llvm-svn: 67326

uncaught language cases a compile warning instead of a runtime error.

llvm-svn: 67325

llvm-svn: 67324

llvm-svn: 67323

llvm-svn: 67321

diagnostics following those notes. 

Make exceeding the template instantiation depth a fatal error.

Thanks to Daniel for pointing out the problem!

llvm-svn: 67320

llvm-svn: 67319

llvm-svn: 67318

llvm-svn: 67316

llvm-svn: 67315

dot-syntax expression after earching the list of protocols
in the qualified-id, must keep searching the protocol list
of each of the protocols in the list.

llvm-svn: 67314

  F f;

where F is a typedef of a function type, then the function "f" has a
prototype. This is a slight tweak to Chris's suggested fix in
PR3817. Fixes PR3817 and PR3840.

llvm-svn: 67313

the inliner; prevents nondeterministic behavior
when the same address is reallocated.
Don't build call graph nodes for debug intrinsic calls;
they're useless, and there were typically a lot of them.

llvm-svn: 67311

 - Apologies for commits w/o test cases; they are coming.

llvm-svn: 67310

the set of blocks in which values are used, the set in which
values are live-through, and the set in which values are
killed. For the live-through and killed sets, conservative
approximations are used.

llvm-svn: 67309

dependent qualified-ids such as

  Fibonacci<N - 1>::value

where N is a template parameter. These references are "unresolved"
because the name is dependent and, therefore, cannot be resolved to a
declaration node (as we would do for a DeclRefExpr or
QualifiedDeclRefExpr). UnresolvedDeclRefExprs instantiate to
DeclRefExprs, QualifiedDeclRefExprs, etc.

Also, be a bit more careful about keeping only a single set of
specializations for a class template, and instantiating from the
definition of that template rather than a previous declaration. In
general, we need a better solution for this for all TagDecls, because
it's too easy to accidentally look at a declaration that isn't the
definition.

We can now process a simple Fibonacci computation described as a
template metaprogram.

llvm-svn: 67308

llvm-svn: 67307

and was deleting Instructions without clearing the
corresponding map entry.  This led to nondeterministic
behavior if the same address got allocated to another
Instruction within a short time.

llvm-svn: 67306

Fix regression in 'tblgen -gen-clang-diags-defs': Emit the diagnostic kind instead of "DIAGNOSTICCONTROLLED".

llvm-svn: 67305

is the first in its block.  This is PR3842.

llvm-svn: 67304

llvm-svn: 67302

 - PR3818.

llvm-svn: 67297

llvm-svn: 67296

llvm-svn: 67295

diagnostic (to suppress more generic unused warning).

llvm-svn: 67294

are forwarded to GCC.
 - The later is unfortunate, as it prevents us from generally warning
   about anything interesting on platforms that use a generic
   toolchain. However, we can't do much better without significantly
   complicating things, and generally we should have proper tool chain
   definitions.

llvm-svn: 67293

 - Make InputInfo a variant of filename, pipe, input argument,
   nothing.

 - Leave a FIXME in InputInfo that this should be revisited.

llvm-svn: 67292

was hoping to clean up in the rewrite, but I don't see it yet.

llvm-svn: 67291

llvm-svn: 67290

llvm-svn: 67289

either.

llvm-svn: 67288

llvm-svn: 67287

in selectiondag patterns.  This is required for the upcoming shuffle_vector rewrite,
and as it turns out, cleans up a hack in the Alpha instruction info.

llvm-svn: 67286

llvm-svn: 67285

  struct N::M::foo

llvm-svn: 67284

QualifiedNameType and QualifiedDeclRefExpr. We now keep track of the
exact nested-name-specifier spelling for a QualifiedDeclRefExpr, and
use that spelling when printing ASTs. This fixes PR3493.

llvm-svn: 67283

llvm-svn: 67280

llvm-svn: 67279

llvm-svn: 67269

specialization names. This way, we keep track of sugared types like

  std::vector<Real>

I believe we are now using QualifiedNameTypes everywhere we can. Next
step: QualifiedDeclRefExprs.

llvm-svn: 67268