devfs and hot unplugging firewire device
M. L. Dodson
mldodson at houston.rr.com
Wed Sep 20 13:04:46 PDT 2006
On Wednesday 20 September 2006 14:42, John-Mark Gurney wrote:
> M. L. Dodson wrote this message on Wed, Sep 20, 2006 at 14:14 -0500:
> > More experimental results on firewire disks and hot unplugging:
> > Transcripts of two sessions can be retrieved from
> > http://siegfried.utmb.edu/bdodson/firewire.on.histidine.txt and
> > http://siegfried.utmb.edu/bdodson/firewire.on.serine.txt
> doh, I think I know what your problem is... you need to do three
> or so fwcontrol -r's before the device will disappear... This is
> necessary, since each plug causes a rescan, and you don't want devices
> disappearing when you add your cd drive... It looks like adding a
> timeout would be a good thing to the device orphan... If the device
> has been orphaned for more than a minute, drop it, or three or so
> Instead of fwcontrol -r, you can plug/unplug another device, and a
> couple times will cause the device to fall off the bus..
OK, that all seems reasonable (at my level of understanding). I
thought it was automatic, but it seems not so. Sort of like your
last statement on device orphan (not a driver guru, sorry.)
> Are you putting a different drive into the same enclosure? If you
> are, then the disk 'reappears', and apparently geom tastes w/ some
> of the old info.. if you do: true > /dev/da4 it will force a retaste
> and the bogus entries should go away...
Yes this is exactly the situation. I physically replace the drive
in the adapter (I don't even put it in the nice aluminum case).
So the adapter is the same, but the drive is not. So, assuming
the driver actually gets to the info for the actual drive, it will
be different, but also alike. I suspected that was causing the
driver confusion. I assume retaste is geom jargon for querying
the device info?
Thanks for all your help. I'll try all that.
M. L. Dodson
More information about the freebsd-firewire