llvm-svn: 27699

vla are present in the function. This causes a crash when a leaf function
allocates space on the stack used to store / load with 128-bit SSE
instructions.

llvm-svn: 27698

llvm-svn: 27697

llvm-svn: 27685

llvm-svn: 27684

llvm-svn: 27668

llvm-svn: 27651

llvm-svn: 27647

llvm-svn: 27645

llvm-svn: 27644

llvm-svn: 27643

llvm-svn: 27639

llvm-svn: 27638

llvm-svn: 27637

Clean up and fix various logical ops issues.

llvm-svn: 27633

llvm-svn: 27612

llvm-svn: 27603

__builtin_ia32_loadlv4si, __builtin_ia32_loaddqu, __builtin_ia32_storedqu.

llvm-svn: 27599

later.

llvm-svn: 27591

llvm-svn: 27590

llvm-svn: 27587

llvm-svn: 27575

llvm-svn: 27569

llvm-svn: 27567

llvm-svn: 27565

llvm-svn: 27564

llvm-svn: 27563

llvm-svn: 27562

llvm-svn: 27556

llvm-svn: 27555

llvm-svn: 27554

letting the code generator handle them directly.

llvm-svn: 27539

llvm-svn: 27506

llvm-svn: 27501

llvm-svn: 27497

llvm-svn: 27496

entry point of a function.

llvm-svn: 27494

llvm-svn: 27491

Also MOVPDI2DIrr.

llvm-svn: 27476

- Normalize shuffle nodes so result vector lower half elements come from the
  first vector, the rest come from the second vector. (Except for the
  exceptions :-).
- Other minor fixes.

llvm-svn: 27474