What steps will reproduce the problem?
1. Running a particular sphere, files included, causes an NaN at iteration 97
every time.
What is the expected output? What do you see instead?
All wavelengths run (200nm to 1000nm) except 514nm
What version of the product are you using? On what operating system?
7.1.0 [10.03.03] on Fedora core 15. Execution fails on a linux cluster running
scientific redhat as well as core I7 laptop.
Please provide any additional information below.
Files attached. Wavelengths might say STARTLEN ENDLEN NUMWAVE in the par file,
but those are variables so qsub will run different wavelengths, inputting 0.514
directly also causes it to fail. Main executable is compiled on Ifort with MKL
only, no MPI or open-MP.
Original issue reported on code.google.com by dara...@gmail.com on 26 Aug 2011 at 7:46
Original issue reported on code.google.com by
dara...@gmail.comon 26 Aug 2011 at 7:46Attachments: