llvm-svn: 188093

llvm-svn: 188092

llvm-svn: 188091

llvm-svn: 188090

llvm-svn: 188089

llvm-svn: 188088

llvm-svn: 188087

Since r187945, clang-cl will add the runtime library dependency to
the .obj file.

llvm-svn: 188086

This includes special members (copy/default ctor, copy assign, default
ctor) and template specializations for member function templates.

Good for a 5% decrease (1.80 to 1.71 GB) in size on Clang+LLVM's .dwo
files (when using fission).

llvm-svn: 188085

No functionality change.

llvm-svn: 188084

Avoids unnecessary static constructors.

llvm-svn: 188083

This is necessary to allow Clang to only emit implicit members when
there is code generated for them, rather than whenever they are ODR
used.

llvm-svn: 188082

This trades some complexity in operator== for not introducing static objects
into any functions using recursive directory iterators.

llvm-svn: 188081

Add a check for arrays of unknown bounds to is_destructible.  This fixes http://llvm.org/bugs/show_bug.cgi?id=16839

llvm-svn: 188080

The compiler was warning about using | on a uintptr_t and bool:

  Object/ELFObjectFile.h(131) : warning C4805: '|' : unsafe
  mix of type 'uintptr_t' and type 'bool' in operation

I think the warning might be useful in other cases, so I added
a cast instead of disabling it altogether.

llvm-svn: 188079

- updated RNBDefs.h to allow version numbers to be passed in via preprocessor defines
- update libdebugserver.cpp to compile against latest DNBProcessKill signature

Review: http://llvm-reviews.chandlerc.com/D1331
llvm-svn: 188078

Partial implementation of N3665.  This paper was not voted into the C++1y draft. However I was looking at it and with some experimentation realized that I could partially implement it, and at the same time offer a performance optimization to cout.  I simply added an xsputn override to the cout filebuf.  The override does nothing special at all if there is a non-trivial codecvt installed.  However if the codecvt returns true for always_noconv(), then this function can dump an entire string to fwrite, instead of doing it a character at a time under overflow().  This just makes sense.  I stopped short of a full implementation of N3665 because in order to do so, xsputn would have to allocate a buffer when always_noconv() returned false, and I don't want to go to that expense.

llvm-svn: 188077

llvm-svn: 188076

Patch by Richard Mitton.

llvm-svn: 188075

llvm-svn: 188074

llvm-svn: 188073

llvm-svn: 188072

In Thumb1, only one variant is supported: CPS{effect} {flags}

Thumb2 supports three:
CPS{effect}.W {flags}
CPS{effect} {flags} {mode}
CPS {mode}

Canonically, .W should be used only when ambiguity is present between encodings of different width.
The wide suffix is still accepted for the latter two forms via aliases.

llvm-svn: 188071

llvm-svn: 188070

%n does not increase the input item count. The new code emits writes to %n
arguments even if it has run out of input items.

llvm-svn: 188069

llvm-svn: 188068

The long encoding for Thumb2 unconditional branches is broken.
Additionally, there is no range checking for target operands; as such 
for instructions originating in assembly code, only short Thumb encodings
are generated, regardless of the bitsize needed for the offset.

Adding range checking is non trivial due to the representation of Thumb
branch instructions. There is no true difference between conditional and
unconditional branches in terms of operands and syntax - even unconditional
branches have a predicate which is expected to match that of the IT block
they are in. Yet, the encodings and the permitted size of the offset differ.

Due to this, for any mnemonic there are really 4 encodings to choose for.

The problem cannot be handled in the parser alone or by manipulating td files.
Because the parser builds first a set of match candidates and then checks them
one by one, whatever tablegen-only solution might be found will ultimately be
dependent of the parser's evaluation order. What's worse is that due to the fact
that all branches have the same syntax and the same kinds of operands, that 
order is governed by the lexicographical ordering of the names of operand 
classes...

To circumvent all this, any necessary disambiguation is added to the instruction
validation pass.

llvm-svn: 188067

llvm-svn: 188066

llvm-svn: 188065

llvm-svn: 188064

llvm-svn: 188063

llvm-svn: 188062

Make sure we can properly generate code when the UUID has curly braces
on it, strip the curly braces at the sema layer.

This fixes PR16813.

llvm-svn: 188061

Revert r188055 which reverted r188053. An unrelated change previously snuck in.

llvm-svn: 188060

Summary:
ExprEngine had code which specificaly disabled using CXXTempObjectRegions in
InitListExprs. This was a hack put in r168757 to silence a false positive.

The underlying problem seems to have been fixed in the mean time, as removing
this code doesn't seem to break anything. Therefore I propose to remove it and
solve PR16629 in the process.

Reviewers: jordan_rose

CC: cfe-commits

Differential Revision: http://llvm-reviews.chandlerc.com/D1325

llvm-svn: 188059

llvm-svn: 188058

llvm-svn: 188057

Use isCharInSet from llvm/Support/UnicodeCharRanges.h, added a test for double-width characters in FixIt-hints.

Summary: This is a follow-up to r187837.

Reviewers: gribozavr, jordan_rose

Reviewed By: jordan_rose

CC: cfe-commits

Differential Revision: http://llvm-reviews.chandlerc.com/D1306

llvm-svn: 188056

This commit reverts r188053.

It is breaking the build bots.

llvm-svn: 188055

llvm-svn: 188054