 |
 |
Dear Satrajit & list, Many thanks for your reply. Yes, it is odd. No, I'm not using NFS - everything is stored locally on an EXT3 formatted SATA disk. To be honest, I wouldn't put my hand on my heart and swear that it crashed "whilst writing to the disk". It might have crashed just before it started writing. I will tinker a bit more tomorrow and see if I can figure out what's going on. I'm no Matlab expert... can anyone suggest any particular functions that I should try to execute? I'll try to get hold of ICC 8.1 (or whatever the latest version is) and try compiling my mex files with that. Just to check that I'm setting OMP_NUM_THREADS correctly... I've written a little three-line batchscript which reads: export OMP_NUM_THREADS=2 echo Using $OMP_NUM_THREADS threads matlab Is that the correct procedure (I'm using BASH)? Many thanks, Jack Satrajit Ghosh wrote: >That's rather interesting and intriguing at the same time. I don't >have Matlab R14 SP2 yet, but I have "successfully" run an spm analysis >with OMP_NUM_THREADS=4 on a dual Xeon P4 machine. Not using the >spm_dev version though (I have it installed but haven't had the time >to check it out), but we do have our mex files compiled with icc 8.1. > >The intriguing part is that it crashed while writing. MKL should not >be used for writing files. By any chance are you writing over NFS? > >Satra > > > > >