[vpp-dev] Way to send api message from vpp to govpp

Hi, We want to push some monitoring information from vpp to govpp based on some condition. Please help me to know if there is any procedure through which we can meet this requirement as there is no request from agent in this case. Thanks, Alok -=-=-=-=-=-=-=-=-=-=-=- Links: You receive all

Re: [vpp-dev] Running multiple VPP processes causing VAPI API failures

Hi, Is there a way by which we can restrict this extra "vpp" daemon from starting, if the main vpp process is already running? Thanks and Regards, Chinmaya Agarwal. -=-=-=-=-=-=-=-=-=-=-=- Links: You receive all messages sent to this group. View/Reply Online (#22358):

Re: [vpp-dev] CNAT SNAT Not Working

Got the correct syntax now, I think the 22.06 or other similar version documentation (Cloud NAT section under SourceNATing outgoing traffic) needs to be updated?: *set cnat snat-policy * Thanks On Tue, Dec 20, 2022 at 7:47 PM Christopher Adigun via lists.fd.io wrote: > Hi, > > I am trying

[vpp-dev] Linux-cp Plugin Bird Routes Not Showing Up in VPP

Hi, I am currently trying to use linx-cp plugins to sync routes learned via BGP, bird is seeing the BGP route but I can't see the routes in VPP. Details: *vpp v22.10-release built by root on ff42e25458af at 2022-10-26T14:00:24* *startup.conf:* unix { nodaemon log /tmp/vpp.log

[vpp-dev] VPP AF_PACKET (aka host-interface) driver is now a plugin

Folks, As part of the ongoing effort to move features from vpp/src/vnet to plugins (tracked Jira VPP-2040 [0]), the gerrit change [1] moving the VPP AF_PACKET (aka host-interface) driver to a plugin has just been merged into master. If you use the VPP AF_PACKET driverand have the stanza

Re: [vpp-dev] VPP crashes with vlib_worker_thread_barrier_syn_int:thread deadlock

Hi Mechthild, Your issue is caused by the main thread waiting too long for a worker to finish. You may examine these lines in src/vlib/threads.h: 171 /* Check for a barrier sync request every 30ms */ 172 #define BARRIER_SYNC_DELAY (0.03) 173 174 #if CLIB_DEBUG > 0 175 /* long barrier

Re: [vpp-dev] VPP crashes with vlib_worker_thread_barrier_syn_int:thread deadlock

Hi, You can try adding the line "path /usr/lib/vpp_plugins/" at the top in plugins{} section in startup.conf file and do a VPP restart. That should fix your problem. Thanks, Chinmaya Agarwal -=-=-=-=-=-=-=-=-=-=-=- Links: You receive all messages sent to this group. View/Reply Online

[vpp-dev] VPP crashes with vlib_worker_thread_barrier_syn_int:thread deadlock

Hi, We have a severe problem with VPP – it’s cyclic restarting due to the following error: Dec 21 09:46:32 843V0N3 vpp[3846]: vlib_worker_thread_barrier_sync_int: worker thread deadlock This happens on both servers of the setup and it cannot recover. Can you please help us to debug this

[vpp-dev] CNAT SNAT Not Working

Hi, I am trying to use the CNAT plugin but each time I try to configure it I am getting error that it is unknown: *vpp# show versionvpp v22.10-release built by root on ff42e25458af at 2022-10-26T14:00:24vpp#vpp#vpp# cnat snatcnat: unknown input `snat'* I checked that it is part of the

Re: [vpp-dev] vpp+dpdk #dpdk

Have never seen two net devices, even SRIOV devices share same PCI address. Will you dump your device via your lspci or or /sys/… , PCI bus addresses are always unique? Personally I thought the PCI addresses in your screenshot are inaccurate. From: on behalf of "first_se...@163.com"

Re: [vpp-dev] mellanox mlx5 + rdma + lcpng + bond - performance (tuning ? or just FIB/RIB processing limit) (max performance pps about 2Mpps when packet drops starts)

HI Pawel, On Sat, Dec 17, 2022 at 6:28 PM Paweł Staszewski wrote: > Hi > > > So without bgp (lcp) and only one static route performance is basically as > it is on "paper" 24Mpps/100Gbit/s without problem. > > And then no matter if with bond or without bond (with lcp) there are > problems

Re: [vpp-dev] Error message on starting vpp

Hi Xiaodong, It seems like a '.runs_after' was introduced on a node that does not belong to the arc in question. [0] should solve this. (CCingJulian as he is the author of the original patch & sr plugin maintainers) Cheers -Nathan [0] https://gerrit.fd.io/r/c/vpp/+/37837 Le sam. 3 déc. 2022 à

Re: [vpp-dev] mellanox mlx5 + rdma + lcpng + bond - performance (tuning ? or just FIB/RIB processing limit) (max performance pps about 2Mpps when packet drops starts)

> Basically it looks like it is lcp problem and routing - dont know how this > tests are done for lcp but it looks like those tests are like > 1. load 900k routes > 2. connect to traffic generator - and push 10Mpps (and dont check if You > have any reply) or more from one single ip src to one

Re: [vpp-dev] vpp+dpdk

Can anyone answer my question? official person？ I use the pci called N10 -=-=-=-=-=-=-=-=-=-=-=- Links: You receive all messages sent to this group. View/Reply Online (#22346): https://lists.fd.io/g/vpp-dev/message/22346 Mute This Topic: https://lists.fd.io/mt/95640416/21656 Mute

Re: [vpp-dev] vpp+dpdk

Can anyone answer my question? official person？ -=-=-=-=-=-=-=-=-=-=-=- Links: You receive all messages sent to this group. View/Reply Online (#22345): https://lists.fd.io/g/vpp-dev/message/22345 Mute This Topic: https://lists.fd.io/mt/95640416/21656 Mute

Re: [vpp-dev] mellanox mlx5 + rdma + lcpng + bond - performance (tuning ? or just FIB/RIB processing limit) (max performance pps about 2Mpps when packet drops starts)

Switched to native linux kernel ip forwarding and Can now do 12Mpps with 12 cores at about 80Gbit/s Basically it looks like it is lcp problem and routing - dont know how this tests are done for lcp but it looks like those tests are like 1. load 900k routes 2. connect to traffic generator -

Re: [vpp-dev] mellanox mlx5 + rdma + lcpng + bond - performance (tuning ? or just FIB/RIB processing limit) (max performance pps about 2Mpps when packet drops starts)

Hi So without bgp (lcp) and only one static route performance is basically as it is on "paper" 24Mpps/100Gbit/s without problem. And then no matter if with bond or without bond (with lcp) there are problems starting. Basically side where im Receiving most traffic that need to be TX-ed to

Re: [vpp-dev] memory growth in charon using vpp_sswan

Thank you. Enabling DPD is not necessary to reproduce the issue. You can use `swanctl --list-sas` command to query SAs and see the problem. -=-=-=-=-=-=-=-=-=-=-=- Links: You receive all messages sent to this group. View/Reply Online (#22342): https://lists.fd.io/g/vpp-dev/message/22342 Mute

Re: [vpp-dev] mellanox mlx5 + rdma + lcpng + bond - performance (tuning ? or just FIB/RIB processing limit) (max performance pps about 2Mpps when packet drops starts)

Hi, > So the hardware is: > Intel 6246R > 96GB ram > Mellanox Connect-X 5 2x 100GB Ethernet NIC > And simple configuration with vpp/frr where one vlan interface all > traffix is RX-ed and second vlan interface where this traffic is TX-ed - > it is normal internet traffic - about 20Gbit/s with

Re: [vpp-dev] Procedure to send API from VPP to agent based on event

Hi Alok, There are a few examples of these sort of “want” APIs. For example, from src/vnet/interface.api: service { rpc want_interface_events returns want_interface_events_reply events sw_interface_event; }; /neale From: vpp-dev@lists.fd.io on behalf of Alok Tiwari via lists.fd.io

Re: [vpp-dev] ipv6 fragment problem

I suppose you should take a look at the updated RFC at https://datatracker.ietf.org/doc/html/rfc8200#section-4.5 Do you mind attaching your packet capture in this thread? It sounds to me that the destination address of the two fragments is your VPP instance running on the CentOS Linux (or you

Re: [vpp-dev] memory growth in charon using vpp_sswan

Hi Mahdi, We haven’t ever tried DPD features with sswan plugin, so I don’t have any info. I think we need to investigated and reproduce this memory growth issue in our end, I will keep you posted. regards Kai -=-=-=-=-=-=-=-=-=-=-=- Links: You receive all messages sent to this group.

Re: [vpp-dev] ipv6 fragment problem

Hi, Xiaodong IXIA test RFC 2460 4.5section https://www.rfc-editor.org/rfc/rfc2460#section-4.5 the software will send two packet, software configure MTU is 1500, and it will send two 768 size icmpv6 request fragment packet it expect receive two reply packet centos7 can receive two packet, but

[vpp-dev] Procedure to send API from VPP to agent based on event

Hi Folks, I understand that VPP handles two types of API exchanges 1. There is a simple "request and reply" type. 2. The second type of message exchange is a "dump" request and a repeated reply message in the form of multiple "details". In our case, I have a requirement to send

Re: [vpp-dev] ipv6 fragment problem

By setting MTU no less than 1280 does not mean the packet can’t be smaller than 1280. A packet smaller than MTU is supposed to be forwarded without any issue. If any packet is larger than the specified MTU, the DUT should reply with an ICMPv6 too big so that the sender could do fragmentation

[vpp-dev] ipv6 fragment problem

Hi, vpp experts Now i meet a problem that use IXIA software to test ipv6 support RFC2460 point out that all machine support ipv6 fragment must set MTU cannot less than 1280, but IXIA test ipv6 fragment will send two request frag packet that data size is 768 and expect receive two reply packet,

Re: [vpp-dev] vpp core & bound checking

Also, gcc once maintained "fat" pointer to store length of array for bound checking but removed later due to sanitizers such as valgrind. https://gcc.gnu.org/wiki/MIRO?action=AttachFile=get=MIRO.pdf There are bounded for loops being discussed in new programming languages for data plane. A

[vpp-dev] SCTP over host-interface not working

Hi , I have a problem with sctp over host-interface.I tried to create an SCTP session , but the session is not started . When I tried it with real interfaces (not host-interface) it worked correctly. Steps to reproduce issue : - create two veth pairs , one is moved to new namespace vnet0 .

Re: [vpp-dev] vpp core & bound checking

Hi Ben, What kind of bound failure is causing crashes? Is a for loop terminator exceeding bound or networking data exceeding bound? I can investigate changing the gcc compiler to check any bound at compile-time. Hemant -Original Message- From: vpp-dev@lists.fd.io On Behalf Of Benoit

Re: [vpp-dev] Running multiple VPP processes causing VAPI API failures

Hoi, Also worth noting, several sockets are created in /run/vpp/* notably the API socket /run/vpp/api.sock - the second invocation of "vpp" may have overwritten the socket with a new one, rendering the existing first invocation of "vpp" unusable. As Klement points out, you can run multiple

Re: [vpp-dev] Running multiple VPP processes causing VAPI API failures

Hi, if you have two vpp instances running with same shm prefix, conflicts are to be expected. It’s possible that the second instance mangled file system entries for the first one. When running more than one vpp instance, you need to ensure that every one of those has a unique shared memory

Re: [vpp-dev] possible use deleted sw if index in ip4-lookup and cause crash

By adding the following code right after process dispatch in the main loop, the crash is fixed. So I think the condition mentioned above is a rare but valid case. A ctrl process node being scheduled adds a packet (pending frame) to a node and the packet is referring to an interface which will be

Re: [vpp-dev] memory growth in charon using vpp_sswan

Hi Kai, Thanks for your response. Yes your understanding is correct and `stat_segment_connect` is called only when querying SAs. This query does not occur unless: 1) you ask for SA status and 2) before sending DPD messages. I'm afraid using 5.9.6 version did not change anything. May I ask if

Re: [vpp-dev] possible use deleted sw if index in ip4-lookup and cause crash

Hi list, During the test, when l3sub if is deleted, I got a new abort in interface drop node, seems the packet reference to a deleted interface. #0 __GI_raise (sig=sig@entry=6) at ../sysdeps/unix/sysv/linux/raise.c:50 > #1 0x7face8d17859 in __GI_abort () at abort.c:79 > #2

Re: [vpp-dev] memory growth in charon using vpp_sswan

Hi Mahdi, Thank you for report your discovery in vpp_sswan, unfortunately we haven't see this memory growth issue in our end. If my understanding is correct, the stat_segment_connect function should be called only if you want to see how many packets and bytes were processed by each SA, and the

[vpp-dev] memory growth in charon using vpp_sswan

Hi, I used the plugin resided in `extras/vpp_sswan` on both 5.9.8 and 5.8.2 Strongswan versions. All functionalities are working great but after Child SAs are established, there's a constant memory growth in charon process( Reaching from 20MB RSS to 46MB RSS in 4 days, but this growth is

[vpp-dev] vpp+dpdk #dpdk

I have the issue about when the same buf_info with different  name  of device like bottom picture,what should I do to bound two device called enp6s0f01d and enp6s0f02d. tks -=-=-=-=-=-=-=-=-=-=-=- Links: You receive all messages sent to this group. View/Reply Online (#22322):

[vpp-dev] VPP & HICN docs verify jobs now upload generated docs to a 7-day retention S3 bucket

Folks, The VPP and HICN docs verify jobs now upload the generated docs for each patch to a 7-day retention bucket in Amazon S3 storage. The upload URL for the docs is the same as the log URL for the docs job except that the domain name is 's3-docs-7day.fd.io' instead of 's3-logs.fd.io'. For

Re: [vpp-dev] [csit-dev] Core isolation on kernel command line deprecated

Hi Juraj, + vpp-dev@lists.fd.io For those curious to learn more, here is the cpusets kernel documentation [0] and cpuset manpage [1]. Thanks for the heads up! -daw- [0] https://www.kernel.org/doc/Documentation/admin-guide/cgroup-v1/cpusets.rst [1]

Re: [vpp-dev] RFC: move the “1”st instance of the community meeting 3 hours earlier to reduce impact for earlier time zone

OK, moving meeting tomorrow (and following 1st monthly meetings) for -3 hrs, new time is 5-6am PT. Wiki updated with: Bi-weekly on the second Tuesday of each month at 5-6am PT and fourth Tuesday of each month at 8-9am PT. — Damjan > On 07.12.2022., at 18:30, Damjan Marion via

[vpp-dev] Running multiple VPP processes causing VAPI API failures

Hi, We have a main VPP process running and we have IPIP tunnels configured in VPP which we try to fetch using VAPI API calls. It works fine in normal scenarios but when we by mistake executed the command "vpp" from linux command line, we saw that an additional VPP daemon started running and

Re: [vpp-dev] Q) Is it a possible acl-plugin race condtion?

Thank you for your answer. It's for fast datapath processing, I see. However, I'm curious whether it is an accepted behavior to stop routing while processing configuration change. Are production CISCO routers made in that way? Kind regards, Chul-Woong -=-=-=-=-=-=-=-=-=-=-=- Links: You receive

[vpp-dev] problem with ikev2 upd-encap when lcp is enabled

Hello guys, last night I faced a problem with ikev2 over natt when lcp is enabled. After tcpdump the traffic , I noticed there is a problem with the NONESP-encap response (using udp port 4500). can guys help me out please Thank you. -=-=-=-=-=-=-=-=-=-=-=- Links: You receive all messages sent

Re: [vpp-dev] Q) Is it a possible acl-plugin race condtion?

API and CLI are executed under the worker barrier by default in VPP, only the VPP main thread runs, and workers are stopped. For an API/CLI to be executed concurrently with workers, it must be explicitly marked as such (mp_safe attribute). It does not seem to be the case for ACL, so it should be

Re: [vpp-dev] Python VPP_API_DIR

Thanks Pim! I also wondered the other day where the env variable had disappeared to. 37792 merged. O. On 12 Dec 2022, at 00:44, Pim van Pelt via lists.fd.io wrote:Of course, the link is https://gerrit.fd.io/r/c/vpp/+/37791groet,PimOn Mon, Dec 12, 2022 at 12:41 AM Pim van Pelt via lists.fd.io

[vpp-dev] Q) Is it a possible acl-plugin race condtion?

Hi, team. I'm reading acl-plugin code and have a question on race condition. When vpp needs to update acl rules, it does not keep a refcnt or something and just `vec_free`s the rules. Is it safe?: ``` /* ACL updater: from acl_add_list() */ { a = am->acls + *acl_list_index; /* Get

Re: [vpp-dev] Python VPP_API_DIR

Of course, the link is https://gerrit.fd.io/r/c/vpp/+/37791 groet, Pim On Mon, Dec 12, 2022 at 12:41 AM Pim van Pelt via lists.fd.io wrote: > Hoi, > > I am starting to work on a few APIs and noticed that the vpp_papi module's > method find_api_dir promises that it will include VPP_API_DIR

[vpp-dev] Python VPP_API_DIR

Hoi, I am starting to work on a few APIs and noticed that the vpp_papi module's method find_api_dir promises that it will include VPP_API_DIR first, if specified in the environment. It doesn't actually do that. I wrote a changelist that prepends VPP_API_DIR if it exists. This is backwards

Re: [vpp-dev] VPP crashes on LX2160A platform

Hoi Pim, VPP runs on this board, but with some tricky things. 1. In the past, VPP had build-data/platform/dpaa.mk file , which is build parameters for DPDK for dpaa2-based system. It was removed from fdio/vpp repository few years ago. For cross-compiling VPP to that platfom, either for native

[vpp-dev] a problem with ikeV2 encapsulation when lcp is enabled

Hi , i have trouble with ikev2 vpn when the linux control plane is enabled : ##vpp1 13:43:37.221702 IP 141.94.64.84.500 > 10.7.129.204.500: isakmp: parent_sa ikev2_init[I] 13:43:37.244029 IP 10.7.129.204.500 >

Re: [vpp-dev] VPP crash while create ipip tunnel after a certain limit

To my knowledge there is no such ratio between heapsize and stackseg. based on your application needs you can tune these values. In your case as the number of ipip tunnels are more you may be required to increase stackseg size to accommodate the counters for those tunnel interfaces. Thanks and

Re: [vpp-dev] VPP crash while create ipip tunnel after a certain limit

On Wed, Dec 7, 2022 at 08:04 PM, Sudhir CR wrote: > > heapsize and statseg Thanks for your response. What should be the ideal ratio between heapsize and  statseg? Thanks and Regards, Chinmaya Agarwal. -=-=-=-=-=-=-=-=-=-=-=- Links: You receive all messages sent to this group. View/Reply

Re: [vpp-dev] Wireguard api - incorrect key lengh issue

Thank, You're the best, now it works fine. That message from the ChatGPT was the key. Best Regards Petr Documentation: Wireguard API "wireguard_interface_create" message with private/public key must be sent in right decoded values. Keys (string) generated from "wg keygen" command cannot be

Re: [vpp-dev] Wireguard api - incorrect key lengh issue

Hoi Petr, I think that the keys in the wireguard plugin are represented as bytestream, and your 44 character string probably refers to its base64 representation. Try sending the raw key (32 bytes). As an aside, I thought I'd ask ChatGPT this question, because I thought it was a fun puzzle. Pim:

[vpp-dev] Wireguard api - incorrect key lengh issue

Hi all, I have an issue with wireguard api message wireguard_interface_create. The public/private key is statically declared as "u8 public_key[32]", but the length of the public/private key is 44 (at this moment). Api wireguard_interface_dump is also impacted. Cli works fine - there is "u8

Re: [vpp-dev] VPP crash while create ipip tunnel after a certain limit

Hi Chinmaya Aggarwal, I can see the "vec_resize_allocate_memory" api in the above stack and you are telling that after 7k tunnels this issue is seen. I suspect this issue could be due to memory exhaust in the system. Can you please increase heapsize and statseg size in startup.conf file and

Re: [vpp-dev] FDIO Maintenance: 2022-12-07 18:00 UTC to 00:00 UTC

Jira maintenance is complete. Thank you, Vanessa On 12/7/22 4:01 PM, Vanessa Valderrama wrote: All standard maintenance is complete. Jenkins and Gerrit are available. We're still working on the Jira migration and will send an update when it's complete. If you have any issues, please open

[vpp-dev] VPP crash while create ipip tunnel after a certain limit

Hi, As per our use case, we need to have a large number of ipip tunnels in VPP (approx 1). When we try to configure that many tunnels inside VPP, after a certain limit it crashes with below core dump:- Dec 07 20:01:27 j3norvmstm01 vpp[2053130]: ipipCouldn't create

Re: [vpp-dev] FDIO Maintenance: 2022-12-07 18:00 UTC to 00:00 UTC

All standard maintenance is complete. Jenkins and Gerrit are available. We're still working on the Jira migration and will send an update when it's complete. If you have any issues, please open a ticket at support.linuxfoundation.org. Thank you, Vanessa On 12/7/22 10:59 AM, Vanessa

Re: [vpp-dev] RFC: move the “1”st instance of the community meeting 3 hours earlier to reduce impact for earlier time zone

Can we get a bit more votes on this? It is significant change so would like to confirm that everybody is ok with it…. — Damjan > On 24.11.2022., at 03:52, Marvin Liu wrote: > > +1, looking forward to join community meeting and have more discussions. > > From: vpp-dev@lists.fd.io

Re: [vpp-dev] FDIO Maintenance: 2022-12-07 18:00 UTC to 00:00 UTC

Jenkins has been placed in shutdown mode in preparation for maintenance On 11/29/22 2:58 PM, Vanessa Valderrama wrote: *What*: * Jenkins o Jenkins sandbox and production o OS and security updates o Jenkins upgrade o Plugin upgrades * Gerrit o OS and security

Re: [vpp-dev] possible use deleted sw if index in ip4-lookup and cause crash

The crash have not been found anymore. Does this fix make any sense? it it does, I will submit a patch later. Zhang Dongya via lists.fd.io 于 2022年11月29日周二 22:51写道： > Hi ben, > > In the beginning I also think it should be a barrier issue, however it > turned out not the case. > > The pkt which

Re: [vpp-dev] crash in vlib_get_next_frame_internal

It was pointed out by Dmitry that my fix doesn't make much sense. Any other ideas much appreciated. On Mon, 5 Dec 2022 at 10:53, Stanislav Zaikin via lists.fd.io wrote: > Hello folks, > > I've got a crash on v22.10 > vnet[2391390]: received signal SIGSEGV, PC 0x7f4847feb994, faulting > address

Re: [vpp-dev] Exclude a rx-queue from RSS #vpp_qos #dpdk

Adding Kai Ji to see if he could help. Thanks Kai's help in advance. On 12/6/2022 9:13 AM, ltham...@usc.edu wrote: Once I enable a flow on an interface to redirect certain packets to queue 0, I don't want other packets to use this queue 0. This can be done by disabling RSS for queue 0. I am

[vpp-dev] Unable to enable ICMP and src-ip match rte flows at once #dpdk #vpp_qos #vpp

Hi, When I enable an ICMP RTE flow on an interface, I am not able to enable another RTE flow with src-ip match on the same interface. It throws out an error saying "Failed to create flow". Below are the logs. Please let me know if I am doing anything wrong. When creating ICMP flow without

Re: [vpp-dev] Exclude a rx-queue from RSS #vpp_qos #dpdk

Hi Fan, Thanks for the info. I am aware of the test flow cli and the NICs on my device support RTE flows. Once I enable a flow on an interface to redirect certain packets to queue 0, I don't want other packets to use this queue 0. This can be done by disabling RSS for queue 0. I am looking

Re: [vpp-dev] bihash os_panic #vpp

Hi, I still have a doubt, why is it os_panic when the memory is not enough, instead of returning operation failure. -=-=-=-=-=-=-=-=-=-=-=- Links: You receive all messages sent to this group. View/Reply Online (#22290): https://lists.fd.io/g/vpp-dev/message/22290 Mute This Topic:

Re: [vpp-dev] bihash os_panic #vpp

Hi, I am using VPP bihash to implement the kernel conntrack session recording function, which involves frequent insertion and deletion. The memory starts to record from 4.xg to about 5.xg. VPP bihash os_panic, the system memory is 8g. OS Version: Ubuntu 18.04 -=-=-=-=-=-=-=-=-=-=-=- Links: You

Re: [vpp-dev] arm64 / clearfog + VPP

+Josua Mayer Hello again Josua, you'll remember we discussed Vector Packet Processing (VPP) on the NXP/DPAA board, and we've still been discussing this in the community off and on. I did not manage to get VPP to run on the NXP LX2160A Clearfog, but Xiaodong references a document from NXP that

Re: [vpp-dev] LACP issues w/ cdma/connectX 6

Type show lacp details to see if the member interface that is not forming the bundle receives and sends LACP PDUs. Type show hardware to see if both member interfaces have the same mac address. From: on behalf of Eyle Brinkhuis Reply-To: "vpp-dev@lists.fd.io" Date: Monday,

Re: [vpp-dev] arm64 / clearfog + VPP

The platform looks interesting indeed, guys, please share if it's possible to use ipsec offloading on it. On Mon, 5 Dec 2022 at 21:59, Xiaodong Xu wrote: > There is a guide to run VPP on dpaa2 board from NXP at >

Re: [vpp-dev] crash in vlib_get_next_frame_internal

Keeping you updated :) So, having the following piece of code in src/vlib/main.c, /* ??? Don't need valid flag: can use frame_index == ~0 */ if (PREDICT_FALSE (!(*nf->flags & VLIB_FRAME_IS_ALLOCATED*))) { nf->frame = vlib_frame_alloc (vm, node, next_index); nf->flags |=

Re: [vpp-dev] arm64 / clearfog + VPP

There is a guide to run VPP on dpaa2 board from NXP at https://docs.nxp.com/bundle/GUID-487B2E69-BB19-42CB-AC38-7EF18C0FE3AE/page/GUID-EBA9A5C6-2407-4AC7-87B6-C1470B2CD92D.html It requires a few scripts as well as some driver tools (restool, etc.) not sure they had any specific changes to VPP

Re: [vpp-dev] LACP issues w/ cdma/connectX 6

Hi Eyle, Once we faced LACP issue with host-if interfaces. In our topology host-if interfaces between two containers are connected via Linux bridge. when one container is sending LACP PDU to another container those PDU's are dropped by Linux. Please enable the packet trace on both sides and

Re: [vpp-dev] LACP issues w/ cdma/connectX 6

Hi, thanks for your reply. That’s the weird thing.. we have two identical hosts connected to the same switch (sn2700), with same OS, same VPP version, same minx_ofed and same everything except for the NIC. The box with a CX5 works like a charm, the box with the CX6 doesn’t.. but also does,

Re: [vpp-dev] LACP issues w/ cdma/connectX 6

I once experienced LACP not UP because the mode on vpp (lacp-static) did not match the mode on switch (lacp-dynamic). try changing the lacp mode on your switch. -=-=-=-=-=-=-=-=-=-=-=- Links: You receive all messages sent to this group. View/Reply Online (#22281):

Re: [vpp-dev] LACP issues w/ cdma/connectX 6

I'm no expert in LACP so maybe someone can jump in to confirm whether it looks good on VPP side, but as we're receiving LACP packets and VPP seems to be happy, I wonder if the issue could be VPP failing to send LACP packets back to the switch? Eg. because the NIC eat them up for some reason

[vpp-dev] LACP issues w/ cdma/connectX 6

Hi Ben, We have a few new boxes that have a connectX 6 fitted (Mellanox ConnectX-6 Dx 100GbE QSFP56 2-port PCIe 4 Ethernet Adapter) and we run into an issue with LACP (seems a popular topic these days.. :-)). We are not able to get LACP up, while running this: create int rdma host-if ens3f0

[vpp-dev] crash in vlib_get_next_frame_internal

Hello folks, I've got a crash on v22.10 vnet[2391390]: received signal SIGSEGV, PC 0x7f4847feb994, faulting address 0x0 vnet[2391390]: #0 0x7f484805e85b 0x7f484805e85b vnet[2391390]: #1 0x7f4847f753c0 0x7f4847f753c0 vnet[2391390]: #2 0x7f4847feb994 vlib_get_next_frame_internal +

Re: [vpp-dev] Exclude a rx-queue from RSS #dpdk #vpp #vpp_qos

Hi, vpp/flow_cli.c at master · FDio/vpp · GitHub contains some useful information to add a flow to redirect to queue X (check out test_flow() function's redirect_to_queue option, or vl_api_flow_enable_t_handler if you are

[vpp-dev] Exclude a rx-queue from RSS #dpdk #vpp #vpp_qos

Hi, I am setting num-rx-queues for a dpdk dev in vpp startup config to 4. This enables RSS and distribute packets coming on this interface between these 4 queues. I would like to configure RTE flow and redirect certain packets to queue 0. When this is done, I don't want other packets to use

[vpp-dev] Gerrit Review Id 37745

Hi community maintainers, I have added a fix as gerrit review id 37745. gerrit link: https://gerrit.fd.io/r/c/vpp/+/37745 short description: fix incorrect using about sw_if_index in nat44-ed static mapping v2 api. Please have a look and review. Best

[vpp-dev] Error message on starting vpp

Hi VPP experts, I got the following error message when starting vpp recently: 0: vnet_feature_arc_init:272: feature node 'ip6-lookup' not found (before 'pt', arc 'ip6-output') I'm using master branch from VPP git repo. By checking the source code, it seems it might have something to do with the

Re: [vpp-dev] PSA: host stack active opens from first worker

Florin, Most Excellent -- very nice work on improving the VPP Hoststack! :D Thanks, -daw- On 12/1/22 18:17, Florin Coras wrote: Hi folks, It’s been many months and patches at this point but once [1] is merged, session layer will accept connects from both main, with worker barrier, and

Re: [vpp-dev] Gerrit review for memif DMA acceleration

Damjan, External memory VFIO mapping action can moved to master thread through rpc call. From the usage of host stack, pre-allocate is not enough for session segments as mapped size may vary and these segments will dynamically allocated and freed when session created and destroyed. Regards,

Re: [E] COMMERCIAL BULK: Re: [vpp-dev] Issues with failsafe dpdk pmd in Azure

Hi All, After I run below two cmds manually, ping traffic recovered , but why tc mirred not working after sometime tc qdisc add dev eth1 handle : ingress tc filter add dev eth1 parent : u32 match u32 0 0 action mirred egress redirect dev dtap0 Best Regards, Kevin From:

Re: [E] COMMERCIAL BULK: Re: [vpp-dev] Issues with failsafe dpdk pmd in Azure

Hi Peter, Stephen and Long, I am facing some issues when running VPP on Azure VM, can you please help and give some suggestion if possible. I'm running CentOS 7.9 with kernel version 3.10.0 on Azure VM, VPP version is 20.09 and DPDK version is 20.11, below is the

[vpp-dev] PSA: host stack active opens from first worker

Hi folks, It’s been many months and patches at this point but once [1] is merged, session layer will accept connects from both main, with worker barrier, and first worker. Preference is now for the latter, especially if CPS performance is critical. There should be no need to change existing

[vpp-dev] RFC: VPP stat segment access from Rust

Folks, I made a first rough cut of stat segment bindings for Rust - so wanted to throw it in here and see if there is any interest/solicit feedback: https://github.com/ayourtch/vpp-stats-client It has two examples - vpp-get-stats and vpp_prometheus_exporter, which copy the functionality of

Re: [vpp-dev] VPP crashes on LX2160A platform

Hoi, VPP does not run on that board, because the soc does not enumerate its DPDK enabled interfaces on the PCIe bus, but rather has a custom bus, which VPP is not integrated with. Incidentally, I did get VPP itself to run but without DPDK (or AVF or etc) interfaces; and its basic functionality

Re: [SUSPECTED SPAM] [vpp-dev] VPP crashes on LX2160A platform

Hi, I think the 1st thing to try would be whether you can reproduce it with a debug build. The backtrace you have is not usable unfortunately. The easiest should be to build debug binaries: ~# make rebuild ~# ./build-root/install-vpp_debug-native/vpp/bin/vpp -c /etc/vpp/startup.conf Best Ben >

[vpp-dev] VPP crashes on LX2160A platform

Dear VPP community, I'm trying to operate VPP on SolidRun LX2160 board, which is based on 16 cores A72 NXP SoC, unfortunately, with little success. Does anybody have any experience with running VPP on such boards? The performance in my tests is quite good (more then 4mpps NDR) , but VPP works

Re: [vpp-dev] Gerrit review for memif DMA acceleration

Sure, I think it is possible as only few batches needed for typical workload. For dynamically mapping extended memory, I think this function is still needed as new segment will be allocated from system memory when new stream established. This action is happened in work thread. >

Re: [vpp-dev] Review request for Generic Flow in Native AVF

CSIT check issue is resolved. From: Xu, Ting Sent: Tuesday, November 29, 2022 8:23 PM To: vpp-dev Cc: Damjan Marion Subject: Review request for Generic Flow in Native AVF Hi, Damjan and community, I have uploaded a patch for enabling generic flow in native AVF days ago. To support generic

Re: [vpp-dev] vpp core & bound checking

Agree with Ben. There is also gdb to use to debug a crash. Hemant -Original Message- From: vpp-dev@lists.fd.io On Behalf Of Benoit Ganne (bganne) via lists.fd.io Sent: Wednesday, November 30, 2022 9:30 AM To: vpp-dev@lists.fd.io Subject: [vpp-dev] vpp core & bound checking Hi

[vpp-dev] vpp core & bound checking

Hi everyone, I'd like to get the discussion started on the topic of bound checking in VPP: some of us recently discussed a patch that added explicit bound checking within VPP core dataplane infrastructure to prevent a crash when the function is misused from a plugin. The bug is clearly in the

Re: [vpp-dev] Gerrit review for memif DMA acceleration

Hi Damjan, VFIO map function now can be called from work thread in some cases. Like allocate physical memory for DMA batch and then do mapping for if no attached device, or a new session segment attached to VPP process. So I use event-logger for vfio logging. These numbers are collected from

[vpp-dev] FDIO Maintenance: 2022-12-07 18:00 UTC to 00:00 UTC

*What*: * Jenkins o Jenkins sandbox and production o OS and security updates o Jenkins upgrade o Plugin upgrades * Gerrit o OS and security updates *     JIRA o OS and security updates o JIRA upgrade o JIRA migration *     Ingress o OS and

Re: [vpp-dev] possible use deleted sw if index in ip4-lookup and cause crash

Hi ben, In the beginning I also think it should be a barrier issue, however it turned out not the case. The pkt which had sw_if_index[VLIB_RX] set as the to-be-deleted interface is actually being put to ip4-lookup node by my process node, the process node add pkt in a timer drive way. Since the

Re: [vpp-dev] possible use deleted sw if index in ip4-lookup and cause crash

Hi Zhang, I'd expect the interface deletion to happen under the worker barrier. VPP workers should drain all their in-flight packets before entering the barrier, so it should not be possible for the interface to disappear between your node and ip4-lookup. Or am I missing something? What I have

<    1   2   3   4   5   6   7   8   9   10   >