Ubuntu Xenial, x86_64
Compiling 6.3.0 gcc
nativ gcc version is 5.4.0
Error Source:
g++ -c -g -O2 -DIN_GCC -DCROSS_DIRECTORY_STRUCTURE -fno-exceptions -fno-rtti -fasynchronous-unwind-tables -W -Wall -Wno-narrowing -Wwrite-strings -Wcast-qual -Wmissing-format-attribute -Woverloaded-virtual -pedantic -Wno-long-long -Wno-variadic-macros -Wno-overlength-strings -DHAVE_CONFIG_H -DGENERATOR_FILE -fno-PIE -I. -Ibuild -I../../gcc-6.3.0/gcc -I../../gcc-6.3.0/gcc/build -I../../gcc-6.3.0/gcc/../include -I../../gcc-6.3.0/gcc/../libcpp/include \
-o build/genmddeps.o ../../gcc-6.3.0/gcc/genmddeps.c
Error:
In file included from ./bconfig.h:3:0,
from ../../gcc-6.3.0/gcc/genmddeps.c:18:
./auto-host.h:2321:16: error: declaration does not declare anything [-fpermissive]
#define rlim_t long
^
In file included from ../../gcc-6.3.0/gcc/genmddeps.c:19:0:
../../gcc-6.3.0/gcc/system.h:488:23: error: conflicting declaration of C function ‘void* sbrk(int)’
extern void *sbrk (int);
^
I am trying to compile gcc from scratch. My config options are
../gcc-6.3.0/configure --target=arm-eabi --prefix=/local/development/arm-project/tools/ --enable-languages=c --with-newlib --without-headers
Can anyone please help with debugging?
Try (and edit accordingly):
tar xzf gcc-6.3.0.tar.gz
cd gcc-6.3.0
./contrib/download_prerequisites
cd ..
mkdir objdir
cd objdir
$PWD/../gcc-6.3.0/configure --prefix=$HOME/GCC-6.3.0 --enable-languages=c,c++
make
make install
Source: Installing GCC
I could resolve this issue, by installing gmp-devel, mpfr-devel and libmpc-devel. See GCC docs for details.
There are some additional requirements for GCC 9.1.0 such as DejaGnu 1.6.2 and GDB 8.3 (solved my libiberty make target error).
If you have more than one compiler installed, use the GCC compiler:
export CC=gcc
export CXX=g++
Related
I am trying to build cross compiler gcc for mips architecture. Build and host are x86_64 and target is mips. I also want -micromips built version of libgcc.
I am able to build and install binutils and bootstrap gcc as below:
TARGET=mips-sde-elf
cd binutils-2.31-build
../binutils-2.31/configure --prefix=$PREFIX --target=$TARGET --disable-nls
make -j24
make install
cd gcc-6.5.0-build
../gcc-6.5.0/configure --prefix=$PREFIX --target=$TARGET \
--enable-languages=c,c++ --disable-nls \
--with-newlib --disable-threads --disable-shared --disable-libssp \
--disable-libstdcxx-pch --disable-libgomp --with-system-zlib
make -j24 all-gcc
However, when trying to build libgcc I get below error.
make -j24 all-target-libgcc
/projects/k2072/users/nima/tools/sources/gcc/gcc-6.5.0/gcc-6.5.0-build-mips/./gcc/xgcc -B/projects/k2072/users/nima/tools/sources/gcc/gcc-6.5.0/gcc-6.5.0-build-mips/./gcc/ -B/projects/k2072/users/nima/jadarite/toolchain/gcc-6.5.0-mips/mips-sde-elf/bin/ -B/projects/k2072/users/nima/jadarite/toolchain/gcc-6.5.0-mips/mips-sde-elf/lib/ -isystem /projects/k2072/users/nima/jadarite/toolchain/gcc-6.5.0-mips/mips-sde-elf/include -isystem /projects/k2072/users/nima/jadarite/toolchain/gcc-6.5.0-mips/mips-sde-elf/sys-include -g -O2 -Os -mcode-readable=pcrel -mno-gpopt -mmicromips -O2 -g -O2 -Os -minterlink-mips16 -mcode-readable=pcrel -mno-gpopt -DIN_GCC -DCROSS_DIRECTORY_STRUCTURE -W -Wall -Wno-narrowing -Wwrite-strings -Wcast-qual -Wstrict-prototypes -Wmissing-prototypes -Wold-style-definition -isystem ./include -g -DIN_LIBGCC2 -fbuilding-libgcc -fno-stack-protector -Dinhibit_libc -I. -I. -I../../.././gcc -I../../../../gcc-6.5.0/libgcc -I../../../../gcc-6.5.0/libgcc/. -I../../../../gcc-6.5.0/libgcc/../gcc -I../../../../gcc-6.5.0/libgcc/../include -DHAVE_CC_TLS -o fixsfsi.o -MT fixsfsi.o -MD -MP -MF fixsfsi.dep -DFUNC=__fixsfsi -DOP_fixsi -DTYPE=sf -c ../../../../gcc-6.5.0/libgcc/config/hardfp.c -fvisibility=hidden -DHIDE_EXPORTS -Wno-missing-prototypes
/tmp/cc8vFnc3.s: Assembler messages:
/tmp/cc8vFnc3.s:29: Error: invalid operands `trunc.w.s $f0,$f12,$2'
If I remove -mmicromips from above compile command, compilation passes. I also can build libgcc with other compile options like -EL -msoft-float, etc. but not micromips.
find -name "libgcc.a"
./libgcc/libgcc.a
./el/libgcc/libgcc.a
./mips32/libgcc/libgcc.a
./mips32r2/libgcc/libgcc.a
./mips64/libgcc/libgcc.a
./mips64r2/libgcc/libgcc.a
./mips16/libgcc/libgcc.a
Greatly appreciate any help.
I am using MacOS Mojave version 10.14.16 and I am trying to install http://www.rpl2.net/index.php (a programming language using Reverse Polish Notation as found on some HP calculators like the 48GX), supposed to work on MacOSX (they say "MacOS X (Xcode 3.1.4 + gfortran 4.5)").
So I have downloaded the latest stable release, cd to the folder, and wanted to run
./configure
make
make install
But the ./configure is failing due to "configure: error: Can not find libgomp !". There is no libgomp package I can install with brew (I already installed gcc which comes with gfortran and thought it would provide libgomp but apparently not).
Could someone help a poor applied mathematician trying to install a software on its macbook?
Thanks
EDIT
in my config.log I have this, maybe it will help:
configure:7433: checking for vim
configure:7449: found /usr/bin/vim
configure:7461: result: yes
configure:7507: checking for main in -lm
configure:7526: gcc -o conftest -g -O2 -O2 -fno-strict-overflow conftest.c -lm >&5
configure:7526: $? = 0
configure:7535: result: yes
configure:7548: checking for pthread_mutex_init in -lc
configure:7573: gcc -o conftest -g -O2 -O2 -fno-strict-overflow conftest.c -lc -lm >&5
configure:7573: $? = 0
configure:7582: result: yes
configure:7687: checking for omp_get_num_procs in -lgomp
configure:7712: gcc -o conftest -g -O2 -O2 -fno-strict-overflow conftest.c -lgomp -lc -lm >&5
ld: library not found for -lgomp
clang: error: linker command failed with exit code 1 (use -v to see invocation)
configure:7712: $? = 1
You don't say how you installed GCC. If you did it using homebrew, and
brew install gcc
you should be able to see:
/usr/local/Cellar/gcc/10.2.0/lib/gcc/10/libgomp.dylib
which would mean you need to compile with:
gcc ... -L /usr/local/Cellar/gcc/10.2.0/lib/gcc/10 -l gomp ...
Failing that, you can always search for libgomp with:
find /usr -name "lib*gomp*lib"
You can test which gcc you are using with:
type gcc
Okay, so, I switched from ubuntu 12.04 64-bit to 32-bit and installed build-essential.
I then compiled and installed GMP-5.0.5, MPFR-3.1.1, MPC-1.0, ISL-0.10 and CLOOG-0.17.0. I checked out a copy of the main gcc trunk and attempted to build it with the following configure line (from a separate directory):
../svnsrc/configure --prefix=/usr/GCC/svn --enable-__cxa_atexit --with-plugin-ld=/usr/bin/ld.gold --enable-threads=posix --enable-werror --enable-build-with-cxx --with-gmp=/usr/GCC/prereq/svn --with-mpfr=/usr/GCC/prereq/svn --with-mpc=/usr/GCC/prereq/svn --with-isl=/usr/GCC/prereq/svn --with-cloog=/usr/GCC/prereq/svn --enable-languages=c,c++
Configure ran fine and so I ran make && make check. This ran fine for a while, but then it failed with the following error:
/home/matt/GCC/svnbuild/./gcc/xgcc -B/home/matt/GCC/svnbuild/./gcc/ -B/usr/GCC/svn/i686-pc-linux-gnu/bin/ -B/usr/GCC/svn/i686-pc-linux-gnu/lib/ -isystem /usr/GCC/svn/i686-pc-linux-gnu/include -isystem /usr/GCC/svn/i686-pc-linux-gnu/sys-include -g -O2 -O2 -g -O2 -DIN_GCC -W -Wall -Wwrite-strings -Wcast-qual -Wstrict-prototypes -Wmissing-prototypes -Wold-style-definition -isystem ./include -fpic -mlong-double-80 -g -DIN_LIBGCC2 -fbuilding-libgcc -fno-stack-protector -fpic -mlong-double-80 -I. -I. -I../.././gcc -I../../../svnsrc/libgcc -I../../../svnsrc/libgcc/. -I../../../svnsrc/libgcc/../gcc -I../../../svnsrc/libgcc/../include -I../../../svnsrc/libgcc/config/libbid -DENABLE_DECIMAL_BID_FORMAT -DHAVE_CC_TLS -DUSE_TLS -o _muldi3.o -MT _muldi3.o -MD -MP -MF _muldi3.dep -DL_muldi3 -c ../../../svnsrc/libgcc/libgcc2.c -fvisibility=hidden -DHIDE_EXPORTS
In file included from /usr/include/stdio.h:28:0,
from ../../../svnsrc/libgcc/../gcc/tsystem.h:88,
from ../../../svnsrc/libgcc/libgcc2.c:29:
/usr/include/features.h:324:26: fatal error: bits/predefs.h: No such file or directory
#include <bits/predefs.h>
^
compilation terminated.
make[3]: *** [_muldi3.o] Error 1
make[3]: Leaving directory `/home/matt/GCC/svnbuild/i686-pc-linux-gnu/libgcc
I looked around, but everything I seemed to find was that this error was caused on x86_64, by not installing gcc-multilib, because Ubuntu and Debian use the mutiarch system, separating the libraries. Okay, fine...but I'm using i686, so why would I need the 64-bit libraries? Any help would be appreciated. Thanks.
Try doing a
sudo apt-get install gcc-multilib
I don't think that installing a 32-bit system changes the architecture of your computer, because your computer will always be a 64-bit machine. Installing the 64-bit version of Ubuntu should only give better multi-core performance. Since your computer is still a 64-bit computer, you probably need a C compiler that will compile on 64-bit machines. Hence the gcc-multilib. I think.
the gcc-multilib trick didn't work out for me. Instead, setting this in the build environment did the trick:
C_INCLUDE_PATH=/usr/include/$(gcc -print-multiarch)
I am having trouble installing pyipopt on ubuntu 12.04. During linking, I receive the error:
/usr/bin/ld: cannot find -lcoinhsl
Even though I know that this library is installed and the .so and .la files are available in /home/mostafa/MyBuilds/CoinIpopt/build/lib/
does anyone have a solution for this?
below is the complete return of running setup.py build:
root#ubuntu:~/MyBuilds/pyipopt# sudo python setup.py build
running build
running build_ext
building 'pyipopt' extension
gcc -pthread -fno-strict-aliasing -DNDEBUG -g -fwrapv -O2 -Wall -Wstrict-prototypes -fPIC -I/usr/lib/python2.7/dist-packages/numpy/core/include -I/home/mostafa/MyBuilds/CoinIpopt/build/include/coin/ -I/usr/include/python2.7 -c src/callback.c -o build/temp.linux-x86_64-2.7/src/callback.o
gcc -pthread -fno-strict-aliasing -DNDEBUG -g -fwrapv -O2 -Wall -Wstrict-prototypes -fPIC -I/usr/lib/python2.7/dist-packages/numpy/core/include -I/home/mostafa/MyBuilds/CoinIpopt/build/include/coin/ -I/usr/include/python2.7 -c src/pyipopt.c -o build/temp.linux-x86_64-2.7/src/pyipopt.o
src/pyipopt.c: In function ‘set_intermediate_callback’:
src/pyipopt.c:452:15: warning: variable ‘myowndata’ set but not used [-Wunused-but-set-variable]
gcc -pthread -shared -Wl,-O1 -Wl,-Bsymbolic-functions -Wl,-Bsymbolic-functions -Wl,-z,relrobuild/temp.linux-x86_64-2.7/src/callback.o build/temp.linux-x86_64-2.7/src/pyipopt.o -L/home/mostafa/MyBuilds/CoinIpopt/build/lib/ -lipopt -lblas -lcoinhsl -lcoinmetis -llapack -ldl -lm -o build/lib.linux-x86_64-2.7/pyipopt.so -Wl,--rpath -Wl,/home/mostafa/MyBuilds/CoinIpopt/build/lib/
/usr/bin/ld: cannot find -lcoinhsl
collect2: ld returned 1 exit status
error: command 'gcc' failed with exit status 1
-Ldir
Add directory dir to the list of directories to be searched for -l.
You may install your coinhsl library in one of your standard libraries directories and run 'ldconfig` before doing your ppyipopt install
I had a similar problem with another library and the reason why it didn't found it, was that I didn't run the make install (after running ./configure and make) for that library. The make install may require root privileges (in this case use: sudo make install). After running the make install you should have the so files in the correct folder, i.e. here /usr/local/lib and not in the folder mentioned by you.
you can add the Path to coinhsl lib to LD_LIBRARY_PATH variable. May be that will help.
export LD_LIBRARY_PATH=/xx/yy/zz:$LD_LIBRARY_PATH
where /xx/yy/zz represent the path to coinhsl lib.
As a requirement for chef-client, I am trying to install yajl-ruby on OpenSUSE 12.1. So far, it is returning the following message:
linux:~ # gem install yajl-ruby
Building native extensions. This could take a while...
ERROR: Error installing yajl-ruby:
ERROR: Failed to build gem native extension.
/usr/bin/ruby extconf.rb
creating Makefile
make
gcc -I. -I/usr/lib64/ruby/1.8/x86_64-linux -I/usr/lib64/ruby/1.8/x86_64-linux -I. -fPIC -fmessage-length=0 -O2 -Wall -D_FORTIFY_SOURCE=2 -fstack-protector -funwind-tables -fasynchronous-unwind-tables -g -g -fno-strict-aliasing -fPIC -Wall -funroll-loops -c yajl.c
gcc -I. -I/usr/lib64/ruby/1.8/x86_64-linux -I/usr/lib64/ruby/1.8/x86_64-linux -I. -fPIC -fmessage-length=0 -O2 -Wall -D_FORTIFY_SOURCE=2 -fstack-protector -funwind-tables -fasynchronous-unwind-tables -g -g -fno-strict-aliasing -fPIC -Wall -funroll-loops -c yajl_alloc.c
gcc -I. -I/usr/lib64/ruby/1.8/x86_64-linux -I/usr/lib64/ruby/1.8/x86_64-linux -I. -fPIC -fmessage-length=0 -O2 -Wall -D_FORTIFY_SOURCE=2 -fstack-protector -funwind-tables -fasynchronous-unwind-tables -g -g -fno-strict-aliasing -fPIC -Wall -funroll-loops -c yajl_buf.c
gcc -I. -I/usr/lib64/ruby/1.8/x86_64-linux -I/usr/lib64/ruby/1.8/x86_64-linux -I. -fPIC -fmessage-length=0 -O2 -Wall -D_FORTIFY_SOURCE=2 -fstack-protector -funwind-tables -fasynchronous-unwind-tables -g -g -fno-strict-aliasing -fPIC -Wall -funroll-loops -c yajl_encode.c
yajl_encode.c: In function ‘hexToDigit’:
yajl_encode.c:201:1: internal compiler error: Aborted
Please submit a full bug report,
with preprocessed source if appropriate.
See <http://bugs.opensuse.org/> for instructions.
make: *** [yajl_encode.o] Error 1
Gem files will remain installed in /usr/lib64/ruby/gems/1.8/gems/yajl-ruby-1.1.0 for inspection.
Results logged to /usr/lib64/ruby/gems/1.8/gems/yajl-ruby-1.1.0/ext/yajl/gem_make.out
The appropriate packages are installed:
zypper install ruby ruby-devel ruby-ri ruby-rdoc ruby-shadow gcc gcc-c++ automake autoconf make curl dmidecode
It may be an issue with the compiler or there may be an issue specific on OpenSUSE. So far, I am not sure which path to take.
gcc clearly recommends you to send bug report to OpenSuse with full preprocessed source (-E option instead of "-c" and add a redirect to file). This may be because opensuse gcc might have some modifications. You should check instructions on bugs.opensuse.org and send a bug report to OpenSuse. If the bug is in basic gcc too, opensuse bugzilla people will forward it upstream or will ask you to do this
To avoid an "internal compiler error" without sending bugs you can try to change build options. Usually "internal compiler error" means that something goes wrong in complex process of optimizations, so you can just change this process (order of optimization stages and which are enabled). Easiest is to change level of optimization (e.g. replace -O2 with -O1 or -O0) or add something like -Osize.