Compile-time tuning
Specialized the dataflow iterators and did a few other minor tweaks.
Showing ~5% compile-time improvement in a single-threaded environment;
less in multi-threaded (presumably because we're blocked by something
else).
Change-Id: I2e2ed58d881414b9fc97e04cd0623e188259afd2
diff --git a/compiler/dex/quick/mir_to_lir.h b/compiler/dex/quick/mir_to_lir.h
index a37ebd1..7e6d21a 100644
--- a/compiler/dex/quick/mir_to_lir.h
+++ b/compiler/dex/quick/mir_to_lir.h
@@ -727,7 +727,7 @@
GrowableArray<LIR*> throw_launchpads_;
GrowableArray<LIR*> suspend_launchpads_;
GrowableArray<LIR*> intrinsic_launchpads_;
- SafeMap<unsigned int, LIR*> boundary_map_; // boundary lookup cache.
+ GrowableArray<LIR*> boundary_map_;
/*
* Holds mapping from native PC to dex PC for safepoints where we may deoptimize.
* Native PC is on the return address of the safepointed operation. Dex PC is for