[−][src]Macro safe_arch::shr_imm_i16_m256i
This is supported with target feature
avx2
only.Shifts all i16
lanes left by an immediate.
let a = m256i::from([1_i16, 2, 3, 4, -1, -2, -3, -4, 1, 2, 3, 4, -1, -2, -3, -4]); let c: [i16; 16] = shr_imm_i16_m256i!(a, 1).into(); assert_eq!(c, [0_i16, 1, 1, 2, -1, -1, -2, -2, 0, 1, 1, 2, -1, -1, -2, -2]);
- Intrinsic:
_mm256_srai_epi16
- Assembly:
vpsraw ymm, ymm, imm8