From: Herbert Xu <[EMAIL PROTECTED]> Date: Sun, 31 Jul 2005 17:17:56 +1000
> On Sat, Jul 30, 2005 at 11:44:55PM -0700, David S. Miller wrote: > > > > I wish I could come up with a sane way to use the inetpeer cache to > > store this info. Unfortunately we need to key on address _and_ ID. > > However, this is even better :) Except that I'm confused by what you > said here that we need to key on the ID. Why do we need to key on > the ID? I would've thought that the saddr is sufficient. Well... Looking at your code, I came away understanding that you needed to make the IPC objects unique per saddr/ID. That's what you hash them by. But thinking some more, this appears to be no more than an implementation detail, ie. so that the IPCs show up on the same hash chains as the queues using them, thus making them easy to find. So do you need these window measurements IPC object to be per saddr or per saddr+ID? If they only need to be per saddr alone, then we can and should use the inetpeer cache for this. I would also suggest that some reasonable non-zero default max_win be choosen. This stuff will not get tested otherwise. - 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
