On Sat, 2016-01-30 at 16:59 +0000, Koul, Vinod wrote: > On Tue, Jan 19, 2016 at 08:57:48AM -0700, Dave Jiang wrote: > > The timer_event() function seems to have a bug where it ends up > > marking the > > last entry as non-responding and eventually attempts to restart the > > channel. This also continuously happen when idle. What needs to > > happen is > > for us to make sure there are no descriptors active and then handle > > that > > case properly. We should only hit the "cleanup" stage if there are > > still > > active descriptors. > > Applied now and for Nth time pls use the right subsystem name > Vinod, I'm not seeing this patch in the slave-dma/topic/ioatdma branch. ��.n��������+%������w��{.n��������)�)��jg��������ݢj����G�������j:+v���w�m������w�������h�����٥