A softlockup panic may occur when shrinking a VxFS filesystem on InfoScale 7.4.2/sles12

Problem

A softlockup panic may occur when shrinking a VxFS filesystem on InfoScale 7.4.2/sles12

Error Message

[259267.374443] Kernel panic - not syncing: softlockup: hung tasks
[259267.374446] CPU: 2 PID: 78989 Comm: fsadm Tainted: P           OEL     4.12.14-120-default #1 SLE12-SP5
[259267.374447] Hardware name: VMware, Inc. VMware Virtual Platform/440BX Desktop Reference Platform, BIOS 6.00 11/12/2020
[259267.374447] Call Trace:
[259267.374450]  <IRQ>
[259267.374454]  dump_stack+0x5a/0x75
[259267.374458]  panic+0xdb/0x23e
[259267.374463]  watchdog_timer_fn+0x208/0x210
[259267.374464]  ? watchdog_park_threads+0x70/0x70
[259267.374469]  __hrtimer_run_queues+0xce/0x210
[259267.374471]  hrtimer_interrupt+0x99/0x190
[259267.374476]  smp_apic_timer_interrupt+0x3a/0x60
[259267.374478]  apic_timer_interrupt+0x8f/0xa0
[259267.374479]  </IRQ>
[259267.374501] RIP: 0010:vx_getblk_bp+0x5/0x80 [vxfs]
[259267.374502] RSP: 0018:ffffb39914a576e0 EFLAGS: 00000202 ORIG_RAX: ffffffffffffff13
[259267.374503] RAX: ffff96f323531400 RBX: 000000000000d05d RCX: ffffb39914a57708
[259267.374504] RDX: 0000000000000400 RSI: 00000001a80d7f46 RDI: ffff96f385ccec00
[259267.374505] RBP: ffff96f296ab2000 R08: 0000000000000022 R09: 0000000000000004
[259267.374506] R10: 0000000000000246 R11: 000010000000f050 R12: 00000000d406bfa3
[259267.374507] R13: ffff96f385ccec00 R14: 0000000000000001 R15: 0000000000000400
[259267.374528]  ? vx_multi_bufinval+0x21d/0x230 [vxfs]
[259267.374564]  ? vx_reorg_start_io+0x32e/0x3d0 [vxfs]
[259267.374598]  ? vx_reorg_copy+0x235/0x4c0 [vxfs]
[259267.374630]  ? vx_reorg_dostruct+0x3b6/0x960 [vxfs]
[259267.374663]  ? vx_trancommit+0x32f/0x1220 [vxfs]
[259267.374695]  ? vx_extmap_reorg+0xdd2/0xe90 [vxfs]
[259267.374734]  ? vx_ilock+0x18/0x50 [vxfs]
[259267.374767]  ? vx_struct_reorg+0xb5f/0xb90 [vxfs]
[259267.374768]  ? __switch_to_asm+0x40/0x70
[259267.374806]  ? vx_aioctl_full+0x107d/0x1160 [vxfs]
[259267.374825]  ? vx_aioctl_common+0x1ba9/0x2410 [vxfs]
[259267.374827]  ? generic_file_read_iter+0xce/0xaf0
[259267.374845]  ? vx_aioctl+0x10f/0x130 [vxfs]
[259267.374882]  ? vx_unlocked_ioctl+0x224/0x2e0 [vxfs]
[259267.374884]  ? do_vfs_ioctl+0x92/0x5e0
[259267.374886]  ? SyS_ioctl+0x74/0x80
[259267.374888]  ? do_syscall_64+0x74/0x160
[259267.374889]  ? entry_SYSCALL_64_after_hwframe+0x3d/0xa2

Cause

The softlockup was due to e4188107 whereby the area of the VxFS filesystem being shrunk held part of the extent map file which needed to be reorganized.
 

Solution

The Product Engineering team currently plans to address this issue through a patch or hotfix in the current software version. Please note that our company reserves the right to withdraw any fix from the targeted release if it fails quality assurance tests. Development plans are subject to change, and any actions you take based on this information, or your reliance on it, are at your own risk.

A hotfix for VxFS 7.4.2/sles12 is available and the fix has been included in the InfoScale 7.4.2update 8/sles12 patch (  https://www.veritas.com/content/support/en_US/downloads/update.UPD184896 ).

As some sles15 kernels have PREEMPT_DYNAMIC enabled, the likelihood of the softlockup occurring on sles15 is reduced. However, a hf is available for VxFS 7.4.2/sles15 and VxFS 8.0.2/sles15 to fix the VxFS code and avoid the risk of a potential softlockup. The fix will also be included in the next GA patches for InfoScale 8.0, 8.0.2 and 9.0 on sles15.

Please contact Technical Support should the hotfix be required.