Teach the inline spiller to attempt folding a load instruction into its single
use before rematerializing the load.

This allows us to produce:

    addps	LCPI0_1(%rip), %xmm2

Instead of:

    movaps	LCPI0_1(%rip), %xmm3
    addps	%xmm3, %xmm2

Saving a register and an instruction. The standard spiller already knows how to
do this.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@122133 91177308-0d34-0410-b5e6-96231b3b80d8
diff --git a/lib/CodeGen/LiveRangeEdit.h b/lib/CodeGen/LiveRangeEdit.h
index ad248bf..37b5827 100644
--- a/lib/CodeGen/LiveRangeEdit.h
+++ b/lib/CodeGen/LiveRangeEdit.h
@@ -117,6 +117,12 @@
                             const TargetInstrInfo&,
                             const TargetRegisterInfo&);
 
+  /// markRematerialized - explicitly mark a value as rematerialized after doing
+  /// it manually.
+  void markRematerialized(VNInfo *ParentVNI) {
+    rematted_.insert(ParentVNI);
+  }
+
   /// didRematerialize - Return true if ParentVNI was rematerialized anywhere.
   bool didRematerialize(VNInfo *ParentVNI) const {
     return rematted_.count(ParentVNI);