search

1inch / profanity2

Vanity address generator for Ethereum
325 stars 85 forks source link

windows11 error #23

Open Nobay3679 opened 1 year ago

Nobay3679 commented 1 year ago

Dispatcher.hpp:14:10: fatal error: CL/cl.h: No such file or directory

include <CL/cl.h>

      ^~~~~~~~~

my gpu is 3080ti and i have installed cuda12.1.1,what shall i do?my cpu is ryzen 5600. anyone can help me,thanks

k06a commented 1 year ago

I am not sure what should be done on Windows11, but ChatGPT provided some thoughts (chat.openai.com):

The error you're experiencing is related to the OpenCL library, not CUDA. Even though both CUDA and OpenCL are computing interfaces that allow software to use GPU for general-purpose computing, they are different and one cannot substitute the other. CUDA is proprietary to NVIDIA while OpenCL is an open standard that can be used by any GPU, including NVIDIA, AMD, and others.

The error you're seeing, "CL/cl.h: No such file or directory", is the compiler complaining that it cannot find the OpenCL header file. It seems like you don't have the OpenCL headers installed on your system.

Here are some steps you can follow to resolve this issue:

  1. Install OpenCL: NVIDIA's GPU drivers should come with OpenCL support. However, you will need to install the OpenCL headers separately. You can download these headers from the official OpenCL-Headers GitHub repository.
  2. Set the path: After installing the OpenCL headers, you will need to set the correct path so that the compiler knows where to find these headers. You can do this by adding the path to the OpenCL headers to your C_INCLUDE_PATH environment variable. Here's how you can do that:
    • Open a new terminal.
    • Type export C_INCLUDE_PATH=$C_INCLUDE_PATH:/path/to/your/opencl/headers (replace "/path/to/your/opencl/headers" with the actual path to the OpenCL headers that you just installed).
    • Press Enter.
  3. Install OpenCL SDK: In some cases, you might also need to install the OpenCL SDK. For NVIDIA GPUs, this is usually included with the CUDA toolkit. If you've already installed the CUDA toolkit, you should already have the OpenCL SDK installed. If not, you can download and install it from NVIDIA's official website.
  4. Rebuild your project: After completing the steps above, try rebuilding your project. The error should be resolved.

Please note that these steps are general guidelines. The specific commands you need to use may vary depending on your exact operating system and setup. If you need further help, let me know the specific details of your system, and I will provide more precise instructions.

bolabola commented 3 months ago

try https://github.com/bolabola/profanity2