llvm/Analysis at 677d6a3a87f954487a82abc644688c4c49be64f3 - llvm

RPCS3/llvm

mirror of https://github.com/RPCS3/llvm.git synced 2024-12-22 20:20:03 +00:00

History

Silviu Baranga 0c35941274 [SCEV] Generalize the SCEV algorithm for creating expressions for PHI nodes Summary: When forming expressions for phi nodes having an incoming value from outside the loop A and a value coming from the previous iteration B we were forming an AddRec if: - B was an AddRec - the value A was equal to the value for B at iteration -1 (or equal to the value of B shifted by one iteration, at iteration 0) In this case, we were computing the expression to be the expression of B, shifted by one iteration. This changes generalizes the logic above by removing the restriction that B needs to be an AddRec. For this we introduce two expression rewriters that allow us to - shift an expression by one iteration - get the value of an expression at iteration 0 This allows us to get SCEV expressions for PHI nodes when these expressions are not AddRecExprs. Reviewers: sanjoy Subscribers: llvm-commits, sanjoy Differential Revision: http://reviews.llvm.org/D14175 git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@251700 91177308-0d34-0410-b5e6-96231b3b80d8		2015-10-30 15:02:28 +00:00
..
AliasAnalysis.cpp
AliasAnalysisEvaluator.cpp
AliasSetTracker.cpp	Revert "r251451 - [AliasSetTracker] Use mod/ref information for UnknownInstr"	2015-10-28 22:13:41 +00:00
Analysis.cpp
AssumptionCache.cpp
BasicAliasAnalysis.cpp	[AliasAnalysis] Take into account readnone attribute for the function arguments	2015-10-28 17:54:48 +00:00
BlockFrequencyInfo.cpp
BlockFrequencyInfoImpl.cpp
BranchProbabilityInfo.cpp	Check the case that the numerator and denominator are both zeros when getting edge probabilities in BPI and return 100% in this case.	2015-10-26 18:00:17 +00:00
CallGraph.cpp
CallGraphSCCPass.cpp
CallPrinter.cpp
CaptureTracking.cpp
CFG.cpp
CFGPrinter.cpp
CFLAliasAnalysis.cpp
CGSCCPassManager.cpp
CMakeLists.txt
CodeMetrics.cpp	Use all_of to simplify control flow. NFC.	2015-10-24 19:30:37 +00:00
ConstantFolding.cpp
CostModel.cpp
Delinearization.cpp
DemandedBits.cpp
DependenceAnalysis.cpp
DivergenceAnalysis.cpp
DominanceFrontier.cpp
DomPrinter.cpp
GlobalsModRef.cpp	[GlobalsAA] An indirect global that is initialized is not fair game	2015-10-28 10:41:29 +00:00
InlineCost.cpp
InstCount.cpp
InstructionSimplify.cpp	[InstSimplify] sgt on i1s also encodes implication	2015-10-29 03:19:10 +00:00
Interval.cpp
IntervalPartition.cpp
IteratedDominanceFrontier.cpp
IVUsers.cpp
LazyCallGraph.cpp
LazyValueInfo.cpp	Fix an unused variable warning which broke the clang-cmake-mips builder	2015-10-29 04:21:49 +00:00
LibCallSemantics.cpp
Lint.cpp
LLVMBuild.txt
Loads.cpp
LoopAccessAnalysis.cpp	Put global classes into the appropriate namespace.	2015-10-28 13:54:36 +00:00
LoopInfo.cpp
LoopPass.cpp
Makefile
MemDepPrinter.cpp
MemDerefPrinter.cpp
MemoryBuiltins.cpp	Use find_if to simplify control flow. NFC.	2015-10-24 19:03:15 +00:00
MemoryDependenceAnalysis.cpp
MemoryLocation.cpp
ModuleDebugInfoPrinter.cpp
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] Generalize the SCEV algorithm for creating expressions for PHI nodes	2015-10-30 15:02:28 +00:00
ScalarEvolutionAliasAnalysis.cpp
ScalarEvolutionExpander.cpp	[ScalarEvolutionExpander] PHI on a catchpad can be used on both edges	2015-10-27 19:48:28 +00:00
ScalarEvolutionNormalization.cpp
ScopedNoAliasAA.cpp
SparsePropagation.cpp
StratifiedSets.h
TargetLibraryInfo.cpp	ARM: teach backend about WatchOS and TvOS libcalls.	2015-10-28 22:51:16 +00:00
TargetTransformInfo.cpp	Scalarizer for masked.gather and masked.scatter intrinsics.	2015-10-25 15:37:55 +00:00
Trace.cpp
TypeBasedAliasAnalysis.cpp
ValueTracking.cpp	[ValueTracking] Expose `implies` via ValueTracking, NFC	2015-10-28 03:20:19 +00:00
VectorUtils.cpp

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))

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