search

RobertCNelson / ti-linux-kernel-dev

vendor bsp...
MIT License
85 stars 74 forks source link

uio-pruss fixes #30

Closed mvduin closed 5 years ago

mvduin commented 5 years ago

presumably Daniel Kulp is still waiting for these :)

(see my mail on Wed, 13 Feb 2019 10:28:46 +0100 for notes on the patches)

RobertCNelson commented 5 years ago

@mvduin i merged these in last night:

https://github.com/RobertCNelson/ti-linux-kernel-dev/commits/ti-linux-4.19.y https://github.com/RobertCNelson/ti-linux-kernel-dev/commit/64222294c7cd812613c5df956e14eaeeca0abe04

should be in the apt repo as: 4.19.15-ti-r9

I need to patch AM335X-PRU-UIO-00A0.dts so it handles both cases..

https://github.com/beagleboard/bb.org-overlays/blob/master/src/arm/AM335X-PRU-UIO-00A0.dts

Regards,

mvduin commented 5 years ago

Note that the different way to do the DT structure for uio-pruss is not a difference between 4.19 and older kernels, but rather merely a suggestion meant to allow a greater degree of compatiblity with the remoteproc DT structure and allowing remoteproc to be enabled by default and allowing the overlay to switch to uio-pruss to be really simple and tiny (ideally only changing the "compatible" property). The old DT declaration should work fine on this kernel, and this alternative DT declaration should work fine on older kernels.

I saw on the mailing list that the overlay based on it doesn't seem to work, I'll need to look into why that's happening since I didn't see any obvious problem apart from a typo in the pruss_soc_bus node label.

However, if the DT structure for remoteproc-pru is going to be completely changed again, as the later discussion about mainline efforts for remoteproc-pru seem to indicate, this whole effort may be for nothing anyway.