Bug 264366 - net/mpich2: Build fails with GCC 11 on i386: error: size of array element is not a multiple of its alignment
Summary: net/mpich2: Build fails with GCC 11 on i386: error: size of array element is ...
Status: Closed FIXED
Alias: None
Product: Ports & Packages
Classification: Unclassified
Component: Individual Port(s) (show other bugs)
Version: Latest
Hardware: Any Any
: --- Affects Some People
Assignee: Thierry Thomas
URL: http://package23.nyi.freebsd.org/data...
Keywords:
Depends on:
Blocks: 258378
  Show dependency treegraph
 
Reported: 2022-05-31 10:25 UTC by Lorenzo Salvadore
Modified: 2022-05-31 15:41 UTC (History)
2 users (show)

See Also:
thierry: maintainer-feedback+


Attachments

Note You need to log in before you can comment on or make changes to this bug.
Description Lorenzo Salvadore freebsd_committer freebsd_triage 2022-05-31 10:25:11 UTC
Updating GCC_DEFAULT from GCC 10 to GCC 11 is blocked by the fact that building this port with GCC 11 on i386 fails.

The error messages are the following:

--- src/mpi/attr/attrutil.lo ---
In file included from src/mpi/attr/attrutil.c:16:
./src/include/mpiimpl.h:1458:1: error: size of array element is not a multiple of its alignment
 1458 | extern MPID_Request MPID_Request_direct[];
      | ^~~~~~
./src/include/mpiimpl.h:1458:1: error: size of array element is not a multiple of its alignment

and later we also have

--- src/mpi/attr/dup_fn.lo ---
In file included from src/mpi/attr/dup_fn.c:8:
./src/include/mpiimpl.h:1458:1: error: size of array element is not a multiple of its alignment
 1458 | extern MPID_Request MPID_Request_direct[];
      | ^~~~~~
./src/include/mpiimpl.h:1458:1: error: size of array element is not a multiple of its alignment
*** [src/mpi/attr/dup_fn.lo] Error code 1

Please also see the log in the URL field.
Comment 1 Thierry Thomas freebsd_committer freebsd_triage 2022-05-31 11:33:11 UTC
MPICH2 has been marked as deprecated since dec. 2020, and no port in the tree depends on it anymore.

Upstream switched to MPICH3 (net/mpich) and MPICH4 (not yet in the ports tree): it’s time to remove it.
Comment 2 Thierry Thomas freebsd_committer freebsd_triage 2022-05-31 15:05:13 UTC
Port removed. Thanks for the notification.
Comment 3 commit-hook freebsd_committer freebsd_triage 2022-05-31 15:05:35 UTC
A commit in branch main references this bug:

URL: https://cgit.FreeBSD.org/ports/commit/?id=c212e29dfc9d48f2999b16d110231f22439a2c78

commit c212e29dfc9d48f2999b16d110231f22439a2c78
Author:     Thierry Thomas <thierry@FreeBSD.org>
AuthorDate: 2022-05-31 15:01:15 +0000
Commit:     Thierry Thomas <thierry@FreeBSD.org>
CommitDate: 2022-05-31 15:03:23 +0000

    net/mpich2: remove

    MPICH2 has been marked as deprecated since dec. 2020, and no port in the
    tree depends on it anymore.

    Upstream switched to MPICH3 (net/mpich) and MPICH4 (not yet in the ports
    tree), and now it does not build with GCC-11: it’s time to remove it.

    PR:             264366
    Reported by:    salvadore

 MOVED                                              |   1 +
 net/Makefile                                       |   1 -
 net/mpich2/Makefile (gone)                         | 189 -----
 net/mpich2/distinfo (gone)                         |   2 -
 net/mpich2/files/patch-Makefile.in (gone)          |  11 -
 net/mpich2/files/patch-configure (gone)            |  47 --
 .../patch-src_mpe2_src_sbin_mpeinstall.in (gone)   |  47 --
 .../files/patch-src_mpi_romio_configure (gone)     |  18 -
 ...patch-src_mpi_romio_util_romioinstall.in (gone) |  19 -
 .../files/patch-src_openpa_Makefile.in (gone)      |  11 -
 net/mpich2/files/patch-src_openpa_configure (gone) |  11 -
 net/mpich2/files/patch-test_mpi_configure (gone)   |  30 -
 net/mpich2/files/pkg-message.in (gone)             |   7 -
 net/mpich2/pkg-descr (gone)                        |   8 -
 net/mpich2/pkg-plist (gone)                        | 863 ---------------------
 15 files changed, 1 insertion(+), 1264 deletions(-)
Comment 4 Gerald Pfeifer freebsd_committer freebsd_triage 2022-05-31 15:41:04 UTC
Thank you, Thierry! (And Lorenzo for post-processing the -exp run.)