Implement irreducible loop support in optimizing.

So we don't fallback to the interpreter in the presence of
irreducible loops.

Implications:
- A loop pre-header does not necessarily dominate a loop header.
- Non-constant redundant phis will be kept in loop headers, to
  satisfy our linear scan register allocation algorithm.
- while-graph optimizations, such as gvn, licm, lse, and dce
  need to know when they are dealing with irreducible loops.

Change-Id: I2cea8934ce0b40162d215353497c7f77d6c9137e

1 files changed, 0 insertions, 1 deletions

diff --git a/compiler/optimizing/find_loops_test.cc b/compiler/optimizing/find_loops_test.cc
index 9b0eb70742..4770fa2eca 100644
--- a/compiler/optimizing/find_loops_test.cc
+++ b/compiler/optimizing/find_loops_test.cc
@@ -33,7 +33,6 @@ static HGraph* TestCode(const uint16_t* data, ArenaAllocator* allocator) {
   const DexFile::CodeItem* item = reinterpret_cast<const DexFile::CodeItem*>(data);
   builder.BuildGraph(*item);
   graph->BuildDominatorTree();
-  graph->AnalyzeNaturalLoops();
   return graph;
 }