- l4-hackers - example.com Site

All *.cpp files are ignored by Makefile
by BogDan 02 Jun '13

02 Jun '13

Hi, I tried to add a few .cpp files to Makefile but it seems that L4Re build systems ignores this extension. This is the content of the Makefile: #################################### PKGDIR ?= ../.. L4DIR ?= $(PKGDIR)/../.. TARGET = test_cpp SRC_CC = main.cpp class1.cpp class2.cpp REQUIRES_LIBS = libstdc++ include $(L4DIR)/mk/prog.mk #################################### I tried to replace SRC_CC with SRC_CXX or SRC_CPP but with no luck ... This bug prevents the port of most of C++ libs out there to L4Re. BTW, is there any chance to use other build-systems (cmake, qbs, premake, SCons, etc.) to build L4Re libs/apps ? Cheers, BogDan.

3 5

How to catch unhandled exceptions ?
by BogDan 02 Jun '13

02 Jun '13

Hi, Is there any way to catch all system unhandled exceptions (seg fault, fpu fault, etc.) and "transform" them into C++ exceptions ? I tried to use a combination of "L4::Thread::control" with "L4::Thread::Attr::exc_handler" to catch the system exception but it didn't worked for me, most probably I didn't understood the meaning of "l4_thread_control_exc_handler" or how to use it. Any example/help will be very appreciated. Of course, after I'll catch the exception, I need a way to "resume" the thread execution to another ip (address) which actually is a simple function that throws the C++ exception, any example/help on the matter will also be highly appreciated. Cheers, BogDan.

2 2

Can't use cxx::Thread
by BogDan 02 Jun '13

02 Jun '13

Hello, I tried to create a very simple application that uses cxx::Thread, but I have some link problems: ... Compiling main.o ==> Linking test_cxx_thread main.o:(.rodata._ZTI10TestThread[_ZTI10TestThread]+0x8): undefined reference to `typeinfo for cxx::Thread' Also I think it will be cleaner/better for developers if run, execute and shutdown methods will be moved to a protected section. This is the main.cc content: /////////////////////////////////////////////// #include <iostream> #include <l4/cxx/thread> #include <thread> class TestThread : public cxx::Thread { public: TestThread():cxx::Thread(true){} private: virtual void run() { std::cout << "TestThread" << std::endl; } }; static void TestStdThread(int param) { std::cout << "TestStdThread param:" << param << std::endl; } int main(int argc, char ** argv) { (void)argc; (void)argv; TestThread cxxTh; cxxTh.start(); std::thread stdTh(TestStdThread, 1); for (;;) { std::cout << "Hello World!" << std::endl; sleep( 1 ); } stdTh.join(); } /////////////////////////////////////////////// This is the Makefile content: #################################### PKGDIR ?= ../.. L4DIR ?= $(PKGDIR)/../.. TARGET = test_cxx_thread SRC_CC = main.cc CXXFLAGS += -std=c++11 REQUIRES_LIBS = cxx_thread cxx_libc_io cxx_io libstdc++ libpthread include $(L4DIR)/mk/prog.mk #################################### Yours, BogDan.

2 2

Kernel debugger in MP mode causes double fault
by Markus Partheymueller 31 May '13

31 May '13

Hi folks, has anyone tried to invoke the kernel debugger in a multiprocessor setup? When I boot the demo scenario in qemu with qemu -smp 2 -cpu core2duo and then hit Escape, the system crashes with a double fault. Without smp or with -cpu qemu32 the problem doesn't occur. Can someone confirm this issue or give a hint what could be wrong? Cheers Markus

2 1

L4Re: FATAL: Exception Operation not permitted: ''"
by Maximin Coste 31 May '13

31 May '13

Hi, I'm running an edited and "old" (September 2012) version of L4Linux & uC-OS/II in top of Fiasco+L4RE (updated here). When launching the image I have generated, I have the following error: MOE: cmdline: moe rom/l4lx_ucos.cfg MOE: Starting: rom/ned rom/l4lx_ucos.cfg MOE: loading 'rom/ned' Ned says: Hi World! Ned: loading file: 'rom/l4lx_ucos.cfg' Ucos | L4Re: FATAL: Exception Operation not permitted: '' Ucos | Exception: Operation not permitted: : backtrace ... l4linux | L4Re: FATAL: Exception Operation not permitted: '' l4linux | Exception: Operation not permitted: : backtrace ... io | Dev_factory: register factory for N2Hw6DeviceE io | GTF: register factory for 7Pci_dev io | GTF: register factory for N2Hw12Msi_resourceE io | Dev_factory: register factory for N2Hw11Gpio_deviceE io | Io service io | Ready. Waiting for request. Can I see more about what's going wrong ? Can I see the backtrace those logs talks about ? :) -- Maximin COSTE-LEENHARDT

2 3

How to boot Fiasco microkernel on raspberry pi
by Rahul Singhal 31 May '13

31 May '13

I want to port Fiasco microkernel to RPi. I found that Fiasco now supports RPi. I am a newbie in this kernel field. I have compiled the L4 microkernel and the L4 runtime environment for RPi succesfully but I don't know how to use the generated binaries to run RPi. I am running on ubuntu 12.04. Could you please tell me the procedure and files to be put on the sd card of RPi so that it boots into the L4 runtime environment.

3 5

sigma0 memory space
by joel＠nider.org 30 May '13

30 May '13

I have been reading about the structure of L4Ka::Pistachio, and discovered that they use idempotent (1:1) page table mapping in their sigma0 process. I am wondering if Fiasco also takes this approach? If so, does the kernel also use idempotent mapping? I would imagine that this would have some implications on demand paging in the kernel. Does the kernel even use demand paging? (I mean at some point during boot the kernel will start to use memory to store kernel objects - either the kernel 'guesses' how much it will ever need and statically allocates that memory, or it's going to have to ask for pages later on, which will be owned by some other process). I tried to read the code, but found it very hard to traverse - the C++ classes don't make it easy to follow a flow. Can you please point out the file(s) where the kernel sets up the memory space for sigma0, and where it passes the initial ownership (from kernel space to sigma0 space)? Thanks, Joel

2 1

Building l4re packages and linking errors
by Bryce Cronkite-Ratcliff 12 May '13

12 May '13

Hi folks, I'm afraid my configuration may have changed recently, possibly due to version changes in gcc, and I'm having some trouble compiling l4re, so I wanted to ping you to see if you had any insights. Two separate issues I've had: in an existing l4re build, executing make in an individual package sub-directory fails to find basic system headers in included files ("stddef.h" and "stdarg.h" for example). I'm running make O=basel4rebuilddir On the theory that l4re needs to be recompiled in order to set up links that may have changed since compilation of the last build, I've tried building l4re again, but I similarly run into a linker problem here: gcc: symbol lookup error: /home/bryce/L4Linux/x86/l4re4/tool/gendep/64/libgendep.so: undefined symbol: dlsym If you have any thoughts on what I might try, let me know! Thank you! Bryce

2 2

Questions about Fiasco.OC and L4Re
by BogDan 12 May '13

12 May '13

Hello, First I'd like to introduce myself, I'm BogDan Vatra a free software hacker. Recently I've watched these http://wiki.tudos.org/HOWTOs#Introduction_to_Fiasco.OC_and_L4Re presentations on youtube. After I've seen them (many many times) I became fascinated. I started to "play" a little bit with Fiasco.OC and L4Re, I created a few apps and I must confess that I'm extremely impressed by your work. To me Fiasco.OC and L4Re seems way too good to be true, and I wonder if there are any "hidden traps" which prevents people to use it to create a general-purpose operating system on top of Fiasco.OC and L4Re, so my first question is: Can Fiasco.OC and L4Re be used to create free (GPL, BSD, apache, etc.) and non-free apps? I checked L4Re licenses and it seems that it has the necessary exceptions, but I'd like to be 100% sure :). I also look to Genode frameowrk and I it looks to me that it cannot be used to create non-GPL apps, so, it can not be used to create a general-purpose operating system where any kind of apps can run on it. I also have a few questions about the completeness of Fiasco.OC and L4Re: - does L4Re have any OpenGL support? I checked the source code and it seems that the only way to do graphics is via goos and it doesn't seem to have any OpenGL support. If I am right wouldn't be easier to port (and upstream) an existing server (e.g. xorg, wayland, mir, etc.) to L4Re, than to create one from scratch and maintain it? - does L4Re have any sound support? If it hasn't, then how hard will be to implement or to port an existing implementation? - does L4Re have any real filesystem support (e.g. ext2/3/4, xfs, raiser, etc.)? It is ok for testing to use qemu or grub to load simple apps, but for a real o.s., apps need to be loaded from disk. Again, if it hasn't, then how hard will to implement it or to port an existing implementation? - it is possible to use a debugger (e.g gdb(server)) to debug (user space) apps? - last but not least, is there any repository for Fiasco.OC and L4Re with logs? http://svn.tudos.org/repos/oc/tudos/trunk/ has no useful logs and IMHO it makes very hard any contribution to it. I've seen that you have a few repos on github (https://github.com/TUD-OS) do you have any plans to move Fiasco.OC and L4Re to github as well? IMHO being able to fork a repository and to create a pull request, to be able to open an issue and watch its progress will make contributors life much more easier :) ! Many thanks for your work! Yours, BogDan.

2 1

Question regarding the "Jdb::handle_user_request()' function
by Christian Prochaska 11 May '13

11 May '13

Hello, the 'Jdb::enter_jdb()' function in file 'jdb.cpp' contains the following code: // check for int $3 user debugging interface if (foreach_cpu(&handle_user_request, true)) { close_debug_console(cpu); leave_trap_handler(cpu); return 0; } I'm not sure why in this case the second argument of 'foreach_cpu()' is true. I understand that it causes the result to be true only if the 'handle_user_request()' function returned true for every currently running CPU. But I found that this can be a problem when using the '-jdb_cmd=JS' kernel argument: if only one CPU is currently running, the result of 'foreach_cpu()' is true, but when two CPUs are currently running, the 'handle_user_request()' function returns true only for one CPU, which makes the end result false and causes the kernel to stop with the JDB prompt shown. Can somebody please explain? Christian

2 4