Re: [LTP] [PATCH v5 3/7] lib: Add support for max_kver to struct tst_test and tst_fs

[Petr Vorel <pvorel@suse.cz>]

> Hi!
> > > Uff, this is quite ugly. What exactly are we trying to do?

> > I hoped that was explained in the comment above the code:
> >  .max_kver = "7.0" should run the test not only kernel <= 7.0,
> > but also all stable kernels: 7.0.x.
> > OTOH if there is something backported to stable kernel and one specifies:
> > .max_kver = "7.0.5" it will be compared just to <= 7.0.5.

> > Sure, this can be avoided if .max_kver = "7.0" is not inclusive,
> > i.e. < 7.0 (one would have to use .max_kver = "7.1"), which is less
> > intuitive, because .min_kver is inclusive).

> Or maybe instead of 7.0 the tests should require something as "7.0.*"
> to make it clear that it will run on any 7.0.X patchlevel. And the
> parser vould set the v3 to INT_MAX if it sees '*' as the value for v3.

For me was quite obvious that 7.0.x should ideally have the same main features
as 7.0, just contain fixes. But for others it'd be confusing that 7.0 means
something different in .max_kver then in .min_kver. But the same applies to
"7.0.*", adding it to .min_kver will break the runtime code.

Of course, only if we agree we want .max_kver at all.

Kind regards,
Petr

-- 
Mailing list info: https://lists.linux.it/listinfo/ltp