qemu-arm Segmentation Fault - cmake

Take a simple file:
int main(void) {
return 0;
}
My CMakeLists.txt file:
cmake_minimum_required(VERSION 3.20)
project(TRA
VERSION 0.0.1
DESCRIPTION "STM32 Traffic Controller"
LANGUAGES C ASM)
set(BSP_DRIVER_FILES
BSP/tra/startup_stm32f429xx.s
BSP/tra/Core/Src/system_stm32f4xx.c
)
set(BSP_COMPILER_FLAGS
-DUSE_HAL_DRIVER
-DSTM32F429xx
-I${CMAKE_SOURCE_DIR}/BSP/tra/Core/Inc
-I${CMAKE_SOURCE_DIR}/BSP/tra/Drivers/STM32F4xx_HAL_Driver/Inc
-I${CMAKE_SOURCE_DIR}/BSP/tra/Drivers/STM32F4xx_HAL_Driver/Inc/Legacy
-I${CMAKE_SOURCE_DIR}/BSP/tra/Drivers/CMSIS/Device/ST/STM32F4xx/Include
-I${CMAKE_SOURCE_DIR}/BSP/tra/Drivers/CMSIS/Include
)
add_executable(tra)
target_sources(tra PRIVATE
${BSP_DRIVER_FILES}
code/tra.c
)
set(COMPILER_FLAGS
${BSP_COMPILER_FLAGS}
-O0 -g -ggdb3
-mcpu=cortex-m4 -mthumb -mfloat-abi=soft
-fdata-sections -ffunction-sections
--specs=rdimon.specs
)
target_compile_options(tra PRIVATE
${COMPILER_FLAGS}
)
target_link_options(tra PRIVATE
-mcpu=cortex-m4 -mthumb -mfloat-abi=soft
--specs=rdimon.specs -lm -lc
-Wl,--gc-sections
)
Running qemu-arm build/tra.
I get
qemu: uncaught target signal 11 (Segmentation fault) - core dumped
Segmentation fault (core dumped)
My understanding was that this issue was commonly caused by qemu not handling hardware FPU, however I'm specifying software floating point here. Also, I have removed FPU initialization code from my startup.c file.
With gdb:
# Start
qemu-arm -g 1234 build/tra
# Attach
arm-none-eabi-gdb -q --nh -ex 'file build/tra' -ex 'target remote localhost:1234'
I get the segmentation fault at SystemInit(). Running gdb list, it shows the commented out FPU initialization code, so I don't see how that could be the issue.
Therefore, I'm at a loss as to what is going on.

You're building a bare-metal binary for a Cortex-M CPU, and then trying to run it on qemu-arm, which is the emulator for running Cortex-A Linux binaries.
If you're building a bare-metal binary, you need to build it for the exact machine type that you want to run it on, which means you need to target a machine type supported by qemu-system-arm, and you need to tell qemu-system-arm to use that machine type.

As suggested by Peter Maydell require full system emulation. For cortex-M, this was provided by xPack QEMU

Related

Compile errors running the ot-br-posix ./script/setup on RPi4

I'm trying to run the ./script/setup, but get compile errors:
Please note that the total 65 steps listed below is because I've restarted the setup script. The initial number of steps were closer to 465.
[1/65] Building CXX object src/common/CMakeFiles/otbr-common.dir/mainloop.cpp.o
FAILED: src/common/CMakeFiles/otbr-common.dir/mainloop.cpp.o
/usr/bin/c++ -DHAVE_LIBSYSTEMD=1 -DOTBR_ENABLE_BACKBONE_ROUTER=1 -DOTBR_ENABLE_BORDER_AGENT=1 -DOTBR_ENABLE_BORDER_ROUTING=1 -DOTBR_ENABLE_BORDER_ROUTING_COUNTERS=1 -DOTBR_ENABLE_DBUS_SERVER=1 -DOTBR_ENABLE_DNSSD_DISCOVERY_PROXY=1 -DOTBR_ENABLE_NAT64=1 -DOTBR_ENABLE_NOTIFY_UPSTART=1 -DOTBR_ENABLE_REST_SERVER=1 -DOTBR_ENABLE_SRP_ADVERTISING_PROXY=1 -DOTBR_ENABLE_SRP_SERVER_AUTO_ENABLE_MODE=1 -DOTBR_ENABLE_VENDOR_INFRA_LINK_SELECT=0 -DOTBR_MESHCOP_SERVICE_INSTANCE_NAME="\"OpenThread BorderRouter\"" -DOTBR_PACKAGE_NAME=\"OpenThread_BorderRouter\" -DOTBR_PACKAGE_VERSION=\"0.3.0-0cdef3c\" -DOTBR_PRODUCT_NAME=\"BorderRouter\" -DOTBR_SYSLOG_FACILITY_ID=LOG_USER -DOTBR_VENDOR_NAME=\"OpenThread\" -I../../include -I../../src -Ithird_party/openthread/repo/etc/cmake -I../../third_party/openthread/repo/etc/cmake -I../../third_party/openthread/repo/include -I../../third_party/openthread/repo/src/posix/platform/include -I../../third_party/openthread/repo/src -Wall -Wextra -Werror -Wfatal-errors -Wuninitialized -Wno-missing-braces -std=c++11 -MD -MT src/common/CMakeFiles/otbr-common.dir/mainloop.cpp.o -MF src/common/CMakeFiles/otbr-common.dir/mainloop.cpp.o.d -o src/common/CMakeFiles/otbr-common.dir/mainloop.cpp.o -c ../../src/common/mainloop.cpp
In file included from /usr/include/c++/8/list:63,
from ../../src/common/mainloop_manager.hpp:41,
from ../../src/common/mainloop.cpp:30:
/usr/include/c++/8/bits/stl_list.h:811:19: error: expected ‘)’ before ‘&’ token
list(_InputIterat&... __args)`
compilation terminated due to -Wfatal-errors.
I receive a lot more errors, but they follow the same pattern as above.
I have followed the guide from openthread.io to setup an Open Thread Border Router
The execution of the bootstrap script ran smoothly.
Additional information:
Git local repository path: ~/src/openthread/ot-br-posix
Command for executing the setup script:
pi#raspberrypi:~/src/openthread/ot-br-posix$> INFRA_IF_NAME=eth0 ./script/setup
RPi OS: Recommended image from the guide Raspberry Pi OS lite
Libgcc versions:
libgcc-8-dev/oldstable,now 8.3.0-6+rpi1 armhf [installed,automatic]
libgcc1/oldstable,now 1:8.3.0-6+rpi1 armhf [installed]
Cmake versions:
cmake-data/oldstable,now 3.16.3-3~bpo10+1 all [installed,automatic]
cmake/oldstable,now 3.16.3-3~bpo10+1 armhf [installed]

Redis build failure

I'm attempting to build redis on my aws linux server in order to get access to the redis-cli and connect to my redis instance which is also running successfully in aws.
my gcc is:
gcc version 4.8.5 20150623 (Red Hat 4.8.5-28) (GCC)
I downloaded the source and started the build:
wget http://download.redis.io/redis-stable.tar.gz
tar xvzf redis-stable.tar.gz
cd redis-stable
make distclean
make
It built the dependencies and then gave me this:
gcc -std=gnu99 -Wall -pipe -g3 -O3 -funroll-loops -c -D_GNU_SOURCE -D_REENTRANT -Iinclude -Iinclude -o src/tsd.o src/tsd.c
gcc -std=gnu99 -Wall -pipe -g3 -O3 -funroll-loops -c -D_GNU_SOURCE -D_REENTRANT -Iinclude -Iinclude -o src/witness.o src/witness.c
ar crus lib/libjemalloc.a src/jemalloc.o src/arena.o src/background_thread.o src/base.o src/bin.o src/bitmap.o src/ckh.o src/ctl.o src/div.o src/extent.o src/extent_dss.o src/extent_mmap.o src/hash.o src/hooks.o src/large.o src/log.o src/malloc_io.o src/mutex.o src/mutex_pool.o src/nstime.o src/pages.o src/prng.o src/prof.o src/rtree.o src/stats.o src/sz.o src/tcache.o src/ticker.o src/tsd.o src/witness.o
make[3]: Leaving directory `/home/ec2-user/redis-stable/deps/jemalloc'
make[2]: Leaving directory `/home/ec2-user/redis-stable/deps'
CC adlist.o
CC quicklist.o
CC ae.o
CC anet.o
CC dict.o
CC server.o
In file included from server.c:30:0:
server.h:1044:5: error: expected specifier-qualifier-list before ‘_Atomic’
_Atomic unsigned int lruclock; /* Clock for LRU eviction */
^
server.c: In function ‘serverLogRaw’:
server.c:1028:31: error: ‘struct redisServer’ has no member named ‘logfile’
int log_to_stdout = server.logfile[0] == '\0';
^
server.c:1031:23: error: ‘struct redisServer’ has no member named ‘verbosity’
if (level < server.verbosity) return;
^
server.c:1033:47: error: ‘struct redisServer’ has no member named ‘logfile’
fp = log_to_stdout ? stdout : fopen(server.logfile,"a");
...and a lot more
I assume that I must have done something wrong as this is code that undoubtedly builds for everyone else. I'm a java developer with little to no c background. Any pointers will be appreciated.
Jay
Appears that gcc >= 4.9 is required to build redis 6x ...
Problem here is latest Redis version(6.0.7) depends on _Atomic which is not supported in older gcc 4.8.5.
One can also compile latest Redis code with clang if updating gcc version is not possible.
$ cd redis-stable
$ CC=clang make
i have successed.
at first i want to install gcc6, but failed again and agin, then i notise that,devtoolset-6 is deprecated. Maybe you should try -7 or -8.
then i use the flow cmd and it worked
yum install -y devtoolset-6-gcc devtoolset-7-gcc-c++
so i hope it can also help you
Thanks for the responses. The solution I ended up with was to build a new server with CentOS8 and gcc 8.3.1. This combination built the redis server 6.0.3 with no issues. Moral: you need to have up to date infra and software in order to properly build new software.
Download older version, seems some issue with 6.0.*
That worked for me on ubunt 14.04 LTS

Compile openjdk 7 on arm ubuntu

I am trying to compile openjdk 7 on my arm ubuntu:
make all ALLOW_DOWNLOADS=true DISABLE_HOTSPOT_OS_VERSION_CHECK=ok
Then I received this error:
g++ -DLINUX -D_GNU_SOURCE -DIA32 -I/home/darklord/Develop/jdk7/hotspot/src/share/vm/prims -I/home/darklord/Develop/jdk7/hotspot/src/share/vm -I/home/darklord/Develop/jdk7/hotspot/src/cpu/x86/vm -I/home/darklord/Develop/jdk7/hotspot/src/os_cpu/linux_x86/vm -I/home/darklord/Develop/jdk7/hotspot/src/os/linux/vm -I/home/darklord/Develop/jdk7/hotspot/src/os/posix/vm -I/home/darklord/Develop/jdk7/hotspot/src/share/vm/adlc -I../generated -DASSERT -DTARGET_OS_FAMILY_linux -DTARGET_ARCH_x86 -DTARGET_ARCH_MODEL_x86_32 -DTARGET_OS_ARCH_linux_x86 -DTARGET_OS_ARCH_MODEL_linux_x86_32 -DTARGET_COMPILER_gcc -DCOMPILER2 -DCOMPILER1 -fno-rtti -fno-exceptions -D_REENTRANT -fcheck-new -fvisibility=hidden -m32 -march=i586 -pipe -Werror -g -c -o ../generated/adfiles/adlparse.o /home/darklord/Develop/jdk7/hotspot/src/share/vm/adlc/adlparse.cpp
g++: error: unrecognized argument in option '-march=i586'
It seems it is trying compile using x86 configuration. So how can I make the build pass on ARM machine?
You have to specify proper architecture option for g++. Reference here.
-march=name
This specifies the name of the target ARM architecture. GCC uses this name to determine what kind of instructions it can emit when
generating assembly code. This option can be used in conjunction with
or instead of the -mcpu= option. Permissible names are: armv2',
armv2a', armv3',armv3m', armv4',armv4t', armv5',armv5t',
armv5te',armv6', armv6j',iwmmxt', `ep9312'.
Please make sure you refer proper version docs of gcc

How to use cmake's 'make install' from a pbuilder env debian/rules script?

This is to compile and link a static library (so only a build time dependency) that the source is fetched from a repository (just like the source of the main program) on a ubuntu launchpad build bot.
currently i am doing:
#!/usr/bin/make -f
export PREFIX=/usr
export CFLAGS= -O3 -fomit-frame-pointer -flto -fwhole-program
export CXXFLAGS= -O3 -fomit-frame-pointer -flto -fwhole-program
%:
dh $#
override_dh_auto_configure:
cd src/munt;cmake -DCMAKE_CXX_FLAGS="-O3 -fomit-frame-pointer -flto" mt32emu;make;make install
#...compile of the program that depends on mt32emu...
But it fails with:
Install the project...
-- Install configuration: ""
-- Installing: /usr/local/lib/libmt32emu.a
CMake Error at cmake_install.cmake:36 (FILE):
file INSTALL cannot copy file
"/tmp/buildd/dosbox-0.74+20121225/src/munt/libmt32emu.a" to
"/usr/local/lib/libmt32emu.a".
make[2]: *** [install] Error 1
make[2]: Leaving directory `/tmp/buildd/dosbox-0.74+20121225/src/munt'
make[1]: *** [override_dh_auto_configure] Error 2
make[1]: Leaving directory `/tmp/buildd/dosbox-0.74+20121225'
make: *** [build] Error 2
dpkg-buildpackage: error: debian/rules build gave error exit status 2
E: Failed autobuilding of package
I: unmounting /var/cache/pbuilder/ccache filesystem
I: unmounting dev/pts filesystem
I: unmounting proc filesystem
I: cleaning the build env
I: removing directory /var/cache/pbuilder/build//2751 and its subdirectories
The idea is to install a static library dependency that is is not packaged in the ubuntu repositories in the launchpad pbuilder env, so it can be used as if it was a system dependency already.
If i try to do 'sudo make install' (and add sudo to the build-deps in debian/control), it asks me for the 'pbuilder' password when testing locally, which i'm assuming will hang the machine on the ubuntu buildbots.
edit: it actually fails on the buildbots because 'no tty present and no askpass program specified'.
There are several things you can do to clean up your rules file, especially when you are using dh.
In the % target, all of the dh command take a parameter builddirectory, which specifies what directory you are building in. This tells the builder to cd to that directory and then call commands (make, cmake, etc.).
In addition, you should just let dh install the files for you. This is done automatically. You shouldn't have to call make install manually.
Here's a slightly easier-to-read rules file:
#!/usr/bin/make -f
export PREFIX=/usr
export CFLAGS= -O3 -fomit-frame-pointer -flto -fwhole-program
export CXXFLAGS= -O3 -fomit-frame-pointer -flto -fwhole-program
%:
dh $# --builddirectory=src/munt
override_dh_auto_configure:
cd src/munt && cmake -DCMAKE_CXX_FLAGS="-O3 -fomit-frame-pointer -flto" mt32emu
#...compile of the program that depends on mt32emu...
Is this just a permissions issue? (i.e. -- must use 'sudo' to install to '/usr/local'?)
Must you install it to '/usr/local'?
If it's just a static library, purely needed for the build of the "the program that depends on mt32emu" then you could put it anywhere, and just tell the dependent program where it is.
To install somewhere else, use -DCMAKE_INSTALL_PREFIX=/directory/where/you/have/write/privileges. Or use DESTDIR= with the make install.
I eventually 'solved' this by depending on launchpad repository dependencies, that is, building a whole package for the library and building that on launchpad and then importing the archive where that was placed to my other builds. Made it explicit i guess.

Can not compile simple C# application with mkbundle

I have written some console "Hello world"-like app. and have followed c# cywgwin mono mkbundle windows 7 - cannot compile file answer. But I have got:
$ mkbundle -o Fur Furries.exe --deps -z
OS is: Windows
Sources: 1 Auto-dependencies: True
embedding: C:\Monotest\Furries.exe
compression ratio: 40.43%
embedding: C:\Soft\Mono\lib\mono\4.0\mscorlib.dll
compression ratio: 34.68%
Compiling:
as -o temp.o temp.s
gcc -mno-cygwin -g -o Fur -Wall temp.c `pkg-config --cflags --libs mono-2|dos2un
ix` -lz temp.o
temp.c: In function `main':
temp.c:173: warning: implicit declaration of function `g_utf16_to_utf8'
temp.c:173: warning: assignment makes pointer from integer without a cast
temp.c:188: warning: assignment makes pointer from integer without a cast
/tmp/ccQwnxrF.o: In function `main':
/cygdrive/c/Monotest/temp.c:173: undefined reference to `_g_utf16_to_utf8'
/cygdrive/c/Monotest/temp.c:188: undefined reference to `_g_utf16_to_utf8'
collect2: ld returned 1 exit status
[Fail]
It's in Windows XP.
First of all, prepare development environment:
Install Mono. For example, you have installed it into "C:\Soft\Mono".
Install Cygwin. When selecting which packages to install select following: gcc-mingw, mingw-zlib, pkg-config, nano.
Start Cygwin Bash shell (either using a link or "bash --login -i" command).
Open "$HOME/.bashrc" with "nano" ("nano ~/.bashrc"). Don't use editors which don't preserve end-of-line-s ("CR", "LF", "CR/LF" or other), or it will corrupt the file!
Add following lines to the end of the file:
export PKG_CONFIG_PATH=/cygdrive/c/Soft/Mono/lib/pkgconfig
export PATH=$PATH:/cygdrive/c/Soft/Mono/bin
Restart Cygwin Bash shell.
After that you can compile your assemblies with "mkbundle":
Perform the following command: "mkbundle -c -o host.c -oo bundle.o --deps YourAssembly.exe <additional arguments>". You also may optionally pass "-z" to compress resultant bundle. You should get "host.c" and "bundle.o" files.
In "host.c" you should remove "_WIN32" "branch" (except "#include <windows.h>" one). It doesn't work. You may do it just by adding "#undef _WIN32" right after following lines in it:
#ifdef _WIN32
#include <windows.h>
#endif
So you'll get:
#ifdef _WIN32
#include <windows.h>
#endif
#undef _WIN32
Perform the following command: "gcc -mno-cygwin -o ResultantBundle.exe -Wall host.c `pkg-config --cflags --libs mono-2|dos2unix` bundle.o <additional arguments>". If you added a -z additional argument in step 2, you must add a -lz additional argument in this step.
You will get "ResultantBundle.exe". This is your Mono application packed as standalone executable.
It still requires "mono-2.0.dll" and some additional DLL-s and resources you depended on during development (for example, it may require GTK# native DLL-s) but it doesn't require full Mono runtime to run.
Just wanted to add that if you pass -z to mkbundle then you'll need to pass -lz to gcc. I had some issues getting an application with winforms and net access to work properly, and I had to copy machine.config from C:\Mono\etc\mono\4.0\machine.config to where my application was. I then passed --machine-config machine.config to mkbundle.
All of these steps are pretty confusing and frustrating, why is not as simple as just typing mkbundle --deps app.exe? I tried making a change to the template used by mkbundle and compiling it myself, but it wont run. I've gone as far now as to download the mono source and attempt to build the whole thing, but I doubt it will work. If anyone can explain what the hell is going on with mkbundle to make this so annoying, I'd be interested in contributing.
after you have the temp.o and temp.c, you can add them to visual c++ to make a windows application with other sources.