Re: Loops and parallel execution

["Yannick Duchêne (Hibou57)" <yannick_duchene@yahoo.fr>]

Le Tue, 25 Jan 2011 22:32:57 +0100, Dmitry A. Kazakov  
<mailbox@dmitry-kazakov.de> a écrit:

> On Tue, 25 Jan 2011 18:36:29 +0100, Georg Bauhaus wrote:
>
>> If a compiler detects two independent paths inside a loop,
>
> I don't see any application for this. Can you remember the last time you
> wrote such loop? I cannot.
I can see one: a kind of compiler optimization. I use to though about  
something similar to what Georg exposed (except not strictly with loops),  
which I called “micro-parallelism”. There are many case in an application  
where some short sequence of instructions or groups of instructions does  
not need to be sequenced. Typically I notice this when I do not know which  
order to give these to make the source clear, as many orders would be  
equivalent. Unfortunately, tasking is inefficient here (too much  
overhead). You talked about Occam which I do not know (just the  
principle), but could be fine, yes.

This kind of parallelism requires to be handled at low level (CPU or  
else). This could be either marked explicitly by the author or detected by  
the compiler as Georg suggested, as a kind of compiler optimization.

-- 
Si les chats miaulent et font autant de vocalises bizarres, c’est pas pour  
les chiens.

“I am fluent in ASCII” [Warren 2010]