x86/vm/macroAssembler_x86.cpp

8992                                            Register carry, Register product,
8997   //    huge_128 product = y[idx] * x[xstart] + carry;
8998   //    z[kdx] = (jlong)product;
8999   //    carry  = (jlong)(product >>> 64);
9021   movq(product, x_xstart);
9022   mulq(y_idx); // product(rax) * y_idx -> rdx:rax
9023   addq(product, carry);
9026   movl(Address(z, kdx, Address::times_4,  4), product);
9027   shrq(product, 32);
9028   movl(Address(z, kdx, Address::times_4,  0), product);
9048                                             Register carry, Register product, int offset) {
9049   //     huge_128 product = (y[idx] * x_xstart) + z[kdx] + carry;
9050   //     z[kdx] = (jlong)product;
9054   movq(product, x_xstart);
9055   mulq(yz_idx);     // product(rax) * yz_idx -> rdx:product(rax)
9059   add2_with_carry(rdx, product, carry, yz_idx);
9061   movl(Address(z, idx, Address::times_4,  offset+4), product);
9062   shrq(product, 32);
9063   movl(Address(z, idx, Address::times_4,  offset), product);
9072                                              Register carry, Register product,
9077   //     huge_128 product = (y[idx+1] * x_xstart) + z[kdx+idx+1] + carry;
9078   //     z[kdx+idx+1] = (jlong)product;
9079   //     jlong carry2  = (jlong)(product >>> 64);
9080   //     product = (y[idx] * x_xstart) + z[kdx+idx] + carry2;
9081   //     z[kdx+idx] = (jlong)product;
9082   //     carry  = (jlong)(product >>> 64);
9086   //     product = (y[idx] * x_xstart) + z[kdx+idx] + carry;
9087   //     z[kdx+idx] = (jlong)product;
9088   //     carry  = (jlong)(product >>> 64);
9103   multiply_add_128_x_128(x_xstart, y, z, yz_idx, idx, carry, product, 8);
9106   multiply_add_128_x_128(x_xstart, y, z, yz_idx, idx, carry2, product, 0);
9119   multiply_add_128_x_128(x_xstart, y, z, yz_idx, idx, carry, product, 0);
9129   movq(product, x_xstart);
9130   mulq(yz_idx); // product(rax) * yz_idx -> rdx:product(rax)
9133   add2_with_carry(rdx, product, yz_idx, carry);
9135   movl(Address(z, idx, Address::times_4,  0), product);
9136   shrq(product, 32);
9139   orq(product, rdx);
9140   movq(carry, product);
9303   const Register product  = xlen;
9313   //    long product = (y[idx] & LONG_MASK) * (x[xstart] & LONG_MASK) + carry;
9314   //    z[kdx] = (int)product;
9315   //    carry = product >>> 32;
9330   multiply_64_x_64_loop(x, xstart, x_xstart, y, y_idx, z, carry, product, idx, kdx);
9352   //     long product = (y[jdx] & LONG_MASK) * (x[i] & LONG_MASK) +
9354   //     z[k] = (int)product;
9355   //     carry = product >>> 32;
9395     multiply_128_x_128_bmi2_loop(y, z, carry, x, jdx, ylen, product, tmp2, x_xstart, tmp3, tmp4);
9397     multiply_128_x_128_loop(x_xstart, y, z, y_idx, jdx, ylen, carry, product, x);
9670   //     huge_128 product = x[j:j+1] * x[j:j+1];
9671   //     z[i:i+1] = (carry << 63) | (jlong)(product >>> 65);
9672   //     z[i+2:i+3] = (jlong)(product >>> 1);
9673   //     carry = (jlong)product;
10068 //    long product = (in[j] & LONG_MASK) * kLong +
10070 //    out[offs--] = (int)product;
10071 //    carry = product >>> 32;
10386 // Output: the 64-bit carry-less product of B * CONST