add sqrPre3

include/mcl/fp_generator.hpp

@@ -258,10 +258,10 @@ struct FpGenerator : Xbyak::CodeGenerator {
 			op.fpDbl_mulPre = getCurr<void3u>();
 			gen_fpDbl_mulPre();
 		}
-		if (op.N == 3) {
+		if (op.N == 3 || op.N == 4) {
 			align(16);
 			op.fpDbl_sqrPre = getCurr<void2u>();
-			gen_fpDbl_sqrPre();
+			gen_fpDbl_sqrPre(op);
 		}
 	}
 	void gen_addSubNC(bool isAdd, int n)
@@ -1056,7 +1056,7 @@ struct FpGenerator : Xbyak::CodeGenerator {
 	/*
 		py[5..0] <- px[2..0]^2
 	*/
-	void sqr3x3(const RegExp& py, const RegExp& px, const Pack& t)
+	void sqrPre3(const RegExp& py, const RegExp& px, const Pack& t)
 	{
 		const Reg64& a = rax;
 		const Reg64& d = rdx;
@@ -1072,6 +1072,27 @@ struct FpGenerator : Xbyak::CodeGenerator {
 		const Reg64& t8 = t[8];
 		const Reg64& t9 = t[9];
 
+		if (useMulx_) {
+			mov(d, ptr [px + 8 * 0]);
+			mulx(t0, a, d);
+			mov(ptr [py + 8 * 0], a);
+
+			mov(t7, ptr [px + 8 * 1]);
+			mov(t9, ptr [px + 8 * 2]);
+			mulx(t2, t1, t7);
+			mulx(t4, t3, t9);
+
+			mov(t5, t2);
+			mov(t6, t4);
+
+			add(t0, t1);
+			adc(t5, t3);
+			adc(t6, 0); // [t6:t5:t0]
+
+			mov(d, t7);
+			mulx(t8, t7, d);
+			mulx(c, t9, t9);
+		} else {
 			mov(t9, ptr [px + 8 * 0]);
 			mov(a, t9);
 			mul(t9);
@@ -1093,7 +1114,6 @@ struct FpGenerator : Xbyak::CodeGenerator {
 			adc(t5, t3);
 			adc(t6, 0); // [t6:t5:t0]
 
-
 			mov(t9, ptr [px + 8 * 1]);
 			mov(a, t9);
 			mul(t9);
@@ -1103,7 +1123,7 @@ struct FpGenerator : Xbyak::CodeGenerator {
 			mul(t9);
 			mov(t9, a);
 			mov(c, d);
-
+		}
 		add(t2, t7);
 		adc(t8, t9);
 		mov(t7, c);
@@ -1130,7 +1150,7 @@ struct FpGenerator : Xbyak::CodeGenerator {
 	/*
 		pz[5..0] <- px[2..0] * py[2..0]
 	*/
-	void mul3x3(const RegExp& pz, const RegExp& px, const RegExp& py, const Pack& t)
+	void mulPre3(const RegExp& pz, const RegExp& px, const RegExp& py, const Pack& t)
 	{
 		const Reg64& a = rax;
 		const Reg64& d = rdx;
@@ -1197,7 +1217,7 @@ struct FpGenerator : Xbyak::CodeGenerator {
 	/*
 		pz[7..0] <- px[3..0] * py[3..0]
 	*/
-	void mul4x4(const RegExp& pz, const RegExp& px, const RegExp& py, const Pack& t)
+	void mulPre4(const RegExp& pz, const RegExp& px, const RegExp& py, const Pack& t)
 	{
 		const Reg64& a = rax;
 		const Reg64& d = rdx;
@@ -1276,21 +1296,28 @@ struct FpGenerator : Xbyak::CodeGenerator {
 		store_mr(pz + 8 * 3, Pack(t7, t8, t3, t2));
 		mov(ptr [pz + 8 * 7], d);
 	}
-	void gen_fpDbl_sqrPre()
+	void gen_fpDbl_sqrPre(mcl::fp::Op& op)
 	{
 		if (pn_ == 3) {
 			StackFrame sf(this, 2, 10 | UseRDX | UseRCX);
-			sqr3x3(sf.p[0], sf.p[1], sf.t);
+			sqrPre3(sf.p[0], sf.p[1], sf.t);
+			return;
 		}
+#ifdef XBYAK64_WIN
+		mov(r8, rdx);
+#else
+		mov(rdx, rsi);
+#endif
+		jmp((void*)op.fpDbl_mulPre);
 	}
 	void gen_fpDbl_mulPre()
 	{
 		if (pn_ == 3) {
 			StackFrame sf(this, 3, 10 | UseRDX);
-			mul3x3(sf.p[0], sf.p[1], sf.p[2], sf.t);
+			mulPre3(sf.p[0], sf.p[1], sf.p[2], sf.t);
 		} else if (pn_ == 4) {
 			StackFrame sf(this, 3, 10 | UseRDX);
-			mul4x4(sf.p[0], sf.p[1], sf.p[2], sf.t);
+			mulPre4(sf.p[0], sf.p[1], sf.p[2], sf.t);
 		}
 	}
 	static inline void debug_put_inner(const uint64_t *ptr, int n)

sample/rawbench.cpp

@@ -29,21 +29,20 @@ void benchRaw(const char *p, mcl::fp::Mode mode)
 	typedef mcl::fp::Unit Unit;
 	const size_t maxN = sizeof(Fp) / sizeof(Unit);
 	const mcl::fp::Op& op = Fp::getOp();
-	Fp fx = -1, fy;
+	cybozu::XorShift rg;
-	mpz_class mp(p);
+	Fp fx, fy;
-	fy.setMpz(mp / 2);
+	fx.setRand(rg);
+	fy.setRand(rg);
 	Unit ux[maxN * 2] = {};
 	Unit uy[maxN * 2] = {};
 	memcpy(ux, fx.getUnit(), sizeof(Unit) * op.N);
+	memcpy(ux + op.N, fx.getUnit(), sizeof(Unit) * op.N);
 	memcpy(uy, fy.getUnit(), sizeof(Unit) * op.N);
-	fy.setMpz(mp - 1);
+	memcpy(ux + op.N, fx.getUnit(), sizeof(Unit) * op.N);
-	memcpy(uy + op.N, fy.getUnit(), sizeof(Unit) * op.N);
 	double fp_sqrT, fp_addT, fp_subT, fp_mulT;
 	double fpDbl_addT, fpDbl_subT;
 	double fpDbl_sqrPreT, fpDbl_mulPreT, fpDbl_modT;
 	double fp2_sqrT, fp2_mulT, fp2_mul2T;
-//	double fp2_mulT, fp2_sqrT;
-//	double fp_addNCT, fp_subNCT, fpDbl_addNCT,fpDbl_subNCT;
 	CYBOZU_BENCH_T(fp_sqrT, op.fp_sqr, ux, ux);
 	CYBOZU_BENCH_T(fp_addT, op.fp_add, ux, ux, ux);
 	CYBOZU_BENCH_T(fp_subT, op.fp_sub, ux, uy, ux);