Re: Document PR 84073 change in /gcc-8/porting_to.html

[Thomas König <tk at tkoenig dot net>]

Am 10.05.2018 um 14:20 schrieb Thomas Koenig:
> Am 10.05.2018 um 12:33 schrieb Jonathan Wakely:
> > Should the fix for PR 84073 be documented, so that users whose code is
> > now rejected understand why, and how to fix it?
> >
> > https://gcc.gnu.org/bugzilla/show_bug.cgi?id=84073
> > https://gcc.gnu.org/gcc-8/porting_to.html#fortran
>
> Sounds like a good idea.
>
> Since I introduced this, I'll do it within a few days (unless, of
> course, somebody beats me to it, which I won't mind :-)

OK, not a few days, but a few hours :-)

Here is the diff.  OK to commit to gcc-docs?

Regards

	Thomas


Index: porting_to.html
===================================================================
RCS file: /cvs/gcc/wwwdocs/htdocs/gcc-8/porting_to.html,v
retrieving revision 1.6
diff -u -r1.6 porting_to.html
--- porting_to.html	4 May 2018 06:35:39 -0000	1.6
+++ porting_to.html	10 May 2018 15:16:57 -0000
@@ -196,7 +196,31 @@
       void foo_ (char*, int*, fortran_charlen_t);
   </code></pre>
 
+<p>
+  Versions of gfortran prior to 8.1 wrongly accepted CHARACTER
+  variables with a length type parameter other than one as C
+  interoperable. For example, the code
+  <pre><code>
+  module mod
+    use iso_c_binding
+    type, bind(C) :: a
+      character(len=2,kind=c_char) :: b ! Wrong
+    end type a
+    character(len=2), bind(C) :: c ! Also wrong
+  end module mod
+  </code></pre>
+  was accepted.  To achieve similar functionality, an array of
+  LEN=1 characters can be used, for example
+  <pre><code>
+  module mod
+    use iso_c_binding
+    type, bind(C) :: a
+      character(kind=c_char) :: b(2)
+    end type a
+    character(kind=c_char), bind(C) :: c(2)
+  end module mod
+  </code></pre>
+</p>
 <h2 id="links">Links</h2>
-
 </body>
 </html>