On Tue, Jan 04, 2022 at 01:10:18AM +0100, Vlastimil Babka wrote:
> From: "Matthew Wilcox (Oracle)" <willy@xxxxxxxxxxxxx>
>
> Make struct slab independent of struct page. It still uses the
> underlying memory in struct page for storing slab-specific data, but
> slab and slub can now be weaned off using struct page directly. Some of
> the wrapper functions (slab_address() and slab_order()) still need to
> cast to struct folio, but this is a significant disentanglement.
>
> [ vbabka@xxxxxxx: Rebase on folios, use folio instead of page where
> possible.
>
> Do not duplicate flags field in struct slab, instead make the related
> accessors go through slab_folio(). For testing pfmemalloc use the
> folio_*_active flag accessors directly so the PageSlabPfmemalloc
> wrappers can be removed later.
>
> Make folio_slab() expect only folio_test_slab() == true folios and
> virt_to_slab() return NULL when folio_test_slab() == false.
>
> Move struct slab to mm/slab.h.
>
> Don't represent with struct slab pages that are not true slab pages,
> but just a compound page obtained directly rom page allocator (with
a typo here: (f)rom
> large kmalloc() for SLUB and SLOB). ]
>
> Signed-off-by: Matthew Wilcox (Oracle) <willy@xxxxxxxxxxxxx>
> Signed-off-by: Vlastimil Babka <vbabka@xxxxxxx>
> Acked-by: Johannes Weiner <hannes@xxxxxxxxxxx>
> Reviewed-by: Roman Gushchin <guro@xxxxxx>
> ---
> include/linux/mm_types.h | 10 +--
> mm/slab.h | 167 +++++++++++++++++++++++++++++++++++++++
> mm/slub.c | 8 +-
> 3 files changed, 176 insertions(+), 9 deletions(-)
>
> diff --git a/include/linux/mm_types.h b/include/linux/mm_types.h
> index c3a6e6209600..1ae3537c7920 100644
> --- a/include/linux/mm_types.h
> +++ b/include/linux/mm_types.h
> @@ -56,11 +56,11 @@ struct mem_cgroup;
> * in each subpage, but you may need to restore some of their values
> * afterwards.
> *
> - * SLUB uses cmpxchg_double() to atomically update its freelist and
> - * counters. That requires that freelist & counters be adjacent and
> - * double-word aligned. We align all struct pages to double-word
> - * boundaries, and ensure that 'freelist' is aligned within the
> - * struct.
> + * SLUB uses cmpxchg_double() to atomically update its freelist and counters.
> + * That requires that freelist & counters in struct slab be adjacent and
> + * double-word aligned. Because struct slab currently just reinterprets the
> + * bits of struct page, we align all struct pages to double-word boundaries,
> + * and ensure that 'freelist' is aligned within struct slab.
> */
> #ifdef CONFIG_HAVE_ALIGNED_STRUCT_PAGE
> #define _struct_page_alignment __aligned(2 * sizeof(unsigned long))
> diff --git a/mm/slab.h b/mm/slab.h
> index 56ad7eea3ddf..0e67a8cb7f80 100644
> --- a/mm/slab.h
> +++ b/mm/slab.h
> @@ -5,6 +5,173 @@
> * Internal slab definitions
> */
>
> +/* Reuses the bits in struct page */
> +struct slab {
> + unsigned long __page_flags;
> + union {
> + struct list_head slab_list;
> + struct { /* Partial pages */
> + struct slab *next;
> +#ifdef CONFIG_64BIT
> + int slabs; /* Nr of slabs left */
> +#else
> + short int slabs;
> +#endif
> + };
> + struct rcu_head rcu_head;
> + };
> + struct kmem_cache *slab_cache; /* not slob */
> + /* Double-word boundary */
> + void *freelist; /* first free object */
> + union {
> + void *s_mem; /* slab: first object */
> + unsigned long counters; /* SLUB */
> + struct { /* SLUB */
> + unsigned inuse:16;
> + unsigned objects:15;
> + unsigned frozen:1;
> + };
> + };
> +
> + union {
> + unsigned int active; /* SLAB */
> + int units; /* SLOB */
> + };
> + atomic_t __page_refcount;
> +#ifdef CONFIG_MEMCG
> + unsigned long memcg_data;
> +#endif
> +};
> +
> +#define SLAB_MATCH(pg, sl) \
> + static_assert(offsetof(struct page, pg) == offsetof(struct slab, sl))
> +SLAB_MATCH(flags, __page_flags);
> +SLAB_MATCH(compound_head, slab_list); /* Ensure bit 0 is clear */
> +SLAB_MATCH(slab_list, slab_list);
> +SLAB_MATCH(rcu_head, rcu_head);
> +SLAB_MATCH(slab_cache, slab_cache);
> +SLAB_MATCH(s_mem, s_mem);
> +SLAB_MATCH(active, active);
> +SLAB_MATCH(_refcount, __page_refcount);
> +#ifdef CONFIG_MEMCG
> +SLAB_MATCH(memcg_data, memcg_data);
> +#endif
> +#undef SLAB_MATCH
> +static_assert(sizeof(struct slab) <= sizeof(struct page));
> +
> +/**
> + * folio_slab - Converts from folio to slab.
> + * @folio: The folio.
> + *
> + * Currently struct slab is a different representation of a folio where
> + * folio_test_slab() is true.
> + *
> + * Return: The slab which contains this folio.
> + */
> +#define folio_slab(folio) (_Generic((folio), \
> + const struct folio *: (const struct slab *)(folio), \
> + struct folio *: (struct slab *)(folio)))
> +
> +/**
> + * slab_folio - The folio allocated for a slab
> + * @slab: The slab.
> + *
> + * Slabs are allocated as folios that contain the individual objects and are
> + * using some fields in the first struct page of the folio - those fields are
> + * now accessed by struct slab. It is occasionally necessary to convert back to
> + * a folio in order to communicate with the rest of the mm. Please use this
> + * helper function instead of casting yourself, as the implementation may change
> + * in the future.
> + */
> +#define slab_folio(s) (_Generic((s), \
> + const struct slab *: (const struct folio *)s, \
> + struct slab *: (struct folio *)s))
> +
> +/**
> + * page_slab - Converts from first struct page to slab.
> + * @p: The first (either head of compound or single) page of slab.
> + *
> + * A temporary wrapper to convert struct page to struct slab in situations where
> + * we know the page is the compound head, or single order-0 page.
> + *
> + * Long-term ideally everything would work with struct slab directly or go
> + * through folio to struct slab.
> + *
> + * Return: The slab which contains this page
> + */
> +#define page_slab(p) (_Generic((p), \
> + const struct page *: (const struct slab *)(p), \
> + struct page *: (struct slab *)(p)))
> +
> +/**
> + * slab_page - The first struct page allocated for a slab
> + * @slab: The slab.
> + *
> + * A convenience wrapper for converting slab to the first struct page of the
> + * underlying folio, to communicate with code not yet converted to folio or
> + * struct slab.
> + */
> +#define slab_page(s) folio_page(slab_folio(s), 0)
> +
> +/*
> + * If network-based swap is enabled, sl*b must keep track of whether pages
> + * were allocated from pfmemalloc reserves.
> + */
> +static inline bool slab_test_pfmemalloc(const struct slab *slab)
> +{
> + return folio_test_active((struct folio *)slab_folio(slab));
> +}
> +
> +static inline void slab_set_pfmemalloc(struct slab *slab)
> +{
> + folio_set_active(slab_folio(slab));
> +}
> +
> +static inline void slab_clear_pfmemalloc(struct slab *slab)
> +{
> + folio_clear_active(slab_folio(slab));
> +}
> +
> +static inline void __slab_clear_pfmemalloc(struct slab *slab)
> +{
> + __folio_clear_active(slab_folio(slab));
> +}
> +
> +static inline void *slab_address(const struct slab *slab)
> +{
> + return folio_address(slab_folio(slab));
> +}
> +
> +static inline int slab_nid(const struct slab *slab)
> +{
> + return folio_nid(slab_folio(slab));
> +}
> +
> +static inline pg_data_t *slab_pgdat(const struct slab *slab)
> +{
> + return folio_pgdat(slab_folio(slab));
> +}
> +
> +static inline struct slab *virt_to_slab(const void *addr)
> +{
> + struct folio *folio = virt_to_folio(addr);
> +
> + if (!folio_test_slab(folio))
> + return NULL;
> +
> + return folio_slab(folio);
> +}
> +
> +static inline int slab_order(const struct slab *slab)
> +{
> + return folio_order((struct folio *)slab_folio(slab));
> +}
> +
> +static inline size_t slab_size(const struct slab *slab)
> +{
> + return PAGE_SIZE << slab_order(slab);
> +}
> +
> #ifdef CONFIG_SLOB
> /*
> * Common fields provided in kmem_cache by all slab allocators
> diff --git a/mm/slub.c b/mm/slub.c
> index 2ccb1c71fc36..a211d96011ba 100644
> --- a/mm/slub.c
> +++ b/mm/slub.c
> @@ -3787,7 +3787,7 @@ static unsigned int slub_min_objects;
> * requested a higher minimum order then we start with that one instead of
> * the smallest order which will fit the object.
> */
> -static inline unsigned int slab_order(unsigned int size,
> +static inline unsigned int calc_slab_order(unsigned int size,
> unsigned int min_objects, unsigned int max_order,
> unsigned int fract_leftover)
> {
> @@ -3851,7 +3851,7 @@ static inline int calculate_order(unsigned int size)
>
> fraction = 16;
> while (fraction >= 4) {
> - order = slab_order(size, min_objects,
> + order = calc_slab_order(size, min_objects,
> slub_max_order, fraction);
> if (order <= slub_max_order)
> return order;
> @@ -3864,14 +3864,14 @@ static inline int calculate_order(unsigned int size)
> * We were unable to place multiple objects in a slab. Now
> * lets see if we can place a single object there.
> */
> - order = slab_order(size, 1, slub_max_order, 1);
> + order = calc_slab_order(size, 1, slub_max_order, 1);
> if (order <= slub_max_order)
> return order;
>
> /*
> * Doh this slab cannot be placed using slub_max_order.
> */
> - order = slab_order(size, 1, MAX_ORDER, 1);
> + order = calc_slab_order(size, 1, MAX_ORDER, 1);
> if (order < MAX_ORDER)
> return order;
> return -ENOSYS;
This patch looks good.
Reviewed-by: Hyeonggon Yoo <42.hyeyoo@xxxxxxxxx>
SL[AUO]B works fine on the top of this patch.
Tested-by: Hyeonggon Yoo <42.hyeyoo@xxxxxxxxx>
> --
> 2.34.1
>
