* Tero Kristo <t-kristo@xxxxxx> [160901 23:01]: > On 02/09/16 05:52, Adam Ford wrote: > > Tony, > > > > There were a bunch of patches from TI that were submitted, but I don't > > see any followup. > > > > For example, > > https://patchwork.kernel.org/patch/7257051/ > > This series has been abandoned for now. > > > I know they are a year old, but I am trying to get caught up on how > > the device tree stuff works regarding the reset, and how hwmods relate > > to the device tree, etc. > > Currently, the relationship is pretty much missing. There are plans to build > a new reset driver implementation on top of the interconnect driver that > Tony is working on. However, this is going to take some time to get into > mainline still, as there are missing dependencies like this one which > requires major rework also: > http://www.spinics.net/lists/arm-kernel/msg515300.html Yeah any standalone Linux driver module under drivers/ that implements the necessary resets using Linux reset framework should be OK to do. When we have a more generic reset driver available switching to use that should be trivial. The omap_device and omap_hwmod calls needed should be passed to it in platform_data using pdata-quirks.c. Similar to what drivers/pwm/pwm-omap-dmtimer.c is doing with the dmtimer callback functions. What I don't want to see is more driver like features stuffed into mach-omap2 though :) > > I was looking through some of the TI drivers, and it looks like they > > are applying some patches to add SGX to the device tree for the AM33x > > and AM43x but none of that stuff has made it to the mainline. similar > > concepts seems reasonable to me to place into the 3430/3630 DTSI > > files, but there seem to be different schools of thought, and I don't > > fully understand the big picture. > > DT is a beast to tackle, I'm not sure if anybody has the grasp of the big > picture of it... We really should get the mainline kernel to handle all the resets, clocks and power for SGX in a way that works with and without the acceleration. Then the SGX firmware or module is just an optional piece that may or may not be available. > > I have some patches that I'd let to post for feedback, but the patches > > still seem to be missing something (since the driver complains about > > NULL pointers) and some of the entries into the hwmod and confusing to > > me but I have tried mapping them to the TRM. > > I guess you could just post the patches as RFC to get some feedback. Yeah it seems there's nothing blocking doing a reset driver for SGX. Regards, Tony -- 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
