[librseq.git] / src / rseq-percpu-alloc.c

// SPDX-License-Identifier: MIT
// SPDX-FileCopyrightText: 2024 Mathieu Desnoyers <mathieu.desnoyers@efficios.com>

#include <rseq/percpu-alloc.h>
#include <sys/mman.h>
#include <assert.h>
#include <string.h>
#include <pthread.h>
#include <unistd.h>
#include <stdlib.h>
#include <rseq/compiler.h>
#include <errno.h>
#include <stdint.h>
#include <stdbool.h>
#include <stdio.h>

#ifdef HAVE_LIBNUMA
# include <numa.h>
# include <numaif.h>
#endif

#include "rseq-alloc-utils.h"

/*
 * rseq-percpu-alloc.c: rseq CPU-Local Storage (CLS) memory allocator.
 *
 * The rseq per-CPU memory allocator allows the application the request
 * memory pools of CPU-Local memory each of containing objects of a
 * given size (rounded to next power of 2), reserving a given virtual
 * address size per CPU, for a given maximum number of CPUs.
 *
 * The per-CPU memory allocator is analogous to TLS (Thread-Local
 * Storage) memory: TLS is Thread-Local Storage, whereas the per-CPU
 * memory allocator provides CPU-Local Storage.
 */

/*
 * Use high bits of per-CPU addresses to index the pool.
 * This leaves the low bits of available to the application for pointer
 * tagging (based on next power of 2 alignment of the allocations).
 */
#if RSEQ_BITS_PER_LONG == 64
# define POOL_INDEX_BITS	16
#else
# define POOL_INDEX_BITS	8
#endif
#define MAX_NR_POOLS		(1UL << POOL_INDEX_BITS)
#define POOL_INDEX_SHIFT	(RSEQ_BITS_PER_LONG - POOL_INDEX_BITS)
#define MAX_POOL_LEN		(1UL << POOL_INDEX_SHIFT)
#define MAX_POOL_LEN_MASK	(MAX_POOL_LEN - 1)

#define POOL_SET_NR_ENTRIES	POOL_INDEX_SHIFT

/*
 * Smallest allocation should hold enough space for a free list pointer.
 */
#if RSEQ_BITS_PER_LONG == 64
# define POOL_SET_MIN_ENTRY	3	/* Smallest item_len=8 */
#else
# define POOL_SET_MIN_ENTRY	2	/* Smallest item_len=4 */
#endif

/*
 * Skip pool index 0 to ensure allocated entries at index 0 do not match
 * a NULL pointer.
 */
#define FIRST_POOL		1

struct free_list_node;

struct free_list_node {
	struct free_list_node *next;
};

/* This lock protects pool create/destroy. */
static pthread_mutex_t pool_lock = PTHREAD_MUTEX_INITIALIZER;

struct rseq_mmap_attr {
	void *(*mmap_func)(void *priv, size_t len);
	int (*munmap_func)(void *priv, void *ptr, size_t len);
	void *mmap_priv;
};

struct rseq_percpu_pool {
	void *base;
	unsigned int index;
	size_t item_len;
	size_t percpu_len;
	int item_order;
	int max_nr_cpus;

	/*
	 * The free list chains freed items on the CPU 0 address range.
	 * We should rethink this decision if false sharing between
	 * malloc/free from other CPUs and data accesses from CPU 0
	 * becomes an issue. This is a NULL-terminated singly-linked
	 * list.
	 */
	struct free_list_node *free_list_head;
	size_t next_unused;
	/* This lock protects allocation/free within the pool. */
	pthread_mutex_t lock;

	struct rseq_mmap_attr mmap_attr;
};

//TODO: the array of pools should grow dynamically on create.
static struct rseq_percpu_pool rseq_percpu_pool[MAX_NR_POOLS];

/*
 * Pool set entries are indexed by item_len rounded to the next power of
 * 2. A pool set can contain NULL pool entries, in which case the next
 * large enough entry will be used for allocation.
 */
struct rseq_percpu_pool_set {
	/* This lock protects add vs malloc/zmalloc within the pool set. */
	pthread_mutex_t lock;
	struct rseq_percpu_pool *entries[POOL_SET_NR_ENTRIES];
};

static
void *__rseq_pool_percpu_ptr(struct rseq_percpu_pool *pool, int cpu, uintptr_t item_offset)
{
	return pool->base + (pool->percpu_len * cpu) + item_offset;
}

ptrdiff_t rseq_percpu_pool_ptr_offset(struct rseq_percpu_pool *pool, int cpu)
{
	uintptr_t rseq_percpu_base = (uintptr_t) pool->index << POOL_INDEX_SHIFT;
	uintptr_t refptr = (uintptr_t) __rseq_pool_percpu_ptr(pool, cpu, 0);

	return (ptrdiff_t) (refptr - rseq_percpu_base);
}

void *__rseq_percpu_ptr(void __rseq_percpu *_ptr, int cpu)
{
	uintptr_t ptr = (uintptr_t) _ptr;
	uintptr_t item_offset = ptr & MAX_POOL_LEN_MASK;
	uintptr_t pool_index = ptr >> POOL_INDEX_SHIFT;
	struct rseq_percpu_pool *pool = &rseq_percpu_pool[pool_index];

	assert(cpu >= 0);
	return __rseq_pool_percpu_ptr(pool, cpu, item_offset);
}

static
void rseq_percpu_zero_item(struct rseq_percpu_pool *pool, uintptr_t item_offset)
{
	int i;

	for (i = 0; i < pool->max_nr_cpus; i++) {
		char *p = __rseq_pool_percpu_ptr(pool, i, item_offset);
		memset(p, 0, pool->item_len);
	}
}

#ifdef HAVE_LIBNUMA
int rseq_percpu_pool_init_numa(struct rseq_percpu_pool *pool, int numa_flags)
{
	unsigned long nr_pages, page;
	long ret, page_len;
	int cpu;

	if (!numa_flags)
		return 0;
	page_len = rseq_get_page_len();
	nr_pages = pool->percpu_len >> rseq_get_count_order_ulong(page_len);
	for (cpu = 0; cpu < pool->max_nr_cpus; cpu++) {
		int node = numa_node_of_cpu(cpu);

		/* TODO: batch move_pages() call with an array of pages. */
		for (page = 0; page < nr_pages; page++) {
			void *pageptr = __rseq_pool_percpu_ptr(pool, cpu, page * page_len);
			int status = -EPERM;

			ret = move_pages(0, 1, &pageptr, &node, &status, numa_flags);
			if (ret)
				return ret;
		}
	}
	return 0;
}
#else
void rseq_percpu_pool_init_numa(struct rseq_percpu_pool *pool __attribute__((unused)),
		int numa_flags __attribute__((unused)))
{
	return 0;
}
#endif

static
void *default_mmap_func(void *priv __attribute__((unused)), size_t len)
{
	void *base;

	base = mmap(NULL, len, PROT_READ | PROT_WRITE,
			MAP_ANONYMOUS | MAP_PRIVATE, -1, 0);
	if (base == MAP_FAILED)
		return NULL;
	return base;
}

static
int default_munmap_func(void *priv __attribute__((unused)), void *ptr, size_t len)
{
	return munmap(ptr, len);
}

struct rseq_percpu_pool *rseq_percpu_pool_create(size_t item_len,
		size_t percpu_len, int max_nr_cpus,
		const struct rseq_mmap_attr *mmap_attr,
		int flags)
{
	void *(*mmap_func)(void *priv, size_t len);
	int (*munmap_func)(void *priv, void *ptr, size_t len);
	void *mmap_priv;
	struct rseq_percpu_pool *pool;
	void *base;
	unsigned int i;
	int order;

	if (flags) {
		errno = EINVAL;
		return NULL;
	}

	/* Make sure each item is large enough to contain free list pointers. */
	if (item_len < sizeof(void *))
		item_len = sizeof(void *);

	/* Align item_len on next power of two. */
	order = rseq_get_count_order_ulong(item_len);
	if (order < 0) {
		errno = EINVAL;
		return NULL;
	}
	item_len = 1UL << order;

	/* Align percpu_len on page size. */
	percpu_len = rseq_align(percpu_len, rseq_get_page_len());

	if (max_nr_cpus < 0 || item_len > percpu_len ||
			percpu_len > (UINTPTR_MAX >> POOL_INDEX_BITS)) {
		errno = EINVAL;
		return NULL;
	}

	if (mmap_attr) {
		mmap_func = mmap_attr->mmap_func;
		munmap_func = mmap_attr->munmap_func;
		mmap_priv = mmap_attr->mmap_priv;
	} else {
		mmap_func = default_mmap_func;
		munmap_func = default_munmap_func;
		mmap_priv = NULL;
	}
	pthread_mutex_lock(&pool_lock);
	/* Linear scan in array of pools to find empty spot. */
	for (i = FIRST_POOL; i < MAX_NR_POOLS; i++) {
		pool = &rseq_percpu_pool[i];
		if (!pool->base)
			goto found_empty;
	}
	errno = ENOMEM;
	pool = NULL;
	goto end;

found_empty:
	base = mmap_func(mmap_priv, percpu_len * max_nr_cpus);
	if (!base) {
		pool = NULL;
		goto end;
	}
	pthread_mutex_init(&pool->lock, NULL);
	pool->base = base;
	pool->percpu_len = percpu_len;
	pool->max_nr_cpus = max_nr_cpus;
	pool->index = i;
	pool->item_len = item_len;
	pool->item_order = order;
	pool->mmap_attr.mmap_func = mmap_func;
	pool->mmap_attr.munmap_func = munmap_func;
	pool->mmap_attr.mmap_priv = mmap_priv;
end:
	pthread_mutex_unlock(&pool_lock);
	return pool;
}

int rseq_percpu_pool_destroy(struct rseq_percpu_pool *pool)
{
	int ret;

	pthread_mutex_lock(&pool_lock);
	if (!pool->base) {
		errno = ENOENT;
		ret = -1;
		goto end;
	}
	ret = pool->mmap_attr.munmap_func(pool->mmap_attr.mmap_priv, pool->base,
			pool->percpu_len * pool->max_nr_cpus);
	if (ret)
		goto end;
	pthread_mutex_destroy(&pool->lock);
	memset(pool, 0, sizeof(*pool));
end:
	pthread_mutex_unlock(&pool_lock);
	return 0;
}

static
void __rseq_percpu *__rseq_percpu_malloc(struct rseq_percpu_pool *pool, bool zeroed)
{
	struct free_list_node *node;
	uintptr_t item_offset;
	void __rseq_percpu *addr;

	pthread_mutex_lock(&pool->lock);
	/* Get first entry from free list. */
	node = pool->free_list_head;
	if (node != NULL) {
		/* Remove node from free list (update head). */
		pool->free_list_head = node->next;
		item_offset = (uintptr_t) ((void *) node - pool->base);
		addr = (void *) (((uintptr_t) pool->index << POOL_INDEX_SHIFT) | item_offset);
		goto end;
	}
	if (pool->next_unused + pool->item_len > pool->percpu_len) {
		errno = ENOMEM;
		addr = NULL;
		goto end;
	}
	item_offset = pool->next_unused;
	addr = (void *) (((uintptr_t) pool->index << POOL_INDEX_SHIFT) | item_offset);
	pool->next_unused += pool->item_len;
end:
	pthread_mutex_unlock(&pool->lock);
	if (zeroed && addr)
		rseq_percpu_zero_item(pool, item_offset);
	return addr;
}

void __rseq_percpu *rseq_percpu_malloc(struct rseq_percpu_pool *pool)
{
	return __rseq_percpu_malloc(pool, false);
}

void __rseq_percpu *rseq_percpu_zmalloc(struct rseq_percpu_pool *pool)
{
	return __rseq_percpu_malloc(pool, true);
}

void rseq_percpu_free(void __rseq_percpu *_ptr)
{
	uintptr_t ptr = (uintptr_t) _ptr;
	uintptr_t item_offset = ptr & MAX_POOL_LEN_MASK;
	uintptr_t pool_index = ptr >> POOL_INDEX_SHIFT;
	struct rseq_percpu_pool *pool = &rseq_percpu_pool[pool_index];
	struct free_list_node *head, *item;

	pthread_mutex_lock(&pool->lock);
	/* Add ptr to head of free list */
	head = pool->free_list_head;
	/* Free-list is in CPU 0 range. */
	item = (struct free_list_node *)__rseq_pool_percpu_ptr(pool, 0, item_offset);
	item->next = head;
	pool->free_list_head = item;
	pthread_mutex_unlock(&pool->lock);
}

struct rseq_percpu_pool_set *rseq_percpu_pool_set_create(void)
{
	struct rseq_percpu_pool_set *pool_set;

	pool_set = calloc(1, sizeof(struct rseq_percpu_pool_set));
	if (!pool_set)
		return NULL;
	pthread_mutex_init(&pool_set->lock, NULL);
	return pool_set;
}

int rseq_percpu_pool_set_destroy(struct rseq_percpu_pool_set *pool_set)
{
	int order, ret;

	for (order = POOL_SET_MIN_ENTRY; order < POOL_SET_NR_ENTRIES; order++) {
		struct rseq_percpu_pool *pool = pool_set->entries[order];

		if (!pool)
			continue;
		ret = rseq_percpu_pool_destroy(pool);
		if (ret)
			return ret;
		pool_set->entries[order] = NULL;
	}
	pthread_mutex_destroy(&pool_set->lock);
	free(pool_set);
	return 0;
}

/* Ownership of pool is handed over to pool set on success. */
int rseq_percpu_pool_set_add_pool(struct rseq_percpu_pool_set *pool_set, struct rseq_percpu_pool *pool)
{
	size_t item_order = pool->item_order;
	int ret = 0;

	pthread_mutex_lock(&pool_set->lock);
	if (pool_set->entries[item_order]) {
		errno = EBUSY;
		ret = -1;
		goto end;
	}
	pool_set->entries[pool->item_order] = pool;
end:
	pthread_mutex_unlock(&pool_set->lock);
	return ret;
}

static
void __rseq_percpu *__rseq_percpu_pool_set_malloc(struct rseq_percpu_pool_set *pool_set, size_t len, bool zeroed)
{
	int order, min_order = POOL_SET_MIN_ENTRY;
	struct rseq_percpu_pool *pool;
	void __rseq_percpu *addr;

	order = rseq_get_count_order_ulong(len);
	if (order > POOL_SET_MIN_ENTRY)
		min_order = order;
again:
	pthread_mutex_lock(&pool_set->lock);
	/* First smallest present pool where @len fits. */
	for (order = min_order; order < POOL_SET_NR_ENTRIES; order++) {
		pool = pool_set->entries[order];

		if (!pool)
			continue;
		if (pool->item_len >= len)
			goto found;
	}
	pool = NULL;
found:
	pthread_mutex_unlock(&pool_set->lock);
	if (pool) {
		addr = __rseq_percpu_malloc(pool, zeroed);
		if (addr == NULL && errno == ENOMEM) {
			/*
			 * If the allocation failed, try again with a
			 * larger pool.
			 */
			min_order = order + 1;
			goto again;
		}
	} else {
		/* Not found. */
		errno = ENOMEM;
		addr = NULL;
	}
	return addr;
}

void __rseq_percpu *rseq_percpu_pool_set_malloc(struct rseq_percpu_pool_set *pool_set, size_t len)
{
	return __rseq_percpu_pool_set_malloc(pool_set, len, false);
}

void __rseq_percpu *rseq_percpu_pool_set_zmalloc(struct rseq_percpu_pool_set *pool_set, size_t len)
{
	return __rseq_percpu_pool_set_malloc(pool_set, len, true);
}

struct rseq_mmap_attr *rseq_mmap_attr_create(void *(*mmap_func)(void *priv, size_t len),
		int (*munmap_func)(void *priv, void *ptr, size_t len),
		void *mmap_priv)
{
	struct rseq_mmap_attr *attr = calloc(1, sizeof(struct rseq_mmap_attr));

	if (!attr)
		return NULL;
	attr->mmap_func = mmap_func;
	attr->munmap_func = munmap_func;
	attr->mmap_priv = mmap_priv;
	return attr;
}

void rseq_mmap_attr_destroy(struct rseq_mmap_attr *attr)
{
	free(attr);
}
Commit	Line	Data
ef6695f1 MD	1	// SPDX-License-Identifier: MIT
	2	// SPDX-FileCopyrightText: 2024 Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
	3
	4	#include <rseq/percpu-alloc.h>
	5	#include <sys/mman.h>
	6	#include <assert.h>
	7	#include <string.h>
	8	#include <pthread.h>
	9	#include <unistd.h>
	10	#include <stdlib.h>
	11	#include <rseq/compiler.h>
	12	#include <errno.h>
	13	#include <stdint.h>
	14	#include <stdbool.h>
367e559c MD	15	#include <stdio.h>
	16
	17	#ifdef HAVE_LIBNUMA
	18	# include <numa.h>
	19	# include <numaif.h>
	20	#endif
ef6695f1	21
19be9217 MD	22	#include "rseq-alloc-utils.h"
19be9217 MD	23
ef6695f1	24	/*
8ab16a24	25	* rseq-percpu-alloc.c: rseq CPU-Local Storage (CLS) memory allocator.
ef6695f1	26	*
8ab16a24 MD	27	* The rseq per-CPU memory allocator allows the application the request
8ab16a24 MD	28	* memory pools of CPU-Local memory each of containing objects of a
8aa1462d MD	29	* given size (rounded to next power of 2), reserving a given virtual
8aa1462d MD	30	* address size per CPU, for a given maximum number of CPUs.
8ab16a24 MD	31	*
	32	* The per-CPU memory allocator is analogous to TLS (Thread-Local
	33	* Storage) memory: TLS is Thread-Local Storage, whereas the per-CPU
	34	* memory allocator provides CPU-Local Storage.
ef6695f1 MD	35	*/
ef6695f1 MD	36
72b100a1	37	/*
8ab16a24	38	* Use high bits of per-CPU addresses to index the pool.
72b100a1 MD	39	* This leaves the low bits of available to the application for pointer
	40	* tagging (based on next power of 2 alignment of the allocations).
	41	*/
ef6695f1	42	#if RSEQ_BITS_PER_LONG == 64
72b100a1	43	# define POOL_INDEX_BITS 16
ef6695f1	44	#else
72b100a1	45	# define POOL_INDEX_BITS 8
ef6695f1	46	#endif
72b100a1 MD	47	#define MAX_NR_POOLS (1UL << POOL_INDEX_BITS)
	48	#define POOL_INDEX_SHIFT (RSEQ_BITS_PER_LONG - POOL_INDEX_BITS)
	49	#define MAX_POOL_LEN (1UL << POOL_INDEX_SHIFT)
	50	#define MAX_POOL_LEN_MASK (MAX_POOL_LEN - 1)
ef6695f1	51
72b100a1	52	#define POOL_SET_NR_ENTRIES POOL_INDEX_SHIFT
ef6695f1	53
72b100a1 MD	54	/*
	55	* Smallest allocation should hold enough space for a free list pointer.
	56	*/
ef6695f1 MD	57	#if RSEQ_BITS_PER_LONG == 64
	58	# define POOL_SET_MIN_ENTRY 3 /* Smallest item_len=8 */
	59	#else
	60	# define POOL_SET_MIN_ENTRY 2 /* Smallest item_len=4 */
	61	#endif
	62
bb1552e2 MD	63	/*
	64	* Skip pool index 0 to ensure allocated entries at index 0 do not match
	65	* a NULL pointer.
	66	*/
	67	#define FIRST_POOL 1
	68
ef6695f1 MD	69	struct free_list_node;
	70
	71	struct free_list_node {
	72	struct free_list_node *next;
	73	};
	74
	75	/* This lock protects pool create/destroy. */
	76	static pthread_mutex_t pool_lock = PTHREAD_MUTEX_INITIALIZER;
	77
9bd07c29 MD	78	struct rseq_mmap_attr {
	79	void (mmap_func)(void *priv, size_t len);
	80	int (munmap_func)(void priv, void *ptr, size_t len);
	81	void *mmap_priv;
	82	};
	83
ef6695f1 MD	84	struct rseq_percpu_pool {
	85	void *base;
	86	unsigned int index;
	87	size_t item_len;
	88	size_t percpu_len;
	89	int item_order;
	90	int max_nr_cpus;
	91
	92	/*
8ab16a24	93	* The free list chains freed items on the CPU 0 address range.
ef6695f1	94	* We should rethink this decision if false sharing between
8ab16a24	95	* malloc/free from other CPUs and data accesses from CPU 0
ef6695f1 MD	96	* becomes an issue. This is a NULL-terminated singly-linked
	97	* list.
	98	*/
	99	struct free_list_node *free_list_head;
	100	size_t next_unused;
	101	/* This lock protects allocation/free within the pool. */
	102	pthread_mutex_t lock;
9bd07c29 MD	103
9bd07c29 MD	104	struct rseq_mmap_attr mmap_attr;
ef6695f1 MD	105	};
	106
	107	//TODO: the array of pools should grow dynamically on create.
	108	static struct rseq_percpu_pool rseq_percpu_pool[MAX_NR_POOLS];
	109
	110	/*
	111	* Pool set entries are indexed by item_len rounded to the next power of
	112	* 2. A pool set can contain NULL pool entries, in which case the next
	113	* large enough entry will be used for allocation.
	114	*/
	115	struct rseq_percpu_pool_set {
	116	/* This lock protects add vs malloc/zmalloc within the pool set. */
	117	pthread_mutex_t lock;
	118	struct rseq_percpu_pool *entries[POOL_SET_NR_ENTRIES];
	119	};
	120
367e559c MD	121	static
	122	void __rseq_pool_percpu_ptr(struct rseq_percpu_pool pool, int cpu, uintptr_t item_offset)
	123	{
	124	return pool->base + (pool->percpu_len * cpu) + item_offset;
	125	}
	126
8fde749f MD	127	ptrdiff_t rseq_percpu_pool_ptr_offset(struct rseq_percpu_pool *pool, int cpu)
	128	{
	129	uintptr_t rseq_percpu_base = (uintptr_t) pool->index << POOL_INDEX_SHIFT;
	130	uintptr_t refptr = (uintptr_t) __rseq_pool_percpu_ptr(pool, cpu, 0);
	131
	132	return (ptrdiff_t) (refptr - rseq_percpu_base);
	133	}
	134
d24ee051	135	void __rseq_percpu_ptr(void __rseq_percpu _ptr, int cpu)
367e559c MD	136	{
367e559c MD	137	uintptr_t ptr = (uintptr_t) _ptr;
72b100a1 MD	138	uintptr_t item_offset = ptr & MAX_POOL_LEN_MASK;
72b100a1 MD	139	uintptr_t pool_index = ptr >> POOL_INDEX_SHIFT;
367e559c MD	140	struct rseq_percpu_pool *pool = &rseq_percpu_pool[pool_index];
	141
	142	assert(cpu >= 0);
	143	return __rseq_pool_percpu_ptr(pool, cpu, item_offset);
	144	}
	145
	146	static
	147	void rseq_percpu_zero_item(struct rseq_percpu_pool *pool, uintptr_t item_offset)
	148	{
	149	int i;
	150
	151	for (i = 0; i < pool->max_nr_cpus; i++) {
	152	char *p = __rseq_pool_percpu_ptr(pool, i, item_offset);
	153	memset(p, 0, pool->item_len);
	154	}
	155	}
	156
	157	#ifdef HAVE_LIBNUMA
9bd07c29	158	int rseq_percpu_pool_init_numa(struct rseq_percpu_pool *pool, int numa_flags)
367e559c MD	159	{
	160	unsigned long nr_pages, page;
	161	long ret, page_len;
	162	int cpu;
	163
	164	if (!numa_flags)
9bd07c29	165	return 0;
367e559c	166	page_len = rseq_get_page_len();
19be9217	167	nr_pages = pool->percpu_len >> rseq_get_count_order_ulong(page_len);
367e559c MD	168	for (cpu = 0; cpu < pool->max_nr_cpus; cpu++) {
	169	int node = numa_node_of_cpu(cpu);
	170
	171	/* TODO: batch move_pages() call with an array of pages. */
	172	for (page = 0; page < nr_pages; page++) {
	173	void pageptr = __rseq_pool_percpu_ptr(pool, cpu, page page_len);
	174	int status = -EPERM;
	175
	176	ret = move_pages(0, 1, &pageptr, &node, &status, numa_flags);
9bd07c29 MD	177	if (ret)
9bd07c29 MD	178	return ret;
367e559c MD	179	}
367e559c MD	180	}
9bd07c29	181	return 0;
367e559c MD	182	}
367e559c MD	183	#else
367e559c MD	184	void rseq_percpu_pool_init_numa(struct rseq_percpu_pool *pool __attribute__((unused)),
	185	int numa_flags __attribute__((unused)))
	186	{
9bd07c29	187	return 0;
367e559c MD	188	}
	189	#endif
	190
9bd07c29 MD	191	static
	192	void default_mmap_func(void priv __attribute__((unused)), size_t len)
	193	{
	194	void *base;
	195
	196	base = mmap(NULL, len, PROT_READ \| PROT_WRITE,
	197	MAP_ANONYMOUS \| MAP_PRIVATE, -1, 0);
	198	if (base == MAP_FAILED)
	199	return NULL;
	200	return base;
	201	}
	202
	203	static
	204	int default_munmap_func(void priv __attribute__((unused)), void ptr, size_t len)
	205	{
	206	return munmap(ptr, len);
	207	}
	208
ef6695f1 MD	209	struct rseq_percpu_pool *rseq_percpu_pool_create(size_t item_len,
ef6695f1 MD	210	size_t percpu_len, int max_nr_cpus,
6b30db4e MD	211	const struct rseq_mmap_attr *mmap_attr,
6b30db4e MD	212	int flags)
ef6695f1	213	{
9bd07c29 MD	214	void (mmap_func)(void *priv, size_t len);
	215	int (munmap_func)(void priv, void *ptr, size_t len);
	216	void *mmap_priv;
ef6695f1 MD	217	struct rseq_percpu_pool *pool;
	218	void *base;
	219	unsigned int i;
	220	int order;
ef6695f1	221
6b30db4e MD	222	if (flags) {
	223	errno = EINVAL;
	224	return NULL;
	225	}
	226
ef6695f1 MD	227	/* Make sure each item is large enough to contain free list pointers. */
	228	if (item_len < sizeof(void *))
	229	item_len = sizeof(void *);
	230
	231	/* Align item_len on next power of two. */
19be9217	232	order = rseq_get_count_order_ulong(item_len);
ef6695f1 MD	233	if (order < 0) {
	234	errno = EINVAL;
	235	return NULL;
	236	}
	237	item_len = 1UL << order;
	238
	239	/* Align percpu_len on page size. */
367e559c	240	percpu_len = rseq_align(percpu_len, rseq_get_page_len());
ef6695f1 MD	241
ef6695f1 MD	242	if (max_nr_cpus < 0 \|\| item_len > percpu_len \|\|
72b100a1	243	percpu_len > (UINTPTR_MAX >> POOL_INDEX_BITS)) {
ef6695f1 MD	244	errno = EINVAL;
	245	return NULL;
	246	}
	247
9bd07c29 MD	248	if (mmap_attr) {
	249	mmap_func = mmap_attr->mmap_func;
	250	munmap_func = mmap_attr->munmap_func;
	251	mmap_priv = mmap_attr->mmap_priv;
	252	} else {
	253	mmap_func = default_mmap_func;
	254	munmap_func = default_munmap_func;
	255	mmap_priv = NULL;
	256	}
ef6695f1 MD	257	pthread_mutex_lock(&pool_lock);
ef6695f1 MD	258	/* Linear scan in array of pools to find empty spot. */
bb1552e2	259	for (i = FIRST_POOL; i < MAX_NR_POOLS; i++) {
ef6695f1 MD	260	pool = &rseq_percpu_pool[i];
	261	if (!pool->base)
	262	goto found_empty;
	263	}
	264	errno = ENOMEM;
	265	pool = NULL;
	266	goto end;
	267
	268	found_empty:
9bd07c29 MD	269	base = mmap_func(mmap_priv, percpu_len * max_nr_cpus);
9bd07c29 MD	270	if (!base) {
ef6695f1 MD	271	pool = NULL;
	272	goto end;
	273	}
ef6695f1 MD	274	pthread_mutex_init(&pool->lock, NULL);
	275	pool->base = base;
	276	pool->percpu_len = percpu_len;
	277	pool->max_nr_cpus = max_nr_cpus;
	278	pool->index = i;
	279	pool->item_len = item_len;
	280	pool->item_order = order;
9bd07c29 MD	281	pool->mmap_attr.mmap_func = mmap_func;
	282	pool->mmap_attr.munmap_func = munmap_func;
	283	pool->mmap_attr.mmap_priv = mmap_priv;
ef6695f1 MD	284	end:
	285	pthread_mutex_unlock(&pool_lock);
	286	return pool;
	287	}
	288
	289	int rseq_percpu_pool_destroy(struct rseq_percpu_pool *pool)
	290	{
	291	int ret;
	292
	293	pthread_mutex_lock(&pool_lock);
	294	if (!pool->base) {
	295	errno = ENOENT;
	296	ret = -1;
	297	goto end;
	298	}
9bd07c29 MD	299	ret = pool->mmap_attr.munmap_func(pool->mmap_attr.mmap_priv, pool->base,
9bd07c29 MD	300	pool->percpu_len * pool->max_nr_cpus);
ef6695f1 MD	301	if (ret)
	302	goto end;
	303	pthread_mutex_destroy(&pool->lock);
	304	memset(pool, 0, sizeof(*pool));
	305	end:
	306	pthread_mutex_unlock(&pool_lock);
	307	return 0;
	308	}
	309
ef6695f1	310	static
d24ee051	311	void __rseq_percpu __rseq_percpu_malloc(struct rseq_percpu_pool pool, bool zeroed)
ef6695f1 MD	312	{
	313	struct free_list_node *node;
	314	uintptr_t item_offset;
d24ee051	315	void __rseq_percpu *addr;
ef6695f1 MD	316
	317	pthread_mutex_lock(&pool->lock);
	318	/* Get first entry from free list. */
	319	node = pool->free_list_head;
	320	if (node != NULL) {
	321	/* Remove node from free list (update head). */
	322	pool->free_list_head = node->next;
	323	item_offset = (uintptr_t) ((void *) node - pool->base);
72b100a1	324	addr = (void *) (((uintptr_t) pool->index << POOL_INDEX_SHIFT) \| item_offset);
ef6695f1 MD	325	goto end;
	326	}
	327	if (pool->next_unused + pool->item_len > pool->percpu_len) {
ea1a3ada	328	errno = ENOMEM;
ef6695f1 MD	329	addr = NULL;
	330	goto end;
	331	}
	332	item_offset = pool->next_unused;
72b100a1	333	addr = (void *) (((uintptr_t) pool->index << POOL_INDEX_SHIFT) \| item_offset);
ef6695f1 MD	334	pool->next_unused += pool->item_len;
	335	end:
	336	pthread_mutex_unlock(&pool->lock);
	337	if (zeroed && addr)
	338	rseq_percpu_zero_item(pool, item_offset);
	339	return addr;
	340	}
	341
d24ee051	342	void __rseq_percpu rseq_percpu_malloc(struct rseq_percpu_pool pool)
ef6695f1 MD	343	{
	344	return __rseq_percpu_malloc(pool, false);
	345	}
	346
d24ee051	347	void __rseq_percpu rseq_percpu_zmalloc(struct rseq_percpu_pool pool)
ef6695f1 MD	348	{
	349	return __rseq_percpu_malloc(pool, true);
	350	}
	351
d24ee051	352	void rseq_percpu_free(void __rseq_percpu *_ptr)
ef6695f1 MD	353	{
ef6695f1 MD	354	uintptr_t ptr = (uintptr_t) _ptr;
72b100a1 MD	355	uintptr_t item_offset = ptr & MAX_POOL_LEN_MASK;
72b100a1 MD	356	uintptr_t pool_index = ptr >> POOL_INDEX_SHIFT;
ef6695f1 MD	357	struct rseq_percpu_pool *pool = &rseq_percpu_pool[pool_index];
	358	struct free_list_node head, item;
	359
	360	pthread_mutex_lock(&pool->lock);
	361	/* Add ptr to head of free list */
	362	head = pool->free_list_head;
8ab16a24	363	/* Free-list is in CPU 0 range. */
ef6695f1 MD	364	item = (struct free_list_node *)__rseq_pool_percpu_ptr(pool, 0, item_offset);
	365	item->next = head;
	366	pool->free_list_head = item;
	367	pthread_mutex_unlock(&pool->lock);
	368	}
	369
	370	struct rseq_percpu_pool_set *rseq_percpu_pool_set_create(void)
	371	{
	372	struct rseq_percpu_pool_set *pool_set;
	373
	374	pool_set = calloc(1, sizeof(struct rseq_percpu_pool_set));
	375	if (!pool_set)
	376	return NULL;
	377	pthread_mutex_init(&pool_set->lock, NULL);
	378	return pool_set;
	379	}
	380
	381	int rseq_percpu_pool_set_destroy(struct rseq_percpu_pool_set *pool_set)
	382	{
	383	int order, ret;
	384
	385	for (order = POOL_SET_MIN_ENTRY; order < POOL_SET_NR_ENTRIES; order++) {
	386	struct rseq_percpu_pool *pool = pool_set->entries[order];
	387
	388	if (!pool)
	389	continue;
	390	ret = rseq_percpu_pool_destroy(pool);
	391	if (ret)
	392	return ret;
	393	pool_set->entries[order] = NULL;
	394	}
	395	pthread_mutex_destroy(&pool_set->lock);
	396	free(pool_set);
	397	return 0;
	398	}
	399
	400	/* Ownership of pool is handed over to pool set on success. */
	401	int rseq_percpu_pool_set_add_pool(struct rseq_percpu_pool_set pool_set, struct rseq_percpu_pool pool)
	402	{
	403	size_t item_order = pool->item_order;
	404	int ret = 0;
	405
	406	pthread_mutex_lock(&pool_set->lock);
	407	if (pool_set->entries[item_order]) {
	408	errno = EBUSY;
	409	ret = -1;
	410	goto end;
	411	}
	412	pool_set->entries[pool->item_order] = pool;
	413	end:
	414	pthread_mutex_unlock(&pool_set->lock);
	415	return ret;
	416	}
	417
	418	static
d24ee051	419	void __rseq_percpu __rseq_percpu_pool_set_malloc(struct rseq_percpu_pool_set pool_set, size_t len, bool zeroed)
ef6695f1 MD	420	{
	421	int order, min_order = POOL_SET_MIN_ENTRY;
	422	struct rseq_percpu_pool *pool;
d24ee051	423	void __rseq_percpu *addr;
ef6695f1	424
d06f5cf5 MD	425	order = rseq_get_count_order_ulong(len);
	426	if (order > POOL_SET_MIN_ENTRY)
	427	min_order = order;
ef6695f1 MD	428	again:
	429	pthread_mutex_lock(&pool_set->lock);
	430	/* First smallest present pool where @len fits. */
	431	for (order = min_order; order < POOL_SET_NR_ENTRIES; order++) {
	432	pool = pool_set->entries[order];
	433
	434	if (!pool)
	435	continue;
	436	if (pool->item_len >= len)
	437	goto found;
	438	}
	439	pool = NULL;
	440	found:
	441	pthread_mutex_unlock(&pool_set->lock);
	442	if (pool) {
	443	addr = __rseq_percpu_malloc(pool, zeroed);
	444	if (addr == NULL && errno == ENOMEM) {
	445	/*
	446	* If the allocation failed, try again with a
	447	* larger pool.
	448	*/
	449	min_order = order + 1;
	450	goto again;
	451	}
	452	} else {
	453	/* Not found. */
	454	errno = ENOMEM;
	455	addr = NULL;
	456	}
	457	return addr;
	458	}
	459
d24ee051	460	void __rseq_percpu rseq_percpu_pool_set_malloc(struct rseq_percpu_pool_set pool_set, size_t len)
ef6695f1 MD	461	{
	462	return __rseq_percpu_pool_set_malloc(pool_set, len, false);
	463	}
	464
d24ee051	465	void __rseq_percpu rseq_percpu_pool_set_zmalloc(struct rseq_percpu_pool_set pool_set, size_t len)
ef6695f1 MD	466	{
	467	return __rseq_percpu_pool_set_malloc(pool_set, len, true);
	468	}
9bd07c29 MD	469
	470	struct rseq_mmap_attr rseq_mmap_attr_create(void (mmap_func)(void priv, size_t len),
	471	int (munmap_func)(void priv, void *ptr, size_t len),
	472	void *mmap_priv)
	473	{
	474	struct rseq_mmap_attr *attr = calloc(1, sizeof(struct rseq_mmap_attr));
	475
	476	if (!attr)
	477	return NULL;
	478	attr->mmap_func = mmap_func;
	479	attr->munmap_func = munmap_func;
	480	attr->mmap_priv = mmap_priv;
	481	return attr;
	482	}
	483
	484	void rseq_mmap_attr_destroy(struct rseq_mmap_attr *attr)
	485	{
	486	free(attr);
	487	}