md-cluster/raid10: support add disk under grow mode

For clustered raid10 scenario, we need to let all the nodes
know about that a new disk is added to the array, and the
reshape caused by add new member just need to be happened in
one node, but other nodes should know about the change.

Since reshape means read data from somewhere (which is already
used by array) and write data to unused region. Obviously, it
is awful if one node is reading data from address while another
node is writing to the same address. Considering we have
implemented suspend writes in the resyncing area, so we can
just broadcast the reading address to other nodes to avoid the
trouble.

For master node, it would call reshape_request then update sb
during the reshape period. To avoid above trouble, we call
resync_info_update to send RESYNC message in reshape_request.

Then from slave node's view, it receives two type messages:
1. RESYNCING message
Slave node add the address (where master node reading data from)
to suspend list.

2. METADATA_UPDATED message
Once slave nodes know the reshaping is started in master node,
it is time to update reshape position and call start_reshape to
follow master node's step. After reshape is done, only reshape
position is need to be updated, so the majority task of reshaping
is happened on the master node.

Reviewed-by: NeilBrown <neilb@suse.com>
Signed-off-by: Guoqing Jiang <gqjiang@suse.com>
Signed-off-by: Shaohua Li <shli@fb.com>

3 files changed, 59 insertions, 0 deletions

diff --git a/drivers/md/md.c b/drivers/md/md.c
index 4c0f3e0331d5..e07096c4ff20 100644
--- a/drivers/md/md.c
+++ b/drivers/md/md.c
@@ -9230,6 +9230,30 @@ static void check_sb_changes(struct mddev *mddev, struct md_rdev *rdev)
         if (mddev->raid_disks != le32_to_cpu(sb->raid_disks))
                 update_raid_disks(mddev, le32_to_cpu(sb->raid_disks));
 
+        /*
+         * Since mddev->delta_disks has already updated in update_raid_disks,
+         * so it is time to check reshape.
+         */
+        if (test_bit(MD_RESYNCING_REMOTE, &mddev->recovery) &&
+            (le32_to_cpu(sb->feature_map) & MD_FEATURE_RESHAPE_ACTIVE)) {
+                /*
+                 * reshape is happening in the remote node, we need to
+                 * update reshape_position and call start_reshape.
+                 */
+                mddev->reshape_position = sb->reshape_position;
+                if (mddev->pers->update_reshape_pos)
+                        mddev->pers->update_reshape_pos(mddev);
+                if (mddev->pers->start_reshape)
+                        mddev->pers->start_reshape(mddev);
+        } else if (test_bit(MD_RESYNCING_REMOTE, &mddev->recovery) &&
+                   mddev->reshape_position != MaxSector &&
+                   !(le32_to_cpu(sb->feature_map) & MD_FEATURE_RESHAPE_ACTIVE)) {
+                /* reshape is just done in another node. */
+                mddev->reshape_position = MaxSector;
+                if (mddev->pers->update_reshape_pos)
+                        mddev->pers->update_reshape_pos(mddev);
+        }
+
         /* Finally set the event to be up to date */
         mddev->events = le64_to_cpu(sb->events);
 }
diff --git a/drivers/md/md.h b/drivers/md/md.h
index 8afd6bfdbfb9..c52afb52c776 100644
--- a/drivers/md/md.h
+++ b/drivers/md/md.h
@@ -557,6 +557,7 @@ struct md_personality
         int (*check_reshape) (struct mddev *mddev);
         int (*start_reshape) (struct mddev *mddev);
         void (*finish_reshape) (struct mddev *mddev);
+        void (*update_reshape_pos) (struct mddev *mddev);
         /* quiesce suspends or resumes internal processing.
          * 1 - stop new actions and wait for action io to complete
          * 0 - return to normal behaviour
diff --git a/drivers/md/raid10.c b/drivers/md/raid10.c
index 72e52921c545..1edd58a3098b 100644
--- a/drivers/md/raid10.c
+++ b/drivers/md/raid10.c
@@ -4605,6 +4605,32 @@ read_more:
         r10_bio->master_bio = read_bio;
         r10_bio->read_slot = r10_bio->devs[r10_bio->read_slot].devnum;
 
+        /*
+         * Broadcast RESYNC message to other nodes, so all nodes would not
+         * write to the region to avoid conflict.
+        */
+        if (mddev_is_clustered(mddev) && conf->cluster_sync_high <= sector_nr) {
+                struct mdp_superblock_1 *sb = NULL;
+                int sb_reshape_pos = 0;
+
+                conf->cluster_sync_low = sector_nr;
+                conf->cluster_sync_high = sector_nr + CLUSTER_RESYNC_WINDOW_SECTORS;
+                sb = page_address(rdev->sb_page);
+                if (sb) {
+                        sb_reshape_pos = le64_to_cpu(sb->reshape_position);
+                        /*
+                         * Set cluster_sync_low again if next address for array
+                         * reshape is less than cluster_sync_low. Since we can't
+                         * update cluster_sync_low until it has finished reshape.
+                         */
+                        if (sb_reshape_pos < conf->cluster_sync_low)
+                                conf->cluster_sync_low = sb_reshape_pos;
+                }
+
+                md_cluster_ops->resync_info_update(mddev, conf->cluster_sync_low,
+                                                          conf->cluster_sync_high);
+        }
+
         /* Now find the locations in the new layout */
         __raid10_find_phys(&conf->geo, r10_bio);
 
@@ -4756,6 +4782,13 @@ static void end_reshape(struct r10conf *conf)
         conf->fullsync = 0;
 }
 
+static void raid10_update_reshape_pos(struct mddev *mddev)
+{
+        struct r10conf *conf = mddev->private;
+
+        conf->reshape_progress = mddev->reshape_position;
+}
+
 static int handle_reshape_read_error(struct mddev *mddev,
                                      struct r10bio *r10_bio)
 {
@@ -4924,6 +4957,7 @@ static struct md_personality raid10_personality =
         .check_reshape  = raid10_check_reshape,
         .start_reshape  = raid10_start_reshape,
         .finish_reshape = raid10_finish_reshape,
+        .update_reshape_pos = raid10_update_reshape_pos,
         .congested      = raid10_congested,
 };