From mboxrd@z Thu Jan 1 00:00:00 1970 From: "Qiu, Michael" Subject: Re: [PATCH] eal: add function to check if primary proc alive Date: Tue, 26 Jan 2016 11:07:22 +0000 Message-ID: <533710CFB86FA344BFBF2D6802E6028622F22C3C@SHSMSX101.ccr.corp.intel.com> References: <1453296322-1210-1-git-send-email-harry.van.haaren@intel.com> <533710CFB86FA344BFBF2D6802E6028622F2180B@SHSMSX101.ccr.corp.intel.com> <533710CFB86FA344BFBF2D6802E6028622F221CB@SHSMSX101.ccr.corp.intel.com> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: quoted-printable Cc: "dev@dpdk.org" To: "Van Haaren, Harry" , "david.marchand@6wind.com" Return-path: Received: from mga04.intel.com (mga04.intel.com [192.55.52.120]) by dpdk.org (Postfix) with ESMTP id B7EA49221 for ; Tue, 26 Jan 2016 12:07:26 +0100 (CET) Content-Language: en-US List-Id: patches and discussions about DPDK List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: dev-bounces@dpdk.org Sender: "dev" On 1/26/2016 5:04 PM, Van Haaren, Harry wrote:=0A= >> From: Qiu, Michael=0A= >> On 1/25/2016 7:51 PM, Van Haaren, Harry wrote:=0A= >>> Not really, the secondary process will need some CPU,=0A= >>> however it can sleep so it doesn't have to use 100% of it.=0A= >>> It shouldn't be run on a core that is used by the primary=0A= >>> for packet-forwarding though - that will impact performance.=0A= >> If not, what will happen if the primary been killed after you check=0A= >> alive? At that time, the secondary may be doing some work need primary= =0A= >> alive.=0A= > What work are you thinking of? Apart from the shared config=0A= > and hugepages, primary and secondary processes are running=0A= > in their own address-space, and if the primary gets killed,=0A= > the secondary will notice when it next polls rte_eal_primary_proc_alive()= .=0A= >=0A= > Whatever work the secondary was performing (in its own address space)=0A= > won't be directly changed by the primary being killed, because the=0A= > shared config and hugepages stay (EAL "cleans up" when the primary=0A= > is re-launched, not on quit).=0A= =0A= OK, when primary quit or be killed, the queues will be freed, it will=0A= be a potential issue when secondary try to access, maybe I'm wrong.=0A= =0A= Thanks,=0A= Michael=0A= =0A= > -Harry=0A= >=0A= >=0A= =0A=