-
-
Notifications
You must be signed in to change notification settings - Fork 58
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Defect: Build failure on Fedora 32, using gcc 10.0.1 #704
Comments
@jerryd thanks for reporting this. Could you post the date that shows when you run |
@jerryd I just noticed that GCC 10.1 released last week so I'll test with the 10.1 release instead of testing with the GCC |
@jerryd arrrgh... scratch that. Now I see that it hasn't actually released. There is a release candidate. I'll build the release candidate if it's easy to figure out the appropriate git tag. Otherwise, I'll build the |
@jerryd my installation succeeded on Lubuntu 19.10 with the following commands:
Please try the above for me. Feel free to change the installation paths. If the above doesn't work for you, then the issue is most likely related to the issue is most likely Fedora-specific. Note that the last step above will install MPICH 3.2 if it doesn't find an MPICH 3.2 or later in your |
FYI, gcc No regression either on https://github.com/uhhpctools/caf-testsuite.
|
Thanks, @neok-m4700. Also, just FYI, the OpenCoarrays installer is able to build MPICH ( @jerryd I'll close this. If you're still encountering difficulties, feel free to reopen it. |
This bug is very real. Let me explain. If someone follows the instructions here: The result is identical to what I have decribed previously: gcc version is 9.3.1 So either the instructions should be changed or the associated cmake configuration files fixed. (suggest this issue be reopened)
|
@jerryd Hi, can you please post the build output with BTW, the options reported as
|
The function target_compile_options in combination with generator expressions does not handle strings with space separated compiler flags and inserts quotes. Also MPI_<lang>_COMPILE_FLAGS is deprecated since cmake 3.10 and should be replaced by MPI_<lang>_COMPILE_OPRIONS and MPI_<lang>_COMPILE_DEFINITIONS. Those variables are already lists of compiler flags and thus well suited to solve the bug. This fixes the build error described in issue sourceryinstitute#676 when using MPICH, which is the same error as described in issue sourceryinstitute#704. Signed-off-by: Elias Lettl <elias.lettl@physik.uni-augsburg.de>
The function target_compile_options in combination with generator expressions does not handle strings with space separated compiler flags and inserts quotes. Also MPI_<lang>_COMPILE_FLAGS is deprecated since cmake 3.10 and should be replaced by MPI_<lang>_COMPILE_OPRIONS and MPI_<lang>_COMPILE_DEFINITIONS. Those variables are already lists of compiler flags and thus well suited to solve the bug. This fixes the build error described in issue sourceryinstitute#676 when using MPICH, which is the same error as described in issue sourceryinstitute#704. Signed-off-by: Elias Lettl <elias.lettl@physik.uni-augsburg.de>
Confirmed the fix for #706 also fixed my issues here. Thanks much! |
System information including:
uname -a
: Linux quasar 5.6.8-300.fc32.x86_64 tests dis_transpose: test passed #1 SMP Wed Apr 29 19:01:34 UTC 2020 x86_64 x86_64 x86_64 GNU/LinuxTo help us debug your issue please explain:
What you were trying to do (and why)
Trying to Build OpenCoarrays to test it.
--- snip ---
-- Configuring done
-- Generating done
-- Build files have been written to: /home/jerry/dev/buildcaf/OpenCoarrays/build
[$ make
Scanning dependencies of target caf_mpi_static
[ 1%] Building C object src/mpi/CMakeFiles/caf_mpi_static.dir/mpi_caf.c.o
gcc: error: unrecognized command-line option ‘-fexceptions -fasynchronous-unwind-tables -fstack-clash-protection -fcf-protection -D_GLIBCXX_ASSERTIONS’
make[2]: *** [src/mpi/CMakeFiles/caf_mpi_static.dir/build.make:83: src/mpi/CMakeFiles/caf_mpi_static.dir/mpi_caf.c.o] Error 1
make[1]: *** [CMakeFiles/Makefile2:828: src/mpi/CMakeFiles/caf_mpi_static.dir/all] Error 2
make: *** [Makefile:161: all] Error 2
What happened (include command output, screenshots, logs, etc.)
After invoking cmake ....:
[jerry@quasar build]$ rm -rf ./*
[jerry@quasar build]$ export FC=gfortran
[jerry@quasar build]$ export CC=gcc
[jerry@quasar build]$ cmake .. -DCMAKE_INSTALL_PREFIX=/home/jerry/dev/usr
---snip ---
OpenCoarrays
""""""""""""
Build Type: Release
Version: 2.8.0-10-g964360c
-- Build from git repository detected
-- Running with CMake from: /usr/bin/cmake
-- Current source dir (for OpenCoarrays): /home/jerry/dev/buildcaf/OpenCoarrays
-- The C compiler identification is GNU 10.0.1
-- The Fortran compiler identification is GNU 10.0.1
-- Check for working C compiler: /usr/bin/gcc
-- Check for working C compiler: /usr/bin/gcc - works
-- Detecting C compiler ABI info
-- Detecting C compiler ABI info - done
-- Detecting C compile features
-- Detecting C compile features - done
-- Check for working Fortran compiler: /usr/bin/gfortran
-- Check for working Fortran compiler: /usr/bin/gfortran - works
-- Detecting Fortran compiler ABI info
-- Detecting Fortran compiler ABI info - done
-- Checking whether /usr/bin/gfortran supports Fortran 90
-- Checking whether /usr/bin/gfortran supports Fortran 90 - yes
-- Building OpenCoarrays version: 2.8.0-10-g964360c
-- Building for target architecture: x86_64
-- Performing Test Check_Simple_Coarray_Fortran_Source_Compiles
-- Performing Test Check_Simple_Coarray_Fortran_Source_Compiles - Failed
-- Found MPI_C: /usr/lib64/mpich/lib/libmpi.so (found version "3.1")
-- Found MPI_Fortran: /usr/lib64/mpich/lib/libmpifort.so (found version "3.1")
-- Found MPI: TRUE (found version "3.1")
-- MPI runtime and compile time environments appear to be consistent
-- Performing Test MPI_C_COMPILES
-- Performing Test MPI_C_COMPILES - Success
-- Performing Test MPI_Fortran_MODULE_COMPILES
-- Performing Test MPI_Fortran_MODULE_COMPILES - Success
-- Performing Test MPI_Fortran_INCLUDE_COMPILES
-- Performing Test MPI_Fortran_INCLUDE_COMPILES - Success
-- Looking for alloca.h
-- Looking for alloca.h - found
-- Looking for signal.h
-- Looking for signal.h - found
-- Looking for SIGKILL
-- Looking for SIGKILL - found
-- Looking for include files mpi.h, mpi-ext.h
-- Looking for include files mpi.h, mpi-ext.h - not found
-- Looking for MPIX_ERR_PROC_FAILED
-- Looking for MPIX_ERR_PROC_FAILED - found
-- Looking for MPIX_ERR_REVOKED
-- Looking for MPIX_ERR_REVOKED - found
-- Looking for MPIX_Comm_failure_ack
-- Looking for MPIX_Comm_failure_ack - found
-- Looking for MPIX_Comm_failure_get_acked
-- Looking for MPIX_Comm_failure_get_acked - found
-- Looking for MPIX_Comm_shrink
-- Looking for MPIX_Comm_shrink - found
-- Looking for MPIX_Comm_agree
-- Looking for MPIX_Comm_agree - found
-- The MPI implementation appears to have the experimental features for ULFM
that will allow you to build OpenCoarrays with failed images support. However,
ULFM support did not make it into the MPI-4 standard, and it is known to
trigger some bugs. Because of this we have disabled it by default. You may
add the
-DCAF_ENABLE_FAILED_IMAGES:BOOL=ON
CMake flag or edit the valuewith ccmake or cmake-gui if you would like to experiment with failed images.
-- Looking for include file mpi.h
-- Looking for include file mpi.h - found
-- Looking for I_MPI_VERSION
-- Looking for I_MPI_VERSION - not found
-- Performing Test HAVE_ERROR_STOP_IN_PURE
-- Performing Test HAVE_ERROR_STOP_IN_PURE - Success
-- Performing Test HAVE_VARIABLE_STOP_CODE
-- Performing Test HAVE_VARIABLE_STOP_CODE - Success
-- Configuring done
-- Generating done
-- Build files have been written to: /home/jerry/dev/buildcaf/OpenCoarrays/build
What you expected to happen
make in build directory should complete with no errors.
Step-by-step reproduction instructions to reproduce the error/bug
See above, followed instructions per install.md at github.
The text was updated successfully, but these errors were encountered: