[xcrysden] problem compiling and executing xcrysden

[Tone Kokalj]

On Wed, 2012-03-07 at 15:39 +0100, rafael wrote:
> In C/shank_Hinv.c add:
> 
> #define M_PI 3.14159265358979323846

I know this, the M_PI isn't standard and it is not defined everywhere.
It's been fixed.

> 
> In C/shankland.h and C/shank_util.c the compiler gives errors on those functions :
> extern void print_complex_vector(FILE *stream, const char *name, size_t m, fftw_complex *mat);
> 
> extern void print_complex_matrix(FILE *stream, const char *name, size_t m, size_t n, fftw_complex *mat);
> 
> shank_util.c: In function ‘print_complex_vector’:
> shank_util.c:124: error: subscripted value is neither array nor pointer
> shank_util.c:124: error: subscripted value is neither array nor pointer
> shank_util.c: In function ‘print_complex_matrix’:
> shank_util.c:164: error: subscripted value is neither array nor pointer
> shank_util.c:164: error: subscripted value is neither array nor pointer

This I see the first time. Probably the problem is that in your system
the fftw_complex is not defined as "typedef double fftw_complex[2]" but
as C99's complex double (no experience with it whatsoever). It's best to
delete these functions and their prototypes as they were merely used for
debugging.


> After those changes it compiles, but when executing it gives a segmentation fault:

> I have debugged (-d)  it and it shows:
> 
> Executing: /export/home_users/home/soft/xcrysden/originals/XCrySDen-1.5.24-src-all/bin/ftnunit
> 
> Program received signal SIGSEGV, Segmentation fault.
> [Switching to Thread 47700007662880 (LWP 29800)]
> 0xffffffffffffffff in ?? ()
> (gdb) bt
> #0  0xffffffffffffffff in ?? ()
> #1  0x000000000043df17 in LoadLights ()

This is known issue that has been fixed. You may try to compile newer
version, which should work:
http://www.xcrysden.org/download/xc-1.5.32-src.tar.gz

Regards, Tone
-- 
Anton Kokalj
J. Stefan Institute, Jamova 39, 1000 Ljubljana, Slovenia 
(tel: +386-1-477-3523 // fax:+386-1-477-3822)

Please, if possible, avoid sending me Word or PowerPoint attachments.
See:  http://www.gnu.org/philosophy/no-word-attachments.html