Reactive Spin-locks: A Self-tuning Approach

1
Reactive Spin-locks A Self-tuning Approach

• Phuong Hoai Ha
• Marina Papatriantafilou
• Philippas Tsigas

I-SPAN 05, Las Vegas, Dec. 7th 9th, 2005
2
Outline

• Mutual exclusion
• Overhead
• Available reactive spin-locks
• New reactive spin-lock
• Model
• Algorithm
• Evaluation
• Conclusions

3
Mutual exclusion
Entry section
Critical section
Exit section
Noncritical sec.
Requests issued
Lock released
Arbitration
Lock sent to winner

• Performance goals
• Low latency
• Low contention

4
Spin-lock categories

• Arbitrating locks
• Determine who is the next lock-holder in advance,
  e.g. ticket-locks, queue-locks.
• Advantages
• Prevent processors from causing bursts in network
  traffic and high contention on the lock.
• Non-arbitrating locks
• E.g. Test-and-set locks
• Advantages
• Exploit locality/cache
• Tolerate failures in the Entry section.

5
Arbitrating vs. non-arbitrating locks
1
3
5
Interconnection Network
Interconnection Network
2
4
6
6
Available reactive spin-lock algorithms

• Drawbacks
• Their reactive schemes rely on
• Fixed experimental thresholds
• The thresholds frequently become inappropriate in
  variable and unpredictable environments like
  multiprogramming systems
• E.g. ticket locks with proportional backoff,
  test-and-test-and-set locks with exponential
  backoff
• Known probability distributions of some inputs
• The assumption is not usually feasible.

7
New reactive spin-lock algorithm

• Ideas
• A non-arbitrating lock with adaptive sensible
  backoff delay.
• Advantages
• Its reactive scheme is self-tuning
• Neither experimentally tuned thresholds nor
  probability distributions of inputs are needed
• It combines advantages of both arbitrating and
  non-arbitrating spin-lock categories.
• It can exploit locality as well as reduce
  contention on the lock.

8
Find sensible backoff delay

• Need to optimize trade-off between
• Latency
• The interval between a pair of lock-release and
  lock-acquisition
• Contention on the lock
• This is an online problem.

?delay?
Load on the lock
9
Reactive scheme

• Bounds for loads on the lock 1 ? lt ? P
• During a load-rising phase

• Similar for load-dropping phase
• In each load-rising/load-dropping phase, the
  reactive scheme is competitive with competitive
  ration c?(ln(P))

10
Algorithm

• The algorithm guarantees mutual exclusion and
  non-livelock. Its space complexity is log(P).

0
1
3
4
2
0
1
Interconnection Network
3
2
11
Evaluation

• Benchmarks
• Spark98 kernel lmv
• SPLASH-2 suite Volrend and Radiosity
• Representatives
• Arbitrating ticket lock with (tuned)
  proportional backoff
• Non-arbitrating test-and-test-and-set lock with
  (tuned) exponential backoff
• System
• A ccNUMA SGI Origin2000 with 28 250MHz MIPS R1000
  processors.

12
Experimental results
13
Experimental results (2)
14
Experimetal results (3)
15
Conclusions

• We have designed and implemented a new reactive
  spin-lock
• It is self-tuning.
• It combines advantages of both arbitrating and
  non-arbitrating locks
• Its reactive scheme is competitive with c
  ?(ln(P))
• ? The lock automatically adjusts its backoff
  delay reasonably according to loads on the lock
  as well as applications

16
Thanks for your attention!
17
Estimate delay bases

• Fairness
• A fair lock helps parallel application gain
  performance since the application threads can
  execute their non-critical section in parallel.
• Definition
• Heuristic to estimate basel

, where ni is lock-acquisitions of a
processor in ?t and N is processors
, where a, b are system documented constants and
DoCS is the delay outside CS
18
NUMA

• Another parameter that makes the problem harder
  is NUMA
• Latency is much different
• E.g. ccNUMA SGI Origin2000

19
Model An online problem

• A sequence of loads on the lock are unfolded
  on-the-fly.
• When observing a load, the algorithm must decide
  how much its current backoff delay should be
  lengthened.
• If increasing delay too soon, it will waste time
  on a long delay when the lock becomes available
• If not increasing delay in time, it will cause
  high contention on the lock
• ? it must increase delay at high loads
  reasonably
• ? Goal is to maximize ?t ?delayt .loadt ,where
  ?t ?delayt ? P

20
Algorithm

• LockType
• ltlock, countergt
• Initial delay L.counter x basel
• The algorithm guarantees mutual exclusion and
  non-livelock. Its space complexity is log(P).

• Acquire( Lock pL)
• L FAA(pL.L, lt1,1gt)
• if L.lock then
• delay ComputeDelay(L)
• cond lt1,0gt
• do
• sleep(delay)
• L pL.L
• if L.lock then
• delay ComputeDelay(L)
• continue
• cond FAA(pL.L, lt1,0gt)
• while cond.lock
• Release( Lock pL)
• do L pL.L
• while not CAS(pL.L,L,lt0,L.counter-1gt)