[Bug]: Failing to build mmcv from source with CUDA 12.8 + PyTorch 2.8 + RTX 5090

[SYangKay]

Checklist

• The issue has not been resolved by following the troubleshooting guide
• The issue exists on a clean installation of Fooocus
• The issue exists in the current version of Fooocus
• The issue has not been reported before recently
• The issue has been reported before but has not been fixed yet

What happened?

I recently purchased an NVIDIA GeForce RTX 5090. Since this GPU requires CUDA 12.8+, I’ve installed CUDA 12.8 manually and set up my environment with the latest PyTorch nightly build (2.8.0+cu128).

However, when attempting to build mmcv from source (since there are no official wheels for this configuration yet), the compilation consistently fails—specifically at the CUDA extension stage. Has anyone successfully compiled mmcv with 5090? I

Enviroment:
GPU: NVIDIA GeForce RTX 5090

CUDA: 12.8 (installed manually via .run file)

PyTorch: 2.8.0.dev20250410+cu128 (nightly build)

Python: 3.12（also tried 3.10）

GCC: 11.4

OS: Ubuntu 25.04 (Plucky)

Steps to reproduce the problem

What I did:

git clone https://github.com/open-mmlab/mmcv.git
cd mmcv
git checkout v2.1.0
pip install -r requirements/build.txt
MMCV_WITH_OPS=1 pip install -v -e .

What should have happened?

.

What browsers do you use to access Fooocus?

No response

Where are you running Fooocus?

None

What operating system are you using?

ubuntu25.04

Console logs

...
ninja: build stopped: subcommand failed.
    Traceback (most recent call last):
      File "/home/yang/anaconda3/envs/univ2xCUDA128_py312/lib/python3.12/site-packages/torch/utils/cpp_extension.py", line 2556, in _run_ninja_build
        subprocess.run(
      File "/home/yang/anaconda3/envs/univ2xCUDA128_py312/lib/python3.12/subprocess.py", line 573, in run
        raise CalledProcessError(retcode, process.args,
    subprocess.CalledProcessError: Command '['ninja', '-v', '-j', '4']' returned non-zero exit status 1.

    The above exception was the direct cause of the following exception:

    Traceback (most recent call last):
      File "<string>", line 2, in <module>
      File "<pip-setuptools-caller>", line 34, in <module>
      File "/home/yang/mmcv/setup.py", line 521, in <module>
        setup(
      File "/home/yang/anaconda3/envs/univ2xCUDA128_py312/lib/python3.12/site-packages/setuptools/__init__.py", line 117, in setup
        return distutils.core.setup(**attrs)
               ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
      File "/home/yang/anaconda3/envs/univ2xCUDA128_py312/lib/python3.12/site-packages/setuptools/_distutils/core.py", line 186, in setup
        return run_commands(dist)
               ^^^^^^^^^^^^^^^^^^
      File "/home/yang/anaconda3/envs/univ2xCUDA128_py312/lib/python3.12/site-packages/setuptools/_distutils/core.py", line 202, in run_commands
        dist.run_commands()
      File "/home/yang/anaconda3/envs/univ2xCUDA128_py312/lib/python3.12/site-packages/setuptools/_distutils/dist.py", line 983, in run_commands
        self.run_command(cmd)
      File "/home/yang/anaconda3/envs/univ2xCUDA128_py312/lib/python3.12/site-packages/setuptools/dist.py", line 999, in run_command
        super().run_command(command)
      File "/home/yang/anaconda3/envs/univ2xCUDA128_py312/lib/python3.12/site-packages/setuptools/_distutils/dist.py", line 1002, in run_command
        cmd_obj.run()
      File "/home/yang/anaconda3/envs/univ2xCUDA128_py312/lib/python3.12/site-packages/setuptools/command/develop.py", line 35, in run
        self.install_for_development()
      File "/home/yang/anaconda3/envs/univ2xCUDA128_py312/lib/python3.12/site-packages/setuptools/command/develop.py", line 112, in install_for_development
        self.run_command('build_ext')
      File "/home/yang/anaconda3/envs/univ2xCUDA128_py312/lib/python3.12/site-packages/setuptools/_distutils/cmd.py", line 339, in run_command
        self.distribution.run_command(command)
      File "/home/yang/anaconda3/envs/univ2xCUDA128_py312/lib/python3.12/site-packages/setuptools/dist.py", line 999, in run_command
        super().run_command(command)
      File "/home/yang/anaconda3/envs/univ2xCUDA128_py312/lib/python3.12/site-packages/setuptools/_distutils/dist.py", line 1002, in run_command
        cmd_obj.run()
      File "/home/yang/anaconda3/envs/univ2xCUDA128_py312/lib/python3.12/site-packages/setuptools/command/build_ext.py", line 99, in run
        _build_ext.run(self)
      File "/home/yang/anaconda3/envs/univ2xCUDA128_py312/lib/python3.12/site-packages/setuptools/_distutils/command/build_ext.py", line 365, in run
        self.build_extensions()
      File "/home/yang/anaconda3/envs/univ2xCUDA128_py312/lib/python3.12/site-packages/torch/utils/cpp_extension.py", line 1052, in build_extensions
        build_ext.build_extensions(self)
      File "/home/yang/anaconda3/envs/univ2xCUDA128_py312/lib/python3.12/site-packages/setuptools/_distutils/command/build_ext.py", line 481, in build_extensions
        self._build_extensions_serial()
      File "/home/yang/anaconda3/envs/univ2xCUDA128_py312/lib/python3.12/site-packages/setuptools/_distutils/command/build_ext.py", line 507, in _build_extensions_serial
        self.build_extension(ext)
      File "/home/yang/anaconda3/envs/univ2xCUDA128_py312/lib/python3.12/site-packages/setuptools/command/build_ext.py", line 264, in build_extension
        _build_ext.build_extension(self, ext)
      File "/home/yang/anaconda3/envs/univ2xCUDA128_py312/lib/python3.12/site-packages/setuptools/_distutils/command/build_ext.py", line 562, in build_extension
        objects = self.compiler.compile(
                  ^^^^^^^^^^^^^^^^^^^^^^
      File "/home/yang/anaconda3/envs/univ2xCUDA128_py312/lib/python3.12/site-packages/torch/utils/cpp_extension.py", line 836, in unix_wrap_ninja_compile
        _write_ninja_file_and_compile_objects(
      File "/home/yang/anaconda3/envs/univ2xCUDA128_py312/lib/python3.12/site-packages/torch/utils/cpp_extension.py", line 2209, in _write_ninja_file_and_compile_objects
        _run_ninja_build(
      File "/home/yang/anaconda3/envs/univ2xCUDA128_py312/lib/python3.12/site-packages/torch/utils/cpp_extension.py", line 2573, in _run_ninja_build
        raise RuntimeError(message) from e
    RuntimeError: Error compiling objects for extension
    error: subprocess-exited-with-error
    
    × python setup.py develop did not run successfully.
    │ exit code: 1
    ╰─> See above for output.
    
    note: This error originates from a subprocess, and is likely not a problem with pip.

Additional information

Is mmcv currently compatible with CUDA 12.8 / PyTorch 2.8?

Has anyone successfully compiled mmcv with 5090? I'd greatly appreciate hearing about any working configurations or suggestions.

Thanks in advance for your help and for maintaining this excellent library!

[EgorT78]

I have the same problem. RTX 5080.

[Ichiruchan]

Same question with cuda 12.8 pytorch 2.7.0 python3.9 rtx 5080

[clode1002]

Open the Fooocus\python_embeded folder
Open cmd
Enter the command to update

python.exe -m pip install --upgrade torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu128

Re-execute, this is how I solved it

[alibakhtiari2]

https://github.com/alibakhtiari2/fooocusrtx508090 use this repository which solved the entire issues ... please like! also you can download entire package from the links. thanks

[JackBo04]

Followed the instruction on Build MMCV from source, i have installed mmcv successfully.

Here are my envirment:
sys.platform: win32
Python: 3.10.18 | packaged by Anaconda, Inc. | (main, Jun 5 2025, 13:08:55) [MSC v.1929 64 bit (AMD64)]
CUDA available: True
MUSA available: False
numpy_random_seed: 2147483648
GPU 0: NVIDIA GeForce RTX 5080 Laptop GPU
CUDA_HOME: C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v12.8
NVCC: Cuda compilation tools, release 12.8, V12.8.93
MSVC: 用于 x64 的 Microsoft (R) C/C++ 优化编译器 19.44.35217 版
GCC: n/a
PyTorch: 2.8.0+cu128
PyTorch compiling details: PyTorch built with:

• C++ Version: 201703
• MSVC 193833145
• Intel(R) oneAPI Math Kernel Library Version 2025.2-Product Build 20250620 for Intel(R) 64 architecture applications
• Intel(R) MKL-DNN v3.7.1 (Git Hash 8d263e693366ef8db40acc569cc7d8edf644556d)
• OpenMP 2019
• LAPACK is enabled (usually provided by MKL)
• CPU capability usage: AVX2
• CUDA Runtime 12.8
• NVCC architecture flags: -gencode;arch=compute_61,code=sm_61;-gencode;arch=compute_70,code=sm_70;-gencode;arch=compute_75,code=sm_75;-gencode;arch=compute_80,code=sm_80;-gencode;arch=compute_86,code=sm_86;-gencode;arch=compute_90,code=sm_90;-gencode;arch=compute_100,code=sm_100;-gencode;arch=compute_120,code=sm_120
• CuDNN 91.0.2 (built against CUDA 12.9)
• Magma 2.5.4
• Build settings: BLAS_INFO=mkl, BUILD_TYPE=Release, COMMIT_SHA=a1cb3cc05d46d198467bebbb6e8fba50a325d4e7, CUDA_VERSION=12.8, CUDNN_VERSION=9.10.2, CXX_COMPILER=C:/actions-runner/_work/pytorch/pytorch/pytorch/.ci/pytorch/windows/tmp_bin/sccache-cl.exe, CXX_FLAGS=/DWIN32 /D_WINDOWS /EHsc /Zc:__cplusplus /bigobj /FS /utf-8 -DUSE_PTHREADPOOL -DNDEBUG -DUSE_KINETO -DLIBKINETO_NOROCTRACER -DLIBKINETO_NOXPUPTI=ON -DUSE_FBGEMM -DUSE_XNNPACK -DSYMBOLICATE_MOBILE_DEBUG_HANDLE /wd4624 /wd4068 /wd4067 /wd4267 /wd4661 /wd4717 /wd4244 /wd4804 /wd4273, LAPACK_INFO=mkl, PERF_WITH_AVX=1, PERF_WITH_AVX2=1, TORCH_VERSION=2.8.0, USE_CUDA=ON, USE_CUDNN=ON, USE_CUSPARSELT=OFF, USE_GFLAGS=OFF, USE_GLOG=OFF, USE_GLOO=ON, USE_MKL=ON, USE_MKLDNN=ON, USE_MPI=OFF, USE_NCCL=OFF, USE_NNPACK=OFF, USE_OPENMP=ON, USE_ROCM=OFF, USE_ROCM_KERNEL_ASSERT=OFF, USE_XCCL=OFF, USE_XPU=OFF,

TorchVision: 0.23.0+cu128
OpenCV: 4.12.0
MMEngine: 0.11.0rc0
MMCV: 2.2.0
MMCV Compiler: MSVC 194435217
MMCV CUDA Compiler: 12.8

[whitebeltmillionaire]

i followed this video to get it working on my RTX 5080:

https://youtu.be/W0KWJxRRtAo?si=rLkV5os2fTjc0n9g

[Jaewoo-Net]

Instead of installing mmcv-full, I just cloned the MMCV repo locally and added it to sys.path.

import sys
sys.path.insert(0, r"<user path>\mmcv_1.x") 

It’s a bit of a hassle, but it works.