the second (store) instruction in SpillSlotToUsesMap
consistently.  I don't think this matters functionally,
but it's cleaner and Evan wants it this way.

llvm-svn: 85463

llvm-svn: 85462

Switch a few ugly switch-on-string-literal constructs to use the new
llvm::StringSwitch.

llvm-svn: 85461

llvm-svn: 85460

cascading "ifs" of strcmps/memcmps.

llvm-svn: 85459

llvm-svn: 85458

llvm-svn: 85457

llvm-svn: 85456

--- Reverse-merging r85338 into '.':
U    lib/CodeGen/SimpleRegisterCoalescing.cpp
U    lib/CodeGen/SimpleRegisterCoalescing.h

llvm-svn: 85454

llvm-svn: 85453

llvm-svn: 85452

llvm-svn: 85451

llvm-svn: 85450

llvm-svn: 85449

llvm-svn: 85448

class template partial specializations of member templates. Also,
fixes a silly little bug in the marking of "used" template parameters
in member templates. Fixes PR5236.

llvm-svn: 85447

using the new LLVM support for this.  This is temporarily hiding
behind horrible and ugly #ifdefs until the time when the optimizer
is stable (hopefully a week or so).  Until then, lets make it "opt in" :)

llvm-svn: 85446

llvm-svn: 85445

Unused ivars checker: also check methods in categories that are defined in the same translation unit.  Fixes <rdar://problem/6260004>.

llvm-svn: 85442

common tail, except when the OptimizeForSize function attribute is present.
Radar 7338114.

llvm-svn: 85441

llvm-svn: 85440

to spill after all, we weren't handling 2-instruction
spill sequences correctly (PPC Altivec).  We need to
remove the store in this case.  Removing the other
instruction(s) would be goodness but is not needed for
correctness, and isn't done here.  7331562.

llvm-svn: 85437

llvm-svn: 85436

that don't have sizes.

llvm-svn: 85435

llvm-svn: 85434

llvm-svn: 85432

I'm going to redo this using the OptimizeForSize function attribute.

llvm-svn: 85426

Remove _clang_initCXLookupHint() and _clang_getCursorWithHint(). Related to <rdar://problem/7310688>.

Localize the optimization to ResolveLocationInAST(). The last valid AST location is now stored with ASTUnit. There still isn't optimal, however it's an improvement (with a much cleaner API). Having the client manage an "hint" is error prone and complex.

I wanted to land the major changes before finishing up the optimizations. 

llvm-svn: 85425

llvm-svn: 85424

categories as well (WIP). No functionality change yet.

llvm-svn: 85423

llvm-svn: 85422

ArraySize * ElementSize
ElementSize * ArraySize
ArraySize << log2(ElementSize)
ElementSize << log2(ArraySize)

Refactor isArrayMallocHelper and delete isSafeToGetMallocArraySize, so that there is only 1 copy of the malloc array determining logic.
Update users of getMallocArraySize() to not bother calling isArrayMalloc() as well.

llvm-svn: 85421

constructor. WIP.

llvm-svn: 85420

llvm-svn: 85419

which should come from the variable and wasn't correct for arrays in any case.
No change in functionality.

llvm-svn: 85415

This fixes parallel build with clang on glibc platforms.

llvm-svn: 85414

llvm-svn: 85412

opcode and operand with a tab.  Check for these instructions in the usual
places.

llvm-svn: 85411

llvm-svn: 85410

to issue involved.

llvm-svn: 85409