The article Porting Qt for Embedded Linux to Another Operating System lists five things you have to do to port Qt for Embedded Linux to another OS. From the article:
There are several issues to be aware of if you plan to do your own port to another operating system. In particular you must resolve Qt for Embedded Linux's shared memory and semaphores (used to share window regions), and you must provide something similar to Unix-domain sockets for inter-application communication. You must also provide a screen driver, and if you want to implement sound you must provide your own sound server. Finally you must modify the event dispatcher used by Qt for Embedded Linux.
Is it really this easy to port Qt to another OS, or have i missed some information?
Another important component to port would be QAtomic, to ensure that you can have atomic operations and implicit sharing working well. See also
http://labs.trolltech.com/blogs/2007/08/28/say-hello-to-qatomicint-and-qatomicpointer/
Since Qt has been ported a large number of times it seems logical that it would be inherently simple. However the issue really is on the platform you are porting to and how many features it currently supports.
Assuming you find all those things easy, then the port is easy.
After investigating this in more detail I have come to the conclusion that the article "Porting Qt for Embedded Linux to Another Operating System" assumes that you are porting Qt to a very "linux-like" OS.
I have attempted this and currently making progress.
Some difficulties:
IDE - I have to manually add all Qt files and fight the compiler with #ifdefs until it builds with all dependencies in place.
Linux(ness) - I've had to disable all Linux/Windows things that are not supported in my target OS: threads, sockets, processes. Even the timers are slightly different.
Tips:
Start small : I compiled QtCore as a standard lib within my IDE, next up is QtGui which is a behemoth compared to QtCore.
I plan to run only a single QThread, so I have to artificially made a Thread object to avoid null pointers. You cannot compile out Thread information as it is key to all QObjects.
So far I have an qeventloop running within a qcoreapplication.
I wrote some inline assembly but had serious difficulties with my IDE and compilation. I left it in C++ and let the assembler handle it for me. Because I am single-threaded, I am not too concerned with shared data/ exclusive access as required by the atomic operations.
Related
I have been working with QML for a couple of weeks now. But now a new requirement has risen where I have to check whether I can run QT5 in an embedded device for my project. I need this for my GUI related operations. The device has an ARM processor (Intel ARM920T-PXA27x processor to be exact) with 64 MB RAM. 32 MB will be taken up for the OS and the rest will be available for application and QT.
I have to check whether this is possible. I have googled for some references and unable to find a suitable answer to my requirement with QT5. I need the GUI libraries but my project does not have advanced graphical requirements like swiping or animations. It contains basic controls like texts, buttons and lists/grids. Since I am new to QT especially to this part, I would like to hear whether
1) Is this a viable option and should I keep digging more into this? Any feedback would be welcome to know if it's worthy to spend time to attempt this or if it's a lost cause.
2) If there are options, could you point me in the right direction to look more into this?
It depends on the OS you are running on your SOC.
It's definitely possible when running Linux. You can e.g. use Qt Lite for configuring a minimal static build of Qt. Then you deploy your statically linked application to the device. You can check this blog post for experiences with Qt Lite.
One possibility is to go with Qt for Device Creation for the whole embedded Linux based software stack which boots to a UI implemented with Qt technologies.
I'm doing and embedded linux+qt project and I was wondering what was the base memory consuption of the linux kernel plus some basic services. Just enough to run some framebuffer based application.
I ended up in here: http://qt-project.org/doc/qt-4.8/requirements-embedded-linux.html but as I'm reading that seems like it's just the qt requirements without counting the linux overhead.
Can someone point me to a more detailed resource on the topic?
The numbers in the table you cited look reasonable.
The actual answer is "it depends". Yes, Virginia: it is possible to have a working OS and a Qt-based GUI in under 4MB.
The actual memory usage will vary wildly, depending on:
Which kernel you use
How you configure your kernel build
Which kernel drivers you load at runtime
What you start up during system init
Etc etc
Book recommendation:
Embedded Linux Primer, Christopher Hallinan
ALSO: here's a list of prebuilt-distros with GUIs that all run on Pentium IVs with 512MB RAM:
http://www.osnews.com/story/26087
I would suggest using Yocto for such builds, but you can also take a look at the upcoming "Boot to Qt" project which is basic a Qt 5 replacement for Qt embedded with Qt 4. I would not suggest looking into the link you pasted in your question.
You should definitely focus on Qt 5 for several reasons. The foremost is probably because you can get hardware acceleration and Qt got a lot of utilization for embedded, including decoupling the QtWidgets module, and so forth.
Here you can find the technology preview that the guy in Norway are working on. This is just for future reference:
http://blog.qt.digia.com/blog/2013/05/21/introducing-boot-to-qt-a-technology-preview/
I would start using the Yocto project for now. We have worked on a "meta-qt5" layer which is not perfect, but good enough. Yocto will also take care of the Linux with "minimal images", et cetera.
Not sure if you had seen the classic example a couple of years ago, but there was a "Qt boot" for an embedded Linux board which happened within a second. Here is the link to the reading material. Unfortunately, the original video does not seem to be available anymore.
http://www.embedded-bits.co.uk/2011/1-second-linux-boot-to-qt/
I have to port the application which is using the Qt3 library to a new embedded OS.
If the OS is not yet selected but it is known that it is not Linux/Windows, what is the general way of porting procedure? I mean what files should be changed and what are independed and can be left as is?
If it's not Linux or Windows you could be pretty stuck out the door. Qt is highly dependent on libraries being available for it to be able to build. This is not a no however, it has been ported to Symbian which is obviously not Linux or Windows. The big thing on that though is that it was ported by Nokia, who had amount of time and amount of developers, where those X's are reasonably big numbers.
Start by getting the Qt everywhere source code, work out what bits you're going to need and which bits you won't. If the platform has gcc/g++/another reasonably well understood C++ compiler you're in luck and you should be able to build the bits as needed. Qt does depend on standard libraries throughout so make sure they're also available in some form for the platform, and it does use some non-standard ones for some components so you'll need to know which aspects of Qt you're working on. You should be able to get some sort of toolchain from there.
I'd stick to platforms that Qt supports though, if you go off the beaten track you're looking into increasing your dev times and having very little useful support.
I would want to compile existing software into presentation that can later be run on different architectures (and OS).
For that I need a (byte)code that can be easily run/emulated on another arch/OS (LLVM IR? Some RISC assemby?)
Some random ideas:
Compiling into JVM bytecode and running with java. Too restricting? C-compilers available?
MS CIL. C-Compilers available?
LLVM? Can Intermediate representation be run later?
Compiling into RISC arch such as MMIX. What about system calls?
Then there is the system call mapping thing, but e.g. BSD have system call translation layers.
Are there any already working systems that compile C/C++ into something that can later be run with an interpreter on another architecture?
Edit
Could I compile existing unix software into not-so-lowlevel binary, which could be "emulated" more easily than running full x86 emulator? Something more like JVM than XEN HVM.
There are several C to JVM compilers listed on Wikipedia's JVM page. I've never tried any of them, but they sound like an interesting exercise to build.
Because of its close association with the Java language, the JVM performs the strict runtime checks mandated by the Java specification. That requires C to bytecode compilers to provide their own "lax machine abstraction", for instance producing compiled code that uses a Java array to represent main memory (so pointers can be compiled to integers), and linking the C library to a centralized Java class that emulates system calls. Most or all of the compilers listed below use a similar approach.
C compiled to LLVM bit code is not platform independent. Have a look at Google portable native client, they are trying to address that.
Adobe has alchemy which will let you compile C to flash.
There are C to Java or even JavaScript compilers. However, due to differences in memory management, they aren't very usable.
Web Assembly is trying to address that now by creating a standard bytecode format for the web, but unlike the JVM bytecode, Web Assembly is more low level, working at the abstraction level of C/C++, and not Java, so it's more like what's typically called an "assembly language", which is what C/C++ code is normally compiled to.
LLVM is not a good solution for this problem. As beautiful as LLVM IR is, it is by no means machine independent, nor was it intended to be. It is very easy, and indeed necessary in some languages, to generate target dependent LLVM IR: sizeof(void*), for example, will be 4 or 8 or whatever when compiled into IR.
LLVM also does nothing to provide OS independence.
One interesting possibility might be QEMU. You could compile a program for a particular architecture and then use QEMU user space emulation to run it on different architectures. Unfortunately, this might solve the target machine problem, but doesn't solve the OS problem: QEMU Linux user mode emulation only works on Linux systems.
JVM is probably your best bet for both target and OS independence if you want to distribute binaries.
As Ankur mentions, C++/CLI may be a solution. You can use Mono to run it on Linux, as long as it has no native bits. But unless you already have a code base you are trying to port at minimal cost, maybe using it would be counter productive. If it makes sense in your situation, you should go with Java or C#.
Most people who go with C++ do it for performance reasons, but unless you play with very low level stuff, you'll be done coding earlier in a higher level language. This in turn gives you the time to optimize so that by the time you would have been done in C++, you'll have an even faster version in whatever higher level language you choose to use.
The real problem is that C and C++ are not architecture independent languages. You can write things that are reasonably portable in them, but the compiler also hardcodes aspects of the machine via your code. Think about, for example, sizeof(long). Also, as Richard mentions, there's no OS independence. So unless the libraries you use happen to have the same conventions and exist on multiple platforms then it you wouldn't be able to run the application.
Your best bet would be to write your code in a more portable language, or provide binaries for the platforms you care about.
I would like to experiment with ideas about distributed file synchronization/replication. To make it efficient when the user is working, I would like to implement some kind of daemon to monitor changes in some directory (e.g. /home/user/dirToBeMonitored or c:\docs and setts\user\dirToBeMonitored). So, I could be able to know which filename was added/changed/deleted at every time (or within a reasonable interval).
Is this possible with any high-medium level language?. Do you know some API (and in which language?) to do this?
Thanks.
The APIs are totally different for Windows, Linux, Mac OS X, and any other Unix you can name, it seems. I don't know of any cross-platform library that handles this in a consistent way.
A bonified answer, albeit one that requires a largish library dependency (well-worth it IMO)!
QT provides the QFileSystemwatcher class, which uses the native mechanism of the underlying platform.
Even better, you can use the QT language bindings for Python or Ruby. Here is a simple PyQT4 application which uses QFileSystemWatcher.
Notes
A good reference on on creating deployable PyQT4 apps, especially on OSX but should work for Windows also.
Same solution previously posted here.
Other cross-platform toolkits may also do the trick (for example Gnome's GIO has GFileMonitor, although it is UNIX only and doesn't support OSX's FSEvents mechanism afaik).
In Linux it is called inotify.
And on OS X it's called fsevents. It's an OS-level API, so it's easiest to access from C or C++. It should be accessible from nearly any language, although bindings for your preferred language may not have been written yet.