Re: [PATCH v2 7/8] selftests: Remove KSFT_TAP_LEVEL

[shuah]

On 4/25/19 10:56 AM, Kees Cook wrote:
> On Thu, Apr 25, 2019 at 9:36 AM shuah <shuah@xxxxxxxxxx> wrote:
> > On 4/24/19 5:12 PM, Kees Cook wrote:
> > > Since sub-testing can now be detected by indentation level, this removes
> > > KSFT_TAP_LEVEL so that subtests report their TAP header for later parsing.
> >
> > Does this take into ksft_print_header() getenv logic to avoid printing
> > TAP headers from tests when they fork? e.g timers tests do that a lot.
>
> I didn't change the ksft_print_header() code, in case you want it back
> in the future. But nothing sets that variable any more in my series:

Right. I want to see the impact of not setting this. I added this for
two reasons, one is to prevent nesting which goes away with your
refactor. I know there is another reason which I can't recall.


> $ git grep KSFT_TAP_LEVEL
> tools/testing/selftests/kselftest.h:    if (!(getenv("KSFT_TAP_LEVEL")))
>
> I don't see the timers tests using print_header() at all, actually...
>
> $ cd tools/testing/kselftest/timers
> $ git grep print_header | wc -l
> 0

I don't see it in there either. I must be thinking about another test.

btw I like the changes. I just want to make sure it gets some road
testing first. :)

thanks,
-- Shuah