l4-hackers September 2014

l4-hackers@os.inf.tu-dresden.de

20 participants
29 discussions

GCC 4.9
by Stefan Wallentowitz 09 Sep '14

09 Sep '14

Dear L4 hackers, we recently started porting FIASCO to OpenRISC and our cross-compiler is GCC 4.9. I added the required libraries to our project mirror at github: https://github.com/TUM-LIS/optimsoc-fiasco/tree/for-upstream-gcc4.9 Are you interested in this? Do you want me to submit a patch for the latest public svn tree? Regards, Stefan

2 1

Preventing access to a part of the address space from Sigma0
by Masti Ramya Jayaram 09 Sep '14

09 Sep '14

Dear all, I would like to prevent access to a part of the address space from sigma0 for security reasons on my board. For example, I would like to prevent even sigma from accessing say 0xc0000000 - 0xffffffff of memory. What is the best way to accomplish this and test it? Thanks, ramya

2 1

Checking executables before running them in L4re/fiasco
by Masti Ramya Jayaram 09 Sep '14

09 Sep '14

Dear all, I am using fiasco/l4re to run applications (say hello world) and L4linux. The module list in the two cases looks like this: entry hello-cfg kernel fiasco -serial_esc -freq=533000 roottask moe rom/hello.cfg module l4re module ned module hello.cfg module hello entry l4linux kernel fiasco -serial_esc -freq=533000 roottask moe rom/l4linux.cfg module l4re module ned module l4linux.cfg module io module vmlinuz module ramdisk-x86.rd I would like to add some functionality that performs a few checks on hello and vmlinuz before they are executed. In this context, I have the following questions: a. Where is it best to implement such functionality - moe or ned? Or should I implement a separate server and do this? b. I need access to an special range of IO memory when I perform the checks. i guess I could always use the io module and do "l4io_request_iomem()" . Would that be the right way to do it from moe/ned? c. Is there a way to allocate/reserve a dedicated range of physical memory to these applications (hello/vmlinuz) so that they get allocated the same chunk everytime (of physical memory)? Thanks in advance, ramya

2 12

Adding a custom library to l4/pkg/bootstrap
by Masti Ramya Jayaram 05 Sep '14

05 Sep '14

Hi all, I have managed to compile a custom C library (libscc-sec) as part of the l4 packages. Now I would like to use it in the bootstrap package (make a call to a function in the library). I did the following: a. Added the library as a requirement in pkg/bootstrap/Control requires: drivers_uart drivers_of libc l4util cxx_io libscc-sec b. Modified the pkg/bootstrap/server/src/Make.rules so that it also looks for libscc-sec by: L4_LIBS = -static -nostdlib $(DRV_LIBS) -lcxx_base -lcxx_io -llibscc-sec But on compiling, I get an error that says that -llibscc-sec not found The package is getting compiled and I can see the library libscc-sec.a in the build folder. What am I missing? Thanks, ramya

3 6

Compiling the Fiasco.OC +L4Re for Raspberry PI
by Waldo Paz Rodriguez 03 Sep '14

03 Sep '14

hello hanzel, I have seen in l4hacker list that you have compiled fiasco.oc + l4re for Raspberry PI whithout problems. Iá¸¿ trying to compile it for rpi. I follow the next steps> 1- Compile Fiasco.OC for Broadcom2835 2- Compile L4Re for RaspberryPI model B with MODULE_SEARCH_PATH pointing to FOC builddir 3- Generate the hello example ELF image I have transformed it into a binary using: /usr/local/genode-gcc/bin/genode-arm-objcopy -Obinary bootstrap.elf l4.img I have used the three elf image and have used to the linaro and genode toolchain (both). When I have put the binary on the PI only can see the colored screen. I am using the Fiasco.OC +L4Re from 2014022815 snapshot. I am doing something wrong? Can you help me?

2 1

Passing data to a thread
by Valentin Hauner 03 Sep '14

03 Sep '14

Hi, I'm trying to pass some data to a newly created thread. Since l4_thread_ex_regs only accepts an instruction pointer, but no data pointer, it seems impossible to me to use a parameterized function such as: > thread1_func(void *data) { /* Read the data ... */ } So far, my efforts are: 1. I've tried filling the stack allocated to each newly created thread with > thread_stack = malloc(8 << 10); > thread_stack[0] = my_data; but then I'm not able to pop the data from the stack in thread1_func. 2. I've tried using IPC: The main thread which is creating the new threads sends the data to the destination thread. I'm using l4_ipc_send in the main thread and l4_ipc_receive in thread1_func just like in your utcb-ipc example. It works great, but unfortunately the call of l4_ipc_send leads to an _immediate_ execution of thread1_func (timeouts: L4_IPC_NEVER), so my scheduling policy is not respected. So how can I pass data to a thread in a way that is equivalent to passing arguments to functions? Best regards, Valentin

2 2

Fault in function writev (ulibc)
by Masti Ramya Jayaram 02 Sep '14

02 Sep '14

Dear all, I am running a custom application package on fiasco using a serial line (no keyboard/screen) and I intermittently get faults caused by printf. I notice that on disassembling the code around the error that the function is called writev in ulibc. Has anyone experienced something similar before? Thanks, ramya

2 2

New package linking fails
by Masti Ramya Jayaram 01 Sep '14

01 Sep '14

Dear all, I am trying to create a new package under l4/pkg. The linking of the package fails and it has to do with the linker path but I am not sure how to fix it. Here are the details. Package structure: (relevant parts) l4/pkg/test/server/src/main.cc l4/pkg/test/server/src/folder1/test1.c l4/pkg/test/include/test1.h l4/pkg/test/server/src/Makefile PKGDIR ?= ../.. L4DIR ?= $(PKGDIR)/../.. TARGET = $(PKGNAME) # list requirements of your program here REQUIRES_LIBS = libsigma0 libirq l4util l4re_c-util # list your .c or .cc files here SRC_C = folder1/test1.c SRC_CC = main.cc include $(L4DIR)/mk/prog.mk The compilation of the package fails with an undefined reference in main.cc to a function in test1.c (defined through test1.h). All the object files are generated and I checked under build. What do you think is the issue? Thanks, ramya

2 3

lwip and C++ compilation/linking issue
by Stark, Josef 01 Sep '14

01 Sep '14

Hello, I want to use TCP/IP (and therefore ankh and lwip) inside a C++ application. The unmodified lwip example (src/l4/pkg/ankh/examples/lwip) compiles fine (and also works without DHCP and with a little tweaking), but it is written and configured as C and not C++. So if I change the filename of main.c to main.cc and the Makefile accordingly, unfortunately it doesn't compile. Following error: ---------------- ... Compiling main.o In file included from /home/josef/Desktop/l4re-snapshot-2014053111/obj/l4/x86/include/contrib/lwip/contrib/src/include/lwip/arch.h:46:0, from /home/josef/Desktop/l4re-snapshot-2014053111/obj/l4/x86/include/contrib/lwip/contrib/src/include/lwip/debug.h:35, from /home/josef/Desktop/l4re-snapshot-2014053111/obj/l4/x86/include/contrib/lwip/contrib/src/include/lwip/opt.h:46, from /home/josef/Desktop/l4re-snapshot-2014053111/obj/l4/x86/include/contrib/lwip/contrib/src/include/lwip/tcpip.h:35, from /home/josef/Desktop/l4re-snapshot-2014053111/obj/l4/x86/include/l4/ankh/lwip-ankh.h:3, from /home/josef/Desktop/test/main.cc:10: /home/josef/Desktop/l4re-snapshot-2014053111/obj/l4/x86/include/contrib/lwip/arch/cc.h:48:13: error: previous declaration of ‘int* __h_errno_location()’ with ‘C++’ linkage extern int *__h_errno_location (void) __THROW __attribute__ ((__const__)); ^ In file included from /home/josef/Desktop/test/main.cc:19:0: /home/josef/Desktop/l4re-snapshot-2014053111/obj/l4/x86/include/uclibc/netdb.h:61:73: error: conflicts with new declaration with ‘C’ linkage extern int *__h_errno_location (void) __THROW __attribute__ ((__const__)); ^ make[1]: *** [main.o] Error 1 make: *** [/home/josef/Desktop/l4re-snapshot-2014053111/obj/l4/x86/ext-pkg/home/josef/Desktop/test/OBJ-x86_586-l4f] Error 2 ----------------- Alright, so I tried surrounding the offending include with 'extern "C" ' (as well as moving it above #include <l4/ankh/lwip-ankh.h> since this file also includes it): extern "C" { #include "arch/cc.h" } I think that it solved the problem, because the error went away, but another error appeared (see below), so I also tried doing the same only to the offending line inside arch/cc.h instead: extern "C" { extern int *__h_errno_location (void) __THROW __attribute__ ((__const__)); } Both approaches led to the same linker error: ------------------ main.o: In function `main': main.cc:(.text.startup+0x11c): undefined reference to `ankhif_init(netif*)' make[1]: *** [test] Error 1 make: *** [/home/josef/Desktop/l4re-snapshot-2014053111/obj/l4/x86/ext-pkg/home/josef/Desktop/test/OBJ-x86_586-l4f] Error 2 ------------------ This is the Makefile (only the line SRC_CC is different to the original example Makefile) ------------------ PKGDIR ?= . L4DIR ?= $(PKGDIR)/../l4re-snapshot-2014053111/src/l4 TARGET = test SRC_CC = main.cc SYSTEMS = x86-l4f arm-l4f REQUIRES_LIBS = lwip libc_be_socket_lwip liblwip_netif_ankh libc_support_misc include $(L4DIR)/mk/prog.mk ------------------ Which library is missing here, or what else could be going wrong? Cheers, Josef

1 1

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

2003

2002

2001

2000

1999

1998

l4-hackers September 2014