long name length protection triggering when it shouldn't

long name length protection triggering when it shouldn't

[Jack Hill]

Hi,

I believe that I'm having a problem where the checks in 
5ed70f7db9afe5e4e5cc1fa2239676e3b2d2c5b2 for file name length are 
triggering when they shouldn't be. I see this problem on a normal Debian 
Stretch system where the patch in question was introduced in Debian's 
linux-image-4.9.0-4-amd64 kernel. I believe that the errors about file 
name length are erroneous because I see no problems when using a prior 
kernel. As an added twist, I only see this problem when using the sec=krb5 
option. If I use sec=ntlmv2, the problem does not occur. I am looking for 
advice on how to troubleshoot further.

More information can be found in a post I made to a different list about 
this issue 
<https://trilug.org/pipermail/trilug/Week-of-Mon-20171023/077662.html>.

Thanks,
Jack

Re: long name length protection triggering when it shouldn't

[Leif Sahlberg]

Thanks for the report.

I had a report yesterday about the name length check failing together with DFS use,
this was fixed by applying this patch :

http://www.spinics.net/lists/linux-cifs/msg13458.html

Can you try that patch and confirm if it fixes the issue for you too ?


regards
ronnie sahlberg


----- Original Message -----
From: "Jack Hill" <jackhill-4+jYJfmkT58@public.gmane.org>
To: linux-cifs-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
Sent: Friday, 27 October, 2017 8:13:27 AM
Subject: long name length protection triggering when it shouldn't

Hi,

I believe that I'm having a problem where the checks in 
5ed70f7db9afe5e4e5cc1fa2239676e3b2d2c5b2 for file name length are 
triggering when they shouldn't be. I see this problem on a normal Debian 
Stretch system where the patch in question was introduced in Debian's 
linux-image-4.9.0-4-amd64 kernel. I believe that the errors about file 
name length are erroneous because I see no problems when using a prior 
kernel. As an added twist, I only see this problem when using the sec=krb5 
option. If I use sec=ntlmv2, the problem does not occur. I am looking for 
advice on how to troubleshoot further.

More information can be found in a post I made to a different list about 
this issue 
<https://trilug.org/pipermail/trilug/Week-of-Mon-20171023/077662.html>.

Thanks,
Jack
--
To unsubscribe from this list: send the line "unsubscribe linux-cifs" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: long name length protection triggering when it shouldn't

[Jack Hill]

On Thu, 26 Oct 2017, Leif Sahlberg wrote:

> Can you try that patch and confirm if it fixes the issue for you too ?

I am happy to confirm that the patch fixes the issue for me.

Is it worth trying to figure out why MaxPathNameComponentLength is not 
being set correctly? For me, it seems to be set correctly when I use 
different mount options (either a different sec= option, or when using 
multiuser).

Do you expect the proposed patch to be the eventual fix that ends up in a 
released kernel?

Thanks,
Jack

Re: long name length protection triggering when it shouldn't

[ronnie sahlberg]

Thanks for testing and confirming the patch fixes the issue.

I will try to figure out why the mount options affect this. But that
is not as urgent.





On Sat, Oct 28, 2017 at 10:58 AM, Jack Hill <jackhill-4+jYJfmkT58@public.gmane.org> wrote:
> On Thu, 26 Oct 2017, Leif Sahlberg wrote:
>
>> Can you try that patch and confirm if it fixes the issue for you too ?
>
>
> I am happy to confirm that the patch fixes the issue for me.
>
> Is it worth trying to figure out why MaxPathNameComponentLength is not being
> set correctly? For me, it seems to be set correctly when I use different
> mount options (either a different sec= option, or when using multiuser).
>
> Do you expect the proposed patch to be the eventual fix that ends up in a
> released kernel?
>
>
> Thanks,
> Jack
> --
> To unsubscribe from this list: send the line "unsubscribe linux-cifs" in
> the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: long name length protection triggering when it shouldn't

[Jack Hill]

On Sat, 28 Oct 2017, ronnie sahlberg wrote:

> Thanks for testing and confirming the patch fixes the issue.

You're welcome, and thanks for your work.

Jack