xfs_repair: allow filesystems with a single AG

Currently xfs_repair bails out on a filesystem with just a single AG. But that's a perfectly valid configureation, so we should allow it.

Skip the geomery validation because we simply can't do it if we don't have a secondary SB, and make sure to take the internal log into account when guestimating the first inode cluster.

I'll also cook up a testcase for repair on single AG filesystems.

Quoted 8 lines · Christoph Hellwig

Currently xfs_repair bails out on a filesystem with just a single AG. But that's a perfectly valid configureation, so we should allow it.

Skip the geomery validation because we simply can't do it if we don't have a secondary SB, and make sure to take the internal log into account when guestimating the first inode cluster.

I'll also cook up a testcase for repair on single AG filesystems.

While I think we should certainly allow this, what's the worst-case scenario for a corrupted superblock when we can't validate it and continue with repair?

I wonder if something like

# xfs_repair --allow-single-sb

should be required, with some man page docs suggesting a run with -n first etc to be sure that garbled geometry doesn't trash the whole thing...?

Currently xfs_repair bails out on a filesystem with just a single AG. But that's a perfectly valid configureation, so we should allow it.

Skip the geomery validation because we simply can't do it if we don't have a secondary SB, and make sure to take the internal log into account when guestimating the first inode cluster.

I'll also cook up a testcase for repair on single AG filesystems.

Signed-off-by Christoph Hellwig <[email protected]>

Reported-By Arkadiusz Miskiewicz <[email protected]>

Index: xfsprogs/repair/sb.c
===================================================================

repair/sb.c +10 -7

@@ -773,14 +773,17 @@ verify_set_primary_sb(xfs_sb_t		*rsb,
 		break;
 	case 1:
 		/*
-		 * just report the geometry info and get out.
-		 * refuse to run further unless the force (-F)
-		 * option is in effect.
+		 * If we only have a single allocation group there is no
+		 * secondary superblock that we can use to verify the geometry
+		 * information.  Not much we can do here, as we don't want
+		 * to prevent the user from checking the filesystem.
+		 *
+		 * XXX(hch): We should allow putting a secondary superblock
+		 *	     into the last sector of a filesystem to so that
+		 *	     we can still have a backup for single allocation
+		 *	     group filesystems.
 		 */
-		if (!force_geo)  {
-			do_warn(_("Only one AG detected - cannot proceed.\n"));
-			exit(1);
-		}
+		break;
 	default:
 		/*
 		 * at least half of the probed superblocks have

Index: xfsprogs/repair/xfs_repair.c
===================================================================

repair/xfs_repair.c +13 -0

@@ -409,6 +409,19 @@ calc_mkfs(xfs_mount_t *mp)
 	fino_bno = inobt_root + XFS_MIN_FREELIST_RAW(1, 1, mp) + 1;
 
 	/*
+	 * If we only have a single allocation group the log is also allocated
+	 * in the first allocation group and we need to add the number of
+	 * blocks used by the log to the above calculation.
+	 * All this of course doesn't apply if we have an external log.
+	 */
+	if (mp->m_sb.sb_agcount == 1 && mp->m_sb.sb_logstart) {
+		/*
+		 * XXX(hch): verify that sb_logstart makes sense?
+		 */
+		 fino_bno += mp->m_sb.sb_logblocks;
+	}
+
+	/*
 	 * ditto the location of the first inode chunks in the fs ('/')
 	 */
 	if (xfs_sb_version_hasdalign(&mp->m_sb) && do_inoalign)  {

_______________________________________________
xfs mailing list
[email protected]
http://oss.sgi.com/mailman/listinfo/xfs