llvm-6502

mirror of https://github.com/c64scene-ar/llvm-6502.git synced 2024-10-05 22:59:16 +00:00

History

Benjamin Kramer 8611d4449a Cost model support for lowered math builtins. We make the cost for calling libm functions extremely high as emitting the calls is expensive and causes spills (on x86) so performance suffers. We still vectorize important calls like ceilf and friends on SSE4.1. and fabs. Differential Revision: http://llvm-reviews.chandlerc.com/D466 git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@176287 91177308-0d34-0410-b5e6-96231b3b80d8		2013-02-28 19:09:33 +00:00
..
IPA	Begin fleshing out an interface in TTI for modelling the costs of	2013-01-22 11:26:02 +00:00
AliasAnalysis.cpp	Move isKnownNonNull out of AliasAnalysis.h and into ValueTracking.cpp since	2013-01-31 02:40:59 +00:00
AliasAnalysisCounter.cpp
AliasAnalysisEvaluator.cpp
AliasDebugger.cpp
AliasSetTracker.cpp
Analysis.cpp
BasicAliasAnalysis.cpp	Fix a problem in alias analysis. It is about the misinterpretation of "Object".	2013-02-28 00:24:45 +00:00
BlockFrequencyInfo.cpp
BranchProbabilityInfo.cpp
CaptureTracking.cpp
CFGPrinter.cpp
CMakeLists.txt
CodeMetrics.cpp	Begin fleshing out an interface in TTI for modelling the costs of	2013-01-22 11:26:02 +00:00
ConstantFolding.cpp	Constant fold vector bitcasts of halves similarly to how floats and doubles are folded. Test case included.	2013-02-26 22:51:07 +00:00
CostModel.cpp	Cost model support for lowered math builtins.	2013-02-28 19:09:33 +00:00
DbgInfoPrinter.cpp
DependenceAnalysis.cpp
DominanceFrontier.cpp
DomPrinter.cpp
InstCount.cpp
InstructionSimplify.cpp	Identify and simplify idempotent intrinsics. Test case included.	2013-02-07 19:26:05 +00:00
Interval.cpp
IntervalPartition.cpp
IVUsers.cpp
LazyValueInfo.cpp	Move isKnownNonNull out of AliasAnalysis.h and into ValueTracking.cpp since	2013-01-31 02:40:59 +00:00
LibCallAliasAnalysis.cpp
LibCallSemantics.cpp
Lint.cpp	Change GetPointerBaseWithConstantOffset's DataLayout argument from a	2013-01-31 02:00:45 +00:00
LLVMBuild.txt
Loads.cpp	Change GetPointerBaseWithConstantOffset's DataLayout argument from a	2013-01-31 02:00:45 +00:00
LoopInfo.cpp	Metadata for annotating loops as parallel. The first consumer for this	2013-02-13 18:08:57 +00:00
LoopPass.cpp
Makefile
MemDepPrinter.cpp
MemoryBuiltins.cpp
MemoryDependenceAnalysis.cpp	Unify clang/llvm attributes for asan/tsan/msan (LLVM part)	2013-02-26 06:58:09 +00:00
ModuleDebugInfoPrinter.cpp
NoAliasAnalysis.cpp
PathNumbering.cpp
PathProfileInfo.cpp
PathProfileVerifier.cpp
PHITransAddr.cpp
PostDominators.cpp
ProfileDataLoader.cpp
ProfileDataLoaderPass.cpp	Revert "Add LLVMContext::emitWarning methods and use them. <rdar://problem/12867368>"	2013-02-08 21:48:29 +00:00
ProfileEstimatorPass.cpp
ProfileInfo.cpp
ProfileInfoLoader.cpp
ProfileInfoLoaderPass.cpp	Revert "Add LLVMContext::emitWarning methods and use them. <rdar://problem/12867368>"	2013-02-08 21:48:29 +00:00
ProfileVerifierPass.cpp
PtrUseVisitor.cpp
README.txt
RegionInfo.cpp
RegionPass.cpp
RegionPrinter.cpp
ScalarEvolution.cpp
ScalarEvolutionAliasAnalysis.cpp
ScalarEvolutionExpander.cpp
ScalarEvolutionNormalization.cpp
SparsePropagation.cpp
TargetTransformInfo.cpp	ARM cost model: Address computation in vector mem ops not free	2013-02-08 14:50:48 +00:00
Trace.cpp
TypeBasedAliasAnalysis.cpp
ValueTracking.cpp	Move isKnownNonNull out of AliasAnalysis.h and into ValueTracking.cpp since	2013-01-31 02:40:59 +00:00

README.txt

Analysis Opportunities:

//===---------------------------------------------------------------------===//

In test/Transforms/LoopStrengthReduce/quadradic-exit-value.ll, the
ScalarEvolution expression for %r is this:

  {1,+,3,+,2}<loop>

Outside the loop, this could be evaluated simply as (%n * %n), however
ScalarEvolution currently evaluates it as

  (-2 + (2 * (trunc i65 (((zext i64 (-2 + %n) to i65) * (zext i64 (-1 + %n) to i65)) /u 2) to i64)) + (3 * %n))

In addition to being much more complicated, it involves i65 arithmetic,
which is very inefficient when expanded into code.

//===---------------------------------------------------------------------===//

In formatValue in test/CodeGen/X86/lsr-delayed-fold.ll,

ScalarEvolution is forming this expression:

((trunc i64 (-1 * %arg5) to i32) + (trunc i64 %arg5 to i32) + (-1 * (trunc i64 undef to i32)))

This could be folded to

(-1 * (trunc i64 undef to i32))

//===---------------------------------------------------------------------===//