Backtrace not giving meaningful info

Backtrace not giving meaningful info

[jlm]

I'm using gdb 5.3 to debug a C++ program that has been compiled with gcc
3.2 using the -gdwarf-2 -g3 flags. I can reliably make my program crash,
and at that point, I want to know where my program crashed at, so I do a
backtrace:

(gdb) bt
#0  0x4007cbd8 in ?? ()
#1  0x4007b208 in ?? ()
#2  0x40079560 in ?? ()
#3  0x4008e300 in ?? ()
#4  0x4008d44c in ?? ()
#5  0x4008e9f7 in ?? ()
#6  0x4006da3f in ?? ()
#7  0x4006d25a in ?? ()
#8  0x40082a2b in ?? ()
#9  0x4008fc30 in ?? ()
#10 0x4008fa6f in ?? ()
#11 0x408d12c3 in
Glib::SignalProxyNormal::slot0_void_callback(_GObject*, void*)
(self=0x818d180, data=0xbfffd764) at signalproxy.cc:101

I have two questions. One, why doesn't the backtrace go all the way back
to main()? Two, what must I do in order to get some more meaningful
information in the backtrace? The ?? doesn't help me much.

Thanks.

Re: Backtrace not giving meaningful info

[Daniel Jacobowitz]

On Sat, Feb 01, 2003 at 10:38:52AM -0500, jlm wrote:
> I'm using gdb 5.3 to debug a C++ program that has been compiled with gcc
> 3.2 using the -gdwarf-2 -g3 flags. I can reliably make my program crash,
> and at that point, I want to know where my program crashed at, so I do a
> backtrace:
> 
> (gdb) bt
> #0  0x4007cbd8 in ?? ()
> #1  0x4007b208 in ?? ()
> #2  0x40079560 in ?? ()
> #3  0x4008e300 in ?? ()
> #4  0x4008d44c in ?? ()
> #5  0x4008e9f7 in ?? ()
> #6  0x4006da3f in ?? ()
> #7  0x4006d25a in ?? ()
> #8  0x40082a2b in ?? ()
> #9  0x4008fc30 in ?? ()
> #10 0x4008fa6f in ?? ()
> #11 0x408d12c3 in
> Glib::SignalProxyNormal::slot0_void_callback(_GObject*, void*)
> (self=0x818d180, data=0xbfffd764) at signalproxy.cc:101
> 
> I have two questions. One, why doesn't the backtrace go all the way back
> to main()? Two, what must I do in order to get some more meaningful
> information in the backtrace? The ?? doesn't help me much.

Not enough information.  You need to figure out where those addresses
actually are; are they in a shared library somewhere?  Try "info
shared".

-- 
Daniel Jacobowitz
MontaVista Software                         Debian GNU/Linux Developer

Re: Backtrace not giving meaningful info

[jlm]

Thanks for the reply. Oddly enough, I ran it again and it now displays
complete info in the backtrace. Not sure what I've done different.

Also, seems like I can't debug my program via libtool as it says I
should be able to like:

libtool gdb foo_test

It says "No executable file specified." Instead I'm doing:

cd .libs
LD_PRELOAD=./foolib.so gdb foo_test

Oh, well. At least it works some way.

On Sat, 2003-02-01 at 12:12, Daniel Jacobowitz wrote:
> On Sat, Feb 01, 2003 at 10:38:52AM -0500, jlm wrote:
> > I'm using gdb 5.3 to debug a C++ program that has been compiled with gcc
> > 3.2 using the -gdwarf-2 -g3 flags. I can reliably make my program crash,
> > and at that point, I want to know where my program crashed at, so I do a
> > backtrace:
> > 
> > (gdb) bt
> > #0  0x4007cbd8 in ?? ()
> > #1  0x4007b208 in ?? ()
> > #2  0x40079560 in ?? ()
> > #3  0x4008e300 in ?? ()
> > #4  0x4008d44c in ?? ()
> > #5  0x4008e9f7 in ?? ()
> > #6  0x4006da3f in ?? ()
> > #7  0x4006d25a in ?? ()
> > #8  0x40082a2b in ?? ()
> > #9  0x4008fc30 in ?? ()
> > #10 0x4008fa6f in ?? ()
> > #11 0x408d12c3 in
> > Glib::SignalProxyNormal::slot0_void_callback(_GObject*, void*)
> > (self=0x818d180, data=0xbfffd764) at signalproxy.cc:101
> > 
> > I have two questions. One, why doesn't the backtrace go all the way back
> > to main()? Two, what must I do in order to get some more meaningful
> > information in the backtrace? The ?? doesn't help me much.
> 
> Not enough information.  You need to figure out where those addresses
> actually are; are they in a shared library somewhere?  Try "info
> shared".
> 
> -- 
> Daniel Jacobowitz
> MontaVista Software                         Debian GNU/Linux Developer