WIP3

Makefile

@@ -22,7 +22,7 @@ endif
 
 ifeq ($(V),1)
 	OPTS += -v              # Be verbose
-	BENCH_OPTS += -gcflags=-m     # Show results from inlining
+	BENCH_OPTS += -gcflags="-m -m"    # Show results from inlining
 endif
 
 all: test

p503/arith_amd64.s

@@ -698,13 +698,13 @@ TEXT ·mulWithMULXADX(SB), NOSPLIT, $104-24
 	MUL(CX, REG_P2, REG_P1, MULS256_MULXADX)
 	RET
 
-TEXT ·fp503Mul(SB), NOSPLIT, $104-24
-	// Actual implementation
-	MOVQ    z+ 0(FP), CX
-	MOVQ    x+ 8(FP), REG_P2
-	MOVQ    y+16(FP), REG_P1
-	MUL(CX, REG_P2, REG_P1, MULS256_MULXADX)
-	RET
+// TEXT ·fp503Mul(SB), NOSPLIT, $104-24
+// 	// Actual implementation
+// 	MOVQ    z+ 0(FP), CX
+// 	MOVQ    x+ 8(FP), REG_P2
+// 	MOVQ    y+16(FP), REG_P1
+// 	MUL(CX, REG_P2, REG_P1, MULS256_MULXADX)
+// 	RET
 
 TEXT ·fp503MulXXX(SB), NOSPLIT, $104-72
 	// Actual implementation

p503/arith_decl.go

@@ -4,7 +4,7 @@ package p503
 
 import (
 	. "github.com/cloudflare/p751sidh/internal/isogeny"
-	//	cpu "github.com/cloudflare/p751sidh/internal/utils"
+	cpu "github.com/cloudflare/p751sidh/internal/utils"
 )
 
 // If choice = 0, leave x,y unchanged. If choice = 1, set x,y = y,x.
@@ -54,8 +54,13 @@ func mulWithMULX(z *FpElementX2, x, y *FpElement)
 
 // Mul implementation for CPUs supporting two independent carry chain
 // (ADOX/ADCX) instructions and carry-less MULX multiplier
-//go:noescape
-func fp503Mul(z *FpElementX2, x, y *FpElement)
+func fp503Mul(z *FpElementX2, x, y *FpElement) {
+	if cpu.HasBMI2 {
+		mulWithMULXADX(z, x, y)
+	} else {
+		mul(z, x, y)
+	}
+}
 
 //go:noescape
 func fp503MulXXX(z, x, y []uint64)

p503/field_ops.go

@@ -36,8 +36,8 @@ func (fp503Ops) Mul(dest, lhs, rhs *Fp2Element) {
 	//
 	// so (a*d + b*c) = (b-a)*(c-d) + a*c + b*d.
 
-	fp503Mul1(ac[:], lhs.A[:], rhs.A[:]) // = a*c*R*R
-	fp503Mul(&bd, &lhs.B, &rhs.B)        // = b*d*R*R
+	fp503Mul(&ac, &lhs.A, &rhs.A) // = a*c*R*R
+	fp503Mul(&bd, &lhs.B, &rhs.B) // = b*d*R*R
 
 	fp503SubReduced(&b_minus_a, &lhs.B, &lhs.A) // = (b-a)*R
 	fp503SubReduced(&c_minus_d, &rhs.A, &rhs.B) // = (c-d)*R