[bug#27806] openmpi changes

[bug#27806] openmpi changes

[Dave Love]

[-- Attachment #1: Type: text/plain, Size: 281 bytes --]

I assume the only one of these that's questionable is the SGE one.  It
used to be the default, and I don't know of any way it could be
problematic -- the support just looks at the environment to determine
whether to use qrsh for remote startup.  Debian and Fedora both enable
it.


[-- Attachment #2: Version update --]
[-- Type: text/x-diff, Size: 1356 bytes --]

From a03995a070bbfc31c091d3b90267acf4a3b242b7 Mon Sep 17 00:00:00 2001
From: Dave Love <fx@gnu.org>
Date: Sun, 23 Jul 2017 22:00:21 +0100
Subject: [PATCH 1/5] * mpi.scm (gnu, openmpi): Update to 1.10.7.

---
 gnu/packages/mpi.scm | 5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)

diff --git a/gnu/packages/mpi.scm b/gnu/packages/mpi.scm
index 3b1ba003c..6adff2a79 100644
--- a/gnu/packages/mpi.scm
+++ b/gnu/packages/mpi.scm
@@ -3,6 +3,7 @@
 ;;; Copyright © 2014, 2015, 2016, 2017 Ludovic Courtès <ludo@gnu.org>
 ;;; Copyright © 2014 Ian Denhardt <ian@zenhack.net>
 ;;; Copyright © 2016 Andreas Enge <andreas@enge.fr>
+;;; Copyright © 2017 Dave Love <fx@gnu.org>
 ;;;
 ;;; This file is part of GNU Guix.
 ;;;
@@ -110,7 +111,7 @@ bind processes, and much more.")
 (define-public openmpi
   (package
     (name "openmpi")
-    (version "1.10.3")
+    (version "1.10.7")
     (source
      (origin
       (method url-fetch)
@@ -119,7 +120,7 @@ bind processes, and much more.")
                           "/downloads/openmpi-" version ".tar.bz2"))
       (sha256
        (base32
-        "0k95ri9f8kzx5vhzrdbzn59rn2324fs4a96w5v8jy20j8dkbp13l"))))
+        "142s1vny9gllkq336yafxayjgcirj2jv0ddabj879jgya7hyr2d0"))))
     (build-system gnu-build-system)
     (inputs
      `(("hwloc" ,hwloc "lib")
-- 
2.11.0


[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #3: enable SGE --]
[-- Type: text/x-diff, Size: 879 bytes --]

From 7b0b1b2c92353a5253da260f51c9dc946ffe77d8 Mon Sep 17 00:00:00 2001
From: Dave Love <fx@gnu.org>
Date: Sun, 23 Jul 2017 22:02:54 +0100
Subject: [PATCH 2/5] * mpi.scm (gnu, openmpi): Enable SGE support.

---
 gnu/packages/mpi.scm | 1 +
 1 file changed, 1 insertion(+)

diff --git a/gnu/packages/mpi.scm b/gnu/packages/mpi.scm
index 6adff2a79..34d78ed30 100644
--- a/gnu/packages/mpi.scm
+++ b/gnu/packages/mpi.scm
@@ -138,6 +138,7 @@ bind processes, and much more.")
                            "--enable-mpi-ext=all"
                            "--with-devel-headers"
                            "--enable-memchecker"
+                           "--with-sge"
                            ,(string-append "--with-valgrind="
                                            (assoc-ref %build-inputs "valgrind"))
                            ,(string-append "--with-hwloc="
-- 
2.11.0


[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #4: fix description --]
[-- Type: text/x-diff, Size: 1148 bytes --]

From 265475a9655959506409b8fe905f1b0452831423 Mon Sep 17 00:00:00 2001
From: Dave Love <fx@gnu.org>
Date: Sun, 23 Jul 2017 22:30:40 +0100
Subject: [PATCH 4/5] * mpi.scm (gnu, openmpi): It's MPI 3.0, not MPI 2.

---
 gnu/packages/mpi.scm | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/gnu/packages/mpi.scm b/gnu/packages/mpi.scm
index 24d290bc3..223dd2ae6 100644
--- a/gnu/packages/mpi.scm
+++ b/gnu/packages/mpi.scm
@@ -156,9 +156,9 @@ bind processes, and much more.")
                         (for-each delete-file (find-files out "config.log"))
                         #t))))))
     (home-page "http://www.open-mpi.org")
-    (synopsis "MPI-2 implementation")
+    (synopsis "MPI-3 implementation")
     (description
-     "The Open MPI Project is an MPI-2 implementation that is developed and
+     "The Open MPI Project is an MPI-3 implementation that is developed and
 maintained by a consortium of academic, research, and industry partners.  Open
 MPI is therefore able to combine the expertise, technologies, and resources
 from all across the High Performance Computing community in order to build the
-- 
2.11.0

[bug#27806] openmpi changes

[Ludovic Courtès]

Hi Dave,

I adjusted the commit log and applied the first two patches.

Dave Love <fx@gnu.org> skribis:

> I assume the only one of these that's questionable is the SGE one.  It
> used to be the default, and I don't know of any way it could be
> problematic -- the support just looks at the environment to determine
> whether to use qrsh for remote startup.  Debian and Fedora both enable
> it.

To be clear: if there’s no environment variable suggesting that SGE is
being used, then it won’t try to use qrsh, right?

If so that seems reasonable to me.  I’ll push the patch if there are no
objections.

Thank you,
Ludo’.

[bug#27806] openmpi changes

[Dave Love]

Ludovic Courtès <ludo@gnu.org> writes:

> Hi Dave,
>
> I adjusted the commit log and applied the first two patches.
>
> Dave Love <fx@gnu.org> skribis:
>
>> I assume the only one of these that's questionable is the SGE one.  It
>> used to be the default, and I don't know of any way it could be
>> problematic -- the support just looks at the environment to determine
>> whether to use qrsh for remote startup.  Debian and Fedora both enable
>> it.
>
> To be clear: if there’s no environment variable suggesting that SGE is
> being used, then it won’t try to use qrsh, right?

Yes, and that flag is long-established in other distributions which add
more intrusive support for resource managers.

bug#27806: openmpi changes

[Ludovic Courtès]

Dave Love <fx@gnu.org> skribis:

> Ludovic Courtès <ludo@gnu.org> writes:
>
>> Hi Dave,
>>
>> I adjusted the commit log and applied the first two patches.
>>
>> Dave Love <fx@gnu.org> skribis:
>>
>>> I assume the only one of these that's questionable is the SGE one.  It
>>> used to be the default, and I don't know of any way it could be
>>> problematic -- the support just looks at the environment to determine
>>> whether to use qrsh for remote startup.  Debian and Fedora both enable
>>> it.
>>
>> To be clear: if there’s no environment variable suggesting that SGE is
>> being used, then it won’t try to use qrsh, right?
>
> Yes, and that flag is long-established in other distributions which add
> more intrusive support for resource managers.

Sounds good, applied.

Thank you!

Ludo’.