[3.12] gh-105699: Re-enable the Multiple-Interpreters Stress Tests (gh-107572)

[ericsnowcurrently]

We had disabled them due to crashes they exposed, which have since been fixed.
(cherry picked from commit f9e3ff1)

(This is a repeat of gh-107783, which we had to revert.)

• Issue: Crash During Subinterpreter Finalization #105699

[bedevere-bot]

🤖 New build scheduled with the buildbot fleet by @ericsnowcurrently for commit b19e0ef 🤖

If you want to schedule another build, you need to add the 🔨 test-with-buildbots label again.

[ericsnowcurrently]

Looks like the various other backports didn't fix all the issues, though test_create_many_sequential seems to be passing consistently now. test_create_many_threaded is still crashing and we're still leaking references, but neither in CI. Notably, this hasn't been a problem on main. I guess there is at least one PR that we haven't backported that is making the difference.

buildbot: AMD64 Fedora Stable PR

https://buildbot.python.org/all/#/builders/52/builds/1218/steps/6/logs/stdio

test_create_many_threaded (test.test_interpreters.StressTests.test_create_many_threaded) ... mremap_chunk(): invalid pointer Fatal Python error: Aborted Current thread 0x00007f64956fc6c0 (most recent call first): File "<frozen importlib._bootstrap_external>", line 1186 in get_data File "<frozen importlib._bootstrap_external>", line 1086 in get_code File "<frozen importlib._bootstrap_external>", line 990 in exec_module File "<frozen importlib._bootstrap>", line 935 in _load_unlocked File "<frozen importlib._bootstrap>", line 1331 in _find_and_load_unlocked File "<frozen importlib._bootstrap>", line 1360 in _find_and_load Extension modules: _xxsubinterpreters, _xxinterpchannels (total: 2) 1 test failed again: test_interpreters 

buildbot: AMD64 FreeBSD14 PR

https://buildbot.python.org/all/#/builders/1255/builds/34/steps/5/logs/stdio

test_create_many_threaded (test.test_interpreters.StressTests.test_create_many_threaded) ... Debug memory block at address p=0x358a5428fc10: API 'r' 1048 bytes originally requested The 7 pad bytes at p-7 are FORBIDDENBYTE, as expected. The 8 pad bytes at tail=0x358a54290028 are not all FORBIDDENBYTE (0xfd): at tail+0: 0x20 *** OUCH at tail+1: 0x06 *** OUCH at tail+2: 0x29 *** OUCH at tail+3: 0x54 *** OUCH at tail+4: 0x8a *** OUCH at tail+5: 0x35 *** OUCH at tail+6: 0x00 *** OUCH at tail+7: 0x00 *** OUCH Data at p: dd dd dd dd dd dd dd dd ... a8 90 7e 00 00 00 00 00 Enable tracemalloc to get the memory block allocation traceback Fatal Python error: _PyMem_DebugRawRealloc: bad trailing pad byte Python runtime state: initialized Current thread 0x0000358a4db8fd00 (most recent call first): File "<frozen importlib._bootstrap_external>", line 752 in _compile_bytecode File "<frozen importlib._bootstrap_external>", line 1123 in get_code File "<frozen importlib._bootstrap_external>", line 990 in exec_module File "<frozen importlib._bootstrap>", line 935 in _load_unlocked File "<frozen importlib._bootstrap>", line 1331 in _find_and_load_unlocked File "<frozen importlib._bootstrap>", line 1360 in _find_and_load File "/home/buildbot/buildarea/pull_request.opsec-fbsd14/build/Lib/site.py", line 73 in <module> File "<frozen importlib._bootstrap>", line 488 in _call_with_frames_removed File "<frozen importlib._bootstrap_external>", line 994 in exec_module File "<frozen importlib._bootstrap>", line 935 in _load_unlocked File "<frozen importlib._bootstrap>", line 1331 in _find_and_load_unlocked File "<frozen importlib._bootstrap>", line 1360 in _find_and_load Debug memory block at address p=0x358a3569aeb0: API '\xfd' 8247625154782626045 bytes originally requested The 7 pad bytes at p-7 are not all FORBIDDENBYTE (0xfd): at p-7: 0xfd at p-6: 0xfd at p-5: 0xfd at p-4: 0xfd at p-3: 0xfd at p-2: 0xfd at p-1: 0xdd *** OUCH Because memory is corrupted at the start, the count of bytes requested may be bogus, and checking the trailing pad bytes may segfault. The 8 pad bytes at tail=0x7275a8eb9cceafad are Fatal Python error: Bus error Current thread 0x0000358a4db8fd00 (most recent call first): File "<frozen importlib._bootstrap_external>", line 752 in _compile_bytecode File "<frozen importlib._bootstrap_external>", line 1123 in get_code File "<frozen importlib._bootstrap_external>", line 990 in exec_module File "<frozen importlib._bootstrap>", line 935 in _load_unlocked File "<frozen importlib._bootstrap>", line 1331 in _find_and_load_unlocked File "<frozen importlib._bootstrap>", line 1360 in _find_and_load File "/home/buildbot/buildarea/pull_request.opsec-fbsd14/build/Lib/site.py", line 73 in <module> File "<frozen importlib._bootstrap>", line 488 in _call_with_frames_removed File "<frozen importlib._bootstrap_external>", line 994 in exec_module File "<frozen importlib._bootstrap>", line 935 in _load_unlocked File "<frozen importlib._bootstrap>", line 1331 in _find_and_load_unlocked File "<frozen importlib._bootstrap>", line 1360 in _find_and_load Extension modules: _xxsubinterpreters, _xxinterpchannels (total: 2) 1 test failed again: test_interpreters 

buildbot: AMD64 RHEL7 Refleaks PR

https://buildbot.python.org/all/#/builders/18/builds/1511/steps/5/logs/stdio

test_interpreters leaked [14, 14, 14] file descriptors, sum=42 1 test failed again: test_interpreters 

[ericsnowcurrently]

I may bail out on re-enabling the stress tests on 3.12. They are more important on main.

[ericsnowcurrently]

I suspect that at least gh-110709 would help, though I'm not sure if we'll backport that or not.

[hugovk]

Triage: is this backport still needed?

[ericsnowcurrently]

I'm going to say no. :)