Home
last modified time | relevance | path

Searched refs:MulAdd (Results 1 – 25 of 26) sorted by relevance

12

/external/rust/crates/num-traits/src/ops/
Dmul_add.rs23 pub trait MulAdd<A = Self, B = Self> { trait
38 impl MulAdd<f32, f32> for f32 {
48 impl MulAdd<f64, f64> for f64 {
70 mul_add_impl!(MulAdd for isize usize i8 u8 i16 u16 i32 u32 i64 u64);
72 mul_add_impl!(MulAdd for i128 u128);
119 assert_eq!(MulAdd::mul_add(m, x, b), (m*x + b)); in mul_add_integer()
141 let abs_difference = (MulAdd::mul_add(m, x, b) - (m*x + b)).abs(); in mul_add_float()
/external/tensorflow/tensorflow/compiler/xla/service/cpu/
Dllvm_ir_runtime.cc211 llvm::Value* n = vsl.Floor(vsl.MulAdd(input, cephes_LOG2EF, half)); in GenerateVF32Exp()
259 llvm::Value* z = vsl.MulAdd(x, cephes_exp_p0, cephes_exp_p1); in GenerateVF32Exp()
260 z = vsl.MulAdd(z, x, cephes_exp_p2); in GenerateVF32Exp()
261 z = vsl.MulAdd(z, x, cephes_exp_p3); in GenerateVF32Exp()
262 z = vsl.MulAdd(z, x, cephes_exp_p4); in GenerateVF32Exp()
263 z = vsl.MulAdd(z, x, cephes_exp_p5); in GenerateVF32Exp()
264 z = vsl.MulAdd(z, vsl.Mul(x, x), x); in GenerateVF32Exp()
360 y = vsl.MulAdd(tmp0, cephes_log_p0, cephes_log_p1); in GenerateVF32Log()
361 y1 = vsl.MulAdd(tmp0, cephes_log_p3, cephes_log_p4); in GenerateVF32Log()
362 y2 = vsl.MulAdd(tmp0, cephes_log_p6, cephes_log_p7); in GenerateVF32Log()
[all …]
Dvector_support_library.h89 llvm::Value* MulAdd(llvm::Value* a, llvm::Value* b, llvm::Value* c) { in MulAdd() function
93 llvm::Value* MulAdd(llvm::Value* a, llvm::Value* b, const llvm::APFloat& c) { in MulAdd() function
97 llvm::Value* MulAdd(llvm::Value* a, const llvm::APFloat& b, in MulAdd() function
Dtiled_dot_emitter.cc335 accumulator = vsl_.MulAdd(lhs_tile[i], rhs_tile[i], accumulator); in EmitInnerLoopTiled()
929 vsl->MulAdd(lhs_tile[r_m_i][r_k_i], rhs_tile[r_k_i], in EmitTiledGemm()
/external/ruy/ruy/
Dkernel_x86.h250 inline __m256 MulAdd(const __m256&, const __m256&, const __m256&) {
620 accum_data_v[0] = intrin_utils::MulAdd<path>(
624 accum_data_v[1] = intrin_utils::MulAdd<path>(
628 accum_data_v[2] = intrin_utils::MulAdd<path>(
632 accum_data_v[3] = intrin_utils::MulAdd<path>(
636 accum_data_v[4] = intrin_utils::MulAdd<path>(
640 accum_data_v[5] = intrin_utils::MulAdd<path>(
644 accum_data_v[6] = intrin_utils::MulAdd<path>(
648 accum_data_v[7] = intrin_utils::MulAdd<path>(
719 accum_data_v[0] = intrin_utils::MulAdd<path>(
[all …]
Dkernel_avx2_fma.cc72 inline __m256 MulAdd<Path::kAvx2Fma>(const __m256& a, const __m256& b,
Dkernel_avx.cc434 inline __m256 MulAdd<Path::kAvx>(const __m256& a, const __m256& b,
/external/llvm-project/llvm/examples/OrcV2Examples/OrcV2CBindingsReflectProcessSymbols/
DOrcV2CBindingsReflectProcessSymbols.c186 int32_t (*MulAdd)(int32_t, int32_t, int32_t) = in main() local
188 int32_t Result = MulAdd(3, 4, 5); in main()
/external/gemmlowp/internal/
Dsimd_wrappers_sse.h99 inline void MulAdd(Int32x4 lhs, Int32x4 rhs, Int32x4* acc) {
103 inline void MulAdd(Int32x4 lhs, std::int32_t rhs, Int32x4* acc) {
Dsimd_wrappers_msa.h134 inline void MulAdd(Int32x4 lhs, Int32x4 rhs, Int32x4* acc) {
140 inline void MulAdd(Int32x4 lhs, std::int32_t rhs, Int32x4* acc) {
Dsimd_wrappers_common_neon_sse.h700 MulAdd(lhs.buf.reg[i], p, &acc->buf.reg[i]);
732 MulAdd(lhs.buf.reg[i], p, &acc->buf.reg[i]);
757 MulAdd(lhs.buf.reg[0], rhs.buf.reg[0], &acc->buf.reg[0]);
820 MulAdd(lhs.buf.reg[0], p, &acc->buf.reg[0]);
844 MulAdd(lhs.buf.reg[0], p, &acc->buf.reg[0]);
Dsimd_wrappers_neon.h129 inline void MulAdd(Int32x4 lhs, Int32x4 rhs, Int32x4* acc) {
133 inline void MulAdd(Int32x4 lhs, std::int32_t rhs, Int32x4* acc) {
Dsimd_wrappers.h40 inline void MulAdd(std::int32_t lhs, std::int32_t rhs, std::int32_t* acc) { in MulAdd() function
419 MulAdd(lhs.buf.reg[lhs_r + lhs_c * LhsRows],
/external/rust/crates/num-traits/
DREADME.md40 The `FloatCore` trait is always available. `MulAdd` and `MulAddAssign` for `f32`
DRELEASES.md139 - [The new `MulAdd` and `MulAddAssign` traits][59] perform a fused multiply-
/external/tensorflow/tensorflow/core/kernels/
Dsparse_matmul_op.cc409 ALWAYS_INLINE void MulAdd(const Packet a, const bfloat16** binp, float** out) { in MulAdd() function
523 ALWAYS_INLINE void MulAdd(const Packet a, const float** inp, float** out) { in MulAdd() function
732 MulAdd(l, &r, &out); in GEPP()
733 MulAdd(nl, &nr, &nout); in GEPP()
734 MulAdd(n2l, &n2r, &n2out); in GEPP()
735 MulAdd(n3l, &n3r, &n3out); in GEPP()
756 MulAdd(l, &r, &out); in GEPP()
/external/tensorflow/tensorflow/lite/kernels/internal/optimized/
Dneon_tensor_utils.cc1467 inline int64x2x2_t MulAdd(int32x4_t acc, int32x4_t lhs, int32x4_t rhs) { in MulAdd() function
1579 MulAdd(bias_0, rescaled.val[0], layer_norm_weights_s32_0); in NeonApplyLayerNorm()
1581 MulAdd(bias_1, rescaled.val[1], layer_norm_weights_s32_1); in NeonApplyLayerNorm()
1583 MulAdd(bias_2, rescaled.val[2], layer_norm_weights_s32_2); in NeonApplyLayerNorm()
1585 MulAdd(bias_3, rescaled.val[3], layer_norm_weights_s32_3); in NeonApplyLayerNorm()
/external/rust/crates/num-traits/src/
Dlib.rs44 pub use ops::mul_add::{MulAdd, MulAddAssign};
/external/clang/test/SemaCXX/
Dconstant-expression-cxx11.cpp501 constexpr int MulAdd(int x, int y, int c) { return x * y + c; } in MulAdd() function
502 constexpr int InnerProduct = ZipFoldR(MulAdd, 5, xs, ys, 0);
/external/llvm-project/clang/test/SemaCXX/
Dconstant-expression-cxx11.cpp545 constexpr int MulAdd(int x, int y, int c) { return x * y + c; } in MulAdd() function
546 constexpr int InnerProduct = ZipFoldR(MulAdd, 5, xs, ys, 0);
/external/swiftshader/tests/ReactorUnitTests/
DReactorUnitTests.cpp1660 TEST(ReactorUnitTests, MulAdd) in TEST() argument
1667 MulAdd(Short4(0x1aa, 0x2dd, 0x3ee, 0xF422), in TEST()
/external/swiftshader/src/Pipeline/
DSamplerCore.cpp1418 Int2 i01 = MulAdd(uv0uv1, *Pointer<Short4>(mipmap + OFFSET(Mipmap, onePitchP))); in computeIndices()
1419 Int2 i23 = MulAdd(uv2uv3, *Pointer<Short4>(mipmap + OFFSET(Mipmap, onePitchP))); in computeIndices()
/external/swiftshader/src/Shader/
DSamplerCore.cpp1678 uuuu = As<Short4>(MulAdd(uuuu, *Pointer<Short4>(mipmap + OFFSET(Mipmap,onePitchP)))); in computeIndices()
1679 uuu2 = As<Short4>(MulAdd(uuu2, *Pointer<Short4>(mipmap + OFFSET(Mipmap,onePitchP)))); in computeIndices()
/external/swiftshader/src/Reactor/
DReactor.hpp943 RValue<Int2> MulAdd(RValue<Short4> x, RValue<Short4> y);
1070 RValue<Int4> MulAdd(RValue<Short8> x, RValue<Short8> y);
DSubzeroReactor.cpp2708 RValue<Int2> MulAdd(RValue<Short4> x, RValue<Short4> y) in MulAdd() function
3115 RValue<Int4> MulAdd(RValue<Short8> x, RValue<Short8> y) in MulAdd() function

12