python / python/cpython

Assertion failure: `!_PyFrame_IsIncomplete(frame->f_frame)`

Open
#149,110 0 comments 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

interpreter-core topic-free-threading type-crash
Dominant language
Python
Stars
77.2k
Forks
35.9k
PR merge metrics
PR metrics pending

Description

Crash report

What happened?

It's possible to cause an abort on a free-threaded build even with PYTHON_GIL=1 by running the code below (only reproduces with free threading enabled builds).

I attempted to replicate the trigger without pdb (using sys.monitoring directly, frame trace setters, sys.settrace, Bdb directly) without success. pdb's specific call chain seems necessary to reliably trigger it.

MRE:

from threading import Thread
import sys, pdb
sys.stdin = open("/dev/null")

# ONE shared Pdb instance so threads call set_trace on the SAME one
shared_pdb = pdb.Pdb(mode='inline', backend='monitoring', colorize=False)

def thread_target():
    shared_pdb.set_trace(sys._getframe().f_back)

def set_trace():
    return sys._getframe().f_back

for x in range(5):
    threads = []
    for x in range(50):
        threads.append(Thread(target=thread_target))
        threads.append(Thread(target=set_trace))
        threads.append(Thread(target=sys._clear_internal_caches))

    for t in threads:
        t.start()
        sys._clear_internal_caches()
    for t in threads:
        t.join()

print("NO ABORT")

Backtrace:

python: Objects/frameobject.c:2400: PyFrameObject *PyFrame_GetBack(PyFrameObject *): Assertion `!_PyFrame_IsIncomplete(frame->f_frame)' failed.

Thread 23 "Thread-22 (thre" received signal SIGABRT, Aborted.

#0  __pthread_kill_implementation (threadid=<optimized out>, signo=6, no_tid=0) at ./nptl/pthread_kill.c:44
#1  __pthread_kill_internal (threadid=<optimized out>, signo=6) at ./nptl/pthread_kill.c:89
#2  __GI___pthread_kill (threadid=<optimized out>, signo=signo@entry=6) at ./nptl/pthread_kill.c:100
#3  0x00007ffff7c45e2e in __GI_raise (sig=sig@entry=6) at ../sysdeps/posix/raise.c:26
#4  0x00007ffff7c28888 in __GI_abort () at ./stdlib/abort.c:77
#5  0x00007ffff7c287f0 in __assert_fail_base (fmt=<optimized out>, assertion=<optimized out>, file=<optimized out>, line=<optimized out>, function=<optimized out>) at ./assert/assert.c:118
#6  0x00007ffff7c3c19f in __assert_fail (assertion=<optimized out>, file=<optimized out>, line=<optimized out>, function=<optimized out>) at ./assert/assert.c:127
#7  0x0000555555b50d29 in PyFrame_GetBack (frame=frame@entry=0x7bffcc160410) at Objects/frameobject.c:2400
#8  0x0000555555b58c76 in frame_back_get_impl (self=0x7bffcc160410) at Objects/frameobject.c:1119
#9  frame_back_get (self=<optimized out>, _unused_context=<optimized out>) at Objects/clinic/frameobject.c.h:212
#10 0x0000555555c2cfb4 in _PyObject_GenericGetAttrWithDict (obj=0x7bffcc160410, name=0x7bffb60c0610, dict=0x0, suppress=0) at Objects/object.c:1924
#11 0x0000555555c2cab3 in _PyObject_GetAttrStackRef (v=0x7bffcc160410, name=0x7bffb60c0610) at Objects/object.c:1367
#12 0x0000555555edc322 in _PyEval_EvalFrameDefault (tstate=<optimized out>, frame=<optimized out>, throwflag=<optimized out>) at Python/generated_cases.c.h:8041
#13 0x0000555555ea8d04 in _PyEval_EvalFrame (tstate=0x7bffb4249010, frame=0x7e8ff709e328, throwflag=0) at ./Include/internal/pycore_ceval.h:118
#14 _PyEval_Vector (tstate=0x7bffb4249010, func=0x7bffb64cb1b0, locals=0x0, args=0x7bff9d4b9720, argcount=1, kwnames=0x0) at Python/ceval.c:2179
#15 0x0000555555ac6030 in _PyObject_VectorcallTstate (tstate=tstate@entry=0x7bffb4249010, callable=callable@entry=0x7bffb64cb1b0, args=0x7bff9d4b9720, nargsf=nargsf@entry=1,
    kwnames=kwnames@entry=0x0) at ./Include/internal/pycore_call.h:144
#16 0x0000555555acbf7f in _PyObject_VectorcallPrepend (tstate=<optimized out>, callable=<optimized out>, arg=<error reading variable: Cannot access memory at address 0x6>,
    args=<optimized out>, nargsf=0, kwnames=<optimized out>) at Objects/call.c:855
#17 0x0000555555f6b113 in _PyObject_VectorcallTstate (tstate=0x7bffb4249010, callable=0x7bffc82200d0, args=0x7bff9a2ab6c0, nargsf=22, kwnames=0x0) at ./Include/internal/pycore_call.h:144
#18 context_run (self=0x7bffb757cc10, args=args@entry=0x7bff9a2a96b0, nargs=nargs@entry=1, kwnames=kwnames@entry=0x0) at Python/context.c:727
#19 0x0000555555af812d in method_vectorcall_FASTCALL_KEYWORDS (func=func@entry=0x7bffb60a8820, args=args@entry=0x7bff9a2a96a8, nargsf=nargsf@entry=9223372036854775810,
    kwnames=kwnames@entry=0x0) at Objects/descrobject.c:421
#20 0x0000555555ac6030 in _PyObject_VectorcallTstate (tstate=0x7bffb4249010, callable=0x7bffb60a8820, args=0x7bff9a2a96a8, nargsf=9223372036854775810, kwnames=0x0)
    at ./Include/internal/pycore_call.h:144
#21 0x0000555555ea9e3b in _Py_VectorCallInstrumentation_StackRefSteal (callable=..., arguments=arguments@entry=0x7e8ff709e2f8, total_args=total_args@entry=2, kwnames=kwnames@entry=...,
    call_instrumentation=false, frame=frame@entry=0x7e8ff709e298, this_instr=0x7bffc81901fe, tstate=0x7bffb4249010) at Python/ceval.c:775
#22 0x0000555555ee754d in _PyEval_EvalFrameDefault (tstate=<optimized out>, frame=<optimized out>, throwflag=<optimized out>) at Python/generated_cases.c.h:1841
#23 0x0000555555ea8d04 in _PyEval_EvalFrame (tstate=0x7bffb4249010, frame=0x7e8ff709e220, throwflag=0) at ./Include/internal/pycore_ceval.h:118
#24 _PyEval_Vector (tstate=0x7bffb4249010, func=0x7bffb64cb290, locals=0x0, args=0x7bff9d4b8220, argcount=1, kwnames=0x0) at Python/ceval.c:2179
#25 0x0000555555ac6030 in _PyObject_VectorcallTstate (tstate=tstate@entry=0x7bffb4249010, callable=callable@entry=0x7bffb64cb290, args=0x7bff9d4b8220, nargsf=nargsf@entry=1,
    kwnames=kwnames@entry=0x0) at ./Include/internal/pycore_call.h:144
#26 0x0000555555acbf7f in _PyObject_VectorcallPrepend (tstate=<optimized out>, callable=<optimized out>, arg=<error reading variable: Cannot access memory at address 0x6>,
    args=<optimized out>, nargsf=0, kwnames=<optimized out>) at Objects/call.c:855
#27 0x0000555555ac8802 in _PyVectorcall_Call (tstate=<optimized out>, func=<optimized out>, callable=<optimized out>, tuple=0x55555695f428 <_PyRuntime+173928>, kwargs=<optimized out>)
    at Objects/call.c:273
#28 0x00005555562c9019 in thread_run (boot_raw=boot_raw@entry=0x7bffb41f08d0) at ./Modules/_threadmodule.c:387
#29 0x000055555610e226 in pythread_wrapper (arg=0x7bffb4010b50) at Python/thread_pthread.h:234
#30 0x00005555558a034b in asan_thread_start(void*) ()

Found using fusil by @vstinner.

CPython versions tested on:

CPython main branch, 3.15

Operating systems tested on:

Linux

Output from running 'python -VV' on the command line:

Python 3.15.0a7+ free-threading build (heads/main:4d0e8ee649c, Mar 29 2026, 23:20:29) [Clang 21.1.2 (2ubuntu6)]

Linked PRs
  • gh-156001

Contributor guide

Open the contributing guide

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

Research direction

Start with the reproducer and backtrace, then inspect Objects/frameobject.c around PyFrame_GetBack and the pdb/sys.monitoring call chain described in the issue. Run the MRE on a free-threaded CPython 3.15 build; done means the abort no longer occurs under the reported threaded workload and appropriate regression coverage is added.

Written by the indexing model from the issue text.

Assessment

Tech stack
python
Domain
backend
Issue type
Bug
Difficulty
4/5
Estimated time
3-5 days
Activity status
Stale
Clarity
Needs clarification
Newbie friendliness
25/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.