[cfe-dev] plans about kernel compilation?

[James Widman]

On Oct 28, 2007, at 12:30 PM, Scott Wood wrote:

> On Sun, Oct 28, 2007 at 12:18:26AM -0700, Giangiacomo Mariotti wrote:
<snip>
> "Is cfe planning on implementing optimizations that cause
> memory locations to be written to even if the condition controlling
> the write is false (or more generally, optimizations that are hostile
> to significant amounts of existing threaded code), with no option to
> disable them, and with no alternative less sucky than volatile or
> inline assembly?".

People interested in the more general issue of proper concurrency  
support (which subsumes this specific issue of hostile optimizations)  
might want to look at the proposals (N2429 and N2427) for a multi- 
threaded memory model and atomic operations for C++1x (which the ISO C 
++ committee formally adopted earlier this month).

For background information, see the multiple instances of "memory  
model" and "atomic" in this page:

http://open-std.org/JTC1/SC22/WG21/docs/papers/2007/n2389.html

...and give yourself several weeks to digest it all.  (:

Note, N2429 and N2427 will be publicly visible when the next WG21  
mailing is released (any day now), at which time you can grab them  
from here:

http://open-std.org/JTC1/SC22/WG21/docs/papers/2007/

That mailing will also include an updated version of the Working  
Draft of the next C++ Standard, and it will have the proposed wording  
from N2429 and N2427 merged in (with LaTeX-diff-o-vision!).

James Widman
-- 
Gimpel Software
http://gimpel.com