6 Jul
2012
6 Jul
'12
6:43 p.m.
On Fri, Jul 06, 2012 at 05:36:15PM +0100, Liam Girdwood wrote:
I don't think there is anything explicitly stopping a power off from happening during a stream shutdown atm without this patch. It was reported to me that this would fail 1 in 20 times on a certain device and that the above patch fixed the issue (1200 tests completed OK after the patch).
I do think we need this patch atm since we are accessing the same resources as the stream event.
So how is suspend and resume safe, for example? It's the same general path through the PM core.