block/io: optimize bdrv_co_pwritev for small requests

in a read-modify-write cycle a small request might cause
head and tail to fall into the same aligned block. Currently
QEMU reads the same block twice in this case which is
not necessary.

Signed-off-by: Peter Lieven <pl@kamp.de>
Message-id: 1464607873-28206-1-git-send-email-pl@kamp.de
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>

1 files changed, 8 insertions, 0 deletions

diff --git a/block/io.c b/block/io.c
index 7ac9897..e12f303 100644
--- a/block/io.c
+++ b/block/io.c
@@ -1427,6 +1427,14 @@ int coroutine_fn bdrv_co_pwritev(BlockDriverState *bs,
 
         bytes += offset & (align - 1);
         offset = offset & ~(align - 1);
+
+        /* We have read the tail already if the request is smaller
+         * than one aligned block.
+         */
+        if (bytes < align) {
+            qemu_iovec_add(&local_qiov, head_buf + bytes, align - bytes);
+            bytes = align;
+        }
     }
 
     if ((offset + bytes) & (align - 1)) {