> Thanks for the quick response.
> I made this change, and the panic no longer occurs there (the one with
> the oops output), but I still get the other "scsi_free:Bad offset"
> panic. For some reason, that particular panic does not produce the
> "oops" type output, so there is no traceback to follow. Any idea how I
> can get it to supply me with the oops output? I realize this is likely
> no longer an XFS problem, but something in the SCSI layer.
You might ask on linux kernel, there are people out there who may
recognize the error.
> In terms of this modification you described, are there any system level
> impacts to removing this code? Based on the comments there, it doesn't
> seem like there should be. Incidentally, I immediately unmount and ther
> remount the filesystem once I'm able to do so (ie: once the FC link is
> plugged back in).
There should be no problem removing the code, other parts of the system
will eventually have the same effect.