Re: [vpp-dev] undefined refrence to

Gld doesn’t know that e.g. vpp_api.o needs e.g. format until after it’s already processed -lvppinfra. Reorder the command line. From: vpp-dev-boun...@lists.fd.io [mailto:vpp-dev-boun...@lists.fd.io] On Behalf Of Samuel S Sent: Monday, December 11, 2017 2:05 AM To: vpp-dev@lists.fd.io Subject:

Re: [vpp-dev] Build error when trying to cross-compile vpp

Look in config.log and work out the name of the compiler. Fix in .../build-data/platforms/x86_64.mk or override from the command line. From: nikhil ap [mailto:niks3...@gmail.com] Sent: Sunday, December 10, 2017 8:43 AM To: Dave Barach (dbarach) <dbar...@cisco.com> Cc: vpp-dev@lists.fd.io S

Re: [vpp-dev] MACIP ACL replace causes ip4_table_index change?

Folks will miss a clib_warning, unless they check syslog. I'd consider returning VNET_API_ERROR_ENTRY_ALREADY_EXISTS and calling it a day. Thanks… Dave -Original Message- From: Andrew  Yourtchenko [mailto:ayour...@gmail.com] Sent: Sunday, December 10, 2017 9:04 AM To: Dave Barach

Re: [vpp-dev] how to find vlib_buffer_t.data size(capacity) @ runtime?

Interpret b->current_data, b->current_length, the buffer freelist index, and the related vlib_buffer_free_list_t structure. In most cases, b->packet_data is actually VLIB_BUFFER_DATA_SIZE (2048) bytes long. Look at the related vlib_buffer_free_list_t to know for sure. Current_data is a

Re: [vpp-dev] memory issues

Before we crank up the vppinfra memory leakfinder, etc. etc.: cat /proc/`pidof vpp`/maps and have a hard stare at the output. Configure one step at a time, looking for significant changes in the address space layout. HTH… Dave From: vpp-dev-boun...@lists.fd.io

Re: [vpp-dev] some files are never compiled

Merged... I’ll clean out some more junk and push another patch... Thanks… Dave From: Gabriel Ganne [mailto:gabriel.ga...@enea.com] Sent: Tuesday, December 5, 2017 10:14 AM To: Dave Barach (dbarach) <dbar...@cisco.com>; vpp-dev@lists.fd.io Subject: Re: some files are never compiled Thank

Re: [vpp-dev] Build error when trying to cross-compile vpp

See also “bootstrap.sh...” $ make V=0 is_build_tool=yes tools-install Thanks… Dave From: nikhil ap [mailto:niks3...@gmail.com] Sent: Tuesday, December 5, 2017 9:11 AM To: Dave Barach (dbarach) <dbar...@cisco.com> Cc: vpp-dev@lists.fd.io Subject: Re: [vpp-dev] Build error when trying to

Re: [vpp-dev] some files are never compiled

Dear Gabriel, The files mentioned below fall into several buckets: * Code samples which might reasonably move to .../extras * Things we’re not using at the moment, but which would take someone a good long time to build from scratch. * The simulated annealing driver in

Re: [vpp-dev] Build error when trying to cross-compile vpp

Dear Nikhil, The first step in adding a new platform: construct .../build-data/plaforms/xxx.mk. There are several examples. Note the rule: xxx_native_tools = vppapigen This rule builds the missing build-host tool. Then: “make PLATFORM=xxx TAG=xxx_debug vpp-install” or similar. Caveat: the

Re: [vpp-dev] Question on node type "VLIB_NODE_TYPE_PROCESS"

At least for now, process nodes run on the main thread. See line 1587 of .../src/vlib/main.c. The lldp-process is not super-complicated. Set a gdb breakpoint on line 157 [switch(event_type)], cause it to do something, and you can walk through it, etc. HTH... Dave From:

Re: [vpp-dev] problem in elog format

: Thursday, November 30, 2017 12:50 AM To: Dave Barach (dbarach) <dbar...@cisco.com> Cc: Florin Coras <fcoras.li...@gmail.com>; vpp-dev@lists.fd.io Subject: Re: [vpp-dev] problem in elog format Thanks a lot, Now I want to convert elog file to text file. I compiled perftools in te

[vpp-dev] Frequently-asked questions wiki page

Folks, Please see https://wiki.fd.io/view/VPP/FAQ. Additions welcome. I decided to start with a personal favorite... Thanks... Dave ___ vpp-dev mailing list vpp-dev@lists.fd.io https://lists.fd.io/mailman/listinfo/vpp-dev

Re: [vpp-dev] problem in elog format

PMFJI, but we have organized schemes for capturing, serializing, and eventually displaying string data. Please note: a single "format" call will probably cost more than the entire clock-cycle budget available to process a packet. Really. Seriously. Printfs (aka format calls) in the

Re: [vpp-dev] How to enable RSS in VPP

You are sending traffic with more than one flow, correct? From: vpp-dev-boun...@lists.fd.io [mailto:vpp-dev-boun...@lists.fd.io] On Behalf Of Saxena, Nitin Sent: Tuesday, November 28, 2017 11:45 AM To: vpp-dev@lists.fd.io Cc: Athreya, Narayana Prasad Subject:

Re: [vpp-dev] VPP crash observed with 4k sub-interfaces and 4k FIBs

Laying aside the out-of-memory issue for a minute: can you explain the vpp deployment you have in mind? Given where vpp would fit in a normal network design, I’m not seeing why you’d want to go with a full vlan / VRF’s mesh. Thanks… Dave From: vpp-dev-boun...@lists.fd.io

Re: [vpp-dev] Anyone using the l2 classify path these days?

...@lists.fd.io] On Behalf Of Andrew Yourtchenko Sent: Monday, November 20, 2017 1:37 PM To: Dave Barach <d...@barachs.net> Cc: vpp-dev@lists.fd.io Subject: Re: [vpp-dev] Anyone using the l2 classify path these days? Dear Dave, ACL plugin is using l2 classify path for traffic diversion an

[vpp-dev] Anyone using the l2 classify path these days?

Dear John, I'm wondering about copying e0->metadata into vlib_buffer (b)->l2_classify.xxx in l2_input_classify. On the face of it, that seems reasonable. Do you see obvious problems with that idea? Related question, for vpp-dev: is anyone using L2 input or output classification these days?

Re: [vpp-dev] Import/includes in .api files

Would some variant of the usual C / C++ guitar lick work? #ifndef __defined_my_types #define __defined_my_types #include #endif /* __defined_my_types */ -Original Message- From: Ole Troan [mailto:otr...@employees.org] Sent: Monday, November 20, 2017 10:32 AM To: Dave Barach (dbarach

Re: [vpp-dev] Import/includes in .api files

Dear Neale, Since the beginning of time, we've been running .api files through the C preprocessor. Put all of your "typeonly..." definitions in a file, and #include it. Should work immediately. Thanks to Damjan, there's only one copy of the suffix rule, in .../src/suffix-rules.mk. Here's the

Re: [vpp-dev] Please Call DigSafe...

about that?" Guidelines seem like a good idea. I'll try to write something on the wiki. Thanks... Dave -Original Message- From: vpp-dev-boun...@lists.fd.io [mailto:vpp-dev-boun...@lists.fd.io] On Behalf Of Luke, Chris Sent: Friday, November 17, 2017 8:51 AM To: Dave Barach <d...@barachs.

[vpp-dev] Please Call DigSafe...

Folks, At our next project meeting, I'd like to spend a few minutes talking about a good-news / bad-news situation affecting the vpp project. As the community has expanded, committers have begun noticing unacceptable and unfixable patches in mission-critical code. Yesterday's soap-opera episode

Re: [vpp-dev] Discussion Topic: creating demo branches in git.fd.io/vpp

+1... -Original Message- From: Ole Troan [mailto:otr...@employees.org] Sent: Wednesday, November 15, 2017 5:49 PM To: Dave Wallace <dwallac...@gmail.com> Cc: Ed Warnicke <hagb...@gmail.com>; Dave Barach (dbarach) <dbar...@cisco.com>; Keith Burns (krb) <k...@cis

Re: [vpp-dev] Discussion Topic: creating demo branches in git.fd.io/vpp

? Thanks... Dave From: Dave Wallace [mailto:dwallac...@gmail.com] Sent: Wednesday, November 15, 2017 5:17 PM To: Ed Warnicke <hagb...@gmail.com>; Dave Barach (dbarach) <dbar...@cisco.com>; Keith Burns (krb) <k...@cisco.com>; Florin Coras (fcoras) <fco...@cisco.com>; John

Re: [vpp-dev] TCP Options: tcp_header_t and tcp_options_t

Dear Justin, Brief commercial: hopefully you added your node to the ip4 unicast feature arc, configured to grab pkts, pre-ip4/6-lookup. In feature-arc land, the following one-liner sets next0 so pkts will visit the next enabled feature. The last node in the ip4-unicast feature arc is

Re: [vpp-dev] vlib_validate_buffer_enqueue

Dear Justin, Quad-loops are generally not effective for table-lookup-intensive tasks. At a certain point, gcc runs out of registers and starts putting hot variables onto the stack. I've converted a number of dual loops into quad loops, only to discover that they're no faster than the dual loop

Re: [vpp-dev] make test-all

Try increasing the size of the shared-memory API segment. An allocation of 25mb is failing. You might ask yourself how sane it is to generate that much output. Thanks… Dave -Original Message- From: vpp-dev-boun...@lists.fd.io [mailto:vpp-dev-boun...@lists.fd.io] On Behalf Of Klement

Re: [vpp-dev] multi-core multi-threading performance

Please write up what you’ve done, and provide a pointer to your code. Thanks… Dave From: Pragash Vijayaragavan [mailto:pxv3...@rit.edu] Sent: Wednesday, November 8, 2017 1:19 AM To: Dave Barach (dbarach) <dbar...@cisco.com> Cc: vpp-dev@lists.fd.io; John Marshall (jwm) <j...@cisco.co

Re: [vpp-dev] Simple setup, that does not work.

Check host interface IP address, basic connectivity [cable on floor?], and so on. Check “show hardware.” If the MIB stats indicate that packets are reaching the NIC MAC layer - but not VPP - see if /proc/cmdline contains “intel_iommu=on”. If it does, try removing that stanza and reboot. You

Re: [vpp-dev] multi-core multi-threading performance

How many RX queues did you provision? One per worker, or no supper... Thanks… Dave From: Pragash Vijayaragavan [mailto:pxv3...@rit.edu] Sent: Monday, November 6, 2017 7:36 AM To: Dave Barach (dbarach) <dbar...@cisco.com> Cc: vpp-dev@lists.fd.io; John Marshall (jwm) <j...@cisco.com>;

Re: [vpp-dev] multi-core multi-threading performance

To: vpp-dev@lists.fd.io Cc: John Marshall (jwm) <j...@cisco.com>; Neale Ranns (nranns) <nra...@cisco.com>; Dave Barach (dbarach) <dbar...@cisco.com>; Minseok Kwon <mxk...@rit.edu> Subject: multi-core multi-threading performance Hi , We are measuring performance of ip6 lo

Re: [vpp-dev] VPP default graph

Dear Mostafa, First, “show vlib graph” describes the entire graph in detail. Vpp uses ingress flow-hashing (e.g. hardware RSS hashing) across a set of threads running identical graph replicas to achieve multi-core scaling. Historical experiments with pipelining in vpp dissuaded me from

Re: [vpp-dev] gerrit 8872 centos validation failure (stable/1710)

policer.lo' failed 14:09:30 make[5]: *** [vnet/policer/policer.lo] Error 1 14:09:30 Build step 'Execute shell' marked build as failure 14:09:30 $ ssh-agent -k 14:09:30 FATAL: Cannot run program "ssh-agent": error=12, Cannot allocate memory Thanks... Dave From: Dave Barach (dbarach) Sent: W

Re: [vpp-dev] VPP routing API dump

Confirmed. Patch on the way... Thanks... Dave From: vpp-dev-boun...@lists.fd.io [mailto:vpp-dev-boun...@lists.fd.io] On Behalf Of Dave Barach (dbarach) Sent: Wednesday, October 18, 2017 10:36 AM To: Samuel Elias -X (samelias - PANTHEON TECHNOLOGIES at Cisco) <samel...@cisco.com>; Neale

Re: [vpp-dev] VPP routing API dump

Please don't assume that people scrub through jira tickets. If you'd sent email to vpp-dev, this would have been fixed ages ago. Without even looking at the code, I'll bet it's a single missing ntohl(...): (gdb) p/x 33554432 $1 = 0x200 Thanks... Dave From: vpp-dev-boun...@lists.fd.io

Re: [vpp-dev] vnet_buffer(b)-> sw_if_index[VLIB_TX] => fib index in ip[46]_lookup

Dear John, I must've done a seriously defective [read: typo-ridden] search. Scratch that idea... Thanks for your help... Dave From: John Lo (loj) Sent: Wednesday, October 18, 2017 10:15 AM To: Dave Barach (dbarach) <dbar...@cisco.com>; vpp-dev@lists.fd.io Subject: RE: vnet_bu

[vpp-dev] gerrit 8872 centos validation failure (stable/1710)

Please see https://gerrit.fd.io/r/#/c/8872 and https://jenkins.fd.io/job/vpp-verify-1710-centos7/53. I've already pressed the "recheck" button. The validation failure appears unrelated to the patch. Thanks... Dave 12:50:49 Wrote:

[vpp-dev] vnet_buffer(b)-> sw_if_index[VLIB_TX] => fib index in ip[46]_lookup

Folks, Is anyone is actually using the "vnet_buffer(b)->sw_if_index[VLIB_TX] => [fib_index | ~0]" method to select the lookup fib index in ip[46]_lookup? If not, I would like to remove the corresponding code... Thanks... Dave ___ vpp-dev mailing

Re: [vpp-dev] not_last parameter of ip_add_del_route from ip.api

Adding Neale for further comment, but I believe it's a FIB 1.0 historical artifact which has no obvious reason to exist at this point. Thanks... Dave From: vpp-dev-boun...@lists.fd.io [mailto:vpp-dev-boun...@lists.fd.io] On Behalf Of Marek Gradzki -X (mgradzki - PANTHEON TECHNOLOGIES at Cisco)

Re: [vpp-dev] [FD.io Helpdesk #47101] No joy: ping6 gerrit.fd.io

Ack... Thanks… Dave -Original Message- From: Anton Baranov via RT [mailto:fdio-helpd...@rt.linuxfoundation.org] Sent: Tuesday, October 17, 2017 11:57 AM To: Dave Barach (dbarach) <dbar...@cisco.com> Cc: vpp-dev@lists.fd.io Subject: [FD.io Helpdesk #47101] No joy: ping6 gerrit

Re: [vpp-dev] [FD.io Helpdesk #47101] No joy: ping6 gerrit.fd.io

Ack... Thanks… Dave -Original Message- From: Anton Baranov via RT [mailto:fdio-helpd...@rt.linuxfoundation.org] Sent: Tuesday, October 17, 2017 11:57 AM To: Dave Barach (dbarach) <dbar...@cisco.com> Cc: vpp-dev@lists.fd.io Subject: [FD.io Helpdesk #47101] No joy: ping6 gerrit

Re: [vpp-dev] vpp deadlock - syslog in signal handler

In almost all cases, the glibc malloc heap will not be pickled since it's not used on a regular basis. For some effort, one could replace the syslog library code, I guess. Thanks... Dave From: vpp-dev-boun...@lists.fd.io [mailto:vpp-dev-boun...@lists.fd.io] On Behalf Of Gabriel Ganne Sent:

[vpp-dev] No joy: ping6 gerrit.fd.io

It looks like gerrit.fd.io has dropped off the ipv6 radar screen. Appears not to be a DNS problem or other problem on my end: $ ping6 gerrit.fd.io PING gerrit.fd.io(2604:e100:1:0:f816:3eff:fe7e:8731) 56 data bytes ^C --- gerrit.fd.io ping statistics --- 4 packets transmitted, 0 received, 100%

Re: [vpp-dev] jvpp core future test failure (gerrit 8743)

clib_warning ("Trying to register a NULL msg with a NULL msg id!"); clib_warning ("Did you forget to call setup_message_id_table?"); return; } Thanks… Dave From: vpp-dev-boun...@lists.fd.io [mailto:vpp-dev-boun...@lists.fd.io] On Behalf Of Dave Barach (dbara

Re: [vpp-dev] jvpp core future test failure (gerrit 8743)

Yes, I did. I just worked it out myself... Thanks… Dave From: Ole Troan (otroan) Sent: Friday, October 13, 2017 1:48 PM To: Ole Troan (otroan) <otr...@cisco.com> Cc: Dave Barach (dbarach) <dbar...@cisco.com>; Ole Troan <otr...@employees.org>; vpp-dev@lists.fd.io Subject: Re

[vpp-dev] jvpp core future test failure (gerrit 8743)

Dear Ole, See https://gerrit.fd.io/r/#/c/8743. It turns out that the java core future "make test" test fails as shown below. The patch adds three xxx_reply_t binary api messages. See .../src/vnet/dns/dns.api. It sure looks like the Java code knows about them, but isn't doing a very good job

Re: [vpp-dev] [v17.07.01]: vec_add2() causing crash on ARMv8

As a quick hack: try moving "u32 interrupt_pending;" to the start of the structure... Thanks… Dave -Original Message- From: Brian Brooks [mailto:brian.bro...@arm.com] Sent: Friday, September 22, 2017 12:33 PM To: Dave Barach (dbarach) <dbar...@cisco.com> Cc: Saxena,

Re: [vpp-dev] Compile error with linux/memfd.h

Dear Eddie, As discussed in private email: I think that the version of CentOS on your build system is too old. If memory serves, CentOS 7.3 is required. Google tells me that the earliest Linux kernel with memfd support is 3.17; it looks like your system is running a 3.10 derivative:

Re: [vpp-dev] vlan sub interfaces

See https://gerrit.fd.io/r/#/c/8590. The patch cherry-picked easily to stable/1707. Assuming that the cherry-pick patch validates - and that it solves your problem - it will be up to Neale [as the 17.07 release manager] whether to merge it or not. Please let us know whether the cherry-pick

Re: [vpp-dev] [v17.07.01]: vec_add2() causing crash on ARMv8

Dear Nitin, First off: exactly which LDXR / STXR instruction variant pairs is generated? I begin to wonder if __sync_lock_test_and_set(...) might not be doing you any favors. Given that dq->interrupt_pending is a u32, I would have expected a 4-byte instruction with (at worst) a 4-byte

Re: [vpp-dev] Stable branch for 17.10 pulled

+1... Thanks… Dave From: vpp-dev-boun...@lists.fd.io [mailto:vpp-dev-boun...@lists.fd.io] On Behalf Of Luke, Chris Sent: Wednesday, September 27, 2017 11:15 PM To: Florin Coras ; vpp-dev Subject: Re: [vpp-dev] Stable branch for 17.10 pulled Great

Re: [vpp-dev] deadlock issue in VPP during DHCP packet processing

Does this happen w/ master/latest? My guess: yes... Florin and I are working on a patch to fix an obvious issue in this path right now, look for results shortly... HTH... Dave From: vpp-dev-boun...@lists.fd.io [mailto:vpp-dev-boun...@lists.fd.io] On Behalf Of Balaji Kn Sent: Tuesday,

Re: [vpp-dev] Poor L3/L4 Performance

As discussed off-list: please stick to best-practice coding patterns. Single-packet frames simply cannot perform, etc. Thanks… Dave From: vpp-dev-boun...@lists.fd.io [mailto:vpp-dev-boun...@lists.fd.io] On Behalf Of Alessio Silvestro Sent: Monday, September 25, 2017 10:13 AM To:

Re: [vpp-dev] Practical use of the include-what-you-use tool for individual developers

Dear Burt, This is of interest, but I have concerns about boiling the ocean at this point in the release cycle. Please hold any patches on this topic until well after the 17.10 RC1 throttle branch pull. Although we haven’t caused ourselves massive pain with similar work - coding standards

Re: [vpp-dev] some issue about using unformat %u 

Varargs functions effectively bypass strong type-checking. It can’t be helped. From: vpp-dev-boun...@lists.fd.io [mailto:vpp-dev-boun...@lists.fd.io] On Behalf Of wang.hu...@zte.com.cn Sent: Tuesday, September 19, 2017 11:34 PM To: vpp-dev@lists.fd.io Cc: gu.ji...@zte.com.cn;

Re: [vpp-dev] Multiple vppctl Considered Harmful

Give me a minute, I'll try it right away... Thanks... Dave -Original Message- From: vpp-dev-boun...@lists.fd.io [mailto:vpp-dev-boun...@lists.fd.io] On Behalf Of Jon Loeliger Sent: Tuesday, September 19, 2017 2:01 PM To: vpp-dev Subject: [vpp-dev] Multiple vppctl

Re: [vpp-dev] Coverity runs

Very cool! Thanks for working on it... Dave From: vpp-dev-boun...@lists.fd.io [mailto:vpp-dev-boun...@lists.fd.io] On Behalf Of Luke, Chris Sent: Tuesday, September 19, 2017 11:50 AM To: vpp-dev Subject: [vpp-dev] Coverity runs All, Coverity have increased the limits for

Re: [vpp-dev] VPP API Message Multi-Registration Question

How about: only complain if the new registration is actually different from the old one? Thanks... Dave -Original Message- From: vpp-dev-boun...@lists.fd.io [mailto:vpp-dev-boun...@lists.fd.io] On Behalf Of Jon Loeliger Sent: Friday, September 15, 2017 3:35 PM To: vpp-dev

Re: [vpp-dev] net/mlx5: install libmlx5 & libibverbs if no OFED

I typically use "git commit --amend" followed by "git review [--draft]". HTH... Dave From: vpp-dev-boun...@lists.fd.io [mailto:vpp-dev-boun...@lists.fd.io] On Behalf Of Shachar Beiser Sent: Wednesday, September 13, 2017 11:29 AM To: vpp-dev@lists.fd.io Cc: Shahaf Shuler ;

[vpp-dev] Spurious patch verification failure (gerrit 8400)

See gerrit https://gerrit.fd.io/r/#/c/8400, https://jenkins.fd.io/job/vpp-verify-master-centos7/7070/console 12:29:12 make[1]: Leaving directory `/w/workspace/vpp-verify-master-centos7/test' 12:29:12 [vpp-verify-master-centos7] $ /bin/bash /tmp/hudson3100921859131279854.sh 12:29:12 Loaded

Re: [vpp-dev] vpp performance numbers with 10Gbps interface.

+1. If you want to rx-and-drop packets, install a drop adjacency... Sending to an unrouteable address results in 100% icmp error replies... Thanks... Dave From: vpp-dev-boun...@lists.fd.io [mailto:vpp-dev-boun...@lists.fd.io] On Behalf Of Florin Coras Sent: Tuesday, September 12, 2017 1:05 PM

Re: [vpp-dev] VPP 1704 and router plugin

Set a breakpoint in format_fib_table_name, and see if e.g. fib_table->ft_desc is NULL. Thanks... Dave From: vpp-dev-boun...@lists.fd.io [mailto:vpp-dev-boun...@lists.fd.io] On Behalf Of Bhanu Chander Gaddoju Sent: Tuesday, September 12, 2017 11:45 AM To: vpp-dev@lists.fd.io; Ni, Hongjun

Re: [vpp-dev] u32 vs uint32_t

+1, let’s stick with u32... Thanks… Dave From: vpp-dev-boun...@lists.fd.io [mailto:vpp-dev-boun...@lists.fd.io] On Behalf Of Dave Wallace Sent: Monday, September 11, 2017 12:36 PM To: Florin Coras ; Luke, Chris Cc: vpp-dev

Re: [vpp-dev] VPP contribute

That’s right, no need to send patches to a mailing list. In fact, please don’t send patches to this list. ()... Thanks… Dave From: vpp-dev-boun...@lists.fd.io [mailto:vpp-dev-boun...@lists.fd.io] On Behalf Of Shachar Beiser Sent: Monday, September 11, 2017 7:48 AM To: vpp-dev@lists.fd.io Cc:

Re: [vpp-dev] mheap performance

om] Sent: Friday, September 8, 2017 10:39 AM To: Dave Barach (dbarach) <dbar...@cisco.com> Cc: vpp-dev@lists.fd.io; Michał Dubiel <m...@semihalf.com> Subject: Re: [vpp-dev] mheap performance Hi Dave, The perf backtrace (taken from "control-only" lcore 0) is as follows: - 91.8

Re: [vpp-dev] Rearrangement of graph nodes

One could do that, but what problem are you trying to solve? The data structures involved are not super-complicated, but what you’ve described is neither a beginner project nor a worthwhile project IMO. If you want to spoof MAC addresses in the L2 path, add an L2 feature node which does that.

Re: [vpp-dev] mheap performance

Dear Jacek, Use of the clib memory allocator is mainly historical. It’s elegant in a couple of ways - including built-in leak-finding - but it has been known to backfire in terms of performance. Individual mheaps are limited to 4gb in a [typical] 32-bit vector length image. Note that the

Re: [vpp-dev] Packet loss on use of API & cmdline

at... Thanks… Dave -Original Message- From: Colin Tregenza Dancer [mailto:c...@metaswitch.com] Sent: Friday, September 1, 2017 11:01 AM To: Dave Barach (dbarach) <dbar...@cisco.com>; Ole Troan <otr...@employees.org>; Neale Ranns (nranns) <nra...@cisco.com> Cc: vpp-dev@lis

Re: [vpp-dev] Packet loss on use of API & cmdline

Tregenza Dancer [mailto:c...@metaswitch.com] Sent: Friday, September 1, 2017 8:51 AM To: Dave Barach (dbarach) <dbar...@cisco.com>; Ole Troan <otr...@employees.org> Cc: vpp-dev@lists.fd.io Subject: RE: [vpp-dev] Packet loss on use of API & cmdline Hi Dave, Thanks for looking

Re: [vpp-dev] Packet loss on use of API & cmdline

Dear Colin, Please describe the scenario which leads to vlib_node_runtime_update(). I wouldn't mind having a good long stare at the situation. I do like the parallel data structure update approach that you've described, tempered with the realization that it amounts to "collective brain

Re: [vpp-dev] About the order of  VLIB_INIT_FUNCTION called between different plugins

: vpp-dev@lists.fd.io; zhao.qingl...@zte.com.cn; wu.bi...@zte.com.cn; gu.ji...@zte.com.cn; dong.ju...@zte.com.cn; Dave Barach (dbarach) <dbar...@cisco.com> Subject: 答复: Re: [vpp-dev] About the order of VLIB_INIT_FUNCTION called between different plugins Thanks you two for the he

Re: [vpp-dev] Issue forwarding TCP packets

This is a system problem; vpp can’t solve it all by itself. When forwarding packets at L2, vpp doesn’t look past the ethernet header. It’s simply delivering packets generated by the Linux kernel on one interface to another Linux kernel interface. The kernel cheats by not generating L4

Re: [vpp-dev] About the order of  VLIB_INIT_FUNCTION called between different plugins

Explicit dependencies between plugins is probably not a good idea. There is little to guarantee that both A and B will be loaded. Please describe the use-case in more detail. Thanks… Dave From: vpp-dev-boun...@lists.fd.io [mailto:vpp-dev-boun...@lists.fd.io] On Behalf Of wang.hu...@zte.com.cn

Re: [vpp-dev] [EXT] Re: compiling error natively on an am64 box for fd.io_vpp

+1 From: Damjan Marion [mailto:dmarion.li...@gmail.com] Sent: Saturday, August 26, 2017 3:11 PM To: Eric Chen <eri...@marvell.com> Cc: Dave Barach (dbarach) <dbar...@cisco.com>; Sergio Gonzalez Monroy <sergio.gonzalez.mon...@intel.com>; vpp-dev <vpp-dev@lists.fd.io> Sub

Re: [vpp-dev] [EXT] Re: compiling error natively on an am64 box for fd.io_vpp

Just so everyone knows, the function in question is almost too simple for its own good: always_inline uword vlib_process_suspend_time_is_zero (f64 dt) { return dt < 10e-6; } What happens if you try this variation? always_inline int vlib_process_suspend_time_is_zero (f64 dt) { if (dt <

Re: [vpp-dev] 回复：compiling error when build fd.io_odp4vpp with cross_compiler for aarch64 target

of a related idiom which will probably not compile, either: void foo(void) __attribute__((weak)); void foo(void) { return 0;} Thanks… Dave From: Eric Chen [mailto:eri...@marvell.com] Sent: Tuesday, August 22, 2017 8:16 AM To: Dave Barach (dbarach) <dbar...@cisco.com>; vpp-dev <vpp-dev@lists.

Re: [vpp-dev] 回复：compiling error when build fd.io_odp4vpp with cross_compiler for aarch64 target

The multiply-defined constructor function takes its name from the VLIB_REGISTER_NODE macro argument... From: Eric Chen [mailto:eri...@marvell.com] Sent: Tuesday, August 22, 2017 8:05 AM To: Dave Barach (dbarach) <dbar...@cisco.com>; vpp-dev <vpp-dev@lists.fd.io> Cc: odp4vpp-...

Re: [vpp-dev] 回复：compiling error when build fd.io_odp4vpp with cross_compiler for aarch64 target

Did you try changing the data structure names so they don’t conflict? VLIB_REGISTER_NODE (ipsec_output4_node) + VLIB_REGISTER_NODE (ipsec_output6_node). Feel free to submit a patch. ()... Thanks… Dave From: vpp-dev-boun...@lists.fd.io [mailto:vpp-dev-boun...@lists.fd.io] On Behalf Of Eric

Re: [vpp-dev] [discuss] Question about VPP support for ARM 64

+1, pls add the typedef... Thanks… Dave -Original Message- From: Damjan Marion (damarion) Sent: Saturday, August 19, 2017 9:09 AM To: Dave Barach (dbarach) <dbar...@cisco.com> Cc: George Zhao <george.y.z...@huawei.com>; vpp-dev@lists.fd.io; disc...@lists.fd.io; csit-...

Re: [vpp-dev] Question about VPP support for ARM 64

Dear George, This specific issue isn't anywhere near as bad as you might think. As given, the code confuses 128-bit vectors with 256-bit vectors, and 64-bit vectors with 128-bit vectors. Question: does the hardware involved support 256-bit vectors? Probably not... It almost certainly does

Re: [vpp-dev] IP4 checksum calculation error

When computing an ip header checksum from scratch, please set it to zero...: remove_header(buffer, header_size, outer_headers_size); ip4_header_t* ip4_header = vlib_buffer_get_current(buffer) + sizeof(ethernet_header); ip4_header->length -= htons(header_size); ip4_header->checksum = 0;

Re: [vpp-dev] Issues building latest master

Where is that library located, or is it entirely missing? If you find it, how about making a symbolic link? Thanks… Dave From: vpp-dev-boun...@lists.fd.io [mailto:vpp-dev-boun...@lists.fd.io] On Behalf Of Burt Silverman Sent: Friday, August 11, 2017 3:26 PM To: Marco Varlese

[vpp-dev] Spurious make test failure (container POC)

Please see https://gerrit.fd.io/r/#/c/7927, and http://jenkins.ejkern.net:8080/job/vpp-test-debug-master-ubuntu1604/1056/con sole The patch in question is highly unlikely to cause this failure... 14:37:11

[vpp-dev] "vpp-make-test-docs-verify-master" job failure

Please see https://gerrit.fd.io/r/#/c/7885, https://jenkins.fd.io/job/vpp-make-test-docs-verify-master/2814/console. Any idea what this is all about? Thanks... Dave ___ vpp-dev mailing list vpp-dev@lists.fd.io

Re: [vpp-dev] dpdk tx drops

. Thanks… Dave From: SAKTHIVEL ANAND S [mailto:anand.s...@gmail.com] Sent: Friday, August 4, 2017 10:27 AM To: Dave Barach (dbarach) <dbar...@cisco.com> Cc: vpp-dev <vpp-dev@lists.fd.io> Subject: Re: [vpp-dev] dpdk tx drops Sure, Dave.. I will move to latest versions of VPP either 17

Re: [vpp-dev] dpdk tx drops

Does this problem occur w/ vpp 17.07? The software you mention is a full year old, and is so different from master/latest that anything is possible. As of this writing, the community has not announced an LTS plan. As the vpp project tech lead, I would be shocked if the community decided to

Re: [vpp-dev] 回复: vpp cpu usage utility

tor size 150 should be good for O(5mpps). Thanks… Dave P.S. The clocks/pkt figures below look like they may be from a TAG=vpp_debug image... Or not. It’s hard to tell at 20 KPPS... From: SAKTHIVEL ANAND S [mailto:anand.s...@gmail.com] Sent: Tuesday, July 25, 2017 9:20 AM To: Dave Barach (dbar

Re: [vpp-dev] [csit-dev] 17.07 Release

+1... Hey Neale, thanks for all your work to make the 17.07 release happen... Dave From: vpp-dev-boun...@lists.fd.io [mailto:vpp-dev-boun...@lists.fd.io] On Behalf Of Ed Warnicke Sent: Thursday, July 20, 2017 8:04 PM To: Neale Ranns (nranns) Cc: csit-...@lists.fd.io;

Re: [vpp-dev] TWS call scheduled for 7/18 8am PDT / 11am EDT / 5pm CEST

the fd.io<http://fd.io> calendar: https://calendar.google.com/calendar/embed?src=qpen2df76a954fieqdnc3avg0k%40group.calendar.google.com Ed On Wed, Jul 5, 2017 at 12:08 PM, Dave Barach (dbarach) <dbar...@cisco.com<mailto:dbar...@cisco.com>> wrote: Eric Ernst and Manohar Castelino wi

Re: [vpp-dev] VPP classifier node plugin not invoked.

Glad to hear it... Thanks for letting everyone know... Dave From: Dharmaray Kundargi [mailto:dharmaray.kunda...@mavenir.com] Sent: Monday, July 17, 2017 10:25 PM To: Dave Barach (dbarach) <dbar...@cisco.com>; vpp-dev@lists.fd.io Subject: RE: [vpp-dev] VPP classifier node plugin not i

Re: [vpp-dev] 回复: vpp cpu usage utility

The “show runtime” command displays the prevailing vector size, and per-node vector sizes. Those stats are reasonably equivalent to a vector engine load average. Thanks… Dave From: vpp-dev-boun...@lists.fd.io [mailto:vpp-dev-boun...@lists.fd.io] On Behalf Of SAKTHIVEL ANAND S Sent: Monday,

Re: [vpp-dev] Why VPP drop packets while the average vectors per node is still low

This is no surprise. Lcore 1 is running at the max vector size, indicating that it’s completely out of clock cycles / dropping traffic. Moreover: Node Name Clocks/Pkt TenGigabitEthernet6/0/1-output 8.67E+00 TenGigabitEthernet6/0/1-tx 1.08E+02 error-drop 4.15E+01 ip4-arp 3.43E+01

Re: [vpp-dev] Merge multiple frames?

Short answer: yes. If you imagine several frames on the pending frame queue - all of which end up enqueueing buffer indices to a common successor node - the successor frame will be merged by the time it’s processed. We’ve seen this effect when handing off frames from multiple source threads to

Re: [vpp-dev] trace

Folks, If one wants/needs to trace packets which vpp creates from whole cloth, here's how to do it. Typically, it's an input-node function. /* Top of node dispatch function */ u32 n_trace = vlib_get_trace_count (vm, node); /* per-buffer */ VLIB_BUFFER_TRACE_TRAJECTORY_INIT (b0); /*

Re: [vpp-dev] vpp installation without internet

+1. From: vpp-dev-boun...@lists.fd.io [mailto:vpp-dev-boun...@lists.fd.io] On Behalf Of Luke, Chris Sent: Wednesday, July 12, 2017 6:54 AM To: SAKTHIVEL ANAND S ; vpp-dev@lists.fd.io Subject: Re: [vpp-dev] vpp installation without internet You need you install the missing

Re: [vpp-dev] Weekly VPP Meeting

Dear Billy, Thanks for the update. Sorry for any confusion about today’s meeting. Dave From: Billy McFall [mailto:bmcf...@redhat.com] Sent: Tuesday, July 11, 2017 11:03 AM To: Dave Barach (dbarach) <dbar...@cisco.com>; Thomas F Herbert <therb...@redhat.com> Cc: vpp-dev@lists.f

Re: [vpp-dev] set mapping from node to thread

To: Dave Barach (dbarach) <dbar...@cisco.com> Cc: vpp-dev@lists.fd.io Subject: Re: [vpp-dev] set mapping from node to thread Hi Dave, Thanks for the detailed response. However, from the experiment I did, I see the same packet being processed by two threads. That is why I asked this question. M

Re: [vpp-dev] set mapping from node to thread

Dear Yuliang, From a high level: vpp creates N identical graph replicas in a multi-core configuration. When practicable, we use hardware RSS hashing to ensure that all packets belonging to a specific flow are processed [in order!] by the same thread / graph replica. In effect, embarrassing

Re: [vpp-dev] GDB File format not recognized

Please run gdb on .../install-vpp_debug-native/vpp/bin/vpp, not .../build-vpp_debug-native... Thanks… Dave From: vpp-dev-boun...@lists.fd.io [mailto:vpp-dev-boun...@lists.fd.io] On Behalf Of Anand Nandugudi Sent: Thursday, July 6, 2017 10:07 AM To: vpp-dev@lists.fd.io Subject: [vpp-dev] GDB

[vpp-dev] TWS call scheduled for 7/18 8am PDT / 11am EDT / 5pm CEST

Eric Ernst and Manohar Castelino will present their vpp integration with ClearContainer. Dial-in info will be added to https://wiki.fd.io/view/VPP/Meeting shortly... Thanks... Dave ___ vpp-dev mailing list vpp-dev@lists.fd.io

Re: [vpp-dev] how can i include dpdk.h header file in vnet?

Please do not introduce dpdk plugin dependencies into interface_api.c. We’ve moved all dpdk dependencies to the dpdk plugin, on purpose. What are you trying to do? Thanks… Dave From: vpp-dev-boun...@lists.fd.io [mailto:vpp-dev-boun...@lists.fd.io] On Behalf Of Samuel S Sent: Tuesday, July 4,

<    3   4   5   6   7   8   9   10   >