Using M1 Mac, I've followed the manual install procedures to install Repast HPC.
I'm trying to run ./install.sh rhpc and the following error persists:
configure: error: cannot not find the flags to link with Boost mpi
The suggested solution says to add the compiler wrappers with the following commands:
export PATH=$BASE_DIR/MPICH/bin:$PATH
MPI_COMPILER_INVOCATION=$BASE_DIR/MPICH/bin/mpicxx
which didn't work as the command mpicxx still couldn't be invoked.
So I tried installing mpich through homebrew, where mpicxx now works outside of the install. However, the error persists.
So I tried installing boost through homebrew and directing the install to the homebrew directory, but the error persists.
So I tried installing open-mpi through homebrew, and mpicxx still works, but the error persists.
Any suggestions? I'm running out of online resources to solve the issue.
I've just installed os Catalina and updated xcode to current version.
I've tried
xcode-select --install
brew update
in cmd.
The thing is that program compiles but an annoying error:
'iostream' file not found
How can I solve this?
This question already has answers here:
'wchar.h' file not found
(9 answers)
Closed 2 years ago.
Unsurprisingly the new update of OSX Mojave broke my llvm installation in Homebrew, these are the steps I tried:
xcode-select --install # Complained, so I installed commandLineTools from here https://developer.apple.com/download/more/
xcode-select -p /Library/Developer/CommandLineTools
xcode-select --install # Now says installed
sudo xcodebuild -license # Fails, as it says I only have CommandLineTools installed in /Library/Developer/CommandLineTools not xcode
# Try something else (all versions)
brew uninstall --force llvm
brew install llvm # yay v7 how exciting
# Only it still don't work
clang++ -std=c++17 foo.cpp -o f
In file included from foo.cpp:1:
In file included from /usr/local/Cellar/llvm/7.0.0/include/c++/v1/iostream:38:
In file included from /usr/local/Cellar/llvm/7.0.0/include/c++/v1/ios:215:
In file included from /usr/local/Cellar/llvm/7.0.0/include/c++/v1/iosfwd:90:
/usr/local/Cellar/llvm/7.0.0/include/c++/v1/wchar.h:119:15: fatal error: 'wchar.h' file not found
#include_next <wchar.h>
^~~~~~~~~
1 error generated.
Any ideas? The first half of the things I tried I found online but mostly for older updates of MacOS, maybe Mojave broke things in a new and unpredictable way.
Here's how I solved it (very dirty but has worked for a long time):
Disabled SIP on my Mac
Symlinked /Applications/Xcode.app/Contents/Developer/Platforms/MacOSX.platform/Developer/SDKs/MacOSX.sdk/usr/include to /usr/include as is expected in POSIX environments.
Re-enabled SIP.
Now I got to make a note to re-do this everytime a new OSX update comes out. This has not broken anything wrt Xcode, clang (from brew) etc so I'll keep it. Be advised though, this might break your stuff.
This is a well known issue of the latest macos. Try to set this environment variable, it must fix this issue.
export LIBCLANG_SYSTEM_INCLUDE_DIR="/Applications/Xcode.app/Contents/Developer/Platforms/MacOSX.platform/Developer/SDKs/MacOSX.sdk/usr/include"
I did something dumb but it did fix the problem... Not sure if it will work for your computer though.
Here is what I did:
I originally installed c++ using Homebrew on my MacBook on Mojave... so I had to manually install the Xcode CommandLineTools using the installer found at /Library/Developer/CommandLineTools/Packages/macOS_SDK_headers_for_macOS_10.14.pkg . Then I copied the "c++" folder contents from to folder /Library/Developer/CommandLineTools/usr/include/c++/v1 to the folder where I want the missing file to be, at /usr/local/Cellar/gcc/8.2.0/include/c++/8.2.0 , while replacing files with the same names.
The trick worked, since now the missing file "wchar.h" is right where it should be... I could have simlinked, in hindsight. Cheers!
I know about this question: macOS 'wchar.h' File Not Found
but it did not help me. I've tried reinstalling xcode, reinstalling command line tools, restarting the system. wchar.h file is in place, but the compiler does not manage to find it. What else could I try? Maybe it is sysroot related? Any way to fix that?
macbooks-MacBook-Pro:Rack euphorbium$ sudo make
c++ -Iinclude -Idep/include -Idep/lib/libzip/include -DVERSION=dev -MMD -O3 -march=core2 -ffast-math -g -Wall -DARCH_MAC -isysroot /Applications/Xcode.app/Contents/Developer/Platforms/MacOSX.platform/Developer/SDKs/MacOSX10.7.sdk -mmacosx-version-min=10.7 -DAPPLE -stdlib=libc++ -std=c++11 -stdlib=libc++ -c -o build/src/app.cpp.o src/app.cpp
clang: warning: no such sysroot directory: '/Applications/Xcode.app/Contents/Developer/Platforms/MacOSX.platform/Developer/SDKs/MacOSX10.7.sdk' [-Wmissing-sysroot]
In file included from src/app.cpp:1:
In file included from include/app.hpp:2:
In file included from /Library/Developer/CommandLineTools/usr/include/c++/v1/vector:265:
In file included from /Library/Developer/CommandLineTools/usr/include/c++/v1/iosfwd:90:
/Library/Developer/CommandLineTools/usr/include/c++/v1/wchar.h:119:15: fatal error:
'wchar.h' file not found
#include_next <wchar.h>
^~~~~~~~~
1 error generated.
make: *** [build/src/app.cpp.o] Error 1
since wchar.h exists both in the xcode app, and in the /Library/Developer/CommandLineTools I think that it looks for it in some completely unrelated folder.
I had the same problem after upgrading to macOS High Sierra. A project which was compiling before stopped working with this error. In my case, the project files had been generated by the CMake application. The solution was to run CMake, 'Delete Cache', and 'Configure' the project again. This corrected the sysroot directory reference in the make files to point to the new SDK and the error was resolved.
I've got the same problem. I was able to fix it by using a symbolic link.
This is what I made :
cd /Applications/Xcode.app/Contents/Developer/Platforms/MacOSX.platform/Developer/SDKs/
sudo ln -s MacOSX.sdk MacOSX10.07.sdk
Then enter your root password.
NB: you have to adapt the last line with required sdk.
For any one struggling with this on MacOSX Mojave this answer solves it. In short:
Go to the terminal and run open /Library/Developer/CommandLineTools/Packages/macOS_SDK_headers_for_macOS_10.14.pkg This will bring up the prompt to install Xcode tools.
After that's done got to the terminal and run sudo xcode-select -s /Library/Developer/CommandLineTools
My system is macOS Mojave, and I've fixed this problem by running this command line:
open /Library/Developer/CommandLineTools/Packages/macOS_SDK_headers_for_macOS_10.14.pkg
It was caused by incorrect sysroot, (set for an older mac os version) set in makefiles.
If you are using CLion, follow these steps :
Click on CMake at the bottom center of your IDE
Click on the Settings Gear Icon at the left
Click on Reset Cache and Reload Project
Took me 3 hours to find this solution.
FYI this project was running fine on Bug Sur 11.3 but failed to build after I upgraded to Big Sur 11.4
Set an environment variable
SDKROOT="/Applications/Xcode.app/Contents/Developer/Platforms/MacOSX.platform/Developer/SDKs/MacOSX10.14.sdk"
and reopen Terminal. This solves the header not found issue.
To see where the compiler finds headers, run with -v
clang -v test.c
For my case, this problem result from that MacOS changes the headers's path.
Check this note: MacOS release note.
So the solution is to make the compiler can find those headers.
My macOS version is Catalina 10.15, and XCode Version 12.4. For this version, it seems that /Library/Developer/CommandLineTools/Packages/macOS_SDK_headers_for_macOS_10.14.pkg is not provided anymore and thus the sulotions above do not work for me.
This method saved my night: Fixing C++ compilation bugs for the MacOS Catalina upgrade. Note that if you follow its steps and errors that some global symbols are not found arise, do not modify the LIBRARY_PATH as it says.
I did something dumb but it did fix the problem... Not sure if it will work for your computer though.
Here is what I did:
I originally installed c++ using Homebrew on my MacBook on Mojave... so I had to manually install the Xcode CommandLineTools using the installer found at /Library/Developer/CommandLineTools/Packages/macOS_SDK_headers_for_macOS_10.14.pkg . Then I copied the "c++" folder contents from to folder /Library/Developer/CommandLineTools/usr/include/c++/v1 to the folder where I want the missing file to be, at /usr/local/Cellar/gcc/8.2.0/include/c++/8.2.0 , while replacing files with the same names.
The trick worked, since now the missing file "wchar.h" is right where it should be... Your settings is probably not the same, however, I am afraid.
I am getting this error when i tried to compile my program
-bash: g++: command not found
Also I have only been getting this error after I formatted my Mac Book Pro
This error came after I did this
g++ functions.cpp
Please help me.
Thank You
Here is a cleanest solution than installing XCode just for Gcc.
STEP 1 > install this:
[HomeBrew Web Site] (http://brew.sh/)
All the informations concerning the installation and the utility of such a program are on the site itself.
It's a package manager, but for Mac. It's like apt-get or yaourt, if you ever used Ubuntu or Arch Linux.
STEP 2 > Go to the Terminal application and type :
$> homebrew search gcc
Then choose from the version you want, and type again in Terminal:
$> homebrew install [choosen version]
Gcc is now installed.