ICE with 15.0.2

[IanH]

Oh the trouble I had - when compiling the attached the compiler went bad, which made me all sad.  If it got fixed I would be glad, so help me, before I go mad.

>ifort /c /check:all /warn:all /standard-semantics /stand slippery-little-so-and-so.f90
Intel(R) Visual Fortran Intel(R) 64 Compiler XE for applications running on Intel(R) 64, Version 15.0.2.179 Build 20150121
Copyright (C) 1985-2015 Intel Corporation.  All rights reserved.

slippery-little-so-and-so.f90(38): warning #6178: The return value of this FUNCTION has not been defined.   [WHAT]
  FUNCTION ts_What() RESULT(what)
----------------------------^
...
slippery-little-so-and-so.f90(76): remark #7712: This variable has not been used.   [NAMES]
  SUBROUTINE ParseDataComponentDefStmt(decl_spec, attrs, names)
---------------------------------------------------------^
fortcom: Fatal: There has been an internal compiler error (C0000005).
compilation aborted for slippery-little-so-and-so.f90 (code 1)

[Steven_L_Intel1]

What a rad lad! I hope such use of features is a passing fad. The compiler is certainly being a cad here. I will have to speak sternly to its Dad (and Mum.)

[Steven_L_Intel1]

Issue ID is DPD200366616. Removing all of the type-bound procedures did not change the behavior, but some other changes did, such as removing INTENT(OUT) from the "names" argument to the subroutine.

[IanH]

Steve Lionel (Intel) wrote:

...Removing all of the type-bound procedures did not change the behavior, but some other changes did, such as removing INTENT(OUT) from the "names" argument to the subroutine.

Ye gad... a workaround I now had!

[mecej4]

Did not know Fortran bugs could cause one to wax poetic...in this spirit, please enjoy the fine poem (with equations for appendix) by W.J.M. Rankine, http://www.eoht.info/page/The+Mathematician+in+Love .

[Steven_L_Intel1]

Fixed for an update later this year. (Sorry this doesn't rhyme.)