[Paraview] Making a package for Nix system

[Mathieu Westphal]

Hi

Without debug symbols, we can't really use the backtrace.

Regarding 5.3 version, are you saying that it does not have the segfault ?

Mathieu Westphal

On Tue, Mar 7, 2017 at 4:29 PM, Yves Rogez <
yves.rogez at univ-grenoble-alpes.fr> wrote:

> Hi Mathieu,
>
> for info, here is the backtrace for the PV5.2 bug with seg fault when
> disconnecting (sorry it was stuck in my draft e-mails since a while :-).
>
> I did not find the time to rebuild with the last version. I will try to do
> so soon... I think I have to rebuild it from source as the basic goal is to
> package it on nix system (which uses internal nix package dependencies). I
> currently use 5.3 RC2 for my developments and it works completely fine on
> any system (not including nix :-).
>
> Bye,
>
> Yves
>
> Le 10/02/2017 à 15:14, Mathieu Westphal a écrit :
>
> Hi Yves
>
> In order to test with the last version, you do not need to compile, you
> can download a nightly build on the ParaView download page.
> (Highest number means newer).
> Let me know how it goes.
>
> I will defer to others regarding the packaging question.
>
>
> Mathieu Westphal
>
> On Fri, Feb 10, 2017 at 9:44 AM, Yves Rogez <yves.rogez at univ-grenoble-
> alpes.fr> wrote:
>
>> Hi Mathieu :-)
>>
>> yes it's for the case of multi-server (actually single client) option. I
>> will get the source from the git repository master a give it another try.
>>
>> OK for the debug, I can compile it again in debug but it takes a long
>> time on my little test laptop so I was wondering maybe some other tricky
>> test on the release version ;-) I will let the debug build run this
>> week-end and send a quick report.
>>
>> By the way, do you plan to support nix packages build in the future for
>> Paraview ? The problem I will encounter now is that Paraview depends on
>> video card driver OpenGL libs so that to get good performances, I will
>> break the nix packages philosophy of being independent from the operating
>> system specific kernel...
>>
>> Thanks for the answer!
>>
>> Yves
>>
>> Le 10/02/2017 à 03:48, Mathieu Westphal a écrit :
>>
>> Hi Yves !
>>
>> Do you happen to use the multi-clients/multi-server option ? we recently
>> corrected a bug that sound like yours, will be fixed in ParaView 5.3.
>> In any case, you may want to try with the last version of paraview master.
>>
>> In order to debug, the first step would be to build ParaView in Debug,
>> and then run it with a debuger, in order to get a backtrace of the segfault.
>>
>> Best Regards,
>>
>> Mathieu Westphal
>>
>> On Thu, Feb 9, 2017 at 6:09 PM, Yves Rogez <yves.rogez at univ-grenoble-alpe
>> s.fr> wrote:
>>
>>> Dear Paraview team,
>>>
>>> we are currently studying the opportunity of migrating our computing
>>> grid environment from "module" system to "nix" packaging system (
>>> https://nixos.org/). In that purpose, I'm trying to compile and pack
>>> Paraview 5.2.0 with its dependencies with nix packages (
>>> https://nixos.org/nixos/packages.html).
>>>
>>> To do so, I first installed nixos on a local laptop and give it a try.
>>> The main issue I encountered was to update my nixos kernel with native
>>> videocard drivers but OK, that's done. Then using the attached nix package
>>> script (a slight evolution on paraview 4.0.1 existing one), I could have
>>> packetized 5.2.0 succesfully and it runs very well.
>>>
>>> However, I still get an error when I'm trying to use MPI functionality.
>>> The MPI itself with parallelism works, as I can run a auto-MPI session
>>> without problem. The issue occurs when *servermanager.Disconnect()
>>> function is called => Seg fault* with no more info... The bad thing is
>>> that the Disconnect function is called every time I would like to connect
>>> to a server (to disconnect the built-in one). This is also the case when
>>> passing the pvserver in command-line argument.
>>>
>>> Do you have an idea of the cause of this seg fault ? Maybe could you
>>> help me find any useful test to debug this ?
>>>
>>> Thanks in advance for your support,
>>>
>>> Best regards,
>>>
>>> Yves
>>>
>>> --
>>> Yves Rogez
>>>
>>> *IPAG*
>>> *Institut de Planétologie et d'Astrophysique de Grenoble *
>>>
>>>
>>> _______________________________________________
>>> 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
>>>
>>> Search the list archives at: http://markmail.org/search/?q=ParaView
>>>
>>> Follow this link to subscribe/unsubscribe:
>>> http://public.kitware.com/mailman/listinfo/paraview
>>>
>>>
>>
>> --
>> Yves Rogez
>>
>> *IPAG*
>> *Institut de Planétologie et d'Astrophysique de Grenoble *
>> Bat D de Physique - BP. 53 - 38041 Grenoble - FRANCE
>>
>> tel : +33 (0)4 76 63 52 80 <+33%204%2076%2063%2052%2080>
>> lab : +33 (0)4 76 63 57 60 <+33%204%2076%2063%2057%2060>
>>
>
>
> --
> Yves Rogez
>
> *IPAG*
> *Institut de Planétologie et d'Astrophysique de Grenoble *
> Bat D de Physique - BP. 53 - 38041 Grenoble - FRANCE
>
> tel : +33 (0)4 76 63 52 80 <+33%204%2076%2063%2052%2080>
> lab : +33 (0)4 76 63 57 60 <+33%204%2076%2063%2057%2060>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://public.kitware.com/pipermail/paraview/attachments/20170307/245d2e53/attachment-0001.html>