From mboxrd@z Thu Jan  1 00:00:00 1970
Return-Path: <gcc-prs-return-19896-listarch-gcc-prs=gcc.gnu.org@gcc.gnu.org>
Received: (qmail 22587 invoked by alias); 26 Jun 2002 20:26:03 -0000
Mailing-List: contact gcc-prs-help@gcc.gnu.org; run by ezmlm
Precedence: bulk
List-Archive: <http://gcc.gnu.org/ml/gcc-prs/>
List-Post: <mailto:gcc-prs@gcc.gnu.org>
List-Help: <mailto:gcc-prs-help@gcc.gnu.org>
Sender: gcc-prs-owner@gcc.gnu.org
Received: (qmail 22572 invoked by uid 71); 26 Jun 2002 20:26:01 -0000
Date: Wed, 26 Jun 2002 16:36:00 -0000
Message-ID: <20020626202601.22571.qmail@sources.redhat.com>
To: nobody@gcc.gnu.org
Cc: gcc-prs@gcc.gnu.org,
From: Bernd Strieder <strieder@rhrk.uni-kl.de>
Subject: Re: c++/7115: Internal compiler error on template stuff
Reply-To: Bernd Strieder <strieder@rhrk.uni-kl.de>
X-SW-Source: 2002-06/txt/msg00637.txt.bz2
List-Id: <gcc-prs.sourceware.org>

The following reply was made to PR c++/7115; it has been noted by GNATS.

From: Bernd Strieder <strieder@rhrk.uni-kl.de>
To: Reichelt <reichelt@igpm.rwth-aachen.de>, gcc-gnats@gcc.gnu.org,
   strieder@informatik.uni-kl.de, gcc-bugs@gcc.gnu.org, nobody@gcc.gnu.org
Cc:  
Subject: Re: c++/7115: Internal compiler error on template stuff
Date: Wed, 26 Jun 2002 22:18:38 +0200

 > ----------------------------------snip here----------------------------
 > template <unsigned int> class A {};
 >
 > template <typename T> A<sizeof(T*)> foo();
 >
 > void bar()
 > {
 >    foo<void>();
 > }
 > ----------------------------------snip here----------------------------
 
 as I said in the bug report I'm busy preparing my thesis. The code causing ICE 
 in my report is not from me, stripping it down to the essentials has to wait. 
 I know it'd be very nice to have that done. But better than nothing, at least 
 the problem is known to exist.
 
 The code above seems to be quite different from the code the ICE is happening 
 in my example. The only thing in common is the template return type. The 
 project I'm working in has tons of advanced template code, some of the former 
 coders were very interested in STL, Boost, etc., and the techniques coming 
 from there. We were at the outer limits of g++ for several years having to 
 use alternate STL variants most of the time.
 
 Looking at the simplicity of the ICE example above, I really wonder, why I 
 haven't run earlier into the problem. There are obviously people working on 
 the problems at high priority, that makes me feel better.
 
 Bernd Strieder