Re: stabs and #define/#undef

[David Taylor <dtaylor at emc dot com>]

> Date: Tue, 24 Oct 2006 14:46:06 -0400
> From: Daniel Jacobowitz <drow@false.org>
> 
> On Tue, Oct 24, 2006 at 02:40:52PM -0400, David Taylor wrote:
> > > Date: Tue, 24 Oct 2006 11:30:00 -0400
> > > From: Daniel Jacobowitz <drow@false.org>
> 
> The size issue is a real one, and a well known one; there are various
> approaches to it, but so far none of them have been thoroughly
> implemented across the GNU toolchain.

Yep.  And every now and then I double check to see if it has changed
and I somehow missed learning about it.

> > > I don't know what issues there are with grabbing arbitrary numbers for
> > > the new stabs.  Other than that, I don't see what could be
> > > objectionable.
> > 
> > I look upon the stabs.texinfo and stabs.def files as the unofficial
> > ``clearinghouse'' (for want of a better word) for stabs stuff.
> 
> That's because you're a GNU developer :-)  Sun's got the master copy,
> but I don't know how horribly incompatible the two are now.

I'm proud to be a GNU developer.  I've been using and modifying GNU
tools for over 20 years (uh, oh, did I just date myself?).

Did Sun make any effort pull together a list of stab extensions and
document the result?  I wasn't aware of such an effort.

Sun may have had the master copy at one time, but I would tend to
think that their copy is no longer the master copy.  I would argue
that they lost mastership through inaction.  Several companies made
extensions to stabs.  Did Sun make any effort to keep them compatible?
Even just acting as a ``registry''?  I wasn't aware of such an effort.
I know that when stabs.texinfo was created (by John Gilmore at Cygnus)
some reverse engineering of the stabs created by various compilers
occurred.

> -- 
> Daniel Jacobowitz
> CodeSourcery