1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
|
/*
Copyright 2016-2022 melonDS team, RSDuck
This file is part of melonDS.
melonDS is free software: you can redistribute it and/or modify it under
the terms of the GNU General Public License as published by the Free
Software Foundation, either version 3 of the License, or (at your option)
any later version.
melonDS is distributed in the hope that it will be useful, but WITHOUT ANY
WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS
FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.
You should have received a copy of the GNU General Public License along
with melonDS. If not, see http://www.gnu.org/licenses/.
*/
#ifndef ARMJIT_A64_COMPILER_H
#define ARMJIT_A64_COMPILER_H
#include "../ARM.h"
#include "../dolphin/Arm64Emitter.h"
#include "../ARMJIT_Internal.h"
#include "../ARMJIT_RegisterCache.h"
#include <unordered_map>
namespace ARMJIT
{
class ARMJIT;
const Arm64Gen::ARM64Reg RMemBase = Arm64Gen::X26;
const Arm64Gen::ARM64Reg RCPSR = Arm64Gen::W27;
const Arm64Gen::ARM64Reg RCycles = Arm64Gen::W28;
const Arm64Gen::ARM64Reg RCPU = Arm64Gen::X29;
struct Op2
{
Op2()
{}
Op2(Arm64Gen::ARM64Reg rm) : IsImm(false)
{
Reg.Rm = rm;
Reg.ShiftType = Arm64Gen::ST_LSL;
Reg.ShiftAmount = 0;
}
Op2(u32 imm) : IsImm(true), Imm(imm)
{}
Op2(Arm64Gen::ARM64Reg rm, Arm64Gen::ShiftType st, int amount) : IsImm(false)
{
Reg.Rm = rm;
Reg.ShiftType = st;
Reg.ShiftAmount = amount;
}
Arm64Gen::ArithOption ToArithOption()
{
assert(!IsImm);
return Arm64Gen::ArithOption(Reg.Rm, Reg.ShiftType, Reg.ShiftAmount);
}
bool IsSimpleReg()
{ return !IsImm && !Reg.ShiftAmount && Reg.ShiftType == Arm64Gen::ST_LSL; }
bool ImmFits12Bit()
{ return IsImm && (Imm & 0xFFF == Imm); }
bool IsZero()
{ return IsImm && !Imm; }
bool IsImm;
union
{
struct
{
Arm64Gen::ARM64Reg Rm;
Arm64Gen::ShiftType ShiftType;
int ShiftAmount;
} Reg;
u32 Imm;
};
};
struct LoadStorePatch
{
void* PatchFunc;
s32 PatchOffset;
u32 PatchSize;
};
class Compiler : public Arm64Gen::ARM64XEmitter
{
public:
typedef void (Compiler::*CompileFunc)();
Compiler(ARMJIT& jit);
~Compiler();
void PushRegs(bool saveHiRegs, bool saveRegsToBeChanged, bool allowUnload = true);
void PopRegs(bool saveHiRegs, bool saveRegsToBeChanged);
Arm64Gen::ARM64Reg MapReg(int reg)
{
assert(RegCache.Mapping[reg] != Arm64Gen::INVALID_REG);
return RegCache.Mapping[reg];
}
JitBlockEntry CompileBlock(ARM* cpu, bool thumb, FetchedInstr instrs[], int instrsCount, bool hasMemInstr);
bool CanCompile(bool thumb, u16 kind);
bool FlagsNZNeeded()
{
return CurInstr.SetFlags & 0xC;
}
void Reset();
void Comp_AddCycles_C(bool forceNonConstant = false);
void Comp_AddCycles_CI(u32 numI);
void Comp_AddCycles_CI(u32 c, Arm64Gen::ARM64Reg numI, Arm64Gen::ArithOption shift);
void Comp_AddCycles_CD();
void Comp_AddCycles_CDI();
void MovePC();
void LoadReg(int reg, Arm64Gen::ARM64Reg nativeReg);
void SaveReg(int reg, Arm64Gen::ARM64Reg nativeReg);
void LoadCPSR();
void SaveCPSR(bool markClean = true);
void LoadCycles();
void SaveCycles();
void Nop() {}
void A_Comp_ALUTriOp();
void A_Comp_ALUMovOp();
void A_Comp_ALUCmpOp();
void A_Comp_Mul();
void A_Comp_Mul_Long();
void A_Comp_Mul_Short();
void A_Comp_Clz();
void A_Comp_MemWB();
void A_Comp_MemHD();
void A_Comp_LDM_STM();
void A_Comp_BranchImm();
void A_Comp_BranchXchangeReg();
void A_Comp_MRS();
void A_Comp_MSR();
void T_Comp_ShiftImm();
void T_Comp_AddSub_();
void T_Comp_ALUImm8();
void T_Comp_ALU();
void T_Comp_ALU_HiReg();
void T_Comp_AddSP();
void T_Comp_RelAddr();
void T_Comp_MemReg();
void T_Comp_MemImm();
void T_Comp_MemRegHalf();
void T_Comp_MemImmHalf();
void T_Comp_LoadPCRel();
void T_Comp_MemSPRel();
void T_Comp_LDMIA_STMIA();
void T_Comp_PUSH_POP();
void T_Comp_BCOND();
void T_Comp_B();
void T_Comp_BranchXchangeReg();
void T_Comp_BL_LONG_1();
void T_Comp_BL_LONG_2();
void T_Comp_BL_Merged();
s32 Comp_MemAccessBlock(int rn, BitSet16 regs, bool store, bool preinc, bool decrement, bool usermode, bool skipLoadingRn);
void Comp_Mul_Mla(bool S, bool mla, Arm64Gen::ARM64Reg rd, Arm64Gen::ARM64Reg rm, Arm64Gen::ARM64Reg rs, Arm64Gen::ARM64Reg rn);
void Comp_Compare(int op, Arm64Gen::ARM64Reg rn, Op2 op2);
void Comp_Logical(int op, bool S, Arm64Gen::ARM64Reg rd, Arm64Gen::ARM64Reg rn, Op2 op2);
void Comp_Arithmetic(int op, bool S, Arm64Gen::ARM64Reg rd, Arm64Gen::ARM64Reg rn, Op2 op2);
void Comp_RetriveFlags(bool retriveCV);
Arm64Gen::FixupBranch CheckCondition(u32 cond);
void Comp_JumpTo(Arm64Gen::ARM64Reg addr, bool switchThumb, bool restoreCPSR = false);
void Comp_JumpTo(u32 addr, bool forceNonConstantCycles = false);
void A_Comp_GetOp2(bool S, Op2& op2);
void Comp_RegShiftImm(int op, int amount, bool S, Op2& op2, Arm64Gen::ARM64Reg tmp = Arm64Gen::W0);
void Comp_RegShiftReg(int op, bool S, Op2& op2, Arm64Gen::ARM64Reg rs);
bool Comp_MemLoadLiteral(int size, bool signExtend, int rd, u32 addr);
enum
{
memop_Writeback = 1 << 0,
memop_Post = 1 << 1,
memop_SignExtend = 1 << 2,
memop_Store = 1 << 3,
memop_SubtractOffset = 1 << 4
};
void Comp_MemAccess(int rd, int rn, Op2 offset, int size, int flags);
// 0 = switch mode, 1 = stay arm, 2 = stay thumb
void* Gen_JumpTo9(int kind);
void* Gen_JumpTo7(int kind);
void Comp_BranchSpecialBehaviour(bool taken);
JitBlockEntry AddEntryOffset(u32 offset)
{
return (JitBlockEntry)(GetRXBase() + offset);
}
u32 SubEntryOffset(JitBlockEntry entry)
{
return (u8*)entry - GetRXBase();
}
bool IsJITFault(u8* pc);
u8* RewriteMemAccess(u8* pc);
void SwapCodeRegion()
{
ptrdiff_t offset = GetCodeOffset();
SetCodePtrUnsafe(OtherCodeRegion);
OtherCodeRegion = offset;
}
ARMJIT& JIT;
ptrdiff_t OtherCodeRegion;
bool Exit;
FetchedInstr CurInstr;
bool Thumb;
u32 R15;
u32 Num;
ARM* CurCPU;
u32 ConstantCycles;
u32 CodeRegion;
BitSet32 SavedRegs;
u32 JitMemSecondarySize;
u32 JitMemMainSize;
std::unordered_map<ptrdiff_t, LoadStorePatch> LoadStorePatches;
RegisterCache<Compiler, Arm64Gen::ARM64Reg> RegCache;
bool CPSRDirty = false;
bool IrregularCycles = false;
#ifdef __SWITCH__
void* JitRWBase;
void* JitRWStart;
void* JitRXStart;
#endif
void* ReadBanked, *WriteBanked;
void* JumpToFuncs9[3];
void* JumpToFuncs7[3];
// [Console Type][Num][Size][Sign Extend][Output register]
void* PatchedLoadFuncs[2][2][3][2][32];
void* PatchedStoreFuncs[2][2][3][32];
};
}
#endif
|