From: Simon Guinot <simon.guinot@sequanux.org>
To: Neil Brown <neilb@suse.de>
Cc: linux-raid@vger.kernel.org, linux-kernel@vger.kernel.org,
"Rémi Rérolle" <remi.rerolle@seagate.com>,
"Simon Guinot" <simon.guinot@sequanux.org>
Subject: [PATCH 1/2] md: move up mddev locking functions
Date: Tue, 11 Mar 2014 20:12:09 +0100 [thread overview]
Message-ID: <1394565130-24233-2-git-send-email-simon.guinot@sequanux.org> (raw)
In-Reply-To: <1394565130-24233-1-git-send-email-simon.guinot@sequanux.org>
This patch slightly moves up the mddev locking functions. This allows
to use them from mddev_suspend().
Signed-off-by: Simon Guinot <simon.guinot@sequanux.org>
---
drivers/md/md.c | 135 ++++++++++++++++++++++++++++----------------------------
1 file changed, 67 insertions(+), 68 deletions(-)
diff --git a/drivers/md/md.c b/drivers/md/md.c
index 4ad5cc4e63e8..fb4296adae80 100644
--- a/drivers/md/md.c
+++ b/drivers/md/md.c
@@ -218,6 +218,73 @@ static void md_new_event_inintr(struct mddev *mddev)
static LIST_HEAD(all_mddevs);
static DEFINE_SPINLOCK(all_mddevs_lock);
+static inline int __must_check mddev_lock(struct mddev * mddev)
+{
+ return mutex_lock_interruptible(&mddev->reconfig_mutex);
+}
+
+/* Sometimes we need to take the lock in a situation where
+ * failure due to interrupts is not acceptable.
+ */
+static inline void mddev_lock_nointr(struct mddev * mddev)
+{
+ mutex_lock(&mddev->reconfig_mutex);
+}
+
+static inline int mddev_is_locked(struct mddev *mddev)
+{
+ return mutex_is_locked(&mddev->reconfig_mutex);
+}
+
+static inline int mddev_trylock(struct mddev * mddev)
+{
+ return mutex_trylock(&mddev->reconfig_mutex);
+}
+
+static struct attribute_group md_redundancy_group;
+
+static void mddev_unlock(struct mddev * mddev)
+{
+ if (mddev->to_remove) {
+ /* These cannot be removed under reconfig_mutex as
+ * an access to the files will try to take reconfig_mutex
+ * while holding the file unremovable, which leads to
+ * a deadlock.
+ * So hold set sysfs_active while the remove in happeing,
+ * and anything else which might set ->to_remove or my
+ * otherwise change the sysfs namespace will fail with
+ * -EBUSY if sysfs_active is still set.
+ * We set sysfs_active under reconfig_mutex and elsewhere
+ * test it under the same mutex to ensure its correct value
+ * is seen.
+ */
+ struct attribute_group *to_remove = mddev->to_remove;
+ mddev->to_remove = NULL;
+ mddev->sysfs_active = 1;
+ mutex_unlock(&mddev->reconfig_mutex);
+
+ if (mddev->kobj.sd) {
+ if (to_remove != &md_redundancy_group)
+ sysfs_remove_group(&mddev->kobj, to_remove);
+ if (mddev->pers == NULL ||
+ mddev->pers->sync_request == NULL) {
+ sysfs_remove_group(&mddev->kobj, &md_redundancy_group);
+ if (mddev->sysfs_action)
+ sysfs_put(mddev->sysfs_action);
+ mddev->sysfs_action = NULL;
+ }
+ }
+ mddev->sysfs_active = 0;
+ } else
+ mutex_unlock(&mddev->reconfig_mutex);
+
+ /* As we've dropped the mutex we need a spinlock to
+ * make sure the thread doesn't disappear
+ */
+ spin_lock(&pers_lock);
+ md_wakeup_thread(mddev->thread);
+ spin_unlock(&pers_lock);
+}
/*
* iterates through all used mddevs in the system.
@@ -562,74 +629,6 @@ static struct mddev * mddev_find(dev_t unit)
goto retry;
}
-static inline int __must_check mddev_lock(struct mddev * mddev)
-{
- return mutex_lock_interruptible(&mddev->reconfig_mutex);
-}
-
-/* Sometimes we need to take the lock in a situation where
- * failure due to interrupts is not acceptable.
- */
-static inline void mddev_lock_nointr(struct mddev * mddev)
-{
- mutex_lock(&mddev->reconfig_mutex);
-}
-
-static inline int mddev_is_locked(struct mddev *mddev)
-{
- return mutex_is_locked(&mddev->reconfig_mutex);
-}
-
-static inline int mddev_trylock(struct mddev * mddev)
-{
- return mutex_trylock(&mddev->reconfig_mutex);
-}
-
-static struct attribute_group md_redundancy_group;
-
-static void mddev_unlock(struct mddev * mddev)
-{
- if (mddev->to_remove) {
- /* These cannot be removed under reconfig_mutex as
- * an access to the files will try to take reconfig_mutex
- * while holding the file unremovable, which leads to
- * a deadlock.
- * So hold set sysfs_active while the remove in happeing,
- * and anything else which might set ->to_remove or my
- * otherwise change the sysfs namespace will fail with
- * -EBUSY if sysfs_active is still set.
- * We set sysfs_active under reconfig_mutex and elsewhere
- * test it under the same mutex to ensure its correct value
- * is seen.
- */
- struct attribute_group *to_remove = mddev->to_remove;
- mddev->to_remove = NULL;
- mddev->sysfs_active = 1;
- mutex_unlock(&mddev->reconfig_mutex);
-
- if (mddev->kobj.sd) {
- if (to_remove != &md_redundancy_group)
- sysfs_remove_group(&mddev->kobj, to_remove);
- if (mddev->pers == NULL ||
- mddev->pers->sync_request == NULL) {
- sysfs_remove_group(&mddev->kobj, &md_redundancy_group);
- if (mddev->sysfs_action)
- sysfs_put(mddev->sysfs_action);
- mddev->sysfs_action = NULL;
- }
- }
- mddev->sysfs_active = 0;
- } else
- mutex_unlock(&mddev->reconfig_mutex);
-
- /* As we've dropped the mutex we need a spinlock to
- * make sure the thread doesn't disappear
- */
- spin_lock(&pers_lock);
- md_wakeup_thread(mddev->thread);
- spin_unlock(&pers_lock);
-}
-
static struct md_rdev * find_rdev_nr(struct mddev *mddev, int nr)
{
struct md_rdev *rdev;
--
1.8.5.3
next prev parent reply other threads:[~2014-03-11 19:12 UTC|newest]
Thread overview: 5+ messages / expand[flat|nested] mbox.gz Atom feed top
2014-03-11 19:12 [PATCH 0/2] md: fix fix deadlock while suspending RAID array Simon Guinot
2014-03-11 19:12 ` Simon Guinot [this message]
2014-03-11 19:12 ` [PATCH 2/2] md: " Simon Guinot
2014-03-19 23:02 ` NeilBrown
2014-03-18 14:48 ` [PATCH 0/2] md: fix " Simon Guinot
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=1394565130-24233-2-git-send-email-simon.guinot@sequanux.org \
--to=simon.guinot@sequanux.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-raid@vger.kernel.org \
--cc=neilb@suse.de \
--cc=remi.rerolle@seagate.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).