Crispy13

I have a similar issue, and solved it by adding proper channels of the packages which i need. '-c conda-forge' ``` conda create -n snubh python tensorflow=2.0.0 keras matplotlib opencv...

I had the same issue.

It crashed too when not using `do_something`. (I just returned dummy output) I found that it did not crash when I don't use rayon, just using iterator. And I modified...

Could you try this? I wrote a minimal reproducible code newly. [https://gist.github.com/Crispy13/320129862c5fe80740a14dcea50519fd](https://gist.github.com/Crispy13/320129862c5fe80740a14dcea50519fd) The error does not occur when using global thread pool instead of local one. ```Toml [dependencies] anyhow =...

Would you try this python code? [https://gist.github.com/Crispy13/c3acaef979ed93567f5e14e1b34add47](https://gist.github.com/Crispy13/c3acaef979ed93567f5e14e1b34add47) I found somethings: - When not using python multiprocessing code, the error does not occur. - When not using `gc.collect()`, the error does...

@adamreichold It seems that segmentation fault does not occur anymore. I'll leave a comment If I find something

Sadly, In my case, this didn't solve the error. Python 3.9.15 If I return PyBytes instead of PyString, the error seems to go away. \* edit In the code I...

Any update?

In OutStream class, ``` def _setup_stream_redirects(self, name): pr, pw = os.pipe() fno = getattr(sys, name).fileno() self._original_stdstream_copy = os.dup(fno) os.dup2(pw, fno) self._fid = pr self._exc = None self.watch_fd_thread = threading.Thread(target=self._watch_pipe_fd) self.watch_fd_thread.daemon...

It's right. But I just want to occupy maximum N threads of clusters. When I set `--jobs` to 10: - If 10 jobs using 10 threads are running on the...