Convert _mm_cmp_ss to const generics and fix imm width

This commit is contained in:
Rémy Rakic 2021-03-01 21:13:40 +01:00 committed by Amanieu d'Antras
parent 4fbb4ea13b
commit cadf5ceedc

View file

@ -871,16 +871,12 @@ pub unsafe fn _mm_cmp_sd<const IMM8: i32>(a: __m128d, b: __m128d) -> __m128d {
/// [Intel's documentation](https://software.intel.com/sites/landingpage/IntrinsicsGuide/#text=_mm_cmp_ss)
#[inline]
#[target_feature(enable = "avx,sse")]
#[cfg_attr(test, assert_instr(vcmpeqss, imm8 = 0))] // TODO Validate vcmpss
#[rustc_args_required_const(2)]
#[cfg_attr(test, assert_instr(vcmpeqss, IMM8 = 0))] // TODO Validate vcmpss
#[rustc_legacy_const_generics(2)]
#[stable(feature = "simd_x86", since = "1.27.0")]
pub unsafe fn _mm_cmp_ss(a: __m128, b: __m128, imm8: i32) -> __m128 {
macro_rules! call {
($imm8:expr) => {
vcmpss(a, b, $imm8)
};
}
constify_imm6!(imm8, call)
pub unsafe fn _mm_cmp_ss<const IMM8: i32>(a: __m128, b: __m128) -> __m128 {
static_assert_imm5!(IMM8);
vcmpss(a, b, IMM8 as i8)
}
/// Converts packed 32-bit integers in `a` to packed double-precision (64-bit)
@ -3662,7 +3658,7 @@ mod tests {
unsafe fn test_mm_cmp_ss() {
let a = _mm_setr_ps(4., 3., 2., 5.);
let b = _mm_setr_ps(4., 9., 16., 25.);
let r = _mm_cmp_ss(a, b, _CMP_GE_OS);
let r = _mm_cmp_ss::<_CMP_GE_OS>(a, b);
assert!(get_m128(r, 0).is_nan());
assert_eq!(get_m128(r, 1), 3.);
assert_eq!(get_m128(r, 2), 2.);