Begin implementation of an inline spiller.

InlineSpiller inserts loads and spills immediately instead of deferring to
VirtRegMap. This is possible now because SlotIndexes allows instructions to be
inserted and renumbered.

This is work in progress, and is mostly a copy of TrivialSpiller so far. It
works very well for functions that don't require spilling.

llvm-svn: 107227
diff --git a/llvm/lib/CodeGen/CMakeLists.txt b/llvm/lib/CodeGen/CMakeLists.txt
index d7f6737..0501350 100644
--- a/llvm/lib/CodeGen/CMakeLists.txt
+++ b/llvm/lib/CodeGen/CMakeLists.txt
@@ -13,6 +13,7 @@
   GCMetadataPrinter.cpp
   GCStrategy.cpp
   IfConversion.cpp
+  InlineSpiller.cpp
   IntrinsicLowering.cpp
   LLVMTargetMachine.cpp
   LatencyPriorityQueue.cpp