From mboxrd@z Thu Jan 1 00:00:00 1970 From: Eric Dumazet Subject: Re: Optimizing instruction-cache, more packets at each stage Date: Mon, 18 Jan 2016 08:53:13 -0800 Message-ID: <1453135993.1223.230.camel@edumazet-glaptop2.roam.corp.google.com> References: <20160115142223.1e92be75@redhat.com> <20160115.154721.458450438918273509.davem@davemloft.net> <20160118112703.6eac71ca@redhat.com> Mime-Version: 1.0 Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: 7bit Cc: David Miller , netdev@vger.kernel.org, alexander.duyck@gmail.com, alexei.starovoitov@gmail.com, borkmann@iogearbox.net, marek@cloudflare.com, hannes@stressinduktion.org, fw@strlen.de, pabeni@redhat.com, john.r.fastabend@intel.com To: Jesper Dangaard Brouer Return-path: Received: from mail-pa0-f49.google.com ([209.85.220.49]:34469 "EHLO mail-pa0-f49.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1755750AbcARQxQ (ORCPT ); Mon, 18 Jan 2016 11:53:16 -0500 Received: by mail-pa0-f49.google.com with SMTP id uo6so422370996pac.1 for ; Mon, 18 Jan 2016 08:53:16 -0800 (PST) In-Reply-To: <20160118112703.6eac71ca@redhat.com> Sender: netdev-owner@vger.kernel.org List-ID: On Mon, 2016-01-18 at 11:27 +0100, Jesper Dangaard Brouer wrote: > The NAPI drivers actually already have a flush API (calling > napi_complete_done()), BUT it does not always get invoked, e.g. if the > driver have more work to do, and want to keep polling. > I'm not sure we want to delay "flushing" packets queued in the GRO > layer for this long(?). Since linux-3.7 we have this logic : http://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git/commit/?id=2e71a6f8084e7ac87166dd77d99c44190fb844fc (Some arches have quite expensive high resolution timestamps)