Intel Linux Graphics On Ubuntu Still Flaky

[Craig73]

Out of interest, how do you suggest you provide a quantative answer to the "Is Ubuntu 9.10 looking to be faster or slower than 9.04 for Intel Users?".

Why does this question need to be answered on an alpha? (From a publication stand point... I can certainly see it's value from a build quality perspective, but this should be more automated and focused on identifying bottlenecks and regressions, not a "review")

[mtippett]

Thanks, Michael Larabel for continuing to post meaningless benchmarks and drawing wrong conclusions.

Out of interest, how do you suggest you provide a quantative answer to the "Is Ubuntu 9.10 looking to be faster or slower than 9.04 for Intel Users?".

[7oby]

Thanks, Michael Larabel for continuing to post meaningless benchmarks and drawing wrong conclusions.

From my experience on Arch with Linux 2.6.30, I get 850fps on glxgears with 2.7 and 550 with 2.8.

There is a technical reason of why 2.7 (I assume EXA/DRI1) performs faster in glxgears than 2.8 (UXA/DR2):

The difference between DRI1 and DRI2 is due in part to the context switch necessary to get buffer swap commands from the DRI2 application to the X server which owns the ?real? front buffer. For an application like glxgears which draws almost nothing, and spends most of its time clearing and swapping, the impact can be significant (note, glxgears is not a benchmark, this is just one of many reasons). On the other hand, having private back buffers means that partially obscured applications will draw faster, not having to loop over clip rectangles in the main rendering loop.

http://keithp.com/blogs/Sharpening_the_Intel_Driver_Focus/

But as Keith already mentions: glxgears is not a benchmark. But maybe you need one more:

Nobody measures the performance of "draw the same triangles over and over". And if someone does, (by seriously quoting glxgear fps numbers, for example), then everybody gets a good laugh. In fact, the phrase "glxgears is not a benchmark" is a catchphrase among 3D developers.

http://www.cworth.org/intel/performance_measurement/

Now that we laughed about you and Michael for drawing conclusions from this same benchmark (just ported to QT4):

http://zrusin.blogspot.com/2008/08/fast-graphics.html

Let's consider some serious aspects. We need real application benchmarks such as game engines, firefox rendering or compiz performance. But even in this case, performance may degrade from one version to another. If that's the reason one should check whether one of these reasons apply:

. I noticed in Urban Terror, when performing Mesa benchmarks, that in certain revisions of the Mesa Stack on x3100 GM965 hardware, some visual effects were not drawn e.g. some lightning or the shot of your weapon. If it's not drawn the entire scene draws faster (though with slight visual corruption or flaws) and you get higher fps reports.

. In some of the recent versions of the intel driver tearing disappeared. I'm not aware of in which version exactly and whether it applies only to indirect or direct rendering, but this might may a dfficerence. I remember Jesse talking about double buffering:

http://virtuousgeek.org/blog/index.php/2009/05/

I don't know whether this entered mainline or whether it's related. BUT: I don't see tearing with 2.8.0 in compiz enabled X-Org anymore and I know that double buffering has an impact on reported framerates (compared to tearing withough v-sync):

http://www.anandtech.com/video/showdoc.aspx?i=3591&p=3

--

Michael, please remove QGears2 benchmarks from the benchmark suite that you use for performance evaluation. That applies to the GTK benchmarks as well that you used in this article, which I critized already some time ago:

http://www.phoronix.com/forums/showthread.php?t=17726

[mtippett]

Can PTS run system updates every day at a given time, immediately run benches, and then make line graphs of a release day over day? I think that kind of thing would be fascinating to see.

http://phoromatic.com/

Continous integration and ongoing performance management is something I would dearly love to see lots of OSS projects pick up.

Matt

[ethana2]

..plus that would allow for articles like "Ubuntu 9.10 3d performance jumps 32% today with xorg intel driver update"

[ethana2]

Can PTS run system updates every day at a given time, immediately run benches, and then make line graphs of a release day over day? I think that kind of thing would be fascinating to see.

[pvtcupcakes]

Have you tried the driver included in the kernel, b43?

According to http://linuxwireless.org/en/users/Drivers/b43 my chip isn't supported by the kernel driver.

My PCI ID is 14e4:4315.

[slumbergod]

I hope they get it right for Koala

Jaunty was the worst release for me so far. I experience terrible graphics performance (yes, Intel chipset) and regressions with sound (low volume, no physical volume control).

Given the popularity of Intel chipsets on low cost laptops it is hard to believe Jaunty was released - a good argument for releasing when something is ready rather than a regular date.

I hope Koala shapes up to be a better release but this time around I will wait before doing a clean install to see if the important things are fixed.

[Rip-Rip]

[...]

I haven't been able to try 2.6.31 on that system because I wasn't paying attention when I bought the laptop and I ended up with a broadcom wifi chip, and the driver from AUR doesn't support 2.6.31 yet.

Have you tried the driver included in the kernel, b43?

[pvtcupcakes]

From what I've heard on the Arch Linux forums is that 2.8 runs much faster on the 2.6.31 kernel than it does on .30.

From my experience on Arch with Linux 2.6.30, I get 850fps on glxgears with 2.7 and 550 with 2.8.

I haven't been able to try 2.6.31 on that system because I wasn't paying attention when I bought the laptop and I ended up with a broadcom wifi chip, and the driver from AUR doesn't support 2.6.31 yet.