git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@215796 91177308-0d34-0410-b5e6-96231b3b80d8

This reverts commit r215785 / 170ebf4f.

LLD has some StringSavers that need to be updated. One of which takes a
lock and I need to investigate that more closely.

git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@215791 91177308-0d34-0410-b5e6-96231b3b80d8

This fixes PR20671, see the bug for details. In short, ActOnTranslationUnit()
calls DefineUsedVTables() and only then PerformPendingInstantiations(). But
PerformPendingInstantiations() is what does delayed template parsing, so
vtables only references from late-parsed templates weren't marked used.

As a fix, move the SavePendingInstantiationsAndVTableUsesRAII in
PerformPendingInstantiations() up above the delayed template parsing code.
That way, vtables referenced from templates end up in the RAII object, and the
call to DefineUsedVTables() in PerformPendingInstantiations() marks them used.


git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@215786 91177308-0d34-0410-b5e6-96231b3b80d8

There is more cleanup to be done here. Once
llvm::sys::Process::GetArgumentVector is switched over to StringSaver,
we can simplify this code a fair amount.

git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@215785 91177308-0d34-0410-b5e6-96231b3b80d8

git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@215780 91177308-0d34-0410-b5e6-96231b3b80d8

git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@215776 91177308-0d34-0410-b5e6-96231b3b80d8

git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@215774 91177308-0d34-0410-b5e6-96231b3b80d8

I've shied away from ArrayRef'izing CompilerInvocation::CreateFromArgs
in this commit because that is a less localized change.

git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@215773 91177308-0d34-0410-b5e6-96231b3b80d8

A little test case simplification - this could be simplified further,
though there are certainly interesting connections to the if/else
construct so I'm hesitant to remove that entirely though it does appear
somewhat unrelated.

(similar fix to r215766, related to PR19864)

git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@215768 91177308-0d34-0410-b5e6-96231b3b80d8

Thanks to dblaikie for the impetus to look for a better name.

git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@215767 91177308-0d34-0410-b5e6-96231b3b80d8

DebugInfo: Fix PR19864 better - attribute the jump at the end of a range-for loop, to the start of the loop.

This avoids debuggers stepping to strange places (like the last
statement in the loop body, or the first statement in the if).

This is not the whole answer, though - similar bugs no doubt exist in
other loops (patches to follow) and attributing exception handling code
to the correct line is also tricky (based on the previous fix to
PR19864, exception handling is still erroneously attributed to the 'if'
line).

git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@215766 91177308-0d34-0410-b5e6-96231b3b80d8

git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@215764 91177308-0d34-0410-b5e6-96231b3b80d8

This also suggests some refactoring to simplify this code. Basically, a
ton of complexity in this argument handling code comes from the need to
save const char *'s in stable storage for pushing onto argv.
It seems like llvm::cl::StringSaver can be improved to cover all the
needs here.

git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@215761 91177308-0d34-0410-b5e6-96231b3b80d8

git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@215760 91177308-0d34-0410-b5e6-96231b3b80d8

git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@215759 91177308-0d34-0410-b5e6-96231b3b80d8

Reviewers: klimek

Subscribers: klimek, cfe-commits

Differential Revision: http://reviews.llvm.org/D4910

git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@215757 91177308-0d34-0410-b5e6-96231b3b80d8

git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@215756 91177308-0d34-0410-b5e6-96231b3b80d8

git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@215755 91177308-0d34-0410-b5e6-96231b3b80d8

git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@215754 91177308-0d34-0410-b5e6-96231b3b80d8

The core logic in main() is actually pretty simple, but there's lots of
stuff that has been added over time which obscures the flow of the code.
In upcoming patches, I'll be pulling more stuff out of the main
codepath.

I'm open to naming suggestions for these helper functions.

git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@215751 91177308-0d34-0410-b5e6-96231b3b80d8

They can be compared for identity.

git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@215745 91177308-0d34-0410-b5e6-96231b3b80d8

git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@215738 91177308-0d34-0410-b5e6-96231b3b80d8

user provided setter name (as declared in @property attribute declaration).
rdar://18022762


git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@215736 91177308-0d34-0410-b5e6-96231b3b80d8

git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@215731 91177308-0d34-0410-b5e6-96231b3b80d8

git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@215726 91177308-0d34-0410-b5e6-96231b3b80d8

Reviewers: klimek

Subscribers: klimek, cfe-commits

Differential Revision: http://reviews.llvm.org/D4911

git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@215714 91177308-0d34-0410-b5e6-96231b3b80d8

git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@215706 91177308-0d34-0410-b5e6-96231b3b80d8

git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@215705 91177308-0d34-0410-b5e6-96231b3b80d8

git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@215704 91177308-0d34-0410-b5e6-96231b3b80d8

Since __SIZEOF_LONG_LONG__ is always defined as 8 on ARM targets,
there's no point in checking this. NFC.

Patch by Moritz Roth.

git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@215697 91177308-0d34-0410-b5e6-96231b3b80d8

Before:
  typename aaaaaaaaaa<aaaaaa>::aaaaaaaaaaa aaaaaaaaaa<
      aaaaaa>::aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa(bool *aaaaaaaaaaaaaaaaaa,
                                                   bool *aa) {}

After:
  typename aaaaaaaaaa<aaaaaa>::aaaaaaaaaaa
  aaaaaaaaaa<aaaaaa>::aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa(
      bool *aaaaaaaaaaaaaaaaaa, bool *aa) {}

git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@215693 91177308-0d34-0410-b5e6-96231b3b80d8

Before:
  option (MyProto.options) = {
    field_c : "OK" msg_field{field_d : 123}
  };

After:
  option (MyProto.options) = {
    field_c : "OK"
    msg_field{field_d : 123}
  };

(Note that the colon after "msg_field" is optional).

git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@215692 91177308-0d34-0410-b5e6-96231b3b80d8

Summary:
When the CallExpr passed to Sema::ConvertArgumentsForCall has all default parameters, and the number of actual arguments passed is zero, this function will segfault in the call to Call->getLocStart() if the Callee has an invalid getLocStart(), the reason being that since ConvertArgumentsForCall has set the correct number of arguments, but has not filled them in yet, getLocStart() will try to access the first (not yet existent) argument and thus segfaults.

This fixes that by making getLocStart return an invalid source location if the queried argument is NULL rather than segfaulting.

Reviewers: rnk

Reviewed By: rnk

Subscribers: cfe-commits

Differential Revision: http://reviews.llvm.org/D4917

git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@215686 91177308-0d34-0410-b5e6-96231b3b80d8

Previously, any undeclared unqualified id starting a nested name
specifier in a dependent context would have its lookup retried during
template instantiation.  Now we limit that retry hack to methods of a
class with dependent bases.  Free function templates in particular are
no longer affected by this hack.

Also, diagnose this as a Microsoft extension. This has the downside that
template authors may see this warning *and* an error during
instantiation time about this identifier. Fixing that will probably
require formalizing some kind of "delayed" identifier, instead of our
ad-hoc solutions of forming dependent AST nodes when lookup fails.

Based on a patch by Kim Gräsman!

Differential Revision: http://reviews.llvm.org/D4854

git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@215683 91177308-0d34-0410-b5e6-96231b3b80d8

-Wthread-safety umbrella flag, pending updates to documentation.  The flag
works, but is likely to be confusing to existing users of -Wthread-safety.

git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@215679 91177308-0d34-0410-b5e6-96231b3b80d8

git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@215678 91177308-0d34-0410-b5e6-96231b3b80d8

Thread safety analysis: add -Wthread-safety-verbose flag, which adds additional notes that are helpful when compiling statistics on thread safety warnings.

git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@215677 91177308-0d34-0410-b5e6-96231b3b80d8

members from all redefinitions of a class that have them, in case the special
member is defined in one module but only declared in another.


git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@215675 91177308-0d34-0410-b5e6-96231b3b80d8

Fixes PR20662.

git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@215672 91177308-0d34-0410-b5e6-96231b3b80d8

data members and range based for loops.

git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@215671 91177308-0d34-0410-b5e6-96231b3b80d8