[Scilab-users] Segfault with inv, det...

Hello,

   I'm still working on scilab-5.5.2 (I will update my Ubuntu but I
would like to wait a bit. It's not a good time for me!).

--> det(zeros(10,10))

produces Segmentation fault. There is no problem with lower matrix size,
and there is no problem at all if I use "scilab -nwni" and higher matrix
sizes.

"scilab -debug" gives

Thread 1 "scilab-bin" received signal SIGSEGV, Segmentation fault.
[Switching to Thread 0x7ffff7f9f780 (LWP 14031)]
0x00007fffe8c9a343 in dgetrf_parallel () from /usr/lib/libopenblas.so.0

before stopping, but I don't know if it is relevant because there is not
problem with "scilab -nwni"!

Is there a simple way to solve this segfault, still using scilab-5.5.2?

Many thanks,

--Jean-Yves
_______________________________________________
users mailing list
[hidden email]
http://lists.scilab.org/mailman/listinfo/users

Hi,

----- Original Message -----
> From: "Antoine Monmayrant" <[hidden email]>
>> produces Segmentation fault.
> Ouch, that's bad.

"I don't you make it to tell" (sorry, it's franglais)

> So it only occurs for a specific size of 10x10 and with java graphical
> interface enabled?

No, that appends for all size equal or more than 10x10. There is not problem with lower size.

> (1) Can you split your code into one part that does the heavy-lifting
> (ran using "scilab -nwni") and another part that does the gui interface
> (plots, uicontrols ..., ran using "scilab") ?

Yes I can do that using save and load, but my code is not finish and to debug it I overuse plot...

> (2) Can you install a newer scilab version? Which ubuntu are you using?

Hum... 16.04. I know, I have to update. About Scilab, I also used 6.1 but not on my computer. (One day I should "rationalize" my numerical tools.) I wonder if the segfault comes from multiple CPU use, but I don't know how to check that.

Thanks,

--Jean-Yves
_______________________________________________
users mailing list
[hidden email]
http://lists.scilab.org/mailman/listinfo/users