Hi Eric, On Thu, Mar 24, 2016 at 07:13:33AM -0700, Eric Dumazet wrote: > On Thu, 2016-03-24 at 07:12 +0100, Willy Tarreau wrote: > > Hi, > > > > On Wed, Mar 23, 2016 at 10:10:06PM -0700, Tolga Ceylan wrote: > > > I apologize for not properly following up on this. I had the > > > impression that we did not want to merge my original patch and then I > > > also noticed that it fails to keep the hash consistent. Recently, I > > > read the follow ups on it as well as Willy's patch/proposals. > > > > > > Is there any update on Willy's SO_REUSEPORT patch? IMHO, it solves the > > > problem and it is simpler than adding new sock option. > > > > no, Craig's changes were merged, and I haven't checked yet if my patch > > needs to be rebased or still applies. Feel free to check it and resubmit > > if you have time. > > No need for a patch AFAIK. > > BPF solution is generic enough. > > All user space needs to do is to update the BPF filter so that the > listener needing to be dismantled does not receive any new packet.
But that means that any software making use of SO_REUSEPORT needs to also implement BPF on Linux to achieve the same as what it does on other OSes ? Also I found a case where a dying process would still cause trouble in the accept queue, maybe it's not redistributed, I don't remember, all I remember is that my traffic stopped after a segfault of only one of them :-/ I'll have to dig a bit regarding this. Thanks, Willy
