Bug#656899: Same problem too under SPARC64

[Denny Schierz]

hi,

uname -a

2.6.32 #3 SMP Wed Feb 1 14:41:07 CET 2012 sparc64 GNU/Linux

with self compiled kernel, but with Debian sources:

  linux-source-2.6.32   2.6.32-41 

-
[   95.380528] mdadm: sending ioctl 20001261 to a partition!
[   95.451590] mdadm: sending ioctl 20001261 to a partition!
[   95.527711] mdadm: sending ioctl 20001261 to a partition!
[   95.598789] mdadm: sending ioctl 20001261 to a partition!
[   95.691778] mdadm: sending ioctl 20001261 to a partition!
[   95.772963] mdadm: sending ioctl 20001261 to a partition!
--

# cat /proc/mdstat 
Personalities : [raid1] [raid10] [raid6] [raid5] [raid4] [multipath] 
md0 : active raid1 sda2[0] sdb2[1]
  71190951 blocks super 1.2 [2/2] [UU]

---

Disk /dev/sda (Sun disk label): 255 heads, 63 sectors, 8924 cylinders
Units = cylinders of 16065 * 512 bytes

   Device FlagStart   EndBlocks   Id  System
/dev/sda1 061489982+   1  Boot
/dev/sda261  8924  71192047+  fd  Linux raid autodetect
/dev/sda3 0  8924  716820305  Whole disk

Disk /dev/sdb (Sun disk label): 255 heads, 63 sectors, 8924 cylinders
Units = cylinders of 16065 * 512 bytes

   Device FlagStart   EndBlocks   Id  System
/dev/sdb1 061489982+  83  Linux native
/dev/sdb261  8924  71192047+  fd  Linux raid autodetect
/dev/sdb3 0  8924  716820305  Whole disk

---

syslog:

Feb  2 06:26:08 mpt-statusd: detected non-optimal RAID status
Feb  2 06:36:08  mpt-statusd: detected non-optimal RAID status
Feb  2 06:46:08  mpt-statusd: detected non-optimal RAID status


and the kernel remounted my / readonly ... but found nothing about that in 
the log.

cu denny



signature.asc
Description: Message signed with OpenPGP using GPGMail

Bug#656899: Same problem too under SPARC64

[Ben Hutchings]

On Thu, 2012-02-02 at 16:00 +0100, Denny Schierz wrote:
 hi,
 
 uname -a
 
 2.6.32 #3 SMP Wed Feb 1 14:41:07 CET 2012 sparc64 GNU/Linux
 
 with self compiled kernel, but with Debian sources:
 
   linux-source-2.6.32   2.6.32-41 
 
 -
 [   95.380528] mdadm: sending ioctl 20001261 to a partition!
 [   95.451590] mdadm: sending ioctl 20001261 to a partition!
 [   95.527711] mdadm: sending ioctl 20001261 to a partition!
 [   95.598789] mdadm: sending ioctl 20001261 to a partition!
 [   95.691778] mdadm: sending ioctl 20001261 to a partition!
 [   95.772963] mdadm: sending ioctl 20001261 to a partition!
 --
[...]

Does the attached patch fix this?  This change should be in the longterm
update 2.6.32.56 and we can put it into a stable update pretty soon if
necessary.

Ben.

-- 
Ben Hutchings
Lowery's Law:
 If it jams, force it. If it breaks, it needed replacing anyway.
diff -u b/drivers/scsi/sd.c b/drivers/scsi/sd.c
--- b/drivers/scsi/sd.c
+++ b/drivers/scsi/sd.c
@@ -1004,7 +1004,7 @@
 
 	ret = scsi_verify_blk_ioctl(bdev, cmd);
 	if (ret  0)
-		return ret;
+		return -ENOIOCTLCMD;
 
 	/*
 	 * If we are in the middle of error recovery, don't let anyone


signature.asc
Description: This is a digitally signed message part

Bug#656899: Same problem too under SPARC64

[Ben Hutchings]

On Fri, 2012-02-03 at 03:45 +, Ben Hutchings wrote:
 On Thu, 2012-02-02 at 16:00 +0100, Denny Schierz wrote:
  hi,
  
  uname -a
  
  2.6.32 #3 SMP Wed Feb 1 14:41:07 CET 2012 sparc64 GNU/Linux
  
  with self compiled kernel, but with Debian sources:
  
linux-source-2.6.32   2.6.32-41 
  
  -
  [   95.380528] mdadm: sending ioctl 20001261 to a partition!
  [   95.451590] mdadm: sending ioctl 20001261 to a partition!
  [   95.527711] mdadm: sending ioctl 20001261 to a partition!
  [   95.598789] mdadm: sending ioctl 20001261 to a partition!
  [   95.691778] mdadm: sending ioctl 20001261 to a partition!
  [   95.772963] mdadm: sending ioctl 20001261 to a partition!
  --
 [...]
 
 Does the attached patch fix this?  This change should be in the longterm
 update 2.6.32.56 and we can put it into a stable update pretty soon if
 necessary.

Sorry, this patch probably isn't going to make any difference, as only
processes *not* running as root are affected by the ioctl filtering.
And the patch certainly won't remove the warning message.

Denny, I suspect the RAID failure is due to an error condition that's
detected at boot time and it would have showed up whenever you rebooted,
regardless of any upgrade.  You can probably check this by running an
older version of the Debian installer in rescue mode.

Ben.

-- 
Ben Hutchings
Lowery's Law:
 If it jams, force it. If it breaks, it needed replacing anyway.


signature.asc
Description: This is a digitally signed message part

Bug#656899: Same problem too under SPARC64

[Denny Schierz]

hi Ben,

Am 03.02.2012 um 04:45 schrieb Ben Hutchings:
 
 Does the attached patch fix this?  This change should be in the longterm
 update 2.6.32.56 and we can put it into a stable update pretty soon if
 necessary.

as you said, the patch changes nothing:

== Booting =
[...]

[   93.392382] scsi 1:0:1:0: Direct-Access FUJITSU  MAY2073RCSUN72G  0601 
PQ: 0 ANSI: 4
[   93.522739] sd 1:0:1:0: [sdb] 143374738 512-byte logical blocks: (73.4 
GB/68.3 GiB)
[   93.542647] sd 1:0:0:0: [sda] 143374738 512-byte logical blocks: (73.4 
GB/68.3 GiB)
[   93.544892] sd 1:0:0:0: [sda] Write Protect is off
[   93.546724] sd 1:0:0:0: [sda] Write cache: enabled, read cache: enabled, 
supports DPO and FUA
[   93.571398]  sda: sda1 sda2 sda3
[   93.944231] sd 1:0:1:0: [sdb] Write Protect is off
[   93.972728] sd 1:0:0:0: [sda] Attached SCSI disk
[   94.069324] sd 1:0:1:0: [sdb] Write cache: enabled, read cache: enabled, 
doesn't support DPO or FUA
[   94.192651]  sdb: sdb1 sdb2 sdb3
[   94.252601] sd 1:0:1:0: [sdb] Attached SCSI disk
Begin: Loading essential drivers[   94.490764] mdadm: sending ioctl 20001261 to 
a partition!
 ... done.
Begi[   94.572794] mdadm: sending ioctl 20001261 to a partition!
n: Running /scri[   94.666140] md: md0 stopped.
pts/init-premoun[   94.715638] mdadm: sending ioctl 20001261 to a partition!
t ... done.
Beg[   94.802768] mdadm: sending ioctl 20001261 to a partition!
in: Mounting roo[   94.891950] mdadm: sending ioctl 20001261 to a partition!
t file system ..[   94.978148] mdadm: sending ioctl 20001261 to a partition!
. Begin: Running[   95.066478] mdadm: sending ioctl 20001261 to a partition!
 /scripts/local-[   95.153554] mdadm: sending ioctl 20001261 to a partition!
top ... Begin: L[   95.243853] md: bindsdb2
oadingFailure: f[   95.293958] md: bindsda2
ailed to load mo[   95.348444] raid1: raid set md0 active with 2 out of 2 
mirrors
dule raid1.
don[   95.439125] md0: detected capacity change from 0 to 72899533824
e.
Begin: Assem[   95.535673]  md0:bling all MD arr unknown partition table
ays ... mdadm: d[   95.623786] mdadm: sending ioctl 20001261 to a partition!
evice 1 in /dev/[   95.707595] mdadm: sending ioctl 20001261 to a partition!
md/0 has wrong state in superblock, but /dev/sdb2 seems ok
mdadm: /dev/md/0 has been started with 2 drives.
Success: assembled all arrays.
done.
done.
Begin: Running /scripts/local-premount ... done.
[   96.329438] EXT4-fs (dm-0): mounted filesystem with ordered data mode
Begin: Running /scripts/local-bottom ... done.
done.
Begin: Running /scripts/init-bottom ... done.
INIT: version 2.88 booting
Using makefile-style concurrent boot in runlevel S.
Starting the hotplug events dispatcher: udevd[   97.163217] udev[272]: starting 
version 164
.
Synthesizing the initial hotplug events...done.
Waiting for /dev to be fully populated...[   97.500481] mdadm: sending ioctl 
20001261 to a partition!
[   97.571656] mdadm: sending ioctl 20001261 to a partition!
done.
Checking root file system...fsck from util-linux-ng 2.17.2
[   98.109972] mdadm: sending ioctl 20001261 to a partition!8747/1220608 blocks

done.
Cleaning up ifupdown[   98.208631] mdadm: sending ioctl 20001261 to a 
partition!

[   98.292584] mdadm: sending ioctl 20001261 to a partition!
Setting up netwo[   98.371654] mdadm: sending ioctl 20001261 to a partition!
rking
[   98.453822] mdadm: sending ioctl 20001261 to a partition!
[   98.474271] loop: module loaded
Loading kernel m[   98.587144] mdadm: sending ioctl 20001261 to a partition!
odules...done.
[   98.666444] mdadm: sending ioctl 20001261 to a partition!
[   98.754847] mdadm: sending ioctl 20001261 to a partition!
[   98.826007] mdadm: sending ioctl 20001261 to a partition!
Generating udev [   98.913101] mdadm: sending ioctl 20001261 to a partition!
events for MD ar[   98.993103] mdadm: sending ioctl 20001261 to a partition!
rays...done.
Setting up LVM Volume Groups  Reading all physical volumes.  This may take a 
while...
  Found volume group root-vg using metadata type lvm2
  5 logical volume(s) in volume group root-vg now active
.
A[   99.721865] Adding 3903472k swap on /dev/mapper/root--vg-swap_fs.  
Priority:-1 extents:1 across:3903472k 
ctivating lvm and md swap...done.
Checking file systems...fsck from util-linux-ng 2.17.2
/dev/sda1: clean, 35/245280 files, 55661/489982 blocks
/dev/mapper/root--vg-tmp_fs: clean, 13/60928 files, 8233/243712 blocks
/dev/mapper/root--vg-var_fs: clean, 2182/244320 files, 101989/975872 [  
100.153260] EXT4-fs (dm-3): blocks
done.
Mmounted filesystem with ordered data modeounting local fi
lesystems...[  100.289830] EXT4-fs (dm-2): mounted filesystem with ordered data 
mode
done.
Activating swapfile swap...done.
Cleaning up temporary files
Setting kernel variables ...done.
[...]

===


but one thing I've found: the remount to Read-Only from / didn't come from 
the mdadm, it comes from LXC if you shutdown a container ,