Failed to set CUDA shm size 97280: named symbol not found on CUDA 12.9
Nobody has claimed this yet.
- Dominant language
- Cuda
- Stars
- 351
- Forks
- 84
- Avg merge
- 3d 2h
- Merged PRs (30d)
- 35
Description
I'm getting this error [Failed to set CUDA shm size 97280: named symbol not found] when running:
from gpu4pyscf import dft as dftgpu
CuPy itself is working fine, and a couple of commits earlier this was working. I've setup gpu4pyscf for cuda12 as well as cutensor12.
Thank you for your time.
Update: After reverting to version 1.4.0 this error does not occur.
Contributor guide
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
Research direction
Start by reproducing the failure from from gpu4pyscf import dft as dftgpu with CUDA 12.9, CuPy, and the CUDA 12/cuTENSOR 12 setup described, then compare the current version with 1.4.0 where it does not occur. Trace the code that attempts to set the CUDA shared-memory size and identify the version compatibility change; done means the import no longer reports the named-symbol error on the supported setup.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- python
- Domain
- hpc
- Issue type
- Bug
- Difficulty
- 4/5
- Estimated time
- 3-5 days
- Activity status
- Stale
- Clarity
- Needs clarification
- Newbie friendliness
- 25/100