From a81c5e12bdb416238da68ff46d1a01fe4e38f30e Mon Sep 17 00:00:00 2001 From: kferjaoui Date: Thu, 29 Oct 2026 12:27:44 +0100 Subject: [PATCH] perf: use the shared nsys profile flags in the PixelHistogramCUDA probes Picks up --cuda-event-trace=false on nsys 2025.1+, as the cf probes do. --- python/tests/perf/hist/run_probes.py | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/python/tests/perf/hist/run_probes.py b/python/tests/perf/hist/run_probes.py index e4bb81b5..fde48abc 100644 --- a/python/tests/perf/hist/run_probes.py +++ b/python/tests/perf/hist/run_probes.py @@ -62,8 +62,8 @@ def run_one(dtype, n_bins, batch, label, n_frames, chunk, outdir) -> dict | None rep = outdir / f"probe_{label}" print(f"\n--- {label}: {dtype} frames, {n_bins} bins, batch={batch}, N={n_frames} ---") - prof = [common.NSYS, "profile", "--trace=cuda", "--sample=none", - "--cpuctxsw=none", "--force-overwrite=true", "-o", str(rep), + prof = [common.NSYS, "profile", *common.nsys_profile_flags(), + "-o", str(rep), sys.executable, str(HERE / "nsys_hist_probe.py"), dtype, str(n_frames), str(n_bins), str(batch), str(chunk)] p = subprocess.run(prof, capture_output=True, text=True)