BUG: jumbo frames broken after commit xen-netfront: Fix race between device setup and open

BUG: jumbo frames broken after commit xen-netfront: Fix race between device setup and open

[Andrew Jeddeloh]

Hi all,

The patch "xen-netfront: Fix race between device setup and open" seems
to have introduced a regression preventing setting MTU's larger than
1500. We experienced this downstream with Container Linux and
confirmed with Fedora 28 as well.

It's commit f599c64fdf7d9c108e8717fb04bc41c680120da4 in the linux-stable tree.
https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux-stable.git/commit/?id=f599c64fdf7d9c108e8717fb04bc41c680120da4

Downstream bugs:
https://github.com/coreos/bugs/issues/2443
https://bugzilla.redhat.com/show_bug.cgi?id=1584216

We've confirmed that reverting that commit fixes the bug. It be
reliably can be reproduced on AWS with t2.micro instances (and
presumably other systems using the same driver). Both using
systemd-networkd to set the mtu and manual ip link commands cause the
link to repsond with "Invalid argument" when trying to set the MTU >
1500.

I'm not sure why that commit introduced the regression.

Please let me know if there's any more information that would be helpful.

- Andrew

Re: BUG: jumbo frames broken after commit xen-netfront: Fix race between device setup and open

[Javier Martinez Canillas]

Hi Andrew,

On Wed, Jun 6, 2018 at 6:29 PM, Andrew Jeddeloh
<andrew.jeddeloh@redhat.com> wrote:
> Hi all,
>
> The patch "xen-netfront: Fix race between device setup and open" seems
> to have introduced a regression preventing setting MTU's larger than
> 1500. We experienced this downstream with Container Linux and
> confirmed with Fedora 28 as well.
>
> It's commit f599c64fdf7d9c108e8717fb04bc41c680120da4 in the linux-stable tree.
> https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux-stable.git/commit/?id=f599c64fdf7d9c108e8717fb04bc41c680120da4
>
> Downstream bugs:
> https://github.com/coreos/bugs/issues/2443
> https://bugzilla.redhat.com/show_bug.cgi?id=1584216
>
> We've confirmed that reverting that commit fixes the bug. It be
> reliably can be reproduced on AWS with t2.micro instances (and
> presumably other systems using the same driver). Both using
> systemd-networkd to set the mtu and manual ip link commands cause the
> link to repsond with "Invalid argument" when trying to set the MTU >
> 1500.
>
> I'm not sure why that commit introduced the regression.
>
> Please let me know if there's any more information that would be helpful.
>
> - Andrew

I'm adding some relevant people to the CC list to bring more attention
on this regression.

The get_maintainer.pl script is very useful to get some hints on who
should be copied, i.e:

$ ./scripts/get_maintainer.pl -f drivers/net/xen-netfront.c

Best regards,
Javier

Re: BUG: jumbo frames broken after commit xen-netfront: Fix race between device setup and open

[Liam Shepherd]

[-- Attachment #1.1: Type: text/plain, Size: 1076 bytes --]

> The patch "xen-netfront: Fix race between device setup and open" seems
> to have introduced a regression preventing setting MTU's larger than
> 1500. We experienced this downstream with Container Linux and
> confirmed with Fedora 28 as well.
>
> It's commit f599c64fdf7d9c108e8717fb04bc41c680120da4 in the linux-stable
tree.
> https://git.kernel.org/pub/scm/linux/kernel/git/stable/
linux-stable.git/commit/?id=f599c64fdf7d9c108e8717fb04bc41c680120da4
>
> Downstream bugs:
> https://github.com/coreos/bugs/issues/2443
> https://bugzilla.redhat.com/show_bug.cgi?id=1584216
>
> We've confirmed that reverting that commit fixes the bug. It be
> reliably can be reproduced on AWS with t2.micro instances (and
> presumably other systems using the same driver). Both using
> systemd-networkd to set the mtu and manual ip link commands cause the
> link to repsond with "Invalid argument" when trying to set the MTU >
> 1500.

Sorry for the spam, but this issue I believe this issue is linked with:
https://lists.xenproject.org/archives/html/xen-devel/2018-06/msg00254.html

Thanks

[-- Attachment #1.2: Type: text/html, Size: 1842 bytes --]

[-- Attachment #2: Type: text/plain, Size: 157 bytes --]

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xenproject.org
https://lists.xenproject.org/mailman/listinfo/xen-devel

Re: BUG: jumbo frames broken after commit xen-netfront: Fix race between device setup and open

[Juergen Gross]

On 14/06/18 17:43, Javier Martinez Canillas wrote:
> Hi Andrew,
> 
> On Wed, Jun 6, 2018 at 6:29 PM, Andrew Jeddeloh
> <andrew.jeddeloh@redhat.com> wrote:
>> Hi all,
>>
>> The patch "xen-netfront: Fix race between device setup and open" seems
>> to have introduced a regression preventing setting MTU's larger than
>> 1500. We experienced this downstream with Container Linux and
>> confirmed with Fedora 28 as well.
>>
>> It's commit f599c64fdf7d9c108e8717fb04bc41c680120da4 in the linux-stable tree.
>> https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux-stable.git/commit/?id=f599c64fdf7d9c108e8717fb04bc41c680120da4
>>
>> Downstream bugs:
>> https://github.com/coreos/bugs/issues/2443
>> https://bugzilla.redhat.com/show_bug.cgi?id=1584216
>>
>> We've confirmed that reverting that commit fixes the bug. It be
>> reliably can be reproduced on AWS with t2.micro instances (and
>> presumably other systems using the same driver). Both using
>> systemd-networkd to set the mtu and manual ip link commands cause the
>> link to repsond with "Invalid argument" when trying to set the MTU >
>> 1500.
>>
>> I'm not sure why that commit introduced the regression.
>>
>> Please let me know if there's any more information that would be helpful.
>>
>> - Andrew
> 
> I'm adding some relevant people to the CC list to bring more attention
> on this regression.
> 
> The get_maintainer.pl script is very useful to get some hints on who
> should be copied, i.e:
> 
> $ ./scripts/get_maintainer.pl -f drivers/net/xen-netfront.c

Ross, have you made any progress here? If not I'm thinking of reverting
your patch as I think the current problem is more severe than the one
your patch did address.


Juergen

Re: BUG: jumbo frames broken after commit xen-netfront: Fix race between device setup and open

[Ross Lagerwall]

On 06/21/2018 10:05 AM, Juergen Gross wrote:
> On 14/06/18 17:43, Javier Martinez Canillas wrote:
>> Hi Andrew,
>>
>> On Wed, Jun 6, 2018 at 6:29 PM, Andrew Jeddeloh
>> <andrew.jeddeloh@redhat.com> wrote:
>>> Hi all,
>>>
>>> The patch "xen-netfront: Fix race between device setup and open" seems
>>> to have introduced a regression preventing setting MTU's larger than
>>> 1500. We experienced this downstream with Container Linux and
>>> confirmed with Fedora 28 as well.
>>>
>>> It's commit f599c64fdf7d9c108e8717fb04bc41c680120da4 in the linux-stable tree.
>>> https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux-stable.git/commit/?id=f599c64fdf7d9c108e8717fb04bc41c680120da4
>>>
>>> Downstream bugs:
>>> https://github.com/coreos/bugs/issues/2443
>>> https://bugzilla.redhat.com/show_bug.cgi?id=1584216
>>>
>>> We've confirmed that reverting that commit fixes the bug. It be
>>> reliably can be reproduced on AWS with t2.micro instances (and
>>> presumably other systems using the same driver). Both using
>>> systemd-networkd to set the mtu and manual ip link commands cause the
>>> link to repsond with "Invalid argument" when trying to set the MTU >
>>> 1500.
>>>
>>> I'm not sure why that commit introduced the regression.
>>>
>>> Please let me know if there's any more information that would be helpful.
>>>
>>> - Andrew
>>
>> I'm adding some relevant people to the CC list to bring more attention
>> on this regression.
>>
>> The get_maintainer.pl script is very useful to get some hints on who
>> should be copied, i.e:
>>
>> $ ./scripts/get_maintainer.pl -f drivers/net/xen-netfront.c
> 
> Ross, have you made any progress here? If not I'm thinking of reverting
> your patch as I think the current problem is more severe than the one
> your patch did address.
> 
> 

I've sent a patch today which I believe fixes the issue. Sorry for the 
slow response.

Regards,
-- 
Ross Lagerwall

Re: BUG: jumbo frames broken after commit xen-netfront: Fix race between device setup and open

[Brandon Holbrook]

We are experiencing this bug on stable Gentoo (4.14.67) on EC2.  I see the original "Fix race between device setup and open" commit is present in 4.14, but the fix commit(s) have not yet been backported.  As your fix seems to have resolved the issue and not caused any other ill effects, will you please?

Thanks!
-Brandon Holbrook