llvm-svn: 4546

llvm-svn: 4543

llvm-svn: 4538

llvm-svn: 4537

        %A = getelementptr [4 x int]* %M, long 0, long 0
        %B = getelementptr int* %A, long 2

llvm-svn: 4536

  malloc Ty, C
int
  malloc [C x Ty], 1

llvm-svn: 4534

llvm-svn: 4532

into a "oh crap, lets collapse" case

llvm-svn: 4530

llvm-svn: 4526

constant to one particular type.

llvm-svn: 4525

llvm-svn: 4523

llvm-svn: 4521

llvm-svn: 4520

us much more accurate numbers and actually WORKS.

llvm-svn: 4518

llvm-svn: 4517

llvm-svn: 4516

llvm-svn: 4515

respectively.  This is to make presentation easier in the paper.

llvm-svn: 4514

llvm-svn: 4511

llvm-svn: 4510

llvm-svn: 4508

llvm-svn: 4506

* Merge add code into logical code

llvm-svn: 4503

llvm-svn: 4502

llvm-svn: 4500

llvm-svn: 4498

* Turn table into 2d table

llvm-svn: 4496

code a bit

llvm-svn: 4493

llvm-svn: 4492

* Shift amount is always guaranteed to be 8 bits

llvm-svn: 4491

llvm-svn: 4490

llvm-svn: 4489

nodes in the graph are memory objects, which is very nice.  This also greatly
reduces the size and memory footprint for DSGraphs.  For example, the local
DSGraph for llu went from 65 to 13 nodes with this change.  As a side bonus,
dot seems to lay out the graphs slightly better too.  :)

llvm-svn: 4488

llvm-svn: 4486

mod/ref apis yet

llvm-svn: 4485

llvm-svn: 4482

llvm-svn: 4481

llvm-svn: 4479

 Add ISel::visitShiftInst() to instruction select shift instructions.
 Add a comment in visitAdd about how to do 64 bit adds.

X86InstrInfo.def: Add register-to-register move opcodes and shift opcodes.

llvm-svn: 4477

       larger one.
     * Handle the case where we are merging two nodes of different size better.

llvm-svn: 4476