riscv-poky build ld missing or too old - riscv

I'm working on building poky-linux for the first time. I'll need to run it in spike. I'm running into an error during the build (see below).
I guess bitbake may be looking at the version of ld and throwing an error. Any ideas how to fix this?
| checking if riscv64-poky-linux-gcc --sysroot=/home/user/riscv-poky/build/tmp/work/riscv64-poky-linux/glibc-initial/2.26-r0/recipe-sysroot is sufficient to build libc... yes
| checking for riscv64-poky-linux-nm... riscv64-poky-linux-nm
| checking for python3... python3
| configure: error:
| *** These critical programs are missing or too old: GNU ld
| *** Check the INSTALL file for required versions.
| WARNING: exit code 1 from a shell command.
| ERROR: Function failed: do_configure (log file is located at /home/user/riscv-poky/build/tmp/work/riscv64-poky-linux/glibc-initial/2.26-r0/temp/log.do_configure.30937)
ERROR: Task (/home/user/riscv-poky/build/../meta/recipes-core/glibc/glibc-initial_2.26.bb:do_configure) failed with exit code '1'
NOTE: Tasks Summary: Attempted 201 tasks of which 18 didn't need to be rerun and 1 failed.
Summary: 1 task failed:
/home/user/riscv-poky/build/../meta/recipes-core/glibc/glibc-initial_2.26.bb:do_configure
Summary: There were 2 WARNING messages shown.
Summary: There was 1 ERROR message shown, returning a non-zero exit code.
In the build directory:
file tmp/work/x86_64-linux/binutils-cross-riscv64/2.29-r0/sysroot-destdir/home/user/riscv-poky/build/tmp/work/x86_64-linux/binutils-cross-riscv64/2.29-r0/recipe-sysroot-native/usr/bin/riscv64-poky-linux/riscv64-poky-linux-ld
tmp/work/x86_64-linux/binutils-cross-riscv64/2.29-r0/sysroot-destdir/home/user/riscv-poky/build/tmp/work/x86_64-linux/binutils-cross-riscv64/2.29-r0/recipe-sysroot-native/usr/bin/riscv64-poky-linux/
file riscv64-poky-linux-ld
riscv64-poky-linux-ld: ELF 64-bit LSB shared object, x86-64, version 1 (SYSV), dynamically linked, interpreter /home/user/riscv-poky/build/tmp/sysroots-uninative/x86_64-linux/lib/ld-linux-x86-64.so.2, for GNU/Linux 3.2.0, BuildID[sha1]=debebc536bc2af4dd35ec0deb333bd32eb85b098, stripped

I think the solution in this case is that ld of binutils is too new rather than too old. I was using binutils 2.30. I switched to 2.25.1 and got past this error but ran into others.

Related

Yum reports an error:CRITICAL:yum.cli:Config error: Error accessing file for config file:///etc/yum.conf

My environment: python3.8, gcc12.2, cmake4.3. I want to compile and install the glibc library. When I finish compiling and installing, my yum has an error
The following is the error content:
CRITICAL:yum.cli:Config error: Error accessing file for config file:///etc/yum.conf
But I obviously have /etc/yum.conf, but I still get an error
I hope to know why and get a solution, thank you
I downloaded and compiled glibc, and the installation was successful, but there was an error message after installation
collect2: error: ld returned 1 exit status
Execution of gcc failed!
The script has found some problems with your installation!
Please read the FAQ and the README file and check the following:
- Did you change the gcc specs file (necessary after upgrading from
Linux libc5)?
- Are there any symbolic links of the form libXXX.so to old libraries?
Links like libm.so -> libm.so.5 (where libm.so.5 is an old library) are wrong,
libm.so should point to the newly installed glibc file - and there should be
only one such link (check e.g. /lib and /usr/lib)
You should restart this script from your build directory after you've
fixed all problems!
Btw. the script doesn't work if you're installing GNU libc not as your
primary library!
make[1]: *** [Makefile:111: install] Error 1
make[1]: Leaving directory '/opt/glibc-2.29'
make: *** [Makefile:12:install] 错误 2
But I found on the net that this error doesn't affect anything

Linker error : undefined reference to _fbss

I'm sucessfully building pulp-riscv-gnu-toolchain from this this
But when I try to compile my simple source code with builded compiler,
error occured.
My command is riscv32-unknown-elf-g++ -o hello hello.cpp
Error message is /home/jskim/test_toolchains_lib/gcc/riscv32-unknown-elf/7.1.1/../../../../riscv32-unknown-elf/bin/ld : cannot open linker script file riscv.ld : No such file or directory
So I try with this command riscv32-unknown-elf-g++ -o main main.cpp -T /home/jskim/test_toolchains/riscv32-unknown-elf-ld/lib/ldscripts/elf32lriscv.x
But following error occured : /home/jskim/test_toolchains/lib/gcc/riscv32-unknown-elf/7.1.1/../../../../riscv32-unknown-elf/lib/crt0.o: In function '.L0': (.text+0x10):undefined reference to '_fbss' collect2: error: ld returned 1 exit status
I'm using CentOS 7.6 and I try build this pulp-gcc with gcc 4.8.5 and 7.3.1 but both shows me same error. Please help me.
I resolve the problem in this way.
Actually Here is the solution.
I'm not accurate but I think the problem is that the name is not matched.
crt0, which seems to be the helper of start riscv processor has this line la a0 _fbss just basically linker can recognize this if _fbss is replaced with _edata. So i edit the assembly file which is located at pulp-riscv-gnu-toolchain/riscv-newlib/libgloss/riscv/ And rebuild the gnu toolchain. And execute your compiler with -T option like riscv32-unknown-elf-g++ -o main main.cpp -T /home/jskim/test_toolchains/riscv32-unknown-elf-ld/lib/ldscripts/elf32lriscv.x
This is the easiest solution as I think.

Fail to build y86-64 simulator from sources

I am attempting to compile a simulator for Y86-64 code on Linux.I have already rewritten the makefile but it turned out like below.It said "undefined reference for 'matherr'".(Looks like it connects with gcc when linking)
(cd pipe; make all GUIMODE=-DHAS_GUI TKLIBS="-L/usr/lib/ -ltk8.5 -ltcl8.5" TKINC="-I/usr/include/tcl8.5 ")
make[1]: 进入目录“/home/gongchen/桌面/ICS/archlab-handout/sim/pipe”
# Building the pipe-std.hcl version of PIPE
../misc/hcl2c -n pipe-std.hcl < pipe-std.hcl > pipe-std.c
gcc -Wall -O2 -I/usr/include/tcl8.5 -I../misc -DHAS_GUI -o psim psim.c pipe-std.c \
../misc/isa.c -L/usr/lib/ -ltk8.5 -ltcl8.5 -lm
/tmp/cchKTZy7.o:(.data.rel+0x0):对‘matherr’未定义的引用
collect2: error: ld returned 1 exit status
Makefile:42: recipe for target 'psim' failed
make[1]: *** [psim] Error 1
make[1]: 离开目录“/home/gongchen/桌面/ICS/archlab-handout/sim/pipe”
Makefile:28: recipe for target 'all' failed
make: *** [all] Error 2
gcc -Wall -O2 -I/usr/include/tcl8.5 -I../misc -DHAS_GUI -o psim psim.c pipe-std.c \
../misc/isa.c -L/usr/lib/ -ltk8.5 -ltcl8.5 -lm
/tmp/cchKTZy7.o:(.data.rel+0x0):对‘matherr’未定义的引用
You are linking and getting an undefined reference error to matherr.
It looks like matherr is part of SVID math library. According to the matherr(3) man page the symbol is no longer present in Glibc 2.27 or above.
DESCRIPTION
Note: the mechanism described in this page is no longer supported by
glibc. Before glibc 2.27, it had been marked as obsolete. Since
glibc 2.27, the mechanism has been removed altogether. New
applications should use the techniques described in math_error(7) and
fenv(3). This page documents the matherr() mechanism as an aid for
maintaining and porting older applications.
The math_error(7) man page says that you should do the following to check for errors:
set errno to zero
call feclearexcept(FE_ALL_EXCEPT);
After the math calculation completes you should check the following for non-zero value to indicate error:
errno
fetestexcept(FE_INVALID | FE_DIVBYZERO | FE_OVERFLOW | FE_UNDERFLOW);
Since you are a guy or gal trying to use the program (and not a maintainer) I suggest two courses of actions. The strategy is to use a distro where things just work, and punt to the Y86 maintainer to fix it.
First, use a different, older distro that provides Glibc 2.26 or early. Something like Debian 8 (Glibc 2.19) or Fedora 25 (Glibc 2.24) should do just fine.
Second, file a bug report against Y86 project. The Y86 maintainers need to fix the problem, not students trying to learn the class material.
My classmates have a way to solve this problem: comment the code related to matherr, like the code in the picture. And the GUI mode works. 好厉害!

tcl8.5.10 make fail error, Network Simulator2 v2.35 (Cygwin v1.7.32) in Windows 8

during installation of NS2 v2.35 (Cygwin v1.7.32) in Windows 8 i keep getting this error message:
Package gcc4 is not present on your system.
Please install it using Cygwin's setup.exe
before trying to install the ns-2 distribution.
I tried downloading the gcc4 package from the setup but it does not show that it has a gcc4 package. So, i downloaded it from internet but the file was empty!!! If i ignore the warning for gcc4 and continue with the installation then i got the following error.
/usr/lib/gcc/x86_64-pc-cygwin/4.8.1/../../../../x86_64-pc-cygwin/bin/ld: final l ink failed: Invalid operation
collect2: error: ld returned 1 exit status
Makefile:569: recipe for target `tclsh.exe' failed
make: *** [tclsh.exe] Error 1
tcl8.5.10 make failed! Exiting ...
For problems with Tcl/Tk see http://www.scriptics.com
Somebody please help me!!!
Latest gcc ,i.e., 4.8 not compatible, Go for older versions of gcc like 4.7

Gcc compilation "cannot compute suffix of object files: cannot compile"

I'm actually reading the LFS book (version 7.1) and I'm blocked at page 53. Trying to compile gcc, I tried the following command:
./configure --target=$LFS_TGT --prefix=$LFS/build/gcc-build --disable-nls\
--disable-shared --disable-multilib --disable-decimal-float --disable-threads\
--disable-libmudflap --disable-libssp --disable-libgomp --disable-libquadmath\
--disable-target-libiberty --disable-target-zlib\
--enable-languages=c\
--without-ppl --without-cloog\
--with-mpfr-include=$LFS/source/mpfr/src
--with-mpfr-lib=$LFS/source/mpfr/src/.libs\
--with-gmp-include=/mnt/LFS/source/gmp\
--with-gmp-lib=/mnt/LFS/source/gmp/.libs\
--with-mpc-include=/mnt/LFS/source/mpc/src\
--with-mpc-lib=/mnt/LFS/source/mpc/src/.libs
to run the configure script of gcc (of course I already compiled mpfr, mpc and gmp as well).
But once I launch:
make -j4
I get the following error:
checking for suffix of object files... configure: error: in `/mnt/LFS/source/gcc-4.6.2/x86_64-lfs-linux-gnu/libgcc':
configure: error: cannot compute suffix of object files: cannot compile
See `config.log' for more details.
make[1]: *** [configure-target-libgcc] Error 1
I tried to google for it and tried the solutions I found but nothing worked. Does anyone know why I get this error?
This issue is caused by dyanmic link library path issue when the test programs try to link against libmpc/libmpfr/libgmp.
Append below environment variable to allow ld link against the correct so file:
export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/path/to/mpc/lib/
Then try build gcc again.
"Building GCC is not trivial, but is not difficult if you follow the instructions carefully.
Many people rush into trying to build it without reading the installation docs properly and make one or more of these common mistakes:
do not run ./configure from gcc src dir (this is not supported) => you need to run configure from outside the gcc source directory
Note: if GCC links dynamically to the prerequisite libs (GMP/MPFR/MPC) then the shared libraries must be in the dynamic linker's path (LD_LIBRARY_PATH), both when building gcc and when using the installed compiler."
Simple example (without dynamic link to GMP/MPFR/MPC):
tar xzf gcc-4.8.0.tar.gz
cd gcc-4.8.0
./contrib/download_prerequisites
cd ..
mkdir objdir
cd objdir
$PWD/../gcc-4.8.0/configure --prefix=/opt/gcc-4.8.0
make
make install
Sources:
Advogato Doc -
GNU Doc
This error message can arise from a number of different reasons. The best way to figure out which one is to check the logfile '/home/manu/gcc/gcc/i686-pc-linux-gnu/libgcc/config.log' in the example below. Or in the original posters case '/mnt/LFS/source/gcc-4.6.2/x86_64-lfs-linux-gnu/libgcc' and look for the last error line.
Quoting GCC FAQ: http://gcc.gnu.org/wiki/FAQ#configure_suffix
Like any of the GNU projects, GCC is using the GNU autotools to
commonly configure the compilation for the specifics of the build
system. The configure script thereby uses small test programs -
usually called conftest.c - to test if certain functions and/or
features are available. If the compilation of such a test program
fails, you'll see an error message like:
checking for suffix of object files... configure: error: in
`/home/manu/gcc/gcc/i686-pc-linux-gnu/libgcc': configure: error:
cannot compute suffix of object files: cannot compile See `config.log'
for more details. make[2]: *** [configure-stage1-target-libgcc] Error
1 make[2]: Leaving directory `/home/manu/gcc/gcc'
This error message is quite misleading and frequently the problem has
nothing to do with the message. You have to check the file
'config.log' in the directory where the error occurred. In the example
above, you would have to check the 'config.log' file in the directory
'/home/manu/gcc/gcc/i686-pc-linux-gnu/libgcc'. There might be several
test programs that failed during the configuration, but some of these
failures are non-critical. Check for the last error entry in the file.
Common causes for this error message are:
Required libraries for the GCC build are missing, specifically MPFR,
GMP and MPC. If installed as shared libraries they must be in the
runtime linker's search path so they can be found. Please, follow the
instructions in the answer to Why does my ./configure and make fail?
The compiler crashed. For example, if there is an error such as
'conftest.c: internal compiler error:', this indicates a bug in the
compiler. If you are using an unmodified version of GCC, please follow
the procedure to report the bug.
This happens while creating a cross-compiler.
It's not referring to the top-level config.log, but $LFS_TGT/libgcc/config.log.
For me, in there, it says:
configure:3566: /foo/gcc/build/./gcc/xgcc -B/foo/gcc/build/./gcc/ -B/opt/gcc-cross-11.1.0/aarch64-none-elf/bin/ -B/opt/gcc-cross-11.1.0/aarch64-none-elf/lib/ -isystem /opt/gcc-cross-11.1.0/aarch64-none-elf/include -isystem /opt/gcc-cross-11.1.0/aarch64-none-elf/sys-include -o conftest -g -O2 conftest.c >&5
conftest.c:9:10: fatal error: stdio.h: No such file or directory
9 | #include <stdio.h>
| ^~~~~~~~~
compilation terminated.
configure:3569: $? = 1
configure:3782: checking for suffix of object files
configure:3804: /foo/gcc/build/./gcc/xgcc -B/foo/gcc/build/./gcc/ -B/opt/gcc-cross-11.1.0/aarch64-none-elf/bin/ -B/opt/gcc-cross-11.1.0/aarch64-none-elf/lib/ -isystem /opt/gcc-cross-11.1.0/aarch64-none-elf/include -isystem /opt/gcc-cross-11.1.0/aarch64-none-elf/sys-include -c -g -O2 conftest.c >&5
/foo/gcc/build/./gcc/as: line 106: exec: -E: invalid option
exec: usage: exec [-cl] [-a name] [command [arguments ...]] [redirection ...]
configure:3808: $? = 1
configure: failed program was:
| /* confdefs.h */
| #define PACKAGE_NAME "GNU C Runtime Library"
| #define PACKAGE_TARNAME "libgcc"
| #define PACKAGE_VERSION "1.0"
| #define PACKAGE_STRING "GNU C Runtime Library 1.0"
| #define PACKAGE_BUGREPORT ""
| #define PACKAGE_URL "http://www.gnu.org/software/libgcc/"
| /* end confdefs.h. */
|
| int
| main ()
| {
|
| ;
| return 0;
| }
configure:3822: error: in `/foo/gcc/build/aarch64-none-elf/libgcc':
configure:3824: error: cannot compute suffix of object files: cannot compile
See `config.log' for more details
Was your logfile similar?
It's my belief that the gcc compilation/installation process will take the various includes, run them through fixincludes to fix problems and tailor them to the target arch. The expected file is not present.
Referring to the Prerequisites page, I believe this section is relevant (emphasis mine):
C standard library and headers
In order to build GCC, the C standard library and headers must be present for all
target variants for which target libraries will be built (and not only the variant
of the host C++ compiler).
This affects the popular ‘x86_64-pc-linux-gnu’ platform (among other multilib
targets), for which 64-bit (‘x86_64’) and 32-bit (‘i386’) libc headers are usually
packaged separately. If you do a build of a native compiler on ‘x86_64-pc-linux-gnu’,
make sure you either have the 32-bit libc developer package properly installed (the
exact name of the package depends on your distro) or you must build GCC as a 64-bit
only compiler by configuring with the option --disable-multilib. Otherwise, you may
encounter an error such as ‘fatal error: gnu/stubs-32.h: No such file’
I believe that the solution is to obtain headers for the TARGET platform.
The problem is the assembler, "as" is missing. Build and install binutils for the target first.
In my case it was:
export LD_LIBRARY_PATH=/usr/mpc-082/lib:/usr/gmp501/lib:/usr/libelf0812/lib
Whereas each of the above library is the destination of what additional libraries you installed appended with /lib at the end.
Example: Lib-MPC:
1) Downloaded onto /usr/src/libmpc-src
2) cd into /usr/src/mpc-082
3) ../configure --prefix=/usr/mpc-082
4) make
5) make install
6) once installed, verify that you have the new libs installed under /usr/mpc-082-lib
7) add it to your LD_LIBRARY_PATH by doing:
export LD_LIBRARY_PATH=/usr/mpc-082/lib:$LD_LIBDADY_PATH
8) Repeat the same with your LibELF, and LibGMP
9) Go back and compile your GCC, should work after this.
Cheers
H

Resources