[cpp-threads] Memory model counterproposals

[Nelson, Clark]

The "room" seems to have grown quiet. That can't be good. So I'm going
to try stirring up the conversation again.

I have been in consultation with many colleagues on the Intel compiler
team, and we've reached some Intel positions with respect to the
proposed memory model. I will be posting separable counterproposals in
separate messages, to simplify thread management.

Having introduced my sequence of messages, perhaps I should also
introduce myself. I have been working on C and C++ compiler front ends
since 1987. I first became involved with the C standard committee in
that year, and still go to meetings occasionally. I have been on the C++
committee since 1991, and an officer thereof since 1996.

I am not an expert in concurrent programming or memory models (although
I've been learning fast), but I am well-versed in the C and C++
standards, and am fairly fluent in the dialects of standardese used
therein. My hope is that, once the substance of the memory model
proposal has been worked out in sufficient detail, I will be able to
help draft formal wording changes for it.

I should perhaps also mention that I'm going to be traveling for much of
June, starting Memorial Day, so it shouldn't be surprising if I fall
silent for a day or two at a time.

-- 
Clark Nelson		Vice chair, J16 (ANSI C++ standard committee)
Intel Corporation
clark.nelson at intel.com