Skip to content

gh-159137: gc: Count GC survivors in untrack_tuples() instead of walking the young generation again #159149

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Closed
maurycy wants to merge 2 commits into from
+13 −4
Closed

gh-159137: gc: Count GC survivors in untrack_tuples() instead of walking the young generation again #159149

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
e3ee67d
up
maurycy Oct 10, 2026
0665b13
news
maurycy Oct 10, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
2 changes: 2 additions & 0 deletions Misc/NEWS.d/next/Core_and_Builtins/2026-10-10-23-01-13.gh-issue-159137.kTq3Lw.rst
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,2 @@
Avoid walking the surviving objects a second time after each garbage collection
to count them. Patch by Maurycy Pawłowski-Wieroński.
15 changes: 11 additions & 4 deletions Python/gc.c
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -661,18 +661,25 @@ move_unreachable(PyGC_Head *young, PyGC_Head *unreachable)
* and is much faster than a more complex approach that
* would untrack all relevant tuples.
*/
static void
static Py_ssize_t
untrack_tuples(PyGC_Head *head)
{
Py_ssize_t tracked = 0;
PyGC_Head *gc = GC_NEXT(head);
while (gc != head) {
PyObject *op = FROM_GC(gc);
PyGC_Head *next = GC_NEXT(gc);
if (PyTuple_CheckExact(op)) {
_PyTuple_MaybeUntrack(op);
if (!_PyObject_GC_IS_TRACKED(op)) {
gc = next;
continue;
}
}
tracked++;
gc = next;
}
return tracked;
}

/* Return true if object has a pre-PEP 442 finalization method. */
Expand Down Expand Up @@ -1514,11 +1521,11 @@ gc_collect_main(PyThreadState *tstate, int generation, _PyGC_Reason reason)

stats.candidates = deduce_unreachable(young, &unreachable);

untrack_tuples(young);
Py_ssize_t survivors = untrack_tuples(young);
/* Move reachable objects to next generation. */
if (young != old) {
if (generation == NUM_GENERATIONS - 2) {
gcstate->long_lived_pending += gc_list_size(young);
gcstate->long_lived_pending += survivors;
}
gc_list_merge(young, old);
}
Expand All @@ -1531,7 +1538,7 @@ gc_collect_main(PyThreadState *tstate, int generation, _PyGC_Reason reason)
// site; the cost is slightly more work for full collections on dicts
// with only atomic values.
gcstate->long_lived_pending = 0;
gcstate->long_lived_total = gc_list_size(young);
gcstate->long_lived_total = survivors;
}

/* All objects in unreachable are trash, but objects reachable from
Expand Down
Loading