An LOR recently observed on r212598

[b. f.]

An LOR, which resembles another reported in:

http://lists.freebsd.org/pipermail/freebsd-current/2010-August/018986.html

but none that I noticed at:

http://sources.zabbadoz.net/freebsd/lor.html

lock order reversal:
1st 0xffffff0001696098 ufs (ufs) @ /mnt/disk2/usr/src/sys/kern/vfs_lookup.c:501
2nd 0xffffff8026d985b8 bufwait (bufwait) @
/mnt/disk2/usr/src/sys/ufs/ffs/ffs_softdep.c:11309
3rd 0xffffff0001705638 ufs (ufs) @ /mnt/disk2/usr/src/sys/kern/vfs_subr.c:2111
KDB: stack backtrace:
db_trace_self_wrapper() at 0xffffffff801cae5a = db_trace_self_wrapper+0x2a
_witness_debugger() at 0xffffffff802ba265 = _witness_debugger+0x65
witness_checkorder() at 0xffffffff802bb513 = witness_checkorder+0x833
__lockmgr_args() at 0xffffffff8025efad = __lockmgr_args+0xd4d
ffs_lock() at 0xffffffff8041a9af = ffs_lock+0x8f
VOP_LOCK1_APV() at 0xffffffff804ab13b = VOP_LOCK1_APV+0x9b
_vn_lock() at 0xffffffff80313c67 = _vn_lock+0x57
vget() at 0xffffffff8030775b = vget+0x7b
vfs_hash_get() at 0xffffffff802fb065 = vfs_hash_get+0xd5
ffs_vgetf() at 0xffffffff80415ca8 = ffs_vgetf+0x48
softdep_sync_metadata() at 0xffffffff8041309e = softdep_sync_metadata+0x5de
ffs_syncvnode() at 0xffffffff8041a65a = ffs_syncvnode+0x22a
ffs_truncate() at 0xffffffff803fac38 = ffs_truncate+0x408
ufs_direnter() at 0xffffffff8042284d = ufs_direnter+0x6fd
ufs_makeinode() at 0xffffffff80427a74 = ufs_makeinode+0x254
VOP_CREATE_APV() at 0xffffffff804abaad = VOP_CREATE_APV+0x8d
vn_open_cred() at 0xffffffff80313642 = vn_open_cred+0x452
kern_openat() at 0xffffffff803126f1 = kern_openat+0x181
syscallenter() at 0xffffffff802b274a = syscallenter+0x1aa
syscall() at 0xffffffff8046dfac = syscall+0x4c
Xfast_syscall() at 0xffffffff8045a332 = Xfast_syscall+0xe2
--- syscall (5, FreeBSD ELF64, open), rip = 0x800726ddc, rsp =
0x7fffffffeac8, rbp = 0 ---


b.