewed-by: Zorro Lang \u003czlang@kernel.org\u003e\nReviewed-by: Christian Brauner (Amutable) \u003cbrauner@kernel.org\u003e\nSigned-off-by: Zorro Lang \u003czlang@kernel.org\u003e\n" }, { "commit": "acb6d4cb84205a8e3f19ca470cfcf7bf6d93a509", "tree": "e561df373d23656e490ec462d1698bd5551b72d4", "parents": [ "65caee6192ce5006bedbd40cc2889e0961c93b46" ], "author": { "name": "Disha Goel", "email": "disgoel@linux.ibm.com", "time": "Mon Jun 08 10:23:28 2026" }, "committer": { "name": "Zorro Lang", "email": "zlang@kernel.org", "time": "Wed Jun 17 19:34:26 2026" }, "message": "common/defrag: skip defrag tests on DAX-enabled filesystems\n\nOnline defragmentation is not supported on ext4 DAX-enabled filesystems.\nThe ext4 defrag ioctl (EXT4_IOC_MOVE_EXT) returns EOPNOTSUPP when used\non DAX files.\n\nAdd an ext4-specific check in _require_defrag() to skip tests when DAX\nis enabled, avoiding false failures on ext4/301-304, ext4/308, and\ngeneric/018.\n\nXFS defrag works with DAX, so this check is ext4-specific.\n\nSuggested-by: Darrick J. Wong \u003cdjwong@kernel.org\u003e\nSigned-off-by: Disha Goel \u003cdisgoel@linux.ibm.com\u003e\nReviewed-by: Ojaswin Mujoo \u003cojaswin@linux.ibm.com\u003e\nReviewed-by: Zorro Lang \u003czlang@kernel.org\u003e\nSigned-off-by: Zorro Lang \u003czlang@kernel.org\u003e\n" }, { "commit": "65caee6192ce5006bedbd40cc2889e0961c93b46", "tree": "ca3ccef30fd3b7678b5ec04f1828a356bbd3170d", "parents": [ "26eb955e4243495c0df0bc2f157cda32edccf18b" ], "author": { "name": "Joanne Chang", "email": "joannechien@google.com", "time": "Tue Jun 02 07:19:11 2026" }, "committer": { "name": "Zorro Lang", "email": "zlang@kernel.org", "time": "Wed Jun 17 19:28:41 2026" }, "message": "generic/63{4,5}: Modify criteria for passing test\n\nThe two tests pass when a timestamp is not preserved after the remount,\nbecause \"cmp -s\" does not print to the output file, and status is set to\n0 regardless of the return value of cmp. So, use \"diff\" instead to\ncorrectly fail the test and flag the error case in the output file.\n\nAlso, as the tests are focused on timestamps, don\u0027t print file contents\nso the tests can still pass if only the file contents differ.\n\nSigned-off-by: Joanne Chang \u003cjoannechien@google.com\u003e\nReviewed-by: Zorro Lang \u003czlang@kernel.org\u003e\nSigned-off-by: Zorro Lang \u003czlang@kernel.org\u003e\n" }, { "commit": "26eb955e4243495c0df0bc2f157cda32edccf18b", "tree": "9cfc0df258b72aa69f498e4aacde42378a0bf926", "parents": [ "88b140d74b3e01484db5c080165d5d7240d8ce0c" ], "author": { "name": "Christoph Hellwig", "email": "hch@lst.de", "time": "Wed Jun 17 06:04:17 2026" }, "committer": { "name": "Zorro Lang", "email": "zlang@kernel.org", "time": "Wed Jun 17 09:34:48 2026" }, "message": "generic/366: check minimum dio size correctly\n\nmin_dio_alignment requires the file system to be checked to be mounted,\nso move the call into the main loop. Drop the pointless blocksize\nargument to _require_odirect, which is only tested on the test device.\n\nSigned-off-by: Christoph Hellwig \u003chch@lst.de\u003e\nReviewed-by: Zorro Lang \u003czlang@kernel.org\u003e\nSigned-off-by: Zorro Lang \u003czlang@kernel.org\u003e\n" }, { "commit": "88b140d74b3e01484db5c080165d5d7240d8ce0c", "tree": "ba2935649e785c01b33161a2533f0cfc287e11c2", "parents": [ "4f599bae27e4aa4a4eb743b51cf017662278c36e" ], "author": { "name": "Chao Yu", "email": "chao@kernel.org", "time": "Mon Jun 15 09:05:58 2026" }, "committer": { "name": "Zorro Lang", "email": "zlang@kernel.org", "time": "Tue Jun 16 14:18:42 2026" }, "message": "f2fs/009: use common _scratch_shutdown helper\n\nf2fs has supported ioctl(_IOR (\u0027X\u0027, 125, uint32_t)), so that\n\"f2fs_io shutdown\" can be completely replaced w/ _scratch_shutdown(),\nlet\u0027s clean up in f2fs/009.\n\nSuggested-by: Zorro Lang \u003czlang@kernel.org\u003e\nSigned-off-by: Chao Yu \u003cchao@kernel.org\u003e\nReviewed-by: Zorro Lang \u003czlang@kernel.org\u003e\nSigned-off-by: Zorro Lang \u003czlang@kernel.org\u003e\n" }, { "commit": "4f599bae27e4aa4a4eb743b51cf017662278c36e", "tree": "bf59ca027c2b713194161b5fd762db27196c632a", "parents": [ "6c5d1992fb3bcb6adf5c38eb31924d76794ae9dc" ], "author": { "name": "Chao Yu", "email": "chao@kernel.org", "time": "Tue Jun 16 07:35:26 2026" }, "committer": { "name": "Zorro Lang", "email": "zlang@kernel.org", "time": "Tue Jun 16 14:09:33 2026" }, "message": "f2fs/025: test to do sanity check section type correctly in f2fs GC\n\nWithout commit 520760b9f915 (\"f2fs: optimize representative type determination\nin GC\"), f2fs GC will report inconsistent segment type in large section issue,\nand then it will force to shutdown filesystem.\n\n[ 768.190903] F2FS-fs (loop51): Inconsistent segment (3) type [1, 0] in SIT and SSA\n\nThe reason is f2fs kernel will assume all segment type inside large section is\nthe same, during GC it loads type from one segment and compare it to other\nsegments\u0027 type, however due to recovery flow, the chosen segment may has zero\nvalid blocks w/ different segment type, since the segment is invalid(free) one,\nit will never be migrated, so that we should not treat such state as abnormal\ncondition.\n\nThis testcase is created to simulate above condition to see whether f2fs kernel\nmodule can handle it correctly\n\nSigned-off-by: Chao Yu \u003cchao@kernel.org\u003e\nReviewed-by: Zorro Lang \u003czlang@kernel.org\u003e\nSigned-off-by: Zorro Lang \u003czlang@kernel.org\u003e\n" }, { "commit": "6c5d1992fb3bcb6adf5c38eb31924d76794ae9dc", "tree": "6bb11e61a88bd3228a18797c4b10fd9549b123ae", "parents": [ "ec6d49da5d27690f8146a5fbbea9214abaa8a4cc" ], "author": { "name": "Shin\u0027ichiro Kawasaki", "email": "shinichiro.kawasaki@wdc.com", "time": "Mon Jun 15 12:12:57 2026" }, "committer": { "name": "Zorro Lang", "email": "zlang@kernel.org", "time": "Mon Jun 15 20:26:55 2026" }, "message": "xfs/013: fix ENOSPC handling\n\nCommit 000813899afb (\"fstests: scale some tests for high CPU count\nsanity\") replaced \"touch\" with \"echo -n \u003e\" to speed up file creation by\nusing a shell builtin redirection instead of forking and execing an\nexternal binary. However, this broke the ENOSPC filtering the test\nrelies on.\n\nWhen the scratch filesystem fills up, the shell redirection \"\u003e $dir/$i\"\nfails instead of the \"touch\" command. The shell applies redirections\nleft-to-right, so \"\u003e $dir/$i\" is attempted before \"2\u003e\u00261\". Because\nopening the file fails immediately with ENOSPC, the command never runs\nand the shell writes the error message \"No space left on device\" to the\nstderr instead of the pipe. Then the error message is not passed to\n\"filter_enospc\" and leaks into the test output, causing the test case\nfailure. The failure is recreated by running the test case with rather\nsmall size of block devices, such as 128MiB null_blk:\n\nxfs/013 - output mismatch (see /home/test/xfstests/results//xfs/013.out.bad)\n --- tests/xfs/013.out\t2026-05-28 16:52:16.881159200 +0900\n +++ /home/test/xfstests/results//xfs/013.out.bad\t2026-06-11 17:20:36.269003467 +0900\n @@ -5,3 +5,12305 @@\n naming \u003dVERN bsize\u003dXXX\n log \u003dLDEV bsize\u003dXXX blocks\u003dXXX\n realtime \u003dRDEV extsz\u003dXXX blocks\u003dXXX, rtextents\u003dXXX\n +/home/test/xfstests/tests/xfs/013: line 45: /var/scratch/dir2/19013: No space left on device\n +/home/test/xfstests/tests/xfs/013: line 45: /var/scratch/dir2/6452: No space left on device\n +/home/test/xfstests/tests/xfs/013: line 45: /var/scratch/dir2/14964: No space left on device\n +/home/test/xfstests/tests/xfs/013: line 45: /var/scratch/dir2/10051: No space left on device\n ...\n (Run \u0027diff -u /home/test/xfstests/tests/xfs/013.out /home/test/xfstests/results//xfs/013.out.bad\u0027 to see the entire diff)\n\nTo fix the ENOSPC handling, wrap the redirection in a brace group so\nthat \"2\u003e\u00261\" is applied to the group before the inner redirection is\nattempted. The redirection failure is then written to the group\u0027s\nstderr, which is piped into \"filter_enospc\". This keeps the faster\nbuiltin file creation than the \"touch\" command as the trigger commit\nintended.\n\nFixes: 000813899afb (\"fstests: scale some tests for high CPU count sanity\")\nSigned-off-by: Shin\u0027ichiro Kawasaki \u003cshinichiro.kawasaki@wdc.com\u003e\nReviewed-by: Zorro Lang \u003czlang@kernel.org\u003e\nReviewed-by: Christoph Hellwig \u003chch@lst.de\u003e\nSigned-off-by: Zorro Lang \u003czlang@kernel.org\u003e\n" }, { "commit": "ec6d49da5d27690f8146a5fbbea9214abaa8a4cc", "tree": "79141b3d12724f96778cbf455ec6ce89459c8932", "parents": [ "973512e9c24062257ec456e929232a8d6c0cb4e7" ], "author": { "name": "Yao Sang", "email": "sangyao@kylinos.cn", "time": "Fri Jun 12 07:42:31 2026" }, "committer": { "name": "Zorro Lang", "email": "zlang@kernel.org", "time": "Sun Jun 14 16:26:11 2026" }, "message": "xfs: test zoned realtime writeback EIO shutdown\n\nCreate a zoned realtime filesystem with a single user data open zone,\ndirty one zone worth of data, and inject a writeback error on the\nrealtime device.\n\nAfter disabling fault injection, fsync a second write. The filesystem\nmust already be shut down so that the second fsync fails quickly instead\nof waiting for zoned allocation progress.\n\nSigned-off-by: Yao Sang \u003csangyao@kylinos.cn\u003e\nReviewed-by: Zorro Lang \u003czlang@kernel.org\u003e\nSigned-off-by: Zorro Lang \u003czlang@kernel.org\u003e\n" }, { "commit": "973512e9c24062257ec456e929232a8d6c0cb4e7", "tree": "3c84264e6685c3158388ea1da2d11422ea164a54", "parents": [ "ace8d9c9dee5d6bf762cbf794f2d8d0e07fec3f0" ], "author": { "name": "Filipe Manana", "email": "fdmanana@suse.com", "time": "Thu May 28 10:27:07 2026" }, "committer": { "name": "Zorro Lang", "email": "zlang@kernel.org", "time": "Sun Jun 14 15:27:55 2026" }, "message": "generic: add a test case for writes with prealloc extents beyond i_size\n\nTest writing into a file range containing prealloc extents beyond the current\ni_size, with an unmount and mount after fallocate and the write, to verify\nthat the file data, size and extent layout were not lost.\n\nThis used to fail on btrfs when not using the no-holes feature (which is\na default since btrfs-progs 5.15) before this recent kernel fix:\n\n 080ecbd05432 (\"btrfs: mark file extent range dirty after converting prealloc extents\")\n\nSo in order to reproduce the failure when using an unpatched kernel and\na btrfs-progs \u003e\u003d 5.15, one must run the test with:\n\n MKFS_OPTIONS\u003d\"-O ^no-holes\"\n\nSigned-off-by: Filipe Manana \u003cfdmanana@suse.com\u003e\nReviewed-by: Qu Wenruo \u003cwqu@suse.com\u003e\nSigned-off-by: Zorro Lang \u003czlang@kernel.org\u003e\n" }, { "commit": "ace8d9c9dee5d6bf762cbf794f2d8d0e07fec3f0", "tree": "3c51a994a6787d58e9b586c1d95e172a6b00cba3", "parents": [ "120ff814a4c98cd303e838ac6604b5816ab1600b" ], "author": { "name": "Qu Wenruo", "email": "wqu@suse.com", "time": "Sun Jun 07 23:38:01 2026" }, "committer": { "name": "Zorro Lang", "email": "zlang@kernel.org", "time": "Sun Jun 14 09:13:23 2026" }, "message": "fstests: btrfs/138: add requirement for datacsum\n\n[FALSE ALERT]\nIf one has specified \"-o nodatasum\" as mount options, the test case\nbtrfs/138 will fail with the following error:\n\n MOUNT_OPTIONS -- -o nodatasum /dev/mapper/test-scratch1 /mnt/scratch\n\n btrfs/138 34s ... - output mismatch (see /home/adam/xfstests/results//btrfs/138.out.bad)\n --- tests/btrfs/138.out\t2022-05-11 11:25:30.743333331 +0930\n +++ /home/adam/xfstests/results//btrfs/138.out.bad\t2026-06-08 09:04:30.627143086 +0930\n @@ -1,2 +1,5 @@\n QA output created by 138\n +ERROR: failed to set compression for /mnt/scratch/zlib: Invalid argument\n +ERROR: failed to set compression for /mnt/scratch/lzo: Invalid argument\n +ERROR: failed to set compression for /mnt/scratch/zstd: Invalid argument\n Silence is golden\n ...\n\n[CAUSE]\nBtrfs compression requires both datacow and datasum, thus nodatasum\nmount option will reject compression completely, causing the above\nfailures.\n\n[FIX]\nAdd _require_btrfs_no_nodatasum for the test case, so that for nodatasum\nmount option the test case can be properly skipped other than causing\nfalse failure.\n\nSigned-off-by: Qu Wenruo \u003cwqu@suse.com\u003e\nReviewed-by: Anand Jain \u003casj@kernel.org\u003e\nSigned-off-by: Zorro Lang \u003czlang@kernel.org\u003e\n" }, { "commit": "120ff814a4c98cd303e838ac6604b5816ab1600b", "tree": "143ffd00be697e5023f7068b488fe4a5aa7ce1b6", "parents": [ "1104c45b26d01d40a187ed4c9f0517bebdc12b27" ], "author": { "name": "Christoph Hellwig", "email": "hch@lst.de", "time": "Tue Jun 09 08:00:44 2026" }, "committer": { "name": "Zorro Lang", "email": "zlang@kernel.org", "time": "Thu Jun 11 18:03:32 2026" }, "message": "xfs: test that a grown RT section can be filled\n\nThis reproduces issues with zoned growfs.\n\nSigned-off-by: Christoph Hellwig \u003chch@lst.de\u003e\nReviewed-by: Hans Holmberg \u003chans.holmberg@wdc.com\u003e\nReviewed-by: \"Darrick J. Wong\" \u003cdjwong@kernel.org\u003e\nSigned-off-by: Zorro Lang \u003czlang@kernel.org\u003e\n" }, { "commit": "1104c45b26d01d40a187ed4c9f0517bebdc12b27", "tree": "f7939d7da7e7a7a0d3328759ec251a75b5f53a2e", "parents": [ "5d68212a7373f675fcff8b82082a0a64b04ae985" ], "author": { "name": "Christoph Hellwig", "email": "hch@lst.de", "time": "Tue Jun 09 08:00:43 2026" }, "committer": { "name": "Zorro Lang", "email": "zlang@kernel.org", "time": "Thu Jun 11 18:00:49 2026" }, "message": "common: add a _require_external_realtime_device helper\n\nFactor out the check for an RT device on an external section.\n\nSigned-off-by: Christoph Hellwig \u003chch@lst.de\u003e\nReviewed-by: Hans Holmberg \u003chans.holmberg@wdc.com\u003e\nReviewed-by: \"Darrick J. Wong\" \u003cdjwong@kernel.org\u003e\nSigned-off-by: Zorro Lang \u003czlang@kernel.org\u003e\n" }, { "commit": "5d68212a7373f675fcff8b82082a0a64b04ae985", "tree": "46a584495430ab905ded59adaa05464fa86588a5", "parents": [ "80d250af40caf7d0db95caaf7243deefb16ef9a6" ], "author": { "name": "Shin\u0027ichiro Kawasaki", "email": "shinichiro.kawasaki@wdc.com", "time": "Wed Jun 03 05:27:06 2026" }, "committer": { "name": "Zorro Lang", "email": "zlang@kernel.org", "time": "Sat Jun 06 19:59:35 2026" }, "message": "common/filestreams: check zoned device before mkfs with small dsize\n\nThe test case xfs/170 performs file stream tests and must be skipped for\nzoned scratch devices since the filestream feature is not supported. The\ntest calls _require_xfs_scratch_non_zoned() to check if the scratch\ndevice is zoned.\n\nThis check is done after _try_scratch_mkfs_xfs() with a small dsize\nparameter. However, mkfs.xfs creates an inode for each zone, and when a\nzoned device has many zones, the small dsize causes metadata area\nshortage and mkfs fails. This unexpected failure causes the test to fail\neven though it should be skipped for zoned devices.\n\nMove the zoned device check to occur before the constrained mkfs. First\nperform a usual mkfs without the small dsize parameter to ensure the\ndevice can be properly checked for zoned mode, then verify it\u0027s non-\nzoned before proceeding with the small dsize mkfs.\n\nSigned-off-by: Shin\u0027ichiro Kawasaki \u003cshinichiro.kawasaki@wdc.com\u003e\nReviewed-by: Christoph Hellwig \u003chch@lst.de\u003e\nSigned-off-by: Zorro Lang \u003czlang@kernel.org\u003e\n" }, { "commit": "80d250af40caf7d0db95caaf7243deefb16ef9a6", "tree": "f32db26cc39751604ae159d656c337642210e8e8", "parents": [ "26c2b701724afe41ec7e925549c883abf570b4f9" ], "author": { "name": "Qu Wenruo", "email": "wqu@suse.com", "time": "Tue Jun 02 10:39:03 2026" }, "committer": { "name": "Zorro Lang", "email": "zlang@kernel.org", "time": "Sat Jun 06 19:58:09 2026" }, "message": "fstests: btrfs: add a new test case for fstrim crash\n\nThere is a bug report that btrfs/242 can crash very randomly.\n\nIt turns out the crash is caused by a re-appearing missing device, which\nscrewed up several checks in fstrim.\n\nThis new test case is a more reliable version, dedicated for such crash,\nwhich will:\n\n- Create a raid1 btrfs\n- Mount with a single device in degraded mode\n- Let the missing device to re-appear\n- Run fstrim\n\nWhich is reliably crashing in my environment without the submitted fix.\nAnd with the submitted fix, the test no longer crashes.\n\nReported-by: Su Yue \u003cglass.su@suse.com\u003e\nLink: https://lore.kernel.org/linux-btrfs/wlwir19t.fsf@damenly.org/\nReviewed-by: Filipe Manana \u003cfdmanana@suse.com\u003e\nSigned-off-by: Qu Wenruo \u003cwqu@suse.com\u003e\nSigned-off-by: Zorro Lang \u003czlang@kernel.org\u003e\n" }, { "commit": "26c2b701724afe41ec7e925549c883abf570b4f9", "tree": "471ae13303f61e17d553145e145c7bf17cba7b0e", "parents": [ "d65330dd7de46410342e7b7000d7bbceb903de0d" ], "author": { "name": "Boris Burkov", "email": "boris@bur.io", "time": "Wed May 13 00:43:34 2026" }, "committer": { "name": "Zorro Lang", "email": "zlang@kernel.org", "time": "Fri May 29 09:57:16 2026" }, "message": "btrfs: test qgroup deletion races with squota writes\n\nWhen using simple quotas, an extent\u0027s EXTENT_OWNER_REF can long outlive\nthe subvolume that created it, since the extent can pick up additional\nreferences that keep it alive after the owning subvolume is deleted.\n\nSeveral lifecycle bugs around the owning qgroup arise from this:\n\n 1. Freeing an extent whose owner qgroup is gone: must not cause an\n underflow nor a transaction abort.\n\n 2. Creating an extent in the same transaction that the owner subvolume\n is deleted: the qgroup may already be gone by the time the squota\n delta is recorded, leaving behind an EXTENT_OWNER_REF that points\n at a qgroup that never accumulated a usage delta. Manually\n re-creating that qgroup and then freeing the extent would underflow\n it.\n\n 3. Destroying a live subvolume\u0027s qgroup while delayed refs for newly\n allocated tree blocks have not yet flushed: rfer/excl read as 0,\n so the destroy looked safe, but the alloc delta was silently lost.\n\n 4. Destroying the qgroup of a live subvolume whose data predates\n \u0027btrfs quota enable --simple\u0027: pre-existing extents are not\n accounted (generation \u003c qgroup_enable_gen), so rfer/excl stay 0\n permanently and a usage-only check happily destroys a qgroup\n belonging to a still-mounted, still-writeable subvolume.\n\nAdd four cases covering these scenarios. On a fixed kernel all four\n\u0027qgroup destroy\u0027 (and the re-create in case 2) operations are refused\nbecause the subvol check rejects the request; on an unfixed kernel\nthey would succeed and leave the filesystem with broken accounting or\ntrigger a transaction abort.\n\nSigned-off-by: Boris Burkov \u003cboris@bur.io\u003e\nReviewed-by: Anand Jain \u003casj@kernel.org\u003e\nSigned-off-by: Zorro Lang \u003czlang@kernel.org\u003e\n" }, { "commit": "d65330dd7de46410342e7b7000d7bbceb903de0d", "tree": "8ba9dc1da79d40f27b3f9c25b6e404a4edc22554", "parents": [ "239cc1d598663b849b5da29c5f9d0382ec8eb442" ], "author": { "name": "Boris Burkov", "email": "boris@bur.io", "time": "Wed May 13 00:43:33 2026" }, "committer": { "name": "Zorro Lang", "email": "zlang@kernel.org", "time": "Fri May 29 09:52:51 2026" }, "message": "btrfs: inline enable_quota helper in test 301\n\nThe enable_quota helper in btrfs/301 takes a mode argument (\u0027n\u0027, \u0027s\u0027,\nor anything else) to switch between \"no quota\", \"simple quota\", and\n\"full quota\" enable. Only the simple-quota path is ever exercised by\nthis test, so the dispatch is dead weight and the indirection just\nobscures the actual btrfs command being run.\n\nInline the two call sites to call \u0027btrfs quota enable --simple\u0027\ndirectly and drop the helper.\n\nSigned-off-by: Boris Burkov \u003cboris@bur.io\u003e\nReviewed-by: Anand Jain \u003casj@kernel.org\u003e\nSigned-off-by: Zorro Lang \u003czlang@kernel.org\u003e\n" }, { "commit": "239cc1d598663b849b5da29c5f9d0382ec8eb442", "tree": "58453dfec87bb493cd1a7c51994f005754401bd4", "parents": [ "1418baaef1faa80d8f2203a2e8cbb36b86afb322" ], "author": { "name": "Qu Wenruo", "email": "wqu@suse.com", "time": "Thu May 28 11:16:59 2026" }, "committer": { "name": "Zorro Lang", "email": "zlang@kernel.org", "time": "Fri May 29 09:13:01 2026" }, "message": "fstests: cleanup old files to respect mount options needing new inodes\n\n[BUG]\nTest case generic/362 and generic/365 will fail with nodatasum, but\nthat\u0027s only when TEST_DEV is newly formatted.\n\n FSTYP -- btrfs\n PLATFORM -- Linux/x86_64 btrfs-vm 7.1.0-rc4-custom+ #381 SMP PREEMPT_DYNAMIC Tue May 26 10:47:14 ACST 2026\n MKFS_OPTIONS -- -O bgt -K /dev/mapper/test-scratch1\n MOUNT_OPTIONS -- -o nodatasum /dev/mapper/test-scratch1 /mnt/scratch\n\n generic/362 0s ... - output mismatch (see /home/adam/xfstests/results//generic/362.out.bad)\n --- tests/generic/362.out\t2024-08-24 15:31:37.200000000 +0930\n +++ /home/adam/xfstests/results//generic/362.out.bad\t2026-05-28 16:18:33.866141979 +0930\n @@ -1,2 +1,3 @@\n QA output created by 362\n +First write failed: Input/output error\n Silence is golden\n ...\n (Run \u0027diff -u /home/adam/xfstests/tests/generic/362.out /home/adam/xfstests/results//generic/362.out.bad\u0027 to see the entire diff)\n\n generic/364 11s ... - output mismatch (see /home/adam/xfstests/results//generic/364.out.bad)\n --- tests/generic/364.out\t2024-09-30 09:09:51.216666681 +0930\n +++ /home/adam/xfstests/results//generic/364.out.bad\t2026-05-28 16:18:34.318532257 +0930\n @@ -1,2 +1,3 @@\n QA output created by 364\n +Fsync failed: Input/output error\n Silence is golden\n ...\n (Run \u0027diff -u /home/adam/xfstests/tests/generic/364.out /home/adam/xfstests/results//generic/364.out.bad\u0027 to see the entire diff)\n\nBut if one has formated TEST_DEV, run test with default mount option,\nthen change the mount option to \"nodatasum\", the test will not fail\nanymore:\n\n FSTYP -- btrfs\n PLATFORM -- Linux/x86_64 btrfs-vm 7.1.0-rc4-custom+ #381 SMP PREEMPT_DYNAMIC Tue May 26 10:47:14 ACST 2026\n MKFS_OPTIONS -- -O bgt -K /dev/mapper/test-scratch1\n MOUNT_OPTIONS -- /dev/mapper/test-scratch1 /mnt/scratch\n\n generic/362 0s ... 1s\n generic/364 11s ... 10s\n Ran: generic/362 generic/364\n Passed all 2 tests\n\n FSTYP -- btrfs\n PLATFORM -- Linux/x86_64 btrfs-vm 7.1.0-rc4-custom+ #381 SMP PREEMPT_DYNAMIC Tue May 26 10:47:14 ACST 2026\n MKFS_OPTIONS -- -O bgt -K /dev/mapper/test-scratch1\n MOUNT_OPTIONS -- -o nodatasum /dev/mapper/test-scratch1 /mnt/scratch\n\n generic/362 1s ... 0s\n generic/364 10s ... 11s\n Ran: generic/362 generic/364\n Passed all 2 tests\n\n[CAUSE]\nBtrfs\u0027 nodatasum mount option only affects new files, but the test cases\nare using TEST_DEV, and never delete the target files\n\"$TEST_DIR/dio-append-buf-fault\" for generic/362 and\n\"$TEST_DIR/dio-write-fsync-same-fd\" for generic/364.\n\nSo if the files are created with default mount options, then all later\n\"nodatasum\" mount options will not affect those files, thus hide the test\nfailure.\n\n[FIX]\nFor all test cases utilizing \"$here/src/dio*\", add a _cleanup()\nfunction, to remove target files, so that for mount options that only\naffect new inodes, the mount options will be respected, and expose\nfailures.\n\nNow generic/36[24] will properly fail for \"nodatasum\" mount option on\nbtrfs.\n\nReviewed-by: Filipe Manana \u003cfdmanana@suse.com\u003e\nSigned-off-by: Qu Wenruo \u003cwqu@suse.com\u003e\nReviewed-by: Zorro Lang \u003czlang@kernel.org\u003e\nReviewed-by: Anand Jain \u003casj@kernel.org\u003e\nSigned-off-by: Zorro Lang \u003czlang@kernel.org\u003e\n" }, { "commit": "1418baaef1faa80d8f2203a2e8cbb36b86afb322", "tree": "b66ff68cfaee14f0a34012f9b258a89b8492f05f", "parents": [ "c1dcfe62b3f95d323224e9b8087ab0e6df69c993" ], "author": { "name": "Qu Wenruo", "email": "wqu@suse.com", "time": "Sun May 24 05:08:48 2026" }, "committer": { "name": "Zorro Lang", "email": "zlang@kernel.org", "time": "Fri May 29 08:54:30 2026" }, "message": "fstests: btrfs: add a new test case for single data RAID56\n\nAlthough single-data-RAID56, aka 2 disks raid5 or 3 disks raid6, is\nnot recommended for btrfs, btrfs still supports it as there are cases\nlike degraded writes can still lead to such cases.\n\nBut we do not have any proper tests for such corner cases.\n\nFurthermore, RAID56 lib is going to drop the support of\nsingle-data-RAID56, and btrfs will have its own handling of such cases,\nthus coverage on single-data-RAID56 is required.\n\nThis new test case will have the following workload to verify the\nbehavior:\n\n- Create a fs with single-data-RAID56\n\n- Populate the fs with fsx and fsstress\n Fsx and fsstress will utilize read and write paths of RAID56.\n\n- Mount each device in degraded mode and run read-only scrub on them\n Such scrub will utilize the following path of RAID56:\n * Regular data read\n * Reconstruction read\n * P/Q stripe scrub\n\n- Re-scan all devices and add a new disk to the array\n- Balance all data and scrub again\n This should utilized non-single-data-RAID56 handling.\n\nSigned-off-by: Qu Wenruo \u003cwqu@suse.com\u003e\nReviewed-by: Zorro Lang \u003czlang@kernel.org\u003e\nReviewed-by: Christoph Hellwig \u003chch@lst.de\u003e\nSigned-off-by: Zorro Lang \u003czlang@kernel.org\u003e\n" }, { "commit": "c1dcfe62b3f95d323224e9b8087ab0e6df69c993", "tree": "0649a3ad5bc9bd5c0a10773b703e2a00041628b4", "parents": [ "c5ddd1789ae14e1acd512f6cc2af61778ac53aa6" ], "author": { "name": "Andrey Albershteyn", "email": "aalbersh@kernel.org", "time": "Thu May 21 08:26:46 2026" }, "committer": { "name": "Zorro Lang", "email": "zlang@kernel.org", "time": "Fri May 29 07:22:08 2026" }, "message": "generic/791: don\u0027t run if kernel misses FANOTIFY\n\nIf FANOTIFY\u003dno in the kernel, then fs-monitor will fail to start. Fix\nfs-monitor error codes to detect and skip the test on \"Function not\nimplemented\" code.\n\nSigned-off-by: Andrey Albershteyn \u003caalbersh@kernel.org\u003e\nReviewed-by: \"Darrick J. Wong\" \u003cdjwong@kernel.org\u003e\nReviewed-by: Christoph Hellwig \u003chch@lst.de\u003e\nSigned-off-by: Zorro Lang \u003czlang@kernel.org\u003e\n" }, { "commit": "c5ddd1789ae14e1acd512f6cc2af61778ac53aa6", "tree": "51805ff380b5450b44b0217a7fd649738a44635a", "parents": [ "55d8071bc4a86895fadb5d7767ef1f016df7d010" ], "author": { "name": "Joanne Chang", "email": "joannechien@google.com", "time": "Wed May 13 03:20:09 2026" }, "committer": { "name": "Zorro Lang", "email": "zlang@kernel.org", "time": "Wed May 20 10:42:32 2026" }, "message": "f2fs/021: Fix mount syntax for disabling jquota\n\nF2FS has removed the trailing \"\u003d\" in the mount options for disabling\nuser journaled quotas. Update f2fs/021 to try the new syntax first and\nfall back to the legacy syntax if the initial mount attempt fails.\n\nSigned-off-by: Joanne Chang \u003cjoannechien@google.com\u003e\nReviewed-by: Chao Yu \u003cchao@kernel.org\u003e\nSigned-off-by: Zorro Lang \u003czlang@kernel.org\u003e\n" }, { "commit": "55d8071bc4a86895fadb5d7767ef1f016df7d010", "tree": "417f13e6b3929cbb835f64e84f38baebcab0a46f", "parents": [ "d5fb5bbbd22d836cbb256480191632b485671dbe" ], "author": { "name": "Joanne Chang", "email": "joannechien@google.com", "time": "Wed May 13 03:20:08 2026" }, "committer": { "name": "Zorro Lang", "email": "zlang@kernel.org", "time": "Wed May 20 10:42:32 2026" }, "message": "f2fs/015: Fix mount syntax for disabling jquota\n\nF2FS has removed the trailing \"\u003d\" in the mount options for disabling\nuser, group, and project journaled quotas. To maintain compatibility,\nupdate the test cases in f2fs/015 to try the new syntax first, and fall\nback to the legacy syntax if the initial mount attempt fails.\n\nSigned-off-by: Joanne Chang \u003cjoannechien@google.com\u003e\nReviewed-by: Chao Yu \u003cchao@kernel.org\u003e\nSigned-off-by: Zorro Lang \u003czlang@kernel.org\u003e\n" }, { "commit": "d5fb5bbbd22d836cbb256480191632b485671dbe", "tree": "a3d8186c4484537cd5b1d4360e3a28e7b7ff5b9b", "parents": [ "d66e76fc5ab6a6f1550b09aa6904d7974da3aad8" ], "author": { "name": "Viacheslav Dubeyko", "email": "slava@dubeyko.com", "time": "Thu May 14 20:00:05 2026" }, "committer": { "name": "Zorro Lang", "email": "zlang@kernel.org", "time": "Wed May 20 10:38:38 2026" }, "message": "common/rc: exclude HFS from atime testing\n\nHFS hasn\u0027t any field in on-disk layout that can keep\nthe file/folder access times (atime):\n\n/* The catalog record for a file */\nstruct hfs_cat_file {\n s8 type; /* The type of entry */\n u8 reserved;\n u8 Flags; /* Flags such as read-only */\n s8 Typ; /* file version number \u003d 0 */\n struct hfs_finfo UsrWds; /* data used by the Finder */\n __be32 FlNum; /* The CNID */\n __be16 StBlk; /* obsolete */\n __be32 LgLen; /* The logical EOF of the data fork*/\n __be32 PyLen; /* The physical EOF of the data fork */\n __be16 RStBlk; /* obsolete */\n __be32 RLgLen; /* The logical EOF of the rsrc fork */\n __be32 RPyLen; /* The physical EOF of the rsrc fork */\n __be32 CrDat; /* The creation date */\n __be32 MdDat; /* The modified date */\n __be32 BkDat; /* The last backup date */\n struct hfs_fxinfo FndrInfo; /* more data for the Finder */\n __be16 ClpSize; /* number of bytes to allocate\n when extending files */\n hfs_extent_rec ExtRec; /* first extent record\n for the data fork */\n hfs_extent_rec RExtRec; /* first extent record\n for the resource fork */\n u32 Resrv; /* reserved by Apple */\n} __packed;\n\nThis patch disable the generic/003 and atime related test-cases for\nHFS file system.\n\nSigned-off-by: Viacheslav Dubeyko \u003cslava@dubeyko.com\u003e\ncc: John Paul Adrian Glaubitz \u003cglaubitz@physik.fu-berlin.de\u003e\ncc: Yangtao Li \u003cfrank.li@vivo.com\u003e\ncc: linux-fsdevel@vger.kernel.org\ncc: fstests@vger.kernel.org\nReviewed-by: Zorro Lang \u003czlang@kernel.org\u003e\nSigned-off-by: Zorro Lang \u003czlang@kernel.org\u003e\n" }, { "commit": "d66e76fc5ab6a6f1550b09aa6904d7974da3aad8", "tree": "301eefd778e0450e3076a042f16337d63ce541f3", "parents": [ "32b4dfffbc086df51c7a16ee59086da8e5024bc6" ], "author": { "name": "Jan Prusakowski", "email": "jprusakowski@google.com", "time": "Fri May 15 08:26:08 2026" }, "committer": { "name": "Zorro Lang", "email": "zlang@kernel.org", "time": "Wed May 20 10:26:37 2026" }, "message": "generic/563: allow read tolerance for f2fs in scenario 3\n\nF2FS is a log-structured file system that allocates new blocks even for\noverwrites. To perform allocation, it may need to read metadata blocks\n(such as the Node Address Table (NAT) or Segment Info Table (SIT)) if\nthey are not already in memory.\n\nIn scenario 3 (read -\u003e read/write), unlike scenario 2, there is no prior\nwrite in the same mount session to load these metadata blocks into memory.\nTherefore, the write operation in the second cgroup triggers metadata\nreads, which are charged to that cgroup.\n\nRelax the read tolerance for f2fs in scenario 3 to accommodate these\nexpected metadata reads, matching the tolerance already used in scenario 2.\n\nSigned-off-by: Jan Prusakowski \u003cjprusakowski@google.com\u003e\nReviewed-by: Chao Yu \u003cchao@kernel.org\u003e\nReviewed-by: Zorro Lang \u003czlang@kernel.org\u003e\nSigned-off-by: Zorro Lang \u003czlang@kernel.org\u003e\n" }, { "commit": "32b4dfffbc086df51c7a16ee59086da8e5024bc6", "tree": "045eaf6dc7d09fb98f15cac9d23fa7fe2b9921cc", "parents": [ "ce4e96a89a2b876497006b3fcbb388ee41f7aa8a" ], "author": { "name": "Filipe Manana", "email": "fdmanana@suse.com", "time": "Tue May 19 17:14:10 2026" }, "committer": { "name": "Zorro Lang", "email": "zlang@kernel.org", "time": "Wed May 20 09:42:06 2026" }, "message": "generic: add a test case for fallocate i_size extension\n\nTest an i_size expanding fallocate into a range beyond current i_size that\nalready has a preallocated extent, with an unmount and mount after each\nfallocate to verify that no metadata (i_size) and extents were lost.\n\nThis used to fail on btrfs when not using the no-holes feature (which is\na default since btrfs-progs 5.15) before this recent kernel fix:\n\n c562ba61fc5e (\"btrfs: fix incorrect i_size after remount caused by KEEP_SIZE prealloc gap\")\n\nSo in order to reproduce the failure when using an unpatched kernel and\na btrfs-progs \u003e\u003d 5.15, one must run the test with:\n\n MKFS_OPTIONS\u003d\"-O ^no-holes\"\n\nSigned-off-by: Filipe Manana \u003cfdmanana@suse.com\u003e\nReviewed-by: Zorro Lang \u003czlang@kernel.org\u003e\nReviewed-by: Qu Wenruo \u003cwqu@suse.com\u003e\nSigned-off-by: Zorro Lang \u003czlang@kernel.org\u003e\n" }, { "commit": "ce4e96a89a2b876497006b3fcbb388ee41f7aa8a", "tree": "4bf13b60cfec61124be7bd0ff9b6d2527072c472", "parents": [ "748ae5ab407610b1e3fa706c354016cf035e82ec" ], "author": { "name": "Christoph Hellwig", "email": "hch@lst.de", "time": "Mon May 18 06:24:53 2026" }, "committer": { "name": "Zorro Lang", "email": "zlang@kernel.org", "time": "Wed May 20 09:41:20 2026" }, "message": "generic: add a test that stresses metadata eviction\n\nThe fsverity test generic/579 could be used to reproduce a regression in\nthe xfs buffer hashing. Carlos Maiolino figured out that even removing\nthe fsverity-specific bits would reproduce the issue. This test is\nthus a copy of generic/579 with the fsverity bits striped out, and\nrun 5 times as that seems to be the threshold at which the issue\nreproduces reliably.\n\nSigned-off-by: Christoph Hellwig \u003chch@lst.de\u003e\nSigned-off-by: Zorro Lang \u003czlang@kernel.org\u003e\n" }, { "commit": "748ae5ab407610b1e3fa706c354016cf035e82ec", "tree": "8e2e07fdcc67c7037e9302d6ea04eb41e548ad77", "parents": [ "ffc8bad17e5b2f56e48dbac43f7c5ae8ac368fe5" ], "author": { "name": "Qu Wenruo", "email": "wqu@suse.com", "time": "Thu May 14 09:04:39 2026" }, "committer": { "name": "Zorro Lang", "email": "zlang@kernel.org", "time": "Sun May 17 08:11:33 2026" }, "message": "btrfs/340: add support for older kernels/progs\n\nThe test case will fail on older distros like SLE12-SP5, there are two\nproblems on the test case itself:\n\n- Requires \"--sync\" option from \"btrfs qgroup show\" subcommand\n That option was introduecd in v4.9.1, thus there are still some older\n distros using that very old version.\n\n- Not all kernels support quick inherit\n The kernel commit introducing the quick inherit is 6.9, thus some\n supported LTS kernels do not have that commit.\n\n In that case qgroup will just be marked inconsistent, this will cause\n the later \"btrfs qgroup show\" command to output a warning due to the\n inconsistent flag.\n\nFix both problems by:\n\n- Require \"--sync\" option from \"btrfs qgroup show\" subcommand\n So older distros will just skip the test.\n\n- Redirect all output from \"btrfs qgroup show\" to $seqres.full\n So older kernels can still pass the test case by marking qgroup\n inconsitent.\n\nReported-by: Long An \u003clan@suse.com\u003e\nSigned-off-by: Qu Wenruo \u003cwqu@suse.com\u003e\nReviewed-by: Zorro Lang \u003czlang@kernel.org\u003e\nSigned-off-by: Zorro Lang \u003czlang@kernel.org\u003e\n" }, { "commit": "ffc8bad17e5b2f56e48dbac43f7c5ae8ac368fe5", "tree": "907d2844388cd8c2da36db4f29160717662f488c", "parents": [ "393a4e58584dccbbc457aa7f16a315175bdaa474" ], "author": { "name": "Lukas Herbolt", "email": "lukas@herbolt.com", "time": "Thu May 14 11:39:15 2026" }, "committer": { "name": "Zorro Lang", "email": "zlang@kernel.org", "time": "Fri May 15 16:20:23 2026" }, "message": "xfs/21{6,7} Use default -l concurrency\u003d0 on mkfs.xfs that supports it\n\nThe XFS concurrency optimization breaks the log sizing check on\nsystems with non-rotational disks and high amount of CPUs. Default to\nthe old behavior with -l concurrency\u003d0 and with -d concurrency\u003d0.\n\nSigned-off-by: Lukas Herbolt \u003clukas@herbolt.com\u003e\nReviewed-by: Zorro Lang \u003czlang@kernel.org\u003e\nReviewed-by: \"Darrick J. Wong\" \u003cdjwong@kernel.org\u003e\nSigned-off-by: Zorro Lang \u003czlang@kernel.org\u003e\n" }, { "commit": "393a4e58584dccbbc457aa7f16a315175bdaa474", "tree": "6ca86687e41c07ae5b7d362c653d691e76778612", "parents": [ "7f162f5bcf50fc32c7e3b432e7ff189571a3151f" ], "author": { "name": "Lukas Herbolt", "email": "lukas@herbolt.com", "time": "Thu May 14 11:39:13 2026" }, "committer": { "name": "Zorro Lang", "email": "zlang@kernel.org", "time": "Fri May 15 16:20:23 2026" }, "message": "generic/{102,172,347}: Adapt test for XFS on systems with 128+CPUs + SSDs\n\nXFS will now scale by default its log size according the amount of CPUs. This\ncan significantly increase the log area and using hard-coded size of file size\nwill make the test file. Using the X percent of the free space to create the\nfile will let the test proceed.\n\nFor generic/102 we also have to change the 102.out as we cannot know the\namount written data and it has to be filtered out.\n\nFor generic/172 we just use the new function to calculate file big enough\nto consume most of the available space.\n\nThe generic/347 will use -l concurrency\u003d0 in case it is being run on the\nXFS filesystem and mkfs.xfs already supports this option.\n\nSigned-off-by: Lukas Herbolt \u003clukas@herbolt.com\u003e\nReviewed-by: Zorro Lang \u003czlang@kernel.org\u003e\nReviewed-by: \"Darrick J. Wong\" \u003cdjwong@kernel.org\u003e\nSigned-off-by: Zorro Lang \u003czlang@kernel.org\u003e\n" }, { "commit": "7f162f5bcf50fc32c7e3b432e7ff189571a3151f", "tree": "90d90be08e3ea2a23865bdf443026b6d105f2b1d", "parents": [ "55f45a620c552ec77e69291f58b89c27a972ba9a" ], "author": { "name": "Lukas Herbolt", "email": "lukas@herbolt.com", "time": "Thu May 14 11:39:11 2026" }, "committer": { "name": "Zorro Lang", "email": "zlang@kernel.org", "time": "Fri May 15 16:20:23 2026" }, "message": "common/xfs: helper function to check if -l/-d/-r concurrecy flags.\n\nCheck if the requested flag is supported by the mkfs.xfs on scratch\ndevice.\n\nSigned-off-by: Lukas Herbolt \u003clukas@herbolt.com\u003e\nReviewed-by: Zorro Lang \u003czlang@kernel.org\u003e\nReviewed-by: \"Darrick J. Wong\" \u003cdjwong@kernel.org\u003e\nSigned-off-by: Zorro Lang \u003czlang@kernel.org\u003e\n" }, { "commit": "55f45a620c552ec77e69291f58b89c27a972ba9a", "tree": "18c616ac1585b8cbf5336c3d2791278fecfa014c", "parents": [ "8c69a37c4f4c32f49957bc9720e11304de6fca33" ], "author": { "name": "Lukas Herbolt", "email": "lukas@herbolt.com", "time": "Thu May 14 11:39:09 2026" }, "committer": { "name": "Zorro Lang", "email": "zlang@kernel.org", "time": "Fri May 15 16:20:23 2026" }, "message": "common/rc: Add helper to calculate percentage of free space available\n\nIt calculates percentage of filesystem available space and returns it\nin MB. This helper will be used in the following test/generic fixes.\n\nSigned-off-by: Lukas Herbolt \u003clukas@herbolt.com\u003e\nReviewed-by: Zorro Lang \u003czlang@kernel.org\u003e\nReviewed-by: \"Darrick J. Wong\" \u003cdjwong@kernel.org\u003e\nSigned-off-by: Zorro Lang \u003czlang@kernel.org\u003e\n" }, { "commit": "8c69a37c4f4c32f49957bc9720e11304de6fca33", "tree": "369c2655537ba76da11bfad19dbf2911cf03b1cc", "parents": [ "4192c9e2266c8e92b49e3f73fe505101ab36ce1b" ], "author": { "name": "Christian Brauner", "email": "brauner@kernel.org", "time": "Mon Mar 23 13:32:15 2026" }, "committer": { "name": "Zorro Lang", "email": "zlang@kernel.org", "time": "Tue May 12 17:22:35 2026" }, "message": "vfs: fix dumpable race in create_userns_hierarchy()\n\nAll processes in the userns hierarchy share the same mm_struct due to\nCLONE_VM. When a child calls setresuid() in switch_ids(), the kernel\nclears the dumpable flag on the shared mm. The child immediately\nre-sets it via prctl(PR_SET_DUMPABLE, 1), but there is a window\nbetween the two where the parent may be opening /proc/PID/ns/user for\nthe previous level\u0027s child. That open hits ptrace_may_access() which\nchecks get_dumpable(mm) and, finding it zero, falls back to\nptrace_has_cap(mm-\u003euser_ns, mode). Since mm-\u003euser_ns is init_user_ns\n(the mm was created by real root) and the parent is only ns-root, the\ncapability check fails and the open returns -EACCES.\n\nFix this by extending the existing two-way socketpair handshake into a\nthree-way handshake: the child signals the parent after becoming\ndumpable, then blocks until the parent has opened /proc/PID/ns/user\nand signals back. Only then does the child proceed to\ncreate_userns_hierarchy() for the next level. This guarantees no\nconcurrent setresuid() can clear dumpable while the parent\u0027s open() is\nin flight.\n\nSigned-off-by: Christian Brauner \u003cbrauner@kernel.org\u003e\nReviewed-by: Zorro Lang \u003czlang@kernel.org\u003e\nSigned-off-by: Zorro Lang \u003czlang@kernel.org\u003e\n" }, { "commit": "4192c9e2266c8e92b49e3f73fe505101ab36ce1b", "tree": "3dbb1897bc99098a0b03f6d2cdcd099613afcd6e", "parents": [ "0d6ed520323d82dbb0f9ad77a7a63e48d01a3c4e" ], "author": { "name": "David Disseldorp", "email": "ddiss@suse.de", "time": "Mon Mar 23 04:45:58 2026" }, "committer": { "name": "Zorro Lang", "email": "zlang@kernel.org", "time": "Tue May 12 17:22:03 2026" }, "message": "common/rc: add default _udev_wait timeout\n\nIf udevd isn\u0027t running then \"udevadm wait\" loops endlessly. With a\n--timeout parameter provided, udevadm (udev-254.27) prints to stderr:\n Timed out for waiting devices being initialized.\n\nThe error message causes test failure via golden output mismatch and\nis easier to debug than an infinite loop.\n\nThe timeout parameter is also compatible with existing\nargs\u003d\"--removed \u003cdev\u003e\" callers. 60 seconds is relatively arbitrary, but\nshould be long enough to avoid premature timeout on regular systems.\n\nSigned-off-by: David Disseldorp \u003cddiss@suse.de\u003e\nReviewed-by: Zorro Lang \u003czlang@kernel.org\u003e\nSigned-off-by: Zorro Lang \u003czlang@kernel.org\u003e\n" }, { "commit": "0d6ed520323d82dbb0f9ad77a7a63e48d01a3c4e", "tree": "f70165a230fb7e777b6fc15b4c9843ffa8751bda", "parents": [ "3aca6733f6cd93cbf000b7dbfc8af128796e2425" ], "author": { "name": "Ojaswin Mujoo", "email": "ojaswin@linux.ibm.com", "time": "Wed Apr 01 10:09:41 2026" }, "committer": { "name": "Zorro Lang", "email": "zlang@kernel.org", "time": "Tue May 12 17:13:30 2026" }, "message": "ext4/307: Calculate donor size to avoid failures for 64k bs\n\nCalculating size of donor file based on du output causes the following\nissue:\n\ne4compact.c:68: do_defrag_range: Assertion `donor-\u003elength \u003e\u003d len\u0027 failed.\n./common/rc: line 4616: 1583182 Aborted (core dumped)\n\nThe issue is that e4compat asserts that the stat.st_size of all files\ncombined shouldn\u0027t be more than that of the donor file. In case of 64k\nblocksize, fsstress often creates sparse files which have \u003e3G size even\nthough their disk utilization is \u003c100M. Since donor file is not big\nenough, this trips e4compat causing the failure.\n\nFix this by using stat.st_size rather than du value to get the size.\nAlso, make some changes to avoid future ENOSPC like reducing the -n\npassed to fsstress to keep the size around 2G and making sure we use the\nwhole SCRATCH disk instead of 500M.\n\nWhile we are at it, add some new lines to make the code bit more\nreadable.\n\nReported-by: Disha Goel \u003cdisgoel@linux.ibm.com\u003e\nTested-by: Disha Goel \u003cdisgoel@linux.ibm.com\u003e\nSigned-off-by: Ojaswin Mujoo \u003cojaswin@linux.ibm.com\u003e\nSigned-off-by: Zorro Lang \u003czlang@kernel.org\u003e\n" }, { "commit": "3aca6733f6cd93cbf000b7dbfc8af128796e2425", "tree": "f7ed0a686bdf4c7aeef2feec83c27ff3899a9b56", "parents": [ "a974318c5bfa1debda2f0ce18fccb1ea336dcfbb" ], "author": { "name": "Disha Goel", "email": "disgoel@linux.ibm.com", "time": "Fri May 08 14:47:01 2026" }, "committer": { "name": "Zorro Lang", "email": "zlang@kernel.org", "time": "Mon May 11 16:18:14 2026" }, "message": "btrfs/291: fix state transition logic and add size requirement\n\nThis patch fixes two issues in btrfs/291:\n\n1. Add dynamic LOGWRITES_DEV size requirement based on SCRATCH_DEV\n The test creates LVM snapshots at each FUA point during replay,\n requiring significant space. Calculate the required size as 120%\n of SCRATCH_DEV size (adding 20% overhead for LVM snapshots and\n metadata) to ensure the test works regardless of SCRATCH_DEV size.\n\n2. Fix state transition logic for verity enablement\n The original test assumed orphan items would always be created\n during verity enablement (state 0-\u003e1 transition). However, in\n some cases verity completes without creating orphan items,\n causing the test to fail with \"expected to reach verity done state\".\n\n Fix by transitioning to state 1 when either orphan items exist\n OR merkle items appear, handling both verity enablement paths.\n Also improve state 1 validation to only check for cleared merkle\n items when measurement actually fails.\n\nThe test now correctly handles verity enablement with or without\norphan items while maintaining crash consistency validation, and\nworks with any SCRATCH_DEV size.\n\nSuggested-by: Anand Jain \u003canajain.sg@gmail.com\u003e\nSigned-off-by: Disha Goel \u003cdisgoel@linux.ibm.com\u003e\nReviewed-by: Zorro Lang \u003czlang@kernel.org\u003e\nSigned-off-by: Zorro Lang \u003czlang@kernel.org\u003e\n" }, { "commit": "a974318c5bfa1debda2f0ce18fccb1ea336dcfbb", "tree": "b8e953853944022b0e9f114ad702415aa6fcd78f", "parents": [ "8413f58d843f88c5132474297e5cbfac5ba6dad5" ], "author": { "name": "Disha Goel", "email": "disgoel@linux.ibm.com", "time": "Fri May 08 14:47:00 2026" }, "committer": { "name": "Zorro Lang", "email": "zlang@kernel.org", "time": "Mon May 11 16:18:14 2026" }, "message": "common/dmlogwrites: add _require_log_writes_sized helper\n\nAdd a new helper function _require_log_writes_sized() to check if the\nLOGWRITES_DEV meets a minimum size requirement.\n\nThis is useful for tests that use dm-log-writes with additional space\nrequirements, such as creating LVM snapshots during log replay or tests\nthat generate large amounts of logged I/O operations.\n\nThe function takes a size parameter in KB, calls _require_log_writes()\ninternally, and skips the test if LOGWRITES_DEV is smaller than the\nrequired size.\n\nSuggested-by: Darrick J. Wong \u003cdjwong@kernel.org\u003e\nSuggested-by: Anand Jain \u003canajain.sg@gmail.com\u003e\nSigned-off-by: Disha Goel \u003cdisgoel@linux.ibm.com\u003e\nReviewed-by: Zorro Lang \u003czlang@kernel.org\u003e\nSigned-off-by: Zorro Lang \u003czlang@kernel.org\u003e\n" }, { "commit": "8413f58d843f88c5132474297e5cbfac5ba6dad5", "tree": "e49c5ec2aa8bec43ad73b1c20c69ae0d26126013", "parents": [ "b8bcadc36c3624fad6f6a2b8bdc407ae2a88b496" ], "author": { "name": "Qu Wenruo", "email": "wqu@suse.com", "time": "Sun May 10 00:17:04 2026" }, "committer": { "name": "Zorro Lang", "email": "zlang@kernel.org", "time": "Sun May 10 19:10:23 2026" }, "message": "fstests: support new mkfs.btrfs help string\n\n[BUG]\nThe latest btrfs-progs release v7.0 breaks a lot of tests, e.g.\nbtrfs/003:\n\nbtrfs/003 55s ... [failed, exit status 1]- output mismatch (see /home/kts/kernel-test-suite/src/xfstests/results//btrfs/003.out.bad)\n --- tests/btrfs/003.out 2026-05-09 03:07:52.952213772 +0900\n +++ /home/kts/kernel-test-suite/src/xfstests/results//btrfs/003.out.bad 2026-05-09 23:32:17.595751435 +0900\n @@ -1,2 +1,3 @@\n QA output created by 003\n -Silence is golden\n +mkfs failed\n +(see /home/kts/kernel-test-suite/src/xfstests/results//btrfs/003.full for details)\n ...\n (Run \u0027diff -u /home/kts/kernel-test-suite/src/xfstests/tests/btrfs/003.out /home/kts/kernel-test-suite/src/xfstests/results//btrfs/003.out.bad\u0027 to see the entire diff)\n\n[CAUSE]\nCommit ec8324cc4b8b (\"btrfs-progs: help: adjust alignment of text\") from\nbtrfs-progs changed the helpe string output format, the previous help\nstring for \"-f\" option of mkfs.btrfs looks like this:\n\n -f|--force force overwrite of existing filesystem\n\nThere were 4 spaces before the \"-f\", thus can be matched by\n\"[[:space:]]-f[[:space]|,]\".\n\nBut now the help string looks like:\n\n-f, --force force overwrite of existing filesystem\n\nWhich removed the spaces at the beginning.\n\nThe existing regex to catch \"-f\" requires at least one space before\n\"-f\", now it doesn\u0027t match the new btrfs help string, and cause mkfs\nfailure because we can not overwrite an existing scratch device without\n\"-f\" option.\n\n[FIX]\nChange the regex to allow no space before \"-f\".\n\nReported-by: Shin\u0027ichiro Kawasaki \u003cshinichiro.kawasaki@wdc.com\u003e\nLink: https://lore.kernel.org/linux-btrfs/af9EHC5b-aBdr1zn@shinmob/\nSigned-off-by: Qu Wenruo \u003cwqu@suse.com\u003e\nReviewed-by: Zorro Lang \u003czlang@kernel.org\u003e\nSigned-off-by: Zorro Lang \u003czlang@kernel.org\u003e\n" }, { "commit": "b8bcadc36c3624fad6f6a2b8bdc407ae2a88b496", "tree": "3e3f8738e57c375c52051c8b888f1f35e3d4244d", "parents": [ "cdd9140a2f78951f077aafeb0f337c8a93a35026" ], "author": { "name": "Joanne Chang", "email": "joannechien@google.com", "time": "Tue Apr 28 02:24:40 2026" }, "committer": { "name": "Zorro Lang", "email": "zlang@kernel.org", "time": "Sat May 09 16:51:09 2026" }, "message": "f2fs/{022,023}: clear MOUNT_OPTIONS alongside MKFS_OPTIONS\n\nCurrently, f2fs/022 and f2fs/023 only clears MKFS_OPTIONS. This causes\nthe tests to fail when leftover MOUNT_OPTIONS depend on unapplied\nMKFS_OPTIONS. So MOUNT_OPTIONS should also be cleared to ensure reliable\nmounting.\n\nSigned-off-by: Joanne Chang \u003cjoannechien@google.com\u003e\nReviewed-by: Zorro Lang \u003czlang@kernel.org\u003e\nReviewed-by: Chao Yu \u003cchao@kernel.org\u003e\nSigned-off-by: Zorro Lang \u003czlang@kernel.org\u003e\n" }, { "commit": "cdd9140a2f78951f077aafeb0f337c8a93a35026", "tree": "3e3bec0311354236620818c9238f137ef002b1c8", "parents": [ "f61ba0fde3ec9cdfd7eed711aca2b28e973c2905" ], "author": { "name": "Darrick J. Wong", "email": "djwong@kernel.org", "time": "Thu May 07 22:23:34 2026" }, "committer": { "name": "Zorro Lang", "email": "zlang@kernel.org", "time": "Sat May 09 16:51:03 2026" }, "message": "generic/45[34]: don\u0027t warn on mixed bidirectional characters\n\nWe\u0027re dropping this warning from xfs_scrub because the warning is\nincorrect, so let\u0027s make sure the QA system checks that it\u0027s gone.\n\nCc: gedalya@gedalya.net\nSigned-off-by: \"Darrick J. Wong\" \u003cdjwong@kernel.org\u003e\nReviewed-by: Zorro Lang \u003czlang@kernel.org\u003e\nSigned-off-by: Zorro Lang \u003czlang@kernel.org\u003e\n" }, { "commit": "f61ba0fde3ec9cdfd7eed711aca2b28e973c2905", "tree": "e68d2e20cb5d9a66cc84c3b8a9adb18f86402b19", "parents": [ "528c7ba2a275d4e4b592145acd4366158af48cbc" ], "author": { "name": "Darrick J. Wong", "email": "djwong@kernel.org", "time": "Thu May 07 22:23:19 2026" }, "committer": { "name": "Zorro Lang", "email": "zlang@kernel.org", "time": "Sat May 09 16:51:03 2026" }, "message": "generic/45[34]: add detection of confusable variation sequences\n\nArsTechnica recently wrote about a GitHub supply chain attack wherein\nnon-rendering unicode sequences were embedded in javascript files to\nhide payloads that could be decrypted trivially later. While these are\nunlikely to appear in file and attribute names, xfs_scrub will warn about\nthis sort of steganography, so let\u0027s make sure it works.\n\nSigned-off-by: \"Darrick J. Wong\" \u003cdjwong@kernel.org\u003e\nReviewed-by: Zorro Lang \u003czlang@kernel.org\u003e\nSigned-off-by: Zorro Lang \u003czlang@kernel.org\u003e\n" }, { "commit": "528c7ba2a275d4e4b592145acd4366158af48cbc", "tree": "d1907f8da61f0b08f7be4c588fa5982d8efac922", "parents": [ "12c729c1a296dcd799a217a26c433f9bf79477fa" ], "author": { "name": "Filipe Manana", "email": "fdmanana@suse.com", "time": "Mon May 04 16:09:57 2026" }, "committer": { "name": "Zorro Lang", "email": "zlang@kernel.org", "time": "Thu May 07 20:25:16 2026" }, "message": "fstests: add commit IDs for kernel fixes already merged\n\nA few fixes were merged into Linus\u0027 tree, so add their commit IDs in the\nrespective test cases.\n\nSigned-off-by: Filipe Manana \u003cfdmanana@suse.com\u003e\nReviewed-by: Anand Jain \u003casj@kernel.org\u003e\nSigned-off-by: Zorro Lang \u003czlang@kernel.org\u003e\n" }, { "commit": "12c729c1a296dcd799a217a26c433f9bf79477fa", "tree": "b6529aa07eb8ca3be3823035bce105c35b32beb7", "parents": [ "d9b1d9c6f43ee71cf16020afde63d453bf0468ac" ], "author": { "name": "Christoph Hellwig", "email": "hch@lst.de", "time": "Thu May 07 05:09:49 2026" }, "committer": { "name": "Zorro Lang", "email": "zlang@kernel.org", "time": "Thu May 07 18:48:11 2026" }, "message": "xfs/216: pass -t xfs to _mount\n\nWithout this, mount probes different file system type, which ended up\nexposing kernel bugs in blocksize \u003e PAGE_SIZE bugs in tons of\nlegacy file systems. Bypass that by forcing XFS and making the\nmount a little more efficient.\n\nSigned-off-by: Christoph Hellwig \u003chch@lst.de\u003e\nReviewed-by: Zorro Lang \u003czlang@kernel.org\u003e\nReviewed-by: Andrey Albershteyn \u003caalbersh@kernel.org\u003e\nSigned-off-by: Zorro Lang \u003czlang@kernel.org\u003e\n" }, { "commit": "d9b1d9c6f43ee71cf16020afde63d453bf0468ac", "tree": "47adefba321f687c3a9f2c38153e2fdd4db8631d", "parents": [ "95155ec233fad04d8701bfec74accefec691342c" ], "author": { "name": "Johannes Thumshirn", "email": "johannes.thumshirn@wdc.com", "time": "Wed Apr 29 20:46:24 2026" }, "committer": { "name": "Zorro Lang", "email": "zlang@kernel.org", "time": "Mon May 04 02:57:17 2026" }, "message": "generic/793: limit number of overwrites\n\nTestcase generic/793 overwrites a 1GB file \"number of sequential zones\"\ntimes in order to trigger a bug in zoned BTRFS\u0027 garbage collection. But on\nscratch devices with a huge number of sequential zones (like SMR drives)\nthis can cause very long runtimes.\n\nAs the testcase also limits the filesystem size to 16GB, constraint the\nnumber of overwrite loops to the number of sequential zones in the\nfilesystem. This still guarantees that the file will be overwritten often\nenough to trigger the bug while not exploding the overall test runtime.\n\nReported-by: Shin\u0027ichiro Kawasaki \u003cshinichiro.kawasaki@wdc.com\u003e\nReviewed-by: Christoph Hellwig \u003chch@lst.de\u003e\nSigned-off-by: Johannes Thumshirn \u003cjohannes.thumshirn@wdc.com\u003e\nReviewed-by: Hans Holmberg \u003chans.holmberg@wdc.com\u003e\nSigned-off-by: Zorro Lang \u003czlang@kernel.org\u003e\n" }, { "commit": "95155ec233fad04d8701bfec74accefec691342c", "tree": "a240f4c8b6e8c445c5cb0e9fcead6efdd1fd8764", "parents": [ "55e172c4c9144d2b944f0c03801b507f9e74c5ba" ], "author": { "name": "Zhang Yi", "email": "yi.zhang@huawei.com", "time": "Tue Apr 28 08:57:50 2026" }, "committer": { "name": "Zorro Lang", "email": "zlang@kernel.org", "time": "Fri May 01 16:57:03 2026" }, "message": "generic: test post-EOF gap zeroing persistence\n\nTest that extending a file past a non-block-aligned EOF correctly\nzero-fills the gap [old_EOF, block_boundary), and that this zeroing\npersists through a filesystem shutdown+remount cycle.\n\nStale data beyond EOF can persist on disk when append write data blocks\nare flushed before the on-disk file size update, or when concurrent\nappend writeback and mmap writes persist non-zero data past EOF.\nSubsequent post-EOF operations (append write, fallocate, truncate up)\nmust zero-fill and persist the gap to prevent exposing stale data.\n\nThe test pollutes the file\u0027s last physical block (via FIEMAP + raw\ndevice write) with a sentinel pattern beyond i_size, then performs each\nextend operation and verifies the gap is zeroed both in memory and on\ndisk.\n\nSigned-off-by: Zhang Yi \u003cyi.zhang@huawei.com\u003e\nReviewed-by: Brian Foster \u003cbfoster@redhat.com\u003e\nReviewed-by: Zorro Lang \u003czlang@kernel.org\u003e\nSigned-off-by: Zorro Lang \u003czlang@kernel.org\u003e\n" }, { "commit": "55e172c4c9144d2b944f0c03801b507f9e74c5ba", "tree": "498f1fd316151df98c1a0c190850e455ad248be1", "parents": [ "832a6d9ac226e017b41b052b89bfa2f66ac8c039" ], "author": { "name": "Zorro Lang", "email": "zlang@kernel.org", "time": "Fri Apr 17 18:08:39 2026" }, "committer": { "name": "Zorro Lang", "email": "zlang@kernel.org", "time": "Fri May 01 15:49:22 2026" }, "message": "xfs/227: fix bad variable expansion\n\nThe create_data() set cmd_str to \"-c \\\"resvsp $((off * 4096)) 4096\\\".\nThat cause xfs_io can\u0027t split the command line properly, it fails as:\n\n command \"\"resvsp\" not found\n\nFixes: aaf40fb7 (\"fstests: xfs/227 is really slow\")\nReviewed-by: Darrick J. Wong \u003cdjwong@kernel.org\u003e\nSigned-off-by: Zorro Lang \u003czlang@kernel.org\u003e\n" }, { "commit": "832a6d9ac226e017b41b052b89bfa2f66ac8c039", "tree": "9332ce123d79a85b2fc09a3d9591b6a37c880535", "parents": [ "078f3204e049c8ffcf7819bab9afd419267d0084" ], "author": { "name": "Darrick J. Wong", "email": "djwong@kernel.org", "time": "Thu Apr 16 19:15:43 2026" }, "committer": { "name": "Zorro Lang", "email": "zlang@kernel.org", "time": "Fri May 01 15:42:41 2026" }, "message": "check: capture dmesg of mount failures if test fails\n\nCapture the kernel output after a mount failure occurs. If the test\nitself fails, then keep the logging output for further diagnosis. The\nxunit.xsd update adds a \u003cmount-failure\u003e element to the xml output, whose\noutput is the mountfail file.\n\nNote that because the .mountfail file is preserved by ./check, the new\nselftest requires the user to check for the .mountfail file. This is a\nlittle awkward.\n\nSigned-off-by: Darrick J. Wong \u003cdjwong@kernel.org\u003e\nReviewed-by: Christoph Hellwig \u003chch@lst.de\u003e\nSigned-off-by: Zorro Lang \u003czlang@kernel.org\u003e\n" }, { "commit": "078f3204e049c8ffcf7819bab9afd419267d0084", "tree": "60510e19016a84be873400ba11c78ccd5e394987", "parents": [ "cac9fe2b8dc3e6dbc0c2383b9e3b4d3c1b9e7dd0" ], "author": { "name": "Darrick J. Wong", "email": "djwong@kernel.org", "time": "Mon Apr 13 17:51:09 2026" }, "committer": { "name": "Zorro Lang", "email": "zlang@kernel.org", "time": "Fri May 01 15:40:16 2026" }, "message": "treewide: convert all $MOUNT_PROG to _mount\n\nGoing to add some new log scraping functionality when mount failures\noccur, so we need everyone to use _mount instead of $MOUNT_PROG.\n\nSigned-off-by: Darrick J. Wong \u003cdjwong@kernel.org\u003e\nReviewed-by: Christoph Hellwig \u003chch@lst.de\u003e\nSigned-off-by: Zorro Lang \u003czlang@kernel.org\u003e\n" }, { "commit": "cac9fe2b8dc3e6dbc0c2383b9e3b4d3c1b9e7dd0", "tree": "0da288dba10795ca83c4c71a750284632bb62677", "parents": [ "d59865ddf8e4dc199e8b572fe8475abb621d6725" ], "author": { "name": "Johannes Thumshirn", "email": "johannes.thumshirn@wdc.com", "time": "Wed Apr 15 10:40:15 2026" }, "committer": { "name": "Zorro Lang", "email": "zlang@kernel.org", "time": "Thu Apr 16 18:19:30 2026" }, "message": "generic: test premature ENOSPC in zoned garbage collection\n\nThis test stresses garbage collection in zoned file systems by\nconstantly overwriting the same file. It is inspired by a reproducer\nfor a btrfs bugfix.\n\nSigned-off-by: Johannes Thumshirn \u003cjohannes.thumshirn@wdc.com\u003e\nReviewed-by: Christoph Hellwig \u003chch@lst.de\u003e\nReviewed-by: Zorro Lang \u003czlang@kernel.org\u003e\nSigned-off-by: Zorro Lang \u003czlang@kernel.org\u003e\n" }, { "commit": "d59865ddf8e4dc199e8b572fe8475abb621d6725", "tree": "f5ab17ab4d9314936a6e43661e8b5876f5fe51a9", "parents": [ "dde983e4419abd83568e9c9ebb4b576848d48eb7" ], "author": { "name": "Christoph Hellwig", "email": "hch@lst.de", "time": "Tue Apr 14 07:55:47 2026" }, "committer": { "name": "Zorro Lang", "email": "zlang@kernel.org", "time": "Thu Apr 16 17:00:19 2026" }, "message": "xfs: add a test for number of open zones on conventional devices\n\nTest that mounts of zoned file systems on conventional devices don\u0027t\ncreate more open zones than allowed when the last blocks in one or more\nzones have been invalidated.\n\nSigned-off-by: Christoph Hellwig \u003chch@lst.de\u003e\nReviewed-by: Darrick J. Wong \u003cdjwong@kernel.org\u003e\nSigned-off-by: Zorro Lang \u003czlang@kernel.org\u003e\n" }, { "commit": "dde983e4419abd83568e9c9ebb4b576848d48eb7", "tree": "56c14b2e6f3bc3f5b4b33e565abcd4f026be566e", "parents": [ "56b35283cf8e98b50c13f64912161830e6bd0a08" ], "author": { "name": "Darrick J. Wong", "email": "djwong@kernel.org", "time": "Mon Apr 13 17:50:54 2026" }, "committer": { "name": "Zorro Lang", "email": "zlang@kernel.org", "time": "Thu Apr 16 16:49:53 2026" }, "message": "xfs/841: actually compare the new filesystem contents to the proto dir\n\nWe should confirm that the directory tree was correctly copied into the\nnew filesystem, not just that the image files all have the same\nchecksum.\n\nSigned-off-by: Darrick J. Wong \u003cdjwong@kernel.org\u003e\nReviewed-by: Christoph Hellwig \u003chch@lst.de\u003e\nSigned-off-by: Zorro Lang \u003czlang@kernel.org\u003e\n" }, { "commit": "56b35283cf8e98b50c13f64912161830e6bd0a08", "tree": "5ec33c19ed742293e67a97ee108508c9f2dc4387", "parents": [ "c4e03688706d008211c8bb9faeef8fbab3a45615" ], "author": { "name": "Jan Prusakowski", "email": "jprusakowski@google.com", "time": "Fri Apr 10 13:18:20 2026" }, "committer": { "name": "Zorro Lang", "email": "zlang@kernel.org", "time": "Thu Apr 16 16:48:18 2026" }, "message": "generic/050: handle f2fs as nojournal filesystem\n\nF2FS uses a checkpoint mechanism for metadata consistency rather than a\ntraditional journal. Roll-forward recovery is only needed if there are\nfsync\u0027d files since the last checkpoint.\n\nIn this test case, files are created without fsync, so there is no\nroll-forward data to replay during mount.\n\nTherefore, F2FS does not need to write to the device to recover, and\nsuccessfully mounts on the read-only block device. Thus, it should be\ntreated as nojournal in this case.\n\nSigned-off-by: Jan Prusakowski \u003cjprusakowski@google.com\u003e\nReviewed-by: Zorro Lang \u003czlang@kernel.org\u003e\nSigned-off-by: Zorro Lang \u003czlang@kernel.org\u003e\n" }, { "commit": "c4e03688706d008211c8bb9faeef8fbab3a45615", "tree": "e384ccef4714304aeef3d9414fc07c51aab0ed3c", "parents": [ "f3e91c87aa9f1a5a4cea4270d1c22c6ba091f34a" ], "author": { "name": "Filipe Manana", "email": "fdmanana@suse.com", "time": "Thu Apr 09 15:22:27 2026" }, "committer": { "name": "Zorro Lang", "email": "zlang@kernel.org", "time": "Mon Apr 13 19:41:25 2026" }, "message": "generic: test fsync against a removed directory\n\nTest that if we fsync a directory that became empty after removing it\nand we have a power failure, we are able to mount the filesystem and the\ndirectory does not exists anymore.\n\nThis exercises a bug found in btrfs which is fixed by the following patch:\n\n btrfs: fix missing last_unlink_trans update when removing a directory\n\nSigned-off-by: Filipe Manana \u003cfdmanana@suse.com\u003e\nReviewed-by: Boris Burkov \u003cboris@bur.io\u003e\nSigned-off-by: Zorro Lang \u003czlang@kernel.org\u003e\n" }, { "commit": "f3e91c87aa9f1a5a4cea4270d1c22c6ba091f34a", "tree": "7fe0e6e22b07c09afa24c088c34e50b69af1f8c3", "parents": [ "53e68c2afbb9c62de47a227dd05b4373ff85aa1c" ], "author": { "name": "Ojaswin Mujoo", "email": "ojaswin@linux.ibm.com", "time": "Fri Apr 10 06:36:06 2026" }, "committer": { "name": "Zorro Lang", "email": "zlang@kernel.org", "time": "Mon Apr 13 17:42:42 2026" }, "message": "treewide: Use _sysfs_queue_path helper in all queue access locations\n\nThe sysfs queue dir for partitions is present in their parent device\u0027s\nsysfs dir. This requires extra checks when trying to reading queue data\nfrom sysfs. Currently, fstests open codes this handling or straight up\nignores it in some places.\n\nAvoid this by converting all remaining direct sysfs queue accesses to\nuse the _sysfs_queue_path helper function. This ensures proper handling\nof partitions across the entire codebase and provides a consistent\ninterface for accessing queue attributes.\n\nSigned-off-by: Ojaswin Mujoo \u003cojaswin@linux.ibm.com\u003e\nReviewed-by: Darrick J. Wong \u003cdjwong@kernel.org\u003e\nSigned-off-by: Zorro Lang \u003czlang@kernel.org\u003e\n" }, { "commit": "53e68c2afbb9c62de47a227dd05b4373ff85aa1c", "tree": "bd7bb5f5367dc1d8e5f9420c7463b37354629590", "parents": [ "fb11f06d9026a9a449c4a5afb7e9263e88fdea22" ], "author": { "name": "Ojaswin Mujoo", "email": "ojaswin@linux.ibm.com", "time": "Fri Apr 10 06:36:05 2026" }, "committer": { "name": "Zorro Lang", "email": "zlang@kernel.org", "time": "Mon Apr 13 17:42:41 2026" }, "message": "generic/775: Fix an infinite loop due to variable name clash\n\nWe use i as the iteration variable in the main test loop as well as some\ninternal loops. Due to this clash, $i variable of main test loops was\ngetting modified by the following loop in prep_mixed_mapping().\n\n\tfor ((i\u003d0; i\u003cnum_blocks; i++)); do\n\nIf num_blocks is less than 10 (example ext4 with blocksize 4k and\ncluster size 8k) i would always be set as 3 and the main loop would\nnever exit. Use local variable to fix this and while we are at it,\nrename it for clarity.\n\nReported-by: Disha Goel \u003cdisgoel@linux.ibm.com\u003e\nSigned-off-by: Ojaswin Mujoo \u003cojaswin@linux.ibm.com\u003e\nReviewed-by: Darrick J. Wong \u003cdjwong@kernel.org\u003e\nSigned-off-by: Zorro Lang \u003czlang@kernel.org\u003e\n" }, { "commit": "fb11f06d9026a9a449c4a5afb7e9263e88fdea22", "tree": "57cdb8d9d17bcba1e793ddadf5d0f72425a14045", "parents": [ "4dc8a079a670c0831e75dd9fc25e02f94b261807" ], "author": { "name": "Ojaswin Mujoo", "email": "ojaswin@linux.ibm.com", "time": "Fri Apr 10 06:36:04 2026" }, "committer": { "name": "Zorro Lang", "email": "zlang@kernel.org", "time": "Mon Apr 13 17:42:41 2026" }, "message": "generic/765: Ignore mkfs warning\n\nThis test validates atomic writes for all possible block sizes. In ext4, for\nsmaller block sizes with configurations like:\n\nexport MKFS_OPTIONS\u003d\"-O bigalloc,quota -b 65536 -C 131072\"\n\nThe output can get corrupted with warnings like below because clustersize\nmore than 16xbs is experimental:\n\n+ 16 times the block size is considered experimental\n\nHence pipe these to seqres.full to avoid false negatives.\n\nReported-by: Disha Goel \u003cdisgoel@linux.ibm.com\u003e\nSigned-off-by: Ojaswin Mujoo \u003cojaswin@linux.ibm.com\u003e\nReviewed-by: Zorro Lang \u003czlang@kernel.org\u003e\nSigned-off-by: Zorro Lang \u003czlang@kernel.org\u003e\n" }, { "commit": "4dc8a079a670c0831e75dd9fc25e02f94b261807", "tree": "5a8c55ea593089e32f3ccef9aee06fa765e22330", "parents": [ "2c002c3a86c03155e6129fea77c1b2c29a43b18a" ], "author": { "name": "Ojaswin Mujoo", "email": "ojaswin@linux.ibm.com", "time": "Fri Apr 10 06:36:03 2026" }, "committer": { "name": "Zorro Lang", "email": "zlang@kernel.org", "time": "Mon Apr 13 17:42:41 2026" }, "message": "generic/765: Fix sysfs queue path for nvme partitions\n\nThis tests checks atomic write limits reported by statx() are same as\nthe ones reported by sysfs, however nvme partitions don\u0027t have the\n/sys/block/nvme0n1p1 style entry and also use the namespace\u0027s queue\nlimits for atomic writes. This causes the test to fail because it\nexpects /sys/block/nvme0n*p* to be present.\n\nHence, use the new _sysfs_queue_path() helper to get the correct\natomic write limits\n\nAlso, fix a small typo (s/awu_min/awu_max) when printing awu limits to\nseqres.full\n\nReported-by: Disha Goel \u003cdisgoel@linux.ibm.com\u003e\nSigned-off-by: Ojaswin Mujoo \u003cojaswin@linux.ibm.com\u003e\nReviewed-by: Darrick J. Wong \u003cdjwong@kernel.org\u003e\nSigned-off-by: Zorro Lang \u003czlang@kernel.org\u003e\n" }, { "commit": "2c002c3a86c03155e6129fea77c1b2c29a43b18a", "tree": "b3b4e871b9dba098e5ee22c92c05cc30a88fd029", "parents": [ "ea4d04e7b6766118bbf0cbe596de2e41a325119c" ], "author": { "name": "Ojaswin Mujoo", "email": "ojaswin@linux.ibm.com", "time": "Fri Apr 10 06:36:02 2026" }, "committer": { "name": "Zorro Lang", "email": "zlang@kernel.org", "time": "Mon Apr 13 17:42:41 2026" }, "message": "common/rc: Add _sysfs_queue_path helper with partition support\n\nAdd a generic helper function to get the sysfs queue path for block\ndevices, properly handling partitions. Partitions don\u0027t have their own\nqueue directory in sysfs - they inherit from their parent device.\n\nThis helper checks if a device is a partition and returns the parent\ndevice\u0027s queue path accordingly, making it easier to access queue\nattributes for both whole disks and partitions.\n\nSigned-off-by: Ojaswin Mujoo \u003cojaswin@linux.ibm.com\u003e\nReviewed-by: Darrick J. Wong \u003cdjwong@kernel.org\u003e\nSigned-off-by: Zorro Lang \u003czlang@kernel.org\u003e\n" }, { "commit": "ea4d04e7b6766118bbf0cbe596de2e41a325119c", "tree": "1ba9783efd55dc8bf3456a5b6eabdbf0b165ca13", "parents": [ "0bece2e05be1c5179654f7437519f80eeef71111" ], "author": { "name": "Ojaswin Mujoo", "email": "ojaswin@linux.ibm.com", "time": "Fri Apr 10 06:36:01 2026" }, "committer": { "name": "Zorro Lang", "email": "zlang@kernel.org", "time": "Mon Apr 13 17:42:41 2026" }, "message": "ext4/061,062: Minor fixes and refactoring\n\nFix 2 issues in the tests:\n\n1. Use fs atomic write limits instead of bdev\u0027s\n\nThe tests use block device\u0027s limits instead of limits advertised\nby filesystem. This can cause failures because block device might\nadvertise a higher maximum than the FS. Fix the tests to use\nfilesystem\u0027s limits instead.\n\n2. Fix the test loop range\n\nFix the test loop start and end so that our calculations for\nblocksize, clustersize and iosize are correct.\n\nReported-by: Disha Goel \u003cdisgoel@linux.ibm.com\u003e\nReviewed-by: Darrick J. Wong \u003cdjwong@kernel.org\u003e\nSigned-off-by: Ojaswin Mujoo \u003cojaswin@linux.ibm.com\u003e\nSigned-off-by: Zorro Lang \u003czlang@kernel.org\u003e\n" }, { "commit": "0bece2e05be1c5179654f7437519f80eeef71111", "tree": "0df84a792d7bde01afa218a40d1ed933ff8d68c9", "parents": [ "29dbf07039172295ebe3ae158eb1a7e44dedcc96" ], "author": { "name": "Disha Goel", "email": "disgoel@linux.ibm.com", "time": "Mon Mar 23 12:40:05 2026" }, "committer": { "name": "Zorro Lang", "email": "zlang@kernel.org", "time": "Wed Apr 01 15:07:25 2026" }, "message": "generic/730: skip test if DAX mount option is enabled\n\nThe generic/730 test utilizes the scsi_debug driver to create a\nsynthetic SCSI block device. Since scsi_debug does not support\nDAX capabilities, attempting to mount it with \"-o dax\" results\nin failures (specifically on ext4).\n\nSince this test is not DAX-related, exclude it when DAX is\nspecified in MOUNT_OPTIONS to avoid compatibility issues with\nscsi_debug.\n\nSigned-off-by: Disha Goel \u003cdisgoel@linux.ibm.com\u003e\nReviewed-by: Zorro Lang \u003czlang@redhat.com\u003e\nReviewed-by: Ojaswin Mujoo \u003cojaswin@linux.ibm.com\u003e\nSigned-off-by: Zorro Lang \u003czlang@kernel.org\u003e\n" }, { "commit": "29dbf07039172295ebe3ae158eb1a7e44dedcc96", "tree": "7f33d1cad6394232fdf28851eefc768d6e306eb5", "parents": [ "c34c06b9728c2b3e3dbe7666de29f0fb57f24f33" ], "author": { "name": "Long Li", "email": "leo.lilong@huawei.com", "time": "Wed Apr 01 06:06:32 2026" }, "committer": { "name": "Zorro Lang", "email": "zlang@kernel.org", "time": "Wed Apr 01 15:07:08 2026" }, "message": "xfs/018: test log attr replay with zero-length attr value\n\nKernel commit d72f2084e309 (\"xfs: fix ri_total validation in\nxlog_recover_attri_commit_pass2\") fixed a bug where setting a\nzero-length attribute value would cause incorrect ri_total\nvalidation during log recovery.\n\nAdd test cases to xfs/018 to cover this scenario across attr\nfork formats: extent and remote.\n\nUsing echo -n \"\" | attr -s is unreliable for empty values since\nattr may treat stdin EOF as no value provided. Instead, handle\nzero-length values explicitly in test_attr_replay() by using\nattr -s -V \"\" when the value is empty.\n\nSigned-off-by: Long Li \u003cleo.lilong@huawei.com\u003e\nReviewed-by: Darrick J. Wong \u003cdjwong@kernel.org\u003e\nSigned-off-by: Zorro Lang \u003czlang@kernel.org\u003e\n" }, { "commit": "c34c06b9728c2b3e3dbe7666de29f0fb57f24f33", "tree": "6b007f35c20e9f1c8e08f3a1c302412b062a564a", "parents": [ "8deafb5c487e73c9384df2fcbab523b12007875b" ], "author": { "name": "Filipe Manana", "email": "fdmanana@suse.com", "time": "Tue Mar 31 14:15:10 2026" }, "committer": { "name": "Zorro Lang", "email": "zlang@kernel.org", "time": "Wed Apr 01 15:07:08 2026" }, "message": "fstests: add kernel commit IDs for patches already merged upstream\n\nSeveral tests mention patches that are already in Linus\u0027 git tree, so add\ntheir commit IDs. For generic/762, also use \"fixed_by_fs_commit xfs\"\ninstead of \"_fixed_by_git_commit kernel\" since it\u0027s a XFS specific commit\nand it does not make sense for the test to report that commit in case it\nfails when running against a filesystem other than XFS.\n\nSigned-off-by: Filipe Manana \u003cfdmanana@suse.com\u003e\nReviewed-by: Zorro Lang \u003czlang@redhat.com\u003e\nSigned-off-by: Zorro Lang \u003czlang@kernel.org\u003e\n" } ], "next": "8deafb5c487e73c9384df2fcbab523b12007875b" }