summaryrefslogtreecommitdiff
Commit message (Expand)AuthorAgeFilesLines
* Add change note.issue252Jason Madden2021-10-122-0/+7
* Fix a double-addressing issue.Jason Madden2021-10-122-28/+39
* Whoops, can't enable PYTHONFAULTHANDLER twice. GHA complains.Jason Madden2021-10-121-1/+0
* Enable exception handling for old MSVCJason Madden2021-10-122-0/+8
* Always use RAII/stack-based locking/unlocking.Jason Madden2021-10-122-22/+80
* Make waiting for the greenlet cleanup much more deterministic.Jason Madden2021-10-125-51/+63
* More work on the race conditions seen on GHAJason Madden2021-10-122-4/+10
* Try harder to wait for active threads to die.Jason Madden2021-10-121-3/+10
* Looks like my cleanup isn't running?Jason Madden2021-10-121-0/+2
* Yet more debugging.Jason Madden2021-10-121-0/+4
* Yet more debugging.Jason Madden2021-10-122-4/+6
* More debuggingJason Madden2021-10-121-3/+14
* DebuggingJason Madden2021-10-122-0/+19
* Try to clean up some more leaked greenlets seen on linux.Jason Madden2021-10-123-2/+13
* Ensure more tests don't leak greenlets.Jason Madden2021-10-122-6/+61
* Update documentation.Jason Madden2021-10-123-36/+23
* Remove more debugging code.Jason Madden2021-10-129-272/+274
* Py2.7 testing.Jason Madden2021-10-122-10/+10
* More Py2/Py3 cleanups.Jason Madden2021-10-123-26/+54
* Tweak compiler flags.Jason Madden2021-10-123-4/+10
* Fix a faulty assertion.Jason Madden2021-10-121-12/+13
* Simplify thread state cleanup.Jason Madden2021-10-126-216/+247
* Reducing preprocessor reliance.Jason Madden2021-10-123-175/+154
* Try another way to pass the info around.Jason Madden2021-10-121-29/+38
* Passing the thread state through all the calls was too much, see if we can sa...Jason Madden2021-10-122-19/+38
* Eliminate a few unneeded calls te GET_THREAD_STATE.Jason Madden2021-10-121-109/+17
* Test 2.7 non-standard threading on GHA. Use 3.10 final on Appveyor.Jason Madden2021-10-122-4/+4
* Allow using the PyThread_ APIs back on Python 2.7 as well. Add tox.ini to tes...Jason Madden2021-10-123-7/+13
* Add GHA test cases that disable the use of standard threading.Jason Madden2021-10-123-7/+39
* Implement cleanup using the thread dictionary on platforms that require it.Jason Madden2021-10-123-19/+176
* Fix bad descriptor error on Python 2.7 for the main greenlet, and compilation...Jason Madden2021-10-124-18/+89
* Introduce PyMainGreenlet as a subtype of greenlet.Jason Madden2021-10-125-34/+64
* MFC was a no-go. So use the win32 API.Jason Madden2021-10-122-4/+7
* Trying MFC classes for the lock. Not sure if that will work.Jason Madden2021-10-121-6/+7
* Stop including headers we know aren't present. We got to that point of the co...Jason Madden2021-10-121-5/+3
* Use the old idiom.Jason Madden2021-10-122-4/+17
* Thus begins the fight against MSVC 9.Jason Madden2021-10-123-5/+7
* Progress on supporting older compilers.Jason Madden2021-10-124-80/+175
* Separating out concerns.Jason Madden2021-10-127-728/+811
* Queue pending cleanups to reduce the number of Py_AddPendingCall slots we take.Jason Madden2021-10-126-41/+155
* Some platforms don't leak the greenlet when explicitly referenced.Jason Madden2021-10-121-2/+5
* Fix the leak on older versions of CPython.Jason Madden2021-10-122-6/+110
* Sometimes we lose a greenlet.Jason Madden2021-10-121-4/+5
* Some debugging.Jason Madden2021-10-121-4/+19
* The indirect TLS support appears to be working, and I have fixed #252Jason Madden2021-10-123-272/+418
* Checkpoint moving to a more flexible use of TLS.Jason Madden2021-10-123-232/+428
* Hmm, sometimes the current greenlet was null on Windows? How?Jason Madden2021-10-122-31/+44
* More debuggingJason Madden2021-10-121-15/+28
* You can't call PyObject_Print if there's an error set.Jason Madden2021-10-122-18/+44
* Enable debug prints, windows is crashing.Jason Madden2021-10-123-51/+50