[issue10181] Problems with Py_buffer management in memoryobject.c (and elsewhere?)

[Nick Coghlan]

Nick Coghlan <ncoghlan at gmail.com> added the comment:

It's OK if the Py_buffer is on the stack - it's just a unique identifier for the exporter to use as a key, not something the exporter controls the lifecycle of (the latter is true only for the pointers *inside* the struct, such as buf, shape, strides, etc).

PyMemoryView_FromBuffer should be calling PyObject_Getbuffer on the view->obj member (it's one of the things that embedding the reference allows, just as it allowed removal of the separate obj argument from the PyObject_ReleaseBuffer signature). That way the source object knows there is now a *second* Py_buffer struct kicking around, and can decide whether to re-use the same internal pointers or create new ones.

----------

_______________________________________
Python tracker <report at bugs.python.org>
<http://bugs.python.org/issue10181>
_______________________________________