The issue number is wrong.

Thank you @ZackerySpytz, I updated it.

Not quite. Please also fix the commit message and the link in the first post.

A Python core developer has requested some changes be made to your pull request before we can consider merging it. If you could please address their requests along with any other requests in other reviews from core developers that would be appreciated.

Once you have made the requested changes, please leave a comment on this pull request containing the phrase I have made the requested changes; please review again. I will then notify any core developers who have left a review that you're ready for them to take another look at this pull request.

@vstinner any ideas to fix it?

@vstinner any ideas to fix it?

Sorry, I have no idea.

I don't know how to fix it. As Nathaniel J. Smith said:

you might be thinking of PIPE_BUF, which POSIX says has to be >= 512 bytes, and on linux is 4096 bytes. This a magic number where if you send < PIPE_BUF bytes in a single write, then the kernel isn't allowed to break it up into pieces. This means that if you have a non-blocking pipe, then select can only report it as being writable if it as at least PIPE_BUF bytes of free space available.

Some new ideas are needed.

As said here by Victor and in b.p.o by Davin, there may be a fix, so we better work on fixing it than documenting the bug.