https://community.intel.com/t5/Intel-Fortran-Compiler/access-violation-with-unlimited-polymorphic-function/m-p/791056#M32664 Yes, that does seem to be the case. "13.0" is the next major release, due out later this year. Tue, 03 Jul 2012 13:18:53 GMT Steven_L_Intel1 2012-07-03T13:18:53Z https://community.intel.com/t5/Intel-Fortran-Compiler/access-violation-with-unlimited-polymorphic-function/m-p/791054#M32662 when I run the following test case, which is the smallest one I can provide,<BR /><BR />[fortran]module mtest contains function to_str(obj) result( s ) character(:),allocatable :: s class(*) :: obj integer :: j character(len=100) :: temp select type(obj) type is (integer(4)) j = obj write(temp, fmt="(i0)") j end select s = trim(temp) end function end module program test use mtest print *, "i:", to_str(5) print *, "i:" // to_str(5) end program [/fortran]<BR />I get an access violation on the second print statement:<BR /><BR />[bash]C:\TEMP&gt;ifort test.f90 Intel Visual Fortran Compiler XE for applications running on IA-32, Version 12.1.4.325 Build 20120410 Copyright (C) 1985-2012 Intel Corporation. All rights reserved. Microsoft Incremental Linker Version 10.00.40219.01 Copyright (C) Microsoft Corporation. All rights reserved. -out:test.exe -subsystem:console test.obj C:\TEMP&gt;test i:5 forrtl: severe (157): Program Exception - access violation Image PC Routine Line Source test.exe 00409E79 Unknown Unknown Unknown test.exe 00401305 Unknown Unknown Unknown test.exe 0044F2A3 Unknown Unknown Unknown test.exe 0043524F Unknown Unknown Unknown kernel32.dll 7C817067 Unknown Unknown Unknown C:\TEMP&gt;[/bash]<BR />If instead of making the result of the function allocatable, I make it a fixed lenght string, there is no problem.<BR /><BR />Johny Tue, 03 Jul 2012 07:13:48 GMT https://community.intel.com/t5/Intel-Fortran-Compiler/access-violation-with-unlimited-polymorphic-function/m-p/791054#M32662 johnyb_ 2012-07-03T07:13:48Z https://community.intel.com/t5/Intel-Fortran-Compiler/access-violation-with-unlimited-polymorphic-function/m-p/791055#M32663 This problem appears to have been corrected in ifort 13.0, although I see it in 12.1.5. Tue, 03 Jul 2012 10:35:22 GMT https://community.intel.com/t5/Intel-Fortran-Compiler/access-violation-with-unlimited-polymorphic-function/m-p/791055#M32663 TimP 2012-07-03T10:35:22Z https://community.intel.com/t5/Intel-Fortran-Compiler/access-violation-with-unlimited-polymorphic-function/m-p/791056#M32664 Yes, that does seem to be the case. "13.0" is the next major release, due out later this year. Tue, 03 Jul 2012 13:18:53 GMT https://community.intel.com/t5/Intel-Fortran-Compiler/access-violation-with-unlimited-polymorphic-function/m-p/791056#M32664 Steven_L_Intel1 2012-07-03T13:18:53Z https://community.intel.com/t5/Intel-Fortran-Compiler/access-violation-with-unlimited-polymorphic-function/m-p/791057#M32665 <P>Tim, Steve,<BR />thanks for your answers.<BR />I can work around the bug for the moment, so I can wait for the next major version. I think our support runs 'til then<BR /><BR />Johny</P> Wed, 04 Jul 2012 14:10:50 GMT https://community.intel.com/t5/Intel-Fortran-Compiler/access-violation-with-unlimited-polymorphic-function/m-p/791057#M32665 johnyb_ 2012-07-04T14:10:50Z