Hi, On Sun, Aug 14, 2011 at 2:44 AM, Felipe Contreras <felipe.contreras@xxxxxxxxx> wrote: > Yeah, but with the current approach it would be possible that > everything works fine, and then the DSP goes to hibernation, a module > is loaded that uses one dm timer, and then when the DSP wakes up, that > timer would fail, there isn't even a check for that. If I follow the > code correctly, when the DSP goes back to hibernation, there would be > a crash, as a NULL timer would be freed. I think that's too error > prone. Yes, this scenario could trigger such error condition. So, FWIW, I'm fine with just a check or your approach. If possible to apply the same mechanism for the other clocks, so in the enable/disable functions all the clocks are then enabled/disabled avoiding mixed behavior of enable/request, disable/free between gpt and mcbsp clocks. Regards, Omar -- To unsubscribe from this list: send the line "unsubscribe linux-omap" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
