llvm-svn: 54545

X86ISelLowering creates.

llvm-svn: 54544

llvm-svn: 54543

llvm-svn: 54542

llvm-svn: 54541

llvm-svn: 54540

llvm-svn: 54539

llvm-svn: 54538

llvm-svn: 54537

llvm-svn: 54536

Reduce the entries in a phi before testing it for deadness, because removing the entries might make it dead.

llvm-svn: 54535

llvm-svn: 54534

- Move checking from MergeVarDecl->FinializeDeclaratorGroup. Since MergeVarDecl is called before the initializer is attacted, it can't be done there (this removes a long standing FIXME).
- Add Sema::isTentativeDefinition() and Sema::CheckForFileScopedRedefinitions().
- Remove FIXME's and touch-up test case.

Still some more work to do (forthcoming)...

llvm-svn: 54533

llvm-svn: 54530

llvm-svn: 54529

up the passmgr by avoiding useless work.

llvm-svn: 54528

llvm-svn: 54527

llvm-svn: 54526

llvm-svn: 54525

llvm-svn: 54524

llvm-svn: 54523

llvm-svn: 54522

llvm-svn: 54521

It's not legal to output a GV in a coalesced section if it's used in an ARM PIC relative constantpool.

llvm-svn: 54519

llvm-svn: 54518

is narrower than the pointer. This testcase emits:

  .byte (((17) - 16) & 255)

llvm-svn: 54517

llvm-svn: 54516

llvm-svn: 54515

llvm-svn: 54514

Unbreak SerializationTest and the Rewriter by doing the work in HandleTranslationUnit instead of the destructor.

llvm-svn: 54513

llvm-svn: 54512

llvm-svn: 54511

llvm-svn: 54510

llvm-svn: 54509

llvm-svn: 54508

llvm-svn: 54507

I had a lot of simple local codegen tests and they are now ready to be placed
in test/CodeGen. 

llvm-svn: 54506

other from Python.

llvm-svn: 54505

The motivation is that clients may either:

(a) query the ASTConsumer object after AST parsing to collect data/etc.
(b) reuse the ASTConsumer.

llvm-svn: 54502

llvm-svn: 54501