Re: [vpp-dev] VIRL and SSE4.2

Great, SSE4.2 issue removed moving a step further. ... now I have a segfault issue to figure out :-) Many thanks !? From: Nicolás Leiva Sent: Thursday, January 26, 2017 12:03 AM To: Bernier, Daniel Cc: vpp-dev@lists.fd.io; csit-...@lists.fd.io Subject: Re: [

Re: [vpp-dev] VIRL and SSE4.2

Make sure you have host-passthrough configured in Nova settings, so the CPU visible to the guest is exactly the same as the host CPU even in the aspects that libvirt does not understand. virl@virl:~$ cat /etc/nova/nova.conf | grep cpu cpu_mode = host-passthrough virl@virl:~$ sudo service nova-com

[vpp-dev] VIRL and SSE4.2

Hi, Any particular tricks you have to run VPP in VIRL (latest release) ? Keep getting the "ERROR: This binary requires CPU with SSE4.2 extensions.?" Dan B ___ vpp-dev mailing list vpp-dev@lists.fd.io https://lists.fd.io/mailman/listinfo/vpp-dev

Re: [vpp-dev] VPP 17.01 on VFIO-PCI driver

Spent hours on it, finally root caused it. VPP works fine out of container with vfio-pci, while I was running inside of docker container… I am doing more investigation on it why it makes any difference in-and-out container, but appreciate your helps again if you happen to know any context on it

Re: [vpp-dev] VPP 17.01 on VFIO-PCI driver

> On 25 Jan 2017, at 18:59, Yichen Wang (yicwang) wrote: > > Here is the output, sorry for the long list, as I have a lot of Cisco vNICs, > and Intel NICs on my setup… I’ve highlighted the ones I am using for VPP. > > vpp# show pci > Address Socket VID:PID Link Speed Driver

Re: [vpp-dev] VPP 17.01 on VFIO-PCI driver

Here is the output, sorry for the long list, as I have a lot of Cisco vNICs, and Intel NICs on my setup… I’ve highlighted the ones I am using for VPP. vpp# show pci Address Socket VID:PID Link Speed Driver Product Name :01:00.0 08086:1521 5.0 GT/s x4igb

Re: [vpp-dev] VPP 17.01 on VFIO-PCI driver

> On 25 Jan 2017, at 18:03, Yichen Wang (yicwang) wrote: > > Yes, I did! Can you share output of “show pci” from the VPP debug cli? > > Regards, > Yichen > > On Jan 25, 2017, at 07:12, Damjan Marion > wrote: > >> >>> On 25 Jan 2017, at 05:41, Yichen Wang (

Re: [vpp-dev] VPP 17.01 on VFIO-PCI driver

Yes, I did! Regards, Yichen On Jan 25, 2017, at 07:12, Damjan Marion mailto:dmarion.li...@gmail.com>> wrote: On 25 Jan 2017, at 05:41, Yichen Wang (yicwang) mailto:yicw...@cisco.com>> wrote: Hi, VPP guys, I have a RHEL 7.3 setup with Intel X710, and want to bring VPP 17.01 on top of it. Amo

Re: [vpp-dev] SNAT Plugin Use

Dear Jon, Sorry for the delayed response. Anyhow, the constant VL_MSG_FIRST_AVAILABLE is a historical name for a simple thing: the number of binary API messages defined by a given plugin. I’m not sure what you’re trying to do, but it’s actually just as well that you ran into a compile error. I

Re: [vpp-dev] igb_uio -> uio_pci_generic

Looping in the deb_dpdk and rpm_dpdk guys to see if they have thoughts on how these things play out in deployment. Ed On Tue, Jan 24, 2017 at 9:05 PM, Burt Silverman wrote: > Hi Damjan, > > My understanding is that CONFIG_VFIO_NOIOMMU will never be set in a stock > kernel, and you will need to

Re: [vpp-dev] building rpm packages failure

I ran into some strange results on Fedora 25; first there was a problem but it disappeared: The problem occurred when I was on master. It seemed that the '~' character in the filenames like you have in your note was giving the rpm tools a bad time. So I switched to the 17.01 branch and the problem

Re: [vpp-dev] plugin infrastructure changes

Damjan, This looks good to me and I think it an improvement over the existing mechanism. Thanks, -daw- ps. I noticed a typo in the commit comment: s/will nor be/will not be/g On 1/24/17 12:10 PM, Damjan Marion (damarion) wrote: As discussed on the call, let’s continue discussion on the mail

Re: [vpp-dev] Native feature support in VPP without dpdk

Hi Damjan, Thanks for reply. Yes I can add support of VFIO in vpp. Ofcourse I will be needing help as well as I have just started looking into vpp. Having said that my priority will be to first add one of our non nic PF driver for our device to vpp with existing framework. After then I can work

Re: [vpp-dev] How are dpo nodes plugged in the graph?

Hi Nagp, It picks up those next-nodes because it is a sibling of ip4-lookup – by definition siblings share next-nodes. The VLIB code ensures this by tracking sibling relationships and adding a next node to all siblings. Regards, neale From: on behalf of Nagaprabhanjan Bellaru Date: Wednesda

[vpp-dev] building rpm packages failure

Building rpm packages on Fedora23 fails with the following error message: Processing files: vpp-debuginfo-17.04-rc0~144_ge056c8a.x86_64 Provides: vpp-debuginfo = 17.04-rc0~144_ge056c8a vpp-debuginfo(x86-64) = 17.04-rc0~144_ge056c8a Requires(rpmlib): rpmlib(CompressedFileNames) <= 3.0.4-1 rpmlib(F

[vpp-dev] How are dpo nodes plugged in the graph?

Hi, I am trying to understand how the nodes defined in the dpo files, for example "lookup-ip4-dst-itf" in lookup_dpo.c are linked to the graph in the data plane. For example, "show vlib graph" shows a number of next nodes for "lookup-ip4-dst-itf" like "ip4-local", "ip4-arp", "ip4-glean" etc., but

Re: [vpp-dev] Error reading from file descriptor 9: Input/output error

EAL throws this message to stderr, systemd collects it. There is no much of vpp involved here. On 25 Jan 2017, at 16:00, Kinsella, Ray mailto:ray.kinse...@intel.com>> wrote: Hi Damjan, I don't get how the the following message is being blurbed to the Syslog "PCI INTX mask not supported\n"

Re: [vpp-dev] VPP 17.01 on VFIO-PCI driver

> On 25 Jan 2017, at 05:41, Yichen Wang (yicwang) wrote: > > Hi, VPP guys, > I have a RHEL 7.3 setup with Intel X710, and want to bring VPP 17.01 on top > of it. Among all three DPDK drivers: > (1) uio_pci_generic is not supported on X710 > (http://dpdk.org/dev/patchwork/patch/19820/ >

Re: [vpp-dev] Error reading from file descriptor 9: Input/output error

Hi Damjan, I don't get how the the following message is being blurbed to the Syslog "PCI INTX mask not supported\n" Is causing the VPP log to run out of space and resulting in ... ? "file descriptor 9: Input/output error". On the patch ... http://dpdk.org/dev/patchwork/patch/11622/ Steve He

Re: [vpp-dev] Native feature support in VPP without dpdk

Hi, > On 25 Jan 2017, at 06:14, Saxena, Nitin wrote: > > Hi VPP developers, > > I am new to VPP and we want to add support of some Cavium NICs/non-NICs in > VPP. Our existing custom user space PF driver has VFIO NO IOMMU support to > get MSIX interrupts from device. It seems VFIO support in V

[vpp-dev] Native feature support in VPP without dpdk

Hi VPP developers, I am new to VPP and we want to add support of some Cavium NICs/non-NICs in VPP. Our existing custom user space PF driver has VFIO NO IOMMU support to get MSIX interrupts from device. It seems VFIO support in VPP can be used via dpdk only. So my question 1) Is there any plan

Re: [vpp-dev] Error reading from file descriptor 9: Input/output error

This is long standing issue in DPDK e1000 driver likely caused by bad e1000 emulation. Still, bad emulation is still not good excuse for lack of error message throttling. There is even patch submitted to fix this issue: http://dpdk.org/dev/patchwork/patch/11622/

Re: [vpp-dev] Error reading from file descriptor 9: Input/output error

Thank you, Dave, and Damjan, that’s clear enough for this question…We’ll plan to switch to Vmxnet3 then. - Andrew From: "Dave Barach (dbarach)" Date: Wednesday, 25 January 2017 at 9:37 PM To: "Andrew Li (zhaoxili)" , Damjan Marion Cc: "vpp-dev@lists.fd.io" Subject: RE: [vpp-dev] Error reading

Re: [vpp-dev] igb_uio -> uio_pci_generic

As Damjan wrote, the vpp graph dispatcher has known how to switch input nodes between interrupt and polling mode since before the dpdk even existed. Unfortunately, not all dpdk drivers support interrupt-mode processing, again, as discussed. As a clock-cycle / power-consumption mitigation maneu

Re: [vpp-dev] Error reading from file descriptor 9: Input/output error

Vmxnet3 interfaces are strongly preferred: for performance, as well as to make this specific issue disappear. We have a patch floating around which turns off the disk-filling syslog message, but if you simply switch to vmxnet3 interfaces the pain will go away without rebuilding images, etc. etc

Re: [vpp-dev] Error reading from file descriptor 9: Input/output error

Hi Damjan, Yes…and will it hurt anything other than printing this error message? If so we could move to VMXNET3. Thanks, Andrew From: Damjan Marion Date: Wednesday, 25 January 2017 at 6:27 PM To: "Andrew Li (zhaoxili)" Cc: "vpp-dev@lists.fd.io" Subject: Re: [vpp-dev] Error reading from file de

Re: [vpp-dev] Error reading from file descriptor 9: Input/output error

> On 25 Jan 2017, at 05:43, Andrew Li (zhaoxili) wrote: > > Hi vpp-dev, > > I’m encountering this strange issue: VPP keeps generating this error message > into /var/log/syslog: > > Jan 24 23:28:19 localhost vpp[4749]: EAL: Error reading from file descriptor > 9: Input/output error > Jan 24

Re: [vpp-dev] igb_uio -> uio_pci_generic

> On 25 Jan 2017, at 06:00, Stephen Hemminger > wrote: > > On Tue, 24 Jan 2017 23:05:55 -0500 > Burt Silverman wrote: > >> Hi Damjan, >> >> My understanding is that CONFIG_VFIO_NOIOMMU will never be set in a stock >> kernel, and you will need to build a custom kernel for that. I understand >

Re: [vpp-dev] SNAT Plugin Use

You should use SNAT plugin like all other plugins. Documentation for SNAT plugin is here https://wiki.fd.io/view/VPP/SNAT Regards, Matus Fabian Engineer - Software matfa...@cisco.com Tel: Cisco Systems, Inc. Slovakia cisco.com [http://www.cisco.com/assets/swa/

Re: [vpp-dev] ppp interface and route dpo

Hi Ewan, An interface class that is defined as point-2-point (P2P) will have an interface route that links to an adjacency and not to a glean. PPP interfaces are configured this way; VNET_HW_INTERFACE_CLASS (ppp_hw_interface_class) = { .name = "PPP", .flags = VNET_HW_INTERFACE_CLASS_FLAG_P2P,