Oops when starting md multipath on a 2.4 kernel
We have an existing system runing a 2.4.27 based kernel that uses md multipath and external fibre channel arrays.
We need to add more internal disks to the system, which means the external drives change device names.
When I tried to start the md multipath device using mdadm, the kernel Oops'd. Removing the new internal disks and going back the original setup, I can start the multipath device - as this machine is in production, I can't do any more tests.
However, I can reproduce the problem on test system by creating an md multipath device on an external SCSI disk, using /dev/sda1, stopping the multipath device, rmmod'ing the SCSI driver, pluging in a couple of USB storage devices which become /dev/sda and /dev/sdb and then modprobing the SCSI driver, so the original /dev/sda1 is now /dev/sdc1.
When I run 'mdadm -A -s', I get the following Oops:
[events: 00000004]
md: bind<sdc1,1>
md: sdc1's event counter: 00000004
md0: former device sda1 is unavailable, removing from array!
md: unbind<sdc1,0>
md: export_rdev(sdc1)
md: RAID level -4 does not need chunksize! Continuing anyway.
md: multipath personality registered as nr 7
md0: max total readahead window set to 124k
md0: 1 data-disks, max readahead per data-disk: 124k
Unable to handle kernel NULL pointer dereference at virtual address 00000040
printing eip:
e096527e
*pde = 00000000
Oops: 0000
CPU: 0
EIP: 0010:[<e096527e>] Not tainted
EFLAGS: 00010246
eax: deb62a94 ebx: 00000000 ecx: dd65b400 edx: 00000000
esi: 0000001c edi: deb62a94 ebp: 00000000 esp: dd5fbdbc
ds: 0018 es: 0018 ss: 0018
Process mdadm (pid: 1389, stackpage=dd5fb000)
Stack: dd4c4000 dfa96000 c035ad00 00000000 00000286 dd4c4000 00000000 00000000
deb62a94 dd5fbe5c dd4c6000 c02a6e10 dd65b400 c035ef1f 0000007c
00000000
0000000a ffffffff 00000002 00002e2e c0118b49 00002e2e 00002e2e
00000286
Call Trace: [<c02a6e10>] [<c0118b49>] [<c0118cc4>] [<c024a88c>] [<c024abb6>]
[<c0118cc4>] [<c024907e>] [<c024b6f2>] [<c024c60c>] [<c014a326>] [<c013c483>]
[<c013ca18>] [<c01375ac>] [<c013ca63>] [<c01439b6>] [<c01087c7>]
Code: 8b 45 40 85 c0 0f 84 c2 01 00 00 6a 00 ff b4 24 cc 00 00 00
Running through ksymoops gives:
Unable to handle kernel NULL pointer dereference at virtual address 00000040 e096527e *pde = 00000000 Oops: 0000 CPU: 0 EIP: 0010:[<e096527e>] Not tainted
Using defaults from ksymoops -t elf32-i386 -a i386
EFLAGS: 00010246
eax: deb62a94 ebx: 00000000 ecx: dd65b400 edx: 00000000
esi: 0000001c edi: deb62a94 ebp: 00000000 esp: dd5fbdbc
ds: 0018 es: 0018 ss: 0018
Process mdadm (pid: 1389, stackpage=dd5fb000)
Stack: dd4c4000 dfa96000 c035ad00 00000000 00000286 dd4c4000 00000000 00000000
deb62a94 dd5fbe5c dd4c6000 c02a6e10 dd65b400 c035ef1f 0000007c
00000000
0000000a ffffffff 00000002 00002e2e c0118b49 00002e2e 00002e2e
00000286
Call Trace: [<c02a6e10>] [<c0118b49>] [<c0118cc4>] [<c024a88c>] [<c024abb6>]
[<c0118cc4>] [<c024907e>] [<c024b6f2>] [<c024c60c>] [<c014a326>]
[<c013c483>]
[<c013ca18>] [<c01375ac>] [<c013ca63>] [<c01439b6>] [<c01087c7>]
Code: 8b 45 40 85 c0 0f 84 c2 01 00 00 6a 00 ff b4 24 cc 00 00 00Quoted 1 line
Quoted 1 line
EIP;
e096527e<[multipath]multipath_run+2be/6c0> <=====
Trace; c02a6e10 <vsnprintf+2e0/450>
Trace; c0118b49 <call_console_drivers+e9/f0> Trace; c0118cc4 <printk+104/110>
Trace; c024a88c <device_size_calculation+19c/1f0> Trace; c024abb6 <do_md_run+2d6/360>
Trace; c0118cc4 <printk+104/110>
Trace; c024907e <bind_rdev_to_array+9e/b0> Trace; c024b6f2 <add_new_disk+132/290> Trace; c024c60c <md_ioctl+6fc/790>
Trace; c014a326 <iput+236/240>
Trace; c013c483 <bdput+93/a0>
Trace; c013ca18 <blkdev_put+98/a0>
Trace; c01375ac <fput+bc/e0>
Trace; c013ca63 <blkdev_ioctl+23/30> Trace; c01439b6 <sys_ioctl+216/230>
Trace; c01087c7 <system_call+33/38>
Code; e096527e <[multipath]multipath_run+2be/6c0> 00000000 <_EIP>:
Code; e096527e <[multipath]multipath_run+2be/6c0> <=====
0: 8b 45 40 mov 0x40(%ebp),%eax <=====
Code; e0965281 <[multipath]multipath_run+2c1/6c0>
3: 85 c0 test %eax,%eax
Code; e0965283 <[multipath]multipath_run+2c3/6c0>
5: 0f 84 c2 01 00 00 je 1cd <_EIP+0x1cd> e096544b
<[multipath]m
ultipath_run+48b/6c0>
Code; e0965289 <[multipath]multipath_run+2c9/6c0>
b: 6a 00 push $0x0
Code; e096528b <[multipath]multipath_run+2cb/6c0>
d: ff b4 24 cc 00 00 00 pushl 0xcc(%esp,1)
My /etc/mdadm.conf contains:
DEVICE /dev/sd?1
ARRAY /dev/md0 level=multipath num-devices=1
UUID=277e4ba5:6c23c087:e17c877c:da642955
Should md multipath be able to handle changes like this with the underlying devices?
Thanks
James Pearson
Quoted 108 lines · James Pearson
We have an existing system runing a 2.4.27 based kernel that uses md multipath and external fibre channel arrays.
We need to add more internal disks to the system, which means the external drives change device names.
When I tried to start the md multipath device using mdadm, the kernel Oops'd. Removing the new internal disks and going back the original setup, I can start the multipath device - as this machine is in production, I can't do any more tests.
However, I can reproduce the problem on test system by creating an md multipath device on an external SCSI disk, using
/dev/sda1, stopping the multipath device, rmmod'ing the SCSI driver, pluging in a couple of USB storage devices which become/dev/sdaand/dev/sdband then modprobing the SCSI driver, so the original/dev/sda1is now/dev/sdc1.When I run 'mdadm -A -s', I get the following Oops:
[events: 00000004]
md: bind<sdc1,1>
md: sdc1's event counter: 00000004
md0: former device sda1 is unavailable, removing from array!
md: unbind<sdc1,0>
md:export_rdev(sdc1)
md: RAID level -4 does not need chunksize! Continuing anyway.
md: multipath personality registered as nr 7
md0: max total readahead window set to 124k
md0: 1 data-disks, max readahead per data-disk: 124kUnable to handle kernel NULL pointer dereference at virtual address 00000040 printing eip: e096527e *pde = 00000000 Oops: 0000 CPU: 0 EIP: 0010:[<e096527e>] Not tainted EFLAGS: 00010246 eax: deb62a94 ebx: 00000000 ecx: dd65b400 edx: 00000000 esi: 0000001c edi: deb62a94 ebp: 00000000 esp: dd5fbdbc ds: 0018 es: 0018 ss: 0018 Process mdadm (pid: 1389, stackpage=dd5fb000) Stack: dd4c4000 dfa96000 c035ad00 00000000 00000286 dd4c4000 00000000 00000000 deb62a94 dd5fbe5c dd4c6000 c02a6e10 dd65b400 c035ef1f 0000007c 00000000 0000000a ffffffff 00000002 00002e2e c0118b49 00002e2e 00002e2e 00000286 Call Trace: [<c02a6e10>] [<c0118b49>] [<c0118cc4>] [<c024a88c>] [<c024abb6>] [<c0118cc4>] [<c024907e>] [<c024b6f2>] [<c024c60c>] [<c014a326>][<
c013c483>]
[<c013ca18>] [<c01375ac>] [<c013ca63>] [<c01439b6>] [<c01087c7>]Code: 8b 45 40 85 c0 0f 84 c2 01 00 00 6a 00 ff b4 24 cc 00 00 00Running through ksymoops gives:
Unable to handle kernel NULL pointer dereference at virtual address 00000040 e096527e *pde = 00000000 Oops: 0000 CPU: 0 EIP: 0010:[<e096527e>] Not taintedUsing defaults from ksymoops -t elf32-i386 -a i386
EFLAGS: 00010246
eax:deb62a94ebx: 00000000 ecx:dd65b400edx: 00000000
esi:0000001cedi:deb62a94ebp: 00000000 esp:dd5fbdbc
ds: 0018 es: 0018 ss: 0018
Process mdadm (pid: 1389, stackpage=dd5fb000)Stack: dd4c4000 dfa96000 c035ad00 00000000 00000286 dd4c4000 00000000 00000000 deb62a94 dd5fbe5c dd4c6000 c02a6e10 dd65b400 c035ef1f 0000007c 00000000 0000000a ffffffff 00000002 00002e2e c0118b49 00002e2e 00002e2e 00000286 Call Trace: [<c02a6e10>] [<c0118b49>] [<c0118cc4>] [<c024a88c>] [<c024abb6>] [<c0118cc4>] [<c024907e>] [<c024b6f2>] [<c024c60c>] [<c014a326>] [<c013c483>] [<c013ca18>] [<c01375ac>] [<c013ca63>] [<c01439b6>] [<c01087c7>] Code: 8b 45 40 85 c0 0f 84 c2 01 00 00 6a 00 ff b4 24 cc 00 00 00Quoted 1 line
Quoted 1 line
EIP;
e096527e<[multipath]multipath_run+2be/6c0> <=====Trace;
c02a6e10<vsnprintf+2e0/450>
Trace;c0118b49<call_console_drivers+e9/f0> Trace;c0118cc4<printk+104/110>
Trace;c024a88c<device_size_calculation+19c/1f0> Trace;c024abb6<do_md_run+2d6/360>
Trace;c0118cc4<printk+104/110>
Trace;c024907e<bind_rdev_to_array+9e/b0> Trace;c024b6f2<add_new_disk+132/290> Trace;c024c60c<md_ioctl+6fc/790>
Trace;c014a326<iput+236/240>
Trace;c013c483<bdput+93/a0>
Trace;c013ca18<blkdev_put+98/a0>
Trace;c01375ac<fput+bc/e0>
Trace;c013ca63<blkdev_ioctl+23/30> Trace;c01439b6<sys_ioctl+216/230>
Trace;c01087c7<system_call+33/38>
Code;e096527e<[multipath]multipath_run+2be/6c0> 00000000 <_EIP>:
Code;e096527e<[multipath]multipath_run+2be/6c0> <=====
0: 8b 45 40 mov 0x40(%ebp),%eax <=====
Code;e0965281<[multipath]multipath_run+2c1/6c0>
3: 85 c0 test %eax,%eax
Code;e0965283<[multipath]multipath_run+2c3/6c0>
5: 0f 84 c2 01 00 00 je 1cd <_EIP+0x1cd>e096544b
<[multipath]m
ultipath_run+48b/6c0>
Code;e0965289<[multipath]multipath_run+2c9/6c0>
b: 6a 00 push $0x0
Code;e096528b<[multipath]multipath_run+2cb/6c0>
d: ff b4 24 cc 00 00 00 pushl 0xcc(%esp,1)My
/etc/mdadm.confcontains:DEVICE /dev/sd?1
ARRAY/dev/md0level=multipath num-devices=1
UUID=277e4ba5:6c23c087:e17c877c:da642955
Quoted 7 lines
Should md multipath be able to handle changes like this with the underlying devices?
Thanks
James Pearson
Hi James,
My co-worker and I just happened to run into this problem a few days ago. So, I would like to share with you what we know.
The device major/minor numbers no longer match up values recorded in the descriptor array in the md superblock. Because of the exception made in the current code, the descriptor entries are removed and although the real devices are present and accounted for, they are kicked out from the array. This leaves the array with zero devices. When multipath_run() is invoked, it blows up expecting to have had some disks.
Lars Marowsky-Brée suggested some patches for md multipath in 2002 but never made it to mainline 2.4 kernel:
http://marc.theaimsgroup.com/?l=linux-kernel&m=103355467608953&w=2
That patch is large and most of it is not requried for this particular problem. The section that reinitializes the descriptor array from current rdevs for the case of multipath will resolve this issue of device names shift.
Lars, Is it ok with you if I compose a patch from your original patch and post it here?
--
Regards,
Mike T.
-
To unsubscribe from this list: send the line "unsubscribe linux-raid" in the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html
Quoted 121 lines · Mike Tran
Quoted 108 lines · James Pearson
We have an existing system runing a 2.4.27 based kernel that uses md multipath and external fibre channel arrays.
We need to add more internal disks to the system, which means the external drives change device names.
When I tried to start the md multipath device using mdadm, the kernel Oops'd. Removing the new internal disks and going back the original setup, I can start the multipath device - as this machine is in production, I can't do any more tests.
However, I can reproduce the problem on test system by creating an md multipath device on an external SCSI disk, using
/dev/sda1, stopping the multipath device, rmmod'ing the SCSI driver, pluging in a couple of USB storage devices which become/dev/sdaand/dev/sdband then modprobing the SCSI driver, so the original/dev/sda1is now/dev/sdc1.When I run 'mdadm -A -s', I get the following Oops:
[events: 00000004]
md: bind<sdc1,1>
md: sdc1's event counter: 00000004
md0: former device sda1 is unavailable, removing from array!
md: unbind<sdc1,0>
md:export_rdev(sdc1)
md: RAID level -4 does not need chunksize! Continuing anyway.
md: multipath personality registered as nr 7
md0: max total readahead window set to 124k
md0: 1 data-disks, max readahead per data-disk: 124kUnable to handle kernel NULL pointer dereference at virtual address 00000040 printing eip: e096527e *pde = 00000000 Oops: 0000 CPU: 0 EIP: 0010:[<e096527e>] Not tainted EFLAGS: 00010246 eax: deb62a94 ebx: 00000000 ecx: dd65b400 edx: 00000000 esi: 0000001c edi: deb62a94 ebp: 00000000 esp: dd5fbdbc ds: 0018 es: 0018 ss: 0018 Process mdadm (pid: 1389, stackpage=dd5fb000) Stack: dd4c4000 dfa96000 c035ad00 00000000 00000286 dd4c4000 00000000 00000000 deb62a94 dd5fbe5c dd4c6000 c02a6e10 dd65b400 c035ef1f 0000007c 00000000 0000000a ffffffff 00000002 00002e2e c0118b49 00002e2e 00002e2e 00000286 Call Trace: [<c02a6e10>] [<c0118b49>] [<c0118cc4>] [<c024a88c>] [<c024abb6>] [<c0118cc4>] [<c024907e>] [<c024b6f2>] [<c024c60c>] [<c014a326>][<
c013c483>]
[<c013ca18>] [<c01375ac>] [<c013ca63>] [<c01439b6>] [<c01087c7>]Code: 8b 45 40 85 c0 0f 84 c2 01 00 00 6a 00 ff b4 24 cc 00 00 00Running through ksymoops gives:
Unable to handle kernel NULL pointer dereference at virtual address 00000040 e096527e *pde = 00000000 Oops: 0000 CPU: 0 EIP: 0010:[<e096527e>] Not taintedUsing defaults from ksymoops -t elf32-i386 -a i386
EFLAGS: 00010246
eax:deb62a94ebx: 00000000 ecx:dd65b400edx: 00000000
esi:0000001cedi:deb62a94ebp: 00000000 esp:dd5fbdbc
ds: 0018 es: 0018 ss: 0018
Process mdadm (pid: 1389, stackpage=dd5fb000)Stack: dd4c4000 dfa96000 c035ad00 00000000 00000286 dd4c4000 00000000 00000000 deb62a94 dd5fbe5c dd4c6000 c02a6e10 dd65b400 c035ef1f 0000007c 00000000 0000000a ffffffff 00000002 00002e2e c0118b49 00002e2e 00002e2e 00000286 Call Trace: [<c02a6e10>] [<c0118b49>] [<c0118cc4>] [<c024a88c>] [<c024abb6>] [<c0118cc4>] [<c024907e>] [<c024b6f2>] [<c024c60c>] [<c014a326>] [<c013c483>] [<c013ca18>] [<c01375ac>] [<c013ca63>] [<c01439b6>] [<c01087c7>] Code: 8b 45 40 85 c0 0f 84 c2 01 00 00 6a 00 ff b4 24 cc 00 00 00Quoted 1 line
Quoted 1 line
EIP;
e096527e<[multipath]multipath_run+2be/6c0> <=====Trace;
c02a6e10<vsnprintf+2e0/450>
Trace;c0118b49<call_console_drivers+e9/f0> Trace;c0118cc4<printk+104/110>
Trace;c024a88c<device_size_calculation+19c/1f0> Trace;c024abb6<do_md_run+2d6/360>
Trace;c0118cc4<printk+104/110>
Trace;c024907e<bind_rdev_to_array+9e/b0> Trace;c024b6f2<add_new_disk+132/290> Trace;c024c60c<md_ioctl+6fc/790>
Trace;c014a326<iput+236/240>
Trace;c013c483<bdput+93/a0>
Trace;c013ca18<blkdev_put+98/a0>
Trace;c01375ac<fput+bc/e0>
Trace;c013ca63<blkdev_ioctl+23/30> Trace;c01439b6<sys_ioctl+216/230>
Trace;c01087c7<system_call+33/38>
Code;e096527e<[multipath]multipath_run+2be/6c0> 00000000 <_EIP>:
Code;e096527e<[multipath]multipath_run+2be/6c0> <=====
0: 8b 45 40 mov 0x40(%ebp),%eax <=====
Code;e0965281<[multipath]multipath_run+2c1/6c0>
3: 85 c0 test %eax,%eax
Code;e0965283<[multipath]multipath_run+2c3/6c0>
5: 0f 84 c2 01 00 00 je 1cd <_EIP+0x1cd>e096544b
<[multipath]m
ultipath_run+48b/6c0>
Code;e0965289<[multipath]multipath_run+2c9/6c0>
b: 6a 00 push $0x0
Code;e096528b<[multipath]multipath_run+2cb/6c0>
d: ff b4 24 cc 00 00 00 pushl 0xcc(%esp,1)My
/etc/mdadm.confcontains:DEVICE /dev/sd?1
ARRAY/dev/md0level=multipath num-devices=1
UUID=277e4ba5:6c23c087:e17c877c:da642955Quoted 7 lines
Should md multipath be able to handle changes like this with the underlying devices?
Thanks
James Pearson
Hi James,
My co-worker and I just happened to run into this problem a few days ago. So, I would like to share with you what we know.
Quoted 8 lines
The device major/minor numbers no longer match up values recorded in the descriptor array in the md superblock. Because of the exception made in the current code, the descriptor entries are removed and although the real devices are present and accounted for, they are kicked out from the array. This leaves the array with zero devices. When
multipath_run()is invoked, it blows up expecting to have had some disks.Lars Marowsky-Brée suggested some patches for md multipath in 2002 but never made it to mainline 2.4 kernel:
http://marc.theaimsgroup.com/?l=linux-kernel&m=103355467608953&w=2That patch is large and most of it is not requried for this particular problem. The section that reinitializes the descriptor array from current rdevs for the case of multipath will resolve this issue of device names shift.
Lars, Is it ok with you if I compose a patch from your original patch and post it here?
Thanks - that patch applies OK to more recent 2.4 kernels and appears to 'fix' this problem.
However, if you have a cut down patch that fixes just this problem, then I would appreciate it if you could make it available.
Thanks
James Pearson
-
To unsubscribe from this list: send the line "unsubscribe linux-raid" in the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html
Quoted 5 lines · James Pearson
Thanks - that patch applies OK to more recent 2.4 kernels and appears to 'fix' this problem.
However, if you have a cut down patch that fixes just this problem, then I would appreciate it if you could make it available.
There's a bugfix needed for 2.4 md multipath which prevents guaranteed data corruption on failover too. I don't have time to redo the diffs against 2.4 proper, but
- bh->b_rdev = bh->b_dev;
- bh->b_rsector = bh->b_blocknr;
are probably the two most important changes to multipath.c:multipathd().
The patch in the SLES8 2.4 kernel is patches.common/md-multipath-retry-handling - there's also some locking fixes etc in there.
The problem is our kernel has deviated so much from 2.4, and active development is now focused on DM mpath in 2.6, that pulling out smaller chunks and feeding them upstream on 2.4 just isn't worth it :-(
Sincerely,
Lars Marowsky-Brée <[email protected]>
Signature · 9 lines
High Availability & Clustering SUSE Labs, Research and Development SUSE LINUX Products GmbH - A Novell Business -- Charles Darwin "Ignorance more frequently begets confidence than does knowledge" - To unsubscribe from this list: send the line "unsubscribe linux-raid" in the body of a message to [email protected] More majordomo info at http://vger.kernel.org/majordomo-info.html
Quoted 161 lines · James Pearson
Quoted 149 lines · Mike Tran
Quoted 123 lines · James Pearson
We have an existing system runing a 2.4.27 based kernel that uses md multipath and external fibre channel arrays.
We need to add more internal disks to the system, which means the external drives change device names.
When I tried to start the md multipath device using mdadm, the kernel Oops'd. Removing the new internal disks and going back the original setup, I can start the multipath device - as this machine is in production, I can't do any more tests.
However, I can reproduce the problem on test system by creating an md multipath device on an external SCSI disk, using
/dev/sda1, stopping the multipath device, rmmod'ing the SCSI driver, pluging in a couple of USB storage devices which become/dev/sdaand/dev/sdband then modprobing the SCSI driver, so the original/dev/sda1is now/dev/sdc1.When I run 'mdadm -A -s', I get the following Oops:
[events: 00000004]
md: bind<sdc1,1>
md: sdc1's event counter: 00000004
md0: former device sda1 is unavailable, removing from array!
md: unbind<sdc1,0>
md:export_rdev(sdc1)
md: RAID level -4 does not need chunksize! Continuing anyway.
md: multipath personality registered as nr 7
md0: max total readahead window set to 124k
md0: 1 data-disks, max readahead per data-disk: 124kUnable to handle kernel NULL pointer dereference at virtual address00000040
printing eip:e096527e
*pde = 00000000Oops: 0000 CPU: 0 EIP: 0010:[<e096527e>] Not tainted EFLAGS: 00010246 eax: deb62a94 ebx: 00000000 ecx: dd65b400 edx: 00000000 esi: 0000001c edi: deb62a94 ebp: 00000000 esp: dd5fbdbc ds: 0018 es: 0018 ss: 0018 Process mdadm (pid: 1389, stackpage=dd5fb000) Stack: dd4c4000 dfa96000 c035ad00 00000000 00000286 dd4c4000 00000000 00000000 deb62a94 dd5fbe5c dd4c6000 c02a6e10 dd65b400 c035ef1f 0000007c 00000000 0000000a ffffffff 00000002 00002e2e c0118b49 00002e2e 00002e2e 00000286 Call Trace: [<c02a6e10>] [<c0118b49>] [<c0118cc4>] [<c024a88c>][<
c024abb6>]
[<c0118cc4>] [<c024907e>] [<c024b6f2>] [<c024c60c>] [<c014a326>]
[<c013c483>]
[<c013ca18>] [<c01375ac>] [<c013ca63>] [<c01439b6>] [<c01087c7>]Code: 8b 45 40 85 c0 0f 84 c2 01 00 00 6a 00 ff b4 24 cc 00 00 00Running through ksymoops gives:
Unable to handle kernel NULL pointer dereference at virtual address 00000040 e096527e *pde = 00000000 Oops: 0000 CPU: 0 EIP: 0010:[<e096527e>] Not taintedUsing defaults from ksymoops -t elf32-i386 -a i386
EFLAGS: 00010246
eax:deb62a94ebx: 00000000 ecx:dd65b400edx: 00000000
esi:0000001cedi:deb62a94ebp: 00000000 esp:dd5fbdbc
ds: 0018 es: 0018 ss: 0018
Process mdadm (pid: 1389, stackpage=dd5fb000)Stack: dd4c4000 dfa96000 c035ad00 00000000 00000286 dd4c4000 00000000 00000000 deb62a94 dd5fbe5c dd4c6000 c02a6e10 dd65b400 c035ef1f 0000007c 00000000 0000000a ffffffff 00000002 00002e2e c0118b49 00002e2e 00002e2e 00000286 Call Trace: [<c02a6e10>] [<c0118b49>] [<c0118cc4>] [<c024a88c>][<
c024abb6>]
[<c0118cc4>] [<c024907e>] [<c024b6f2>] [<c024c60c>] [<c014a326>]
[<c013c483>]
[<c013ca18>] [<c01375ac>] [<c013ca63>] [<c01439b6>] [<c01087c7>]Code: 8b 45 40 85 c0 0f 84 c2 01 00 00 6a 00 ff b4 24 cc 00 00 00Quoted 1 line
Quoted 1 line
EIP;
e096527e<[multipath]multipath_run+2be/6c0> <=====Trace;
c02a6e10<vsnprintf+2e0/450>
Trace;c0118b49<call_console_drivers+e9/f0> Trace;c0118cc4<printk+104/110>
Trace;c024a88c<device_size_calculation+19c/1f0> Trace;c024abb6<do_md_run+2d6/360>
Trace;c0118cc4<printk+104/110>
Trace;c024907e<bind_rdev_to_array+9e/b0> Trace;c024b6f2<add_new_disk+132/290> Trace;c024c60c<md_ioctl+6fc/790>
Trace;c014a326<iput+236/240>
Trace;c013c483<bdput+93/a0>
Trace;c013ca18<blkdev_put+98/a0>
Trace;c01375ac<fput+bc/e0>
Trace;c013ca63<blkdev_ioctl+23/30> Trace;c01439b6<sys_ioctl+216/230>
Trace;c01087c7<system_call+33/38>
Code;e096527e<[multipath]multipath_run+2be/6c0> 00000000 <_EIP>:
Code;e096527e<[multipath]multipath_run+2be/6c0> <=====
0: 8b 45 40 mov 0x40(%ebp),%eax <=====
Code;e0965281<[multipath]multipath_run+2c1/6c0>
3: 85 c0 test %eax,%eax
Code;e0965283<[multipath]multipath_run+2c3/6c0>
5: 0f 84 c2 01 00 00 je 1cd <_EIP+0x1cd>e096544b
<[multipath]m
ultipath_run+48b/6c0>
Code;e0965289<[multipath]multipath_run+2c9/6c0>
b: 6a 00 push $0x0
Code;e096528b<[multipath]multipath_run+2cb/6c0>
d: ff b4 24 cc 00 00 00 pushl 0xcc(%esp,1)My
/etc/mdadm.confcontains:DEVICE /dev/sd?1
ARRAY/dev/md0level=multipath num-devices=1
UUID=277e4ba5:6c23c087:e17c877c:da642955Quoted 8 lines
Should md multipath be able to handle changes like this with the underlying devices?
Thanks
James Pearson
Hi James,
My co-worker and I just happened to run into this problem a few days ago. So, I would like to share with you what we know.
The device major/minor numbers no longer match up values recorded in the descriptor array in the md superblock. Because of the exception made in the current code, the descriptor entries are removed and although the real devices are present and accounted for, they are kicked out from the array. This leaves the array with zero devices. When
multipath_run()is invoked, it blows up expecting to have had some disks.Lars Marowsky-Brée suggested some patches for md multipath in 2002 but never made it to mainline 2.4 kernel:
http://marc.theaimsgroup.com/?l=linux-kernel&m=103355467608953&w=2Quoted 7 lines
That patch is large and most of it is not requried for this particular problem. The section that reinitializes the descriptor array from current rdevs for the case of multipath will resolve this issue of device names shift.
Lars, Is it ok with you if I compose a patch from your original patch and post it here?
Thanks - that patch applies OK to more recent 2.4 kernels and appears to 'fix' this problem.
Quoted 7 lines
However, if you have a cut down patch that fixes just this problem, then I would appreciate it if you could make it available.
Thanks
James Pearson
James,
Here is the reduced patch from the patch that Lars originally produced that worked for us for that particular problem with the multipath disks major/minor numbers shifting. Hopefully, Marcelo can review it and consider it for inclusion in 2.4 mainline. Let us know if this works for you. The credit, of couse, should still go to Lars. We simply picked out the part that fixes that particular issue.
regards,
Signature · 2 lines
Luciano Chavez <[email protected]> IBM
drivers/md/md.c +134 -118
@@ -1320,148 +1320,164 @@memcpy (sb, freshest->sb,sizeof (*sb));/* -* at this point we have picked the 'best' superblock -* from all available superblocks. -* now we validate this superblock and kick out possibly -* failed disks. +* For multipathing, lots of things are different from "true" +* RAIDs. +* All rdev's could be read, so they are no longer faulty. +* As there is just one sb, trying to find changed devices via the +* this_disk pointer is useless too. +* +* [email protected], 2002-09-12 */ - ITERATE_RDEV(mddev,rdev,tmp) { -/* -* Kick all non-fresh devices -*/ -__u64 ev1, ev2; - ev1 =md_event (rdev->sb); - ev2 =md_event (sb); - ++ev1; -if (ev1 < ev2) { -printk (KERN_WARNING "md: kicking non-fresh %s from array!\n" , -partition_name (rdev->dev)); -kick_rdev_from_array (rdev); -continue ; - } - } -/* -* Fix up changed device names ... but only if this disk has a -* recent update time. Use faulty checksum ones too. -*/ -if (mddev->sb->level !=-4 ) - ITERATE_RDEV(mddev,rdev,tmp) { -__u64 ev1, ev2, ev3; -if (rdev->faulty || rdev->alias_device) { -MD_BUG (); - goto abort; - } - ev1 =md_event (rdev->sb); - ev2 =md_event (sb); - ev3 = ev2; - --ev3; -if ((rdev->dev != rdev->old_dev) && - ((ev1 == ev2) || (ev1 == ev3))) { +if (sb->level ==-4 ) { +int desc_nr =0 ; + +/* ... and initialize from the current rdevs instead */ + ITERATE_RDEV(mddev,rdev,tmp) {mdp_disk_t *desc; -printk (KERN_WARNING "md: device name has changed from %s to %s since last import!\n" , -partition_name (rdev->old_dev),partition_name (rdev->dev)); -if (rdev->desc_nr ==-1 ) { -MD_BUG (); - goto abort; - } + rdev->desc_nr=desc_nr; + desc = &sb->disks[rdev->desc_nr]; -if (rdev->old_dev !=MKDEV (desc->major, desc->minor)) { -MD_BUG (); - goto abort; - } - desc->major =MAJOR (rdev->dev); - desc->minor =MINOR (rdev->dev); - desc = &rdev->sb->this_disk; + + desc->number = desc_nr; desc->major =MAJOR (rdev->dev); desc->minor =MINOR (rdev->dev); - } - } + desc->raid_disk = desc_nr; -/* -* Remove unavailable and faulty devices ... -* -* note that if an array becomes completely unrunnable due to -* missing devices, we do not write the superblock back, so the -* administrator has a chance to fix things up. The removal thus -* only happens if it's nonfatal to the contents of the array. -*/ -for (i =0 ; i <MD_SB_DISKS ; i++) { -int found; -mdp_disk_t *desc; -kdev_t dev; +/* We could read from it, so it isn't faulty +* any longer */ +if (disk_faulty (desc)) +mark_disk_spare (desc); - desc = sb->disks + i; - dev =MKDEV (desc->major, desc->minor); +memcpy (&rdev->sb->this_disk,desc,sizeof (*desc)); + + desc_nr++; + } +/* Kick out all old info about disks we used to have, +* if any */ +for (i = desc_nr; i <MD_SB_DISKS ; i++) +memset (&(sb->disks[i]),0 ,sizeof (mdp_disk_t)); + }else {/* -* We kick faulty devices/descriptors immediately. -* -* Note: multipath devices are a special case. Since we -* were able to read the superblock on the path, we don't -* care if it was previously marked as faulty, it's up now -* so enable it. +* at this point we have picked the 'best' superblock +* from all available superblocks. +* now we validate this superblock and kick out possibly +* failed disks. */ -if (disk_faulty (desc) && mddev->sb->level !=-4 ) { - found =0 ; - ITERATE_RDEV(mddev,rdev,tmp) { -if (rdev->desc_nr != desc->number) -continue ; -printk (KERN_WARNING "md%d: kicking faulty %s!\n" , -mdidx (mddev),partition_name (rdev->dev)); -kick_rdev_from_array (rdev); - found =1 ; -break ; - } -if (!found) { -if (dev ==MKDEV (0 ,0 )) -continue ; -printk (KERN_WARNING "md%d: removing former faulty %s!\n" , -mdidx (mddev),partition_name (dev)); - } -remove_descriptor (desc, sb); -continue ; - }else if (disk_faulty (desc)) { + ITERATE_RDEV(mddev,rdev,tmp) {/* -* multipath entry marked as faulty, unfaulty it +* Kick all non-fresh devices */ - rdev =find_rdev (mddev, dev); -if (rdev) -mark_disk_spare (desc); -else -remove_descriptor (desc, sb); +__u64 ev1, ev2; + ev1 =md_event (rdev->sb); + ev2 =md_event (sb); + ++ev1; +if (ev1 < ev2) { +printk (KERN_WARNING "md: kicking non-fresh %s from array!\n" , +partition_name (rdev->dev)); +kick_rdev_from_array (rdev); +continue ; + } } -if (dev ==MKDEV (0 ,0 )) -continue ;/* -* Is this device present in the rdev ring? +* Fix up changed device names ... but only if this disk has a +* recent update time. Use faulty checksum ones too. */ - found =0 ; ITERATE_RDEV(mddev,rdev,tmp) { +__u64 ev1, ev2, ev3; +if (rdev->faulty || rdev->alias_device) { +MD_BUG (); + goto abort; + } + ev1 =md_event (rdev->sb); + ev2 =md_event (sb); + ev3 = ev2; + --ev3; +if ((rdev->dev != rdev->old_dev) && + ((ev1 == ev2) || (ev1 == ev3))) { +mdp_disk_t *desc; + +printk (KERN_WARNING "md: device name has changed from %s to %s since last import!\n" , +partition_name (rdev->old_dev),partition_name (rdev->dev)); +if (rdev->desc_nr ==-1 ) { +MD_BUG (); + goto abort; + } + desc = &sb->disks[rdev->desc_nr]; +if (rdev->old_dev !=MKDEV (desc->major, desc->minor)) { +MD_BUG (); + goto abort; + } + desc->major =MAJOR (rdev->dev); + desc->minor =MINOR (rdev->dev); + desc = &rdev->sb->this_disk; + desc->major =MAJOR (rdev->dev); + desc->minor =MINOR (rdev->dev); + } + } + +/* +* Remove unavailable and faulty devices ... +* +* note that if an array becomes completely unrunnable due to +* missing devices, we do not write the superblock back, so the +* administrator has a chance to fix things up. The removal thus +* only happens if it's nonfatal to the contents of the array. +*/ +for (i =0 ; i <MD_SB_DISKS ; i++) { +int found; +mdp_disk_t *desc; +kdev_t dev; + + desc = sb->disks + i; + dev =MKDEV (desc->major, desc->minor); +/* -* Multi-path IO special-case: since we have no -* this_disk descriptor at auto-detect time, -* we cannot check rdev->number. -* We can check the device though. +* We kick faulty devices/descriptors immediately. */ -if ((sb->level ==-4 ) && (rdev->dev == -MKDEV (desc->major,desc->minor))) { - found =1 ; -break ; +if (disk_faulty (desc)) { + found =0 ; + ITERATE_RDEV(mddev,rdev,tmp) { +if (rdev->desc_nr != desc->number) +continue ; +printk (KERN_WARNING "md%d: kicking faulty %s!\n" , +mdidx (mddev),partition_name (rdev->dev)); +kick_rdev_from_array (rdev); + found =1 ; +break ; + } +if (!found) { +if (dev ==MKDEV (0 ,0 )) +continue ; +printk (KERN_WARNING "md%d: removing former faulty %s!\n" , +mdidx (mddev),partition_name (dev)); + } +remove_descriptor (desc, sb); +continue ; } -if (rdev->desc_nr == desc->number) { - found =1 ; -break ; + +if (dev ==MKDEV (0 ,0 )) +continue ; +/* +* Is this device present in the rdev ring? +*/ + found =0 ; + ITERATE_RDEV(mddev,rdev,tmp) { +if (rdev->desc_nr == desc->number) { + found =1 ; +break ; + } } - } -if (found) -continue ; +if (found) +continue ; -printk (KERN_WARNING "md%d: former device %s is unavailable, removing from array!\n" , -mdidx (mddev),partition_name (dev)); -remove_descriptor (desc, sb); +printk (KERN_WARNING "md%d: former device %s is unavailable, removing from array!\n" , +mdidx (mddev),partition_name (dev)); +remove_descriptor (desc, sb); + } }/*
Quoted 148 lines · Luciano Chavez
Quoted 136 lines · James Pearson
Quoted 125 lines · Mike Tran
Quoted 110 lines · James Pearson
We have an existing system runing a 2.4.27 based kernel that uses md multipath and external fibre channel arrays.
We need to add more internal disks to the system, which means the external drives change device names.
When I tried to start the md multipath device using mdadm, the kernel Oops'd. Removing the new internal disks and going back the original setup, I can start the multipath device - as this machine is in production, I can't do any more tests.
However, I can reproduce the problem on test system by creating an md multipath device on an external SCSI disk, using
/dev/sda1, stopping the multipath device, rmmod'ing the SCSI driver, pluging in a couple of USB storage devices which become/dev/sdaand/dev/sdband then modprobing the SCSI driver, so the original/dev/sda1is now/dev/sdc1.When I run 'mdadm -A -s', I get the following Oops:
[events: 00000004]
md: bind<sdc1,1>
md: sdc1's event counter: 00000004
md0: former device sda1 is unavailable, removing from array!
md: unbind<sdc1,0>
md:export_rdev(sdc1)
md: RAID level -4 does not need chunksize! Continuing anyway.
md: multipath personality registered as nr 7
md0: max total readahead window set to 124k
md0: 1 data-disks, max readahead per data-disk: 124kUnable to handle kernel NULL pointer dereference at virtual address 00000040 printing eip: e096527e *pde = 00000000 Oops: 0000 CPU: 0 EIP: 0010:[<e096527e>] Not tainted EFLAGS: 00010246 eax: deb62a94 ebx: 00000000 ecx: dd65b400 edx: 00000000 esi: 0000001c edi: deb62a94 ebp: 00000000 esp: dd5fbdbc ds: 0018 es: 0018 ss: 0018 Process mdadm (pid: 1389, stackpage=dd5fb000) Stack: dd4c4000 dfa96000 c035ad00 00000000 00000286 dd4c4000 00000000 00000000 deb62a94 dd5fbe5c dd4c6000 c02a6e10 dd65b400 c035ef1f 0000007c 00000000 0000000a ffffffff 00000002 00002e2e c0118b49 00002e2e 00002e2e 00000286 Call Trace: [<c02a6e10>] [<c0118b49>] [<c0118cc4>] [<c024a88c>] [<c024abb6>] [<c0118cc4>] [<c024907e>] [<c024b6f2>] [<c024c60c>] [<c014a326>][<
c013c483>]
[<c013ca18>] [<c01375ac>] [<c013ca63>] [<c01439b6>] [<c01087c7>]Code: 8b 45 40 85 c0 0f 84 c2 01 00 00 6a 00 ff b4 24 cc 00 00 00Running through ksymoops gives:
Unable to handle kernel NULL pointer dereference at virtual address 00000040 e096527e *pde = 00000000 Oops: 0000 CPU: 0 EIP: 0010:[<e096527e>] Not taintedUsing defaults from ksymoops -t elf32-i386 -a i386
EFLAGS: 00010246
eax:deb62a94ebx: 00000000 ecx:dd65b400edx: 00000000
esi:0000001cedi:deb62a94ebp: 00000000 esp:dd5fbdbc
ds: 0018 es: 0018 ss: 0018
Process mdadm (pid: 1389, stackpage=dd5fb000)Stack: dd4c4000 dfa96000 c035ad00 00000000 00000286 dd4c4000 00000000 00000000 deb62a94 dd5fbe5c dd4c6000 c02a6e10 dd65b400 c035ef1f 0000007c 00000000 0000000a ffffffff 00000002 00002e2e c0118b49 00002e2e 00002e2e 00000286 Call Trace: [<c02a6e10>] [<c0118b49>] [<c0118cc4>] [<c024a88c>] [<c024abb6>] [<c0118cc4>] [<c024907e>] [<c024b6f2>] [<c024c60c>] [<c014a326>] [<c013c483>] [<c013ca18>] [<c01375ac>] [<c013ca63>] [<c01439b6>] [<c01087c7>] Code: 8b 45 40 85 c0 0f 84 c2 01 00 00 6a 00 ff b4 24 cc 00 00 00Quoted 1 line
Quoted 1 line
EIP;
e096527e<[multipath]multipath_run+2be/6c0> <=====Trace;
c02a6e10<vsnprintf+2e0/450>
Trace;c0118b49<call_console_drivers+e9/f0> Trace;c0118cc4<printk+104/110>
Trace;c024a88c<device_size_calculation+19c/1f0> Trace;c024abb6<do_md_run+2d6/360>
Trace;c0118cc4<printk+104/110>
Trace;c024907e<bind_rdev_to_array+9e/b0> Trace;c024b6f2<add_new_disk+132/290> Trace;c024c60c<md_ioctl+6fc/790>
Trace;c014a326<iput+236/240>
Trace;c013c483<bdput+93/a0>
Trace;c013ca18<blkdev_put+98/a0>
Trace;c01375ac<fput+bc/e0>
Trace;c013ca63<blkdev_ioctl+23/30> Trace;c01439b6<sys_ioctl+216/230>
Trace;c01087c7<system_call+33/38>
Code;e096527e<[multipath]multipath_run+2be/6c0> 00000000 <_EIP>:Code; e096527e <[multipath]multipath_run+2be/6c0> <===== 0: 8b 45 40 mov 0x40(%ebp),%eax <===== Code; e0965281 <[multipath]multipath_run+2c1/6c0> 3: 85 c0 test %eax,%eax Code; e0965283 <[multipath]multipath_run+2c3/6c0> 5: 0f 84 c2 01 00 00 je 1cd <_EIP+0x1cd> e096544b <[multipath]m ultipath_run+48b/6c0> Code; e0965289 <[multipath]multipath_run+2c9/6c0> b: 6a 00 push $0x0 Code; e096528b <[multipath]multipath_run+2cb/6c0> d: ff b4 24 cc 00 00 00 pushl 0xcc(%esp,1)My
/etc/mdadm.confcontains:DEVICE /dev/sd?1
ARRAY/dev/md0level=multipath num-devices=1
UUID=277e4ba5:6c23c087:e17c877c:da642955Quoted 7 lines
Should md multipath be able to handle changes like this with the underlying devices?
Thanks
James Pearson
Hi James,
My co-worker and I just happened to run into this problem a few days ago. So, I would like to share with you what we know.
Quoted 8 lines
The device major/minor numbers no longer match up values recorded in the descriptor array in the md superblock. Because of the exception made in the current code, the descriptor entries are removed and although the real devices are present and accounted for, they are kicked out from the array. This leaves the array with zero devices. When
multipath_run()is invoked, it blows up expecting to have had some disks.Lars Marowsky-Brée suggested some patches for md multipath in 2002 but never made it to mainline 2.4 kernel:
http://marc.theaimsgroup.com/?l=linux-kernel&m=103355467608953&w=2That patch is large and most of it is not requried for this particular problem. The section that reinitializes the descriptor array from current rdevs for the case of multipath will resolve this issue of device names shift.
Lars, Is it ok with you if I compose a patch from your original patch and post it here?
Quoted 8 lines
Thanks - that patch applies OK to more recent 2.4 kernels and appears to 'fix' this problem.
However, if you have a cut down patch that fixes just this problem, then I would appreciate it if you could make it available.
Thanks
James Pearson
Quoted 8 lines
James,
Here is the reduced patch from the patch that Lars originally produced that worked for us for that particular problem with the multipath disks major/minor numbers shifting. Hopefully, Marcelo can review it and consider it for inclusion in 2.4 mainline. Let us know if this works for you. The credit, of couse, should still go to Lars. We simply picked out the part that fixes that particular issue.
Patch appears to work fine.
Thanks
James Pearson
-
To unsubscribe from this list: send the line "unsubscribe linux-raid" in the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html