R600 Gallium3D Patch Boosts Unigine By ~30%

[phoronix]

Phoronix: R600 Gallium3D Patch Boosts Unigine By ~30%

A rather simple patch by Vadim Girlin has led to a reported significant performance improvement within the R600 Gallium3D graphics driver...

http://www.phoronix.com/vr.php?view=MTMwNDY

[Drago]

And right after:

Please ignore this, there are some issues with this patch undetected by
piglit.

I suspect it's something similar to the hw bug discussed here recently:

http://www.mail-archive.com/mesa-dev.../msg33361.html

[archibald]

He's posted a v2 patch now.

[Drago]

I am curious, how many spots like this, there are in r600g at the moment.

[Xake]

He's posted a v2 patch now.

Link to patch:
http://lists.freedesktop.org/archive...ry/034728.html

No a nice one-liner anymore tho.

[przemoli]

Well, AMD could host farm of GPU's of various generations/configs. With some automated tools that would allow previlaged devs to submit patches to test performance/piglit regresions.

QA would also had easier time (if dev's could detect bad code earlier).


Great idea, so why its not already done?

[Nille]

Well, AMD could host farm of GPU's of various generations/configs. With some automated tools that would allow previlaged devs to submit patches to test performance/piglit regresions.

QA would also had easier time (if dev's could detect bad code earlier).


Great idea, so why its not already done?

Or Stop developing the Driver in the Kernel Tree. For an user its much more easy to update an out of tree kernel module. To tell an user he has to recompile his kernel is is a mess.

[GreatEmerald]

Well, AMD could host farm of GPU's of various generations/configs. With some automated tools that would allow previlaged devs to submit patches to test performance/piglit regresions.

From what I recall, they don't have such resources to spare. Though who knows, maybe the situation has changed.

Or Stop developing the Driver in the Kernel Tree. For an user its much more easy to update an out of tree kernel module. To tell an user he has to recompile his kernel is is a mess.

Not for Gentoo users it isn't!

[agd5f]

Well, AMD could host farm of GPU's of various generations/configs. With some automated tools that would allow previlaged devs to submit patches to test performance/piglit regresions.

QA would also had easier time (if dev's could detect bad code earlier).


Great idea, so why its not already done?

What automated tools would you suggest? There weren't any piglit tests which triggered the issue. Unfortunately, automated testing is often a challenge for GPU development. For best results you generally need physical access to the hardware.

[Veto]

No a nice one-liner anymore tho.

Yes, now it is an even nicer patch, as it in total removes 24 lines from the driver

Thanks for the link.