Resolve "underflow in stats channel stale cached RRSIG gauge [ISC-support #14769]"

Closes #1074 and #602

See merge request !2237

When updating the statistics for RRset types, if a header is marked
stale or ancient, the appropriate statistic counters are decremented,
then incremented.

Also fix some out of date comments.

Having the decrement/increment logic in stats makes the code hard
to follow. Remove it here and adjust the unit test. The caller
will be responsible for maintaining the correct increments and
decrements for statistics counters (in the following commit).

The stale RR types are now printed with '#'.  This used to be the
prefix for RR types that were marked ancient, but commit
df507515 changed the meaning.  It is
probably better to keep '#' for stale RR types and introduce a new
prefix for reintroducing ancient type stat counters.

Clarify relationship between ACLs and RPZ

Closes #1110

See merge request !2255

In the ARM section about RPZ, add text explicitly stating that ACLs take
precedence over RPZ to prevent users from expecting RPZ actions to be
applied to queries coming from clients which are not permitted access to
the resolver by ACLs.

Remove DLV validation code

Closes #7

See merge request !2233

- this required modification to the code that generates grammar text for
  the documentation, because the "dnssec-lookaside" option spanned more
  than one line in doc/misc/options, so grepping out only the lines
  marked "// obsolete" didn't remove the whole option.  this commit adds
  an option to cfg_test to print named.conf clauses only if they don't
  have the obsolete, ancient, test-only, or not-yet-implemented flags
  set.

Add PKG_CHECK_VAR and AS_VAR_COPY compat macros (Courtesy of ycflash)

Closes #1182

See merge request !2246

Implement a convenience function for "rndc dumpdb"

See merge request !1967

Add a helper shell function, rndc_dumpdb(), which provides a convenient
way to call "rndc dumpdb" for a given server with optional additional
arguments.  Since database dumping is an asynchronous process, the
function waits until the dump is complete before returning, which
prevents false positives in system tests caused by inspecting the dump
before its preparation is finished.  The function also renames the dump
file before returning so that it does not get overwritten by subsequent
calls; this retains forensic data in case of an unexpected test failure.

configure.ac: autodetect 'pause' instruction presence on sparc

See merge request !2243

The change fixes the following build failure on sparc T3 and older CPUs:

```
sparc-unknown-linux-gnu-gcc ... -O2 -mcpu=niagara2 ... -c rwlock.c
{standard input}: Assembler messages:
{standard input}:398: Error: Architecture mismatch on "pause ".
{standard input}:398: (Requires v9e|v9v|v9m|m8; requested architecture is v9b.)
make[1]: *** [Makefile:280: rwlock.o] Error 1
```

`pause` insutruction exists only on `-mcpu=niagara4` (`T4`) and upper.

The change adds `pause` configure-time autodetection and uses it if available.
config.h.in got new `HAVE_SPARC_PAUSE` knob. Fallback is a fall-through no-op.

Build-tested on:

- sparc-unknown-linux-gnu-gcc (no `pause`, build succeeds)
- sparc-unknown-linux-gnu-gcc -mcpu=niagara4 (`pause`, build succeeds)

Reported-by: Rolf Eike Beer
Bug: https://bugs.gentoo.org/691708Signed-off-by: Sergei Trofimovich <slyfox@gentoo.org>