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

Various ZeroGS cleanups. A few things got mirrored from the dx version to the opengl version, and vice-versa. A few defines became inlined functions, and if statements became switch statements... 

git-svn-id: http://pcsx2-playground.googlecode.com/svn/trunk@574 a6443dda-0b58-4228-96e9-037be469359c
This commit is contained in:
arcum42 2009-01-10 11:30:47 +00:00 committed by Gregory Hainaut
parent 0c2f86a870
commit 84c7b1bdc3
15 changed files with 774 additions and 766 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

10
plugins/zerogs/dx/GS.h
View File

@ -70,7 +70,8 @@ struct Vector_16F
/////////////////////
// define when releasing
#define ZEROGS_CACHEDCLEAR // much better performance
// The only code that uses it is commented out!
//#define ZEROGS_CACHEDCLEAR // much better performance
//#define RELEASE_TO_PUBLIC
#if !defined(_RELEASE)
@ -79,6 +80,9 @@ struct Vector_16F
#define GS_LOG 0&&
#endif
#define ERROR_LOG __LogToConsole
#define DEBUG_LOG printf
#ifdef RELEASE_TO_PUBLIC
#define WARN_LOG 0&&
#define PRIM_LOG 0&&
@ -632,8 +636,8 @@ extern GSinternal gs;
extern FILE *gsLog;
void __Log(char *fmt, ...);
void ERROR_LOG(char *fmt, ...);
void __Log(const char *fmt, ...);
void __LogToConsole(const char *fmt, ...);
void LoadConfig();
void SaveConfig();

165
plugins/zerogs/dx/GSmain.cpp
View File

@ -127,8 +127,7 @@ void SysMessage(char *fmt, ...) {
#endif
#ifdef GS_LOG
void __Log(char *fmt, ...) {
void __Log(const char *fmt, ...) {
va_list list;
if (!conf.log) return;
@ -137,9 +136,8 @@ void __Log(char *fmt, ...) {
vfprintf(gsLog, fmt, list);
va_end(list);
}
#endif
void ERROR_LOG(char *fmt, ...) {
void __LogToConsole(const char *fmt, ...) {
va_list list;
printf("ZeroGS: ");
@ -193,7 +191,7 @@ void CALLBACK GSsetGameCRC(int crc, int options)
break;
}
printf("ZeroGS: Set game options: 0x%8.8x\n", g_GameSettings);
DEBUG_LOG("ZeroGS: Set game options: 0x%8.8x\n", g_GameSettings);
}
void CALLBACK GSsetFrameSkip(int frameskip)
@ -256,7 +254,7 @@ void CALLBACK GSreset() {
gs.q = 1;
}
void CALLBACK GSgifSoftReset(int mask)
void CALLBACK GSgifSoftReset(u32 mask)
{
if( mask & 1 ) memset(&gs.path1, 0, sizeof(gs.path1));
if( mask & 2 ) memset(&gs.path2, 0, sizeof(gs.path2));
@ -283,10 +281,7 @@ s32 CALLBACK GSinit()
#endif
GSreset();
#ifdef GS_LOG
GS_LOG("GSinit ok\n");
#endif
return 0;
}
@ -318,7 +313,7 @@ LRESULT WINAPI MsgProc( HWND hWnd, UINT msg, WPARAM wParam, LPARAM lParam )
case WM_ACTIVATE:
if( wParam != WA_INACTIVE ) {
//printf("restoring device\n");
//DEBUG_LOG("restoring device\n");
ZeroGS::Restore();
}
@ -350,9 +345,7 @@ s32 CALLBACK GSopen(void *pDsp, char *Title, int multithread) {
g_GSMultiThreaded = multithread;
#ifdef GS_LOG
GS_LOG("GSopen\n");
#endif
#ifdef _DEBUG
g_hCurrentThread = GetCurrentThread();
@ -389,12 +382,12 @@ s32 CALLBACK GSopen(void *pDsp, char *Title, int multithread) {
if( pDsp != NULL )
*(int*)pDsp = (int)GShwnd;
printf("creating zerogs\n");
DEBUG_LOG("creating zerogs\n");
//if (conf.record) recOpen();
if( FAILED(ZeroGS::Create(conf.width, conf.height)) )
return -1;
printf("init zerogs\n");
DEBUG_LOG("initialization successful\n");
if( FAILED(ZeroGS::InitDeviceObjects()) )
return -1;
@ -430,9 +423,7 @@ s32 CALLBACK GSopen(void *pDsp, char *Title, int multithread) {
luPerfFreq.QuadPart = 1;
#endif
#ifdef GS_LOG
GS_LOG("GSopen ok\n");
#endif
gs.path1.mode = 0;
gs.path2.mode = 0;
@ -508,9 +499,7 @@ float fFPS = 0;
void CALLBACK GSvsync(int interlace)
{
#ifdef GS_LOG
GS_LOG("\nGSvsync\n\n");
#endif
static u32 dwTime = timeGetTime();
static int nToNextUpdate = 1;
@ -694,22 +683,21 @@ void CALLBACK GSvsync(int interlace)
}
void GIFtag(pathInfo *path, u32 *data) {
path->tag.nloop = data[0] & 0x7fff;
path->tag.eop = (data[0] >> 15) & 0x1;
u32 tagpre = (data[1] >> 14) & 0x1;
u32 tagprim = (data[1] >> 15) & 0x7ff;
u32 tagflg = (data[1] >> 26) & 0x3;
path->tag.nreg = (data[1] >> 28)<<2;
path->tag.nloop = data[0] & 0x7fff;
path->tag.eop = (data[0] >> 15) & 0x1;
u32 tagpre = (data[1] >> 14) & 0x1;
u32 tagprim = (data[1] >> 15) & 0x7ff;
u32 tagflg = (data[1] >> 26) & 0x3;
path->tag.nreg = (data[1] >> 28)<<2;
if (path->tag.nreg == 0) path->tag.nreg = 64;
gs.q = 1;
#ifdef GS_LOG
// GS_LOG("GIFtag: %8.8lx_%8.8lx_%8.8lx_%8.8lx: EOP=%d, NLOOP=%x, FLG=%x, NREG=%d, PRE=%d\n",
// data[3], data[2], data[1], data[0],
// path->tag.eop, path->tag.nloop, tagflg, path->tag.nreg, tagpre);
#endif
path->mode = tagflg+1;
@ -759,9 +747,6 @@ static int nPath3Hack = 0;
void CALLBACK GSgetLastTag(u64* ptag)
{
// int mode = s_pLastPath->mode > 0 ? s_pLastPath->mode-1 : 0;
// *(u32*)ptag = s_pLastPath->tag.nloop|(s_pLastPath->tag.eop<<15);
// *((u32*)ptag+1) = (mode<<26)|(s_pLastPath->regn<<28);
*(u32*)ptag = nPath3Hack;
nPath3Hack = 0;
}
@ -785,20 +770,23 @@ void _GSgifTransfer(pathInfo *path, u32 *pMem, u32 size)
while(size > 0)
{
//LOG(_T("Transfer(%08x, %d) START\n"), pMem, size);
if(path->tag.nloop == 0) {
if (path->tag.nloop == 0)
{
GIFtag(path, pMem);
pMem+= 4;
size--;
if( (g_GameSettings & GAME_PATH3HACK) && path == &gs.path3 && gs.path3.tag.eop )
if ((g_GameSettings & GAME_PATH3HACK) && path == &gs.path3 && gs.path3.tag.eop)
nPath3Hack = 1;
if( path == &gs.path1 ) {
if (path == &gs.path1)
{
// if too much data, just ignore
if( path->tag.nloop * (path->tag.nreg / 4) > (int)size * (path->mode==2?2:1)) {
if (path->tag.nloop * (path->tag.nreg / 4) > (int)size * (path->mode==2?2:1))
{
static int lasttime = 0;
if( timeGetTime() - lasttime > 5000 ) {
if( timeGetTime() - lasttime > 5000 )
{
ERROR_LOG("VU1 too much data, ignore if gfx are fine\n");
lasttime = timeGetTime();
}
@ -806,13 +794,16 @@ void _GSgifTransfer(pathInfo *path, u32 *pMem, u32 size)
return;
}
if( path->mode == 1 ) {
if (path->mode == 1)
{
// check if 0xb is in any reg, if yes, exit (kh2)
for(int i = 0; i < path->tag.nreg; i += 4) {
if( ((path->regs >> i)&0xf) == 11 ) {
for(int i = 0; i < path->tag.nreg; i += 4)
{
if (((path->regs >> i)&0xf) == 11)
{
static int lasttime = 0;
if( timeGetTime() - lasttime > 5000 ) {
if( timeGetTime() - lasttime > 5000 )
{
ERROR_LOG("Invalid unpack type\n");
lasttime = timeGetTime();
}
@ -823,11 +814,13 @@ void _GSgifTransfer(pathInfo *path, u32 *pMem, u32 size)
}
}
if(path->tag.nloop == 0 ) {
//bAfter0Tag = 1;
if( path == &gs.path1 ) {
if(path->tag.nloop == 0 )
{
if( path == &gs.path1 )
{
// ffx hack
if( g_GameSettings & GAME_FFXHACK ) {
if( g_GameSettings & GAME_FFXHACK )
{
if( path->tag.eop )
return;
continue;
@ -836,32 +829,19 @@ void _GSgifTransfer(pathInfo *path, u32 *pMem, u32 size)
return;
}
if( !path->tag.eop ) {
//printf("contuing from eop\n");
if( !path->tag.eop )
{
//DEBUG_LOG("continuing from eop\n");
continue;
}
break;
}
// else {
// if( bAfter0Tag ) {
// // hack!! if all 0s, then break
// if( pMem[-3] == 0 && pMem[-2] == 0 ) {
// path->tag.nloop = 0;
// return;
// }
// }
// }
}
switch(path->mode) {
case 1: // PACKED
{
// if( path->tag.nreg == 9 && gs.regs == 0x412412412L && gs.primC == 0) {
// // draw 3 points
// }
//__Log("%8.8x%8.8x %d\n", ((u32*)&gs.regs)[1], *(u32*)&gs.regs, path->tag.nreg/4);
assert( path->tag.nloop > 0 );
for(; size > 0; size--, pMem += 4)
{
@ -870,9 +850,11 @@ void _GSgifTransfer(pathInfo *path, u32 *pMem, u32 size)
g_GIFPackedRegHandlers[reg](pMem);
path->regn += 4;
if (path->tag.nreg == path->regn) {
if (path->tag.nreg == path->regn)
{
path->regn = 0;
if( path->tag.nloop-- <= 1 ) {
if( path->tag.nloop-- <= 1 )
{
size--;
pMem += 4;
break;
@ -883,7 +865,7 @@ void _GSgifTransfer(pathInfo *path, u32 *pMem, u32 size)
}
case 2: // REGLIST
{
//__Log("%8.8x%8.8x %d L\n", ((u32*)&gs.regs)[1], *(u32*)&gs.regs, path->tag.nreg/4);
//GS_LOG("%8.8x%8.8x %d L\n", ((u32*)&gs.regs)[1], *(u32*)&gs.regs, path->tag.nreg/4);
assert( path->tag.nloop > 0 );
size *= 2;
for(; size > 0; pMem+= 2, size--)
@ -891,9 +873,11 @@ void _GSgifTransfer(pathInfo *path, u32 *pMem, u32 size)
int reg = (int)((path->regs >> path->regn) & 0xf);
g_GIFRegHandlers[reg](pMem);
path->regn += 4;
if (path->tag.nreg == path->regn) {
if (path->tag.nreg == path->regn)
{
path->regn = 0;
if( path->tag.nloop-- <= 1 ) {
if( path->tag.nloop-- <= 1 )
{
size--;
pMem += 2;
break;
@ -912,9 +896,12 @@ void _GSgifTransfer(pathInfo *path, u32 *pMem, u32 size)
{
int process = min((int)size, path->tag.nloop);
if( process > 0 ) {
if( gs.imageTransfer ) ZeroGS::TransferLocalHost(pMem, process);
else ZeroGS::TransferHostLocal(pMem, process*4);
if( process > 0 )
{
if (gs.imageTransfer)
ZeroGS::TransferLocalHost(pMem, process);
else
ZeroGS::TransferHostLocal(pMem, process*4);
path->tag.nloop -= process;
pMem += process*4; size -= process;
@ -923,7 +910,8 @@ void _GSgifTransfer(pathInfo *path, u32 *pMem, u32 size)
}
break;
}
else {
else
{
// simulate
int process = min((int)size, path->tag.nloop);
path->tag.nloop -= process;
@ -933,11 +921,8 @@ void _GSgifTransfer(pathInfo *path, u32 *pMem, u32 size)
break;
}
default: // GIF_IMAGE
GS_LOG(_T("*** WARNING **** Unexpected GIFTag flag\n"));
GS_LOG("*** WARNING **** Unexpected GIFTag flag\n");
assert(0);
// ZeroGS::TransferLocalHost(pMem, size);
// pMem+= size*4; path->tag.nloop-= size;
// size = 0; path->mode = 0;
path->tag.nloop = 0;
break;
@ -950,29 +935,15 @@ void _GSgifTransfer(pathInfo *path, u32 *pMem, u32 size)
void CALLBACK GSgifTransfer2(u32 *pMem, u32 size)
{
#ifdef GS_LOG
//GS_LOG("GSgifTransfer2 size = %lx (mode %d, gs.path2.tag.nloop = %d)\n", size, gs.path2.mode, gs.path2.tag.nloop);
#endif
// if(!g_GSMultiThreaded)
// CSR->FINISH = 0;
//DVProfileFunc _pf("Transf2");
//assert( ((u32)pMem & 0xf) == 0 );
_GSgifTransfer(&gs.path2, pMem, size);
}
void CALLBACK GSgifTransfer3(u32 *pMem, u32 size)
{
#ifdef GS_LOG
//GS_LOG("GSgifTransfer3 size = %lx (mode %d, gs.path3.tag.nloop = %d)\n", size, gs.path3.mode, gs.path3.tag.nloop);
#endif
// if(!g_GSMultiThreaded)
// CSR->FINISH = 0;
//DVProfileFunc _pf("Transf3");
//assert( ((u32)pMem & 0xf) == 0 );
nPath3Hack = 0;
_GSgifTransfer(&gs.path3, pMem, size);
}
@ -982,20 +953,13 @@ void CALLBACK GSgifTransfer1(u32 *pMem, u32 addr)
{
pathInfo *path = &gs.path1;
#ifdef GS_LOG
//GS_LOG("GSgifTransfer1 0x%x (mode %d)\n", addr, path->mode);
#endif
addr &= 0x3fff;
#ifdef _DEBUG
PRIM_LOG("count: %d\n", count);
count++;
// for(int i = addr; i < 0x4000; i += 16 ) {
// u32* mem = (u32*)((u8*)pMem+i);
// PRIM_LOG("%x: %x %x %x %x\n", i, mem[0], mem[1], mem[2], mem[3]);
// }
#endif
gs.path1.tag.nloop = 0;
@ -1006,25 +970,20 @@ void CALLBACK GSgifTransfer1(u32 *pMem, u32 addr)
assert( (addr&0xf) == 0 ); //BUG
gs.path1.tag.nloop = 0;
ERROR_LOG("Transfer1 - 2\n");
//_GSgifTransfer(&gs.path1, (u32*)((u8*)pMem+0x4000-addr), addr/16);
return;
}
}
void CALLBACK GSreadFIFO(u64 *pMem)
{
#ifdef GS_LOG
//GS_LOG("GSreadFIFO\n");
#endif
ZeroGS::TransferLocalHost((u32*)pMem, 1);
}
void CALLBACK GSreadFIFO2(u64 *pMem, int qwc)
{
#ifdef GS_LOG
//GS_LOG("GSreadFIFO2\n");
#endif
ZeroGS::TransferLocalHost((u32*)pMem, qwc);
}
@ -1037,14 +996,14 @@ int CALLBACK GSsetupRecording(int start, void* pData)
if( !ZeroGS::StartCapture() )
return 0;
conf.options |= GSOPTION_CAPTUREAVI;
printf("ZeroGS: started recording at zerogs.avi\n");
DEBUG_LOG("ZeroGS: started recording at zerogs.avi\n");
}
else {
if( !(conf.options & GSOPTION_CAPTUREAVI) )
return 1;
conf.options &= ~GSOPTION_CAPTUREAVI;
ZeroGS::StopCapture();
printf("ZeroGS: stopped recording\n");
DEBUG_LOG("ZeroGS: stopped recording\n");
}
return 1;
@ -1053,7 +1012,7 @@ int CALLBACK GSsetupRecording(int start, void* pData)
s32 CALLBACK GSfreeze(int mode, freezeData *data) {
if (mode == FREEZE_LOAD) {
if( !ZeroGS::Load(data->data) )
printf("GS: Bad load format!");
DEBUG_LOG("GS: Bad load format!");
g_nRealFrame += 100;
} else if (mode == FREEZE_SAVE) {

30
plugins/zerogs/dx/Mem.cpp
View File

@ -242,7 +242,7 @@ u32 g_pageTable8[64][128];
u32 g_pageTable4[128][128];
BLOCK m_Blocks[0x40]; // do so blocks are indexable
static __declspec(align(16)) u32 tempblock[64];
static PCSX2_ALIGNED16(u32 tempblock[64]);
#define DSTPSM gs.dstbuf.psm
@ -503,13 +503,13 @@ End: \
#define SwizzleBlock4HLu SwizzleBlock4HL
// ------------------------
// | Y |
// | Y |
// ------------------------
// | block | |
// | block | |
// | aligned area | X |
// | | |
// | | |
// ------------------------
// | Y |
// | Y |
// ------------------------
#define DEFINE_TRANSFERLOCAL(psm, T, widthlimit, blockbits, blockwidth, blockheight, TransSfx, SwizzleBlock) \
int TransferHostLocal##psm(const void* pbyMem, u32 nQWordSize) \
@ -521,16 +521,18 @@ int TransferHostLocal##psm(const void* pbyMem, u32 nQWordSize) \
int nSize = nQWordSize*4*2/TRANSMIT_PITCH##TransSfx(2, T); \
nSize = min(nSize, gs.imageWnew * gs.imageHnew); \
\
int pitch, area, fracX; \
int endY = ROUND_UPPOW2(i, blockheight); \
int alignedY = ROUND_DOWNPOW2(gs.imageEndY, blockheight); \
int alignedX = ROUND_DOWNPOW2(gs.imageEndX, blockwidth); \
bool bCanAlign = MOD_POW2(gs.trxpos.dx, blockwidth) == 0 && (j == gs.trxpos.dx) && (alignedY > endY) && alignedX > gs.trxpos.dx; \
bool bAligned, bCanAlign = MOD_POW2(gs.trxpos.dx, blockwidth) == 0 && (j == gs.trxpos.dx) && (alignedY > endY) && alignedX > gs.trxpos.dx; \
\
if( (gs.imageEndX-gs.trxpos.dx)%widthlimit ) { \
/* hack */ \
if( abs((int)nSize - (gs.imageEndY-i)*(gs.imageEndX-gs.trxpos.dx)+(j-gs.trxpos.dx)) <= widthlimit ) { \
/* don't transfer */ \
/*printf("bad texture %s: %d %d %d\n", #psm, gs.trxpos.dx, gs.imageEndX, nQWordSize);*/ \
int testwidth = (int)nSize - (gs.imageEndY-i)*(gs.imageEndX-gs.trxpos.dx)+(j-gs.trxpos.dx); \
if ((testwidth <= widthlimit) && (testwidth >= -widthlimit)) { \
/* Don't transfer */ \
/*DEBUG_LOG("bad texture %s: %d %d %d\n", #psm, gs.trxpos.dx, gs.imageEndX, nQWordSize);*/ \
gs.imageTransfer = -1; \
} \
bCanAlign = false; \
@ -559,12 +561,12 @@ int TransferHostLocal##psm(const void* pbyMem, u32 nQWordSize) \
assert( MOD_POW2(i, blockheight) == 0 && j == gs.trxpos.dx); \
\
/* can align! */ \
int pitch = gs.imageEndX-gs.trxpos.dx; \
int area = pitch*blockheight; \
int fracX = gs.imageEndX-alignedX; \
pitch = gs.imageEndX-gs.trxpos.dx; \
area = pitch*blockheight; \
fracX = gs.imageEndX-alignedX; \
\
/* on top of checking whether pbuf is alinged, make sure that the width is at least aligned to its limits (due to bugs in pcsx2) */ \
bool bAligned = !((u32)pbuf & 0xf) && (TRANSMIT_PITCH##TransSfx(pitch, T)&0xf) == 0; \
bAligned = !((uptr)pbuf & 0xf) && (TRANSMIT_PITCH##TransSfx(pitch, T)&0xf) == 0; \
\
/* transfer aligning to blocks */ \
for(; i < alignedY && nSize >= area; i += blockheight, nSize -= area) { \
@ -638,7 +640,7 @@ DEFINE_TRANSFERLOCAL(4HH, u8, 8, 32, 8, 8, _4, SwizzleBlock4HH);
// /* hack */
// if( abs((int)nSize - (gs.imageEndY-i)*(gs.imageEndX-gs.trxpos.dx)+(j-gs.trxpos.dx)) <= widthlimit ) {
// /* don't transfer */
// /*printf("bad texture %s: %d %d %d\n", #psm, gs.trxpos.dx, gs.imageEndX, nQWordSize);*/
// /*DEBUG_LOG("bad texture %s: %d %d %d\n", #psm, gs.trxpos.dx, gs.imageEndX, nQWordSize);*/
// gs.imageTransfer = -1;
// }
// bCanAlign = false;

136
plugins/zerogs/dx/Mem.h
View File

@ -83,7 +83,7 @@ extern u32 g_pageTable16SZ[64][64];
extern u32 g_pageTable8[64][128];
extern u32 g_pageTable4[128][128];
__forceinline u32 getPixelAddress32(int x, int y, u32 bp, u32 bw) {
static __forceinline u32 getPixelAddress32(int x, int y, u32 bp, u32 bw) {
u32 basepage = ((y>>5) * (bw>>6)) + (x>>6);
u32 word = bp * 64 + basepage * 2048 + g_pageTable32[y&31][x&63];
//assert (word < 0x100000);
@ -91,7 +91,7 @@ __forceinline u32 getPixelAddress32(int x, int y, u32 bp, u32 bw) {
return word;
}
__forceinline u32 getPixelAddress32_0(int x, int y, u32 bw) {
static __forceinline u32 getPixelAddress32_0(int x, int y, u32 bw) {
u32 basepage = ((y>>5) * (bw>>6)) + (x>>6);
u32 word = basepage * 2048 + g_pageTable32[y&31][x&63];
//assert (word < 0x100000);
@ -108,7 +108,7 @@ __forceinline u32 getPixelAddress32_0(int x, int y, u32 bw) {
#define getPixelAddress4HH getPixelAddress32
#define getPixelAddress4HH_0 getPixelAddress32_0
__forceinline u32 getPixelAddress16(int x, int y, u32 bp, u32 bw) {
static __forceinline u32 getPixelAddress16(int x, int y, u32 bp, u32 bw) {
u32 basepage = ((y>>6) * (bw>>6)) + (x>>6);
u32 word = bp * 128 + basepage * 4096 + g_pageTable16[y&63][x&63];
//assert (word < 0x200000);
@ -116,7 +116,7 @@ __forceinline u32 getPixelAddress16(int x, int y, u32 bp, u32 bw) {
return word;
}
__forceinline u32 getPixelAddress16_0(int x, int y, u32 bw) {
static __forceinline u32 getPixelAddress16_0(int x, int y, u32 bw) {
u32 basepage = ((y>>6) * (bw>>6)) + (x>>6);
u32 word = basepage * 4096 + g_pageTable16[y&63][x&63];
//assert (word < 0x200000);
@ -124,7 +124,7 @@ __forceinline u32 getPixelAddress16_0(int x, int y, u32 bw) {
return word;
}
__forceinline u32 getPixelAddress16S(int x, int y, u32 bp, u32 bw) {
static __forceinline u32 getPixelAddress16S(int x, int y, u32 bp, u32 bw) {
u32 basepage = ((y>>6) * (bw>>6)) + (x>>6);
u32 word = bp * 128 + basepage * 4096 + g_pageTable16S[y&63][x&63];
//assert (word < 0x200000);
@ -132,7 +132,7 @@ __forceinline u32 getPixelAddress16S(int x, int y, u32 bp, u32 bw) {
return word;
}
__forceinline u32 getPixelAddress16S_0(int x, int y, u32 bw) {
static __forceinline u32 getPixelAddress16S_0(int x, int y, u32 bw) {
u32 basepage = ((y>>6) * (bw>>6)) + (x>>6);
u32 word = basepage * 4096 + g_pageTable16S[y&63][x&63];
//assert (word < 0x200000);
@ -140,7 +140,7 @@ __forceinline u32 getPixelAddress16S_0(int x, int y, u32 bw) {
return word;
}
__forceinline u32 getPixelAddress8(int x, int y, u32 bp, u32 bw) {
static __forceinline u32 getPixelAddress8(int x, int y, u32 bp, u32 bw) {
u32 basepage = ((y>>6) * ((bw+127)>>7)) + (x>>7);
u32 word = bp * 256 + basepage * 8192 + g_pageTable8[y&63][x&127];
//assert (word < 0x400000);
@ -148,7 +148,7 @@ __forceinline u32 getPixelAddress8(int x, int y, u32 bp, u32 bw) {
return word;
}
__forceinline u32 getPixelAddress8_0(int x, int y, u32 bw) {
static __forceinline u32 getPixelAddress8_0(int x, int y, u32 bw) {
u32 basepage = ((y>>6) * ((bw+127)>>7)) + (x>>7);
u32 word = basepage * 8192 + g_pageTable8[y&63][x&127];
//assert (word < 0x400000);
@ -156,7 +156,7 @@ __forceinline u32 getPixelAddress8_0(int x, int y, u32 bw) {
return word;
}
__forceinline u32 getPixelAddress4(int x, int y, u32 bp, u32 bw) {
static __forceinline u32 getPixelAddress4(int x, int y, u32 bp, u32 bw) {
u32 basepage = ((y>>7) * ((bw+127)>>7)) + (x>>7);
u32 word = bp * 512 + basepage * 16384 + g_pageTable4[y&127][x&127];
//assert (word < 0x800000);
@ -164,7 +164,7 @@ __forceinline u32 getPixelAddress4(int x, int y, u32 bp, u32 bw) {
return word;
}
__forceinline u32 getPixelAddress4_0(int x, int y, u32 bw) {
static __forceinline u32 getPixelAddress4_0(int x, int y, u32 bw) {
u32 basepage = ((y>>7) * ((bw+127)>>7)) + (x>>7);
u32 word = basepage * 16384 + g_pageTable4[y&127][x&127];
//assert (word < 0x800000);
@ -172,7 +172,7 @@ __forceinline u32 getPixelAddress4_0(int x, int y, u32 bw) {
return word;
}
__forceinline u32 getPixelAddress32Z(int x, int y, u32 bp, u32 bw) {
static __forceinline u32 getPixelAddress32Z(int x, int y, u32 bp, u32 bw) {
u32 basepage = ((y>>5) * (bw>>6)) + (x>>6);
u32 word = bp * 64 + basepage * 2048 + g_pageTable32Z[y&31][x&63];
//assert (word < 0x100000);
@ -180,7 +180,7 @@ __forceinline u32 getPixelAddress32Z(int x, int y, u32 bp, u32 bw) {
return word;
}
__forceinline u32 getPixelAddress32Z_0(int x, int y, u32 bw) {
static __forceinline u32 getPixelAddress32Z_0(int x, int y, u32 bw) {
u32 basepage = ((y>>5) * (bw>>6)) + (x>>6);
u32 word = basepage * 2048 + g_pageTable32Z[y&31][x&63];
//assert (word < 0x100000);
@ -191,7 +191,7 @@ __forceinline u32 getPixelAddress32Z_0(int x, int y, u32 bw) {
#define getPixelAddress24Z getPixelAddress32Z
#define getPixelAddress24Z_0 getPixelAddress32Z_0
__forceinline u32 getPixelAddress16Z(int x, int y, u32 bp, u32 bw) {
static __forceinline u32 getPixelAddress16Z(int x, int y, u32 bp, u32 bw) {
u32 basepage = ((y>>6) * (bw>>6)) + (x>>6);
u32 word = bp * 128 + basepage * 4096 + g_pageTable16Z[y&63][x&63];
//assert (word < 0x200000);
@ -199,7 +199,7 @@ __forceinline u32 getPixelAddress16Z(int x, int y, u32 bp, u32 bw) {
return word;
}
__forceinline u32 getPixelAddress16Z_0(int x, int y, u32 bw) {
static __forceinline u32 getPixelAddress16Z_0(int x, int y, u32 bw) {
u32 basepage = ((y>>6) * (bw>>6)) + (x>>6);
u32 word = basepage * 4096 + g_pageTable16Z[y&63][x&63];
//assert (word < 0x200000);
@ -207,7 +207,7 @@ __forceinline u32 getPixelAddress16Z_0(int x, int y, u32 bw) {
return word;
}
__forceinline u32 getPixelAddress16SZ(int x, int y, u32 bp, u32 bw) {
static __forceinline u32 getPixelAddress16SZ(int x, int y, u32 bp, u32 bw) {
u32 basepage = ((y>>6) * (bw>>6)) + (x>>6);
u32 word = bp * 128 + basepage * 4096 + g_pageTable16SZ[y&63][x&63];
//assert (word < 0x200000);
@ -215,7 +215,7 @@ __forceinline u32 getPixelAddress16SZ(int x, int y, u32 bp, u32 bw) {
return word;
}
__forceinline u32 getPixelAddress16SZ_0(int x, int y, u32 bw) {
static __forceinline u32 getPixelAddress16SZ_0(int x, int y, u32 bw) {
u32 basepage = ((y>>6) * (bw>>6)) + (x>>6);
u32 word = basepage * 4096 + g_pageTable16SZ[y&63][x&63];
//assert (word < 0x200000);
@ -223,95 +223,95 @@ __forceinline u32 getPixelAddress16SZ_0(int x, int y, u32 bw) {
return word;
}
__forceinline void writePixel32(void* pmem, int x, int y, u32 pixel, u32 bp, u32 bw) {
static __forceinline void writePixel32(void* pmem, int x, int y, u32 pixel, u32 bp, u32 bw) {
((u32*)pmem)[getPixelAddress32(x, y, bp, bw)] = pixel;
}
__forceinline void writePixel24(void* pmem, int x, int y, u32 pixel, u32 bp, u32 bw) {
static __forceinline void writePixel24(void* pmem, int x, int y, u32 pixel, u32 bp, u32 bw) {
u8 *buf = (u8*)&((u32*)pmem)[getPixelAddress32(x, y, bp, bw)];
u8 *pix = (u8*)&pixel;
buf[0] = pix[0]; buf[1] = pix[1]; buf[2] = pix[2];
}
__forceinline void writePixel16(void* pmem, int x, int y, u32 pixel, u32 bp, u32 bw) {
static __forceinline void writePixel16(void* pmem, int x, int y, u32 pixel, u32 bp, u32 bw) {
((u16*)pmem)[getPixelAddress16(x, y, bp, bw)] = pixel;
}
__forceinline void writePixel16S(void* pmem, int x, int y, u32 pixel, u32 bp, u32 bw) {
static __forceinline void writePixel16S(void* pmem, int x, int y, u32 pixel, u32 bp, u32 bw) {
((u16*)pmem)[getPixelAddress16S(x, y, bp, bw)] = pixel;
}
__forceinline void writePixel8(void* pmem, int x, int y, u32 pixel, u32 bp, u32 bw) {
static __forceinline void writePixel8(void* pmem, int x, int y, u32 pixel, u32 bp, u32 bw) {
((u8*)pmem)[getPixelAddress8(x, y, bp, bw)] = pixel;
}
__forceinline void writePixel8H(void* pmem, int x, int y, u32 pixel, u32 bp, u32 bw) {
static __forceinline void writePixel8H(void* pmem, int x, int y, u32 pixel, u32 bp, u32 bw) {
((u8*)pmem)[4*getPixelAddress32(x, y, bp, bw)+3] = pixel;
}
__forceinline void writePixel4(void* pmem, int x, int y, u32 pixel, u32 bp, u32 bw) {
static __forceinline void writePixel4(void* pmem, int x, int y, u32 pixel, u32 bp, u32 bw) {
u32 addr = getPixelAddress4(x, y, bp, bw);
u8 pix = ((u8*)pmem)[addr/2];
if (addr & 0x1) ((u8*)pmem)[addr/2] = (pix & 0x0f) | (pixel << 4);
else ((u8*)pmem)[addr/2] = (pix & 0xf0) | (pixel);
}
__forceinline void writePixel4HL(void* pmem, int x, int y, u32 pixel, u32 bp, u32 bw) {
static __forceinline void writePixel4HL(void* pmem, int x, int y, u32 pixel, u32 bp, u32 bw) {
u8 *p = (u8*)pmem + 4*getPixelAddress4HL(x, y, bp, bw)+3;
*p = (*p & 0xf0) | pixel;
}
__forceinline void writePixel4HH(void* pmem, int x, int y, u32 pixel, u32 bp, u32 bw) {
static __forceinline void writePixel4HH(void* pmem, int x, int y, u32 pixel, u32 bp, u32 bw) {
u8 *p = (u8*)pmem + 4*getPixelAddress4HH(x, y, bp, bw)+3;
*p = (*p & 0x0f) | (pixel<<4);
}
__forceinline void writePixel32Z(void* pmem, int x, int y, u32 pixel, u32 bp, u32 bw) {
static __forceinline void writePixel32Z(void* pmem, int x, int y, u32 pixel, u32 bp, u32 bw) {
((u32*)pmem)[getPixelAddress32Z(x, y, bp, bw)] = pixel;
}
__forceinline void writePixel24Z(void* pmem, int x, int y, u32 pixel, u32 bp, u32 bw) {
static __forceinline void writePixel24Z(void* pmem, int x, int y, u32 pixel, u32 bp, u32 bw) {
u8 *buf = (u8*)pmem + 4*getPixelAddress32Z(x, y, bp, bw);
u8 *pix = (u8*)&pixel;
buf[0] = pix[0]; buf[1] = pix[1]; buf[2] = pix[2];
}
__forceinline void writePixel16Z(void* pmem, int x, int y, u32 pixel, u32 bp, u32 bw) {
static __forceinline void writePixel16Z(void* pmem, int x, int y, u32 pixel, u32 bp, u32 bw) {
((u16*)pmem)[getPixelAddress16Z(x, y, bp, bw)] = pixel;
}
__forceinline void writePixel16SZ(void* pmem, int x, int y, u32 pixel, u32 bp, u32 bw) {
static __forceinline void writePixel16SZ(void* pmem, int x, int y, u32 pixel, u32 bp, u32 bw) {
((u16*)pmem)[getPixelAddress16SZ(x, y, bp, bw)] = pixel;
}
///////////////
__forceinline u32 readPixel32(const void* pmem, int x, int y, u32 bp, u32 bw) {
static __forceinline u32 readPixel32(const void* pmem, int x, int y, u32 bp, u32 bw) {
return ((const u32*)pmem)[getPixelAddress32(x, y, bp, bw)];
}
__forceinline u32 readPixel24(const void* pmem, int x, int y, u32 bp, u32 bw) {
static __forceinline u32 readPixel24(const void* pmem, int x, int y, u32 bp, u32 bw) {
return ((const u32*)pmem)[getPixelAddress32(x, y, bp, bw)] & 0xffffff;
}
__forceinline u32 readPixel16(const void* pmem, int x, int y, u32 bp, u32 bw) {
static __forceinline u32 readPixel16(const void* pmem, int x, int y, u32 bp, u32 bw) {
return ((const u16*)pmem)[getPixelAddress16(x, y, bp, bw)];
}
__forceinline u32 readPixel16S(const void* pmem, int x, int y, u32 bp, u32 bw) {
static __forceinline u32 readPixel16S(const void* pmem, int x, int y, u32 bp, u32 bw) {
return ((const u16*)pmem)[getPixelAddress16S(x, y, bp, bw)];
}
__forceinline u32 readPixel8(const void* pmem, int x, int y, u32 bp, u32 bw) {
static __forceinline u32 readPixel8(const void* pmem, int x, int y, u32 bp, u32 bw) {
return ((const u8*)pmem)[getPixelAddress8(x, y, bp, bw)];
}
__forceinline u32 readPixel8H(const void* pmem, int x, int y, u32 bp, u32 bw) {
static __forceinline u32 readPixel8H(const void* pmem, int x, int y, u32 bp, u32 bw) {
return ((const u8*)pmem)[4*getPixelAddress32(x, y, bp, bw) + 3];
}
__forceinline u32 readPixel4(const void* pmem, int x, int y, u32 bp, u32 bw) {
static __forceinline u32 readPixel4(const void* pmem, int x, int y, u32 bp, u32 bw) {
u32 addr = getPixelAddress4(x, y, bp, bw);
u8 pix = ((const u8*)pmem)[addr/2];
if (addr & 0x1)
@ -319,31 +319,31 @@ __forceinline u32 readPixel4(const void* pmem, int x, int y, u32 bp, u32 bw) {
else return pix & 0xf;
}
__forceinline u32 readPixel4HL(const void* pmem, int x, int y, u32 bp, u32 bw) {
static __forceinline u32 readPixel4HL(const void* pmem, int x, int y, u32 bp, u32 bw) {
const u8 *p = (const u8*)pmem+4*getPixelAddress4HL(x, y, bp, bw)+3;
return *p & 0x0f;
}
__forceinline u32 readPixel4HH(const void* pmem, int x, int y, u32 bp, u32 bw) {
static __forceinline u32 readPixel4HH(const void* pmem, int x, int y, u32 bp, u32 bw) {
const u8 *p = (const u8*)pmem+4*getPixelAddress4HH(x, y, bp, bw) + 3;
return *p >> 4;
}
///////////////
__forceinline u32 readPixel32Z(const void* pmem, int x, int y, u32 bp, u32 bw) {
static __forceinline u32 readPixel32Z(const void* pmem, int x, int y, u32 bp, u32 bw) {
return ((const u32*)pmem)[getPixelAddress32Z(x, y, bp, bw)];
}
__forceinline u32 readPixel24Z(const void* pmem, int x, int y, u32 bp, u32 bw) {
static __forceinline u32 readPixel24Z(const void* pmem, int x, int y, u32 bp, u32 bw) {
return ((const u32*)pmem)[getPixelAddress32Z(x, y, bp, bw)] & 0xffffff;
}
__forceinline u32 readPixel16Z(const void* pmem, int x, int y, u32 bp, u32 bw) {
static __forceinline u32 readPixel16Z(const void* pmem, int x, int y, u32 bp, u32 bw) {
return ((const u16*)pmem)[getPixelAddress16Z(x, y, bp, bw)];
}
__forceinline u32 readPixel16SZ(const void* pmem, int x, int y, u32 bp, u32 bw) {
static __forceinline u32 readPixel16SZ(const void* pmem, int x, int y, u32 bp, u32 bw) {
return ((const u16*)pmem)[getPixelAddress16SZ(x, y, bp, bw)];
}
@ -351,11 +351,11 @@ __forceinline u32 readPixel16SZ(const void* pmem, int x, int y, u32 bp, u32 bw)
// Functions that take 0 bps //
///////////////////////////////
__forceinline void writePixel32_0(void* pmem, int x, int y, u32 pixel, u32 bw) {
static __forceinline void writePixel32_0(void* pmem, int x, int y, u32 pixel, u32 bw) {
((u32*)pmem)[getPixelAddress32_0(x, y, bw)] = pixel;
}
__forceinline void writePixel24_0(void* pmem, int x, int y, u32 pixel, u32 bw) {
static __forceinline void writePixel24_0(void* pmem, int x, int y, u32 pixel, u32 bw) {
u8 *buf = (u8*)&((u32*)pmem)[getPixelAddress32_0(x, y, bw)];
u8 *pix = (u8*)&pixel;
#if defined(_MSC_VER) && defined(__x86_64__)
@ -365,44 +365,44 @@ __forceinline void writePixel24_0(void* pmem, int x, int y, u32 pixel, u32 bw) {
#endif
}
__forceinline void writePixel16_0(void* pmem, int x, int y, u32 pixel, u32 bw) {
static __forceinline void writePixel16_0(void* pmem, int x, int y, u32 pixel, u32 bw) {
((u16*)pmem)[getPixelAddress16_0(x, y, bw)] = pixel;
}
__forceinline void writePixel16S_0(void* pmem, int x, int y, u32 pixel, u32 bw) {
static __forceinline void writePixel16S_0(void* pmem, int x, int y, u32 pixel, u32 bw) {
((u16*)pmem)[getPixelAddress16S_0(x, y, bw)] = pixel;
}
__forceinline void writePixel8_0(void* pmem, int x, int y, u32 pixel, u32 bw) {
static __forceinline void writePixel8_0(void* pmem, int x, int y, u32 pixel, u32 bw) {
((u8*)pmem)[getPixelAddress8_0(x, y, bw)] = pixel;
}
__forceinline void writePixel8H_0(void* pmem, int x, int y, u32 pixel, u32 bw) {
static __forceinline void writePixel8H_0(void* pmem, int x, int y, u32 pixel, u32 bw) {
((u8*)pmem)[4*getPixelAddress32_0(x, y, bw)+3] = pixel;
}
__forceinline void writePixel4_0(void* pmem, int x, int y, u32 pixel, u32 bw) {
static __forceinline void writePixel4_0(void* pmem, int x, int y, u32 pixel, u32 bw) {
u32 addr = getPixelAddress4_0(x, y, bw);
u8 pix = ((u8*)pmem)[addr/2];
if (addr & 0x1) ((u8*)pmem)[addr/2] = (pix & 0x0f) | (pixel << 4);
else ((u8*)pmem)[addr/2] = (pix & 0xf0) | (pixel);
}
__forceinline void writePixel4HL_0(void* pmem, int x, int y, u32 pixel, u32 bw) {
static __forceinline void writePixel4HL_0(void* pmem, int x, int y, u32 pixel, u32 bw) {
u8 *p = (u8*)pmem + 4*getPixelAddress4HL_0(x, y, bw)+3;
*p = (*p & 0xf0) | pixel;
}
__forceinline void writePixel4HH_0(void* pmem, int x, int y, u32 pixel, u32 bw) {
static __forceinline void writePixel4HH_0(void* pmem, int x, int y, u32 pixel, u32 bw) {
u8 *p = (u8*)pmem + 4*getPixelAddress4HH_0(x, y, bw)+3;
*p = (*p & 0x0f) | (pixel<<4);
}
__forceinline void writePixel32Z_0(void* pmem, int x, int y, u32 pixel, u32 bw) {
static __forceinline void writePixel32Z_0(void* pmem, int x, int y, u32 pixel, u32 bw) {
((u32*)pmem)[getPixelAddress32Z_0(x, y, bw)] = pixel;
}
__forceinline void writePixel24Z_0(void* pmem, int x, int y, u32 pixel, u32 bw) {
static __forceinline void writePixel24Z_0(void* pmem, int x, int y, u32 pixel, u32 bw) {
u8 *buf = (u8*)pmem + 4*getPixelAddress32Z_0(x, y, bw);
u8 *pix = (u8*)&pixel;
#if defined(_MSC_VER) && defined(__x86_64__)
@ -412,42 +412,42 @@ __forceinline void writePixel24Z_0(void* pmem, int x, int y, u32 pixel, u32 bw)
#endif
}
__forceinline void writePixel16Z_0(void* pmem, int x, int y, u32 pixel, u32 bw) {
static __forceinline void writePixel16Z_0(void* pmem, int x, int y, u32 pixel, u32 bw) {
((u16*)pmem)[getPixelAddress16Z_0(x, y, bw)] = pixel;
}
__forceinline void writePixel16SZ_0(void* pmem, int x, int y, u32 pixel, u32 bw) {
static __forceinline void writePixel16SZ_0(void* pmem, int x, int y, u32 pixel, u32 bw) {
((u16*)pmem)[getPixelAddress16SZ_0(x, y, bw)] = pixel;
}
///////////////
__forceinline u32 readPixel32_0(const void* pmem, int x, int y, u32 bw) {
static __forceinline u32 readPixel32_0(const void* pmem, int x, int y, u32 bw) {
return ((const u32*)pmem)[getPixelAddress32_0(x, y, bw)];
}
__forceinline u32 readPixel24_0(const void* pmem, int x, int y, u32 bw) {
static __forceinline u32 readPixel24_0(const void* pmem, int x, int y, u32 bw) {
return ((const u32*)pmem)[getPixelAddress32_0(x, y, bw)] & 0xffffff;
}
__forceinline u32 readPixel16_0(const void* pmem, int x, int y, u32 bw) {
static __forceinline u32 readPixel16_0(const void* pmem, int x, int y, u32 bw) {
return ((const u16*)pmem)[getPixelAddress16_0(x, y, bw)];
}
__forceinline u32 readPixel16S_0(const void* pmem, int x, int y, u32 bw) {
static __forceinline u32 readPixel16S_0(const void* pmem, int x, int y, u32 bw) {
return ((const u16*)pmem)[getPixelAddress16S_0(x, y, bw)];
}
__forceinline u32 readPixel8_0(const void* pmem, int x, int y, u32 bw) {
static __forceinline u32 readPixel8_0(const void* pmem, int x, int y, u32 bw) {
return ((const u8*)pmem)[getPixelAddress8_0(x, y, bw)];
}
__forceinline u32 readPixel8H_0(const void* pmem, int x, int y, u32 bw) {
static __forceinline u32 readPixel8H_0(const void* pmem, int x, int y, u32 bw) {
return ((const u8*)pmem)[4*getPixelAddress32_0(x, y, bw) + 3];
}
__forceinline u32 readPixel4_0(const void* pmem, int x, int y, u32 bw) {
static __forceinline u32 readPixel4_0(const void* pmem, int x, int y, u32 bw) {
u32 addr = getPixelAddress4_0(x, y, bw);
u8 pix = ((const u8*)pmem)[addr/2];
if (addr & 0x1)
@ -455,31 +455,31 @@ __forceinline u32 readPixel4_0(const void* pmem, int x, int y, u32 bw) {
else return pix & 0xf;
}
__forceinline u32 readPixel4HL_0(const void* pmem, int x, int y, u32 bw) {
static __forceinline u32 readPixel4HL_0(const void* pmem, int x, int y, u32 bw) {
const u8 *p = (const u8*)pmem+4*getPixelAddress4HL_0(x, y, bw)+3;
return *p & 0x0f;
}
__forceinline u32 readPixel4HH_0(const void* pmem, int x, int y, u32 bw) {
static __forceinline u32 readPixel4HH_0(const void* pmem, int x, int y, u32 bw) {
const u8 *p = (const u8*)pmem+4*getPixelAddress4HH_0(x, y, bw) + 3;
return *p >> 4;
}
///////////////
__forceinline u32 readPixel32Z_0(const void* pmem, int x, int y, u32 bw) {
static __forceinline u32 readPixel32Z_0(const void* pmem, int x, int y, u32 bw) {
return ((const u32*)pmem)[getPixelAddress32Z_0(x, y, bw)];
}
__forceinline u32 readPixel24Z_0(const void* pmem, int x, int y, u32 bw) {
static __forceinline u32 readPixel24Z_0(const void* pmem, int x, int y, u32 bw) {
return ((const u32*)pmem)[getPixelAddress32Z_0(x, y, bw)] & 0xffffff;
}
__forceinline u32 readPixel16Z_0(const void* pmem, int x, int y, u32 bw) {
static __forceinline u32 readPixel16Z_0(const void* pmem, int x, int y, u32 bw) {
return ((const u16*)pmem)[getPixelAddress16Z_0(x, y, bw)];
}
__forceinline u32 readPixel16SZ_0(const void* pmem, int x, int y, u32 bw) {
static __forceinline u32 readPixel16SZ_0(const void* pmem, int x, int y, u32 bw) {
return ((const u16*)pmem)[getPixelAddress16SZ_0(x, y, bw)];
}

6
plugins/zerogs/dx/Regs.cpp
View File

@ -78,7 +78,7 @@ u32 s_uTex1Data[2][2] = {0}, s_uClampData[2] = {0};
void __fastcall GIFPackedRegHandlerNull(u32* data)
{
#ifdef _DEBUG
printf("Unexpected packed reg handler %8.8lx_%8.8lx %x\n", data[0], data[1], data[2]);
DEBUG_LOG("Unexpected packed reg handler %8.8lx_%8.8lx %x\n", data[0], data[1], data[2]);
#endif
}
@ -373,16 +373,14 @@ void __fastcall GIFRegHandlerNull(u32* data)
// 0x7f happens on a lot of games
if( data[2] != 0x7f && (data[0] || data[1]) ) {
printf("Unexpected reg handler %x %x %x\n", data[0], data[1], data[2]);
DEBUG_LOG("Unexpected reg handler %x %x %x\n", data[0], data[1], data[2]);
}
#endif
}
void __fastcall GIFRegHandlerPRIM(u32 *data) {
if (data[0] & ~0x3ff) {
#ifdef WARN_LOG
//WARN_LOG("warning: unknown bits in prim %8.8lx_%8.8lx\n", data[1], data[0]);
#endif
}
gs.nTriFanVert = gs.primIndex;

20
plugins/zerogs/dx/common.h
View File

@ -338,13 +338,13 @@ static inline void put_bits(PutBitContext *s, int n, unsigned int value)
#ifdef STATS
st_out_bit_counts[st_current_index] += n;
#endif
// printf("put_bits=%d %x\n", n, value);
// DEBUG_LOG("put_bits=%d %x\n", n, value);
assert(n == 32 || value < (1U << n));
bit_buf = s->bit_buf;
bit_left = s->bit_left;
// printf("n=%d value=%x cnt=%d buf=%x\n", n, value, bit_cnt, bit_buf);
// DEBUG_LOG("n=%d value=%x cnt=%d buf=%x\n", n, value, bit_cnt, bit_buf);
/* XXX: optimize */
if (n < bit_left) {
bit_buf = (bit_buf<<n) | value;
@ -361,7 +361,7 @@ static inline void put_bits(PutBitContext *s, int n, unsigned int value)
} else
#endif
*(uint32_t *)s->buf_ptr = be2me_32(bit_buf);
//printf("bitbuf = %08x\n", bit_buf);
//DEBUG_LOG("bitbuf = %08x\n", bit_buf);
s->buf_ptr+=4;
bit_left+=32 - n;
bit_buf = value;
@ -403,7 +403,7 @@ static inline void put_bits(PutBitContext *s, int n, unsigned int value)
ptr[0] |= be2me_32(value>>(index&31));
ptr[1] = be2me_32(value<<(32-(index&31)));
//if(n>24) printf("%d %d\n", n, value);
//if(n>24) DEBUG_LOG("%d %d\n", n, value);
index+= n;
s->index= index;
# endif
@ -431,7 +431,7 @@ static inline void put_bits(PutBitContext *s, int n, unsigned int value)
ptr[0] |= be2me_32(value<<(32-n-(index&7) ));
ptr[1] = 0;
//if(n>24) printf("%d %d\n", n, value);
//if(n>24) DEBUG_LOG("%d %d\n", n, value);
index+= n;
s->index= index;
# endif
@ -885,17 +885,17 @@ static inline void print_bin(int bits, int n){
int i;
for(i=n-1; i>=0; i--){
printf("%d", (bits>>i)&1);
DEBUG_LOG("%d", (bits>>i)&1);
}
for(i=n; i<24; i++)
printf(" ");
DEBUG_LOG(" ");
}
static inline int get_bits_trace(GetBitContext *s, int n, char *file, char *func, int line){
int r= get_bits(s, n);
print_bin(r, n);
printf("%5d %2d %3d bit @%5d in %s %s:%d\n", r, n, r, get_bits_count(s)-n, file, func, line);
DEBUG_LOG("%5d %2d %3d bit @%5d in %s %s:%d\n", r, n, r, get_bits_count(s)-n, file, func, line);
return r;
}
static inline int get_vlc_trace(GetBitContext *s, VLC_TYPE (*table)[2], int bits, int max_depth, char *file, char *func, int line){
@ -907,7 +907,7 @@ static inline int get_vlc_trace(GetBitContext *s, VLC_TYPE (*table)[2], int bits
print_bin(bits2, len);
printf("%5d %2d %3d vlc @%5d in %s %s:%d\n", bits2, len, r, pos, file, func, line);
DEBUG_LOG("%5d %2d %3d vlc @%5d in %s %s:%d\n", bits2, len, r, pos, file, func, line);
return r;
}
static inline int get_xbits_trace(GetBitContext *s, int n, char *file, char *func, int line){
@ -915,7 +915,7 @@ static inline int get_xbits_trace(GetBitContext *s, int n, char *file, char *fun
int r= get_xbits(s, n);
print_bin(show, n);
printf("%5d %2d %3d xbt @%5d in %s %s:%d\n", show, n, r, get_bits_count(s)-n, file, func, line);
DEBUG_LOG("%5d %2d %3d xbt @%5d in %s %s:%d\n", show, n, r, get_bits_count(s)-n, file, func, line);
return r;
}

20
plugins/zerogs/dx/targets.cpp
View File

@ -386,12 +386,12 @@ void ZeroGS::CRenderTarget::Update(int context, ZeroGS::CRenderTarget* pdepth)
if( ittarg == s_DepthRTs.mapTargets.end() )
nUpdateTarg = 0;
else if( ittarg->second == this ) {
printf("zerogs: updating self");
DEBUG_LOG("zerogs: updating self");
nUpdateTarg = 0;
}
}
else if( ittarg->second == this ) {
printf("zerogs: updating self");
DEBUG_LOG("zerogs: updating self");
nUpdateTarg = 0;
}
}
@ -500,7 +500,7 @@ void ZeroGS::CRenderTarget::ConvertTo32()
V(pd3dDevice->CreateTexture(fbw<<s_AAx, (fbh<<s_AAy)/2, 1, D3DUSAGE_RENDERTARGET, g_RenderFormat, D3DPOOL_DEFAULT, &ptexConv, NULL));
if( FAILED(hr) ) {
printf("Failed to create feedback\n");
DEBUG_LOG("Failed to create feedback\n");
return;
}
@ -696,7 +696,7 @@ void ZeroGS::CRenderTarget::ConvertTo16()
V(pd3dDevice->CreateTexture(fbw<<s_AAx, (fbh<<s_AAy)*2, 1, D3DUSAGE_RENDERTARGET, g_RenderFormat, D3DPOOL_DEFAULT, &ptexConv, NULL));
if( FAILED(hr) ) {
printf("Failed to create feedback\n");
DEBUG_LOG("Failed to create feedback\n");
return;
}
@ -901,7 +901,7 @@ void ZeroGS::CRenderTarget::_CreateFeedback()
V(pd3dDevice->CreateTexture(fbw<<s_AAx, fbh<<s_AAy, 1, D3DUSAGE_RENDERTARGET, g_RenderFormat, D3DPOOL_DEFAULT, &ptexFeedback, NULL));
if( FAILED(hr) ) {
printf("Failed to create feedback\n");
DEBUG_LOG("Failed to create feedback\n");
return;
}
@ -1069,7 +1069,7 @@ void ZeroGS::CDepthTarget::Update(int context, ZeroGS::CRenderTarget* prndr)
if( nUpdateTarg ) {
CRenderTargetMngr::MAPTARGETS::iterator ittarg = s_DepthRTs.mapTargets.find(nUpdateTarg);
if( ittarg == s_DepthRTs.mapTargets.end() || ittarg->second == this) {
printf("zerogs: updating self");
DEBUG_LOG("zerogs: updating self");
nUpdateTarg = 0;
}
else pusetarg = ittarg->second;
@ -2082,7 +2082,7 @@ ZeroGS::CMemoryTarget* ZeroGS::CMemoryTargetMngr::GetMemoryTarget(const tex0Info
listClearedTargets.pop_front();
else {
if( listTargets.size() == 0 ) {
printf("Failed to create %dx%x texture\n", GPU_TEXWIDTH*channels*widthmult, (targ->realheight+widthmult-1)/widthmult);
DEBUG_LOG("Failed to create %dx%x texture\n", GPU_TEXWIDTH*channels*widthmult, (targ->realheight+widthmult-1)/widthmult);
channels = 1;
}
DestroyOldest();
@ -2776,7 +2776,7 @@ void TransferHostLocal(const void* pbyMem, u32 nQWordSize)
if( (ptarg->status & CRenderTarget::TS_Virtual) )
continue;
//printf("zerogs: resolving to alpha channel\n");
//DEBUG_LOG("zerogs: resolving to alpha channel\n");
ptarg->Resolve();
}
}
@ -2839,7 +2839,7 @@ void TransferHostLocal(const void* pbyMem, u32 nQWordSize)
// const T* pbuf = (const T*)pbyMem; \
// u32 nSize = nQWordSize*(4/sizeof(T)); \
// assert( (nSize%widthlimit) == 0 && widthlimit <= 4 ); \
// if( ((gs.imageEndX-gs.trxpos.dx)%widthlimit) ) printf("Bad Transmission! %d %d, psm: %d\n", gs.trxpos.dx, gs.imageEndX, DSTPSM); \
// if( ((gs.imageEndX-gs.trxpos.dx)%widthlimit) ) DEBUG_LOG("Bad Transmission! %d %d, psm: %d\n", gs.trxpos.dx, gs.imageEndX, DSTPSM); \
// for(; i < gs.imageEndY; ++i) { \
// for(; j < gs.imageEndX && nSize > 0; j += widthlimit, nSize -= widthlimit, pbuf += widthlimit) { \
// /* write as many pixel at one time as possible */ \
@ -2907,7 +2907,7 @@ void TransferHostLocal(const void* pbyMem, u32 nQWordSize)
//// // hack
//// if( abs((int)nQWordSize*8 - (gs.imageEndY-i)*(gs.imageEndX-gs.trxpos.dx)+(j-gs.trxpos.dx)) <= 8 ) {
//// // don't transfer
//// printf("bad texture 4: %d %d %d\n", gs.trxpos.dx, gs.imageEndX, nQWordSize);
//// DEBUG_LOG("bad texture 4: %d %d %d\n", gs.trxpos.dx, gs.imageEndX, nQWordSize);
//// gs.imageEndX = gs.trxpos.dx + (gs.imageEndX-gs.trxpos.dx)&~7;
//// //i = gs.imageEndY;
//// //goto End;

328
plugins/zerogs/dx/zerogs.cpp
View File

@ -125,7 +125,7 @@ extern int s_frameskipping;
// static const char* pnames[4] = {NULL};
//
// if( ptargs[index] == psurf && psurf != NULL )
// printf("duplicate targets\n");
// DEBUG_LOG("duplicate targets\n");
// pd3dDevice->SetRenderTarget(index, psurf);
// ptargs[index] = psurf;
// counters[index] = counter;
@ -475,7 +475,7 @@ void ZeroGS::VB::CheckFrame(int tbp)
maxpos = min(gsfb.fbh, maxpos);
maxpos = min(maxmin, maxpos);
//? alteir aris crashes without it
//? atelier iris crashes without it
if( maxpos > 256 )
maxpos &= ~0x1f;
}
@ -505,8 +505,7 @@ void ZeroGS::VB::CheckFrame(int tbp)
}
frame = gsfb;
if( frame.fbw > 1024 )
frame.fbw = 1024;
if (frame.fbw > 1024) frame.fbw = 1024;
// if( fbh > 256 && (fbh % m_Blocks[gsfb.psm].height) <= 2 ) {
// // dragon ball z
@ -600,7 +599,7 @@ void ZeroGS::VB::CheckFrame(int tbp)
//CRenderTargetMngr::MAPTARGETS::iterator it = s_RTs.mapTargets.find(key);
// if( it != s_RTs.mapTargets.end() ) {
//#ifdef _DEBUG
// printf("zbuf resolve\n");
// DEBUG_LOG("zbuf resolve\n");
//#endif
// if( it->second->status & CRenderTarget::TS_Resolved )
// it->second->Resolve();
@ -868,7 +867,7 @@ HRESULT ZeroGS::Create(LONG _width, LONG _height)
D3DADAPTER_IDENTIFIER9 id;
HRESULT hr = pD3D->GetAdapterIdentifier(i, 0, &id);
if( strcmp(id.Description, "NVIDIA NVPerfHUD") == 0 ) {
printf("Using %s adapter\n", id.Description);
DEBUG_LOG("Using %s adapter\n", id.Description);
adapter = i;
devtype = D3DDEVTYPE_REF;
break;
@ -1177,7 +1176,7 @@ void ZeroGS::ChangeDeviceSize(int nNewWidth, int nNewHeight)
int oldwidth = width, oldheight = height;
if( FAILED(Create(nNewWidth&~7, nNewHeight&~7)) ) {
printf("Failed to recreate, changing to old\n");
DEBUG_LOG("Failed to recreate, changing to old\n");
if( FAILED(Create(oldwidth, oldheight)) ) {
MessageBox(NULL, "failed to create dev, exiting...\n", "Error", MB_OK);
exit(0);
@ -1246,7 +1245,7 @@ void ZeroGS::SetAA(int mode)
assert( (header) != NULL && (header)->index == (Index) ); \
hr = pd3dDevice->CreateVertexShader((DWORD*)(s_lpShaderResources + (header)->offset), &(ptr)); \
if( FAILED(hr) || ptr == NULL ) { \
printf("errors 0x%x for %d, failed.. try updating your drivers or dx\n", hr, Index); \
DEBUG_LOG("errors 0x%x for %d, failed.. try updating your drivers or dx\n", hr, Index); \
return E_FAIL; \
} \
} \
@ -1256,7 +1255,7 @@ void ZeroGS::SetAA(int mode)
header = mapShaderResources[index]; \
hr = pd3dDevice->CreatePixelShader((DWORD*)(s_lpShaderResources + (header)->offset), &(ptr)); \
if( FAILED(hr) || ptr == NULL ) { \
printf("errors 0x%x for %s, failed.. try updating your drivers or dx\n", hr, index); \
DEBUG_LOG("errors 0x%x for %s, failed.. try updating your drivers or dx\n", hr, index); \
return E_FAIL; \
} \
} \
@ -1357,7 +1356,7 @@ LPD3DPS ZeroGS::LoadShadeEffect(int type, int texfilter, int fog, int testaem, i
index += NUM_SHADERS*g_nPixelShaderVer;
assert( mapShaderResources.find(index) != mapShaderResources.end() );
SHADERHEADER* header = mapShaderResources[index];
if( header == NULL ) printf("%d %d\n", index%NUM_SHADERS, g_nPixelShaderVer);
if( header == NULL ) DEBUG_LOG("%d %d\n", index%NUM_SHADERS, g_nPixelShaderVer);
assert( header != NULL );
HRESULT hr = pd3dDevice->CreatePixelShader((DWORD*)(s_lpShaderResources + header->offset), pps);
@ -1425,7 +1424,7 @@ public:
V(D3DXCompileShaderFromFile(EFFECT_NAME"ps2hw.fx", pmacros, pInclude, name, shaderver, ShaderFlagsVS, &pShader, &pError, NULL)); \
if( FAILED(hr) ) \
{ \
printf("Failed to load vs %s: \n%s\n", name, pError->GetBufferPointer()); \
DEBUG_LOG("Failed to load vs %s: \n%s\n", name, pError->GetBufferPointer()); \
SAFE_RELEASE(pShader); \
SAFE_RELEASE(pError); \
return hr; \
@ -1441,7 +1440,7 @@ public:
V(D3DXCompileShaderFromFile(EFFECT_NAME"ps2hw.fx", pmacros, pInclude, name, shmodel, ShaderFlagsPS, &pShader, &pError, NULL)); \
if( FAILED(hr) ) \
{ \
printf("Failed to load ps %s: \n%s\n", name, pError->GetBufferPointer()); \
DEBUG_LOG("Failed to load ps %s: \n%s\n", name, pError->GetBufferPointer()); \
SAFE_RELEASE(pShader); \
SAFE_RELEASE(pError); \
return hr; \
@ -1450,7 +1449,7 @@ public:
SAFE_RELEASE(pShader); \
SAFE_RELEASE(pError); \
if( FAILED(hr) || ptr == NULL ) { \
printf("errors 0x%x for %s, failed.. try updating your drivers or dx\n", hr, name); \
DEBUG_LOG("errors 0x%x for %s, failed.. try updating your drivers or dx\n", hr, name); \
return E_FAIL; \
} \
} \
@ -1566,7 +1565,7 @@ LPD3DPS ZeroGS::LoadShadeEffect(int type, int texfilter, int fog, int testaem, i
if( SUCCEEDED(hr) )
return *pps;
printf("Failed to create shader %d,%d,%d,%d\n", type, fog, texfilter, 4*clamp.wms+clamp.wmt);
DEBUG_LOG("Failed to create shader %d,%d,%d,%d\n", type, fog, texfilter, 4*clamp.wms+clamp.wmt);
return NULL;
}
@ -1606,7 +1605,7 @@ HRESULT ZeroGS::InitDeviceObjects()
if( blockfmt == D3DFMT_R32F ) {
if( FAILED(hr = pd3dDevice->CreateTexture(BLOCK_TEXWIDTH, BLOCK_TEXHEIGHT, 1, 0, D3DFMT_A32B32G32R32F, D3DPOOL_MANAGED, &ptexBilinearBlocks, NULL)) ) {
printf("Failed to create bilinear block texture, fmt = D3DFMT_A32B32G32R32F\n");
DEBUG_LOG("Failed to create bilinear block texture, fmt = D3DFMT_A32B32G32R32F\n");
}
}
else ptexBilinearBlocks = NULL;
@ -1716,7 +1715,7 @@ void ZeroGS::DeleteDeviceObjects()
if( s_aviinit ) {
StopCapture();
STOP_AVI();
printf("zerogs.avi stopped");
DEBUG_LOG("zerogs.avi stopped");
s_aviinit = 0;
}
@ -1926,7 +1925,7 @@ void ZeroGS::Flush(int context)
// kh2 hack
// if( curvb.dwCount == 2 && curvb.curprim.tme == 0 && curvb.curprim.abe == 0 && (curvb.tex0.tbp0 == 0x2a00 || curvb.tex0.tbp0==0x1d00) ) {
// // skip
// printf("skipping\n");
// DEBUG_LOG("skipping\n");
// g_SaveFrameNum++;
// curvb.dwCount = 0;
// return;
@ -3378,7 +3377,7 @@ void ZeroGS::RenderCRTC(int interlace)
pd3dDevice->EndScene();
if( pd3dDevice->Present(NULL, NULL, NULL, NULL) == D3DERR_DEVICELOST ) {
// device is lost, need to recreate
printf("device lost\n");
DEBUG_LOG("device lost\n");
g_bIsLost = TRUE;
Reset();
return;
@ -3462,7 +3461,7 @@ void ZeroGS::RenderCRTC(int interlace)
if( total / ARRAYSIZE(s_nResolveCounts) > 3 ) {
if( s_nLastResolveReset > (int)(fFPS * 8) ) {
// reset
printf("ZeroGS: video mem reset\n");
DEBUG_LOG("ZeroGS: video mem reset\n");
s_nLastResolveReset = 0;
memset(s_nResolveCounts, 0, sizeof(s_nResolveCounts));
@ -3495,52 +3494,63 @@ void ZeroGS::RenderCRTC(int interlace)
// Internal Definitions //
//////////////////////////
#define MOVZ(p, gsz) p.z = curvb.zprimmask==0xffff?min(0xffff, gsz):gsz;
#define MOVFOG(p, gsf) p.f = ((s16)(gsf).f<<7)|0x7f;
__forceinline void MOVZ(VertexGPU *p, u32 gsz, const VB& curvb)
{
p->z = curvb.zprimmask==0xffff?min((u32)0xffff, gsz):gsz;
}
#define SET_VERTEX(p, Index) { \
int index = Index; \
p.x = (((int)gs.gsvertex[index].x - curvb.offset.x)>>1)&0xffff; \
p.y = (((int)gs.gsvertex[index].y - curvb.offset.y)>>1)&0xffff; \
/*x = ((int)gs.gsvertex[index].x - curvb.offset.x); \
y = ((int)gs.gsvertex[index].y - curvb.offset.y); \
p.x = (x&0x7fff) | (x < 0 ? 0x8000 : 0); \
p.y = (y&0x7fff) | (y < 0 ? 0x8000 : 0);*/ \
p.f = ((s16)gs.gsvertex[index].f<<7)|0x7f; \
MOVZ(p, gs.gsvertex[index].z); \
p.rgba = prim->iip ? gs.gsvertex[index].rgba : gs.rgba; \
if( (g_GameSettings & GAME_TEXAHACK) && !(p.rgba&0xffffff) ) \
p.rgba = 0; \
if( prim->tme ) { \
if( prim->fst ) { \
p.s = (float)gs.gsvertex[index].u * fiTexWidth[prim->ctxt]; \
p.t = (float)gs.gsvertex[index].v * fiTexHeight[prim->ctxt]; \
p.q = 1; \
} \
else { \
p.s = gs.gsvertex[index].s; \
p.t = gs.gsvertex[index].t; \
p.q = gs.gsvertex[index].q; \
} \
} \
} \
__forceinline void MOVFOG(VertexGPU *p, Vertex gsf)
{
p->f = ((s16)(gsf).f<<7)|0x7f;
}
__forceinline void SET_VERTEX(VertexGPU *p, int Index, const VB& curvb)
{
int index = Index;
p->x = (((int)gs.gsvertex[index].x - curvb.offset.x)>>1)&0xffff;
p->y = (((int)gs.gsvertex[index].y - curvb.offset.y)>>1)&0xffff;
/*x = ((int)gs.gsvertex[index].x - curvb.offset.x);
y = ((int)gs.gsvertex[index].y - curvb.offset.y);
p.x = (x&0x7fff) | (x < 0 ? 0x8000 : 0);
p.y = (y&0x7fff) | (y < 0 ? 0x8000 : 0);*/
p->f = ((s16)gs.gsvertex[index].f<<7)|0x7f;
MOVZ(p, gs.gsvertex[index].z, curvb);
p->rgba = prim->iip ? gs.gsvertex[index].rgba : gs.rgba;
if ((g_GameSettings & GAME_TEXAHACK) && !(p->rgba&0xffffff))
p->rgba = 0;
if (prim->tme )
{
if( prim->fst )
{
p->s = (float)gs.gsvertex[index].u * fiTexWidth[prim->ctxt];
p->t = (float)gs.gsvertex[index].v * fiTexHeight[prim->ctxt];
p->q = 1;
}
else
{
p->s = gs.gsvertex[index].s;
p->t = gs.gsvertex[index].t;
p->q = gs.gsvertex[index].q;
}
}
}
#define OUTPUT_VERT(fn, vert, id) { \
fn("%c%d(%d): xyzf=(%4d,%4d,0x%x,%3d), rgba=0x%8.8x, stq = (%2.5f,%2.5f,%2.5f)\n", id==0?'*':' ', id, prim->prim, vert.x/8, vert.y/8, vert.z, vert.f/128, \
vert.rgba, Clamp(vert.s, -10, 10), Clamp(vert.t, -10, 10), Clamp(vert.q, -10, 10)); \
} \
//#define OUTPUT_VERT(fn, vert, id) { \
// fn("%c%d(%d): xyzf=(%4d,%4d,0x%x,%3d)\n", id==0?'*':' ', id, prim->prim, vert.x/8, vert.y/8, vert.z, vert.f/128); \
//} \
void ZeroGS::KickPoint()
{
assert( gs.primC >= 1 );
VB& curvb = vb[prim->ctxt];
if( curvb.bNeedTexCheck )
curvb.FlushTexData();
if (curvb.bNeedTexCheck) curvb.FlushTexData();
if( !(g_GameSettings&GAME_DOPARALLELCTX) && vb[!prim->ctxt].dwCount > 0 && vb[prim->ctxt].gsfb.fbp == vb[!prim->ctxt].gsfb.fbp )
{
@ -3555,7 +3565,7 @@ void ZeroGS::KickPoint()
int last = (gs.primIndex+2)%ARRAYSIZE(gs.gsvertex);
VertexGPU* p = curvb.pbuf+curvb.dwCount;
SET_VERTEX(p[0], last);
SET_VERTEX(&p[0], last, curvb);
curvb.dwCount++;
#ifdef PRIM_LOG
@ -3584,8 +3594,8 @@ void ZeroGS::KickLine()
int last = (gs.primIndex+2)%ARRAYSIZE(gs.gsvertex);
VertexGPU* p = curvb.pbuf+curvb.dwCount*2;
SET_VERTEX(p[0], next);
SET_VERTEX(p[1], last);
SET_VERTEX(&p[0], next, curvb);
SET_VERTEX(&p[1], last, curvb);
curvb.dwCount++;
@ -3599,8 +3609,7 @@ void ZeroGS::KickTriangle()
{
assert( gs.primC >= 3 );
VB& curvb = vb[prim->ctxt];
if( curvb.bNeedTexCheck )
curvb.FlushTexData();
if (curvb.bNeedTexCheck) curvb.FlushTexData();
if( !(g_GameSettings&GAME_DOPARALLELCTX) && vb[!prim->ctxt].dwCount > 0 && vb[prim->ctxt].gsfb.fbp == vb[!prim->ctxt].gsfb.fbp )
{
@ -3613,9 +3622,9 @@ void ZeroGS::KickTriangle()
curvb.Lock();
VertexGPU* p = curvb.pbuf+curvb.dwCount*3;
SET_VERTEX(p[0], 0);
SET_VERTEX(p[1], 1);
SET_VERTEX(p[2], 2);
SET_VERTEX(&p[0], 0, curvb);
SET_VERTEX(&p[1], 1, curvb);
SET_VERTEX(&p[2], 2, curvb);
curvb.dwCount++;
@ -3630,8 +3639,7 @@ void ZeroGS::KickTriangleFan()
{
assert( gs.primC >= 3 );
VB& curvb = vb[prim->ctxt];
if( curvb.bNeedTexCheck )
curvb.FlushTexData();
if (curvb.bNeedTexCheck) curvb.FlushTexData();
if( !(g_GameSettings&GAME_DOPARALLELCTX) && vb[!prim->ctxt].dwCount > 0 && vb[prim->ctxt].gsfb.fbp == vb[!prim->ctxt].gsfb.fbp )
{
@ -3644,9 +3652,9 @@ void ZeroGS::KickTriangleFan()
curvb.Lock();
VertexGPU* p = curvb.pbuf+curvb.dwCount*3;
SET_VERTEX(p[0], 0);
SET_VERTEX(p[1], 1);
SET_VERTEX(p[2], 2);
SET_VERTEX(&p[0], 0, curvb);
SET_VERTEX(&p[1], 1, curvb);
SET_VERTEX(&p[2], 2, curvb);
curvb.dwCount++;
@ -3661,6 +3669,13 @@ void ZeroGS::KickTriangleFan()
#endif
}
__forceinline void SetKickVertex(VertexGPU *p, Vertex v, int next, const VB& curvb)
{
SET_VERTEX(p, next, curvb);
MOVZ(p, v.z, curvb);
MOVFOG(p, v);
}
void ZeroGS::KickSprite()
{
assert( gs.primC >= 2 );
@ -3674,18 +3689,17 @@ void ZeroGS::KickSprite()
Flush(!prim->ctxt);
}
if( curvb.dwCount >= POINT_BUFFERFLUSH/3 )
Flush(prim->ctxt);
if (curvb.dwCount >= POINT_BUFFERFLUSH/3) Flush(prim->ctxt);
curvb.Lock();
int next = (gs.primIndex+1)%ARRAYSIZE(gs.gsvertex);
int last = (gs.primIndex+2)%ARRAYSIZE(gs.gsvertex);
// sprite is too small and AA shows lines (tek4)
if( s_AAx ) {
if( s_AAx )
{
gs.gsvertex[last].x += 4;
if( s_AAy )
gs.gsvertex[last].y += 4;
if (s_AAy) gs.gsvertex[last].y += 4;
}
// might be bad sprite (KH dialog text)
@ -3693,23 +3707,20 @@ void ZeroGS::KickSprite()
// return;
VertexGPU* p = curvb.pbuf+curvb.dwCount*3;
SET_VERTEX(p[0], next); MOVZ(p[0], gs.gsvertex[last].z); MOVFOG(p[0], gs.gsvertex[last]);
SET_VERTEX(p[3], next); MOVZ(p[3], gs.gsvertex[last].z); MOVFOG(p[3], gs.gsvertex[last]);
SET_VERTEX(p[1], last); MOVZ(p[1], gs.gsvertex[last].z); MOVFOG(p[1], gs.gsvertex[last]);
SET_VERTEX(p[4], last); MOVZ(p[4], gs.gsvertex[last].z); MOVFOG(p[4], gs.gsvertex[last]);
SetKickVertex(&p[0], gs.gsvertex[last], next, curvb);
SetKickVertex(&p[3], gs.gsvertex[last], next, curvb);
SetKickVertex(&p[1], gs.gsvertex[last], last, curvb);
SetKickVertex(&p[4], gs.gsvertex[last], last, curvb);
if( g_MaxRenderedHeight < p[0].y )
g_MaxRenderedHeight = p[0].y;
if( g_MaxRenderedHeight < p[1].y )
g_MaxRenderedHeight = p[1].y;
if (g_MaxRenderedHeight < p[0].y) g_MaxRenderedHeight = p[0].y;
if (g_MaxRenderedHeight < p[1].y) g_MaxRenderedHeight = p[1].y;
SET_VERTEX(p[2], next); MOVZ(p[2], gs.gsvertex[last].z); MOVFOG(p[2], gs.gsvertex[last]);
SetKickVertex(&p[2], gs.gsvertex[last], next, curvb);
p[2].s = p[1].s;
p[2].x = p[1].x;
SET_VERTEX(p[5], last); MOVZ(p[5], gs.gsvertex[last].z); MOVFOG(p[5], gs.gsvertex[last]);
SetKickVertex(&p[5], gs.gsvertex[last], last, curvb);
p[5].s = p[0].s;
p[5].x = p[0].x;
@ -4397,7 +4408,7 @@ void ZeroGS::SetTexVariablesInt(int context, int bilinear, const tex0Info& tex0,
//if( a.fix <= 0x80 ) { \
// dwTemp = (a.fix*2)>255?255:(a.fix*2); \
// dwTemp = dwTemp|(dwTemp<<8)|(dwTemp<<16)|0x80000000; \
// printf("bfactor: %8.8x\n", dwTemp); \
// DEBUG_LOG("bfactor: %8.8x\n", dwTemp); \
// SETRS(D3DRS_BLENDFACTOR, dwTemp); \
// } \
// else { \
@ -4664,12 +4675,13 @@ bool ZeroGS::CheckChangeInClut(u32 highdword, u32 psm)
// processing the CLUT after tex0/2 are written
switch(cld) {
case 0: return false;
case 1: break; // Seems to rarely not be 1.
// note sure about changing cbp[0,1]
case 4: return gs.cbp[0] != cbp;
case 5: return gs.cbp[1] != cbp;
// default: load
default: break;
}
int cpsm = (highdword >> 19) & 0xe;
@ -4775,9 +4787,10 @@ void ZeroGS::texClutWrite(int ctx)
// processing the CLUT after tex0/2 are written
switch(tex0.cld) {
case 0: return;
case 1: break; // tex0.cld is usually 1.
case 2: gs.cbp[0] = tex0.cbp; break;
case 3: gs.cbp[1] = tex0.cbp; break;
// note sure about changing cbp[0,1]
// not sure about changing cbp[0,1]
case 4:
if( gs.cbp[0] == tex0.cbp )
return;
@ -4788,75 +4801,97 @@ void ZeroGS::texClutWrite(int ctx)
return;
gs.cbp[1] = tex0.cbp;
break;
default: //DEBUG_LOG("cld isn't 0-5!");
break;
}
Flush(!ctx);
int entries = (tex0.psm&3)==3 ? 256 : 16;
if (tex0.csm)
{
switch (tex0.cpsm)
{
// 16bit psm
// eggomania uses non16bit textures for csm2
case PSMCT16:
{
u16* src = (u16*)g_pbyGSMemory + tex0.cbp*128;
u16 *dst = (u16*)(g_pbyGSClut+32*(tex0.csa&15)+(tex0.csa>=16?2:0));
if( tex0.csm ) {
// 16bit psm
// eggomania uses non16bit textures for csm2
if( tex0.cpsm == PSMCT16 ) {
u16* src = (u16*)g_pbyGSMemory + tex0.cbp*128;
u16 *dst = (u16*)(g_pbyGSClut+32*(tex0.csa&15)+(tex0.csa>=16?2:0));
for(int i = 0; i < entries; ++i) {
*dst = src[getPixelAddress16_0(gs.clut.cou+i, gs.clut.cov, gs.clut.cbw)];
dst += 2;
// check for wrapping
if( ((u32)dst & 0x3ff) == 0 )
dst = (u16*)(g_pbyGSClut+2);
}
}
else if( tex0.cpsm == PSMCT16S ) {
u16* src = (u16*)g_pbyGSMemory + tex0.cbp*128;
u16 *dst = (u16*)(g_pbyGSClut+32*(tex0.csa&15)+(tex0.csa>=16?2:0));
for(int i = 0; i < entries; ++i) {
*dst = src[getPixelAddress16S_0(gs.clut.cou+i, gs.clut.cov, gs.clut.cbw)];
dst += 2;
// check for wrapping
if( ((u32)dst & 0x3ff) == 0 )
dst = (u16*)(g_pbyGSClut+2);
}
}
else if( tex0.cpsm == PSMCT32 || tex0.cpsm == PSMCT24 ) {
u32* src = (u32*)g_pbyGSMemory + tex0.cbp*64;
u32 *dst = (u32*)(g_pbyGSClut+64*tex0.csa);
// check if address exceeds src
if( src+getPixelAddress32_0(gs.clut.cou+entries-1, gs.clut.cov, gs.clut.cbw) >= (u32*)g_pbyGSMemory + 0x00100000 ) {
ERROR_LOG("texClutWrite out of bounds\n");
}
else {
for(int i = 0; i < entries; ++i) {
*dst++ = src[getPixelAddress32_0(gs.clut.cou+i, gs.clut.cov, gs.clut.cbw)];
for (int i = 0; i < entries; ++i)
{
*dst = src[getPixelAddress16_0(gs.clut.cou+i, gs.clut.cov, gs.clut.cbw)];
dst += 2;
// check for wrapping
if (((u32)dst & 0x3ff) == 0) dst = (u16*)(g_pbyGSClut+2);
}
break;
}
}
else {
case PSMCT16S:
{
u16* src = (u16*)g_pbyGSMemory + tex0.cbp*128;
u16 *dst = (u16*)(g_pbyGSClut+32*(tex0.csa&15)+(tex0.csa>=16?2:0));
for (int i = 0; i < entries; ++i)
{
*dst = src[getPixelAddress16S_0(gs.clut.cou+i, gs.clut.cov, gs.clut.cbw)];
dst += 2;
// check for wrapping
if (((u32)dst & 0x3ff) == 0) dst = (u16*)(g_pbyGSClut+2);
}
break;
}
case PSMCT32:
case PSMCT24:
{
u32* src = (u32*)g_pbyGSMemory + tex0.cbp*64;
u32 *dst = (u32*)(g_pbyGSClut+64*tex0.csa);
// check if address exceeds src
if( src+getPixelAddress32_0(gs.clut.cou+entries-1, gs.clut.cov, gs.clut.cbw) >= (u32*)g_pbyGSMemory + 0x00100000 )
ERROR_LOG("texClutWrite out of bounds\n");
else
for(int i = 0; i < entries; ++i)
{
*dst = src[getPixelAddress32_0(gs.clut.cou+i, gs.clut.cov, gs.clut.cbw)];
dst++;
}
}
break;
default:
{
#ifndef RELEASE_TO_PUBLIC
//printf("unknown cpsm: %x (%x)\n", tex0.cpsm, tex0.psm);
//DEBUG_LOG("unknown cpsm: %x (%x)\n", tex0.cpsm, tex0.psm);
#endif
break;
}
}
}
else {
if( tex0.cpsm <= 1 ) {
if( entries == 16 )
WriteCLUT_T32_I4_CSM1((u32*)(g_pbyGSMemory + tex0.cbp*256), (u32*)(g_pbyGSClut+64*tex0.csa));
else
WriteCLUT_T32_I8_CSM1((u32*)(g_pbyGSMemory + tex0.cbp*256), (u32*)(g_pbyGSClut+64*tex0.csa));
}
else {
if( entries == 16 )
WriteCLUT_T16_I4_CSM1((u32*)(g_pbyGSMemory + 256 * tex0.cbp), (u32*)(g_pbyGSClut+32*(tex0.csa&15)+(tex0.csa>=16?2:0)));
else if(entries == 256) {
// sse2 for 256 is more complicated, so use regular
WriteCLUT_T16_I8_CSM1_c((u32*)(g_pbyGSMemory + 256 * tex0.cbp), (u32*)(g_pbyGSClut+32*(tex0.csa&15)+(tex0.csa>=16?2:0)));
}
else
{
switch (tex0.cpsm)
{
case PSMCT24:
case PSMCT32:
if( entries == 16 )
WriteCLUT_T32_I4_CSM1((u32*)(g_pbyGSMemory + tex0.cbp*256), (u32*)(g_pbyGSClut+64*tex0.csa));
else
WriteCLUT_T32_I8_CSM1((u32*)(g_pbyGSMemory + tex0.cbp*256), (u32*)(g_pbyGSClut+64*tex0.csa));
break;
default:
if( entries == 16 )
WriteCLUT_T16_I4_CSM1((u32*)(g_pbyGSMemory + 256 * tex0.cbp), (u32*)(g_pbyGSClut+32*(tex0.csa&15)+(tex0.csa>=16?2:0)));
else // sse2 for 256 is more complicated, so use regular
WriteCLUT_T16_I8_CSM1_c((u32*)(g_pbyGSMemory + 256 * tex0.cbp), (u32*)(g_pbyGSClut+32*(tex0.csa&15)+(tex0.csa>=16?2:0)));
break;
}
}
}
@ -4870,12 +4905,15 @@ void ZeroGS::SetTexFlush()
// if( PSMT_ISCLUT(vb[1].tex0.psm) )
// texClutWrite(1);
if( !s_bForceTexFlush ) {
if( s_ptexCurSet[0] != s_ptexNextSet[0] ) {
if( !s_bForceTexFlush )
{
if( s_ptexCurSet[0] != s_ptexNextSet[0] )
{
s_ptexCurSet[0] = s_ptexNextSet[0];
pd3dDevice->SetTexture(SAMP_MEMORY0, s_ptexNextSet[0]);
}
if( s_ptexCurSet[1] != s_ptexNextSet[1] ) {
if( s_ptexCurSet[1] != s_ptexNextSet[1] )
{
s_ptexCurSet[1] = s_ptexNextSet[1];
pd3dDevice->SetTexture(SAMP_MEMORY1, s_ptexNextSet[1]);
}
@ -5019,7 +5057,7 @@ bool ZeroGS::StartCapture()
s_aviinit = 1;
}
else {
printf("ZeroGS: Continuing from previous capture");
DEBUG_LOG("ZeroGS: Continuing from previous capture");
}
s_avicapturing = 1;

12
plugins/zerogs/opengl/GS.h
View File

@ -142,7 +142,8 @@ struct Vector_16F
/////////////////////
// define when releasing
#define ZEROGS_CACHEDCLEAR // much better performance
// The only code that uses it is commented out!
//#define ZEROGS_CACHEDCLEAR // much better performance
//#define RELEASE_TO_PUBLIC
#if !defined(_RELEASE)
@ -151,6 +152,9 @@ struct Vector_16F
#define GS_LOG 0&&
#endif
#define ERROR_LOG __LogToConsole
#define DEBUG_LOG printf
#ifdef RELEASE_TO_PUBLIC
#define WARN_LOG 0&&
#define PRIM_LOG 0&&
@ -693,8 +697,8 @@ extern GSinternal gs;
extern FILE *gsLog;
void __Log(char *fmt, ...);
void ERROR_LOG(char *fmt, ...);
void __Log(const char *fmt, ...);
void __LogToConsole(const char *fmt, ...);
void LoadConfig();
void SaveConfig();
@ -803,7 +807,7 @@ inline float Clamp(float fx, float fmin, float fmax)
return fx > fmax ? fmax : fx;
}
// IMPORTANT: For every Reigster there must be an End
// IMPORTANT: For every Register there must be an End
void DVProfRegister(char* pname); // first checks if this profiler exists in g_listProfilers
void DVProfEnd(u32 dwUserData);
void DVProfWrite(char* pfilename, u32 frames = 0);

191
plugins/zerogs/opengl/GSmain.cpp
View File

@ -128,8 +128,7 @@ bool THR_bShift = false;
#endif
#ifdef GS_LOG
void __Log(char *fmt, ...) {
void __Log(const char *fmt, ...) {
va_list list;
if (!conf.log) return;
@ -138,9 +137,8 @@ void __Log(char *fmt, ...) {
vfprintf(gsLog, fmt, list);
va_end(list);
}
#endif
void ERROR_LOG(char *fmt, ...) {
void __LogToConsole(const char *fmt, ...) {
va_list list;
va_start(list, fmt);
@ -167,9 +165,7 @@ void CALLBACK GSsetGameCRC(int crc, int options)
g_GameSettings = conf.gamesettings|options;
conf.mrtdepth = 0;//!(conf.gamesettings&GAME_DISABLEMRTDEPTH);
if( !conf.mrtdepth ) {
ERROR_LOG("Disabling MRT depth writing\n");
}
if( !conf.mrtdepth ) ERROR_LOG("Disabling MRT depth writing\n");
g_GameSettings |= GAME_PATH3HACK;
g_LastCRC = crc;
@ -287,11 +283,7 @@ s32 CALLBACK GSinit()
#endif
GSreset();
#ifdef GS_LOG
GS_LOG("GSinit ok\n");
#endif
return 0;
}
@ -396,7 +388,7 @@ LRESULT WINAPI MsgProc( HWND hWnd, UINT msg, WPARAM wParam, LPARAM lParam )
case WM_ACTIVATE:
if( wParam != WA_INACTIVE ) {
//printf("restoring device\n");
//DEBUG_LOG("restoring device\n");
ZeroGS::Restore();
}
@ -440,9 +432,7 @@ s32 CALLBACK GSopen(void *pDsp, char *Title, int multithread) {
g_GSMultiThreaded = multithread;
#ifdef GS_LOG
GS_LOG("GSopen\n");
#endif
#ifdef _DEBUG
g_hCurrentThread = GetCurrentThread();
@ -478,7 +468,7 @@ s32 CALLBACK GSopen(void *pDsp, char *Title, int multithread) {
if( pDsp != NULL )
*(HWND*)pDsp = GShwnd;
ERROR_LOG("creating\n");
ERROR_LOG("creating zerogs\n");
//if (conf.record) recOpen();
if( !ZeroGS::Create(conf.width, conf.height) )
return -1;
@ -486,10 +476,10 @@ s32 CALLBACK GSopen(void *pDsp, char *Title, int multithread) {
ERROR_LOG("initialization successful\n");
if( conf.bilinear == 2 ) {
ZeroGS::AddMessage("bilinear filtering - forced", 1000);
ZeroGS::AddMessage("forced bilinear filtering - on", 1000);
}
else if( conf.bilinear == 1 ) {
ZeroGS::AddMessage("bilinear filtering - normal", 1000);
ZeroGS::AddMessage("normal bilinear filtering - on", 1000);
}
if( conf.aa ) {
char strtitle[64];
@ -502,15 +492,12 @@ s32 CALLBACK GSopen(void *pDsp, char *Title, int multithread) {
ShowWindow( GShwnd, SW_SHOWDEFAULT );
UpdateWindow( GShwnd );
//SetForegroundWindow(GShwnd);
SetFocus(GShwnd);
conf.winstyle = GetWindowLong( GShwnd, GWL_STYLE );
conf.winstyle &= ~WS_MAXIMIZE & ~WS_MINIMIZE; // remove minimize/maximize style
#ifdef GS_LOG
GS_LOG("GSopen ok\n");
#endif
LARGE_INTEGER temp;
QueryPerformanceFrequency(&temp);
@ -598,9 +585,7 @@ void ProcessMessages()
s32 CALLBACK GSopen(void *pDsp, char *Title, int multithread)
{
#ifdef GS_LOG
GS_LOG("GSopen\n");
#endif
assert( GSirq != NULL );
LoadConfig();
@ -613,7 +598,7 @@ s32 CALLBACK GSopen(void *pDsp, char *Title, int multithread)
if( pDsp != NULL )
*(Display**)pDsp = GLWin.dpy;
ERROR_LOG("creating\n");
ERROR_LOG("creating zerogs\n");
//if (conf.record) recOpen();
if( !ZeroGS::Create(conf.width, conf.height) )
return -1;
@ -632,9 +617,7 @@ s32 CALLBACK GSopen(void *pDsp, char *Title, int multithread)
ZeroGS::AddMessage(strtitle);
}
#ifdef GS_LOG
GS_LOG("GSopen ok\n");
#endif
gs.path1.mode = 0;
gs.path2.mode = 0;
@ -758,9 +741,7 @@ float fFPS = 0;
void CALLBACK GSvsync(int interlace)
{
#ifdef GS_LOG
GS_LOG("\nGSvsync\n\n");
#endif
static u32 dwTime = timeGetTime();
static int nToNextUpdate = 1;
@ -771,10 +752,6 @@ void CALLBACK GSvsync(int interlace)
g_nRealFrame++;
ZeroGS::RenderCRTC(!interlace);
#ifndef _WIN32
//XLockDisplay(GLWin.dpy);
#endif
ProcessMessages();
@ -805,7 +782,7 @@ void CALLBACK GSvsync(int interlace)
#endif
// if( g_nFrame > 100 && fFPS > 60.0f ) {
// printf("set profile\n");
// DEBUG_LOG("set profile\n");
// g_bWriteProfile = 1;
// }
@ -844,29 +821,24 @@ void CALLBACK GSvsync(int interlace)
}
#endif
GL_REPORT_ERRORD();
#ifndef _WIN32
//XUnlockDisplay(GLWin.dpy);
#endif
}
void GIFtag(pathInfo *path, u32 *data) {
path->tag.nloop = data[0] & 0x7fff;
path->tag.eop = (data[0] >> 15) & 0x1;
u32 tagpre = (data[1] >> 14) & 0x1;
u32 tagprim = (data[1] >> 15) & 0x7ff;
u32 tagflg = (data[1] >> 26) & 0x3;
path->tag.nloop = data[0] & 0x7fff;
path->tag.eop = (data[0] >> 15) & 0x1;
u32 tagpre = (data[1] >> 14) & 0x1;
u32 tagprim = (data[1] >> 15) & 0x7ff;
u32 tagflg = (data[1] >> 26) & 0x3;
path->tag.nreg = (data[1] >> 28)<<2;
if (path->tag.nreg == 0) path->tag.nreg = 64;
gs.q = 1;
#ifdef GS_LOG
// GS_LOG("GIFtag: %8.8lx_%8.8lx_%8.8lx_%8.8lx: EOP=%d, NLOOP=%x, FLG=%x, NREG=%d, PRE=%d\n",
// data[3], data[2], data[1], data[0],
// path->tag.eop, path->tag.nloop, tagflg, path->tag.nreg, tagpre);
#endif
path->mode = tagflg+1;
@ -915,9 +887,6 @@ static int nPath3Hack = 0;
void CALLBACK GSgetLastTag(u64* ptag)
{
// int mode = s_pLastPath->mode > 0 ? s_pLastPath->mode-1 : 0;
// *(u32*)ptag = s_pLastPath->tag.nloop|(s_pLastPath->tag.eop<<15);
// *((u32*)ptag+1) = (mode<<26)|(s_pLastPath->regn<<28);
*(u32*)ptag = nPath3Hack;
nPath3Hack = 0;
}
@ -931,47 +900,36 @@ void _GSgifTransfer(pathInfo *path, u32 *pMem, u32 size)
#ifdef _DEBUG
if( conf.log & 0x20 ) {
static int nSaveIndex=0;
__Log("%d: p:%d %x\n", nSaveIndex++, (path==&gs.path3)?3:(path==&gs.path2?2:1), size);
GS_LOG("%d: p:%d %x\n", nSaveIndex++, (path==&gs.path3)?3:(path==&gs.path2?2:1), size);
int vals[4] = {0};
for(int i = 0; i < size; i++) {
for(int j = 0; j < 4; ++j )
vals[j] ^= pMem[4*i+j];
}
__Log("%x %x %x %x\n", vals[0], vals[1], vals[2], vals[3]);
}
#endif
//s_pLastPath = path;
#ifdef _DEBUG
if( conf.log & 0x20 ) {
static int nSaveIndex=0;
__Log("%d: p:%d %x\n", nSaveIndex++, (path==&gs.path3)?3:(path==&gs.path2?2:1), size);
int vals[4] = {0};
for(int i = 0; i < size; ++i) {
for(int j = 0; j < 4; ++j )
vals[j] ^= pMem[4*i+j];
}
__Log("%x %x %x %x\n", vals[0], vals[1], vals[2], vals[3]);
GS_LOG("%x %x %x %x\n", vals[0], vals[1], vals[2], vals[3]);
}
#endif
while(size > 0)
{
//LOG(_T("Transfer(%08x, %d) START\n"), pMem, size);
if(path->tag.nloop == 0) {
if (path->tag.nloop == 0)
{
GIFtag(path, pMem);
pMem+= 4;
size--;
if( (g_GameSettings & GAME_PATH3HACK) && path == &gs.path3 && gs.path3.tag.eop )
if ((g_GameSettings & GAME_PATH3HACK) && path == &gs.path3 && gs.path3.tag.eop)
nPath3Hack = 1;
if( path == &gs.path1 ) {
if (path == &gs.path1)
{
// if too much data, just ignore
if( path->tag.nloop * (path->tag.nreg / 4) > (int)size * (path->mode==2?2:1)) {
if (path->tag.nloop * (path->tag.nreg / 4) > (int)size * (path->mode==2?2:1))
{
static int lasttime = 0;
if( timeGetTime() - lasttime > 5000 ) {
if( timeGetTime() - lasttime > 5000 )
{
ERROR_LOG("VU1 too much data, ignore if gfx are fine\n");
lasttime = timeGetTime();
}
@ -979,13 +937,16 @@ void _GSgifTransfer(pathInfo *path, u32 *pMem, u32 size)
return;
}
if( path->mode == 1 ) {
if (path->mode == 1)
{
// check if 0xb is in any reg, if yes, exit (kh2)
for(int i = 0; i < path->tag.nreg; i += 4) {
if( ((path->regs >> i)&0xf) == 11 ) {
for(int i = 0; i < path->tag.nreg; i += 4)
{
if (((path->regs >> i)&0xf) == 11)
{
static int lasttime = 0;
if( timeGetTime() - lasttime > 5000 ) {
if( timeGetTime() - lasttime > 5000 )
{
ERROR_LOG("Invalid unpack type\n");
lasttime = timeGetTime();
}
@ -996,11 +957,13 @@ void _GSgifTransfer(pathInfo *path, u32 *pMem, u32 size)
}
}
if(path->tag.nloop == 0 ) {
if( path == &gs.path1 ) {
if(path->tag.nloop == 0 )
{
if( path == &gs.path1 )
{
// ffx hack
if( g_GameSettings & GAME_FFXHACK ) {
if( g_GameSettings & GAME_FFXHACK )
{
if( path->tag.eop )
return;
continue;
@ -1009,8 +972,9 @@ void _GSgifTransfer(pathInfo *path, u32 *pMem, u32 size)
return;
}
if( !path->tag.eop ) {
//printf("contuing from eop\n");
if( !path->tag.eop )
{
//DEBUG_LOG("continuing from eop\n");
continue;
}
@ -1025,12 +989,15 @@ void _GSgifTransfer(pathInfo *path, u32 *pMem, u32 size)
for(; size > 0; size--, pMem += 4)
{
int reg = (int)((path->regs >> path->regn) & 0xf);
g_GIFPackedRegHandlers[reg](pMem);
path->regn += 4;
if (path->tag.nreg == path->regn) {
if (path->tag.nreg == path->regn)
{
path->regn = 0;
if( path->tag.nloop-- <= 1 ) {
if( path->tag.nloop-- <= 1 )
{
size--;
pMem += 4;
break;
@ -1041,7 +1008,7 @@ void _GSgifTransfer(pathInfo *path, u32 *pMem, u32 size)
}
case 2: // REGLIST
{
//__Log("%8.8x%8.8x %d L\n", ((u32*)&gs.regs)[1], *(u32*)&gs.regs, path->tag.nreg/4);
//GS_LOG("%8.8x%8.8x %d L\n", ((u32*)&gs.regs)[1], *(u32*)&gs.regs, path->tag.nreg/4);
assert( path->tag.nloop > 0 );
size *= 2;
for(; size > 0; pMem+= 2, size--)
@ -1049,9 +1016,11 @@ void _GSgifTransfer(pathInfo *path, u32 *pMem, u32 size)
int reg = (int)((path->regs >> path->regn) & 0xf);
g_GIFRegHandlers[reg](pMem);
path->regn += 4;
if (path->tag.nreg == path->regn) {
if (path->tag.nreg == path->regn)
{
path->regn = 0;
if( path->tag.nloop-- <= 1 ) {
if( path->tag.nloop-- <= 1 )
{
size--;
pMem += 2;
break;
@ -1070,9 +1039,12 @@ void _GSgifTransfer(pathInfo *path, u32 *pMem, u32 size)
{
int process = min((int)size, path->tag.nloop);
if( process > 0 ) {
if( gs.imageTransfer ) ZeroGS::TransferLocalHost(pMem, process);
else ZeroGS::TransferHostLocal(pMem, process*4);
if( process > 0 )
{
if ( gs.imageTransfer )
ZeroGS::TransferLocalHost(pMem, process);
else
ZeroGS::TransferHostLocal(pMem, process*4);
path->tag.nloop -= process;
pMem += process*4; size -= process;
@ -1081,7 +1053,8 @@ void _GSgifTransfer(pathInfo *path, u32 *pMem, u32 size)
}
break;
}
else {
else
{
// simulate
int process = min((int)size, path->tag.nloop);
path->tag.nloop -= process;
@ -1093,9 +1066,6 @@ void _GSgifTransfer(pathInfo *path, u32 *pMem, u32 size)
default: // GIF_IMAGE
GS_LOG("*** WARNING **** Unexpected GIFTag flag\n");
assert(0);
// ZeroGS::TransferLocalHost(pMem, size);
// pMem+= size*4; path->tag.nloop-= size;
// size = 0; path->mode = 0;
path->tag.nloop = 0;
break;
}
@ -1107,47 +1077,31 @@ void _GSgifTransfer(pathInfo *path, u32 *pMem, u32 size)
void CALLBACK GSgifTransfer2(u32 *pMem, u32 size)
{
#ifdef GS_LOG
//GS_LOG("GSgifTransfer2 size = %lx (mode %d, gs.path2.tag.nloop = %d)\n", size, gs.path2.mode, gs.path2.tag.nloop);
#endif
// if(!g_GSMultiThreaded)
// CSR->FINISH = 0;
//DVProfileFunc _pf("Transf2");
//assert( ((u32)pMem & 0xf) == 0 );
_GSgifTransfer(&gs.path2, pMem, size);
}
void CALLBACK GSgifTransfer3(u32 *pMem, u32 size)
{
#ifdef GS_LOG
//GS_LOG("GSgifTransfer3 size = %lx (mode %d, gs.path3.tag.nloop = %d)\n", size, gs.path3.mode, gs.path3.tag.nloop);
#endif
nPath3Hack = 0;
_GSgifTransfer(&gs.path3, pMem, size);
}
static int s_trcount = 0;
static int count = 0;
void CALLBACK GSgifTransfer1(u32 *pMem, u32 addr)
{
pathInfo *path = &gs.path1;
#ifdef GS_LOG
//GS_LOG("GSgifTransfer1 0x%x (mode %d)\n", addr, path->mode);
#endif
addr &= 0x3fff;
#ifdef _DEBUG
PRIM_LOG("count :%d\n", s_trcount);
s_trcount++;
// for(int i = addr; i < 0x4000; i += 16 ) {
// u32* mem = (u32*)((u8*)pMem+i);
// PRIM_LOG("%x: %x %x %x %x\n", i, mem[0], mem[1], mem[2], mem[3]);
// }
PRIM_LOG("count: %d\n", count);
count++;
#endif
gs.path1.tag.nloop = 0;
@ -1158,25 +1112,20 @@ void CALLBACK GSgifTransfer1(u32 *pMem, u32 addr)
assert( (addr&0xf) == 0 ); //BUG
gs.path1.tag.nloop = 0;
ERROR_LOG("Transfer1 - 2\n");
//_GSgifTransfer(&gs.path1, (u32*)((u8*)pMem+0x4000-addr), addr/16);
return;
}
}
void CALLBACK GSreadFIFO(u64 *pMem)
{
#ifdef GS_LOG
//GS_LOG("GSreadFIFO\n");
#endif
ZeroGS::TransferLocalHost((u32*)pMem, 1);
}
void CALLBACK GSreadFIFO2(u64 *pMem, int qwc)
{
#ifdef GS_LOG
//GS_LOG("GSreadFIFO2\n");
#endif
ZeroGS::TransferLocalHost((u32*)pMem, qwc);
}
@ -1188,14 +1137,14 @@ int CALLBACK GSsetupRecording(int start, void* pData)
return 1;
ZeroGS::StartCapture();
conf.options |= GSOPTION_CAPTUREAVI;
printf("ZeroGS: started recording at zerogs.avi\n");
WARN_LOG("ZeroGS: started recording at zerogs.avi\n");
}
else {
if( !(conf.options & GSOPTION_CAPTUREAVI) )
return 1;
conf.options &= ~GSOPTION_CAPTUREAVI;
ZeroGS::StopCapture();
printf("ZeroGS: stopped recording\n");
WARN_LOG("ZeroGS: stopped recording\n");
}
return 1;
@ -1227,7 +1176,7 @@ using namespace std;
#ifdef _WIN32
inline u64 GET_PROFILE_TIME()
__forceinline u64 GET_PROFILE_TIME()
{
LARGE_INTEGER lu;
QueryPerformanceCounter(&lu);

2
plugins/zerogs/opengl/Linux/Linux.cpp
View File

@ -38,7 +38,7 @@ static int prevbilinearfilter;
//static map<string, int> mapConfOpts;
struct confOptsStruct{
int value;
char *desc;
const char *desc;
}confOpts;
static map<string, confOptsStruct> mapConfOpts;

4
plugins/zerogs/opengl/Mem.cpp
View File

@ -530,7 +530,7 @@ int TransferHostLocal##psm(const void* pbyMem, u32 nQWordSize) \
int testwidth = (int)nSize - (gs.imageEndY-i)*(gs.imageEndX-gs.trxpos.dx)+(j-gs.trxpos.dx); \
if( testwidth <= widthlimit && testwidth >= -widthlimit ) { \
/* don't transfer */ \
/*printf("bad texture %s: %d %d %d\n", #psm, gs.trxpos.dx, gs.imageEndX, nQWordSize);*/ \
/*DEBUG_LOG("bad texture %s: %d %d %d\n", #psm, gs.trxpos.dx, gs.imageEndX, nQWordSize);*/ \
gs.imageTransfer = -1; \
} \
bCanAlign = false; \
@ -638,7 +638,7 @@ DEFINE_TRANSFERLOCAL(4HH, u8, 8, 32, 8, 8, _4, SwizzleBlock4HH);
// /* hack */
// if( abs((int)nSize - (gs.imageEndY-i)*(gs.imageEndX-gs.trxpos.dx)+(j-gs.trxpos.dx)) <= widthlimit ) {
// /* don't transfer */
// /*printf("bad texture %s: %d %d %d\n", #psm, gs.trxpos.dx, gs.imageEndX, nQWordSize);*/
// /*DEBUG_LOG("bad texture %s: %d %d %d\n", #psm, gs.trxpos.dx, gs.imageEndX, nQWordSize);*/
// gs.imageTransfer = -1;
// }
// bCanAlign = false;

14
plugins/zerogs/opengl/Regs.cpp
View File

@ -75,9 +75,7 @@ u32 s_uTex1Data[2][2] = {0}, s_uClampData[2] = {0};
void __fastcall GIFPackedRegHandlerNull(u32* data)
{
#ifdef _DEBUG
printf("Unexpected packed reg handler %8.8lx_%8.8lx %x\n", data[0], data[1], data[2]);
#endif
DEBUG_LOG("Unexpected packed reg handler %8.8lx_%8.8lx %x\n", data[0], data[1], data[2]);
}
void __fastcall GIFPackedRegHandlerRGBA(u32* data)
@ -122,7 +120,7 @@ void __forceinline KICK_VERTEX3()
{
/* tri fans need special processing */
if (gs.nTriFanVert == gs.primIndex)
gs.primIndex = (gs.primIndex+1)%ARRAY_SIZE(gs.gsvertex);
gs.primIndex = (gs.primIndex+1)%ARRAYSIZE(gs.gsvertex);
}
}
}
@ -134,7 +132,7 @@ void __fastcall GIFPackedRegHandlerXYZF2(u32* data)
gs.vertexregs.z = (data[2] >> 4) & 0xffffff;
gs.vertexregs.f = (data[3] >> 4) & 0xff;
gs.gsvertex[gs.primIndex] = gs.vertexregs;
gs.primIndex = (gs.primIndex+1)%ARRAY_SIZE(gs.gsvertex);
gs.primIndex = (gs.primIndex+1)%ARRAYSIZE(gs.gsvertex);
if( data[3] & 0x8000 ) {
KICK_VERTEX3();
@ -150,7 +148,7 @@ void __fastcall GIFPackedRegHandlerXYZ2(u32* data)
gs.vertexregs.y = (data[1] >> 0) & 0xffff;
gs.vertexregs.z = data[2];
gs.gsvertex[gs.primIndex] = gs.vertexregs;
gs.primIndex = (gs.primIndex+1)%ARRAY_SIZE(gs.gsvertex);
gs.primIndex = (gs.primIndex+1)%ARRAYSIZE(gs.gsvertex);
if( data[3] & 0x8000 ) {
KICK_VERTEX3();
@ -371,7 +369,7 @@ void __fastcall GIFRegHandlerNull(u32* data)
// 0x7f happens on a lot of games
if( data[2] != 0x7f && (data[0] || data[1]) ) {
printf("Unexpected reg handler %x %x %x\n", data[0], data[1], data[2]);
DEBUG_LOG("Unexpected reg handler %x %x %x\n", data[0], data[1], data[2]);
}
#endif
}
@ -379,9 +377,7 @@ void __fastcall GIFRegHandlerNull(u32* data)
void __fastcall GIFRegHandlerPRIM(u32 *data)
{
if (data[0] & ~0x3ff) {
#ifdef WARN_LOG
//WARN_LOG("warning: unknown bits in prim %8.8lx_%8.8lx\n", data[1], data[0]);
#endif
}
gs.nTriFanVert = gs.primIndex;

20
plugins/zerogs/opengl/common.h
View File

@ -338,13 +338,13 @@ static inline void put_bits(PutBitContext *s, int n, unsigned int value)
#ifdef STATS
st_out_bit_counts[st_current_index] += n;
#endif
// printf("put_bits=%d %x\n", n, value);
// DEBUG_LOG("put_bits=%d %x\n", n, value);
assert(n == 32 || value < (1U << n));
bit_buf = s->bit_buf;
bit_left = s->bit_left;
// printf("n=%d value=%x cnt=%d buf=%x\n", n, value, bit_cnt, bit_buf);
// DEBUG_LOG("n=%d value=%x cnt=%d buf=%x\n", n, value, bit_cnt, bit_buf);
/* XXX: optimize */
if (n < bit_left) {
bit_buf = (bit_buf<<n) | value;
@ -361,7 +361,7 @@ static inline void put_bits(PutBitContext *s, int n, unsigned int value)
} else
#endif
*(uint32_t *)s->buf_ptr = be2me_32(bit_buf);
//printf("bitbuf = %08x\n", bit_buf);
//DEBUG_LOG("bitbuf = %08x\n", bit_buf);
s->buf_ptr+=4;
bit_left+=32 - n;
bit_buf = value;
@ -403,7 +403,7 @@ static inline void put_bits(PutBitContext *s, int n, unsigned int value)
ptr[0] |= be2me_32(value>>(index&31));
ptr[1] = be2me_32(value<<(32-(index&31)));
//if(n>24) printf("%d %d\n", n, value);
//if(n>24) DEBUG_LOG("%d %d\n", n, value);
index+= n;
s->index= index;
# endif
@ -431,7 +431,7 @@ static inline void put_bits(PutBitContext *s, int n, unsigned int value)
ptr[0] |= be2me_32(value<<(32-n-(index&7) ));
ptr[1] = 0;
//if(n>24) printf("%d %d\n", n, value);
//if(n>24) DEBUG_LOG("%d %d\n", n, value);
index+= n;
s->index= index;
# endif
@ -885,17 +885,17 @@ static inline void print_bin(int bits, int n){
int i;
for(i=n-1; i>=0; i--){
printf("%d", (bits>>i)&1);
DEBUG_LOG("%d", (bits>>i)&1);
}
for(i=n; i<24; i++)
printf(" ");
DEBUG_LOG(" ");
}
static inline int get_bits_trace(GetBitContext *s, int n, char *file, char *func, int line){
int r= get_bits(s, n);
print_bin(r, n);
printf("%5d %2d %3d bit @%5d in %s %s:%d\n", r, n, r, get_bits_count(s)-n, file, func, line);
DEBUG_LOG("%5d %2d %3d bit @%5d in %s %s:%d\n", r, n, r, get_bits_count(s)-n, file, func, line);
return r;
}
static inline int get_vlc_trace(GetBitContext *s, VLC_TYPE (*table)[2], int bits, int max_depth, char *file, char *func, int line){
@ -907,7 +907,7 @@ static inline int get_vlc_trace(GetBitContext *s, VLC_TYPE (*table)[2], int bits
print_bin(bits2, len);
printf("%5d %2d %3d vlc @%5d in %s %s:%d\n", bits2, len, r, pos, file, func, line);
DEBUG_LOG("%5d %2d %3d vlc @%5d in %s %s:%d\n", bits2, len, r, pos, file, func, line);
return r;
}
static inline int get_xbits_trace(GetBitContext *s, int n, char *file, char *func, int line){
@ -915,7 +915,7 @@ static inline int get_xbits_trace(GetBitContext *s, int n, char *file, char *fun
int r= get_xbits(s, n);
print_bin(show, n);
printf("%5d %2d %3d xbt @%5d in %s %s:%d\n", show, n, r, get_bits_count(s)-n, file, func, line);
DEBUG_LOG("%5d %2d %3d xbt @%5d in %s %s:%d\n", show, n, r, get_bits_count(s)-n, file, func, line);
return r;
}

582
plugins/zerogs/opengl/zerogs.cpp
View File

File diff suppressed because it is too large Load Diff