On Fri, Jun 1, 2018 at 4:12 AM, Ondrej Mosnacek <omosnace(a)redhat.com> wrote:
I'm actually playing with the idea of unifying the filtering
logic in
these two functions, where sharing this function wouldn't be
necessary. However, that is quite a big change (a lot of LOC being
moved around) so I'd prefer the simple & dirty approach now and keep
the cleanup for a later patch.
[Resend as I forgot the reply-all - oops]
I've had similar thoughts in the past, and deferred the work for
exactly the same reason.
If I recall correctly, I believe part of the reason may stem from the
fact that some fields are simply not always valid when the filter is
run and this may have been a crude effort at optimization (smaller
function size). Regardless, we might preserve some of this idea by
creating some helper functions (two?) with the different filter fields
in each (no overlap) and a couple of wrapper functions which call the
appropriate helpers ... or that just might be extra work for no noticeable
perf advantage :)
--
paul moore
www.paul-moore.com