Pending review, this can be merged since the test_threading failure is not related. (It has failed elsewhere today, including on current main on my machine.) I will look at the patch.

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.

And if you don't make the requested changes, you will be put in the comfy chair!

Thanks @aisk for the PR, and @terryjreedy for merging it 🌮🎉.. I'm working now to backport this PR to: 3.11, 3.12.
🐍🍒⛏🤖

⚠️⚠️⚠️ Buildbot failure ⚠️⚠️⚠️

Hi! The buildbot AMD64 Windows10 3.x has failed when building commit bcb2ab5.

What do you need to do:

1. Don't panic.
2. Check the buildbot page in the devguide if you don't know what the buildbots are or how they work.
3. Go to the page of the buildbot that failed (https://buildbot.python.org/all/#builders/146/builds/6284) and take a look at the build logs.
4. Check if the failure is related to this commit (bcb2ab5) or if it is a false positive.
5. If the failure is related to this commit, please, reflect that on the issue and make a new Pull Request with a fix.

You can take a look at the buildbot page here:

https://buildbot.python.org/all/#builders/146/builds/6284

Failed tests:

• test_msvcrt

Failed subtests:

• test_kbhit - test.test_msvcrt.TestConsoleIO.test_kbhit
• test_getwch - test.test_msvcrt.TestConsoleIO.test_getwch

Summary of the results of the build (if available):

==

Traceback (most recent call last):
  File "D:\buildarea\3.x.bolen-windows10\build\Lib\test\test_msvcrt.py", line 65, in test_kbhit
    self.assertEqual(msvcrt.kbhit(), 0)
AssertionError: 1 != 0


Traceback (most recent call last):
  File "D:\buildarea\3.x.bolen-windows10\build\Lib\test\test_msvcrt.py", line 77, in test_getwch
    self.assertEqual(msvcrt.getwch(), c)
AssertionError: '\n' != '\u5b57'
+ \u5b57
- 
- 

This change broke "AMD64 Windows10 3.x" buildbot and the Windows x64 CI job on GitHub Action :-(

buildbot failure: https://buildbot.python.org/all/#builders/146/builds/6284

GHA failure: https://github.com/python/cpython/actions/runs/6127105160/job/16632346105?pr=109168

FAIL: test_getwch (test.test_msvcrt.TestConsoleIO.test_getwch)
----------------------------------------------------------------------
Traceback (most recent call last):
  File "D:\a\cpython\cpython\Lib\test\test_msvcrt.py", line 77, in test_getwch
    self.assertEqual(msvcrt.getwch(), c)
AssertionError: '\n' != '\u5b57'
+ \u5b57
- 

The GitHub Action job also started to log a new error:

ResourceUnavailable: D:\a\_temp\0ad13265-0ee9-4c5f-87fb-4d47fe5a3611.ps1:2
Line |
   2 |  .\PCbuild\rt.bat -p x64 -d -q -uall -u-cpu -rwW --slowest --timeout=1 �
     |  ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
     | The Win32 internal error "No process is on the other end of the pipe." 0xE9 occurred while setting the console window title. Contact Microsoft Customer Support Services.

Error: Process completed with exit code 1.

"AMD64 Windows10 3.x" buildbot failures:

======================================================================
FAIL: test_getwch (test.test_msvcrt.TestConsoleIO.test_getwch)
----------------------------------------------------------------------
Traceback (most recent call last):
  File "D:\buildarea\3.x.bolen-windows10\build\Lib\test\test_msvcrt.py", line 77, in test_getwch
    self.assertEqual(msvcrt.getwch(), c)
AssertionError: '\n' != '\u5b57'
+ \u5b57
- 
- 


======================================================================
FAIL: test_kbhit (test.test_msvcrt.TestConsoleIO.test_kbhit)
----------------------------------------------------------------------
Traceback (most recent call last):
  File "D:\buildarea\3.x.bolen-windows10\build\Lib\test\test_msvcrt.py", line 65, in test_kbhit
    self.assertEqual(msvcrt.kbhit(), 0)
AssertionError: 1 != 0

vstinner removed needs backport to 3.11 needs backport to 3.12 labels

I disabled backport, to avoid backporting a change which breaks the CI.

Ah, another strange thing in the GHA job:

FAILED (failures=1)
test test_msvcrt failed
0:16:43 load avg: 14.11 [327/463/2] test_devpoll process crashed (Exit code 3221225794) -- running: test_regrtest (2 min 58 sec), test_fstring (32.5 sec)

== Tests result: FAILURE ==

I don't know why test_devpoll is run on Windows, nor why it does crash :-( The test starts with:

if not hasattr(select, 'devpoll') :
    raise unittest.SkipTest('test works only on Solaris OS family')

But it's a bug in libregrtest which reports a test_msvcrt crash as a test_devpoll crash?

Backports are #109166 and #109167

Backports are #109166 and #109167

I close them to prevent merging them by mistake. I would prefer to see two broken CI being fixed yet.

The regression is preventing to merge any new PR in main, since the Windows x64 job now fails :-(

Funny, it passed above. Feel free to revert: I don't know how.

Funny, it passed above.

I don't know why it passed before!? It's surprising :-( At least, we know a buildbot which should be tested if a fix is proposed.

Feel free to revert: I don't know how.

I don't think that it would be productive to revert. Instead, I wrote PR #109169 to keep the test but always skip the test (on all platforms) for now, until someone is available to fix it.

Instead, I wrote PR #109169 to keep the test but always skip the test (on all platforms) for now, until someone is available to fix it.

Merged.

The strange part is that I re-ran Windows x64 job on my PR #109168 before PR #109169 was merged, and all tests pass Windows x64!

It seems like test_msvcrt is not reliable :-(

Similar failure on "AMD64 Windows11 Non-Debug 3.x" buildbot:

FAIL: test_getwch (test.test_msvcrt.TestConsoleIO.test_getwch)
----------------------------------------------------------------------
Traceback (most recent call last):
  File "b:\uildarea\3.x.ware-win11.nondebug\build\Lib\test\test_msvcrt.py", line 77, in test_getwch
    self.assertEqual(msvcrt.getwch(), c)
AssertionError: '\n' != '\u5b57'
+ \u5b57
- 
- 


======================================================================
FAIL: test_kbhit (test.test_msvcrt.TestConsoleIO.test_kbhit)
----------------------------------------------------------------------
Traceback (most recent call last):
  File "b:\uildarea\3.x.ware-win11.nondebug\build\Lib\test\test_msvcrt.py", line 65, in test_kbhit
    self.assertEqual(msvcrt.kbhit(), 0)
AssertionError: 1 != 0

logs: https://buildbot.python.org/all/#/builders/914/builds/2794

@terryjreedy: if you want, next week if can have a look at the failing tests. But when I ran the test manually on my Windows VM: oops, it just worked. I don't know which kind of "terminal" the buildbot and the GitHub Action Windows jobs use.

These days, I run tests in a SSH connection on a Linux terminal to my Windows VM. In this case, apparently, Python seems to be satisfied with the terminal that it gets.

A first step would be to only skip tests which are known to fail, and run the other tests which are fine.

For the test_kbhit, maybe some parallel tests may simulate keyboard hit, so I think maybe we should not check the result value, just test the function exists and not raise exceptions?

I am done with this as these errors are out of my league.

@zooba knows more and wrote some of the changes.

I figured out what happened:

The test_winconsoleio may have left over contents in the stdin, and if it runs before test_msvcrt, the call of msvcrt.getwch() in test_msvcrt.test_getwch may get a invalid \n result which is written in test_winconsoleio, also the test_kbhit.

One should paste the test case in test_winconsoleio:

I tried to call FlushConsoleInputBuffer via ctypes in #109226 to fix the error.

Tests still fails😂Looks like there are more issues...

Is it possible to create a "new terminal" just to run test_msvcrt?