This is the mail archive of the gcc@gcc.gnu.org mailing list for the GCC project.

Index Nav:	[Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav:	[Date Prev] [Date Next]	[Thread Prev] [Thread Next]
Other format:	[Raw text]

Re: Visibility attribute (was: proposed change: weak symbol support for Darwin)

From: Richard Henderson <rth at redhat dot com>
To: Matt Austern <austern at apple dot com>
Cc: Alexandre Oliva <aoliva at redhat dot com>, GCC list <gcc at gcc dot gnu dot org>
Date: Thu, 16 Oct 2003 10:56:33 -0700
Subject: Re: Visibility attribute (was: proposed change: weak symbol support for Darwin)
References: <D0088458-FF78-11D7-A525-000393B2ABA2@apple.com> <8FDE5768-FF8C-11D7-BEF9-000393B2ABA2@apple.com>

On Wed, Oct 15, 2003 at 08:55:20PM -0700, Matt Austern wrote:
> Caveat: I haven't tested this code yet, so it's quite possible
> there are typos and such.  I'm posting it to show the general
> idea I have in mind.  If people agree it's the right approach,
> then I'll test it and submit it as an official patch.

If there hadn't been two free bits, I'd be more hesitant to approve
this scheme.  The thing you're missing in your calculations of how
much memory is used by these attributes are the auto variables.
I'm sure they dwarf global variables by a large margin.

But since you'd not be having to add another word of flags, then I
think this is ok.

r~

References:
- Visibility attribute (was: proposed change: weak symbol support for Darwin)
  - From: Matt Austern
- Re: Visibility attribute (was: proposed change: weak symbol support for Darwin)
  - From: Matt Austern

Index Nav:	[Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav:	[Date Prev] [Date Next]	[Thread Prev] [Thread Next]