[issue35537] use os.posix_spawn in subprocess

[STINNER Victor]

STINNER Victor <vstinner at redhat.com> added the comment:

Serhiy Storchaka, Alexey Izbyshev: I read and understood your valid concerns.

restore_signals=True will be quickly implemented, and so posix_spawn() code path will be tested by more tests of test_subprocess. If not, I will ensure that missing tests will be added.

Enhance _posixsubprocess to use vfork() is an interesting project, but IMHO it's complementary and doesn't replace all advantages of posix_spawn().

I am going to merge the PR tomorrow, except if someone sees a good reason to not merge it. I prefer to merge the PR early in the 3.8 development cycle to have more time to handle any issue if someone notice bugs. If something goes badly, we will be able to easily revert the change. Don't worry, I like to revert changes ;-)

Again, I'm mentoring Joannah who is learning Python, so I prefer to move step by step (with small steps :-)). We will support more and more subprocess.Popen options.

----------

_______________________________________
Python tracker <report at bugs.python.org>
<https://bugs.python.org/issue35537>
_______________________________________