LLVM  6.0.0svn
Public Member Functions | Public Attributes | List of all members
llvm::TargetTransformInfo::ReductionFlags Struct Reference

Flags describing the kind of vector reduction. More...

#include "llvm/Analysis/TargetTransformInfo.h"

Collaboration diagram for llvm::TargetTransformInfo::ReductionFlags:
Collaboration graph
[legend]

Public Member Functions

 ReductionFlags ()
 

Public Attributes

bool IsMaxOp
 If the op a min/max kind, true if it's a max operation. More...
 
bool IsSigned
 Whether the operation is a signed int reduction. More...
 
bool NoNaN
 If op is an fp min/max, whether NaNs may be present. More...
 

Detailed Description

Flags describing the kind of vector reduction.

Definition at line 903 of file TargetTransformInfo.h.

Constructor & Destructor Documentation

◆ ReductionFlags()

llvm::TargetTransformInfo::ReductionFlags::ReductionFlags ( )
inline

Definition at line 904 of file TargetTransformInfo.h.

Member Data Documentation

◆ IsMaxOp

bool llvm::TargetTransformInfo::ReductionFlags::IsMaxOp

If the op a min/max kind, true if it's a max operation.

Definition at line 905 of file TargetTransformInfo.h.

Referenced by createRdxShuffleMask(), llvm::createSimpleTargetReduction(), and llvm::createTargetReduction().

◆ IsSigned

bool llvm::TargetTransformInfo::ReductionFlags::IsSigned

Whether the operation is a signed int reduction.

Definition at line 906 of file TargetTransformInfo.h.

Referenced by createRdxShuffleMask(), llvm::createSimpleTargetReduction(), and llvm::createTargetReduction().

◆ NoNaN

bool llvm::TargetTransformInfo::ReductionFlags::NoNaN

If op is an fp min/max, whether NaNs may be present.

Definition at line 907 of file TargetTransformInfo.h.

Referenced by createRdxShuffleMask(), llvm::createSimpleTargetReduction(), llvm::createTargetReduction(), and llvm::AArch64TTIImpl::useReductionIntrinsic().


The documentation for this struct was generated from the following file: