You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I am able to install the software and verify the following works, python -c "import open3d.ml.torch as ml3d" with out error,
However, when I run the example in readme, I got OpenGL error.
INFO - 2021-12-11 14:35:22,038 - semantickitti - Found 42296 pointclouds for all {'idx': 0, 'name': '00_001256', 'path': '/mnt/c/Users/PeterHuang/Documents/KITTI_DATASET/data_odometry_velodyne/dataset/sequences/00/velodyne/001256.bin', 'split': 'all'} (122985, 3) INFO - 2021-12-11 14:35:22,260 - semantickitti - Found 42296 pointclouds for all FEngine (64 bits) created at 0x7f7075c42010 (threading is enabled) FEngine resolved backend: OpenGL X Error of failed request: GLXBadFBConfig Major opcode of failed request: 148 (GLX) Minor opcode of failed request: 0 () Serial number of failed request: 27 Current serial number in output stream: 27 Segmentation fault
My environment is WSL2 with CUDA driver install from Nvidia link
+-----------------------------------------------------------------------------+
| NVIDIA-SMI 510.00 Driver Version: 510.06 CUDA Version: 11.6 |
|-------------------------------+----------------------+----------------------+
I google online and tried to add the following export MESA_GL_VERSION_OVERRIDE=4.5, it seems work but render with a black screen.
The text was updated successfully, but these errors were encountered:
I have the same issue. I'm running OpenGL from within conda
CUDA 11.2
Pytorch 1.11
Open3D built from source
with the same result as above. Any solutions?
EDIT!!!
With below settings (either set directly in terminal or in .bashrc and .profile) it works for me. I assume the numbers set depends on the version of mesa drivers in use. I dont know if ALWAYS_SOFTWARE implies any performance hit, but I cannot see any. glxgears is actually faster with this setting, wheter it is a true performance gain or an artefact of the measurement method used I dont know.
Hi,
I am able to install the software and verify the following works,
python -c "import open3d.ml.torch as ml3d"
with out error,However, when I run the example in readme, I got OpenGL error.
INFO - 2021-12-11 14:35:22,038 - semantickitti - Found 42296 pointclouds for all {'idx': 0, 'name': '00_001256', 'path': '/mnt/c/Users/PeterHuang/Documents/KITTI_DATASET/data_odometry_velodyne/dataset/sequences/00/velodyne/001256.bin', 'split': 'all'} (122985, 3) INFO - 2021-12-11 14:35:22,260 - semantickitti - Found 42296 pointclouds for all FEngine (64 bits) created at 0x7f7075c42010 (threading is enabled) FEngine resolved backend: OpenGL X Error of failed request: GLXBadFBConfig Major opcode of failed request: 148 (GLX) Minor opcode of failed request: 0 () Serial number of failed request: 27 Current serial number in output stream: 27 Segmentation fault
My environment is WSL2 with CUDA driver install from Nvidia link
+-----------------------------------------------------------------------------+
| NVIDIA-SMI 510.00 Driver Version: 510.06 CUDA Version: 11.6 |
|-------------------------------+----------------------+----------------------+
I google online and tried to add the following
export MESA_GL_VERSION_OVERRIDE=4.5
, it seems work but render with a black screen.The text was updated successfully, but these errors were encountered: