RadeonHD Driver Gets "CS" Branch

[phoronix]

Phoronix: RadeonHD Driver Gets "CS" Branch

There are already separate branches of the xf86-video-radeonhd driver for AtomBIOS support and another one for improved 2D acceleration, but today we have yet another branch. This new branch is called "CS" and what it does is change all calls within the driver for the command sequences (the "CS") into macros. As Luc Verhaegen explained in his mailing list message, inline functions should yield the same performance as macros, but it reality it's not -- at least not with GCC on OpenSuSE 10.3...

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

[d2kx]

You've forget to mention the HDMI-Audio branch :)

So we got 4 different branches + master. I really hope all that stuff gets merged within the next 1-2 months or so and then we'll see RadeonHD 2.0, ehm 1.3 ;)

[etymxris]

If macros are performing any better than inline functions, then something is wrong with the compiler. On the other hand, last I heard compilers ignored keywords like "inline" and "register", trusting their own optimization routines to do better. Using a macro is one way to force a function to be "inline", so it seems likely GCC is doing suboptimal optimization.

[grantek]

Using a macro is one way to force a function to be "inline", so it seems likely GCC is doing suboptimal optimization.
Makes sense. I'm not a developer, but it sounds like relying on macros would just make the code get more and more unreadable - inlines are there for a reason. Good to know people are keeping an eye on things like this though :)