On (06/26/07 10:01), James Carlson wrote:
> 
> Sowmini.Varadhan at Sun.COM writes:
> > On (06/26/07 01:06), Peter Memishian wrote:
> > > While I can understand that from an implementation standpoint, from a
> > > usability standpoint I don't think a log message goes far enough.  The
> > > person running the dladm set-linkprop subcommand needs the feedback.
> > 
> > What we could do is to have a distinct error code to
> > track this case (EINTR?) that could be caught by dladm to emit 
> > the information to the user- would that suffice?
> 
> I have trouble with that on two fronts:
> 
>   - We shouldn't use errors to report success.  Telling the user that
>     the operation was successful, but that something else needs to be
>     done is something that the API itself needs to encompass, rather
>     than overloading an errno.  (Partial success is itself a
>     troublesome concept.)
> 
>   - EINTR already has a defined meaning.  If this isn't a case where a
>     potentially long-running operation was stopped and not restartable
>     due to a UNIX signal, then don't call it EINTR.
> 

We could fail the operation, but the question of finding the right
error code to be returned by the SETPROP ioctl remains: attempts
to modunload a used module seem to return EBUSY- we could return that
back to dladm and then have dladm react appropriately.

--Sowmini


Reply via email to