Re: Comments on the thread library

From: williamkempf_at_[hidden]
Date: 2001-03-27 09:50:45

--- In boost_at_y..., Jens Maurer <Jens.Maurer_at_g...> wrote:
> williamkempf_at_h... wrote:
> > [Jens Maurer wrote:]
> [ Interfaces vs. performance]
> >  Unless you think the interface prevents creating
> > efficient implementations
> 
> So I believe.

I've been thinking about this a bit.  With an increase in size (some 
people dislike this as well, but some tradeoffs have to be made) I 
could get the pthread implementation to run as fast (or close to, 
basically as fast as the current fast_mutex) as standard pthreads 
mutexes.  This helps to illustrate why I think the interface doesn't 
really prevent creating efficient implementations.  Only if an 
interface truly makes writing a portable implementation prohibitive 
should we consider the interface to be in need of redesign (well, 
from an implementation stand point, not from a usability/safety/etc. 
stand point).

If you're interested in the faster implementation I have in mind I 
can go into depth on it in private.  I could implement it as well, 
but it requires the error checked and recursive mutex types that are 
a new edition (or a soon to be addition?) of the pthreads standard 
and my version of pthreads (pthreads-win32) doesn't support them.  So 
I'll need detailed documentation on the new standard and someone 
willing to compile and test what I code.

Bill Kempf