segmentation fault when running HadCEM

[ggjhtw]

Hi there

I am trying to run HadCEM, the high res ocean coupled GCM at Bristol. I can compile successfully but my run is refusing to compile, it falls over at timestep 1 with a seg fault. I have attached the lates output (.leave) file. I am using a local compile override to increase the buffer size as recommended here http://cms.ncas.ac.uk/ticket/248

As far as I can tell there is no useful information in the output file I have attached and so I would be very grateful to hear of any tips you may have. I have already tried increasing the number of processors used but this does not fix the issue.

Many thanks

Jonny Williams

[ggjhtw]

latest output file from TDDLD

[grenville]

Jonny

We are struggling to provide any help - have you been in contact with Gethin, he is better placed to assist you with a problem ob the Bristol machine.

Grenville

[ggjhtw]

Thanks for your message

I have been in touch with Gethin and he has suggested running with the "-g" and "-traceback" compiler flags but this still has not brought up any more useful information. I will be seeing Gethin about this next week so hopefully things will start to make sense soon.

Thanks

Jonny

[annette]

Hi Jonny,

I'm assuming you got help from Gethin so I'm closing this ticket.

Annette