LI77169: COMPILER ERROR IN XLFHOT WITH -QINTSIZE=8

December 2012 Update for XL Fortran for Linux, V14.1
April 2013 Update for XL Fortran for Linux, V14.1

APAR status

• Closed as program error.

Error description

• Compiling a test case involving Fortran arrays and using the
  -qintsize=8 option is encountering an internal compiler error in
  the xlfhot compiler component.
  
  
     =====  TEST CASE  =====
  
    subroutine test( )
      real(4) :: a(0:6)
      real(4) :: b(3)
      real(4) :: c(2)
      real(4) :: d(2,2)
      integer :: l
  
      c(:) = MATMUL(d(:,:),b(1:2))
      l = minloc(a, MASK=a>1.0e-3, DIM=1)
  
    end subroutine test
  
  
  > xlf90_r -qintsize=8 -c test.F90
  ** test   === End of Compilation 1 ===
  1585-100  Internal compiler error -- please contact your service
  representative. For more information visit:
  http://www.ibm.com/support/docview.wss?uid=swg21110810
  1501-511  Compilation failed for file apar.F90.
  >
  

Local fix

• - Compile with -qintsize=4
  

Problem summary

• PROBLEM DESCRIPTION:
  Internal compiler error in the xlfhot compiler component when
  calling maxloc/minloc.
  
  USERS AFFECTED:
  Users who call intrinsic subroutines MAXLOC or MINLOC.
  

Problem conclusion

• The issue was corrected.
  

Temporary fix

Comments

APAR Information

• APAR is sysrouted FROM one or more of the following:

  LI77023

• APAR is sysrouted TO one or more of the following:

Fix information

• Fixed component name

  XL FORTRAN LINU

• Fixed component ID

  5725C7500

Applicable component levels

• RE10 PSN LI77023

     UP06/09/13