On 11-05-30 09:07 AM, Bruce Stenning wrote:

quoted

quoted

Mark, did you get a chance to clean up the sata_mv spinlock addition in
mv_set_main_irq_mask?

Not yet.  I've been waiting to hear back from you as to
whether they visibly fix anything or not.

Having now tested with Tejun's two patches and without the spinlock in
mv_set_main_irq_mask for a reasonable time, the hotplug seems to be very
reliable.  So I think it's probably best to leave the spinlock out and
continue to observe the behaviour with further testing.

Peachy.  I'll have another look at some point -- I think I thought
the spinlock was still needed for ATAPI under some conditions.
Gotta look closer to make sure.

Thanks.