swarm-support
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

GCC and objc runtime work.


From: Scott Christley
Subject: GCC and objc runtime work.
Date: Mon, 3 Feb 1997 18:16:29 -0500

FYI for those of you not on the GNU ObjC or GNUstep mailing lists.

Scott

>Return-Path: address@hidden
>Date: Thu, 30 Jan 1997 12:31:40 -0500
>X-Sender: address@hidden
>To: GNUstep Discussion <address@hidden>
>From: Scott Christley <address@hidden>
>Subject: GCC and objc runtime work.
>
>
>Hi all,
>
>Just to let you know; I've been submitting patches and adding enhancements
>for the ObjC runtime to the GCC developers recently.  I hope to catch up on
>my backlog of things to do.
>
>One important note is that the next ObjC snapshot that is released will be
>*incompatible* with code compiled with earlier versions of the ObjC runtime.  
>The reason for this is because one of the fundamental structure,
>objc_module, has changed.  Static instances, essentially constant strings
>like @"Hello World!", were being put into a separate member of objc_module,
>but this has been altered and that member has been removed.  The reason for
>this change is simple: GCC compiled code on NEXTSTEP (with NeXT runtime)
>could not be debugged by the NeXT debugger.  The reason it could not be
>debugged is because the NeXT gdb made assumptions about the objc_module
>structure which became invalid when the static instances member was added.
>
>So the solution, which will be in the next ObjC snapshot, was to remove that
>static instances member.  This now allows NeXT gdb to debug GCC generated
>code.  As I indicated above, this also makes code previously compiled by GCC
>obsolete; you must recompile *ALL* your ObjC code with the new compiler.
>Because of this the ObjC runtime version has been incremented to 8.
>
>Please feel free to comment on whether such a change is acceptable.
>
>thanks
>Scott
>



reply via email to

[Prev in Thread] Current Thread [Next in Thread]