Tom Stellard 0ffcaa0d54 SelectionDAG: Optimize expansion of vec_type = BITCAST scalar_type ... The legalizer can now do this type of expansion for more type combinations without loading and storing to and from the stack. NOTE: This is a candidate for the 3.4 branch. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@195398 91177308-0d34-0410-b5e6-96231b3b80d8		2013-11-22 00:41:05 +00:00
..
Analysis	Don't speculate loads under ThreadSanitizer	2013-11-21 07:29:28 +00:00
AsmParser	Make it explicit that nulls are not allowed in names.	2013-11-19 21:12:39 +00:00
Bitcode	Fix spacing, forward declare order.	2013-11-18 02:51:33 +00:00
CodeGen	SelectionDAG: Optimize expansion of vec_type = BITCAST scalar_type	2013-11-22 00:41:05 +00:00
DebugInfo	DebugInfo: Make pubnames header printing similar to unit header printing	2013-11-01 17:53:30 +00:00
ExecutionEngine	[mips] Resolve relocation for the stubs in MCJIT when load address is known	2013-11-19 21:56:00 +00:00
IR	[PM] Widen the interface for invalidate on an analysis result now that	2013-11-21 10:53:05 +00:00
IRReader	[llvm-c] Expose IRReader interface	2013-11-06 09:21:15 +00:00
Linker	Revert "Move copying of global initializers below the cloning of functions."	2013-11-09 00:43:18 +00:00
LTO	Use array_pod_sort instead of std::sort	2013-11-16 16:15:56 +00:00
MC	Support multiple COFF sections with the same name but different COMDAT.	2013-11-19 19:52:52 +00:00
Object	Path: Recognize COFF import library file magic.	2013-11-15 21:22:02 +00:00
Option	Use startswith_lower() where possible.	2013-11-04 19:22:50 +00:00
Support	revert r194655	2013-11-21 00:20:10 +00:00
TableGen	Fix most memory leaks in tablegen.	2013-10-31 04:07:41 +00:00
Target	SHLD/SHRD are VectorPath (microcode) instructions known to have poor latency on certain architectures. While generating SHLD/SHRD instructions is acceptable when optimizing for size, optimizing for speed on these platforms should be implemented using alternative sequences of instructions composed of add, adc, shr, shl, or and lea which are directPath instructions. These alternative instructions not only have a lower latency but they also increase the decode bandwidth by allowing simultaneous decoding of a third directPath instruction.	2013-11-21 23:21:26 +00:00
Transforms	Introduce two command-line flags for the instrumentation pass to control whether the labels of pointers should be ignored in load and store instructions	2013-11-21 23:20:54 +00:00
CMakeLists.txt	Move LTO support library to a component, allowing it to be tested	2013-09-24 23:52:22 +00:00
LLVMBuild.txt	Move LTO support library to a component, allowing it to be tested	2013-09-24 23:52:22 +00:00
Makefile	Reformat Makefile. No other changes.	2013-10-30 04:03:03 +00:00