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

Auto-Vectorizing LLVMpipe For Performance Gains

Mesa

Published on 03 June 2012 01:58 PM EDT
Written by Michael Larabel in Mesa
7 Comments

Here's a hint that may allow for some notable performance gains out of the Gallium3D LLVMpipe driver for multi-core systems.

Recently I wrote about the Gallium3D LLVMpipe Driver On LLVM 3.1. There wasn't really any major performance changes to note for this Gallium3D "software rasterizer" driver that uses the CPU and not the GPU as a means of fallback support, but this Sunday I received a message from Hal Finkel. This developer from the Argonne National Laboratory wrote the new basic-block auto-vectorization pass for LLVM 3.1. Below is the message he sent over to me regarding this work and LLVMpipe.
Michael,

I looked through your benchmark results on "Gallium3D LLVMpipe Driver On LLVM 3.1", and, as author of LLVM's new basic-block autovectorization pass, I wonder if turning on autovectorization would improve the 3.1 results.

If you're curious about the autovectorization pass, you can see my presentation at the developer's conference earlier this year: http://llvm.org/devmtg/2012-04-12/

It seems like llvmpipe uses its own pass manager, which is probably not a bad thing, but it is either intentionally minimalistic, or is pretty out of date (or both). There are many other optimization passes available (see PassManagerBuilder::populateModulePassManager in lib/Transforms/IPO/PassManagerBuilder.cpp to get an idea of what is available). It looks like the current pass manager used by llvmpipe is in create_pass_manager in src/gallium/auxiliary/gallivm/lp_bld_init.c and it adds only a few optimization passes.

To add autovectorization to llvmpipe, I think adding something like this after the call to LLVMAddGVNPass(gallivm->passmgr); will work:

#if HAVE_LLVM >= 0x301
LLVMAddBBVectorizePass(gallivm->passmgr);
/* BBVectorize depends on other passes for cleanup... */
LLVMAddInstructionCombiningPass(gallivm->passmgr);
LLVMAddEarlyCSEPass(gallivm->passmgr);
#endif

I've never really used the C API, and I'm not really sure how to pass tuning parameters through (and some tuning might be helpful or necessary here -- integration between the optimization passes and the backends is not as good yet as it should be), but this would be a start.

Thanks in advance,
Hal

P.S. Please feel free to forward or repost this message elsewhere if you feel that would be helpful or more appropriate.
I'm still out of the country on this nearly month-long business trip, but hopefully a few Phoronix readers would be interested in poking around with making these code changes as it could yield some interesting performance gains for this promising GPU fallback driver. LLVMpipe is now used as the fallback for Ubuntu and Fedora when no hardware GPU driver is available.

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. Intel Xeon E5-1680 v3 & E5-2687W v3 Compared To The Core i7 5960X On Linux
  2. Intel 120GB 530 Series SSD Linux Performance
  3. Btrfs/EXT4/XFS/F2FS RAID 0/1/5/6/10 Linux Benchmarks On Four SSDs
  4. AMD's Windows Catalyst Driver Remains Largely Faster Than Linux Drivers
Latest Linux Articles
  1. NVIDIA vs. Nouveau Drivers With Linux 3.18 + Mesa 10.4-devel
  2. Is The Open-Source NVIDIA Driver Fast Enough For Steam On Linux Gaming?
  3. Linux 3.18 File-System Performance Minimally Changed But Possible Regressions
  4. AMD Radeon Gallium3D Is Catching Up & Sometimes Beating Catalyst On Linux
Latest Linux News
  1. Linux 3.18 Kernel: Not Much Change With Intel Haswell Performance
  2. More File-System Tests Of The Linux 3.18 Kernel
  3. Using NVIDIA's NVENC On Linux With FFmpeg
  4. There's Talk Again About An "Open To The Core" Ubuntu Laptop
  5. PowerVR SGX Driver Code Gets Leaked
  6. V2 Of KDBUS Published For Linux Kernel Review
  7. VirtualBox 4.3.20 Arrives, Still No Sign Of VirtualBox 4.4
  8. Scientific Linux 6.6 vs. Scientific Linux 7.0 Benchmarks
  9. Qualcomm Looks To Get Into The ARM Server Business
  10. HHVM 3.4 Adds New Features, Support
Latest Forum Discussions
  1. Roadmap to Catalyst 14.10 ?
  2. Updated and Optimized Ubuntu Free Graphics Drivers
  3. Cant get working Kaveri APU - A10-7850k
  4. Debian Developer Resigns From The Systemd Maintainership Team
  5. Script for Fan Speed Control
  6. Debian Init System Coupling Vote Results
  7. The Slides Announcing The New "AMDGPU" Kernel Driver
  8. Ubuntu Developers Still Thinking What To Do About Adobe Flash Support