From mboxrd@z Thu Jan 1 00:00:00 1970 From: "Zhou, Danny" Subject: Re: [PATCH v2] librte_pmd_packet: add PMD for AF_PACKET-based virtual devices Date: Tue, 15 Jul 2014 15:40:56 +0000 Message-ID: References: <1405024369-30058-1-git-send-email-linville@tuxdriver.com> <1405362290-6753-1-git-send-email-linville@tuxdriver.com> <20140715121743.GA14273@localhost.localdomain> <20140715140111.GA26012@tuxdriver.com> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: quoted-printable Cc: "dev-VfR2kkLFssw@public.gmane.org" To: "John W. Linville" , Neil Horman Return-path: In-Reply-To: <20140715140111.GA26012-2XuSBdqkA4R54TAoqtyWWQ@public.gmane.org> 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-VfR2kkLFssw@public.gmane.org Sender: "dev" > -----Original Message----- > From: John W. Linville [mailto:linville-2XuSBdqkA4R54TAoqtyWWQ@public.gmane.org] > Sent: Tuesday, July 15, 2014 10:01 PM > To: Neil Horman > Cc: Zhou, Danny; dev-VfR2kkLFssw@public.gmane.org > Subject: Re: [dpdk-dev] [PATCH v2] librte_pmd_packet: add PMD for > AF_PACKET-based virtual devices >=20 > On Tue, Jul 15, 2014 at 08:17:44AM -0400, Neil Horman wrote: > > On Tue, Jul 15, 2014 at 12:15:49AM +0000, Zhou, Danny wrote: > > > According to my performance measurement results for 64B small > > > packet, 1 queue perf. is better than 16 queues (1.35M pps vs. 0.93M > > > pps) which make sense to me as for 16 queues case more CPU cycles > > > (16 queues' 87% vs. 1 queue' 80%) in kernel land needed for > > > NAPI-enabled ixgbe driver to switch between polling and interrupt > > > modes in order to service per-queue rx interrupts, so more context > > > switch overhead involved. Also, since the > > > eth_packet_rx/eth_packet_tx routines involves in two memory copies > > > between DPDK mbuf and pbuf for each packet, it can hardly achieve > > > high performance unless packet are directly DMA to mbuf which needs i= xgbe > driver to support. > > > > I thought 16 queues would be spread out between as many cpus as you > > had though, obviating the need for context switches, no? >=20 > I think Danny is testing the single CPU case. Having more queues than CP= Us > probably does not provide any benefit. >=20 > It would be cool to hack the DPDK memory management to work directly out = of the > mmap'ed AF_PACKET buffers. But at this point I don't have enough knowled= ge of > DPDK internals to know if that is at all reasonable... >=20 > John >=20 > P.S. Danny, have you run any performance tests on the PCAP driver? No, I do not have PCAP driver performance results in hand. But I remember i= t is less than 1M pps for 64B. >=20 > -- > John W. Linville Someday the world will need a hero, and you > linville-2XuSBdqkA4R54TAoqtyWWQ@public.gmane.org might be all we have. Be ready.