[PATCH] Tree SRA and atomicity/volatility

[Mark Mitchell]

Mike Stump wrote:
> On Jan 25, 2007, at 3:44 AM, Richard Kenner wrote:
>> But we still can't try to express it in any semantically-meaningful way.
> 
> :-(  I disagree fundamentally with this notion.  gcc, being an
> implementation, is free to talk about opcodes produced or not on i686,
> if it wants.

Sure, that's doable in principle.

But, once you write it down, you get to figure out how to make sure that
the entire compiler honors it, including the machine-independent
TREE-SSA bits.  Since the sensible choices may be different for each
CPU, how to enforce the right constraints (either separately for each
CPU, or some superset of all of them) may not be easy.

-- 
Mark Mitchell
CodeSourcery
mark@codesourcery.com
(650) 331-3385 x713