I see the same issue under Fedora 42 running the default Gnome desktop. Also using an AMD RX580. I'll gather my system info later.

I see something similar, however I also see a crash some Arcade Maps, I've attached the Proton log, and the crash logs from SC2.

It happons considently at the same place in the arcade map "Direct Strike".
This map has server graphis issues, where the entire map goes red, but also some objects on the map like the fire from the photon cannon.

I am curius as to if this reproduce with AMD graphics.

I play on a laptop with a dual Grafhics card setup Intel/Nvidia it does not matter if use use the NVidia or Intel.
I've tested both with XWayland and native Xorg.

The Video below is with it started from Heroic, but I have also had this issue directly from Steam. Heroic was an attempt to woraround the issue.

Battle.net is the main app and it does not crash. I've tried to keep track of when the crash happoned in the steam-0.log it happons close to the laster shader error.

0428:err:d3dcompiler:D3DCompile2 Failed to compile shader, vkd3d result -4.
0428:err:d3dcompiler:D3DCompile2 Shader log:
0428:err:d3dcompiler:D3DCompile2     <anonymous>:1556:19: W5300: Implicit truncation of vector type.
0428:err:d3dcompiler:D3DCompile2     <anonymous>:1798:48: W5300: Implicit truncation of vector type.
0428:err:d3dcompiler:D3DCompile2     <anonymous>:1799:48: W5300: Implicit truncation of vector type.
0428:err:d3dcompiler:D3DCompile2     <anonymous>:1800:48: W5300: Implicit truncation of vector type.
0428:err:d3dcompiler:D3DCompile2     <anonymous>:3608:8: W5300: Implicit truncation of vector type.
0428:err:d3dcompiler:D3DCompile2     <anonymous>:3645:8: E5005: Function "tex3Dbias" is not defined.
0428:err:d3dcompiler:D3DCompile2 

I belive the norm is to have only one issue for each game, if not let me know and I will open a new issue.

SC2-CrashInfo-2025-06-10 21.29.21.tar.gz

https://github.com/user-attachments/assets/c4748101-d203-4030-822e-e4c6255f6b24

I can confirm the right-and-bottom-mouse-scroll issue with Pop OS 22.04 (based on Ubuntu). It happens to me when I do Fullscreen mode using 1920x1080 resolution on a 4K monitor. If I play the game in native resolution of 4K, it works fine. I also tried a couple other resolutions and have the same issue.

Encounter the problem on CachyOS (current version, rolling release, ARCH based).

I have this issue including the graphical glitch reported by @KrauseStefan (see attached images).

Short system details:

• Ubuntu 25.10
• NVIDIA GA106M [GeForce RTX 3060 Mobile / Max-Q] Driver Version: 580.95.05
• Proton 10.0-2 (beta) - same results with other Proton versions.
• Vulkan v: 1.4.321

Detailed system info

Steam log

I don't know if it's customary to submit already reported issues. I hope my contribution would be helpful.

I have the same issues on the Direct Strike arcade map: both red textures and a crash after destruction of buildings.

System Information
GPU: AMD RX 570
Video Driver Version: Mesa 25.2.6-cachyos1.2
Kernel Version: 6.17.7-3-cachyos
Full system information report: https://gist.github.com/CuriousLurker/6819583ad585428375df4db6959c74a3
Proton version: Experimental

Steam Runtime Diagnostics: https://gist.github.com/CuriousLurker/edb5c3d71d7a760abe1bd93b8f4256b3
Steam log: steam-13454620091438071808.tar.gz
SC2 crash log: 2025-11-06 20.54.59 CACHYOS-DESKTOP B95299 Crash.tar.gz

Starcraft 2 on Proton 10.0-2 beta and 10.0-3 works well at native resolution windowed mode only. 4k scrolling issues on the right and bottom of display similar to OP.

Proton experimental stuttered terribly. Earlier protons would not launch battle.net.

Nixos 25.11

niri + xwayland-satellite
Starcraft 2 installed using lutris
https://gist.github.com/bigFin/4f94aa75ebd3bfee157ac0a8203984cb

With the latest build of GE-Proton-29 the issue seems to be fixed.
There are still red graphical glitches but the game does no longer crash in that map.
I haven't tested it with regular Proton yet. I do expect the fix to come from regular proton based on reading the change log.

With the latest build of GE-Proton-29 the issue seems to be fixed. There are still red graphical glitches but the game does no longer crash in that map. I haven't tested it with regular Proton yet. I do expect the fix to come from regular proton based on reading the change log.

Unfortunately I can not confirm this. I still have the same graphical issues and crashes as before. The only thing that changed is that the game crashes less frequent.

I've found a way to make Direct Strike work: no red textures and no crashes.

https://old.reddit.com/r/wine_gaming/comments/1dgvzup/custom_assets_in_starcraft_2_are_red_on_my_nixos/lp7jgpx

Tested with Proton 10.0-4.

I've found a way to make Direct Strike work: no red textures and no crashes.

https://old.reddit.com/r/wine_gaming/comments/1dgvzup/custom_assets_in_starcraft_2_are_red_on_my_nixos/lp7jgpx

Tested with Proton 10.0-4.

Tested it with proton GE-Proton-29 and it made the game crash again.

I've found a way to make Direct Strike work: no red textures and no crashes.

https://old.reddit.com/r/wine_gaming/comments/1dgvzup/custom_assets_in_starcraft_2_are_red_on_my_nixos/lp7jgpx

Tested with Proton 10.0-4.

Thanks! It fixed red tile in many mods.
For the record I am using GE-Proton10-29.

On the original cursor issue, I've noticed that Command and Conquer 4 has the same issue if not using native resolution. Maybe this is common to more older games?

I've found a way to make Direct Strike work: no red textures and no crashes.

https://old.reddit.com/r/wine_gaming/comments/1dgvzup/custom_assets_in_starcraft_2_are_red_on_my_nixos/lp7jgpx

Tested with Proton 10.0-4.

Tried this on Proton 11 beta and it seems the workaround is broken. Make sure to stay on Proton 10 if you have this problem.

Aside from the OP's report and the shader issue, there are more issues I've encountered (GNOME, Wayland, CachyOS):

The cursor occasionally exits the window border, when you try to edge-scroll while pressing keyboard keys at the same time. This makes the majority of high APM games effectively unplayable on multi-monitor setup.

I believe this happens when all of the followings are true:

• You're using XWayland (which is default)
• You have fractional scaling enabled on your monitor

Can be worked around by either:

• Using winewayland (use Proton-GE and add PROTON_ENABLE_WAYLAND=1 option)
• Using gamescope (any backend) with --force-grab-cursor, though gamescope does have some decent issues on its own.

On exiting the app the Battle.net launcher process lingers. You can manually click "STOP" button in Steam to force killing it as a workaround.

Typing Korean doesn't work. Neither ibus nor fcitx is able to switch and input Korean at all. The only viable way to type it is to copy-paste from the outside.

I reckon that any games that try to do something nonstandard with CJK IMEs have problem with Proton so it's not really a SC2-specific problem. On Windows I remember SC2 had some fancy Korean/English mode-switching indicator, which may imply that it tweaks at least a part of the Windows IME themselves.

I believe this problem is likely the same one as described in https://github.com/ValveSoftware/Proton/issues/7199#issuecomment-2134139472. I'm not sure if it had been solved in Diablo 4 but the developer checks out.

with any of the protos, I can't even see the game attempt to launch.

it did generate 3164 lines of logfile though.

I launched with

PROTON_ENABLE_WAYLAND=1 PROTON_LOG=1 %command%

with proton 10.0

The first thing I notice is:

 201   │ 138035.785:00cc:00d0:trace:seh:dispatch_exception code=6ba (RPC_S_SERVER_UNAVAILABLE) flags=0 addr=00006FFFFFC0D157
 202   │ 138035.785:00cc:00d0:trace:seh:dispatch_exception rip=00006fffffc0d157 rsp=000000000021d570 rbp=000000000021d660 eflags=00000202
 203   │ 138035.785:00cc:00d0:trace:seh:dispatch_exception rax=000000000021d590 rbx=000000000021e420 rcx=000000000021d590 rdx=0000000000000000
 204   │ 138035.785:00cc:00d0:trace:seh:dispatch_exception rsi=0000000000000064 rdi=00000000007ebf40  r8=0000000000000000  r9=0000000000000000
 205   │ 138035.785:00cc:00d0:trace:seh:dispatch_exception r10=00000000007ec690 r11=0000000000000010 r12=000000000021e420 r13=0000000000000005
 206   │ 138035.785:00cc:00d0:trace:seh:dispatch_exception r14=00006ffffeeecdb4 r15=00000000007ebf40 mxcsr=00001fa0
 207   │ 138035.785:00cc:00d0:trace:unwind:RtlVirtualUnwind2 type 1 base 6fffffc00000 rip 6fffffc0d157 rva d157 rsp 21d570
 208   │ 138035.785:00cc:00d0:trace:unwind:dump_unwind_info **** func d0e8-d15f
 209   │ 138035.785:00cc:00d0:trace:unwind:dump_unwind_info unwind info at 00006FFFFFCC1CE8 flags 0 prolog 0xf bytes function 00006FFFFFC0D0E8-00006FFFFFC0D15F
 210   │ 138035.785:00cc:00d0:trace:unwind:dump_unwind_info     0xf: subq $0xc8,%rsp
 211   │ 138035.785:00cc:00d0:trace:unwind:RtlVirtualUnwind2 inside epilog.

steam-10858654163093422080.log

and then with
2 │ Proton: 1789159601 experimental-11.0-20260910b-x86_64

I see the same thing:

 185   │ 138195.558:00dc:00e0:trace:seh:dispatch_exception code=6ba (RPC_S_SERVER_UNAVAILABLE) flags=0 addr=00006FFFFFBFD947
 186   │ 138195.558:00dc:00e0:trace:seh:dispatch_exception rip=00006fffffbfd947 rsp=000000000031d570 rbp=000000000031d660 eflags=00000202
 187   │ 138195.558:00dc:00e0:trace:seh:dispatch_exception rax=000000000031d590 rbx=000000000031e430 rcx=000000000031d590 rdx=0000000000000000
 188   │ 138195.558:00dc:00e0:trace:seh:dispatch_exception rsi=0000000000000064 rdi=00000000007ed260  r8=0000000000000000  r9=0000000000000000
 189   │ 138195.558:00dc:00e0:trace:seh:dispatch_exception r10=00000000007ed990 r11=000000000000c267 r12=000000000031d738 r13=0000000000000005
 190   │ 138195.558:00dc:00e0:trace:seh:dispatch_exception r14=00006fffff8fe260 r15=00006ffffe71e414 mxcsr=00001f80 cs=0033 ss=002b
 191   │ 138195.558:00dc:00e0:trace:unwind:RtlVirtualUnwind2 type 1 base 6fffffbf0000 rip 6fffffbfd947 rva d947 rsp 31d570
 192   │ 138195.558:00dc:00e0:trace:unwind:dump_unwind_info **** func d8d8-d94f
 193   │ 138195.558:00dc:00e0:trace:unwind:dump_unwind_info unwind info at 00006FFFFFCAEE64 flags 0 prolog 0xf bytes function 00006FFFFFBFD8D8-00006FFFFFBFD94F
 194   │ 138195.558:00dc:00e0:trace:unwind:dump_unwind_info     0xf: subq $0xc8,%rsp
 195   │ 138195.558:00dc:00e0:trace:unwind:RtlVirtualUnwind2 inside epilog.
 196   │ 138195.558:00dc:00e0:trace:unwind:RtlVirtualUnwind2 type 1 base 6fffff8b0000 rip 6fffff8e2496 rva 32496 rsp 31d640
 197   │ 138195.558:00dc:00e0:trace:unwind:dump_unwind_info **** func 32480-324cb
 198   │ 138195.558:00dc:00e0:trace:unwind:dump_unwind_info unwind info at 00006FFFFF92EC40 flags 0 prolog 0x8 bytes function 00006FFFFF8E2480-00006FFFFF8E24CB
 199   │ 138195.558:00dc:00e0:trace:unwind:dump_unwind_info     frame register rbp offset 0x0(%rsp)
 200   │ 138195.558:00dc:00e0:trace:unwind:dump_unwind_info     0x8: subq $0x20,%rsp

steam-10858654163093422080.log

I've been trying to launch directly, and haven't looked at battle.net issues

• https://github.com/ValveSoftware/Proton/issues/7302

though, my battle.net also does not launch.

I am also on Ubuntu 26.04