Re: [deal.II] Re: 'make test' runs wrong

2022-11-11 Thread Marc Fehling 
Hello Meng,

when you terminal exits at 57%, you must have encountered some compilation 
problem. The original problem you posted was a runtime error while testing, 
and for this you compiled the library. Let me ask you a couple of questions:

   - What exactly is the error message you encountered while compiling your 
   code?
   - How did you configure deal.II, i.e., how did the cmake command look 
   like? Exactly like in the 'linux-debug-intel-oneapi' CI stage?
   - How did you install the Intel oneAPI suite?

Marc

On Friday, November 11, 2022 at 7:27:23 AM UTC-7 dengm...@gmail.com wrote:

>   Dear Marc:
>  I have tried the 'linux-debug-intel-oneapi' commands under the 
> file '.github/workflows/linux.yml' at 
> https://github.com/dealii/dealii/pull/14145/files, but I still can not 
> handle this problem. I can not even pass the 'make -j 2' command (the 
> terminal exit automatically when progress reaches 57%). Is it better to 
> create a new virtual box environment without intel mkl and mpi lib?
>   
>  Best regards
>  Meng 
>
> Marc Fehling  于2022年11月10日周四 05:15写道：
>
>> Hello,
>>
>> at some point, the quicktests failed on our ubuntu CI when using Intel 
>> OneAPI in deal.ii 9.4.0, see also 
>> https://github.com/dealii/dealii/issues/13477
>>
>> Do you face the same problem when working on an up-to-date master branch? 
>> We fixed it at some point, see 
>> https://github.com/dealii/dealii/pull/14145
>>
>> Best,
>> Marc
>>
>> On Monday, November 7, 2022 at 10:57:13 PM UTC-7 dengm...@gmail.com 
>> wrote:
>>
>>> when I execute ' make test ' after ' make --job=3 install ' in my VBOX 
>>> ubuntu , it sends wrong massage:
>>> umfpack.debug: RUN failed. Output:
>>> /home/dm/Desktop/deal.II/dealii-9.4.0/build/tests/quick_tests/umfpack.debug:
>>>  
>>> symbol lookup error: 
>>> /opt/intel/oneapi/mkl/2022.2.0/lib/intel64/libmkl_intel_lp64.so.2: 
>>> undefined symbol: mkl_blas_dgemm
>>> (I have installed the Intel oneapi library)
>>> i do not know how to slove this, it's really hard for a  green hand. 
>>>
>> -- 
>> The deal.II project is located at http://www.dealii.org/
>> For mailing list/forum options, see 
>> https://groups.google.com/d/forum/dealii?hl=en
>> --- 
>> You received this message because you are subscribed to the Google Groups 
>> "deal.II User Group" group.
>> To unsubscribe from this group and stop receiving emails from it, send an 
>> email to dealii+un...@googlegroups.com.
>> To view this discussion on the web visit 
>> https://groups.google.com/d/msgid/dealii/061fc36d-1c57-49c3-8b9c-092e41a5bd43n%40googlegroups.com
>>  
>> 
>> .
>>
>

-- 
The deal.II project is located at http://www.dealii.org/
For mailing list/forum options, see 
https://groups.google.com/d/forum/dealii?hl=en
--- 
You received this message because you are subscribed to the Google Groups 
"deal.II User Group" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to dealii+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/dealii/66e0c9bb-0717-45c1-90ef-aba4fdef23e2n%40googlegroups.com.

Re: [deal.II] SEGMENTATION FAULT(CORE DUMPED)

2022-11-11 Thread Wolfgang Bangerth 

On 11/11/22 09:34, Alberto SALVADORI wrote:


as far as I understand, your problem is not in compilation. In fact, you 
seamlessly compiled step-1 and step-2, both in debug and release mode.
Your "segmentation fault" issue arises in running the executable file from the 
command line.
That said, I have no idea on what is causing the issue. Perhaps someone in the 
community has had similar problems and can address us.


Does this happen with the unmodified step-1 and step-2 programs?

In any case, you will have to find out more information. The program seems to 
run for the majority of us, so we don't really know what is specific about 
your situation to help. Can you run the program in a debugger and obtain a 
backtrace?


Best
 Wolfgang

--

Wolfgang Bangerth  email: bange...@colostate.edu
   www: http://www.math.colostate.edu/~bangerth/


--
The deal.II project is located at http://www.dealii.org/
For mailing list/forum options, see 
https://groups.google.com/d/forum/dealii?hl=en
--- 
You received this message because you are subscribed to the Google Groups "deal.II User Group" group.

To unsubscribe from this group and stop receiving emails from it, send an email 
to dealii+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/dealii/e39e9fbe-d600-c7d9-a44f-6467ed2da53b%40colostate.edu.

Re: [deal.II] SEGMENTATION FAULT(CORE DUMPED)

2022-11-11 Thread Alberto SALVADORI 
Erkin

as far as I understand, your problem is not in compilation. In fact, you
seamlessly compiled step-1 and step-2, both in debug and release mode.
Your "segmentation fault" issue arises in running the executable file from
the command line.
That said, I have no idea on what is causing the issue. Perhaps someone in
the community has had similar problems and can address us.

dealii-candi 9.4.0 was compiled with gcc 13.3.0

Alberto


*Alberto Salvadori* Dipartimento di Ingegneria Meccanica e Industriale
(DIMI)
 Universita` di Brescia, via Branze 43, 25123 Brescia
 Italy
 tel 030 3715426

e-mail:
 alberto.salvad...@unibs.it
web-page:
 http://m4lab.unibs.it/faculty.html



On Fri, Nov 11, 2022 at 5:21 PM erkin yildiz  wrote:

> Dear everyone,
>
> My name is Erkin Yildiz. I am PhD student at University of Brescia. The
> deal.ii has been installed with candi inside of my computer. I am going to
> share the problem that I have deal with it. After the everything when I
> want to compile the system by using IDE it gives me error which is called
> segmentation faul (core dumped).
>
> What does it mean? I am going to share also the properities of the
> computer.
>
> Best regards,
> Erkin
>
>
> Informativa sulla Privacy: https://www.unibs.it/it/node/1452
>
> --
> The deal.II project is located at http://www.dealii.org/
> For mailing list/forum options, see
> https://groups.google.com/d/forum/dealii?hl=en
> ---
> You received this message because you are subscribed to the Google Groups
> "deal.II User Group" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to dealii+unsubscr...@googlegroups.com.
> To view this discussion on the web visit
> https://groups.google.com/d/msgid/dealii/CAP3xP0raSGJCG8Z72JSZ1QMSx_hUDyHS-c9tF6wHFXQ_Pc-ePg%40mail.gmail.com
> 
> .
>

-- 


Informativa sulla Privacy: https://www.unibs.it/it/node/1452 


-- 
The deal.II project is located at http://www.dealii.org/
For mailing list/forum options, see 
https://groups.google.com/d/forum/dealii?hl=en
--- 
You received this message because you are subscribed to the Google Groups 
"deal.II User Group" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to dealii+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/dealii/CABcATpcRX%3DcpmYHESg0Hrda1FeAsofvS6RAcONNdcwj%3DVLLHXA%40mail.gmail.com.

Re: [deal.II] Re: 'make test' runs wrong

2022-11-11 Thread meng deng 
  Dear Marc:
 I have tried the 'linux-debug-intel-oneapi' commands under the
file '.github/workflows/linux.yml' at
https://github.com/dealii/dealii/pull/14145/files, but I still can not
handle this problem. I can not even pass the 'make -j 2' command (the
terminal exit automatically when progress reaches 57%). Is it better to
create a new virtual box environment without intel mkl and mpi lib?

 Best regards
 Meng

Marc Fehling  于2022年11月10日周四 05:15写道：

> Hello,
>
> at some point, the quicktests failed on our ubuntu CI when using Intel
> OneAPI in deal.ii 9.4.0, see also
> https://github.com/dealii/dealii/issues/13477
>
> Do you face the same problem when working on an up-to-date master branch?
> We fixed it at some point, see https://github.com/dealii/dealii/pull/14145
>
> Best,
> Marc
>
> On Monday, November 7, 2022 at 10:57:13 PM UTC-7 dengm...@gmail.com wrote:
>
>> when I execute ' make test ' after ' make --job=3 install ' in my VBOX
>> ubuntu , it sends wrong massage:
>> umfpack.debug: RUN failed. Output:
>> /home/dm/Desktop/deal.II/dealii-9.4.0/build/tests/quick_tests/umfpack.debug:
>> symbol lookup error:
>> /opt/intel/oneapi/mkl/2022.2.0/lib/intel64/libmkl_intel_lp64.so.2:
>> undefined symbol: mkl_blas_dgemm
>> (I have installed the Intel oneapi library)
>> i do not know how to slove this, it's really hard for a  green hand.
>>
> --
> The deal.II project is located at http://www.dealii.org/
> For mailing list/forum options, see
> https://groups.google.com/d/forum/dealii?hl=en
> ---
> You received this message because you are subscribed to the Google Groups
> "deal.II User Group" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to dealii+unsubscr...@googlegroups.com.
> To view this discussion on the web visit
> https://groups.google.com/d/msgid/dealii/061fc36d-1c57-49c3-8b9c-092e41a5bd43n%40googlegroups.com
> 
> .
>

-- 
The deal.II project is located at http://www.dealii.org/
For mailing list/forum options, see 
https://groups.google.com/d/forum/dealii?hl=en
--- 
You received this message because you are subscribed to the Google Groups 
"deal.II User Group" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to dealii+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/dealii/CAD-ZJzcCoWTWFiyKBNzynhiHiqNH0UdVLZ7BHJmNM3eHcGaN1g%40mail.gmail.com.

Re: [deal.II] 3D Mixed Simplex Mesh

2022-11-11 Thread Alex Quinlan 
Dear Wolfgang,

Thanks very much for your reply.  I'm happy to hear that you think my 
problems can be resolved without too much difficulty. I'd like to try 
writing a patch for the mixed 3D meshes.  I have done some development work 
before, but I have never sent it out of our group; your advice on writing 
the steps to write a patch would be most welcome!

I will re-review the relevant code and see if I can put together a general 
plan on how to approach this. 

Best regards,
Alex

On Wednesday, November 9, 2022 at 4:44:25 PM UTC-5 Wolfgang Bangerth wrote:

> Alex:
>
> > I am coming from a background of solid mechanics FEA and I've been using 
> > CalculiX as an openSource program for some time.  My group has decided 
> > to switch to deal.ii
>
> Good choice -- we applaud you :-)
>
>
> > Part 1:  Abaqus .inp format
> > I had success importing hex meshes from the abaqus <.inp> format, but it 
> > seems that the UCD reader does not accept tet meshes.  I am basing this 
> > on the following bit of code in grid_in::read_ucd around Line 948
> > 
> >   if (((cell_type == "line") && (dim == 1)) ||
> >   ((cell_type == "quad") && (dim == 2)) ||
> >   ((cell_type == "hex") && (dim == 3)))
> > // found a cell
> > {
> > 
> > The absence of
> >((cell_type == "tet") && (dim == 3))
> > suggests that simplex elements are not supported when importing from 
> > .inp.  Am I correct in this conclusion.  How difficult would it be for 
> > me to add this functionality in?
>
> Not very difficult in all likelihood. The key difficulty in adding 
> support for other cells in this function is that it uses the 
> GeometryInfo class, which only provides information about hypercube 
> cells. For example, we read vertex indices like this in that function:
> for (const unsigned int i : GeometryInfo::vertex_indices())
> in >> cells.back().vertices[GeometryInfo::ucd_to_deal[i]];
>
> This needs to be changed to account for the number of vertices a cell 
> has, given the cell type, not the dimension. The right approach will be 
> to use the ReferenceCell objects defined in namespace ReferenceCells 
> (plural), and the other readers in that function will serve as guidance. 
> You might want to take a look at the gmsh reader, for example.
>
>
> > Part 2:  GMSH format
> > While looking at grid_in.cc, I saw that the MSH reader supported tet 
> > elements.  I found the FAQ page supplied these examples of simplex 
> > meshes: 
> https://www.dealii.org/developer/doxygen/deal.II/group__simplex.html
> > 
> > I first ran the 2D simplex mesh example, and then I modified it to work 
> > for 3D without much problem.  Then, I switched to the mixed mesh, and 
> > had success running the 2D example.  However, I started having problems 
> > when I switched to 3D.  The mesh seems to import, but then I have issues 
> > operating with it. I found several places within fe_simplex_p.cc (L714, 
> > L759) where the dimension is asserted to be 2:
> > 
> >AssertDimension(dim, 2);
> > This seems like a problem for a 3D analysis, so I am wondering if I need 
> > to do something different so that this function doesn't get called.
>
> No, I think this is just not implemented. The function is called on 
> mixed meshes when the library wants to know how elements of type 
> FE_SimplexP on tetrahedra connect to another element types on other cell 
> shapes (namely, on wedges and pyramids). The comparison is simply not 
> implemented, but if you want to compute on mixed 3d meshes, you will 
> have to implement the missing parts of this function. As long as you 
> want to use the same polynomial degree on all cells, this should not 
> even be particularly difficult and I would be happy to walk you through 
> the necessary steps if you were interested in writing up a patch!
>
> Best
> Wolfgang
>
> -- 
> 
> Wolfgang Bangerth email: bang...@colostate.edu
> www: http://www.math.colostate.edu/~bangerth/
>

-- 
The deal.II project is located at http://www.dealii.org/
For mailing list/forum options, see 
https://groups.google.com/d/forum/dealii?hl=en
--- 
You received this message because you are subscribed to the Google Groups 
"deal.II User Group" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to dealii+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/dealii/885a1c8c-a301-4467-8d16-9a5d156d6f95n%40googlegroups.com.