1. Computers
  2. Display Drivers
  3. Graphics Cards
  4. Memory
  5. Motherboards
  6. Processors
  7. Software
  8. Storage
  9. Operating Systems


Facebook RSS Twitter Twitter Google Plus


Phoronix Test Suite

OpenBenchmarking.org

The Big RadeonSI Performance Patch To Land In Linux 3.13

AMD

Published on 01 January 2014 06:24 AM EST
Written by Michael Larabel in AMD
24 Comments

Just before Christmas was a patch by Marek Olšák that provided much better RadeonSI performance. That patch will now be merged into the Linux 3.13 kernel to be released later this month.

The patch for the Radeon HD 7000 series graphics processors and newer sharply improves the OpenGL performance for some workloads. The open-source AMD Linux driver finally correctly sets up the render back-ends for the hardware. It turns out the open-source driver was only enabling the render back-end on the first shader engine for modern Radeon GPUs rather than eabling all the render back-ends.

With this simple fix, the performance improved sharply for some workloads. The RadeonSI performance went up by 6x for Marek with the Unigine Sanctuary benchmark as an example.

At the time the patch was published we didn't know if it would go into the current Linux 3.13 kernel or be held off until the Linux 3.14 merge window, but it's just been called for landing in 3.13. This performance boosting patch is technically a fix and Red Hat's David Airlie went ahead today to send in a batch of Radeon DRM fixes for Linux 3.13.

There's a handful of open-source AMD Radeon kernel driver patches and among them is a fix for a UVD check, Sea Islands / Hawaii fixes, and some other minor work. Assuming Linus is okay with the pull request, it will be part of Linux 3.13-rc7 when released this forthcoming weekend. The stable release of Linux 3.13 is expected mid-January. New RadeonSI benchmarks will come on the Linux 3.13 kernel when I'm back from Russia.

About The Author
Michael Larabel is the principal author of Phoronix.com and founded the web-site in 2004 with a focus on enriching the Linux hardware experience and being the largest web-site devoted to Linux hardware reviews, particularly for products relevant to Linux gamers and enthusiasts but also commonly reviewing servers/workstations and embedded Linux devices. Michael has written more than 10,000 articles covering the state of Linux hardware support, Linux performance, graphics hardware drivers, and other topics. Michael is also the lead developer of the Phoronix Test Suite, Phoromatic, and OpenBenchmarking.org automated testing software. He can be followed via and or contacted via .
Latest Linux Hardware Reviews
  1. MSI X99S SLI PLUS On Linux
  2. NVIDIA GeForce GTX 970 Offers Great Linux Performance
  3. CompuLab Intense-PC2: An Excellent, Fanless, Mini PC Powered By Intel's i7 Haswell
  4. From The Atom 330 To Haswell ULT: Intel Linux Performance Benchmarks
Latest Linux Articles
  1. RunAbove: A POWER8 Compute Cloud With Offerings Up To 176 Threads
  2. 6-Way Ubuntu 14.10 Linux Desktop Benchmarks
  3. Ubuntu 14.10 XMir System Compositor Benchmarks
  4. Btrfs RAID HDD Testing On Ubuntu Linux 14.10
Latest Linux News
  1. Fedora 21 Beta & Final Release Slip Further
  2. Mesa 10.3.2 Has A Couple Bug-Fixes
  3. RadeonSI/R600g HyperZ Support Gets Turned Back On
  4. openSUSE Factory & Tumbleweed Are Merging
  5. More Fedora Delays: Fedora 21 Beta Slips
  6. Mono Brings C# To The Unreal Engine 4
  7. Coreboot Now Has Support For Intel Broadwell Hardware
  8. Enlightenment's EFL 1.12 Alpha Has Evas GL-DRM Engine, OpenGL ES 1.1 Support
  9. GTK+ Lands Experimental Backend For Mir Display Server
  10. Ubuntu 14.10 Officially Released
Latest Forum Discussions
  1. HOPE: The Ease Of Python With The Speed Of C++
  2. Updated and Optimized Ubuntu Free Graphics Drivers
  3. Ubuntu 16.04 Might Be The Distribution's Last 32-Bit Release
  4. Linux hacker compares Solaris kernel code:
  5. Advertisements On Phoronix
  6. Users/Developers Threatening Fork Of Debian GNU/Linux
  7. AMD Releases UVD Video Decode Support For R600 GPUs
  8. Proof that strlcpy is un-needed