[Paraview] Has Paraview been ported to BlueGene/L,P or Q?

[David E DeMarle]

I just pushed a working BlueGeneQ config to ParaViewSuperbuild for review:
http://review.source.kitware.com/#/t/3393/

I you check it out (either from the link above or in a few days from
ParaViewSuperBuild directly) you should be able to build on Mira. On other
BlueGeneQ's you will need to change compiler and library paths to match
your machine's setup. Once you do get it working on your machine, please
post your config to the Paraview developers mailing list so that we can
commit it as a machine specific variant of this one.

My tools config script:
soft add +mpiwrapper-xl
soft add +cmake-2.8.11.1
cmake \
 -D ParaView_FROM_GIT:BOOL=OFF \
 -D cross_target:STRING=bgq_gnu \
 -D CROSS_BUILD_STAGE:STRING=TOOLS \
 -D ENABLE_paraview:BOOL=ON \
 -D ENABLE_boost:BOOL=ON \
 -D ENABLE_python:BOOL=ON \
../ParaViewSuperbuild
make

My cross config script:
soft add +mpiwrapper-gcc
soft add +cmake-2.8.11.1
cmake \
 -D ParaView_FROM_GIT:BOOL=OFF \
 -D cross_target:STRING=bgq_gnu \
 -D CROSS_BUILD_STAGE:STRING=CROSS \
 -D ENABLE_paraview:BOOL=ON \
 -D ENABLE_python:BOOL=ON \
 ../ParaViewSuperbuild
make

Note there is also an XLC compiler target (bgq_xlc), that I could not quite
get to work. It fails at the link stage. Bonus points will be given to
whoever fixes the TOC link error to fix it.


David E DeMarle
Kitware, Inc.
R&D Engineer
21 Corporate Drive
Clifton Park, NY 12065-8662
Phone: 518-881-4909


On Thu, Sep 12, 2013 at 10:56 AM, David E DeMarle
<dave.demarle at kitware.com>wrote:

> Short story:
>
> Not done yet. It has been taking more patience than I expected. I expect
> to have it early next week.
>
> Long story:
>
> I've had a nearly working bgq_xlc target for mira since last Thursday.
>
> Unfortunately it fails at final link time of the executables with this
> message:
>
> /bgsys/drivers/ppcfloor/gnu-linux/powerpc64-bgq-linux/bin/ld:
> /bgsys/drivers/V1R2M0/ppc64/comm/sys/lib/libpami.a(component-pami--CAPI.cnk.o)(.text._ZNSt8_Rb_treeIjSt4pairIKjjESt10_Select1stIS2_ESt4lessIjESaIS2_EE8_M_eraseEPSt13_Rb_tree_nodeIS2_E[_ZNSt8_Rb_treeIjSt4pairIKjjESt10_Select1stIS2_ESt4lessIjESaIS2_EE8_M_eraseEPSt13_Rb_tree_nodeIS2_E]+0xbc):
> sibling call optimization to
> `_ZNSt8_Rb_treeIjSt4pairIKjjESt10_Select1stIS2_ESt4lessIjESaIS2_EE8_M_eraseEPSt13_Rb_tree_nodeIS2_E'
> does not allow automatic multiple TOCs; recompile with -mminimal-toc or
> -fno-optimize-sibling-calls, or make
> `_ZNSt8_Rb_treeIjSt4pairIKjjESt10_Select1stIS2_ESt4lessIjESaIS2_EE8_M_eraseEPSt13_Rb_tree_nodeIS2_E'
> extern
>
> Which would seem to indicate that the global symbol table of contents is
> overflowing.
> See:
>  https://trac.mcs.anl.gov/projects/ITAPS/wiki/BuildingOnBgq
> This was never an issue on intrepid and I have not yet found a combination
> of compiler flags that gets around the issue.
>
> Since I'm stalled on that front I've made a bgq_gnu target for comparison
> and in the hopes that the different compiler flags available might resolve
> it. So far that target will build mesa and python fully but can not yet
> compile ParaView.
>
>
> On Sep 12, 2013 1:19 AM, "Benson Muite" <benson_muite at yahoo.com> wrote:
>
> Please let us know if you have finished porting Paraview to blue gene q?
>>
>> Thanks,
>> Benson
>> _______________________________________________
>> Powered by 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
>>
>> Follow this link to subscribe/unsubscribe:
>> http://www.paraview.org/mailman/listinfo/paraview
>>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.paraview.org/pipermail/paraview/attachments/20131007/56c3773a/attachment.htm>