Re: [RFC PATCH] rework memory hotplug onlining

["Kani, Toshimitsu" <toshi.kani@hpe.com>]

On Wed, 2017-03-15 at 10:13 +0100, Michal Hocko wrote:
 :
> @@ -388,39 +389,44 @@ static ssize_t show_valid_zones(struct device
> *dev,
>  				struct device_attribute *attr, char
> *buf)
>  {
>  	struct memory_block *mem = to_memory_block(dev);
> -	unsigned long start_pfn, end_pfn;
> -	unsigned long valid_start, valid_end, valid_pages;
> -	unsigned long nr_pages = PAGES_PER_SECTION *
> sections_per_block;
> +	unsigned long start_pfn, nr_pages;
> +	bool append = false;
>  	struct zone *zone;
> -	int zone_shift = 0;
> +	int nid;
>  
>  	start_pfn = section_nr_to_pfn(mem->start_section_nr);
> -	end_pfn = start_pfn + nr_pages;
> +	zone = page_zone(pfn_to_page(start_pfn));
> +	nr_pages = PAGES_PER_SECTION * sections_per_block;
>  
> -	/* The block contains more than one zone can not be
> offlined. */
> -	if (!test_pages_in_a_zone(start_pfn, end_pfn, &valid_start,
> &valid_end))
> +	/*
> +	 * The block contains more than one zone can not be
> offlined.
> +	 * This can happen e.g. for ZONE_DMA and ZONE_DMA32
> +	 */
> +	if (!test_pages_in_a_zone(start_pfn, start_pfn + nr_pages,
> NULL, NULL))
>  		return sprintf(buf, "none\n");
>  
> -	zone = page_zone(pfn_to_page(valid_start));

Please do not remove the fix made in a96dfddbcc043. zone needs to be
set from valid_start, not from start_pfn.

Thanks,
-Toshi