From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 25602 invoked by alias); 20 Jun 2006 18:28:06 -0000 Received: (qmail 25568 invoked by uid 48); 20 Jun 2006 18:27:59 -0000 Date: Tue, 20 Jun 2006 18:30:00 -0000 Message-ID: <20060620182759.25567.qmail@sourceware.org> X-Bugzilla-Reason: CC References: Subject: [Bug c++/19134] Class visibility of templated classes can't be overridden for function specializations In-Reply-To: Reply-To: gcc-bugzilla@gcc.gnu.org To: gcc-bugs@gcc.gnu.org From: "jason at gcc dot gnu dot org" Mailing-List: contact gcc-bugs-help@gcc.gnu.org; run by ezmlm Precedence: bulk List-Archive: List-Post: List-Help: Sender: gcc-bugs-owner@gcc.gnu.org X-SW-Source: 2006-06/txt/msg01875.txt.bz2 List-Id: ------- Comment #2 from jason at gcc dot gnu dot org 2006-06-20 18:27 ------- This is clearly a bug, since the specialization is not inline and so should not be affected by -fvisibility-inlines-hidden. The broader issue here is the question of when #pragma visibility should override other specified visibilities; suppose the user gave A::foo an explicit visibility. Should that or the #pragma take precedence? After some thought I've concluded that the #pragma should win, that the template's visibility should only be used for implicit instantiations, or if the specialization/explicit instantiation has no explicit visibility. -- jason at gcc dot gnu dot org changed: What |Removed |Added ---------------------------------------------------------------------------- AssignedTo|unassigned at gcc dot gnu |jason at gcc dot gnu dot org |dot org | Status|NEW |ASSIGNED Last reconfirmed|2005-07-02 01:51:33 |2006-06-20 18:27:58 date| | http://gcc.gnu.org/bugzilla/show_bug.cgi?id=19134