freebsd stable - Mar 2010 - 8.0-RELEASE-p2: boot0cfg yields "vnode_pager_getpages: I/O read error"

Greetings.  I'm testing NanoBSD images for eventual use in an Alix board
and have run into an issue with the boot0cfg lines in the update scripts,
in multiple environments.  Specifically, any use of boot0cfg yields a
successful output[1], followed by:

vnode_pager_getpages: I/O read error

At this point, it looks like the command succeeded, but all reads and
writes to disk fail:

# man init
/usr/bin/man: Device not configured.

# touch /var/tmp/test
/usr/bin/touch: Input/output error.

I've tried the following combinations and it happens under all of them:

- VM using a file-loop-backed disk (KVM on Debian Lenny/5.0)
- VM using a block-backed disk (same)
- Alix board with CF disk in LBA mode
- Alix board with CF disk in CHS mode
- All of the above in packet and nopacket modes

I've tried using the 'sysctl kern.geom.debugflags=0x10' mentioned in
the
boot0cfg man page, but that results in no change.  The NanoBSD images I'm
using are mostly vanilla, built on 8.0-RELEASE-p2.  Any ideas?  Bug or
user error?  Thanks.

Brian Conway


[1]
# boot0cfg -s 1 -v ad0
#   flag     start chs   type       end chs       offset         size
1   0x80      0:  1: 1   0xa5    791: 15:63           63       798273
2   0x00    792:  1: 1   0xa5    559: 15:63       798399       798273
3   0x00    560:  0: 1   0xa5    562: 15:63      1596672         3024

version=2.0  drive=0x80  mask=0x3  ticks=182  bell=# (0x23)
options=packet,update,nosetdrv
volume serial ID 9090-9090
default_selection=F1 (Slice 1)

Dnia poniedzia?ek, 8 marca 2010 o 04:00:02 Brian Conway
napisa?(a):
> Greetings.  I'm testing NanoBSD images for eventual use in an Alix
board
> and have run into an issue with the boot0cfg lines in the update scripts,
> in multiple environments.  Specifically, any use of boot0cfg yields a
> successful output[1], followed by:
> ...
> Any ideas?  Bug or
> user error?  Thanks.
> 
> Brian Conway
Brian,
try changing in updatep1 and updatep2 to use gpart not boot0cfg:

#boot0cfg -s 1 -v ${NANO_DRIVE}
gpart set -a active -i 1 ${NANO_DRIVE}

I think is the problem. I was having similar errors as you while I was using 
boot0cfg.

Maciej Milewski

On Sun, Mar 07, 2010 at 10:00:02PM -0500, Brian Conway wrote:
> Greetings.  I'm testing NanoBSD images for eventual use in an Alix
board
> and have run into an issue with the boot0cfg lines in the update scripts,
> in multiple environments.  Specifically, any use of boot0cfg yields a
> successful output[1], followed by:
> 
> vnode_pager_getpages: I/O read error
> 
> At this point, it looks like the command succeeded, but all reads and
> writes to disk fail:
> 
> # man init
> /usr/bin/man: Device not configured.
> 
> # touch /var/tmp/test
> /usr/bin/touch: Input/output error.
boot0cfg is broken since RELENG_6 (at least) for unknown reason.

For RELENG_6, there was a workaround (use kern.geom.debugflags=16)
but it does not work for RELENG_8 anymore.

Eugene Grosbein