On Jul 30, 2011, at 9:44 PM, Dave Abrahams <dave@boostpro.com> wrote:
No program of substantial size can be understood "dynamically." You can't think about code paths and timing and control flow; there are just too many combinations and too much to keep track of. You have to think of static things like preconditions, postconditions, and invariants. That's why purely functional languages are easier to reason about, and why const and value semantics are important in languages with mutation.
Right, sequential consistency in itself is too low level to help understanding --
That's also why it's always seemed to me that thinking in terms of interleavings of instructions doesn't lead to any fundamentally improved ability to understand parallelism: it doesn't change anything from a static point of view. The basic issue is still one of protecting one thread from observing the broken invariants of another.
-- but without sequential consistency, the compiler may be rearranging instructions in ways that wouldn't affect a single thread (as allowed by c++03), and threads may see each others' actions in the wrong order, so a weak memory model can actually break invariants even if the code is good.