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

Removed an extraneous FlushLockX, further optimized simultaneous handling of carry/overflow.

This commit is contained in:
calc84maniac 2012-01-02 11:50:01 -05:00
parent 4cb1af0f13
commit a6d041bfa9
4 changed files with 37 additions and 32 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
Source/Core/Core/Src/PowerPC/Jit64/Jit.h
View File

@ -149,7 +149,7 @@ public:
void GenerateConstantOverflow(bool overflow);
void GenerateOverflow();
void GenerateOverflowFinalizeCarry(bool oe, bool inv = false);
void FinalizeCarryOverflow(bool oe, bool inv = false);
void GenerateCarry();
void GenerateRC();
void ComputeRC(const Gen::OpArg & arg);

58
Source/Core/Core/Src/PowerPC/Jit64/Jit_Integer.cpp
View File

@ -50,8 +50,8 @@ void Jit64::GenerateOverflow()
SetJumpTarget(exit);
}
// Assumes CA is clear
void Jit64::GenerateOverflowFinalizeCarry(bool oe, bool inv)
// Assumes CA,OV are clear
void Jit64::FinalizeCarryOverflow(bool oe, bool inv)
{
// USES_XER
if (oe)
@ -69,8 +69,6 @@ void Jit64::GenerateOverflowFinalizeCarry(bool oe, bool inv)
FixupBranch carry2 = J_CC(inv ? CC_C : CC_NC);
JitSetCA();
SetJumpTarget(carry2);
//XER[OV] = 0
AND(32, M(&PowerPC::ppcState.spr[SPR_XER]), Imm32(~XER_OV_MASK));
SetJumpTarget(exit);
}
else
@ -878,7 +876,7 @@ void Jit64::subfcx(UGeckoInstruction inst)
gpr.Lock(a, b, d);
gpr.BindToRegister(d, (d == a || d == b), true);
JitClearCA();
JitClearCAOV(inst.OE);
if (d == b)
{
SUB(32, gpr.R(d), gpr.R(a));
@ -897,7 +895,7 @@ void Jit64::subfcx(UGeckoInstruction inst)
if (inst.Rc) {
GenerateRC();
}
GenerateOverflowFinalizeCarry(inst.OE, true);
FinalizeCarryOverflow(inst.OE, true);
gpr.UnlockAll();
}
@ -907,13 +905,12 @@ void Jit64::subfex(UGeckoInstruction inst)
INSTRUCTION_START;
JITDISABLE(Integer)
int a = inst.RA, b = inst.RB, d = inst.RD;
gpr.FlushLockX(ECX);
gpr.Lock(a, b, d);
gpr.BindToRegister(d, (d == a || d == b), true);
// Get CA and clear it
// Get CA and clear it (along with OV if applicable)
MOV(32, R(EAX), M(&PowerPC::ppcState.spr[SPR_XER]));
JitClearCA();
JitClearCAOV(inst.OE);
SHR(32, R(EAX), Imm8(30));
// Convert carry to borrow
@ -937,10 +934,9 @@ void Jit64::subfex(UGeckoInstruction inst)
if (inst.Rc) {
GenerateRC();
}
GenerateOverflowFinalizeCarry(inst.OE, true);
FinalizeCarryOverflow(inst.OE, true);
gpr.UnlockAll();
gpr.UnlockAllX();
}
void Jit64::subfmex(UGeckoInstruction inst)
@ -953,7 +949,7 @@ void Jit64::subfmex(UGeckoInstruction inst)
gpr.BindToRegister(d, d == a);
MOV(32, R(EAX), M(&PowerPC::ppcState.spr[SPR_XER]));
JitClearCA();
JitClearCAOV(inst.OE);
SHR(32, R(EAX), Imm8(30)); // shift the carry flag out into the x86 carry flag
if (d != a)
{
@ -965,7 +961,7 @@ void Jit64::subfmex(UGeckoInstruction inst)
{
GenerateRC();
}
GenerateOverflowFinalizeCarry(inst.OE);
FinalizeCarryOverflow(inst.OE);
gpr.UnlockAll();
}
@ -979,7 +975,7 @@ void Jit64::subfzex(UGeckoInstruction inst)
gpr.Lock(a, d);
gpr.BindToRegister(d, d == a);
MOV(32, R(EAX), M(&PowerPC::ppcState.spr[SPR_XER]));
JitClearCA();
JitClearCAOV(inst.OE);
SHR(32, R(EAX), Imm8(30)); // shift the carry flag out into the x86 carry flag
if (d != a)
{
@ -991,7 +987,7 @@ void Jit64::subfzex(UGeckoInstruction inst)
{
GenerateRC();
}
GenerateOverflowFinalizeCarry(inst.OE);
FinalizeCarryOverflow(inst.OE);
gpr.UnlockAll();
}
@ -1269,14 +1265,14 @@ void Jit64::addex(UGeckoInstruction inst)
gpr.Lock(a, b, d);
gpr.BindToRegister(d, true);
MOV(32, R(EAX), M(&PowerPC::ppcState.spr[SPR_XER]));
JitClearCA();
JitClearCAOV(inst.OE);
SHR(32, R(EAX), Imm8(30)); // shift the carry flag out into the x86 carry flag
ADC(32, gpr.R(d), gpr.R((d == a) ? b : a));
if (inst.Rc)
{
GenerateRC();
}
GenerateOverflowFinalizeCarry(inst.OE);
FinalizeCarryOverflow(inst.OE);
gpr.UnlockAll();
}
else
@ -1284,7 +1280,7 @@ void Jit64::addex(UGeckoInstruction inst)
gpr.Lock(a, b, d);
gpr.BindToRegister(d, false);
MOV(32, R(EAX), M(&PowerPC::ppcState.spr[SPR_XER]));
JitClearCA();
JitClearCAOV(inst.OE);
SHR(32, R(EAX), Imm8(30)); // shift the carry flag out into the x86 carry flag
MOV(32, gpr.R(d), gpr.R(a));
ADC(32, gpr.R(d), gpr.R(b));
@ -1292,7 +1288,7 @@ void Jit64::addex(UGeckoInstruction inst)
{
GenerateRC();
}
GenerateOverflowFinalizeCarry(inst.OE);
FinalizeCarryOverflow(inst.OE);
gpr.UnlockAll();
}
}
@ -1308,27 +1304,27 @@ void Jit64::addcx(UGeckoInstruction inst)
int operand = ((d == a) ? b : a);
gpr.Lock(a, b, d);
gpr.BindToRegister(d, true);
JitClearCA();
JitClearCAOV(inst.OE);
ADD(32, gpr.R(d), gpr.R(operand));
if (inst.Rc)
{
GenerateRC();
}
GenerateOverflowFinalizeCarry(inst.OE);
FinalizeCarryOverflow(inst.OE);
gpr.UnlockAll();
}
else
{
gpr.Lock(a, b, d);
gpr.BindToRegister(d, false);
JitClearCA();
JitClearCAOV(inst.OE);
MOV(32, gpr.R(d), gpr.R(a));
ADD(32, gpr.R(d), gpr.R(b));
if (inst.Rc)
{
GenerateRC();
}
GenerateOverflowFinalizeCarry(inst.OE);
FinalizeCarryOverflow(inst.OE);
gpr.UnlockAll();
}
}
@ -1345,14 +1341,14 @@ void Jit64::addmex(UGeckoInstruction inst)
gpr.Lock(d);
gpr.BindToRegister(d, true);
MOV(32, R(EAX), M(&PowerPC::ppcState.spr[SPR_XER]));
JitClearCA();
JitClearCAOV(inst.OE);
SHR(32, R(EAX), Imm8(30)); // shift the carry flag out into the x86 carry flag
ADC(32, gpr.R(d), Imm32(0xFFFFFFFF));
if (inst.Rc)
{
GenerateRC();
}
GenerateOverflowFinalizeCarry(inst.OE);
FinalizeCarryOverflow(inst.OE);
gpr.UnlockAll();
}
else
@ -1360,7 +1356,7 @@ void Jit64::addmex(UGeckoInstruction inst)
gpr.Lock(a, d);
gpr.BindToRegister(d, false);
MOV(32, R(EAX), M(&PowerPC::ppcState.spr[SPR_XER]));
JitClearCA();
JitClearCAOV(inst.OE);
SHR(32, R(EAX), Imm8(30)); // shift the carry flag out into the x86 carry flag
MOV(32, gpr.R(d), gpr.R(a));
ADC(32, gpr.R(d), Imm32(0xFFFFFFFF));
@ -1368,7 +1364,7 @@ void Jit64::addmex(UGeckoInstruction inst)
{
GenerateRC();
}
GenerateOverflowFinalizeCarry(inst.OE);
FinalizeCarryOverflow(inst.OE);
gpr.UnlockAll();
}
}
@ -1385,14 +1381,14 @@ void Jit64::addzex(UGeckoInstruction inst)
gpr.Lock(d);
gpr.BindToRegister(d, true);
MOV(32, R(EAX), M(&PowerPC::ppcState.spr[SPR_XER]));
JitClearCA();
JitClearCAOV(inst.OE);
SHR(32, R(EAX), Imm8(30)); // shift the carry flag out into the x86 carry flag
ADC(32, gpr.R(d), Imm8(0));
if (inst.Rc)
{
GenerateRC();
}
GenerateOverflowFinalizeCarry(inst.OE);
FinalizeCarryOverflow(inst.OE);
gpr.UnlockAll();
}
else
@ -1400,7 +1396,7 @@ void Jit64::addzex(UGeckoInstruction inst)
gpr.Lock(a, d);
gpr.BindToRegister(d, false);
MOV(32, R(EAX), M(&PowerPC::ppcState.spr[SPR_XER]));
JitClearCA();
JitClearCAOV(inst.OE);
SHR(32, R(EAX), Imm8(30)); // shift the carry flag out into the x86 carry flag
MOV(32, gpr.R(d), gpr.R(a));
ADC(32, gpr.R(d), Imm8(0));
@ -1408,7 +1404,7 @@ void Jit64::addzex(UGeckoInstruction inst)
{
GenerateRC();
}
GenerateOverflowFinalizeCarry(inst.OE);
FinalizeCarryOverflow(inst.OE);
gpr.UnlockAll();
}
}

8
Source/Core/Core/Src/PowerPC/JitCommon/Jit_Util.cpp
View File

@ -352,3 +352,11 @@ void EmuCodeBlock::JitSetCA()
{
OR(32, M(&PowerPC::ppcState.spr[SPR_XER]), Imm32(XER_CA_MASK)); //XER.CA = 1
}
void EmuCodeBlock::JitClearCAOV(bool oe)
{
if (oe)
AND(32, M(&PowerPC::ppcState.spr[SPR_XER]), Imm32(~XER_CA_MASK & ~XER_OV_MASK)); //XER.CA, XER.OV = 0
else
AND(32, M(&PowerPC::ppcState.spr[SPR_XER]), Imm32(~XER_CA_MASK)); //XER.CA = 0
}

1
Source/Core/Core/Src/PowerPC/JitCommon/Jit_Util.h
View File

@ -38,6 +38,7 @@ public:
void WriteFloatToConstRamAddress(const Gen::X64Reg& xmm_reg, u32 address);
void JitClearCA();
void JitSetCA();
void JitClearCAOV(bool oe);
void ForceSinglePrecisionS(Gen::X64Reg xmm);
void ForceSinglePrecisionP(Gen::X64Reg xmm);