33390848a7eca75301d04a59b89b516d83e19ee0 - platform/external/llvm

commit	33390848a7eca75301d04a59b89b516d83e19ee0	[log] [tgz]
author	Cameron Zwarich <zwarich@apple.com>	Fri Jul 08 21:39:21 2011 +0000
committer	Cameron Zwarich <zwarich@apple.com>	Fri Jul 08 21:39:21 2011 +0000
tree	78f81419440b86ae4fa8a150780b7f211668b781
parent	cc0ddc707d5a7b1dd11141881df0bf4210f8aeee [diff]

Add an intrinsic and codegen support for fused multiply-accumulate. The intent
is to use this for architectures that have a native FMA instruction.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@134742 91177308-0d34-0410-b5e6-96231b3b80d8

docs/LangRef.html[diff]
include/llvm/CodeGen/ISDOpcodes.h[diff]
include/llvm/CodeGen/RuntimeLibcalls.h[diff]
include/llvm/Intrinsics.td[diff]
lib/CodeGen/SelectionDAG/LegalizeDAG.cpp[diff]
lib/CodeGen/SelectionDAG/LegalizeFloatTypes.cpp[diff]
lib/CodeGen/SelectionDAG/LegalizeTypes.h[diff]
lib/CodeGen/SelectionDAG/SelectionDAG.cpp[diff]
lib/CodeGen/SelectionDAG/SelectionDAGBuilder.cpp[diff]
lib/CodeGen/SelectionDAG/TargetLowering.cpp[diff]
lib/Target/ARM/ARMISelLowering.cpp[diff]
lib/Target/Alpha/AlphaISelLowering.cpp[diff]
lib/Target/CellSPU/SPUISelLowering.cpp[diff]
lib/Target/MBlaze/MBlazeISelLowering.cpp[diff]
lib/Target/Mips/MipsISelLowering.cpp[diff]
lib/Target/PowerPC/PPCISelLowering.cpp[diff]
lib/Target/Sparc/SparcISelLowering.cpp[diff]
lib/Target/SystemZ/SystemZISelLowering.cpp[diff]
lib/Target/X86/X86ISelLowering.cpp[diff]
test/CodeGen/X86/fma.ll[Added - diff]

20 files changed

tree: 78f81419440b86ae4fa8a150780b7f211668b781