AceFour
September 22, 2025, 6:49pm
61
So while the benchmark is slower, it isn’t crashing so that is good
❯ MESA_LOADER_DRIVER_OVERRIDE=zink glmark2
=======================================================
glmark2 2023.01
=======================================================
OpenGL Information
GL_VENDOR: Mesa
GL_RENDERER: zink Vulkan 1.4(AMD Radeon RX 6700 (RADV NAVI22) (MESA_RADV))
GL_VERSION: 4.6 (Compatibility Profile) Mesa 25.2.3-arch1.2
Surface Config: buf=32 r=8 g=8 b=8 a=8 depth=24 stencil=0 samples=0
Surface Size: 800x600 windowed
=======================================================
[build] use-vbo=false: FPS: 10288 FrameTime: 0.097 ms
[build] use-vbo=true: FPS: 15848 FrameTime: 0.063 ms
[texture] texture-filter=nearest: FPS: 15701 FrameTime: 0.064 ms
[texture] texture-filter=linear: FPS: 15632 FrameTime: 0.064 ms
[texture] texture-filter=mipmap: FPS: 15494 FrameTime: 0.065 ms
[shading] shading=gouraud: FPS: 15546 FrameTime: 0.064 ms
[shading] shading=blinn-phong-inf: FPS: 15349 FrameTime: 0.065 ms
[shading] shading=phong: FPS: 15424 FrameTime: 0.065 ms
[shading] shading=cel: FPS: 14841 FrameTime: 0.067 ms
[bump] bump-render=high-poly: FPS: 14304 FrameTime: 0.070 ms
[bump] bump-render=normals: FPS: 14866 FrameTime: 0.067 ms
[bump] bump-render=height: FPS: 15240 FrameTime: 0.066 ms
[effect2d] kernel=0,1,0;1,-4,1;0,1,0;: FPS: 15088 FrameTime: 0.066 ms
[effect2d] kernel=1,1,1,1,1;1,1,1,1,1;1,1,1,1,1;: FPS: 14651 FrameTime: 0.068 ms
[pulsar] light=false:quads=5:texture=false: FPS: 14425 FrameTime: 0.069 ms
[desktop] blur-radius=5:effect=blur:passes=1:separable=true:windows=4: FPS: 8285 FrameTime: 0.121 ms
[desktop] effect=shadow:windows=4: FPS: 8534 FrameTime: 0.117 ms
[buffer] columns=200:interleave=false:update-dispersion=0.9:update-fraction=0.5:update-method=map: FPS: 1785 FrameTime: 0.561 ms
[buffer] columns=200:interleave=false:update-dispersion=0.9:update-fraction=0.5:update-method=subdata: FPS: 2223 FrameTime: 0.450 ms
[buffer] columns=200:interleave=true:update-dispersion=0.9:update-fraction=0.5:update-method=map: FPS: 2255 FrameTime: 0.444 ms
[ideas] speed=duration: FPS: 6927 FrameTime: 0.144 ms
[jellyfish] <default>: FPS: 12674 FrameTime: 0.079 ms
[terrain] <default>: FPS: 2819 FrameTime: 0.355 ms
[shadow] <default>: FPS: 12624 FrameTime: 0.079 ms
[refract] <default>: FPS: 5535 FrameTime: 0.181 ms
[conditionals] fragment-steps=0:vertex-steps=0: FPS: 14075 FrameTime: 0.071 ms
[conditionals] fragment-steps=5:vertex-steps=0: FPS: 14138 FrameTime: 0.071 ms
[conditionals] fragment-steps=0:vertex-steps=5: FPS: 13890 FrameTime: 0.072 ms
[function] fragment-complexity=low:fragment-steps=5: FPS: 14170 FrameTime: 0.071 ms
[function] fragment-complexity=medium:fragment-steps=5: FPS: 13769 FrameTime: 0.073 ms
[loop] fragment-loop=false:fragment-steps=5:vertex-steps=5: FPS: 13853 FrameTime: 0.072 ms
[loop] fragment-steps=5:fragment-uniform=false:vertex-steps=5: FPS: 13675 FrameTime: 0.073 ms
[loop] fragment-steps=5:fragment-uniform=true:vertex-steps=5: FPS: 13570 FrameTime: 0.074 ms
=======================================================
glmark2 Score: 12044
=======================================================
AceFour:
OVERRIDE=zink
Are you using an environmental variable for this?
$MESA_LOADER_DRIVER_OVERRIDE=zink
Edit: I tried setting it in environment and desktop fails to load.
AceFour
September 22, 2025, 10:43pm
63
you have to open it from command line
MESA_LOADER_DRIVER_OVERRIDE=zink glmark2
you can add it to your shell profile as
.zshrc
export MESA_LOADER_DRIVER_OVERRIDE=zink
I don’t think that menu launchers will pick it up.
AceFour
September 23, 2025, 1:19am
65
you can put in .bashrc if you are using Bash
glmark2 Score: 6960
Edit: The score really drops using zink
AceFour
September 23, 2025, 2:19am
67
I’ll take the lower score rather than kernel faulting any day. It is good enough to run shaders in MC and plenty smooth. Mesa meanwhile is still trying to figure out the regression for 6 weeks now and the amdgpu ring timeout tickets are piling up
AceFour
September 24, 2025, 3:05am
68
So, I did a fresh install and you do not need Oracle JDK or need to twiddle the kernel to run MultiMC launcher of Minecraft. The Zink environment trick does not work with minecraft-launcher
Is the quick and dirty solution until Mesa figures out the regression bug with ring timeout
Mesa updated. vkmark successfully ran
[acefour@acefour-2 ~]$ vkmark
=======================================================
vkmark 2025.01
=======================================================
Vendor ID: 0x1002
Device ID: 0x73DF
Device Name: AMD Radeon RX 6700 (RADV NAVI22)
Driver Version: 104865796
Device UUID: 80b0907ec2d37b93becd8b7531372d8c
=======================================================
[vertex] device-local=true: FPS: 33327 FrameTime: 0.030 ms
[vertex] device-local=false: FPS: 26739 FrameTime: 0.037 ms
[texture] anisotropy=0: FPS: 32967 FrameTime: 0.030 ms
[texture] anisotropy=16: FPS: 33517 FrameTime: 0.030 ms
[shading] shading=gouraud: FPS: 33864 FrameTime: 0.030 ms
[shading] shading=blinn-phong-inf: FPS: 33893 FrameTime: 0.030 ms
[shading] shading=phong: FPS: 33900 FrameTime: 0.029 ms
[shading] shading=cel: FPS: 33723 FrameTime: 0.030 ms
[effect2d] kernel=edge: FPS: 34083 FrameTime: 0.029 ms
[effect2d] kernel=blur: FPS: 24878 FrameTime: 0.040 ms
[desktop] <default>: FPS: 30047 FrameTime: 0.033 ms
[cube] <default>: FPS: 33975 FrameTime: 0.029 ms
[clear] <default>: FPS: 30716 FrameTime: 0.033 ms
=======================================================
vkmark Score: 31971
=======================================================
[acefour@acefour-2 ~]$ inxi -Fxxc0z
System:
Kernel: 6.16.8-arch3-1 arch: x86_64 bits: 64 compiler: gcc v: 15.2.1
Desktop: KDE Plasma v: 6.4.5 tk: Qt v: N/A wm: kwin_wayland dm: SDDM
Distro: EndeavourOS base: Arch Linux
Machine:
Type: Desktop System: ASUS product: N/A v: N/A serial: <superuser required>
Mobo: ASUSTeK model: PRIME Z690M-PLUS D4 v: Rev 1.xx
serial: <superuser required> part-nu: SKU UEFI: American Megatrends v: 3801
date: 05/14/2025
Battery:
Device-1: hidpp_battery_0 model: Logitech M720 Triathlon Multi-Device Mouse
serial: <filter> charge: 55% (should be ignored) status: discharging
CPU:
Info: 6-core model: 12th Gen Intel Core i5-12400F bits: 64 type: MT MCP
arch: Alder Lake rev: 5 cache: L1: 480 KiB L2: 7.5 MiB L3: 18 MiB
Speed (MHz): avg: 800 min/max: 800/5600 cores: 1: 800 2: 800 3: 800 4: 800
5: 800 6: 800 7: 800 8: 800 9: 800 10: 800 11: 800 12: 800 bogomips: 59904
Flags-basic: avx avx2 ht lm nx pae sse sse2 sse3 sse4_1 sse4_2 ssse3 vmx
Graphics:
Device-1: Advanced Micro Devices [AMD/ATI] Navi 22 [Radeon RX 6700/6700
XT/6750 XT / 6800M/6850M XT] vendor: XFX driver: amdgpu v: kernel
arch: RDNA-2 pcie: speed: 16 GT/s lanes: 16 ports: active: DP-3,HDMI-A-1
empty: DP-1,DP-2,Writeback-1 bus-ID: 03:00.0 chip-ID: 1002:73df
Device-2: Microdia USB Live camera driver: snd-usb-audio,uvcvideo
type: USB rev: 2.0 speed: 480 Mb/s lanes: 1 bus-ID: 1-6:3 chip-ID: 0c45:636b
Display: wayland server: X.org v: 1.21.1.18 with: Xwayland v: 24.1.8
compositor: kwin_wayland driver: X: loaded: amdgpu
unloaded: modesetting,radeon alternate: fbdev,vesa dri: radeonsi
gpu: amdgpu d-rect: 7680x2160 display-ID: 0
Monitor-1: DP-3 pos: primary,left model: Dell S3221QS res: 3840x2160
hz: 60 dpi: 139 diag: 806mm (31.7")
Monitor-2: HDMI-A-1 pos: right model: LG (GoldStar) Ultra HD
res: 3840x2160 hz: 60 dpi: 163 diag: 690mm (27.2")
API: EGL v: 1.5 platforms: device: 0 drv: radeonsi device: 1 drv: swrast
gbm: drv: kms_swrast surfaceless: drv: radeonsi wayland: drv: radeonsi x11:
drv: radeonsi
API: OpenGL v: 4.6 compat-v: 4.5 vendor: amd mesa v: 25.2.4-arch1.1
glx-v: 1.4 direct-render: yes renderer: AMD Radeon RX 6700 (radeonsi navi22
LLVM 20.1.8 DRM 3.64 6.16.8-arch3-1) device-ID: 1002:73df display-ID: :1.0
API: Vulkan v: 1.4.321 surfaces: N/A device: 0 type: discrete-gpu
driver: mesa radv device-ID: 1002:73df
Info: Tools: api: clinfo, eglinfo, glxinfo, vulkaninfo
de: kscreen-console,kscreen-doctor wl: wayland-info
x11: xdpyinfo, xprop, xrandr
Audio:
Device-1: Intel Alder Lake-S HD Audio vendor: ASUSTeK driver: snd_hda_intel
v: kernel bus-ID: 00:1f.3 chip-ID: 8086:7ad0
Device-2: Advanced Micro Devices [AMD/ATI] Navi 21/23 HDMI/DP Audio
driver: snd_hda_intel v: kernel pcie: speed: 16 GT/s lanes: 16
bus-ID: 03:00.1 chip-ID: 1002:ab28
Device-3: Microdia USB Live camera driver: snd-usb-audio,uvcvideo
type: USB rev: 2.0 speed: 480 Mb/s lanes: 1 bus-ID: 1-6:3 chip-ID: 0c45:636b
API: ALSA v: k6.16.8-arch3-1 status: kernel-api
Server-1: PipeWire v: 1.4.8 status: active with: 1: pipewire-pulse
status: active 2: wireplumber status: active 3: pipewire-alsa type: plugin
4: pw-jack type: plugin
Sensors:
System Temperatures: cpu: 44.0 C mobo: N/A gpu: amdgpu temp: 50.0 C
mem: 48.0 C
Fan Speeds (rpm): N/A gpu: amdgpu fan: 623
Info:
Memory: total: 64 GiB available: 62.6 GiB used: 19.99 GiB (31.9%)
Processes: 546 Power: uptime: 2h 52m wakeups: 0 Init: systemd v: 258
default: graphical
Packages: pm: pacman pkgs: 1179 Compilers: gcc: 15.2.1 Shell: Bash
v: 5.3.3 running-in: konsole inxi: 3.3.39
Minecraft has a lot of tearing without the zink workaround.
I did add this param to kernel
acefour@acefour-2 ~]$ cat /etc/dracut.conf.d/amd.conf
kernel_cmdline="amdgpu.ppfeaturemask=0xfffd3fff"
Maybe it’s a minecraft issue!
AceFour
October 2, 2025, 11:54pm
71
It is not beyond the realm of possibilities. Searching for minecraft amd 6700 yields a lot threads about opengl calls being poorly handled. Many of the fixes don’t work, or at least, don’t work anymore like use shaders or sodium. One user reported rolling back to 6.9.7 kernel and that fixed his issue, so who knows. At least the zink fix works and vkmark doesn’t fail. I did see some minor tearing in the brave browser after the last package updates.
well fugg, google earth just crashed my amdgpu
Oct 02 22:02:06 acefour-2 kernel: amdgpu 0000:03:00.0: amdgpu: Dumping IP State
Oct 02 22:02:06 acefour-2 kernel: amdgpu 0000:03:00.0: amdgpu: Dumping IP State Completed
Oct 02 22:02:06 acefour-2 kernel: amdgpu 0000:03:00.0: amdgpu: [drm] AMDGPU device coredump file has been created
Oct 02 22:02:06 acefour-2 kernel: amdgpu 0000:03:00.0: amdgpu: [drm] Check your /sys/class/drm/card1/device/devcoredump/data
Oct 02 22:02:06 acefour-2 kernel: amdgpu 0000:03:00.0: amdgpu: ring gfx_0.1.0 timeout, signaled seq=2742593, emitted seq=2742595
Oct 02 22:02:06 acefour-2 kernel: amdgpu 0000:03:00.0: amdgpu: Process information: process kwin_wayland pid 7083 thread kwin_wayla:cs0 pid 7129
Oct 02 22:02:06 acefour-2 kernel: amdgpu 0000:03:00.0: amdgpu: Starting gfx_0.1.0 ring reset
Oct 02 22:02:06 acefour-2 kernel: amdgpu 0000:03:00.0: amdgpu: Ring gfx_0.1.0 reset failed
Oct 02 22:02:06 acefour-2 kernel: amdgpu 0000:03:00.0: amdgpu: GPU reset begin!
Oct 02 22:02:06 acefour-2 kernel: amdgpu 0000:03:00.0: amdgpu: MODE1 reset
Oct 02 22:02:06 acefour-2 kernel: amdgpu 0000:03:00.0: amdgpu: GPU mode1 reset
Oct 02 22:02:06 acefour-2 kernel: amdgpu 0000:03:00.0: amdgpu: GPU smu mode1 reset
Oct 02 22:02:17 acefour-2 kernel: amdgpu 0000:03:00.0: amdgpu: GPU reset succeeded, trying to resume
Oct 02 22:02:17 acefour-2 kernel: amdgpu 0000:03:00.0: amdgpu: PSP is resuming...
Oct 02 22:02:23 acefour-2 kernel: [drm:psp_v11_0_memory_training [amdgpu]] *ERROR* send training msg failed.
Oct 02 22:02:23 acefour-2 kernel: amdgpu 0000:03:00.0: amdgpu: Failed to process memory training!
Oct 02 22:02:23 acefour-2 kernel: amdgpu 0000:03:00.0: amdgpu: resume of IP block <psp> failed -62
Oct 02 22:02:23 acefour-2 kernel: amdgpu 0000:03:00.0: amdgpu: GPU reset end with ret = -62
Oct 02 22:02:23 acefour-2 kernel: amdgpu 0000:03:00.0: amdgpu: GPU Recovery Failed: -62
Oct 02 22:02:33 acefour-2 kernel: amdgpu 0000:03:00.0: amdgpu: Dumping IP State
Oct 02 22:02:33 acefour-2 kernel: amdgpu 0000:03:00.0: amdgpu: Dumping IP State Completed
Oct 02 22:02:33 acefour-2 kernel: amdgpu 0000:03:00.0: amdgpu: [drm] AMDGPU device coredump file has been created
Oct 02 22:02:33 acefour-2 kernel: amdgpu 0000:03:00.0: amdgpu: [drm] Check your /sys/class/drm/card1/device/devcoredump/data
Oct 02 22:02:33 acefour-2 kernel: amdgpu 0000:03:00.0: amdgpu: ring gfx_0.1.0 timeout, signaled seq=2742595, emitted seq=2742595
Oct 02 22:02:33 acefour-2 kernel: amdgpu 0000:03:00.0: amdgpu: Process information: process kwin_wayland pid 7083 thread kwin_wayla:cs0 pid 7129
Oct 02 22:02:33 acefour-2 kernel: amdgpu 0000:03:00.0: amdgpu: Starting gfx_0.1.0 ring reset
Oct 02 22:02:33 acefour-2 kernel: amdgpu 0000:03:00.0: amdgpu: Ring gfx_0.1.0 reset failed
Oct 02 22:02:33 acefour-2 kernel: amdgpu 0000:03:00.0: amdgpu: GPU reset begin!
AceFour
November 13, 2025, 10:24pm
73
Looks like no movement via Mesa on the ring timeout. Minecraft still screen tearing without going through zink. Played with flags on Brave/Chromium and still would occasionally lock up the display. Back to Firefox/uBlock. So far no issues.
AceFour
November 13, 2025, 11:15pm
74
well poop, now zink interface doesn’t stop tearing with Minecraft. I might as well go back to nVidia and dump AMD at this point. Been going on for months now
I can’t believe you are having these issues. If i was a gamer I’d try my 6750XT and see. I have no issues using it for all the things I’m doing. Do you have it set up via the arch wiki with hardware acceleration etc and using vulkan? enviromental variables set in environment file?
AceFour
November 14, 2025, 12:25am
76
$ cat /etc/modprobe.d/amdgpu.conf
options amdgpu si_support=1
options amdgpu cik_support=1
$ alias mc
alias mc='MESA_LOADER_DRIVER_OVERRIDE=zink multimc'
Zink use to work but now I get square tearing that is like 1/5 the size of the monitor. If I use vanilla MC without zink I get the vertical line tearing. Every mesa release seems to just keep getting worse and worse.
I switched to Firefox because Brave was causing the the ring timeout at times.
AceFour
November 14, 2025, 12:57am
77
Playing with multiMC I was playing with shaders. If I turn it off the tearing stops with zink. Before the update I could run shaders no problem. Having flashbacks to Apple making older phones crappier to make people upgrade to the new iPhone and a new charger for $89.95.
AceFour
November 14, 2025, 1:28am
78
dropping resolution I can go back to using shaders under zink. I can make it work but every release seem less capable then the previous.
1 Like
AceFour
November 14, 2025, 6:12pm
79
Did a system update before I went and tried Cosmic DE. Under Cosmic I can play both vanilla minecraft and multimc w/shaders on 4k without the zink trick. Firefox and Brave both have tearing issues with WebGL2. Seems chromium derived browsers can fix this via
alias brave='brave --password-store=kwallet6 --ozone-platform=wayland --use-angle=vulkan --enable-features=Vulkan,VulkanFromANGLE,DefaultANGLEVulkan &'
Cosmic does experience screen tearing if you open the Cosmic terminal full screen. Not sure why but I tried Ghostty and it did not experience tearing.
Seems like there is some issue with Plasma. Both monitors have AMD Freesync
Turning on “Allow in fullscreen windows” does not make a difference in Plasma for screen tearing with Minecraft.
Going to fire up Hyprland and see what happens there next . . .
Update: Hyprland vanilla minecraft after going into a game immediately starts tearing and the minecraft-launcher core dumps and locks up the screen. At lease amdgpu isn’t doing a ring timeout, which is nice. I guess.
In short, nothing seems entirely stable. Cosmic has better results then Plasma for running Minecraft. Ghostty doesn’t tear fullscreen on Cosmic but Cosmic terminal does. WebGL2 is an issue for both Chromium-based browsers and Firefox. There is mitigation for Chromium-based browsers.
AceFour
December 24, 2025, 4:42am
80
I replaced the 6700XT with a 9070 w/16gb card. I can run Minecraft with shaders on a 4k resolution without using Zink. There is no tearing, no issues and smooth play . Clearly the issue is the support of the 6700 card. Likely the RNDA4 goes down a completely different path then the older card.
My UPS is under powered https://www.amazon.com/dp/B01FWAZEIU?th=1
