On 2012/11/29 7:52, Andrew Morton wrote:
> On Wed, 21 Nov 2012 23:09:46 +0800
> Jiang Liu <liu...@gmail.com> wrote:
>
>> Subject: Re: [RFT PATCH v2 4/5] mm: provide more accurate estimation of
>> pages occupied by memmap
>
> How are people to test this? "does it boot"?
>
I have tested this in x86_64, it does boot.
Node 0, zone DMA
pages free 3972
min 1
low 1
high 1
scanned 0
spanned 4080
present 3979
managed 3972
Node 0, zone DMA32
pages free 448783
min 172
low 215
high 258
scanned 0
spanned 1044480
present 500799
managed 444545
Node 0, zone Normal
pages free 2375547
min 1394
low 1742
high 2091
scanned 0
spanned 3670016
present 3670016
managed 3585105
Thanks,
Jianguo Wu
>> If SPARSEMEM is enabled, it won't build page structures for
>> non-existing pages (holes) within a zone, so provide a more accurate
>> estimation of pages occupied by memmap if there are bigger holes within
>> the zone.
>>
>> And pages for highmem zones' memmap will be allocated from lowmem, so
>> charge nr_kernel_pages for that.
>>
>> ...
>>
>> --- a/mm/page_alloc.c
>> +++ b/mm/page_alloc.c
>> @@ -4442,6 +4442,26 @@ void __init set_pageblock_order(void)
>>
>> #endif /* CONFIG_HUGETLB_PAGE_SIZE_VARIABLE */
>>
>> +static unsigned long calc_memmap_size(unsigned long spanned_pages,
>> + unsigned long present_pages)
>> +{
>> + unsigned long pages = spanned_pages;
>> +
>> + /*
>> + * Provide a more accurate estimation if there are holes within
>> + * the zone and SPARSEMEM is in use. If there are holes within the
>> + * zone, each populated memory region may cost us one or two extra
>> + * memmap pages due to alignment because memmap pages for each
>> + * populated regions may not naturally algined on page boundary.
>> + * So the (present_pages >> 4) heuristic is a tradeoff for that.
>> + */
>> + if (spanned_pages > present_pages + (present_pages >> 4) &&
>> + IS_ENABLED(CONFIG_SPARSEMEM))
>> + pages = present_pages;
>> +
>> + return PAGE_ALIGN(pages * sizeof(struct page)) >> PAGE_SHIFT;
>> +}
>> +
>
> I spose we should do this, although it makes no difference as the
> compiler will inline calc_memmap_size() into its caller:
>
> ---
> a/mm/page_alloc.c~mm-provide-more-accurate-estimation-of-pages-occupied-by-memmap-fix
> +++ a/mm/page_alloc.c
> @@ -4526,8 +4526,8 @@ void __init set_pageblock_order(void)
>
> #endif /* CONFIG_HUGETLB_PAGE_SIZE_VARIABLE */
>
> -static unsigned long calc_memmap_size(unsigned long spanned_pages,
> - unsigned long present_pages)
> +static unsigned long __paginginit calc_memmap_size(unsigned long
> spanned_pages,
> + unsigned long present_pages)
> {
> unsigned long pages = spanned_pages;
>
>
>
> .
>
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majord...@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
