Re: [vpp-dev] router_plugin

[Wang]

HI Gulakh,

>From my experience, VPP and VPPSB are compatible with the latest versions.
What sequence of command lines did you use?

2018-03-04 4:58 GMT-05:00 Gulakh :

> Hi,
> I am using vppsb router plugin with vpp. there is a problem when I use
> tap-inject.
> The problem is that after I enable tap-inject and set interfaces' and
> taps' IP address, I want to make taps' state up but at this point vpp
> crashes and everything rolls back to its first state (No tap, No IP
> address, ...)
>
> Q1) What is the problem??
>
> Q2) I have heard that VPP and VPPSB are not compatible in some versions
> ... To which versions Should checkout both source code so that they can
> work correctly??
>
> Thanks in advance
> 
>
>

Re: [vpp-dev] vppsb's librtnl is broken with top of VPP tree

[Wang]

I just submitted a patch to fix it.

https://gerrit.fd.io/r/#/c/11127/

2018-03-13 18:35 GMT-04:00 Jon Loeliger :

> Hey VPP-dev
>
> Is it me?  It's me, isn't it.
>
> jdl
>
>
>
>
> librtnl/mapper.c: In function 'mapper_add_del_route':
> librtnl/mapper.c:101:31: error: passing argument 10 of
> 'fib_table_entry_path_add' from incompatible pointer type [-Werror]
>FIB_ROUTE_PATH_FLAG_NONE);
>^
> In file included from /work/vpp/build-root/rpmbuild/
> vpp-18.04.0/build-root/install-vpp-native/vpp/
> include/vnet/fib/fib.h:647:0,
>  from librtnl/mapper.c:25:
> work/vpp/build-root/rpmbuild/vpp-18.04.0/build-root/
> install-vpp-native/vpp/include/vnet/fib/fib_table.h:332:25: note:
> expected 'struct fib_mpls_label_t *' but argument is of type 'mpls_label_t
> *'
>  extern fib_node_index_t fib_table_entry_path_add(u32 fib_index,
>  ^
> librtnl/mapper.c:132:31: error: passing argument 10 of
> 'fib_table_entry_path_add' from incompatible pointer type [-Werror]
>FIB_ROUTE_PATH_FLAG_NONE);
>^
> In file included from work/vpp/build-root/rpmbuild/vpp-18.04.0/build-root/
> install-vpp-native/vpp/include/vnet/fib/fib.h:647:0,
>  from librtnl/mapper.c:25:
> work/vpp/build-root/rpmbuild/vpp-18.04.0/build-root/
> install-vpp-native/vpp/include/vnet/fib/fib_table.h:332:25: note:
> expected 'struct fib_mpls_label_t *' but argument is of type 'mpls_label_t
> *'
>  extern fib_node_index_t fib_table_entry_path_add(u32 fib_index,
>  ^
> cc1: all warnings being treated as errors
> make[3]: *** [librtnl/mapper.lo] Error 1
> make[3]: Leaving directory `/vppsb/netlink/build-root/
> rpmbuild/BUILD/netlink-1.0.1-22~gd8e6dbc-dirty'
> error: Bad exit status from /var/tmp/rpm-tmp.IlCmHt (%build)
>
> 
>
>

Re: [vpp-dev] Issue in installing router and netlink plugins

[Wang]

Did you checkout the latest vppsb? There should be a patch merged to fix
this issue.

2018-03-30 3:25 GMT-04:00 Hamid Rasool :

> Hi,
>
> I want to enable dynamic routing protocols (using Quagga or FRR as VPP
> does not have BGP and IS-IS built-in) to perform routing on my VPP VM
> (built using stable/1801 branch) using KVM hypervisor. I cannot enable
> tap-inject on the device. After linking the folders to vppsb/netlink,
> vppsb/router, vppsb/netlink.mk and vppsb/router.mk in my vpp build
> directory, moved to build-root directory and then ran ./bootstrap.sh.
> However I am getting errors in the
>
> make V=0 PLATFORM=vpp TAG=vpp_debug install-deb netlink-install router-install
>
> command. Attached is the output of the failed build.
>
> Has anyone been successful in enabling tap-inject module in the 18.01 or
> master version builds without vagrant? The README of vppsb indicated that
> it was compliant with 17.04.
>
> Regards,
> Hamid.
>
>
> 
>  Virus-free.
> www.avast.com
> 
> <#m_6376853032253994508_DAB4FAD8-2DD7-40BB-A1B8-4E2AA1F9FDF2>
> 
>
>

Re: [vpp-dev] Issue in installing router and netlink plugins

[Wang]

Here is the patch

https://gerrit.fd.io/r/#/c/11167/

it should be merged into vppsb.

Chad

2018-04-01 12:39 GMT-04:00 Hamid via Lists.Fd.Io <
14mseesrasool=seecs.edu...@lists.fd.io>:

> Not working with the latest code.
>
> I have retried with 'git checkout origin/master' after make wipe-all and
> make vpp-wipe commands but I am still getting the same error with librtnl
> files.
> 
>
>

Re: [vpp-dev] Issue in installing router and netlink plugins

[Wang]

Would you share the command lines you used to install router plugin? It
seems to me the router plugin is not loaded?

2018-04-02 13:05 GMT-04:00 Hamid Rasool <14mseesras...@seecs.edu.pk>:

> Hi Wang,
>
> My copy of the code contains the patch. I have checked the source and it
> is according to the diff in your commit
> <https://gerrit.fd.io/r/gitweb?p=vppsb.git;a=blobdiff;f=netlink/librtnl/mapper.c;h=05b7f1df43c2dae6ad193b8cb6abb6f70ae1673d;hp=65cc13a976e459f4e84d23fb8e1f8d11b9e0b543;hb=ce5ab87645e7472b99e23686047e4141a7a315b0;hpb=042a782c530ceb58507d697726e68ad7fe4fe023>.
> The make netlink-install and router-install now build without errors (using
> vpp 18.04-rc0 ; vpp stable/18.01 had the above issue).
>
> The problem I am facing that after netlink and router are built, I am
> unable to perform 'sudo vppctl enable tap-inject' command. Does the build
> command supposed to have .deb packages for vppsb. This is my build-root
> directory:
> xflow@vpphamid1:~/install1/vpp/build-root$ ls
> autowank  platforms.mk
> bin   python
> build-config.mk   sbin
> build-config.mk.READMEscripts
> build-tool-native tools
> build-vpp_debug-nativevagrant
> build-vpp-native  vpp_18.04-rc0~503-gf0404e9_amd64.changes
> config.site   vpp_18.04-rc0~503-gf0404e9_amd64.deb
> copyimg   vpp-api-java_18.04-rc0~503-gf0404e9_amd64.deb
> deb   vpp-api-lua_18.04-rc0~503-gf0404e9_amd64.deb
> install-vpp_debug-native  vpp-api-python_18.04-rc0~503-gf0404e9_amd64.deb
> install-vpp-nativevpp-dbg_18.04-rc0~503-gf0404e9_amd64.deb
> lib64 vpp-dev_18.04-rc0~503-gf0404e9_amd64.deb
> Makefile  vpp-lib_18.04-rc0~503-gf0404e9_amd64.deb
> path_setupvpp-plugins_18.04-rc0~503-gf0404e9_amd64.deb
>
>
>
>
> <https://www.avast.com/sig-email?utm_medium=email&utm_source=link&utm_campaign=sig-email&utm_content=webmail&utm_term=icon>
>  Virus-free.
> www.avast.com
> <https://www.avast.com/sig-email?utm_medium=email&utm_source=link&utm_campaign=sig-email&utm_content=webmail&utm_term=link>
> <#m_-6449746801624537005_DAB4FAD8-2DD7-40BB-A1B8-4E2AA1F9FDF2>
>
> On Mon, Apr 2, 2018 at 6:14 PM, Wang  wrote:
>
>> Here is the patch
>>
>> https://gerrit.fd.io/r/#/c/11167/
>>
>> it should be merged into vppsb.
>>
>> Chad
>>
>> 2018-04-01 12:39 GMT-04:00 Hamid via Lists.Fd.Io <
>> 14mseesrasool=seecs.edu...@lists.fd.io>:
>>
>>> Not working with the latest code.
>>>
>>> I have retried with 'git checkout origin/master' after make wipe-all and
>>> make vpp-wipe commands but I am still getting the same error with librtnl
>>> files.
>>> 
>>>
>>>
>>
>

Re: [vpp-dev] Issue in installing router and netlink plugins

[Wang]

 sudo ln -sf /root/xflow/install1/vpp/build-root/install-vpp_debug-
native/router/lib64/router.so.0.0.0 /usr/lib/vpp_plugins/router.so

can you copy instead of ln?

2018-04-03 14:16 GMT-04:00 Hamid Rasool <14mseesras...@seecs.edu.pk>:

> I am following the FRR tutorial
> <https://github.com/FRRouting/frr/wiki/Alternate-forwarding-planes%3A-VPP>
> to set up vppsb. Please let me know if there is better documentation
> available.
>
> I had already built and ran stable/1801 branch vpp code on my machine but
> due to conflicts with vppsb I had to check out master for vpp and vppsb
> repos.
>
> Here the steps,
>
> In the home directory, git clone https://gerrit.fd.io/r/vppsb
> cd vppsb
> git checkout master
>
> Then, in vpp directory,
> ln -sf /home/xflow/vppsb/netlink
> ln -sf /home/xflow/vppsb/router
> ln -sf /home/xflow/netlink/netlink.mk build-data/packages/
> ln -sf /home/xflow/router/router.mk build-data/packages/
>
> cd build-root
> sudo make V=0 PLATFORM=vpp TAG=vpp_debug install-deb netlink-install
> router-install
> sudo dpkg -i *.deb
> sudo ln -sf /root/xflow/install1/vpp/build-root/install-vpp_debug-
> native/router/lib64/router.so.0.0.0 /usr/lib/vpp_plugins/router.so
>
> And then, I restart vpp service and check tap-inject which does not work.
>
> The contents of the build-root directory is given earlier in the thread.
>
>
> <https://www.avast.com/sig-email?utm_medium=email&utm_source=link&utm_campaign=sig-email&utm_content=webmail&utm_term=icon>
>  Virus-free.
> www.avast.com
> <https://www.avast.com/sig-email?utm_medium=email&utm_source=link&utm_campaign=sig-email&utm_content=webmail&utm_term=link>
> <#m_-8555950005750188709_DAB4FAD8-2DD7-40BB-A1B8-4E2AA1F9FDF2>
>
> On Tue, Apr 3, 2018 at 9:48 PM, Wang  wrote:
>
>> Would you share the command lines you used to install router plugin? It
>> seems to me the router plugin is not loaded?
>>
>> 2018-04-02 13:05 GMT-04:00 Hamid Rasool <14mseesras...@seecs.edu.pk>:
>>
>>> Hi Wang,
>>>
>>> My copy of the code contains the patch. I have checked the source and it
>>> is according to the diff in your commit
>>> <https://gerrit.fd.io/r/gitweb?p=vppsb.git;a=blobdiff;f=netlink/librtnl/mapper.c;h=05b7f1df43c2dae6ad193b8cb6abb6f70ae1673d;hp=65cc13a976e459f4e84d23fb8e1f8d11b9e0b543;hb=ce5ab87645e7472b99e23686047e4141a7a315b0;hpb=042a782c530ceb58507d697726e68ad7fe4fe023>.
>>> The make netlink-install and router-install now build without errors (using
>>> vpp 18.04-rc0 ; vpp stable/18.01 had the above issue).
>>>
>>> The problem I am facing that after netlink and router are built, I am
>>> unable to perform 'sudo vppctl enable tap-inject' command. Does the build
>>> command supposed to have .deb packages for vppsb. This is my build-root
>>> directory:
>>> xflow@vpphamid1:~/install1/vpp/build-root$ ls
>>> autowank  platforms.mk
>>> bin   python
>>> build-config.mk   sbin
>>> build-config.mk.READMEscripts
>>> build-tool-native tools
>>> build-vpp_debug-nativevagrant
>>> build-vpp-native  vpp_18.04-rc0~503-gf0404e9_amd64.changes
>>> config.site   vpp_18.04-rc0~503-gf0404e9_amd64.deb
>>> copyimg   vpp-api-java_18.04-rc0~503-gf0404e9_amd64.deb
>>> deb   vpp-api-lua_18.04-rc0~503-gf0404e9_amd64.deb
>>> install-vpp_debug-native  vpp-api-python_18.04-rc0~503-g
>>> f0404e9_amd64.deb
>>> install-vpp-nativevpp-dbg_18.04-rc0~503-gf0404e9_amd64.deb
>>> lib64 vpp-dev_18.04-rc0~503-gf0404e9_amd64.deb
>>> Makefile  vpp-lib_18.04-rc0~503-gf0404e9_amd64.deb
>>> path_setupvpp-plugins_18.04-rc0~503-gf0404e9_amd64.deb
>>>
>>>
>>>
>>>
>>> <https://www.avast.com/sig-email?utm_medium=email&utm_source=link&utm_campaign=sig-email&utm_content=webmail&utm_term=icon>
>>>  Virus-free.
>>> www.avast.com
>>> <https://www.avast.com/sig-email?utm_medium=email&utm_source=link&utm_campaign=sig-email&utm_content=webmail&utm_term=link>
>>> <#m_-8555950005750188709_m_8696160008385872788_m_-6449746801624537005_DAB4FAD8-2DD7-40BB-A1B8-4E2AA1F9FDF2>
>>>
>>> On Mon, Apr 2, 2018 at 6:14 PM, Wang  wrote:
>>>
>>>> Here is the patch
>>>>
>>>> https://gerrit.fd.io/r/#/c/11167/
>>>>
>>>> it should be merged into vppsb.
>>>>
>>>> Chad
>>>>
>>>> 2018-04-01 12:39 GMT-04:00 Hamid via Lists.Fd.Io <
>>>> 14mseesrasool=seecs.edu...@lists.fd.io>:
>>>>
>>>>> Not working with the latest code.
>>>>>
>>>>> I have retried with 'git checkout origin/master' after make wipe-all
>>>>> and make vpp-wipe commands but I am still getting the same error with
>>>>> librtnl files.
>>>>> 
>>>>>
>>>>>
>>>>
>>>
>>
>

Re: [vpp-dev] router plugin improvements

[Wang]

Hi Tom,

You can enable router-plugin by adding "enable tap-inject" in the startup
config file.

unix {
  nodaemon
  log /var/log/vpp/vpp.log
  full-coredump
  cli-listen /run/vpp/cli.sock
  gid vpp
  exec /vpp.init
}

in vpp.init you can write

"enable tap-inject"

Thanks
Chad

2018-05-17 2:11 GMT-04:00 Glaza :

> Hi
>
> At the beginning I would like to thank for router plugin.
>
> Actually my systemd service to start vpp looks like:
>
> ExecStart=/usr/bin/vpp -c /etc/vpp/startup.conf
> ExecStartPost=-/bin/sleep 5
> ExecStartPost=-/sbin/ip link set dev vpp0 up
> ExecStartPost=-/sbin/ip address add 10.0.0.14/30 dev vpp0
>
> Is there any better way to start vpp with router plugin enabled? It would
> be nice to have it in configuration like in https://github.com/iqiyi/dpvs
> where rte_kni.ko is used to make similar functionality.
>
> Thanks,
> Tom
>
>
>
> 
>
>

Re: [vpp-dev] router_plugin

[Wang]

I will take a look at it. Thanks

2018-05-17 0:12 GMT-04:00 Demian Pecile :

> Hi Wang trying last version.
> I have cannot find -lrtnl error.
>
> Any idea ?
>
> its working with last versions ?
>
>  Arch for platform 'vpp' is native 
>  Finding source for router 
>  Makefile fragment found in /vpp/build-data/packages/router.mk 
>  Source found in /vpp/router 
>  Configuring router: nothing to do 
>  Building router in /vpp/build-root/build-vpp-native/router 
> make[1]: Entering directory '/vpp/build-root/build-vpp-native/router'
>   CC   router/tap_inject_netlink.lo
>   CCLD router.la
> /usr/bin/ld: cannot find -lrtnl
> collect2: error: ld returned 1 exit status
> Makefile:444: recipe for target 'router.la' failed
> make[1]: *** [router.la] Error 1
> make[1]: Leaving directory '/vpp/build-root/build-vpp-native/router'
> Makefile:691: recipe for target 'router-build' failed
> make: *** [router-build] Error 2
>
>
>
> El 4 mar. 2018, a las 23:38, Wang  escribió:
>
> HI Gulakh,
>
> From my experience, VPP and VPPSB are compatible with the latest versions.
> What sequence of command lines did you use?
>
> 2018-03-04 4:58 GMT-05:00 Gulakh :
>
>> Hi,
>> I am using vppsb router plugin with vpp. there is a problem when I use
>> tap-inject.
>> The problem is that after I enable tap-inject and set interfaces' and
>> taps' IP address, I want to make taps' state up but at this point vpp
>> crashes and everything rolls back to its first state (No tap, No IP
>> address, ...)
>>
>> Q1) What is the problem??
>>
>> Q2) I have heard that VPP and VPPSB are not compatible in some versions
>> ... To which versions Should checkout both source code so that they can
>> work correctly??
>>
>> Thanks in advance
>>
>>
>>
>>
> 
>
>
> --
> Demian Pecile
> Siete Capas S.R.L.
> Periodistas Neuquinos 136
> <https://maps.google.com/?q=Periodistas+Neuquinos+136&entry=gmail&source=g>
> Piso 4 - Dpto. A - 8300 Neuquen
> Argentina
> Tel +54-299-4479172
> Cel. +549-299-5833500
>
>

[vpp-dev] Failed to build vppsb router plugin (librtnl)

[Wang]

Hi Yusuke,

It seems to me your OS kernel does not support the latest netlink macros
such as RTNLGRP_MPLS_ROUTE.

Thanks
Chad

2018-05-22 7:01 GMT-04:00 Yusuke Tatsumi :

> Hi vpp routing guys,
>
> I have tried vppsb plugins (router, netlink) and these plugins work well
> with
> - vpp: v18.04
> - vppsb: ce5ab87645e7472b99e23686047e4141a7a315b0
> - Bird: 1.4.5-2.el7.x86_64
>
> I noticed that vppsb current master can't be built.
> It seems to be caused by 2ab698c963b12219aecf72d001ae28c173548933 commit
> but I don't know how to fix this error.
> Can someone help to fix this issue?
>
> 0. References
> - https://github.com/FRRouting/frr/wiki/Alternate-forwarding-planes:-VPP
> - https://wiki.fd.io/view/VPP_Sandbox/router
>
> 1. build env.
> - CentOS 7.3.1611 w/ kernel 3.10.0-693.21.1.el7.x86_64
> - On baremetal server, NOT on vagrant.
>
> 2. build steps.
>  cd ~/
>  mkdir router-build-working-dir
>  cd router-build-working-dir
>  git clone https://gerrit.fd.io/r/vpp
>  git clone https://gerrit.fd.io/r/vppsb
>  cd vpp
>  git checkout v18.04
>  ln -sf ../vppsb/netlink
>  ln -sf ../vppsb/router
>  cd build-data/packages/
>  ln -sf ../../../vppsb/router/router.mk
>  ln -sf ../../../vppsb/netlink/netlink.mk
>  cd ../../build-root
>  make V=0 PLATFORM=vpp TAG=vpp_debug netlink-install router-install
>
> 3. result.
> ---
>  Building netlink in /home/ytatsumi/router-build-wo
> rking-dir/vpp/build-root/build-vpp_debug-native/netlink 
> make[1]: Entering directory `/home/ytatsumi/router-build-w
> orking-dir/vpp/build-root/build-vpp_debug-native/netlink'
>   CC   librtnl/netns.lo
>   CC   librtnl/rtnl.lo
> /home/ytatsumi/router-build-working-dir/vpp/build-data/../netlink/librtnl/rtnl.c:
> In function 'rtnl_socket_open':
> /home/ytatsumi/router-build-working-dir/vpp/build-data/../netlink/librtnl/rtnl.c:266:39:
> error: 'RTNLGRP_MPLS_ROUTE' undeclared (first use in this function)
>  grpmask(RTNLGRP_NOTIFY) | grpmask(RTNLGRP_MPLS_ROUTE),
>^
> /home/ytatsumi/router-build-working-dir/vpp/build-data/../netlink/librtnl/rtnl.c:266:39:
> note: each undeclared identifier is reported only once for each function it
> appears in
> /home/ytatsumi/router-build-working-dir/vpp/build-data/../netlink/librtnl/netns.c:69:5:
> error: 'RTA_VIA' undeclared here (not in a function)
>_(RTA_VIA, via, 1)\
>  ^
> /home/ytatsumi/router-build-working-dir/vpp/build-data/../netlink/librtnl/netns.c:82:13:
> note: in definition of macro '_'
>  .type = t, .unique = u, \
>  ^
> /home/ytatsumi/router-build-working-dir/vpp/build-data/../netlink/librtnl/netns.c:86:3:
> note: in expansion of macro 'ns_foreach_rta'
>ns_foreach_rta
>^
> make[1]: *** [librtnl/rtnl.lo] Error 1
> make[1]: *** Waiting for unfinished jobs
> make[1]: *** [librtnl/netns.lo] Error 1
> ---
>
> 4. deploy *.so files (If plugin build success)
> cd router-build-working-dir
> sudo cp vpp/build-root/install-vpp_debug-native/netlink/lib64/librtnl.so.0.0.0
> /usr/lib64/librtnl.so
> sudo cp 
> vpp/build-root/install-vpp_debug-native/netlink/lib64/testrtnl_plugin.so.0.0.0
> /usr/lib/vpp_plugins/testrtnl_plugin.so
> sudo cp vpp/build-root/install-vpp_debug-native/router/lib64/router.so.0.0.0
> /usr/lib/vpp_plugins/router.so
> sudo ln -sf /usr/lib64/librtnl.so /usr/lib64/librtnl.so.0
>
> # BTW, does anyone have plan to make "make pkg-rpm"?
> # From operation perspective, it is too hard to usebinary w/o rpm package..
>
> Thanks,
> Tatsumi.
>
> —
> 立見 祐介
> ヤフー株式会社
> テクノロジーグループ システム統括本部 サイトオペレーション本部 インフラ技術1部 プライベートクラウド
> TEL: 03-6898-3081
> maill: ytats...@yahoo-corp.jp
>
> —
> Yusuke Tatsumi
> Private Cloud
> Site Operations Division
> System Management Group
> Yahoo Japan Corporation
> Direct: +81 (3) 6898 3081
> mail: ytats...@yahoo-corp.jp
>
> 
>
>

Re: [vpp-dev] ACL Match in fa_node.c

[Wang, Yipeng1]

Hi, Andrew and Pan,

I came across the thread you are talking about improvements of ACL plugin. Any 
update? I am also interested in the performance of ACL,  wondering what is the 
direction of the optimizations you are working on. Could you share?

Thanks
Yipeng




>On 5/23/17, 张攀 yunshan.net.cn> wrote:

> Hi Andrew!

>

>

> -- Original --

> From:  "Andrew   Yourtchenko" gmail.com>;

> Date:  Tue, May 23, 2017 07:56 PM

> To:  "张攀" yunshan.net.cn>;

> Cc:  "vpp-dev" lists.fd.io>;

> Subject:  Re: [vpp-dev] ACL Match in fa_node.c

>

>

> Hi!

>

> On 5/23/17, 张攀  yunshan.net.cn> wrote:

>> Hi guys,

>>

>>

>> I looked into the source code of vpp/src/plugin/acl/fa_node.c,

>> in function full_acl_match_5tuple(), it seems that every ingress packet

>> is

>> matching against each ACL rule stored in acl_main->acls in a for-loop

>> manner. This seems not fairly effective.

>

> You're absolutely right on both counts. First make it work, then make

> it right, then make it fast :-) I have some ideas that I wanted to

> experiment with there, would you be interested to help ? ACL matching

> is a fairly distinct function operation to not affect much else.

>

> [PAN]: I would be very pleased to help as I am also a addicted person to

> high performance programming :D

>

>



Great! :-) I will try to sketch the idea tomorrow/thursday and will

add you to the draft, so we can work together. (I also will have

limited connectivity the next week, so I won't be a lot in your way!

:-) (or if you have some good idea on how you'd like to do it, feel

free to shoot a doc/code into a gerrit draft, let's see!)



>

>>

>>

>> Besides, I notice that in vpp/src/plugin/acl/acl.c，when you call the

>> function acl_hook_l2_input_classify(), you will create a

>> vnet_classify_table, but I didn't see any code which adds

>> classify_session

>> to it, why?

>

> I had used classify table for storing the sessions in the pre-1704

> version of the ACL plugin.

>

> in 1704 as I was adding the L3, I moved to the new data path while

> keeping the old one still around, and potentially switchable (not

> terribly gracefully, but still).

>

> In the current master the classifier is used merely as a hook to get

> into the packet processing within the L2 packet path - that's why you

> see no sessions added.

>

> [PAN]: Cool. Correct me if I am wrong: ingress packet will first check

> against if there is matched session existing

> in fa_node.c and if not, the packet will check against ACL then to decide

> whether to create a new session or drop.



Exactly, that is the idea!



>

>

> Is there any 'match then action' alike behavior existing in fa_node?  And a

> long standing mystery question for me: what

> does "fa" stand for :p

>



"FA" is "feature arc" :) When adding the routed path, the feature arc

support appeared a bit earlier, and I was "wow, this is so excellent

and easy", and I thought I might use the same mechanism in L2 as well,

so I just bluntly called everything "fa_*". But then I decided to

leave for L2 data path the old mechanism of (ab)using the L2

classifier, since it allowed for a while to retain (just in case) the

ability to switch between the old and the new data path. As the time

goes, I suspect there might be a better way to hook in L2 forwarding

path...



>

>

>>

>>

>> Is there any document/idea could basically explain the relationships

>> between

>> acl/fa_node and vnet_classify?

>

> In a gerrit draft that I am working now on, which aims to provide the

> multicore support, there is text file with the history and the

> description of operation. If you like, I can add you to the reviewers

> list so you can have a look at that - just send me your email that you

> use in gerrit.

>

> [PAN]: Thank you Andrew, I believe this will help a lot. My email is

>  dazhangpan at gmail.com

>



Excellent, I've just added you, it's 6771. As you notice, it's still a

good amount of work to do (lots of clib_warning()s to ensure

everything indeed does what I think it does). Looking forward to your

feedback! :-)



--a



>

>

> --a

>

>>

>>

>> Any help will be much appreciated.

>>

>>

>> Best Regards,

>>

>>

>> Pan

___
vpp-dev mailing list
vpp-dev@lists.fd.io
https://lists.fd.io/mailman/listinfo/vpp-dev

Re: [vpp-dev] ACL Match in fa_node.c

[Wang, Yipeng1]

Thank you Andrew, seems I was looking at an older version...

I was thinking of similar thing like using  tuple space search like you already 
done. It is very good work and lots of considerations. One quick question:

I just start looking at the code, but it seems all the rules are stored in one 
hash table (acl_lookup_hash). If a packet say having src IP= 1.1.0.0, 
dst=2.2.2.2, and there are two rules one is src=1.1.0.0/16, dst=2.2.0.0/32  and 
the other is src=1.1.0.0/32, dst=2.2.0.0/16. The two rules will have same key 
value in the table right? When do lookup, assume packet header is masked by the 
second rule’s mask type and become src=1.1.0.0 dst=2.2.0.0.  Will it match the 
first rule as well by mistake?

Thank you.
Yipeng

From: Andrew Yourtchenko [mailto:ayour...@gmail.com]
Sent: Sunday, August 27, 2017 6:30 AM
To: Wang, Yipeng1 
Cc: vpp-dev@lists.fd.io; zhang...@yunshan.net.cn
Subject: Re: [vpp-dev] ACL Match in fa_node.c

Hi Yipeng,

It's already there - just have a look through hash_* files in the ACL plugin 
directory on the master or latest stable/1707 :-)

There are several things more that can be taken care of (e.g. the determination 
of the "ACE not shadowed" shortcut for applied ACEs would allow some good 
further improvement, or more distinction between the v4 and v6 lookups, and 
potentially more cleverness with handling of port ranges).

Also, if you have ideas for a different lookup mechanism - feel free to add one 
- using a similar API as for the hash-based one it should be fairly 
straightforward.

Do you have some specific ideas in mind ?

--a

On 26 Aug 2017, at 03:23, Wang, Yipeng1 
mailto:yipeng1.w...@intel.com>> wrote:
Hi, Andrew and Pan,

I came across the thread you are talking about improvements of ACL plugin. Any 
update? I am also interested in the performance of ACL,  wondering what is the 
direction of the optimizations you are working on. Could you share?

Thanks
Yipeng




>On 5/23/17, 张攀 yunshan.net.cn<https://lists.fd.io/mailman/listinfo/vpp-dev>> wrote:

> Hi Andrew!

>

>

> -- Original --

> From:  "Andrew   Yourtchenko" gmail.com<https://lists.fd.io/mailman/listinfo/vpp-dev>>;

> Date:  Tue, May 23, 2017 07:56 PM

> To:  "张攀" yunshan.net.cn<https://lists.fd.io/mailman/listinfo/vpp-dev>>;

> Cc:  "vpp-dev" lists.fd.io<https://lists.fd.io/mailman/listinfo/vpp-dev>>;

> Subject:  Re: [vpp-dev] ACL Match in fa_node.c

>

>

> Hi!

>

> On 5/23/17, 张攀  yunshan.net.cn<https://lists.fd.io/mailman/listinfo/vpp-dev>> wrote:

>> Hi guys,

>>

>>

>> I looked into the source code of vpp/src/plugin/acl/fa_node.c,

>> in function full_acl_match_5tuple(), it seems that every ingress packet

>> is

>> matching against each ACL rule stored in acl_main->acls in a for-loop

>> manner. This seems not fairly effective.

>

> You're absolutely right on both counts. First make it work, then make

> it right, then make it fast :-) I have some ideas that I wanted to

> experiment with there, would you be interested to help ? ACL matching

> is a fairly distinct function operation to not affect much else.

>

> [PAN]: I would be very pleased to help as I am also a addicted person to

> high performance programming :D

>

>



Great! :-) I will try to sketch the idea tomorrow/thursday and will

add you to the draft, so we can work together. (I also will have

limited connectivity the next week, so I won't be a lot in your way!

:-) (or if you have some good idea on how you'd like to do it, feel

free to shoot a doc/code into a gerrit draft, let's see!)



>

>>

>>

>> Besides, I notice that in vpp/src/plugin/acl/acl.c，when you call the

>> function acl_hook_l2_input_classify(), you will create a

>> vnet_classify_table, but I didn't see any code which adds

>> classify_session

>> to it, why?

>

> I had used classify table for storing the sessions in the pre-1704

> version of the ACL plugin.

>

> in 1704 as I was adding the L3, I moved to the new data path while

> keeping the old one still around, and potentially switchable (not

> terribly gracefully, but still).

>

> In the current master the classifier is used merely as a hook to get

> into the packet processing within the L2 packet path - that's why you

> see no sessions added.

>

> [PAN]: Cool. Correct me if I am wrong: ingress packet will first check

> against if there is matched session existing

> in fa_node.c and if not, the packet will check against ACL then to decide

> whether to create a new session or drop.



Exactly, that is the idea!



>

>

> Is there any 'match then action' alike behavior existing in fa_node?  And a

Re: [vpp-dev] ACL Match in fa_node.c

[Wang, Yipeng1]

Thank you Andrew for the detailed explanation.  It is very helpful for me to 
understand the plugin.

Thanks
Yipeng

> -Original Message-
> From: Andrew 👽 Yourtchenko [mailto:ayour...@gmail.com]
> Sent: Tuesday, August 29, 2017 5:05 AM
> To: Wang, Yipeng1 
> Cc: vpp-dev@lists.fd.io; zhang...@yunshan.net.cn
> Subject: Re: [vpp-dev] ACL Match in fa_node.c
> 
> Hi Yipeng,
> 
> yeah, this case should be handled as well - note that the ACL lookup hash is 
> 48x8,
> while the session lookup hash is 40x8, and the fa_5tuple_t (being 48 bytes in 
> size)
> contains the fa_packet_info_t which is sitting at the last 8 bytes.
> 
> That data structure has a bunch of flag-type information derived from the
> packet:
> 
> typedef union {
>   u64 as_u64;
>   struct {
> u32 sw_if_index;
> u16 mask_type_index_lsb;
> u8 tcp_flags;
> u8 tcp_flags_valid:1;
> u8 is_input:1;
> u8 l4_valid:1;
> u8 is_nonfirst_fragment:1;
> u8 is_ip6:1;
> u8 flags_reserved:3;
>   };
> } fa_packet_info_t;
> 
> They are filled during the initial extraction of the 5-tuple, and are also 
> used for
> the lookup in the hash table (this way we can match non-initial fragments
> differently from the initial ones).
> 
> Note the "mask_type_index_lsb" field - when preparing for the lookup iteration
> in multi_acl_match_get_applied_ace_index, it is filled with the index of the 
> mask
> used:
> 
> kv_key->pkt.mask_type_index_lsb = mask_type_index;
> 
> So in the example you cite, there will be two lookups in
> multi_acl_match_get_applied_ace_index, which will both match, but the one
> which is earlier in the ACLs will win, because of the check in "if 
> (result_val-
> >applied_entry_index < curr_match_index)" - thus the end result will be the
> same as if we just scanned the entire list downwards.
> 
> Hopefully this clarifies the logic! :-)
> 
> --a
> 
> On 8/29/17, Wang, Yipeng1  wrote:
> > Thank you Andrew, seems I was looking at an older version...
> >
> > I was thinking of similar thing like using  tuple space search like
> > you already done. It is very good work and lots of considerations. One
> > quick
> > question:
> >
> > I just start looking at the code, but it seems all the rules are
> > stored in one hash table (acl_lookup_hash). If a packet say having src
> > IP= 1.1.0.0, dst=2.2.2.2, and there are two rules one is
> > src=1.1.0.0/16, dst=2.2.0.0/32 and the other is src=1.1.0.0/32,
> > dst=2.2.0.0/16. The two rules will have same key value in the table
> > right? When do lookup, assume packet header is masked by the second rule’s
> mask type and become src=1.1.0.0 dst=2.2.0.0.
> > Will it match the first rule as well by mistake?
> >
> > Thank you.
> > Yipeng
> >
> > From: Andrew Yourtchenko [mailto:ayour...@gmail.com]
> > Sent: Sunday, August 27, 2017 6:30 AM
> > To: Wang, Yipeng1 
> > Cc: vpp-dev@lists.fd.io; zhang...@yunshan.net.cn
> > Subject: Re: [vpp-dev] ACL Match in fa_node.c
> >
> > Hi Yipeng,
> >
> > It's already there - just have a look through hash_* files in the ACL
> > plugin directory on the master or latest stable/1707 :-)
> >
> > There are several things more that can be taken care of (e.g. the
> > determination of the "ACE not shadowed" shortcut for applied ACEs
> > would allow some good further improvement, or more distinction between
> > the v4 and
> > v6 lookups, and potentially more cleverness with handling of port ranges).
> >
> > Also, if you have ideas for a different lookup mechanism - feel free
> > to add one - using a similar API as for the hash-based one it should
> > be fairly straightforward.
> >
> > Do you have some specific ideas in mind ?
> >
> > --a
> >
> > On 26 Aug 2017, at 03:23, Wang, Yipeng1
> > mailto:yipeng1.w...@intel.com>> wrote:
> > Hi, Andrew and Pan,
> >
> > I came across the thread you are talking about improvements of ACL plugin.
> > Any update? I am also interested in the performance of ACL,  wondering
> > what is the direction of the optimizations you are working on. Could you 
> > share?
> >
> > Thanks
> > Yipeng
> >
> >
> >
> >
> >>On 5/23/17, 张攀  >> yunshan.net.cn<https://lists.fd.io/mailman/listinfo/vpp-dev>> wrote:
> >
> >> Hi Andrew!
> >
> >>
> >
> >>
> >
> >> -- Original --
> >
> >> From:  "Andrew   Yourtchenko" >> gmail.com<https

[vpp-dev] FileNotFoundError error(2): No such file or directory as: /dev/shm/vpe-api

[Bojian Wang]

hi vpp team

I tried to install vpp onto ubuntu 16.04, and everything looks good when I
followed installation guide from FD.io wiki. but after I complete the
installation and am going to start vpp service, encounter the error message
as below:

cisco@server-1:~$ sudo vppctl
_____   _  ___
 __/ __/ _ \  (_)__| | / / _ \/ _ \
 _/ _// // / / / _ \   | |/ / ___/ ___/
 /_/ /(_)_/\___/   |___/_/  /_/

vpp# pwd
FileNotFoundError error(2): No such file or directory as:
/dev/shm/vpe-api
cisco@server-1:~$ systemctl status vpp
‚óè vpp.service - vector packet processing engine
   Loaded: loaded (/lib/systemd/system/vpp.service; enabled; vendor preset:
enabled)
   Active: inactive (dead) (Result: exit-code) since Mon 2018-02-12
07:49:57 UTC; 2min 39s ago
  Process: 3141 ExecStopPost=/bin/rm -f /dev/shm/db /dev/shm/global_vm
/dev/shm/vpe-api (code=exited, status=0/SUCCESS)
  Process: 3135 ExecStart=/usr/bin/vpp -c /etc/vpp/startup.conf
(code=exited, status=1/FAILURE)
  Process: 3133 ExecStartPre=/sbin/modprobe uio_pci_generic (code=exited,
status=1/FAILURE)
  Process: 3129 ExecStartPre=/bin/rm -f /dev/shm/db /dev/shm/global_vm
/dev/shm/vpe-api (code=exited, status=0/SUCCESS)
 Main PID: 3135 (code=exited, status=1/FAILURE)

regards/Bojian Wang

[vpp-dev] Issue with deploying VPP image under "server model" in Cisco VIRL

[Bojian Wang]

hi, Team

I am using “server mode” image to deploy VPP in my VIRL environment. The
particular image work well in VIRL 1.2 environment, I can start VPP service
and config i/f and IP for VPP. But When I snapshot the image , download to
local and then upload to other VIRL 1.3 server ( also I tried VIRL 1.5
server), this image seem can’t work well. I can still boot this image but
when I tried to enable VPP service, encountering the following error:



*** error message in VIRL 1.3 ***



cisco@vpp-1:~$ systemctl status vpp.service

‚óè vpp.service - vector packet processing engine

   Loaded: loaded (/lib/systemd/system/vpp.service; enabled; vendor preset:
enabled)

   Active: inactive (dead) (Result: exit-code) since Wed 2018-02-14
05:32:29 UTC; 1min 13s ago

  Process: 1360 ExecStopPost=/bin/rm -f /dev/shm/db /dev/shm/global_vm
/dev/shm/vpe-api (code=exited, status=0/SUCCESS)

  Process: 1357 ExecStart=/usr/bin/vpp -c /etc/vpp/startup.conf
(code=exited, status=1/FAILURE)

  Process: 1353 ExecStartPre=/sbin/modprobe uio_pci_generic (code=exited,
status=1/FAILURE)

  Process: 1350 ExecStartPre=/bin/rm -f /dev/shm/db /dev/shm/global_vm
/dev/shm/vpe-api (code=exited, status=0/SUCCESS)

 Main PID: 1357 (code=exited, status=1/FAILURE)

cisco@vpp-1:~$

cisco@vpp-1:~$

cisco@vpp-1:~$



** as comparison, this image has used to work in VIRL 1.2
*





cisco@vpp-1:~$ sudo systemctl status vpp.service

[sudo] password for cisco:

‚óè vpp.service - vector packet processing engine

   Loaded: loaded (/lib/systemd/system/vpp.service; enabled; vendor preset:
enabled)

   Active: active (running) since Wed 2018-02-14 00:09:32 UTC; 5h 25min ago

  Process: 1119 ExecStartPre=/sbin/modprobe uio_pci_generic (code=exited,
status=1/FAILURE)

  Process: 1075 ExecStartPre=/bin/rm -f /dev/shm/db /dev/shm/global_vm
/dev/shm/vpe-api (code=exited, status=0/SUCCESS)

 Main PID: 1135 (vpp_main)

Tasks: 3

   Memory: 81.8M

  CPU: 5h 24min 52.306s

   CGroup: /system.slice/vpp.service

   └─1135 /usr/bin/vpp -c /etc/vpp/startup.conf



Feb 14 00:09:33 vpp-1 /usr/bin/vpp[1135]: load_one_plugin:63: Loaded
plugin: /usr/lib/vpp_api_test_plugins/ioam_trace_test_plugin.so

Feb 14 00:09:33 vpp-1 vpp[1135]: /usr/bin/vpp[1135]: vlib_pci_bind_to_uio:
Skipping PCI device :00:03.0 as host interface ens3 is up

Feb 14 00:09:33 vpp-1 /usr/bin/vpp[1135]: vlib_pci_bind_to_uio: Skipping
PCI device :00:03.0 as host interface ens3 is up

Feb 14 00:09:33 vpp-1 vpp[1135]: /usr/bin/vpp[1135]: dpdk_config:1215: EAL
init args: -c 1 -n 4 --huge-dir /run/vpp/hugepages --file-prefix vpp -b
:00:03.0 --master-lcore 0 --socket-mem 256

Feb 14 00:09:33 vpp-1 /usr/bin/vpp[1135]: dpdk_config:1215: EAL init args:
-c 1 -n 4 --huge-dir /run/vpp/hugepages --file-prefix vpp -b :00:03.0
--master-lcore 0 --socket-mem 256

Feb 14 00:09:34 vpp-1 vpp[1135]: EAL: WARNING: cpu flags constant_tsc=yes
nonstop_tsc=no -> using unreliable clock cycles !

Feb 14 00:09:34 vpp-1 vnet[1135]: EAL: WARNING: cpu flags constant_tsc=yes
nonstop_tsc=no -> using unreliable clock cycles !

Feb 14 00:09:34 vpp-1 vpp[1135]: EAL:   Invalid NUMA socket, default to 0

Feb 14 00:09:34 vpp-1 vnet[1135]: EAL:   Invalid NUMA socket, default to 0

Feb 14 00:09:34 vpp-1 vnet[1135]: dpdk_ipsec_process:271: not enough
Cryptodevs, default to OpenSSL IPsec

cisco@vpp-1:~$

cisco@vpp-1:~$

Re: [vpp-dev] Very poor performance vm to vm via VPP vhostuser

[Yichen Wang]

Hi, Sara,

May I know what image you are using to achieving the good performance with VPP? 
I assume you are running the same iperf3 testings?

Thanks very much!

Regards,
Yichen

From:  on behalf of Sara Gittlin 
Date: Sunday, March 25, 2018 at 5:31 AM
To: "vpp-dev@lists.fd.io" 
Subject: Re: [vpp-dev] Very poor performance vm to vm via VPP vhostuser

Fyi
When i launch another image (not based on the clear-linux-kvm) - i'm achieving 
good performance (exactly like ovs-dpdk - which is expected in single flow/mac )
any idea why ?
this is the  launch cmdline:  [anyway i cant find a way to set the 
tx/rx-queue-size)

 sudo qemu-system-x86_64 -name vhost1 -m 2048M -smp 4 -cpu host -hda 
/var/lib/libvirt/images/vm1.qcow2 -boot c -enable-kvm -no-reboot -net none 
-chardev socket,id=char3,path=/var/run/vpp/sock2.sock -netdev 
type=vhost-user,id=mynet1,chardev=char3,vhostforce -device 
virtio-net-pci,mac=c0:00:00:00:00:ff,netdev=mynet1 -object 
memory-backend-file,id=mem,size=2048M,mem-path=/dev/hugepages,share=on -numa 
node,memdev=mem -mem-prealloc
Thank you
-Sara


On Sun, Mar 25, 2018 at 11:56 AM, Sara Gittlin 
mailto:sara.gitt...@gmail.com>> wrote:
Thank Steven and Alec
I'm using the iperf as a basic tool to check sanity - i'll run pktgen on the 
VMs later.
the resut i'm got 20Mbps (~20K pps) throughput and latency of 1ms is not iperf 
issue.
Steven -  you suggest to set the rx-queue-size when lanuching  the VM but 
unfortunately i dont find this option in the qemy-system-x86_64 command line
i'm using this:

qemu-system-x86_64 \

-enable-kvm -m 1024 \

-bios OVMF.fd \

-smp 4 -cpu host \

-vga none -nographic \

-drive file="1-clear-14200-kvm.img",if=virtio,aio=threads \

-chardev socket,id=char1,path=/var/run/vpp/sock1.sock \

-netdev type=vhost-user,id=mynet1,chardev=char1,vhostforce \

-device virtio-net-pci,mac=00:00:00:00:00:01,netdev=mynet1 \

-object 
memory-backend-file,id=mem,size=1024M,mem-path=/dev/hugepages,share=on \

-numa node,memdev=mem -mem-prealloc \

-debugcon file:debug.log -global isa-debugcon.iobase=0x402



-Sara


On Thu, Mar 22, 2018 at 6:10 PM, steven luong 
mailto:slu...@cisco.com>> wrote:
Sara,

Iperf3 is not blasting traffic fast enough. You could try specifying multiple 
parallel streams using -P. Then, you will likely encounter vhost-user dropping 
packets as you are using the default virtqueue size 256. You’ll need to specify 
rx_queue_size=1024 when you launch the VM in the qemu start command to reduce 
the drops.

Steven

From: mailto:vpp-dev@lists.fd.io>> on behalf of Sara 
Gittlin mailto:sara.gitt...@gmail.com>>
Date: Thursday, March 22, 2018 at 8:12 AM
To: "vpp-dev@lists.fd.io" 
mailto:vpp-dev@lists.fd.io>>
Cc: "vpp-dev@lists.fd.io" 
mailto:vpp-dev@lists.fd.io>>

Subject: Re: [vpp-dev] Very poor performance vm to vm via VPP vhostuser

Hi John and Steven

setting this in the startup config didnt help
vhost-user {
  coalesce-frame 0
}
John

I'm using ping -f  for latency and iperf3 for pps testing.
later i'll run pktgen in the vm's

output :
sho int
  Name   Idx   State  Counter  Count
VirtualEthernet0/0/0  1 up   rx packets 
11748
 rx bytes  
868648
 tx packets 
40958
 tx bytes
58047352
 drops  
   29
VirtualEthernet0/0/1  2 up   rx packets 
40958
 rx bytes
58047352
 tx packets 
11719
 tx bytes  
862806
 tx-error   
   29
local00 up


 show hardware
  NameIdx   Link  Hardware
VirtualEthernet0/0/0   1 up   VirtualEthernet0/0/0
  Ethernet address 02:fe:25:2f:bd:c2
VirtualEthernet0/0/1   2 up   VirtualEthernet0/0/1
  Ethernet address 02:fe:40:16:70:1b
local0 0down  local0
  local

-Sara


On Thu, Mar 22, 2018 at 4:53 PM, John DeNisco 
mailto:jdeni...@cisco.com>> wrote:

Hi Sara,

Can you also send the results from show hardware and show interfaces?

What are you using to test your performance.

John


From: mailto:vpp-dev@lists.fd.io>> on behalf of Sara 
Gittlin mailto:sara.gitt...@gmail.com>>
Date: Thursday, March 22, 2018 at 9:27 AM
To: "vpp-dev@lists.fd.io" 
mailto:vpp-dev@lists.fd.io>>
Subject: Re: [vpp-dev] Very poor performance vm to vm via VPP vhostuser

this i

Re: [vpp-dev] virtio devices add to white list

[Yichen Wang]

Hi, Avu,

If I am understanding correctly, you want to run VPP in a VM, while the 
interface is a virtio interface.

If that is the case, there is no difference as you want to use with a physical 
NIC interface. You need to get the PCI address of your virtio interface by 
using "lspci" + "ip" command. Or just this "basename $(readlink 
/sys/class/net/ens3/device)". Once you have the pci address, put below in the 
startup.conf file for VPP to consume:
dpdk {
  dev {{PCI_ADDRESS_1}}
  dev {{PCI_ADDRESS_2}}
  uio-driver igb_uio
}

For uio-driver, you have the choice of igb_uio or vfio-pci. Up to you. If you 
want igb_uio, you may need to compile it manually from dpdk if on RedHat distro.

Thanks very much!

Regards,
Yichen

On 3/28/18, 10:54 PM, "vpp-dev@lists.fd.io on behalf of Avi Cohen (A)" 
 wrote:

Thank you Steven
My interfaces are not vhost-user.
Just virtio interfaces created on the host and then assigned to the VM.
In my VM is see them as ens3 , ens4 etc...

Regards
Avu


> -Original Message-
> From: vpp-dev@lists.fd.io [mailto:vpp-dev@lists.fd.io] On Behalf Of steven
> luong
> Sent: Wednesday, 28 March, 2018 7:08 PM
> To: vpp-dev@lists.fd.io
> Subject: Re: [vpp-dev] virtio devices add to white list
> 
> Avi,
> 
> Yes, you can. As an example, I have it like this in my startup.conf
> 
> dpdk {
>  vdev virtio_user0,path=/tmp/sock0
> }
> 
> Steven
> 
> On 3/28/18, 8:09 AM, "vpp-dev@lists.fd.io on behalf of Avi Cohen (A)" 
 d...@lists.fd.io on behalf of avi.co...@huawei.com> wrote:
> 
> Hi
> In the startup.conf , in the dpdk part we can add pci devices to the 
white-list .
> Is there option to also add virtual interfaces to white-list - I'm 
running VPP on
> virtual-machine and I have few virtio devices that I would automatically 
add to
> VPP
> 
> Best Regards
> Avi
> 
> 
> 
> 
> 
> 
> 







-=-=-=-=-=-=-=-=-=-=-=-
Links:

You receive all messages sent to this group.

View/Reply Online (#8720): https://lists.fd.io/g/vpp-dev/message/8720
View All Messages In Topic (5): https://lists.fd.io/g/vpp-dev/topic/16303673
Mute This Topic: https://lists.fd.io/mt/16303673/21656
New Topic: https://lists.fd.io/g/vpp-dev/post

Change Your Subscription: https://lists.fd.io/g/vpp-dev/editsub/21656
Group Home: https://lists.fd.io/g/vpp-dev
Contact Group Owner: vpp-dev+ow...@lists.fd.io
Terms of Service: https://lists.fd.io/static/tos
Unsubscribe: https://lists.fd.io/g/vpp-dev/unsub
-=-=-=-=-=-=-=-=-=-=-=-

Re: [vpp-dev] virtio devices add to white list

[Yichen Wang]

This is a general questions of Linux. Proper way is (1) Install the kernel 
modules; (2) Create /etc/modules-load.d/.conf so KO will automatically 
being loaded

More practical way, or for development:
(1) If you have a systemd managed VPP service, you can add these commands in 
the systemd service file as ExecPre task;
(2) Just put this two line in /etc/rc.local.

Normally for long term, you may want to deal with kernel update case for this 
particular KO.

Thanks very much!

Regards,
Yichen 

On 3/29/18, 5:11 AM, "vpp-dev@lists.fd.io on behalf of Avi Cohen (A)" 
 wrote:

Yichen
Also do I have to insmod the igb_uio or this can be done auto by vpp ? 
where to put it ?
Regards
Avi

> -Original Message-
> From: Avi Cohen (A)
> Sent: Thursday, 29 March, 2018 2:34 PM
    > To: 'Yichen Wang (yicwang)'; vpp-dev@lists.fd.io
> Subject: RE: [vpp-dev] virtio devices add to white list
> 
> Great Yichen,
> I see that the interfaces should be down before VPP  startup, otherwise 
it does
> not acquire the interface
> 
> Best Regards
> Avi
> 
> 
> > -Original Message-
> > From: Yichen Wang (yicwang) [mailto:yicw...@cisco.com]
> > Sent: Thursday, 29 March, 2018 9:45 AM
> > To: Avi Cohen (A); vpp-dev@lists.fd.io
> > Subject: Re: [vpp-dev] virtio devices add to white list
> >
> > Hi, Avu,
> >
> > If I am understanding correctly, you want to run VPP in a VM, while
> > the interface is a virtio interface.
> >
> > If that is the case, there is no difference as you want to use with a
> > physical NIC interface. You need to get the PCI address of your virtio
> > interface by using "lspci" + "ip" command. Or just this "basename
> > $(readlink /sys/class/net/ens3/device)". Once you have the pci
> > address, put below in the startup.conf file for VPP to consume:
> > dpdk {
> >   dev {{PCI_ADDRESS_1}}
> >   dev {{PCI_ADDRESS_2}}
> >   uio-driver igb_uio
> > }
> >
> > For uio-driver, you have the choice of igb_uio or vfio-pci. Up to you.
> > If you want igb_uio, you may need to compile it manually from dpdk if on
> RedHat distro.
> >
> > Thanks very much!
> >
> > Regards,
> > Yichen
> >
> > On 3/28/18, 10:54 PM, "vpp-dev@lists.fd.io on behalf of Avi Cohen
> > (A)"  wrote:
> >
> > Thank you Steven
> > My interfaces are not vhost-user.
> > Just virtio interfaces created on the host and then assigned to the 
VM.
> > In my VM is see them as ens3 , ens4 etc...
> >
> > Regards
> > Avu
> >
> >
> > > -Original Message-
> > > From: vpp-dev@lists.fd.io [mailto:vpp-dev@lists.fd.io] On Behalf 
Of
> steven
> > > luong
> > > Sent: Wednesday, 28 March, 2018 7:08 PM
> > > To: vpp-dev@lists.fd.io
> > > Subject: Re: [vpp-dev] virtio devices add to white list
> > >
> > > Avi,
> > >
> > > Yes, you can. As an example, I have it like this in my 
startup.conf
> > >
> > > dpdk {
> > >  vdev virtio_user0,path=/tmp/sock0
> > > }
> > >
> > > Steven
> > >
> > > On 3/28/18, 8:09 AM, "vpp-dev@lists.fd.io on behalf of Avi Cohen 
(A)"
> >  > > d...@lists.fd.io on behalf of avi.co...@huawei.com> wrote:
> > >
> > > Hi
> > > In the startup.conf , in the dpdk part we can add pci devices 
to the
> white-
> > list .
> > > Is there option to also add virtual interfaces to white-list 
- I'm running
> > VPP on
> > > virtual-machine and I have few virtio devices that I would
> > automatically add to
> > > VPP
> > >
> > > Best Regards
> > > Avi
> > >
> > >
> > >
> > >
> > >
> > >
> > >
> >
> >
> > 
> >
> >







-=-=-=-=-=-=-=-=-=-=-=-
Links:

You receive all messages sent to this group.

View/Reply Online (#8733): https://lists.fd.io/g/vpp-dev/message/8733
View All Messages In Topic (8): https://lists.fd.io/g/vpp-dev/topic/16303673
Mute This Topic: https://lists.fd.io/mt/16303673/21656
New Topic: https://lists.fd.io/g/vpp-dev/post

Change Your Subscription: https://lists.fd.io/g/vpp-dev/editsub/21656
Group Home: https://lists.fd.io/g/vpp-dev
Contact Group Owner: vpp-dev+ow...@lists.fd.io
Terms of Service: https://lists.fd.io/static/tos
Unsubscribe: https://lists.fd.io/g/vpp-dev/unsub
-=-=-=-=-=-=-=-=-=-=-=-

Re: [vpp-dev] What RPM contains vpp debug symbols?

[Yichen Wang]

Hi, Thomas,

Just want to revive this thread… Do we know when can we have this fixed? We are 
testing and will likely to ship with VPP 18.04 once the final release comes 
out. Having the debuginfo RPM is *very* important in our product, as it will 
give clues when VPP crashes. Given it has been broken for 6months+, we would 
really need your helps so that we have the debuginfo RPM before we finalize our 
code.

Thanks very much!

Regards,
Yichen

From: Dave Wallace 
Date: Thursday, January 25, 2018 at 8:20 PM
To: Billy McFall 
Cc: "vpp-dev@lists.fd.io" , "Yichen Wang (yicwang)" 

Subject: Re: [vpp-dev] What RPM contains vpp debug symbols?

Thanks for the info Billy!
-daw-
On 1/25/18 8:26 PM, Billy McFall wrote:
'make pkg-rpm' is currently not generating an RPM with debug symbols. There is 
a bug in  rpmbuild for packages with short names (i.e. - vpp) that was not 
allowing RPMs to be built upstream in CentOS and RHEL. The work around is to 
add a -0 to the package name to make it long enough. Much later after this 
change was merged, it was discovered that the debug RPM was not being built.

Thomas Herbert is tracking the issue. rpmbuild has been fixed in Fedora but 
that fix has not been pulled into CentOS or RHEL. Not sure when it will make 
it. Thomas can give more details, I am just regurgitating what he previously 
reported to me.

Billy McFall


On Thu, Jan 25, 2018 at 6:57 PM, Dave Wallace 
mailto:dwallac...@gmail.com>> wrote:
Does anyone know what RPM contains debug symbols?

Back in VPP 17.07 there used to be an RPM package called 
vpp-debuginfo-*.x86_64.rpm, but that package is no longer installed on 
nexus.fd.io<http://nexus.fd.io> starting with release 17.10

https://nexus.fd.io/content/repositories/fd.io.stable.1707.centos7/io/fd/vpp/vpp-debuginfo/17.07.01-release.x86_64/vpp-debuginfo-17.07.01-release.x86_64.rpm

Thanks,
-daw-

___
vpp-dev mailing list
vpp-dev@lists.fd.io<mailto:vpp-dev@lists.fd.io>
https://lists.fd.io/mailman/listinfo/vpp-dev



--
Billy McFall
Networking Group
CTO Office
Red Hat

Re: [vpp-dev] [csit-dev] vHost user test scenarios for CSIT

[Damon Wang]

About

VPP in guest with layer 2 and layer 3 vRouted traffic.


Does this mean vpp run in guest vm? There are lots needs for running vpp in
a vm as a VNF, the point is testing vpp routing performance in a vm with
vhost-user.

+-+
| |
| |
|++   |
||   VPP in Guest VM  |   |
|||   |
|| Routing form   |   |
|| eth0 to eth1   |   |
|||   |
|++---+--+---++   |
| |   |  |   ||
| | eth0  |  |  eth1 ||
| |   |  |   ||
| +---+---+  +---+---+|
| |  ||
| |  ||
| |  ||
| |  ||
| +---+--+---+|
| |  ||
| |  ||
| |   vSwitch, eg. OVS DPDK  ||
| |  ||
| |  ||
+--+--+---+-++-+--+
   |  | |  |
   |  eth0| |   eth1   |
   |  | |  |
   ++-+ +-++
| |
| |
| |
| |
   ++-+ +-++
   |  | |  |
   |  eth0| |   eth1   |
   |  | |  |
+--+--+-+--+---+
|  |
|  Traffic Generator   |
|  |
|  |
+--+


2016-11-17 9:06 GMT+08:00 Alec Hothan (ahothan) :

> Few comments inline…
>
>
> On 11/16/16, 8:18 AM, "vpp-dev-boun...@lists.fd.io on behalf of Thomas F
> Herbert" 
> wrote:
>
> +Irene Liew from Intel
>
> On 11/15/2016 02:06 PM, Maciek Konstantynowicz (mkonstan) wrote:
>
> On 11 Nov 2016, at 13:58, Thomas F Herbert <  therb...@redhat.com>therb...@redhat.com> wrote:
>
>
> On 11/09/2016 07:39 AM, Maciek Konstantynowicz (mkonstan) wrote:
>
>
> Some inputs from my side with MK.
>
>
> On 8 Nov 2016, at 21:25, Thomas F Herbert  wrote:
>
> All:
>
> Soliciting opinions from people as to vhost-user testing scenarios and
> guest modes in
> fd.io  CSIT testing of VPP - vhost-user.
> I will forward to this mailing list as well as summarize any
> additional feedback.
>
> I asked some people that happen to be here at OVSCON as well as some
> Red Hat and Intel people. I am also including some people that are involved
> in upstream vhost-user work in DPDK.
> So far, I have the following feedback with an attempt to condense
> feedback and to keep the list small. If I left out anything, let me know.
>
> In addition to the PVP tests done now with small packets.
>
> We should standardize on a basic limited set of sizes: 64, IMIX, 1518
> bytes (this can be extended if needed to the list defined in RFC-2455)
>
>
> Testpmd in guest is OK for now.
>
>
> I’d like to suggest to define/document the testpmd config used for
> testing: testpmd options and config, VM (vcpu, RAM).
> Having a testpmd image capable of auto-configuring itself on the virtual
> interfaces at init time would also be good to have.
>
>
>
> MK: vhost should be tested also with IRQ drivers, not only PMD, e.g.
> Linux guest with kernel IP routing. It’s done today in CSIT functional
> tests in VIRL (no testpmd there).
>
>
>
> Yes, as long as testPMD in guest is in the suite to maximize perf test
>
>
> Agree. testpmd is already used in csit perf tests with vhost.
>
>
>
> 1 Add multiple VMs (How many?)
>
>
>
>
>
> MK: For performance test, we should aim for a box-full, so for 1vCPU
> VMs fill up all cores :)
>
>
> This will depend on the testpmd settings (mostly number of vCPU).
> I’d suggest a minimum of 10 chains (10 x PVP) and 2 networks per chain.
>
>
>
> 2 Both multi-queue and single-queue
>
>
>
>
> MK: vhost single-queue for sure. vhost multi-queue seems to matter
> only to huge VMs that generate lots of traffic and coming close to
> overloading worker thread dealing with it.
>
> +1 for both single and multi-queue
>
>
> 3 Tests that cause the equivalent of multiple flows in OVS. Varying
> variety of traffic including layer 2 and layer 3 traffic.
>
>
>
>
> MK: Yes. Many flows is must.
>
> 4 Multiple IF's (Guest or Host or Both?)
>
>
>
>
> MK: What do you mean by multiple IF’s (interfaces)? With multiple VMs
> we surely have multiple vhost interfaces, minimum 2 vhost interfaces per
> VM. What matters IMV is the ratio and speed between: i) physical interfaces
> 10GE, 40GE; and ii) vhost inte

Re: [vpp-dev] vHost user test scenarios for CSIT

[Damon Wang]

Maciek, really thanks!

2016-11-30 12:36 GMT+08:00 Alec Hothan (ahothan) :

>
>
> Thanks for putting this together!
>
> A few preliminary general comments which we can discuss Wednesday.
>
>
>
> *2p1nic vs 1p1nic*:
>
> I know lots of vswitch benchmarks love pairing phys interfaces as it is
> easier to scale with more cores…
>
> But a lot of openstack deployments will come with 1 physical port for all
> the tenant traffic. In that case, looping traffic will come in through a
> vlan and go out to another vlan (vlan case) or come in through a vxlan
> tunnel and go out another vxlan tunnel (VxLAN overlay). The maximum
> throughput is 10G in and 10 G out, less wires on the compute nodes and less
> ports on the TOR.
>
>
>
> I’m not sure there are/will be many deployments with 2 independent
> physical ports for tenant traffic. At 30 nodes per rack that makes it 60
> ports per TOR just for the tenant traffic…. If you factor in bonding, that
> would make it 4 wires per compute node instead of 2.
>
> For those deployments that really need more than a 10G link, they might
> just use 1 40G link (or 2 bonded links) rather than 2 individual 10G links.
>
> Neutron supports multiple interfaces but it looks harder to manage (I
> guess you need to split your vlan range in 2 but in the case of vxlan
> overlay I’m not sure how that can be managed.
>
>
>
> I’d like to hear from others whether we should focus on 1p1nic first. It
> just seems to me this might be more representative of real NFV deployments
> than 2p1nic.
>
>
>
>
>
> *Test topologies of interest for OpenStack*:
>
>
>
> Could we mark all those test topologies that are applicable to OpenStack?
> Or more generally describe the use case for each test topology.
>
> I don’t think OpenStack will use any l2xc in vpp.
>
>
>
> OpenStack ML2/VPP/VLAN will use dot1q-l2bdbase-eth-2vhost-1vm
>
> OpenStack VPP/VxLAN will use ethip4vxlan-l2bdbase-eth-2vhost-1vm
>
>
>
> *VM images*:
>
>
>
> I think vsperf is using testpmd. I know some other perf teams use l2fwd.
>
> In our team, we use VPP l2 x-connect in the VM (what you call vswitch VPP)
> and evaluating testpmd and they don’t seem to differ to much in results.
> Seems like testpmd is the way to go for L2…
>
>
>
> Would also be good to have the VM image creation scripted – so that
> anybody can recreate them from scratch. We use DIB (openstack disk image
> builder) for creating our VM images but any other scripted solution should
> work.
>
> For easier reuse inside openstack, the config of the VM instance must be
> as automatic as possible. That is easy for the L2 case (just cross connect
> between the 2 virtual interfaces). For the L3 case, the L3 config should be
> done through config drive (not SSH).
>
>
>
>
>
> That’s it for now,
>
>
>
>   Alec
>
>
>
>
>
>
>
>
>
> *From: *"Maciek Konstantynowicz (mkonstan)" 
> *Date: *Tuesday, November 29, 2016 at 6:27 PM
> *To: *Thomas F Herbert , vpp-dev ,
> "csit-...@lists.fd.io" , "Pierre Pfister
> (ppfister)" , Andrew Theurer ,
> Douglas Shakshober , Rashid Khan ,
> Karl Rister , Irene Liew ,
> "Alec Hothan (ahothan)" , Damon Wang <
> damon.dev...@gmail.com>
> *Subject: *Re: [vpp-dev] vHost user test scenarios for CSIT
>
>
>
> All,
>
>
>
> Here is the first draft:
>
> https://wiki.fd.io/view/CSIT/vhostuser_test_scenarios
>
>
>
> I did my best to capture all inputs as per this thread. But it’s hardly
> readable yet - requires more TLC :)
>
> See what you think - feel free to add/edit things directly on FD.io
> <http://fd.io> wiki page.
>
>
>
> Suggest to discuss next steps on csit weekly call tomorrow, details here:
>
> https://wiki.fd.io/view/CSIT/Meeting
>
>
> -Maciek
>
>
>
> On 28 Nov 2016, at 07:37, Thomas F Herbert  wrote:
>
>
>
> All,
>
> At last week's CSIT meeting, Maciek (mkons...@cisco.com) offered to
> compile a summary suggestions on this mailing list.
>
>
>
> On 11/22/2016 11:34 AM, Pierre Pfister (ppfister) wrote:
>
> Hello Thomas,
>
>
>
> Sorry I haven't reached out faster, I was travelling.
>
>
>
> Please have a look at vppsb/vhost-test
>
> It includes a standalone script which provides VPP and VM configuration
> for PVP tests.
>
> - Runs testpmd in the VM
>
> - Supports various CPU configuration for VPP
>
> - Can run with or without gdb, debug or release
>
>
>
> Not committed yet:
>
> - Supports VM restart
>
> - Support for VPP restart
>
> - Support for multiple

[vpp-dev] vpp api's change

[Wang, Drenfong]

Hi vpp-dev
Vpp have changed it's api  string type many times, Could we keep it stable?
In the interface.api.h.vapi.h, the name_filter, interface_name, admin_up_down

-=-=-=-=-=-=-=-=-=-=-=-
Links: You receive all messages sent to this group.

View/Reply Online (#13898): https://lists.fd.io/g/vpp-dev/message/13898
Mute This Topic: https://lists.fd.io/mt/33136938/21656
Group Owner: vpp-dev+ow...@lists.fd.io
Unsubscribe: https://lists.fd.io/g/vpp-dev/unsub  [arch...@mail-archive.com]
-=-=-=-=-=-=-=-=-=-=-=-

[vpp-dev] vppctl start error

[Evan Wang]

Hi 

First, thank you to read my email.

 

I build vpp,(latest from master) and install on my Centos-7.4.

The SElinux is disabled.

 

I start the service of vpp, #systemctl start vpp

 

Then I run vppctl.

 

If I use the default /etc/vpp/startup.conf, the vppctl start correctly.

 

If I want to enable more dpdk rx/tx queue via updating startup.conf, and
then restart vpp service.

After that I run vppctl again, it failed with below log:

 

[root@h3c_skylake_8hdd vpp]# vppctl   

clib_socket_init: connect (fd 3, '/run/vpp/cli.sock'): Connection refused

 

 

would you like to tell me the similar issue appeared before? How to fix it?

 

Thank you!

-=-=-=-=-=-=-=-=-=-=-=-
Links: You receive all messages sent to this group.

View/Reply Online (#10501): https://lists.fd.io/g/vpp-dev/message/10501
Mute This Topic: https://lists.fd.io/mt/25673019/21656
Group Owner: vpp-dev+ow...@lists.fd.io
Unsubscribe: https://lists.fd.io/g/vpp-dev/unsub  [arch...@mail-archive.com]
-=-=-=-=-=-=-=-=-=-=-=-

答复: [vpp-dev] vppctl start error

[Evan Wang]

Hi Ed Warnicke

Thank you for the kindly reply, and you are right, I made mistake when changing 
the configuration file with a syntactically error.

Now it is ok.

 

Thx again.

 

发件人: Ed Warnicke [mailto:hagb...@gmail.com] 
发送时间: 2018年9月14日 20:20
收件人: Evan Wang 
抄送: vpp-dev@lists.fd.io
主题: Re: [vpp-dev] vppctl start error

 

Could you attach your new startup.conf file?

 

I can't be completely certain from the info provided... but it looks a lot like 
vpp didn't really start... which might be because your edit of the startup.conf 
file produced a syntactically invalid file.

 

You might also use systemctl to grab any logs from the startup

 

Another quick way to check is to simply try manually

 

/usr/bin/vpp -c /etc/vpp/startup.conf

 

That way you will see the output  from startup, and can find out if vpp is 
failing to start due to issues with the config file.

 

Ed

 

On Fri, Sep 14, 2018 at 6:30 AM Evan Wang mailto:evan.w...@ucloud.cn> > wrote:

Hi 

First, thank you to read my email.

 

I build vpp,(latest from master) and install on my Centos-7.4.

The SElinux is disabled.

 

I start the service of vpp, #systemctl start vpp

 

Then I run vppctl.

 

If I use the default /etc/vpp/startup.conf, the vppctl start correctly.

 

If I want to enable more dpdk rx/tx queue via updating startup.conf, and then 
restart vpp service.

After that I run vppctl again, it failed with below log:

 

[root@h3c_skylake_8hdd vpp]# vppctl   

clib_socket_init: connect (fd 3, '/run/vpp/cli.sock'): Connection refused

 

 

would you like to tell me the similar issue appeared before? How to fix it?

 

Thank you!

-=-=-=-=-=-=-=-=-=-=-=-
Links: You receive all messages sent to this group.

View/Reply Online (#10501): https://lists.fd.io/g/vpp-dev/message/10501
Mute This Topic: https://lists.fd.io/mt/25673019/464962
Group Owner: vpp-dev+ow...@lists.fd.io <mailto:vpp-dev%2bow...@lists.fd.io> 
Unsubscribe: https://lists.fd.io/g/vpp-dev/unsub  [hagb...@gmail.com 
<mailto:hagb...@gmail.com> ]
-=-=-=-=-=-=-=-=-=-=-=-

-=-=-=-=-=-=-=-=-=-=-=-
Links: You receive all messages sent to this group.

View/Reply Online (#10537): https://lists.fd.io/g/vpp-dev/message/10537
Mute This Topic: https://lists.fd.io/mt/25745519/21656
Group Owner: vpp-dev+ow...@lists.fd.io
Unsubscribe: https://lists.fd.io/g/vpp-dev/unsub  [arch...@mail-archive.com]
-=-=-=-=-=-=-=-=-=-=-=-

Re: [vpp-dev] [sweetcomb-dev] cannot find vom library

[Wang, Drenfong]

Hi Dave Damarion and Neale


In this file(the rpm package) vpp\extras\rpm\vpp.spec

I found there is no vom in lib package


Is this mean the vom will not be put into rpm package or I am wrong ?


Thanks

Best Regards
Junfeng Wang

Network Platforms Group  |  +86 021 2207 1161  | the 4 floor, No. 1 Building, 
No. 588 Zi Xing Rd., Zizhu Science Park, Shanghai, 200241, PRC

From: sweetcomb-...@lists.fd.io [mailto:sweetcomb-...@lists.fd.io] On Behalf Of 
Andrej Kozemcák
Sent: Monday, February 11, 2019 2:50 PM
To: sweetcomb-...@lists.fd.io
Subject: Re: [sweetcomb-dev] cannot find vom library


Hi Junfeng,



Yes, you right.



That should be that problem.



Theoretical you can have an old version of VPP which don't contain VOM library.

If that right, this hot path can help you, but this path fix problem only if 
you use the old version of VPP



Best Regards,

Andrej Kozemcak


Od: Drenfong Wong mailto:drenfong.w...@intel.com>>
Odoslané: pondelok, 11. februára 2019 7:30:47
Komu: sweetcomb-...@lists.fd.io<mailto:sweetcomb-...@lists.fd.io>
Predmet: Re: [sweetcomb-dev] cannot find vom library



Hi Andrej

When I use # find / -name "*vom*"

On ubuntu:
can found /usr/include/vom   /usr/lib/x86_64-linux-gnu/libvom.so

On centos:
Can only found /usr/include/vom

So this maybe the vpp's installation problem

Thanks

Best Regards
Junfeng Wang

Network Platforms Group  |  +86 021 2207 1161  | the 4 floor, No. 1 Building, 
No. 588 Zi Xing Rd., Zizhu Science Park, Shanghai, 200241, PRC

From: sweetcomb-...@lists.fd.io<mailto:sweetcomb-...@lists.fd.io> 
[mailto:sweetcomb-...@lists.fd.io] On Behalf Of Drenfong Wong
Sent: Monday, February 11, 2019 2:21 PM
To: andrej.kozem...@pantheon.tech<mailto:andrej.kozem...@pantheon.tech>
Cc: sweetcomb-...@lists.fd.io<mailto:sweetcomb-...@lists.fd.io>
Subject: [sweetcomb-dev] cannot find vom library



Hi Andrej

Could you help to debug this issue, when compile plugins on centos, the vom 
library cannot be found

https://logs.fd.io/production/vex-yul-rot-jenkins-1/sweetcomb-merge-1901-centos7/26/console-timestamp.log.gz


Thanks

Best Regards
Junfeng Wang

Network Platforms Group  |  +86 021 2207 1161  | the 4 floor, No. 1 Building, 
No. 588 Zi Xing Rd., Zizhu Science Park, Shanghai, 200241, PRC

_._,_._,_

Links:

You receive all messages sent to this group.

View/Reply Online (#179)<https://lists.fd.io/g/sweetcomb-dev/message/179> | 
Reply To 
Group<mailto:sweetcomb-...@lists.fd.io?subject=Re:%20Re%3A%20%5Bsweetcomb-dev%5D%20cannot%20find%20vom%20library>
 | Reply To 
Sender<mailto:andrej.kozem...@pantheon.tech?subject=Private:%20Re:%20Re%3A%20%5Bsweetcomb-dev%5D%20cannot%20find%20vom%20library>
 | Mute This Topic<https://lists.fd.io/mt/29732273/1274056> | New 
Topic<https://lists.fd.io/g/sweetcomb-dev/post>

Your Subscription<https://lists.fd.io/g/sweetcomb-dev/editsub/1274056> | 
Contact Group Owner<mailto:sweetcomb-dev+ow...@lists.fd.io> | 
Unsubscribe<https://lists.fd.io/g/sweetcomb-dev/leave/3374981/1134849542/xyzzy> 
[drenfong.w...@intel.com]
_._,_._,_
-=-=-=-=-=-=-=-=-=-=-=-
Links: You receive all messages sent to this group.

View/Reply Online (#12224): https://lists.fd.io/g/vpp-dev/message/12224
Mute This Topic: https://lists.fd.io/mt/29732429/21656
Group Owner: vpp-dev+ow...@lists.fd.io
Unsubscribe: https://lists.fd.io/g/vpp-dev/unsub  [arch...@mail-archive.com]
-=-=-=-=-=-=-=-=-=-=-=-

Re: [vpp-dev] [sweetcomb-dev] cannot find vom library

[Wang, Drenfong]

Hi Mohsin

It seem that the vom will not be put into vpp-lib(I found libvom.so in 
vpp-lib_19.04-rc0~27-ge67c1d8~b6157_amd64.deb<https://packagecloud.io/fdio/master/packages/ubuntu/xenial/vpp-lib_19.04-rc0~27-ge67c1d8~b6157_amd64.deb>,
 and after 19.04-rc0-27, the vpp-lib was put into vpp-dev and libvppinfra (no 
libvom.so))


Thanks for you vom deb package patch, this help a lot



From: Wang, Drenfong
Sent: Tuesday, February 12, 2019 4:29 PM
To: 'syka...@cisco.com' 
Subject: RE: [vpp-dev] [sweetcomb-dev] cannot find vom library

Hi Mohsin

Thanks for your patch.
I will test this manually.


Best Regards
Junfeng Wang

Network Platforms Group  |  +86 021 2207 1161  | the 4 floor, No. 1 Building, 
No. 588 Zi Xing Rd., Zizhu Science Park, Shanghai, 200241, PRC

From: Mohsin Kazmi (sykazmi) [mailto:syka...@cisco.com]
Sent: Monday, February 11, 2019 6:55 PM
To: Wang, Drenfong mailto:drenfong.w...@intel.com>>; 
sweetcomb-...@lists.fd.io<mailto:sweetcomb-...@lists.fd.io>; 
vpp-dev@lists.fd.io<mailto:vpp-dev@lists.fd.io>; 
d...@barachs.net<mailto:d...@barachs.net>; Damjan Marion (damarion) 
mailto:damar...@cisco.com>>; Neale Ranns (nranns) 
mailto:nra...@cisco.com>>
Cc: Ni, Hongjun mailto:hongjun...@intel.com>>
Subject: Re: [vpp-dev] [sweetcomb-dev] cannot find vom library


Hi,



Does this fix your issue?



https://gerrit.fd.io/r/#/c/17459/​



-br,

Mohsin


From: vpp-dev@lists.fd.io<mailto:vpp-dev@lists.fd.io> 
mailto:vpp-dev@lists.fd.io>> on behalf of Wang, Drenfong 
mailto:drenfong.w...@intel.com>>
Sent: Monday, February 11, 2019 8:08 AM
To: sweetcomb-...@lists.fd.io<mailto:sweetcomb-...@lists.fd.io>; 
vpp-dev@lists.fd.io<mailto:vpp-dev@lists.fd.io>; 
d...@barachs.net<mailto:d...@barachs.net>; Damjan Marion (damarion); Neale 
Ranns (nranns)
Cc: Ni, Hongjun
Subject: Re: [vpp-dev] [sweetcomb-dev] cannot find vom library


Hi Dave Damarion and Neale


In this file(the rpm package) vpp\extras\rpm\vpp.spec

I found there is no vom in lib package


Is this mean the vom will not be put into rpm package or I am wrong ?


Thanks

Best Regards
Junfeng Wang

Network Platforms Group  |  +86 021 2207 1161  | the 4 floor, No. 1 Building, 
No. 588 Zi Xing Rd., Zizhu Science Park, Shanghai, 200241, PRC

From: sweetcomb-...@lists.fd.io<mailto:sweetcomb-...@lists.fd.io> 
[mailto:sweetcomb-...@lists.fd.io] On Behalf Of Andrej Kozemcák
Sent: Monday, February 11, 2019 2:50 PM
To: sweetcomb-...@lists.fd.io<mailto:sweetcomb-...@lists.fd.io>
Subject: Re: [sweetcomb-dev] cannot find vom library


Hi Junfeng,



Yes, you right.



That should be that problem.



Theoretical you can have an old version of VPP which don't contain VOM library.

If that right, this hot path can help you, but this path fix problem only if 
you use the old version of VPP



Best Regards,

Andrej Kozemcak


Od: Drenfong Wong mailto:drenfong.w...@intel.com>>
Odoslané: pondelok, 11. februára 2019 7:30:47
Komu: sweetcomb-...@lists.fd.io<mailto:sweetcomb-...@lists.fd.io>
Predmet: Re: [sweetcomb-dev] cannot find vom library



Hi Andrej

When I use # find / -name “*vom*”

On ubuntu:
can found /usr/include/vom   /usr/lib/x86_64-linux-gnu/libvom.so

On centos:
Can only found /usr/include/vom

So this maybe the vpp’s installation problem

Thanks

Best Regards
Junfeng Wang

Network Platforms Group  |  +86 021 2207 1161  | the 4 floor, No. 1 Building, 
No. 588 Zi Xing Rd., Zizhu Science Park, Shanghai, 200241, PRC

From: sweetcomb-...@lists.fd.io<mailto:sweetcomb-...@lists.fd.io> 
[mailto:sweetcomb-...@lists.fd.io] On Behalf Of Drenfong Wong
Sent: Monday, February 11, 2019 2:21 PM
To: andrej.kozem...@pantheon.tech<mailto:andrej.kozem...@pantheon.tech>
Cc: sweetcomb-...@lists.fd.io<mailto:sweetcomb-...@lists.fd.io>
Subject: [sweetcomb-dev] cannot find vom library



Hi Andrej

Could you help to debug this issue, when compile plugins on centos, the vom 
library cannot be found

https://logs.fd.io/production/vex-yul-rot-jenkins-1/sweetcomb-merge-1901-centos7/26/console-timestamp.log.gz


Thanks

Best Regards
Junfeng Wang

Network Platforms Group  |  +86 021 2207 1161  | the 4 floor, No. 1 Building, 
No. 588 Zi Xing Rd., Zizhu Science Park, Shanghai, 200241, PRC

_._,_._,_

Links:

You receive all messages sent to this group.

View/Reply Online (#179)<https://lists.fd.io/g/sweetcomb-dev/message/179> | 
Reply To 
Group<mailto:sweetcomb-...@lists.fd.io?subject=Re:%20Re%3A%20%5Bsweetcomb-dev%5D%20cannot%20find%20vom%20library>
 | Reply To 
Sender<mailto:andrej.kozem...@pantheon.tech?subject=Private:%20Re:%20Re%3A%20%5Bsweetcomb-dev%5D%20cannot%20find%20vom%20library>
 | Mute This Topic<https://lists.fd.io/mt/29732273/1274056> | New 
Topic<https://lists.fd.io/g/sweetcomb-dev/post>

Your Subscription<https:

[vpp-dev] Questions about no-multi-seg option in startup.conf

[Jieqiang Wang]

7;:
vpp# sh hardware-interfaces
  NameIdx   Link  Hardware
eth0   1 up   eth0
  Link speed: 40 Gbps
  Ethernet address 3c:fd:fe:bb:d4:10
  Intel X710/XL710 Family
carrier up full duplex mtu 9206
flags: admin-up pmd maybe-multiseg rx-ip4-cksum
Devargs:
rx: queues 1 (max 320), desc 1024 (min 64 max 4096 align 32)
tx: queues 2 (max 320), desc 1024 (min 64 max 4096 align 32)
pci: device 8086:1583 subsystem 8086:0001 address 0001:01:00.00 numa 0
max rx packet len: 9728
promiscuous: unicast off all-multicast on
vlan offload: strip off filter off qinq off
rx offload avail:  vlan-strip ipv4-cksum udp-cksum tcp-cksum qinq-strip
   outer-ipv4-cksum vlan-filter vlan-extend jumbo-frame
   scatter keep-crc rss-hash
rx offload active: ipv4-cksum jumbo-frame scatter
tx offload avail:  vlan-insert ipv4-cksum udp-cksum tcp-cksum sctp-cksum
   tcp-tso outer-ipv4-cksum qinq-insert vxlan-tnl-tso
   gre-tnl-tso ipip-tnl-tso geneve-tnl-tso multi-segs
   mbuf-fast-free
tx offload active: multi-segs
rss avail: ipv4-frag ipv4-tcp ipv4-udp ipv4-sctp ipv4-other 
ipv6-frag
   ipv6-tcp ipv6-udp ipv6-sctp ipv6-other l2-payload
rss active:none
tx burst mode: Scalar
rx burst mode: Vector Neon Scattered

So I am wondering why no-multi-seg option will change the vector rates 
highlight in red above? Is this phenomenon expected?
I also saw performance drop when no-multi-seg option was not set in 
startup.conf when I sent small packets(like 64 bytes) as traffic input(simple 
IPv4 routing test case with 1 flow). How will VPP behave differently if 
no-multi-seg option is set?

Look forward to getting your feedback.

Thanks,
Jieqiang Wang

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.

-=-=-=-=-=-=-=-=-=-=-=-
Links: You receive all messages sent to this group.
View/Reply Online (#18488): https://lists.fd.io/g/vpp-dev/message/18488
Mute This Topic: https://lists.fd.io/mt/79516636/21656
Group Owner: vpp-dev+ow...@lists.fd.io
Unsubscribe: https://lists.fd.io/g/vpp-dev/unsub [arch...@mail-archive.com]
-=-=-=-=-=-=-=-=-=-=-=-

Re: [vpp-dev] Questions about no-multi-seg option in startup.conf

[Jieqiang Wang]

Hi Ben,

Thank you for your quick reply and your insightful explanation. Your answers 
are really helpful.

Thanks,
Jieqiang Wang
-Original Message-
From: Benoit Ganne (bganne) 
Sent: Friday, January 8, 2021 4:20 PM
To: Jieqiang Wang ; vpp-dev 
Cc: Lijian Zhang ; Tianyu Li ; 
Govindarajan Mohandoss ; nd 
Subject: RE: Questions about no-multi-seg option in startup.conf

Yes this is expected: 'no-multi-seg' tells DPDK that all packets will consist 
of 1 and only 1 buffer (no chained buffers). Lots of DPDK PMDs supports 
vectorization (SSE, NEON...) only for this simpler case. When you set this 
option, DPDK can select the vectorized PMD instead of the more generic, 
non-vectorized (and hence slower) version.
You can see in the 'show hardware' output that in case of 'no-multi-seg' you 
get the 'Vector NEON' for both RX and TX, whereas you only get RX and not TX 
otherwise. So in the 2nd case, the DPDK TX is slower - this is reflected by the 
'show run' output, where in the 2nd case the cycles/packet for 'eth0-tx' and 
'eth1-tx' grows from 0.8 to 1.8. As TX cost is bigger, VPP is slower, process 
bigger vectors and less pps overall.

ben

> -Original Message-
> From: vpp-dev@lists.fd.io  On Behalf Of Jieqiang
> Wang
> Sent: vendredi 8 janvier 2021 04:26
> To: vpp-dev 
> Cc: Lijian Zhang ; Tianyu Li
> ; Govindarajan Mohandoss
> ; nd 
> Subject: [vpp-dev] Questions about no-multi-seg option in startup.conf
>
> Hi VPP dev,
>
>
>
> I was trying to do some benchmarking on VPP and found out no-multi-seg
> option in startup.conf will have impact on both the performance and
> how the runtime shows.
>
> The VPP version is v21.01-rc0~547-gf0419a0c8, the DPDK version is DPDK
> 20.11.0.
>
> With no-multi-seg option set in the startup.conf, the runtime shows
> like the following:
>
> Thread 1 vpp_wk_0 (lcore 2)
>
> Time 1.1, 10 sec internal node vector rate 85.33 loops/sec 97035.37
>
>   vector rates in 1.2537e7, out 1.2537e7, drop 0.e0, punt 0.e0
>
>  Name State Calls  Vectors
> Suspends Clocks   Vectors/Call
>
> dpdk-input   polling11252714403456
> 0 9.34e-1  128.00
>
> eth0-output  active 112527 7201728
> 0 2.04e-1   64.00
>
> eth0-tx  active 112527 7201728
> 0 7.86e-1   64.00
>
> eth1-output  active 112527 7201728
> 0 1.91e-1   64.00
>
> eth1-tx  active 112527 7201728
> 0 7.93e-1   64.00
>
> ethernet-input   active 22505414403456
> 0 5.65e-1   64.00
>
> ip4-input-no-checksumactive 11252714403456
> 0 3.83e-1  128.00
>
> ip4-lookup   active 11252714403456
> 0 5.34e-1  128.00
>
> ip4-rewrite  active 11252714403456
> 0 5.73e-1  128.00
>
> unix-epoll-input polling   110   0
> 0  2.84e10.00
>
>
>
> Output for command 'show hardware-interfaces':
>
> vpp# sh hardware-interfaces
>
>   NameIdx   Link  Hardware
>
> eth0   1 up   eth0
>
>   Link speed: 40 Gbps
>
>   Ethernet address 3c:fd:fe:bb:d4:10
>
>   Intel X710/XL710 Family
>
> carrier up full duplex mtu 9206
>
> flags: admin-up pmd rx-ip4-cksum
>
> Devargs:
>
> rx: queues 1 (max 320), desc 1024 (min 64 max 4096 align 32)
>
> tx: queues 2 (max 320), desc 1024 (min 64 max 4096 align 32)
>
> pci: device 8086:1583 subsystem 8086:0001 address 0001:01:00.00
> numa 0
>
> max rx packet len: 9728
>
> promiscuous: unicast off all-multicast on
>
> vlan offload: strip off filter off qinq off
>
> rx offload avail:  vlan-strip ipv4-cksum udp-cksum tcp-cksum qinq-
> strip
>
>outer-ipv4-cksum vlan-filter vlan-extend jumbo-
> frame
>
>scatter keep-crc rss-hash
>
> rx offload active: ipv4-cksum
>
> tx offload avail:  vlan-insert ipv4-cksum udp-cksum tcp-cksum
> sctp- cksum
>
>tcp-tso outer-ipv4-cksum qinq-insert
> vxlan-tnl-tso
>
>gre-tnl-tso ipip-tnl-tso geneve-tnl-tso
> multi-segs
>
>mbuf-fast-free
>
>

Re: [vpp-dev] a compiling error on vpp master

[Jieqiang Wang]

Hi Xiaoming,

We met the same problem on CentOS 7 and proposed a 
patch(https://gerrit.fd.io/r/c/vpp/+/31421) to resolve this issue.
It looks to me that the root cause is the different declaration scheme for 
struct udphdr in musl libc and glibc. In older glibc(less than 2.19), field 
names of uh_sport/uh_dport are only used if macro __FAVOR_BSD is defined 
according to the header file. But we won’t encounter this issue with musl libc 
if macro _GNU_SOURCE is defined like Ben mentioned in 
https://gerrit.fd.io/r/c/vpp/+/30790/8/src/vppinfra/unix-formats.c#320 .
Anyway, this is just my thought and needs confirmation from VPP maintainers. 
But you can check your glibc version and try if this patch works for you.

Thanks,
Jieqiang Wang
From: vpp-dev@lists.fd.io  On Behalf Of jiangxiaoming via 
lists.fd.io
Sent: Thursday, March 4, 2021 10:25 AM
To: vpp-dev@lists.fd.io
Subject: [vpp-dev] a compiling error on vpp master


Hi guys:
I have a compiling error on vpp master, is thre anyone have the save 
problem?

VPP library version : 21.06

GIT toplevel dir: /home/dev/code/vpp

Build type  : debug

C flags : -Wno-address-of-packed-member -g -fPIC -Werror -Wall 
-march=corei7 -mtune=corei7-avx -O0 -DCLIB_DEBUG -fstack-protector -fno-common

Linker flags (apps) : -z execstack

Linker flags (libs) : -z execstack

Host processor  : x86_64

Target processor: x86_64

Prefix path : 
/opt/vpp/external/x86_64;/home/dev/code/vpp/build-root/install-vpp_debug-native/external

Install prefix  : /home/dev/code/vpp/build-root/install-vpp_debug-native/vpp

-- Configuring done

-- Generating done

-- Build files have been written to: 
/home/dev/code/vpp/build-root/build-vpp_debug-native/vpp

 Building vpp in /home/dev/code/vpp/build-root/build-vpp_debug-native/vpp 


[554/2614] Building C object vppinfra/CMakeFiles/vppinfra.dir/unix-formats.c.o

FAILED: vppinfra/CMakeFiles/vppinfra.dir/unix-formats.c.o

ccache /opt/rh/devtoolset-9/root/bin/cc -Dvppinfra_EXPORTS 
-I/home/dev/code/vpp/src -I. -Iinclude -Wno-address-of-packed-member -g -fPIC 
-Werror -Wall -march=corei7 -mtune=corei7-avx  -O0 -DCLIB_DEBUG 
-fstack-protector -fno-common  -fPIC   -fvisibility=hidden -MD -MT 
vppinfra/CMakeFiles/vppinfra.dir/unix-formats.c.o -MF 
vppinfra/CMakeFiles/vppinfra.dir/unix-formats.c.o.d -o 
vppinfra/CMakeFiles/vppinfra.dir/unix-formats.c.o   -c 
/home/dev/code/vpp/src/vppinfra/unix-formats.c

/home/dev/code/vpp/src/vppinfra/unix-formats.c: In function 
‘format_udp4_packet’:

/home/dev/code/vpp/src/vppinfra/unix-formats.c:319:19: error: ‘struct udphdr’ 
has no member named ‘uh_sport’

  319 |   u16 source = udp->uh_sport;

  |   ^~

/home/dev/code/vpp/src/vppinfra/unix-formats.c:320:17: error: ‘struct udphdr’ 
has no member named ‘uh_dport’

  320 |   u16 dest = udp->uh_dport;

  | ^~

[565/2614] Building C object svm/CMakeFiles/svm.dir/fifo_segment.c.o

ninja: build stopped: subcommand failed.

make[1]: *** [vpp-build] Error 1

make[1]: Leaving directory `/home/dev/code/vpp/build-root'

make: *** [build] Error 2

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.

-=-=-=-=-=-=-=-=-=-=-=-
Links: You receive all messages sent to this group.
View/Reply Online (#18847): https://lists.fd.io/g/vpp-dev/message/18847
Mute This Topic: https://lists.fd.io/mt/81068821/21656
Group Owner: vpp-dev+ow...@lists.fd.io
Unsubscribe: https://lists.fd.io/g/vpp-dev/unsub [arch...@mail-archive.com]
-=-=-=-=-=-=-=-=-=-=-=-

[vpp-dev] Is there a convenient way to name generated files when making VPP, such as rpm

[Tao Wang]

Hi,


I got VPP code via "git clone -b v20.09 https://github.com/FDio/vpp.git";.
When "make pkg-rpm", the generated VPP rpms are named by git tag v20.09 and 
looked like "vpp-20.09-release.x86_64.rpm".

We are using VPP as a sub-project in our project and we have our own git tag 
format not like VPP's format "vXX.XX-X".
If using our tag format for the whole project (include VPP as a sub-folder), it 
would cause VPP make failed.
Is there a convenient way to purify VPP original naming processing (most 
related with git operations in Makefile) to customize VPP generated files' name 
as we want?
I have found all the naming processing codes in VPP Makefile and I only know to 
modify them one by one to achieve the goal of customizing VPP generated files' 
name.
So as a VPP 3rd-party user like us, I want to know if there is a more 
convenient and elegant way to name VPP generated files?

Thanks!


Best Regards,
Tao Wang

-=-=-=-=-=-=-=-=-=-=-=-
Links: You receive all messages sent to this group.
View/Reply Online (#18946): https://lists.fd.io/g/vpp-dev/message/18946
Mute This Topic: https://lists.fd.io/mt/81394971/21656
Group Owner: vpp-dev+ow...@lists.fd.io
Unsubscribe: https://lists.fd.io/g/vpp-dev/unsub [arch...@mail-archive.com]
-=-=-=-=-=-=-=-=-=-=-=-

Re: [vpp-dev] Is there a convenient way to name generated files when making VPP, such as rpm

[Tao Wang]

Hi, Hemant


Thanks for your feedback.
It's indeed just like what you described to build vpp.
I know how to build vpp.
Why to choose vpp 20.09 not the latest version? It's just because our project 
is set up based on vpp 20.09.

For our case,
1) our project ABC dir hierarchy is like this:
├── ABC
│   ├── vpp
│   │   ├── xxx
│   │   └── ...
│   ├── apps
│   │   ├── xxx
│   │   │   ├── xxx
│   │   │   │   ├── xxx
│   │   │   │   ├── ...
2) when building vpp rpm in the sub-folder "vpp", vpp official build processing 
captures vpp version info from git tag info via commands "git describe".
I regard that is the vpp official naming processing for vpp generated files, 
such as rpms.
3) vpp has its own tag format like this vXX.XX.X-rcX, eg. v20.09.1-rc0
4) but our project "ABC" has our own tag format like this ABC-XX.XX.XX-rcX, eg. 
ABC-21.04.01-rc0
5) so, when applying the tag ABC-21.04.01-rc0 for the whole project "ABC", it 
means the tag format is applied to the sub-folder "vpp" also
6) it causes the tag of our project's "vpp" does not comply with vpp official 
tag format rule and it causes the vpp "make pkg-rpm" failed
7) so, is there a convenient and elegant way to eliminate vpp official naming 
processing to customize vpp generated files' name as we demand
8) I have found all the naming processing codes in vpp Makefile and my solution 
for my case is just to modify them one by one to customize the generated vpp 
rpm file name
9) but I regard there are a lot of vpp 3rd-party users like us, so I want to 
know how other vpp 3rd-party users handle the same vpp usage scenario as ours

Thanks in advance!


Best Regards,
Tao Wang
____
From: hem...@mnkcg.com
Sent: Wednesday, March 17, 2021 12:36 PM
To: Tao Wang; vpp-dev@lists.fd.io
Cc: zhiyong.y...@intel.com; zhongbao@intel.com; shaohe.f...@intel.com; 
Jinhui Li; Cong Hu
Subject: RE: [vpp-dev] Is there a convenient way to name generated files when 
making VPP, such as rpm


Github is just a mirror which may be up to date with latest VPP software and 
also github is not the correct repo to work with VPP software.  Please use the 
command below to get latest VPP software.



git clone https://gerrit.fd.io/r/vpp



In vpp root directory, type “sudo make install-deps”.  Thereafter, type “sudo 
make install-ext-deps”.  Then, you can build an image using “sudo make 
build-release”.  The README has a note with alternative way to build using 
vagrant.



Hemant



From: vpp-dev@lists.fd.io  On Behalf Of Tao Wang
Sent: Tuesday, March 16, 2021 10:51 PM
To: vpp-dev@lists.fd.io
Cc: Yang, Zhiyong ; Shi, Zhongbao 
; Feng, Shaohe ; Jinhui Li 
; Cong Hu 
Subject: [vpp-dev] Is there a convenient way to name generated files when 
making VPP, such as rpm



Hi,





I got VPP code via "git clone -b v20.09 https://github.com/FDio/vpp.git";.

When "make pkg-rpm", the generated VPP rpms are named by git tag v20.09 and 
looked like "vpp-20.09-release.x86_64.rpm".



We are using VPP as a sub-project in our project and we have our own git tag 
format not like VPP's format "vXX.XX-X".

If using our tag format for the whole project (include VPP as a sub-folder), it 
would cause VPP make failed.

Is there a convenient way to purify VPP original naming processing (most 
related with git operations in Makefile) to customize VPP generated files' name 
as we want?

I have found all the naming processing codes in VPP Makefile and I only know to 
modify them one by one to achieve the goal of customizing VPP generated files' 
name.

So as a VPP 3rd-party user like us, I want to know if there is a more 
convenient and elegant way to name VPP generated files?



Thanks!





Best Regards,

Tao Wang

-=-=-=-=-=-=-=-=-=-=-=-
Links: You receive all messages sent to this group.
View/Reply Online (#18963): https://lists.fd.io/g/vpp-dev/message/18963
Mute This Topic: https://lists.fd.io/mt/81394971/21656
Group Owner: vpp-dev+ow...@lists.fd.io
Unsubscribe: https://lists.fd.io/g/vpp-dev/unsub [arch...@mail-archive.com]
-=-=-=-=-=-=-=-=-=-=-=-

[vpp-dev] Enable DPDK tx offload flag mbuf-fast-free on VPP vector mode

[Jieqiang Wang]

Hi VPP maintainers,



Recently VPP has upgraded the DPDK version to DPDK-21.08, which includes two 
optimization patches[1][2] from Arm DPDK team. With the mbuf-fast-free flag, 
the two patches add code segment to accelerate mbuf free in PMD TX path for 
i40e driver, which shows quite obvious performance improvement from DPDK L3FWD 
benchmarking results.



I tried to verify the benefits that those optimization patches can bring up to 
VPP, but find out that mbuf-fast-free flag is not enabled in VPP+DPDK by 
default.

Applying DPDK mbuf-fast-free has some constraints, e.g,

  *   mbufs to be freed should be coming from the same mempool
  *   ref_cnt == 1 always in mbuf meta-data when user apps call DPDK 
rte_eth_tx_burst ()
  *   No TX checksum offload
  *   No jumble frame

But VPP vector mode(set by adding ‘no-tx-checksum-offload’ and ‘no-multi-seg’ 
parameters in dpdk section of the startup.conf) seems to satisfy all the 
requirements. So I made a few code changes shown as below to set mbuf-fast-free 
flag by default in VPP vector mode and did some benchmarking for IPv4 routing 
test cases with 1 flow/10k flows. The benchmarking results show both throughput 
improvement and CPU cycles saved regarding DPDK transmit function.



So any thought on enabling mbuf-fast-free tx offload flag in VPP vector mode?  
Any feedback is welcome :)



Code Changes:



diff --git a/src/plugins/dpdk/device/init.c b/src/plugins/dpdk/device/init.c

index f7c1cc106..0fbdd2317 100644

--- a/src/plugins/dpdk/device/init.c

+++ b/src/plugins/dpdk/device/init.c

@@ -398,6 +398,8 @@ dpdk_lib_init (dpdk_main_t * dm)

  xd->port_conf.rxmode.offloads |= DEV_RX_OFFLOAD_SCATTER;

  xd->flags |= DPDK_DEVICE_FLAG_MAYBE_MULTISEG;

}

+  if (dm->conf->no_multi_seg && dm->conf->no_tx_checksum_offload)

+   xd->port_conf.txmode.offloads |= DEV_TX_OFFLOAD_MBUF_FAST_FREE;



   xd->tx_q_used = clib_min (dev_info.max_tx_queues, tm->n_vlib_mains);



Benchmark Results:



1 flow, bidirectional

Throughput(Mpps):



Original

Patched

Ratio

N1SDP

11.62

12.44

+7.06%

ThunderX2

9.52

10.16

+6.30%

Dell 8268

17.82

18.20

+2.13%



CPU cycles overhead for DPDK transmit function(recorded by Perf tools):



Original

Patched

Ratio

N1SDP

13.08%

5.53%

-7.55%

ThunderX2

11.01%

6.68%

-4.33%

Dell 8268

10.78%

7.35%

-3.43%



10k flows, bidirectional

Throughput(Mpps):



Original

Patched

Ratio

N1SDP

8.48

9.0

+6.13%

ThunderX2

8.84

9.26

+4.75%

Dell 8268

15.04

15.40

+2.39%



CPU cycles overhead for DPDK transmit function(recorded by Perf tools):



Original

Patched

Ratio

N1SDP

10.58%

4.54%

-6.04%

ThunderX2

12.92%

6.63%

-6.29%

Dell 8268

10.36%

7.97%

-2.39%



[1] 
http://git.dpdk.org/dpdk/commit/?h=v21.08-rc1&id=be8ff6210851fdacbe00033259b7dc5426e95589

[2] 
http://git.dpdk.org/dpdk/commit/?h=v21.08-rc1&id=95e7bb6a5fc9e371e763b11ec15786e4d574ef8e



Best Regards,

Jieqiang Wang

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.

-=-=-=-=-=-=-=-=-=-=-=-
Links: You receive all messages sent to this group.
View/Reply Online (#20152): https://lists.fd.io/g/vpp-dev/message/20152
Mute This Topic: https://lists.fd.io/mt/85669132/21656
Group Owner: vpp-dev+ow...@lists.fd.io
Unsubscribe: https://lists.fd.io/g/vpp-dev/unsub [arch...@mail-archive.com]
-=-=-=-=-=-=-=-=-=-=-=-

Re: [vpp-dev] Enable DPDK tx offload flag mbuf-fast-free on VPP vector mode

[Jieqiang Wang]

Hi Ben,

Thanks for your quick feedback. A few comments inline.

Best Regards,
Jieqiang Wang

-Original Message-
From: Benoit Ganne (bganne) 
Sent: Friday, September 17, 2021 3:34 PM
To: Jieqiang Wang ; vpp-dev 
Cc: Lijian Zhang ; Honnappa Nagarahalli 
; Govindarajan Mohandoss 
; Ruifeng Wang ; Tianyu 
Li ; Feifei Wang ; nd 
Subject: RE: Enable DPDK tx offload flag mbuf-fast-free on VPP vector mode

Hi Jieqiang,

This looks like an interesting optimization but you need to check that the 
'mbufs to be freed should be coming from the same mempool' rule holds true. 
This won't be the case on NUMA systems (VPP creates 1 buffer pool per NUMA).
This should be easy to check with eg. 'vec_len (vm->buffer_main->buffer_pools) 
== 1'.
>>> Jieqiang: That's a really good point here. Like you said, it holds true on 
>>> SMP systems and we can check by if the numbers of buffer pool equal to 1. 
>>> But I am wondering that is this check too strict? If the worker CPUs and 
>>> NICs used reside in the same NUMA node, I think mbufs come from the same 
>>> mempool and we still meet the requirement here.  What do you think?

For the rest, I think we do not use DPDK mbuf refcounting at all as we maintain 
our own anyway, but someone more knowledgeable than me should confirm.
>>> Jieqiang: This echoes with the experiments(IPv4 multicasting and L2 flood) 
>>> I have done. All the mbufs in the two test cases are copied instead of ref 
>>> counting. But this also needs double-check from VPP experts like you 
>>> mentioned.

I'd be curious to see if we can measure a real performance difference in CSIT.
>>> Jieqiang: Let me trigger some performance test cases in CSIT and come back 
>>> to you with related performance figures.

Best
ben

> -Original Message-
> From: vpp-dev@lists.fd.io  On Behalf Of Jieqiang
> Wang
> Sent: vendredi 17 septembre 2021 06:07
> To: vpp-dev 
> Cc: Lijian Zhang ; Honnappa Nagarahalli
> ; Govindarajan Mohandoss
> ; Ruifeng Wang ;
> Tianyu Li ; Feifei Wang ; nd
> 
> Subject: [vpp-dev] Enable DPDK tx offload flag mbuf-fast-free on VPP
> vector mode
>
> Hi VPP maintainers,
>
>
>
> Recently VPP has upgraded the DPDK version to DPDK-21.08, which
> includes two optimization patches[1][2] from Arm DPDK team. With the
> mbuf-fast-free flag, the two patches add code segment to accelerate
> mbuf free in PMD TX path for i40e driver, which shows quite obvious
> performance improvement from DPDK L3FWD benchmarking results.
>
>
>
> I tried to verify the benefits that those optimization patches can
> bring up to VPP, but find out that mbuf-fast-free flag is not enabled
> in
> VPP+DPDK by default.
>
> Applying DPDK mbuf-fast-free has some constraints, e.g,
>
> * mbufs to be freed should be coming from the same mempool
> * ref_cnt == 1 always in mbuf meta-data when user apps call DPDK
> rte_eth_tx_burst ()
> * No TX checksum offload
> * No jumble frame
>
> But VPP vector mode(set by adding ‘no-tx-checksum-offload’ and
> ‘no-multi- seg’ parameters in dpdk section of the startup.conf) seems
> to satisfy all the requirements. So I made a few code changes shown as
> below to set mbuf- fast-free flag by default in VPP vector mode and
> did some benchmarking for
> IPv4 routing test cases with 1 flow/10k flows. The benchmarking
> results show both throughput improvement and CPU cycles saved
> regarding DPDK transmit function.
>
>
>
> So any thought on enabling mbuf-fast-free tx offload flag in VPP
> vector mode?  Any feedback is welcome :)
>
>
>
> Code Changes:
>
>
>
> diff --git a/src/plugins/dpdk/device/init.c
> b/src/plugins/dpdk/device/init.c
>
> index f7c1cc106..0fbdd2317 100644
>
> --- a/src/plugins/dpdk/device/init.c
>
> +++ b/src/plugins/dpdk/device/init.c
>
> @@ -398,6 +398,8 @@ dpdk_lib_init (dpdk_main_t * dm)
>
>   xd->port_conf.rxmode.offloads |= DEV_RX_OFFLOAD_SCATTER;
>
>   xd->flags |= DPDK_DEVICE_FLAG_MAYBE_MULTISEG;
>
> }
>
> +  if (dm->conf->no_multi_seg && dm->conf->no_tx_checksum_offload)
>
> +   xd->port_conf.txmode.offloads |=
> + DEV_TX_OFFLOAD_MBUF_FAST_FREE;
>
>
>
>xd->tx_q_used = clib_min (dev_info.max_tx_queues, tm-
> >n_vlib_mains);
>
>
>
> Benchmark Results:
>
>
>
> 1 flow, bidirectional
>
> Throughput(Mpps):
>
>
>
> Original
>
> Patched
>
> Ratio
>
> N1SDP
>
> 11.62
>
> 12.44
>
> +7.06%
>
> ThunderX2
>
> 9.52
>
> 10.16
>
> +6.30%
>
> Dell 8268
>
> 17.82
>
> 18.20
>
> +2.1

Re: [vpp-dev] Enable DPDK tx offload flag mbuf-fast-free on VPP vector mode

[Jieqiang Wang]

Thanks for your explanation, Damjan. Based on your words, it seems 
inappropriate to apply mbuf-fast-free on VPP, even for SMP systems …

From: vpp-dev@lists.fd.io  On Behalf Of Damjan Marion via 
lists.fd.io
Sent: Thursday, September 23, 2021 12:18 AM
To: Jieqiang Wang 
Cc: Benoit Ganne (bganne) ; vpp-dev ; 
Lijian Zhang ; Honnappa Nagarahalli 
; Govindarajan Mohandoss 
; Ruifeng Wang ; Tianyu 
Li ; Feifei Wang ; nd 
Subject: Re: [vpp-dev] Enable DPDK tx offload flag mbuf-fast-free on VPP vector 
mode


—
Damjan




On 22.09.2021., at 11:50, Jieqiang Wang 
mailto:jieqiang.w...@arm.com>> wrote:

Hi Ben,

Thanks for your quick feedback. A few comments inline.

Best Regards,
Jieqiang Wang

-Original Message-
From: Benoit Ganne (bganne) mailto:bga...@cisco.com>>
Sent: Friday, September 17, 2021 3:34 PM
To: Jieqiang Wang mailto:jieqiang.w...@arm.com>>; 
vpp-dev mailto:vpp-dev@lists.fd.io>>
Cc: Lijian Zhang mailto:lijian.zh...@arm.com>>; Honnappa 
Nagarahalli 
mailto:honnappa.nagaraha...@arm.com>>; 
Govindarajan Mohandoss 
mailto:govindarajan.mohand...@arm.com>>; 
Ruifeng Wang mailto:ruifeng.w...@arm.com>>; Tianyu Li 
mailto:tianyu...@arm.com>>; Feifei Wang 
mailto:feifei.wa...@arm.com>>; nd 
mailto:n...@arm.com>>
Subject: RE: Enable DPDK tx offload flag mbuf-fast-free on VPP vector mode

Hi Jieqiang,

This looks like an interesting optimization but you need to check that the 
'mbufs to be freed should be coming from the same mempool' rule holds true. 
This won't be the case on NUMA systems (VPP creates 1 buffer pool per NUMA).
This should be easy to check with eg. 'vec_len (vm->buffer_main->buffer_pools) 
== 1'.

Jieqiang: That's a really good point here. Like you said, it holds true on SMP 
systems and we can check by if the numbers of buffer pool equal to 1. But I am 
wondering that is this check too strict? If the worker CPUs and NICs used 
reside in the same NUMA node, I think mbufs come from the same mempool and we 
still meet the requirement here.  What do you think?

Please note that VPP is not using DPDK mempools. We are faking them by 
registering our own mempool handlers.
There is special trick how refcnt > 1 is handled. All packets which have vpp 
ref count > 1 are sent to DPDK code as members of another fake mempool which 
have cache turned off.
In reality that means that DPDK will have 2 fake mempools per numa, and all 
packets going to DPDK code will always have set refcnt to 1.



For the rest, I think we do not use DPDK mbuf refcounting at all as we maintain 
our own anyway, but someone more knowledgeable than me should confirm.

Jieqiang: This echoes with the experiments(IPv4 multicasting and L2 flood) I 
have done. All the mbufs in the two test cases are copied instead of ref 
counting. But this also needs double-check from VPP experts like you mentioned.

see above….



I'd be curious to see if we can measure a real performance difference in CSIT.

Jieqiang: Let me trigger some performance test cases in CSIT and come back to 
you with related performance figures.

—
Damjan
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.

-=-=-=-=-=-=-=-=-=-=-=-
Links: You receive all messages sent to this group.
View/Reply Online (#20188): https://lists.fd.io/g/vpp-dev/message/20188
Mute This Topic: https://lists.fd.io/mt/85669132/21656
Group Owner: vpp-dev+ow...@lists.fd.io
Unsubscribe: https://lists.fd.io/g/vpp-dev/unsub [arch...@mail-archive.com]
-=-=-=-=-=-=-=-=-=-=-=-

[vpp-dev] IPv6 routes calculation error using VPP layer 3 IP CLI "ip route add"

[Jieqiang Wang]

Hi VPP Maintainers,

It seems that adding static IPv6 routes using VPP CLI 'ip route add' has some 
problems.
For example, like shown in the attached ipv6_cli.txt file, I add 4 static IPv6 
routes using CLI 'ip route add 2001:1::0/128 count 4 via eth0', which returns 3 
weird routes.
2001:1::/128
  unicast-ip6-chain
  [@0]: dpo-load-balance: [proto:ip6 index:8 buckets:1 uRPF:7 to:[0:0]]
[0] [@3]: arp-ipv6: via 2001:1:: eth0
2001:1:0:1::/128
  unicast-ip6-chain
  [@0]: dpo-load-balance: [proto:ip6 index:9 buckets:1 uRPF:8 to:[0:0]]
[0] [@4]: ipv6-glean: [src:2001:1:0:1::/128] eth0: mtu:9000 next:2 flags:[] 
6891d06246ac86dd
2001:1:0:2::/128
  unicast-ip6-chain
  [@0]: dpo-load-balance: [proto:ip6 index:10 buckets:1 uRPF:9 to:[0:0]]
[0] [@4]: ipv6-glean: [src:2001:1:0:2::/128] eth0: mtu:9000 next:2 flags:[] 
6891d06246ac86dd
2001:1:0:3::/128
  unicast-ip6-chain
  [@0]: dpo-load-balance: [proto:ip6 index:11 buckets:1 uRPF:10 to:[0:0]]
[0] [@4]: ipv6-glean: [src:2001:1:0:3::/128] eth0: mtu:9000 next:2 flags:[] 
6891d06246ac86dd

With some debugging, I find the logic of calculating IPv6 routes using VPP CLI 
is probably wrong. I made a quick patch https://gerrit.fd.io/r/c/vpp/+/33422 to 
resolve this issue.
Could you please take your time to double-confirm this issue and possibly help 
review the patch?

Thanks,
Jieqiang Wang

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.
snowball@net-x86-dell-8268:~/tasks/vpp_debug$ sudo 
./vpp/build-root/install-vpp-native/vpp/bin/vpp -c startup.conf
dpdk/cryptodev   [warn  ]: dpdk_cryptodev_init: Failed to configure 
cryptodev
vat-plug/load[error ]: vat_plugin_register: lb plugin not loaded...
vat-plug/load[error ]: vat_plugin_register: avf plugin not loaded...
vat-plug/load[error ]: vat_plugin_register: ct6 plugin not loaded...
vat-plug/load[error ]: vat_plugin_register: dns plugin not loaded...
vat-plug/load[error ]: vat_plugin_register: mactime plugin not loaded...
vat-plug/load[error ]: pot_vat_plugin_register: pot plugin not loaded...
vat-plug/load[error ]: vat_plugin_register: lacp plugin not loaded...
vat-plug/load[error ]: vat_plugin_register: l2tp plugin not loaded...
vat-plug/load[error ]: vat_plugin_register: lldp plugin not loaded...
vat-plug/load[error ]: vat_plugin_register: builtinurl plugin not 
loaded...
vat-plug/load[error ]: vat_plugin_register: nsim plugin not loaded...
vat-plug/load[error ]: vat_plugin_register: af_xdp plugin not loaded...
vat-plug/load[error ]: vat_plugin_register: vrrp plugin not loaded...
vat-plug/load[error ]: vat_plugin_register: tls_openssl plugin not 
loaded...
vat-plug/load[error ]: vat_plugin_register: vmxnet3 plugin not loaded...
vat-plug/load[error ]: vat_plugin_register: cdp plugin not loaded...
vat-plug/load[error ]: vat_plugin_register: stn plugin not loaded...
vat-plug/load[error ]: vat_plugin_register: rdma plugin not loaded...
vat-plug/load[error ]: vat_plugin_register: mdata plugin not loaded...
vat-plug/load[error ]: vat_plugin_register: dhcp plugin not loaded...
vat-plug/load[error ]: vat_plugin_register: pppoe plugin not loaded...
vat-plug/load[error ]: vat_plugin_register: oddbuf plugin not loaded...
vat-plug/load[error ]: vat_plugin_register: flowprobe plugin not 
loaded...
vat-plug/load[error ]: vat_plugin_register: geneve plugin not loaded...
vat-plug/load[error ]: vat_plugin_register: memif plugin not loaded...
vat-plug/load[error ]: vat_plugin_register: arping plugin not loaded...
vat-plug/load[error ]: vat_plugin_register: ikev2 plugin not loaded...
vat-plug/load[error ]: vat_plugin_register_gpe: lisp_gpe plugin not 
loaded...
vat-plug/load[error ]: vat_plugin_register: acl plugin not loaded...
vat-plug/load[error ]: vat_plugin_register: http_static plugin not 
loaded...
vat-plug/load[error ]: vat_plugin_register: tracedump plugin not 
loaded...
vat-plug/load[error ]: vat_plugin_register: adl plugin not loaded...
vat-plug/load[error ]: vat_plugin_register: nsh plugin not loaded...
vat-plug/load[error ]: vat_plugin_register: gtpu plugin not loaded...
_____   _  ___
 __/ __/ _ \  (_)__| | / / _ \/ _ \
 _/ _// // / / / _ \   | |/ / ___/ ___/
 /_/ /(_)_/\___/   |___/_/  /_/

vpp# show int
  Name   IdxState  MTU (L3/IP4/IP6/MPLS) 
Counter  Count
eth0  1 down 9000/0/0/0
eth1   

[vpp-dev] UDP multithreading error

[nengjie wang]

The attachment is my server program.

My VPP version is 22.06. When I start this server by LDP and send a request
to this server, the server will have a direct segment error. Two threads
share an event queue, resulting in an event queue crash. Can you tell me if
LDP does not support this use?

In addition, I noticed that there is a configuration item called ‘muti
thread workers’ in the VCL configuration file. When this configuration is
turned on, the program can not receive RPC response. Is it suitable for me
to use this configuration item in this case?


udp_recv.c
Description: Binary data

-=-=-=-=-=-=-=-=-=-=-=-
Links: You receive all messages sent to this group.
View/Reply Online (#21855): https://lists.fd.io/g/vpp-dev/message/21855
Mute This Topic: https://lists.fd.io/mt/93520105/21656
Group Owner: vpp-dev+ow...@lists.fd.io
Unsubscribe: https://lists.fd.io/g/vpp-dev/leave/1480452/21656/631435203/xyzzy 
[arch...@mail-archive.com]
-=-=-=-=-=-=-=-=-=-=-=-

[vpp-dev] create host-interface failed

[nengjie wang]

Hi,

  Why do I always return invalid argument when I execute the command create
host interface name vpp? Does af packet v3 have any special dependencies?

  My vpp version is 22.06, the system version is centos7, the GLIBC version
is 2.17, and the kernel version is 3.10.0-1127.


Error message:
af_ packet:Failed to set packet rx ring options: Invalid argument (errno 22)
af_ packet:Failed to set queue 0 error.
af_ packet:Failed to init device error.
create host-interface: Invalid argument (errno 22)

  Can you give me some advice?

Thanks and regards,
Moho

-=-=-=-=-=-=-=-=-=-=-=-
Links: You receive all messages sent to this group.
View/Reply Online (#21926): https://lists.fd.io/g/vpp-dev/message/21926
Mute This Topic: https://lists.fd.io/mt/93925263/21656
Group Owner: vpp-dev+ow...@lists.fd.io
Unsubscribe: https://lists.fd.io/g/vpp-dev/leave/1480452/21656/631435203/xyzzy 
[arch...@mail-archive.com]
-=-=-=-=-=-=-=-=-=-=-=-

Re: [vpp-dev] UDP multithreading error

[nengjie wang]

Thank you very much for your reply.

However, after the socket is set to the non blocking mode, the CPU
utilization rate is always 100%. The resource consumption is very high when
multithreading. I can only reduce the CPU utilization rate by using methods
such as select.

Why does this mode fail in blocking mode? Do you have any ideas?

Florin Coras  于2022年9月8日周四 02:06写道：

> Hi,
>
> Just tested on master and this seems to work fine after a few app fixes.
> Having said that, I would recommend you build vcl native apps and register
> multiple workers for better performance.
>
> Comments:
> - make sure sockets are not blocking otherwise there’s a good change that
> only one of the threads will output something
> - only print a message if data was received
>
> See diff here [1].
>
> Example output, with 2 clients (c1 and c2) and updated ips.
>
> [snip]
> vppcom_session_create:1410: vcl<233836:0>: created session 1
> ldp<233836>: fd 33: calling vls_bind: vlsh 1, addr 0x7ffd7d7cf750, len 16
> vppcom_session_bind:1599: vcl<233836:0>: session 1 handle 1: binding to
> local IPv4 address 7.0.0.2 port 12342, proto UDP
> vppcom_session_listen:1628: vcl<233836:0>: session 1: sending vpp listen
> request...
> vcl_session_bound_handler:569: vcl<233836:0>: session 1 [0x0]: listen
> succeeded!
> ldp<233836>: fd 33 vlsh 1, cmd 4
>
> [recv from 7.0.0.1:60414]c1 s1
>
> [recv from 7.0.0.1:56760]c2 s1
>
> [recv from 7.0.0.1:56760]c2 s2
>
> [recv from 7.0.0.1:60414]c1 s2
>
>
> Regards,
> Florin
>
> [1] https://pastebin.com/4LCukbEC
>
> > On Sep 7, 2022, at 12:05 AM, nengjie wang  wrote:
> >
> > The attachment is my server program.
> >
> > My VPP version is 22.06. When I start this server by LDP and send a
> request to this server, the server will have a direct segment error. Two
> threads share an event queue, resulting in an event queue crash. Can you
> tell me if LDP does not support this use?
> >
> > In addition, I noticed that there is a configuration item called ‘muti
> thread workers’ in the VCL configuration file. When this configuration is
> turned on, the program can not receive RPC response. Is it suitable for me
> to use this configuration item in this case?
> > 
> >
> >
>
>
> 
>
>

-=-=-=-=-=-=-=-=-=-=-=-
Links: You receive all messages sent to this group.
View/Reply Online (#21927): https://lists.fd.io/g/vpp-dev/message/21927
Mute This Topic: https://lists.fd.io/mt/93520105/21656
Group Owner: vpp-dev+ow...@lists.fd.io
Unsubscribe: https://lists.fd.io/g/vpp-dev/leave/1480452/21656/631435203/xyzzy 
[arch...@mail-archive.com]
-=-=-=-=-=-=-=-=-=-=-=-

[vpp-dev] VPP interface parsing error?

[Yichen Wang (yicwang)]

Hi, VPP folks,

I am on latest VPP 17.07-rc1, and saw an interface parsing error.

vpp# show int
  Name   Idx   State  Counter  Count
BondEthernet0 3 up   rx packets
1955066091
 rx bytes
125124243370
 tx packets
1508472227
 tx bytes 
9654800
 drops  
  317
 rx-miss   
1328405166
BondEthernet0.15274 up   rx packets
1422663050
 rx bytes 
91050435200
 tx packets 
532269842
 tx bytes 
34065269888
 drops  
1
BondEthernet0.15146 up   rx packets 
532402727
 rx bytes 
34073774528
 tx packets 
976202368
 tx bytes 
62476951552
 drops  
2
TenGigabitEthernet81/0/1  2 bond-slave   rx-miss   
1109342837
TenGigabitEthernete/0/1   1 bond-slave   rx-miss
219062329
VirtualEthernet0/0/0  5 up   rx packets 
532269844
 rx bytes 
31936190640
 tx packets
1422663049
 tx bytes 
85359782940
 drops  
430668918
VirtualEthernet0/0/1  7 up   rx packets 
976202370
 rx bytes 
58572142200
 tx packets 
532402725
 tx bytes 
31944163500
 drops  
27063
local00down

vpp# show interface rx-pl
Thread 1 (vpp_wk_0):
  node dpdk-input:
TenGigabitEthernete/0/1 queue 0 (polling)
TenGigabitEthernet81/0/1 queue 0 (polling)
BondEthernet0 queue 0 (polling)
Thread 2 (vpp_wk_1):
  node dpdk-input:
TenGigabitEthernete/0/1 queue 1 (polling)
TenGigabitEthernet81/0/1 queue 1 (polling)
BondEthernet0 queue 1 (polling)
Thread 3 (vpp_wk_2):
  node vhost-user-input:
BondEthernet0.1527 queue 0 (polling)
Thread 4 (vpp_wk_3):
  node vhost-user-input:
VirtualEthernet0/0/0 queue 0 (polling)

It is clearly wrong for the CLI above. BondEthernet0.1527 cannot be in 
vhost-user-input node. Also “show vhost” also confirms that Thread 3 should 
have VirtualEthernet0/0/0 and Thread 4 should have VirtualEthernet0/0/1. So I 
believe this is just the names shown here are wrong, but this will be really 
confusing… ☺

Thanks very much!

Regards,
Yichen
___
vpp-dev mailing list
vpp-dev@lists.fd.io
https://lists.fd.io/mailman/listinfo/vpp-dev

[vpp-dev] vppctl bug?

[Yichen Wang (yicwang)]

Hi, VPP folks,

Using a pretty new VPP 17.07-rc2 RPM:
neutron_vpp_8422 [root@sjc04-pod6-compute-1 /]# vppctl show ver verbose
Version:  v17.07-rc2~3-g0786710~b20
Compiled by:  jenkins
Compile host: centos7-basebuild-4c-4g-489
Compile date: Thu Jul  6 18:21:54 UTC 2017
Compile location: /w/workspace/vpp-merge-1707-centos7
Compiler: GCC 4.8.5 20150623 (Red Hat 4.8.5-11)
Current PID:  69

When I use “telnet 0 5002” I can see all information are returned correctly, 
but with vppctl, the same CLI just like break in the middle. Looks like a bug 
in vppctl, 100% reproducible:
[root@sjc04-pod6-compute-1 /]# vppctl show int rx-placement
Thread 1 (vpp_wk_0
[root@sjc04-pod6-compute-1 /]#

Thanks very much!

Regards,
Yichen
___
vpp-dev mailing list
vpp-dev@lists.fd.io
https://lists.fd.io/mailman/listinfo/vpp-dev

Re: [vpp-dev] vppctl bug?

[Yichen Wang (yicwang)]

[Resend]

This is 100% reproducible, and kind of blocking our automation. Anyone knows 
the reason?

Thanks very much!

Regards,
Yichen

From: "Yichen Wang (yicwang)" 
Date: Tuesday, July 11, 2017 at 14:51
To: Denis Lotarev via vpp-dev 
Subject: vppctl bug?

Hi, VPP folks,

Using a pretty new VPP 17.07-rc2 RPM:
neutron_vpp_8422 [root@sjc04-pod6-compute-1 /]# vppctl show ver verbose
Version:  v17.07-rc2~3-g0786710~b20
Compiled by:  jenkins
Compile host: centos7-basebuild-4c-4g-489
Compile date: Thu Jul  6 18:21:54 UTC 2017
Compile location: /w/workspace/vpp-merge-1707-centos7
Compiler: GCC 4.8.5 20150623 (Red Hat 4.8.5-11)
Current PID:  69

When I use “telnet 0 5002” I can see all information are returned correctly, 
but with vppctl, the same CLI just like break in the middle. Looks like a bug 
in vppctl, 100% reproducible:
[root@sjc04-pod6-compute-1 /]# vppctl show int rx-placement
Thread 1 (vpp_wk_0
[root@sjc04-pod6-compute-1 /]#

Thanks very much!

Regards,
Yichen
___
vpp-dev mailing list
vpp-dev@lists.fd.io
https://lists.fd.io/mailman/listinfo/vpp-dev

[vpp-dev] VPP VCL HOWTO

[WANG, CHENGWEI (CHENGWEI)]

Hi VPP developers,

I was trying to install and run VPP VCL by following the directions on

https://wiki.fd.io/view/VPP/VPPCommunicationsLibrary

but I was noticed that this page is out of date.

Would you please point to me the latest HowTO?

Thanks!
Chengwei
___
vpp-dev mailing list
vpp-dev@lists.fd.io
https://lists.fd.io/mailman/listinfo/vpp-dev

[vpp-dev] Does the latest VPP from gerrit support honeycomb?

[WANG, CHENGWEI (CHENGWEI)]

Hi VPP developers,

Does the latest VPP support honeycomb? We could not install honeycomb with the 
latest VPP, so we thought it might be a version issue. We can be wrong. Thanks 
for enlightening.

Chengwei
___
vpp-dev mailing list
vpp-dev@lists.fd.io
https://lists.fd.io/mailman/listinfo/vpp-dev

[vpp-dev] Is there a way to change the MAC address in VPP 1609?

[WANG, CHENGWEI (CHENGWEI)]

Hi VPP developers,

As far as I know, the VPP 1609 version does not have

"Set interface mac address" command line.

I know the latest VPP has this command line but If I use VPP 1609, is there any 
other way to change the mac address of an interface?

Thanks
Chengwei
___
vpp-dev mailing list
vpp-dev@lists.fd.io
https://lists.fd.io/mailman/listinfo/vpp-dev

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

[Yichen Wang (yicwang)]

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/), and driver bind failed;
(2)   igb_uio will work perfectly, but does not come with RHEL 7.3 kernel 
directly. Have to build from source to have it;
(3)   vfio-pci, which is the only option left.
According to 
https://wiki.fd.io/view/VPP/Command-line_Arguments#.22dpdk.22_parameters, 
vfio-pci should be supported. However, when I bring it up, VPP is complaining:
EAL: Detected 72 lcore(s)
EAL: No free hugepages reported in hugepages-1048576kB
EAL: Probing VFIO support...
EAL: VFIO support initialized
EAL: Initializing pmd_bond for eth_bond0
EAL: Create bonded device eth_bond0 on port 0 in mode 2 on socket 0.
EAL: PCI device :16:00.1 on NUMA socket 0
EAL:   probe driver: 8086:1572 net_i40e
EAL:   :16:00.1 not managed by VFIO driver, skipping
EAL: PCI device :81:00.1 on NUMA socket 1
EAL:   probe driver: 8086:1572 net_i40e
EAL:   :81:00.1 not managed by VFIO driver, skipping
DPDK physical memory layout:
Segment 0: phys:0x4b80, len:534773760, virt:0x7f9c41a0, socket_id:0, 
hugepage_sz:2097152, nchannel:0, nrank:0
Segment 1: phys:0x5f7a80, len:534773760, virt:0x7f5c6f20, socket_id:1, 
hugepage_sz:2097152, nchannel:0, nrank:0
PMD: bond_ethdev_parse_slave_port_kvarg(142) - Invalid slave port value 
(:16:00.1) specified
EAL: Failed to parse slave ports for bonded device eth_bond0
Apparently VPP is not recognizing the interfaces bound to vfio-pci, therefore 
it couldn’t set up bonding afterwards. However I do have those interfaces bound 
to vfio-pci already, here is the output from dpdk-devbind.py:
[root@sjc04-pod6-compute-4 tools]# ./dpdk-devbind.py --status
Network devices using DPDK-compatible driver

:16:00.1 'Ethernet Controller X710 for 10GbE SFP+' drv=vfio-pci unused=i40e
:81:00.1 'Ethernet Controller X710 for 10GbE SFP+' drv=vfio-pci unused=i40e
Do we ever tested vfio-pci on X710 before, or did I miss anything? Appreciate 
your helps!
Thanks very much!
Regards,
Yichen
___
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

[Yichen Wang (yicwang)]

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. Among all three DPDK drivers:
(1) uio_pci_generic is not supported on X710 
(http://dpdk.org/dev/patchwork/patch/19820/), and driver bind failed;
(2)   igb_uio will work perfectly, but does not come with RHEL 7.3 kernel 
directly. Have to build from source to have it;
(3)   vfio-pci, which is the only option left.
According to 
https://wiki.fd.io/view/VPP/Command-line_Arguments#.22dpdk.22_parameters, 
vfio-pci should be supported. However, when I bring it up, VPP is complaining:
EAL: Detected 72 lcore(s)
EAL: No free hugepages reported in hugepages-1048576kB
EAL: Probing VFIO support...
EAL: VFIO support initialized
EAL: Initializing pmd_bond for eth_bond0
EAL: Create bonded device eth_bond0 on port 0 in mode 2 on socket 0.
EAL: PCI device :16:00.1 on NUMA socket 0
EAL:   probe driver: 8086:1572 net_i40e
EAL:   :16:00.1 not managed by VFIO driver, skipping
EAL: PCI device :81:00.1 on NUMA socket 1
EAL:   probe driver: 8086:1572 net_i40e
EAL:   :81:00.1 not managed by VFIO driver, skipping
DPDK physical memory layout:
Segment 0: phys:0x4b80, len:534773760, virt:0x7f9c41a0, socket_id:0, 
hugepage_sz:2097152, nchannel:0, nrank:0
Segment 1: phys:0x5f7a80, len:534773760, virt:0x7f5c6f20, socket_id:1, 
hugepage_sz:2097152, nchannel:0, nrank:0
PMD: bond_ethdev_parse_slave_port_kvarg(142) - Invalid slave port value 
(:16:00.1) specified
EAL: Failed to parse slave ports for bonded device eth_bond0
Apparently VPP is not recognizing the interfaces bound to vfio-pci, therefore 
it couldn't set up bonding afterwards. However I do have those interfaces bound 
to vfio-pci already, here is the output from dpdk-devbind.py:
[root@sjc04-pod6-compute-4 tools]# ./dpdk-devbind.py --status
Network devices using DPDK-compatible driver

:16:00.1 'Ethernet Controller X710 for 10GbE SFP+' drv=vfio-pci unused=i40e
:81:00.1 'Ethernet Controller X710 for 10GbE SFP+' drv=vfio-pci unused=i40e
Do we ever tested vfio-pci on X710 before, or did I miss anything? Appreciate 
your helps!
Thanks very much!
Regards,
Yichen

Have you specified:

dpdk {
  uio-driver vfio-pci
}

in startup.conf?



___
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

[Yichen Wang (yicwang)]

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 Cisco  
I350-TX 1Gig LOM
:01:00.1   08086:1521   5.0 GT/s x4igb Cisco  
I350-TX 1Gig LOM
:09:00.0   01137:0043   5.0 GT/s x16   enic
:10:00.0   01137:0043   5.0 GT/s x16   enic
:11:00.0   01137:0043   5.0 GT/s x16   enic
:12:00.0   01137:0043   5.0 GT/s x16   enic
:81:00.0   18086:1572   8.0 GT/s x8i40eCisco(R) 
Ethernet Converged NIC X710-DA4
:81:00.1   18086:1572   8.0 GT/s x8vfio-pciCisco(R) 
Ethernet Converged NIC X710-DA4
:81:00.2   18086:1572   8.0 GT/s x8i40eCisco(R) 
Ethernet Converged NIC X710-DA4
:81:00.3   18086:1572   8.0 GT/s x8i40eCisco(R) 
Ethernet Converged NIC X710-DA4
:81:0a.0   18086:154c   unknownpci-stub
:81:0a.1   18086:154c   unknownpci-stub
:81:0a.2   18086:154c   unknownpci-stub
:81:0a.3   18086:154c   unknownpci-stub
:81:0a.4   18086:154c   unknownpci-stub
:81:0a.5   18086:154c   unknownpci-stub
:81:0a.6   18086:154c   unknownpci-stub
:81:0a.7   18086:154c   unknownpci-stub
:81:0b.0   18086:154c   unknownpci-stub
:81:0b.1   18086:154c   unknownpci-stub
:81:0b.2   18086:154c   unknownpci-stub
:81:0b.3   18086:154c   unknownpci-stub
:81:0b.4   18086:154c   unknownpci-stub
:81:0b.5   18086:154c   unknownpci-stub
:81:0b.6   18086:154c   unknownpci-stub
:81:0b.7   18086:154c   unknownpci-stub
:13:00.0   01137:0043   5.0 GT/s x16   enic
:14:00.0   01137:0043   5.0 GT/s x16   enic
:16:00.0   08086:1572   8.0 GT/s x8i40eCisco(R) 
Ethernet Converged NIC X710-DA4
:16:00.1   08086:1572   8.0 GT/s x8vfio-pciCisco(R) 
Ethernet Converged NIC X710-DA4
:16:00.2   08086:1572   8.0 GT/s x8i40eCisco(R) 
Ethernet Converged NIC X710-DA4
:16:00.3   08086:1572   8.0 GT/s x8i40eCisco(R) 
Ethernet Converged NIC X710-DA4
:16:0a.0   08086:154c   unknownpci-stub
:16:0a.1   08086:154c   unknownpci-stub
:16:0a.2   08086:154c   unknownpci-stub
:16:0a.3   08086:154c   unknownpci-stub
:16:0a.4   08086:154c   unknownpci-stub
:16:0a.5   08086:154c   unknownpci-stub
:16:0a.6   08086:154c   unknownpci-stub
:16:0a.7   08086:154c   unknownpci-stub
:16:0b.0   08086:154c   unknownpci-stub
:16:0b.1   08086:154c   unknownpci-stub
:16:0b.2   08086:154c   unknownpci-stub
:16:0b.3   08086:154c   unknownpci-stub
:16:0b.4   08086:154c   unknownpci-stub
:16:0b.5   08086:154c   unknownpci-stub
:16:0b.6   08086:154c   unknownpci-stub
:16:0b.7   08086:154c   unknownpci-stub
:0a:00.0   01137:0043   5.0 GT/s x16   enic
:0d:00.0   01137:0043   5.0 GT/s x16   enic
:0e:00.0   01137:0043   5.0 GT/s x16   enic
:0f:00.0   01137:0043   5.0 GT/s x16   enic

The VPP console will come up, just found no interfaces in “show int”.

Thanks very much!

Regards,
Yichen

From: Damjan Marion 
Date: Wednesday, January 25, 2017 at 09:38
To: "Yichen Wang (yicwang)" 
Cc: "vpp-dev@lists.fd.io" , "Ian Wells (iawells)" 

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


On 25 Jan 2017, at 18:03, Yichen Wang (yicwang) 
mailto:yicw...@cisco.com>> 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 
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. Among all three DPDK drivers:
(1) uio_pci_generic is not supported on X710 
(http://dpdk.org/dev/patchwork/patch/19820/), and driver bind failed;
(2)   igb_uio will work perfectly, but does not come with RHEL 7.3 kernel 
directly. Have to build from source to have it;
(3)   vfio-pci, which is the only option left.
According to 
https://wiki.fd.io/view/VPP/Command-line_Arguments#.22dpdk.22_parameters, 
vfio-pci should be supported. However, when I bring it up, VPP is complaining:
EAL: Detected 72 lcore(s)
EAL: No free hugepages reported in hugepages-1048576kB
EAL: Probing V

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

[Yichen Wang (yicwang)]

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.

Thanks very much!

Regards,
Yichen

From: Damjan Marion 
Date: Wednesday, January 25, 2017 at 10:39
To: "Yichen Wang (yicwang)" 
Cc: "vpp-dev@lists.fd.io" , "Ian Wells (iawells)" 

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



On 25 Jan 2017, at 18:59, Yichen Wang (yicwang) 
mailto:yicw...@cisco.com>> 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  Product Name
:01:00.0   08086:1521   5.0 GT/s x4igb Cisco  
I350-TX 1Gig LOM
:01:00.1   08086:1521   5.0 GT/s x4igb Cisco  
I350-TX 1Gig LOM
:09:00.0   01137:0043   5.0 GT/s x16   enic
:10:00.0   01137:0043   5.0 GT/s x16   enic
:11:00.0   01137:0043   5.0 GT/s x16   enic
:12:00.0   01137:0043   5.0 GT/s x16   enic
:81:00.0   18086:1572   8.0 GT/s x8i40eCisco(R) 
Ethernet Converged NIC X710-DA4
:81:00.1   18086:1572   8.0 GT/s x8vfio-pciCisco(R) 
Ethernet Converged NIC X710-DA4
:81:00.2   18086:1572   8.0 GT/s x8i40eCisco(R) 
Ethernet Converged NIC X710-DA4
:81:00.3   18086:1572   8.0 GT/s x8i40eCisco(R) 
Ethernet Converged NIC X710-DA4
:81:0a.0   18086:154c   unknownpci-stub
:81:0a.1   18086:154c   unknownpci-stub
:81:0a.2   18086:154c   unknownpci-stub
:81:0a.3   18086:154c   unknownpci-stub
:81:0a.4   18086:154c   unknownpci-stub
:81:0a.5   18086:154c   unknownpci-stub
:81:0a.6   18086:154c   unknownpci-stub
:81:0a.7   18086:154c   unknownpci-stub
:81:0b.0   18086:154c   unknownpci-stub
:81:0b.1   18086:154c   unknownpci-stub
:81:0b.2   18086:154c   unknownpci-stub
:81:0b.3   18086:154c   unknownpci-stub
:81:0b.4   18086:154c   unknownpci-stub
:81:0b.5   18086:154c   unknownpci-stub
:81:0b.6   18086:154c   unknownpci-stub
:81:0b.7   18086:154c   unknownpci-stub
:13:00.0   01137:0043   5.0 GT/s x16   enic
:14:00.0   01137:0043   5.0 GT/s x16   enic
:16:00.0   08086:1572   8.0 GT/s x8i40eCisco(R) 
Ethernet Converged NIC X710-DA4
:16:00.1   08086:1572   8.0 GT/s x8vfio-pciCisco(R) 
Ethernet Converged NIC X710-DA4
:16:00.2   08086:1572   8.0 GT/s x8i40eCisco(R) 
Ethernet Converged NIC X710-DA4
:16:00.3   08086:1572   8.0 GT/s x8i40eCisco(R) 
Ethernet Converged NIC X710-DA4
:16:0a.0   08086:154c   unknownpci-stub
:16:0a.1   08086:154c   unknownpci-stub
:16:0a.2   08086:154c   unknownpci-stub
:16:0a.3   08086:154c   unknownpci-stub
:16:0a.4   08086:154c   unknownpci-stub
:16:0a.5   08086:154c   unknownpci-stub
:16:0a.6   08086:154c   unknownpci-stub
:16:0a.7   08086:154c   unknownpci-stub
:16:0b.0   08086:154c   unknownpci-stub
:16:0b.1   08086:154c   unknownpci-stub
:16:0b.2   08086:154c   unknownpci-stub
:16:0b.3   08086:154c   unknownpci-stub
:16:0b.4   08086:154c   unknownpci-stub
:16:0b.5   08086:154c   unknownpci-stub
:16:0b.6   08086:154c   unknownpci-stub
:16:0b.7   08086:154c   unknownpci-stub
:0a:00.0   01137:0043   5.0 GT/s x16   enic
:0d:00.0   01137:0043   5.0 GT/s x16   enic
:0e:00.0   01137:0043   5.0 GT/s x16   enic
:0f:00.0   01137:0043   5.0 GT/s x16   enic

The VPP console will come up, just found no interfaces in “show int”.

This is interesting problem. Have you tried to use testpmd?

Error message you’re getting is comping straight from DPDK and this device 
looks properly bound to vfio-pci.


Thanks very much!

Regards,
Yichen

From: Damjan Marion mailto:dmarion.li...@gmail.com>>
Date: Wednesday, January 25, 2017 at 09:38
To: "Yichen Wang (yicwang)" mailto:yicw...@cisco.com>>
Cc: "vpp-dev@lists.fd.io<mailto:vpp-dev@lists.fd.io>" 
mailto:vpp-dev@lists.fd.io>>, "Ian Wells (iawells)" 
mailto:iawe...@cisco.com>>
Subject: Re: [vpp-dev] VPP 17.01 on VFIO-PCI driver


On 25 Jan 2017, at 18:03, Yichen Wang (yicwa

[vpp-dev] Use RSS in VPP 17.01

[Yichen Wang (yicwang)]

Hi, VPP folks,

From what I saw on the VPP docs, there are some places do mention that VPP 
supports RSS. Like the example given in the bottom of the link, we do see two 
queues per interfaces are shown:
https://wiki.fd.io/view/VPP/Using_VPP_In_A_Multi-thread_Model

I want to try exactly the same thing on both Cisco VIC (enic driver) and Intel 
X710 (i40e), but could not get it working straight, and need some helps from 
you guys! ☺

On Cisco VIC, I went to CIMC and configure 1 TX Queue and 4 RX Queue, and 
enabled all RSS related features. From the RHEL OS, I can clearly see all 5 
queues are shown in /proc/interrupts. However, when I bind the interface to 
VPP, I can only see “TenGigabitEthernet9/0/0 queue 0” from “show dpdk interface 
placement”, but not other queues, am I doing something wrong?

On the Intel X710 side, I did the same type of check of Cisco VIC, and I see 16 
TxRx queues were shown, which makes sense because my server has 16 CPU cores. 
When I bind the interfaces to VPP, same I could only see 1 queue are shown in 
the VPP. How can I configure VPP/DPDK to use RSS? I looked online, but there 
were really limited documents from RHEL/Intel/VPP, which doesn’t seem to be 
very helpful. Only reference I found which is on OVS-DPDK and they are doing 
the RSS configuration from OVS context.

Very looking forward to your helps, and thanks in advance!

Regards,
Yichen
___
vpp-dev mailing list
vpp-dev@lists.fd.io
https://lists.fd.io/mailman/listinfo/vpp-dev

Re: [vpp-dev] Vpp in a container

[Yichen Wang (yicwang)]

Hi, Raghav,

The pros and cons of running VPP in a container are the general pros/cons of 
running applications in container, and shouldn’t be anything special. Folks can 
comment more about some specific pros/cons for VPP.

We were running VPP inside Docker in our project, and it seems to work 
properly. Regarding of the performance, people are measuring on having 
different setups, so it is hard to compare apple-to-apple. The thing I can tell 
is, there is not significant performance loss when compared to the data we are 
aware of so far when running VPP inside container.

Hope that helps.

Regards,
Yichen

From:  on behalf of "Raghav Kaushik (rakaushi)" 

Date: Thursday, February 9, 2017 at 17:40
To: "vpp-dev@lists.fd.io" 
Subject: [vpp-dev] Vpp in a container

Hi Folks,

I’m trying to find some data about pros and cons of running VPP in a container.

Has this been tried before? Are there some performance number comparisons 
available ?

Any pointers will be much appreciated.

Thanks,
Raghav
___
vpp-dev mailing list
vpp-dev@lists.fd.io
https://lists.fd.io/mailman/listinfo/vpp-dev

Re: [vpp-dev] Use RSS in VPP 17.01

[Yichen Wang (yicwang)]

Hi, John/all,

Thanks for your pointer, and I am able to bring up my VPP with multiple queues!

I am doing a PVP test, which basically I am expecting Traffic Generator -> VPP 
on Host -> Loopback VM (testpmd) -> VPP on Host -> Traffic Generator. I can see 
the packets are delivered to the Loopback VM with no problem, but:

(1) VPP shows all packets are dropped:

VirtualEthernet0/0/2  8 up   tx packets   
3237064

 tx bytes   
194223840

 drops3237051

But I did check testpmd and it got all packets, and does its job correctly by 
forwarding the packets to the other interfaces;

(2) VPP show err:

   CountNode  Reason

692521171vhost-user-inputno available buffer

Why it says “no available buffer”? It works pretty well without RSS,. Did I 
miss anything?

Thanks very much for your helps!

Regards,
Yichen

From: "John Lo (loj)" 
Date: Thursday, February 9, 2017 at 20:09
To: "Yichen Wang (yicwang)" , "vpp-dev@lists.fd.io" 

Cc: "Ian Wells (iawells)" 
Subject: RE: Use RSS in VPP 17.01

For VPP, the number of queues on a device can be specified in the DPDK portion 
of the startup config, which default to 1. This is usually documented in as 
comments in the startup.conf template file when installing VPP rpm/deb on the 
target Linux OS. Following is the dpdk portion from the startup.conf in 
/etc/vpp/ directory after installing the vpp deb packages on my Ubuntu server:

dpdk {
## Change default settings for all intefaces
# dev default {
   ## Number of receive queues, enables RSS
   ## Default is 1
   # num-rx-queues 3

   ## Number of transmit queues, Default is equal
   ## to number of worker threads or 1 if no workers treads
   # num-tx-queues 3

   ## Number of descriptors in transmit and receive rings
   ## increasing or reducing number can impact performance
   ## Default is 1024 for both rx and tx
   # num-rx-desc 512
   # num-tx-desc 512

   ## VLAN strip offload mode for interface
   ## Default is off
   # vlan-strip-offload on
# }

## Whitelist specific interface by specifying PCI address
# dev :02:00.0

## Whitelist specific interface by specifying PCI address and in
## addition specify custom parameters for this interface
# dev :02:00.1 {
#   num-rx-queues 2
# }

## Change UIO driver used by VPP, Options are: uio_pci_generic, vfio-pci
## and igb_uio (default)
# uio-driver uio_pci_generic

## Disable mutli-segment buffers, improves performance but
## disables Jumbo MTU support
# no-multi-seg

## Increase number of buffers allocated, needed only in scenarios with
## large number of interfaces and worker threads. Value is per CPU 
socket.
## Default is 32768
# num-mbufs 128000

## Change hugepages allocation per-socket, needed only if there is need 
for
## larger number of mbufs. Default is 256M on each detected CPU socket
# socket-mem 2048,2048
}

Regards,
John

From: vpp-dev-boun...@lists.fd.io [mailto:vpp-dev-boun...@lists.fd.io] On 
Behalf Of Yichen Wang (yicwang)
Sent: Thursday, February 09, 2017 10:38 PM
To: vpp-dev@lists.fd.io
Cc: Ian Wells (iawells) 
Subject: [vpp-dev] Use RSS in VPP 17.01

Hi, VPP folks,

From what I saw on the VPP docs, there are some places do mention that VPP 
supports RSS. Like the example given in the bottom of the link, we do see two 
queues per interfaces are shown:
https://wiki.fd.io/view/VPP/Using_VPP_In_A_Multi-thread_Model

I want to try exactly the same thing on both Cisco VIC (enic driver) and Intel 
X710 (i40e), but could not get it working straight, and need some helps from 
you guys! ☺

On Cisco VIC, I went to CIMC and configure 1 TX Queue and 4 RX Queue, and 
enabled all RSS related features. From the RHEL OS, I can clearly see all 5 
queues are shown in /proc/interrupts. However, when I bind the interface to 
VPP, I can only see “TenGigabitEthernet9/0/0 queue 0” from “show dpdk interface 
placement”, but not other queues, am I doing something wrong?

On the Intel X710 side, I did the same type of check of Cisco VIC, and I see 16 
TxRx queues were shown, which makes sense because my server has 16 CPU cores. 
When I bind the interfaces to VPP, same I could only see 1 queue are shown in 
the VPP. How can I configure VPP/DPDK to use RSS? I looked online, but there 
were really limited documents from RHEL/Intel/VPP, which doesn’t seem to be 
very helpful. Only reference I found which is on OVS-DPDK and they are doing 
the RSS configuration from OVS context.

Very 

Re: [vpp-dev] Use RSS in VPP 17.01

[Yichen Wang (yicwang)]

Ok, figured it out. I need a larger num-mbufs when RSS is enabled…

Regards,
Yichen

From:  on behalf of "Yichen Wang (yicwang)" 

Date: Monday, February 13, 2017 at 17:23
To: "John Lo (loj)" , "vpp-dev@lists.fd.io" 

Cc: "Ian Wells (iawells)" 
Subject: Re: [vpp-dev] Use RSS in VPP 17.01

Hi, John/all,

Thanks for your pointer, and I am able to bring up my VPP with multiple queues!

I am doing a PVP test, which basically I am expecting Traffic Generator -> VPP 
on Host -> Loopback VM (testpmd) -> VPP on Host -> Traffic Generator. I can see 
the packets are delivered to the Loopback VM with no problem, but:

(1)   VPP shows all packets are dropped:

VirtualEthernet0/0/2  8 up   tx packets   
3237064

 tx bytes   
194223840

 drops3237051

But I did check testpmd and it got all packets, and does its job correctly by 
forwarding the packets to the other interfaces;

(2)   VPP show err:

   CountNode  Reason

692521171vhost-user-inputno available buffer

Why it says “no available buffer”? It works pretty well without RSS,. Did I 
miss anything?

Thanks very much for your helps!

Regards,
Yichen

From: "John Lo (loj)" 
Date: Thursday, February 9, 2017 at 20:09
To: "Yichen Wang (yicwang)" , "vpp-dev@lists.fd.io" 

Cc: "Ian Wells (iawells)" 
Subject: RE: Use RSS in VPP 17.01

For VPP, the number of queues on a device can be specified in the DPDK portion 
of the startup config, which default to 1. This is usually documented in as 
comments in the startup.conf template file when installing VPP rpm/deb on the 
target Linux OS. Following is the dpdk portion from the startup.conf in 
/etc/vpp/ directory after installing the vpp deb packages on my Ubuntu server:

dpdk {
## Change default settings for all intefaces
# dev default {
   ## Number of receive queues, enables RSS
   ## Default is 1
   # num-rx-queues 3

   ## Number of transmit queues, Default is equal
   ## to number of worker threads or 1 if no workers treads
   # num-tx-queues 3

   ## Number of descriptors in transmit and receive rings
   ## increasing or reducing number can impact performance
   ## Default is 1024 for both rx and tx
   # num-rx-desc 512
   # num-tx-desc 512

   ## VLAN strip offload mode for interface
   ## Default is off
   # vlan-strip-offload on
# }

## Whitelist specific interface by specifying PCI address
# dev :02:00.0

## Whitelist specific interface by specifying PCI address and in
## addition specify custom parameters for this interface
# dev :02:00.1 {
#   num-rx-queues 2
# }

## Change UIO driver used by VPP, Options are: uio_pci_generic, vfio-pci
## and igb_uio (default)
# uio-driver uio_pci_generic

## Disable mutli-segment buffers, improves performance but
## disables Jumbo MTU support
# no-multi-seg

## Increase number of buffers allocated, needed only in scenarios with
## large number of interfaces and worker threads. Value is per CPU 
socket.
## Default is 32768
# num-mbufs 128000

## Change hugepages allocation per-socket, needed only if there is need 
for
## larger number of mbufs. Default is 256M on each detected CPU socket
# socket-mem 2048,2048
}

Regards,
John

From: vpp-dev-boun...@lists.fd.io [mailto:vpp-dev-boun...@lists.fd.io] On 
Behalf Of Yichen Wang (yicwang)
Sent: Thursday, February 09, 2017 10:38 PM
To: vpp-dev@lists.fd.io
Cc: Ian Wells (iawells) 
Subject: [vpp-dev] Use RSS in VPP 17.01

Hi, VPP folks,

From what I saw on the VPP docs, there are some places do mention that VPP 
supports RSS. Like the example given in the bottom of the link, we do see two 
queues per interfaces are shown:
https://wiki.fd.io/view/VPP/Using_VPP_In_A_Multi-thread_Model

I want to try exactly the same thing on both Cisco VIC (enic driver) and Intel 
X710 (i40e), but could not get it working straight, and need some helps from 
you guys! ☺

On Cisco VIC, I went to CIMC and configure 1 TX Queue and 4 RX Queue, and 
enabled all RSS related features. From the RHEL OS, I can clearly see all 5 
queues are shown in /proc/interrupts. However, when I bind the interface to 
VPP, I can only see “TenGigabitEthernet9/0/0 queue 0” from “show dpdk interface 
placement”, but not other queues, am I doing something wrong?

On the Intel X710 side, I did the same type of check of Cisco VIC, and I see 16 
TxRx queues were shown, which makes sense because my server has 16 CPU cores. 
When I bind the interfaces to VPP, sam

Re: [vpp-dev] Use RSS in VPP 17.01

[Yichen Wang (yicwang)]

Hi, John/all,

I’ve been able to make the RSS work, and got some sets of data. But the data 
doesn’t look promising when compared with 1 RX queue case, i.e. even slower. My 
deployment is running on bonding, and even in the case of 1 RX queue, I saw 
three interfaces in total already in “show dpdk interface placement”, 2 
physical + 1 bonding. So if I want to set up with 2 RX queue, do I have to:

(1) Set the num-rx-queues to 2 for all my 3 interfaces including bonding, 
i.e. I am expected to see 6 interfaces in total in “show dpdk interface 
placement”? How does VPP work internally when bonding interface is created?

(2) In the loopback VM, is it a must to have 2 RX queues as well, or 
actually the loopback VM doesn’t really matter.

Thanks very much!

Regards,
Yichen

From: "Yichen Wang (yicwang)" 
Date: Monday, February 13, 2017 at 23:39
To: "John Lo (loj)" , "vpp-dev@lists.fd.io" 

Cc: "Ian Wells (iawells)" 
Subject: Re: [vpp-dev] Use RSS in VPP 17.01

Ok, figured it out. I need a larger num-mbufs when RSS is enabled…

Regards,
Yichen

From:  on behalf of "Yichen Wang (yicwang)" 

Date: Monday, February 13, 2017 at 17:23
To: "John Lo (loj)" , "vpp-dev@lists.fd.io" 

Cc: "Ian Wells (iawells)" 
Subject: Re: [vpp-dev] Use RSS in VPP 17.01

Hi, John/all,

Thanks for your pointer, and I am able to bring up my VPP with multiple queues!

I am doing a PVP test, which basically I am expecting Traffic Generator -> VPP 
on Host -> Loopback VM (testpmd) -> VPP on Host -> Traffic Generator. I can see 
the packets are delivered to the Loopback VM with no problem, but:

(1)   VPP shows all packets are dropped:

VirtualEthernet0/0/2  8 up   tx packets   
3237064

 tx bytes   
194223840

 drops3237051

But I did check testpmd and it got all packets, and does its job correctly by 
forwarding the packets to the other interfaces;

(2)   VPP show err:

   CountNode  Reason

692521171vhost-user-inputno available buffer

Why it says “no available buffer”? It works pretty well without RSS,. Did I 
miss anything?

Thanks very much for your helps!

Regards,
Yichen

From: "John Lo (loj)" 
Date: Thursday, February 9, 2017 at 20:09
To: "Yichen Wang (yicwang)" , "vpp-dev@lists.fd.io" 

Cc: "Ian Wells (iawells)" 
Subject: RE: Use RSS in VPP 17.01

For VPP, the number of queues on a device can be specified in the DPDK portion 
of the startup config, which default to 1. This is usually documented in as 
comments in the startup.conf template file when installing VPP rpm/deb on the 
target Linux OS. Following is the dpdk portion from the startup.conf in 
/etc/vpp/ directory after installing the vpp deb packages on my Ubuntu server:

dpdk {
## Change default settings for all intefaces
# dev default {
   ## Number of receive queues, enables RSS
   ## Default is 1
   # num-rx-queues 3

   ## Number of transmit queues, Default is equal
   ## to number of worker threads or 1 if no workers treads
   # num-tx-queues 3

   ## Number of descriptors in transmit and receive rings
   ## increasing or reducing number can impact performance
   ## Default is 1024 for both rx and tx
   # num-rx-desc 512
   # num-tx-desc 512

   ## VLAN strip offload mode for interface
   ## Default is off
   # vlan-strip-offload on
# }

## Whitelist specific interface by specifying PCI address
# dev :02:00.0

## Whitelist specific interface by specifying PCI address and in
## addition specify custom parameters for this interface
# dev :02:00.1 {
#   num-rx-queues 2
# }

## Change UIO driver used by VPP, Options are: uio_pci_generic, vfio-pci
## and igb_uio (default)
# uio-driver uio_pci_generic

## Disable mutli-segment buffers, improves performance but
## disables Jumbo MTU support
# no-multi-seg

## Increase number of buffers allocated, needed only in scenarios with
## large number of interfaces and worker threads. Value is per CPU 
socket.
## Default is 32768
# num-mbufs 128000

## Change hugepages allocation per-socket, needed only if there is need 
for
## larger number of mbufs. Default is 256M on each detected CPU socket
# socket-mem 2048,2048
}

Regards,
John

From: vpp-dev-boun...@lists.fd.io [mailto:vpp-dev-boun...@lists.fd.io] On 
Behalf Of Yichen Wang (yicwang)
Sent: Thursday, February 09, 2017 10:38 PM
To: vpp-dev@lists.fd.io
Cc: Ian Wells (iawells) 
Subject: [vpp-dev] Use RSS in V

Re: [vpp-dev] "vppctl show int" no NIC (just local0) #vpp #vnet

[Yichen Wang via Lists.Fd.Io]

If you want to use vfio-pci, you might want to check:
# dmesg | grep Virtualization
[5.208330] DMAR: Intel(R) Virtualization Technology for Directed I/O
If you don’t see above, vfio-pci will not work and the fix is to enable Intel 
VT-d in BIOS.

Also, uio_pci_generic won’t work with i40e, and if you want to use UIO you have 
to use igb_uio (Built-in in Ubuntu, and complied as KO for RHEL/CentOS).

Hope that helps.

Regards,
Yichen

From:  on behalf of "steven luong via Lists.Fd.Io" 

Reply-To: "Steven Luong (sluong)" 
Date: Monday, January 6, 2020 at 8:32 PM
To: Gencli Liu <18600640...@163.com>, "vpp-dev@lists.fd.io" 

Cc: "vpp-dev@lists.fd.io" 
Subject: Re: [vpp-dev] "vppctl show int" no NIC (just local0) #vpp #vnet

It is likely a resource problem – when VPP requests more descriptors and/or 
TX/RX queues for the NIC than the firmware has, DPDK fails to initialize the 
interface. There are few ways to figure out what the problem is.

  1.  Bypass VPP and run testpmd with debug options turned on, something like 
this

--log-level=lib.eal,debug --log-level=pmd,debug

  1.  Reduce your RX/TX queues and descriptors to the minimum for the 
interface. What do you have in the dpdk section for the NIC, anyway?
  2.  Run VPP with bare minimum config.

unix { interactive }

I would start with (3) since it is the easiest. I hope DPDK will discover the 
NIC in show hardware if the interface is already bound to DPDK. If that is the 
case, you can proceed to check and see if your startup.conf oversubscribes the 
descriptors and/or TX/RX queues. If (3) still fails, try (1). It is a bit more 
work. I am sure you’ll figure out how to compile testpmd APP and run it.

Steven

From:  on behalf of Gencli Liu <18600640...@163.com>
Date: Monday, January 6, 2020 at 7:22 PM
To: "vpp-dev@lists.fd.io" 
Subject: Re: [vpp-dev] "vppctl show int" no NIC (just local0) #vpp #vnet

Hi Ezpeer :
Thank you for you advice.
I did a test to update X710's driver(i40e) and X710's Firmware:
i40e's new version : 2.10.19.30
Firmware's new version : 6.80  (inter delete NVM's 7.0 and 7.1 verison 
files because they introduced some serious errors).
(I will try again when inter republish NVM's 7.1).
UIO-driver use vfio-pci.
Even so, NIC still have no driver when use "vppctl show pci".
I aslo switch UIO-driver to uio_pci_generic by modidfy vpp.service and 
startup.conf, the result has little different but still not OK.

This is my environment:
[root@localhost i40e]# cat /etc/redhat-release
CentOS Linux release 7.7.1908 (Core)

[root@localhost i40e]# uname -a
Linux localhost.localdomain 3.10.0-1062.4.1.el7.x86_64 #1 SMP Fri Oct 18 
17:15:30 UTC 2019 x86_64 x86_64 x86_64 GNU/Linux
[root@localhost i40e]# uname -r
3.10.0-1062.4.1.el7.x86_64

[root@localhost i40e]# modinfo i40e
filename:   
/lib/modules/3.10.0-1062.4.1.el7.x86_64/updates/drivers/net/ethernet/intel/i40e/i40e.ko
version:2.10.19.30
license:GPL
description:Intel(R) 40-10 Gigabit Ethernet Connection Network Driver
author: Intel Corporation, 
retpoline:  Y
rhelversion:7.7
srcversion: 9EB781BDF574D047F098566
alias:  pci:v8086d158Bsv*sd*bc*sc*i*
alias:  pci:v8086d158Asv*sd*bc*sc*i*
alias:  pci:v8086d37D3sv*sd*bc*sc*i*
alias:  pci:v8086d37D2sv*sd*bc*sc*i*
alias:  pci:v8086d37D1sv*sd*bc*sc*i*
alias:  pci:v8086d37D0sv*sd*bc*sc*i*
alias:  pci:v8086d37CFsv*sd*bc*sc*i*
alias:  pci:v8086d37CEsv*sd*bc*sc*i*
alias:  pci:v8086d0D58sv*sd*bc*sc*i*
alias:  pci:v8086d0CF8sv*sd*bc*sc*i*
alias:  pci:v8086d1588sv*sd*bc*sc*i*
alias:  pci:v8086d1587sv*sd*bc*sc*i*
alias:  pci:v8086d104Fsv*sd*bc*sc*i*
alias:  pci:v8086d104Esv*sd*bc*sc*i*
alias:  pci:v8086d15FFsv*sd*bc*sc*i*
alias:  pci:v8086d1589sv*sd*bc*sc*i*
alias:  pci:v8086d1586sv*sd*bc*sc*i*
alias:  pci:v8086d1585sv*sd*bc*sc*i*
alias:  pci:v8086d1584sv*sd*bc*sc*i*
alias:  pci:v8086d1583sv*sd*bc*sc*i*
alias:  pci:v8086d1581sv*sd*bc*sc*i*
alias:  pci:v8086d1580sv*sd*bc*sc*i*
alias:  pci:v8086d1574sv*sd*bc*sc*i*
alias:  pci:v8086d1572sv*sd*bc*sc*i*
depends:ptp
vermagic:   3.10.0-1062.4.1.el7.x86_64 SMP mod_unload modversions
parm:   debug:Debug level (0=none,...,16=all) (int)

[root@localhost i40e]# ethtool -i p1p3
driver: i40e
version: 2.10.19.30
firmware-version: 6.80 0x80003c64 1.2007.0
expansion-rom-version:
bus-info: :3b:00.2
supports-statistics: yes
supports-test: yes
supports-eeprom-access: yes
supports-register-dump: yes
supports-priv-flags: yes

This is vfio-pci error:
[root@localhost ~]# lsmod | grep vfio
vfio_pci   41412  0
vfio_iommu_type1   22440  0
vfio   

Re: [vpp-dev] tx-errors on VPP controlled dpdk device

[Yichen Wang via Lists.Fd.Io]

Hi, Chakri,

You can change the VPP worker assignments by doing:
vpp# show interface rx-placement
vpp# set interface rx-placement ?
  set interface rx-placement   set interface rx-placement 
 [queue ] [worker  | main]

Thanks very much!

Regards,
Yichen

From:  on behalf of "chakravarthy.arise...@viasat.com" 

Date: Wednesday, August 1, 2018 at 11:22 AM
To: "vpp-dev@lists.fd.io" 
Subject: Re: [vpp-dev] tx-errors on VPP controlled dpdk device

Hi Damjan,

Thanks for your feedback. I'm running the test in AWS instances. Thus, I have 
got only VFs. I do not have access to PF. So, I'm trying to get help from AWS 
to find out.
Once I get the info, I'll post it over here. In the mean time, I looked at the 
counters that you suggested me to focus on. It looks like the packets are 
scheduled on only one core in transmit direction. Is there a way to change?

I have 3 dedicated cores (1 main core thread for stats/mgmt and 2 cores for the 
worker threads). All the Tx queues are pinned to worker thread 1. So, worker 
thread 2 is not used for transmit path at all. Is there way to spread the 
transmit queues across the threads?

Thanks
Chakri
vpp# sh threads
ID NameTypeLWPSched Policy (Priority)  
lcore  Core   Socket State
0  vpp_main  1733other (0)1 
 1  0
1  vpp_wk_0workers 1745other (0)2  
2  0
2  vpp_wk_1workers 1746other (0)3  
3  0
3   stats 1747other (0) 
   0  0  0

vpp# sh run
Thread 0 vpp_main (lcore 1)
Time 5125.9, average vectors/node 0.00, last 128 main loops 0.00 per node 0.00
  vector rates in 0.e0, out 0.e0, drop 0.e0, punt 0.e0
 Name State Calls  Vectors
Suspends Clocks   Vectors/Call
api-rx-from-ringany wait 0   0  
   364  1.19e40.00
cdp-process any wait 0   0  
   992  1.98e30.00
dhcp-client-process any wait 0   0  
51  3.41e30.00
dns-resolver-processany wait 0   0  
 5  4.06e30.00
dpdk-processany wait 0   0  
  1709  5.13e40.00
fib-walkany wait 0   0  
  2563  1.37e30.00
ikev2-manager-process   any wait 0   0  
  5124  7.25e20.00
ip-route-resolver-process   any wait 0   0  
51  2.64e30.00
ip4-reassembly-expire-walk  any wait 0   0  
   513  3.85e30.00
ip6-icmp-neighbor-discovery-ev  any wait 0   0  
  5124  6.92e20.00
ip6-reassembly-expire-walk  any wait 0   0  
   513  3.84e30.00
lisp-retry-service  any wait 0   0  
  2563  1.57e30.00
memif-process   any wait 0   0  
  1709  2.10e30.00
rd-cp-process   any wait 0   0  
 237212380  3.21e20.00
unix-cli-local:17active  0   0  
   580  2.05e50.00
unix-epoll-input polling  96172305   0  
 0  1.19e40.00
vpe-oam-process any wait 0   0  
  2513  1.23e30.00
---
Thread 1 vpp_wk_0 (lcore 2)
Time 5125.9, average vectors/node 4.82, last 128 main loops 0.00 per node 0.00
  vector rates in 9.5578e3, out 8.4052e3, drop 0.e0, punt 0.e0
 Name State Calls  Vectors
Suspends Clocks   Vectors/Call
VirtualFunctionEthernet0/6/0-o   active 91  91  
 0  8.59e21.00
VirtualFunctionEthernet0/6/0-t   active 91  91  
 0  2.82e31.00
VirtualFunctionEthernet0/7/0-o   active533416432661561  
 0  4.33e16.12
VirtualFunctionEthernet0/7/0-t   active533416426753703  
 0  3.83e25.02
arp-inputactive182 182  
 0  7.25e3

Re: [vpp-dev] tx-drops with vhost-user interface

[Yichen Wang via Lists.Fd.Io]

Hi, Vijay,

Sorry to ask dumb question, can you make sure the interface in your VM (either 
Linux Kernel or DPDK) is “UP”?

Regards,
Yichen

From:  on behalf of "steven luong via Lists.Fd.Io" 

Reply-To: "Steven Luong (sluong)" 
Date: Monday, August 6, 2018 at 12:10 PM
To: "Chandra Mohan, Vijay Mohan" , "vpp-dev@lists.fd.io" 

Cc: "vpp-dev@lists.fd.io" 
Subject: Re: [vpp-dev] tx-drops with vhost-user interface

Vijay,

From the show output, I can’t really tell what your problem is. If you could 
provide additional information about your environment, I could try setting it 
up and see what’s wrong. Things I need from you are exact VPP version, VPP 
configuration, qemu startup command line or the XML startup file if you use 
virsh, and the version of the VM distro.

Steven

From:  on behalf of "Chandra Mohan, Vijay Mohan" 

Date: Monday, August 6, 2018 at 10:31 AM
To: "vpp-dev@lists.fd.io" 
Subject: [vpp-dev] tx-drops with vhost-user interface

Hi,

I am trying to pass traffic with vhost-user interface and seeing tx-drops on 
virtual interface. Here is the setup: created a bridge domain with a physical 
interface and a vhost-user interface. Physical interface GigabitEthernet5/0/0 
is connected to traffic generator. As shown below, observing drops on  
VirtualEthernet0/0/0 .

Following is the config and vhost-user commands o/p:

DBGvpp# show bridge-domain 1 detail
  BD-ID   Index   BSN  Age(min)  Learning  U-Forwrd  UU-Flood  Flooding  
ARP-Term  BVI-Intf
1   1  0 offonononon   off  
 N/A

   Interface   If-idx ISN  SHG  BVI  TxFlood
VLAN-Tag-Rewrite
 GigabitEthernet5/0/03 10-  * none
 GigabitEthernet5/0/14 10-  * none
 VirtualEthernet0/0/05 10-  * none

Virtual interface is operationally up. Connected to virtual interface server in 
VM.

DBGvpp# show hardware-interfaces VirtualEthernet0/0/0
  NameIdx   Link  Hardware
VirtualEthernet0/0/0   5 up   VirtualEthernet0/0/0
  Ethernet address 02:fe:98:19:c2:6b

DBGvpp# show interface VirtualEthernet0/0/0

  Name   Idx   State  Counter  Count

VirtualEthernet0/0/0  5 up   tx packets 
1

 tx bytes   
   60

 drops  
1





DBGvpp# show errors

   CountNode  Reason

 3l2-output   L2 output packets

 2l2-learnL2 learn packets

 2l2-learnL2 learn misses

 2l2-inputL2 input packets

 3l2-floodL2 flood packets

 1 VirtualEthernet0/0/0-txtx packet drops (no available 
descriptors)


DBGvpp# show vhost-user VirtualEthernet0/0/0
Virtio vhost-user interfaces
Global:
  coalesce frames 32 time 1e-3
  number of rx virtqueues in interrupt mode: 0
Interface: VirtualEthernet0/0/0 (ifindex 5)
virtio_net_hdr_sz 12
features mask (0x):
 features (0x150208000):
   VIRTIO_NET_F_MRG_RXBUF (15)
   VIRTIO_NET_F_GUEST_ANNOUNCE (21)
   VIRTIO_F_INDIRECT_DESC (28)
   VHOST_USER_F_PROTOCOL_FEATURES (30)
   VIRTIO_F_VERSION_1 (32)
  protocol features (0x3)
   VHOST_USER_PROTOCOL_F_MQ (0)
   VHOST_USER_PROTOCOL_F_LOG_SHMFD (1)

socket filename /socket/vnet-0 type client errno "Success"

rx placement:
   thread 0 on vring 1, polling
tx placement: lock-free
   thread 0 on vring 0

Memory regions (total 3)
region fdguest_phys_addrmemory_sizeuserspace_addr 
mmap_offsetmmap_addr
== = == == == 
== ==
  0 500x0001 0x0001c000 0x7fe39360 
0xc000 0x7fc0e4a0
  1 510x 0x000a 0x7fe2d360 
0x 0x7fc02480
  2 520x000c 0xbff4 0x7fe2d36c 
0x000c 0x7fbf648c

Virtqueue 0 (TX)
  qsz 256 last_avail_idx 0 last_used_idx 0
  avail.flags 0 avail.idx 0 used.flags 1 used.idx 0
  kickfd 53 callfd 54 errfd -1

Virtqueue 1 (RX)
  qsz 256 last_avail_idx 0 last_used_idx 0
  avail.flags 0 avail.idx 0 used.flags 1 used.idx 0
  kickfd 46 callfd 55 errfd -1

Tried dumping descriptors from Rx queue and didn’t find any entries. It’s all 
zeros.

Any idea what is going on here ?

Thanks,
Vijay

-=-=-=-=-=-=-=-=-=-=-=-
Links: You receive all messages sent to this group.

View/Reply Online (#10055): https://lists.fd.io/g/vpp-dev/message/10055
Mute This Topic: https://lists.fd.io/mt/24211191/21656
Gr

Re: [vpp-dev] packet capture

[Yichen Wang via Lists.Fd.Io]

You can create a span ports, with SRC interface is your VirtualEthernet, and 
DST interface to be a tap interface. Then you run easily run tcpdump from 
kernel side for your packet capture.

Thanks very much!

Regards,
Yichen

On Nov 13, 2018, at 09:36, Bly, Mike mailto:m...@ciena.com>> 
wrote:


Ed, et al,

pcap works fine for DPDK controlled NICs, and I am assuming tcpdump can still 
be used for tap based internal connections, but what can be used for vhost-user 
interfaces?

Regards,

Mike

-=-=-=-=-=-=-=-=-=-=-=-
Links: You receive all messages sent to this group.

View/Reply Online (#11225): https://lists.fd.io/g/vpp-dev/message/11225
Mute This Topic: https://lists.fd.io/mt/10642670/675839
Group Owner: vpp-dev+ow...@lists.fd.io
Unsubscribe: https://lists.fd.io/g/vpp-dev/unsub  
[yicw...@cisco.com]
-=-=-=-=-=-=-=-=-=-=-=-
-=-=-=-=-=-=-=-=-=-=-=-
Links: You receive all messages sent to this group.

View/Reply Online (#11227): https://lists.fd.io/g/vpp-dev/message/11227
Mute This Topic: https://lists.fd.io/mt/10642670/21656
Group Owner: vpp-dev+ow...@lists.fd.io
Unsubscribe: https://lists.fd.io/g/vpp-dev/unsub  [arch...@mail-archive.com]
-=-=-=-=-=-=-=-=-=-=-=-

Re: [vpp-dev] Andrew Yourtchenko is now a vpp project committer!

[Yichen Wang via Lists.Fd.Io]

Congratulations, Andrew! ☺

Regards,
Yichen

From:  on behalf of Andrew Yourtchenko 
Date: Thursday, November 15, 2018 at 09:23
To: "Dave Barach (dbarach)" 
Cc: "vpp-dev@lists.fd.io" 
Subject: Re: [vpp-dev] Andrew Yourtchenko is now a vpp project committer!

Thank you very much Dave and all the committers and TSC!

Looking forward to continue enjoying working with this amazing group of 
talented people !
--a

On 15 Nov 2018, at 17:43, Dave Barach via Lists.Fd.Io 
mailto:dbarach=cisco@lists.fd.io>> wrote:
The TSC formally approved Andrew’s vpp committer nomination a short while ago.

Congratulations, and thanks in advance for your participation as a committer.

Dave

-=-=-=-=-=-=-=-=-=-=-=-
Links: You receive all messages sent to this group.

View/Reply Online (#11261): https://lists.fd.io/g/vpp-dev/message/11261
Mute This Topic: https://lists.fd.io/mt/28147575/675608
Group Owner: vpp-dev+ow...@lists.fd.io
Unsubscribe: https://lists.fd.io/g/vpp-dev/unsub  
[ayour...@gmail.com]
-=-=-=-=-=-=-=-=-=-=-=-
-=-=-=-=-=-=-=-=-=-=-=-
Links: You receive all messages sent to this group.

View/Reply Online (#11267): https://lists.fd.io/g/vpp-dev/message/11267
Mute This Topic: https://lists.fd.io/mt/28147575/21656
Group Owner: vpp-dev+ow...@lists.fd.io
Unsubscribe: https://lists.fd.io/g/vpp-dev/unsub  [arch...@mail-archive.com]
-=-=-=-=-=-=-=-=-=-=-=-

Re: [vpp-dev] RPM package question related to https://gerrit.fd.io/r/#/c/18953/

[Yichen Wang via Lists.Fd.Io]

Hi, Tom,

epel-release is nothing but some yum repo files, so technically it should not 
bring any code for dependencies. So it should be removed from runtime 
dependencies for sure IMO. The reason is, we want to stick with whatever RedHat 
supports. When we bring in epel-release, the RPMs that we are going to build 
will all be messed, and I don’t think RedHat will support any versions that we 
pull from epel-release. We can survive with build time, as that is safe in a 
virtual environment, and technically we are allowed to mess up with anything.

Python3.6 is another problem. There is no problem for us to have Python 3.6, as 
this is technically supported and shipped by RedHat, and our subscription does 
have it. However, the question is more about “Is that really needed?”. As you 
mention, if this is implicitly needed by vpp-devel, we can just put that for 
vpp-devel RPM, right? Normally we don’t install vpp-devel as well. Python3.6 is 
just a question that will save us time, as majority of code is still in Python 
2.7, and pulling Python3.6 may or may not cause any issue. But if not needed, 
it will make our life easier. That’s it. 😊

Thanks very much!

Regards,
Yichen

From:  on behalf of Thomas F Herbert 
Date: Friday, May 10, 2019 at 12:26 PM
To: "vpp-dev@lists.fd.io" 
Subject: Re: [vpp-dev] RPM package question related to 
https://gerrit.fd.io/r/#/c/18953/


Jerome,

python3 is an implicit dependency of vpp-devel and I am not sure why.  Because 
of that I put python36 explicitly in Requires because I was concerned about 
cross versions of Python between 34 and 36 having side affects.

As for epel-release, it is harmless but is not strictly required installation 
time but I wanted to make sure the epel repo was installed so further 
dependency checks were resolved from the correct repo.

--Tom

On 05/10/2019 10:20 AM, Thomas F Herbert wrote:

Jerome,

Let me take a look at this today. I will submit a patch to take them out if I 
can verify that they are not needed.

--Tom

On 05/10/2019 05:11 AM, Jerome Tollet (jtollet) wrote:
Hi Thomas,

In https://gerrit.fd.io/r/#/c/18953/ you are putting epel-release and Python3 
into the dependencies.

  1.  I can understand that those are needed as BuildRequires:, but are they 
really needed for Requires: ?
  2.  Why do we need epel-release as runtime requirement?
  3.  Do we really need Python3 for VPP 19.04 itself?

Jerome




--
Thomas F Herbert
NFV and Fast Data Planes
Networking Group Office of the CTO
Red Hat




-=-=-=-=-=-=-=-=-=-=-=-

Links: You receive all messages sent to this group.



View/Reply Online (#12980): https://lists.fd.io/g/vpp-dev/message/12980

Mute This Topic: https://lists.fd.io/mt/31575010/675058

Group Owner: vpp-dev+ow...@lists.fd.io

Unsubscribe: https://lists.fd.io/g/vpp-dev/unsub  
[therb...@redhat.com]

-=-=-=-=-=-=-=-=-=-=-=-

--
Thomas F Herbert
NFV and Fast Data Planes
Networking Group Office of the CTO
Red Hat
-=-=-=-=-=-=-=-=-=-=-=-
Links: You receive all messages sent to this group.

View/Reply Online (#12996): https://lists.fd.io/g/vpp-dev/message/12996
Mute This Topic: https://lists.fd.io/mt/31575010/21656
Group Owner: vpp-dev+ow...@lists.fd.io
Unsubscribe: https://lists.fd.io/g/vpp-dev/unsub  [arch...@mail-archive.com]
-=-=-=-=-=-=-=-=-=-=-=-

Re: [vpp-dev] wireguard_handoff.c:112:12: error: 'peeri' may be used uninitialized

[Yichen Wang via lists.fd.io]

You need this: https://gerrit.fd.io/r/c/vpp/+/29084, and merged in stable/2009 
already.

Regards,
Yichen

From:  on behalf of "Heckman, Steve" 

Date: Friday, September 25, 2020 at 1:04 PM
To: "vpp-dev@lists.fd.io" 
Subject: [vpp-dev] wireguard_handoff.c:112:12: error: 'peeri' may be used 
uninitialized

I had been building in the master branch fine for a while. We need 20.09, so I 
saw that the branch was available and checked it out.

I ran onto this building on Ubuntu Focal (20.04):

$ git fetch
remote: Enumerating objects: 60, done.
remote: Counting objects: 100% (55/55), done.
remote: Compressing objects: 100% (28/28), done.
remote: Total 30 (delta 26), reused 6 (delta 2), pack-reused 0
Unpacking objects: 100% (30/30), 3.05 KiB | 283.00 KiB/s, done.
From https://github.com/FDio/vpp
   561ae5d17..277d402be  master  -> origin/master
   df143bf70..a26b40af9  stable/2009 -> origin/stable/2009

$ make pkg-deb
make[1]: Entering directory '/home/labuser/Development/git/vpp/build-root'
 Arch for platform 'vpp' is native 
 Finding source for external 
 Makefile fragment found in 
/home/labuser/Development/git/vpp/build-data/packages/external.mk 
 Source found in /home/labuser/Development/git/vpp/build 
 Arch for platform 'vpp' is native 
 Finding source for vpp 
 Makefile fragment found in 
/home/labuser/Development/git/vpp/build-data/packages/vpp.mk 
 Source found in /home/labuser/Development/git/vpp/src 
 Configuring external: nothing to do 
 Building external: nothing to do 
 Installing external: nothing to do 
 Configuring vpp: nothing to do 
 Building vpp in 
/home/labuser/Development/git/vpp/build-root/build-vpp-native/vpp 
[11/18] Building C object 
plugins/wireguard/CMakeFiles/wireguard_plugin.dir/wireguard_handoff.c.o
FAILED: plugins/wireguard/CMakeFiles/wireguard_plugin.dir/wireguard_handoff.c.o
ccache /usr/lib/ccache/gcc-9 -DHAVE_FCNTL64 -DHAVE_GETCPU -DHAVE_MEMFD_CREATE 
-Dwireguard_plugin_EXPORTS -I/home/labuser/Development/git/vpp/src -I. 
-Iinclude -I/home/labuser/Development/git/vpp/src/plugins -Iplugins 
-Wno-address-of-packed-member -g -fPIC -Werror -Wall -march=corei7 
-mtune=corei7-avx  -O2 -fstack-protector -DFORTIFY_SOURCE=2 -fno-common  -fPIC 
-MD -MT plugins/wireguard/CMakeFiles/wireguard_plugin.dir/wireguard_handoff.c.o 
-MF plugins/wireguard/CMakeFiles/wireguard_plugin.dir/wireguard_handoff.c.o.d 
-o plugins/wireguard/CMakeFiles/wireguard_plugin.dir/wireguard_handoff.c.o   -c 
/home/labuser/Development/git/vpp/src/plugins/wireguard/wireguard_handoff.c
/home/labuser/Development/git/vpp/src/plugins/wireguard/wireguard_handoff.c: In 
function 'wg_handshake_handoff_fn':
/home/labuser/Development/git/vpp/src/plugins/wireguard/wireguard_handoff.c:112:12:
 error: 'peeri' may be used uninitialized in this function 
[-Werror=maybe-uninitialized]
  112 |t->peer = peeri;
  |^~~
/home/labuser/Development/git/vpp/src/plugins/wireguard/wireguard_handoff.c:82:15:
 note: 'peeri' was declared here
   82 |   index_t peeri;
  |   ^
cc1: all warnings being treated as errors
[12/18] Linking C shared library lib/libvnet.so.20.09
ninja: build stopped: subcommand failed.
make[1]: *** [Makefile:693: vpp-build] Error 1
make[1]: Leaving directory '/home/labuser/Development/git/vpp/build-root'
make: *** [Makefile:560: pkg-deb] Error 2

Anyone else see that?

Thanks,
Steve Heckman
Distinguished Software Engineer
Commscope, Inc.
Lowell, Ma




-=-=-=-=-=-=-=-=-=-=-=-
Links: You receive all messages sent to this group.
View/Reply Online (#17531): https://lists.fd.io/g/vpp-dev/message/17531
Mute This Topic: https://lists.fd.io/mt/77124338/21656
Group Owner: vpp-dev+ow...@lists.fd.io
Unsubscribe: https://lists.fd.io/g/vpp-dev/unsub [arch...@mail-archive.com]
-=-=-=-=-=-=-=-=-=-=-=-

Re: [vpp-dev] VPP on ESXI with i40evf (SR-IOV Passthrough) Driver

[Yichen Wang via lists.fd.io]

I am not sure if it worth to try to add “iommu=pt intel_iommu=on” in the GRUB? 
We normally do it in both Linux host/guest, but in the case of ESXi not sure if 
that will help with the vfio-pci IOMMU issue…

Regards,
Yichen

From:  on behalf of "Damjan Marion via lists.fd.io" 

Reply-To: "dmar...@me.com" 
Date: Tuesday, September 29, 2020 at 7:48 AM
To: Joshua Moore 
Cc: "Benoit Ganne (bganne)" , "vpp-dev@lists.fd.io" 

Subject: Re: [vpp-dev] VPP on ESXI with i40evf (SR-IOV Passthrough) Driver

Joshua,

What is your motivation for using IOMMU inside VM?
It is typically used by hypervisor to protect from misbehaving VMs.
So unless you want to do nested virtualisation, IOMMU inside your VM doesn’t 
bring lot of value.
Can you simply try to turn off iommu in the VM and turn 
“enable_unsafe_noiommu_mode”.

Another possible problem with AVF may be version of ESXi PF driver. AVF 
communicates
with PF over the virtua lchannel and PF driver needs to support it. In linux, 
that works well with recent kernels
but I’m not sure what is the state of ESXi i40e PF driver….

—
Damjan

On 29.09.2020., at 14:04, Joshua Moore mailto:j...@jcm.me>> wrote:

Ben,

Of course there's a PCI switch. My main purpose of leveraging SR-IOV with VF 
allocation is to allow the internal eswitch on the Intel NIC to handle 
switching in hardware instead of the vswitch on the ESXI hypervisor. I don't 
really care so much about isolation of the PCI devices nor the risk of bad 
firmware on the NIC. I will control/trust all of the VMs with access to the VFs 
as well as the device attached to the PF.

So just to confirm, I need to expect 100% CPU utilization with VPP/DPDK + 
IOMMU? If so, what's the best way to monitor CPU-related performance impact if 
I always see 100%? Also I want to confirm that enable_unsafe_noiommu_mode still 
enables the performance benefits of SR-IOV and the only tradeoff is the 
aforementioned isolation/security concern?


Thanks for your help,

--Josh

On Tue, Sep 29, 2020 at 5:23 AM Benoit Ganne (bganne) 
mailto:bga...@cisco.com>> wrote:
Hi Joshua,

Glad it solves the vfio issue. Looking at the dmesg output, I suspect the issue 
is that the PCIe topology advertised is not fully supported by vfio + IOMMU: it 
looks like your VF is behind a PCIe switch, so the CPU PCIe IOMMU root-complex 
port cannot guarantee full isolation: all devices behind the PCIe switch can 
talk peer-to-peer directly w/o going through the CPU PCIe root-complex port.
As the CPU IOMMU cannot fully isolate your device, vfio refuses to bind unless 
you allow for unsafe IOMMU config - rightfully, as it seems to be your case.
Anyway, it still means you should benefit from the IOMMU to prevent the device 
to read/write everywhere in host memory. You might not however prevent a malign 
firmware running on the NIC to harm other devices behind the same PCIe switch.

Regarding the VM crash, note that VPP is polling the interfaces so it will 
always uses 100% CPU.
Does the VM also crashes if you run stress-test the CPU eg.
~# stress-ng --matrix 0 -t 1m

Best
ben

> -Original Message-
> From: Joshua Moore mailto:j...@jcm.me>>
> Sent: mardi 29 septembre 2020 12:07
> To: Benoit Ganne (bganne) mailto:bga...@cisco.com>>
> Cc: Damjan Marion mailto:dmar...@me.com>>; 
> vpp-dev@lists.fd.io
> Subject: Re: [vpp-dev] VPP on ESXI with i40evf (SR-IOV Passthrough) Driver
>
> Hello Ben,
>
> echo 1 | sudo tee /sys/module/vfio/parameters/enable_unsafe_noiommu_mode
> sudo dpdk-devbind --bind=vfio-pci :13:00.0
>
>
> The above commands successfully resulted in vfio-pci driver binding to the
> NIC. However, as soon as I assigned the NIC to VPP and restarted the
> service, my VM CPU shot up and the VM crashes.
>
>
> Regarding IOMMU I do have it enabled in the host's BIOS, ESXI "Expose
> IOMMU to the guest OS" option, and I have set the GRUB_CMDLINE_LINUX per
> the below wiki:
>
> https://wiki.fd.io/view/VPP/How_To_Optimize_Performance_(System_Tuning)
>
> root@test:~# cat /etc/default/grub | grep GRUB_CMDLINE_LINUX
> GRUB_CMDLINE_LINUX_DEFAULT="maybe-ubiquity"
> GRUB_CMDLINE_LINUX="intel_iommu=on isolcpus=1-7 nohz_full=1-7
> hugepagesz=1GB hugepages=16 default_hugepagesz=1GB"
>
> Full dmesg output can be found at: http://jcm.me/dmesg.txt




-=-=-=-=-=-=-=-=-=-=-=-
Links: You receive all messages sent to this group.
View/Reply Online (#17588): https://lists.fd.io/g/vpp-dev/message/17588
Mute This Topic: https://lists.fd.io/mt/77164974/21656
Group Owner: vpp-dev+ow...@lists.fd.io
Unsubscribe: https://lists.fd.io/g/vpp-dev/unsub [arch...@mail-archive.com]
-=-=-=-=-=-=-=-=-=-=-=-