[petsc-users] runtime error on Summit with nvhpc21.7

Barry Smith bsmith at petsc.dev
Fri Aug 27 23:19:06 CDT 2021 


> On Aug 27, 2021, at 5:05 PM, Mark Adams <mfadams at lbl.gov> wrote:
> 
> 
> 
> On Fri, Aug 27, 2021 at 5:03 PM Junchao Zhang <junchao.zhang at gmail.com <mailto:junchao.zhang at gmail.com>> wrote:
> I don't understand the configure options
> 
> --with-cc=/gpfs/alpine/world-shared/phy122/lib/install/summit/kokkos/nvhpc21.7/bin/nvcc_wrapper --with-cxx=/gpfs/alpine/world-shared/phy122/lib/install/summit/kokkos/nvhpc21.7/bin/nvcc_wrapper --with-fc=/sw/summit/spack-envs/base/opt/linux-rhel8-ppc64le/nvhpc-21.7/spectrum-mpi-10.4.0.3-20210112-nv7jd363ym3n4zpgornfbq6bh4tqjyak/bin/mpifort COPTFLAGS="-g -fast" CXXOPTFLAGS="-g -fast" FOPTFLAGS="-g -fast" CUDAFLAGS="-ccbin nvc++" --with-ssl=0 --with-batch=0 --with-mpiexec="jsrun -g 1" --with-cuda=0 --with-cudac=/gpfs/alpine/world-shared/phy122/lib/install/summit/kokkos/nvhpc21.7/bin/nvcc_wrapper --with-cuda-gencodearch=70 --download-metis --download-parmetis --with-x=0 --with-debugging=0 PETSC_ARCH=arch-summit-opt-nvhpc --prefix=/gpfs/alpine/world-shared/phy122/lib/install/summit/petsc/current/opt-nvhpc21.7b
> 
> Why do you need to use nvcc_wrapper if you do not want to use cuda?
> 
> That code that is having a problem links with nvcc_wrapper.
> They get a segv that I sent earlier, in PetscInitialize so I figure I should use the same compiler / linker.
> They use CUDA, but we don't need PETSc to use CUDA now.
>  
> In addition, nvcc_wrapper is a C++ compiler. Using it for --with-cc=, you also need --with-clanguage=c++ 
> 
> I rebuilt PETSc with mpicc, mpiCC, mpif90 and --with-nvcc=nvcc_wrapper and that built make check works. I gave it to them to test.

   This is an odd way to do it. The Kokkos nvcc_wrapper wraps the nvcc compiler to allow it to compile Kokkos code and link it against the Kokkos libraries; so using nvcc_wrapper as nvcc is strangely recursive; sure everything in PETSc/Kokkos may build ok (assuming the nvcc that the nvcc_wrapper uses is correct for the situation and uses a correct underlying C++) but it is freakish. PETSc should just be built with the same nvcc that the nvcc_wrapper is using and using the same inner C++ compiler. 

  I suspect the crashes came from the /gpfs/alpine/world-shared/phy122/lib/install/summit/kokkos/nvhpc21.7/bin/nvcc_wrapper not using the same nvcc and internal C++ compiler as PETSc is ending up using.

  But if it works, I guess it works. Perhaps when PETSc does not build Kokkos we should have a --with-kokkos-nvcc-wrapper= to allow setting the wrapper.

 Barry



> 
> Thanks,
> Mark
>  
> 
> --Junchao Zhang
> 
> 
> On Fri, Aug 27, 2021 at 3:28 PM Mark Adams <mfadams at lbl.gov <mailto:mfadams at lbl.gov>> wrote:
> 
> 
> On Fri, Aug 27, 2021 at 3:56 PM Junchao Zhang <junchao.zhang at gmail.com <mailto:junchao.zhang at gmail.com>> wrote:
> 
> 
> On Fri, Aug 27, 2021, 1:52 PM Mark Adams <mfadams at lbl.gov <mailto:mfadams at lbl.gov>> wrote:
> I think the problem is that I build with MPICC and they use nvcc_wrapper. I could just try building PETSc with CC=nvcc_wrapper, but it was not clear if this was the way to go.
> --with-nvcc=nvcc_wrapper
> 
> What do I specify for cc and CC?
>  
> I will try it.
> Thanks,
> Mark
> 
> On Fri, Aug 27, 2021 at 10:50 AM Junchao Zhang <junchao.zhang at gmail.com <mailto:junchao.zhang at gmail.com>> wrote:
> 
> 
> 
> On Fri, Aug 27, 2021 at 7:06 AM Mark Adams <mfadams at lbl.gov <mailto:mfadams at lbl.gov>> wrote:
> I have a user (cc'ed) that has a C++ code and is using a PETSc that I built. He is getting this runtime error.
> 
> 'make check' runs clean and I built snes/tutorial/ex1 manually, to get a link line, and it ran fine.
> I appended the users link line and my test.
> 
> I see that they are using Kokkos' "nvcc_wrapper". Should I rebuild PETSc using that, maybe we just need to make sure we are both using the same underlying compiler or should they use mpiCC?
> It looks like they used nvcc_wrapper to replace nvcc.  You can ask them to use nvcc directly to see what happens. But the error happened in petsc initialization, petscsys_petscinitializenohelp, so I doubt it helps.  The easy way is to just attach a debugger. 
>  
> Thanks,
> Mark
> 
> 
> [e13n16:591873] *** Process received signal ***
> [e13n16:591873] Signal: Segmentation fault (11)
> [e13n16:591873] Signal code: Invalid permissions (2)
> [e13n16:591873] Failing at address: 0x102c87e0
> [e13n16:591873] [ 0] linux-vdso64.so.1(__kernel_sigtramp_rt64+0x0)[0x2000000504d8]
> [e13n16:591873] [ 1] [e13n16:591872] *** Process received signal ***
> [e13n16:591872] Signal: Segmentation fault (11)
> [e13n16:591872] Signal code: Invalid permissions (2)
> [e13n16:591872] Failing at address: 0x102c87e0
> [e13n16:591872] [ 0] linux-vdso64.so.1(__kernel_sigtramp_rt64+0x0)[0x2000000504d8]
> [e13n16:591872] [ 1] [e13n16:591871] *** Process received signal ***
> [e13n16:591871] Signal: Segmentation fault (11)
> [e13n16:591871] Signal code: Invalid permissions (2)
> [e13n16:591871] Failing at address: 0x102c87e0
> [e13n16:591871] [ 0] linux-vdso64.so.1(__kernel_sigtramp_rt64+0x0)[0x2000000504d8]
> [e13n16:591871] [ 1] /autofs/nccs-svm1_sw/summit/nvhpc_sdk/rhel8/Linux_ppc64le/21.7/compilers/lib/libnvf.so(pgf90_str_copy_klen+0x1fc)[0x200004a79ee4]
> [e13n16:591871] [ 2] [e13n16:591874] *** Process received signal ***
> [e13n16:591874] Signal: Segmentation fault (11)
> [e13n16:591874] Signal code: Invalid permissions (2)
> [e13n16:591874] Failing at address: 0x102c87e0
> [e13n16:591874] [ 0] linux-vdso64.so.1(__kernel_sigtramp_rt64+0x0)[0x2000000504d8]
> [e13n16:591874] [ 1] /autofs/nccs-svm1_sw/summit/nvhpc_sdk/rhel8/Linux_ppc64le/21.7/compilers/lib/libnvf.so(pgf90_str_copy_klen+0x1fc)[0x200004a79ee4]
> [e13n16:591874] [ 2] /gpfs/alpine/world-shared/phy122/lib/install/summit/petsc/current/opt-nvhpc21.7b/lib/libpetsc.so.3.015(petscsys_petscinitializenohelp_+0xf4)[0x20000097b3ec]
> [e13n16:591874] [ 3] /ccs/home/scheinberg/new_install/build/bin/xgc-es-cpp[0x10131dd8]
> [e13n16:591874] [ 4] /ccs/home/scheinberg/new_install/build/bin/xgc-es-cpp[0x10015c60]
> [e13n16:591874] [ 5] /ccs/home/scheinberg/new_install/build/bin/xgc-es-cpp[0x1005a8b0]
> [e13n16:591874] [ 6] /ccs/home/scheinberg/new_install/build/bin/xgc-es-cpp[0x10015b14]
> [e13n16:591874] [ 7] /ccs/home/scheinberg/new_install/build/bin/xgc-es-cpp[0x10014cd0]
> [e13n16:591874] [ 8] /gpfs/alpine/world-shared/phy122/lib/install/summit/petsc/current/opt-nvhpc21.7b/lib/libpetsc.so.3.015(petscsys_petscinitializenohelp_+0xf4)[0x20000097b3ec]
> [e13n16:591871] [ 3] /ccs/home/scheinberg/new_install/build/bin/xgc-es-cpp[0x10131dd8]
> [e13n16:591871] [ 4] /ccs/home/scheinberg/new_install/build/bin/xgc-es-cpp[0x10015c60]
> [e13n16:591871] [ 5] /ccs/home/scheinberg/new_install/build/bin/xgc-es-cpp[0x1005a8b0]
> [e13n16:591871] [ 6] /ccs/home/scheinberg/new_install/build/bin/xgc-es-cpp[0x10015b14]
> [e13n16:591871] [ 7] /ccs/home/scheinberg/new_install/build/bin/xgc-es-cpp[0x10014cd0]
> [e13n16:591871] [ 8] /usr/lib/gcc/ppc64le-redhat-linux/8/../../../../lib64/power9/libc.so.6(+0x24078)[0x200005934078]
> [e13n16:591871] [ 9] /usr/lib/gcc/ppc64le-redhat-linux/8/../../../../lib64/power9/libc.so.6(__libc_start_main+0xb4)[0x200005934264]
> [e13n16:591871] *** End of error message ***
> /usr/lib/gcc/ppc64le-redhat-linux/8/../../../../lib64/power9/libc.so.6(+0x24078)[0x200005934078]
> [e13n16:591874] [ 9] /usr/lib/gcc/ppc64le-redhat-linux/8/../../../../lib64/power9/libc.so.6(__libc_start_main+0xb4)[0x200005934264]
> [e13n16:591874] *** End of error message ***
> /autofs/nccs-svm1_sw/summit/nvhpc_sdk/rhel8/Linux_ppc64le/21.7/compilers/lib/libnvf.so(pgf90_str_copy_klen+0x1fc)[0x200004a79ee4]
> [e13n16:591872] [ 2] /gpfs/alpine/world-shared/phy122/lib/install/summit/petsc/current/opt-nvhpc21.7b/lib/libpetsc.so.3.015(petscsys_petscinitializenohelp_+0xf4)[0x20000097b3ec]
> [e13n16:591872] [ 3] /ccs/home/scheinberg/new_install/build/bin/xgc-es-cpp[0x10131dd8]
> [e13n16:591872] [ 4] /ccs/home/scheinberg/new_install/build/bin/xgc-es-cpp[0x10015c60]
> [e13n16:591872] [ 5] /ccs/home/scheinberg/new_install/build/bin/xgc-es-cpp[0x1005a8b0]
> [e13n16:591872] [ 6] /ccs/home/scheinberg/new_install/build/bin/xgc-es-cpp[0x10015b14]
> [e13n16:591872] [ 7] /ccs/home/scheinberg/new_install/build/bin/xgc-es-cpp[0x10014cd0]
> [e13n16:591872] [ 8] /usr/lib/gcc/ppc64le-redhat-linux/8/../../../../lib64/power9/libc.so.6(+0x24078)[0x200005934078]
> [e13n16:591872] [ 9] /usr/lib/gcc/ppc64le-redhat-linux/8/../../../../lib64/power9/libc.so.6(__libc_start_main+0xb4)[0x200005934264]
> [e13n16:591872] *** End of error message ***
> /autofs/nccs-svm1_sw/summit/nvhpc_sdk/rhel8/Linux_ppc64le/21.7/compilers/lib/libnvf.so(pgf90_str_copy_klen+0x1fc)[0x200004a79ee4]
> [e13n16:591873] [ 2] /gpfs/alpine/world-shared/phy122/lib/install/summit/petsc/current/opt-nvhpc21.7b/lib/libpetsc.so.3.015(petscsys_petscinitializenohelp_+0xf4)[0x20000097b3ec]
> [e13n16:591873] [ 3] /ccs/home/scheinberg/new_install/build/bin/xgc-es-cpp[0x10131dd8]
> [e13n16:591873] [ 4] /ccs/home/scheinberg/new_install/build/bin/xgc-es-cpp[0x10015c60]
> [e13n16:591873] [ 5] /ccs/home/scheinberg/new_install/build/bin/xgc-es-cpp[0x1005a8b0]
> [e13n16:591873] [ 6] /ccs/home/scheinberg/new_install/build/bin/xgc-es-cpp[0x10015b14]
> [e13n16:591873] [ 7] /ccs/home/scheinberg/new_install/build/bin/xgc-es-cpp[0x10014cd0]
> [e13n16:591873] [ 8] /usr/lib/gcc/ppc64le-redhat-linux/8/../../../../lib64/power9/libc.so.6(+0x24078)[0x200005934078]
> [e13n16:591873] [ 9] /usr/lib/gcc/ppc64le-redhat-linux/8/../../../../lib64/power9/libc.so.6(__libc_start_main+0xb4)[0x200005934264]
> [e13n16:591873] *** End of error message ***
> ERROR:  One or more process (first noticed rank 1) terminated with signal 11 (core dumped)
> 
> 
> 
> /gpfs/alpine/world-shared/phy122/lib/install/summit/kokkos/nvhpc21.7/bin/nvcc_wrapper -arch=sm_70 CMakeFiles/xgc-es-cpp.dir/xgc-es-cpp_build_info.F90.o -o bin/xgc-es-cpp  -Wl,-rpath,/gpfs/alpine/world-shared/phy122/lib/install/summit/petsc/current/opt-nvhpc21.7b/lib:/gpfs/alpine/world-shared/phy122/lib/install/summit/adios2/devel/nvhpc/lib64 liblibxgc-es-cpp.a /sw/summit/spack-envs/base/opt/linux-rhel8-ppc64le/nvhpc-21.7/netlib-lapack-3.9.1-b5iqtudpwjumes5gsdol3bzsh7qlv7mf/lib64/liblapack.so /sw/summit/spack-envs/base/opt/linux-rhel8-ppc64le/nvhpc-21.7/netlib-lapack-3.9.1-b5iqtudpwjumes5gsdol3bzsh7qlv7mf/lib64/libblas.so /gpfs/alpine/world-shared/phy122/lib/install/summit/petsc/current/opt-nvhpc21.7b/lib/libpetsc.so /gpfs/alpine/world-shared/phy122/lib/install/summit/petsc/current/opt-nvhpc21.7b/lib/libparmetis.so /gpfs/alpine/world-shared/phy122/lib/install/summit/petsc/current/opt-nvhpc21.7b/lib/libmetis.so /sw/summit/spack-envs/base/opt/linux-rhel8-ppc64le/nvhpc-21.7/fftw-3.3.9-bzi7deue27ijd7xm4zn7pt22u4sj47g4/lib/libfftw3.so libs/pspline/libpspline.a libs/camtimers/libtimers.a /autofs/nccs-svm1_sw/summit/nvhpc_sdk/rhel8/Linux_ppc64le/21.7/compilers/lib/libacchost.so /gpfs/alpine/world-shared/phy122/lib/install/summit/adios2/devel/nvhpc/lib64/libadios2_fortran_mpi.so.2.7.1 /gpfs/alpine/world-shared/phy122/lib/install/summit/adios2/devel/nvhpc/lib64/libadios2_fortran.so.2.7.1 /gpfs/alpine/world-shared/phy122/lib/install/summit/kokkos/DEFAULT/install/lib64/libkokkoscontainers.a /gpfs/alpine/world-shared/phy122/lib/install/summit/kokkos/DEFAULT/install/lib64/libkokkoscore.a /usr/lib64/libcuda.so /autofs/nccs-svm1_sw/summit/nvhpc_sdk/rhel8/Linux_ppc64le/21.7/cuda/11.0/lib64/libcudart.so /usr/lib64/libdl.so -lmpi_ibm_usempif08 -lmpi_ibm_usempi_ignore_tkr -lmpi_ibm_mpifh -lnvf -Wl,-rpath-link,/gpfs/alpine/world-shared/phy122/lib/install/summit/adios2/devel/nvhpc/lib64
> 
> 
> 19:39 main= /gpfs/alpine/csc314/scratch/adams/petsc/src/snes/tutorials$ make PETSC_DIR=/gpfs/alpine/world-shared/phy122/lib/install/summit/petsc/current/opt-nvhpc21.7 PETSC_ARCH="" ex1
> mpicc -fPIC -g -fast  -fPIC -g -fast    -I/gpfs/alpine/world-shared/phy122/lib/install/summit/petsc/current/opt-nvhpc21.7/include     ex1.c  -Wl,-rpath,/gpfs/alpine/world-shared/phy122/lib/install/summit/petsc/current/opt-nvhpc21.7/lib -L/gpfs/alpine/world-shared/phy122/lib/install/summit/petsc/current/opt-nvhpc21.7/lib -Wl,-rpath,/gpfs/alpine/world-shared/phy122/lib/install/summit/petsc/current/opt-nvhpc21.7/lib -L/gpfs/alpine/world-shared/phy122/lib/install/summit/petsc/current/opt-nvhpc21.7/lib -Wl,-rpath,/sw/summit/spack-envs/base/opt/linux-rhel8-ppc64le/nvhpc-21.7/spectrum-mpi-10.4.0.3-20210112-nv7jd363ym3n4zpgornfbq6bh4tqjyak/lib -L/sw/summit/spack-envs/base/opt/linux-rhel8-ppc64le/nvhpc-21.7/spectrum-mpi-10.4.0.3-20210112-nv7jd363ym3n4zpgornfbq6bh4tqjyak/lib -Wl,-rpath,/sw/summit/spack-envs/base/opt/linux-rhel8-ppc64le/nvhpc-21.7/hdf5-1.10.7-nfhjvzsshg5qihqv44y5ji6ihsqpd73v/lib -L/sw/summit/spack-envs/base/opt/linux-rhel8-ppc64le/nvhpc-21.7/hdf5-1.10.7-nfhjvzsshg5qihqv44y5ji6ihsqpd73v/lib -Wl,-rpath,/sw/summit/spack-envs/base/opt/linux-rhel8-ppc64le/nvhpc-21.7/netlib-lapack-3.9.1-b5iqtudpwjumes5gsdol3bzsh7qlv7mf/lib64 -L/sw/summit/spack-envs/base/opt/linux-rhel8-ppc64le/nvhpc-21.7/netlib-lapack-3.9.1-b5iqtudpwjumes5gsdol3bzsh7qlv7mf/lib64 -Wl,-rpath,/autofs/nccs-svm1_sw/summit/nvhpc_sdk/rhel8/Linux_ppc64le/21.7/compilers/lib -L/autofs/nccs-svm1_sw/summit/nvhpc_sdk/rhel8/Linux_ppc64le/21.7/compilers/lib -Wl,-rpath,/usr/lib/gcc/ppc64le-redhat-linux/8 -L/usr/lib/gcc/ppc64le-redhat-linux/8 -lpetsc -llapack -lblas -lparmetis -lmetis -lstdc++ -ldl -lpthread -lmpiprofilesupport -lmpi_ibm_usempif08 -lmpi_ibm_usempi_ignore_tkr -lmpi_ibm_mpifh -lmpi_ibm -lnvf -lnvomp -latomic -lnvhpcatm -lnvcpumath -lnvc -lrt -lm -lgcc_s -lstdc++ -ldl -o ex1
> 19:40 main= /gpfs/alpine/csc314/scratch/adams/petsc/src/snes/tutorials$ mpicc --version
> 
> nvc 21.7-0 linuxpower target on Linuxpower
> NVIDIA Compilers and Tools
> Copyright (c) 2021, NVIDIA CORPORATION & AFFILIATES.  All rights reserved.
> 19:40 main= /gpfs/alpine/csc314/scratch/adams/petsc/src/snes/tutorials$ jsrun -n 1 ./ex1 -ksp_monitor
>     0 KSP Residual norm 6.041522986797e+00
>     1 KSP Residual norm 1.042493382631e+00
>     2 KSP Residual norm 7.950907844730e-16
>     0 KSP Residual norm 4.786756692342e+00
>     1 KSP Residual norm 1.426392207750e-01
>     2 KSP Residual norm 1.801079604472e-15
>     0 KSP Residual norm 2.986456323228e+00
>     1 KSP Residual norm 7.669888809223e-02
>     2 KSP Residual norm 3.744083117256e-16
>     0 KSP Residual norm 2.306244667700e-01
>     1 KSP Residual norm 1.355550749587e-02
>     2 KSP Residual norm 5.845524837731e-17
>     0 KSP Residual norm 1.936314002654e-03
>     1 KSP Residual norm 2.125593590819e-04
>     2 KSP Residual norm 6.987141455073e-20
>     0 KSP Residual norm 1.435593531990e-07
>     1 KSP Residual norm 2.588271385567e-08
>     2 KSP Residual norm 3.942196167935e-23
> 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.mcs.anl.gov/pipermail/petsc-users/attachments/20210827/d471be0e/attachment-0001.html>

More information about the petsc-users mailing list