Error when building openmpi-5.0.8.tar.gz

0 views
Skip to first unread message

jde...@intec.unl.edu.ar

unread,
Jul 28, 2025, 8:51:38 PMJul 28
to us...@lists.open-mpi.org

Hi there,

Today I tried to update the openmpi package by building

https://download.open-mpi.org/release/open-mpi/v5.0/openmpi-5.0.8.tar.gz

on a x86_64-pc-linux-gnu machine:

$ cat /proc/cpuinfo 
processor : 0
vendor_id : GenuineIntel
cpu family : 6
model : 60
model name : Intel(R) Core(TM) i7-4702MQ CPU @ 2.20GHz
stepping : 3
microcode : 0x28
cpu MHz : 798.197
cache size : 6144 KB
physical id : 0
siblings : 4
core id : 0
cpu cores : 4
apicid : 0
initial apicid : 0
fpu : yes
fpu_exception : yes
cpuid level : 13
wp : yes


using a Fedora 42 release:

$ uname --a
Linux pitonisa 6.5.6-300.fc39.x86_64 #1 SMP PREEMPT_DYNAMIC Fri Oct  6 19:57:21 UTC 2023 x86_64 GNU/Linux

However, this time, there is the following error:

...
CC       pevent.o
CCLD     pevent
/usr/bin/ld: /home/bigpack/openmpi-paq/openmpi-5.0.8/BUILD/3rd-party/libevent-2.1.12-stable-ompi/.libs/libevent_pthreads.so: undefined reference to `__pthread_cond_timedwait64'
collect2: error: ld returned 1 exit status
make[4]: *** [Makefile:711: pevent] Error 1
make[4]: Leaving directory '/home/bigpack/openmpi-paq/openmpi-5.0.8/BUILD/3rd-party/openpmix/src/tools/pevent'
make[3]: *** [Makefile:1291: all-recursive] Error 1
make[3]: Leaving directory '/home/bigpack/openmpi-paq/openmpi-5.0.8/BUILD/3rd-party/openpmix/src'
make[2]: *** [Makefile:789: all-recursive] Error 1
make[2]: Leaving directory '/home/bigpack/openmpi-paq/openmpi-5.0.8/BUILD/3rd-party/openpmix'
make[1]: *** [Makefile:1412: all-recursive] Error 1
make[1]: Leaving directory '/home/bigpack/openmpi-paq/openmpi-5.0.8/BUILD/3rd-party'
make: *** [Makefile:1540: all-recursive] Error 1
 
(see the 3 attached files).

Please,  any clue to solve it?

Thanks in advance.

Regards.
Jorge D'Elia.
--


config.log.gz
configure.eco.gz
make-all.eco.gz

Gilles Gouaillardet

unread,
Jul 28, 2025, 9:18:23 PMJul 28
to us...@lists.open-mpi.org
Jorge,

The logs suggest you built on a Fedora 39 with a custom built gcc.
You might have more luck using the distro compiler, also, using the distro provided hwloc and libevent libraries will make things easier.

Cheers,

Gilles 

To unsubscribe from this group and stop receiving emails from it, send an email to users+un...@lists.open-mpi.org.

jde...@intec.unl.edu.ar

unread,
Jul 29, 2025, 1:23:54 PMJul 29
to us...@lists.open-mpi.org
Dear Gilles,

Thank you very much for your reply!

Gilles Gouaillardet <gilles.go...@gmail.com> escribió:

> Jorge,
>
> The logs suggest you built on a Fedora 39 with a custom built gcc.

I almost agree (that is, it is a custom built of gcc). Last night I
also noticed
that the messages suggested I had Fedora 39 but, on the other hand,
since I routinely update to the latest versions, without any
difficulties of any
kind, I am almost certain that is not the case. So now I am surprised by the
discrepancy with what is being reported, on the same laptop, for example.

$ cat /etc/fedora-release
Fedora release 42 (Adams)

$ uname -r
6.5.6-300.fc39.x86_64

$ cat /etc/os-release
NAME="Fedora Linux"
VERSION="42 (Workstation Edition)"
RELEASE_TYPE=stable
ID=fedora
VERSION_ID=42
VERSION_CODENAME=""
PLATFORM_ID="platform:f42"
PRETTY_NAME="Fedora Linux 42 (Workstation Edition)"
ANSI_COLOR="0;38;2;60;110;180"
LOGO=fedora-logo-icon
CPE_NAME="cpe:/o:fedoraproject:fedora:42"
DEFAULT_HOSTNAME="fedora"
HOME_URL="https://fedoraproject.org/"
DOCUMENTATION_URL="https://docs.fedoraproject.org/en-US/fedora/f42/system-administrators-guide/"
SUPPORT_URL="https://ask.fedoraproject.org/"
BUG_REPORT_URL="https://bugzilla.redhat.com/"
REDHAT_BUGZILLA_PRODUCT="Fedora"
REDHAT_BUGZILLA_PRODUCT_VERSION=42
REDHAT_SUPPORT_PRODUCT="Fedora"
REDHAT_SUPPORT_PRODUCT_VERSION=42
SUPPORT_END=2026-05-13
VARIANT="Workstation Edition"
VARIANT_ID=workstation

Perhaps it would be more advisable to reinstall Fedora 42 from scratch?

> You might have more luck using the distro compiler, also, using the distro
> provided hwloc and libevent libraries will make things easier.

Ok. I will also keep this suggestion in mind.


> Cheers,
>
> Gilles

Best regards.
Jorge D'Elia.

> ---

Gilles Gouaillardet

unread,
Jul 29, 2025, 11:38:41 PMJul 29
to us...@lists.open-mpi.org
Jorge,

you'd probably retry from a fresh install.

FWIW, I built Open 5.0.8 on a Fedora 42 container without any issues, so that suggests the issue could come from your environment.

Cheers,

Gilles

jde...@intec.unl.edu.ar

unread,
Jul 30, 2025, 7:58:36 AMJul 30
to us...@lists.open-mpi.org
Dear Gilles,

> Gilles Gouaillardet <gilles.go...@gmail.com> escribió:
>
> Jorge,
>
> you'd probably retry from a fresh install.

Ok. I will do that. Thank you very much for your time in responding.

> FWIW, I built Open 5.0.8 on a Fedora 42 container without any issues, so
> that suggests the issue could come from your environment

You are right! The problem was with my beta version of gfortran, since using
the system one, the OpenMPI compilation was completed successfully, as it
almost always does.

> Cheers,
>
> Gilles

Regards,
Jorge.
Reply all
Reply to author
Forward
0 new messages