[PATCH V1] 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>

diff --git a/drivers/md/md.c b/drivers/md/md.c
index 5c330412979021cdbd0e378ae20dda0fd4706630..6fd7dc8606c241befa437a697bd5f88dc5f0fcc2 100644 (file)
--- a/drivers/md/md.c
+++ b/drivers/md/md.c
@@ -7269,6 +7269,7 @@ 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;
@@ -7277,6 +7278,7 @@ 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;
@@ -7436,6 +7438,7 @@ 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) {
@@ -7447,6 +7450,7 @@ 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
                            )) {
@@ -7457,6 +7461,7 @@ 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");
                }