Add the memory envelope, pprof, and a per-packet fix on the UDP path

A cgroup OOM kill is the worst outcome available to a resolver: an outage
rather than a cache miss, with no chance to serve stale on the way out. So:

- memory.limit / memory.reserve / memory.cache_max_bytes, a new config section.
  limit becomes a GOMEMLIMIT through debug.SetMemoryLimit, reduced by the
  reserve. The reserve exists because the Go heap does not account for
  goroutine stacks, runtime metadata, or the bbolt mmap backing the cache
  snapshot and the query log — handing the runtime the full cgroup limit is how
  you get killed by a mechanism that believed it had headroom.
- config.Size, so a budget reads as "512MiB" rather than a byte count converted
  by hand. Strict about junk: "512MiBx" is an error.
- The cache byte budget is now actually fed to the LRU. It already enforced
  maxBytes per shard on insert, but had no setter and nothing ever called one,
  so the footprint of cache.size was unbounded: one cached ANY or TXT answer
  can be tens of kilobytes, so 500k entries is anywhere from 150MB to several GB
  depending on what the traffic asked for.

- api.pprof mounts net/http/pprof on the admin listener. The pprof handlers sit
  on the app-level mux, not inside the API's own handler, so they would bypass
  the auth middleware entirely; the option is therefore refused unless every
  admin address is loopback. Profiling means port-forwarding, so this is safe
  by construction rather than by policy, and cmdline cannot leak an argument
  that happens to be a key.

- serveUDP was calling net.ResolveUDPAddr per packet. The peer already is a
  *net.UDPAddr; that call builds a string, re-parses the port and allocates on
  the busiest path in the process.

Also fixes a flaky test: TestParallelPolicyAnswersFromTheFastestUpstream
asserted the slow upstreams were hit, but the fan-out cancels on the first
answer, so whether they were dispatched before cancellation is a real race. It
now asserts the answer came from the fast upstream and that nothing waited.

deploy/config.example.yaml

@@ -70,6 +70,21 @@ upstream:
       proto: tls
       server_name: dns.quad9.net
 
+# The memory envelope. A cgroup kill is an outage; a cache miss is not, so the
+# runtime gets a soft ceiling to work towards instead of finding out the hard
+# way. Set limit to your container's memory limit, minus the reserve.
+memory:
+  # Soft heap ceiling (GOMEMLIMIT). 0 = unbounded. The Go heap excludes
+  # goroutine stacks, runtime metadata and the bbolt mmap for the cache
+  # snapshot and the query log, which is what the reserve is for.
+  limit: "0"
+  reserve: "256MiB"
+  # Byte budget for the in-memory LRU. 0 derives it from the limit (60%).
+  # cache.size alone does not bound memory: one cached ANY or TXT answer can
+  # be tens of kilobytes, so 500k entries is anywhere from 150MB to several GB
+  # depending on what the traffic asked for.
+  cache_max_bytes: "0"
+
 cache:
   enabled: true
   size: 200000