Subject: Re: [boost] Synchronization (RE: [compute] review)
From: Thomas M (firespot71_at_[hidden])
Date: 2014-12-31 04:04:49
On 31/12/2014 09:15, Gruenke,Matt wrote:
> I think our positions are fairly entrenched, at this point. But don't let me stop you from trying to convince Kyle. In the end, his is the only opinion that matters.
Let me throw in a proposal to dissolve our entrenchment ;):
boost::compute::...::guarantee is a true guarantee and refers only to
events (or whatever can fulfill a true guarantee);
boost::compute::...::wait (... being e.g. a [event,] wait_list,
command_queue, all sorts of stuff) does basically the same, i.e. wait,
but the name makes clear it isn't a full guarantee. Thus it provides
convenience but doesn't prevent you from shooting yourself in the foot
if you are careless - sounds like perfect C++ philosophy ;).
I am very open to different names (wait, waiter, wait_lock, wait_on_me,
eaahhhh ... anything short + concise much appreciated).
Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk