Re: ZFS disaster on -current

[Jaromír Doleček <jaromir.dolecek%gmail.com@localhost>]

By chance, do you have the kernel crash dump from the original panic
which happened yesterday? The subsequent ones might be a result of the
first one.

The messages about redzone don't mean anything beyond that there is no
overflow protection for items on the pool.

Jaromir

Le mer. 24 juin 2020 à 11:34, Chavdar Ivanov <ci4ic4%gmail.com@localhost> a écrit :
>
> Hi,
>
> On
>
> NetBSD ymir 9.99.68 NetBSD 9.99.68 (GENERIC) #1: Tue Jun 23 22:53:46
> BST 2020  sysbuild@ymir:/home/sysbuild/amd64/obj/home/sysbuild/src/sys/arch/amd64/compile/GENERIC
> amd64
>
> I suddenly got a panic with ZFS; it took place with the previous
> kernel, so it was something with the module. In single user I disabled
> zfs in /etc/rc.conf and was able to complete boot, but obviously
> without my two pools.
>
> 'modload solaris' didn't show any problem.
>
> I set aside the contents of /etc/zfs and did 'modload zfs', which resulted in:
>
> .....
>
> WARNING: ZFS on NetBSD is under development
> pool redzone disabled for 'zio_buf_4096'
> pool redzone disabled for 'zio_data_buf_4096'
> pool redzone disabled for 'zio_buf_8192'
> pool redzone disabled for 'zio_data_buf_8192'
> pool redzone disabled for 'zio_buf_16384'
> pool redzone disabled for 'zio_data_buf_16384'
> pool redzone disabled for 'zio_buf_32768'
> pool redzone disabled for 'zio_data_buf_32768'
> pool redzone disabled for 'zio_buf_65536'
> pool redzone disabled for 'zio_data_buf_65536'
> pool redzone disabled for 'zio_buf_131072'
> pool redzone disabled for 'zio_data_buf_131072'
> pool redzone disabled for 'zio_buf_262144'
> pool redzone disabled for 'zio_data_buf_262144'
> pool redzone disabled for 'zio_buf_524288'
> pool redzone disabled for 'zio_data_buf_524288'
> pool redzone disabled for 'zio_buf_1048576'
> pool redzone disabled for 'zio_data_buf_1048576'
> pool redzone disabled for 'zio_buf_2097152'
> pool redzone disabled for 'zio_data_buf_2097152'
> pool redzone disabled for 'zio_buf_4194304'
> pool redzone disabled for 'zio_data_buf_4194304'
> pool redzone disabled for 'zio_buf_8388608'
> pool redzone disabled for 'zio_data_buf_8388608'
> pool redzone disabled for 'zio_buf_16777216'
> pool redzone disabled for 'zio_data_buf_16777216'
>
> I have no idea what that means, it is a first for me, ZFS otherwise
> has been very reliable on this hardware so far, inasmuch as I have the
> mercurial repo on a zfs and build from it from time to time (the panic
> is from the last cvs update from yesterday, though).
>
> Subsequent 'zpool import' repeated the panic (without getting me into
> the debugger, though):
>
>
> ZFS filesystem version: 5
> uvm_fault(0xffffa97e4c3e1610, 0x0, 1) -> e
> fatal page fault in supervisor mode
> trap type 6 code 0 rip 0xffffffff81d49882 cs 0x8 rflags 0x10286 cr2
> 0xa0 ilevel 0 rsp 0xffffde819c16d760
> curlwp 0xffffa97e3a41e140 pid 17394.17394 lowest kstack 0xffffde819c16a2c0
> panic: trap
> cpu0: Begin traceback...
> vpanic() at netbsd:vpanic+0x152
> snprintf() at netbsd:snprintf
> startlwp() at netbsd:startlwp
> alltraps() at netbsd:alltraps+0xc3
> vdev_open() at zfs:vdev_open+0x9e
> vdev_open_children() at zfs:vdev_open_children+0x39
> vdev_root_open() at zfs:vdev_root_open+0x33
> vdev_open() at zfs:vdev_open+0x9e
> spa_load() at zfs:spa_load+0x38e
> spa_tryimport() at zfs:spa_tryimport+0x86
> zfs_ioc_pool_tryimport() at zfs:zfs_ioc_pool_tryimport+0x41
> zfsdev_ioctl() at zfs:zfsdev_ioctl+0x8c1
> nb_zfsdev_ioctl() at zfs:nb_zfsdev_ioctl+0x38
> VOP_IOCTL() at netbsd:VOP_IOCTL+0x44
> vn_ioctl() at netbsd:vn_ioctl+0xa5
> sys_ioctl() at netbsd:sys_ioctl+0x550
> syscall() at netbsd:syscall+0x26e
> --- syscall (number 54) ---
> netbsd:syscall+0x26e:
> cpu0: End traceback...
>
> The above panic did not leave a crash dump.
>
> When I had /etc/zfs populated before, I also got a crash dump (with
> 'reboot 0x104'), as follows:
>
> # crash -M netbsd.18.core -N netbsd.18
> Crash version 9.99.68, image version 9.99.68.
> crash: _kvm_kvatop(0)
> Kernel compiled without options LOCKDEBUG.
> System panicked: reboot forced via kernel debugger
> Backtrace from time of crash is available.
> crash> bt
> _KERNEL_OPT_NARCNET() at 0
> _KERNEL_OPT_NARCNET() at 0
> sys_reboot() at sys_reboot
> db_fncall() at db_fncall
> db_command() at db_command+0x127
> db_command_loop() at db_command_loop+0xa6
> db_trap() at db_trap+0xe6
> kdb_trap() at kdb_trap+0xe1
> trap() at trap+0x2b7
> --- trap (number 6) ---
> vdev_disk_open.part.4() at vdev_disk_open.part.4+0x49a
> vdev_open() at vdev_open+0x9e
> vdev_open_children() at vdev_open_children+0x39
> vdev_root_open() at vdev_root_open+0x33
> vdev_open() at vdev_open+0x9e
> spa_load() at spa_load+0x38e
> spa_load_best() at spa_load_best+0x58
> spa_open_common() at spa_open_common+0xc2
> pool_status_check.part.25() at pool_status_check.part.25+0x1e
> zfsdev_ioctl() at zfsdev_ioctl+0x80e
> nb_zfsdev_ioctl() at nb_zfsdev_ioctl+0x38
> VOP_IOCTL() at VOP_IOCTL+0x44
> vn_ioctl() at vn_ioctl+0xa5
> sys_ioctl() at sys_ioctl+0x550
> syscall() at syscall+0x26e
> --- syscall (number 54) ---
> syscall+0x26e:
> .....
>
> Any idea what is going on? I've restarted a build, but the cvs log
> doesn't show anything relevant as far as I can see.
>
>
> Chavdar
>
>
>
> --
> ----