Manual sample called at wrong phase

Puzzle

Difficulty: Hard · Puzzle 3 of 6 · Topic: Coverage Closure and Sampling Timing Puzzles

Manual sampling is used, but closure still misses the latest state. Compute observed bins and explain timing.

Code

bit clk;
bit [1:0] state, next_state;

covergroup cg with function sample(bit [1:0] s);
  cp_state: coverpoint s {
    bins s0 = {0};
    bins s1 = {1};
    bins s2 = {2};
    bins s3 = {3};
  }
endgroup

cg c = new();

always_ff @(posedge clk) state <= next_state;
always @(posedge clk) c.sample(state);

initial begin
  state = 0;
  next_state = 1;
  @(posedge clk); next_state = 2;
  @(posedge clk); next_state = 3;
  @(posedge clk);
end

Hint

sample(state) in an always block at posedge still reads pre-NBA state.

Step-by-step solution

1) c.sample(state) executes in active region on posedge.
2) state <= next_state updates in NBA after active region.
3) Sampled states over three edges are 0,1,2.
4) Hit bins are s0,s1,s2 -> 3/4 = 75%, leaving s3 uncovered.

Answer

Answer: Observed closure is 75% because manual sample timing still captures old state values.

Why candidates get it wrong

Using sample() does not automatically fix timing; call site region still controls observed value.

Interviewer follow-up

Would clocking-block input skew or monitor-after-NBA sampling be a cleaner fix here?

Related topics

• Back to Coverage Closure and Sampling Timing Puzzles

• Functional Coverage Puzzles section