[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: Slow performance due to llvmpipe being used despite Radeon kernel and X modules being loaded



Le 06/05/2020 à 22:15, EoflaOE ViceCity a écrit :
> Hello. Sorry for the length of this problem, but I am trying to get the
> X server to use my graphics card, AMD Radeon 9200 SE (RV280), instead of
> my CPU to render things on the desktop. I actually have a newer
> computer, but I use the older computer for experiments. Here's how the
> problem goes.
> 
> Last month, I have installed Debian 10 on my old computer for testing
> purposes, and upgraded it to testing to test newer versions of packages
> to see if there are bugs.
|...]
> "glxinfo | grep OpenGL":
> 
> OpenGL vendor string: *VMware, Inc.*
> OpenGL renderer string: *llvmpipe (LLVM 9.0.1, 128 bits)*
[...]
> X server logs "cat /var/log/Xorg.0.log | grep -i radeon":
[...]
> [    57.917] (II) RADEON(0): [DRI2] Setup complete
> [    57.917] (II) RADEON(0): [DRI2]   DRI driver: r200
> [    57.918] (II) RADEON(0): Front buffer size: 5175K
> [    57.918] (II) RADEON(0): VRAM usage limit set to 108490K
> [    57.935] (==) RADEON(0): DRI3 disabled
> [    57.935] (==) RADEON(0): Backing store enabled
> [    57.935] (II) RADEON(0): Direct rendering enabled
> [    57.936] (II) RADEON(0): Render acceleration enabled for R200 type
> cards.
[...]
> By the way I am
> using Linux 5.6 from testing, if that matters.

Hello,

what puzzles me is VMWare as OpenGL vendor: am I wrong or does it mean
that you have installed Debian as a VMWare guest?
If it is the case, I do not know VMWare at all, but I think there is a
form of passthrough for your graphic card, VMWare recognises it as a
Radeon (not a generic card) but by default 3D acceleration (DRI3?) is
disabled in VMWare and you have to enable it on your VMWare host.

Debian Testing is for testing a possibly broken Debian in order for the
contributors to build the next stable release without bugs. It is not
really for testing new functionnalities


Reply to: