llvm-svn: 52522

Change around the data structures used to store availability sets, resulting in a GVN+PRE that is faster that GVN alone was before.

llvm-svn: 52521

propagating constants across aggregate return values when
insertvalue and extractvalue are used.

llvm-svn: 52520

return statements and aggregate returns so that it handles both
correctly.

llvm-svn: 52519

llvm-svn: 52518

llvm-svn: 52517

llvm-svn: 52516

object.

llvm-svn: 52515

llvm-svn: 52514

llvm-svn: 52512

llvm-svn: 52511

llvm-svn: 52508

llvm-svn: 52507

once benchmarking is completed.

llvm-svn: 52506

llvm-svn: 52505

Remove an extra space at the end of a line.

llvm-svn: 52504

I choose to remove extraneous whitespace at end of lines as a semantic
nop for the test.

llvm-svn: 52503

llvm-svn: 52502

have complex RValues yet, so this logic is only fully implemented when __imag__
and __real__ are used on non-complex types.

llvm-svn: 52501

using BasicBlock::getInstList.

llvm-svn: 52500

This fixes a failure on povray.

llvm-svn: 52499

directly.

llvm-svn: 52498

llvm-svn: 52497

GVN expects that all inputs which to an instruction fall somewhere in the value
hierarchy, which isn't true for these.

llvm-svn: 52496

llvm-svn: 52495

llvm-svn: 52494

llvm-svn: 52493

llvm-svn: 52492

llvm-svn: 52491

using getOperand() directly. This makes things work with invoke instructions as
well.

llvm-svn: 52489

this still compiles on windows - I can't test!

llvm-svn: 52488

llvm-svn: 52487

shuffle could be skipped.  The check is invalid because the loop index i 
doesn't correspond to the element actually inserted. The correct check is
already done a few lines earlier, for whether the element is already in 
the right spot, so this shouldn't have any effect on the codegen for 
code that was already correct.

llvm-svn: 52486

llvm-svn: 52485

llvm-svn: 52483

llvm-svn: 52482

llvm-svn: 52481

llvm-svn: 52480

llvm-svn: 52479

llvm-svn: 52478