summaryrefslogtreecommitdiff
path: root/fs/ocfs2/dlm/dlmmaster.c
diff options
context:
space:
mode:
authorXue jiufei <xuejiufei@huawei.com>2014-06-23 13:22:09 -0700
committerLinus Torvalds <torvalds@linux-foundation.org>2014-06-23 16:47:45 -0700
commitac4fef4d23ed879a7fd11ab24ccd2e1464277e9a (patch)
tree6f5139c8f615683b688d27f3d5f628cc140465bc /fs/ocfs2/dlm/dlmmaster.c
parentb9aaac5a6b7d228435fcb80963d41c274406011b (diff)
ocfs2/dlm: do not purge lockres that is queued for assert master
When workqueue is delayed, it may occur that a lockres is purged while it is still queued for master assert. it may trigger BUG() as follows. N1 N2 dlm_get_lockres() ->dlm_do_master_requery is the master of lockres, so queue assert_master work dlm_thread() start running and purge the lockres dlm_assert_master_worker() send assert master message to other nodes receiving the assert_master message, set master to N2 dlmlock_remote() send create_lock message to N2, but receive DLM_IVLOCKID, if it is RECOVERY lockres, it triggers the BUG(). Another BUG() is triggered when N3 become the new master and send assert_master to N1, N1 will trigger the BUG() because owner doesn't match. So we should not purge lockres when it is queued for assert master. Signed-off-by: joyce.xue <xuejiufei@huawei.com> Reviewed-by: Mark Fasheh <mfasheh@suse.de> Cc: Joel Becker <jlbec@evilplan.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Diffstat (limited to 'fs/ocfs2/dlm/dlmmaster.c')
-rw-r--r--fs/ocfs2/dlm/dlmmaster.c43
1 files changed, 42 insertions, 1 deletions
diff --git a/fs/ocfs2/dlm/dlmmaster.c b/fs/ocfs2/dlm/dlmmaster.c
index 4f4b00e173f2..82abf0cc9a12 100644
--- a/fs/ocfs2/dlm/dlmmaster.c
+++ b/fs/ocfs2/dlm/dlmmaster.c
@@ -581,6 +581,7 @@ static void dlm_init_lockres(struct dlm_ctxt *dlm,
atomic_set(&res->asts_reserved, 0);
res->migration_pending = 0;
res->inflight_locks = 0;
+ res->inflight_assert_workers = 0;
res->dlm = dlm;
@@ -683,6 +684,43 @@ void dlm_lockres_drop_inflight_ref(struct dlm_ctxt *dlm,
wake_up(&res->wq);
}
+void __dlm_lockres_grab_inflight_worker(struct dlm_ctxt *dlm,
+ struct dlm_lock_resource *res)
+{
+ assert_spin_locked(&res->spinlock);
+ res->inflight_assert_workers++;
+ mlog(0, "%s:%.*s: inflight assert worker++: now %u\n",
+ dlm->name, res->lockname.len, res->lockname.name,
+ res->inflight_assert_workers);
+}
+
+static void dlm_lockres_grab_inflight_worker(struct dlm_ctxt *dlm,
+ struct dlm_lock_resource *res)
+{
+ spin_lock(&res->spinlock);
+ __dlm_lockres_grab_inflight_worker(dlm, res);
+ spin_unlock(&res->spinlock);
+}
+
+static void __dlm_lockres_drop_inflight_worker(struct dlm_ctxt *dlm,
+ struct dlm_lock_resource *res)
+{
+ assert_spin_locked(&res->spinlock);
+ BUG_ON(res->inflight_assert_workers == 0);
+ res->inflight_assert_workers--;
+ mlog(0, "%s:%.*s: inflight assert worker--: now %u\n",
+ dlm->name, res->lockname.len, res->lockname.name,
+ res->inflight_assert_workers);
+}
+
+static void dlm_lockres_drop_inflight_worker(struct dlm_ctxt *dlm,
+ struct dlm_lock_resource *res)
+{
+ spin_lock(&res->spinlock);
+ __dlm_lockres_drop_inflight_worker(dlm, res);
+ spin_unlock(&res->spinlock);
+}
+
/*
* lookup a lock resource by name.
* may already exist in the hashtable.
@@ -1603,7 +1641,8 @@ send_response:
mlog(ML_ERROR, "failed to dispatch assert master work\n");
response = DLM_MASTER_RESP_ERROR;
dlm_lockres_put(res);
- }
+ } else
+ dlm_lockres_grab_inflight_worker(dlm, res);
} else {
if (res)
dlm_lockres_put(res);
@@ -2118,6 +2157,8 @@ static void dlm_assert_master_worker(struct dlm_work_item *item, void *data)
dlm_lockres_release_ast(dlm, res);
put:
+ dlm_lockres_drop_inflight_worker(dlm, res);
+
dlm_lockres_put(res);
mlog(0, "finished with dlm_assert_master_worker\n");