On Tue, Sep 19, 2023 at 12:44:49AM +0800, Zorro Lang wrote: > On Mon, Sep 18, 2023 at 08:17:56AM -0700, Darrick J. Wong wrote: > > On Mon, Sep 18, 2023 at 12:44:40AM +0800, Zorro Lang wrote: > > > Hi Darrick, > > > > > > Hi Darrick, > > > > > > Do these two changes cover new xfs bug fix? I hit xfs/270 fails on > > > latest upstream mainline kernel testing, e.g. [1][2]. But the xfs/270 > > > always test passed on my side before. > > > > Yes, the changes in xfs/270 reflect the corrections to the ro mounting > > behavior in this fix series: > > https://lore.kernel.org/linux-xfs/169454023347.3411463.3256623887249978768.stg-ugh@frogsfrogsfrogs/T/#u > > > > The ro mount behavior regressed some 7-8 years ago; this test was > > written to test the broken fix; and we didn't notice any of this until > > 6.1 added the incore unlinked list. xfs/270 didn't check the usability > > of the mounted ro filesystem, so nobody noticed that the mount failed > > when it should have succeeded. > > > > (That's the first + line) > > > > The second diff here checks what happens when you try to ro mount with a > > dirty log. > > Thanks Darrick, > > With the "[PATCH 1/2] xfs/270: actually test file readability", the x/270 only > fails as [1], but with the "[PATCH 2/2] xfs/270: actually test log recovery with > unknown rocompat feature", I hit one more kernel assertion failure as [2]. > > Are these two kind of failures all as known, and will be fixed by : > [GIT PULL 3/8] xfs: fix ro mounting with unknown rocompat features > ? Yes. That second assertion is log recovery complaining that the fs shut down during recovery; and the reason the log shut down during recovery is that someone tried to write a superblock with unknown rocompat bits and triggered the verifier. The verifier failure is addressed by "xfs: fix log recovery when unknown rocompat bits are set". --D > Thanks, > Zorro > > > > > > --D > > > > > Thanks, > > > Zorro > > > > > > [1] > > > --- /dev/fd/63 2023-09-16 20:33:38.373115157 -0400 > > > +++ xfs/270.out.bad 2023-09-16 20:33:37.959115275 -0400 > > > @@ -1,6 +1,9 @@ > > > QA output created by 270 > > > rw mount test > > > ro mount test > > > +cat: /mnt/xfstests/scratch/testfile: Input/output error > > > rw remount test > > > rw mount test > > > ro mount test > > > +ro mount test failed > > > +(see /var/lib/xfstests/results//xfs/270.full for details) > > > > > > [2] > > > [57829.436755] run fstests xfs/270 at 2023-09-16 20:33:31 > > > [57832.373519] XFS (vda3): Mounting V5 Filesystem b31f9652-417d-4dd2-a5ed-54ee957fbfcb > > > [57832.398699] XFS (vda3): Ending clean mount > > > [57832.551895] XFS (vda3): Unmounting Filesystem b31f9652-417d-4dd2-a5ed-54ee957fbfcb > > > [57832.909487] XFS (vda3): Mounting V5 Filesystem 08403da7-afbe-4bfa-a45c-487e28280acc > > > [57832.934603] XFS (vda3): Ending clean mount > > > [57832.963287] XFS (vda3): User initiated shutdown received. > > > [57832.967959] XFS (vda3): Metadata I/O Error (0x4) detected at xfs_fs_goingdown+0x4b/0x180 [xfs] (fs/xfs/xfs_fsops.c:492). Shutting down filesystem. > > > [57832.974742] XFS (vda3): Please unmount the filesystem and rectify the problem(s) > > > [57833.002720] XFS (vda3): Unmounting Filesystem 08403da7-afbe-4bfa-a45c-487e28280acc > > > [57833.376473] XFS (vda3): Mounting V5 Filesystem 801b25ca-acc4-4232-9955-dc3448680467 > > > [57833.403380] XFS (vda3): Ending clean mount > > > [57833.443719] XFS (vda3): Unmounting Filesystem 801b25ca-acc4-4232-9955-dc3448680467 > > > [57833.516114] [TTM] Buffer eviction failed > > > [57833.519490] qxl 0000:00:02.0: object_init failed for (3149824, 0x00000001) > > > [57833.523184] [drm:qxl_alloc_bo_reserved [qxl]] *ERROR* failed to allocate VRAM BO > > > [57833.955746] XFS (vda3): Superblock has unknown read-only compatible features (0x80000000) enabled. > > > [57833.960383] XFS (vda3): Attempted to mount read-only compatible filesystem read-write. > > > [57833.964353] XFS (vda3): Filesystem can only be safely mounted read only. > > > [57833.967997] XFS (vda3): SB validate failed with error -22. > > > [57834.039637] XFS (vda3): Superblock has unknown read-only compatible features (0x80000000) enabled. > > > [57834.046289] XFS (vda3): Mounting V5 Filesystem 801b25ca-acc4-4232-9955-dc3448680467 > > > [57834.075451] XFS (vda3): Ending clean mount > > > [57834.082359] XFS (vda3): Corruption detected in superblock read-only compatible features (0x80000000)! > > > [57834.086487] XFS (vda3): Metadata corruption detected at xfs_sb_write_verify+0x11d/0x380 [xfs], xfs_sb block 0x0 > > > [57834.090716] XFS (vda3): Unmount and run xfs_repair > > > [57834.093708] XFS (vda3): First 128 bytes of corrupted metadata buffer: > > > [57834.096953] 00000000: 58 46 53 42 00 00 10 00 00 00 00 00 00 3c 00 00 XFSB.........<.. > > > [57834.100635] 00000010: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ................ > > > [57834.104330] 00000020: 80 1b 25 ca ac c4 42 32 99 55 dc 34 48 68 04 67 ..%...B2.U.4Hh.g > > > [57834.108273] 00000030: 00 00 00 00 00 20 00 06 00 00 00 00 00 00 00 80 ..... .......... > > > [57834.112176] 00000040: 00 00 00 00 00 00 00 81 00 00 00 00 00 00 00 82 ................ > > > [57834.116043] 00000050: 00 00 00 01 00 0f 00 00 00 00 00 04 00 00 00 00 ................ > > > [57834.119693] 00000060: 00 00 40 00 b4 a5 02 00 02 00 00 08 00 00 00 00 ..@............. > > > [57834.123459] 00000070: 00 00 00 00 00 00 00 00 0c 09 09 03 14 00 00 19 ................ > > > [57834.127159] XFS (vda3): Corruption of in-memory data (0x8) detected at _xfs_buf_ioapply+0x50b/0x5f0 [xfs] (fs/xfs/xfs_buf.c:1558). Shutting down filesystem. > > > [57834.134557] XFS (vda3): Please unmount the filesystem and rectify the problem(s) > > > [57834.253491] XFS (vda3): ro->rw transition prohibited on unknown (0x80000000) ro-compat filesystem > > > [57834.281980] XFS (vda3): Unmounting Filesystem 801b25ca-acc4-4232-9955-dc3448680467 > > > [57834.655120] XFS (vda3): Mounting V5 Filesystem 29902d21-a146-4095-8747-562850110ff6 > > > [57834.682299] XFS (vda3): Ending clean mount > > > [57834.723067] XFS (vda3): User initiated shutdown received. > > > [57834.730954] XFS (vda3): Metadata I/O Error (0x4) detected at xfs_fs_goingdown+0x4b/0x180 [xfs] (fs/xfs/xfs_fsops.c:492). Shutting down filesystem. > > > [57834.738710] XFS (vda3): Please unmount the filesystem and rectify the problem(s) > > > [57834.768607] XFS (vda3): Unmounting Filesystem 29902d21-a146-4095-8747-562850110ff6 > > > [57835.270472] XFS (vda3): Superblock has unknown read-only compatible features (0x80000000) enabled. > > > [57835.275655] XFS (vda3): Attempted to mount read-only compatible filesystem read-write. > > > [57835.279434] XFS (vda3): Filesystem can only be safely mounted read only. > > > [57835.283124] XFS (vda3): SB validate failed with error -22. > > > [57835.357336] XFS (vda3): Superblock has unknown read-only compatible features (0x80000000) enabled. > > > [57835.369968] XFS (vda3): Mounting V5 Filesystem 29902d21-a146-4095-8747-562850110ff6 > > > [57835.398595] XFS (vda3): Starting recovery (logdev: internal) > > > [57835.410361] XFS (vda3): Superblock has unknown read-only compatible features (0x80000000) enabled. > > > [57835.414593] XFS (vda3): Attempted to mount read-only compatible filesystem read-write. > > > [57835.418495] XFS (vda3): Filesystem can only be safely mounted read only. > > > [57835.422152] XFS (vda3): metadata I/O error in "xlog_do_recover+0x3e8/0x540 [xfs]" at daddr 0x0 len 1 error 22 > > > [57835.426851] XFS: Assertion failed: 0, file: fs/xfs/xfs_log_recover.c, line: 3321 > > > [57835.430765] ------------[ cut here ]------------ > > > [57835.434014] WARNING: CPU: 2 PID: 3426232 at fs/xfs/xfs_message.c:104 assfail+0x54/0x70 [xfs] > > > [57835.438380] Modules linked in: ext2 overlay dm_zero dm_log_writes dm_thin_pool dm_persistent_data dm_bio_prison sd_mod t10_pi sg dm_snapshot dm_bufio ext4 mbcache jbd2 loop dm_flakey dm_mod tls rfkill intel_rapl_msr sunrpc intel_rapl_common intel_uncore_frequency_common isst_if_common nfit snd_hda_codec_generic ledtrig_audio kvm_intel snd_hda_intel snd_intel_dspcfg snd_intel_sdw_acpi snd_hda_codec kvm snd_hda_core snd_hwdep irqbypass snd_seq snd_seq_device snd_pcm qxl snd_timer drm_ttm_helper ttm pcspkr virtio_balloon snd drm_kms_helper soundcore i2c_piix4 joydev drm fuse xfs libcrc32c ata_generic crct10dif_pclmul crc32_pclmul crc32c_intel ghash_clmulni_intel ata_piix virtio_console virtio_net net_failover failover virtio_blk libata serio_raw [last unloaded: scsi_debug] > > > [57835.467794] CPU: 2 PID: 3426232 Comm: mount Kdump: loaded Tainted: G W 6.6.0-rc1+ #1 > > > [57835.472233] Hardware name: Red Hat KVM, BIOS 0.5.1 01/01/2011 > > > [57835.475946] RIP: 0010:assfail+0x54/0x70 [xfs] > > > [57835.479708] Code: c0 48 ba 00 00 00 00 00 fc ff df 48 89 c1 83 e0 07 48 c1 e9 03 0f b6 14 11 38 c2 7f 04 84 d2 75 10 80 3d 6e 3f 1c 00 00 75 15 <0f> 0b c3 cc cc cc cc 48 c7 c7 10 47 99 c0 e8 79 fd 34 df eb e2 0f > > > [57835.489526] RSP: 0018:ffffc9000293f918 EFLAGS: 00010246 > > > [57835.493381] RAX: 0000000000000000 RBX: ffff88817b73a000 RCX: 1ffffffff81328e2 > > > [57835.497512] RDX: 0000000000000004 RSI: ffffc9000293f680 RDI: ffffffffc0afd780 > > > [57835.501588] RBP: ffff8881964ea000 R08: 00000000ffffffea R09: fffff52000527ec1 > > > [57835.505701] R10: ffffc9000293f60f R11: fffffffffff6c2b0 R12: 00000000ffffffea > > > [57835.509789] R13: ffff88817b73a0e0 R14: ffff8881b927ba80 R15: > > > > > > > > > On Tue, Aug 29, 2023 at 04:09:48PM -0700, Darrick J. Wong wrote: > > > > Hi all, > > > > > > > > Dave pointed out some failures in xfs/270 when he upgraded Debian > > > > unstable and util-linux started using the new mount apis. Upon further > > > > inquiry I noticed that XFS is quite a hot mess when it encounters a > > > > filesystem with unrecognized rocompat bits set in the superblock. > > > > > > > > Whereas we used to allow readonly mounts under these conditions, a > > > > change to the sb write verifier several years ago resulted in the > > > > filesystem going down immediately because the post-mount log cleaning > > > > writes the superblock, which trips the sb write verifier on the > > > > unrecognized rocompat bit. I made the observation that the ROCOMPAT > > > > features RMAPBT and REFLINK both protect new log intent item types, > > > > which means that we actually cannot support recovering the log if we > > > > don't recognize all the rocompat bits. > > > > > > > > Therefore -- fix inode inactivation to work when we're recovering the > > > > log, disallow recovery when there's unrecognized rocompat bits, and > > > > don't clean the log if doing so would trip the rocompat checks. > > > > > > > > v2: change direction of series to allow log recovery on ro mounts > > > > > > > > If you're going to start using this code, I strongly recommend pulling > > > > from my git trees, which are linked below. > > > > > > > > This has been running on the djcloud for months with no problems. Enjoy! > > > > Comments and questions are, as always, welcome. > > > > > > > > --D > > > > > > > > fstests git tree: > > > > https://git.kernel.org/cgit/linux/kernel/git/djwong/xfstests-dev.git/log/?h=fix-ro-mounts > > > > --- > > > > tests/xfs/270 | 80 ++++++++++++++++++++++++++++++++++++++--------------- > > > > tests/xfs/270.out | 2 + > > > > 2 files changed, 59 insertions(+), 23 deletions(-) > > > > > > > > > >
