Some Of The AMDGPU Changes Being Worked On For Linux 4.8

Written by Michael Larabel in Radeon on 26 June 2016 at 10:39 AM EDT. 26 Comments
RADEON
The AMD developers still have a few more weeks to get their new feature material ready for the Linux 4.8 kernel while here is an early look at some of the code merged so far.

One of the changes we're looking forward to most with the AMDGPU DRM of Linux 4.8 is the OverDrive overclocking support. Finally the ability with the open-source AMD stack to overclock your GPU easily, but it's only supported for AMDGPU-capable hardware. There are commits though in the 4.8 W.I.P. branch for enabling the overclocking for Sea Islands with that experimental AMDGPU support. Another addition since the original AMDGPU overclocking support is there's now support for video memory overclocking too. Similar to the GPU core re-clocking, the memory overclocking can be done up to 20% in 1% steps.

Another change being looked for with Linux 4.8 AMDGPU is the experimental GCN 1.0 support for allowing the original GCN GPUs of the Radeon HD 7000 series (and re-brands in the Rx 200/300 series) to work with this newer DRM driver and thus too the AMDGPU-PRO driver with Vulkan support. However, for Linux 4.8 this support would still be very experimental and disabled by default: while I've tried the GCN 1.1 experimental AMDGPU support, I haven't yet given the GCN 1.0 support a go but will try it out assuming it's all tidied up for Linux 4.8.

Some other work queued up currently in the drm-next-4.8-wip branch includes:

- Various PowerPlay improvements and fixes, a seemingly never-ending task with GPU power management only becoming increasingly complicated.

- Various bits of hybrid platform code and other changes around hybrid GPU laptops... Perhaps for seeing more dGPU+iGPU hybrid integration in the future?

- VCE changes and other alterations.

Not part of drm-next-4.8-wip is the DAL Display Abstraction Layer code as the big changes for making the display code more like AMD's binary driver... The DAL code is necessary for eventually supporting FreeSync/AdaptiveSync, HDMI 2.0, and other newer display-related features. It's not clear though if the cleaned up DAL code will be ready for Linux 4.8 or not: for now it's just in its own branch.

Stay tuned for more AMDGPU Linux 4.8 coverage (and benchmarks!) over the weeks ahead.
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