Uh, It's xxx->resid that's supposed to be set. Seems like this a bug in siop2.

-matt


On Sun, 2 Jun 2002, Ignatios Souvatzis wrote:

> Hi,
> 
> I noticed that the Amiga siop2 driver doesn't handle ignore wide residue
> messages. What's worse, it seems to synthesize the data length passed up
> to the mid-layer, setting xs->resid to 0 (with a big /* XXXX */), thus
> pretending the full data where transferred if all went well. This might
> work for normal disk transfers, but is certainly a dangerous assumption
> when looking at, say, VPP data or when reading from tape drives...
> 
> How is a driver supposed to tell the mid layer how many bytes where read 
> or written?
> 
> Regards,
> 	-is
> -- 
> seal your e-mail: http://www.gnupg.org/
>