I am trying to build LLVM for target X86:
$ git clone ssh://xxx/llvm-project.git
Cloning into 'llvm-project'...
...
Updating files: 100% (118553/118553), done.
$ cd llvm-project
$ mkdir build
$ cd build
$ cmake ../llvm -DCMAKE_BUILD_TYPE=Release
...
-- Configuring done
-- Generating done
-- Build files have been written to: /xxx/llvm-project/build
$ time cmake --build . --target X86 --config Release
Built target X86
real 0m1.183s
user 0m1.152s
sys 0m0.031s
Here we see that cmake quickly exists w/o actually building target X86. Why? How to fix? Any ideas?
The X86 is listed as a target:
$ cmake --build . --target help | grep -P '... X86' -w
... X86
Note: LLVM branch is origin/release/14.x:
$ git branch
* (HEAD detached at origin/release/14.x)
The last commit is:
$ git log | head
commit f28c006a5895fc0e329fe15fead81e37457cb1d1
Author: Tom Stellard <tstellar#redhat.com>
Date: Wed Jun 22 09:46:24 2022 -0700
Bump version to 14.0.6
Related
I am trying to find a way to enable incremental compilation with CMake through a toolchain upgrade. Here is the problematic scenario :
Branch main uses g++-9 (using CMAKE_CXX_COMPILER=g++-9)
A new branch uses g++-10 (using CMAKE_CXX_COMPILER=g++-10)
Commits are happening on both branches
Incremental builds on one branch work fine
Switching to the other branch and explicitly invoking CMake fails
My question is the following : I'm looking for the proper way to make the invocation of CMake succeed and rebuild all the project from scratch when a toolchain change happens.
Here is a script that will make it quick and easy to reproduce the problem. This script requires Docker. It will create folders Sources and Build at the location where it is executed to avoid littering your filesystem. It then creates Dockerfiles to build docker containers with both g++ and cmake. It then creates a dummy Hello World C++ CMake project. Finally, it creates a folder for build artifacts and then executes the build with g++-9 and then g++-10. The second build fails because CMake generates an error.
#!/bin/bash
set -e
mkdir -p Sources
mkdir -p Build
# Creates a script that will be executed inside the docker container to perform builds
cat << EOF > Sources/Compile.sh
cd /Build \
&& cmake /Sources \
&& make \
&& ./IncrementalBuild
EOF
# Creates a Dockerfile that will be used to have both gcc-9 and cmake
cat << EOF > Sources/Dockerfile-gcc9
FROM gcc:9
RUN apt-get update && apt-get install -y cmake
RUN ln -s /usr/local/bin/g++ /usr/local/bin/g++-9
ADD Compile.sh /Compile.sh
RUN chmod +x /Compile.sh
ENTRYPOINT /Compile.sh
EOF
# Creates a Dockerfile that will be used to have both gcc-10 and cmake
cat << EOF > Sources/Dockerfile-gcc10
FROM gcc:10
RUN apt-get update && apt-get install -y cmake
RUN ln -s /usr/local/bin/g++ /usr/local/bin/g++-10
ADD Compile.sh /Compile.sh
RUN chmod +x /Compile.sh
ENTRYPOINT /Compile.sh
EOF
# Creates a dummy C++ program that will be compiled
cat << EOF > Sources/main.cpp
#include <iostream>
int main()
{
std::cout << "Hello World!\n";
}
EOF
# Creates CMakeLists.txt that will be used to compile the dummy C++ program
cat << EOF > Sources/CMakeLists.txt
cmake_minimum_required(VERSION 3.9)
project(IncrementalBuild CXX)
add_executable(IncrementalBuild main.cpp)
set_target_properties(IncrementalBuild PROPERTIES CXX_STANDARD 17)
EOF
# Build the docker images with both Dockerfiles created earlier
docker build -t cmake-gcc:9 -f Sources/Dockerfile-gcc9 Sources
docker build -t cmake-gcc:10 -f Sources/Dockerfile-gcc10 Sources
# Run a build with g++-9
echo ""
echo "### Compiling with g++-9 and then running the result..."
docker run --rm --user $(id -u):$(id -g) -v $(pwd)/Sources:/Sources -v $(pwd)/Build:/Build -e CXX=g++-9 cmake-gcc:9
echo ""
# Run a build with g++-10
echo "### Compiling with g++-10 and then running the result..."
docker run --rm --user $(id -u):$(id -g) -v $(pwd)/Sources:/Sources -v $(pwd)/Build:/Build -e CXX=g++-10 cmake-gcc:10
echo ""
# Print success if we reach this point
echo "SUCCESS!"
I'm looking for the proper way to make the invocation of CMake succeed and rebuild all the project from scratch when a toolchain change happens.
The proper way is to use a fresh binary directory. Either remove the binary directory when changing and let it recreate or just use a separate different directory for each toolchain.
Use Build/gcc10 binary directory for gcc10 build and Build/gcc9 for gcc9 builds.
No need to cd Build and mkdir with nowadays cmake - use cmake -S. -BBuild. Also do not use make - prefer cmake --build Build to let you switch generator later.
"If you change the toolchain, you should start with a fresh build. There are too many things that assume the toolchain doesn’t change and while you may be able to find workarounds which appear to work, I recommend you always use a fresh build tree for a different toolchain. This same logic also applies if you update the existing toolchain in-place (e.g. you update to a newer version of GCC on Linux, a newer version of Xcode on macOS, etc.). CMake queries compiler capabilities and caches the results. If you change the toolchain in a way that CMake can’t catch, then you end up with stale cached capabilities being used for the new/updated toolchain. Please don’t do that." - Craig Scott
So essentially I don't think it's possible. You just need to blow away your build. The best thing you can do is alert users if CMake isn't doing it for you.
Perhaps reply on this also:
https://discourse.cmake.org/t/how-to-change-toolchain-without-breaking-developer-workflows/1166
Or start another discourse.
I am trying to convert an MSYS2 build from azure pipeline CI (which works) to github actions.
I am now stuck with cmake not being recognised as a command.
My current (simplified) yaml is the following:
- uses: msys2/setup-msys2#v2
with:
msystem: MSYS
install: >-
git
base-devel
mingw-w64-${{ matrix.arch }}-gcc
mingw-w64-${{ matrix.arch }}-cmake
update: true
- name: Move Checkout
run: |
Copy-Item -Path ".\temp" -Destination "C:\_" -Recurse
- name: CI-Build
shell: msys2 {0}
run: |
cd /C/_
mkdir build
cd build
cmake -DCMAKE_FIND_LIBRARY_SUFFIXES='.a' -DCMAKE_TOOLCHAIN_FILE=cmake/Toolchains/mingw64-x86_64.cmake -DCMAKE_BUILD_TYPE=Release -G 'MSYS Makefiles' ..
make VERBOSE=1 -j4
Why is cmake not a recognised command?!
The name of the package for CMake is wrong. Here is the list of packages in the MSYS repository that contain cmake in the name:
cmake
icmake
mingw-w64-cmake
mingw-w64-cmake-doc-qt
mingw-w64-extra-cmake-modules
mingw-w64-python-sphinxcontrib-moderncmakedomain
I wager you either want mingw-w64-cmake or cmake. No value of matrix.arch in mingw-w64-${{ matrix.arch }}-cmake will expand to one of the above names.
The answer is to change msystem from MSYS to MINGW64 this is because MSYS will target cygwin (which would require the cmake package as specified by Alex Reinking).
Specifying MINGW64 will allow use of the mingw-w64-x86_64-cmake package correctly.
I'm trying to speed up build process using ramdisk for build directory.
I've created ramdisk:
sudo mount -t tmpfs -o size=1024m tmpfs /mnt/ramdisk
On ramdisk I've created build dir:
mkdir -p /mnt/ramdisk/rust/hello3/build/
Then I've symlinked ramdisk build dir to project in which I want to use this directory:
cd /home/wakatana/rust/hello3
ln -s /mnt/ramdisk/rust/hello3/build/ build
After this I did classic combo for building project:
cd /home/wakatana/rust/hello3/build
cmake ..
make
But above command does not worked because relative path (cmake ..) is translated to /mnt/ramdisk/rust/hello3 and not to /home/wakatana/rust/hello3/ (I suspect that this is whole problem)
So instead of classic combo I did a little bit modified combo (when build dir is not symlinked this works):
cd /home/wakatana/rust/hello3/build
cmake /home/wakatana/rust/hello3
make
But this ends up with errors during make phase:
-- Configuring done
-- Generating done
-- Build files have been written to: /home/wakatana/rust/hello3/build
make[2]: *** No rule to make target '../src/lib.rs', needed by 'src/x86_64-unknown-linux-gnu/debug/libtest_lib.a'. Stop.
CMakeFiles/Makefile2:122: recipe for target 'src/CMakeFiles/test-lib_target.dir/all' failed
make[1]: *** [src/CMakeFiles/test-lib_target.dir/all] Error 2
Makefile:83: recipe for target 'all' failed
make: *** [all] Error 2
Is it possible to somehow tell to cmake/make to deal symlinks correctly?
Just save yourself the trouble of creating symlinks and work on the ram disc:
cmake -S /home/wakatana/rust/hello3 -B /mnt/ramdisk/rust/hello3/build/
cmake --build /mnt/ramdisk/rust/hello3/build/
You could create the symlink, and then work from parent dir:
ln -s /mnt/ramdisk/rust/hello3/build/
cd /home/wakatana/rust/hello3
cmake -S . -B build
cmake --build build
# or expand the symlink before cmake has to:
cmake -S . -B "$(readlink -f "./build")"
cmake --build "$(readlink -f "./build")"
The other way is to rebind your RAM disk into your project tree instead of symlinking:
$ cd /home/wakatana/rust/hello3
$ mkdir -p build
$ mount --bind /mnt/ramdisk/rust/hello3/build build
I have a Makefile generated by CMake. The following path to CMake executable is set in the Makefile:
CMAKE_COMMAND = /home/xyz/opt/cmake/cmake-3.1.1/bin/cmake
How can I integrate Fortify sourceanalyzer with it and run scans?
I had the same challenge but solved it by running it like this:
sourceanalyzer -b project_ID -clean
Go to your build directory and perform make clean or remove all contents including the Makefile
Run cmake by changing CC and CXX variables:
CC="sourceanalyzer -b project_ID gcc" CXX="sourceanalyzer -b project_ID g++" cmake ..
Run make and fortify should be translating files while compilers do their job.
Run sourceanalyzer -b project_ID -scan -f results.fpr
Hope it helps.
I was tasked with integrating our CMake build system with HP Fortify SCA and came across this Thread that gave some insights but lacked specifics as related to HP Fortify so I thought I would share my implementation.
I created a fortify_tools directory at the same level as the source directory. Inside the fortify_tools are a toolchain file and fortify_cc, fortify_cxx, and fortify_ar scripts that will be set as the cmake_compilers via the toolchain file.
fortify_cc
#!/bin/bash
sourceanalyzer -b <PROJECT_ID> gcc $#
fortify_cxx
#!/bin/bash
sourceanalyzer -b <PROJECT_ID> g++ $#
fortify_ar
#!/bin/bash
sourceanalyzer -b <PROJECT_ID> ar $#
NOTE: insert your project name in place of PROJECT_ID
Setting cmake to use the scripts is accomplished in a toolchain file.
fortify_linux_toolchain.cmake
INCLUDE (CMakeForceCompiler)
SET(CMAKE_SYSTEM_NAME Linux)
SET(CMAKE_SYSTEM_VERSION 1)
#specify the compilers
SET(CMAKE_C_COMPILER ${CMAKE_SOURCE_DIR}/fortify_tools/fortify_cc)
SET(CMAKE_CXX_COMPILER ${CMAKE_SOURCE_DIR}/fortify_tools/fortify_cxx)
SET(CMAKE_AR_COMPILER ${CMAKE_SOURCE_DIR}/fortify_tools/fortify_ar)
To generate makefiles using the toolchain file
ccmake -DCMAKE_TOOLCHAIN_FILE=../fortify_tools/foritfy_linux_toolchain.cmake ../
configure and generate your makefiles and build your project.
Once the project is built from within the build directory generate a fortify report by
sourceanalyzer -Xmx2400M -debug -verbose -b <PROJECT_ID> -scan -f <PROJECT_ID>.fpr
I understand the last step is outside of CMake but I am pretty confident a cmake_custom_command can be created to perform the scan step as a post build action.
Finally, this is just the linux implementation but the concept scales well to Windows by creating the necessary batch files and windows specific toolchain file
Fortify doesn't support CMake, I received confirmation from Fortify support team.
This answer is late, but might help someone. This is actually easy to fix - you simply need to run cmake inside sourceanalyzer as well. Make a simple build script that calls cmake and then make, and use sourceanalyzer on that instead. I am using fortify 4.21.
Our old Fortify script for building hand-created Makefiles used a build command that looked like this:
$SOURCEANALYZER $MEMORY $LAUNCHERSWITCHES -b $BUILDID make -f Makefile -j12
I was able to get it working for a project that had been converted to CMake by replacing the above line with this, inspired by a couple of the other answers here:
CC="$SOURCEANALYZER $MEMORY $LAUNCHERSWITCHES -b $BUILDID gcc" \
CXX="$SOURCEANALYZER $MEMORY $LAUNCHERSWITCHES -b $BUILDID g++" \
AR="$SOURCEANALYZER $MEMORY $LAUNCHERSWITCHES -b $BUILDID ar" \
cmake -G "Unix Makefiles" -DCMAKE_BUILD_TYPE=Debug ..
make -f Makefile -j12 VERBOSE=1
This is with cmake 2.8.12.2 on Linux.
Below is the script i use for my example project to generate HP Fortify report for Android JNI C/C++ Code.
#!/bin/sh
# Configure NDK version and CMake version
NDK_VERSION=21.0.6113669
CMAKE_VERSION=3.10.2
CMAKE_VERSION_PATH=$CMAKE_VERSION.4988404
PROJECTID="JNI_EXAMPLE"
REPORT_NAME=$PROJECTID"_$(date +'%Y%m%d_%H:%M:%S')"
WORKING_DIR="$( cd "$( dirname "${BASH_SOURCE[0]}" )" && pwd )"
BUILD_HOME=${WORKING_DIR}/../hpfortify_build
FPR="$BUILD_HOME/$REPORT_NAME.fpr"
# Following exports need to be configured according to host machine.
export ANDROID_SDK_HOME=/Library/Android/sdk
export ANDROID_CMAKE_HOME=$ANDROID_SDK_HOME/cmake/$CMAKE_VERSION_PATH/bin
export ANDROID_NDK_HOME=$ANDROID_SDK_HOME/ndk/$NDK_VERSION
# E.g. JniExample/app/hpfortify/build/CMakeFiles/3.10.2
export CMAKE_FILES_PATH=${BUILD_HOME}/CMakeFiles/$CMAKE_VERSION
export HPFORTIFY_HOME="/Applications/Fortify/Fortify_SCA_and_Apps_20.1.0/bin"
export PATH=$PATH:$ANDROID_SDK_HOME:$ANDROID_NDK_HOME:$ANDROID_CMAKE_HOME:$HPFORTIFY_HOME
echo "[========Start Android JNI C/C++ HP Fortify scanning========]"
echo "[========Build Dir: $BUILD_HOME========]"
echo "[========HP Fortify report path: $FPR========]"
function create_build_folder {
rm -rf $BUILD_HOME
mkdir $BUILD_HOME
}
# The standalone cmake build command can be found from below file.
# JniExample/app/.cxx/cmake/release/x86/build_command.txt
# This file is generated after running command
# `➜ JniExample git:(master) ✗ ./gradlew :app:externalNativeBuildRelease`
function configure_cmake_files {
cd $BUILD_HOME
$ANDROID_CMAKE_HOME/cmake -H$BUILD_HOME/. \
-DCMAKE_CXX_FLAGS=-std=c++11 -frtti -fexceptions \
-DCMAKE_FIND_ROOT_PATH=$BUILD_HOME/.cxx/cmake/release/prefab/x86/prefab \
-DCMAKE_BUILD_TYPE=Release \
-DCMAKE_TOOLCHAIN_FILE=$ANDROID_SDK_HOME/ndk/$NDK_VERSION/build/cmake/android.toolchain.cmake \
-DANDROID_ABI=x86 \
-DANDROID_NDK=$ANDROID_SDK_HOME/ndk/$NDK_VERSION \
-DANDROID_PLATFORM=android-16 \
-DCMAKE_ANDROID_ARCH_ABI=x86 \
-DCMAKE_ANDROID_NDK=$ANDROID_SDK_HOME/ndk/$NDK_VERSION \
-DCMAKE_EXPORT_COMPILE_COMMANDS=ON \
-DCMAKE_LIBRARY_OUTPUT_DIRECTORY=$BUILD_HOME/intermediates/cmake/release/obj/x86 \
-DCMAKE_MAKE_PROGRAM=$ANDROID_SDK_HOME/cmake/$CMAKE_VERSION_PATH/bin/ninja \
-DCMAKE_SYSTEM_NAME=Android \
-DCMAKE_SYSTEM_VERSION=16 \
-B$BUILD_HOME/.cxx/cmake/release/x86 \
-GNinja ..
}
function build {
cmake --build .
}
function cleanup {
rm -rf $BUILD_HOME/CMakeFiles/native-lib.dir
rm -rf $FPR
$HPFORTIFY_HOME/sourceanalyzer -clean
}
function replace_compiler_paths {
FORTIFY_TOOLS_PATH="$WORKING_DIR"
CLANG_PATH="$ANDROID_SDK_HOME/ndk/$NDK_VERSION/toolchains/llvm/prebuilt/darwin-x86_64/bin/clang"
CLANGXX_PATH="$ANDROID_SDK_HOME/ndk/$NDK_VERSION/toolchains/llvm/prebuilt/darwin-x86_64/bin/clang++"
HPFORTIFY_CCPATH="$FORTIFY_TOOLS_PATH/fortify_cc"
HPFORTIFY_CXXPATH="$FORTIFY_TOOLS_PATH/fortify_cxx"\"
sed -i '' 's+'$CLANG_PATH'+'$HPFORTIFY_CCPATH'+g' $CMAKE_FILES_PATH/CMakeCCompiler.cmake
sed -i '' 's+'$CLANG_PATH.*[^")"]'+'$HPFORTIFY_CXXPATH'+g' $CMAKE_FILES_PATH/CMakeCXXCompiler.cmake
}
function scan {
$HPFORTIFY_HOME/sourceanalyzer -b $PROJECTID -scan -f $FPR
# copy the file to $WORKING_DIR
cp $FPR $WORKING_DIR
}
create_build_folder
configure_cmake_files
echo "[========Compile C/C++ using normal compiler ========"]
build
echo "[========Replace the compiler with HP Fortify analyser wrapper compilers ========"]
replace_compiler_paths
echo "[========Clean up the build intermediates and the older build ID and fpr file ========"]
cleanup
echo "[========Recompile C/C++ using HP Fortify analyser wrapper compilers ========"]
build
echo "[========Scan the compiled files and generate final report ========"]
scan
echo "[========Change directory to original working dir ========"]
cd $WORKING_DIR
Need to configure below vars before using it. For my case, I use NDK 21 and CMake 3.10.2 and my project ID is "JNI_EXAMPLE"
# Configure NDK version and CMake version
NDK_VERSION=21.0.6113669
CMAKE_VERSION=3.10.2
CMAKE_VERSION_PATH=$CMAKE_VERSION.4988404
PROJECTID="JNI_EXAMPLE"
# Following exports need to be configured according to host machine.
export ANDROID_SDK_HOME=/Library/Android/sdk
export ANDROID_NDK_HOME=$ANDROID_SDK_HOME/ndk/$NDK_VERSION
export HPFORTIFY_HOME="/Applications/Fortify/Fortify_SCA_and_Apps_20.1.0/bin"
Here is a more detailed explanation: Using HP Fortify to Scan Android JNI C/C++ Code
On recent version of CMake one can use:
CMAKE_<LANG>_COMPILER_LAUNCHER='sourceanalyzer;-b;<PROJECT_ID>'
You can add other arguments (like -Xmx2G for instance), semicolon separated, as mentioned on cmake documentation
You need to check if you don't use the compiler launcher for another tool like ccache. We can probably use both with
CCACHE_PREFIX='.../sourceanalyzer -b ID'
Here is what I've used in CMake project:
project(myFortifiedProject LANGUAGES CXX)
set(CMAKE_CXX_COMPILER_LAUNCHER ${FORTIFY_TOOL} -b ${PROJECT_NAME})
So when running cmake (assuming sourceanalyzer is on the path):
cmake <other args> -DFORTIFY_TOOL=sourceanalyzer
So the normal build command works:
make myFortifiedProject
And you can finally collect results with:
sourceanalyzer -b myFortifiedProject -scan
Folks, Am trying to build the rabbitmq C library. Following the instructions closely from the site (https://github.com/alanxz/rabbitmq-c), and am running into the following issue:
if [ ! -d "build" ]; then
mkdir build
fi
cd build
cmake ..
cmake --build .
cmake -DCMAKE_INSTALL_PREFIX=/usr/local ..
cmake --build . --target install
rabbitmq-c-0.4.1/build/install" does not exist
Thanks for all your help!
correct build steps:
if [ ! -d "build" ]; then
mkdir build
fi
cd build
cmake ..
cmake --build .
cmake -DCMAKE_INSTALL_PREFIX=/usr/local ..
make
make install