diff --git a/Lib/test/test_profiling/test_sampling_profiler/test_binary_format.py b/Lib/test/test_profiling/test_sampling_profiler/test_binary_format.py index 5efc60a9211175..fce1809fedd686 100644 --- a/Lib/test/test_profiling/test_sampling_profiler/test_binary_format.py +++ b/Lib/test/test_profiling/test_sampling_profiler/test_binary_format.py @@ -1175,6 +1175,49 @@ def test_alternating_threads_status_changes(self): self.assertEqual(count, 100) self.assert_samples_equal(samples, collector) + def test_rle_alternating_status_batches_correctly(self): + """A repeat record whose status alternates every sample replays as N + single-status batches with the right cumulative timestamps.""" + class BatchCollector: + def __init__(self): + self.batches = [] + + def collect(self, stack_frames, timestamps_us): + for interp in stack_frames: + for thread in interp.threads: + self.batches.append( + (thread.status, list(timestamps_us)) + ) + + def export(self, filename): + pass + + num_samples = 2000 + frame = make_frame("rle.py", 42, "rle_func") + with tempfile.NamedTemporaryFile(suffix=".bin", delete=False) as f: + filename = f.name + self.temp_files.append(filename) + + writer = BinaryCollector(filename, 1000, compression="none") + expected = [] + for i in range(num_samples): + status = THREAD_STATUS_HAS_GIL if i % 2 else 0 + ts = 1000 + i + expected.append((status, [ts])) + sample = [ + make_interpreter(0, [make_thread(1, [frame], status)]) + ] + writer.collect(sample, timestamp_us=ts) + writer.export(None) + + collector = BatchCollector() + with BinaryReader(filename) as reader: + count = reader.replay_samples(collector) + + self.assertEqual(count, num_samples) + self.assertEqual(len(collector.batches), num_samples) + self.assertEqual(collector.batches, expected) + class TestBinaryStress(BinaryFormatTestBase): """Randomized stress tests for binary format.""" diff --git a/Misc/NEWS.d/next/Library/2026-07-01-18-00-00.gh-issue-152721.rlequad.rst b/Misc/NEWS.d/next/Library/2026-07-01-18-00-00.gh-issue-152721.rlequad.rst new file mode 100644 index 00000000000000..4dac0ed245bd67 --- /dev/null +++ b/Misc/NEWS.d/next/Library/2026-07-01-18-00-00.gh-issue-152721.rlequad.rst @@ -0,0 +1,2 @@ +Fix quadratic replay time in the :mod:`profiling.sampling` binary reader when a +profile's run-length-encoded samples alternate thread status. diff --git a/Modules/_remote_debugging/binary_io_reader.c b/Modules/_remote_debugging/binary_io_reader.c index ce1c3d232c94e0..b2a89b1177ef09 100644 --- a/Modules/_remote_debugging/binary_io_reader.c +++ b/Modules/_remote_debugging/binary_io_reader.c @@ -28,6 +28,9 @@ /* Progress callback frequency */ #define PROGRESS_CALLBACK_INTERVAL 1000 +/* Cap per-batch RLE samples to bound the timestamp list (gh-152089) */ +#define MAX_RLE_BATCH_SAMPLES 8192 + /* ============================================================================ * BINARY READER IMPLEMENTATION * ============================================================================ */ @@ -1095,7 +1098,8 @@ binary_reader_replay(BinaryReader *reader, PyObject *collector, PyObject *progre ts->prev_timestamp += delta; /* Start new batch on first sample or status change */ - if (i == 0 || status != batch_status) { + if (i == 0 || status != batch_status + || batch_idx >= MAX_RLE_BATCH_SAMPLES) { if (timestamps_list) { int rc = emit_batch(state, collector, thread_id, interpreter_id, batch_status, ts->current_stack, ts->current_stack_depth, @@ -1105,7 +1109,8 @@ binary_reader_replay(BinaryReader *reader, PyObject *collector, PyObject *progre return -1; } } - timestamps_list = PyList_New(count - i); + /* Append per element; the old alloc(count - i) + trim per batch was O(count^2). */ + timestamps_list = PyList_New(0); if (!timestamps_list) { return -1; } @@ -1118,7 +1123,13 @@ binary_reader_replay(BinaryReader *reader, PyObject *collector, PyObject *progre Py_DECREF(timestamps_list); return -1; } - PyList_SET_ITEM(timestamps_list, batch_idx++, ts_obj); + int append_rc = PyList_Append(timestamps_list, ts_obj); + Py_DECREF(ts_obj); + if (append_rc < 0) { + Py_DECREF(timestamps_list); + return -1; + } + batch_idx++; } /* Emit final batch */