In short: This is a set of the most common coverage interview questions with clear, LRM-correct answers: code vs functional coverage, the code-coverage types, what high-one-low-other numbers mean, covergroups vs cover properties, coverpoints and bins, ignore vs illegal bins, transition and wildcard bins, cross coverage, and the two ways to sample a covergroup. Coverage comes…
In short: Good functional coverage guidelines keep your coverage honest and useful. The essentials: base coverage on the test plan and the spec, keep a traceability link from each scenario to a cover point, add controls to enable or disable coverage, turn coverage on after the early bug storm settles, collect coverage only from passing…
In short: There are two families of coverage metrics. Code coverage is measured automatically by the simulator and tells you how much of the RTL was exercised (statement, block, branch, condition or expression, toggle, and FSM). Functional coverage is written by you and tells you whether the features from the test plan were actually tested….
In short: Functional coverage options in SystemVerilog tune how a covergroup collects and reports data. The everyday ones are option.comment to label a group, option.per_instance to report each instance separately, option.at_least to set how many hits a bin needs, option.cross_num_print_missing to show empty bins, and option.goal to set the target percentage. They change reporting, not…