Intel® C++ Compiler 16.0 User and Reference Guide
Moves high and low bytes between two vector masks. Corresponding instruction is KMERGE2L1H. This intrinsic only applies to Intel® Many Integrated Core Architecture (Intel® MIC Architecture).
extern __mmask16__cdecl _mm512_kswapb(__mmask16 k1, __mmask16 k2); |
k1 |
vector mask |
k2 |
vector mask |
Moves high byte from vector mask k2 to low byte of vector mask k1 and moves low byte from vector mask k2 to high byte of vector mask k1.
Returns the result of the move operation.