Arnold Schwaighofer a3fb330d05 LoopVectorizer: Recognize min/max reductions ... A min/max operation is represented by a select(cmp(lt/le/gt/ge, X, Y), X, Y) sequence in LLVM. If we see such a sequence we can treat it just as any other commutative binary instruction and reduce it. This appears to help bzip2 by about 1.5% on an imac12,2. radar://12960601 git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@179773 91177308-0d34-0410-b5e6-96231b3b80d8		2013-04-18 17:22:34 +00:00
..
Analysis	X86 cost model: Exit before calling getSimpleVT on non-simple VTs	2013-04-17 20:04:53 +00:00
Archive
Assembler
Bindings/Ocaml
Bitcode
BugPoint
CodeGen	Fix for PR14824, An ARM Load/Store Optimization bug	2013-04-18 09:11:08 +00:00
DebugInfo	PR15149/r174304 improvement - print hex for unknown dwarf language codes & add a test case	2013-04-17 03:41:36 +00:00
ExecutionEngine
Feature
FileCheck
Instrumentation
Integer
JitListener
Linker
MC	[mips] DSP-ASE move from HI/LO register instructions.	2013-04-18 00:52:44 +00:00
Object	llvm-objdump: Don't print contents of BSS sections: it makes no sense and crashes llvm-objdump on relocated objects with large bss	2013-04-16 10:53:11 +00:00
Other
TableGen
tools
Transforms	LoopVectorizer: Recognize min/max reductions	2013-04-18 17:22:34 +00:00
Unit
Verifier
YAMLParser
CMakeLists.txt
lit.cfg
lit.site.cfg.in
Makefile	This appears to be no longer necessary for the testsuite.	2013-04-17 06:37:30 +00:00
Makefile.tests
TestRunner.sh