On 10/23/25 18:47, Colin Watson wrote:
On Thu, Oct 23, 2025 at 04:57:23PM +0100, Colin Watson wrote:
On Sun, Sep 07, 2025 at 02:47:53PM +0200, Matthias Klose wrote:
8696s
======================================================================
8696s ERROR: test_many_expressions
(test_fstring.TestCase.test_many_expressions)
8696s
----------------------------------------------------------------------
8696s Traceback (most recent call last):
8696s File "/usr/lib/python3/dist-packages/setuptools/_distutils/
spawn.py", line 87, in spawn
8696s subprocess.check_call(cmd, env=_inject_macos_ver(env))
8696s ~~~~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
8696s File "/usr/lib/python3.13/subprocess.py", line 419, in
check_call
8696s raise CalledProcessError(retcode, cmd)
8696s subprocess.CalledProcessError: Command '['/usr/bin/i686-linux-
gnu-gcc', '-fno-strict-overflow', '-Wsign-compare', '-DNDEBUG', '-g',
'-O2', '-Wall', '-fPIC', '-I/usr/include/python3.13', '-c', '/home/
debci/.cython/inline/
_cython_inline_b81c8cc73be357772ff34f66b468dda67d89a335.c', '-o', '/
home/debci/.cython/inline/home/debci/.cython/inline/
_cython_inline_b81c8cc73be357772ff34f66b468dda67d89a335.o']' returned
non-zero exit status 1.
I can reproduce this with 3.0.11 but not with upstream master, so I'm
in the process of (slowly) bisecting it to narrow things down a bit.
Good news: it bisects cleanly.
Bad news: the change that fixed it is https://github.com/cython/cython/
pull/5830, which doesn't look terribly backportable. (Presumably it
tips it back under some kind of memory threshold.)
Stefano/Julian, what do you think about bumping to 3.1.x at this point?
I'm happy to do the work for that, but I'd like to check that I'm not
stepping in an obvious bear-trap.
I did a few upstream updates in the past. Afaicr, I didn't see too many
issues.
Matthias
