OPTIM: patterns: cache the current generation

This makes a significant difference when loading large files and during
commit and clear operations, thanks to improved cache locality. In the
measurements below, master refers to the code before any of the changes
to the patterns code, not the code before this one commit.

Timing the replacement of 10M entries from the CLI with this command
which also reports timestamps at start, end of upload and end of clear:

  $ (echo "prompt i"; echo "show activity"; echo "prepare acl #0";
     awk '{print "add acl @1 #0",$0}' < bad-ip.map; echo "show activity";
     echo "commit acl @1 #0"; echo "clear acl @0 #0";echo "show activity") |
    socat -t 10 - /tmp/sock1 | grep ^uptim

master, on a 3.7 GHz EPYC, 3 samples:

  uptime_now: 6.087030
  uptime_now: 25.981777  => 21.9 sec insertion time
  uptime_now: 29.286368  => 3.3 sec commit+clear

  uptime_now: 5.748087
  uptime_now: 25.740675  => 20.0s insertion time
  uptime_now: 29.039023  => 3.3 s commit+clear

  uptime_now: 7.065362
  uptime_now: 26.769596  => 19.7s insertion time
  uptime_now: 30.065044  => 3.3s commit+clear

And after this commit:

  uptime_now: 6.119215
  uptime_now: 25.023019  => 18.9 sec insertion time
  uptime_now: 27.155503  => 2.1 sec commit+clear

  uptime_now: 5.675931
  uptime_now: 24.551035  => 18.9s insertion
  uptime_now: 26.652352  => 2.1s commit+clear

  uptime_now: 6.722256
  uptime_now: 25.593952  => 18.9s insertion
  uptime_now: 27.724153  => 2.1s commit+clear

Now timing the startup time with a 10M entries file (on another machine)
on master, 20 samples:

Standard Deviation, s: 0.061652677408033
Mean:        4.217

And after this commit:

Standard Deviation, s: 0.081821371548669
Mean:        3.78

include/haproxy/pattern-t.h

@@ -112,6 +112,11 @@ struct pat_ref {
 	unsigned int flags; /* flags PAT_REF_*. */
 	unsigned int curr_gen; /* current generation number (anything below can be removed) */
 	unsigned int next_gen; /* next generation number (insertions use this one) */
+	/* We keep a cached pointer to the current generation for performance. */
+	struct {
+		struct pat_ref_gen *data;
+		unsigned int id;
+	} cached_gen;
 	int unique_id; /* Each pattern reference have unique id. */
 	unsigned long long revision; /* updated for each update */
 	unsigned long long entry_cnt; /* the total number of entries */

src/pattern.c

@@ -1669,7 +1669,21 @@ struct pat_ref_gen *pat_ref_gen_new(struct pat_ref *ref, unsigned int gen_id)
  */
 struct pat_ref_gen *pat_ref_gen_get(struct pat_ref *ref, unsigned int gen_id)
 {
-	return cebu32_item_lookup(&ref->gen_root, gen_node, gen_id, gen_id, struct pat_ref_gen);
+	struct pat_ref_gen *gen;
+
+	/* We optimistically try to use the cached generation if it's the current one. */
+	if (likely(gen_id == ref->curr_gen && gen_id == ref->cached_gen.id && ref->cached_gen.data))
+		return ref->cached_gen.data;
+
+	gen = cebu32_item_lookup(&ref->gen_root, gen_node, gen_id, gen_id, struct pat_ref_gen);
+	if (unlikely(!gen))
+		return NULL;
+
+	if (gen_id == ref->curr_gen) {
+		ref->cached_gen.id = gen_id;
+		ref->cached_gen.data = gen;
+	}
+	return gen;
 }
 
 /* This function removes all elements belonging to <gen_id> and matching <key>
@@ -1937,6 +1951,8 @@ static struct pat_ref *_pat_ref_new(const char *display, unsigned int flags)
 	ref->revision = 0;
 	ref->entry_cnt = 0;
 	ceb_init_root(&ref->gen_root);
+	ref->cached_gen.id = ref->curr_gen;
+	ref->cached_gen.data = NULL;
 	LIST_INIT(&ref->pat);
 	HA_RWLOCK_INIT(&ref->lock);
 	event_hdl_sub_list_init(&ref->e_subs);
@@ -2244,6 +2260,8 @@ int pat_ref_purge_range(struct pat_ref *ref, uint from, uint to, int budget)
 		BUG_ON(!LIST_ISEMPTY(&gen->head));
 		BUG_ON(!ceb_isempty(&gen->elt_root));
 		cebu32_item_delete(&ref->gen_root, gen_node, gen_id, gen);
+		if (gen->gen_id == ref->cached_gen.id)
+			ref->cached_gen.data = NULL;
 		free(gen);
 	}