Re: [boost] [thread] best practice to lock multiple mutexes

From: Anthony Williams (anthony_w.geo_at_[hidden])
Date: 2008-02-15 07:34:35

Kowalke Oliver (QD IT PA AS <Oliver.Kowalke <at> qimonda.com> writes:

> 
> > Andrey Tcherepanov <moyt63c02 <at> sneakemail.com> writes:
> >
> > > I sent you results to anthony_w.geo<at>yahoo.com, please
> > let me know
> > > if you received them or not - these corp. filters play
> > strange games
> > > sometimes...
> >
> > Received, thank you.
> 
> Would you share the data?

Preliminary results indicate that the effects of adding yield() to before a retry
when locking multiple mutexes are unpredictable in general. In some cases there
was a 58% drop in run time (i.e. better performance) with the yield. In other
cases there was a 16% increase in run time (i.e. worse performance) with the
yield. In many cases the standard deviation in the measurements was larger than
the effect.

I will share more details when I've analyzed all of Andrey's data, and my own
data.

Anthony