---------- Forwarded message ----------
From: Martin Michlmayr <t...@cyrius.com>
Date: Apr 11, 2006 12:20 PM
Subject: Bug#361933: FTBFS with GCC 4.2: invalid pure specifier
To: sub...@bugs.debian.org

Package: krusader
Version: 1.70.0-1
Tags: patch

Your package fails to build with GCC 4.2.  Version 4.2 has not been
released yet but I'm building with a snapshot in order to find errors
and give people an advance warning.  The bug below is in your package
and not because I'm using a snapshot of the compiler so please take a
look at it.  You can reproduce this with the gcc-snapshot package.

> Automatic build of krusader_1.70.0-1 on em64t by sbuild/amd64 1.112
...
> if x86_64-linux-gnu-g++ -DHAVE_CONFIG_H -I. -I. -I../.. -I/usr/include/kde -I/usr/include/qt3 -I.   -DQT_THREAD_SUPPORT  -D_REENTRANT -D_LARGEFILE64_SOURCE -DKDE_NO_COMPAT -DQT_NO_ASCII_CAST  -Wnon-virtual-dtor -Wno-long-long -Wundef -ansi -D_XOPEN_SOURCE=500 -D_BSD_SOURCE -Wcast-align -Wconversion -Wchar-subscripts -Wall -W -Wpointer-arith -Wwrite-strings -O2 -Wformat-security -Wmissing-format-attribute -fno-exceptions -fno-check-new -fno-common  -MT krbookmarkhandler.o -MD -MP -MF ".deps/krbookmarkhandler.Tpo" \
>         -c -o krbookmarkhandler.o `test -f 'krbookmarkhandler.cpp' || echo './'`krbookmarkhandler.cpp; \
>       then mv -f ".deps/krbookmarkhandler.Tpo" ".deps/krbookmarkhandler.Po"; \
>       else rm -f ".deps/krbookmarkhandler.Tpo"; exit 1; \
>       fi
> ../VFS/vfs.h:137: error: invalid pure specifier (only `= 0' is allowed) before ';' token
> /usr/include/kde/kbookmarkmanager.h:312: warning: 'class KBookmarkOwner' has virtual functions but non-virtual destructor
> make[4]: *** [krbookmarkhandler.o] Error 1

--- ./krusader/VFS/vfs.h~       2006-04-11 09:49:09.000000000 +0000
+++ ./krusader/VFS/vfs.h        2006-04-11 09:49:13.000000000 +0000
@@ -134,7 +134,7 @@

 protected:
        /// Feel the vfs dictionary with vfiles, must be implemented
for each vfs
-       virtual bool populateVfsList(const KURL& origin, bool showHidden) = 0L;
+       virtual bool populateVfsList(const KURL& origin, bool showHidden) = 0;
        /// Set the vfile list pointer
        void setVfsFilesP(vfileDict* dict);
        /// clear and delete all current vfiles

--
Martin Michlmayr
http://www.cyrius.com/

> Hi guys,

> There is a gcc 4.2 patch in the Debian bugtracker.
> http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=361933
> http://bugs.debian.org/cgi-bin/pkgreport.cgi?pkg=krusader
> Can someone take a look at it?
[...]
> --- ./krusader/VFS/vfs.h~       2006-04-11 09:49:09.000000000 +0000
> +++ ./krusader/VFS/vfs.h        2006-04-11 09:49:13.000000000 +0000
> @@ -134,7 +134,7 @@

>  protected:
>         /// Feel the vfs dictionary with vfiles, must be implemented
> for each vfs
> -       virtual bool populateVfsList(const KURL& origin, bool showHidden) =
> 0L; +       virtual bool populateVfsList(const KURL& origin, bool
> showHidden) = 0; /// Set the vfile list pointer
>         void setVfsFilesP(vfileDict* dict);
>         /// clear and delete all current vfiles

> Just out of curiosity, where is "0L" defined, and how is it different from
> 0 or NULL (i suppose it's not)?

> On Tuesday 11 April 2006 22:13, Dirk Eschler wrote:
> > Just out of curiosity, where is "0L" defined, and how is it different
> > from 0 or NULL (i suppose it's not)?

> Hi!

> 0L is just the "long" version of 0, whatever this means (the meaning is
> compiler dependent). I think, that

> virtual bool populateVfsList(const KURL& origin, bool showHidden) = 0L;

> is syntactically wrong, but gcc < 4.2 just allows it. From the bug report I
> further guess, that gcc 4.2 will not allow it any more. Can't check it, as
> I am still using gcc 3.4.2,

> In any case the fix should not harm.

> Heiner

> Just out of curiosity, where is "0L" defined, and how is it  
> different from 0
> or NULL (i suppose it's not)?

> I changed all the "0L" to  "0" in vfs files. I'll comit as soon as I'll have
> a version that works ;)

> On 4/12/06, Jonas Bähr < jonas.ba...@web.de> wrote:

> > Am 11.04.2006 um 22:13 schrieb Dirk Eschler:
> > > Just out of curiosity, where is "0L" defined, and how is it
> > > different from 0
> > > or NULL (i suppose it's not)?

> > From what I know 0 is the mathemtical zero while 0L is a null-adress
> > (used for pointers) - if there is the need to distinguish between
> > numerical values and adresses...

> > bye,
> > Jonas