Date: Tue, 17 Jun 1997 08:40:16 -0400 (EDT)
Message-Id: <199706171240.IAA28784@delorie.com>
From: DJ Delorie <dj AT delorie DOT com>
To: robert DOT hoehne AT mathematik DOT tu-chemnitz DOT de
CC: djgpp-workers AT delorie DOT com
In-reply-to: 
	<Pine DOT HPP DOT 3 DOT 95q DOT 970617113806 DOT 22944A-100000 AT newton DOT mathematik DOT tu-chemnitz DOT de>
	(message from Robert Hoehne on Tue, 17 Jun 1997 12:00:06 +0200
	(METDST))
Subject: Re: Latest stub for binutils
Precedence: bulk


> > Because one of these days, the stub will grow to 2.5K.
> 
> What do you mean with this? After applying all the patches
> to stub.asm which I have seen the last time it is still
> 2048 bytes (Or have I missed some patches).

What I mean is that there is nothing in DJGPP that requires the stub
to be 2K.  While I know of no plans to grow it beyond 2K at the
moment, at some point in the future we may choose to do so to add some
useful functionality.  I'd rather not have binutils throw up because
of that.  The Right Thing is to handle whatever size it is, so that
you don't break in the future.

> But my opionion in general is, not to grow the stub. If

That's a preferred opinion anyway, and a really neat challenge, but
it's not a functional requirement.

> we relly need some few bytes more we can reduce the size
> of the exe header in the stub (I mean here to remove some
> duplicated compile time information) and patching djasm
> to emit the smallest needed exe header. Currently it

Nope, tried it.  Not all tools work when you do this (I don't recall
which).

> BTW: If someone didn't know. The exe header must not be
> 512 byte aligned, it have to be 16 byte aligned.

That's what the spec says, but that's not what Reality says.  EXE
headers have to be a whole number of 512 byte blocks.