[thirdparty/kernel/linux.git] / include / linux / gfp.h

/* SPDX-License-Identifier: GPL-2.0 */
#ifndef __LINUX_GFP_H
#define __LINUX_GFP_H

#include <linux/gfp_types.h>

#include <linux/mmzone.h>
#include <linux/topology.h>
#include <linux/alloc_tag.h>
#include <linux/sched.h>

struct vm_area_struct;
struct mempolicy;

/* Convert GFP flags to their corresponding migrate type */
#define GFP_MOVABLE_MASK (__GFP_RECLAIMABLE|__GFP_MOVABLE)
#define GFP_MOVABLE_SHIFT 3

static inline int gfp_migratetype(const gfp_t gfp_flags)
{
	VM_WARN_ON((gfp_flags & GFP_MOVABLE_MASK) == GFP_MOVABLE_MASK);
	BUILD_BUG_ON((1UL << GFP_MOVABLE_SHIFT) != ___GFP_MOVABLE);
	BUILD_BUG_ON((___GFP_MOVABLE >> GFP_MOVABLE_SHIFT) != MIGRATE_MOVABLE);
	BUILD_BUG_ON((___GFP_RECLAIMABLE >> GFP_MOVABLE_SHIFT) != MIGRATE_RECLAIMABLE);
	BUILD_BUG_ON(((___GFP_MOVABLE | ___GFP_RECLAIMABLE) >>
		      GFP_MOVABLE_SHIFT) != MIGRATE_HIGHATOMIC);

	if (unlikely(page_group_by_mobility_disabled))
		return MIGRATE_UNMOVABLE;

	/* Group based on mobility */
	return (__force unsigned long)(gfp_flags & GFP_MOVABLE_MASK) >> GFP_MOVABLE_SHIFT;
}
#undef GFP_MOVABLE_MASK
#undef GFP_MOVABLE_SHIFT

static inline bool gfpflags_allow_blocking(const gfp_t gfp_flags)
{
	return !!(gfp_flags & __GFP_DIRECT_RECLAIM);
}

#ifdef CONFIG_HIGHMEM
#define OPT_ZONE_HIGHMEM ZONE_HIGHMEM
#else
#define OPT_ZONE_HIGHMEM ZONE_NORMAL
#endif

#ifdef CONFIG_ZONE_DMA
#define OPT_ZONE_DMA ZONE_DMA
#else
#define OPT_ZONE_DMA ZONE_NORMAL
#endif

#ifdef CONFIG_ZONE_DMA32
#define OPT_ZONE_DMA32 ZONE_DMA32
#else
#define OPT_ZONE_DMA32 ZONE_NORMAL
#endif

/*
 * GFP_ZONE_TABLE is a word size bitstring that is used for looking up the
 * zone to use given the lowest 4 bits of gfp_t. Entries are GFP_ZONES_SHIFT
 * bits long and there are 16 of them to cover all possible combinations of
 * __GFP_DMA, __GFP_DMA32, __GFP_MOVABLE and __GFP_HIGHMEM.
 *
 * The zone fallback order is MOVABLE=>HIGHMEM=>NORMAL=>DMA32=>DMA.
 * But GFP_MOVABLE is not only a zone specifier but also an allocation
 * policy. Therefore __GFP_MOVABLE plus another zone selector is valid.
 * Only 1 bit of the lowest 3 bits (DMA,DMA32,HIGHMEM) can be set to "1".
 *
 *       bit       result
 *       =================
 *       0x0    => NORMAL
 *       0x1    => DMA or NORMAL
 *       0x2    => HIGHMEM or NORMAL
 *       0x3    => BAD (DMA+HIGHMEM)
 *       0x4    => DMA32 or NORMAL
 *       0x5    => BAD (DMA+DMA32)
 *       0x6    => BAD (HIGHMEM+DMA32)
 *       0x7    => BAD (HIGHMEM+DMA32+DMA)
 *       0x8    => NORMAL (MOVABLE+0)
 *       0x9    => DMA or NORMAL (MOVABLE+DMA)
 *       0xa    => MOVABLE (Movable is valid only if HIGHMEM is set too)
 *       0xb    => BAD (MOVABLE+HIGHMEM+DMA)
 *       0xc    => DMA32 or NORMAL (MOVABLE+DMA32)
 *       0xd    => BAD (MOVABLE+DMA32+DMA)
 *       0xe    => BAD (MOVABLE+DMA32+HIGHMEM)
 *       0xf    => BAD (MOVABLE+DMA32+HIGHMEM+DMA)
 *
 * GFP_ZONES_SHIFT must be <= 2 on 32 bit platforms.
 */

#if defined(CONFIG_ZONE_DEVICE) && (MAX_NR_ZONES-1) <= 4
/* ZONE_DEVICE is not a valid GFP zone specifier */
#define GFP_ZONES_SHIFT 2
#else
#define GFP_ZONES_SHIFT ZONES_SHIFT
#endif

#if 16 * GFP_ZONES_SHIFT > BITS_PER_LONG
#error GFP_ZONES_SHIFT too large to create GFP_ZONE_TABLE integer
#endif

#define GFP_ZONE_TABLE ( \
	(ZONE_NORMAL << 0 * GFP_ZONES_SHIFT)				       \
	| (OPT_ZONE_DMA << ___GFP_DMA * GFP_ZONES_SHIFT)		       \
	| (OPT_ZONE_HIGHMEM << ___GFP_HIGHMEM * GFP_ZONES_SHIFT)	       \
	| (OPT_ZONE_DMA32 << ___GFP_DMA32 * GFP_ZONES_SHIFT)		       \
	| (ZONE_NORMAL << ___GFP_MOVABLE * GFP_ZONES_SHIFT)		       \
	| (OPT_ZONE_DMA << (___GFP_MOVABLE | ___GFP_DMA) * GFP_ZONES_SHIFT)    \
	| (ZONE_MOVABLE << (___GFP_MOVABLE | ___GFP_HIGHMEM) * GFP_ZONES_SHIFT)\
	| (OPT_ZONE_DMA32 << (___GFP_MOVABLE | ___GFP_DMA32) * GFP_ZONES_SHIFT)\
)

/*
 * GFP_ZONE_BAD is a bitmap for all combinations of __GFP_DMA, __GFP_DMA32
 * __GFP_HIGHMEM and __GFP_MOVABLE that are not permitted. One flag per
 * entry starting with bit 0. Bit is set if the combination is not
 * allowed.
 */
#define GFP_ZONE_BAD ( \
	1 << (___GFP_DMA | ___GFP_HIGHMEM)				      \
	| 1 << (___GFP_DMA | ___GFP_DMA32)				      \
	| 1 << (___GFP_DMA32 | ___GFP_HIGHMEM)				      \
	| 1 << (___GFP_DMA | ___GFP_DMA32 | ___GFP_HIGHMEM)		      \
	| 1 << (___GFP_MOVABLE | ___GFP_HIGHMEM | ___GFP_DMA)		      \
	| 1 << (___GFP_MOVABLE | ___GFP_DMA32 | ___GFP_DMA)		      \
	| 1 << (___GFP_MOVABLE | ___GFP_DMA32 | ___GFP_HIGHMEM)		      \
	| 1 << (___GFP_MOVABLE | ___GFP_DMA32 | ___GFP_DMA | ___GFP_HIGHMEM)  \
)

static inline enum zone_type gfp_zone(gfp_t flags)
{
	enum zone_type z;
	int bit = (__force int) (flags & GFP_ZONEMASK);

	z = (GFP_ZONE_TABLE >> (bit * GFP_ZONES_SHIFT)) &
					 ((1 << GFP_ZONES_SHIFT) - 1);
	VM_BUG_ON((GFP_ZONE_BAD >> bit) & 1);
	return z;
}

/*
 * There is only one page-allocator function, and two main namespaces to
 * it. The alloc_page*() variants return 'struct page *' and as such
 * can allocate highmem pages, the *get*page*() variants return
 * virtual kernel addresses to the allocated page(s).
 */

static inline int gfp_zonelist(gfp_t flags)
{
#ifdef CONFIG_NUMA
	if (unlikely(flags & __GFP_THISNODE))
		return ZONELIST_NOFALLBACK;
#endif
	return ZONELIST_FALLBACK;
}

/*
 * gfp flag masking for nested internal allocations.
 *
 * For code that needs to do allocations inside the public allocation API (e.g.
 * memory allocation tracking code) the allocations need to obey the caller
 * allocation context constrains to prevent allocation context mismatches (e.g.
 * GFP_KERNEL allocations in GFP_NOFS contexts) from potential deadlock
 * situations.
 *
 * It is also assumed that these nested allocations are for internal kernel
 * object storage purposes only and are not going to be used for DMA, etc. Hence
 * we strip out all the zone information and leave just the context information
 * intact.
 *
 * Further, internal allocations must fail before the higher level allocation
 * can fail, so we must make them fail faster and fail silently. We also don't
 * want them to deplete emergency reserves.  Hence nested allocations must be
 * prepared for these allocations to fail.
 */
static inline gfp_t gfp_nested_mask(gfp_t flags)
{
	return ((flags & (GFP_KERNEL | GFP_ATOMIC | __GFP_NOLOCKDEP)) |
		(__GFP_NORETRY | __GFP_NOMEMALLOC | __GFP_NOWARN));
}

/*
 * We get the zone list from the current node and the gfp_mask.
 * This zone list contains a maximum of MAX_NUMNODES*MAX_NR_ZONES zones.
 * There are two zonelists per node, one for all zones with memory and
 * one containing just zones from the node the zonelist belongs to.
 *
 * For the case of non-NUMA systems the NODE_DATA() gets optimized to
 * &contig_page_data at compile-time.
 */
static inline struct zonelist *node_zonelist(int nid, gfp_t flags)
{
	return NODE_DATA(nid)->node_zonelists + gfp_zonelist(flags);
}

#ifndef HAVE_ARCH_FREE_PAGE
static inline void arch_free_page(struct page *page, int order) { }
#endif
#ifndef HAVE_ARCH_ALLOC_PAGE
static inline void arch_alloc_page(struct page *page, int order) { }
#endif

struct page *__alloc_pages_noprof(gfp_t gfp, unsigned int order, int preferred_nid,
		nodemask_t *nodemask);
#define __alloc_pages(...)			alloc_hooks(__alloc_pages_noprof(__VA_ARGS__))

struct folio *__folio_alloc_noprof(gfp_t gfp, unsigned int order, int preferred_nid,
		nodemask_t *nodemask);
#define __folio_alloc(...)			alloc_hooks(__folio_alloc_noprof(__VA_ARGS__))

unsigned long alloc_pages_bulk_noprof(gfp_t gfp, int preferred_nid,
				nodemask_t *nodemask, int nr_pages,
				struct list_head *page_list,
				struct page **page_array);
#define __alloc_pages_bulk(...)			alloc_hooks(alloc_pages_bulk_noprof(__VA_ARGS__))

unsigned long alloc_pages_bulk_array_mempolicy_noprof(gfp_t gfp,
				unsigned long nr_pages,
				struct page **page_array);
#define  alloc_pages_bulk_array_mempolicy(...)				\
	alloc_hooks(alloc_pages_bulk_array_mempolicy_noprof(__VA_ARGS__))

/* Bulk allocate order-0 pages */
#define alloc_pages_bulk_list(_gfp, _nr_pages, _list)			\
	__alloc_pages_bulk(_gfp, numa_mem_id(), NULL, _nr_pages, _list, NULL)

#define alloc_pages_bulk_array(_gfp, _nr_pages, _page_array)		\
	__alloc_pages_bulk(_gfp, numa_mem_id(), NULL, _nr_pages, NULL, _page_array)

static inline unsigned long
alloc_pages_bulk_array_node_noprof(gfp_t gfp, int nid, unsigned long nr_pages,
				   struct page **page_array)
{
	if (nid == NUMA_NO_NODE)
		nid = numa_mem_id();

	return alloc_pages_bulk_noprof(gfp, nid, NULL, nr_pages, NULL, page_array);
}

#define alloc_pages_bulk_array_node(...)				\
	alloc_hooks(alloc_pages_bulk_array_node_noprof(__VA_ARGS__))

static inline void warn_if_node_offline(int this_node, gfp_t gfp_mask)
{
	gfp_t warn_gfp = gfp_mask & (__GFP_THISNODE|__GFP_NOWARN);

	if (warn_gfp != (__GFP_THISNODE|__GFP_NOWARN))
		return;

	if (node_online(this_node))
		return;

	pr_warn("%pGg allocation from offline node %d\n", &gfp_mask, this_node);
	dump_stack();
}

/*
 * Allocate pages, preferring the node given as nid. The node must be valid and
 * online. For more general interface, see alloc_pages_node().
 */
static inline struct page *
__alloc_pages_node_noprof(int nid, gfp_t gfp_mask, unsigned int order)
{
	VM_BUG_ON(nid < 0 || nid >= MAX_NUMNODES);
	warn_if_node_offline(nid, gfp_mask);

	return __alloc_pages_noprof(gfp_mask, order, nid, NULL);
}

#define  __alloc_pages_node(...)		alloc_hooks(__alloc_pages_node_noprof(__VA_ARGS__))

static inline
struct folio *__folio_alloc_node_noprof(gfp_t gfp, unsigned int order, int nid)
{
	VM_BUG_ON(nid < 0 || nid >= MAX_NUMNODES);
	warn_if_node_offline(nid, gfp);

	return __folio_alloc_noprof(gfp, order, nid, NULL);
}

#define  __folio_alloc_node(...)		alloc_hooks(__folio_alloc_node_noprof(__VA_ARGS__))

/*
 * Allocate pages, preferring the node given as nid. When nid == NUMA_NO_NODE,
 * prefer the current CPU's closest node. Otherwise node must be valid and
 * online.
 */
static inline struct page *alloc_pages_node_noprof(int nid, gfp_t gfp_mask,
						   unsigned int order)
{
	if (nid == NUMA_NO_NODE)
		nid = numa_mem_id();

	return __alloc_pages_node_noprof(nid, gfp_mask, order);
}

#define  alloc_pages_node(...)			alloc_hooks(alloc_pages_node_noprof(__VA_ARGS__))

#ifdef CONFIG_NUMA
struct page *alloc_pages_noprof(gfp_t gfp, unsigned int order);
struct page *alloc_pages_mpol_noprof(gfp_t gfp, unsigned int order,
		struct mempolicy *mpol, pgoff_t ilx, int nid);
struct folio *folio_alloc_noprof(gfp_t gfp, unsigned int order);
struct folio *vma_alloc_folio_noprof(gfp_t gfp, int order, struct vm_area_struct *vma,
		unsigned long addr, bool hugepage);
#else
static inline struct page *alloc_pages_noprof(gfp_t gfp_mask, unsigned int order)
{
	return alloc_pages_node_noprof(numa_node_id(), gfp_mask, order);
}
static inline struct page *alloc_pages_mpol_noprof(gfp_t gfp, unsigned int order,
		struct mempolicy *mpol, pgoff_t ilx, int nid)
{
	return alloc_pages_noprof(gfp, order);
}
static inline struct folio *folio_alloc_noprof(gfp_t gfp, unsigned int order)
{
	return __folio_alloc_node(gfp, order, numa_node_id());
}
#define vma_alloc_folio_noprof(gfp, order, vma, addr, hugepage)		\
	folio_alloc_noprof(gfp, order)
#endif

#define alloc_pages(...)			alloc_hooks(alloc_pages_noprof(__VA_ARGS__))
#define alloc_pages_mpol(...)			alloc_hooks(alloc_pages_mpol_noprof(__VA_ARGS__))
#define folio_alloc(...)			alloc_hooks(folio_alloc_noprof(__VA_ARGS__))
#define vma_alloc_folio(...)			alloc_hooks(vma_alloc_folio_noprof(__VA_ARGS__))

#define alloc_page(gfp_mask) alloc_pages(gfp_mask, 0)

static inline struct page *alloc_page_vma_noprof(gfp_t gfp,
		struct vm_area_struct *vma, unsigned long addr)
{
	struct folio *folio = vma_alloc_folio_noprof(gfp, 0, vma, addr, false);

	return &folio->page;
}
#define alloc_page_vma(...)			alloc_hooks(alloc_page_vma_noprof(__VA_ARGS__))

extern unsigned long get_free_pages_noprof(gfp_t gfp_mask, unsigned int order);
#define __get_free_pages(...)			alloc_hooks(get_free_pages_noprof(__VA_ARGS__))

extern unsigned long get_zeroed_page_noprof(gfp_t gfp_mask);
#define get_zeroed_page(...)			alloc_hooks(get_zeroed_page_noprof(__VA_ARGS__))

void *alloc_pages_exact_noprof(size_t size, gfp_t gfp_mask) __alloc_size(1);
#define alloc_pages_exact(...)			alloc_hooks(alloc_pages_exact_noprof(__VA_ARGS__))

void free_pages_exact(void *virt, size_t size);

__meminit void *alloc_pages_exact_nid_noprof(int nid, size_t size, gfp_t gfp_mask) __alloc_size(2);
#define alloc_pages_exact_nid(...)					\
	alloc_hooks(alloc_pages_exact_nid_noprof(__VA_ARGS__))

#define __get_free_page(gfp_mask)					\
	__get_free_pages((gfp_mask), 0)

#define __get_dma_pages(gfp_mask, order)				\
	__get_free_pages((gfp_mask) | GFP_DMA, (order))

extern void __free_pages(struct page *page, unsigned int order);
extern void free_pages(unsigned long addr, unsigned int order);

struct page_frag_cache;
void page_frag_cache_drain(struct page_frag_cache *nc);
extern void __page_frag_cache_drain(struct page *page, unsigned int count);
void *__page_frag_alloc_align(struct page_frag_cache *nc, unsigned int fragsz,
			      gfp_t gfp_mask, unsigned int align_mask);

static inline void *page_frag_alloc_align(struct page_frag_cache *nc,
					  unsigned int fragsz, gfp_t gfp_mask,
					  unsigned int align)
{
	WARN_ON_ONCE(!is_power_of_2(align));
	return __page_frag_alloc_align(nc, fragsz, gfp_mask, -align);
}

static inline void *page_frag_alloc(struct page_frag_cache *nc,
			     unsigned int fragsz, gfp_t gfp_mask)
{
	return __page_frag_alloc_align(nc, fragsz, gfp_mask, ~0u);
}

extern void page_frag_free(void *addr);

#define __free_page(page) __free_pages((page), 0)
#define free_page(addr) free_pages((addr), 0)

void page_alloc_init_cpuhp(void);
int decay_pcp_high(struct zone *zone, struct per_cpu_pages *pcp);
void drain_zone_pages(struct zone *zone, struct per_cpu_pages *pcp);
void drain_all_pages(struct zone *zone);
void drain_local_pages(struct zone *zone);

void page_alloc_init_late(void);
void setup_pcp_cacheinfo(unsigned int cpu);

/*
 * gfp_allowed_mask is set to GFP_BOOT_MASK during early boot to restrict what
 * GFP flags are used before interrupts are enabled. Once interrupts are
 * enabled, it is set to __GFP_BITS_MASK while the system is running. During
 * hibernation, it is used by PM to avoid I/O during memory allocation while
 * devices are suspended.
 */
extern gfp_t gfp_allowed_mask;

/* Returns true if the gfp_mask allows use of ALLOC_NO_WATERMARK */
bool gfp_pfmemalloc_allowed(gfp_t gfp_mask);

static inline bool gfp_has_io_fs(gfp_t gfp)
{
	return (gfp & (__GFP_IO | __GFP_FS)) == (__GFP_IO | __GFP_FS);
}

/*
 * Check if the gfp flags allow compaction - GFP_NOIO is a really
 * tricky context because the migration might require IO.
 */
static inline bool gfp_compaction_allowed(gfp_t gfp_mask)
{
	return IS_ENABLED(CONFIG_COMPACTION) && (gfp_mask & __GFP_IO);
}

extern gfp_t vma_thp_gfp_mask(struct vm_area_struct *vma);

#ifdef CONFIG_CONTIG_ALLOC
/* The below functions must be run on a range from a single zone. */
extern int alloc_contig_range_noprof(unsigned long start, unsigned long end,
			      unsigned migratetype, gfp_t gfp_mask);
#define alloc_contig_range(...)			alloc_hooks(alloc_contig_range_noprof(__VA_ARGS__))

extern struct page *alloc_contig_pages_noprof(unsigned long nr_pages, gfp_t gfp_mask,
					      int nid, nodemask_t *nodemask);
#define alloc_contig_pages(...)			alloc_hooks(alloc_contig_pages_noprof(__VA_ARGS__))

#endif
void free_contig_range(unsigned long pfn, unsigned long nr_pages);

#endif /* __LINUX_GFP_H */
Commit	Line	Data
b2441318	1	/* SPDX-License-Identifier: GPL-2.0 */
1da177e4 LT	2	#ifndef __LINUX_GFP_H
	3	#define __LINUX_GFP_H
	4
cb5a065b IM	5	#include <linux/gfp_types.h>
cb5a065b IM	6
1da177e4	7	#include <linux/mmzone.h>
082edb7b	8	#include <linux/topology.h>
b951aaff SB	9	#include <linux/alloc_tag.h>
b951aaff SB	10	#include <linux/sched.h>
1da177e4 LT	11
1da177e4 LT	12	struct vm_area_struct;
ddc1a5cb	13	struct mempolicy;
1da177e4	14
dd56b046	15	/* Convert GFP flags to their corresponding migrate type */
e12ba74d	16	#define GFP_MOVABLE_MASK (__GFP_RECLAIMABLE\|__GFP_MOVABLE)
016c13da	17	#define GFP_MOVABLE_SHIFT 3
6cb06229	18
01c0bfe0	19	static inline int gfp_migratetype(const gfp_t gfp_flags)
467c996c	20	{
016c13da MG	21	VM_WARN_ON((gfp_flags & GFP_MOVABLE_MASK) == GFP_MOVABLE_MASK);
	22	BUILD_BUG_ON((1UL << GFP_MOVABLE_SHIFT) != ___GFP_MOVABLE);
	23	BUILD_BUG_ON((___GFP_MOVABLE >> GFP_MOVABLE_SHIFT) != MIGRATE_MOVABLE);
4d86d4f7 PC	24	BUILD_BUG_ON((___GFP_RECLAIMABLE >> GFP_MOVABLE_SHIFT) != MIGRATE_RECLAIMABLE);
	25	BUILD_BUG_ON(((___GFP_MOVABLE \| ___GFP_RECLAIMABLE) >>
	26	GFP_MOVABLE_SHIFT) != MIGRATE_HIGHATOMIC);
467c996c MG	27
	28	if (unlikely(page_group_by_mobility_disabled))
	29	return MIGRATE_UNMOVABLE;
	30
	31	/* Group based on mobility */
fe573327	32	return (__force unsigned long)(gfp_flags & GFP_MOVABLE_MASK) >> GFP_MOVABLE_SHIFT;
467c996c	33	}
dd56b046 MG	34	#undef GFP_MOVABLE_MASK
dd56b046 MG	35	#undef GFP_MOVABLE_SHIFT
a2f1b424	36
d0164adc MG	37	static inline bool gfpflags_allow_blocking(const gfp_t gfp_flags)
d0164adc MG	38	{
543dfb2d	39	return !!(gfp_flags & __GFP_DIRECT_RECLAIM);
d0164adc MG	40	}
d0164adc MG	41
b70d94ee CL	42	#ifdef CONFIG_HIGHMEM
	43	#define OPT_ZONE_HIGHMEM ZONE_HIGHMEM
	44	#else
	45	#define OPT_ZONE_HIGHMEM ZONE_NORMAL
	46	#endif
	47
4b51d669	48	#ifdef CONFIG_ZONE_DMA
b70d94ee CL	49	#define OPT_ZONE_DMA ZONE_DMA
	50	#else
	51	#define OPT_ZONE_DMA ZONE_NORMAL
4b51d669	52	#endif
b70d94ee	53
4e4785bc	54	#ifdef CONFIG_ZONE_DMA32
b70d94ee CL	55	#define OPT_ZONE_DMA32 ZONE_DMA32
	56	#else
	57	#define OPT_ZONE_DMA32 ZONE_NORMAL
4e4785bc	58	#endif
b70d94ee CL	59
	60	/*
	61	* GFP_ZONE_TABLE is a word size bitstring that is used for looking up the
ac2e8e40 HL	62	* zone to use given the lowest 4 bits of gfp_t. Entries are GFP_ZONES_SHIFT
ac2e8e40 HL	63	* bits long and there are 16 of them to cover all possible combinations of
263ff5d8	64	* __GFP_DMA, __GFP_DMA32, __GFP_MOVABLE and __GFP_HIGHMEM.
b70d94ee CL	65	*
	66	* The zone fallback order is MOVABLE=>HIGHMEM=>NORMAL=>DMA32=>DMA.
	67	* But GFP_MOVABLE is not only a zone specifier but also an allocation
	68	* policy. Therefore __GFP_MOVABLE plus another zone selector is valid.
263ff5d8	69	* Only 1 bit of the lowest 3 bits (DMA,DMA32,HIGHMEM) can be set to "1".
b70d94ee CL	70	*
	71	* bit result
	72	* =================
	73	* 0x0 => NORMAL
	74	* 0x1 => DMA or NORMAL
	75	* 0x2 => HIGHMEM or NORMAL
	76	* 0x3 => BAD (DMA+HIGHMEM)
4b33b695	77	* 0x4 => DMA32 or NORMAL
b70d94ee CL	78	* 0x5 => BAD (DMA+DMA32)
	79	* 0x6 => BAD (HIGHMEM+DMA32)
	80	* 0x7 => BAD (HIGHMEM+DMA32+DMA)
	81	* 0x8 => NORMAL (MOVABLE+0)
	82	* 0x9 => DMA or NORMAL (MOVABLE+DMA)
	83	* 0xa => MOVABLE (Movable is valid only if HIGHMEM is set too)
	84	* 0xb => BAD (MOVABLE+HIGHMEM+DMA)
4b33b695	85	* 0xc => DMA32 or NORMAL (MOVABLE+DMA32)
b70d94ee CL	86	* 0xd => BAD (MOVABLE+DMA32+DMA)
	87	* 0xe => BAD (MOVABLE+DMA32+HIGHMEM)
	88	* 0xf => BAD (MOVABLE+DMA32+HIGHMEM+DMA)
	89	*
b11a7b94	90	* GFP_ZONES_SHIFT must be <= 2 on 32 bit platforms.
b70d94ee CL	91	*/
b70d94ee CL	92
b11a7b94 DW	93	#if defined(CONFIG_ZONE_DEVICE) && (MAX_NR_ZONES-1) <= 4
	94	/* ZONE_DEVICE is not a valid GFP zone specifier */
	95	#define GFP_ZONES_SHIFT 2
	96	#else
	97	#define GFP_ZONES_SHIFT ZONES_SHIFT
	98	#endif
	99
	100	#if 16 * GFP_ZONES_SHIFT > BITS_PER_LONG
	101	#error GFP_ZONES_SHIFT too large to create GFP_ZONE_TABLE integer
b70d94ee CL	102	#endif
	103
	104	#define GFP_ZONE_TABLE ( \
b11a7b94 DW	105	(ZONE_NORMAL << 0 * GFP_ZONES_SHIFT) \
	106	\| (OPT_ZONE_DMA << ___GFP_DMA * GFP_ZONES_SHIFT) \
	107	\| (OPT_ZONE_HIGHMEM << ___GFP_HIGHMEM * GFP_ZONES_SHIFT) \
	108	\| (OPT_ZONE_DMA32 << ___GFP_DMA32 * GFP_ZONES_SHIFT) \
	109	\| (ZONE_NORMAL << ___GFP_MOVABLE * GFP_ZONES_SHIFT) \
	110	\| (OPT_ZONE_DMA << (___GFP_MOVABLE \| ___GFP_DMA) * GFP_ZONES_SHIFT) \
	111	\| (ZONE_MOVABLE << (___GFP_MOVABLE \| ___GFP_HIGHMEM) * GFP_ZONES_SHIFT)\
	112	\| (OPT_ZONE_DMA32 << (___GFP_MOVABLE \| ___GFP_DMA32) * GFP_ZONES_SHIFT)\
b70d94ee CL	113	)
	114
	115	/*
263ff5d8	116	* GFP_ZONE_BAD is a bitmap for all combinations of __GFP_DMA, __GFP_DMA32
b70d94ee CL	117	* __GFP_HIGHMEM and __GFP_MOVABLE that are not permitted. One flag per
	118	* entry starting with bit 0. Bit is set if the combination is not
	119	* allowed.
	120	*/
	121	#define GFP_ZONE_BAD ( \
16b56cf4 NK	122	1 << (___GFP_DMA \| ___GFP_HIGHMEM) \
	123	\| 1 << (___GFP_DMA \| ___GFP_DMA32) \
	124	\| 1 << (___GFP_DMA32 \| ___GFP_HIGHMEM) \
	125	\| 1 << (___GFP_DMA \| ___GFP_DMA32 \| ___GFP_HIGHMEM) \
	126	\| 1 << (___GFP_MOVABLE \| ___GFP_HIGHMEM \| ___GFP_DMA) \
	127	\| 1 << (___GFP_MOVABLE \| ___GFP_DMA32 \| ___GFP_DMA) \
	128	\| 1 << (___GFP_MOVABLE \| ___GFP_DMA32 \| ___GFP_HIGHMEM) \
	129	\| 1 << (___GFP_MOVABLE \| ___GFP_DMA32 \| ___GFP_DMA \| ___GFP_HIGHMEM) \
b70d94ee CL	130	)
	131
	132	static inline enum zone_type gfp_zone(gfp_t flags)
	133	{
	134	enum zone_type z;
16b56cf4	135	int bit = (__force int) (flags & GFP_ZONEMASK);
b70d94ee	136
b11a7b94 DW	137	z = (GFP_ZONE_TABLE >> (bit * GFP_ZONES_SHIFT)) &
b11a7b94 DW	138	((1 << GFP_ZONES_SHIFT) - 1);
82d4b577	139	VM_BUG_ON((GFP_ZONE_BAD >> bit) & 1);
b70d94ee	140	return z;
4e4785bc CL	141	}
4e4785bc CL	142
1da177e4 LT	143	/*
	144	* There is only one page-allocator function, and two main namespaces to
	145	* it. The alloc_page() variants return 'struct page ' and as such
	146	* can allocate highmem pages, the getpage*() variants return
	147	* virtual kernel addresses to the allocated page(s).
	148	*/
	149
54a6eb5c MG	150	static inline int gfp_zonelist(gfp_t flags)
54a6eb5c MG	151	{
c00eb15a YB	152	#ifdef CONFIG_NUMA
	153	if (unlikely(flags & __GFP_THISNODE))
	154	return ZONELIST_NOFALLBACK;
	155	#endif
	156	return ZONELIST_FALLBACK;
54a6eb5c MG	157	}
54a6eb5c MG	158
1c00f936 DC	159	/*
	160	* gfp flag masking for nested internal allocations.
	161	*
	162	* For code that needs to do allocations inside the public allocation API (e.g.
	163	* memory allocation tracking code) the allocations need to obey the caller
	164	* allocation context constrains to prevent allocation context mismatches (e.g.
	165	* GFP_KERNEL allocations in GFP_NOFS contexts) from potential deadlock
	166	* situations.
	167	*
	168	* It is also assumed that these nested allocations are for internal kernel
	169	* object storage purposes only and are not going to be used for DMA, etc. Hence
	170	* we strip out all the zone information and leave just the context information
	171	* intact.
	172	*
	173	* Further, internal allocations must fail before the higher level allocation
	174	* can fail, so we must make them fail faster and fail silently. We also don't
	175	* want them to deplete emergency reserves. Hence nested allocations must be
	176	* prepared for these allocations to fail.
	177	*/
	178	static inline gfp_t gfp_nested_mask(gfp_t flags)
	179	{
	180	return ((flags & (GFP_KERNEL \| GFP_ATOMIC \| __GFP_NOLOCKDEP)) \|
	181	(__GFP_NORETRY \| __GFP_NOMEMALLOC \| __GFP_NOWARN));
	182	}
	183
1da177e4 LT	184	/*
1da177e4 LT	185	* We get the zone list from the current node and the gfp_mask.
cb152a1a	186	* This zone list contains a maximum of MAX_NUMNODES*MAX_NR_ZONES zones.
54a6eb5c MG	187	* There are two zonelists per node, one for all zones with memory and
54a6eb5c MG	188	* one containing just zones from the node the zonelist belongs to.
1da177e4	189	*
d3c251ab MR	190	* For the case of non-NUMA systems the NODE_DATA() gets optimized to
d3c251ab MR	191	* &contig_page_data at compile-time.
1da177e4	192	*/
0e88460d MG	193	static inline struct zonelist *node_zonelist(int nid, gfp_t flags)
0e88460d MG	194	{
54a6eb5c	195	return NODE_DATA(nid)->node_zonelists + gfp_zonelist(flags);
0e88460d	196	}
1da177e4 LT	197
	198	#ifndef HAVE_ARCH_FREE_PAGE
	199	static inline void arch_free_page(struct page *page, int order) { }
	200	#endif
cc102509 NP	201	#ifndef HAVE_ARCH_ALLOC_PAGE
	202	static inline void arch_alloc_page(struct page *page, int order) { }
	203	#endif
1da177e4	204
b951aaff	205	struct page *__alloc_pages_noprof(gfp_t gfp, unsigned int order, int preferred_nid,
84172f4b	206	nodemask_t *nodemask);
b951aaff SB	207	#define __alloc_pages(...) alloc_hooks(__alloc_pages_noprof(__VA_ARGS__))
	208
	209	struct folio *__folio_alloc_noprof(gfp_t gfp, unsigned int order, int preferred_nid,
cc09cb13	210	nodemask_t *nodemask);
b951aaff	211	#define __folio_alloc(...) alloc_hooks(__folio_alloc_noprof(__VA_ARGS__))
e4048e5d	212
b951aaff	213	unsigned long alloc_pages_bulk_noprof(gfp_t gfp, int preferred_nid,
387ba26f	214	nodemask_t *nodemask, int nr_pages,
0f87d9d3 MG	215	struct list_head *page_list,
0f87d9d3 MG	216	struct page **page_array);
b951aaff	217	#define __alloc_pages_bulk(...) alloc_hooks(alloc_pages_bulk_noprof(__VA_ARGS__))
387ba26f	218
b951aaff	219	unsigned long alloc_pages_bulk_array_mempolicy_noprof(gfp_t gfp,
c00b6b96 CW	220	unsigned long nr_pages,
c00b6b96 CW	221	struct page **page_array);
b951aaff SB	222	#define alloc_pages_bulk_array_mempolicy(...) \
b951aaff SB	223	alloc_hooks(alloc_pages_bulk_array_mempolicy_noprof(__VA_ARGS__))
c00b6b96	224
387ba26f	225	/* Bulk allocate order-0 pages */
b951aaff SB	226	#define alloc_pages_bulk_list(_gfp, _nr_pages, _list) \
b951aaff SB	227	__alloc_pages_bulk(_gfp, numa_mem_id(), NULL, _nr_pages, _list, NULL)
0f87d9d3	228
b951aaff SB	229	#define alloc_pages_bulk_array(_gfp, _nr_pages, _page_array) \
b951aaff SB	230	__alloc_pages_bulk(_gfp, numa_mem_id(), NULL, _nr_pages, NULL, _page_array)
387ba26f	231
a2afc59f	232	static inline unsigned long
b951aaff SB	233	alloc_pages_bulk_array_node_noprof(gfp_t gfp, int nid, unsigned long nr_pages,
b951aaff SB	234	struct page **page_array)
a2afc59f URS	235	{
	236	if (nid == NUMA_NO_NODE)
	237	nid = numa_mem_id();
	238
b951aaff	239	return alloc_pages_bulk_noprof(gfp, nid, NULL, nr_pages, NULL, page_array);
a2afc59f URS	240	}
a2afc59f URS	241
b951aaff SB	242	#define alloc_pages_bulk_array_node(...) \
	243	alloc_hooks(alloc_pages_bulk_array_node_noprof(__VA_ARGS__))
	244
dec1d352 YS	245	static inline void warn_if_node_offline(int this_node, gfp_t gfp_mask)
	246	{
	247	gfp_t warn_gfp = gfp_mask & (__GFP_THISNODE\|__GFP_NOWARN);
	248
	249	if (warn_gfp != (__GFP_THISNODE\|__GFP_NOWARN))
	250	return;
	251
	252	if (node_online(this_node))
	253	return;
	254
	255	pr_warn("%pGg allocation from offline node %d\n", &gfp_mask, this_node);
	256	dump_stack();
	257	}
	258
96db800f VB	259	/*
	260	* Allocate pages, preferring the node given as nid. The node must be valid and
	261	* online. For more general interface, see alloc_pages_node().
	262	*/
	263	static inline struct page *
b951aaff	264	__alloc_pages_node_noprof(int nid, gfp_t gfp_mask, unsigned int order)
1da177e4	265	{
0bc35a97	266	VM_BUG_ON(nid < 0 \|\| nid >= MAX_NUMNODES);
dec1d352	267	warn_if_node_offline(nid, gfp_mask);
819a6928	268
b951aaff	269	return __alloc_pages_noprof(gfp_mask, order, nid, NULL);
1da177e4 LT	270	}
1da177e4 LT	271
b951aaff SB	272	#define __alloc_pages_node(...) alloc_hooks(__alloc_pages_node_noprof(__VA_ARGS__))
b951aaff SB	273
cc09cb13	274	static inline
b951aaff	275	struct folio *__folio_alloc_node_noprof(gfp_t gfp, unsigned int order, int nid)
cc09cb13 MWO	276	{
cc09cb13 MWO	277	VM_BUG_ON(nid < 0 \|\| nid >= MAX_NUMNODES);
dec1d352	278	warn_if_node_offline(nid, gfp);
cc09cb13	279
b951aaff	280	return __folio_alloc_noprof(gfp, order, nid, NULL);
cc09cb13 MWO	281	}
cc09cb13 MWO	282
b951aaff SB	283	#define __folio_alloc_node(...) alloc_hooks(__folio_alloc_node_noprof(__VA_ARGS__))
b951aaff SB	284
96db800f VB	285	/*
96db800f VB	286	* Allocate pages, preferring the node given as nid. When nid == NUMA_NO_NODE,
82c1fc71 VB	287	* prefer the current CPU's closest node. Otherwise node must be valid and
82c1fc71 VB	288	* online.
96db800f	289	*/
b951aaff SB	290	static inline struct page *alloc_pages_node_noprof(int nid, gfp_t gfp_mask,
b951aaff SB	291	unsigned int order)
6484eb3e	292	{
0bc35a97	293	if (nid == NUMA_NO_NODE)
82c1fc71	294	nid = numa_mem_id();
6484eb3e	295
b951aaff	296	return __alloc_pages_node_noprof(nid, gfp_mask, order);
6484eb3e MG	297	}
6484eb3e MG	298
b951aaff SB	299	#define alloc_pages_node(...) alloc_hooks(alloc_pages_node_noprof(__VA_ARGS__))
b951aaff SB	300
1da177e4	301	#ifdef CONFIG_NUMA
b951aaff SB	302	struct page *alloc_pages_noprof(gfp_t gfp, unsigned int order);
b951aaff SB	303	struct page *alloc_pages_mpol_noprof(gfp_t gfp, unsigned int order,
ddc1a5cb	304	struct mempolicy *mpol, pgoff_t ilx, int nid);
b951aaff SB	305	struct folio *folio_alloc_noprof(gfp_t gfp, unsigned int order);
b951aaff SB	306	struct folio vma_alloc_folio_noprof(gfp_t gfp, int order, struct vm_area_struct vma,
f584b680	307	unsigned long addr, bool hugepage);
1da177e4	308	#else
b951aaff	309	static inline struct page *alloc_pages_noprof(gfp_t gfp_mask, unsigned int order)
43ee5b6d	310	{
b951aaff	311	return alloc_pages_node_noprof(numa_node_id(), gfp_mask, order);
43ee5b6d	312	}
b951aaff	313	static inline struct page *alloc_pages_mpol_noprof(gfp_t gfp, unsigned int order,
ddc1a5cb HD	314	struct mempolicy *mpol, pgoff_t ilx, int nid)
ddc1a5cb HD	315	{
b951aaff	316	return alloc_pages_noprof(gfp, order);
ddc1a5cb	317	}
b951aaff	318	static inline struct folio *folio_alloc_noprof(gfp_t gfp, unsigned int order)
cc09cb13 MWO	319	{
	320	return __folio_alloc_node(gfp, order, numa_node_id());
	321	}
b951aaff SB	322	#define vma_alloc_folio_noprof(gfp, order, vma, addr, hugepage) \
b951aaff SB	323	folio_alloc_noprof(gfp, order)
1da177e4	324	#endif
b951aaff SB	325
	326	#define alloc_pages(...) alloc_hooks(alloc_pages_noprof(__VA_ARGS__))
	327	#define alloc_pages_mpol(...) alloc_hooks(alloc_pages_mpol_noprof(__VA_ARGS__))
	328	#define folio_alloc(...) alloc_hooks(folio_alloc_noprof(__VA_ARGS__))
	329	#define vma_alloc_folio(...) alloc_hooks(vma_alloc_folio_noprof(__VA_ARGS__))
	330
1da177e4	331	#define alloc_page(gfp_mask) alloc_pages(gfp_mask, 0)
b951aaff SB	332
b951aaff SB	333	static inline struct page *alloc_page_vma_noprof(gfp_t gfp,
adf88aa8 MWO	334	struct vm_area_struct *vma, unsigned long addr)
adf88aa8 MWO	335	{
b951aaff	336	struct folio *folio = vma_alloc_folio_noprof(gfp, 0, vma, addr, false);
adf88aa8 MWO	337
	338	return &folio->page;
	339	}
b951aaff SB	340	#define alloc_page_vma(...) alloc_hooks(alloc_page_vma_noprof(__VA_ARGS__))
	341
	342	extern unsigned long get_free_pages_noprof(gfp_t gfp_mask, unsigned int order);
	343	#define __get_free_pages(...) alloc_hooks(get_free_pages_noprof(__VA_ARGS__))
1da177e4	344
b951aaff SB	345	extern unsigned long get_zeroed_page_noprof(gfp_t gfp_mask);
	346	#define get_zeroed_page(...) alloc_hooks(get_zeroed_page_noprof(__VA_ARGS__))
	347
	348	void *alloc_pages_exact_noprof(size_t size, gfp_t gfp_mask) __alloc_size(1);
	349	#define alloc_pages_exact(...) alloc_hooks(alloc_pages_exact_noprof(__VA_ARGS__))
1da177e4	350
2be0ffe2 TT	351	void free_pages_exact(void *virt, size_t size);
2be0ffe2 TT	352
b951aaff SB	353	__meminit void *alloc_pages_exact_nid_noprof(int nid, size_t size, gfp_t gfp_mask) __alloc_size(2);
	354	#define alloc_pages_exact_nid(...) \
	355	alloc_hooks(alloc_pages_exact_nid_noprof(__VA_ARGS__))
	356
	357	#define __get_free_page(gfp_mask) \
	358	__get_free_pages((gfp_mask), 0)
1da177e4	359
b951aaff SB	360	#define __get_dma_pages(gfp_mask, order) \
b951aaff SB	361	__get_free_pages((gfp_mask) \| GFP_DMA, (order))
1da177e4	362
b3c97528 HH	363	extern void __free_pages(struct page *page, unsigned int order);
b3c97528 HH	364	extern void free_pages(unsigned long addr, unsigned int order);
1da177e4	365
b63ae8ca	366	struct page_frag_cache;
a0727489	367	void page_frag_cache_drain(struct page_frag_cache *nc);
2976db80	368	extern void __page_frag_cache_drain(struct page *page, unsigned int count);
411c5f36 YL	369	void __page_frag_alloc_align(struct page_frag_cache nc, unsigned int fragsz,
	370	gfp_t gfp_mask, unsigned int align_mask);
	371
	372	static inline void page_frag_alloc_align(struct page_frag_cache nc,
	373	unsigned int fragsz, gfp_t gfp_mask,
	374	unsigned int align)
	375	{
	376	WARN_ON_ONCE(!is_power_of_2(align));
	377	return __page_frag_alloc_align(nc, fragsz, gfp_mask, -align);
	378	}
b358e212 KH	379
	380	static inline void page_frag_alloc(struct page_frag_cache nc,
	381	unsigned int fragsz, gfp_t gfp_mask)
	382	{
411c5f36	383	return __page_frag_alloc_align(nc, fragsz, gfp_mask, ~0u);
b358e212 KH	384	}
b358e212 KH	385
8c2dd3e4	386	extern void page_frag_free(void *addr);
b63ae8ca	387
1da177e4	388	#define __free_page(page) __free_pages((page), 0)
fd23855e	389	#define free_page(addr) free_pages((addr), 0)
1da177e4	390
c4fbed4b	391	void page_alloc_init_cpuhp(void);
51a755c5	392	int decay_pcp_high(struct zone zone, struct per_cpu_pages pcp);
4037d452	393	void drain_zone_pages(struct zone zone, struct per_cpu_pages pcp);
93481ff0 VB	394	void drain_all_pages(struct zone *zone);
93481ff0 VB	395	void drain_local_pages(struct zone *zone);
1da177e4	396
0e1cc95b	397	void page_alloc_init_late(void);
5cec4eb7	398	void setup_pcp_cacheinfo(unsigned int cpu);
0e1cc95b	399
f90ac398 MG	400	/*
	401	* gfp_allowed_mask is set to GFP_BOOT_MASK during early boot to restrict what
	402	* GFP flags are used before interrupts are enabled. Once interrupts are
	403	* enabled, it is set to __GFP_BITS_MASK while the system is running. During
	404	* hibernation, it is used by PM to avoid I/O during memory allocation while
	405	* devices are suspended.
	406	*/
dcce284a BH	407	extern gfp_t gfp_allowed_mask;
dcce284a BH	408
c93bdd0e MG	409	/* Returns true if the gfp_mask allows use of ALLOC_NO_WATERMARK */
	410	bool gfp_pfmemalloc_allowed(gfp_t gfp_mask);
	411
07f44ac3	412	static inline bool gfp_has_io_fs(gfp_t gfp)
f90ac398	413	{
07f44ac3	414	return (gfp & (__GFP_IO \| __GFP_FS)) == (__GFP_IO \| __GFP_FS);
f90ac398	415	}
07f44ac3	416
803de900 VB	417	/*
	418	* Check if the gfp flags allow compaction - GFP_NOIO is a really
	419	* tricky context because the migration might require IO.
	420	*/
	421	static inline bool gfp_compaction_allowed(gfp_t gfp_mask)
	422	{
	423	return IS_ENABLED(CONFIG_COMPACTION) && (gfp_mask & __GFP_IO);
	424	}
	425
07f44ac3	426	extern gfp_t vma_thp_gfp_mask(struct vm_area_struct *vma);
f90ac398	427
8df995f6	428	#ifdef CONFIG_CONTIG_ALLOC
041d3a8c	429	/* The below functions must be run on a range from a single zone. */
b951aaff	430	extern int alloc_contig_range_noprof(unsigned long start, unsigned long end,
ca96b625	431	unsigned migratetype, gfp_t gfp_mask);
b951aaff SB	432	#define alloc_contig_range(...) alloc_hooks(alloc_contig_range_noprof(__VA_ARGS__))
	433
	434	extern struct page *alloc_contig_pages_noprof(unsigned long nr_pages, gfp_t gfp_mask,
	435	int nid, nodemask_t *nodemask);
	436	#define alloc_contig_pages(...) alloc_hooks(alloc_contig_pages_noprof(__VA_ARGS__))
	437
080fe206	438	#endif
78fa5150	439	void free_contig_range(unsigned long pfn, unsigned long nr_pages);
041d3a8c	440
1da177e4	441	#endif /* __LINUX_GFP_H */