Without -ffreestanding, clang tries to seek /usr/include/stdlib.h in host filesystem, even on Windows hosts.

llvm-svn: 139899

Tweak *mmintrin.h so that they don't make any bad assumptions about alignment (which probably has little effect in practice, but better to get it right).  Make the load in _mm_loadh_pi and _mm_loadl_pi a single LLVM IR instruction to make optimizing easier for CodeGen.

rdar://10054986

llvm-svn: 139874

Make clang use Acquire loads and Release stores where necessary.

llvm-svn: 139650

llvm-svn: 139648

llvm-svn: 139643

llvm-svn: 139464

llvm-svn: 139462

llvm-svn: 139460

a test failure in CodeGen/palignr.c, which has been marked
XFAIL for the time being.  A bug has been filed at PR10901
for this issue.

llvm-svn: 139457

llvm-svn: 139455

annotate global, local variables, struct fields, or arbitrary statements (using
the __builtin_annotation), rdar://8037476.

llvm-svn: 139423

along with the new insert point.

Fixes PR10829

llvm-svn: 139416

feature akin to the ARC runtime checks.  Removes a terrible hack where
IR gen needed to find the declarations of those symbols in the translation
unit.

llvm-svn: 139404

These functions return a second value by writing to a pointer argument,
so they cannot be marked 'readnone' which implies that they don't access
memory.

<rdar://problem/10070234>

llvm-svn: 139319

llvm-svn: 139224

Switch clang over to using fence/atomicrmw/cmpxchg instead of the intrinsics (which will go away).  LLVM CodeGen does almost exactly the same thing with these and the old intrinsics, so I'm reasonably confident this will not break anything.

There are still a few issues which need to be resolved with code generation for atomic load and store, so I'm not converting the places which need those for now.

I'm not entirely sure what to do about __builtin_llvm_memory_barrier: the fence instruction doesn't expose all the possibilities which can be expressed by __builtin_llvm_memory_barrier.  I would appreciate hearing from anyone who is using this intrinsic.

llvm-svn: 139216

llvm-svn: 139043

// rdar://10033896

llvm-svn: 139041

llvm-svn: 139029

// rdar://10033896

llvm-svn: 139020

emit call results into potentially aliased slots.  This allows us
to properly mark indirect return slots as noalias, at the cost
of requiring an extra memcpy when assigning an aggregate call
result into a l-value.  It also brings us into compliance with
the x86-64 ABI.

llvm-svn: 138599

llvm-svn: 138422

After talking with John making this the case for all of these is
the right way to go.

Fixes rdar://9804564 and PR10414

llvm-svn: 138418

is captured by a given statement expression. // rdar://10001085

llvm-svn: 138314

llvm-svn: 138310

Remove one outdated test.

llvm-svn: 138144

llvm-svn: 137752

test over from llvm/test/FrontendC++ and update others to account for
the change.

llvm-svn: 137669

llvm-svn: 137535

llvm-svn: 137302

llvm-svn: 137298

This fixes cases where the anonymous bitfield is followed by a bitfield member.
E.g.,
struct t4
{
 char foo;
 long : 0;
 char bar : 1;
};

rdar://9859156

llvm-svn: 136991

llvm-svn: 136955

llvm-svn: 136924

alignment.  This fixes cases where the anonymous bitfield is followed by a 
non-bitfield member.  E.g.,

struct t4
{
  int foo : 1;
  long : 0;
  char bar;
};

Part of rdar://9859156

llvm-svn: 136858

Patch by Craig Topper and Sundeep!

llvm-svn: 136856

structures.  Alignment can be enforced with the use of anonymous bitfields 
(e.g., int :0), but this is not currently supported.  Add this test case to 
document the current state, which will hopefully be fixed shortly. 

llvm-svn: 136848

designed to be executed, and its output inspected for correct values,
but we aren't executing it. We're just compiling it, and dumping it to
/dev/null. It also isn't freestanding. If there is a desire to have this
test actually stick around, complain and I'll revert this and try to add
the file checks necessary to make this actually test things.

llvm-svn: 136846

A homogeneous aggregate is an aggregate data structure where after flattening
any nesting there are 1 to 4 elements of the same base type that is either a
float, double, or Neon vector.  All Neon vectors of the same size, either 64
or 128 bits, are treated as equivalent for this purpose.  When using the
AAPCS-VFP ABI, check for homogeneous aggregates and pass them as arguments by
expanding them into a sequence of their base types.  This requires extending
the existing support for expanded arguments to handle not only structs, but
also constant arrays and complex types.

llvm-svn: 136767

function, be sure to drop parameter attributes when dropping their
associated arguments.  Patch by Aaron Landwehr!

llvm-svn: 136753