System crash during a BTRFS maintenance task

This document (000019638) is provided subject to the disclaimer at the end of this document.

Environment

SUSE Linux Enterprise Server 12 SP4 (SLES12 SP4) with kernel 4.12.14-95.48.1 or older.
SUSE Linux Enterprise Server 12 SP5 (SLES12 SP5) with kernel 4.12.14-122.17.1 or older.

Situation

During a BTRFS maintenance a kernel panic may happen.
If a kernel dump has been saved, the dmesg.txt in the crash directory (usually /var/crash/<date-time-of-crash>) contains at the bottom of the file messages similar to this one:

kernel BUG at ../fs/btrfs/relocation.c:4837! 
invalid opcode: 0000 [#1] SMP PTI
CPU: 1 PID: 30984 Comm: btrfs Not tainted 4.12.14-122.12-default #1 SLE12-SP5
Hardware name: VMware, Inc. VMware Virtual Platform/440BX Desktop Reference Platform, BIOS 6.00 09/17/2015
task: ffff939fa65a8840 task.stack: ffffa22307570000
RIP: 0010:btrfs_reloc_cow_block+0x29b/0x2c0 [btrfs]
RSP: 0018:ffffa22307573818 EFLAGS: 00010206
RAX: 0000000000000001 RBX: ffff93a3c8e42800 RCX: ffff939efd030040
RDX: 0000000000000001 RSI: ffff939f69fb0180 RDI: 000000454fd14000
RBP: ffff93a3c5c41000 R08: 0000000000000000 R09: 0000000000000004
R10: ffff939f1cbc3360 R11: 0000000000000002 R12: ffff939f62e8d7e8
R13: 0000000000000001 R14: ffff939f19593e10 R15: ffff93a3e0920001
FS:  00007f47f51cd8c0(0000) GS:ffff93a3e6e40000(0000) knlGS:0000000000000000
CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: 00000000043c5000 CR3: 00000004cc738000 CR4: 00000000000406e0
Call Trace:
 ? update_ref_for_cow+0x277/0x300 [btrfs]
 __btrfs_cow_block+0x330/0x5b0 [btrfs]
 btrfs_cow_block+0xe4/0x200 [btrfs]
 btrfs_search_slot+0x4b2/0x980 [btrfs]
 do_relocation+0x468/0x5d0 [btrfs]
 relocate_tree_blocks+0x55f/0x5b0 [btrfs]
 ? tree_insert+0x49/0x50 [btrfs]
 relocate_block_group+0x27f/0x6f0 [btrfs]
 btrfs_relocate_block_group+0x189/0x270 [btrfs]
 btrfs_relocate_chunk+0x49/0xf0 [btrfs]
 __btrfs_balance+0x7f0/0xb90 [btrfs]
 btrfs_balance+0x40f/0x690 [btrfs]
 btrfs_ioctl_balance+0x3e9/0x530 [btrfs]
 btrfs_ioctl+0x82b/0x2150 [btrfs]
 ? __handle_mm_fault+0xc7e/0xf40
 do_vfs_ioctl+0x92/0x5e0
 ? __audit_syscall_entry+0xac/0xf0
 ? syscall_trace_enter+0x1ac/0x290
  SyS_ioctl+0x74/0x80
 do_syscall_64+0x74/0x160
  entry_SYSCALL_64_after_hwframe+0x3d/0xa2
RIP: 0033:0x7f47f4253407
RSP: 002b:00007ffda35aaae8 EFLAGS: 00000206 ORIG_RAX: 0000000000000010
RAX: ffffffffffffffda RBX: 0000000000000004 RCX: 00007f47f4253407
RDX: 00007ffda35aab90 RSI: 00000000c4009420 RDI: 0000000000000004
RBP: 00007ffda35aab90 R08: 0000000000000000 R09: 0000000000000000
R10: 000000000fa99fa0 R11: 0000000000000206 R12: 00007ffda35ab4f8
R13: 00007ffda35ab973 R14: 0000000000000001 R15: 0000000000000001




The line number at the following message may vary, depending on the kernel version:
kernel BUG at ../fs/btrfs/relocation.c:4837!

Resolution

This is a known problem and has been fixed with the following kernel release:

SLES12 SP4 - 4.12.14-95.51.1
SLES12 SP5 - 4.12.14-122.20.1

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.