[Bug 209158] node / npm triggering zfs rename deadlock

[bugzilla-noreply at freebsd.org]

https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=209158

            Bug ID: 209158
           Summary: node / npm triggering zfs rename deadlock
           Product: Base System
           Version: 11.0-CURRENT
          Hardware: Any
                OS: Any
            Status: New
          Severity: Affects Some People
          Priority: ---
         Component: kern
          Assignee: freebsd-bugs at FreeBSD.org
          Reporter: peter at FreeBSD.org

The freebsd.org cluster package builder has been deadlocking regularly with
things like this:

root at bob.nyi:/home/peter # ps -axHl -O lwp | grep node
    0 92502 92499   0   52  0 771080 197588 kqread   I+J   1       0:07.57 npm
(node)       92502 101429  1  I+J       0:07.57 npm (node)
    0 92502 92499   0   22  0 771080 197588 usem     I+J   1       0:00.11 npm
(node)       92502 101858  1  I+J       0:00.11 npm (node)
    0 92502 92499   0   22  0 771080 197588 usem     I+J   1       0:00.11 npm
(node)       92502 101859  1  I+J       0:00.11 npm (node)
    0 92502 92499   0   22  0 771080 197588 usem     I+J   1       0:00.11 npm
(node)       92502 101860  1  I+J       0:00.11 npm (node)
    0 92502 92499   0   22  0 771080 197588 usem     I+J   1       0:00.10 npm
(node)       92502 101861  1  I+J       0:00.10 npm (node)
    0 92502 92499   0   28  0 771080 197588 zfs      D+J   1       0:00.89 npm
(node)       92502 101862  1  D+J       0:00.89 npm (node)
    0 92502 92499   0   28  0 771080 197588 zfs      D+J   1       0:00.90 npm
(node)       92502 101863  1  D+J       0:00.90 npm (node)
    0 92502 92499   0   28  0 771080 197588 zfs      D+J   1       0:00.89 npm
(node)       92502 101864  1  D+J       0:00.89 npm (node)
    0 92502 92499   0   28  0 771080 197588 zfs      D+J   1       0:00.91 npm
(node)       92502 101865  1  D+J       0:00.91 npm (node)
    0 84004 81177   0   20  0  14864   2532 piperd   S+    5       0:00.00 grep
node        84004 101732  5  S+        0:00.00 grep node
root at bob.nyi:/home/peter # kgdb


(kgdb) tid 101862
[Switching to thread 580 (Thread 101862)]#0  0xffffffff807e9f83 in sched_switch
()
(kgdb) where
#0  0xffffffff807e9f83 in sched_switch ()
#1  0xffffffff807c3f88 in mi_switch ()
#2  0xffffffff8080be0a in sleepq_wait ()
#3  0xffffffff807941d5 in sleeplk ()
#4  0xffffffff8079350f in __lockmgr_args ()
#5  0xffffffff808710bc in vop_stdlock ()
#6  0xffffffff80b9b64f in VOP_LOCK1_APV ()
#7  0xffffffff808953c3 in _vn_lock ()
#8  0xffffffff803cb542 in zfs_lookup ()
#9  0xffffffff803cb86d in zfs_freebsd_lookup ()
#10 0xffffffff80b99810 in VOP_CACHEDLOOKUP_APV ()
#11 0xffffffff8086dc56 in vfs_cache_lookup ()
#12 0xffffffff80b99710 in VOP_LOOKUP_APV ()
#13 0xffffffff80877131 in lookup ()
#14 0xffffffff80876811 in namei ()
#15 0xffffffff80891da1 in kern_renameat ()
#16 0xffffffff80af24a8 in amd64_syscall ()
#17 0xffffffff80ad69bb in Xfast_syscall ()

(kgdb) tid 101863
[Switching to thread 579 (Thread 101863)]#0  0xffffffff807e9f83 in sched_switch
()
(kgdb) where
#0  0xffffffff807e9f83 in sched_switch ()
#1  0xffffffff807c3f88 in mi_switch ()
#2  0xffffffff8080be0a in sleepq_wait ()
#3  0xffffffff807941d5 in sleeplk ()
#4  0xffffffff8079350f in __lockmgr_args ()
#5  0xffffffff808710bc in vop_stdlock ()
#6  0xffffffff80b9b64f in VOP_LOCK1_APV ()
#7  0xffffffff808953c3 in _vn_lock ()
#8  0xffffffff808839b7 in vget ()
#9  0xffffffff8086c3d8 in cache_lookup ()
#10 0xffffffff8086dc2c in vfs_cache_lookup ()
#11 0xffffffff80b99710 in VOP_LOOKUP_APV ()
#12 0xffffffff80877131 in lookup ()
#13 0xffffffff80876811 in namei ()
#14 0xffffffff80891da1 in kern_renameat ()
#15 0xffffffff80af24a8 in amd64_syscall ()
#16 0xffffffff80ad69bb in Xfast_syscall ()


(kgdb) tid 101864
[Switching to thread 578 (Thread 101864)]#0  0xffffffff807e9f83 in sched_switch
()
(kgdb) where
#0  0xffffffff807e9f83 in sched_switch ()
#1  0xffffffff807c3f88 in mi_switch ()
#2  0xffffffff8080be0a in sleepq_wait ()
#3  0xffffffff807941d5 in sleeplk ()
#4  0xffffffff80793d35 in __lockmgr_args ()
#5  0xffffffff808710bc in vop_stdlock ()
#6  0xffffffff80b9b64f in VOP_LOCK1_APV ()
#7  0xffffffff808953c3 in _vn_lock ()
#8  0xffffffff808840d8 in vputx ()
#9  0xffffffff803d2303 in zfs_freebsd_rename ()
#10 0xffffffff80b9adbf in VOP_RENAME_APV ()
#11 0xffffffff80892198 in kern_renameat ()
#12 0xffffffff80af24a8 in amd64_syscall ()
#13 0xffffffff80ad69bb in Xfast_syscall ()

(kgdb) tid 101865
[Switching to thread 577 (Thread 101865)]#0  0xffffffff807e9f83 in sched_switch
()
(kgdb) where
#0  0xffffffff807e9f83 in sched_switch ()
#1  0xffffffff807c3f88 in mi_switch ()
#2  0xffffffff8080be0a in sleepq_wait ()
#3  0xffffffff807941d5 in sleeplk ()
#4  0xffffffff8079350f in __lockmgr_args ()
#5  0xffffffff808710bc in vop_stdlock ()
#6  0xffffffff80b9b64f in VOP_LOCK1_APV ()
#7  0xffffffff808953c3 in _vn_lock ()
#8  0xffffffff808839b7 in vget ()
#9  0xffffffff8086c3d8 in cache_lookup ()
#10 0xffffffff8086dc2c in vfs_cache_lookup ()
#11 0xffffffff80b99710 in VOP_LOOKUP_APV ()
#12 0xffffffff80877131 in lookup ()
#13 0xffffffff80876811 in namei ()
#14 0xffffffff80891da1 in kern_renameat ()
#15 0xffffffff80af24a8 in amd64_syscall ()
#16 0xffffffff80ad69bb in Xfast_syscall ()
(kgdb) 

Build info is:
FreeBSD bob.nyi.freebsd.org 11.0-CURRENT FreeBSD 11.0-CURRENT #0 r296980: Thu
Mar 17 07:14:40 UTC 2016 

This is a relatively new development.  This used to build robustly for a while
but we've been having lockups reliably for a few weeks.

-- 
You are receiving this mail because:
You are the assignee for the bug.