gh-145376: Fix various refleaks Objects/ by eendebakpt · Pull Request #145609 · python/cpython · GitHub

Issue: Various reference leaks in the core #145376

Be careful here. PyCurses_ConvertToString has a weird return type so you might not necessarily have bytesobj set to something (it might be in wstr instead).

Good catch!

Could you move this line after the if instead so we don't have to worry about cleaning up the ref?

I check all changes and they LGTM.

Merged. Thans for the fix.

⚠️⚠️⚠️ Buildbot failure ⚠️⚠️⚠️

Hi! The buildbot AMD64 FreeBSD 3.x (tier-3) has failed when building commit 8060aa5.

What do you need to do:

Don't panic.
Check the buildbot page in the devguide if you don't know what the buildbots are or how they work.
Go to the page of the buildbot that failed (https://buildbot.python.org/#/builders/1223/builds/8317) and take a look at the build logs.
Check if the failure is related to this commit (8060aa5) or if it is a false positive.
If the failure is related to this commit, please, reflect that on the issue and make a new Pull Request with a fix.

You can take a look at the buildbot page here:

https://buildbot.python.org/#/builders/1223/builds/8317

Summary of the results of the build (if available):

Click to see traceback logs

remote: Enumerating objects: 21, done.        
remote: Counting objects:   4% (1/21)        
remote: Counting objects:   9% (2/21)        
remote: Counting objects:  14% (3/21)        
remote: Counting objects:  19% (4/21)        
remote: Counting objects:  23% (5/21)        
remote: Counting objects:  28% (6/21)        
remote: Counting objects:  33% (7/21)        
remote: Counting objects:  38% (8/21)        
remote: Counting objects:  42% (9/21)        
remote: Counting objects:  47% (10/21)        
remote: Counting objects:  52% (11/21)        
remote: Counting objects:  57% (12/21)        
remote: Counting objects:  61% (13/21)        
remote: Counting objects:  66% (14/21)        
remote: Counting objects:  71% (15/21)        
remote: Counting objects:  76% (16/21)        
remote: Counting objects:  80% (17/21)        
remote: Counting objects:  85% (18/21)        
remote: Counting objects:  90% (19/21)        
remote: Counting objects:  95% (20/21)        
remote: Counting objects: 100% (21/21)        
remote: Counting objects: 100% (21/21), done.        
remote: Compressing objects:  10% (1/10)        
remote: Compressing objects:  20% (2/10)        
remote: Compressing objects:  30% (3/10)        
remote: Compressing objects:  40% (4/10)        
remote: Compressing objects:  50% (5/10)        
remote: Compressing objects:  60% (6/10)        
remote: Compressing objects:  70% (7/10)        
remote: Compressing objects:  80% (8/10)        
remote: Compressing objects:  90% (9/10)        
remote: Compressing objects: 100% (10/10)        
remote: Compressing objects: 100% (10/10), done.        
remote: Total 11 (delta 10), reused 1 (delta 1), pack-reused 0 (from 0)        
From https://github.com/python/cpython
 * branch                    main       -> FETCH_HEAD
Note: switching to '8060aa5d7dd02c22a987e23cf5dbb7049b50b042'.

You are in 'detached HEAD' state. You can look around, make experimental
changes and commit them, and you can discard any commits you make in this
state without impacting any branches by switching back to a branch.

If you want to create a new branch to retain commits you create, you may
do so (now or later) by using -c with the switch command. Example:

  git switch -c <new-branch-name>

Or undo this operation with:

  git switch -

Turn off this advice by setting config variable advice.detachedHead to false

HEAD is now at 8060aa5d7dd gh-145376: Fix various refleaks in Objects/ (#145609)
Switched to and reset branch 'main'

1 error

⚠️⚠️⚠️ Buildbot failure ⚠️⚠️⚠️

Hi! The buildbot ARM64 MacOS M1 NoGIL 3.x (tier-2) has failed when building commit 8060aa5.

What do you need to do:

Don't panic.
Check the buildbot page in the devguide if you don't know what the buildbots are or how they work.
Go to the page of the buildbot that failed (https://buildbot.python.org/#/builders/1270/builds/6984) and take a look at the build logs.
Check if the failure is related to this commit (8060aa5) or if it is a false positive.
If the failure is related to this commit, please, reflect that on the issue and make a new Pull Request with a fix.

You can take a look at the buildbot page here:

https://buildbot.python.org/#/builders/1270/builds/6984

Summary of the results of the build (if available):

Click to see traceback logs

remote: Enumerating objects: 21, done.        
remote: Counting objects:   4% (1/21)        
remote: Counting objects:   9% (2/21)        
remote: Counting objects:  14% (3/21)        
remote: Counting objects:  19% (4/21)        
remote: Counting objects:  23% (5/21)        
remote: Counting objects:  28% (6/21)        
remote: Counting objects:  33% (7/21)        
remote: Counting objects:  38% (8/21)        
remote: Counting objects:  42% (9/21)        
remote: Counting objects:  47% (10/21)        
remote: Counting objects:  52% (11/21)        
remote: Counting objects:  57% (12/21)        
remote: Counting objects:  61% (13/21)        
remote: Counting objects:  66% (14/21)        
remote: Counting objects:  71% (15/21)        
remote: Counting objects:  76% (16/21)        
remote: Counting objects:  80% (17/21)        
remote: Counting objects:  85% (18/21)        
remote: Counting objects:  90% (19/21)        
remote: Counting objects:  95% (20/21)        
remote: Counting objects: 100% (21/21)        
remote: Counting objects: 100% (21/21), done.        
remote: Compressing objects:  10% (1/10)        
remote: Compressing objects:  20% (2/10)        
remote: Compressing objects:  30% (3/10)        
remote: Compressing objects:  40% (4/10)        
remote: Compressing objects:  50% (5/10)        
remote: Compressing objects:  60% (6/10)        
remote: Compressing objects:  70% (7/10)        
remote: Compressing objects:  80% (8/10)        
remote: Compressing objects:  90% (9/10)        
remote: Compressing objects: 100% (10/10)        
remote: Compressing objects: 100% (10/10), done.        
remote: Total 11 (delta 10), reused 1 (delta 1), pack-reused 0 (from 0)        
From https://github.com/python/cpython
 * branch                    main       -> FETCH_HEAD
Note: switching to '8060aa5d7dd02c22a987e23cf5dbb7049b50b042'.

You are in 'detached HEAD' state. You can look around, make experimental
changes and commit them, and you can discard any commits you make in this
state without impacting any branches by switching back to a branch.

If you want to create a new branch to retain commits you create, you may
do so (now or later) by using -c with the switch command. Example:

  git switch -c <new-branch-name>

Or undo this operation with:

  git switch -

Turn off this advice by setting config variable advice.detachedHead to false

HEAD is now at 8060aa5d7dd gh-145376: Fix various refleaks in Objects/ (#145609)
Switched to and reset branch 'main'

configure: WARNING: pkg-config is missing. Some dependencies may not be detected correctly.

./Modules/_cursesmodule.c:1135:16: error: use of undeclared identifier 'wstr'; did you mean 'str'?
 1135 |         assert(wstr == NULL);
      |                ^~~~
      |                str
/Library/Developer/CommandLineTools/SDKs/MacOSX.sdk/usr/include/assert.h:75:25: note: expanded from macro 'assert'
   75 |     (__builtin_expect(!(e), 0) ? __assert_rtn(__func__, __ASSERT_FILE_NAME, __LINE__, #e) : (void)0)
      |                         ^
./Modules/_cursesmodule.c:1089:52: note: 'str' declared here
 1089 |                            int y, int x, PyObject *str, int group_right_1,
      |                                                    ^
1 error generated.
make: *** [Modules/_cursesmodule.o] Error 1
make: *** Waiting for unfinished jobs....

find: build: No such file or directory
find: build: No such file or directory
find: build: No such file or directory
find: build: No such file or directory
make: [clean-retain-profile] Error 1 (ignored)

Oh, macOS and FreeBSD fail to build _cursesmodule.c with this change!

./Modules/_cursesmodule.c:1135:16: error: use of undeclared identifier 'wstr'; did you mean 'str'?

I wrote #145669 to fix it.

⚠️⚠️⚠️ Buildbot failure ⚠️⚠️⚠️

Hi! The buildbot x86-64 MacOS Intel NoGIL 3.x (tier-2) has failed when building commit 8060aa5.

What do you need to do:

Don't panic.
Check the buildbot page in the devguide if you don't know what the buildbots are or how they work.
Go to the page of the buildbot that failed (https://buildbot.python.org/#/builders/1258/builds/7049) and take a look at the build logs.
Check if the failure is related to this commit (8060aa5) or if it is a false positive.
If the failure is related to this commit, please, reflect that on the issue and make a new Pull Request with a fix.

You can take a look at the buildbot page here:

https://buildbot.python.org/#/builders/1258/builds/7049

Summary of the results of the build (if available):

Click to see traceback logs

remote: Enumerating objects: 72, done.        
remote: Counting objects:   1% (1/72)        
remote: Counting objects:   2% (2/72)        
remote: Counting objects:   4% (3/72)        
remote: Counting objects:   5% (4/72)        
remote: Counting objects:   6% (5/72)        
remote: Counting objects:   8% (6/72)        
remote: Counting objects:   9% (7/72)        
remote: Counting objects:  11% (8/72)        
remote: Counting objects:  12% (9/72)        
remote: Counting objects:  13% (10/72)        
remote: Counting objects:  15% (11/72)        
remote: Counting objects:  16% (12/72)        
remote: Counting objects:  18% (13/72)        
remote: Counting objects:  19% (14/72)        
remote: Counting objects:  20% (15/72)        
remote: Counting objects:  22% (16/72)        
remote: Counting objects:  23% (17/72)        
remote: Counting objects:  25% (18/72)        
remote: Counting objects:  26% (19/72)        
remote: Counting objects:  27% (20/72)        
remote: Counting objects:  29% (21/72)        
remote: Counting objects:  30% (22/72)        
remote: Counting objects:  31% (23/72)        
remote: Counting objects:  33% (24/72)        
remote: Counting objects:  34% (25/72)        
remote: Counting objects:  36% (26/72)        
remote: Counting objects:  37% (27/72)        
remote: Counting objects:  38% (28/72)        
remote: Counting objects:  40% (29/72)        
remote: Counting objects:  41% (30/72)        
remote: Counting objects:  43% (31/72)        
remote: Counting objects:  44% (32/72)        
remote: Counting objects:  45% (33/72)        
remote: Counting objects:  47% (34/72)        
remote: Counting objects:  48% (35/72)        
remote: Counting objects:  50% (36/72)        
remote: Counting objects:  51% (37/72)        
remote: Counting objects:  52% (38/72)        
remote: Counting objects:  54% (39/72)        
remote: Counting objects:  55% (40/72)        
remote: Counting objects:  56% (41/72)        
remote: Counting objects:  58% (42/72)        
remote: Counting objects:  59% (43/72)        
remote: Counting objects:  61% (44/72)        
remote: Counting objects:  62% (45/72)        
remote: Counting objects:  63% (46/72)        
remote: Counting objects:  65% (47/72)        
remote: Counting objects:  66% (48/72)        
remote: Counting objects:  68% (49/72)        
remote: Counting objects:  69% (50/72)        
remote: Counting objects:  70% (51/72)        
remote: Counting objects:  72% (52/72)        
remote: Counting objects:  73% (53/72)        
remote: Counting objects:  75% (54/72)        
remote: Counting objects:  76% (55/72)        
remote: Counting objects:  77% (56/72)        
remote: Counting objects:  79% (57/72)        
remote: Counting objects:  80% (58/72)        
remote: Counting objects:  81% (59/72)        
remote: Counting objects:  83% (60/72)        
remote: Counting objects:  84% (61/72)        
remote: Counting objects:  86% (62/72)        
remote: Counting objects:  87% (63/72)        
remote: Counting objects:  88% (64/72)        
remote: Counting objects:  90% (65/72)        
remote: Counting objects:  91% (66/72)        
remote: Counting objects:  93% (67/72)        
remote: Counting objects:  94% (68/72)        
remote: Counting objects:  95% (69/72)        
remote: Counting objects:  97% (70/72)        
remote: Counting objects:  98% (71/72)        
remote: Counting objects: 100% (72/72)        
remote: Counting objects: 100% (72/72), done.        
remote: Compressing objects:   2% (1/37)        
remote: Compressing objects:   5% (2/37)        
remote: Compressing objects:   8% (3/37)        
remote: Compressing objects:  10% (4/37)        
remote: Compressing objects:  13% (5/37)        
remote: Compressing objects:  16% (6/37)        
remote: Compressing objects:  18% (7/37)        
remote: Compressing objects:  21% (8/37)        
remote: Compressing objects:  24% (9/37)        
remote: Compressing objects:  27% (10/37)        
remote: Compressing objects:  29% (11/37)        
remote: Compressing objects:  32% (12/37)        
remote: Compressing objects:  35% (13/37)        
remote: Compressing objects:  37% (14/37)        
remote: Compressing objects:  40% (15/37)        
remote: Compressing objects:  43% (16/37)        
remote: Compressing objects:  45% (17/37)        
remote: Compressing objects:  48% (18/37)        
remote: Compressing objects:  51% (19/37)        
remote: Compressing objects:  54% (20/37)        
remote: Compressing objects:  56% (21/37)        
remote: Compressing objects:  59% (22/37)        
remote: Compressing objects:  62% (23/37)        
remote: Compressing objects:  64% (24/37)        
remote: Compressing objects:  67% (25/37)        
remote: Compressing objects:  70% (26/37)        
remote: Compressing objects:  72% (27/37)        
remote: Compressing objects:  75% (28/37)        
remote: Compressing objects:  78% (29/37)        
remote: Compressing objects:  81% (30/37)        
remote: Compressing objects:  83% (31/37)        
remote: Compressing objects:  86% (32/37)        
remote: Compressing objects:  89% (33/37)        
remote: Compressing objects:  91% (34/37)        
remote: Compressing objects:  94% (35/37)        
remote: Compressing objects:  97% (36/37)        
remote: Compressing objects: 100% (37/37)        
remote: Compressing objects: 100% (37/37), done.        
remote: Total 41 (delta 35), reused 6 (delta 4), pack-reused 0 (from 0)        
From https://github.com/python/cpython
 * branch                    main       -> FETCH_HEAD
Note: switching to '8060aa5d7dd02c22a987e23cf5dbb7049b50b042'.

You are in 'detached HEAD' state. You can look around, make experimental
changes and commit them, and you can discard any commits you make in this
state without impacting any branches by switching back to a branch.

If you want to create a new branch to retain commits you create, you may
do so (now or later) by using -c with the switch command. Example:

  git switch -c <new-branch-name>

Or undo this operation with:

  git switch -

Turn off this advice by setting config variable advice.detachedHead to false

HEAD is now at 8060aa5d7dd gh-145376: Fix various refleaks in Objects/ (#145609)
Switched to and reset branch 'main'

configure: WARNING: pkg-config is missing. Some dependencies may not be detected correctly.

./Modules/_cursesmodule.c:1135:16: error: use of undeclared identifier 'wstr'; did you mean 'str'?
        assert(wstr == NULL);
               ^~~~
               str
/Library/Developer/CommandLineTools/SDKs/MacOSX.sdk/usr/include/assert.h:99:25: note: expanded from macro 'assert'
    (__builtin_expect(!(e), 0) ? __assert_rtn(__func__, __ASSERT_FILE_NAME, __LINE__, #e) : (void)0)
                        ^
./Modules/_cursesmodule.c:1089:52: note: 'str' declared here
                           int y, int x, PyObject *str, int group_right_1,
                                                   ^
1 error generated.
make: *** [Modules/_cursesmodule.o] Error 1
make: *** Waiting for unfinished jobs....

find: build: No such file or directory
find: build: No such file or directory
find: build: No such file or directory
find: build: No such file or directory
make: [clean-retain-profile] Error 1 (ignored)

eendebakpt added 7 commits March 5, 2026 16:02

Fix refleak in binascii.c

577de24

Fix refleak in _cursesmodule.c

e4fa631

Fix refleak in object.c

efb497b

Fix refleak in genericaliasobject.c

c9d5cac

Fix refleak in typevarobject.c

3bafd1a

Fix decref on null in structseq.c

f0e38ea

Fix refleak in Objects/unicodeobject.c

de71a74

eendebakpt requested review from JelleZijlstra and ZeroIntensity as code owners March 6, 2026 19:29

bedevere-app Bot mentioned this pull request Mar 6, 2026

Various reference leaks in the core #145376

Closed

bedevere-app Bot added the awaiting review label Mar 6, 2026

picnixz reviewed Mar 6, 2026

View reviewed changes

eendebakpt added 2 commits March 6, 2026 20:50

missing commit

b128a9d

review comment

a337697

JelleZijlstra reviewed Mar 6, 2026

View reviewed changes

review comment

eadf7a1

eendebakpt added the skip news label Mar 6, 2026

JelleZijlstra approved these changes Mar 7, 2026

View reviewed changes

bedevere-app Bot added awaiting merge and removed awaiting review labels Mar 7, 2026

vstinner approved these changes Mar 9, 2026

View reviewed changes

vstinner merged commit 8060aa5 into python:main Mar 9, 2026
54 checks passed

bedevere-app Bot removed the awaiting merge label Mar 9, 2026

eendebakpt deleted the refleaks_misc branch March 19, 2026 21:39

ljfp pushed a commit to ljfp/cpython that referenced this pull request Apr 25, 2026

pythongh-145376 : Fix various refleaks in Objects/ (python#145609 )

0b43869

This was referenced Sep 4, 2026

window.insnstr() leaks its bytes argument when setting attr fails #156953

Closed

gh-156953: Fix a reference leak in curses window.insnstr() #156954

Merged

                       attr_old = getattrs(self->win);
                       if (curses_wattrset(self, attr, "addstr") < 0) {
                           curses_release_wstr(strtype, wstr);
+                          Py_DECREF(bytesobj);

gh-145376: Fix various refleaks Objects/ by eendebakpt · Pull Request #145609 · python/cpython · GitHub

Repository navigation

gh-145376: Fix various refleaks Objects/ - #145609

gh-145376: Fix various refleaks Objects/#145609
vstinner merged 10 commits into
python:mainfrom
eendebakpt:refleaks_misc

eendebakpt commented Mar 6, 2026 •
edited by bedevere-app Bot
Loading

Uh oh!

Uh oh!

picnixz Mar 6, 2026

Uh oh!

eendebakpt Mar 6, 2026

Uh oh!

JelleZijlstra Mar 6, 2026

Uh oh!

vstinner left a comment

Uh oh!

Uh oh!

vstinner commented Mar 9, 2026

Uh oh!

bedevere-bot commented Mar 9, 2026

Uh oh!

bedevere-bot commented Mar 9, 2026

Uh oh!

vstinner commented Mar 9, 2026

Uh oh!

bedevere-bot commented Mar 9, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

5 participants

gh-145376: Fix various refleaks Objects/ by eendebakpt · Pull Request #145609 · python/cpython · GitHub

Repository navigation

gh-145376: Fix various refleaks Objects/ - #145609

gh-145376: Fix various refleaks Objects/#145609vstinner merged 10 commits intopython:mainpython/cpython:mainfrom eendebakpt:refleaks_misceendebakpt/cpython:refleaks_miscCopy head branch name to clipboard

Conversation

eendebakpt commented Mar 6, 2026 • edited by bedevere-app Bot Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

picnixz Mar 6, 2026

Choose a reason for hiding this comment

Uh oh!

eendebakpt Mar 6, 2026

Choose a reason for hiding this comment

Uh oh!

JelleZijlstra Mar 6, 2026

Choose a reason for hiding this comment

Uh oh!

vstinner left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

vstinner commented Mar 9, 2026

Uh oh!

bedevere-bot commented Mar 9, 2026

⚠️⚠️⚠️ Buildbot failure ⚠️⚠️⚠️

Uh oh!

bedevere-bot commented Mar 9, 2026

⚠️⚠️⚠️ Buildbot failure ⚠️⚠️⚠️

Uh oh!

vstinner commented Mar 9, 2026

Uh oh!

bedevere-bot commented Mar 9, 2026

⚠️⚠️⚠️ Buildbot failure ⚠️⚠️⚠️

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

5 participants

gh-145376: Fix various refleaks Objects/#145609
vstinner merged 10 commits into
python:mainfrom
eendebakpt:refleaks_misc

eendebakpt commented Mar 6, 2026 •
edited by bedevere-app Bot
Loading