it is definitely profitable to tail duplicate indirect branches for x86.
This is likely to be true to various degrees for all modern x86 processors.

llvm-svn: 89865

Fix a thinko where we weren't always performing unary conversions on the switch condition, fixing PR5612

llvm-svn: 89864

llvm-svn: 89863

llvm-svn: 89862

llvm-svn: 89861

llvm-svn: 89860

llvm-svn: 89859

llvm-svn: 89858

llvm-svn: 89856

llvm-svn: 89855

llvm-svn: 89854

llvm-svn: 89853

#ifndef foo
#define foo(parm1)
#endif

foo(parm1) parm1 ...

#undef foo

llvm-svn: 89851

llvm-svn: 89850

llvm-svn: 89849

API change Path::isSpecialFile to Path::isRegularFile, improve semantics in regards to comments from 89765 post review.

llvm-svn: 89848

still parse the body of the switch to try to avoid spurious
diagnostics. Fixes PR5606.

llvm-svn: 89847

llvm-svn: 89846

Don't crash when we re-use a template specialization node for an explicit instantiation. lib/Support/CommandLine.cpp is our test case

llvm-svn: 89845

llvm-svn: 89844

llvm-svn: 89843

llvm-svn: 89842

llvm-svn: 89841

llvm-svn: 89840

llvm-svn: 89839

The attributes are currently ignored.

llvm-svn: 89837

llvm-svn: 89836

llvm-svn: 89835

llvm-svn: 89834

llvm-svn: 89833

statements, e.g., 

  for(; X x = X(); ) { ... }  

Daniel or Anders, please review!

llvm-svn: 89832

Add really basic support for blocks in the retain/release checker.  For now, anytime we pass a tracked object to a block call we stop tracking it.

llvm-svn: 89831

llvm-svn: 89830

llvm-svn: 89829

Split CodeTextRegion into FunctionTextRegion and BlockTextRegion.  This a precursor to having basic static analysis support for blocks.

llvm-svn: 89828

to build casted expression-list AST to Sema.

llvm-svn: 89827

using two swc/lwc instead of sdc/ldc.

llvm-svn: 89826

llvm-svn: 89824

unnecessary save/restore.

llvm-svn: 89823

llvm-svn: 89822