kernel/bpf/queue_stack_maps.c

   1 // SPDX-License-Identifier: GPL-2.0
   2 /*
   3  * queue_stack_maps.c: BPF queue and stack maps
   4  *
   5  * Copyright (c) 2018 Politecnico di Torino
   6  */
   7 #include <linux/bpf.h>
   8 #include <linux/list.h>
   9 #include <linux/slab.h>
  10 #include <linux/capability.h>
  11 #include "percpu_freelist.h"
  12
  13 #define QUEUE_STACK_CREATE_FLAG_MASK \
  14         (BPF_F_NUMA_NODE | BPF_F_ACCESS_MASK)
  15
  16 struct bpf_queue_stack {
  17         struct bpf_map map;
  18         raw_spinlock_t lock;
  19         u32 head, tail;
  20         u32 size; /* max_entries + 1 */
  21
  22         char elements[] __aligned(8);
  23 };
  24
  25 static struct bpf_queue_stack *bpf_queue_stack(struct bpf_map *map)
  26 {
  27         return container_of(map, struct bpf_queue_stack, map);
  28 }
  29
  30 static bool queue_stack_map_is_empty(struct bpf_queue_stack *qs)
  31 {
  32         return qs->head == qs->tail;
  33 }
  34
  35 static bool queue_stack_map_is_full(struct bpf_queue_stack *qs)
  36 {
  37         u32 head = qs->head + 1;
  38
  39         if (unlikely(head >= qs->size))
  40                 head = 0;
  41
  42         return head == qs->tail;
  43 }
  44
  45 /* Called from syscall */
  46 static int queue_stack_map_alloc_check(union bpf_attr *attr)
  47 {
  48         if (!bpf_capable())
  49                 return -EPERM;
  50
  51         /* check sanity of attributes */
  52         if (attr->max_entries == 0 || attr->key_size != 0 ||
  53             attr->value_size == 0 ||
  54             attr->map_flags & ~QUEUE_STACK_CREATE_FLAG_MASK ||
  55             !bpf_map_flags_access_ok(attr->map_flags))
  56                 return -EINVAL;
  57
  58         if (attr->value_size > KMALLOC_MAX_SIZE)
  59                 /* if value_size is bigger, the user space won't be able to
  60                  * access the elements.
  61                  */
  62                 return -E2BIG;
  63
  64         return 0;
  65 }
  66
  67 static struct bpf_map *queue_stack_map_alloc(union bpf_attr *attr)
  68 {
  69         int numa_node = bpf_map_attr_numa_node(attr);
  70         struct bpf_queue_stack *qs;
  71         u64 size, queue_size;
  72
  73         size = (u64) attr->max_entries + 1;
  74         queue_size = sizeof(*qs) + size * attr->value_size;
  75
  76         qs = bpf_map_area_alloc(queue_size, numa_node);
  77         if (!qs)
  78                 return ERR_PTR(-ENOMEM);
  79
  80         memset(qs, 0, sizeof(*qs));
  81
  82         bpf_map_init_from_attr(&qs->map, attr);
  83
  84         qs->size = size;
  85
  86         raw_spin_lock_init(&qs->lock);
  87
  88         return &qs->map;
  89 }
  90
  91 /* Called when map->refcnt goes to zero, either from workqueue or from syscall */
  92 static void queue_stack_map_free(struct bpf_map *map)
  93 {
  94         struct bpf_queue_stack *qs = bpf_queue_stack(map);
  95
  96         bpf_map_area_free(qs);
  97 }
  98
  99 static int __queue_map_get(struct bpf_map *map, void *value, bool delete)
 100 {
 101         struct bpf_queue_stack *qs = bpf_queue_stack(map);
 102         unsigned long flags;
 103         int err = 0;
 104         void *ptr;
 105
 106         raw_spin_lock_irqsave(&qs->lock, flags);
 107
 108         if (queue_stack_map_is_empty(qs)) {
 109                 memset(value, 0, qs->map.value_size);
 110                 err = -ENOENT;
 111                 goto out;
 112         }
 113
 114         ptr = &qs->elements[qs->tail * qs->map.value_size];
 115         memcpy(value, ptr, qs->map.value_size);
 116
 117         if (delete) {
 118                 if (unlikely(++qs->tail >= qs->size))
 119                         qs->tail = 0;
 120         }
 121
 122 out:
 123         raw_spin_unlock_irqrestore(&qs->lock, flags);
 124         return err;
 125 }
 126
 127
 128 static int __stack_map_get(struct bpf_map *map, void *value, bool delete)
 129 {
 130         struct bpf_queue_stack *qs = bpf_queue_stack(map);
 131         unsigned long flags;
 132         int err = 0;
 133         void *ptr;
 134         u32 index;
 135
 136         raw_spin_lock_irqsave(&qs->lock, flags);
 137
 138         if (queue_stack_map_is_empty(qs)) {
 139                 memset(value, 0, qs->map.value_size);
 140                 err = -ENOENT;
 141                 goto out;
 142         }
 143
 144         index = qs->head - 1;
 145         if (unlikely(index >= qs->size))
 146                 index = qs->size - 1;
 147
 148         ptr = &qs->elements[index * qs->map.value_size];
 149         memcpy(value, ptr, qs->map.value_size);
 150
 151         if (delete)
 152                 qs->head = index;
 153
 154 out:
 155         raw_spin_unlock_irqrestore(&qs->lock, flags);
 156         return err;
 157 }
 158
 159 /* Called from syscall or from eBPF program */
 160 static int queue_map_peek_elem(struct bpf_map *map, void *value)
 161 {
 162         return __queue_map_get(map, value, false);
 163 }
 164
 165 /* Called from syscall or from eBPF program */
 166 static int stack_map_peek_elem(struct bpf_map *map, void *value)
 167 {
 168         return __stack_map_get(map, value, false);
 169 }
 170
 171 /* Called from syscall or from eBPF program */
 172 static int queue_map_pop_elem(struct bpf_map *map, void *value)
 173 {
 174         return __queue_map_get(map, value, true);
 175 }
 176
 177 /* Called from syscall or from eBPF program */
 178 static int stack_map_pop_elem(struct bpf_map *map, void *value)
 179 {
 180         return __stack_map_get(map, value, true);
 181 }
 182
 183 /* Called from syscall or from eBPF program */
 184 static int queue_stack_map_push_elem(struct bpf_map *map, void *value,
 185                                      u64 flags)
 186 {
 187         struct bpf_queue_stack *qs = bpf_queue_stack(map);
 188         unsigned long irq_flags;
 189         int err = 0;
 190         void *dst;
 191
 192         /* BPF_EXIST is used to force making room for a new element in case the
 193          * map is full
 194          */
 195         bool replace = (flags & BPF_EXIST);
 196
 197         /* Check supported flags for queue and stack maps */
 198         if (flags & BPF_NOEXIST || flags > BPF_EXIST)
 199                 return -EINVAL;
 200
 201         raw_spin_lock_irqsave(&qs->lock, irq_flags);
 202
 203         if (queue_stack_map_is_full(qs)) {
 204                 if (!replace) {
 205                         err = -E2BIG;
 206                         goto out;
 207                 }
 208                 /* advance tail pointer to overwrite oldest element */
 209                 if (unlikely(++qs->tail >= qs->size))
 210                         qs->tail = 0;
 211         }
 212
 213         dst = &qs->elements[qs->head * qs->map.value_size];
 214         memcpy(dst, value, qs->map.value_size);
 215
 216         if (unlikely(++qs->head >= qs->size))
 217                 qs->head = 0;
 218
 219 out:
 220         raw_spin_unlock_irqrestore(&qs->lock, irq_flags);
 221         return err;
 222 }
 223
 224 /* Called from syscall or from eBPF program */
 225 static void *queue_stack_map_lookup_elem(struct bpf_map *map, void *key)
 226 {
 227         return NULL;
 228 }
 229
 230 /* Called from syscall or from eBPF program */
 231 static int queue_stack_map_update_elem(struct bpf_map *map, void *key,
 232                                        void *value, u64 flags)
 233 {
 234         return -EINVAL;
 235 }
 236
 237 /* Called from syscall or from eBPF program */
 238 static int queue_stack_map_delete_elem(struct bpf_map *map, void *key)
 239 {
 240         return -EINVAL;
 241 }
 242
 243 /* Called from syscall */
 244 static int queue_stack_map_get_next_key(struct bpf_map *map, void *key,
 245                                         void *next_key)
 246 {
 247         return -EINVAL;
 248 }
 249
 250 static int queue_map_btf_id;
 251 const struct bpf_map_ops queue_map_ops = {
 252         .map_meta_equal = bpf_map_meta_equal,
 253         .map_alloc_check = queue_stack_map_alloc_check,
 254         .map_alloc = queue_stack_map_alloc,
 255         .map_free = queue_stack_map_free,
 256         .map_lookup_elem = queue_stack_map_lookup_elem,
 257         .map_update_elem = queue_stack_map_update_elem,
 258         .map_delete_elem = queue_stack_map_delete_elem,
 259         .map_push_elem = queue_stack_map_push_elem,
 260         .map_pop_elem = queue_map_pop_elem,
 261         .map_peek_elem = queue_map_peek_elem,
 262         .map_get_next_key = queue_stack_map_get_next_key,
 263         .map_btf_name = "bpf_queue_stack",
 264         .map_btf_id = &queue_map_btf_id,
 265 };
 266
 267 static int stack_map_btf_id;
 268 const struct bpf_map_ops stack_map_ops = {
 269         .map_meta_equal = bpf_map_meta_equal,
 270         .map_alloc_check = queue_stack_map_alloc_check,
 271         .map_alloc = queue_stack_map_alloc,
 272         .map_free = queue_stack_map_free,
 273         .map_lookup_elem = queue_stack_map_lookup_elem,
 274         .map_update_elem = queue_stack_map_update_elem,
 275         .map_delete_elem = queue_stack_map_delete_elem,
 276         .map_push_elem = queue_stack_map_push_elem,
 277         .map_pop_elem = stack_map_pop_elem,
 278         .map_peek_elem = stack_map_peek_elem,
 279         .map_get_next_key = queue_stack_map_get_next_key,
 280         .map_btf_name = "bpf_queue_stack",
 281         .map_btf_id = &stack_map_btf_id,
 282 };