I ran recon-all:
$ recon-all -sd /share/users/mfakhri/FS/subjects -subject 102109 -i T1w.nii.gz -all
Then I tried segmentation fornix and I used -f on the mri_cc command:
$ mri_cc -f -force -sdir /share/users/mfakhri/FS/subjects 102109
including fornix in segmentation
processing regardless of existence of cc in input volume
using /share/users/mfakhri/freeSurfer-analys/subjects as SUBJECTS_DIR...
reading aseg from /share/users/mfakhri/freeSurfer-analys/subjects/102109/mri/aseg.mgz
reading norm from /share/users/mfakhri/freeSurfer-analys/subjects/102109/mri/norm.mgz
128569 voxels in left wm, 173189 in right wm, xrange [111, 140]
MRIprincipleComponents: unsupported input type 1
searching rotation angles z=[-7 7], y=[-7 7]
searching scale 1 Z rot 5.8 global minimum found at slice 128.0, rotations (-0.50, -1.00)
final transformation (x=128.0, yr=-0.500, zr=-1.000):
0.99981 0.01745 -0.00873 0.02437;
-0.01745 0.99985 0.00015 130.23384;
0.00873 -0.00000 0.99996 126.88302;
0.00000 0.00000 0.00000 1.00000;
mri_cc: no WM voxels found with norm > 40 -- check skull stripping
There was no fornix.mgz in mri directory and this error file has been created:
$ cat error.log
MRIprincipleComponents: unsupported input type 1
I used an HCP 3T preprocess image.