On 5 February 2014 15:34, Linus Walleij <linus.walleij@xxxxxxxxxx> wrote: > On Tue, Feb 4, 2014 at 4:58 PM, Ulf Hansson <ulf.hansson@xxxxxxxxxx> wrote: > >> Since the device is active while a successful probe has been completed, >> the reference counting for the clock will be screwed up and never reach >> zero. >> >> The issue is resolved by implementing runtime PM callbacks and let them >> handle the resources accordingly, including the clock. >> >> Cc: Alessandro Rubini <rubini@xxxxxxxx> >> Cc: Linus Walleij <linus.walleij@xxxxxxxxxx> >> Cc: Wolfram Sang <wsa@xxxxxxxxxxxxx> >> Signed-off-by: Ulf Hansson <ulf.hansson@xxxxxxxxxx> > > Hm do I read it right as patch 13 breaks runtime PM by leaving > the device active after probe() and this patch > 14 fixes it again? Maybe these two patches should be squashed > then. You are right; but the driver will still be working, you just don't get the benefit from inactivating the device at request inactivity - as you pointed out. The reason for why I wanted to do this as separate steps was to make it easier for reviewing, otherwise the patch(es) would have been quite big and messy. I am for sure open to adopt to your proposal, but just wanted to give you some more background, before I go ahead and send a v2. Kind regards Ulf Hansson > > Yours, > Linus Walleij -- To unsubscribe from this list: send the line "unsubscribe linux-i2c" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
