On Fri, Mar 01, 2019 at 04:57:32PM +0100, Andrea Bolognani wrote:
Basically I see no reason not to cover both active and inactive
when possible, because even when the feature is only going to
manifest itself only in one of the two output files, the fact that
it does *not* show up in the other one represent some amount of
useful (negative) coverage.
k
Reviewed-by: Ján Tomko <jtomko(a)redhat.com>
Jano