Dear FreeSurfer developers, 

Apologies for emailing once again about how to resolve the recon-all stopping exactly at the fill step. I understand this may have to do with specific issues on UCSD's computer cluster, but if UCSD's cluster is not having the most up-to-date version of the FreeSurfer module, would it make a difference if I asked UCSD's supercomputer team to update the module version, so that the entire processing stream may be able to run all the way for 1 subject? Would it also make a difference if I increased the number of cores for 1 subject from 1 to 2? 

Sincerely,

Yilei