lib/Analysis/README.txt File Reference

Functions
Analysis the ScalarEvolution expression for r is< loop > Outside the this could be evaluated simply	as (%n *%n)
Analysis the ScalarEvolution expression for r is< loop > Outside the this could be evaluated simply 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

Variables
Analysis	Opportunities
Analysis the ScalarEvolution expression for r is	this
Analysis the ScalarEvolution expression for r is< loop > Outside the	loop
Analysis the ScalarEvolution expression for r is< loop > Outside the this could be evaluated simply however ScalarEvolution currently evaluates it it involves i65	arithmetic
Analysis the ScalarEvolution expression for r is< loop > Outside the this could be evaluated simply however ScalarEvolution currently evaluates it it involves i65 which is very inefficient when expanded into code In formatValue in test CodeGen X86 lsr delayed fold	ll

Function Documentation

as() [1/2]

Analysis the ScalarEvolution expression for r is<loop> Outside the this could be evaluated simply as

(

%n *%

n

)

as() [2/2]

Analysis the ScalarEvolution expression for r is<loop> Outside the this could be evaluated simply 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

)

Variable Documentation

arithmetic

Analysis the ScalarEvolution expression for r is<loop> Outside the this could be evaluated simply however ScalarEvolution currently evaluates it it involves i65 arithmetic

Definition at line 15 of file README.txt.

ll

Analysis the ScalarEvolution expression for r is<loop> Outside the this could be evaluated simply however ScalarEvolution currently evaluates it it involves i65 which is very inefficient when expanded into code In formatValue in test CodeGen X86 lsr delayed fold ll

Definition at line 20 of file README.txt.

Referenced by llvm::RISCVMatInt::generateInstSeq(), llvm::sys::toFILETIME(), and llvm::sys::toTimePoint().

loop

is twice as slow as this loop

Definition at line 8 of file README.txt.

Opportunities

Analysis Opportunities

Definition at line 5 of file README.txt.

this

LLVM currently emits rax rax movq rax rax ret It could narrow the loads and stores to emit this

Definition at line 8 of file README.txt.

Referenced by llvm::BranchProbabilityInfo::copyEdgeProbabilities(), llvm::orc::DumpObjects::DumpObjects(), llvm::CondOpInit::Fold(), llvm::GCStatepointInst::getGCResultLocality(), llvm::IRAttribute< Attribute::Returned, AbstractAttribute >::initialize(), llvm::simple_ilist< MachineInstr, Options... >::insert(), llvm::orc::MaterializationUnit::MaterializationUnit(), llvm::SmallVectorImpl< llvm::VectorizationFactor >::operator=(), llvm::OwningArrayRef< T >::OwningArrayRef(), llvm::PMStack::push(), llvm::PriorityQueue< T, Sequence, Compare >::reheapify(), llvm::IntervalMap< KeyT, ValT, N, Traits >::iterator::setStopUnchecked(), llvm::SmallVectorMemoryBuffer::SmallVectorMemoryBuffer(), llvm::BasicBlock::splitBasicBlock(), llvm::BasicBlock::splitBasicBlockBefore(), llvm::SmallPtrSetImplBase::swap(), llvm::SmallVectorImpl< llvm::VectorizationFactor >::swap(), and llvm::SmallDenseMap< llvm::Value *, llvm::Value * >::swap().