vrepeat 0.0.3 ----------------------------------------------- Usage: vrepeat [options] CMD1 CMD2 ... Description: Repeat command(s) and collect statistics. Note: quote each command (argument), when it contains spaces. The arguments should be at least 1 in number. Options: -h, --help Show this help screen. -r, --runsRun count. Default: 10 -R, --repeats Repeats count (it repeats everything, including reporting). Default: 1 -w, --warmup Warmup run count. These are done *at the start* of each series, and the timings are ignored. Default: 2 -s, --series Series count. `-s 2 -r 4 a b` => aaaabbbbaaaabbbb, while `-s 3 -r 2 a b` => aabbaabbaabb. Default: 1 -e, --ignore Ignore failed commands (returning a non 0 exit code). -N, --no_vexe_reset Do not reset the VEXE env variable at the start. By default, VEXE will be set to "", to allow for measuring different V executables. Use this option to override it -n, --newline Use \n, do not overwrite the last line. Produces more output, but easier to diagnose. -O, --output Show command stdout/stderr in the progress indicator for each command. Note: slower, for verbose commands. -v, --verbose Be more verbose. -m, --max_time Fail with exit code 2, when first cmd takes above M milliseconds (regression). Default: 60000 -f, --fail_percent Fail with exit code 3, when first cmd is X% slower than the rest (regression). Default: 100000 -t, --template Command template. {T} will be substituted with the current command. Default: {T}. Here is an example, that will produce and run 24 permutations = (3 for opt) x (2 for source) x (4 = v names): v repeat -p opt=-check-syntax,-check,"-o x.c" -p source=examples/hello_world.v,examples/hanoi.v --template "./{T} {opt} {source}" vold vnew vold_prod vnew_prod -p, --parameter A parameter substitution list. `pp=val1,val2,val2` means that {pp} in the template, will be substituted with *each* of val1, val2, val3. -i, --nmins Ignore the BOTTOM X results (minimum execution time). Makes the results more robust to performance flukes. Default: 0 -a, --nmaxs Ignore the TOP X results (maximum execution time). Makes the results more robust to performance flukes. Default: 0