From mboxrd@z Thu Jan  1 00:00:00 1970
Return-Path: <gnatsid@sources.redhat.com>
Mailing-List: contact gcc-prs-help@gcc.gnu.org; run by ezmlm
Received: (qmail 31377 invoked by uid 71); 11 Jan 2003 21:16:01 -0000
Date: Sat, 11 Jan 2003 21:16:00 -0000
Message-ID: <20030111211601.31375.qmail@sources.redhat.com>
To: nobody@gcc.gnu.org
Cc: gcc-prs@gcc.gnu.org,
From: Nathanael Nerode <neroden@twcny.rr.com>
Subject: Re: c++/5050: template instantiation depth exceeds...
Reply-To: Nathanael Nerode <neroden@twcny.rr.com>
X-SW-Source: 2003-01/txt/msg00736.txt.bz2
List-Id: <gcc-prs.sourceware.org>

The following reply was made to PR c++/5050; it has been noted by GNATS.

>>From neroden@twcny.rr.com  Mon Jan 13 01:00:41 2003
From: Nathanael Nerode <neroden@twcny.rr.com>
To: gcc-gnats@gcc.gnu.org, gcc-prs@gcc.gnu.org, georg.wild@gmx.de,
   gcc-bugs@gcc.gnu.org, nobody@gcc.gnu.org
Cc:  
Subject: Re: c++/5050: template instantiation depth exceeds...
Date: Sat, 11 Jan 2003 16:07:43 -0500

 The following testcase fails when compiled with -ftemplate-depth-1 -O2 or 
 -ftemplate-depth-1 -O1, but not -ftemplate-depth-1 -O0.  (It's always OK with 
 template depth higher than 1).
 
 This is the error message:
 ----
 test.cxx:5: error: template instantiation depth exceeds maximum of 1 (use
    -ftemplate-depth-NN to increase the maximum) instantiating `ostream&
    ostream::foo(const T&) [with T = int]'
 test.cxx:5:   instantiated from `ostream& ostream::foo(const T&) [with T = 
 int]'test.cxx:12:   instantiated from here
 ----
 
 The problem appears to be that the compiler doesn't realize that it's already 
 instantiated `ostream& ostream::foo(const T&) [with T = int]'.  It doesn't 
 have to instantiate it repeatedly, does it now?  At any rate, behaving 
 differently with optimization is a no-no.
 
 This is the testcase:
 ----
 struct ostream
 {
   template<class T> ostream& foo( const T & )
   {
     return *this;
   }
 };
 
 void foo()
 {
   ostream os;
   (os.foo(1)).foo(2);
 }
 ----