Message301353
I reverted because of the crash in test_threading. I'm pretty sure there is a bug with the locking of bufferedio.c, related to threads and flush. Here is the stacktrace I get (my patch applied, I'm trying to write a Python test that triggers the SEGV without my flush patch).
Attached is the script that triggers the crash.
Thread 3 received signal SIGSEGV, Segmentation fault.
0x00000001000c06d4 in PyObject_GetAttr (
v=<unknown at remote 0xdbdbdbdbdbdbdbdb>, name='seek')
at ../Objects/object.c:882
882 PyTypeObject *tp = Py_TYPE(v);
(gdb) bt
#0 0x00000001000c06d4 in PyObject_GetAttr (
v=<unknown at remote 0xdbdbdbdbdbdbdbdb>, name='seek')
at ../Objects/object.c:882
#1 0x00000001000440e9 in PyObject_CallMethodObjArgs (
callable=<unknown at remote 0xdbdbdbdbdbdbdbdb>, name='seek')
at ../Objects/call.c:1212
#2 0x00000001003005a2 in _buffered_raw_seek (self=0x1053a39b8, target=0,
whence=1) at ../Modules/_io/bufferedio.c:742
#3 0x000000010030015e in buffered_flush_and_rewind_unlocked (self=0x1053a39b8)
at ../Modules/_io/bufferedio.c:851
#4 0x00000001002fed12 in buffered_flush (self=0x1053a39b8, args=0x0)
at ../Modules/_io/bufferedio.c:869
#5 0x00000001002feb4d in _PyIO_atexit_flush ()
at ../Modules/_io/bufferedio.c:1863
[...] |
|
Date |
User |
Action |
Args |
2017-09-05 18:22:15 | nascheme | set | recipients:
+ nascheme, tim.peters, arigo, vstinner, nikratio, neologix, martin.panter, serhiy.storchaka, xgdomingo |
2017-09-05 18:22:15 | nascheme | set | messageid: <1504635735.13.0.250216311091.issue17852@psf.upfronthosting.co.za> |
2017-09-05 18:22:15 | nascheme | link | issue17852 messages |
2017-09-05 18:22:15 | nascheme | create | |
|