[RFC/NOT FOR MERGING 1/5] arm: omap: fix up _od_suspend_noirq and _od_resume_noirq

Felipe Balbi [off-list ref] writes:

[...]

ps: I'll continue reading the code and further test my series to see
if I can understand what you say here.

OK.  And please get it working in cases where drivers are using I2C in
their suspend/resume (and even late/early) paths, and also where device
runtime PM is disabled from userspace.  All of that works with today's
code and will not work with your proposed code.

Kevin