Re: Syntax check for IMPLICIT statement

From: glen herrmannsfeldt <gah@xxxxxxxxxxxxxxxx>
Date: Sun, 27 Apr 2008 15:46:23 -0800

James Van Buskirk wrote:

In attempting to create a digestible version of a program that gives
gfortran an ICE and causes incorrect output for ifort, the critical
factor for both seems to be a line such as:

implicit character(len=*,kind=kind('A')) (Q)

In the sense that if the LEN is changed to 3 both ifort and gfortran
get happy with the code. Is this syntactically incorrect or simply
obscure enough that it didn't get sufficient testing on either
compiler?

It is an interesting case. Since len=* can only be used in a
restricted set of cases, such variables would have to be used
in those cases. Still, I don't see any reason why it couldn't
work for those cases.

-- glen

.

References:
- Syntax check for IMPLICIT statement
  - From: James Van Buskirk

Prev by Date: Re: Syntax check for IMPLICIT statement
Next by Date: Re: right sequence of included module files
Previous by thread: Re: Syntax check for IMPLICIT statement
Next by thread: Re: Syntax check for IMPLICIT statement
Index(es):
- Date
- Thread

Relevant Pages

Re: Syntax check for IMPLICIT statement
... gfortran an ICE and causes incorrect output for ifort, ...
(comp.lang.fortran)
Syntax check for IMPLICIT statement
... gfortran an ICE and causes incorrect output for ifort, ... In the sense that if the LEN is changed to 3 both ifort and gfortran ... end program startest ...
(comp.lang.fortran)
Re: Syntax check for IMPLICIT statement
... -|gfortran an ICE and causes incorrect output for ifort, ... -|In the sense that if the LEN is changed to 3 both ifort and gfortran ...
(comp.lang.fortran)
Re: Syntax check for IMPLICIT statement
... gfortran an ICE and causes incorrect output for ifort, ... implicit character) ... "For a scalar named constant or for a dummy argument ... ...
(comp.lang.fortran)