lkml.org 
[lkml]   [2010]   [Mar]   [31]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
Date
From
SubjectRe: RFC: Ideal Adaptive Spinning Conditions
Steven Rostedt wrote:
> On Wed, 2010-03-31 at 16:21 -0700, Darren Hart wrote:
>
>> o What type of lock hold times do we expect to benefit?
>
> 0 (that's a zero) :-p
>
> I haven't seen your patches but you are not doing a heuristic approach,
> are you? That is, do not "spin" hoping the lock will suddenly become
> free. I was against that for -rt and I would be against that for futex
> too.

I'm not sure what you're getting at here. Adaptive spinning is indeed
hoping the lock will become free while you are spinning and checking
it's owner...

>
>> o How much contention is a good match for adaptive spinning?
>> - this is related to the number of threads to run in the test
>> o How many spinners should be allowed?
>>
>> I can share the kernel patches if people are interested, but they are
>> really early, and I'm not sure they are of much value until I better
>> understand the conditions where this is expected to be useful.
>
> Again, I don't know how you implemented your adaptive spinners, but the
> trick to it in -rt was that it would only spin while the owner of the
> lock was actually running. If it was not running, it would sleep. No
> point waiting for a sleeping task to release its lock.

It does exactly this.

> Is this what you did? Because, IIRC, this only benefited spinlocks
> converted to mutexes. It did not help with semaphores, because
> semaphores could be held for a long time. Thus, it was good for short
> held locks, but hurt performance on long held locks.

Trouble is, I'm still seeing performance penalties even on the shortest
critical section possible (lock();unlock();)

> If userspace is going to do this, I guess the blocked task would need to
> go into kernel, and spin there (with preempt enabled) if the task is
> still active and holding the lock.

It is currently under preempt_disable() just like mutexes. I asked Peter
why it was done that way for mutexes, but didn't really get an answer.
He did point out that since we check need_resched() at every iteration
that we won't run longer than our timeslice anyway, so it shouldn't be a
problem.

> Then the application would need to determine which to use. An adaptive
> spinner for short held locks, and a normal futex for long held locks.

Yes, this was intended to be an optional thing (and certainly not the
default).


--
Darren Hart
IBM Linux Technology Center
Real-Time Linux Team


\
 
 \ /
  Last update: 2010-04-01 04:15    [from the cache]
©2003-2011 Jasper Spaans