simple 'benchmark suite'

tools/measure_perf.py

+#!/usr/bin/env python
+
+import subprocess
+import time
+
+def run_tests(executables, benchmarks):
+    for b in benchmarks:
+        for e in executables:
+            start = time.time()
+            subprocess.check_call(e + [b], stdout=open("/dev/null", 'w'))
+            elapsed = time.time() - start
+
+            print "%s %s: % 4.1fs" % (" ".join(e).rjust(25), b.ljust(35), elapsed)
+
+def main():
+    executables = [["./pyston_release", "-q"]]
+
+    benchmarks = []
+
+    benchmarks += ["../microbenchmarks/%s" % s for s in [
+        ]]
+
+    benchmarks += ["../minibenchmarks/%s" % s for s in [
+        "fannkuch_med.py",
+        "nbody_med.py",
+        "interp2.py",
+        "raytrace.py",
+        "chaos.py",
+        "nbody.py",
+        "fannkuch.py",
+        "spectral_norm.py",
+        ]]
+
+    run_tests(executables, benchmarks)
+
+if __name__ == "__main__":
+    main()