[gclist] memory protections and system calls

[Stefan Monnier]

Giuliano Carlini <giuliano@ix.netcom.com> said:
> The kernel believes the pages it wrote aren't dirty. But,
> since the driver didn't pass them on, the write never took
> place. So, the driver had better tell the kernel that the
> page is still dirty. Ideally, the interface between the
> kernel and driver supports some way to do this. If not,
> the VDB logic can touch every page that the flush logic
> marked as dirty.

But then, at the next flush, the kernel will again tell you those pages are
dirty and you won't be able to tell if this is because the VDB purposefully
dirtied them or because the mutator touched them again in the mean time.

I'm sure there is some way to get what you want with such a scheme, but on
another hand, I'm wondering if the efficiency will really be improved compared
to more traditional methods.

	
	Stefan