--- /dev/null
+
+ +---Using our compiler in Oct 2011---+
+ +-Actual-------+ | and old/current CVS snapshots |
+ |SPLASH results| | |
+ |from April | |16 April 2011 October 2011 |
+ | | | +-------------
+Benchmark Single OoOJava Single OoOJava Single OoOJava-and-notes (@ denotes rechecked after all CVS commits)
+------------------------------------------------------------------------
+Barnes-Hut 122 N/A 109 9.2 (bad!) 103 117 @ Same flagged sites, conflict graphs are actually the same, but HA! SERIALIZED!! GOOD!
+Crypt 83.3 4.4 79.9 4.3 81 4.71 @ Same flagged sites, old conflict graph empty?!?! Performance same
+KMeans 55.4 4.3 55.5 *** 55.7 4.29 @ Flagged sites print differently (2d arrays?) no old graph to compare, performance same
+Lab(512-) 184 16.3 191.7 48.2 (high?) 210 17.5 @ Same flagged sites, same graphs,
+MergeSort 161 13.0 162 13.0 152 13.4 @ Same: no flagged sites and no conflict graphs, performance same
+MolDyn 70.0 5.0 71 *** 63.5 5.78 @ Same flagged sites, no old conflict graph, matching performance
+Monte 21.7 0.98 24 1.30 25 1.68 @ Fixed a conflict where String cached the hashcode, before that: 19.9 Newer added a flagged site! Newer parallel root has a self fine conflict! PERFORMANCE LIKE SERIALIZED! BAD!!
+Power 31.8 1.58 35.9 7.1 (high?) 33.4 1.96 @ Same flagged sites, conflict graphs same, performance same
+RayTracer 42.8 2.29 38.3 2.28 46.4 2.24 @ Same flagged sites, conflict graphs same except for a heap root name, matching performance
+SOR 235 23.8 266 *** 113.6 24.5 @ Older version didn't compile, but conflict graph looks right and binary performs on target
+Tracking 439 22.2 452 21.5 478 21.4 @ Equivalent flagged sites and conflict graphs
+
+
+***
+Exception in thread "main" java.lang.NullPointerException
+ at Analysis.OoOJava.ConflictGraph.writeGraph(ConflictGraph.java:772)
+ at Analysis.OoOJava.OoOJavaAnalysis.writeConflictGraph(OoOJavaAnalysis.java:1350)
+ at Analysis.OoOJava.OoOJavaAnalysis.<init>(OoOJavaAnalysis.java:273)
+ at Main.Main.main(Main.java:539)