[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);
}

static inline bool gfpflags_allow_spinning(const gfp_t gfp_flags)
{
	/*
	 * !__GFP_DIRECT_RECLAIM -> direct claim is not allowed.
	 * !__GFP_KSWAPD_RECLAIM -> it's not safe to wake up kswapd.
	 * All GFP_* flags including GFP_NOWAIT use one or both flags.
	 * alloc_pages_nolock() is the only API that doesn't specify either flag.
	 *
	 * This is stronger than GFP_NOWAIT or GFP_ATOMIC because
	 * those are guaranteed to never block on a sleeping lock.
	 * Here we are enforcing that the allocation doesn't ever spin
	 * on any locks (i.e. only trylocks). There is no high level
	 * GFP_$FOO flag for this use in alloc_pages_nolock() as the
	 * regular page allocator doesn't fully support this
	 * allocation mode.
	 */
	return !!(gfp_flags & __GFP_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 page **page_array);
#define __alloc_pages_bulk(...)			alloc_hooks(alloc_pages_bulk_noprof(__VA_ARGS__))

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

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

static inline unsigned long
alloc_pages_bulk_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, page_array);
}

#define alloc_pages_bulk_node(...)				\
	alloc_hooks(alloc_pages_bulk_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 folio *folio_alloc_noprof(gfp_t gfp, unsigned int order);
struct folio *folio_alloc_mpol_noprof(gfp_t gfp, unsigned int order,
		struct mempolicy *mpol, pgoff_t ilx, int nid);
struct folio *vma_alloc_folio_noprof(gfp_t gfp, int order, struct vm_area_struct *vma,
		unsigned long addr);
#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 folio *folio_alloc_noprof(gfp_t gfp, unsigned int order)
{
	return __folio_alloc_node_noprof(gfp, order, numa_node_id());
}
static inline struct folio *folio_alloc_mpol_noprof(gfp_t gfp, unsigned int order,
		struct mempolicy *mpol, pgoff_t ilx, int nid)
{
	return folio_alloc_noprof(gfp, order);
}
#define vma_alloc_folio_noprof(gfp, order, vma, addr)		\
	folio_alloc_noprof(gfp, order)
#endif

#define alloc_pages(...)			alloc_hooks(alloc_pages_noprof(__VA_ARGS__))
#define folio_alloc(...)			alloc_hooks(folio_alloc_noprof(__VA_ARGS__))
#define folio_alloc_mpol(...)			alloc_hooks(folio_alloc_mpol_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);

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

struct page *alloc_pages_nolock_noprof(int nid, unsigned int order);
#define alloc_pages_nolock(...)			alloc_hooks(alloc_pages_nolock_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_nolock(struct page *page, unsigned int order);
extern void free_pages(unsigned long addr, unsigned int order);

#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);

#ifdef CONFIG_CONTIG_ALLOC
static inline struct folio *folio_alloc_gigantic_noprof(int order, gfp_t gfp,
							int nid, nodemask_t *node)
{
	struct page *page;

	if (WARN_ON(!order || !(gfp & __GFP_COMP)))
		return NULL;

	page = alloc_contig_pages_noprof(1 << order, gfp, nid, node);

	return page ? page_folio(page) : NULL;
}
#else
static inline struct folio *folio_alloc_gigantic_noprof(int order, gfp_t gfp,
							int nid, nodemask_t *node)
{
	return NULL;
}
#endif
/* This should be paired with folio_put() rather than free_contig_range(). */
#define folio_alloc_gigantic(...) alloc_hooks(folio_alloc_gigantic_noprof(__VA_ARGS__))

#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
97769a53 AS	42	static inline bool gfpflags_allow_spinning(const gfp_t gfp_flags)
	43	{
	44	/*
	45	* !__GFP_DIRECT_RECLAIM -> direct claim is not allowed.
	46	* !__GFP_KSWAPD_RECLAIM -> it's not safe to wake up kswapd.
	47	* All GFP_* flags including GFP_NOWAIT use one or both flags.
2aad4edf	48	* alloc_pages_nolock() is the only API that doesn't specify either flag.
97769a53 AS	49	*
	50	* This is stronger than GFP_NOWAIT or GFP_ATOMIC because
	51	* those are guaranteed to never block on a sleeping lock.
	52	* Here we are enforcing that the allocation doesn't ever spin
	53	* on any locks (i.e. only trylocks). There is no high level
2aad4edf	54	* GFP_$FOO flag for this use in alloc_pages_nolock() as the
97769a53 AS	55	* regular page allocator doesn't fully support this
	56	* allocation mode.
	57	*/
f90b474a	58	return !!(gfp_flags & __GFP_RECLAIM);
97769a53 AS	59	}
97769a53 AS	60
b70d94ee CL	61	#ifdef CONFIG_HIGHMEM
	62	#define OPT_ZONE_HIGHMEM ZONE_HIGHMEM
	63	#else
	64	#define OPT_ZONE_HIGHMEM ZONE_NORMAL
	65	#endif
	66
4b51d669	67	#ifdef CONFIG_ZONE_DMA
b70d94ee CL	68	#define OPT_ZONE_DMA ZONE_DMA
	69	#else
	70	#define OPT_ZONE_DMA ZONE_NORMAL
4b51d669	71	#endif
b70d94ee	72
4e4785bc	73	#ifdef CONFIG_ZONE_DMA32
b70d94ee CL	74	#define OPT_ZONE_DMA32 ZONE_DMA32
	75	#else
	76	#define OPT_ZONE_DMA32 ZONE_NORMAL
4e4785bc	77	#endif
b70d94ee CL	78
	79	/*
	80	* GFP_ZONE_TABLE is a word size bitstring that is used for looking up the
ac2e8e40 HL	81	* zone to use given the lowest 4 bits of gfp_t. Entries are GFP_ZONES_SHIFT
ac2e8e40 HL	82	* bits long and there are 16 of them to cover all possible combinations of
263ff5d8	83	* __GFP_DMA, __GFP_DMA32, __GFP_MOVABLE and __GFP_HIGHMEM.
b70d94ee CL	84	*
	85	* The zone fallback order is MOVABLE=>HIGHMEM=>NORMAL=>DMA32=>DMA.
	86	* But GFP_MOVABLE is not only a zone specifier but also an allocation
	87	* policy. Therefore __GFP_MOVABLE plus another zone selector is valid.
263ff5d8	88	* Only 1 bit of the lowest 3 bits (DMA,DMA32,HIGHMEM) can be set to "1".
b70d94ee CL	89	*
	90	* bit result
	91	* =================
	92	* 0x0 => NORMAL
	93	* 0x1 => DMA or NORMAL
	94	* 0x2 => HIGHMEM or NORMAL
	95	* 0x3 => BAD (DMA+HIGHMEM)
4b33b695	96	* 0x4 => DMA32 or NORMAL
b70d94ee CL	97	* 0x5 => BAD (DMA+DMA32)
	98	* 0x6 => BAD (HIGHMEM+DMA32)
	99	* 0x7 => BAD (HIGHMEM+DMA32+DMA)
	100	* 0x8 => NORMAL (MOVABLE+0)
	101	* 0x9 => DMA or NORMAL (MOVABLE+DMA)
	102	* 0xa => MOVABLE (Movable is valid only if HIGHMEM is set too)
	103	* 0xb => BAD (MOVABLE+HIGHMEM+DMA)
4b33b695	104	* 0xc => DMA32 or NORMAL (MOVABLE+DMA32)
b70d94ee CL	105	* 0xd => BAD (MOVABLE+DMA32+DMA)
	106	* 0xe => BAD (MOVABLE+DMA32+HIGHMEM)
	107	* 0xf => BAD (MOVABLE+DMA32+HIGHMEM+DMA)
	108	*
b11a7b94	109	* GFP_ZONES_SHIFT must be <= 2 on 32 bit platforms.
b70d94ee CL	110	*/
b70d94ee CL	111
b11a7b94 DW	112	#if defined(CONFIG_ZONE_DEVICE) && (MAX_NR_ZONES-1) <= 4
	113	/* ZONE_DEVICE is not a valid GFP zone specifier */
	114	#define GFP_ZONES_SHIFT 2
	115	#else
	116	#define GFP_ZONES_SHIFT ZONES_SHIFT
	117	#endif
	118
	119	#if 16 * GFP_ZONES_SHIFT > BITS_PER_LONG
	120	#error GFP_ZONES_SHIFT too large to create GFP_ZONE_TABLE integer
b70d94ee CL	121	#endif
	122
	123	#define GFP_ZONE_TABLE ( \
b11a7b94 DW	124	(ZONE_NORMAL << 0 * GFP_ZONES_SHIFT) \
	125	\| (OPT_ZONE_DMA << ___GFP_DMA * GFP_ZONES_SHIFT) \
	126	\| (OPT_ZONE_HIGHMEM << ___GFP_HIGHMEM * GFP_ZONES_SHIFT) \
	127	\| (OPT_ZONE_DMA32 << ___GFP_DMA32 * GFP_ZONES_SHIFT) \
	128	\| (ZONE_NORMAL << ___GFP_MOVABLE * GFP_ZONES_SHIFT) \
	129	\| (OPT_ZONE_DMA << (___GFP_MOVABLE \| ___GFP_DMA) * GFP_ZONES_SHIFT) \
	130	\| (ZONE_MOVABLE << (___GFP_MOVABLE \| ___GFP_HIGHMEM) * GFP_ZONES_SHIFT)\
	131	\| (OPT_ZONE_DMA32 << (___GFP_MOVABLE \| ___GFP_DMA32) * GFP_ZONES_SHIFT)\
b70d94ee CL	132	)
	133
	134	/*
263ff5d8	135	* GFP_ZONE_BAD is a bitmap for all combinations of __GFP_DMA, __GFP_DMA32
b70d94ee CL	136	* __GFP_HIGHMEM and __GFP_MOVABLE that are not permitted. One flag per
	137	* entry starting with bit 0. Bit is set if the combination is not
	138	* allowed.
	139	*/
	140	#define GFP_ZONE_BAD ( \
16b56cf4 NK	141	1 << (___GFP_DMA \| ___GFP_HIGHMEM) \
	142	\| 1 << (___GFP_DMA \| ___GFP_DMA32) \
	143	\| 1 << (___GFP_DMA32 \| ___GFP_HIGHMEM) \
	144	\| 1 << (___GFP_DMA \| ___GFP_DMA32 \| ___GFP_HIGHMEM) \
	145	\| 1 << (___GFP_MOVABLE \| ___GFP_HIGHMEM \| ___GFP_DMA) \
	146	\| 1 << (___GFP_MOVABLE \| ___GFP_DMA32 \| ___GFP_DMA) \
	147	\| 1 << (___GFP_MOVABLE \| ___GFP_DMA32 \| ___GFP_HIGHMEM) \
	148	\| 1 << (___GFP_MOVABLE \| ___GFP_DMA32 \| ___GFP_DMA \| ___GFP_HIGHMEM) \
b70d94ee CL	149	)
	150
	151	static inline enum zone_type gfp_zone(gfp_t flags)
	152	{
	153	enum zone_type z;
16b56cf4	154	int bit = (__force int) (flags & GFP_ZONEMASK);
b70d94ee	155
b11a7b94 DW	156	z = (GFP_ZONE_TABLE >> (bit * GFP_ZONES_SHIFT)) &
b11a7b94 DW	157	((1 << GFP_ZONES_SHIFT) - 1);
82d4b577	158	VM_BUG_ON((GFP_ZONE_BAD >> bit) & 1);
b70d94ee	159	return z;
4e4785bc CL	160	}
4e4785bc CL	161
1da177e4 LT	162	/*
	163	* There is only one page-allocator function, and two main namespaces to
	164	* it. The alloc_page() variants return 'struct page ' and as such
	165	* can allocate highmem pages, the getpage*() variants return
	166	* virtual kernel addresses to the allocated page(s).
	167	*/
	168
54a6eb5c MG	169	static inline int gfp_zonelist(gfp_t flags)
54a6eb5c MG	170	{
c00eb15a YB	171	#ifdef CONFIG_NUMA
	172	if (unlikely(flags & __GFP_THISNODE))
	173	return ZONELIST_NOFALLBACK;
	174	#endif
	175	return ZONELIST_FALLBACK;
54a6eb5c MG	176	}
54a6eb5c MG	177
1c00f936 DC	178	/*
	179	* gfp flag masking for nested internal allocations.
	180	*
	181	* For code that needs to do allocations inside the public allocation API (e.g.
	182	* memory allocation tracking code) the allocations need to obey the caller
	183	* allocation context constrains to prevent allocation context mismatches (e.g.
	184	* GFP_KERNEL allocations in GFP_NOFS contexts) from potential deadlock
	185	* situations.
	186	*
	187	* It is also assumed that these nested allocations are for internal kernel
	188	* object storage purposes only and are not going to be used for DMA, etc. Hence
	189	* we strip out all the zone information and leave just the context information
	190	* intact.
	191	*
	192	* Further, internal allocations must fail before the higher level allocation
	193	* can fail, so we must make them fail faster and fail silently. We also don't
	194	* want them to deplete emergency reserves. Hence nested allocations must be
	195	* prepared for these allocations to fail.
	196	*/
	197	static inline gfp_t gfp_nested_mask(gfp_t flags)
	198	{
	199	return ((flags & (GFP_KERNEL \| GFP_ATOMIC \| __GFP_NOLOCKDEP)) \|
	200	(__GFP_NORETRY \| __GFP_NOMEMALLOC \| __GFP_NOWARN));
	201	}
	202
1da177e4 LT	203	/*
1da177e4 LT	204	* We get the zone list from the current node and the gfp_mask.
cb152a1a	205	* This zone list contains a maximum of MAX_NUMNODES*MAX_NR_ZONES zones.
54a6eb5c MG	206	* There are two zonelists per node, one for all zones with memory and
54a6eb5c MG	207	* one containing just zones from the node the zonelist belongs to.
1da177e4	208	*
d3c251ab MR	209	* For the case of non-NUMA systems the NODE_DATA() gets optimized to
d3c251ab MR	210	* &contig_page_data at compile-time.
1da177e4	211	*/
0e88460d MG	212	static inline struct zonelist *node_zonelist(int nid, gfp_t flags)
0e88460d MG	213	{
54a6eb5c	214	return NODE_DATA(nid)->node_zonelists + gfp_zonelist(flags);
0e88460d	215	}
1da177e4 LT	216
	217	#ifndef HAVE_ARCH_FREE_PAGE
	218	static inline void arch_free_page(struct page *page, int order) { }
	219	#endif
cc102509 NP	220	#ifndef HAVE_ARCH_ALLOC_PAGE
	221	static inline void arch_alloc_page(struct page *page, int order) { }
	222	#endif
1da177e4	223
b951aaff	224	struct page *__alloc_pages_noprof(gfp_t gfp, unsigned int order, int preferred_nid,
84172f4b	225	nodemask_t *nodemask);
b951aaff SB	226	#define __alloc_pages(...) alloc_hooks(__alloc_pages_noprof(__VA_ARGS__))
	227
	228	struct folio *__folio_alloc_noprof(gfp_t gfp, unsigned int order, int preferred_nid,
cc09cb13	229	nodemask_t *nodemask);
b951aaff	230	#define __folio_alloc(...) alloc_hooks(__folio_alloc_noprof(__VA_ARGS__))
e4048e5d	231
b951aaff	232	unsigned long alloc_pages_bulk_noprof(gfp_t gfp, int preferred_nid,
387ba26f	233	nodemask_t *nodemask, int nr_pages,
0f87d9d3	234	struct page **page_array);
b951aaff	235	#define __alloc_pages_bulk(...) alloc_hooks(alloc_pages_bulk_noprof(__VA_ARGS__))
387ba26f	236
6bf9b5b4	237	unsigned long alloc_pages_bulk_mempolicy_noprof(gfp_t gfp,
c00b6b96 CW	238	unsigned long nr_pages,
c00b6b96 CW	239	struct page **page_array);
6bf9b5b4 LC	240	#define alloc_pages_bulk_mempolicy(...) \
6bf9b5b4 LC	241	alloc_hooks(alloc_pages_bulk_mempolicy_noprof(__VA_ARGS__))
c00b6b96	242
387ba26f	243	/* Bulk allocate order-0 pages */
6bf9b5b4	244	#define alloc_pages_bulk(_gfp, _nr_pages, _page_array) \
c8b97953	245	__alloc_pages_bulk(_gfp, numa_mem_id(), NULL, _nr_pages, _page_array)
387ba26f	246
a2afc59f	247	static inline unsigned long
6bf9b5b4	248	alloc_pages_bulk_node_noprof(gfp_t gfp, int nid, unsigned long nr_pages,
b951aaff	249	struct page **page_array)
a2afc59f URS	250	{
	251	if (nid == NUMA_NO_NODE)
	252	nid = numa_mem_id();
	253
c8b97953	254	return alloc_pages_bulk_noprof(gfp, nid, NULL, nr_pages, page_array);
a2afc59f URS	255	}
a2afc59f URS	256
6bf9b5b4 LC	257	#define alloc_pages_bulk_node(...) \
6bf9b5b4 LC	258	alloc_hooks(alloc_pages_bulk_node_noprof(__VA_ARGS__))
b951aaff	259
dec1d352 YS	260	static inline void warn_if_node_offline(int this_node, gfp_t gfp_mask)
	261	{
	262	gfp_t warn_gfp = gfp_mask & (__GFP_THISNODE\|__GFP_NOWARN);
	263
	264	if (warn_gfp != (__GFP_THISNODE\|__GFP_NOWARN))
	265	return;
	266
	267	if (node_online(this_node))
	268	return;
	269
	270	pr_warn("%pGg allocation from offline node %d\n", &gfp_mask, this_node);
	271	dump_stack();
	272	}
	273
96db800f VB	274	/*
	275	* Allocate pages, preferring the node given as nid. The node must be valid and
	276	* online. For more general interface, see alloc_pages_node().
	277	*/
	278	static inline struct page *
b951aaff	279	__alloc_pages_node_noprof(int nid, gfp_t gfp_mask, unsigned int order)
1da177e4	280	{
0bc35a97	281	VM_BUG_ON(nid < 0 \|\| nid >= MAX_NUMNODES);
dec1d352	282	warn_if_node_offline(nid, gfp_mask);
819a6928	283
b951aaff	284	return __alloc_pages_noprof(gfp_mask, order, nid, NULL);
1da177e4 LT	285	}
1da177e4 LT	286
b951aaff SB	287	#define __alloc_pages_node(...) alloc_hooks(__alloc_pages_node_noprof(__VA_ARGS__))
b951aaff SB	288
cc09cb13	289	static inline
b951aaff	290	struct folio *__folio_alloc_node_noprof(gfp_t gfp, unsigned int order, int nid)
cc09cb13 MWO	291	{
cc09cb13 MWO	292	VM_BUG_ON(nid < 0 \|\| nid >= MAX_NUMNODES);
dec1d352	293	warn_if_node_offline(nid, gfp);
cc09cb13	294
b951aaff	295	return __folio_alloc_noprof(gfp, order, nid, NULL);
cc09cb13 MWO	296	}
cc09cb13 MWO	297
b951aaff SB	298	#define __folio_alloc_node(...) alloc_hooks(__folio_alloc_node_noprof(__VA_ARGS__))
b951aaff SB	299
96db800f VB	300	/*
96db800f VB	301	* Allocate pages, preferring the node given as nid. When nid == NUMA_NO_NODE,
82c1fc71 VB	302	* prefer the current CPU's closest node. Otherwise node must be valid and
82c1fc71 VB	303	* online.
96db800f	304	*/
b951aaff SB	305	static inline struct page *alloc_pages_node_noprof(int nid, gfp_t gfp_mask,
b951aaff SB	306	unsigned int order)
6484eb3e	307	{
0bc35a97	308	if (nid == NUMA_NO_NODE)
82c1fc71	309	nid = numa_mem_id();
6484eb3e	310
b951aaff	311	return __alloc_pages_node_noprof(nid, gfp_mask, order);
6484eb3e MG	312	}
6484eb3e MG	313
b951aaff SB	314	#define alloc_pages_node(...) alloc_hooks(alloc_pages_node_noprof(__VA_ARGS__))
b951aaff SB	315
1da177e4	316	#ifdef CONFIG_NUMA
b951aaff	317	struct page *alloc_pages_noprof(gfp_t gfp, unsigned int order);
b951aaff	318	struct folio *folio_alloc_noprof(gfp_t gfp, unsigned int order);
a19621ed KW	319	struct folio *folio_alloc_mpol_noprof(gfp_t gfp, unsigned int order,
a19621ed KW	320	struct mempolicy *mpol, pgoff_t ilx, int nid);
b951aaff	321	struct folio vma_alloc_folio_noprof(gfp_t gfp, int order, struct vm_area_struct vma,
6359c39c	322	unsigned long addr);
1da177e4	323	#else
b951aaff	324	static inline struct page *alloc_pages_noprof(gfp_t gfp_mask, unsigned int order)
43ee5b6d	325	{
b951aaff	326	return alloc_pages_node_noprof(numa_node_id(), gfp_mask, order);
43ee5b6d	327	}
b951aaff	328	static inline struct folio *folio_alloc_noprof(gfp_t gfp, unsigned int order)
cc09cb13	329	{
0a6fff20	330	return __folio_alloc_node_noprof(gfp, order, numa_node_id());
cc09cb13	331	}
a19621ed KW	332	static inline struct folio *folio_alloc_mpol_noprof(gfp_t gfp, unsigned int order,
	333	struct mempolicy *mpol, pgoff_t ilx, int nid)
	334	{
	335	return folio_alloc_noprof(gfp, order);
	336	}
6359c39c	337	#define vma_alloc_folio_noprof(gfp, order, vma, addr) \
b951aaff	338	folio_alloc_noprof(gfp, order)
1da177e4	339	#endif
b951aaff SB	340
b951aaff SB	341	#define alloc_pages(...) alloc_hooks(alloc_pages_noprof(__VA_ARGS__))
b951aaff	342	#define folio_alloc(...) alloc_hooks(folio_alloc_noprof(__VA_ARGS__))
a19621ed	343	#define folio_alloc_mpol(...) alloc_hooks(folio_alloc_mpol_noprof(__VA_ARGS__))
b951aaff SB	344	#define vma_alloc_folio(...) alloc_hooks(vma_alloc_folio_noprof(__VA_ARGS__))
b951aaff SB	345
1da177e4	346	#define alloc_page(gfp_mask) alloc_pages(gfp_mask, 0)
b951aaff SB	347
b951aaff SB	348	static inline struct page *alloc_page_vma_noprof(gfp_t gfp,
adf88aa8 MWO	349	struct vm_area_struct *vma, unsigned long addr)
adf88aa8 MWO	350	{
6359c39c	351	struct folio *folio = vma_alloc_folio_noprof(gfp, 0, vma, addr);
adf88aa8 MWO	352
	353	return &folio->page;
	354	}
b951aaff SB	355	#define alloc_page_vma(...) alloc_hooks(alloc_page_vma_noprof(__VA_ARGS__))
b951aaff SB	356
2aad4edf AS	357	struct page *alloc_pages_nolock_noprof(int nid, unsigned int order);
2aad4edf AS	358	#define alloc_pages_nolock(...) alloc_hooks(alloc_pages_nolock_noprof(__VA_ARGS__))
97769a53	359
b951aaff SB	360	extern unsigned long get_free_pages_noprof(gfp_t gfp_mask, unsigned int order);
b951aaff SB	361	#define __get_free_pages(...) alloc_hooks(get_free_pages_noprof(__VA_ARGS__))
1da177e4	362
b951aaff SB	363	extern unsigned long get_zeroed_page_noprof(gfp_t gfp_mask);
	364	#define get_zeroed_page(...) alloc_hooks(get_zeroed_page_noprof(__VA_ARGS__))
	365
	366	void *alloc_pages_exact_noprof(size_t size, gfp_t gfp_mask) __alloc_size(1);
	367	#define alloc_pages_exact(...) alloc_hooks(alloc_pages_exact_noprof(__VA_ARGS__))
1da177e4	368
2be0ffe2 TT	369	void free_pages_exact(void *virt, size_t size);
2be0ffe2 TT	370
b951aaff SB	371	__meminit void *alloc_pages_exact_nid_noprof(int nid, size_t size, gfp_t gfp_mask) __alloc_size(2);
	372	#define alloc_pages_exact_nid(...) \
	373	alloc_hooks(alloc_pages_exact_nid_noprof(__VA_ARGS__))
	374
	375	#define __get_free_page(gfp_mask) \
	376	__get_free_pages((gfp_mask), 0)
1da177e4	377
b951aaff SB	378	#define __get_dma_pages(gfp_mask, order) \
b951aaff SB	379	__get_free_pages((gfp_mask) \| GFP_DMA, (order))
1da177e4	380
b3c97528	381	extern void __free_pages(struct page *page, unsigned int order);
8c57b687	382	extern void free_pages_nolock(struct page *page, unsigned int order);
b3c97528	383	extern void free_pages(unsigned long addr, unsigned int order);
1da177e4 LT	384
1da177e4 LT	385	#define __free_page(page) __free_pages((page), 0)
fd23855e	386	#define free_page(addr) free_pages((addr), 0)
1da177e4	387
c4fbed4b	388	void page_alloc_init_cpuhp(void);
51a755c5	389	int decay_pcp_high(struct zone zone, struct per_cpu_pages pcp);
4037d452	390	void drain_zone_pages(struct zone zone, struct per_cpu_pages pcp);
93481ff0 VB	391	void drain_all_pages(struct zone *zone);
93481ff0 VB	392	void drain_local_pages(struct zone *zone);
1da177e4	393
0e1cc95b	394	void page_alloc_init_late(void);
5cec4eb7	395	void setup_pcp_cacheinfo(unsigned int cpu);
0e1cc95b	396
f90ac398 MG	397	/*
	398	* gfp_allowed_mask is set to GFP_BOOT_MASK during early boot to restrict what
	399	* GFP flags are used before interrupts are enabled. Once interrupts are
	400	* enabled, it is set to __GFP_BITS_MASK while the system is running. During
	401	* hibernation, it is used by PM to avoid I/O during memory allocation while
	402	* devices are suspended.
	403	*/
dcce284a BH	404	extern gfp_t gfp_allowed_mask;
dcce284a BH	405
c93bdd0e MG	406	/* Returns true if the gfp_mask allows use of ALLOC_NO_WATERMARK */
	407	bool gfp_pfmemalloc_allowed(gfp_t gfp_mask);
	408
07f44ac3	409	static inline bool gfp_has_io_fs(gfp_t gfp)
f90ac398	410	{
07f44ac3	411	return (gfp & (__GFP_IO \| __GFP_FS)) == (__GFP_IO \| __GFP_FS);
f90ac398	412	}
07f44ac3	413
803de900 VB	414	/*
	415	* Check if the gfp flags allow compaction - GFP_NOIO is a really
	416	* tricky context because the migration might require IO.
	417	*/
	418	static inline bool gfp_compaction_allowed(gfp_t gfp_mask)
	419	{
	420	return IS_ENABLED(CONFIG_COMPACTION) && (gfp_mask & __GFP_IO);
	421	}
	422
07f44ac3	423	extern gfp_t vma_thp_gfp_mask(struct vm_area_struct *vma);
f90ac398	424
8df995f6	425	#ifdef CONFIG_CONTIG_ALLOC
041d3a8c	426	/* The below functions must be run on a range from a single zone. */
b951aaff	427	extern int alloc_contig_range_noprof(unsigned long start, unsigned long end,
ca96b625	428	unsigned migratetype, gfp_t gfp_mask);
b951aaff SB	429	#define alloc_contig_range(...) alloc_hooks(alloc_contig_range_noprof(__VA_ARGS__))
	430
	431	extern struct page *alloc_contig_pages_noprof(unsigned long nr_pages, gfp_t gfp_mask,
	432	int nid, nodemask_t *nodemask);
	433	#define alloc_contig_pages(...) alloc_hooks(alloc_contig_pages_noprof(__VA_ARGS__))
	434
080fe206	435	#endif
78fa5150	436	void free_contig_range(unsigned long pfn, unsigned long nr_pages);
041d3a8c	437
e98337d1 YZ	438	#ifdef CONFIG_CONTIG_ALLOC
	439	static inline struct folio *folio_alloc_gigantic_noprof(int order, gfp_t gfp,
	440	int nid, nodemask_t *node)
	441	{
	442	struct page *page;
	443
	444	if (WARN_ON(!order \|\| !(gfp & __GFP_COMP)))
	445	return NULL;
	446
	447	page = alloc_contig_pages_noprof(1 << order, gfp, nid, node);
	448
	449	return page ? page_folio(page) : NULL;
	450	}
	451	#else
	452	static inline struct folio *folio_alloc_gigantic_noprof(int order, gfp_t gfp,
	453	int nid, nodemask_t *node)
	454	{
	455	return NULL;
	456	}
	457	#endif
	458	/* This should be paired with folio_put() rather than free_contig_range(). */
	459	#define folio_alloc_gigantic(...) alloc_hooks(folio_alloc_gigantic_noprof(__VA_ARGS__))
	460
1da177e4	461	#endif /* __LINUX_GFP_H */