Re: [boost] [fiber] new version in vault

Subject: Re: [boost] [fiber] new version in vault
From: Phil Endecott (spam_from_boost_dev_at_[hidden])
Date: 2009-11-30 12:48:21

Helge Bahmann wrote:
> why would you ever want to use a mutex that does not properly inform the 
> system scheduler until when the calling thread should be suspended (in 
> case of contention) for "true" inter-thread (or even inter-process) 
> coordination?

When you believe that the probability of contention is very small, and 
you care only about average and not worst-case performance, and you 
wish to avoid the overhead (e.g. time or code or data size) of 
"properly informing" the system scheduler.

Earlier in this thread, Stefan Strasser wrote:
> a mutex template with a parameter on what to do to waste some time (this_fiber::yield,
> this_thread::yield, spinning...) could clean this up

Yes.  I implemented something like this that could either spin, or call 
sched_yield, or (on Linux) call futex_wait.  It's basically a 
Futex-based mutex that implements the other behaviours by replacing the 
futex with something trivial.  See 
http://svn.chezphil.org/libpbe/trunk/include/Mutex.hh and Futex.hh, 
Yield.hh, Spin.hh in the same directory.  I've discussed the relative 
performance of these things on this list in the past.

Regards,  Phil.