Header files will be on the way.

llvm-svn: 9298

Change the Opcode enum for PHI nodes from "Instruction::PHINode" to "Instruction::PHI" to be more consistent with the other instructions.

llvm-svn: 9269

multiple times.  This reduces the time to construct post-dominance sets a LOT.
For example, optimizing perlbmk goes from taking 12.9894s to 1.4074s.

llvm-svn: 9091

llvm-svn: 9071

llvm-svn: 9063

llvm-svn: 8811

llvm-svn: 8708

llvm-svn: 8706

not have their arguments marked complete

llvm-svn: 8639

llvm-svn: 8638

Do not mark incoming arguments to functions complete if the function is pointed to by a global in the globals graph

llvm-svn: 8637

llvm-svn: 8636

llvm-svn: 8635

llvm-svn: 8634

llvm-svn: 8630

CallSite class.  Now we can represent function calls by invoke instructions
too!

llvm-svn: 8629

llvm-svn: 8628

llvm-svn: 8615

VS: ----------------------------------------------------------------------

llvm-svn: 8611

llvm-svn: 8518

llvm-svn: 8517

llvm-svn: 8478

llvm-svn: 8469

llvm-svn: 8466

unify all exit nodes of a function to compute post-dominance information.
This does not work with functions that have both unwind and return nodes,
because we cannot unify these blocks.  The new implementation is better
anyway. :)

llvm-svn: 8460

llvm-svn: 8442

llvm-svn: 8435

  - make sure to check the indvar type before anything else (efficiency)
  - Make sure to insert the 'add' into the program, even though it'll be
    dead
  - Wrap code at 80 columns
  - Other minor cleanups to reduce indentation level

llvm-svn: 8434

llvm-svn: 8429

llvm-svn: 8427

llvm-svn: 8426

Volatile loads and stores must not be value numbered

llvm-svn: 8398

llvm-svn: 8298

llvm-svn: 8274

  * Increases consistency with other iterators (e.g. df_iterator, po_iterator...)
  * It's shorter
  * We don't name classes by the implementation, we name it for the interface!

llvm-svn: 8273

llvm-svn: 8271

llvm-svn: 8270

llvm-svn: 8269

llvm-svn: 8267

llvm-svn: 8264