enum Bools { False }; template<class T> struct Container { static const int a = False < false; }; template<class CharType> class BasicString { Container<int> b; }; struct InplaceString { InplaceString(BasicString<int>); }; int operator<(int, InplaceString); Container<int> c; Produces: Reading specs from /usr/lib/gcc-lib/i386-linux/3.2.2/specs Configured with: ../src/configure -v --enable-languages=c,c++,java,f77,proto,pascal,objc,ada --prefix=/usr --mandir=/usr/share/man --infodir=/usr/share/info --with-gxx-include-dir=/usr/include/c++/3.2 --enable-shared --with-system-zlib --enable-nls --without-included-gettext --enable-__cxa_atexit --enable-clocale=gnu --enable-java-gc=boehm --enable-objc-gc i386-linux Thread model: posix gcc version 3.2.2 20030109 (Debian prerelease) /usr/lib/gcc-lib/i386-linux/3.2.2/cc1plus -v -D__GNUC__=3 -D__GNUC_MINOR__=2 -D__GNUC_PATCHLEVEL__=2 -D__GXX_ABI_VERSION=102 -D__ELF__ -Dunix -D__gnu_linux__ -Dlinux -D__ELF__ -D__unix__ -D__gnu_linux__ -D__linux__ -D__unix -D__linux -Asystem=posix -D__NO_INLINE__ -D__STDC_HOSTED__=1 -D_GNU_SOURCE -Acpu=i386 -Amachine=i386 -Di386 -D__i386 -D__i386__ -D__tune_i386__ Whole.cpp -D__GNUG__=3 -D__DEPRECATED -D__EXCEPTIONS -quiet -dumpbase Whole.cpp -W -Wall -pedantic -version -o /tmp/ccWxcPJh.s GNU CPP version 3.2.2 20030109 (Debian prerelease) (cpplib) (i386 Linux/ELF) GNU C++ version 3.2.2 20030109 (Debian prerelease) (i386-linux) compiled by GNU C version 3.2.2 20030109 (Debian prerelease). ignoring nonexistent directory "/usr/i386-linux/include" #include "..." search starts here: #include <...> search starts here: /usr/include/c++/3.2 /usr/include/c++/3.2/i386-linux /usr/include/c++/3.2/backward /usr/local/include /usr/lib/gcc-lib/i386-linux/3.2.2/include /usr/include End of search list. Whole.cpp: In instantiation of `BasicString<int>': Whole.cpp:10: instantiated from `Container<int>' Whole.cpp:25: instantiated from here Whole.cpp:16: `BasicString<CharType>::b' has incomplete type Whole.cpp:9: declaration of `struct Container<int>' When compiled with: gcc -v -c -W -Wall -pedantic Whole.cpp Release: GNU C version 3.2.2 20030109 (Debian prerelease) Environment: Linux trevithick 2.4.16 #2 Mon Dec 10 15:54:50 GMT 2001 i686 unknown unknown GNU/Linux
State-Changed-From-To: open->analyzed State-Changed-Why: True. Very weird problem: -------------------------- enum { E }; template <typename T> struct A { static const int a = (E < 0); }; template <typename T> class B { A<int> b; }; struct C { C(B<int>); }; int operator<(C, C); A<int> c; ---------------------------- Changing any bit of the code makes the bug go away (e.g. changing op< to function foo). Used to work with 2.95, but does no longer with 3.2.2 and 3.3 (just accidentially ruined my 3.4 installation, but likely to show same problem). W.
Responsible-Changed-From-To: unassigned->nathan Responsible-Changed-Why: fixing
Responsible-Changed-From-To: nathan->jason Responsible-Changed-Why: theft
State-Changed-From-To: analyzed->open State-Changed-Why: an->an
State-Changed-From-To: open->analyzed State-Changed-Why: Core issues 63 and 212 are relevant here. [temp.inst] says A class template specialization is implicitly instantiated if the class type is used in a context that requires a completely-defined object type or if the completeness of the class type might affect the semantics of the program. [Note: in particular, if the semantics of an expression depend on the member or base class lists of a class template specialization, the class template specialization is implicitly generated. For instance, deleting a pointer to class type depends on whether or not the class declares a destructor, and conversion between pointer to class types depends on the inheritance relationship between the two classes involved. ] ... If the overload resolution process can determine the correct function to call without instantiating a class template definition, it is unspecified whether that instantiation actually takes place. In this case, we need to do overload resolution for E<0, because E is an enum. We consider op<(C,C) as a candidate. To test whether or not it is a valid candidate, we need to determine whether or not there is a conversion from E (and int) to C. Hmm... No, I'm making this harder than it actually is. The bug is that we're instantiating B<int> when we don't need to, because we can't use a user-defined conversion from E (or int) to B<int>, because we're already considering a UDC to C. This should be simple to fix.
From: jason@gcc.gnu.org To: gcc-gnats@gcc.gnu.org Cc: Subject: c++/9420 Date: 13 Mar 2003 21:39:40 -0000 CVSROOT: /cvs/gcc Module name: gcc Changes by: jason@gcc.gnu.org 2003-03-13 21:39:40 Modified files: gcc/cp : ChangeLog call.c search.c Log message: PR c++/9420 * search.c (lookup_conversions): Call complete_type here. * call.c (implicit_conversion): Not here. Patches: http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/cp/ChangeLog.diff?cvsroot=gcc&r1=1.3271&r2=1.3272 http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/cp/call.c.diff?cvsroot=gcc&r1=1.368&r2=1.369 http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/cp/search.c.diff?cvsroot=gcc&r1=1.256&r2=1.257
From: jason@gcc.gnu.org To: gcc-gnats@gcc.gnu.org Cc: Subject: c++/9420 Date: 13 Mar 2003 21:40:34 -0000 CVSROOT: /cvs/gcc Module name: gcc Branch: gcc-3_3-branch Changes by: jason@gcc.gnu.org 2003-03-13 21:40:34 Modified files: gcc/cp : ChangeLog call.c search.c Log message: PR c++/9420 * search.c (lookup_conversions): Call complete_type here. * call.c (implicit_conversion): Not here. Patches: http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/cp/ChangeLog.diff?cvsroot=gcc&only_with_tag=gcc-3_3-branch&r1=1.3076.2.83&r2=1.3076.2.84 http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/cp/call.c.diff?cvsroot=gcc&only_with_tag=gcc-3_3-branch&r1=1.341.2.13&r2=1.341.2.14 http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/cp/search.c.diff?cvsroot=gcc&only_with_tag=gcc-3_3-branch&r1=1.243.2.2&r2=1.243.2.3
From: jason@gcc.gnu.org To: gcc-gnats@gcc.gnu.org Cc: Subject: c++/9420 Date: 13 Mar 2003 21:41:38 -0000 CVSROOT: /cvs/gcc Module name: gcc Branch: gcc-3_2-branch Changes by: jason@gcc.gnu.org 2003-03-13 21:41:38 Modified files: gcc/cp : ChangeLog call.c search.c Log message: PR c++/9420 * search.c (lookup_conversions): Call complete_type here. * call.c (implicit_conversion): Not here. Patches: http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/cp/ChangeLog.diff?cvsroot=gcc&only_with_tag=gcc-3_2-branch&r1=1.2685.2.114.2.69&r2=1.2685.2.114.2.70 http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/cp/call.c.diff?cvsroot=gcc&only_with_tag=gcc-3_2-branch&r1=1.307.2.6.4.2&r2=1.307.2.6.4.3 http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/cp/search.c.diff?cvsroot=gcc&only_with_tag=gcc-3_2-branch&r1=1.223.2.3.4.2&r2=1.223.2.3.4.3
State-Changed-From-To: analyzed->closed State-Changed-Why: fixed.
From: jason@gcc.gnu.org To: gcc-gnats@gcc.gnu.org Cc: Subject: c++/9420 Date: 21 Mar 2003 07:10:44 -0000 CVSROOT: /cvs/gcc Module name: gcc Changes by: jason@gcc.gnu.org 2003-03-21 07:10:44 Added files: gcc/testsuite/g++.dg/template: overload1.C Log message: PR c++/9420 * search.c (lookup_conversions): Call complete_type here. * call.c (implicit_conversion): Not here. Patches: http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/testsuite/g++.dg/template/overload1.C.diff?cvsroot=gcc&r1=NONE&r2=1.1