> -----Original Message----- > From: Paul Walmsley [mailto:paul@xxxxxxxxx] > Sent: Wednesday, October 13, 2010 8:36 PM > To: Nayak, Rajendra > Cc: linux-omap@xxxxxxxxxxxxxxx; khilman@xxxxxxxxxxxxxxxxxxx; Cousson, Benoit > Subject: Re: [RFC 0/8] Split powerdomain framework into plat specfic/independent > > Hi Rajendra, > > On Thu, 23 Sep 2010, Rajendra Nayak wrote: > > > OMAP4 powerdomains have some inherent differences as compared > > to OMAP2/3 powerdomains, starting with register offsets being different > > to clubbing of multiple controls into one register and in some cases > > splitting of control into multiple registers. > > There are also new features like lowpowerstatechange bits and features > > like HW SAR which are no longer present in the older form. > > > > Supporting all these in the existing powerdomain framework would mean adding > > a lot of cpu_is_* checks which makes code unmaintainable going fwd. > > > > Hence this RFC series is an attempt to split the existing powerdomain framework > > into platform independent part (which does error checking, usecounting et al) > > which can be reused across OMAP's and hook up platform specific functions to > > do low level programming which varies across OMAP's. > > > > The series has a dependency on the following patch series > > http://marc.info/?l=linux-omap&m=128146137929118&w=2 > > http://marc.info/?l=linux-omap&m=128464495603506&w=2 > > > > This series along with all the dependent series patches can be > > found here > > git://gitorious.org/omap-pm/linux.git powerdomain-split > > I'm happy with the basic approach here, to split the powerdomain code via > function pointers. There were some minor comments on those patches; > you've probably seen them already. Similar issues exist in some of the > other powerdomain split patches (e.g., many functions should be static, > etc.) Also please make sure the changes don't add any sparse warnings; I > noticed some when doing a test build: > > arch/arm/mach-omap2/powerdomains.c:52:5: warning: symbol > '_get_mem_bank_onstate_mask' was not declared. Should it be static? > arch/arm/mach-omap2/powerdomains.c:72:5: warning: symbol > '_get_mem_bank_retst_mask' was not declared. Should it be static? > arch/arm/mach-omap2/powerdomains.c:92:5: warning: symbol > '_get_mem_bank_stst_mask' was not declared. Should it be static? Hi Paul, Thanks for the review. I'll fix all the issues you reported in the various patches before I repost again. Will also fix all the sparse warnings too. > > ... > > Aside from those minor issues, the other major issue to clear up is the > PRM/CM API for OMAP4 that is referenced in your mail, which of course will > impact these patches too. Will take a close look at those - Ok, thanks, I was about to check with you if you had a look at those as well. Thanks, will wait for your feedback. Regards, Rajendra > > > - Paul -- 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
