René's URL Explorer Experiment

{"@context":"https://schema.org","@type":"DiscussionForumPosting","headline":"Reduce redundant object creation while calling callback function from gc","articleBody":"https://github.com/python/cpython/blob/4b27972f5fe816d3616f97f8643d8ad922473ab5/Modules/gcmodule.c#L1391-L1402\r\n\r\nThe phase object can be created earlier with one-time creation and it can be replaced with vectorcall either.\r\nFrom my microbenchmark, there is 4-5% performance improvement by doing this.\r\n\r\n## microbenchmark\r\n```python\r\n\r\nimport pyperf\r\nimport gc\r\n\r\n\r\ndef benchamark_collection(loops):\r\n    def callback_foo(phase, info):\r\n        pass\r\n    for _ in range(100):\r\n        gc.callbacks.append(callback_foo)\r\n\r\n    total_time = 0\r\n    for _ in range(loops):\r\n        t0 = pyperf.perf_counter()\r\n        collected = gc.collect()\r\n        total_time += pyperf.perf_counter() - t0\r\n    return total_time\r\n\r\n\r\nif __name__ == \"__main__\":\r\n    runner = pyperf.Runner()\r\n    runner.metadata[\"description\"] = \"GC callback benchmark\"\r\n    runner.bench_time_func(\"create_gc_cycles\", benchamark_collection)\r\n`````\n\n\u003c!-- gh-linked-prs --\u003e\n### Linked PRs\n* gh-104030\n\u003c!-- /gh-linked-prs --\u003e\n","author":{"url":"https://github.com/corona10","@type":"Person","name":"corona10"},"datePublished":"2023-05-01T05:08:39.000Z","interactionStatistic":{"@type":"InteractionCounter","interactionType":"https://schema.org/CommentAction","userInteractionCount":0},"url":"https://github.com/104028/cpython/issues/104028"}