Re: [Flightgear-devel] FlightGear/SimGear (git) segfaults
On Tue, May 25, 2010 at 10:27 PM, Curtis Olson wrote: > Hi Torsten, > > On Tue, May 25, 2010 at 2:57 PM, Torsten Dreyer wrote: > > > There's no difference between Technique.cxx in the cvs vs. git >> > repositories, The CVS version of flightgear runs fine against the >> > fgdata-git tree. The git version of flightgear crashes. I do observer >> > there are a bunch of shader/effect code differences between simgear-cvs >> > and simgear-git. Could something new in that code be breaking >> FlightGear? >> > Is anyone else having trouble running the git version of FlightGear? >> For me, the lates git sources of SimGear+FlightGear run on the current git >> and >> latest cvs base, I have. >> >> I observe no difference between cvs and git in the data/Shader and >> data/Effects files - if yours differ, maybe that is the problem? >> > > I see quite a few differences in data-cvs/Shaders versus data-git/Shaders, > but the Effects directories are identical. > > My problem is that I cannot run simgear-git + flightgear-git + fgdata-git > ... this should be the new default for developers. When I run all the git > pieces together I get the segfault I reported. > > The segfault also happens when I run the git compiled code with data-cvs. > The problem does not happen when I run the CVS compiled code. > > If I am the only one seeing this problem, then it must be something on my > end. I thought my system was pretty main stream and pretty current ... hmmm > > > Tim: is there a way to check out a version of the git code that represents > the first conversion/import of the cvs repository before any subsequent > changes were made? I'm still a git-novice. > > There's a "last-cvs" tag in each of the repos. If you want to get the code that was the last thing from cvs, then git checkout last-cvs. Tim > Thanks, > > Curt. > -- > Curtis Olson: http://baron.flightgear.org/~curt/ > > > -- > > > ___ > Flightgear-devel mailing list > Flightgear-devel@lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/flightgear-devel > > -- ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] FlightGear/SimGear (git) segfaults
Hi Torsten, On Tue, May 25, 2010 at 2:57 PM, Torsten Dreyer wrote: > > There's no difference between Technique.cxx in the cvs vs. git > > repositories, The CVS version of flightgear runs fine against the > > fgdata-git tree. The git version of flightgear crashes. I do observer > > there are a bunch of shader/effect code differences between simgear-cvs > > and simgear-git. Could something new in that code be breaking > FlightGear? > > Is anyone else having trouble running the git version of FlightGear? > For me, the lates git sources of SimGear+FlightGear run on the current git > and > latest cvs base, I have. > > I observe no difference between cvs and git in the data/Shader and > data/Effects files - if yours differ, maybe that is the problem? > I see quite a few differences in data-cvs/Shaders versus data-git/Shaders, but the Effects directories are identical. My problem is that I cannot run simgear-git + flightgear-git + fgdata-git ... this should be the new default for developers. When I run all the git pieces together I get the segfault I reported. The segfault also happens when I run the git compiled code with data-cvs. The problem does not happen when I run the CVS compiled code. If I am the only one seeing this problem, then it must be something on my end. I thought my system was pretty main stream and pretty current ... hmmm Tim: is there a way to check out a version of the git code that represents the first conversion/import of the cvs repository before any subsequent changes were made? I'm still a git-novice. Thanks, Curt. -- Curtis Olson: http://baron.flightgear.org/~curt/ -- ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] FlightGear/SimGear (git) segfaults
> There's no difference between Technique.cxx in the cvs vs. git > repositories, The CVS version of flightgear runs fine against the > fgdata-git tree. The git version of flightgear crashes. I do observer > there are a bunch of shader/effect code differences between simgear-cvs > and simgear-git. Could something new in that code be breaking FlightGear? > Is anyone else having trouble running the git version of FlightGear? For me, the lates git sources of SimGear+FlightGear run on the current git and latest cvs base, I have. I observe no difference between cvs and git in the data/Shader and data/Effects files - if yours differ, maybe that is the problem? Torsten -- ___ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
Re: [Flightgear-devel] FlightGear/SimGear (git) segfaults
On Tue, May 25, 2010 at 10:34 AM, Curtis Olson wrote:
> Hi,
>
> I am having a problem running the latest flightgear and simgear from the
> gitorious repository. Here is a back trace of my crash. Please note that
> the CVS version of FlightGear runs just fine on my system, and I've just
> reinstalled the latest nvidia graphics drivers so I'm 99.99% confident this
> isn't a system configuration problem.
>
> OS = Fedora 12 (with all the latest system updates installed)
> Graphics drivers = NVIDIA-Linux-x86_64-195.36.15-pkg2.run
> CPU = 64-bit dual core athlon
> kernel = 2.6.32.12-115.fc12.x86_64
> OSG = 2.9.7
>
> Here's a back-trace from gdb:
>
> Program received signal SIGSEGV, Segmentation fault.
> 0x00a0850d in SGExpression::getValue (this=0x0,
> binding=0x7fffd830) at
> ../../../simgear/structure/SGExpression.hxx:123
> 123 { T value; eval(value, binding); return value; }
> Missing separate debuginfos, use: debuginfo-install
> alsa-lib-1.0.23-1.fc12.x86_64 dbus-libs-1.2.16-9.fc12.x86_64
> flac-1.2.1-6.fc12.x86_64 freealut-1.1.0-11.fc12.x86_64
> freeglut-2.6.0-1.fc12.x86_64 glibc-2.11.1-6.x86_64
> libICE-1.0.6-1.fc12.x86_64 libSM-1.1.0-7.fc12.x86_64
> libX11-1.3-1.fc12.x86_64 libXau-1.0.5-1.fc12.x86_64
> libXext-1.1-2.fc12.x86_64 libXi-1.3-2.fc12.x86_64 libXmu-1.0.5-1.fc12.x86_64
> libXrandr-1.3.0-5.fc12.x86_64 libXrender-0.9.5-1.fc12.x86_64
> libXt-1.0.7-1.fc12.x86_64 libXtst-1.0.99.2-3.fc12.x86_64
> libXxf86vm-1.1.0-1.fc12.x86_64 libasyncns-0.8-1.fc12.x86_64
> libcap-ng-0.6.2-3.fc12.x86_64 libgcc-4.4.3-4.fc12.x86_64
> libogg-1.1.4-2.fc12.x86_64 libpng-1.2.43-1.fc12.x86_64
> libsndfile-1.0.20-3.fc12.x86_64 libstdc++-4.4.3-4.fc12.x86_64
> libuuid-2.16.2-9.fc12.x86_64 libvorbis-1.2.3-4.fc12.x86_64
> libxcb-1.5-1.fc12.x86_64 mesa-libGLU-7.7-4.fc12.x86_64
> openal-soft-1.12.854-1.fc12.x86_64 pulseaudio-libs-0.9.21-5.fc12.x86_64
> tcp_wrappers-libs-7.6-56.fc12.x86_64 zlib-1.2.3-23.fc12.x86_64
> (gdb) where
> #0 0x00a0850d in SGExpression::getValue (this=0x0,
> binding=0x7fffd830) at
> ../../../simgear/structure/SGExpression.hxx:123
> #1 0x00a2a180 in simgear::Technique::validateInContext (
> this=0x7fffe4873a10, gc=0x11707a0) at Technique.cxx:118
> #2 0x00a29a96 in simgear::(anonymous
> namespace)::ValidateOperation::operator() (this=0x12ddb5a0, gc=0x11707a0) at
> Technique.cxx:51
> #3 0x76579429 in osg::GraphicsContext::runOperations() ()
>from /usr/local/lib64/libosg.so.63
> #4 0x77459cdc in osgViewer::ViewerBase::renderingTraversals() ()
>from /usr/local/lib64/libosgViewer.so.63
> #5 0x00496695 in fgOSMainLoop () at fg_os_osgviewer.cxx:204
> #6 0x0043297c in fgMainInit (argc=1, argv=0x7fffe218)
> at main.cxx:945
> #7 0x0042ecae in main (argc=1, argv=0x7fffe218)
> at bootstrap.cxx:229
>
> There's no difference between Technique.cxx in the cvs vs. git
> repositories, The CVS version of flightgear runs fine against the fgdata-git
> tree. The git version of flightgear crashes. I do observer there are a
> bunch of shader/effect code differences between simgear-cvs and simgear-git.
> Could something new in that code be breaking FlightGear? Is anyone else
> having trouble running the git version of FlightGear? There is always a
> non-zero possibility I'm doing something stupid on my end, but I've tried to
> rule out all the stupid things I normally do before posting. (I've done a
> complete clean rebuild from scratch, reinstalled my graphics drivers,
> verified the cvs version of flightgear does run fine, made sure I've pulled
> the latest changes for simgear/flightgear/fgdata, rebooted, etc. etc. etc.)
>
Here is a follow up to my own message:
If I change line #118 in simgear/scene/material/Technique.cxx
from:
if (_validExpression->getValue(&binding))
newVal = VALID;
to:
if (_validExpression && _validExpression->getValue(&binding))
newVal = VALID;
Then flightgear comes up and runs. However, the urban areas are missing
(pointing to a problem with the urban shader effect ??? and there are other
weird shading/lighting issues.)
Thanks,
Curt.
--
Curtis Olson: http://baron.flightgear.org/~curt/
--
___
Flightgear-devel mailing list
Flightgear-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/flightgear-devel
[Flightgear-devel] FlightGear/SimGear (git) segfaults
Hi,
I am having a problem running the latest flightgear and simgear from the
gitorious repository. Here is a back trace of my crash. Please note that
the CVS version of FlightGear runs just fine on my system, and I've just
reinstalled the latest nvidia graphics drivers so I'm 99.99% confident this
isn't a system configuration problem.
OS = Fedora 12 (with all the latest system updates installed)
Graphics drivers = NVIDIA-Linux-x86_64-195.36.15-pkg2.run
CPU = 64-bit dual core athlon
kernel = 2.6.32.12-115.fc12.x86_64
OSG = 2.9.7
Here's a back-trace from gdb:
Program received signal SIGSEGV, Segmentation fault.
0x00a0850d in SGExpression::getValue (this=0x0,
binding=0x7fffd830) at
../../../simgear/structure/SGExpression.hxx:123
123 { T value; eval(value, binding); return value; }
Missing separate debuginfos, use: debuginfo-install
alsa-lib-1.0.23-1.fc12.x86_64 dbus-libs-1.2.16-9.fc12.x86_64
flac-1.2.1-6.fc12.x86_64 freealut-1.1.0-11.fc12.x86_64
freeglut-2.6.0-1.fc12.x86_64 glibc-2.11.1-6.x86_64
libICE-1.0.6-1.fc12.x86_64 libSM-1.1.0-7.fc12.x86_64
libX11-1.3-1.fc12.x86_64 libXau-1.0.5-1.fc12.x86_64
libXext-1.1-2.fc12.x86_64 libXi-1.3-2.fc12.x86_64 libXmu-1.0.5-1.fc12.x86_64
libXrandr-1.3.0-5.fc12.x86_64 libXrender-0.9.5-1.fc12.x86_64
libXt-1.0.7-1.fc12.x86_64 libXtst-1.0.99.2-3.fc12.x86_64
libXxf86vm-1.1.0-1.fc12.x86_64 libasyncns-0.8-1.fc12.x86_64
libcap-ng-0.6.2-3.fc12.x86_64 libgcc-4.4.3-4.fc12.x86_64
libogg-1.1.4-2.fc12.x86_64 libpng-1.2.43-1.fc12.x86_64
libsndfile-1.0.20-3.fc12.x86_64 libstdc++-4.4.3-4.fc12.x86_64
libuuid-2.16.2-9.fc12.x86_64 libvorbis-1.2.3-4.fc12.x86_64
libxcb-1.5-1.fc12.x86_64 mesa-libGLU-7.7-4.fc12.x86_64
openal-soft-1.12.854-1.fc12.x86_64 pulseaudio-libs-0.9.21-5.fc12.x86_64
tcp_wrappers-libs-7.6-56.fc12.x86_64 zlib-1.2.3-23.fc12.x86_64
(gdb) where
#0 0x00a0850d in SGExpression::getValue (this=0x0,
binding=0x7fffd830) at
../../../simgear/structure/SGExpression.hxx:123
#1 0x00a2a180 in simgear::Technique::validateInContext (
this=0x7fffe4873a10, gc=0x11707a0) at Technique.cxx:118
#2 0x00a29a96 in simgear::(anonymous
namespace)::ValidateOperation::operator() (this=0x12ddb5a0, gc=0x11707a0) at
Technique.cxx:51
#3 0x76579429 in osg::GraphicsContext::runOperations() ()
from /usr/local/lib64/libosg.so.63
#4 0x77459cdc in osgViewer::ViewerBase::renderingTraversals() ()
from /usr/local/lib64/libosgViewer.so.63
#5 0x00496695 in fgOSMainLoop () at fg_os_osgviewer.cxx:204
#6 0x0043297c in fgMainInit (argc=1, argv=0x7fffe218)
at main.cxx:945
#7 0x0042ecae in main (argc=1, argv=0x7fffe218)
at bootstrap.cxx:229
There's no difference between Technique.cxx in the cvs vs. git repositories,
The CVS version of flightgear runs fine against the fgdata-git tree. The
git version of flightgear crashes. I do observer there are a bunch of
shader/effect code differences between simgear-cvs and simgear-git. Could
something new in that code be breaking FlightGear? Is anyone else having
trouble running the git version of FlightGear? There is always a non-zero
possibility I'm doing something stupid on my end, but I've tried to rule out
all the stupid things I normally do before posting. (I've done a complete
clean rebuild from scratch, reinstalled my graphics drivers, verified the
cvs version of flightgear does run fine, made sure I've pulled the latest
changes for simgear/flightgear/fgdata, rebooted, etc. etc. etc.)
Thanks,
Curt.
--
Curtis Olson: http://baron.flightgear.org/~curt/
--
___
Flightgear-devel mailing list
Flightgear-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/flightgear-devel
