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

reverted some stuff for now, this should fix what r79 broke. plz comment if it doesn't fix the problems. 

git-svn-id: http://pcsx2-playground.googlecode.com/svn/trunk@85 a6443dda-0b58-4228-96e9-037be469359c
This commit is contained in:
cottonvibes 2008-09-01 20:27:22 +00:00 committed by Gregory Hainaut
parent a490bdb7cf
commit e78cf6c9b6
1 changed files with 76 additions and 102 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

162
pcsx2/x86/iFPU.c
View File

@ -752,7 +752,7 @@ FPURECOMPILE_CONSTCODE(C_LE, XMMINFO_READS|XMMINFO_READT);
// Doesnt seem to like negatives - Ruins katamari graphics
// I REPEAT THE SIGN BIT (THATS 0x80000000) MUST *NOT* BE SET, jeez.
static PCSX2_ALIGNED16(u32 s_overflowmask[]) = {0x7fffffff, 0x7fffffff, 0x7fffffff, 0x7fffffff};
static PCSX2_ALIGNED16(u32 s_overflowmask[]) = {0x7f7fffff, 0x7f7fffff, 0x7f7fffff, 0x7f7fffff};
static u32 s_signbit = 0x80000000;
extern int g_VuNanHandling;
@ -790,9 +790,9 @@ void ClampValues2(regd) {
SSE_ANDPS_XMM_to_XMM(regd, t5reg);
// clamp infinities
//SSE_MAXSS_M32_to_XMM(regd, (uptr)&g_minvals[0]);
SSE_MINSS_M32_to_XMM(regd, (uptr)&g_maxvals[0]); // after above calculations, unordered floats will be positive
// not necessary since above ORPS handles that (i think) Lets enable it for now ;)
SSE_MAXSS_M32_to_XMM(regd, (uptr)&g_minvals[0]);
SSE_MINSS_M32_to_XMM(regd, (uptr)&g_maxvals[0]);
_freeXMMreg(t5reg);
}
@ -811,67 +811,23 @@ static void (*recComOpM32_to_XMM[] )(x86SSERegType, uptr) = {
SSE_ADDSS_M32_to_XMM, SSE_MULSS_M32_to_XMM, SSE_MAXSS_M32_to_XMM, SSE_MINSS_M32_to_XMM };
int recCommutativeOp(int info, int regd, int op) {
if (CHECK_FPU_EXTRA_OVERFLOW) {
int t0reg = _allocTempXMMreg(XMMT_FPS, -1);
if (t0reg < 0) goto allocationError;
switch(info & (PROCESS_EE_S|PROCESS_EE_T) ) {
case PROCESS_EE_S:
if (regd != EEREC_S) SSE_MOVSS_XMM_to_XMM(regd, EEREC_S);
SSE_MOVSS_M32_to_XMM(t0reg, (uptr)&fpuRegs.fpr[_Ft_]);
ClampValues (regd);
ClampValues (t0reg);
recComOpXMM_to_XMM[op](regd, t0reg);
break;
case PROCESS_EE_T:
if (regd != EEREC_T) SSE_MOVSS_XMM_to_XMM(regd, EEREC_T);
SSE_MOVSS_M32_to_XMM(t0reg, (uptr)&fpuRegs.fpr[_Fs_]);
ClampValues (regd);
ClampValues (t0reg);
recComOpXMM_to_XMM[op](regd, t0reg);
break;
case (PROCESS_EE_S|PROCESS_EE_T):
if (regd == EEREC_S) {
ClampValues (regd);
ClampValues (EEREC_T);
recComOpXMM_to_XMM[op](regd, EEREC_T);
}
else if (regd == EEREC_T) {
ClampValues (regd);
ClampValues (EEREC_S);
recComOpXMM_to_XMM[op](regd, EEREC_S);
}
if (regd == EEREC_S) recComOpM32_to_XMM[op](regd, (uptr)&fpuRegs.fpr[_Ft_]);
else {
ClampValues (EEREC_S);
ClampValues (EEREC_T);
SSE_MOVSS_XMM_to_XMM(regd, EEREC_S);
recComOpXMM_to_XMM[op](regd, EEREC_T);
}
break;
default:
SSE_MOVSS_M32_to_XMM(regd, (uptr)&fpuRegs.fpr[_Fs_]);
SSE_MOVSS_M32_to_XMM(t0reg, (uptr)&fpuRegs.fpr[_Ft_]);
ClampValues (regd);
ClampValues (t0reg);
recComOpXMM_to_XMM[op](regd, t0reg);
break;
}
_freeXMMreg(t0reg);
return regd;
allocationError:
SysPrintf("recCommutativeOp() allocation error! Skipping Pre-Opcode Overflow checks! \n");
} // End of pre-opcode overflow checking
switch(info & (PROCESS_EE_S|PROCESS_EE_T) ) {
case PROCESS_EE_S:
if (regd != EEREC_S) SSE_MOVSS_XMM_to_XMM(regd, EEREC_S);
recComOpM32_to_XMM[op](regd, (uptr)&fpuRegs.fpr[_Ft_]);
}
break;
case PROCESS_EE_T:
if (regd != EEREC_T) SSE_MOVSS_XMM_to_XMM(regd, EEREC_T);
if (regd == EEREC_T) recComOpM32_to_XMM[op](regd, (uptr)&fpuRegs.fpr[_Fs_]);
else {
SSE_MOVSS_XMM_to_XMM(regd, EEREC_T);
recComOpM32_to_XMM[op](regd, (uptr)&fpuRegs.fpr[_Fs_]);
}
break;
case (PROCESS_EE_S|PROCESS_EE_T):
// SysPrintf("Hello2 :)\n");
if (regd == EEREC_S) recComOpXMM_to_XMM[op](regd, EEREC_T);
else if (regd == EEREC_T) recComOpXMM_to_XMM[op](regd, EEREC_S);
else {
@ -880,6 +836,17 @@ allocationError:
}
break;
default:
SysPrintf("But we dont have regs2 :(\n");
/*if (regd == EEREC_S) {
recComOpXMM_to_XMM[op](regd, EEREC_T);
}
else if (regd == EEREC_T) {
recComOpXMM_to_XMM[op](regd, EEREC_S);
}
else {
SSE_MOVSS_XMM_to_XMM(regd, EEREC_S);
recComOpXMM_to_XMM[op](regd, EEREC_T);
}*/
SSE_MOVSS_M32_to_XMM(regd, (uptr)&fpuRegs.fpr[_Fs_]);
recComOpM32_to_XMM[op](regd, (uptr)&fpuRegs.fpr[_Ft_]);
break;
@ -996,18 +963,32 @@ void recSQRT_S_xmm(int info)
{
SysPrintf("FPU: SQRT \n");
if( info & PROCESS_EE_T ) {
if (CHECK_FPU_EXTRA_OVERFLOW) { ClampValues(EEREC_T); }
if( EEREC_D != EEREC_T ) SSE_MOVSS_XMM_to_XMM(EEREC_D, EEREC_T);
//if( CHECK_OVERFLOW ) {
if( EEREC_D == EEREC_T ) SSE_ANDPS_M128_to_XMM(EEREC_D, (uptr)&s_pos[0]);
else {
SSE_MOVSS_XMM_to_XMM(EEREC_D, EEREC_T);
SSE_ANDPS_M128_to_XMM(EEREC_D, (uptr)&s_pos[0]);
}
SSE_SQRTSS_XMM_to_XMM(EEREC_D, EEREC_D);
//}
/*else {
SSE_SQRTSS_XMM_to_XMM(EEREC_D, EEREC_T);
}*/
}
else {
//if( CHECK_OVERFLOW ) {
SSE_MOVSS_M32_to_XMM(EEREC_D, (uptr)&fpuRegs.fpr[_Ft_]);
SSE_ANDPS_M128_to_XMM(EEREC_D, (uptr)&s_pos[0]);
if (CHECK_FPU_EXTRA_OVERFLOW) { ClampValues(EEREC_D); }
SSE_SQRTSS_XMM_to_XMM(EEREC_D, EEREC_D);
/*}
else {
SSE_SQRTSS_M32_to_XMM(EEREC_D, (uptr)&fpuRegs.fpr[_Ft_]);
}*/
}
//ClampValues(EEREC_D); // No need to clamp since sqrt of a number is always smaller than that number
ClampValues(EEREC_D);
}
FPURECOMPILE_CONSTCODE(SQRT_S, XMMINFO_WRITED|XMMINFO_READT);
@ -1022,6 +1003,7 @@ void recABS_S_xmm(int info)
else {
SSE_MOVSS_M32_to_XMM(EEREC_D, (uptr)&fpuRegs.fpr[_Fs_]);
SSE_ANDPS_M128_to_XMM(EEREC_D, (uptr)&s_pos[0]);
//xmmregs[EEREC_D].mode &= ~MODE_WRITE;
}
ClampValues(EEREC_D);
}
@ -1059,51 +1041,43 @@ void recRSQRT_S_xmm(int info)
int t0reg = _allocTempXMMreg(XMMT_FPS, -1);
switch(info & (PROCESS_EE_S|PROCESS_EE_T) ) {
case PROCESS_EE_S:
SSE_MOVSS_M32_to_XMM(t0reg, (uptr)&fpuRegs.fpr[_Ft_]);
SSE_ANDPS_M128_to_XMM(t0reg, (uptr)&s_pos[0]);
if (CHECK_FPU_EXTRA_OVERFLOW) {
ClampValues(EEREC_S);
SSE_MINSS_M32_to_XMM(t0reg, (uptr)&g_maxvals[0]);
}
if( EEREC_D != EEREC_S ) SSE_MOVSS_XMM_to_XMM(EEREC_D, EEREC_S);
SSE_SQRTSS_XMM_to_XMM(t0reg, t0reg);
if( EEREC_D == EEREC_S ) {
SSE_SQRTSS_M32_to_XMM(t0reg, (uptr)&fpuRegs.fpr[_Ft_]);
SSE_DIVSS_XMM_to_XMM(EEREC_D, t0reg);
}
else {
SSE_SQRTSS_M32_to_XMM(t0reg, (uptr)&fpuRegs.fpr[_Ft_]);
SSE_MOVSS_XMM_to_XMM(EEREC_D, EEREC_S);
SSE_DIVSS_XMM_to_XMM(EEREC_D, t0reg);
}
break;
case PROCESS_EE_T:
SSE_SQRTSS_XMM_to_XMM(t0reg, EEREC_T);
SSE_MOVSS_M32_to_XMM(EEREC_D, (uptr)&fpuRegs.fpr[_Fs_]);
SSE_MOVSS_XMM_to_XMM(t0reg, EEREC_T);
SSE_ANDPS_M128_to_XMM(t0reg, (uptr)&s_pos[0]);
if (CHECK_FPU_EXTRA_OVERFLOW) {
ClampValues(EEREC_D);
SSE_MINSS_M32_to_XMM(t0reg, (uptr)&g_maxvals[0]);
}
SSE_SQRTSS_XMM_to_XMM(t0reg, t0reg);
SSE_DIVSS_XMM_to_XMM(EEREC_D, t0reg);
break;
case (PROCESS_EE_S | PROCESS_EE_T):
SSE_MOVSS_XMM_to_XMM(t0reg, EEREC_T);
SSE_ANDPS_M128_to_XMM(t0reg, (uptr)&s_pos[0]);
if (CHECK_FPU_EXTRA_OVERFLOW) {
ClampValues(EEREC_S);
SSE_MINSS_M32_to_XMM(t0reg, (uptr)&g_maxvals[0]);
}
if( EEREC_D != EEREC_S ) SSE_MOVSS_XMM_to_XMM(EEREC_D, EEREC_S);
SSE_SQRTSS_XMM_to_XMM(t0reg, t0reg);
SSE_DIVSS_XMM_to_XMM(EEREC_D, t0reg);
break;
default:
SSE_MOVSS_M32_to_XMM(t0reg, (uptr)&fpuRegs.fpr[_Ft_]);
SSE_ANDPS_M128_to_XMM(t0reg, (uptr)&s_pos[0]);
SSE_MOVSS_M32_to_XMM(EEREC_D, (uptr)&fpuRegs.fpr[_Fs_]);
if (CHECK_FPU_EXTRA_OVERFLOW) {
ClampValues(EEREC_D);
SSE_MINSS_M32_to_XMM(t0reg, (uptr)&g_maxvals[0]);
}
SSE_SQRTSS_XMM_to_XMM(t0reg, t0reg);
if( (info & PROCESS_EE_T) && (info & PROCESS_EE_S) ) {
if( EEREC_D == EEREC_T ){
SSE_SQRTSS_XMM_to_XMM(t0reg, EEREC_T);
SSE_MOVSS_XMM_to_XMM(EEREC_D, EEREC_S);
SSE_DIVSS_XMM_to_XMM(EEREC_D, t0reg);
}
else if( EEREC_D == EEREC_S ){
SSE_SQRTSS_XMM_to_XMM(t0reg, EEREC_T);
SSE_DIVSS_XMM_to_XMM(EEREC_D, t0reg);
} else {
SSE_SQRTSS_XMM_to_XMM(t0reg, EEREC_T);
SSE_MOVSS_XMM_to_XMM(EEREC_D, EEREC_S);
SSE_DIVSS_XMM_to_XMM(EEREC_D, t0reg);
}
}else{
SSE_SQRTSS_M32_to_XMM(t0reg, (uptr)&fpuRegs.fpr[_Ft_]);
SSE_MOVSS_M32_to_XMM(EEREC_D, (uptr)&fpuRegs.fpr[_Fs_]);
SSE_DIVSS_XMM_to_XMM(EEREC_D, t0reg);
}
break;
}