Skip to content

Fix flaky test_max_pending_count by preventing premature GC #857

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

Open
xyaz1313 wants to merge 1 commit into
base: main
Choose a base branch
from
Open

Fix flaky test_max_pending_count by preventing premature GC #857

Changes from all commits
Commits
Show all changes
1 commit
Select commit
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
11 changes: 9 additions & 2 deletions numba_cuda/numba/cuda/tests/cudadrv/test_deallocations.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -2,6 +2,7 @@
# SPDX-License-Identifier: BSD-2-Clause

from contextlib import contextmanager
import gc

import numpy as np

Expand All @@ -24,12 +25,18 @@ def test_max_pending_count(self):
deallocs = cuda.current_context().memory_manager.deallocations
deallocs.clear()
self.assertEqual(len(deallocs), 0)
# deallocate to maximum count
# Allocate device arrays, keeping references to prevent GC from
# collecting them prematurely (which caused flaky counts).
arrays = [cuda.to_device(np.arange(1))
for _ in range(config.CUDA_DEALLOCS_COUNT)]
# Delete arrays one by one and check pending count increments
for i in range(config.CUDA_DEALLOCS_COUNT):
cuda.to_device(np.arange(1))
del arrays[i]
gc.collect() # Ensure finalizers run
self.assertEqual(len(deallocs), i + 1)
# one more to trigger .clear()
cuda.to_device(np.arange(1))
gc.collect()
self.assertEqual(len(deallocs), 0)

@skip_if_external_memmgr("Deallocation specific to Numba memory management")
Expand Down