On Mon, Jun 26, 2023 at 01:24:45PM -0700, John Hubbard wrote: > On 6/26/23 09:24, Peter Xu wrote: > > On Mon, Jun 26, 2023 at 10:10:26AM +0200, David Hildenbrand wrote: > > > On 23.06.23 16:29, Peter Xu wrote: > > > > Allows to specify optional tests in run_vmtests.sh, where we can run time > > > > consuming test matrix only when user specified "-a". > > > > > > I'd have used something like "-e: extended tests that might be more > > > time-consuming". > > > > '-a' stands for "all" here, I worry '-e' may be misread into running "extra > > tests only". > > > > It's getting to the point where long options would help. This one > could be "--stress-tests" or "--all", or "--include-long-running", > for example. > > I hesitate to suggest it because it's extra work, but at some point > it would be nice. The options are already hopelessly un-guessable. Thanks for being considerate, John. Yes I do hope this series can converge soon, indeed hopefully any further test patches can be worked on top. To me I would be fine with either option (-a/-e), as long as there's easy way to dump a help message, where "run_vmtest.sh -h" works for me always. What's more "hopeless" along my way as an user is there's no help message for gup_test.c even though its parameter list is crazily long. I used to "convince" myself that gup is just so special so someone should just understand what option goes to what ioctl, better read into the .c file. But I know that's an excuse. Thanks, -- Peter Xu
