DNSSEC: constraint checks for KSK and ZSK

fixes #285

See merge request !256

TSIG+OPT bugfixes

See merge request !257

fixes #282

Bulk updates

Pending:
 - ~~renaming (update vs. request vs. query ...)~~
 - ~~proper logging (waiting for !250)~~
 - ~~`rrset_is_nsec3_rel()` and `mm_realloc` copypasta~~
 - partial zone adjustment - next MR

Pros:
 - faster for lots of small updates
 - improved terrible empty node cleanup (now takes time proportional to size of update, not size of the zone)
 - spinlocks instead of mutexes for ddns q access (faster for this case + less memory :fisted_hand_sign: )

Cons:
 - if one transfer fails (no matter the reason, except for acl issues), they all fail
 - might actually be slower for small singular updates, but oh well

Not sure if good or bad:
 - most of logging is lost (only logs errors in detail, i.e. who sent the faulty update)

See merge request !251

Trie next

See merge request !255

knotc fixes

Fixed zone name comparison in `knotc checkzone` and `knotc memstats`. See commit message. Fixes #280.

See merge request !254

EDNS OPT writing

Fixes a problem, where OPT RR was not included into the packet, when a previous RRs didn't fit.

See merge request !253

Zone names were compared only partially and case-sensitive.
Changed to dname comparison.

fixes #286

- same way of OPT size computation for reservation and reclaim
- correct max_size setting