Bettering Radeon Gallium3D Performance With PCI-E 2.0

[bridgman]

mesa ppl already knows very clearly (prolly) what all the missing bits are and globally whats needed to face catalyst very close, what ppl seems to don't understand no matter how many times is explained is that mesa have like 5 DEVELOPERS and the code to handle GPU is EXTREMELY COMPLEX so is not like you can't magically shave out of your ass a couple of houndred thousand of lines code to magically make all work

Yep, and we're actually talking about kernel driver code where the number of active open source developers is even smaller.

[DanL]

Missing features from r600g are 2D Tiling,HiZ, apparently PCI-E2, power saving.

The first 3 just need kinks worked out before they become enabled by default, and power saving is present, but not as good as Catalyst at the moment. No reason for unnecessary panic..

[Drago]

The first 3 just need kinks worked out before they become enabled by default, and power saving is present, but not as good as Catalyst at the moment. No reason for unnecessary panic..

Not panicking. Just I am concerned that Intel have way more weaker chips, but with stronger performance. They also have very good inertia so I expect event better performance/power numbers from them.

[ChrisXY]

The main problem is the sheer size of the Catalyst code relative to the size of the open source dev team. We tried sanitizing a much smaller code base (the "tcore" diagnostic platform) to kickstart r6xx support... spent ~9 months of part-time work on it and eventually gave up. The Catalyst power management code alone is bigger than the entire open driver stack.

Oh dear, so I guess that suspend to ram etc. is so buggy (http://ati.cchtml.com/show_bug.cgi?id=153) because it is so complex that there inevitably are bugs all over the place?
Well, thanks for the answer.

The challenge enabling things like tiling isn't "not knowing how to enable", it's that enabling tiling requires changes in a lot of different places and since the driver architectures are different you can't just look at the proprietary driver and find the corresponding areas in the open driver.

I understand that and thus I was speaking about some low level functionality and that it could help, at least in some areas. Of course I assmue that there are some helpful comments on that code.

[werfu]

Michael, could you test this with the 2D tiling patch? I wonder how much these two combined would increase the performance.

[Azpegath]

The main problem is the sheer size of the Catalyst code relative to the size of the open source dev team. We tried sanitizing a much smaller code base (the "tcore" diagnostic platform) to kickstart r6xx support... spent ~9 months of part-time work on it and eventually gave up. The Catalyst power management code alone is bigger than the entire open driver stack.

The challenge enabling things like tiling isn't "not knowing how to enable", it's that enabling tiling requires changes in a lot of different places and since the driver architectures are different you can't just look at the proprietary driver and find the corresponding areas in the open driver.

Cool, thanks for the answer. I should have understood that since I'm a software developer myself. I prefer C++ over C though. So I guess the register stuff is the easy part and the framework/pipeline of it is the hard part?

[darkbasic]

Just I am concerned that Intel have way more weaker chips, but with stronger performance.

Intel has 30+ developers while amd has 5, that's the difference.

[HokTar]

Intel has 30+ developers while amd has 5, that's the difference.

As much as I love AMD I still think that they are going to lose if things remain this way.

It appears though, that they understood recently that the good quality on the software side is essantial. What they still lack is the realisation that:
"If we pay 5 more guys we could conquer the _whole_ linux world."
This would mean roughly 2.5% increase in sales for them. I guess that means a higher revenue than the salary of those guys, i.e. profit.

Just my 2 cents though...

[Ansla]

Now hold on, making a great mesa driver might make a difference for the technical users and those that care about their freedom, but there are hordes of Ubuntu users out there that just use Linux because it's free as in beer and couldn't care less what driver they use.

And don't get me wrong, I appreciate AMD's effort for making a free as in speech driver, and I'm using the radeon driver on my HD3650 from the days there was no 3D acceleration, not even the classical r600 driver. I just don't live the illusion that the majority of Linux users share my view regarding freedom.

[darkbasic]

AMD has a shared codebase and a proprietary driver with lots of developers, while intel has only the open source one.
I would be glad if AMD will drop catalyst for linux and shift 50 developers to mesa, unfortunately the vast majority of the users do not care about the "free speech" if there is the "free beer".