Samsung DRM Driver Could Make It Into The Kernel

Written by Michael Larabel in Linux Kernel on 30 August 2011 at 02:31 PM EDT. 1 Comment
LINUX KERNEL
Last week I wrote about Samsung releasing code to a new DRM driver for one of their ARM SoCs, the Exynos 4210 that's used by the Samsung Galaxy S II and other mobile devices. It looks like this open-source kernel driver from Samsung stands a chance as being the first ARM driver to be accepted into the DRM area of the mainline Linux kernel.

Samsung's Inki Dae sent another message last week to the mailing lists to target David Airlie in an attempt to achieve a response out of him. Inki was successful. David looked at the code and provided feedback. He didn't make any comments about reject the code outright based upon any missing open-source user-space driver suppot or anything else, like he has done in the past. His message commenting on Samsung's new open-source Direct Rendering Manager code is below.
I was sort of hoping someone else would take a look at these ARM drivers before me, it might be worth getting some inter-company review between the vendors submitting drm components as I'm guessing its going to be a lot of the same thing.

But I've done a quick review and some things that stick out.

1. include/drm/samsung_drm.h should only contain public information for use on the userspace ioctl interface, it shouldn't contain any kernel private data structures, they should be in drivers/gpu/drm/samsung/samsung_drv.h or something very similiar.

2. I'm not sure why samsung_drm_fn_encoder exists, it looks like from the crtc mode set functions you call the encoder mode set functions, is it not possible to use the helper drm_crtc_helper_set_mode so that the crtc mode set is called then the encoder ones without having the explicit calls? If not please either document it or point at the explaination I missed.

3. Not terribly urgent but is samsung the best name for this? what happens if you bring out another chip, I also think there is a lot of samsung_drm_ in function names that seems not that useful. dropping _drm_ might help.

4 ioctls: drm_samsung_gem_map_off needs explicit padding before the 64-bit, drm_samsung_gem_mmap needs explicit padding before the 64-bit,, though I'm not sure you need these ioctls all now that the dumb interface is upstream, what is usr_addr in the gem create ioctl for? this seems wrong, it also looks too short for 64-bit addresses, but passing in userspace addr to kernel is generally not a great plan.

5. you probably don't need master_create/set ops.

Dave.

Since that message he's also exchanged several more emails on the mailing list with Inki.

It's a different story from what he's said to Qualcomm or Texas Instruments in the past when they have published similar open-source DRM driver code for their ARM SoCs. It's a similar situation to VIA's failed attempts in the past at only partially open-source graphics stacks on Linux.

Let's see what happens here and whether the Samsung DRM driver manages to make it into the mainline Linux kernel.
Related News
About The Author
Michael Larabel

Michael Larabel is the principal author of Phoronix.com and founded the site in 2004 with a focus on enriching the Linux hardware experience. Michael has written more than 20,000 articles covering the state of Linux hardware support, Linux performance, graphics drivers, and other topics. Michael is also the lead developer of the Phoronix Test Suite, Phoromatic, and OpenBenchmarking.org automated benchmarking software. He can be followed via Twitter, LinkedIn, or contacted via MichaelLarabel.com.

Popular News This Week