Re: Crash in ZFS during vnode sync

It looks like you tried to lock a destroyed vnode. Did you get a core?

Cheers,
Kip


On Tue, Jun 16, 2009 at 3:10 PM, Artem Belevich<fbsdlist_at_src.cx> wrote:
> Got a new crash on fresh (as of yesterday) -current/amd64 while the
> box was pretty much idle. Haven't seen this one before.
>
> Fatal trap 12: page fault while in kernel mode
> cpuid = 3; apic id = 03
> fault virtual address   = 0xd8
> fault code              = supervisor write data, page not present
> instruction pointer     = 0x20:0xffffffff80360a35
> stack pointer           = 0x28:0xffffff842cfc8900
> frame pointer           = 0x28:0xffffff842cfc8910
> code segment            = base 0x0, limit 0xfffff, type 0x1b
>                        = DPL 0, pres 1, long 1, def32 0, gran 1
> processor eflags        = interrupt enabled, resume, IOPL = 0
> current process         = 48 (syncer)
> [thread pid 48 tid 100076 ]
> Stopped at      _mtx_lock_flags+0x15:   lock cmpxchgq   %rsi,0x18(%rdi)
> db> where
> Tracing pid 48 tid 100076 td 0xffffff0007b0a390
> _mtx_lock_flags() at _mtx_lock_flags+0x15
> vn_rele_async() at vn_rele_async+0x31
> zfs_get_data() at zfs_get_data+0xd0
> zil_commit() at zil_commit+0x532
> zfs_sync() at zfs_sync+0xa6
> sync_fsync() at sync_fsync+0x184
> VOP_FSYNC_APV() at VOP_FSYNC_APV+0x4a
> sync_vnode() at sync_vnode+0x16b
> sched_sync() at sched_sync+0x1c9
> fork_exit() at fork_exit+0x118
> fork_trampoline() at fork_trampoline+0xe
> --- trap 0, rip = 0, rsp = 0xffffff842cfc8d40, rbp = 0 ---
>
> db> show lockedbufs
> buf at 0xffffff83c11d8550
> b_flags = 0xa00200a0<remfree,vmio,clusterok,delwri,cache>
> b_error = 0, b_bufsize = 4096, b_bcount = 4096, b_resid = 0
> b_bufobj = (0xffffff0012c7a2f0), b_data = 0xffffff83da5fa000, b_blkno
> = 19214240, b_dep = 0
> b_npages = 1, pages(OBJ, IDX, PA): (0xffffff003d5b47d0, 0x0, 0x3c082000)
> lock type bufwait: EXCL by thread 0xffffff0010e85720 (pid 886)
>
> db> show lockedvnods
> Locked vnodes
>
> 0xffffff0012c7a1d8: tag ufs, type VREG
>    usecount 1, writecount 1, refcount 3 mountedhere 0
>    flags ()
>    v_object 0xffffff003d5b47d0 ref 0 pages 1
>    lock type ufs: EXCL by thread 0xffffff0010e85720 (pid 886)
>        ino 1201240, on dev ad4s1a
>
> 0xffffff0007db9ce8: tag syncer, type VNON
>    usecount 1, writecount 0, refcount 2 mountedhere 0
>    flags ()
>    lock type syncer: EXCL by thread 0xffffff0007b0a390 (pid 48)
> db> show vnodebufs
> usage: show vnodebufs <addr>
> db> show vnodebufs 0xffffff0007db9ce8
> Clean buffers:
> Dirty buffers:
> db> show vnodebufs 0xffffff0012c7a1d8
> Clean buffers:
> Dirty buffers:
> buf at 0xffffff83c11d8550
> b_flags = 0xa00200a0<remfree,vmio,clusterok,delwri,cache>
> b_error = 0, b_bufsize = 4096, b_bcount = 4096, b_resid = 0
> b_bufobj = (0xffffff0012c7a2f0), b_data = 0xffffff83da5fa000, b_blkno
> = 19214240, b_dep = 0
> b_npages = 1, pages(OBJ, IDX, PA): (0xffffff003d5b47d0, 0x0, 0x3c082000)
> lock type bufwait: EXCL by thread 0xffffff0010e85720 (pid 886)
>
> db> bt 886
> Tracing pid 886 tid 100183 td 0xffffff0010e85720
> cpustop_handler() at cpustop_handler+0x3b
> ipi_nmi_handler() at ipi_nmi_handler+0x30
> trap() at trap+0x195
> nmi_calltrap() at nmi_calltrap+0x8
> --- trap 0x13, rip = 0xffffffff803605ca, rsp = 0xffffff800001bfe0, rbp
> = 0xffffff842f52b7f0 ---
> _mtx_lock_sleep() at _mtx_lock_sleep+0x11a
> bdwrite() at bdwrite+0x313
> ffs_write() at ffs_write+0x634
> VOP_WRITE_APV() at VOP_WRITE_APV+0xc6
> vn_write() at vn_write+0x188
> dofilewrite() at dofilewrite+0x85
> kern_writev() at kern_writev+0x60
> writev() at writev+0x41
> syscall() at syscall+0x28f
> Xfast_syscall() at Xfast_syscall+0xd0
> --- syscall (121, FreeBSD ELF64, writev), rip = 0x80083501c, rsp =
> 0x7fffffffcaf8, rbp = 0 ---
> db> bt 48
> Tracing pid 48 tid 100076 td 0xffffff0007b0a390
> _mtx_lock_flags() at _mtx_lock_flags+0x15
> vn_rele_async() at vn_rele_async+0x31
> zfs_get_data() at zfs_get_data+0xd0
> zil_commit() at zil_commit+0x532
> zfs_sync() at zfs_sync+0xa6
> sync_fsync() at sync_fsync+0x184
> VOP_FSYNC_APV() at VOP_FSYNC_APV+0x4a
> sync_vnode() at sync_vnode+0x16b
> sched_sync() at sched_sync+0x1c9
> fork_exit() at fork_exit+0x118
> fork_trampoline() at fork_trampoline+0xe
> --- trap 0, rip = 0, rsp = 0xffffff842cfc8d40, rbp = 0 ---
>
>
>
> --Artem
> _______________________________________________
> freebsd-current_at_freebsd.org mailing list
> http://lists.freebsd.org/mailman/listinfo/freebsd-current
> To unsubscribe, send any mail to "freebsd-current-unsubscribe_at_freebsd.org"
>



-- 
When bad men combine, the good must associate; else they will fall one
by one, an unpitied sacrifice in a contemptible struggle.

    Edmund Burke