[concurrency-interest] regarding StampedLock
dl at cs.oswego.edu
Fri Oct 19 12:57:35 EDT 2012
On 10/15/12 14:38, Doug Lea wrote:
> In your tests, the reader blocks are so quick that
> lots of them can get in and out before writer queuing
> stabilizes so that phase-fair rules can kick in, at which point the writer
> will often hit an expensive context switch and have the lock stolen away
> while it is in the process of waking up. There are some mechanics for
> reducing/avoiding this that I'll put in.
In case anyone is waiting for these updates, they won't happen
immediately. Adaptively bridging the factor-of-10,000 rate gap to
improve performance in some programs without hurting others is
an interesting challenge. I'll be out to OOPSLA/Splash most of
next week though, so probably won't commit updates for a while.
More information about the Concurrency-interest