Add support for sharded bins within an arena.

This makes it possible to have multiple set of bins in an arena, which improves
arena scalability because the bins (especially the small ones) are always the
limiting factor in production workload.

A bin shard is picked on allocation; each extent tracks the bin shard id for
deallocation.  The shard size will be determined using runtime options.

src/jemalloc.c

@@ -379,6 +379,9 @@ arena_bind(tsd_t *tsd, unsigned ind, bool internal) {
 		tsd_iarena_set(tsd, arena);
 	} else {
 		tsd_arena_set(tsd, arena);
+		unsigned binshard = atomic_fetch_add_u(&arena->binshard_next, 1,
+		    ATOMIC_RELAXED) % BIN_SHARDS_MAX;
+		tsd_binshard_set(tsd, binshard);
 	}
 }