diff --git a/src/AssemblyGeneratorX86.cpp b/src/AssemblyGeneratorX86.cpp
index 4a35dfb..f1c3de8 100644
--- a/src/AssemblyGeneratorX86.cpp
+++ b/src/AssemblyGeneratorX86.cpp
@@ -30,12 +30,20 @@ namespace RandomX {
static const char* regR[8] = { "r8", "r9", "r10", "r11", "r12", "r13", "r14", "r15" };
static const char* regR32[8] = { "r8d", "r9d", "r10d", "r11d", "r12d", "r13d", "r14d", "r15d" };
- static const char* regF[8] = { "xmm8", "xmm9", "xmm2", "xmm3", "xmm4", "xmm5", "xmm6", "xmm7" };
+ static const char* regFE[8] = { "xmm0", "xmm1", "xmm2", "xmm3", "xmm4", "xmm5", "xmm6", "xmm7" };
+ static const char* regF[4] = { "xmm0", "xmm1", "xmm2", "xmm3" };
+ static const char* regE[4] = { "xmm4", "xmm5", "xmm6", "xmm7" };
+ static const char* regA[4] = { "xmm8", "xmm9", "xmm10", "xmm11" };
+ static const char* regA4 = "xmm12";
+ static const char* dblMin = "xmm13";
+ static const char* absMask = "xmm14";
+ static const char* signMask = "xmm15";
static const char* regMx = "rbp";
- static const char* regIc = "ebx";
+ static const char* regIc = "rbx";
+ static const char* regIc32 = "ebx";
static const char* regIc8 = "bl";
- static const char* regStackBeginAddr = "rdi";
+ static const char* regDatasetAddr = "rdi";
static const char* regScratchpadAddr = "rsi";
void AssemblyGeneratorX86::generateProgram(const void* seed) {
@@ -49,226 +57,217 @@ namespace RandomX {
for (unsigned j = 0; j < sizeof(instr) / sizeof(Pcg32::result_type); ++j) {
*(((uint32_t*)&instr) + j) = gen();
}
+ instr.src %= RegistersCount;
+ instr.dst %= RegistersCount;
generateCode(instr, i);
- asmCode << std::endl;
+ //asmCode << std::endl;
}
- if(ProgramLength > 0)
- asmCode << "\tjmp rx_i_0" << std::endl;
}
void AssemblyGeneratorX86::generateCode(Instruction& instr, int i) {
- asmCode << "rx_i_" << i << ": ;" << instr.getName() << std::endl;
- asmCode << "\tdec " << regIc << std::endl;
- asmCode << "\tjz rx_finish" << std::endl;
+ asmCode << "\t; " << instr;
auto generator = engine[instr.opcode];
(this->*generator)(instr, i);
}
- void AssemblyGeneratorX86::gena(Instruction& instr, int i) {
- asmCode << "\txor " << regR[instr.rega % RegistersCount] << ", 0" << std::hex << instr.addra << "h" << std::dec << std::endl;
- asmCode << "\tmov eax, " << regR32[instr.rega % RegistersCount] << std::endl;
- asmCode << "\ttest " << regIc8 << ", 63" << std::endl;
- asmCode << "\tjnz short rx_body_" << i << std::endl;
- asmCode << "\tcall rx_read" << std::endl;
- asmCode << "rx_body_" << i << ":" << std::endl;
- if ((instr.loca & 192) == 0)
- asmCode << "\txor " << regMx << ", rax" << std::endl;
- if (instr.loca & 15) {
- if (instr.loca & 3) {
- asmCode << "\tand eax, " << (ScratchpadL1 - 1) << std::endl;
- }
- else {
- asmCode << "\tand eax, " << (ScratchpadL2 - 1) << std::endl;
- }
+ void AssemblyGeneratorX86::genAddressReg(Instruction& instr, const char* reg = "eax") {
+ asmCode << "\tmov " << reg << ", " << regR32[instr.src] << std::endl;
+ asmCode << "\tand " << reg << ", " << ((instr.alt % 4) ? ScratchpadL1Mask : ScratchpadL2Mask) << std::endl;
+ }
+
+ int32_t AssemblyGeneratorX86::genAddressImm(Instruction& instr) {
+ return instr.imm32 & ((instr.alt % 4) ? ScratchpadL1Mask : ScratchpadL2Mask);
+ }
+
+ //1 uOP
+ void AssemblyGeneratorX86::h_IADD_R(Instruction& instr, int i) {
+ if (instr.src != instr.dst) {
+ asmCode << "\tadd " << regR[instr.dst] << ", " << regR[instr.src] << std::endl;
}
else {
- asmCode << "\tand eax, " << (ScratchpadL3 - 1) << std::endl;
+ asmCode << "\tadd " << regR[instr.dst] << ", " << instr.imm32 << std::endl;
}
}
- void AssemblyGeneratorX86::genar(Instruction& instr, int i) {
- gena(instr, i);
- asmCode << "\tmov rax, qword ptr [" << regScratchpadAddr << "+rax*8]" << std::endl;
- }
-
-
- void AssemblyGeneratorX86::genaf(Instruction& instr, int i) {
- gena(instr, i);
- asmCode << "\tcvtdq2pd xmm0, qword ptr [" << regScratchpadAddr << "+rax*8]" << std::endl;
- }
-
- void AssemblyGeneratorX86::genbiashift(Instruction& instr, const char* instrx86) {
- if (instr.locb & 1) {
- asmCode << "\tmov rcx, " << regR[instr.regb % RegistersCount] << std::endl;
- asmCode << "\t" << instrx86 << " rax, cl" << std::endl;
- } else {
- asmCode << "\t" << instrx86 << " rax, " << (instr.imm8 & 63) << std::endl;;
- }
- }
-
- void AssemblyGeneratorX86::genbia(Instruction& instr) {
- if (instr.locb & 3) {
- asmCode << regR[instr.regb % RegistersCount] << std::endl;
- } else {
- asmCode << instr.imm32 << std::endl;;
- }
- }
-
- void AssemblyGeneratorX86::genbia32(Instruction& instr) {
- if (instr.locb & 3) {
- asmCode << regR32[instr.regb % RegistersCount] << std::endl;
+ //2.75 uOP
+ void AssemblyGeneratorX86::h_IADD_M(Instruction& instr, int i) {
+ if (instr.src != instr.dst) {
+ genAddressReg(instr);
+ asmCode << "\tadd " << regR[instr.dst] << ", qword ptr [rsi+rax]" << std::endl;
}
else {
- asmCode << instr.imm32 << std::endl;;
+ asmCode << "\tadd " << regR[instr.dst] << ", qword ptr [rsi+" << genAddressImm(instr) << "]" << std::endl;
}
}
- void AssemblyGeneratorX86::genbf(Instruction& instr, const char* instrx86) {
- asmCode << "\t" << instrx86 << " xmm0, " << regF[instr.regb % RegistersCount] << std::endl;
+ //1 uOP
+ void AssemblyGeneratorX86::h_IADD_RC(Instruction& instr, int i) {
+ asmCode << "\tlea " << regR[instr.dst] << ", [" << regR[instr.dst] << "+" << regR[instr.src] << std::showpos << instr.imm32 << std::noshowpos << "]" << std::endl;
}
- void AssemblyGeneratorX86::gencr(Instruction& instr, bool rax = true) {
- if (instr.locc & 16) { //write to register
- asmCode << "\tmov " << regR[instr.regc % RegistersCount] << ", " << (rax ? "rax" : "rcx") << std::endl;
- if (trace) {
- asmCode << "\tmov qword ptr [" << regScratchpadAddr << " + " << regIc << " * 8 + 262136], " << (rax ? "rax" : "rcx") << std::endl;
- }
+ //1 uOP
+ void AssemblyGeneratorX86::h_ISUB_R(Instruction& instr, int i) {
+ if (instr.src != instr.dst) {
+ asmCode << "\tsub " << regR[instr.dst] << ", " << regR[instr.src] << std::endl;
}
- else { //write to scratchpad
- if (rax)
- asmCode << "\tmov rcx, rax" << std::endl;
- asmCode << "\tmov eax, " << regR32[instr.regc % RegistersCount] << std::endl;
- asmCode << "\txor eax, 0" << std::hex << instr.addrc << "h" << std::dec << std::endl;
- if (instr.locc & 15) {
- if (instr.locc & 3) {
- asmCode << "\tand eax, " << (ScratchpadL1 - 1) << std::endl;
- }
- else {
- asmCode << "\tand eax, " << (ScratchpadL2 - 1) << std::endl;
- }
- }
- else {
- asmCode << "\tand eax, " << (ScratchpadL3 - 1) << std::endl;
- }
- asmCode << "\tmov qword ptr [" << regScratchpadAddr << " + rax * 8], rcx" << std::endl;
- if (trace) {
- asmCode << "\tmov qword ptr [" << regScratchpadAddr << " + " << regIc << " * 8 + 262136], rcx" << std::endl;
- }
+ else {
+ asmCode << "\tsub " << regR[instr.dst] << ", " << instr.imm32 << std::endl;
}
}
- void AssemblyGeneratorX86::gencf(Instruction& instr, bool move = true) {
- if(move)
- asmCode << "\tmovaps " << regF[instr.regc % RegistersCount] << ", xmm0" << std::endl;
- const char* store = (instr.locc & 128) ? "movhpd" : "movlpd";
- if (instr.locc & 16) { //write to scratchpad
- asmCode << "\tmov eax, " << regR32[instr.regc % RegistersCount] << std::endl;
- asmCode << "\txor eax, 0" << std::hex << instr.addrc << "h" << std::dec << std::endl;
- if (instr.locc & 15) {
- if (instr.locc & 3) {
- asmCode << "\tand eax, " << (ScratchpadL1 - 1) << std::endl;
- }
- else {
- asmCode << "\tand eax, " << (ScratchpadL2 - 1) << std::endl;
- }
- }
- else {
- asmCode << "\tand eax, " << (ScratchpadL3 - 1) << std::endl;
- }
- asmCode << "\t" << store << " qword ptr [" << regScratchpadAddr << " + rax * 8], " << regF[instr.regc % RegistersCount] << std::endl;
+ //2.75 uOP
+ void AssemblyGeneratorX86::h_ISUB_M(Instruction& instr, int i) {
+ if (instr.src != instr.dst) {
+ genAddressReg(instr);
+ asmCode << "\tsub " << regR[instr.dst] << ", qword ptr [rsi+rax]" << std::endl;
}
- if (trace) {
- asmCode << "\t" << store << " qword ptr [" << regScratchpadAddr << " + " << regIc << " * 8 + 262136], " << regF[instr.regc % RegistersCount] << std::endl;
+ else {
+ asmCode << "\tsub " << regR[instr.dst] << ", qword ptr [rsi+" << genAddressImm(instr) << "]" << std::endl;
}
}
- void AssemblyGeneratorX86::h_ADD_64(Instruction& instr, int i) {
- genar(instr, i);
- asmCode << "\tadd rax, ";
- genbia(instr);
- gencr(instr);
+ //1 uOP
+ void AssemblyGeneratorX86::h_IMUL_9C(Instruction& instr, int i) {
+ asmCode << "\tlea " << regR[instr.dst] << ", [" << regR[instr.dst] << "+" << regR[instr.dst] << "*8" << std::showpos << instr.imm32 << std::noshowpos << "]" << std::endl;
}
- void AssemblyGeneratorX86::h_ADD_32(Instruction& instr, int i) {
- genar(instr, i);
- asmCode << "\tadd eax, ";
- genbia32(instr);
- gencr(instr);
- }
-
- void AssemblyGeneratorX86::h_SUB_64(Instruction& instr, int i) {
- genar(instr, i);
- asmCode << "\tsub rax, ";
- genbia(instr);
- gencr(instr);
- }
-
- void AssemblyGeneratorX86::h_SUB_32(Instruction& instr, int i) {
- genar(instr, i);
- asmCode << "\tsub eax, ";
- genbia32(instr);
- gencr(instr);
- }
-
- void AssemblyGeneratorX86::h_MUL_64(Instruction& instr, int i) {
- genar(instr, i);
- asmCode << "\timul rax, ";
- if ((instr.locb & 3) == 0) {
- asmCode << "rax, ";
+ //1 uOP
+ void AssemblyGeneratorX86::h_IMUL_R(Instruction& instr, int i) {
+ if (instr.src != instr.dst) {
+ asmCode << "\timul " << regR[instr.dst] << ", " << regR[instr.src] << std::endl;
+ }
+ else {
+ asmCode << "\timul " << regR[instr.dst] << ", " << instr.imm32 << std::endl;
}
- genbia(instr);
- gencr(instr);
}
- void AssemblyGeneratorX86::h_MULH_64(Instruction& instr, int i) {
- genar(instr, i);
- asmCode << "\tmov rcx, ";
- genbia(instr);
- asmCode << "\tmul rcx" << std::endl;
- asmCode << "\tmov rax, rdx" << std::endl;
- gencr(instr);
+ //2.75 uOP
+ void AssemblyGeneratorX86::h_IMUL_M(Instruction& instr, int i) {
+ if (instr.src != instr.dst) {
+ genAddressReg(instr);
+ asmCode << "\timul " << regR[instr.dst] << ", qword ptr [rsi+rax]" << std::endl;
+ }
+ else {
+ asmCode << "\timul " << regR[instr.dst] << ", qword ptr [rsi+" << genAddressImm(instr) << "]" << std::endl;
+ }
}
- void AssemblyGeneratorX86::h_MUL_32(Instruction& instr, int i) {
- genar(instr, i);
- asmCode << "\tmov ecx, eax" << std::endl;
- asmCode << "\tmov eax, ";
- genbia32(instr);
- asmCode << "\timul rax, rcx" << std::endl;
- gencr(instr);
+ //4 uOPs
+ void AssemblyGeneratorX86::h_IMULH_R(Instruction& instr, int i) {
+ if (instr.src != instr.dst) {
+ asmCode << "\tmov rax, " << regR[instr.dst] << std::endl;
+ asmCode << "\tmul " << regR[instr.src] << std::endl;
+ asmCode << "\tmov " << regR[instr.dst] << ", rdx" << std::endl;
+ }
+ else {
+ asmCode << "\tmov eax, " << instr.imm32 << std::endl;
+ asmCode << "\tmul " << regR[instr.dst] << std::endl;
+ asmCode << "\tadd " << regR[instr.dst] << ", rdx" << std::endl;
+ }
}
- void AssemblyGeneratorX86::h_IMUL_32(Instruction& instr, int i) {
- genar(instr, i);
- asmCode << "\tmovsxd rcx, eax" << std::endl;
- if ((instr.locb & 3) == 0) {
+ //5.75 uOPs
+ void AssemblyGeneratorX86::h_IMULH_M(Instruction& instr, int i) {
+ if (instr.src != instr.dst) {
+ genAddressReg(instr, "ecx");
+ asmCode << "\tmov rax, " << regR[instr.dst] << std::endl;
+ asmCode << "\tmul qword ptr [rsi+rcx]" << std::endl;
+ }
+ else {
+ asmCode << "\tmov rax, " << regR[instr.dst] << std::endl;
+ asmCode << "\tmul qword ptr [rsi+" << genAddressImm(instr) << "]" << std::endl;
+ }
+ asmCode << "\tmov " << regR[instr.dst] << ", rdx" << std::endl;
+ }
+
+ //4 uOPs
+ void AssemblyGeneratorX86::h_ISMULH_R(Instruction& instr, int i) {
+ if (instr.src != instr.dst) {
+ asmCode << "\tmov rax, " << regR[instr.dst] << std::endl;
+ asmCode << "\timul " << regR[instr.src] << std::endl;
+ asmCode << "\tmov " << regR[instr.dst] << ", rdx" << std::endl;
+ }
+ else {
asmCode << "\tmov rax, " << instr.imm32 << std::endl;
+ asmCode << "\timul " << regR[instr.dst] << std::endl;
+ asmCode << "\tadd " << regR[instr.dst] << ", rdx" << std::endl;
+ }
+ }
+
+ //5.75 uOPs
+ void AssemblyGeneratorX86::h_ISMULH_M(Instruction& instr, int i) {
+ if (instr.src != instr.dst) {
+ genAddressReg(instr, "ecx");
+ asmCode << "\tmov rax, " << regR[instr.dst] << std::endl;
+ asmCode << "\timul qword ptr [rsi+rcx]" << std::endl;
}
else {
- asmCode << "\tmovsxd rax, " << regR32[instr.regb % RegistersCount] << std::endl;
+ asmCode << "\tmov rax, " << regR[instr.dst] << std::endl;
+ asmCode << "\timul qword ptr [rsi+" << genAddressImm(instr) << "]" << std::endl;
}
- asmCode << "\timul rax, rcx" << std::endl;
- gencr(instr);
+ asmCode << "\tmov " << regR[instr.dst] << ", rdx" << std::endl;
}
- void AssemblyGeneratorX86::h_IMULH_64(Instruction& instr, int i) {
- genar(instr, i);
- asmCode << "\tmov rcx, ";
- genbia(instr);
- asmCode << "\timul rcx" << std::endl;
- asmCode << "\tmov rax, rdx" << std::endl;
- gencr(instr);
+ //1 uOP
+ void AssemblyGeneratorX86::h_INEG_R(Instruction& instr, int i) {
+ asmCode << "\tneg " << regR[instr.dst] << std::endl;
}
- void AssemblyGeneratorX86::h_DIV_64(Instruction& instr, int i) {
- genar(instr, i);
- if (instr.locb & 3) {
-#ifdef MAGIC_DIVISION
- if (instr.imm32 != 0) {
- uint32_t divisor = instr.imm32;
- asmCode << "\t; magic divide by " << divisor << std::endl;
- if (divisor & (divisor - 1)) {
- magicu_info mi = compute_unsigned_magic_info(divisor, sizeof(uint64_t) * 8);
+ //1 uOP
+ void AssemblyGeneratorX86::h_IXOR_R(Instruction& instr, int i) {
+ if (instr.src != instr.dst) {
+ asmCode << "\txor " << regR[instr.dst] << ", " << regR[instr.src] << std::endl;
+ }
+ else {
+ asmCode << "\txor " << regR[instr.dst] << ", " << instr.imm32 << std::endl;
+ }
+ }
+
+ //2.75 uOP
+ void AssemblyGeneratorX86::h_IXOR_M(Instruction& instr, int i) {
+ if (instr.src != instr.dst) {
+ genAddressReg(instr);
+ asmCode << "\txor " << regR[instr.dst] << ", qword ptr [rsi+rax]" << std::endl;
+ }
+ else {
+ asmCode << "\txor " << regR[instr.dst] << ", qword ptr [rsi+" << genAddressImm(instr) << "]" << std::endl;
+ }
+ }
+
+ //1.75 uOPs
+ void AssemblyGeneratorX86::h_IROR_R(Instruction& instr, int i) {
+ if (instr.src != instr.dst) {
+ asmCode << "\tmov ecx, " << regR32[instr.src] << std::endl;
+ asmCode << "\tror " << regR[instr.dst] << ", cl" << std::endl;
+ }
+ else {
+ asmCode << "\tror " << regR[instr.dst] << ", " << (instr.imm32 & 63) << std::endl;
+ }
+ }
+
+ //1.75 uOPs
+ void AssemblyGeneratorX86::h_IROL_R(Instruction& instr, int i) {
+ if (instr.src != instr.dst) {
+ asmCode << "\tmov ecx, " << regR32[instr.src] << std::endl;
+ asmCode << "\trol " << regR[instr.dst] << ", cl" << std::endl;
+ }
+ else {
+ asmCode << "\trol " << regR[instr.dst] << ", " << (instr.imm32 & 63) << std::endl;
+ }
+ }
+
+ //~6 uOPs
+ void AssemblyGeneratorX86::h_IDIV_C(Instruction& instr, int i) {
+ if (instr.imm32 != 0) {
+ uint32_t divisor = instr.imm32;
+ if (divisor & (divisor - 1)) {
+ magicu_info mi = compute_unsigned_magic_info(divisor, sizeof(uint64_t) * 8);
+ if (mi.pre_shift == 0 && !mi.increment) {
+ asmCode << "\tmov rax, " << mi.multiplier << std::endl;
+ asmCode << "\tmul " << regR[instr.dst] << std::endl;
+ }
+ else {
+ asmCode << "\tmov rax, " << regR[instr.dst] << std::endl;
if (mi.pre_shift > 0)
asmCode << "\tshr rax, " << mi.pre_shift << std::endl;
if (mi.increment) {
@@ -277,326 +276,249 @@ namespace RandomX {
}
asmCode << "\tmov rcx, " << mi.multiplier << std::endl;
asmCode << "\tmul rcx" << std::endl;
- asmCode << "\tmov rax, rdx" << std::endl;
- if (mi.post_shift > 0)
- asmCode << "\tshr rax, " << mi.post_shift << std::endl;
- }
- else { //divisor is a power of two
- int shift = 0;
- while (divisor >>= 1)
- ++shift;
- if(shift > 0)
- asmCode << "\tshr rax, " << shift << std::endl;
}
+ if (mi.post_shift > 0)
+ asmCode << "\tshr rdx, " << mi.post_shift << std::endl;
+ asmCode << "\tadd " << regR[instr.dst] << ", rdx" << std::endl;
}
-#else
- if (instr.imm32 == 0) {
- asmCode << "\tmov ecx, 1" << std::endl;
- }
- else {
- asmCode << "\tmov ecx, " << instr.imm32 << std::endl;
- }
-#endif
- }
- else {
- asmCode << "\tmov ecx, 1" << std::endl;
- asmCode << "\tmov edx, " << regR32[instr.regb % RegistersCount] << std::endl;
- asmCode << "\ttest edx, edx" << std::endl;
- asmCode << "\tcmovne ecx, edx" << std::endl;
-#ifdef MAGIC_DIVISION
- asmCode << "\txor edx, edx" << std::endl;
- asmCode << "\tdiv rcx" << std::endl;
-#endif
- }
-#ifndef MAGIC_DIVISION
- asmCode << "\txor edx, edx" << std::endl;
- asmCode << "\tdiv rcx" << std::endl;
-#endif
- gencr(instr);
- }
-
- void AssemblyGeneratorX86::h_IDIV_64(Instruction& instr, int i) {
- genar(instr, i);
- if (instr.locb & 3) {
-#ifdef MAGIC_DIVISION
- int64_t divisor = instr.imm32;
- asmCode << "\t; magic divide by " << divisor << std::endl;
- if ((divisor & -divisor) == divisor || (divisor & -divisor) == -divisor) {
- // +/- power of two
- bool negative = divisor < 0;
- if (negative)
- divisor = -divisor;
+ else { //divisor is a power of two
int shift = 0;
- uint64_t unsignedDivisor = divisor;
- while (unsignedDivisor >>= 1)
+ while (divisor >>= 1)
++shift;
- if (shift > 0) {
- asmCode << "\tmov rcx, rax" << std::endl;
- asmCode << "\tsar rcx, 63" << std::endl;
- uint32_t mask = (1ULL << shift) + 0xFFFFFFFF;
- asmCode << "\tand ecx, 0" << std::hex << mask << std::dec << "h" << std::endl;
- asmCode << "\tadd rax, rcx" << std::endl;
- asmCode << "\tsar rax, " << shift << std::endl;
- }
- if (negative)
- asmCode << "\tneg rax" << std::endl;
+ if(shift > 0)
+ asmCode << "\tshr " << regR[instr.dst] << ", " << shift << std::endl;
}
- else if (divisor != 0) {
- magics_info mi = compute_signed_magic_info(divisor);
- if ((divisor >= 0) != (mi.multiplier >= 0))
- asmCode << "\tmov rcx, rax" << std::endl;
- asmCode << "\tmov rdx, " << mi.multiplier << std::endl;
- asmCode << "\timul rdx" << std::endl;
- asmCode << "\tmov rax, rdx" << std::endl;
- asmCode << "\txor edx, edx" << std::endl;
- bool haveSF = false;
- if (divisor > 0 && mi.multiplier < 0) {
- asmCode << "\tadd rax, rcx" << std::endl;
- haveSF = true;
- }
- if (divisor < 0 && mi.multiplier > 0) {
- asmCode << "\tsub rax, rcx" << std::endl;
- haveSF = true;
- }
- if (mi.shift > 0) {
- asmCode << "\tsar rax, " << mi.shift << std::endl;
- haveSF = true;
- }
- if (!haveSF)
- asmCode << "\ttest rax, rax" << std::endl;
- asmCode << "\tsets dl" << std::endl;
- asmCode << "\tadd rax, rdx" << std::endl;
+ }
+ }
+
+ //~8.5 uOPs
+ void AssemblyGeneratorX86::h_ISDIV_C(Instruction& instr, int i) {
+ int64_t divisor = instr.imm32;
+ if ((divisor & -divisor) == divisor || (divisor & -divisor) == -divisor) {
+ asmCode << "\tmov rax, " << regR[instr.dst] << std::endl;
+ // +/- power of two
+ bool negative = divisor < 0;
+ if (negative)
+ divisor = -divisor;
+ int shift = 0;
+ uint64_t unsignedDivisor = divisor;
+ while (unsignedDivisor >>= 1)
+ ++shift;
+ if (shift > 0) {
+ asmCode << "\tmov rcx, rax" << std::endl;
+ asmCode << "\tsar rcx, 63" << std::endl;
+ uint32_t mask = (1ULL << shift) + 0xFFFFFFFF;
+ asmCode << "\tand ecx, 0" << std::hex << mask << std::dec << "h" << std::endl;
+ asmCode << "\tadd rax, rcx" << std::endl;
+ asmCode << "\tsar rax, " << shift << std::endl;
}
-#else
- asmCode << "\tmov edx, " << instr.imm32 << std::endl;
-#endif
+ if (negative)
+ asmCode << "\tneg rax" << std::endl;
+ asmCode << "\tadd " << regR[instr.dst] << ", rax" << std::endl;
}
- else {
- asmCode << "\tmov edx, " << regR32[instr.regb % RegistersCount] << std::endl;
-#ifndef MAGIC_DIVISION
+ else if (divisor != 0) {
+ magics_info mi = compute_signed_magic_info(divisor);
+ asmCode << "\tmov rax, " << mi.multiplier << std::endl;
+ asmCode << "\timul " << regR[instr.dst] << std::endl;
+ //asmCode << "\tmov rax, rdx" << std::endl;
+ asmCode << "\txor eax, eax" << std::endl;
+ bool haveSF = false;
+ if (divisor > 0 && mi.multiplier < 0) {
+ asmCode << "\tadd rdx, " << regR[instr.dst] << std::endl;
+ haveSF = true;
+ }
+ if (divisor < 0 && mi.multiplier > 0) {
+ asmCode << "\tsub rdx, " << regR[instr.dst] << std::endl;
+ haveSF = true;
+ }
+ if (mi.shift > 0) {
+ asmCode << "\tsar rdx, " << mi.shift << std::endl;
+ haveSF = true;
+ }
+ if (!haveSF)
+ asmCode << "\ttest rdx, rdx" << std::endl;
+ asmCode << "\tsets al" << std::endl;
+ asmCode << "\tadd rdx, rax" << std::endl;
+ asmCode << "\tadd " << regR[instr.dst] << ", rdx" << std::endl;
}
-#endif
- asmCode << "\tcmp edx, -1" << std::endl;
- asmCode << "\tjne short body_idiv_" << i << std::endl;
- asmCode << "\tneg rax" << std::endl;
- asmCode << "\tjmp short result_idiv_" << i << std::endl;
- asmCode << "body_idiv_" << i << ":" << std::endl;
- asmCode << "\tmov ecx, 1" << std::endl;
- asmCode << "\ttest edx, edx" << std::endl;
- asmCode << "\tcmovne ecx, edx" << std::endl;
- asmCode << "\tmovsxd rcx, ecx" << std::endl;
- asmCode << "\tcqo" << std::endl;
- asmCode << "\tidiv rcx" << std::endl;
- asmCode << "result_idiv_" << i << ":" << std::endl;
-#ifdef MAGIC_DIVISION
- }
-#endif
- gencr(instr);
}
- void AssemblyGeneratorX86::h_AND_64(Instruction& instr, int i) {
- genar(instr, i);
- asmCode << "\tand rax, ";
- genbia(instr);
- gencr(instr);
+ //1 uOPs
+ void AssemblyGeneratorX86::h_FPSWAP_R(Instruction& instr, int i) {
+ asmCode << "\tshufpd " << regFE[instr.dst] << ", " << regFE[instr.dst] << ", 1" << std::endl;
}
- void AssemblyGeneratorX86::h_AND_32(Instruction& instr, int i) {
- genar(instr, i);
- asmCode << "\tand eax, ";
- genbia32(instr);
- gencr(instr);
+ //1 uOP
+ void AssemblyGeneratorX86::h_FPADD_R(Instruction& instr, int i) {
+ instr.dst %= 4;
+ instr.src %= 4;
+ asmCode << "\taddpd " << regF[instr.dst] << ", " << regA[instr.src] << std::endl;
}
- void AssemblyGeneratorX86::h_OR_64(Instruction& instr, int i) {
- genar(instr, i);
- asmCode << "\tor rax, ";
- genbia(instr);
- gencr(instr);
+ //5 uOPs
+ void AssemblyGeneratorX86::h_FPADD_M(Instruction& instr, int i) {
+ instr.dst %= 4;
+ genAddressReg(instr);
+ asmCode << "\tcvtdq2pd xmm12, qword ptr [rsi+rax]" << std::endl;
+ asmCode << "\taddpd " << regF[instr.dst] << ", xmm12" << std::endl;
}
- void AssemblyGeneratorX86::h_OR_32(Instruction& instr, int i) {
- genar(instr, i);
- asmCode << "\tor eax, ";
- genbia32(instr);
- gencr(instr);
+ //1 uOP
+ void AssemblyGeneratorX86::h_FPSUB_R(Instruction& instr, int i) {
+ instr.dst %= 4;
+ instr.src %= 4;
+ asmCode << "\tsubpd " << regF[instr.dst] << ", " << regA[instr.src] << std::endl;
}
- void AssemblyGeneratorX86::h_XOR_64(Instruction& instr, int i) {
- genar(instr, i);
- asmCode << "\txor rax, ";
- genbia(instr);
- gencr(instr);
+ //5 uOPs
+ void AssemblyGeneratorX86::h_FPSUB_M(Instruction& instr, int i) {
+ instr.dst %= 4;
+ genAddressReg(instr);
+ asmCode << "\tcvtdq2pd xmm12, qword ptr [rsi+rax]" << std::endl;
+ asmCode << "\tsubpd " << regF[instr.dst] << ", xmm12" << std::endl;
}
- void AssemblyGeneratorX86::h_XOR_32(Instruction& instr, int i) {
- genar(instr, i);
- asmCode << "\txor eax, ";
- genbia32(instr);
- gencr(instr);
+ //1 uOP
+ void AssemblyGeneratorX86::h_FPNEG_R(Instruction& instr, int i) {
+ instr.dst %= 4;
+ asmCode << "\txorps " << regF[instr.dst] << ", " << signMask << std::endl;
}
- void AssemblyGeneratorX86::h_SHL_64(Instruction& instr, int i) {
- genar(instr, i);
- genbiashift(instr, "shl");
- gencr(instr);
+ //1 uOPs
+ void AssemblyGeneratorX86::h_FPMUL_R(Instruction& instr, int i) {
+ instr.dst %= 4;
+ instr.src %= 4;
+ asmCode << "\tmulpd " << regE[instr.dst] << ", " << regA[instr.src] << std::endl;
}
- void AssemblyGeneratorX86::h_SHR_64(Instruction& instr, int i) {
- genar(instr, i);
- genbiashift(instr, "shr");
- gencr(instr);
+ //6 uOPs
+ void AssemblyGeneratorX86::h_FPMUL_M(Instruction& instr, int i) {
+ instr.dst %= 4;
+ genAddressReg(instr);
+ asmCode << "\tcvtdq2pd xmm12, qword ptr [rsi+rax]" << std::endl;
+ asmCode << "\tmulpd " << regE[instr.dst] << ", xmm12" << std::endl;
+ asmCode << "\tmaxpd " << regE[instr.dst] << ", " << dblMin << std::endl;
}
- void AssemblyGeneratorX86::h_SAR_64(Instruction& instr, int i) {
- genar(instr, i);
- genbiashift(instr, "sar");
- gencr(instr);
+ //2 uOPs
+ void AssemblyGeneratorX86::h_FPDIV_R(Instruction& instr, int i) {
+ instr.dst %= 4;
+ instr.src %= 4;
+ asmCode << "\tdivpd " << regE[instr.dst] << ", " << regA[instr.src] << std::endl;
+ asmCode << "\tmaxpd " << regE[instr.dst] << ", " << dblMin << std::endl;
}
- void AssemblyGeneratorX86::h_ROL_64(Instruction& instr, int i) {
- genar(instr, i);
- genbiashift(instr, "rol");
- gencr(instr);
+ //6 uOPs
+ void AssemblyGeneratorX86::h_FPDIV_M(Instruction& instr, int i) {
+ instr.dst %= 4;
+ genAddressReg(instr);
+ asmCode << "\tcvtdq2pd xmm12, qword ptr [rsi+rax]" << std::endl;
+ asmCode << "\tdivpd " << regE[instr.dst] << ", xmm12" << std::endl;
+ asmCode << "\tmaxpd " << regE[instr.dst] << ", " << dblMin << std::endl;
}
- void AssemblyGeneratorX86::h_ROR_64(Instruction& instr, int i) {
- genar(instr, i);
- genbiashift(instr, "ror");
- gencr(instr);
- }
+ //1 uOP
+ void AssemblyGeneratorX86::h_FPSQRT_R(Instruction& instr, int i) {
+ instr.dst %= 4;
+ asmCode << "\tsqrtpd " << regE[instr.dst] << ", " << regE[instr.dst] << std::endl;
+ }
- void AssemblyGeneratorX86::h_FPADD(Instruction& instr, int i) {
- genaf(instr, i);
- genbf(instr, "addpd");
- gencf(instr);
- }
-
- void AssemblyGeneratorX86::h_FPSUB(Instruction& instr, int i) {
- genaf(instr, i);
- genbf(instr, "subpd");
- gencf(instr);
- }
-
- void AssemblyGeneratorX86::h_FPMUL(Instruction& instr, int i) {
- genaf(instr, i);
- genbf(instr, "mulpd");
- asmCode << "\tmovaps xmm1, xmm0" << std::endl;
- asmCode << "\tcmpeqpd xmm1, xmm1" << std::endl;
- asmCode << "\tandps xmm0, xmm1" << std::endl;
- gencf(instr);
- }
-
- void AssemblyGeneratorX86::h_FPDIV(Instruction& instr, int i) {
- genaf(instr, i);
- genbf(instr, "divpd");
- asmCode << "\tmovaps xmm1, xmm0" << std::endl;
- asmCode << "\tcmpeqpd xmm1, xmm1" << std::endl;
- asmCode << "\tandps xmm0, xmm1" << std::endl;
- gencf(instr);
- }
-
- void AssemblyGeneratorX86::h_FPSQRT(Instruction& instr, int i) {
- genaf(instr, i);
- asmCode << "\tandps xmm0, xmm10" << std::endl;
- asmCode << "\tsqrtpd " << regF[instr.regc % RegistersCount] << ", xmm0" << std::endl;
- gencf(instr, false);
- }
-
- void AssemblyGeneratorX86::h_FPROUND(Instruction& instr, int i) {
- genar(instr, i);
- asmCode << "\tmov rcx, rax" << std::endl;
- int rotate = (13 - (instr.imm8 & 63)) & 63;
+ //6 uOPs
+ void AssemblyGeneratorX86::h_CFROUND(Instruction& instr, int i) {
+ asmCode << "\tmov rax, " << regR[instr.dst] << std::endl;
+ int rotate = (13 - (instr.alt & 63)) & 63;
if (rotate != 0)
asmCode << "\trol rax, " << rotate << std::endl;
asmCode << "\tand eax, 24576" << std::endl;
asmCode << "\tor eax, 40896" << std::endl;
- asmCode << "\tmov dword ptr [rsp - 8], eax" << std::endl;
- asmCode << "\tldmxcsr dword ptr [rsp - 8]" << std::endl;
- gencr(instr, false);
+ asmCode << "\tmov dword ptr [rsp-8], eax" << std::endl;
+ asmCode << "\tldmxcsr dword ptr [rsp-8]" << std::endl;
}
- static inline const char* jumpCondition(Instruction& instr, bool invert = false) {
- switch ((instr.locb & 7) ^ invert)
+ static inline const char* condition(Instruction& instr, bool invert = false) {
+ switch (((instr.alt >> 2) & 7) ^ invert)
{
case 0:
- return "jbe";
+ return "be";
case 1:
- return "ja";
+ return "a";
case 2:
- return "js";
+ return "s";
case 3:
- return "jns";
+ return "ns";
case 4:
- return "jo";
+ return "o";
case 5:
- return "jno";
+ return "no";
case 6:
- return "jl";
+ return "l";
case 7:
- return "jge";
+ return "ge";
}
}
- void AssemblyGeneratorX86::h_JUMP(Instruction& instr, int i) {
- genar(instr, i);
- gencr(instr);
- asmCode << "\tcmp " << regR32[instr.regb % RegistersCount] << ", " << instr.imm32 << std::endl;
- asmCode << "\t" << jumpCondition(instr);
- asmCode << " rx_i_" << wrapInstr(i + (instr.imm8 & 127) + 2) << std::endl;
+ //4 uOPs
+ void AssemblyGeneratorX86::h_COND_R(Instruction& instr, int i) {
+ asmCode << "\txor ecx, ecx" << std::endl;
+ asmCode << "\tcmp " << regR32[instr.src] << ", " << instr.imm32 << std::endl;
+ asmCode << "\tset" << condition(instr) << " cl" << std::endl;
+ asmCode << "\tadd " << regR[instr.dst] << ", rcx" << std::endl;
}
- void AssemblyGeneratorX86::h_CALL(Instruction& instr, int i) {
- genar(instr, i);
- gencr(instr);
- asmCode << "\tcmp " << regR32[instr.regb % RegistersCount] << ", " << instr.imm32 << std::endl;
- asmCode << "\t" << jumpCondition(instr, true);
- asmCode << " short rx_i_" << wrapInstr(i + 1) << std::endl;
- asmCode << "\tcall rx_i_" << wrapInstr(i + (instr.imm8 & 127) + 2) << std::endl;
- }
-
- void AssemblyGeneratorX86::h_RET(Instruction& instr, int i) {
- genar(instr, i);
- gencr(instr);
- asmCode << "\tcmp rsp, " << regStackBeginAddr << std::endl;
- asmCode << "\tje short rx_i_" << wrapInstr(i + 1) << std::endl;
- asmCode << "\tret" << std::endl;
+ //6 uOPs
+ void AssemblyGeneratorX86::h_COND_M(Instruction& instr, int i) {
+ asmCode << "\txor ecx, ecx" << std::endl;
+ genAddressReg(instr);
+ asmCode << "\tcmp dword ptr [rsi+rax], " << instr.imm32 << std::endl;
+ asmCode << "\tset" << condition(instr) << " cl" << std::endl;
+ asmCode << "\tadd " << regR[instr.dst] << ", rcx" << std::endl;
}
#include "instructionWeights.hpp"
#define INST_HANDLE(x) REPN(&AssemblyGeneratorX86::h_##x, WT(x))
InstructionGenerator AssemblyGeneratorX86::engine[256] = {
- INST_HANDLE(ADD_64)
- INST_HANDLE(ADD_32)
- INST_HANDLE(SUB_64)
- INST_HANDLE(SUB_32)
- INST_HANDLE(MUL_64)
- INST_HANDLE(MULH_64)
- INST_HANDLE(MUL_32)
- INST_HANDLE(IMUL_32)
- INST_HANDLE(IMULH_64)
- INST_HANDLE(DIV_64)
- INST_HANDLE(IDIV_64)
- INST_HANDLE(AND_64)
- INST_HANDLE(AND_32)
- INST_HANDLE(OR_64)
- INST_HANDLE(OR_32)
- INST_HANDLE(XOR_64)
- INST_HANDLE(XOR_32)
- INST_HANDLE(SHL_64)
- INST_HANDLE(SHR_64)
- INST_HANDLE(SAR_64)
- INST_HANDLE(ROL_64)
- INST_HANDLE(ROR_64)
- INST_HANDLE(FPADD)
- INST_HANDLE(FPSUB)
- INST_HANDLE(FPMUL)
- INST_HANDLE(FPDIV)
- INST_HANDLE(FPSQRT)
- INST_HANDLE(FPROUND)
- INST_HANDLE(JUMP)
- INST_HANDLE(CALL)
- INST_HANDLE(RET)
+ //Integer
+ INST_HANDLE(IADD_R)
+ INST_HANDLE(IADD_M)
+ INST_HANDLE(IADD_RC)
+ INST_HANDLE(ISUB_R)
+ INST_HANDLE(ISUB_M)
+ INST_HANDLE(IMUL_9C)
+ INST_HANDLE(IMUL_R)
+ INST_HANDLE(IMUL_M)
+ INST_HANDLE(IMULH_R)
+ INST_HANDLE(IMULH_M)
+ INST_HANDLE(ISMULH_R)
+ INST_HANDLE(ISMULH_M)
+ INST_HANDLE(IDIV_C)
+ INST_HANDLE(ISDIV_C)
+ INST_HANDLE(INEG_R)
+ INST_HANDLE(IXOR_R)
+ INST_HANDLE(IXOR_M)
+ INST_HANDLE(IROR_R)
+ INST_HANDLE(IROL_R)
+
+ //Common floating point
+ INST_HANDLE(FPSWAP_R)
+
+ //Floating point group F
+ INST_HANDLE(FPADD_R)
+ INST_HANDLE(FPADD_M)
+ INST_HANDLE(FPSUB_R)
+ INST_HANDLE(FPSUB_M)
+ INST_HANDLE(FPNEG_R)
+
+ //Floating point group E
+ INST_HANDLE(FPMUL_R)
+ INST_HANDLE(FPMUL_M)
+ INST_HANDLE(FPDIV_R)
+ INST_HANDLE(FPDIV_M)
+ INST_HANDLE(FPSQRT_R)
+
+ //Control
+ INST_HANDLE(COND_R)
+ INST_HANDLE(COND_M)
+ INST_HANDLE(CFROUND)
};
}
\ No newline at end of file
diff --git a/src/AssemblyGeneratorX86.hpp b/src/AssemblyGeneratorX86.hpp
index d2e2eb0..2d3c9a6 100644
--- a/src/AssemblyGeneratorX86.hpp
+++ b/src/AssemblyGeneratorX86.hpp
@@ -47,39 +47,43 @@ namespace RandomX {
void genbf(Instruction&, const char*);
void gencr(Instruction&, bool);
void gencf(Instruction&, bool);
+ void genAddressReg(Instruction&, const char*);
+ int32_t genAddressImm(Instruction&);
void generateCode(Instruction&, int);
- void h_ADD_64(Instruction&, int);
- void h_ADD_32(Instruction&, int);
- void h_SUB_64(Instruction&, int);
- void h_SUB_32(Instruction&, int);
- void h_MUL_64(Instruction&, int);
- void h_MULH_64(Instruction&, int);
- void h_MUL_32(Instruction&, int);
- void h_IMUL_32(Instruction&, int);
- void h_IMULH_64(Instruction&, int);
- void h_DIV_64(Instruction&, int);
- void h_IDIV_64(Instruction&, int);
- void h_AND_64(Instruction&, int);
- void h_AND_32(Instruction&, int);
- void h_OR_64(Instruction&, int);
- void h_OR_32(Instruction&, int);
- void h_XOR_64(Instruction&, int);
- void h_XOR_32(Instruction&, int);
- void h_SHL_64(Instruction&, int);
- void h_SHR_64(Instruction&, int);
- void h_SAR_64(Instruction&, int);
- void h_ROL_64(Instruction&, int);
- void h_ROR_64(Instruction&, int);
- void h_FPADD(Instruction&, int);
- void h_FPSUB(Instruction&, int);
- void h_FPMUL(Instruction&, int);
- void h_FPDIV(Instruction&, int);
- void h_FPSQRT(Instruction&, int);
- void h_FPROUND(Instruction&, int);
- void h_JUMP(Instruction&, int);
- void h_CALL(Instruction&, int);
- void h_RET(Instruction&, int);
+ void h_IADD_R(Instruction&, int);
+ void h_IADD_M(Instruction&, int);
+ void h_IADD_RC(Instruction&, int);
+ void h_ISUB_R(Instruction&, int);
+ void h_ISUB_M(Instruction&, int);
+ void h_IMUL_9C(Instruction&, int);
+ void h_IMUL_R(Instruction&, int);
+ void h_IMUL_M(Instruction&, int);
+ void h_IMULH_R(Instruction&, int);
+ void h_IMULH_M(Instruction&, int);
+ void h_ISMULH_R(Instruction&, int);
+ void h_ISMULH_M(Instruction&, int);
+ void h_IDIV_C(Instruction&, int);
+ void h_ISDIV_C(Instruction&, int);
+ void h_INEG_R(Instruction&, int);
+ void h_IXOR_R(Instruction&, int);
+ void h_IXOR_M(Instruction&, int);
+ void h_IROR_R(Instruction&, int);
+ void h_IROL_R(Instruction&, int);
+ void h_FPSWAP_R(Instruction&, int);
+ void h_FPADD_R(Instruction&, int);
+ void h_FPADD_M(Instruction&, int);
+ void h_FPSUB_R(Instruction&, int);
+ void h_FPSUB_M(Instruction&, int);
+ void h_FPNEG_R(Instruction&, int);
+ void h_FPMUL_R(Instruction&, int);
+ void h_FPMUL_M(Instruction&, int);
+ void h_FPDIV_R(Instruction&, int);
+ void h_FPDIV_M(Instruction&, int);
+ void h_FPSQRT_R(Instruction&, int);
+ void h_COND_R(Instruction&, int);
+ void h_COND_M(Instruction&, int);
+ void h_CFROUND(Instruction&, int);
};
}
\ No newline at end of file
diff --git a/src/CompiledVirtualMachine.cpp b/src/CompiledVirtualMachine.cpp
index 5e87b50..f0a63d1 100644
--- a/src/CompiledVirtualMachine.cpp
+++ b/src/CompiledVirtualMachine.cpp
@@ -25,6 +25,12 @@ along with RandomX. If not, see.
namespace RandomX {
+ constexpr int mantissaSize = 52;
+ constexpr int exponentSize = 11;
+ constexpr uint64_t mantissaMask = (1ULL << mantissaSize) - 1;
+ constexpr uint64_t exponentMask = (1ULL << exponentSize) - 1;
+ constexpr int exponentBias = 1023;
+
CompiledVirtualMachine::CompiledVirtualMachine() {
totalSize = 0;
}
@@ -37,25 +43,42 @@ namespace RandomX {
memcpy(scratchpad, mem.ds.dataset + ScratchpadSize * index, ScratchpadSize);
}
+ static uint64_t getSmallPositiveFloatBits(uint64_t entropy) {
+ auto exponent = entropy >> 60; //0..15
+ auto mantissa = entropy & mantissaMask;
+ exponent += exponentBias;
+ exponent &= exponentMask;
+ exponent <<= mantissaSize;
+ return exponent | mantissa;
+ }
+
void CompiledVirtualMachine::initializeProgram(const void* seed) {
Pcg32 gen(seed);
for (unsigned i = 0; i < sizeof(reg) / sizeof(Pcg32::result_type); ++i) {
*(((uint32_t*)®) + i) = gen();
}
FPINIT();
- for (int i = 0; i < RegistersCount; ++i) {
+ /*for (int i = 0; i < RegistersCount / 2; ++i) {
reg.f[i].lo.f64 = (double)reg.f[i].lo.i64;
reg.f[i].hi.f64 = (double)reg.f[i].hi.i64;
}
+ for (int i = 0; i < RegistersCount / 2; ++i) {
+ reg.g[i].lo.f64 = std::abs((double)reg.g[i].lo.i64);
+ reg.g[i].hi.f64 = std::abs((double)reg.g[i].hi.i64);
+ }*/
+ for (int i = 0; i < RegistersCount / 2; ++i) {
+ reg.a[i].lo.u64 = getSmallPositiveFloatBits(reg.f[i].lo.u64);
+ reg.a[i].hi.u64 = getSmallPositiveFloatBits(reg.f[i].hi.u64);
+ }
compiler.generateProgram(gen);
mem.ma = (gen() ^ *(((uint32_t*)seed) + 4)) & ~7;
mem.mx = *(((uint32_t*)seed) + 5);
}
void CompiledVirtualMachine::execute() {
- //executeProgram(reg, mem, scratchpad, readDataset);
+ executeProgram(reg, mem, scratchpad, InstructionCount);
totalSize += compiler.getCodeSize();
- compiler.getProgramFunc()(reg, mem, scratchpad);
+ //compiler.getProgramFunc()(reg, mem, scratchpad);
#ifdef TRACEVM
for (int32_t i = InstructionCount - 1; i >= 0; --i) {
std::cout << std::hex << tracepad[i].u64 << std::endl;
diff --git a/src/Instruction.cpp b/src/Instruction.cpp
index b668a81..c766ffd 100644
--- a/src/Instruction.cpp
+++ b/src/Instruction.cpp
@@ -18,54 +18,391 @@ along with RandomX. If not, see.
*/
#include "Instruction.hpp"
+#include "common.hpp"
namespace RandomX {
void Instruction::print(std::ostream& os) const {
- os << " A: loc = " << std::dec << (loca & 7) << ", reg: " << (rega & 7) << std::endl;
- os << " B: loc = " << (locb & 7) << ", reg: " << (regb & 7) << std::endl;
- os << " C: loc = " << (locc & 7) << ", reg: " << (regc & 7) << std::endl;
- os << " addra = " << std::hex << addra << std::endl;
- os << " addrc = " << addrc << std::endl;
- os << " imm8 = " << std::dec << (int)imm8 << std::endl;
- os << " imm32 = " << imm32 << std::endl;
+ os << names[opcode] << " ";
+ auto handler = engine[opcode];
+ (this->*handler)(os);
+ }
+
+ void Instruction::genAddressReg(std::ostream& os) const {
+ os << ((alt % 4) ? "L1" : "L2") << "[r" << (int)src << "]";
+ }
+
+ void Instruction::genAddressImm(std::ostream& os) const {
+ os << ((alt % 4) ? "L1" : "L2") << "[" << (imm32 & ((alt % 4) ? ScratchpadL1Mask : ScratchpadL2Mask)) << "]";
+ }
+
+ void Instruction::h_IADD_R(std::ostream& os) const {
+ if (src != dst) {
+ os << "r" << (int)dst << ", r" << (int)src << std::endl;
}
+ else {
+ os << "r" << (int)dst << ", " << imm32 << std::endl;
+ }
+ }
+
+ void Instruction::h_IADD_M(std::ostream& os) const {
+ if (src != dst) {
+ os << "r" << (int)dst << ", ";
+ genAddressReg(os);
+ os << std::endl;
+ }
+ else {
+ os << "r" << (int)dst << ", ";
+ genAddressImm(os);
+ os << std::endl;
+ }
+ }
+
+ void Instruction::h_IADD_RC(std::ostream& os) const {
+ os << "r" << (int)dst << ", r" << (int)src << ", " << imm32 << std::endl;
+ }
+
+ //1 uOP
+ void Instruction::h_ISUB_R(std::ostream& os) const {
+ if (src != dst) {
+ os << "r" << (int)dst << ", r" << (int)src << std::endl;
+ }
+ else {
+ os << "r" << (int)dst << ", " << imm32 << std::endl;
+ }
+ }
+
+ void Instruction::h_ISUB_M(std::ostream& os) const {
+ if (src != dst) {
+ os << "r" << (int)dst << ", ";
+ genAddressReg(os);
+ os << std::endl;
+ }
+ else {
+ os << "r" << (int)dst << ", ";
+ genAddressImm(os);
+ os << std::endl;
+ }
+ }
+
+ void Instruction::h_IMUL_9C(std::ostream& os) const {
+ os << "r" << (int)dst << ", " << imm32 << std::endl;
+ }
+
+ void Instruction::h_IMUL_R(std::ostream& os) const {
+ if (src != dst) {
+ os << "r" << (int)dst << ", r" << (int)src << std::endl;
+ }
+ else {
+ os << "r" << (int)dst << ", " << imm32 << std::endl;
+ }
+ }
+
+ void Instruction::h_IMUL_M(std::ostream& os) const {
+ if (src != dst) {
+ os << "r" << (int)dst << ", ";
+ genAddressReg(os);
+ os << std::endl;
+ }
+ else {
+ os << "r" << (int)dst << ", ";
+ genAddressImm(os);
+ os << std::endl;
+ }
+ }
+
+ void Instruction::h_IMULH_R(std::ostream& os) const {
+ if (src != dst) {
+ os << "r" << (int)dst << ", r" << (int)src << std::endl;
+ }
+ else {
+ os << "r" << (int)dst << ", " << imm32 << std::endl;
+ }
+ }
+
+ void Instruction::h_IMULH_M(std::ostream& os) const {
+ if (src != dst) {
+ os << "r" << (int)dst << ", ";
+ genAddressReg(os);
+ os << std::endl;
+ }
+ else {
+ os << "r" << (int)dst << ", ";
+ genAddressImm(os);
+ os << std::endl;
+ }
+ }
+
+ void Instruction::h_ISMULH_R(std::ostream& os) const {
+ if (src != dst) {
+ os << "r" << (int)dst << ", r" << (int)src << std::endl;
+ }
+ else {
+ os << "r" << (int)dst << ", " << imm32 << std::endl;
+ }
+ }
+
+ void Instruction::h_ISMULH_M(std::ostream& os) const {
+ if (src != dst) {
+ os << "r" << (int)dst << ", ";
+ genAddressReg(os);
+ os << std::endl;
+ }
+ else {
+ os << "r" << (int)dst << ", ";
+ genAddressImm(os);
+ os << std::endl;
+ }
+ }
+
+ void Instruction::h_INEG_R(std::ostream& os) const {
+ os << "r" << (int)dst << std::endl;
+ }
+
+ void Instruction::h_IXOR_R(std::ostream& os) const {
+ if (src != dst) {
+ os << "r" << (int)dst << ", r" << (int)src << std::endl;
+ }
+ else {
+ os << "r" << (int)dst << ", " << imm32 << std::endl;
+ }
+ }
+
+ void Instruction::h_IXOR_M(std::ostream& os) const {
+ if (src != dst) {
+ os << "r" << (int)dst << ", ";
+ genAddressReg(os);
+ os << std::endl;
+ }
+ else {
+ os << "r" << (int)dst << ", ";
+ genAddressImm(os);
+ os << std::endl;
+ }
+ }
+
+ void Instruction::h_IROR_R(std::ostream& os) const {
+ if (src != dst) {
+ os << "r" << (int)dst << ", r" << (int)src << std::endl;
+ }
+ else {
+ os << "r" << (int)dst << ", " << (imm32 & 63) << std::endl;
+ }
+ }
+
+ void Instruction::h_IROL_R(std::ostream& os) const {
+ if (src != dst) {
+ os << "r" << (int)dst << ", r" << (int)src << std::endl;
+ }
+ else {
+ os << "r" << (int)dst << ", " << (imm32 & 63) << std::endl;
+ }
+ }
+
+ void Instruction::h_IDIV_C(std::ostream& os) const {
+ os << "r" << (int)dst << ", " << (uint32_t)imm32 << std::endl;
+ }
+
+ void Instruction::h_ISDIV_C(std::ostream& os) const {
+ os << "r" << (int)dst << ", " << imm32 << std::endl;
+ }
+
+ void Instruction::h_FPSWAP_R(std::ostream& os) const {
+ const char reg = (dst >= 4) ? 'e' : 'f';
+ auto dstIndex = dst % 4;
+ os << reg << dstIndex << std::endl;
+ }
+
+ void Instruction::h_FPADD_R(std::ostream& os) const {
+ auto dstIndex = dst % 4;
+ auto srcIndex = src % 4;
+ os << "f" << dstIndex << ", a" << srcIndex << std::endl;
+ }
+
+ void Instruction::h_FPADD_M(std::ostream& os) const {
+ auto dstIndex = dst % 4;
+ os << "f" << dstIndex << ", ";
+ genAddressReg(os);
+ os << std::endl;
+ }
+
+ void Instruction::h_FPSUB_R(std::ostream& os) const {
+ auto dstIndex = dst % 4;
+ auto srcIndex = src % 4;
+ os << "f" << dstIndex << ", a" << srcIndex << std::endl;
+ }
+
+ void Instruction::h_FPSUB_M(std::ostream& os) const {
+ auto dstIndex = dst % 4;
+ os << "f" << dstIndex << ", ";
+ genAddressReg(os);
+ os << std::endl;
+ }
+
+ void Instruction::h_FPNEG_R(std::ostream& os) const {
+ auto dstIndex = dst % 4;
+ os << "f" << dstIndex << std::endl;
+ }
+
+ void Instruction::h_FPMUL_R(std::ostream& os) const {
+ auto dstIndex = dst % 4;
+ auto srcIndex = src % 4;
+ os << "e" << dstIndex << ", a" << srcIndex << std::endl;
+ }
+
+ void Instruction::h_FPMUL_M(std::ostream& os) const {
+ auto dstIndex = dst % 4;
+ os << "e" << dstIndex << ", ";
+ genAddressReg(os);
+ os << std::endl;
+ }
+
+ void Instruction::h_FPDIV_R(std::ostream& os) const {
+ auto dstIndex = dst % 4;
+ auto srcIndex = src % 4;
+ os << "e" << dstIndex << ", a" << srcIndex << std::endl;
+ }
+
+ void Instruction::h_FPDIV_M(std::ostream& os) const {
+ auto dstIndex = dst % 4;
+ os << "e" << dstIndex << ", ";
+ genAddressReg(os);
+ os << std::endl;
+ }
+
+ void Instruction::h_FPSQRT_R(std::ostream& os) const {
+ auto dstIndex = dst % 4;
+ os << "e" << dstIndex << std::endl;
+ }
+
+ void Instruction::h_CFROUND(std::ostream& os) const {
+ os << "r" << (int)dst << ", " << (alt & 63) << std::endl;
+ }
+
+ static inline const char* condition(int index) {
+ switch (index)
+ {
+ case 0:
+ return "be";
+ case 1:
+ return "ab";
+ case 2:
+ return "sg";
+ case 3:
+ return "ns";
+ case 4:
+ return "of";
+ case 5:
+ return "no";
+ case 6:
+ return "lt";
+ case 7:
+ return "ge";
+ }
+ }
+
+ void Instruction::h_COND_R(std::ostream& os) const {
+ os << "r" << (int)dst << ", " << condition((alt >> 2) & 7) << "(r" << (int)src << ", " << imm32 << ")" << std::endl;
+ }
+
+ void Instruction::h_COND_M(std::ostream& os) const {
+ os << "r" << (int)dst << ", " << condition((alt >> 2) & 7) << "(";
+ genAddressReg(os);
+ os << ", " << imm32 << ")" << std::endl;
+ }
#include "instructionWeights.hpp"
#define INST_NAME(x) REPN(#x, WT(x))
+#define INST_HANDLE(x) REPN(&Instruction::h_##x, WT(x))
const char* Instruction::names[256] = {
- INST_NAME(ADD_64)
- INST_NAME(ADD_32)
- INST_NAME(SUB_64)
- INST_NAME(SUB_32)
- INST_NAME(MUL_64)
- INST_NAME(MULH_64)
- INST_NAME(MUL_32)
- INST_NAME(IMUL_32)
- INST_NAME(IMULH_64)
- INST_NAME(DIV_64)
- INST_NAME(IDIV_64)
- INST_NAME(AND_64)
- INST_NAME(AND_32)
- INST_NAME(OR_64)
- INST_NAME(OR_32)
- INST_NAME(XOR_64)
- INST_NAME(XOR_32)
- INST_NAME(SHL_64)
- INST_NAME(SHR_64)
- INST_NAME(SAR_64)
- INST_NAME(ROL_64)
- INST_NAME(ROR_64)
- INST_NAME(FPADD)
- INST_NAME(FPSUB)
- INST_NAME(FPMUL)
- INST_NAME(FPDIV)
- INST_NAME(FPSQRT)
- INST_NAME(FPROUND)
- INST_NAME(JUMP)
- INST_NAME(CALL)
- INST_NAME(RET)
+ //Integer
+ INST_NAME(IADD_R)
+ INST_NAME(IADD_M)
+ INST_NAME(IADD_RC)
+ INST_NAME(ISUB_R)
+ INST_NAME(ISUB_M)
+ INST_NAME(IMUL_9C)
+ INST_NAME(IMUL_R)
+ INST_NAME(IMUL_M)
+ INST_NAME(IMULH_R)
+ INST_NAME(IMULH_M)
+ INST_NAME(ISMULH_R)
+ INST_NAME(ISMULH_M)
+ INST_NAME(IDIV_C)
+ INST_NAME(ISDIV_C)
+ INST_NAME(INEG_R)
+ INST_NAME(IXOR_R)
+ INST_NAME(IXOR_M)
+ INST_NAME(IROR_R)
+ INST_NAME(IROL_R)
+
+ //Common floating point
+ INST_NAME(FPSWAP_R)
+
+ //Floating point group F
+ INST_NAME(FPADD_R)
+ INST_NAME(FPADD_M)
+ INST_NAME(FPSUB_R)
+ INST_NAME(FPSUB_M)
+ INST_NAME(FPNEG_R)
+
+ //Floating point group E
+ INST_NAME(FPMUL_R)
+ INST_NAME(FPMUL_M)
+ INST_NAME(FPDIV_R)
+ INST_NAME(FPDIV_M)
+ INST_NAME(FPSQRT_R)
+
+ //Control
+ INST_NAME(COND_R)
+ INST_NAME(COND_M)
+ INST_NAME(CFROUND)
+ };
+
+ InstructionVisualizer Instruction::engine[256] = {
+ //Integer
+ INST_HANDLE(IADD_R)
+ INST_HANDLE(IADD_M)
+ INST_HANDLE(IADD_RC)
+ INST_HANDLE(ISUB_R)
+ INST_HANDLE(ISUB_M)
+ INST_HANDLE(IMUL_9C)
+ INST_HANDLE(IMUL_R)
+ INST_HANDLE(IMUL_M)
+ INST_HANDLE(IMULH_R)
+ INST_HANDLE(IMULH_M)
+ INST_HANDLE(ISMULH_R)
+ INST_HANDLE(ISMULH_M)
+ INST_HANDLE(IDIV_C)
+ INST_HANDLE(ISDIV_C)
+ INST_HANDLE(INEG_R)
+ INST_HANDLE(IXOR_R)
+ INST_HANDLE(IXOR_M)
+ INST_HANDLE(IROR_R)
+ INST_HANDLE(IROL_R)
+
+ //Common floating point
+ INST_HANDLE(FPSWAP_R)
+
+ //Floating point group F
+ INST_HANDLE(FPADD_R)
+ INST_HANDLE(FPADD_M)
+ INST_HANDLE(FPSUB_R)
+ INST_HANDLE(FPSUB_M)
+ INST_HANDLE(FPNEG_R)
+
+ //Floating point group E
+ INST_HANDLE(FPMUL_R)
+ INST_HANDLE(FPMUL_M)
+ INST_HANDLE(FPDIV_R)
+ INST_HANDLE(FPDIV_M)
+ INST_HANDLE(FPSQRT_R)
+
+ //Control
+ INST_HANDLE(COND_R)
+ INST_HANDLE(COND_M)
+ INST_HANDLE(CFROUND)
};
}
\ No newline at end of file
diff --git a/src/Instruction.hpp b/src/Instruction.hpp
index 33c2059..becb983 100644
--- a/src/Instruction.hpp
+++ b/src/Instruction.hpp
@@ -24,21 +24,17 @@ along with RandomX. If not, see.
namespace RandomX {
+ class Instruction;
+
+ typedef void(Instruction::*InstructionVisualizer)(std::ostream&) const;
+
class Instruction {
public:
uint8_t opcode;
- uint8_t loca;
- uint8_t rega;
- uint8_t locb;
- uint8_t regb;
- uint8_t locc;
- uint8_t regc;
- uint8_t imm8;
- int32_t addra;
- union {
- uint32_t addrc;
- int32_t imm32;
- };
+ uint8_t dst;
+ uint8_t src;
+ uint8_t alt;
+ int32_t imm32;
const char* getName() const {
return names[opcode];
}
@@ -49,8 +45,46 @@ namespace RandomX {
private:
void print(std::ostream&) const;
static const char* names[256];
+ static InstructionVisualizer engine[256];
+
+ void genAddressReg(std::ostream& os) const;
+ void genAddressImm(std::ostream& os) const;
+
+ void h_IADD_R(std::ostream&) const;
+ void h_IADD_M(std::ostream&) const;
+ void h_IADD_RC(std::ostream&) const;
+ void h_ISUB_R(std::ostream&) const;
+ void h_ISUB_M(std::ostream&) const;
+ void h_IMUL_9C(std::ostream&) const;
+ void h_IMUL_R(std::ostream&) const;
+ void h_IMUL_M(std::ostream&) const;
+ void h_IMULH_R(std::ostream&) const;
+ void h_IMULH_M(std::ostream&) const;
+ void h_ISMULH_R(std::ostream&) const;
+ void h_ISMULH_M(std::ostream&) const;
+ void h_IDIV_C(std::ostream&) const;
+ void h_ISDIV_C(std::ostream&) const;
+ void h_INEG_R(std::ostream&) const;
+ void h_IXOR_R(std::ostream&) const;
+ void h_IXOR_M(std::ostream&) const;
+ void h_IROR_R(std::ostream&) const;
+ void h_IROL_R(std::ostream&) const;
+ void h_FPSWAP_R(std::ostream&) const;
+ void h_FPADD_R(std::ostream&) const;
+ void h_FPADD_M(std::ostream&) const;
+ void h_FPSUB_R(std::ostream&) const;
+ void h_FPSUB_M(std::ostream&) const;
+ void h_FPNEG_R(std::ostream&) const;
+ void h_FPMUL_R(std::ostream&) const;
+ void h_FPMUL_M(std::ostream&) const;
+ void h_FPDIV_R(std::ostream&) const;
+ void h_FPDIV_M(std::ostream&) const;
+ void h_FPSQRT_R(std::ostream&) const;
+ void h_COND_R(std::ostream&) const;
+ void h_COND_M(std::ostream&) const;
+ void h_CFROUND(std::ostream&) const;
};
- static_assert(sizeof(Instruction) == 16, "Invalid alignment of struct Instruction");
+ static_assert(sizeof(Instruction) == 8, "Invalid alignment of struct Instruction");
}
\ No newline at end of file
diff --git a/src/InterpretedVirtualMachine.cpp b/src/InterpretedVirtualMachine.cpp
index d7e4fc4..d145e78 100644
--- a/src/InterpretedVirtualMachine.cpp
+++ b/src/InterpretedVirtualMachine.cpp
@@ -130,333 +130,10 @@ namespace RandomX {
#endif
}
- convertible_t InterpretedVirtualMachine::loada(Instruction& instr) {
- convertible_t& rega = reg.r[instr.rega % RegistersCount];
- rega.i64 ^= instr.addra; //sign-extend addra
- addr_t addr = rega.u32;
-
- if ((ic % 64) == 0) {
- addr = currentTransform->apply(addr);
-#ifdef STATS
- datasetAccess[mem.ma / (DatasetBlockCount / 256) / CacheLineSize]++;
-#endif
- readDataset(addr, mem, reg);
- }
-
- if ((instr.loca & 192) == 0) {
- mem.mx ^= addr;
- }
-
- if (instr.loca & 3) {
- return scratchpad[addr % ScratchpadL1];
- }
- else {
- return scratchpad[addr % ScratchpadL2];
- }
- }
-
- convertible_t InterpretedVirtualMachine::loadbia(Instruction& instr) {
- if (instr.locb & 3) {
- return reg.r[instr.regb % RegistersCount];
- }
- else {
- convertible_t temp;
- temp.i64 = instr.imm32; //sign-extend imm32
- return temp;
- }
- }
-
- convertible_t InterpretedVirtualMachine::loadbiashift(Instruction& instr) {
- if (instr.locb & 1) {
- return reg.r[instr.regb % RegistersCount];
- }
- else {
- convertible_t temp;
- temp.u64 = instr.imm8;
- return temp;
- }
- }
-
- convertible_t InterpretedVirtualMachine::loadbiadiv(Instruction& instr) {
- if (instr.locb & 3) {
- convertible_t temp;
- temp.u64 = instr.imm32;
- return temp;
- }
- else {
- return reg.r[instr.regb % RegistersCount];
- }
- }
-
- convertible_t& InterpretedVirtualMachine::getcr(Instruction& inst) {
- addr_t addr;
- switch (inst.locc & 7)
- {
- case 0:
- addr = reg.r[inst.regc % RegistersCount].u32 ^ inst.addrc;
- return scratchpad[addr % ScratchpadL2];
-
- case 1:
- case 2:
- case 3:
- addr = reg.r[inst.regc % RegistersCount].u32 ^ inst.addrc;
- return scratchpad[addr % ScratchpadL1];
-
- case 4:
- case 5:
- case 6:
- case 7:
- return reg.r[inst.regc % RegistersCount];
- }
- }
-
- void InterpretedVirtualMachine::writecf(Instruction& inst, fpu_reg_t& regc) {
- addr_t addr;
- switch (inst.locc & 7)
- {
- case 4:
- addr = reg.r[inst.regc % RegistersCount].u32 ^ inst.addrc;
- scratchpad[addr % ScratchpadL2] = (inst.locc & 8) ? regc.hi : regc.lo;
- break;
-
- case 5:
- case 6:
- case 7:
- addr = reg.r[inst.regc % RegistersCount].u32 ^ inst.addrc;
- scratchpad[addr % ScratchpadL1] = (inst.locc & 8) ? regc.hi : regc.lo;
-
- default:
- break;
- }
- }
-
-#define ALU_RETIRE(x) x(a, b, c); \
- if(trace) std::cout << std::hex << /*a.u64 << " " << b.u64 << " " <<*/ c.u64 << std::endl;
-
-#define CHECK_NOP_FPDIV(b, c)
-#ifndef STATS
-#define CHECK_NOP_FPADD(b, c)
-#define CHECK_NOP_FPSUB(b, c)
-#define CHECK_NOP_FPMUL(b, c)
-#else
-#define CHECK_NOP_FPADD(b, c) bool loeq = (b.lo.u64 == c.lo.u64); bool hieq = (b.hi.u64 == c.hi.u64); count_FPADD_nop += loeq + hieq; if(loeq && hieq) count_FPADD_nop2++;
-#define CHECK_NOP_FPSUB(b, c) bool loeq = ((b.lo.u64 & INT64_MAX) == (c.lo.u64 & INT64_MAX)); bool hieq = ((b.hi.u64 & INT64_MAX) == (c.hi.u64 & INT64_MAX)); count_FPSUB_nop += loeq + hieq; if(loeq && hieq) count_FPSUB_nop2++;
-#define CHECK_NOP_FPMUL(b, c) bool loeq = (b.lo.u64 == c.lo.u64); bool hieq = (b.hi.u64 == c.hi.u64); count_FPMUL_nop += loeq + hieq; if(loeq && hieq) count_FPMUL_nop2++;
-#endif
-
-#define FPU_RETIRE(x) x(a, b, c); \
- writecf(inst, c); \
- if(trace) { \
- std::cout << std::hex << ((inst.locc & 8) ? c.hi.u64 : c.lo.u64) << std::endl; \
- } \
- if(fpuCheck) { \
- if(c.hi.f64 != c.hi.f64 || c.lo.f64 != c.lo.f64) { \
- std::stringstream ss; \
- ss << "NaN result of " << #x << "(" << std::hex << a.u64 << ", " << b.hi.u64 << " " << b.lo.u64 << ") = " << c.hi.u64 << " " << c.lo.u64 << std::endl; \
- throw std::runtime_error(ss.str()); \
- } else if (std::fpclassify(c.hi.f64) == FP_SUBNORMAL || std::fpclassify(c.lo.f64) == FP_SUBNORMAL) {\
- std::stringstream ss; \
- ss << "Denormal result of " << #x << "(" << std::hex << a.u64 << ", " << b.hi.u64 << " " << b.lo.u64 << ") = " << c.hi.u64 << " " << c.lo.u64 << std::endl; \
- throw std::runtime_error(ss.str()); \
- } \
- }
-
-#ifdef STATS
-#define INC_COUNT(x) count_##x++;
-#else
-#define INC_COUNT(x)
-#endif
-
-#define FPU_RETIRE_FPSQRT(x) FPSQRT(a, b, c); \
- writecf(inst, c); \
- if(trace) std::cout << std::hex << ((inst.locc & 8) ? c.hi.u64 : c.lo.u64) << std::endl;
-
-#define FPU_RETIRE_FPROUND(x) FPROUND(a, b, c); \
- writecflo(inst, c); \
- if(trace) std::cout << std::hex << c.lo.u64 << std::endl;
-
-#define ALU_INST(x) void InterpretedVirtualMachine::h_##x(Instruction& inst) { \
- INC_COUNT(x) \
- convertible_t a = loada(inst); \
- convertible_t b = loadbia(inst); \
- convertible_t& c = getcr(inst); \
- ALU_RETIRE(x) \
- }
-
-#define ALU_INST_SR(x) void InterpretedVirtualMachine::h_##x(Instruction& inst) { \
- INC_COUNT(x) \
- convertible_t a = loada(inst); \
- convertible_t b = loadbiashift(inst); \
- convertible_t& c = getcr(inst); \
- ALU_RETIRE(x) \
- }
-
-#define ALU_INST_DIV(x) void InterpretedVirtualMachine::h_##x(Instruction& inst) { \
- INC_COUNT(x) \
- convertible_t a = loada(inst); \
- convertible_t b = loadbiadiv(inst); \
- convertible_t& c = getcr(inst); \
- ALU_RETIRE(x) \
- }
-
-#define FPU_INST(x) void InterpretedVirtualMachine::h_##x(Instruction& inst) { \
- INC_COUNT(x) \
- convertible_t a = loada(inst); \
- fpu_reg_t& b = reg.f[inst.regb % RegistersCount]; \
- fpu_reg_t btemp = b; \
- fpu_reg_t& c = reg.f[inst.regc % RegistersCount]; \
- FPU_RETIRE(x) \
- CHECK_NOP_##x(btemp, c) \
- }
-
-#define FPU_INST_NB(x) void InterpretedVirtualMachine::h_##x(Instruction& inst) { \
- INC_COUNT(x) \
- convertible_t a = loada(inst); \
- fpu_reg_t b; \
- fpu_reg_t& c = reg.f[inst.regc % RegistersCount]; \
- FPU_RETIRE_##x(x) \
- }
-
- ALU_INST(ADD_64)
- ALU_INST(ADD_32)
- ALU_INST(SUB_64)
- ALU_INST(SUB_32)
- ALU_INST(MUL_64)
- ALU_INST(MULH_64)
- ALU_INST(MUL_32)
- ALU_INST(IMUL_32)
- ALU_INST(IMULH_64)
- ALU_INST_DIV(DIV_64)
- ALU_INST_DIV(IDIV_64)
- ALU_INST(AND_64)
- ALU_INST(AND_32)
- ALU_INST(OR_64)
- ALU_INST(OR_32)
- ALU_INST(XOR_64)
- ALU_INST(XOR_32)
-
- ALU_INST_SR(SHL_64)
- ALU_INST_SR(SHR_64)
- ALU_INST_SR(SAR_64)
- ALU_INST_SR(ROL_64)
- ALU_INST_SR(ROR_64)
-
- FPU_INST(FPADD)
- FPU_INST(FPSUB)
- FPU_INST(FPMUL)
- FPU_INST(FPDIV)
- FPU_INST_NB(FPSQRT)
-
- void InterpretedVirtualMachine::h_FPROUND(Instruction& inst) {
- convertible_t a = loada(inst);
- convertible_t& c = getcr(inst);
- c.u64 = a.u64;
- if (trace) std::cout << std::hex << a.u64 << std::endl;
- FPROUND(a, inst.imm8);
- }
-
- void InterpretedVirtualMachine::h_JUMP(Instruction& inst) {
- convertible_t a = loada(inst);
- convertible_t& c = getcr(inst);
- c.u64 = a.u64;
- if (trace) std::cout << std::hex << a.u64 << std::endl;
- if (JMP_COND(inst.locb, reg.r[inst.regb % RegistersCount], inst.imm32)) {
-#ifdef STATS
- count_JUMP_taken++;
- count_jump_taken[inst.locb & 7]++;
-#endif
- pc += (inst.imm8 & 127) + 1;
- pc = pc % ProgramLength;
- }
-#ifdef STATS
- else {
- count_JUMP_not_taken++;
- count_jump_not_taken[inst.locb & 7]++;
- }
-#endif
- }
-
- void InterpretedVirtualMachine::h_CALL(Instruction& inst) {
- convertible_t a = loada(inst);
- convertible_t& c = getcr(inst);
- c.u64 = a.u64;
- if (trace) std::cout << std::hex << a.u64 << std::endl;
- if (JMP_COND(inst.locb, reg.r[inst.regb % RegistersCount], inst.imm32)) {
-#ifdef STATS
- count_CALL_taken++;
- count_jump_taken[inst.locb & 7]++;
- count_retdepth = std::max(0, count_retdepth - 1);
-#endif
- stackPush(pc);
-#ifdef STATS
- count_max_stack = std::max(count_max_stack, (int)stack.size());
-#endif
- pc += (inst.imm8 & 127) + 1;
- pc = pc % ProgramLength;
- }
-#ifdef STATS
- else {
- count_CALL_not_taken++;
- count_jump_not_taken[inst.locb & 7]++;
- }
-#endif
- }
-
- void InterpretedVirtualMachine::h_RET(Instruction& inst) {
- convertible_t a = loada(inst);
- convertible_t& c = getcr(inst);
- c.u64 = a.u64;
- if (trace) std::cout << std::hex << a.u64 << std::endl;
- if (stack.size() > 0) {
-#ifdef STATS
- count_RET_taken++;
- count_retdepth++;
- count_retdepth_max = std::max(count_retdepth_max, count_retdepth);
-#endif
- auto raddr = stackPopAddress();
- pc = raddr;
- }
-#ifdef STATS
- else {
- count_RET_stack_empty++;
- }
-#endif
- }
-
#include "instructionWeights.hpp"
#define INST_HANDLE(x) REPN(&InterpretedVirtualMachine::h_##x, WT(x))
InstructionHandler InterpretedVirtualMachine::engine[256] = {
- INST_HANDLE(ADD_64)
- INST_HANDLE(ADD_32)
- INST_HANDLE(SUB_64)
- INST_HANDLE(SUB_32)
- INST_HANDLE(MUL_64)
- INST_HANDLE(MULH_64)
- INST_HANDLE(MUL_32)
- INST_HANDLE(IMUL_32)
- INST_HANDLE(IMULH_64)
- INST_HANDLE(DIV_64)
- INST_HANDLE(IDIV_64)
- INST_HANDLE(AND_64)
- INST_HANDLE(AND_32)
- INST_HANDLE(OR_64)
- INST_HANDLE(OR_32)
- INST_HANDLE(XOR_64)
- INST_HANDLE(XOR_32)
- INST_HANDLE(SHL_64)
- INST_HANDLE(SHR_64)
- INST_HANDLE(SAR_64)
- INST_HANDLE(ROL_64)
- INST_HANDLE(ROR_64)
- INST_HANDLE(FPADD)
- INST_HANDLE(FPSUB)
- INST_HANDLE(FPMUL)
- INST_HANDLE(FPDIV)
- INST_HANDLE(FPSQRT)
- INST_HANDLE(FPROUND)
- INST_HANDLE(JUMP)
- INST_HANDLE(CALL)
- INST_HANDLE(RET)
+
};
}
\ No newline at end of file
diff --git a/src/JitCompilerX86.cpp b/src/JitCompilerX86.cpp
index ee91fc3..8776d61 100644
--- a/src/JitCompilerX86.cpp
+++ b/src/JitCompilerX86.cpp
@@ -38,9 +38,9 @@ along with RandomX. If not, see.
namespace RandomX {
-#if !defined(_M_X64) && !defined(__x86_64__)
+#if true || !defined(_M_X64) && !defined(__x86_64__)
JitCompilerX86::JitCompilerX86() {
- throw std::runtime_error("JIT compiler only supports x86-64 CPUs");
+ //throw std::runtime_error("JIT compiler only supports x86-64 CPUs");
}
void JitCompilerX86::generateProgram(Pcg32& gen) {
diff --git a/src/Program.cpp b/src/Program.cpp
index 6e94fca..b78a5ee 100644
--- a/src/Program.cpp
+++ b/src/Program.cpp
@@ -30,7 +30,6 @@ namespace RandomX {
void Program::print(std::ostream& os) const {
for (int i = 0; i < RandomX::ProgramLength; ++i) {
auto instr = programBuffer[i];
- os << std::dec << instr.getName() << " (" << i << "):" << std::endl;
os << instr;
}
}
diff --git a/src/common.hpp b/src/common.hpp
index cffa53c..bf235ec 100644
--- a/src/common.hpp
+++ b/src/common.hpp
@@ -46,7 +46,7 @@ namespace RandomX {
constexpr int CacheBlockCount = CacheSize / CacheLineSize;
constexpr int BlockExpansionRatio = DatasetSize / CacheSize;
constexpr int DatasetBlockCount = BlockExpansionRatio * CacheBlockCount;
- constexpr int DatasetIterations = 10;
+ constexpr int DatasetIterations = 3;
#ifdef TRACE
@@ -72,13 +72,15 @@ namespace RandomX {
convertible_t hi;
};
- constexpr int ProgramLength = 512;
- constexpr uint32_t InstructionCount = 1024 * 1024;
+ constexpr int ProgramLength = 256;
+ constexpr uint32_t InstructionCount = 1024;
constexpr uint32_t ScratchpadSize = 1024 * 1024;
constexpr uint32_t ScratchpadLength = ScratchpadSize / sizeof(convertible_t);
constexpr uint32_t ScratchpadL1 = ScratchpadSize / 64 / sizeof(convertible_t);
constexpr uint32_t ScratchpadL2 = ScratchpadSize / 4 / sizeof(convertible_t);
constexpr uint32_t ScratchpadL3 = ScratchpadSize / sizeof(convertible_t);
+ constexpr int ScratchpadL1Mask = (ScratchpadL1 - 1) * 8;
+ constexpr int ScratchpadL2Mask = (ScratchpadL2 - 1) * 8;
constexpr uint32_t TransformationCount = 90;
constexpr int RegistersCount = 8;
@@ -118,17 +120,19 @@ namespace RandomX {
struct RegisterFile {
convertible_t r[RegistersCount];
- fpu_reg_t f[RegistersCount];
+ fpu_reg_t f[RegistersCount / 2];
+ fpu_reg_t g[RegistersCount / 2];
+ fpu_reg_t a[RegistersCount / 2];
};
- static_assert(sizeof(RegisterFile) == 3 * RegistersCount * sizeof(convertible_t), "Invalid alignment of struct RandomX::RegisterFile");
+ static_assert(sizeof(RegisterFile) == 256, "Invalid alignment of struct RandomX::RegisterFile");
typedef void(*DatasetReadFunc)(addr_t, MemoryRegisters&, RegisterFile&);
typedef void(*ProgramFunc)(RegisterFile&, MemoryRegisters&, convertible_t*);
extern "C" {
- void executeProgram(RegisterFile&, MemoryRegisters&, convertible_t*, DatasetReadFunc);
+ void executeProgram(RegisterFile&, MemoryRegisters&, convertible_t*, uint64_t);
}
}
diff --git a/src/executeProgram-win64.asm b/src/executeProgram-win64.asm
index 2da88b5..17e593d 100644
--- a/src/executeProgram-win64.asm
+++ b/src/executeProgram-win64.asm
@@ -21,16 +21,24 @@ _RANDOMX_EXECUTE_PROGRAM SEGMENT PAGE READ EXECUTE
PUBLIC executeProgram
+ALIGN 16
+minDbl:
+db 0, 0, 0, 0, 0, 0, 16, 0, 0, 0, 0, 0, 0, 0, 16, 0
+absMask:
+db 255, 255, 255, 255, 255, 255, 255, 127, 255, 255, 255, 255, 255, 255, 255, 127
+signMask:
+db 0, 0, 0, 0, 0, 0, 0, 128, 0, 0, 0, 0, 0, 0, 0, 128
+
executeProgram PROC
; REGISTER ALLOCATION:
; rax -> temporary
; rbx -> "ic"
; rcx -> temporary
; rdx -> temporary
- ; rsi -> convertible_t& scratchpad
- ; rdi -> beginning of VM stack
+ ; rsi -> scratchpad pointer
+ ; rdi -> dataset pointer
; rbp -> "ma", "mx"
- ; rsp -> end of VM stack
+ ; rsp -> stack pointer
; r8 -> "r0"
; r9 -> "r1"
; r10 -> "r2"
@@ -39,32 +47,22 @@ executeProgram PROC
; r13 -> "r5"
; r14 -> "r6"
; r15 -> "r7"
- ; xmm0 -> temporary
- ; xmm1 -> temporary
+ ; xmm0 -> "f0"
+ ; xmm1 -> "f1"
; xmm2 -> "f2"
; xmm3 -> "f3"
- ; xmm4 -> "f4"
- ; xmm5 -> "f5"
- ; xmm6 -> "f6"
- ; xmm7 -> "f7"
- ; xmm8 -> "f0"
- ; xmm9 -> "f1"
- ; xmm10 -> absolute value mask
-
- ; STACK STRUCTURE:
- ; |
- ; |
- ; | saved registers
- ; |
- ; v
- ; [rbx+8] RegisterFile& registerFile
- ; [rbx+0] uint8_t* dataset
- ; |
- ; |
- ; | VM stack
- ; |
- ; v
- ; [rsp] last element of VM stack
+ ; xmm4 -> "e0"
+ ; xmm5 -> "e1"
+ ; xmm6 -> "e2"
+ ; xmm7 -> "e3"
+ ; xmm8 -> "a0"
+ ; xmm9 -> "a1"
+ ; xmm10 -> "a2"
+ ; xmm11 -> "a3"
+ ; xmm12 -> temporary
+ ; xmm13 -> DBL_MIN
+ ; xmm14 -> absolute value mask
+ ; xmm15 -> sign mask
; store callee-saved registers
push rbx
@@ -81,100 +79,117 @@ executeProgram PROC
movdqu xmmword ptr [rsp+32], xmm8
movdqu xmmword ptr [rsp+16], xmm9
movdqu xmmword ptr [rsp+0], xmm10
+ sub rsp, 80
+ movdqu xmmword ptr [rsp+64], xmm11
+ movdqu xmmword ptr [rsp+48], xmm12
+ movdqu xmmword ptr [rsp+32], xmm13
+ movdqu xmmword ptr [rsp+16], xmm14
+ movdqu xmmword ptr [rsp+0], xmm15
; function arguments
push rcx ; RegisterFile& registerFile
mov rbp, qword ptr [rdx] ; "mx", "ma"
- mov rax, qword ptr [rdx+8] ; uint8_t* dataset
- push rax
+ mov eax, ebp ; "mx"
+ mov rdi, qword ptr [rdx+8] ; uint8_t* dataset
mov rsi, r8 ; convertible_t* scratchpad
+ mov rbx, r9 ; loop counter
+
+ ;# zero integer registers
+ xor r8, r8
+ xor r9, r9
+ xor r10, r10
+ xor r11, r11
+ xor r12, r12
+ xor r13, r13
+ xor r14, r14
+ xor r15, r15
+
+ ;# load constant registers
+ lea rcx, [rcx+120]
+ movapd xmm8, xmmword ptr [rcx+72]
+ movapd xmm9, xmmword ptr [rcx+88]
+ movapd xmm10, xmmword ptr [rcx+104]
+ movapd xmm11, xmmword ptr [rcx+120]
+ movapd xmm13, xmmword ptr [minDbl]
+ movapd xmm14, xmmword ptr [absMask]
+ movapd xmm15, xmmword ptr [signMask]
- mov rdi, rsp ; beginning of VM stack
- mov ebx, 1048577 ; number of VM instructions to execute + 1
-
- xorps xmm10, xmm10
- cmpeqpd xmm10, xmm10
- psrlq xmm10, 1 ; mask for absolute value = 0x7fffffffffffffff7fffffffffffffff
-
- ; reset rounding mode
- mov dword ptr [rsp-8], 40896
- ldmxcsr dword ptr [rsp-8]
-
- ; load integer registers
- mov r8, qword ptr [rcx+0]
- mov r9, qword ptr [rcx+8]
- mov r10, qword ptr [rcx+16]
- mov r11, qword ptr [rcx+24]
- mov r12, qword ptr [rcx+32]
- mov r13, qword ptr [rcx+40]
- mov r14, qword ptr [rcx+48]
- mov r15, qword ptr [rcx+56]
-
- ; load register f0 hi, lo
- xorps xmm8, xmm8
- cvtsi2sd xmm8, qword ptr [rcx+72]
- pslldq xmm8, 8
- cvtsi2sd xmm8, qword ptr [rcx+64]
-
- ; load register f1 hi, lo
- xorps xmm9, xmm9
- cvtsi2sd xmm9, qword ptr [rcx+88]
- pslldq xmm9, 8
- cvtsi2sd xmm9, qword ptr [rcx+80]
-
- ; load register f2 hi, lo
- xorps xmm2, xmm2
- cvtsi2sd xmm2, qword ptr [rcx+104]
- pslldq xmm2, 8
- cvtsi2sd xmm2, qword ptr [rcx+96]
-
- ; load register f3 hi, lo
- xorps xmm3, xmm3
- cvtsi2sd xmm3, qword ptr [rcx+120]
- pslldq xmm3, 8
- cvtsi2sd xmm3, qword ptr [rcx+112]
-
- lea rcx, [rcx+64]
-
- ; load register f4 hi, lo
- xorps xmm4, xmm4
- cvtsi2sd xmm4, qword ptr [rcx+72]
- pslldq xmm4, 8
- cvtsi2sd xmm4, qword ptr [rcx+64]
-
- ; load register f5 hi, lo
- xorps xmm5, xmm5
- cvtsi2sd xmm5, qword ptr [rcx+88]
- pslldq xmm5, 8
- cvtsi2sd xmm5, qword ptr [rcx+80]
-
- ; load register f6 hi, lo
- xorps xmm6, xmm6
- cvtsi2sd xmm6, qword ptr [rcx+104]
- pslldq xmm6, 8
- cvtsi2sd xmm6, qword ptr [rcx+96]
-
- ; load register f7 hi, lo
- xorps xmm7, xmm7
- cvtsi2sd xmm7, qword ptr [rcx+120]
- pslldq xmm7, 8
- cvtsi2sd xmm7, qword ptr [rcx+112]
-
- jmp program_begin
-
- ; program body
-ALIGN 64
program_begin:
+ xor eax, r8d ;# read address register 1
+ and eax, 262080
+ lea rcx, [rsi+rax]
+ xor r8, qword ptr [rcx+0]
+ xor r9, qword ptr [rcx+8]
+ xor r10, qword ptr [rcx+16]
+ xor r11, qword ptr [rcx+24]
+ xor r12, qword ptr [rcx+32]
+ xor r13, qword ptr [rcx+40]
+ xor r14, qword ptr [rcx+48]
+ xor r15, qword ptr [rcx+56]
+ xor eax, r9d ;# read address register 2
+ and eax, 262080
+ lea rcx, [rsi+rax]
+ cvtdq2pd xmm0, qword ptr [rcx+0]
+ cvtdq2pd xmm1, qword ptr [rcx+8]
+ cvtdq2pd xmm2, qword ptr [rcx+16]
+ cvtdq2pd xmm3, qword ptr [rcx+24]
+ cvtdq2pd xmm4, qword ptr [rcx+32]
+ cvtdq2pd xmm5, qword ptr [rcx+40]
+ cvtdq2pd xmm6, qword ptr [rcx+48]
+ cvtdq2pd xmm7, qword ptr [rcx+56]
+ andps xmm4, xmm14
+ andps xmm5, xmm14
+ andps xmm6, xmm14
+ andps xmm7, xmm14
+
+ ;# 256 instructions
include program.inc
-
-ALIGN 64
+
+ mov eax, r8d ;# read address register 1
+ xor eax, r9d ;# read address register 2
+ xor rbp, rax ;# modify "mx"
+ and rbp, -64 ;# align "mx" to the start of a cache line
+ mov edx, ebp ;# edx = mx
+ prefetchnta byte ptr [rdi+rdx]
+ ror rbp, 32 ;# swap "ma" and "mx"
+ mov edx, ebp ;# edx = ma
+ lea rcx, [rdi+rdx] ;# dataset cache line
+ xor r8, qword ptr [rcx+0]
+ xor r9, qword ptr [rcx+8]
+ xor r10, qword ptr [rcx+16]
+ xor r11, qword ptr [rcx+24]
+ xor r12, qword ptr [rcx+32]
+ xor r13, qword ptr [rcx+40]
+ xor r14, qword ptr [rcx+48]
+ xor r15, qword ptr [rcx+56]
+ mov eax, r12d ;# write address register 1
+ and eax, 262080
+ lea rcx, [rsi+rax]
+ mov qword ptr [rcx+0], r8
+ mov qword ptr [rcx+8], r9
+ mov qword ptr [rcx+16], r10
+ mov qword ptr [rcx+24], r11
+ mov qword ptr [rcx+32], r12
+ mov qword ptr [rcx+40], r13
+ mov qword ptr [rcx+48], r14
+ mov qword ptr [rcx+56], r15
+ xor eax, r13d ;# write address register 2
+ and eax, 262080
+ lea rcx, [rsi+rax]
+ mulpd xmm0, xmm4
+ mulpd xmm1, xmm5
+ mulpd xmm2, xmm6
+ mulpd xmm3, xmm7
+ movapd xmmword ptr [rcx+0], xmm0
+ movapd xmmword ptr [rcx+16], xmm1
+ movapd xmmword ptr [rcx+32], xmm2
+ movapd xmmword ptr [rcx+48], xmm3
+ dec ebx
+ jnz program_begin
+
rx_finish:
- ; unroll the stack
- mov rsp, rdi
-
; save VM register values
pop rcx
- pop rcx
mov qword ptr [rcx+0], r8
mov qword ptr [rcx+8], r9
mov qword ptr [rcx+16], r10
@@ -183,8 +198,8 @@ rx_finish:
mov qword ptr [rcx+40], r13
mov qword ptr [rcx+48], r14
mov qword ptr [rcx+56], r15
- movdqa xmmword ptr [rcx+64], xmm8
- movdqa xmmword ptr [rcx+80], xmm9
+ movdqa xmmword ptr [rcx+64], xmm0
+ movdqa xmmword ptr [rcx+80], xmm1
movdqa xmmword ptr [rcx+96], xmm2
movdqa xmmword ptr [rcx+112], xmm3
lea rcx, [rcx+64]
@@ -194,6 +209,12 @@ rx_finish:
movdqa xmmword ptr [rcx+112], xmm7
; load callee-saved registers
+ movdqu xmm15, xmmword ptr [rsp]
+ movdqu xmm14, xmmword ptr [rsp+16]
+ movdqu xmm13, xmmword ptr [rsp+32]
+ movdqu xmm12, xmmword ptr [rsp+48]
+ movdqu xmm11, xmmword ptr [rsp+64]
+ add rsp, 80
movdqu xmm10, xmmword ptr [rsp]
movdqu xmm9, xmmword ptr [rsp+16]
movdqu xmm8, xmmword ptr [rsp+32]
diff --git a/src/instructionWeights.hpp b/src/instructionWeights.hpp
index de027b7..242b5bd 100644
--- a/src/instructionWeights.hpp
+++ b/src/instructionWeights.hpp
@@ -19,46 +19,58 @@ along with RandomX. If not, see.
#pragma once
-#define WT_ADD_64 12
-#define WT_ADD_32 2
-#define WT_SUB_64 12
-#define WT_SUB_32 2
-#define WT_MUL_64 23
-#define WT_MULH_64 5
-#define WT_MUL_32 15
-#define WT_IMUL_32 15
-#define WT_IMULH_64 3
-#define WT_DIV_64 8
-#define WT_IDIV_64 8
-#define WT_AND_64 4
-#define WT_AND_32 2
-#define WT_OR_64 4
-#define WT_OR_32 2
-#define WT_XOR_64 4
-#define WT_XOR_32 2
-#define WT_SHL_64 3
-#define WT_SHR_64 3
-#define WT_SAR_64 3
-#define WT_ROL_64 6
-#define WT_ROR_64 6
-#define WT_FPADD 20
-#define WT_FPSUB 20
-#define WT_FPMUL 22
-#define WT_FPDIV 8
-#define WT_FPSQRT 6
-#define WT_FPROUND 2
-#define WT_JUMP 11
-#define WT_CALL 11
-#define WT_RET 12
+//Integer
+#define WT_IADD_R 10
+#define WT_IADD_M 3
+#define WT_IADD_RC 12
+#define WT_ISUB_R 10
+#define WT_ISUB_M 3
+#define WT_IMUL_9C 12
+#define WT_IMUL_R 24
+#define WT_IMUL_M 8
+#define WT_IMULH_R 6
+#define WT_IMULH_M 2
+#define WT_ISMULH_R 6
+#define WT_ISMULH_M 2
+#define WT_IDIV_C 4
+#define WT_ISDIV_C 2
+#define WT_INEG_R 4
+#define WT_IXOR_R 15
+#define WT_IXOR_M 5
+#define WT_IROR_R 10
+#define WT_IROL_R 10
+
+//Common floating point
+#define WT_FPSWAP_R 6
+
+//Floating point group F
+#define WT_FPADD_R 18
+#define WT_FPADD_M 3
+#define WT_FPSUB_R 18
+#define WT_FPSUB_M 3
+#define WT_FPNEG_R 5
+
+//Floating point group E
+#define WT_FPMUL_R 18
+#define WT_FPMUL_M 3
+#define WT_FPDIV_R 6
+#define WT_FPDIV_M 1
+#define WT_FPSQRT_R 6
+
+//Control
+#define WT_COND_R 15
+#define WT_COND_M 5
+#define WT_CFROUND 1
+
#define WT_NOP 0
-
-constexpr int wtSum = WT_ADD_64 + WT_ADD_32 + WT_SUB_64 + WT_SUB_32 + \
-WT_MUL_64 + WT_MULH_64 + WT_MUL_32 + WT_IMUL_32 + WT_IMULH_64 + \
-WT_DIV_64 + WT_IDIV_64 + WT_AND_64 + WT_AND_32 + WT_OR_64 + \
-WT_OR_32 + WT_XOR_64 + WT_XOR_32 + WT_SHL_64 + WT_SHR_64 + \
-WT_SAR_64 + WT_ROL_64 + WT_ROR_64 + WT_FPADD + WT_FPSUB + WT_FPMUL \
-+ WT_FPDIV + WT_FPSQRT + WT_FPROUND + WT_JUMP + WT_CALL + WT_RET + WT_NOP;
+constexpr int wtSum = WT_IADD_R + WT_IADD_M + WT_IADD_RC + WT_ISUB_R + \
+WT_ISUB_M + WT_IMUL_9C + WT_IMUL_R + WT_IMUL_M + WT_IMULH_R + \
+WT_IMULH_M + WT_ISMULH_R + WT_ISMULH_M + WT_IDIV_C + WT_ISDIV_C + \
+WT_INEG_R + WT_IXOR_R + WT_IXOR_M + WT_IROR_R + WT_IROL_R + \
+WT_FPSWAP_R + WT_FPADD_R + WT_FPADD_M + WT_FPSUB_R + WT_FPSUB_M + \
+WT_FPNEG_R + WT_FPMUL_R + WT_FPMUL_M + WT_FPDIV_R + WT_FPDIV_M + \
+WT_FPSQRT_R + WT_COND_R + WT_COND_M + WT_CFROUND + WT_NOP;
static_assert(wtSum == 256,
"Sum of instruction weights must be 256");
diff --git a/src/main.cpp b/src/main.cpp
index 84c76c8..0b09a74 100644
--- a/src/main.cpp
+++ b/src/main.cpp
@@ -34,6 +34,7 @@ along with RandomX. If not, see.
#include
#include "dataset.hpp"
#include "Cache.hpp"
+#include "Pcg32.hpp"
const uint8_t seed[32] = { 191, 182, 222, 175, 249, 89, 134, 104, 241, 68, 191, 62, 162, 166, 61, 64, 123, 191, 227, 193, 118, 60, 188, 53, 223, 133, 175, 24, 123, 230, 55, 74 };
@@ -130,6 +131,27 @@ void generateAsm(int nonce) {
asmX86.printCode(std::cout);
}
+void generateNative(int nonce) {
+ uint64_t hash[4];
+ unsigned char blockTemplate[] = {
+ 0x07, 0x07, 0xf7, 0xa4, 0xf0, 0xd6, 0x05, 0xb3, 0x03, 0x26, 0x08, 0x16, 0xba, 0x3f, 0x10, 0x90, 0x2e, 0x1a, 0x14,
+ 0x5a, 0xc5, 0xfa, 0xd3, 0xaa, 0x3a, 0xf6, 0xea, 0x44, 0xc1, 0x18, 0x69, 0xdc, 0x4f, 0x85, 0x3f, 0x00, 0x2b, 0x2e,
+ 0xea, 0x00, 0x00, 0x00, 0x00, 0x77, 0xb2, 0x06, 0xa0, 0x2c, 0xa5, 0xb1, 0xd4, 0xce, 0x6b, 0xbf, 0xdf, 0x0a, 0xca,
+ 0xc3, 0x8b, 0xde, 0xd3, 0x4d, 0x2d, 0xcd, 0xee, 0xf9, 0x5c, 0xd2, 0x0c, 0xef, 0xc1, 0x2f, 0x61, 0xd5, 0x61, 0x09
+ };
+ int* noncePtr = (int*)(blockTemplate + 39);
+ *noncePtr = nonce;
+ blake2b(hash, sizeof(hash), blockTemplate, sizeof(blockTemplate), nullptr, 0);
+ RandomX::Program prog;
+ Pcg32 gen(hash);
+ prog.initialize(gen);
+ for (int i = 0; i < RandomX::ProgramLength; ++i) {
+ prog(i).dst %= 8;
+ prog(i).src %= 8;
+ }
+ std::cout << prog << std::endl;
+}
+
void mine(RandomX::VirtualMachine* vm, std::atomic& atomicNonce, AtomicHash& result, int noncesCount, int thread, uint8_t* scratchpad) {
uint64_t hash[4];
unsigned char blockTemplate[] = {
@@ -147,18 +169,16 @@ void mine(RandomX::VirtualMachine* vm, std::atomic& atomicNonce, AtomicHash
blake2b(hash, sizeof(hash), blockTemplate, sizeof(blockTemplate), nullptr, 0);
int spIndex = ((uint8_t*)hash)[24] | ((((uint8_t*)hash)[25] & 15) << 8);
vm->initializeScratchpad(scratchpad, spIndex);
- vm->initializeProgram(hash);
+ //vm->initializeProgram(hash);
//dump((char*)((RandomX::CompiledVirtualMachine*)vm)->getProgram(), RandomX::CodeSize, "code-1337-jmp.txt");
- vm->setScratchpad(scratchpad + 3 * RandomX::ScratchpadSize / 4);
- vm->execute();
- vm->setScratchpad(scratchpad + 2 * RandomX::ScratchpadSize / 4);
- vm->execute();
- vm->getResult(nullptr, 0, hash);
- vm->initializeProgram(hash);
- vm->setScratchpad(scratchpad + 1 * RandomX::ScratchpadSize / 4);
- vm->execute();
- vm->setScratchpad(scratchpad + 0 * RandomX::ScratchpadSize / 4);
- vm->execute();
+ for (int chain = 0; chain < 16; ++chain) {
+ vm->initializeProgram(hash);
+ int segment = hash[3] & 3;
+ vm->setScratchpad(scratchpad);// +segment * RandomX::ScratchpadSize / 4);
+ vm->execute();
+ vm->getResult(nullptr, 0, hash);
+ }
+ //vm->initializeProgram(hash);
vm->getResult(scratchpad, RandomX::ScratchpadSize, hash);
result.xorWith(hash);
if (RandomX::trace) {
@@ -171,7 +191,7 @@ void mine(RandomX::VirtualMachine* vm, std::atomic& atomicNonce, AtomicHash
}
int main(int argc, char** argv) {
- bool softAes, lightClient, genAsm, compiled, help, largePages, async, aesBench;
+ bool softAes, lightClient, genAsm, compiled, help, largePages, async, aesBench, genNative;
int programCount, threadCount;
readOption("--help", argc, argv, help);
@@ -189,12 +209,18 @@ int main(int argc, char** argv) {
readOption("--largePages", argc, argv, largePages);
readOption("--async", argc, argv, async);
readOption("--aesBench", argc, argv, aesBench);
+ readOption("--genNative", argc, argv, genNative);
if (genAsm) {
generateAsm(programCount);
return 0;
}
+ if (genNative) {
+ generateNative(programCount);
+ return 0;
+ }
+
if (softAes)
std::cout << "Using software AES." << std::endl;
diff --git a/src/program.inc b/src/program.inc
index f06ca58..a91240e 100644
--- a/src/program.inc
+++ b/src/program.inc
@@ -1,8923 +1,745 @@
-rx_i_0: ;CALL
- dec ebx
- jz rx_finish
- xor r9, 0ca9788ah
- mov eax, r9d
- test bl, 63
- jnz short rx_body_0
- call rx_read
-rx_body_0:
- xor rbp, rax
- and eax, 32767
- mov rax, qword ptr [rsi+rax*8]
- mov r12, rax
- cmp r11d, 445530481
- ja short rx_i_1
- call rx_i_30
-
-rx_i_1: ;IDIV_64
- dec ebx
- jz rx_finish
- xor r15, 06afc2fa4h
- mov eax, r15d
- test bl, 63
- jnz short rx_body_1
- call rx_read
-rx_body_1:
- and eax, 131071
- mov rax, qword ptr [rsi+rax*8]
- mov edx, r10d
- cmp edx, -1
- jne short body_idiv_1
- neg rax
- jmp short result_idiv_1
-body_idiv_1:
- mov ecx, 1
- test edx, edx
- cmovne ecx, edx
- movsxd rcx, ecx
- cqo
- idiv rcx
-result_idiv_1:
- mov r12, rax
-
-rx_i_2: ;JUMP
- dec ebx
- jz rx_finish
- xor r15, 097210f7bh
- mov eax, r15d
- test bl, 63
- jnz short rx_body_2
- call rx_read
-rx_body_2:
- xor rbp, rax
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- mov rcx, rax
- mov eax, r9d
- xor eax, 05060ccf7h
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
- cmp r11d, 1348521207
- jno rx_i_47
-
-rx_i_3: ;FPDIV
- dec ebx
- jz rx_finish
- xor r13, 082c73195h
+ ; ISUB_R r0, r4
+ sub r8, r12
+ ; IROR_R r5, 15
+ ror r13, 15
+ ; ISUB_M r6, L1[r5]
mov eax, r13d
- test bl, 63
- jnz short rx_body_3
- call rx_read
-rx_body_3:
- xor rbp, rax
- and eax, 32767
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- divpd xmm0, xmm9
- movaps xmm1, xmm0
- cmpeqpd xmm1, xmm1
- andps xmm0, xmm1
- movaps xmm8, xmm0
-
-rx_i_4: ;MUL_32
- dec ebx
- jz rx_finish
- xor r14, 077daefb4h
+ and eax, 16376
+ sub r14, qword ptr [rsi+rax]
+ ; IMUL_R r7, r6
+ imul r15, r14
+ ; FPADD_R f3, a1
+ addpd xmm3, xmm9
+ ; FPMUL_R e1, a3
+ mulpd xmm5, xmm11
+ ; IMUL_R r2, r4
+ imul r10, r12
+ ; IADD_RC r4, r5, 1789610138
+ lea r12, [r12+r13+1789610138]
+ ; IADD_R r1, r4
+ add r9, r12
+ ; IADD_R r6, r0
+ add r14, r8
+ ; IXOR_R r7, r2
+ xor r15, r10
+ ; ISMULH_M r6, L1[6816]
+ mov rax, r14
+ imul qword ptr [rsi+6816]
+ mov r14, rdx
+ ; ISUB_R r0, r4
+ sub r8, r12
+ ; IXOR_R r7, r2
+ xor r15, r10
+ ; INEG_R r4
+ neg r12
+ ; IROL_R r3, r0
+ mov ecx, r8d
+ rol r11, cl
+ ; IADD_RC r2, r5, -1667142135
+ lea r10, [r10+r13-1667142135]
+ ; ISUB_R r6, r2
+ sub r14, r10
+ ; IDIV_C r3, 2650709570
+ mov rax, 3736177069856446853
+ mul r11
+ shr rdx, 29
+ add r11, rdx
+ ; IMULH_R r3, r0
+ mov rax, r11
+ mul r8
+ mov r11, rdx
+ ; FPSUB_R f0, a2
+ subpd xmm0, xmm10
+ ; FPADD_M f3, L2[r4]
+ mov eax, r12d
+ and eax, 262136
+ cvtdq2pd xmm12, qword ptr [rsi+rax]
+ addpd xmm3, xmm12
+ ; FPMUL_M e1, L1[r5]
+ mov eax, r13d
+ and eax, 16376
+ cvtdq2pd xmm12, qword ptr [rsi+rax]
+ mulpd xmm5, xmm12
+ maxpd xmm5, xmm13
+ ; IMUL_9C r7, -778247271
+ lea r15, [r15+r15*8-778247271]
+ ; IXOR_R r4, 1846379510
+ xor r12, 1846379510
+ ; COND_M r6, of(L1[r1], -397786451)
+ xor ecx, ecx
+ mov eax, r9d
+ and eax, 16376
+ cmp dword ptr [rsi+rax], -397786451
+ seto cl
+ add r14, rcx
+ ; COND_R r6, of(r3, -1033710571)
+ xor ecx, ecx
+ cmp r11d, -1033710571
+ seto cl
+ add r14, rcx
+ ; COND_M r6, sg(L1[r6], 1413230028)
+ xor ecx, ecx
mov eax, r14d
- test bl, 63
- jnz short rx_body_4
- call rx_read
-rx_body_4:
- and eax, 32767
- mov rax, qword ptr [rsi+rax*8]
- mov ecx, eax
+ and eax, 16376
+ cmp dword ptr [rsi+rax], 1413230028
+ sets cl
+ add r14, rcx
+ ; IDIV_C r0, 2791108943
+ mov rax, 1774119268816201525
+ mul r8
+ shr rdx, 28
+ add r8, rdx
+ ; FPSUB_M f1, L1[r6]
mov eax, r14d
- imul rax, rcx
- mov r9, rax
-
-rx_i_5: ;IMUL_32
- dec ebx
- jz rx_finish
- xor r15, 0379f9ee0h
- mov eax, r15d
- test bl, 63
- jnz short rx_body_5
- call rx_read
-rx_body_5:
- xor rbp, rax
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- movsxd rcx, eax
- mov rax, 1037420699
- imul rax, rcx
- mov rcx, rax
- mov eax, r12d
- xor eax, 03dd5c89bh
- and eax, 32767
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_6: ;MUL_64
- dec ebx
- jz rx_finish
- xor r8, 03bae7272h
- mov eax, r8d
- test bl, 63
- jnz short rx_body_6
- call rx_read
-rx_body_6:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- imul rax, r15
- mov rcx, rax
- mov eax, r9d
- xor eax, 098a649d1h
- and eax, 131071
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_7: ;FPADD
- dec ebx
- jz rx_finish
- xor r10, 0e264ed81h
+ and eax, 16376
+ cvtdq2pd xmm12, qword ptr [rsi+rax]
+ subpd xmm1, xmm12
+ ; FPSWAP_R f0
+ shufpd xmm0, xmm0, 1
+ ; IADD_RC r6, r5, -640194892
+ lea r14, [r14+r13-640194892]
+ ; FPADD_M f0, L1[r2]
mov eax, r10d
- test bl, 63
- jnz short rx_body_7
- call rx_read
-rx_body_7:
- and eax, 32767
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- addpd xmm0, xmm6
- movaps xmm6, xmm0
-
-rx_i_8: ;XOR_64
- dec ebx
- jz rx_finish
- xor r13, 068c1e5d2h
- mov eax, r13d
- test bl, 63
- jnz short rx_body_8
- call rx_read
-rx_body_8:
- and eax, 131071
- mov rax, qword ptr [rsi+rax*8]
- xor rax, r11
- mov rcx, rax
- mov eax, r12d
- xor eax, 050267ebdh
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_9: ;IDIV_64
- dec ebx
- jz rx_finish
- xor r14, 085121c54h
- mov eax, r14d
- test bl, 63
- jnz short rx_body_9
- call rx_read
-rx_body_9:
- xor rbp, rax
- and eax, 32767
- mov rax, qword ptr [rsi+rax*8]
- ; magic divide by 565870810
- mov rdx, 8750690209911200579
- imul rdx
- mov rax, rdx
- xor edx, edx
- sar rax, 28
- sets dl
- add rax, rdx
- mov r10, rax
-
-rx_i_10: ;AND_64
- dec ebx
- jz rx_finish
- xor r8, 052efde3eh
- mov eax, r8d
- test bl, 63
- jnz short rx_body_10
- call rx_read
-rx_body_10:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- and rax, r10
- mov rcx, rax
- mov eax, r13d
- xor eax, 0d49dbd9fh
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_11: ;FPADD
- dec ebx
- jz rx_finish
- xor r10, 0a9bf8aa1h
- mov eax, r10d
- test bl, 63
- jnz short rx_body_11
- call rx_read
-rx_body_11:
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- addpd xmm0, xmm5
- movaps xmm4, xmm0
- mov eax, r12d
- xor eax, 0852d40d8h
- and eax, 2047
- movlpd qword ptr [rsi + rax * 8], xmm4
-
-rx_i_12: ;FPSQRT
- dec ebx
- jz rx_finish
- xor r10, 0db2691ch
- mov eax, r10d
- test bl, 63
- jnz short rx_body_12
- call rx_read
-rx_body_12:
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- andps xmm0, xmm10
- sqrtpd xmm8, xmm0
-
-rx_i_13: ;FPADD
- dec ebx
- jz rx_finish
- xor r12, 061c0d34dh
- mov eax, r12d
- test bl, 63
- jnz short rx_body_13
- call rx_read
-rx_body_13:
- xor rbp, rax
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- addpd xmm0, xmm3
- movaps xmm9, xmm0
- mov eax, r9d
- xor eax, 04f2f223ch
- and eax, 2047
- movhpd qword ptr [rsi + rax * 8], xmm9
-
-rx_i_14: ;XOR_64
- dec ebx
- jz rx_finish
- xor r10, 0e761d1beh
- mov eax, r10d
- test bl, 63
- jnz short rx_body_14
- call rx_read
-rx_body_14:
- xor rbp, rax
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- xor rax, r9
- mov r10, rax
-
-rx_i_15: ;RET
- dec ebx
- jz rx_finish
- xor r11, 074ddb688h
- mov eax, r11d
- test bl, 63
- jnz short rx_body_15
- call rx_read
-rx_body_15:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- mov r14, rax
- cmp rsp, rdi
- je short rx_i_16
- ret
-
-rx_i_16: ;ADD_64
- dec ebx
- jz rx_finish
- xor r14, 06be90627h
- mov eax, r14d
- test bl, 63
- jnz short rx_body_16
- call rx_read
-rx_body_16:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- add rax, r10
- mov rcx, rax
- mov eax, r9d
- xor eax, 0d7e75aeh
- and eax, 131071
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_17: ;FPMUL
- dec ebx
- jz rx_finish
- xor r11, 0fbc6fc35h
- mov eax, r11d
- test bl, 63
- jnz short rx_body_17
- call rx_read
-rx_body_17:
- and eax, 32767
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- mulpd xmm0, xmm4
- movaps xmm1, xmm0
- cmpeqpd xmm1, xmm1
- andps xmm0, xmm1
- movaps xmm4, xmm0
-
-rx_i_18: ;FPSUB
- dec ebx
- jz rx_finish
- xor r14, 0c28ca080h
- mov eax, r14d
- test bl, 63
- jnz short rx_body_18
- call rx_read
-rx_body_18:
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- subpd xmm0, xmm4
- movaps xmm3, xmm0
- mov eax, r11d
- xor eax, 0869baa81h
- and eax, 2047
- movlpd qword ptr [rsi + rax * 8], xmm3
-
-rx_i_19: ;FPSUB
- dec ebx
- jz rx_finish
- xor r13, 0ac009c30h
- mov eax, r13d
- test bl, 63
- jnz short rx_body_19
- call rx_read
-rx_body_19:
- xor rbp, rax
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- subpd xmm0, xmm8
- movaps xmm7, xmm0
-
-rx_i_20: ;FPSUB
- dec ebx
- jz rx_finish
- xor r13, 0ecca967dh
- mov eax, r13d
- test bl, 63
- jnz short rx_body_20
- call rx_read
-rx_body_20:
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- subpd xmm0, xmm2
- movaps xmm7, xmm0
-
-rx_i_21: ;ROR_64
- dec ebx
- jz rx_finish
- xor r8, 0977f0284h
- mov eax, r8d
- test bl, 63
- jnz short rx_body_21
- call rx_read
-rx_body_21:
- xor rbp, rax
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- mov rcx, r9
- ror rax, cl
- mov r15, rax
-
-rx_i_22: ;ADD_64
- dec ebx
- jz rx_finish
- xor r13, 080bdfefah
- mov eax, r13d
- test bl, 63
- jnz short rx_body_22
- call rx_read
-rx_body_22:
- xor rbp, rax
- and eax, 131071
- mov rax, qword ptr [rsi+rax*8]
- add rax, r8
- mov r10, rax
-
-rx_i_23: ;MUL_64
- dec ebx
- jz rx_finish
- xor r15, 0e1e0d3c4h
- mov eax, r15d
- test bl, 63
- jnz short rx_body_23
- call rx_read
-rx_body_23:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- imul rax, rax, 1283724485
- mov rcx, rax
- mov eax, r8d
- xor eax, 04c8414c5h
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_24: ;DIV_64
- dec ebx
- jz rx_finish
- xor r8, 070d3b8c7h
- mov eax, r8d
- test bl, 63
- jnz short rx_body_24
- call rx_read
-rx_body_24:
- xor rbp, rax
- and eax, 131071
- mov rax, qword ptr [rsi+rax*8]
- mov ecx, 1
- mov edx, r15d
- test edx, edx
- cmovne ecx, edx
- xor edx, edx
- div rcx
- mov rcx, rax
- mov eax, r15d
- xor eax, 099b77a68h
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_25: ;FPMUL
- dec ebx
- jz rx_finish
- xor r12, 01cf77a04h
- mov eax, r12d
- test bl, 63
- jnz short rx_body_25
- call rx_read
-rx_body_25:
- xor rbp, rax
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- mulpd xmm0, xmm9
- movaps xmm1, xmm0
- cmpeqpd xmm1, xmm1
- andps xmm0, xmm1
- movaps xmm6, xmm0
- mov eax, r14d
- xor eax, 0baf5c2d4h
- and eax, 2047
- movlpd qword ptr [rsi + rax * 8], xmm6
-
-rx_i_26: ;IMULH_64
- dec ebx
- jz rx_finish
- xor r11, 0e311468ch
- mov eax, r11d
- test bl, 63
- jnz short rx_body_26
- call rx_read
-rx_body_26:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- mov rcx, 812644844
- imul rcx
- mov rax, rdx
- mov r9, rax
-
-rx_i_27: ;FPMUL
- dec ebx
- jz rx_finish
- xor r12, 01fd9911ah
- mov eax, r12d
- test bl, 63
- jnz short rx_body_27
- call rx_read
-rx_body_27:
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- mulpd xmm0, xmm3
- movaps xmm1, xmm0
- cmpeqpd xmm1, xmm1
- andps xmm0, xmm1
- movaps xmm6, xmm0
-
-rx_i_28: ;AND_32
- dec ebx
- jz rx_finish
- xor r13, 067df757eh
- mov eax, r13d
- test bl, 63
- jnz short rx_body_28
- call rx_read
-rx_body_28:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- and eax, 565865719
- mov rcx, rax
- mov eax, r14d
- xor eax, 021ba6cf7h
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_29: ;SUB_64
- dec ebx
- jz rx_finish
- xor r12, 0be2e7c42h
- mov eax, r12d
- test bl, 63
- jnz short rx_body_29
- call rx_read
-rx_body_29:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- sub rax, r13
- mov rcx, rax
- mov eax, r14d
- xor eax, 073e1a073h
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_30: ;FPADD
- dec ebx
- jz rx_finish
- xor r11, 084d067f7h
- mov eax, r11d
- test bl, 63
- jnz short rx_body_30
- call rx_read
-rx_body_30:
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- addpd xmm0, xmm3
- movaps xmm7, xmm0
-
-rx_i_31: ;ROR_64
- dec ebx
- jz rx_finish
- xor r14, 0d352ce37h
- mov eax, r14d
- test bl, 63
- jnz short rx_body_31
- call rx_read
-rx_body_31:
- xor rbp, rax
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- ror rax, 55
- mov rcx, rax
- mov eax, r14d
- xor eax, 01e2da792h
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_32: ;AND_32
- dec ebx
- jz rx_finish
- xor r12, 0a1f248dah
- mov eax, r12d
- test bl, 63
- jnz short rx_body_32
- call rx_read
-rx_body_32:
- and eax, 131071
- mov rax, qword ptr [rsi+rax*8]
- and eax, r14d
- mov r9, rax
-
-rx_i_33: ;MUL_64
- dec ebx
- jz rx_finish
- xor r9, 0554720fch
- mov eax, r9d
- test bl, 63
- jnz short rx_body_33
- call rx_read
-rx_body_33:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- imul rax, r15
- mov r12, rax
-
-rx_i_34: ;CALL
- dec ebx
- jz rx_finish
- xor r13, 0665e91f1h
- mov eax, r13d
- test bl, 63
- jnz short rx_body_34
- call rx_read
-rx_body_34:
- xor rbp, rax
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- mov rcx, rax
- mov eax, r15d
- xor eax, 0e9563b32h
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
- cmp r14d, -380224718
- jns short rx_i_35
- call rx_i_108
-
-rx_i_35: ;CALL
- dec ebx
- jz rx_finish
- xor r15, 05ef1be79h
- mov eax, r15d
- test bl, 63
- jnz short rx_body_35
- call rx_read
-rx_body_35:
- and eax, 32767
- mov rax, qword ptr [rsi+rax*8]
- mov rcx, rax
- mov eax, r8d
- xor eax, 0865c0f66h
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
- cmp r9d, -2040787098
- jns short rx_i_36
- call rx_i_58
-
-rx_i_36: ;FPMUL
- dec ebx
- jz rx_finish
- xor r8, 012ec7e3ah
- mov eax, r8d
- test bl, 63
- jnz short rx_body_36
- call rx_read
-rx_body_36:
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- mulpd xmm0, xmm6
- movaps xmm1, xmm0
- cmpeqpd xmm1, xmm1
- andps xmm0, xmm1
- movaps xmm7, xmm0
-
-rx_i_37: ;FPSUB
- dec ebx
- jz rx_finish
- xor r12, 0d0706601h
- mov eax, r12d
- test bl, 63
- jnz short rx_body_37
- call rx_read
-rx_body_37:
- and eax, 32767
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- subpd xmm0, xmm2
- movaps xmm9, xmm0
-
-rx_i_38: ;SUB_64
- dec ebx
- jz rx_finish
- xor r9, 064056913h
- mov eax, r9d
- test bl, 63
- jnz short rx_body_38
- call rx_read
-rx_body_38:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- sub rax, r14
- mov rcx, rax
- mov eax, r10d
- xor eax, 087c32de2h
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_39: ;ADD_64
- dec ebx
- jz rx_finish
- xor r14, 02c1f1eb0h
- mov eax, r14d
- test bl, 63
- jnz short rx_body_39
- call rx_read
-rx_body_39:
- xor rbp, rax
- and eax, 131071
- mov rax, qword ptr [rsi+rax*8]
- add rax, r14
- mov rcx, rax
- mov eax, r14d
- xor eax, 0f4101ad9h
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_40: ;CALL
- dec ebx
- jz rx_finish
- xor r10, 068fd9009h
- mov eax, r10d
- test bl, 63
- jnz short rx_body_40
- call rx_read
-rx_body_40:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- mov rcx, rax
- mov eax, r9d
- xor eax, 0b2a27eceh
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
- cmp r12d, -1297973554
- jns short rx_i_41
- call rx_i_90
-
-rx_i_41: ;JUMP
- dec ebx
- jz rx_finish
- xor r9, 037a30933h
- mov eax, r9d
- test bl, 63
- jnz short rx_body_41
- call rx_read
-rx_body_41:
- and eax, 32767
- mov rax, qword ptr [rsi+rax*8]
- mov r9, rax
- cmp r14d, -1070581824
- jo rx_i_127
-
-rx_i_42: ;FPADD
- dec ebx
- jz rx_finish
- xor r15, 0bc1de9f6h
- mov eax, r15d
- test bl, 63
- jnz short rx_body_42
- call rx_read
-rx_body_42:
- xor rbp, rax
- and eax, 32767
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- addpd xmm0, xmm6
- movaps xmm6, xmm0
-
-rx_i_43: ;SUB_64
- dec ebx
- jz rx_finish
- xor r12, 02b2a2eech
- mov eax, r12d
- test bl, 63
- jnz short rx_body_43
- call rx_read
-rx_body_43:
- and eax, 32767
- mov rax, qword ptr [rsi+rax*8]
- sub rax, r8
- mov r11, rax
-
-rx_i_44: ;SAR_64
- dec ebx
- jz rx_finish
- xor r11, 0685817abh
- mov eax, r11d
- test bl, 63
- jnz short rx_body_44
- call rx_read
-rx_body_44:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- mov rcx, r9
- sar rax, cl
- mov rcx, rax
- mov eax, r15d
- xor eax, 0372116f6h
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_45: ;FPSUB
- dec ebx
- jz rx_finish
- xor r12, 08cd244ebh
- mov eax, r12d
- test bl, 63
- jnz short rx_body_45
- call rx_read
-rx_body_45:
- xor rbp, rax
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- subpd xmm0, xmm2
- movaps xmm5, xmm0
- mov eax, r13d
- xor eax, 0977132cdh
- and eax, 2047
- movhpd qword ptr [rsi + rax * 8], xmm5
-
-rx_i_46: ;ADD_64
- dec ebx
- jz rx_finish
- xor r8, 06d8f4254h
- mov eax, r8d
- test bl, 63
- jnz short rx_body_46
- call rx_read
-rx_body_46:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- add rax, r9
- mov r8, rax
-
-rx_i_47: ;JUMP
- dec ebx
- jz rx_finish
- xor r12, 05ba232c6h
- mov eax, r12d
- test bl, 63
- jnz short rx_body_47
- call rx_read
-rx_body_47:
- xor rbp, rax
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- mov rcx, rax
- mov eax, r13d
- xor eax, 071ba231h
- and eax, 32767
- mov qword ptr [rsi + rax * 8], rcx
- cmp r10d, 119251505
- jbe rx_i_131
-
-rx_i_48: ;FPDIV
- dec ebx
- jz rx_finish
- xor r8, 0aaed618fh
- mov eax, r8d
- test bl, 63
- jnz short rx_body_48
- call rx_read
-rx_body_48:
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- divpd xmm0, xmm3
- movaps xmm1, xmm0
- cmpeqpd xmm1, xmm1
- andps xmm0, xmm1
- movaps xmm9, xmm0
-
-rx_i_49: ;FPSUB
- dec ebx
- jz rx_finish
- xor r8, 0f96c6a45h
- mov eax, r8d
- test bl, 63
- jnz short rx_body_49
- call rx_read
-rx_body_49:
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- subpd xmm0, xmm3
- movaps xmm5, xmm0
-
-rx_i_50: ;AND_64
- dec ebx
- jz rx_finish
- xor r9, 0da3e4842h
- mov eax, r9d
- test bl, 63
- jnz short rx_body_50
- call rx_read
-rx_body_50:
- xor rbp, rax
- and eax, 32767
- mov rax, qword ptr [rsi+rax*8]
- and rax, r10
- mov r15, rax
-
-rx_i_51: ;SUB_64
- dec ebx
- jz rx_finish
- xor r10, 0302b676ah
- mov eax, r10d
- test bl, 63
- jnz short rx_body_51
- call rx_read
-rx_body_51:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- sub rax, r15
- mov rcx, rax
- mov eax, r15d
- xor eax, 018fd1fbfh
- and eax, 32767
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_52: ;FPSQRT
- dec ebx
- jz rx_finish
- xor r11, 0fa88f48bh
- mov eax, r11d
- test bl, 63
- jnz short rx_body_52
- call rx_read
-rx_body_52:
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- andps xmm0, xmm10
- sqrtpd xmm7, xmm0
-
-rx_i_53: ;RET
- dec ebx
- jz rx_finish
- xor r13, 03dff9b9eh
- mov eax, r13d
- test bl, 63
- jnz short rx_body_53
- call rx_read
-rx_body_53:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- mov rcx, rax
- mov eax, r13d
- xor eax, 078ed00edh
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
- cmp rsp, rdi
- je short rx_i_54
- ret
-
-rx_i_54: ;DIV_64
- dec ebx
- jz rx_finish
- xor r11, 060638de0h
- mov eax, r11d
- test bl, 63
- jnz short rx_body_54
- call rx_read
-rx_body_54:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- ; magic divide by 282209221
- mov rcx, 1096650948274100047
- mul rcx
- mov rax, rdx
- shr rax, 24
- mov rcx, rax
- mov eax, r12d
- xor eax, 010d22bc5h
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_55: ;FPMUL
- dec ebx
- jz rx_finish
- xor r10, 0dda983d4h
- mov eax, r10d
- test bl, 63
- jnz short rx_body_55
- call rx_read
-rx_body_55:
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- mulpd xmm0, xmm5
- movaps xmm1, xmm0
- cmpeqpd xmm1, xmm1
- andps xmm0, xmm1
- movaps xmm3, xmm0
-
-rx_i_56: ;IDIV_64
- dec ebx
- jz rx_finish
- xor r14, 0f1456b8eh
- mov eax, r14d
- test bl, 63
- jnz short rx_body_56
- call rx_read
-rx_body_56:
- and eax, 32767
- mov rax, qword ptr [rsi+rax*8]
- ; magic divide by -50768751
- mov rcx, rax
- mov rdx, 6254795139557318139
- imul rdx
- mov rax, rdx
- xor edx, edx
- sub rax, rcx
- sar rax, 25
- sets dl
- add rax, rdx
- mov r8, rax
-
-rx_i_57: ;MUL_64
- dec ebx
- jz rx_finish
- xor r9, 010dc4571h
- mov eax, r9d
- test bl, 63
- jnz short rx_body_57
- call rx_read
-rx_body_57:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- imul rax, rax, 172123015
- mov rcx, rax
- mov eax, r15d
- xor eax, 0a426387h
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_58: ;DIV_64
- dec ebx
- jz rx_finish
- xor r14, 0bcec0ebah
- mov eax, r14d
- test bl, 63
- jnz short rx_body_58
- call rx_read
-rx_body_58:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- ; magic divide by 1506547423
- mov rcx, 6573653217342526495
- mul rcx
- mov rax, rdx
- shr rax, 29
- mov r8, rax
-
-rx_i_59: ;FPSUB
- dec ebx
- jz rx_finish
- xor r11, 0980dd402h
- mov eax, r11d
- test bl, 63
- jnz short rx_body_59
- call rx_read
-rx_body_59:
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- subpd xmm0, xmm8
- movaps xmm7, xmm0
-
-rx_i_60: ;CALL
- dec ebx
- jz rx_finish
- xor r15, 03de14d1eh
- mov eax, r15d
- test bl, 63
- jnz short rx_body_60
- call rx_read
-rx_body_60:
- and eax, 131071
- mov rax, qword ptr [rsi+rax*8]
- mov rcx, rax
- mov eax, r11d
- xor eax, 07bb60f45h
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
- cmp r11d, 2075529029
- jo short rx_i_61
- call rx_i_116
-
-rx_i_61: ;JUMP
- dec ebx
- jz rx_finish
- xor r13, 05058ce64h
- mov eax, r13d
- test bl, 63
- jnz short rx_body_61
- call rx_read
-rx_body_61:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- mov r11, rax
- cmp r15d, 1933164545
- jns rx_i_120
-
-rx_i_62: ;FPSUB
- dec ebx
- jz rx_finish
- xor r15, 0c3089414h
- mov eax, r15d
- test bl, 63
- jnz short rx_body_62
- call rx_read
-rx_body_62:
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- subpd xmm0, xmm8
- movaps xmm2, xmm0
- mov eax, r10d
- xor eax, 05c4789e3h
- and eax, 2047
- movhpd qword ptr [rsi + rax * 8], xmm2
-
-rx_i_63: ;FPSUB
- dec ebx
- jz rx_finish
- xor r9, 065cf272eh
- mov eax, r9d
- test bl, 63
- jnz short rx_body_63
- call rx_read
-rx_body_63:
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- subpd xmm0, xmm7
- movaps xmm8, xmm0
- mov eax, r8d
- xor eax, 0be13d69eh
- and eax, 2047
- movlpd qword ptr [rsi + rax * 8], xmm8
-
-rx_i_64: ;SUB_64
- dec ebx
- jz rx_finish
- xor r13, 0ae54dfbfh
- mov eax, r13d
- test bl, 63
- jnz short rx_body_64
- call rx_read
-rx_body_64:
- and eax, 32767
- mov rax, qword ptr [rsi+rax*8]
- sub rax, r15
- mov r9, rax
-
-rx_i_65: ;JUMP
- dec ebx
- jz rx_finish
- xor r13, 07b366ce6h
- mov eax, r13d
- test bl, 63
- jnz short rx_body_65
- call rx_read
-rx_body_65:
- and eax, 32767
- mov rax, qword ptr [rsi+rax*8]
- mov rcx, rax
- mov eax, r11d
- xor eax, 0594a879fh
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
- cmp r8d, 1498056607
- js rx_i_129
-
-rx_i_66: ;FPDIV
- dec ebx
- jz rx_finish
- xor r15, 015a1b689h
- mov eax, r15d
- test bl, 63
- jnz short rx_body_66
- call rx_read
-rx_body_66:
- xor rbp, rax
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- divpd xmm0, xmm3
- movaps xmm1, xmm0
- cmpeqpd xmm1, xmm1
- andps xmm0, xmm1
- movaps xmm9, xmm0
- mov eax, r9d
- xor eax, 07305e78h
- and eax, 2047
- movlpd qword ptr [rsi + rax * 8], xmm9
-
-rx_i_67: ;JUMP
- dec ebx
- jz rx_finish
- xor r14, 088393ba0h
- mov eax, r14d
- test bl, 63
- jnz short rx_body_67
- call rx_read
-rx_body_67:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- mov rcx, rax
- mov eax, r9d
- xor eax, 07916db59h
- and eax, 32767
- mov qword ptr [rsi + rax * 8], rcx
- cmp r13d, 2031541081
- jns rx_i_79
-
-rx_i_68: ;FPADD
- dec ebx
- jz rx_finish
- xor r13, 03aa5c3a4h
- mov eax, r13d
- test bl, 63
- jnz short rx_body_68
- call rx_read
-rx_body_68:
- and eax, 32767
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- addpd xmm0, xmm2
- movaps xmm4, xmm0
- mov eax, r12d
- xor eax, 03c51ef39h
- and eax, 2047
- movhpd qword ptr [rsi + rax * 8], xmm4
-
-rx_i_69: ;FPADD
- dec ebx
- jz rx_finish
- xor r15, 0376c9c27h
- mov eax, r15d
- test bl, 63
- jnz short rx_body_69
- call rx_read
-rx_body_69:
- xor rbp, rax
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- addpd xmm0, xmm5
- movaps xmm8, xmm0
-
-rx_i_70: ;MULH_64
- dec ebx
- jz rx_finish
- xor r8, 0bbbec3fah
- mov eax, r8d
- test bl, 63
- jnz short rx_body_70
- call rx_read
-rx_body_70:
- xor rbp, rax
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- mov rcx, r9
- mul rcx
- mov rax, rdx
- mov r13, rax
-
-rx_i_71: ;FPMUL
- dec ebx
- jz rx_finish
- xor r14, 0e9efb350h
- mov eax, r14d
- test bl, 63
- jnz short rx_body_71
- call rx_read
-rx_body_71:
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- mulpd xmm0, xmm5
- movaps xmm1, xmm0
- cmpeqpd xmm1, xmm1
- andps xmm0, xmm1
- movaps xmm7, xmm0
- mov eax, r15d
- xor eax, 056660eedh
- and eax, 131071
- movlpd qword ptr [rsi + rax * 8], xmm7
-
-rx_i_72: ;JUMP
- dec ebx
- jz rx_finish
- xor r13, 0f4e51e28h
- mov eax, r13d
- test bl, 63
- jnz short rx_body_72
- call rx_read
-rx_body_72:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- mov r11, rax
- cmp r9d, -631091751
- jno rx_i_191
-
-rx_i_73: ;FPDIV
- dec ebx
- jz rx_finish
- xor r12, 0c24ddbd4h
- mov eax, r12d
- test bl, 63
- jnz short rx_body_73
- call rx_read
-rx_body_73:
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- divpd xmm0, xmm3
- movaps xmm1, xmm0
- cmpeqpd xmm1, xmm1
- andps xmm0, xmm1
- movaps xmm2, xmm0
-
-rx_i_74: ;MUL_64
- dec ebx
- jz rx_finish
- xor r8, 04c4b0c7fh
- mov eax, r8d
- test bl, 63
- jnz short rx_body_74
- call rx_read
-rx_body_74:
- xor rbp, rax
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- imul rax, r13
- mov rcx, rax
- mov eax, r9d
- xor eax, 0aaaacb32h
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_75: ;CALL
- dec ebx
- jz rx_finish
- xor r14, 03bcc02e3h
- mov eax, r14d
- test bl, 63
- jnz short rx_body_75
- call rx_read
-rx_body_75:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- mov r13, rax
- cmp r11d, -1160798683
- jno short rx_i_76
- call rx_i_202
-
-rx_i_76: ;FPADD
- dec ebx
- jz rx_finish
- xor r11, 04b0ff63eh
- mov eax, r11d
- test bl, 63
- jnz short rx_body_76
- call rx_read
-rx_body_76:
- and eax, 32767
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- addpd xmm0, xmm3
- movaps xmm7, xmm0
- mov eax, r15d
- xor eax, 083bc0396h
- and eax, 32767
- movlpd qword ptr [rsi + rax * 8], xmm7
-
-rx_i_77: ;RET
- dec ebx
- jz rx_finish
- xor r14, 0b956b3e8h
- mov eax, r14d
- test bl, 63
- jnz short rx_body_77
- call rx_read
-rx_body_77:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- mov rcx, rax
- mov eax, r11d
- xor eax, 03a92bc7ah
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
- cmp rsp, rdi
- je short rx_i_78
- ret
-
-rx_i_78: ;MUL_32
- dec ebx
- jz rx_finish
- xor r9, 0edeca680h
- mov eax, r9d
- test bl, 63
- jnz short rx_body_78
- call rx_read
-rx_body_78:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- mov ecx, eax
- mov eax, r8d
- imul rax, rcx
- mov rcx, rax
- mov eax, r15d
- xor eax, 0697e6195h
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_79: ;CALL
- dec ebx
- jz rx_finish
- xor r11, 0fbdddcb5h
- mov eax, r11d
- test bl, 63
- jnz short rx_body_79
- call rx_read
-rx_body_79:
- and eax, 32767
- mov rax, qword ptr [rsi+rax*8]
- mov r11, rax
- cmp r13d, 1800043331
- ja short rx_i_80
- call rx_i_93
-
-rx_i_80: ;ROR_64
- dec ebx
- jz rx_finish
- xor r13, 09cec97a1h
- mov eax, r13d
- test bl, 63
- jnz short rx_body_80
- call rx_read
-rx_body_80:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- mov rcx, r11
- ror rax, cl
- mov rcx, rax
- mov eax, r11d
- xor eax, 01a681d13h
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_81: ;AND_64
- dec ebx
- jz rx_finish
- xor r15, 078228167h
- mov eax, r15d
- test bl, 63
- jnz short rx_body_81
- call rx_read
-rx_body_81:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- and rax, 338325607
- mov rcx, rax
- mov eax, r8d
- xor eax, 0142a7067h
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_82: ;JUMP
- dec ebx
- jz rx_finish
- xor r11, 078cae1ffh
- mov eax, r11d
- test bl, 63
- jnz short rx_body_82
- call rx_read
-rx_body_82:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- mov r10, rax
- cmp r12d, -68969733
- jo rx_i_145
-
-rx_i_83: ;IDIV_64
- dec ebx
- jz rx_finish
- xor r10, 0d9b6a533h
- mov eax, r10d
- test bl, 63
- jnz short rx_body_83
- call rx_read
-rx_body_83:
- and eax, 32767
- mov rax, qword ptr [rsi+rax*8]
- ; magic divide by 91850728
- mov rdx, 842358619687110887
- imul rdx
- mov rax, rdx
- xor edx, edx
- sar rax, 22
- sets dl
- add rax, rdx
- mov r12, rax
-
-rx_i_84: ;SAR_64
- dec ebx
- jz rx_finish
- xor r15, 0e9e75336h
- mov eax, r15d
- test bl, 63
- jnz short rx_body_84
- call rx_read
-rx_body_84:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- sar rax, 45
- mov rcx, rax
- mov eax, r13d
- xor eax, 0ec5c52e6h
- and eax, 131071
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_85: ;MUL_64
- dec ebx
- jz rx_finish
- xor r13, 04c0d378ah
- mov eax, r13d
- test bl, 63
- jnz short rx_body_85
- call rx_read
-rx_body_85:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- imul rax, rax, 20014507
- mov rcx, rax
- mov eax, r10d
- xor eax, 013165abh
- and eax, 32767
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_86: ;AND_64
- dec ebx
- jz rx_finish
- xor r11, 04386e368h
- mov eax, r11d
- test bl, 63
- jnz short rx_body_86
- call rx_read
-rx_body_86:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- and rax, r8
- mov r12, rax
-
-rx_i_87: ;SUB_64
- dec ebx
- jz rx_finish
- xor r9, 0d75a0ecfh
- mov eax, r9d
- test bl, 63
- jnz short rx_body_87
- call rx_read
-rx_body_87:
- xor rbp, rax
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- sub rax, r12
- mov r8, rax
-
-rx_i_88: ;ROR_64
- dec ebx
- jz rx_finish
- xor r9, 031bb7f7ah
- mov eax, r9d
- test bl, 63
- jnz short rx_body_88
- call rx_read
-rx_body_88:
- xor rbp, rax
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- mov rcx, r14
- ror rax, cl
- mov r9, rax
-
-rx_i_89: ;MUL_64
- dec ebx
- jz rx_finish
- xor r9, 03b45ecebh
- mov eax, r9d
- test bl, 63
- jnz short rx_body_89
- call rx_read
-rx_body_89:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- imul rax, r8
- mov r10, rax
-
-rx_i_90: ;FPADD
- dec ebx
- jz rx_finish
- xor r12, 0ee08e76bh
- mov eax, r12d
- test bl, 63
- jnz short rx_body_90
- call rx_read
-rx_body_90:
- and eax, 32767
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- addpd xmm0, xmm3
- movaps xmm6, xmm0
-
-rx_i_91: ;FPMUL
- dec ebx
- jz rx_finish
- xor r9, 042e28e94h
- mov eax, r9d
- test bl, 63
- jnz short rx_body_91
- call rx_read
-rx_body_91:
- xor rbp, rax
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- mulpd xmm0, xmm2
- movaps xmm1, xmm0
- cmpeqpd xmm1, xmm1
- andps xmm0, xmm1
- movaps xmm4, xmm0
-
-rx_i_92: ;JUMP
- dec ebx
- jz rx_finish
- xor r8, 0729260e1h
- mov eax, r8d
- test bl, 63
- jnz short rx_body_92
- call rx_read
-rx_body_92:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- mov r12, rax
- cmp r14d, 1288893603
- jge rx_i_170
-
-rx_i_93: ;FPADD
- dec ebx
- jz rx_finish
- xor r8, 0bfcebaf4h
- mov eax, r8d
- test bl, 63
- jnz short rx_body_93
- call rx_read
-rx_body_93:
- xor rbp, rax
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- addpd xmm0, xmm2
- movaps xmm2, xmm0
- mov eax, r10d
- xor eax, 07e48a0d8h
- and eax, 2047
- movhpd qword ptr [rsi + rax * 8], xmm2
-
-rx_i_94: ;CALL
- dec ebx
- jz rx_finish
- xor r13, 0ea326630h
- mov eax, r13d
- test bl, 63
- jnz short rx_body_94
- call rx_read
-rx_body_94:
- xor rbp, rax
- and eax, 32767
- mov rax, qword ptr [rsi+rax*8]
- mov rcx, rax
- mov eax, r8d
- xor eax, 0eb8c5be0h
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
- cmp r13d, -343122976
- jns short rx_i_95
- call rx_i_157
-
-rx_i_95: ;MUL_64
- dec ebx
- jz rx_finish
- xor r13, 0b5451a2dh
- mov eax, r13d
- test bl, 63
- jnz short rx_body_95
- call rx_read
-rx_body_95:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- imul rax, r10
- mov r15, rax
-
-rx_i_96: ;IMUL_32
- dec ebx
- jz rx_finish
- xor r11, 04f912ef8h
- mov eax, r11d
- test bl, 63
- jnz short rx_body_96
- call rx_read
-rx_body_96:
- xor rbp, rax
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- movsxd rcx, eax
- movsxd rax, r11d
- imul rax, rcx
- mov r11, rax
-
-rx_i_97: ;FPDIV
- dec ebx
- jz rx_finish
- xor r15, 0acc45b3bh
- mov eax, r15d
- test bl, 63
- jnz short rx_body_97
- call rx_read
-rx_body_97:
- and eax, 131071
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- divpd xmm0, xmm9
- movaps xmm1, xmm0
- cmpeqpd xmm1, xmm1
- andps xmm0, xmm1
- movaps xmm5, xmm0
-
-rx_i_98: ;SUB_64
- dec ebx
- jz rx_finish
- xor r14, 09900a4e8h
- mov eax, r14d
- test bl, 63
- jnz short rx_body_98
- call rx_read
-rx_body_98:
- and eax, 32767
- mov rax, qword ptr [rsi+rax*8]
- sub rax, r15
- mov rcx, rax
- mov eax, r14d
- xor eax, 0d067d49ah
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_99: ;FPMUL
- dec ebx
- jz rx_finish
- xor r9, 0841b2984h
- mov eax, r9d
- test bl, 63
- jnz short rx_body_99
- call rx_read
-rx_body_99:
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- mulpd xmm0, xmm6
- movaps xmm1, xmm0
- cmpeqpd xmm1, xmm1
- andps xmm0, xmm1
- movaps xmm4, xmm0
- mov eax, r12d
- xor eax, 04c21df83h
- and eax, 2047
- movhpd qword ptr [rsi + rax * 8], xmm4
-
-rx_i_100: ;ADD_64
- dec ebx
- jz rx_finish
- xor r15, 07ebea48fh
- mov eax, r15d
- test bl, 63
- jnz short rx_body_100
- call rx_read
-rx_body_100:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- add rax, r9
- mov r14, rax
-
-rx_i_101: ;SUB_64
- dec ebx
- jz rx_finish
- xor r10, 0631209d3h
- mov eax, r10d
- test bl, 63
- jnz short rx_body_101
- call rx_read
-rx_body_101:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- sub rax, 1732300336
- mov r11, rax
-
-rx_i_102: ;FPMUL
- dec ebx
- jz rx_finish
- xor r10, 0e50bf07ah
- mov eax, r10d
- test bl, 63
- jnz short rx_body_102
- call rx_read
-rx_body_102:
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- mulpd xmm0, xmm3
- movaps xmm1, xmm0
- cmpeqpd xmm1, xmm1
- andps xmm0, xmm1
- movaps xmm7, xmm0
-
-rx_i_103: ;MUL_64
- dec ebx
- jz rx_finish
- xor r10, 02b7096f1h
- mov eax, r10d
- test bl, 63
- jnz short rx_body_103
- call rx_read
-rx_body_103:
- and eax, 32767
- mov rax, qword ptr [rsi+rax*8]
- imul rax, r13
- mov rcx, rax
- mov eax, r15d
- xor eax, 0e4dd92b6h
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_104: ;DIV_64
- dec ebx
- jz rx_finish
- xor r11, 075deaf71h
- mov eax, r11d
- test bl, 63
- jnz short rx_body_104
- call rx_read
-rx_body_104:
- and eax, 32767
- mov rax, qword ptr [rsi+rax*8]
- ; magic divide by 2381897207
- mov rcx, 16631314374404138087
- mul rcx
- mov rax, rdx
- shr rax, 31
- mov r15, rax
-
-rx_i_105: ;MUL_32
- dec ebx
- jz rx_finish
- xor r13, 036a51f72h
- mov eax, r13d
- test bl, 63
- jnz short rx_body_105
- call rx_read
-rx_body_105:
- xor rbp, rax
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- mov ecx, eax
- mov eax, r15d
- imul rax, rcx
- mov r14, rax
-
-rx_i_106: ;FPMUL
- dec ebx
- jz rx_finish
- xor r11, 07b512986h
- mov eax, r11d
- test bl, 63
- jnz short rx_body_106
- call rx_read
-rx_body_106:
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- mulpd xmm0, xmm3
- movaps xmm1, xmm0
- cmpeqpd xmm1, xmm1
- andps xmm0, xmm1
- movaps xmm4, xmm0
- mov eax, r12d
- xor eax, 03cb2505h
- and eax, 2047
- movlpd qword ptr [rsi + rax * 8], xmm4
-
-rx_i_107: ;JUMP
- dec ebx
- jz rx_finish
- xor r12, 0f1d2e50h
- mov eax, r12d
- test bl, 63
- jnz short rx_body_107
- call rx_read
-rx_body_107:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- mov rcx, rax
- mov eax, r14d
- xor eax, 07243ab81h
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
- cmp r11d, 1917037441
- jl rx_i_143
-
-rx_i_108: ;FPMUL
- dec ebx
- jz rx_finish
- xor r9, 07327ba60h
- mov eax, r9d
- test bl, 63
- jnz short rx_body_108
- call rx_read
-rx_body_108:
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- mulpd xmm0, xmm5
- movaps xmm1, xmm0
- cmpeqpd xmm1, xmm1
- andps xmm0, xmm1
- movaps xmm9, xmm0
-
-rx_i_109: ;ROR_64
- dec ebx
- jz rx_finish
- xor r15, 0594e37deh
- mov eax, r15d
- test bl, 63
- jnz short rx_body_109
- call rx_read
-rx_body_109:
- xor rbp, rax
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- mov rcx, r10
- ror rax, cl
- mov rcx, rax
- mov eax, r11d
- xor eax, 094ab5a5ch
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_110: ;SHR_64
- dec ebx
- jz rx_finish
- xor r9, 04cdf5ebah
- mov eax, r9d
- test bl, 63
- jnz short rx_body_110
- call rx_read
-rx_body_110:
- and eax, 131071
- mov rax, qword ptr [rsi+rax*8]
- mov rcx, r9
- shr rax, cl
- mov rcx, rax
- mov eax, r14d
- xor eax, 0ec68532fh
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_111: ;CALL
- dec ebx
- jz rx_finish
- xor r8, 02e16c97ch
- mov eax, r8d
- test bl, 63
- jnz short rx_body_111
- call rx_read
-rx_body_111:
- and eax, 131071
- mov rax, qword ptr [rsi+rax*8]
- mov rcx, rax
- mov eax, r12d
- xor eax, 05d237d0bh
- and eax, 131071
- mov qword ptr [rsi + rax * 8], rcx
- cmp r14d, 1562606859
- jl short rx_i_112
- call rx_i_212
-
-rx_i_112: ;SUB_64
- dec ebx
- jz rx_finish
- xor r12, 0d42ddbd4h
- mov eax, r12d
- test bl, 63
- jnz short rx_body_112
- call rx_read
-rx_body_112:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- sub rax, -1025977295
- mov r14, rax
-
-rx_i_113: ;MULH_64
- dec ebx
- jz rx_finish
- xor r10, 07a4f8cbbh
- mov eax, r10d
- test bl, 63
- jnz short rx_body_113
- call rx_read
-rx_body_113:
- and eax, 32767
- mov rax, qword ptr [rsi+rax*8]
- mov rcx, r9
- mul rcx
- mov rax, rdx
- mov rcx, rax
- mov eax, r13d
- xor eax, 0dea3f7e3h
- and eax, 32767
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_114: ;DIV_64
- dec ebx
- jz rx_finish
- xor r13, 06e83e2cdh
- mov eax, r13d
- test bl, 63
- jnz short rx_body_114
- call rx_read
-rx_body_114:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- ; magic divide by 770835683
- mov rcx, 12847770974664443757
- mul rcx
- mov rax, rdx
- shr rax, 29
- mov r14, rax
-
-rx_i_115: ;IDIV_64
- dec ebx
- jz rx_finish
- xor r14, 0336c980eh
- mov eax, r14d
- test bl, 63
- jnz short rx_body_115
- call rx_read
-rx_body_115:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- ; magic divide by 587029837
- mov rdx, 527204905636414983
- imul rdx
- mov rax, rdx
- xor edx, edx
- sar rax, 24
- sets dl
- add rax, rdx
- mov r14, rax
-
-rx_i_116: ;DIV_64
- dec ebx
- jz rx_finish
- xor r10, 0d122702eh
- mov eax, r10d
- test bl, 63
- jnz short rx_body_116
- call rx_read
-rx_body_116:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- ; magic divide by 2444190605
- mov rcx, 16207443550472271289
- mul rcx
- mov rax, rdx
- shr rax, 31
- mov rcx, rax
- mov eax, r8d
- xor eax, 091af638dh
- and eax, 131071
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_117: ;IDIV_64
- dec ebx
- jz rx_finish
- xor r11, 015f2012bh
- mov eax, r11d
- test bl, 63
- jnz short rx_body_117
- call rx_read
-rx_body_117:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- ; magic divide by -1205826972
- mov rdx, -8213052572424165513
- imul rdx
- mov rax, rdx
- xor edx, edx
- sar rax, 29
- sets dl
- add rax, rdx
- mov r15, rax
-
-rx_i_118: ;FPSUB
- dec ebx
- jz rx_finish
- xor r9, 037ddf43dh
- mov eax, r9d
- test bl, 63
- jnz short rx_body_118
- call rx_read
-rx_body_118:
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- subpd xmm0, xmm5
- movaps xmm6, xmm0
-
-rx_i_119: ;FPSUB
- dec ebx
- jz rx_finish
- xor r9, 0bba475f3h
- mov eax, r9d
- test bl, 63
- jnz short rx_body_119
- call rx_read
-rx_body_119:
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- subpd xmm0, xmm3
- movaps xmm5, xmm0
- mov eax, r13d
- xor eax, 02401488h
- and eax, 2047
- movlpd qword ptr [rsi + rax * 8], xmm5
-
-rx_i_120: ;FPADD
- dec ebx
- jz rx_finish
- xor r12, 0e5561e3eh
- mov eax, r12d
- test bl, 63
- jnz short rx_body_120
- call rx_read
-rx_body_120:
- xor rbp, rax
- and eax, 32767
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- addpd xmm0, xmm4
- movaps xmm8, xmm0
-
-rx_i_121: ;FPSUB
- dec ebx
- jz rx_finish
- xor r9, 03ab8f73h
- mov eax, r9d
- test bl, 63
- jnz short rx_body_121
- call rx_read
-rx_body_121:
- and eax, 131071
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- subpd xmm0, xmm5
- movaps xmm8, xmm0
-
-rx_i_122: ;CALL
- dec ebx
- jz rx_finish
- xor r10, 04e0dbd40h
- mov eax, r10d
- test bl, 63
- jnz short rx_body_122
- call rx_read
-rx_body_122:
- and eax, 131071
- mov rax, qword ptr [rsi+rax*8]
- mov rcx, rax
- mov eax, r14d
- xor eax, 078f6ec29h
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
- cmp r11d, 2029448233
- jno short rx_i_123
- call rx_i_192
-
-rx_i_123: ;ADD_32
- dec ebx
- jz rx_finish
- xor r13, 073e9f58ah
- mov eax, r13d
- test bl, 63
- jnz short rx_body_123
- call rx_read
-rx_body_123:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- add eax, 1530846772
- mov r13, rax
-
-rx_i_124: ;JUMP
- dec ebx
- jz rx_finish
- xor r12, 0e3fa3670h
- mov eax, r12d
- test bl, 63
- jnz short rx_body_124
- call rx_read
-rx_body_124:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- mov r11, rax
- cmp r11d, 1719505436
- jns rx_i_237
-
-rx_i_125: ;IMUL_32
- dec ebx
- jz rx_finish
- xor r8, 0ebec27cdh
- mov eax, r8d
- test bl, 63
- jnz short rx_body_125
- call rx_read
-rx_body_125:
- xor rbp, rax
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- movsxd rcx, eax
- mov rax, 1774711622
- imul rax, rcx
- mov rcx, rax
- mov eax, r14d
- xor eax, 069c7f346h
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_126: ;FPMUL
- dec ebx
- jz rx_finish
- xor r8, 01feb5264h
- mov eax, r8d
- test bl, 63
- jnz short rx_body_126
- call rx_read
-rx_body_126:
- and eax, 32767
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- mulpd xmm0, xmm6
- movaps xmm1, xmm0
- cmpeqpd xmm1, xmm1
- andps xmm0, xmm1
- movaps xmm2, xmm0
-
-rx_i_127: ;IMUL_32
- dec ebx
- jz rx_finish
- xor r9, 0405f500fh
- mov eax, r9d
- test bl, 63
- jnz short rx_body_127
- call rx_read
-rx_body_127:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- movsxd rcx, eax
- mov rax, -1027270754
- imul rax, rcx
- mov r8, rax
-
-rx_i_128: ;MUL_64
- dec ebx
- jz rx_finish
- xor r13, 0459f1154h
- mov eax, r13d
- test bl, 63
- jnz short rx_body_128
- call rx_read
-rx_body_128:
- xor rbp, rax
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- imul rax, r9
- mov rcx, rax
- mov eax, r9d
- xor eax, 0cb2ee635h
- and eax, 32767
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_129: ;JUMP
- dec ebx
- jz rx_finish
- xor r9, 081918b4ch
- mov eax, r9d
- test bl, 63
- jnz short rx_body_129
- call rx_read
-rx_body_129:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- mov r9, rax
- cmp r13d, -590624856
- jge rx_i_154
-
-rx_i_130: ;IDIV_64
- dec ebx
- jz rx_finish
- xor r9, 077c3b332h
- mov eax, r9d
- test bl, 63
- jnz short rx_body_130
- call rx_read
-rx_body_130:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- ; magic divide by -281794782
- mov rdx, -8786110448882479839
- imul rdx
- mov rax, rdx
- xor edx, edx
- sar rax, 27
- sets dl
- add rax, rdx
- mov rcx, rax
- mov eax, r11d
- xor eax, 0ef342722h
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_131: ;RET
- dec ebx
- jz rx_finish
- xor r12, 05792310bh
- mov eax, r12d
- test bl, 63
- jnz short rx_body_131
- call rx_read
-rx_body_131:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- mov r15, rax
- cmp rsp, rdi
- je short rx_i_132
- ret
-
-rx_i_132: ;FPADD
- dec ebx
- jz rx_finish
- xor r10, 0ebc6e10h
- mov eax, r10d
- test bl, 63
- jnz short rx_body_132
- call rx_read
-rx_body_132:
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- addpd xmm0, xmm6
- movaps xmm7, xmm0
- mov eax, r15d
- xor eax, 0b0c38959h
- and eax, 2047
- movlpd qword ptr [rsi + rax * 8], xmm7
-
-rx_i_133: ;OR_64
- dec ebx
- jz rx_finish
- xor r14, 0822f8b60h
- mov eax, r14d
- test bl, 63
- jnz short rx_body_133
- call rx_read
-rx_body_133:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- or rax, r13
- mov r15, rax
-
-rx_i_134: ;ADD_64
- dec ebx
- jz rx_finish
- xor r10, 0d0f18593h
- mov eax, r10d
- test bl, 63
- jnz short rx_body_134
- call rx_read
-rx_body_134:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- add rax, r8
- mov rcx, rax
- mov eax, r13d
- xor eax, 05a5de2cbh
- and eax, 32767
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_135: ;FPMUL
- dec ebx
- jz rx_finish
- xor r11, 088212ef9h
- mov eax, r11d
- test bl, 63
- jnz short rx_body_135
- call rx_read
-rx_body_135:
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- mulpd xmm0, xmm3
- movaps xmm1, xmm0
- cmpeqpd xmm1, xmm1
- andps xmm0, xmm1
- movaps xmm8, xmm0
- mov eax, r8d
- xor eax, 0b29f3d2ah
- and eax, 2047
- movlpd qword ptr [rsi + rax * 8], xmm8
-
-rx_i_136: ;FPDIV
- dec ebx
- jz rx_finish
- xor r8, 01ae56e03h
- mov eax, r8d
- test bl, 63
- jnz short rx_body_136
- call rx_read
-rx_body_136:
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- divpd xmm0, xmm8
- movaps xmm1, xmm0
- cmpeqpd xmm1, xmm1
- andps xmm0, xmm1
- movaps xmm5, xmm0
-
-rx_i_137: ;SHR_64
- dec ebx
- jz rx_finish
- xor r11, 015a24231h
- mov eax, r11d
- test bl, 63
- jnz short rx_body_137
- call rx_read
-rx_body_137:
- xor rbp, rax
- and eax, 32767
- mov rax, qword ptr [rsi+rax*8]
- mov rcx, r9
- shr rax, cl
- mov r11, rax
-
-rx_i_138: ;RET
- dec ebx
- jz rx_finish
- xor r13, 02fd380c5h
- mov eax, r13d
- test bl, 63
- jnz short rx_body_138
- call rx_read
-rx_body_138:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- mov rcx, rax
- mov eax, r10d
- xor eax, 08e1fd158h
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
- cmp rsp, rdi
- je short rx_i_139
- ret
-
-rx_i_139: ;ADD_64
- dec ebx
- jz rx_finish
- xor r9, 093172470h
- mov eax, r9d
- test bl, 63
- jnz short rx_body_139
- call rx_read
-rx_body_139:
- xor rbp, rax
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- add rax, r8
- mov r11, rax
-
-rx_i_140: ;IMUL_32
- dec ebx
- jz rx_finish
- xor r14, 052543553h
- mov eax, r14d
- test bl, 63
- jnz short rx_body_140
- call rx_read
-rx_body_140:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- movsxd rcx, eax
- mov rax, -140239781
- imul rax, rcx
- mov rcx, rax
- mov eax, r14d
- xor eax, 0f7a41c5bh
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_141: ;FPADD
- dec ebx
- jz rx_finish
- xor r8, 02f636da1h
- mov eax, r8d
- test bl, 63
- jnz short rx_body_141
- call rx_read
-rx_body_141:
- xor rbp, rax
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- addpd xmm0, xmm2
- movaps xmm9, xmm0
- mov eax, r9d
- xor eax, 099ff9ffdh
- and eax, 2047
- movlpd qword ptr [rsi + rax * 8], xmm9
-
-rx_i_142: ;JUMP
- dec ebx
- jz rx_finish
- xor r11, 0b11a4f2ch
- mov eax, r11d
- test bl, 63
- jnz short rx_body_142
- call rx_read
-rx_body_142:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- mov rcx, rax
- mov eax, r10d
- xor eax, 0516a9452h
- and eax, 131071
- mov qword ptr [rsi + rax * 8], rcx
- cmp r12d, 1365939282
- js rx_i_257
-
-rx_i_143: ;IMUL_32
- dec ebx
- jz rx_finish
- xor r15, 037f4b5d0h
- mov eax, r15d
- test bl, 63
- jnz short rx_body_143
- call rx_read
-rx_body_143:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- movsxd rcx, eax
- movsxd rax, r11d
- imul rax, rcx
- mov r9, rax
-
-rx_i_144: ;DIV_64
- dec ebx
- jz rx_finish
- xor r10, 02e59e00ah
- mov eax, r10d
- test bl, 63
- jnz short rx_body_144
- call rx_read
-rx_body_144:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- mov ecx, 1
- mov edx, r11d
- test edx, edx
- cmovne ecx, edx
- xor edx, edx
- div rcx
- mov r15, rax
-
-rx_i_145: ;DIV_64
- dec ebx
- jz rx_finish
- xor r13, 08d5c798h
- mov eax, r13d
- test bl, 63
- jnz short rx_body_145
- call rx_read
-rx_body_145:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- ; magic divide by 3712555397
- mov rcx, 10670300378317066981
- mul rcx
- mov rax, rdx
- shr rax, 31
- mov rcx, rax
- mov eax, r10d
- xor eax, 0dd491985h
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_146: ;IMULH_64
- dec ebx
- jz rx_finish
- xor r13, 02327e6e2h
- mov eax, r13d
- test bl, 63
- jnz short rx_body_146
- call rx_read
-rx_body_146:
- xor rbp, rax
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- mov rcx, r12
- imul rcx
- mov rax, rdx
- mov r10, rax
-
-rx_i_147: ;MUL_64
- dec ebx
- jz rx_finish
- xor r13, 03a7df043h
- mov eax, r13d
- test bl, 63
- jnz short rx_body_147
- call rx_read
-rx_body_147:
- xor rbp, rax
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- imul rax, r11
- mov r12, rax
-
-rx_i_148: ;SUB_64
- dec ebx
- jz rx_finish
- xor r10, 0783e5c4eh
- mov eax, r10d
- test bl, 63
- jnz short rx_body_148
- call rx_read
-rx_body_148:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- sub rax, r14
- mov rcx, rax
- mov eax, r10d
- xor eax, 08c783d2ch
- and eax, 32767
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_149: ;MUL_32
- dec ebx
- jz rx_finish
- xor r12, 0aa0f5b2fh
- mov eax, r12d
- test bl, 63
- jnz short rx_body_149
- call rx_read
-rx_body_149:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- mov ecx, eax
- mov eax, r14d
- imul rax, rcx
- mov r8, rax
-
-rx_i_150: ;DIV_64
- dec ebx
- jz rx_finish
- xor r9, 01504ca7ah
- mov eax, r9d
- test bl, 63
- jnz short rx_body_150
- call rx_read
-rx_body_150:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- mov ecx, 1
- mov edx, r8d
- test edx, edx
- cmovne ecx, edx
- xor edx, edx
- div rcx
- mov rcx, rax
- mov eax, r9d
- xor eax, 0c854a524h
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_151: ;AND_64
- dec ebx
- jz rx_finish
- xor r9, 0ea72a7cfh
- mov eax, r9d
- test bl, 63
- jnz short rx_body_151
- call rx_read
-rx_body_151:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- and rax, -2018584590
- mov r11, rax
-
-rx_i_152: ;SAR_64
- dec ebx
- jz rx_finish
- xor r13, 0ad0e7a88h
- mov eax, r13d
- test bl, 63
- jnz short rx_body_152
- call rx_read
-rx_body_152:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- mov rcx, r10
- sar rax, cl
- mov r10, rax
-
-rx_i_153: ;FPMUL
- dec ebx
- jz rx_finish
- xor r15, 0fd95ab87h
- mov eax, r15d
- test bl, 63
- jnz short rx_body_153
- call rx_read
-rx_body_153:
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- mulpd xmm0, xmm2
- movaps xmm1, xmm0
- cmpeqpd xmm1, xmm1
- andps xmm0, xmm1
- movaps xmm8, xmm0
-
-rx_i_154: ;MUL_32
- dec ebx
- jz rx_finish
- xor r10, 0256697b0h
- mov eax, r10d
- test bl, 63
- jnz short rx_body_154
- call rx_read
-rx_body_154:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- mov ecx, eax
- mov eax, -820047839
- imul rax, rcx
- mov rcx, rax
- mov eax, r10d
- xor eax, 0cf1f1021h
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_155: ;ROL_64
- dec ebx
- jz rx_finish
- xor r11, 0d23f3b78h
- mov eax, r11d
- test bl, 63
- jnz short rx_body_155
- call rx_read
-rx_body_155:
- and eax, 32767
- mov rax, qword ptr [rsi+rax*8]
- mov rcx, r10
- rol rax, cl
- mov r13, rax
-
-rx_i_156: ;IMUL_32
- dec ebx
- jz rx_finish
- xor r10, 098917533h
- mov eax, r10d
- test bl, 63
- jnz short rx_body_156
- call rx_read
-rx_body_156:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- movsxd rcx, eax
- movsxd rax, r15d
- imul rax, rcx
- mov rcx, rax
- mov eax, r15d
- xor eax, 0b803e8a9h
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_157: ;ADD_64
- dec ebx
- jz rx_finish
- xor r10, 0dfac3efch
- mov eax, r10d
- test bl, 63
- jnz short rx_body_157
- call rx_read
-rx_body_157:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- add rax, r12
- mov r14, rax
-
-rx_i_158: ;ADD_64
- dec ebx
- jz rx_finish
- xor r15, 0a64de090h
- mov eax, r15d
- test bl, 63
- jnz short rx_body_158
- call rx_read
-rx_body_158:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- add rax, r13
- mov rcx, rax
- mov eax, r10d
- xor eax, 04984392fh
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_159: ;CALL
- dec ebx
- jz rx_finish
- xor r13, 0952a3abbh
- mov eax, r13d
- test bl, 63
- jnz short rx_body_159
- call rx_read
-rx_body_159:
- and eax, 131071
- mov rax, qword ptr [rsi+rax*8]
- mov r13, rax
- cmp r15d, -8571241
- ja short rx_i_160
- call rx_i_181
-
-rx_i_160: ;SUB_64
- dec ebx
- jz rx_finish
- xor r14, 0b1685b90h
- mov eax, r14d
- test bl, 63
- jnz short rx_body_160
- call rx_read
-rx_body_160:
- xor rbp, rax
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- sub rax, r14
- mov rcx, rax
- mov eax, r10d
- xor eax, 05a86b929h
- and eax, 32767
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_161: ;IDIV_64
- dec ebx
- jz rx_finish
- xor r15, 0ea992531h
- mov eax, r15d
- test bl, 63
- jnz short rx_body_161
- call rx_read
-rx_body_161:
- and eax, 131071
- mov rax, qword ptr [rsi+rax*8]
- mov edx, r14d
- cmp edx, -1
- jne short body_idiv_161
- neg rax
- jmp short result_idiv_161
-body_idiv_161:
- mov ecx, 1
- test edx, edx
- cmovne ecx, edx
- movsxd rcx, ecx
- cqo
- idiv rcx
-result_idiv_161:
- mov rcx, rax
- mov eax, r8d
- xor eax, 0db9043dah
- and eax, 32767
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_162: ;SHL_64
- dec ebx
- jz rx_finish
- xor r9, 01fd57a4ah
- mov eax, r9d
- test bl, 63
- jnz short rx_body_162
- call rx_read
-rx_body_162:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- shl rax, 7
- mov rcx, rax
- mov eax, r13d
- xor eax, 0170a46d8h
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_163: ;SUB_64
- dec ebx
- jz rx_finish
- xor r12, 0e3486c0ah
- mov eax, r12d
- test bl, 63
- jnz short rx_body_163
- call rx_read
-rx_body_163:
- xor rbp, rax
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- sub rax, r8
- mov r14, rax
-
-rx_i_164: ;MUL_32
- dec ebx
- jz rx_finish
- xor r12, 01f0c2737h
- mov eax, r12d
- test bl, 63
- jnz short rx_body_164
- call rx_read
-rx_body_164:
- xor rbp, rax
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- mov ecx, eax
- mov eax, r9d
- imul rax, rcx
- mov r13, rax
-
-rx_i_165: ;RET
- dec ebx
- jz rx_finish
- xor r12, 0debb493eh
- mov eax, r12d
- test bl, 63
- jnz short rx_body_165
- call rx_read
-rx_body_165:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- mov r9, rax
- cmp rsp, rdi
- je short rx_i_166
- ret
-
-rx_i_166: ;SHR_64
- dec ebx
- jz rx_finish
- xor r9, 0fe684081h
- mov eax, r9d
- test bl, 63
- jnz short rx_body_166
- call rx_read
-rx_body_166:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- shr rax, 62
- mov rcx, rax
- mov eax, r13d
- xor eax, 0bb67f8abh
- and eax, 131071
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_167: ;FPMUL
- dec ebx
- jz rx_finish
- xor r11, 0d10371ch
- mov eax, r11d
- test bl, 63
- jnz short rx_body_167
- call rx_read
-rx_body_167:
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- mulpd xmm0, xmm4
- movaps xmm1, xmm0
- cmpeqpd xmm1, xmm1
- andps xmm0, xmm1
- movaps xmm2, xmm0
-
-rx_i_168: ;FPDIV
- dec ebx
- jz rx_finish
- xor r12, 071b15effh
- mov eax, r12d
- test bl, 63
- jnz short rx_body_168
- call rx_read
-rx_body_168:
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- divpd xmm0, xmm5
- movaps xmm1, xmm0
- cmpeqpd xmm1, xmm1
- andps xmm0, xmm1
- movaps xmm7, xmm0
- mov eax, r15d
- xor eax, 08d1a76f8h
- and eax, 131071
- movhpd qword ptr [rsi + rax * 8], xmm7
-
-rx_i_169: ;CALL
- dec ebx
- jz rx_finish
- xor r11, 072790347h
- mov eax, r11d
- test bl, 63
- jnz short rx_body_169
- call rx_read
-rx_body_169:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- mov rcx, rax
- mov eax, r14d
- xor eax, 0b353bf8dh
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
- cmp r10d, -1286357107
- ja short rx_i_170
- call rx_i_197
-
-rx_i_170: ;FPSQRT
- dec ebx
- jz rx_finish
- xor r8, 04ae8a020h
- mov eax, r8d
- test bl, 63
- jnz short rx_body_170
- call rx_read
-rx_body_170:
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- andps xmm0, xmm10
- sqrtpd xmm6, xmm0
-
-rx_i_171: ;DIV_64
- dec ebx
- jz rx_finish
- xor r15, 09901e05bh
- mov eax, r15d
- test bl, 63
- jnz short rx_body_171
- call rx_read
-rx_body_171:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- ; magic divide by 2064150457
- add rax, 1
- sbb rax, 0
- mov rcx, 4797867461985617359
- mul rcx
- mov rax, rdx
- shr rax, 29
- mov rcx, rax
- mov eax, r12d
- xor eax, 07b086fb9h
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_172: ;SUB_64
- dec ebx
- jz rx_finish
- xor r13, 050e8c510h
- mov eax, r13d
- test bl, 63
- jnz short rx_body_172
- call rx_read
-rx_body_172:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- sub rax, -478081934
- mov r12, rax
-
-rx_i_173: ;MUL_64
- dec ebx
- jz rx_finish
- xor r14, 05422cf8fh
- mov eax, r14d
- test bl, 63
- jnz short rx_body_173
- call rx_read
-rx_body_173:
- xor rbp, rax
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- imul rax, rax, -1386172772
- mov r12, rax
-
-rx_i_174: ;FPDIV
- dec ebx
- jz rx_finish
- xor r12, 0a025c3dbh
- mov eax, r12d
- test bl, 63
- jnz short rx_body_174
- call rx_read
-rx_body_174:
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- divpd xmm0, xmm9
- movaps xmm1, xmm0
- cmpeqpd xmm1, xmm1
- andps xmm0, xmm1
- movaps xmm6, xmm0
- mov eax, r14d
- xor eax, 02be6989fh
- and eax, 32767
- movlpd qword ptr [rsi + rax * 8], xmm6
-
-rx_i_175: ;XOR_32
- dec ebx
- jz rx_finish
- xor r13, 08f74c11h
- mov eax, r13d
- test bl, 63
- jnz short rx_body_175
- call rx_read
-rx_body_175:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- xor eax, r8d
- mov r8, rax
-
-rx_i_176: ;SUB_64
- dec ebx
- jz rx_finish
- xor r9, 01f2ed5f1h
- mov eax, r9d
- test bl, 63
- jnz short rx_body_176
- call rx_read
-rx_body_176:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- sub rax, -2101315181
- mov r10, rax
-
-rx_i_177: ;ADD_64
- dec ebx
- jz rx_finish
- xor r10, 0d2072c79h
- mov eax, r10d
- test bl, 63
- jnz short rx_body_177
- call rx_read
-rx_body_177:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- add rax, 794235831
- mov r13, rax
-
-rx_i_178: ;RET
- dec ebx
- jz rx_finish
- xor r15, 0a8e51933h
- mov eax, r15d
- test bl, 63
- jnz short rx_body_178
- call rx_read
-rx_body_178:
- and eax, 32767
- mov rax, qword ptr [rsi+rax*8]
- mov rcx, rax
- mov eax, r12d
- xor eax, 0c366b275h
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
- cmp rsp, rdi
- je short rx_i_179
- ret
-
-rx_i_179: ;FPADD
- dec ebx
- jz rx_finish
- xor r12, 0934ad492h
- mov eax, r12d
- test bl, 63
- jnz short rx_body_179
- call rx_read
-rx_body_179:
- and eax, 32767
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- addpd xmm0, xmm2
- movaps xmm8, xmm0
-
-rx_i_180: ;AND_32
- dec ebx
- jz rx_finish
- xor r15, 01cb3ce1fh
- mov eax, r15d
- test bl, 63
- jnz short rx_body_180
- call rx_read
-rx_body_180:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- and eax, r9d
- mov rcx, rax
- mov eax, r9d
- xor eax, 076edfe13h
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_181: ;CALL
- dec ebx
- jz rx_finish
- xor r10, 023c7845fh
- mov eax, r10d
- test bl, 63
- jnz short rx_body_181
- call rx_read
-rx_body_181:
- xor rbp, rax
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- mov r10, rax
- cmp r12d, -1612576918
- jbe short rx_i_182
- call rx_i_211
-
-rx_i_182: ;FPSUB
- dec ebx
- jz rx_finish
- xor r8, 0f8884327h
- mov eax, r8d
- test bl, 63
- jnz short rx_body_182
- call rx_read
-rx_body_182:
- xor rbp, rax
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- subpd xmm0, xmm7
- movaps xmm6, xmm0
- mov eax, r14d
- xor eax, 07c8d12a5h
- and eax, 131071
- movhpd qword ptr [rsi + rax * 8], xmm6
-
-rx_i_183: ;ADD_64
- dec ebx
- jz rx_finish
- xor r13, 013070461h
- mov eax, r13d
- test bl, 63
- jnz short rx_body_183
- call rx_read
-rx_body_183:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- add rax, r11
- mov r10, rax
-
-rx_i_184: ;XOR_32
- dec ebx
- jz rx_finish
- xor r12, 04764cdf7h
- mov eax, r12d
- test bl, 63
- jnz short rx_body_184
- call rx_read
-rx_body_184:
- xor rbp, rax
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- xor eax, r13d
- mov rcx, rax
- mov eax, r12d
- xor eax, 02f185447h
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_185: ;JUMP
- dec ebx
- jz rx_finish
- xor r10, 03c41026fh
- mov eax, r10d
- test bl, 63
- jnz short rx_body_185
- call rx_read
-rx_body_185:
- and eax, 32767
- mov rax, qword ptr [rsi+rax*8]
- mov rcx, rax
- mov eax, r9d
- xor eax, 0a5fae4a3h
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
- cmp r15d, -1510284125
- jbe rx_i_246
-
-rx_i_186: ;OR_64
- dec ebx
- jz rx_finish
- xor r9, 0cded414bh
- mov eax, r9d
- test bl, 63
- jnz short rx_body_186
- call rx_read
-rx_body_186:
- xor rbp, rax
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- or rax, -1252263008
- mov rcx, rax
- mov eax, r10d
- xor eax, 0b55bfba0h
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_187: ;FPMUL
- dec ebx
- jz rx_finish
- xor r13, 05c6d64a8h
- mov eax, r13d
- test bl, 63
- jnz short rx_body_187
- call rx_read
-rx_body_187:
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- mulpd xmm0, xmm6
- movaps xmm1, xmm0
- cmpeqpd xmm1, xmm1
- andps xmm0, xmm1
- movaps xmm5, xmm0
-
-rx_i_188: ;FPSUB
- dec ebx
- jz rx_finish
- xor r9, 04659becbh
- mov eax, r9d
- test bl, 63
- jnz short rx_body_188
- call rx_read
-rx_body_188:
- xor rbp, rax
- and eax, 32767
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- subpd xmm0, xmm3
- movaps xmm4, xmm0
-
-rx_i_189: ;FPDIV
- dec ebx
- jz rx_finish
- xor r11, 0c52741d5h
- mov eax, r11d
- test bl, 63
- jnz short rx_body_189
- call rx_read
-rx_body_189:
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- divpd xmm0, xmm7
- movaps xmm1, xmm0
- cmpeqpd xmm1, xmm1
- andps xmm0, xmm1
- movaps xmm5, xmm0
-
-rx_i_190: ;RET
- dec ebx
- jz rx_finish
- xor r12, 0217bf5f3h
- mov eax, r12d
- test bl, 63
- jnz short rx_body_190
- call rx_read
-rx_body_190:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- mov r13, rax
- cmp rsp, rdi
- je short rx_i_191
- ret
-
-rx_i_191: ;FPSQRT
- dec ebx
- jz rx_finish
- xor r15, 0884f3526h
- mov eax, r15d
- test bl, 63
- jnz short rx_body_191
- call rx_read
-rx_body_191:
- and eax, 131071
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- andps xmm0, xmm10
- sqrtpd xmm6, xmm0
-
-rx_i_192: ;FPSQRT
- dec ebx
- jz rx_finish
- xor r8, 0d76edad3h
- mov eax, r8d
- test bl, 63
- jnz short rx_body_192
- call rx_read
-rx_body_192:
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- andps xmm0, xmm10
- sqrtpd xmm8, xmm0
-
-rx_i_193: ;MUL_32
- dec ebx
- jz rx_finish
- xor r12, 0e9939ach
- mov eax, r12d
- test bl, 63
- jnz short rx_body_193
- call rx_read
-rx_body_193:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- mov ecx, eax
- mov eax, r12d
- imul rax, rcx
- mov rcx, rax
- mov eax, r15d
- xor eax, 074e097dch
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_194: ;FPMUL
- dec ebx
- jz rx_finish
- xor r12, 0f21ca520h
- mov eax, r12d
- test bl, 63
- jnz short rx_body_194
- call rx_read
-rx_body_194:
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- mulpd xmm0, xmm2
- movaps xmm1, xmm0
- cmpeqpd xmm1, xmm1
- andps xmm0, xmm1
- movaps xmm5, xmm0
-
-rx_i_195: ;SHL_64
- dec ebx
- jz rx_finish
- xor r10, 09405152ch
- mov eax, r10d
- test bl, 63
- jnz short rx_body_195
- call rx_read
-rx_body_195:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- shl rax, 27
- mov r9, rax
-
-rx_i_196: ;SUB_64
- dec ebx
- jz rx_finish
- xor r8, 0c2a9f41bh
- mov eax, r8d
- test bl, 63
- jnz short rx_body_196
- call rx_read
-rx_body_196:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- sub rax, r8
- mov rcx, rax
- mov eax, r13d
- xor eax, 08e47b269h
- and eax, 131071
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_197: ;MUL_64
- dec ebx
- jz rx_finish
- xor r12, 0229208efh
- mov eax, r12d
- test bl, 63
- jnz short rx_body_197
- call rx_read
-rx_body_197:
- xor rbp, rax
- and eax, 32767
- mov rax, qword ptr [rsi+rax*8]
- imul rax, r15
- mov rcx, rax
- mov eax, r11d
- xor eax, 0b1d1e60dh
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_198: ;MULH_64
- dec ebx
- jz rx_finish
- xor r14, 0c8d95bbbh
- mov eax, r14d
- test bl, 63
- jnz short rx_body_198
- call rx_read
-rx_body_198:
- and eax, 32767
- mov rax, qword ptr [rsi+rax*8]
- mov rcx, r14
- mul rcx
- mov rax, rdx
- mov r8, rax
-
-rx_i_199: ;MULH_64
- dec ebx
- jz rx_finish
- xor r13, 050049e2eh
- mov eax, r13d
- test bl, 63
- jnz short rx_body_199
- call rx_read
-rx_body_199:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- mov rcx, r10
- mul rcx
- mov rax, rdx
- mov r10, rax
-
-rx_i_200: ;FPSUB
- dec ebx
- jz rx_finish
- xor r10, 0c63b99e8h
- mov eax, r10d
- test bl, 63
- jnz short rx_body_200
- call rx_read
-rx_body_200:
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- subpd xmm0, xmm2
- movaps xmm4, xmm0
-
-rx_i_201: ;FPADD
- dec ebx
- jz rx_finish
- xor r8, 0cdda801dh
- mov eax, r8d
- test bl, 63
- jnz short rx_body_201
- call rx_read
-rx_body_201:
- xor rbp, rax
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- addpd xmm0, xmm9
- movaps xmm4, xmm0
- mov eax, r12d
- xor eax, 040cfe68eh
- and eax, 2047
- movlpd qword ptr [rsi + rax * 8], xmm4
-
-rx_i_202: ;FPADD
- dec ebx
- jz rx_finish
- xor r13, 0fa44b04ah
- mov eax, r13d
- test bl, 63
- jnz short rx_body_202
- call rx_read
-rx_body_202:
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- addpd xmm0, xmm9
- movaps xmm5, xmm0
-
-rx_i_203: ;FPSUB
- dec ebx
- jz rx_finish
- xor r10, 0d73e472ch
- mov eax, r10d
- test bl, 63
- jnz short rx_body_203
- call rx_read
-rx_body_203:
- and eax, 32767
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- subpd xmm0, xmm2
- movaps xmm7, xmm0
- mov eax, r15d
- xor eax, 09bdff355h
- and eax, 2047
- movhpd qword ptr [rsi + rax * 8], xmm7
-
-rx_i_204: ;MUL_64
- dec ebx
- jz rx_finish
- xor r9, 01af8ab1dh
- mov eax, r9d
- test bl, 63
- jnz short rx_body_204
- call rx_read
-rx_body_204:
- and eax, 32767
- mov rax, qword ptr [rsi+rax*8]
- imul rax, r15
- mov r8, rax
-
-rx_i_205: ;FPMUL
- dec ebx
- jz rx_finish
- xor r14, 094e997c5h
- mov eax, r14d
- test bl, 63
- jnz short rx_body_205
- call rx_read
-rx_body_205:
- xor rbp, rax
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- mulpd xmm0, xmm8
- movaps xmm1, xmm0
- cmpeqpd xmm1, xmm1
- andps xmm0, xmm1
- movaps xmm5, xmm0
-
-rx_i_206: ;FPSUB
- dec ebx
- jz rx_finish
- xor r11, 0e836a177h
- mov eax, r11d
- test bl, 63
- jnz short rx_body_206
- call rx_read
-rx_body_206:
- xor rbp, rax
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- subpd xmm0, xmm7
- movaps xmm4, xmm0
- mov eax, r12d
- xor eax, 0d01fb731h
- and eax, 2047
- movlpd qword ptr [rsi + rax * 8], xmm4
-
-rx_i_207: ;IDIV_64
- dec ebx
- jz rx_finish
- xor r9, 039ccdd30h
- mov eax, r9d
- test bl, 63
- jnz short rx_body_207
- call rx_read
-rx_body_207:
- xor rbp, rax
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- ; magic divide by 314297476
- mov rdx, 1969376361274661135
- imul rdx
- mov rax, rdx
- xor edx, edx
- sar rax, 25
- sets dl
- add rax, rdx
- mov rcx, rax
- mov eax, r9d
- xor eax, 012bbcc84h
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_208: ;MUL_64
- dec ebx
- jz rx_finish
- xor r9, 0f4f126c5h
- mov eax, r9d
- test bl, 63
- jnz short rx_body_208
- call rx_read
-rx_body_208:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- imul rax, rax, -486588965
- mov rcx, rax
- mov eax, r10d
- xor eax, 0e2ff3ddbh
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_209: ;XOR_64
- dec ebx
- jz rx_finish
- xor r8, 0b84811f1h
- mov eax, r8d
- test bl, 63
- jnz short rx_body_209
- call rx_read
-rx_body_209:
- xor rbp, rax
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- xor rax, r15
- mov rcx, rax
- mov eax, r12d
- xor eax, 0c36b836ah
- and eax, 32767
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_210: ;MUL_32
- dec ebx
- jz rx_finish
- xor r12, 0c5efc90ah
- mov eax, r12d
- test bl, 63
- jnz short rx_body_210
- call rx_read
-rx_body_210:
- xor rbp, rax
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- mov ecx, eax
- mov eax, r12d
- imul rax, rcx
- mov rcx, rax
- mov eax, r15d
- xor eax, 0c2c6bee0h
- and eax, 131071
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_211: ;ROR_64
- dec ebx
- jz rx_finish
- xor r12, 0ce533072h
- mov eax, r12d
- test bl, 63
- jnz short rx_body_211
- call rx_read
-rx_body_211:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- mov rcx, r9
- ror rax, cl
- mov rcx, rax
- mov eax, r11d
- xor eax, 0212e615h
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_212: ;MUL_64
- dec ebx
- jz rx_finish
- xor r13, 06b465fdbh
- mov eax, r13d
- test bl, 63
- jnz short rx_body_212
- call rx_read
-rx_body_212:
- xor rbp, rax
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- imul rax, r13
- mov r15, rax
-
-rx_i_213: ;IMUL_32
- dec ebx
- jz rx_finish
- xor r13, 02dd1d503h
- mov eax, r13d
- test bl, 63
- jnz short rx_body_213
- call rx_read
-rx_body_213:
- xor rbp, rax
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- movsxd rcx, eax
- movsxd rax, r14d
- imul rax, rcx
- mov rcx, rax
- mov eax, r14d
- xor eax, 07bf8b75h
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_214: ;SHL_64
- dec ebx
- jz rx_finish
- xor r9, 0a159f313h
- mov eax, r9d
- test bl, 63
- jnz short rx_body_214
- call rx_read
-rx_body_214:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- mov rcx, r14
- shl rax, cl
- mov rcx, rax
- mov eax, r14d
- xor eax, 0936ebe0bh
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_215: ;ADD_32
- dec ebx
- jz rx_finish
- xor r15, 08359265eh
- mov eax, r15d
- test bl, 63
- jnz short rx_body_215
- call rx_read
-rx_body_215:
- xor rbp, rax
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- add eax, r12d
- mov rcx, rax
- mov eax, r10d
- xor eax, 01194f02bh
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_216: ;MUL_64
- dec ebx
- jz rx_finish
- xor r12, 080696de3h
- mov eax, r12d
- test bl, 63
- jnz short rx_body_216
- call rx_read
-rx_body_216:
- and eax, 131071
- mov rax, qword ptr [rsi+rax*8]
- imul rax, r13
- mov rcx, rax
- mov eax, r15d
- xor eax, 03b609d2bh
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_217: ;IMUL_32
- dec ebx
- jz rx_finish
- xor r8, 040d5b526h
- mov eax, r8d
- test bl, 63
- jnz short rx_body_217
- call rx_read
-rx_body_217:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- movsxd rcx, eax
- movsxd rax, r9d
- imul rax, rcx
- mov r10, rax
-
-rx_i_218: ;FPSQRT
- dec ebx
- jz rx_finish
- xor r11, 083c0bd93h
- mov eax, r11d
- test bl, 63
- jnz short rx_body_218
- call rx_read
-rx_body_218:
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- andps xmm0, xmm10
- sqrtpd xmm3, xmm0
-
-rx_i_219: ;OR_64
- dec ebx
- jz rx_finish
- xor r8, 0ca37f668h
- mov eax, r8d
- test bl, 63
- jnz short rx_body_219
- call rx_read
-rx_body_219:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- or rax, r10
- mov rcx, rax
- mov eax, r15d
- xor eax, 0d3d68798h
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_220: ;IMUL_32
- dec ebx
- jz rx_finish
- xor r9, 0bb44c384h
- mov eax, r9d
- test bl, 63
- jnz short rx_body_220
- call rx_read
-rx_body_220:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- movsxd rcx, eax
- movsxd rax, r11d
- imul rax, rcx
- mov r11, rax
-
-rx_i_221: ;DIV_64
- dec ebx
- jz rx_finish
- xor r9, 0a3deb512h
- mov eax, r9d
- test bl, 63
- jnz short rx_body_221
- call rx_read
-rx_body_221:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- mov ecx, 1
- mov edx, r15d
- test edx, edx
- cmovne ecx, edx
- xor edx, edx
- div rcx
- mov r11, rax
-
-rx_i_222: ;FPMUL
- dec ebx
- jz rx_finish
- xor r9, 084a02d64h
- mov eax, r9d
- test bl, 63
- jnz short rx_body_222
- call rx_read
-rx_body_222:
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- mulpd xmm0, xmm5
- movaps xmm1, xmm0
- cmpeqpd xmm1, xmm1
- andps xmm0, xmm1
- movaps xmm7, xmm0
-
-rx_i_223: ;FPSUB
- dec ebx
- jz rx_finish
- xor r8, 01e5cc085h
- mov eax, r8d
- test bl, 63
- jnz short rx_body_223
- call rx_read
-rx_body_223:
- xor rbp, rax
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- subpd xmm0, xmm3
- movaps xmm2, xmm0
- mov eax, r10d
- xor eax, 07fca59eeh
- and eax, 32767
- movhpd qword ptr [rsi + rax * 8], xmm2
-
-rx_i_224: ;XOR_32
- dec ebx
- jz rx_finish
- xor r12, 053982440h
- mov eax, r12d
- test bl, 63
- jnz short rx_body_224
- call rx_read
-rx_body_224:
- and eax, 32767
- mov rax, qword ptr [rsi+rax*8]
- xor eax, -452933987
- mov rcx, rax
- mov eax, r11d
- xor eax, 0e500c69dh
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_225: ;DIV_64
- dec ebx
- jz rx_finish
- xor r13, 0c558367eh
- mov eax, r13d
- test bl, 63
- jnz short rx_body_225
- call rx_read
-rx_body_225:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- ; magic divide by 4264577610
- shr rax, 1
- mov rcx, 9289098447696480965
- mul rcx
- mov rax, rdx
- shr rax, 30
- mov rcx, rax
- mov eax, r12d
- xor eax, 0fe304a4ah
- and eax, 32767
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_226: ;JUMP
- dec ebx
- jz rx_finish
- xor r10, 040139b65h
- mov eax, r10d
- test bl, 63
- jnz short rx_body_226
- call rx_read
-rx_body_226:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- mov r8, rax
- cmp r8d, -1752488808
- jno rx_i_328
-
-rx_i_227: ;FPMUL
- dec ebx
- jz rx_finish
- xor r11, 0fa312dbdh
- mov eax, r11d
- test bl, 63
- jnz short rx_body_227
- call rx_read
-rx_body_227:
- and eax, 32767
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- mulpd xmm0, xmm7
- movaps xmm1, xmm0
- cmpeqpd xmm1, xmm1
- andps xmm0, xmm1
- movaps xmm3, xmm0
- mov eax, r11d
- xor eax, 0aabe2a0ah
- and eax, 2047
- movhpd qword ptr [rsi + rax * 8], xmm3
-
-rx_i_228: ;FPSQRT
- dec ebx
- jz rx_finish
- xor r11, 0b64246c0h
- mov eax, r11d
- test bl, 63
- jnz short rx_body_228
- call rx_read
-rx_body_228:
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- andps xmm0, xmm10
- sqrtpd xmm7, xmm0
- mov eax, r15d
- xor eax, 0ffdff798h
- and eax, 2047
- movhpd qword ptr [rsi + rax * 8], xmm7
-
-rx_i_229: ;IMULH_64
- dec ebx
- jz rx_finish
- xor r11, 05c535836h
- mov eax, r11d
- test bl, 63
- jnz short rx_body_229
- call rx_read
-rx_body_229:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- mov rcx, 334017248
- imul rcx
- mov rax, rdx
- mov r13, rax
-
-rx_i_230: ;FPMUL
- dec ebx
- jz rx_finish
- xor r15, 0f394972eh
- mov eax, r15d
- test bl, 63
- jnz short rx_body_230
- call rx_read
-rx_body_230:
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- mulpd xmm0, xmm6
- movaps xmm1, xmm0
- cmpeqpd xmm1, xmm1
- andps xmm0, xmm1
- movaps xmm5, xmm0
- mov eax, r13d
- xor eax, 01dc2b4f6h
- and eax, 2047
- movhpd qword ptr [rsi + rax * 8], xmm5
-
-rx_i_231: ;RET
- dec ebx
- jz rx_finish
- xor r9, 0bb56428dh
- mov eax, r9d
- test bl, 63
- jnz short rx_body_231
- call rx_read
-rx_body_231:
- xor rbp, rax
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- mov r9, rax
- cmp rsp, rdi
- je short rx_i_232
- ret
-
-rx_i_232: ;FPMUL
- dec ebx
- jz rx_finish
- xor r15, 09ab46ab3h
- mov eax, r15d
- test bl, 63
- jnz short rx_body_232
- call rx_read
-rx_body_232:
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- mulpd xmm0, xmm3
- movaps xmm1, xmm0
- cmpeqpd xmm1, xmm1
- andps xmm0, xmm1
- movaps xmm7, xmm0
- mov eax, r15d
- xor eax, 07e732935h
- and eax, 2047
- movhpd qword ptr [rsi + rax * 8], xmm7
-
-rx_i_233: ;JUMP
- dec ebx
- jz rx_finish
- xor r13, 08eb2cd76h
- mov eax, r13d
- test bl, 63
- jnz short rx_body_233
- call rx_read
-rx_body_233:
- xor rbp, rax
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- mov r14, rax
- cmp r12d, 392389867
- jo rx_i_268
-
-rx_i_234: ;FPDIV
- dec ebx
- jz rx_finish
- xor r15, 0ba687578h
- mov eax, r15d
- test bl, 63
- jnz short rx_body_234
- call rx_read
-rx_body_234:
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- divpd xmm0, xmm4
- movaps xmm1, xmm0
- cmpeqpd xmm1, xmm1
- andps xmm0, xmm1
- movaps xmm4, xmm0
-
-rx_i_235: ;IMUL_32
- dec ebx
- jz rx_finish
- xor r13, 0b6cb9ff2h
- mov eax, r13d
- test bl, 63
- jnz short rx_body_235
- call rx_read
-rx_body_235:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- movsxd rcx, eax
- mov rax, 212286089
- imul rax, rcx
- mov r15, rax
-
-rx_i_236: ;FPADD
- dec ebx
- jz rx_finish
- xor r15, 03ad196ach
- mov eax, r15d
- test bl, 63
- jnz short rx_body_236
- call rx_read
-rx_body_236:
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- addpd xmm0, xmm4
- movaps xmm3, xmm0
-
-rx_i_237: ;JUMP
- dec ebx
- jz rx_finish
- xor r15, 0fab4600h
- mov eax, r15d
- test bl, 63
- jnz short rx_body_237
- call rx_read
-rx_body_237:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- mov r11, rax
- cmp r12d, -121899164
- jge rx_i_295
-
-rx_i_238: ;FPADD
- dec ebx
- jz rx_finish
- xor r8, 0158f119fh
- mov eax, r8d
- test bl, 63
- jnz short rx_body_238
- call rx_read
-rx_body_238:
- xor rbp, rax
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- addpd xmm0, xmm6
- movaps xmm7, xmm0
-
-rx_i_239: ;ADD_64
- dec ebx
- jz rx_finish
- xor r13, 044f30b3fh
- mov eax, r13d
- test bl, 63
- jnz short rx_body_239
- call rx_read
-rx_body_239:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- add rax, r10
- mov rcx, rax
- mov eax, r10d
- xor eax, 0e42cdf41h
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_240: ;IMULH_64
- dec ebx
- jz rx_finish
- xor r9, 0d65d29f9h
- mov eax, r9d
- test bl, 63
- jnz short rx_body_240
- call rx_read
-rx_body_240:
- xor rbp, rax
- and eax, 131071
- mov rax, qword ptr [rsi+rax*8]
- mov rcx, r14
- imul rcx
- mov rax, rdx
- mov rcx, rax
- mov eax, r8d
- xor eax, 0e6bcdcfbh
- and eax, 32767
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_241: ;FPADD
- dec ebx
- jz rx_finish
- xor r11, 0ce5260adh
- mov eax, r11d
- test bl, 63
- jnz short rx_body_241
- call rx_read
-rx_body_241:
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- addpd xmm0, xmm3
- movaps xmm7, xmm0
-
-rx_i_242: ;MUL_32
- dec ebx
- jz rx_finish
- xor r12, 01119b0f9h
- mov eax, r12d
- test bl, 63
- jnz short rx_body_242
- call rx_read
-rx_body_242:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- mov ecx, eax
- mov eax, r12d
- imul rax, rcx
- mov r10, rax
-
-rx_i_243: ;OR_64
- dec ebx
- jz rx_finish
- xor r12, 0d6c2ce3dh
- mov eax, r12d
- test bl, 63
- jnz short rx_body_243
- call rx_read
-rx_body_243:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- or rax, r9
- mov r14, rax
-
-rx_i_244: ;ROR_64
- dec ebx
- jz rx_finish
- xor r11, 0c6a6248h
- mov eax, r11d
- test bl, 63
- jnz short rx_body_244
- call rx_read
-rx_body_244:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- mov rcx, r14
- ror rax, cl
- mov rcx, rax
- mov eax, r9d
- xor eax, 0b4a1fad6h
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_245: ;AND_32
- dec ebx
- jz rx_finish
- xor r13, 084505739h
- mov eax, r13d
- test bl, 63
- jnz short rx_body_245
- call rx_read
-rx_body_245:
- xor rbp, rax
- and eax, 131071
- mov rax, qword ptr [rsi+rax*8]
- and eax, r10d
- mov rcx, rax
- mov eax, r12d
- xor eax, 0a3d1ad8bh
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_246: ;IDIV_64
- dec ebx
- jz rx_finish
- xor r15, 027eeaa2eh
- mov eax, r15d
- test bl, 63
- jnz short rx_body_246
- call rx_read
-rx_body_246:
- xor rbp, rax
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- ; magic divide by -156808488
- mov rdx, -3947299202596036367
- imul rdx
- mov rax, rdx
- xor edx, edx
- sar rax, 25
- sets dl
- add rax, rdx
- mov r12, rax
-
-rx_i_247: ;IMUL_32
- dec ebx
- jz rx_finish
- xor r10, 0c4de0296h
- mov eax, r10d
- test bl, 63
- jnz short rx_body_247
- call rx_read
-rx_body_247:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- movsxd rcx, eax
- movsxd rax, r14d
- imul rax, rcx
- mov rcx, rax
- mov eax, r9d
- xor eax, 03814cf80h
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_248: ;MUL_32
- dec ebx
- jz rx_finish
- xor r8, 0649df46fh
- mov eax, r8d
- test bl, 63
- jnz short rx_body_248
- call rx_read
-rx_body_248:
- xor rbp, rax
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- mov ecx, eax
- mov eax, r15d
- imul rax, rcx
- mov rcx, rax
- mov eax, r9d
- xor eax, 07b10fc32h
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_249: ;IMULH_64
- dec ebx
- jz rx_finish
- xor r15, 0499552cch
- mov eax, r15d
- test bl, 63
- jnz short rx_body_249
- call rx_read
-rx_body_249:
- xor rbp, rax
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- mov rcx, -508571655
- imul rcx
- mov rax, rdx
- mov r13, rax
-
-rx_i_250: ;MUL_64
- dec ebx
- jz rx_finish
- xor r13, 083eafe6fh
- mov eax, r13d
- test bl, 63
- jnz short rx_body_250
- call rx_read
-rx_body_250:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- imul rax, r8
- mov r14, rax
-
-rx_i_251: ;FPMUL
- dec ebx
- jz rx_finish
- xor r13, 0a25a4d8ah
- mov eax, r13d
- test bl, 63
- jnz short rx_body_251
- call rx_read
-rx_body_251:
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- mulpd xmm0, xmm2
- movaps xmm1, xmm0
- cmpeqpd xmm1, xmm1
- andps xmm0, xmm1
- movaps xmm4, xmm0
-
-rx_i_252: ;SHL_64
- dec ebx
- jz rx_finish
- xor r14, 08a75ad41h
- mov eax, r14d
- test bl, 63
- jnz short rx_body_252
- call rx_read
-rx_body_252:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- shl rax, 53
- mov rcx, rax
- mov eax, r14d
- xor eax, 0b178001h
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_253: ;CALL
- dec ebx
- jz rx_finish
- xor r14, 057f3f596h
- mov eax, r14d
- test bl, 63
- jnz short rx_body_253
- call rx_read
-rx_body_253:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- mov rcx, rax
- mov eax, r13d
- xor eax, 0654b460bh
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
- cmp r15d, 1699431947
- js short rx_i_254
- call rx_i_367
-
-rx_i_254: ;FPADD
- dec ebx
- jz rx_finish
- xor r14, 04cfb709eh
- mov eax, r14d
- test bl, 63
- jnz short rx_body_254
- call rx_read
-rx_body_254:
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- addpd xmm0, xmm4
- movaps xmm8, xmm0
-
-rx_i_255: ;FPADD
- dec ebx
- jz rx_finish
- xor r9, 0b96ec9ech
- mov eax, r9d
- test bl, 63
- jnz short rx_body_255
- call rx_read
-rx_body_255:
- and eax, 32767
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- addpd xmm0, xmm5
- movaps xmm6, xmm0
- mov eax, r14d
- xor eax, 0ae781d10h
- and eax, 2047
- movlpd qword ptr [rsi + rax * 8], xmm6
-
-rx_i_256: ;MULH_64
- dec ebx
- jz rx_finish
- xor r8, 08375472ch
- mov eax, r8d
- test bl, 63
- jnz short rx_body_256
- call rx_read
-rx_body_256:
- xor rbp, rax
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- mov rcx, r15
- mul rcx
- mov rax, rdx
- mov r9, rax
-
-rx_i_257: ;FPADD
- dec ebx
- jz rx_finish
- xor r12, 0d75a8c3fh
- mov eax, r12d
- test bl, 63
- jnz short rx_body_257
- call rx_read
-rx_body_257:
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- addpd xmm0, xmm5
- movaps xmm3, xmm0
-
-rx_i_258: ;MUL_32
- dec ebx
- jz rx_finish
- xor r11, 064fdbda0h
- mov eax, r11d
- test bl, 63
- jnz short rx_body_258
- call rx_read
-rx_body_258:
- xor rbp, rax
- and eax, 32767
- mov rax, qword ptr [rsi+rax*8]
- mov ecx, eax
- mov eax, r14d
- imul rax, rcx
- mov rcx, rax
- mov eax, r9d
- xor eax, 01c58ef2dh
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_259: ;FPADD
- dec ebx
- jz rx_finish
- xor r11, 02e36a073h
- mov eax, r11d
- test bl, 63
- jnz short rx_body_259
- call rx_read
-rx_body_259:
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- addpd xmm0, xmm9
- movaps xmm3, xmm0
- mov eax, r11d
- xor eax, 06c1856f0h
- and eax, 32767
- movlpd qword ptr [rsi + rax * 8], xmm3
-
-rx_i_260: ;FPSUB
- dec ebx
- jz rx_finish
- xor r13, 0f94e9fa9h
- mov eax, r13d
- test bl, 63
- jnz short rx_body_260
- call rx_read
-rx_body_260:
- xor rbp, rax
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- subpd xmm0, xmm5
- movaps xmm9, xmm0
-
-rx_i_261: ;FPDIV
- dec ebx
- jz rx_finish
- xor r14, 02346171ch
- mov eax, r14d
- test bl, 63
- jnz short rx_body_261
- call rx_read
-rx_body_261:
- xor rbp, rax
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- divpd xmm0, xmm3
- movaps xmm1, xmm0
- cmpeqpd xmm1, xmm1
- andps xmm0, xmm1
- movaps xmm3, xmm0
- mov eax, r11d
- xor eax, 0745a48e9h
- and eax, 2047
- movhpd qword ptr [rsi + rax * 8], xmm3
-
-rx_i_262: ;AND_64
- dec ebx
- jz rx_finish
- xor r10, 01c42baa6h
- mov eax, r10d
- test bl, 63
- jnz short rx_body_262
- call rx_read
-rx_body_262:
- xor rbp, rax
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- and rax, -1569587450
- mov rcx, rax
- mov eax, r11d
- xor eax, 0a271ff06h
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_263: ;FPMUL
- dec ebx
- jz rx_finish
- xor r11, 0b39b140h
- mov eax, r11d
- test bl, 63
- jnz short rx_body_263
- call rx_read
-rx_body_263:
- xor rbp, rax
- and eax, 131071
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- mulpd xmm0, xmm8
- movaps xmm1, xmm0
- cmpeqpd xmm1, xmm1
- andps xmm0, xmm1
- movaps xmm6, xmm0
-
-rx_i_264: ;FPMUL
- dec ebx
- jz rx_finish
- xor r11, 01a07d201h
- mov eax, r11d
- test bl, 63
- jnz short rx_body_264
- call rx_read
-rx_body_264:
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- mulpd xmm0, xmm3
- movaps xmm1, xmm0
- cmpeqpd xmm1, xmm1
- andps xmm0, xmm1
- movaps xmm7, xmm0
-
-rx_i_265: ;FPADD
- dec ebx
- jz rx_finish
- xor r13, 07a3eb340h
- mov eax, r13d
- test bl, 63
- jnz short rx_body_265
- call rx_read
-rx_body_265:
- xor rbp, rax
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
+ and eax, 16376
+ cvtdq2pd xmm12, qword ptr [rsi+rax]
+ addpd xmm0, xmm12
+ ; IMUL_R r6, r5
+ imul r14, r13
+ ; IROL_R r4, r1
+ mov ecx, r9d
+ rol r12, cl
+ ; FPDIV_R e2, a0
+ divpd xmm6, xmm8
+ maxpd xmm6, xmm13
+ ; IADD_RC r0, r2, -487084195
+ lea r8, [r8+r10-487084195]
+ ; FPADD_R f0, a0
addpd xmm0, xmm8
- movaps xmm2, xmm0
- mov eax, r10d
- xor eax, 04c559414h
- and eax, 2047
- movhpd qword ptr [rsi + rax * 8], xmm2
-
-rx_i_266: ;CALL
- dec ebx
- jz rx_finish
- xor r13, 03d0a3a89h
- mov eax, r13d
- test bl, 63
- jnz short rx_body_266
- call rx_read
-rx_body_266:
- and eax, 32767
- mov rax, qword ptr [rsi+rax*8]
- mov r10, rax
- cmp r12d, 136160027
- jbe short rx_i_267
- call rx_i_295
-
-rx_i_267: ;ROL_64
- dec ebx
- jz rx_finish
- xor r8, 0c6c7b37h
- mov eax, r8d
- test bl, 63
- jnz short rx_body_267
- call rx_read
-rx_body_267:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- mov rcx, r10
- rol rax, cl
- mov r11, rax
-
-rx_i_268: ;JUMP
- dec ebx
- jz rx_finish
- xor r12, 0c2510cebh
+ ; IXOR_R r5, r3
+ xor r13, r11
+ ; IMUL_R r2, r4
+ imul r10, r12
+ ; FPMUL_R e0, a0
+ mulpd xmm4, xmm8
+ ; FPSUB_R f3, a3
+ subpd xmm3, xmm11
+ ; IMUL_M r4, L1[4856]
+ imul r12, qword ptr [rsi+4856]
+ ; IMUL_9C r2, 7951348
+ lea r10, [r10+r10*8+7951348]
+ ; COND_R r3, ab(r7, 984532162)
+ xor ecx, ecx
+ cmp r15d, 984532162
+ seta cl
+ add r11, rcx
+ ; IXOR_M r7, L1[r4]
mov eax, r12d
- test bl, 63
- jnz short rx_body_268
- call rx_read
-rx_body_268:
- xor rbp, rax
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- mov rcx, rax
+ and eax, 16376
+ xor r15, qword ptr [rsi+rax]
+ ; IMUL_R r4, 248971329
+ imul r12, 248971329
+ ; IXOR_R r3, r1
+ xor r11, r9
+ ; IMUL_R r3, 2098482639
+ imul r11, 2098482639
+ ; IXOR_R r6, r3
+ xor r14, r11
+ ; IXOR_R r5, r4
+ xor r13, r12
+ ; IADD_R r5, r4
+ add r13, r12
+ ; IMUL_9C r7, 66530302
+ lea r15, [r15+r15*8+66530302]
+ ; IMULH_R r0, r5
+ mov rax, r8
+ mul r13
+ mov r8, rdx
+ ; IMUL_R r2, r7
+ imul r10, r15
+ ; IMUL_R r1, 770985098
+ imul r9, 770985098
+ ; COND_R r7, be(r5, 58538265)
+ xor ecx, ecx
+ cmp r13d, 58538265
+ setbe cl
+ add r15, rcx
+ ; IMUL_9C r3, 245704334
+ lea r11, [r11+r11*8+245704334]
+ ; ISMULH_R r2, r4
+ mov rax, r10
+ imul r12
+ mov r10, rdx
+ ; FPDIV_R e3, a3
+ divpd xmm7, xmm11
+ maxpd xmm7, xmm13
+ ; IMULH_R r5, r2
+ mov rax, r13
+ mul r10
+ mov r13, rdx
+ ; ISUB_M r7, L1[r5]
mov eax, r13d
- xor eax, 0850bf8dah
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
- cmp r15d, -2062812966
- jl rx_i_381
-
-rx_i_269: ;ROL_64
- dec ebx
- jz rx_finish
- xor r11, 0c80cc899h
+ and eax, 16376
+ sub r15, qword ptr [rsi+rax]
+ ; FPMUL_R e3, a3
+ mulpd xmm7, xmm11
+ ; IMUL_R r3, r4
+ imul r11, r12
+ ; FPSWAP_R f1
+ shufpd xmm1, xmm1, 1
+ ; IMULH_R r1, 633797287
+ mov eax, 633797287
+ mul r9
+ add r9, rdx
+ ; IADD_R r4, r3
+ add r12, r11
+ ; IROR_R r2, r7
+ mov ecx, r15d
+ ror r10, cl
+ ; FPSUB_R f0, a2
+ subpd xmm0, xmm10
+ ; FPSUB_R f2, a2
+ subpd xmm2, xmm10
+ ; FPMUL_R e0, a2
+ mulpd xmm4, xmm10
+ ; IMUL_M r4, L1[r3]
mov eax, r11d
- test bl, 63
- jnz short rx_body_269
- call rx_read
-rx_body_269:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- rol rax, 50
- mov r10, rax
-
-rx_i_270: ;FPMUL
- dec ebx
- jz rx_finish
- xor r11, 0eb355caah
- mov eax, r11d
- test bl, 63
- jnz short rx_body_270
- call rx_read
-rx_body_270:
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- mulpd xmm0, xmm9
- movaps xmm1, xmm0
- cmpeqpd xmm1, xmm1
- andps xmm0, xmm1
- movaps xmm7, xmm0
+ and eax, 16376
+ imul r12, qword ptr [rsi+rax]
+ ; IMUL_9C r1, -1901091890
+ lea r9, [r9+r9*8-1901091890]
+ ; IROR_R r2, r6
+ mov ecx, r14d
+ ror r10, cl
+ ; IMULH_R r5, r3
+ mov rax, r13
+ mul r11
+ mov r13, rdx
+ ; FPSUB_M f1, L1[r7]
mov eax, r15d
- xor eax, 03981662bh
- and eax, 32767
- movhpd qword ptr [rsi + rax * 8], xmm7
-
-rx_i_271: ;MUL_32
- dec ebx
- jz rx_finish
- xor r13, 0c6f12299h
- mov eax, r13d
- test bl, 63
- jnz short rx_body_271
- call rx_read
-rx_body_271:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- mov ecx, eax
- mov eax, r10d
- imul rax, rcx
- mov rcx, rax
+ and eax, 16376
+ cvtdq2pd xmm12, qword ptr [rsi+rax]
+ subpd xmm1, xmm12
+ ; IMUL_M r2, L1[r1]
mov eax, r9d
- xor eax, 086ddd754h
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_272: ;AND_64
- dec ebx
- jz rx_finish
- xor r12, 0695a5dd2h
- mov eax, r12d
- test bl, 63
- jnz short rx_body_272
- call rx_read
-rx_body_272:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- and rax, r12
- mov rcx, rax
- mov eax, r13d
- xor eax, 0d45957b7h
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_273: ;JUMP
- dec ebx
- jz rx_finish
- xor r9, 0d315e4dch
- mov eax, r9d
- test bl, 63
- jnz short rx_body_273
- call rx_read
-rx_body_273:
- xor rbp, rax
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- mov r13, rax
- cmp r12d, 1670848568
- jl rx_i_372
-
-rx_i_274: ;FPADD
- dec ebx
- jz rx_finish
- xor r15, 0b66ca7e0h
- mov eax, r15d
- test bl, 63
- jnz short rx_body_274
- call rx_read
-rx_body_274:
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- addpd xmm0, xmm4
- movaps xmm6, xmm0
-
-rx_i_275: ;IDIV_64
- dec ebx
- jz rx_finish
- xor r10, 0788eceb7h
- mov eax, r10d
- test bl, 63
- jnz short rx_body_275
- call rx_read
-rx_body_275:
- xor rbp, rax
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- ; magic divide by -333089764
- mov rdx, -7433071640624659213
- imul rdx
- mov rax, rdx
- xor edx, edx
- sar rax, 27
- sets dl
- add rax, rdx
- mov r13, rax
-
-rx_i_276: ;JUMP
- dec ebx
- jz rx_finish
- xor r9, 0c6ac5edah
- mov eax, r9d
- test bl, 63
- jnz short rx_body_276
- call rx_read
-rx_body_276:
- xor rbp, rax
- and eax, 131071
- mov rax, qword ptr [rsi+rax*8]
- mov r12, rax
- cmp r11d, -1236180570
- jns rx_i_404
-
-rx_i_277: ;IMUL_32
- dec ebx
- jz rx_finish
- xor r11, 0c9549789h
- mov eax, r11d
- test bl, 63
- jnz short rx_body_277
- call rx_read
-rx_body_277:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- movsxd rcx, eax
- movsxd rax, r10d
- imul rax, rcx
- mov r9, rax
-
-rx_i_278: ;FPSUB
- dec ebx
- jz rx_finish
- xor r9, 0a2bc66c9h
- mov eax, r9d
- test bl, 63
- jnz short rx_body_278
- call rx_read
-rx_body_278:
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- subpd xmm0, xmm7
- movaps xmm4, xmm0
-
-rx_i_279: ;FPADD
- dec ebx
- jz rx_finish
- xor r15, 0f1a91458h
- mov eax, r15d
- test bl, 63
- jnz short rx_body_279
- call rx_read
-rx_body_279:
- and eax, 32767
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- addpd xmm0, xmm5
- movaps xmm9, xmm0
- mov eax, r9d
- xor eax, 0475ade01h
- and eax, 2047
- movlpd qword ptr [rsi + rax * 8], xmm9
-
-rx_i_280: ;IDIV_64
- dec ebx
- jz rx_finish
- xor r12, 066246b43h
- mov eax, r12d
- test bl, 63
- jnz short rx_body_280
- call rx_read
-rx_body_280:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- ; magic divide by 555412224
- mov rdx, 2228867111296024113
- imul rdx
- mov rax, rdx
- xor edx, edx
- sar rax, 26
- sets dl
- add rax, rdx
- mov rcx, rax
- mov eax, r13d
- xor eax, 0211aeb00h
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_281: ;SUB_64
- dec ebx
- jz rx_finish
- xor r10, 05a762727h
- mov eax, r10d
- test bl, 63
- jnz short rx_body_281
- call rx_read
-rx_body_281:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- sub rax, -202979002
- mov rcx, rax
- mov eax, r11d
- xor eax, 0f3e6c946h
- and eax, 32767
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_282: ;SUB_64
- dec ebx
- jz rx_finish
- xor r15, 0de1ab603h
- mov eax, r15d
- test bl, 63
- jnz short rx_body_282
- call rx_read
-rx_body_282:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- sub rax, r12
- mov r11, rax
-
-rx_i_283: ;ADD_64
- dec ebx
- jz rx_finish
- xor r9, 0df4d084fh
- mov eax, r9d
- test bl, 63
- jnz short rx_body_283
- call rx_read
-rx_body_283:
- xor rbp, rax
- and eax, 32767
- mov rax, qword ptr [rsi+rax*8]
- add rax, r12
- mov rcx, rax
- mov eax, r12d
- xor eax, 0bb0da7d0h
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_284: ;FPADD
- dec ebx
- jz rx_finish
- xor r15, 0e68f36ach
- mov eax, r15d
- test bl, 63
- jnz short rx_body_284
- call rx_read
-rx_body_284:
- xor rbp, rax
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- addpd xmm0, xmm6
- movaps xmm9, xmm0
- mov eax, r9d
- xor eax, 0936f2960h
- and eax, 32767
- movlpd qword ptr [rsi + rax * 8], xmm9
-
-rx_i_285: ;IMUL_32
- dec ebx
- jz rx_finish
- xor r8, 09adb333bh
- mov eax, r8d
- test bl, 63
- jnz short rx_body_285
- call rx_read
-rx_body_285:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- movsxd rcx, eax
- movsxd rax, r8d
- imul rax, rcx
- mov rcx, rax
- mov eax, r14d
- xor eax, 09308cd6dh
- and eax, 32767
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_286: ;ROL_64
- dec ebx
- jz rx_finish
- xor r14, 082f5e36ch
- mov eax, r14d
- test bl, 63
- jnz short rx_body_286
- call rx_read
-rx_body_286:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- mov rcx, r9
- rol rax, cl
- mov r15, rax
-
-rx_i_287: ;IDIV_64
- dec ebx
- jz rx_finish
- xor r11, 049547c9ch
- mov eax, r11d
- test bl, 63
- jnz short rx_body_287
- call rx_read
-rx_body_287:
- xor rbp, rax
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- ; magic divide by 1227278330
- mov rdx, 8069498232143512385
- imul rdx
- mov rax, rdx
- xor edx, edx
- sar rax, 29
- sets dl
- add rax, rdx
- mov r8, rax
-
-rx_i_288: ;MUL_64
- dec ebx
- jz rx_finish
- xor r10, 08716ac8bh
- mov eax, r10d
- test bl, 63
- jnz short rx_body_288
- call rx_read
-rx_body_288:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- imul rax, r8
- mov rcx, rax
- mov eax, r9d
- xor eax, 062eafa1bh
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_289: ;FPMUL
- dec ebx
- jz rx_finish
- xor r14, 0efef52b5h
- mov eax, r14d
- test bl, 63
- jnz short rx_body_289
- call rx_read
-rx_body_289:
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- mulpd xmm0, xmm9
- movaps xmm1, xmm0
- cmpeqpd xmm1, xmm1
- andps xmm0, xmm1
- movaps xmm8, xmm0
-
-rx_i_290: ;FPSUB
- dec ebx
- jz rx_finish
- xor r15, 060665748h
- mov eax, r15d
- test bl, 63
- jnz short rx_body_290
- call rx_read
-rx_body_290:
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- subpd xmm0, xmm8
- movaps xmm9, xmm0
- mov eax, r9d
- xor eax, 02f4d18d7h
- and eax, 2047
- movhpd qword ptr [rsi + rax * 8], xmm9
-
-rx_i_291: ;RET
- dec ebx
- jz rx_finish
- xor r13, 0ddf4bd1ah
- mov eax, r13d
- test bl, 63
- jnz short rx_body_291
- call rx_read
-rx_body_291:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- mov rcx, rax
- mov eax, r14d
- xor eax, 0768a9d75h
- and eax, 131071
- mov qword ptr [rsi + rax * 8], rcx
- cmp rsp, rdi
- je short rx_i_292
- ret
-
-rx_i_292: ;ROL_64
- dec ebx
- jz rx_finish
- xor r13, 05a87cc3dh
- mov eax, r13d
- test bl, 63
- jnz short rx_body_292
- call rx_read
-rx_body_292:
- and eax, 32767
- mov rax, qword ptr [rsi+rax*8]
- mov rcx, r8
- rol rax, cl
- mov rcx, rax
- mov eax, r10d
- xor eax, 035600fe9h
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_293: ;FPADD
- dec ebx
- jz rx_finish
- xor r9, 0c61f4279h
- mov eax, r9d
- test bl, 63
- jnz short rx_body_293
- call rx_read
-rx_body_293:
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- addpd xmm0, xmm5
- movaps xmm8, xmm0
- mov eax, r8d
- xor eax, 014844990h
- and eax, 2047
- movhpd qword ptr [rsi + rax * 8], xmm8
-
-rx_i_294: ;RET
- dec ebx
- jz rx_finish
- xor r14, 0f3b9d85h
- mov eax, r14d
- test bl, 63
- jnz short rx_body_294
- call rx_read
-rx_body_294:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- mov r8, rax
- cmp rsp, rdi
- je short rx_i_295
- ret
-
-rx_i_295: ;FPSUB
- dec ebx
- jz rx_finish
- xor r9, 0f42798fdh
- mov eax, r9d
- test bl, 63
- jnz short rx_body_295
- call rx_read
-rx_body_295:
- xor rbp, rax
- and eax, 131071
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- subpd xmm0, xmm8
- movaps xmm7, xmm0
- mov eax, r15d
- xor eax, 08a66e69fh
- and eax, 2047
- movlpd qword ptr [rsi + rax * 8], xmm7
-
-rx_i_296: ;FPSQRT
- dec ebx
- jz rx_finish
- xor r14, 018738758h
- mov eax, r14d
- test bl, 63
- jnz short rx_body_296
- call rx_read
-rx_body_296:
- and eax, 131071
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- andps xmm0, xmm10
- sqrtpd xmm8, xmm0
- mov eax, r8d
- xor eax, 0f3a594cah
- and eax, 2047
- movlpd qword ptr [rsi + rax * 8], xmm8
-
-rx_i_297: ;ADD_64
- dec ebx
- jz rx_finish
- xor r15, 0de3b9d9bh
- mov eax, r15d
- test bl, 63
- jnz short rx_body_297
- call rx_read
-rx_body_297:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- add rax, r10
- mov r14, rax
-
-rx_i_298: ;FPSUB
- dec ebx
- jz rx_finish
- xor r14, 084f53637h
- mov eax, r14d
- test bl, 63
- jnz short rx_body_298
- call rx_read
-rx_body_298:
- and eax, 32767
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- subpd xmm0, xmm7
- movaps xmm6, xmm0
- mov eax, r14d
- xor eax, 0d10f7c42h
- and eax, 2047
- movhpd qword ptr [rsi + rax * 8], xmm6
-
-rx_i_299: ;ADD_64
- dec ebx
- jz rx_finish
- xor r12, 042f4897h
- mov eax, r12d
- test bl, 63
- jnz short rx_body_299
- call rx_read
-rx_body_299:
- xor rbp, rax
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- add rax, r10
- mov r12, rax
-
-rx_i_300: ;FPSUB
- dec ebx
- jz rx_finish
- xor r12, 095765693h
- mov eax, r12d
- test bl, 63
- jnz short rx_body_300
- call rx_read
-rx_body_300:
- xor rbp, rax
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- subpd xmm0, xmm8
- movaps xmm2, xmm0
-
-rx_i_301: ;FPMUL
- dec ebx
- jz rx_finish
- xor r8, 0a0ec5eech
- mov eax, r8d
- test bl, 63
- jnz short rx_body_301
- call rx_read
-rx_body_301:
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- mulpd xmm0, xmm5
- movaps xmm1, xmm0
- cmpeqpd xmm1, xmm1
- andps xmm0, xmm1
- movaps xmm7, xmm0
-
-rx_i_302: ;ADD_64
- dec ebx
- jz rx_finish
- xor r15, 0f6f8c345h
- mov eax, r15d
- test bl, 63
- jnz short rx_body_302
- call rx_read
-rx_body_302:
- xor rbp, rax
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- add rax, r10
- mov rcx, rax
- mov eax, r11d
- xor eax, 0afbbe406h
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_303: ;FPADD
- dec ebx
- jz rx_finish
- xor r14, 082a3e965h
- mov eax, r14d
- test bl, 63
- jnz short rx_body_303
- call rx_read
-rx_body_303:
- and eax, 32767
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- addpd xmm0, xmm3
- movaps xmm9, xmm0
-
-rx_i_304: ;MUL_64
- dec ebx
- jz rx_finish
- xor r12, 04940c652h
- mov eax, r12d
- test bl, 63
- jnz short rx_body_304
- call rx_read
-rx_body_304:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- imul rax, rax, 2007686513
- mov r13, rax
-
-rx_i_305: ;MUL_64
- dec ebx
- jz rx_finish
- xor r11, 03c6c62b8h
- mov eax, r11d
- test bl, 63
- jnz short rx_body_305
- call rx_read
-rx_body_305:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- imul rax, r15
- mov rcx, rax
- mov eax, r10d
- xor eax, 0fc12db20h
- and eax, 32767
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_306: ;ADD_64
- dec ebx
- jz rx_finish
- xor r15, 08b34cdfch
- mov eax, r15d
- test bl, 63
- jnz short rx_body_306
- call rx_read
-rx_body_306:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- add rax, 400578979
- mov rcx, rax
- mov eax, r13d
- xor eax, 017e059a3h
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_307: ;SHL_64
- dec ebx
- jz rx_finish
- xor r15, 04c36adb1h
- mov eax, r15d
- test bl, 63
- jnz short rx_body_307
- call rx_read
-rx_body_307:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- shl rax, 33
- mov r10, rax
-
-rx_i_308: ;MUL_64
- dec ebx
- jz rx_finish
- xor r11, 0a4213b21h
- mov eax, r11d
- test bl, 63
- jnz short rx_body_308
- call rx_read
-rx_body_308:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- imul rax, r13
- mov rcx, rax
- mov eax, r15d
- xor eax, 0c2d34e82h
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_309: ;DIV_64
- dec ebx
- jz rx_finish
- xor r9, 090c42304h
- mov eax, r9d
- test bl, 63
- jnz short rx_body_309
- call rx_read
-rx_body_309:
- and eax, 32767
- mov rax, qword ptr [rsi+rax*8]
- ; magic divide by 2642117268
- mov rcx, 14993309243657753043
- mul rcx
- mov rax, rdx
- shr rax, 31
- mov r9, rax
-
-rx_i_310: ;FPMUL
- dec ebx
- jz rx_finish
- xor r9, 0f78e1c8ch
- mov eax, r9d
- test bl, 63
- jnz short rx_body_310
- call rx_read
-rx_body_310:
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- mulpd xmm0, xmm6
- movaps xmm1, xmm0
- cmpeqpd xmm1, xmm1
- andps xmm0, xmm1
- movaps xmm7, xmm0
-
-rx_i_311: ;FPMUL
- dec ebx
- jz rx_finish
- xor r8, 0ff8848cfh
- mov eax, r8d
- test bl, 63
- jnz short rx_body_311
- call rx_read
-rx_body_311:
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- mulpd xmm0, xmm4
- movaps xmm1, xmm0
- cmpeqpd xmm1, xmm1
- andps xmm0, xmm1
- movaps xmm4, xmm0
-
-rx_i_312: ;MUL_32
- dec ebx
- jz rx_finish
- xor r13, 0b18904cdh
- mov eax, r13d
- test bl, 63
- jnz short rx_body_312
- call rx_read
-rx_body_312:
- and eax, 32767
- mov rax, qword ptr [rsi+rax*8]
- mov ecx, eax
- mov eax, r14d
- imul rax, rcx
- mov rcx, rax
- mov eax, r10d
- xor eax, 0bb93ffb8h
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_313: ;ROR_64
- dec ebx
- jz rx_finish
- xor r8, 0a0d0befh
- mov eax, r8d
- test bl, 63
- jnz short rx_body_313
- call rx_read
-rx_body_313:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- ror rax, 62
- mov r14, rax
-
-rx_i_314: ;IMUL_32
- dec ebx
- jz rx_finish
- xor r15, 01e3c65f7h
- mov eax, r15d
- test bl, 63
- jnz short rx_body_314
- call rx_read
-rx_body_314:
- and eax, 32767
- mov rax, qword ptr [rsi+rax*8]
- movsxd rcx, eax
- mov rax, 2143811925
- imul rax, rcx
- mov rcx, rax
- mov eax, r9d
- xor eax, 07fc7f955h
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_315: ;XOR_64
- dec ebx
- jz rx_finish
- xor r9, 02e36ddafh
- mov eax, r9d
- test bl, 63
- jnz short rx_body_315
- call rx_read
-rx_body_315:
- and eax, 32767
- mov rax, qword ptr [rsi+rax*8]
- xor rax, r15
- mov r9, rax
-
-rx_i_316: ;RET
- dec ebx
- jz rx_finish
- xor r14, 05b0cb5bbh
- mov eax, r14d
- test bl, 63
- jnz short rx_body_316
- call rx_read
-rx_body_316:
- xor rbp, rax
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- mov rcx, rax
- mov eax, r8d
- xor eax, 03602c513h
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
- cmp rsp, rdi
- je short rx_i_317
- ret
-
-rx_i_317: ;FPADD
- dec ebx
- jz rx_finish
- xor r9, 0c74e7415h
- mov eax, r9d
- test bl, 63
- jnz short rx_body_317
- call rx_read
-rx_body_317:
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- addpd xmm0, xmm7
- movaps xmm5, xmm0
- mov eax, r13d
- xor eax, 0b5bc8h
- and eax, 2047
- movlpd qword ptr [rsi + rax * 8], xmm5
-
-rx_i_318: ;ROR_64
- dec ebx
- jz rx_finish
- xor r9, 057621d9ah
- mov eax, r9d
- test bl, 63
- jnz short rx_body_318
- call rx_read
-rx_body_318:
- and eax, 32767
- mov rax, qword ptr [rsi+rax*8]
- mov rcx, r11
- ror rax, cl
- mov rcx, rax
- mov eax, r15d
- xor eax, 061cb9db8h
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_319: ;SHR_64
- dec ebx
- jz rx_finish
- xor r13, 08ee02d99h
- mov eax, r13d
- test bl, 63
- jnz short rx_body_319
- call rx_read
-rx_body_319:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- shr rax, 46
- mov r11, rax
-
-rx_i_320: ;FPADD
- dec ebx
- jz rx_finish
- xor r15, 013461188h
- mov eax, r15d
- test bl, 63
- jnz short rx_body_320
- call rx_read
-rx_body_320:
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- addpd xmm0, xmm4
- movaps xmm2, xmm0
-
-rx_i_321: ;IMUL_32
- dec ebx
- jz rx_finish
- xor r11, 0a7bae383h
- mov eax, r11d
- test bl, 63
- jnz short rx_body_321
- call rx_read
-rx_body_321:
- and eax, 32767
- mov rax, qword ptr [rsi+rax*8]
- movsxd rcx, eax
- movsxd rax, r9d
- imul rax, rcx
- mov rcx, rax
- mov eax, r12d
- xor eax, 0f213dach
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_322: ;CALL
- dec ebx
- jz rx_finish
- xor r14, 08215399bh
- mov eax, r14d
- test bl, 63
- jnz short rx_body_322
- call rx_read
-rx_body_322:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- mov r11, rax
- cmp r11d, 1411981860
- jno short rx_i_323
- call rx_i_343
-
-rx_i_323: ;MULH_64
- dec ebx
- jz rx_finish
- xor r14, 07b07664bh
- mov eax, r14d
- test bl, 63
- jnz short rx_body_323
- call rx_read
-rx_body_323:
- xor rbp, rax
- and eax, 32767
- mov rax, qword ptr [rsi+rax*8]
- mov rcx, r14
- mul rcx
- mov rax, rdx
- mov r14, rax
-
-rx_i_324: ;FPDIV
- dec ebx
- jz rx_finish
- xor r9, 0f956baffh
- mov eax, r9d
- test bl, 63
- jnz short rx_body_324
- call rx_read
-rx_body_324:
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- divpd xmm0, xmm2
- movaps xmm1, xmm0
- cmpeqpd xmm1, xmm1
- andps xmm0, xmm1
- movaps xmm9, xmm0
-
-rx_i_325: ;OR_32
- dec ebx
- jz rx_finish
- xor r11, 0708ab9d1h
- mov eax, r11d
- test bl, 63
- jnz short rx_body_325
- call rx_read
-rx_body_325:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- or eax, r8d
- mov rcx, rax
- mov eax, r13d
- xor eax, 0ef376c54h
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_326: ;MULH_64
- dec ebx
- jz rx_finish
- xor r11, 0d1b27540h
- mov eax, r11d
- test bl, 63
- jnz short rx_body_326
- call rx_read
-rx_body_326:
- xor rbp, rax
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- mov rcx, -1233771581
- mul rcx
- mov rax, rdx
- mov r9, rax
-
-rx_i_327: ;IDIV_64
- dec ebx
- jz rx_finish
- xor r9, 09665f98dh
- mov eax, r9d
- test bl, 63
- jnz short rx_body_327
- call rx_read
-rx_body_327:
- xor rbp, rax
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- ; magic divide by 1572662125
- mov rcx, rax
- mov rdx, -5852150286715358951
- imul rdx
- mov rax, rdx
- xor edx, edx
- add rax, rcx
- sar rax, 30
- sets dl
- add rax, rdx
- mov r12, rax
-
-rx_i_328: ;SHR_64
- dec ebx
- jz rx_finish
- xor r12, 0fb9c32adh
- mov eax, r12d
- test bl, 63
- jnz short rx_body_328
- call rx_read
-rx_body_328:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- shr rax, 18
- mov rcx, rax
- mov eax, r9d
- xor eax, 04d159415h
- and eax, 32767
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_329: ;RET
- dec ebx
- jz rx_finish
- xor r11, 0e1110623h
- mov eax, r11d
- test bl, 63
- jnz short rx_body_329
- call rx_read
-rx_body_329:
- and eax, 131071
- mov rax, qword ptr [rsi+rax*8]
- mov r11, rax
- cmp rsp, rdi
- je short rx_i_330
- ret
-
-rx_i_330: ;IMUL_32
- dec ebx
- jz rx_finish
- xor r9, 0f6a93f19h
- mov eax, r9d
- test bl, 63
- jnz short rx_body_330
- call rx_read
-rx_body_330:
- xor rbp, rax
- and eax, 32767
- mov rax, qword ptr [rsi+rax*8]
- movsxd rcx, eax
- mov rax, -1349816041
- imul rax, rcx
- mov r11, rax
-
-rx_i_331: ;FPADD
- dec ebx
- jz rx_finish
- xor r9, 0bc9bbe4ah
- mov eax, r9d
- test bl, 63
- jnz short rx_body_331
- call rx_read
-rx_body_331:
- xor rbp, rax
- and eax, 32767
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- addpd xmm0, xmm3
- movaps xmm9, xmm0
-
-rx_i_332: ;FPADD
- dec ebx
- jz rx_finish
- xor r12, 0f253cd4eh
- mov eax, r12d
- test bl, 63
- jnz short rx_body_332
- call rx_read
-rx_body_332:
- xor rbp, rax
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- addpd xmm0, xmm6
- movaps xmm3, xmm0
-
-rx_i_333: ;OR_64
- dec ebx
- jz rx_finish
- xor r14, 0f009758bh
- mov eax, r14d
- test bl, 63
- jnz short rx_body_333
- call rx_read
-rx_body_333:
- xor rbp, rax
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- or rax, r12
- mov rcx, rax
- mov eax, r11d
- xor eax, 0f58fcaa8h
- and eax, 32767
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_334: ;ADD_64
- dec ebx
- jz rx_finish
- xor r8, 0dda04168h
- mov eax, r8d
- test bl, 63
- jnz short rx_body_334
- call rx_read
-rx_body_334:
- xor rbp, rax
- and eax, 32767
- mov rax, qword ptr [rsi+rax*8]
- add rax, r13
- mov r8, rax
-
-rx_i_335: ;SUB_64
- dec ebx
- jz rx_finish
- xor r15, 03e6cfb73h
- mov eax, r15d
- test bl, 63
- jnz short rx_body_335
- call rx_read
-rx_body_335:
- and eax, 32767
- mov rax, qword ptr [rsi+rax*8]
- sub rax, r8
- mov r12, rax
-
-rx_i_336: ;ROR_64
- dec ebx
- jz rx_finish
- xor r15, 0aea0a435h
- mov eax, r15d
- test bl, 63
- jnz short rx_body_336
- call rx_read
-rx_body_336:
- xor rbp, rax
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- ror rax, 42
- mov rcx, rax
- mov eax, r11d
- xor eax, 02644c5ah
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_337: ;ADD_64
- dec ebx
- jz rx_finish
- xor r8, 03d6c4ab2h
- mov eax, r8d
- test bl, 63
- jnz short rx_body_337
- call rx_read
-rx_body_337:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- add rax, r12
- mov rcx, rax
- mov eax, r13d
- xor eax, 0dab07c39h
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_338: ;MUL_64
- dec ebx
- jz rx_finish
- xor r12, 0d428a742h
- mov eax, r12d
- test bl, 63
- jnz short rx_body_338
- call rx_read
-rx_body_338:
- xor rbp, rax
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- imul rax, r12
- mov rcx, rax
- mov eax, r11d
- xor eax, 0184d2abbh
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_339: ;FPADD
- dec ebx
- jz rx_finish
- xor r9, 04596ef73h
- mov eax, r9d
- test bl, 63
- jnz short rx_body_339
- call rx_read
-rx_body_339:
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- addpd xmm0, xmm6
- movaps xmm2, xmm0
-
-rx_i_340: ;FPADD
- dec ebx
- jz rx_finish
- xor r15, 0e51629cch
- mov eax, r15d
- test bl, 63
- jnz short rx_body_340
- call rx_read
-rx_body_340:
- and eax, 131071
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- addpd xmm0, xmm5
- movaps xmm5, xmm0
- mov eax, r13d
- xor eax, 038b653beh
- and eax, 2047
- movhpd qword ptr [rsi + rax * 8], xmm5
-
-rx_i_341: ;MUL_32
- dec ebx
- jz rx_finish
- xor r12, 019eb9ea5h
- mov eax, r12d
- test bl, 63
- jnz short rx_body_341
- call rx_read
-rx_body_341:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- mov ecx, eax
- mov eax, r15d
- imul rax, rcx
- mov r8, rax
-
-rx_i_342: ;FPSUB
- dec ebx
- jz rx_finish
- xor r9, 09ccc7abah
- mov eax, r9d
- test bl, 63
- jnz short rx_body_342
- call rx_read
-rx_body_342:
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- subpd xmm0, xmm2
- movaps xmm3, xmm0
-
-rx_i_343: ;XOR_64
- dec ebx
- jz rx_finish
- xor r14, 056f6cf0bh
- mov eax, r14d
- test bl, 63
- jnz short rx_body_343
- call rx_read
-rx_body_343:
- xor rbp, rax
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- xor rax, r13
- mov rcx, rax
- mov eax, r15d
- xor eax, 0d9a469a9h
- and eax, 32767
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_344: ;FPSUB
- dec ebx
- jz rx_finish
- xor r10, 03ef9bcc4h
- mov eax, r10d
- test bl, 63
- jnz short rx_body_344
- call rx_read
-rx_body_344:
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- subpd xmm0, xmm6
- movaps xmm5, xmm0
- mov eax, r13d
- xor eax, 0627d9feah
- and eax, 2047
- movlpd qword ptr [rsi + rax * 8], xmm5
-
-rx_i_345: ;MULH_64
- dec ebx
- jz rx_finish
- xor r12, 0bbbcdbach
- mov eax, r12d
- test bl, 63
- jnz short rx_body_345
- call rx_read
-rx_body_345:
- and eax, 131071
- mov rax, qword ptr [rsi+rax*8]
- mov rcx, r13
- mul rcx
- mov rax, rdx
- mov rcx, rax
- mov eax, r9d
- xor eax, 0ef03b0ddh
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_346: ;AND_32
- dec ebx
- jz rx_finish
- xor r12, 0ae9d1e96h
- mov eax, r12d
- test bl, 63
- jnz short rx_body_346
- call rx_read
-rx_body_346:
- and eax, 32767
- mov rax, qword ptr [rsi+rax*8]
- and eax, r15d
- mov r13, rax
-
-rx_i_347: ;ADD_64
- dec ebx
- jz rx_finish
- xor r14, 070c34d69h
- mov eax, r14d
- test bl, 63
- jnz short rx_body_347
- call rx_read
-rx_body_347:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- add rax, r10
- mov rcx, rax
- mov eax, r13d
- xor eax, 0d529429ah
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_348: ;FPSUB
- dec ebx
- jz rx_finish
- xor r13, 0523ff904h
- mov eax, r13d
- test bl, 63
- jnz short rx_body_348
- call rx_read
-rx_body_348:
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- subpd xmm0, xmm3
- movaps xmm9, xmm0
- mov eax, r9d
- xor eax, 039c35461h
- and eax, 2047
- movhpd qword ptr [rsi + rax * 8], xmm9
-
-rx_i_349: ;OR_64
- dec ebx
- jz rx_finish
- xor r8, 018e0e5ddh
- mov eax, r8d
- test bl, 63
- jnz short rx_body_349
- call rx_read
-rx_body_349:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- or rax, r15
- mov rcx, rax
- mov eax, r13d
- xor eax, 05c449453h
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_350: ;CALL
- dec ebx
- jz rx_finish
- xor r9, 09bd050f0h
- mov eax, r9d
- test bl, 63
- jnz short rx_body_350
- call rx_read
-rx_body_350:
- and eax, 32767
- mov rax, qword ptr [rsi+rax*8]
- mov r12, rax
- cmp r9d, -980411581
- ja short rx_i_351
- call rx_i_352
-
-rx_i_351: ;MUL_64
- dec ebx
- jz rx_finish
- xor r11, 0a3a5906fh
- mov eax, r11d
- test bl, 63
- jnz short rx_body_351
- call rx_read
-rx_body_351:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- imul rax, r10
- mov rcx, rax
- mov eax, r13d
- xor eax, 0985ba4h
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_352: ;FPADD
- dec ebx
- jz rx_finish
- xor r10, 0afc9af2bh
- mov eax, r10d
- test bl, 63
- jnz short rx_body_352
- call rx_read
-rx_body_352:
- xor rbp, rax
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- addpd xmm0, xmm6
- movaps xmm2, xmm0
- mov eax, r10d
- xor eax, 03bf686f2h
- and eax, 2047
- movlpd qword ptr [rsi + rax * 8], xmm2
-
-rx_i_353: ;FPSUB
- dec ebx
- jz rx_finish
- xor r13, 02e65278bh
- mov eax, r13d
- test bl, 63
- jnz short rx_body_353
- call rx_read
-rx_body_353:
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- subpd xmm0, xmm2
- movaps xmm7, xmm0
-
-rx_i_354: ;MUL_32
- dec ebx
- jz rx_finish
- xor r13, 02412fc10h
- mov eax, r13d
- test bl, 63
- jnz short rx_body_354
- call rx_read
-rx_body_354:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- mov ecx, eax
- mov eax, r13d
- imul rax, rcx
- mov rcx, rax
- mov eax, r13d
- xor eax, 049cc2e0ch
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_355: ;MUL_64
- dec ebx
- jz rx_finish
- xor r10, 06bd6e65fh
- mov eax, r10d
- test bl, 63
- jnz short rx_body_355
- call rx_read
-rx_body_355:
- and eax, 32767
- mov rax, qword ptr [rsi+rax*8]
- imul rax, r14
- mov r8, rax
-
-rx_i_356: ;MUL_64
- dec ebx
- jz rx_finish
- xor r10, 01cd85d80h
- mov eax, r10d
- test bl, 63
- jnz short rx_body_356
- call rx_read
-rx_body_356:
- xor rbp, rax
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- imul rax, r10
- mov r11, rax
-
-rx_i_357: ;ADD_64
- dec ebx
- jz rx_finish
- xor r10, 0f7daed36h
- mov eax, r10d
- test bl, 63
- jnz short rx_body_357
- call rx_read
-rx_body_357:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- add rax, r11
- mov r11, rax
-
-rx_i_358: ;DIV_64
- dec ebx
- jz rx_finish
- xor r13, 088fa6e5ah
- mov eax, r13d
- test bl, 63
- jnz short rx_body_358
- call rx_read
-rx_body_358:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- ; magic divide by 3667831238
- shr rax, 1
- mov rcx, 2700102505175032865
- mul rcx
- mov rax, rdx
- shr rax, 28
- mov r9, rax
-
-rx_i_359: ;FPSUB
- dec ebx
- jz rx_finish
- xor r10, 0714fc2cdh
- mov eax, r10d
- test bl, 63
- jnz short rx_body_359
- call rx_read
-rx_body_359:
- xor rbp, rax
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- subpd xmm0, xmm9
- movaps xmm4, xmm0
- mov eax, r12d
- xor eax, 0f16b9be3h
- and eax, 32767
- movlpd qword ptr [rsi + rax * 8], xmm4
-
-rx_i_360: ;FPMUL
- dec ebx
- jz rx_finish
- xor r10, 0c2d110b5h
- mov eax, r10d
- test bl, 63
- jnz short rx_body_360
- call rx_read
-rx_body_360:
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- mulpd xmm0, xmm8
- movaps xmm1, xmm0
- cmpeqpd xmm1, xmm1
- andps xmm0, xmm1
- movaps xmm8, xmm0
- mov eax, r8d
- xor eax, 0c41a4103h
- and eax, 131071
- movlpd qword ptr [rsi + rax * 8], xmm8
-
-rx_i_361: ;FPDIV
- dec ebx
- jz rx_finish
- xor r15, 01d125a7fh
- mov eax, r15d
- test bl, 63
- jnz short rx_body_361
- call rx_read
-rx_body_361:
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- divpd xmm0, xmm6
- movaps xmm1, xmm0
- cmpeqpd xmm1, xmm1
- andps xmm0, xmm1
- movaps xmm6, xmm0
-
-rx_i_362: ;SUB_64
- dec ebx
- jz rx_finish
- xor r9, 0ed8954bdh
- mov eax, r9d
- test bl, 63
- jnz short rx_body_362
- call rx_read
-rx_body_362:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- sub rax, r9
- mov rcx, rax
- mov eax, r15d
- xor eax, 04080bf8dh
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_363: ;FPMUL
- dec ebx
- jz rx_finish
- xor r12, 09f75887bh
- mov eax, r12d
- test bl, 63
- jnz short rx_body_363
- call rx_read
-rx_body_363:
- and eax, 32767
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- mulpd xmm0, xmm6
- movaps xmm1, xmm0
- cmpeqpd xmm1, xmm1
- andps xmm0, xmm1
- movaps xmm3, xmm0
-
-rx_i_364: ;MUL_32
- dec ebx
- jz rx_finish
- xor r11, 0badaf867h
- mov eax, r11d
- test bl, 63
- jnz short rx_body_364
- call rx_read
-rx_body_364:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- mov ecx, eax
- mov eax, r8d
- imul rax, rcx
- mov rcx, rax
- mov eax, r8d
- xor eax, 0bb8ee9ch
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_365: ;IMUL_32
- dec ebx
- jz rx_finish
- xor r15, 02db4444ah
- mov eax, r15d
- test bl, 63
- jnz short rx_body_365
- call rx_read
-rx_body_365:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- movsxd rcx, eax
- movsxd rax, r9d
- imul rax, rcx
- mov r12, rax
-
-rx_i_366: ;IMUL_32
- dec ebx
- jz rx_finish
- xor r12, 0bff7218fh
- mov eax, r12d
- test bl, 63
- jnz short rx_body_366
- call rx_read
-rx_body_366:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- movsxd rcx, eax
- movsxd rax, r8d
- imul rax, rcx
- mov r15, rax
-
-rx_i_367: ;ROR_64
- dec ebx
- jz rx_finish
- xor r9, 04d14cb3ah
- mov eax, r9d
- test bl, 63
- jnz short rx_body_367
- call rx_read
-rx_body_367:
- xor rbp, rax
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- ror rax, 18
- mov rcx, rax
- mov eax, r12d
- xor eax, 0ad9b92e8h
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_368: ;SUB_32
- dec ebx
- jz rx_finish
- xor r10, 0a14836bah
- mov eax, r10d
- test bl, 63
- jnz short rx_body_368
- call rx_read
-rx_body_368:
- and eax, 32767
- mov rax, qword ptr [rsi+rax*8]
- sub eax, r10d
- mov r8, rax
-
-rx_i_369: ;IDIV_64
- dec ebx
- jz rx_finish
- xor r9, 053fe22e2h
- mov eax, r9d
- test bl, 63
- jnz short rx_body_369
- call rx_read
-rx_body_369:
- and eax, 32767
- mov rax, qword ptr [rsi+rax*8]
- ; magic divide by 470792991
- mov rdx, 1314739240972876203
- imul rdx
- mov rax, rdx
- xor edx, edx
- sar rax, 25
- sets dl
- add rax, rdx
- mov r9, rax
-
-rx_i_370: ;FPSUB
- dec ebx
- jz rx_finish
- xor r15, 010e1fb24h
- mov eax, r15d
- test bl, 63
- jnz short rx_body_370
- call rx_read
-rx_body_370:
- and eax, 32767
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- subpd xmm0, xmm6
- movaps xmm6, xmm0
-
-rx_i_371: ;FPADD
- dec ebx
- jz rx_finish
- xor r8, 0ebbd5cc9h
- mov eax, r8d
- test bl, 63
- jnz short rx_body_371
- call rx_read
-rx_body_371:
- xor rbp, rax
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
+ and eax, 16376
+ imul r10, qword ptr [rsi+rax]
+ ; IMUL_R r6, r0
+ imul r14, r8
+ ; IADD_R r7, r6
+ add r15, r14
+ ; FPSUB_R f2, a3
+ subpd xmm2, xmm11
+ ; COND_R r5, no(r2, -1589295370)
+ xor ecx, ecx
+ cmp r10d, -1589295370
+ setno cl
+ add r13, rcx
+ ; IMUL_9C r7, 420978486
+ lea r15, [r15+r15*8+420978486]
+ ; IROL_R r4, r2
+ mov ecx, r10d
+ rol r12, cl
+ ; IMUL_9C r0, -1084530831
+ lea r8, [r8+r8*8-1084530831]
+ ; FPNEG_R f3
+ xorps xmm3, xmm15
+ ; IROR_R r6, r4
+ mov ecx, r12d
+ ror r14, cl
+ ; IROL_R r4, r5
+ mov ecx, r13d
+ rol r12, cl
+ ; FPSUB_R f2, a3
+ subpd xmm2, xmm11
+ ; FPMUL_R e2, a2
+ mulpd xmm6, xmm10
+ ; ISMULH_M r6, L2[98600]
+ mov rax, r14
+ imul qword ptr [rsi+98600]
+ mov r14, rdx
+ ; IXOR_R r0, r6
+ xor r8, r14
+ ; FPSWAP_R f1
+ shufpd xmm1, xmm1, 1
+ ; FPADD_R f0, a1
addpd xmm0, xmm9
- movaps xmm5, xmm0
- mov eax, r13d
- xor eax, 0c40fe413h
- and eax, 2047
- movhpd qword ptr [rsi + rax * 8], xmm5
-
-rx_i_372: ;SHL_64
- dec ebx
- jz rx_finish
- xor r10, 098ab79d7h
- mov eax, r10d
- test bl, 63
- jnz short rx_body_372
- call rx_read
-rx_body_372:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- mov rcx, r13
- shl rax, cl
- mov r9, rax
-
-rx_i_373: ;FPMUL
- dec ebx
- jz rx_finish
- xor r15, 056438b3h
- mov eax, r15d
- test bl, 63
- jnz short rx_body_373
- call rx_read
-rx_body_373:
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- mulpd xmm0, xmm8
- movaps xmm1, xmm0
- cmpeqpd xmm1, xmm1
- andps xmm0, xmm1
- movaps xmm4, xmm0
-
-rx_i_374: ;FPMUL
- dec ebx
- jz rx_finish
- xor r11, 0dbcce604h
- mov eax, r11d
- test bl, 63
- jnz short rx_body_374
- call rx_read
-rx_body_374:
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- mulpd xmm0, xmm2
- movaps xmm1, xmm0
- cmpeqpd xmm1, xmm1
- andps xmm0, xmm1
- movaps xmm2, xmm0
- mov eax, r10d
- xor eax, 03507e810h
- and eax, 2047
- movlpd qword ptr [rsi + rax * 8], xmm2
-
-rx_i_375: ;ADD_64
- dec ebx
- jz rx_finish
- xor r9, 0edea6200h
+ ; COND_R r1, ab(r3, -991705199)
+ xor ecx, ecx
+ cmp r11d, -991705199
+ seta cl
+ add r9, rcx
+ ; IMULH_M r4, L2[r2]
+ mov ecx, r10d
+ and ecx, 262136
+ mov rax, r12
+ mul qword ptr [rsi+rcx]
+ mov r12, rdx
+ ; IROR_R r2, r6
+ mov ecx, r14d
+ ror r10, cl
+ ; FPDIV_R e0, a1
+ divpd xmm4, xmm9
+ maxpd xmm4, xmm13
+ ; IMUL_R r1, r7
+ imul r9, r15
+ ; COND_R r6, ns(r2, 939392855)
+ xor ecx, ecx
+ cmp r10d, 939392855
+ setns cl
+ add r14, rcx
+ ; FPMUL_R e3, a1
+ mulpd xmm7, xmm9
+ ; COND_R r2, ab(r2, -499266314)
+ xor ecx, ecx
+ cmp r10d, -499266314
+ seta cl
+ add r10, rcx
+ ; COND_M r7, lt(L1[r1], -1624420482)
+ xor ecx, ecx
mov eax, r9d
- test bl, 63
- jnz short rx_body_375
- call rx_read
-rx_body_375:
- and eax, 131071
- mov rax, qword ptr [rsi+rax*8]
- add rax, -332030999
- mov r12, rax
-
-rx_i_376: ;ADD_64
- dec ebx
- jz rx_finish
- xor r14, 05e61b279h
- mov eax, r14d
- test bl, 63
- jnz short rx_body_376
- call rx_read
-rx_body_376:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- add rax, r9
- mov rcx, rax
- mov eax, r8d
- xor eax, 01c614282h
- and eax, 32767
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_377: ;FPSUB
- dec ebx
- jz rx_finish
- xor r14, 0fc1fb433h
- mov eax, r14d
- test bl, 63
- jnz short rx_body_377
- call rx_read
-rx_body_377:
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- subpd xmm0, xmm3
- movaps xmm7, xmm0
-
-rx_i_378: ;MUL_32
- dec ebx
- jz rx_finish
- xor r12, 082aa21ach
+ and eax, 16376
+ cmp dword ptr [rsi+rax], -1624420482
+ setl cl
+ add r15, rcx
+ ; COND_R r1, lt(r1, 1525413977)
+ xor ecx, ecx
+ cmp r9d, 1525413977
+ setl cl
+ add r9, rcx
+ ; IMUL_R r4, r5
+ imul r12, r13
+ ; IMUL_R r4, r2
+ imul r12, r10
+ ; FPSQRT_R e1
+ sqrtpd xmm5, xmm5
+ ; ISUB_R r2, r6
+ sub r10, r14
+ ; FPDIV_R e1, a0
+ divpd xmm5, xmm8
+ maxpd xmm5, xmm13
+ ; FPMUL_R e2, a3
+ mulpd xmm6, xmm11
+ ; IADD_R r6, 671627590
+ add r14, 671627590
+ ; COND_M r6, sg(L1[r4], -780452820)
+ xor ecx, ecx
mov eax, r12d
- test bl, 63
- jnz short rx_body_378
- call rx_read
-rx_body_378:
- and eax, 131071
- mov rax, qword ptr [rsi+rax*8]
- mov ecx, eax
- mov eax, r14d
- imul rax, rcx
- mov r15, rax
-
-rx_i_379: ;ROR_64
- dec ebx
- jz rx_finish
- xor r10, 05dba41fbh
- mov eax, r10d
- test bl, 63
- jnz short rx_body_379
- call rx_read
-rx_body_379:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- mov rcx, r9
- ror rax, cl
- mov rcx, rax
- mov eax, r13d
- xor eax, 03a2dc429h
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_380: ;MUL_64
- dec ebx
- jz rx_finish
- xor r11, 0229e3d6eh
- mov eax, r11d
- test bl, 63
- jnz short rx_body_380
- call rx_read
-rx_body_380:
- and eax, 32767
- mov rax, qword ptr [rsi+rax*8]
- imul rax, r10
- mov rcx, rax
- mov eax, r13d
- xor eax, 0a9fd85e0h
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_381: ;XOR_32
- dec ebx
- jz rx_finish
- xor r8, 019816ff9h
- mov eax, r8d
- test bl, 63
- jnz short rx_body_381
- call rx_read
-rx_body_381:
- xor rbp, rax
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- xor eax, r14d
- mov rcx, rax
- mov eax, r9d
- xor eax, 032349ff8h
- and eax, 32767
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_382: ;ROL_64
- dec ebx
- jz rx_finish
- xor r14, 036b5b81fh
- mov eax, r14d
- test bl, 63
- jnz short rx_body_382
- call rx_read
-rx_body_382:
- and eax, 32767
- mov rax, qword ptr [rsi+rax*8]
- rol rax, 55
- mov rcx, rax
- mov eax, r11d
- xor eax, 0a6a2e0b1h
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_383: ;FPSUB
- dec ebx
- jz rx_finish
- xor r15, 05f798ec3h
- mov eax, r15d
- test bl, 63
- jnz short rx_body_383
- call rx_read
-rx_body_383:
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- subpd xmm0, xmm4
- movaps xmm5, xmm0
-
-rx_i_384: ;XOR_64
- dec ebx
- jz rx_finish
- xor r10, 05b459fd7h
- mov eax, r10d
- test bl, 63
- jnz short rx_body_384
- call rx_read
-rx_body_384:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- xor rax, 1413715044
- mov r9, rax
-
-rx_i_385: ;MUL_64
- dec ebx
- jz rx_finish
- xor r15, 0c91749bbh
- mov eax, r15d
- test bl, 63
- jnz short rx_body_385
- call rx_read
-rx_body_385:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- imul rax, r12
- mov rcx, rax
- mov eax, r13d
- xor eax, 0fb9b50b9h
- and eax, 32767
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_386: ;FPADD
- dec ebx
- jz rx_finish
- xor r9, 0575b4bdch
- mov eax, r9d
- test bl, 63
- jnz short rx_body_386
- call rx_read
-rx_body_386:
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
+ and eax, 16376
+ cmp dword ptr [rsi+rax], -780452820
+ sets cl
+ add r14, rcx
+ ; IMULH_R r4, r7
+ mov rax, r12
+ mul r15
+ mov r12, rdx
+ ; FPMUL_R e3, a1
+ mulpd xmm7, xmm9
+ ; FPADD_R f0, a0
addpd xmm0, xmm8
- movaps xmm9, xmm0
-
-rx_i_387: ;SUB_32
- dec ebx
- jz rx_finish
- xor r9, 0d4f7bc6ah
- mov eax, r9d
- test bl, 63
- jnz short rx_body_387
- call rx_read
-rx_body_387:
- xor rbp, rax
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- sub eax, r15d
- mov rcx, rax
- mov eax, r9d
- xor eax, 028cbb7adh
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_388: ;RET
- dec ebx
- jz rx_finish
- xor r8, 08a949356h
- mov eax, r8d
- test bl, 63
- jnz short rx_body_388
- call rx_read
-rx_body_388:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- mov r11, rax
- cmp rsp, rdi
- je short rx_i_389
- ret
-
-rx_i_389: ;JUMP
- dec ebx
- jz rx_finish
- xor r11, 06531ad2eh
- mov eax, r11d
- test bl, 63
- jnz short rx_body_389
- call rx_read
-rx_body_389:
- and eax, 32767
- mov rax, qword ptr [rsi+rax*8]
- mov rcx, rax
- mov eax, r14d
- xor eax, 0eb1a1f50h
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
- cmp r9d, -350609584
- jge rx_i_421
-
-rx_i_390: ;FPADD
- dec ebx
- jz rx_finish
- xor r15, 02914abeah
- mov eax, r15d
- test bl, 63
- jnz short rx_body_390
- call rx_read
-rx_body_390:
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- addpd xmm0, xmm4
- movaps xmm3, xmm0
- mov eax, r11d
- xor eax, 0e5c5acbbh
- and eax, 2047
- movlpd qword ptr [rsi + rax * 8], xmm3
-
-rx_i_391: ;FPADD
- dec ebx
- jz rx_finish
- xor r8, 0473a41f0h
- mov eax, r8d
- test bl, 63
- jnz short rx_body_391
- call rx_read
-rx_body_391:
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- addpd xmm0, xmm3
- movaps xmm6, xmm0
-
-rx_i_392: ;SAR_64
- dec ebx
- jz rx_finish
- xor r14, 01ebc1f0dh
- mov eax, r14d
- test bl, 63
- jnz short rx_body_392
- call rx_read
-rx_body_392:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- mov rcx, r9
- sar rax, cl
- mov r13, rax
-
-rx_i_393: ;AND_64
- dec ebx
- jz rx_finish
- xor r14, 0742e95b1h
- mov eax, r14d
- test bl, 63
- jnz short rx_body_393
- call rx_read
-rx_body_393:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- and rax, r12
- mov r13, rax
-
-rx_i_394: ;FPADD
- dec ebx
- jz rx_finish
- xor r12, 0db885c2ch
- mov eax, r12d
- test bl, 63
- jnz short rx_body_394
- call rx_read
-rx_body_394:
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- addpd xmm0, xmm9
- movaps xmm6, xmm0
-
-rx_i_395: ;DIV_64
- dec ebx
- jz rx_finish
- xor r8, 04ae4fe8ch
- mov eax, r8d
- test bl, 63
- jnz short rx_body_395
- call rx_read
-rx_body_395:
- xor rbp, rax
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- ; magic divide by 939698704
- mov rcx, 5269518980991934091
- mul rcx
- mov rax, rdx
- shr rax, 28
- mov rcx, rax
- mov eax, r8d
- xor eax, 03802aa10h
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_396: ;ROR_64
- dec ebx
- jz rx_finish
- xor r10, 07b41862bh
+ ; FPMUL_R e0, a1
+ mulpd xmm4, xmm9
+ ; IMUL_R r7, r3
+ imul r15, r11
+ ; IROL_R r0, r7
+ mov ecx, r15d
+ rol r8, cl
+ ; IMUL_R r1, r7
+ imul r9, r15
+ ; COND_R r0, no(r7, 449007464)
+ xor ecx, ecx
+ cmp r15d, 449007464
+ setno cl
+ add r8, rcx
+ ; ISMULH_M r6, L2[134288]
+ mov rax, r14
+ imul qword ptr [rsi+134288]
+ mov r14, rdx
+ ; IMULH_R r5, r2
+ mov rax, r13
+ mul r10
+ mov r13, rdx
+ ; IMULH_R r7, r4
+ mov rax, r15
+ mul r12
+ mov r15, rdx
+ ; FPDIV_R e3, a0
+ divpd xmm7, xmm8
+ maxpd xmm7, xmm13
+ ; IXOR_R r3, r4
+ xor r11, r12
+ ; IDIV_C r1, 72349044
+ mov rax, 8555331009525020641
+ mul r9
+ shr rdx, 25
+ add r9, rdx
+ ; IADD_R r5, r4
+ add r13, r12
+ ; IROR_R r2, r4
+ mov ecx, r12d
+ ror r10, cl
+ ; FPSUB_M f1, L1[r2]
mov eax, r10d
- test bl, 63
- jnz short rx_body_396
- call rx_read
-rx_body_396:
- and eax, 32767
- mov rax, qword ptr [rsi+rax*8]
- ror rax, 62
- mov rcx, rax
+ and eax, 16376
+ cvtdq2pd xmm12, qword ptr [rsi+rax]
+ subpd xmm1, xmm12
+ ; FPMUL_R e2, a3
+ mulpd xmm6, xmm11
+ ; IADD_R r5, r6
+ add r13, r14
+ ; IXOR_M r1, L1[r4]
mov eax, r12d
- xor eax, 01ee1c837h
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_397: ;SUB_32
- dec ebx
- jz rx_finish
- xor r8, 0916f3819h
+ and eax, 16376
+ xor r9, qword ptr [rsi+rax]
+ ; ISUB_R r2, -1544880589
+ sub r10, -1544880589
+ ; FPNEG_R f0
+ xorps xmm0, xmm15
+ ; IROR_R r1, r6
+ mov ecx, r14d
+ ror r9, cl
+ ; IMUL_R r6, r4
+ imul r14, r12
+ ; IMULH_M r4, L2[r1]
+ mov ecx, r9d
+ and ecx, 262136
+ mov rax, r12
+ mul qword ptr [rsi+rcx]
+ mov r12, rdx
+ ; IXOR_R r3, r0
+ xor r11, r8
+ ; FPSWAP_R f0
+ shufpd xmm0, xmm0, 1
+ ; FPSWAP_R f0
+ shufpd xmm0, xmm0, 1
+ ; COND_R r0, ns(r2, -308295242)
+ xor ecx, ecx
+ cmp r10d, -308295242
+ setns cl
+ add r8, rcx
+ ; IMUL_9C r1, 591587965
+ lea r9, [r9+r9*8+591587965]
+ ; FPADD_R f3, a1
+ addpd xmm3, xmm9
+ ; IMUL_R r5, r4
+ imul r13, r12
+ ; IMUL_M r7, L1[r0]
mov eax, r8d
- test bl, 63
- jnz short rx_body_397
- call rx_read
-rx_body_397:
- xor rbp, rax
- and eax, 32767
- mov rax, qword ptr [rsi+rax*8]
- sub eax, r12d
- mov r10, rax
-
-rx_i_398: ;SHR_64
- dec ebx
- jz rx_finish
- xor r8, 04eb6fd2ah
- mov eax, r8d
- test bl, 63
- jnz short rx_body_398
- call rx_read
-rx_body_398:
- and eax, 32767
- mov rax, qword ptr [rsi+rax*8]
- mov rcx, r8
- shr rax, cl
- mov rcx, rax
- mov eax, r11d
- xor eax, 0724e7136h
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_399: ;FPMUL
- dec ebx
- jz rx_finish
- xor r11, 0899a98cfh
- mov eax, r11d
- test bl, 63
- jnz short rx_body_399
- call rx_read
-rx_body_399:
- xor rbp, rax
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- mulpd xmm0, xmm2
- movaps xmm1, xmm0
- cmpeqpd xmm1, xmm1
- andps xmm0, xmm1
- movaps xmm6, xmm0
-
-rx_i_400: ;AND_64
- dec ebx
- jz rx_finish
- xor r13, 0aae75db6h
- mov eax, r13d
- test bl, 63
- jnz short rx_body_400
- call rx_read
-rx_body_400:
- and eax, 32767
- mov rax, qword ptr [rsi+rax*8]
- and rax, -1800645748
- mov r14, rax
-
-rx_i_401: ;FPSUB
- dec ebx
- jz rx_finish
- xor r13, 032e81f25h
- mov eax, r13d
- test bl, 63
- jnz short rx_body_401
- call rx_read
-rx_body_401:
- and eax, 32767
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- subpd xmm0, xmm4
- movaps xmm6, xmm0
- mov eax, r14d
- xor eax, 03ea60344h
- and eax, 32767
- movlpd qword ptr [rsi + rax * 8], xmm6
-
-rx_i_402: ;RET
- dec ebx
- jz rx_finish
- xor r9, 0fa1a07ffh
+ and eax, 16376
+ imul r15, qword ptr [rsi+rax]
+ ; COND_R r6, sg(r5, -1119525789)
+ xor ecx, ecx
+ cmp r13d, -1119525789
+ sets cl
+ add r14, rcx
+ ; IMUL_M r0, L1[r1]
mov eax, r9d
- test bl, 63
- jnz short rx_body_402
- call rx_read
-rx_body_402:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- mov r14, rax
- cmp rsp, rdi
- je short rx_i_403
- ret
-
-rx_i_403: ;DIV_64
- dec ebx
- jz rx_finish
- xor r9, 0e59500f7h
- mov eax, r9d
- test bl, 63
- jnz short rx_body_403
- call rx_read
-rx_body_403:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- ; magic divide by 536056992
- mov rcx, 4618688153536407095
- mul rcx
- mov rax, rdx
- shr rax, 27
- mov rcx, rax
- mov eax, r11d
- xor eax, 01ff394a0h
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_404: ;MUL_32
- dec ebx
- jz rx_finish
- xor r15, 05b8ceb2fh
+ and eax, 16376
+ imul r8, qword ptr [rsi+rax]
+ ; IADD_M r3, L2[r7]
mov eax, r15d
- test bl, 63
- jnz short rx_body_404
- call rx_read
-rx_body_404:
- and eax, 131071
- mov rax, qword ptr [rsi+rax*8]
- mov ecx, eax
- mov eax, r8d
- imul rax, rcx
- mov rcx, rax
+ and eax, 262136
+ add r11, qword ptr [rsi+rax]
+ ; IADD_R r0, r1
+ add r8, r9
+ ; FPSUB_R f2, a1
+ subpd xmm2, xmm9
+ ; IXOR_M r0, L2[r7]
mov eax, r15d
- xor eax, 08f83c4f1h
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_405: ;CALL
- dec ebx
- jz rx_finish
- xor r8, 0f61082a3h
- mov eax, r8d
- test bl, 63
- jnz short rx_body_405
- call rx_read
-rx_body_405:
- and eax, 32767
- mov rax, qword ptr [rsi+rax*8]
- mov r12, rax
- cmp r10d, 1795880641
- ja short rx_i_406
- call rx_i_494
-
-rx_i_406: ;FPDIV
- dec ebx
- jz rx_finish
- xor r9, 0af6886b7h
- mov eax, r9d
- test bl, 63
- jnz short rx_body_406
- call rx_read
-rx_body_406:
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- divpd xmm0, xmm7
- movaps xmm1, xmm0
- cmpeqpd xmm1, xmm1
- andps xmm0, xmm1
- movaps xmm9, xmm0
-
-rx_i_407: ;FPSUB
- dec ebx
- jz rx_finish
- xor r14, 09699566fh
- mov eax, r14d
- test bl, 63
- jnz short rx_body_407
- call rx_read
-rx_body_407:
- xor rbp, rax
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
+ and eax, 262136
+ xor r8, qword ptr [rsi+rax]
+ ; COND_R r6, be(r6, 1481939391)
+ xor ecx, ecx
+ cmp r14d, 1481939391
+ setbe cl
+ add r14, rcx
+ ; FPADD_R f0, a1
+ addpd xmm0, xmm9
+ ; IXOR_R r3, r2
+ xor r11, r10
+ ; FPSUB_R f0, a1
subpd xmm0, xmm9
- movaps xmm8, xmm0
-
-rx_i_408: ;MUL_64
- dec ebx
- jz rx_finish
- xor r15, 066e79fa6h
+ ; IXOR_R r7, r3
+ xor r15, r11
+ ; IXOR_M r6, L1[r4]
+ mov eax, r12d
+ and eax, 16376
+ xor r14, qword ptr [rsi+rax]
+ ; IMULH_R r2, r7
+ mov rax, r10
+ mul r15
+ mov r10, rdx
+ ; ISUB_R r5, r1
+ sub r13, r9
+ ; FPMUL_R e1, a3
+ mulpd xmm5, xmm11
+ ; FPADD_R f3, a2
+ addpd xmm3, xmm10
+ ; FPSWAP_R f1
+ shufpd xmm1, xmm1, 1
+ ; FPSUB_R f1, a3
+ subpd xmm1, xmm11
+ ; FPSUB_M f0, L1[r4]
+ mov eax, r12d
+ and eax, 16376
+ cvtdq2pd xmm12, qword ptr [rsi+rax]
+ subpd xmm0, xmm12
+ ; FPMUL_R e1, a2
+ mulpd xmm5, xmm10
+ ; FPADD_R f3, a0
+ addpd xmm3, xmm8
+ ; IROL_R r2, r4
+ mov ecx, r12d
+ rol r10, cl
+ ; COND_M r7, ab(L2[r7], -2012390318)
+ xor ecx, ecx
mov eax, r15d
- test bl, 63
- jnz short rx_body_408
- call rx_read
-rx_body_408:
- and eax, 131071
- mov rax, qword ptr [rsi+rax*8]
- imul rax, rax, 693109961
- mov r10, rax
-
-rx_i_409: ;MUL_64
- dec ebx
- jz rx_finish
- xor r11, 04b6caa9ah
- mov eax, r11d
- test bl, 63
- jnz short rx_body_409
- call rx_read
-rx_body_409:
- xor rbp, rax
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- imul rax, r15
- mov rcx, rax
+ and eax, 262136
+ cmp dword ptr [rsi+rax], -2012390318
+ seta cl
+ add r15, rcx
+ ; IMUL_9C r4, -38079585
+ lea r12, [r12+r12*8-38079585]
+ ; IXOR_R r0, r1
+ xor r8, r9
+ ; FPMUL_R e1, a3
+ mulpd xmm5, xmm11
+ ; FPMUL_R e1, a1
+ mulpd xmm5, xmm9
+ ; FPSUB_R f1, a2
+ subpd xmm1, xmm10
+ ; IMUL_9C r4, -847745598
+ lea r12, [r12+r12*8-847745598]
+ ; FPSQRT_R e1
+ sqrtpd xmm5, xmm5
+ ; IADD_R r7, r6
+ add r15, r14
+ ; FPSUB_R f3, a0
+ subpd xmm3, xmm8
+ ; FPSUB_R f1, a1
+ subpd xmm1, xmm9
+ ; IADD_R r7, r6
+ add r15, r14
+ ; IROL_R r2, r5
+ mov ecx, r13d
+ rol r10, cl
+ ; IADD_RC r4, r2, 1338806320
+ lea r12, [r12+r10+1338806320]
+ ; FPSQRT_R e3
+ sqrtpd xmm7, xmm7
+ ; IMUL_R r5, r0
+ imul r13, r8
+ ; FPADD_R f2, a1
+ addpd xmm2, xmm9
+ ; INEG_R r6
+ neg r14
+ ; IXOR_M r6, L1[r2]
+ mov eax, r10d
+ and eax, 16376
+ xor r14, qword ptr [rsi+rax]
+ ; FPSUB_R f2, a2
+ subpd xmm2, xmm10
+ ; FPADD_R f2, a2
+ addpd xmm2, xmm10
+ ; FPADD_R f1, a2
+ addpd xmm1, xmm10
+ ; COND_R r3, be(r4, 174667458)
+ xor ecx, ecx
+ cmp r12d, 174667458
+ setbe cl
+ add r11, rcx
+ ; INEG_R r6
+ neg r14
+ ; IXOR_R r6, r3
+ xor r14, r11
+ ; COND_M r5, sg(L1[r0], -864345921)
+ xor ecx, ecx
mov eax, r8d
- xor eax, 05a68b80fh
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_410: ;RET
- dec ebx
- jz rx_finish
- xor r15, 0d17f245eh
+ and eax, 16376
+ cmp dword ptr [rsi+rax], -864345921
+ sets cl
+ add r13, rcx
+ ; IROL_R r7, r3
+ mov ecx, r11d
+ rol r15, cl
+ ; FPSUB_R f1, a2
+ subpd xmm1, xmm10
+ ; IADD_M r1, L1[r0]
+ mov eax, r8d
+ and eax, 16376
+ add r9, qword ptr [rsi+rax]
+ ; IMULH_R r1, r3
+ mov rax, r9
+ mul r11
+ mov r9, rdx
+ ; IMUL_R r0, -1489192296
+ imul r8, -1489192296
+ ; FPMUL_R e0, a2
+ mulpd xmm4, xmm10
+ ; COND_R r1, ge(r1, -1358904097)
+ xor ecx, ecx
+ cmp r9d, -1358904097
+ setge cl
+ add r9, rcx
+ ; FPSUB_R f1, a1
+ subpd xmm1, xmm9
+ ; FPADD_R f2, a3
+ addpd xmm2, xmm11
+ ; IROR_R r4, r7
+ mov ecx, r15d
+ ror r12, cl
+ ; ISDIV_C r1, -1368098113
+ mov rax, -7238896260565957085
+ imul r9
+ xor eax, eax
+ sar rdx, 29
+ sets al
+ add rdx, rax
+ add r9, rdx
+ ; IADD_M r4, L1[r1]
+ mov eax, r9d
+ and eax, 16376
+ add r12, qword ptr [rsi+rax]
+ ; IMUL_R r0, -1011605520
+ imul r8, -1011605520
+ ; FPSUB_R f3, a1
+ subpd xmm3, xmm9
+ ; IADD_RC r1, r4, 272540736
+ lea r9, [r9+r12+272540736]
+ ; FPSWAP_R f2
+ shufpd xmm2, xmm2, 1
+ ; IROR_R r3, r2
+ mov ecx, r10d
+ ror r11, cl
+ ; IMUL_R r3, 2085105439
+ imul r11, 2085105439
+ ; FPMUL_R e0, a0
+ mulpd xmm4, xmm8
+ ; IMUL_9C r6, -483723153
+ lea r14, [r14+r14*8-483723153]
+ ; FPSUB_M f3, L1[r7]
mov eax, r15d
- test bl, 63
- jnz short rx_body_410
- call rx_read
-rx_body_410:
- and eax, 32767
- mov rax, qword ptr [rsi+rax*8]
- mov r8, rax
- cmp rsp, rdi
- je short rx_i_411
- ret
-
-rx_i_411: ;RET
- dec ebx
- jz rx_finish
- xor r12, 0364f10e7h
+ and eax, 16376
+ cvtdq2pd xmm12, qword ptr [rsi+rax]
+ subpd xmm3, xmm12
+ ; IMUL_R r3, r2
+ imul r11, r10
+ ; ISMULH_R r7, r1
+ mov rax, r15
+ imul r9
+ mov r15, rdx
+ ; COND_R r1, of(r7, 778804236)
+ xor ecx, ecx
+ cmp r15d, 778804236
+ seto cl
+ add r9, rcx
+ ; FPSUB_R f3, a2
+ subpd xmm3, xmm10
+ ; IROL_R r5, r7
+ mov ecx, r15d
+ rol r13, cl
+ ; FPADD_R f1, a0
+ addpd xmm1, xmm8
+ ; FPADD_R f2, a3
+ addpd xmm2, xmm11
+ ; IMUL_R r6, r0
+ imul r14, r8
+ ; ISUB_M r2, L2[r4]
mov eax, r12d
- test bl, 63
- jnz short rx_body_411
- call rx_read
-rx_body_411:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- mov rcx, rax
- mov eax, r12d
- xor eax, 0b492f6bah
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
- cmp rsp, rdi
- je short rx_i_412
- ret
-
-rx_i_412: ;FPDIV
- dec ebx
- jz rx_finish
- xor r10, 0ac90e7ah
- mov eax, r10d
- test bl, 63
- jnz short rx_body_412
- call rx_read
-rx_body_412:
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- divpd xmm0, xmm4
- movaps xmm1, xmm0
- cmpeqpd xmm1, xmm1
- andps xmm0, xmm1
- movaps xmm3, xmm0
-
-rx_i_413: ;FPMUL
- dec ebx
- jz rx_finish
- xor r11, 04b6037abh
- mov eax, r11d
- test bl, 63
- jnz short rx_body_413
- call rx_read
-rx_body_413:
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- mulpd xmm0, xmm2
- movaps xmm1, xmm0
- cmpeqpd xmm1, xmm1
- andps xmm0, xmm1
- movaps xmm4, xmm0
- mov eax, r12d
- xor eax, 043989376h
- and eax, 131071
- movlpd qword ptr [rsi + rax * 8], xmm4
-
-rx_i_414: ;AND_64
- dec ebx
- jz rx_finish
- xor r14, 06c01554dh
+ and eax, 262136
+ sub r10, qword ptr [rsi+rax]
+ ; IXOR_R r0, r6
+ xor r8, r14
+ ; INEG_R r6
+ neg r14
+ ; FPMUL_R e2, a3
+ mulpd xmm6, xmm11
+ ; IADD_RC r4, r6, -1312075035
+ lea r12, [r12+r14-1312075035]
+ ; IMUL_R r1, r5
+ imul r9, r13
+ ; IXOR_M r7, L2[r6]
mov eax, r14d
- test bl, 63
- jnz short rx_body_414
- call rx_read
-rx_body_414:
- xor rbp, rax
- and eax, 32767
- mov rax, qword ptr [rsi+rax*8]
- and rax, -378293327
- mov r10, rax
-
-rx_i_415: ;DIV_64
- dec ebx
- jz rx_finish
- xor r8, 08c3e59a1h
- mov eax, r8d
- test bl, 63
- jnz short rx_body_415
- call rx_read
-rx_body_415:
- xor rbp, rax
- and eax, 32767
- mov rax, qword ptr [rsi+rax*8]
- ; magic divide by 3756873911
- add rax, 1
- sbb rax, 0
- mov rcx, 10544426615208851175
- mul rcx
- mov rax, rdx
- shr rax, 31
- mov r9, rax
-
-rx_i_416: ;FPADD
- dec ebx
- jz rx_finish
- xor r12, 0f3fafde9h
+ and eax, 262136
+ xor r15, qword ptr [rsi+rax]
+ ; IROR_R r2, 23
+ ror r10, 23
+ ; FPMUL_R e0, a2
+ mulpd xmm4, xmm10
+ ; ISMULH_M r5, L1[r2]
+ mov ecx, r10d
+ and ecx, 16376
+ mov rax, r13
+ imul qword ptr [rsi+rcx]
+ mov r13, rdx
+ ; ISUB_M r7, L1[r4]
mov eax, r12d
- test bl, 63
- jnz short rx_body_416
- call rx_read
-rx_body_416:
- xor rbp, rax
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- addpd xmm0, xmm3
- movaps xmm5, xmm0
- mov eax, r13d
- xor eax, 0f84b5382h
- and eax, 2047
- movlpd qword ptr [rsi + rax * 8], xmm5
-
-rx_i_417: ;SUB_64
- dec ebx
- jz rx_finish
- xor r10, 03c6481fah
+ and eax, 16376
+ sub r15, qword ptr [rsi+rax]
+ ; COND_R r0, sg(r2, 1538841628)
+ xor ecx, ecx
+ cmp r10d, 1538841628
+ sets cl
+ add r8, rcx
+ ; IMUL_R r6, r2
+ imul r14, r10
+ ; ISUB_R r0, r1
+ sub r8, r9
+ ; IMUL_R r5, r7
+ imul r13, r15
+ ; IADD_RC r1, r0, 516706834
+ lea r9, [r9+r8+516706834]
+ ; INEG_R r5
+ neg r13
+ ; FPSQRT_R e3
+ sqrtpd xmm7, xmm7
+ ; IADD_RC r5, r4, -1679394922
+ lea r13, [r13+r12-1679394922]
+ ; FPSUB_R f1, a1
+ subpd xmm1, xmm9
+ ; IMUL_R r0, r2
+ imul r8, r10
+ ; ISUB_R r3, r2
+ sub r11, r10
+ ; FPDIV_R e0, a3
+ divpd xmm4, xmm11
+ maxpd xmm4, xmm13
+ ; ISUB_R r1, r5
+ sub r9, r13
+ ; COND_M r2, be(L2[r2], 1840094725)
+ xor ecx, ecx
mov eax, r10d
- test bl, 63
- jnz short rx_body_417
- call rx_read
-rx_body_417:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- sub rax, r12
- mov rcx, rax
- mov eax, r10d
- xor eax, 0dfa7569ch
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_418: ;MULH_64
- dec ebx
- jz rx_finish
- xor r10, 02bd61c5fh
- mov eax, r10d
- test bl, 63
- jnz short rx_body_418
- call rx_read
-rx_body_418:
- and eax, 131071
- mov rax, qword ptr [rsi+rax*8]
- mov rcx, r11
- mul rcx
- mov rax, rdx
- mov r10, rax
-
-rx_i_419: ;OR_64
- dec ebx
- jz rx_finish
- xor r9, 0b6ab9d32h
- mov eax, r9d
- test bl, 63
- jnz short rx_body_419
- call rx_read
-rx_body_419:
- and eax, 131071
- mov rax, qword ptr [rsi+rax*8]
- or rax, r14
- mov rcx, rax
- mov eax, r14d
- xor eax, 0beeca8dbh
- and eax, 32767
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_420: ;ROR_64
- dec ebx
- jz rx_finish
- xor r9, 0f9690ceah
- mov eax, r9d
- test bl, 63
- jnz short rx_body_420
- call rx_read
-rx_body_420:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- ror rax, 38
- mov rcx, rax
- mov eax, r9d
- xor eax, 08f7bb3ech
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_421: ;CALL
- dec ebx
- jz rx_finish
- xor r12, 01ada0f39h
- mov eax, r12d
- test bl, 63
- jnz short rx_body_421
- call rx_read
-rx_body_421:
- xor rbp, rax
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- mov r10, rax
- cmp r8d, -1600409762
- jo short rx_i_422
- call rx_i_31
-
-rx_i_422: ;IMUL_32
- dec ebx
- jz rx_finish
- xor r11, 04dd16ca4h
- mov eax, r11d
- test bl, 63
- jnz short rx_body_422
- call rx_read
-rx_body_422:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- movsxd rcx, eax
- movsxd rax, r10d
- imul rax, rcx
- mov rcx, rax
- mov eax, r13d
- xor eax, 07c614e2h
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_423: ;MUL_64
- dec ebx
- jz rx_finish
- xor r12, 04df5ce05h
- mov eax, r12d
- test bl, 63
- jnz short rx_body_423
- call rx_read
-rx_body_423:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- imul rax, r10
- mov r15, rax
-
-rx_i_424: ;FPADD
- dec ebx
- jz rx_finish
- xor r13, 01ad12ce2h
- mov eax, r13d
- test bl, 63
- jnz short rx_body_424
- call rx_read
-rx_body_424:
- xor rbp, rax
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- addpd xmm0, xmm7
- movaps xmm9, xmm0
-
-rx_i_425: ;IMUL_32
- dec ebx
- jz rx_finish
- xor r8, 0a3c5391dh
- mov eax, r8d
- test bl, 63
- jnz short rx_body_425
- call rx_read
-rx_body_425:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- movsxd rcx, eax
- mov rax, 1776029069
- imul rax, rcx
- mov rcx, rax
- mov eax, r14d
- xor eax, 069dc0d8dh
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_426: ;IDIV_64
- dec ebx
- jz rx_finish
- xor r12, 09dd55ba0h
- mov eax, r12d
- test bl, 63
- jnz short rx_body_426
- call rx_read
-rx_body_426:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- ; magic divide by -590728721
- mov rdx, -4191230239118101979
- imul rdx
- mov rax, rdx
- xor edx, edx
- sar rax, 27
- sets dl
- add rax, rdx
- mov r14, rax
-
-rx_i_427: ;MUL_32
- dec ebx
- jz rx_finish
- xor r11, 0d6cae9aeh
- mov eax, r11d
- test bl, 63
- jnz short rx_body_427
- call rx_read
-rx_body_427:
- xor rbp, rax
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- mov ecx, eax
- mov eax, -2146332428
- imul rax, rcx
- mov r9, rax
-
-rx_i_428: ;RET
- dec ebx
- jz rx_finish
- xor r11, 0f807a961h
- mov eax, r11d
- test bl, 63
- jnz short rx_body_428
- call rx_read
-rx_body_428:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- mov r8, rax
- cmp rsp, rdi
- je short rx_i_429
- ret
-
-rx_i_429: ;MUL_64
- dec ebx
- jz rx_finish
- xor r12, 0650a4102h
- mov eax, r12d
- test bl, 63
- jnz short rx_body_429
- call rx_read
-rx_body_429:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- imul rax, r9
- mov rcx, rax
+ and eax, 262136
+ cmp dword ptr [rsi+rax], 1840094725
+ setbe cl
+ add r10, rcx
+ ; IMUL_M r6, L1[r7]
mov eax, r15d
- xor eax, 076a3ad84h
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_430: ;FPADD
- dec ebx
- jz rx_finish
- xor r14, 019cc0e5h
- mov eax, r14d
- test bl, 63
- jnz short rx_body_430
- call rx_read
-rx_body_430:
- xor rbp, rax
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- addpd xmm0, xmm8
- movaps xmm5, xmm0
- mov eax, r13d
- xor eax, 058891433h
- and eax, 32767
- movhpd qword ptr [rsi + rax * 8], xmm5
-
-rx_i_431: ;ROR_64
- dec ebx
- jz rx_finish
- xor r12, 0ed17ab58h
- mov eax, r12d
- test bl, 63
- jnz short rx_body_431
- call rx_read
-rx_body_431:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- mov rcx, r13
- ror rax, cl
- mov r13, rax
-
-rx_i_432: ;SUB_64
- dec ebx
- jz rx_finish
- xor r10, 01c3b321fh
- mov eax, r10d
- test bl, 63
- jnz short rx_body_432
- call rx_read
-rx_body_432:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- sub rax, 876274173
- mov r8, rax
-
-rx_i_433: ;ADD_32
- dec ebx
- jz rx_finish
- xor r13, 0bbb88499h
- mov eax, r13d
- test bl, 63
- jnz short rx_body_433
- call rx_read
-rx_body_433:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- add eax, 1193456495
- mov rcx, rax
- mov eax, r12d
- xor eax, 04722b36fh
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_434: ;FPDIV
- dec ebx
- jz rx_finish
- xor r13, 0167edabdh
- mov eax, r13d
- test bl, 63
- jnz short rx_body_434
- call rx_read
-rx_body_434:
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- divpd xmm0, xmm3
- movaps xmm1, xmm0
- cmpeqpd xmm1, xmm1
- andps xmm0, xmm1
- movaps xmm9, xmm0
-
-rx_i_435: ;MUL_64
- dec ebx
- jz rx_finish
- xor r15, 0b940480ah
- mov eax, r15d
- test bl, 63
- jnz short rx_body_435
- call rx_read
-rx_body_435:
- xor rbp, rax
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- imul rax, rax, 1971717631
- mov rcx, rax
- mov eax, r9d
- xor eax, 0758605ffh
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_436: ;FPADD
- dec ebx
- jz rx_finish
- xor r15, 0bfc3ca8bh
- mov eax, r15d
- test bl, 63
- jnz short rx_body_436
- call rx_read
-rx_body_436:
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- addpd xmm0, xmm2
- movaps xmm7, xmm0
-
-rx_i_437: ;FPMUL
- dec ebx
- jz rx_finish
- xor r8, 098a6bcf7h
- mov eax, r8d
- test bl, 63
- jnz short rx_body_437
- call rx_read
-rx_body_437:
- xor rbp, rax
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- mulpd xmm0, xmm3
- movaps xmm1, xmm0
- cmpeqpd xmm1, xmm1
- andps xmm0, xmm1
- movaps xmm8, xmm0
-
-rx_i_438: ;FPMUL
- dec ebx
- jz rx_finish
- xor r10, 0325b38ebh
- mov eax, r10d
- test bl, 63
- jnz short rx_body_438
- call rx_read
-rx_body_438:
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- mulpd xmm0, xmm9
- movaps xmm1, xmm0
- cmpeqpd xmm1, xmm1
- andps xmm0, xmm1
- movaps xmm4, xmm0
- mov eax, r12d
- xor eax, 0b7c490eeh
- and eax, 2047
- movlpd qword ptr [rsi + rax * 8], xmm4
-
-rx_i_439: ;OR_64
- dec ebx
- jz rx_finish
- xor r13, 05e807e81h
- mov eax, r13d
- test bl, 63
- jnz short rx_body_439
- call rx_read
-rx_body_439:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- or rax, -1299288575
- mov r10, rax
-
-rx_i_440: ;CALL
- dec ebx
- jz rx_finish
- xor r10, 062f83728h
- mov eax, r10d
- test bl, 63
- jnz short rx_body_440
- call rx_read
-rx_body_440:
- xor rbp, rax
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- mov rcx, rax
- mov eax, r9d
- xor eax, 07ed31f7ah
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
- cmp r12d, 2127765370
- jns short rx_i_441
- call rx_i_41
-
-rx_i_441: ;ADD_64
- dec ebx
- jz rx_finish
- xor r14, 0d18ec075h
- mov eax, r14d
- test bl, 63
- jnz short rx_body_441
- call rx_read
-rx_body_441:
- xor rbp, rax
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- add rax, r14
- mov rcx, rax
- mov eax, r9d
- xor eax, 01f93242ch
- and eax, 131071
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_442: ;FPSQRT
- dec ebx
- jz rx_finish
- xor r14, 0a53dd1bh
- mov eax, r14d
- test bl, 63
- jnz short rx_body_442
- call rx_read
-rx_body_442:
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- andps xmm0, xmm10
- sqrtpd xmm3, xmm0
-
-rx_i_443: ;RET
- dec ebx
- jz rx_finish
- xor r14, 0232d1285h
- mov eax, r14d
- test bl, 63
- jnz short rx_body_443
- call rx_read
-rx_body_443:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- mov r9, rax
- cmp rsp, rdi
- je short rx_i_444
- ret
-
-rx_i_444: ;FPSUB
- dec ebx
- jz rx_finish
- xor r8, 042455dd8h
- mov eax, r8d
- test bl, 63
- jnz short rx_body_444
- call rx_read
-rx_body_444:
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- subpd xmm0, xmm7
- movaps xmm5, xmm0
-
-rx_i_445: ;ADD_64
- dec ebx
- jz rx_finish
- xor r13, 09ae009b2h
- mov eax, r13d
- test bl, 63
- jnz short rx_body_445
- call rx_read
-rx_body_445:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- add rax, r11
- mov r9, rax
-
-rx_i_446: ;MUL_32
- dec ebx
- jz rx_finish
- xor r12, 01734708eh
- mov eax, r12d
- test bl, 63
- jnz short rx_body_446
- call rx_read
-rx_body_446:
- xor rbp, rax
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- mov ecx, eax
- mov eax, r15d
- imul rax, rcx
- mov rcx, rax
- mov eax, r13d
- xor eax, 03166163h
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_447: ;FPADD
- dec ebx
- jz rx_finish
- xor r8, 01596d0e8h
- mov eax, r8d
- test bl, 63
- jnz short rx_body_447
- call rx_read
-rx_body_447:
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- addpd xmm0, xmm7
- movaps xmm5, xmm0
- mov eax, r13d
- xor eax, 0b384d4afh
- and eax, 2047
- movlpd qword ptr [rsi + rax * 8], xmm5
-
-rx_i_448: ;FPSUB
- dec ebx
- jz rx_finish
- xor r9, 0390cfdb0h
- mov eax, r9d
- test bl, 63
- jnz short rx_body_448
- call rx_read
-rx_body_448:
- xor rbp, rax
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- subpd xmm0, xmm3
- movaps xmm9, xmm0
-
-rx_i_449: ;ROL_64
- dec ebx
- jz rx_finish
- xor r8, 04f27744bh
- mov eax, r8d
- test bl, 63
- jnz short rx_body_449
- call rx_read
-rx_body_449:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- rol rax, 28
- mov rcx, rax
- mov eax, r8d
- xor eax, 089e19790h
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_450: ;SAR_64
- dec ebx
- jz rx_finish
- xor r8, 04e2c76ffh
- mov eax, r8d
- test bl, 63
- jnz short rx_body_450
- call rx_read
-rx_body_450:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- mov rcx, r12
- sar rax, cl
- mov r8, rax
-
-rx_i_451: ;ADD_64
- dec ebx
- jz rx_finish
- xor r8, 0c4d99ac9h
- mov eax, r8d
- test bl, 63
- jnz short rx_body_451
- call rx_read
-rx_body_451:
- xor rbp, rax
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- add rax, r10
- mov rcx, rax
- mov eax, r8d
- xor eax, 0eedd10b3h
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_452: ;RET
- dec ebx
- jz rx_finish
- xor r13, 040130b88h
- mov eax, r13d
- test bl, 63
- jnz short rx_body_452
- call rx_read
-rx_body_452:
- and eax, 32767
- mov rax, qword ptr [rsi+rax*8]
- mov r11, rax
- cmp rsp, rdi
- je short rx_i_453
- ret
-
-rx_i_453: ;DIV_64
- dec ebx
- jz rx_finish
- xor r11, 0a2096aa4h
- mov eax, r11d
- test bl, 63
- jnz short rx_body_453
- call rx_read
-rx_body_453:
- and eax, 131071
- mov rax, qword ptr [rsi+rax*8]
- ; magic divide by 380157076
- shr rax, 2
- mov rcx, 3256390890604862173
- mul rcx
- mov rax, rdx
- shr rax, 24
- mov rcx, rax
- mov eax, r8d
- xor eax, 016a8bc94h
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_454: ;FPADD
- dec ebx
- jz rx_finish
- xor r13, 081314291h
- mov eax, r13d
- test bl, 63
- jnz short rx_body_454
- call rx_read
-rx_body_454:
- and eax, 32767
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- addpd xmm0, xmm9
- movaps xmm4, xmm0
-
-rx_i_455: ;OR_64
- dec ebx
- jz rx_finish
- xor r8, 059263cdbh
- mov eax, r8d
- test bl, 63
- jnz short rx_body_455
- call rx_read
-rx_body_455:
- xor rbp, rax
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- or rax, r9
- mov r8, rax
-
-rx_i_456: ;AND_64
- dec ebx
- jz rx_finish
- xor r9, 010e8fe6h
- mov eax, r9d
- test bl, 63
- jnz short rx_body_456
- call rx_read
-rx_body_456:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- and rax, 401943615
- mov r9, rax
-
-rx_i_457: ;SUB_64
- dec ebx
- jz rx_finish
- xor r9, 09de1a3efh
- mov eax, r9d
- test bl, 63
- jnz short rx_body_457
- call rx_read
-rx_body_457:
- xor rbp, rax
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- sub rax, 1482178870
- mov rcx, rax
- mov eax, r10d
- xor eax, 058584136h
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_458: ;SAR_64
- dec ebx
- jz rx_finish
- xor r11, 05c79df6eh
- mov eax, r11d
- test bl, 63
- jnz short rx_body_458
- call rx_read
-rx_body_458:
- and eax, 131071
- mov rax, qword ptr [rsi+rax*8]
- mov rcx, r8
- sar rax, cl
- mov rcx, rax
- mov eax, r14d
- xor eax, 028f0a8ch
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_459: ;MUL_64
- dec ebx
- jz rx_finish
- xor r9, 0346f46adh
- mov eax, r9d
- test bl, 63
- jnz short rx_body_459
- call rx_read
-rx_body_459:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- imul rax, r9
- mov rcx, rax
- mov eax, r13d
- xor eax, 016bb0164h
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_460: ;ADD_32
- dec ebx
- jz rx_finish
- xor r11, 098ab71fch
- mov eax, r11d
- test bl, 63
- jnz short rx_body_460
- call rx_read
-rx_body_460:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- add eax, -347784553
- mov r12, rax
-
-rx_i_461: ;XOR_64
- dec ebx
- jz rx_finish
- xor r11, 0c814e926h
- mov eax, r11d
- test bl, 63
- jnz short rx_body_461
- call rx_read
-rx_body_461:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- xor rax, 1659853721
- mov rcx, rax
- mov eax, r12d
- xor eax, 062ef5b99h
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_462: ;ADD_64
- dec ebx
- jz rx_finish
- xor r10, 0c64b4a9eh
- mov eax, r10d
- test bl, 63
- jnz short rx_body_462
- call rx_read
-rx_body_462:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- add rax, r8
- mov rcx, rax
- mov eax, r15d
- xor eax, 098a05350h
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_463: ;ADD_32
- dec ebx
- jz rx_finish
- xor r9, 08c29341h
- mov eax, r9d
- test bl, 63
- jnz short rx_body_463
- call rx_read
-rx_body_463:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- add eax, r15d
- mov rcx, rax
- mov eax, r10d
- xor eax, 0c8204c90h
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_464: ;MUL_64
- dec ebx
- jz rx_finish
- xor r12, 06ff587fdh
- mov eax, r12d
- test bl, 63
- jnz short rx_body_464
- call rx_read
-rx_body_464:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- imul rax, r15
- mov r13, rax
-
-rx_i_465: ;FPADD
- dec ebx
- jz rx_finish
- xor r12, 0b62c0003h
- mov eax, r12d
- test bl, 63
- jnz short rx_body_465
- call rx_read
-rx_body_465:
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- addpd xmm0, xmm5
- movaps xmm2, xmm0
- mov eax, r10d
- xor eax, 0d11c1242h
- and eax, 32767
- movhpd qword ptr [rsi + rax * 8], xmm2
-
-rx_i_466: ;IMUL_32
- dec ebx
- jz rx_finish
- xor r13, 05c541c42h
- mov eax, r13d
- test bl, 63
- jnz short rx_body_466
- call rx_read
-rx_body_466:
- xor rbp, rax
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- movsxd rcx, eax
- movsxd rax, r13d
- imul rax, rcx
- mov r9, rax
-
-rx_i_467: ;FPADD
- dec ebx
- jz rx_finish
- xor r8, 0cbb33f81h
- mov eax, r8d
- test bl, 63
- jnz short rx_body_467
- call rx_read
-rx_body_467:
- xor rbp, rax
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- addpd xmm0, xmm9
- movaps xmm8, xmm0
-
-rx_i_468: ;DIV_64
- dec ebx
- jz rx_finish
- xor r8, 091044dc3h
- mov eax, r8d
- test bl, 63
- jnz short rx_body_468
- call rx_read
-rx_body_468:
- xor rbp, rax
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- ; magic divide by 4281572471
- add rax, 1
- sbb rax, 0
- mov rcx, 9252227195836753313
- mul rcx
- mov rax, rdx
- shr rax, 31
- mov r8, rax
-
-rx_i_469: ;IMUL_32
- dec ebx
- jz rx_finish
- xor r9, 0c0186beh
- mov eax, r9d
- test bl, 63
- jnz short rx_body_469
- call rx_read
-rx_body_469:
- and eax, 131071
- mov rax, qword ptr [rsi+rax*8]
- movsxd rcx, eax
- movsxd rax, r9d
- imul rax, rcx
- mov rcx, rax
- mov eax, r9d
- xor eax, 01186619dh
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_470: ;OR_64
- dec ebx
- jz rx_finish
- xor r14, 090849e3eh
- mov eax, r14d
- test bl, 63
- jnz short rx_body_470
- call rx_read
-rx_body_470:
- and eax, 131071
- mov rax, qword ptr [rsi+rax*8]
- or rax, r11
- mov r14, rax
-
-rx_i_471: ;IMUL_32
- dec ebx
- jz rx_finish
- xor r14, 0cedba9b6h
- mov eax, r14d
- test bl, 63
- jnz short rx_body_471
- call rx_read
-rx_body_471:
- and eax, 32767
- mov rax, qword ptr [rsi+rax*8]
- movsxd rcx, eax
- mov rax, 1914863189
- imul rax, rcx
- mov r14, rax
-
-rx_i_472: ;JUMP
- dec ebx
- jz rx_finish
- xor r9, 038f4b9d6h
- mov eax, r9d
- test bl, 63
- jnz short rx_body_472
- call rx_read
-rx_body_472:
- xor rbp, rax
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- mov r10, rax
- cmp r10d, 1738497427
- jl rx_i_8
-
-rx_i_473: ;MUL_64
- dec ebx
- jz rx_finish
- xor r14, 01fb7637dh
- mov eax, r14d
- test bl, 63
- jnz short rx_body_473
- call rx_read
-rx_body_473:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- imul rax, r11
- mov r12, rax
-
-rx_i_474: ;JUMP
- dec ebx
- jz rx_finish
- xor r9, 0b5c0b4d4h
- mov eax, r9d
- test bl, 63
- jnz short rx_body_474
- call rx_read
-rx_body_474:
- xor rbp, rax
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- mov r15, rax
- cmp r15d, -233120543
- jo rx_i_69
-
-rx_i_475: ;FPSUB
- dec ebx
- jz rx_finish
- xor r10, 0910dcdeeh
- mov eax, r10d
- test bl, 63
- jnz short rx_body_475
- call rx_read
-rx_body_475:
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- subpd xmm0, xmm9
- movaps xmm7, xmm0
-
-rx_i_476: ;FPADD
- dec ebx
- jz rx_finish
- xor r8, 07ab3b5a4h
- mov eax, r8d
- test bl, 63
- jnz short rx_body_476
- call rx_read
-rx_body_476:
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- addpd xmm0, xmm2
- movaps xmm9, xmm0
- mov eax, r9d
- xor eax, 0b01bb14ch
- and eax, 2047
- movlpd qword ptr [rsi + rax * 8], xmm9
-
-rx_i_477: ;FPADD
- dec ebx
- jz rx_finish
- xor r12, 07a29ec63h
- mov eax, r12d
- test bl, 63
- jnz short rx_body_477
- call rx_read
-rx_body_477:
- and eax, 32767
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- addpd xmm0, xmm9
- movaps xmm6, xmm0
-
-rx_i_478: ;MUL_64
- dec ebx
- jz rx_finish
- xor r14, 02d3d7e7fh
- mov eax, r14d
- test bl, 63
- jnz short rx_body_478
- call rx_read
-rx_body_478:
- and eax, 131071
- mov rax, qword ptr [rsi+rax*8]
- imul rax, r10
- mov r12, rax
-
-rx_i_479: ;MUL_64
- dec ebx
- jz rx_finish
- xor r12, 09b49c793h
- mov eax, r12d
- test bl, 63
- jnz short rx_body_479
- call rx_read
-rx_body_479:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- imul rax, r14
- mov r13, rax
-
-rx_i_480: ;FPADD
- dec ebx
- jz rx_finish
- xor r9, 0a9cc4f01h
- mov eax, r9d
- test bl, 63
- jnz short rx_body_480
- call rx_read
-rx_body_480:
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- addpd xmm0, xmm4
- movaps xmm6, xmm0
-
-rx_i_481: ;DIV_64
- dec ebx
- jz rx_finish
- xor r14, 0225ba1f9h
- mov eax, r14d
- test bl, 63
- jnz short rx_body_481
- call rx_read
-rx_body_481:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- ; magic divide by 2101516912
- shr rax, 4
- mov rcx, 147267437180322377
- mul rcx
- mov rax, rdx
- shr rax, 20
- mov r12, rax
-
-rx_i_482: ;AND_32
- dec ebx
- jz rx_finish
- xor r14, 044a0f592h
- mov eax, r14d
- test bl, 63
- jnz short rx_body_482
- call rx_read
-rx_body_482:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- and eax, 1304556205
- mov rcx, rax
- mov eax, r11d
- xor eax, 04dc1f2adh
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_483: ;FPADD
- dec ebx
- jz rx_finish
- xor r11, 07f71f219h
- mov eax, r11d
- test bl, 63
- jnz short rx_body_483
- call rx_read
-rx_body_483:
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- addpd xmm0, xmm6
- movaps xmm6, xmm0
- mov eax, r14d
- xor eax, 0545908cah
- and eax, 2047
- movlpd qword ptr [rsi + rax * 8], xmm6
-
-rx_i_484: ;SHR_64
- dec ebx
- jz rx_finish
- xor r12, 07027bacdh
- mov eax, r12d
- test bl, 63
- jnz short rx_body_484
- call rx_read
-rx_body_484:
- xor rbp, rax
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- shr rax, 37
- mov rcx, rax
- mov eax, r11d
- xor eax, 074a50ee0h
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_485: ;JUMP
- dec ebx
- jz rx_finish
- xor r13, 03a04647h
- mov eax, r13d
- test bl, 63
- jnz short rx_body_485
- call rx_read
-rx_body_485:
- xor rbp, rax
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- mov rcx, rax
- mov eax, r15d
- xor eax, 02112cbaeh
- and eax, 131071
- mov qword ptr [rsi + rax * 8], rcx
- cmp r8d, 554879918
- jno rx_i_58
-
-rx_i_486: ;ADD_64
- dec ebx
- jz rx_finish
- xor r15, 0ad072937h
- mov eax, r15d
- test bl, 63
- jnz short rx_body_486
- call rx_read
-rx_body_486:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- add rax, r8
- mov rcx, rax
- mov eax, r8d
- xor eax, 03832b3b2h
- and eax, 32767
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_487: ;SUB_64
- dec ebx
- jz rx_finish
- xor r11, 07f78ad34h
- mov eax, r11d
- test bl, 63
- jnz short rx_body_487
- call rx_read
-rx_body_487:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- sub rax, r9
- mov rcx, rax
- mov eax, r11d
- xor eax, 0ec228e26h
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_488: ;DIV_64
- dec ebx
- jz rx_finish
- xor r12, 0d8b1788eh
- mov eax, r12d
- test bl, 63
- jnz short rx_body_488
- call rx_read
-rx_body_488:
- and eax, 32767
- mov rax, qword ptr [rsi+rax*8]
- ; magic divide by 297357073
- mov rcx, 16652572300311555393
- mul rcx
- mov rax, rdx
- shr rax, 28
- mov r12, rax
-
-rx_i_489: ;JUMP
- dec ebx
- jz rx_finish
- xor r10, 0b2ec9f3ah
- mov eax, r10d
- test bl, 63
- jnz short rx_body_489
- call rx_read
-rx_body_489:
- xor rbp, rax
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- mov r8, rax
- cmp r15d, -1127175870
- jge rx_i_75
-
-rx_i_490: ;ROR_64
- dec ebx
- jz rx_finish
- xor r11, 015c7f598h
- mov eax, r11d
- test bl, 63
- jnz short rx_body_490
- call rx_read
-rx_body_490:
- xor rbp, rax
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- mov rcx, r9
- ror rax, cl
- mov r15, rax
-
-rx_i_491: ;FPADD
- dec ebx
- jz rx_finish
- xor r8, 0902da6bdh
- mov eax, r8d
- test bl, 63
- jnz short rx_body_491
- call rx_read
-rx_body_491:
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- addpd xmm0, xmm9
- movaps xmm7, xmm0
- mov eax, r15d
- xor eax, 0b0f0fca4h
- and eax, 32767
- movhpd qword ptr [rsi + rax * 8], xmm7
-
-rx_i_492: ;IDIV_64
- dec ebx
- jz rx_finish
- xor r9, 0491090d9h
- mov eax, r9d
- test bl, 63
- jnz short rx_body_492
- call rx_read
-rx_body_492:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- ; magic divide by -1779388031
- mov rcx, rax
- mov rdx, 7315366159790064091
- imul rdx
- mov rax, rdx
- xor edx, edx
- sub rax, rcx
- sar rax, 30
- sets dl
- add rax, rdx
- mov rcx, rax
- mov eax, r12d
- xor eax, 095f0b181h
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_493: ;FPSUB
- dec ebx
- jz rx_finish
- xor r8, 09de81282h
- mov eax, r8d
- test bl, 63
- jnz short rx_body_493
- call rx_read
-rx_body_493:
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- subpd xmm0, xmm9
- movaps xmm4, xmm0
- mov eax, r12d
- xor eax, 02feb2fd7h
- and eax, 2047
- movhpd qword ptr [rsi + rax * 8], xmm4
-
-rx_i_494: ;MUL_32
- dec ebx
- jz rx_finish
- xor r10, 0b0d50e46h
- mov eax, r10d
- test bl, 63
- jnz short rx_body_494
- call rx_read
-rx_body_494:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- mov ecx, eax
- mov eax, r11d
- imul rax, rcx
- mov r14, rax
-
-rx_i_495: ;FPMUL
- dec ebx
- jz rx_finish
- xor r11, 0e276cad1h
- mov eax, r11d
- test bl, 63
- jnz short rx_body_495
- call rx_read
-rx_body_495:
- and eax, 32767
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- mulpd xmm0, xmm2
- movaps xmm1, xmm0
- cmpeqpd xmm1, xmm1
- andps xmm0, xmm1
- movaps xmm8, xmm0
- mov eax, r8d
- xor eax, 02d12bd27h
- and eax, 2047
- movlpd qword ptr [rsi + rax * 8], xmm8
-
-rx_i_496: ;IDIV_64
- dec ebx
- jz rx_finish
- xor r14, 0fe757b73h
- mov eax, r14d
- test bl, 63
- jnz short rx_body_496
- call rx_read
-rx_body_496:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- ; magic divide by -359802064
- mov rdx, -860153514353783887
- imul rdx
- mov rax, rdx
- xor edx, edx
- sar rax, 24
- sets dl
- add rax, rdx
- mov r9, rax
-
-rx_i_497: ;FPMUL
- dec ebx
- jz rx_finish
- xor r8, 08d25742eh
- mov eax, r8d
- test bl, 63
- jnz short rx_body_497
- call rx_read
-rx_body_497:
- xor rbp, rax
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- mulpd xmm0, xmm3
- movaps xmm1, xmm0
- cmpeqpd xmm1, xmm1
- andps xmm0, xmm1
- movaps xmm8, xmm0
-
-rx_i_498: ;FPMUL
- dec ebx
- jz rx_finish
- xor r15, 0e066fd15h
- mov eax, r15d
- test bl, 63
- jnz short rx_body_498
- call rx_read
-rx_body_498:
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- mulpd xmm0, xmm9
- movaps xmm1, xmm0
- cmpeqpd xmm1, xmm1
- andps xmm0, xmm1
- movaps xmm8, xmm0
-
-rx_i_499: ;IMUL_32
- dec ebx
- jz rx_finish
- xor r12, 08925556bh
- mov eax, r12d
- test bl, 63
- jnz short rx_body_499
- call rx_read
-rx_body_499:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- movsxd rcx, eax
- movsxd rax, r13d
- imul rax, rcx
- mov r8, rax
-
-rx_i_500: ;FPSQRT
- dec ebx
- jz rx_finish
- xor r10, 04bc870ebh
- mov eax, r10d
- test bl, 63
- jnz short rx_body_500
- call rx_read
-rx_body_500:
- and eax, 32767
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- andps xmm0, xmm10
- sqrtpd xmm2, xmm0
- mov eax, r10d
- xor eax, 04a250342h
- and eax, 32767
- movhpd qword ptr [rsi + rax * 8], xmm2
-
-rx_i_501: ;XOR_64
- dec ebx
- jz rx_finish
- xor r8, 07d46c503h
- mov eax, r8d
- test bl, 63
- jnz short rx_body_501
- call rx_read
-rx_body_501:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- xor rax, r10
- mov rcx, rax
- mov eax, r12d
- xor eax, 03e22874bh
- and eax, 2047
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_502: ;RET
- dec ebx
- jz rx_finish
- xor r10, 09e70b20ch
- mov eax, r10d
- test bl, 63
- jnz short rx_body_502
- call rx_read
-rx_body_502:
- xor rbp, rax
- and eax, 131071
- mov rax, qword ptr [rsi+rax*8]
- mov r9, rax
- cmp rsp, rdi
- je short rx_i_503
- ret
-
-rx_i_503: ;FPSUB
- dec ebx
- jz rx_finish
- xor r13, 0442e4850h
- mov eax, r13d
- test bl, 63
- jnz short rx_body_503
- call rx_read
-rx_body_503:
- and eax, 32767
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- subpd xmm0, xmm2
- movaps xmm9, xmm0
- mov eax, r9d
- xor eax, 080465282h
- and eax, 2047
- movlpd qword ptr [rsi + rax * 8], xmm9
-
-rx_i_504: ;FPADD
- dec ebx
- jz rx_finish
- xor r13, 099d48347h
- mov eax, r13d
- test bl, 63
- jnz short rx_body_504
- call rx_read
-rx_body_504:
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- addpd xmm0, xmm9
- movaps xmm4, xmm0
-
-rx_i_505: ;FPSUB
- dec ebx
- jz rx_finish
- xor r12, 032c0a28ah
- mov eax, r12d
- test bl, 63
- jnz short rx_body_505
- call rx_read
-rx_body_505:
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- subpd xmm0, xmm4
- movaps xmm8, xmm0
- mov eax, r8d
- xor eax, 021b54eaeh
- and eax, 32767
- movhpd qword ptr [rsi + rax * 8], xmm8
-
-rx_i_506: ;FPSUB
- dec ebx
- jz rx_finish
- xor r9, 0a973d58ch
- mov eax, r9d
- test bl, 63
- jnz short rx_body_506
- call rx_read
-rx_body_506:
- and eax, 32767
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- subpd xmm0, xmm9
- movaps xmm3, xmm0
- mov eax, r11d
- xor eax, 05e890759h
- and eax, 2047
- movlpd qword ptr [rsi + rax * 8], xmm3
-
-rx_i_507: ;RET
- dec ebx
- jz rx_finish
- xor r10, 0d3b7165ch
- mov eax, r10d
- test bl, 63
- jnz short rx_body_507
- call rx_read
-rx_body_507:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- mov r14, rax
- cmp rsp, rdi
- je short rx_i_508
- ret
-
-rx_i_508: ;RET
- dec ebx
- jz rx_finish
- xor r13, 0da34d818h
- mov eax, r13d
- test bl, 63
- jnz short rx_body_508
- call rx_read
-rx_body_508:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- mov r8, rax
- cmp rsp, rdi
- je short rx_i_509
- ret
-
-rx_i_509: ;FPROUND
- dec ebx
- jz rx_finish
- xor r11, 01b2873f2h
- mov eax, r11d
- test bl, 63
- jnz short rx_body_509
- call rx_read
-rx_body_509:
- and eax, 2047
- mov rax, qword ptr [rsi+rax*8]
- mov rcx, rax
- rol rax, 34
- and eax, 24576
- or eax, 40896
- mov dword ptr [rsp - 8], eax
- ldmxcsr dword ptr [rsp - 8]
- mov eax, r10d
- xor eax, 06cd84each
- and eax, 32767
- mov qword ptr [rsi + rax * 8], rcx
-
-rx_i_510: ;FPADD
- dec ebx
- jz rx_finish
- xor r8, 0db65513ch
- mov eax, r8d
- test bl, 63
- jnz short rx_body_510
- call rx_read
-rx_body_510:
- and eax, 2047
- cvtdq2pd xmm0, qword ptr [rsi+rax*8]
- addpd xmm0, xmm2
- movaps xmm9, xmm0
- mov eax, r9d
- xor eax, 097614097h
- and eax, 2047
- movlpd qword ptr [rsi + rax * 8], xmm9
-
-rx_i_511: ;SHR_64
- dec ebx
- jz rx_finish
- xor r11, 02bd79286h
- mov eax, r11d
- test bl, 63
- jnz short rx_body_511
- call rx_read
-rx_body_511:
- and eax, 131071
- mov rax, qword ptr [rsi+rax*8]
- shr rax, 56
- mov r11, rax
-
- jmp rx_i_0
+ and eax, 16376
+ imul r14, qword ptr [rsi+rax]
+ ; IMULH_M r6, L1[r5]
+ mov ecx, r13d
+ and ecx, 16376
+ mov rax, r14
+ mul qword ptr [rsi+rcx]
+ mov r14, rdx
+ ; IMUL_9C r7, -1048659408
+ lea r15, [r15+r15*8-1048659408]
+ ; IMUL_R r6, r3
+ imul r14, r11
+ ; FPADD_R f3, a0
+ addpd xmm3, xmm8
+ ; IMULH_R r0, r3
+ mov rax, r8
+ mul r11
+ mov r8, rdx
+ ; FPSWAP_R f0
+ shufpd xmm0, xmm0, 1
+ ; FPSQRT_R e3
+ sqrtpd xmm7, xmm7
+ ; IMULH_R r2, r0
+ mov rax, r10
+ mul r8
+ mov r10, rdx
+ ; FPDIV_R e1, a1
+ divpd xmm5, xmm9
+ maxpd xmm5, xmm13