https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=281749

--- Comment #28 from [email protected] ---
(In reply to Konstantin Belousov from comment #27)

~ 6 min before usual place for failure:

reclaim with v_usecount 1
0xfffff8001538bc08: type VREG state VSTATE_DESTROYING op 0xffffffff82410158
    usecount 0, writecount 0, refcount 1 seqc users 1
    hold count flags ()
    flags (VIRF_DOOMED)
    lock type zfs: EXCL by thread 0xfffff80004278780 (pid 0, kernel, tid
100153)
KDB: stack backtrace:
#0 0xffffffff80bbe18d at kdb_backtrace+0x5d
#1 0xffffffff821622ab at zfs_freebsd_reclaim+0x6b
#2 0xffffffff811416fc at VOP_RECLAIM_APV+0x1c
#3 0xffffffff80c60315 at vgonel+0x315
#4 0xffffffff80c5c579 at vnlru_free_impl+0x419
#5 0xffffffff82158d60 at zfs_prune_task+0x50
#6 0xffffffff8218e89b at arc_prune_task+0x1b
#7 0xffffffff821447d7 at taskq_run+0x17
#8 0xffffffff80bd4d82 at taskqueue_run_locked+0x182
#9 0xffffffff80bd5f52 at taskqueue_thread_loop+0xc2
#10 0xffffffff80b2782b at fork_exit+0x7b
#11 0xffffffff8105103e at fork_trampoline+0xe

~ at the moment of usual failure:

reclaim with v_usecount 1
0xfffff80058035dc0: type VREG state VSTATE_DESTROYING op 0xffffffff82410158
    usecount 0, writecount 0, refcount 1 seqc users 1
    hold count flags ()
    flags (VIRF_DOOMED)
    lock type zfs: EXCL by thread 0xfffff80004278780 (pid 0, kernel, tid
100153)
KDB: stack backtrace:
#0 0xffffffff80bbe18d at kdb_backtrace+0x5d
#1 0xffffffff821622ab at zfs_freebsd_reclaim+0x6b
#2 0xffffffff811416fc at VOP_RECLAIM_APV+0x1c
#3 0xffffffff80c60315 at vgonel+0x315
#4 0xffffffff80c5c579 at vnlru_free_impl+0x419
#5 0xffffffff82158d60 at zfs_prune_task+0x50
#6 0xffffffff8218e89b at arc_prune_task+0x1b
#7 0xffffffff821447d7 at taskq_run+0x17
#8 0xffffffff80bd4d82 at taskqueue_run_locked+0x182
#9 0xffffffff80bd5f52 at taskqueue_thread_loop+0xc2
#10 0xffffffff80b2782b at fork_exit+0x7b
#11 0xffffffff8105103e at fork_trampoline+0xe


this run it was not failed

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

Reply via email to