findRegisterUseOperandIdx, findRegisterDefOperandIndx. Fix some naming inconsistencies.

llvm-svn: 47927

llvm-svn: 47924

llvm-svn: 47922

llvm-svn: 47918

class (cosmetic).  First piece of byval implementation;
this doesn't work yet.  No functional change.

llvm-svn: 47917

llvm-svn: 47916

llvm-svn: 47915

llvm-svn: 47911

llvm-svn: 47909

llvm-svn: 47908

Update test to use multiple return value directly, instead of relying on -sretpromotion.

llvm-svn: 47907

llvm-svn: 47906

llvm-svn: 47904

llvm-svn: 47903

llvm-svn: 47900

llvm-svn: 47899

llvm-svn: 47897

new function and new call sites.

llvm-svn: 47896

llvm-svn: 47895

llvm-svn: 47892

llvm-svn: 47890

llvm-svn: 47889

llvm-svn: 47888

llvm-svn: 47886

The basic idea is that all these algorithms are computing the longest paths from the root node or to the exit node. Therefore the existing implementation that uses and iterative and potentially
exponential algorithm was changed to a well-known graph algorithm based on dynamic programming. It has a linear run-time.

llvm-svn: 47884

Refactor ExpandConstantFP so it can optimize load from constpool of types larger than f64 into extload from smaller types.

llvm-svn: 47883

llvm-svn: 47882

llvm-svn: 47881

llvm-svn: 47878

llvm-svn: 47877

llvm-svn: 47875

llvm-svn: 47874

llvm-svn: 47872

llvm-svn: 47871

- Cleaned up how the prologue-epilogue inserter loops over the instructions.
- Instead of restarting the processing of an instruction if we remove an
  implicit kill, just update the end iterator and make sure that the iterator
  isn't incremented.

llvm-svn: 47870

llvm-svn: 47869

llvm-svn: 47868

llvm-svn: 47867

llvm-svn: 47866

llvm-svn: 47865