AMD's Initial Radeon Driver Changes For Linux 3.12

[halfmanhalfamazing]

Easier said than done, at least in terms of making it effective.

This is true, but I would at least like to see Crossfire implemented, testable, and verifiable.(even if the results shown by benchmarks are fairly low. At least it works) Optimizations can and obviously would come later.

[agd5f]

Easier said than done, at least in terms of making it effective. Why do you think it took so long to eliminate micro stuttering?.

RIght. My point is that you don't really need any hw details to implement it. It mostly sits above the hw specific layers. Unfortunately, there is not much demand for crossfire outside of windows.

[bobjimbob]

Is there any interface to over/underclock the memory and shader clock on AMD gpus using the new DPM changes? Sort of like how you can use amdoverdrivectrl with the catalyst driver.

[CrvenaZvezda]

RIght. My point is that you don't really need any hw details to implement it. It mostly sits above the hw specific layers. Unfortunately, there is not much demand for crossfire outside of windows.

Chicken and egg situation. Provide it and it will be used!

[CrvenaZvezda]

Me being a bit slow when it comes to understanding all this graphical stuff have a question for the bright minds out there. Is anything of this something that will help speed up 3D rendering? You see, the day the open source driver provides about 80% of the performance of the closed source one, I will drop the proprietary driver. (And yes, selfish me only cares about getting 3D up to speed).

[Ericg]

Me being a bit slow when it comes to understanding all this graphical stuff have a question for the bright minds out there. Is anything of this something that will help speed up 3D rendering? You see, the day the open source driver provides about 80% of the performance of the closed source one, I will drop the proprietary driver. (And yes, selfish me only cares about getting 3D up to speed).

Run Kernel 3.11 with raden.dpm=1, run Mesa 9.2 + LLVM 3.3, set R600_DEBUG=sb as an environment variable:

If you're running R600G.... BAM, 60-80% performance of closed source (in theory. Your mileage may vary).

If you're running RadeonSI... you gotta wait a little longer, though its coming up pretty quick.

[schmidtbag]

This is true, but I would at least like to see Crossfire implemented, testable, and verifiable.(even if the results shown by benchmarks are fairly low. At least it works) Optimizations can and obviously would come later.

I agree - I would much rather it work and be improved upon than not have it at all. I'm sure it'd still be a better experience than only using 1 card. On the other hand, I'd rather have video decoding and openCL worked on first before CF, thankfully, both of which are supposedly WIP.

[Ericg]

Alex, any chance we could ever see a Radeon Settings Panel / Radeon Control Panel? For display management (multi-monitor) the various DE's handle that themselves, fine, but what about for tuning power settings? Enabling CrossFire if it would ever get written and merged? Fine-tuning MSAA? Changing handling setting of module options? etc etc etc.

Not asking AMD themselves to write it-- there's no reason another contributor couldn't do it. I'm just wondering if the idea has come up already in team talks for the paid staff, or if it's come up on mailing lists, or anywhere else you've noticed.

[CrvenaZvezda]

Run Kernel 3.11 with raden.dpm=1, run Mesa 9.2 + LLVM 3.3, set R600_DEBUG=sb as an environment variable:

If you're running R600G.... BAM, 60-80% performance of closed source (in theory. Your mileage may vary).

If you're running RadeonSI... you gotta wait a little longer, though its coming up pretty quick.

Uh... I have a HD6950 card. Guess that is related to R600? Northern Island?

Darn those milages. I wish they would come in just one size!

[Ericg]

On the other hand, I'd rather have video decoding and openCL worked on first before CF, thankfully, both of which are supposedly WIP.

Video Decode IS done already.. (?) o.O