Steve Stagg <stest...@gmail.com> added the comment:
Reproducible on master, but doesn't occur with a debug build.
My configure:
> ./configure --prefix=$PREFIX --exec-prefix=$EPREFIX
> --cache-file=../config.cache --without-ensurepip > /dev/null
> PYTHONMALLOC=pymalloc_debug $EPREFIX/bin/python3 ../test.py
Crash comes out of the child process, I modified example to invoke lldb for
child:
```
import os
import time
import subprocess
def daemon():
pid = os.fork()
if pid != 0:
subprocess.run(["lldb", '-p', str(pid)])
time.sleep(10)
os._exit(0)
time.sleep(1)
daemon()
```
Attaching to the child gives this backtrace:
(lldb) bt
* thread #1, name = 'python3', stop reason = signal SIGSEGV: invalid address
(fault address: 0x0)
frame #0: 0x0000564a5a325f9c
python3.10`_PyObject_DebugDumpAddress(p=0x0000564a5aa2c630) at
obmalloc.c:2457:12
frame #1: 0x0000564a5a326571
python3.10`_PyMem_DebugCheckAddress(func="_PyMem_DebugRawFree",
api=<unavailable>, p=<unavailable>) at obmalloc.c:2375:9
* frame #2: 0x0000564a5a326d29 python3.10`_PyMem_DebugRawFree at
obmalloc.c:2208:5
frame #3: 0x0000564a5a326d17
python3.10`_PyMem_DebugRawFree(ctx=0x0000564a5a5d37e0, p=0x0000564a5aa2c630) at
obmalloc.c:2197
frame #4: 0x0000564a5a3d4aa3
python3.10`PyInterpreterState_Delete(interp=0x0000564a5aa4e6b0) at
pystate.c:396:9
frame #5: 0x0000564a5a3d1921 python3.10`Py_FinalizeEx [inlined]
finalize_interp_delete(tstate=<unavailable>) at pylifecycle.c:1633:5
frame #6: 0x0000564a5a3d190b python3.10`Py_FinalizeEx at pylifecycle.c:1782
frame #7: 0x0000564a5a2dab80 python3.10`Py_BytesMain at main.c:679:9
frame #8: 0x0000564a5a2dab69 python3.10`Py_BytesMain at main.c:707
frame #9: 0x0000564a5a2dab4f python3.10`Py_BytesMain(argc=<unavailable>,
argv=<unavailable>) at main.c:731
frame #10: 0x00007ff2e6251152 libc.so.6`__libc_start_main + 242
frame #11: 0x0000564a5a2d978e python3.10`_start + 46
The call stack is trying to free an interpreter mutex:
PyThread_free_lock(interp->id_mutex);
----------
nosy: +stestagg
_______________________________________
Python tracker <rep...@bugs.python.org>
<https://bugs.python.org/issue42540>
_______________________________________
_______________________________________________
Python-bugs-list mailing list
Unsubscribe:
https://mail.python.org/mailman/options/python-bugs-list/archive%40mail-archive.com
