On Thu, 2008-05-08 at 16:15 -0700, Dave Hansen wrote:
quoted text
> Here's one quick stab at a solution.  I figured that we already pass
> that 'private' variable around.  This patch just sticks that variable
> *in* the mm_walk and also makes the caller fill in an 'mm' as well.
> Then, we just pass the actual mm_walk around.
> 
> Maybe we should just stick the VMA in the mm_walk as well, and have the
> common code keep it up to date with the addresses currently being
> walked.
> 
> Sadly, I didn't quite get enough time to flesh this idea out very far
> today, and I'll be offline for a couple of days now.  But, if someone
> wants to go this route, I thought this might be useful.  

This much looks reasonable. But the real test of course is to actually
teach it about detecting huge pages efficiently. And I suspect that
means tracking the current VMA all the time in the walk. Am I wrong?

-- 
Mathematics is the supreme nostalgia of our time.

--
unsubscribe notice
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/