Big kudos to Alan for handling this. He's already patched a bunch of these and is streaming patches out for others.
Type: Posts; User: mattst88; Keyword(s):
Big kudos to Alan for handling this. He's already patched a bunch of these and is streaming patches out for others.
It's just a toy.
No, it's a toy, as acknowledged by its author.
Seriously not trolling here, but I think this is the most important part of your post. Gallium isn't some silver bullet. It actually is a bad fit for i915 since Gallium doesn't have any ability to...
How many 3D drivers have you written?
Please don't answer if you're not sure. (Maybe being sure is the actual problem)
The i915g driver is for Gen3 hardware. Gen4, 5 (Ironlake), 6 (Sandy Bridge), 7 (Ivy Bridge), et al. is i965 and...
X.Org is in GSoC this year. Check again. Please retract this article before people believe it.
The softpipe driver is much slower than swrast.
Just a heads-up: I've read that shipping the binary kernel modules is not okay according to the GPL. Other Live CD projects have run afoul of this in the past. See...
How do you handle the kernel module portions of the proprietary graphics drivers? Compile at boot-time?
Turn it off so you get a useful graph. It's like this in /every/ article that uses nexuiz. Comparing performance of non-MSAA vs MSAA is silly.
Are you building only pbzip2 with the various compilers? pbzip2 is sort of just a front end for libbzip2, which is where the work actually happens.
I read this in every article about Intel and wonder: can you not turn off MSAA to get an accurate performance comparison?
The plan is to only increase the major version for new "desktop" GL versions.
Oh! Fantastic! I'm really glad to hear that.
I'm sure you can help smooth out the problems with R600g?
No, but this part is:
Please stop with the conspiracy theory.
LLVM is a nightmare to use in your project. See all of the difficulties with using it in r600g.
If that's the last time you used Gallium...
Updated to 0.4 on Feb 3 2010
Updated to 0.3 on Mar 24 2009
... then you haven't used Gallium since before Mesa 7.5, which didn't have an R600...
... where by "most" you mean 39%.
One would have thought. :)
I think when LLVMpipe was first written, LLVM wasn't able to generate good SSE code given vectorized IR, so the authors worked around this by using intrinsics.
From the first Google result for "transform feedback" (http://www.opengl.org/wiki/Transform_Feedback)
From the article: