Date: Wed, 11 Mar 1998 19:34:13 -0800 (PST)
Message-Id: <199803120334.TAA03678@adit.ap.net>
Mime-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
To: sl AT psycode DOT com, djgpp AT delorie DOT com
From: Nate Eldredge <eldredge AT ap DOT net>
Subject: Re: long double - update
Precedence: bulk

At 07:23  3/11/1998 GMT, Gili wrote:
>Hi,
>
>	In my last message, I meant to say DJGPP reports "long double" as 
>VOID... Why is this so?

This came up a while ago, from me. :) Search the archives for the thread
"GDB doesn't understand long doubles" around 7/97. Here's a summary.

There are two things necessary to fix this. First, you have to use stabs
debugging which is complete enough to describe the `long double' type. I
think the buglist has this patch, or I can mail it. You have to apply it and
rebuild GCC. As another option, you can get PGCC which has it already.

Also, due to an oversight by the GDB maintainer, `long double' handling was
not enabled in `gdb'. There is also a patch for this. I think it was posted
as part of the aforementioned thread, or I can mail it too. This requires
rebuilding GDB, of course. :)

If you're using RHIDE, I'm not sure what you have to do to make it work. You
might be able to apply the patch to the parts of GDB code that are used,
then recompile RHIDE. Or not. :)

Hope this helps.

Nate Eldredge
eldredge AT ap DOT net