BIRD Filter performance improvement: Call for testing

[Maria Matejka]

Hello!

We are finishing a significant rework of BIRD filters in 2.0 branch. We 
dropped the old interpreter and wrote it a bit differently. Currently, 
our tests are promising at least 10% reduction of the interpreting time; 
the actual improvement depends on the filter.

See the "faster-filters" branch in our git repository:
https://gitlab.labs.nic.cz/labs/bird/tree/faster-filters

Anyway, reworking a 20-year-old code with almost no documentation and 
not many tests always yields unexpected bugs. Please try to run your 
filters on this commit and report if it doesn't work. The frontend and 
overall behavior should not change in any way -- it should only do it 
faster.

There is a known bug -- we still haven't reimplemented caching of 
function and named filter comparison results so reconfiguration may take 
significantly longer time than before on this branch. This will be fixed 
before merging this branch into master.

Thank you for using BIRD, thank you for your testing!
Maria