[Patch, Fortran] PR34246 - Fix for character bind(C) functions

[Patch, Fortran] PR34246 - Fix for character bind(C) functions

[Tobias Burnus]

[-- Attachment #1: Type: text/plain, Size: 579 bytes --]

I implemented a while a go a patch which should generate for
   character(len=1,kind=c_char) function foo() bind(C)
an interface matching the C prototype
  char foo()

However, the test case failed on some systems (e.g. working on x86-64
but failing on x86-32). It turned out that the returned value was still
regarded as array by the middle end.

The attached patch fixes this; as the patch was created on a
try-and-error bases, it should be carefully reviewed.

Build and regression tested (check-gfortran + gomp) with -m32 and -m64
on x86-64-linux.

OK for the trunk?

Tobias

[-- Attachment #2: bind-c-fixes.diff --]
[-- Type: application/unknown, Size: 6892 bytes --]

Re: [Patch, Fortran] PR34246 - Fix for character bind(C) functions

[Tobias Burnus]

Tobias Burnus wrote:
> Build and regression tested (check-gfortran + gomp) with -m32 and -m64
> on x86-64-linux.
>   
As Dominique wrote in the associated PR: "The patch [...] fix the
problem for ppc Darwin9 without regression for ppc/Intel Darwin9."

> OK for the trunk?
>   

Tobias

Re: [Patch, Fortran] PR34246 - Fix for character bind(C) functions

[Tobias Burnus]

* PING *

http://gcc.gnu.org/ml/gcc-patches/2007-12/msg00424.html

Tobias Burnus wrote:
> I implemented a while a go a patch which should generate for
>    character(len=1,kind=c_char) function foo() bind(C)
> an interface matching the C prototype
>   char foo()
>
> However, the test case failed on some systems (e.g. working on x86-64
> but failing on x86-32). It turned out that the returned value was still
> regarded as array by the middle end.
>
> The attached patch fixes this; as the patch was created on a
> try-and-error bases, it should be carefully reviewed.
>
> Build and regression tested (check-gfortran + gomp) with -m32 and -m64
> on x86-64-linux.
>
> OK for the trunk?
>
> Tobias

Re: [Patch, Fortran] PR34246 - Fix for character bind(C) functions

[Jerry DeLisle]

Tobias Burnus wrote:
> I implemented a while a go a patch which should generate for
>    character(len=1,kind=c_char) function foo() bind(C)
> an interface matching the C prototype
>   char foo()
> 
> However, the test case failed on some systems (e.g. working on x86-64
> but failing on x86-32). It turned out that the returned value was still
> regarded as array by the middle end.
> 
> The attached patch fixes this; as the patch was created on a
> try-and-error bases, it should be carefully reviewed.
> 
> Build and regression tested (check-gfortran + gomp) with -m32 and -m64
> on x86-64-linux.
> 
> OK for the trunk?
> 
> Tobias
OK to commit

Jerry

Re: [Patch, Fortran] PR34246 - Fix for character bind(C) functions

[Tobias Burnus]

Jerry DeLisle wrote:
> Tobias Burnus wrote:
>> Build and regression tested (check-gfortran + gomp) with -m32 and -m64
>> on x86-64-linux.
> OK to commit
Thanks for the review. Committed as Rev. 130991.

Tobias