Re: address@hidden (Marcus G. Daniels)] Re: Error

[David Sumpter]

Yes, the problem with the debugger is that there is no stack left after
the crash. I should have mentioned that. I will put in breakpoints and get
back to you.

I don't use the Archiver.

Regards,

David.

On 29 Nov 2001, Marcus G. Daniels wrote:

> >>>>> "DS" == David Sumpter <address@hidden> writes:
>
> DS> I don't actually know what a 'backtrace' is
>
> A backtrace is the annotated listing of the call stack upon a crash.
> To get one, it's necessary to use GDB, the debugger.  The usage
> is like this:
>
> $ gdb emigrate
> (gdb) run
> (gdb) bt
>
> Sometimes it can happen that a crash is so obnoxious that it wipes
> out the call stack, and then you must add breakpoints to prevent
> it from doing that.  The usage for that is like this:
>
> (gdb) break main.m:300
>
> That means, "set a break point in the file main.m on line 300".  When
> line 300 is hit, the debugger will have the program stopped on that line.
> Then you can look at variables like so:
>
> (gdb) print someVariable
>
> ..or restart (to see if the next break point somewhere else gets hit):
>
> (gdb) c
>
> DS> or what it is to 'serialize an object'.
>
> To use Archiver methods like -putDeep:object:.  I was just speculating.
> The backtrace should show what code triggered the crash.  If the backtrace
> is not there, you'll have to incrementally set breakpoints to see how
> far it gets through your code.  That can pretty much be done by bisection.
>


                  ==================================
   Swarm-Support is for discussion of the technical details of the day
   to day usage of Swarm.  For list administration needs (esp.
   [un]subscribing), please send a message to <address@hidden>
   with "help" in the body of the message.