Re: [PATCH v1 26] mm allow reuse of the lower 16 bit of the page type with an actual type

[David Hildenbrand]

On 29.05.24 17:55, wang wei wrote:
> ---
> > As long as the owner sets a page type first, we can allow reuse of the
> > lower 18 bit: sufficient to store an offset into a 64 KiB page, which
> > is the maximum base page size in *common* configurations (ignoring the
> > 256 KiB variant). Restrict it to the head page.
> >
> > We'll use that for zsmalloc next, to set a proper type while still
> > reusing that field to store information (offset into a base page) that
> > cannot go elsewhere for now.
> >
> > Fear of running out of bits for storing the actual type? Actually, we
> > don't need one bit per type, we could store a single value instead.
> > Further, we could likely limit PAGE_TYPE_BASE to a single (highest) bit.
> >
> > Signed-off-by: David Hildenbrand <david@xxxxxxxxxx>
> > ---
> >   include/linux/mm_types.h   |  5 +++++
> >   include/linux/page-flags.h | 20 ++++++++++++--------
> >   2 files changed, 17 insertions(+), 8 deletions(-)
> >
> > diff --git a/include/linux/mm_types.h b/include/linux/mm_types.h
> > index 6b2aeba792c4..598cfedbbfa0 100644
> > --- a/include/linux/mm_types.h
> > +++ b/include/linux/mm_types.h
> > @@ -157,6 +157,11 @@ struct page {
> >   		 *
> >   		 * See page-flags.h for a list of page types which are currently
> >   		 * stored here.
> > +		 *
> > +		 * Owners of typed folios may reuse the lower 16 bit of the
> > +		 * head page page_type field after setting the page type,
> > +		 * but must reset these 16 bit to -1 before clearing the
> > +		 * page type.
> >   		 */
> >   		unsigned int page_type;
> >   
> > diff --git a/include/linux/page-flags.h b/include/linux/page-flags.h
> > index 104078afe0b1..b43e380ffa0b 100644
> > --- a/include/linux/page-flags.h
> > +++ b/include/linux/page-flags.h
> > @@ -945,14 +945,18 @@ PAGEFLAG_FALSE(HasHWPoisoned, has_hwpoisoned)
> >    */
> >   
> >   #define PAGE_TYPE_BASE	0xf0000000
> > -/* Reserve		0x0000007f to catch underflows of _mapcount */
> > -#define PAGE_MAPCOUNT_RESERVE	-128
> > -#define PG_buddy	0x00000080
> > -#define PG_offline	0x00000100
> > -#define PG_table	0x00000200
> > -#define PG_guard	0x00000400
> > -#define PG_hugetlb	0x00000800
> > -#define PG_slab		0x00001000
> > +/*
> > + * Reserve		0x0000ffff to catch underflows of _mapcount and
> > + * allow owners that set a type to reuse the lower 16 bit for their own
> > + * purposes.
> > + */
> > +#define PAGE_MAPCOUNT_RESERVE	-65536
> > +#define PG_buddy	0x00010000
> > +#define PG_offline	0x00020000
> > +#define PG_table	0x00040000
> > +#define PG_guard	0x00080000
> > +#define PG_hugetlb	0x00100800
>
> Every PG_XX occupies one bit in my understanding.   But why PG_hugetlb occupies two bits?

Because it's wrong (although not harmful). Same issue in v2, fat fingers.

Thanks for pointing that out!

--
Cheers,

David / dhildenb