x11-wm/fluxbox-devel having issues on recent -CURRENT?

[Jeremy Messenger]

On Sat, 14 Aug 2004 16:17:54 -0500, Jeremy Messenger <mezz7 at cox.net> wrote:

> On Sat, 14 Aug 2004 23:48:06 +0300, Ruslan Ermilov <ru at freebsd.org>  
> wrote:
>
>> On Sat, Aug 14, 2004 at 12:32:15PM -0500, Jeremy Messenger wrote:
>>> On Sat, 14 Aug 2004 20:18:06 +0300, Ruslan Ermilov <ru at FreeBSD.org>  
>>> wrote:
>>>
>>> >Hi!
>>> >
>>> >I've just upgraded my June 2004 -CURRENT, including
>>> >recompiling kernel, world, switching to X.Org, and
>>> >upgrading all other ports.  All ports have been
>>> >forcibly recompiled, including those that didn't
>>> >change the version.
>>> >
>>> >Everything is fine so far, except for my WM, which is
>>> >fluxbox-devel -- it now fails with signal 10.  Before
>>> >I upgraded ports, the old binary also exhibited this
>>> >same behavior (SIGBUS), and a recompile of all ports
>>> >didn't help.
>>> >
>>> >Can someone investigate or at least test/confirm this?
>>>
>>> Right now, I am working on upgrade my -CURRENT from June to today. I  
>>> will
>>> find out if I have the same problem as your this afternoon. I am  
>>> planning
>>> to do the 'rm -rf /usr/local/* /usr/X11R6/* /compat/linux/*  
>>> /var/db/pkg/*'
>>> (I always do that) right after I am done with upgrade -CURRENT.
>>>
>>> Do you have mem and io in your kernel config as what in  
>>> /usr/src/UPDATING
>>> has said. They will be need for the X stuff.
>>>
>> Sure, I do.  My kernel is GENERIC based, with all extra
>> stuff removed.
>>
>> $ file /dev/io /dev/mem
>> /dev/io:  character special (248/14)
>> /dev/mem: character special (245/0)
>>
>> Note that old fluxbox binary started exhibit this problem right
>> after a world/kernel upgrade.  The new binary is exhibiting this
>> same problem.
>
> I get the same problem......running in plain X right now.
>
> I am not skill with C++ and read gdb, but it complains about the Time  
> stuff in C++.. I have no idea if the bug is in our GCC 3.4.x stl_list.h  
> or is in fluxbox. I searched in Fluxbox, Debian and Gentoo, I can't find  
> anything else that has same problem as our. Here's backtraces of the  
> crash:
>
> =================================================
> Program terminated with signal 10, Bus error.
>
> #0  0x0812c288 in std::_List_iterator<FbTk::Timer*>::operator++ (
>      this=0xbfbfe834) at stl_list.h:139
> 139             _M_node = _M_node->_M_next;
>
> (gdb) bt
> #0  0x0812c288 in std::_List_iterator<FbTk::Timer*>::operator++ (
>      this=0xbfbfe834) at stl_list.h:139
> #1  0x0812bfa8 in FbTk::Timer::updateTimers (fd=-1077942220) at  
> Timer.cc:146
> #2  0x0806b32b in Fluxbox::eventLoop (this=0x816cc00) at fluxbox.cc:668
> #3  0x0807e8f8 in main (argc=1, argv=0xbfbfed00) at main.cc:238
> =================================================
>
> It will be great if someone can take a look at and cook up patch(es). :-)

The workaround is 'ln -s aj /etc/malloc.conf' to disable the memory debug  
and I am in Fluxbox right now.

Cheers,
Mezz

> Cheers,
> Mezz
>
>> Cheers,


-- 
mezz7 at cox.net  -  mezz at FreeBSD.org
FreeBSD GNOME Team
http://www.FreeBSD.org/gnome/ - gnome at FreeBSD.org