Home
last modified time | relevance | path

Searched refs:MulAdd (Results 1 – 16 of 16) sorted by relevance

/external/tensorflow/tensorflow/compiler/xla/service/cpu/
Dllvm_ir_runtime.cc143 llvm::Value* fx = vsl.Floor(vsl.MulAdd(input_clamped, cephes_LOG2EF, half)); in GenerateVF32Exp()
150 llvm::Value* y = vsl.MulAdd(x, cephes_exp_p0, cephes_exp_p1); in GenerateVF32Exp()
151 y = vsl.MulAdd(y, x, cephes_exp_p2); in GenerateVF32Exp()
152 y = vsl.MulAdd(y, x, cephes_exp_p3); in GenerateVF32Exp()
153 y = vsl.MulAdd(y, x, cephes_exp_p4); in GenerateVF32Exp()
154 y = vsl.MulAdd(y, x, cephes_exp_p5); in GenerateVF32Exp()
155 y = vsl.MulAdd(y, z, x); in GenerateVF32Exp()
247 y = vsl.MulAdd(tmp0, cephes_log_p0, cephes_log_p1); in GenerateVF32Log()
248 y1 = vsl.MulAdd(tmp0, cephes_log_p3, cephes_log_p4); in GenerateVF32Log()
249 y2 = vsl.MulAdd(tmp0, cephes_log_p6, cephes_log_p7); in GenerateVF32Log()
[all …]
Dvector_support_library.h87 llvm::Value* MulAdd(llvm::Value* a, llvm::Value* b, llvm::Value* c) { in MulAdd() function
91 llvm::Value* MulAdd(llvm::Value* a, llvm::Value* b, const llvm::APFloat& c) { in MulAdd() function
95 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/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_neon.h116 inline void MulAdd(Int32x4 lhs, Int32x4 rhs, Int32x4* acc) {
120 inline void MulAdd(Int32x4 lhs, std::int32_t rhs, Int32x4* acc) {
Dsimd_wrappers_msa.h139 inline void MulAdd(Int32x4 lhs, Int32x4 rhs, Int32x4* acc) {
145 inline void MulAdd(Int32x4 lhs, std::int32_t rhs, Int32x4* acc) {
Dsimd_wrappers_common_neon_sse.h496 MulAdd(lhs.buf.reg[i], p, &acc->buf.reg[i]);
528 MulAdd(lhs.buf.reg[i], p, &acc->buf.reg[i]);
553 MulAdd(lhs.buf.reg[0], rhs.buf.reg[0], &acc->buf.reg[0]);
616 MulAdd(lhs.buf.reg[0], p, &acc->buf.reg[0]);
640 MulAdd(lhs.buf.reg[0], p, &acc->buf.reg[0]);
Dsimd_wrappers.h40 inline void MulAdd(std::int32_t lhs, std::int32_t rhs, std::int32_t* acc) { in MulAdd() function
268 MulAdd(lhs.buf.reg[lhs_r + lhs_c * LhsRows],
/external/tensorflow/tensorflow/core/kernels/
Dsparse_matmul_op.cc410 ALWAYS_INLINE void MulAdd(const Packet a, const bfloat16** binp, float** out) { in MulAdd() function
524 ALWAYS_INLINE void MulAdd(const Packet a, const float** inp, float** out) { in MulAdd() function
733 MulAdd(l, &r, &out); in GEPP()
734 MulAdd(nl, &nr, &nout); in GEPP()
735 MulAdd(n2l, &n2r, &n2out); in GEPP()
736 MulAdd(n3l, &n3r, &n3out); in GEPP()
757 MulAdd(l, &r, &out); in GEPP()
/external/swiftshader/src/Reactor/
DReactorUnitTests.cpp959 TEST(ReactorUnitTests, MulAdd) in TEST() argument
969 MulAdd(Short4(0x1aa, 0x2dd, 0x3ee, 0xF422), in TEST()
DReactor.hpp800 RValue<Int2> MulAdd(RValue<Short4> x, RValue<Short4> y);
927 RValue<Int4> MulAdd(RValue<Short8> x, RValue<Short8> y);
DLLVMReactor.cpp3545 RValue<Int2> MulAdd(RValue<Short4> x, RValue<Short4> y) in MulAdd() function
3934 RValue<Int4> MulAdd(RValue<Short8> x, RValue<Short8> y) in MulAdd() function
DSubzeroReactor.cpp3625 RValue<Int2> MulAdd(RValue<Short4> x, RValue<Short4> y) in MulAdd() function
4195 RValue<Int4> MulAdd(RValue<Short8> x, RValue<Short8> y) in MulAdd() function
/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/swiftshader/src/Pipeline/
DSamplerCore.cpp1586 uuuu = As<Short4>(MulAdd(uuuu, *Pointer<Short4>(mipmap + OFFSET(Mipmap,onePitchP)))); in computeIndices()
1587 uuu2 = As<Short4>(MulAdd(uuu2, *Pointer<Short4>(mipmap + OFFSET(Mipmap,onePitchP)))); in computeIndices()
/external/swiftshader/src/Shader/
DSamplerCore.cpp1677 uuuu = As<Short4>(MulAdd(uuuu, *Pointer<Short4>(mipmap + OFFSET(Mipmap,onePitchP)))); in computeIndices()
1678 uuu2 = As<Short4>(MulAdd(uuu2, *Pointer<Short4>(mipmap + OFFSET(Mipmap,onePitchP)))); in computeIndices()