cluster/afr: Handle NULL fdctx in fsync

Problem:
If fdctx is NULL in afr_fsync, process crashes because
of NULL dereference.

Fix:
if fdctx is NULL, always say witnessed unstable write so
that fsyncs are done properly. Handled fdctx being null
in afr_delayed_changelog_post_op otherwise fsync stub is
never resumed and the mount was hanging.

Change-Id: Icacc900e9be63c29db3325cb0e19cc250adebaac
BUG: 978794
Signed-off-by: Pranith Kumar K <pkarampu@redhat.com>
Reviewed-on: http://review.gluster.org/5258
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Vijay Bellur <vbellur@redhat.com>

1 files changed, 29 insertions, 0 deletions

diff --git a/tests/bugs/bug-978794.t b/tests/bugs/bug-978794.t
new file mode 100644
index 00000000000..d22d3cde33d
--- /dev/null
+++ b/tests/bugs/bug-978794.t
@@ -0,0 +1,29 @@
+#!/bin/bash
+. $(dirname $0)/../include.rc
+. $(dirname $0)/../volume.rc
+. $(dirname $0)/../fileio.rc
+
+
+# This test opens 100 fds and triggers graph switches to check if fsync
+# as part of graph-switch causes crash or not.
+
+cleanup;
+
+TEST glusterd
+TEST pidof glusterd
+TEST $CLI volume create $V0 replica 2 $H0:$B0/${V0}{0,1}
+TEST $CLI volume start $V0
+TEST glusterfs --volfile-id=/$V0 --volfile-server=$H0 $M0
+TEST touch $M0/{1..100}
+for i in {1..100}; do fd[$i]=`fd_available`; fd_open ${fd[$i]} 'w' $M0/$i; done
+TEST $CLI volume add-brick $V0 $H0:$B0/${V0}{2,3}
+TEST $CLI volume rebalance $V0 start force
+EXPECT_WITHIN 120 "completed" rebalance_status_field $V0
+TEST cat $M0/{1..100}
+for i in {1..100}; do fd_write ${fd[$i]} 'abc'; done
+TEST $CLI volume add-brick $V0 $H0:$B0/${V0}{4,5}
+TEST $CLI volume rebalance $V0 start force
+EXPECT_WITHIN 120 "completed" rebalance_status_field $V0
+for i in {1..100}; do fd_write ${fd[$i]} 'abc'; done
+TEST cat $M0/{1..100}
+cleanup