Type Coverage vs Instance Coverage

Where the interviewer goes next. Open a node to practise that follow-up.

WhyWhat IfBreak ItCompare

1. Type Coverage vs Instance Coverage
2. WhyWhy the Type Figure Is an AverageL2 Mid-Level

   Why does the type figure read 50% when the instances together hit every bin?

   Reveal the Model Answer

   With merging off, each instance is scored against its own bins, and the type is the weighted average of those scores. Port A scores 50 and port B scores 50, so the type is 50. The rule asks "how well was each instance covered", not "was each bin hit somewhere". One idle instance therefore pulls the type figure down.

   Figure	Bins hit	Score
   Instance A	b[0], b[1]	50%
   Instance B	b[2], b[3]	50%
   Type, default (average)	each instance scored on its own	50%
   Type, merge_instances = 1 (union)	all four, from either instance	100%

   The same samples as the root question. Only the rule for combining instances changes.

   Self-Check

   Tick each key point you covered in your answer.

   Key point 1 With merging off, each instance is scored against its own bins, and the type is the weighted average of those scores. Key point 2 Port A scores 50 and port B scores 50, so the type is 50. Key point 3 The rule asks "how well was each instance covered", not "was each bin hit somewhere". Key point 4 One idle instance therefore pulls the type figure down.

   Tool as Truth

   "The average rule is a tool bug."

   It is the standard's default. Tools that show the union have merge_instances set, or show a merged database view.

   Source Mehta, Introduction to SystemVerilog, Table 15.2 (PDF p613)

   • Lesson C6Coverage in a Class
3. CompareWhen to Merge InstancesL2 Mid-LevelL3 Senior

   Compare the merged view with the averaged view. When do you want each?

   Reveal the Model Answer

   Merge when the instances are interchangeable and a case seen on any one of them is enough. Identical lanes behind one arbiter are an example. Average, and read per-instance numbers, when every instance must see every case, such as ports with their own logic. The choice belongs in the coverage plan, written next to the covergroup.

   Self-Check

   Tick each key point you covered in your answer.

   Key point 1 Merge when the instances are interchangeable and a case seen on any one of them is enough. Key point 2 Identical lanes behind one arbiter are an example. Key point 3 Average, and read per-instance numbers, when every instance must see every case, such as ports with their own logic. Key point 4 The choice belongs in the coverage plan, written next to the covergroup.

   Half-Truth

   "Always merge; the higher number is the true one."

   A merged 100% can hide an instance that never saw traffic at all, as the next question shows.

   Source IEEE 1800-2017 §19.7.1

   • Lesson C6Coverage in a Class
   • Lesson C7Closing Holes
4. Break ItA Dead Port Behind a Merged 100 PercentL2 Mid-LevelL3 Senior

   Four ports share one covergroup type. Port 3 never sees traffic. What do the averaged and merged figures say?

   module tb;
     covergroup avg_cg with function sample(bit [1:0] x);
       cp: coverpoint x { bins b[] = {[0:3]}; }
     endgroup
     covergroup union_cg with function sample(bit [1:0] x);
       type_option.merge_instances = 1;
       cp: coverpoint x { bins b[] = {[0:3]}; }
     endgroup
     avg_cg a[4];
     union_cg u[4];
     initial begin
       foreach (a[i]) begin a[i] = new(); u[i] = new(); end
       for (int p = 0; p < 3; p++)              // ports 0-2 see every value; port 3 is dead
         for (int v = 0; v < 4; v++) begin a[p].sample(2'(v)); u[p].sample(2'(v)); end
       $display("average over ports=%0.2f union over ports=%0.2f", a[0].get_coverage(), u[0].get_coverage());
     end
   endmodule
   

   Reveal the Model Answer

   The average reads 75%, and the union reads 100%. Merging asks whether any instance hit each bin, so three busy ports cover for the dead one. The averaged figure at least shows a gap. Only the per-instance report says which port is dead.

   RUNVerilator (coverage build) · simulator service

   Edit the code, then run it

   OutputWaves

   RunReset codeExpected: average over ports=75.00 union over ports=100.00

   Self-Check

   Tick each key point you covered in your answer.

   Key point 1 The average reads 75%, and the union reads 100%. Key point 2 Merging asks whether any instance hit each bin, so three busy ports cover for the dead one. Key point 3 The averaged figure at least shows a gap. Key point 4 Only the per-instance report says which port is dead.

   Misconception

   "A merged 100% means every port saw every case."

   It means every case was seen on some port. The run reads 100% with port 3 idle.

   Source Measured: the run on this question, Verilator coverage build (master b8e7f1e)

   • Lesson C6Coverage in a Class
5. What IfCoverage Without the Per-Instance OptionL2 Mid-Level

   What if option.per_instance is left at 0?

   Reveal the Model Answer

   Each instance still contributes to the type figure. But the per-instance data is not saved to the coverage database, so the report cannot show which port is weak. Set per_instance = 1 whenever instances watch different things. It can only be set on the covergroup, not on a coverpoint.

   Self-Check

   Tick each key point you covered in your answer.

   Key point 1 Each instance still contributes to the type figure. Key point 2 But the per-instance data is not saved to the coverage database, so the report cannot show which port is weak. Key point 3 Set per_instance = 1 whenever instances watch different things. Key point 4 It can only be set on the covergroup, not on a coverpoint.

   Spec Misquote

   "Set per_instance on the one coverpoint you care about."

   It is a covergroup-level option. It cannot be given in a coverpoint or a cross.

   Source Spear & Tumbush, SystemVerilog for Verification (3rd ed.), §9.10.1 (PDF p393)

   • Lesson C6Coverage in a Class