tools/x264_perf_O3: support to disable loop vectorize optimization for get_ref

tools/x264_perf_O3.py

@@ -1,19 +1,38 @@
 from pathlib import Path
 from subprocess import run
 
-from speccpu import get_benchspec_dir, x264
+import click
+
+from speccpu import create_spec_env, get_benchspec_dir, x264
 from speccpu.flaglib.perf import record as recordflags
 
 
-def main():
+@click.command()
+@click.option(
+    "--disable-get-ref-vectorize",
+    type=bool,
+    default=False,
+    help="Enable or disable MC vectorization",
+)
+@click.option(
+    "--spec",
+    type=Path,
+    help="Path to SPEC CPU directory",
+)
+def main(disable_get_ref_vectorize: bool, spec: Path):
     x264_run_id = "c56635a4-9f49-4d2b-be79-38d9e08097c5"
     x264_outputroot = Path("/home/lyc/work-ts/local/specOutput") / x264_run_id
 
-
     x264_dir = get_benchspec_dir(x264_outputroot) / "625.x264_s"
 
     x264_run = x264_dir / "run" / "run_base_refspeed_mytest-m64.0000"
+    x264_exe = x264_run / "x264_s_base.mytest-m64"
 
+    if disable_get_ref_vectorize:
+        mc_src = x264.MCBuilder().with_disabled_vectorize_get_ref().build()
+        x264_exe = x264.recompile_mc_src(
+            mc_src, x264_dir / "build", create_spec_env(spec)
+        )
 
     run(
         [
@@ -21,7 +40,7 @@ def main():
             "record",
             *recordflags(
                 [
-                    str(x264_run / "x264_s_base.mytest-m64"),
+                    str(x264_exe),
                     *x264.specinvoke_args["ref"],
                 ],
                 output=str(Path.cwd() / f"x264-{x264_run_id}perf"),
@@ -31,5 +50,6 @@ def main():
         cwd=x264_run,
     )
 
+
 if __name__ == "__main__":
     main()