This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: Change in status of __GXX_WEAK__ on mainline?
- From: Neil Booth <neil at daikokuya dot demon dot co dot uk>
- To: rittle at labs dot mot dot com
- Cc: gcc-patches at gcc dot gnu dot org
- Date: Thu, 9 May 2002 18:30:56 +0100
- Subject: Re: Change in status of __GXX_WEAK__ on mainline?
- References: <200205090802.g4982jQ48987@latour.rsch.comm.mot.com>
Loren James Rittle wrote:-
> Regarding http://gcc.gnu.org/ml/gcc-patches/2002-05/msg00404.html .
>
> It appears that __GXX_WEAK__ is no longer defined in every case that
> it had been (and I suspect that you have good reason for the change
> based on the fact that the correct value for __GXX_WEAK__ is target
> dependent). Running this:
>
> cc1plus -E -lang-c++ -dM [...]
>
> under the debugger reveals that the callback is never made and thus
> __GXX_WEAK__ is never defined either way. This subtly changes
> behavior verses 3.0, 3.1.
Yah, stupid bug. I got confused in the tangle that is front-end
initialization ordering, and with -E we exit just before the
built-ins are registered. They should be there with -c.
I'll post a patch soon, possibly including Zack's suggestions
on how to do the target stuff better.
In the near term I want to untangle the initialization web
a bit more.
Thanks,
Neil.