llvm-svn: 154695

- Don't copy offsets into HashData, the underlying vector won't change once the table is finalized.
- Allocate HashData and HashDataContents in a BumpPtrAllocator.
- Allocate string map entries in the same allocator.
- Random cleanups.

llvm-svn: 154694

llvm-svn: 154692

llvm-svn: 154691

llvm-svn: 154689

llvm-svn: 154687

targets so if the branch target has the high bit set it does not get printed as:
	 beq     0xffffffff8008c404

llvm-svn: 154685

jump into these scopes, and the cleanup-entering code sometimes wants
to do some operations first (e.g. a GEP), which can leave us with
unparented IR.

llvm-svn: 154684

llvm-svn: 154683

their argument as "escape" points for objc_retainBlock optimization.
This fixes rdar://11229925.

llvm-svn: 154682

ThreadPlanCallFunction's destructor wasn't calling DoTakedown, so if the that plan got discarded we weren't doing the takedown.

llvm-svn: 154681

__typeof which is a regression by reverting
r154360. // rdar://11233924

llvm-svn: 154679

llvm-svn: 154678

llvm-svn: 154676

llvm-svn: 154675

llvm-svn: 154674

As has been suggested by Duncan and others, Early-CSE and GVN should
do similar redundancy elimination, but Early-CSE is much less expensive.
Most of my autovectorization benchmarks show a performance regresion, but
all of these are < 0.1%, and so I think that it is still worth using
the less expensive pass.

llvm-svn: 154673

llvm-svn: 154672

Added a --memory option to allow dumping the matching malloc block memory with a default format that makes sense, or that format can be overridden with the --format option.

llvm-svn: 154671

call to 'super' use __rw_objc_super as type of the
'super' meta-data instead of objc_super.
// rdar://11239894

llvm-svn: 154670

Make control flow more explicit for rebuilding property reference expressions without their OpaqueValueExprs

llvm-svn: 154669

out of the tree and use the tooling infrastructure.

llvm-svn: 154668

rebuilt. Fixes <rdar://problem/11052352>. 

llvm-svn: 154667

llvm-svn: 154666

For example, if llc cannot be found, the full python stacktrace is displayed
and no interesting information are provided.
+ fail the process when an exception occurs

llvm-svn: 154665

  - Handle unions
  - Handle C++ classes

llvm-svn: 154664

llvm-svn: 154661

Silence various build warnings from Hexagon backend that show up in release builds. Mostly converting 'assert(0)' to 'llvm_unreachable' to silence warnings about missing returns. Also fold some variable declarations into asserts to prevent the variables from being unused in release builds.

llvm-svn: 154660

GNU __atomic builtins.

llvm-svn: 154659

Fix target specific intrinsic handling to adjust intrinsic number before doing attribute table lookup. Also fix attribute table lookup to handle 'invalid' intrinsic correctly. Fixes PR12542

llvm-svn: 154658

Remove getElfArchType from ELF.h. It's only used in ELFObjectFile.cpp and there's already a copy there. ELF.h was hiding the one there and causing an unused function warning.

llvm-svn: 154657

in the wrong namespace scope. Patch by Jonathan Sauer!

llvm-svn: 154656

http://lists.cs.uiuc.edu/pipermail/cfe-commits/Week-of-Mon-20120409/056126.html

llvm-svn: 154655

in -std=gnu++11 mode.

llvm-svn: 154654

and add a test case.

llvm-svn: 154653

The less locks there are, the better. I removed the thread ID mutex and now just shared the m_thread_list's mutex to make sure we don't deadlock due to lock inversion.

llvm-svn: 154652

llvm-svn: 154650

the MC disassembler.

llvm-svn: 154649

Thanks to Nico Weber for the suggestion.

llvm-svn: 154648

library return value optimization for phi uses. Even when the
phi itself is not dominated, the specific use may be dominated.

llvm-svn: 154647