I have two setup errors reported in my PAR/Timing report. However, the actual failing paths are over 100.
Why are the number of timing errors lower than the number of actual failed paths in timing report?
Error counts reflect the number of path endpoints (register setup inputs, output pads) that fail to meet timing constraints, not the number of paths that fail the specification.
If an error is generated at the endpoints of A and B, the timing report would list two errors, one for each endpoint.