llvm-mirror

mirror of https://github.com/RPCS3/llvm-mirror.git synced 2024-11-24 03:33:20 +01:00

History

Teresa Johnson fcf8fe19ba [ThinLTO] Refine fix to avoid renaming of uses in inline assembly. Summary: Refine the workaround from r266877 that attempts to prevent renaming of locals in inline assembly, so that in addition to looking for a llvm.used local value, that there is at least one inline assembly call in the module. Otherwise, debug functions added to the llvm.used can block importing/exporting unnecessarily. Reviewers: joker.eph Subscribers: llvm-commits, joker.eph Differential Revision: http://reviews.llvm.org/D19573 llvm-svn: 267717		2016-04-27 14:19:38 +00:00
..
AliasAnalysis.cpp
AliasAnalysisEvaluator.cpp
AliasSetTracker.cpp
Analysis.cpp	Port DemandedBits to the new pass manager.	2016-04-18 23:55:01 +00:00
AssumptionCache.cpp
BasicAliasAnalysis.cpp
BlockFrequencyInfo.cpp
BlockFrequencyInfoImpl.cpp
BranchProbabilityInfo.cpp
CallGraph.cpp
CallGraphSCCPass.cpp	Re-commit optimization bisect support (r267022) without new pass manager support.	2016-04-22 22:06:11 +00:00
CallPrinter.cpp
CaptureTracking.cpp
CFG.cpp
CFGPrinter.cpp
CFLAliasAnalysis.cpp
CGSCCPassManager.cpp
CMakeLists.txt
CodeMetrics.cpp
ConstantFolding.cpp	CodeGen: Use PLT relocations for relative references to unnamed_addr functions.	2016-04-22 20:40:10 +00:00
CostModel.cpp
Delinearization.cpp
DemandedBits.cpp	Port DemandedBits to the new pass manager.	2016-04-18 23:55:01 +00:00
DependenceAnalysis.cpp	[DependenceAnalysis] Refactor uses of getConstantPart. NFC.	2016-04-19 16:46:57 +00:00
DivergenceAnalysis.cpp
DominanceFrontier.cpp
DomPrinter.cpp
EHPersonalities.cpp
GlobalsModRef.cpp
InlineCost.cpp	Introduce llvm.load.relative intrinsic.	2016-04-22 21:18:02 +00:00
InstCount.cpp
InstructionSimplify.cpp	Introduce llvm.load.relative intrinsic.	2016-04-22 21:18:02 +00:00
Interval.cpp
IntervalPartition.cpp
IteratedDominanceFrontier.cpp	Correct IDF calculator for ReverseIDF	2016-04-19 06:13:28 +00:00
IVUsers.cpp
LazyCallGraph.cpp
LazyValueInfo.cpp	[LVI] Delete stale and misleading comment.	2016-04-27 03:03:15 +00:00
Lint.cpp
LLVMBuild.txt
Loads.cpp	NFC. Introduce Value::getPointerDerferecnceableBytes	2016-04-27 12:51:01 +00:00
LoopAccessAnalysis.cpp	[ValueTracking, VectorUtils] Refactor getIntrinsicIDForCall	2016-04-19 19:10:21 +00:00
LoopInfo.cpp
LoopPass.cpp	Re-commit optimization bisect support (r267022) without new pass manager support.	2016-04-22 22:06:11 +00:00
LoopPassManager.cpp
LoopUnrollAnalyzer.cpp
MemDepPrinter.cpp
MemDerefPrinter.cpp
MemoryBuiltins.cpp
MemoryDependenceAnalysis.cpp
MemoryLocation.cpp
ModuleDebugInfoPrinter.cpp
ModuleSummaryAnalysis.cpp	[ThinLTO] Refine fix to avoid renaming of uses in inline assembly.	2016-04-27 14:19:38 +00:00
ObjCARCAliasAnalysis.cpp
ObjCARCAnalysisUtils.cpp
ObjCARCInstKind.cpp
OrderedBasicBlock.cpp
PHITransAddr.cpp
PostDominators.cpp
PtrUseVisitor.cpp
README.txt
RegionInfo.cpp
RegionPass.cpp
RegionPrinter.cpp
ScalarEvolution.cpp	[SCEV] Extract out a `isSCEVExprNeverPoison` helper; NFCI	2016-04-22 05:38:54 +00:00
ScalarEvolutionAliasAnalysis.cpp
ScalarEvolutionExpander.cpp	[SCEV] Improve the run-time checking of the NoWrap predicate	2016-04-25 09:27:16 +00:00
ScalarEvolutionNormalization.cpp	Remove emacs mode markers from .cpp files. NFC	2016-04-24 17:55:41 +00:00
ScopedNoAliasAA.cpp
SparsePropagation.cpp
StratifiedSets.h
TargetLibraryInfo.cpp
TargetTransformInfo.cpp
Trace.cpp
TypeBasedAliasAnalysis.cpp
ValueTracking.cpp	[ValueTracking] Improve isImpliedCondition when the dominating cond is false.	2016-04-25 17:23:36 +00:00
VectorUtils.cpp	[ValueTracking, VectorUtils] Refactor getIntrinsicIDForCall	2016-04-19 19:10:21 +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))

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