I've verified that the same problem occurs with 4.1.6, but other than
that, I haven't got anything to go on.
RAH
On 7 July 2010 18:04, Rolf Heckemann <[log in to unmask]> wrote:
> I am still running into this trouble when I run fast (output
> appended). I've tried the FSL 4.1.5 binary download as well as a
> version compiled from current sources. Occasionally, it works when I
> on the command line, but when I queue a script for cluster execution,
> it always aborts after reading the input image. Could this be an
> incompatibility with the kernel? It is a bit dated (2.6.18 in a RHEL5
> installation). Cluster admins advise to check the environment, but the
> FSL initialization is identical between the situation where it works
> and where it doesn't. Does "std::logic_error" in the error message
> indicate a C library problem? What other information might be relevant
> for solving this problem? Any hint appreciated.
>
> Thanks
>
> Rolf Heckemann
>
> -- job log excerpt --
> Starting Single Image Segmentation
> T1-weighted image
> Imagesize : 192 x 192 x 160
> Pixelsize : 1.24982 x 1.24668 x 1.19789
>
> terminate called after throwing an instance of 'std::logic_error'
> what(): basic_string::_S_construct NULL not valid
> /sps/cermep/soundray/adni/scripts/pipeline.sh: line 48: 19184 Aborted
> fast -v -N -o fast mricvpre
>
> -- log excerpt end --
>
|