From mboxrd@z Thu Jan 1 00:00:00 1970 From: Eric Dumazet Subject: Re: [PATCH 1/1] net: Add SO_REUSEPORT_LISTEN_OFF socket option as drain mode Date: Sat, 26 Sep 2015 19:24:15 -0700 Message-ID: <1443320655.3273.7.camel@edumazet-glaptop2.roam.corp.google.com> References: <1443313848-751-1-git-send-email-tolga.ceylan@gmail.com> <877fnc7i0i.fsf@bytheb.org> Mime-Version: 1.0 Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: 7bit Cc: Aaron Conole , "David S. Miller" , netdev@vger.kernel.org To: Tolga Ceylan Return-path: Received: from mail-pa0-f45.google.com ([209.85.220.45]:36627 "EHLO mail-pa0-f45.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1754732AbbI0CYR (ORCPT ); Sat, 26 Sep 2015 22:24:17 -0400 Received: by pablk4 with SMTP id lk4so44075862pab.3 for ; Sat, 26 Sep 2015 19:24:17 -0700 (PDT) In-Reply-To: Sender: netdev-owner@vger.kernel.org List-ID: On Sat, 2015-09-26 at 19:02 -0700, Tolga Ceylan wrote: > By keeping hiscore/matches as is, I'm trying to keep the hashing consistent. > Otherwise, this would behave similar to removing a listener which > drops connections. Right, this problem will soon disappear when listener rewrite is complete. Only SYN packet will have to select a listener. Then when ACK packet comes, the SYN_RECV will be found in ehash table, and req->rsk_listener will be used to get the listener that was chosen at SYN time.