[LTP] [Automated-testing] [PATCH v2 01/30] Introduce a concept of max runtime

Li Wang liwang@redhat.com
Tue May 10 15:15:51 CEST 2022


Richard Palethorpe <rpalethorpe@suse.de> wrote:


> >> > +  int max_iteration_runtime;
> >>
> >> It's not immediately clear if iteration refers to the inner test loop
> >> (e.g. Fuzzy Sync) or the outer loop performed by adding '-i N'. Perhaps
> >> it would be better to call it max_runtime and document that it is scaled
> >> by '-i N'?
> >
> > It's not only the -i N paramater, it's carthesian product of:
> >
> >       -i parameter x all_filesystems x test_variants
> >
> > And every single instance of that product is limited by the runtime
> > value, which is the reason I want to have iteration explicitly in the
> > name. Maybe we should call it instance instead or whatever else that may
> > be more fitting.
>
> I suppose it is the innermost runtime, so I would vote for inner_runtime
> or just runtime. However whatever it is called it will require
>

I do really have the same feelings here. I'm even afraid 'iteration' will
confused people who are not familiar with ltp-library.

>From my comprehension, max_iteration_runtime takes effect on every
onefold test iteration, or maybe just directly rename to 'max_test_runtime'
(or max_runtime) with explanation in code comments.



> explanation. What is considered to be an iteration, instance, inner, test
> etc. is arbitrary.
>

Yes, includes "onefold" which hovers in my mind.


-- 
Regards,
Li Wang
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.linux.it/pipermail/ltp/attachments/20220510/da2e5237/attachment-0001.htm>


More information about the ltp mailing list