Windows 10 Radeon Software vs. AMDGPU On Ubuntu Linux

[liam]

We are talking about game performances here, so I assumed we could talk about gamers hardware.

Intel graphics are plenty for many Steam games unless you want to run them at high res or effects settings.

[SystemCrasher]

It's so much more complicated then that. For desktop usage their drivers have been the very best available for years.

I can admit at least these MAJOR improvements:
- Fucking Catalyst kernel module is gone. No more GPF kernel crashes after week of uptime. No more black screens after kernel upgrades. Regardless if it was "pro" or "non-pro".
- Bonus: bye-bye tainted kernels, so you do not collect middle fingers of whole kernel team anymore when trying to report bug.
- Now Linux driver performance is on par with Windows. Historically, AMD's Windows driver had better performance in most cases. So it seems reworking Linux driver seriously fixed it in the process.
- As bonus, all-open stack exposes really neat performance in these runs and now it hit OpenGL 4.3 level. So there're quite few reasons to mess with blobs at all.

[duby229]

Real title of this article should be:

FINALLY, AFTER SO MANY YEARS, AMD HAS GOT THEIR LINUX DRIVERS RIGHT!!!

It's so much more complicated then that. For desktop usage their drivers have been the very best available for years.

[Passso]

https://www.techpowerup.com/198029/g...n-market-share

We are talking about game performances here, so I assumed we could talk about gamers hardware.

[SystemCrasher]

Real title of this article should be:

FINALLY, AFTER SO MANY YEARS, AMD HAS GOT THEIR LINUX DRIVERS RIGHT!!!

[Amarildo]

Don't wonder when your bug report gets closed as it will be a duplicate of this "closed" bug: https://github.com/ValveSoftware/ste...time/issues/13

Sorry, I fail to see how the two bugs are identical, or even related. Could you be more specific?

[BenPope]

Installing the amdgpu hybrid stack actually replaces the existing amdgpu kernel module with a new one built from a DKMS package, so the existing kernel amdgpu driver itself shouldn't matter much.

Yes, but the amdgpu pro driver failed to build... so it's using the one from the agd5f kernel tree based on 4.6-rc1.

[liam]

Please read the Steam statistics about hardware first. Then write.

https://www.techpowerup.com/198029/g...n-market-share

[liam]

What is IGP?

Integrated Graphics Processor

[stiiixy]

I thought only the kernel ran in long mode and then context switched to compatibility mode for 32bit code in userspace? If so then that 32bit code doesn't have access to additional registers or extensions. However if your right that long mode supports 2, 4 or 8 byte pointers, then in that case there is absolutely no reason to build a linux system that is totally 64bit since it will have access to all the additional registers and extensions from long mode..

EDIT: It should be pretty easy to write a tool to determine how much address space app needs and then compile it appropriately.

Thanks mate (and bridgman), I didn't realise there was actually another way to go about it in ARCH called PRIME,and I've been plugging away at that when I can. Very hackish methods all these things to get a game running, but at least with PRIME, I can simply specify

xrandr --setprovideroffloadsink provider sink

with provider being my DGP (0x55) and sink being IGP (0x7e) in my case.

Then run DRI_PRIME=1 ./nameof3Dapp (or just add DRI_PRIME=1 to the config/shortcut etc).

But then I run in to the black screen issue mentioned in th article, and currently installing MATE to see if it works better 'natively' compared to Plasma.