In oom_kill.c, badness() counts half the total_vm of all the children against the parent.
The problem with this calculation is that, if the parent and children share memory, the shared memory is over-counted N/2 times where N is the number of children. There's already a hard limit on the amount of shared memory, so it's unlikely that process is runaway due to shared memory consumption, and shouldn't add to a process's badness() score. Shouldn't it just use (total_vm - shared_vm) instead, and never even count shared memory? Regards, Jeff Davis - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/