From: NeilBrown <neilb@suse.de>
Date: Fri, 13 Nov 2009 06:40:48 +0000 (+1100)
Subject: md: allow v0.91 metadata to record devices as being active but not in-sync.
X-Git-Tag: v2.6.32-rc8~38^2~2
X-Git-Url: http://ftp.safe.ca/?p=safe%2Fjmp%2Flinux-2.6;a=commitdiff_plain;h=0261cd9f1cb42fa44ece314d27868d83742bdf03

md: allow v0.91 metadata to record devices as being active but not in-sync.

This is a combination that didn't really make sense before.
However when a reshape is converting e.g. raid5 -> raid6, the extra
device is not fully in-sync, but is certainly active and contains
important data.
So allow that start to be meaningful and in particular get
the 'recovery_offset' value (which is needed for any non-in-sync
active device) from the reshape_position.

Signed-off-by: NeilBrown <neilb@suse.de>
---

diff --git a/drivers/md/md.c b/drivers/md/md.c
index 01b9a0f..b182f86 100644
--- a/drivers/md/md.c
+++ b/drivers/md/md.c
@@ -944,6 +944,14 @@ static int super_90_validate(mddev_t *mddev, mdk_rdev_t *rdev)
 			    desc->raid_disk < mddev->raid_disks */) {
 			set_bit(In_sync, &rdev->flags);
 			rdev->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
+			 * that is, so set to zero for now */
+			if (mddev->minor_version >= 91) {
+				rdev->recovery_offset = 0;
+				rdev->raid_disk = desc->raid_disk;
+			}
 		}
 		if (desc->state & (1<<MD_DISK_WRITEMOSTLY))
 			set_bit(WriteMostly, &rdev->flags);
@@ -1032,8 +1040,19 @@ static void super_90_sync(mddev_t *mddev, mdk_rdev_t *rdev)
 	list_for_each_entry(rdev2, &mddev->disks, same_set) {
 		mdp_disk_t *d;
 		int desc_nr;
-		if (rdev2->raid_disk >= 0 && test_bit(In_sync, &rdev2->flags)
-		    && !test_bit(Faulty, &rdev2->flags))
+		int is_active = test_bit(In_sync, &rdev2->flags);
+
+		if (rdev2->raid_disk >= 0 &&
+		    sb->minor_version >= 91)
+			/* we have nowhere to store the recovery_offset,
+			 * but if it is not below the reshape_position,
+			 * we can piggy-back on that.
+			 */
+			is_active = 1;
+		if (rdev2->raid_disk < 0 ||
+		    test_bit(Faulty, &rdev2->flags))
+			is_active = 0;
+		if (is_active)
 			desc_nr = rdev2->raid_disk;
 		else
 			desc_nr = next_spare++;
@@ -1043,16 +1062,16 @@ static void super_90_sync(mddev_t *mddev, mdk_rdev_t *rdev)
 		d->number = rdev2->desc_nr;
 		d->major = MAJOR(rdev2->bdev->bd_dev);
 		d->minor = MINOR(rdev2->bdev->bd_dev);
-		if (rdev2->raid_disk >= 0 && test_bit(In_sync, &rdev2->flags)
-		    && !test_bit(Faulty, &rdev2->flags))
+		if (is_active)
 			d->raid_disk = rdev2->raid_disk;
 		else
 			d->raid_disk = rdev2->desc_nr; /* compatibility */
 		if (test_bit(Faulty, &rdev2->flags))
 			d->state = (1<<MD_DISK_FAULTY);
-		else if (test_bit(In_sync, &rdev2->flags)) {
+		else if (is_active) {
 			d->state = (1<<MD_DISK_ACTIVE);
-			d->state |= (1<<MD_DISK_SYNC);
+			if (test_bit(In_sync, &rdev2->flags))
+				d->state |= (1<<MD_DISK_SYNC);
 			active++;
 			working++;
 		} else {