Linux Receiving Fix For AMD Radeon Polaris GPUs Producing Lots Of Log Spam
([Radeon] 3 Hours Ago
Radeon RX 500 Series)
- Reference: 0001561924
- News link: https://www.phoronix.com/news/Linux-Fixing-AMD-Polaris-Spam
- Source link:
For those using an AMD Radeon RX 500 "Polaris" graphics card on Linux and routinely suspend/resume your system, going into the Linux 6.16 kernel and then to be back-ported to the stable series is a fix where the AMDGPU driver could end up producing a lot of spam in the kernel log.
Sent out today was this week's [1]drm-fixes-6.16 for the Radeon and AMDGPU open-source kernel graphics drivers. There is a memory leak fix in the DC display code, a DCN 4.0.1 degamma LUT fix, fixing the reset counter handling on soft recovery, and then catching my eye was a GFX8 (Polaris) fix with not too often seeing new commits from AMD around the aging pre-GFX9 hardware.
Thie fix is for [2]this 5 month old bug report over a lot of "log spam" after the system wakes from sleep. Reporter DeKay explained:
"After waking from sleep, I get thousands of lines like these show up in journalctl and dmesg, so much so that earlier dmesg boot events are lost off the top. My gpu is an AMD RX560 running under Arch Linux kernel 6.12.10-arch1-1 and KDE."
Other AMD Radeon RX 500 "Polaris" graphics card owners on Linux also reported similar problems too after waking/resuming the system that their kernel log would be flooded with countless messages about " scheduler comp_1.0.X is not ready, skipping ."
After months of users running into this issue under Linux with the Polaris GPUs and AMD working to analyze and debug the problem, the fix is now ready for going initially into Linux 6.16 and in the coming days be back-ported to the supported stable kernel series. There is one line of code needed to reset the compute ring wptr on the GPU on resume to avoid a disconnect between the driver and firmware pointer.
Also on the topic of old hardware fixes, today's pull request also drops the console locks when suspending/resuming for the Radeon DRM driver used mostly by pre-GCN hardware. This fixes a circular dependency issue when Radeon is used with another DRM driver.
[1] https://lists.freedesktop.org/archives/amd-gfx/2025-July/127548.html
[2] https://gitlab.freedesktop.org/drm/amd/-/issues/3911
Sent out today was this week's [1]drm-fixes-6.16 for the Radeon and AMDGPU open-source kernel graphics drivers. There is a memory leak fix in the DC display code, a DCN 4.0.1 degamma LUT fix, fixing the reset counter handling on soft recovery, and then catching my eye was a GFX8 (Polaris) fix with not too often seeing new commits from AMD around the aging pre-GFX9 hardware.
Thie fix is for [2]this 5 month old bug report over a lot of "log spam" after the system wakes from sleep. Reporter DeKay explained:
"After waking from sleep, I get thousands of lines like these show up in journalctl and dmesg, so much so that earlier dmesg boot events are lost off the top. My gpu is an AMD RX560 running under Arch Linux kernel 6.12.10-arch1-1 and KDE."
Other AMD Radeon RX 500 "Polaris" graphics card owners on Linux also reported similar problems too after waking/resuming the system that their kernel log would be flooded with countless messages about " scheduler comp_1.0.X is not ready, skipping ."
After months of users running into this issue under Linux with the Polaris GPUs and AMD working to analyze and debug the problem, the fix is now ready for going initially into Linux 6.16 and in the coming days be back-ported to the supported stable kernel series. There is one line of code needed to reset the compute ring wptr on the GPU on resume to avoid a disconnect between the driver and firmware pointer.
Also on the topic of old hardware fixes, today's pull request also drops the console locks when suspending/resuming for the Radeon DRM driver used mostly by pre-GCN hardware. This fixes a circular dependency issue when Radeon is used with another DRM driver.
[1] https://lists.freedesktop.org/archives/amd-gfx/2025-July/127548.html
[2] https://gitlab.freedesktop.org/drm/amd/-/issues/3911
touchdown