[CPU] [ARM64] jit equal (openvinotoolkit#23266)

### Details: - *[CPU] [AARCH64] jit eltwise Equal ### Tickets: - *CVS-134691*
alvoron · Apr 29, 2024 · ed37a5c · ed37a5c
1 parent a647402
commit ed37a5c
Show file tree

Hide file tree

Showing 4 changed files with 81 additions and 0 deletions.
diff --git a/src/plugins/intel_cpu/src/emitters/plugin/aarch64/jit_eltwise_emitters.cpp b/src/plugins/intel_cpu/src/emitters/plugin/aarch64/jit_eltwise_emitters.cpp
@@ -172,6 +172,57 @@ std::set<std::vector<element::Type>> jit_divide_emitter::get_supported_precision
     return {{element::f32, element::f32}};
 }
 
+/// EQUAL ///
+jit_equal_emitter::jit_equal_emitter(dnnl::impl::cpu::aarch64::jit_generator *host,
+                                     dnnl::impl::cpu::aarch64::cpu_isa_t host_isa,
+                                     const std::shared_ptr<ov::Node>& node)
+                                     : jit_emitter(host, host_isa, get_arithmetic_binary_exec_precision(node)) {
+    prepare_table();
+}
+jit_equal_emitter::jit_equal_emitter(dnnl::impl::cpu::aarch64::jit_generator *host,
+                                     dnnl::impl::cpu::aarch64::cpu_isa_t host_isa,
+                                     const ov::element::Type exec_prc)
+                                     : jit_emitter(host, host_isa, exec_prc) {
+    prepare_table();
+}
+
+size_t jit_equal_emitter::get_inputs_count() const { return 2; }
+
+size_t jit_equal_emitter::get_aux_vecs_count() const { return 1; }
+
+size_t jit_equal_emitter::get_aux_gprs_count() const { return 1; }
+
+std::set<std::vector<element::Type>> jit_equal_emitter::get_supported_precisions(const std::shared_ptr<ov::Node>& node) {
+    return {{element::f32, element::f32}};
+}
+
+void jit_equal_emitter::emit_impl(const std::vector<size_t>& in_vec_idxs, const std::vector<size_t>& out_vec_idxs) const {
+    if (host_isa_ == dnnl::impl::cpu::aarch64::asimd) {
+        emit_isa<dnnl::impl::cpu::aarch64::asimd>(in_vec_idxs, out_vec_idxs);
+    } else {
+        OV_CPU_JIT_EMITTER_THROW("Can't create jit eltwise kernel");
+    }
+}
+
+template <dnnl::impl::cpu::aarch64::cpu_isa_t isa>
+void jit_equal_emitter::emit_isa(const std::vector<size_t> &in_vec_idxs, const std::vector<size_t> &out_vec_idxs) const {
+    OV_CPU_JIT_EMITTER_ASSERT(exec_prc_ == ov::element::f32, "unsupported precision: " + exec_prc_.to_string());
+
+    using TReg = typename dnnl::impl::cpu::aarch64::cpu_isa_traits<isa>::TReg;
+    const TReg src1 = TReg(in_vec_idxs[0]);
+    const TReg src2 = TReg(in_vec_idxs[1]);
+    const TReg dst = TReg(out_vec_idxs[0]);
+    const TReg aux = TReg(aux_vec_idxs[0]);
+
+    h->fcmeq(dst.s, src1.s, src2.s);
+
+    h->ld1r(aux.s, table_val2("one"));
+    h->and_(dst.b16, dst.b16, aux.b16);
+}
+
+void jit_equal_emitter::register_table_entries() {
+    push_arg_entry_of("one", 0x3f800000, true);
+}
 
 /// MUL_ADD ///
 jit_mul_add_emitter::jit_mul_add_emitter(dnnl::impl::cpu::aarch64::jit_generator* host,

diff --git a/src/plugins/intel_cpu/src/emitters/plugin/aarch64/jit_eltwise_emitters.hpp b/src/plugins/intel_cpu/src/emitters/plugin/aarch64/jit_eltwise_emitters.hpp
@@ -84,6 +84,33 @@ class jit_divide_emitter : public jit_emitter {
     void emit_isa(const std::vector<size_t> &in_vec_idxs, const std::vector<size_t> &out_vec_idxs) const;
 };
 
+class jit_equal_emitter : public jit_emitter {
+public:
+    jit_equal_emitter(dnnl::impl::cpu::aarch64::jit_generator *host,
+                      dnnl::impl::cpu::aarch64::cpu_isa_t host_isa,
+                      const ov::element::Type exec_prc = ov::element::f32);
+
+    jit_equal_emitter(dnnl::impl::cpu::aarch64::jit_generator *host,
+                      dnnl::impl::cpu::aarch64::cpu_isa_t host_isa,
+                      const std::shared_ptr<ov::Node>& n);
+
+    size_t get_inputs_count() const override;
+
+    size_t get_aux_vecs_count() const override;
+
+    size_t get_aux_gprs_count() const override;
+
+    static std::set<std::vector<element::Type>> get_supported_precisions(
+            const std::shared_ptr<ov::Node>& node = nullptr);
+
+private:
+    void emit_impl(const std::vector<size_t> &in_vec_idxs, const std::vector<size_t> &out_vec_idxs) const override;
+
+    template <dnnl::impl::cpu::aarch64::cpu_isa_t isa>
+    void emit_isa(const std::vector<size_t> &in_vec_idxs, const std::vector<size_t> &out_vec_idxs) const;
+
+    void register_table_entries() override;
+};
 
 class jit_mul_add_emitter : public jit_emitter {
 public:

diff --git a/src/plugins/intel_cpu/src/nodes/executors/aarch64/jit_eltwise.cpp b/src/plugins/intel_cpu/src/nodes/executors/aarch64/jit_eltwise.cpp
@@ -21,6 +21,7 @@ bool JitEltwiseExecutor::isSupported(
                                      Algorithm::EltwiseAdd,
                                      Algorithm::EltwiseClamp,
                                      Algorithm::EltwiseDivide,
+                                     Algorithm::EltwiseEqual,
                                      Algorithm::EltwiseMultiply,
                                      Algorithm::EltwiseMulAdd,
                                      Algorithm::EltwisePowerStatic,

diff --git a/src/plugins/intel_cpu/src/nodes/kernels/aarch64/jit_uni_eltwise_generic.cpp b/src/plugins/intel_cpu/src/nodes/kernels/aarch64/jit_uni_eltwise_generic.cpp
@@ -612,6 +612,7 @@ std::shared_ptr<jit_emitter> jit_uni_eltwise_generic<isa>::create_eltwise_emitte
     OV_CASE(Algorithm::EltwiseAdd, ov::intel_cpu::aarch64::jit_add_emitter),
     OV_CASE(Algorithm::EltwiseClamp, ov::intel_cpu::aarch64::jit_clamp_emitter),
     OV_CASE(Algorithm::EltwiseDivide, ov::intel_cpu::aarch64::jit_divide_emitter),
+    OV_CASE(Algorithm::EltwiseEqual, ov::intel_cpu::aarch64::jit_equal_emitter),
     OV_CASE(Algorithm::EltwiseMulAdd, ov::intel_cpu::aarch64::jit_mul_add_emitter),
     OV_CASE(Algorithm::EltwiseMultiply, ov::intel_cpu::aarch64::jit_multiply_emitter),
     OV_CASE(Algorithm::EltwisePowerStatic, ov::intel_cpu::aarch64::jit_power_static_emitter),
@@ -767,6 +768,7 @@ std::set<std::vector<element::Type>> eltwise_precision_helper::get_supported_pre
         OV_CASE(Algorithm::EltwiseAdd, jit_add_emitter),
         OV_CASE(Algorithm::EltwiseClamp, jit_clamp_emitter),
         OV_CASE(Algorithm::EltwiseDivide, jit_divide_emitter),
+        OV_CASE(Algorithm::EltwiseEqual, jit_equal_emitter),
         OV_CASE(Algorithm::EltwiseMulAdd, jit_mul_add_emitter),
         OV_CASE(Algorithm::EltwiseMultiply, jit_multiply_emitter),
         OV_CASE(Algorithm::EltwisePrelu, jit_prelu_emitter),