This is the mail archive of the libstdc++@gcc.gnu.org mailing list for the libstdc++ 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: question about __cxa_guard_acquire and static variables in the scope of a function

From: Ian Lance Taylor <iant at google dot com>
To: Brendan Miller <catphive at catphive dot net>
Cc: libstdc++ at gcc dot gnu dot org
Date: Mon, 09 May 2011 22:36:01 -0700
Subject: Re: question about __cxa_guard_acquire and static variables in the scope of a function
References: <BANLkTi=wWHQ9sRwiEdFWfBOeQ12u53bGfQ@mail.gmail.com> <mcrei47o39c.fsf@coign.corp.google.com> <BANLkTikfCTfQMs-RedBdgKWHaiG6dmph9g@mail.gmail.com> <BANLkTim5JxMmoOXufjOfFwoKOnZxqi=gzg@mail.gmail.com>

Brendan Miller <catphive@catphive.net> writes:

> Do you know if G++ 3.4 did thread-safe static local initialization in
> this manner? This bug appeared after a migration from 3.4 to 4.1. I'm
> wondering if statics in the scope of the function were previously
> getting double-created and double destroyed, and we just didn't notice
> somehow.

According to http://gcc.gnu.org/PR13684 support for thread-safe
initialization of function statics was indeed introduced in gcc 4.0.

Ian

References:
- question about __cxa_guard_acquire and static variables in the scope of a function
  - From: Brendan Miller
- Re: question about __cxa_guard_acquire and static variables in the scope of a function
  - From: Ian Lance Taylor
- Re: question about __cxa_guard_acquire and static variables in the scope of a function
  - From: Brendan Miller
- Re: question about __cxa_guard_acquire and static variables in the scope of a function
  - From: Brendan Miller

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