Re: [patch 1/2] mm: speculative get_page

[Oleg Nesterov <oleg@tv-sign.ru>]

Nick Piggin wrote:
>
> --- linux-2.6.orig/mm/vmscan.c
> +++ linux-2.6/mm/vmscan.c
> @@ -380,6 +380,8 @@ int remove_mapping(struct address_space 
>  	if (!mapping)
>  		return 0;		/* truncate got there first */
>
> +	SetPageNoNewRefs(page);
> +	smp_wmb();
>  	write_lock_irq(&mapping->tree_lock);
>

Is it enough?

PG_nonewrefs could be already set by another add_to_page_cache()/remove_mapping(),
and it will be cleared when we take ->tree_lock. For example:

CPU_0					CPU_1					CPU_3

add_to_page_cache:

    SetPageNoNewRefs();
    write_lock_irq(->tree_lock);
    ...
    write_unlock_irq(->tree_lock);

					remove_mapping:
	
					    SetPageNoNewRefs();

    ClearPageNoNewRefs();
					    write_lock_irq(->tree_lock);

					    check page_count()

										page_cache_get_speculative:

										    increment page_count()

										    no PG_nonewrefs => return

Oleg.