llvm-svn: 86805

tons of std::string trashing. I plan to move this and other fun string munging
utilities to a StringRefExtras.h at some point if no one beats me to it.

On a synthetic benchmark on x86_64, llvm-gcc actually generates code thats 10%
faster using the StringRef version. gcc miscompiles the synthetic benchmark,
which I'm crossing my fingers and hoping won't happen here. clang compiles the
sythetic benchmark correctly (wootness), but the StringRef version is
slower. Silly clang.

llvm-svn: 86799

parameters. Rather than storing them as either declarations (for the
non-dependent case) or expressions (for the dependent case), we now
(always) store them as TemplateNames. 

The primary change here is to add a new kind of TemplateArgument,
which stores a TemplateName. However, making that change ripples to
every switch on a TemplateArgument's kind, also affecting
TemplateArgumentLocInfo/TemplateArgumentLoc, default template
arguments for template template parameters, type-checking of template
template arguments, etc.

This change is light on testing. It should fix several pre-existing
problems with template template parameters, such as:
  - the inability to use dependent template names as template template
  arguments
  - template template parameter default arguments cannot be
  instantiation

However, there are enough pieces missing that more implementation is
required before we can adequately test template template parameters. 

llvm-svn: 86777

llvm-svn: 86775

llvm-svn: 86774

llvm-svn: 86760

llvm-svn: 86757

 - This is conceptually better since the only thing we want this option to do is
   preserve the internal module as constructed by IRgen, before running any
   passes.

 - This also fixes bugs in -disable-llvm-optzns handling with regards to debug
   info.

llvm-svn: 86691

llvm-svn: 86632

ApplyHeaderSearchOptions now.

llvm-svn: 86617

unless we start implementing command-line switches which override the default
calling convention, so the effect is mostly to silence unknown attribute
warnings.)

llvm-svn: 86571

llvm-svn: 86464

llvm-svn: 86341

initialize HeaderSearch. Not used yet.

llvm-svn: 86338

llvm-svn: 86337

llvm-svn: 86336

filenames.

Also, move InitializePreprocessor to Utils.h.

llvm-svn: 86335

llvm-svn: 86183

llvm-svn: 86177

llvm-svn: 86167

llvm-svn: 86142

invariants (diagnostics with source informations must occur between
{Begin,End}SourceFile).

llvm-svn: 86113

llvm-svn: 86112

llvm-svn: 86111

migrate work in the destructors of PathDiagnosticClients from their
destructors to FlushReports().  The destructors now currently call
FlushReports(); this will be fixed in a subsequent patch.

llvm-svn: 86108

llvm-svn: 86076

llvm-svn: 86048

llvm-svn: 86047

through to indexing.

llvm-svn: 86018

llvm-svn: 86010

Roman Divacky! PR5363

llvm-svn: 85932

Fix buffer overflow in PrintMacroDefinition() by inverting the check to see if the target buffer needs to be resized.  Fixes <rdar://problem/7255377>.

llvm-svn: 85872

llvm-svn: 85828

llvm-svn: 85826

llvm-svn: 85588

types.  Preserve it through template instantiation.  Preserve it through PCH,
although TSTs themselves aren't serializable, so that's pretty much meaningless.

llvm-svn: 85500

llvm-svn: 85481

FunctionDecl::isInlineSpecified/setInlineSpecified.

llvm-svn: 85305

The user visible changes are:

*) Frameworks are only searched on OS X
*) The Cygwin c++ headers are now marked as c++ aware. I am almost
sure that not marking them was a bug.

llvm-svn: 85240

Audit the code for places where it is assumed that every base specifier refers to a RecordType. Add assertions or conditions as appropriate. This fixes another crash in the Apache stdlib vector.

llvm-svn: 85055