ShuriZma
/
pcsx2
mirror of https://github.com/PCSX2/pcsx2.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

GS: Switch integer typedefs to match rest of pcsx2 

Had to capitalize the names of some struct params that had the same names
This commit is contained in:
TellowKrinkle 2021-08-30 21:45:52 -05:00 committed by tellowkrinkle
parent 2351431d71
commit b74be70ffc
86 changed files with 2419 additions and 2433 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

1
common/Pcsx2Defs.h
View File

@ -194,7 +194,6 @@ static const int __pagesize = PCSX2_PAGESIZE;
#define __fc __fastcall #define __fc __fastcall
// Makes sure that if anyone includes xbyak, it doesn't do anything bad // Makes sure that if anyone includes xbyak, it doesn't do anything bad
#define MIE_INTEGER_TYPE_DEFINED
#define XBYAK_ENABLE_OMITTED_OPERAND #define XBYAK_ENABLE_OMITTED_OPERAND
#ifdef __x86_64__ #ifdef __x86_64__

82
pcsx2/GS/GS.cpp
View File

@ -46,7 +46,7 @@ static HRESULT s_hr = E_FAIL;
#undef None #undef None
static GSRenderer* s_gs = NULL; static GSRenderer* s_gs = NULL;
static uint8* s_basemem = NULL; static u8* s_basemem = NULL;
static int s_vsync = 0; static int s_vsync = 0;
static bool s_exclusive = true; static bool s_exclusive = true;
static std::string s_renderer_name; static std::string s_renderer_name;
@ -59,7 +59,7 @@ static int s_new_gs_window_width = 0;
static int s_new_gs_window_height = 0; static int s_new_gs_window_height = 0;
#endif #endif
void GSsetBaseMem(uint8* mem) void GSsetBaseMem(u8* mem)
{ {
s_basemem = mem; s_basemem = mem;
@ -267,19 +267,19 @@ int _GSopen(const WindowInfo& wi, const char* title, GSRendererType renderer, in
return 0; return 0;
} }
void GSosdLog(const char* utf8, uint32 color) void GSosdLog(const char* utf8, u32 color)
{ {
if (s_gs && s_gs->m_dev) if (s_gs && s_gs->m_dev)
s_gs->m_dev->m_osd.Log(utf8); s_gs->m_dev->m_osd.Log(utf8);
} }
void GSosdMonitor(const char* key, const char* value, uint32 color) void GSosdMonitor(const char* key, const char* value, u32 color)
{ {
if (s_gs && s_gs->m_dev) if (s_gs && s_gs->m_dev)
s_gs->m_dev->m_osd.Monitor(key, value); s_gs->m_dev->m_osd.Monitor(key, value);
} }
int GSopen2(const WindowInfo& wi, uint32 flags) int GSopen2(const WindowInfo& wi, u32 flags)
{ {
static bool stored_toggle_state = false; static bool stored_toggle_state = false;
const bool toggle_state = !!(flags & 4); const bool toggle_state = !!(flags & 4);
@ -344,7 +344,7 @@ void GSreset()
} }
} }
void GSgifSoftReset(uint32 mask) void GSgifSoftReset(u32 mask)
{ {
try try
{ {
@ -355,7 +355,7 @@ void GSgifSoftReset(uint32 mask)
} }
} }
void GSwriteCSR(uint32 csr) void GSwriteCSR(u32 csr)
{ {
try try
{ {
@ -366,7 +366,7 @@ void GSwriteCSR(uint32 csr)
} }
} }
void GSinitReadFIFO(uint8* mem) void GSinitReadFIFO(u8* mem)
{ {
GL_PERF("Init Read FIFO1"); GL_PERF("Init Read FIFO1");
try try
@ -382,7 +382,7 @@ void GSinitReadFIFO(uint8* mem)
} }
} }
void GSreadFIFO(uint8* mem) void GSreadFIFO(u8* mem)
{ {
try try
{ {
@ -397,7 +397,7 @@ void GSreadFIFO(uint8* mem)
} }
} }
void GSinitReadFIFO2(uint8* mem, uint32 size) void GSinitReadFIFO2(u8* mem, u32 size)
{ {
GL_PERF("Init Read FIFO2"); GL_PERF("Init Read FIFO2");
try try
@ -413,7 +413,7 @@ void GSinitReadFIFO2(uint8* mem, uint32 size)
} }
} }
void GSreadFIFO2(uint8* mem, uint32 size) void GSreadFIFO2(u8* mem, u32 size)
{ {
try try
{ {
@ -428,7 +428,7 @@ void GSreadFIFO2(uint8* mem, uint32 size)
} }
} }
void GSgifTransfer(const uint8* mem, uint32 size) void GSgifTransfer(const u8* mem, u32 size)
{ {
try try
{ {
@ -439,33 +439,33 @@ void GSgifTransfer(const uint8* mem, uint32 size)
} }
} }
void GSgifTransfer1(uint8* mem, uint32 addr) void GSgifTransfer1(u8* mem, u32 addr)
{ {
try try
{ {
s_gs->Transfer<0>(const_cast<uint8*>(mem) + addr, (0x4000 - addr) / 16); s_gs->Transfer<0>(const_cast<u8*>(mem) + addr, (0x4000 - addr) / 16);
} }
catch (GSRecoverableError) catch (GSRecoverableError)
{ {
} }
} }
void GSgifTransfer2(uint8* mem, uint32 size) void GSgifTransfer2(u8* mem, u32 size)
{ {
try try
{ {
s_gs->Transfer<1>(const_cast<uint8*>(mem), size); s_gs->Transfer<1>(const_cast<u8*>(mem), size);
} }
catch (GSRecoverableError) catch (GSRecoverableError)
{ {
} }
} }
void GSgifTransfer3(uint8* mem, uint32 size) void GSgifTransfer3(u8* mem, u32 size)
{ {
try try
{ {
s_gs->Transfer<2>(const_cast<uint8*>(mem), size); s_gs->Transfer<2>(const_cast<u8*>(mem), size);
} }
catch (GSRecoverableError) catch (GSRecoverableError)
{ {
@ -487,7 +487,7 @@ void GSvsync(int field)
} }
} }
uint32 GSmakeSnapshot(char* path) u32 GSmakeSnapshot(char* path)
{ {
try try
{ {
@ -629,7 +629,7 @@ void GSendRecording()
pt(" - Capture ended\n"); pt(" - Capture ended\n");
} }
void GSsetGameCRC(uint32 crc, int options) void GSsetGameCRC(u32 crc, int options)
{ {
s_gs->SetGameCRC(crc, options); s_gs->SetGameCRC(crc, options);
} }
@ -746,7 +746,7 @@ void vmfree(void* ptr, size_t size)
} }
static HANDLE s_fh = NULL; static HANDLE s_fh = NULL;
static uint8* s_Next[8]; static u8* s_Next[8];
void* fifo_alloc(size_t size, size_t repeat) void* fifo_alloc(size_t size, size_t repeat)
{ {
@ -770,8 +770,8 @@ void* fifo_alloc(size_t size, size_t repeat)
void* fifo = MapViewOfFile(s_fh, FILE_MAP_ALL_ACCESS, 0, 0, size); void* fifo = MapViewOfFile(s_fh, FILE_MAP_ALL_ACCESS, 0, 0, size);
for (size_t i = 1; i < repeat; i++) for (size_t i = 1; i < repeat; i++)
{ {
void* base = (uint8*)fifo + size * i; void* base = (u8*)fifo + size * i;
s_Next[i] = (uint8*)MapViewOfFileEx(s_fh, FILE_MAP_ALL_ACCESS, 0, 0, size, base); s_Next[i] = (u8*)MapViewOfFileEx(s_fh, FILE_MAP_ALL_ACCESS, 0, 0, size, base);
errorID = ::GetLastError(); errorID = ::GetLastError();
if (s_Next[i] != base) if (s_Next[i] != base)
{ {
@ -881,8 +881,8 @@ void* fifo_alloc(size_t size, size_t repeat)
for (size_t i = 1; i < repeat; i++) for (size_t i = 1; i < repeat; i++)
{ {
void* base = (uint8*)fifo + size * i; void* base = (u8*)fifo + size * i;
uint8* next = (uint8*)mmap(base, size, PROT_READ | PROT_WRITE, MAP_SHARED | MAP_FIXED, s_shm_fd, 0); u8* next = (u8*)mmap(base, size, PROT_READ | PROT_WRITE, MAP_SHARED | MAP_FIXED, s_shm_fd, 0);
if (next != base) if (next != base)
fprintf(stderr, "Fail to mmap contiguous segment\n"); fprintf(stderr, "Fail to mmap contiguous segment\n");
} }
@ -1081,16 +1081,16 @@ void GSApp::Init()
m_section = "Settings"; m_section = "Settings";
#ifdef _WIN32 #ifdef _WIN32
m_gs_renderers.push_back(GSSetting(static_cast<uint32>(GSRendererType::DX1011_HW), "Direct3D 11", "")); m_gs_renderers.push_back(GSSetting(static_cast<u32>(GSRendererType::DX1011_HW), "Direct3D 11", ""));
m_gs_renderers.push_back(GSSetting(static_cast<uint32>(GSRendererType::OGL_HW), "OpenGL", "")); m_gs_renderers.push_back(GSSetting(static_cast<u32>(GSRendererType::OGL_HW), "OpenGL", ""));
m_gs_renderers.push_back(GSSetting(static_cast<uint32>(GSRendererType::OGL_SW), "Software", "")); m_gs_renderers.push_back(GSSetting(static_cast<u32>(GSRendererType::OGL_SW), "Software", ""));
#else // Linux #else // Linux
m_gs_renderers.push_back(GSSetting(static_cast<uint32>(GSRendererType::OGL_HW), "OpenGL", "")); m_gs_renderers.push_back(GSSetting(static_cast<u32>(GSRendererType::OGL_HW), "OpenGL", ""));
m_gs_renderers.push_back(GSSetting(static_cast<uint32>(GSRendererType::OGL_SW), "Software", "")); m_gs_renderers.push_back(GSSetting(static_cast<u32>(GSRendererType::OGL_SW), "Software", ""));
#endif #endif
// The null renderer goes third, it has use for benchmarking purposes in a release build // The null renderer goes third, it has use for benchmarking purposes in a release build
m_gs_renderers.push_back(GSSetting(static_cast<uint32>(GSRendererType::Null), "Null", "")); m_gs_renderers.push_back(GSSetting(static_cast<u32>(GSRendererType::Null), "Null", ""));
m_gs_interlace.push_back(GSSetting(0, "None", "")); m_gs_interlace.push_back(GSSetting(0, "None", ""));
m_gs_interlace.push_back(GSSetting(1, "Weave tff", "saw-tooth")); m_gs_interlace.push_back(GSSetting(1, "Weave tff", "saw-tooth"));
@ -1120,14 +1120,14 @@ void GSApp::Init()
m_gs_dithering.push_back(GSSetting(2, "Unscaled", "Default")); m_gs_dithering.push_back(GSSetting(2, "Unscaled", "Default"));
m_gs_dithering.push_back(GSSetting(1, "Scaled", "")); m_gs_dithering.push_back(GSSetting(1, "Scaled", ""));
m_gs_bifilter.push_back(GSSetting(static_cast<uint32>(BiFiltering::Nearest), "Nearest", "")); m_gs_bifilter.push_back(GSSetting(static_cast<u32>(BiFiltering::Nearest), "Nearest", ""));
m_gs_bifilter.push_back(GSSetting(static_cast<uint32>(BiFiltering::Forced_But_Sprite), "Bilinear", "Forced excluding sprite")); m_gs_bifilter.push_back(GSSetting(static_cast<u32>(BiFiltering::Forced_But_Sprite), "Bilinear", "Forced excluding sprite"));
m_gs_bifilter.push_back(GSSetting(static_cast<uint32>(BiFiltering::Forced), "Bilinear", "Forced")); m_gs_bifilter.push_back(GSSetting(static_cast<u32>(BiFiltering::Forced), "Bilinear", "Forced"));
m_gs_bifilter.push_back(GSSetting(static_cast<uint32>(BiFiltering::PS2), "Bilinear", "PS2")); m_gs_bifilter.push_back(GSSetting(static_cast<u32>(BiFiltering::PS2), "Bilinear", "PS2"));
m_gs_trifilter.push_back(GSSetting(static_cast<uint32>(TriFiltering::None), "None", "Default")); m_gs_trifilter.push_back(GSSetting(static_cast<u32>(TriFiltering::None), "None", "Default"));
m_gs_trifilter.push_back(GSSetting(static_cast<uint32>(TriFiltering::PS2), "Trilinear", "")); m_gs_trifilter.push_back(GSSetting(static_cast<u32>(TriFiltering::PS2), "Trilinear", ""));
m_gs_trifilter.push_back(GSSetting(static_cast<uint32>(TriFiltering::Forced), "Trilinear", "Ultra/Slow")); m_gs_trifilter.push_back(GSSetting(static_cast<u32>(TriFiltering::Forced), "Trilinear", "Ultra/Slow"));
m_gs_generic_list.push_back(GSSetting(-1, "Automatic", "Default")); m_gs_generic_list.push_back(GSSetting(-1, "Automatic", "Default"));
m_gs_generic_list.push_back(GSSetting(0, "Force-Disabled", "")); m_gs_generic_list.push_back(GSSetting(0, "Force-Disabled", ""));
@ -1202,7 +1202,7 @@ void GSApp::Init()
m_default_configuration["capture_threads"] = "4"; m_default_configuration["capture_threads"] = "4";
m_default_configuration["CaptureHeight"] = "480"; m_default_configuration["CaptureHeight"] = "480";
m_default_configuration["CaptureWidth"] = "640"; m_default_configuration["CaptureWidth"] = "640";
m_default_configuration["crc_hack_level"] = std::to_string(static_cast<int8>(CRCHackLevel::Automatic)); m_default_configuration["crc_hack_level"] = std::to_string(static_cast<s8>(CRCHackLevel::Automatic));
m_default_configuration["CrcHacksExclusions"] = ""; m_default_configuration["CrcHacksExclusions"] = "";
m_default_configuration["debug_glsl_shader"] = "0"; m_default_configuration["debug_glsl_shader"] = "0";
m_default_configuration["debug_opengl"] = "0"; m_default_configuration["debug_opengl"] = "0";
@ -1211,7 +1211,7 @@ void GSApp::Init()
m_default_configuration["dump"] = "0"; m_default_configuration["dump"] = "0";
m_default_configuration["extrathreads"] = "2"; m_default_configuration["extrathreads"] = "2";
m_default_configuration["extrathreads_height"] = "4"; m_default_configuration["extrathreads_height"] = "4";
m_default_configuration["filter"] = std::to_string(static_cast<int8>(BiFiltering::PS2)); m_default_configuration["filter"] = std::to_string(static_cast<s8>(BiFiltering::PS2));
m_default_configuration["force_texture_clear"] = "0"; m_default_configuration["force_texture_clear"] = "0";
m_default_configuration["fxaa"] = "0"; m_default_configuration["fxaa"] = "0";
m_default_configuration["interlace"] = "7"; m_default_configuration["interlace"] = "7";
@ -1288,7 +1288,7 @@ void GSApp::Init()
m_default_configuration["UserHacks_TCOffsetX"] = "0"; m_default_configuration["UserHacks_TCOffsetX"] = "0";
m_default_configuration["UserHacks_TCOffsetY"] = "0"; m_default_configuration["UserHacks_TCOffsetY"] = "0";
m_default_configuration["UserHacks_TextureInsideRt"] = "0"; m_default_configuration["UserHacks_TextureInsideRt"] = "0";
m_default_configuration["UserHacks_TriFilter"] = std::to_string(static_cast<int8>(TriFiltering::None)); m_default_configuration["UserHacks_TriFilter"] = std::to_string(static_cast<s8>(TriFiltering::None));
m_default_configuration["UserHacks_WildHack"] = "0"; m_default_configuration["UserHacks_WildHack"] = "0";
m_default_configuration["wrap_gs_mem"] = "0"; m_default_configuration["wrap_gs_mem"] = "0";
m_default_configuration["vsync"] = "0"; m_default_configuration["vsync"] = "0";

1042
pcsx2/GS/GS.h
View File

File diff suppressed because it is too large Load Diff

158
pcsx2/GS/GSBlock.h
View File

@ -47,11 +47,11 @@ class GSBlock
static const GSVector4i m_uw8hmask3; static const GSVector4i m_uw8hmask3;
public: public:
template <int i, int alignment, uint32 mask> template <int i, int alignment, u32 mask>
__forceinline static void WriteColumn32(uint8* RESTRICT dst, const uint8* RESTRICT src, int srcpitch) __forceinline static void WriteColumn32(u8* RESTRICT dst, const u8* RESTRICT src, int srcpitch)
{ {
const uint8* RESTRICT s0 = &src[srcpitch * 0]; const u8* RESTRICT s0 = &src[srcpitch * 0];
const uint8* RESTRICT s1 = &src[srcpitch * 1]; const u8* RESTRICT s1 = &src[srcpitch * 1];
#if _M_SSE >= 0x501 #if _M_SSE >= 0x501
@ -173,12 +173,12 @@ public:
} }
template <int i, int alignment> template <int i, int alignment>
__forceinline static void WriteColumn16(uint8* RESTRICT dst, const uint8* RESTRICT src, int srcpitch) __forceinline static void WriteColumn16(u8* RESTRICT dst, const u8* RESTRICT src, int srcpitch)
{ {
const uint8* RESTRICT s0 = &src[srcpitch * 0]; const u8* RESTRICT s0 = &src[srcpitch * 0];
const uint8* RESTRICT s1 = &src[srcpitch * 1]; const u8* RESTRICT s1 = &src[srcpitch * 1];
// for(int j = 0; j < 16; j++) {((uint16*)s0)[j] = columnTable16[0][j]; ((uint16*)s1)[j] = columnTable16[1][j];} // for(int j = 0; j < 16; j++) {((u16*)s0)[j] = columnTable16[0][j]; ((u16*)s1)[j] = columnTable16[1][j];}
#if _M_SSE >= 0x501 #if _M_SSE >= 0x501
@ -247,7 +247,7 @@ public:
} }
template <int i, int alignment> template <int i, int alignment>
__forceinline static void WriteColumn8(uint8* RESTRICT dst, const uint8* RESTRICT src, int srcpitch) __forceinline static void WriteColumn8(u8* RESTRICT dst, const u8* RESTRICT src, int srcpitch)
{ {
// TODO: read unaligned as WriteColumn32 does and try saving a few shuffles // TODO: read unaligned as WriteColumn32 does and try saving a few shuffles
@ -310,7 +310,7 @@ public:
} }
template <int i, int alignment> template <int i, int alignment>
__forceinline static void WriteColumn4(uint8* RESTRICT dst, const uint8* RESTRICT src, int srcpitch) __forceinline static void WriteColumn4(u8* RESTRICT dst, const u8* RESTRICT src, int srcpitch)
{ {
//printf("WriteColumn4\n"); //printf("WriteColumn4\n");
@ -345,8 +345,8 @@ public:
((GSVector4i*)dst)[i * 4 + 3] = v3; ((GSVector4i*)dst)[i * 4 + 3] = v3;
} }
template <int alignment, uint32 mask> template <int alignment, u32 mask>
static void WriteColumn32(int y, uint8* RESTRICT dst, const uint8* RESTRICT src, int srcpitch) static void WriteColumn32(int y, u8* RESTRICT dst, const u8* RESTRICT src, int srcpitch)
{ {
switch ((y >> 1) & 3) switch ((y >> 1) & 3)
{ {
@ -359,7 +359,7 @@ public:
} }
template <int alignment> template <int alignment>
static void WriteColumn16(int y, uint8* RESTRICT dst, const uint8* RESTRICT src, int srcpitch) static void WriteColumn16(int y, u8* RESTRICT dst, const u8* RESTRICT src, int srcpitch)
{ {
switch ((y >> 1) & 3) switch ((y >> 1) & 3)
{ {
@ -372,7 +372,7 @@ public:
} }
template <int alignment> template <int alignment>
static void WriteColumn8(int y, uint8* RESTRICT dst, const uint8* RESTRICT src, int srcpitch) static void WriteColumn8(int y, u8* RESTRICT dst, const u8* RESTRICT src, int srcpitch)
{ {
switch ((y >> 2) & 3) switch ((y >> 2) & 3)
{ {
@ -385,7 +385,7 @@ public:
} }
template <int alignment> template <int alignment>
static void WriteColumn4(int y, uint8* RESTRICT dst, const uint8* RESTRICT src, int srcpitch) static void WriteColumn4(int y, u8* RESTRICT dst, const u8* RESTRICT src, int srcpitch)
{ {
switch ((y >> 2) & 3) switch ((y >> 2) & 3)
{ {
@ -397,8 +397,8 @@ public:
} }
} }
template <int alignment, uint32 mask> template <int alignment, u32 mask>
static void WriteBlock32(uint8* RESTRICT dst, const uint8* RESTRICT src, int srcpitch) static void WriteBlock32(u8* RESTRICT dst, const u8* RESTRICT src, int srcpitch)
{ {
WriteColumn32<0, alignment, mask>(dst, src, srcpitch); WriteColumn32<0, alignment, mask>(dst, src, srcpitch);
src += srcpitch * 2; src += srcpitch * 2;
@ -410,7 +410,7 @@ public:
} }
template <int alignment> template <int alignment>
static void WriteBlock16(uint8* RESTRICT dst, const uint8* RESTRICT src, int srcpitch) static void WriteBlock16(u8* RESTRICT dst, const u8* RESTRICT src, int srcpitch)
{ {
WriteColumn16<0, alignment>(dst, src, srcpitch); WriteColumn16<0, alignment>(dst, src, srcpitch);
src += srcpitch * 2; src += srcpitch * 2;
@ -422,7 +422,7 @@ public:
} }
template <int alignment> template <int alignment>
static void WriteBlock8(uint8* RESTRICT dst, const uint8* RESTRICT src, int srcpitch) static void WriteBlock8(u8* RESTRICT dst, const u8* RESTRICT src, int srcpitch)
{ {
WriteColumn8<0, alignment>(dst, src, srcpitch); WriteColumn8<0, alignment>(dst, src, srcpitch);
src += srcpitch * 4; src += srcpitch * 4;
@ -434,7 +434,7 @@ public:
} }
template <int alignment> template <int alignment>
static void WriteBlock4(uint8* RESTRICT dst, const uint8* RESTRICT src, int srcpitch) static void WriteBlock4(u8* RESTRICT dst, const u8* RESTRICT src, int srcpitch)
{ {
WriteColumn4<0, alignment>(dst, src, srcpitch); WriteColumn4<0, alignment>(dst, src, srcpitch);
src += srcpitch * 4; src += srcpitch * 4;
@ -446,7 +446,7 @@ public:
} }
template <int i> template <int i>
__forceinline static void ReadColumn32(const uint8* RESTRICT src, uint8* RESTRICT dst, int dstpitch) __forceinline static void ReadColumn32(const u8* RESTRICT src, u8* RESTRICT dst, int dstpitch)
{ {
#if _M_SSE >= 0x501 #if _M_SSE >= 0x501
@ -484,7 +484,7 @@ public:
} }
template <int i> template <int i>
__forceinline static void ReadColumn16(const uint8* RESTRICT src, uint8* RESTRICT dst, int dstpitch) __forceinline static void ReadColumn16(const u8* RESTRICT src, u8* RESTRICT dst, int dstpitch)
{ {
#if _M_SSE >= 0x501 #if _M_SSE >= 0x501
@ -526,10 +526,10 @@ public:
} }
template <int i> template <int i>
__forceinline static void ReadColumn8(const uint8* RESTRICT src, uint8* RESTRICT dst, int dstpitch) __forceinline static void ReadColumn8(const u8* RESTRICT src, u8* RESTRICT dst, int dstpitch)
{ {
//for(int j = 0; j < 64; j++) ((uint8*)src)[j] = (uint8)j; //for(int j = 0; j < 64; j++) ((u8*)src)[j] = (u8)j;
#if 0 //_M_SSE >= 0x501 #if 0 //_M_SSE >= 0x501
@ -593,7 +593,7 @@ public:
} }
template <int i> template <int i>
__forceinline static void ReadColumn4(const uint8* RESTRICT src, uint8* RESTRICT dst, int dstpitch) __forceinline static void ReadColumn4(const u8* RESTRICT src, u8* RESTRICT dst, int dstpitch)
{ {
//printf("ReadColumn4\n"); //printf("ReadColumn4\n");
@ -628,7 +628,7 @@ public:
GSVector4i::store<true>(&dst[dstpitch * 3], v3); GSVector4i::store<true>(&dst[dstpitch * 3], v3);
} }
static void ReadColumn32(int y, const uint8* RESTRICT src, uint8* RESTRICT dst, int dstpitch) static void ReadColumn32(int y, const u8* RESTRICT src, u8* RESTRICT dst, int dstpitch)
{ {
switch ((y >> 1) & 3) switch ((y >> 1) & 3)
{ {
@ -640,7 +640,7 @@ public:
} }
} }
static void ReadColumn16(int y, const uint8* RESTRICT src, uint8* RESTRICT dst, int dstpitch) static void ReadColumn16(int y, const u8* RESTRICT src, u8* RESTRICT dst, int dstpitch)
{ {
switch ((y >> 1) & 3) switch ((y >> 1) & 3)
{ {
@ -652,7 +652,7 @@ public:
} }
} }
static void ReadColumn8(int y, const uint8* RESTRICT src, uint8* RESTRICT dst, int dstpitch) static void ReadColumn8(int y, const u8* RESTRICT src, u8* RESTRICT dst, int dstpitch)
{ {
switch ((y >> 2) & 3) switch ((y >> 2) & 3)
{ {
@ -664,7 +664,7 @@ public:
} }
} }
static void ReadColumn4(int y, const uint8* RESTRICT src, uint8* RESTRICT dst, int dstpitch) static void ReadColumn4(int y, const u8* RESTRICT src, u8* RESTRICT dst, int dstpitch)
{ {
switch ((y >> 2) & 3) switch ((y >> 2) & 3)
{ {
@ -676,7 +676,7 @@ public:
} }
} }
static void ReadBlock32(const uint8* RESTRICT src, uint8* RESTRICT dst, int dstpitch) static void ReadBlock32(const u8* RESTRICT src, u8* RESTRICT dst, int dstpitch)
{ {
ReadColumn32<0>(src, dst, dstpitch); ReadColumn32<0>(src, dst, dstpitch);
dst += dstpitch * 2; dst += dstpitch * 2;
@ -687,7 +687,7 @@ public:
ReadColumn32<3>(src, dst, dstpitch); ReadColumn32<3>(src, dst, dstpitch);
} }
static void ReadBlock16(const uint8* RESTRICT src, uint8* RESTRICT dst, int dstpitch) static void ReadBlock16(const u8* RESTRICT src, u8* RESTRICT dst, int dstpitch)
{ {
ReadColumn16<0>(src, dst, dstpitch); ReadColumn16<0>(src, dst, dstpitch);
dst += dstpitch * 2; dst += dstpitch * 2;
@ -698,7 +698,7 @@ public:
ReadColumn16<3>(src, dst, dstpitch); ReadColumn16<3>(src, dst, dstpitch);
} }
static void ReadBlock8(const uint8* RESTRICT src, uint8* RESTRICT dst, int dstpitch) static void ReadBlock8(const u8* RESTRICT src, u8* RESTRICT dst, int dstpitch)
{ {
ReadColumn8<0>(src, dst, dstpitch); ReadColumn8<0>(src, dst, dstpitch);
dst += dstpitch * 4; dst += dstpitch * 4;
@ -709,7 +709,7 @@ public:
ReadColumn8<3>(src, dst, dstpitch); ReadColumn8<3>(src, dst, dstpitch);
} }
static void ReadBlock4(const uint8* RESTRICT src, uint8* RESTRICT dst, int dstpitch) static void ReadBlock4(const u8* RESTRICT src, u8* RESTRICT dst, int dstpitch)
{ {
ReadColumn4<0>(src, dst, dstpitch); ReadColumn4<0>(src, dst, dstpitch);
dst += dstpitch * 4; dst += dstpitch * 4;
@ -720,7 +720,7 @@ public:
ReadColumn4<3>(src, dst, dstpitch); ReadColumn4<3>(src, dst, dstpitch);
} }
__forceinline static void ReadBlock4P(const uint8* RESTRICT src, uint8* RESTRICT dst, int dstpitch) __forceinline static void ReadBlock4P(const u8* RESTRICT src, u8* RESTRICT dst, int dstpitch)
{ {
//printf("ReadBlock4P\n"); //printf("ReadBlock4P\n");
@ -784,12 +784,12 @@ public:
} }
} }
__forceinline static void ReadBlock8HP(const uint8* RESTRICT src, uint8* RESTRICT dst, int dstpitch) __forceinline static void ReadBlock8HP(const u8* RESTRICT src, u8* RESTRICT dst, int dstpitch)
{ {
#if _M_SSE >= 0x501 #if _M_SSE >= 0x501
uint8* RESTRICT d0 = &dst[dstpitch * 0]; u8* RESTRICT d0 = &dst[dstpitch * 0];
uint8* RESTRICT d1 = &dst[dstpitch * 4]; u8* RESTRICT d1 = &dst[dstpitch * 4];
const GSVector8i* s = (const GSVector8i*)src; const GSVector8i* s = (const GSVector8i*)src;
@ -855,12 +855,12 @@ public:
#endif #endif
} }
__forceinline static void ReadBlock4HLP(const uint8* RESTRICT src, uint8* RESTRICT dst, int dstpitch) __forceinline static void ReadBlock4HLP(const u8* RESTRICT src, u8* RESTRICT dst, int dstpitch)
{ {
#if _M_SSE >= 0x501 #if _M_SSE >= 0x501
uint8* RESTRICT d0 = &dst[dstpitch * 0]; u8* RESTRICT d0 = &dst[dstpitch * 0];
uint8* RESTRICT d1 = &dst[dstpitch * 4]; u8* RESTRICT d1 = &dst[dstpitch * 4];
const GSVector8i* s = (const GSVector8i*)src; const GSVector8i* s = (const GSVector8i*)src;
@ -929,12 +929,12 @@ public:
#endif #endif
} }
__forceinline static void ReadBlock4HHP(const uint8* RESTRICT src, uint8* RESTRICT dst, int dstpitch) __forceinline static void ReadBlock4HHP(const u8* RESTRICT src, u8* RESTRICT dst, int dstpitch)
{ {
#if _M_SSE >= 0x501 #if _M_SSE >= 0x501
uint8* RESTRICT d0 = &dst[dstpitch * 0]; u8* RESTRICT d0 = &dst[dstpitch * 0];
uint8* RESTRICT d1 = &dst[dstpitch * 4]; u8* RESTRICT d1 = &dst[dstpitch * 4];
const GSVector8i* s = (const GSVector8i*)src; const GSVector8i* s = (const GSVector8i*)src;
@ -1013,7 +1013,7 @@ public:
} }
template <bool AEM> template <bool AEM>
static void ExpandBlock24(const uint32* RESTRICT src, uint8* RESTRICT dst, int dstpitch, const GIFRegTEXA& TEXA) static void ExpandBlock24(const u32* RESTRICT src, u8* RESTRICT dst, int dstpitch, const GIFRegTEXA& TEXA)
{ {
#if _M_SSE >= 0x501 #if _M_SSE >= 0x501
@ -1061,7 +1061,7 @@ public:
} }
template <bool AEM> template <bool AEM>
static void ExpandBlock16(const uint16* RESTRICT src, uint8* RESTRICT dst, int dstpitch, const GIFRegTEXA& TEXA) // do not inline, uses too many xmm regs static void ExpandBlock16(const u16* RESTRICT src, u8* RESTRICT dst, int dstpitch, const GIFRegTEXA& TEXA) // do not inline, uses too many xmm regs
{ {
#if _M_SSE >= 0x501 #if _M_SSE >= 0x501
@ -1101,7 +1101,7 @@ public:
#endif #endif
} }
__forceinline static void ExpandBlock8_32(const uint8* RESTRICT src, uint8* RESTRICT dst, int dstpitch, const uint32* RESTRICT pal) __forceinline static void ExpandBlock8_32(const u8* RESTRICT src, u8* RESTRICT dst, int dstpitch, const u32* RESTRICT pal)
{ {
for (int j = 0; j < 16; j++, dst += dstpitch) for (int j = 0; j < 16; j++, dst += dstpitch)
{ {
@ -1109,7 +1109,7 @@ public:
} }
} }
__forceinline static void ExpandBlock8_16(const uint8* RESTRICT src, uint8* RESTRICT dst, int dstpitch, const uint32* RESTRICT pal) __forceinline static void ExpandBlock8_16(const u8* RESTRICT src, u8* RESTRICT dst, int dstpitch, const u32* RESTRICT pal)
{ {
for (int j = 0; j < 16; j++, dst += dstpitch) for (int j = 0; j < 16; j++, dst += dstpitch)
{ {
@ -1117,7 +1117,7 @@ public:
} }
} }
__forceinline static void ExpandBlock4_32(const uint8* RESTRICT src, uint8* RESTRICT dst, int dstpitch, const uint64* RESTRICT pal) __forceinline static void ExpandBlock4_32(const u8* RESTRICT src, u8* RESTRICT dst, int dstpitch, const u64* RESTRICT pal)
{ {
for (int j = 0; j < 16; j++, dst += dstpitch) for (int j = 0; j < 16; j++, dst += dstpitch)
{ {
@ -1125,7 +1125,7 @@ public:
} }
} }
__forceinline static void ExpandBlock4_16(const uint8* RESTRICT src, uint8* RESTRICT dst, int dstpitch, const uint64* RESTRICT pal) __forceinline static void ExpandBlock4_16(const u8* RESTRICT src, u8* RESTRICT dst, int dstpitch, const u64* RESTRICT pal)
{ {
for (int j = 0; j < 16; j++, dst += dstpitch) for (int j = 0; j < 16; j++, dst += dstpitch)
{ {
@ -1133,7 +1133,7 @@ public:
} }
} }
__forceinline static void ExpandBlock8H_32(uint32* RESTRICT src, uint8* RESTRICT dst, int dstpitch, const uint32* RESTRICT pal) __forceinline static void ExpandBlock8H_32(u32* RESTRICT src, u8* RESTRICT dst, int dstpitch, const u32* RESTRICT pal)
{ {
for (int j = 0; j < 8; j++, dst += dstpitch) for (int j = 0; j < 8; j++, dst += dstpitch)
{ {
@ -1144,7 +1144,7 @@ public:
} }
} }
__forceinline static void ExpandBlock8H_16(uint32* RESTRICT src, uint8* RESTRICT dst, int dstpitch, const uint32* RESTRICT pal) __forceinline static void ExpandBlock8H_16(u32* RESTRICT src, u8* RESTRICT dst, int dstpitch, const u32* RESTRICT pal)
{ {
for (int j = 0; j < 8; j++, dst += dstpitch) for (int j = 0; j < 8; j++, dst += dstpitch)
{ {
@ -1158,7 +1158,7 @@ public:
} }
} }
__forceinline static void ExpandBlock4HL_32(uint32* RESTRICT src, uint8* RESTRICT dst, int dstpitch, const uint32* RESTRICT pal) __forceinline static void ExpandBlock4HL_32(u32* RESTRICT src, u8* RESTRICT dst, int dstpitch, const u32* RESTRICT pal)
{ {
for (int j = 0; j < 8; j++, dst += dstpitch) for (int j = 0; j < 8; j++, dst += dstpitch)
{ {
@ -1169,7 +1169,7 @@ public:
} }
} }
__forceinline static void ExpandBlock4HL_16(uint32* RESTRICT src, uint8* RESTRICT dst, int dstpitch, const uint32* RESTRICT pal) __forceinline static void ExpandBlock4HL_16(u32* RESTRICT src, u8* RESTRICT dst, int dstpitch, const u32* RESTRICT pal)
{ {
for (int j = 0; j < 8; j++, dst += dstpitch) for (int j = 0; j < 8; j++, dst += dstpitch)
{ {
@ -1182,7 +1182,7 @@ public:
} }
} }
__forceinline static void ExpandBlock4HH_32(uint32* RESTRICT src, uint8* RESTRICT dst, int dstpitch, const uint32* RESTRICT pal) __forceinline static void ExpandBlock4HH_32(u32* RESTRICT src, u8* RESTRICT dst, int dstpitch, const u32* RESTRICT pal)
{ {
for (int j = 0; j < 8; j++, dst += dstpitch) for (int j = 0; j < 8; j++, dst += dstpitch)
{ {
@ -1193,7 +1193,7 @@ public:
} }
} }
__forceinline static void ExpandBlock4HH_16(uint32* RESTRICT src, uint8* RESTRICT dst, int dstpitch, const uint32* RESTRICT pal) __forceinline static void ExpandBlock4HH_16(u32* RESTRICT src, u8* RESTRICT dst, int dstpitch, const u32* RESTRICT pal)
{ {
for (int j = 0; j < 8; j++, dst += dstpitch) for (int j = 0; j < 8; j++, dst += dstpitch)
{ {
@ -1206,14 +1206,14 @@ public:
} }
} }
__forceinline static void UnpackAndWriteBlock24(const uint8* RESTRICT src, int srcpitch, uint8* RESTRICT dst) __forceinline static void UnpackAndWriteBlock24(const u8* RESTRICT src, int srcpitch, u8* RESTRICT dst)
{ {
#if _M_SSE >= 0x501 #if _M_SSE >= 0x501
const uint8* RESTRICT s0 = &src[srcpitch * 0]; const u8* RESTRICT s0 = &src[srcpitch * 0];
const uint8* RESTRICT s1 = &src[srcpitch * 1]; const u8* RESTRICT s1 = &src[srcpitch * 1];
const uint8* RESTRICT s2 = &src[srcpitch * 2]; const u8* RESTRICT s2 = &src[srcpitch * 2];
const uint8* RESTRICT s3 = &src[srcpitch * 3]; const u8* RESTRICT s3 = &src[srcpitch * 3];
GSVector8i v0, v1, v2, v3, v4, v5, v6; GSVector8i v0, v1, v2, v3, v4, v5, v6;
GSVector8i mask = GSVector8i::x00ffffff(); GSVector8i mask = GSVector8i::x00ffffff();
@ -1293,7 +1293,7 @@ public:
#endif #endif
} }
__forceinline static void UnpackAndWriteBlock8H(const uint8* RESTRICT src, int srcpitch, uint8* RESTRICT dst) __forceinline static void UnpackAndWriteBlock8H(const u8* RESTRICT src, int srcpitch, u8* RESTRICT dst)
{ {
#if _M_SSE >= 0x501 #if _M_SSE >= 0x501
@ -1367,13 +1367,13 @@ public:
#endif #endif
} }
__forceinline static void UnpackAndWriteBlock4HL(const uint8* RESTRICT src, int srcpitch, uint8* RESTRICT dst) __forceinline static void UnpackAndWriteBlock4HL(const u8* RESTRICT src, int srcpitch, u8* RESTRICT dst)
{ {
//printf("4HL\n"); //printf("4HL\n");
if (0) if (0)
{ {
uint8* s = (uint8*)src; u8* s = (u8*)src;
for (int j = 0; j < 8; j++, s += srcpitch) for (int j = 0; j < 8; j++, s += srcpitch)
for (int i = 0; i < 4; i++) for (int i = 0; i < 4; i++)
s[i] = (columnTable32[j][i * 2] & 0x0f) | (columnTable32[j][i * 2 + 1] << 4); s[i] = (columnTable32[j][i * 2] & 0x0f) | (columnTable32[j][i * 2 + 1] << 4);
@ -1385,7 +1385,7 @@ public:
GSVector8i v0, v1, v2, v3; GSVector8i v0, v1, v2, v3;
GSVector8i mask(0x0f000000); GSVector8i mask(0x0f000000);
v6 = GSVector4i(*(uint32*)&src[srcpitch * 0], *(uint32*)&src[srcpitch * 2], *(uint32*)&src[srcpitch * 1], *(uint32*)&src[srcpitch * 3]); v6 = GSVector4i(*(u32*)&src[srcpitch * 0], *(u32*)&src[srcpitch * 2], *(u32*)&src[srcpitch * 1], *(u32*)&src[srcpitch * 3]);
v4 = v6.upl8(v6 >> 4); v4 = v6.upl8(v6 >> 4);
v5 = v6.uph8(v6 >> 4); v5 = v6.uph8(v6 >> 4);
@ -1405,7 +1405,7 @@ public:
src += srcpitch * 4; src += srcpitch * 4;
v6 = GSVector4i(*(uint32*)&src[srcpitch * 0], *(uint32*)&src[srcpitch * 2], *(uint32*)&src[srcpitch * 1], *(uint32*)&src[srcpitch * 3]); v6 = GSVector4i(*(u32*)&src[srcpitch * 0], *(u32*)&src[srcpitch * 2], *(u32*)&src[srcpitch * 1], *(u32*)&src[srcpitch * 3]);
v4 = v6.upl8(v6 >> 4); v4 = v6.upl8(v6 >> 4);
v5 = v6.uph8(v6 >> 4); v5 = v6.uph8(v6 >> 4);
@ -1434,7 +1434,7 @@ public:
for (int i = 0; i < 2; i++, src += srcpitch * 4) for (int i = 0; i < 2; i++, src += srcpitch * 4)
{ {
GSVector4i v(*(uint32*)&src[srcpitch * 0], *(uint32*)&src[srcpitch * 1], *(uint32*)&src[srcpitch * 2], *(uint32*)&src[srcpitch * 3]); GSVector4i v(*(u32*)&src[srcpitch * 0], *(u32*)&src[srcpitch * 1], *(u32*)&src[srcpitch * 2], *(u32*)&src[srcpitch * 3]);
v4 = v.upl8(v >> 4); v4 = v.upl8(v >> 4);
v5 = v.uph8(v >> 4); v5 = v.uph8(v >> 4);
@ -1463,7 +1463,7 @@ public:
#endif #endif
} }
__forceinline static void UnpackAndWriteBlock4HH(const uint8* RESTRICT src, int srcpitch, uint8* RESTRICT dst) __forceinline static void UnpackAndWriteBlock4HH(const u8* RESTRICT src, int srcpitch, u8* RESTRICT dst)
{ {
#if _M_SSE >= 0x501 #if _M_SSE >= 0x501
@ -1472,7 +1472,7 @@ public:
GSVector8i v0, v1, v2, v3; GSVector8i v0, v1, v2, v3;
GSVector8i mask = GSVector8i::xf0000000(); GSVector8i mask = GSVector8i::xf0000000();
v6 = GSVector4i(*(uint32*)&src[srcpitch * 0], *(uint32*)&src[srcpitch * 2], *(uint32*)&src[srcpitch * 1], *(uint32*)&src[srcpitch * 3]); v6 = GSVector4i(*(u32*)&src[srcpitch * 0], *(u32*)&src[srcpitch * 2], *(u32*)&src[srcpitch * 1], *(u32*)&src[srcpitch * 3]);
v4 = (v6 << 4).upl8(v6); v4 = (v6 << 4).upl8(v6);
v5 = (v6 << 4).uph8(v6); v5 = (v6 << 4).uph8(v6);
@ -1492,7 +1492,7 @@ public:
src += srcpitch * 4; src += srcpitch * 4;
v6 = GSVector4i(*(uint32*)&src[srcpitch * 0], *(uint32*)&src[srcpitch * 2], *(uint32*)&src[srcpitch * 1], *(uint32*)&src[srcpitch * 3]); v6 = GSVector4i(*(u32*)&src[srcpitch * 0], *(u32*)&src[srcpitch * 2], *(u32*)&src[srcpitch * 1], *(u32*)&src[srcpitch * 3]);
v4 = (v6 << 4).upl8(v6); v4 = (v6 << 4).upl8(v6);
v5 = (v6 << 4).uph8(v6); v5 = (v6 << 4).uph8(v6);
@ -1521,7 +1521,7 @@ public:
for (int i = 0; i < 2; i++, src += srcpitch * 4) for (int i = 0; i < 2; i++, src += srcpitch * 4)
{ {
GSVector4i v(*(uint32*)&src[srcpitch * 0], *(uint32*)&src[srcpitch * 1], *(uint32*)&src[srcpitch * 2], *(uint32*)&src[srcpitch * 3]); GSVector4i v(*(u32*)&src[srcpitch * 0], *(u32*)&src[srcpitch * 1], *(u32*)&src[srcpitch * 2], *(u32*)&src[srcpitch * 3]);
v4 = (v << 4).upl8(v); v4 = (v << 4).upl8(v);
v5 = (v << 4).uph8(v); v5 = (v << 4).uph8(v);
@ -1551,7 +1551,7 @@ public:
} }
template <bool AEM> template <bool AEM>
__forceinline static void ReadAndExpandBlock24(const uint8* RESTRICT src, uint8* RESTRICT dst, int dstpitch, const GIFRegTEXA& TEXA) __forceinline static void ReadAndExpandBlock24(const u8* RESTRICT src, u8* RESTRICT dst, int dstpitch, const GIFRegTEXA& TEXA)
{ {
#if _M_SSE >= 0x501 #if _M_SSE >= 0x501
@ -1628,7 +1628,7 @@ public:
} }
template <bool AEM> template <bool AEM>
__forceinline static void ReadAndExpandBlock16(const uint8* RESTRICT src, uint8* RESTRICT dst, int dstpitch, const GIFRegTEXA& TEXA) __forceinline static void ReadAndExpandBlock16(const u8* RESTRICT src, u8* RESTRICT dst, int dstpitch, const GIFRegTEXA& TEXA)
{ {
#if _M_SSE >= 0x501 #if _M_SSE >= 0x501
@ -1656,16 +1656,16 @@ public:
#else #else
alignas(32) uint16 block[16 * 8]; alignas(32) u16 block[16 * 8];
ReadBlock16(src, (uint8*)block, sizeof(block) / 8); ReadBlock16(src, (u8*)block, sizeof(block) / 8);
ExpandBlock16<AEM>(block, dst, dstpitch, TEXA); ExpandBlock16<AEM>(block, dst, dstpitch, TEXA);
#endif #endif
} }
__forceinline static void ReadAndExpandBlock8_32(const uint8* RESTRICT src, uint8* RESTRICT dst, int dstpitch, const uint32* RESTRICT pal) __forceinline static void ReadAndExpandBlock8_32(const u8* RESTRICT src, u8* RESTRICT dst, int dstpitch, const u32* RESTRICT pal)
{ {
//printf("ReadAndExpandBlock8_32\n"); //printf("ReadAndExpandBlock8_32\n");
@ -1714,7 +1714,7 @@ public:
// TODO: ReadAndExpandBlock8_16 // TODO: ReadAndExpandBlock8_16
__forceinline static void ReadAndExpandBlock4_32(const uint8* RESTRICT src, uint8* RESTRICT dst, int dstpitch, const uint64* RESTRICT pal) __forceinline static void ReadAndExpandBlock4_32(const u8* RESTRICT src, u8* RESTRICT dst, int dstpitch, const u64* RESTRICT pal)
{ {
//printf("ReadAndExpandBlock4_32\n"); //printf("ReadAndExpandBlock4_32\n");
@ -1779,7 +1779,7 @@ public:
// TODO: ReadAndExpandBlock4_16 // TODO: ReadAndExpandBlock4_16
__forceinline static void ReadAndExpandBlock8H_32(const uint8* RESTRICT src, uint8* RESTRICT dst, int dstpitch, const uint32* RESTRICT pal) __forceinline static void ReadAndExpandBlock8H_32(const u8* RESTRICT src, u8* RESTRICT dst, int dstpitch, const u32* RESTRICT pal)
{ {
//printf("ReadAndExpandBlock8H_32\n"); //printf("ReadAndExpandBlock8H_32\n");
@ -1810,7 +1810,7 @@ public:
// TODO: ReadAndExpandBlock8H_16 // TODO: ReadAndExpandBlock8H_16
__forceinline static void ReadAndExpandBlock4HL_32(const uint8* RESTRICT src, uint8* RESTRICT dst, int dstpitch, const uint32* RESTRICT pal) __forceinline static void ReadAndExpandBlock4HL_32(const u8* RESTRICT src, u8* RESTRICT dst, int dstpitch, const u32* RESTRICT pal)
{ {
//printf("ReadAndExpandBlock4HL_32\n"); //printf("ReadAndExpandBlock4HL_32\n");
const GSVector4i* s = (const GSVector4i*)src; const GSVector4i* s = (const GSVector4i*)src;
@ -1840,7 +1840,7 @@ public:
// TODO: ReadAndExpandBlock4HL_16 // TODO: ReadAndExpandBlock4HL_16
__forceinline static void ReadAndExpandBlock4HH_32(const uint8* RESTRICT src, uint8* RESTRICT dst, int dstpitch, const uint32* RESTRICT pal) __forceinline static void ReadAndExpandBlock4HH_32(const u8* RESTRICT src, u8* RESTRICT dst, int dstpitch, const u32* RESTRICT pal)
{ {
//printf("ReadAndExpandBlock4HH_32\n"); //printf("ReadAndExpandBlock4HH_32\n");

18
pcsx2/GS/GSCapture.cpp
View File

@ -126,7 +126,7 @@ GSSource : public CBaseFilter, private CCritSec, public IGSSource
vih.bmiHeader.biPlanes = 1; vih.bmiHeader.biPlanes = 1;
vih.bmiHeader.biBitCount = 16; vih.bmiHeader.biBitCount = 16;
vih.bmiHeader.biSizeImage = m_size.x * m_size.y * 2; vih.bmiHeader.biSizeImage = m_size.x * m_size.y * 2;
mt.SetFormat((uint8*)&vih, sizeof(vih)); mt.SetFormat((u8*)&vih, sizeof(vih));
m_mts.push_back(mt); m_mts.push_back(mt);
@ -139,7 +139,7 @@ GSSource : public CBaseFilter, private CCritSec, public IGSSource
vih.bmiHeader.biPlanes = 1; vih.bmiHeader.biPlanes = 1;
vih.bmiHeader.biBitCount = 32; vih.bmiHeader.biBitCount = 32;
vih.bmiHeader.biSizeImage = m_size.x * m_size.y * 4; vih.bmiHeader.biSizeImage = m_size.x * m_size.y * 4;
mt.SetFormat((uint8*)&vih, sizeof(vih)); mt.SetFormat((u8*)&vih, sizeof(vih));
if (colorspace == 1) if (colorspace == 1)
m_mts.insert(m_mts.begin(), mt); m_mts.insert(m_mts.begin(), mt);
@ -272,8 +272,8 @@ public:
const CMediaType& mt = m_output->CurrentMediaType(); const CMediaType& mt = m_output->CurrentMediaType();
uint8* src = (uint8*)bits; u8* src = (u8*)bits;
uint8* dst = NULL; u8* dst = NULL;
sample->GetPointer(&dst); sample->GetPointer(&dst);
@ -300,8 +300,8 @@ public:
for (int j = 0; j < h; j++, dst += dstpitch, src += srcpitch) for (int j = 0; j < h; j++, dst += dstpitch, src += srcpitch)
{ {
uint32* s = (uint32*)src; u32* s = (u32*)src;
uint16* d = (uint16*)dst; u16* d = (u16*)dst;
for (int i = 0; i < w; i += 2) for (int i = 0; i < w; i += 2)
{ {
@ -314,7 +314,7 @@ public:
GSVector4 c = lo.hadd(hi) + offset; GSVector4 c = lo.hadd(hi) + offset;
*((uint32*)&d[i]) = GSVector4i(c).rgba32(); *((u32*)&d[i]) = GSVector4i(c).rgba32();
} }
} }
} }
@ -568,8 +568,8 @@ bool GSCapture::DeliverFrame(const void* bits, int pitch, bool rgba)
#elif defined(__unix__) #elif defined(__unix__)
std::string out_file = m_out_dir + format("/frame.%010d.png", m_frame); std::string out_file = m_out_dir + format("/frame.%010d.png", m_frame);
//GSPng::Save(GSPng::RGB_PNG, out_file, (uint8*)bits, m_size.x, m_size.y, pitch, m_compression_level); //GSPng::Save(GSPng::RGB_PNG, out_file, (u8*)bits, m_size.x, m_size.y, pitch, m_compression_level);
m_workers[m_frame % m_threads]->Push(std::make_shared<GSPng::Transaction>(GSPng::RGB_PNG, out_file, static_cast<const uint8*>(bits), m_size.x, m_size.y, pitch, m_compression_level)); m_workers[m_frame % m_threads]->Push(std::make_shared<GSPng::Transaction>(GSPng::RGB_PNG, out_file, static_cast<const u8*>(bits), m_size.x, m_size.y, pitch, m_compression_level));
m_frame++; m_frame++;

2
pcsx2/GS/GSCapture.h
View File

@ -28,7 +28,7 @@ class GSCapture
std::recursive_mutex m_lock; std::recursive_mutex m_lock;
bool m_capturing; bool m_capturing;
GSVector2i m_size; GSVector2i m_size;
uint64 m_frame; u64 m_frame;
std::string m_out_dir; std::string m_out_dir;
int m_threads; int m_threads;

68
pcsx2/GS/GSClut.cpp
View File

@ -22,11 +22,11 @@
GSClut::GSClut(GSLocalMemory* mem) GSClut::GSClut(GSLocalMemory* mem)
: m_mem(mem) : m_mem(mem)
{ {
uint8* p = (uint8*)vmalloc(CLUT_ALLOC_SIZE, false); u8* p = (u8*)vmalloc(CLUT_ALLOC_SIZE, false);
m_clut = (uint16*)&p[0]; // 1k + 1k for mirrored area simulating wrapping memory m_clut = (u16*)&p[0]; // 1k + 1k for mirrored area simulating wrapping memory
m_buff32 = (uint32*)&p[2048]; // 1k m_buff32 = (u32*)&p[2048]; // 1k
m_buff64 = (uint64*)&p[4096]; // 2k m_buff64 = (u64*)&p[4096]; // 2k
m_write.dirty = true; m_write.dirty = true;
m_read.dirty = true; m_read.dirty = true;
@ -107,7 +107,7 @@ void GSClut::Invalidate()
m_write.dirty = true; m_write.dirty = true;
} }
void GSClut::Invalidate(uint32 block) void GSClut::Invalidate(u32 block)
{ {
if (block == m_write.TEX0.CBP) if (block == m_write.TEX0.CBP)
{ {
@ -163,34 +163,34 @@ void GSClut::Write(const GIFRegTEX0& TEX0, const GIFRegTEXCLUT& TEXCLUT)
void GSClut::WriteCLUT32_I8_CSM1(const GIFRegTEX0& TEX0, const GIFRegTEXCLUT& TEXCLUT) void GSClut::WriteCLUT32_I8_CSM1(const GIFRegTEX0& TEX0, const GIFRegTEXCLUT& TEXCLUT)
{ {
ALIGN_STACK(32); ALIGN_STACK(32);
WriteCLUT_T32_I8_CSM1((uint32*)m_mem->BlockPtr32(0, 0, TEX0.CBP, 1), m_clut, (TEX0.CSA & 15)); WriteCLUT_T32_I8_CSM1((u32*)m_mem->BlockPtr32(0, 0, TEX0.CBP, 1), m_clut, (TEX0.CSA & 15));
} }
void GSClut::WriteCLUT32_I4_CSM1(const GIFRegTEX0& TEX0, const GIFRegTEXCLUT& TEXCLUT) void GSClut::WriteCLUT32_I4_CSM1(const GIFRegTEX0& TEX0, const GIFRegTEXCLUT& TEXCLUT)
{ {
ALIGN_STACK(32); ALIGN_STACK(32);
WriteCLUT_T32_I4_CSM1((uint32*)m_mem->BlockPtr32(0, 0, TEX0.CBP, 1), m_clut + ((TEX0.CSA & 15) << 4)); WriteCLUT_T32_I4_CSM1((u32*)m_mem->BlockPtr32(0, 0, TEX0.CBP, 1), m_clut + ((TEX0.CSA & 15) << 4));
} }
void GSClut::WriteCLUT16_I8_CSM1(const GIFRegTEX0& TEX0, const GIFRegTEXCLUT& TEXCLUT) void GSClut::WriteCLUT16_I8_CSM1(const GIFRegTEX0& TEX0, const GIFRegTEXCLUT& TEXCLUT)
{ {
WriteCLUT_T16_I8_CSM1((uint16*)m_mem->BlockPtr16(0, 0, TEX0.CBP, 1), m_clut + (TEX0.CSA << 4)); WriteCLUT_T16_I8_CSM1((u16*)m_mem->BlockPtr16(0, 0, TEX0.CBP, 1), m_clut + (TEX0.CSA << 4));
} }
void GSClut::WriteCLUT16_I4_CSM1(const GIFRegTEX0& TEX0, const GIFRegTEXCLUT& TEXCLUT) void GSClut::WriteCLUT16_I4_CSM1(const GIFRegTEX0& TEX0, const GIFRegTEXCLUT& TEXCLUT)
{ {
WriteCLUT_T16_I4_CSM1((uint16*)m_mem->BlockPtr16(0, 0, TEX0.CBP, 1), m_clut + (TEX0.CSA << 4)); WriteCLUT_T16_I4_CSM1((u16*)m_mem->BlockPtr16(0, 0, TEX0.CBP, 1), m_clut + (TEX0.CSA << 4));
} }
void GSClut::WriteCLUT16S_I8_CSM1(const GIFRegTEX0& TEX0, const GIFRegTEXCLUT& TEXCLUT) void GSClut::WriteCLUT16S_I8_CSM1(const GIFRegTEX0& TEX0, const GIFRegTEXCLUT& TEXCLUT)
{ {
WriteCLUT_T16_I8_CSM1((uint16*)m_mem->BlockPtr16S(0, 0, TEX0.CBP, 1), m_clut + (TEX0.CSA << 4)); WriteCLUT_T16_I8_CSM1((u16*)m_mem->BlockPtr16S(0, 0, TEX0.CBP, 1), m_clut + (TEX0.CSA << 4));
} }
void GSClut::WriteCLUT16S_I4_CSM1(const GIFRegTEX0& TEX0, const GIFRegTEXCLUT& TEXCLUT) void GSClut::WriteCLUT16S_I4_CSM1(const GIFRegTEX0& TEX0, const GIFRegTEXCLUT& TEXCLUT)
{ {
WriteCLUT_T16_I4_CSM1((uint16*)m_mem->BlockPtr16S(0, 0, TEX0.CBP, 1), m_clut + (TEX0.CSA << 4)); WriteCLUT_T16_I4_CSM1((u16*)m_mem->BlockPtr16S(0, 0, TEX0.CBP, 1), m_clut + (TEX0.CSA << 4));
} }
template <int n> template <int n>
@ -199,14 +199,14 @@ void GSClut::WriteCLUT32_CSM2(const GIFRegTEX0& TEX0, const GIFRegTEXCLUT& TEXCL
GSOffset off = GSOffset::fromKnownPSM(TEX0.CBP, TEXCLUT.CBW, PSM_PSMCT32); GSOffset off = GSOffset::fromKnownPSM(TEX0.CBP, TEXCLUT.CBW, PSM_PSMCT32);
auto pa = off.paMulti(m_mem->m_vm32, TEXCLUT.COU << 4, TEXCLUT.COV); auto pa = off.paMulti(m_mem->m_vm32, TEXCLUT.COU << 4, TEXCLUT.COV);
uint16* RESTRICT clut = m_clut + ((TEX0.CSA & 15) << 4); u16* RESTRICT clut = m_clut + ((TEX0.CSA & 15) << 4);
for (int i = 0; i < n; i++) for (int i = 0; i < n; i++)
{ {
uint32 c = *pa.value(i); u32 c = *pa.value(i);
clut[i] = (uint16)(c & 0xffff); clut[i] = (u16)(c & 0xffff);
clut[i + 256] = (uint16)(c >> 16); clut[i + 256] = (u16)(c >> 16);
} }
} }
@ -216,7 +216,7 @@ void GSClut::WriteCLUT16_CSM2(const GIFRegTEX0& TEX0, const GIFRegTEXCLUT& TEXCL
GSOffset off = GSOffset::fromKnownPSM(TEX0.CBP, TEXCLUT.CBW, PSM_PSMCT16); GSOffset off = GSOffset::fromKnownPSM(TEX0.CBP, TEXCLUT.CBW, PSM_PSMCT16);
auto pa = off.paMulti(m_mem->m_vm16, TEXCLUT.COU << 4, TEXCLUT.COV); auto pa = off.paMulti(m_mem->m_vm16, TEXCLUT.COU << 4, TEXCLUT.COV);
uint16* RESTRICT clut = m_clut + (TEX0.CSA << 4); u16* RESTRICT clut = m_clut + (TEX0.CSA << 4);
for (int i = 0; i < n; i++) for (int i = 0; i < n; i++)
{ {
@ -230,7 +230,7 @@ void GSClut::WriteCLUT16S_CSM2(const GIFRegTEX0& TEX0, const GIFRegTEXCLUT& TEXC
GSOffset off = GSOffset::fromKnownPSM(TEX0.CBP, TEXCLUT.CBW, PSM_PSMCT16S); GSOffset off = GSOffset::fromKnownPSM(TEX0.CBP, TEXCLUT.CBW, PSM_PSMCT16S);
auto pa = off.paMulti(m_mem->m_vm16, TEXCLUT.COU << 4, TEXCLUT.COV); auto pa = off.paMulti(m_mem->m_vm16, TEXCLUT.COU << 4, TEXCLUT.COV);
uint16* RESTRICT clut = m_clut + (TEX0.CSA << 4); u16* RESTRICT clut = m_clut + (TEX0.CSA << 4);
for (int i = 0; i < n; i++) for (int i = 0; i < n; i++)
{ {
@ -252,7 +252,7 @@ void GSClut::Read(const GIFRegTEX0& TEX0)
m_read.TEX0 = TEX0; m_read.TEX0 = TEX0;
m_read.dirty = false; m_read.dirty = false;
uint16* clut = m_clut; u16* clut = m_clut;
if(TEX0.CPSM == PSM_PSMCT32 || TEX0.CPSM == PSM_PSMCT24) if(TEX0.CPSM == PSM_PSMCT32 || TEX0.CPSM == PSM_PSMCT24)
{ {
@ -301,7 +301,7 @@ void GSClut::Read32(const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA)
m_read.dirty = false; m_read.dirty = false;
m_read.adirty = true; m_read.adirty = true;
uint16* clut = m_clut; u16* clut = m_clut;
if (TEX0.CPSM == PSM_PSMCT32 || TEX0.CPSM == PSM_PSMCT24) if (TEX0.CPSM == PSM_PSMCT32 || TEX0.CPSM == PSM_PSMCT24)
{ {
@ -317,7 +317,7 @@ void GSClut::Read32(const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA)
clut += (TEX0.CSA & 15) << 4; clut += (TEX0.CSA & 15) << 4;
// TODO: merge these functions // TODO: merge these functions
ReadCLUT_T32_I4(clut, m_buff32); ReadCLUT_T32_I4(clut, m_buff32);
ExpandCLUT64_T32_I8(m_buff32, (uint64*)m_buff64); // sw renderer does not need m_buff64 anymore ExpandCLUT64_T32_I8(m_buff32, (u64*)m_buff64); // sw renderer does not need m_buff64 anymore
break; break;
} }
} }
@ -336,7 +336,7 @@ void GSClut::Read32(const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA)
clut += TEX0.CSA << 4; clut += TEX0.CSA << 4;
// TODO: merge these functions // TODO: merge these functions
Expand16(clut, m_buff32, 16, TEXA); Expand16(clut, m_buff32, 16, TEXA);
ExpandCLUT64_T32_I8(m_buff32, (uint64*)m_buff64); // sw renderer does not need m_buff64 anymore ExpandCLUT64_T32_I8(m_buff32, (u64*)m_buff64); // sw renderer does not need m_buff64 anymore
break; break;
} }
} }
@ -412,7 +412,7 @@ void GSClut::GetAlphaMinMax32(int& amin_out, int& amax_out)
// //
void GSClut::WriteCLUT_T32_I8_CSM1(const uint32* RESTRICT src, uint16* RESTRICT clut, uint16 offset) void GSClut::WriteCLUT_T32_I8_CSM1(const u32* RESTRICT src, u16* RESTRICT clut, u16 offset)
{ {
// This is required when CSA is offset from the base of the CLUT so we point to the right data // This is required when CSA is offset from the base of the CLUT so we point to the right data
for (int i = offset; i < 16; i ++) for (int i = offset; i < 16; i ++)
@ -425,7 +425,7 @@ void GSClut::WriteCLUT_T32_I8_CSM1(const uint32* RESTRICT src, uint16* RESTRICT
} }
} }
__forceinline void GSClut::WriteCLUT_T32_I4_CSM1(const uint32* RESTRICT src, uint16* RESTRICT clut) __forceinline void GSClut::WriteCLUT_T32_I4_CSM1(const u32* RESTRICT src, u16* RESTRICT clut)
{ {
// 1 block // 1 block
@ -466,7 +466,7 @@ __forceinline void GSClut::WriteCLUT_T32_I4_CSM1(const uint32* RESTRICT src, uin
#endif #endif
} }
void GSClut::WriteCLUT_T16_I8_CSM1(const uint16* RESTRICT src, uint16* RESTRICT clut) void GSClut::WriteCLUT_T16_I8_CSM1(const u16* RESTRICT src, u16* RESTRICT clut)
{ {
// 2 blocks // 2 blocks
@ -491,7 +491,7 @@ void GSClut::WriteCLUT_T16_I8_CSM1(const uint16* RESTRICT src, uint16* RESTRICT
} }
} }
__forceinline void GSClut::WriteCLUT_T16_I4_CSM1(const uint16* RESTRICT src, uint16* RESTRICT clut) __forceinline void GSClut::WriteCLUT_T16_I4_CSM1(const u16* RESTRICT src, u16* RESTRICT clut)
{ {
// 1 block (half) // 1 block (half)
@ -501,7 +501,7 @@ __forceinline void GSClut::WriteCLUT_T16_I4_CSM1(const uint16* RESTRICT src, uin
} }
} }
void GSClut::ReadCLUT_T32_I8(const uint16* RESTRICT clut, uint32* RESTRICT dst, int offset) void GSClut::ReadCLUT_T32_I8(const u16* RESTRICT clut, u32* RESTRICT dst, int offset)
{ {
// Okay this deserves a small explanation // Okay this deserves a small explanation
// T32 I8 can address up to 256 colors however the offset can be "more than zero" when reading // T32 I8 can address up to 256 colors however the offset can be "more than zero" when reading
@ -516,7 +516,7 @@ void GSClut::ReadCLUT_T32_I8(const uint16* RESTRICT clut, uint32* RESTRICT dst,
} }
} }
__forceinline void GSClut::ReadCLUT_T32_I4(const uint16* RESTRICT clut, uint32* RESTRICT dst) __forceinline void GSClut::ReadCLUT_T32_I4(const u16* RESTRICT clut, u32* RESTRICT dst)
{ {
GSVector4i* s = (GSVector4i*)clut; GSVector4i* s = (GSVector4i*)clut;
GSVector4i* d = (GSVector4i*)dst; GSVector4i* d = (GSVector4i*)dst;
@ -535,7 +535,7 @@ __forceinline void GSClut::ReadCLUT_T32_I4(const uint16* RESTRICT clut, uint32*
} }
#if 0 #if 0
__forceinline void GSClut::ReadCLUT_T32_I4(const uint16* RESTRICT clut, uint32* RESTRICT dst32, uint64* RESTRICT dst64) __forceinline void GSClut::ReadCLUT_T32_I4(const u16* RESTRICT clut, u32* RESTRICT dst32, u64* RESTRICT dst64)
{ {
GSVector4i* s = (GSVector4i*)clut; GSVector4i* s = (GSVector4i*)clut;
GSVector4i* d32 = (GSVector4i*)dst32; GSVector4i* d32 = (GSVector4i*)dst32;
@ -561,7 +561,7 @@ __forceinline void GSClut::ReadCLUT_T32_I4(const uint16* RESTRICT clut, uint32*
#endif #endif
#if 0 #if 0
void GSClut::ReadCLUT_T16_I8(const uint16* RESTRICT clut, uint32* RESTRICT dst) void GSClut::ReadCLUT_T16_I8(const u16* RESTRICT clut, u32* RESTRICT dst)
{ {
for(int i = 0; i < 256; i += 16) for(int i = 0; i < 256; i += 16)
{ {
@ -571,7 +571,7 @@ void GSClut::ReadCLUT_T16_I8(const uint16* RESTRICT clut, uint32* RESTRICT dst)
#endif #endif
#if 0 #if 0
__forceinline void GSClut::ReadCLUT_T16_I4(const uint16* RESTRICT clut, uint32* RESTRICT dst) __forceinline void GSClut::ReadCLUT_T16_I4(const u16* RESTRICT clut, u32* RESTRICT dst)
{ {
GSVector4i* s = (GSVector4i*)clut; GSVector4i* s = (GSVector4i*)clut;
GSVector4i* d = (GSVector4i*)dst; GSVector4i* d = (GSVector4i*)dst;
@ -587,7 +587,7 @@ __forceinline void GSClut::ReadCLUT_T16_I4(const uint16* RESTRICT clut, uint32*
#endif #endif
#if 0 #if 0
__forceinline void GSClut::ReadCLUT_T16_I4(const uint16* RESTRICT clut, uint32* RESTRICT dst32, uint64* RESTRICT dst64) __forceinline void GSClut::ReadCLUT_T16_I4(const u16* RESTRICT clut, u32* RESTRICT dst32, u64* RESTRICT dst64)
{ {
GSVector4i* s = (GSVector4i*)clut; GSVector4i* s = (GSVector4i*)clut;
GSVector4i* d32 = (GSVector4i*)dst32; GSVector4i* d32 = (GSVector4i*)dst32;
@ -613,7 +613,7 @@ __forceinline void GSClut::ReadCLUT_T16_I4(const uint16* RESTRICT clut, uint32*
} }
#endif #endif
void GSClut::ExpandCLUT64_T32_I8(const uint32* RESTRICT src, uint64* RESTRICT dst) void GSClut::ExpandCLUT64_T32_I8(const u32* RESTRICT src, u64* RESTRICT dst)
{ {
GSVector4i* s = (GSVector4i*)src; GSVector4i* s = (GSVector4i*)src;
GSVector4i* d = (GSVector4i*)dst; GSVector4i* d = (GSVector4i*)dst;
@ -656,7 +656,7 @@ __forceinline void GSClut::ExpandCLUT64_T32(const GSVector4i& hi, const GSVector
} }
#if 0 #if 0
void GSClut::ExpandCLUT64_T16_I8(const uint32* RESTRICT src, uint64* RESTRICT dst) void GSClut::ExpandCLUT64_T16_I8(const u32* RESTRICT src, u64* RESTRICT dst)
{ {
GSVector4i* s = (GSVector4i*)src; GSVector4i* s = (GSVector4i*)src;
GSVector4i* d = (GSVector4i*)dst; GSVector4i* d = (GSVector4i*)dst;
@ -705,7 +705,7 @@ CONSTINIT const GSVector4i GSClut::m_bm = GSVector4i::cxpr(0x00007c00);
CONSTINIT const GSVector4i GSClut::m_gm = GSVector4i::cxpr(0x000003e0); CONSTINIT const GSVector4i GSClut::m_gm = GSVector4i::cxpr(0x000003e0);
CONSTINIT const GSVector4i GSClut::m_rm = GSVector4i::cxpr(0x0000001f); CONSTINIT const GSVector4i GSClut::m_rm = GSVector4i::cxpr(0x0000001f);
void GSClut::Expand16(const uint16* RESTRICT src, uint32* RESTRICT dst, int w, const GIFRegTEXA& TEXA) void GSClut::Expand16(const u16* RESTRICT src, u32* RESTRICT dst, int w, const GIFRegTEXA& TEXA)
{ {
ASSERT((w & 7) == 0); ASSERT((w & 7) == 0);

42
pcsx2/GS/GSClut.h
View File

@ -30,10 +30,10 @@ class alignas(32) GSClut : public GSAlignedClass<32>
GSLocalMemory* m_mem; GSLocalMemory* m_mem;
uint32 m_CBP[2]; u32 m_CBP[2];
uint16* m_clut; u16* m_clut;
uint32* m_buff32; u32* m_buff32;
uint64* m_buff64; u64* m_buff64;
struct alignas(32) WriteState struct alignas(32) WriteState
{ {
@ -74,42 +74,42 @@ class alignas(32) GSClut : public GSAlignedClass<32>
void WriteCLUT_NULL(const GIFRegTEX0& TEX0, const GIFRegTEXCLUT& TEXCLUT); void WriteCLUT_NULL(const GIFRegTEX0& TEX0, const GIFRegTEXCLUT& TEXCLUT);
static void WriteCLUT_T32_I8_CSM1(const uint32* RESTRICT src, uint16* RESTRICT clut, uint16 offset); static void WriteCLUT_T32_I8_CSM1(const u32* RESTRICT src, u16* RESTRICT clut, u16 offset);
static void WriteCLUT_T32_I4_CSM1(const uint32* RESTRICT src, uint16* RESTRICT clut); static void WriteCLUT_T32_I4_CSM1(const u32* RESTRICT src, u16* RESTRICT clut);
static void WriteCLUT_T16_I8_CSM1(const uint16* RESTRICT src, uint16* RESTRICT clut); static void WriteCLUT_T16_I8_CSM1(const u16* RESTRICT src, u16* RESTRICT clut);
static void WriteCLUT_T16_I4_CSM1(const uint16* RESTRICT src, uint16* RESTRICT clut); static void WriteCLUT_T16_I4_CSM1(const u16* RESTRICT src, u16* RESTRICT clut);
static void ReadCLUT_T32_I8(const uint16* RESTRICT clut, uint32* RESTRICT dst, int offset); static void ReadCLUT_T32_I8(const u16* RESTRICT clut, u32* RESTRICT dst, int offset);
static void ReadCLUT_T32_I4(const uint16* RESTRICT clut, uint32* RESTRICT dst); static void ReadCLUT_T32_I4(const u16* RESTRICT clut, u32* RESTRICT dst);
//static void ReadCLUT_T32_I4(const uint16* RESTRICT clut, uint32* RESTRICT dst32, uint64* RESTRICT dst64); //static void ReadCLUT_T32_I4(const u16* RESTRICT clut, u32* RESTRICT dst32, u64* RESTRICT dst64);
//static void ReadCLUT_T16_I8(const uint16* RESTRICT clut, uint32* RESTRICT dst); //static void ReadCLUT_T16_I8(const u16* RESTRICT clut, u32* RESTRICT dst);
//static void ReadCLUT_T16_I4(const uint16* RESTRICT clut, uint32* RESTRICT dst); //static void ReadCLUT_T16_I4(const u16* RESTRICT clut, u32* RESTRICT dst);
//static void ReadCLUT_T16_I4(const uint16* RESTRICT clut, uint32* RESTRICT dst32, uint64* RESTRICT dst64); //static void ReadCLUT_T16_I4(const u16* RESTRICT clut, u32* RESTRICT dst32, u64* RESTRICT dst64);
public: public:
static void ExpandCLUT64_T32_I8(const uint32* RESTRICT src, uint64* RESTRICT dst); static void ExpandCLUT64_T32_I8(const u32* RESTRICT src, u64* RESTRICT dst);
private: private:
static void ExpandCLUT64_T32(const GSVector4i& hi, const GSVector4i& lo0, const GSVector4i& lo1, const GSVector4i& lo2, const GSVector4i& lo3, GSVector4i* dst); static void ExpandCLUT64_T32(const GSVector4i& hi, const GSVector4i& lo0, const GSVector4i& lo1, const GSVector4i& lo2, const GSVector4i& lo3, GSVector4i* dst);
static void ExpandCLUT64_T32(const GSVector4i& hi, const GSVector4i& lo, GSVector4i* dst); static void ExpandCLUT64_T32(const GSVector4i& hi, const GSVector4i& lo, GSVector4i* dst);
//static void ExpandCLUT64_T16_I8(const uint32* RESTRICT src, uint64* RESTRICT dst); //static void ExpandCLUT64_T16_I8(const u32* RESTRICT src, u64* RESTRICT dst);
static void ExpandCLUT64_T16(const GSVector4i& hi, const GSVector4i& lo0, const GSVector4i& lo1, const GSVector4i& lo2, const GSVector4i& lo3, GSVector4i* dst); static void ExpandCLUT64_T16(const GSVector4i& hi, const GSVector4i& lo0, const GSVector4i& lo1, const GSVector4i& lo2, const GSVector4i& lo3, GSVector4i* dst);
static void ExpandCLUT64_T16(const GSVector4i& hi, const GSVector4i& lo, GSVector4i* dst); static void ExpandCLUT64_T16(const GSVector4i& hi, const GSVector4i& lo, GSVector4i* dst);
static void Expand16(const uint16* RESTRICT src, uint32* RESTRICT dst, int w, const GIFRegTEXA& TEXA); static void Expand16(const u16* RESTRICT src, u32* RESTRICT dst, int w, const GIFRegTEXA& TEXA);
public: public:
GSClut(GSLocalMemory* mem); GSClut(GSLocalMemory* mem);
virtual ~GSClut(); virtual ~GSClut();
void Invalidate(); void Invalidate();
void Invalidate(uint32 block); void Invalidate(u32 block);
bool WriteTest(const GIFRegTEX0& TEX0, const GIFRegTEXCLUT& TEXCLUT); bool WriteTest(const GIFRegTEX0& TEX0, const GIFRegTEXCLUT& TEXCLUT);
void Write(const GIFRegTEX0& TEX0, const GIFRegTEXCLUT& TEXCLUT); void Write(const GIFRegTEX0& TEX0, const GIFRegTEXCLUT& TEXCLUT);
//void Read(const GIFRegTEX0& TEX0); //void Read(const GIFRegTEX0& TEX0);
void Read32(const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA); void Read32(const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA);
void GetAlphaMinMax32(int& amin, int& amax); void GetAlphaMinMax32(int& amin, int& amax);
uint32 operator[](size_t i) const { return m_buff32[i]; } u32 operator[](size_t i) const { return m_buff32[i]; }
operator const uint32*() const { return m_buff32; } operator const u32*() const { return m_buff32; }
operator const uint64*() const { return m_buff64; } operator const u64*() const { return m_buff64; }
}; };

4
pcsx2/GS/GSCodeBuffer.cpp
View File

@ -42,14 +42,14 @@ void* GSCodeBuffer::GetBuffer(size_t size)
if (m_ptr == NULL || m_pos + size > m_blocksize) if (m_ptr == NULL || m_pos + size > m_blocksize)
{ {
m_ptr = (uint8*)vmalloc(m_blocksize, true); m_ptr = (u8*)vmalloc(m_blocksize, true);
m_pos = 0; m_pos = 0;
m_buffers.push_back(m_ptr); m_buffers.push_back(m_ptr);
} }
uint8* ptr = &m_ptr[m_pos]; u8* ptr = &m_ptr[m_pos];
m_reserved = size; m_reserved = size;

2
pcsx2/GS/GSCodeBuffer.h
View File

@ -22,7 +22,7 @@ class GSCodeBuffer
std::vector<void*> m_buffers; std::vector<void*> m_buffers;
size_t m_blocksize; size_t m_blocksize;
size_t m_pos, m_reserved; size_t m_pos, m_reserved;
uint8* m_ptr; u8* m_ptr;
public: public:
GSCodeBuffer(size_t blocksize = 4096 * 64); // 256k GSCodeBuffer(size_t blocksize = 4096 * 64); // 256k

6
pcsx2/GS/GSCrc.cpp
View File

@ -505,7 +505,7 @@ const CRC::Game CRC::m_games[] =
{0xB1BE3E51, Whiplash, EU, 0}, {0xB1BE3E51, Whiplash, EU, 0},
}; };
std::map<uint32, const CRC::Game*> CRC::m_map; std::map<u32, const CRC::Game*> CRC::m_map;
std::string ToLower(std::string str) std::string ToLower(std::string str)
{ {
@ -517,14 +517,14 @@ std::string ToLower(std::string str)
// The list is case insensitive and order insensitive. // The list is case insensitive and order insensitive.
// E.g. Disable all CRC hacks: CrcHacksExclusions=all // E.g. Disable all CRC hacks: CrcHacksExclusions=all
// E.g. Disable hacks for these CRCs: CrcHacksExclusions=0x0F0C4A9C, 0x0EE5646B, 0x7ACF7E03 // E.g. Disable hacks for these CRCs: CrcHacksExclusions=0x0F0C4A9C, 0x0EE5646B, 0x7ACF7E03
bool IsCrcExcluded(std::string exclusionList, uint32 crc) bool IsCrcExcluded(std::string exclusionList, u32 crc)
{ {
std::string target = format("0x%08x", crc); std::string target = format("0x%08x", crc);
exclusionList = ToLower(exclusionList); exclusionList = ToLower(exclusionList);
return exclusionList.find(target) != std::string::npos || exclusionList.find("all") != std::string::npos; return exclusionList.find(target) != std::string::npos || exclusionList.find("all") != std::string::npos;
} }
const CRC::Game& CRC::Lookup(uint32 crc) const CRC::Game& CRC::Lookup(u32 crc)
{ {
printf("GS Lookup CRC:%08X\n", crc); printf("GS Lookup CRC:%08X\n", crc);
if (m_map.empty()) if (m_map.empty())

8
pcsx2/GS/GSCrc.h
View File

@ -170,16 +170,16 @@ public:
struct Game struct Game
{ {
uint32 crc; u32 crc;
Title title; Title title;
Region region; Region region;
uint32 flags; u32 flags;
}; };
private: private:
static const Game m_games[]; static const Game m_games[];
static std::map<uint32, const Game*> m_map; static std::map<u32, const Game*> m_map;
public: public:
static const Game& Lookup(uint32 crc); static const Game& Lookup(u32 crc);
}; };

10
pcsx2/GS/GSDrawingContext.h
View File

@ -15,7 +15,7 @@
#pragma once #pragma once
#ifdef __clang__ #ifdef __clang__
// Ignore format for this file, as it spams a lot of warnings about uint64 and %llu. // Ignore format for this file, as it spams a lot of warnings about u64 and %llu.
#pragma clang diagnostic ignored "-Wformat" #pragma clang diagnostic ignored "-Wformat"
#endif #endif
@ -102,10 +102,10 @@ public:
{ {
ASSERT(XYOFFSET.OFX <= 0xf800 && XYOFFSET.OFY <= 0xf800); ASSERT(XYOFFSET.OFX <= 0xf800 && XYOFFSET.OFY <= 0xf800);
scissor.ex.u16[0] = (uint16)((SCISSOR.SCAX0 << 4) + XYOFFSET.OFX - 0x8000); scissor.ex.U16[0] = (u16)((SCISSOR.SCAX0 << 4) + XYOFFSET.OFX - 0x8000);
scissor.ex.u16[1] = (uint16)((SCISSOR.SCAY0 << 4) + XYOFFSET.OFY - 0x8000); scissor.ex.U16[1] = (u16)((SCISSOR.SCAY0 << 4) + XYOFFSET.OFY - 0x8000);
scissor.ex.u16[2] = (uint16)((SCISSOR.SCAX1 << 4) + XYOFFSET.OFX - 0x8000); scissor.ex.U16[2] = (u16)((SCISSOR.SCAX1 << 4) + XYOFFSET.OFX - 0x8000);
scissor.ex.u16[3] = (uint16)((SCISSOR.SCAY1 << 4) + XYOFFSET.OFY - 0x8000); scissor.ex.U16[3] = (u16)((SCISSOR.SCAY1 << 4) + XYOFFSET.OFY - 0x8000);
scissor.ofex = GSVector4( scissor.ofex = GSVector4(
(int)((SCISSOR.SCAX0 << 4) + XYOFFSET.OFX), (int)((SCISSOR.SCAX0 << 4) + XYOFFSET.OFX),

20
pcsx2/GS/GSDump.cpp
View File

@ -31,7 +31,7 @@ GSDumpBase::~GSDumpBase()
fclose(m_gs); fclose(m_gs);
} }
void GSDumpBase::AddHeader(uint32 crc, const freezeData& fd, const GSPrivRegSet* regs) void GSDumpBase::AddHeader(u32 crc, const freezeData& fd, const GSPrivRegSet* regs)
{ {
AppendRawData(&crc, 4); AppendRawData(&crc, 4);
AppendRawData(&fd.size, 4); AppendRawData(&fd.size, 4);
@ -39,18 +39,18 @@ void GSDumpBase::AddHeader(uint32 crc, const freezeData& fd, const GSPrivRegSet*
AppendRawData(regs, sizeof(*regs)); AppendRawData(regs, sizeof(*regs));
} }
void GSDumpBase::Transfer(int index, const uint8* mem, size_t size) void GSDumpBase::Transfer(int index, const u8* mem, size_t size)
{ {
if (size == 0) if (size == 0)
return; return;
AppendRawData(0); AppendRawData(0);
AppendRawData(static_cast<uint8>(index)); AppendRawData(static_cast<u8>(index));
AppendRawData(&size, 4); AppendRawData(&size, 4);
AppendRawData(mem, size); AppendRawData(mem, size);
} }
void GSDumpBase::ReadFIFO(uint32 size) void GSDumpBase::ReadFIFO(u32 size)
{ {
if (size == 0) if (size == 0)
return; return;
@ -69,7 +69,7 @@ bool GSDumpBase::VSync(int field, bool last, const GSPrivRegSet* regs)
AppendRawData(regs, sizeof(*regs)); AppendRawData(regs, sizeof(*regs));
AppendRawData(1); AppendRawData(1);
AppendRawData(static_cast<uint8>(field)); AppendRawData(static_cast<u8>(field));
if (last) if (last)
m_extra_frames--; m_extra_frames--;
@ -91,7 +91,7 @@ void GSDumpBase::Write(const void* data, size_t size)
// GSDump implementation // GSDump implementation
////////////////////////////////////////////////////////////////////// //////////////////////////////////////////////////////////////////////
GSDump::GSDump(const std::string& fn, uint32 crc, const freezeData& fd, const GSPrivRegSet* regs) GSDump::GSDump(const std::string& fn, u32 crc, const freezeData& fd, const GSPrivRegSet* regs)
: GSDumpBase(fn + ".gs") : GSDumpBase(fn + ".gs")
{ {
AddHeader(crc, fd, regs); AddHeader(crc, fd, regs);
@ -102,7 +102,7 @@ void GSDump::AppendRawData(const void* data, size_t size)
Write(data, size); Write(data, size);
} }
void GSDump::AppendRawData(uint8 c) void GSDump::AppendRawData(u8 c)
{ {
Write(&c, 1); Write(&c, 1);
} }
@ -111,7 +111,7 @@ void GSDump::AppendRawData(uint8 c)
// GSDumpXz implementation // GSDumpXz implementation
////////////////////////////////////////////////////////////////////// //////////////////////////////////////////////////////////////////////
GSDumpXz::GSDumpXz(const std::string& fn, uint32 crc, const freezeData& fd, const GSPrivRegSet* regs) GSDumpXz::GSDumpXz(const std::string& fn, u32 crc, const freezeData& fd, const GSPrivRegSet* regs)
: GSDumpBase(fn + ".gs.xz") : GSDumpBase(fn + ".gs.xz")
{ {
m_strm = LZMA_STREAM_INIT; m_strm = LZMA_STREAM_INIT;
@ -150,7 +150,7 @@ void GSDumpXz::AppendRawData(const void* data, size_t size)
Flush(); Flush();
} }
void GSDumpXz::AppendRawData(uint8 c) void GSDumpXz::AppendRawData(u8 c)
{ {
m_in_buff.push_back(c); m_in_buff.push_back(c);
} }
@ -170,7 +170,7 @@ void GSDumpXz::Flush()
void GSDumpXz::Compress(lzma_action action, lzma_ret expected_status) void GSDumpXz::Compress(lzma_action action, lzma_ret expected_status)
{ {
std::vector<uint8> out_buff(1024 * 1024); std::vector<u8> out_buff(1024 * 1024);
do do
{ {
m_strm.next_out = out_buff.data(); m_strm.next_out = out_buff.data();

18
pcsx2/GS/GSDump.h
View File

@ -45,28 +45,28 @@ class GSDumpBase
FILE* m_gs; FILE* m_gs;
protected: protected:
void AddHeader(uint32 crc, const freezeData& fd, const GSPrivRegSet* regs); void AddHeader(u32 crc, const freezeData& fd, const GSPrivRegSet* regs);
void Write(const void* data, size_t size); void Write(const void* data, size_t size);
virtual void AppendRawData(const void* data, size_t size) = 0; virtual void AppendRawData(const void* data, size_t size) = 0;
virtual void AppendRawData(uint8 c) = 0; virtual void AppendRawData(u8 c) = 0;
public: public:
GSDumpBase(const std::string& fn); GSDumpBase(const std::string& fn);
virtual ~GSDumpBase(); virtual ~GSDumpBase();
void ReadFIFO(uint32 size); void ReadFIFO(u32 size);
void Transfer(int index, const uint8* mem, size_t size); void Transfer(int index, const u8* mem, size_t size);
bool VSync(int field, bool last, const GSPrivRegSet* regs); bool VSync(int field, bool last, const GSPrivRegSet* regs);
}; };
class GSDump final : public GSDumpBase class GSDump final : public GSDumpBase
{ {
void AppendRawData(const void* data, size_t size) final; void AppendRawData(const void* data, size_t size) final;
void AppendRawData(uint8 c) final; void AppendRawData(u8 c) final;
public: public:
GSDump(const std::string& fn, uint32 crc, const freezeData& fd, const GSPrivRegSet* regs); GSDump(const std::string& fn, u32 crc, const freezeData& fd, const GSPrivRegSet* regs);
virtual ~GSDump() = default; virtual ~GSDump() = default;
}; };
@ -74,14 +74,14 @@ class GSDumpXz final : public GSDumpBase
{ {
lzma_stream m_strm; lzma_stream m_strm;
std::vector<uint8> m_in_buff; std::vector<u8> m_in_buff;
void Flush(); void Flush();
void Compress(lzma_action action, lzma_ret expected_status); void Compress(lzma_action action, lzma_ret expected_status);
void AppendRawData(const void* data, size_t size); void AppendRawData(const void* data, size_t size);
void AppendRawData(uint8 c); void AppendRawData(u8 c);
public: public:
GSDumpXz(const std::string& fn, uint32 crc, const freezeData& fd, const GSPrivRegSet* regs); GSDumpXz(const std::string& fn, u32 crc, const freezeData& fd, const GSPrivRegSet* regs);
virtual ~GSDumpXz(); virtual ~GSDumpXz();
}; };

314
pcsx2/GS/GSLocalMemory.cpp
View File

@ -18,7 +18,7 @@
#include "GS.h" #include "GS.h"
template <typename Fn> template <typename Fn>
static void foreachBlock(const GSOffset& off, GSLocalMemory* mem, const GSVector4i& r, uint8* dst, int dstpitch, int bpp, Fn&& fn) static void foreachBlock(const GSOffset& off, GSLocalMemory* mem, const GSVector4i& r, u8* dst, int dstpitch, int bpp, Fn&& fn)
{ {
ASSERT(off.isBlockAligned(r)); ASSERT(off.isBlockAligned(r));
GSOffset::BNHelper bn = off.bnMulti(r.left, r.top); GSOffset::BNHelper bn = off.bnMulti(r.left, r.top);
@ -32,8 +32,8 @@ static void foreachBlock(const GSOffset& off, GSLocalMemory* mem, const GSVector
{ {
for (int x = 0; bn.blkX() < right; bn.nextBlockX(), x += xAdd) for (int x = 0; bn.blkX() < right; bn.nextBlockX(), x += xAdd)
{ {
const uint8* src = mem->BlockPtr(bn.value()); const u8* src = mem->BlockPtr(bn.value());
uint8* read_dst = dst + x; u8* read_dst = dst + x;
fn(read_dst, src); fn(read_dst, src);
} }
} }
@ -70,19 +70,19 @@ GSLocalMemory::GSLocalMemory()
} }
if (m_use_fifo_alloc) if (m_use_fifo_alloc)
m_vm8 = (uint8*)fifo_alloc(m_vmsize, 4); m_vm8 = (u8*)fifo_alloc(m_vmsize, 4);
else else
m_vm8 = nullptr; m_vm8 = nullptr;
// Either we don't use fifo alloc or we get an error. // Either we don't use fifo alloc or we get an error.
if (m_vm8 == nullptr) if (m_vm8 == nullptr)
{ {
m_vm8 = (uint8*)vmalloc(m_vmsize * 4, false); m_vm8 = (u8*)vmalloc(m_vmsize * 4, false);
m_use_fifo_alloc = false; m_use_fifo_alloc = false;
} }
m_vm16 = (uint16*)m_vm8; m_vm16 = (u16*)m_vm8;
m_vm32 = (uint32*)m_vm8; m_vm32 = (u32*)m_vm8;
memset(m_vm8, 0, m_vmsize); memset(m_vm8, 0, m_vmsize);
@ -341,20 +341,20 @@ GSLocalMemory::~GSLocalMemory()
GSPixelOffset* GSLocalMemory::GetPixelOffset(const GIFRegFRAME& FRAME, const GIFRegZBUF& ZBUF) GSPixelOffset* GSLocalMemory::GetPixelOffset(const GIFRegFRAME& FRAME, const GIFRegZBUF& ZBUF)
{ {
uint32 fbp = FRAME.Block(); u32 fbp = FRAME.Block();
uint32 zbp = ZBUF.Block(); u32 zbp = ZBUF.Block();
uint32 fpsm = FRAME.PSM; u32 fpsm = FRAME.PSM;
uint32 zpsm = ZBUF.PSM; u32 zpsm = ZBUF.PSM;
uint32 bw = FRAME.FBW; u32 bw = FRAME.FBW;
ASSERT(m_psm[fpsm].trbpp > 8 || m_psm[zpsm].trbpp > 8); ASSERT(m_psm[fpsm].trbpp > 8 || m_psm[zpsm].trbpp > 8);
// "(psm & 0x0f) ^ ((psm & 0xf0) >> 2)" creates 4 bit unique identifiers for render target formats (only) // "(psm & 0x0f) ^ ((psm & 0xf0) >> 2)" creates 4 bit unique identifiers for render target formats (only)
uint32 fpsm_hash = (fpsm & 0x0f) ^ ((fpsm & 0x30) >> 2); u32 fpsm_hash = (fpsm & 0x0f) ^ ((fpsm & 0x30) >> 2);
uint32 zpsm_hash = (zpsm & 0x0f) ^ ((zpsm & 0x30) >> 2); u32 zpsm_hash = (zpsm & 0x0f) ^ ((zpsm & 0x30) >> 2);
uint32 hash = (FRAME.FBP << 0) | (ZBUF.ZBP << 9) | (bw << 18) | (fpsm_hash << 24) | (zpsm_hash << 28); u32 hash = (FRAME.FBP << 0) | (ZBUF.ZBP << 9) | (bw << 18) | (fpsm_hash << 24) | (zpsm_hash << 28);
auto it = m_pomap.find(hash); auto it = m_pomap.find(hash);
@ -394,20 +394,20 @@ GSPixelOffset* GSLocalMemory::GetPixelOffset(const GIFRegFRAME& FRAME, const GIF
GSPixelOffset4* GSLocalMemory::GetPixelOffset4(const GIFRegFRAME& FRAME, const GIFRegZBUF& ZBUF) GSPixelOffset4* GSLocalMemory::GetPixelOffset4(const GIFRegFRAME& FRAME, const GIFRegZBUF& ZBUF)
{ {
uint32 fbp = FRAME.Block(); u32 fbp = FRAME.Block();
uint32 zbp = ZBUF.Block(); u32 zbp = ZBUF.Block();
uint32 fpsm = FRAME.PSM; u32 fpsm = FRAME.PSM;
uint32 zpsm = ZBUF.PSM; u32 zpsm = ZBUF.PSM;
uint32 bw = FRAME.FBW; u32 bw = FRAME.FBW;
ASSERT(m_psm[fpsm].trbpp > 8 || m_psm[zpsm].trbpp > 8); ASSERT(m_psm[fpsm].trbpp > 8 || m_psm[zpsm].trbpp > 8);
// "(psm & 0x0f) ^ ((psm & 0xf0) >> 2)" creates 4 bit unique identifiers for render target formats (only) // "(psm & 0x0f) ^ ((psm & 0xf0) >> 2)" creates 4 bit unique identifiers for render target formats (only)
uint32 fpsm_hash = (fpsm & 0x0f) ^ ((fpsm & 0x30) >> 2); u32 fpsm_hash = (fpsm & 0x0f) ^ ((fpsm & 0x30) >> 2);
uint32 zpsm_hash = (zpsm & 0x0f) ^ ((zpsm & 0x30) >> 2); u32 zpsm_hash = (zpsm & 0x0f) ^ ((zpsm & 0x30) >> 2);
uint32 hash = (FRAME.FBP << 0) | (ZBUF.ZBP << 9) | (bw << 18) | (fpsm_hash << 24) | (zpsm_hash << 28); u32 hash = (FRAME.FBP << 0) | (ZBUF.ZBP << 9) | (bw << 18) | (fpsm_hash << 24) | (zpsm_hash << 28);
auto it = m_po4map.find(hash); auto it = m_po4map.find(hash);
@ -449,7 +449,7 @@ static bool cmp_vec2x(const GSVector2i& a, const GSVector2i& b) { return a.x < b
std::vector<GSVector2i>* GSLocalMemory::GetPage2TileMap(const GIFRegTEX0& TEX0) std::vector<GSVector2i>* GSLocalMemory::GetPage2TileMap(const GIFRegTEX0& TEX0)
{ {
uint64 hash = TEX0.u64 & 0x3ffffffffull; // TBP0 TBW PSM TW TH u64 hash = TEX0.U64 & 0x3ffffffffull; // TBP0 TBW PSM TW TH
auto it = m_p2tmap.find(hash); auto it = m_p2tmap.find(hash);
@ -466,34 +466,34 @@ std::vector<GSVector2i>* GSLocalMemory::GetPage2TileMap(const GIFRegTEX0& TEX0)
GSOffset off = GetOffset(TEX0.TBP0, TEX0.TBW, TEX0.PSM); GSOffset off = GetOffset(TEX0.TBP0, TEX0.TBW, TEX0.PSM);
GSOffset::BNHelper bn = off.bnMulti(0, 0); GSOffset::BNHelper bn = off.bnMulti(0, 0);
std::unordered_map<uint32, std::unordered_set<uint32>> tmp; // key = page, value = y:x, 7 bits each, max 128x128 tiles for the worst case (1024x1024 32bpp 8x8 blocks) std::unordered_map<u32, std::unordered_set<u32>> tmp; // key = page, value = y:x, 7 bits each, max 128x128 tiles for the worst case (1024x1024 32bpp 8x8 blocks)
for (; bn.blkY() < (th >> off.blockShiftY()); bn.nextBlockY()) for (; bn.blkY() < (th >> off.blockShiftY()); bn.nextBlockY())
{ {
for (; bn.blkX() < (tw >> off.blockShiftX()); bn.nextBlockX()) for (; bn.blkX() < (tw >> off.blockShiftX()); bn.nextBlockX())
{ {
uint32 page = (bn.value() >> 5) % MAX_PAGES; u32 page = (bn.value() >> 5) % MAX_PAGES;
tmp[page].insert((bn.blkY() << 7) + bn.blkX()); tmp[page].insert((bn.blkY() << 7) + bn.blkX());
} }
} }
// combine the lower 5 bits of the address into a 9:5 pointer:mask form, so the "valid bits" can be tested against an uint32 array // combine the lower 5 bits of the address into a 9:5 pointer:mask form, so the "valid bits" can be tested against an u32 array
auto p2t = new std::vector<GSVector2i>[MAX_PAGES]; auto p2t = new std::vector<GSVector2i>[MAX_PAGES];
for (const auto& i : tmp) for (const auto& i : tmp)
{ {
uint32 page = i.first; u32 page = i.first;
auto& tiles = i.second; auto& tiles = i.second;
std::unordered_map<uint32, uint32> m; std::unordered_map<u32, u32> m;
for (const auto addr : tiles) for (const auto addr : tiles)
{ {
uint32 row = addr >> 5; u32 row = addr >> 5;
uint32 col = 1 << (addr & 31); u32 col = 1 << (addr & 31);
auto k = m.find(row); auto k = m.find(row);
@ -528,10 +528,10 @@ std::vector<GSVector2i>* GSLocalMemory::GetPage2TileMap(const GIFRegTEX0& TEX0)
//////////////////// ////////////////////
template <int psm, int bsx, int bsy, int alignment> template <int psm, int bsx, int bsy, int alignment>
void GSLocalMemory::WriteImageColumn(int l, int r, int y, int h, const uint8* src, int srcpitch, const GIFRegBITBLTBUF& BITBLTBUF) void GSLocalMemory::WriteImageColumn(int l, int r, int y, int h, const u8* src, int srcpitch, const GIFRegBITBLTBUF& BITBLTBUF)
{ {
uint32 bp = BITBLTBUF.DBP; u32 bp = BITBLTBUF.DBP;
uint32 bw = BITBLTBUF.DBW; u32 bw = BITBLTBUF.DBW;
const int csy = bsy / 4; const int csy = bsy / 4;
@ -557,10 +557,10 @@ void GSLocalMemory::WriteImageColumn(int l, int r, int y, int h, const uint8* sr
} }
template <int psm, int bsx, int bsy, int alignment> template <int psm, int bsx, int bsy, int alignment>
void GSLocalMemory::WriteImageBlock(int l, int r, int y, int h, const uint8* src, int srcpitch, const GIFRegBITBLTBUF& BITBLTBUF) void GSLocalMemory::WriteImageBlock(int l, int r, int y, int h, const u8* src, int srcpitch, const GIFRegBITBLTBUF& BITBLTBUF)
{ {
uint32 bp = BITBLTBUF.DBP; u32 bp = BITBLTBUF.DBP;
uint32 bw = BITBLTBUF.DBW; u32 bw = BITBLTBUF.DBW;
for (int offset = srcpitch * bsy; h >= bsy; h -= bsy, y += bsy, src += offset) for (int offset = srcpitch * bsy; h >= bsy; h -= bsy, y += bsy, src += offset)
{ {
@ -584,10 +584,10 @@ void GSLocalMemory::WriteImageBlock(int l, int r, int y, int h, const uint8* src
} }
template <int psm, int bsx, int bsy> template <int psm, int bsx, int bsy>
void GSLocalMemory::WriteImageLeftRight(int l, int r, int y, int h, const uint8* src, int srcpitch, const GIFRegBITBLTBUF& BITBLTBUF) void GSLocalMemory::WriteImageLeftRight(int l, int r, int y, int h, const u8* src, int srcpitch, const GIFRegBITBLTBUF& BITBLTBUF)
{ {
uint32 bp = BITBLTBUF.DBP; u32 bp = BITBLTBUF.DBP;
uint32 bw = BITBLTBUF.DBW; u32 bw = BITBLTBUF.DBW;
for (; h > 0; y++, h--, src += srcpitch) for (; h > 0; y++, h--, src += srcpitch)
{ {
@ -595,14 +595,14 @@ void GSLocalMemory::WriteImageLeftRight(int l, int r, int y, int h, const uint8*
{ {
switch (psm) switch (psm)
{ {
case PSM_PSMCT32: WritePixel32(x, y, *(uint32*)&src[x * 4], bp, bw); break; case PSM_PSMCT32: WritePixel32(x, y, *(u32*)&src[x * 4], bp, bw); break;
case PSM_PSMCT16: WritePixel16(x, y, *(uint16*)&src[x * 2], bp, bw); break; case PSM_PSMCT16: WritePixel16(x, y, *(u16*)&src[x * 2], bp, bw); break;
case PSM_PSMCT16S: WritePixel16S(x, y, *(uint16*)&src[x * 2], bp, bw); break; case PSM_PSMCT16S: WritePixel16S(x, y, *(u16*)&src[x * 2], bp, bw); break;
case PSM_PSMT8: WritePixel8(x, y, src[x], bp, bw); break; case PSM_PSMT8: WritePixel8(x, y, src[x], bp, bw); break;
case PSM_PSMT4: WritePixel4(x, y, src[x >> 1] >> ((x & 1) << 2), bp, bw); break; case PSM_PSMT4: WritePixel4(x, y, src[x >> 1] >> ((x & 1) << 2), bp, bw); break;
case PSM_PSMZ32: WritePixel32Z(x, y, *(uint32*)&src[x * 4], bp, bw); break; case PSM_PSMZ32: WritePixel32Z(x, y, *(u32*)&src[x * 4], bp, bw); break;
case PSM_PSMZ16: WritePixel16Z(x, y, *(uint16*)&src[x * 2], bp, bw); break; case PSM_PSMZ16: WritePixel16Z(x, y, *(u16*)&src[x * 2], bp, bw); break;
case PSM_PSMZ16S: WritePixel16SZ(x, y, *(uint16*)&src[x * 2], bp, bw); break; case PSM_PSMZ16S: WritePixel16SZ(x, y, *(u16*)&src[x * 2], bp, bw); break;
// TODO // TODO
default: __assume(0); default: __assume(0);
} }
@ -611,12 +611,12 @@ void GSLocalMemory::WriteImageLeftRight(int l, int r, int y, int h, const uint8*
} }
template <int psm, int bsx, int bsy, int trbpp> template <int psm, int bsx, int bsy, int trbpp>
void GSLocalMemory::WriteImageTopBottom(int l, int r, int y, int h, const uint8* src, int srcpitch, const GIFRegBITBLTBUF& BITBLTBUF) void GSLocalMemory::WriteImageTopBottom(int l, int r, int y, int h, const u8* src, int srcpitch, const GIFRegBITBLTBUF& BITBLTBUF)
{ {
alignas(32) uint8 buff[64]; // merge buffer for one column alignas(32) u8 buff[64]; // merge buffer for one column
uint32 bp = BITBLTBUF.DBP; u32 bp = BITBLTBUF.DBP;
uint32 bw = BITBLTBUF.DBW; u32 bw = BITBLTBUF.DBW;
const int csy = bsy / 4; const int csy = bsy / 4;
@ -630,7 +630,7 @@ void GSLocalMemory::WriteImageTopBottom(int l, int r, int y, int h, const uint8*
for (int x = l; x < r; x += bsx) for (int x = l; x < r; x += bsx)
{ {
uint8* dst = NULL; u8* dst = NULL;
switch (psm) switch (psm)
{ {
@ -719,7 +719,7 @@ void GSLocalMemory::WriteImageTopBottom(int l, int r, int y, int h, const uint8*
{ {
for (int x = l; x < r; x += bsx) for (int x = l; x < r; x += bsx)
{ {
uint8* dst = NULL; u8* dst = NULL;
switch (psm) switch (psm)
{ {
@ -772,7 +772,7 @@ void GSLocalMemory::WriteImageTopBottom(int l, int r, int y, int h, const uint8*
} }
template <int psm, int bsx, int bsy, int trbpp> template <int psm, int bsx, int bsy, int trbpp>
void GSLocalMemory::WriteImage(int& tx, int& ty, const uint8* src, int len, GIFRegBITBLTBUF& BITBLTBUF, GIFRegTRXPOS& TRXPOS, GIFRegTRXREG& TRXREG) void GSLocalMemory::WriteImage(int& tx, int& ty, const u8* src, int len, GIFRegBITBLTBUF& BITBLTBUF, GIFRegTRXPOS& TRXPOS, GIFRegTRXREG& TRXREG)
{ {
if (TRXREG.RRW == 0) if (TRXREG.RRW == 0)
return; return;
@ -797,7 +797,7 @@ void GSLocalMemory::WriteImage(int& tx, int& ty, const uint8* src, int len, GIFR
if (ra - la >= bsx && h > 0) // "transfer width" >= "block width" && there is at least one full row if (ra - la >= bsx && h > 0) // "transfer width" >= "block width" && there is at least one full row
{ {
const uint8* s = &src[-l * trbpp >> 3]; const u8* s = &src[-l * trbpp >> 3];
src += srcpitch * h; src += srcpitch * h;
len -= srcpitch * h; len -= srcpitch * h;
@ -890,13 +890,13 @@ static bool IsTopLeftAligned(int dsax, int tx, int ty, int bw, int bh)
return ((dsax & (bw - 1)) == 0 && (tx & (bw - 1)) == 0 && dsax == tx && (ty & (bh - 1)) == 0); return ((dsax & (bw - 1)) == 0 && (tx & (bw - 1)) == 0 && dsax == tx && (ty & (bh - 1)) == 0);
} }
void GSLocalMemory::WriteImage24(int& tx, int& ty, const uint8* src, int len, GIFRegBITBLTBUF& BITBLTBUF, GIFRegTRXPOS& TRXPOS, GIFRegTRXREG& TRXREG) void GSLocalMemory::WriteImage24(int& tx, int& ty, const u8* src, int len, GIFRegBITBLTBUF& BITBLTBUF, GIFRegTRXPOS& TRXPOS, GIFRegTRXREG& TRXREG)
{ {
if (TRXREG.RRW == 0) if (TRXREG.RRW == 0)
return; return;
uint32 bp = BITBLTBUF.DBP; u32 bp = BITBLTBUF.DBP;
uint32 bw = BITBLTBUF.DBW; u32 bw = BITBLTBUF.DBW;
int tw = TRXPOS.DSAX + TRXREG.RRW, srcpitch = TRXREG.RRW * 3; int tw = TRXPOS.DSAX + TRXREG.RRW, srcpitch = TRXREG.RRW * 3;
int th = len / srcpitch; int th = len / srcpitch;
@ -925,13 +925,13 @@ void GSLocalMemory::WriteImage24(int& tx, int& ty, const uint8* src, int len, GI
} }
} }
void GSLocalMemory::WriteImage8H(int& tx, int& ty, const uint8* src, int len, GIFRegBITBLTBUF& BITBLTBUF, GIFRegTRXPOS& TRXPOS, GIFRegTRXREG& TRXREG) void GSLocalMemory::WriteImage8H(int& tx, int& ty, const u8* src, int len, GIFRegBITBLTBUF& BITBLTBUF, GIFRegTRXPOS& TRXPOS, GIFRegTRXREG& TRXREG)
{ {
if (TRXREG.RRW == 0) if (TRXREG.RRW == 0)
return; return;
uint32 bp = BITBLTBUF.DBP; u32 bp = BITBLTBUF.DBP;
uint32 bw = BITBLTBUF.DBW; u32 bw = BITBLTBUF.DBW;
int tw = TRXPOS.DSAX + TRXREG.RRW, srcpitch = TRXREG.RRW; int tw = TRXPOS.DSAX + TRXREG.RRW, srcpitch = TRXREG.RRW;
int th = len / srcpitch; int th = len / srcpitch;
@ -960,13 +960,13 @@ void GSLocalMemory::WriteImage8H(int& tx, int& ty, const uint8* src, int len, GI
} }
} }
void GSLocalMemory::WriteImage4HL(int& tx, int& ty, const uint8* src, int len, GIFRegBITBLTBUF& BITBLTBUF, GIFRegTRXPOS& TRXPOS, GIFRegTRXREG& TRXREG) void GSLocalMemory::WriteImage4HL(int& tx, int& ty, const u8* src, int len, GIFRegBITBLTBUF& BITBLTBUF, GIFRegTRXPOS& TRXPOS, GIFRegTRXREG& TRXREG)
{ {
if (TRXREG.RRW == 0) if (TRXREG.RRW == 0)
return; return;
uint32 bp = BITBLTBUF.DBP; u32 bp = BITBLTBUF.DBP;
uint32 bw = BITBLTBUF.DBW; u32 bw = BITBLTBUF.DBW;
int tw = TRXPOS.DSAX + TRXREG.RRW, srcpitch = TRXREG.RRW / 2; int tw = TRXPOS.DSAX + TRXREG.RRW, srcpitch = TRXREG.RRW / 2;
int th = len / srcpitch; int th = len / srcpitch;
@ -995,13 +995,13 @@ void GSLocalMemory::WriteImage4HL(int& tx, int& ty, const uint8* src, int len, G
} }
} }
void GSLocalMemory::WriteImage4HH(int& tx, int& ty, const uint8* src, int len, GIFRegBITBLTBUF& BITBLTBUF, GIFRegTRXPOS& TRXPOS, GIFRegTRXREG& TRXREG) void GSLocalMemory::WriteImage4HH(int& tx, int& ty, const u8* src, int len, GIFRegBITBLTBUF& BITBLTBUF, GIFRegTRXPOS& TRXPOS, GIFRegTRXREG& TRXREG)
{ {
if (TRXREG.RRW == 0) if (TRXREG.RRW == 0)
return; return;
uint32 bp = BITBLTBUF.DBP; u32 bp = BITBLTBUF.DBP;
uint32 bw = BITBLTBUF.DBW; u32 bw = BITBLTBUF.DBW;
int tw = TRXPOS.DSAX + TRXREG.RRW, srcpitch = TRXREG.RRW / 2; int tw = TRXPOS.DSAX + TRXREG.RRW, srcpitch = TRXREG.RRW / 2;
int th = len / srcpitch; int th = len / srcpitch;
@ -1030,13 +1030,13 @@ void GSLocalMemory::WriteImage4HH(int& tx, int& ty, const uint8* src, int len, G
} }
} }
void GSLocalMemory::WriteImage24Z(int& tx, int& ty, const uint8* src, int len, GIFRegBITBLTBUF& BITBLTBUF, GIFRegTRXPOS& TRXPOS, GIFRegTRXREG& TRXREG) void GSLocalMemory::WriteImage24Z(int& tx, int& ty, const u8* src, int len, GIFRegBITBLTBUF& BITBLTBUF, GIFRegTRXPOS& TRXPOS, GIFRegTRXREG& TRXREG)
{ {
if (TRXREG.RRW == 0) if (TRXREG.RRW == 0)
return; return;
uint32 bp = BITBLTBUF.DBP; u32 bp = BITBLTBUF.DBP;
uint32 bw = BITBLTBUF.DBW; u32 bw = BITBLTBUF.DBW;
int tw = TRXPOS.DSAX + TRXREG.RRW, srcpitch = TRXREG.RRW * 3; int tw = TRXPOS.DSAX + TRXREG.RRW, srcpitch = TRXREG.RRW * 3;
int th = len / srcpitch; int th = len / srcpitch;
@ -1122,17 +1122,17 @@ static void readWriteHelper(VM* vm, int& tx, int& ty, int len, int xinc, int sx,
readWriteHelperImpl(tx, ty, len, xinc, sx, w, [&](int x, int y){ return off.paMulti(vm, x, y); }, std::forward<Fn>(fn)); readWriteHelperImpl(tx, ty, len, xinc, sx, w, [&](int x, int y){ return off.paMulti(vm, x, y); }, std::forward<Fn>(fn));
} }
void GSLocalMemory::WriteImageX(int& tx, int& ty, const uint8* src, int len, GIFRegBITBLTBUF& BITBLTBUF, GIFRegTRXPOS& TRXPOS, GIFRegTRXREG& TRXREG) void GSLocalMemory::WriteImageX(int& tx, int& ty, const u8* src, int len, GIFRegBITBLTBUF& BITBLTBUF, GIFRegTRXPOS& TRXPOS, GIFRegTRXREG& TRXREG)
{ {
if (len <= 0) if (len <= 0)
return; return;
const uint8* pb = (uint8*)src; const u8* pb = (u8*)src;
const uint16* pw = (uint16*)src; const u16* pw = (u16*)src;
const uint32* pd = (uint32*)src; const u32* pd = (u32*)src;
uint32 bp = BITBLTBUF.DBP; u32 bp = BITBLTBUF.DBP;
uint32 bw = BITBLTBUF.DBW; u32 bw = BITBLTBUF.DBW;
int sx = TRXPOS.DSAX; int sx = TRXPOS.DSAX;
int w = TRXREG.RRW; int w = TRXREG.RRW;
@ -1154,7 +1154,7 @@ void GSLocalMemory::WriteImageX(int& tx, int& ty, const uint8* src, int len, GIF
case PSM_PSMZ24: case PSM_PSMZ24:
readWriteHelper(m_vm32, tx, ty, len / 3, 1, sx, w, off.assertSizesMatch(swizzle32), [&](auto& pa, int x) readWriteHelper(m_vm32, tx, ty, len / 3, 1, sx, w, off.assertSizesMatch(swizzle32), [&](auto& pa, int x)
{ {
WritePixel24(pa.value(x), *(uint32*)pb); WritePixel24(pa.value(x), *(u32*)pb);
pb += 3; pb += 3;
}); });
break; break;
@ -1217,17 +1217,17 @@ void GSLocalMemory::WriteImageX(int& tx, int& ty, const uint8* src, int len, GIF
// //
void GSLocalMemory::ReadImageX(int& tx, int& ty, uint8* dst, int len, GIFRegBITBLTBUF& BITBLTBUF, GIFRegTRXPOS& TRXPOS, GIFRegTRXREG& TRXREG) const void GSLocalMemory::ReadImageX(int& tx, int& ty, u8* dst, int len, GIFRegBITBLTBUF& BITBLTBUF, GIFRegTRXPOS& TRXPOS, GIFRegTRXREG& TRXREG) const
{ {
if (len <= 0) if (len <= 0)
return; return;
uint8* RESTRICT pb = (uint8*)dst; u8* RESTRICT pb = (u8*)dst;
uint16* RESTRICT pw = (uint16*)dst; u16* RESTRICT pw = (u16*)dst;
uint32* RESTRICT pd = (uint32*)dst; u32* RESTRICT pd = (u32*)dst;
uint32 bp = BITBLTBUF.SBP; u32 bp = BITBLTBUF.SBP;
uint32 bw = BITBLTBUF.SBW; u32 bw = BITBLTBUF.SBW;
int sx = TRXPOS.SSAX; int sx = TRXPOS.SSAX;
int w = TRXREG.RRW; int w = TRXREG.RRW;
@ -1262,7 +1262,7 @@ void GSLocalMemory::ReadImageX(int& tx, int& ty, uint8* dst, int len, GIFRegBITB
for (int ex8 = ex - 8; len >= 8 && x <= ex8; len -= 8, x += 8, pd += 8) for (int ex8 = ex - 8; len >= 8 && x <= ex8; len -= 8, x += 8, pd += 8)
{ {
uint32* ps = pa.value(x); u32* ps = pa.value(x);
GSVector4i::store<false>(&pd[0], GSVector4i::load(ps + 0, ps + 4)); GSVector4i::store<false>(&pd[0], GSVector4i::load(ps + 0, ps + 4));
GSVector4i::store<false>(&pd[4], GSVector4i::load(ps + 8, ps + 12)); GSVector4i::store<false>(&pd[4], GSVector4i::load(ps + 8, ps + 12));
@ -1293,10 +1293,10 @@ void GSLocalMemory::ReadImageX(int& tx, int& ty, uint8* dst, int len, GIFRegBITB
case PSM_PSMZ24: case PSM_PSMZ24:
readWriteHelper(m_vm32, tx, ty, len / 3, 1, sx, w, off.assertSizesMatch(swizzle32), [&](auto& pa, int x) readWriteHelper(m_vm32, tx, ty, len / 3, 1, sx, w, off.assertSizesMatch(swizzle32), [&](auto& pa, int x)
{ {
uint32 c = *pa.value(x); u32 c = *pa.value(x);
pb[0] = (uint8)(c); pb[0] = (u8)(c);
pb[1] = (uint8)(c >> 8); pb[1] = (u8)(c >> 8);
pb[2] = (uint8)(c >> 16); pb[2] = (u8)(c >> 16);
pb += 3; pb += 3;
}); });
break; break;
@ -1323,8 +1323,8 @@ void GSLocalMemory::ReadImageX(int& tx, int& ty, uint8* dst, int len, GIFRegBITB
case PSM_PSMT4: case PSM_PSMT4:
readWriteHelper(tx, ty, len * 2, 2, sx, w, GSOffset::fromKnownPSM(bp, bw, PSM_PSMT4), [&](GSOffset::PAHelper& pa, int x) readWriteHelper(tx, ty, len * 2, 2, sx, w, GSOffset::fromKnownPSM(bp, bw, PSM_PSMT4), [&](GSOffset::PAHelper& pa, int x)
{ {
uint8 low = ReadPixel4(pa.value(x)); u8 low = ReadPixel4(pa.value(x));
uint8 high = ReadPixel4(pa.value(x + 1)); u8 high = ReadPixel4(pa.value(x + 1));
*pb = low | (high << 4); *pb = low | (high << 4);
}); });
break; break;
@ -1332,7 +1332,7 @@ void GSLocalMemory::ReadImageX(int& tx, int& ty, uint8* dst, int len, GIFRegBITB
case PSM_PSMT8H: case PSM_PSMT8H:
readWriteHelper(m_vm32, tx, ty, len, 1, sx, w, GSOffset::fromKnownPSM(bp, bw, PSM_PSMT8H), [&](auto& pa, int x) readWriteHelper(m_vm32, tx, ty, len, 1, sx, w, GSOffset::fromKnownPSM(bp, bw, PSM_PSMT8H), [&](auto& pa, int x)
{ {
*pb = (uint8)(*pa.value(x) >> 24); *pb = (u8)(*pa.value(x) >> 24);
pb++; pb++;
}); });
break; break;
@ -1340,9 +1340,9 @@ void GSLocalMemory::ReadImageX(int& tx, int& ty, uint8* dst, int len, GIFRegBITB
case PSM_PSMT4HL: case PSM_PSMT4HL:
readWriteHelper(m_vm32, tx, ty, len * 2, 2, sx, w, GSOffset::fromKnownPSM(bp, bw, PSM_PSMT4HL), [&](auto& pa, int x) readWriteHelper(m_vm32, tx, ty, len * 2, 2, sx, w, GSOffset::fromKnownPSM(bp, bw, PSM_PSMT4HL), [&](auto& pa, int x)
{ {
uint32 c0 = *pa.value(x) >> 24 & 0x0f; u32 c0 = *pa.value(x) >> 24 & 0x0f;
uint32 c1 = *pa.value(x + 1) >> 20 & 0xf0; u32 c1 = *pa.value(x + 1) >> 20 & 0xf0;
*pb = (uint8)(c0 | c1); *pb = (u8)(c0 | c1);
pb++; pb++;
}); });
break; break;
@ -1350,9 +1350,9 @@ void GSLocalMemory::ReadImageX(int& tx, int& ty, uint8* dst, int len, GIFRegBITB
case PSM_PSMT4HH: case PSM_PSMT4HH:
readWriteHelper(m_vm32, tx, ty, len * 2, 2, sx, w, GSOffset::fromKnownPSM(bp, bw, PSM_PSMT4HH), [&](auto& pa, int x) readWriteHelper(m_vm32, tx, ty, len * 2, 2, sx, w, GSOffset::fromKnownPSM(bp, bw, PSM_PSMT4HH), [&](auto& pa, int x)
{ {
uint32 c0 = *pa.value(x) >> 28 & 0x0f; u32 c0 = *pa.value(x) >> 28 & 0x0f;
uint32 c1 = *pa.value(x + 1) >> 24 & 0xf0; u32 c1 = *pa.value(x + 1) >> 24 & 0xf0;
*pb = (uint8)(c0 | c1); *pb = (u8)(c0 | c1);
pb++; pb++;
}); });
break; break;
@ -1361,35 +1361,35 @@ void GSLocalMemory::ReadImageX(int& tx, int& ty, uint8* dst, int len, GIFRegBITB
/////////////////// ///////////////////
void GSLocalMemory::ReadTexture32(const GSOffset& off, const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEXA& TEXA) void GSLocalMemory::ReadTexture32(const GSOffset& off, const GSVector4i& r, u8* dst, int dstpitch, const GIFRegTEXA& TEXA)
{ {
foreachBlock(off.assertSizesMatch(swizzle32), this, r, dst, dstpitch, 32, [&](uint8* read_dst, const uint8* src) foreachBlock(off.assertSizesMatch(swizzle32), this, r, dst, dstpitch, 32, [&](u8* read_dst, const u8* src)
{ {
GSBlock::ReadBlock32(src, read_dst, dstpitch); GSBlock::ReadBlock32(src, read_dst, dstpitch);
}); });
} }
void GSLocalMemory::ReadTexture24(const GSOffset& off, const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEXA& TEXA) void GSLocalMemory::ReadTexture24(const GSOffset& off, const GSVector4i& r, u8* dst, int dstpitch, const GIFRegTEXA& TEXA)
{ {
if (TEXA.AEM) if (TEXA.AEM)
{ {
foreachBlock(off.assertSizesMatch(swizzle32), this, r, dst, dstpitch, 32, [&](uint8* read_dst, const uint8* src) foreachBlock(off.assertSizesMatch(swizzle32), this, r, dst, dstpitch, 32, [&](u8* read_dst, const u8* src)
{ {
GSBlock::ReadAndExpandBlock24<true>(src, read_dst, dstpitch, TEXA); GSBlock::ReadAndExpandBlock24<true>(src, read_dst, dstpitch, TEXA);
}); });
} }
else else
{ {
foreachBlock(off.assertSizesMatch(swizzle32), this, r, dst, dstpitch, 32, [&](uint8* read_dst, const uint8* src) foreachBlock(off.assertSizesMatch(swizzle32), this, r, dst, dstpitch, 32, [&](u8* read_dst, const u8* src)
{ {
GSBlock::ReadAndExpandBlock24<false>(src, read_dst, dstpitch, TEXA); GSBlock::ReadAndExpandBlock24<false>(src, read_dst, dstpitch, TEXA);
}); });
} }
} }
void GSLocalMemory::ReadTextureGPU24(const GSOffset& off, const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEXA& TEXA) void GSLocalMemory::ReadTextureGPU24(const GSOffset& off, const GSVector4i& r, u8* dst, int dstpitch, const GIFRegTEXA& TEXA)
{ {
foreachBlock(off.assertSizesMatch(swizzle16), this, r, dst, dstpitch, 16, [&](uint8* read_dst, const uint8* src) foreachBlock(off.assertSizesMatch(swizzle16), this, r, dst, dstpitch, 16, [&](u8* read_dst, const u8* src)
{ {
GSBlock::ReadBlock16(src, read_dst, dstpitch); GSBlock::ReadBlock16(src, read_dst, dstpitch);
}); });
@ -1398,78 +1398,78 @@ void GSLocalMemory::ReadTextureGPU24(const GSOffset& off, const GSVector4i& r, u
ASSERT(dstpitch >= r.width() * 4); ASSERT(dstpitch >= r.width() * 4);
for (int y = r.top; y < r.bottom; y++) for (int y = r.top; y < r.bottom; y++)
{ {
uint8* line = dst + y * dstpitch; u8* line = dst + y * dstpitch;
for (int x = r.right; x >= r.left; x--) for (int x = r.right; x >= r.left; x--)
{ {
*(uint32*)&line[x * 4] = *(uint32*)&line[x * 3] & 0xFFFFFF; *(u32*)&line[x * 4] = *(u32*)&line[x * 3] & 0xFFFFFF;
} }
} }
} }
void GSLocalMemory::ReadTexture16(const GSOffset& off, const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEXA& TEXA) void GSLocalMemory::ReadTexture16(const GSOffset& off, const GSVector4i& r, u8* dst, int dstpitch, const GIFRegTEXA& TEXA)
{ {
if (TEXA.AEM) if (TEXA.AEM)
{ {
foreachBlock(off.assertSizesMatch(swizzle16), this, r, dst, dstpitch, 32, [&](uint8* read_dst, const uint8* src) foreachBlock(off.assertSizesMatch(swizzle16), this, r, dst, dstpitch, 32, [&](u8* read_dst, const u8* src)
{ {
GSBlock::ReadAndExpandBlock16<true>(src, read_dst, dstpitch, TEXA); GSBlock::ReadAndExpandBlock16<true>(src, read_dst, dstpitch, TEXA);
}); });
} }
else else
{ {
foreachBlock(off.assertSizesMatch(swizzle16), this, r, dst, dstpitch, 32, [&](uint8* read_dst, const uint8* src) foreachBlock(off.assertSizesMatch(swizzle16), this, r, dst, dstpitch, 32, [&](u8* read_dst, const u8* src)
{ {
GSBlock::ReadAndExpandBlock16<false>(src, read_dst, dstpitch, TEXA); GSBlock::ReadAndExpandBlock16<false>(src, read_dst, dstpitch, TEXA);
}); });
} }
} }
void GSLocalMemory::ReadTexture8(const GSOffset& off, const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEXA& TEXA) void GSLocalMemory::ReadTexture8(const GSOffset& off, const GSVector4i& r, u8* dst, int dstpitch, const GIFRegTEXA& TEXA)
{ {
const uint32* pal = m_clut; const u32* pal = m_clut;
foreachBlock(off.assertSizesMatch(swizzle8), this, r, dst, dstpitch, 32, [&](uint8* read_dst, const uint8* src) foreachBlock(off.assertSizesMatch(swizzle8), this, r, dst, dstpitch, 32, [&](u8* read_dst, const u8* src)
{ {
GSBlock::ReadAndExpandBlock8_32(src, read_dst, dstpitch, pal); GSBlock::ReadAndExpandBlock8_32(src, read_dst, dstpitch, pal);
}); });
} }
void GSLocalMemory::ReadTexture4(const GSOffset& off, const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEXA& TEXA) void GSLocalMemory::ReadTexture4(const GSOffset& off, const GSVector4i& r, u8* dst, int dstpitch, const GIFRegTEXA& TEXA)
{ {
const uint64* pal = m_clut; const u64* pal = m_clut;
foreachBlock(off.assertSizesMatch(swizzle4), this, r, dst, dstpitch, 32, [&](uint8* read_dst, const uint8* src) foreachBlock(off.assertSizesMatch(swizzle4), this, r, dst, dstpitch, 32, [&](u8* read_dst, const u8* src)
{ {
GSBlock::ReadAndExpandBlock4_32(src, read_dst, dstpitch, pal); GSBlock::ReadAndExpandBlock4_32(src, read_dst, dstpitch, pal);
}); });
} }
void GSLocalMemory::ReadTexture8H(const GSOffset& off, const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEXA& TEXA) void GSLocalMemory::ReadTexture8H(const GSOffset& off, const GSVector4i& r, u8* dst, int dstpitch, const GIFRegTEXA& TEXA)
{ {
const uint32* pal = m_clut; const u32* pal = m_clut;
foreachBlock(off.assertSizesMatch(swizzle32), this, r, dst, dstpitch, 32, [&](uint8* read_dst, const uint8* src) foreachBlock(off.assertSizesMatch(swizzle32), this, r, dst, dstpitch, 32, [&](u8* read_dst, const u8* src)
{ {
GSBlock::ReadAndExpandBlock8H_32(src, read_dst, dstpitch, pal); GSBlock::ReadAndExpandBlock8H_32(src, read_dst, dstpitch, pal);
}); });
} }
void GSLocalMemory::ReadTexture4HL(const GSOffset& off, const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEXA& TEXA) void GSLocalMemory::ReadTexture4HL(const GSOffset& off, const GSVector4i& r, u8* dst, int dstpitch, const GIFRegTEXA& TEXA)
{ {
const uint32* pal = m_clut; const u32* pal = m_clut;
foreachBlock(off.assertSizesMatch(swizzle32), this, r, dst, dstpitch, 32, [&](uint8* read_dst, const uint8* src) foreachBlock(off.assertSizesMatch(swizzle32), this, r, dst, dstpitch, 32, [&](u8* read_dst, const u8* src)
{ {
GSBlock::ReadAndExpandBlock4HL_32(src, read_dst, dstpitch, pal); GSBlock::ReadAndExpandBlock4HL_32(src, read_dst, dstpitch, pal);
}); });
} }
void GSLocalMemory::ReadTexture4HH(const GSOffset& off, const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEXA& TEXA) void GSLocalMemory::ReadTexture4HH(const GSOffset& off, const GSVector4i& r, u8* dst, int dstpitch, const GIFRegTEXA& TEXA)
{ {
const uint32* pal = m_clut; const u32* pal = m_clut;
foreachBlock(off.assertSizesMatch(swizzle32), this, r, dst, dstpitch, 32, [&](uint8* read_dst, const uint8* src) foreachBlock(off.assertSizesMatch(swizzle32), this, r, dst, dstpitch, 32, [&](u8* read_dst, const u8* src)
{ {
GSBlock::ReadAndExpandBlock4HH_32(src, read_dst, dstpitch, pal); GSBlock::ReadAndExpandBlock4HH_32(src, read_dst, dstpitch, pal);
}); });
@ -1477,14 +1477,14 @@ void GSLocalMemory::ReadTexture4HH(const GSOffset& off, const GSVector4i& r, uin
/////////////////// ///////////////////
void GSLocalMemory::ReadTextureBlock32(uint32 bp, uint8* dst, int dstpitch, const GIFRegTEXA& TEXA) const void GSLocalMemory::ReadTextureBlock32(u32 bp, u8* dst, int dstpitch, const GIFRegTEXA& TEXA) const
{ {
ALIGN_STACK(32); ALIGN_STACK(32);
GSBlock::ReadBlock32(BlockPtr(bp), dst, dstpitch); GSBlock::ReadBlock32(BlockPtr(bp), dst, dstpitch);
} }
void GSLocalMemory::ReadTextureBlock24(uint32 bp, uint8* dst, int dstpitch, const GIFRegTEXA& TEXA) const void GSLocalMemory::ReadTextureBlock24(u32 bp, u8* dst, int dstpitch, const GIFRegTEXA& TEXA) const
{ {
ALIGN_STACK(32); ALIGN_STACK(32);
@ -1498,7 +1498,7 @@ void GSLocalMemory::ReadTextureBlock24(uint32 bp, uint8* dst, int dstpitch, cons
} }
} }
void GSLocalMemory::ReadTextureBlock16(uint32 bp, uint8* dst, int dstpitch, const GIFRegTEXA& TEXA) const void GSLocalMemory::ReadTextureBlock16(u32 bp, u8* dst, int dstpitch, const GIFRegTEXA& TEXA) const
{ {
ALIGN_STACK(32); ALIGN_STACK(32);
@ -1512,35 +1512,35 @@ void GSLocalMemory::ReadTextureBlock16(uint32 bp, uint8* dst, int dstpitch, cons
} }
} }
void GSLocalMemory::ReadTextureBlock8(uint32 bp, uint8* dst, int dstpitch, const GIFRegTEXA& TEXA) const void GSLocalMemory::ReadTextureBlock8(u32 bp, u8* dst, int dstpitch, const GIFRegTEXA& TEXA) const
{ {
ALIGN_STACK(32); ALIGN_STACK(32);
GSBlock::ReadAndExpandBlock8_32(BlockPtr(bp), dst, dstpitch, m_clut); GSBlock::ReadAndExpandBlock8_32(BlockPtr(bp), dst, dstpitch, m_clut);
} }
void GSLocalMemory::ReadTextureBlock4(uint32 bp, uint8* dst, int dstpitch, const GIFRegTEXA& TEXA) const void GSLocalMemory::ReadTextureBlock4(u32 bp, u8* dst, int dstpitch, const GIFRegTEXA& TEXA) const
{ {
ALIGN_STACK(32); ALIGN_STACK(32);
GSBlock::ReadAndExpandBlock4_32(BlockPtr(bp), dst, dstpitch, m_clut); GSBlock::ReadAndExpandBlock4_32(BlockPtr(bp), dst, dstpitch, m_clut);
} }
void GSLocalMemory::ReadTextureBlock8H(uint32 bp, uint8* dst, int dstpitch, const GIFRegTEXA& TEXA) const void GSLocalMemory::ReadTextureBlock8H(u32 bp, u8* dst, int dstpitch, const GIFRegTEXA& TEXA) const
{ {
ALIGN_STACK(32); ALIGN_STACK(32);
GSBlock::ReadAndExpandBlock8H_32(BlockPtr(bp), dst, dstpitch, m_clut); GSBlock::ReadAndExpandBlock8H_32(BlockPtr(bp), dst, dstpitch, m_clut);
} }
void GSLocalMemory::ReadTextureBlock4HL(uint32 bp, uint8* dst, int dstpitch, const GIFRegTEXA& TEXA) const void GSLocalMemory::ReadTextureBlock4HL(u32 bp, u8* dst, int dstpitch, const GIFRegTEXA& TEXA) const
{ {
ALIGN_STACK(32); ALIGN_STACK(32);
GSBlock::ReadAndExpandBlock4HL_32(BlockPtr(bp), dst, dstpitch, m_clut); GSBlock::ReadAndExpandBlock4HL_32(BlockPtr(bp), dst, dstpitch, m_clut);
} }
void GSLocalMemory::ReadTextureBlock4HH(uint32 bp, uint8* dst, int dstpitch, const GIFRegTEXA& TEXA) const void GSLocalMemory::ReadTextureBlock4HH(u32 bp, u8* dst, int dstpitch, const GIFRegTEXA& TEXA) const
{ {
ALIGN_STACK(32); ALIGN_STACK(32);
@ -1549,7 +1549,7 @@ void GSLocalMemory::ReadTextureBlock4HH(uint32 bp, uint8* dst, int dstpitch, con
/////////////////// ///////////////////
void GSLocalMemory::ReadTexture(const GSOffset& off, const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEXA& TEXA) void GSLocalMemory::ReadTexture(const GSOffset& off, const GSVector4i& r, u8* dst, int dstpitch, const GIFRegTEXA& TEXA)
{ {
const psm_t& psm = m_psm[off.psm()]; const psm_t& psm = m_psm[off.psm()];
@ -1566,7 +1566,7 @@ void GSLocalMemory::ReadTexture(const GSOffset& off, const GSVector4i& r, uint8*
GSVector4i cr = r.ralign<Align_Inside>(psm.bs); GSVector4i cr = r.ralign<Align_Inside>(psm.bs);
bool aligned = ((size_t)(dst + (cr.left - r.left) * sizeof(uint32)) & 0xf) == 0; bool aligned = ((size_t)(dst + (cr.left - r.left) * sizeof(u32)) & 0xf) == 0;
if (cr.rempty() || !aligned) if (cr.rempty() || !aligned)
{ {
@ -1579,7 +1579,7 @@ void GSLocalMemory::ReadTexture(const GSOffset& off, const GSVector4i& r, uint8*
{ {
for (int x = r.left, i = 0; x < r.right; x++, i++) for (int x = r.left, i = 0; x < r.right; x++, i++)
{ {
((uint32*)dst)[i] = (this->*rt)(x, y, TEX0, TEXA); ((u32*)dst)[i] = (this->*rt)(x, y, TEX0, TEXA);
} }
} }
} }
@ -1589,7 +1589,7 @@ void GSLocalMemory::ReadTexture(const GSOffset& off, const GSVector4i& r, uint8*
{ {
for (int x = r.left, i = 0; x < r.right; x++, i++) for (int x = r.left, i = 0; x < r.right; x++, i++)
{ {
((uint32*)dst)[i] = (this->*rt)(x, y, TEX0, TEXA); ((u32*)dst)[i] = (this->*rt)(x, y, TEX0, TEXA);
} }
} }
@ -1597,7 +1597,7 @@ void GSLocalMemory::ReadTexture(const GSOffset& off, const GSVector4i& r, uint8*
{ {
for (int x = r.left, i = 0; x < r.right; x++, i++) for (int x = r.left, i = 0; x < r.right; x++, i++)
{ {
((uint32*)dst)[i] = (this->*rt)(x, y, TEX0, TEXA); ((u32*)dst)[i] = (this->*rt)(x, y, TEX0, TEXA);
} }
} }
@ -1605,18 +1605,18 @@ void GSLocalMemory::ReadTexture(const GSOffset& off, const GSVector4i& r, uint8*
{ {
for (int x = r.left, i = 0; x < cr.left; x++, i++) for (int x = r.left, i = 0; x < cr.left; x++, i++)
{ {
((uint32*)dst)[i] = (this->*rt)(x, y, TEX0, TEXA); ((u32*)dst)[i] = (this->*rt)(x, y, TEX0, TEXA);
} }
for (int x = cr.right, i = x - r.left; x < r.right; x++, i++) for (int x = cr.right, i = x - r.left; x < r.right; x++, i++)
{ {
((uint32*)dst)[i] = (this->*rt)(x, y, TEX0, TEXA); ((u32*)dst)[i] = (this->*rt)(x, y, TEX0, TEXA);
} }
} }
if (!cr.rempty()) if (!cr.rempty())
{ {
(this->*rtx)(off, cr, dst + (cr.left - r.left) * sizeof(uint32), dstpitch, TEXA); (this->*rtx)(off, cr, dst + (cr.left - r.left) * sizeof(u32), dstpitch, TEXA);
} }
} }
} }
@ -1628,41 +1628,41 @@ void GSLocalMemory::ReadTexture(const GSOffset& off, const GSVector4i& r, uint8*
// 32/8 // 32/8
void GSLocalMemory::ReadTexture8P(const GSOffset& off, const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEXA& TEXA) void GSLocalMemory::ReadTexture8P(const GSOffset& off, const GSVector4i& r, u8* dst, int dstpitch, const GIFRegTEXA& TEXA)
{ {
foreachBlock(off.assertSizesMatch(swizzle8), this, r, dst, dstpitch, 8, [&](uint8* read_dst, const uint8* src) foreachBlock(off.assertSizesMatch(swizzle8), this, r, dst, dstpitch, 8, [&](u8* read_dst, const u8* src)
{ {
GSBlock::ReadBlock8(src, read_dst, dstpitch); GSBlock::ReadBlock8(src, read_dst, dstpitch);
}); });
} }
void GSLocalMemory::ReadTexture4P(const GSOffset& off, const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEXA& TEXA) void GSLocalMemory::ReadTexture4P(const GSOffset& off, const GSVector4i& r, u8* dst, int dstpitch, const GIFRegTEXA& TEXA)
{ {
foreachBlock(off.assertSizesMatch(swizzle4), this, r, dst, dstpitch, 8, [&](uint8* read_dst, const uint8* src) foreachBlock(off.assertSizesMatch(swizzle4), this, r, dst, dstpitch, 8, [&](u8* read_dst, const u8* src)
{ {
GSBlock::ReadBlock4P(src, read_dst, dstpitch); GSBlock::ReadBlock4P(src, read_dst, dstpitch);
}); });
} }
void GSLocalMemory::ReadTexture8HP(const GSOffset& off, const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEXA& TEXA) void GSLocalMemory::ReadTexture8HP(const GSOffset& off, const GSVector4i& r, u8* dst, int dstpitch, const GIFRegTEXA& TEXA)
{ {
foreachBlock(off.assertSizesMatch(swizzle32), this, r, dst, dstpitch, 8, [&](uint8* read_dst, const uint8* src) foreachBlock(off.assertSizesMatch(swizzle32), this, r, dst, dstpitch, 8, [&](u8* read_dst, const u8* src)
{ {
GSBlock::ReadBlock8HP(src, read_dst, dstpitch); GSBlock::ReadBlock8HP(src, read_dst, dstpitch);
}); });
} }
void GSLocalMemory::ReadTexture4HLP(const GSOffset& off, const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEXA& TEXA) void GSLocalMemory::ReadTexture4HLP(const GSOffset& off, const GSVector4i& r, u8* dst, int dstpitch, const GIFRegTEXA& TEXA)
{ {
foreachBlock(off.assertSizesMatch(swizzle32), this, r, dst, dstpitch, 8, [&](uint8* read_dst, const uint8* src) foreachBlock(off.assertSizesMatch(swizzle32), this, r, dst, dstpitch, 8, [&](u8* read_dst, const u8* src)
{ {
GSBlock::ReadBlock4HLP(src, read_dst, dstpitch); GSBlock::ReadBlock4HLP(src, read_dst, dstpitch);
}); });
} }
void GSLocalMemory::ReadTexture4HHP(const GSOffset& off, const GSVector4i& r, uint8* dst, int dstpitch, const GIFRegTEXA& TEXA) void GSLocalMemory::ReadTexture4HHP(const GSOffset& off, const GSVector4i& r, u8* dst, int dstpitch, const GIFRegTEXA& TEXA)
{ {
foreachBlock(off.assertSizesMatch(swizzle32), this, r, dst, dstpitch, 8, [&](uint8* read_dst, const uint8* src) foreachBlock(off.assertSizesMatch(swizzle32), this, r, dst, dstpitch, 8, [&](u8* read_dst, const u8* src)
{ {
GSBlock::ReadBlock4HHP(src, read_dst, dstpitch); GSBlock::ReadBlock4HHP(src, read_dst, dstpitch);
}); });
@ -1670,33 +1670,33 @@ void GSLocalMemory::ReadTexture4HHP(const GSOffset& off, const GSVector4i& r, ui
// //
void GSLocalMemory::ReadTextureBlock8P(uint32 bp, uint8* dst, int dstpitch, const GIFRegTEXA& TEXA) const void GSLocalMemory::ReadTextureBlock8P(u32 bp, u8* dst, int dstpitch, const GIFRegTEXA& TEXA) const
{ {
GSBlock::ReadBlock8(BlockPtr(bp), dst, dstpitch); GSBlock::ReadBlock8(BlockPtr(bp), dst, dstpitch);
} }
void GSLocalMemory::ReadTextureBlock4P(uint32 bp, uint8* dst, int dstpitch, const GIFRegTEXA& TEXA) const void GSLocalMemory::ReadTextureBlock4P(u32 bp, u8* dst, int dstpitch, const GIFRegTEXA& TEXA) const
{ {
ALIGN_STACK(32); ALIGN_STACK(32);
GSBlock::ReadBlock4P(BlockPtr(bp), dst, dstpitch); GSBlock::ReadBlock4P(BlockPtr(bp), dst, dstpitch);
} }
void GSLocalMemory::ReadTextureBlock8HP(uint32 bp, uint8* dst, int dstpitch, const GIFRegTEXA& TEXA) const void GSLocalMemory::ReadTextureBlock8HP(u32 bp, u8* dst, int dstpitch, const GIFRegTEXA& TEXA) const
{ {
ALIGN_STACK(32); ALIGN_STACK(32);
GSBlock::ReadBlock8HP(BlockPtr(bp), dst, dstpitch); GSBlock::ReadBlock8HP(BlockPtr(bp), dst, dstpitch);
} }
void GSLocalMemory::ReadTextureBlock4HLP(uint32 bp, uint8* dst, int dstpitch, const GIFRegTEXA& TEXA) const void GSLocalMemory::ReadTextureBlock4HLP(u32 bp, u8* dst, int dstpitch, const GIFRegTEXA& TEXA) const
{ {
ALIGN_STACK(32); ALIGN_STACK(32);
GSBlock::ReadBlock4HLP(BlockPtr(bp), dst, dstpitch); GSBlock::ReadBlock4HLP(BlockPtr(bp), dst, dstpitch);
} }
void GSLocalMemory::ReadTextureBlock4HHP(uint32 bp, uint8* dst, int dstpitch, const GIFRegTEXA& TEXA) const void GSLocalMemory::ReadTextureBlock4HHP(u32 bp, u8* dst, int dstpitch, const GIFRegTEXA& TEXA) const
{ {
ALIGN_STACK(32); ALIGN_STACK(32);
@ -1707,7 +1707,7 @@ void GSLocalMemory::ReadTextureBlock4HHP(uint32 bp, uint8* dst, int dstpitch, co
#include "Renderers/SW/GSTextureSW.h" #include "Renderers/SW/GSTextureSW.h"
void GSLocalMemory::SaveBMP(const std::string& fn, uint32 bp, uint32 bw, uint32 psm, int w, int h) void GSLocalMemory::SaveBMP(const std::string& fn, u32 bp, u32 bw, u32 psm, int w, int h)
{ {
int pitch = w * 4; int pitch = w * 4;
int size = pitch * h; int size = pitch * h;
@ -1721,13 +1721,13 @@ void GSLocalMemory::SaveBMP(const std::string& fn, uint32 bp, uint32 bw, uint32
readPixel rp = m_psm[psm].rp; readPixel rp = m_psm[psm].rp;
uint8* p = (uint8*)bits; u8* p = (u8*)bits;
for (int j = 0; j < h; j++, p += pitch) for (int j = 0; j < h; j++, p += pitch)
{ {
for (int i = 0; i < w; i++) for (int i = 0; i < w; i++)
{ {
((uint32*)p)[i] = (this->*rp)(i, j, TEX0.TBP0, TEX0.TBW); ((u32*)p)[i] = (this->*rp)(i, j, TEX0.TBP0, TEX0.TBW);
} }
} }

440
pcsx2/GS/GSLocalMemory.h
View File

File diff suppressed because it is too large Load Diff

2
pcsx2/GS/GSPerfMon.cpp
View File

@ -40,7 +40,7 @@ void GSPerfMon::Put(counter_t c, double val)
# else # else
clock_gettime(CLOCK_MONOTONIC, &ts); clock_gettime(CLOCK_MONOTONIC, &ts);
# endif # endif
uint64 now = (uint64)ts.tv_sec * (uint64)1e6 + (uint64)ts.tv_nsec / (uint64)1e3; u64 now = (u64)ts.tv_sec * (u64)1e6 + (u64)ts.tv_nsec / (u64)1e3;
#else #else
clock_t now = clock(); clock_t now = clock();
#endif #endif

8
pcsx2/GS/GSPerfMon.h
View File

@ -45,8 +45,8 @@ public:
protected: protected:
double m_counters[CounterLast]; double m_counters[CounterLast];
double m_stats[CounterLast]; double m_stats[CounterLast];
uint64 m_begin[TimerLast], m_total[TimerLast], m_start[TimerLast]; u64 m_begin[TimerLast], m_total[TimerLast], m_start[TimerLast];
uint64 m_frame; u64 m_frame;
clock_t m_lastframe; clock_t m_lastframe;
int m_count; int m_count;
@ -55,8 +55,8 @@ protected:
public: public:
GSPerfMon(); GSPerfMon();
void SetFrame(uint64 frame) { m_frame = frame; } void SetFrame(u64 frame) { m_frame = frame; }
uint64 GetFrame() { return m_frame; } u64 GetFrame() { return m_frame; }
void Put(counter_t c, double val = 0); void Put(counter_t c, double val = 0);
double Get(counter_t c) { return m_stats[c]; } double Get(counter_t c) { return m_stats[c]; }

12
pcsx2/GS/GSPng.cpp
View File

@ -38,8 +38,8 @@ struct
namespace GSPng namespace GSPng
{ {
bool SaveFile(const std::string& file, const Format fmt, const uint8* const image, bool SaveFile(const std::string& file, const Format fmt, const u8* const image,
uint8* const row, const int width, const int height, const int pitch, u8* const row, const int width, const int height, const int pitch,
const int compression, const bool rb_swapped = false, const bool first_image = false) const int compression, const bool rb_swapped = false, const bool first_image = false)
{ {
const int channel_bit_depth = pixel[fmt].channel_bit_depth; const int channel_bit_depth = pixel[fmt].channel_bit_depth;
@ -105,7 +105,7 @@ namespace GSPng
return success; return success;
} }
bool Save(GSPng::Format fmt, const std::string& file, uint8* image, int w, int h, int pitch, int compression, bool rb_swapped) bool Save(GSPng::Format fmt, const std::string& file, u8* image, int w, int h, int pitch, int compression, bool rb_swapped)
{ {
std::string root = file; std::string root = file;
root.replace(file.length() - 4, 4, ""); root.replace(file.length() - 4, 4, "");
@ -115,7 +115,7 @@ namespace GSPng
if (compression < 0 || compression > Z_BEST_COMPRESSION) if (compression < 0 || compression > Z_BEST_COMPRESSION)
compression = Z_BEST_SPEED; compression = Z_BEST_SPEED;
std::unique_ptr<uint8[]> row(new uint8[pixel[fmt].bytes_per_pixel_out * w]); std::unique_ptr<u8[]> row(new u8[pixel[fmt].bytes_per_pixel_out * w]);
std::string filename = root + pixel[fmt].extension[0]; std::string filename = root + pixel[fmt].extension[0];
if (!SaveFile(filename, fmt, image, row.get(), w, h, pitch, compression, rb_swapped, true)) if (!SaveFile(filename, fmt, image, row.get(), w, h, pitch, compression, rb_swapped, true))
@ -129,11 +129,11 @@ namespace GSPng
return SaveFile(filename, fmt, image, row.get(), w, h, pitch, compression); return SaveFile(filename, fmt, image, row.get(), w, h, pitch, compression);
} }
Transaction::Transaction(GSPng::Format fmt, const std::string& file, const uint8* image, int w, int h, int pitch, int compression) Transaction::Transaction(GSPng::Format fmt, const std::string& file, const u8* image, int w, int h, int pitch, int compression)
: m_fmt(fmt), m_file(file), m_w(w), m_h(h), m_pitch(pitch), m_compression(compression) : m_fmt(fmt), m_file(file), m_w(w), m_h(h), m_pitch(pitch), m_compression(compression)
{ {
// Note: yes it would be better to use shared pointer // Note: yes it would be better to use shared pointer
m_image = (uint8*)_aligned_malloc(pitch * h, 32); m_image = (u8*)_aligned_malloc(pitch * h, 32);
if (m_image) if (m_image)
memcpy(m_image, image, pitch * h); memcpy(m_image, image, pitch * h);
} }

6
pcsx2/GS/GSPng.h
View File

@ -37,17 +37,17 @@ namespace GSPng
public: public:
Format m_fmt; Format m_fmt;
const std::string m_file; const std::string m_file;
uint8* m_image; u8* m_image;
int m_w; int m_w;
int m_h; int m_h;
int m_pitch; int m_pitch;
int m_compression; int m_compression;
Transaction(GSPng::Format fmt, const std::string& file, const uint8* image, int w, int h, int pitch, int compression); Transaction(GSPng::Format fmt, const std::string& file, const u8* image, int w, int h, int pitch, int compression);
~Transaction(); ~Transaction();
}; };
bool Save(GSPng::Format fmt, const std::string& file, uint8* image, int w, int h, int pitch, int compression, bool rb_swapped = false); bool Save(GSPng::Format fmt, const std::string& file, u8* image, int w, int h, int pitch, int compression, bool rb_swapped = false);
void Process(std::shared_ptr<Transaction>& item); void Process(std::shared_ptr<Transaction>& item);

262
pcsx2/GS/GSState.cpp
View File

@ -148,7 +148,7 @@ GSState::~GSState()
_aligned_free(m_index.buff); _aligned_free(m_index.buff);
} }
void GSState::SetRegsMem(uint8* basemem) void GSState::SetRegsMem(u8* basemem)
{ {
ASSERT(basemem); ASSERT(basemem);
@ -338,8 +338,8 @@ GSVideoMode GSState::GetVideoMode()
// Other videomodes can't be detected on our side without the help of the data from core // Other videomodes can't be detected on our side without the help of the data from core
// You can only identify a limited number of video modes based on the info from CRTC registers. // You can only identify a limited number of video modes based on the info from CRTC registers.
const uint8 Colorburst = m_regs->SMODE1.CMOD; // Subcarrier frequency const u8 Colorburst = m_regs->SMODE1.CMOD; // Subcarrier frequency
const uint8 PLL_Divider = m_regs->SMODE1.LC; // Phased lock loop divider const u8 PLL_Divider = m_regs->SMODE1.LC; // Phased lock loop divider
switch (Colorburst) switch (Colorburst)
{ {
@ -438,13 +438,13 @@ GSVector4i GSState::GetDisplayRect(int i)
const auto& DISP = m_regs->DISP[i].DISPLAY; const auto& DISP = m_regs->DISP[i].DISPLAY;
const uint32 DW = DISP.DW + 1; const u32 DW = DISP.DW + 1;
const uint32 DH = DISP.DH + 1; const u32 DH = DISP.DH + 1;
const uint32 DX = DISP.DX; const u32 DX = DISP.DX;
const uint32 DY = DISP.DY; const u32 DY = DISP.DY;
const uint32 MAGH = DISP.MAGH + 1; const u32 MAGH = DISP.MAGH + 1;
const uint32 MAGV = DISP.MAGV + 1; const u32 MAGV = DISP.MAGV + 1;
const GSVector2i magnification(MAGH, MAGV); const GSVector2i magnification(MAGH, MAGV);
@ -478,8 +478,8 @@ GSVector4i GSState::GetFrameRect(int i)
if (isinterlaced() && m_regs->SMODE2.FFMD && h > 1) if (isinterlaced() && m_regs->SMODE2.FFMD && h > 1)
h >>= 1; h >>= 1;
const uint32 DBX = m_regs->DISP[i].DISPFB.DBX; const u32 DBX = m_regs->DISP[i].DISPFB.DBX;
const uint32 DBY = m_regs->DISP[i].DISPFB.DBY; const u32 DBY = m_regs->DISP[i].DISPFB.DBY;
rectangle.left = DBX; rectangle.left = DBX;
rectangle.top = DBY; rectangle.top = DBY;
@ -560,16 +560,16 @@ void GSState::GIFPackedRegHandlerRGBA(const GIFPackedReg* RESTRICT r)
const GSVector4i mask = GSVector4i::load(0x0c080400); const GSVector4i mask = GSVector4i::load(0x0c080400);
const GSVector4i v = GSVector4i::load<false>(r).shuffle8(mask); const GSVector4i v = GSVector4i::load<false>(r).shuffle8(mask);
m_v.RGBAQ.u32[0] = (uint32)GSVector4i::store(v); m_v.RGBAQ.U32[0] = (u32)GSVector4i::store(v);
m_v.RGBAQ.Q = m_q; m_v.RGBAQ.Q = m_q;
} }
void GSState::GIFPackedRegHandlerSTQ(const GIFPackedReg* RESTRICT r) void GSState::GIFPackedRegHandlerSTQ(const GIFPackedReg* RESTRICT r)
{ {
const GSVector4i st = GSVector4i::loadl(&r->u64[0]); const GSVector4i st = GSVector4i::loadl(&r->U64[0]);
GSVector4i q = GSVector4i::loadl(&r->u64[1]); GSVector4i q = GSVector4i::loadl(&r->U64[1]);
GSVector4i::storel(&m_v.ST, st); GSVector4i::storel(&m_v.ST, st);
// Vexx (character shadow) // Vexx (character shadow)
@ -593,23 +593,23 @@ void GSState::GIFPackedRegHandlerUV(const GIFPackedReg* RESTRICT r)
{ {
const GSVector4i v = GSVector4i::loadl(r) & GSVector4i::x00003fff(); const GSVector4i v = GSVector4i::loadl(r) & GSVector4i::x00003fff();
m_v.UV = (uint32)GSVector4i::store(v.ps32(v)); m_v.UV = (u32)GSVector4i::store(v.ps32(v));
} }
void GSState::GIFPackedRegHandlerUV_Hack(const GIFPackedReg* RESTRICT r) void GSState::GIFPackedRegHandlerUV_Hack(const GIFPackedReg* RESTRICT r)
{ {
const GSVector4i v = GSVector4i::loadl(r) & GSVector4i::x00003fff(); const GSVector4i v = GSVector4i::loadl(r) & GSVector4i::x00003fff();
m_v.UV = (uint32)GSVector4i::store(v.ps32(v)); m_v.UV = (u32)GSVector4i::store(v.ps32(v));
m_isPackedUV_HackFlag = true; m_isPackedUV_HackFlag = true;
} }
template <uint32 prim, uint32 adc, bool auto_flush> template <u32 prim, u32 adc, bool auto_flush>
void GSState::GIFPackedRegHandlerXYZF2(const GIFPackedReg* RESTRICT r) void GSState::GIFPackedRegHandlerXYZF2(const GIFPackedReg* RESTRICT r)
{ {
GSVector4i xy = GSVector4i::loadl(&r->u64[0]); GSVector4i xy = GSVector4i::loadl(&r->U64[0]);
GSVector4i zf = GSVector4i::loadl(&r->u64[1]); GSVector4i zf = GSVector4i::loadl(&r->U64[1]);
xy = xy.upl16(xy.srl<4>()).upl32(GSVector4i::load((int)m_v.UV)); xy = xy.upl16(xy.srl<4>()).upl32(GSVector4i::load((int)m_v.UV));
zf = zf.srl32(4) & GSVector4i::x00ffffff().upl32(GSVector4i::x000000ff()); zf = zf.srl32(4) & GSVector4i::x00ffffff().upl32(GSVector4i::x000000ff());
@ -619,11 +619,11 @@ void GSState::GIFPackedRegHandlerXYZF2(const GIFPackedReg* RESTRICT r)
VertexKick<prim, auto_flush>(adc ? 1 : r->XYZF2.Skip()); VertexKick<prim, auto_flush>(adc ? 1 : r->XYZF2.Skip());
} }
template <uint32 prim, uint32 adc, bool auto_flush> template <u32 prim, u32 adc, bool auto_flush>
void GSState::GIFPackedRegHandlerXYZ2(const GIFPackedReg* RESTRICT r) void GSState::GIFPackedRegHandlerXYZ2(const GIFPackedReg* RESTRICT r)
{ {
const GSVector4i xy = GSVector4i::loadl(&r->u64[0]); const GSVector4i xy = GSVector4i::loadl(&r->U64[0]);
const GSVector4i z = GSVector4i::loadl(&r->u64[1]); const GSVector4i z = GSVector4i::loadl(&r->U64[1]);
const GSVector4i xyz = xy.upl16(xy.srl<4>()).upl32(z); const GSVector4i xyz = xy.upl16(xy.srl<4>()).upl32(z);
m_v.m[1] = xyz.upl64(GSVector4i::loadl(&m_v.UV)); m_v.m[1] = xyz.upl64(GSVector4i::loadl(&m_v.UV));
@ -645,8 +645,8 @@ void GSState::GIFPackedRegHandlerNOP(const GIFPackedReg* RESTRICT r)
{ {
} }
template <uint32 prim, bool auto_flush> template <u32 prim, bool auto_flush>
void GSState::GIFPackedRegHandlerSTQRGBAXYZF2(const GIFPackedReg* RESTRICT r, uint32 size) void GSState::GIFPackedRegHandlerSTQRGBAXYZF2(const GIFPackedReg* RESTRICT r, u32 size)
{ {
ASSERT(size > 0 && size % 3 == 0); ASSERT(size > 0 && size % 3 == 0);
@ -654,16 +654,16 @@ void GSState::GIFPackedRegHandlerSTQRGBAXYZF2(const GIFPackedReg* RESTRICT r, ui
while (r < r_end) while (r < r_end)
{ {
GSVector4i st = GSVector4i::loadl(&r[0].u64[0]); GSVector4i st = GSVector4i::loadl(&r[0].U64[0]);
GSVector4i q = GSVector4i::loadl(&r[0].u64[1]); GSVector4i q = GSVector4i::loadl(&r[0].U64[1]);
GSVector4i rgba = (GSVector4i::load<false>(&r[1]) & GSVector4i::x000000ff()).ps32().pu16(); GSVector4i rgba = (GSVector4i::load<false>(&r[1]) & GSVector4i::x000000ff()).ps32().pu16();
q = q.blend8(GSVector4i::cast(GSVector4::m_one), q == GSVector4i::zero()); // see GIFPackedRegHandlerSTQ q = q.blend8(GSVector4i::cast(GSVector4::m_one), q == GSVector4i::zero()); // see GIFPackedRegHandlerSTQ
m_v.m[0] = st.upl64(rgba.upl32(q)); // TODO: only store the last one m_v.m[0] = st.upl64(rgba.upl32(q)); // TODO: only store the last one
GSVector4i xy = GSVector4i::loadl(&r[2].u64[0]); GSVector4i xy = GSVector4i::loadl(&r[2].U64[0]);
GSVector4i zf = GSVector4i::loadl(&r[2].u64[1]); GSVector4i zf = GSVector4i::loadl(&r[2].U64[1]);
xy = xy.upl16(xy.srl<4>()).upl32(GSVector4i::load((int)m_v.UV)); xy = xy.upl16(xy.srl<4>()).upl32(GSVector4i::load((int)m_v.UV));
zf = zf.srl32(4) & GSVector4i::x00ffffff().upl32(GSVector4i::x000000ff()); zf = zf.srl32(4) & GSVector4i::x00ffffff().upl32(GSVector4i::x000000ff());
@ -677,8 +677,8 @@ void GSState::GIFPackedRegHandlerSTQRGBAXYZF2(const GIFPackedReg* RESTRICT r, ui
m_q = r[-3].STQ.Q; // remember the last one, STQ outputs this to the temp Q each time m_q = r[-3].STQ.Q; // remember the last one, STQ outputs this to the temp Q each time
} }
template <uint32 prim, bool auto_flush> template <u32 prim, bool auto_flush>
void GSState::GIFPackedRegHandlerSTQRGBAXYZ2(const GIFPackedReg* RESTRICT r, uint32 size) void GSState::GIFPackedRegHandlerSTQRGBAXYZ2(const GIFPackedReg* RESTRICT r, u32 size)
{ {
ASSERT(size > 0 && size % 3 == 0); ASSERT(size > 0 && size % 3 == 0);
@ -686,16 +686,16 @@ void GSState::GIFPackedRegHandlerSTQRGBAXYZ2(const GIFPackedReg* RESTRICT r, uin
while (r < r_end) while (r < r_end)
{ {
GSVector4i st = GSVector4i::loadl(&r[0].u64[0]); GSVector4i st = GSVector4i::loadl(&r[0].U64[0]);
GSVector4i q = GSVector4i::loadl(&r[0].u64[1]); GSVector4i q = GSVector4i::loadl(&r[0].U64[1]);
GSVector4i rgba = (GSVector4i::load<false>(&r[1]) & GSVector4i::x000000ff()).ps32().pu16(); GSVector4i rgba = (GSVector4i::load<false>(&r[1]) & GSVector4i::x000000ff()).ps32().pu16();
q = q.blend8(GSVector4i::cast(GSVector4::m_one), q == GSVector4i::zero()); // see GIFPackedRegHandlerSTQ q = q.blend8(GSVector4i::cast(GSVector4::m_one), q == GSVector4i::zero()); // see GIFPackedRegHandlerSTQ
m_v.m[0] = st.upl64(rgba.upl32(q)); // TODO: only store the last one m_v.m[0] = st.upl64(rgba.upl32(q)); // TODO: only store the last one
GSVector4i xy = GSVector4i::loadl(&r[2].u64[0]); GSVector4i xy = GSVector4i::loadl(&r[2].U64[0]);
GSVector4i z = GSVector4i::loadl(&r[2].u64[1]); GSVector4i z = GSVector4i::loadl(&r[2].U64[1]);
GSVector4i xyz = xy.upl16(xy.srl<4>()).upl32(z); GSVector4i xyz = xy.upl16(xy.srl<4>()).upl32(z);
m_v.m[1] = xyz.upl64(GSVector4i::loadl(&m_v.UV)); // TODO: only store the last one m_v.m[1] = xyz.upl64(GSVector4i::loadl(&m_v.UV)); // TODO: only store the last one
@ -708,7 +708,7 @@ void GSState::GIFPackedRegHandlerSTQRGBAXYZ2(const GIFPackedReg* RESTRICT r, uin
m_q = r[-3].STQ.Q; // remember the last one, STQ outputs this to the temp Q each time m_q = r[-3].STQ.Q; // remember the last one, STQ outputs this to the temp Q each time
} }
void GSState::GIFPackedRegHandlerNOP(const GIFPackedReg* RESTRICT r, uint32 size) void GSState::GIFPackedRegHandlerNOP(const GIFPackedReg* RESTRICT r, u32 size)
{ {
} }
@ -716,11 +716,11 @@ void GSState::GIFRegHandlerNull(const GIFReg* RESTRICT r)
{ {
} }
__forceinline void GSState::ApplyPRIM(uint32 prim) __forceinline void GSState::ApplyPRIM(u32 prim)
{ {
if (GSUtil::GetPrimClass(m_env.PRIM.PRIM) == GSUtil::GetPrimClass(prim & 7)) // NOTE: assume strips/fans are converted to lists if (GSUtil::GetPrimClass(m_env.PRIM.PRIM) == GSUtil::GetPrimClass(prim & 7)) // NOTE: assume strips/fans are converted to lists
{ {
if (m_env.PRMODECONT.AC == 1 && (m_env.PRIM.u32[0] ^ prim) & 0x7f8) // all fields except PRIM if (m_env.PRMODECONT.AC == 1 && (m_env.PRIM.U32[0] ^ prim) & 0x7f8) // all fields except PRIM
Flush(); Flush();
} }
else else
@ -730,7 +730,7 @@ __forceinline void GSState::ApplyPRIM(uint32 prim)
if (m_env.PRMODECONT.AC == 1) if (m_env.PRMODECONT.AC == 1)
{ {
m_env.PRIM.u32[0] = prim; m_env.PRIM.U32[0] = prim;
UpdateContext(); UpdateContext();
} }
@ -753,7 +753,7 @@ void GSState::GIFRegHandlerPRIM(const GIFReg* RESTRICT r)
{ {
ALIGN_STACK(32); ALIGN_STACK(32);
ApplyPRIM(r->PRIM.u32[0]); ApplyPRIM(r->PRIM.U32[0]);
} }
void GSState::GIFRegHandlerRGBAQ(const GIFReg* RESTRICT r) void GSState::GIFRegHandlerRGBAQ(const GIFReg* RESTRICT r)
@ -782,17 +782,17 @@ void GSState::GIFRegHandlerST(const GIFReg* RESTRICT r)
void GSState::GIFRegHandlerUV(const GIFReg* RESTRICT r) void GSState::GIFRegHandlerUV(const GIFReg* RESTRICT r)
{ {
m_v.UV = r->UV.u32[0] & 0x3fff3fff; m_v.UV = r->UV.U32[0] & 0x3fff3fff;
} }
void GSState::GIFRegHandlerUV_Hack(const GIFReg* RESTRICT r) void GSState::GIFRegHandlerUV_Hack(const GIFReg* RESTRICT r)
{ {
m_v.UV = r->UV.u32[0] & 0x3fff3fff; m_v.UV = r->UV.U32[0] & 0x3fff3fff;
m_isPackedUV_HackFlag = false; m_isPackedUV_HackFlag = false;
} }
template <uint32 prim, uint32 adc, bool auto_flush> template <u32 prim, u32 adc, bool auto_flush>
void GSState::GIFRegHandlerXYZF2(const GIFReg* RESTRICT r) void GSState::GIFRegHandlerXYZF2(const GIFReg* RESTRICT r)
{ {
GSVector4i xyzf = GSVector4i::loadl(&r->XYZF); GSVector4i xyzf = GSVector4i::loadl(&r->XYZF);
@ -804,7 +804,7 @@ void GSState::GIFRegHandlerXYZF2(const GIFReg* RESTRICT r)
VertexKick<prim, auto_flush>(adc); VertexKick<prim, auto_flush>(adc);
} }
template <uint32 prim, uint32 adc, bool auto_flush> template <u32 prim, u32 adc, bool auto_flush>
void GSState::GIFRegHandlerXYZ2(const GIFReg* RESTRICT r) void GSState::GIFRegHandlerXYZ2(const GIFReg* RESTRICT r)
{ {
m_v.m[1] = GSVector4i::load(&r->XYZ, &m_v.UV); m_v.m[1] = GSVector4i::load(&r->XYZ, &m_v.UV);
@ -828,21 +828,21 @@ void GSState::ApplyTEX0(GIFRegTEX0& TEX0)
// extremely broken for the same reasons as MLB Power Pros in that it spams TEX0 with // extremely broken for the same reasons as MLB Power Pros in that it spams TEX0 with
// complete garbage making for a nice 1G heap of GSOffset. // complete garbage making for a nice 1G heap of GSOffset.
GL_REG("Apply TEX0_%d = 0x%x_%x", i, TEX0.u32[1], TEX0.u32[0]); GL_REG("Apply TEX0_%d = 0x%x_%x", i, TEX0.U32[1], TEX0.U32[0]);
// even if TEX0 did not change, a new palette may have been uploaded and will overwrite the currently queued for drawing // even if TEX0 did not change, a new palette may have been uploaded and will overwrite the currently queued for drawing
const bool wt = m_mem.m_clut.WriteTest(TEX0, m_env.TEXCLUT); const bool wt = m_mem.m_clut.WriteTest(TEX0, m_env.TEXCLUT);
// clut loading already covered with WriteTest, for drawing only have to check CPSM and CSA (MGS3 intro skybox would be drawn piece by piece without this) // clut loading already covered with WriteTest, for drawing only have to check CPSM and CSA (MGS3 intro skybox would be drawn piece by piece without this)
constexpr uint64 mask = 0x1f78001fffffffffull; // TBP0 TBW PSM TW TH TCC TFX CPSM CSA constexpr u64 mask = 0x1f78001fffffffffull; // TBP0 TBW PSM TW TH TCC TFX CPSM CSA
if (wt || PRIM->CTXT == i && ((TEX0.u64 ^ m_env.CTXT[i].TEX0.u64) & mask)) if (wt || PRIM->CTXT == i && ((TEX0.U64 ^ m_env.CTXT[i].TEX0.U64) & mask))
Flush(); Flush();
TEX0.CPSM &= 0xa; // 1010b TEX0.CPSM &= 0xa; // 1010b
if ((TEX0.u32[0] ^ m_env.CTXT[i].TEX0.u32[0]) & 0x3ffffff) // TBP0 TBW PSM if ((TEX0.U32[0] ^ m_env.CTXT[i].TEX0.U32[0]) & 0x3ffffff) // TBP0 TBW PSM
m_env.CTXT[i].offset.tex = m_mem.GetOffset(TEX0.TBP0, TEX0.TBW, TEX0.PSM); m_env.CTXT[i].offset.tex = m_mem.GetOffset(TEX0.TBP0, TEX0.TBW, TEX0.PSM);
m_env.CTXT[i].TEX0 = (GSVector4i)TEX0; m_env.CTXT[i].TEX0 = (GSVector4i)TEX0;
@ -895,7 +895,7 @@ void GSState::ApplyTEX0(GIFRegTEX0& TEX0)
template <int i> template <int i>
void GSState::GIFRegHandlerTEX0(const GIFReg* RESTRICT r) void GSState::GIFRegHandlerTEX0(const GIFReg* RESTRICT r)
{ {
GL_REG("TEX0_%d = 0x%x_%x", i, r->u32[1], r->u32[0]); GL_REG("TEX0_%d = 0x%x_%x", i, r->U32[1], r->U32[0]);
GIFRegTEX0 TEX0 = r->TEX0; GIFRegTEX0 TEX0 = r->TEX0;
@ -910,8 +910,8 @@ void GSState::GIFRegHandlerTEX0(const GIFReg* RESTRICT r)
// Sets TW/TH to 0 // Sets TW/TH to 0
// there used to be a case to force this to 10 // there used to be a case to force this to 10
// but GetSizeFixedTEX0 sorts this now // but GetSizeFixedTEX0 sorts this now
TEX0.TW = std::clamp<uint32>(TEX0.TW, 0, 10); TEX0.TW = std::clamp<u32>(TEX0.TW, 0, 10);
TEX0.TH = std::clamp<uint32>(TEX0.TH, 0, 10); TEX0.TH = std::clamp<u32>(TEX0.TH, 0, 10);
ApplyTEX0<i>(TEX0); ApplyTEX0<i>(TEX0);
@ -921,36 +921,36 @@ void GSState::GIFRegHandlerTEX0(const GIFReg* RESTRICT r)
// NOTE 2: Mipmap levels are tightly packed, if (tbw << 6) > (1 << tw) then the left-over space to the right is used. (common for PSM_PSMT4) // NOTE 2: Mipmap levels are tightly packed, if (tbw << 6) > (1 << tw) then the left-over space to the right is used. (common for PSM_PSMT4)
// NOTE 3: Non-rectangular textures are treated as rectangular when calculating the occupied space (height is extended, not sure about width) // NOTE 3: Non-rectangular textures are treated as rectangular when calculating the occupied space (height is extended, not sure about width)
uint32 bp = TEX0.TBP0; u32 bp = TEX0.TBP0;
uint32 bw = TEX0.TBW; u32 bw = TEX0.TBW;
uint32 w = 1u << TEX0.TW; u32 w = 1u << TEX0.TW;
uint32 h = 1u << TEX0.TH; u32 h = 1u << TEX0.TH;
const uint32 bpp = GSLocalMemory::m_psm[TEX0.PSM].bpp; const u32 bpp = GSLocalMemory::m_psm[TEX0.PSM].bpp;
if (h < w) if (h < w)
h = w; h = w;
bp += ((w * h * bpp >> 3) + 255) >> 8; bp += ((w * h * bpp >> 3) + 255) >> 8;
bw = std::max<uint32>(bw >> 1, 1); bw = std::max<u32>(bw >> 1, 1);
w = std::max<uint32>(w >> 1, 1); w = std::max<u32>(w >> 1, 1);
h = std::max<uint32>(h >> 1, 1); h = std::max<u32>(h >> 1, 1);
m_env.CTXT[i].MIPTBP1.TBP1 = bp; m_env.CTXT[i].MIPTBP1.TBP1 = bp;
m_env.CTXT[i].MIPTBP1.TBW1 = bw; m_env.CTXT[i].MIPTBP1.TBW1 = bw;
bp += ((w * h * bpp >> 3) + 255) >> 8; bp += ((w * h * bpp >> 3) + 255) >> 8;
bw = std::max<uint32>(bw >> 1, 1); bw = std::max<u32>(bw >> 1, 1);
w = std::max<uint32>(w >> 1, 1); w = std::max<u32>(w >> 1, 1);
h = std::max<uint32>(h >> 1, 1); h = std::max<u32>(h >> 1, 1);
m_env.CTXT[i].MIPTBP1.TBP2 = bp; m_env.CTXT[i].MIPTBP1.TBP2 = bp;
m_env.CTXT[i].MIPTBP1.TBW2 = bw; m_env.CTXT[i].MIPTBP1.TBW2 = bw;
bp += ((w * h * bpp >> 3) + 255) >> 8; bp += ((w * h * bpp >> 3) + 255) >> 8;
bw = std::max<uint32>(bw >> 1, 1); bw = std::max<u32>(bw >> 1, 1);
w = std::max<uint32>(w >> 1, 1); w = std::max<u32>(w >> 1, 1);
h = std::max<uint32>(h >> 1, 1); h = std::max<u32>(h >> 1, 1);
m_env.CTXT[i].MIPTBP1.TBP3 = bp; m_env.CTXT[i].MIPTBP1.TBP3 = bp;
m_env.CTXT[i].MIPTBP1.TBW3 = bw; m_env.CTXT[i].MIPTBP1.TBW3 = bw;
@ -960,7 +960,7 @@ void GSState::GIFRegHandlerTEX0(const GIFReg* RESTRICT r)
template <int i> template <int i>
void GSState::GIFRegHandlerCLAMP(const GIFReg* RESTRICT r) void GSState::GIFRegHandlerCLAMP(const GIFReg* RESTRICT r)
{ {
GL_REG("CLAMP_%d = 0x%x_%x", i, r->u32[1], r->u32[0]); GL_REG("CLAMP_%d = 0x%x_%x", i, r->U32[1], r->U32[0]);
if (PRIM->CTXT == i && r->CLAMP != m_env.CTXT[i].CLAMP) if (PRIM->CTXT == i && r->CLAMP != m_env.CTXT[i].CLAMP)
Flush(); Flush();
@ -980,7 +980,7 @@ void GSState::GIFRegHandlerNOP(const GIFReg* RESTRICT r)
template <int i> template <int i>
void GSState::GIFRegHandlerTEX1(const GIFReg* RESTRICT r) void GSState::GIFRegHandlerTEX1(const GIFReg* RESTRICT r)
{ {
GL_REG("TEX1_%d = 0x%x_%x", i, r->u32[1], r->u32[0]); GL_REG("TEX1_%d = 0x%x_%x", i, r->U32[1], r->U32[0]);
if (PRIM->CTXT == i && r->TEX1 != m_env.CTXT[i].TEX1) if (PRIM->CTXT == i && r->TEX1 != m_env.CTXT[i].TEX1)
Flush(); Flush();
@ -991,7 +991,7 @@ void GSState::GIFRegHandlerTEX1(const GIFReg* RESTRICT r)
template <int i> template <int i>
void GSState::GIFRegHandlerTEX2(const GIFReg* RESTRICT r) void GSState::GIFRegHandlerTEX2(const GIFReg* RESTRICT r)
{ {
GL_REG("TEX2_%d = 0x%x_%x", i, r->u32[1], r->u32[0]); GL_REG("TEX2_%d = 0x%x_%x", i, r->U32[1], r->U32[0]);
// TEX2 is a masked write to TEX0, for performing CLUT swaps (palette swaps). // TEX2 is a masked write to TEX0, for performing CLUT swaps (palette swaps).
// It only applies the following fields: // It only applies the following fields:
@ -999,11 +999,11 @@ void GSState::GIFRegHandlerTEX2(const GIFReg* RESTRICT r)
// It ignores these fields (uses existing values in the context): // It ignores these fields (uses existing values in the context):
// TFX, TCC, TH, TW, TBW, and TBP0 // TFX, TCC, TH, TW, TBW, and TBP0
constexpr uint64 mask = 0xFFFFFFE003F00000ull; // TEX2 bits constexpr u64 mask = 0xFFFFFFE003F00000ull; // TEX2 bits
GIFRegTEX0 TEX0; GIFRegTEX0 TEX0;
TEX0.u64 = (m_env.CTXT[i].TEX0.u64 & ~mask) | (r->u64 & mask); TEX0.U64 = (m_env.CTXT[i].TEX0.U64 & ~mask) | (r->U64 & mask);
ApplyTEX0<i>(TEX0); ApplyTEX0<i>(TEX0);
} }
@ -1011,7 +1011,7 @@ void GSState::GIFRegHandlerTEX2(const GIFReg* RESTRICT r)
template <int i> template <int i>
void GSState::GIFRegHandlerXYOFFSET(const GIFReg* RESTRICT r) void GSState::GIFRegHandlerXYOFFSET(const GIFReg* RESTRICT r)
{ {
GL_REG("XYOFFSET_%d = 0x%x_%x", i, r->u32[1], r->u32[0]); GL_REG("XYOFFSET_%d = 0x%x_%x", i, r->U32[1], r->U32[0]);
const GSVector4i o = (GSVector4i)r->XYOFFSET & GSVector4i::x0000ffff(); const GSVector4i o = (GSVector4i)r->XYOFFSET & GSVector4i::x0000ffff();
@ -1027,18 +1027,18 @@ void GSState::GIFRegHandlerXYOFFSET(const GIFReg* RESTRICT r)
void GSState::GIFRegHandlerPRMODECONT(const GIFReg* RESTRICT r) void GSState::GIFRegHandlerPRMODECONT(const GIFReg* RESTRICT r)
{ {
GL_REG("PRMODECONT = 0x%x_%x", r->u32[1], r->u32[0]); GL_REG("PRMODECONT = 0x%x_%x", r->U32[1], r->U32[0]);
m_env.PRMODECONT.AC = r->PRMODECONT.AC; m_env.PRMODECONT.AC = r->PRMODECONT.AC;
} }
void GSState::GIFRegHandlerPRMODE(const GIFReg* RESTRICT r) void GSState::GIFRegHandlerPRMODE(const GIFReg* RESTRICT r)
{ {
GL_REG("PRMODE = 0x%x_%x", r->u32[1], r->u32[0]); GL_REG("PRMODE = 0x%x_%x", r->U32[1], r->U32[0]);
if (!m_env.PRMODECONT.AC) if (!m_env.PRMODECONT.AC)
{ {
if ((m_env.PRIM.u32[0] ^ r->PRMODE.u32[0]) & 0x7f8) if ((m_env.PRIM.U32[0] ^ r->PRMODE.U32[0]) & 0x7f8)
Flush(); Flush();
} }
else else
@ -1046,7 +1046,7 @@ void GSState::GIFRegHandlerPRMODE(const GIFReg* RESTRICT r)
return; return;
} }
const uint32 _PRIM = m_env.PRIM.PRIM; const u32 _PRIM = m_env.PRIM.PRIM;
m_env.PRIM = (GSVector4i)r->PRMODE; m_env.PRIM = (GSVector4i)r->PRMODE;
m_env.PRIM.PRIM = _PRIM; m_env.PRIM.PRIM = _PRIM;
@ -1055,7 +1055,7 @@ void GSState::GIFRegHandlerPRMODE(const GIFReg* RESTRICT r)
void GSState::GIFRegHandlerTEXCLUT(const GIFReg* RESTRICT r) void GSState::GIFRegHandlerTEXCLUT(const GIFReg* RESTRICT r)
{ {
GL_REG("TEXCLUT = 0x%x_%x", r->u32[1], r->u32[0]); GL_REG("TEXCLUT = 0x%x_%x", r->U32[1], r->U32[0]);
if (r->TEXCLUT != m_env.TEXCLUT) if (r->TEXCLUT != m_env.TEXCLUT)
Flush(); Flush();
@ -1074,7 +1074,7 @@ void GSState::GIFRegHandlerSCANMSK(const GIFReg* RESTRICT r)
template <int i> template <int i>
void GSState::GIFRegHandlerMIPTBP1(const GIFReg* RESTRICT r) void GSState::GIFRegHandlerMIPTBP1(const GIFReg* RESTRICT r)
{ {
GL_REG("MIPTBP1_%d = 0x%x_%x", i, r->u32[1], r->u32[0]); GL_REG("MIPTBP1_%d = 0x%x_%x", i, r->U32[1], r->U32[0]);
if (PRIM->CTXT == i && r->MIPTBP1 != m_env.CTXT[i].MIPTBP1) if (PRIM->CTXT == i && r->MIPTBP1 != m_env.CTXT[i].MIPTBP1)
Flush(); Flush();
@ -1085,7 +1085,7 @@ void GSState::GIFRegHandlerMIPTBP1(const GIFReg* RESTRICT r)
template <int i> template <int i>
void GSState::GIFRegHandlerMIPTBP2(const GIFReg* RESTRICT r) void GSState::GIFRegHandlerMIPTBP2(const GIFReg* RESTRICT r)
{ {
GL_REG("MIPTBP2_%d = 0x%x_%x", i, r->u32[1], r->u32[0]); GL_REG("MIPTBP2_%d = 0x%x_%x", i, r->U32[1], r->U32[0]);
if (PRIM->CTXT == i && r->MIPTBP2 != m_env.CTXT[i].MIPTBP2) if (PRIM->CTXT == i && r->MIPTBP2 != m_env.CTXT[i].MIPTBP2)
Flush(); Flush();
@ -1095,7 +1095,7 @@ void GSState::GIFRegHandlerMIPTBP2(const GIFReg* RESTRICT r)
void GSState::GIFRegHandlerTEXA(const GIFReg* RESTRICT r) void GSState::GIFRegHandlerTEXA(const GIFReg* RESTRICT r)
{ {
GL_REG("TEXA = 0x%x_%x", r->u32[1], r->u32[0]); GL_REG("TEXA = 0x%x_%x", r->U32[1], r->U32[0]);
if (r->TEXA != m_env.TEXA) if (r->TEXA != m_env.TEXA)
Flush(); Flush();
@ -1104,7 +1104,7 @@ void GSState::GIFRegHandlerTEXA(const GIFReg* RESTRICT r)
void GSState::GIFRegHandlerFOGCOL(const GIFReg* RESTRICT r) void GSState::GIFRegHandlerFOGCOL(const GIFReg* RESTRICT r)
{ {
GL_REG("FOGCOL = 0x%x_%x", r->u32[1], r->u32[0]); GL_REG("FOGCOL = 0x%x_%x", r->U32[1], r->U32[0]);
if (r->FOGCOL != m_env.FOGCOL) if (r->FOGCOL != m_env.FOGCOL)
Flush(); Flush();
@ -1114,7 +1114,7 @@ void GSState::GIFRegHandlerFOGCOL(const GIFReg* RESTRICT r)
void GSState::GIFRegHandlerTEXFLUSH(const GIFReg* RESTRICT r) void GSState::GIFRegHandlerTEXFLUSH(const GIFReg* RESTRICT r)
{ {
GL_REG("TEXFLUSH = 0x%x_%x", r->u32[1], r->u32[0]); GL_REG("TEXFLUSH = 0x%x_%x", r->U32[1], r->U32[0]);
} }
template <int i> template <int i>
@ -1133,8 +1133,7 @@ void GSState::GIFRegHandlerSCISSOR(const GIFReg* RESTRICT r)
template <int i> template <int i>
void GSState::GIFRegHandlerALPHA(const GIFReg* RESTRICT r) void GSState::GIFRegHandlerALPHA(const GIFReg* RESTRICT r)
{ {
GL_REG("ALPHA = 0x%x_%x", r->u32[1], r->u32[0]); GL_REG("ALPHA = 0x%x_%x", r->U32[1], r->U32[0]);
if (PRIM->CTXT == i && r->ALPHA != m_env.CTXT[i].ALPHA) if (PRIM->CTXT == i && r->ALPHA != m_env.CTXT[i].ALPHA)
Flush(); Flush();
@ -1142,10 +1141,10 @@ void GSState::GIFRegHandlerALPHA(const GIFReg* RESTRICT r)
// value of 4 is not allowed by the spec // value of 4 is not allowed by the spec
// acts has 3 on real hw, so just clamp it // acts has 3 on real hw, so just clamp it
m_env.CTXT[i].ALPHA.A = std::clamp<uint32>(r->ALPHA.A, 0, 3); m_env.CTXT[i].ALPHA.A = std::clamp<u32>(r->ALPHA.A, 0, 3);
m_env.CTXT[i].ALPHA.B = std::clamp<uint32>(r->ALPHA.B, 0, 3); m_env.CTXT[i].ALPHA.B = std::clamp<u32>(r->ALPHA.B, 0, 3);
m_env.CTXT[i].ALPHA.C = std::clamp<uint32>(r->ALPHA.C, 0, 3); m_env.CTXT[i].ALPHA.C = std::clamp<u32>(r->ALPHA.C, 0, 3);
m_env.CTXT[i].ALPHA.D = std::clamp<uint32>(r->ALPHA.D, 0, 3); m_env.CTXT[i].ALPHA.D = std::clamp<u32>(r->ALPHA.D, 0, 3);
} }
void GSState::GIFRegHandlerDIMX(const GIFReg* RESTRICT r) void GSState::GIFRegHandlerDIMX(const GIFReg* RESTRICT r)
@ -1210,12 +1209,12 @@ void GSState::GIFRegHandlerFBA(const GIFReg* RESTRICT r)
template <int i> template <int i>
void GSState::GIFRegHandlerFRAME(const GIFReg* RESTRICT r) void GSState::GIFRegHandlerFRAME(const GIFReg* RESTRICT r)
{ {
GL_REG("FRAME_%d = 0x%x_%x", i, r->u32[1], r->u32[0]); GL_REG("FRAME_%d = 0x%x_%x", i, r->U32[1], r->U32[0]);
if (PRIM->CTXT == i && r->FRAME != m_env.CTXT[i].FRAME) if (PRIM->CTXT == i && r->FRAME != m_env.CTXT[i].FRAME)
Flush(); Flush();
if ((m_env.CTXT[i].FRAME.u32[0] ^ r->FRAME.u32[0]) & 0x3f3f01ff) // FBP FBW PSM if ((m_env.CTXT[i].FRAME.U32[0] ^ r->FRAME.U32[0]) & 0x3f3f01ff) // FBP FBW PSM
{ {
m_env.CTXT[i].offset.fb = m_mem.GetOffset(r->FRAME.Block(), r->FRAME.FBW, r->FRAME.PSM); m_env.CTXT[i].offset.fb = m_mem.GetOffset(r->FRAME.Block(), r->FRAME.FBW, r->FRAME.PSM);
m_env.CTXT[i].offset.zb = m_mem.GetOffset(m_env.CTXT[i].ZBUF.Block(), r->FRAME.FBW, m_env.CTXT[i].ZBUF.PSM); m_env.CTXT[i].offset.zb = m_mem.GetOffset(m_env.CTXT[i].ZBUF.Block(), r->FRAME.FBW, m_env.CTXT[i].ZBUF.PSM);
@ -1251,7 +1250,7 @@ void GSState::GIFRegHandlerFRAME(const GIFReg* RESTRICT r)
template <int i> template <int i>
void GSState::GIFRegHandlerZBUF(const GIFReg* RESTRICT r) void GSState::GIFRegHandlerZBUF(const GIFReg* RESTRICT r)
{ {
GL_REG("ZBUF_%d = 0x%x_%x", i, r->u32[1], r->u32[0]); GL_REG("ZBUF_%d = 0x%x_%x", i, r->U32[1], r->U32[0]);
GIFRegZBUF ZBUF = r->ZBUF; GIFRegZBUF ZBUF = r->ZBUF;
@ -1264,7 +1263,7 @@ void GSState::GIFRegHandlerZBUF(const GIFReg* RESTRICT r)
if (PRIM->CTXT == i && ZBUF != m_env.CTXT[i].ZBUF) if (PRIM->CTXT == i && ZBUF != m_env.CTXT[i].ZBUF)
Flush(); Flush();
if ((m_env.CTXT[i].ZBUF.u32[0] ^ ZBUF.u32[0]) & 0x3f0001ff) // ZBP PSM if ((m_env.CTXT[i].ZBUF.U32[0] ^ ZBUF.U32[0]) & 0x3f0001ff) // ZBP PSM
{ {
m_env.CTXT[i].offset.zb = m_mem.GetOffset(ZBUF.Block(), m_env.CTXT[i].FRAME.FBW, ZBUF.PSM); m_env.CTXT[i].offset.zb = m_mem.GetOffset(ZBUF.Block(), m_env.CTXT[i].FRAME.FBW, ZBUF.PSM);
m_env.CTXT[i].offset.fzb = m_mem.GetPixelOffset(m_env.CTXT[i].FRAME, ZBUF); m_env.CTXT[i].offset.fzb = m_mem.GetPixelOffset(m_env.CTXT[i].FRAME, ZBUF);
@ -1286,7 +1285,7 @@ void GSState::GIFRegHandlerBITBLTBUF(const GIFReg* RESTRICT r)
// documentation on this problem, nothing in the game to suggest // documentation on this problem, nothing in the game to suggest
// it is broken and the code here for it was likely incorrect to begin with. // it is broken and the code here for it was likely incorrect to begin with.
GL_REG("BITBLTBUF = 0x%x_%x", r->u32[1], r->u32[0]); GL_REG("BITBLTBUF = 0x%x_%x", r->U32[1], r->U32[0]);
if (r->BITBLTBUF != m_env.BITBLTBUF) if (r->BITBLTBUF != m_env.BITBLTBUF)
FlushWrite(); FlushWrite();
@ -1296,7 +1295,7 @@ void GSState::GIFRegHandlerBITBLTBUF(const GIFReg* RESTRICT r)
void GSState::GIFRegHandlerTRXPOS(const GIFReg* RESTRICT r) void GSState::GIFRegHandlerTRXPOS(const GIFReg* RESTRICT r)
{ {
GL_REG("TRXPOS = 0x%x_%x", r->u32[1], r->u32[0]); GL_REG("TRXPOS = 0x%x_%x", r->U32[1], r->U32[0]);
if (r->TRXPOS != m_env.TRXPOS) if (r->TRXPOS != m_env.TRXPOS)
FlushWrite(); FlushWrite();
@ -1306,8 +1305,7 @@ void GSState::GIFRegHandlerTRXPOS(const GIFReg* RESTRICT r)
void GSState::GIFRegHandlerTRXREG(const GIFReg* RESTRICT r) void GSState::GIFRegHandlerTRXREG(const GIFReg* RESTRICT r)
{ {
GL_REG("TRXREG = 0x%x_%x", r->u32[1], r->u32[0]); GL_REG("TRXREG = 0x%x_%x", r->U32[1], r->U32[0]);
if (r->TRXREG != m_env.TRXREG) if (r->TRXREG != m_env.TRXREG)
FlushWrite(); FlushWrite();
@ -1316,7 +1314,7 @@ void GSState::GIFRegHandlerTRXREG(const GIFReg* RESTRICT r)
void GSState::GIFRegHandlerTRXDIR(const GIFReg* RESTRICT r) void GSState::GIFRegHandlerTRXDIR(const GIFReg* RESTRICT r)
{ {
GL_REG("TRXDIR = 0x%x_%x", r->u32[1], r->u32[0]); GL_REG("TRXDIR = 0x%x_%x", r->U32[1], r->U32[0]);
Flush(); Flush();
@ -1341,14 +1339,14 @@ void GSState::GIFRegHandlerTRXDIR(const GIFReg* RESTRICT r)
void GSState::GIFRegHandlerHWREG(const GIFReg* RESTRICT r) void GSState::GIFRegHandlerHWREG(const GIFReg* RESTRICT r)
{ {
GL_REG("HWREG = 0x%x_%x", r->u32[1], r->u32[0]); GL_REG("HWREG = 0x%x_%x", r->U32[1], r->U32[0]);
// don't bother if not host -> local // don't bother if not host -> local
// real hw ignores // real hw ignores
if (m_env.TRXDIR.XDIR != 0) if (m_env.TRXDIR.XDIR != 0)
return; return;
Write(reinterpret_cast<const uint8*>(r), 8); // haunting ground Write(reinterpret_cast<const u8*>(r), 8); // haunting ground
} }
void GSState::Flush() void GSState::Flush()
@ -1482,7 +1480,7 @@ void GSState::FlushPrim()
} }
} }
void GSState::Write(const uint8* mem, int len) void GSState::Write(const u8* mem, int len)
{ {
int w = m_env.TRXREG.RRW; int w = m_env.TRXREG.RRW;
int h = m_env.TRXREG.RRH; int h = m_env.TRXREG.RRH;
@ -1555,7 +1553,7 @@ void GSState::Write(const uint8* mem, int len)
m_mem.m_clut.Invalidate(); m_mem.m_clut.Invalidate();
} }
void GSState::InitReadFIFO(uint8* mem, int len) void GSState::InitReadFIFO(u8* mem, int len)
{ {
if (len <= 0) if (len <= 0)
return; return;
@ -1565,7 +1563,7 @@ void GSState::InitReadFIFO(uint8* mem, int len)
const int w = m_env.TRXREG.RRW; const int w = m_env.TRXREG.RRW;
const int h = m_env.TRXREG.RRH; const int h = m_env.TRXREG.RRH;
const uint16 bpp = GSLocalMemory::m_psm[m_env.BITBLTBUF.SPSM].trbpp; const u16 bpp = GSLocalMemory::m_psm[m_env.BITBLTBUF.SPSM].trbpp;
if (!m_tr.Update(w, h, bpp, len)) if (!m_tr.Update(w, h, bpp, len))
return; return;
@ -1575,7 +1573,7 @@ void GSState::InitReadFIFO(uint8* mem, int len)
} }
// NOTE: called from outside MTGS // NOTE: called from outside MTGS
void GSState::Read(uint8* mem, int len) void GSState::Read(u8* mem, int len)
{ {
if (len <= 0) if (len <= 0)
return; return;
@ -1587,7 +1585,7 @@ void GSState::Read(uint8* mem, int len)
const GSVector4i r(sx, sy, sx + w, sy + h); const GSVector4i r(sx, sy, sx + w, sy + h);
const uint16 bpp = GSLocalMemory::m_psm[m_env.BITBLTBUF.SPSM].trbpp; const u16 bpp = GSLocalMemory::m_psm[m_env.BITBLTBUF.SPSM].trbpp;
if (!m_tr.Update(w, h, bpp, len)) if (!m_tr.Update(w, h, bpp, len))
return; return;
@ -1710,21 +1708,21 @@ void GSState::Move()
{ {
if (spsm.trbpp == 32) if (spsm.trbpp == 32)
{ {
copyFast(m_mem.m_vm32, dpo.assertSizesMatch(GSLocalMemory::swizzle32), spo.assertSizesMatch(GSLocalMemory::swizzle32), [](uint32* d, uint32* s) copyFast(m_mem.m_vm32, dpo.assertSizesMatch(GSLocalMemory::swizzle32), spo.assertSizesMatch(GSLocalMemory::swizzle32), [](u32* d, u32* s)
{ {
*d = *s; *d = *s;
}); });
} }
else if (spsm.trbpp == 24) else if (spsm.trbpp == 24)
{ {
copyFast(m_mem.m_vm32, dpo.assertSizesMatch(GSLocalMemory::swizzle32), spo.assertSizesMatch(GSLocalMemory::swizzle32), [](uint32* d, uint32* s) copyFast(m_mem.m_vm32, dpo.assertSizesMatch(GSLocalMemory::swizzle32), spo.assertSizesMatch(GSLocalMemory::swizzle32), [](u32* d, u32* s)
{ {
*d = (*d & 0xff000000) | (*s & 0x00ffffff); *d = (*d & 0xff000000) | (*s & 0x00ffffff);
}); });
} }
else // if(spsm.trbpp == 16) else // if(spsm.trbpp == 16)
{ {
copyFast(m_mem.m_vm16, dpo.assertSizesMatch(GSLocalMemory::swizzle16), spo.assertSizesMatch(GSLocalMemory::swizzle16), [](uint16* d, uint16* s) copyFast(m_mem.m_vm16, dpo.assertSizesMatch(GSLocalMemory::swizzle16), spo.assertSizesMatch(GSLocalMemory::swizzle16), [](u16* d, u16* s)
{ {
*d = *s; *d = *s;
}); });
@ -1732,28 +1730,28 @@ void GSState::Move()
} }
else if (m_env.BITBLTBUF.SPSM == PSM_PSMT8 && m_env.BITBLTBUF.DPSM == PSM_PSMT8) else if (m_env.BITBLTBUF.SPSM == PSM_PSMT8 && m_env.BITBLTBUF.DPSM == PSM_PSMT8)
{ {
copyFast(m_mem.m_vm8, GSOffset::fromKnownPSM(dbp, dbw, PSM_PSMT8), GSOffset::fromKnownPSM(sbp, sbw, PSM_PSMT8), [](uint8* d, uint8* s) copyFast(m_mem.m_vm8, GSOffset::fromKnownPSM(dbp, dbw, PSM_PSMT8), GSOffset::fromKnownPSM(sbp, sbw, PSM_PSMT8), [](u8* d, u8* s)
{ {
*d = *s; *d = *s;
}); });
} }
else if (m_env.BITBLTBUF.SPSM == PSM_PSMT4 && m_env.BITBLTBUF.DPSM == PSM_PSMT4) else if (m_env.BITBLTBUF.SPSM == PSM_PSMT4 && m_env.BITBLTBUF.DPSM == PSM_PSMT4)
{ {
copy(GSOffset::fromKnownPSM(dbp, dbw, PSM_PSMT4), GSOffset::fromKnownPSM(sbp, sbw, PSM_PSMT4), [&](uint32 doff, uint32 soff) copy(GSOffset::fromKnownPSM(dbp, dbw, PSM_PSMT4), GSOffset::fromKnownPSM(sbp, sbw, PSM_PSMT4), [&](u32 doff, u32 soff)
{ {
m_mem.WritePixel4(doff, m_mem.ReadPixel4(soff)); m_mem.WritePixel4(doff, m_mem.ReadPixel4(soff));
}); });
} }
else else
{ {
copy(dpo, spo, [&](uint32 doff, uint32 soff) copy(dpo, spo, [&](u32 doff, u32 soff)
{ {
(m_mem.*dpsm.wpa)(doff, (m_mem.*spsm.rpa)(soff)); (m_mem.*dpsm.wpa)(doff, (m_mem.*spsm.rpa)(soff));
}); });
} }
} }
void GSState::SoftReset(uint32 mask) void GSState::SoftReset(u32 mask)
{ {
if (mask & 1) if (mask & 1)
{ {
@ -1772,7 +1770,7 @@ void GSState::SoftReset(uint32 mask)
m_q = 1.0f; m_q = 1.0f;
} }
void GSState::ReadFIFO(uint8* mem, int size) void GSState::ReadFIFO(u8* mem, int size)
{ {
GSPerfMonAutoTimer pmat(&m_perfmon); GSPerfMonAutoTimer pmat(&m_perfmon);
@ -1786,17 +1784,17 @@ void GSState::ReadFIFO(uint8* mem, int size)
m_dump->ReadFIFO(size); m_dump->ReadFIFO(size);
} }
template void GSState::Transfer<0>(const uint8* mem, uint32 size); template void GSState::Transfer<0>(const u8* mem, u32 size);
template void GSState::Transfer<1>(const uint8* mem, uint32 size); template void GSState::Transfer<1>(const u8* mem, u32 size);
template void GSState::Transfer<2>(const uint8* mem, uint32 size); template void GSState::Transfer<2>(const u8* mem, u32 size);
template void GSState::Transfer<3>(const uint8* mem, uint32 size); template void GSState::Transfer<3>(const u8* mem, u32 size);
template <int index> template <int index>
void GSState::Transfer(const uint8* mem, uint32 size) void GSState::Transfer(const u8* mem, u32 size)
{ {
GSPerfMonAutoTimer pmat(&m_perfmon); GSPerfMonAutoTimer pmat(&m_perfmon);
const uint8* start = mem; const u8* start = mem;
GIFPath& path = m_path[index]; GIFPath& path = m_path[index];
@ -1823,7 +1821,7 @@ void GSState::Transfer(const uint8* mem, uint32 size)
} }
else else
{ {
uint32 total; u32 total;
switch (path.tag.FLG) switch (path.tag.FLG)
{ {
@ -1852,7 +1850,7 @@ void GSState::Transfer(const uint8* mem, uint32 size)
{ {
case GIFPath::TYPE_UNKNOWN: case GIFPath::TYPE_UNKNOWN:
{ {
uint32 reg = 0; u32 reg = 0;
do do
{ {
@ -2058,9 +2056,9 @@ int GSState::Freeze(freezeData* fd, bool sizeonly)
path.tag.NLOOP = path.nloop; path.tag.NLOOP = path.nloop;
path.tag.REGS = 0; path.tag.REGS = 0;
for (size_t j = 0; j < std::size(path.regs.u8); j++) for (size_t j = 0; j < std::size(path.regs.U8); j++)
{ {
path.tag.u32[2 + (j >> 3)] |= path.regs.u8[j] << ((j & 7) << 2); path.tag.U32[2 + (j >> 3)] |= path.regs.U8[j] << ((j & 7) << 2);
} }
WriteState(data, &path.tag); WriteState(data, &path.tag);
@ -2145,7 +2143,7 @@ int GSState::Defrost(const freezeData* fd)
m_env.CTXT[i].XYOFFSET.OFY &= 0xffff; m_env.CTXT[i].XYOFFSET.OFY &= 0xffff;
if (version <= 4) if (version <= 4)
data += sizeof(uint32) * 7; // skip data += sizeof(u32) * 7; // skip
} }
ReadState(&m_v.RGBAQ, data); ReadState(&m_v.RGBAQ, data);
@ -2196,7 +2194,7 @@ int GSState::Defrost(const freezeData* fd)
return 0; return 0;
} }
void GSState::SetGameCRC(uint32 crc, int options) void GSState::SetGameCRC(u32 crc, int options)
{ {
m_crc = crc; m_crc = crc;
m_options = options; m_options = options;
@ -2229,7 +2227,7 @@ void GSState::UpdateVertexKick()
if (m_frameskip) if (m_frameskip)
return; return;
const uint32 prim = PRIM->PRIM; const u32 prim = PRIM->PRIM;
m_fpGIFPackedRegHandlers[GIF_REG_XYZF2] = m_fpGIFPackedRegHandlerXYZ[prim][0]; m_fpGIFPackedRegHandlers[GIF_REG_XYZF2] = m_fpGIFPackedRegHandlerXYZ[prim][0];
m_fpGIFPackedRegHandlers[GIF_REG_XYZF3] = m_fpGIFPackedRegHandlerXYZ[prim][1]; m_fpGIFPackedRegHandlers[GIF_REG_XYZF3] = m_fpGIFPackedRegHandlerXYZ[prim][1];
@ -2250,12 +2248,12 @@ void GSState::GrowVertexBuffer()
const size_t maxcount = std::max<size_t>(m_vertex.maxcount * 3 / 2, 10000); const size_t maxcount = std::max<size_t>(m_vertex.maxcount * 3 / 2, 10000);
GSVertex* vertex = (GSVertex*)_aligned_malloc(sizeof(GSVertex) * maxcount, 32); GSVertex* vertex = (GSVertex*)_aligned_malloc(sizeof(GSVertex) * maxcount, 32);
uint32* index = (uint32*)_aligned_malloc(sizeof(uint32) * maxcount * 3, 32); // worst case is slightly less than vertex number * 3 u32* index = (u32*)_aligned_malloc(sizeof(u32) * maxcount * 3, 32); // worst case is slightly less than vertex number * 3
if (vertex == NULL || index == NULL) if (vertex == NULL || index == NULL)
{ {
const size_t vert_byte_count = sizeof(GSVertex) * maxcount; const size_t vert_byte_count = sizeof(GSVertex) * maxcount;
const size_t idx_byte_count = sizeof(uint32) * maxcount * 3; const size_t idx_byte_count = sizeof(u32) * maxcount * 3;
Console.Error("GS: failed to allocate %zu bytes for verticles and %zu for indices.", Console.Error("GS: failed to allocate %zu bytes for verticles and %zu for indices.",
vert_byte_count, idx_byte_count); vert_byte_count, idx_byte_count);
@ -2272,7 +2270,7 @@ void GSState::GrowVertexBuffer()
if (m_index.buff != NULL) if (m_index.buff != NULL)
{ {
memcpy(index, m_index.buff, sizeof(uint32) * m_index.tail); memcpy(index, m_index.buff, sizeof(u32) * m_index.tail);
_aligned_free(m_index.buff); _aligned_free(m_index.buff);
} }
@ -2282,8 +2280,8 @@ void GSState::GrowVertexBuffer()
m_index.buff = index; m_index.buff = index;
} }
template <uint32 prim, bool auto_flush> template <u32 prim, bool auto_flush>
__forceinline void GSState::VertexKick(uint32 skip) __forceinline void GSState::VertexKick(u32 skip)
{ {
ASSERT(m_vertex.tail < m_vertex.maxcount + 3); ASSERT(m_vertex.tail < m_vertex.maxcount + 3);
@ -2441,7 +2439,7 @@ __forceinline void GSState::VertexKick(uint32 skip)
if (tail >= m_vertex.maxcount) if (tail >= m_vertex.maxcount)
GrowVertexBuffer(); GrowVertexBuffer();
uint32* RESTRICT buff = &m_index.buff[m_index.tail]; u32* RESTRICT buff = &m_index.buff[m_index.tail];
switch (prim) switch (prim)
{ {
@ -2726,7 +2724,7 @@ void GSState::GetAlphaMinMax()
m_vt.m_alpha.valid = true; m_vt.m_alpha.valid = true;
} }
bool GSState::TryAlphaTest(uint32& fm, uint32& zm) bool GSState::TryAlphaTest(u32& fm, u32& zm)
{ {
// Shortcut for the easy case // Shortcut for the easy case
if (m_context->TEST.ATST == ATST_ALWAYS) if (m_context->TEST.ATST == ATST_ALWAYS)
@ -2901,7 +2899,7 @@ bool GSState::IsMipMapActive()
return m_mipmap && IsMipMapDraw(); return m_mipmap && IsMipMapDraw();
} }
GIFRegTEX0 GSState::GetTex0Layer(uint32 lod) GIFRegTEX0 GSState::GetTex0Layer(u32 lod)
{ {
// Shortcut // Shortcut
if (lod == 0) if (lod == 0)
@ -2962,7 +2960,7 @@ GSState::GSTransferBuffer::GSTransferBuffer()
start = end = total = 0; start = end = total = 0;
constexpr size_t alloc_size = 1024 * 1024 * 4; constexpr size_t alloc_size = 1024 * 1024 * 4;
buff = reinterpret_cast<uint8*>(_aligned_malloc(alloc_size, 32)); buff = reinterpret_cast<u8*>(_aligned_malloc(alloc_size, 32));
} }
GSState::GSTransferBuffer::~GSTransferBuffer() GSState::GSTransferBuffer::~GSTransferBuffer()

64
pcsx2/GS/GSState.h
View File

@ -31,12 +31,12 @@
struct GSFrameInfo struct GSFrameInfo
{ {
uint32 FBP; u32 FBP;
uint32 FPSM; u32 FPSM;
uint32 FBMSK; u32 FBMSK;
uint32 TBP0; u32 TBP0;
uint32 TPSM; u32 TPSM;
uint32 TZTST; u32 TZTST;
bool TME; bool TME;
}; };
@ -56,8 +56,8 @@ class GSState : public GSAlignedClass<32>
void GIFPackedRegHandlerSTQ(const GIFPackedReg* RESTRICT r); void GIFPackedRegHandlerSTQ(const GIFPackedReg* RESTRICT r);
void GIFPackedRegHandlerUV(const GIFPackedReg* RESTRICT r); void GIFPackedRegHandlerUV(const GIFPackedReg* RESTRICT r);
void GIFPackedRegHandlerUV_Hack(const GIFPackedReg* RESTRICT r); void GIFPackedRegHandlerUV_Hack(const GIFPackedReg* RESTRICT r);
template<uint32 prim, uint32 adc, bool auto_flush> void GIFPackedRegHandlerXYZF2(const GIFPackedReg* RESTRICT r); template<u32 prim, u32 adc, bool auto_flush> void GIFPackedRegHandlerXYZF2(const GIFPackedReg* RESTRICT r);
template<uint32 prim, uint32 adc, bool auto_flush> void GIFPackedRegHandlerXYZ2(const GIFPackedReg* RESTRICT r); template<u32 prim, u32 adc, bool auto_flush> void GIFPackedRegHandlerXYZ2(const GIFPackedReg* RESTRICT r);
void GIFPackedRegHandlerFOG(const GIFPackedReg* RESTRICT r); void GIFPackedRegHandlerFOG(const GIFPackedReg* RESTRICT r);
void GIFPackedRegHandlerA_D(const GIFPackedReg* RESTRICT r); void GIFPackedRegHandlerA_D(const GIFPackedReg* RESTRICT r);
void GIFPackedRegHandlerNOP(const GIFPackedReg* RESTRICT r); void GIFPackedRegHandlerNOP(const GIFPackedReg* RESTRICT r);
@ -67,18 +67,18 @@ class GSState : public GSAlignedClass<32>
GIFRegHandler m_fpGIFRegHandlers[256]; GIFRegHandler m_fpGIFRegHandlers[256];
GIFRegHandler m_fpGIFRegHandlerXYZ[8][4]; GIFRegHandler m_fpGIFRegHandlerXYZ[8][4];
typedef void (GSState::*GIFPackedRegHandlerC)(const GIFPackedReg* RESTRICT r, uint32 size); typedef void (GSState::*GIFPackedRegHandlerC)(const GIFPackedReg* RESTRICT r, u32 size);
GIFPackedRegHandlerC m_fpGIFPackedRegHandlersC[2]; GIFPackedRegHandlerC m_fpGIFPackedRegHandlersC[2];
GIFPackedRegHandlerC m_fpGIFPackedRegHandlerSTQRGBAXYZF2[8]; GIFPackedRegHandlerC m_fpGIFPackedRegHandlerSTQRGBAXYZF2[8];
GIFPackedRegHandlerC m_fpGIFPackedRegHandlerSTQRGBAXYZ2[8]; GIFPackedRegHandlerC m_fpGIFPackedRegHandlerSTQRGBAXYZ2[8];
template<uint32 prim, bool auto_flush> void GIFPackedRegHandlerSTQRGBAXYZF2(const GIFPackedReg* RESTRICT r, uint32 size); template<u32 prim, bool auto_flush> void GIFPackedRegHandlerSTQRGBAXYZF2(const GIFPackedReg* RESTRICT r, u32 size);
template<uint32 prim, bool auto_flush> void GIFPackedRegHandlerSTQRGBAXYZ2(const GIFPackedReg* RESTRICT r, uint32 size); template<u32 prim, bool auto_flush> void GIFPackedRegHandlerSTQRGBAXYZ2(const GIFPackedReg* RESTRICT r, u32 size);
void GIFPackedRegHandlerNOP(const GIFPackedReg* RESTRICT r, uint32 size); void GIFPackedRegHandlerNOP(const GIFPackedReg* RESTRICT r, u32 size);
template<int i> void ApplyTEX0(GIFRegTEX0& TEX0); template<int i> void ApplyTEX0(GIFRegTEX0& TEX0);
void ApplyPRIM(uint32 prim); void ApplyPRIM(u32 prim);
void GIFRegHandlerNull(const GIFReg* RESTRICT r); void GIFRegHandlerNull(const GIFReg* RESTRICT r);
void GIFRegHandlerPRIM(const GIFReg* RESTRICT r); void GIFRegHandlerPRIM(const GIFReg* RESTRICT r);
@ -86,8 +86,8 @@ class GSState : public GSAlignedClass<32>
void GIFRegHandlerST(const GIFReg* RESTRICT r); void GIFRegHandlerST(const GIFReg* RESTRICT r);
void GIFRegHandlerUV(const GIFReg* RESTRICT r); void GIFRegHandlerUV(const GIFReg* RESTRICT r);
void GIFRegHandlerUV_Hack(const GIFReg* RESTRICT r); void GIFRegHandlerUV_Hack(const GIFReg* RESTRICT r);
template<uint32 prim, uint32 adc, bool auto_flush> void GIFRegHandlerXYZF2(const GIFReg* RESTRICT r); template<u32 prim, u32 adc, bool auto_flush> void GIFRegHandlerXYZF2(const GIFReg* RESTRICT r);
template<uint32 prim, uint32 adc, bool auto_flush> void GIFRegHandlerXYZ2(const GIFReg* RESTRICT r); template<u32 prim, u32 adc, bool auto_flush> void GIFRegHandlerXYZ2(const GIFReg* RESTRICT r);
template<int i> void GIFRegHandlerTEX0(const GIFReg* RESTRICT r); template<int i> void GIFRegHandlerTEX0(const GIFReg* RESTRICT r);
template<int i> void GIFRegHandlerCLAMP(const GIFReg* RESTRICT r); template<int i> void GIFRegHandlerCLAMP(const GIFReg* RESTRICT r);
void GIFRegHandlerFOG(const GIFReg* RESTRICT r); void GIFRegHandlerFOG(const GIFReg* RESTRICT r);
@ -128,7 +128,7 @@ class GSState : public GSAlignedClass<32>
int x, y; int x, y;
int start, end, total; int start, end, total;
bool overflow; bool overflow;
uint8* buff; u8* buff;
GIFRegBITBLTBUF m_blit; GIFRegBITBLTBUF m_blit;
GSTransferBuffer(); GSTransferBuffer();
@ -163,12 +163,12 @@ protected:
GSVertex* buff; GSVertex* buff;
size_t head, tail, next, maxcount; // head: first vertex, tail: last vertex + 1, next: last indexed + 1 size_t head, tail, next, maxcount; // head: first vertex, tail: last vertex + 1, next: last indexed + 1
size_t xy_tail; size_t xy_tail;
uint64 xy[4]; u64 xy[4];
} m_vertex; } m_vertex;
struct struct
{ {
uint32* buff; u32* buff;
size_t tail; size_t tail;
} m_index; } m_index;
@ -179,8 +179,8 @@ protected:
void GrowVertexBuffer(); void GrowVertexBuffer();
template <uint32 prim, bool auto_flush> template <u32 prim, bool auto_flush>
void VertexKick(uint32 skip); void VertexKick(u32 skip);
// following functions need m_vt to be initialized // following functions need m_vt to be initialized
@ -188,11 +188,11 @@ protected:
void GetTextureMinMax(GSVector4i& r, const GIFRegTEX0& TEX0, const GIFRegCLAMP& CLAMP, bool linear); void GetTextureMinMax(GSVector4i& r, const GIFRegTEX0& TEX0, const GIFRegCLAMP& CLAMP, bool linear);
void GetAlphaMinMax(); void GetAlphaMinMax();
bool TryAlphaTest(uint32& fm, uint32& zm); bool TryAlphaTest(u32& fm, u32& zm);
bool IsOpaque(); bool IsOpaque();
bool IsMipMapDraw(); bool IsMipMapDraw();
bool IsMipMapActive(); bool IsMipMapActive();
GIFRegTEX0 GetTex0Layer(uint32 lod); GIFRegTEX0 GetTex0Layer(u32 lod);
public: public:
GIFPath m_path[4]; GIFPath m_path[4];
@ -202,7 +202,7 @@ public:
GSDrawingEnvironment m_env; GSDrawingEnvironment m_env;
GSDrawingContext* m_context; GSDrawingContext* m_context;
GSPerfMon m_perfmon; GSPerfMon m_perfmon;
uint32 m_crc; u32 m_crc;
CRC::Game m_game; CRC::Game m_game;
std::unique_ptr<GSDumpBase> m_dump; std::unique_ptr<GSDumpBase> m_dump;
int m_options; int m_options;
@ -248,17 +248,17 @@ public:
virtual void InvalidateLocalMem(const GIFRegBITBLTBUF& BITBLTBUF, const GSVector4i& r, bool clut = false) {} virtual void InvalidateLocalMem(const GIFRegBITBLTBUF& BITBLTBUF, const GSVector4i& r, bool clut = false) {}
void Move(); void Move();
void Write(const uint8* mem, int len); void Write(const u8* mem, int len);
void Read(uint8* mem, int len); void Read(u8* mem, int len);
void InitReadFIFO(uint8* mem, int len); void InitReadFIFO(u8* mem, int len);
void SoftReset(uint32 mask); void SoftReset(u32 mask);
void WriteCSR(uint32 csr) { m_regs->CSR.u32[1] = csr; } void WriteCSR(u32 csr) { m_regs->CSR.U32[1] = csr; }
void ReadFIFO(uint8* mem, int size); void ReadFIFO(u8* mem, int size);
template<int index> void Transfer(const uint8* mem, uint32 size); template<int index> void Transfer(const u8* mem, u32 size);
int Freeze(freezeData* fd, bool sizeonly); int Freeze(freezeData* fd, bool sizeonly);
int Defrost(const freezeData* fd); int Defrost(const freezeData* fd);
virtual void SetGameCRC(uint32 crc, int options); virtual void SetGameCRC(u32 crc, int options);
void SetFrameSkip(int skip); void SetFrameSkip(int skip);
void SetRegsMem(uint8* basemem); void SetRegsMem(u8* basemem);
}; };

40
pcsx2/GS/GSTables.cpp
View File

@ -20,7 +20,7 @@
#include "GS_types.h" #include "GS_types.h"
template <int Width, int Height> template <int Width, int Height>
static constexpr GSSizedBlockSwizzleTable<Height, Width> makeSwizzleTable(const uint8 (&arr)[Height][Width]) { static constexpr GSSizedBlockSwizzleTable<Height, Width> makeSwizzleTable(const u8 (&arr)[Height][Width]) {
GSSizedBlockSwizzleTable<Height, Width> table = {}; GSSizedBlockSwizzleTable<Height, Width> table = {};
for (int y = 0; y < 8; y++) { for (int y = 0; y < 8; y++) {
for (int x = 0; x < 8; x++) { for (int x = 0; x < 8; x++) {
@ -30,7 +30,7 @@ static constexpr GSSizedBlockSwizzleTable<Height, Width> makeSwizzleTable(const
return table; return table;
} }
static constexpr uint8 _blockTable32[4][8] = static constexpr u8 _blockTable32[4][8] =
{ {
{ 0, 1, 4, 5, 16, 17, 20, 21}, { 0, 1, 4, 5, 16, 17, 20, 21},
{ 2, 3, 6, 7, 18, 19, 22, 23}, { 2, 3, 6, 7, 18, 19, 22, 23},
@ -38,7 +38,7 @@ static constexpr uint8 _blockTable32[4][8] =
{ 10, 11, 14, 15, 26, 27, 30, 31} { 10, 11, 14, 15, 26, 27, 30, 31}
}; };
static constexpr uint8 _blockTable32Z[4][8] = static constexpr u8 _blockTable32Z[4][8] =
{ {
{ 24, 25, 28, 29, 8, 9, 12, 13}, { 24, 25, 28, 29, 8, 9, 12, 13},
{ 26, 27, 30, 31, 10, 11, 14, 15}, { 26, 27, 30, 31, 10, 11, 14, 15},
@ -46,7 +46,7 @@ static constexpr uint8 _blockTable32Z[4][8] =
{ 18, 19, 22, 23, 2, 3, 6, 7} { 18, 19, 22, 23, 2, 3, 6, 7}
}; };
static constexpr uint8 _blockTable16[8][4] = static constexpr u8 _blockTable16[8][4] =
{ {
{ 0, 2, 8, 10 }, { 0, 2, 8, 10 },
{ 1, 3, 9, 11 }, { 1, 3, 9, 11 },
@ -58,7 +58,7 @@ static constexpr uint8 _blockTable16[8][4] =
{ 21, 23, 29, 31 } { 21, 23, 29, 31 }
}; };
static constexpr uint8 _blockTable16S[8][4] = static constexpr u8 _blockTable16S[8][4] =
{ {
{ 0, 2, 16, 18 }, { 0, 2, 16, 18 },
{ 1, 3, 17, 19 }, { 1, 3, 17, 19 },
@ -70,7 +70,7 @@ static constexpr uint8 _blockTable16S[8][4] =
{ 13, 15, 29, 31 } { 13, 15, 29, 31 }
}; };
static constexpr uint8 _blockTable16Z[8][4] = static constexpr u8 _blockTable16Z[8][4] =
{ {
{ 24, 26, 16, 18 }, { 24, 26, 16, 18 },
{ 25, 27, 17, 19 }, { 25, 27, 17, 19 },
@ -82,7 +82,7 @@ static constexpr uint8 _blockTable16Z[8][4] =
{ 13, 15, 5, 7 } { 13, 15, 5, 7 }
}; };
static constexpr uint8 _blockTable16SZ[8][4] = static constexpr u8 _blockTable16SZ[8][4] =
{ {
{ 24, 26, 8, 10 }, { 24, 26, 8, 10 },
{ 25, 27, 9, 11 }, { 25, 27, 9, 11 },
@ -94,7 +94,7 @@ static constexpr uint8 _blockTable16SZ[8][4] =
{ 21, 23, 5, 7 } { 21, 23, 5, 7 }
}; };
static constexpr uint8 _blockTable8[4][8] = static constexpr u8 _blockTable8[4][8] =
{ {
{ 0, 1, 4, 5, 16, 17, 20, 21}, { 0, 1, 4, 5, 16, 17, 20, 21},
{ 2, 3, 6, 7, 18, 19, 22, 23}, { 2, 3, 6, 7, 18, 19, 22, 23},
@ -102,7 +102,7 @@ static constexpr uint8 _blockTable8[4][8] =
{ 10, 11, 14, 15, 26, 27, 30, 31} { 10, 11, 14, 15, 26, 27, 30, 31}
}; };
static constexpr uint8 _blockTable4[8][4] = static constexpr u8 _blockTable4[8][4] =
{ {
{ 0, 2, 8, 10 }, { 0, 2, 8, 10 },
{ 1, 3, 9, 11 }, { 1, 3, 9, 11 },
@ -123,7 +123,7 @@ constexpr GSSizedBlockSwizzleTable<8, 4> blockTable16SZ = makeSwizzleTable(_bloc
constexpr GSSizedBlockSwizzleTable<4, 8> blockTable8 = makeSwizzleTable(_blockTable8); constexpr GSSizedBlockSwizzleTable<4, 8> blockTable8 = makeSwizzleTable(_blockTable8);
constexpr GSSizedBlockSwizzleTable<8, 4> blockTable4 = makeSwizzleTable(_blockTable4); constexpr GSSizedBlockSwizzleTable<8, 4> blockTable4 = makeSwizzleTable(_blockTable4);
constexpr uint8 columnTable32[8][8] = constexpr u8 columnTable32[8][8] =
{ {
{ 0, 1, 4, 5, 8, 9, 12, 13 }, { 0, 1, 4, 5, 8, 9, 12, 13 },
{ 2, 3, 6, 7, 10, 11, 14, 15 }, { 2, 3, 6, 7, 10, 11, 14, 15 },
@ -135,7 +135,7 @@ constexpr uint8 columnTable32[8][8] =
{ 50, 51, 54, 55, 58, 59, 62, 63 }, { 50, 51, 54, 55, 58, 59, 62, 63 },
}; };
constexpr uint8 columnTable16[8][16] = constexpr u8 columnTable16[8][16] =
{ {
{ 0, 2, 8, 10, 16, 18, 24, 26, { 0, 2, 8, 10, 16, 18, 24, 26,
1, 3, 9, 11, 17, 19, 25, 27 }, 1, 3, 9, 11, 17, 19, 25, 27 },
@ -155,7 +155,7 @@ constexpr uint8 columnTable16[8][16] =
101, 103, 109, 111, 117, 119, 125, 127 }, 101, 103, 109, 111, 117, 119, 125, 127 },
}; };
constexpr uint8 columnTable8[16][16] = constexpr u8 columnTable8[16][16] =
{ {
{ 0, 4, 16, 20, 32, 36, 48, 52, // column 0 { 0, 4, 16, 20, 32, 36, 48, 52, // column 0
2, 6, 18, 22, 34, 38, 50, 54 }, 2, 6, 18, 22, 34, 38, 50, 54 },
@ -191,7 +191,7 @@ constexpr uint8 columnTable8[16][16] =
203, 207, 219, 223, 235, 239, 251, 255 }, 203, 207, 219, 223, 235, 239, 251, 255 },
}; };
constexpr uint16 columnTable4[16][32] = constexpr u16 columnTable4[16][32] =
{ {
{ 0, 8, 32, 40, 64, 72, 96, 104, // column 0 { 0, 8, 32, 40, 64, 72, 96, 104, // column 0
2, 10, 34, 42, 66, 74, 98, 106, 2, 10, 34, 42, 66, 74, 98, 106,
@ -259,7 +259,7 @@ constexpr uint16 columnTable4[16][32] =
407, 415, 439, 447, 471, 479, 503, 511 }, 407, 415, 439, 447, 471, 479, 503, 511 },
}; };
constexpr uint8 clutTableT32I8[128] = constexpr u8 clutTableT32I8[128] =
{ {
0, 1, 4, 5, 8, 9, 12, 13, 2, 3, 6, 7, 10, 11, 14, 15, 0, 1, 4, 5, 8, 9, 12, 13, 2, 3, 6, 7, 10, 11, 14, 15,
64, 65, 68, 69, 72, 73, 76, 77, 66, 67, 70, 71, 74, 75, 78, 79, 64, 65, 68, 69, 72, 73, 76, 77, 66, 67, 70, 71, 74, 75, 78, 79,
@ -271,13 +271,13 @@ constexpr uint8 clutTableT32I8[128] =
112, 113, 116, 117, 120, 121, 124, 125, 114, 115, 118, 119, 122, 123, 126, 127 112, 113, 116, 117, 120, 121, 124, 125, 114, 115, 118, 119, 122, 123, 126, 127
}; };
constexpr uint8 clutTableT32I4[16] = constexpr u8 clutTableT32I4[16] =
{ {
0, 1, 4, 5, 8, 9, 12, 13, 0, 1, 4, 5, 8, 9, 12, 13,
2, 3, 6, 7, 10, 11, 14, 15 2, 3, 6, 7, 10, 11, 14, 15
}; };
constexpr uint8 clutTableT16I8[32] = constexpr u8 clutTableT16I8[32] =
{ {
0, 2, 8, 10, 16, 18, 24, 26, 0, 2, 8, 10, 16, 18, 24, 26,
4, 6, 12, 14, 20, 22, 28, 30, 4, 6, 12, 14, 20, 22, 28, 30,
@ -285,14 +285,14 @@ constexpr uint8 clutTableT16I8[32] =
5, 7, 13, 15, 21, 23, 29, 31 5, 7, 13, 15, 21, 23, 29, 31
}; };
constexpr uint8 clutTableT16I4[16] = constexpr u8 clutTableT16I4[16] =
{ {
0, 2, 8, 10, 16, 18, 24, 26, 0, 2, 8, 10, 16, 18, 24, 26,
4, 6, 12, 14, 20, 22, 28, 30 4, 6, 12, 14, 20, 22, 28, 30
}; };
template <int BlocksHigh, int BlocksWide, int ColHeight, int ColWidth, typename Col> template <int BlocksHigh, int BlocksWide, int ColHeight, int ColWidth, typename Col>
constexpr int pxOffset(const uint8 (&blockTable)[BlocksHigh][BlocksWide], Col (&colTable)[ColHeight][ColWidth], int x, int y) constexpr int pxOffset(const u8 (&blockTable)[BlocksHigh][BlocksWide], Col (&colTable)[ColHeight][ColWidth], int x, int y)
{ {
int blockSize = ColHeight * ColWidth; int blockSize = ColHeight * ColWidth;
int pageSize = blockSize * BlocksHigh * BlocksWide; int pageSize = blockSize * BlocksHigh * BlocksWide;
@ -305,7 +305,7 @@ constexpr int pxOffset(const uint8 (&blockTable)[BlocksHigh][BlocksWide], Col (&
} }
template <int BlocksHigh, int BlocksWide, int ColHeight, int ColWidth, typename Col> template <int BlocksHigh, int BlocksWide, int ColHeight, int ColWidth, typename Col>
constexpr GSPixelColOffsetTable<BlocksHigh * ColHeight> makeColOffsetTable(const uint8 (&blockTable)[BlocksHigh][BlocksWide], Col (&colTable)[ColHeight][ColWidth]) constexpr GSPixelColOffsetTable<BlocksHigh * ColHeight> makeColOffsetTable(const u8 (&blockTable)[BlocksHigh][BlocksWide], Col (&colTable)[ColHeight][ColWidth])
{ {
constexpr int size = BlocksHigh * ColHeight; constexpr int size = BlocksHigh * ColHeight;
GSPixelColOffsetTable<size> table = {}; GSPixelColOffsetTable<size> table = {};
@ -317,7 +317,7 @@ constexpr GSPixelColOffsetTable<BlocksHigh * ColHeight> makeColOffsetTable(const
} }
template <int BlocksHigh, int BlocksWide, int ColHeight, int ColWidth, typename Col> template <int BlocksHigh, int BlocksWide, int ColHeight, int ColWidth, typename Col>
constexpr GSSizedPixelRowOffsetTable<BlocksWide * ColWidth> makeRowOffsetTable(const uint8 (&blockTable)[BlocksHigh][BlocksWide], Col (&colTable)[ColHeight][ColWidth], int y) constexpr GSSizedPixelRowOffsetTable<BlocksWide * ColWidth> makeRowOffsetTable(const u8 (&blockTable)[BlocksHigh][BlocksWide], Col (&colTable)[ColHeight][ColWidth], int y)
{ {
int base = pxOffset(blockTable, colTable, 0, y); int base = pxOffset(blockTable, colTable, 0, y);
GSSizedPixelRowOffsetTable<BlocksWide * ColWidth> table = {}; GSSizedPixelRowOffsetTable<BlocksWide * ColWidth> table = {};

20
pcsx2/GS/GSTables.h
View File

@ -21,9 +21,9 @@
struct alignas(64) GSBlockSwizzleTable struct alignas(64) GSBlockSwizzleTable
{ {
// Some swizzles are 4x8 and others are 8x4. An 8x8 table can store either at the cost of 2x size // Some swizzles are 4x8 and others are 8x4. An 8x8 table can store either at the cost of 2x size
uint8 value[8][8]; u8 value[8][8];
constexpr uint8 lookup(int x, int y) const constexpr u8 lookup(int x, int y) const
{ {
return value[y & 7][x & 7]; return value[y & 7][x & 7];
} }
@ -111,14 +111,14 @@ extern const GSSizedBlockSwizzleTable<8, 4> blockTable16Z;
extern const GSSizedBlockSwizzleTable<8, 4> blockTable16SZ; extern const GSSizedBlockSwizzleTable<8, 4> blockTable16SZ;
extern const GSSizedBlockSwizzleTable<4, 8> blockTable8; extern const GSSizedBlockSwizzleTable<4, 8> blockTable8;
extern const GSSizedBlockSwizzleTable<8, 4> blockTable4; extern const GSSizedBlockSwizzleTable<8, 4> blockTable4;
extern const uint8 columnTable32[8][8]; extern const u8 columnTable32[8][8];
extern const uint8 columnTable16[8][16]; extern const u8 columnTable16[8][16];
extern const uint8 columnTable8[16][16]; extern const u8 columnTable8[16][16];
extern const uint16 columnTable4[16][32]; extern const u16 columnTable4[16][32];
extern const uint8 clutTableT32I8[128]; extern const u8 clutTableT32I8[128];
extern const uint8 clutTableT32I4[16]; extern const u8 clutTableT32I4[16];
extern const uint8 clutTableT16I8[32]; extern const u8 clutTableT16I8[32];
extern const uint8 clutTableT16I4[16]; extern const u8 clutTableT16I4[16];
extern const GSPixelColOffsetTable< 32> pixelColOffset32; extern const GSPixelColOffsetTable< 32> pixelColOffset32;
extern const GSPixelColOffsetTable< 32> pixelColOffset32Z; extern const GSPixelColOffsetTable< 32> pixelColOffset32Z;
extern const GSPixelColOffsetTable< 64> pixelColOffset16; extern const GSPixelColOffsetTable< 64> pixelColOffset16;

4
pcsx2/GS/GSThread_CXX11.h
View File

@ -50,7 +50,7 @@ private:
l.unlock(); l.unlock();
uint32 waited = 0; u32 waited = 0;
while (true) while (true)
{ {
while (m_queue.consume_one(*this)) while (m_queue.consume_one(*this))
@ -110,7 +110,7 @@ public:
void Wait() void Wait()
{ {
uint32 waited = 0; u32 waited = 0;
while (true) while (true)
{ {
if (IsEmpty()) if (IsEmpty())

26
pcsx2/GS/GSUtil.cpp
View File

@ -32,11 +32,11 @@ Xbyak::util::Cpu g_cpu;
static class GSUtilMaps static class GSUtilMaps
{ {
public: public:
uint8 PrimClassField[8]; u8 PrimClassField[8];
uint8 VertexCountField[8]; u8 VertexCountField[8];
uint8 ClassVertexCountField[4]; u8 ClassVertexCountField[4];
uint32 CompatibleBitsField[64][2]; u32 CompatibleBitsField[64][2];
uint32 SharedBitsField[64][2]; u32 SharedBitsField[64][2];
// Defer init to avoid AVX2 illegal instructions // Defer init to avoid AVX2 illegal instructions
void Init() void Init()
@ -105,42 +105,42 @@ void GSUtil::Init()
s_maps.Init(); s_maps.Init();
} }
GS_PRIM_CLASS GSUtil::GetPrimClass(uint32 prim) GS_PRIM_CLASS GSUtil::GetPrimClass(u32 prim)
{ {
return (GS_PRIM_CLASS)s_maps.PrimClassField[prim]; return (GS_PRIM_CLASS)s_maps.PrimClassField[prim];
} }
int GSUtil::GetVertexCount(uint32 prim) int GSUtil::GetVertexCount(u32 prim)
{ {
return s_maps.VertexCountField[prim]; return s_maps.VertexCountField[prim];
} }
int GSUtil::GetClassVertexCount(uint32 primclass) int GSUtil::GetClassVertexCount(u32 primclass)
{ {
return s_maps.ClassVertexCountField[primclass]; return s_maps.ClassVertexCountField[primclass];
} }
const uint32* GSUtil::HasSharedBitsPtr(uint32 dpsm) const u32* GSUtil::HasSharedBitsPtr(u32 dpsm)
{ {
return s_maps.SharedBitsField[dpsm]; return s_maps.SharedBitsField[dpsm];
} }
bool GSUtil::HasSharedBits(uint32 spsm, const uint32* RESTRICT ptr) bool GSUtil::HasSharedBits(u32 spsm, const u32* RESTRICT ptr)
{ {
return (ptr[spsm >> 5] & (1 << (spsm & 0x1f))) == 0; return (ptr[spsm >> 5] & (1 << (spsm & 0x1f))) == 0;
} }
bool GSUtil::HasSharedBits(uint32 spsm, uint32 dpsm) bool GSUtil::HasSharedBits(u32 spsm, u32 dpsm)
{ {
return (s_maps.SharedBitsField[dpsm][spsm >> 5] & (1 << (spsm & 0x1f))) == 0; return (s_maps.SharedBitsField[dpsm][spsm >> 5] & (1 << (spsm & 0x1f))) == 0;
} }
bool GSUtil::HasSharedBits(uint32 sbp, uint32 spsm, uint32 dbp, uint32 dpsm) bool GSUtil::HasSharedBits(u32 sbp, u32 spsm, u32 dbp, u32 dpsm)
{ {
return ((sbp ^ dbp) | (s_maps.SharedBitsField[dpsm][spsm >> 5] & (1 << (spsm & 0x1f)))) == 0; return ((sbp ^ dbp) | (s_maps.SharedBitsField[dpsm][spsm >> 5] & (1 << (spsm & 0x1f)))) == 0;
} }
bool GSUtil::HasCompatibleBits(uint32 spsm, uint32 dpsm) bool GSUtil::HasCompatibleBits(u32 spsm, u32 dpsm)
{ {
return (s_maps.CompatibleBitsField[spsm][dpsm >> 5] & (1 << (dpsm & 0x1f))) != 0; return (s_maps.CompatibleBitsField[spsm][dpsm >> 5] & (1 << (dpsm & 0x1f))) != 0;
} }

16
pcsx2/GS/GSUtil.h
View File

@ -24,15 +24,15 @@ class GSUtil
public: public:
static void Init(); static void Init();
static GS_PRIM_CLASS GetPrimClass(uint32 prim); static GS_PRIM_CLASS GetPrimClass(u32 prim);
static int GetVertexCount(uint32 prim); static int GetVertexCount(u32 prim);
static int GetClassVertexCount(uint32 primclass); static int GetClassVertexCount(u32 primclass);
static const uint32* HasSharedBitsPtr(uint32 dpsm); static const u32* HasSharedBitsPtr(u32 dpsm);
static bool HasSharedBits(uint32 spsm, const uint32* ptr); static bool HasSharedBits(u32 spsm, const u32* ptr);
static bool HasSharedBits(uint32 spsm, uint32 dpsm); static bool HasSharedBits(u32 spsm, u32 dpsm);
static bool HasSharedBits(uint32 sbp, uint32 spsm, uint32 dbp, uint32 dpsm); static bool HasSharedBits(u32 sbp, u32 spsm, u32 dbp, u32 dpsm);
static bool HasCompatibleBits(uint32 spsm, uint32 dpsm); static bool HasCompatibleBits(u32 spsm, u32 dpsm);
static bool CheckSSE(); static bool CheckSSE();
static CRCHackLevel GetRecommendedCRCHackLevel(GSRendererType type); static CRCHackLevel GetRecommendedCRCHackLevel(GSRendererType type);

28
pcsx2/GS/GSVector4.h
View File

@ -50,15 +50,15 @@ public:
struct { float r, g, b, a; }; struct { float r, g, b, a; };
struct { float left, top, right, bottom; }; struct { float left, top, right, bottom; };
float v[4]; float v[4];
float f32[4]; float F32[4];
int8 i8[16]; s8 I8[16];
int16 i16[8]; s16 I16[8];
int32 i32[4]; s32 I32[4];
int64 i64[2]; s64 I64[2];
uint8 u8[16]; u8 U8[16];
uint16 u16[8]; u16 U16[8];
uint32 u32[4]; u32 U32[4];
uint64 u64[2]; u64 U64[2];
__m128 m; __m128 m;
}; };
@ -154,7 +154,7 @@ public:
#endif #endif
} }
__forceinline explicit GSVector4(uint32 u) __forceinline explicit GSVector4(u32 u)
{ {
GSVector4i v((int)u); GSVector4i v((int)u);
@ -216,17 +216,17 @@ public:
return *this; return *this;
} }
__forceinline uint32 rgba32() const __forceinline u32 rgba32() const
{ {
return GSVector4i(*this).rgba32(); return GSVector4i(*this).rgba32();
} }
__forceinline static GSVector4 rgba32(uint32 rgba) __forceinline static GSVector4 rgba32(u32 rgba)
{ {
return GSVector4(GSVector4i::load((int)rgba).u8to32()); return GSVector4(GSVector4i::load((int)rgba).u8to32());
} }
__forceinline static GSVector4 rgba32(uint32 rgba, int shift) __forceinline static GSVector4 rgba32(u32 rgba, int shift)
{ {
return GSVector4(GSVector4i::load((int)rgba).u8to32() << shift); return GSVector4(GSVector4i::load((int)rgba).u8to32() << shift);
} }
@ -634,7 +634,7 @@ GSVector.h:2973:15: error: shadows template parm 'int i'
return GSVector4(_mm_load_ss(&f)); return GSVector4(_mm_load_ss(&f));
} }
__forceinline static GSVector4 load(uint32 u) __forceinline static GSVector4 load(u32 u)
{ {
GSVector4i v = GSVector4i::load((int)u); GSVector4i v = GSVector4i::load((int)u);

52
pcsx2/GS/GSVector4i.h
View File

@ -54,15 +54,15 @@ public:
struct { int r, g, b, a; }; struct { int r, g, b, a; };
struct { int left, top, right, bottom; }; struct { int left, top, right, bottom; };
int v[4]; int v[4];
float f32[4]; float F32[4];
int8 i8[16]; s8 I8[16];
int16 i16[8]; s16 I16[8];
int32 i32[4]; s32 I32[4];
int64 i64[2]; s64 I64[2];
uint8 u8[16]; u8 U8[16];
uint16 u16[8]; u16 U16[8];
uint32 u32[4]; u32 U32[4];
uint64 u64[2]; u64 U64[2];
__m128i m; __m128i m;
}; };
@ -286,14 +286,14 @@ public:
// //
__forceinline uint32 rgba32() const __forceinline u32 rgba32() const
{ {
GSVector4i v = *this; GSVector4i v = *this;
v = v.ps32(v); v = v.ps32(v);
v = v.pu16(v); v = v.pu16(v);
return (uint32)store(v); return (u32)store(v);
} }
__forceinline GSVector4i sat_i8(const GSVector4i& a, const GSVector4i& b) const __forceinline GSVector4i sat_i8(const GSVector4i& a, const GSVector4i& b) const
@ -1073,13 +1073,13 @@ public:
#ifdef _M_AMD64 #ifdef _M_AMD64
template <int i> template <int i>
__forceinline GSVector4i insert64(int64 a) const __forceinline GSVector4i insert64(s64 a) const
{ {
return GSVector4i(_mm_insert_epi64(m, a, i)); return GSVector4i(_mm_insert_epi64(m, a, i));
} }
template <int i> template <int i>
__forceinline int64 extract64() const __forceinline s64 extract64() const
{ {
if (i == 0) if (i == 0)
return GSVector4i::storeq(*this); return GSVector4i::storeq(*this);
@ -1321,8 +1321,8 @@ public:
{ {
GSVector4i v; GSVector4i v;
v = loadq((int64)ptr[extract8<src + 0>() & 0xf]); v = loadq((s64)ptr[extract8<src + 0>() & 0xf]);
v = v.insert64<1>((int64)ptr[extract8<src + 0>() >> 4]); v = v.insert64<1>((s64)ptr[extract8<src + 0>() >> 4]);
return v; return v;
} }
@ -1332,8 +1332,8 @@ public:
{ {
GSVector4i v; GSVector4i v;
v = loadq((int64)ptr[extract8<src + 0>()]); v = loadq((s64)ptr[extract8<src + 0>()]);
v = v.insert64<1>((int64)ptr[extract8<src + 1>()]); v = v.insert64<1>((s64)ptr[extract8<src + 1>()]);
return v; return v;
} }
@ -1343,8 +1343,8 @@ public:
{ {
GSVector4i v; GSVector4i v;
v = loadq((int64)ptr[extract16<src + 0>()]); v = loadq((s64)ptr[extract16<src + 0>()]);
v = v.insert64<1>((int64)ptr[extract16<src + 1>()]); v = v.insert64<1>((s64)ptr[extract16<src + 1>()]);
return v; return v;
} }
@ -1354,8 +1354,8 @@ public:
{ {
GSVector4i v; GSVector4i v;
v = loadq((int64)ptr[extract32<src + 0>()]); v = loadq((s64)ptr[extract32<src + 0>()]);
v = v.insert64<1>((int64)ptr[extract32<src + 1>()]); v = v.insert64<1>((s64)ptr[extract32<src + 1>()]);
return v; return v;
} }
@ -1365,8 +1365,8 @@ public:
{ {
GSVector4i v; GSVector4i v;
v = loadq((int64)ptr[extract64<0>()]); v = loadq((s64)ptr[extract64<0>()]);
v = v.insert64<1>((int64)ptr[extract64<1>()]); v = v.insert64<1>((s64)ptr[extract64<1>()]);
return v; return v;
} }
@ -1422,7 +1422,7 @@ public:
dst[1] = gather8_4<8>(ptr); dst[1] = gather8_4<8>(ptr);
} }
__forceinline void gather8_8(const uint8* RESTRICT ptr, GSVector4i* RESTRICT dst) const __forceinline void gather8_8(const u8* RESTRICT ptr, GSVector4i* RESTRICT dst) const
{ {
dst[0] = gather8_8<>(ptr); dst[0] = gather8_8<>(ptr);
} }
@ -1590,7 +1590,7 @@ public:
#ifdef _M_AMD64 #ifdef _M_AMD64
__forceinline static GSVector4i loadq(int64 i) __forceinline static GSVector4i loadq(s64 i)
{ {
return GSVector4i(_mm_cvtsi64_si128(i)); return GSVector4i(_mm_cvtsi64_si128(i));
} }
@ -1634,7 +1634,7 @@ public:
#ifdef _M_AMD64 #ifdef _M_AMD64
__forceinline static int64 storeq(const GSVector4i& v) __forceinline static s64 storeq(const GSVector4i& v)
{ {
return _mm_cvtsi128_si64(v.m); return _mm_cvtsi128_si64(v.m);
} }

20
pcsx2/GS/GSVector8.h
View File

@ -59,15 +59,15 @@ public:
struct { float x0, y0, z0, w0, x1, y1, z1, w1; }; struct { float x0, y0, z0, w0, x1, y1, z1, w1; };
struct { float r0, g0, b0, a0, r1, g1, b1, a1; }; struct { float r0, g0, b0, a0, r1, g1, b1, a1; };
float v[8]; float v[8];
float f32[8]; float F32[8];
int8 i8[32]; s8 I8[32];
int16 i16[16]; s16 I16[16];
int32 i32[8]; s32 I32[8];
int64 i64[4]; s64 I64[4];
uint8 u8[32]; u8 U8[32];
uint16 u16[16]; u16 U16[16];
uint32 u32[8]; u32 U32[8];
uint64 u64[4]; u64 U64[4];
__m256 m; __m256 m;
__m128 m0, m1; __m128 m0, m1;
}; };
@ -103,7 +103,7 @@ public:
return GSVector8(cxpr_setr_epi32(x, x, x, x, x, x, x, x)); return GSVector8(cxpr_setr_epi32(x, x, x, x, x, x, x, x));
} }
static constexpr GSVector8 cxpr(uint32 x) static constexpr GSVector8 cxpr(u32 x)
{ {
return cxpr(static_cast<int>(x)); return cxpr(static_cast<int>(x));
} }

36
pcsx2/GS/GSVector8i.h
View File

@ -69,15 +69,15 @@ public:
struct { int x0, y0, z0, w0, x1, y1, z1, w1; }; struct { int x0, y0, z0, w0, x1, y1, z1, w1; };
struct { int r0, g0, b0, a0, r1, g1, b1, a1; }; struct { int r0, g0, b0, a0, r1, g1, b1, a1; };
int v[8]; int v[8];
float f32[8]; float F32[8];
int8 i8[32]; s8 I8[32];
int16 i16[16]; s16 I16[16];
int32 i32[8]; s32 I32[8];
int64 i64[4]; s64 I64[4];
uint8 u8[32]; u8 U8[32];
uint16 u16[16]; u16 U16[16];
uint32 u32[8]; u32 U32[8];
uint64 u64[4]; u64 U64[4];
__m256i m; __m256i m;
__m128i m0, m1; __m128i m0, m1;
}; };
@ -1055,17 +1055,17 @@ public:
return cast(v0).insert<1>(v1); return cast(v0).insert<1>(v1);
} }
__forceinline GSVector8i gather32_32(const uint8* ptr) const __forceinline GSVector8i gather32_32(const u8* ptr) const
{ {
return GSVector8i(_mm256_i32gather_epi32((const int*)ptr, m, 1)) & GSVector8i::x000000ff(); return GSVector8i(_mm256_i32gather_epi32((const int*)ptr, m, 1)) & GSVector8i::x000000ff();
} }
__forceinline GSVector8i gather32_32(const uint16* ptr) const __forceinline GSVector8i gather32_32(const u16* ptr) const
{ {
return GSVector8i(_mm256_i32gather_epi32((const int*)ptr, m, 2)) & GSVector8i::x0000ffff(); return GSVector8i(_mm256_i32gather_epi32((const int*)ptr, m, 2)) & GSVector8i::x0000ffff();
} }
__forceinline GSVector8i gather32_32(const uint32* ptr) const __forceinline GSVector8i gather32_32(const u32* ptr) const
{ {
return GSVector8i(_mm256_i32gather_epi32((const int*)ptr, m, 4)); return GSVector8i(_mm256_i32gather_epi32((const int*)ptr, m, 4));
} }
@ -1092,14 +1092,14 @@ public:
return cast(v0).insert<1>(v1); return cast(v0).insert<1>(v1);
} }
__forceinline GSVector8i gather32_32(const uint8* ptr1, const uint32* ptr2) const __forceinline GSVector8i gather32_32(const u8* ptr1, const u32* ptr2) const
{ {
return gather32_32<uint8>(ptr1).gather32_32<uint32>(ptr2); return gather32_32<u8>(ptr1).gather32_32<u32>(ptr2);
} }
__forceinline GSVector8i gather32_32(const uint32* ptr1, const uint32* ptr2) const __forceinline GSVector8i gather32_32(const u32* ptr1, const u32* ptr2) const
{ {
return gather32_32<uint32>(ptr1).gather32_32<uint32>(ptr2); return gather32_32<u32>(ptr1).gather32_32<u32>(ptr2);
} }
template <class T> template <class T>
@ -1169,7 +1169,7 @@ public:
#ifdef _M_AMD64 #ifdef _M_AMD64
__forceinline static GSVector8i loadq(int64 i) __forceinline static GSVector8i loadq(s64 i)
{ {
return cast(GSVector4i::loadq(i)); return cast(GSVector4i::loadq(i));
} }
@ -1213,7 +1213,7 @@ public:
#ifdef _M_AMD64 #ifdef _M_AMD64
__forceinline static int64 storeq(const GSVector8i& v) __forceinline static s64 storeq(const GSVector8i& v)
{ {
return GSVector4i::storeq(GSVector4i::cast(v)); return GSVector4i::storeq(GSVector4i::cast(v));
} }

11
pcsx2/GS/GS_types.h
View File

@ -15,17 +15,6 @@
#pragma once #pragma once
typedef unsigned char uint8;
typedef signed char int8;
typedef unsigned short uint16;
typedef signed short int16;
typedef unsigned int uint32;
typedef signed int int32;
typedef unsigned long long uint64;
typedef signed long long int64;
typedef signed long long sint64;
#include <cfloat> #include <cfloat>
// clang-format off // clang-format off

4
pcsx2/GS/Renderers/Common/GSDevice.cpp
View File

@ -126,7 +126,7 @@ GSTexture* GSDevice::FetchSurface(int type, int w, int h, int format)
void GSDevice::PrintMemoryUsage() void GSDevice::PrintMemoryUsage()
{ {
#ifdef ENABLE_OGL_DEBUG #ifdef ENABLE_OGL_DEBUG
uint32 pool = 0; u32 pool = 0;
for (auto t : m_pool) for (auto t : m_pool)
{ {
if (t) if (t)
@ -422,7 +422,7 @@ HWBlend GSDevice::GetBlend(size_t index)
return blend; return blend;
} }
uint16 GSDevice::GetBlendFlags(size_t index) { return m_blendMap[index].flags; } u16 GSDevice::GetBlendFlags(size_t index) { return m_blendMap[index].flags; }
// clang-format off // clang-format off

20
pcsx2/GS/Renderers/Common/GSDevice.h
View File

@ -120,7 +120,7 @@ enum HWBlendFlags
// Determines the HW blend function for DX11/OGL // Determines the HW blend function for DX11/OGL
struct HWBlend struct HWBlend
{ {
uint16 flags, op, src, dst; u16 flags, op, src, dst;
}; };
class GSDevice : public GSAlignedClass<32> class GSDevice : public GSAlignedClass<32>
@ -130,7 +130,7 @@ private:
static std::array<HWBlend, 3*3*3*3 + 1> m_blendMap; static std::array<HWBlend, 3*3*3*3 + 1> m_blendMap;
protected: protected:
enum : uint16 enum : u16
{ {
// HW blend factors // HW blend factors
SRC_COLOR, INV_SRC_COLOR, DST_COLOR, INV_DST_COLOR, SRC_COLOR, INV_SRC_COLOR, DST_COLOR, INV_DST_COLOR,
@ -172,7 +172,7 @@ protected:
virtual void DoFXAA(GSTexture* sTex, GSTexture* dTex) {} virtual void DoFXAA(GSTexture* sTex, GSTexture* dTex) {}
virtual void DoShadeBoost(GSTexture* sTex, GSTexture* dTex) {} virtual void DoShadeBoost(GSTexture* sTex, GSTexture* dTex) {}
virtual void DoExternalFX(GSTexture* sTex, GSTexture* dTex) {} virtual void DoExternalFX(GSTexture* sTex, GSTexture* dTex) {}
virtual uint16 ConvertBlendEnum(uint16 generic) = 0; // Convert blend factors/ops from the generic enum to DX11/OGl specific. virtual u16 ConvertBlendEnum(u16 generic) = 0; // Convert blend factors/ops from the generic enum to DX11/OGl specific.
public: public:
GSOsdManager m_osd; GSOsdManager m_osd;
@ -208,9 +208,9 @@ public:
virtual bool HasColorSparse() { return false; } virtual bool HasColorSparse() { return false; }
virtual void ClearRenderTarget(GSTexture* t, const GSVector4& c) {} virtual void ClearRenderTarget(GSTexture* t, const GSVector4& c) {}
virtual void ClearRenderTarget(GSTexture* t, uint32 c) {} virtual void ClearRenderTarget(GSTexture* t, u32 c) {}
virtual void ClearDepth(GSTexture* t) {} virtual void ClearDepth(GSTexture* t) {}
virtual void ClearStencil(GSTexture* t, uint8 c) {} virtual void ClearStencil(GSTexture* t, u8 c) {}
GSTexture* CreateSparseRenderTarget(int w, int h, int format = 0); GSTexture* CreateSparseRenderTarget(int w, int h, int format = 0);
GSTexture* CreateSparseDepthStencil(int w, int h, int format = 0); GSTexture* CreateSparseDepthStencil(int w, int h, int format = 0);
@ -257,15 +257,15 @@ public:
// Convert the GS blend equations to HW specific blend factors/ops // Convert the GS blend equations to HW specific blend factors/ops
// Index is computed as ((((A * 3 + B) * 3) + C) * 3) + D. A, B, C, D taken from ALPHA register. // Index is computed as ((((A * 3 + B) * 3) + C) * 3) + D. A, B, C, D taken from ALPHA register.
HWBlend GetBlend(size_t index); HWBlend GetBlend(size_t index);
uint16 GetBlendFlags(size_t index); u16 GetBlendFlags(size_t index);
}; };
struct GSAdapter struct GSAdapter
{ {
uint32 vendor; u32 vendor;
uint32 device; u32 device;
uint32 subsys; u32 subsys;
uint32 rev; u32 rev;
operator std::string() const; operator std::string() const;
bool operator==(const GSAdapter&) const; bool operator==(const GSAdapter&) const;

2
pcsx2/GS/Renderers/Common/GSDirtyRect.cpp
View File

@ -22,7 +22,7 @@ GSDirtyRect::GSDirtyRect()
left = top = right = bottom = 0; left = top = right = bottom = 0;
} }
GSDirtyRect::GSDirtyRect(const GSVector4i& r, uint32 psm) GSDirtyRect::GSDirtyRect(const GSVector4i& r, u32 psm)
: psm(psm) : psm(psm)
{ {
left = r.left; left = r.left;

4
pcsx2/GS/Renderers/Common/GSDirtyRect.h
View File

@ -24,11 +24,11 @@ class GSDirtyRect
int right; int right;
int bottom; int bottom;
uint32 psm; u32 psm;
public: public:
GSDirtyRect(); GSDirtyRect();
GSDirtyRect(const GSVector4i& r, uint32 psm); GSDirtyRect(const GSVector4i& r, u32 psm);
const GSVector4i GetDirtyRect(const GIFRegTEX0& TEX0) const; const GSVector4i GetDirtyRect(const GIFRegTEX0& TEX0) const;
}; };

56
pcsx2/GS/Renderers/Common/GSFastList.h
View File

@ -21,8 +21,8 @@ template <class T>
struct Element struct Element
{ {
T data; T data;
uint16 next_index; u16 next_index;
uint16 prev_index; u16 prev_index;
}; };
template <class T> template <class T>
@ -47,11 +47,11 @@ private:
// the relevant iterator (or the index alone) are erased from the list. // the relevant iterator (or the index alone) are erased from the list.
// m_buffer[0] is always present as auxiliary Element<T> of the list // m_buffer[0] is always present as auxiliary Element<T> of the list
Element<T>* m_buffer; Element<T>* m_buffer;
uint16 m_capacity; u16 m_capacity;
uint16 m_free_indexes_stack_top; u16 m_free_indexes_stack_top;
// m_free_indexes_stack has dynamic size (m_capacity - 1) // m_free_indexes_stack has dynamic size (m_capacity - 1)
// m_buffer indexes that are free to be used are stacked here // m_buffer indexes that are free to be used are stacked here
uint16* m_free_indexes_stack; u16* m_free_indexes_stack;
public: public:
__forceinline FastList() __forceinline FastList()
@ -74,8 +74,8 @@ public:
// Initialize m_buffer and m_free_indexes_stack as a contiguous block of memory starting at m_buffer // Initialize m_buffer and m_free_indexes_stack as a contiguous block of memory starting at m_buffer
// This should increase cache locality and reduce memory fragmentation // This should increase cache locality and reduce memory fragmentation
_aligned_free(m_buffer); _aligned_free(m_buffer);
m_buffer = (Element<T>*)_aligned_malloc(m_capacity * sizeof(Element<T>) + (m_capacity - 1) * sizeof(uint16), 64); m_buffer = (Element<T>*)_aligned_malloc(m_capacity * sizeof(Element<T>) + (m_capacity - 1) * sizeof(u16), 64);
m_free_indexes_stack = (uint16*)&m_buffer[m_capacity]; m_free_indexes_stack = (u16*)&m_buffer[m_capacity];
// Initialize m_buffer[0], data field is unused but initialized using default T constructor // Initialize m_buffer[0], data field is unused but initialized using default T constructor
m_buffer[0] = {T(), 0, 0}; m_buffer[0] = {T(), 0, 0};
@ -84,14 +84,14 @@ public:
m_free_indexes_stack_top = 0; m_free_indexes_stack_top = 0;
// m_buffer index 0 is reserved for auxiliary element // m_buffer index 0 is reserved for auxiliary element
for (uint16 i = 0; i < m_capacity - 1; i++) for (u16 i = 0; i < m_capacity - 1; i++)
{ {
m_free_indexes_stack[i] = i + 1; m_free_indexes_stack[i] = i + 1;
} }
} }
// Insert the element in front of the list and return its position in m_buffer // Insert the element in front of the list and return its position in m_buffer
__forceinline uint16 InsertFront(const T& data) __forceinline u16 InsertFront(const T& data)
{ {
if (Full()) if (Full())
{ {
@ -99,7 +99,7 @@ public:
} }
// Pop a free index from the stack // Pop a free index from the stack
const uint16 free_index = m_free_indexes_stack[m_free_indexes_stack_top++]; const u16 free_index = m_free_indexes_stack[m_free_indexes_stack_top++];
m_buffer[free_index].data = data; m_buffer[free_index].data = data;
ListInsertFront(free_index); ListInsertFront(free_index);
return free_index; return free_index;
@ -120,7 +120,7 @@ public:
EraseIndex(LastIndex()); EraseIndex(LastIndex());
} }
__forceinline uint16 size() const __forceinline u16 size() const
{ {
return m_free_indexes_stack_top; return m_free_indexes_stack_top;
} }
@ -130,13 +130,13 @@ public:
return size() == 0; return size() == 0;
} }
__forceinline void EraseIndex(const uint16 index) __forceinline void EraseIndex(const u16 index)
{ {
ListRemove(index); ListRemove(index);
m_free_indexes_stack[--m_free_indexes_stack_top] = index; m_free_indexes_stack[--m_free_indexes_stack_top] = index;
} }
__forceinline void MoveFront(const uint16 index) __forceinline void MoveFront(const u16 index)
{ {
if (FirstIndex() != index) if (FirstIndex() != index)
{ {
@ -163,29 +163,29 @@ public:
private: private:
// Accessed by FastListIterator<T> using class friendship // Accessed by FastListIterator<T> using class friendship
__forceinline const T& Data(const uint16 index) const __forceinline const T& Data(const u16 index) const
{ {
return m_buffer[index].data; return m_buffer[index].data;
} }
// Accessed by FastListIterator<T> using class friendship // Accessed by FastListIterator<T> using class friendship
__forceinline uint16 NextIndex(const uint16 index) const __forceinline u16 NextIndex(const u16 index) const
{ {
return m_buffer[index].next_index; return m_buffer[index].next_index;
} }
// Accessed by FastListIterator<T> using class friendship // Accessed by FastListIterator<T> using class friendship
__forceinline uint16 PrevIndex(const uint16 index) const __forceinline u16 PrevIndex(const u16 index) const
{ {
return m_buffer[index].prev_index; return m_buffer[index].prev_index;
} }
__forceinline uint16 FirstIndex() const __forceinline u16 FirstIndex() const
{ {
return m_buffer[0].next_index; return m_buffer[0].next_index;
} }
__forceinline uint16 LastIndex() const __forceinline u16 LastIndex() const
{ {
return m_buffer[0].prev_index; return m_buffer[0].prev_index;
} }
@ -196,7 +196,7 @@ private:
return size() == m_capacity - 1; return size() == m_capacity - 1;
} }
__forceinline void ListInsertFront(const uint16 index) __forceinline void ListInsertFront(const u16 index)
{ {
// Update prev / next indexes to add m_buffer[index] to the chain // Update prev / next indexes to add m_buffer[index] to the chain
Element<T>& head = m_buffer[0]; Element<T>& head = m_buffer[0];
@ -206,7 +206,7 @@ private:
head.next_index = index; head.next_index = index;
} }
__forceinline void ListRemove(const uint16 index) __forceinline void ListRemove(const u16 index)
{ {
// Update prev / next indexes to remove m_buffer[index] from the chain // Update prev / next indexes to remove m_buffer[index] from the chain
const Element<T>& to_remove = m_buffer[index]; const Element<T>& to_remove = m_buffer[index];
@ -221,13 +221,13 @@ private:
throw std::runtime_error("FastList size maxed out at USHRT_MAX (65535) elements, cannot grow futhermore."); throw std::runtime_error("FastList size maxed out at USHRT_MAX (65535) elements, cannot grow futhermore.");
} }
const uint16 new_capacity = m_capacity <= (USHRT_MAX / 2) ? (m_capacity * 2) : USHRT_MAX; const u16 new_capacity = m_capacity <= (USHRT_MAX / 2) ? (m_capacity * 2) : USHRT_MAX;
Element<T>* new_buffer = (Element<T>*)_aligned_malloc(new_capacity * sizeof(Element<T>) + (new_capacity - 1) * sizeof(uint16), 64); Element<T>* new_buffer = (Element<T>*)_aligned_malloc(new_capacity * sizeof(Element<T>) + (new_capacity - 1) * sizeof(u16), 64);
uint16* new_free_indexes_stack = (uint16*)&new_buffer[new_capacity]; u16* new_free_indexes_stack = (u16*)&new_buffer[new_capacity];
memcpy(new_buffer, m_buffer, m_capacity * sizeof(Element<T>)); memcpy(new_buffer, m_buffer, m_capacity * sizeof(Element<T>));
memcpy(new_free_indexes_stack, m_free_indexes_stack, (m_capacity - 1) * sizeof(uint16)); memcpy(new_free_indexes_stack, m_free_indexes_stack, (m_capacity - 1) * sizeof(u16));
_aligned_free(m_buffer); _aligned_free(m_buffer);
@ -235,7 +235,7 @@ private:
m_free_indexes_stack = new_free_indexes_stack; m_free_indexes_stack = new_free_indexes_stack;
// Initialize the additional space in the stack // Initialize the additional space in the stack
for (uint16 i = m_capacity - 1; i < new_capacity - 1; i++) for (u16 i = m_capacity - 1; i < new_capacity - 1; i++)
{ {
m_free_indexes_stack[i] = i + 1; m_free_indexes_stack[i] = i + 1;
} }
@ -251,10 +251,10 @@ class FastListIterator
{ {
private: private:
const FastList<T>* m_fastlist; const FastList<T>* m_fastlist;
uint16 m_index; u16 m_index;
public: public:
__forceinline FastListIterator(const FastList<T>* fastlist, const uint16 index) __forceinline FastListIterator(const FastList<T>* fastlist, const u16 index)
{ {
m_fastlist = fastlist; m_fastlist = fastlist;
m_index = index; m_index = index;
@ -305,7 +305,7 @@ public:
return m_fastlist->Data(m_index); return m_fastlist->Data(m_index);
} }
__forceinline uint16 Index() const __forceinline u16 Index() const
{ {
return m_index; return m_index;
} }

22
pcsx2/GS/Renderers/Common/GSFunctionMap.h
View File

@ -29,8 +29,8 @@ class GSFunctionMap
protected: protected:
struct ActivePtr struct ActivePtr
{ {
uint64 frame, frames, prims; u64 frame, frames, prims;
uint64 ticks, actual, total; u64 ticks, actual, total;
VALUE f; VALUE f;
}; };
@ -68,7 +68,7 @@ public:
memset(p, 0, sizeof(*p)); memset(p, 0, sizeof(*p));
p->frame = (uint64)-1; p->frame = (u64)-1;
p->f = GetDefaultFunction(key); p->f = GetDefaultFunction(key);
@ -80,7 +80,7 @@ public:
return m_active->f; return m_active->f;
} }
void UpdateStats(uint64 frame, uint64 ticks, int actual, int total, int prims) void UpdateStats(u64 frame, u64 ticks, int actual, int total, int prims)
{ {
if (m_active) if (m_active)
{ {
@ -101,7 +101,7 @@ public:
virtual void PrintStats() virtual void PrintStats()
{ {
uint64 totalTicks = 0; u64 totalTicks = 0;
for (const auto& i : m_map_active) for (const auto& i : m_map_active)
{ {
@ -128,10 +128,10 @@ public:
if (p->frames && p->actual) if (p->frames && p->actual)
{ {
uint64 tpf = p->ticks / p->frames; u64 tpf = p->ticks / p->frames;
printf("%016llx | %6llu | %5llu | %5.2f%% %5.1f %6.1f | %8llu %6llu %5.2f%%\n", printf("%016llx | %6llu | %5llu | %5.2f%% %5.1f %6.1f | %8llu %6llu %5.2f%%\n",
(uint64)key, (u64)key,
p->frames, p->frames,
p->prims / p->frames, p->prims / p->frames,
(double)(p->ticks * 100) / totalTicks, (double)(p->ticks * 100) / totalTicks,
@ -162,7 +162,7 @@ class GSCodeGeneratorFunctionMap : public GSFunctionMap<KEY, VALUE>
{ {
std::string m_name; std::string m_name;
void* m_param; void* m_param;
std::unordered_map<uint64, VALUE> m_cgmap; std::unordered_map<u64, VALUE> m_cgmap;
GSCodeBuffer m_cb; GSCodeBuffer m_cb;
size_t m_total_code_size; size_t m_total_code_size;
@ -201,7 +201,7 @@ public:
ASSERT(cg->getSize() < MAX_SIZE); ASSERT(cg->getSize() < MAX_SIZE);
#if 0 #if 0
fprintf(stderr, "%s Location:%p Size:%zu Key:%llx\n", m_name.c_str(), code_ptr, cg->getSize(), (uint64)key); fprintf(stderr, "%s Location:%p Size:%zu Key:%llx\n", m_name.c_str(), code_ptr, cg->getSize(), (u64)key);
GSScanlineSelector sel(key); GSScanlineSelector sel(key);
sel.Print(); sel.Print();
#endif #endif
@ -220,7 +220,7 @@ public:
// if(iJIT_IsProfilingActive()) // always > 0 // if(iJIT_IsProfilingActive()) // always > 0
{ {
std::string name = format("%s<%016llx>()", m_name.c_str(), (uint64)key); std::string name = format("%s<%016llx>()", m_name.c_str(), (u64)key);
iJIT_Method_Load ml; iJIT_Method_Load ml;
@ -233,7 +233,7 @@ public:
iJIT_NotifyEvent(iJVM_EVENT_TYPE_METHOD_LOAD_FINISHED, &ml); iJIT_NotifyEvent(iJVM_EVENT_TYPE_METHOD_LOAD_FINISHED, &ml);
/* /*
name = format("c:/temp1/%s_%016llx.bin", m_name.c_str(), (uint64)key); name = format("c:/temp1/%s_%016llx.bin", m_name.c_str(), (u64)key);
if(FILE* fp = fopen(name.c_str(), "wb")) if(FILE* fp = fopen(name.c_str(), "wb"))
{ {

12
pcsx2/GS/Renderers/Common/GSOsdManager.cpp
View File

@ -276,7 +276,7 @@ void GSOsdManager::Monitor(const char* key, const char* value)
} }
} }
void GSOsdManager::RenderGlyph(GSVertexPT1* dst, const glyph_info g, float x, float y, uint32 color) void GSOsdManager::RenderGlyph(GSVertexPT1* dst, const glyph_info g, float x, float y, u32 color)
{ {
float x2 = x + g.bl * (2.0f / m_real_size.x); float x2 = x + g.bl * (2.0f / m_real_size.x);
float y2 = -y - g.bt * (2.0f / m_real_size.y); float y2 = -y - g.bt * (2.0f / m_real_size.y);
@ -309,7 +309,7 @@ void GSOsdManager::RenderGlyph(GSVertexPT1* dst, const glyph_info g, float x, fl
++dst; ++dst;
} }
void GSOsdManager::RenderString(GSVertexPT1* dst, const std::u32string msg, float x, float y, uint32 color) void GSOsdManager::RenderString(GSVertexPT1* dst, const std::u32string msg, float x, float y, u32 color)
{ {
char32_t p = 0; char32_t p = 0;
for (const auto& c : msg) for (const auto& c : msg)
@ -435,8 +435,8 @@ size_t GSOsdManager::GeneratePrimitives(GSVertexPT1* dst, size_t count)
ratio = ratio > 1.0f ? 1.0f : ratio < 0.0f ? 0.0f : ratio; ratio = ratio > 1.0f ? 1.0f : ratio < 0.0f ? 0.0f : ratio;
y += offset += ((m_size + 2) * (2.0f / m_real_size.y)) * ratio; y += offset += ((m_size + 2) * (2.0f / m_real_size.y)) * ratio;
uint32 color = m_color; u32 color = m_color;
((uint8*)&color)[3] = (uint8)(((uint8*)&color)[3] * (1.0f - ratio) * opacity); ((u8*)&color)[3] = (u8)(((u8*)&color)[3] * (1.0f - ratio) * opacity);
RenderString(dst, it->msg, x, y, color); RenderString(dst, it->msg, x, y, color);
dst += it->msg.size() * 6; dst += it->msg.size() * 6;
drawn += it->msg.size() * 6; drawn += it->msg.size() * 6;
@ -473,8 +473,8 @@ size_t GSOsdManager::GeneratePrimitives(GSVertexPT1* dst, size_t count)
float x = 1.0f - 8 * (2.0f / m_real_size.x) - first_max - m_char_info[' '].ax * (2.0f / m_real_size.x) - second_max; float x = 1.0f - 8 * (2.0f / m_real_size.x) - first_max - m_char_info[' '].ax * (2.0f / m_real_size.x) - second_max;
float y = -1.0f + ((m_size + 2) * (2.0f / m_real_size.y)) * line++; float y = -1.0f + ((m_size + 2) * (2.0f / m_real_size.y)) * line++;
uint32 color = m_color; u32 color = m_color;
((uint8*)&color)[3] = (uint8)(((uint8*)&color)[3] * opacity); ((u8*)&color)[3] = (u8)(((u8*)&color)[3] * opacity);
// Render the key // Render the key
RenderString(dst, pair.first, x, y, color); RenderString(dst, pair.first, x, y, color);

26
pcsx2/GS/Renderers/Common/GSOsdManager.h
View File

@ -26,14 +26,14 @@ class GSOsdManager
{ {
struct glyph_info struct glyph_info
{ {
int32 ax; // advance.x s32 ax; // advance.x
int32 ay; // advance.y s32 ay; // advance.y
uint32 bw; // bitmap.width; u32 bw; // bitmap.width;
uint32 bh; // bitmap.rows; u32 bh; // bitmap.rows;
int32 bl; // bitmap_left; s32 bl; // bitmap_left;
int32 bt; // bitmap_top; s32 bt; // bitmap_top;
float tx; // x offset of glyph float tx; // x offset of glyph
float ty; // y offset of glyph float ty; // y offset of glyph
@ -47,10 +47,10 @@ class GSOsdManager
FT_Face m_face; FT_Face m_face;
FT_UInt m_size; FT_UInt m_size;
uint32 m_atlas_h; u32 m_atlas_h;
uint32 m_atlas_w; u32 m_atlas_w;
int32 m_max_width; s32 m_max_width;
int32 m_onscreen_messages; s32 m_onscreen_messages;
struct log_info struct log_info
{ {
@ -62,15 +62,15 @@ class GSOsdManager
std::map<std::u32string, std::u32string> m_monitor; std::map<std::u32string, std::u32string> m_monitor;
void AddGlyph(char32_t codepoint); void AddGlyph(char32_t codepoint);
void RenderGlyph(GSVertexPT1* dst, const glyph_info g, float x, float y, uint32 color); void RenderGlyph(GSVertexPT1* dst, const glyph_info g, float x, float y, u32 color);
void RenderString(GSVertexPT1* dst, const std::u32string msg, float x, float y, uint32 color); void RenderString(GSVertexPT1* dst, const std::u32string msg, float x, float y, u32 color);
float StringSize(const std::u32string msg); float StringSize(const std::u32string msg);
bool m_log_enabled; bool m_log_enabled;
int m_log_timeout; int m_log_timeout;
bool m_monitor_enabled; bool m_monitor_enabled;
int m_opacity; int m_opacity;
uint32 m_color; u32 m_color;
int m_max_onscreen_messages; int m_max_onscreen_messages;
public: public:

6
pcsx2/GS/Renderers/Common/GSTexture.h
View File

@ -31,7 +31,7 @@ protected:
public: public:
struct GSMap struct GSMap
{ {
uint8* bits; u8* bits;
int pitch; int pitch;
}; };
@ -61,7 +61,7 @@ public:
virtual void Unmap() = 0; virtual void Unmap() = 0;
virtual void GenerateMipmap() {} virtual void GenerateMipmap() {}
virtual bool Save(const std::string& fn) = 0; virtual bool Save(const std::string& fn) = 0;
virtual uint32 GetID() { return 0; } virtual u32 GetID() { return 0; }
GSVector2 GetScale() const { return m_scale; } GSVector2 GetScale() const { return m_scale; }
void SetScale(const GSVector2& scale) { m_scale = scale; } void SetScale(const GSVector2& scale) { m_scale = scale; }
@ -90,5 +90,5 @@ public:
float OffsetHack_mody; float OffsetHack_mody;
// Typical size of a RGBA texture // Typical size of a RGBA texture
virtual uint32 GetMemUsage() { return m_size.x * m_size.y * 4; } virtual u32 GetMemUsage() { return m_size.x * m_size.y * 4; }
}; };

6
pcsx2/GS/Renderers/Common/GSVertex.h
View File

@ -31,8 +31,8 @@ struct alignas(32) GSVertex
GIFRegST ST; // S:0, T:4 GIFRegST ST; // S:0, T:4
GIFRegRGBAQ RGBAQ; // RGBA:8, Q:12 GIFRegRGBAQ RGBAQ; // RGBA:8, Q:12
GIFRegXYZ XYZ; // XY:16, Z:20 GIFRegXYZ XYZ; // XY:16, Z:20
union { uint32 UV; struct { uint16 U, V; }; }; // UV:24 union { u32 UV; struct { u16 U, V; }; }; // UV:24
uint32 FOG; // FOG:28 u32 FOG; // FOG:28
}; };
#if _M_SSE >= 0x500 #if _M_SSE >= 0x500
@ -73,7 +73,7 @@ struct alignas(32) GSVertexPT1
GSVector4 p; GSVector4 p;
GSVector2 t; GSVector2 t;
char pad[4]; char pad[4];
union { uint32 c; struct { uint8 r, g, b, a; }; }; union { u32 c; struct { u8 r, g, b, a; }; };
}; };
struct GSVertexPT2 struct GSVertexPT2

24
pcsx2/GS/Renderers/Common/GSVertexTrace.cpp
View File

@ -47,14 +47,14 @@ GSVertexTrace::GSVertexTrace(const GSState* state)
InitUpdate(GS_SPRITE_CLASS); InitUpdate(GS_SPRITE_CLASS);
} }
void GSVertexTrace::Update(const void* vertex, const uint32* index, int v_count, int i_count, GS_PRIM_CLASS primclass) void GSVertexTrace::Update(const void* vertex, const u32* index, int v_count, int i_count, GS_PRIM_CLASS primclass)
{ {
m_primclass = primclass; m_primclass = primclass;
uint32 iip = m_state->PRIM->IIP; u32 iip = m_state->PRIM->IIP;
uint32 tme = m_state->PRIM->TME; u32 tme = m_state->PRIM->TME;
uint32 fst = m_state->PRIM->FST; u32 fst = m_state->PRIM->FST;
uint32 color = !(m_state->PRIM->TME && m_state->m_context->TEX0.TFX == TFX_DECAL && m_state->m_context->TEX0.TCC); u32 color = !(m_state->PRIM->TME && m_state->m_context->TEX0.TFX == TFX_DECAL && m_state->m_context->TEX0.TCC);
(this->*m_fmm[color][fst][tme][iip][primclass])(vertex, index, i_count); (this->*m_fmm[color][fst][tme][iip][primclass])(vertex, index, i_count);
@ -148,8 +148,8 @@ void GSVertexTrace::Update(const void* vertex, const uint32* index, int v_count,
} }
} }
template <GS_PRIM_CLASS primclass, uint32 iip, uint32 tme, uint32 fst, uint32 color> template <GS_PRIM_CLASS primclass, u32 iip, u32 tme, u32 fst, u32 color>
void GSVertexTrace::FindMinMax(const void* vertex, const uint32* index, int count) void GSVertexTrace::FindMinMax(const void* vertex, const u32* index, int count)
{ {
const GSDrawingContext* context = m_state->m_context; const GSDrawingContext* context = m_state->m_context;
@ -184,8 +184,8 @@ void GSVertexTrace::FindMinMax(const void* vertex, const uint32* index, int coun
{ {
if (color) if (color)
{ {
GSVector4i c0 = GSVector4i::load(v0.RGBAQ.u32[0]); GSVector4i c0 = GSVector4i::load(v0.RGBAQ.U32[0]);
GSVector4i c1 = GSVector4i::load(v1.RGBAQ.u32[0]); GSVector4i c1 = GSVector4i::load(v1.RGBAQ.U32[0]);
if (iip || finalVertex) if (iip || finalVertex)
{ {
cmin = cmin.min_u8(c0.min_u8(c1)); cmin = cmin.min_u8(c0.min_u8(c1));
@ -304,8 +304,8 @@ void GSVertexTrace::FindMinMax(const void* vertex, const uint32* index, int coun
m_max.p = (GSVector4(pmax) - o) * s; m_max.p = (GSVector4(pmax) - o) * s;
// Fix signed int conversion // Fix signed int conversion
m_min.p = m_min.p.insert32<0, 2>(GSVector4::load((float)(uint32)pmin.extract32<2>())); m_min.p = m_min.p.insert32<0, 2>(GSVector4::load((float)(u32)pmin.extract32<2>()));
m_max.p = m_max.p.insert32<0, 2>(GSVector4::load((float)(uint32)pmax.extract32<2>())); m_max.p = m_max.p.insert32<0, 2>(GSVector4::load((float)(u32)pmax.extract32<2>()));
if (tme) if (tme)
{ {
@ -354,7 +354,7 @@ void GSVertexTrace::CorrectDepthTrace(const void* vertex, int count)
const GSVertex* RESTRICT v = (GSVertex*)vertex; const GSVertex* RESTRICT v = (GSVertex*)vertex;
uint32 z = v[0].XYZ.Z; u32 z = v[0].XYZ.Z;
// ought to check only 1/2 for sprite // ought to check only 1/2 for sprite
if (z & 1) if (z & 1)

16
pcsx2/GS/Renderers/Common/GSVertexTrace.h
View File

@ -45,12 +45,12 @@ protected:
static const GSVector4 s_minmax; static const GSVector4 s_minmax;
typedef void (GSVertexTrace::*FindMinMaxPtr)(const void* vertex, const uint32* index, int count); typedef void (GSVertexTrace::*FindMinMaxPtr)(const void* vertex, const u32* index, int count);
FindMinMaxPtr m_fmm[2][2][2][2][4]; FindMinMaxPtr m_fmm[2][2][2][2][4];
template <GS_PRIM_CLASS primclass, uint32 iip, uint32 tme, uint32 fst, uint32 color> template <GS_PRIM_CLASS primclass, u32 iip, u32 tme, u32 fst, u32 color>
void FindMinMax(const void* vertex, const uint32* index, int count); void FindMinMax(const void* vertex, const u32* index, int count);
public: public:
GS_PRIM_CLASS m_primclass; GS_PRIM_CLASS m_primclass;
@ -61,14 +61,14 @@ public:
union union
{ {
uint32 value; u32 value;
struct { uint32 r:4, g:4, b:4, a:4, x:1, y:1, z:1, f:1, s:1, t:1, q:1, _pad:1; }; struct { u32 r:4, g:4, b:4, a:4, x:1, y:1, z:1, f:1, s:1, t:1, q:1, _pad:1; };
struct { uint32 rgba:16, xyzf:4, stq:4; }; struct { u32 rgba:16, xyzf:4, stq:4; };
} m_eq; } m_eq;
union union
{ {
struct { uint32 mmag:1, mmin:1, linear:1, opt_linear:1; }; struct { u32 mmag:1, mmin:1, linear:1, opt_linear:1; };
} m_filter; } m_filter;
GSVector2 m_lod; // x = min, y = max GSVector2 m_lod; // x = min, y = max
@ -77,7 +77,7 @@ public:
GSVertexTrace(const GSState* state); GSVertexTrace(const GSState* state);
virtual ~GSVertexTrace() {} virtual ~GSVertexTrace() {}
void Update(const void* vertex, const uint32* index, int v_count, int i_count, GS_PRIM_CLASS primclass); void Update(const void* vertex, const u32* index, int v_count, int i_count, GS_PRIM_CLASS primclass);
bool IsLinear() const { return m_filter.opt_linear; } bool IsLinear() const { return m_filter.opt_linear; }
bool IsRealLinear() const { return m_filter.linear; } bool IsRealLinear() const { return m_filter.linear; }

22
pcsx2/GS/Renderers/DX11/GSDevice11.cpp
View File

@ -113,7 +113,7 @@ bool GSDevice11::Create(const WindowInfo& wi)
// device creation // device creation
{ {
uint32 flags = D3D11_CREATE_DEVICE_SINGLETHREADED; u32 flags = D3D11_CREATE_DEVICE_SINGLETHREADED;
if(enable_debugging) if(enable_debugging)
flags |= D3D11_CREATE_DEVICE_DEBUG; flags |= D3D11_CREATE_DEVICE_DEBUG;
@ -536,7 +536,7 @@ void GSDevice11::ClearRenderTarget(GSTexture* t, const GSVector4& c)
m_ctx->ClearRenderTargetView(*(GSTexture11*)t, c.v); m_ctx->ClearRenderTargetView(*(GSTexture11*)t, c.v);
} }
void GSDevice11::ClearRenderTarget(GSTexture* t, uint32 c) void GSDevice11::ClearRenderTarget(GSTexture* t, u32 c)
{ {
if (!t) if (!t)
return; return;
@ -552,7 +552,7 @@ void GSDevice11::ClearDepth(GSTexture* t)
m_ctx->ClearDepthStencilView(*(GSTexture11*)t, D3D11_CLEAR_DEPTH, 0.0f, 0); m_ctx->ClearDepthStencilView(*(GSTexture11*)t, D3D11_CLEAR_DEPTH, 0.0f, 0);
} }
void GSDevice11::ClearStencil(GSTexture* t, uint8 c) void GSDevice11::ClearStencil(GSTexture* t, u8 c)
{ {
if (!t) if (!t)
return; return;
@ -714,7 +714,7 @@ void GSDevice11::StretchRect(GSTexture* sTex, const GSVector4& sRect, GSTexture*
{ {
D3D11_BLEND_DESC bd = {}; D3D11_BLEND_DESC bd = {};
uint8 write_mask = 0; u8 write_mask = 0;
if (red) write_mask |= D3D11_COLOR_WRITE_ENABLE_RED; if (red) write_mask |= D3D11_COLOR_WRITE_ENABLE_RED;
if (green) write_mask |= D3D11_COLOR_WRITE_ENABLE_GREEN; if (green) write_mask |= D3D11_COLOR_WRITE_ENABLE_GREEN;
@ -1087,7 +1087,7 @@ bool GSDevice11::IAMapVertexBuffer(void** vertex, size_t stride, size_t count)
return false; return false;
} }
*vertex = (uint8*)m.pData + m_vertex.start * stride; *vertex = (u8*)m.pData + m_vertex.start * stride;
m_vertex.count = count; m_vertex.count = count;
m_vertex.stride = stride; m_vertex.stride = stride;
@ -1109,8 +1109,8 @@ void GSDevice11::IASetVertexBuffer(ID3D11Buffer* vb, size_t stride)
m_state.vb = vb; m_state.vb = vb;
m_state.vb_stride = stride; m_state.vb_stride = stride;
const uint32 stride2 = stride; const u32 stride2 = stride;
const uint32 offset = 0; const u32 offset = 0;
m_ctx->IASetVertexBuffers(0, 1, &vb, &stride2, &offset); m_ctx->IASetVertexBuffers(0, 1, &vb, &stride2, &offset);
} }
@ -1135,7 +1135,7 @@ void GSDevice11::IASetIndexBuffer(const void* index, size_t count)
memset(&bd, 0, sizeof(bd)); memset(&bd, 0, sizeof(bd));
bd.Usage = D3D11_USAGE_DYNAMIC; bd.Usage = D3D11_USAGE_DYNAMIC;
bd.ByteWidth = m_index.limit * sizeof(uint32); bd.ByteWidth = m_index.limit * sizeof(u32);
bd.BindFlags = D3D11_BIND_INDEX_BUFFER; bd.BindFlags = D3D11_BIND_INDEX_BUFFER;
bd.CPUAccessFlags = D3D11_CPU_ACCESS_WRITE; bd.CPUAccessFlags = D3D11_CPU_ACCESS_WRITE;
@ -1158,7 +1158,7 @@ void GSDevice11::IASetIndexBuffer(const void* index, size_t count)
if (SUCCEEDED(m_ctx->Map(m_ib.get(), 0, type, 0, &m))) if (SUCCEEDED(m_ctx->Map(m_ib.get(), 0, type, 0, &m)))
{ {
memcpy((uint8*)m.pData + m_index.start * sizeof(uint32), index, count * sizeof(uint32)); memcpy((u8*)m.pData + m_index.start * sizeof(u32), index, count * sizeof(u32));
m_ctx->Unmap(m_ib.get(), 0); m_ctx->Unmap(m_ib.get(), 0);
} }
@ -1297,7 +1297,7 @@ void GSDevice11::PSUpdateShaderState()
m_ctx->PSSetSamplers(0, std::size(m_state.ps_ss), m_state.ps_ss); m_ctx->PSSetSamplers(0, std::size(m_state.ps_ss), m_state.ps_ss);
} }
void GSDevice11::OMSetDepthStencilState(ID3D11DepthStencilState* dss, uint8 sref) void GSDevice11::OMSetDepthStencilState(ID3D11DepthStencilState* dss, u8 sref)
{ {
if (m_state.dss != dss || m_state.sref != sref) if (m_state.dss != dss || m_state.sref != sref)
{ {
@ -1464,7 +1464,7 @@ void GSDevice11::CompileShader(const std::vector<char>& source, const char* fn,
throw GSRecoverableError(); throw GSRecoverableError();
} }
uint16 GSDevice11::ConvertBlendEnum(uint16 generic) u16 GSDevice11::ConvertBlendEnum(u16 generic)
{ {
switch (generic) switch (generic)
{ {

168
pcsx2/GS/Renderers/DX11/GSDevice11.h
View File

@ -73,22 +73,22 @@ public:
{ {
struct struct
{ {
uint32 tme : 1; u32 tme : 1;
uint32 fst : 1; u32 fst : 1;
uint32 _free : 30; u32 _free : 30;
}; };
uint32 key; u32 key;
}; };
operator uint32() const { return key; } operator u32() const { return key; }
VSSelector() VSSelector()
: key(0) : key(0)
{ {
} }
VSSelector(uint32 k) VSSelector(u32 k)
: key(k) : key(k)
{ {
} }
@ -179,25 +179,25 @@ public:
{ {
struct struct
{ {
uint32 iip : 1; u32 iip : 1;
uint32 prim : 2; u32 prim : 2;
uint32 point : 1; u32 point : 1;
uint32 line : 1; u32 line : 1;
uint32 cpu_sprite : 1; u32 cpu_sprite : 1;
uint32 _free : 26; u32 _free : 26;
}; };
uint32 key; u32 key;
}; };
operator uint32() { return key; } operator u32() { return key; }
GSSelector() GSSelector()
: key(0) : key(0)
{ {
} }
GSSelector(uint32 k) GSSelector(u32 k)
: key(k) : key(k)
{ {
} }
@ -211,61 +211,61 @@ public:
{ {
// *** Word 1 // *** Word 1
// Format // Format
uint32 fmt : 4; u32 fmt : 4;
uint32 dfmt : 2; u32 dfmt : 2;
uint32 depth_fmt : 2; u32 depth_fmt : 2;
// Alpha extension/Correction // Alpha extension/Correction
uint32 aem : 1; u32 aem : 1;
uint32 fba : 1; u32 fba : 1;
// Fog // Fog
uint32 fog : 1; u32 fog : 1;
// Pixel test // Pixel test
uint32 atst : 3; u32 atst : 3;
// Color sampling // Color sampling
uint32 fst : 1; u32 fst : 1;
uint32 tfx : 3; u32 tfx : 3;
uint32 tcc : 1; u32 tcc : 1;
uint32 wms : 2; u32 wms : 2;
uint32 wmt : 2; u32 wmt : 2;
uint32 ltf : 1; u32 ltf : 1;
// Shuffle and fbmask effect // Shuffle and fbmask effect
uint32 shuffle : 1; u32 shuffle : 1;
uint32 read_ba : 1; u32 read_ba : 1;
uint32 fbmask : 1; u32 fbmask : 1;
// Blend and Colclip // Blend and Colclip
uint32 hdr : 1; u32 hdr : 1;
uint32 blend_a : 2; u32 blend_a : 2;
uint32 blend_b : 2; // bit30/31 u32 blend_b : 2; // bit30/31
uint32 blend_c : 2; // bit0 u32 blend_c : 2; // bit0
uint32 blend_d : 2; u32 blend_d : 2;
uint32 clr1 : 1; u32 clr1 : 1;
uint32 colclip : 1; u32 colclip : 1;
uint32 pabe : 1; u32 pabe : 1;
// Others ways to fetch the texture // Others ways to fetch the texture
uint32 channel : 3; u32 channel : 3;
// Dithering // Dithering
uint32 dither : 2; u32 dither : 2;
// Depth clamp // Depth clamp
uint32 zclamp : 1; u32 zclamp : 1;
// Hack // Hack
uint32 tcoffsethack : 1; u32 tcoffsethack : 1;
uint32 urban_chaos_hle : 1; u32 urban_chaos_hle : 1;
uint32 tales_of_abyss_hle : 1; u32 tales_of_abyss_hle : 1;
uint32 point_sampler : 1; u32 point_sampler : 1;
uint32 invalid_tex0 : 1; // Lupin the 3rd u32 invalid_tex0 : 1; // Lupin the 3rd
uint32 _free : 14; u32 _free : 14;
}; };
uint64 key; u64 key;
}; };
operator uint64() { return key; } operator u64() { return key; }
PSSelector() PSSelector()
: key(0) : key(0)
@ -279,15 +279,15 @@ public:
{ {
struct struct
{ {
uint32 tau : 1; u32 tau : 1;
uint32 tav : 1; u32 tav : 1;
uint32 ltf : 1; u32 ltf : 1;
}; };
uint32 key; u32 key;
}; };
operator uint32() { return key & 0x7; } operator u32() { return key & 0x7; }
PSSamplerSelector() PSSamplerSelector()
: key(0) : key(0)
@ -301,17 +301,17 @@ public:
{ {
struct struct
{ {
uint32 ztst : 2; u32 ztst : 2;
uint32 zwe : 1; u32 zwe : 1;
uint32 date : 1; u32 date : 1;
uint32 fba : 1; u32 fba : 1;
uint32 date_one : 1; u32 date_one : 1;
}; };
uint32 key; u32 key;
}; };
operator uint32() { return key & 0x3f; } operator u32() { return key & 0x3f; }
OMDepthStencilSelector() OMDepthStencilSelector()
: key(0) : key(0)
@ -326,26 +326,26 @@ public:
struct struct
{ {
// Color mask // Color mask
uint32 wr : 1; u32 wr : 1;
uint32 wg : 1; u32 wg : 1;
uint32 wb : 1; u32 wb : 1;
uint32 wa : 1; u32 wa : 1;
// Alpha blending // Alpha blending
uint32 blend_index : 7; u32 blend_index : 7;
uint32 abe : 1; u32 abe : 1;
uint32 accu_blend : 1; u32 accu_blend : 1;
}; };
struct struct
{ {
// Color mask // Color mask
uint32 wrgba : 4; u32 wrgba : 4;
}; };
uint32 key; u32 key;
}; };
operator uint32() { return key & 0x1fff; } operator u32() { return key & 0x1fff; }
OMBlendSelector() OMBlendSelector()
: key(0) : key(0)
@ -405,7 +405,7 @@ private:
void BeforeDraw(); void BeforeDraw();
void AfterDraw(); void AfterDraw();
uint16 ConvertBlendEnum(uint16 generic) final; u16 ConvertBlendEnum(u16 generic) final;
wil::com_ptr_nothrow<ID3D11Device> m_dev; wil::com_ptr_nothrow<ID3D11Device> m_dev;
wil::com_ptr_nothrow<ID3D11DeviceContext> m_ctx; wil::com_ptr_nothrow<ID3D11DeviceContext> m_ctx;
@ -432,7 +432,7 @@ private:
GSVector2i viewport; GSVector2i viewport;
GSVector4i scissor; GSVector4i scissor;
ID3D11DepthStencilState* dss; ID3D11DepthStencilState* dss;
uint8 sref; u8 sref;
ID3D11BlendState* bs; ID3D11BlendState* bs;
float bf; float bf;
ID3D11RenderTargetView* rt_view; ID3D11RenderTargetView* rt_view;
@ -494,16 +494,16 @@ private:
// Shaders... // Shaders...
std::unordered_map<uint32, GSVertexShader11> m_vs; std::unordered_map<u32, GSVertexShader11> m_vs;
wil::com_ptr_nothrow<ID3D11Buffer> m_vs_cb; wil::com_ptr_nothrow<ID3D11Buffer> m_vs_cb;
std::unordered_map<uint32, wil::com_ptr_nothrow<ID3D11GeometryShader>> m_gs; std::unordered_map<u32, wil::com_ptr_nothrow<ID3D11GeometryShader>> m_gs;
wil::com_ptr_nothrow<ID3D11Buffer> m_gs_cb; wil::com_ptr_nothrow<ID3D11Buffer> m_gs_cb;
std::unordered_map<uint64, wil::com_ptr_nothrow<ID3D11PixelShader>> m_ps; std::unordered_map<u64, wil::com_ptr_nothrow<ID3D11PixelShader>> m_ps;
wil::com_ptr_nothrow<ID3D11Buffer> m_ps_cb; wil::com_ptr_nothrow<ID3D11Buffer> m_ps_cb;
std::unordered_map<uint32, wil::com_ptr_nothrow<ID3D11SamplerState>> m_ps_ss; std::unordered_map<u32, wil::com_ptr_nothrow<ID3D11SamplerState>> m_ps_ss;
wil::com_ptr_nothrow<ID3D11SamplerState> m_palette_ss; wil::com_ptr_nothrow<ID3D11SamplerState> m_palette_ss;
std::unordered_map<uint32, wil::com_ptr_nothrow<ID3D11DepthStencilState>> m_om_dss; std::unordered_map<u32, wil::com_ptr_nothrow<ID3D11DepthStencilState>> m_om_dss;
std::unordered_map<uint32, wil::com_ptr_nothrow<ID3D11BlendState>> m_om_bs; std::unordered_map<u32, wil::com_ptr_nothrow<ID3D11BlendState>> m_om_bs;
VSConstantBuffer m_vs_cb_cache; VSConstantBuffer m_vs_cb_cache;
GSConstantBuffer m_gs_cb_cache; GSConstantBuffer m_gs_cb_cache;
@ -535,9 +535,9 @@ public:
void DrawIndexedPrimitive(int offset, int count) final; void DrawIndexedPrimitive(int offset, int count) final;
void ClearRenderTarget(GSTexture* t, const GSVector4& c) final; void ClearRenderTarget(GSTexture* t, const GSVector4& c) final;
void ClearRenderTarget(GSTexture* t, uint32 c) final; void ClearRenderTarget(GSTexture* t, u32 c) final;
void ClearDepth(GSTexture* t) final; void ClearDepth(GSTexture* t) final;
void ClearStencil(GSTexture* t, uint8 c) final; void ClearStencil(GSTexture* t, u8 c) final;
GSTexture* CopyOffscreen(GSTexture* src, const GSVector4& sRect, int w, int h, int format = 0, int ps_shader = 0) final; GSTexture* CopyOffscreen(GSTexture* src, const GSVector4& sRect, int w, int h, int format = 0, int ps_shader = 0) final;
@ -571,7 +571,7 @@ public:
void PSUpdateShaderState(); void PSUpdateShaderState();
void PSSetSamplerState(ID3D11SamplerState* ss0, ID3D11SamplerState* ss1); void PSSetSamplerState(ID3D11SamplerState* ss0, ID3D11SamplerState* ss1);
void OMSetDepthStencilState(ID3D11DepthStencilState* dss, uint8 sref); void OMSetDepthStencilState(ID3D11DepthStencilState* dss, u8 sref);
void OMSetBlendState(ID3D11BlendState* bs, float bf); void OMSetBlendState(ID3D11BlendState* bs, float bf);
void OMSetRenderTargets(GSTexture* rt, GSTexture* ds, const GSVector4i* scissor = NULL) final; void OMSetRenderTargets(GSTexture* rt, GSTexture* ds, const GSVector4i* scissor = NULL) final;
@ -579,7 +579,7 @@ public:
void SetupVS(VSSelector sel, const VSConstantBuffer* cb); void SetupVS(VSSelector sel, const VSConstantBuffer* cb);
void SetupGS(GSSelector sel, const GSConstantBuffer* cb); void SetupGS(GSSelector sel, const GSConstantBuffer* cb);
void SetupPS(PSSelector sel, const PSConstantBuffer* cb, PSSamplerSelector ssel); void SetupPS(PSSelector sel, const PSConstantBuffer* cb, PSSamplerSelector ssel);
void SetupOM(OMDepthStencilSelector dssel, OMBlendSelector bsel, uint8 afix); void SetupOM(OMDepthStencilSelector dssel, OMBlendSelector bsel, u8 afix);
ID3D11Device* operator->() { return m_dev.get(); } ID3D11Device* operator->() { return m_dev.get(); }
operator ID3D11Device*() { return m_dev.get(); } operator ID3D11Device*() { return m_dev.get(); }

34
pcsx2/GS/Renderers/DX11/GSRendererDX11.cpp
View File

@ -117,8 +117,8 @@ void GSRendererDX11::EmulateZbuffer()
// On the real GS we appear to do clamping on the max z value the format allows. // On the real GS we appear to do clamping on the max z value the format allows.
// Clamping is done after rasterization. // Clamping is done after rasterization.
const uint32 max_z = 0xFFFFFFFF >> (GSLocalMemory::m_psm[m_context->ZBUF.PSM].fmt * 8); const u32 max_z = 0xFFFFFFFF >> (GSLocalMemory::m_psm[m_context->ZBUF.PSM].fmt * 8);
const bool clamp_z = (uint32)(GSVector4i(m_vt.m_max.p).z) > max_z; const bool clamp_z = (u32)(GSVector4i(m_vt.m_max.p).z) > max_z;
vs_cb.MaxDepth = GSVector2i(0xFFFFFFFF); vs_cb.MaxDepth = GSVector2i(0xFFFFFFFF);
//ps_cb.Af_MaxDepth.y = 1.0f; //ps_cb.Af_MaxDepth.y = 1.0f;
@ -197,11 +197,11 @@ void GSRendererDX11::EmulateTextureShuffleAndFbmask()
// Please bang my head against the wall! // Please bang my head against the wall!
// 1/ Reduce the frame mask to a 16 bit format // 1/ Reduce the frame mask to a 16 bit format
const uint32& m = m_context->FRAME.FBMSK; const u32& m = m_context->FRAME.FBMSK;
const uint32 fbmask = ((m >> 3) & 0x1F) | ((m >> 6) & 0x3E0) | ((m >> 9) & 0x7C00) | ((m >> 16) & 0x8000); const u32 fbmask = ((m >> 3) & 0x1F) | ((m >> 6) & 0x3E0) | ((m >> 9) & 0x7C00) | ((m >> 16) & 0x8000);
// FIXME GSVector will be nice here // FIXME GSVector will be nice here
const uint8 rg_mask = fbmask & 0xFF; const u8 rg_mask = fbmask & 0xFF;
const uint8 ba_mask = (fbmask >> 8) & 0xFF; const u8 ba_mask = (fbmask >> 8) & 0xFF;
m_om_bsel.wrgba = 0; m_om_bsel.wrgba = 0;
// 2 Select the new mask (Please someone put SSE here) // 2 Select the new mask (Please someone put SSE here)
@ -421,10 +421,10 @@ void GSRendererDX11::EmulateChannelShuffle(GSTexture** rt, const GSTextureCache:
// the rendered size of the framebuffer // the rendered size of the framebuffer
GSVertex* s = &m_vertex.buff[0]; GSVertex* s = &m_vertex.buff[0];
s[0].XYZ.X = (uint16)(m_context->XYOFFSET.OFX + 0); s[0].XYZ.X = (u16)(m_context->XYOFFSET.OFX + 0);
s[1].XYZ.X = (uint16)(m_context->XYOFFSET.OFX + 16384); s[1].XYZ.X = (u16)(m_context->XYOFFSET.OFX + 16384);
s[0].XYZ.Y = (uint16)(m_context->XYOFFSET.OFY + 0); s[0].XYZ.Y = (u16)(m_context->XYOFFSET.OFY + 0);
s[1].XYZ.Y = (uint16)(m_context->XYOFFSET.OFY + 16384); s[1].XYZ.Y = (u16)(m_context->XYOFFSET.OFY + 16384);
m_vertex.head = m_vertex.tail = m_vertex.next = 2; m_vertex.head = m_vertex.tail = m_vertex.next = 2;
m_index.tail = 2; m_index.tail = 2;
@ -448,7 +448,7 @@ void GSRendererDX11::EmulateBlending()
return; return;
m_om_bsel.abe = 1; m_om_bsel.abe = 1;
m_om_bsel.blend_index = uint8(((ALPHA.A * 3 + ALPHA.B) * 3 + ALPHA.C) * 3 + ALPHA.D); m_om_bsel.blend_index = u8(((ALPHA.A * 3 + ALPHA.B) * 3 + ALPHA.C) * 3 + ALPHA.D);
const int blend_flag = m_dev->GetBlendFlags(m_om_bsel.blend_index); const int blend_flag = m_dev->GetBlendFlags(m_om_bsel.blend_index);
// Do the multiplication in shader for blending accumulation: Cs*As + Cd or Cs*Af + Cd // Do the multiplication in shader for blending accumulation: Cs*As + Cd or Cs*Af + Cd
@ -567,8 +567,8 @@ void GSRendererDX11::EmulateTextureSampler(const GSTextureCache::Source* tex)
const GSLocalMemory::psm_t& psm = GSLocalMemory::m_psm[tex->m_TEX0.PSM]; const GSLocalMemory::psm_t& psm = GSLocalMemory::m_psm[tex->m_TEX0.PSM];
const GSLocalMemory::psm_t& cpsm = psm.pal > 0 ? GSLocalMemory::m_psm[m_context->TEX0.CPSM] : psm; const GSLocalMemory::psm_t& cpsm = psm.pal > 0 ? GSLocalMemory::m_psm[m_context->TEX0.CPSM] : psm;
const uint8 wms = m_context->CLAMP.WMS; const u8 wms = m_context->CLAMP.WMS;
const uint8 wmt = m_context->CLAMP.WMT; const u8 wmt = m_context->CLAMP.WMT;
const bool complex_wms_wmt = !!((wms | wmt) & 2); const bool complex_wms_wmt = !!((wms | wmt) & 2);
bool bilinear = m_vt.IsLinear(); bool bilinear = m_vt.IsLinear();
@ -937,7 +937,7 @@ void GSRendererDX11::DrawPrims(GSTexture* rt, GSTexture* ds, GSTextureCache::Sou
{ {
m_ps_sel.fog = 1; m_ps_sel.fog = 1;
const GSVector4 fc = GSVector4::rgba32(m_env.FOGCOL.u32[0]); const GSVector4 fc = GSVector4::rgba32(m_env.FOGCOL.U32[0]);
// Blend AREF to avoid to load a random value for alpha (dirty cache) // Blend AREF to avoid to load a random value for alpha (dirty cache)
ps_cb.FogColor_AREF = fc.blend32<8>(ps_cb.FogColor_AREF); ps_cb.FogColor_AREF = fc.blend32<8>(ps_cb.FogColor_AREF);
} }
@ -948,7 +948,7 @@ void GSRendererDX11::DrawPrims(GSTexture* rt, GSTexture* ds, GSTextureCache::Sou
// pass to handle the depth based on the alpha test. // pass to handle the depth based on the alpha test.
bool ate_RGBA_then_Z = false; bool ate_RGBA_then_Z = false;
bool ate_RGB_then_ZA = false; bool ate_RGB_then_ZA = false;
uint8 ps_atst = 0; u8 ps_atst = 0;
if (ate_first_pass & ate_second_pass) if (ate_first_pass & ate_second_pass)
{ {
// fprintf(stdout, "%d: Complex Alpha Test\n", s_n); // fprintf(stdout, "%d: Complex Alpha Test\n", s_n);
@ -1025,7 +1025,7 @@ void GSRendererDX11::DrawPrims(GSTexture* rt, GSTexture* ds, GSTextureCache::Sou
if (!tex->m_palette) if (!tex->m_palette)
{ {
const uint16 pal = GSLocalMemory::m_psm[tex->m_TEX0.PSM].pal; const u16 pal = GSLocalMemory::m_psm[tex->m_TEX0.PSM].pal;
m_tc->AttachPaletteToSource(tex, pal, true); m_tc->AttachPaletteToSource(tex, pal, true);
} }
} }
@ -1045,7 +1045,7 @@ void GSRendererDX11::DrawPrims(GSTexture* rt, GSTexture* ds, GSTextureCache::Sou
SetupIA(sx, sy); SetupIA(sx, sy);
const uint8 afix = m_context->ALPHA.FIX; const u8 afix = m_context->ALPHA.FIX;
dev->SetupOM(m_om_dssel, m_om_bsel, afix); dev->SetupOM(m_om_dssel, m_om_bsel, afix);
dev->SetupVS(m_vs_sel, &vs_cb); dev->SetupVS(m_vs_sel, &vs_cb);
dev->SetupGS(m_gs_sel, &gs_cb); dev->SetupGS(m_gs_sel, &gs_cb);

14
pcsx2/GS/Renderers/DX11/GSTexture11.cpp
View File

@ -80,7 +80,7 @@ bool GSTexture11::Map(GSMap& m, const GSVector4i* r, int layer)
if (SUCCEEDED(m_ctx->Map(m_texture.get(), subresource, D3D11_MAP_READ_WRITE, 0, &map))) if (SUCCEEDED(m_ctx->Map(m_texture.get(), subresource, D3D11_MAP_READ_WRITE, 0, &map)))
{ {
m.bits = (uint8*)map.pData; m.bits = (u8*)map.pData;
m.pitch = (int)map.RowPitch; m.pitch = (int)map.RowPitch;
m_layer = layer; m_layer = layer;
@ -152,14 +152,14 @@ bool GSTexture11::Save(const std::string& fn)
m_ctx->Unmap(dst.get(), 0); m_ctx->Unmap(dst.get(), 0);
}); });
const uint8* s = static_cast<const uint8*>(sm.pData); const u8* s = static_cast<const u8*>(sm.pData);
uint8* d = static_cast<uint8*>(dm.pData); u8* d = static_cast<u8*>(dm.pData);
for (uint32 y = 0; y < desc.Height; y++, s += sm.RowPitch, d += dm.RowPitch) for (u32 y = 0; y < desc.Height; y++, s += sm.RowPitch, d += dm.RowPitch)
{ {
for (uint32 x = 0; x < desc.Width; x++) for (u32 x = 0; x < desc.Width; x++)
{ {
reinterpret_cast<uint32*>(d)[x] = static_cast<uint32>(ldexpf(reinterpret_cast<const float*>(s)[x * 2], 32)); reinterpret_cast<u32*>(d)[x] = static_cast<u32>(ldexpf(reinterpret_cast<const float*>(s)[x * 2], 32));
} }
} }
@ -193,7 +193,7 @@ bool GSTexture11::Save(const std::string& fn)
} }
int compression = theApp.GetConfigI("png_compression_level"); int compression = theApp.GetConfigI("png_compression_level");
bool success = GSPng::Save(format, fn, static_cast<uint8*>(sm.pData), desc.Width, desc.Height, sm.RowPitch, compression); bool success = GSPng::Save(format, fn, static_cast<u8*>(sm.pData), desc.Width, desc.Height, sm.RowPitch, compression);
m_ctx->Unmap(res.get(), 0); m_ctx->Unmap(res.get(), 0);

2
pcsx2/GS/Renderers/DX11/GSTextureFX11.cpp
View File

@ -279,7 +279,7 @@ void GSDevice11::SetupPS(PSSelector sel, const PSConstantBuffer* cb, PSSamplerSe
PSSetShader(i->second.get(), m_ps_cb.get()); PSSetShader(i->second.get(), m_ps_cb.get());
} }
void GSDevice11::SetupOM(OMDepthStencilSelector dssel, OMBlendSelector bsel, uint8 afix) void GSDevice11::SetupOM(OMDepthStencilSelector dssel, OMBlendSelector bsel, u8 afix)
{ {
auto i = std::as_const(m_om_dss).find(dssel); auto i = std::as_const(m_om_dss).find(dssel);

148
pcsx2/GS/Renderers/HW/GSRendererHW.cpp
View File

@ -188,7 +188,7 @@ GSRendererHW::~GSRendererHW()
delete m_tc; delete m_tc;
} }
void GSRendererHW::SetGameCRC(uint32 crc, int options) void GSRendererHW::SetGameCRC(u32 crc, int options)
{ {
GSRenderer::SetGameCRC(crc, options); GSRenderer::SetGameCRC(crc, options);
@ -389,7 +389,7 @@ void GSRendererHW::Lines2Sprites()
int i = (int)count * 2 - 4; int i = (int)count * 2 - 4;
GSVertex* s = &m_vertex.buff[count - 2]; GSVertex* s = &m_vertex.buff[count - 2];
GSVertex* q = &m_vertex.buff[count * 2 - 4]; GSVertex* q = &m_vertex.buff[count * 2 - 4];
uint32* RESTRICT index = &m_index.buff[count * 3 - 6]; u32* RESTRICT index = &m_index.buff[count * 3 - 6];
for (; i >= 0; i -= 4, s -= 2, q -= 4, index -= 6) for (; i >= 0; i -= 4, s -= 2, q -= 4, index -= 6)
{ {
@ -402,13 +402,13 @@ void GSRendererHW::Lines2Sprites()
if (PRIM->TME && !PRIM->FST) if (PRIM->TME && !PRIM->FST)
{ {
GSVector4 st0 = GSVector4::loadl(&v0.ST.u64); GSVector4 st0 = GSVector4::loadl(&v0.ST.U64);
GSVector4 st1 = GSVector4::loadl(&v1.ST.u64); GSVector4 st1 = GSVector4::loadl(&v1.ST.U64);
GSVector4 Q = GSVector4(v1.RGBAQ.Q, v1.RGBAQ.Q, v1.RGBAQ.Q, v1.RGBAQ.Q); GSVector4 Q = GSVector4(v1.RGBAQ.Q, v1.RGBAQ.Q, v1.RGBAQ.Q, v1.RGBAQ.Q);
GSVector4 st = st0.upld(st1) / Q; GSVector4 st = st0.upld(st1) / Q;
GSVector4::storel(&v0.ST.u64, st); GSVector4::storel(&v0.ST.U64, st);
GSVector4::storeh(&v1.ST.u64, st); GSVector4::storeh(&v1.ST.U64, st);
v0.RGBAQ.Q = 1.0f; v0.RGBAQ.Q = 1.0f;
v1.RGBAQ.Q = 1.0f; v1.RGBAQ.Q = 1.0f;
@ -419,7 +419,7 @@ void GSRendererHW::Lines2Sprites()
// swap x, s, u // swap x, s, u
const uint16 x = v0.XYZ.X; const u16 x = v0.XYZ.X;
v0.XYZ.X = v1.XYZ.X; v0.XYZ.X = v1.XYZ.X;
v1.XYZ.X = x; v1.XYZ.X = x;
@ -427,7 +427,7 @@ void GSRendererHW::Lines2Sprites()
v0.ST.S = v1.ST.S; v0.ST.S = v1.ST.S;
v1.ST.S = s; v1.ST.S = s;
const uint16 u = v0.U; const u16 u = v0.U;
v0.U = v1.U; v0.U = v1.U;
v1.U = u; v1.U = u;
@ -447,9 +447,9 @@ void GSRendererHW::Lines2Sprites()
} }
} }
void GSRendererHW::EmulateAtst(GSVector4& FogColor_AREF, uint8& ps_atst, const bool pass_2) void GSRendererHW::EmulateAtst(GSVector4& FogColor_AREF, u8& ps_atst, const bool pass_2)
{ {
static const uint32 inverted_atst[] = {ATST_ALWAYS, ATST_NEVER, ATST_GEQUAL, ATST_GREATER, ATST_NOTEQUAL, ATST_LESS, ATST_LEQUAL, ATST_EQUAL}; static const u32 inverted_atst[] = {ATST_ALWAYS, ATST_NEVER, ATST_GEQUAL, ATST_GREATER, ATST_NOTEQUAL, ATST_LESS, ATST_LEQUAL, ATST_EQUAL};
if (!m_context->TEST.ATE) if (!m_context->TEST.ATE)
return; return;
@ -570,10 +570,10 @@ void GSRendererHW::ConvertSpriteTextureShuffle(bool& write_ba, bool& read_ba)
GSVector4i tmp(v[i].XYZ.Y, v[i].V, v[i + 1].XYZ.Y, v[i + 1].V); GSVector4i tmp(v[i].XYZ.Y, v[i].V, v[i + 1].XYZ.Y, v[i + 1].V);
tmp = GSVector4i(tmp - offset).srl32(1) + offset; tmp = GSVector4i(tmp - offset).srl32(1) + offset;
v[i].XYZ.Y = (uint16)tmp.x; v[i].XYZ.Y = (u16)tmp.x;
v[i].V = (uint16)tmp.y; v[i].V = (u16)tmp.y;
v[i + 1].XYZ.Y = (uint16)tmp.z; v[i + 1].XYZ.Y = (u16)tmp.z;
v[i + 1].V = (uint16)tmp.w; v[i + 1].V = (u16)tmp.w;
} }
} }
} }
@ -603,9 +603,9 @@ void GSRendererHW::ConvertSpriteTextureShuffle(bool& write_ba, bool& read_ba)
tmp = GSVector4i(tmp - offset).srl32(1) + offset; tmp = GSVector4i(tmp - offset).srl32(1) + offset;
//fprintf(stderr, "Before %d, After %d\n", v[i + 1].XYZ.Y, tmp.y); //fprintf(stderr, "Before %d, After %d\n", v[i + 1].XYZ.Y, tmp.y);
v[i].XYZ.Y = (uint16)tmp.x; v[i].XYZ.Y = (u16)tmp.x;
v[i].ST.T /= 2.0f; v[i].ST.T /= 2.0f;
v[i + 1].XYZ.Y = (uint16)tmp.y; v[i + 1].XYZ.Y = (u16)tmp.y;
v[i + 1].ST.T /= 2.0f; v[i + 1].ST.T /= 2.0f;
} }
} }
@ -757,15 +757,15 @@ void GSRendererHW::MergeSprite(GSTextureCache::Source* tex)
// Replace all sprite with a single fullscreen sprite. // Replace all sprite with a single fullscreen sprite.
GSVertex* s = &m_vertex.buff[0]; GSVertex* s = &m_vertex.buff[0];
s[0].XYZ.X = static_cast<uint16>((16.0f * m_vt.m_min.p.x) + m_context->XYOFFSET.OFX); s[0].XYZ.X = static_cast<u16>((16.0f * m_vt.m_min.p.x) + m_context->XYOFFSET.OFX);
s[1].XYZ.X = static_cast<uint16>((16.0f * m_vt.m_max.p.x) + m_context->XYOFFSET.OFX); s[1].XYZ.X = static_cast<u16>((16.0f * m_vt.m_max.p.x) + m_context->XYOFFSET.OFX);
s[0].XYZ.Y = static_cast<uint16>((16.0f * m_vt.m_min.p.y) + m_context->XYOFFSET.OFY); s[0].XYZ.Y = static_cast<u16>((16.0f * m_vt.m_min.p.y) + m_context->XYOFFSET.OFY);
s[1].XYZ.Y = static_cast<uint16>((16.0f * m_vt.m_max.p.y) + m_context->XYOFFSET.OFY); s[1].XYZ.Y = static_cast<u16>((16.0f * m_vt.m_max.p.y) + m_context->XYOFFSET.OFY);
s[0].U = static_cast<uint16>(16.0f * m_vt.m_min.t.x); s[0].U = static_cast<u16>(16.0f * m_vt.m_min.t.x);
s[0].V = static_cast<uint16>(16.0f * m_vt.m_min.t.y); s[0].V = static_cast<u16>(16.0f * m_vt.m_min.t.y);
s[1].U = static_cast<uint16>(16.0f * m_vt.m_max.t.x); s[1].U = static_cast<u16>(16.0f * m_vt.m_max.t.x);
s[1].V = static_cast<uint16>(16.0f * m_vt.m_max.t.y); s[1].V = static_cast<u16>(16.0f * m_vt.m_max.t.y);
m_vertex.head = m_vertex.tail = m_vertex.next = 2; m_vertex.head = m_vertex.tail = m_vertex.next = 2;
m_index.tail = 2; m_index.tail = 2;
@ -791,10 +791,10 @@ void GSRendererHW::InvalidateLocalMem(const GIFRegBITBLTBUF& BITBLTBUF, const GS
m_tc->InvalidateLocalMem(m_mem.GetOffset(BITBLTBUF.SBP, BITBLTBUF.SBW, BITBLTBUF.SPSM), r); m_tc->InvalidateLocalMem(m_mem.GetOffset(BITBLTBUF.SBP, BITBLTBUF.SBW, BITBLTBUF.SPSM), r);
} }
uint16 GSRendererHW::Interpolate_UV(float alpha, int t0, int t1) u16 GSRendererHW::Interpolate_UV(float alpha, int t0, int t1)
{ {
const float t = (1.0f - alpha) * t0 + alpha * t1; const float t = (1.0f - alpha) * t0 + alpha * t1;
return (uint16)t & ~0xF; // cheap rounding return (u16)t & ~0xF; // cheap rounding
} }
float GSRendererHW::alpha0(int L, int X0, int X1) float GSRendererHW::alpha0(int L, int X0, int X1)
@ -899,28 +899,28 @@ void GSRendererHW::SwSpriteRender()
const bool fb_mask_enabled = m_context->FRAME.FBMSK != 0x0; const bool fb_mask_enabled = m_context->FRAME.FBMSK != 0x0;
const GSVector4i fb_mask = GSVector4i(m_context->FRAME.FBMSK).u8to16(); // 0x00AA00BB00GG00RR00AA00BB00GG00RR const GSVector4i fb_mask = GSVector4i(m_context->FRAME.FBMSK).u8to16(); // 0x00AA00BB00GG00RR00AA00BB00GG00RR
const uint8 tex0_tfx = m_context->TEX0.TFX; const u8 tex0_tfx = m_context->TEX0.TFX;
const uint8 tex0_tcc = m_context->TEX0.TCC; const u8 tex0_tcc = m_context->TEX0.TCC;
const uint8 alpha_b = m_context->ALPHA.B; const u8 alpha_b = m_context->ALPHA.B;
const uint8 alpha_c = m_context->ALPHA.C; const u8 alpha_c = m_context->ALPHA.C;
const uint8 alpha_fix = m_context->ALPHA.FIX; const u8 alpha_fix = m_context->ALPHA.FIX;
for (int y = 0; y < h; y++, ++sy, ++dy) for (int y = 0; y < h; y++, ++sy, ++dy)
{ {
auto spa = texture_mapping_enabled ? spo.paMulti(m_mem.m_vm32, sx, sy) : GSOffset::PAPtrHelper<uint32>(); auto spa = texture_mapping_enabled ? spo.paMulti(m_mem.m_vm32, sx, sy) : GSOffset::PAPtrHelper<u32>();
auto dpa = dpo.paMulti(m_mem.m_vm32, dx, dy); auto dpa = dpo.paMulti(m_mem.m_vm32, dx, dy);
ASSERT(w % 2 == 0); ASSERT(w % 2 == 0);
for (int x = 0; x < w; x += 2) for (int x = 0; x < w; x += 2)
{ {
uint32* di = dpa.value(x); u32* di = dpa.value(x);
ASSERT(*di + 1 == *dpa.value(x + 1)); // Destination pixel pair is adjacent in memory ASSERT(*di + 1 == *dpa.value(x + 1)); // Destination pixel pair is adjacent in memory
GSVector4i sc; GSVector4i sc;
if (texture_mapping_enabled) if (texture_mapping_enabled)
{ {
uint32* si = spa.value(x); u32* si = spa.value(x);
// Read 2 source pixel colors // Read 2 source pixel colors
ASSERT((*si + 1) == *spa.value(x + 1)); // Source pixel pair is adjacent in memory ASSERT((*si + 1) == *spa.value(x + 1)); // Source pixel pair is adjacent in memory
sc = GSVector4i::loadl(si).u8to16(); // 0x00AA00BB00GG00RR00aa00bb00gg00rr sc = GSVector4i::loadl(si).u8to16(); // 0x00AA00BB00GG00RR00aa00bb00gg00rr
@ -1079,8 +1079,8 @@ void GSRendererHW::RoundSpriteOffset()
const int Lx = (v[i + 1].XYZ.X - v[i].XYZ.X); const int Lx = (v[i + 1].XYZ.X - v[i].XYZ.X);
const float ax0 = alpha0(Lx, X0, X1); const float ax0 = alpha0(Lx, X0, X1);
const float ax1 = alpha1(Lx, X0, X1); const float ax1 = alpha1(Lx, X0, X1);
const uint16 tx0 = Interpolate_UV(ax0, v[i].U, v[i + 1].U); const u16 tx0 = Interpolate_UV(ax0, v[i].U, v[i + 1].U);
const uint16 tx1 = Interpolate_UV(ax1, v[i].U, v[i + 1].U); const u16 tx1 = Interpolate_UV(ax1, v[i].U, v[i + 1].U);
#ifdef DEBUG_U #ifdef DEBUG_U
if (debug) if (debug)
{ {
@ -1096,8 +1096,8 @@ void GSRendererHW::RoundSpriteOffset()
const int Ly = (v[i + 1].XYZ.Y - v[i].XYZ.Y); const int Ly = (v[i + 1].XYZ.Y - v[i].XYZ.Y);
const float ay0 = alpha0(Ly, Y0, Y1); const float ay0 = alpha0(Ly, Y0, Y1);
const float ay1 = alpha1(Ly, Y0, Y1); const float ay1 = alpha1(Ly, Y0, Y1);
const uint16 ty0 = Interpolate_UV(ay0, v[i].V, v[i + 1].V); const u16 ty0 = Interpolate_UV(ay0, v[i].V, v[i + 1].V);
const uint16 ty1 = Interpolate_UV(ay1, v[i].V, v[i + 1].V); const u16 ty1 = Interpolate_UV(ay1, v[i].V, v[i + 1].V);
#ifdef DEBUG_V #ifdef DEBUG_V
if (debug) if (debug)
{ {
@ -1206,8 +1206,8 @@ void GSRendererHW::Draw()
const GIFRegFRAME FRAME = context->FRAME; const GIFRegFRAME FRAME = context->FRAME;
const GIFRegZBUF ZBUF = context->ZBUF; const GIFRegZBUF ZBUF = context->ZBUF;
uint32 fm = context->FRAME.FBMSK; u32 fm = context->FRAME.FBMSK;
uint32 zm = context->ZBUF.ZMSK || context->TEST.ZTE == 0 ? 0xffffffff : 0; u32 zm = context->ZBUF.ZMSK || context->TEST.ZTE == 0 ? 0xffffffff : 0;
// Note required to compute TryAlphaTest below. So do it now. // Note required to compute TryAlphaTest below. So do it now.
if (PRIM->TME && tex_psm.pal > 0) if (PRIM->TME && tex_psm.pal > 0)
@ -1469,7 +1469,7 @@ void GSRendererHW::Draw()
if (s_dump) if (s_dump)
{ {
const uint64 frame = m_perfmon.GetFrame(); const u64 frame = m_perfmon.GetFrame();
std::string s; std::string s;
@ -1651,7 +1651,7 @@ void GSRendererHW::Draw()
if (s_dump) if (s_dump)
{ {
const uint64 frame = m_perfmon.GetFrame(); const u64 frame = m_perfmon.GetFrame();
std::string s; std::string s;
@ -1715,7 +1715,7 @@ GSRendererHW::Hacks::Hacks()
void GSRendererHW::Hacks::SetGameCRC(const CRC::Game& game) void GSRendererHW::Hacks::SetGameCRC(const CRC::Game& game)
{ {
const uint32 hash = (uint32)((game.region << 24) | game.title); const u32 hash = (u32)((game.region << 24) | game.title);
m_oi = m_oi_map[hash]; m_oi = m_oi_map[hash];
m_oo = m_oo_map[hash]; m_oo = m_oo_map[hash];
@ -1744,7 +1744,7 @@ void GSRendererHW::OI_DoubleHalfClear(GSTexture* rt, GSTexture* ds)
//const GSLocalMemory::psm_t& depth_psm = GSLocalMemory::m_psm[m_context->ZBUF.PSM]; //const GSLocalMemory::psm_t& depth_psm = GSLocalMemory::m_psm[m_context->ZBUF.PSM];
// Z and color must be constant and the same // Z and color must be constant and the same
if (m_vt.m_eq.rgba != 0xFFFF || !m_vt.m_eq.z || v[1].XYZ.Z != v[1].RGBAQ.u32[0]) if (m_vt.m_eq.rgba != 0xFFFF || !m_vt.m_eq.z || v[1].XYZ.Z != v[1].RGBAQ.U32[0])
return; return;
// Format doesn't have the same size. It smells fishy (xmen...) // Format doesn't have the same size. It smells fishy (xmen...)
@ -1752,12 +1752,12 @@ void GSRendererHW::OI_DoubleHalfClear(GSTexture* rt, GSTexture* ds)
// return; // return;
// Size of the current draw // Size of the current draw
const uint32 w_pages = static_cast<uint32>(roundf(m_vt.m_max.p.x / frame_psm.pgs.x)); const u32 w_pages = static_cast<u32>(roundf(m_vt.m_max.p.x / frame_psm.pgs.x));
const uint32 h_pages = static_cast<uint32>(roundf(m_vt.m_max.p.y / frame_psm.pgs.y)); const u32 h_pages = static_cast<u32>(roundf(m_vt.m_max.p.y / frame_psm.pgs.y));
const uint32 written_pages = w_pages * h_pages; const u32 written_pages = w_pages * h_pages;
// Frame and depth pointer can be inverted // Frame and depth pointer can be inverted
uint32 base = 0, half = 0; u32 base = 0, half = 0;
if (m_context->FRAME.FBP > m_context->ZBUF.ZBP) if (m_context->FRAME.FBP > m_context->ZBUF.ZBP)
{ {
base = m_context->ZBUF.ZBP; base = m_context->ZBUF.ZBP;
@ -1772,7 +1772,7 @@ void GSRendererHW::OI_DoubleHalfClear(GSTexture* rt, GSTexture* ds)
// If both buffers are side by side we can expect a fast clear in on-going // If both buffers are side by side we can expect a fast clear in on-going
if (half <= (base + written_pages)) if (half <= (base + written_pages))
{ {
const uint32 color = v[1].RGBAQ.u32[0]; const u32 color = v[1].RGBAQ.U32[0];
const bool clear_depth = (m_context->FRAME.FBP > m_context->ZBUF.ZBP); const bool clear_depth = (m_context->FRAME.FBP > m_context->ZBUF.ZBP);
GL_INS("OI_DoubleHalfClear:%s: base %x half %x. w_pages %d h_pages %d fbw %d. Color %x", GL_INS("OI_DoubleHalfClear:%s: base %x half %x. w_pages %d h_pages %d fbw %d. Color %x",
@ -1953,7 +1953,7 @@ bool GSRendererHW::OI_BigMuthaTruckers(GSTexture* rt, GSTexture* ds, GSTextureCa
const size_t count = m_vertex.next; const size_t count = m_vertex.next;
GSVertex* v = &m_vertex.buff[0]; GSVertex* v = &m_vertex.buff[0];
const uint16 offset = (uint16)m_r.y * 16; const u16 offset = (u16)m_r.y * 16;
for (size_t i = 0; i < count; i++) for (size_t i = 0; i < count; i++)
v[i].V += offset; v[i].V += offset;
@ -1978,7 +1978,7 @@ bool GSRendererHW::OI_DBZBTGames(GSTexture* rt, GSTexture* ds, GSTextureCache::S
bool GSRendererHW::OI_FFXII(GSTexture* rt, GSTexture* ds, GSTextureCache::Source* t) bool GSRendererHW::OI_FFXII(GSTexture* rt, GSTexture* ds, GSTextureCache::Source* t)
{ {
static uint32* video = NULL; static u32* video = NULL;
static size_t lines = 0; static size_t lines = 0;
if (lines == 0) if (lines == 0)
@ -1997,7 +1997,7 @@ bool GSRendererHW::OI_FFXII(GSTexture* rt, GSTexture* ds, GSTextureCache::Source
// incoming pixels are stored in columns, one column is 16x512, total res 448x512 or 448x454 // incoming pixels are stored in columns, one column is 16x512, total res 448x512 or 448x454
if (!video) if (!video)
video = new uint32[512 * 512]; video = new u32[512 * 512];
const int ox = m_context->XYOFFSET.OFX - 8; const int ox = m_context->XYOFFSET.OFX - 8;
const int oy = m_context->XYOFFSET.OFY - 8; const int oy = m_context->XYOFFSET.OFY - 8;
@ -2012,7 +2012,7 @@ bool GSRendererHW::OI_FFXII(GSTexture* rt, GSTexture* ds, GSTextureCache::Source
if (x < 0 || x >= 448 || y < 0 || y >= (int)lines) if (x < 0 || x >= 448 || y < 0 || y >= (int)lines)
return false; // le sigh return false; // le sigh
video[(y << 8) + (y << 7) + (y << 6) + x] = v->RGBAQ.u32[0]; video[(y << 8) + (y << 7) + (y << 6) + x] = v->RGBAQ.U32[0];
} }
return false; return false;
@ -2062,9 +2062,9 @@ bool GSRendererHW::OI_FFXII(GSTexture* rt, GSTexture* ds, GSTextureCache::Source
bool GSRendererHW::OI_FFX(GSTexture* rt, GSTexture* ds, GSTextureCache::Source* t) bool GSRendererHW::OI_FFX(GSTexture* rt, GSTexture* ds, GSTextureCache::Source* t)
{ {
const uint32 FBP = m_context->FRAME.Block(); const u32 FBP = m_context->FRAME.Block();
const uint32 ZBP = m_context->ZBUF.Block(); const u32 ZBP = m_context->ZBUF.Block();
const uint32 TBP = m_context->TEX0.TBP0; const u32 TBP = m_context->TEX0.TBP0;
if ((FBP == 0x00d00 || FBP == 0x00000) && ZBP == 0x02100 && PRIM->TME && TBP == 0x01a00 && m_context->TEX0.PSM == PSM_PSMCT16S) if ((FBP == 0x00d00 || FBP == 0x00000) && ZBP == 0x02100 && PRIM->TME && TBP == 0x01a00 && m_context->TEX0.PSM == PSM_PSMCT16S)
{ {
@ -2086,15 +2086,15 @@ bool GSRendererHW::OI_MetalSlug6(GSTexture* rt, GSTexture* ds, GSTextureCache::S
for (int i = (int)m_vertex.next; i > 0; i--, v++) for (int i = (int)m_vertex.next; i > 0; i--, v++)
{ {
const uint32 c = v->RGBAQ.u32[0]; const u32 c = v->RGBAQ.U32[0];
const uint32 r = (c >> 0) & 0xff; const u32 r = (c >> 0) & 0xff;
const uint32 g = (c >> 8) & 0xff; const u32 g = (c >> 8) & 0xff;
const uint32 b = (c >> 16) & 0xff; const u32 b = (c >> 16) & 0xff;
if (r == 0 && g != 0 && b != 0) if (r == 0 && g != 0 && b != 0)
{ {
v->RGBAQ.u32[0] = (c & 0xffffff00) | ((g + b + 1) >> 1); v->RGBAQ.U32[0] = (c & 0xffffff00) | ((g + b + 1) >> 1);
} }
} }
@ -2107,8 +2107,8 @@ bool GSRendererHW::OI_RozenMaidenGebetGarden(GSTexture* rt, GSTexture* ds, GSTex
{ {
if (!PRIM->TME) if (!PRIM->TME)
{ {
const uint32 FBP = m_context->FRAME.Block(); const u32 FBP = m_context->FRAME.Block();
const uint32 ZBP = m_context->ZBUF.Block(); const u32 ZBP = m_context->ZBUF.Block();
if (FBP == 0x008c0 && ZBP == 0x01a40) if (FBP == 0x008c0 && ZBP == 0x01a40)
{ {
@ -2193,8 +2193,8 @@ bool GSRendererHW::OI_PointListPalette(GSTexture* rt, GSTexture* ds, GSTextureCa
{ {
if (m_vt.m_primclass == GS_POINT_CLASS && !PRIM->TME) if (m_vt.m_primclass == GS_POINT_CLASS && !PRIM->TME)
{ {
const uint32 FBP = m_context->FRAME.Block(); const u32 FBP = m_context->FRAME.Block();
const uint32 FBW = m_context->FRAME.FBW; const u32 FBW = m_context->FRAME.FBW;
if (FBP >= 0x03f40 && (FBP & 0x1f) == 0) if (FBP >= 0x03f40 && (FBP & 0x1f) == 0)
{ {
@ -2204,12 +2204,12 @@ bool GSRendererHW::OI_PointListPalette(GSTexture* rt, GSTexture* ds, GSTextureCa
for (int i = 0; i < 16; i++, v++) for (int i = 0; i < 16; i++, v++)
{ {
uint32 c = v->RGBAQ.u32[0]; u32 c = v->RGBAQ.U32[0];
const uint32 a = c >> 24; const u32 a = c >> 24;
c = (a >= 0x80 ? 0xff000000 : (a << 25)) | (c & 0x00ffffff); c = (a >= 0x80 ? 0xff000000 : (a << 25)) | (c & 0x00ffffff);
v->RGBAQ.u32[0] = c; v->RGBAQ.U32[0] = c;
m_mem.WritePixel32(i & 7, i >> 3, c, FBP, FBW); m_mem.WritePixel32(i & 7, i >> 3, c, FBP, FBW);
} }
@ -2224,12 +2224,12 @@ bool GSRendererHW::OI_PointListPalette(GSTexture* rt, GSTexture* ds, GSTextureCa
for (int i = 0; i < 256; i++, v++) for (int i = 0; i < 256; i++, v++)
{ {
uint32 c = v->RGBAQ.u32[0]; u32 c = v->RGBAQ.U32[0];
const uint32 a = c >> 24; const u32 a = c >> 24;
c = (a >= 0x80 ? 0xff000000 : (a << 25)) | (c & 0x00ffffff); c = (a >= 0x80 ? 0xff000000 : (a << 25)) | (c & 0x00ffffff);
v->RGBAQ.u32[0] = c; v->RGBAQ.U32[0] = c;
m_mem.WritePixel32(i & 15, i >> 4, c, FBP, FBW); m_mem.WritePixel32(i & 15, i >> 4, c, FBP, FBW);
} }
@ -2329,7 +2329,7 @@ void GSRendererHW::OO_MajokkoALaMode2()
{ {
// palette readback // palette readback
const uint32 FBP = m_context->FRAME.Block(); const u32 FBP = m_context->FRAME.Block();
if (!PRIM->TME && FBP == 0x03f40) if (!PRIM->TME && FBP == 0x03f40)
{ {
@ -2349,7 +2349,7 @@ bool GSRendererHW::CU_MajokkoALaMode2()
{ {
// palette should stay 16 x 16 // palette should stay 16 x 16
const uint32 FBP = m_context->FRAME.Block(); const u32 FBP = m_context->FRAME.Block();
return FBP != 0x03f40; return FBP != 0x03f40;
} }
@ -2358,7 +2358,7 @@ bool GSRendererHW::CU_TalesOfAbyss()
{ {
// full image blur and brightening // full image blur and brightening
const uint32 FBP = m_context->FRAME.Block(); const u32 FBP = m_context->FRAME.Block();
return FBP != 0x036e0 && FBP != 0x03560 && FBP != 0x038e0; return FBP != 0x036e0 && FBP != 0x03560 && FBP != 0x038e0;
} }

10
pcsx2/GS/Renderers/HW/GSRendererHW.h
View File

@ -84,11 +84,11 @@ private:
}; };
template <class T> template <class T>
class FunctionMap : public GSFunctionMap<uint32, T> class FunctionMap : public GSFunctionMap<u32, T>
{ {
std::list<HackEntry<T>>& m_tbl; std::list<HackEntry<T>>& m_tbl;
T GetDefaultFunction(uint32 key) T GetDefaultFunction(u32 key)
{ {
CRC::Title title = (CRC::Title)(key & 0xffffff); CRC::Title title = (CRC::Title)(key & 0xffffff);
CRC::Region region = (CRC::Region)(key >> 24); CRC::Region region = (CRC::Region)(key >> 24);
@ -132,7 +132,7 @@ private:
#pragma endregion #pragma endregion
uint16 Interpolate_UV(float alpha, int t0, int t1); u16 Interpolate_UV(float alpha, int t0, int t1);
float alpha0(int L, int X0, int X1); float alpha0(int L, int X0, int X1);
float alpha1(int L, int X0, int X1); float alpha1(int L, int X0, int X1);
void SwSpriteRender(); void SwSpriteRender();
@ -168,13 +168,13 @@ public:
GSRendererHW(GSTextureCache* tc); GSRendererHW(GSTextureCache* tc);
virtual ~GSRendererHW(); virtual ~GSRendererHW();
void SetGameCRC(uint32 crc, int options); void SetGameCRC(u32 crc, int options);
bool CanUpscale(); bool CanUpscale();
int GetUpscaleMultiplier(); int GetUpscaleMultiplier();
GSVector2i GetCustomResolution(); GSVector2i GetCustomResolution();
void SetScaling(); void SetScaling();
void Lines2Sprites(); void Lines2Sprites();
void EmulateAtst(GSVector4& FogColor_AREF, uint8& atst, const bool pass_2); void EmulateAtst(GSVector4& FogColor_AREF, u8& atst, const bool pass_2);
void ConvertSpriteTextureShuffle(bool& write_ba, bool& read_ba); void ConvertSpriteTextureShuffle(bool& write_ba, bool& read_ba);
GSVector4 RealignTargetTextureCoordinate(const GSTextureCache::Source* tex); GSVector4 RealignTargetTextureCoordinate(const GSTextureCache::Source* tex);
GSVector4i ComputeBoundingBox(const GSVector2& rtscale, const GSVector2i& rtsize); GSVector4i ComputeBoundingBox(const GSVector2& rtscale, const GSVector2i& rtsize);

154
pcsx2/GS/Renderers/HW/GSTextureCache.cpp
View File

@ -54,7 +54,7 @@ GSTextureCache::GSTextureCache(GSRenderer* r)
// In theory 4MB is enough but 9MB is safer for overflow (8MB // In theory 4MB is enough but 9MB is safer for overflow (8MB
// isn't enough in custom resolution) // isn't enough in custom resolution)
// Test: onimusha 3 PAL 60Hz // Test: onimusha 3 PAL 60Hz
m_temp = (uint8*)_aligned_malloc(9 * 1024 * 1024, 32); m_temp = (u8*)_aligned_malloc(9 * 1024 * 1024, 32);
m_texture_inside_rt_cache.reserve(m_texture_inside_rt_cache_size); m_texture_inside_rt_cache.reserve(m_texture_inside_rt_cache_size);
} }
@ -119,8 +119,8 @@ GSTextureCache::Source* GSTextureCache::LookupDepthSource(const GIFRegTEX0& TEX0
Target* dst = NULL; Target* dst = NULL;
// Check only current frame, I guess it is only used as a postprocessing effect // Check only current frame, I guess it is only used as a postprocessing effect
uint32 bp = TEX0.TBP0; u32 bp = TEX0.TBP0;
uint32 psm = TEX0.PSM; u32 psm = TEX0.PSM;
for (auto t : m_dst[DepthStencil]) for (auto t : m_dst[DepthStencil])
{ {
@ -224,7 +224,7 @@ GSTextureCache::Source* GSTextureCache::LookupSource(const GIFRegTEX0& TEX0, con
if (psm_s.pal > 0) if (psm_s.pal > 0)
m_renderer->m_mem.m_clut.Read32(TEX0, TEXA); m_renderer->m_mem.m_clut.Read32(TEX0, TEXA);
const uint32* clut = m_renderer->m_mem.m_clut; const u32* clut = m_renderer->m_mem.m_clut;
Source* src = NULL; Source* src = NULL;
@ -234,7 +234,7 @@ GSTextureCache::Source* GSTextureCache::LookupSource(const GIFRegTEX0& TEX0, con
{ {
Source* s = *i; Source* s = *i;
if (((TEX0.u32[0] ^ s->m_TEX0.u32[0]) | ((TEX0.u32[1] ^ s->m_TEX0.u32[1]) & 3)) != 0) // TBP0 TBW PSM TW TH if (((TEX0.U32[0] ^ s->m_TEX0.U32[0]) | ((TEX0.U32[1] ^ s->m_TEX0.U32[1]) & 3)) != 0) // TBP0 TBW PSM TW TH
continue; continue;
// Target are converted (AEM & palette) on the fly by the GPU. They don't need extra check // Target are converted (AEM & palette) on the fly by the GPU. They don't need extra check
@ -248,7 +248,7 @@ GSTextureCache::Source* GSTextureCache::LookupSource(const GIFRegTEX0& TEX0, con
// We request a 24/16 bit RGBA texture. Alpha expansion was done by // We request a 24/16 bit RGBA texture. Alpha expansion was done by
// the CPU. We need to check that TEXA is identical // the CPU. We need to check that TEXA is identical
if (psm_s.pal == 0 && psm_s.fmt > 0 && s->m_TEXA.u64 != TEXA.u64) if (psm_s.pal == 0 && psm_s.fmt > 0 && s->m_TEXA.U64 != TEXA.U64)
continue; continue;
} }
@ -270,13 +270,13 @@ GSTextureCache::Source* GSTextureCache::LookupSource(const GIFRegTEX0& TEX0, con
if (src == NULL) if (src == NULL)
#endif #endif
{ {
uint32 bp = TEX0.TBP0; u32 bp = TEX0.TBP0;
uint32 psm = TEX0.PSM; u32 psm = TEX0.PSM;
uint32 bw = TEX0.TBW; u32 bw = TEX0.TBW;
int tw = 1 << TEX0.TW; int tw = 1 << TEX0.TW;
int th = 1 << TEX0.TH; int th = 1 << TEX0.TH;
uint32 bp_end = psm_s.info.bn(tw - 1, th - 1, bp, bw); // Valid only for color formats u32 bp_end = psm_s.info.bn(tw - 1, th - 1, bp, bw); // Valid only for color formats
// Arc the Lad finds the wrong surface here when looking for a depth stencil. // Arc the Lad finds the wrong surface here when looking for a depth stencil.
// Since we're currently not caching depth stencils (check ToDo in CreateSource) we should not look for it here. // Since we're currently not caching depth stencils (check ToDo in CreateSource) we should not look for it here.
@ -297,7 +297,7 @@ GSTextureCache::Source* GSTextureCache::LookupSource(const GIFRegTEX0& TEX0, con
// because of the previous draw call format // because of the previous draw call format
// //
// Solution: consider the RT as 32 bits if the alpha was used in the past // Solution: consider the RT as 32 bits if the alpha was used in the past
uint32 t_psm = (t->m_dirty_alpha) ? t->m_TEX0.PSM & ~0x1 : t->m_TEX0.PSM; u32 t_psm = (t->m_dirty_alpha) ? t->m_TEX0.PSM & ~0x1 : t->m_TEX0.PSM;
if (GSUtil::HasSharedBits(bp, psm, t->m_TEX0.TBP0, t_psm)) if (GSUtil::HasSharedBits(bp, psm, t->m_TEX0.TBP0, t_psm))
{ {
@ -376,7 +376,7 @@ GSTextureCache::Source* GSTextureCache::LookupSource(const GIFRegTEX0& TEX0, con
{ {
if (candidate_x_offset == 0 && candidate_y_offset == 0) if (candidate_x_offset == 0 && candidate_y_offset == 0)
continue; continue;
uint32 candidate_bp = psm_s.info.bn(candidate_x_offset, candidate_y_offset, t->m_TEX0.TBP0, bw); u32 candidate_bp = psm_s.info.bn(candidate_x_offset, candidate_y_offset, t->m_TEX0.TBP0, bw);
if (bp == candidate_bp && bp_end <= t->m_end_block) if (bp == candidate_bp && bp_end <= t->m_end_block)
{ {
// SWEEP HIT: <x,y> offset found // SWEEP HIT: <x,y> offset found
@ -442,8 +442,8 @@ GSTextureCache::Source* GSTextureCache::LookupSource(const GIFRegTEX0& TEX0, con
if (psm_s.bpp > 8) if (psm_s.bpp > 8)
{ {
GIFRegTEX0 depth_TEX0; GIFRegTEX0 depth_TEX0;
depth_TEX0.u32[0] = TEX0.u32[0] | (0x30u << 20u); depth_TEX0.U32[0] = TEX0.U32[0] | (0x30u << 20u);
depth_TEX0.u32[1] = TEX0.u32[1]; depth_TEX0.U32[1] = TEX0.U32[1];
return LookupDepthSource(depth_TEX0, TEXA, r); return LookupDepthSource(depth_TEX0, TEXA, r);
} }
else else
@ -521,10 +521,10 @@ bool GSTextureCache::ShallSearchTextureInsideRt()
return m_texture_inside_rt || (m_renderer->m_game.flags & CRC::Flags::TextureInsideRt); return m_texture_inside_rt || (m_renderer->m_game.flags & CRC::Flags::TextureInsideRt);
} }
GSTextureCache::Target* GSTextureCache::LookupTarget(const GIFRegTEX0& TEX0, int w, int h, int type, bool used, uint32 fbmask) GSTextureCache::Target* GSTextureCache::LookupTarget(const GIFRegTEX0& TEX0, int w, int h, int type, bool used, u32 fbmask)
{ {
const GSLocalMemory::psm_t& psm_s = GSLocalMemory::m_psm[TEX0.PSM]; const GSLocalMemory::psm_t& psm_s = GSLocalMemory::m_psm[TEX0.PSM];
uint32 bp = TEX0.TBP0; u32 bp = TEX0.TBP0;
Target* dst = NULL; Target* dst = NULL;
@ -655,7 +655,7 @@ GSTextureCache::Target* GSTextureCache::LookupTarget(const GIFRegTEX0& TEX0, int
GSTextureCache::Target* GSTextureCache::LookupTarget(const GIFRegTEX0& TEX0, int w, int h, int real_h) GSTextureCache::Target* GSTextureCache::LookupTarget(const GIFRegTEX0& TEX0, int w, int h, int real_h)
{ {
uint32 bp = TEX0.TBP0; u32 bp = TEX0.TBP0;
Target* dst = NULL; Target* dst = NULL;
@ -771,7 +771,7 @@ GSTextureCache::Target* GSTextureCache::LookupTarget(const GIFRegTEX0& TEX0, int
// Goal: Depth And Target at the same address is not possible. On GS it is // Goal: Depth And Target at the same address is not possible. On GS it is
// the same memory but not on the Dx/GL. Therefore a write to the Depth/Target // the same memory but not on the Dx/GL. Therefore a write to the Depth/Target
// must invalidate the Target/Depth respectively // must invalidate the Target/Depth respectively
void GSTextureCache::InvalidateVideoMemType(int type, uint32 bp) void GSTextureCache::InvalidateVideoMemType(int type, u32 bp)
{ {
if (!m_can_convert_depth) if (!m_can_convert_depth)
return; return;
@ -799,9 +799,9 @@ void GSTextureCache::InvalidateVideoMemType(int type, uint32 bp)
// Called each time you want to write to the GS memory // Called each time you want to write to the GS memory
void GSTextureCache::InvalidateVideoMem(const GSOffset& off, const GSVector4i& rect, bool target) void GSTextureCache::InvalidateVideoMem(const GSOffset& off, const GSVector4i& rect, bool target)
{ {
uint32 bp = off.bp(); u32 bp = off.bp();
uint32 bw = off.bw(); u32 bw = off.bw();
uint32 psm = off.psm(); u32 psm = off.psm();
if (!target) if (!target)
{ {
@ -820,7 +820,7 @@ void GSTextureCache::InvalidateVideoMem(const GSOffset& off, const GSVector4i& r
} }
} }
uint32 bbp = bp + bw * 0x10; u32 bbp = bp + bw * 0x10;
if (bw >= 16 && bbp < 16384) if (bw >= 16 && bbp < 16384)
{ {
// Detect half of the render target (fix snow engine game) // Detect half of the render target (fix snow engine game)
@ -844,7 +844,7 @@ void GSTextureCache::InvalidateVideoMem(const GSOffset& off, const GSVector4i& r
// we are screwed. // we are screwed.
if (m_renderer->m_game.title == CRC::HauntingGround) if (m_renderer->m_game.title == CRC::HauntingGround)
{ {
uint32 end_block = GSLocalMemory::m_psm[psm].info.bn(rect.z - 1, rect.w - 1, bp, bw); // Valid only for color formats u32 end_block = GSLocalMemory::m_psm[psm].info.bn(rect.z - 1, rect.w - 1, bp, bw); // Valid only for color formats
auto type = RenderTarget; auto type = RenderTarget;
for (auto t : m_dst[type]) for (auto t : m_dst[type])
@ -872,7 +872,7 @@ void GSTextureCache::InvalidateVideoMem(const GSOffset& off, const GSVector4i& r
GSVector4i r = rect.ralign<Align_Outside>((bp & 31) == 0 ? GSLocalMemory::m_psm[psm].pgs : GSLocalMemory::m_psm[psm].bs); GSVector4i r = rect.ralign<Align_Outside>((bp & 31) == 0 ? GSLocalMemory::m_psm[psm].pgs : GSLocalMemory::m_psm[psm].bs);
off.loopPages(rect, [&](uint32 page) off.loopPages(rect, [&](u32 page)
{ {
auto& list = m_src.m_map[page]; auto& list = m_src.m_map[page];
for (auto i = list.begin(); i != list.end();) for (auto i = list.begin(); i != list.end();)
@ -892,7 +892,7 @@ void GSTextureCache::InvalidateVideoMem(const GSOffset& off, const GSVector4i& r
} }
else else
{ {
uint32* RESTRICT valid = s->m_valid; u32* RESTRICT valid = s->m_valid;
// Invalidate data of input texture // Invalidate data of input texture
if (s->m_repeating) if (s->m_repeating)
@ -982,8 +982,8 @@ void GSTextureCache::InvalidateVideoMem(const GSOffset& off, const GSVector4i& r
{ {
if (bp < t->m_TEX0.TBP0) if (bp < t->m_TEX0.TBP0)
{ {
uint32 rowsize = bw * 8192; u32 rowsize = bw * 8192;
uint32 offset = (uint32)((t->m_TEX0.TBP0 - bp) * 256); u32 offset = (u32)((t->m_TEX0.TBP0 - bp) * 256);
if (rowsize > 0 && offset % rowsize == 0) if (rowsize > 0 && offset % rowsize == 0)
{ {
@ -1011,8 +1011,8 @@ void GSTextureCache::InvalidateVideoMem(const GSOffset& off, const GSVector4i& r
// (128 pixels) target // (128 pixels) target
if (bw > 2 && t->m_TEX0.TBW == bw && t->Inside(bp, bw, psm, rect) && GSUtil::HasCompatibleBits(psm, t->m_TEX0.PSM)) if (bw > 2 && t->m_TEX0.TBW == bw && t->Inside(bp, bw, psm, rect) && GSUtil::HasCompatibleBits(psm, t->m_TEX0.PSM))
{ {
uint32 rowsize = bw * 8192u; u32 rowsize = bw * 8192u;
uint32 offset = (uint32)((bp - t->m_TEX0.TBP0) * 256); u32 offset = (u32)((bp - t->m_TEX0.TBP0) * 256);
if (rowsize > 0 && offset % rowsize == 0) if (rowsize > 0 && offset % rowsize == 0)
{ {
@ -1038,9 +1038,9 @@ void GSTextureCache::InvalidateVideoMem(const GSOffset& off, const GSVector4i& r
// Called each time you want to read from the GS memory // Called each time you want to read from the GS memory
void GSTextureCache::InvalidateLocalMem(const GSOffset& off, const GSVector4i& r) void GSTextureCache::InvalidateLocalMem(const GSOffset& off, const GSVector4i& r)
{ {
uint32 bp = off.bp(); u32 bp = off.bp();
uint32 psm = off.psm(); u32 psm = off.psm();
//uint32 bw = off->bw; //u32 bw = off->bw;
// No depth handling please. // No depth handling please.
if (psm == PSM_PSMZ32 || psm == PSM_PSMZ24 || psm == PSM_PSMZ16 || psm == PSM_PSMZ16S) if (psm == PSM_PSMZ32 || psm == PSM_PSMZ24 || psm == PSM_PSMZ16 || psm == PSM_PSMZ16S)
@ -1155,8 +1155,8 @@ void GSTextureCache::InvalidateLocalMem(const GSOffset& off, const GSVector4i& r
// && ((t->m_TEX0.TBP0 == 0) || (t->m_TEX0.TBP0==3328) || (t->m_TEX0.TBP0==3584))) // && ((t->m_TEX0.TBP0 == 0) || (t->m_TEX0.TBP0==3328) || (t->m_TEX0.TBP0==3584)))
// { // {
// //printf("first : %d-%d child : %d-%d\n", psm, bp, t->m_TEX0.PSM, t->m_TEX0.TBP0); // //printf("first : %d-%d child : %d-%d\n", psm, bp, t->m_TEX0.PSM, t->m_TEX0.TBP0);
// uint32 rowsize = bw * 8192; // u32 rowsize = bw * 8192;
// uint32 offset = (uint32)((bp - t->m_TEX0.TBP0) * 256); // u32 offset = (u32)((bp - t->m_TEX0.TBP0) * 256);
// if (rowsize > 0 && offset % rowsize == 0) // if (rowsize > 0 && offset % rowsize == 0)
// { // {
@ -1656,10 +1656,10 @@ GSTextureCache::Target* GSTextureCache::CreateTarget(const GIFRegTEX0& TEX0, int
void GSTextureCache::PrintMemoryUsage() void GSTextureCache::PrintMemoryUsage()
{ {
#ifdef ENABLE_OGL_DEBUG #ifdef ENABLE_OGL_DEBUG
uint32 tex = 0; u32 tex = 0;
uint32 tex_rt = 0; u32 tex_rt = 0;
uint32 rt = 0; u32 rt = 0;
uint32 dss = 0; u32 dss = 0;
for (auto s : m_src.m_surfaces) for (auto s : m_src.m_surfaces)
{ {
if (s && !s->m_shared_texture) if (s && !s->m_shared_texture)
@ -1687,7 +1687,7 @@ void GSTextureCache::PrintMemoryUsage()
// GSTextureCache::Surface // GSTextureCache::Surface
GSTextureCache::Surface::Surface(GSRenderer* r, uint8* temp) GSTextureCache::Surface::Surface(GSRenderer* r, u8* temp)
: m_renderer(r) : m_renderer(r)
, m_texture(NULL) , m_texture(NULL)
, m_age(0) , m_age(0)
@ -1712,24 +1712,24 @@ void GSTextureCache::Surface::UpdateAge()
m_age = 0; m_age = 0;
} }
bool GSTextureCache::Surface::Inside(uint32 bp, uint32 bw, uint32 psm, const GSVector4i& rect) bool GSTextureCache::Surface::Inside(u32 bp, u32 bw, u32 psm, const GSVector4i& rect)
{ {
// Valid only for color formats. // Valid only for color formats.
uint32 const end_block = GSLocalMemory::m_psm[psm].info.bn(rect.z - 1, rect.w - 1, bp, bw); u32 const end_block = GSLocalMemory::m_psm[psm].info.bn(rect.z - 1, rect.w - 1, bp, bw);
return bp >= m_TEX0.TBP0 && end_block <= m_end_block; return bp >= m_TEX0.TBP0 && end_block <= m_end_block;
} }
bool GSTextureCache::Surface::Overlaps(uint32 bp, uint32 bw, uint32 psm, const GSVector4i& rect) bool GSTextureCache::Surface::Overlaps(u32 bp, u32 bw, u32 psm, const GSVector4i& rect)
{ {
// Valid only for color formats. // Valid only for color formats.
uint32 const end_block = GSLocalMemory::m_psm[psm].info.bn(rect.z - 1, rect.w - 1, bp, bw); u32 const end_block = GSLocalMemory::m_psm[psm].info.bn(rect.z - 1, rect.w - 1, bp, bw);
return (m_TEX0.TBP0 <= bp && bp <= m_end_block) return (m_TEX0.TBP0 <= bp && bp <= m_end_block)
|| (m_TEX0.TBP0 <= end_block && end_block <= m_end_block); || (m_TEX0.TBP0 <= end_block && end_block <= m_end_block);
} }
// GSTextureCache::Source // GSTextureCache::Source
GSTextureCache::Source::Source(GSRenderer* r, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA, uint8* temp, bool dummy_container) GSTextureCache::Source::Source(GSRenderer* r, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA, u8* temp, bool dummy_container)
: Surface(r, temp) : Surface(r, temp)
, m_palette_obj(nullptr) , m_palette_obj(nullptr)
, m_palette(nullptr) , m_palette(nullptr)
@ -1801,7 +1801,7 @@ void GSTextureCache::Source::Update(const GSVector4i& rect, int layer)
const GSOffset& off = m_renderer->m_context->offset.tex; const GSOffset& off = m_renderer->m_context->offset.tex;
GSOffset::BNHelper bn = off.bnMulti(r.left, r.top); GSOffset::BNHelper bn = off.bnMulti(r.left, r.top);
uint32 blocks = 0; u32 blocks = 0;
if (m_repeating) if (m_repeating)
{ {
@ -1810,14 +1810,14 @@ void GSTextureCache::Source::Update(const GSVector4i& rect, int layer)
for (int x = r.left; x < r.right; bn.nextBlockX(), x += bs.x) for (int x = r.left; x < r.right; bn.nextBlockX(), x += bs.x)
{ {
int i = (bn.blkY() << 7) + bn.blkX(); int i = (bn.blkY() << 7) + bn.blkX();
uint32 block = bn.valueNoWrap(); u32 block = bn.valueNoWrap();
if (block < MAX_BLOCKS || m_wrap_gs_mem) if (block < MAX_BLOCKS || m_wrap_gs_mem)
{ {
uint32 addr = i % MAX_BLOCKS; u32 addr = i % MAX_BLOCKS;
uint32 row = addr >> 5u; u32 row = addr >> 5u;
uint32 col = 1 << (addr & 31u); u32 col = 1 << (addr & 31u);
if ((m_valid[row] & col) == 0) if ((m_valid[row] & col) == 0)
{ {
@ -1837,14 +1837,14 @@ void GSTextureCache::Source::Update(const GSVector4i& rect, int layer)
{ {
for (int x = r.left; x < r.right; x += bs.x, bn.nextBlockX()) for (int x = r.left; x < r.right; x += bs.x, bn.nextBlockX())
{ {
uint32 block = bn.valueNoWrap(); u32 block = bn.valueNoWrap();
if (block < MAX_BLOCKS || m_wrap_gs_mem) if (block < MAX_BLOCKS || m_wrap_gs_mem)
{ {
block %= MAX_BLOCKS; block %= MAX_BLOCKS;
uint32 row = block >> 5u; u32 row = block >> 5u;
uint32 col = 1 << (block & 31u); u32 col = 1 << (block & 31u);
if ((m_valid[row] & col) == 0) if ((m_valid[row] & col) == 0)
{ {
@ -1921,7 +1921,7 @@ void GSTextureCache::Source::Write(const GSVector4i& r, int layer)
} }
} }
void GSTextureCache::Source::Flush(uint32 count, int layer) void GSTextureCache::Source::Flush(u32 count, int layer)
{ {
// This function as written will not work for paletted formats copied from framebuffers // This function as written will not work for paletted formats copied from framebuffers
// because they are 8 or 4 bit formats on the GS and the GS local memory module reads // because they are 8 or 4 bit formats on the GS and the GS local memory module reads
@ -1935,7 +1935,7 @@ void GSTextureCache::Source::Flush(uint32 count, int layer)
GSVector4i tr(0, 0, tw, th); GSVector4i tr(0, 0, tw, th);
int pitch = std::max(tw, psm.bs.x) * sizeof(uint32); int pitch = std::max(tw, psm.bs.x) * sizeof(u32);
GSLocalMemory& mem = m_renderer->m_mem; GSLocalMemory& mem = m_renderer->m_mem;
@ -1949,9 +1949,9 @@ void GSTextureCache::Source::Flush(uint32 count, int layer)
rtx = psm.rtxP; rtx = psm.rtxP;
} }
uint8* buff = m_temp; u8* buff = m_temp;
for (uint32 i = 0; i < count; i++) for (u32 i = 0; i < count; i++)
{ {
GSVector4i r = m_write.rect[i]; GSVector4i r = m_write.rect[i];
@ -1996,7 +1996,7 @@ bool GSTextureCache::Source::ClutMatch(PaletteKey palette_key)
// GSTextureCache::Target // GSTextureCache::Target
GSTextureCache::Target::Target(GSRenderer* r, const GIFRegTEX0& TEX0, uint8* temp, bool depth_supported) GSTextureCache::Target::Target(GSRenderer* r, const GIFRegTEX0& TEX0, u8* temp, bool depth_supported)
: Surface(r, temp) : Surface(r, temp)
, m_type(-1) , m_type(-1)
, m_used(false) , m_used(false)
@ -2135,7 +2135,7 @@ void GSTextureCache::SourceMap::Add(Source* s, const GIFRegTEX0& TEX0, const GSO
} }
// The source pointer will be stored/duplicated in all m_map[array of pages] // The source pointer will be stored/duplicated in all m_map[array of pages]
s->m_pages.loopPages([this, s](uint32 page) s->m_pages.loopPages([this, s](u32 page)
{ {
s->m_erase_it[page] = m_map[page].InsertFront(s); s->m_erase_it[page] = m_map[page].InsertFront(s);
}); });
@ -2169,7 +2169,7 @@ void GSTextureCache::SourceMap::RemoveAt(Source* s)
} }
else else
{ {
s->m_pages.loopPages([this, s](uint32 page) s->m_pages.loopPages([this, s](u32 page)
{ {
m_map[page].EraseIndex(s->m_erase_it[page]); m_map[page].EraseIndex(s->m_erase_it[page]);
}); });
@ -2178,7 +2178,7 @@ void GSTextureCache::SourceMap::RemoveAt(Source* s)
delete s; delete s;
} }
void GSTextureCache::AttachPaletteToSource(Source* s, uint16 pal, bool need_gs_texture) void GSTextureCache::AttachPaletteToSource(Source* s, u16 pal, bool need_gs_texture)
{ {
s->m_palette_obj = m_palette_map.LookupPalette(pal, need_gs_texture); s->m_palette_obj = m_palette_map.LookupPalette(pal, need_gs_texture);
s->m_palette = need_gs_texture ? s->m_palette_obj->GetPaletteGSTexture() : nullptr; s->m_palette = need_gs_texture ? s->m_palette_obj->GetPaletteGSTexture() : nullptr;
@ -2186,14 +2186,14 @@ void GSTextureCache::AttachPaletteToSource(Source* s, uint16 pal, bool need_gs_t
// GSTextureCache::Palette // GSTextureCache::Palette
GSTextureCache::Palette::Palette(const GSRenderer* renderer, uint16 pal, bool need_gs_texture) GSTextureCache::Palette::Palette(const GSRenderer* renderer, u16 pal, bool need_gs_texture)
: m_pal(pal) : m_pal(pal)
, m_tex_palette(nullptr) , m_tex_palette(nullptr)
, m_renderer(renderer) , m_renderer(renderer)
{ {
uint16 palette_size = pal * sizeof(uint32); u16 palette_size = pal * sizeof(u32);
m_clut = (uint32*)_aligned_malloc(palette_size, 64); m_clut = (u32*)_aligned_malloc(palette_size, 64);
memcpy(m_clut, (const uint32*)m_renderer->m_mem.m_clut, palette_size); memcpy(m_clut, (const u32*)m_renderer->m_mem.m_clut, palette_size);
if (need_gs_texture) if (need_gs_texture)
{ {
InitializeTexture(); InitializeTexture();
@ -2234,20 +2234,20 @@ void GSTextureCache::Palette::InitializeTexture()
// Hashes the content of the clut. // Hashes the content of the clut.
// The hashing function is implemented by taking two things into account: // The hashing function is implemented by taking two things into account:
// 1) The clut can be an array of 16 or 256 uint32 (depending on the pal parameter) and in order to speed up the computation of the hash // 1) The clut can be an array of 16 or 256 u32 (depending on the pal parameter) and in order to speed up the computation of the hash
// the array is hashed in blocks of 16 uint32, so for clut of size 16 uint32 the hashing is computed in one pass and for clut of 256 uint32 // the array is hashed in blocks of 16 u32, so for clut of size 16 u32 the hashing is computed in one pass and for clut of 256 u32
// it is computed in 16 passes, // it is computed in 16 passes,
// 2) The clut can contain many 0s, so as a way to increase the spread of hashing values for small changes in the input clut the hashing function // 2) The clut can contain many 0s, so as a way to increase the spread of hashing values for small changes in the input clut the hashing function
// is using addition in combination with logical XOR operator; The addition constants are large prime numbers, which may help in achieving what intended. // is using addition in combination with logical XOR operator; The addition constants are large prime numbers, which may help in achieving what intended.
std::size_t GSTextureCache::PaletteKeyHash::operator()(const PaletteKey& key) const std::size_t GSTextureCache::PaletteKeyHash::operator()(const PaletteKey& key) const
{ {
uint16 pal = key.pal; u16 pal = key.pal;
const uint32* clut = key.clut; const u32* clut = key.clut;
ASSERT((pal & 15) == 0); ASSERT((pal & 15) == 0);
size_t clut_hash = 3831179159; size_t clut_hash = 3831179159;
for (uint16 i = 0; i < pal; i += 16) for (u16 i = 0; i < pal; i += 16)
{ {
clut_hash = (clut_hash + 1488000301) ^ (clut[i ] + 33644011); clut_hash = (clut_hash + 1488000301) ^ (clut[i ] + 33644011);
clut_hash = (clut_hash + 3831179159) ^ (clut[i + 1] + 47627467); clut_hash = (clut_hash + 3831179159) ^ (clut[i + 1] + 47627467);
@ -2295,7 +2295,7 @@ GSTextureCache::PaletteMap::PaletteMap(const GSRenderer* renderer)
} }
} }
std::shared_ptr<GSTextureCache::Palette> GSTextureCache::PaletteMap::LookupPalette(uint16 pal, bool need_gs_texture) std::shared_ptr<GSTextureCache::Palette> GSTextureCache::PaletteMap::LookupPalette(u16 pal, bool need_gs_texture)
{ {
ASSERT(pal == 16 || pal == 256); ASSERT(pal == 16 || pal == 256);
@ -2304,7 +2304,7 @@ std::shared_ptr<GSTextureCache::Palette> GSTextureCache::PaletteMap::LookupPalet
// pal == 256 : index 1 // pal == 256 : index 1
auto& map = m_maps[pal == 16 ? 0 : 1]; auto& map = m_maps[pal == 16 ? 0 : 1];
const uint32* clut = (const uint32*)m_renderer->m_mem.m_clut; const u32* clut = (const u32*)m_renderer->m_mem.m_clut;
// Create PaletteKey for searching into map (clut is actually not copied, so do not store this key into the map) // Create PaletteKey for searching into map (clut is actually not copied, so do not store this key into the map)
PaletteKey palette_key = {clut, pal}; PaletteKey palette_key = {clut, pal};
@ -2327,9 +2327,9 @@ std::shared_ptr<GSTextureCache::Palette> GSTextureCache::PaletteMap::LookupPalet
if (map.size() > MAX_SIZE) if (map.size() > MAX_SIZE)
{ {
// If the map is too big, try to clean it by disposing and removing unused palettes, before adding the new one // If the map is too big, try to clean it by disposing and removing unused palettes, before adding the new one
GL_INS("WARNING, %u-bit PaletteMap (Size %u): Max size %u exceeded, clearing unused palettes.", pal * sizeof(uint32), map.size(), MAX_SIZE); GL_INS("WARNING, %u-bit PaletteMap (Size %u): Max size %u exceeded, clearing unused palettes.", pal * sizeof(u32), map.size(), MAX_SIZE);
uint32 current_size = map.size(); u32 current_size = map.size();
for (auto it = map.begin(); it != map.end();) for (auto it = map.begin(); it != map.end();)
{ {
@ -2347,16 +2347,16 @@ std::shared_ptr<GSTextureCache::Palette> GSTextureCache::PaletteMap::LookupPalet
} }
} }
uint32 cleared_palette_count = current_size - (uint32)map.size(); u32 cleared_palette_count = current_size - (u32)map.size();
if (cleared_palette_count == 0) if (cleared_palette_count == 0)
{ {
GL_INS("ERROR, %u-bit PaletteMap (Size %u): Max size %u exceeded, could not clear any palette, negative performance impact.", pal * sizeof(uint32), map.size(), MAX_SIZE); GL_INS("ERROR, %u-bit PaletteMap (Size %u): Max size %u exceeded, could not clear any palette, negative performance impact.", pal * sizeof(u32), map.size(), MAX_SIZE);
} }
else else
{ {
map.reserve(MAX_SIZE); // Ensure map capacity is not modified by the clearing map.reserve(MAX_SIZE); // Ensure map capacity is not modified by the clearing
GL_INS("INFO, %u-bit PaletteMap (Size %u): Cleared %u palettes.", pal * sizeof(uint32), map.size(), cleared_palette_count); GL_INS("INFO, %u-bit PaletteMap (Size %u): Cleared %u palettes.", pal * sizeof(u32), map.size(), cleared_palette_count);
} }
} }
@ -2364,7 +2364,7 @@ std::shared_ptr<GSTextureCache::Palette> GSTextureCache::PaletteMap::LookupPalet
map.emplace(palette->GetPaletteKey(), palette); map.emplace(palette->GetPaletteKey(), palette);
GL_CACHE("TC, %u-bit PaletteMap (Size %u): Added new palette.", pal * sizeof(uint32), map.size()); GL_CACHE("TC, %u-bit PaletteMap (Size %u): Added new palette.", pal * sizeof(u32), map.size());
return palette; return palette;
} }

60
pcsx2/GS/Renderers/HW/GSTextureCache.h
View File

@ -38,36 +38,36 @@ public:
GIFRegTEX0 m_TEX0; GIFRegTEX0 m_TEX0;
GIFRegTEXA m_TEXA; GIFRegTEXA m_TEXA;
int m_age; int m_age;
uint8* m_temp; u8* m_temp;
bool m_32_bits_fmt; // Allow to detect the casting of 32 bits as 16 bits texture bool m_32_bits_fmt; // Allow to detect the casting of 32 bits as 16 bits texture
bool m_shared_texture; bool m_shared_texture;
uint32 m_end_block; // Hint of the surface area. u32 m_end_block; // Hint of the surface area.
public: public:
Surface(GSRenderer* r, uint8* temp); Surface(GSRenderer* r, u8* temp);
virtual ~Surface(); virtual ~Surface();
void UpdateAge(); void UpdateAge();
bool Inside(uint32 bp, uint32 bw, uint32 psm, const GSVector4i& rect); bool Inside(u32 bp, u32 bw, u32 psm, const GSVector4i& rect);
bool Overlaps(uint32 bp, uint32 bw, uint32 psm, const GSVector4i& rect); bool Overlaps(u32 bp, u32 bw, u32 psm, const GSVector4i& rect);
}; };
struct PaletteKey struct PaletteKey
{ {
const uint32* clut; const u32* clut;
uint16 pal; u16 pal;
}; };
class Palette class Palette
{ {
private: private:
uint32* m_clut; u32* m_clut;
uint16 m_pal; u16 m_pal;
GSTexture* m_tex_palette; GSTexture* m_tex_palette;
const GSRenderer* m_renderer; const GSRenderer* m_renderer;
public: public:
Palette(const GSRenderer* renderer, uint16 pal, bool need_gs_texture); Palette(const GSRenderer* renderer, u16 pal, bool need_gs_texture);
~Palette(); ~Palette();
// Disable copy constructor and copy operator // Disable copy constructor and copy operator
@ -102,16 +102,16 @@ public:
struct struct
{ {
GSVector4i* rect; GSVector4i* rect;
uint32 count; u32 count;
} m_write; } m_write;
void Write(const GSVector4i& r, int layer); void Write(const GSVector4i& r, int layer);
void Flush(uint32 count, int layer); void Flush(u32 count, int layer);
public: public:
std::shared_ptr<Palette> m_palette_obj; std::shared_ptr<Palette> m_palette_obj;
GSTexture* m_palette; GSTexture* m_palette;
uint32 m_valid[MAX_PAGES]; // each uint32 bits map to the 32 blocks of that page u32 m_valid[MAX_PAGES]; // each u32 bits map to the 32 blocks of that page
GSVector4i m_valid_rect; GSVector4i m_valid_rect;
bool m_target; bool m_target;
bool m_complete; bool m_complete;
@ -124,11 +124,11 @@ public:
GIFRegTEX0 m_from_target_TEX0; // TEX0 of the target texture, if any, else equal to texture TEX0 GIFRegTEX0 m_from_target_TEX0; // TEX0 of the target texture, if any, else equal to texture TEX0
GIFRegTEX0 m_layer_TEX0[7]; // Detect already loaded value GIFRegTEX0 m_layer_TEX0[7]; // Detect already loaded value
// Keep a GSTextureCache::SourceMap::m_map iterator to allow fast erase // Keep a GSTextureCache::SourceMap::m_map iterator to allow fast erase
std::array<uint16, MAX_PAGES> m_erase_it; std::array<u16, MAX_PAGES> m_erase_it;
GSOffset::PageLooper m_pages; GSOffset::PageLooper m_pages;
public: public:
Source(GSRenderer* r, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA, uint8* temp, bool dummy_container = false); Source(GSRenderer* r, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA, u8* temp, bool dummy_container = false);
virtual ~Source(); virtual ~Source();
void Update(const GSVector4i& rect, int layer = 0); void Update(const GSVector4i& rect, int layer = 0);
@ -148,7 +148,7 @@ public:
bool m_dirty_alpha; bool m_dirty_alpha;
public: public:
Target(GSRenderer* r, const GIFRegTEX0& TEX0, uint8* temp, bool depth_supported); Target(GSRenderer* r, const GIFRegTEX0& TEX0, u8* temp, bool depth_supported);
void UpdateValidity(const GSVector4i& rect); void UpdateValidity(const GSVector4i& rect);
@ -158,7 +158,7 @@ public:
class PaletteMap class PaletteMap
{ {
private: private:
static const uint16 MAX_SIZE = 65535; // Max size of each map. static const u16 MAX_SIZE = 65535; // Max size of each map.
const GSRenderer* m_renderer; const GSRenderer* m_renderer;
// Array of 2 maps, the first for 64B palettes and the second for 1024B palettes. // Array of 2 maps, the first for 64B palettes and the second for 1024B palettes.
@ -170,7 +170,7 @@ public:
PaletteMap(const GSRenderer* renderer); PaletteMap(const GSRenderer* renderer);
// Retrieves a shared pointer to a valid Palette from m_maps or creates a new one adding it to the data structure // Retrieves a shared pointer to a valid Palette from m_maps or creates a new one adding it to the data structure
std::shared_ptr<Palette> LookupPalette(uint16 pal, bool need_gs_texture); std::shared_ptr<Palette> LookupPalette(u16 pal, bool need_gs_texture);
void Clear(); // Clears m_maps, thus deletes Palette objects void Clear(); // Clears m_maps, thus deletes Palette objects
}; };
@ -180,7 +180,7 @@ public:
public: public:
std::unordered_set<Source*> m_surfaces; std::unordered_set<Source*> m_surfaces;
std::array<FastList<Source*>, MAX_PAGES> m_map; std::array<FastList<Source*>, MAX_PAGES> m_map;
uint32 m_pages[16]; // bitmap of all pages u32 m_pages[16]; // bitmap of all pages
bool m_used; bool m_used;
SourceMap() SourceMap()
@ -197,12 +197,12 @@ public:
struct TexInsideRtCacheEntry struct TexInsideRtCacheEntry
{ {
uint32 psm; u32 psm;
uint32 bp; u32 bp;
uint32 bp_end; u32 bp_end;
uint32 bw; u32 bw;
uint32 t_tex0_tbp0; u32 t_tex0_tbp0;
uint32 m_end_block; u32 m_end_block;
bool has_valid_offset; bool has_valid_offset;
int x_offset; int x_offset;
int y_offset; int y_offset;
@ -215,14 +215,14 @@ protected:
FastList<Target*> m_dst[2]; FastList<Target*> m_dst[2];
bool m_paltex; bool m_paltex;
bool m_preload_frame; bool m_preload_frame;
uint8* m_temp; u8* m_temp;
bool m_can_convert_depth; bool m_can_convert_depth;
bool m_cpu_fb_conversion; bool m_cpu_fb_conversion;
CRCHackLevel m_crc_hack_level; CRCHackLevel m_crc_hack_level;
static bool m_disable_partial_invalidation; static bool m_disable_partial_invalidation;
bool m_texture_inside_rt; bool m_texture_inside_rt;
static bool m_wrap_gs_mem; static bool m_wrap_gs_mem;
uint8 m_texture_inside_rt_cache_size = 255; u8 m_texture_inside_rt_cache_size = 255;
std::vector<TexInsideRtCacheEntry> m_texture_inside_rt_cache; std::vector<TexInsideRtCacheEntry> m_texture_inside_rt_cache;
virtual Source* CreateSource(const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA, Target* t = NULL, bool half_right = false, int x_offset = 0, int y_offset = 0); virtual Source* CreateSource(const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA, Target* t = NULL, bool half_right = false, int x_offset = 0, int y_offset = 0);
@ -244,10 +244,10 @@ public:
Source* LookupSource(const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA, const GSVector4i& r); Source* LookupSource(const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA, const GSVector4i& r);
Source* LookupDepthSource(const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA, const GSVector4i& r, bool palette = false); Source* LookupDepthSource(const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA, const GSVector4i& r, bool palette = false);
Target* LookupTarget(const GIFRegTEX0& TEX0, int w, int h, int type, bool used, uint32 fbmask = 0); Target* LookupTarget(const GIFRegTEX0& TEX0, int w, int h, int type, bool used, u32 fbmask = 0);
Target* LookupTarget(const GIFRegTEX0& TEX0, int w, int h, int real_h); Target* LookupTarget(const GIFRegTEX0& TEX0, int w, int h, int real_h);
void InvalidateVideoMemType(int type, uint32 bp); void InvalidateVideoMemType(int type, u32 bp);
void InvalidateVideoMemSubTarget(GSTextureCache::Target* rt); void InvalidateVideoMemSubTarget(GSTextureCache::Target* rt);
void InvalidateVideoMem(const GSOffset& off, const GSVector4i& r, bool target = true); void InvalidateVideoMem(const GSOffset& off, const GSVector4i& r, bool target = true);
void InvalidateLocalMem(const GSOffset& off, const GSVector4i& r); void InvalidateLocalMem(const GSOffset& off, const GSVector4i& r);
@ -265,5 +265,5 @@ public:
void PrintMemoryUsage(); void PrintMemoryUsage();
void AttachPaletteToSource(Source* s, uint16 pal, bool need_gs_texture); void AttachPaletteToSource(Source* s, u16 pal, bool need_gs_texture);
}; };

4
pcsx2/GS/Renderers/HW/GSVertexHW.h
View File

@ -25,8 +25,8 @@ struct alignas(32) GSVertexHW9
GSVector4 t; GSVector4 t;
GSVector4 p; GSVector4 p;
// t.z = union {struct {uint8 r, g, b, a;}; uint32 c0;}; // t.z = union {struct {u8 r, g, b, a;}; u32 c0;};
// t.w = union {struct {uint8 ta0, ta1, res, f;}; uint32 c1;} // t.w = union {struct {u8 ta0, ta1, res, f;}; u32 c1;}
GSVertexHW9& operator=(GSVertexHW9& v) GSVertexHW9& operator=(GSVertexHW9& v)
{ {

2
pcsx2/GS/Renderers/Null/GSDeviceNull.h
View File

@ -25,7 +25,7 @@ private:
void DoMerge(GSTexture* sTex[3], GSVector4* sRect, GSTexture* dTex, GSVector4* dRect, const GSRegPMODE& PMODE, const GSRegEXTBUF& EXTBUF, const GSVector4& c) {} void DoMerge(GSTexture* sTex[3], GSVector4* sRect, GSTexture* dTex, GSVector4* dRect, const GSRegPMODE& PMODE, const GSRegEXTBUF& EXTBUF, const GSVector4& c) {}
void DoInterlace(GSTexture* sTex, GSTexture* dTex, int shader, bool linear, float yoffset = 0) {} void DoInterlace(GSTexture* sTex, GSTexture* dTex, int shader, bool linear, float yoffset = 0) {}
uint16 ConvertBlendEnum(uint16 generic) { return 0xFFFF; } u16 ConvertBlendEnum(u16 generic) { return 0xFFFF; }
public: public:
GSDeviceNull() {} GSDeviceNull() {}

12
pcsx2/GS/Renderers/OpenGL/GLState.cpp
View File

@ -23,11 +23,11 @@ namespace GLState
GSVector4i scissor; GSVector4i scissor;
bool blend; bool blend;
uint16 eq_RGB; u16 eq_RGB;
uint16 f_sRGB; u16 f_sRGB;
uint16 f_dRGB; u16 f_dRGB;
uint8 bf; u8 bf;
uint32 wrgba; u32 wrgba;
bool depth; bool depth;
GLenum depth_func; GLenum depth_func;
@ -50,7 +50,7 @@ namespace GLState
GLuint program; GLuint program;
GLuint pipeline; GLuint pipeline;
int64 available_vram; s64 available_vram;
void Clear() void Clear()
{ {

12
pcsx2/GS/Renderers/OpenGL/GLState.h
View File

@ -25,11 +25,11 @@ namespace GLState
extern GSVector4i scissor; extern GSVector4i scissor;
extern bool blend; extern bool blend;
extern uint16 eq_RGB; extern u16 eq_RGB;
extern uint16 f_sRGB; extern u16 f_sRGB;
extern uint16 f_dRGB; extern u16 f_dRGB;
extern uint8 bf; extern u8 bf;
extern uint32 wrgba; extern u32 wrgba;
extern bool depth; extern bool depth;
extern GLenum depth_func; extern GLenum depth_func;
@ -52,7 +52,7 @@ namespace GLState
extern GLuint program; extern GLuint program;
extern GLuint pipeline; extern GLuint pipeline;
extern int64 available_vram; extern s64 available_vram;
extern void Clear(); extern void Clear();
} // namespace GLState } // namespace GLState

48
pcsx2/GS/Renderers/OpenGL/GSDeviceOGL.cpp
View File

@ -30,17 +30,17 @@
// TODO port those value into PerfMon API // TODO port those value into PerfMon API
#ifdef ENABLE_OGL_DEBUG_MEM_BW #ifdef ENABLE_OGL_DEBUG_MEM_BW
uint64 g_real_texture_upload_byte = 0; u64 g_real_texture_upload_byte = 0;
uint64 g_vertex_upload_byte = 0; u64 g_vertex_upload_byte = 0;
uint64 g_uniform_upload_byte = 0; u64 g_uniform_upload_byte = 0;
#endif #endif
static constexpr uint32 g_merge_cb_index = 10; static constexpr u32 g_merge_cb_index = 10;
static constexpr uint32 g_interlace_cb_index = 11; static constexpr u32 g_interlace_cb_index = 11;
static constexpr uint32 g_fx_cb_index = 14; static constexpr u32 g_fx_cb_index = 14;
static constexpr uint32 g_convert_index = 15; static constexpr u32 g_convert_index = 15;
static constexpr uint32 g_vs_cb_index = 20; static constexpr u32 g_vs_cb_index = 20;
static constexpr uint32 g_ps_cb_index = 21; static constexpr u32 g_ps_cb_index = 21;
static constexpr u32 VERTEX_BUFFER_SIZE = 32 * 1024 * 1024; static constexpr u32 VERTEX_BUFFER_SIZE = 32 * 1024 * 1024;
static constexpr u32 INDEX_BUFFER_SIZE = 16 * 1024 * 1024; static constexpr u32 INDEX_BUFFER_SIZE = 16 * 1024 * 1024;
@ -179,10 +179,10 @@ void GSDeviceOGL::GenerateProfilerData()
const int first_query = replay > 1 ? m_profiler.last_query / replay : 0; const int first_query = replay > 1 ? m_profiler.last_query / replay : 0;
glGetQueryObjectui64v(m_profiler.timer_query[first_query], GL_QUERY_RESULT, &time_start); glGetQueryObjectui64v(m_profiler.timer_query[first_query], GL_QUERY_RESULT, &time_start);
for (uint32 q = first_query + 1; q < m_profiler.last_query; q++) for (u32 q = first_query + 1; q < m_profiler.last_query; q++)
{ {
glGetQueryObjectui64v(m_profiler.timer_query[q], GL_QUERY_RESULT, &time_end); glGetQueryObjectui64v(m_profiler.timer_query[q], GL_QUERY_RESULT, &time_end);
uint64 t = time_end - time_start; u64 t = time_end - time_start;
times.push_back((double)t * ms); times.push_back((double)t * ms);
time_start = time_end; time_start = time_end;
@ -207,7 +207,7 @@ void GSDeviceOGL::GenerateProfilerData()
sd += pow(t - mean, 2); sd += pow(t - mean, 2);
sd = sqrt(sd / frames); sd = sqrt(sd / frames);
uint32 time_repartition[16] = {0}; u32 time_repartition[16] = {0};
for (auto t : times) for (auto t : times)
{ {
size_t slot = std::min<size_t>(t / 2.0, std::size(time_repartition) - 1); size_t slot = std::min<size_t>(t / 2.0, std::size(time_repartition) - 1);
@ -221,7 +221,7 @@ void GSDeviceOGL::GenerateProfilerData()
fprintf(stderr, "SD %4.2f ms\n", sd); fprintf(stderr, "SD %4.2f ms\n", sd);
fprintf(stderr, "\n"); fprintf(stderr, "\n");
fprintf(stderr, "Frame Repartition\n"); fprintf(stderr, "Frame Repartition\n");
for (uint32 i = 0; i < std::size(time_repartition); i++) for (u32 i = 0; i < std::size(time_repartition); i++)
{ {
fprintf(stderr, "%3u ms => %3u ms\t%4u\n", 2 * i, 2 * (i + 1), time_repartition[i]); fprintf(stderr, "%3u ms => %3u ms\t%4u\n", 2 * i, 2 * (i + 1), time_repartition[i]);
} }
@ -420,7 +420,7 @@ bool GSDeviceOGL::Create(const WindowInfo& wi)
{ {
GL_PUSH("GSDeviceOGL::Sampler"); GL_PUSH("GSDeviceOGL::Sampler");
for (uint32 key = 0; key < std::size(m_ps_ss); key++) for (u32 key = 0; key < std::size(m_ps_ss); key++)
{ {
m_ps_ss[key] = CreateSampler(PSSamplerSelector(key)); m_ps_ss[key] = CreateSampler(PSSamplerSelector(key));
} }
@ -606,7 +606,7 @@ bool GSDeviceOGL::Create(const WindowInfo& wi)
// When VRAM is below 2GB, we add a factor 2 because RAM can be used. Potentially // When VRAM is below 2GB, we add a factor 2 because RAM can be used. Potentially
// low VRAM gpu can go higher but perf will be bad anyway. // low VRAM gpu can go higher but perf will be bad anyway.
if (vram[0] > 0 && vram[0] < 1800000) if (vram[0] > 0 && vram[0] < 1800000)
GLState::available_vram = (int64)(vram[0]) * 1024ul * 2ul; GLState::available_vram = (s64)(vram[0]) * 1024ul * 2ul;
fprintf(stdout, "Available VRAM/RAM:%lldMB for textures\n", GLState::available_vram >> 20u); fprintf(stdout, "Available VRAM/RAM:%lldMB for textures\n", GLState::available_vram >> 20u);
@ -654,14 +654,14 @@ void GSDeviceOGL::CreateTextureFX()
m_gs[2] = CompileGS(GSSelector(2)); m_gs[2] = CompileGS(GSSelector(2));
m_gs[4] = CompileGS(GSSelector(4)); m_gs[4] = CompileGS(GSSelector(4));
for (uint32 key = 0; key < std::size(m_vs); key++) for (u32 key = 0; key < std::size(m_vs); key++)
m_vs[key] = CompileVS(VSSelector(key)); m_vs[key] = CompileVS(VSSelector(key));
// Enable all bits for stencil operations. Technically 1 bit is // Enable all bits for stencil operations. Technically 1 bit is
// enough but buffer is polluted with noise. Clear will be limited // enough but buffer is polluted with noise. Clear will be limited
// to the mask. // to the mask.
glStencilMask(0xFF); glStencilMask(0xFF);
for (uint32 key = 0; key < std::size(m_om_dss); key++) for (u32 key = 0; key < std::size(m_om_dss); key++)
{ {
m_om_dss[key] = CreateDepthStencil(OMDepthStencilSelector(key)); m_om_dss[key] = CreateDepthStencil(OMDepthStencilSelector(key));
} }
@ -747,7 +747,7 @@ void GSDeviceOGL::ClearRenderTarget(GSTexture* t, const GSVector4& c)
// TODO: check size of scissor before toggling it // TODO: check size of scissor before toggling it
glDisable(GL_SCISSOR_TEST); glDisable(GL_SCISSOR_TEST);
const uint32 old_color_mask = GLState::wrgba; const u32 old_color_mask = GLState::wrgba;
OMSetColorMaskState(); OMSetColorMaskState();
if (T->IsBackbuffer()) if (T->IsBackbuffer())
@ -773,7 +773,7 @@ void GSDeviceOGL::ClearRenderTarget(GSTexture* t, const GSVector4& c)
T->WasCleaned(); T->WasCleaned();
} }
void GSDeviceOGL::ClearRenderTarget(GSTexture* t, uint32 c) void GSDeviceOGL::ClearRenderTarget(GSTexture* t, u32 c)
{ {
if (!t) if (!t)
return; return;
@ -828,7 +828,7 @@ void GSDeviceOGL::ClearDepth(GSTexture* t)
} }
} }
void GSDeviceOGL::ClearStencil(GSTexture* t, uint8 c) void GSDeviceOGL::ClearStencil(GSTexture* t, u8 c)
{ {
if (!t) if (!t)
return; return;
@ -1404,7 +1404,7 @@ void GSDeviceOGL::StretchRect(GSTexture* sTex, const GSVector4& sRect, GSTexture
else else
OMSetRenderTargets(dTex, NULL); OMSetRenderTargets(dTex, NULL);
OMSetBlendState((uint8)bs); OMSetBlendState((u8)bs);
OMSetColorMaskState(cms); OMSetColorMaskState(cms);
// ************************************ // ************************************
@ -1475,7 +1475,7 @@ void GSDeviceOGL::RenderOsd(GSTexture* dt)
m_shader->BindPipeline(m_convert.ps[ShaderConvert_OSD]); m_shader->BindPipeline(m_convert.ps[ShaderConvert_OSD]);
OMSetDepthStencilState(m_convert.dss); OMSetDepthStencilState(m_convert.dss);
OMSetBlendState((uint8)GSDeviceOGL::m_MERGE_BLEND); OMSetBlendState((u8)GSDeviceOGL::m_MERGE_BLEND);
OMSetRenderTargets(dt, NULL); OMSetRenderTargets(dt, NULL);
if (m_osd.m_texture_dirty) if (m_osd.m_texture_dirty)
@ -1839,7 +1839,7 @@ void GSDeviceOGL::OMSetColorMaskState(OMColorMaskSelector sel)
} }
} }
void GSDeviceOGL::OMSetBlendState(uint8 blend_index, uint8 blend_factor, bool is_blend_constant, bool accumulation_blend) void GSDeviceOGL::OMSetBlendState(u8 blend_index, u8 blend_factor, bool is_blend_constant, bool accumulation_blend)
{ {
if (blend_index) if (blend_index)
{ {
@ -2137,7 +2137,7 @@ void GSDeviceOGL::DebugOutputToFile(GLenum gl_source, GLenum gl_type, GLuint id,
#endif #endif
} }
uint16 GSDeviceOGL::ConvertBlendEnum(uint16 generic) u16 GSDeviceOGL::ConvertBlendEnum(u16 generic)
{ {
switch (generic) switch (generic)
{ {

176
pcsx2/GS/Renderers/OpenGL/GSDeviceOGL.h
View File

@ -25,8 +25,8 @@
#include "GLState.h" #include "GLState.h"
#ifdef ENABLE_OGL_DEBUG_MEM_BW #ifdef ENABLE_OGL_DEBUG_MEM_BW
extern uint64 g_real_texture_upload_byte; extern u64 g_real_texture_upload_byte;
extern uint64 g_vertex_upload_byte; extern u64 g_vertex_upload_byte;
#endif #endif
class GSDepthStencilOGL class GSDepthStencilOGL
@ -164,20 +164,20 @@ public:
{ {
struct struct
{ {
uint32 int_fst : 1; u32 int_fst : 1;
uint32 _free : 31; u32 _free : 31;
}; };
uint32 key; u32 key;
}; };
operator uint32() const { return key; } operator u32() const { return key; }
VSSelector() VSSelector()
: key(0) : key(0)
{ {
} }
VSSelector(uint32 k) VSSelector(u32 k)
: key(k) : key(k)
{ {
} }
@ -189,23 +189,23 @@ public:
{ {
struct struct
{ {
uint32 sprite : 1; u32 sprite : 1;
uint32 point : 1; u32 point : 1;
uint32 line : 1; u32 line : 1;
uint32 _free : 29; u32 _free : 29;
}; };
uint32 key; u32 key;
}; };
operator uint32() const { return key; } operator u32() const { return key; }
GSSelector() GSSelector()
: key(0) : key(0)
{ {
} }
GSSelector(uint32 k) GSSelector(u32 k)
: key(k) : key(k)
{ {
} }
@ -287,72 +287,72 @@ public:
{ {
// *** Word 1 // *** Word 1
// Format // Format
uint32 tex_fmt : 4; u32 tex_fmt : 4;
uint32 dfmt : 2; u32 dfmt : 2;
uint32 depth_fmt : 2; u32 depth_fmt : 2;
// Alpha extension/Correction // Alpha extension/Correction
uint32 aem : 1; u32 aem : 1;
uint32 fba : 1; u32 fba : 1;
// Fog // Fog
uint32 fog : 1; u32 fog : 1;
// Flat/goround shading // Flat/goround shading
uint32 iip : 1; u32 iip : 1;
// Pixel test // Pixel test
uint32 date : 3; u32 date : 3;
uint32 atst : 3; u32 atst : 3;
// Color sampling // Color sampling
uint32 fst : 1; // Investigate to do it on the VS u32 fst : 1; // Investigate to do it on the VS
uint32 tfx : 3; u32 tfx : 3;
uint32 tcc : 1; u32 tcc : 1;
uint32 wms : 2; u32 wms : 2;
uint32 wmt : 2; u32 wmt : 2;
uint32 ltf : 1; u32 ltf : 1;
// Shuffle and fbmask effect // Shuffle and fbmask effect
uint32 shuffle : 1; u32 shuffle : 1;
uint32 read_ba : 1; u32 read_ba : 1;
uint32 write_rg : 1; u32 write_rg : 1;
uint32 fbmask : 1; u32 fbmask : 1;
//uint32 _free1:0; //u32 _free1:0;
// *** Word 2 // *** Word 2
// Blend and Colclip // Blend and Colclip
uint32 blend_a : 2; u32 blend_a : 2;
uint32 blend_b : 2; u32 blend_b : 2;
uint32 blend_c : 2; u32 blend_c : 2;
uint32 blend_d : 2; u32 blend_d : 2;
uint32 clr1 : 1; // useful? u32 clr1 : 1; // useful?
uint32 hdr : 1; u32 hdr : 1;
uint32 colclip : 1; u32 colclip : 1;
uint32 pabe : 1; u32 pabe : 1;
// Others ways to fetch the texture // Others ways to fetch the texture
uint32 channel : 3; u32 channel : 3;
// Dithering // Dithering
uint32 dither : 2; u32 dither : 2;
// Depth clamp // Depth clamp
uint32 zclamp : 1; u32 zclamp : 1;
// Hack // Hack
uint32 tcoffsethack : 1; u32 tcoffsethack : 1;
uint32 urban_chaos_hle : 1; u32 urban_chaos_hle : 1;
uint32 tales_of_abyss_hle : 1; u32 tales_of_abyss_hle : 1;
uint32 tex_is_fb : 1; // Jak Shadows u32 tex_is_fb : 1; // Jak Shadows
uint32 automatic_lod : 1; u32 automatic_lod : 1;
uint32 manual_lod : 1; u32 manual_lod : 1;
uint32 point_sampler : 1; u32 point_sampler : 1;
uint32 invalid_tex0 : 1; // Lupin the 3rd u32 invalid_tex0 : 1; // Lupin the 3rd
uint32 _free2 : 6; u32 _free2 : 6;
}; };
uint64 key; u64 key;
}; };
// FIXME is the & useful ? // FIXME is the & useful ?
operator uint64() const { return key; } operator u64() const { return key; }
PSSelector() PSSelector()
: key(0) : key(0)
@ -366,25 +366,25 @@ public:
{ {
struct struct
{ {
uint32 tau : 1; u32 tau : 1;
uint32 tav : 1; u32 tav : 1;
uint32 biln : 1; u32 biln : 1;
uint32 triln : 3; u32 triln : 3;
uint32 aniso : 1; u32 aniso : 1;
uint32 _free : 25; u32 _free : 25;
}; };
uint32 key; u32 key;
}; };
operator uint32() { return key; } operator u32() { return key; }
PSSamplerSelector() PSSamplerSelector()
: key(0) : key(0)
{ {
} }
PSSamplerSelector(uint32 k) PSSamplerSelector(u32 k)
: key(k) : key(k)
{ {
} }
@ -396,25 +396,25 @@ public:
{ {
struct struct
{ {
uint32 ztst : 2; u32 ztst : 2;
uint32 zwe : 1; u32 zwe : 1;
uint32 date : 1; u32 date : 1;
uint32 date_one : 1; u32 date_one : 1;
uint32 _free : 27; u32 _free : 27;
}; };
uint32 key; u32 key;
}; };
// FIXME is the & useful ? // FIXME is the & useful ?
operator uint32() { return key; } operator u32() { return key; }
OMDepthStencilSelector() OMDepthStencilSelector()
: key(0) : key(0)
{ {
} }
OMDepthStencilSelector(uint32 k) OMDepthStencilSelector(u32 k)
: key(k) : key(k)
{ {
} }
@ -426,30 +426,30 @@ public:
{ {
struct struct
{ {
uint32 wr : 1; u32 wr : 1;
uint32 wg : 1; u32 wg : 1;
uint32 wb : 1; u32 wb : 1;
uint32 wa : 1; u32 wa : 1;
uint32 _free : 28; u32 _free : 28;
}; };
struct struct
{ {
uint32 wrgba : 4; u32 wrgba : 4;
}; };
uint32 key; u32 key;
}; };
// FIXME is the & useful ? // FIXME is the & useful ?
operator uint32() { return key & 0xf; } operator u32() { return key & 0xf; }
OMColorMaskSelector() OMColorMaskSelector()
: key(0xF) : key(0xF)
{ {
} }
OMColorMaskSelector(uint32 c) { wrgba = c; } OMColorMaskSelector(u32 c) { wrgba = c; }
}; };
struct alignas(32) MiscConstantBuffer struct alignas(32) MiscConstantBuffer
@ -539,7 +539,7 @@ private:
struct struct
{ {
uint16 last_query; u16 last_query;
GLuint timer_query[1 << 16]; GLuint timer_query[1 << 16];
GLuint timer() { return timer_query[last_query]; } GLuint timer() { return timer_query[last_query]; }
@ -549,7 +549,7 @@ private:
GLuint m_gs[1 << 3]; GLuint m_gs[1 << 3];
GLuint m_ps_ss[1 << 7]; GLuint m_ps_ss[1 << 7];
GSDepthStencilOGL* m_om_dss[1 << 5]; GSDepthStencilOGL* m_om_dss[1 << 5];
std::unordered_map<uint64, GLuint> m_ps; std::unordered_map<u64, GLuint> m_ps;
GLuint m_apitrace; GLuint m_apitrace;
GLuint m_palette_ss; GLuint m_palette_ss;
@ -574,7 +574,7 @@ private:
void OMAttachDs(GSTextureOGL* ds = NULL); void OMAttachDs(GSTextureOGL* ds = NULL);
void OMSetFBO(GLuint fbo); void OMSetFBO(GLuint fbo);
uint16 ConvertBlendEnum(uint16 generic) final; u16 ConvertBlendEnum(u16 generic) final;
public: public:
GSShaderOGL* m_shader; GSShaderOGL* m_shader;
@ -597,9 +597,9 @@ public:
void DrawIndexedPrimitive(int offset, int count) final; void DrawIndexedPrimitive(int offset, int count) final;
void ClearRenderTarget(GSTexture* t, const GSVector4& c) final; void ClearRenderTarget(GSTexture* t, const GSVector4& c) final;
void ClearRenderTarget(GSTexture* t, uint32 c) final; void ClearRenderTarget(GSTexture* t, u32 c) final;
void ClearDepth(GSTexture* t) final; void ClearDepth(GSTexture* t) final;
void ClearStencil(GSTexture* t, uint8 c) final; void ClearStencil(GSTexture* t, u8 c) final;
void InitPrimDateTexture(GSTexture* rt, const GSVector4i& area); void InitPrimDateTexture(GSTexture* rt, const GSVector4i& area);
void RecycleDateTexture(); void RecycleDateTexture();
@ -624,7 +624,7 @@ public:
void PSSetSamplerState(GLuint ss); void PSSetSamplerState(GLuint ss);
void OMSetDepthStencilState(GSDepthStencilOGL* dss); void OMSetDepthStencilState(GSDepthStencilOGL* dss);
void OMSetBlendState(uint8 blend_index = 0, uint8 blend_factor = 0, bool is_blend_constant = false, bool accumulation_blend = false); void OMSetBlendState(u8 blend_index = 0, u8 blend_factor = 0, bool is_blend_constant = false, bool accumulation_blend = false);
void OMSetRenderTargets(GSTexture* rt, GSTexture* ds, const GSVector4i* scissor = NULL) final; void OMSetRenderTargets(GSTexture* rt, GSTexture* ds, const GSVector4i* scissor = NULL) final;
void OMSetColorMaskState(OMColorMaskSelector sel = OMColorMaskSelector()); void OMSetColorMaskState(OMColorMaskSelector sel = OMColorMaskSelector());

36
pcsx2/GS/Renderers/OpenGL/GSRendererOGL.cpp
View File

@ -131,8 +131,8 @@ void GSRendererOGL::EmulateZbuffer()
// On the real GS we appear to do clamping on the max z value the format allows. // On the real GS we appear to do clamping on the max z value the format allows.
// Clamping is done after rasterization. // Clamping is done after rasterization.
const uint32 max_z = 0xFFFFFFFF >> (GSLocalMemory::m_psm[m_context->ZBUF.PSM].fmt * 8); const u32 max_z = 0xFFFFFFFF >> (GSLocalMemory::m_psm[m_context->ZBUF.PSM].fmt * 8);
const bool clamp_z = (uint32)(GSVector4i(m_vt.m_max.p).z) > max_z; const bool clamp_z = (u32)(GSVector4i(m_vt.m_max.p).z) > max_z;
vs_cb.MaxDepth = GSVector2i(0xFFFFFFFF); vs_cb.MaxDepth = GSVector2i(0xFFFFFFFF);
//ps_cb.MaxDepth = GSVector4(0.0f, 0.0f, 0.0f, 1.0f); //ps_cb.MaxDepth = GSVector4(0.0f, 0.0f, 0.0f, 1.0f);
@ -184,11 +184,11 @@ void GSRendererOGL::EmulateTextureShuffleAndFbmask()
// Please bang my head against the wall! // Please bang my head against the wall!
// 1/ Reduce the frame mask to a 16 bit format // 1/ Reduce the frame mask to a 16 bit format
const uint32& m = m_context->FRAME.FBMSK; const u32& m = m_context->FRAME.FBMSK;
const uint32 fbmask = ((m >> 3) & 0x1F) | ((m >> 6) & 0x3E0) | ((m >> 9) & 0x7C00) | ((m >> 16) & 0x8000); const u32 fbmask = ((m >> 3) & 0x1F) | ((m >> 6) & 0x3E0) | ((m >> 9) & 0x7C00) | ((m >> 16) & 0x8000);
// FIXME GSVector will be nice here // FIXME GSVector will be nice here
const uint8 rg_mask = fbmask & 0xFF; const u8 rg_mask = fbmask & 0xFF;
const uint8 ba_mask = (fbmask >> 8) & 0xFF; const u8 ba_mask = (fbmask >> 8) & 0xFF;
m_om_csel.wrgba = 0; m_om_csel.wrgba = 0;
// 2 Select the new mask (Please someone put SSE here) // 2 Select the new mask (Please someone put SSE here)
@ -443,10 +443,10 @@ void GSRendererOGL::EmulateChannelShuffle(GSTexture** rt, const GSTextureCache::
// the rendered size of the framebuffer // the rendered size of the framebuffer
GSVertex* s = &m_vertex.buff[0]; GSVertex* s = &m_vertex.buff[0];
s[0].XYZ.X = (uint16)(m_context->XYOFFSET.OFX + 0); s[0].XYZ.X = (u16)(m_context->XYOFFSET.OFX + 0);
s[1].XYZ.X = (uint16)(m_context->XYOFFSET.OFX + 16384); s[1].XYZ.X = (u16)(m_context->XYOFFSET.OFX + 16384);
s[0].XYZ.Y = (uint16)(m_context->XYOFFSET.OFY + 0); s[0].XYZ.Y = (u16)(m_context->XYOFFSET.OFY + 0);
s[1].XYZ.Y = (uint16)(m_context->XYOFFSET.OFY + 16384); s[1].XYZ.Y = (u16)(m_context->XYOFFSET.OFY + 16384);
m_vertex.head = m_vertex.tail = m_vertex.next = 2; m_vertex.head = m_vertex.tail = m_vertex.next = 2;
m_index.tail = 2; m_index.tail = 2;
@ -473,7 +473,7 @@ void GSRendererOGL::EmulateBlending(bool& DATE_GL42, bool& DATE_GL45)
} }
// Compute the blending equation to detect special case // Compute the blending equation to detect special case
const uint8 blend_index = uint8(((ALPHA.A * 3 + ALPHA.B) * 3 + ALPHA.C) * 3 + ALPHA.D); const u8 blend_index = u8(((ALPHA.A * 3 + ALPHA.B) * 3 + ALPHA.C) * 3 + ALPHA.D);
const int blend_flag = m_dev->GetBlendFlags(blend_index); const int blend_flag = m_dev->GetBlendFlags(blend_index);
// SW Blend is (nearly) free. Let's use it. // SW Blend is (nearly) free. Let's use it.
@ -627,7 +627,7 @@ void GSRendererOGL::EmulateBlending(bool& DATE_GL42, bool& DATE_GL45)
if (m_ps_sel.dfmt == 1 && ALPHA.C == 1) if (m_ps_sel.dfmt == 1 && ALPHA.C == 1)
{ {
// 24 bits doesn't have an alpha channel so use 1.0f fix factor as equivalent // 24 bits doesn't have an alpha channel so use 1.0f fix factor as equivalent
const uint8 hacked_blend_index = blend_index + 3; // +3 <=> +1 on C const u8 hacked_blend_index = blend_index + 3; // +3 <=> +1 on C
dev->OMSetBlendState(hacked_blend_index, 128, true); dev->OMSetBlendState(hacked_blend_index, 128, true);
} }
else else
@ -646,8 +646,8 @@ void GSRendererOGL::EmulateTextureSampler(const GSTextureCache::Source* tex)
const GSLocalMemory::psm_t& psm = GSLocalMemory::m_psm[tex->m_TEX0.PSM]; const GSLocalMemory::psm_t& psm = GSLocalMemory::m_psm[tex->m_TEX0.PSM];
const GSLocalMemory::psm_t& cpsm = psm.pal > 0 ? GSLocalMemory::m_psm[m_context->TEX0.CPSM] : psm; const GSLocalMemory::psm_t& cpsm = psm.pal > 0 ? GSLocalMemory::m_psm[m_context->TEX0.CPSM] : psm;
const uint8 wms = m_context->CLAMP.WMS; const u8 wms = m_context->CLAMP.WMS;
const uint8 wmt = m_context->CLAMP.WMT; const u8 wmt = m_context->CLAMP.WMT;
const bool complex_wms_wmt = !!((wms | wmt) & 2); const bool complex_wms_wmt = !!((wms | wmt) & 2);
const bool need_mipmap = IsMipMapDraw(); const bool need_mipmap = IsMipMapDraw();
@ -660,7 +660,7 @@ void GSRendererOGL::EmulateTextureSampler(const GSTextureCache::Source* tex)
switch (UserHacks_tri_filter) switch (UserHacks_tri_filter)
{ {
case TriFiltering::Forced: case TriFiltering::Forced:
trilinear = static_cast<uint8>(GS_MIN_FILTER::Linear_Mipmap_Linear); trilinear = static_cast<u8>(GS_MIN_FILTER::Linear_Mipmap_Linear);
trilinear_auto = m_mipmap != 2; trilinear_auto = m_mipmap != 2;
break; break;
@ -1323,7 +1323,7 @@ void GSRendererOGL::DrawPrims(GSTexture* rt, GSTexture* ds, GSTextureCache::Sour
{ {
m_ps_sel.fog = 1; m_ps_sel.fog = 1;
const GSVector4 fc = GSVector4::rgba32(m_env.FOGCOL.u32[0]); const GSVector4 fc = GSVector4::rgba32(m_env.FOGCOL.U32[0]);
// Blend AREF to avoid to load a random value for alpha (dirty cache) // Blend AREF to avoid to load a random value for alpha (dirty cache)
ps_cb.FogColor_AREF = fc.blend32<8>(ps_cb.FogColor_AREF); ps_cb.FogColor_AREF = fc.blend32<8>(ps_cb.FogColor_AREF);
} }
@ -1334,7 +1334,7 @@ void GSRendererOGL::DrawPrims(GSTexture* rt, GSTexture* ds, GSTextureCache::Sour
// pass to handle the depth based on the alpha test. // pass to handle the depth based on the alpha test.
bool ate_RGBA_then_Z = false; bool ate_RGBA_then_Z = false;
bool ate_RGB_then_ZA = false; bool ate_RGB_then_ZA = false;
uint8 ps_atst = 0; u8 ps_atst = 0;
if (ate_first_pass & ate_second_pass) if (ate_first_pass & ate_second_pass)
{ {
GL_DBG("Complex Alpha Test"); GL_DBG("Complex Alpha Test");
@ -1405,7 +1405,7 @@ void GSRendererOGL::DrawPrims(GSTexture* rt, GSTexture* ds, GSTextureCache::Sour
// We need the palette to convert the depth to the correct alpha value. // We need the palette to convert the depth to the correct alpha value.
if (!tex->m_palette) if (!tex->m_palette)
{ {
const uint16 pal = GSLocalMemory::m_psm[tex->m_TEX0.PSM].pal; const u16 pal = GSLocalMemory::m_psm[tex->m_TEX0.PSM].pal;
m_tc->AttachPaletteToSource(tex, pal, true); m_tc->AttachPaletteToSource(tex, pal, true);
dev->PSSetShaderResource(1, tex->m_palette); dev->PSSetShaderResource(1, tex->m_palette);
} }

2
pcsx2/GS/Renderers/OpenGL/GSShaderOGL.cpp
View File

@ -63,7 +63,7 @@ GLuint GSShaderOGL::LinkPipeline(const std::string& pretty_print, GLuint vs, GLu
GLuint GSShaderOGL::LinkProgram(GLuint vs, GLuint gs, GLuint ps) GLuint GSShaderOGL::LinkProgram(GLuint vs, GLuint gs, GLuint ps)
{ {
uint32 hash = ((vs ^ gs) << 24) ^ ps; u32 hash = ((vs ^ gs) << 24) ^ ps;
auto it = m_program.find(hash); auto it = m_program.find(hash);
if (it != m_program.end()) if (it != m_program.end())
return it->second; return it->second;

2
pcsx2/GS/Renderers/OpenGL/GSShaderOGL.h
View File

@ -20,7 +20,7 @@
class GSShaderOGL class GSShaderOGL
{ {
GLuint m_pipeline; GLuint m_pipeline;
std::unordered_map<uint32, GLuint> m_program; std::unordered_map<u32, GLuint> m_program;
const bool m_debug_shader; const bool m_debug_shader;
std::vector<GLuint> m_shad_to_delete; std::vector<GLuint> m_shad_to_delete;

34
pcsx2/GS/Renderers/OpenGL/GSTextureOGL.cpp
View File

@ -20,20 +20,20 @@
#include "GS/GSPng.h" #include "GS/GSPng.h"
#ifdef ENABLE_OGL_DEBUG_MEM_BW #ifdef ENABLE_OGL_DEBUG_MEM_BW
extern uint64 g_real_texture_upload_byte; extern u64 g_real_texture_upload_byte;
#endif #endif
// FIXME OGL4: investigate, only 1 unpack buffer always bound // FIXME OGL4: investigate, only 1 unpack buffer always bound
namespace PboPool namespace PboPool
{ {
const uint32 m_pbo_size = 64 * 1024 * 1024; const u32 m_pbo_size = 64 * 1024 * 1024;
const uint32 m_seg_size = 16 * 1024 * 1024; const u32 m_seg_size = 16 * 1024 * 1024;
GLuint m_buffer; GLuint m_buffer;
uptr m_offset; uptr m_offset;
char* m_map; char* m_map;
uint32 m_size; u32 m_size;
GLsync m_fence[m_pbo_size / m_seg_size]; GLsync m_fence[m_pbo_size / m_seg_size];
// Option for buffer storage // Option for buffer storage
@ -61,7 +61,7 @@ namespace PboPool
UnbindPbo(); UnbindPbo();
} }
char* Map(uint32 size) char* Map(u32 size)
{ {
char* map; char* map;
// Note: keep offset aligned for SSE/AVX // Note: keep offset aligned for SSE/AVX
@ -113,8 +113,8 @@ namespace PboPool
void Sync() void Sync()
{ {
uint32 segment_current = m_offset / m_seg_size; u32 segment_current = m_offset / m_seg_size;
uint32 segment_next = (m_offset + m_size) / m_seg_size; u32 segment_next = (m_offset + m_size) / m_seg_size;
if (segment_current != segment_next) if (segment_current != segment_next)
{ {
@ -256,7 +256,7 @@ GSTextureOGL::GSTextureOGL(int type, int w, int h, int format, GLuint fbo_read,
return; // backbuffer isn't a real texture return; // backbuffer isn't a real texture
case GSTexture::Offscreen: case GSTexture::Offscreen:
// Offscreen is only used to read color. So it only requires 4B by pixel // Offscreen is only used to read color. So it only requires 4B by pixel
m_local_buffer = (uint8*)_aligned_malloc(m_size.x * m_size.y * 4, 32); m_local_buffer = (u8*)_aligned_malloc(m_size.x * m_size.y * 4, 32);
break; break;
case GSTexture::Texture: case GSTexture::Texture:
// Only 32 bits input texture will be supported for mipmap // Only 32 bits input texture will be supported for mipmap
@ -391,8 +391,8 @@ bool GSTextureOGL::Update(const GSVector4i& r, const void* data, int pitch, int
m_clean = false; m_clean = false;
uint32 row_byte = r.width() << m_int_shift; u32 row_byte = r.width() << m_int_shift;
uint32 map_size = r.height() * row_byte; u32 map_size = r.height() * row_byte;
#ifdef ENABLE_OGL_DEBUG_MEM_BW #ifdef ENABLE_OGL_DEBUG_MEM_BW
g_real_texture_upload_byte += map_size; g_real_texture_upload_byte += map_size;
#endif #endif
@ -457,7 +457,7 @@ bool GSTextureOGL::Map(GSMap& m, const GSVector4i* _r, int layer)
ASSERT(r.width() != 0); ASSERT(r.width() != 0);
ASSERT(r.height() != 0); ASSERT(r.height() != 0);
uint32 row_byte = r.width() << m_int_shift; u32 row_byte = r.width() << m_int_shift;
m.pitch = row_byte; m.pitch = row_byte;
if (m_type == GSTexture::Offscreen) if (m_type == GSTexture::Offscreen)
@ -495,9 +495,9 @@ bool GSTextureOGL::Map(GSMap& m, const GSVector4i* _r, int layer)
m_clean = false; m_clean = false;
uint32 map_size = r.height() * row_byte; u32 map_size = r.height() * row_byte;
m.bits = (uint8*)PboPool::Map(map_size); m.bits = (u8*)PboPool::Map(map_size);
#ifdef ENABLE_OGL_DEBUG_MEM_BW #ifdef ENABLE_OGL_DEBUG_MEM_BW
g_real_texture_upload_byte += map_size; g_real_texture_upload_byte += map_size;
@ -586,9 +586,9 @@ void GSTextureOGL::CommitPages(const GSVector2i& region, bool commit)
bool GSTextureOGL::Save(const std::string& fn) bool GSTextureOGL::Save(const std::string& fn)
{ {
// Collect the texture data // Collect the texture data
uint32 pitch = 4 * m_committed_size.x; u32 pitch = 4 * m_committed_size.x;
uint32 buf_size = pitch * m_committed_size.y * 2; // Note *2 for security (depth/stencil) u32 buf_size = pitch * m_committed_size.y * 2; // Note *2 for security (depth/stencil)
std::unique_ptr<uint8[]> image(new uint8[buf_size]); std::unique_ptr<u8[]> image(new u8[buf_size]);
#ifdef ENABLE_OGL_DEBUG #ifdef ENABLE_OGL_DEBUG
GSPng::Format fmt = GSPng::RGB_A_PNG; GSPng::Format fmt = GSPng::RGB_A_PNG;
#else #else
@ -646,7 +646,7 @@ bool GSTextureOGL::Save(const std::string& fn)
return GSPng::Save(fmt, fn, image.get(), m_committed_size.x, m_committed_size.y, pitch, compression); return GSPng::Save(fmt, fn, image.get(), m_committed_size.x, m_committed_size.y, pitch, compression);
} }
uint32 GSTextureOGL::GetMemUsage() u32 GSTextureOGL::GetMemUsage()
{ {
return m_mem_usage; return m_mem_usage;
} }

12
pcsx2/GS/Renderers/OpenGL/GSTextureOGL.h
View File

@ -24,7 +24,7 @@ namespace PboPool
inline void UnbindPbo(); inline void UnbindPbo();
inline void Sync(); inline void Sync();
inline char* Map(uint32 size); inline char* Map(u32 size);
inline void Unmap(); inline void Unmap();
inline uptr Offset(); inline uptr Offset();
inline void EndTransfer(); inline void EndTransfer();
@ -41,7 +41,7 @@ private:
bool m_clean; bool m_clean;
bool m_generate_mipmap; bool m_generate_mipmap;
uint8* m_local_buffer; u8* m_local_buffer;
// Avoid alignment constrain // Avoid alignment constrain
//GSVector4i m_r; //GSVector4i m_r;
int m_r_x; int m_r_x;
@ -54,10 +54,10 @@ private:
// internal opengl format/type/alignment // internal opengl format/type/alignment
GLenum m_int_format; GLenum m_int_format;
GLenum m_int_type; GLenum m_int_type;
uint32 m_int_shift; u32 m_int_shift;
// Allow to track size of allocated memory // Allow to track size of allocated memory
uint32 m_mem_usage; u32 m_mem_usage;
public: public:
explicit GSTextureOGL(int type, int w, int h, int format, GLuint fbo_read, bool mipmap); explicit GSTextureOGL(int type, int w, int h, int format, GLuint fbo_read, bool mipmap);
@ -72,7 +72,7 @@ public:
bool IsBackbuffer() { return (m_type == GSTexture::Backbuffer); } bool IsBackbuffer() { return (m_type == GSTexture::Backbuffer); }
bool IsDss() { return (m_type == GSTexture::DepthStencil || m_type == GSTexture::SparseDepthStencil); } bool IsDss() { return (m_type == GSTexture::DepthStencil || m_type == GSTexture::SparseDepthStencil); }
uint32 GetID() final { return m_texture_id; } u32 GetID() final { return m_texture_id; }
bool HasBeenCleaned() { return m_clean; } bool HasBeenCleaned() { return m_clean; }
void WasAttached() { m_clean = false; } void WasAttached() { m_clean = false; }
void WasCleaned() { m_clean = true; } void WasCleaned() { m_clean = true; }
@ -82,5 +82,5 @@ public:
void CommitPages(const GSVector2i& region, bool commit) final; void CommitPages(const GSVector2i& region, bool commit) final;
uint32 GetMemUsage(); u32 GetMemUsage();
}; };

10
pcsx2/GS/Renderers/OpenGL/GSUniformBufferOGL.h
View File

@ -18,7 +18,7 @@
#include "GLState.h" #include "GLState.h"
#ifdef ENABLE_OGL_DEBUG_MEM_BW #ifdef ENABLE_OGL_DEBUG_MEM_BW
extern uint64 g_uniform_upload_byte; extern u64 g_uniform_upload_byte;
#endif #endif
@ -26,11 +26,11 @@ class GSUniformBufferOGL
{ {
GLuint m_buffer; // data object GLuint m_buffer; // data object
GLuint m_index; // GLSL slot GLuint m_index; // GLSL slot
uint32 m_size; // size of the data u32 m_size; // size of the data
uint8* m_cache; // content of the previous upload u8* m_cache; // content of the previous upload
public: public:
GSUniformBufferOGL(const std::string& pretty_name, GLuint index, uint32 size) GSUniformBufferOGL(const std::string& pretty_name, GLuint index, u32 size)
: m_index(index), m_size(size) : m_index(index), m_size(size)
{ {
glGenBuffers(1, &m_buffer); glGenBuffers(1, &m_buffer);
@ -38,7 +38,7 @@ public:
glObjectLabel(GL_BUFFER, m_buffer, pretty_name.size(), pretty_name.c_str()); glObjectLabel(GL_BUFFER, m_buffer, pretty_name.size(), pretty_name.c_str());
allocate(); allocate();
attach(); attach();
m_cache = (uint8*)_aligned_malloc(m_size, 32); m_cache = (u8*)_aligned_malloc(m_size, 32);
memset(m_cache, 0, m_size); memset(m_cache, 0, m_size);
} }

244
pcsx2/GS/Renderers/SW/GSDrawScanline.cpp
View File

@ -103,14 +103,14 @@ void GSDrawScanline::BeginDraw(const GSRasterizerData* data)
m_sp = m_sp_map[sel]; m_sp = m_sp_map[sel];
} }
void GSDrawScanline::EndDraw(uint64 frame, uint64 ticks, int actual, int total, int prims) void GSDrawScanline::EndDraw(u64 frame, u64 ticks, int actual, int total, int prims)
{ {
m_ds_map.UpdateStats(frame, ticks, actual, total, prims); m_ds_map.UpdateStats(frame, ticks, actual, total, prims);
} }
#ifndef ENABLE_JIT_RASTERIZER #ifndef ENABLE_JIT_RASTERIZER
void GSDrawScanline::SetupPrim(const GSVertexSW* vertex, const uint32* index, const GSVertexSW& dscan) void GSDrawScanline::SetupPrim(const GSVertexSW* vertex, const u32* index, const GSVertexSW& dscan)
{ {
GSScanlineSelector sel = m_global.sel; GSScanlineSelector sel = m_global.sel;
@ -162,7 +162,7 @@ void GSDrawScanline::SetupPrim(const GSVertexSW* vertex, const uint32* index, co
if (has_z) if (has_z)
{ {
m_local.p.z = vertex[index[1]].t.u32[3]; // uint32 z is bypassed in t.w m_local.p.z = vertex[index[1]].t.U32[3]; // u32 z is bypassed in t.w
} }
} }
} }
@ -315,7 +315,7 @@ void GSDrawScanline::SetupPrim(const GSVertexSW* vertex, const uint32* index, co
if (has_z) if (has_z)
{ {
m_local.p.z = vertex[index[1]].t.u32[3]; // uint32 z is bypassed in t.w m_local.p.z = vertex[index[1]].t.U32[3]; // u32 z is bypassed in t.w
} }
} }
} }
@ -564,8 +564,8 @@ void GSDrawScanline::DrawScanline(int pixels, int left, int top, const GSVertexS
if (sel.ztest) if (sel.ztest)
{ {
zd = GSVector8i::load( zd = GSVector8i::load(
(uint8*)m_global.vm + za * 2, (uint8*)m_global.vm + za * 2 + 16, (u8*)m_global.vm + za * 2, (u8*)m_global.vm + za * 2 + 16,
(uint8*)m_global.vm + za * 2 + 32, (uint8*)m_global.vm + za * 2 + 48); (u8*)m_global.vm + za * 2 + 32, (u8*)m_global.vm + za * 2 + 48);
switch (sel.zpsm) switch (sel.zpsm)
{ {
@ -717,24 +717,24 @@ void GSDrawScanline::DrawScanline(int pixels, int left, int top, const GSVertexS
{ {
for (int i = 0; i < 8; i++) for (int i = 0; i < 8; i++)
{ {
const uint8* tex = (const uint8*)m_global.tex[lodi.u32[i]]; const u8* tex = (const u8*)m_global.tex[lodi.U32[i]];
c00.u32[i] = m_global.clut[tex[addr00.u32[i]]]; c00.U32[i] = m_global.clut[tex[addr00.U32[i]]];
c01.u32[i] = m_global.clut[tex[addr01.u32[i]]]; c01.U32[i] = m_global.clut[tex[addr01.U32[i]]];
c10.u32[i] = m_global.clut[tex[addr10.u32[i]]]; c10.U32[i] = m_global.clut[tex[addr10.U32[i]]];
c11.u32[i] = m_global.clut[tex[addr11.u32[i]]]; c11.U32[i] = m_global.clut[tex[addr11.U32[i]]];
} }
} }
else else
{ {
for (int i = 0; i < 8; i++) for (int i = 0; i < 8; i++)
{ {
const uint32* tex = (const uint32*)m_global.tex[lodi.u32[i]]; const u32* tex = (const u32*)m_global.tex[lodi.U32[i]];
c00.u32[i] = tex[addr00.u32[i]]; c00.U32[i] = tex[addr00.U32[i]];
c01.u32[i] = tex[addr01.u32[i]]; c01.U32[i] = tex[addr01.U32[i]];
c10.u32[i] = tex[addr10.u32[i]]; c10.U32[i] = tex[addr10.U32[i]];
c11.u32[i] = tex[addr11.u32[i]]; c11.U32[i] = tex[addr11.U32[i]];
} }
} }
@ -765,14 +765,14 @@ void GSDrawScanline::DrawScanline(int pixels, int left, int top, const GSVertexS
{ {
for (int i = 0; i < 8; i++) for (int i = 0; i < 8; i++)
{ {
c00.u32[i] = m_global.clut[((const uint8*)m_global.tex[lodi.u32[i]])[addr00.u32[i]]]; c00.U32[i] = m_global.clut[((const u8*)m_global.tex[lodi.U32[i]])[addr00.U32[i]]];
} }
} }
else else
{ {
for (int i = 0; i < 8; i++) for (int i = 0; i < 8; i++)
{ {
c00.u32[i] = ((const uint32*)m_global.tex[lodi.u32[i]])[addr00.u32[i]]; c00.U32[i] = ((const u32*)m_global.tex[lodi.U32[i]])[addr00.U32[i]];
} }
} }
@ -838,24 +838,24 @@ void GSDrawScanline::DrawScanline(int pixels, int left, int top, const GSVertexS
{ {
for (int i = 0; i < 8; i++) for (int i = 0; i < 8; i++)
{ {
const uint8* tex = (const uint8*)m_global.tex[lodi.u32[i]]; const u8* tex = (const u8*)m_global.tex[lodi.U32[i]];
c00.u32[i] = m_global.clut[tex[addr00.u32[i]]]; c00.U32[i] = m_global.clut[tex[addr00.U32[i]]];
c01.u32[i] = m_global.clut[tex[addr01.u32[i]]]; c01.U32[i] = m_global.clut[tex[addr01.U32[i]]];
c10.u32[i] = m_global.clut[tex[addr10.u32[i]]]; c10.U32[i] = m_global.clut[tex[addr10.U32[i]]];
c11.u32[i] = m_global.clut[tex[addr11.u32[i]]]; c11.U32[i] = m_global.clut[tex[addr11.U32[i]]];
} }
} }
else else
{ {
for (int i = 0; i < 8; i++) for (int i = 0; i < 8; i++)
{ {
const uint32* tex = (const uint32*)m_global.tex[lodi.u32[i]]; const u32* tex = (const u32*)m_global.tex[lodi.U32[i]];
c00.u32[i] = tex[addr00.u32[i]]; c00.U32[i] = tex[addr00.U32[i]];
c01.u32[i] = tex[addr01.u32[i]]; c01.U32[i] = tex[addr01.U32[i]];
c10.u32[i] = tex[addr10.u32[i]]; c10.U32[i] = tex[addr10.U32[i]];
c11.u32[i] = tex[addr11.u32[i]]; c11.U32[i] = tex[addr11.U32[i]];
} }
} }
@ -886,14 +886,14 @@ void GSDrawScanline::DrawScanline(int pixels, int left, int top, const GSVertexS
{ {
for (int i = 0; i < 8; i++) for (int i = 0; i < 8; i++)
{ {
c00.u32[i] = m_global.clut[((const uint8*)m_global.tex[lodi.u32[i]])[addr00.u32[i]]]; c00.U32[i] = m_global.clut[((const u8*)m_global.tex[lodi.U32[i]])[addr00.U32[i]]];
} }
} }
else else
{ {
for (int i = 0; i < 8; i++) for (int i = 0; i < 8; i++)
{ {
c00.u32[i] = ((const uint32*)m_global.tex[lodi.u32[i]])[addr00.u32[i]]; c00.U32[i] = ((const u32*)m_global.tex[lodi.U32[i]])[addr00.U32[i]];
} }
} }
@ -977,7 +977,7 @@ void GSDrawScanline::DrawScanline(int pixels, int left, int top, const GSVertexS
if (sel.tlu) if (sel.tlu)
{ {
const uint8* tex = (const uint8*)m_global.tex[0]; const u8* tex = (const u8*)m_global.tex[0];
c00 = addr00.gather32_32(tex, m_global.clut); c00 = addr00.gather32_32(tex, m_global.clut);
c01 = addr01.gather32_32(tex, m_global.clut); c01 = addr01.gather32_32(tex, m_global.clut);
@ -986,7 +986,7 @@ void GSDrawScanline::DrawScanline(int pixels, int left, int top, const GSVertexS
} }
else else
{ {
const uint32* tex = (const uint32*)m_global.tex[0]; const u32* tex = (const u32*)m_global.tex[0];
c00 = addr00.gather32_32(tex); c00 = addr00.gather32_32(tex);
c01 = addr01.gather32_32(tex); c01 = addr01.gather32_32(tex);
@ -1019,11 +1019,11 @@ void GSDrawScanline::DrawScanline(int pixels, int left, int top, const GSVertexS
if (sel.tlu) if (sel.tlu)
{ {
c00 = addr00.gather32_32((const uint8*)m_global.tex[0], m_global.clut); c00 = addr00.gather32_32((const u8*)m_global.tex[0], m_global.clut);
} }
else else
{ {
c00 = addr00.gather32_32((const uint32*)m_global.tex[0]); c00 = addr00.gather32_32((const u32*)m_global.tex[0]);
} }
rb = c00.sll16(8).srl16(8); rb = c00.sll16(8).srl16(8);
@ -1149,8 +1149,8 @@ void GSDrawScanline::DrawScanline(int pixels, int left, int top, const GSVertexS
if (sel.rfb) if (sel.rfb)
{ {
fd = GSVector8i::load( fd = GSVector8i::load(
(uint8*)m_global.vm + fa * 2, (uint8*)m_global.vm + fa * 2 + 16, (u8*)m_global.vm + fa * 2, (u8*)m_global.vm + fa * 2 + 16,
(uint8*)m_global.vm + fa * 2 + 32, (uint8*)m_global.vm + fa * 2 + 48); (u8*)m_global.vm + fa * 2 + 32, (u8*)m_global.vm + fa * 2 + 48);
} }
} }
@ -1234,10 +1234,10 @@ void GSDrawScanline::DrawScanline(int pixels, int left, int top, const GSVertexS
{ {
if (fast) if (fast)
{ {
GSVector4i::storel((uint8*)m_global.vm + za * 2, zs.extract<0>()); GSVector4i::storel((u8*)m_global.vm + za * 2, zs.extract<0>());
GSVector4i::storeh((uint8*)m_global.vm + za * 2 + 16, zs.extract<0>()); GSVector4i::storeh((u8*)m_global.vm + za * 2 + 16, zs.extract<0>());
GSVector4i::storel((uint8*)m_global.vm + za * 2 + 32, zs.extract<1>()); GSVector4i::storel((u8*)m_global.vm + za * 2 + 32, zs.extract<1>());
GSVector4i::storeh((uint8*)m_global.vm + za * 2 + 48, zs.extract<1>()); GSVector4i::storeh((u8*)m_global.vm + za * 2 + 48, zs.extract<1>());
} }
else else
{ {
@ -1255,10 +1255,10 @@ void GSDrawScanline::DrawScanline(int pixels, int left, int top, const GSVertexS
{ {
if (fast) if (fast)
{ {
if (fzm & 0x00000f00) GSVector4i::storel((uint8*)m_global.vm + za * 2, zs.extract<0>()); if (fzm & 0x00000f00) GSVector4i::storel((u8*)m_global.vm + za * 2, zs.extract<0>());
if (fzm & 0x0000f000) GSVector4i::storeh((uint8*)m_global.vm + za * 2 + 16, zs.extract<0>()); if (fzm & 0x0000f000) GSVector4i::storeh((u8*)m_global.vm + za * 2 + 16, zs.extract<0>());
if (fzm & 0x0f000000) GSVector4i::storel((uint8*)m_global.vm + za * 2 + 32, zs.extract<1>()); if (fzm & 0x0f000000) GSVector4i::storel((u8*)m_global.vm + za * 2 + 32, zs.extract<1>());
if (fzm & 0xf0000000) GSVector4i::storeh((uint8*)m_global.vm + za * 2 + 48, zs.extract<1>()); if (fzm & 0xf0000000) GSVector4i::storeh((u8*)m_global.vm + za * 2 + 48, zs.extract<1>());
} }
else else
{ {
@ -1443,10 +1443,10 @@ void GSDrawScanline::DrawScanline(int pixels, int left, int top, const GSVertexS
{ {
if (fast) if (fast)
{ {
GSVector4i::storel((uint8*)m_global.vm + fa * 2, fs.extract<0>()); GSVector4i::storel((u8*)m_global.vm + fa * 2, fs.extract<0>());
GSVector4i::storeh((uint8*)m_global.vm + fa * 2 + 16, fs.extract<0>()); GSVector4i::storeh((u8*)m_global.vm + fa * 2 + 16, fs.extract<0>());
GSVector4i::storel((uint8*)m_global.vm + fa * 2 + 32, fs.extract<1>()); GSVector4i::storel((u8*)m_global.vm + fa * 2 + 32, fs.extract<1>());
GSVector4i::storeh((uint8*)m_global.vm + fa * 2 + 48, fs.extract<1>()); GSVector4i::storeh((u8*)m_global.vm + fa * 2 + 48, fs.extract<1>());
} }
else else
{ {
@ -1464,10 +1464,10 @@ void GSDrawScanline::DrawScanline(int pixels, int left, int top, const GSVertexS
{ {
if (fast) if (fast)
{ {
if (fzm & 0x0000000f) GSVector4i::storel((uint8*)m_global.vm + fa * 2, fs.extract<0>()); if (fzm & 0x0000000f) GSVector4i::storel((u8*)m_global.vm + fa * 2, fs.extract<0>());
if (fzm & 0x000000f0) GSVector4i::storeh((uint8*)m_global.vm + fa * 2 + 16, fs.extract<0>()); if (fzm & 0x000000f0) GSVector4i::storeh((u8*)m_global.vm + fa * 2 + 16, fs.extract<0>());
if (fzm & 0x000f0000) GSVector4i::storel((uint8*)m_global.vm + fa * 2 + 32, fs.extract<1>()); if (fzm & 0x000f0000) GSVector4i::storel((u8*)m_global.vm + fa * 2 + 32, fs.extract<1>());
if (fzm & 0x00f00000) GSVector4i::storeh((uint8*)m_global.vm + fa * 2 + 48, fs.extract<1>()); if (fzm & 0x00f00000) GSVector4i::storeh((u8*)m_global.vm + fa * 2 + 48, fs.extract<1>());
} }
else else
{ {
@ -1688,7 +1688,7 @@ void GSDrawScanline::DrawScanline(int pixels, int left, int top, const GSVertexS
if (sel.ztest) if (sel.ztest)
{ {
zd = GSVector4i::load((uint8*)m_global.vm + za * 2, (uint8*)m_global.vm + za * 2 + 16); zd = GSVector4i::load((u8*)m_global.vm + za * 2, (u8*)m_global.vm + za * 2 + 16);
switch (sel.zpsm) switch (sel.zpsm)
{ {
@ -1710,10 +1710,10 @@ void GSDrawScanline::DrawScanline(int pixels, int left, int top, const GSVertexS
{ {
const unsigned int z_max = 0xffffffff >> (sel.zpsm * 8); const unsigned int z_max = 0xffffffff >> (sel.zpsm * 8);
zso.u32[0] = std::min(z_max, zso.u32[0]); zso.U32[0] = std::min(z_max, zso.U32[0]);
zso.u32[1] = std::min(z_max, zso.u32[1]); zso.U32[1] = std::min(z_max, zso.U32[1]);
zso.u32[2] = std::min(z_max, zso.u32[2]); zso.U32[2] = std::min(z_max, zso.U32[2]);
zso.u32[3] = std::min(z_max, zso.u32[3]); zso.U32[3] = std::min(z_max, zso.U32[3]);
} }
switch (sel.ztst) switch (sel.ztst)
@ -1860,24 +1860,24 @@ void GSDrawScanline::DrawScanline(int pixels, int left, int top, const GSVertexS
{ {
for (int i = 0; i < 4; i++) for (int i = 0; i < 4; i++)
{ {
const uint8* tex = (const uint8*)m_global.tex[lodi.u32[i]]; const u8* tex = (const u8*)m_global.tex[lodi.U32[i]];
c00.u32[i] = m_global.clut[tex[addr00.u32[i]]]; c00.U32[i] = m_global.clut[tex[addr00.U32[i]]];
c01.u32[i] = m_global.clut[tex[addr01.u32[i]]]; c01.U32[i] = m_global.clut[tex[addr01.U32[i]]];
c10.u32[i] = m_global.clut[tex[addr10.u32[i]]]; c10.U32[i] = m_global.clut[tex[addr10.U32[i]]];
c11.u32[i] = m_global.clut[tex[addr11.u32[i]]]; c11.U32[i] = m_global.clut[tex[addr11.U32[i]]];
} }
} }
else else
{ {
for (int i = 0; i < 4; i++) for (int i = 0; i < 4; i++)
{ {
const uint32* tex = (const uint32*)m_global.tex[lodi.u32[i]]; const u32* tex = (const u32*)m_global.tex[lodi.U32[i]];
c00.u32[i] = tex[addr00.u32[i]]; c00.U32[i] = tex[addr00.U32[i]];
c01.u32[i] = tex[addr01.u32[i]]; c01.U32[i] = tex[addr01.U32[i]];
c10.u32[i] = tex[addr10.u32[i]]; c10.U32[i] = tex[addr10.U32[i]];
c11.u32[i] = tex[addr11.u32[i]]; c11.U32[i] = tex[addr11.U32[i]];
} }
} }
@ -1908,14 +1908,14 @@ void GSDrawScanline::DrawScanline(int pixels, int left, int top, const GSVertexS
{ {
for (int i = 0; i < 4; i++) for (int i = 0; i < 4; i++)
{ {
c00.u32[i] = m_global.clut[((const uint8*)m_global.tex[lodi.u32[i]])[addr00.u32[i]]]; c00.U32[i] = m_global.clut[((const u8*)m_global.tex[lodi.U32[i]])[addr00.U32[i]]];
} }
} }
else else
{ {
for (int i = 0; i < 4; i++) for (int i = 0; i < 4; i++)
{ {
c00.u32[i] = ((const uint32*)m_global.tex[lodi.u32[i]])[addr00.u32[i]]; c00.U32[i] = ((const u32*)m_global.tex[lodi.U32[i]])[addr00.U32[i]];
} }
} }
@ -1981,24 +1981,24 @@ void GSDrawScanline::DrawScanline(int pixels, int left, int top, const GSVertexS
{ {
for (int i = 0; i < 4; i++) for (int i = 0; i < 4; i++)
{ {
const uint8* tex = (const uint8*)m_global.tex[lodi.u32[i]]; const u8* tex = (const u8*)m_global.tex[lodi.U32[i]];
c00.u32[i] = m_global.clut[tex[addr00.u32[i]]]; c00.U32[i] = m_global.clut[tex[addr00.U32[i]]];
c01.u32[i] = m_global.clut[tex[addr01.u32[i]]]; c01.U32[i] = m_global.clut[tex[addr01.U32[i]]];
c10.u32[i] = m_global.clut[tex[addr10.u32[i]]]; c10.U32[i] = m_global.clut[tex[addr10.U32[i]]];
c11.u32[i] = m_global.clut[tex[addr11.u32[i]]]; c11.U32[i] = m_global.clut[tex[addr11.U32[i]]];
} }
} }
else else
{ {
for (int i = 0; i < 4; i++) for (int i = 0; i < 4; i++)
{ {
const uint32* tex = (const uint32*)m_global.tex[lodi.u32[i]]; const u32* tex = (const u32*)m_global.tex[lodi.U32[i]];
c00.u32[i] = tex[addr00.u32[i]]; c00.U32[i] = tex[addr00.U32[i]];
c01.u32[i] = tex[addr01.u32[i]]; c01.U32[i] = tex[addr01.U32[i]];
c10.u32[i] = tex[addr10.u32[i]]; c10.U32[i] = tex[addr10.U32[i]];
c11.u32[i] = tex[addr11.u32[i]]; c11.U32[i] = tex[addr11.U32[i]];
} }
} }
@ -2029,14 +2029,14 @@ void GSDrawScanline::DrawScanline(int pixels, int left, int top, const GSVertexS
{ {
for (int i = 0; i < 4; i++) for (int i = 0; i < 4; i++)
{ {
c00.u32[i] = m_global.clut[((const uint8*)m_global.tex[lodi.u32[i]])[addr00.u32[i]]]; c00.U32[i] = m_global.clut[((const u8*)m_global.tex[lodi.U32[i]])[addr00.U32[i]]];
} }
} }
else else
{ {
for (int i = 0; i < 4; i++) for (int i = 0; i < 4; i++)
{ {
c00.u32[i] = ((const uint32*)m_global.tex[lodi.u32[i]])[addr00.u32[i]]; c00.U32[i] = ((const u32*)m_global.tex[lodi.U32[i]])[addr00.U32[i]];
} }
} }
@ -2117,7 +2117,7 @@ void GSDrawScanline::DrawScanline(int pixels, int left, int top, const GSVertexS
if (sel.tlu) if (sel.tlu)
{ {
const uint8* tex = (const uint8*)m_global.tex[0]; const u8* tex = (const u8*)m_global.tex[0];
c00 = addr00.gather32_32(tex, m_global.clut); c00 = addr00.gather32_32(tex, m_global.clut);
c01 = addr01.gather32_32(tex, m_global.clut); c01 = addr01.gather32_32(tex, m_global.clut);
@ -2126,7 +2126,7 @@ void GSDrawScanline::DrawScanline(int pixels, int left, int top, const GSVertexS
} }
else else
{ {
const uint32* tex = (const uint32*)m_global.tex[0]; const u32* tex = (const u32*)m_global.tex[0];
c00 = addr00.gather32_32(tex); c00 = addr00.gather32_32(tex);
c01 = addr01.gather32_32(tex); c01 = addr01.gather32_32(tex);
@ -2159,11 +2159,11 @@ void GSDrawScanline::DrawScanline(int pixels, int left, int top, const GSVertexS
if (sel.tlu) if (sel.tlu)
{ {
c00 = addr00.gather32_32((const uint8*)m_global.tex[0], m_global.clut); c00 = addr00.gather32_32((const u8*)m_global.tex[0], m_global.clut);
} }
else else
{ {
c00 = addr00.gather32_32((const uint32*)m_global.tex[0]); c00 = addr00.gather32_32((const u32*)m_global.tex[0]);
} }
rb = c00.sll16(8).srl16(8); rb = c00.sll16(8).srl16(8);
@ -2285,7 +2285,7 @@ void GSDrawScanline::DrawScanline(int pixels, int left, int top, const GSVertexS
if (sel.rfb) if (sel.rfb)
{ {
fd = GSVector4i::load((uint8*)m_global.vm + fa * 2, (uint8*)m_global.vm + fa * 2 + 16); fd = GSVector4i::load((u8*)m_global.vm + fa * 2, (u8*)m_global.vm + fa * 2 + 16);
} }
} }
@ -2364,10 +2364,10 @@ void GSDrawScanline::DrawScanline(int pixels, int left, int top, const GSVertexS
{ {
const unsigned int z_max = 0xffffffff >> (sel.zpsm * 8); const unsigned int z_max = 0xffffffff >> (sel.zpsm * 8);
zs.u32[0] = std::min(z_max, zs.u32[0]); zs.U32[0] = std::min(z_max, zs.U32[0]);
zs.u32[1] = std::min(z_max, zs.u32[1]); zs.U32[1] = std::min(z_max, zs.U32[1]);
zs.u32[2] = std::min(z_max, zs.u32[2]); zs.U32[2] = std::min(z_max, zs.U32[2]);
zs.u32[3] = std::min(z_max, zs.u32[3]); zs.U32[3] = std::min(z_max, zs.U32[3]);
} }
bool fast = sel.ztest ? sel.zpsm < 2 : sel.zpsm == 0 && sel.notest; bool fast = sel.ztest ? sel.zpsm < 2 : sel.zpsm == 0 && sel.notest;
@ -2376,8 +2376,8 @@ void GSDrawScanline::DrawScanline(int pixels, int left, int top, const GSVertexS
{ {
if (fast) if (fast)
{ {
GSVector4i::storel((uint8*)m_global.vm + za * 2, zs); GSVector4i::storel((u8*)m_global.vm + za * 2, zs);
GSVector4i::storeh((uint8*)m_global.vm + za * 2 + 16, zs); GSVector4i::storeh((u8*)m_global.vm + za * 2 + 16, zs);
} }
else else
{ {
@ -2391,8 +2391,8 @@ void GSDrawScanline::DrawScanline(int pixels, int left, int top, const GSVertexS
{ {
if (fast) if (fast)
{ {
if (fzm & 0x0f00) GSVector4i::storel((uint8*)m_global.vm + za * 2, zs); if (fzm & 0x0f00) GSVector4i::storel((u8*)m_global.vm + za * 2, zs);
if (fzm & 0xf000) GSVector4i::storeh((uint8*)m_global.vm + za * 2 + 16, zs); if (fzm & 0xf000) GSVector4i::storeh((u8*)m_global.vm + za * 2 + 16, zs);
} }
else else
{ {
@ -2573,8 +2573,8 @@ void GSDrawScanline::DrawScanline(int pixels, int left, int top, const GSVertexS
{ {
if (fast) if (fast)
{ {
GSVector4i::storel((uint8*)m_global.vm + fa * 2, fs); GSVector4i::storel((u8*)m_global.vm + fa * 2, fs);
GSVector4i::storeh((uint8*)m_global.vm + fa * 2 + 16, fs); GSVector4i::storeh((u8*)m_global.vm + fa * 2 + 16, fs);
} }
else else
{ {
@ -2588,8 +2588,8 @@ void GSDrawScanline::DrawScanline(int pixels, int left, int top, const GSVertexS
{ {
if (fast) if (fast)
{ {
if (fzm & 0x000f) GSVector4i::storel((uint8*)m_global.vm + fa * 2, fs); if (fzm & 0x000f) GSVector4i::storel((u8*)m_global.vm + fa * 2, fs);
if (fzm & 0x00f0) GSVector4i::storeh((uint8*)m_global.vm + fa * 2 + 16, fs); if (fzm & 0x00f0) GSVector4i::storeh((u8*)m_global.vm + fa * 2 + 16, fs);
} }
else else
{ {
@ -2675,8 +2675,8 @@ void GSDrawScanline::DrawScanline(int pixels, int left, int top, const GSVertexS
void GSDrawScanline::DrawEdge(int pixels, int left, int top, const GSVertexSW& scan) void GSDrawScanline::DrawEdge(int pixels, int left, int top, const GSVertexSW& scan)
{ {
uint32 zwrite = m_global.sel.zwrite; u32 zwrite = m_global.sel.zwrite;
uint32 edge = m_global.sel.edge; u32 edge = m_global.sel.edge;
m_global.sel.zwrite = 0; m_global.sel.zwrite = 0;
m_global.sel.edge = 1; m_global.sel.edge = 1;
@ -2774,20 +2774,20 @@ bool GSDrawScanline::TestAlpha(T& test, T& fm, T& zm, const T& ga)
static const int s_offsets[] = {0, 2, 8, 10, 16, 18, 24, 26}; // columnTable16[0] static const int s_offsets[] = {0, 2, 8, 10, 16, 18, 24, 26}; // columnTable16[0]
template <class T> template <class T>
void GSDrawScanline::WritePixel(const T& src, int addr, int i, uint32 psm) void GSDrawScanline::WritePixel(const T& src, int addr, int i, u32 psm)
{ {
uint8* dst = (uint8*)m_global.vm + addr * 2 + s_offsets[i] * 2; u8* dst = (u8*)m_global.vm + addr * 2 + s_offsets[i] * 2;
switch (psm) switch (psm)
{ {
case 0: case 0:
*(uint32*)dst = src.u32[i]; *(u32*)dst = src.U32[i];
break; break;
case 1: case 1:
*(uint32*)dst = (src.u32[i] & 0xffffff) | (*(uint32*)dst & 0xff000000); *(u32*)dst = (src.U32[i] & 0xffffff) | (*(u32*)dst & 0xff000000);
break; break;
case 2: case 2:
*(uint16*)dst = src.u16[i * 2]; *(u16*)dst = src.u16[i * 2];
break; break;
} }
} }
@ -2801,38 +2801,38 @@ void GSDrawScanline::DrawRect(const GSVector4i& r, const GSVertexSW& v)
// FIXME: sometimes the frame and z buffer may overlap, the outcome is undefined // FIXME: sometimes the frame and z buffer may overlap, the outcome is undefined
uint32 m; u32 m;
#if _M_SSE >= 0x501 #if _M_SSE >= 0x501
m = m_global.zm; m = m_global.zm;
#else #else
m = m_global.zm.u32[0]; m = m_global.zm.U32[0];
#endif #endif
if (m != 0xffffffff) if (m != 0xffffffff)
{ {
uint32 z = v.t.u32[3]; // (uint32)v.p.z; u32 z = v.t.U32[3]; // (u32)v.p.z;
if (m_global.sel.zpsm != 2) if (m_global.sel.zpsm != 2)
{ {
if (m == 0) if (m == 0)
{ {
DrawRectT<uint32, false>(m_global.zbo, r, z, m); DrawRectT<u32, false>(m_global.zbo, r, z, m);
} }
else else
{ {
DrawRectT<uint32, true>(m_global.zbo, r, z, m); DrawRectT<u32, true>(m_global.zbo, r, z, m);
} }
} }
else else
{ {
if ((m & 0xffff) == 0) if ((m & 0xffff) == 0)
{ {
DrawRectT<uint16, false>(m_global.zbo, r, z, m); DrawRectT<u16, false>(m_global.zbo, r, z, m);
} }
else else
{ {
DrawRectT<uint16, true>(m_global.zbo, r, z, m); DrawRectT<u16, true>(m_global.zbo, r, z, m);
} }
} }
} }
@ -2840,12 +2840,12 @@ void GSDrawScanline::DrawRect(const GSVector4i& r, const GSVertexSW& v)
#if _M_SSE >= 0x501 #if _M_SSE >= 0x501
m = m_global.fm; m = m_global.fm;
#else #else
m = m_global.fm.u32[0]; m = m_global.fm.U32[0];
#endif #endif
if (m != 0xffffffff) if (m != 0xffffffff)
{ {
uint32 c = (GSVector4i(v.c) >> 7).rgba32(); u32 c = (GSVector4i(v.c) >> 7).rgba32();
if (m_global.sel.fba) if (m_global.sel.fba)
{ {
@ -2856,11 +2856,11 @@ void GSDrawScanline::DrawRect(const GSVector4i& r, const GSVertexSW& v)
{ {
if (m == 0) if (m == 0)
{ {
DrawRectT<uint32, false>(m_global.fbo, r, c, m); DrawRectT<u32, false>(m_global.fbo, r, c, m);
} }
else else
{ {
DrawRectT<uint32, true>(m_global.fbo, r, c, m); DrawRectT<u32, true>(m_global.fbo, r, c, m);
} }
} }
else else
@ -2869,18 +2869,18 @@ void GSDrawScanline::DrawRect(const GSVector4i& r, const GSVertexSW& v)
if ((m & 0xffff) == 0) if ((m & 0xffff) == 0)
{ {
DrawRectT<uint16, false>(m_global.fbo, r, c, m); DrawRectT<u16, false>(m_global.fbo, r, c, m);
} }
else else
{ {
DrawRectT<uint16, true>(m_global.fbo, r, c, m); DrawRectT<u16, true>(m_global.fbo, r, c, m);
} }
} }
} }
} }
template <class T, bool masked> template <class T, bool masked>
void GSDrawScanline::DrawRectT(const GSOffset& off, const GSVector4i& r, uint32 c, uint32 m) void GSDrawScanline::DrawRectT(const GSOffset& off, const GSVector4i& r, u32 c, u32 m)
{ {
if (m == 0xffffffff) if (m == 0xffffffff)
return; return;
@ -2897,7 +2897,7 @@ void GSDrawScanline::DrawRectT(const GSOffset& off, const GSVector4i& r, uint32
#endif #endif
if (sizeof(T) == sizeof(uint16)) if (sizeof(T) == sizeof(u16))
{ {
color = color.xxzzlh(); color = color.xxzzlh();
mask = mask.xxzzlh(); mask = mask.xxzzlh();
@ -2909,7 +2909,7 @@ void GSDrawScanline::DrawRectT(const GSOffset& off, const GSVector4i& r, uint32
c = c & (~m); c = c & (~m);
if (masked) if (masked)
ASSERT(mask.u32[0] != 0); ASSERT(mask.U32[0] != 0);
GSVector4i br = r.ralign<Align_Inside>(GSVector2i(8 * 4 / sizeof(T), 8)); GSVector4i br = r.ralign<Align_Inside>(GSVector2i(8 * 4 / sizeof(T), 8));
@ -2933,7 +2933,7 @@ void GSDrawScanline::DrawRectT(const GSOffset& off, const GSVector4i& r, uint32
} }
template <class T, bool masked> template <class T, bool masked>
void GSDrawScanline::FillRect(const GSOffset& off, const GSVector4i& r, uint32 c, uint32 m) void GSDrawScanline::FillRect(const GSOffset& off, const GSVector4i& r, u32 c, u32 m)
{ {
if (r.x >= r.z) if (r.x >= r.z)
return; return;

14
pcsx2/GS/Renderers/SW/GSDrawScanline.h
View File

@ -34,14 +34,14 @@ protected:
GSScanlineGlobalData m_global; GSScanlineGlobalData m_global;
GSScanlineLocalData m_local; GSScanlineLocalData m_local;
GSCodeGeneratorFunctionMap<GSSetupPrimCodeGenerator, uint64, SetupPrimPtr> m_sp_map; GSCodeGeneratorFunctionMap<GSSetupPrimCodeGenerator, u64, SetupPrimPtr> m_sp_map;
GSCodeGeneratorFunctionMap<GSDrawScanlineCodeGenerator, uint64, DrawScanlinePtr> m_ds_map; GSCodeGeneratorFunctionMap<GSDrawScanlineCodeGenerator, u64, DrawScanlinePtr> m_ds_map;
template <class T, bool masked> template <class T, bool masked>
void DrawRectT(const GSOffset& off, const GSVector4i& r, uint32 c, uint32 m); void DrawRectT(const GSOffset& off, const GSVector4i& r, u32 c, u32 m);
template <class T, bool masked> template <class T, bool masked>
__forceinline void FillRect(const GSOffset& off, const GSVector4i& r, uint32 c, uint32 m); __forceinline void FillRect(const GSOffset& off, const GSVector4i& r, u32 c, u32 m);
#if _M_SSE >= 0x501 #if _M_SSE >= 0x501
@ -62,13 +62,13 @@ public:
// IDrawScanline // IDrawScanline
void BeginDraw(const GSRasterizerData* data); void BeginDraw(const GSRasterizerData* data);
void EndDraw(uint64 frame, uint64 ticks, int actual, int total, int prims); void EndDraw(u64 frame, u64 ticks, int actual, int total, int prims);
void DrawRect(const GSVector4i& r, const GSVertexSW& v); void DrawRect(const GSVector4i& r, const GSVertexSW& v);
#ifndef ENABLE_JIT_RASTERIZER #ifndef ENABLE_JIT_RASTERIZER
void SetupPrim(const GSVertexSW* vertex, const uint32* index, const GSVertexSW& dscan); void SetupPrim(const GSVertexSW* vertex, const u32* index, const GSVertexSW& dscan);
void DrawScanline(int pixels, int left, int top, const GSVertexSW& scan); void DrawScanline(int pixels, int left, int top, const GSVertexSW& scan);
void DrawEdge(int pixels, int left, int top, const GSVertexSW& scan); void DrawEdge(int pixels, int left, int top, const GSVertexSW& scan);
@ -76,7 +76,7 @@ public:
bool IsRect() const { return m_global.sel.IsSolidRect(); } bool IsRect() const { return m_global.sel.IsSolidRect(); }
template<class T> bool TestAlpha(T& test, T& fm, T& zm, const T& ga); template<class T> bool TestAlpha(T& test, T& fm, T& zm, const T& ga);
template<class T> void WritePixel(const T& src, int addr, int i, uint32 psm); template<class T> void WritePixel(const T& src, int addr, int i, u32 psm);
#endif #endif

54
pcsx2/GS/Renderers/SW/GSDrawScanlineCodeGenerator.all.cpp
View File

@ -100,7 +100,7 @@ using namespace Xbyak;
#define _rip_local_d_p(x) _rip_local_d(x) #define _rip_local_d_p(x) _rip_local_d(x)
#endif #endif
GSDrawScanlineCodeGenerator2::GSDrawScanlineCodeGenerator2(Xbyak::CodeGenerator* base, CPUInfo cpu, void* param, uint64 key) GSDrawScanlineCodeGenerator2::GSDrawScanlineCodeGenerator2(Xbyak::CodeGenerator* base, CPUInfo cpu, void* param, u64 key)
: _parent(base, cpu) : _parent(base, cpu)
, m_local(*(GSScanlineLocalData*)param) , m_local(*(GSScanlineLocalData*)param)
, m_rip(false) , m_rip(false)
@ -201,7 +201,7 @@ void GSDrawScanlineCodeGenerator2::broadcastGPRToVec(const XYm& vec, const Xbyak
#endif #endif
} }
void GSDrawScanlineCodeGenerator2::modulate16(const XYm& a, const Operand& f, uint8 shift) void GSDrawScanlineCodeGenerator2::modulate16(const XYm& a, const Operand& f, u8 shift)
{ {
if (shift == 0) if (shift == 0)
{ {
@ -214,7 +214,7 @@ void GSDrawScanlineCodeGenerator2::modulate16(const XYm& a, const Operand& f, ui
} }
} }
void GSDrawScanlineCodeGenerator2::lerp16(const XYm& a, const XYm& b, const XYm& f, uint8 shift) void GSDrawScanlineCodeGenerator2::lerp16(const XYm& a, const XYm& b, const XYm& f, u8 shift)
{ {
psubw(a, b); psubw(a, b);
modulate16(a, f, shift); modulate16(a, f, shift);
@ -251,7 +251,7 @@ void GSDrawScanlineCodeGenerator2::clamp16(const XYm& a, const XYm& temp)
void GSDrawScanlineCodeGenerator2::alltrue(const XYm& test) void GSDrawScanlineCodeGenerator2::alltrue(const XYm& test)
{ {
uint32 mask = test.isYMM() ? 0xffffffff : 0xffff; u32 mask = test.isYMM() ? 0xffffffff : 0xffff;
pmovmskb(eax, test); pmovmskb(eax, test);
cmp(eax, mask); cmp(eax, mask);
je("step", GSCodeGenerator::T_NEAR); je("step", GSCodeGenerator::T_NEAR);
@ -1215,7 +1215,7 @@ void GSDrawScanlineCodeGenerator2::TestZ(const XYm& temp1, const XYm& temp2)
if (m_sel.zclamp) if (m_sel.zclamp)
{ {
const uint8 amt = (uint8)((m_sel.zpsm & 0x3) * 8); const u8 amt = (u8)((m_sel.zpsm & 0x3) * 8);
pcmpeqd(temp1, temp1); pcmpeqd(temp1, temp1);
psrld(temp1, amt); psrld(temp1, amt);
pminsd(xym0, temp1); pminsd(xym0, temp1);
@ -1244,8 +1244,8 @@ void GSDrawScanlineCodeGenerator2::TestZ(const XYm& temp1, const XYm& temp2)
if (m_sel.zpsm) if (m_sel.zpsm)
{ {
pslld(temp2, static_cast<uint8>(m_sel.zpsm * 8)); pslld(temp2, static_cast<u8>(m_sel.zpsm * 8));
psrld(temp2, static_cast<uint8>(m_sel.zpsm * 8)); psrld(temp2, static_cast<u8>(m_sel.zpsm * 8));
} }
if (m_sel.zoverflow || m_sel.zpsm == 0) if (m_sel.zoverflow || m_sel.zpsm == 0)
@ -1415,7 +1415,7 @@ void GSDrawScanlineCodeGenerator2::SampleTexture_TexelReadHelper(int mip_offset)
THREEARG(punpcklwd, xym5, xym2, xym0); THREEARG(punpcklwd, xym5, xym2, xym0);
punpckhwd(xym2, xym0); punpckhwd(xym2, xym0);
pslld(xym2, static_cast<uint8>(m_sel.tw + 3)); pslld(xym2, static_cast<u8>(m_sel.tw + 3));
// xym0 = 0 // xym0 = 0
// xym2 = y0 // xym2 = y0
@ -1432,7 +1432,7 @@ void GSDrawScanlineCodeGenerator2::SampleTexture_TexelReadHelper(int mip_offset)
THREEARG(punpcklwd, xym1, xym3, xym0); THREEARG(punpcklwd, xym1, xym3, xym0);
punpckhwd(xym3, xym0); punpckhwd(xym3, xym0);
pslld(xym3, static_cast<uint8>(m_sel.tw + 3)); pslld(xym3, static_cast<u8>(m_sel.tw + 3));
// xym1 = x1 // xym1 = x1
// xym2 = y0 // xym2 = y0
@ -1460,10 +1460,10 @@ void GSDrawScanlineCodeGenerator2::SampleTexture_TexelReadHelper(int mip_offset)
// xym7 = used[x86] vf[x64&&!needsMoreRegs] // xym7 = used[x86] vf[x64&&!needsMoreRegs]
// Free: xym4, xym5 // Free: xym4, xym5
// c00 = addr00.gather32_32((const uint32/uint8*)tex[, clut]); // c00 = addr00.gather32_32((const u32/u8*)tex[, clut]);
// c01 = addr01.gather32_32((const uint32/uint8*)tex[, clut]); // c01 = addr01.gather32_32((const u32/u8*)tex[, clut]);
// c10 = addr10.gather32_32((const uint32/uint8*)tex[, clut]); // c10 = addr10.gather32_32((const u32/u8*)tex[, clut]);
// c11 = addr11.gather32_32((const uint32/uint8*)tex[, clut]); // c11 = addr11.gather32_32((const u32/u8*)tex[, clut]);
const XYm& tmp1 = is64 ? xym7 : xym4; // OK to destroy if needsMoreRegs const XYm& tmp1 = is64 ? xym7 : xym4; // OK to destroy if needsMoreRegs
const XYm& tmp2 = is64 ? xym4 : xym7; const XYm& tmp2 = is64 ? xym4 : xym7;
@ -1559,7 +1559,7 @@ void GSDrawScanlineCodeGenerator2::SampleTexture_TexelReadHelper(int mip_offset)
paddd(xym2, xym5); paddd(xym2, xym5);
// c00 = addr00.gather32_32((const uint32/uint8*)tex[, clut]); // c00 = addr00.gather32_32((const u32/u8*)tex[, clut]);
ReadTexel1(xym5, xym2, xym0, xym1, mip_offset); ReadTexel1(xym5, xym2, xym0, xym1, mip_offset);
@ -1875,25 +1875,25 @@ void GSDrawScanlineCodeGenerator2::SampleTextureLOD()
movdqa(xym5, xym2); movdqa(xym5, xym2);
movdqa(xym3, xym6); movdqa(xym3, xym6);
movd(xym0, _rip_local(temp.lod.i.u32[0])); movd(xym0, _rip_local(temp.lod.i.U32[0]));
psrad(xym2, xym0); psrad(xym2, xym0);
THREEARG(psrlw, xym1, xym4, xym0); THREEARG(psrlw, xym1, xym4, xym0);
movq(_rip_local(temp.uv_minmax[0].u32[0]), xym1); movq(_rip_local(temp.uv_minmax[0].U32[0]), xym1);
movd(xym0, _rip_local(temp.lod.i.u32[1])); movd(xym0, _rip_local(temp.lod.i.U32[1]));
psrad(xym5, xym0); psrad(xym5, xym0);
THREEARG(psrlw, xym1, xym4, xym0); THREEARG(psrlw, xym1, xym4, xym0);
movq(_rip_local(temp.uv_minmax[1].u32[0]), xym1); movq(_rip_local(temp.uv_minmax[1].U32[0]), xym1);
movd(xym0, _rip_local(temp.lod.i.u32[2])); movd(xym0, _rip_local(temp.lod.i.U32[2]));
psrad(xym3, xym0); psrad(xym3, xym0);
THREEARG(psrlw, xym1, xym4, xym0); THREEARG(psrlw, xym1, xym4, xym0);
movq(_rip_local(temp.uv_minmax[0].u32[2]), xym1); movq(_rip_local(temp.uv_minmax[0].U32[2]), xym1);
movd(xym0, _rip_local(temp.lod.i.u32[3])); movd(xym0, _rip_local(temp.lod.i.U32[3]));
psrad(xym6, xym0); psrad(xym6, xym0);
THREEARG(psrlw, xym1, xym4, xym0); THREEARG(psrlw, xym1, xym4, xym0);
movq(_rip_local(temp.uv_minmax[1].u32[2]), xym1); movq(_rip_local(temp.uv_minmax[1].U32[2]), xym1);
punpckldq(xym2, xym3); punpckldq(xym2, xym3);
punpckhdq(xym5, xym6); punpckhdq(xym5, xym6);
@ -1931,7 +1931,7 @@ void GSDrawScanlineCodeGenerator2::SampleTextureLOD()
{ {
// lod = K // lod = K
movd(Xmm(xym0.getIdx()), _rip_global(lod.i.u32[0])); movd(Xmm(xym0.getIdx()), _rip_global(lod.i.U32[0]));
psrad(xym2, Xmm(xym0.getIdx())); psrad(xym2, Xmm(xym0.getIdx()));
psrad(xym3, Xmm(xym0.getIdx())); psrad(xym3, Xmm(xym0.getIdx()));
@ -2743,7 +2743,7 @@ void GSDrawScanlineCodeGenerator2::WriteZBuf()
// Clamp Z to ZPSM_FMT_MAX // Clamp Z to ZPSM_FMT_MAX
if (m_sel.zclamp) if (m_sel.zclamp)
{ {
const uint8 amt = (uint8)((m_sel.zpsm & 0x3) * 8); const u8 amt = (u8)((m_sel.zpsm & 0x3) * 8);
pcmpeqd(xym7, xym7); pcmpeqd(xym7, xym7);
psrld(xym7, amt); psrld(xym7, amt);
pminsd(xym1, xym7); pminsd(xym1, xym7);
@ -3273,7 +3273,7 @@ void GSDrawScanlineCodeGenerator2::WritePixel(const XYm& src_, const AddressReg&
} }
} }
void GSDrawScanlineCodeGenerator2::WritePixel(const Xmm& src, const AddressReg& addr, uint8 i, uint8 j, int psm) void GSDrawScanlineCodeGenerator2::WritePixel(const Xmm& src, const AddressReg& addr, u8 i, u8 j, int psm)
{ {
constexpr int s_offsets[8] = {0, 2, 8, 10, 16, 18, 24, 26}; constexpr int s_offsets[8] = {0, 2, 8, 10, 16, 18, 24, 26};
@ -3356,7 +3356,7 @@ void GSDrawScanlineCodeGenerator2::ReadTexelImpl(
void GSDrawScanlineCodeGenerator2::ReadTexelImplLoadTexLOD(int lod, int mip_offset) void GSDrawScanlineCodeGenerator2::ReadTexelImplLoadTexLOD(int lod, int mip_offset)
{ {
AddressReg texIn = is64 ? _64_m_local__gd__tex : t2; AddressReg texIn = is64 ? _64_m_local__gd__tex : t2;
Address lod_addr = m_sel.lcm ? _rip_global(lod.i.u32[lod]) : _rip_local(temp.lod.i.u32[lod]); Address lod_addr = m_sel.lcm ? _rip_global(lod.i.U32[lod]) : _rip_local(temp.lod.i.U32[lod]);
mov(ebx, lod_addr); mov(ebx, lod_addr);
mov(rbx, ptr[texIn + rbx * wordsize + mip_offset]); mov(rbx, ptr[texIn + rbx * wordsize + mip_offset]);
} }
@ -3483,7 +3483,7 @@ void GSDrawScanlineCodeGenerator2::ReadTexelImplSSE4(
} }
} }
void GSDrawScanlineCodeGenerator2::ReadTexelImpl(const Xmm& dst, const Xmm& addr, uint8 i, bool texInRBX, bool preserveDst) void GSDrawScanlineCodeGenerator2::ReadTexelImpl(const Xmm& dst, const Xmm& addr, u8 i, bool texInRBX, bool preserveDst)
{ {
ASSERT(i < 4); ASSERT(i < 4);

10
pcsx2/GS/Renderers/SW/GSDrawScanlineCodeGenerator.all.h
View File

@ -92,7 +92,7 @@ class GSDrawScanlineCodeGenerator2 : public GSNewCodeGenerator
} }
public: public:
GSDrawScanlineCodeGenerator2(Xbyak::CodeGenerator* base, CPUInfo cpu, void* param, uint64 key); GSDrawScanlineCodeGenerator2(Xbyak::CodeGenerator* base, CPUInfo cpu, void* param, u64 key);
void Generate(); void Generate();
private: private:
@ -120,8 +120,8 @@ private:
void pbroadcastwLocal(const XYm& reg, const Xbyak::Address& mem); void pbroadcastwLocal(const XYm& reg, const Xbyak::Address& mem);
/// Broadcast a 32-bit GPR to a vector register /// Broadcast a 32-bit GPR to a vector register
void broadcastGPRToVec(const XYm& vec, const Xbyak::Reg32& gpr); void broadcastGPRToVec(const XYm& vec, const Xbyak::Reg32& gpr);
void modulate16(const XYm& a, const Xbyak::Operand& f, uint8 shift); void modulate16(const XYm& a, const Xbyak::Operand& f, u8 shift);
void lerp16(const XYm& a, const XYm& b, const XYm& f, uint8 shift); void lerp16(const XYm& a, const XYm& b, const XYm& f, u8 shift);
void lerp16_4(const XYm& a, const XYm& b, const XYm& f); void lerp16_4(const XYm& a, const XYm& b, const XYm& f);
void mix16(const XYm& a, const XYm& b, const XYm& temp); void mix16(const XYm& a, const XYm& b, const XYm& temp);
void clamp16(const XYm& a, const XYm& temp); void clamp16(const XYm& a, const XYm& temp);
@ -159,7 +159,7 @@ private:
#else #else
void WritePixel(const XYm& src_, const AddressReg& addr, const Xbyak::Reg32& mask, bool fast, int psm, int fz); void WritePixel(const XYm& src_, const AddressReg& addr, const Xbyak::Reg32& mask, bool fast, int psm, int fz);
#endif #endif
void WritePixel(const Xmm& src, const AddressReg& addr, uint8 i, uint8 j, int psm); void WritePixel(const Xmm& src, const AddressReg& addr, u8 i, u8 j, int psm);
void ReadTexel1(const XYm& dst, const XYm& src, const XYm& tmp1, const XYm& tmp2, int mip_offset); void ReadTexel1(const XYm& dst, const XYm& src, const XYm& tmp1, const XYm& tmp2, int mip_offset);
void ReadTexel4( void ReadTexel4(
const XYm& d0, const XYm& d1, const XYm& d0, const XYm& d1,
@ -185,5 +185,5 @@ private:
const Xmm& d2s0, const Xmm& d3s1, const Xmm& d2s0, const Xmm& d3s1,
const Xmm& s2, const Xmm& s3, const Xmm& s2, const Xmm& s3,
int pixels, int mip_offset); int pixels, int mip_offset);
void ReadTexelImpl(const Xmm& dst, const Xmm& addr, uint8 i, bool texInA3, bool preserveDst); void ReadTexelImpl(const Xmm& dst, const Xmm& addr, u8 i, bool texInA3, bool preserveDst);
}; };

2
pcsx2/GS/Renderers/SW/GSDrawScanlineCodeGenerator.cpp
View File

@ -18,7 +18,7 @@
#include "GSDrawScanlineCodeGenerator.all.h" #include "GSDrawScanlineCodeGenerator.all.h"
GSDrawScanlineCodeGenerator::GSDrawScanlineCodeGenerator(void* param, uint64 key, void* code, size_t maxsize) GSDrawScanlineCodeGenerator::GSDrawScanlineCodeGenerator(void* param, u64 key, void* code, size_t maxsize)
: GSCodeGenerator(code, maxsize) : GSCodeGenerator(code, maxsize)
, m_local(*(GSScanlineLocalData*)param) , m_local(*(GSScanlineLocalData*)param)
, m_rip(false) , m_rip(false)

2
pcsx2/GS/Renderers/SW/GSDrawScanlineCodeGenerator.h
View File

@ -34,5 +34,5 @@ class GSDrawScanlineCodeGenerator : public GSCodeGenerator
bool m_rip; bool m_rip;
public: public:
GSDrawScanlineCodeGenerator(void* param, uint64 key, void* code, size_t maxsize); GSDrawScanlineCodeGenerator(void* param, u64 key, void* code, size_t maxsize);
}; };

20
pcsx2/GS/Renderers/SW/GSNewCodeGenerator.h
View File

@ -89,7 +89,7 @@ public:
"used SSE instruction in AVX code", "used SSE instruction in AVX code",
"used AVX instruction in SSE code", "used AVX instruction in SSE code",
}; };
if (static_cast<uint32>(value) < (sizeof(tbl) / sizeof(*tbl))) if (static_cast<u32>(value) < (sizeof(tbl) / sizeof(*tbl)))
{ {
return tbl[value]; return tbl[value];
} }
@ -320,11 +320,11 @@ public:
FORWARD(2, BASE, name, ARGS_OO) \ FORWARD(2, BASE, name, ARGS_OO) \
FORWARD(2, BASE, name, ARGS_OI) FORWARD(2, BASE, name, ARGS_OI)
#define ARGS_OI const Operand&, uint32 #define ARGS_OI const Operand&, u32
#define ARGS_OO const Operand&, const Operand& #define ARGS_OO const Operand&, const Operand&
#define ARGS_XI const Xmm&, int #define ARGS_XI const Xmm&, int
#define ARGS_XO const Xmm&, const Operand& #define ARGS_XO const Xmm&, const Operand&
#define ARGS_XOI const Xmm&, const Operand&, uint8 #define ARGS_XOI const Xmm&, const Operand&, u8
#define ARGS_XXO const Xmm&, const Xmm&, const Operand& #define ARGS_XXO const Xmm&, const Xmm&, const Operand&
// For instructions that are ifdef'd out without XBYAK64 // For instructions that are ifdef'd out without XBYAK64
@ -334,7 +334,7 @@ public:
#define REQUIRE64(action) require64() #define REQUIRE64(action) require64()
#endif #endif
const uint8 *getCurr() { return actual.getCurr(); } const u8 *getCurr() { return actual.getCurr(); }
void align(int x = 16) { return actual.align(x); } void align(int x = 16) { return actual.align(x); }
void db(int code) { actual.db(code); } void db(int code) { actual.db(code); }
void L(const std::string& label) { actual.L(label); } void L(const std::string& label) { actual.L(label); }
@ -374,7 +374,7 @@ public:
SFORWARD(2, cvtdq2ps, ARGS_XO) SFORWARD(2, cvtdq2ps, ARGS_XO)
SFORWARD(2, cvtps2dq, ARGS_XO) SFORWARD(2, cvtps2dq, ARGS_XO)
SFORWARD(2, cvttps2dq, ARGS_XO) SFORWARD(2, cvttps2dq, ARGS_XO)
SFORWARD(3, extractps, const Operand&, const Xmm&, uint8) SFORWARD(3, extractps, const Operand&, const Xmm&, u8)
AFORWARD(2, maxps, ARGS_XO) AFORWARD(2, maxps, ARGS_XO)
AFORWARD(2, minps, ARGS_XO) AFORWARD(2, minps, ARGS_XO)
SFORWARD(2, movaps, ARGS_XO) SFORWARD(2, movaps, ARGS_XO)
@ -403,8 +403,8 @@ public:
AFORWARD(2, pcmpeqd, ARGS_XO) AFORWARD(2, pcmpeqd, ARGS_XO)
AFORWARD(2, pcmpeqw, ARGS_XO) AFORWARD(2, pcmpeqw, ARGS_XO)
AFORWARD(2, pcmpgtd, ARGS_XO) AFORWARD(2, pcmpgtd, ARGS_XO)
SFORWARD(3, pextrd, const Operand&, const Xmm&, uint8) SFORWARD(3, pextrd, const Operand&, const Xmm&, u8)
SFORWARD(3, pextrw, const Operand&, const Xmm&, uint8) SFORWARD(3, pextrw, const Operand&, const Xmm&, u8)
AFORWARD(3, pinsrd, ARGS_XOI) AFORWARD(3, pinsrd, ARGS_XOI)
AFORWARD(2, pmaxsw, ARGS_XO) AFORWARD(2, pmaxsw, ARGS_XO)
AFORWARD(2, pminsd, ARGS_XO) AFORWARD(2, pminsd, ARGS_XO)
@ -448,12 +448,12 @@ public:
FORWARD(2, AVX2, vbroadcasti128, const Ymm&, const Address&) FORWARD(2, AVX2, vbroadcasti128, const Ymm&, const Address&)
FORWARD(2, AVX, vbroadcastf128, const Ymm&, const Address&) FORWARD(2, AVX, vbroadcastf128, const Ymm&, const Address&)
FORWARD(3, FMA, vfmadd213ps, ARGS_XXO) FORWARD(3, FMA, vfmadd213ps, ARGS_XXO)
FORWARD(3, AVX2, vextracti128, const Operand&, const Ymm&, uint8) FORWARD(3, AVX2, vextracti128, const Operand&, const Ymm&, u8)
FORWARD(4, AVX2, vinserti128, const Ymm&, const Ymm&, const Operand&, uint8); FORWARD(4, AVX2, vinserti128, const Ymm&, const Ymm&, const Operand&, u8);
FORWARD(2, AVX2, vpbroadcastd, ARGS_XO) FORWARD(2, AVX2, vpbroadcastd, ARGS_XO)
FORWARD(2, AVX2, vpbroadcastq, ARGS_XO) FORWARD(2, AVX2, vpbroadcastq, ARGS_XO)
FORWARD(2, AVX2, vpbroadcastw, ARGS_XO) FORWARD(2, AVX2, vpbroadcastw, ARGS_XO)
FORWARD(3, AVX2, vpermq, const Ymm&, const Operand&, uint8) FORWARD(3, AVX2, vpermq, const Ymm&, const Operand&, u8)
FORWARD(3, AVX2, vpgatherdd, const Xmm&, const Address&, const Xmm&); FORWARD(3, AVX2, vpgatherdd, const Xmm&, const Address&, const Xmm&);
FORWARD(3, AVX2, vpsravd, ARGS_XXO) FORWARD(3, AVX2, vpsravd, ARGS_XXO)
FORWARD(3, AVX2, vpsrlvd, ARGS_XXO) FORWARD(3, AVX2, vpsrlvd, ARGS_XXO)

56
pcsx2/GS/Renderers/SW/GSRasterizer.cpp
View File

@ -50,7 +50,7 @@ GSRasterizer::GSRasterizer(IDrawScanline* ds, int id, int threads, GSPerfMon* pe
m_edge.count = 0; m_edge.count = 0;
int rows = (2048 >> m_thread_height) + 16; int rows = (2048 >> m_thread_height) + 16;
m_scanline = (uint8*)_aligned_malloc(rows, 64); m_scanline = (u8*)_aligned_malloc(rows, 64);
int row = 0; int row = 0;
@ -148,10 +148,10 @@ void GSRasterizer::Draw(GSRasterizerData* data)
const GSVertexSW* vertex = data->vertex; const GSVertexSW* vertex = data->vertex;
const GSVertexSW* vertex_end = data->vertex + data->vertex_count; const GSVertexSW* vertex_end = data->vertex + data->vertex_count;
const uint32* index = data->index; const u32* index = data->index;
const uint32* index_end = data->index + data->index_count; const u32* index_end = data->index + data->index_count;
uint32 tmp_index[] = {0, 1, 2}; u32 tmp_index[] = {0, 1, 2};
bool scissor_test = !data->bbox.eq(data->bbox.rintersect(data->scissor)); bool scissor_test = !data->bbox.eq(data->bbox.rintersect(data->scissor));
@ -247,7 +247,7 @@ void GSRasterizer::Draw(GSRasterizerData* data)
data->pixels = m_pixels.actual; data->pixels = m_pixels.actual;
uint64 ticks = __rdtsc() - data->start; u64 ticks = __rdtsc() - data->start;
m_pixels.sum += m_pixels.actual; m_pixels.sum += m_pixels.actual;
@ -255,7 +255,7 @@ void GSRasterizer::Draw(GSRasterizerData* data)
} }
template <bool scissor_test> template <bool scissor_test>
void GSRasterizer::DrawPoint(const GSVertexSW* vertex, int vertex_count, const uint32* index, int index_count) void GSRasterizer::DrawPoint(const GSVertexSW* vertex, int vertex_count, const u32* index, int index_count)
{ {
m_primcount++; m_primcount++;
@ -280,7 +280,7 @@ void GSRasterizer::DrawPoint(const GSVertexSW* vertex, int vertex_count, const u
} }
else else
{ {
uint32 tmp_index[1] = {0}; u32 tmp_index[1] = {0};
for (int i = 0; i < vertex_count; i++, vertex++) for (int i = 0; i < vertex_count; i++, vertex++)
{ {
@ -301,7 +301,7 @@ void GSRasterizer::DrawPoint(const GSVertexSW* vertex, int vertex_count, const u
} }
} }
void GSRasterizer::DrawLine(const GSVertexSW* vertex, const uint32* index) void GSRasterizer::DrawLine(const GSVertexSW* vertex, const u32* index)
{ {
m_primcount++; m_primcount++;
@ -405,7 +405,7 @@ void GSRasterizer::DrawLine(const GSVertexSW* vertex, const uint32* index)
} }
} }
static const uint8 s_ysort[8][4] = static const u8 s_ysort[8][4] =
{ {
{0, 1, 2, 0}, // y0 <= y1 <= y2 {0, 1, 2, 0}, // y0 <= y1 <= y2
{1, 0, 2, 0}, // y1 < y0 <= y2 {1, 0, 2, 0}, // y1 < y0 <= y2
@ -419,7 +419,7 @@ static const uint8 s_ysort[8][4] =
#if _M_SSE >= 0x501 #if _M_SSE >= 0x501
void GSRasterizer::DrawTriangle(const GSVertexSW* vertex, const uint32* index) void GSRasterizer::DrawTriangle(const GSVertexSW* vertex, const u32* index)
{ {
m_primcount++; m_primcount++;
@ -612,7 +612,7 @@ void GSRasterizer::DrawTriangleSection(int top, int bottom, GSVertexSW2& edge, c
#else #else
void GSRasterizer::DrawTriangle(const GSVertexSW* vertex, const uint32* index) void GSRasterizer::DrawTriangle(const GSVertexSW* vertex, const u32* index)
{ {
m_primcount++; m_primcount++;
@ -807,7 +807,7 @@ void GSRasterizer::DrawTriangleSection(int top, int bottom, GSVertexSW& edge, co
#endif #endif
void GSRasterizer::DrawSprite(const GSVertexSW* vertex, const uint32* index) void GSRasterizer::DrawSprite(const GSVertexSW* vertex, const u32* index)
{ {
m_primcount++; m_primcount++;
@ -971,7 +971,7 @@ void GSRasterizer::DrawEdge(const GSVertexSW& v0, const GSVertexSW& v1, const GS
{ {
AddScanline(e, 1, xi, top, edge); AddScanline(e, 1, xi, top, edge);
e->t.u32[3] = (0x10000 - xf) & 0xffff; e->t.U32[3] = (0x10000 - xf) & 0xffff;
e++; e++;
} }
@ -994,7 +994,7 @@ void GSRasterizer::DrawEdge(const GSVertexSW& v0, const GSVertexSW& v1, const GS
{ {
AddScanline(e, 1, xi, top, edge); AddScanline(e, 1, xi, top, edge);
e->t.u32[3] = xf; e->t.U32[3] = xf;
e++; e++;
} }
@ -1061,7 +1061,7 @@ void GSRasterizer::DrawEdge(const GSVertexSW& v0, const GSVertexSW& v1, const GS
{ {
AddScanline(e, 1, left, yi, edge); AddScanline(e, 1, left, yi, edge);
e->t.u32[3] = (0x10000 - yf) & 0xffff; e->t.U32[3] = (0x10000 - yf) & 0xffff;
e++; e++;
} }
@ -1084,7 +1084,7 @@ void GSRasterizer::DrawEdge(const GSVertexSW& v0, const GSVertexSW& v1, const GS
{ {
AddScanline(e, 1, left, yi, edge); AddScanline(e, 1, left, yi, edge);
e->t.u32[3] = yf; e->t.U32[3] = yf;
e++; e++;
} }
@ -1105,12 +1105,12 @@ void GSRasterizer::AddScanline(GSVertexSW* e, int pixels, int left, int top, con
{ {
*e = scan; *e = scan;
e->_pad.i32[0] = pixels; e->_pad.I32[0] = pixels;
e->_pad.i32[1] = left; e->_pad.I32[1] = left;
e->_pad.i32[2] = top; e->_pad.I32[2] = top;
} }
void GSRasterizer::Flush(const GSVertexSW* vertex, const uint32* index, const GSVertexSW& dscan, bool edge) void GSRasterizer::Flush(const GSVertexSW* vertex, const u32* index, const GSVertexSW& dscan, bool edge)
{ {
// TODO: on win64 this could be the place where xmm6-15 are preserved (not by each DrawScanline) // TODO: on win64 this could be the place where xmm6-15 are preserved (not by each DrawScanline)
@ -1127,9 +1127,9 @@ void GSRasterizer::Flush(const GSVertexSW* vertex, const uint32* index, const GS
{ {
do do
{ {
int pixels = e->_pad.i32[0]; int pixels = e->_pad.I32[0];
int left = e->_pad.i32[1]; int left = e->_pad.I32[1];
int top = e->_pad.i32[2]; int top = e->_pad.I32[2];
DrawScanline(pixels, left, top, *e++); DrawScanline(pixels, left, top, *e++);
} while (e < ee); } while (e < ee);
@ -1138,9 +1138,9 @@ void GSRasterizer::Flush(const GSVertexSW* vertex, const uint32* index, const GS
{ {
do do
{ {
int pixels = e->_pad.i32[0]; int pixels = e->_pad.I32[0];
int left = e->_pad.i32[1]; int left = e->_pad.I32[1];
int top = e->_pad.i32[2]; int top = e->_pad.I32[2];
DrawEdge(pixels, left, top, *e++); DrawEdge(pixels, left, top, *e++);
} while (e < ee); } while (e < ee);
@ -1185,7 +1185,7 @@ GSRasterizerList::GSRasterizerList(int threads, GSPerfMon* perfmon)
m_thread_height = compute_best_thread_height(threads); m_thread_height = compute_best_thread_height(threads);
int rows = (2048 >> m_thread_height) + 16; int rows = (2048 >> m_thread_height) + 16;
m_scanline = (uint8*)_aligned_malloc(rows, 64); m_scanline = (u8*)_aligned_malloc(rows, 64);
int row = 0; int row = 0;
@ -1193,7 +1193,7 @@ GSRasterizerList::GSRasterizerList(int threads, GSPerfMon* perfmon)
{ {
for (int i = 0; i < threads; i++, row++) for (int i = 0; i < threads; i++, row++)
{ {
m_scanline[row] = (uint8)i; m_scanline[row] = (u8)i;
} }
} }
} }

30
pcsx2/GS/Renderers/SW/GSRasterizer.h
View File

@ -30,13 +30,13 @@ public:
GSVector4i scissor; GSVector4i scissor;
GSVector4i bbox; GSVector4i bbox;
GS_PRIM_CLASS primclass; GS_PRIM_CLASS primclass;
uint8* buff; u8* buff;
GSVertexSW* vertex; GSVertexSW* vertex;
int vertex_count; int vertex_count;
uint32* index; u32* index;
int index_count; int index_count;
uint64 frame; u64 frame;
uint64 start; u64 start;
int pixels; int pixels;
int counter; int counter;
@ -66,7 +66,7 @@ public:
class IDrawScanline : public GSAlignedClass<32> class IDrawScanline : public GSAlignedClass<32>
{ {
public: public:
typedef void (*SetupPrimPtr)(const GSVertexSW* vertex, const uint32* index, const GSVertexSW& dscan); typedef void (*SetupPrimPtr)(const GSVertexSW* vertex, const u32* index, const GSVertexSW& dscan);
typedef void(__fastcall* DrawScanlinePtr)(int pixels, int left, int top, const GSVertexSW& scan); typedef void(__fastcall* DrawScanlinePtr)(int pixels, int left, int top, const GSVertexSW& scan);
typedef void (IDrawScanline::*DrawRectPtr)(const GSVector4i& r, const GSVertexSW& v); // TODO: jit typedef void (IDrawScanline::*DrawRectPtr)(const GSVector4i& r, const GSVertexSW& v); // TODO: jit
@ -87,18 +87,18 @@ public:
virtual ~IDrawScanline() {} virtual ~IDrawScanline() {}
virtual void BeginDraw(const GSRasterizerData* data) = 0; virtual void BeginDraw(const GSRasterizerData* data) = 0;
virtual void EndDraw(uint64 frame, uint64 ticks, int actual, int total, int prims) = 0; virtual void EndDraw(u64 frame, u64 ticks, int actual, int total, int prims) = 0;
#ifdef ENABLE_JIT_RASTERIZER #ifdef ENABLE_JIT_RASTERIZER
__forceinline void SetupPrim(const GSVertexSW* vertex, const uint32* index, const GSVertexSW& dscan) { m_sp(vertex, index, dscan); } __forceinline void SetupPrim(const GSVertexSW* vertex, const u32* index, const GSVertexSW& dscan) { m_sp(vertex, index, dscan); }
__forceinline void DrawScanline(int pixels, int left, int top, const GSVertexSW& scan) { m_ds(pixels, left, top, scan); } __forceinline void DrawScanline(int pixels, int left, int top, const GSVertexSW& scan) { m_ds(pixels, left, top, scan); }
__forceinline void DrawEdge(int pixels, int left, int top, const GSVertexSW& scan) { m_de(pixels, left, top, scan); } __forceinline void DrawEdge(int pixels, int left, int top, const GSVertexSW& scan) { m_de(pixels, left, top, scan); }
__forceinline void DrawRect(const GSVector4i& r, const GSVertexSW& v) { (this->*m_dr)(r, v); } __forceinline void DrawRect(const GSVector4i& r, const GSVertexSW& v) { (this->*m_dr)(r, v); }
#else #else
virtual void SetupPrim(const GSVertexSW* vertex, const uint32* index, const GSVertexSW& dscan) = 0; virtual void SetupPrim(const GSVertexSW* vertex, const u32* index, const GSVertexSW& dscan) = 0;
virtual void DrawScanline(int pixels, int left, int top, const GSVertexSW& scan) = 0; virtual void DrawScanline(int pixels, int left, int top, const GSVertexSW& scan) = 0;
virtual void DrawEdge(int pixels, int left, int top, const GSVertexSW& scan) = 0; virtual void DrawEdge(int pixels, int left, int top, const GSVertexSW& scan) = 0;
virtual void DrawRect(const GSVector4i& r, const GSVertexSW& v) = 0; virtual void DrawRect(const GSVector4i& r, const GSVertexSW& v) = 0;
@ -131,7 +131,7 @@ protected:
int m_id; int m_id;
int m_threads; int m_threads;
int m_thread_height; int m_thread_height;
uint8* m_scanline; u8* m_scanline;
GSVector4i m_scissor; GSVector4i m_scissor;
GSVector4 m_fscissor_x; GSVector4 m_fscissor_x;
GSVector4 m_fscissor_y; GSVector4 m_fscissor_y;
@ -142,10 +142,10 @@ protected:
typedef void (GSRasterizer::*DrawPrimPtr)(const GSVertexSW* v, int count); typedef void (GSRasterizer::*DrawPrimPtr)(const GSVertexSW* v, int count);
template <bool scissor_test> template <bool scissor_test>
void DrawPoint(const GSVertexSW* vertex, int vertex_count, const uint32* index, int index_count); void DrawPoint(const GSVertexSW* vertex, int vertex_count, const u32* index, int index_count);
void DrawLine(const GSVertexSW* vertex, const uint32* index); void DrawLine(const GSVertexSW* vertex, const u32* index);
void DrawTriangle(const GSVertexSW* vertex, const uint32* index); void DrawTriangle(const GSVertexSW* vertex, const u32* index);
void DrawSprite(const GSVertexSW* vertex, const uint32* index); void DrawSprite(const GSVertexSW* vertex, const u32* index);
#if _M_SSE >= 0x501 #if _M_SSE >= 0x501
__forceinline void DrawTriangleSection(int top, int bottom, GSVertexSW2& edge, const GSVertexSW2& dedge, const GSVertexSW2& dscan, const GSVector4& p0); __forceinline void DrawTriangleSection(int top, int bottom, GSVertexSW2& edge, const GSVertexSW2& dedge, const GSVertexSW2& dscan, const GSVector4& p0);
@ -156,7 +156,7 @@ protected:
void DrawEdge(const GSVertexSW& v0, const GSVertexSW& v1, const GSVertexSW& dv, int orientation, int side); void DrawEdge(const GSVertexSW& v0, const GSVertexSW& v1, const GSVertexSW& dv, int orientation, int side);
__forceinline void AddScanline(GSVertexSW* e, int pixels, int left, int top, const GSVertexSW& scan); __forceinline void AddScanline(GSVertexSW* e, int pixels, int left, int top, const GSVertexSW& scan);
__forceinline void Flush(const GSVertexSW* vertex, const uint32* index, const GSVertexSW& dscan, bool edge = false); __forceinline void Flush(const GSVertexSW* vertex, const u32* index, const GSVertexSW& dscan, bool edge = false);
__forceinline void DrawScanline(int pixels, int left, int top, const GSVertexSW& scan); __forceinline void DrawScanline(int pixels, int left, int top, const GSVertexSW& scan);
__forceinline void DrawEdge(int pixels, int left, int top, const GSVertexSW& scan); __forceinline void DrawEdge(int pixels, int left, int top, const GSVertexSW& scan);
@ -189,7 +189,7 @@ protected:
// Worker threads depend on the rasterizers, so don't change the order. // Worker threads depend on the rasterizers, so don't change the order.
std::vector<std::unique_ptr<GSRasterizer>> m_r; std::vector<std::unique_ptr<GSRasterizer>> m_r;
std::vector<std::unique_ptr<GSWorker>> m_workers; std::vector<std::unique_ptr<GSWorker>> m_workers;
uint8* m_scanline; u8* m_scanline;
int m_thread_height; int m_thread_height;
GSRasterizerList(int threads, GSPerfMon* perfmon); GSRasterizerList(int threads, GSPerfMon* perfmon);

142
pcsx2/GS/Renderers/SW/GSRendererSW.cpp
View File

@ -36,7 +36,7 @@ GSRendererSW::GSRendererSW(int threads)
m_rl = GSRasterizerList::Create<GSDrawScanline>(threads, &m_perfmon); m_rl = GSRasterizerList::Create<GSDrawScanline>(threads, &m_perfmon);
m_output = (uint8*)_aligned_malloc(1024 * 1024 * sizeof(uint32), 32); m_output = (u8*)_aligned_malloc(1024 * 1024 * sizeof(u32), 32);
std::fill(std::begin(m_fzb_pages), std::end(m_fzb_pages), 0); std::fill(std::begin(m_fzb_pages), std::end(m_fzb_pages), 0);
std::fill(std::begin(m_tex_pages), std::end(m_tex_pages), 0); std::fill(std::begin(m_tex_pages), std::end(m_tex_pages), 0);
@ -194,7 +194,7 @@ GSTexture* GSRendererSW::GetFeedbackOutput()
} }
template <uint32 primclass, uint32 tme, uint32 fst, uint32 q_div> template <u32 primclass, u32 tme, u32 fst, u32 q_div>
void GSRendererSW::ConvertVertexBuffer(GSVertexSW* RESTRICT dst, const GSVertex* RESTRICT src, size_t count) void GSRendererSW::ConvertVertexBuffer(GSVertexSW* RESTRICT dst, const GSVertex* RESTRICT src, size_t count)
{ {
// FIXME q_div wasn't added to AVX2 code path. // FIXME q_div wasn't added to AVX2 code path.
@ -339,20 +339,20 @@ void GSRendererSW::Draw()
std::shared_ptr<GSRasterizerData> data(sd); std::shared_ptr<GSRasterizerData> data(sd);
sd->primclass = m_vt.m_primclass; sd->primclass = m_vt.m_primclass;
sd->buff = (uint8*)_aligned_malloc(sizeof(GSVertexSW) * ((m_vertex.next + 1) & ~1) + sizeof(uint32) * m_index.tail, 64); sd->buff = (u8*)_aligned_malloc(sizeof(GSVertexSW) * ((m_vertex.next + 1) & ~1) + sizeof(u32) * m_index.tail, 64);
sd->vertex = (GSVertexSW*)sd->buff; sd->vertex = (GSVertexSW*)sd->buff;
sd->vertex_count = m_vertex.next; sd->vertex_count = m_vertex.next;
sd->index = (uint32*)(sd->buff + sizeof(GSVertexSW) * ((m_vertex.next + 1) & ~1)); sd->index = (u32*)(sd->buff + sizeof(GSVertexSW) * ((m_vertex.next + 1) & ~1));
sd->index_count = m_index.tail; sd->index_count = m_index.tail;
// skip per pixel division if q is constant. // skip per pixel division if q is constant.
// Optimize the division by 1 with a nop. It also means that GS_SPRITE_CLASS must be processed when !m_vt.m_eq.q. // Optimize the division by 1 with a nop. It also means that GS_SPRITE_CLASS must be processed when !m_vt.m_eq.q.
// If you have both GS_SPRITE_CLASS && m_vt.m_eq.q, it will depends on the first part of the 'OR' // If you have both GS_SPRITE_CLASS && m_vt.m_eq.q, it will depends on the first part of the 'OR'
uint32 q_div = !IsMipMapActive() && ((m_vt.m_eq.q && m_vt.m_min.t.z != 1.0f) || (!m_vt.m_eq.q && m_vt.m_primclass == GS_SPRITE_CLASS)); u32 q_div = !IsMipMapActive() && ((m_vt.m_eq.q && m_vt.m_min.t.z != 1.0f) || (!m_vt.m_eq.q && m_vt.m_primclass == GS_SPRITE_CLASS));
(this->*m_cvb[m_vt.m_primclass][PRIM->TME][PRIM->FST][q_div])(sd->vertex, m_vertex.buff, m_vertex.next); (this->*m_cvb[m_vt.m_primclass][PRIM->TME][PRIM->FST][q_div])(sd->vertex, m_vertex.buff, m_vertex.next);
memcpy(sd->index, m_index.buff, sizeof(uint32) * m_index.tail); memcpy(sd->index, m_index.buff, sizeof(u32) * m_index.tail);
GSVector4i scissor = GSVector4i(context->scissor.in); GSVector4i scissor = GSVector4i(context->scissor.in);
GSVector4i bbox = GSVector4i(m_vt.m_min.p.floor().xyxy(m_vt.m_max.p.ceil())); GSVector4i bbox = GSVector4i(m_vt.m_min.p.floor().xyxy(m_vt.m_max.p.ceil()));
@ -382,7 +382,7 @@ void GSRendererSW::Draw()
{ {
int n = GSUtil::GetVertexCount(PRIM->PRIM); int n = GSUtil::GetVertexCount(PRIM->PRIM);
for (uint32 i = 0, j = 0; i < m_index.tail; i += n, j++) for (u32 i = 0, j = 0; i < m_index.tail; i += n, j++)
{ {
for (int k = 0; k < n; k++) for (int k = 0; k < n; k++)
{ {
@ -443,7 +443,7 @@ void GSRendererSW::Draw()
{ {
Sync(2); Sync(2);
uint64 frame = m_perfmon.GetFrame(); u64 frame = m_perfmon.GetFrame();
// Dump the texture in 32 bits format. It helps to debug texture shuffle effect // Dump the texture in 32 bits format. It helps to debug texture shuffle effect
// It will breaks the few games that really uses 16 bits RT // It will breaks the few games that really uses 16 bits RT
bool texture_shuffle = ((context->FRAME.PSM & 0x2) && ((context->TEX0.PSM & 3) == 2) && (m_vt.m_primclass == GS_SPRITE_CLASS)); bool texture_shuffle = ((context->FRAME.PSM & 0x2) && ((context->TEX0.PSM & 3) == 2) && (m_vt.m_primclass == GS_SPRITE_CLASS));
@ -594,7 +594,7 @@ void GSRendererSW::Sync(int reason)
GSPerfMonAutoTimer pmat(&m_perfmon, GSPerfMon::Sync); GSPerfMonAutoTimer pmat(&m_perfmon, GSPerfMon::Sync);
uint64 t = __rdtsc(); u64 t = __rdtsc();
m_rl->Sync(); m_rl->Sync();
@ -645,7 +645,7 @@ void GSRendererSW::InvalidateVideoMem(const GIFRegBITBLTBUF& BITBLTBUF, const GS
if (!m_rl->IsSynced()) if (!m_rl->IsSynced())
{ {
pages.loopPagesWithBreak([&](uint32 page) pages.loopPagesWithBreak([&](u32 page)
{ {
if (m_fzb_pages[page] | m_tex_pages[page]) if (m_fzb_pages[page] | m_tex_pages[page])
{ {
@ -673,7 +673,7 @@ void GSRendererSW::InvalidateLocalMem(const GIFRegBITBLTBUF& BITBLTBUF, const GS
GSOffset off = m_mem.GetOffset(BITBLTBUF.SBP, BITBLTBUF.SBW, BITBLTBUF.SPSM); GSOffset off = m_mem.GetOffset(BITBLTBUF.SBP, BITBLTBUF.SBW, BITBLTBUF.SPSM);
GSOffset::PageLooper pages = off.pageLooperForRect(r); GSOffset::PageLooper pages = off.pageLooperForRect(r);
pages.loopPagesWithBreak([&](uint32 page) pages.loopPagesWithBreak([&](u32 page)
{ {
if (m_fzb_pages[page]) if (m_fzb_pages[page])
{ {
@ -688,7 +688,7 @@ void GSRendererSW::InvalidateLocalMem(const GIFRegBITBLTBUF& BITBLTBUF, const GS
void GSRendererSW::UsePages(const GSOffset::PageLooper& pages, const int type) void GSRendererSW::UsePages(const GSOffset::PageLooper& pages, const int type)
{ {
pages.loopPages([=](uint32 page) pages.loopPages([=](u32 page)
{ {
switch (type) switch (type)
{ {
@ -712,7 +712,7 @@ void GSRendererSW::UsePages(const GSOffset::PageLooper& pages, const int type)
void GSRendererSW::ReleasePages(const GSOffset::PageLooper& pages, const int type) void GSRendererSW::ReleasePages(const GSOffset::PageLooper& pages, const int type)
{ {
pages.loopPages([=](uint32 page) pages.loopPages([=](u32 page)
{ {
switch (type) switch (type)
{ {
@ -767,14 +767,14 @@ bool GSRendererSW::CheckTargetPages(const GSOffset::PageLooper* fb_pages, const
memset(m_fzb_cur_pages, 0, sizeof(m_fzb_cur_pages)); memset(m_fzb_cur_pages, 0, sizeof(m_fzb_cur_pages));
uint32 used = 0; u32 used = 0;
requirePages(); requirePages();
fb_pages->loopPages([&](uint32 i) fb_pages->loopPages([&](u32 i)
{ {
uint32 row = i >> 5; u32 row = i >> 5;
uint32 col = 1 << (i & 31); u32 col = 1 << (i & 31);
m_fzb_cur_pages[row] |= col; m_fzb_cur_pages[row] |= col;
@ -782,10 +782,10 @@ bool GSRendererSW::CheckTargetPages(const GSOffset::PageLooper* fb_pages, const
used |= m_tex_pages[i]; used |= m_tex_pages[i];
}); });
zb_pages->loopPages([&](uint32 i) zb_pages->loopPages([&](u32 i)
{ {
uint32 row = i >> 5; u32 row = i >> 5;
uint32 col = 1 << (i & 31); u32 col = 1 << (i & 31);
m_fzb_cur_pages[row] |= col; m_fzb_cur_pages[row] |= col;
@ -825,12 +825,12 @@ bool GSRendererSW::CheckTargetPages(const GSOffset::PageLooper* fb_pages, const
requirePages(); requirePages();
uint32 used = 0; u32 used = 0;
fb_pages->loopPages([&](uint32 i) fb_pages->loopPages([&](u32 i)
{ {
uint32 row = i >> 5; u32 row = i >> 5;
uint32 col = 1 << (i & 31); u32 col = 1 << (i & 31);
if ((m_fzb_cur_pages[row] & col) == 0) if ((m_fzb_cur_pages[row] & col) == 0)
{ {
@ -840,10 +840,10 @@ bool GSRendererSW::CheckTargetPages(const GSOffset::PageLooper* fb_pages, const
} }
}); });
zb_pages->loopPages([&](uint32 i) zb_pages->loopPages([&](u32 i)
{ {
uint32 row = i >> 5; u32 row = i >> 5;
uint32 col = 1 << (i & 31); u32 col = 1 << (i & 31);
if ((m_fzb_cur_pages[row] & col) == 0) if ((m_fzb_cur_pages[row] & col) == 0)
{ {
@ -875,7 +875,7 @@ bool GSRendererSW::CheckTargetPages(const GSOffset::PageLooper* fb_pages, const
if (fb && !res) if (fb && !res)
{ {
fb_pages->loopPagesWithBreak([&](uint32 page) fb_pages->loopPagesWithBreak([&](u32 page)
{ {
if (m_fzb_pages[page] & 0xffff0000) if (m_fzb_pages[page] & 0xffff0000)
{ {
@ -895,7 +895,7 @@ bool GSRendererSW::CheckTargetPages(const GSOffset::PageLooper* fb_pages, const
if (zb && !res) if (zb && !res)
{ {
zb_pages->loopPagesWithBreak([&](uint32 page) zb_pages->loopPagesWithBreak([&](u32 page)
{ {
if (m_fzb_pages[page] & 0x0000ffff) if (m_fzb_pages[page] & 0x0000ffff)
{ {
@ -927,7 +927,7 @@ bool GSRendererSW::CheckSourcePages(SharedData* sd)
GSOffset::PageLooper pages = sd->m_tex[i].t->m_offset.pageLooperForRect(sd->m_tex[i].r); GSOffset::PageLooper pages = sd->m_tex[i].t->m_offset.pageLooperForRect(sd->m_tex[i].r);
bool ret = false; bool ret = false;
pages.loopPagesWithBreak([&](uint32 pages) pages.loopPagesWithBreak([&](u32 pages)
{ {
// TODO: 8H 4HL 4HH texture at the same place as the render target (24 bit, or 32-bit where the alpha channel is masked, Valkyrie Profile 2) // TODO: 8H 4HL 4HH texture at the same place as the render target (24 bit, or 32-bit where the alpha channel is masked, Valkyrie Profile 2)
@ -972,8 +972,8 @@ bool GSRendererSW::GetScanlineGlobalData(SharedData* data)
gd.sel.ababcd = 0xff; gd.sel.ababcd = 0xff;
gd.sel.prim = primclass; gd.sel.prim = primclass;
uint32 fm = context->FRAME.FBMSK; u32 fm = context->FRAME.FBMSK;
uint32 zm = context->ZBUF.ZMSK || context->TEST.ZTE == 0 ? 0xffffffff : 0; u32 zm = context->ZBUF.ZMSK || context->TEST.ZTE == 0 ? 0xffffffff : 0;
if (context->TEST.ZTE && context->TEST.ZTST == ZTST_NEVER) if (context->TEST.ZTE && context->TEST.ZTST == ZTST_NEVER)
{ {
@ -1049,9 +1049,9 @@ bool GSRendererSW::GetScanlineGlobalData(SharedData* data)
{ {
gd.sel.tlu = 1; gd.sel.tlu = 1;
gd.clut = (uint32*)_aligned_malloc(sizeof(uint32) * 256, 32); // FIXME: might address uninitialized data of the texture (0xCD) that is not in 0-15 range for 4-bpp formats gd.clut = (u32*)_aligned_malloc(sizeof(u32) * 256, 32); // FIXME: might address uninitialized data of the texture (0xCD) that is not in 0-15 range for 4-bpp formats
memcpy(gd.clut, (const uint32*)m_mem.m_clut, sizeof(uint32) * GSLocalMemory::m_psm[context->TEX0.PSM].pal); memcpy(gd.clut, (const u32*)m_mem.m_clut, sizeof(u32) * GSLocalMemory::m_psm[context->TEX0.PSM].pal);
} }
gd.sel.wms = context->CLAMP.WMS; gd.sel.wms = context->CLAMP.WMS;
@ -1216,30 +1216,30 @@ bool GSRendererSW::GetScanlineGlobalData(SharedData* data)
} }
} }
uint16 tw = 1u << TEX0.TW; u16 tw = 1u << TEX0.TW;
uint16 th = 1u << TEX0.TH; u16 th = 1u << TEX0.TH;
switch (context->CLAMP.WMS) switch (context->CLAMP.WMS)
{ {
case CLAMP_REPEAT: case CLAMP_REPEAT:
gd.t.min.u16[0] = gd.t.minmax.u16[0] = tw - 1; gd.t.min.U16[0] = gd.t.minmax.U16[0] = tw - 1;
gd.t.max.u16[0] = gd.t.minmax.u16[2] = 0; gd.t.max.U16[0] = gd.t.minmax.U16[2] = 0;
gd.t.mask.u32[0] = 0xffffffff; gd.t.mask.U32[0] = 0xffffffff;
break; break;
case CLAMP_CLAMP: case CLAMP_CLAMP:
gd.t.min.u16[0] = gd.t.minmax.u16[0] = 0; gd.t.min.U16[0] = gd.t.minmax.U16[0] = 0;
gd.t.max.u16[0] = gd.t.minmax.u16[2] = tw - 1; gd.t.max.U16[0] = gd.t.minmax.U16[2] = tw - 1;
gd.t.mask.u32[0] = 0; gd.t.mask.U32[0] = 0;
break; break;
case CLAMP_REGION_CLAMP: case CLAMP_REGION_CLAMP:
gd.t.min.u16[0] = gd.t.minmax.u16[0] = std::min<uint16>(context->CLAMP.MINU, tw - 1); gd.t.min.U16[0] = gd.t.minmax.U16[0] = std::min<u16>(context->CLAMP.MINU, tw - 1);
gd.t.max.u16[0] = gd.t.minmax.u16[2] = std::min<uint16>(context->CLAMP.MAXU, tw - 1); gd.t.max.U16[0] = gd.t.minmax.U16[2] = std::min<u16>(context->CLAMP.MAXU, tw - 1);
gd.t.mask.u32[0] = 0; gd.t.mask.U32[0] = 0;
break; break;
case CLAMP_REGION_REPEAT: case CLAMP_REGION_REPEAT:
gd.t.min.u16[0] = gd.t.minmax.u16[0] = context->CLAMP.MINU & (tw - 1); gd.t.min.U16[0] = gd.t.minmax.U16[0] = context->CLAMP.MINU & (tw - 1);
gd.t.max.u16[0] = gd.t.minmax.u16[2] = context->CLAMP.MAXU & (tw - 1); gd.t.max.U16[0] = gd.t.minmax.U16[2] = context->CLAMP.MAXU & (tw - 1);
gd.t.mask.u32[0] = 0xffffffff; gd.t.mask.U32[0] = 0xffffffff;
break; break;
default: default:
__assume(0); __assume(0);
@ -1248,24 +1248,24 @@ bool GSRendererSW::GetScanlineGlobalData(SharedData* data)
switch (context->CLAMP.WMT) switch (context->CLAMP.WMT)
{ {
case CLAMP_REPEAT: case CLAMP_REPEAT:
gd.t.min.u16[4] = gd.t.minmax.u16[1] = th - 1; gd.t.min.U16[4] = gd.t.minmax.U16[1] = th - 1;
gd.t.max.u16[4] = gd.t.minmax.u16[3] = 0; gd.t.max.U16[4] = gd.t.minmax.U16[3] = 0;
gd.t.mask.u32[2] = 0xffffffff; gd.t.mask.U32[2] = 0xffffffff;
break; break;
case CLAMP_CLAMP: case CLAMP_CLAMP:
gd.t.min.u16[4] = gd.t.minmax.u16[1] = 0; gd.t.min.U16[4] = gd.t.minmax.U16[1] = 0;
gd.t.max.u16[4] = gd.t.minmax.u16[3] = th - 1; gd.t.max.U16[4] = gd.t.minmax.U16[3] = th - 1;
gd.t.mask.u32[2] = 0; gd.t.mask.U32[2] = 0;
break; break;
case CLAMP_REGION_CLAMP: case CLAMP_REGION_CLAMP:
gd.t.min.u16[4] = gd.t.minmax.u16[1] = std::min<uint16>(context->CLAMP.MINV, th - 1); gd.t.min.U16[4] = gd.t.minmax.U16[1] = std::min<u16>(context->CLAMP.MINV, th - 1);
gd.t.max.u16[4] = gd.t.minmax.u16[3] = std::min<uint16>(context->CLAMP.MAXV, th - 1); // ffx anima summon scene, when the anchor appears (th = 256, maxv > 256) gd.t.max.U16[4] = gd.t.minmax.U16[3] = std::min<u16>(context->CLAMP.MAXV, th - 1); // ffx anima summon scene, when the anchor appears (th = 256, maxv > 256)
gd.t.mask.u32[2] = 0; gd.t.mask.U32[2] = 0;
break; break;
case CLAMP_REGION_REPEAT: case CLAMP_REGION_REPEAT:
gd.t.min.u16[4] = gd.t.minmax.u16[1] = context->CLAMP.MINV & (th - 1); // skygunner main menu water texture 64x64, MINV = 127 gd.t.min.U16[4] = gd.t.minmax.U16[1] = context->CLAMP.MINV & (th - 1); // skygunner main menu water texture 64x64, MINV = 127
gd.t.max.u16[4] = gd.t.minmax.u16[3] = context->CLAMP.MAXV & (th - 1); gd.t.max.U16[4] = gd.t.minmax.U16[3] = context->CLAMP.MAXV & (th - 1);
gd.t.mask.u32[2] = 0xffffffff; gd.t.mask.U32[2] = 0xffffffff;
break; break;
default: default:
__assume(0); __assume(0);
@ -1281,8 +1281,8 @@ bool GSRendererSW::GetScanlineGlobalData(SharedData* data)
{ {
gd.sel.fge = 1; gd.sel.fge = 1;
gd.frb = env.FOGCOL.u32[0] & 0x00ff00ff; gd.frb = env.FOGCOL.U32[0] & 0x00ff00ff;
gd.fga = (env.FOGCOL.u32[0] >> 8) & 0x00ff00ff; gd.fga = (env.FOGCOL.U32[0] >> 8) & 0x00ff00ff;
} }
if (context->FRAME.PSM != PSM_PSMCT24) if (context->FRAME.PSM != PSM_PSMCT24)
@ -1294,7 +1294,7 @@ bool GSRendererSW::GetScanlineGlobalData(SharedData* data)
if (!IsOpaque()) if (!IsOpaque())
{ {
gd.sel.abe = PRIM->ABE; gd.sel.abe = PRIM->ABE;
gd.sel.ababcd = context->ALPHA.u32[0]; gd.sel.ababcd = context->ALPHA.U32[0];
if (env.PABE.PABE) if (env.PABE.PABE)
{ {
@ -1337,12 +1337,12 @@ bool GSRendererSW::GetScanlineGlobalData(SharedData* data)
if (zwrite || ztest) if (zwrite || ztest)
{ {
uint32_t z_max = 0xffffffff >> (GSLocalMemory::m_psm[context->ZBUF.PSM].fmt * 8); u32 z_max = 0xffffffff >> (GSLocalMemory::m_psm[context->ZBUF.PSM].fmt * 8);
gd.sel.zpsm = GSLocalMemory::m_psm[context->ZBUF.PSM].fmt; gd.sel.zpsm = GSLocalMemory::m_psm[context->ZBUF.PSM].fmt;
gd.sel.ztst = ztest ? context->TEST.ZTST : (int)ZTST_ALWAYS; gd.sel.ztst = ztest ? context->TEST.ZTST : (int)ZTST_ALWAYS;
gd.sel.zoverflow = (uint32)GSVector4i(m_vt.m_max.p).z == 0x80000000U; gd.sel.zoverflow = (u32)GSVector4i(m_vt.m_max.p).z == 0x80000000U;
gd.sel.zclamp = (uint32)GSVector4i(m_vt.m_max.p).z > z_max; gd.sel.zclamp = (u32)GSVector4i(m_vt.m_max.p).z > z_max;
} }
#if _M_SSE >= 0x501 #if _M_SSE >= 0x501
@ -1356,8 +1356,8 @@ bool GSRendererSW::GetScanlineGlobalData(SharedData* data)
} }
else if (gd.sel.fpsm == 2) else if (gd.sel.fpsm == 2)
{ {
uint32 rb = gd.fm & 0x00f800f8; u32 rb = gd.fm & 0x00f800f8;
uint32 ga = gd.fm & 0x8000f800; u32 ga = gd.fm & 0x8000f800;
gd.fm = (ga >> 16) | (rb >> 9) | (ga >> 6) | (rb >> 3) | 0xffff0000; gd.fm = (ga >> 16) | (rb >> 9) | (ga >> 6) | (rb >> 3) | 0xffff0000;
} }
@ -1403,7 +1403,7 @@ bool GSRendererSW::GetScanlineGlobalData(SharedData* data)
{ {
gd.sel.notest = 1; gd.sel.notest = 1;
uint32 ofx = context->XYOFFSET.OFX; u32 ofx = context->XYOFFSET.OFX;
for (int i = 0, j = m_vertex.tail; i < j; i++) for (int i = 0, j = m_vertex.tail; i < j; i++)
{ {
@ -1551,7 +1551,7 @@ void GSRendererSW::SharedData::UpdateSource()
if (m_parent->s_dump) if (m_parent->s_dump)
{ {
uint64 frame = m_parent->m_perfmon.GetFrame(); u64 frame = m_parent->m_perfmon.GetFrame();
std::string s; std::string s;
@ -1570,7 +1570,7 @@ void GSRendererSW::SharedData::UpdateSource()
{ {
GSTextureSW* t = new GSTextureSW(0, 256, 1); GSTextureSW* t = new GSTextureSW(0, 256, 1);
t->Update(GSVector4i(0, 0, 256, 1), global.clut, sizeof(uint32) * 256); t->Update(GSVector4i(0, 0, 256, 1), global.clut, sizeof(u32) * 256);
s = format("%05d_f%lld_itexp_%05x_%s.bmp", m_parent->s_n, frame, (int)m_parent->m_context->TEX0.CBP, psm_str(m_parent->m_context->TEX0.CPSM)); s = format("%05d_f%lld_itexp_%05x_%s.bmp", m_parent->s_n, frame, (int)m_parent->m_context->TEX0.CBP, psm_str(m_parent->m_context->TEX0.CPSM));

10
pcsx2/GS/Renderers/SW/GSRendererSW.h
View File

@ -63,19 +63,19 @@ class GSRendererSW : public GSRenderer
ConvertVertexBufferPtr m_cvb[4][2][2][2]; ConvertVertexBufferPtr m_cvb[4][2][2][2];
template <uint32 primclass, uint32 tme, uint32 fst, uint32 q_div> template <u32 primclass, u32 tme, u32 fst, u32 q_div>
void ConvertVertexBuffer(GSVertexSW* RESTRICT dst, const GSVertex* RESTRICT src, size_t count); void ConvertVertexBuffer(GSVertexSW* RESTRICT dst, const GSVertex* RESTRICT src, size_t count);
protected: protected:
IRasterizer* m_rl; IRasterizer* m_rl;
GSTextureCacheSW* m_tc; GSTextureCacheSW* m_tc;
GSTexture* m_texture[2]; GSTexture* m_texture[2];
uint8* m_output; u8* m_output;
GSPixelOffset4* m_fzb; GSPixelOffset4* m_fzb;
GSVector4i m_fzb_bbox; GSVector4i m_fzb_bbox;
uint32 m_fzb_cur_pages[16]; u32 m_fzb_cur_pages[16];
std::atomic<uint32> m_fzb_pages[512]; // uint16 frame/zbuf pages interleaved std::atomic<u32> m_fzb_pages[512]; // u16 frame/zbuf pages interleaved
std::atomic<uint16> m_tex_pages[512]; std::atomic<u16> m_tex_pages[512];
void Reset(); void Reset();
void VSync(int field); void VSync(int field);

122
pcsx2/GS/Renderers/SW/GSScanlineEnvironment.h
View File

@ -22,79 +22,79 @@ union GSScanlineSelector
{ {
struct struct
{ {
uint32 fpsm : 2; // 0 u32 fpsm : 2; // 0
uint32 zpsm : 2; // 2 u32 zpsm : 2; // 2
uint32 ztst : 2; // 4 (0: off, 1: write, 2: test (ge), 3: test (g)) u32 ztst : 2; // 4 (0: off, 1: write, 2: test (ge), 3: test (g))
uint32 atst : 3; // 6 u32 atst : 3; // 6
uint32 afail : 2; // 9 u32 afail : 2; // 9
uint32 iip : 1; // 11 u32 iip : 1; // 11
uint32 tfx : 3; // 12 u32 tfx : 3; // 12
uint32 tcc : 1; // 15 u32 tcc : 1; // 15
uint32 fst : 1; // 16 u32 fst : 1; // 16
uint32 ltf : 1; // 17 u32 ltf : 1; // 17
uint32 tlu : 1; // 18 u32 tlu : 1; // 18
uint32 fge : 1; // 19 u32 fge : 1; // 19
uint32 date : 1; // 20 u32 date : 1; // 20
uint32 abe : 1; // 21 u32 abe : 1; // 21
uint32 aba : 2; // 22 u32 aba : 2; // 22
uint32 abb : 2; // 24 u32 abb : 2; // 24
uint32 abc : 2; // 26 u32 abc : 2; // 26
uint32 abd : 2; // 28 u32 abd : 2; // 28
uint32 pabe : 1; // 30 u32 pabe : 1; // 30
uint32 aa1 : 1; // 31 u32 aa1 : 1; // 31
uint32 fwrite : 1; // 32 u32 fwrite : 1; // 32
uint32 ftest : 1; // 33 u32 ftest : 1; // 33
uint32 rfb : 1; // 34 u32 rfb : 1; // 34
uint32 zwrite : 1; // 35 u32 zwrite : 1; // 35
uint32 ztest : 1; // 36 u32 ztest : 1; // 36
uint32 zoverflow : 1; // 37 (z max >= 0x80000000) u32 zoverflow : 1; // 37 (z max >= 0x80000000)
uint32 zclamp : 1; // 38 u32 zclamp : 1; // 38
uint32 wms : 2; // 39 u32 wms : 2; // 39
uint32 wmt : 2; // 41 u32 wmt : 2; // 41
uint32 datm : 1; // 43 u32 datm : 1; // 43
uint32 colclamp : 1; // 44 u32 colclamp : 1; // 44
uint32 fba : 1; // 45 u32 fba : 1; // 45
uint32 dthe : 1; // 46 u32 dthe : 1; // 46
uint32 prim : 2; // 47 u32 prim : 2; // 47
uint32 edge : 1; // 49 u32 edge : 1; // 49
uint32 tw : 3; // 50 (encodes values between 3 -> 10, texture cache makes sure it is at least 3) u32 tw : 3; // 50 (encodes values between 3 -> 10, texture cache makes sure it is at least 3)
uint32 lcm : 1; // 53 u32 lcm : 1; // 53
uint32 mmin : 2; // 54 u32 mmin : 2; // 54
uint32 notest : 1; // 55 (no ztest, no atest, no date, no scissor test, and horizontally aligned to 4 pixels) u32 notest : 1; // 55 (no ztest, no atest, no date, no scissor test, and horizontally aligned to 4 pixels)
// TODO: 1D texture flag? could save 2 texture reads and 4 lerps with bilinear, and also the texture coordinate clamp/wrap code in one direction // TODO: 1D texture flag? could save 2 texture reads and 4 lerps with bilinear, and also the texture coordinate clamp/wrap code in one direction
uint32 breakpoint : 1; // Insert a trap to stop the program, helpful to stop debugger on a program u32 breakpoint : 1; // Insert a trap to stop the program, helpful to stop debugger on a program
}; };
struct struct
{ {
uint32 _pad1 : 22; u32 _pad1 : 22;
uint32 ababcd : 8; u32 ababcd : 8;
uint32 _pad2 : 2; u32 _pad2 : 2;
uint32 fb : 2; u32 fb : 2;
uint32 _pad3 : 1; u32 _pad3 : 1;
uint32 zb : 2; u32 zb : 2;
}; };
struct struct
{ {
uint32 lo; u32 lo;
uint32 hi; u32 hi;
}; };
uint64 key; u64 key;
GSScanlineSelector() = default; GSScanlineSelector() = default;
GSScanlineSelector(uint64 k) GSScanlineSelector(u64 k)
: key(k) : key(k)
{ {
} }
operator uint32() const { return lo; } operator u32() const { return lo; }
operator uint64() const { return key; } operator u64() const { return key; }
bool IsSolidRect() const bool IsSolidRect() const
{ {
@ -124,7 +124,7 @@ struct alignas(32) GSScanlineGlobalData // per batch variables, this is like a p
void* vm; void* vm;
const void* tex[7]; const void* tex[7];
uint32* clut; u32* clut;
GSVector4i* dimx; GSVector4i* dimx;
GSOffset fbo; GSOffset fbo;
@ -138,8 +138,8 @@ struct alignas(32) GSScanlineGlobalData // per batch variables, this is like a p
#if _M_SSE >= 0x501 #if _M_SSE >= 0x501
uint32 fm, zm; u32 fm, zm;
uint32 frb, fga; u32 frb, fga;
GSVector8 mxl; GSVector8 mxl;
GSVector8 k; // TEX1.K * 0x10000 GSVector8 k; // TEX1.K * 0x10000
GSVector8 l; // TEX1.L * -0x10000 GSVector8 l; // TEX1.L * -0x10000
@ -162,9 +162,9 @@ struct alignas(32) GSScanlineLocalData // per prim variables, each thread has it
#if _M_SSE >= 0x501 #if _M_SSE >= 0x501
struct skip { GSVector8 z, s, t, q; GSVector8i rb, ga, f, _pad; } d[8]; struct skip { GSVector8 z, s, t, q; GSVector8i rb, ga, f, _pad; } d[8];
struct step { GSVector4 stq; struct { uint32 rb, ga; } c; struct { uint32 z, f; } p; } d8; struct step { GSVector4 stq; struct { u32 rb, ga; } c; struct { u32 z, f; } p; } d8;
struct { GSVector8i rb, ga; } c; struct { GSVector8i rb, ga; } c;
struct { uint32 z, f; } p; struct { u32 z, f; } p;
// these should be stored on stack as normal local variables (no free regs to use, esp cannot be saved to anywhere, and we need an aligned stack) // these should be stored on stack as normal local variables (no free regs to use, esp cannot be saved to anywhere, and we need an aligned stack)
@ -231,11 +231,11 @@ struct alignas(32) GSScanlineLocalData // per prim variables, each thread has it
// absolute addressing. Otherwise we need to store a base address in a register. // absolute addressing. Otherwise we need to store a base address in a register.
struct GSScanlineConstantData : public GSAlignedClass<32> struct GSScanlineConstantData : public GSAlignedClass<32>
{ {
alignas(32) uint8 m_test_256b[16][8]; alignas(32) u8 m_test_256b[16][8];
alignas(32) float m_shift_256b[9][8]; alignas(32) float m_shift_256b[9][8];
alignas(32) float m_log2_coef_256b[4][8]; alignas(32) float m_log2_coef_256b[4][8];
alignas(16) uint32 m_test_128b[8][4]; alignas(16) u32 m_test_128b[8][4];
alignas(16) float m_shift_128b[5][4]; alignas(16) float m_shift_128b[5][4];
alignas(16) float m_log2_coef_128b[4][4]; alignas(16) float m_log2_coef_128b[4][4];
@ -245,7 +245,7 @@ struct GSScanlineConstantData : public GSAlignedClass<32>
// So it must be defered to post global constructor // So it must be defered to post global constructor
void Init() void Init()
{ {
uint8 I_hate_vs2013_m_test_256b[16][8] = { u8 I_hate_vs2013_m_test_256b[16][8] = {
{0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00}, {0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00},
{0xff, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00}, {0xff, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00},
{0xff, 0xff, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00}, {0xff, 0xff, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00},
@ -264,7 +264,7 @@ struct GSScanlineConstantData : public GSAlignedClass<32>
{0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00} {0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00}
}; };
uint32 I_hate_vs2013_m_test_128b[8][4] = { u32 I_hate_vs2013_m_test_128b[8][4] = {
{0x00000000, 0x00000000, 0x00000000, 0x00000000}, {0x00000000, 0x00000000, 0x00000000, 0x00000000},
{0xffffffff, 0x00000000, 0x00000000, 0x00000000}, {0xffffffff, 0x00000000, 0x00000000, 0x00000000},
{0xffffffff, 0xffffffff, 0x00000000, 0x00000000}, {0xffffffff, 0xffffffff, 0x00000000, 0x00000000},

12
pcsx2/GS/Renderers/SW/GSSetupPrimCodeGenerator.all.cpp
View File

@ -48,7 +48,7 @@ using namespace Xbyak;
#define _rip_local_d_p(x) _rip_local_d(x) #define _rip_local_d_p(x) _rip_local_d(x)
#endif #endif
GSSetupPrimCodeGenerator2::GSSetupPrimCodeGenerator2(Xbyak::CodeGenerator* base, CPUInfo cpu, void* param, uint64 key) GSSetupPrimCodeGenerator2::GSSetupPrimCodeGenerator2(Xbyak::CodeGenerator* base, CPUInfo cpu, void* param, u64 key)
: _parent(base, cpu) : _parent(base, cpu)
, m_local(*(GSScanlineLocalData*)param) , m_local(*(GSScanlineLocalData*)param)
, m_rip(false), many_regs(false) , m_rip(false), many_regs(false)
@ -212,7 +212,7 @@ void GSSetupPrimCodeGenerator2::Depth_XMM()
if (is32) if (is32)
mov(_index, ptr[rsp + _32_index]); mov(_index, ptr[rsp + _32_index]);
mov(eax, ptr[_index + sizeof(uint32) * 1]); mov(eax, ptr[_index + sizeof(u32) * 1]);
shl(eax, 6); // * sizeof(GSVertexSW) shl(eax, 6); // * sizeof(GSVertexSW)
if (is64) if (is64)
add(rax, _64_vertex); add(rax, _64_vertex);
@ -232,7 +232,7 @@ void GSSetupPrimCodeGenerator2::Depth_XMM()
if (m_en.z) if (m_en.z)
{ {
// uint32 z is bypassed in t.w // u32 z is bypassed in t.w
movdqa(xmm0, ptr[rax + offsetof(GSVertexSW, t)]); movdqa(xmm0, ptr[rax + offsetof(GSVertexSW, t)]);
pshufd(xmm0, xmm0, _MM_SHUFFLE(3, 3, 3, 3)); pshufd(xmm0, xmm0, _MM_SHUFFLE(3, 3, 3, 3));
@ -314,7 +314,7 @@ void GSSetupPrimCodeGenerator2::Depth_YMM()
if (is32) if (is32)
mov(_index, ptr[rsp + _32_index]); mov(_index, ptr[rsp + _32_index]);
mov(eax, ptr[_index + sizeof(uint32) * 1]); mov(eax, ptr[_index + sizeof(u32) * 1]);
shl(eax, 6); // * sizeof(GSVertexSW) shl(eax, 6); // * sizeof(GSVertexSW)
if (is64) if (is64)
add(rax, _64_vertex); add(rax, _64_vertex);
@ -332,7 +332,7 @@ void GSSetupPrimCodeGenerator2::Depth_YMM()
if (m_en.z) if (m_en.z)
{ {
// m_local.p.z = vertex[index[1]].t.u32[3]; // uint32 z is bypassed in t.w // m_local.p.z = vertex[index[1]].t.u32[3]; // u32 z is bypassed in t.w
mov(t1.cvt32(), ptr[rax + offsetof(GSVertexSW, t.w)]); mov(t1.cvt32(), ptr[rax + offsetof(GSVertexSW, t.w)]);
mov(_rip_local(p.z), t1.cvt32()); mov(_rip_local(p.z), t1.cvt32());
@ -524,7 +524,7 @@ void GSSetupPrimCodeGenerator2::Color()
{ {
if (is32) if (is32)
mov(_index, ptr[rsp + _32_index]); mov(_index, ptr[rsp + _32_index]);
mov(eax, ptr[_index + sizeof(uint32) * last]); mov(eax, ptr[_index + sizeof(u32) * last]);
shl(eax, 6); // * sizeof(GSVertexSW) shl(eax, 6); // * sizeof(GSVertexSW)
if (is64) if (is64)
add(rax, _64_vertex); add(rax, _64_vertex);

4
pcsx2/GS/Renderers/SW/GSSetupPrimCodeGenerator.all.h
View File

@ -57,7 +57,7 @@ class GSSetupPrimCodeGenerator2 : public GSNewCodeGenerator
bool m_rip; bool m_rip;
bool many_regs; bool many_regs;
struct {uint32 z:1, f:1, t:1, c:1;} m_en; struct {u32 z:1, f:1, t:1, c:1;} m_en;
const XYm xym0{0}, xym1{1}, xym2{2}, xym3{3}, xym4{4}, xym5{5}, xym6{6}, xym7{7}, xym8{8}, xym9{9}, xym10{10}, xym11{11}, xym12{12}, xym13{13}, xym14{14}, xym15{15}; const XYm xym0{0}, xym1{1}, xym2{2}, xym3{3}, xym4{4}, xym5{5}, xym6{6}, xym7{7}, xym8{8}, xym9{9}, xym10{10}, xym11{11}, xym12{12}, xym13{13}, xym14{14}, xym15{15};
const AddressReg _64_vertex, _index, _dscan, _64_t0, t1; const AddressReg _64_vertex, _index, _dscan, _64_t0, t1;
@ -69,7 +69,7 @@ class GSSetupPrimCodeGenerator2 : public GSNewCodeGenerator
} }
public: public:
GSSetupPrimCodeGenerator2(Xbyak::CodeGenerator* base, CPUInfo cpu, void* param, uint64 key); GSSetupPrimCodeGenerator2(Xbyak::CodeGenerator* base, CPUInfo cpu, void* param, u64 key);
void Generate(); void Generate();
private: private:

2
pcsx2/GS/Renderers/SW/GSSetupPrimCodeGenerator.cpp
View File

@ -19,7 +19,7 @@
using namespace Xbyak; using namespace Xbyak;
GSSetupPrimCodeGenerator::GSSetupPrimCodeGenerator(void* param, uint64 key, void* code, size_t maxsize) GSSetupPrimCodeGenerator::GSSetupPrimCodeGenerator(void* param, u64 key, void* code, size_t maxsize)
: GSCodeGenerator(code, maxsize) : GSCodeGenerator(code, maxsize)
, m_local(*(GSScanlineLocalData*)param) , m_local(*(GSScanlineLocalData*)param)
, m_rip(false) , m_rip(false)

4
pcsx2/GS/Renderers/SW/GSSetupPrimCodeGenerator.h
View File

@ -29,9 +29,9 @@ class GSSetupPrimCodeGenerator : public GSCodeGenerator
struct struct
{ {
uint32 z : 1, f : 1, t : 1, c : 1; u32 z : 1, f : 1, t : 1, c : 1;
} m_en; } m_en;
public: public:
GSSetupPrimCodeGenerator(void* param, uint64 key, void* code, size_t maxsize); GSSetupPrimCodeGenerator(void* param, u64 key, void* code, size_t maxsize);
}; };

44
pcsx2/GS/Renderers/SW/GSTextureCacheSW.cpp
View File

@ -26,7 +26,7 @@ GSTextureCacheSW::~GSTextureCacheSW()
RemoveAll(); RemoveAll();
} }
GSTextureCacheSW::Texture* GSTextureCacheSW::Lookup(const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA, uint32 tw0) GSTextureCacheSW::Texture* GSTextureCacheSW::Lookup(const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA, u32 tw0)
{ {
const GSLocalMemory::psm_t& psm = GSLocalMemory::m_psm[TEX0.PSM]; const GSLocalMemory::psm_t& psm = GSLocalMemory::m_psm[TEX0.PSM];
@ -36,7 +36,7 @@ GSTextureCacheSW::Texture* GSTextureCacheSW::Lookup(const GIFRegTEX0& TEX0, cons
{ {
Texture* t = *i; Texture* t = *i;
if (((TEX0.u32[0] ^ t->m_TEX0.u32[0]) | ((TEX0.u32[1] ^ t->m_TEX0.u32[1]) & 3)) != 0) // TBP0 TBW PSM TW TH if (((TEX0.U32[0] ^ t->m_TEX0.U32[0]) | ((TEX0.U32[1] ^ t->m_TEX0.U32[1]) & 3)) != 0) // TBP0 TBW PSM TW TH
{ {
continue; continue;
} }
@ -62,7 +62,7 @@ GSTextureCacheSW::Texture* GSTextureCacheSW::Lookup(const GIFRegTEX0& TEX0, cons
m_textures.insert(t); m_textures.insert(t);
t->m_pages.loopPages([&](uint32 page) t->m_pages.loopPages([&](u32 page)
{ {
t->m_erase_it[page] = m_map[page].InsertFront(t); t->m_erase_it[page] = m_map[page].InsertFront(t);
}); });
@ -70,15 +70,15 @@ GSTextureCacheSW::Texture* GSTextureCacheSW::Lookup(const GIFRegTEX0& TEX0, cons
return t; return t;
} }
void GSTextureCacheSW::InvalidatePages(const GSOffset::PageLooper& pages, uint32 psm) void GSTextureCacheSW::InvalidatePages(const GSOffset::PageLooper& pages, u32 psm)
{ {
pages.loopPages([&](uint32 page) pages.loopPages([&](u32 page)
{ {
for (Texture* t : m_map[page]) for (Texture* t : m_map[page])
{ {
if (GSUtil::HasSharedBits(psm, t->m_sharedbits)) if (GSUtil::HasSharedBits(psm, t->m_sharedbits))
{ {
uint32* RESTRICT valid = t->m_valid; u32* RESTRICT valid = t->m_valid;
if (t->m_repeating) if (t->m_repeating)
{ {
@ -121,7 +121,7 @@ void GSTextureCacheSW::IncAge()
{ {
i = m_textures.erase(i); i = m_textures.erase(i);
t->m_pages.loopPages([&](uint32 page) t->m_pages.loopPages([&](u32 page)
{ {
m_map[page].EraseIndex(t->m_erase_it[page]); m_map[page].EraseIndex(t->m_erase_it[page]);
}); });
@ -137,7 +137,7 @@ void GSTextureCacheSW::IncAge()
// //
GSTextureCacheSW::Texture::Texture(GSState* state, uint32 tw0, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA) GSTextureCacheSW::Texture::Texture(GSState* state, u32 tw0, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA)
: m_state(state) : m_state(state)
, m_buff(NULL) , m_buff(NULL)
, m_tw(tw0) , m_tw(tw0)
@ -203,7 +203,7 @@ bool GSTextureCacheSW::Texture::Update(const GSVector4i& rect)
if (m_buff == NULL) if (m_buff == NULL)
{ {
uint32 pitch = (1 << m_tw) << shift; u32 pitch = (1 << m_tw) << shift;
m_buff = _aligned_malloc(pitch * th * 4, 32); m_buff = _aligned_malloc(pitch * th * 4, 32);
@ -217,13 +217,13 @@ bool GSTextureCacheSW::Texture::Update(const GSVector4i& rect)
GSOffset off = m_offset; GSOffset off = m_offset;
uint32 blocks = 0; u32 blocks = 0;
GSLocalMemory::readTextureBlock rtxbP = psm.rtxbP; GSLocalMemory::readTextureBlock rtxbP = psm.rtxbP;
uint32 pitch = (1 << m_tw) << shift; u32 pitch = (1 << m_tw) << shift;
uint8* dst = (uint8*)m_buff + pitch * r.top; u8* dst = (u8*)m_buff + pitch * r.top;
int block_pitch = pitch * bs.y; int block_pitch = pitch * bs.y;
@ -240,10 +240,10 @@ bool GSTextureCacheSW::Texture::Update(const GSVector4i& rect)
for (; bn.blkX() < right; bn.nextBlockX()) for (; bn.blkX() < right; bn.nextBlockX())
{ {
int i = (bn.blkY() << 7) + bn.blkX(); int i = (bn.blkY() << 7) + bn.blkX();
uint32 block = bn.value(); u32 block = bn.value();
uint32 row = i >> 5; u32 row = i >> 5;
uint32 col = 1 << (i & 31); u32 col = 1 << (i & 31);
if ((m_valid[row] & col) == 0) if ((m_valid[row] & col) == 0)
{ {
@ -262,10 +262,10 @@ bool GSTextureCacheSW::Texture::Update(const GSVector4i& rect)
{ {
for (; bn.blkX() < right; bn.nextBlockX()) for (; bn.blkX() < right; bn.nextBlockX())
{ {
uint32 block = bn.value(); u32 block = bn.value();
uint32 row = block >> 5; u32 row = block >> 5;
uint32 col = 1 << (block & 31); u32 col = 1 << (block & 31);
if ((m_valid[row] & col) == 0) if ((m_valid[row] & col) == 0)
{ {
@ -291,7 +291,7 @@ bool GSTextureCacheSW::Texture::Update(const GSVector4i& rect)
bool GSTextureCacheSW::Texture::Save(const std::string& fn, bool dds) const bool GSTextureCacheSW::Texture::Save(const std::string& fn, bool dds) const
{ {
const uint32* RESTRICT clut = m_state->m_mem.m_clut; const u32* RESTRICT clut = m_state->m_mem.m_clut;
int w = 1 << m_TEX0.TW; int w = 1 << m_TEX0.TW;
int h = 1 << m_TEX0.TH; int h = 1 << m_TEX0.TH;
@ -304,20 +304,20 @@ bool GSTextureCacheSW::Texture::Save(const std::string& fn, bool dds) const
{ {
const GSLocalMemory::psm_t& psm = GSLocalMemory::m_psm[m_TEX0.PSM]; const GSLocalMemory::psm_t& psm = GSLocalMemory::m_psm[m_TEX0.PSM];
const uint8* RESTRICT src = (uint8*)m_buff; const u8* RESTRICT src = (u8*)m_buff;
int pitch = 1 << (m_tw + (psm.pal == 0 ? 2 : 0)); int pitch = 1 << (m_tw + (psm.pal == 0 ? 2 : 0));
for (int j = 0; j < h; j++, src += pitch, m.bits += m.pitch) for (int j = 0; j < h; j++, src += pitch, m.bits += m.pitch)
{ {
if (psm.pal == 0) if (psm.pal == 0)
{ {
memcpy(m.bits, src, sizeof(uint32) * w); memcpy(m.bits, src, sizeof(u32) * w);
} }
else else
{ {
for (int i = 0; i < w; i++) for (int i = 0; i < w; i++)
{ {
((uint32*)m.bits)[i] = clut[src[i]]; ((u32*)m.bits)[i] = clut[src[i]];
} }
} }
} }

20
pcsx2/GS/Renderers/SW/GSTextureCacheSW.h
View File

@ -30,20 +30,20 @@ public:
GIFRegTEX0 m_TEX0; GIFRegTEX0 m_TEX0;
GIFRegTEXA m_TEXA; GIFRegTEXA m_TEXA;
void* m_buff; void* m_buff;
uint32 m_tw; u32 m_tw;
uint32 m_age; u32 m_age;
bool m_complete; bool m_complete;
bool m_repeating; bool m_repeating;
std::vector<GSVector2i>* m_p2t; std::vector<GSVector2i>* m_p2t;
uint32 m_valid[MAX_PAGES]; u32 m_valid[MAX_PAGES];
std::array<uint16, MAX_PAGES> m_erase_it; std::array<u16, MAX_PAGES> m_erase_it;
const uint32* RESTRICT m_sharedbits; const u32* RESTRICT m_sharedbits;
// m_valid // m_valid
// fast mode: each uint32 bits map to the 32 blocks of that page // fast mode: each u32 bits map to the 32 blocks of that page
// repeating mode: 1 bpp image of the texture tiles (8x8), also having 512 elements is just a coincidence (worst case: (1024*1024)/(8*8)/(sizeof(uint32)*8)) // repeating mode: 1 bpp image of the texture tiles (8x8), also having 512 elements is just a coincidence (worst case: (1024*1024)/(8*8)/(sizeof(u32)*8))
Texture(GSState* state, uint32 tw0, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA); Texture(GSState* state, u32 tw0, const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA);
virtual ~Texture(); virtual ~Texture();
bool Update(const GSVector4i& r); bool Update(const GSVector4i& r);
@ -59,9 +59,9 @@ public:
GSTextureCacheSW(GSState* state); GSTextureCacheSW(GSState* state);
virtual ~GSTextureCacheSW(); virtual ~GSTextureCacheSW();
Texture* Lookup(const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA, uint32 tw0 = 0); Texture* Lookup(const GIFRegTEX0& TEX0, const GIFRegTEXA& TEXA, u32 tw0 = 0);
void InvalidatePages(const GSOffset::PageLooper& pages, uint32 psm); void InvalidatePages(const GSOffset::PageLooper& pages, u32 psm);
void RemoveAll(); void RemoveAll();
void IncAge(); void IncAge();

8
pcsx2/GS/Renderers/SW/GSTextureSW.cpp
View File

@ -38,8 +38,8 @@ bool GSTextureSW::Update(const GSVector4i& r, const void* data, int pitch, int l
if (m_data != NULL && Map(m, &r)) if (m_data != NULL && Map(m, &r))
{ {
uint8* RESTRICT src = (uint8*)data; u8* RESTRICT src = (u8*)data;
uint8* RESTRICT dst = m.bits; u8* RESTRICT dst = m.bits;
int rowbytes = r.width() << 2; int rowbytes = r.width() << 2;
@ -64,7 +64,7 @@ bool GSTextureSW::Map(GSMap& m, const GSVector4i* r, int layer)
{ {
if (!m_mapped.test_and_set(std::memory_order_acquire)) if (!m_mapped.test_and_set(std::memory_order_acquire))
{ {
m.bits = (uint8*)m_data + m_pitch * r2.top + (r2.left << 2); m.bits = (u8*)m_data + m_pitch * r2.top + (r2.left << 2);
m.pitch = m_pitch; m.pitch = m_pitch;
return true; return true;
@ -87,5 +87,5 @@ bool GSTextureSW::Save(const std::string& fn)
GSPng::Format fmt = GSPng::RGB_PNG; GSPng::Format fmt = GSPng::RGB_PNG;
#endif #endif
int compression = theApp.GetConfigI("png_compression_level"); int compression = theApp.GetConfigI("png_compression_level");
return GSPng::Save(fmt, fn, static_cast<uint8*>(m_data), m_size.x, m_size.y, m_pitch, compression); return GSPng::Save(fmt, fn, static_cast<u8*>(m_data), m_size.x, m_size.y, m_pitch, compression);
} }

50
tests/ctest/GS/swizzle_test_main.cpp
View File

@ -19,7 +19,7 @@
#include <gtest/gtest.h> #include <gtest/gtest.h>
#include <string.h> #include <string.h>
static void swizzle(const uint8* table, uint8* dst, const uint8* src, int bpp, bool deswizzle) static void swizzle(const u8* table, u8* dst, const u8* src, int bpp, bool deswizzle)
{ {
int pxbytes = bpp / 8; int pxbytes = bpp / 8;
for (int i = 0; i < (256 / pxbytes); i++) for (int i = 0; i < (256 / pxbytes); i++)
@ -30,21 +30,21 @@ static void swizzle(const uint8* table, uint8* dst, const uint8* src, int bpp, b
} }
} }
static void swizzle4(const uint16* table, uint8* dst, const uint8* src, bool deswizzle) static void swizzle4(const u16* table, u8* dst, const u8* src, bool deswizzle)
{ {
for (int i = 0; i < 512; i++) for (int i = 0; i < 512; i++)
{ {
int soff = (deswizzle ? table[i] : i); int soff = (deswizzle ? table[i] : i);
int doff = (deswizzle ? i : table[i]); int doff = (deswizzle ? i : table[i]);
int spx = src[soff >> 1] >> ((soff & 1) * 4) & 0xF; int spx = src[soff >> 1] >> ((soff & 1) * 4) & 0xF;
uint8* dpx = &dst[doff >> 1]; u8* dpx = &dst[doff >> 1];
int dshift = (doff & 1) * 4; int dshift = (doff & 1) * 4;
*dpx &= (0xF0 >> dshift); *dpx &= (0xF0 >> dshift);
*dpx |= (spx << dshift); *dpx |= (spx << dshift);
} }
} }
static void swizzleH(const uint8* table, uint32* dst, const uint8* src, int bpp, int shift) static void swizzleH(const u8* table, u32* dst, const u8* src, int bpp, int shift)
{ {
for (int i = 0; i < 64; i++) for (int i = 0; i < 64; i++)
{ {
@ -58,7 +58,7 @@ static void swizzleH(const uint8* table, uint32* dst, const uint8* src, int bpp,
} }
} }
static void expand16(uint32* dst, const uint16* src, const GIFRegTEXA& texa) static void expand16(u32* dst, const u16* src, const GIFRegTEXA& texa)
{ {
for (int i = 0; i < 128; i++) for (int i = 0; i < 128; i++)
{ {
@ -77,7 +77,7 @@ static void expand16(uint32* dst, const uint16* src, const GIFRegTEXA& texa)
} }
} }
static void expand8(uint32* dst, const uint8* src, const uint32* palette) static void expand8(u32* dst, const u8* src, const u32* palette)
{ {
for (int i = 0; i < 256; i++) for (int i = 0; i < 256; i++)
{ {
@ -85,7 +85,7 @@ static void expand8(uint32* dst, const uint8* src, const uint32* palette)
} }
} }
static void expand4(uint32* dst, const uint8* src, const uint32* palette) static void expand4(u32* dst, const u8* src, const u32* palette)
{ {
for (int i = 0; i < 512; i++) for (int i = 0; i < 512; i++)
{ {
@ -93,7 +93,7 @@ static void expand4(uint32* dst, const uint8* src, const uint32* palette)
} }
} }
static void expand4P(uint8* dst, const uint8* src) static void expand4P(u8* dst, const u8* src)
{ {
for (int i = 0; i < 512; i++) for (int i = 0; i < 512; i++)
{ {
@ -101,7 +101,7 @@ static void expand4P(uint8* dst, const uint8* src)
} }
} }
static void expandH(uint32* dst, const uint32* src, const uint32* palette, int shift, int mask) static void expandH(u32* dst, const u32* src, const u32* palette, int shift, int mask)
{ {
for (int i = 0; i < 64; i++) for (int i = 0; i < 64; i++)
{ {
@ -109,7 +109,7 @@ static void expandH(uint32* dst, const uint32* src, const uint32* palette, int s
} }
} }
static void expandHP(uint8* dst, const uint32* src, int shift, int mask) static void expandHP(u8* dst, const u32* src, int shift, int mask)
{ {
for (int i = 0; i < 64; i++) for (int i = 0; i < 64; i++)
{ {
@ -117,7 +117,7 @@ static void expandHP(uint8* dst, const uint32* src, int shift, int mask)
} }
} }
static std::string image2hex(const uint8* bin, int rows, int columns, int bpp) static std::string image2hex(const u8* bin, int rows, int columns, int bpp)
{ {
std::string out; std::string out;
const char* hex = "0123456789ABCDEF"; const char* hex = "0123456789ABCDEF";
@ -159,10 +159,10 @@ static std::string image2hex(const uint8* bin, int rows, int columns, int bpp)
struct TestData struct TestData
{ {
alignas(64) uint8 block[256]; alignas(64) u8 block[256];
alignas(64) uint8 output[256 * (32 / 4)]; alignas(64) u8 output[256 * (32 / 4)];
alignas(64) uint32 clut32[256]; alignas(64) u32 clut32[256];
alignas(64) uint64 clut64[256]; alignas(64) u64 clut64[256];
/// Get some input data with pixel values counting up from 0 /// Get some input data with pixel values counting up from 0
static TestData Linear() static TestData Linear()
@ -202,39 +202,39 @@ struct TestData
} }
}; };
static TestData swizzle(const uint8* table, TestData data, int bpp, bool deswizzle) static TestData swizzle(const u8* table, TestData data, int bpp, bool deswizzle)
{ {
swizzle(table, data.output, data.block, bpp, deswizzle); swizzle(table, data.output, data.block, bpp, deswizzle);
return data; return data;
} }
static TestData swizzle4(const uint16* table, TestData data, bool deswizzle) static TestData swizzle4(const u16* table, TestData data, bool deswizzle)
{ {
swizzle4(table, data.output, data.block, deswizzle); swizzle4(table, data.output, data.block, deswizzle);
return data; return data;
} }
static TestData swizzleH(const uint8* table, TestData data, int bpp, int shift) static TestData swizzleH(const u8* table, TestData data, int bpp, int shift)
{ {
swizzleH(table, reinterpret_cast<uint32*>(data.output), data.block, bpp, shift); swizzleH(table, reinterpret_cast<u32*>(data.output), data.block, bpp, shift);
return data; return data;
} }
static TestData expand16(TestData data, const GIFRegTEXA& texa) static TestData expand16(TestData data, const GIFRegTEXA& texa)
{ {
expand16(reinterpret_cast<uint32*>(data.output), reinterpret_cast<const uint16*>(data.block), texa); expand16(reinterpret_cast<u32*>(data.output), reinterpret_cast<const u16*>(data.block), texa);
return data; return data;
} }
static TestData expand8(TestData data) static TestData expand8(TestData data)
{ {
expand8(reinterpret_cast<uint32*>(data.output), data.block, data.clut32); expand8(reinterpret_cast<u32*>(data.output), data.block, data.clut32);
return data; return data;
} }
static TestData expand4(TestData data) static TestData expand4(TestData data)
{ {
expand4(reinterpret_cast<uint32*>(data.output), data.block, data.clut32); expand4(reinterpret_cast<u32*>(data.output), data.block, data.clut32);
return data; return data;
} }
@ -246,13 +246,13 @@ static TestData expand4P(TestData data)
static TestData expandH(TestData data, int shift, int mask) static TestData expandH(TestData data, int shift, int mask)
{ {
expandH(reinterpret_cast<uint32*>(data.output), reinterpret_cast<const uint32*>(data.block), data.clut32, shift, mask); expandH(reinterpret_cast<u32*>(data.output), reinterpret_cast<const u32*>(data.block), data.clut32, shift, mask);
return data; return data;
} }
static TestData expandHP(TestData data, int shift, int mask) static TestData expandHP(TestData data, int shift, int mask)
{ {
expandHP(data.output, reinterpret_cast<uint32*>(data.block), shift, mask); expandHP(data.output, reinterpret_cast<u32*>(data.block), shift, mask);
return data; return data;
} }
@ -318,7 +318,7 @@ TEST(ReadAndExpandTest, Read16AEM)
runTest([](TestData data) runTest([](TestData data)
{ {
// Actually test AEM // Actually test AEM
uint8 idx = data.block[0] >> 1; u8 idx = data.block[0] >> 1;
data.block[idx * 2 + 0] = 0; data.block[idx * 2 + 0] = 0;
data.block[idx * 2 + 1] = 0; data.block[idx * 2 + 1] = 0;
GIFRegTEXA texa = {0}; GIFRegTEXA texa = {0};