test/bench/shootout/timing.log: update after recent compiler changes

Moving panic out of line speeds up fannkuch almost a factor of two.
Changes to bitwhacking code affect mandelbrot badly.

R=golang-dev, bradfitz, rsc, r
CC=golang-dev
https://golang.org/cl/6258056

diff --git a/test/bench/shootout/timing.log b/test/bench/shootout/timing.log
index 7ad781fc2244b6be6534e71afb5fbe74deb781f6..467f50a780dc1837a96073531d6e74f52764cf1f 100644 (file)
--- a/test/bench/shootout/timing.log
+++ b/test/bench/shootout/timing.log
@@ -993,3 +993,32 @@ chameneos 6000000
        gccgo -O2 chameneosredux.go     7.73u 7.53s 15.30r
        gc chameneosredux       6.51u 0.00s 6.53r # - 14%
 
+# After http://codereview.appspot.com/6248049, moving panicindex
+# calls out of line (putting the likely code into a single path and shortening
+# loops). Significant changes since the last run (note: some are slower for
+# unrelated and as yet undiagnosed reasons):
+
+nbody -n 50000000
+       gc nbody        19.10u 0.01s 19.19r # -12%
+       gc_B nbody      19.19u 0.00s 19.23r # -12%
+
+binary-tree 15 # too slow to use 20
+       gc binary-tree  1.49u 0.01s 1.51r # -19%
+       
+fannkuch 12
+       gc fannkuch     60.79u 0.00s 60.92r # -41%
+       gc fannkuch-parallel    183.51u 0.01s 51.75r # -14%
+       gc_B fannkuch   51.68u 0.00s 51.79r # -3%
+
+k-nucleotide 1000000
+       gc k-nucleotide 9.74u 0.04s 9.80r # +6%
+       gc k-nucleotide-parallel        9.89u 0.05s 3.59r # +1%
+       gc_B k-nucleotide       9.39u 0.02s 9.43r # +2%
+
+mandelbrot (much slower, due to unrelated http://codereview.appspot.com/6209077)
+       gc mandelbrot   100.98u 0.00s 101.20r # +65%
+       gc_B mandelbrot 100.90u 0.01s 101.17r # +65%
+
+meteor 2098
+       gc meteor-contest       0.13u 0.00s 0.13r # -13%
+       gc_B meteor-contest     0.13u 0.00s 0.13r # -7%