My Oracle Support Banner

XFS Freeze Operation causes the System Hang with Deadlock under Memory Pressure (Doc ID 2646027.1)

Last updated on MARCH 18, 2020

Applies to:

Linux OS - Version Oracle Linux 7.6 and later
Linux x86-64

Symptoms

When a XFS Freeze Operation (xfs_freeze) is used under memory pressure, a deadlock situation may happen.

From the vmcore analysis, the following 3 deadlock threads can be seen.

PID: 934 TASK: ffff90e92e6ac740 CPU: 0 COMMAND: "systemd-logind"
#0 [ffffa3ca8776bc90] __schedule at ffffffffb486487c
#1 [ffffa3ca8776bd28] schedule at ffffffffb4864e96
#2 [ffffa3ca8776bd40] rwsem_down_write_failed at ffffffffb4868f18
#3 [ffffa3ca8776bdd8] call_rwsem_down_write_failed at ffffffffb4859367
#4 [ffffa3ca8776be20] down_write at ffffffffb4867cfd
#5 [ffffa3ca8776be38] unregister_shrinker at ffffffffb41ef010
#6 [ffffa3ca8776be50] deactivate_locked_super at ffffffffb4285abb
#7 [ffffa3ca8776be70] deactivate_super at ffffffffb4285b45
#8 [ffffa3ca8776be88] cleanup_mnt at ffffffffb42a753f
#9 [ffffa3ca8776bea0] __cleanup_mnt at ffffffffb42a75c2
#10 [ffffa3ca8776beb0] task_work_run at ffffffffb40b59e0
#11 [ffffa3ca8776bee8] exit_to_usermode_loop at ffffffffb40040a4
#12 [ffffa3ca8776bf28] do_syscall_64 at ffffffffb4003a62
#13 [ffffa3ca8776bf50] entry_SYSCALL_64_after_hwframe at ffffffffb4a001ad
RIP: 00007fa44e8470c7 RSP: 00007ffdca56d978 RFLAGS: 00000246
RAX: 0000000000000000 RBX: 0000000000000000 RCX: 00007fa44e8470c7
RDX: 000000000000fe40 RSI: 0000000000000002 RDI: 0000564c571c1760
RBP: 0000000000000000 R8: 0000000000000001 R9: 0000000000008030
R10: 0000000000000076 R11: 0000000000000246 R12: 0000564c571bd470
R13: 0000000000000001 R14: 0000000000000000 R15: 0000000000000000
ORIG_RAX: 00000000000000a6 CS: 0033 SS: 002b
PID: 128032 TASK: ffff90e89d892f80 CPU: 1 COMMAND: "writers.bash"
#0 [ffffa3ca86893510] __schedule at ffffffffb486487c
#1 [ffffa3ca868935a8] schedule at ffffffffb4864e96
#2 [ffffa3ca868935c0] rwsem_down_read_failed at ffffffffb48689ee
#3 [ffffa3ca86893650] call_rwsem_down_read_failed at ffffffffb4859308
#4 [ffffa3ca868936a0] __percpu_down_read at ffffffffb40ebd38
#5 [ffffa3ca868936c0] __sb_start_write at ffffffffb42859ef
#6 [ffffa3ca868936e0] xfs_trans_alloc at ffffffffc0849e9c [xfs]
#7 [ffffa3ca86893728] xfs_free_eofblocks at ffffffffc08219d1 [xfs]
#8 [ffffa3ca86893790] xfs_inactive at ffffffffc083c878 [xfs]
#9 [ffffa3ca868937b0] __dta_xfs_fs_destroy_inode_3543 at ffffffffc084685e [xfs]
#10 [ffffa3ca868937e0] destroy_inode at ffffffffb42a25de
#11 [ffffa3ca868937f8] evict at ffffffffb42a2b73
#12 [ffffa3ca86893820] dispose_list at ffffffffb42a2c1d
#13 [ffffa3ca86893848] prune_icache_sb at ffffffffb42a421a
#14 [ffffa3ca86893880] super_cache_scan at ffffffffb42870a1
#15 [ffffa3ca868938d8] shrink_slab at ffffffffb41eebb3
#16 [ffffa3ca868939c0] shrink_node at ffffffffb41f4788
#17 [ffffa3ca86893a48] do_try_to_free_pages at ffffffffb41f4cc2
#18 [ffffa3ca86893ab0] try_to_free_pages at ffffffffb41f5069
#19 [ffffa3ca86893b38] __alloc_pages_slowpath at ffffffffb41e38fd
#20 [ffffa3ca86893c58] __alloc_pages_nodemask at ffffffffb41e1391
#21 [ffffa3ca86893cc8] alloc_pages_vma at ffffffffb4245bbf
#22 [ffffa3ca86893d10] wp_page_copy at ffffffffb42141e8
#23 [ffffa3ca86893d88] do_wp_page at ffffffffb42177c7
#24 [ffffa3ca86893dc0] __handle_mm_fault at ffffffffb421aaa3
#25 [ffffa3ca86893e70] handle_mm_fault at ffffffffb421b28c
#26 [ffffa3ca86893ea0] __do_page_fault at ffffffffb4078634
#27 [ffffa3ca86893f18] do_page_fault at ffffffffb4078928
#28 [ffffa3ca86893f50] page_fault at ffffffffb4a03e6d
RIP: 00007f97581401df RSP: 00007ffd6f8fa260 RFLAGS: 00010246
RAX: 0000000000000000 RBX: 0000000000a90580 RCX: 0000000000a90500
RDX: 00000000ffffffff RSI: 00007f9758485768 RDI: 00007f9758485760
RBP: 00007f9758485760 R8: 0000000000a8e540 R9: 0000000000a90590
R10: 0000000000000008 R11: 0000000000000246 R12: 0000000000000000
R13: 0000000000000000 R14: 0000000000000000 R15: 0000000000000000
ORIG_RAX: ffffffffffffffff CS: 0033 SS: 002b
PID: 34176 TASK: ffff90e8cd0c4740 CPU: 0 COMMAND: "fsfreeze"
#0 [ffffa3ca8b803c90] __schedule at ffffffffb486487c
#1 [ffffa3ca8b803d28] schedule at ffffffffb4864e96
#2 [ffffa3ca8b803d40] rwsem_down_write_failed at ffffffffb4868f18
#3 [ffffa3ca8b803dd8] call_rwsem_down_write_failed at ffffffffb4859367
#4 [ffffa3ca8b803e20] down_write at ffffffffb4867cfd
#5 [ffffa3ca8b803e38] thaw_super at ffffffffb4285d2d
#6 [ffffa3ca8b803e60] do_vfs_ioctl at ffffffffb4299566
#7 [ffffa3ca8b803ee8] sys_ioctl at ffffffffb4299709
#8 [ffffa3ca8b803f28] do_syscall_64 at ffffffffb4003949
#9 [ffffa3ca8b803f50] entry_SYSCALL_64_after_hwframe at ffffffffb4a001ad
RIP: 00007fd701f60a57 RSP: 00007ffd6ce74e78 RFLAGS: 00000206
RAX: ffffffffffffffda RBX: 0000000000000003 RCX: 00007fd701f60a57
RDX: 0000000000000000 RSI: 00000000c0045878 RDI: 0000000000000003
RBP: 00007ffd6ce7689f R8: 0000000000000000 R9: 0000000000000000
R10: 00007ffd6ce748e0 R11: 0000000000000206 R12: 0000000000000002
R13: 00007ffd6ce75020 R14: 0000000000000000 R15: 0000000000000000
ORIG_RAX: 0000000000000010 CS: 0033 SS: 002b

 

Cause

To view full details, sign in with your My Oracle Support account.

Don't have a My Oracle Support account? Click to get started!


In this Document
Symptoms
Cause
Solution
References


My Oracle Support provides customers with access to over a million knowledge articles and a vibrant support community of peers and Oracle experts.