Re: What are the arguments in favor of delay/force in eval.c?

[Rob Browning]

Ken Raeburn <address@hidden> writes:

> Of course, the compiler might reorder these accesses, unless you
> make everything volatile.  Even then, the CPU still might reorder
> them so another compiler might see something different.  I'd be
> worried about the reordering or caching in another processor if the
> code decides it can bypass all the mutex stuff, too.
>
> Once every thread agrees on the new data value and no mutex needed,
> everything should be fine, but managing the transition to that state
> (without adding *another* mutex) is very tricky.

Yeah.  I definitely wondered if there was a way to do it completely
safely without full protection.  I'm always hesitant about code that
relies on so many assumptions, though I (obviously) wasn't thinking
about the reordering issues.  (Of course, I pretty much always fully
protect things in real code, so it's not an issue...)

> Actually, I think some memory models will allow some accesses to be
> moved into the region where the lock is held from outside it.

Interesting.

Thanks
-- 
Rob Browning
rlb @defaultvalue.org and @debian.org; previously @cs.utexas.edu
GPG starting 2002-11-03 = 14DD 432F AE39 534D B592  F9A0 25C8 D377 8C7E 73A4