[Patch, gfortran] PR22304, 17917, 16511, 18870 and 23270 - modules, equivalences and commons

[Paul Thomas]

Steven Bosscher wrote:

>On Monday 22 August 2005 08:18, Paul Thomas wrote:
>
>>=================fortran.diff===============
>>
>
>Mostly looks good to me.  One question:
>
>+         /* Check to see if the symbol is already in an equivalence group.
>+            If it is, set the other members as being in common.  */
>+         if (sym->attr.in_equivalence)
>+           {
>+             for (e1 = gfc_current_ns->equiv; e1; e1 = e1->next)
>+               {
>
>
>What if the equivalenced sym is defined in the parent name space, or
>is that impossible?
>
I'll have a think about that and I'll check the standard too;  I have 
not thought through the rules concerning the relative scopes of common 
and equivalences.  An equivalence can be made with a host associated 
common variable but the reverse cannot be right, can it?

I just realised that I need to generalize the common block name 
demangling in module.c; you can be sure somebody will succeed in nesting 
modules 10 or more times!  This is a trivial modification to the patch.

Paul