Corrupted XFS journal can cause a kernel crash

This document (7017195) is provided subject to the disclaimer at the end of this document.

Environment

SUSE Linux Enterprise Server 12 Service Pack 1 (SLES 12 SP1)

Situation

When installing SLES 12 SP1, a storage disk that is accessible to the installer and having a corrupted XFS journal can cause the kernel to crash with the following call trace:


| [c0000001503e3500] [c000000000256b84] cache_alloc_refill+0x104/0x480 (unreliable)
| [c0000001503e35a0] [c000000000258bbc] __kmalloc+0x25c/0x2b0
| [c0000001503e3600] [d000000009405ffc] kmem_alloc+0x9c/0x130 [xfs]
| [c0000001503e3680] [d000000009451dec] xlog_recover_add_to_trans+0x6c/0x2c0 [xfs]
| [c0000001503e3710] [d000000009458084] xlog_recover_process_data+0x2f4/0x3c0 [xfs]
| [c0000001503e37e0] [d00000000945856c] xlog_do_recovery_pass+0x41c/0x5c0 [xfs]
| [c0000001503e3900] [d0000000094587b8] xlog_do_log_recovery+0xa8/0x120 [xfs]
| [c0000001503e3940] [d000000009458874] xlog_do_recover+0x44/0x180 [xfs]
| [c0000001503e3990] [d000000009458a5c] xlog_recover+0xac/0x120 [xfs]
| [c0000001503e3a10] [d00000000945e9fc] xfs_log_mount+0x10c/0x2e0 [xfs]
| [c0000001503e3a90] [d0000000093fd608] xfs_mountfs+0x428/0x7e0 [xfs]
| [c0000001503e3b30] [d000000009402044] xfs_fs_fill_super+0x344/0x410 [xfs]
| [c0000001503e3b90] [c0000000002825d4] mount_bdev+0x234/0x270
| [c0000001503e3c10] [d0000000093ff758] xfs_fs_mount+0x48/0x60 [xfs]
| [c0000001503e3c50] [c00000000028359c] mount_fs+0x8c/0x230
| [c0000001503e3cf0] [c0000000002aa368] vfs_kern_mount+0x78/0x140
| [c0000001503e3d40] [c0000000002ad9b8] do_mount+0x2f8/0xc60
| [c0000001503e3dd0] [c0000000002ae3dc] SyS_mount+0xbc/0x140
| [c0000001503e3e30] [c00000000000a17c] syscall_exit+0x0/0x7c

The crash can also be observed when a storage disk with a corrupt XFS journal is being mounted on a machine with SLES 12 SP1 already installed.

Resolution

This issue is resolved by installing SLES12 SP1 kernel maintenance update version 3.12.53 or greater.

To workaround this issue during an installation of SLES 12 SP1, the disk with a corrupt XFS journal must be excluded. The disk can then be added to the system after the installation is done and the kernel has been updated to version 3.12.53 or greater.

Disclaimer

This Support Knowledgebase provides a valuable tool for SUSE customers and parties interested in our products and solutions to acquire information, ideas and learn from one another. Materials are provided for informational, personal or non-commercial use within your organization and are presented "AS IS" WITHOUT WARRANTY OF ANY KIND.