On Fri, 2006-30-06 at 22:08 +0200, Thomas Graf wrote: > * jamal <[EMAIL PROTECTED]> 2006-06-30 15:34
> > So assuming it is taken in mirred (i havent thought of where it is > > decremented), why would using the ifindex be better? > > The issue exists regardless of mirred/ifb. As soon as the packet is > queued for the first time we leave netif_receive_skb() and the dev > reference is dropped. Therefore in order to allow functionality > like tcf_match_indev() at egress we have to either take a reference > or ensure that we can catch the unlikely case of the device having > disappeared. I think everyone would agree to use device pointers > if only mirred would acquire it. Thomas, this is certainly more reasonable explanation. On Fri, 2006-30-06 at 13:17 -0700, David Miller wrote: > We have a > lot of bits of state that sit in the sk_buff but which are used by > a tiny minority, yet the space consumption is eaten by everyone. > > > Maybe when the IFB and mirred are in more widespread use we can > investigate a different approach. > Not unreasonable. So saving the bits for the majority trumps the need for slightly more cycles in mirred/ifb until they get more widespread. > Another part of this issue is that if you use a pointer there is no > clean place to clean up the input_dev reference within the scope it is > actually used. > > > The only reliable place is kfree_skb() which is far > beyond the scope that this ->input_dev thing is used. > And i cant think of a good place to do it. > We have a lot > of problems in some parts of the networking because object references > are held for too long. One example is all the awful side effects of > the way the bridge netfilter code holds onto object references for > long periods of time in the netfilter call chains. > > And we know the refcounting is broken. And one way we know to fix > the refcounting without incurring the cost upon everyone is to > use an interface index and only make the input_dev users eat the > atomic operation incurred by grabbing the reference. > > And even for the input_dev users, it isn't such a big deal, there > are other costs already associated with hitting these actions and > devices that use input_dev, the atomic reference grab there should > be lost in the noise I think. Alright guys - lets get the change in. I wish we had this specific discussion sooner. Thomas, can you post your latest incarnation so i can mow comment more peacefully? ;-> cheers, jamal - To unsubscribe from this list: send the line "unsubscribe netdev" in the body of a message to [EMAIL PROTECTED] More majordomo info at http://vger.kernel.org/majordomo-info.html
