On Thu, 29 Aug 1996, Chris Mason wrote:

> MacOS formatter utils report that indeed these are defective blocks, but
> when I reformatted the drive the last time trough, the defective block list
> was reset and the util reported no errors.  In other words, somehow MacBSD
> is corrupting the disk.

This conclusion is incorrect.  If the blocks were bad before the format,
there is every reason to believe that the blocks are _still_ bad, whether
they are marked as such by the formatter utility or not.  Another
possibility is a head crash on the disk.  In any event, it is simply not
possible that the SCSI driver -- be it ncrscsi or sbc -- is causing medium
errors. 

> Right now I'm using GENERIC_SBC_2.  I've just done an fsck and its
> corrected some of the problems that have cropped up, so I'll see what
> happens.

Now, timing issues, those are another story.  In polled mode (which is the
default, and what the GENERICSBC #2 kernel users), this should not be a
problem.  If, however, any disk corruption result from this sort of thing,
fsck would be able to repair it "permanently"... or until the next timing
problem occured. These would be filesystem errors that were not
accompanied by "medium error" message in the log file, to be absolutely
clear.