Ubuntu 16, DPDK 16
I have a couple of DPDK-based applications sending packets between two machines and want to capture them with dpdk-pdump. I'm running it with sudo ./build/app/pdump/dpdk-pdump -c 0x00f -- --pdump 'port=0,queue=*,rx-dev=/tmp/rx.pcap', and I see this output and a crash:
EAL: no driver found for eth_pcap_rx_0
EAL: Driver, cannot attach the device
EAL: Error - exiting with code: 1
Cause: vdev creation failed:create_mp_ring_vdev:736
My DPDK was compiled like this: hostname:~/dpdk-stable-16.07.1/x86_64-native-linuxapp-gcc# CONFIG_RTE_LIBRTE_PMD_PCAP=y make (also did export CONFIG_RTE_LIBRTE_PMD_PCAP=y before) so that it builds with libpcap support, as the documentation said to do. And my system has libpcap installed. The make script output shows == Build lib/librte_pdump. Looking around online, I've only found one person with a similar issue, and the suggested fix was what I tried. Is there something else I have to do?
I tried installing the DPDK Ubuntu package's version of the tool and using the dpdk-pdump command, same result.
Unfortunately, setting CONFIG_RTE_LIBRTE_PMD_PCAP=y in command line does not work, so the compiled dpdk-pdump have no pcap support.
In config/common_base set CONFIG_RTE_LIBRTE_PMD_PCAP=y, remove your build directory and recompile the DPDK.
Make sure the .config file you your build directory has CONFIG_RTE_LIBRTE_PMD_PCAP=y
Also if you get following error:
Error: missing pcap.h
rte_eth_pcap.c:19:18: fatal error: pcap.h: No such file or directory
while compiling with option CONFIG_RTE_LIBRTE_PMD_PCAP=y
Install libpcap-dev* package. Eg:
yum install libpcap-devel
Related
I did compile Ignite Application successfully.
but The Binary didn't work.
/tmp/tmp.Nw0IPD6ru3/cmake-build-debug-local-container/planet_engine: error while loading shared libraries: libjvm.so: cannot open shared object file: No such file or directory
how can I make to it work?
Also, I compiled C++ Examples successfully. such as ignite-compute-example.
and, I execute that but I got an error message.
An error occurred: JVM library is not found (did you set JAVA_HOME environment variable?)
and I using a nightly release version 2.8.0.20190213 because I couldn't build to version 2.7 in my environment.
I posted environment values down.
IGNITE_HOME=
TERM=xterm-256color
SHELL=/bin/bash
LIBRARY_PATH=/root/jre1.8.0_201/lib/amd64/server:/root/jre1.8.0_201/lib/amd64/
LC_NUMERIC=ko_KR.UTF-8
SSH_TTY=/dev/pts/0
JRE_HOME=/root/jre1.8.0_201
USER=root
LS_COLORS=rs=0:d...
LD_LIBRARY_PATH=/root/jre1.8.0_201/lib/amd64/server:/root/jre1.8.0_201/lib/amd64/
CLASS_PATH=/root/jdk-11.0.2/lib:
LC_TELEPHONE=ko_KR.UTF-8
MAIL=/var/mail/root
PATH=/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin:/root/jdk-11.0.2/bin
LC_IDENTIFICATION=ko_KR.UTF-8
JAVA_HOME=/root/jdk-11.0.2
LANG=en_US.UTF-8
LC_MEASUREMENT=ko_KR.UTF-8
JDK_HOME=/root/jdk-11.0.2/lib
SHLVL=1
HOME=/root
LOGNAME=root
LESSOPEN=| /usr/bin/lesspipe %s
LESSCLOSE=/usr/bin/lesspipe %s %s
LC_TIME=ko_KR.UTF-8
LC_NAME=ko_KR.UTF-8
_=/usr/bin/env
Thank you for reading. :)
I got it.
I am working on a docker container environment.
and therefore I am using remote build and debug with ssh and gdb.
finally, I found out why it couldn't find libjvm.so and why couldn't read environment values such as JAVA_HOME.
because it is working in gdb for now.
I confirmed that it is working when without gdb.
I will find a solution.
and, if I have been found, I will update the answer.
[Solved]
I share how I make solved that.
I was using an Oracle JDK-11 through source install.
but Ignite C++ client need something different with latest released jdk versions.
Ignite need a directory structure like this
JAVA_HOME/ (as JDK install directory)
- jre/
- lib/
- lib/
...
I solved by apt install openjdk-8-jdk.
openjdk-8-jdk have structure for what Ignite need.
i added JAVA_HOME, IGNITE_HOME, at /etc/environment.
It works finally.
but I got another problem. HAHA
I am so sad.
This also GDB problem..
I have Debian Linux(64-bit) in VirtualBox.
Ram: 8Gb (for Debian 4Gb)
CPU: i5-3470
If any other details are needed please tell me.
I have followed instructions written here, and here you can see the console log.
For some reason I am unable to to build phantomjs and I can't understand why its not working... Have searched a lot, but couldn't find any thread which would be about this error.
--- EDIT ---
When I executed this command:
./configure --with-openssl-includes=/usr/include/openssl-1.0/ --with-openssl-libraries=/usr/lib/openssl-1.0/
Output was this:
--with-openssl-includes=/usr/include/openssl-1.0/: invalid command-line switch
--with-openssl-libraries=/usr/lib/openssl-1.0/: invalid command-line switch
The output of the console has changed but I still get errors: https://pastebin.com/wbgi8syg
Looks like you are on stretch (or later) and that you are already aware of the libssl vs libssl1.0 case.
Therefore, if you don't have libssl-dev installed, this line will get you further:
python build.py --qt-config "-I /usr/include/openssl-1.0/ -L /usr/lib/openssl-1.0/"
If you already have it installed and my suggestion doesn't work, you can uninstall libssl-dev for the time to build phantomjs which will likely avoid having to play further with configuration related variables.
I ln ing the curl from /usr/bin/curl to /usr/local/bin/curl.
When I run from command line, the "no version information available" is gone, but when I compile it in home environment, it still complains about it:
cmake: ~/local/lib/libcurl.so.4: no version information available (required by cmake) /bin/bash: line 14: 63003 Segmentation fault.
When I look at the local version:
./curl --version
curl 7.27.0-DEV (x86_64-unknown-linux-gnu) libcurl/7.27.0-DEV OpenSSL/1.0.1f zlib/1.2.8 libssh2/1.4.0_DEV
Protocols: dict file ftp ftps gopher http https imap imaps pop3 pop3s rtsp scp sftp smtp smtps telnet tftp
Features: IPv6 Largefile NTLM NTLM_WB SSL libz TLS-SRP
It was installed via apt-get libcurl4-openssl-dev, couldn't remove it.
How do I install it in local/bin and local/lib?
Thanks
You have a library compiled without version symbols.
The library has version symbols for very important purpose: asserting the library is modern enough to fit the requirements of the program. If cmake complains about the library, it means that:
You've compiled it without version symbols (probable)
You compiled too old version
I think the problem is with the name "libcurl.so.4". I had a similar problem on Linux Mint 18 (base - Ubuntu 16.04), 64 bit, installing a new .deb file for software previously only available for Windoze. The install was faultless but when I tried to run the programme, it failed with the message "error while loading shared libraries: libcurl.so.4: cannot open shared object file: No such file or directory".
The supplier advised me to install "curl" but it was already there. I fully uninstalled it then reinstalled without any change. I then did a scan for "libcurl.so.4" on my hard disk without result. The nearest I could find for it were in "/usr/lib/x86_64-linux-gnu/". The files were "libcurl-gnutls.so.3", "libcurl-gnutls.so.4" and "libcurl-gnutls.so.4.4.0". The first two being symlinks to the third. I then added a further symlink to "libcurl-gnutls.so.4.4.0" called "libcurl.so.4" and this allowed me to run the new version without problems.
I'm trying to make a local install of pyMPI on a server running CentOS 5.9 (i.e. I don't have sudo privileges). The ./configure step completes successfully, however there a lot of "no"s. When i try running make i get a lot of "deprecated" error messages and the process exits with error code 2. Can anyone help me with this please?
The results of running
./configure --prefix=/inside/home/aarjunrao/apps/py_modules/pyMPI
are as follows
./configure output
I then run make, which gives the output
make output
Thanks in advance,
Arjun
From your make output, line 305:
pyMPI_util.c:22:31: error: numpy/arrayobject.h: No such file or directory
Is NumPy installed on this system? If not, that's probably your problem...
You can use this option in the configure process
./configure --prefix=/inside/home/aarjunrao/apps/py_modules/pyMPI --with-includes='-I/[path_to_numpy]/core/include'
In my case, it is
--with-includes='-I/usr/local/lib/python2.7/site-packages/numpy/core/include'
Doing that, you don't need to modify the system PATH.
I have trouble on updating tizen-2.1 on my test device
I follow manual step by step, but some errors when I use lthor tool.
Can you give me some insight?
my environment
current tizen version on test device => 2.0
ubuntu 13.04 (on vmware 9)
modemmanager off
Error Message:
line 345: failed to read signature bytes
line 629: handshake failed
line 958: failed to open port (null)
Unfortunatelly, sometimes it's needed to uninstall modemmanger at all:
sudo apt-get purge modemmanager
Also VMWare may be at the bottom of your problem.
please follow these steps
I updated my device by following these steps.
https://wiki.tizen.org/wiki/Flash_Tizen_2.1_Image_to_Reference_Device