From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from userp1040.oracle.com ([156.151.31.81]:32061 "EHLO userp1040.oracle.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1755451Ab3CMLrG (ORCPT ); Wed, 13 Mar 2013 07:47:06 -0400 Message-ID: <51406728.5030705@oracle.com> Date: Wed, 13 Mar 2013 19:46:48 +0800 From: Anand Jain MIME-Version: 1.0 To: dsterba@suse.cz, Eric Sandeen , linux-btrfs@vger.kernel.org Subject: Re: [PATCH 3/3] btrfs-progs: use BTRFS_SCAN_BACKUP_SB flag in btrfs_scan_one_device References: <1362756300-30212-1-git-send-email-anand.jain@oracle.com> <1362756300-30212-4-git-send-email-anand.jain@oracle.com> <513DF252.8070109@redhat.com> <20130311181635.GH7229@twin.jikos.cz> In-Reply-To: <20130311181635.GH7229@twin.jikos.cz> Content-Type: text/plain; charset=ISO-8859-1; format=flowed Sender: linux-btrfs-owner@vger.kernel.org List-ID: Thanks Eric and David. I have sent out V3 patch-set which will disable access to backup super-block unless requested by the user. here below are some test cases and results before and after this fix.. which finds the patch works awesome. -------- the original problem which all this started.. mkfs.btrfs /dev/sdb -f && yes| mkfs.ext4 /dev/sdb && mount /dev/sdb /ext4 ~/before/mkfs.btrfs -f /dev/sdc /dev/sdd (run twice) ~/before/mkfs.btrfs -f /dev/sdc /dev/sdd :: ERROR: unable to scan the device '/dev/sdb' - Device or resource busy ERROR: unable to scan the device '/dev/sdb' - Device or resource busy adding device /dev/sdd id 2 fs created label (null) on /dev/sdc nodesize 4096 leafsize 4096 sectorsize 4096 size 3.11GB after the fix. # mkfs.btrfs -f /dev/sdc /dev/sdd WARNING! - Btrfs v0.20-rc1-238-g170881a-dirty IS EXPERIMENTAL WARNING! - see http://btrfs.wiki.kernel.org before using adding device /dev/sdd id 2 fs created label (null) on /dev/sdc nodesize 4096 leafsize 4096 sectorsize 4096 size 104.98GB Btrfs v0.20-rc1-238-g170881a-dirty ------------- # umount /ext4 # mkfs.btrfs /dev/sdb /dev/sdc -f && wipefs /dev/sdb There is an existing bug.. # ~/before/btrfs check /dev/sdc --repair enabling repair mode warning, device 1 is missing warning, device 1 is missing warning devid 1 not found already Checking filesystem on /dev/sdc UUID: e03c477d-b696-46c1-bf79-43b8b8454f11 checking extents checking fs roots checking root refs btrfs: extent-tree.c:2553: btrfs_reserve_extent: Assertion `!(ret)' failed. Aborted (core dumped) # and its still there even after this patch :-) # btrfs check /dev/sdc --repair enabling repair mode warning, device 1 is missing warning, device 1 is missing warning devid 1 not found already Checking filesystem on /dev/sdc UUID: e03c477d-b696-46c1-bf79-43b8b8454f11 checking extents checking fs roots checking root refs btrfs: extent-tree.c:2553: btrfs_reserve_extent: Assertion `!(ret)' failed. Aborted (core dumped) ------------------------- # mkfs.btrfs /dev/sdb /dev/sdc -f && yes|mkfs.ext4 /dev/sdb without the fix # ~/before/btrfs restore -o /dev/sdb ~/tmp ERROR: device scan failed '/dev/sdb' - Invalid argument ERROR: device scan failed '/dev/sdb' - Invalid argument Root objectid is 5 with the fix # btrfs restore -o /dev/sdb ~/tmp No valid Btrfs found on /dev/sdb Could not open root, trying backup super Root objectid is 5 now I say to use sb backup 1.. # btrfs restore -u 1 -o /dev/sdb ~/tmp Root objectid is 5 # echo $? 0 -------------- # mkfs.btrfs /dev/sdb /dev/sdc -f && yes|mkfs.ext4 /dev/sdb # ~/before/btrfs check /dev/sdc --repair enabling repair mode ERROR: device scan failed '/dev/sdb' - Invalid argument ERROR: device scan failed '/dev/sdb' - Invalid argument Checking filesystem on /dev/sdc UUID: 5dc5c663-eeea-4ec2-b8fe-170157f5d68d checking extents checking fs roots checking root refs found 28672 bytes used err is 0 total csum bytes: 0 total tree bytes: 28672 total fs tree bytes: 8192 btree space waste bytes: 22420 file data blocks allocated: 0 referenced 0 Btrfs v0.20-rc1-235-gdd21bc1 # mkfs.btrfs /dev/sdb /dev/sdc -f && yes|mkfs.ext4 /dev/sdb since now new code doesn't read backup SB of devid1 it hits the earlier known bug. # btrfs check /dev/sdc --repair enabling repair mode warning, device 1 is missing warning, device 1 is missing warning devid 1 not found already Checking filesystem on /dev/sdc UUID: daa0dc6d-35b6-43ab-8333-894710a5b917 checking extents checking fs roots checking root refs btrfs: extent-tree.c:2553: btrfs_reserve_extent: Assertion `!(ret)' failed. Aborted (core dumped) So user would need to fix the devid1 first. # btrfs check /dev/sdb --repair enabling repair mode No valid Btrfs found on /dev/sdb # btrfs check -s 1 /dev/sdb --repair using SB copy 1, bytenr 67108864 enabling repair mode Checking filesystem on /dev/sdb UUID: daa0dc6d-35b6-43ab-8333-894710a5b917 checking extents checking fs roots checking root refs found 28672 bytes used err is 0 total csum bytes: 0 total tree bytes: 28672 total fs tree bytes: 8192 btree space waste bytes: 22420 file data blocks allocated: 0 referenced 0 Btrfs v0.20-rc1-238-g170881a-dirty # btrfs check /dev/sdc --repair enabling repair mode warning, device 1 is missing warning, device 1 is missing warning devid 1 not found already Checking filesystem on /dev/sdc UUID: daa0dc6d-35b6-43ab-8333-894710a5b917 checking extents checking fs roots checking root refs btrfs: extent-tree.c:2553: btrfs_reserve_extent: Assertion `!(ret)' failed. Aborted (core dumped) Need to write back the backup SB to primary SB. IMO when -s 1 was given it had the chance to fix primary.. this can be fixed later. # ./btrfs-select-super -s 1 /dev/sdb using SB copy 1, bytenr 67108864 ERROR: device scan failed '/dev/sdb' - Invalid argument # btrfs check /dev/sdc --repair enabling repair mode Checking filesystem on /dev/sdc UUID: daa0dc6d-35b6-43ab-8333-894710a5b917 checking extents checking fs roots checking root refs found 28672 bytes used err is 0 total csum bytes: 0 total tree bytes: 28672 total fs tree bytes: 8192 btree space waste bytes: 22420 file data blocks allocated: 0 referenced 0 Btrfs v0.20-rc1-238-g170881a-dirty --------------------------------------------------------------- I am not sure if I have a wipe -a working for btrfs, so I would clean using dd. # dd if=/dev/zero bs=1 count=8 of=/dev/sdb seek=$((64*1024+64)) 8+0 records in 8+0 records out 8 bytes (8 B) copied, 0.000324503 s, 24.7 kB/s # dd if=/dev/zero bs=1 count=8 of=/dev/sdb seek=$((64*1024*1024+64)) 8+0 records in 8+0 records out 8 bytes (8 B) copied, 0.000364635 s, 21.9 kB/s # dd if=/dev/zero bs=1 count=8 of=/dev/sdb seek=$((256*1024*1024*1024+64)) dd: `/dev/sdb': cannot seek: Invalid argument 0+0 records in 0+0 records out 0 bytes (0 B) copied, 0.000430846 s, 0.0 kB/s This behavior is same before and after the patch. # btrfs check /dev/sdb No valid Btrfs found on /dev/sdb # btrfs check /dev/sdb -s 1 using SB copy 1, bytenr 67108864 No valid Btrfs found on /dev/sdb # btrfs check /dev/sdb -s 2 using SB copy 2, bytenr 274877906944 No valid Btrfs found on /dev/sdb # ------------------------ # mkfs.btrfs /dev/sdb /dev/sdc -f && mount /dev/sdc /btrfs # dd if=/dev/zero bs=1 count=8 of=/dev/sdb seek=$((64*1024+64)) # dd if=/dev/zero bs=1 count=8 of=/dev/sdb seek=$((64*1024*1024+64)) # dd if=/dev/zero bs=1 count=8 of=/dev/sdb seek=$((256*1024*1024*1024+64)) # mkfs.btrfs /dev/sdb WARNING! - Btrfs v0.20-rc1-238-g170881a-dirty IS EXPERIMENTAL WARNING! - see http://btrfs.wiki.kernel.org before using /dev/sdb is mounted # umount /btrfs # mkfs.btrfs /dev/sdb WARNING! - Btrfs v0.20-rc1-238-g170881a-dirty IS EXPERIMENTAL WARNING! - see http://btrfs.wiki.kernel.org before using /dev/sdb appears to contain an existing filesystem (btrfs). Use the -f option to force overwrite. # ------------------------- # mkfs.btrfs /dev/sdd /dev/sde -f WARNING! - Btrfs v0.20-rc1-238-g170881a-dirty IS EXPERIMENTAL WARNING! - see http://btrfs.wiki.kernel.org before using adding device /dev/sde id 2 fs created label (null) on /dev/sdd nodesize 4096 leafsize 4096 sectorsize 4096 size 93.22GB Btrfs v0.20-rc1-238-g170881a-dirty # mount /dev/sde /btrfs # umount /btrfs # dd if=/dev/zero bs=1 count=8 of=/dev/sde seek=$((64*1024+64)) 8+0 records in 8+0 records out 8 bytes (8 B) copied, 0.0149415 s, 0.5 kB/s remove /dev/sdd # x="0 0 3 0" # echo "scsi remove-single-device $x" > /proc/scsi/scsi # lsscsi [0:0:0:0] disk Sun boot1 V1.0 /dev/sda [0:0:2:0] disk Sun boot3 V1.0 /dev/sdc [0:0:4:0] disk Sun boot5 V1.0 /dev/sde [0:1:0:0] disk SEAGATE ST914602SSUN146G 0603 - [0:1:1:0] disk SEAGATE ST914602SSUN146G 0603 - [0:3:0:0] enclosu ADAPTEC Virtual SGPIO 0 0001 - # btrfs check /dev/sde No valid Btrfs found on /dev/sde # btrfs check /dev/sde -s 1 using SB copy 1, bytenr 67108864 warning, device 1 is missing warning, device 1 is missing warning devid 1 not found already Checking filesystem on /dev/sde UUID: fa08122d-832f-4f32-87ee-58eb9380d5da checking extents checking fs roots checking root refs found 815104 bytes used err is 0 total csum bytes: 0 total tree bytes: 28672 total fs tree bytes: 8192 btree space waste bytes: 21476 file data blocks allocated: 786432 referenced 786432 Btrfs v0.20-rc1-238-g170881a-dirty # ./btrfs-select-super -s 1 /dev/sde using SB copy 1, bytenr 67108864 ERROR: device scan failed '/dev/sdc' - Invalid argument warning, device 1 is missing warning, device 1 is missing warning devid 1 not found already # btrfs check /dev/sde warning, device 1 is missing warning, device 1 is missing warning devid 1 not found already Checking filesystem on /dev/sde UUID: fa08122d-832f-4f32-87ee-58eb9380d5da checking extents checking fs roots checking root refs found 815104 bytes used err is 0 total csum bytes: 0 total tree bytes: 28672 total fs tree bytes: 8192 btree space waste bytes: 21476 file data blocks allocated: 786432 referenced 786432 Btrfs v0.20-rc1-238-g170881a-dirty # --------------- Thanks, Anand On 03/12/2013 02:16 AM, David Sterba wrote: > On Mon, Mar 11, 2013 at 10:03:46AM -0500, Eric Sandeen wrote: >> IMHO, nothing should be checking the backup superblocks unless explicitly >> told to. > > That's the whole point I believe. > > update the infrastructure, every SB access looks to the first copy > unless told by command line options. > > > david > -- > To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in > the body of a message to majordomo@vger.kernel.org > More majordomo info at http://vger.kernel.org/majordomo-info.html >