[RFC] [Ping] Comments on gfortran manual

Steve Kargl sgk@troutmask.apl.washington.edu
Wed Dec 10 21:20:00 GMT 2008


On Wed, Dec 10, 2008 at 09:17:09PM +0100, Daniel Kraft wrote:
> 
> some time ago, I posted a patch for PR 37605, based on suggested manual 
> improvements by Markus Arjen; according to my opinion, those are mostly 
> good points, although some of them may be a bit controversial.
> 
> Comments welcome...  Hope we can apply some of them :)
> 

I haven't received official reviewer status, again; however,
see below.

>  @item -fall-intrinsics
>  @opindex @code{fall-intrinsics}
> -Accept all of the intrinsic procedures provided in libgfortran 
> -without regard to the setting of @option{-std}.  In particular, 
> -this option can be quite useful with @option{-std=f95}.  Additionally,
> -@command{gfortran} will ignore @option{-Wintrinsics-std} and will never try
> -to link to an @code{EXTERNAL} version if the intrinsic is not included in the
> -selected standard.
> +This option causes all intrinsic proceduers (including the GNU-specific

procedures

>  
>  @item -fdefault-integer-8
>  @opindex @code{fdefault-integer-8}
>  Set the default integer and logical types to an 8 byte wide type.
> -Do nothing if this is already the default.
> +Do nothing if this is already the default.  This option does also affect
> +the kind of integer constants like @code{42}.

This option also affects the kind ...

>  @opindex @code{fdefault-real-8}
>  Set the default real type to an 8 byte wide type.
> -Do nothing if this is already the default.
> +Do nothing if this is already the default.  This option does also affect
> +the kind of non-double real constants like @code{1.0}.

Ditto.

>  @item @emph{Class}:
>  Elemental function
> @@ -971,7 +973,7 @@ is @code{.TRUE.}; otherwise, it returns 
>  program test_allocated
>    integer :: i = 4
>    real(4), allocatable :: x(:)
> -  if (allocated(x) .eqv. .false.) allocate(x(i))
> +  if (.not. allocated(x)) allocate(x(i))

I prefer the original code.  Hopefully, this isn't too controversal! :)

Otherwise, the patch looks fine to me.

-- 
Steve



More information about the Fortran mailing list