[Mono-devel-list] Patch for monodis AIX symbol clash

Bernie Solomon bernard at ugsolutions.com
Thu May 6 19:06:24 EDT 2004


Thanks for the hints, I'd assumed the calling convention was similar
to one of the existing platforms.

I'd got as far as realizing the function descriptors were needed (at
least in the interpreter which I was looking at first as I am more
familiar with this code) and found the magic _ptrgl helper
function that is used by default by the IBM compiler to call
a function pointer. Will either need to use it or inline the equivalent.

There are also just plain porting issues re semaphores, AIO and
so on....

Low priority for me but we will see what progress can be made
but I know others are interested in this as well.

Bernie

----- Original Message ----- 
From: "Paolo Molaro" <lupus at ximian.com>
To: <mono-devel-list at lists.ximian.com>
Sent: Thursday, May 06, 2004 7:43 AM
Subject: Re: [Mono-devel-list] Patch for monodis AIX symbol clash


> On 05/05/04 Bernie Solomon wrote:
> > I have vaguely tried to build on AIX using IBM's compiler - nothing
running
> > but found some symbol clashes. This renames map_t in monodis since there
is
> > one in a system header at least when using _ALL_SOURCE. Just renames
map_t
> > to dis_map_t which is straightforward but just wanted to check if there
were
> > any problems with my choice of name.
>
> Looks fine to me.
> As for AIX: the calling convention is similar (or the same) as on
> MacOSX, so you may want to grep for __APPLE__ in the ppc sources and
> check for AIX, too, except in the ucontext code which I assume is
> different (or, better, in mini-ppc.h, add a call conv specific define
> for Darwin and AIX and change the call conv specific places to use
> that instead of __APPLE__).
> The other issue you may run into with the port is related to function
> descriptors: there may be a few places where we return a pointer to the
> code instead of a pointer to a function descriptor for the function
> pointer.
>
> lupus
>
> -- 
> -----------------------------------------------------------------
> lupus at debian.org                                     debian/rules
> lupus at ximian.com                             Monkeys do it better
> _______________________________________________
> Mono-devel-list mailing list
> Mono-devel-list at lists.ximian.com
> http://lists.ximian.com/mailman/listinfo/mono-devel-list
>




More information about the Mono-devel-list mailing list