mirror/eigen
2
0
Fork 0
mirror of https://gitlab.com/libeigen/eigen.git synced 2024-12-15 07:10:37 +08:00
Code Issues Packages Projects Releases Wiki Activity

LinearVectorization: If the destination isn't aligned,

Browse Source 
we have to do runtime checks and we don't unroll, so it's only good for large enough sizes
This commit is contained in:
Benoit Jacob 2009-08-09 22:19:12 +02:00
parent 1f1705868b
commit 216ee335ac
1 changed files with 4 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

5
Eigen/src/Core/Assign.h
View File

@ -57,7 +57,10 @@ private:
&& ((int(Derived::Flags)&RowMajorBit)==(int(OtherDerived::Flags)&RowMajorBit)),
MayInnerVectorize = MightVectorize && int(InnerSize)!=Dynamic && int(InnerSize)%int(PacketSize)==0
&& int(DstIsAligned) && int(SrcIsAligned),
MayLinearVectorize = MightVectorize && (int(Derived::Flags) & int(OtherDerived::Flags) & LinearAccessBit),
MayLinearVectorize = MightVectorize && (int(Derived::Flags) & int(OtherDerived::Flags) & LinearAccessBit)
&& (DstIsAligned || InnerMaxSize == Dynamic),/* If the destination isn't aligned,
we have to do runtime checks and we don't unroll, so it's only good for large enough sizes. See remark below
about InnerMaxSize. */
MaySliceVectorize = MightVectorize && int(InnerMaxSize)>=3*PacketSize /* slice vectorization can be slow, so we only
want it if the slices are big, which is indicated by InnerMaxSize rather than InnerSize, think of the case
of a dynamic block in a fixed-size matrix */