fellow_storage: in disk LRU, get the disk reserve upfront

an unintended side effect of the last rework was that the main LRU loop would
only run once the disk buddy had waiting requests, such that the reserve would
not be built up unless it was needed at least once. Which, to some extent,
defeats the purpose for freshly loaded caches.

src/fellow_storage.c

@@ -1755,8 +1755,6 @@ sfedsk_lru_thread(struct worker *wrk, void *arg)
 	wrk->vsl = &vsl;
 	VSL_Setup(wrk->vsl, NULL, (size_t)0);
 
-	buddy_wait_needspace(buddy);
-
 	while (!stvfe->shutdown) {
 		sfedsk_reserve_resize(r,
 		    stvfe->tune.dsk_reserve_chunks,
@@ -1774,7 +1772,8 @@ sfedsk_lru_thread(struct worker *wrk, void *arg)
 		n = 0;
 
 		sfedsk_reserve_release(r);
-		nuked += sfedsk_nuke(wrk, stvfe->ffd, stv, ++n);
+		if (buddy->waiting)
+			nuked += sfedsk_nuke(wrk, stvfe->ffd, stv, ++n);
 
 		sfedsk_reserve_req(r);
 		while (buddy->waiting) {