Below assertion checks if src_data has changed more than once before clock changes. I am observing a race around where src_data may have #0 glitch and assertion is false firing. How do I rewrite this?
property data_src_no_glitch (src_data, clk_dest, kill=0);
logic data;
@(src_data) disable iff (kill)
(1,data = !src_data) |=> @(clk_dest) (src_data == data);
endproperty
In reply to vshankr:
Because you are observing a race around where src_data with a #0 glitch, you could use trigger an automatic task upon the detection of src_data transition. That task does a #0 to let the race condition settle, and then continue with the assertion.
Now, I am not sure if #0 will do the trick, that needs further analysis. If that #0 solution fails, use a #1, as shown in the comments.
Please let us know if that worked, and whether the #0 solution worked.
always @(src_data) begin
if(kill==0)
fork
glitch_t();
join_none
end
task automatic glitch_t();
logic data;
#0; // May work, needs analysis
data = $sampled(src_data);
// #1; // If #0 does not work
// data = src_data;
@(clk_dest);
assert(src_data == data);
endtask
Ben Cohen
http://www.systemverilog.us/ ben@systemverilog.us
For training, consulting, services: contact Home - My cvcblr