https://bugs.freedesktop.org/show_bug.cgi?id=109171
Bug ID: 109171 Summary: AMD's GPU is Slower than Intel's Product: Mesa Version: 18.2 Hardware: x86-64 (AMD64) OS: Linux (All) Status: NEW Severity: critical Priority: medium Component: Drivers/Gallium/r600 Assignee: dri-devel@lists.freedesktop.org Reporter: ali.sherif10@yahoo.com QA Contact: dri-devel@lists.freedesktop.org
glxgears and CS: Go are much slower with AMD's GPU. Performance's much higher with Intel's.
Outputs: $ glxgears 306 frames in 5.0 seconds = 61.029 FPS 299 frames in 5.0 seconds = 59.761 FPS 299 frames in 5.0 seconds = 59.762 FPS 299 frames in 5.0 seconds = 59.762 FPS 299 frames in 5.0 seconds = 59.762 FPS
$ DRI_PRIME=1 glxgears 304 frames in 5.0 seconds = 60.666 FPS 299 frames in 5.0 seconds = 59.762 FPS 299 frames in 5.0 seconds = 59.762 FPS 299 frames in 5.0 seconds = 59.763 FPS 299 frames in 5.0 seconds = 59.760 FPS
$ vblank_mode=0 glxgears ATTENTION: default value of option vblank_mode overridden by environment. 32606 frames in 5.0 seconds = 6521.047 FPS 31516 frames in 5.0 seconds = 6302.010 FPS 32145 frames in 5.0 seconds = 6428.891 FPS 31443 frames in 5.0 seconds = 6288.595 FPS 31029 frames in 5.0 seconds = 6205.721 FPS
$ DRI_PRIME=1 vblank_mode=0 glxgears ATTENTION: default value of option vblank_mode overridden by environment. 12031 frames in 5.0 seconds = 2406.195 FPS 10110 frames in 5.0 seconds = 2021.870 FPS 10122 frames in 5.0 seconds = 2022.474 FPS 9782 frames in 5.0 seconds = 1956.206 FPS 10175 frames in 5.0 seconds = 2034.901 FPS
Laptop: HP Pavilion g6. AMD's GPU: Radeon HD 7670M. OS: 64-bit Linux (Fedora 29 KDE Spin).
https://bugs.freedesktop.org/show_bug.cgi?id=109171
--- Comment #1 from Sherif ali.sherif10@yahoo.com --- CS: Go is faster with Intel's GPU even without vblank_mode=0.
https://bugs.freedesktop.org/show_bug.cgi?id=109171
--- Comment #2 from Alex Deucher alexdeucher@gmail.com --- Please attach your xorg log and dmesg output. Note that using the dGPU involves extra copies to copy the rendered frame to the integrated GPU for display so there is additional overhead. You need a relatively advanced 3D game to see an advantage on the dGPU as you have to overcome the overhead of the additional copies.
https://bugs.freedesktop.org/show_bug.cgi?id=109171
--- Comment #3 from Sherif ali.sherif10@yahoo.com --- Created attachment 142914 --> https://bugs.freedesktop.org/attachment.cgi?id=142914&action=edit Xorg log files.
I attached all files.
https://bugs.freedesktop.org/show_bug.cgi?id=109171
Sherif ali.sherif10@yahoo.com changed:
What |Removed |Added ---------------------------------------------------------------------------- URL| |https://bugzilla.redhat.com | |/show_bug.cgi?id=1662535
--- Comment #4 from Sherif ali.sherif10@yahoo.com --- I reported it in Fedora Bugzilla, also. Please, check.
https://bugs.freedesktop.org/show_bug.cgi?id=109171
--- Comment #5 from Sherif ali.sherif10@yahoo.com --- Created attachment 142951 --> https://bugs.freedesktop.org/attachment.cgi?id=142951&action=edit I used something like dmesg | vim.
https://bugs.freedesktop.org/show_bug.cgi?id=109171
--- Comment #6 from Sherif ali.sherif10@yahoo.com --- Sorry, I didn't pay attention to the dmesg part.
https://bugs.freedesktop.org/show_bug.cgi?id=109171
--- Comment #7 from Christian König christian.koenig@amd.com --- The dmesg is only a fraction. Please provide the full dmesg directly after boot.
That glxgears is getting slower numbers on the secondary GPU is perfectly normal. It is a test how fast we can page flip and not how fast the GPU is.
https://bugs.freedesktop.org/show_bug.cgi?id=109171
Sherif ali.sherif10@yahoo.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Attachment #142951|0 |1 is obsolete| |
--- Comment #8 from Sherif ali.sherif10@yahoo.com --- Created attachment 142953 --> https://bugs.freedesktop.org/attachment.cgi?id=142953&action=edit Output of dmesg.
https://bugs.freedesktop.org/show_bug.cgi?id=109171
--- Comment #9 from Sherif ali.sherif10@yahoo.com --- Created attachment 142954 --> https://bugs.freedesktop.org/attachment.cgi?id=142954&action=edit Output of DRI_PRIME=1 dmesg
https://bugs.freedesktop.org/show_bug.cgi?id=109171
--- Comment #10 from Sherif ali.sherif10@yahoo.com --- I restarted, logged into KDE then $ dmesg >> dmesg. After a minute, $ DRI_PRIME=1 dmesg >> dmesg2. Did I do it correctly?
(In reply to Alex Deucher from comment #2)
You need a relatively advanced 3D game to see an advantage on the dGPU as you have to overcome the overhead of the additional copies.
Counter-Strike: Global Offensive is much slower, when I run Steam with DRI_PRIME=1.
Additional Info: I installed Steam from RPMFusion.
https://bugs.freedesktop.org/show_bug.cgi?id=109171
--- Comment #11 from Sherif ali.sherif10@yahoo.com --- Created attachment 142955 --> https://bugs.freedesktop.org/attachment.cgi?id=142955&action=edit Steam output wit Intel's GPU.
https://bugs.freedesktop.org/show_bug.cgi?id=109171
--- Comment #12 from Sherif ali.sherif10@yahoo.com --- Created attachment 142956 --> https://bugs.freedesktop.org/attachment.cgi?id=142956&action=edit Steam output wit AMD's GPU.
https://bugs.freedesktop.org/show_bug.cgi?id=109171
--- Comment #13 from Sherif ali.sherif10@yahoo.com --- I found that only CS: GO's main GUI is slow with AMD's GPU, but it's much faster during matches. On Intel's GPU, the main GUI's smooth, but there's some lag during matches. I tried the tutorial only.
On both cases, fullscreen was enabled.
https://bugs.freedesktop.org/show_bug.cgi?id=109171
--- Comment #14 from Sherif ali.sherif10@yahoo.com --- Note: I played CS: GO during recording Steam's output.
https://bugs.freedesktop.org/show_bug.cgi?id=109171
--- Comment #15 from Sherif ali.sherif10@yahoo.com --- Created attachment 142957 --> https://bugs.freedesktop.org/attachment.cgi?id=142957&action=edit glxinfo with(out) DRI_PRIME=1.
This can help.
https://bugs.freedesktop.org/show_bug.cgi?id=109171
GitLab Migration User gitlab-migration@fdo.invalid changed:
What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |RESOLVED Resolution|--- |MOVED
--- Comment #16 from GitLab Migration User gitlab-migration@fdo.invalid --- -- GitLab Migration Automatic Message --
This bug has been migrated to freedesktop.org's GitLab instance and has been closed from further activity.
You can subscribe and participate further through the new bug through this link to our GitLab instance: https://gitlab.freedesktop.org/mesa/mesa/issues/649.
dri-devel@lists.freedesktop.org
-
bugzilla-daemon＠freedesktop.org