On Mon, May 10, 2004 at 03:14:49PM -0700, Philippe Troin wrote:
> Paul Stewart <[EMAIL PROTECTED]> writes:
> 
> > Hello Philippe.  I'm also over on the iRiver list, sharing in your
> > pain with H140 read errors under Linux.  It appears that at least
> > from the perspective of the USB storage bulk-only specification,
> > there are provisions for returning short responses.  However at the
> > very least the device should return a CSW, which appears to be 
> > absent if I read your trace correctly.
> 
> That's correct.  And now I know what a CSW is :-)
>  
> > It seems like we need to retry the transfer at a later point if 
> > the bulk transfer returns zero bytes initially.  It seems that an 
> > immediate retry doesn't work most of the time, but a try later on 
> > works.  This might take a bit of tweaking to put in...  Something 
> > like putting the thread back to sleep and waking it up later with 
> > the same request might work.
> 
> Do you have a patch?  I plan to return the H120 unit very soon since
> nobody seems to be interested in fixing this...

Retries should be generated by the SCSI layer...

Matt

-- 
Matthew Dharm                              Home: [EMAIL PROTECTED] 
Maintainer, Linux USB Mass Storage Driver

Dudes! May the Open Source be with you.
                                        -- Eric S. Raymond
User Friendly, 12/3/1998

Attachment: pgp00000.pgp
Description: PGP signature

Reply via email to