solib-aix5.c

solib-aix5.c

[Joel Brobecker]

Hello,

I have recently investigated a regression in core file loading on
AIX, and will post a fix soon. However, this made me realize that
we are still using the SOLIB_ADD and SOLIB_CREATE_INFERIOR_HOOK
on this platform.

I have checked, and it appears that we are the last one to use
these macros. I haven't thoroughly check what "we" includes,
but wouldn't it be nice if we could get rid of them entirely?

In any case, I think what's involved is to move to the target_so_ops,
right? It just so happens that I've found a file called solib-aix5.c.
This file seems to have been created in 2001, but doesn't seem to be
used at all. It's never included on any configuration as far as I
can tell.

Does anyone know the history of this file?

-- 
Joel

Re: solib-aix5.c

[Mark Kettenis]

> Date: Fri, 18 May 2007 22:43:42 -0700
> From: Joel Brobecker <brobecker@adacore.com>
> 
> Hello,
> 
> I have recently investigated a regression in core file loading on
> AIX, and will post a fix soon. However, this made me realize that
> we are still using the SOLIB_ADD and SOLIB_CREATE_INFERIOR_HOOK
> on this platform.
> 
> I have checked, and it appears that we are the last one to use
> these macros. I haven't thoroughly check what "we" includes,
> but wouldn't it be nice if we could get rid of them entirely?
> 
> In any case, I think what's involved is to move to the target_so_ops,
> right? It just so happens that I've found a file called solib-aix5.c.
> This file seems to have been created in 2001, but doesn't seem to be
> used at all. It's never included on any configuration as far as I
> can tell.
> 
> Does anyone know the history of this file?

I vageuly remember an ia64 port of AIX.  The fact that there are some
section name strings that start with ".IA_64." suggest that this file
was part of that effort.  So I think the file should be deleted.

Mark

Re: solib-aix5.c

[Joel Brobecker]

> I vageuly remember an ia64 port of AIX.  The fact that there are some
> section name strings that start with ".IA_64." suggest that this file
> was part of that effort.  So I think the file should be deleted.

Thanks Mark. I have sent a patch that proposes just that. I've given
a week before going ahead with the deletion.

-- 
Joel