[LU-12374] client went down w/ panic during lustre_rmmod

• Type: Bug
• Resolution: Fixed
• Priority: Minor
• Fix Version/s: Lustre 2.13.0, Lustre 2.12.3
• Severity: 3
• Assignee: Wang Shilong (Inactive)
• Reporter: Wang Shilong (Inactive)

Description

client went down with panic during lustre_rmmod.

crash> bt
PID: 19 TASK: ffff8e8a5b0b30c0 CPU: 22 COMMAND: "rcuos/1"
#0 [ffff8e8a5b0cbaa0] machine_kexec at ffffffffa5263674
#1 [ffff8e8a5b0cbb00] __crash_kexec at ffffffffa531ce12
#2 [ffff8e8a5b0cbbd0] crash_kexec at ffffffffa531cf00
#3 [ffff8e8a5b0cbbe8] oops_end at ffffffffa596c758
#4 [ffff8e8a5b0cbc10] no_context at ffffffffa595aa7e
#5 [ffff8e8a5b0cbc60] __bad_area_nosemaphore at ffffffffa595ab15
#6 [ffff8e8a5b0cbcb0] bad_area_nosemaphore at ffffffffa595ac86
#7 [ffff8e8a5b0cbcc0] __do_page_fault at ffffffffa596f6b0
#8 [ffff8e8a5b0cbd30] do_page_fault at ffffffffa596f915
#9 [ffff8e8a5b0cbd60] page_fault at ffffffffa596b758
[exception RIP: unknown or invalid address]
RIP: ffffffffc13157a0 RSP: ffff8e8a5b0cbe10 RFLAGS: 00010292
RAX: ffff8e8a5b0cbfd8 RBX: ffff8eb8d8d637a8 RCX: dead000000000200
RDX: ffffffffc13157a0 RSI: 0000000000000297 RDI: ffff8eb8d8d637a8
RBP: ffff8e8a5b0cbec0 R8: ffff8e8a5b0cbe80 R9: 0000000000000000
R10: 0000000000000004 R11: 0000000000000005 R12: 0000000000000000
R13: ffff8eb8dd25f218 R14: 0000000000000000 R15: ffff8ea0dbcf0a70
ORIG_RAX: ffffffffffffffff CS: 0010 SS: 0018
#10 [ffff8e8a5b0cbe10] rcu_nocb_kthread at ffffffffa53532bc
#11 [ffff8e8a5b0cbec8] kthread at ffffffffa52c1c31
#12 [ffff8e8a5b0cbf50] ret_from_fork_nospec_begin at ffffffffa5974c1d
crash> ps |grep lustre_rmmod
41960 41832 30 ffff8ea4ef4d5140 IN 0.0 115300 1596 lustre_rmmod
crash>

Issue Links

• is related to: LU-11063 RHEL7.[345] RCU breakage