Re: [patch 1/4] mm: page refcount use atomic primitives

[Andrew Morton]

Nick Piggin <nickpiggin@xxxxxxxxxxxx> wrote:
>
> > That's from a two-minute-peek.  I haven't thought about this dreadfully
>  > hard.  But I'd like to gain some confidence that you have, please.  This
>  > stuff is tricky.
>  > 
> 
>  Right, no blam. This is how I avoid taking the LRU lock.
> 
>    "Instead, use atomic_inc_not_zero to ensure we never **pick up a 0 refcount**
>     page from the LRU (ie. we guarantee the page will not be touched)."
> 
>  I don't understand what you're asking?

Well if you don't understand me, how do you expect me to?

Ho hum.  Please redo the patches into something which vaguely applies and
let's give them a spin.

I would prefer that the BUG_ONs be split into a separate patch tho.  Or at
least, let's take a real close look at whether they're really needed.
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/