coreylowman / cudarc

Safe rust wrapper around CUDA toolkit

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Issue for building my rust application with cudarc v0.11.1 (Unsupported cuda toolkit version: `10010`)

jay11ca39 opened this issue · comments

I am using: cudarc v0.11.1

It is giving weird error:

error: failed to run custom build command for `cudarc v0.11.1`
Unsupported cuda toolkit version: `10010`. Please raise a github issue.

I have removed all cuda-toolkit, installed freshly and rebooted my machine :
Following is my cuda toolkit details:
$ nvcc --version

nvcc: NVIDIA (R) Cuda compiler driver
Copyright (c) 2005-2024 NVIDIA Corporation
Built on Wed_Apr_17_19:19:55_PDT_2024
Cuda compilation tools, release 12.5, V12.5.40
Build cuda_12.5.r12.5/compiler.34177558_0

$ nvidia-smi

Thu May 30 11:09:47 2024
+-----------------------------------------------------------------------------------------+
| NVIDIA-SMI 550.54.15 Driver Version: 550.54.15 CUDA Version: 12.4 |
|-----------------------------------------+------------------------+----------------------+
| GPU Name Persistence-M | Bus-Id Disp.A | Volatile Uncorr. ECC |
| Fan Temp Perf Pwr:Usage/Cap | Memory-Usage | GPU-Util Compute M. |
| | | MIG M. |
|=========================================+========================+======================|
| 0 Quadro P400 Off | 00000000:9E:00.0 Off | N/A |
| 34% 36C P8 N/A / N/A | 243MiB / 2048MiB | 0% Default |
| | | N/A |
+-----------------------------------------+------------------------+----------------------+

+-----------------------------------------------------------------------------------------+
| Processes: |
| GPU GI CI PID Type Process name GPU Memory |
| ID ID Usage |
|=========================================================================================|
| 0 N/A N/A 2482 G /usr/lib/xorg/Xorg 69MiB |
| 0 N/A N/A 2620 G /usr/bin/gnome-shell 126MiB |
+-----------------------------------------------------------------------------------------+

Even i treid for CUDA different versions.

Currently we extract the version from the cuda.h file present on the system. We should move to either extracting from nvcc or nvidia-smi. Likely just running nvcc --version from build.rs and extracting the version from that.