CMake issueshttps://gitlab.kitware.com/cmake/cmake/-/issues2017-06-15T17:16:57-04:00https://gitlab.kitware.com/cmake/cmake/-/issues/16980Regex with `*` behaves unexpectedly2017-06-15T17:16:57-04:00shadow_walkerRegex with `*` behaves unexpectedlyWhen I check if regex "Darwin-12*" matches "Darwin-15.0.0", the result is True. I would expect the result to be False. Sorry, if I am missing something very obvious.
[CMakeLists.txt](/uploads/238cae0574eee289d4d2089fe879bc84/CMakeList...When I check if regex "Darwin-12*" matches "Darwin-15.0.0", the result is True. I would expect the result to be False. Sorry, if I am missing something very obvious.
[CMakeLists.txt](/uploads/238cae0574eee289d4d2089fe879bc84/CMakeLists.txt)
```
set(str Darwin-15.0.0)
if(str MATCHES "Darwin-12*")
message("${str} matches Darwin-12*")
else()
message("${str} DOES NOT match Darwin-12*")
endif()
```
```
$ cmake .
Darwin-15.0.0 matches Darwin-12*
-- Configuring done
-- Generating done
-- Build files have been written to: <REDACTED>
```
```
$ cmake --version
cmake version 3.6.3
CMake suite maintained and supported by Kitware (kitware.com/cmake).
```https://gitlab.kitware.com/cmake/cmake/-/issues/18580regex: How to escape all special characters in string?2023-08-28T05:35:22-04:00Andru1999regex: How to escape all special characters in string?Does cmake have the ability to escape all the quantifier using a special command? If not, then let's add it.Does cmake have the ability to escape all the quantifier using a special command? If not, then let's add it.https://gitlab.kitware.com/cmake/cmake/-/issues/7089regexp makes cmake segfault2016-06-10T12:16:02-04:00Kitware Robotregexp makes cmake segfaultThis issue was created automatically from an original [Mantis Issue](https://cmake.org/Bug/view.php?id=7089). Further discussion may take place here.This issue was created automatically from an original [Mantis Issue](https://cmake.org/Bug/view.php?id=7089). Further discussion may take place here.https://gitlab.kitware.com/cmake/cmake/-/issues/15942regexpr in FindMPI.cmake filtering MPI_COMPILE_FLAGS fails for Intel-flag "-f...2017-10-13T13:17:56-04:00Kitware Robotregexpr in FindMPI.cmake filtering MPI_COMPILE_FLAGS fails for Intel-flag "-fp-model"This issue was created automatically from an original [Mantis Issue](https://cmake.org/Bug/view.php?id=15942). Further discussion may take place here.This issue was created automatically from an original [Mantis Issue](https://cmake.org/Bug/view.php?id=15942). Further discussion may take place here.https://gitlab.kitware.com/cmake/cmake/-/issues/66Register files for make clean2018-03-31T13:07:52-04:00Kitware RobotRegister files for make cleanThis issue was created automatically from an original [Mantis Issue](https://cmake.org/Bug/view.php?id=66). Further discussion may take place here.This issue was created automatically from an original [Mantis Issue](https://cmake.org/Bug/view.php?id=66). Further discussion may take place here.https://gitlab.kitware.com/cmake/cmake/-/issues/6079Regression bug from 2.4.7 to 2.52016-06-10T12:09:41-04:00Kitware RobotRegression bug from 2.4.7 to 2.5This issue was created automatically from an original [Mantis Issue](https://cmake.org/Bug/view.php?id=6079). Further discussion may take place here.This issue was created automatically from an original [Mantis Issue](https://cmake.org/Bug/view.php?id=6079). Further discussion may take place here.https://gitlab.kitware.com/cmake/cmake/-/issues/15325Regression building x265 using CMake 3.1 (Segmentation fault)2018-12-07T09:57:43-05:00Kitware RobotRegression building x265 using CMake 3.1 (Segmentation fault)This issue was created automatically from an original [Mantis Issue](https://cmake.org/Bug/view.php?id=15325). Further discussion may take place here.This issue was created automatically from an original [Mantis Issue](https://cmake.org/Bug/view.php?id=15325). Further discussion may take place here.https://gitlab.kitware.com/cmake/cmake/-/issues/11734Regression from 2.8.3 : add_test in subdir incorrectly keeps cwd at top level...2016-06-10T12:45:55-04:00Kitware RobotRegression from 2.8.3 : add_test in subdir incorrectly keeps cwd at top level CMAKE_BINARY_DIRThis issue was created automatically from an original [Mantis Issue](https://cmake.org/Bug/view.php?id=11734). Further discussion may take place here.This issue was created automatically from an original [Mantis Issue](https://cmake.org/Bug/view.php?id=11734). Further discussion may take place here.https://gitlab.kitware.com/cmake/cmake/-/issues/25252Regression from 3.27.4 to 3.27.5?: lapack 3.11.0 fails to configure with Make...2023-11-27T08:36:56-05:00Heiko BeckerRegression from 3.27.4 to 3.27.5?: lapack 3.11.0 fails to configure with Makefile generatorWith 3.27.4 lapack configuring lapack 3.11.0 worked fine. After updating to cmake 3.27.5 this fails with "CMake Error: CMake can not determine linker language for target: run_test_zcomplexabs", full output below. I don't see anything obv...With 3.27.4 lapack configuring lapack 3.11.0 worked fine. After updating to cmake 3.27.5 this fails with "CMake Error: CMake can not determine linker language for target: run_test_zcomplexabs", full output below. I don't see anything obvious why that would fail from a quick look, at least there's project(... Fortran) and the files have ".f" as extension.
Curiously if I use ninja as generator it works fine again.
```
cmake -DCMAKE_BUILD_TYPE:STRING=None -DCMAKE_C_FLAGS:STRING=-march=native -O2 -pipe -D_FORTIFY_SOURCE=2 -DCMAKE_CXX_FLAGS:STRING=-march=native -O2 -pipe -D_FORTIFY_SOURCE=2 -DCMAKE_AR:PATH=x86_64-pc-linux-gnu-ar -DCMAKE_RANLIB:PATH=x86_64-pc-linux-gnu-ranlib -DCMAKE_NM:PATH=x86_64-pc-linux-gnu-nm -DCMAKE_C_COMPILER:PATH=x86_64-pc-linux-gnu-cc -DCMAKE_CXX_COMPILER:PATH=x86_64-pc-linux-gnu-c++ -DCMAKE_INSTALL_PREFIX:PATH=/usr/x86_64-pc-linux-gnu -DCMAKE_FIND_ROOT_PATH:PATH=/usr/x86_64-pc-linux-gnu -DCMAKE_FIND_ROOT_PATH_MODE_PROGRAM:STRING=NEVER -DCMAKE_SYSTEM_PREFIX_PATH:PATH=/usr/x86_64-pc-linux-gnu -DCMAKE_INSTALL_LIBDIR:STRING=lib -DCMAKE_INSTALL_DATAROOTDIR:PATH=/usr/share/ -DBLAS++:BOOL=FALSE -DBLAS_LIBRARIES:STRING=blas -DBUILD_COMPLEX:BOOL=TRUE -DBUILD_COMPLEX16:BOOL=TRUE -DBUILD_DEPRECATED:BOOL=FALSE -DBUILD_DOUBLE:BOOL=TRUE -DBUILD_SHARED_LIBS:BOOL=TRUE -DBUILD_SINGLE:BOOL=TRUE -DCMAKE_Fortran_COMPILER=x86_64-pc-linux-gnu-gfortran -DLAPACK++:BOOL=FALSE -DLAPACKE:BOOL=TRUE -DBUILD_TESTING:BOOL=FALSE /var/tmp/paludis/build/sci-libs-lapack-3.11.0/work/lapack-3.11.0
CMake Deprecation Warning at CMakeLists.txt:1 (cmake_minimum_required):
Compatibility with CMake < 3.5 will be removed from a future version of
CMake.
Update the VERSION argument <min> value or use a ...<max> suffix to tell
CMake that the project does not need compatibility with older versions.
-- The Fortran compiler identification is GNU 13.2.0
-- The C compiler identification is GNU 13.2.0
-- Detecting Fortran compiler ABI info
-- Detecting Fortran compiler ABI info - done
-- Check for working Fortran compiler: /usr/bin/x86_64-pc-linux-gnu-gfortran - skipped
-- Detecting C compiler ABI info
-- Detecting C compiler ABI info - done
-- Check for working C compiler: /usr/bin/x86_64-pc-linux-gnu-cc - skipped
-- Detecting C compile features
-- Detecting C compile features - done
-- Checking if build type is 'Coverage'
-- Checking if build type is 'Coverage': 0
-- Performing Test _frecursiveFlag
-- Performing Test _frecursiveFlag - Success
-- Build tests: FALSE
-- Reducing RELEASE optimization level to O2
CMake Deprecation Warning at /var/tmp/paludis/build/sci-libs-lapack-3.11.0/work/lapack-3.11.0/INSTALL/CMakeLists.txt:1 (cmake_minimum_required):
Compatibility with CMake < 3.5 will be removed from a future version of
CMake.
Update the VERSION argument <min> value or use a ...<max> suffix to tell
CMake that the project does not need compatibility with older versions.
-- Looking for Fortran NONE - found
CMake Deprecation Warning at /var/tmp/paludis/build/sci-libs-lapack-3.11.0/work/lapack-3.11.0/INSTALL/CMakeLists.txt:1 (cmake_minimum_required):
Compatibility with CMake < 3.5 will be removed from a future version of
CMake.
Update the VERSION argument <min> value or use a ...<max> suffix to tell
CMake that the project does not need compatibility with older versions.
-- Looking for Fortran INT_CPU_TIME - found
CMake Deprecation Warning at /var/tmp/paludis/build/sci-libs-lapack-3.11.0/work/lapack-3.11.0/INSTALL/CMakeLists.txt:1 (cmake_minimum_required):
Compatibility with CMake < 3.5 will be removed from a future version of
CMake.
Update the VERSION argument <min> value or use a ...<max> suffix to tell
CMake that the project does not need compatibility with older versions.
-- Looking for Fortran EXT_ETIME - not found
CMake Deprecation Warning at /var/tmp/paludis/build/sci-libs-lapack-3.11.0/work/lapack-3.11.0/INSTALL/CMakeLists.txt:1 (cmake_minimum_required):
Compatibility with CMake < 3.5 will be removed from a future version of
CMake.
Update the VERSION argument <min> value or use a ...<max> suffix to tell
CMake that the project does not need compatibility with older versions.
-- Looking for Fortran EXT_ETIME_ - not found
CMake Deprecation Warning at /var/tmp/paludis/build/sci-libs-lapack-3.11.0/work/lapack-3.11.0/INSTALL/CMakeLists.txt:1 (cmake_minimum_required):
Compatibility with CMake < 3.5 will be removed from a future version of
CMake.
Update the VERSION argument <min> value or use a ...<max> suffix to tell
CMake that the project does not need compatibility with older versions.
-- Looking for Fortran INT_ETIME - found
-- --> Will use second_INT_ETIME.f and dsecnd_INT_ETIME.f as timing function.
-- Build deprecated routines: FALSE
-- Build single precision real: TRUE
-- Build double precision real: TRUE
-- Build single precision complex: TRUE
-- Build double precision complex: TRUE
-- Looking for Fortran dgemm
-- Looking for Fortran dgemm - found
-- --> BLAS supplied by user is WORKING, will use blas.
-- Using supplied NETLIB LAPACK implementation
-- Detecting Fortran/C Interface
-- Detecting Fortran/C Interface - Found GLOBAL and MODULE mangling
-- Verifying Fortran/C Compiler Compatibility
-- Verifying Fortran/C Compiler Compatibility - Success
-- LAPACKE enabled
-- Build LAPACKE single precision real: ON
-- Build LAPACKE double precision real: ON
-- Build LAPACKE single precision complex: ON
-- Build LAPACKE double precision complex: ON
-- Build html documentation: OFF
-- Build man documentation: OFF
-- Configuring done (2.7s)
CMake Error: CMake can not determine linker language for target: run_test_zcomplexabs
CMake Error: CMake can not determine linker language for target: run_test_zcomplexdiv
CMake Error: CMake can not determine linker language for target: run_test_zcomplexmult
CMake Error: CMake can not determine linker language for target: run_test_zminMax
-- Generating done (0.5s)
CMake Warning:
Manually-specified variables were not used by the project:
CMAKE_CXX_COMPILER
CMAKE_FIND_ROOT_PATH
CMake Generate step failed. Build files cannot be regenerated correctly.
```3.27.6Ben BoeckelBen Boeckelhttps://gitlab.kitware.com/cmake/cmake/-/issues/12307regression in 2.8.5 rc2: UseSWIG.cmake broken2018-12-07T09:59:39-05:00Kitware Robotregression in 2.8.5 rc2: UseSWIG.cmake brokenThis issue was created automatically from an original [Mantis Issue](https://cmake.org/Bug/view.php?id=12307). Further discussion may take place here.This issue was created automatically from an original [Mantis Issue](https://cmake.org/Bug/view.php?id=12307). Further discussion may take place here.https://gitlab.kitware.com/cmake/cmake/-/issues/19251Regression in 3.14.1 and onwards on macOS with Homebrew clang for `-nostdinc+...2019-12-17T07:01:10-05:00Nikolaus DemmelRegression in 3.14.1 and onwards on macOS with Homebrew clang for `-nostdinc++` with `include_directories("/usr/local/opt/llvm/include/c++/v1")`I noticed a regression on macOS with brewed clang in version 3.14.1 and later. It was working fine in 3.14.0. It is also working as expected with Apple's clang.
Tested macOS versions that are affected are 10.12 and 10.13, with clang 8.0...I noticed a regression on macOS with brewed clang in version 3.14.1 and later. It was working fine in 3.14.0. It is also working as expected with Apple's clang.
Tested macOS versions that are affected are 10.12 and 10.13, with clang 8.0.0 from Homebrew and cmake also from homebrew.
It seems if you use `include_directories("/usr/local/opt/llvm/include/c++/v1")`, which may be relevant in combination with `-nostdinc++`, this include path is not actually added. As mentioned, this used to work fine on 3.14.0 and earlier, and continues to work fine with latest cmake with Apple clang.
The [release notes](https://blog.kitware.com/cmake-3-14-1-available-for-download/) hint at some changed behaviour of `include_directories`, not sure if it is related.
To reproduce:
```cmake
# CMakeLists.txt
cmake_minimum_required(VERSION 3.14)
set(CMAKE_CXX_COMPILER "/usr/local/opt/llvm/bin/clang++")
project(Foo CXX)
include_directories("/usr/local/opt/llvm/include/c++/v1")
add_compile_options("-nostdinc++")
add_executable(foobar foobar.cpp)
```
```cpp
// foobar.cpp
#include <iostream>
int main() {
std::cout << "Foobar";
return 0;
}
```
With cmake 3.14.3:
```bash
$ cmake ..
-- The CXX compiler identification is Clang 8.0.0
-- Check for working CXX compiler: /usr/local/opt/llvm/bin/clang++
-- Check for working CXX compiler: /usr/local/opt/llvm/bin/clang++ -- works
-- Detecting CXX compiler ABI info
-- Detecting CXX compiler ABI info - done
-- Detecting CXX compile features
-- Detecting CXX compile features - done
-- Configuring done
-- Generating done
-- Build files have been written to: /Users/demmeln/work/slam/dba/cmake_test/build
$ VERBOSE=1 make
/usr/local/Cellar/cmake/3.14.3/bin/cmake -S/Users/demmeln/work/slam/dba/cmake_test -B/Users/demmeln/work/slam/dba/cmake_test/build --check-build-system CMakeFiles/Makefile.cmake 0
/usr/local/Cellar/cmake/3.14.3/bin/cmake -E cmake_progress_start /Users/demmeln/work/slam/dba/cmake_test/build/CMakeFiles /Users/demmeln/work/slam/dba/cmake_test/build/CMakeFiles/progress.marks
/Applications/Xcode.app/Contents/Developer/usr/bin/make -f CMakeFiles/Makefile2 all
/Applications/Xcode.app/Contents/Developer/usr/bin/make -f CMakeFiles/foobar.dir/build.make CMakeFiles/foobar.dir/depend
cd /Users/demmeln/work/slam/dba/cmake_test/build && /usr/local/Cellar/cmake/3.14.3/bin/cmake -E cmake_depends "Unix Makefiles" /Users/demmeln/work/slam/dba/cmake_test /Users/demmeln/work/slam/dba/cmake_test /Users/demmeln/work/slam/dba/cmake_test/build /Users/demmeln/work/slam/dba/cmake_test/build /Users/demmeln/work/slam/dba/cmake_test/build/CMakeFiles/foobar.dir/DependInfo.cmake --color=
Dependee "/Users/demmeln/work/slam/dba/cmake_test/build/CMakeFiles/foobar.dir/DependInfo.cmake" is newer than depender "/Users/demmeln/work/slam/dba/cmake_test/build/CMakeFiles/foobar.dir/depend.internal".
Dependee "/Users/demmeln/work/slam/dba/cmake_test/build/CMakeFiles/CMakeDirectoryInformation.cmake" is newer than depender "/Users/demmeln/work/slam/dba/cmake_test/build/CMakeFiles/foobar.dir/depend.internal".
Scanning dependencies of target foobar
/Applications/Xcode.app/Contents/Developer/usr/bin/make -f CMakeFiles/foobar.dir/build.make CMakeFiles/foobar.dir/build
[ 50%] Building CXX object CMakeFiles/foobar.dir/foobar.cpp.o
/usr/local/opt/llvm/bin/clang++ -nostdinc++ -o CMakeFiles/foobar.dir/foobar.cpp.o -c /Users/demmeln/work/slam/dba/cmake_test/foobar.cpp
/Users/demmeln/work/slam/dba/cmake_test/foobar.cpp:1:10: fatal error: 'iostream' file not found
#include <iostream>
^~~~~~~~~~
1 error generated.
make[2]: *** [CMakeFiles/foobar.dir/foobar.cpp.o] Error 1
make[1]: *** [CMakeFiles/foobar.dir/all] Error 2
make: *** [all] Error 2
```https://gitlab.kitware.com/cmake/cmake/-/issues/20900Regression in build directory inside symlinked home directory2020-07-03T07:25:41-04:00Holger SebertRegression in build directory inside symlinked home directoryCompiling CMake on FreeBSD-12.1-RELEASE works, but leads to broken ninja files. Here is the minimal CMakeLists.txt:
```text
cmake_minimum_required(VERSION 3.16)
project(NinjaOmitLocalOutputTest)
add_library(foo foo.cpp)
```
The command `...Compiling CMake on FreeBSD-12.1-RELEASE works, but leads to broken ninja files. Here is the minimal CMakeLists.txt:
```text
cmake_minimum_required(VERSION 3.16)
project(NinjaOmitLocalOutputTest)
add_library(foo foo.cpp)
```
The command `cmake -G ninja --debug-trycompile` produces the following output:
```
debug trycompile on
-- The C compiler identification is Clang 8.0.1
-- The CXX compiler identification is Clang 8.0.1
-- Detecting C compiler ABI info
CMake Error:
Running
'/home/hsebert/local/bin/ninja' '-C' '/usr/home/hsebert/Documents/cmake_test/build.broken/CMakeFiles/CMakeTmp' '-t' 'cleandead'
failed with:
ninja: error: build.ninja:120: unknown target 'all'
default all
^ near here
<snip>
```
Looking into the respective `build.ninja` file, I found this:
```
#############################################
# Folder: /home/hsebert/Documents/cmake_test/build.broken/CMakeFiles/CMakeTmp
build /home/hsebert/Documents/cmake_test/build.broken/CMakeFiles/CMakeTmp/all: phony /home/hsebert/Documents/cmake_test/build.broken/CMakeFiles/CMakeTmp/cmTC_91198
```
which should rather be (I suppose)
```
#############################################
# Folder: /home/hsebert/Documents/cmake_test/build.broken/CMakeFiles/CMakeTmp
build all: phony /home/hsebert/Documents/cmake_test/build.broken/CMakeFiles/CMakeTmp/cmTC_91198
```
(without the path before `all`).
Removing the path before `all` by manually editing the `build.ninja` file makes ninja work again.3.18.0Brad KingBrad Kinghttps://gitlab.kitware.com/cmake/cmake/-/issues/18817Regression in cmake-3.13.3 with missing source directory2019-10-12T10:09:37-04:00Tobias C. BernerRegression in cmake-3.13.3 with missing source directoryGood morning
Scalapack [1] fails to build on FreeBSD [2] with cmake-3.13.3, and works all things being the same with 3.13.2.
I suspect the change [3] which made source/binary directories required. A possible fix for scalapack is in i...Good morning
Scalapack [1] fails to build on FreeBSD [2] with cmake-3.13.3, and works all things being the same with 3.13.2.
I suspect the change [3] which made source/binary directories required. A possible fix for scalapack is in its `CMAKE/FortranMangling.cmake` to append the value it passes as `WORKING_DIRECTORY` as additional argument to the called cmake (line 21).
```
17 # Configure:
18 EXECUTE_PROCESS(COMMAND ${CMAKE_COMMAND}
19 "-DCMAKE_Fortran_COMPILER=${CMAKE_Fortran_COMPILER}"
20 "-DCMAKE_C_COMPILER=${CMAKE_C_COMPILER}"
21 "${PROJECT_SOURCE_DIR}/BLACS/INSTALL"
22 WORKING_DIRECTORY ${PROJECT_SOURCE_DIR}/BLACS/INSTALL/
23 RESULT_VARIABLE RESVAR OUTPUT_VARIABLE LOG1 ERROR_VARIABLE LOG1
24 )
```
mfg Tobias
[1] http://www.netlib.org/scalapack/ [v2.0.2]
[2] http://package23.nyi.freebsd.org/data/112i386-default-PR234959/2019-01-17_09h09m39s/logs/errors/scalapack-2.0.2_16.log
[3] https://gitlab.kitware.com/cmake/cmake/commit/27eb7c5bdb5bb8deefe1772675dc4819592bf0363.13.4Craig ScottCraig Scotthttps://gitlab.kitware.com/cmake/cmake/-/issues/14446Regression in cmMakefile::GetSourceFileWithOutput2018-12-07T09:58:17-05:00Kitware RobotRegression in cmMakefile::GetSourceFileWithOutputThis issue was created automatically from an original [Mantis Issue](https://cmake.org/Bug/view.php?id=14446). Further discussion may take place here.This issue was created automatically from an original [Mantis Issue](https://cmake.org/Bug/view.php?id=14446). Further discussion may take place here.https://gitlab.kitware.com/cmake/cmake/-/issues/11791Regression in next: FIND_PACKAGE(Qt4 COMPONENTS QtMain) gives errors2016-06-10T12:46:30-04:00Kitware RobotRegression in next: FIND_PACKAGE(Qt4 COMPONENTS QtMain) gives errorsThis issue was created automatically from an original [Mantis Issue](https://cmake.org/Bug/view.php?id=11791). Further discussion may take place here.This issue was created automatically from an original [Mantis Issue](https://cmake.org/Bug/view.php?id=11791). Further discussion may take place here.https://gitlab.kitware.com/cmake/cmake/-/issues/18033Regression in Windows: after 3.11.x GetRealPath doesn't respect DOS and subst...2022-06-22T12:31:25-04:00Francisco FacioniRegression in Windows: after 3.11.x GetRealPath doesn't respect DOS and subst mountsAfter 640709e7db054e90d5e6609eeb16a01af34b5a5a (from !1202) DOS mounted paths get translated to their real path instead of the mounted ones.
Example: if `C:\path` is mounted as `X:` then `cmSystemTools::GetRealPath("X:\file.txt")` will ...After 640709e7db054e90d5e6609eeb16a01af34b5a5a (from !1202) DOS mounted paths get translated to their real path instead of the mounted ones.
Example: if `C:\path` is mounted as `X:` then `cmSystemTools::GetRealPath("X:\file.txt")` will return `C:\path\file.txt` instead of `X:\file.txt` as it did previously.
Reverting 640709e7db054e90d5e6609eeb16a01af34b5a5a fixes the issue. Not sure what is the best solution for this, but we rely heavily on being able to mount paths as drives in Windows and this issue is blocking us to upgrade to 3.113.11.3Brad KingBrad Kinghttps://gitlab.kitware.com/cmake/cmake/-/issues/8112Regression of CMAKE_INCLUDE_PATH and CMAKE_LIBRARY_PATH with regard to PATH_S...2016-06-10T12:23:03-04:00Kitware RobotRegression of CMAKE_INCLUDE_PATH and CMAKE_LIBRARY_PATH with regard to PATH_SUFFIXESThis issue was created automatically from an original [Mantis Issue](https://cmake.org/Bug/view.php?id=8112). Further discussion may take place here.This issue was created automatically from an original [Mantis Issue](https://cmake.org/Bug/view.php?id=8112). Further discussion may take place here.https://gitlab.kitware.com/cmake/cmake/-/issues/16718Regression when building ParaView with an external HDF52017-04-03T23:02:56-04:00Ben BoeckelRegression when building ParaView with an external HDF5Bisecting now.Bisecting now.3.8.0Brad KingBrad Kinghttps://gitlab.kitware.com/cmake/cmake/-/issues/20096Regression when setting CMAKE_CXX_FLAGS_RELEASE before project() since 3.112019-12-16T16:25:41-05:00DamienRegression when setting CMAKE_CXX_FLAGS_RELEASE before project() since 3.11Depending on the version of cmake, cmake does not generate `compile_commands.json` with the same flags.
With the following `CMakeLists.txt`:
```cmake
cmake_minimum_required(VERSION 3.9 FATAL_ERROR)
set(CMAKE_CXX_FLAGS_RELEASE "${CMAKE_C...Depending on the version of cmake, cmake does not generate `compile_commands.json` with the same flags.
With the following `CMakeLists.txt`:
```cmake
cmake_minimum_required(VERSION 3.9 FATAL_ERROR)
set(CMAKE_CXX_FLAGS_RELEASE "${CMAKE_CXX_FLAGS_RELEASE} -DFOO")
project(foo)
add_executable(foo foo.cpp)
```
invoking `cmake -DCMAKE_EXPORT_COMPILE_COMMANDS=ON -DCMAKE_BUILD_TYPE=Release ..` on cmake 3.9 or 3.10 generates `compile_commands.json` containing:
```json
[
{
"directory": "/home/foo/build-cmake-3.9.6",
"command": "/usr/bin/c++ -O3 -DNDEBUG -o CMakeFiles/foo.dir/foo.cpp.o -c /home/foo/foo.cpp",
"file": "/home/foo/foo.cpp"
}
]
```
With cmake >= 3.11, it generates the following file:
```json
[
{
"directory": "/home/foo/build-cmake-3.11.4",
"command": "/usr/bin/c++ -DFOO -o CMakeFiles/foo.dir/foo.cpp.o -c /home/foo/foo.cpp",
"file": "/home/foo/foo.cpp"
}
]
```
Even though we should not specify `CMAKE_CXX_RELEASE_FLAGS` directly, some old codebase do it that way.
Maybe cmake should emit a warning for theses cases?https://gitlab.kitware.com/cmake/cmake/-/issues/14191Regression.. framework paths are not added to compile args in XCode generator2018-12-07T09:58:28-05:00Kitware RobotRegression.. framework paths are not added to compile args in XCode generatorThis issue was created automatically from an original [Mantis Issue](https://cmake.org/Bug/view.php?id=14191). Further discussion may take place here.This issue was created automatically from an original [Mantis Issue](https://cmake.org/Bug/view.php?id=14191). Further discussion may take place here.