[Bug 13037] reiserfs panic (device null superblock)

[Bug 13037] reiserfs panic (device null superblock)

[bugzilla-daemon]

http://bugzilla.kernel.org/show_bug.cgi?id=13037





--- Comment #3 from Pawel Zawora <pzawora@gmail.com>  2009-06-30 12:20:51 ---
Is any chance to solve that issue ? Is any workaround ?

-- 
Configure bugmail: http://bugzilla.kernel.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug.

[Bug 13037] reiserfs panic (device null superblock)

[bugzilla-daemon]

http://bugzilla.kernel.org/show_bug.cgi?id=13037





--- Comment #4 from Jeff Mahoney <jeffm@suse.com>  2009-07-02 15:00:56 ---
So the problem exists on 2.6.28.9 and not on 2.6.28? That's strange since the
only reiserfs change is this one:

--- a/fs/reiserfs/inode.c
+++ b/fs/reiserfs/inode.c
@@ -2556,7 +2556,7 @@ static int reiserfs_write_begin(struct file *file,
        }

        index = pos >> PAGE_CACHE_SHIFT;
-       page = __grab_cache_page(mapping, index);
+       page = grab_cache_page_write_begin(mapping, index, flags);
        if (!page)
                return -ENOMEM;
        *pagep = page;

I'm surprised that reiserfsck doesn't come up with anything. This is a problem
where it's seeing an invalid bit on disk. Either that or memory corruption.

-- 
Configure bugmail: http://bugzilla.kernel.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug.

[Bug 13037] reiserfs panic (device null superblock)

[bugzilla-daemon]

http://bugzilla.kernel.org/show_bug.cgi?id=13037





--- Comment #5 from Pawel Zawora <pzawora@gmail.com>  2009-07-02 17:33:59 ---
bugzilla-daemon@bugzilla.kernel.org wrote:
> http://bugzilla.kernel.org/show_bug.cgi?id=13037
>
>
>
>
>
> --- Comment #4 from Jeff Mahoney <jeffm@suse.com>  2009-07-02 15:00:56 ---
> So the problem exists on 2.6.28.9 and not on 2.6.28?
/look at
Description from  Pawel Zawora 2009-04-07 11:38:17/

[...]

On 2.6.25.4 works stable. 
Upgrade to 2.6.28.7 and 2.6.28.9 causes kernel panic[...]



> That's strange since the
> only reiserfs change is this one:
>
> --- a/fs/reiserfs/inode.c
> +++ b/fs/reiserfs/inode.c
> @@ -2556,7 +2556,7 @@ static int reiserfs_write_begin(struct file *file,
>         }
>
>         index = pos >> PAGE_CACHE_SHIFT;
> -       page = __grab_cache_page(mapping, index);
> +       page = grab_cache_page_write_begin(mapping, index, flags);
>         if (!page)
>                 return -ENOMEM;
>         *pagep = page;
>
> I'm surprised that reiserfsck doesn't come up with anything. This is a problem
> where it's seeing an invalid bit on disk. Either that or memory corruption.
>
>   
/look at
Description from  Pawel Zawora 2009-04-07 11:38:17/

[...]

We tested:
1. Completely different hardware [...]

-- 
Configure bugmail: http://bugzilla.kernel.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug.

[Bug 13037] reiserfs panic (device null superblock)

[bugzilla-daemon]

http://bugzilla.kernel.org/show_bug.cgi?id=13037





--- Comment #6 from Jeff Mahoney <jeffm@suse.com>  2009-07-02 18:44:07 ---
Ok, I'm going through the reiserfs differences between 2.6.25 and .28. Can you
try to reproduce with 2.6.30?

-- 
Configure bugmail: http://bugzilla.kernel.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug.

[Bug 13037] reiserfs panic (device null superblock)

[bugzilla-daemon]

http://bugzilla.kernel.org/show_bug.cgi?id=13037


Roland Kletzing <devzero@web.de> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |devzero@web.de




--- Comment #7 from Roland Kletzing <devzero@web.de>  2009-09-18 16:16:54 ---
isn`t this an invalid bugreport, as DRBD is NOT part of the mainline kernel?

i`t not against drbd - i whished it was in mainline (please vote for that on
lkml, as there is discussion going on about inclusion) - but until it IS, this
is an unsupported secnario. isn`t it ?

-- 
Configure bugmail: http://bugzilla.kernel.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug.

[Bug 13037] reiserfs panic (device null superblock)

[bugzilla-daemon]

http://bugzilla.kernel.org/show_bug.cgi?id=13037





--- Comment #8 from Pawel Zawora <pzawora@gmail.com>  2009-09-23 16:39:39 ---
It seems that problem is related to external journaling and network device. I
was able to reproduce problem on iSCSI device (using external journaling).
Currently  I have to abandon external journaling: "standard" reiserfs works
over my drbd without any problem...

-- 
Configure bugmail: http://bugzilla.kernel.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug.

[Bug 13037] reiserfs panic (device null superblock)

[bugzilla-daemon]

https://bugzilla.kernel.org/show_bug.cgi?id=13037





--- Comment #9 from Pawel Zawora <pzawora@gmail.com>  2010-05-07 15:46:55 ---
Is there someone who will works on that issue?
Reiserfs is still fast and easy recoverable filesystem...  
There is no maintainer?

-- 
Configure bugmail: https://bugzilla.kernel.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug.

[Bug 13037] reiserfs panic (device null superblock)

[bugzilla-daemon]

https://bugzilla.kernel.org/show_bug.cgi?id=13037


Christian Kujau <lists@nerdbynature.de> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |lists@nerdbynature.de




--- Comment #10 from Christian Kujau <lists@nerdbynature.de>  2010-05-08 22:12:25 ---
A clean testcase would be nice. I tested 2.6.34-rc6 with reiserfs and an
external log; copied 11000000 files around, copied 2 kernel trees around - no
panic. It's a multicore machine too, but no DRBD or NFS involved.
If these variables make a difference, please describe your (exact) setup so
that one can try to reproduce this.

-- 
Configure bugmail: https://bugzilla.kernel.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug.

[Bug 13037] reiserfs panic (device null superblock)

[bugzilla-daemon]

https://bugzilla.kernel.org/show_bug.cgi?id=13037


Alan <alan@lxorguk.ukuu.org.uk> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |alan@lxorguk.ukuu.org.uk
     Kernel Version|2.6.28.9                    |2.6.34




--- Comment #11 from Pawel Zawora <pzawora@gmail.com>  2010-05-10 10:19:58 ---
I can confirm that:
1. reiser over local hard disk works stable,
2. is possible to reproduce problem using internal journaling.

Problem appears in that case:
NFS(v3) ---> resier ----> storage "depends on network" (DRBD/iSCSI). 

My environment:
multiple (4+) nfs clients(noacl,udp) [multiple simultaneous file access on the
same time] uses failover nfs server . NFS works over DRBD (active/passive ;
async protocol). DRBD uses dedicated network connection. "panic time" is depend
on load.

NFS server works stable with kernel <= 2.6.25 and new kernel with ie XFS
(always over DRBD)

Additonally, on the same cluster - when I try switch nodes - stopping nfsd
causes bug reported as 14055.

-- 
Configure bugmail: https://bugzilla.kernel.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug.

[Bug 13037] reiserfs panic (device null superblock)

[bugzilla-daemon]

https://bugzilla.kernel.org/show_bug.cgi?id=13037

Alan <alan@lxorguk.ukuu.org.uk> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|---                         |OBSOLETE

-- 
You are receiving this mail because:
You are the assignee for the bug.