[Paraview] Fwd: User trying to build with catalyst -- need guidance
Cook, Rich
cook47 at llnl.gov
Thu Feb 12 14:36:37 EST 2015
Dear Paraview list,
Got this response from Todd Gamblin. He asked me to share it with the list.
Note the two attachments — hopefully they are not too large for the list.
— Rich
Begin forwarded message:
From: Todd Gamblin <tgamblin at llnl.gov<mailto:tgamblin at llnl.gov>>
Subject: Re: [Paraview] User trying to build with catalyst -- need guidance
Date: February 12, 2015 at 11:16:28 AM PST
To: "Cook, Rich" <cook47 at llnl.gov<mailto:cook47 at llnl.gov>>, David E DeMarle <dave.demarle at kitware.com<mailto:dave.demarle at kitware.com>>
Cc: "tim.gallagher at gatech.edu<mailto:tim.gallagher at gatech.edu>" <tim.gallagher at gatech.edu<mailto:tim.gallagher at gatech.edu>>, ParaView list <paraview at paraview.org<mailto:paraview at paraview.org>>
Rich,
The platform files are meant to work with CMake's cross-compile support. That is similar to the build instructions here:
http://www.visitusers.org/index.php?title=Building_on_BlueGeneP#Example_toolchain_file
In fact I wrote the BlueGene/P support for Cmake, and I contributed it back. I have not contributed the BG/Q files back because I don't feel like I've tested them adequately. They're used by several projects though, so maybe I should just thrown them over the fence to kitware already.
There are TWO platforms you can use in your toolchain file (more on that at http://www.vtk.org/Wiki/CMake_Cross_Compiling). They are:
BlueGeneQ-static
BlueGeneQ-dynamic
There are two because a lot of people like to build backend binaries as full static executables. CMake does not have good support for this, in that it will link an executable exactly one way per platform. If you choose BlueGeneQ-static, you get a platform that has no notion of dynamic libraries and that will link executables fully static. If you use BlueGeneQ-dynamic, you get a system that is much like a standard Linux install: it supports static and dynamic libraries, but it will link *executables* as dynamic executables. This is probably the one that will be most familiar if you are used to CMake on Linux.
For anyone interested, the annoyances stem from the fact that on most platforms, dynamic executables are the default, and you supply a flag (-static, -full-static, etc) to make your exe static. On BlueGene, all is reversed, the default is full static, and you supply -dynamic (gcc) or -qnostaticlink (xlc). CMake knows nothing of this w/o these platform files.
Anyway, that's the deal. If you need help using these, feel free to email me. I attached some example toolchain files that I use to build dynamically on BG/Q. These are more complicated maybe than what you need: all you really need to do in a toolchain file is set CMAKE_SYSTEM_NAME and the three CMAKE_<C|CXX|Fortran>_COMPILER variables. My toolchains ensure that the discovered MPI compiler is consistent with the regular compiler.
-Todd
From: <Cook>, Rich <cook47 at llnl.gov<mailto:cook47 at llnl.gov>>
Date: Thursday, February 12, 2015 at 11:06 AM
To: David E DeMarle <dave.demarle at kitware.com<mailto:dave.demarle at kitware.com>>
Cc: "tim.gallagher at gatech.edu<mailto:tim.gallagher at gatech.edu>" <tim.gallagher at gatech.edu<mailto:tim.gallagher at gatech.edu>>, ParaView list <paraview at paraview.org<mailto:paraview at paraview.org>>, Todd Gamblin <tgamblin at llnl.gov<mailto:tgamblin at llnl.gov>>
Subject: Re: [Paraview] User trying to build with catalyst -- need guidance
Thanks for your message.
I’ve poked around a bit. I have built Catalyst successfully on “normal” Linux and our BG/Q machine front end nodes. I’m ready to try a cross compile. I came across this and was wondering if it seems right to you:
http://www.vtk.org/Wiki/CmakeBlueGene
We have a build of cmake that includes some extra files I don’t really understand, provided for me by our local wizard Todd Gamblin. How could I use these files to good effect? :-) It’s probably a dumb question, but I ask it to try to move in the right direction.
rcook at rzuseqlac2 (bgqos_0 ): cp cmake-3.0-bgq-experimental/share/cmake-3.0/Modules/Platform/BlueGene
BlueGeneL.cmake BlueGeneP-static-GNU-Fortran.cmake BlueGeneQ-dynamic-XL-CXX.cmake
BlueGeneP-base.cmake BlueGeneP-static-XL-C.cmake BlueGeneQ-dynamic-XL-Fortran.cmake
BlueGeneP-dynamic-GNU-C.cmake BlueGeneP-static-XL-CXX.cmake BlueGeneQ-dynamic.cmake
BlueGeneP-dynamic-GNU-CXX.cmake BlueGeneP-static-XL-Fortran.cmake BlueGeneQ-static-GNU-C.cmake
BlueGeneP-dynamic-GNU-Fortran.cmake BlueGeneP-static.cmake BlueGeneQ-static-GNU-CXX.cmake
BlueGeneP-dynamic-XL-C.cmake BlueGeneQ-base.cmake BlueGeneQ-static-GNU-Fortran.cmake
BlueGeneP-dynamic-XL-CXX.cmake BlueGeneQ-base.cmake~ BlueGeneQ-static-XL-C.cmake
BlueGeneP-dynamic-XL-Fortran.cmake BlueGeneQ-dynamic-GNU-C.cmake BlueGeneQ-static-XL-CXX.cmake
BlueGeneP-dynamic.cmake BlueGeneQ-dynamic-GNU-CXX.cmake BlueGeneQ-static-XL-Fortran.cmake
BlueGeneP-static-GNU-C.cmake BlueGeneQ-dynamic-GNU-Fortran.cmake BlueGeneQ-static.cmake
BlueGeneP-static-GNU-CXX.cmake BlueGeneQ-dynamic-XL-C.cmake
On Feb 7, 2015, at 4:35 AM, David E DeMarle <dave.demarle at kitware.com<mailto:dave.demarle at kitware.com>> wrote:
I've done the BGQ ParaView + Catalyst build and installs on Mira at ANL.
I can chip in with tips regarding compilation.
David E DeMarle
Kitware, Inc.
R&D Engineer
21 Corporate Drive
Clifton Park, NY 12065-8662
Phone: 518-881-4909
On Fri, Feb 6, 2015 at 8:25 PM, Tim Gallagher <tim.gallagher at gatech.edu<mailto:tim.gallagher at gatech.edu>> wrote:
Hi Rich,
There may be a few who are willing to help out if you can provide some specific questions or issues you run into.
Have you looked through the documentation at http://www.paraview.org/in-situ/ ? I got Catalyst up and running in our code (not on a BG/Q but on different clusters) with the information in the documentation and by asking some specific questions here on the list. For me, the hardest part was converting our data into VTK data -- I did not have any exposure to VTK prior to taking the Catalyst leap. But it turned out to be (relatively) straight forward for what we are doing.
I will be happy to answer things I can and for things I can't, hopefully somebody else will jump in.
Tim
________________________________
From: "Rich Cook" <cook47 at llnl.gov<mailto:cook47 at llnl.gov>>
To: paraview at paraview.org<mailto:paraview at paraview.org>
Sent: Friday, February 6, 2015 7:23:24 PM
Subject: [Paraview] User trying to build with catalyst -- need guidance
Hello,
We have not used catalyst here and I have a user that is interested in trying it out on our BG/Q machine with a huge code on hundreds of thousands of nodes. ;-)
Is there a brave soul willing to hold our hands through this process?
Thanks.
--
✐Richard Cook
✇ Lawrence Livermore National Laboratory
Bldg-453 Rm-4024, Mail Stop L-557
7000 East Avenue, Livermore, CA, 94550, USA
☎ (office) (925) 423-9605<tel:%28925%29%20423-9605>
☎ (fax) (925) 423-6961<tel:%28925%29%20423-6961>
---
Information Management & Graphics Grp., Services & Development Div., Integrated Computing & Communications Dept.
(opinions expressed herein are mine and not those of LLNL)
_______________________________________________
Powered by www.kitware.com<http://www.kitware.com/>
Visit other Kitware open-source projects at http://www.kitware.com/opensource/opensource.html
Please keep messages on-topic and check the ParaView Wiki at: http://paraview.org/Wiki/ParaView
Search the list archives at: http://markmail.org/search/?q=ParaView
Follow this link to subscribe/unsubscribe:
http://public.kitware.com/mailman/listinfo/paraview
_______________________________________________
Powered by www.kitware.com<http://www.kitware.com/>
Visit other Kitware open-source projects at http://www.kitware.com/opensource/opensource.html
Please keep messages on-topic and check the ParaView Wiki at: http://paraview.org/Wiki/ParaView
Search the list archives at: http://markmail.org/search/?q=ParaView
Follow this link to subscribe/unsubscribe:
http://public.kitware.com/mailman/listinfo/paraview
_______________________________________________
Powered by www.kitware.com<http://www.kitware.com/>
Visit other Kitware open-source projects at http://www.kitware.com/opensource/opensource.html
Please keep messages on-topic and check the ParaView Wiki at: http://paraview.org/Wiki/ParaView
Search the list archives at: http://markmail.org/search/?q=ParaView
Follow this link to subscribe/unsubscribe:
http://public.kitware.com/mailman/listinfo/paraview
--
✐Richard Cook
✇ Lawrence Livermore National Laboratory
Bldg-453 Rm-4024, Mail Stop L-557
7000 East Avenue, Livermore, CA, 94550, USA
☎ (office) (925) 423-9605
☎ (fax) (925) 423-6961
---
Information Management & Graphics Grp., Services & Development Div., Integrated Computing & Communications Dept.
(opinions expressed herein are mine and not those of LLNL)
--
✐Richard Cook
✇ Lawrence Livermore National Laboratory
Bldg-453 Rm-4024, Mail Stop L-557
7000 East Avenue, Livermore, CA, 94550, USA
☎ (office) (925) 423-9605
☎ (fax) (925) 423-6961
---
Information Management & Graphics Grp., Services & Development Div., Integrated Computing & Communications Dept.
(opinions expressed herein are mine and not those of LLNL)
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://public.kitware.com/pipermail/paraview/attachments/20150212/6324a2ca/attachment.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: BlueGeneQ-gnu.cmake
Type: application/octet-stream
Size: 1355 bytes
Desc: BlueGeneQ-gnu.cmake
URL: <http://public.kitware.com/pipermail/paraview/attachments/20150212/6324a2ca/attachment.obj>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: BlueGeneQ-xl.cmake
Type: application/octet-stream
Size: 1542 bytes
Desc: BlueGeneQ-xl.cmake
URL: <http://public.kitware.com/pipermail/paraview/attachments/20150212/6324a2ca/attachment-0001.obj>
More information about the ParaView
mailing list