On Wed, Aug 10, 2011 at 6:27 PM, Felipe Contreras <felipe.contreras@xxxxxxxxx> wrote: > On Wed, Aug 10, 2011 at 9:42 PM, Felipe Contreras > <felipe.contreras@xxxxxxxxx> wrote: >> I am trying to use a more recent version of the tidspbridge code in >> the Nokia N9, but I'm stuck with this warning that is caused by using >> the dm timer framework. > > Actually, the problem only happens on the 'dspbridge' branch, which > has some unmerged patches. These patches introduce some new mutexes > that trigger this. The only lock introduced is a spin_lock_bh which makes sense to warn if the underlying functions have a mutex, but I couldn't find the clk_set_parent mutex which is causing this on the omap files. > My proposed solution is to request the dm timers at initialization > time, and just enable/disable them on wake/hibernate, which is a bit > similar to what was happening before, and probably more efficient and > proper. When I made these changes I wanted to avoid keeping an array with the clocks requested and nobody else being able to use them (even if that meant a conflict between two drivers trying to use the same clock, but other warnings could help to point to such cases). Also I wanted to maintain uniformity, not just requesting some clock, freeing them and then some others to be held and freed only on module removal. That said, I have no problem on changing to your approach if needed. Regards, Omar > > I'm attaching the patch. > > -- > Felipe Contreras > -- To unsubscribe from this list: send the line "unsubscribe linux-omap" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
