Re: [PATCH -next 2/3] net: sched: remove AT INGRESS/EGRESS

[Jamal Hadi Salim <jhs@mojatatu.com>]

On 05/15/15 16:09, Alexei Starovoitov wrote:
> On Fri, May 15, 2015 at 07:21:15PM +0200, Florian Westphal wrote:
>> So, from ifb point of view it makes no difference, G_TC_FROM+AT_STACK
>> causes skb to be dropped and IFB doesn't care about G_TC_AT() at all.
>
> yes. your change is technically correct. It's not causing ifb regression,
> but it removes information in a way that will be very hard to add it later.
>
>> AT_STACK cannot even happen for the G_TC_AT case from looking at the
>> code since dev_queue_xmit forces AT_EGRESS & rx sets AT_INGRESS.
>
> yes, if we only consider ingress and egress hooks.
> I want to use this stack/ingress/egress indication with socket filters.
> If we make stack==egress, I would need to refactor this code all over again.
> It's not broken today. You're doing this aliasing only two squeeze a bit.
> That's why I'm saying keep the stack/ingress/egress flag as-is. It's useful.
>

My point as well. Using ifb or mirred as examples is fine
but they are not the only potential consumers/producers. Using examples
as such is out of place when it is an architectural issue.
So i would rather this be left alone.

cheers,
jamal