Re: [PATCH] mm: inactive-clean list

["Martin Schwidefsky" <schwidefsky@googlemail.com>]

On 7/26/06, Peter Zijlstra <a.p.zijlstra@chello.nl> wrote:
> Wouldn't we typically have all free pages > min_free in state U?
> Also wouldn't all R/O mapped pages not also be V, all R/W mapped pages
> and unmapped page-cache pages P like you state in your paper.

Ahh, ok, I misunderstood. You want to keep the state changes for clean
page cache pages, I assumed that you only want to make pages volatile
if the get on the inactive_clean list and leave them stable if they
are on one of the other two lists.

> This patch would just increase the number of V pages with the tail end
> of the guest LRU, which are typically the pages you would want to evict
> (perhaps even add 5th guest state to indicate that these V pages are
> preferable over the others?)

Yes, that would help for architectures that cannot implement the
potential-volatile state.

> But isn't it so that for the gross over-commit scenario you outline the
> host OS will have to swap out S pages eventually?

My point was that you really have to distinguish between host memory
pressure and guest memory pressure.

-- 
blue skies,
  Martin

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>