Re: 6.0-BETA2 reboots when issuing "zpool create"

[christos%astron.com@localhost (Christos Zoulas)]

In article <0E012154-8E9D-4E49-BE0B-67F649D0BC4C%carnat.net@localhost>,
Joel Carnat  <joel%carnat.net@localhost> wrote:
>Hello,
>
>I'm trying ZFS on NetBSD and it reboots each time I try to create the zpool.
>
>I have installed "NetBSD 6.0_BETA2 (GENERIC) amd64" in a VMware ESXi
>virtual machine.
>I configured a 8GB virtual disk on which I installed NetBSD using UFS.
>I attached two 4GB virtual disks on which I plan to install the ZFS.
>I have configured 2 vCPU and 4GB of RAM.
>
>I have the sd0 configured the "MBR-way" by the installer and did
>configure sd1 & sd2 using the "GPT-way", by hand. Here's my
>configuration:
>
>NetBSD installed on sd0 :
>#        size    offset     fstype [fsize bsize cpg/sgs]
> a:  16241715        63     4.2BSD   2048 16384     0  # (Cyl.      0*-   
> 1011*)
> b:    535438  16241778       swap                     # (Cyl.   1011*-   
> 1044*)
> c:  16777153        63     unused      0     0        # (Cyl.      0*-   
> 1044*)
> d:  16777216         0     unused      0     0        # (Cyl.      0 -   
> 1044*)
>
>Nothing yet on sd1 & sd2:
>5 partitions:
>#        size    offset     fstype [fsize bsize cpg/sgs]
> d:   8388608         0     unused      0     0        # (Cyl.      0 -    
> 522*)
> e:   8388607         1    unknown                     # (Cyl.      0*-    
> 522*)
>5 partitions:
>#        size    offset     fstype [fsize bsize cpg/sgs]
> d:   8388608         0     unused      0     0        # (Cyl.      0 -    
> 522*)
> e:   8388607         1    unknown                     # (Cyl.      0*-    
> 522*)
>
># gpt show sd0 sd1 sd2
>     start      size  index  contents
>         0         1         MBR
>         1        62         
>        63  16777153      1  MBR part 169
>    start     size  index  contents
>        0        1         PMBR
>        1        1         Pri GPT header
>        2       32         Pri GPT table
>       34  8388541      1  GPT part - 516e7cba-6ecf-11d6-8ff8-00022d09712b
>  8388575       32         Sec GPT table
>  8388607        1         Sec GPT header
>    start     size  index  contents
>        0        1         PMBR
>        1        1         Pri GPT header
>        2       32         Pri GPT table
>       34  8388541      1  GPT part - 516e7cba-6ecf-11d6-8ff8-00022d09712b
>  8388575       32         Sec GPT table
>  8388607        1         Sec GPT header
>
>I used the GUID type from FreeBSD-ZFS as I couldn't find the alias from
>`gpt` manpage.
>
># dkctl sd1 listwedges 
>/dev/rsd1d: 1 wedge:
>dk0: 35dae1b9-abef-11e1-99f6-005056915a7c, 8388541 blocks at 34, type: 
># dkctl sd2 listwedges 
>/dev/rsd2d: 1 wedge:
>dk1: 36cd77b7-abef-11e1-99f6-005056915a7c, 8388541 blocks at 34, type: 
>
>To create the ZFS pool, I run:
># zpool create tank /dev/dk0 /dev/dk1
>
>Then NetBSD reboots. In the logs, I can see:
>Jun  1 15:47:51 znetbsd /netbsd: WARNING: ZFS on NetBSD is under development
>Jun  1 15:48:08 znetbsd /netbsd: zfsdev_ioctl called 
>Jun  1 15:48:08 znetbsd /netbsd: zfsdev_ioctl -> zvol_ioctl
>Jun  1 15:48:08 znetbsd /netbsd: zfsdev_ioctl called 
>Jun  1 15:48:08 znetbsd /netbsd: zfsdev_ioctl -> zvol_ioctl
>Jun  1 15:48:08 znetbsd /netbsd: zfsdev_ioctl -> calling zfs_ioc_vec
>zvec_func on 4
>Jun  1 15:48:08 znetbsd /netbsd: zfsdev_ioctl 0
>Jun  1 15:50:12 znetbsd syslogd[356]: restart
>Jun  1 15:50:12 znetbsd /netbsd: zfsdev_ioctl called 
>Jun  1 15:50:12 znetbsd /netbsd: zfsdev_ioctl -> zvol_ioctl
>Jun  1 15:50:12 znetbsd /netbsd: zfsdev_ioctl -> calling zfs_ioc_vec
>zvec_func on 0
>Jun  1 15:50:12 znetbsd /netbsd: vnode @ 0xfffffe81082af170, flags
>(0x30<MPSAFE,LOCKSWORK>)
>Jun  1 15:50:12 znetbsd /netbsd:        tag VT_UFS(1), type VBLK(3),
>usecount 1, writecount 1, holdcount 0
>Jun  1 15:50:12 znetbsd /netbsd:        freelisthd 0x0, mount
>0xfffffe810753e000, data 0xfffffe81088aea00 lock 0xfffffe81082af280
>Jun  1 15:50:12 znetbsd /netbsd:        tag VT_UFS, ino 757668, on dev
>4, 0 flags 0x0, nlink 1
>Jun  1 15:50:12 znetbsd /netbsd:        mode 060640, owner 0, group 5, size 0
>Jun  1 15:50:12 znetbsd /netbsd: panic: vrelel: bad ref count
>Jun  1 15:50:12 znetbsd /netbsd: cpu1: Begin traceback...
>Jun  1 15:50:12 znetbsd /netbsd: vfs_vnode_sysinit() at 
>netbsd:vfs_vnode_sysinit
>Jun  1 15:50:12 znetbsd /netbsd: vrelel() at netbsd:vrelel+0x204
>Jun  1 15:50:12 znetbsd /netbsd: vdev_disk_open() at zfs:vdev_disk_open+0x173
>Jun  1 15:50:12 znetbsd /netbsd: cpu1: End traceback...
>Jun  1 15:50:12 znetbsd /netbsd: 
>Jun  1 15:50:12 znetbsd /netbsd: dumping to dev 4,1 offset 8
>Jun  1 15:50:12 znetbsd /netbsd: dump failed: insufficient space (535438
>< 8387713)
>Jun  1 15:50:12 znetbsd /netbsd: 
>Jun  1 15:50:12 znetbsd /netbsd: 
>Jun  1 15:50:12 znetbsd /netbsd: rebooting...
>
>Am I doing something wrong or is this a bug ?

Bug, zfs has not been used for a while so it is probably broken.

christos