[PATCH] mieq: Provide better adaptability and diagnostics during mieq overflow

[Jeremy Huddleston]

On Oct 16, 2011, at 1:47 PM, Jeremy Huddleston wrote:
>> I'm not sure we need to be even this verbose; just a single warning when
>> the queue is blocked and another one when it starts working again should
>> be sufficient. I fear flooding the disk with errors if this happens
>> because your cat falls asleep on the keyboard while watching dancing
>> mice on the screen.
> 
> That's why we shut up after DROP_BACKTRACE_MAX reports.  What about lowering DROP_BACKTRACE_MAX to 5 rather than 20?

Also, if your cat falls asleep on the keyboard, the system either can keep up with the generated events (no logging before or after this patch) or it fills up the queue and we have dropped events.  Both before and after this patch, we print a backtrace when the first event is dropped.  After this patch, we also print a backtrace after 100 consecutive events have been dropped. If your system can process 1 event for every 100 kitten-generated events that come in, the backtrace spam will be no worse than it is now.