Bug#881873: trilinos: FTBFS on sparc64: Bus errors in 11 tests

[Nico Schlömer]

Chances are slim for to ever be resolved. Upstream only supports amd64, and
PRs to fix building for other architectures are rejected. (See, e.g., [1].)
I'd say we'll have to live with trilinos not running on sparc64.

Cheers,
Nico

[1] https://github.com/kokkos/kokkos/pull/576

On Thu, Nov 16, 2017 at 1:21 AM Aaron M. Ucko <ucko at debian.org> wrote:

> Source: trilinos
> Version: 12.10.1-4+b1
> Severity: important
> Tags: upstream
> Justification: fails to build from source (but built successfully in the
> past)
> User: debian-sparc at lists.debian.org
> Usertags: sparc64
>
> Builds of trilinos for sparc64 (admittedly not a release architecture)
> have been failing lately with bus errors (indicating attempts to use
> underaligned pointers, about which sparc64 is notoriously strict), per
> the below excerpts from
>
> https://buildd.debian.org/status/fetch.php?pkg=trilinos&arch=sparc64&ver=12.12.1-1&stamp=1510746436&raw=0
> .
>
> Could you please take a look?
>
> Thanks!
>
> ------------------------------------------------------------
>
> [...]
> 521/871 Test #521: EpetraExt_Transpose_test_LL_MPI_4
> ..................................................***Failed    2.31 sec
> Epetra::MpiComm::Processor 0 of 4 total processors.
> EpetraExt in Trilinos 12.12.1
>
> Epetra::MpiComm::Processor 1 of 4 total processors.
> Epetra::MpiComm::Processor 2 of 4 total processors.
> Epetra::MpiComm::Processor 3 of 4 total processors.
> [landau:24458] *** Process received signal ***
> [landau:24458] Signal: Bus error (10)
> [landau:24458] Signal code: Invalid address alignment (1)
> [landau:24458] Failing at address: 0x100005ef4bc
> [landau:24459] *** Process received signal ***
> [landau:24459] Signal: Bus error (10)
> [landau:24459] Signal code: Invalid address alignment (1)
> [landau:24459] Failing at address: 0x100005eca7c
> [landau:24458] *** End of error message ***
> [landau:24459] *** End of error message ***
> [landau:24456] *** Process received signal ***
> [landau:24456] Signal: Bus error (10)
> [landau:24456] Signal code: Invalid address alignment (1)
> [landau:24456] Failing at address: 0x1000060a96c
> [landau:24457] *** Process received signal ***
> [landau:24457] Signal: Bus error (10)
> [landau:24457] Signal code: Invalid address alignment (1)
> [landau:24457] Failing at address: 0x100005ef4dc
> [landau:24456] *** End of error message ***
> [landau:24457] *** End of error message ***
> --------------------------------------------------------------------------
> mpiexec noticed that process rank 1 with PID 0 on node landau exited on
> signal 10 (Bus error).
> --------------------------------------------------------------------------
> [...]
> The following tests FAILED:
>         517 - EpetraExt_Permutation_test_LL_MPI_4 (Failed)
>         521 - EpetraExt_Transpose_test_LL_MPI_4 (Failed)
>         523 - EpetraExt_inout_test_LL_MPI_4 (Failed)
>         525 - EpetraExt_MatrixMatrix_test_LL_MPI_4 (Failed)
>         621 - AztecOO_AztecOO_test_LL_MPI_4 (Failed)
>         633 - ML_AztecSimple_MPI_4 (Failed)
>         636 - ML_MatrixFree_MPI_4 (Failed)
>         638 - ML_ML_Operator2Epetra_RowMatrix_MPI_4 (Failed)
>         639 - ML_MLP_Maxwell_MPI_4 (Failed)
>         640 - ML_MLP_NonSym_MPI_4 (Failed)
>         648 - Komplex_simple_MPI_4 (Failed)
> Errors while running CTest
>
> --
> Aaron M. Ucko, KB1CJC (amu at alum.mit.edu, ucko at debian.org)
> http://www.mit.edu/~amu/ |
> http://stuff.mit.edu/cgi/finger/?amu@monk.mit.edu
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.alioth.debian.org/pipermail/debian-science-maintainers/attachments/20171117/cceb5d32/attachment-0002.html>