md: Update checkpoint of resync/recovery based on time.

md will current only only checkpoint recovery or resync ever 1/16th
of the device size.  As devices get larger this can become a long time
an so a lot of work that might need to be duplicated after a shutdown.

So add a time-based checkpoint.  Every 5 minutes limits the amount of
duplicated effort to at most 5 minutes, and has almost zero impact on
performance.

[changelog entry re-written by NeilBrown]

Signed-off-by: Jianpeng Ma <majianpeng@gmail.com>
Signed-off-by: NeilBrown <neilb@suse.de>

1 files changed, 5 insertions, 0 deletions

diff --git a/drivers/md/md.c b/drivers/md/md.c
index f97083476475..6aefa4434819 100644
--- a/drivers/md/md.c
+++ b/drivers/md/md.c
@@ -7277,6 +7277,7 @@ EXPORT_SYMBOL_GPL(md_allow_write);
 
 #define SYNC_MARKS      10
 #define SYNC_MARK_STEP  (3*HZ)
+#define UPDATE_FREQUENCY (5*60*HZ)
 void md_do_sync(struct md_thread *thread)
 {
         struct mddev *mddev = thread->mddev;
@@ -7285,6 +7286,7 @@ void md_do_sync(struct md_thread *thread)
                  window;
         sector_t max_sectors,j, io_sectors;
         unsigned long mark[SYNC_MARKS];
+        unsigned long update_time;
         sector_t mark_cnt[SYNC_MARKS];
         int last_mark,m;
         struct list_head *tmp;
@@ -7444,6 +7446,7 @@ void md_do_sync(struct md_thread *thread)
         mddev->curr_resync_completed = j;
         sysfs_notify(&mddev->kobj, NULL, "sync_completed");
         md_new_event(mddev);
+        update_time = jiffies;
 
         blk_start_plug(&plug);
         while (j < max_sectors) {
@@ -7455,6 +7458,7 @@ void md_do_sync(struct md_thread *thread)
                     ((mddev->curr_resync > mddev->curr_resync_completed &&
                       (mddev->curr_resync - mddev->curr_resync_completed)
                       > (max_sectors >> 4)) ||
+                     time_after_eq(jiffies, update_time + UPDATE_FREQUENCY) ||
                      (j - mddev->curr_resync_completed)*2
                      >= mddev->resync_max - mddev->curr_resync_completed
                             )) {
@@ -7465,6 +7469,7 @@ void md_do_sync(struct md_thread *thread)
                         if (test_bit(MD_RECOVERY_SYNC, &mddev->recovery) &&
                             j > mddev->recovery_cp)
                                 mddev->recovery_cp = j;
+                        update_time = jiffies;
                         set_bit(MD_CHANGE_CLEAN, &mddev->flags);
                         sysfs_notify(&mddev->kobj, NULL, "sync_completed");
                 }