[ovs-dev] [PATCH] linux: Signal datapath that unaligned Netlink message can be received

[Jesse Gross]

On Wed, Nov 13, 2013 at 5:46 PM, Thomas Graf <tgraf at redhat.com> wrote:
> On 11/13/2013 07:11 AM, Jesse Gross wrote:
>>
>> On Mon, Nov 11, 2013 at 11:53 PM, Thomas Graf <tgraf at redhat.com> wrote:
>>>
>>> On 11/11/2013 04:50 PM, Ben Pfaff wrote:
>>>>
>>>>
>>>> On Mon, Nov 11, 2013 at 04:36:24PM +0100, Thomas Graf wrote:
>>>>>
>>>>>
>>>>> Following commit (''netlink: Do not enforce alignment of last Netlink
>>>>> attribute''), signal the ability to receive unaligned Netlink messages
>>>>> to the datapath to enable utilization of zerocopy optimizations.
>>>>>
>>>>> Signed-off-by: Thomas Graf <tgraf at redhat.com>
>>>>
>>>>
>>>>
>>>> Seems OK from a userspace point of view.  I am a little concerned that
>>>> downgrading userspace without deleting and re-creating the datapath
>>>> (e.g. via "force-reload-kmod") will result in a totally broken setup
>>>> since userspace will then drop every packet from the kernel.
>>>
>>>
>>>
>>> Is that something that occurs occasionally in installations? Utilizing
>>> the version field in the genl header could be used to track this and
>>> clear user_features.
>>
>>
>> It's probably a good idea. I could see us having more of these
>> features flags in the future (although obviously we should try to
>> avoid them if possible) and, as Ben said, it would potentially lead to
>> a bad state otherwise.
>>
>> I'm not sure exactly what you have in mind though, can you elaborate a
>> little?
>
>
> My initial thought was to use a version field to notice the replacement
> of user space. On second thought that is not required, modifying user
> space to provide the user features in OVS_DP_CMD_GET as well will
> overwrite the features. Resetting user_features to 0 if not features are
> provided will provide backwards compatibility to versions not aware of
> user features yet. Thoughts?

Hmm, it doesn't really seem ideal to make DP_CMD_GET change settings
as it's probably not the expected behavior for most users of the
command. One example of where this could be a problem is if it is
called from both ovs-dpctl and ovs-vswitch. Usually these would be
have the same capabilities but I'm not sure that it's strictly
required in all cases. DP_CMD_SET seems like the ideal place to put it
but I guess we don't use that at all on existing versions of OVS
userspace.

>> (By the way, it might be a good idea to keep the same CC list on all
>> of the patches. Otherwise, some people might miss parts of the
>> discussion.)
>
>
> I did not CC netdev because this is a pure user space patch and the
> respective kernel bits are included in v5 of the kernel series.

I know but I think the outcome of this discussion likely affects all
of the patches, so it would be helpful even for those who are looking
at just the kernel patches.