Solve-before is a biasing directive: it does not add or remove legal tuples, but it can heavily change observed frequencies by steering variable choice order.
1. Classic a->b Bias Shift [Medium]
2. Three-Branch Kind Selection [Hard]
3. Reverse Bias with solve b before a [Hard]
4. m<n Cardinality Bias [Medium]
5. Mode Branch Collapse Under Extra Relation [Hard]
6. When solve-before Is Mostly Redundant [Medium]Work each puzzle before reading the solution: predict the exact output or answer, name the rule that governs it, then check yourself.
CONSTRAINT SOLVE ORDER
rand bit [3:0] burst_len;
rand bit [7:0] addr;
constraints:
burst_len inside {[1:8]};
addr % burst_len == 0;
solve burst_len before addr;
solver picks burst_len first
then chooses addr satisfying alignment
Explicit solve order reduces ambiguity in interview puzzles.