Dear Damjan, Is the 30% faster because of using different applications or something else?
Thank you, Tina On May 23, 2018, at 7:56 AM, Damjan Marion <dmarion.li...@gmail.com<mailto:dmarion.li...@gmail.com>> wrote: I addef -2 on it as there is no value in having it in vpp. We already have native driver which is 30% faster. At the end dpdk driver is just a wrapper around musdk.... — Damjan On 22 May 2018, at 20:05, Szymon Śliwa <s...@semihalf.com<mailto:s...@semihalf.com>> wrote: Hi, I improved the patch Natalie submitted, and pushed it for review on gerrit: https://gerrit.fd.io/r/#/c/12694/ . There was a small change in dpdk/device/init.c - I needed to set ignore_offload_bitfield manually, don't know what changed since I last tested it, but it did not work without it. I also added stuff for mrvl_pmd in dpdk/Makefile. And fixed the hard coded linker flag to some kind of autotools magic (hope I done it right). Everything was tested on VPP from today morning (not IPsec though). Regards, Szymon 2018-05-17 21:07 GMT+02:00 Damjan Marion <dmarion.li...@gmail.com<mailto:dmarion.li...@gmail.com>>: Fine if this is going to stay out of vpp repo, as it is duplicate functionality and also it will make dpdk plugin more messy due to external dependencies on musdk. Will be better to do that work with some NIC which have native DPDK drivers, not wrapper like in musdk case... -- Damjan On 17 May 2018, at 14:09, Nitin Saxena <nitin.sax...@cavium.com<mailto:nitin.sax...@cavium.com>> wrote: Hi Damjan, This request was discussed in FD.io<http://FD.io> ARM call in order to optimize dpdk-input node for ARM platforms. Macchiatobin is one common platform where almost all ARM stakeholders are working on and profiling data/details can be easily shared among them. Thanks, Nitin On Wednesday 16 May 2018 11:37 PM, Damjan Marion wrote: Hi Natalie, Who requested that? Thanks, -- Damjan On 16 May 2018, at 15:33, Natalie Samsonov <nsams...@marvell.com<mailto:nsams...@marvell.com> <mailto:nsams...@marvell.com>> wrote: Dear Damjan, As I understood, it was a request to use DPDK plugin with all platforms and I was asked to upstream Marvell DPDK plugin. Best Regards, Natalie *From:*Damjan Marion [mailto:dmarion.li...@gmail.com] *Sent:*Wednesday, May 16, 2018 16:03 *To:*Natalie Samsonov <nsams...@marvell.com<mailto:nsams...@marvell.com> <mailto:nsams...@marvell.com>> *Cc:*Tina Tsou <tina.t...@arm.com<mailto:tina.t...@arm.com> <mailto:tina.t...@arm.com>>;vpp-dev@lists.fd.io<mailto:vpp-dev@lists.fd.io> <mailto:vpp-dev@lists.fd.io>; Maen Suleiman <m...@marvell.com<mailto:m...@marvell.com> <mailto:m...@marvell.com>> *Subject:*[EXT] Re: [vpp-dev] Marvell's AI from last call External Email ------------------------------------------------------------------------ Dear Natalie, What is the point of using DPDK marvel driver (which is actually just a wrapper around musdk) if we already have native musdk support in vpp? -- Damjan On 15 May 2018, at 08:48, Natalie Samsonov <nsams...@marvell.com<mailto:nsams...@marvell.com> <mailto:nsams...@marvell.com>> wrote: Hi, 1.We are working now on upstreaming Marvell’s PMD support in VPP DPDK plugin. I’m attaching the patch as well in case you want play with it. 2.Regarding the L3FWD performance difference between MUSDK and DPDK, currently we don’t have numbers with VPP L3FW. We only have the numbers with two different applications. One is our own L3FW application running on top of MUSDK and other is a standard DPDL L3FWD application, so it’s not a fair comparison. Best Regards, Natalie Samsonov <0001-plugins-dpdk-Add-support-for-net_mrvl-dpdk-driver.patch> IMPORTANT NOTICE: The contents of this email and any attachments are confidential and may also be privileged. If you are not the intended recipient, please notify the sender immediately and do not disclose the contents to any other person, use it for any purpose, or store or copy the information in any medium. Thank you.
