https://bugzilla.kernel.org/show_bug.cgi?id=216419 --- Comment #1 from Darrick J. Wong (djwong@xxxxxxxxxx) --- On Sat, Aug 27, 2022 at 04:50:27AM +0000, bugzilla-daemon@xxxxxxxxxx wrote: > https://bugzilla.kernel.org/show_bug.cgi?id=216419 > > Bug ID: 216419 > Summary: Internal error XFS_WANT_CORRUPTED_RETURN at line 442 > of file fs/xfs/libxfs/xfs_alloc.c > Product: File System > Version: 2.5 > Kernel Version: 5.3.18 > Hardware: All > OS: Linux > Tree: Mainline > Status: NEW > Severity: normal > Priority: P1 > Component: XFS > Assignee: filesystem_xfs@xxxxxxxxxxxxxxxxxxxxxx > Reporter: sunjunchao2870@xxxxxxxxx > Regression: No > > Created attachment 301678 > --> https://bugzilla.kernel.org/attachment.cgi?id=301678&action=edit > messages about corrupted xfs > > Recently I ran pressure test on xfs(almost 1.5GB/s bandwidth on write) and I > discovered a lot of error logs about xfs, just like: > Aug 21 20:31:20 rg21-oss005 kernel: XFS (bcache4): metadata I/O error in > "xfs_trans_read_buf_map" at daddr 0x8804a66e8 len 8 error 74 > Aug 21 20:31:20 rg21-oss005 kernel: XFS (bcache4): Metadata CRC error > detected > at xfs_dir3_block_read_verify+0xbf/0xe0 [xfs], xfs_dir3_block block > 0x8804a66e8 > Aug 21 20:31:20 rg21-oss005 kernel: XFS (bcache4): Unmount and run xfs_repair > Aug 21 20:31:20 rg21-oss005 kernel: XFS (bcache4): First 128 bytes of > corrupted > metadata buffer: > Aug 21 20:31:20 rg21-oss005 kernel: 00000000: 38 30 30 2e 30 2c 20 38 30 30 > 2e > 30 2c 20 38 30 800.0, 800.0, 80 > Aug 21 20:31:20 rg21-oss005 kernel: 00000010: 30 2e 30 2c 20 31 34 34 30 2e > 30 > 2c 20 31 34 34 0.0, 1440.0, 144 > Aug 21 20:31:20 rg21-oss005 kernel: 00000020: 30 2e 30 2c 20 32 30 32 39 2e > 39 > 33 37 35 2c 20 0.0, 2029.9375, > Aug 21 20:31:20 rg21-oss005 kernel: 00000030: 32 30 32 39 2e 39 33 37 35 2c > 20 > 32 30 32 39 2e 2029.9375, 2029. > Aug 21 20:31:20 rg21-oss005 kernel: 00000040: 39 33 37 35 2c 20 32 30 32 39 > 2e > 39 33 37 35 5d 9375, 2029.9375] > Aug 21 20:31:20 rg21-oss005 kernel: 00000050: 2c 20 22 65 6c 61 70 73 65 64 > 22 > 3a 20 5b 31 34 , "elapsed": [14 > Aug 21 20:31:20 rg21-oss005 kernel: 00000060: 36 35 2e 31 30 35 31 30 34 34 > 34 > 36 34 31 31 31 65.1051044464111 > Aug 21 20:31:20 rg21-oss005 kernel: 00000070: 2c 20 31 34 36 37 2e 34 37 33 > 33 > 31 36 31 39 32 , 1467.473316192 > Aug 21 20:31:20 rg21-oss005 kernel: XFS (bcache4): metadata I/O error in > "xfs_trans_read_buf_map" at daddr 0x8804a66e8 len 8 error 74 > > and > > Aug 21 22:41:37 rg21-oss005 kernel: XFS (bcache4): Internal error > XFS_WANT_CORRUPTED_RETURN at line 442 of file fs/xfs/libxfs/xfs_alloc.c. > Caller xfs_alloc_ag_vextent_near+0x7f6/0xb60 [xfs] > Aug 21 22:41:37 rg21-oss005 kernel: CPU: 23 PID: 10127 Comm: w16-5 Kdump: > loaded Tainted: G OE N 5.3.18.20211010 #4 SLE15-SP2 > (unreleased) Please contact your distribution partner for support... ^^^^^^^^^ --D > Aug 21 22:41:37 rg21-oss005 kernel: Hardware name: New H3C Technologies Co., > Ltd. UniServer R4300 G3/RS01M2C7S, BIOS 2.00.56 12/17/2021 > Aug 21 22:41:37 rg21-oss005 kernel: Call Trace: > Aug 21 22:41:37 rg21-oss005 kernel: dump_stack+0x64/0x83 > Aug 21 22:41:37 rg21-oss005 kernel: xfs_alloc_fixup_trees+0x211/0x350 [xfs] > Aug 21 22:41:37 rg21-oss005 kernel: xfs_alloc_ag_vextent_near+0x7f6/0xb60 > [xfs] > Aug 21 22:41:37 rg21-oss005 kernel: xfs_alloc_ag_vextent+0x13c/0x150 [xfs] > Aug 21 22:41:37 rg21-oss005 kernel: xfs_alloc_vextent+0x41f/0x5a0 [xfs] > Aug 21 22:41:37 rg21-oss005 kernel: xfs_bmap_btalloc+0x23c/0x8e0 [xfs] > Aug 21 22:41:37 rg21-oss005 kernel: xfs_bmapi_allocate+0x10c/0x2e0 [xfs] > Aug 21 22:41:37 rg21-oss005 kernel: xfs_bmapi_convert_delalloc+0x267/0x4a0 > [xfs] > Aug 21 22:41:37 rg21-oss005 kernel: xfs_map_blocks+0x195/0x3f0 [xfs] > Aug 21 22:41:37 rg21-oss005 kernel: xfs_do_writepage+0x127/0x3c0 [xfs] > Aug 21 22:41:37 rg21-oss005 kernel: write_cache_pages+0x1d3/0x410 > Aug 21 22:41:37 rg21-oss005 kernel: ? xfs_vm_writepages+0xa0/0xa0 [xfs] > Aug 21 22:41:37 rg21-oss005 kernel: ? pagevec_lookup_entries+0x1a/0x30 > Aug 21 22:41:37 rg21-oss005 kernel: xfs_vm_writepages+0x64/0xa0 [xfs] > Aug 21 22:41:37 rg21-oss005 kernel: do_writepages+0x1a/0x70 > Aug 21 22:41:37 rg21-oss005 kernel: __filemap_fdatawrite_range+0xcf/0x100 > Aug 21 22:41:37 rg21-oss005 kernel: filemap_write_and_wait_range+0x41/0xa0 > Aug 21 22:41:37 rg21-oss005 kernel: xfs_setattr_size+0x10d/0x390 [xfs] > Aug 21 22:41:37 rg21-oss005 kernel: xfs_vn_setattr+0x6e/0xc0 [xfs] > Aug 21 22:41:37 rg21-oss005 kernel: notify_change+0x26c/0x450 > Aug 21 22:41:37 rg21-oss005 kernel: do_truncate+0x72/0xc0 > Aug 21 22:41:37 rg21-oss005 kernel: do_sys_ftruncate+0x10c/0x120 > Aug 21 22:41:37 rg21-oss005 kernel: do_syscall_64+0x5b/0x1f0 > Aug 21 22:41:37 rg21-oss005 kernel: entry_SYSCALL_64_after_hwframe+0x44/0xa9 > Aug 21 22:41:37 rg21-oss005 kernel: RIP: 0033:0x7fb5168a86e7 > Aug 21 22:41:37 rg21-oss005 kernel: Code: 73 01 c3 48 8b 0d 89 f7 2c 00 f7 d8 > 64 89 01 48 83 c8 ff c3 66 2e 0f 1f 84 00 00 00 00 00 0f 1f 44 00 00 b8 4d 00 > 00 00 0f 05 <48> 3d 01 f0 ff ff 73 01 c3 48 8b 0d 59 f7 2c 00 f7 d8 64 89 01 > 48 > Aug 21 22:41:37 rg21-oss005 kernel: RSP: 002b:00007fb237bfb498 EFLAGS: > 00000206 > ORIG_RAX: 000000000000004d > Aug 21 22:41:37 rg21-oss005 kernel: RAX: ffffffffffffffda RBX: > 00007fb2200038c0 > RCX: 00007fb5168a86e7 > Aug 21 22:41:37 rg21-oss005 kernel: RDX: 00000000000000e0 RSI: > 0000000002300000 > RDI: 0000000000000074 > Aug 21 22:41:37 rg21-oss005 kernel: RBP: 00007fb237bfc6b0 R08: > 00007fb2360f6050 > R09: 0000000000004000 > Aug 21 22:41:37 rg21-oss005 kernel: R10: 00000000022e4000 R11: > 0000000000000206 > R12: 0000000000100000 > Aug 21 22:41:37 rg21-oss005 kernel: R13: 0000000000001000 R14: > 0000000000000074 > R15: 00007fb237bfb680 > Aug 21 22:41:37 rg21-oss005 kernel: XFS (bcache4): page discard on page > 00000000d976f7c9, inode 0x400079d77, offset 28315648. > Aug 21 22:41:37 rg21-oss005 kernel: XFS (bcache4): writeback error on sector > 42005378320 > > > > It seems xfs corrupted, and I repaired that by xfs_repair. But after one day, > I > got some similar questions and xfs corrupted again and again. Is this a known > issue? Or am I wrong in something? Please let me known if any information is > missed. Thanks! > > -- > You may reply to this email to add a comment. > > You are receiving this mail because: > You are watching the assignee of the bug. -- You may reply to this email to add a comment. You are receiving this mail because: You are watching the assignee of the bug.
