There is one thing that I know will crash the image sampling routines
(when they are compiled correctly), and that is when the offset into
the file is not a multiple of the number of bytes per voxel. For
example, if your data is short int (2 bytes per voxel), and the offset
is an odd number, then you will get a Segmentation fault.
Otherwise, it looks like there could be a compilation problem. Do you
really need to recompile the .mexlx files that came with the spm99b
distribution? These files definately worked on some Linux systems,
so I think it may be a good idea to try the original ones.
Good luck,
-John
| After following the recent advice on compiling SPM99b for SuSe
| linux (path to gccopt's and new spm_list_files.mexlx I got the
| following output when trying to a previously succesful script:
|
| ****
| Segmentation violation occurred within signal handler.
| Unable to complete stack trace (stack was probably corrupted)
|
...........................................
| Error in ==> /home/jdj/csd/warp.m
| On line 2 ==> G0 = spm_slice_vol(VG,spm_matrix([0 0
| 1]),VG.dim(1:2),1);
| ****
|
| I had the warning og "unsigned...." while compiling mentioned by
| Yasuhiko as well.
|
| Should I just revert to my original installation, - or does
| anyone know what happened.
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|