!buildbot AMD64 FreeBSD14

🤖 New build scheduled with the buildbot fleet by @vstinner for commit f41875a 🤖

The command will test the builders whose names match following regular expression: AMD64 FreeBSD14

The builders matched are:

• AMD64 FreeBSD14 PR

!buildbot ARM64 MacOS M1 NoGIL

🤖 New build scheduled with the buildbot fleet by @vstinner for commit f41875a 🤖

The command will test the builders whose names match following regular expression: ARM64 MacOS M1 NoGIL

The builders matched are:

• ARM64 MacOS M1 NoGIL PR

!buildbot s390x Fedora PR

🤖 New build scheduled with the buildbot fleet by @vstinner for commit f41875a 🤖

The command will test the builders whose names match following regular expression: s390x Fedora PR

The builders matched are:

• s390x Fedora PR

Thanks @skirpichev for the PR, and @vstinner for merging it 🌮🎉.. I'm working now to backport this PR to: 3.13.
🐍🍒⛏🤖

GH-121449 is a backport of this pull request to the 3.13 branch.

I still see this failure on my Mac. I changed the failing test to skip it if readline.backend == "editline". I'm not sure that's the right change, but it would seem that editline and the real readline must differ in some way related to this.

diff --git a/Lib/test/test_pyrepl/test_pyrepl.py b/Lib/test/test_pyrepl/test_pyrepl.py
index 015b6905662..0f055edfcc8 100644
--- a/Lib/test/test_pyrepl/test_pyrepl.py
+++ b/Lib/test/test_pyrepl/test_pyrepl.py
@@ -904,8 +904,10 @@ def test_python_basic_repl(self):
 
     def test_not_wiping_history_file(self):
         # skip, if readline module is not available
-        import_module('readline')
+        readline = import_module('readline')
 
+        if readline.backend == "editline":
+            self.skipTest("editline's history mgmt isn't 100% compatible w/readline")
         hfile = tempfile.NamedTemporaryFile(delete=False)
         self.addCleanup(unlink, hfile.name)
         env = os.environ.copy()

If you think that is the right thing to do, I will create a PR, but I'm kind of busy with some other stuff right now. Might take a couple days to get around to it.

I still see this failure on my Mac.

Here is the recent test on main: https://buildbot.python.org/#/builders/1270/builds/2021 So, I doubt it's a good idea just disable this on Mac.

@smontanaro, could you, please, test original issue #121245? I worry it might be valid on your system.

Is it possible that you have some PYTHON* variables defined?

Rebuilding the readline module to use the readline 8.2.10 library in /opt/homebrew, I still get the same failure.

Note, however, that if I unset PYTHONSTARTUP, with the current code, the failing test passes. I believe my personal readline setup is interfering. @vstinner modified another readline test to run in isolated mode (see #121414). It would seem the same change needs to be applied here.

Note, however, that if I unset PYTHONSTARTUP, with the current code, the failing test passes.

See #121672 in this case.