Does A Greedy Intel Driver Improve Performance?
Phoronix: Does A Greedy Intel Driver Improve Performance?
As we have outlined before and shared benchmarks of in the past, the Intel graphics driver stack has been going through some significant changes. The Intel graphics driver now has a proper memory manager in the form of the Graphics Execution Manager, there is upstream kernel mode-setting support, and a new 3D component is coming soon in the form of Gallium3D. With all of this invasive work going on, regressions are currently prevalent from stability problems to graphical corruption to slower 2D performance. While these are problems users will face with the new distribution updates in H1'09, some have been trying out different driver configurations in order to circumvent the situation. Canonical, for example, had been toying with the idea of enabling greedy migration heuristics by default.
There is xf86-video-intel 2.7 around a corner (currently 126.96.36.1993), so why didn't you test performance using that version?
Ubuntu 9.04 is shipping with 2.6 series. And 2.6.3 is still the latest *stable* release.
Originally Posted by Zajec
And there's serious performance degredations present with it. I've got Jaunty on my eeePC 701 and previously with eeeBuntu NBR 2.0, it was able to actually play Caster with the settings trimmed back to minimal values. Now...heh...the thing is down into the single digits with no other changes other than the OS version and driver.
Originally Posted by Michael
If someone is interested :
I've made a DEB-package for driver-version 2.7.0. It is based on the latest Ubuntu-package (I've included the quirks-patches) and compiles cleanly with libdrm 2.4.5 (normally 2.4.6 is needed). I've disabled KMS because Ubuntu 9.04 doesn't provide a KMS-capable kernel.
On my Dell Latitude D505 (855GM-chipset) the driver works stable (EXA and UXA). At the moment there are only packages for the 32bit-version of Ubuntu because i don't own a 64bit-capable machine and i don't have the slightest idea how to crosscompile a package.
Here are the links to the 32bit-packages, the source-code and my patch for the source-code :
Unfortunately, this is _not_ the only issue with displays. Ati users are also in trouble:
Of course this is not our (Ubuntu's) fault, but it still hurts.
Generally speaking, most users should run the upgrade a month after the release. Maybe it would be nice to have a .0 release for enthusiasts, and a .1 release about a month later for the general public. And let all these inevitable quirks be fixed in the time inbetween. The .0 release would be in a fixed schedule, and the .1 a bit more flexible.
If I had an Intel based IGP, I really wouldn't know what to use.
It's a shame that you only tested one, 4 year old, chipset. (support for different chips is at different states, so this really means nothing to users of 950,965(that's me),4500 and a few others)
What does "greedy" actually mean? What changes this option in the driver? thx
I assume it comes from the term "greedy algorithm" which uses a heuristic to "guess" what is going to be the optimal path through a series of options. google it for more info.
Originally Posted by bugmenot