Skip to content

S
simde
Commit 2015796f authored by Evan Nemerson's avatar Evan Nemerson
Browse files
Options

xop: add a bunch of NEON implementations 

This isn't complete, it's just to the point I got tired of working
on it.
parent 214b02a5
Show whitespace changes
Inline Side-by-side
Showing with 450 additions and 230 deletions
x86/xop.h
View file @ 2015796f
......@@ -122,7 +122,9 @@ simde_mm_comeq_epi8 (simde__m128i a, simde__m128i b) {
a_ = simde__m128i_to_private(a),
b_ = simde__m128i_to_private(b);
#if defined(SIMDE_VECTOR_SUBSCRIPT_OPS)
#if defined(SIMDE_ARM_NEON_A32V7_NATIVE)
r_.neon_u8 = vceqq_s8(a_.neon_i8, b_.neon_i8);
#elif defined(SIMDE_VECTOR_SUBSCRIPT_OPS)
r_.i8 = HEDLEY_REINTERPRET_CAST(__typeof__(r_.i8), a_.i8 == b_.i8);
#else
SIMDE_VECTORIZE
......@@ -151,7 +153,9 @@ simde_mm_comeq_epi16 (simde__m128i a, simde__m128i b) {
a_ = simde__m128i_to_private(a),
b_ = simde__m128i_to_private(b);
#if defined(SIMDE_VECTOR_SUBSCRIPT_OPS)
#if defined(SIMDE_ARM_NEON_A32V7_NATIVE)
r_.neon_u16 = vceqq_s16(a_.neon_i16, b_.neon_i16);
#elif defined(SIMDE_VECTOR_SUBSCRIPT_OPS)
r_.i16 = HEDLEY_REINTERPRET_CAST(__typeof__(r_.i16), a_.i16 == b_.i16);
#else
SIMDE_VECTORIZE
......@@ -180,7 +184,9 @@ simde_mm_comeq_epi32 (simde__m128i a, simde__m128i b) {
a_ = simde__m128i_to_private(a),
b_ = simde__m128i_to_private(b);
#if defined(SIMDE_VECTOR_SUBSCRIPT_OPS)
#if defined(SIMDE_ARM_NEON_A32V7_NATIVE)
r_.neon_u32 = vceqq_s32(a_.neon_i32, b_.neon_i32);
#elif defined(SIMDE_VECTOR_SUBSCRIPT_OPS)
r_.i32 = HEDLEY_REINTERPRET_CAST(__typeof__(r_.i32), a_.i32 == b_.i32);
#else
SIMDE_VECTORIZE
......@@ -209,7 +215,9 @@ simde_mm_comeq_epi64 (simde__m128i a, simde__m128i b) {
a_ = simde__m128i_to_private(a),
b_ = simde__m128i_to_private(b);
#if defined(SIMDE_VECTOR_SUBSCRIPT_OPS)
#if defined(SIMDE_ARM_NEON_A64V8_NATIVE)
r_.neon_u64 = vceqq_s64(a_.neon_i64, b_.neon_i64);
#elif defined(SIMDE_VECTOR_SUBSCRIPT_OPS)
r_.i64 = HEDLEY_REINTERPRET_CAST(__typeof__(r_.i64), a_.i64 == b_.i64);
#else
SIMDE_VECTORIZE
......@@ -238,7 +246,9 @@ simde_mm_comeq_epu8 (simde__m128i a, simde__m128i b) {
a_ = simde__m128i_to_private(a),
b_ = simde__m128i_to_private(b);
#if defined(SIMDE_VECTOR_SUBSCRIPT_OPS)
#if defined(SIMDE_ARM_NEON_A32V7_NATIVE)
r_.neon_u8 = vceqq_u8(a_.neon_u8, b_.neon_u8);
#elif defined(SIMDE_VECTOR_SUBSCRIPT_OPS)
r_.u8 = HEDLEY_REINTERPRET_CAST(__typeof__(r_.u8), a_.u8 == b_.u8);
#else
SIMDE_VECTORIZE
......@@ -267,7 +277,9 @@ simde_mm_comeq_epu16 (simde__m128i a, simde__m128i b) {
a_ = simde__m128i_to_private(a),
b_ = simde__m128i_to_private(b);
#if defined(SIMDE_VECTOR_SUBSCRIPT_OPS)
#if defined(SIMDE_ARM_NEON_A32V7_NATIVE)
r_.neon_u16 = vceqq_u16(a_.neon_u16, b_.neon_u16);
#elif defined(SIMDE_VECTOR_SUBSCRIPT_OPS)
r_.u16 = HEDLEY_REINTERPRET_CAST(__typeof__(r_.u16), a_.u16 == b_.u16);
#else
SIMDE_VECTORIZE
......@@ -296,7 +308,9 @@ simde_mm_comeq_epu32 (simde__m128i a, simde__m128i b) {
a_ = simde__m128i_to_private(a),
b_ = simde__m128i_to_private(b);
#if defined(SIMDE_VECTOR_SUBSCRIPT_OPS)
#if defined(SIMDE_ARM_NEON_A32V7_NATIVE)
r_.neon_u32 = vceqq_u32(a_.neon_u32, b_.neon_u32);
#elif defined(SIMDE_VECTOR_SUBSCRIPT_OPS)
r_.u32 = HEDLEY_REINTERPRET_CAST(__typeof__(r_.u32), a_.u32 == b_.u32);
#else
SIMDE_VECTORIZE
......@@ -325,7 +339,9 @@ simde_mm_comeq_epu64 (simde__m128i a, simde__m128i b) {
a_ = simde__m128i_to_private(a),
b_ = simde__m128i_to_private(b);
#if defined(SIMDE_VECTOR_SUBSCRIPT_OPS)
#if defined(SIMDE_ARM_NEON_A64V8_NATIVE)
r_.neon_u64 = vceqq_u64(a_.neon_u64, b_.neon_u64);
#elif defined(SIMDE_VECTOR_SUBSCRIPT_OPS)
r_.u64 = HEDLEY_REINTERPRET_CAST(__typeof__(r_.u64), a_.u64 == b_.u64);
#else
SIMDE_VECTORIZE
......@@ -354,7 +370,9 @@ simde_mm_comge_epi8 (simde__m128i a, simde__m128i b) {
a_ = simde__m128i_to_private(a),
b_ = simde__m128i_to_private(b);
#if defined(SIMDE_VECTOR_SUBSCRIPT_OPS)
#if defined(SIMDE_ARM_NEON_A32V7_NATIVE)
r_.neon_u8 = vcgeq_s8(a_.neon_i8, b_.neon_i8);
#elif defined(SIMDE_VECTOR_SUBSCRIPT_OPS)
r_.i8 = HEDLEY_REINTERPRET_CAST(__typeof__(r_.i8), a_.i8 >= b_.i8);
#else
SIMDE_VECTORIZE
......@@ -383,7 +401,9 @@ simde_mm_comge_epi16 (simde__m128i a, simde__m128i b) {
a_ = simde__m128i_to_private(a),
b_ = simde__m128i_to_private(b);
#if defined(SIMDE_VECTOR_SUBSCRIPT_OPS)
#if defined(SIMDE_ARM_NEON_A32V7_NATIVE)
r_.neon_u16 = vcgeq_s16(a_.neon_i16, b_.neon_i16);
#elif defined(SIMDE_VECTOR_SUBSCRIPT_OPS)
r_.i16 = HEDLEY_REINTERPRET_CAST(__typeof__(r_.i16), a_.i16 >= b_.i16);
#else
SIMDE_VECTORIZE
......@@ -412,7 +432,9 @@ simde_mm_comge_epi32 (simde__m128i a, simde__m128i b) {
a_ = simde__m128i_to_private(a),
b_ = simde__m128i_to_private(b);
#if defined(SIMDE_VECTOR_SUBSCRIPT_OPS)
#if defined(SIMDE_ARM_NEON_A32V7_NATIVE)
r_.neon_u32 = vcgeq_s32(a_.neon_i32, b_.neon_i32);
#elif defined(SIMDE_VECTOR_SUBSCRIPT_OPS)
r_.i32 = HEDLEY_REINTERPRET_CAST(__typeof__(r_.i32), a_.i32 >= b_.i32);
#else
SIMDE_VECTORIZE
......@@ -441,7 +463,9 @@ simde_mm_comge_epi64 (simde__m128i a, simde__m128i b) {
a_ = simde__m128i_to_private(a),
b_ = simde__m128i_to_private(b);
#if defined(SIMDE_VECTOR_SUBSCRIPT_OPS)
#if defined(SIMDE_ARM_NEON_A64V8_NATIVE)
r_.neon_u64 = vcgeq_s64(a_.neon_i64, b_.neon_i64);
#elif defined(SIMDE_VECTOR_SUBSCRIPT_OPS)
r_.i64 = HEDLEY_REINTERPRET_CAST(__typeof__(r_.i64), a_.i64 >= b_.i64);
#else
SIMDE_VECTORIZE
......@@ -470,7 +494,9 @@ simde_mm_comge_epu8 (simde__m128i a, simde__m128i b) {
a_ = simde__m128i_to_private(a),
b_ = simde__m128i_to_private(b);
#if defined(SIMDE_VECTOR_SUBSCRIPT_OPS)
#if defined(SIMDE_ARM_NEON_A32V7_NATIVE)
r_.neon_u8 = vcgeq_u8(a_.neon_u8, b_.neon_u8);
#elif defined(SIMDE_VECTOR_SUBSCRIPT_OPS)
r_.u8 = HEDLEY_REINTERPRET_CAST(__typeof__(r_.u8), a_.u8 >= b_.u8);
#else
SIMDE_VECTORIZE
......@@ -499,7 +525,9 @@ simde_mm_comge_epu16 (simde__m128i a, simde__m128i b) {
a_ = simde__m128i_to_private(a),
b_ = simde__m128i_to_private(b);
#if defined(SIMDE_VECTOR_SUBSCRIPT_OPS)
#if defined(SIMDE_ARM_NEON_A32V7_NATIVE)
r_.neon_u16 = vcgeq_u16(a_.neon_u16, b_.neon_u16);
#elif defined(SIMDE_VECTOR_SUBSCRIPT_OPS)
r_.u16 = HEDLEY_REINTERPRET_CAST(__typeof__(r_.u16), a_.u16 >= b_.u16);
#else
SIMDE_VECTORIZE
......@@ -528,7 +556,9 @@ simde_mm_comge_epu32 (simde__m128i a, simde__m128i b) {
a_ = simde__m128i_to_private(a),
b_ = simde__m128i_to_private(b);
#if defined(SIMDE_VECTOR_SUBSCRIPT_OPS)
#if defined(SIMDE_ARM_NEON_A32V7_NATIVE)
r_.neon_u32 = vcgeq_u32(a_.neon_u32, b_.neon_u32);
#elif defined(SIMDE_VECTOR_SUBSCRIPT_OPS)
r_.u32 = HEDLEY_REINTERPRET_CAST(__typeof__(r_.u32), a_.u32 >= b_.u32);
#else
SIMDE_VECTORIZE
......@@ -557,7 +587,9 @@ simde_mm_comge_epu64 (simde__m128i a, simde__m128i b) {
a_ = simde__m128i_to_private(a),
b_ = simde__m128i_to_private(b);
#if defined(SIMDE_VECTOR_SUBSCRIPT_OPS)
#if defined(SIMDE_ARM_NEON_A64V8_NATIVE)
r_.neon_u64 = vcgeq_u64(a_.neon_u64, b_.neon_u64);
#elif defined(SIMDE_VECTOR_SUBSCRIPT_OPS)
r_.u64 = HEDLEY_REINTERPRET_CAST(__typeof__(r_.u64), a_.u64 >= b_.u64);
#else
SIMDE_VECTORIZE
......@@ -586,7 +618,9 @@ simde_mm_comgt_epi8 (simde__m128i a, simde__m128i b) {
a_ = simde__m128i_to_private(a),
b_ = simde__m128i_to_private(b);
#if defined(SIMDE_VECTOR_SUBSCRIPT_OPS)
#if defined(SIMDE_ARM_NEON_A32V7_NATIVE)
r_.neon_u8 = vcgtq_s8(a_.neon_i8, b_.neon_i8);
#elif defined(SIMDE_VECTOR_SUBSCRIPT_OPS)
r_.i8 = HEDLEY_REINTERPRET_CAST(__typeof__(r_.i8), a_.i8 > b_.i8);
#else
SIMDE_VECTORIZE
......@@ -615,7 +649,9 @@ simde_mm_comgt_epi16 (simde__m128i a, simde__m128i b) {
a_ = simde__m128i_to_private(a),
b_ = simde__m128i_to_private(b);
#if defined(SIMDE_VECTOR_SUBSCRIPT_OPS)
#if defined(SIMDE_ARM_NEON_A32V7_NATIVE)
r_.neon_u16 = vcgtq_s16(a_.neon_i16, b_.neon_i16);
#elif defined(SIMDE_VECTOR_SUBSCRIPT_OPS)
r_.i16 = HEDLEY_REINTERPRET_CAST(__typeof__(r_.i16), a_.i16 > b_.i16);
#else
SIMDE_VECTORIZE
......@@ -644,7 +680,9 @@ simde_mm_comgt_epi32 (simde__m128i a, simde__m128i b) {
a_ = simde__m128i_to_private(a),
b_ = simde__m128i_to_private(b);
#if defined(SIMDE_VECTOR_SUBSCRIPT_OPS)
#if defined(SIMDE_ARM_NEON_A32V7_NATIVE)
r_.neon_u32 = vcgtq_s32(a_.neon_i32, b_.neon_i32);
#elif defined(SIMDE_VECTOR_SUBSCRIPT_OPS)
r_.i32 = HEDLEY_REINTERPRET_CAST(__typeof__(r_.i32), a_.i32 > b_.i32);
#else
SIMDE_VECTORIZE
......@@ -673,7 +711,9 @@ simde_mm_comgt_epi64 (simde__m128i a, simde__m128i b) {
a_ = simde__m128i_to_private(a),
b_ = simde__m128i_to_private(b);
#if defined(SIMDE_VECTOR_SUBSCRIPT_OPS)
#if defined(SIMDE_ARM_NEON_A64V8_NATIVE)
r_.neon_u64 = vcgtq_s64(a_.neon_i64, b_.neon_i64);
#elif defined(SIMDE_VECTOR_SUBSCRIPT_OPS)
r_.i64 = HEDLEY_REINTERPRET_CAST(__typeof__(r_.i64), a_.i64 > b_.i64);
#else
SIMDE_VECTORIZE
......@@ -702,7 +742,9 @@ simde_mm_comgt_epu8 (simde__m128i a, simde__m128i b) {
a_ = simde__m128i_to_private(a),
b_ = simde__m128i_to_private(b);
#if defined(SIMDE_VECTOR_SUBSCRIPT_OPS)
#if defined(SIMDE_ARM_NEON_A32V7_NATIVE)
r_.neon_u8 = vcgtq_u8(a_.neon_u8, b_.neon_u8);
#elif defined(SIMDE_VECTOR_SUBSCRIPT_OPS)
r_.u8 = HEDLEY_REINTERPRET_CAST(__typeof__(r_.u8), a_.u8 > b_.u8);
#else
SIMDE_VECTORIZE
......@@ -731,7 +773,9 @@ simde_mm_comgt_epu16 (simde__m128i a, simde__m128i b) {
a_ = simde__m128i_to_private(a),
b_ = simde__m128i_to_private(b);
#if defined(SIMDE_VECTOR_SUBSCRIPT_OPS)
#if defined(SIMDE_ARM_NEON_A32V7_NATIVE)
r_.neon_u16 = vcgtq_u16(a_.neon_u16, b_.neon_u16);
#elif defined(SIMDE_VECTOR_SUBSCRIPT_OPS)
r_.u16 = HEDLEY_REINTERPRET_CAST(__typeof__(r_.u16), a_.u16 > b_.u16);
#else
SIMDE_VECTORIZE
......@@ -760,7 +804,9 @@ simde_mm_comgt_epu32 (simde__m128i a, simde__m128i b) {
a_ = simde__m128i_to_private(a),
b_ = simde__m128i_to_private(b);
#if defined(SIMDE_VECTOR_SUBSCRIPT_OPS)
#if defined(SIMDE_ARM_NEON_A32V7_NATIVE)
r_.neon_u32 = vcgtq_u32(a_.neon_u32, b_.neon_u32);
#elif defined(SIMDE_VECTOR_SUBSCRIPT_OPS)
r_.u32 = HEDLEY_REINTERPRET_CAST(__typeof__(r_.u32), a_.u32 > b_.u32);
#else
SIMDE_VECTORIZE
......@@ -789,7 +835,9 @@ simde_mm_comgt_epu64 (simde__m128i a, simde__m128i b) {
a_ = simde__m128i_to_private(a),
b_ = simde__m128i_to_private(b);
#if defined(SIMDE_VECTOR_SUBSCRIPT_OPS)
#if defined(SIMDE_ARM_NEON_A64V8_NATIVE)
r_.neon_u64 = vcgtq_u64(a_.neon_u64, b_.neon_u64);
#elif defined(SIMDE_VECTOR_SUBSCRIPT_OPS)
r_.u64 = HEDLEY_REINTERPRET_CAST(__typeof__(r_.u64), a_.u64 > b_.u64);
#else
SIMDE_VECTORIZE
......@@ -818,7 +866,9 @@ simde_mm_comle_epi8 (simde__m128i a, simde__m128i b) {
a_ = simde__m128i_to_private(a),
b_ = simde__m128i_to_private(b);
#if defined(SIMDE_VECTOR_SUBSCRIPT_OPS)
#if defined(SIMDE_ARM_NEON_A32V7_NATIVE)
r_.neon_u8 = vcleq_s8(a_.neon_i8, b_.neon_i8);
#elif defined(SIMDE_VECTOR_SUBSCRIPT_OPS)
r_.i8 = HEDLEY_REINTERPRET_CAST(__typeof__(r_.i8), a_.i8 <= b_.i8);
#else
SIMDE_VECTORIZE
......@@ -876,7 +926,9 @@ simde_mm_comle_epi32 (simde__m128i a, simde__m128i b) {
a_ = simde__m128i_to_private(a),
b_ = simde__m128i_to_private(b);
#if defined(SIMDE_VECTOR_SUBSCRIPT_OPS)
#if defined(SIMDE_ARM_NEON_A32V7_NATIVE)
r_.neon_u32 = vcleq_s32(a_.neon_i32, b_.neon_i32);
#elif defined(SIMDE_VECTOR_SUBSCRIPT_OPS)
r_.i32 = HEDLEY_REINTERPRET_CAST(__typeof__(r_.i32), a_.i32 <= b_.i32);
#else
SIMDE_VECTORIZE
......@@ -905,7 +957,9 @@ simde_mm_comle_epi64 (simde__m128i a, simde__m128i b) {
a_ = simde__m128i_to_private(a),
b_ = simde__m128i_to_private(b);
#if defined(SIMDE_VECTOR_SUBSCRIPT_OPS)
#if defined(SIMDE_ARM_NEON_A64V8_NATIVE)
r_.neon_u64 = vcleq_s64(a_.neon_i64, b_.neon_i64);
#elif defined(SIMDE_VECTOR_SUBSCRIPT_OPS)
r_.i64 = HEDLEY_REINTERPRET_CAST(__typeof__(r_.i64), a_.i64 <= b_.i64);
#else
SIMDE_VECTORIZE
......@@ -934,7 +988,9 @@ simde_mm_comle_epu8 (simde__m128i a, simde__m128i b) {
a_ = simde__m128i_to_private(a),
b_ = simde__m128i_to_private(b);
#if defined(SIMDE_VECTOR_SUBSCRIPT_OPS)
#if defined(SIMDE_ARM_NEON_A32V7_NATIVE)
r_.neon_u8 = vcleq_u8(a_.neon_u8, b_.neon_u8);
#elif defined(SIMDE_VECTOR_SUBSCRIPT_OPS)
r_.u8 = HEDLEY_REINTERPRET_CAST(__typeof__(r_.u8), a_.u8 <= b_.u8);
#else
SIMDE_VECTORIZE
......@@ -963,7 +1019,9 @@ simde_mm_comle_epu16 (simde__m128i a, simde__m128i b) {
a_ = simde__m128i_to_private(a),
b_ = simde__m128i_to_private(b);
#if defined(SIMDE_VECTOR_SUBSCRIPT_OPS)
#if defined(SIMDE_ARM_NEON_A32V7_NATIVE)
r_.neon_u16 = vcleq_u16(a_.neon_u16, b_.neon_u16);
#elif defined(SIMDE_VECTOR_SUBSCRIPT_OPS)
r_.u16 = HEDLEY_REINTERPRET_CAST(__typeof__(r_.u16), a_.u16 <= b_.u16);
#else
SIMDE_VECTORIZE
......@@ -992,7 +1050,9 @@ simde_mm_comle_epu32 (simde__m128i a, simde__m128i b) {
a_ = simde__m128i_to_private(a),
b_ = simde__m128i_to_private(b);
#if defined(SIMDE_VECTOR_SUBSCRIPT_OPS)
#if defined(SIMDE_ARM_NEON_A32V7_NATIVE)
r_.neon_u32 = vcleq_u32(a_.neon_u32, b_.neon_u32);
#elif defined(SIMDE_VECTOR_SUBSCRIPT_OPS)
r_.u32 = HEDLEY_REINTERPRET_CAST(__typeof__(r_.u32), a_.u32 <= b_.u32);
#else
SIMDE_VECTORIZE
......@@ -1021,7 +1081,9 @@ simde_mm_comle_epu64 (simde__m128i a, simde__m128i b) {
a_ = simde__m128i_to_private(a),
b_ = simde__m128i_to_private(b);
#if defined(SIMDE_VECTOR_SUBSCRIPT_OPS)
#if defined(SIMDE_ARM_NEON_A64V8_NATIVE)
r_.neon_u64 = vcleq_u64(a_.neon_u64, b_.neon_u64);
#elif defined(SIMDE_VECTOR_SUBSCRIPT_OPS)
r_.u64 = HEDLEY_REINTERPRET_CAST(__typeof__(r_.u64), a_.u64 <= b_.u64);
#else
SIMDE_VECTORIZE
......@@ -1050,7 +1112,9 @@ simde_mm_comlt_epi8 (simde__m128i a, simde__m128i b) {
a_ = simde__m128i_to_private(a),
b_ = simde__m128i_to_private(b);
#if defined(SIMDE_VECTOR_SUBSCRIPT_OPS)
#if defined(SIMDE_ARM_NEON_A32V7_NATIVE)
r_.neon_u8 = vcltq_s8(a_.neon_i8, b_.neon_i8);
#elif defined(SIMDE_VECTOR_SUBSCRIPT_OPS)
r_.i8 = HEDLEY_REINTERPRET_CAST(__typeof__(r_.i8), a_.i8 < b_.i8);
#else
SIMDE_VECTORIZE
......@@ -1079,7 +1143,9 @@ simde_mm_comlt_epi16 (simde__m128i a, simde__m128i b) {
a_ = simde__m128i_to_private(a),
b_ = simde__m128i_to_private(b);
#if defined(SIMDE_VECTOR_SUBSCRIPT_OPS)
#if defined(SIMDE_ARM_NEON_A32V7_NATIVE)
r_.neon_u16 = vcltq_s16(a_.neon_i16, b_.neon_i16);
#elif defined(SIMDE_VECTOR_SUBSCRIPT_OPS)
r_.i16 = HEDLEY_REINTERPRET_CAST(__typeof__(r_.i16), a_.i16 < b_.i16);
#else
SIMDE_VECTORIZE
......@@ -1108,7 +1174,9 @@ simde_mm_comlt_epi32 (simde__m128i a, simde__m128i b) {
a_ = simde__m128i_to_private(a),
b_ = simde__m128i_to_private(b);
#if defined(SIMDE_VECTOR_SUBSCRIPT_OPS)
#if defined(SIMDE_ARM_NEON_A32V7_NATIVE)
r_.neon_u32 = vcltq_s32(a_.neon_i32, b_.neon_i32);
#elif defined(SIMDE_VECTOR_SUBSCRIPT_OPS)
r_.i32 = HEDLEY_REINTERPRET_CAST(__typeof__(r_.i32), a_.i32 < b_.i32);
#else
SIMDE_VECTORIZE
......@@ -1137,7 +1205,9 @@ simde_mm_comlt_epi64 (simde__m128i a, simde__m128i b) {
a_ = simde__m128i_to_private(a),
b_ = simde__m128i_to_private(b);
#if defined(SIMDE_VECTOR_SUBSCRIPT_OPS)
#if defined(SIMDE_ARM_NEON_A64V8_NATIVE)
r_.neon_u64 = vcltq_s64(a_.neon_i64, b_.neon_i64);
#elif defined(SIMDE_VECTOR_SUBSCRIPT_OPS)
r_.i64 = HEDLEY_REINTERPRET_CAST(__typeof__(r_.i64), a_.i64 < b_.i64);
#else
SIMDE_VECTORIZE
......@@ -1166,7 +1236,9 @@ simde_mm_comlt_epu8 (simde__m128i a, simde__m128i b) {
a_ = simde__m128i_to_private(a),
b_ = simde__m128i_to_private(b);
#if defined(SIMDE_VECTOR_SUBSCRIPT_OPS)
#if defined(SIMDE_ARM_NEON_A32V7_NATIVE)
r_.neon_u8 = vcltq_u8(a_.neon_u8, b_.neon_u8);
#elif defined(SIMDE_VECTOR_SUBSCRIPT_OPS)
r_.u8 = HEDLEY_REINTERPRET_CAST(__typeof__(r_.u8), a_.u8 < b_.u8);
#else
SIMDE_VECTORIZE
......@@ -1195,7 +1267,9 @@ simde_mm_comlt_epu16 (simde__m128i a, simde__m128i b) {
a_ = simde__m128i_to_private(a),
b_ = simde__m128i_to_private(b);
#if defined(SIMDE_VECTOR_SUBSCRIPT_OPS)
#if defined(SIMDE_ARM_NEON_A32V7_NATIVE)
r_.neon_u16 = vcltq_u16(a_.neon_u16, b_.neon_u16);
#elif defined(SIMDE_VECTOR_SUBSCRIPT_OPS)
r_.u16 = HEDLEY_REINTERPRET_CAST(__typeof__(r_.u16), a_.u16 < b_.u16);
#else
SIMDE_VECTORIZE
......@@ -1224,7 +1298,9 @@ simde_mm_comlt_epu32 (simde__m128i a, simde__m128i b) {
a_ = simde__m128i_to_private(a),
b_ = simde__m128i_to_private(b);
#if defined(SIMDE_VECTOR_SUBSCRIPT_OPS)
#if defined(SIMDE_ARM_NEON_A32V7_NATIVE)
r_.neon_u32 = vcltq_u32(a_.neon_u32, b_.neon_u32);
#elif defined(SIMDE_VECTOR_SUBSCRIPT_OPS)
r_.u32 = HEDLEY_REINTERPRET_CAST(__typeof__(r_.u32), a_.u32 < b_.u32);
#else
SIMDE_VECTORIZE
......@@ -1253,7 +1329,9 @@ simde_mm_comlt_epu64 (simde__m128i a, simde__m128i b) {
a_ = simde__m128i_to_private(a),
b_ = simde__m128i_to_private(b);
#if defined(SIMDE_VECTOR_SUBSCRIPT_OPS)
#if defined(SIMDE_ARM_NEON_A64V8_NATIVE)
r_.neon_u64 = vcltq_u64(a_.neon_u64, b_.neon_u64);
#elif defined(SIMDE_VECTOR_SUBSCRIPT_OPS)
r_.u64 = HEDLEY_REINTERPRET_CAST(__typeof__(r_.u64), a_.u64 < b_.u64);
#else
SIMDE_VECTORIZE
......@@ -1282,7 +1360,9 @@ simde_mm_comneq_epi8 (simde__m128i a, simde__m128i b) {
a_ = simde__m128i_to_private(a),
b_ = simde__m128i_to_private(b);
#if defined(SIMDE_VECTOR_SUBSCRIPT_OPS)
#if defined(SIMDE_ARM_NEON_A32V7_NATIVE)
r_.neon_u8 = vmvnq_u8(vceqq_s8(a_.neon_i8, b_.neon_i8));
#elif defined(SIMDE_VECTOR_SUBSCRIPT_OPS)
r_.i8 = HEDLEY_REINTERPRET_CAST(__typeof__(r_.i8), a_.i8 != b_.i8);
#else
SIMDE_VECTORIZE
......@@ -1311,7 +1391,9 @@ simde_mm_comneq_epi16 (simde__m128i a, simde__m128i b) {
a_ = simde__m128i_to_private(a),
b_ = simde__m128i_to_private(b);
#if defined(SIMDE_VECTOR_SUBSCRIPT_OPS)
#if defined(SIMDE_ARM_NEON_A32V7_NATIVE)
r_.neon_u16 = vmvnq_u16(vceqq_s16(a_.neon_i16, b_.neon_i16));
#elif defined(SIMDE_VECTOR_SUBSCRIPT_OPS)
r_.i16 = HEDLEY_REINTERPRET_CAST(__typeof__(r_.i16), a_.i16 != b_.i16);
#else
SIMDE_VECTORIZE
......@@ -1340,7 +1422,9 @@ simde_mm_comneq_epi32 (simde__m128i a, simde__m128i b) {
a_ = simde__m128i_to_private(a),
b_ = simde__m128i_to_private(b);
#if defined(SIMDE_VECTOR_SUBSCRIPT_OPS)
#if defined(SIMDE_ARM_NEON_A32V7_NATIVE)
r_.neon_u32 = vmvnq_u32(vceqq_s32(a_.neon_i32, b_.neon_i32));
#elif defined(SIMDE_VECTOR_SUBSCRIPT_OPS)
r_.i32 = HEDLEY_REINTERPRET_CAST(__typeof__(r_.i32), a_.i32 != b_.i32);
#else
SIMDE_VECTORIZE
......@@ -1369,7 +1453,9 @@ simde_mm_comneq_epi64 (simde__m128i a, simde__m128i b) {
a_ = simde__m128i_to_private(a),
b_ = simde__m128i_to_private(b);
#if defined(SIMDE_VECTOR_SUBSCRIPT_OPS)
#if defined(SIMDE_ARM_NEON_A64V8_NATIVE)
r_.neon_u32 = vmvnq_u32(vreinterpretq_u32_u64(vceqq_s64(a_.neon_i64, b_.neon_i64)));
#elif defined(SIMDE_VECTOR_SUBSCRIPT_OPS)
r_.i64 = HEDLEY_REINTERPRET_CAST(__typeof__(r_.i64), a_.i64 != b_.i64);
#else
SIMDE_VECTORIZE
......@@ -1398,7 +1484,9 @@ simde_mm_comneq_epu8 (simde__m128i a, simde__m128i b) {
a_ = simde__m128i_to_private(a),
b_ = simde__m128i_to_private(b);
#if defined(SIMDE_VECTOR_SUBSCRIPT_OPS)
#if defined(SIMDE_ARM_NEON_A32V7_NATIVE)
r_.neon_u8 = vmvnq_u8(vceqq_u8(a_.neon_u8, b_.neon_u8));
#elif defined(SIMDE_VECTOR_SUBSCRIPT_OPS)
r_.u8 = HEDLEY_REINTERPRET_CAST(__typeof__(r_.u8), a_.u8 != b_.u8);
#else
SIMDE_VECTORIZE
......@@ -1427,7 +1515,9 @@ simde_mm_comneq_epu16 (simde__m128i a, simde__m128i b) {
a_ = simde__m128i_to_private(a),
b_ = simde__m128i_to_private(b);
#if defined(SIMDE_VECTOR_SUBSCRIPT_OPS)
#if defined(SIMDE_ARM_NEON_A32V7_NATIVE)
r_.neon_u16 = vmvnq_u16(vceqq_u16(a_.neon_u16, b_.neon_u16));
#elif defined(SIMDE_VECTOR_SUBSCRIPT_OPS)
r_.u16 = HEDLEY_REINTERPRET_CAST(__typeof__(r_.u16), a_.u16 != b_.u16);
#else
SIMDE_VECTORIZE
......@@ -1456,7 +1546,9 @@ simde_mm_comneq_epu32 (simde__m128i a, simde__m128i b) {
a_ = simde__m128i_to_private(a),
b_ = simde__m128i_to_private(b);
#if defined(SIMDE_VECTOR_SUBSCRIPT_OPS)
#if defined(SIMDE_ARM_NEON_A32V7_NATIVE)
r_.neon_u32 = vmvnq_u32(vceqq_u32(a_.neon_u32, b_.neon_u32));
#elif defined(SIMDE_VECTOR_SUBSCRIPT_OPS)
r_.u32 = HEDLEY_REINTERPRET_CAST(__typeof__(r_.u32), a_.u32 != b_.u32);
#else
SIMDE_VECTORIZE
......@@ -1485,7 +1577,9 @@ simde_mm_comneq_epu64 (simde__m128i a, simde__m128i b) {
a_ = simde__m128i_to_private(a),
b_ = simde__m128i_to_private(b);
#if defined(SIMDE_VECTOR_SUBSCRIPT_OPS)
#if defined(SIMDE_ARM_NEON_A64V8_NATIVE)
r_.neon_u32 = vmvnq_u32(vreinterpretq_u32_u64(vceqq_u64(a_.neon_u64, b_.neon_u64)));
#elif defined(SIMDE_VECTOR_SUBSCRIPT_OPS)
r_.u64 = HEDLEY_REINTERPRET_CAST(__typeof__(r_.u64), a_.u64 != b_.u64);
#else
SIMDE_VECTORIZE
......@@ -2143,10 +2237,14 @@ simde_mm_haddw_epi8 (simde__m128i a) {
r_,
a_ = simde__m128i_to_private(a);
#if defined(SIMDE_ARM_NEON_A32V7_NATIVE)
r_.neon_i16 = vpaddlq_s8(a_.neon_i8);
#else
SIMDE_VECTORIZE
for (size_t i = 0 ; i < (sizeof(r_.i16) / sizeof(r_.i16[0])) ; i++) {
r_.i16[i] = HEDLEY_STATIC_CAST(int16_t, a_.i8[i * 2]) + HEDLEY_STATIC_CAST(int16_t, a_.i8[(i * 2) + 1]);
}
#endif
return simde__m128i_from_private(r_);
#endif
......@@ -2165,10 +2263,14 @@ simde_mm_haddw_epu8 (simde__m128i a) {
r_,
a_ = simde__m128i_to_private(a);
#if defined(SIMDE_ARM_NEON_A32V7_NATIVE)
r_.neon_u16 = vpaddlq_u8(a_.neon_u8);
#else
SIMDE_VECTORIZE
for (size_t i = 0 ; i < (sizeof(r_.u16) / sizeof(r_.u16[0])) ; i++) {
r_.u16[i] = HEDLEY_STATIC_CAST(uint16_t, a_.u8[i * 2]) + HEDLEY_STATIC_CAST(uint16_t, a_.u8[(i * 2) + 1]);
}
#endif
return simde__m128i_from_private(r_);
#endif
......@@ -2187,12 +2289,16 @@ simde_mm_haddd_epi8 (simde__m128i a) {
r_,
a_ = simde__m128i_to_private(a);
#if defined(SIMDE_ARM_NEON_A32V7_NATIVE)
r_.neon_i32 = vpaddlq_s16(vpaddlq_s8(a_.neon_i8));
#else
SIMDE_VECTORIZE
for (size_t i = 0 ; i < (sizeof(r_.i32) / sizeof(r_.i32[0])) ; i++) {
r_.i32[i] =
HEDLEY_STATIC_CAST(int32_t, a_.i8[(i * 4) ]) + HEDLEY_STATIC_CAST(int32_t, a_.i8[(i * 4) + 1]) +
HEDLEY_STATIC_CAST(int32_t, a_.i8[(i * 4) + 2]) + HEDLEY_STATIC_CAST(int32_t, a_.i8[(i * 4) + 3]);
}
#endif
return simde__m128i_from_private(r_);
#endif
......@@ -2211,11 +2317,15 @@ simde_mm_haddd_epi16 (simde__m128i a) {
r_,
a_ = simde__m128i_to_private(a);
#if defined(SIMDE_ARM_NEON_A32V7_NATIVE)
r_.neon_i32 = vpaddlq_s16(a_.neon_i16);
#else
SIMDE_VECTORIZE
for (size_t i = 0 ; i < (sizeof(r_.i32) / sizeof(r_.i32[0])) ; i++) {
r_.i32[i] =
HEDLEY_STATIC_CAST(int32_t, a_.i16[(i * 2) ]) + HEDLEY_STATIC_CAST(int32_t, a_.i16[(i * 2) + 1]);
}
#endif
return simde__m128i_from_private(r_);
#endif
......@@ -2234,12 +2344,16 @@ simde_mm_haddd_epu8 (simde__m128i a) {
r_,
a_ = simde__m128i_to_private(a);
#if defined(SIMDE_ARM_NEON_A32V7_NATIVE)
r_.neon_u32 = vpaddlq_u16(vpaddlq_u8(a_.neon_u8));
#else
SIMDE_VECTORIZE
for (size_t i = 0 ; i < (sizeof(r_.u32) / sizeof(r_.u32[0])) ; i++) {
r_.u32[i] =
HEDLEY_STATIC_CAST(uint32_t, a_.u8[(i * 4) ]) + HEDLEY_STATIC_CAST(uint32_t, a_.u8[(i * 4) + 1]) +
HEDLEY_STATIC_CAST(uint32_t, a_.u8[(i * 4) + 2]) + HEDLEY_STATIC_CAST(uint32_t, a_.u8[(i * 4) + 3]);
}
#endif
return simde__m128i_from_private(r_);
#endif
......@@ -2258,11 +2372,15 @@ simde_mm_haddd_epu16 (simde__m128i a) {
r_,
a_ = simde__m128i_to_private(a);
#if defined(SIMDE_ARM_NEON_A32V7_NATIVE)
r_.neon_u32 = vpaddlq_u16(a_.neon_u16);
#else
SIMDE_VECTORIZE
for (size_t i = 0 ; i < (sizeof(r_.u32) / sizeof(r_.u32[0])) ; i++) {
r_.u32[i] =
HEDLEY_STATIC_CAST(uint32_t, a_.u16[(i * 2) ]) + HEDLEY_STATIC_CAST(uint32_t, a_.u16[(i * 2) + 1]);
}
#endif
return simde__m128i_from_private(r_);
#endif
......@@ -2281,6 +2399,9 @@ simde_mm_haddq_epi8 (simde__m128i a) {
r_,
a_ = simde__m128i_to_private(a);
#if defined(SIMDE_ARM_NEON_A32V7_NATIVE)
r_.neon_i64 = vpaddlq_s32(vpaddlq_s16(vpaddlq_s8(a_.neon_i8)));
#else
SIMDE_VECTORIZE
for (size_t i = 0 ; i < (sizeof(r_.i64) / sizeof(r_.i64[0])) ; i++) {
r_.i64[i] =
......@@ -2289,6 +2410,7 @@ simde_mm_haddq_epi8 (simde__m128i a) {
HEDLEY_STATIC_CAST(int64_t, a_.i8[(i * 8) + 4]) + HEDLEY_STATIC_CAST(int64_t, a_.i8[(i * 8) + 5]) +
HEDLEY_STATIC_CAST(int64_t, a_.i8[(i * 8) + 6]) + HEDLEY_STATIC_CAST(int64_t, a_.i8[(i * 8) + 7]);
}
#endif
return simde__m128i_from_private(r_);
#endif
......@@ -2307,12 +2429,16 @@ simde_mm_haddq_epi16 (simde__m128i a) {
r_,
a_ = simde__m128i_to_private(a);
#if defined(SIMDE_ARM_NEON_A32V7_NATIVE)
r_.neon_i64 = vpaddlq_s32(vpaddlq_s16(a_.neon_i16));
#else
SIMDE_VECTORIZE
for (size_t i = 0 ; i < (sizeof(r_.i64) / sizeof(r_.i64[0])) ; i++) {
r_.i64[i] =
HEDLEY_STATIC_CAST(int64_t, a_.i16[(i * 4) ]) + HEDLEY_STATIC_CAST(int64_t, a_.i16[(i * 4) + 1]) +
HEDLEY_STATIC_CAST(int64_t, a_.i16[(i * 4) + 2]) + HEDLEY_STATIC_CAST(int64_t, a_.i16[(i * 4) + 3]);
}
#endif
return simde__m128i_from_private(r_);
#endif
......@@ -2331,10 +2457,14 @@ simde_mm_haddq_epi32 (simde__m128i a) {
r_,
a_ = simde__m128i_to_private(a);
#if defined(SIMDE_ARM_NEON_A32V7_NATIVE)
r_.neon_i64 = vpaddlq_s32(a_.neon_i32);
#else
SIMDE_VECTORIZE
for (size_t i = 0 ; i < (sizeof(r_.i64) / sizeof(r_.i64[0])) ; i++) {
r_.i64[i] = HEDLEY_STATIC_CAST(int64_t, a_.i32[(i * 2) ]) + HEDLEY_STATIC_CAST(int64_t, a_.i32[(i * 2) + 1]);
}
#endif
return simde__m128i_from_private(r_);
#endif
......@@ -2353,6 +2483,9 @@ simde_mm_haddq_epu8 (simde__m128i a) {
r_,
a_ = simde__m128i_to_private(a);
#if defined(SIMDE_ARM_NEON_A32V7_NATIVE)
r_.neon_u64 = vpaddlq_u32(vpaddlq_u16(vpaddlq_u8(a_.neon_u8)));
#else
SIMDE_VECTORIZE
for (size_t i = 0 ; i < (sizeof(r_.i64) / sizeof(r_.i64[0])) ; i++) {
r_.u64[i] =
......@@ -2361,6 +2494,7 @@ simde_mm_haddq_epu8 (simde__m128i a) {
HEDLEY_STATIC_CAST(uint64_t, a_.u8[(i * 8) + 4]) + HEDLEY_STATIC_CAST(uint64_t, a_.u8[(i * 8) + 5]) +
HEDLEY_STATIC_CAST(uint64_t, a_.u8[(i * 8) + 6]) + HEDLEY_STATIC_CAST(uint64_t, a_.u8[(i * 8) + 7]);
}
#endif
return simde__m128i_from_private(r_);
#endif
......@@ -2379,12 +2513,16 @@ simde_mm_haddq_epu16 (simde__m128i a) {
r_,
a_ = simde__m128i_to_private(a);
#if defined(SIMDE_ARM_NEON_A32V7_NATIVE)
r_.neon_u64 = vpaddlq_u32(vpaddlq_u16(a_.neon_u16));
#else
SIMDE_VECTORIZE
for (size_t i = 0 ; i < (sizeof(r_.i64) / sizeof(r_.i64[0])) ; i++) {
r_.u64[i] =
HEDLEY_STATIC_CAST(uint64_t, a_.u16[(i * 4) ]) + HEDLEY_STATIC_CAST(uint64_t, a_.u16[(i * 4) + 1]) +
HEDLEY_STATIC_CAST(uint64_t, a_.u16[(i * 4) + 2]) + HEDLEY_STATIC_CAST(uint64_t, a_.u16[(i * 4) + 3]);
}
#endif
return simde__m128i_from_private(r_);
#endif
......@@ -2403,10 +2541,14 @@ simde_mm_haddq_epu32 (simde__m128i a) {
r_,
a_ = simde__m128i_to_private(a);
#if defined(SIMDE_ARM_NEON_A32V7_NATIVE)
r_.neon_u64 = vpaddlq_u32(a_.neon_u32);
#else
SIMDE_VECTORIZE
for (size_t i = 0 ; i < (sizeof(r_.i64) / sizeof(r_.i64[0])) ; i++) {
r_.u64[i] = HEDLEY_STATIC_CAST(uint64_t, a_.u32[(i * 2) ]) + HEDLEY_STATIC_CAST(uint64_t, a_.u32[(i * 2) + 1]);
}
#endif
return simde__m128i_from_private(r_);
#endif
......@@ -2494,10 +2636,14 @@ simde_mm_macc_epi16 (simde__m128i a, simde__m128i b, simde__m128i c) {
b_ = simde__m128i_to_private(b),
c_ = simde__m128i_to_private(c);
#if defined(SIMDE_ARM_NEON_A32V7_NATIVE)
r_.neon_i16 = vmlaq_s16(c_.neon_i16, a_.neon_i16, b_.neon_i16);
#else
SIMDE_VECTORIZE
for (size_t i = 0 ; i < (sizeof(r_.i16) / sizeof(r_.i16[0])) ; i++) {
r_.i16[i] = (a_.i16[i] * b_.i16[i]) + c_.i16[i];
}
#endif
return simde__m128i_from_private(r_);
#endif
......@@ -2518,10 +2664,14 @@ simde_mm_macc_epi32 (simde__m128i a, simde__m128i b, simde__m128i c) {
b_ = simde__m128i_to_private(b),
c_ = simde__m128i_to_private(c);
#if defined(SIMDE_ARM_NEON_A32V7_NATIVE)
r_.neon_i32 = vmlaq_s32(c_.neon_i32, a_.neon_i32, b_.neon_i32);
#else
SIMDE_VECTORIZE
for (size_t i = 0 ; i < (sizeof(r_.i32) / sizeof(r_.i32[0])) ; i++) {
r_.i32[i] = (a_.i32[i] * b_.i32[i]) + c_.i32[i];
}
#endif
return simde__m128i_from_private(r_);
#endif
......@@ -2542,10 +2692,17 @@ simde_mm_maccd_epi16 (simde__m128i a, simde__m128i b, simde__m128i c) {
b_ = simde__m128i_to_private(b),
c_ = simde__m128i_to_private(c);
#if defined(SIMDE_ARM_NEON_A64V8_NATIVE)
int16x8_t even = vuzp1q_s16(a_.neon_i16, b_.neon_i16);
int32x4_t a_even = vmovl_s16(vget_low_s16(even));
int32x4_t b_even = vmovl_high_s16(even);
r_.neon_i32 = vmlaq_s32(c_.neon_i32, a_even, b_even);
#else
SIMDE_VECTORIZE
for (size_t i = 0 ; i < (sizeof(r_.i32) / sizeof(r_.i32[0])) ; i++) {
r_.i32[i] = (HEDLEY_STATIC_CAST(int32_t, a_.i16[i * 2]) * HEDLEY_STATIC_CAST(int32_t, b_.i16[i * 2])) + c_.i32[i];
}
#endif
return simde__m128i_from_private(r_);
#endif
......@@ -2566,10 +2723,15 @@ simde_mm_macclo_epi32 (simde__m128i a, simde__m128i b, simde__m128i c) {
b_ = simde__m128i_to_private(b),
c_ = simde__m128i_to_private(c);
#if defined(SIMDE_ARM_NEON_A64V8_NATIVE)
int32x4_t even = vuzp1q_s32(a_.neon_i32, b_.neon_i32);
r_.neon_i64 = vaddq_s64(vmull_s32(vget_low_s32(even), vget_high_s32(even)), c_.neon_i64);
#else
SIMDE_VECTORIZE
for (size_t i = 0 ; i < (sizeof(r_.i64) / sizeof(r_.i64[0])) ; i++) {
r_.i64[i] = (HEDLEY_STATIC_CAST(int64_t, a_.i32[(i * 2) + 0]) * HEDLEY_STATIC_CAST(int64_t, b_.i32[(i * 2) + 0])) + c_.i64[i];
}
#endif
return simde__m128i_from_private(r_);
#endif
......@@ -2590,10 +2752,15 @@ simde_mm_macchi_epi32 (simde__m128i a, simde__m128i b, simde__m128i c) {
b_ = simde__m128i_to_private(b),
c_ = simde__m128i_to_private(c);
#if defined(SIMDE_ARM_NEON_A64V8_NATIVE)
int32x4_t even = vuzp2q_s32(a_.neon_i32, b_.neon_i32);
r_.neon_i64 = vaddq_s64(vmull_s32(vget_low_s32(even), vget_high_s32(even)), c_.neon_i64);
#else
SIMDE_VECTORIZE
for (size_t i = 0 ; i < (sizeof(r_.i64) / sizeof(r_.i64[0])) ; i++) {
r_.i64[i] = (HEDLEY_STATIC_CAST(int64_t, a_.i32[(i * 2) + 1]) * HEDLEY_STATIC_CAST(int64_t, b_.i32[(i * 2) + 1])) + c_.i64[i];
}
#endif
return simde__m128i_from_private(r_);
#endif
......@@ -2614,6 +2781,13 @@ simde_mm_maccs_epi16 (simde__m128i a, simde__m128i b, simde__m128i c) {
b_ = simde__m128i_to_private(b),
c_ = simde__m128i_to_private(c);
#if defined(SIMDE_ARM_NEON_A64V8_NATIVE)
int32x4_t c_lo = vmovl_s16(vget_low_s16(c_.i16));
int32x4_t c_hi = vmovl_high_s16(c_.i16);
int32x4_t lo = vmlal_s16(c_lo, vget_low_s16(a_.neon_i16), vget_low_s16(b_.neon_i16));
int32x4_t hi = vmlal_high_s16(c_hi, a_.neon_i16, b_.neon_i16);
r_.neon_i16 = vcombine_s16(vqmovn_s32(lo), vqmovn_s32(hi));
#else
SIMDE_VECTORIZE
for (size_t i = 0 ; i < (sizeof(r_.i16) / sizeof(r_.i16[0])) ; i++) {
int32_t tmp = HEDLEY_STATIC_CAST(int32_t, a_.i16[i]) * HEDLEY_STATIC_CAST(int32_t, b_.i16[i]);
......@@ -2625,6 +2799,7 @@ simde_mm_maccs_epi16 (simde__m128i a, simde__m128i b, simde__m128i c) {
else
r_.i16[i] = HEDLEY_STATIC_CAST(int16_t, tmp);
}
#endif
return simde__m128i_from_private(r_);
#endif
......@@ -2645,6 +2820,13 @@ simde_mm_maccs_epi32 (simde__m128i a, simde__m128i b, simde__m128i c) {
b_ = simde__m128i_to_private(b),
c_ = simde__m128i_to_private(c);
#if defined(SIMDE_ARM_NEON_A64V8_NATIVE)
int64x2_t c_lo = vmovl_s32(vget_low_s32(c_.i32));
int64x2_t c_hi = vmovl_high_s32(c_.i32);
int64x2_t lo = vmlal_s32(c_lo, vget_low_s32(a_.neon_i32), vget_low_s32(b_.neon_i32));
int64x2_t hi = vmlal_high_s32(c_hi, a_.neon_i32, b_.neon_i32);
r_.neon_i32 = vcombine_s32(vqmovn_s64(lo), vqmovn_s64(hi));
#else
SIMDE_VECTORIZE
for (size_t i = 0 ; i < (sizeof(r_.i32) / sizeof(r_.i32[0])) ; i++) {
int64_t tmp = HEDLEY_STATIC_CAST(int64_t, a_.i32[i]) * HEDLEY_STATIC_CAST(int64_t, b_.i32[i]);
......@@ -2656,6 +2838,7 @@ simde_mm_maccs_epi32 (simde__m128i a, simde__m128i b, simde__m128i c) {
else
r_.i32[i] = HEDLEY_STATIC_CAST(int32_t, tmp);
}
#endif
return simde__m128i_from_private(r_);
#endif
......@@ -2676,6 +2859,10 @@ simde_mm_maccsd_epi16 (simde__m128i a, simde__m128i b, simde__m128i c) {
b_ = simde__m128i_to_private(b),
c_ = simde__m128i_to_private(c);
#if defined(SIMDE_ARM_NEON_A64V8_NATIVE)
int16x8_t even = vuzp1q_s16(a_.neon_i16, b_.neon_i16);
r_.neon_i32 = vmlal_s16(c_.neon_i32, vget_low_s16(even), vget_high_s16(even));
#else
SIMDE_VECTORIZE
for (size_t i = 0 ; i < (sizeof(r_.i32) / sizeof(r_.i32[0])) ; i++) {
int64_t tmp = HEDLEY_STATIC_CAST(int32_t, a_.i16[i * 2]) * HEDLEY_STATIC_CAST(int32_t, b_.i16[i * 2]);
......@@ -2687,6 +2874,7 @@ simde_mm_maccsd_epi16 (simde__m128i a, simde__m128i b, simde__m128i c) {
else
r_.i32[i] = HEDLEY_STATIC_CAST(int32_t, tmp);
}
#endif
return simde__m128i_from_private(r_);
#endif
......@@ -2813,6 +3001,9 @@ simde_mm_sha_epi8 (simde__m128i a, simde__m128i b) {
a_ = simde__m128i_to_private(a),
b_ = simde__m128i_to_private(b);
#if defined(SIMDE_ARM_NEON_A32V7_NATIVE)
r_.neon_i8 = vshlq_s8(a_.neon_i8, b_.neon_i8);
#else
SIMDE_VECTORIZE
for (size_t i = 0 ; i < (sizeof(r_.i8) / sizeof(r_.i8[0])) ; i++) {
if (b_.i8[i] < 0) {
......@@ -2821,6 +3012,7 @@ simde_mm_sha_epi8 (simde__m128i a, simde__m128i b) {
r_.i8[i] = HEDLEY_STATIC_CAST(int8_t, a_.i8[i] << b_.i8[i]);
}
}
#endif
return simde__m128i_from_private(r_);
#endif
......@@ -2840,6 +3032,9 @@ simde_mm_sha_epi16 (simde__m128i a, simde__m128i b) {
a_ = simde__m128i_to_private(a),
b_ = simde__m128i_to_private(b);
#if defined(SIMDE_ARM_NEON_A32V7_NATIVE)
r_.neon_i16 = vshlq_s16(a_.neon_i16, b_.neon_i16);
#else
SIMDE_VECTORIZE
for (size_t i = 0 ; i < (sizeof(r_.i16) / sizeof(r_.i16[0])) ; i++) {
if (b_.i16[i] < 0) {
......@@ -2848,6 +3043,7 @@ simde_mm_sha_epi16 (simde__m128i a, simde__m128i b) {
r_.i16[i] = HEDLEY_STATIC_CAST(int16_t, a_.i16[i] << b_.i16[i]);
}
}
#endif
return simde__m128i_from_private(r_);
#endif
......@@ -2867,6 +3063,9 @@ simde_mm_sha_epi32 (simde__m128i a, simde__m128i b) {
a_ = simde__m128i_to_private(a),
b_ = simde__m128i_to_private(b);
#if defined(SIMDE_ARM_NEON_A32V7_NATIVE)
r_.neon_i32 = vshlq_s32(a_.neon_i32, b_.neon_i32);
#else
SIMDE_VECTORIZE
for (size_t i = 0 ; i < (sizeof(r_.i32) / sizeof(r_.i32[0])) ; i++) {
if (b_.i32[i] < 0) {
......@@ -2875,6 +3074,7 @@ simde_mm_sha_epi32 (simde__m128i a, simde__m128i b) {
r_.i32[i] = HEDLEY_STATIC_CAST(int32_t, a_.i32[i] << b_.i32[i]);
}
}
#endif
return simde__m128i_from_private(r_);
#endif
......@@ -2894,6 +3094,9 @@ simde_mm_sha_epi64 (simde__m128i a, simde__m128i b) {
a_ = simde__m128i_to_private(a),
b_ = simde__m128i_to_private(b);
#if defined(SIMDE_ARM_NEON_A32V7_NATIVE)
r_.neon_i64 = vshlq_s64(a_.neon_i64, b_.neon_i64);
#else
SIMDE_VECTORIZE
for (size_t i = 0 ; i < (sizeof(r_.i64) / sizeof(r_.i64[0])) ; i++) {
if (b_.i64[i] < 0) {
......@@ -2902,6 +3105,7 @@ simde_mm_sha_epi64 (simde__m128i a, simde__m128i b) {
r_.i64[i] = HEDLEY_STATIC_CAST(int64_t, a_.i64[i] << b_.i64[i]);
}
}
#endif
return simde__m128i_from_private(r_);
#endif
......@@ -2921,6 +3125,9 @@ simde_mm_shl_epi8 (simde__m128i a, simde__m128i b) {
a_ = simde__m128i_to_private(a),
b_ = simde__m128i_to_private(b);
#if defined(SIMDE_ARM_NEON_A32V7_NATIVE)
r_.neon_u8 = vshlq_u8(a_.neon_u8, b_.neon_i8);
#else
SIMDE_VECTORIZE
for (size_t i = 0 ; i < (sizeof(r_.u8) / sizeof(r_.u8[0])) ; i++) {
if (HEDLEY_UNLIKELY(b_.i8[i] < -7 || b_.i8[i] > 7)) {
......@@ -2933,6 +3140,7 @@ simde_mm_shl_epi8 (simde__m128i a, simde__m128i b) {
}
}
}
#endif
return simde__m128i_from_private(r_);
#endif
......@@ -2952,6 +3160,9 @@ simde_mm_shl_epi16 (simde__m128i a, simde__m128i b) {
a_ = simde__m128i_to_private(a),
b_ = simde__m128i_to_private(b);
#if defined(SIMDE_ARM_NEON_A32V7_NATIVE)
r_.neon_u16 = vshlq_u16(a_.neon_u16, b_.neon_i16);
#else
SIMDE_VECTORIZE
for (size_t i = 0 ; i < (sizeof(r_.u16) / sizeof(r_.u16[0])) ; i++) {
if (HEDLEY_UNLIKELY(b_.i16[i] < -15 || b_.i16[i] > 15)) {
......@@ -2964,6 +3175,7 @@ simde_mm_shl_epi16 (simde__m128i a, simde__m128i b) {
}
}
}
#endif
return simde__m128i_from_private(r_);
#endif
......@@ -2983,6 +3195,9 @@ simde_mm_shl_epi32 (simde__m128i a, simde__m128i b) {
a_ = simde__m128i_to_private(a),
b_ = simde__m128i_to_private(b);
#if defined(SIMDE_ARM_NEON_A32V7_NATIVE)
r_.neon_u32 = vshlq_u32(a_.neon_u32, b_.neon_i32);
#else
SIMDE_VECTORIZE
for (size_t i = 0 ; i < (sizeof(r_.u32) / sizeof(r_.u32[0])) ; i++) {
if (HEDLEY_UNLIKELY(b_.i32[i] < -31 || b_.i32[i] > 31)) {
......@@ -2995,6 +3210,7 @@ simde_mm_shl_epi32 (simde__m128i a, simde__m128i b) {
}
}
}
#endif
return simde__m128i_from_private(r_);
#endif
......@@ -3014,6 +3230,9 @@ simde_mm_shl_epi64 (simde__m128i a, simde__m128i b) {
a_ = simde__m128i_to_private(a),
b_ = simde__m128i_to_private(b);
#if defined(SIMDE_ARM_NEON_A32V7_NATIVE)
r_.neon_u64 = vshlq_u64(a_.neon_u64, b_.neon_i64);
#else
SIMDE_VECTORIZE
for (size_t i = 0 ; i < (sizeof(r_.u64) / sizeof(r_.u64[0])) ; i++) {
if (HEDLEY_UNLIKELY(b_.i64[i] < -63 || b_.i64[i] > 63)) {
......@@ -3026,6 +3245,7 @@ simde_mm_shl_epi64 (simde__m128i a, simde__m128i b) {
}
}
}
#endif
return simde__m128i_from_private(r_);
#endif
......
Markdown is supported
0% or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment