I run @ 35k jobs every Sunday.  THis is to monitor several ASIC cell 
libraries.  I get several core dumps.

If I know which program caused the dump I know I can do a :
gdb program core
 > bt

To get the trace.

of 50+ different programs with N levels of wrapper scripts, determining
which program failed is impossible.

What other tool/s exist to evalute a core dump that will provide a back 
trace like gdb?

-- 
    ___  _ ____       ___       __  __
   / _ )(_) / /_ __  / _ \___ _/ /_/ /____  ___
  / _  / / / / // / / ___/ _ `/ __/ __/ _ \/ _ \
/____/_/_/_/\_, / /_/   \_,_/\__/\__/\___/_//_/
            /___/
Texas Instruments ASIC Circuit Design Methodlogy Group
Dallas, Texas, 214-480-4455,  b-patton@ti.com

"Billy N. Patton" <b-patton@ti.com> writes:

> I run @ 35k jobs every Sunday.  THis is to monitor several ASIC cell
> libraries.  I get several core dumps.
>
> If I know which program caused the dump I know I can do a :
> gdb program core
>  > bt
>
> To get the trace.
>
> of 50+ different programs with N levels of wrapper scripts, determining
> which program failed is impossible.
>
> What other tool/s exist to evalute a core dump that will provide a
> back trace like gdb?

The core dump will often contain the command line used to start the
program that caused it.  Try the "file" command on a core file.  If
that gives nothing, examine it with "strings" and look for familiar
things.

-- 
M�ns Rullg�rd
mru@kth.se

Billy N. Patton wrote:

> I run @ 35k jobs every Sunday.  THis is to monitor several ASIC cell 
> libraries.  I get several core dumps.
> 
> If I know which program caused the dump I know I can do a :
> gdb program core
>  > bt
> 
> To get the trace.
> 
> of 50+ different programs with N levels of wrapper scripts, determining
> which program failed is impossible.
> 
> What other tool/s exist to evalute a core dump that will provide a back 
> trace like gdb?
> 

Assuming that you're on a Linux platform, maybe this patch will help?
    http://www.sinz.org/Michael.Sinz/Linux/core.html

boa@home

boa <root@localhost.com> writes:

> Billy N. Patton wrote:
>
>> I run @ 35k jobs every Sunday.  THis is to monitor several ASIC cell
>> libraries.  I get several core dumps.
>> If I know which program caused the dump I know I can do a :
>> gdb program core
>>  > bt
>> To get the trace.
>> of 50+ different programs with N levels of wrapper scripts,
>> determining
>> which program failed is impossible.
>> What other tool/s exist to evalute a core dump that will provide a
>> back trace like gdb?
>>
>
> Assuming that you're on a Linux platform, maybe this patch will help?
>     http://www.sinz.org/Michael.Sinz/Linux/core.html

The functionality of that patch is already in recent Linux kernels.
The command

  echo 'core.%e.%p' > /proc/sys/kernel/core_pattern

will cause core dumps to contain the name of the executable that
caused it (%e) and its pid (%p).  Other % codes are available as well.

-- 
M�ns Rullg�rd
mru@kth.se

On Mon, 09 Aug 2004 10:09:04 -0500, Billy N. Patton wrote:

> What other tool/s exist to evalute a core dump that will provide a back 
> trace like gdb?

ptrace on Solaris.

-- 
mail1dotstofanetdotdk