[GRASS-user] m.eigensystem segfaults under Jaunty (beta)

[Nikos Alexandris]

Nikos:
> > On Mon, 2009-04-13 at 13:52 +0200, Nikos Alexandris wrote:
> >> m.eigensystem, compiled against latest grass6_dev _AND_ grass-6.4.0RC4
> >> source code, segfaults under Ubuntu-J 64-bit Beta.
> >
> > Example with spearfish landsat data below (tried with other data as
> > well):
> >
> > # launch fresh-baked grass6_dev
> > grass65 /geo/grassdb/nc_spm_08/PERMANENT/
> >
> >
> > # one-liner using 3 bands
> > (echo 3; r.covar  lsat7_2002_10,lsat7_2002_20,lsat7_2002_30) |
> > m.eigensystem
> >
> > r.covar: complete ...
> >  100%
> > Segmentation fault (core dumped)


Markus:
> Here's my take:
> 
> GRASS 6.5.svn (nc_spm_07):~ > (echo 3; r.covar
> lsat7_2002_10,lsat7_2002_20,lsat7_2002_30) | m.eigensystem
> r.covar: complete ...
>  100%
> ERROR: No non-null values
>  Incomplete input file
> 
> BUT:
> 
> GRASS 6.5.svn (nc_spm_07):~ > g.region rast=lsat7_2002_10
> GRASS 6.5.svn (nc_spm_07):~ > (echo 3; r.covar
> lsat7_2002_10,lsat7_2002_20,lsat7_2002_30) | m.eigensystem
> r.covar: complete ...
>  100%
> E      2134.5009768265         0.0000000000    98.07
> V         0.5053399527         0.0000000000
> V         0.5791606282         0.0000000000
> V         0.8626950892         0.0000000000
> ...

> Does it work for you if you set the region to the lsat7_2002_10 map?
> If yes, then m.eigensystem "just" suffers from not trapping bad input.

Unfortunately it does not work :-(.

# although the region is already set correctly...
g.region rast=lsat7_2002_10 # or g.region rast=lsat7_2002_10 -a
(echo 3; r.covar  lsat7_2002_10,lsat7_2002_20,lsat7_2002_30) |
m.eigensystem

[...]
Segmentation fault (core dumped)

Nikos

P.S. I just remind that I work with Jaunty (devel branch).
Brainstorming:
* Maybe it's Ubuntu's fault?
* Maybe in my new system I forgot to add proper entries in ld.so.conf?
(BTW, could some-Ubuntero post his osgeo-relevant entries in
ld.so.conf?)