Hi,all. I'm a student of Hefei Unversity.
I'm reading the source code Floating-Point Unit which 's downloaded from www.opencores.org
I have troubles with it.
In the file pre_norm_fmul.v at line 99-102
assign {co2,exp_tmp2} = op_div ? ({co1,exp_tmp1} + 8'h7f) : ({co1,exp_tmp1} - 8'h7f);
assign exp_tmp3 = exp_tmp2 + 1;
assign exp_tmp4 = 8'h7f - exp_tmp1;
And at line 103,109
assign exp_tmp5 = op_div ? (exp_tmp4+1) : (exp_tmp4-1);
assign exp_out_div = (expa_dn | expb_dn) ? (co2 ? exp_tmp5 : exp_tmp3 ) : co2 ? exp_tmp4 : exp_tmp2;
Why we have to plus 1 with {exp_tmp4+1} and (exp_tmp2 + 1) when one of the division operand is denormalized number?
All your helps will be appreciated.