Description

assignee = None
closed_at = None
created_at = <Date 2022-01-13.08:18:32.782>
labels = ['type-bug', 'expert-asyncio']
title = 'asyncio subprocess cannot read from /dev/stdin'
updated_at = <Date 2022-02-01.18:11:58.688>
user = 'https://github.com/xopham'

activity = <Date 2022-02-01.18:11:58.688>
actor = 'xoph'
assignee = 'none'
closed = False
closed_date = None
closer = None
components = ['asyncio']
creation = <Date 2022-01-13.08:18:32.782>
creator = 'xoph'
dependencies = []
files = []
hgrepos = []
issue_num = 46364
keywords = ['patch']
message_count = 3.0
messages = ['410466', '410467', '412297']
nosy_count = 3.0
nosy_names = ['asvetlov', 'yselivanov', 'xoph']
pr_nums = ['30596']
priority = 'normal'
resolution = None
stage = 'patch review'
status = 'open'
superseder = None
type = 'behavior'
url = 'https://bugs.python.org/issue46364'
versions = []

Activity

1. xopham

   added

   topic-asyncio

   type-bugAn unexpected behavior, bug, or errorAn unexpected behavior, bug, or error

   on Jan 13, 2022
2. xopham commented on Jan 13, 2022

   

   xophammannequin

   on Jan 13, 2022

   MannequinAuthor

   More actions

   asyncio.create_subprocess_exec and asyncio.create_subprocess_shell fail with "No such device or address" when called on a program that attempts to invoke /dev/stdin on Linux.
   This happens in contrast to the subprocess module and was discussed in the this thread: https://stackoverflow.com/questions/70624413/how-to-read-from-dev-stdin-with-asyncio-create-subprocess-exec

   It seems while subprocess uses a pipe, asyncio uses socket.socketpair() to communicate with the subprocess which in turn appears to fail for /dev/stdin
3. asvetlov commented on Jan 13, 2022

   

   asvetlov

   on Jan 13, 2022

   Contributor

   More actions

   True. Would you make a pull request that uses socketpair on AIX only?
   A test for desired behavior is required.
4. xopham commented on Feb 1, 2022

   

   xophammannequin

   on Feb 1, 2022

   MannequinAuthor

   More actions

   Created a PR: #30596

   It's ready for review. Happy about any feedback!
5. ezio-melotti

   transferred this issue fromon Apr 10, 2022
6. ezio-melotti

   added this to asyncioon Jun 20, 2022
7. ezio-melotti

   moved this to Todo in asyncioon Jul 17, 2022
8. tifv

   added a commit that references this issue on Jul 18, 2022

   pythongh-90522: add test

   6477c7e
9. xopham

   added a commit that references this issue on Aug 12, 2022

   pythongh-90522: add test

   ef62e67
10. xopham commented on Aug 19, 2022

    

    xopham

    on Aug 19, 2022

    ContributorAuthor

    More actions

    @1st1 @asvetlov the associated PR is ready, including tests. It's a small change. Would be happy to move this forward and resolve the issue 🙂

    Reacted by July Tikhonov and Oleg Iarygin
11. xopham

    added a commit that references this issue on Oct 13, 2022

    pythongh-90522: add test

    0ac3aed
12. kumaraditya303

    closed this as completedon Oct 13, 2022
13. Repository owner moved this from Todo to Done in asyncioon Oct 13, 2022

Issue actions