From: Per Oberg <pero@wolfram.com>
To: Laurentiu-Cristian Duca <laurentiu.duca@gmail.com>
Cc: xenomai <xenomai@xenomai.org>
Subject: Re: RTNet - vnic0 question
Date: Thu, 22 Aug 2019 09:47:19 -0500 (CDT) [thread overview]
Message-ID: <382769722.821292.1566485239775.JavaMail.zimbra@wolfram.com> (raw)
In-Reply-To: <CAPm8-H53Xb8HSU6xT33su+K_2Q5gzhCFXibKxuY+43j_E6WdoQ@mail.gmail.com>
----- Den 22 aug 2019, på kl 16:27, Laurentiu-Cristian Duca laurentiu.duca@gmail.com skrev:
> Hi,
> I have used rtifconfig, rtroute and rtping and I mentioned that worked,
> I have written that rtt-sender and rtt-responder which use posix skin
> worked only on vnic0 and not on rteth0.
> I have ported frag-ip.c (from rtnet 0.9.13 to xenomai 3)
> which uses alchemy instead of posix and it works on rteth0 link
> between two qemu-x86_64-rt_8139too virtual machines,
> but do not work on bbb (I think the drivers I use have problems on bbb).
Sorry for the blunt answer. I obviously didn't read enough of the discussion and I didn't provide a good explanation of my thoughts.
My thoughts were basically these:
What I have found myself is that the routing table for rteth0 needed external poking (using ping from other source) to be set up properly, and "Network unreachable" seems to me like a failure in lookup. Perhaps when using virtual devices there are some messages sent in the virtual environment that pokes it for you?
You write that you had it working and already tried rtping, so maybe it's already taken care of. But otherwise it would have been a good idea trying rtroute and see if there is a valid mac-address in the table.
> Thx
> On 8/22/19, Per Oberg <pero@wolfram.com> wrote:
> > ----- Den 22 aug 2019, på kl 13:21, xenomai xenomai@xenomai.org skrev:
> >> Hi,
> >> I have made a test between two qemu-x86_64-rt_8139too with different
> >> mac and IP addresses,
> >> one is master one is slave. In each of them I do:
> >> /usr/xenomai/sbin/rtnet start
> >> Then the applications rtt-sender and rtt-responder (taken from rtnet
> >> 0.9.13 release) work,
> >> but if I turn off vnic0, then rtt-sender reports error: "sendto
> >> failed: Network is unreachable",
> >> even if rteth0 is up and fully configured.
> >> Do you have any simple apps that work on rtnet?
> > You have :
> > rtping
> > rtifconfig
> > rtroute
> > Perhaps these can be of assistance?
> >> Thank you,
> >> Kind regards
> >> On 8/22/19, Laurentiu-Cristian Duca <laurentiu.duca@gmail.com> wrote:
> >> > The rt drivers for bbb are taken from:
> >> > https://github.com/G-Ray/xenomai-boneblack-rtnet-drivers
> >> > and I am not 100% sure that they are working correctly.
> >> > On 8/22/19, Laurentiu-Cristian Duca <laurentiu.duca@gmail.com> wrote:
> >> >> Hello Jan and friends
> >> >> Thank you for your fast answer. Please find below the details.
> >> >> On 8/21/19, Jan Kiszka <jan.kiszka@siemens.com> wrote:
> >> >>> On 21.08.19 16:20, Laurentiu-Cristian Duca via Xenomai wrote:
> >> >>>> Hello rtnet community,
> >> >>>> I successfully tested the examples rtt-sender and rtt-responder
> >> >>>> within a rtnet made from one beaglebone black and one
> >> >>>> qemu-x86_64-8139too,
> >> >>>> with tdma and also with nomac.
> >> >>> Are we talking about some legacy Xenomai or even RTnet version? Or
> >> >>> where
> >> >>> did
> >> >>> you
> >> >>> get the examples from? They are not part of RTnet in Xenomai 3.
> >> >> I got the examples from rtnet-0.9.13. I know they are quite old, but
> >> >> I needed a start point.
> >> >> Are these obsolete? Do you have any examples for Xenomai 3?
> >> >> rtnet-0.9.13/examples/xenomai/posix/rtt-sender.c
> >> >> rtnet-0.9.13/examples/xenomai/posix/rtt-responder.c
> >> >>>> I have noted that the presence of vnic0 as an up and fully
> >> >>>> configured interface is mandatory,
> >> >>>> even if rteth0 is up and fully configured.
> >> >>>> Otherwise the rtt-sender announces "sendto failed: Network is
> >> >>>> unreachable".
> >> >>>> Could anybody tell me why is this happening?
> >> >>>> I have also noted that if I disable vnic0 (ifconfig vnic0 down),
> >> >>>> rtping works if rteth0 is up and fully configured.
> >> >>>> Why does not work rtt-sender in this case?
> >> >>>> Could it be a problem with rtnet drivers?
> >> >>> I bet your examples are falling back to normal network here, and that
> >> >>> is
> >> >>> provided between your test nodes via the vnics. This leads to my
> >> >>> first
> >> >>> question:
> >> >>> Where did you pick the examples up, how did you build them etc.?
> >> >>> Also,
> >> >>> did
> >> >>> you
> >> >>> make sure that rtudp was loaded?
> >> >> I am 100% sure that rtudp was loaded. The modules were loaded by a
> >> >> script:
> >> >> modprobe rt_smsc && modprobe rt_davinci_mdio && modprobe rt_ticpsw
> >> >> modprobe rtipv4 && modprobe rtpacket && modprobe rtudp
> >> >> modprobe rtmac
> >> >> modprobe nomac
> >> >> #modprobe tdma
> >> >> I used robertcnelson Xenomai 3.0.9 patch for beaglebone black linux
> >> >> 4.14.108. It's on github.
> >> >> I have also copied the rtnet header files to xenomai-userspace
> >> >> include. The files compile without warnings.
> >> >> The makefile is:
> >> >> # Tested with linaro toolchain
> >> >> (gcc-linaro-7.4.1-2019.02-x86_64_arm-linux-gnueabihf)
> >> >> CC = arm-linux-gnueabihf-gcc
> >> >> XENO_INSTALL =
> >> >> /home/laur/lucru/raspberry-pi/robertcnelson/ti-linux-kernel-dev/ignore/xenomai-userspace
> >> >> RTNET_INSTALL =
> >> >> /home/laur/lucru/raspberry-pi/robertcnelson/rtnet-hiddeate2m
> >> >> #XENO_CONFIG :=
> >> >> /home/laur/lucru/raspberry-pi/robertcnelson/ti-linux-kernel-dev/ignore/xenomai-userspace/usr/xenomai/bin/xeno-config
> >> >> #/usr/xenomai/bin/xeno-config
> >> >> #CFLAGS := $(shell $(XENO_CONFIG) --posix --alchemy --cflags)
> >> >> #LDFLAGS := $(shell $(XENO_CONFIG) --posix --alchemy --ldflags)
> >> >> LIBCFLAGS = -Wall -I$(XENO_INSTALL)/usr/xenomai/include/cobalt
> >> >> LIBCFLAGS += -I$(XENO_INSTALL)/usr/xenomai/include
> >> >> CFLAGS = -Wall -I$(XENO_INSTALL)/usr/xenomai/include/cobalt
> >> >> CFLAGS += -I$(XENO_INSTALL)/usr/xenomai/include/alchemy
> >> >> -I$(XENO_INSTALL)/usr/xenomai/include
> >> >> CFLAGS += -march=armv7-a -mfpu=vfpv4 -D_GNU_SOURCE -D_REENTRANT
> >> >> -D__COBALT__ -D__COBALT_WRAP_
> >> >> CFLAGS += -I$(RTNET_INSTALL)/stack/include
> >> >> LFLAGS = -lalchemy -lcopperplate
> >> >> $(XENO_INSTALL)/usr/xenomai/lib/xenomai/bootstrap.o
> >> >> LFLAGS += -Wl,--wrap=main -Wl,--wrap=malloc -Wl,--wrap=free
> >> >> #-Wl,--dynamic-list=$(XENO_INSTALL)/usr/xenomai/lib/dynlist.ld
> >> >> LFLAGS += -L$(XENO_INSTALL)/usr/xenomai/lib -lcobalt -lpthread -lrt
> >> >> LFLAGS += -march=armv7-a -mfpu=vfpv4 -Llib
> >> >> all: rtt-sender rtt-responder
> >> >> rtt-sender: rtt-sender.o
> >> >> $(CC) -o rtt-sender rtt-sender.o $(LFLAGS)
> >> >> rtt-sender.o: rtt-sender.c
> >> >> $(CC) -c -o rtt-sender.o rtt-sender.c $(CFLAGS) $(LIBCFLAGS)
> >> >> (same for rtt-responder)
> >> >> Kind regards,
> >> >> L-C.
> > Per Öberg
Per Öberg
prev parent reply other threads:[~2019-08-22 14:47 UTC|newest]
Thread overview: 10+ messages / expand[flat|nested] mbox.gz Atom feed top
2019-08-21 14:20 RTNet - vnic0 question Laurentiu-Cristian Duca
2019-08-21 16:59 ` Jan Kiszka
2019-08-22 7:39 ` Laurentiu-Cristian Duca
2019-08-22 7:54 ` Laurentiu-Cristian Duca
2019-08-22 11:21 ` Laurentiu-Cristian Duca
2019-08-22 12:10 ` Per Oberg
2019-08-22 14:27 ` Laurentiu-Cristian Duca
2019-08-22 14:38 ` Jan Kiszka
2019-08-27 13:53 ` Laurentiu-Cristian Duca
2019-08-22 14:47 ` Per Oberg [this message]
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=382769722.821292.1566485239775.JavaMail.zimbra@wolfram.com \
--to=pero@wolfram.com \
--cc=laurentiu.duca@gmail.com \
--cc=xenomai@xenomai.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.