-
Notifications
You must be signed in to change notification settings - Fork 242
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
How to run tests when distributing pyopencl? #784
Comments
The only thing that's used from compyte in pyopencl is So tl;dr: that stuff is safe to ignore and should really never have been included in the distribution. |
OK, so now, are there any tests for pyopencl that I would want to run when distributing pyopencl? |
Can you assume that an ICD is installed? |
I'm not sure what do you mean exactly.. Maybe I can. |
In OpenCL-speak, an ICD is something like pocl or a GPU driver. It's hard to test pyopencl without one. (because none of the OpenCL API calls do anything) |
OK I see. What command would you run to test pyopencl then? Assuming there is an ICD available. |
If you're willing to ship the |
OK - now I see that indeed the
And in all of these files:
Does that mean that ICD is not available in our build environment? |
Yes. |
https://portablecl.org/ is an ICD that doesn't need special hardware and is fairly straightforward to install. |
Then this issue can be marked as resolved, unless you'd want to keep it open for discussing about the |
Hello,
We are trying to run tests on NixOS'. However, these fail like this:
I tried to debug this a bit myself and I don't understand what's going on in the
compyte
Git submodule - there is a Makefile there from more then a decade ago. I tried tomake pygpu_ndarray.so
and I encounteredpython -c print "..."
errors. I also tried to override the Makefile variables so it won't rely upon Python, and I got this:I'm also surprised no body mentioned
pygpu_ndarray
in the issue tracker of Pyopencl and compyte at all.The text was updated successfully, but these errors were encountered: