Bug 111691

Summary: inconsistent cursor movement speed when using AMD 5700 XT
Product: DRI Reporter: Michael Haworth <a>
Component: DRM/AMDgpuAssignee: Default DRI bug account <dri-devel>
Status: RESOLVED MOVED QA Contact:
Severity: not set    
Priority: not set CC: jaapbuurman, ninuje
Version: DRI git   
Hardware: x86-64 (AMD64)   
OS: Linux (All)   
Whiteboard:
i915 platform: i915 features:
Attachments:
Description Flags
video of the cursors motion when moving the mouse at a reasonable speed
none
dmesg output after booting with HW cursor enabled
none
Xorg log after booting with HW cursor enabled
none
xorg log on xubuntu 19.10 (where bug is not present) none

Description Michael Haworth 2019-09-16 10:17:19 UTC
Created attachment 145372 [details]
video of the cursors motion when moving the mouse at a reasonable speed

Overview: After updating kernel to 5.3rc + firmware from https://people.freedesktop.org/~agd5f/radeon_ucode/navi10/ I am experiencing a very laggy and erratic cursor when using a Radeon 5700 XT

Steps to Reproduce: Use a navi10 card with the latest kernel and firmware. Tried many desktop environments with both Xorg and Wayland

Build Date & Hardware: Built September 2019. AMD Radeon 5700 XT


Additional Builds and Platforms: Issue doesn't occur if I use a Radeon RX 580 instead of the navi card and isn't triggered on Windows 7/10
Comment 1 Michael Haworth 2019-09-16 10:19:44 UTC
I forgot to mention that I have been using oibaf's PPA in addition to the other software. I was under the impression that I had to do this for the display to function.
Comment 2 Michael Haworth 2019-09-16 10:22:04 UTC
A workaround is to enable SWCursor in Xorg
Comment 3 Alex Deucher 2019-09-16 18:27:25 UTC
Please attach your xorg log (if using X) and your dmesg output.  If this is a regression, can you bisect?
Comment 4 Michael Haworth 2019-09-16 18:33:30 UTC
Created attachment 145382 [details]
dmesg output after booting with HW cursor enabled
Comment 5 Michael Haworth 2019-09-16 18:33:55 UTC
Created attachment 145383 [details]
Xorg log after booting with HW cursor enabled
Comment 6 Michael Haworth 2019-09-16 18:36:13 UTC
I have attached the logs but do not know how to answer your question, this is my first bug report.
Comment 7 Felix Schwarz 2019-09-16 20:20:32 UTC
>> If this is a regression, can you bisect?
> do not know how to answer your question, this is my first bug report.

Regression: Was the bug always present (as far as you know) or did it work in the past?

If it worked previously you can do a "git bisection" to find the problematic change. This increases the likelihood that AMD's driver developers can build a fix quickly.
Comment 8 Michael Haworth 2019-09-16 20:24:25 UTC
Thanks for the explanation. I only recently bought the graphics card and the bug occurred the first time I used it on linux
Comment 9 Michael Haworth 2019-09-16 22:18:27 UTC
I updated to the 5.3 release but has not fixed the issue
Comment 10 Michael Haworth 2019-09-19 21:01:17 UTC
issue occurs with the closed source drivers too, on kernel 4.15 (linux mint 19.2) but to a much lesser extent
Comment 11 Michael Haworth 2019-09-27 19:28:00 UTC
Issue is not present in Ubuntu 19.10 beta
Comment 12 Andre Klapper 2019-09-28 08:28:30 UTC
Nearly nobody knows which versions are shipped by some distribution in some release. Please always provide version information - thanks.
Comment 13 Michael Haworth 2019-09-28 09:38:27 UTC
running Xubuntu 19.10 - version information below (sorry for not providing it earlier):

linux-firmware (1.182) eoan; urgency=medium

  * Rebase against git://git.kernel.org/pub/scm/linux/kernel/git/firmware/linux-firmware.git
    6c6918ad8ae0dfb2cb591484eba525409980c16f

linux-generic (5.3.0-12.13) eoan; urgency=medium

  * Master version: 5.3.0-12.13

xserver-xorg-video-amdgpu (19.0.1-1)

xorg-server 2:1.20.5+git20190820-0ubuntu3

I will provide xorg log too
Comment 14 Michael Haworth 2019-09-28 09:39:19 UTC
Created attachment 145575 [details]
xorg log on xubuntu 19.10 (where bug is not present)
Comment 15 Jaap Buurman 2019-10-11 11:38:47 UTC
I am running into the same issue.

Arch
Kernel: 5.3.5
llvm: 9.0.0
Mesa: 19.2.0
xf86-video-amdgpu: 19.0.1
Linux-firmware: 20190923

Let me know if I can supply any additional information.
Comment 16 takios+fdbugs 2019-10-11 16:17:57 UTC
I ran into the same issue but after installing linux kernel 5.4rc2 it was fixed.
Comment 17 Jaap Buurman 2019-10-11 16:19:52 UTC
(In reply to takios+fdbugs from comment #16)
> I ran into the same issue but after installing linux kernel 5.4rc2 it was
> fixed.

That's good to hear! Does anyone know whether the fix will be backported to the 5.3 kernel? It's gonna take a while before 5.4 becomes mainline.
Comment 18 Daniel Suarez 2019-11-11 20:40:27 UTC
(In reply to Jaap Buurman from comment #17)
> (In reply to takios+fdbugs from comment #16)
> > I ran into the same issue but after installing linux kernel 5.4rc2 it was
> > fixed.
> 
> That's good to hear! Does anyone know whether the fix will be backported to
> the 5.3 kernel? It's gonna take a while before 5.4 becomes mainline.

I do not believe it will get backported since most effort is going towards 5.4, but you can try a ditro like Manjaro that officially supports the RC kernels
Comment 19 Martin Peres 2019-11-19 09:51:39 UTC
-- 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/drm/amd/issues/908.

Use of freedesktop.org services, including Bugzilla, is subject to our Code of Conduct. How we collect and use information is described in our Privacy Policy.