gh-82616: Add process_group support to subprocess.Popen#23930
Conversation
Adds a setpgid parameter to subprocess APIs to help us deprecate uses of preexec_fn.
ZackerySpytz
left a comment
There was a problem hiding this comment.
It is subprocess.Popen, not subprocess.POpen. Please fix the title and the news entry.
|
This PR is stale because it has been open for 30 days with no activity. |
Co-authored-by: Zackery Spytz <zspytz@gmail.com>
Co-authored-by: Serhiy Storchaka <storchaka@gmail.com>
Co-authored-by: Serhiy Storchaka <storchaka@gmail.com>
Co-authored-by: Serhiy Storchaka <storchaka@gmail.com>
as suggested by Serhiy on the issue.
|
The Address sanitizer failure blocking this makes no logical sense. I cannot reproduce it on my systems because an address sanitizer build doesn't even work at all. Update: Zach Ware pointed out that I also needed to |
|
(post-merge review ;-)) LGTM. It might help to add a build check (C11 static_assert()) that the pid_t type is unsigned. A macro like that can be used: |
Most recent one in python/cpython#23930. This makes me wish we could have had type evaluation functions.
Adds a
process_group=parameter to subprocess APIs to help us deprecateuses of
preexec_fn.https://bugs.python.org/issue38435