the algorithm instead of hand coded depth first iteration

llvm-svn: 4694

llvm-svn: 4693

llvm-svn: 4692

 values.
X86InstrInfo.def: add LEAVE instruction.

llvm-svn: 4691

llvm-svn: 4690

death when handling moderately sized SCC's, but what can you do

llvm-svn: 4689

llvm-svn: 4688

llvm-svn: 4687

llvm-svn: 4686

llvm-svn: 4685

llvm-svn: 4684

llvm-svn: 4683

llvm-svn: 4682

* Implement the first step of the Globals graph: Deleting nodes from function
  graphs.  In practice, these nodes need to be moved to the globals graph, but
  this will be taken care of later.  Note that the graphs computed right now are
  not strictly correct!

llvm-svn: 4681

llvm-svn: 4680

llvm-svn: 4679

llvm-svn: 4678

llvm-svn: 4677

llvm-svn: 4676

* Significantly improve DEBUG output
* Aggressively fold calls together if we inlined a graph that provides
  call nodes.
* Add a bailout if the current graph has over 200 call nodes in it, this is
  a really whacky case that should never happen.

llvm-svn: 4675

llvm-svn: 4674

llvm-svn: 4673

llvm-svn: 4672

llvm-svn: 4671

llvm-svn: 4670

llvm-svn: 4669

llvm-svn: 4668

llvm-svn: 4667

This slows stuff down a bit, but it should get much better before it gets any
worse.

llvm-svn: 4666

llvm-svn: 4665

llvm-svn: 4664

llvm-svn: 4663

llvm-svn: 4662

llvm-svn: 4661

llvm-svn: 4660

llvm-svn: 4659

llvm-svn: 4658

llvm-svn: 4657

llvm-svn: 4656

llvm-svn: 4655