test_embed failures after switching configure options due to missing Makefile dependencies

[smontanaro]

Bug report

Bug description:

test_embed started failing for me recently on free threaded builds on both M1 Mac and Ubuntu (Intel). git bisect narrowed it down to a suspect commit. @colesbury this is probably for you.

Unfortunately, on my Mac I get this commit as the culprit:

% git bisect bad
5f1997896d9c3ecf92e9863177c452b468a6a2c8 is the first bad commit
commit 5f1997896d9c3ecf92e9863177c452b468a6a2c8
Author: Sam Gross <[email protected]>
Date:   Tue Jan 23 13:05:15 2024 -0500

    gh-112984: Fix link error on free-threaded Windows build (GH-114455)
    
    The test_peg_generator test tried to link the python313_d.lib library,
    which failed because the library is now named python313t_d.lib. The
    underlying problem is that the "compiler" attribute was not set when
    we call get_libraries() from distutils.

 Tools/peg_generator/pegen/build.py | 3 +++
 1 file changed, 3 insertions(+)

while in my Linux environment it says this:

% git bisect bad
b331381485c1965d1c88b7aee7ae9604aca05758 is the first bad commit
commit b331381485c1965d1c88b7aee7ae9604aca05758
Author: Sam Gross <[email protected]>
Date:   Tue Jan 16 16:42:15 2024 -0500

    gh-112529: Track if debug allocator is used as underlying allocator (#113747)
    
    * gh-112529: Track if debug allocator is used as underlying allocator
    
    The GC implementation for free-threaded builds will need to accurately
    detect if the debug allocator is used because it affects the offset of
    the Python object from the beginning of the memory allocation. The
    current implementation of `_PyMem_DebugEnabled` only considers if the
    debug allocator is the outer-most allocator; it doesn't handle the case
    of "hooks" like tracemalloc being used on top of the debug allocator.
    
    This change enables more accurate detection of the debug allocator by
    tracking when debug hooks are enabled.
    
    * Simplify _PyMem_DebugEnabled

 Include/internal/pycore_pymem.h        |  3 +++
 Include/internal/pycore_pymem_init.h   |  2 ++
 Include/internal/pycore_runtime_init.h |  1 +
 Objects/obmalloc.c                     | 21 +++++++++++++++------
 4 files changed, 21 insertions(+), 6 deletions(-)

(I backed up farther on my Linux box to find a good commit.) 16 January was a long time ago. My inclination is that the later commit is the culprit.

I then tried with the GIL enabled on main. Still an error. Hmmm... I went through the bisect process again on my Mac and landed on this as the putative culprit commit:

% git bisect good
441affc9e7f419ef0b68f734505fa2f79fe653c7 is the first bad commit
commit 441affc9e7f419ef0b68f734505fa2f79fe653c7
Author: Sam Gross <[email protected]>
Date:   Tue Jan 23 13:08:23 2024 -0500

    gh-111964: Implement stop-the-world pauses (gh-112471)
    
    The `--disable-gil` builds occasionally need to pause all but one thread.  Some
    examples include:
    
    * Cyclic garbage collection, where this is often called a "stop the world event"
    * Before calling `fork()`, to ensure a consistent state for internal data structures
    * During interpreter shutdown, to ensure that daemon threads aren't accessing Python objects
    
    This adds the following functions to implement global and per-interpreter pauses:
    
    * `_PyEval_StopTheWorldAll()` and `_PyEval_StartTheWorldAll()` (for the global runtime)
    * `_PyEval_StopTheWorld()` and `_PyEval_StartTheWorld()` (per-interpreter)
    
    (The function names may change.)
    
    These functions are no-ops outside of the `--disable-gil` build.

 Include/cpython/pystate.h              |   2 +-
 Include/internal/pycore_ceval.h        |   1 +
 Include/internal/pycore_interp.h       |  17 +++
 Include/internal/pycore_llist.h        |   3 +-
 Include/internal/pycore_pystate.h      |  51 +++++--
 Include/internal/pycore_runtime.h      |   7 +
 Include/internal/pycore_runtime_init.h |   3 +
 Include/pymacro.h                      |   3 +
 Python/ceval_gil.c                     |   9 ++
 Python/pystate.c                       | 269 +++++++++++++++++++++++++++++++--
 10 files changed, 336 insertions(+), 29 deletions(-)

CPython versions tested on:

CPython main branch

Operating systems tested on:

Linux, macOS

Linked PRs

• gh-114505: Add missing header file dependencies #114513

[smontanaro]

I should add that I originally tried this with a --disable-gil build. I later reproduced it on my Mac with the GIL enabled.

[colesbury]

Hi Skip - I tried running test_embed on an M1 Mac (built from d22c066) and haven't seen any errors either with --disable-gil or with the default build.

What is the error you are seeing? Have you tried running make clean?

[smontanaro]

Hrm... I guess it's a make dependency problem. After a make clean I'm not able to reproduce it either. Sorry for the noise.

FWIW, here's the error I was seeing:

======================================================================
FAIL: test_init_isolated_config (test.test_embed.InitConfigTests.test_init_isolated_config)
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/Users/skip/src/python/cpython/Lib/test/test_embed.py", line 1082, in test_init_isolated_config
    self.check_all_configs("test_init_isolated_config", api=API_ISOLATED)
    ~~~~~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/Users/skip/src/python/cpython/Lib/test/test_embed.py", line 800, in check_all_configs
    out, err = self.run_embedded_interpreter(testname,
               ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^^^^^^^^^^
                                             env=env, cwd=cwd)
                                             ^^^^^^^^^^^^^^^^^
  File "/Users/skip/src/python/cpython/Lib/test/test_embed.py", line 119, in run_embedded_interpreter
    self.assertEqual(p.returncode, returncode,
    ~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^
                     "bad returncode %d, stderr is %r" %
                     ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                     (p.returncode, err))
                     ^^^^^^^^^^^^^^^^^^^^
AssertionError: -6 != 0 : bad returncode -6, stderr is 'Assertion failed: (rt_preconfig->isolated == 1), function check_preinit_isolated_config, file _testembed.c, line 1065.\n'

======================================================================
FAIL: test_preinit_isolated_config (test.test_embed.InitConfigTests.test_preinit_isolated_config)
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/Users/skip/src/python/cpython/Lib/test/test_embed.py", line 1079, in test_preinit_isolated_config
    self.check_all_configs("test_preinit_isolated_config", api=API_ISOLATED)
    ~~~~~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/Users/skip/src/python/cpython/Lib/test/test_embed.py", line 800, in check_all_configs
    out, err = self.run_embedded_interpreter(testname,
               ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^^^^^^^^^^
                                             env=env, cwd=cwd)
                                             ^^^^^^^^^^^^^^^^^
  File "/Users/skip/src/python/cpython/Lib/test/test_embed.py", line 119, in run_embedded_interpreter
    self.assertEqual(p.returncode, returncode,
    ~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^
                     "bad returncode %d, stderr is %r" %
                     ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                     (p.returncode, err))
                     ^^^^^^^^^^^^^^^^^^^^
AssertionError: -6 != 0 : bad returncode -6, stderr is 'Assertion failed: (rt_preconfig->isolated == 1), function check_preinit_isolated_config, file _testembed.c, line 1054.\n'

----------------------------------------------------------------------
Ran 69 tests in 3.743s

FAILED (failures=2, skipped=5)
test test_embed failed

== Tests result: FAILURE ==

10 slowest tests:
- test_embed: 3.8 sec

1 test failed:
    test_embed

0:00:03 load avg: 20.85 Re-running 1 failed tests in verbose mode in subprocesses
0:00:03 load avg: 20.85 Run 1 test in parallel using 1 worker process
0:00:03 load avg: 20.85 [1/1/1] test_embed failed (2 failures)
...

That -6 != 0 error was seen on both Linux and MacOS.

[erlend-aasland]

Have you tried running make clean?

I would recommend git clean -fdx between each build when bisecting; trust only a completely clean build.

[colesbury]

Yeah, I think we don't have all the proper dependencies in the Makefile for _testembed.o:

cpython/Makefile.pre.in

Lines 1401 to 1402 in afe8f37

	Programs/_testembed.o: $(srcdir)/Programs/_testembed.c Programs/test_frozenmain.h
	$(CC) -c $(PY_CORE_CFLAGS) -o $@ $(srcdir)/Programs/_testembed.c

We are also missing a bunch of header dependencies in the definition of PYTHON_HEADERS.

[smontanaro]

Can you change the title to reflect this deficiency and assign it to me? I'll see if I can improve that aspect of the make environment.

[colesbury]

Thanks Skip - I updated the title. I think you should be able to do that too -- there's an "Edit" button next to the tile and "New Issue" on this page.

[erlend-aasland]

Can you change the title to reflect this deficiency and assign it to me? I'll see if I can improve that aspect of the make environment.

That would be a very welcome improvement :)

[smontanaro]

I guess the bigger issue was how you assigned it to me. I didn't see an obvious way to assign it to myself.