From: Jonathan Brassow Subject: [PATCH] DM RAID: Clean-up handling of 'rebuild' device requests Date: Mon, 28 Nov 2011 14:01:05 -0600 Message-ID: <1322510465.20158.2.camel@f14.redhat.com> Reply-To: device-mapper development Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Return-path: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: dm-devel-bounces@redhat.com Errors-To: dm-devel-bounces@redhat.com To: dm-devel@redhat.com List-Id: dm-devel.ids dm-raid: Change the way the 'rebuild' parameter is processed. 1) We already allowed the rebuild flag to clear the superblock area for a device. There is no reason to shame the user if they didn't clear the superblock area ahead of time. Change the message to simply print-out that it is clearing the superblock for the purpose of a rebuild. 2) Set the 'FirstUse' flag in 'super_load' vs 'super_init_validation'. This allows the devices that had 'rebuild' set to be processed like a new device. 3) Since the rebuild devices are processed like new devices, the MD_CHANGE_DEVS flag also gets set on the MD device. This was a necessary missing action that forces the new superblock to be written when the array is instantiated. Signed-off-by: Jonathan Brassow Index: linux-upstream/drivers/md/dm-raid.c =================================================================== --- linux-upstream.orig/drivers/md/dm-raid.c +++ linux-upstream/drivers/md/dm-raid.c @@ -667,7 +667,14 @@ static int super_load(struct md_rdev *rd return ret; sb = page_address(rdev->sb_page); - if (sb->magic != cpu_to_le32(DM_RAID_MAGIC)) { + + /* + * Two cases that we want to write new superblocks and rebuild: + * 1) New device (no matching magic number) + * 2) Device specified for rebuild (!In_sync w/ offset == 0) + */ + if ((sb->magic != cpu_to_le32(DM_RAID_MAGIC)) || + (!test_bit(In_sync, &rdev->flags) && !rdev->recovery_offset)) { super_sync(rdev->mddev, rdev); set_bit(FirstUse, &rdev->flags); @@ -743,12 +750,9 @@ static int super_init_validation(struct * recovery_cp must be MaxSector. */ rdev_for_each(r, t, mddev) { - if (!test_bit(In_sync, &r->flags)) { - if (!test_bit(FirstUse, &r->flags)) - DMERR("Superblock area of " - "rebuild device %d should have been " - "cleared.", r->raid_disk); - set_bit(FirstUse, &r->flags); + if (!test_bit(In_sync, &r->flags)) + DMINFO("Device %d specified for rebuild: " + "Clearing superblock", r->raid_disk); rebuilds++; } else if (test_bit(FirstUse, &r->flags)) new_devs++;