Fix assertion on region vs object liveliness during log read

This concerns an open issue from before the public release, which I
had not understood before and believe to have understood now.

With this code *) added to the test

@@ -5946,6 +5949,15 @@ t_cache(unsigned chksum)
	test_bocdone(fbo, TRUST_ME(hash), 1);
	fellow_cache_obj_deref(fc, fco);

+	// === max out region alloc
+	fbo = fellow_busy_obj_alloc(fc, &fco, &priv2, 1234).r.ptr;
+	CHECK_OBJ_NOTNULL(fbo, FELLOW_BUSY_MAGIC);
+	for (u = 0; u < FCO_MAX_REGIONS; u++)
+		AN(fellow_busy_region_alloc(fbo, 1234, INT8_MAX));
+
+	test_bocdone(fbo, TRUST_ME(hash), 1);
+	fellow_cache_obj_delete(fc, fco, hash);
+
	// === alloc space, dont use

we tripped here

   assert(flivs->oob || u == obj_alive);

with u == 1 and obj_alive == 0.

So the offset of a region from a dead object was not taken by a
subsequent allocation, which is fine, why should it be?

*) Note: The added test code is not correct yet, as it does not
   register the regions with the segment list, so obj_delete leaks.

src/fellow_log.c

@@ -5017,19 +5017,25 @@ fellow_logs_iter_block(const struct flics *flics, struct flivs *flivs,
 		case DLE_REG_ADD:
 			DLE_REG_FOREACH(e, off, sz) {
 				bit = off_bit(off);
-				// if object is not alive, this storage
-				// must also have been DEL'ed
+				/* mark any regions not already marked because
+				 * they can not be valid for an earlier object.
+				 *
+				 * We can only assert that they were free if the
+				 * object is alive, it is not, they might have
+				 * been reused or not.
+				 */
+
 				u = bitf_set(flivs->bitf, bit);
 				FDBG(D_LOGS_ITER_BLOCK,
 				    "  u %u = bitf_set(%zu)", u, bit);
-				assert(flivs->oob || u == obj_alive);
-				// ^^^ XXX WIP bugs/20230106_iter_not_alive
 
 				// semantically a break, but sanity
 				// check the other regions too
 				if (! obj_alive)
 					continue;
 
+				AN(u);
+
 				switch (flics->ffd->phase) {
 				case FP_INIT:
 				case FP_OPEN: