On Wed, Nov 30, 2011 at 02:14:03PM -0800, Rakesh Iyer wrote:
Thanks Dmitry.
Since there can be multiple wake causes I wanted to isolate wake key
generation to the case where keyboard actually generated the wake
interrupt.
I do not think you can guarantee this though because if user touches
keyboard "too early", before your resume method had a chance to disable
kbc interrupt as a wakeup source, you are still going to get that
interrupt and deliver KEY_POWER even though KBC is not the actual wakeup
source.
So don't over-complicate it. If hardware can't detect actual key pressed
just emit KEY_POWER if a key was at any time between calls to
tegra_kbc_suspend() and tegra_kbc_resume().
BTW, could you please have your MUA wrap long lines around 75 column or so?
Thanks.
--
Dmitry