N48 model fails when branch is included

[anmcr]

Hi Willie,

I am trying to create a vn7.3 branch which implements a gravity wave scheme and run this in a vn7.3 N48L60 model.

The branch is: ​https://puma.nerc.ac.uk/trac/UM/log/UM/branches/dev/anmcr/vn7.3_gravity_wave_temperature_parameterization/src

The model run is: xhldy

(I previously successfully implemented this in a vn7.8 branch/model.)

My problem is that the model runs successfully when the branch is not included. However, it fails with a segmentation fault when the branch is included. This is extremely puzzling as I have only introduced a few changes so far to the code. I included the 'flush buffers' option but from what I could determine that didn't help understand the problem. I was wondering whether the 'FCM Options for Atmosphere and Reconfiguration' page in the UMUI was setup wrongly?

I would really appreciate some help.

Thanks,

Andrew

*
UM Executable : /work/n02/n02/anmcr/xhldy/bin/xhldy.exe
*

_pmiu_daemon(SIGCHLD): [NID 02533] [c3-1c0s2n1] [Fri May 31 10:25:30 2013] PE RANK 24 exit signal Segmentation fault [NID 02533] 2013-05-31 10:25:30 Apid 4717765: initiated application termination
diff: /work/n02/n02/anmcr/tmp/tmp.hector-xe6-12.20833/xhldy.xhist: No such file or directory
qsexecute: Copying /work/n02/n02/anmcr/xhldy/xhldy.thist to backup thist file /work/n02/n02/anmcr/xhldy/xhldy.thist_keep
xhldy: Run failed
*

Ending script : qsexecute
Completion code : 137
Completion time : Fri May 31 10:25:33 UTC 2013

*

[anmcr]

Hi Willie,

Please refrain from looking at this problem for the time being. There are a number of uncertainties which I want to clarify myself before involving you.

Thanks,

Andrew