Add regression test for cross-thread committed memory growth (#814)#815
Draft
Add regression test for cross-thread committed memory growth (#814)#815
Conversation
Add a perf test that reproduces the workload described in issue #814: a pool of worker threads each allocate large chunks (512KB–16MB) and hand them to a random different worker for deallocation. Every thread both allocates and frees, but always frees objects originally allocated by another thread — the pattern reported to cause unbounded committed memory growth in a game engine. The test runs for a configurable duration (default 30 s), samples snmalloc's committed memory once per second, and compares the average committed memory in the first quarter of the run against the last quarter. If the ratio exceeds 1.5x the test exits with code 1 (FAIL), making it suitable as a CTest regression gate. The CMake build system auto-discovers the new test directory, so no CMakeLists.txt changes are needed.
…20s default - Track per-worker requested/returned bytes (WorkerStats) to compute a live-requested lower bound on expected committed memory. - Print Live(MB) column showing sum(requested) - sum(returned) across all workers at each sample point. - Compare 2nd vs 4th quarter average committed (skip warm-up Q1) for a more accurate growth ratio. - Increase default duration from 30s to 120s for better signal.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Add a perf test that reproduces the workload described in issue #814: a pool of worker threads each allocate large chunks (512KB–16MB) and hand them to a random different worker for deallocation. Every thread both allocates and frees, but always frees objects originally allocated by another thread — the pattern reported to cause unbounded committed memory growth in a game engine.
The test runs for a configurable duration (default 30 s), samples snmalloc's committed memory once per second, and compares the average committed memory in the first quarter of the run against the last quarter. If the ratio exceeds 1.5x the test exits with code 1 (FAIL), making it suitable as a CTest regression gate.
The CMake build system auto-discovers the new test directory, so no CMakeLists.txt changes are needed.