> However we have now noticed one interesting data point. When watching "top",
> the shared memory field usually wobbles around 1200000K, but sometimes
> spikes upward. I have just now seen it hit over 1800000K. So I suspect our
> crashes happen when the shared memory number passes 2097152K, indicating
> some signed 32 bit int overflowing and doing nasty things.
Is it possible that your system actually running out of memory or
processes? -- what are the other stats of the system prior to the crash?
Also, please explain what you mean by "crashes" -- the term itself is far
too broad. If there is an Oops, the results of running it through
ksymoops (against the correct System.map et al) will help pinpoint what is
going wrong. Also, what is the configuration of your hardware?
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to firstname.lastname@example.org
Please read the FAQ at http://www.tux.org/lkml/