[ovs-dev] Revert "dp-packet: Handle multi-seg mbufs in resize__()."

[Aaron Conole]

Ian Stokes <ian.stokes at intel.com> writes:

> On 7/25/2018 2:56 PM, Aaron Conole wrote:
>> 0-day Robot <robot at bytheb.org> writes:
>>
>>> Bleep bloop.  Greetings Tiago Lam, I am a robot and I have tried out your patch.
>>> Thanks for your contribution.
>>>
>>> I encountered some error that I wasn't expecting.  See the details below.
>>>
>>
>> Hi Ian (and all),
>>
>> Given there are currently two trees (mainline and dpdk_merge), do you
>> have any preferences, suggestions, or comments on managing the robot
>> w.r.t. these kind of tree-specific patches?  I suspect that we'll have
>> others in the future, and I'd like to make the robot be a bit friendlier
>> this way (plus it helps to build confidence when there are fewer
>> false-positives).
>>
>> -Aaron
>
> Is it that you'd like to see 0-day running on dpdk_merge also?

I think I was wondering if you would like to see it running against
dpdk_merge. :)

> From my side I think the revert here was an exception, as the pull
> request had not been merged at this point but the assumption was that
> it had been. I encourage people to base patches on what is in master,
> not dpdk_merge.

Good to know.  As a side, maybe it's time to formally document things so
that folks will base their contributions on the mainline always?
The growing pains of multi-tree projects :)

> It's a valid point however, I'm in favor of keeping the mainline as
> clean as possible so typically with a pull request I'll rebase
> dpdk_merge to the head of master, then apply dpdk specific patches on
> top of that so as to ensure its a clean pull request when applying to
> mainline.
>
> If an error does occurs (such as the revert, or a major bug discovery)
> I'll typically rework the patches at the head of dpdk_merge to take
> this into account. I didnt want to see a comit and a revert going into
> the mainline if possible.
>
> I'm happy for 0-day to work on master as is for the moment and
> coordinate with a submitter if a patch is needed and how we handle it?

Okay, that makes sense to me.  For now, I'll leave it running as-is.

Thanks, Ian!

> Ian
>> _______________________________________________
>> dev mailing list
>> dev at openvswitch.org
>> https://mail.openvswitch.org/mailman/listinfo/ovs-dev
>>