Date: Tue, 22 Jul 1997 13:35:43 +1100
From: Bill Currie <billc AT blackmagic DOT tait DOT co DOT nz>
Subject: Re: libgcc.a problem in latest gcc snapshot
In-reply-to: <199707212020.AA243056404@typhoon.rose.hp.com>
To: djgpp-workers AT delorie DOT com
Message-id: <199707220140.NAA19989@teleng1.tait.co.nz gatekeeper.tait.co.nz>
Organization: Tait Electronics Limited
MIME-version: 1.0
Content-type: text/plain; charset=US-ASCII
Content-transfer-encoding: 7BIT
References: <199707130006 DOT UAA23204 AT delorie DOT com>; from "DJ Delorie" at Jul 12,
 97 8:06 pm
Comments: Authenticated sender is <billc AT blackmagic DOT tait DOT co DOT nz>
Precedence: bulk

On 21 Jul 97 at 13:20, Andrew Crabtree wrote:
> We have added the line 
> 
> #define HAS_INIT_SECTION 
> 
> into go32.h.  This will 
> 
> a) suppress generatoin of __main in libgcc.a

That's fine, but is there another solution (see below)?

> b) Suppress generation of a call to __main when compiling a function
> named main.

I *DON'T* like that one personally, as I use the feature in my 
kernel.  However, I CAN live with it (it just breaks my kernel, but 
that's repairable).  What about having another #define, maybe 
CALL___MAIN_ANYWAY or something similar, as I think that the 
generated call to __main is a good idea.

> Item b) seemed redundant as the crt code calls __main anyway before
> calling main.  As a result the check to see if the constructors have
> already been processed can probably be removed.

That is the only point I can agree on about removing the call to 
__main.

Bill
--
Leave others their otherness.