Skip to content
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

xl/16.1.1 with cpp14, openmp build, nightly test failures #2856

Closed
ndellingwood opened this issue Mar 11, 2020 · 1 comment
Closed

xl/16.1.1 with cpp14, openmp build, nightly test failures #2856

ndellingwood opened this issue Mar 11, 2020 · 1 comment
Assignees
Labels
Bug Broken / incorrect code; it could be Kokkos' responsibility, or others’ (e.g., Trilinos)

Comments

@ndellingwood
Copy link
Contributor

Various tests are failing on White with the xl/16.1.1 compiler in an OpenMP build with C++14 enabled.

Sample failure output:

04:37:20 1: /home/jenkins/white/workspace/Kokkos_develop_white_xl_16.1_cpp14/kokkos/core/unit_test/TestMemoryPool.hpp:311: Failure
04:37:20 1: Value of: result
04:37:20 1:   Actual: 62365
04:37:20 1: Expected: nfill
04:37:20 1: Which is: 70000
04:37:20 1: [  FAILED  ] openmp.memory_pool (9 ms)
04:37:20 1: [ RUN      ] openmp.cxx11
04:37:20 1: CXX11 ( test = 'AddTest' FAILED : relative error 0.0486693 > tolerance 1e-14
04:37:20 1: /home/jenkins/white/workspace/Kokkos_develop_white_xl_16.1_cpp14/kokkos/core/unit_test/TestCXX11.hpp:373: Failure
04:37:20 1: Value of: (TestCXX11::Test<TEST_EXECSPACE>(1))
04:37:20 1:   Actual: false
04:37:20 1: Expected: true
04:37:20 1: [  FAILED  ] openmp.cxx11 (0 ms)
  # Reproducer instructions:
  #   Load modules:
        module load cmake/3.12.3 ibm/xl/16.1.1 gcc/7.2.0

  #   Use generate_makefile line below to call cmake which generates makefile for this build:
        <KOKKOS_PATH>/generate_makefile.bash --with-devices=OpenMP --arch=Power8,Kepler37 --compiler=/home/projects/ppc64le/ibm/xlC/16.1.1/bin/xlC --cxxflags="-O3 -Wall -Wshadow -pedantic -Wsign-compare -Wtype-limits -Wuninitialized -qgcc_cpp_stdinc=/home/projects/ppc64le/gcc/7.2.0/include/c++/7.2.0 -I/home/projects/ppc64le/gcc/7.2.0/include/c++/7.2.0/powerpc64le-unknown-linux-gnu" --cxxstandard="14" --ldflags=""  --kokkos-path=<KOKKOS_PATH> --no-examples --with-options=enable_large_mem_tests

If testing on White, I suggest the rhel7T queue as the Kepler queue is down to a single node and tends to be congested.

@ndellingwood ndellingwood added Bug Broken / incorrect code; it could be Kokkos' responsibility, or others’ (e.g., Trilinos) DevelopOnly labels Mar 11, 2020
@ndellingwood ndellingwood added this to the Tentative 3.1 Release milestone Mar 11, 2020
@ndellingwood ndellingwood added the Blocks Promotion Overview issue for release-blocking bugs label Mar 11, 2020
@crtrott crtrott self-assigned this Mar 11, 2020
@crtrott
Copy link
Member

crtrott commented Mar 12, 2020

This was potentially caused by #2605 too. We had no successful testing since the revert due to wide spread file system issues though. But on a power 9 system I can't reproduce this right now.

@ndellingwood ndellingwood added InDevelop and removed Blocks Promotion Overview issue for release-blocking bugs labels Apr 9, 2020
@crtrott crtrott closed this as completed Apr 14, 2020
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
Bug Broken / incorrect code; it could be Kokkos' responsibility, or others’ (e.g., Trilinos)
Projects
None yet
Development

No branches or pull requests

2 participants