catastrophic error: **Internal compiler error: internal abort** with 11.1.051 Fortran

[cteaton77]

Hello we have recently gone from the Intel 9.0 Fortran compiler to the 11.1.051 version. We are having a problem with 4 of our 1,000 of so files.

I have narrowed the code down to a particular function that seems to cause the problem. This is just one example, but hopefully it is representative of what is breaking in the other files.

We can get rid of the problem by disabling Process OpenMP directives.

Any suggestions on what we can do to fix this?

thanks,

Chris

/nologo /Qopenmp /iface:cvf /module:"Release/" /object:"Release/" /libs:static /threads /c

1>------ Build started: Project: pbplus, Configuration: Release Win32 ------
1>Compiling with Intel Visual Fortran 11.1.051 [IA-32]...
1>bcs3.f90
1>C:\trees\Softdev\Source\PBplus\bcs3.f90(271): warning #6075: The data type of the actual argument does not match the definition. [P1]
1>C:\trees\Softdev\Source\PBplus\bcs3.f90(271): warning #6075: The data type of the actual argument does not match the definition. [HS1]
1>C:\trees\Softdev\Source\PBplus\bcs3.f90(271): warning #6075: The data type of the actual argument does not match the definition. [P1]
1>C:\trees\Softdev\Source\PBplus\bcs3.f90(271): warning #6075: The data type of the actual argument does not match the definition. [HS1]
1>C:\trees\Softdev\Source\PBplus\bcs3.f90(271): warning #6075: The data type of the actual argument does not match the definition. [GA1]
1>0_1855
1>: catastrophic error: **Internal compiler error: internal abort** Please report this error along with the circumstances in which it occurred in a Software Problem Report. Note: File and line given may not be explicit cause of this error.
1>in file C:\trees\Softdev\Source\PBplus\bcs3.f90, line 10, column 12
1>compilation aborted for C:\trees\Softdev\Source\PBplus\bcs3.f90 (code 3)
1>
1>Build log written to "file://C:\trees\Softdev\tools\build\output\pbplusr.htm"
1>pbplus - 1 error(s), 5 warning(s)
========== Build: 0 succeeded, 1 failed, 0 up-to-date, 0 skipped ==========

[TimP]

We would need the source for the modules to look at this. As with any internal error, you should follow the advice and file a problem report on premier.intel.com or here.
I would hesitate myself to attempt the combination of /iface:cvf and /Qopenmp, but the first thing to look at would be the reported data type disagreements.