Files
Xenia-Canary/src/xenia/cpu/ppc/ppc_context.h
chss95cs@gmail.com 33a6cfc0a7 Add special cases to DOT_PRODUCT_3/4 that detect whether they're calculating lengthsquared
Add alternate path to DOT_PRODUCT_3/4 for use_fast_dot_product that skips all the status register stuff and just remaps inf to qnan
Add OPCODE_TO_SINGLE to replace the CONVERT_F32_F64 - CONVERT_F64_F32 sequence we used to emit with the idea that a backend could implement a more correct rounding behavior if possible on its arch
Remove some impossible sequences like MUL_HI_I8/I16, MUL_ADD_F32, DIV_V128. These instructions have no equivalent in PPC. Many other instructions are unused/dead code and should be removed to make the x64 backend a better reference for future ones
Add backend_flags to Instr. Basically, flags field that a backend can use for whatever it wants when generating code.
Add backend instr flag to x64 that tells it to not generate code for an instruction. this allows sequences to consume subsequent instructions
Generate actual x64 code for VSL instruction instead of using callnativesafe
Detect repeated COMPARE instructions w/ identical operands and reuse the results in FLAGS if so. this eliminates a ton of garbage compare/set instructions.
If a COMPARE instructions destination is stored to context with no intervening instruction and no additional uses besides the store, do setx [ctx address]
Detect prefetchw and use it in CACHE_CONTROL if prefetch for write is requested instead of doing prefetch to all cache levels
Fixed an accident in an earlier commit by me, VECTOR_DENORMFLUSH was not being emitted at all, so denormal inputs to MUL_ADD_V128 were not becoming zero and outputs from DOT_PRODUCT_X were not either. I believe this introduced a bug into RDR where a wagon wouldnt spawn? (https://discord.com/channels/308194948048486401/308207592482668545/1000443975817252874)
Compute fresx in double precision using RECIP_F64 and then round to single instead of doing (double)(1.0f / (float)value), matching original behavior better
Refactor some of ppc_emit_fpu, much of the InstrEmit function are identical except for whether they round to single or not
Added "tail emitters" to X64Emitter. These are callbacks that get invoked with their label and the X64Emitter after the epilog code. This allows us to move cold code out of the critical path and in the future place constant pools near functions
guest_to_host_thunk/host_to_guest_thunk now gets directly rel32 called, instead of doing a mov
Add X64BackendContext structure, represents data before the start of the PPCContext
Instead of doing branchless sequence, do a compare and jump to tail emitted code for address translation. This makes converting addresses a 3 uop affair in most cases.
Do qnan move for dot product in a tail emitter
Detect whether EFLAGS bits are independent variables for the current cpu (not really detecting it ehe, just checking if zen) and if so generate inc/dec for add/sub 1
Detect whether low 32 bits of membase are 0. If they are then we can use membasereg.cvt32() in place of immediate 0 in many places, particularly in stores
Detect LOAD MODIFY STORE pattern for context variables (currently only done for 64 bit ones) and turn them into modify [context ptr]. This is done for add, sub, and, or, xor, not, neg
Tail emit error handling for TRAP opcodes
Stub out unused trap opcodes like TRAP_TRUE_I32, TRAP_TRUE_I64, TRAP_TRUE_I16 (the call_true/return_true opcodes for these types are also probably unused)
Remove BackpropTruncations. It was poorly written and causes crashes on the game Viva pinata (https://discord.com/channels/308194948048486401/701111856600711208/1000249460451983420)
2022-07-23 12:10:07 -07:00

443 lines
9.2 KiB
C++

/**
******************************************************************************
* Xenia : Xbox 360 Emulator Research Project *
******************************************************************************
* Copyright 2020 Ben Vanik. All rights reserved. *
* Released under the BSD license - see LICENSE in the root for more details. *
******************************************************************************
*/
#ifndef XENIA_CPU_PPC_PPC_CONTEXT_H_
#define XENIA_CPU_PPC_PPC_CONTEXT_H_
#include <cstdint>
#include <mutex>
#include <string>
#include "xenia/base/vec128.h"
namespace xe {
namespace cpu {
class Processor;
class ThreadState;
} // namespace cpu
namespace kernel {
class KernelState;
} // namespace kernel
} // namespace xe
namespace xe {
namespace cpu {
namespace ppc {
// Map:
// 0-31: GPR
// 32-63: FPR
// 64: LR
// 65: CTR
// 66: XER
// 67: FPSCR
// 68: VSCR
// 69-76: CR0-7
// 100: invalid
// 128-256: VR
enum class PPCRegister {
kR0 = 0,
kR1,
kR2,
kR3,
kR4,
kR5,
kR6,
kR7,
kR8,
kR9,
kR10,
kR11,
kR12,
kR13,
kR14,
kR15,
kR16,
kR17,
kR18,
kR19,
kR20,
kR21,
kR22,
kR23,
kR24,
kR25,
kR26,
kR27,
kR28,
kR29,
kR30,
kR31,
kFR0 = 32,
kFR1,
kFR2,
kFR3,
kFR4,
kFR5,
kFR6,
kFR7,
kFR8,
kFR9,
kFR10,
kFR11,
kFR12,
kFR13,
kFR14,
kFR15,
kFR16,
kFR17,
kFR18,
kFR19,
kFR20,
kFR21,
kFR22,
kFR23,
kFR24,
kFR25,
kFR26,
kFR27,
kFR28,
kFR29,
kFR30,
kFR31,
kVR0 = 64,
kVR1,
kVR2,
kVR3,
kVR4,
kVR5,
kVR6,
kVR7,
kVR8,
kVR9,
kVR10,
kVR11,
kVR12,
kVR13,
kVR14,
kVR15,
kVR16,
kVR17,
kVR18,
kVR19,
kVR20,
kVR21,
kVR22,
kVR23,
kVR24,
kVR25,
kVR26,
kVR27,
kVR28,
kVR29,
kVR30,
kVR31,
kVR32,
kVR33,
kVR34,
kVR35,
kVR36,
kVR37,
kVR38,
kVR39,
kVR40,
kVR41,
kVR42,
kVR43,
kVR44,
kVR45,
kVR46,
kVR47,
kVR48,
kVR49,
kVR50,
kVR51,
kVR52,
kVR53,
kVR54,
kVR55,
kVR56,
kVR57,
kVR58,
kVR59,
kVR60,
kVR61,
kVR62,
kVR63,
kVR64,
kVR65,
kVR66,
kVR67,
kVR68,
kVR69,
kVR70,
kVR71,
kVR72,
kVR73,
kVR74,
kVR75,
kVR76,
kVR77,
kVR78,
kVR79,
kVR80,
kVR81,
kVR82,
kVR83,
kVR84,
kVR85,
kVR86,
kVR87,
kVR88,
kVR89,
kVR90,
kVR91,
kVR92,
kVR93,
kVR94,
kVR95,
kVR96,
kVR97,
kVR98,
kVR99,
kVR100,
kVR101,
kVR102,
kVR103,
kVR104,
kVR105,
kVR106,
kVR107,
kVR108,
kVR109,
kVR110,
kVR111,
kVR112,
kVR113,
kVR114,
kVR115,
kVR116,
kVR117,
kVR118,
kVR119,
kVR120,
kVR121,
kVR122,
kVR123,
kVR124,
kVR125,
kVR126,
kVR127,
kVR128,
kLR,
kCTR,
kXER,
kFPSCR,
kVSCR,
kCR,
};
#pragma pack(push, 8)
typedef struct alignas(64) PPCContext_s {
union {
uint32_t value;
struct {
uint8_t cr0_lt; // Negative (LT) - result is negative
uint8_t cr0_gt; // Positive (GT) - result is positive (and not zero)
uint8_t cr0_eq; // Zero (EQ) - result is zero or a stwcx/stdcx completed
// successfully
uint8_t cr0_so; // Summary Overflow (SO) - copy of XER[SO]
};
} cr0; // 0xA24
union {
uint32_t value;
struct {
uint8_t cr1_fx; // FP exception summary - copy of FPSCR[FX]
uint8_t cr1_fex; // FP enabled exception summary - copy of FPSCR[FEX]
uint8_t
cr1_vx; // FP invalid operation exception summary - copy of FPSCR[VX]
uint8_t cr1_ox; // FP overflow exception - copy of FPSCR[OX]
};
} cr1;
union {
uint32_t value;
struct {
uint8_t cr2_0;
uint8_t cr2_1;
uint8_t cr2_2;
uint8_t cr2_3;
};
} cr2;
union {
uint32_t value;
struct {
uint8_t cr3_0;
uint8_t cr3_1;
uint8_t cr3_2;
uint8_t cr3_3;
};
} cr3;
union {
uint32_t value;
struct {
uint8_t cr4_0;
uint8_t cr4_1;
uint8_t cr4_2;
uint8_t cr4_3;
};
} cr4;
union {
uint32_t value;
struct {
uint8_t cr5_0;
uint8_t cr5_1;
uint8_t cr5_2;
uint8_t cr5_3;
};
} cr5;
union {
uint32_t value;
struct {
uint8_t cr6_all_equal;
uint8_t cr6_1;
uint8_t cr6_none_equal;
uint8_t cr6_3;
};
} cr6;
union {
uint32_t value;
struct {
uint8_t cr7_0;
uint8_t cr7_1;
uint8_t cr7_2;
uint8_t cr7_3;
};
} cr7;
union {
uint32_t value;
struct {
uint32_t rn : 2; // FP rounding control: 00 = nearest
// 01 = toward zero
// 10 = toward +infinity
// 11 = toward -infinity
uint32_t ni : 1; // Floating-point non-IEEE mode
uint32_t xe : 1; // IEEE floating-point inexact exception enable
uint32_t ze : 1; // IEEE floating-point zero divide exception enable
uint32_t ue : 1; // IEEE floating-point underflow exception enable
uint32_t oe : 1; // IEEE floating-point overflow exception enable
uint32_t ve : 1; // FP invalid op exception enable
uint32_t vxcvi : 1; // FP invalid op exception: invalid integer convert
// -- sticky
uint32_t vxsqrt : 1; // FP invalid op exception: invalid sqrt -- sticky
uint32_t vxsoft : 1; // FP invalid op exception: software request
// -- sticky
uint32_t reserved : 1;
uint32_t fprf_un : 1; // FP result unordered or NaN (FU or ?)
uint32_t fprf_eq : 1; // FP result equal or zero (FE or =)
uint32_t fprf_gt : 1; // FP result greater than or positive (FG or >)
uint32_t fprf_lt : 1; // FP result less than or negative (FL or <)
uint32_t fprf_c : 1; // FP result class
uint32_t fi : 1; // FP fraction inexact
uint32_t fr : 1; // FP fraction rounded
uint32_t vxvc : 1; // FP invalid op exception: invalid compare --
// sticky
uint32_t vximz : 1; // FP invalid op exception: infinity * 0 -- sticky
uint32_t vxzdz : 1; // FP invalid op exception: 0 / 0 -- sticky
uint32_t vxidi : 1; // FP invalid op exception: infinity / infinity
// -- sticky
uint32_t vxisi : 1; // FP invalid op exception: infinity - infinity
// -- sticky
uint32_t vxsnan : 1; // FP invalid op exception: SNaN -- sticky
uint32_t
xx : 1; // FP inexact exception -- sticky
uint32_t
zx : 1; // FP zero divide exception -- sticky
uint32_t
ux : 1; // FP underflow exception -- sticky
uint32_t
ox : 1; // FP overflow exception -- sticky
uint32_t vx : 1; // FP invalid operation exception summary
uint32_t fex : 1; // FP enabled exception summary
uint32_t
fx : 1; // FP exception summary -- sticky
} bits;
} fpscr; // Floating-point status and control register
// Most frequently used registers first.
uint64_t r[32]; // 0x20 General purpose registers
uint64_t ctr; // 0x18 Count register
uint64_t lr; // 0x10 Link register
double f[32]; // 0x120 Floating-point registers
vec128_t v[128]; // 0x220 VMX128 vector registers
vec128_t vscr_vec;
// XER register:
// Split to make it easier to do individual updates.
uint8_t xer_ca;
uint8_t xer_ov;
uint8_t xer_so;
// Condition registers:
// These are split to make it easier to do DCE on unused stores.
uint64_t cr() const;
void set_cr(uint64_t value);
uint8_t vscr_sat;
// uint32_t get_fprf() {
// return fpscr.value & 0x000F8000;
// }
// void set_fprf(const uint32_t v) {
// fpscr.value = (fpscr.value & ~0x000F8000) | v;
// }
// Thread ID assigned to this context.
uint32_t thread_id;
// Global interrupt lock, held while interrupts are disabled or interrupts are
// executing. This is shared among all threads and comes from the processor.
std::recursive_mutex* global_mutex;
// Used to shuttle data into externs. Contents volatile.
uint64_t scratch;
// Processor-specific data pointer. Used on callbacks to get access to the
// current runtime and its data.
Processor* processor;
// Shared kernel state, for easy access from kernel exports.
xe::kernel::KernelState* kernel_state;
uint8_t* physical_membase;
// Value of last reserved load
uint64_t reserved_val;
ThreadState* thread_state;
uint8_t* virtual_membase;
static std::string GetRegisterName(PPCRegister reg);
std::string GetStringFromValue(PPCRegister reg) const;
void SetValueFromString(PPCRegister reg, std::string value);
void SetRegFromString(const char* name, const char* value);
bool CompareRegWithString(const char* name, const char* value,
std::string& result) const;
} PPCContext;
#pragma pack(pop)
constexpr size_t ppcctx_size = sizeof(PPCContext);
static_assert(sizeof(PPCContext) % 64 == 0, "64b padded");
} // namespace ppc
} // namespace cpu
} // namespace xe
#endif // XENIA_CPU_PPC_PPC_CONTEXT_H_