[DGD]callout order

[balduin at uni-paderborn.de]

Hi !

dworkin wrote
[...]
> I am considering removing some guarantees on callout order as they
> exist now: at present the order in which short-term callouts (integer
> delay <= 127 seconds, including 0) are executed, when the delay
> has passed, is the order in which they were created.  I'd like
> to remove that guarantee: if you start two callouts with a 5
> second delay immediately after eachother, they could be executed
> in any order.  However, they <will> both be executed before a callout
> delayed for 6 seconds.

I don't have a problem with that, in fact i never took for granted that
call_outs are executed in any given order. But what I'd like to see
is a function that allows to execute code immediatly after a given thread
ends. Without having this code to concurr with any other call_out(,0,), 
probably executed even before a dump_state().

> 
> What I propose above is already the case for long-term (>= 128 seconds)
> and millisecond callouts.
> 
> This change would substantially speed up DGD on a multi-processor
> system, support for which I will be adding immediately after I
> complete the implementation of atomic functions.
> 
> Comments?

Whow, go ahead, i will immediatly test your first multi-processor dgd
as fast as i can grab a release.

> 
Ludger


List config page:  http://list.imaginary.com/mailman/listinfo/dgd