Re: sata_mv on Orion: CONFIG_SATA_PMP=y needed even when no multiplier

[Mark Lord <liml@rtr.ca>]

Mark Lord wrote:
> Tejun Heo wrote:
>> Mark Lord wrote:
> ..
>>> I suppose we could just always ask for the extra soft-reset here (?)
>>
>> Hmm... sata_mv doesn't do device classification, right?  If so, it 
>> should return -EAGAIN on success to tell EH to perform follow-up SRST.
> ..
> 
> Right now, mv_hardreset() is just a wrapper around sata_link_hardreset(),
> and it just returns whatever it gets back from sata_link_hardreset().
> 
> And sata_link_hardreset() only returns -EAGAIN for hosts that support PMP.
> 
> Should sata_link_hardreset() instead always return -EAGAIN instead of 0 ?
> Because that's essentially what we'd be doing by changing mv_hardreset()
> to do that.
...

Mmm.. or maybe sata_mv should do what sata_std_hardreset() does:

   rc = sata_link_hardreset(link, timing, deadline, &online, NULL);
   return online ? -EAGAIN : rc;

Or, for that matter, perhaps sata_mv should just be calling sata_std_hardreset()
instead of sata_link_hardreset().   Those all got changed around at the same time
that I was working on sata_mv, so I might not have it quite right there (?).

Cheers