Hi Carsten,
On Mon, Jun 9, 2008 at 9:43 PM, Carsten Neumann <[EMAIL PROTECTED]> wrote:
> Ákos Balázs wrote:
>> I've successfully built the dailybuild I have from early April 2008
>> (BTW what's up with the dailybuild page?) of 1.x.
> I don't know what is going on with it, but I have no access to it and
> Dirk is traveling, so I'm not sure how to fix it for now. Sorry.
Yeah, he's currently in Germany: just happened to bump into him here.
:-) Will try to ask him at lunch. BTW is the machine still hosted in
Darmstadt?
>> All the test
>> programs seem to work, but all of them produce a segfault when
>> exiting. First I thought it might be related to the NVidia driver, but
>> apparently other OpenGL based programs work, plus all even the test
>> programs that don't open a window and therefore probably don't create
>> an OpenGL context at all seem to crash at exit.
>>
>> Has anyone seen something like this? I've only built the release
>> versions so far, will build the debug versions on Monday so that I can
>> see where does it exactly crash, but maybe someone has an immediate
>> idea.
> hm, just a wild guess but it might be related to the log locking, which
> we enabled around that time IIRC. My initial attempts at it were broken,
> so you might be hitting that.
May be. I've done a debug built and obtained a stacktrace, it looks
the same for all the programs I've tried:
Program received signal EXC_BAD_ACCESS, Could not access memory.
Reason: KERN_INVALID_ADDRESS at address: 0xc000001b
0x90a56f8a in std::ostream::flush ()
(gdb) bt
#0 0x90a56f8a in std::ostream::flush ()
#1 0x0059ed05 in osg::ConnectionFactory::subPointType (this=0xa098e0,
type=0x6a5b98) at
/Users/edhellon/work/OpenSG/dailybuild_06_04_2008/OpenSG/Source/Base/Network/Base/OSGConnectionFactory.cpp:160
#2 0x005a03b6 in osg::ConnectionType::~ConnectionType (this=0x6a5b98)
at
/Users/edhellon/work/OpenSG/dailybuild_06_04_2008/OpenSG/Source/Base/Network/Base/OSGConnectionType.cpp:98
#3 0x005a0421 in osg::ConnectionType::~ConnectionType (this=0x6a5b98)
at
/Users/edhellon/work/OpenSG/dailybuild_06_04_2008/OpenSG/Source/Base/Network/Base/OSGConnectionType.cpp:99
#4 0x005b527f in __tcf_1 () at
/Users/edhellon/work/OpenSG/dailybuild_06_04_2008/OpenSG/Source/Base/Network/Socket/OSGPointSockPipeline.cpp:276
#5 0x9360ffdc in __cxa_finalize ()
#6 0x9360fed0 in exit ()
#7 0x00001e9b in start ()
(gdb)
Forget it, just checked the CVS logs, it seems to be already fixed by
Patrick Daehne in revision 1.6 of OSGConnectionFactory.cpp. I will do
a CVS update and a rebuild and will let you know.
>> Oh, and I had to patch a bit both the build system and the source to
>> get it correctly compile on OS X 10.5, I will post the patches on
>> Monday -- as far as I could tell these issues are still not fixed in
>> the HEAD branch of the CVS tree on SF.
> ok, looking forward for your patches :)
hmmm, sorry, no time yesterday in the end (had to watch the EM 2008
games :) but will post the patches ASAP. Nothing big actually, really,
just the usual fixes for gcc options regarding inlining, and Apple
seems to have deprecated finite() when specifying the "-ansi" switch
(which seems to contradict the gcc man page they ship, actually) so
you have to use std::isfinite().
Yours,
Ákos
-------------------------------------------------------------------------
Check out the new SourceForge.net Marketplace.
It's the best place to buy or sell services for
just about anything Open Source.
http://sourceforge.net/services/buy/index.php
_______________________________________________
Opensg-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/opensg-users
