Panicking Thread on WSL2: Error in wgpuSurfaceGetPreferredFormat: unsupported format

[gselzer]

Reporting here because the error asks me to!

I was running the following script from the pygfx README:

import pygfx as gfx
import pylinalg as la

cube = gfx.Mesh(
    gfx.box_geometry(200, 200, 200),
    gfx.MeshPhongMaterial(color="#336699"),
)

rot = la.quat_from_euler((0, 0.01), order="XY")

def animate():
    cube.local.rotation = la.quat_mul(rot, cube.local.rotation)

if __name__ == "__main__":
    gfx.show(cube, before_render=animate)

When running the script (with the RUST_BACKTRACE environment variable set), I receive the following error, observed both with Qt and glfw windows:

thread '<unnamed>' panicked at src/lib.rs:3849:17:
Error in wgpuSurfaceGetPreferredFormat: unsupported format.
Please report it to https://github.com/gfx-rs/wgpu-native
stack backtrace:
   0: rust_begin_unwind
             at /rustc/07dca489ac2d933c78d3c5158e3f43beefeb02ce/library/std/src/panicking.rs:645:5
   1: core::panicking::panic_fmt
             at /rustc/07dca489ac2d933c78d3c5158e3f43beefeb02ce/library/core/src/panicking.rs:72:14
   2: wgpuSurfaceGetPreferredFormat
   3: ffi_call_unix64
   4: ffi_call_int
   5: cdata_call
             at /project/src/c/_cffi_backend.c:3201:5
   6: _PyObject_Call
             at /usr/local/src/conda/python-3.11.9/Objects/call.c:343:19
   7: PyObject_Call
             at /usr/local/src/conda/python-3.11.9/Objects/call.c:355:12
   8: do_call_core
             at /usr/local/src/conda/python-3.11.9/Python/ceval.c:7349:12
   9: _PyEval_EvalFrameDefault
             at /usr/local/src/conda/python-3.11.9/Python/ceval.c:5376:22
  10: _PyEval_EvalFrame
             at /usr/local/src/conda/python-3.11.9/Include/internal/pycore_ceval.h:73:16
  11: _PyEval_Vector
             at /usr/local/src/conda/python-3.11.9/Python/ceval.c:6434:24
  12: _PyFunction_Vectorcall
             at /usr/local/src/conda/python-3.11.9/Objects/call.c:393:16
  13: _PyObject_FastCallDictTstate
             at /usr/local/src/conda/python-3.11.9/Objects/call.c:141:15
  14: _PyObject_Call_Prepend
             at /usr/local/src/conda/python-3.11.9/Objects/call.c:482:24
  15: slot_tp_init
             at /usr/local/src/conda/python-3.11.9/Objects/typeobject.c:7855:15
  16: type_call
             at /usr/local/src/conda/python-3.11.9/Objects/typeobject.c:1103:19
  17: _PyObject_MakeTpCall
             at /usr/local/src/conda/python-3.11.9/Objects/call.c:214:18
  18: _PyEval_EvalFrameDefault
             at /usr/local/src/conda/python-3.11.9/Python/ceval.c:4769:23
  19: _PyEval_EvalFrame
             at /usr/local/src/conda/python-3.11.9/Include/internal/pycore_ceval.h:73:16
  20: _PyEval_Vector
             at /usr/local/src/conda/python-3.11.9/Python/ceval.c:6434:24
  21: PyEval_EvalCode
             at /usr/local/src/conda/python-3.11.9/Python/ceval.c:1148:21
  22: run_eval_code_obj
             at /usr/local/src/conda/python-3.11.9/Python/pythonrun.c:1741:9
  23: run_mod
             at /usr/local/src/conda/python-3.11.9/Python/pythonrun.c:1762:19
  24: pyrun_file
             at /usr/local/src/conda/python-3.11.9/Python/pythonrun.c:1657:15
  25: _PyRun_SimpleFileObject
             at /usr/local/src/conda/python-3.11.9/Python/pythonrun.c:440:13
  26: _PyRun_AnyFileObject
             at /usr/local/src/conda/python-3.11.9/Python/pythonrun.c:79:15
  27: pymain_run_file_obj
             at /usr/local/src/conda/python-3.11.9/Modules/main.c:360:15
  28: pymain_run_file
             at /usr/local/src/conda/python-3.11.9/Modules/main.c:379:15
  29: pymain_run_python
             at /usr/local/src/conda/python-3.11.9/Modules/main.c:601:21
  30: Py_RunMain
             at /usr/local/src/conda/python-3.11.9/Modules/main.c:680:5
  31: Py_BytesMain
             at /usr/local/src/conda/python-3.11.9/Modules/main.c:734:12
  32: <unknown>
  33: __libc_start_main
  34: <unknown>
note: Some details are omitted, run with `RUST_BACKTRACE=full` for a verbose backtrace.
fatal runtime error: failed to initiate panic, error 5

pygfx: 0.2.0 wgpu: 0.15.3 System: Ubuntu 22.04.3 LTS, WSL2

Notably, the script runs just fine on Windows.

[almarklein]

Can you try the cube.py example from wgpu-py. That's a bit closer to the metal, and you could also try replacing this line https://github.com/pygfx/wgpu-py/blob/b5e9f66e13fb83b8976e65ad7aea8fa35437bd3f/examples/cube.py#L30 with render_texture_format = wgpu.TextureFormat.bgra8unorm_srgb.

[gselzer]

@almarklein thanks for the reply!

Can you try the cube.py example from wgpu-py. That's a bit closer to the metal, and you could also try replacing this line https://github.com/pygfx/wgpu-py/blob/b5e9f66e13fb83b8976e65ad7aea8fa35437bd3f/examples/cube.py#L30 with render_texture_format = wgpu.TextureFormat.bgra8unorm_srgb.

Cube example raises a simliar panic as the one I reported above.

Using your suggested Texture Format raises a different, non-panic exception, encountered on the following line:

Available adapters on this system:
llvmpipe (LLVM 15.0.7, 256 bits) (CPU) via Vulkan
D3D12 (Intel(R) Iris(R) Xe Graphics) (IntegratedGPU) via OpenGL
Traceback (most recent call last):
  File "/home/g/code/vispy/vispy/tmp.py", line 32, in <module>
    present_context.configure(device=device, format=render_texture_format)
  File "/home/g/miniforge3/envs/pymmcore-plus-sandbox/lib/python3.11/site-packages/wgpu/backends/wgpu_native/_api.py", line 461, in configure
    raise ValueError(
ValueError: Given format 'bgra8unorm-srgb' is not in supported formats []

[almarklein]

It looks like wgpu reports that the surface simply has zero supported formats. Could this be because it's on WSL2? Ideas @rajveermalviya ?