md: Change handling of save_raid_disk and metadata update during recovery.

[~andy/linux] / drivers / md / md.c

diff --git a/drivers/md/md.c b/drivers/md/md.c
index 2a456a5d59a8092efa09ef755c175f89b6caf63f..539f08885e7fb2d85de65d25e558d4ba2678e30c 100644 (file)
--- a/drivers/md/md.c
+++ b/drivers/md/md.c
@@ -1183,6 +1183,7 @@ static int super_90_validate(struct mddev *mddev, struct md_rdev *rdev)
                            desc->raid_disk < mddev->raid_disks */) {
                        set_bit(In_sync, &rdev->flags);
                        rdev->raid_disk = desc->raid_disk;
+                       rdev->saved_raid_disk = desc->raid_disk;
                } else if (desc->state & (1<<MD_DISK_ACTIVE)) {
                        /* active but not in sync implies recovery up to
                         * reshape position.  We don't know exactly where
@@ -1681,10 +1682,14 @@ static int super_1_validate(struct mddev *mddev, struct md_rdev *rdev)
                        set_bit(Faulty, &rdev->flags);
                        break;
                default:
+                       rdev->saved_raid_disk = role;
                        if ((le32_to_cpu(sb->feature_map) &
-                            MD_FEATURE_RECOVERY_OFFSET))
+                            MD_FEATURE_RECOVERY_OFFSET)) {
                                rdev->recovery_offset = le64_to_cpu(sb->recovery_offset);
-                       else
+                               if (!(le32_to_cpu(sb->feature_map) &
+                                     MD_FEATURE_RECOVERY_BITMAP))
+                                       rdev->saved_raid_disk = -1;
+                       } else
                                set_bit(In_sync, &rdev->flags);
                        rdev->raid_disk = role;
                        break;
@@ -1746,6 +1751,9 @@ static void super_1_sync(struct mddev *mddev, struct md_rdev *rdev)
                        cpu_to_le32(MD_FEATURE_RECOVERY_OFFSET);
                sb->recovery_offset =
                        cpu_to_le64(rdev->recovery_offset);
+               if (rdev->saved_raid_disk >= 0 && mddev->bitmap)
+                       sb->feature_map |=
+                               cpu_to_le32(MD_FEATURE_RECOVERY_BITMAP);
        }
        if (test_bit(Replacement, &rdev->flags))
                sb->feature_map |=
@@ -2487,8 +2495,7 @@ repeat:
                if (rdev->sb_loaded != 1)
                        continue; /* no noise on spare devices */
 
-               if (!test_bit(Faulty, &rdev->flags) &&
-                   rdev->saved_raid_disk == -1) {
+               if (!test_bit(Faulty, &rdev->flags)) {
                        md_super_write(mddev,rdev,
                                       rdev->sb_start, rdev->sb_size,
                                       rdev->sb_page);
@@ -2504,11 +2511,9 @@ repeat:
                                rdev->badblocks.size = 0;
                        }
 
-               } else if (test_bit(Faulty, &rdev->flags))
+               } else
                        pr_debug("md: %s (skipping faulty)\n",
                                 bdevname(rdev->bdev, b));
-               else
-                       pr_debug("(skipping incremental s/r ");
 
                if (mddev->level == LEVEL_MULTIPATH)
                        /* only need to write one superblock... */
@@ -2624,6 +2629,8 @@ state_store(struct md_rdev *rdev, const char *buf, size_t len)
         *  blocked - sets the Blocked flags
         *  -blocked - clears the Blocked and possibly simulates an error
         *  insync - sets Insync providing device isn't active
+        *  -insync - clear Insync for a device with a slot assigned,
+        *            so that it gets rebuilt based on bitmap
         *  write_error - sets WriteErrorSeen
         *  -write_error - clears WriteErrorSeen
         */
@@ -2672,6 +2679,11 @@ state_store(struct md_rdev *rdev, const char *buf, size_t len)
        } else if (cmd_match(buf, "insync") && rdev->raid_disk == -1) {
                set_bit(In_sync, &rdev->flags);
                err = 0;
+       } else if (cmd_match(buf, "-insync") && rdev->raid_disk >= 0) {
+               clear_bit(In_sync, &rdev->flags);
+               rdev->saved_raid_disk = rdev->raid_disk;
+               rdev->raid_disk = -1;
+               err = 0;
        } else if (cmd_match(buf, "write_error")) {
                set_bit(WriteErrorSeen, &rdev->flags);
                err = 0;
@@ -5780,6 +5792,7 @@ static int add_new_disk(struct mddev * mddev, mdu_disk_info_t *info)
                                clear_bit(Bitmap_sync, &rdev->flags);
                        } else
                                rdev->raid_disk = -1;
+                       rdev->saved_raid_disk = rdev->raid_disk;
                } else
                        super_types[mddev->major_version].
                                validate_super(mddev, rdev);
@@ -5792,11 +5805,6 @@ static int add_new_disk(struct mddev * mddev, mdu_disk_info_t *info)
                        return -EINVAL;
                }
 
-               if (test_bit(In_sync, &rdev->flags))
-                       rdev->saved_raid_disk = rdev->raid_disk;
-               else
-                       rdev->saved_raid_disk = -1;
-
                clear_bit(In_sync, &rdev->flags); /* just to be sure */
                if (info->state & (1<<MD_DISK_WRITEMOSTLY))
                        set_bit(WriteMostly, &rdev->flags);
@@ -7948,14 +7956,10 @@ void md_reap_sync_thread(struct mddev *mddev)
                mddev->pers->finish_reshape(mddev);
 
        /* If array is no-longer degraded, then any saved_raid_disk
-        * information must be scrapped.  Also if any device is now
-        * In_sync we must scrape the saved_raid_disk for that device
-        * do the superblock for an incrementally recovered device
-        * written out.
+        * information must be scrapped.
         */
-       rdev_for_each(rdev, mddev)
-               if (!mddev->degraded ||
-                   test_bit(In_sync, &rdev->flags))
+       if (!mddev->degraded)
+               rdev_for_each(rdev, mddev)
                        rdev->saved_raid_disk = -1;
 
        md_update_sb(mddev, 1);