Re: [boost] [lockfree] review

Subject: Re: [boost] [lockfree] review
From: Alexander Terekhov (terekhov_at_[hidden])
Date: 2011-08-30 08:42:48

Hans Boehm wrote:
[...]
> > > plus a number of real lock-free algorithms (which do matter) don't
> > > work.

I'm really interested to know the exact number/names of algorithms.

[... "based on Dekker's example" ...]

Yes, acquire/release is free from implicit store-load barrier (same as
with TSO and even IBM 370).

Note that computer hardware vendors never provided and still do not
provide implicit store-load barrier for hardware.

In this respect I really like "Figure 8":

http://www.hpl.hp.com/techreports/Compaq-DEC/WRL-95-7.pdf
(Figure 8: Simple categorization of relaxed models. ...)

Why the heck is C++'s default mode for atomics so special in this
respect? 

regards,
alexander.