mirror of
https://github.com/rust-lang/rust.git
synced 2025-10-02 10:18:25 +00:00
Change void* type for gather/scatter intrinsics
This commit is contained in:
sayantn
Amanieu d'Antras
parent
d6b0426127
commit
e804386587
@ -16515,7 +16515,10 @@ pub fn _mm512_setr_epi64(
|
||||
#[unstable(feature = "stdarch_x86_avx512", issue = "111137")]
|
||||
#[cfg_attr(test, assert_instr(vgatherdpd, SCALE = 1))]
|
||||
#[rustc_legacy_const_generics(2)]
|
||||
pub unsafe fn _mm512_i32gather_pd<const SCALE: i32>(offsets: __m256i, slice: *const u8) -> __m512d {
|
||||
pub unsafe fn _mm512_i32gather_pd<const SCALE: i32>(
|
||||
offsets: __m256i,
|
||||
slice: *const f64,
|
||||
) -> __m512d {
|
||||
static_assert_imm8_scale!(SCALE);
|
||||
let zero = f64x8::ZERO;
|
||||
let neg_one = -1;
|
||||
@ -16537,7 +16540,7 @@ pub unsafe fn _mm512_mask_i32gather_pd<const SCALE: i32>(
|
||||
src: __m512d,
|
||||
mask: __mmask8,
|
||||
offsets: __m256i,
|
||||
slice: *const u8,
|
||||
slice: *const f64,
|
||||
) -> __m512d {
|
||||
static_assert_imm8_scale!(SCALE);
|
||||
let src = src.as_f64x8();
|
||||
@ -16555,7 +16558,10 @@ pub unsafe fn _mm512_mask_i32gather_pd<const SCALE: i32>(
|
||||
#[unstable(feature = "stdarch_x86_avx512", issue = "111137")]
|
||||
#[cfg_attr(test, assert_instr(vgatherqpd, SCALE = 1))]
|
||||
#[rustc_legacy_const_generics(2)]
|
||||
pub unsafe fn _mm512_i64gather_pd<const SCALE: i32>(offsets: __m512i, slice: *const u8) -> __m512d {
|
||||
pub unsafe fn _mm512_i64gather_pd<const SCALE: i32>(
|
||||
offsets: __m512i,
|
||||
slice: *const f64,
|
||||
) -> __m512d {
|
||||
static_assert_imm8_scale!(SCALE);
|
||||
let zero = f64x8::ZERO;
|
||||
let neg_one = -1;
|
||||
@ -16577,7 +16583,7 @@ pub unsafe fn _mm512_mask_i64gather_pd<const SCALE: i32>(
|
||||
src: __m512d,
|
||||
mask: __mmask8,
|
||||
offsets: __m512i,
|
||||
slice: *const u8,
|
||||
slice: *const f64,
|
||||
) -> __m512d {
|
||||
static_assert_imm8_scale!(SCALE);
|
||||
let src = src.as_f64x8();
|
||||
@ -16595,7 +16601,7 @@ pub unsafe fn _mm512_mask_i64gather_pd<const SCALE: i32>(
|
||||
#[unstable(feature = "stdarch_x86_avx512", issue = "111137")]
|
||||
#[cfg_attr(test, assert_instr(vgatherqps, SCALE = 1))]
|
||||
#[rustc_legacy_const_generics(2)]
|
||||
pub unsafe fn _mm512_i64gather_ps<const SCALE: i32>(offsets: __m512i, slice: *const u8) -> __m256 {
|
||||
pub unsafe fn _mm512_i64gather_ps<const SCALE: i32>(offsets: __m512i, slice: *const f32) -> __m256 {
|
||||
static_assert_imm8_scale!(SCALE);
|
||||
let zero = f32x8::ZERO;
|
||||
let neg_one = -1;
|
||||
@ -16617,7 +16623,7 @@ pub unsafe fn _mm512_mask_i64gather_ps<const SCALE: i32>(
|
||||
src: __m256,
|
||||
mask: __mmask8,
|
||||
offsets: __m512i,
|
||||
slice: *const u8,
|
||||
slice: *const f32,
|
||||
) -> __m256 {
|
||||
static_assert_imm8_scale!(SCALE);
|
||||
let src = src.as_f32x8();
|
||||
@ -16635,7 +16641,7 @@ pub unsafe fn _mm512_mask_i64gather_ps<const SCALE: i32>(
|
||||
#[unstable(feature = "stdarch_x86_avx512", issue = "111137")]
|
||||
#[cfg_attr(test, assert_instr(vgatherdps, SCALE = 1))]
|
||||
#[rustc_legacy_const_generics(2)]
|
||||
pub unsafe fn _mm512_i32gather_ps<const SCALE: i32>(offsets: __m512i, slice: *const u8) -> __m512 {
|
||||
pub unsafe fn _mm512_i32gather_ps<const SCALE: i32>(offsets: __m512i, slice: *const f32) -> __m512 {
|
||||
static_assert_imm8_scale!(SCALE);
|
||||
let zero = f32x16::ZERO;
|
||||
let neg_one = -1;
|
||||
@ -16657,7 +16663,7 @@ pub unsafe fn _mm512_mask_i32gather_ps<const SCALE: i32>(
|
||||
src: __m512,
|
||||
mask: __mmask16,
|
||||
offsets: __m512i,
|
||||
slice: *const u8,
|
||||
slice: *const f32,
|
||||
) -> __m512 {
|
||||
static_assert_imm8_scale!(SCALE);
|
||||
let src = src.as_f32x16();
|
||||
@ -16677,7 +16683,7 @@ pub unsafe fn _mm512_mask_i32gather_ps<const SCALE: i32>(
|
||||
#[rustc_legacy_const_generics(2)]
|
||||
pub unsafe fn _mm512_i32gather_epi32<const SCALE: i32>(
|
||||
offsets: __m512i,
|
||||
slice: *const u8,
|
||||
slice: *const i32,
|
||||
) -> __m512i {
|
||||
static_assert_imm8_scale!(SCALE);
|
||||
let zero = i32x16::ZERO;
|
||||
@ -16700,7 +16706,7 @@ pub unsafe fn _mm512_mask_i32gather_epi32<const SCALE: i32>(
|
||||
src: __m512i,
|
||||
mask: __mmask16,
|
||||
offsets: __m512i,
|
||||
slice: *const u8,
|
||||
slice: *const i32,
|
||||
) -> __m512i {
|
||||
static_assert_imm8_scale!(SCALE);
|
||||
let src = src.as_i32x16();
|
||||
@ -16721,7 +16727,7 @@ pub unsafe fn _mm512_mask_i32gather_epi32<const SCALE: i32>(
|
||||
#[rustc_legacy_const_generics(2)]
|
||||
pub unsafe fn _mm512_i32gather_epi64<const SCALE: i32>(
|
||||
offsets: __m256i,
|
||||
slice: *const u8,
|
||||
slice: *const i64,
|
||||
) -> __m512i {
|
||||
static_assert_imm8_scale!(SCALE);
|
||||
let zero = i64x8::ZERO;
|
||||
@ -16744,7 +16750,7 @@ pub unsafe fn _mm512_mask_i32gather_epi64<const SCALE: i32>(
|
||||
src: __m512i,
|
||||
mask: __mmask8,
|
||||
offsets: __m256i,
|
||||
slice: *const u8,
|
||||
slice: *const i64,
|
||||
) -> __m512i {
|
||||
static_assert_imm8_scale!(SCALE);
|
||||
let src = src.as_i64x8();
|
||||
@ -16765,7 +16771,7 @@ pub unsafe fn _mm512_mask_i32gather_epi64<const SCALE: i32>(
|
||||
#[rustc_legacy_const_generics(2)]
|
||||
pub unsafe fn _mm512_i64gather_epi64<const SCALE: i32>(
|
||||
offsets: __m512i,
|
||||
slice: *const u8,
|
||||
slice: *const i64,
|
||||
) -> __m512i {
|
||||
static_assert_imm8_scale!(SCALE);
|
||||
let zero = i64x8::ZERO;
|
||||
@ -16788,7 +16794,7 @@ pub unsafe fn _mm512_mask_i64gather_epi64<const SCALE: i32>(
|
||||
src: __m512i,
|
||||
mask: __mmask8,
|
||||
offsets: __m512i,
|
||||
slice: *const u8,
|
||||
slice: *const i64,
|
||||
) -> __m512i {
|
||||
static_assert_imm8_scale!(SCALE);
|
||||
let src = src.as_i64x8();
|
||||
@ -16809,7 +16815,7 @@ pub unsafe fn _mm512_mask_i64gather_epi64<const SCALE: i32>(
|
||||
#[rustc_legacy_const_generics(2)]
|
||||
pub unsafe fn _mm512_i64gather_epi32<const SCALE: i32>(
|
||||
offsets: __m512i,
|
||||
slice: *const u8,
|
||||
slice: *const i32,
|
||||
) -> __m256i {
|
||||
static_assert_imm8_scale!(SCALE);
|
||||
let zeros = i32x8::ZERO;
|
||||
@ -16832,7 +16838,7 @@ pub unsafe fn _mm512_mask_i64gather_epi32<const SCALE: i32>(
|
||||
src: __m256i,
|
||||
mask: __mmask8,
|
||||
offsets: __m512i,
|
||||
slice: *const u8,
|
||||
slice: *const i32,
|
||||
) -> __m256i {
|
||||
static_assert_imm8_scale!(SCALE);
|
||||
let src = src.as_i32x8();
|
||||
@ -16852,7 +16858,7 @@ pub unsafe fn _mm512_mask_i64gather_epi32<const SCALE: i32>(
|
||||
#[cfg_attr(test, assert_instr(vscatterdpd, SCALE = 1))]
|
||||
#[rustc_legacy_const_generics(3)]
|
||||
pub unsafe fn _mm512_i32scatter_pd<const SCALE: i32>(
|
||||
slice: *mut u8,
|
||||
slice: *mut f64,
|
||||
offsets: __m256i,
|
||||
src: __m512d,
|
||||
) {
|
||||
@ -16873,7 +16879,7 @@ pub unsafe fn _mm512_i32scatter_pd<const SCALE: i32>(
|
||||
#[cfg_attr(test, assert_instr(vscatterdpd, SCALE = 1))]
|
||||
#[rustc_legacy_const_generics(4)]
|
||||
pub unsafe fn _mm512_mask_i32scatter_pd<const SCALE: i32>(
|
||||
slice: *mut u8,
|
||||
slice: *mut f64,
|
||||
mask: __mmask8,
|
||||
offsets: __m256i,
|
||||
src: __m512d,
|
||||
@ -16894,7 +16900,7 @@ pub unsafe fn _mm512_mask_i32scatter_pd<const SCALE: i32>(
|
||||
#[cfg_attr(test, assert_instr(vscatterqpd, SCALE = 1))]
|
||||
#[rustc_legacy_const_generics(3)]
|
||||
pub unsafe fn _mm512_i64scatter_pd<const SCALE: i32>(
|
||||
slice: *mut u8,
|
||||
slice: *mut f64,
|
||||
offsets: __m512i,
|
||||
src: __m512d,
|
||||
) {
|
||||
@ -16915,7 +16921,7 @@ pub unsafe fn _mm512_i64scatter_pd<const SCALE: i32>(
|
||||
#[cfg_attr(test, assert_instr(vscatterqpd, SCALE = 1))]
|
||||
#[rustc_legacy_const_generics(4)]
|
||||
pub unsafe fn _mm512_mask_i64scatter_pd<const SCALE: i32>(
|
||||
slice: *mut u8,
|
||||
slice: *mut f64,
|
||||
mask: __mmask8,
|
||||
offsets: __m512i,
|
||||
src: __m512d,
|
||||
@ -16936,7 +16942,7 @@ pub unsafe fn _mm512_mask_i64scatter_pd<const SCALE: i32>(
|
||||
#[cfg_attr(test, assert_instr(vscatterdps, SCALE = 1))]
|
||||
#[rustc_legacy_const_generics(3)]
|
||||
pub unsafe fn _mm512_i32scatter_ps<const SCALE: i32>(
|
||||
slice: *mut u8,
|
||||
slice: *mut f32,
|
||||
offsets: __m512i,
|
||||
src: __m512,
|
||||
) {
|
||||
@ -16957,7 +16963,7 @@ pub unsafe fn _mm512_i32scatter_ps<const SCALE: i32>(
|
||||
#[cfg_attr(test, assert_instr(vscatterdps, SCALE = 1))]
|
||||
#[rustc_legacy_const_generics(4)]
|
||||
pub unsafe fn _mm512_mask_i32scatter_ps<const SCALE: i32>(
|
||||
slice: *mut u8,
|
||||
slice: *mut f32,
|
||||
mask: __mmask16,
|
||||
offsets: __m512i,
|
||||
src: __m512,
|
||||
@ -16978,7 +16984,7 @@ pub unsafe fn _mm512_mask_i32scatter_ps<const SCALE: i32>(
|
||||
#[cfg_attr(test, assert_instr(vscatterqps, SCALE = 1))]
|
||||
#[rustc_legacy_const_generics(3)]
|
||||
pub unsafe fn _mm512_i64scatter_ps<const SCALE: i32>(
|
||||
slice: *mut u8,
|
||||
slice: *mut f32,
|
||||
offsets: __m512i,
|
||||
src: __m256,
|
||||
) {
|
||||
@ -16999,7 +17005,7 @@ pub unsafe fn _mm512_i64scatter_ps<const SCALE: i32>(
|
||||
#[cfg_attr(test, assert_instr(vscatterqps, SCALE = 1))]
|
||||
#[rustc_legacy_const_generics(4)]
|
||||
pub unsafe fn _mm512_mask_i64scatter_ps<const SCALE: i32>(
|
||||
slice: *mut u8,
|
||||
slice: *mut f32,
|
||||
mask: __mmask8,
|
||||
offsets: __m512i,
|
||||
src: __m256,
|
||||
@ -17020,7 +17026,7 @@ pub unsafe fn _mm512_mask_i64scatter_ps<const SCALE: i32>(
|
||||
#[cfg_attr(test, assert_instr(vpscatterdq, SCALE = 1))]
|
||||
#[rustc_legacy_const_generics(3)]
|
||||
pub unsafe fn _mm512_i32scatter_epi64<const SCALE: i32>(
|
||||
slice: *mut u8,
|
||||
slice: *mut i64,
|
||||
offsets: __m256i,
|
||||
src: __m512i,
|
||||
) {
|
||||
@ -17041,7 +17047,7 @@ pub unsafe fn _mm512_i32scatter_epi64<const SCALE: i32>(
|
||||
#[cfg_attr(test, assert_instr(vpscatterdq, SCALE = 1))]
|
||||
#[rustc_legacy_const_generics(4)]
|
||||
pub unsafe fn _mm512_mask_i32scatter_epi64<const SCALE: i32>(
|
||||
slice: *mut u8,
|
||||
slice: *mut i64,
|
||||
mask: __mmask8,
|
||||
offsets: __m256i,
|
||||
src: __m512i,
|
||||
@ -17063,7 +17069,7 @@ pub unsafe fn _mm512_mask_i32scatter_epi64<const SCALE: i32>(
|
||||
#[cfg_attr(test, assert_instr(vpscatterqq, SCALE = 1))]
|
||||
#[rustc_legacy_const_generics(3)]
|
||||
pub unsafe fn _mm512_i64scatter_epi64<const SCALE: i32>(
|
||||
slice: *mut u8,
|
||||
slice: *mut i64,
|
||||
offsets: __m512i,
|
||||
src: __m512i,
|
||||
) {
|
||||
@ -17084,7 +17090,7 @@ pub unsafe fn _mm512_i64scatter_epi64<const SCALE: i32>(
|
||||
#[cfg_attr(test, assert_instr(vpscatterqq, SCALE = 1))]
|
||||
#[rustc_legacy_const_generics(4)]
|
||||
pub unsafe fn _mm512_mask_i64scatter_epi64<const SCALE: i32>(
|
||||
slice: *mut u8,
|
||||
slice: *mut i64,
|
||||
mask: __mmask8,
|
||||
offsets: __m512i,
|
||||
src: __m512i,
|
||||
@ -17106,7 +17112,7 @@ pub unsafe fn _mm512_mask_i64scatter_epi64<const SCALE: i32>(
|
||||
#[cfg_attr(test, assert_instr(vpscatterdd, SCALE = 1))]
|
||||
#[rustc_legacy_const_generics(3)]
|
||||
pub unsafe fn _mm512_i32scatter_epi32<const SCALE: i32>(
|
||||
slice: *mut u8,
|
||||
slice: *mut i32,
|
||||
offsets: __m512i,
|
||||
src: __m512i,
|
||||
) {
|
||||
@ -17127,7 +17133,7 @@ pub unsafe fn _mm512_i32scatter_epi32<const SCALE: i32>(
|
||||
#[cfg_attr(test, assert_instr(vpscatterdd, SCALE = 1))]
|
||||
#[rustc_legacy_const_generics(4)]
|
||||
pub unsafe fn _mm512_mask_i32scatter_epi32<const SCALE: i32>(
|
||||
slice: *mut u8,
|
||||
slice: *mut i32,
|
||||
mask: __mmask16,
|
||||
offsets: __m512i,
|
||||
src: __m512i,
|
||||
@ -17149,7 +17155,7 @@ pub unsafe fn _mm512_mask_i32scatter_epi32<const SCALE: i32>(
|
||||
#[cfg_attr(test, assert_instr(vpscatterqd, SCALE = 1))]
|
||||
#[rustc_legacy_const_generics(3)]
|
||||
pub unsafe fn _mm512_i64scatter_epi32<const SCALE: i32>(
|
||||
slice: *mut u8,
|
||||
slice: *mut i32,
|
||||
offsets: __m512i,
|
||||
src: __m256i,
|
||||
) {
|
||||
@ -17170,7 +17176,7 @@ pub unsafe fn _mm512_i64scatter_epi32<const SCALE: i32>(
|
||||
#[cfg_attr(test, assert_instr(vpscatterqd, SCALE = 1))]
|
||||
#[rustc_legacy_const_generics(4)]
|
||||
pub unsafe fn _mm512_mask_i64scatter_epi32<const SCALE: i32>(
|
||||
slice: *mut u8,
|
||||
slice: *mut i32,
|
||||
mask: __mmask8,
|
||||
offsets: __m512i,
|
||||
src: __m256i,
|
||||
@ -17194,9 +17200,9 @@ pub unsafe fn _mm512_mask_i64scatter_epi32<const SCALE: i32>(
|
||||
#[unstable(feature = "stdarch_x86_avx512", issue = "111137")]
|
||||
pub unsafe fn _mm512_i32logather_epi64<const SCALE: i32>(
|
||||
vindex: __m512i,
|
||||
base_addr: *const u8,
|
||||
base_addr: *const i64,
|
||||
) -> __m512i {
|
||||
_mm512_i32gather_epi64::<SCALE>(_mm512_castsi512_si256(vindex), base_addr as _)
|
||||
_mm512_i32gather_epi64::<SCALE>(_mm512_castsi512_si256(vindex), base_addr)
|
||||
}
|
||||
|
||||
/// Loads 8 64-bit integer elements from memory starting at location base_addr at packed 32-bit integer
|
||||
@ -17213,9 +17219,9 @@ pub unsafe fn _mm512_mask_i32logather_epi64<const SCALE: i32>(
|
||||
src: __m512i,
|
||||
k: __mmask8,
|
||||
vindex: __m512i,
|
||||
base_addr: *const u8,
|
||||
base_addr: *const i64,
|
||||
) -> __m512i {
|
||||
_mm512_mask_i32gather_epi64::<SCALE>(src, k, _mm512_castsi512_si256(vindex), base_addr as _)
|
||||
_mm512_mask_i32gather_epi64::<SCALE>(src, k, _mm512_castsi512_si256(vindex), base_addr)
|
||||
}
|
||||
|
||||
/// Loads 8 double-precision (64-bit) floating-point elements from memory starting at location base_addr
|
||||
@ -17229,9 +17235,9 @@ pub unsafe fn _mm512_mask_i32logather_epi64<const SCALE: i32>(
|
||||
#[unstable(feature = "stdarch_x86_avx512", issue = "111137")]
|
||||
pub unsafe fn _mm512_i32logather_pd<const SCALE: i32>(
|
||||
vindex: __m512i,
|
||||
base_addr: *const u8,
|
||||
base_addr: *const f64,
|
||||
) -> __m512d {
|
||||
_mm512_i32gather_pd::<SCALE>(_mm512_castsi512_si256(vindex), base_addr as _)
|
||||
_mm512_i32gather_pd::<SCALE>(_mm512_castsi512_si256(vindex), base_addr)
|
||||
}
|
||||
|
||||
/// Loads 8 double-precision (64-bit) floating-point elements from memory starting at location base_addr
|
||||
@ -17248,9 +17254,9 @@ pub unsafe fn _mm512_mask_i32logather_pd<const SCALE: i32>(
|
||||
src: __m512d,
|
||||
k: __mmask8,
|
||||
vindex: __m512i,
|
||||
base_addr: *const u8,
|
||||
base_addr: *const f64,
|
||||
) -> __m512d {
|
||||
_mm512_mask_i32gather_pd::<SCALE>(src, k, _mm512_castsi512_si256(vindex), base_addr as _)
|
||||
_mm512_mask_i32gather_pd::<SCALE>(src, k, _mm512_castsi512_si256(vindex), base_addr)
|
||||
}
|
||||
|
||||
/// Stores 8 64-bit integer elements from a to memory starting at location base_addr at packed 32-bit integer
|
||||
@ -17263,11 +17269,11 @@ pub unsafe fn _mm512_mask_i32logather_pd<const SCALE: i32>(
|
||||
#[rustc_legacy_const_generics(3)]
|
||||
#[unstable(feature = "stdarch_x86_avx512", issue = "111137")]
|
||||
pub unsafe fn _mm512_i32loscatter_epi64<const SCALE: i32>(
|
||||
base_addr: *mut u8,
|
||||
base_addr: *mut i64,
|
||||
vindex: __m512i,
|
||||
a: __m512i,
|
||||
) {
|
||||
_mm512_i32scatter_epi64::<SCALE>(base_addr as _, _mm512_castsi512_si256(vindex), a)
|
||||
_mm512_i32scatter_epi64::<SCALE>(base_addr, _mm512_castsi512_si256(vindex), a)
|
||||
}
|
||||
|
||||
/// Stores 8 64-bit integer elements from a to memory starting at location base_addr at packed 32-bit integer
|
||||
@ -17281,12 +17287,12 @@ pub unsafe fn _mm512_i32loscatter_epi64<const SCALE: i32>(
|
||||
#[rustc_legacy_const_generics(4)]
|
||||
#[unstable(feature = "stdarch_x86_avx512", issue = "111137")]
|
||||
pub unsafe fn _mm512_mask_i32loscatter_epi64<const SCALE: i32>(
|
||||
base_addr: *mut u8,
|
||||
base_addr: *mut i64,
|
||||
k: __mmask8,
|
||||
vindex: __m512i,
|
||||
a: __m512i,
|
||||
) {
|
||||
_mm512_mask_i32scatter_epi64::<SCALE>(base_addr as _, k, _mm512_castsi512_si256(vindex), a)
|
||||
_mm512_mask_i32scatter_epi64::<SCALE>(base_addr, k, _mm512_castsi512_si256(vindex), a)
|
||||
}
|
||||
|
||||
/// Stores 8 double-precision (64-bit) floating-point elements from a to memory starting at location base_addr
|
||||
@ -17299,11 +17305,11 @@ pub unsafe fn _mm512_mask_i32loscatter_epi64<const SCALE: i32>(
|
||||
#[rustc_legacy_const_generics(3)]
|
||||
#[unstable(feature = "stdarch_x86_avx512", issue = "111137")]
|
||||
pub unsafe fn _mm512_i32loscatter_pd<const SCALE: i32>(
|
||||
base_addr: *mut u8,
|
||||
base_addr: *mut f64,
|
||||
vindex: __m512i,
|
||||
a: __m512d,
|
||||
) {
|
||||
_mm512_i32scatter_pd::<SCALE>(base_addr as _, _mm512_castsi512_si256(vindex), a)
|
||||
_mm512_i32scatter_pd::<SCALE>(base_addr, _mm512_castsi512_si256(vindex), a)
|
||||
}
|
||||
|
||||
/// Stores 8 double-precision (64-bit) floating-point elements from a to memory starting at location base_addr
|
||||
@ -17317,12 +17323,12 @@ pub unsafe fn _mm512_i32loscatter_pd<const SCALE: i32>(
|
||||
#[rustc_legacy_const_generics(4)]
|
||||
#[unstable(feature = "stdarch_x86_avx512", issue = "111137")]
|
||||
pub unsafe fn _mm512_mask_i32loscatter_pd<const SCALE: i32>(
|
||||
base_addr: *mut u8,
|
||||
base_addr: *mut f64,
|
||||
k: __mmask8,
|
||||
vindex: __m512i,
|
||||
a: __m512d,
|
||||
) {
|
||||
_mm512_mask_i32scatter_pd::<SCALE>(base_addr as _, k, _mm512_castsi512_si256(vindex), a)
|
||||
_mm512_mask_i32scatter_pd::<SCALE>(base_addr, k, _mm512_castsi512_si256(vindex), a)
|
||||
}
|
||||
|
||||
/// Stores 8 32-bit integer elements from a to memory starting at location base_addr at packed 32-bit integer
|
||||
@ -17335,7 +17341,7 @@ pub unsafe fn _mm512_mask_i32loscatter_pd<const SCALE: i32>(
|
||||
#[rustc_legacy_const_generics(3)]
|
||||
#[unstable(feature = "stdarch_x86_avx512", issue = "111137")]
|
||||
pub unsafe fn _mm256_i32scatter_epi32<const SCALE: i32>(
|
||||
base_addr: *mut u8,
|
||||
base_addr: *mut i32,
|
||||
vindex: __m256i,
|
||||
a: __m256i,
|
||||
) {
|
||||
@ -17354,7 +17360,7 @@ pub unsafe fn _mm256_i32scatter_epi32<const SCALE: i32>(
|
||||
#[rustc_legacy_const_generics(4)]
|
||||
#[unstable(feature = "stdarch_x86_avx512", issue = "111137")]
|
||||
pub unsafe fn _mm256_mask_i32scatter_epi32<const SCALE: i32>(
|
||||
base_addr: *mut u8,
|
||||
base_addr: *mut i32,
|
||||
k: __mmask8,
|
||||
vindex: __m256i,
|
||||
a: __m256i,
|
||||
@ -17372,7 +17378,7 @@ pub unsafe fn _mm256_mask_i32scatter_epi32<const SCALE: i32>(
|
||||
#[cfg_attr(test, assert_instr(vpscatterdq, SCALE = 1))]
|
||||
#[rustc_legacy_const_generics(3)]
|
||||
pub unsafe fn _mm256_i32scatter_epi64<const SCALE: i32>(
|
||||
slice: *mut u8,
|
||||
slice: *mut i64,
|
||||
offsets: __m128i,
|
||||
src: __m256i,
|
||||
) {
|
||||
@ -17394,7 +17400,7 @@ pub unsafe fn _mm256_i32scatter_epi64<const SCALE: i32>(
|
||||
#[rustc_legacy_const_generics(4)]
|
||||
#[unstable(feature = "stdarch_x86_avx512", issue = "111137")]
|
||||
pub unsafe fn _mm256_mask_i32scatter_epi64<const SCALE: i32>(
|
||||
base_addr: *mut u8,
|
||||
base_addr: *mut i64,
|
||||
k: __mmask8,
|
||||
vindex: __m128i,
|
||||
a: __m256i,
|
||||
@ -17413,7 +17419,7 @@ pub unsafe fn _mm256_mask_i32scatter_epi64<const SCALE: i32>(
|
||||
#[rustc_legacy_const_generics(3)]
|
||||
#[unstable(feature = "stdarch_x86_avx512", issue = "111137")]
|
||||
pub unsafe fn _mm256_i32scatter_pd<const SCALE: i32>(
|
||||
base_addr: *mut u8,
|
||||
base_addr: *mut f64,
|
||||
vindex: __m128i,
|
||||
a: __m256d,
|
||||
) {
|
||||
@ -17432,7 +17438,7 @@ pub unsafe fn _mm256_i32scatter_pd<const SCALE: i32>(
|
||||
#[rustc_legacy_const_generics(4)]
|
||||
#[unstable(feature = "stdarch_x86_avx512", issue = "111137")]
|
||||
pub unsafe fn _mm256_mask_i32scatter_pd<const SCALE: i32>(
|
||||
base_addr: *mut u8,
|
||||
base_addr: *mut f64,
|
||||
k: __mmask8,
|
||||
vindex: __m128i,
|
||||
a: __m256d,
|
||||
@ -17451,7 +17457,7 @@ pub unsafe fn _mm256_mask_i32scatter_pd<const SCALE: i32>(
|
||||
#[rustc_legacy_const_generics(3)]
|
||||
#[unstable(feature = "stdarch_x86_avx512", issue = "111137")]
|
||||
pub unsafe fn _mm256_i32scatter_ps<const SCALE: i32>(
|
||||
base_addr: *mut u8,
|
||||
base_addr: *mut f32,
|
||||
vindex: __m256i,
|
||||
a: __m256,
|
||||
) {
|
||||
@ -17470,7 +17476,7 @@ pub unsafe fn _mm256_i32scatter_ps<const SCALE: i32>(
|
||||
#[rustc_legacy_const_generics(4)]
|
||||
#[unstable(feature = "stdarch_x86_avx512", issue = "111137")]
|
||||
pub unsafe fn _mm256_mask_i32scatter_ps<const SCALE: i32>(
|
||||
base_addr: *mut u8,
|
||||
base_addr: *mut f32,
|
||||
k: __mmask8,
|
||||
vindex: __m256i,
|
||||
a: __m256,
|
||||
@ -17489,7 +17495,7 @@ pub unsafe fn _mm256_mask_i32scatter_ps<const SCALE: i32>(
|
||||
#[rustc_legacy_const_generics(3)]
|
||||
#[unstable(feature = "stdarch_x86_avx512", issue = "111137")]
|
||||
pub unsafe fn _mm256_i64scatter_epi32<const SCALE: i32>(
|
||||
base_addr: *mut u8,
|
||||
base_addr: *mut i32,
|
||||
vindex: __m256i,
|
||||
a: __m128i,
|
||||
) {
|
||||
@ -17508,7 +17514,7 @@ pub unsafe fn _mm256_i64scatter_epi32<const SCALE: i32>(
|
||||
#[rustc_legacy_const_generics(4)]
|
||||
#[unstable(feature = "stdarch_x86_avx512", issue = "111137")]
|
||||
pub unsafe fn _mm256_mask_i64scatter_epi32<const SCALE: i32>(
|
||||
base_addr: *mut u8,
|
||||
base_addr: *mut i32,
|
||||
k: __mmask8,
|
||||
vindex: __m256i,
|
||||
a: __m128i,
|
||||
@ -17527,7 +17533,7 @@ pub unsafe fn _mm256_mask_i64scatter_epi32<const SCALE: i32>(
|
||||
#[rustc_legacy_const_generics(3)]
|
||||
#[unstable(feature = "stdarch_x86_avx512", issue = "111137")]
|
||||
pub unsafe fn _mm256_i64scatter_epi64<const SCALE: i32>(
|
||||
base_addr: *mut u8,
|
||||
base_addr: *mut i64,
|
||||
vindex: __m256i,
|
||||
a: __m256i,
|
||||
) {
|
||||
@ -17546,7 +17552,7 @@ pub unsafe fn _mm256_i64scatter_epi64<const SCALE: i32>(
|
||||
#[rustc_legacy_const_generics(4)]
|
||||
#[unstable(feature = "stdarch_x86_avx512", issue = "111137")]
|
||||
pub unsafe fn _mm256_mask_i64scatter_epi64<const SCALE: i32>(
|
||||
base_addr: *mut u8,
|
||||
base_addr: *mut i64,
|
||||
k: __mmask8,
|
||||
vindex: __m256i,
|
||||
a: __m256i,
|
||||
@ -17565,7 +17571,7 @@ pub unsafe fn _mm256_mask_i64scatter_epi64<const SCALE: i32>(
|
||||
#[rustc_legacy_const_generics(3)]
|
||||
#[unstable(feature = "stdarch_x86_avx512", issue = "111137")]
|
||||
pub unsafe fn _mm256_i64scatter_pd<const SCALE: i32>(
|
||||
base_addr: *mut u8,
|
||||
base_addr: *mut f64,
|
||||
vindex: __m256i,
|
||||
a: __m256d,
|
||||
) {
|
||||
@ -17584,7 +17590,7 @@ pub unsafe fn _mm256_i64scatter_pd<const SCALE: i32>(
|
||||
#[rustc_legacy_const_generics(4)]
|
||||
#[unstable(feature = "stdarch_x86_avx512", issue = "111137")]
|
||||
pub unsafe fn _mm256_mask_i64scatter_pd<const SCALE: i32>(
|
||||
base_addr: *mut u8,
|
||||
base_addr: *mut f64,
|
||||
k: __mmask8,
|
||||
vindex: __m256i,
|
||||
a: __m256d,
|
||||
@ -17603,7 +17609,7 @@ pub unsafe fn _mm256_mask_i64scatter_pd<const SCALE: i32>(
|
||||
#[rustc_legacy_const_generics(3)]
|
||||
#[unstable(feature = "stdarch_x86_avx512", issue = "111137")]
|
||||
pub unsafe fn _mm256_i64scatter_ps<const SCALE: i32>(
|
||||
base_addr: *mut u8,
|
||||
base_addr: *mut f32,
|
||||
vindex: __m256i,
|
||||
a: __m128,
|
||||
) {
|
||||
@ -17622,7 +17628,7 @@ pub unsafe fn _mm256_i64scatter_ps<const SCALE: i32>(
|
||||
#[rustc_legacy_const_generics(4)]
|
||||
#[unstable(feature = "stdarch_x86_avx512", issue = "111137")]
|
||||
pub unsafe fn _mm256_mask_i64scatter_ps<const SCALE: i32>(
|
||||
base_addr: *mut u8,
|
||||
base_addr: *mut f32,
|
||||
k: __mmask8,
|
||||
vindex: __m256i,
|
||||
a: __m128,
|
||||
@ -17645,7 +17651,7 @@ pub unsafe fn _mm256_mmask_i32gather_epi32<const SCALE: i32>(
|
||||
src: __m256i,
|
||||
k: __mmask8,
|
||||
vindex: __m256i,
|
||||
base_addr: *const u8,
|
||||
base_addr: *const i32,
|
||||
) -> __m256i {
|
||||
static_assert_imm8_scale!(SCALE);
|
||||
transmute(vpgatherdd_256(
|
||||
@ -17671,7 +17677,7 @@ pub unsafe fn _mm256_mmask_i32gather_epi64<const SCALE: i32>(
|
||||
src: __m256i,
|
||||
k: __mmask8,
|
||||
vindex: __m128i,
|
||||
base_addr: *const u8,
|
||||
base_addr: *const i64,
|
||||
) -> __m256i {
|
||||
static_assert_imm8_scale!(SCALE);
|
||||
transmute(vpgatherdq_256(
|
||||
@ -17697,7 +17703,7 @@ pub unsafe fn _mm256_mmask_i32gather_pd<const SCALE: i32>(
|
||||
src: __m256d,
|
||||
k: __mmask8,
|
||||
vindex: __m128i,
|
||||
base_addr: *const u8,
|
||||
base_addr: *const f64,
|
||||
) -> __m256d {
|
||||
static_assert_imm8_scale!(SCALE);
|
||||
transmute(vgatherdpd_256(
|
||||
@ -17723,7 +17729,7 @@ pub unsafe fn _mm256_mmask_i32gather_ps<const SCALE: i32>(
|
||||
src: __m256,
|
||||
k: __mmask8,
|
||||
vindex: __m256i,
|
||||
base_addr: *const u8,
|
||||
base_addr: *const f32,
|
||||
) -> __m256 {
|
||||
static_assert_imm8_scale!(SCALE);
|
||||
transmute(vgatherdps_256(
|
||||
@ -17749,7 +17755,7 @@ pub unsafe fn _mm256_mmask_i64gather_epi32<const SCALE: i32>(
|
||||
src: __m128i,
|
||||
k: __mmask8,
|
||||
vindex: __m256i,
|
||||
base_addr: *const u8,
|
||||
base_addr: *const i32,
|
||||
) -> __m128i {
|
||||
static_assert_imm8_scale!(SCALE);
|
||||
transmute(vpgatherqd_256(
|
||||
@ -17775,7 +17781,7 @@ pub unsafe fn _mm256_mmask_i64gather_epi64<const SCALE: i32>(
|
||||
src: __m256i,
|
||||
k: __mmask8,
|
||||
vindex: __m256i,
|
||||
base_addr: *const u8,
|
||||
base_addr: *const i64,
|
||||
) -> __m256i {
|
||||
static_assert_imm8_scale!(SCALE);
|
||||
transmute(vpgatherqq_256(
|
||||
@ -17801,7 +17807,7 @@ pub unsafe fn _mm256_mmask_i64gather_pd<const SCALE: i32>(
|
||||
src: __m256d,
|
||||
k: __mmask8,
|
||||
vindex: __m256i,
|
||||
base_addr: *const u8,
|
||||
base_addr: *const f64,
|
||||
) -> __m256d {
|
||||
static_assert_imm8_scale!(SCALE);
|
||||
transmute(vgatherqpd_256(
|
||||
@ -17827,7 +17833,7 @@ pub unsafe fn _mm256_mmask_i64gather_ps<const SCALE: i32>(
|
||||
src: __m128,
|
||||
k: __mmask8,
|
||||
vindex: __m256i,
|
||||
base_addr: *const u8,
|
||||
base_addr: *const f32,
|
||||
) -> __m128 {
|
||||
static_assert_imm8_scale!(SCALE);
|
||||
transmute(vgatherqps_256(
|
||||
@ -17849,7 +17855,7 @@ pub unsafe fn _mm256_mmask_i64gather_ps<const SCALE: i32>(
|
||||
#[rustc_legacy_const_generics(3)]
|
||||
#[unstable(feature = "stdarch_x86_avx512", issue = "111137")]
|
||||
pub unsafe fn _mm_i32scatter_epi32<const SCALE: i32>(
|
||||
base_addr: *mut u8,
|
||||
base_addr: *mut i32,
|
||||
vindex: __m128i,
|
||||
a: __m128i,
|
||||
) {
|
||||
@ -17868,7 +17874,7 @@ pub unsafe fn _mm_i32scatter_epi32<const SCALE: i32>(
|
||||
#[rustc_legacy_const_generics(4)]
|
||||
#[unstable(feature = "stdarch_x86_avx512", issue = "111137")]
|
||||
pub unsafe fn _mm_mask_i32scatter_epi32<const SCALE: i32>(
|
||||
base_addr: *mut u8,
|
||||
base_addr: *mut i32,
|
||||
k: __mmask8,
|
||||
vindex: __m128i,
|
||||
a: __m128i,
|
||||
@ -17887,7 +17893,7 @@ pub unsafe fn _mm_mask_i32scatter_epi32<const SCALE: i32>(
|
||||
#[rustc_legacy_const_generics(3)]
|
||||
#[unstable(feature = "stdarch_x86_avx512", issue = "111137")]
|
||||
pub unsafe fn _mm_i32scatter_epi64<const SCALE: i32>(
|
||||
base_addr: *mut u8,
|
||||
base_addr: *mut i64,
|
||||
vindex: __m128i,
|
||||
a: __m128i,
|
||||
) {
|
||||
@ -17906,7 +17912,7 @@ pub unsafe fn _mm_i32scatter_epi64<const SCALE: i32>(
|
||||
#[rustc_legacy_const_generics(4)]
|
||||
#[unstable(feature = "stdarch_x86_avx512", issue = "111137")]
|
||||
pub unsafe fn _mm_mask_i32scatter_epi64<const SCALE: i32>(
|
||||
base_addr: *mut u8,
|
||||
base_addr: *mut i64,
|
||||
k: __mmask8,
|
||||
vindex: __m128i,
|
||||
a: __m128i,
|
||||
@ -17924,7 +17930,11 @@ pub unsafe fn _mm_mask_i32scatter_epi64<const SCALE: i32>(
|
||||
#[cfg_attr(test, assert_instr(vscatterdpd, SCALE = 1))]
|
||||
#[rustc_legacy_const_generics(3)]
|
||||
#[unstable(feature = "stdarch_x86_avx512", issue = "111137")]
|
||||
pub unsafe fn _mm_i32scatter_pd<const SCALE: i32>(base_addr: *mut u8, vindex: __m128i, a: __m128d) {
|
||||
pub unsafe fn _mm_i32scatter_pd<const SCALE: i32>(
|
||||
base_addr: *mut f64,
|
||||
vindex: __m128i,
|
||||
a: __m128d,
|
||||
) {
|
||||
static_assert_imm8_scale!(SCALE);
|
||||
vscatterdpd_128(base_addr as _, 0xff, vindex.as_i32x4(), a.as_f64x2(), SCALE)
|
||||
}
|
||||
@ -17940,7 +17950,7 @@ pub unsafe fn _mm_i32scatter_pd<const SCALE: i32>(base_addr: *mut u8, vindex: __
|
||||
#[rustc_legacy_const_generics(4)]
|
||||
#[unstable(feature = "stdarch_x86_avx512", issue = "111137")]
|
||||
pub unsafe fn _mm_mask_i32scatter_pd<const SCALE: i32>(
|
||||
base_addr: *mut u8,
|
||||
base_addr: *mut f64,
|
||||
k: __mmask8,
|
||||
vindex: __m128i,
|
||||
a: __m128d,
|
||||
@ -17958,7 +17968,7 @@ pub unsafe fn _mm_mask_i32scatter_pd<const SCALE: i32>(
|
||||
#[cfg_attr(test, assert_instr(vscatterdps, SCALE = 1))]
|
||||
#[rustc_legacy_const_generics(3)]
|
||||
#[unstable(feature = "stdarch_x86_avx512", issue = "111137")]
|
||||
pub unsafe fn _mm_i32scatter_ps<const SCALE: i32>(base_addr: *mut u8, vindex: __m128i, a: __m128) {
|
||||
pub unsafe fn _mm_i32scatter_ps<const SCALE: i32>(base_addr: *mut f32, vindex: __m128i, a: __m128) {
|
||||
static_assert_imm8_scale!(SCALE);
|
||||
vscatterdps_128(base_addr as _, 0xff, vindex.as_i32x4(), a.as_f32x4(), SCALE)
|
||||
}
|
||||
@ -17974,7 +17984,7 @@ pub unsafe fn _mm_i32scatter_ps<const SCALE: i32>(base_addr: *mut u8, vindex: __
|
||||
#[rustc_legacy_const_generics(4)]
|
||||
#[unstable(feature = "stdarch_x86_avx512", issue = "111137")]
|
||||
pub unsafe fn _mm_mask_i32scatter_ps<const SCALE: i32>(
|
||||
base_addr: *mut u8,
|
||||
base_addr: *mut f32,
|
||||
k: __mmask8,
|
||||
vindex: __m128i,
|
||||
a: __m128,
|
||||
@ -17993,7 +18003,7 @@ pub unsafe fn _mm_mask_i32scatter_ps<const SCALE: i32>(
|
||||
#[rustc_legacy_const_generics(3)]
|
||||
#[unstable(feature = "stdarch_x86_avx512", issue = "111137")]
|
||||
pub unsafe fn _mm_i64scatter_epi32<const SCALE: i32>(
|
||||
base_addr: *mut u8,
|
||||
base_addr: *mut i32,
|
||||
vindex: __m128i,
|
||||
a: __m128i,
|
||||
) {
|
||||
@ -18012,7 +18022,7 @@ pub unsafe fn _mm_i64scatter_epi32<const SCALE: i32>(
|
||||
#[rustc_legacy_const_generics(4)]
|
||||
#[unstable(feature = "stdarch_x86_avx512", issue = "111137")]
|
||||
pub unsafe fn _mm_mask_i64scatter_epi32<const SCALE: i32>(
|
||||
base_addr: *mut u8,
|
||||
base_addr: *mut i32,
|
||||
k: __mmask8,
|
||||
vindex: __m128i,
|
||||
a: __m128i,
|
||||
@ -18031,7 +18041,7 @@ pub unsafe fn _mm_mask_i64scatter_epi32<const SCALE: i32>(
|
||||
#[rustc_legacy_const_generics(3)]
|
||||
#[unstable(feature = "stdarch_x86_avx512", issue = "111137")]
|
||||
pub unsafe fn _mm_i64scatter_epi64<const SCALE: i32>(
|
||||
base_addr: *mut u8,
|
||||
base_addr: *mut i64,
|
||||
vindex: __m128i,
|
||||
a: __m128i,
|
||||
) {
|
||||
@ -18050,7 +18060,7 @@ pub unsafe fn _mm_i64scatter_epi64<const SCALE: i32>(
|
||||
#[rustc_legacy_const_generics(4)]
|
||||
#[unstable(feature = "stdarch_x86_avx512", issue = "111137")]
|
||||
pub unsafe fn _mm_mask_i64scatter_epi64<const SCALE: i32>(
|
||||
base_addr: *mut u8,
|
||||
base_addr: *mut i64,
|
||||
k: __mmask8,
|
||||
vindex: __m128i,
|
||||
a: __m128i,
|
||||
@ -18068,7 +18078,11 @@ pub unsafe fn _mm_mask_i64scatter_epi64<const SCALE: i32>(
|
||||
#[cfg_attr(test, assert_instr(vscatterqpd, SCALE = 1))]
|
||||
#[rustc_legacy_const_generics(3)]
|
||||
#[unstable(feature = "stdarch_x86_avx512", issue = "111137")]
|
||||
pub unsafe fn _mm_i64scatter_pd<const SCALE: i32>(base_addr: *mut u8, vindex: __m128i, a: __m128d) {
|
||||
pub unsafe fn _mm_i64scatter_pd<const SCALE: i32>(
|
||||
base_addr: *mut f64,
|
||||
vindex: __m128i,
|
||||
a: __m128d,
|
||||
) {
|
||||
static_assert_imm8_scale!(SCALE);
|
||||
vscatterqpd_128(base_addr as _, 0xff, vindex.as_i64x2(), a.as_f64x2(), SCALE)
|
||||
}
|
||||
@ -18084,7 +18098,7 @@ pub unsafe fn _mm_i64scatter_pd<const SCALE: i32>(base_addr: *mut u8, vindex: __
|
||||
#[rustc_legacy_const_generics(4)]
|
||||
#[unstable(feature = "stdarch_x86_avx512", issue = "111137")]
|
||||
pub unsafe fn _mm_mask_i64scatter_pd<const SCALE: i32>(
|
||||
base_addr: *mut u8,
|
||||
base_addr: *mut f64,
|
||||
k: __mmask8,
|
||||
vindex: __m128i,
|
||||
a: __m128d,
|
||||
@ -18102,7 +18116,7 @@ pub unsafe fn _mm_mask_i64scatter_pd<const SCALE: i32>(
|
||||
#[cfg_attr(test, assert_instr(vscatterqps, SCALE = 1))]
|
||||
#[rustc_legacy_const_generics(3)]
|
||||
#[unstable(feature = "stdarch_x86_avx512", issue = "111137")]
|
||||
pub unsafe fn _mm_i64scatter_ps<const SCALE: i32>(base_addr: *mut u8, vindex: __m128i, a: __m128) {
|
||||
pub unsafe fn _mm_i64scatter_ps<const SCALE: i32>(base_addr: *mut f32, vindex: __m128i, a: __m128) {
|
||||
static_assert_imm8_scale!(SCALE);
|
||||
vscatterqps_128(base_addr as _, 0xff, vindex.as_i64x2(), a.as_f32x4(), SCALE)
|
||||
}
|
||||
@ -18117,7 +18131,7 @@ pub unsafe fn _mm_i64scatter_ps<const SCALE: i32>(base_addr: *mut u8, vindex: __
|
||||
#[rustc_legacy_const_generics(4)]
|
||||
#[unstable(feature = "stdarch_x86_avx512", issue = "111137")]
|
||||
pub unsafe fn _mm_mask_i64scatter_ps<const SCALE: i32>(
|
||||
base_addr: *mut u8,
|
||||
base_addr: *mut f32,
|
||||
k: __mmask8,
|
||||
vindex: __m128i,
|
||||
a: __m128,
|
||||
@ -18140,7 +18154,7 @@ pub unsafe fn _mm_mmask_i32gather_epi32<const SCALE: i32>(
|
||||
src: __m128i,
|
||||
k: __mmask8,
|
||||
vindex: __m128i,
|
||||
base_addr: *const u8,
|
||||
base_addr: *const i32,
|
||||
) -> __m128i {
|
||||
static_assert_imm8_scale!(SCALE);
|
||||
transmute(vpgatherdd_128(
|
||||
@ -18166,7 +18180,7 @@ pub unsafe fn _mm_mmask_i32gather_epi64<const SCALE: i32>(
|
||||
src: __m128i,
|
||||
k: __mmask8,
|
||||
vindex: __m128i,
|
||||
base_addr: *const u8,
|
||||
base_addr: *const i64,
|
||||
) -> __m128i {
|
||||
static_assert_imm8_scale!(SCALE);
|
||||
transmute(vpgatherdq_128(
|
||||
@ -18192,7 +18206,7 @@ pub unsafe fn _mm_mmask_i32gather_pd<const SCALE: i32>(
|
||||
src: __m128d,
|
||||
k: __mmask8,
|
||||
vindex: __m128i,
|
||||
base_addr: *const u8,
|
||||
base_addr: *const f64,
|
||||
) -> __m128d {
|
||||
static_assert_imm8_scale!(SCALE);
|
||||
transmute(vgatherdpd_128(
|
||||
@ -18218,7 +18232,7 @@ pub unsafe fn _mm_mmask_i32gather_ps<const SCALE: i32>(
|
||||
src: __m128,
|
||||
k: __mmask8,
|
||||
vindex: __m128i,
|
||||
base_addr: *const u8,
|
||||
base_addr: *const f32,
|
||||
) -> __m128 {
|
||||
static_assert_imm8_scale!(SCALE);
|
||||
transmute(vgatherdps_128(
|
||||
@ -18244,7 +18258,7 @@ pub unsafe fn _mm_mmask_i64gather_epi32<const SCALE: i32>(
|
||||
src: __m128i,
|
||||
k: __mmask8,
|
||||
vindex: __m128i,
|
||||
base_addr: *const u8,
|
||||
base_addr: *const i32,
|
||||
) -> __m128i {
|
||||
static_assert_imm8_scale!(SCALE);
|
||||
transmute(vpgatherqd_128(
|
||||
@ -18270,7 +18284,7 @@ pub unsafe fn _mm_mmask_i64gather_epi64<const SCALE: i32>(
|
||||
src: __m128i,
|
||||
k: __mmask8,
|
||||
vindex: __m128i,
|
||||
base_addr: *const u8,
|
||||
base_addr: *const i64,
|
||||
) -> __m128i {
|
||||
static_assert_imm8_scale!(SCALE);
|
||||
transmute(vpgatherqq_128(
|
||||
@ -18296,7 +18310,7 @@ pub unsafe fn _mm_mmask_i64gather_pd<const SCALE: i32>(
|
||||
src: __m128d,
|
||||
k: __mmask8,
|
||||
vindex: __m128i,
|
||||
base_addr: *const u8,
|
||||
base_addr: *const f64,
|
||||
) -> __m128d {
|
||||
static_assert_imm8_scale!(SCALE);
|
||||
transmute(vgatherqpd_128(
|
||||
@ -18322,7 +18336,7 @@ pub unsafe fn _mm_mmask_i64gather_ps<const SCALE: i32>(
|
||||
src: __m128,
|
||||
k: __mmask8,
|
||||
vindex: __m128i,
|
||||
base_addr: *const u8,
|
||||
base_addr: *const f32,
|
||||
) -> __m128 {
|
||||
static_assert_imm8_scale!(SCALE);
|
||||
transmute(vgatherqps_128(
|
||||
@ -49637,7 +49651,7 @@ mod tests {
|
||||
#[rustfmt::skip]
|
||||
let index = _mm512_setr_epi32(0, 16, 32, 48, 64, 80, 96, 112,
|
||||
120, 128, 136, 144, 152, 160, 168, 176);
|
||||
let r = _mm512_i32gather_ps::<4>(index, arr.as_ptr() as *const u8);
|
||||
let r = _mm512_i32gather_ps::<4>(index, arr.as_ptr());
|
||||
#[rustfmt::skip]
|
||||
assert_eq_m512(r, _mm512_setr_ps(0., 16., 32., 48., 64., 80., 96., 112.,
|
||||
120., 128., 136., 144., 152., 160., 168., 176.));
|
||||
@ -49652,7 +49666,7 @@ mod tests {
|
||||
let index = _mm512_setr_epi32(0, 16, 32, 48, 64, 80, 96, 112,
|
||||
120, 128, 136, 144, 152, 160, 168, 176);
|
||||
// A multiplier of 4 is word-addressing
|
||||
let r = _mm512_mask_i32gather_ps::<4>(src, mask, index, arr.as_ptr() as *const u8);
|
||||
let r = _mm512_mask_i32gather_ps::<4>(src, mask, index, arr.as_ptr());
|
||||
#[rustfmt::skip]
|
||||
assert_eq_m512(r, _mm512_setr_ps(2., 16., 2., 48., 2., 80., 2., 112.,
|
||||
2., 128., 2., 144., 2., 160., 2., 176.));
|
||||
@ -49665,7 +49679,7 @@ mod tests {
|
||||
#[rustfmt::skip]
|
||||
let index = _mm512_setr_epi32(0, 16, 32, 48, 64, 80, 96, 112,
|
||||
120, 128, 136, 144, 152, 160, 168, 176);
|
||||
let r = _mm512_i32gather_epi32::<4>(index, arr.as_ptr() as *const u8);
|
||||
let r = _mm512_i32gather_epi32::<4>(index, arr.as_ptr());
|
||||
#[rustfmt::skip]
|
||||
assert_eq_m512i(r, _mm512_setr_epi32(0, 16, 32, 48, 64, 80, 96, 112,
|
||||
120, 128, 136, 144, 152, 160, 168, 176));
|
||||
@ -49680,7 +49694,7 @@ mod tests {
|
||||
0, 16, 32, 48, 64, 80, 96, 112, 128, 144, 160, 176, 192, 208, 224, 240,
|
||||
);
|
||||
// A multiplier of 4 is word-addressing
|
||||
let r = _mm512_mask_i32gather_epi32::<4>(src, mask, index, arr.as_ptr() as *const u8);
|
||||
let r = _mm512_mask_i32gather_epi32::<4>(src, mask, index, arr.as_ptr());
|
||||
assert_eq_m512i(
|
||||
r,
|
||||
_mm512_setr_epi32(2, 16, 2, 48, 2, 80, 2, 112, 2, 144, 2, 176, 2, 208, 2, 240),
|
||||
@ -49697,7 +49711,7 @@ mod tests {
|
||||
1., 2., 3., 4., 5., 6., 7., 8., 9., 10., 11., 12., 13., 14., 15., 16.,
|
||||
);
|
||||
// A multiplier of 4 is word-addressing
|
||||
_mm512_i32scatter_ps::<4>(arr.as_mut_ptr() as *mut u8, index, src);
|
||||
_mm512_i32scatter_ps::<4>(arr.as_mut_ptr(), index, src);
|
||||
let mut expected = [0f32; 256];
|
||||
for i in 0..16 {
|
||||
expected[i * 16] = (i + 1) as f32;
|
||||
@ -49716,7 +49730,7 @@ mod tests {
|
||||
1., 2., 3., 4., 5., 6., 7., 8., 9., 10., 11., 12., 13., 14., 15., 16.,
|
||||
);
|
||||
// A multiplier of 4 is word-addressing
|
||||
_mm512_mask_i32scatter_ps::<4>(arr.as_mut_ptr() as *mut u8, mask, index, src);
|
||||
_mm512_mask_i32scatter_ps::<4>(arr.as_mut_ptr(), mask, index, src);
|
||||
let mut expected = [0f32; 256];
|
||||
for i in 0..8 {
|
||||
expected[i * 32 + 16] = 2. * (i + 1) as f32;
|
||||
@ -49733,7 +49747,7 @@ mod tests {
|
||||
128, 144, 160, 176, 192, 208, 224, 240);
|
||||
let src = _mm512_setr_epi32(1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16);
|
||||
// A multiplier of 4 is word-addressing
|
||||
_mm512_i32scatter_epi32::<4>(arr.as_mut_ptr() as *mut u8, index, src);
|
||||
_mm512_i32scatter_epi32::<4>(arr.as_mut_ptr(), index, src);
|
||||
let mut expected = [0i32; 256];
|
||||
for i in 0..16 {
|
||||
expected[i * 16] = (i + 1) as i32;
|
||||
@ -49750,7 +49764,7 @@ mod tests {
|
||||
128, 144, 160, 176, 192, 208, 224, 240);
|
||||
let src = _mm512_setr_epi32(1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16);
|
||||
// A multiplier of 4 is word-addressing
|
||||
_mm512_mask_i32scatter_epi32::<4>(arr.as_mut_ptr() as *mut u8, mask, index, src);
|
||||
_mm512_mask_i32scatter_epi32::<4>(arr.as_mut_ptr(), mask, index, src);
|
||||
let mut expected = [0i32; 256];
|
||||
for i in 0..8 {
|
||||
expected[i * 32 + 16] = 2 * (i + 1) as i32;
|
||||
|
||||
@ -7441,7 +7441,7 @@ mod tests {
|
||||
let arr: [f64; 128] = core::array::from_fn(|i| i as f64);
|
||||
// A multiplier of 8 is word-addressing
|
||||
let index = _mm256_setr_epi32(0, 16, 32, 48, 64, 80, 96, 112);
|
||||
let r = _mm512_i32gather_pd::<8>(index, arr.as_ptr() as *const u8);
|
||||
let r = _mm512_i32gather_pd::<8>(index, arr.as_ptr());
|
||||
assert_eq_m512d(r, _mm512_setr_pd(0., 16., 32., 48., 64., 80., 96., 112.));
|
||||
}
|
||||
|
||||
@ -7452,7 +7452,7 @@ mod tests {
|
||||
let mask = 0b10101010;
|
||||
let index = _mm256_setr_epi32(0, 16, 32, 48, 64, 80, 96, 112);
|
||||
// A multiplier of 8 is word-addressing
|
||||
let r = _mm512_mask_i32gather_pd::<8>(src, mask, index, arr.as_ptr() as *const u8);
|
||||
let r = _mm512_mask_i32gather_pd::<8>(src, mask, index, arr.as_ptr());
|
||||
assert_eq_m512d(r, _mm512_setr_pd(2., 16., 2., 48., 2., 80., 2., 112.));
|
||||
}
|
||||
|
||||
@ -7461,7 +7461,7 @@ mod tests {
|
||||
let arr: [f64; 128] = core::array::from_fn(|i| i as f64);
|
||||
// A multiplier of 8 is word-addressing
|
||||
let index = _mm512_setr_epi64(0, 16, 32, 48, 64, 80, 96, 112);
|
||||
let r = _mm512_i64gather_pd::<8>(index, arr.as_ptr() as *const u8);
|
||||
let r = _mm512_i64gather_pd::<8>(index, arr.as_ptr());
|
||||
assert_eq_m512d(r, _mm512_setr_pd(0., 16., 32., 48., 64., 80., 96., 112.));
|
||||
}
|
||||
|
||||
@ -7472,7 +7472,7 @@ mod tests {
|
||||
let mask = 0b10101010;
|
||||
let index = _mm512_setr_epi64(0, 16, 32, 48, 64, 80, 96, 112);
|
||||
// A multiplier of 8 is word-addressing
|
||||
let r = _mm512_mask_i64gather_pd::<8>(src, mask, index, arr.as_ptr() as *const u8);
|
||||
let r = _mm512_mask_i64gather_pd::<8>(src, mask, index, arr.as_ptr());
|
||||
assert_eq_m512d(r, _mm512_setr_pd(2., 16., 2., 48., 2., 80., 2., 112.));
|
||||
}
|
||||
|
||||
@ -7482,7 +7482,7 @@ mod tests {
|
||||
// A multiplier of 4 is word-addressing
|
||||
#[rustfmt::skip]
|
||||
let index = _mm512_setr_epi64(0, 16, 32, 48, 64, 80, 96, 112);
|
||||
let r = _mm512_i64gather_ps::<4>(index, arr.as_ptr() as *const u8);
|
||||
let r = _mm512_i64gather_ps::<4>(index, arr.as_ptr());
|
||||
assert_eq_m256(r, _mm256_setr_ps(0., 16., 32., 48., 64., 80., 96., 112.));
|
||||
}
|
||||
|
||||
@ -7494,7 +7494,7 @@ mod tests {
|
||||
#[rustfmt::skip]
|
||||
let index = _mm512_setr_epi64(0, 16, 32, 48, 64, 80, 96, 112);
|
||||
// A multiplier of 4 is word-addressing
|
||||
let r = _mm512_mask_i64gather_ps::<4>(src, mask, index, arr.as_ptr() as *const u8);
|
||||
let r = _mm512_mask_i64gather_ps::<4>(src, mask, index, arr.as_ptr());
|
||||
assert_eq_m256(r, _mm256_setr_ps(2., 16., 2., 48., 2., 80., 2., 112.));
|
||||
}
|
||||
|
||||
@ -7506,7 +7506,7 @@ mod tests {
|
||||
}
|
||||
// A multiplier of 8 is word-addressing
|
||||
let index = _mm256_setr_epi32(0, 16, 32, 48, 64, 80, 96, 112);
|
||||
let r = _mm512_i32gather_epi64::<8>(index, arr.as_ptr() as *const u8);
|
||||
let r = _mm512_i32gather_epi64::<8>(index, arr.as_ptr());
|
||||
assert_eq_m512i(r, _mm512_setr_epi64(0, 16, 32, 48, 64, 80, 96, 112));
|
||||
}
|
||||
|
||||
@ -7520,7 +7520,7 @@ mod tests {
|
||||
let mask = 0b10101010;
|
||||
let index = _mm256_setr_epi32(0, 16, 32, 48, 64, 80, 96, 112);
|
||||
// A multiplier of 8 is word-addressing
|
||||
let r = _mm512_mask_i32gather_epi64::<8>(src, mask, index, arr.as_ptr() as *const u8);
|
||||
let r = _mm512_mask_i32gather_epi64::<8>(src, mask, index, arr.as_ptr());
|
||||
assert_eq_m512i(r, _mm512_setr_epi64(2, 16, 2, 48, 2, 80, 2, 112));
|
||||
}
|
||||
|
||||
@ -7532,7 +7532,7 @@ mod tests {
|
||||
}
|
||||
// A multiplier of 8 is word-addressing
|
||||
let index = _mm512_setr_epi64(0, 16, 32, 48, 64, 80, 96, 112);
|
||||
let r = _mm512_i64gather_epi64::<8>(index, arr.as_ptr() as *const u8);
|
||||
let r = _mm512_i64gather_epi64::<8>(index, arr.as_ptr());
|
||||
assert_eq_m512i(r, _mm512_setr_epi64(0, 16, 32, 48, 64, 80, 96, 112));
|
||||
}
|
||||
|
||||
@ -7546,7 +7546,7 @@ mod tests {
|
||||
let mask = 0b10101010;
|
||||
let index = _mm512_setr_epi64(0, 16, 32, 48, 64, 80, 96, 112);
|
||||
// A multiplier of 8 is word-addressing
|
||||
let r = _mm512_mask_i64gather_epi64::<8>(src, mask, index, arr.as_ptr() as *const u8);
|
||||
let r = _mm512_mask_i64gather_epi64::<8>(src, mask, index, arr.as_ptr());
|
||||
assert_eq_m512i(r, _mm512_setr_epi64(2, 16, 2, 48, 2, 80, 2, 112));
|
||||
}
|
||||
|
||||
@ -7558,7 +7558,7 @@ mod tests {
|
||||
}
|
||||
// A multiplier of 8 is word-addressing
|
||||
let index = _mm512_setr_epi64(0, 16, 32, 48, 64, 80, 96, 112);
|
||||
let r = _mm512_i64gather_epi32::<8>(index, arr.as_ptr() as *const u8);
|
||||
let r = _mm512_i64gather_epi32::<8>(index, arr.as_ptr() as *const i32);
|
||||
assert_eq_m256i(r, _mm256_setr_epi32(0, 16, 32, 48, 64, 80, 96, 112));
|
||||
}
|
||||
|
||||
@ -7572,7 +7572,7 @@ mod tests {
|
||||
let mask = 0b10101010;
|
||||
let index = _mm512_setr_epi64(0, 16, 32, 48, 64, 80, 96, 112);
|
||||
// A multiplier of 8 is word-addressing
|
||||
let r = _mm512_mask_i64gather_epi32::<8>(src, mask, index, arr.as_ptr() as *const u8);
|
||||
let r = _mm512_mask_i64gather_epi32::<8>(src, mask, index, arr.as_ptr() as *const i32);
|
||||
assert_eq_m256i(r, _mm256_setr_epi32(2, 16, 2, 48, 2, 80, 2, 112));
|
||||
}
|
||||
|
||||
@ -7582,7 +7582,7 @@ mod tests {
|
||||
let index = _mm256_setr_epi32(0, 16, 32, 48, 64, 80, 96, 112);
|
||||
let src = _mm512_setr_pd(1., 2., 3., 4., 5., 6., 7., 8.);
|
||||
// A multiplier of 8 is word-addressing
|
||||
_mm512_i32scatter_pd::<8>(arr.as_mut_ptr() as *mut u8, index, src);
|
||||
_mm512_i32scatter_pd::<8>(arr.as_mut_ptr(), index, src);
|
||||
let mut expected = [0f64; 128];
|
||||
for i in 0..8 {
|
||||
expected[i * 16] = (i + 1) as f64;
|
||||
@ -7597,7 +7597,7 @@ mod tests {
|
||||
let index = _mm256_setr_epi32(0, 16, 32, 48, 64, 80, 96, 112);
|
||||
let src = _mm512_setr_pd(1., 2., 3., 4., 5., 6., 7., 8.);
|
||||
// A multiplier of 8 is word-addressing
|
||||
_mm512_mask_i32scatter_pd::<8>(arr.as_mut_ptr() as *mut u8, mask, index, src);
|
||||
_mm512_mask_i32scatter_pd::<8>(arr.as_mut_ptr(), mask, index, src);
|
||||
let mut expected = [0f64; 128];
|
||||
for i in 0..4 {
|
||||
expected[i * 32 + 16] = 2. * (i + 1) as f64;
|
||||
@ -7611,7 +7611,7 @@ mod tests {
|
||||
let index = _mm512_setr_epi64(0, 16, 32, 48, 64, 80, 96, 112);
|
||||
let src = _mm512_setr_pd(1., 2., 3., 4., 5., 6., 7., 8.);
|
||||
// A multiplier of 8 is word-addressing
|
||||
_mm512_i64scatter_pd::<8>(arr.as_mut_ptr() as *mut u8, index, src);
|
||||
_mm512_i64scatter_pd::<8>(arr.as_mut_ptr(), index, src);
|
||||
let mut expected = [0f64; 128];
|
||||
for i in 0..8 {
|
||||
expected[i * 16] = (i + 1) as f64;
|
||||
@ -7626,7 +7626,7 @@ mod tests {
|
||||
let index = _mm512_setr_epi64(0, 16, 32, 48, 64, 80, 96, 112);
|
||||
let src = _mm512_setr_pd(1., 2., 3., 4., 5., 6., 7., 8.);
|
||||
// A multiplier of 8 is word-addressing
|
||||
_mm512_mask_i64scatter_pd::<8>(arr.as_mut_ptr() as *mut u8, mask, index, src);
|
||||
_mm512_mask_i64scatter_pd::<8>(arr.as_mut_ptr(), mask, index, src);
|
||||
let mut expected = [0f64; 128];
|
||||
for i in 0..4 {
|
||||
expected[i * 32 + 16] = 2. * (i + 1) as f64;
|
||||
@ -7640,7 +7640,7 @@ mod tests {
|
||||
let index = _mm512_setr_epi64(0, 16, 32, 48, 64, 80, 96, 112);
|
||||
let src = _mm256_setr_ps(1., 2., 3., 4., 5., 6., 7., 8.);
|
||||
// A multiplier of 4 is word-addressing
|
||||
_mm512_i64scatter_ps::<4>(arr.as_mut_ptr() as *mut u8, index, src);
|
||||
_mm512_i64scatter_ps::<4>(arr.as_mut_ptr(), index, src);
|
||||
let mut expected = [0f32; 128];
|
||||
for i in 0..8 {
|
||||
expected[i * 16] = (i + 1) as f32;
|
||||
@ -7655,7 +7655,7 @@ mod tests {
|
||||
let index = _mm512_setr_epi64(0, 16, 32, 48, 64, 80, 96, 112);
|
||||
let src = _mm256_setr_ps(1., 2., 3., 4., 5., 6., 7., 8.);
|
||||
// A multiplier of 4 is word-addressing
|
||||
_mm512_mask_i64scatter_ps::<4>(arr.as_mut_ptr() as *mut u8, mask, index, src);
|
||||
_mm512_mask_i64scatter_ps::<4>(arr.as_mut_ptr(), mask, index, src);
|
||||
let mut expected = [0f32; 128];
|
||||
for i in 0..4 {
|
||||
expected[i * 32 + 16] = 2. * (i + 1) as f32;
|
||||
@ -7669,7 +7669,7 @@ mod tests {
|
||||
let index = _mm256_setr_epi32(0, 16, 32, 48, 64, 80, 96, 112);
|
||||
let src = _mm512_setr_epi64(1, 2, 3, 4, 5, 6, 7, 8);
|
||||
// A multiplier of 8 is word-addressing
|
||||
_mm512_i32scatter_epi64::<8>(arr.as_mut_ptr() as *mut u8, index, src);
|
||||
_mm512_i32scatter_epi64::<8>(arr.as_mut_ptr(), index, src);
|
||||
let mut expected = [0i64; 128];
|
||||
for i in 0..8 {
|
||||
expected[i * 16] = (i + 1) as i64;
|
||||
@ -7684,7 +7684,7 @@ mod tests {
|
||||
let index = _mm256_setr_epi32(0, 16, 32, 48, 64, 80, 96, 112);
|
||||
let src = _mm512_setr_epi64(1, 2, 3, 4, 5, 6, 7, 8);
|
||||
// A multiplier of 8 is word-addressing
|
||||
_mm512_mask_i32scatter_epi64::<8>(arr.as_mut_ptr() as *mut u8, mask, index, src);
|
||||
_mm512_mask_i32scatter_epi64::<8>(arr.as_mut_ptr(), mask, index, src);
|
||||
let mut expected = [0i64; 128];
|
||||
for i in 0..4 {
|
||||
expected[i * 32 + 16] = 2 * (i + 1) as i64;
|
||||
@ -7698,7 +7698,7 @@ mod tests {
|
||||
let index = _mm512_setr_epi64(0, 16, 32, 48, 64, 80, 96, 112);
|
||||
let src = _mm512_setr_epi64(1, 2, 3, 4, 5, 6, 7, 8);
|
||||
// A multiplier of 8 is word-addressing
|
||||
_mm512_i64scatter_epi64::<8>(arr.as_mut_ptr() as *mut u8, index, src);
|
||||
_mm512_i64scatter_epi64::<8>(arr.as_mut_ptr(), index, src);
|
||||
let mut expected = [0i64; 128];
|
||||
for i in 0..8 {
|
||||
expected[i * 16] = (i + 1) as i64;
|
||||
@ -7713,7 +7713,7 @@ mod tests {
|
||||
let index = _mm512_setr_epi64(0, 16, 32, 48, 64, 80, 96, 112);
|
||||
let src = _mm512_setr_epi64(1, 2, 3, 4, 5, 6, 7, 8);
|
||||
// A multiplier of 8 is word-addressing
|
||||
_mm512_mask_i64scatter_epi64::<8>(arr.as_mut_ptr() as *mut u8, mask, index, src);
|
||||
_mm512_mask_i64scatter_epi64::<8>(arr.as_mut_ptr(), mask, index, src);
|
||||
let mut expected = [0i64; 128];
|
||||
for i in 0..4 {
|
||||
expected[i * 32 + 16] = 2 * (i + 1) as i64;
|
||||
@ -7727,7 +7727,7 @@ mod tests {
|
||||
let index = _mm512_setr_epi64(0, 16, 32, 48, 64, 80, 96, 112);
|
||||
let src = _mm256_setr_epi32(1, 2, 3, 4, 5, 6, 7, 8);
|
||||
// A multiplier of 4 is word-addressing
|
||||
_mm512_i64scatter_epi32::<4>(arr.as_mut_ptr() as *mut u8, index, src);
|
||||
_mm512_i64scatter_epi32::<4>(arr.as_mut_ptr(), index, src);
|
||||
let mut expected = [0i32; 128];
|
||||
for i in 0..8 {
|
||||
expected[i * 16] = (i + 1) as i32;
|
||||
@ -7742,7 +7742,7 @@ mod tests {
|
||||
let index = _mm512_setr_epi64(0, 16, 32, 48, 64, 80, 96, 112);
|
||||
let src = _mm256_setr_epi32(1, 2, 3, 4, 5, 6, 7, 8);
|
||||
// A multiplier of 4 is word-addressing
|
||||
_mm512_mask_i64scatter_epi32::<4>(arr.as_mut_ptr() as *mut u8, mask, index, src);
|
||||
_mm512_mask_i64scatter_epi32::<4>(arr.as_mut_ptr(), mask, index, src);
|
||||
let mut expected = [0i32; 128];
|
||||
for i in 0..4 {
|
||||
expected[i * 32 + 16] = 2 * (i + 1) as i32;
|
||||
@ -7754,7 +7754,7 @@ mod tests {
|
||||
unsafe fn test_mm512_i32logather_epi64() {
|
||||
let base_addr: [i64; 8] = [1, 2, 3, 4, 5, 6, 7, 8];
|
||||
let vindex = _mm512_setr_epi32(1, 2, 3, 4, 5, 6, 7, 0, -1, -1, -1, -1, -1, -1, -1, -1);
|
||||
let r = _mm512_i32logather_epi64::<8>(vindex, base_addr.as_ptr().cast());
|
||||
let r = _mm512_i32logather_epi64::<8>(vindex, base_addr.as_ptr());
|
||||
let expected = _mm512_setr_epi64(2, 3, 4, 5, 6, 7, 8, 1);
|
||||
assert_eq_m512i(expected, r);
|
||||
}
|
||||
@ -7764,8 +7764,7 @@ mod tests {
|
||||
let base_addr: [i64; 8] = [1, 2, 3, 4, 5, 6, 7, 8];
|
||||
let src = _mm512_setr_epi64(9, 10, 11, 12, 13, 14, 15, 16);
|
||||
let vindex = _mm512_setr_epi32(1, 2, 3, 4, 5, 6, 7, 0, -1, -1, -1, -1, -1, -1, -1, -1);
|
||||
let r =
|
||||
_mm512_mask_i32logather_epi64::<8>(src, 0b01010101, vindex, base_addr.as_ptr().cast());
|
||||
let r = _mm512_mask_i32logather_epi64::<8>(src, 0b01010101, vindex, base_addr.as_ptr());
|
||||
let expected = _mm512_setr_epi64(2, 10, 4, 12, 6, 14, 8, 16);
|
||||
assert_eq_m512i(expected, r);
|
||||
}
|
||||
@ -7774,7 +7773,7 @@ mod tests {
|
||||
unsafe fn test_mm512_i32logather_pd() {
|
||||
let base_addr: [f64; 8] = [1., 2., 3., 4., 5., 6., 7., 8.];
|
||||
let vindex = _mm512_setr_epi32(1, 2, 3, 4, 5, 6, 7, 0, -1, -1, -1, -1, -1, -1, -1, -1);
|
||||
let r = _mm512_i32logather_pd::<8>(vindex, base_addr.as_ptr().cast());
|
||||
let r = _mm512_i32logather_pd::<8>(vindex, base_addr.as_ptr());
|
||||
let expected = _mm512_setr_pd(2., 3., 4., 5., 6., 7., 8., 1.);
|
||||
assert_eq_m512d(expected, r);
|
||||
}
|
||||
@ -7784,7 +7783,7 @@ mod tests {
|
||||
let base_addr: [f64; 8] = [1., 2., 3., 4., 5., 6., 7., 8.];
|
||||
let src = _mm512_setr_pd(9., 10., 11., 12., 13., 14., 15., 16.);
|
||||
let vindex = _mm512_setr_epi32(1, 2, 3, 4, 5, 6, 7, 0, -1, -1, -1, -1, -1, -1, -1, -1);
|
||||
let r = _mm512_mask_i32logather_pd::<8>(src, 0b01010101, vindex, base_addr.as_ptr().cast());
|
||||
let r = _mm512_mask_i32logather_pd::<8>(src, 0b01010101, vindex, base_addr.as_ptr());
|
||||
let expected = _mm512_setr_pd(2., 10., 4., 12., 6., 14., 8., 16.);
|
||||
assert_eq_m512d(expected, r);
|
||||
}
|
||||
@ -7794,7 +7793,7 @@ mod tests {
|
||||
let mut base_addr: [i64; 8] = [0; 8];
|
||||
let vindex = _mm512_setr_epi32(1, 2, 3, 4, 5, 6, 7, 0, -1, -1, -1, -1, -1, -1, -1, -1);
|
||||
let src = _mm512_setr_epi64(2, 3, 4, 5, 6, 7, 8, 1);
|
||||
_mm512_i32loscatter_epi64::<8>(base_addr.as_mut_ptr().cast(), vindex, src);
|
||||
_mm512_i32loscatter_epi64::<8>(base_addr.as_mut_ptr(), vindex, src);
|
||||
let expected = [1, 2, 3, 4, 5, 6, 7, 8];
|
||||
assert_eq!(expected, base_addr);
|
||||
}
|
||||
@ -7804,7 +7803,7 @@ mod tests {
|
||||
let mut base_addr: [i64; 8] = [0; 8];
|
||||
let vindex = _mm512_setr_epi32(1, 2, 3, 4, 5, 6, 7, 0, -1, -1, -1, -1, -1, -1, -1, -1);
|
||||
let src = _mm512_setr_epi64(2, 3, 4, 5, 6, 7, 8, 1);
|
||||
_mm512_mask_i32loscatter_epi64::<8>(base_addr.as_mut_ptr().cast(), 0b01010101, vindex, src);
|
||||
_mm512_mask_i32loscatter_epi64::<8>(base_addr.as_mut_ptr(), 0b01010101, vindex, src);
|
||||
let expected = [0, 2, 0, 4, 0, 6, 0, 8];
|
||||
assert_eq!(expected, base_addr);
|
||||
}
|
||||
@ -7814,7 +7813,7 @@ mod tests {
|
||||
let mut base_addr: [f64; 8] = [0.; 8];
|
||||
let vindex = _mm512_setr_epi32(1, 2, 3, 4, 5, 6, 7, 0, -1, -1, -1, -1, -1, -1, -1, -1);
|
||||
let src = _mm512_setr_pd(2., 3., 4., 5., 6., 7., 8., 1.);
|
||||
_mm512_i32loscatter_pd::<8>(base_addr.as_mut_ptr().cast(), vindex, src);
|
||||
_mm512_i32loscatter_pd::<8>(base_addr.as_mut_ptr(), vindex, src);
|
||||
let expected = [1., 2., 3., 4., 5., 6., 7., 8.];
|
||||
assert_eq!(expected, base_addr);
|
||||
}
|
||||
@ -7824,7 +7823,7 @@ mod tests {
|
||||
let mut base_addr: [f64; 8] = [0.; 8];
|
||||
let vindex = _mm512_setr_epi32(1, 2, 3, 4, 5, 6, 7, 0, -1, -1, -1, -1, -1, -1, -1, -1);
|
||||
let src = _mm512_setr_pd(2., 3., 4., 5., 6., 7., 8., 1.);
|
||||
_mm512_mask_i32loscatter_pd::<8>(base_addr.as_mut_ptr().cast(), 0b01010101, vindex, src);
|
||||
_mm512_mask_i32loscatter_pd::<8>(base_addr.as_mut_ptr(), 0b01010101, vindex, src);
|
||||
let expected = [0., 2., 0., 4., 0., 6., 0., 8.];
|
||||
assert_eq!(expected, base_addr);
|
||||
}
|
||||
@ -7834,7 +7833,7 @@ mod tests {
|
||||
let base_addr: [i32; 4] = [1, 2, 3, 4];
|
||||
let src = _mm_setr_epi32(5, 6, 7, 8);
|
||||
let vindex = _mm_setr_epi32(1, 2, 3, 0);
|
||||
let r = _mm_mmask_i32gather_epi32::<4>(src, 0b0101, vindex, base_addr.as_ptr().cast());
|
||||
let r = _mm_mmask_i32gather_epi32::<4>(src, 0b0101, vindex, base_addr.as_ptr());
|
||||
let expected = _mm_setr_epi32(2, 6, 4, 8);
|
||||
assert_eq_m128i(expected, r);
|
||||
}
|
||||
@ -7844,7 +7843,7 @@ mod tests {
|
||||
let base_addr: [i64; 2] = [1, 2];
|
||||
let src = _mm_setr_epi64x(5, 6);
|
||||
let vindex = _mm_setr_epi32(1, 0, -1, -1);
|
||||
let r = _mm_mmask_i32gather_epi64::<8>(src, 0b01, vindex, base_addr.as_ptr().cast());
|
||||
let r = _mm_mmask_i32gather_epi64::<8>(src, 0b01, vindex, base_addr.as_ptr());
|
||||
let expected = _mm_setr_epi64x(2, 6);
|
||||
assert_eq_m128i(expected, r);
|
||||
}
|
||||
@ -7854,7 +7853,7 @@ mod tests {
|
||||
let base_addr: [f64; 2] = [1., 2.];
|
||||
let src = _mm_setr_pd(5., 6.);
|
||||
let vindex = _mm_setr_epi32(1, 0, -1, -1);
|
||||
let r = _mm_mmask_i32gather_pd::<8>(src, 0b01, vindex, base_addr.as_ptr().cast());
|
||||
let r = _mm_mmask_i32gather_pd::<8>(src, 0b01, vindex, base_addr.as_ptr());
|
||||
let expected = _mm_setr_pd(2., 6.);
|
||||
assert_eq_m128d(expected, r);
|
||||
}
|
||||
@ -7864,7 +7863,7 @@ mod tests {
|
||||
let base_addr: [f32; 4] = [1., 2., 3., 4.];
|
||||
let src = _mm_setr_ps(5., 6., 7., 8.);
|
||||
let vindex = _mm_setr_epi32(1, 2, 3, 0);
|
||||
let r = _mm_mmask_i32gather_ps::<4>(src, 0b0101, vindex, base_addr.as_ptr().cast());
|
||||
let r = _mm_mmask_i32gather_ps::<4>(src, 0b0101, vindex, base_addr.as_ptr());
|
||||
let expected = _mm_setr_ps(2., 6., 4., 8.);
|
||||
assert_eq_m128(expected, r);
|
||||
}
|
||||
@ -7874,7 +7873,7 @@ mod tests {
|
||||
let base_addr: [i32; 2] = [1, 2];
|
||||
let src = _mm_setr_epi32(5, 6, 7, 8);
|
||||
let vindex = _mm_setr_epi64x(1, 0);
|
||||
let r = _mm_mmask_i64gather_epi32::<4>(src, 0b01, vindex, base_addr.as_ptr().cast());
|
||||
let r = _mm_mmask_i64gather_epi32::<4>(src, 0b01, vindex, base_addr.as_ptr());
|
||||
let expected = _mm_setr_epi32(2, 6, 0, 0);
|
||||
assert_eq_m128i(expected, r);
|
||||
}
|
||||
@ -7884,7 +7883,7 @@ mod tests {
|
||||
let base_addr: [i64; 2] = [1, 2];
|
||||
let src = _mm_setr_epi64x(5, 6);
|
||||
let vindex = _mm_setr_epi64x(1, 0);
|
||||
let r = _mm_mmask_i64gather_epi64::<8>(src, 0b01, vindex, base_addr.as_ptr().cast());
|
||||
let r = _mm_mmask_i64gather_epi64::<8>(src, 0b01, vindex, base_addr.as_ptr());
|
||||
let expected = _mm_setr_epi64x(2, 6);
|
||||
assert_eq_m128i(expected, r);
|
||||
}
|
||||
@ -7894,7 +7893,7 @@ mod tests {
|
||||
let base_addr: [f64; 2] = [1., 2.];
|
||||
let src = _mm_setr_pd(5., 6.);
|
||||
let vindex = _mm_setr_epi64x(1, 0);
|
||||
let r = _mm_mmask_i64gather_pd::<8>(src, 0b01, vindex, base_addr.as_ptr().cast());
|
||||
let r = _mm_mmask_i64gather_pd::<8>(src, 0b01, vindex, base_addr.as_ptr());
|
||||
let expected = _mm_setr_pd(2., 6.);
|
||||
assert_eq_m128d(expected, r);
|
||||
}
|
||||
@ -7904,7 +7903,7 @@ mod tests {
|
||||
let base_addr: [f32; 2] = [1., 2.];
|
||||
let src = _mm_setr_ps(5., 6., 7., 8.);
|
||||
let vindex = _mm_setr_epi64x(1, 0);
|
||||
let r = _mm_mmask_i64gather_ps::<4>(src, 0b01, vindex, base_addr.as_ptr().cast());
|
||||
let r = _mm_mmask_i64gather_ps::<4>(src, 0b01, vindex, base_addr.as_ptr());
|
||||
let expected = _mm_setr_ps(2., 6., 0., 0.);
|
||||
assert_eq_m128(expected, r);
|
||||
}
|
||||
@ -7914,8 +7913,7 @@ mod tests {
|
||||
let base_addr: [i32; 8] = [1, 2, 3, 4, 5, 6, 7, 8];
|
||||
let src = _mm256_setr_epi32(9, 10, 11, 12, 13, 14, 15, 16);
|
||||
let vindex = _mm256_setr_epi32(1, 2, 3, 4, 5, 6, 7, 0);
|
||||
let r =
|
||||
_mm256_mmask_i32gather_epi32::<4>(src, 0b01010101, vindex, base_addr.as_ptr().cast());
|
||||
let r = _mm256_mmask_i32gather_epi32::<4>(src, 0b01010101, vindex, base_addr.as_ptr());
|
||||
let expected = _mm256_setr_epi32(2, 10, 4, 12, 6, 14, 8, 16);
|
||||
assert_eq_m256i(expected, r);
|
||||
}
|
||||
@ -7925,7 +7923,7 @@ mod tests {
|
||||
let base_addr: [i64; 4] = [1, 2, 3, 4];
|
||||
let src = _mm256_setr_epi64x(9, 10, 11, 12);
|
||||
let vindex = _mm_setr_epi32(1, 2, 3, 4);
|
||||
let r = _mm256_mmask_i32gather_epi64::<8>(src, 0b0101, vindex, base_addr.as_ptr().cast());
|
||||
let r = _mm256_mmask_i32gather_epi64::<8>(src, 0b0101, vindex, base_addr.as_ptr());
|
||||
let expected = _mm256_setr_epi64x(2, 10, 4, 12);
|
||||
assert_eq_m256i(expected, r);
|
||||
}
|
||||
@ -7935,7 +7933,7 @@ mod tests {
|
||||
let base_addr: [f64; 4] = [1., 2., 3., 4.];
|
||||
let src = _mm256_setr_pd(9., 10., 11., 12.);
|
||||
let vindex = _mm_setr_epi32(1, 2, 3, 4);
|
||||
let r = _mm256_mmask_i32gather_pd::<8>(src, 0b0101, vindex, base_addr.as_ptr().cast());
|
||||
let r = _mm256_mmask_i32gather_pd::<8>(src, 0b0101, vindex, base_addr.as_ptr());
|
||||
let expected = _mm256_setr_pd(2., 10., 4., 12.);
|
||||
assert_eq_m256d(expected, r);
|
||||
}
|
||||
@ -7945,7 +7943,7 @@ mod tests {
|
||||
let base_addr: [f32; 8] = [1., 2., 3., 4., 5., 6., 7., 8.];
|
||||
let src = _mm256_setr_ps(9., 10., 11., 12., 13., 14., 15., 16.);
|
||||
let vindex = _mm256_setr_epi32(1, 2, 3, 4, 5, 6, 7, 0);
|
||||
let r = _mm256_mmask_i32gather_ps::<4>(src, 0b01010101, vindex, base_addr.as_ptr().cast());
|
||||
let r = _mm256_mmask_i32gather_ps::<4>(src, 0b01010101, vindex, base_addr.as_ptr());
|
||||
let expected = _mm256_setr_ps(2., 10., 4., 12., 6., 14., 8., 16.);
|
||||
assert_eq_m256(expected, r);
|
||||
}
|
||||
@ -7955,7 +7953,7 @@ mod tests {
|
||||
let base_addr: [i32; 4] = [1, 2, 3, 4];
|
||||
let src = _mm_setr_epi32(9, 10, 11, 12);
|
||||
let vindex = _mm256_setr_epi64x(1, 2, 3, 0);
|
||||
let r = _mm256_mmask_i64gather_epi32::<4>(src, 0b0101, vindex, base_addr.as_ptr().cast());
|
||||
let r = _mm256_mmask_i64gather_epi32::<4>(src, 0b0101, vindex, base_addr.as_ptr());
|
||||
let expected = _mm_setr_epi32(2, 10, 4, 12);
|
||||
assert_eq_m128i(expected, r);
|
||||
}
|
||||
@ -7965,7 +7963,7 @@ mod tests {
|
||||
let base_addr: [i64; 4] = [1, 2, 3, 4];
|
||||
let src = _mm256_setr_epi64x(9, 10, 11, 12);
|
||||
let vindex = _mm256_setr_epi64x(1, 2, 3, 0);
|
||||
let r = _mm256_mmask_i64gather_epi64::<8>(src, 0b0101, vindex, base_addr.as_ptr().cast());
|
||||
let r = _mm256_mmask_i64gather_epi64::<8>(src, 0b0101, vindex, base_addr.as_ptr());
|
||||
let expected = _mm256_setr_epi64x(2, 10, 4, 12);
|
||||
assert_eq_m256i(expected, r);
|
||||
}
|
||||
@ -7975,7 +7973,7 @@ mod tests {
|
||||
let base_addr: [f64; 4] = [1., 2., 3., 4.];
|
||||
let src = _mm256_setr_pd(9., 10., 11., 12.);
|
||||
let vindex = _mm256_setr_epi64x(1, 2, 3, 0);
|
||||
let r = _mm256_mmask_i64gather_pd::<8>(src, 0b0101, vindex, base_addr.as_ptr().cast());
|
||||
let r = _mm256_mmask_i64gather_pd::<8>(src, 0b0101, vindex, base_addr.as_ptr());
|
||||
let expected = _mm256_setr_pd(2., 10., 4., 12.);
|
||||
assert_eq_m256d(expected, r);
|
||||
}
|
||||
@ -7985,7 +7983,7 @@ mod tests {
|
||||
let base_addr: [f32; 4] = [1., 2., 3., 4.];
|
||||
let src = _mm_setr_ps(9., 10., 11., 12.);
|
||||
let vindex = _mm256_setr_epi64x(1, 2, 3, 0);
|
||||
let r = _mm256_mmask_i64gather_ps::<4>(src, 0b0101, vindex, base_addr.as_ptr().cast());
|
||||
let r = _mm256_mmask_i64gather_ps::<4>(src, 0b0101, vindex, base_addr.as_ptr());
|
||||
let expected = _mm_setr_ps(2., 10., 4., 12.);
|
||||
assert_eq_m128(expected, r);
|
||||
}
|
||||
@ -7995,7 +7993,7 @@ mod tests {
|
||||
let mut base_addr: [i32; 4] = [0; 4];
|
||||
let vindex = _mm_setr_epi32(1, 2, 3, 0);
|
||||
let src = _mm_setr_epi32(2, 3, 4, 1);
|
||||
_mm_i32scatter_epi32::<4>(base_addr.as_mut_ptr().cast(), vindex, src);
|
||||
_mm_i32scatter_epi32::<4>(base_addr.as_mut_ptr(), vindex, src);
|
||||
let expected = [1, 2, 3, 4];
|
||||
assert_eq!(expected, base_addr);
|
||||
}
|
||||
@ -8005,7 +8003,7 @@ mod tests {
|
||||
let mut base_addr: [i32; 4] = [0; 4];
|
||||
let vindex = _mm_setr_epi32(1, 2, 3, 0);
|
||||
let src = _mm_setr_epi32(2, 3, 4, 1);
|
||||
_mm_mask_i32scatter_epi32::<4>(base_addr.as_mut_ptr().cast(), 0b0101, vindex, src);
|
||||
_mm_mask_i32scatter_epi32::<4>(base_addr.as_mut_ptr(), 0b0101, vindex, src);
|
||||
let expected = [0, 2, 0, 4];
|
||||
assert_eq!(expected, base_addr);
|
||||
}
|
||||
@ -8015,7 +8013,7 @@ mod tests {
|
||||
let mut base_addr: [i64; 2] = [0; 2];
|
||||
let vindex = _mm_setr_epi32(1, 0, -1, -1);
|
||||
let src = _mm_setr_epi64x(2, 1);
|
||||
_mm_i32scatter_epi64::<8>(base_addr.as_mut_ptr().cast(), vindex, src);
|
||||
_mm_i32scatter_epi64::<8>(base_addr.as_mut_ptr(), vindex, src);
|
||||
let expected = [1, 2];
|
||||
assert_eq!(expected, base_addr);
|
||||
}
|
||||
@ -8025,7 +8023,7 @@ mod tests {
|
||||
let mut base_addr: [i64; 2] = [0; 2];
|
||||
let vindex = _mm_setr_epi32(1, 0, -1, -1);
|
||||
let src = _mm_setr_epi64x(2, 1);
|
||||
_mm_mask_i32scatter_epi64::<8>(base_addr.as_mut_ptr().cast(), 0b01, vindex, src);
|
||||
_mm_mask_i32scatter_epi64::<8>(base_addr.as_mut_ptr(), 0b01, vindex, src);
|
||||
let expected = [0, 2];
|
||||
assert_eq!(expected, base_addr);
|
||||
}
|
||||
@ -8035,7 +8033,7 @@ mod tests {
|
||||
let mut base_addr: [f64; 2] = [0.; 2];
|
||||
let vindex = _mm_setr_epi32(1, 0, -1, -1);
|
||||
let src = _mm_setr_pd(2., 1.);
|
||||
_mm_i32scatter_pd::<8>(base_addr.as_mut_ptr().cast(), vindex, src);
|
||||
_mm_i32scatter_pd::<8>(base_addr.as_mut_ptr(), vindex, src);
|
||||
let expected = [1., 2.];
|
||||
assert_eq!(expected, base_addr);
|
||||
}
|
||||
@ -8045,7 +8043,7 @@ mod tests {
|
||||
let mut base_addr: [f64; 2] = [0.; 2];
|
||||
let vindex = _mm_setr_epi32(1, 0, -1, -1);
|
||||
let src = _mm_setr_pd(2., 1.);
|
||||
_mm_mask_i32scatter_pd::<8>(base_addr.as_mut_ptr().cast(), 0b01, vindex, src);
|
||||
_mm_mask_i32scatter_pd::<8>(base_addr.as_mut_ptr(), 0b01, vindex, src);
|
||||
let expected = [0., 2.];
|
||||
assert_eq!(expected, base_addr);
|
||||
}
|
||||
@ -8055,7 +8053,7 @@ mod tests {
|
||||
let mut base_addr: [f32; 4] = [0.; 4];
|
||||
let vindex = _mm_setr_epi32(1, 2, 3, 0);
|
||||
let src = _mm_setr_ps(2., 3., 4., 1.);
|
||||
_mm_i32scatter_ps::<4>(base_addr.as_mut_ptr().cast(), vindex, src);
|
||||
_mm_i32scatter_ps::<4>(base_addr.as_mut_ptr(), vindex, src);
|
||||
let expected = [1., 2., 3., 4.];
|
||||
assert_eq!(expected, base_addr);
|
||||
}
|
||||
@ -8065,7 +8063,7 @@ mod tests {
|
||||
let mut base_addr: [f32; 4] = [0.; 4];
|
||||
let vindex = _mm_setr_epi32(1, 2, 3, 0);
|
||||
let src = _mm_setr_ps(2., 3., 4., 1.);
|
||||
_mm_mask_i32scatter_ps::<4>(base_addr.as_mut_ptr().cast(), 0b0101, vindex, src);
|
||||
_mm_mask_i32scatter_ps::<4>(base_addr.as_mut_ptr(), 0b0101, vindex, src);
|
||||
let expected = [0., 2., 0., 4.];
|
||||
assert_eq!(expected, base_addr);
|
||||
}
|
||||
@ -8075,7 +8073,7 @@ mod tests {
|
||||
let mut base_addr: [i32; 2] = [0; 2];
|
||||
let vindex = _mm_setr_epi64x(1, 0);
|
||||
let src = _mm_setr_epi32(2, 1, -1, -1);
|
||||
_mm_i64scatter_epi32::<4>(base_addr.as_mut_ptr().cast(), vindex, src);
|
||||
_mm_i64scatter_epi32::<4>(base_addr.as_mut_ptr(), vindex, src);
|
||||
let expected = [1, 2];
|
||||
assert_eq!(expected, base_addr);
|
||||
}
|
||||
@ -8085,7 +8083,7 @@ mod tests {
|
||||
let mut base_addr: [i32; 2] = [0; 2];
|
||||
let vindex = _mm_setr_epi64x(1, 0);
|
||||
let src = _mm_setr_epi32(2, 1, -1, -1);
|
||||
_mm_mask_i64scatter_epi32::<4>(base_addr.as_mut_ptr().cast(), 0b01, vindex, src);
|
||||
_mm_mask_i64scatter_epi32::<4>(base_addr.as_mut_ptr(), 0b01, vindex, src);
|
||||
let expected = [0, 2];
|
||||
assert_eq!(expected, base_addr);
|
||||
}
|
||||
@ -8095,7 +8093,7 @@ mod tests {
|
||||
let mut base_addr: [i64; 2] = [0; 2];
|
||||
let vindex = _mm_setr_epi64x(1, 0);
|
||||
let src = _mm_setr_epi64x(2, 1);
|
||||
_mm_i64scatter_epi64::<8>(base_addr.as_mut_ptr().cast(), vindex, src);
|
||||
_mm_i64scatter_epi64::<8>(base_addr.as_mut_ptr(), vindex, src);
|
||||
let expected = [1, 2];
|
||||
assert_eq!(expected, base_addr);
|
||||
}
|
||||
@ -8105,7 +8103,7 @@ mod tests {
|
||||
let mut base_addr: [i64; 2] = [0; 2];
|
||||
let vindex = _mm_setr_epi64x(1, 0);
|
||||
let src = _mm_setr_epi64x(2, 1);
|
||||
_mm_mask_i64scatter_epi64::<8>(base_addr.as_mut_ptr().cast(), 0b01, vindex, src);
|
||||
_mm_mask_i64scatter_epi64::<8>(base_addr.as_mut_ptr(), 0b01, vindex, src);
|
||||
let expected = [0, 2];
|
||||
assert_eq!(expected, base_addr);
|
||||
}
|
||||
@ -8115,7 +8113,7 @@ mod tests {
|
||||
let mut base_addr: [f64; 2] = [0.; 2];
|
||||
let vindex = _mm_setr_epi64x(1, 0);
|
||||
let src = _mm_setr_pd(2., 1.);
|
||||
_mm_i64scatter_pd::<8>(base_addr.as_mut_ptr().cast(), vindex, src);
|
||||
_mm_i64scatter_pd::<8>(base_addr.as_mut_ptr(), vindex, src);
|
||||
let expected = [1., 2.];
|
||||
assert_eq!(expected, base_addr);
|
||||
}
|
||||
@ -8125,7 +8123,7 @@ mod tests {
|
||||
let mut base_addr: [f64; 2] = [0.; 2];
|
||||
let vindex = _mm_setr_epi64x(1, 0);
|
||||
let src = _mm_setr_pd(2., 1.);
|
||||
_mm_mask_i64scatter_pd::<8>(base_addr.as_mut_ptr().cast(), 0b01, vindex, src);
|
||||
_mm_mask_i64scatter_pd::<8>(base_addr.as_mut_ptr(), 0b01, vindex, src);
|
||||
let expected = [0., 2.];
|
||||
assert_eq!(expected, base_addr);
|
||||
}
|
||||
@ -8135,7 +8133,7 @@ mod tests {
|
||||
let mut base_addr: [f32; 2] = [0.; 2];
|
||||
let vindex = _mm_setr_epi64x(1, 0);
|
||||
let src = _mm_setr_ps(2., 1., -1., -1.);
|
||||
_mm_i64scatter_ps::<4>(base_addr.as_mut_ptr().cast(), vindex, src);
|
||||
_mm_i64scatter_ps::<4>(base_addr.as_mut_ptr(), vindex, src);
|
||||
let expected = [1., 2.];
|
||||
assert_eq!(expected, base_addr);
|
||||
}
|
||||
@ -8145,7 +8143,7 @@ mod tests {
|
||||
let mut base_addr: [f32; 2] = [0.; 2];
|
||||
let vindex = _mm_setr_epi64x(1, 0);
|
||||
let src = _mm_setr_ps(2., 1., -1., -1.);
|
||||
_mm_mask_i64scatter_ps::<4>(base_addr.as_mut_ptr().cast(), 0b01, vindex, src);
|
||||
_mm_mask_i64scatter_ps::<4>(base_addr.as_mut_ptr(), 0b01, vindex, src);
|
||||
let expected = [0., 2.];
|
||||
assert_eq!(expected, base_addr);
|
||||
}
|
||||
@ -8155,7 +8153,7 @@ mod tests {
|
||||
let mut base_addr: [i32; 8] = [0; 8];
|
||||
let vindex = _mm256_setr_epi32(1, 2, 3, 4, 5, 6, 7, 0);
|
||||
let src = _mm256_setr_epi32(2, 3, 4, 5, 6, 7, 8, 1);
|
||||
_mm256_i32scatter_epi32::<4>(base_addr.as_mut_ptr().cast(), vindex, src);
|
||||
_mm256_i32scatter_epi32::<4>(base_addr.as_mut_ptr(), vindex, src);
|
||||
let expected = [1, 2, 3, 4, 5, 6, 7, 8];
|
||||
assert_eq!(expected, base_addr);
|
||||
}
|
||||
@ -8165,7 +8163,7 @@ mod tests {
|
||||
let mut base_addr: [i32; 8] = [0; 8];
|
||||
let vindex = _mm256_setr_epi32(1, 2, 3, 4, 5, 6, 7, 0);
|
||||
let src = _mm256_setr_epi32(2, 3, 4, 5, 6, 7, 8, 1);
|
||||
_mm256_mask_i32scatter_epi32::<4>(base_addr.as_mut_ptr().cast(), 0b01010101, vindex, src);
|
||||
_mm256_mask_i32scatter_epi32::<4>(base_addr.as_mut_ptr(), 0b01010101, vindex, src);
|
||||
let expected = [0, 2, 0, 4, 0, 6, 0, 8];
|
||||
assert_eq!(expected, base_addr);
|
||||
}
|
||||
@ -8175,7 +8173,7 @@ mod tests {
|
||||
let mut base_addr: [i64; 4] = [0; 4];
|
||||
let vindex = _mm_setr_epi32(1, 2, 3, 0);
|
||||
let src = _mm256_setr_epi64x(2, 3, 4, 1);
|
||||
_mm256_i32scatter_epi64::<8>(base_addr.as_mut_ptr().cast(), vindex, src);
|
||||
_mm256_i32scatter_epi64::<8>(base_addr.as_mut_ptr(), vindex, src);
|
||||
let expected = [1, 2, 3, 4];
|
||||
assert_eq!(expected, base_addr);
|
||||
}
|
||||
@ -8185,7 +8183,7 @@ mod tests {
|
||||
let mut base_addr: [i64; 4] = [0; 4];
|
||||
let vindex = _mm_setr_epi32(1, 2, 3, 0);
|
||||
let src = _mm256_setr_epi64x(2, 3, 4, 1);
|
||||
_mm256_mask_i32scatter_epi64::<8>(base_addr.as_mut_ptr().cast(), 0b0101, vindex, src);
|
||||
_mm256_mask_i32scatter_epi64::<8>(base_addr.as_mut_ptr(), 0b0101, vindex, src);
|
||||
let expected = [0, 2, 0, 4];
|
||||
assert_eq!(expected, base_addr);
|
||||
}
|
||||
@ -8195,7 +8193,7 @@ mod tests {
|
||||
let mut base_addr: [f64; 4] = [0.; 4];
|
||||
let vindex = _mm_setr_epi32(1, 2, 3, 0);
|
||||
let src = _mm256_setr_pd(2., 3., 4., 1.);
|
||||
_mm256_i32scatter_pd::<8>(base_addr.as_mut_ptr().cast(), vindex, src);
|
||||
_mm256_i32scatter_pd::<8>(base_addr.as_mut_ptr(), vindex, src);
|
||||
let expected = [1., 2., 3., 4.];
|
||||
assert_eq!(expected, base_addr);
|
||||
}
|
||||
@ -8205,7 +8203,7 @@ mod tests {
|
||||
let mut base_addr: [f64; 4] = [0.; 4];
|
||||
let vindex = _mm_setr_epi32(1, 2, 3, 0);
|
||||
let src = _mm256_setr_pd(2., 3., 4., 1.);
|
||||
_mm256_mask_i32scatter_pd::<8>(base_addr.as_mut_ptr().cast(), 0b0101, vindex, src);
|
||||
_mm256_mask_i32scatter_pd::<8>(base_addr.as_mut_ptr(), 0b0101, vindex, src);
|
||||
let expected = [0., 2., 0., 4.];
|
||||
assert_eq!(expected, base_addr);
|
||||
}
|
||||
@ -8215,7 +8213,7 @@ mod tests {
|
||||
let mut base_addr: [f32; 8] = [0.; 8];
|
||||
let vindex = _mm256_setr_epi32(1, 2, 3, 4, 5, 6, 7, 0);
|
||||
let src = _mm256_setr_ps(2., 3., 4., 5., 6., 7., 8., 1.);
|
||||
_mm256_i32scatter_ps::<4>(base_addr.as_mut_ptr().cast(), vindex, src);
|
||||
_mm256_i32scatter_ps::<4>(base_addr.as_mut_ptr(), vindex, src);
|
||||
let expected = [1., 2., 3., 4., 5., 6., 7., 8.];
|
||||
assert_eq!(expected, base_addr);
|
||||
}
|
||||
@ -8225,7 +8223,7 @@ mod tests {
|
||||
let mut base_addr: [f32; 8] = [0.; 8];
|
||||
let vindex = _mm256_setr_epi32(1, 2, 3, 4, 5, 6, 7, 0);
|
||||
let src = _mm256_setr_ps(2., 3., 4., 5., 6., 7., 8., 1.);
|
||||
_mm256_mask_i32scatter_ps::<4>(base_addr.as_mut_ptr().cast(), 0b01010101, vindex, src);
|
||||
_mm256_mask_i32scatter_ps::<4>(base_addr.as_mut_ptr(), 0b01010101, vindex, src);
|
||||
let expected = [0., 2., 0., 4., 0., 6., 0., 8.];
|
||||
assert_eq!(expected, base_addr);
|
||||
}
|
||||
@ -8235,7 +8233,7 @@ mod tests {
|
||||
let mut base_addr: [i32; 4] = [0; 4];
|
||||
let vindex = _mm256_setr_epi64x(1, 2, 3, 0);
|
||||
let src = _mm_setr_epi32(2, 3, 4, 1);
|
||||
_mm256_i64scatter_epi32::<4>(base_addr.as_mut_ptr().cast(), vindex, src);
|
||||
_mm256_i64scatter_epi32::<4>(base_addr.as_mut_ptr(), vindex, src);
|
||||
let expected = [1, 2, 3, 4];
|
||||
assert_eq!(expected, base_addr);
|
||||
}
|
||||
@ -8245,7 +8243,7 @@ mod tests {
|
||||
let mut base_addr: [i32; 4] = [0; 4];
|
||||
let vindex = _mm256_setr_epi64x(1, 2, 3, 0);
|
||||
let src = _mm_setr_epi32(2, 3, 4, 1);
|
||||
_mm256_mask_i64scatter_epi32::<4>(base_addr.as_mut_ptr().cast(), 0b0101, vindex, src);
|
||||
_mm256_mask_i64scatter_epi32::<4>(base_addr.as_mut_ptr(), 0b0101, vindex, src);
|
||||
let expected = [0, 2, 0, 4];
|
||||
assert_eq!(expected, base_addr);
|
||||
}
|
||||
@ -8255,7 +8253,7 @@ mod tests {
|
||||
let mut base_addr: [i64; 4] = [0; 4];
|
||||
let vindex = _mm256_setr_epi64x(1, 2, 3, 0);
|
||||
let src = _mm256_setr_epi64x(2, 3, 4, 1);
|
||||
_mm256_i64scatter_epi64::<8>(base_addr.as_mut_ptr().cast(), vindex, src);
|
||||
_mm256_i64scatter_epi64::<8>(base_addr.as_mut_ptr(), vindex, src);
|
||||
let expected = [1, 2, 3, 4];
|
||||
assert_eq!(expected, base_addr);
|
||||
}
|
||||
@ -8265,7 +8263,7 @@ mod tests {
|
||||
let mut base_addr: [i64; 4] = [0; 4];
|
||||
let vindex = _mm256_setr_epi64x(1, 2, 3, 0);
|
||||
let src = _mm256_setr_epi64x(2, 3, 4, 1);
|
||||
_mm256_mask_i64scatter_epi64::<8>(base_addr.as_mut_ptr().cast(), 0b0101, vindex, src);
|
||||
_mm256_mask_i64scatter_epi64::<8>(base_addr.as_mut_ptr(), 0b0101, vindex, src);
|
||||
let expected = [0, 2, 0, 4];
|
||||
assert_eq!(expected, base_addr);
|
||||
}
|
||||
@ -8275,7 +8273,7 @@ mod tests {
|
||||
let mut base_addr: [f64; 4] = [0.; 4];
|
||||
let vindex = _mm256_setr_epi64x(1, 2, 3, 0);
|
||||
let src = _mm256_setr_pd(2., 3., 4., 1.);
|
||||
_mm256_i64scatter_pd::<8>(base_addr.as_mut_ptr().cast(), vindex, src);
|
||||
_mm256_i64scatter_pd::<8>(base_addr.as_mut_ptr(), vindex, src);
|
||||
let expected = [1., 2., 3., 4.];
|
||||
assert_eq!(expected, base_addr);
|
||||
}
|
||||
@ -8285,7 +8283,7 @@ mod tests {
|
||||
let mut base_addr: [f64; 4] = [0.; 4];
|
||||
let vindex = _mm256_setr_epi64x(1, 2, 3, 0);
|
||||
let src = _mm256_setr_pd(2., 3., 4., 1.);
|
||||
_mm256_mask_i64scatter_pd::<8>(base_addr.as_mut_ptr().cast(), 0b0101, vindex, src);
|
||||
_mm256_mask_i64scatter_pd::<8>(base_addr.as_mut_ptr(), 0b0101, vindex, src);
|
||||
let expected = [0., 2., 0., 4.];
|
||||
assert_eq!(expected, base_addr);
|
||||
}
|
||||
@ -8295,7 +8293,7 @@ mod tests {
|
||||
let mut base_addr: [f32; 4] = [0.; 4];
|
||||
let vindex = _mm256_setr_epi64x(1, 2, 3, 0);
|
||||
let src = _mm_setr_ps(2., 3., 4., 1.);
|
||||
_mm256_i64scatter_ps::<4>(base_addr.as_mut_ptr().cast(), vindex, src);
|
||||
_mm256_i64scatter_ps::<4>(base_addr.as_mut_ptr(), vindex, src);
|
||||
let expected = [1., 2., 3., 4.];
|
||||
assert_eq!(expected, base_addr);
|
||||
}
|
||||
@ -8305,7 +8303,7 @@ mod tests {
|
||||
let mut base_addr: [f32; 4] = [0.; 4];
|
||||
let vindex = _mm256_setr_epi64x(1, 2, 3, 0);
|
||||
let src = _mm_setr_ps(2., 3., 4., 1.);
|
||||
_mm256_mask_i64scatter_ps::<4>(base_addr.as_mut_ptr().cast(), 0b0101, vindex, src);
|
||||
_mm256_mask_i64scatter_ps::<4>(base_addr.as_mut_ptr(), 0b0101, vindex, src);
|
||||
let expected = [0., 2., 0., 4.];
|
||||
assert_eq!(expected, base_addr);
|
||||
}
|
||||
|
||||
Loading…
x
Reference in New Issue
Block a user