Re: An egcs 1.1.2 patch for frame.h and crtstuff.c

[hjl at lucon dot org (H.J. Lu)]

> 
> 
>   In message <m10G6Wk-000390C@ocean.lucon.org>you write:
>   > > 
>   > >   > I did #2. Now I cannot find a solution for #1, which I like.
>   > > Send it anyway.  It may be the case that you do not like it, but it's sti
>   > ll
>   > > a better solution, or it may be the case that someone can tweak it to mak
>   > e
>   > > it better.  We'll never know if you don't actually send the patch.
>   > > 
>   > > Or it may be the case that the pragmas really are better.
>   > > 
>   > > But until you send the attribute based patch we'll never know.
>   > > 
>   > 
>   > Here it is. BTW, I am working on the libio patch now.
> 
> Thu Feb 25 11:25:13 1999  H.J. Lu  (hjl@gnu.org)
> 
> 	* frame.h (TARGET_ATTRIBUTE_WEAK): New. Use __attribute__ only
> 	if SUPPORTS_WEAK is not zero and IN_CRTSTUFF is defined.
> 	(__register_frame_info): Append TARGET_ATTRIBUTE_WEAK.
> 	(__deregister_frame_info): Likewise.
> 
> 	* crtstuff.c (IN_CRTSTUFF): Defined before #include "frame.h".
> 	(__do_global_dtors_aux): Check if __deregister_frame_info is
> 	zero before calling it.
> 	(__do_global_dtors): Likewise.
> 	(frame_dummy): Check if __register_frame_info is zero before
> 	calling it.
> 	(__frame_dummy): Likewise.
> Thanks.  This looks pretty good.
> 
> I assume the IN_CRTSTUFF is needed because we only want the external
> references in crtstuff to be weak, not the definitions in frame.c?

Yes.

> 
> What parts of this patch are you not happy with?
> 

It has more codes than the pragma version and it touches frame.h. If
you find it is ok, that is fine with me.

Thanks.


-- 
H.J. Lu (hjl@gnu.org)