LI75466: XLF COMBINATION OF OPTIONS CAUSES CORE DUMP

May 2010 Update for XL Fortran for Linux, V12.1
January 2012 Update for XL Fortran for Linux, V12.1
January 2011 Update for XL Fortran for Linux, V12.1
June 2011 Update for XL Fortran for Linux, V12.1
XL Fortran for Linux Fix Pack 8 (January 2014 Update) for 12.1

APAR status

• Closed as program error.

Error description

• The following error occurs at runtime while
  using "-qrealsize=8 -qintsize=8 -qflttrap=en:nanq -qsigtrap
  -qarch=auto -qlist -O2" flags during compilation of test.f
  source file.
  
  $cat test.f
    real,allocatable :: r(:,:,:)
    integer,allocatable :: i(:,:,:)
  
    allocate(r(1,1,1))
    allocate(i(1,1,1))
  
    i(:,:,:)=-1
  
    r(:,:,:)=-1.0
  
    r(:,:,:)=r(:,:,:)-real(i(:,:,:))
  
    write (0,*) "i,r",i,r
  
    deallocate(r)
    deallocate(i)
  
    end
  

Local fix

• Do not use optimization (-O2) or -qarch=auto flags
  

Problem summary

• -qflttrap=nanq trap on conversion of integer to floating point
  when integer value resembles a NaN
  

Problem conclusion

• A minor optimization improving conversion of integer to
  floating point will be suppressed when -qflttrap=nanq is used.
  In that situation the performance impact will be negligible.
  

Temporary fix

Comments

APAR Information

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

  IZ69967

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

Fix information

• Fixed component name

  XL FORTRAN RHEL

• Fixed component ID

  5724U8410

Applicable component levels

• RC10 PSY

     UP