On Sun, Aug 4, 2013 at 12:38 PM, Maxime Ripard
[off-list ref] wrote:
The current code use no locking at all, which is obviously not that
great and can lead to concurrency issues, especially with the newer SMP
SoCs from Allwinner.
Add some locking where it's needed.
Signed-off-by: Maxime Ripard <redacted>
This does not apply to my fixes branch which is based off
v3.11-rc4.
It is also too big to go into fixes.
I think it may be requiring the other queued Allwinner things,
so it is better if I apply this on the devel branch after looking over
that Allwinner A20/A31 stuff?
Yours,
Linus Walleij