Possible leak in pystate.c

[tregnago]

BPO	216236
Nosy	@gvanrossum

^{Note: these values reflect the state of the issue at the time it was migrated and might not reflect the current state.}

Show more details

GitHub fields:

assignee = None
closed_at = <Date 2000-10-06.14:37:03.000>
created_at = <Date 2000-10-06.13:26:11.000>
labels = ['interpreter-core', 'invalid']
title = 'Possible leak in pystate.c'
updated_at = <Date 2000-10-06.14:37:03.000>
user = 'https://bugs.python.org/tregnago'

bugs.python.org fields:

activity = <Date 2000-10-06.14:37:03.000>
actor = 'gvanrossum'
assignee = 'none'
closed = True
closed_date = None
closer = None
components = ['Interpreter Core']
creation = <Date 2000-10-06.13:26:11.000>
creator = 'tregnago'
dependencies = []
files = []
hgrepos = []
issue_num = 216236
keywords = []
message_count = 2.0
messages = ['1864', '1865']
nosy_count = 2.0
nosy_names = ['gvanrossum', 'tregnago']
pr_nums = []
priority = 'normal'
resolution = 'not a bug'
stage = None
status = 'closed'
superseder = None
type = None
url = 'https://bugs.python.org/issue216236'
versions = []

[tregnago]

Possible leak in pystate.c
It seems that head_mutex allocated with the HEAD_INIT() macro is never deleted.
Please, give me explanations...

[gvanrossum]

It would only be a leak if head was ever overwritten. The way the code is written however, head_mutex is a global that is initialized the first time HEAD_INIT() is used and never again. While you may argue that this allocates some dynamic memory that is never deleted, the memory is still live (from the static global head_mutex pointer variable) so it is not a leak.