This is the mail archive of the
gcc-bugs@gcc.gnu.org
mailing list for the GCC project.
[Bug libobjc/36610] objc_msg_sendv is broken for targets which pass argument via registers
- From: "nicola at gcc dot gnu.org" <gcc-bugzilla at gcc dot gnu dot org>
- To: gcc-bugs at gcc dot gnu dot org
- Date: Tue, 14 Jun 2011 18:28:51 +0000
- Subject: [Bug libobjc/36610] objc_msg_sendv is broken for targets which pass argument via registers
- Auto-submitted: auto-generated
- References: <bug-36610-4@http.gcc.gnu.org/bugzilla/>
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=36610
Nicola Pero <nicola at gcc dot gnu.org> changed:
What |Removed |Added
----------------------------------------------------------------------------
CC| |nicola at gcc dot gnu.org
--- Comment #23 from Nicola Pero <nicola at gcc dot gnu.org> 2011-06-14 18:26:40 UTC ---
Yes, in a strict sense this bug could be closed ... because objc_msg_sendv() no
longer exists in the GNU Objective-C runtime. So if the bug is about it not
working on some platforms, it can certainly be closed. ;-)
Forwarding in libobjc still uses __builtin_apply(), so the problem, in a wider
sense, still exists. But that forwarding is not used by any user of libobjc as
far as I know (they all have their own libffi-based implementation of
forwarding
that are used through the forwarding hooks that libobjc has) and I'm planning
on
entirely removing it for 4.7.0 - it's unused (the forwarding hooks override any
implementation in libobjc), and it often doesn't work.
I also have some other plans for a new forwarding infrastructure, that may or
may
not happen for 4.7.0.
Summarizing, I would close the bug, or leave it open but just to remind me/us
to:
* remove the existing forwarding code from libobjc;
* remove the forward-1.m testcase;
* add new testcases for the libobjc forwarding hooks. As these are the
official
ways to use forwarding (and the only ones available), we should have testcases.
Let me know if you have any comments.
Thanks