All images including the latest are available at: file://xsj-pvstd2t01-w/xrlabs/grahams/public/pynq-z1_images/
LATEST CHANGELOG (for a full list, check full_img_changelog.txt):
08-02-2016 - <anurag.dubey@xilinx.com>
software upgrades and Pynq-Z1 revB fixes
MAC address generation (temporary fix)
opencv added
xterm.js addition (build of Jupyter from github)
Latest version of notebook/ipython with an pre alpha release of Jupyterlab
New Sigrok decoders with source in /opt
Username/Password changed to xilinx/xilinx.
New configuration for Jupyter which doesn’t require restart on any network related change
Removed redundant source files and *.bat files and updated scripts
Added REVISION in /home/xilinx folder
Add revision text in /home/xpp folder
Modify username/password to xilinx/xilinx
Samba username/password to samba/samba
Staging - to be added in next image
Digilent u-boot with MAC address generation (removal of temporary MAC address generation)
clean up unneeded packages
Fix/Remove broken symlinks in linux headers
new devicetree with usb0 and fclk3 enabled
08-02-2016
opencv steps TO BE ADDED
sigrok steps TO BE ADDED
06-24-2016
# use the latest update_pynq.sh in /home/xpp/scripts
sudo ./scripts/update_pynq.sh
# add `start_pl_server.py` and `stop_pl_server.py` into /home/xpp/scripts
sudo chown xpp:xpp /home/xpp/scripts/start_pl_server.py
sudo chown xpp:xpp /home/xpp/scripts/stop_pl_server.py
sudo chmod 755 /home/xpp/scripts/start_pl_server.py
sudo chmod 755 /home/xpp/scripts/stop_pl_server.py
# add `3_pl_server.sh` and `4_resizefs.sh` into /home/xpp/scripts
sudo mv -rf ./scripts/3_pl_server.sh /root/
sudo mv -rf ./scripts/4_resizefs.sh /root/
sudo chown xpp:xpp /root/3_pl_server.sh
sudo chown xpp:xpp /root/4_resizefs.sh
sudo chmod 755 /root/3_pl_server.sh
sudo chmod 755 /root/4_resizefs.sh
# use the new `rc.local` file in /etc folder
sudo vi /etc/rc.local
add `/root/3_pl_server.sh`
add `/root/4_resizefs.sh`
# install sigrok-cli with superuser privilege
sudo apt-get install git-core gcc make autoconf automake libtool pkg-config libglib2.0-dev
download the sigrok-cli package
tar xzf <sigrok-cli tar file>
pushd <sigrok-cli folder>
./configure
make
sudo make install
popd
# install PulseView with superuser privilege
sudo apt-get install git-core g++ make cmake libtool pkg-config libglib2.0-dev libqt4-dev libboost-test-dev libboost-thread-dev libboost-filesystem-dev libboost-system-dev
download the sigrok-cli package
tar xzf <pulseview tar file>
pushd <pulseview folder>
cmake .
make
sudo make install
popd
# enable the X11 forwarding on the target
sudo vi /etc/ssh/ssh_config
set ForwardingX11 to yes
# make sure there is a file .Xauthority in /home/xpp/ with mode 0600
sudo chmod 0600 /home/xpp/.Xauthority
04-15-2016
add back in chroot* scripts
pip install sphinx_rtd_theme
pip install nbsphinx
rm /etc/network/interfaces.d/wlx000f6002a692
sudo chown -R xpp:xpp /home/xpp/jupyter_notebooks
cd /home/xpp
ln -s /usr/local/lib/python3.4/dist-packages/pynq
chmod -R a+rw /usr/local/lib/python3.4/dist-packages/pynq
chmod a+x /usr/local/lib/python3.4/dist-packages/pynq
# modified smb.conf to allow symlink-following
# C:\Users\grahams\Documents\PyXi\smb.conf
# Add cpufreq-utils to disable clock scaling
# http://askubuntu.com/questions/523640/how-i-can-disable-cpu-frequency-scaling-and-set-the-system-to-performance
# get trusty armhf debian packages below
dpkg -i libcpufreq0_008-1_armhf.deb
dpkg -i cpufrequtils_008-1_armhf.deb
# add line GOVERNOR="performance"
vi /etc/default/cpufrequtils
sudo update-rc.d ondemand disable
sudo shutdown -r now
cpufreq-info
cpufreq stats: 325 MHz:0.00%, 650 MHz:100.00%
# modify jupyter start call to put jupyter terminal at /home/xpp
vi /root/2_jupyter_server.sh
su -c "cd ~xpp ; jupyter notebook 2> /root/jupyter.log &" -s /bin/bash root
# seaborn for nice matplotlib
pip install seaborn
03-25-2016
# Microblaze cross compiler
# https://briolidz.wordpress.com/2012/02/07/building-embedded-arm-systems-with-crosstool-ng/
# http://crosstool-ng.org/download/crosstool-ng/
# use crosstool-ng version 1.20.0
# once downloaded onto Zybo
./configure --with-libtool=/usr/share/libtool
make
make install
./ct-ng menuconfig
Enable Experimental Features
Target: microblaze
vendor: xilinx
endinanness: little-endian
./ct-ng build
# will take 1.5 hours on Zybo
export PATH=/home/xpp/x-tools/microblazeel-xilinx-elf/bin:${PATH}
cd /home/xpp/x-tools/microblazeel-xilinx-elf/bin
# make symbolic link for each microblazeel-xilinx-elf-* to mb-*
# this link will allow for SDK makefiles to run with alteration
03-15-2016
# cffi used for SDSoC interaction
pip install cffi
#usb wifi and webcam support - limited support for known USB chipsets (see notebooks usb_*
sudo apt-get install libusb-1.0-0-dev usbutils module-init-tools linux-firmware wpasupplicant
<from Linux kernel build> copy kernel loadable modules to /lib/modules/3.17.0-xilinx
#samba support
# reference: https://www.howtoforge.com/tutorial/samba-server-ubuntu/
apt-get install samba
<modify samba settings for xpp user> see /etc/samba/smb.conf for changes
#To get bash in the Jupyter terminal and consequently all the nice feature such as history and TAB autocompletion
<on target terminal>
sudo rm /bin/sh
sudo ln -s /bin/bash /bin/sh
# reveal.js
# https://github.com/hakimel/reveal.js/
<copy reveal.js into /home/xpp/jupyter_notebooks>
02-26-2016
<on target terminal>
# install Pillow to enable saving an AV frame as JPEG.
# First of all, install this packages
sudo apt-get install libtiff5-dev libjpeg8-dev zlib1g-dev libfreetype6-dev liblcms2-dev libwebp-dev tcl8.6-dev tk8.5-dev python-tk
#Then install Pillow using pip
sudo -H pip install pillow
02-11-2016
<on target terminal>
# autologin XPP user
sudo sed -i.bak 's/ExecStart=-\/sbin\/agetty/ExecStart=-\/sbin\/agetty --autologin xpp/' /lib/systemd/system/serial-getty@.service
# fix date/time due to no RTC
sudo apt-get install ntp
sudo sed -i.bak 's/\/var\/lib\/ntp\/ntp.conf.dhcp -nt \/etc\/ntp.conf/ -e \/var\/lib\/ntp\/ntp.conf.dhcp /' /etc/init.d/ntp
# speed up boot when no network DHCP server
sudo sed -i.bak 's/timeout 300;/timeout 3;\nretry 3;\nreboot 3;\nselect-timeout 0;\ninitial-interval 2;\nlink-timeout 3;\n/' /etc/dhcp/dhclient.conf
# reboot
shutdown -r now
02-03-2016
<from repository>
copy <Pynq Repo Head>/ubuntu/* --> <Zybo Linux Root>/home/xpp
<on target terminal>
sudo mv 0_network.sh 1_jupyter_config.sh 2_jupyter_server.sh /root/
sudo mv rc.local /etc/rc.local
sudo chmod +x /root/0_network.sh /root/1_jupyter_config.sh /root/2_jupyter_server.sh /etc/rc.local
sudo cp ~xpp/.bash* /root
sudo cp -r ~xpp/.jupyter /root
source <Path to Vivado>/settings64.sh
cd <Path to the folder containing the .tcl file>
vivado -mode tcl
source pmod.tcl
write_sysdef -hwdef "./project_1.runs/impl_1/top.hwdef" -bitfile "./project_1.runs/impl_1/top.bit" \
-meminfo "./project_1.runs/impl_1/top_bd.bmm" -file "./project_1.runs/impl_1/top.hdf"
exit
Note: This is an example using the "pmod.tcl" file. The bitstream can also be generated using Vivado GUI.
source <Path to Vivado>/settings64.sh
cd <Path to the workspace>
xsdk -batch
sdk set_workspace <Path to the workspace>
sdk create_hw_project -name hw_0 -hwspec ../mipy_zybo_video.runs/impl_1/top.hdf
sdk create_bsp_project -name bsp_0 -hwproject hw_0 -proc ps7_cortexa9_0 -os standalone
sdk create_project -type app -name zybo_fsbl -hwproject hw_0 -proc ps7_cortexa9_0 -os standalone -lang C -app {Zynq FSBL} -bsp bsp_0
sdk build -type bsp bsp_0
sdk build -type app zybo_fsbl
sdk clean -type bsp bsp_0
sdk clean -type all
sdk build -type all
exit
Note: This step should be done after the bitstream has been generated. If Vivado GUI is used to generate the bitstream, first choose "export hardware" and "launch SDK". Then create an application project with proper settings, using the "Zynq FSBL" as the template.
References:
source <Path to Vivado>/settings64.sh
git clone https://github.com/xilinx/u-boot-xlnx
cd u-boot-xlnx
sed -i.bak '/ramdisk_image/d' include/configs/zynq-common.h
make clean zynq_zybo_config
make
mv u-boot u-boot.elf
source <Path to Vivado>/settings64.sh
git clone https://github.com/Xilinx/linux-xlnx.git --branch xlnx_3.17 --single-branch
cd linux-xlnx
export PATH=${PATH}:`pwd`/../u-boot-xlnx/tools/
make ARCH=arm xilinx_zynq_defconfig
make ARCH=arm menuconfig
-- enable drivers
make ARCH=arm UIMAGE_LOADADDR=0x8000 CROSS_COMPILE=arm-xilinx-linux-gnueabi- uImage modules -j32
rm -rf ../linux_modules
make ARCH=arm INSTALL_MOD_PATH=../linux_modules CROSS_COMPILE=arm-xilinx-linux-gnueabi- modules_install
make ARCH=arm headers_install
source <Path to Vivado>/settings64.sh
mkdir sdcard
echo "image : {" > boot.bif
echo "[bootloader]fsbl/fsbl.elf" >> boot.bif
echo "u-boot-xlnx/u-boot.elf" >> boot.bif
echo "}" >> boot.bif
bootgen -image boot.bif -w on -o i sdcard/boot.bin
cp linux-xlnx/arch/arm/boot/uImage sdcard/
cp devicetree/devicetree.dtb sdcard/
TBD - Use standard Zybo Device Tree TBD
Follow step 8 at link below to partition SDCard. Many other online references exist as well on how to partition an SDCard.
References:
References:
- http://hunterhu.com/1/post/2015/02/embedded-rootfs-from-ubuntu-core-rootfs.html
- http://gnu-linux.org/building-ubuntu-rootfs-for-arm.html
# These steps require root access from an Ubuntu machine
# For corporate users - this typcailly means having a Virtual Machine
# All steps were done using VMWare and Ubuntu 32b 14.04.3 - ubuntu-14.04.3-desktop-i386.iso
# SDCard prepartitioned into BOOT (FAT32) and rootfs (EXT4) and accessible from VM
apt-get install qemu-user-static
export TGZ_PATH=<PATH_TO>/ubuntu-core-15.10-core-armhf.tar.gz
export CHROOT_IN_PATH=<PATH_TO>/chroot_in.sh
export CHROOT_OUT_PATH=/chroot_out.sh
tar -zxpf $TGZ_PATH
cp /usr/bin/qemu-arm-static usr/bin
cp $CHROOT_IN_PATH .
cp $CHROOT_OUT_PATH .
./chroot_in.sh
chmod 755 /
apt-get install apt-utils
echo 'debconf debconf/frontend select Noninteractive' | debconf-set-selections
apt-get update
apt-get upgrade
apt-get install sudo net-tools build-essential ssh ethtool \
iputils-ping rsyslog language-pack-en-base bash-completion \
nano vim policykit-1 lsof resolvconf python3-dev \
python3-setuptools gfortran ntp vim lsof git sysstat imagemagick
adduser xpp; adduser xpp sudo
apt-get upgrade
echo "zybo" > /etc/hostname
echo "127.0.0.1 localhost" > /etc/hosts
echo "127.0.1.1 zybo" >> /etc/hosts
./chroot_out.sh
exit
cd ..
# Disconnect the SDCard from VM
# a 1024MB swap partition will be created, but only if a swap partition has not been created yet
if ! grep -Fxq /swap /etc/fstab; then
dd if=/dev/zero of=/swap bs=1M count=1024
mkswap /swap
echo '/swap none swap sw 0 0' >> /etc/fstab
else
echo swap partition already configured
fi
Sometimes you may find that ubuntu official repos does not include certain pakcages for the armhf architecture - i.e. the usual apt-get will not work. However, it may be that these packages are available as a yet-unofficial build on ubuntu's development website: https://launchpad.net/ubuntu/wily/armhf. If so, simply wget the .deb package and then install it using dpkg.
For instance, the i2c-tools and libi2c-dev are not available using normal apt-get, but there is still an armhf build for ubuntu 15.10:
- https://launchpad.net/ubuntu/wily/armhf/i2c-tools/3.1.1-1
- https://launchpad.net/ubuntu/wily/armhf/libi2c-dev/3.1.1-1
To install them, simply execute this commands
wget http://launchpadlibrarian.net/173848656/i2c-tools_3.1.1-1_armhf.deb
wget http://launchpadlibrarian.net/173859774/libi2c-dev_3.1.1-1_all.deb
dpkg -i i2c-tools*
dpkg -i libi2c-dev*
When you install packages in this way, you need to manually recursively resolve every dependency listed on the package website (the depends on list at the bottom of the package page, if present), possibly installing them using the same technique if they are not available with normal apt-get.
-
configure_jupyter.sh : set up the ethernet connection and configure Jupyter to be launched with the DHCP given IP address. A message at the end will provide further instructions
-
xilinx_proxy.sh : set or unset the Xilinx internal network proxy
-
py_package.sh: run this script to know where to put your packages to be used by Jupyter. The script will create the specified path if it does not exists. To copy the package, you can use WinSCP creating a New Site with protocol SFTP, Host name the ip of the board (that you can get typing
hostname -Ifrom terminal or executing the configure_jupyter.sh script), Port number 22 and User name xpp. The password is also xpp. If you leave the password entry empty, it will be asked at every log in. You can obviously copy the package using your method of choice instead. In this repository, the pynq package is located under /python/pynq
Remember to run
chmod +x <selected_scritp.sh>
to make it executable before you launch it.
An easy alternative to script #3 is to install the pynq package using pip directly (and avoid all the hassle of manually copying files). To do that, simply type from a terminal while being in the home directory (/home/xpp).
sudo -H pip install -e 'git+https://github.com/Xilinx/Pynq@master#egg=pynq&subdirectory=python'
https://www.digitalocean.com/community/tutorials/how-to-install-and-configure-vnc-on-ubuntu-14-04
# in /etc/apt/sources.list uncomment 'universe' repository entries
sudo apt-get update
sudo apt-get install xfce4 xfce4-goodies tightvncserver
vncserver