> Hi,
> my previous patch was still not right, I forgot a case where b_resid would
> not be properly updated. I reworked the code, and now the code path used
> to handle drives which don't issue a save data pointer after the last data
> transfers is also used, in some case, for drives which do.
> This means that I can test part of the code on my system, and it's more
> likely to work on yours :)
> 
> Please test this new patch

my disk seems to start working with this patch.

YAMAMOTO Takashi