Improve precision of the io_limit setting

For async engines, we look only at completions. But we could have
a bunch inflight with a high queue depth, making us go higher than
we should.

Signed-off-by: Jens Axboe <axboe@fb.com>
diff --git a/fio.h b/fio.h
index be2f23a..d28f8ce 100644
--- a/fio.h
+++ b/fio.h
@@ -235,7 +235,15 @@
 	uint64_t total_io_size;
 	uint64_t fill_device_size;
 
+	/*
+	 * Issue side
+	 */
 	uint64_t io_issues[DDIR_RWDIR_CNT];
+	uint64_t io_issue_bytes[DDIR_RWDIR_CNT];
+
+	/*
+	 * Completions
+	 */
 	uint64_t io_blocks[DDIR_RWDIR_CNT];
 	uint64_t this_io_blocks[DDIR_RWDIR_CNT];
 	uint64_t io_bytes[DDIR_RWDIR_CNT];