'operator delete' or 'operator delete[]' is an explicit exception
specification. Therefore we should diagnose 'void operator delete(void*)'
instead of 'void operator delete(void*) noexcept'.

This diagnostic remains an ExtWarn, since in practice people don't always
include the exception specification in such a declaration.

llvm-svn: 201002

llvm-svn: 201001

llvm-svn: 201000

llvm-svn: 200999

llvm-svn: 200998

Now that the command interpreter runs in a separate thread in Xcode, we need to lock the target API lock instead of trying to lock it.

What was happening was:
1 - Xcode ran and stopped and was doing work on thread 2
2 - Users would type something in Xcode console on thread 1
3 - thread 3 would be running command interpreter thread and try to execute command but get "failed to get API lock" error for any command that wanted the target API lock (like "expression")

<rdar://problem/15775016>

llvm-svn: 200997

When a user says

type formatter add ... unsigned int

he most probably means to deal with the "unsigned int" type. However, given how the LLDB command parser works, that command will try to add the formatter to the TWO types 'unsigned' AND 'int'

Since this is unlikely to be what the user wants, warn about it, and suggest they can use quotes to override the debugger's understanding

llvm-svn: 200996

llvm-svn: 200995

This is a small simplification and a small step in fixing pr18743 since
private functions on MachO should be using a 'l' prefix.

llvm-svn: 200994

llvm-svn: 200993

llvm-svn: 200992

values whose size differs from the register's
size.

llvm-svn: 200991

llvm-svn: 200990

Stores of <4 x i64> do work (although they do expand to 4 stores
instead of 2), but 3 x i64 vectors fail to select.

llvm-svn: 200989

llvm-svn: 200988

Hopefully the last tweak needed to get this test working everywhere.

Remove matching of the prefix of sys_header.h, which was never the point
of the test anyway.  This avoids dealing with path separators.

llvm-svn: 200987

internal discussions. // rdar://16006401

llvm-svn: 200986

Fix a bug where we will crash if we have a class "B" that is contained inside a class "A" in the DWARF where "A" is a forward declaration.

<rdar://problem/14673945>
<rdar://problem/15682781>

llvm-svn: 200985

Provide a filter for libc++ std::atomic<T>
This just hides some implementation clutter and promotes the actual content to only child status

llvm-svn: 200984

llvm-svn: 200983

Add a darwin triple to get the behaviour from isysroot that the test
expects.

llvm-svn: 200982

An intermediate solution until the problems with analyzer plugins linking with
llvm/Support and causing assertions due to duplicate GeneralCategory are solved.

llvm-svn: 200981

This means always walking the whole call stack for the end path node, but
we'll assume that's always fairly tractable.

<rdar://problem/15952973>

llvm-svn: 200980

We don't stat the system headers to check for stalenes during regular
PCH loading for performance reasons.  When explicitly saying
-verify-pch, we want to check all the dependencies - user or system.

llvm-svn: 200979

Differential Revision: http://llvm-reviews.chandlerc.com/D2694

llvm-svn: 200978

Thanks to John McCall for noticing it.

llvm-svn: 200977

No functionality change.

llvm-svn: 200976

llvm-svn: 200975

Before (81 columns):
  #define A                                                                       \
    void aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa() { return aaaaaaaa; } \
    int i;

After:
  #define A                                                    \
    void aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa() { \
      return aaaaaaaa;                                         \
    }                                                          \
    int i;

llvm-svn: 200974

llvm-svn: 200973

remove TSAN_OPTIONS from the test, so that it runs in the presubmit script
add a CHECK, so that it can run with FileCheck as well

llvm-svn: 200972

An HFA is defined as a struct containing floating point values of the
same machine type. In the 32-bit ABI, double and long double have the
same machine type, so a struct with a mixture of these types must be an
HFA (assuming it meets the other criteria).

llvm-svn: 200971

According to the AAPCS, when a CPRC is allocated to the stack, all other
VFP registers should be marked as unavailable.

I have also modified the rules for allocating non-CPRCs to the stack, to make
it more explicit that all GPRs must be made unavailable. I cannot think of a
case where the old version would produce incorrect answers, so there is no test
for this.

llvm-svn: 200970

llvm-svn: 200969

Before:
  for (aaaaaaaaa aaaaaaaaaaaaaaaaaaaaa : aaaaaaaaaaaa.aaaaaaaaaaaa()
           .aaaaaaaaa()
           .a()) {
  }

After:
  for (aaaaaaaaa aaaaaaaaaaaaaaaaaaaaa :
       aaaaaaaaaaaa.aaaaaaaaaaaa().aaaaaaaaa().a()) {
  }

llvm-svn: 200968

llvm-svn: 200967

when global and thread state are initialized,
this allows to do more things in the callback.

llvm-svn: 200966

llvm-svn: 200965

    hack of passing -fconst-strings to -cc1"

Passing or not a language option based on diagnostic settings is a bad idea, it breaks
using a PCH that was compiled with different diagnostic settings.

Also add a test case to make sure we don't regress.

llvm-svn: 200964

[Sparc] Emit correct encoding for atomic instructions. Also, add support for parsing CAS instructions to test the CAS encoding. 

llvm-svn: 200963