Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

[SOLVED] No latch inferred how do I get rid of this problem ?

Status
Not open for further replies.
X

Xenon02

Guest
Newbie level 1
Hello !
I have a problem because I can't use yosys to sythesize this code. Whenever I try to do it this appears :

1669937840911.png

I don't know where are the latches in my code
Here is the code :

`include "definictions.vh"
module exe_unit_w31 (i_oper, i_argA, i_argB, i_clk, i_rsn, o_status, o_result);
input logic [`OPER_BITS-1:0] i_oper;
input logic [`BITS-1:0] i_argA, i_argB;
input logic i_clk;
input logic i_rsn;
output logic [3:0] o_status;
output logic [`BITS-1:0] o_result;
logic [`BITS-1:0] s_argA, s_argB;
logic [`BITS-1:0] s_oper;
logic [`BITS-1:0] s_y_sub;
logic [`BITS-1:0] s_y_cmp;
logic [`BITS-1:0] s_y_to0;
logic [`BITS-1:0] s_y_ZM2U2;
logic s_carry_sub;
logic s_err_to0;
logic s_err_ZM2U2;
logic sign3;

sub sub_exe (.i_a(s_argA), .i_b(s_argB), .o_y(s_y_sub), .o_carry(s_carry_sub));
cmp cmp_exe (.i_a(s_argA), .i_b(s_argB), .o_y(s_y_cmp));
to0 to0_exe (.i_a(s_argA), .i_b(s_argB), .o_y(s_y_to0), .o_err(s_err_to0));
ZM2U2 ZM2U2_exe (.i_a(s_argA), .o_y(s_y_ZM2U2), .o_err(s_err_ZM2U2));
integer ones;
integer i;
always_ff @(posedge i_clk)
begin
if(!i_rsn)
begin
s_oper <= '0;
s_argA <= '0;
s_argB <= '0;
end
else
begin
s_oper <= i_oper;
s_argA <= i_argA;
s_argB <= i_argB;
end
end
always_comb
begin
o_status = '0;
o_result = '0;
s_y_cmp = '0;
s_y_sub = '0;
s_y_to0 = '0;
s_y_ZM2U2 = '0;
s_oper = '0;
sign3 = '0;
s_carry_sub ='0;
s_err_to0 ='0;
s_err_ZM2U2 = '0;
case(s_oper)
`SUBTRACT:
begin
o_result = s_y_sub;
o_status[3] = 0;
end
`COMPARE:
begin
o_result = s_y_cmp;
o_status[3] = 0;
end
`CHANGE:
begin
o_result = s_y_to0;
if(s_err_to0)
o_status[3] = 1;
end
`CONVERSION:
begin
o_result = s_y_ZM2U2;
if(s_err_ZM2U2)
o_status[3] = 1;
end
default:
begin
o_result = '0;
o_status = 4'b1000;
end
endcase
for(int i = 0; i<`BITS ;i++)
begin
if(i ==`BITS-1)
begin
sign3 = o_result;
end
end
//NEG
if(i_oper!= 'd0)
begin
if(sign3 == 1)
begin
o_status[2] = 1;
end
end
else
begin
if(s_carry_sub == 1)
o_status[2] = 1;
else
begin
if(sign3 == 1)
begin
o_status[2] = 1;
end
end
end
//EVEN
ones = 0;
for(i = `BITS-1; i>=0; i = i-1)
begin
if(o_result == 1)
ones = ones + 1;
end
if(!(ones % 2))
o_status[1] = 1;
//ONES
if(ones == `BITS)
o_status[0] = 1;
end
endmodule


Can someone help me with this ?
 

Solution
The warning are related to the unshown module code, so we can hardly guess where the compiler is expecting latches.
An obvious fault is however to overwrite the module result in the always_comb section
Code:
s_y_cmp = '0;
s_y_sub = '0;
// etc

Deleting the erroneous assignments might also remove the warnings.
The warning are related to the unshown module code, so we can hardly guess where the compiler is expecting latches.
An obvious fault is however to overwrite the module result in the always_comb section
Code:
s_y_cmp = '0;
s_y_sub = '0;
// etc

Deleting the erroneous assignments might also remove the warnings.
 

Solution
Status
Not open for further replies.

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top