The patch titled
Subject: ocfs2: old mle put and release after the function dlm_add_migration_mle called
has been added to the -mm tree. Its filename is
ocfs2-old-mle-put-and-release-after-the-function-dlm_add_migration_mle-called.patch
This patch should soon appear at
http://ozlabs.org/~akpm/mmots/broken-out/ocfs2-old-mle-put-and-release-after-the-function-dlm_add_migration_mle-called.patch
and later at
http://ozlabs.org/~akpm/mmotm/broken-out/ocfs2-old-mle-put-and-release-after-the-function-dlm_add_migration_mle-called.patch
Before you just go and hit "reply", please:
a) Consider who else should be cc'ed
b) Prefer to cc a suitable mailing list as well
c) Ideally: find the original patch on the mailing list and do a
reply-to-all to that, adding suitable additional cc's
*** Remember to use Documentation/SubmitChecklist when testing your code ***
The -mm tree is included into linux-next and is updated
there every 3-4 working days
------------------------------------------------------
From: Guozhonghua <guozhonghua@xxxxxxx>
Subject: ocfs2: old mle put and release after the function dlm_add_migration_mle called
If the old mle is found after the dlm_add_migration_mle called, it should
be put once. If the return value is not - EEXIST and its type is BLOCK,
it should be put again to release it to avoid memory leak, for it had been
unhashed from the map.
Link: http://lkml.kernel.org/r/71604351584F6A4EBAE558C676F37CA4A3D4B7FE@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
Signed-off-by: Guozhonghua <guozhonghua@xxxxxxx>
Cc: Mark Fasheh <mfasheh@xxxxxxxxxxx>
Cc: Joel Becker <jlbec@xxxxxxxxxxxx>
Cc: Junxiao Bi <junxiao.bi@xxxxxxxxxx>
Cc: Joseph Qi <joseph.qi@xxxxxxxxxx>
Signed-off-by: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx>
---
fs/ocfs2/dlm/dlmmaster.c | 62 ++++++++++++++++++++++++++-----------
1 file changed, 44 insertions(+), 18 deletions(-)
diff -puN fs/ocfs2/dlm/dlmmaster.c~ocfs2-old-mle-put-and-release-after-the-function-dlm_add_migration_mle-called fs/ocfs2/dlm/dlmmaster.c
--- a/fs/ocfs2/dlm/dlmmaster.c~ocfs2-old-mle-put-and-release-after-the-function-dlm_add_migration_mle-called
+++ a/fs/ocfs2/dlm/dlmmaster.c
@@ -2612,20 +2612,45 @@ static int dlm_migrate_lockres(struct dl
spin_lock(&dlm->master_lock);
ret = dlm_add_migration_mle(dlm, res, mle, &oldmle, name,
namelen, target, dlm->node_num);
+ if (ret == -EEXIST) {
+ if(oldmle)
+ __dlm_put_mle(oldmle);
+
+ spin_unlock(&dlm->master_lock);
+ spin_unlock(&dlm->spinlock);
+ mlog(0, "another process is already migrating it\n");
+ goto fail;
+ }
+
+ /* If an old one mle found, it should be put. if its type is BLOCK,
+ * it should be put again. Because it had been unhasded from the map
+ * in the function dlm_add_migration_mle.
+ * otherwise the memory will be leaked. It will not found it again from
+ * the hash map.
+ */
+ if (oldmle) {
+ /* master is known, detach if not already detached */
+ __dlm_mle_detach_hb_events(dlm, oldmle);
+ __dlm_put_mle(oldmle);
+
+ /* if the type of the mle is BLOCK, should put it once for release.
+ * otherwise memory leak may be caused because oldmle had been unhashed
+ * from the hash map, it will not be found anymore.
+ */
+ if (oldmle->type == DLM_MLE_BLOCK)
+ __dlm_put_mle(oldmle);
+ }
+
/* get an extra reference on the mle.
* otherwise the assert_master from the new
* master will destroy this.
*/
dlm_get_mle_inuse(mle);
+ mle_added = 1;
+
spin_unlock(&dlm->master_lock);
spin_unlock(&dlm->spinlock);
- if (ret == -EEXIST) {
- mlog(0, "another process is already migrating it\n");
- goto fail;
- }
- mle_added = 1;
-
/*
* set the MIGRATING flag and flush asts
* if we fail after this we need to re-dirty the lockres
@@ -2642,12 +2667,6 @@ static int dlm_migrate_lockres(struct dl
}
fail:
- if (ret != -EEXIST && oldmle) {
- /* master is known, detach if not already detached */
- dlm_mle_detach_hb_events(dlm, oldmle);
- dlm_put_mle(oldmle);
- }
-
if (ret < 0) {
if (mle_added) {
dlm_mle_detach_hb_events(dlm, mle);
@@ -3182,16 +3201,23 @@ int dlm_migrate_request_handler(struct o
if (ret < 0)
kmem_cache_free(dlm_mle_cache, mle);
+ /* If an old one mle found, it should be put. if its type is BLOCK,
+ * it should be put again. Because it had been unhasded from the map
+ * in the function dlm_add_migration_mle.
+ * otherwise the memory will be leaked. It will not found it again from
+ * the hash map.
+ */
+ if (oldmle) {
+ __dlm_mle_detach_hb_events(dlm, oldmle);
+ __dlm_put_mle(oldmle);
+ if (ret >= 0 && oldmle->type == DLM_MLE_BLOCK)
+ __dlm_put_mle(oldmle);
+ }
+
spin_unlock(&dlm->master_lock);
unlock:
spin_unlock(&dlm->spinlock);
- if (oldmle) {
- /* master is known, detach if not already detached */
- dlm_mle_detach_hb_events(dlm, oldmle);
- dlm_put_mle(oldmle);
- }
-
if (res)
dlm_lockres_put(res);
leave:
_
Patches currently in -mm which might be from guozhonghua@xxxxxxx are
ocfs2-delete-redundant-code-and-set-the-node-bit-into-maybe_map-directly.patch
ocfs2-old-mle-put-and-release-after-the-function-dlm_add_migration_mle-called.patch
--
To unsubscribe from this list: send the line "unsubscribe mm-commits" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
