Nadav Rotem 589dc766e0 When unsafe math is used, we can use commutative FMAX and FMIN. In some cases ...

this allows for better code generation.

Added a new DAGCombine transformation to convert FMAX and FMIN to FMANC and
FMINC, which are commutative.

For example:

  movaps  %xmm0, %xmm1
  movsd LC(%rip), %xmm0
  minsd %xmm1, %xmm0

becomes:

  minsd LC(%rip), %xmm0

llvm-svn: 162187

2012-08-19 13:06:16 +00:00

..

Analysis

MemoryBuiltins: Properly guard ObjectSizeOffsetVisitor against cycles in the IR.

2012-08-17 19:26:41 +00:00

Archive

Remove references to compression in llvm-ar. It has been a long time since we

2012-08-10 01:57:52 +00:00

AsmParser

Change the linker_private_weak_def_auto' linkage to linkonce_odr_auto_hide' to

2012-08-17 18:33:14 +00:00

Bitcode

Change the linker_private_weak_def_auto' linkage to linkonce_odr_auto_hide' to

2012-08-17 18:33:14 +00:00

CodeGen

Make atomic load and store of pointers work. Tighten verification of atomic operations

2012-08-17 23:24:29 +00:00

DebugInfo

Fix the representation of debug line table in DebugInfo LLVM library,

2012-08-07 11:46:57 +00:00

ExecutionEngine

Correct MCJIT functionality for MIPS32 architecture.

2012-08-17 21:28:04 +00:00

Linker

Move the "findUsedStructTypes" functionality outside of the Module class.

2012-08-03 00:30:35 +00:00

MC

When emitting the PC range in an FDE, use the same data encoding for both ends

2012-08-12 08:09:45 +00:00

Object

…

Support

Properly test the LLVM_USE_RVALUE_REFERENCES macro.

2012-08-15 19:16:27 +00:00

TableGen

TableGen: Allow use of #NAME# outside of 'def' names.

2012-08-02 18:46:42 +00:00

Target

When unsafe math is used, we can use commutative FMAX and FMIN. In some cases

2012-08-19 13:06:16 +00:00

Transforms

InstCombine: Fix a crasher when encountering a function pointer.

2012-08-18 22:04:34 +00:00

VMCore

Make atomic load and store of pointers work. Tighten verification of atomic operations

2012-08-17 23:24:29 +00:00

CMakeLists.txt

…

LLVMBuild.txt

…

Makefile

…