DSPLLE trying to fix mul

git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@4021 8ced0084-cf51-0410-be5f-012b33b47a6e
This commit is contained in:
nakeee 2009-08-20 15:11:07 +00:00
parent fdc141e333
commit 3c59b8bded
1 changed files with 76 additions and 85 deletions

View File

@ -27,13 +27,12 @@
namespace DSPInterpreter {
// Sets prod as a side effect.
// Only MULX family instructions have unsigned support.
s64 dsp_multiply_conditional_unsigned(u16 a, u16 b)
inline s64 dsp_get_multiply_prod(u16 a, u16 b, bool sign)
{
s64 prod;
// FIXME: Makes AX games sound horrible. Verified through DSPSpy :/
if (g_dsp.r[DSP_REG_SR] & SR_MUL_UNSIGNED)
if (sign && g_dsp.r[DSP_REG_SR] & SR_MUL_UNSIGNED)
prod = (u64)a * (u64)b; // won't overflow 32-bits
else
prod = (s32)(s16)a * (s32)(s16)b; // won't overflow 32-bits
@ -42,55 +41,34 @@ s64 dsp_multiply_conditional_unsigned(u16 a, u16 b)
if ((g_dsp.r[DSP_REG_SR] & SR_MUL_MODIFY) == 0)
prod <<= 1;
// Store the product, and return it, in case the caller wants to read it.
dsp_set_long_prod(prod);
return prod;
}
// Sets prod as a side effect.
s64 dsp_multiply(u16 a, u16 b)
s64 dsp_multiply(u16 a, u16 b, bool sign = false)
{
s64 prod;
prod = (s32)(s16)a * (s32)(s16)b; // won't overflow 32-bits
// Conditionally multiply by 2.
if ((g_dsp.r[DSP_REG_SR] & SR_MUL_MODIFY) == 0)
prod <<= 1;
s64 prod = dsp_get_multiply_prod(a, b, sign);
// Store the product, and return it, in case the caller wants to read it.
dsp_set_long_prod(prod);
// dsp_set_long_prod(prod);
return prod;
}
s64 dsp_multiply_add(u16 a, u16 b)
s64 dsp_multiply_add(u16 a, u16 b, bool sign = false)
{
s64 prod = (s32)(s16)a * (s32)(s16)b; // won't overflow 32-bits
// Conditionally multiply by 2.
if ((g_dsp.r[DSP_REG_SR] & SR_MUL_MODIFY) == 0)
prod <<= 1;
// Add the original prod value.
prod += dsp_get_long_prod();
s64 prod = dsp_get_multiply_prod(a, b, sign) + dsp_get_long_prod();
// Store the product, and return it, in case the caller wants to read it.
dsp_set_long_prod(prod);
// dsp_set_long_prod(prod);
return prod;
}
s64 dsp_multiply_sub(u16 a, u16 b)
s64 dsp_multiply_sub(u16 a, u16 b, bool sign = false)
{
s64 prod = (s32)(s16)a * (s32)(s16)b; // won't overflow 32-bits
// Conditionally multiply by 2.
if ((g_dsp.r[DSP_REG_SR] & SR_MUL_MODIFY) == 0)
prod <<= 1;
// Subtract from the original prod value.
prod = dsp_get_long_prod() - prod;
s64 prod = dsp_get_long_prod() - dsp_get_multiply_prod(a, b, sign);
// Store the product, and return it, in case the caller wants to read it.
dsp_set_long_prod(prod);
// dsp_set_long_prod(prod);
return prod;
}
@ -188,10 +166,11 @@ void mulc(const UDSPInstruction& opc)
u8 treg = (opc.hex >> 12) & 0x1;
u16 accm = dsp_get_acc_m(sreg);
u16 axh = dsp_get_ax_h(treg);
s64 prod = dsp_multiply(accm, axh);
zeroWriteBackLog();
s64 prod = dsp_multiply(accm, axh);
dsp_set_long_prod(prod);
Update_SR_Register64(prod);
}
@ -212,10 +191,11 @@ void mulcmvz(const UDSPInstruction& opc)
u16 accm = dsp_get_acc_m(sreg);
u16 axh = dsp_get_ax_h(treg);
s64 prod = dsp_multiply(accm, axh);
zeroWriteBackLog();
dsp_set_long_prod(prod);
dsp_multiply(accm, axh);
// update acc
u8 rreg = (opc.hex >> 8) & 0x1;
s64 acc = TempProd & ~0xffff; // clear lower 4 bytes
@ -239,13 +219,13 @@ void mulcmv(const UDSPInstruction& opc)
u8 treg = (opc.hex >> 11) & 0x1;
u16 accm = dsp_get_acc_m(sreg);
u16 axh = dsp_get_ax_h(treg);
u8 rreg = (opc.hex >> 8) & 0x1;
s64 prod = dsp_multiply(accm, axh);
zeroWriteBackLog();
dsp_multiply(accm, axh);
dsp_set_long_prod(prod);
// update acc
u8 rreg = (opc.hex >> 8) & 0x1;
dsp_set_long_acc(rreg, old_prod);
Update_SR_Register64(old_prod);
@ -265,14 +245,13 @@ void mulcac(const UDSPInstruction& opc)
u8 treg = (opc.hex >> 11) & 0x1;
u16 accm = dsp_get_acc_m(sreg);
u16 axh = dsp_get_ax_h(treg);
zeroWriteBackLog();
dsp_multiply(accm, axh);
// update acc
s64 prod = dsp_multiply(accm, axh);
u8 rreg = (opc.hex >> 8) & 0x1;
s64 acc = old_prod + dsp_get_long_acc(rreg);
zeroWriteBackLog();
dsp_set_long_prod(prod);
// update acc
dsp_set_long_acc(rreg, acc);
Update_SR_Register64(acc);
@ -288,10 +267,11 @@ void mul(const UDSPInstruction& opc)
u8 sreg = (opc.hex >> 11) & 0x1;
u16 axl = dsp_get_ax_l(sreg);
u16 axh = dsp_get_ax_h(sreg);
zeroWriteBackLog();
s64 prod = dsp_multiply(axh, axl);
zeroWriteBackLog();
dsp_set_long_prod(prod);
// FIXME: no update in duddie's docs
Update_SR_Register64(prod);
}
@ -310,14 +290,12 @@ void mulac(const UDSPInstruction& opc)
s64 acR = dsp_get_long_acc(rreg) + dsp_get_long_prod();
u16 axl = dsp_get_ax_l(sreg);
u16 axh = dsp_get_ax_h(sreg);
s64 prod = dsp_multiply(axl, axh);
zeroWriteBackLog();
dsp_set_long_acc(rreg, acR);
// calculate new prod
s64 prod = dsp_multiply(axl, axh);
dsp_set_long_prod(prod);
// FIXME: no update in duddie's docs
Update_SR_Register64(prod);
@ -332,15 +310,14 @@ void mulmv(const UDSPInstruction& opc)
{
u8 rreg = (opc.hex >> 8) & 0x1;
u8 sreg = ((opc.hex >> 11) & 0x1);
s64 prod = dsp_get_long_prod();
s64 acc = prod;
s64 acc = dsp_get_long_prod();
u16 axl = dsp_get_ax_l(sreg);
u16 axh = dsp_get_ax_h(sreg);
s64 prod = dsp_multiply(axl, axh);
zeroWriteBackLog();
dsp_set_long_acc(rreg, acc);
prod = dsp_multiply(axl, axh);
dsp_set_long_prod(prod);
Update_SR_Register64(prod);
}
@ -356,17 +333,15 @@ void mulmvz(const UDSPInstruction& opc)
u8 rreg = (opc.hex >> 8) & 0x1;
// overwrite acc and clear low part
s64 prod = dsp_get_long_prod();
s64 acc = prod & ~0xffff;
s64 acc = dsp_get_long_prod() & ~0xffff;
u16 axl = dsp_get_ax_l(sreg);
u16 axh = dsp_get_ax_h(sreg);
s64 prod = dsp_multiply(axl, axh);
zeroWriteBackLog();
dsp_set_long_acc(rreg, acc);
// math prod
prod = dsp_multiply(axl, axh);
dsp_set_long_prod(prod);
Update_SR_Register64(prod);
}
@ -381,9 +356,10 @@ void mulx(const UDSPInstruction& opc)
u16 val1 = (sreg == 0) ? dsp_get_ax_l(0) : dsp_get_ax_h(0);
u16 val2 = (treg == 0) ? dsp_get_ax_l(1) : dsp_get_ax_h(1);
s64 prod = dsp_multiply(val1, val2, true);
zeroWriteBackLog();
s64 prod = dsp_multiply_conditional_unsigned(val1, val2);
dsp_set_long_prod(prod);
Update_SR_Register64(prod);
}
@ -397,7 +373,6 @@ void mulxac(const UDSPInstruction& opc)
// add old prod to acc
u8 rreg = (opc.hex >> 8) & 0x1;
s64 acR = dsp_get_long_acc(rreg) + dsp_get_long_prod();
dsp_set_long_acc(rreg, acR);
// math new prod
u8 sreg = (opc.hex >> 12) & 0x1;
@ -405,9 +380,12 @@ void mulxac(const UDSPInstruction& opc)
u16 val1 = (sreg == 0) ? dsp_get_ax_l(0) : dsp_get_ax_h(0);
u16 val2 = (treg == 0) ? dsp_get_ax_l(1) : dsp_get_ax_h(1);
zeroWriteBackLog();
s64 prod = dsp_multiply(val1, val2, true);
s64 prod = dsp_multiply_conditional_unsigned(val1, val2);
zeroWriteBackLog();
dsp_set_long_acc(rreg, acR);
dsp_set_long_prod(prod);
Update_SR_Register64(prod);
}
@ -421,7 +399,6 @@ void mulxmv(const UDSPInstruction& opc)
// add old prod to acc
u8 rreg = ((opc.hex >> 8) & 0x1);
s64 acR = dsp_get_long_prod();
dsp_set_long_acc(rreg, acR);
// math new prod
u8 sreg = (opc.hex >> 12) & 0x1;
@ -429,8 +406,11 @@ void mulxmv(const UDSPInstruction& opc)
s16 val1 = (sreg == 0) ? dsp_get_ax_l(0) : dsp_get_ax_h(0);
s16 val2 = (treg == 0) ? dsp_get_ax_l(1) : dsp_get_ax_h(1);
s64 prod = dsp_multiply(val1, val2, true);
zeroWriteBackLog();
s64 prod = dsp_multiply_conditional_unsigned(val1, val2);
dsp_set_long_acc(rreg, acR);
dsp_set_long_prod(prod);
Update_SR_Register64(prod);
}
@ -444,9 +424,7 @@ void mulxmvz(const UDSPInstruction& opc)
{
// overwrite acc and clear low part
u8 rreg = (opc.hex >> 8) & 0x1;
s64 prod = dsp_get_long_prod();
s64 acc = prod & ~0xffff;
dsp_set_long_acc(rreg, acc);
s64 acc = dsp_get_long_prod() & ~0xffff;
// math prod
u8 sreg = (opc.hex >> 12) & 0x1;
@ -454,8 +432,12 @@ void mulxmvz(const UDSPInstruction& opc)
u16 val1 = (sreg == 0) ? dsp_get_ax_l(0) : dsp_get_ax_h(0);
u16 val2 = (treg == 0) ? dsp_get_ax_l(1) : dsp_get_ax_h(1);
s64 prod = dsp_multiply(val1, val2, true);
zeroWriteBackLog();
prod = dsp_multiply_conditional_unsigned(val1, val2);
dsp_set_long_acc(rreg, acc);
dsp_set_long_prod(prod);
Update_SR_Register64(prod);
}
@ -471,8 +453,10 @@ void maddx(const UDSPInstruction& opc)
u16 val1 = (sreg == 0) ? dsp_get_ax_l(0) : dsp_get_ax_h(0);
u16 val2 = (treg == 0) ? dsp_get_ax_l(1) : dsp_get_ax_h(1);
s64 prod = dsp_multiply_add(val1, val2, true);
zeroWriteBackLog();
s64 prod = dsp_multiply_add(val1, val2);
dsp_set_long_prod(prod);
Update_SR_Register64(prod);
}
@ -488,8 +472,10 @@ void msubx(const UDSPInstruction& opc)
u16 val1 = (sreg == 0) ? dsp_get_ax_l(0) : dsp_get_ax_h(0);
u16 val2 = (treg == 0) ? dsp_get_ax_l(1) : dsp_get_ax_h(1);
s64 prod = dsp_multiply_sub(val1, val2, true);
zeroWriteBackLog();
s64 prod = dsp_multiply_sub(val1, val2);
dsp_set_long_prod(prod);
Update_SR_Register64(prod);
}
@ -504,8 +490,11 @@ void maddc(const UDSPInstruction& opc)
u32 treg = (opc.hex >> 8) & 0x1;
u16 accm = dsp_get_acc_m(sreg);
u16 axh = dsp_get_ax_h(treg);
zeroWriteBackLog();
s64 prod = dsp_multiply_add(accm, axh);
zeroWriteBackLog();
dsp_set_long_prod(prod);
Update_SR_Register64(prod);
}
@ -520,8 +509,10 @@ void msubc(const UDSPInstruction& opc)
u32 treg = (opc.hex >> 8) & 0x1;
u16 accm = dsp_get_acc_m(sreg);
u16 axh = dsp_get_ax_h(treg);
zeroWriteBackLog();
s64 prod = dsp_multiply_sub(accm, axh);
zeroWriteBackLog();
dsp_set_long_prod(prod);
Update_SR_Register64(prod);
}
@ -535,10 +526,10 @@ void madd(const UDSPInstruction& opc)
u8 sreg = (opc.hex >> 8) & 0x1;
u16 axl = dsp_get_ax_l(sreg);
u16 axh = dsp_get_ax_h(sreg);
s64 prod = dsp_multiply_add(axl, axh);
zeroWriteBackLog();
s64 prod = dsp_multiply_add(axl, axh);
dsp_set_long_prod(prod);
Update_SR_Register64(prod);
}
@ -552,10 +543,10 @@ void msub(const UDSPInstruction& opc)
u8 sreg = (opc.hex >> 8) & 0x1;
u16 axl = dsp_get_ax_l(sreg);
u16 axh = dsp_get_ax_h(sreg);
s64 prod = dsp_multiply_sub(axl, axh);
zeroWriteBackLog();
s64 prod = dsp_multiply_sub(axl, axh);
dsp_set_long_prod(prod);
Update_SR_Register64(prod);
}