This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
[objc-improvements-branch]: FYI: Yet another round of ObjC++ work committed -- ADDENDUM
- From: Ziemowit Laski <zlaski at apple dot com>
- To: GCC Patch List <gcc-patches at gcc dot gnu dot org>, GNUStep <discuss-gnustep at gnu dot org>, objc-language at lists dot apple dot com
- Cc: David Ayers <d dot ayers at inode dot at>
- Date: Fri, 25 Jun 2004 21:12:32 -0700
- Subject: [objc-improvements-branch]: FYI: Yet another round of ObjC++ work committed -- ADDENDUM
A couple of days ago, I proudly committed
http://gcc.gnu.org/ml/gcc-cvs/2004-06/msg00966.html
but it turns out the resulting ObjC++ was not quite usable (i.e., there
were many DejaGNU failures) on i686-pc-linux-gnu, since I failed to
appropriately placate the GNU runtime gods. :-( Aside from getting GNU
runtime-specific metadata initialized with values of the correct type,
I had to figure out how to call the GNU ObjC static initializer before
all C++ initializers, including those for global and static variables.
I think I did the latter in a fairly clean fashion, though I'm always
open to suggestions.
Anyway, I just committed
http://gcc.gnu.org/ml/gcc-cvs/2004-06/msg01104.html
and now do not see any failures in the ObjC++ DejaGNU run. Since I was
doing the work on GNU/Linux, I'm curious if I now didn't break Darwin
in some bizarre fashion... :-)
David, if you want to whip-up an identifier cleanup patch for
objc-improvements-branch, go for it. :-)
Enjoy,
--Zem