bpo-26228: pty.spawn hangs on FreeBSD, OS X, and Solaris by diekmann · Pull Request #4167 · python/cpython (original) (raw)
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service andprivacy statement. We’ll occasionally send you account related emails.
Already on GitHub?Sign in to your account
Conversation12 Commits7 Checks0 Files changed
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.Learn more about bidirectional Unicode characters
[ Show hidden characters]({{ revealButtonHref }})
issue26228 as github PR.
This PR contains:
- original patch of issue26228 by Chris
- Update of docs
- Update of pty test suite
According to the bpo discussion, this fixes pty.spawn() on FreeBSD, OS X, and Solaris.
https://bugs.python.org/issue26228
Bpo 26228: pty.spawn hangs on FreeBSD, OS X, and Solaris Bpo-26228: pty.spawn hangs on FreeBSD, OS X, and Solaris
Bpo-26228: pty.spawn hangs on FreeBSD, OS X, and Solaris bpo-26228: pty.spawn hangs on FreeBSD, OS X, and Solaris
| # If we wanted to be really clever, we would use |
|---|
| # the same method as subprocess() to pass the error |
| # back to the parent. For now just dump stack trace. |
| traceback.print_exc() |
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Thanks for catching this ... again 🎃 👻
Thank you so much for your feedback! 🥇
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I can’t see anything seriously wrong with 005dc23, although I think it would be better to test the API (pty.spawn), without depending on the internals (_ copy, select, etc) if possible.
| return rfds_result, [], [] |
|---|
| def test__mock_stdin_stdout(self): |
| self.assertGreater(pty.STDIN_FILENO, 2, "replaced by our mock") |
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Why is the method called stdin_stdout, while the body only looks at STDIN_FILENO?
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Fixed in the following commit: A docstring explains the name and the method also checks for STDOUT_FILENO.
| # called with three empty lists as file descriptors to wait |
|---|
| # on. Behavior of real select is platform-dependent and |
| # likely infinite blocking on Linux. |
| raise self.fail("mock select on no waitables") |
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I suspect the raise statement never executes because TestCase.fail raises its own exception.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Your suspicion is appropriate. Fixed in the following commit.
| # Test that STDIN was not touched. This test simulated the |
| # scenario where the child process immediately closed its end of |
| # the pipe. This means, nothing should be copied. |
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
What is the point of this test? I don’t think it matters.
You seem to be modelling a situation where input (or EOF) to the child becomes available at about the same instant that the slave terminal is closed. If _ copy were written differently, it might copy the input to the child (STDIN_FILENO) before checking the pseudoterminal master (master_fd), and the test would fail.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
True, this part of the test is overly fitted for an unnecessary implementation detail. Removed in the following commit
- Linux -> Unix
- More details about the behavior of spawn
ambv added a commit to ambv/cpython that referenced this pull request
Co-authored-by: Cornelius Diekmann c.diekmann@googlemail.com
The doc part of this was included in GH-27754.
miss-islington pushed a commit to miss-islington/cpython that referenced this pull request
Co-authored-by: Cornelius Diekmann c.diekmann@googlemail.com (cherry picked from commit dd8eb30)
Co-authored-by: Łukasz Langa lukasz@langa.pl
ambv added a commit that referenced this pull request
Co-authored-by: Cornelius Diekmann c.diekmann@googlemail.com
miss-islington added a commit that referenced this pull request
Co-authored-by: Cornelius Diekmann c.diekmann@googlemail.com (cherry picked from commit dd8eb30)
Co-authored-by: Łukasz Langa lukasz@langa.pl