On Sun, May 13, 2012 at 07:04:42PM +0100, Mark Brown wrote:
> Rather than using the pointer passed back by the regmap API (or complaining
> because that wasn't actually being set) the da9052 driver was having some
> fun and games peering through genirq and regmap internals. Fix the driver
> to use the API as expected.

Sorry, I also meant to say that since this fix seems fairly obvious and
blocks some other regmap work I've gone ahead and dropped this into my
regmap tree immediately at

   git://git.kernel.org/pub/scm/linux/kernel/git/broonie/regmap.git topic/irq

Hopefully this should be OK - please let me know if there's any issue.