Developer Guide and Reference

Contents

_mm256_srai_epi16/32

Arithmetic shift of word/doubleword elements to right according to specified number. The corresponding Intel® AVX2 instruction is
VPSRAW
or
VPSRAD
.

Syntax

extern __m256i _mm256_srai_epi16(__m256i s1, const int count);
extern __m256i _mm256_srai_epi32(__m256i s1, const int count);
Arguments
s1
integer source vector used for the operation
count
8-bit immediate used for the operation
Description
Performs an arithmetic shift of 16-bit [word] or 32-bit [doubleword] elements within a 128-bit lane of source vector
s1
to the right by the number of bits specified in
count
. The empty high-order bits are filled with the initial value of the sign bit. If the value specified by
count
is greater than 15 or 31, the whole destination vector is filled with the initial value of the sign bit. The
count
argument is an 8-bit immediate.
Returns
Result of the right-shift operation.