Re: [PATCH] [RFC] for_each_page_in_zone [1/1]

[Dave Hansen <haveblue@us.ibm.com>]

On Fri, 2006-02-24 at 17:15 +0900, KAMEZAWA Hiroyuki wrote:
> +struct page *next_page_in_zone(struct page *page, struct zone *zone)
> +{
> +       unsigned long pfn = page_to_pfn(page);
> +
> +       if (!populated_zone(zone))
> +               return NULL;
> +
> +       pfn = next_valid_pfn(pfn, zone->zone_start_pfn + zone->spanned_pages);
> +
> +       if (pfn == END_PFN)
> +               return NULL;
> +
> +       return pfn_to_page(pfn);
> +} 

If there can be a case where a node spans other nodes, then I don't
think this patch will work.  The next_valid_pfn() could be a pfn in
another zone.  I believe that you may have to do a pfn_to_page() and
check the zone on each one.  

There are some ppc64 machines which have memory laid out like this:

  0-100 MB Node0
100-200 MB Node1
200-300 MB Node0

Node0's ZONE_DMA has a start_pfn of 0, a spanned_pages of 300MB and a
present_pages of 200MB.  The next_valid_pfn() after the first 100MB is a
page in Node1.

Sorry if I missed this on the first go around.  

-- Dave

--
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>