On Mon, Jul 07, 2014 at 06:00:29PM +0100, Jonathan Cameron wrote: > On 02/07/14 14:50, Josef Gajdusek wrote: > > This patch changes hmc5843.c to use regmap. This provides transparent caching > > to the code as well as abstraction necessary to add support for SPI-based > > hmc5983. > > > > Signed-off-by: Josef Gajdusek <atx@xxxxxxxx> > Hi Josef, > > A few little bits and pieces inline. > Peter, could you also take a look at this series? > > Thanks, > > Jonathan Hello, > Whilst minor, we do have a slight change in how this works in the > event > of an error. I don't particularly mind it, but perhaps should have > been broken out for separate discussion. It could be changed to use _check version of the regmap function (all other update_bits calls then probably should be changed as well). Though that would double the amount of data sent and be of any use only on defective chip. > Why move these? You now have the device going to sleep whilst > the userspace interface is still available... strange, I remember this causing issues. Now I am unable to reproduce them so it was probably something else. Do you want me to resend or can you apply without these lines swapped? _______________________________________________ devel mailing list devel@xxxxxxxxxxxxxxxxxxxxxx http://driverdev.linuxdriverproject.org/mailman/listinfo/driverdev-devel
