I have successfully compiled opencv 3.1 on a raspberry pi. Developing with the library works perfectly fine. Now I wanted to set up another, identical raspberry with opencv and to save the compilation time, my idea was to copy the binaries to the second raspberry.
So I copied the opencv directories including the build folder and tried to run sudo make install. Instead of using the already compiled files, compilation using cmake starts over again.
How can I convince the second raspberry's build environent that there is no need to recompile everything? On my original raspberry, I can run sudo make install on the exactly same files without recompilation. Installed dev-libraries are the same on both systems. Is this a cmake, make or a opencv specific problem?
I also tried to copy all .so and .h files from /usr/... directories, but then I run into further problems when other cmake projects try to locate the opencv package.
Build directory is not intended to be copied into other place or on another machine.
For deliver program to other machine you should use installed files, or, more generally, a package.
CMake is shipped with CPack, which can build the program from sources and create package contained all its deliverables.
You may create .deb package on the first Raspberry PI machine:
cpack -G DEB <source-dir>
and install it on the second machine using dpkg.
There are also "archive" packages like .tgz or .zip. Full list of CPack generators is described in wiki.
Related
I am following this tutorial. I already have .fastq files. I want to install ea-utils.
My setup is Ubuntu 18.04 bionic, via Oracle VM Virtual Box.
In terminal, I entered the command:
>>>sudo apt install ea-utils
E: Unable to locate package ea-utils
First, I installed latest Ubuntu updates via. Software Updater.
Then,
>>>sudo apt-get update
E: Some index files failed to download. They have been ignored, or old ones used instead.
Still throwing an error:
>>>sudo apt-get install ea-utils
Second command said: E: Unable to locate package ea-utils.
You cannot install it using Git-Bash. Git-Bash is not a Linux environment (apt-get is a Linux utility that can be used in a Linux environment). Git-Bash is a subset of the MSYS (or MSYS2, not sure) collection of open source tools compiled for Windows
What you can try is
build your own version of ea-utils for Windows. build guide - I will elaborate if required
check if there are any precompiled binaries for it
Expanding on building/compiling your own binaries of programs
Normally a program is written in a programming language (e.g C/C++, Java) that humans can read. These are plain text files.
That is compiled into something computers can read
This compiled file is executable on the platform which it is compiled for - ends in .exe for Windows
This executable file is distributed as a 'precompiled binary' that is copied into (usually) C:\Program Files by the installation procedure
But things change in the world of open source software
You are given the original files of code written in a programming language
You use a compiler combined with other libraries to compile it into an executable file
MinGW is a collection of tools, including the C/C++ compiler for Windows
GSL is a library that provides some other code that ea-utils depends on for the compilation of the binaries
General instructions for building
(Sorry I cannot test these. I do not use Windows any more)
Install MinGW - accepting the defaults should work fine
Install GSL - try the link that says Setup (again, accept defaults)
Unzip the file you downloaded earlier from ea-utils' GitHub
Open command prompt
cd into the unzipped folder
(based on instructions on their wiki) make
make test
Since your updated question is based on using Ubuntu 18.04 in a VM and you there is still an error, I suggest trying
sudo apt-get update
sudo apt-get install ea-utils
Commonly, software in the Linux word is distributed as "packages" - e.g ea-utils. The first command contacts Ubuntu's repositories (they serve the packages) and generates a list of all the available packages.
That should fix the error of ea-utils not being found.
Following the constant errors being thrown,
Download the .deb file 64-bit version or 32-bit version according to the virtual machine you are running. Open it inside the virtual machine, and follow the onscreen instructions.
I've built from source a version of GCC for Red Hat. Given that was a fairly expensive exercise, I'd like to "back up" the built version so I can avoid having to do it again on other machines or even if I scrap the VM it's currently built on.
What are my best options for doing that please?
This is what package managers are for. However, you have to build (usually) or deploy the package under the control of the package manager, because it is what keeps track of where the installed files are. You built it without using a package manager, and no longer have the complete build directory? Sorry, it's going to be quite tricky to find all the installed files.
However, if you still have the build directory and you haven't run make clean, you can just tar that up and copy that tar file somewhere, and run make install from an untarred copy of that tar file on each machine. Alternatively, you could use something like GNU Stow or XStow as a poor-man's package manager to deploy and undeploy it on various machines, by installing it to /usr/local/stow/gcc, tarring up the /usr/local/stow/gcc directory, untarring it on another machine, and then using GNU Stow to install it.
When trying to install Android Studio on my Linux Laptop, I get "Unable to Run mksdcard tool" From what I can tell from searching, this is usually caused by lacking the 32 bit compatibility libraries on 64 bit Linux, however I am running it on an ARMv7 processor, using the crouton project to use Linux on my Chromebook. I have tried install the recomended packages ending in i386, but the command line returned:
Reading Package Lists... Done
Building Dependendency Tree
Reading State information... Done
E: unable to locate package [Name of package here]
E: Couldn't find any package by Regex '[Name of package]'
Does anyone know what is causing this and how I can fix it?
I've discovered a workaround.
After a little searching, I've found that we can create executable binary of the tool for the ARMv7 platform ourselves! Whupee!
Head over to GitHub and pick up the source code, mksdcard.c. Download this to wherever you'd like, but make sure you download it as mksdcard.c and not as mksdcard.c.txt, which your browser might try to do. You can always rename the file later in case you accidentally save the filename incorrectly.
Over in your chroot environment, head to the directory where you downloaded the file.
Make sure you have the gcc compilation tools installed. Try running gcc -v in an attempt to see what version of GCC you have installed. If this doesn't work, you'll need to install GCC via sudo apt-get install gcc.
Run gcc -o mkdscard mksdcard.c. This uses GCC to compile the source code into something that can be executed. After compilation has completed, you can use ./mkscard to have Linux execute the binary file, which verifies that it works.
Navigate to your Android SDK Tools directory. This is usually ~/Downloads/Android/Sdk/tools. By running ls, you'll list the files and find the version of mksdcard that your Linux distribution doesn't understand how to run. (Running ./mksdcard on this file will confirm this.)
Backup the broken binary somewhere, then delete the copy in the tools folder. (I created a backups/ directory within the Android SDK Tools folder to move it to.)
Within the directory, use rm -r mksdcard to delete the old mksdard binary.
Finally, copy your compatible binary over to take it's place, e.g. cp ~/Downloads/mksdcard . (Copies the mksdcard binary we've created to the current directory ., the Android SDK Tools folder.)
Head back over to your Android Studio installer. In the dialogue complaining about mksdcard failing, hit Retry and the installation should continue. After it's finished, be sure to apply any updates that are recommended by the environment. Enjoy!
For newer versions eg. 3.1 C4 of Android Studio running with Ubuntu on ARM32 you will also need to place mksdcard in ~/Downloads/Android/Sdk/emulator (referencing like path from Alext T.).
I want to run QT-Creator on my Odroid XU3 (ARM Architecture with Lubuntu 14.4) and build an application with it. I've read a lot of on the internet to figure out how to do this but haven't been successful yet. I tried to follow these instructions https://wiki.qt.io/ODROID-XU3 but didn't understand few things.
It states that I need to install some dependencies. I followed the given hyperlink on this site and installed the dependencies using following commands:
sudo apt-get build-dep qt5-default
sudo apt-get install "^libxcb.*" libx11-xcb-dev libglu1-mesa-dev
libxrender-dev libxi-dev
When I searched for my qt5 folder, I could only find qt4 in user/share. What did I do wrong?
It states that I have to setup a toolchain and a sysroot. Do I have to do this only if I want to cross-compile, so compile on a different machine for my Odroid XU3, or also if I just want to compile on Odroid itself?
Let's say I want to setup the toolchain. According to this site I should copy the raspberry pi toolchain (or a different one) into the folder ~/Code/Odroid/toolchain. Problem is I don't see such a folder in my home directory on my Odroid. Should I mkdir one?
Let's also say I want to setup the sysroot, I don't get the part "One is to mount the ODROID-XU3's SD Card or eMMC on your development machine and either make a copy or a link to ~/Code/Odroid/sysroot". Am I right in assuming that I should copy the content of my SD-card to this folder on my developing machine?
Where should I build my QT now - on my development machine or Odroid? (sorry for that stupid question)
What If I want to make it simple and build everything on Adroid XU3 assuming that I have plenty of time and resources. I tried to first install qt and qt creator afterwards but as I started qtcreator, there was no compiler preconfigured and also the QuickView was missing. The terminal was also mentioning that the toolchain is missing. I struggled a lot in order to add these missing components but failed.
Thanks in advance!
to your second point: you can compile it on the Odroid or on your PC. But you should take a look at this: https://wiki.qt.io/Category:QtonPi it should be pretty the same on the Odroid.
I've got a Raspberry PI with Emdebian installed on it, and want to cross-compile projects.
There is plenty of documentation on how to obtain a toolchain and build a simple project with it. I myself managed to build a toolchain with crosstool-ng and wrote a hello world program which works fine.
What I don't get is how to handle cross-compiling more complex projects like Qt, which have dependencies on other libraries. Let's use libdbus as an example, as that is one of Qt's dependencies.
The installed Emdebian already contains libdbus.so, so naturally I'd prefer to use that, instead of cross-compiling my own libdbus.so, as compiling all of Qt's dependencies would take a lot of time.
For cross-compiling, there are two important directories, as far as I understand:
The "staging" directory, where all the installed libraries and applications live. This initially is a copy of the toolchain's sysroot directory, and gets populated with more libraries as they are cross-compiled.
The "rootfs" directory, which is equivalent to what is on the device - essentially a copy of the staging directory without unneeded stuff like documentation and header files. As far as I understand it, the best approach is to copy required files from the staging directory into the rootfs.
Getting the rootfs directory is easy, as that can be a NFS mount from the device. But how do I get a staging directory for the existing Emdebian installation on the PI? The staging directory needs to include things like dbus headers, which are not installed on the rootfs.
Some people simply install the dbus headers on the device, with apt-get install libdbus-dev, and then use the rootfs as the staging directory. With this setup, there is no distinction between rootfs and staging anymore, with the disadvantage that the rootfs is polluted with headers, documentation and so on. The advantage of course is that it is easy.
What is the best way to get the dbus headers into my staging directory on my host machine? What is the usual approach people use in this situation?
As a side question, why does the approach of obtaining a toolchain, compiling a program and then copying that on a target work at all? The toolchain ships its own versions of libc, libstdc++ etc, are they not incompatible with the versions that are installed on the target? Especially when creating using a custom toolchain compiled with crosstool-ng?
(Note that I am not asking how to compile Qt, I can figure that out myself. My question is more general, about the approach to take when combining a custom toolchain with an existing installation/rootfs)
In my experience, you don't need to compile your dbus. You can do it as
Create Debian cross rootfs by debootstrap by https://wiki.debian.org/EmDebian/CrossDebootstrap
Create your cros-compile toolchain by crosstool-ng, and make sure the kernel version and eglibc version are the same as rootfs created by 1st step
Build QT by
CPPFLAGS=-I<rootfs>/usr/include \
LDFLAGS=-L<rootfs>/lib -L<rootfs>/usr/lib -Wl,-rpath-link,<rootfs>/lib,<rootfs>/usr/lib \
./configure <your options>
make
Install QT into the stage directory by
make install DESTDIR=<stage directory>
Copy QT dependent libraries from rootfs the your stage directory
So you can see that, the stage directory is kept minimum without pollution.