A64: Implement DUP (general)

src/CMakeLists.txt

@@ -75,6 +75,7 @@ add_library(dynarmic
     frontend/A64/translate/impl/load_store_register_immediate.cpp
     frontend/A64/translate/impl/load_store_register_pair.cpp
     frontend/A64/translate/impl/move_wide.cpp
+    frontend/A64/translate/impl/simd_copy.cpp
     frontend/A64/translate/impl/simd_three_same.cpp
     frontend/A64/translate/impl/system.cpp
     frontend/A64/translate/translate.cpp

src/common/bit_util.h

@@ -98,6 +98,17 @@ inline int HighestSetBit(T value) {
     return result;
 }
 
+template <typename T>
+inline size_t LowestSetBit(T value) {
+    auto x = static_cast<std::make_unsigned_t<T>>(value);
+    size_t result = 0;
+    while ((x & 1) == 0) {
+        x >>= 1;
+        result++;
+    }
+    return result;
+}
+
 template <typename T>
 inline T Ones(size_t count) {
     ASSERT_MSG(count <= BitSize<T>(), "count larger than bitsize of T");

src/frontend/A64/decoder/a64.inc

@@ -689,7 +689,7 @@ INST(ADD_vector,             "ADD (vector)",                              "0Q001
 //INST(EXT,                    "EXT",                                       "0Q101110000mmmmm0iiii0nnnnnddddd")
 
 // Data Processing - FP and SIMD - SIMD Copy
-//INST(DUP_gen,                "DUP (general)",                             "0Q001110000iiiii000011nnnnnddddd")
+INST(DUP_gen,                "DUP (general)",                             "0Q001110000iiiii000011nnnnnddddd")
 //INST(SMOV,                   "SMOV",                                      "0Q001110000iiiii001011nnnnnddddd")
 //INST(UMOV,                   "UMOV",                                      "0Q001110000iiiii001111nnnnnddddd")
 //INST(INS_gen,                "INS (general)",                             "01001110000iiiii000111nnnnnddddd")

src/frontend/A64/translate/impl/simd_copy.cpp

@@ -0,0 +1,41 @@
+/* This file is part of the dynarmic project.
+ * Copyright (c) 2018 MerryMage
+ * This software may be used and distributed according to the terms of the GNU
+ * General Public License version 2 or any later version.
+ */
+
+#include "common/bit_util.h"
+#include "frontend/A64/translate/impl/impl.h"
+
+namespace Dynarmic {
+namespace A64 {
+
+bool TranslatorVisitor::DUP_gen(bool Q, Imm<5> imm5, Reg Rn, Vec Vd) {
+    const size_t size = Common::LowestSetBit(imm5.ZeroExtend());
+    if (size > 3) return UnallocatedEncoding();
+    if (size == 3 && !Q) return ReservedValue();
+    const size_t esize = 8 << size;
+    const size_t datasize = Q ? 128 : 64;
+
+    const IR::UAny element = X(esize, Rn);
+
+    const IR::U128 result = [&]{
+        switch (esize) {
+        case 8:
+            return Q ? ir.VectorBroadcast8(element) : ir.VectorLowerBroadcast8(element);
+        case 16:
+            return Q ? ir.VectorBroadcast16(element) : ir.VectorLowerBroadcast16(element);
+        case 32:
+            return Q ? ir.VectorBroadcast32(element) : ir.VectorLowerBroadcast32(element);
+        default:
+            return ir.VectorBroadcast64(element);
+        }
+    }();
+
+    V(datasize, Vd, result);
+
+    return true;
+}
+
+} // namespace A64
+} // namespace Dynarmic