Killing greenlets across threads leaks a list #251
Comments
jamadden
added a commit
that referenced
this issue
Aug 7, 2021
All tested versions of Python are demonstrating a growth of one extra list.
jamadden
added a commit
that referenced
this issue
Aug 10, 2021
kraj
pushed a commit
to YoeDistro/meta-openembedded
that referenced
this issue
Oct 27, 2021
1.1.2 (2021-09-29) ================== - Fix a potential crash due to a reference counting error when Python subclasses of ``greenlet.greenlet`` were deallocated. The crash became more common on Python 3.10; on earlier versions, silent memory corruption could result. See `issue 245 <https://github.com/python-greenlet/greenlet/issues/245>`_. Patch by fygao-wish. - Fix a leak of a list object when the last reference to a greenlet was deleted from some other thread than the one to which it belonged. For this to work correctly, you must call a greenlet API like ``getcurrent()`` before the thread owning the greenlet exits: this is a long-standing limitation that can also lead to the leak of a thread's main greenlet if not called; we hope to lift this limitation. Note that in some cases this may also fix leaks of greenlet objects themselves. See `issue 251 <https://github.com/python-greenlet/greenlet/issues/251>`_. - Python 3.10: Tracing or profiling into a spawned greenlet didn't work as expected. See `issue 256 <https://github.com/python-greenlet/greenlet/issues/256>`_, reported by Joe Rickerby. Signed-off-by: Zheng Ruoqin <zhengrq.fnst@fujitsu.com> Signed-off-by: Khem Raj <raj.khem@gmail.com>
kraj
pushed a commit
to YoeDistro/meta-openembedded
that referenced
this issue
Oct 27, 2021
1.1.2 (2021-09-29) ================== - Fix a potential crash due to a reference counting error when Python subclasses of ``greenlet.greenlet`` were deallocated. The crash became more common on Python 3.10; on earlier versions, silent memory corruption could result. See `issue 245 <https://github.com/python-greenlet/greenlet/issues/245>`_. Patch by fygao-wish. - Fix a leak of a list object when the last reference to a greenlet was deleted from some other thread than the one to which it belonged. For this to work correctly, you must call a greenlet API like ``getcurrent()`` before the thread owning the greenlet exits: this is a long-standing limitation that can also lead to the leak of a thread's main greenlet if not called; we hope to lift this limitation. Note that in some cases this may also fix leaks of greenlet objects themselves. See `issue 251 <https://github.com/python-greenlet/greenlet/issues/251>`_. - Python 3.10: Tracing or profiling into a spawned greenlet didn't work as expected. See `issue 256 <https://github.com/python-greenlet/greenlet/issues/256>`_, reported by Joe Rickerby. Signed-off-by: Zheng Ruoqin <zhengrq.fnst@fujitsu.com> Signed-off-by: Khem Raj <raj.khem@gmail.com>
halstead
pushed a commit
to openembedded/meta-openembedded
that referenced
this issue
Oct 28, 2021
1.1.2 (2021-09-29) ================== - Fix a potential crash due to a reference counting error when Python subclasses of ``greenlet.greenlet`` were deallocated. The crash became more common on Python 3.10; on earlier versions, silent memory corruption could result. See `issue 245 <https://github.com/python-greenlet/greenlet/issues/245>`_. Patch by fygao-wish. - Fix a leak of a list object when the last reference to a greenlet was deleted from some other thread than the one to which it belonged. For this to work correctly, you must call a greenlet API like ``getcurrent()`` before the thread owning the greenlet exits: this is a long-standing limitation that can also lead to the leak of a thread's main greenlet if not called; we hope to lift this limitation. Note that in some cases this may also fix leaks of greenlet objects themselves. See `issue 251 <https://github.com/python-greenlet/greenlet/issues/251>`_. - Python 3.10: Tracing or profiling into a spawned greenlet didn't work as expected. See `issue 256 <https://github.com/python-greenlet/greenlet/issues/256>`_, reported by Joe Rickerby. Signed-off-by: Zheng Ruoqin <zhengrq.fnst@fujitsu.com> Signed-off-by: Khem Raj <raj.khem@gmail.com> Signed-off-by: Trevor Gamblin <trevor.gamblin@windriver.com>
HerrMuellerluedenscheid
pushed a commit
to HerrMuellerluedenscheid/meta-openembedded
that referenced
this issue
Jan 16, 2022
1.1.2 (2021-09-29) ================== - Fix a potential crash due to a reference counting error when Python subclasses of ``greenlet.greenlet`` were deallocated. The crash became more common on Python 3.10; on earlier versions, silent memory corruption could result. See `issue 245 <https://github.com/python-greenlet/greenlet/issues/245>`_. Patch by fygao-wish. - Fix a leak of a list object when the last reference to a greenlet was deleted from some other thread than the one to which it belonged. For this to work correctly, you must call a greenlet API like ``getcurrent()`` before the thread owning the greenlet exits: this is a long-standing limitation that can also lead to the leak of a thread's main greenlet if not called; we hope to lift this limitation. Note that in some cases this may also fix leaks of greenlet objects themselves. See `issue 251 <https://github.com/python-greenlet/greenlet/issues/251>`_. - Python 3.10: Tracing or profiling into a spawned greenlet didn't work as expected. See `issue 256 <https://github.com/python-greenlet/greenlet/issues/256>`_, reported by Joe Rickerby. Signed-off-by: Zheng Ruoqin <zhengrq.fnst@fujitsu.com> Signed-off-by: Khem Raj <raj.khem@gmail.com> Signed-off-by: Trevor Gamblin <trevor.gamblin@windriver.com>
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
See #245 (comment)
The text was updated successfully, but these errors were encountered: