visualpython-users

[Visualpython-users] Visual 5 from cvs - segmentation fault

[LUK S. <shu...@po...>]

Hi,

I checked out vpython-core2 from cvs and it built fine in
debian/sid/amd64. However it seg faulted when I tried with the examples.
Here's a gdb backtrace when running the bounce example.

<gdb-bt>
#0  0x00007f70947af020 in ?? ()
#1  0x00007f70a0f5e1b3 in gdk_gl_get_proc_address () from
/usr/lib/libgdkglext-x11-1.0.so.0
#2  0x00007f70a18dbd0c in cvisual::display_kernel::realize (this=0x19933b0)
    at ./core/display_kernel.cpp:327
#3  0x00007f70a18dc815 in cvisual::display_kernel::render_scene
(this=0x7f70a199d068)
    at ./core/display_kernel.cpp:727
#4  0x00007f70a1900606 in cvisual::render_surface::paint (this=0x1d34eb0)
    at ./gtk2/render_surface.cpp:181
#5  0x00007f70a18c6e74 in cvisual::render_manager::paint_displays
(displays=@0x1725610,
    swap_single_threaded=true) at ../include/gtk2/display.hpp:38
#6  0x00007f70a18f9231 in cvisual::gui_main::poll (this=0x1725540) at
./gtk2/display.cpp:390
#7  0x00007f709e4e09b2 in ?? () from /usr/lib/libglibmm-2.4.so.1
#8  0x00007f709b7695bb in ?? () from /usr/lib/libglib-2.0.so.0
#9  0x00007f709b768e4a in g_main_context_dispatch () from
/usr/lib/libglib-2.0.so.0
#10 0x00007f709b76c510 in ?? () from /usr/lib/libglib-2.0.so.0
#11 0x00007f709b76c9dd in g_main_loop_run () from /usr/lib/libglib-2.0.so.0
#12 0x00007f709dde7737 in gtk_main () from /usr/lib/libgtk-x11-2.0.so.0
#13 0x00007f70a18fa5b2 in cvisual::gui_main::run (this=0x1725540) at
./gtk2/display.cpp:378
#14 0x00007f70a18fa669 in cvisual::gui_main::thread_proc () at
./gtk2/display.cpp:407
#15 0x00007f70a18c7f97 in boost::function0<void,
std::allocator<boost::function_base> >::operator() (this=<value
optimized out>) at /usr/include/boost/function/function_template.hpp:825
#16 0x00007f709b2e60d2 in ?? () from
/usr/lib/libboost_thread-gcc42-mt-1_34_1.so.1.34.1
#17 0x00007f70a6ed9faa in start_thread () from /lib/libpthread.so.0
---Type <return> to continue, or q <return> to quit---
#18 0x00007f70a65c42cd in clone () from /lib/libc.so.6
#19 0x0000000000000000 in ?? ()
</gdb-bt>

Please let me know what further information should I provide so that the
source of the fault can be traced.

Regards,
ST
--

Re: [Visualpython-users] Visual 5 from cvs - segmentation fault

[Bruce S. <Bru...@nc...>]

I don't know what the problem is. Since the build went okay, I assume 
that there's nothing unusual to be seen in config.log or src/build.log? 
Does someone reading this have experience running Visual 5 on Debian?

Since the fault is in the thread area, I draw your attention to this 
comment in INSTALL.txt in CVS about matching Boost and threadpool, 
though I think a mismatch would have stopped the build:

You also need the threadpool resource:
sourceforge.net/projects/threadpool.
If you are using the 1_35_0 Boost libraries, you can use version 0.2.4 
which is included in the package (in dependencies). If you use a 
different version of the Boost libraries, you need to get an appropriate 
threadpool version and replace the files in
"dependencies/threadpool/include". Copy into the include directory the 
contents of the boost directory in the threadpool package, which 
includes a directory named "include" and a file "threadpool.hpp".

Bruce Sherwood

LUK ShunTim wrote:
> Hi,
> 
> I checked out vpython-core2 from cvs and it built fine in
> debian/sid/amd64. However it seg faulted when I tried with the examples.
> Here's a gdb backtrace when running the bounce example.
> 
> <gdb-bt>
> #0  0x00007f70947af020 in ?? ()
> #1  0x00007f70a0f5e1b3 in gdk_gl_get_proc_address () from
> /usr/lib/libgdkglext-x11-1.0.so.0
> #2  0x00007f70a18dbd0c in cvisual::display_kernel::realize (this=0x19933b0)
>     at ./core/display_kernel.cpp:327
> #3  0x00007f70a18dc815 in cvisual::display_kernel::render_scene
> (this=0x7f70a199d068)
>     at ./core/display_kernel.cpp:727
> #4  0x00007f70a1900606 in cvisual::render_surface::paint (this=0x1d34eb0)
>     at ./gtk2/render_surface.cpp:181
> #5  0x00007f70a18c6e74 in cvisual::render_manager::paint_displays
> (displays=@0x1725610,
>     swap_single_threaded=true) at ../include/gtk2/display.hpp:38
> #6  0x00007f70a18f9231 in cvisual::gui_main::poll (this=0x1725540) at
> ./gtk2/display.cpp:390
> #7  0x00007f709e4e09b2 in ?? () from /usr/lib/libglibmm-2.4.so.1
> #8  0x00007f709b7695bb in ?? () from /usr/lib/libglib-2.0.so.0
> #9  0x00007f709b768e4a in g_main_context_dispatch () from
> /usr/lib/libglib-2.0.so.0
> #10 0x00007f709b76c510 in ?? () from /usr/lib/libglib-2.0.so.0
> #11 0x00007f709b76c9dd in g_main_loop_run () from /usr/lib/libglib-2.0.so.0
> #12 0x00007f709dde7737 in gtk_main () from /usr/lib/libgtk-x11-2.0.so.0
> #13 0x00007f70a18fa5b2 in cvisual::gui_main::run (this=0x1725540) at
> ./gtk2/display.cpp:378
> #14 0x00007f70a18fa669 in cvisual::gui_main::thread_proc () at
> ./gtk2/display.cpp:407
> #15 0x00007f70a18c7f97 in boost::function0<void,
> std::allocator<boost::function_base> >::operator() (this=<value
> optimized out>) at /usr/include/boost/function/function_template.hpp:825
> #16 0x00007f709b2e60d2 in ?? () from
> /usr/lib/libboost_thread-gcc42-mt-1_34_1.so.1.34.1
> #17 0x00007f70a6ed9faa in start_thread () from /lib/libpthread.so.0
> ---Type <return> to continue, or q <return> to quit---
> #18 0x00007f70a65c42cd in clone () from /lib/libc.so.6
> #19 0x0000000000000000 in ?? ()
> </gdb-bt>
> 
> Please let me know what further information should I provide so that the
> source of the fault can be traced.
> 
> Regards,
> ST
> --
> 
> 
> ------------------------------------------------------------------------------
> _______________________________________________
> Visualpython-users mailing list
> Vis...@li...
> https://lists.sourceforge.net/lists/listinfo/visualpython-users

[Visualpython-users] Visual 5 from cvs - segmentation fault on amd64 boxes [Was: [Re: Visual 5 from cvs - segmentation fault]]

[LUK S. <shu...@po...>]

Bruce Sherwood wrote:
> I don't know what the problem is. Since the build went okay, I assume
> that there's nothing unusual to be seen in config.log or src/build.log?
> Does someone reading this have experience running Visual 5 on Debian?

Yes, that would be very much appreciated.

I tried the cvs build because the debian supplied package (which uses
boost 1.35) didn't work. Same seg fault. I filed a bug

http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=520472

but hasn't yet got a reply.

> 
> Since the fault is in the thread area, I draw your attention to this
> comment in INSTALL.txt in CVS about matching Boost and threadpool,
> though I think a mismatch would have stopped the build:

Yes, with threadpool 0.2.3 and 0.2.5, the build couldn't complete (with
boost 1.34).

I guess it's a 64-bit related problem as both the debian supplied and
locally built versions worked on my i386 laptop. I've changed the
subject line.

Regards,
ST
--