On Sat, Jul 11, 2015 at 09:46:21AM +0200, Wolfram Sang wrote: > So, I picked up the old series from Ian and reworked it significantly. While > his series was already an improvement from the driver found in the BSP, it was > still not ready for upstream. It should be now :) My idea was to start simple > and improve incrementally, so e.g. clock handling was broken, so clocks are > always on for starters. Major changes include: > > * clock handling simplified and adapted to current EMEV clock handling > * switch from wait_event to completion mechanism. The old one was basically > some kind of completion with custom code around wait_events > * dropped using signals; they always cause trouble with I2C > * simplified the bus free logic. If it turns out to be too simple, we'd need > to switch to the bus_recovery infrastructure probably. But we'd need a test > case for that > * error handling improved to match subsystem standards > * removed some BSP angst code > * lots of refactoring to make code simpler, more readable... > * bugfixes > > I also added all the glue code to add clocks, disable reset etc which was > missing before. I'd prefer to take the first three patches via i2c myself, > and the latter two via Simon's tree. > > Since V1, there are minor updates fixing issues found in reviews. Thanks! > > A branch for testing can be found here: > > git://git.kernel.org/pub/scm/linux/kernel/git/wsa/linux.git renesas/emma-i2c > > Please test, comment... > > Thanks! > > Wolfram > > Wolfram Sang (5): > clk: shmobile: emev2: deassert reset for IIC0/1 > i2c: emev2: add binding documentation > i2c: emev2: add driver Those patches applied to for-next... > ARM: shmobile: emev2: add IIC cores to dtsi > ARM: shmobile: emev2: kzm9d: enable IIC busses ... so Simon, you can pick up these, I'd say. Thanks, Wolfram
Attachment:
signature.asc
Description: Digital signature
