llvm-6502/lib
Andrew Trick 12d1ebe943 Enable SCEV-based unrolling by default.
This changes loop unrolling to use the same mechanism for trip count
computation as indvars. This is a stronger check that tends to unroll
more loops. A very common side-effect is that many single iteration
loops will be removed sooner. The real goal was simply to remove
dependence on canonical IVs.

x86 is break even.
ARM performance changes to expect (+ is good):
External/SPEC/CFP2000/183.equake/183.equake +13%
SingleSource/Benchmarks/Dhrystone/fldry     +21%
MultiSource/Applications/spiff/spiff         +3%
SingleSource/Benchmarks/Stanford/Puzzle     -14%

The Puzzle regression is actually an improvement in loop optimization
that defeats GVN: rdar://problem/10065079.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@139009 91177308-0d34-0410-b5e6-96231b3b80d8
2011-09-02 17:26:28 +00:00
..
Analysis After r138010, subroutine type does not have context info. Update type verifier accordingly. 2011-08-31 18:04:31 +00:00
Archive Rewrite the CMake build to use explicit dependencies between libraries, 2011-07-29 00:14:25 +00:00
AsmParser Auto upgrade the old EH scheme to use the new one. This is on a trial basis. If 2011-08-27 06:11:03 +00:00
Bitcode Don't forget to add the landingpad and resume instructions to the InstructionList. 2011-09-01 00:50:20 +00:00
CodeGen Revert r131152, r129796, r129761. This code is currently considered 2011-09-01 23:07:08 +00:00
CompilerDriver
ExecutionEngine Move TargetRegistry and TargetSelect from Target to Support where they belong. 2011-08-24 18:08:43 +00:00
Linker switch to the new struct api. 2011-08-12 18:07:26 +00:00
MC Don't drop alignment info on local common symbols. 2011-09-01 23:04:27 +00:00
Object Teach macho-dump to dump the uleb128s referred to by linkedit_data segments. 2011-08-30 22:10:58 +00:00
Support Add AMDIL as valid target triple to LLVM. 2011-08-29 15:44:55 +00:00
Target Pass signed (not unsigned) 10 bit field to SPU 'ori' instruction. 2011-09-02 10:05:01 +00:00
Transforms Enable SCEV-based unrolling by default. 2011-09-02 17:26:28 +00:00
VMCore Return undef value (instead of arbitrary) for wrong or undef index in 2011-09-02 17:01:40 +00:00
CMakeLists.txt
Makefile