[Drbd-dev] [PATCH] drbd:don't increase unacked_cnt when resync write error

[Rui Xu]

In drbd_endio_write_sec_final, if a write error has occured and
protocol is not 'C', inc_unacked will be called.

But when it is a resync write, we shouldn't increase the unacked_cnt
there because of there is no corresponding dec_unacked, then unacked_cnt
will never be zero, which leads to the problem that in do_start_resync,
start_resync_timer will be added over and over again and will never end.

Fix this problem by adding a determine, when resync write error, don't
increase unacked_cnt in drbd_endio_write_sec_final.

Signed-off-by: Rui Xu <rui.xu at easystack.cn>
---
 drbd/drbd_sender.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drbd/drbd_sender.c b/drbd/drbd_sender.c
index 7238730..aaccd07 100644
--- a/drbd/drbd_sender.c
+++ b/drbd/drbd_sender.c
@@ -148,7 +148,7 @@ void drbd_endio_write_sec_final(struct drbd_peer_request *peer_req) __releases(l
 	sector = peer_req->i.sector;
 	block_id = peer_req->block_id;
 
-	if (peer_req->flags & EE_WAS_ERROR) {
+	if ((peer_req->flags & EE_WAS_ERROR) && (block_id != ID_SYNCER)) {
                 /* In protocol != C, we usually do not send write acks.
                  * In case of a write error, send the neg ack anyways. */
                 if (!__test_and_set_bit(__EE_SEND_WRITE_ACK, &peer_req->flags))
-- 
1.8.3.1