From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mga04.intel.com (mga04.intel.com [192.55.52.120]) by dpdk.org (Postfix) with ESMTP id C2D1811DE for ; Fri, 17 Mar 2017 06:26:21 +0100 (CET) DKIM-Signature: v=1; a=rsa-sha256; c=simple/simple; d=intel.com; i=@intel.com; q=dns/txt; s=intel; t=1489728381; x=1521264381; h=date:from:to:cc:subject:message-id:references: mime-version:content-transfer-encoding:in-reply-to; bh=zWhMTKYCwfv6FFJ5LKHNPujJHEAOxNjCwCWjWnhsDaU=; b=BvG4o276LHug9vbcfxPXp7wD7L9dAa2np5g/K2hOiWJe0JowQzYNHaaT DUusEBfYYAQDhY0ZlvRtGrKrhA7w9w==; Received: from orsmga003.jf.intel.com ([10.7.209.27]) by fmsmga104.fm.intel.com with ESMTP/TLS/DHE-RSA-AES256-GCM-SHA384; 16 Mar 2017 22:26:20 -0700 X-ExtLoop1: 1 X-IronPort-AV: E=Sophos;i="5.36,175,1486454400"; d="scan'208";a="945328643" Received: from yliu-dev.sh.intel.com (HELO yliu-dev) ([10.239.67.162]) by orsmga003.jf.intel.com with ESMTP; 16 Mar 2017 22:26:18 -0700 Date: Fri, 17 Mar 2017 13:24:33 +0800 From: Yuanhan Liu To: Gopakumar Choorakkot Edakkunni Cc: dev@dpdk.org Message-ID: <20170317052433.GZ18844@yliu-dev.sh.intel.com> References: <20170317020611.GV18844@yliu-dev.sh.intel.com> <20170317043526.GW18844@yliu-dev.sh.intel.com> <20170317051343.GY18844@yliu-dev.sh.intel.com> MIME-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Disposition: inline Content-Transfer-Encoding: 8bit In-Reply-To: User-Agent: Mutt/1.5.23 (2014-03-12) Subject: Re: [dpdk-dev] virtio "how to restart applications" - //dpdk.org/doc/virtio-net-pmd X-BeenThere: dev@dpdk.org X-Mailman-Version: 2.1.15 Precedence: list List-Id: DPDK patches and discussions List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Fri, 17 Mar 2017 05:26:22 -0000 On Thu, Mar 16, 2017 at 10:20:30PM -0700, Gopakumar Choorakkot Edakkunni wrote: > >> When I was saying dpdk version, I meant the DPDK version with OVS. > > Oh I see! My apologies for the misuderstanding. The dpdk version used by host > ovs should be dpdk2.2, the guest process uses dpdk16.07. The OVS process is not > getting restarted, what is getting restarted is the guest process using > dpdk16.07 - so the above clarifications you had about virtio being > reset-before-opened on guest restart - does that still hold good or does that > need the HOST side dpdk to be 16.04 or above ? Yes, the HOST dpdk should be >= v16.04. --yliu > > >> And yes, the fixes are not included in the DPDK required for OVS 2.4. > > Thanks for the info. > > Rgds, > Gopa. > > On Thu, Mar 16, 2017 at 10:13 PM, Yuanhan Liu > wrote: > > On Thu, Mar 16, 2017 at 09:56:01PM -0700, Gopakumar Choorakkot Edakkunni > wrote: > > Hi Yuanhan, > > > > Thanks for the confirmation about not having to do anything special to > close > > the ports on dpdk going down or coming up. > > > > As for the question about if I met any issue of ovs getting stuck - yes, > my > > guest process runs dpdk 16.07 as I mentioned earlier - and if I kill my > guest > > process, then the host OVS-dpdk on the host reports stall ! The OVS-dpdk > and > > emu versions I use are as below. But maybe that is because of the ovs > missing > > the fixes you mentioned ? > > When I was saying dpdk version, I meant the DPDK version with OVS. > > > ~# ovs-vswitchd --version > > ovs-vswitchd (Open vSwitch) 2.4.1 > > And yes, the fixes are not included in the DPDK required for OVS 2.4. > >         --yliu > > > Compiled Nov 14 2016 06:53:31 > > # kvm --version > > QEMU emulator version 2.2.0, Copyright (c) 2003-2008 Fabrice Bellard > > ~# > > > > > > Rgds, > > Gopa. > > > > On Thu, Mar 16, 2017 at 9:35 PM, Yuanhan Liu > > > wrote: > > > >     On Thu, Mar 16, 2017 at 07:48:28PM -0700, Gopakumar Choorakkot > Edakkunni > >     wrote: > >     > Thanks a lot for the response Yuanhan. I am using dpdk v16.07. So > what > >     you are > >     > saying is that in 16.07, we dont really need to call > rte_eth_dev_close() > >     on > >     > exit, > > > >     It's not about "don't really need", it's more like "it's hard to". > Just > >     think that it may crash at any time. > > > >     > because dpdk will ensure that it will do virtio reset before init > when it > >     > comes up right ? > > > >     No, It just handles the abnormal case well when guest APP restarts. > > > >     > Regarding the vhost commits you mentioned - do we still need those > fixes > >     if we > >     > have the "virtio reset before init" mechanism ? > > > >     Yes, we still need them: just think some malicious guest may also > forge > >     data like that. > > > >     I'm a bit confused then. Have you actually met any issue (like got > stucked) > >     with DPDK v16.07? > > > >             --yliu > > > >     > Or that is a seperate problem > >     > altogether (and hence we would need those fixes) ? > >     > > >     > Rgds, > >     > Gopa. > >     > > >     > On Thu, Mar 16, 2017 at 7:06 PM, Yuanhan Liu < > yuanhan.liu@linux.intel.com > >     > > >     > wrote: > >     > > >     >     On Thu, Mar 16, 2017 at 12:39:16PM -0700, Gopakumar Choorakkot > >     Edakkunni > >     >     wrote: > >     >     > So the doc says we should call rte_eth_dev_close() *before* > going > >     down. > >     >     And I > >     >     > know that especially in dpdk-virtionet  in the guest + > ovs-dpdk in > >     the > >     >     host, > >     >     > the ovs ends up getting stalled/stuck (!!) if I dont close > the port > >     >     before > >     >     > starting() it when the guest dpdk process comes back up. > >     > > >     >     I'm assuming you were using an old version, something like dpdk > v2.2? > >     >     IIRC, DPDK v16.04 should have fixed your issue. > >     > > >     >     > Considering that this not done properly can screw up the HOST > ovs, > >     and I > >     >     want > >     >     > to do everything possible to avoid that, I want to be 200% > sure > >     that I > >     >     call > >     >     > close even if my process gets a kill -9 .. So obviously the > only > >     way of > >     >     doing > >     >     > that is to close the port when the dpdk process comes back up > and > >     >     *before* we > >     >     > init the port. rte_eth_dev_close() is not capable of doing > that as > >     it > >     >     expects > >     >     > the port parameters to be initialized etc.. before it can be > >     called. > >     > > >     >     We do virtio reset before init, which is basically what > >     rte_eth_dev_close() > >     >     mainly does. So I see no big issue here. > >     > > >     >     The stuck issue is due to hugepage reset by the guest DPDK > >     application, > >     >     leading all virtio vring elements being mem zeroed. The old > vhost > >     doesn't > >     >     handle it well, as a result, it got stuck. And here are some > relevant > >     >     commits: > >     > > >     >         a436f53 vhost: avoid dead loop chain > >     >         c687b0b vhost: check for ring descriptors overflow > >     >         623bc47 vhost: do sanity check for ring descriptor length > >     > > >     >             --yliu > >     > > >     >     > Any other > >     >     > suggestions on what can be done to close on restart rather > than > >     close on > >     >     going > >     >     > down ? Thought of bouncing this by the alias before I add a > version > >     of > >     >     close > >     >     > myself that can do this close-on-restart > >     > > >     > > > > > > >