[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [Pan-users] Re: Debugging Pan with gdb
From: |
Douglas Bollinger |
Subject: |
Re: [Pan-users] Re: Debugging Pan with gdb |
Date: |
Tue, 22 Aug 2006 18:15:54 -0400 |
On Tue, 22 Aug 2006 14:09:46 +0000 (UTC)
"Duncan" <address@hidden> wrote:
<snip>
> For x86 (and therefore for 32-bit x86 compatible code on amd64), omitting
> the frame-pointer DOES affect debugging, but ONLY on the libraries and
> executables where the flag is used. (As a result it's not enabled by any
> -O level and must be invoked directly, again, gcc manpage.)
>
> Assuming you are on x86, then, compiling pan without -fomit-frame-pointer
> should enable you to debug pan's own functions. As soon as you hit the
> gtk or other libraries (glibc certainly, gcc's libstdc++ also since pan
> is now C++, other dependencies) compiled with -fomit-frame-pointer,
> however, you'll be in black-box territory -- you won't be able to trace
> into them. Whether that kills debugging then depends on whether you can
> get enough info from what /is/ frame-pointered to be useful troubleshooting
> your particular problem, or not.
<snip>
Ok guys, I compiled gtk+ without the -fomit-frame-pointer flag. Like you
said, it doesn't take very long. I got some more information on the
backtrace:
(gdb) run
Starting program: /home/doug/pan-0.109/pan/gui/pan
[Thread debugging using libthread_db enabled]
[New Thread -1218734416 (LWP 9557)]
Program received signal SIGSEGV, Segmentation fault.
[Switching to Thread -1218734416 (LWP 9557)]
0xb7e9ce0e in gtk_tree_view_set_fixed_height_mode ()
from /usr/lib/libgtk-x11-2.0.so.0
(gdb) bt
#0 0xb7e9ce0e in gtk_tree_view_set_fixed_height_mode ()
from /usr/lib/libgtk-x11-2.0.so.0
#1 0xb7ea9879 in gtk_tree_view_set_model () from /usr/lib/libgtk-x11-2.0.so.0
#2 0xb7dc6bf1 in gtk_marshal_VOID__UINT_STRING ()
from /usr/lib/libgtk-x11-2.0.so.0
#3 0xb7b2eea9 in g_closure_invoke () from /usr/lib/libgobject-2.0.so.0
#4 0xbfac0068 in ?? ()
#5 0x00000000 in ?? ()
I'll add this to the bug report and see if it helps.
Thanks for the gdb info!
--
We can found no scientific discipline, nor a healthy profession on the
technical mistakes of the Department of Defense and IBM.
-- Edsger Dijkstra