Subject: Re: opinion sought about minor change to error reporting in scsi_base
To: None <bouyer@antioche.lip6.fr, mjacob@feral.com>
From: Matthew Jacob <mjacob@feral.com>
List: tech-kern
Date: 07/02/1998 23:38:35
>I'm not sure this means more code duplication: If the semantic you propose is
>fully implemented, it would mean the generic error handler would never
>print the sense if there is a device-specific error handler, thus you
>would have to print the sense in each specific error handler ...

It just means that a device specific handler has to handle things
fully if it wants to control (i.e., by not returning -1) the
returned to generic handler.