Re: pi_test predicted results

[Esben Nielsen <simlo@phys.au.dk>]

The documentation is wrong... To my defence, I belived that was the result
when I wrote it because I assumed a slightly different behaviour, which
you can only obtain on UP.

Anyways, I think we found that the general bound is

 2^{depth} ms

(I have to find the old mails to check it though...)

It does require many threads though. But N=10 for d=3 giving 8 ms of
maximum latency should be enough. Notice, that the worst case should be
hit very rarely.

Esben


On Wed, 1 Jun 2005, Daniel Walker wrote:

> 
> 	What would you predict the results of the pi test with --tasks 10 
> and --depth 3 ?
> 
> 	From the docmentation it looks like depth N adds N milliseconds to 
> the locking latency . It seems like increasing the numbers of tasks would 
> also increase the locking latency by M tasks == M milliseconds. But I 
> thought I would ask to be sure..
> 
> Daniel
> 
> -
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at  http://www.tux.org/lkml/
> 

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/