amdgpu random errors

[j4miel]

Are there any new upstream bugs related/found, and/or is it still related to the change that was reverted for some releases?

The original bug still gets some posts but they don't seem to have re-opened it.

please show us your hardware, cpu, gpu, ram , motherboard, openrc or systemd....

openrc

Code: Select all

         -/oyddmdhs+:.                jamie@gentoo
     -odNMMMMMMMMNNmhy+-`             ------------
   -yNMMMMMMMMMMMNNNmmdhy+-           OS: Gentoo Linux x86_64
 `omMMMMMMMMMMMMNmdmmmmddhhy/`        Host: B550 AORUS ELITE AX V2
 omMMMMMMMMMMMNhhyyyohmdddhhhdo`      Kernel: 6.16.4-gentoo-x86_64
.ydMMMMMMMMMMdhs++so/smdddhhhhdm+`    Uptime: 27 mins
 oyhdmNMMMMMMMNdyooydmddddhhhhyhNd.   Packages: 1276 (emerge), 92 (pkg)
  :oyhhdNNMMMMMMMNNNmmdddhhhhhyymMh   Shell: zsh 5.9
    .:+sydNMMMMMNNNmmmdddhhhhhhmMmy   Resolution: 1920x1080
       /mMMMMMMNNNmmmdddhhhhhmMNhs:   WM: dwm
    `oNMMMMMMMNNNmmmddddhhdmMNhs+`    Theme: Adwaita [GTK2/3]
  `sNMMMMMMMMNNNmmmdddddmNMmhs/.      Icons: Adwaita [GTK2/3]
 /NMMMMMMMMNNNNmmmdddmNMNdso:`        Terminal: tmux
+MMMMMMMNNNNNmmmmdmNMNdso/-           CPU: AMD Ryzen 5 5600 (12) @ 4.470GHz
yMMNNNNNNNmmmmmNNMmhs+/-`             GPU: AMD ATI Radeon RX 7700 XT / 7800 XT
/hMMNNNNNNNNMNdhs++/-`                Memory: 4164MiB / 32007MiB
`/ohdmmddhys+++/:.`
  `-//////:--.

[j4miel]

Just to update on this, I switched my display to 120hz down from 240hz and have not had this issue again when screen wakes up.

So might point to freesync still being an issue in latest kernel even with the reversed patches?

[papu]

Just to update on this, I switched my display to 120hz down from 240hz and have not had this issue again when screen wakes up.

So might point to freesync still being an issue in latest kernel even with the reversed patches?

have you problems with 6.15.7? are you connected to dp or hdmi?

[Krog]

I have the same error (flip done timed out) on 6.16.9
some on internet suggests adding

Code: Select all

amdgpu.runpm=0 amdgpu.dcdebugmask=0x10

on kernel parameters.
I can say the first option only was not enough. I will try the second one too.
Also amdgpu.dcdebugmask=0x12 (12 instead of 10) seems suggested here and there, but I don't know exactly what dey do.

[papu]

Any dosen't works.

amdgpu.runpm=0 and/or amdgpu.dcdebugmask=0x10 amdgpu.dcdebugmask=0x12

are they just for debug puposes? , aren't they?

[j4miel]

I have stopped getting these hangs with 6.17 even on 240hz. Here is my only kernel boot parameter:

Code: Select all

$ grep GRUB_CMDLINE_LINUX_DEFAULT /etc/default/grub
GRUB_CMDLINE_LINUX_DEFAULT="amdgpu.ppfeaturemask=0xffffffff"

[papu]

I have stopped getting these hangs with 6.17 even on 240hz. Here is my only kernel boot parameter:

Code: Select all

$ grep GRUB_CMDLINE_LINUX_DEFAULT /etc/default/grub
GRUB_CMDLINE_LINUX_DEFAULT="amdgpu.ppfeaturemask=0xffffffff"

might you list your motherboard?
my motherboardar is ROG STRIX B650E-F GAMING WIFI

And what are the source of amdgpu.ppfeaturemask=0xffffffff

I see that defaults from amdgpu are 0xfff7bfff

Code: Select all

cat /sys/module/amdgpu/parameters/ppfeaturemask
0xfff7bfff

p.d: It dosen't works for me

thanks