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 Huge Disaster Within The Linux 2.6.35 Kernel

Michael Larabel

Published on 28 May 2010
Written by Michael Larabel
Page 3 of 5 - 184 Comments

The performance setback with the computational biology HMMer test is huge. The Atom Z520 setup took 87.5% more time since the regression(s) was introduced while the Atom 330 had a lower but still significant 30% change.

The performance drop witnessed by the LU.A test within NASA's NAS Parallel Benchmarks is hugely devastating and the most significant. With the kernel code prior to last week the LU.A test consistently had around ~640 Mop/s but after the 22nd of May, this number has since become around ~215 Mop/s.

While you can find many more examples of this severe 2.6.35 regression on the Phoromatic Tracker for the Linux kernel, the last one we are sharing in this article is with the x264 video encoding test. Both of the systems here had dropped by very significant margins on the same dates as the others.

Since there is still some people that do not seem to trust the results of the Phoronix Test Suite / Phoromatic for whatever reasons, we manually confirmed this very serious Linux 2.6.35 kernel problem on an entirely different system. For those not familiar with our automated testing software, each test is run multiple times and if the standard deviation between the runs exceeds a predefined threshold, the run count will dynamically increase. Additionally, all included test profiles are designed to be deployed in a turnkey fashion and be fully reproducible (see The Five Stages of Benchmark Loss for some historical information).

For verifying this Linux 2.6.35 kernel problem we used a ZaReason Verix notebook that we are currently reviewing we did a clean installation of Ubuntu 10.04 LTS (x86_64) and then using the Ubuntu mainline daily kernel PPA had installed the 64-bit kernels from the 20th of May and then again from the 26th. For reference, there are no major changes in the kernel configuration between the two dates. We ran a few tests to just verify this is indeed a real, major regression and is not isolated to a couple Atom systems. The ZaReason Verix boasts an Intel Core i7 Q720 processor that's clocked at 1.60GHz with four physical cores plus Hyper Threading, has 6GB of system memory, sports an 80GB Intel SSD, and its display is powered by a NVIDIA GeForce GTS 250M graphics processor. Ubuntu 10.04 LTS was left at its defaults, including the EXT4 file-system. This notebook is on the completely different end of the hardware spectrum from the Atom computers that first illustrated this problem.

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