just simple threading.

llvm-svn: 60157

llvm-svn: 60156

llvm-svn: 60153

ImmutableList::getInternalPointer() returns a const ImmutableListImpl<T>* pointer, which must be converted to void* explicitly.

llvm-svn: 60152

llvm-svn: 60151

GDMContext.

llvm-svn: 60150

llvm-svn: 60149

1. Make it fold blocks separated by an unconditional branch.  This enables
   jump threading to see a broader scope.
2. Make jump threading able to eliminate locally redundant loads when they
   feed the branch condition of a block.  This frequently occurs due to
   reg2mem running.
3. Make jump threading able to eliminate *partially redundant* loads when
   they feed the branch condition of a block.  This is common in code with
   lots of loads and stores like C++ code and 255.vortex.

This implements thread-loads.ll and rdar://6402033.

Per the fixme's, several pieces of this should be moved into Transforms/Utils.

llvm-svn: 60148

llvm-svn: 60147

llvm-svn: 60146

llvm-svn: 60145

llvm-svn: 60144

 - Creator function pointers are saved in ManagerRegistry.
 - The Register* class is used to notify ManagerRegistry new module is 
   available.
 - AnalysisManager queries ManagerRegistry for configurable module. Then it
   passes them to GRExprEngine, in turn to GRStateManager.

llvm-svn: 60143

  id<P0>

The intended overloading behavior of these entities isn't entirely
clear, and GCC seems to have some strange limitations (e.g., the
inability to overload on id<P0> vs. id<P1>). We'll want to revisit
these semantics and determine just how Objective-C++ overloading
should really work.

llvm-svn: 60142

llvm-svn: 60141

llvm-svn: 60140

On x86 favors folding short immediate into some arithmetic operations (e.g. add, and, xor, etc.) because materializing an immediate in a register is expensive in turns of code size.
e.g.
movl 4(%esp), %eax
addl $4, %eax

is 2 bytes shorter than

movl $4, %eax
addl 4(%esp), %eax

llvm-svn: 60139

llvm-svn: 60138

llvm-svn: 60137

llvm-svn: 60136

llvm-svn: 60135

llvm-svn: 60134

Support block pointer conversions in C++. I'm storing the test case locally until we can enable blocks in C++

llvm-svn: 60133

  - a table including the IdentifierInfo data
  - an index from persistent IdentifierInfo IDs to indices within this file.
- Enhance PTH generation to write out file map information, mapping inodes to tokens.

llvm-svn: 60132

llvm-svn: 60131

converting a pointer to one Objective-C interface into a pointer to another
Objective-C interface, and conversions with 'id'. The semantics seems
to match GCC, although they seem somewhat ad hoc.

Fixed a few cases where we assumed the C++ definition of isObjectType,
but were getting the C definition, causing failures in trouble with
conversions to void pointers.

llvm-svn: 60130

llvm-svn: 60128

llvm-svn: 60127

llvm-svn: 60125

llvm-svn: 60124

the conditional for the BRCOND statement. For instance, it will generate:

    addl %eax, %ecx
    jo LOF

instead of

    addl %eax, %ecx
    ; About 10 instructions to compare the signs of LHS, RHS, and sum.
    jl LOF

llvm-svn: 60123

llvm-svn: 60122

performance in most cases on the Grawp tester, but does speed some 
things up (like shootout/hash by 15%).  This also doesn't impact 
compile time in a noticable way on the Grawp tester.

It also, of course, gets the testcase it was designed for right :)

llvm-svn: 60120

llvm-svn: 60119

-Change Parser::ParseCXXScopeSpecifier to MaybeParseCXXScopeSpecifier
-Remove Parser::isTokenCXXScopeSpecifier and fold it into MaybeParseCXXScopeSpecifier
-Rename Parser::TryAnnotateScopeToken to TryAnnotateCXXScopeToken and only allow it to be called when in C++

llvm-svn: 60117

Fixes building error on MSVC where passing bool couldn't resolve to one of the overloads.

llvm-svn: 60116

llvm-svn: 60115

Implemented anonymous category (also know as continuation class)
used to override main class's property attribute. This is work in 
propgress.

llvm-svn: 60114

llvm-svn: 60113

llvm-svn: 60112