On Jul 1, Matthew Jacob wrote
> I would like to make the following change to scsi_base.c- the
> comment covers most of it, but basically it's to quiesce the
> tape driver from getting sense data *always* printed out
> for non-error 'errors' (e.g., filemarks and short reads).
> 
> Opinions?

Why not fix st.c's error handler to return 0 instead of -1 in this case ?
I think your change will break the semantic of per-driver error handler.
If error messages are printed when they should'nt be, it's a problem in
st_interpret_sense(), not in the generic handler.

--
Manuel Bouyer, LIP6, Universite Paris VI.           Manuel.Bouyer@lip6.fr
--