linux/kernel/bpf/lpm_trie.c
Eric Dumazet 8d75839b84 bpf, lpm: make longest_prefix_match() faster
At LPC 2018 in Vancouver, Vlad Dumitrescu mentioned that longest_prefix_match()
has a high cost [1].

One reason for that cost is a loop handling one byte at a time.

We can handle more bytes at a time, if enough attention is paid
to endianness.

I was able to remove ~55 % of longest_prefix_match() cpu costs.

[1] https://linuxplumbersconf.org/event/2/contributions/88/attachments/76/87/lpc-bpf-2018-shaping.pdf

Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Vlad Dumitrescu <vladum@google.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Daniel Borkmann <daniel@iogearbox.net>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
2018-11-22 10:35:13 +01:00

748 lines
21 KiB
C

/*
* Longest prefix match list implementation
*
* Copyright (c) 2016,2017 Daniel Mack
* Copyright (c) 2016 David Herrmann
*
* This file is subject to the terms and conditions of version 2 of the GNU
* General Public License. See the file COPYING in the main directory of the
* Linux distribution for more details.
*/
#include <linux/bpf.h>
#include <linux/btf.h>
#include <linux/err.h>
#include <linux/slab.h>
#include <linux/spinlock.h>
#include <linux/vmalloc.h>
#include <net/ipv6.h>
#include <uapi/linux/btf.h>
/* Intermediate node */
#define LPM_TREE_NODE_FLAG_IM BIT(0)
struct lpm_trie_node;
struct lpm_trie_node {
struct rcu_head rcu;
struct lpm_trie_node __rcu *child[2];
u32 prefixlen;
u32 flags;
u8 data[0];
};
struct lpm_trie {
struct bpf_map map;
struct lpm_trie_node __rcu *root;
size_t n_entries;
size_t max_prefixlen;
size_t data_size;
raw_spinlock_t lock;
};
/* This trie implements a longest prefix match algorithm that can be used to
* match IP addresses to a stored set of ranges.
*
* Data stored in @data of struct bpf_lpm_key and struct lpm_trie_node is
* interpreted as big endian, so data[0] stores the most significant byte.
*
* Match ranges are internally stored in instances of struct lpm_trie_node
* which each contain their prefix length as well as two pointers that may
* lead to more nodes containing more specific matches. Each node also stores
* a value that is defined by and returned to userspace via the update_elem
* and lookup functions.
*
* For instance, let's start with a trie that was created with a prefix length
* of 32, so it can be used for IPv4 addresses, and one single element that
* matches 192.168.0.0/16. The data array would hence contain
* [0xc0, 0xa8, 0x00, 0x00] in big-endian notation. This documentation will
* stick to IP-address notation for readability though.
*
* As the trie is empty initially, the new node (1) will be places as root
* node, denoted as (R) in the example below. As there are no other node, both
* child pointers are %NULL.
*
* +----------------+
* | (1) (R) |
* | 192.168.0.0/16 |
* | value: 1 |
* | [0] [1] |
* +----------------+
*
* Next, let's add a new node (2) matching 192.168.0.0/24. As there is already
* a node with the same data and a smaller prefix (ie, a less specific one),
* node (2) will become a child of (1). In child index depends on the next bit
* that is outside of what (1) matches, and that bit is 0, so (2) will be
* child[0] of (1):
*
* +----------------+
* | (1) (R) |
* | 192.168.0.0/16 |
* | value: 1 |
* | [0] [1] |
* +----------------+
* |
* +----------------+
* | (2) |
* | 192.168.0.0/24 |
* | value: 2 |
* | [0] [1] |
* +----------------+
*
* The child[1] slot of (1) could be filled with another node which has bit #17
* (the next bit after the ones that (1) matches on) set to 1. For instance,
* 192.168.128.0/24:
*
* +----------------+
* | (1) (R) |
* | 192.168.0.0/16 |
* | value: 1 |
* | [0] [1] |
* +----------------+
* | |
* +----------------+ +------------------+
* | (2) | | (3) |
* | 192.168.0.0/24 | | 192.168.128.0/24 |
* | value: 2 | | value: 3 |
* | [0] [1] | | [0] [1] |
* +----------------+ +------------------+
*
* Let's add another node (4) to the game for 192.168.1.0/24. In order to place
* it, node (1) is looked at first, and because (4) of the semantics laid out
* above (bit #17 is 0), it would normally be attached to (1) as child[0].
* However, that slot is already allocated, so a new node is needed in between.
* That node does not have a value attached to it and it will never be
* returned to users as result of a lookup. It is only there to differentiate
* the traversal further. It will get a prefix as wide as necessary to
* distinguish its two children:
*
* +----------------+
* | (1) (R) |
* | 192.168.0.0/16 |
* | value: 1 |
* | [0] [1] |
* +----------------+
* | |
* +----------------+ +------------------+
* | (4) (I) | | (3) |
* | 192.168.0.0/23 | | 192.168.128.0/24 |
* | value: --- | | value: 3 |
* | [0] [1] | | [0] [1] |
* +----------------+ +------------------+
* | |
* +----------------+ +----------------+
* | (2) | | (5) |
* | 192.168.0.0/24 | | 192.168.1.0/24 |
* | value: 2 | | value: 5 |
* | [0] [1] | | [0] [1] |
* +----------------+ +----------------+
*
* 192.168.1.1/32 would be a child of (5) etc.
*
* An intermediate node will be turned into a 'real' node on demand. In the
* example above, (4) would be re-used if 192.168.0.0/23 is added to the trie.
*
* A fully populated trie would have a height of 32 nodes, as the trie was
* created with a prefix length of 32.
*
* The lookup starts at the root node. If the current node matches and if there
* is a child that can be used to become more specific, the trie is traversed
* downwards. The last node in the traversal that is a non-intermediate one is
* returned.
*/
static inline int extract_bit(const u8 *data, size_t index)
{
return !!(data[index / 8] & (1 << (7 - (index % 8))));
}
/**
* longest_prefix_match() - determine the longest prefix
* @trie: The trie to get internal sizes from
* @node: The node to operate on
* @key: The key to compare to @node
*
* Determine the longest prefix of @node that matches the bits in @key.
*/
static size_t longest_prefix_match(const struct lpm_trie *trie,
const struct lpm_trie_node *node,
const struct bpf_lpm_trie_key *key)
{
u32 limit = min(node->prefixlen, key->prefixlen);
u32 prefixlen = 0, i = 0;
BUILD_BUG_ON(offsetof(struct lpm_trie_node, data) % sizeof(u32));
BUILD_BUG_ON(offsetof(struct bpf_lpm_trie_key, data) % sizeof(u32));
#if defined(CONFIG_HAVE_EFFICIENT_UNALIGNED_ACCESS) && defined(CONFIG_64BIT)
/* data_size >= 16 has very small probability.
* We do not use a loop for optimal code generation.
*/
if (trie->data_size >= 8) {
u64 diff = be64_to_cpu(*(__be64 *)node->data ^
*(__be64 *)key->data);
prefixlen = 64 - fls64(diff);
if (prefixlen >= limit)
return limit;
if (diff)
return prefixlen;
i = 8;
}
#endif
while (trie->data_size >= i + 4) {
u32 diff = be32_to_cpu(*(__be32 *)&node->data[i] ^
*(__be32 *)&key->data[i]);
prefixlen += 32 - fls(diff);
if (prefixlen >= limit)
return limit;
if (diff)
return prefixlen;
i += 4;
}
if (trie->data_size >= i + 2) {
u16 diff = be16_to_cpu(*(__be16 *)&node->data[i] ^
*(__be16 *)&key->data[i]);
prefixlen += 16 - fls(diff);
if (prefixlen >= limit)
return limit;
if (diff)
return prefixlen;
i += 2;
}
if (trie->data_size >= i + 1) {
prefixlen += 8 - fls(node->data[i] ^ key->data[i]);
if (prefixlen >= limit)
return limit;
}
return prefixlen;
}
/* Called from syscall or from eBPF program */
static void *trie_lookup_elem(struct bpf_map *map, void *_key)
{
struct lpm_trie *trie = container_of(map, struct lpm_trie, map);
struct lpm_trie_node *node, *found = NULL;
struct bpf_lpm_trie_key *key = _key;
/* Start walking the trie from the root node ... */
for (node = rcu_dereference(trie->root); node;) {
unsigned int next_bit;
size_t matchlen;
/* Determine the longest prefix of @node that matches @key.
* If it's the maximum possible prefix for this trie, we have
* an exact match and can return it directly.
*/
matchlen = longest_prefix_match(trie, node, key);
if (matchlen == trie->max_prefixlen) {
found = node;
break;
}
/* If the number of bits that match is smaller than the prefix
* length of @node, bail out and return the node we have seen
* last in the traversal (ie, the parent).
*/
if (matchlen < node->prefixlen)
break;
/* Consider this node as return candidate unless it is an
* artificially added intermediate one.
*/
if (!(node->flags & LPM_TREE_NODE_FLAG_IM))
found = node;
/* If the node match is fully satisfied, let's see if we can
* become more specific. Determine the next bit in the key and
* traverse down.
*/
next_bit = extract_bit(key->data, node->prefixlen);
node = rcu_dereference(node->child[next_bit]);
}
if (!found)
return NULL;
return found->data + trie->data_size;
}
static struct lpm_trie_node *lpm_trie_node_alloc(const struct lpm_trie *trie,
const void *value)
{
struct lpm_trie_node *node;
size_t size = sizeof(struct lpm_trie_node) + trie->data_size;
if (value)
size += trie->map.value_size;
node = kmalloc_node(size, GFP_ATOMIC | __GFP_NOWARN,
trie->map.numa_node);
if (!node)
return NULL;
node->flags = 0;
if (value)
memcpy(node->data + trie->data_size, value,
trie->map.value_size);
return node;
}
/* Called from syscall or from eBPF program */
static int trie_update_elem(struct bpf_map *map,
void *_key, void *value, u64 flags)
{
struct lpm_trie *trie = container_of(map, struct lpm_trie, map);
struct lpm_trie_node *node, *im_node = NULL, *new_node = NULL;
struct lpm_trie_node __rcu **slot;
struct bpf_lpm_trie_key *key = _key;
unsigned long irq_flags;
unsigned int next_bit;
size_t matchlen = 0;
int ret = 0;
if (unlikely(flags > BPF_EXIST))
return -EINVAL;
if (key->prefixlen > trie->max_prefixlen)
return -EINVAL;
raw_spin_lock_irqsave(&trie->lock, irq_flags);
/* Allocate and fill a new node */
if (trie->n_entries == trie->map.max_entries) {
ret = -ENOSPC;
goto out;
}
new_node = lpm_trie_node_alloc(trie, value);
if (!new_node) {
ret = -ENOMEM;
goto out;
}
trie->n_entries++;
new_node->prefixlen = key->prefixlen;
RCU_INIT_POINTER(new_node->child[0], NULL);
RCU_INIT_POINTER(new_node->child[1], NULL);
memcpy(new_node->data, key->data, trie->data_size);
/* Now find a slot to attach the new node. To do that, walk the tree
* from the root and match as many bits as possible for each node until
* we either find an empty slot or a slot that needs to be replaced by
* an intermediate node.
*/
slot = &trie->root;
while ((node = rcu_dereference_protected(*slot,
lockdep_is_held(&trie->lock)))) {
matchlen = longest_prefix_match(trie, node, key);
if (node->prefixlen != matchlen ||
node->prefixlen == key->prefixlen ||
node->prefixlen == trie->max_prefixlen)
break;
next_bit = extract_bit(key->data, node->prefixlen);
slot = &node->child[next_bit];
}
/* If the slot is empty (a free child pointer or an empty root),
* simply assign the @new_node to that slot and be done.
*/
if (!node) {
rcu_assign_pointer(*slot, new_node);
goto out;
}
/* If the slot we picked already exists, replace it with @new_node
* which already has the correct data array set.
*/
if (node->prefixlen == matchlen) {
new_node->child[0] = node->child[0];
new_node->child[1] = node->child[1];
if (!(node->flags & LPM_TREE_NODE_FLAG_IM))
trie->n_entries--;
rcu_assign_pointer(*slot, new_node);
kfree_rcu(node, rcu);
goto out;
}
/* If the new node matches the prefix completely, it must be inserted
* as an ancestor. Simply insert it between @node and *@slot.
*/
if (matchlen == key->prefixlen) {
next_bit = extract_bit(node->data, matchlen);
rcu_assign_pointer(new_node->child[next_bit], node);
rcu_assign_pointer(*slot, new_node);
goto out;
}
im_node = lpm_trie_node_alloc(trie, NULL);
if (!im_node) {
ret = -ENOMEM;
goto out;
}
im_node->prefixlen = matchlen;
im_node->flags |= LPM_TREE_NODE_FLAG_IM;
memcpy(im_node->data, node->data, trie->data_size);
/* Now determine which child to install in which slot */
if (extract_bit(key->data, matchlen)) {
rcu_assign_pointer(im_node->child[0], node);
rcu_assign_pointer(im_node->child[1], new_node);
} else {
rcu_assign_pointer(im_node->child[0], new_node);
rcu_assign_pointer(im_node->child[1], node);
}
/* Finally, assign the intermediate node to the determined spot */
rcu_assign_pointer(*slot, im_node);
out:
if (ret) {
if (new_node)
trie->n_entries--;
kfree(new_node);
kfree(im_node);
}
raw_spin_unlock_irqrestore(&trie->lock, irq_flags);
return ret;
}
/* Called from syscall or from eBPF program */
static int trie_delete_elem(struct bpf_map *map, void *_key)
{
struct lpm_trie *trie = container_of(map, struct lpm_trie, map);
struct bpf_lpm_trie_key *key = _key;
struct lpm_trie_node __rcu **trim, **trim2;
struct lpm_trie_node *node, *parent;
unsigned long irq_flags;
unsigned int next_bit;
size_t matchlen = 0;
int ret = 0;
if (key->prefixlen > trie->max_prefixlen)
return -EINVAL;
raw_spin_lock_irqsave(&trie->lock, irq_flags);
/* Walk the tree looking for an exact key/length match and keeping
* track of the path we traverse. We will need to know the node
* we wish to delete, and the slot that points to the node we want
* to delete. We may also need to know the nodes parent and the
* slot that contains it.
*/
trim = &trie->root;
trim2 = trim;
parent = NULL;
while ((node = rcu_dereference_protected(
*trim, lockdep_is_held(&trie->lock)))) {
matchlen = longest_prefix_match(trie, node, key);
if (node->prefixlen != matchlen ||
node->prefixlen == key->prefixlen)
break;
parent = node;
trim2 = trim;
next_bit = extract_bit(key->data, node->prefixlen);
trim = &node->child[next_bit];
}
if (!node || node->prefixlen != key->prefixlen ||
(node->flags & LPM_TREE_NODE_FLAG_IM)) {
ret = -ENOENT;
goto out;
}
trie->n_entries--;
/* If the node we are removing has two children, simply mark it
* as intermediate and we are done.
*/
if (rcu_access_pointer(node->child[0]) &&
rcu_access_pointer(node->child[1])) {
node->flags |= LPM_TREE_NODE_FLAG_IM;
goto out;
}
/* If the parent of the node we are about to delete is an intermediate
* node, and the deleted node doesn't have any children, we can delete
* the intermediate parent as well and promote its other child
* up the tree. Doing this maintains the invariant that all
* intermediate nodes have exactly 2 children and that there are no
* unnecessary intermediate nodes in the tree.
*/
if (parent && (parent->flags & LPM_TREE_NODE_FLAG_IM) &&
!node->child[0] && !node->child[1]) {
if (node == rcu_access_pointer(parent->child[0]))
rcu_assign_pointer(
*trim2, rcu_access_pointer(parent->child[1]));
else
rcu_assign_pointer(
*trim2, rcu_access_pointer(parent->child[0]));
kfree_rcu(parent, rcu);
kfree_rcu(node, rcu);
goto out;
}
/* The node we are removing has either zero or one child. If there
* is a child, move it into the removed node's slot then delete
* the node. Otherwise just clear the slot and delete the node.
*/
if (node->child[0])
rcu_assign_pointer(*trim, rcu_access_pointer(node->child[0]));
else if (node->child[1])
rcu_assign_pointer(*trim, rcu_access_pointer(node->child[1]));
else
RCU_INIT_POINTER(*trim, NULL);
kfree_rcu(node, rcu);
out:
raw_spin_unlock_irqrestore(&trie->lock, irq_flags);
return ret;
}
#define LPM_DATA_SIZE_MAX 256
#define LPM_DATA_SIZE_MIN 1
#define LPM_VAL_SIZE_MAX (KMALLOC_MAX_SIZE - LPM_DATA_SIZE_MAX - \
sizeof(struct lpm_trie_node))
#define LPM_VAL_SIZE_MIN 1
#define LPM_KEY_SIZE(X) (sizeof(struct bpf_lpm_trie_key) + (X))
#define LPM_KEY_SIZE_MAX LPM_KEY_SIZE(LPM_DATA_SIZE_MAX)
#define LPM_KEY_SIZE_MIN LPM_KEY_SIZE(LPM_DATA_SIZE_MIN)
#define LPM_CREATE_FLAG_MASK (BPF_F_NO_PREALLOC | BPF_F_NUMA_NODE | \
BPF_F_RDONLY | BPF_F_WRONLY)
static struct bpf_map *trie_alloc(union bpf_attr *attr)
{
struct lpm_trie *trie;
u64 cost = sizeof(*trie), cost_per_node;
int ret;
if (!capable(CAP_SYS_ADMIN))
return ERR_PTR(-EPERM);
/* check sanity of attributes */
if (attr->max_entries == 0 ||
!(attr->map_flags & BPF_F_NO_PREALLOC) ||
attr->map_flags & ~LPM_CREATE_FLAG_MASK ||
attr->key_size < LPM_KEY_SIZE_MIN ||
attr->key_size > LPM_KEY_SIZE_MAX ||
attr->value_size < LPM_VAL_SIZE_MIN ||
attr->value_size > LPM_VAL_SIZE_MAX)
return ERR_PTR(-EINVAL);
trie = kzalloc(sizeof(*trie), GFP_USER | __GFP_NOWARN);
if (!trie)
return ERR_PTR(-ENOMEM);
/* copy mandatory map attributes */
bpf_map_init_from_attr(&trie->map, attr);
trie->data_size = attr->key_size -
offsetof(struct bpf_lpm_trie_key, data);
trie->max_prefixlen = trie->data_size * 8;
cost_per_node = sizeof(struct lpm_trie_node) +
attr->value_size + trie->data_size;
cost += (u64) attr->max_entries * cost_per_node;
if (cost >= U32_MAX - PAGE_SIZE) {
ret = -E2BIG;
goto out_err;
}
trie->map.pages = round_up(cost, PAGE_SIZE) >> PAGE_SHIFT;
ret = bpf_map_precharge_memlock(trie->map.pages);
if (ret)
goto out_err;
raw_spin_lock_init(&trie->lock);
return &trie->map;
out_err:
kfree(trie);
return ERR_PTR(ret);
}
static void trie_free(struct bpf_map *map)
{
struct lpm_trie *trie = container_of(map, struct lpm_trie, map);
struct lpm_trie_node __rcu **slot;
struct lpm_trie_node *node;
/* Wait for outstanding programs to complete
* update/lookup/delete/get_next_key and free the trie.
*/
synchronize_rcu();
/* Always start at the root and walk down to a node that has no
* children. Then free that node, nullify its reference in the parent
* and start over.
*/
for (;;) {
slot = &trie->root;
for (;;) {
node = rcu_dereference_protected(*slot, 1);
if (!node)
goto out;
if (rcu_access_pointer(node->child[0])) {
slot = &node->child[0];
continue;
}
if (rcu_access_pointer(node->child[1])) {
slot = &node->child[1];
continue;
}
kfree(node);
RCU_INIT_POINTER(*slot, NULL);
break;
}
}
out:
kfree(trie);
}
static int trie_get_next_key(struct bpf_map *map, void *_key, void *_next_key)
{
struct lpm_trie_node *node, *next_node = NULL, *parent, *search_root;
struct lpm_trie *trie = container_of(map, struct lpm_trie, map);
struct bpf_lpm_trie_key *key = _key, *next_key = _next_key;
struct lpm_trie_node **node_stack = NULL;
int err = 0, stack_ptr = -1;
unsigned int next_bit;
size_t matchlen;
/* The get_next_key follows postorder. For the 4 node example in
* the top of this file, the trie_get_next_key() returns the following
* one after another:
* 192.168.0.0/24
* 192.168.1.0/24
* 192.168.128.0/24
* 192.168.0.0/16
*
* The idea is to return more specific keys before less specific ones.
*/
/* Empty trie */
search_root = rcu_dereference(trie->root);
if (!search_root)
return -ENOENT;
/* For invalid key, find the leftmost node in the trie */
if (!key || key->prefixlen > trie->max_prefixlen)
goto find_leftmost;
node_stack = kmalloc_array(trie->max_prefixlen,
sizeof(struct lpm_trie_node *),
GFP_ATOMIC | __GFP_NOWARN);
if (!node_stack)
return -ENOMEM;
/* Try to find the exact node for the given key */
for (node = search_root; node;) {
node_stack[++stack_ptr] = node;
matchlen = longest_prefix_match(trie, node, key);
if (node->prefixlen != matchlen ||
node->prefixlen == key->prefixlen)
break;
next_bit = extract_bit(key->data, node->prefixlen);
node = rcu_dereference(node->child[next_bit]);
}
if (!node || node->prefixlen != key->prefixlen ||
(node->flags & LPM_TREE_NODE_FLAG_IM))
goto find_leftmost;
/* The node with the exactly-matching key has been found,
* find the first node in postorder after the matched node.
*/
node = node_stack[stack_ptr];
while (stack_ptr > 0) {
parent = node_stack[stack_ptr - 1];
if (rcu_dereference(parent->child[0]) == node) {
search_root = rcu_dereference(parent->child[1]);
if (search_root)
goto find_leftmost;
}
if (!(parent->flags & LPM_TREE_NODE_FLAG_IM)) {
next_node = parent;
goto do_copy;
}
node = parent;
stack_ptr--;
}
/* did not find anything */
err = -ENOENT;
goto free_stack;
find_leftmost:
/* Find the leftmost non-intermediate node, all intermediate nodes
* have exact two children, so this function will never return NULL.
*/
for (node = search_root; node;) {
if (!(node->flags & LPM_TREE_NODE_FLAG_IM))
next_node = node;
node = rcu_dereference(node->child[0]);
}
do_copy:
next_key->prefixlen = next_node->prefixlen;
memcpy((void *)next_key + offsetof(struct bpf_lpm_trie_key, data),
next_node->data, trie->data_size);
free_stack:
kfree(node_stack);
return err;
}
static int trie_check_btf(const struct bpf_map *map,
const struct btf_type *key_type,
const struct btf_type *value_type)
{
/* Keys must have struct bpf_lpm_trie_key embedded. */
return BTF_INFO_KIND(key_type->info) != BTF_KIND_STRUCT ?
-EINVAL : 0;
}
const struct bpf_map_ops trie_map_ops = {
.map_alloc = trie_alloc,
.map_free = trie_free,
.map_get_next_key = trie_get_next_key,
.map_lookup_elem = trie_lookup_elem,
.map_update_elem = trie_update_elem,
.map_delete_elem = trie_delete_elem,
.map_check_btf = trie_check_btf,
};