You are not logged in.
Pages: 1
Hi all,
After recently switching to i3, I've noticed that every time after I suspend my entire screen will flicker when objects like my cursor move on screen. The more movement, the worse the flickering. The only solution is a reboot. I'm using i3-gaps 4.17.1 with compton (picom) 7.5-2 and the -git amdgpu drivers (which I've updated today, but the issue started before that) on a fresh install of Arch. I'm not really sure what logs I should include, so if you need something let me know.
Even fully closing and reopening X fails to resolve the issue. Once X is started again the issue persists.
Last edited by ClayativeName (2019-12-26 22:19:14)
Offline
Does it also happen w/ the modesetting driver?
Also try to pass
amdgpu.dc=0 amdgpu.dpm=0 amdgpu.aspm=0 amdgpu.runpm=0 amdgpu.bapm=0 to the kernel.
(first one is the DisplayCore codepath, the others are powermanagement related - if the above "works", try to isolate the crucial parameter)
Online
I tested the modsetting driver (which to my knowledge is done by removing xf86-video-amdgpu(-git in my case), but correct me if I'm wrong) to the same result.
Passing all of the above kernel params caused my system to hang at boot, so I tested them one-by-one:
amdgpu.dc=0 caused my system to hang at boot
amdgpu.dpm=0 caused my system to hang at boot
amdgpu.aspm=0 caused my mouse and keyboard to rapidly disconnect and reconnect, rendering them unusable.
amdgpu.runpm=0 boots
amdgpu.bapm=0 boots
When I tested those last two together, my issue persisted.
Offline
Just an update with some more stuff I've tried, most of it related to compositors. I tried using xcompmgr instead of picom to a small amount of improvement. Next, I tried not starting any compositor at all and this did greatly reduce the flickering, but it is still present and becomes very noticeable when there is a large amount of movement, like when scrolling a webpage.
I tried a fix to a related problem, putting
unredir-if-possible = false;in my picom config, but this also had no effect.
I'm frankly out of ideas as to what to try next. Hopefully you all are able to shed some light on this.
Offline
Another update with more troubleshooting that I've done. I decided to look into the journals to find out why amdgpu.dc=0 will cause my system to fail to boot since that seems to be a common solution to flickering. The journals produced no obvious error, just that LightDM failed to start, so I looked into the X logs, the relevant portion of which I've included below.
[ 9.624] (II) AMDGPU: Driver for AMD Radeon:
All GPUs supported by the amdgpu kernel driver
[ 9.624] (II) AMDGPU(0): [KMS] Kernel modesetting enabled.
[ 9.625] (II) AMDGPU(0): Creating default Display subsection in Screen section
"Default Screen Section" for depth/fbbpp 24/32
[ 9.625] (==) AMDGPU(0): Depth 24, (--) framebuffer bpp 32
[ 9.625] (II) AMDGPU(0): Pixel depth = 24 bits stored in 4 bytes (32 bpp pixmaps)
[ 9.625] (==) AMDGPU(0): Default visual is TrueColor
[ 9.625] (**) AMDGPU(0): Option "TearFree" "true"
[ 9.625] (==) AMDGPU(0): RGB weight 888
[ 9.625] (II) AMDGPU(0): Using 8 bits per RGB (8 bit DAC)
[ 9.625] (--) AMDGPU(0): Chipset: "Unknown AMD Radeon GPU" (ChipID = 0x731f)
[ 9.625] (II) Loading sub module "fb"
[ 9.625] (II) LoadModule: "fb"
[ 9.625] (II) Loading /usr/lib/xorg/modules/libfb.so
[ 9.625] (II) Module fb: vendor="X.Org Foundation"
[ 9.625] compiled for 1.20.6, module version = 1.0.0
[ 9.625] ABI class: X.Org ANSI C Emulation, version 0.4
[ 9.625] (II) Loading sub module "dri2"
[ 9.625] (II) LoadModule: "dri2"
[ 9.625] (II) Module "dri2" already built-in
[ 9.658] (II) Loading sub module "glamoregl"
[ 9.658] (II) LoadModule: "glamoregl"
[ 9.658] (II) Loading /usr/lib/xorg/modules/libglamoregl.so
[ 9.661] (II) Module glamoregl: vendor="X.Org Foundation"
[ 9.661] compiled for 1.20.6, module version = 1.0.1
[ 9.661] ABI class: X.Org ANSI C Emulation, version 0.4
[ 9.669] (II) AMDGPU(0): glamor X acceleration enabled on AMD NAVI10 (DRM 3.35.0, 5.4.6-arch1-1, LLVM 9.0.0)
[ 9.669] (II) AMDGPU(0): glamor detected, initialising EGL layer.
[ 9.669] (**) AMDGPU(0): TearFree property default: on
[ 9.669] (==) AMDGPU(0): VariableRefresh: disabled
[ 9.669] (II) AMDGPU(0): KMS Pageflipping: enabled
[ 9.669] (WW) AMDGPU(0): No outputs definitely connected, trying again...
[ 9.669] (WW) AMDGPU(0): Unable to find connected outputs - setting 1024x768 initial framebuffer
[ 9.669] (II) AMDGPU(0): mem size init: gart size :1fe6ad000 vram size: s:1f7937000 visible:fb35000
[ 9.669] (==) AMDGPU(0): DPI set to (96, 96)
[ 9.669] (==) AMDGPU(0): Using gamma correction (1.0, 1.0, 1.0)
[ 9.669] (II) Loading sub module "ramdac"
[ 9.669] (II) LoadModule: "ramdac"
[ 9.669] (II) Module "ramdac" already built-in
[ 9.669] (EE) AMDGPU(0): No modes.
[ 9.670] (II) UnloadModule: "amdgpu"
[ 9.670] (II) UnloadSubModule: "glamoregl"
[ 9.670] (II) Unloading glamoregl
[ 9.670] (II) UnloadSubModule: "fb"
[ 9.670] (II) Unloading fb
[ 9.670] (EE) Screen(s) found, but none have a usable configuration.
[ 9.670] (EE)
Fatal server error:
[ 9.670] (EE) no screens found(EE)
[ 9.670] (EE)
Please consult the The X.Org Foundation support
at http://wiki.x.org
for help.
[ 9.670] (EE) Please also check the log file at "/var/log/Xorg.0.log" for additional information.
[ 9.670] (EE)
[ 9.670] (EE) Server terminated with error (1). Closing log file.The concerning part of this is the fatal error of X being unable to detect any display when amdgpu.dc=0. Any help for troubleshooting this, or any new ideas for the main problem of flickering would be appreciated.
Last edited by ClayativeName (2019-12-25 23:44:07)
Offline
[ 9.669] (II) AMDGPU(0): glamor X acceleration enabled on AMD NAVI10 (DRM 3.35.0, 5.4.6-arch1-1, LLVM 9.0.0)The concerning part of this is the fatal error of X being unable to detect any display when amdgpu.dc=0.
Without the Display Controller functionality NAVI10 cards can not generate any display signal on linux.
Using amdgpu.dc=0 with NAVI effectively turns all outputs off. All that's left is the compute functionality.
You could try a newer kernel like linux mainline and/or mesa-git / llvm-git to see if the issue may already have been solved in development versions.
Disliking systemd intensely, but not satisfied with alternatives so focusing on taming systemd.
clean chroot building not flexible enough ?
Try clean chroot manager by graysky
Offline
Thank you! Changing my kernel to linux-mainline resolved the issue.
Offline
Pages: 1