md: Check MD_RECOVERY_RUNNING as well as ->sync_thread.
A recent change to md started the ->sync_thread from a asynchronously
from a work_queue rather than synchronously. This means that there
can be a small window between the time when MD_RECOVERY_RUNNING is set
and when ->sync_thread is set.
So code that checks ->sync_thread might now conclude that the thread
has not been started and (because a lock is held) will not be started.
That is no longer the case.
Most of those places are best fixed by testing MD_RECOVERY_RUNNING
as well. To make this completely reliable, we wake_up(&resync_wait)
after clearing that flag as well as after clearing ->sync_thread.
Other places are better served by flushing the relevant workqueue
to ensure that that if the sync thread was starting, it has now
started. This is particularly best if we are about to stop the
sync thread.
Fixes: ac05f256691fe427a3e84c19261adb0b67dd73c0
Signed-off-by: NeilBrown <neilb@suse.de>
diff --git a/drivers/md/md.c b/drivers/md/md.c
index ad5d783..b97ac98 100644
--- a/drivers/md/md.c
+++ b/drivers/md/md.c
@@ -2695,7 +2695,8 @@
if (kstrtoull(buf, 10, &new_offset) < 0)
return -EINVAL;
- if (mddev->sync_thread)
+ if (mddev->sync_thread ||
+ test_bit(MD_RECOVERY_RUNNING,&mddev->recovery))
return -EBUSY;
if (new_offset == rdev->data_offset)
/* reset is always permitted */
@@ -3272,6 +3273,7 @@
*/
if (mddev->sync_thread ||
+ test_bit(MD_RECOVERY_RUNNING, &mddev->recovery) ||
mddev->reshape_position != MaxSector ||
mddev->sysfs_active)
return -EBUSY;
@@ -4026,6 +4028,7 @@
clear_bit(MD_RECOVERY_FROZEN, &mddev->recovery);
if (cmd_match(page, "idle") || cmd_match(page, "frozen")) {
+ flush_workqueue(md_misc_wq);
if (mddev->sync_thread) {
set_bit(MD_RECOVERY_INTR, &mddev->recovery);
md_reap_sync_thread(mddev);
@@ -5044,6 +5047,7 @@
static void __md_stop_writes(struct mddev *mddev)
{
set_bit(MD_RECOVERY_FROZEN, &mddev->recovery);
+ flush_workqueue(md_misc_wq);
if (mddev->sync_thread) {
set_bit(MD_RECOVERY_INTR, &mddev->recovery);
md_reap_sync_thread(mddev);
@@ -5104,19 +5108,22 @@
set_bit(MD_RECOVERY_FROZEN, &mddev->recovery);
md_wakeup_thread(mddev->thread);
}
- if (mddev->sync_thread) {
+ if (test_bit(MD_RECOVERY_RUNNING, &mddev->recovery))
set_bit(MD_RECOVERY_INTR, &mddev->recovery);
+ if (mddev->sync_thread)
/* Thread might be blocked waiting for metadata update
* which will now never happen */
wake_up_process(mddev->sync_thread->tsk);
- }
+
mddev_unlock(mddev);
- wait_event(resync_wait, mddev->sync_thread == NULL);
+ wait_event(resync_wait, !test_bit(MD_RECOVERY_RUNNING,
+ &mddev->recovery));
mddev_lock_nointr(mddev);
mutex_lock(&mddev->open_mutex);
if ((mddev->pers && atomic_read(&mddev->openers) > !!bdev) ||
mddev->sync_thread ||
+ test_bit(MD_RECOVERY_RUNNING, &mddev->recovery) ||
(bdev && !test_bit(MD_STILL_CLOSED, &mddev->flags))) {
printk("md: %s still in use.\n",mdname(mddev));
if (did_freeze) {
@@ -5162,20 +5169,24 @@
set_bit(MD_RECOVERY_FROZEN, &mddev->recovery);
md_wakeup_thread(mddev->thread);
}
- if (mddev->sync_thread) {
+ if (test_bit(MD_RECOVERY_RUNNING, &mddev->recovery))
set_bit(MD_RECOVERY_INTR, &mddev->recovery);
+ if (mddev->sync_thread)
/* Thread might be blocked waiting for metadata update
* which will now never happen */
wake_up_process(mddev->sync_thread->tsk);
- }
+
mddev_unlock(mddev);
- wait_event(resync_wait, mddev->sync_thread == NULL);
+ wait_event(resync_wait, (mddev->sync_thread == NULL &&
+ !test_bit(MD_RECOVERY_RUNNING,
+ &mddev->recovery)));
mddev_lock_nointr(mddev);
mutex_lock(&mddev->open_mutex);
if ((mddev->pers && atomic_read(&mddev->openers) > !!bdev) ||
mddev->sysfs_active ||
mddev->sync_thread ||
+ test_bit(MD_RECOVERY_RUNNING, &mddev->recovery) ||
(bdev && !test_bit(MD_STILL_CLOSED, &mddev->flags))) {
printk("md: %s still in use.\n",mdname(mddev));
mutex_unlock(&mddev->open_mutex);
@@ -5950,7 +5961,8 @@
* of each device. If num_sectors is zero, we find the largest size
* that fits.
*/
- if (mddev->sync_thread)
+ if (test_bit(MD_RECOVERY_RUNNING, &mddev->recovery) ||
+ mddev->sync_thread)
return -EBUSY;
if (mddev->ro)
return -EROFS;
@@ -5981,7 +5993,9 @@
if (raid_disks <= 0 ||
(mddev->max_disks && raid_disks >= mddev->max_disks))
return -EINVAL;
- if (mddev->sync_thread || mddev->reshape_position != MaxSector)
+ if (mddev->sync_thread ||
+ test_bit(MD_RECOVERY_RUNNING, &mddev->recovery) ||
+ mddev->reshape_position != MaxSector)
return -EBUSY;
rdev_for_each(rdev, mddev) {
@@ -7593,6 +7607,7 @@
clear_bit(MD_RECOVERY_REQUESTED, &mddev->recovery);
clear_bit(MD_RECOVERY_CHECK, &mddev->recovery);
clear_bit(MD_RECOVERY_RUNNING, &mddev->recovery);
+ wake_up(&resync_wait);
if (test_and_clear_bit(MD_RECOVERY_RECOVER,
&mddev->recovery))
if (mddev->sysfs_action)
@@ -7761,6 +7776,7 @@
not_running:
if (!mddev->sync_thread) {
clear_bit(MD_RECOVERY_RUNNING, &mddev->recovery);
+ wake_up(&resync_wait);
if (test_and_clear_bit(MD_RECOVERY_RECOVER,
&mddev->recovery))
if (mddev->sysfs_action)
@@ -7779,7 +7795,6 @@
/* resync has finished, collect result */
md_unregister_thread(&mddev->sync_thread);
- wake_up(&resync_wait);
if (!test_bit(MD_RECOVERY_INTR, &mddev->recovery) &&
!test_bit(MD_RECOVERY_REQUESTED, &mddev->recovery)) {
/* success...*/
@@ -7807,6 +7822,7 @@
clear_bit(MD_RECOVERY_RESHAPE, &mddev->recovery);
clear_bit(MD_RECOVERY_REQUESTED, &mddev->recovery);
clear_bit(MD_RECOVERY_CHECK, &mddev->recovery);
+ wake_up(&resync_wait);
/* flag recovery needed just to double check */
set_bit(MD_RECOVERY_NEEDED, &mddev->recovery);
sysfs_notify_dirent_safe(mddev->sysfs_action);