Round 1 of rodinia schedule optimization

• Write better CPU schedules for all rodinia benchmarks.
• Add const-inline pass (inter-procedural constant propagation).
• Add module name to generated function names to avoid linker name collisions.

juno_samples/rodinia/srad/src/cpu.sch

+macro simpl!(X) {
+  ccp(X);
+  simplify-cfg(X);
+  lift-dc-math(X);
+  gvn(X);
+  phi-elim(X);
+  dce(X);
+  infer-schedules(X);
+}
+
+phi-elim(*);
+let loop1 = outline(srad@loop1);
+let loop2 = outline(srad@loop2);
+let loop3 = outline(srad@loop3);
+simpl!(*);
+const-inline(*);
+crc(*);
+slf(*);
+write-predication(*);
+simpl!(*);
+predication(*);
+simpl!(*);
+predication(*);
+simpl!(*);
+fixpoint {
+  forkify(*);
+  fork-guard-elim(*);
+  fork-coalesce(*);
+}
+simpl!(*);
+
+fork-split(*);
+unforkify(*);
+
+gcm(*);