[Ocfs2-devel] [PATCH]ocfs2-1.2: Add dput for uuid entry.

[Ocfs2-devel] [PATCH]ocfs2-1.2: Add dput for uuid entry.

[Tao Ma]

In ocfs2-1.2, when we mount a device, a debugfs dir will be created 
using its uuid. When 2 devices have the same uuid, after the 1st device 
is mounted, the 2nd one can't be mounted. This is OK. But the problem is 
that the dentry's reference is added. So when the 1st volume is 
umounted, none of these 2 volumes can be mounted then. So this fix 
solves this problem by dputting the entry(Copied from the latest debugfs 
code in Linux kernel).
-------------- next part --------------
A non-text attachment was scrubbed...
Name: debugfs.patch
Type: text/x-patch
Size: 446 bytes
Desc: not available
Url : http://oss.oracle.com/pipermail/ocfs2-devel/attachments/20080507/ffa93e59/attachment.bin 

[Ocfs2-devel] [PATCH]ocfs2-1.2: Add dput for uuid entry.

[Tao Ma]

Some more information.
The git commit log in the kernel is: 
65c333367b1aea57d58168ad3dc1df27b0227401
http://www.uwsg.indiana.edu/hypermail/linux/kernel/0611.3/0125.html.

In this patch the debugfs_remove is modified(dput is removed), but in 
the latest fs/debugfs/inode.c, dput is still there. So now I really 
don't know whether this function should be modified or not.

Any advice is welcomed.
Thanks.

Regards,
Tao

Tao Ma wrote:
> In ocfs2-1.2, when we mount a device, a debugfs dir will be created 
> using its uuid. When 2 devices have the same uuid, after the 1st device 
> is mounted, the 2nd one can't be mounted. This is OK. But the problem is 
> that the dentry's reference is added. So when the 1st volume is 
> umounted, none of these 2 volumes can be mounted then. So this fix 
> solves this problem by dputting the entry(Copied from the latest debugfs 
> code in Linux kernel).
> 
> 
> ------------------------------------------------------------------------
> 
> _______________________________________________
> Ocfs2-devel mailing list
> Ocfs2-devel at oss.oracle.com
> http://oss.oracle.com/mailman/listinfo/ocfs2-devel

[Ocfs2-devel] [PATCH]ocfs2-1.2: Add dput for uuid entry.

[Tao Ma]

aha, sorry for the noise. :(

dput is added in the next commit 29a7f3ada7fea5510504c5359c3f70d109aeb055.
http://www.uwsg.indiana.edu/hypermail/linux/kernel/0611.3/0126.html

So here my question is should we patch these 2 patches into our 
ocfs2-1.2 or just the 1st one?

Regards,
Tao

Tao Ma wrote:
> Some more information.
> The git commit log in the kernel is: 
> 65c333367b1aea57d58168ad3dc1df27b0227401
> http://www.uwsg.indiana.edu/hypermail/linux/kernel/0611.3/0125.html.
> 
> In this patch the debugfs_remove is modified(dput is removed), but in 
> the latest fs/debugfs/inode.c, dput is still there. So now I really 
> don't know whether this function should be modified or not.
> 
> Any advice is welcomed.
> Thanks.
> 
> Regards,
> Tao
> 
> Tao Ma wrote:
>> In ocfs2-1.2, when we mount a device, a debugfs dir will be created 
>> using its uuid. When 2 devices have the same uuid, after the 1st 
>> device is mounted, the 2nd one can't be mounted. This is OK. But the 
>> problem is that the dentry's reference is added. So when the 1st 
>> volume is umounted, none of these 2 volumes can be mounted then. So 
>> this fix solves this problem by dputting the entry(Copied from the 
>> latest debugfs code in Linux kernel).
>>
>>
>> ------------------------------------------------------------------------
>>
>> _______________________________________________
>> Ocfs2-devel mailing list
>> Ocfs2-devel at oss.oracle.com
>> http://oss.oracle.com/mailman/listinfo/ocfs2-devel
> 

[Ocfs2-devel] [PATCH]ocfs2-1.2: Add dput for uuid entry.

[Joel Becker]

On Wed, May 07, 2008 at 05:23:32PM +0800, Tao Ma wrote:
> aha, sorry for the noise. :(
> 
> dput is added in the next commit 29a7f3ada7fea5510504c5359c3f70d109aeb055.
> http://www.uwsg.indiana.edu/hypermail/linux/kernel/0611.3/0126.html
> 
> So here my question is should we patch these 2 patches into our 
> ocfs2-1.2 or just the 1st one?

	Well, this is all stuff for debugfs, which is usually part of
the kernel, right?  Or does El4 still not have it?

Joel

-- 

"I don't want to achieve immortality through my work; I want to
 achieve immortality through not dying."
        - Woody Allen

Joel Becker
Principal Software Developer
Oracle
E-mail: joel.becker at oracle.com
Phone: (650) 506-8127