bcov produces coverage information without recompiling a program by instrumenting it with breakpoints. Effectively it debugs the program automatically, removing each breakpoint after it has been hit. Currently only the main binary itself is instrumented. Usage: bcov binary [argument(s)] Executes the binary with the given arguments and stores the coverage summary in .bcovdump. The result file is more or less human readable (and easily machine readable), a nicer presentation can be generated with bcov-report: Usage: bcov-report [dumpfile] [output directory] Converts the coverage dump into an lcov-style html report. If not output directory is given bcov-report uses a temporary directory and tries to open the result in the standard browser.