From mboxrd@z Thu Jan 1 00:00:00 1970 From: Alexei Starovoitov Subject: Re: [PATCH net-next] net: use jump label patching for ingress qdisc in __netif_receive_skb_core Date: Fri, 10 Apr 2015 15:40:00 -0700 Message-ID: <55285140.9030902@plumgrid.com> References: <66b2c2faaa0f629dd9a8c0981bf51f6484bc4ee9.1428699635.git.daniel@iogearbox.net> <1428705014.25985.331.camel@edumazet-glaptop2.roam.corp.google.com> Mime-Version: 1.0 Content-Type: text/plain; charset=utf-8; format=flowed Content-Transfer-Encoding: 7bit Cc: davem@davemloft.net, jhs@mojatatu.com, edumazet@google.com, netdev@vger.kernel.org To: Eric Dumazet , Daniel Borkmann Return-path: Received: from mail-pd0-f175.google.com ([209.85.192.175]:36691 "EHLO mail-pd0-f175.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1755125AbbDJWkC (ORCPT ); Fri, 10 Apr 2015 18:40:02 -0400 Received: by pdea3 with SMTP id a3so36096811pde.3 for ; Fri, 10 Apr 2015 15:40:01 -0700 (PDT) In-Reply-To: <1428705014.25985.331.camel@edumazet-glaptop2.roam.corp.google.com> Sender: netdev-owner@vger.kernel.org List-ID: On 4/10/15 3:30 PM, Eric Dumazet wrote: > On Fri, 2015-04-10 at 23:07 +0200, Daniel Borkmann wrote: >> Even if we make use of classifier and actions from the egress >> path, we're going into handle_ing() executing additional code >> on a per-packet cost for ingress qdisc, just to realize that >> nothing is attached on ingress. >> >> Instead, this can just be blinded out as a no-op entirely with >> the use of a static key. On input fast-path, we already make >> use of static keys in various places, e.g. skb time stamping, >> in RPS, etc. It makes sense to not waste time when we're assured >> that no ingress qdisc is attached anywhere. >> >> Enabling/disabling of that code path is being done via two >> helpers, namely net_{inc,dec}_ingress_queue(), that are being >> invoked under RTNL mutex when a ingress qdisc is being either >> initialized or destructed. >> >> Signed-off-by: Daniel Borkmann > > My concern about jump labels is they add a conditional on arches where > CONFIG_JUMP_LABEL=n > > They look great on x86, but not sure about say MIPS. mips has HAVE_ARCH_JUMP_LABEL, so it should be fast there. I think all major archs are covered: arch/arm/Kconfig: select HAVE_ARCH_JUMP_LABEL if !XIP_KERNEL arch/arm64/Kconfig: select HAVE_ARCH_JUMP_LABEL arch/mips/Kconfig: select HAVE_ARCH_JUMP_LABEL arch/powerpc/Kconfig: select HAVE_ARCH_JUMP_LABEL arch/s390/Kconfig: select HAVE_ARCH_JUMP_LABEL if !MARCH_G5 arch/sparc/Kconfig: select HAVE_ARCH_JUMP_LABEL if SPARC64 arch/x86/Kconfig: select HAVE_ARCH_JUMP_LABEL