[SparcV9]: Do not emit .register directives for global registers that are clobbered by calls but not used in the function itself.

llvm-svn: 195574

llvm-svn: 195573

llvm-svn: 195572

Secondary allocator is mmap-based, so the memory is already zeroed.

llvm-svn: 195571

Performance improvement. Also, the allocator was using CompactSizeClassMap for
no good reason, so I switched it to DefaultSizeClassMap.

llvm-svn: 195570

to what is needed for constant islands. The prescan method for Mips16 constant
islands will eventually go away. It is only temporary and should be done
earlier when the instructions are first created or from the DAG. If we keep
it here we need to handle better the situation where constant islands
is called multiple times since don't want to prescan more than once.

llvm-svn: 195569

llvm-svn: 195567

llvm-svn: 195566

I had to move some code and I moved a declaration forward past it's first use
in the function but by nutty coincidence there was another variable of the same
name and type and  with completely unrelated function that was declared globally
in the class so no compilation error ensued.
It required some unusual conditions for it to even matter. Caused test
case casts.c in test-suite to fail during compilation with a duplicate 
symbol error. I would have noticed it during final code review for this port.

llvm-svn: 195565

llvm-svn: 195564

llvm-svn: 195563

This matches other empty() container functions in LLVM.

No actual usage problems discovered in this instance.

llvm-svn: 195562

llvm-svn: 195561

This refines some diagnostics and reduces some boilerplate checking logic.

llvm-svn: 195560

llvm-svn: 195559

Full discourse at:

  http://lists.cs.uiuc.edu/pipermail/cfe-commits/Week-of-Mon-20131104/092514.html
  http://lists.cs.uiuc.edu/pipermail/llvmdev/2013-November/068124.html

Patch by Dimitry Andric and Alexey Dokuchaev!

llvm-svn: 195558

the installed SDK to using the current OS installed headers/libraries.
This change is to address the removal of the Python framework
from the Mac OS X 10.9 (Mavericks) SDK, and is the recommended
workaround via https://developer.apple.com/library/mac/technotes/tn2328/_index.html

llvm-svn: 195557

[CodeGen] If there is a function definition with duplicate mangled name, emit an error instead of asserting.

rdar://15522601 & http://llvm.org/PR18031

llvm-svn: 195556

This fixes llvm.org/PR17241.

llvm-svn: 195555

Clang knows how to use the gnu assembler directly from doing so on linux and
hurd. The existing support worked out of the box on cygwin and mingw and I was
able to bootstrap clang with it in both systems (with pending patches for the
new mingw abi, but that is independent of the assembler).

llvm-svn: 195554

This fixes llvm.org/PR17979.

Before:
  void f() { g(/*aaa=*/x, /*bbb=*/ !y); }

After:
  void f() { g(/*aaa=*/x, /*bbb=*/!y); }

llvm-svn: 195553

With Style.BreakBeforeBinaryOperators, clang-format breaks incorrectly.
This fixes llvm.org/PR17994.

Before:
  return boost::fusion::at_c<0>(iiii).second == boost::fusion::at_c
                                                <1>(iiii).second;

After:
  return boost::fusion::at_c<0>(iiii).second ==
         boost::fusion::at_c<1>(iiii).second;

llvm-svn: 195552

llvm-svn: 195551

Also disallow breaking between "@" and "{" or "[".

Before:
  - (NSAttributedString *)attributedStringForSegment:(NSUInteger)segment
                                               index:(NSUInteger)index
                                          attributes:(NSDictionary *)attributes
                                  nonDigitAttributes:(NSDictionary *)
      nonDigitAttributes;
  [mailComposeViewController
     setToRecipients:@
     [ NSBundle.mainBundle.infoDictionary[@"ABBFeedbackEmail"] ]];

After:
  - (NSAttributedString *)attributedStringForSegment:(NSUInteger)segment
                                               index:(NSUInteger)index
                                          attributes:(NSDictionary *)attributes
                                  nonDigitAttributes:
                                      (NSDictionary *)nonDigitAttributes;
  [mailComposeViewController
      setToRecipients:
          @[ NSBundle.mainBundle.infoDictionary[@"ABBFeedbackEmail"] ]];

This fixes llvm.org/PR18030.

llvm-svn: 195550

llvm-svn: 195549

The following patch enables the non-fragile ABI for the ObjFW runtime.

However, I noticed that it is not possible anymore to disable it with
-fno-objc-nonfragile-abi like it was before. I think this functionality should
be restored, but I guess this is not in scope for 3.4 anymore.

Patch by Jonathan Schleifer!

llvm-svn: 195547

Add a severe penalty for not using column layout for braced lists. If
there are solutions with column layout, these are generally preferable
over bin-packed solutions.

Before:
  std::vector<MyValues> aaaaaaaaaaaaaaaaaaa{ aaaaaaa, aaaaaaaaaa, aaaaa,
                                             aaaaaaaaaaaaaaa, aaa, aaaaaaaaaa, a,
                                             aaaaaaaaaaaaaaaaaaaaa, aaaaaaaaaaaa,
                                             aaaaaaaaaaaaaaaaaaa +
                                                 aaaaaaaaaaaaaaaaaaa,
                                             aaaaaaa, a };
After:
  std::vector<MyValues> aaaaaaaaaaaaaaaaaaa{
    aaaaaaa,      aaaaaaaaaa,
    aaaaa,        aaaaaaaaaaaaaaa,
    aaa,          aaaaaaaaaa,
    a,            aaaaaaaaaaaaaaaaaaaaa,
    aaaaaaaaaaaa, aaaaaaaaaaaaaaaaaaa + aaaaaaaaaaaaaaaaaaa,
    aaaaaaa,      a
  };

llvm-svn: 195546

llvm-svn: 195545

SmallPtrSet.cpp use different methods if SmallPtrSet.h is included
in C++11 mode.  Building llvm in C++03 mode and lldb in C++11 mode
resulted in a link-time failure with the C++11-mode-specific method
not being found in the llvm build.

llvm-svn: 195544

module. Use the marker to diagnose cases where we try to transition between
submodules when not at the top level (most likely because a closing brace was
missing at the end of a header file, but is also possible if submodule headers
attempt to do something fundamentally non-modular, like our .def files).

llvm-svn: 195543

llvm-svn: 195542

Improved the detection of a valid GDB server where we actually can connect to a socket, but then it doesn't read or write anything (which happens with some USB mux software).

Host::MakeDirectory() now can make as many intermediate directories as needed.

The testsuite now has very initial support for remote test suite running. When running on a remote platform, the setUp function for the test will make a new directory and select it as the working directory on the remote host. 

Added a common function that can be used to create the short option string for getopt_long calls.

llvm-svn: 195541

This will allow the completer to order results by relevance.

Differential Revision: http://llvm-reviews.chandlerc.com/D2209

llvm-svn: 195540

The new names will be more appropriate when the objects are taught to return
type information for the code completer.

Differential Revision: http://llvm-reviews.chandlerc.com/D2208

llvm-svn: 195539

proxy. This lets a function pass query a module analysis manager.
However, the interface is const to indicate that only cached results can
be safely queried.

With this, I think the new pass manager is largely functionally complete
for modules and analyses. Still lots to test, and need to generalize to
SCCs and Loops, and need to build an adaptor layer to support the use of
existing Pass objects in the new managers.

llvm-svn: 195538

for a TestModuleAnalysis.

llvm-svn: 195537

This avoids the need for an extra list of SkeletonCUs and associated
cleanup while staging things to be cleaner for further type unit
improvements.

Also hopefully fixes a memory leak introduced in r195166.

llvm-svn: 195536

We are going to drop debug info without a version number or with a different
version number, to make sure we don't crash when we see bitcode files with
different debug info metadata format.

Make tests more robust by removing hard-coded metadata numbers in CHECK lines.

llvm-svn: 195535

The looked-up matchers will be used during code completion.

Differential Revision: http://llvm-reviews.chandlerc.com/D2207

llvm-svn: 195534

This is still an experimental attribute, but I wanted it in tree
for review.  It may still get yanked.

This attribute can only be applied to a class @interface, not
a class extension or category.  It does not change the type
system rules for Objective-C, but rather the implementation checking
for Objective-C classes that explicitly conform to a protocol.
During protocol conformance checking, clang recursively searches
up the class hierarchy for the set of methods that compose
a protocol.  This attribute will cause the compiler to not consider
the methods contributed by a super class, its categories, and those
from its ancestor classes.  Thus this attribute is used to force
subclasses to redeclare (and hopefully re-implement) methods if
they decide to explicitly conform to a protocol where some of those
methods may be provided by a super class.

This attribute intentionally leaves out properties, which are associated
with state.  This attribute only considers methods (at least right now)
that are non-property accessors.  These represent methods that "do something"
as dictated by the protocol.  This may be further refined, and this
should be considered a WIP until documentation gets written or this
gets removed.

llvm-svn: 195533