Hi Peter,
> > 3. Go with your patch and just fix up the warnings about uninitialized
> > variables. But that seems the least principled to me.
>
> IMO the 3rd option is the lesser evil. Initializing four bools/ints in
> order to make Clang 11 happy doesn't strike me as such a big deal. At
> least until somebody reports a bottleneck for this particular reason.
> We can optimize the code when and if this will happen.
Since the first patch was applied, cfbot now complains that it can't
apply the patchset. Here is the rebased version.
--
Best regards,
Aleksander Alekseev
