^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1) // SPDX-License-Identifier: GPL-2.0-only
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2) /* Copyright (c) 2011-2014 PLUMgrid, http://plumgrid.com
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 3) * Copyright (c) 2016 Facebook
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 4) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 5) #include <linux/bpf.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 6) #include <linux/btf.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 7) #include <linux/jhash.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 8) #include <linux/filter.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 9) #include <linux/rculist_nulls.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 10) #include <linux/random.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 11) #include <uapi/linux/btf.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 12) #include <linux/rcupdate_trace.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 13) #include "percpu_freelist.h"
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 14) #include "bpf_lru_list.h"
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 15) #include "map_in_map.h"
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 16)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 17) #define HTAB_CREATE_FLAG_MASK \
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 18) (BPF_F_NO_PREALLOC | BPF_F_NO_COMMON_LRU | BPF_F_NUMA_NODE | \
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 19) BPF_F_ACCESS_MASK | BPF_F_ZERO_SEED)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 20)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 21) #define BATCH_OPS(_name) \
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 22) .map_lookup_batch = \
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 23) _name##_map_lookup_batch, \
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 24) .map_lookup_and_delete_batch = \
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 25) _name##_map_lookup_and_delete_batch, \
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 26) .map_update_batch = \
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 27) generic_map_update_batch, \
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 28) .map_delete_batch = \
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 29) generic_map_delete_batch
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 30)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 31) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 32) * The bucket lock has two protection scopes:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 33) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 34) * 1) Serializing concurrent operations from BPF programs on differrent
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 35) * CPUs
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 36) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 37) * 2) Serializing concurrent operations from BPF programs and sys_bpf()
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 38) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 39) * BPF programs can execute in any context including perf, kprobes and
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 40) * tracing. As there are almost no limits where perf, kprobes and tracing
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 41) * can be invoked from the lock operations need to be protected against
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 42) * deadlocks. Deadlocks can be caused by recursion and by an invocation in
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 43) * the lock held section when functions which acquire this lock are invoked
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 44) * from sys_bpf(). BPF recursion is prevented by incrementing the per CPU
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 45) * variable bpf_prog_active, which prevents BPF programs attached to perf
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 46) * events, kprobes and tracing to be invoked before the prior invocation
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 47) * from one of these contexts completed. sys_bpf() uses the same mechanism
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 48) * by pinning the task to the current CPU and incrementing the recursion
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 49) * protection accross the map operation.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 50) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 51) * This has subtle implications on PREEMPT_RT. PREEMPT_RT forbids certain
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 52) * operations like memory allocations (even with GFP_ATOMIC) from atomic
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 53) * contexts. This is required because even with GFP_ATOMIC the memory
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 54) * allocator calls into code pathes which acquire locks with long held lock
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 55) * sections. To ensure the deterministic behaviour these locks are regular
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 56) * spinlocks, which are converted to 'sleepable' spinlocks on RT. The only
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 57) * true atomic contexts on an RT kernel are the low level hardware
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 58) * handling, scheduling, low level interrupt handling, NMIs etc. None of
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 59) * these contexts should ever do memory allocations.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 60) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 61) * As regular device interrupt handlers and soft interrupts are forced into
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 62) * thread context, the existing code which does
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 63) * spin_lock*(); alloc(GPF_ATOMIC); spin_unlock*();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 64) * just works.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 65) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 66) * In theory the BPF locks could be converted to regular spinlocks as well,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 67) * but the bucket locks and percpu_freelist locks can be taken from
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 68) * arbitrary contexts (perf, kprobes, tracepoints) which are required to be
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 69) * atomic contexts even on RT. These mechanisms require preallocated maps,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 70) * so there is no need to invoke memory allocations within the lock held
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 71) * sections.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 72) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 73) * BPF maps which need dynamic allocation are only used from (forced)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 74) * thread context on RT and can therefore use regular spinlocks which in
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 75) * turn allows to invoke memory allocations from the lock held section.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 76) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 77) * On a non RT kernel this distinction is neither possible nor required.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 78) * spinlock maps to raw_spinlock and the extra code is optimized out by the
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 79) * compiler.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 80) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 81) struct bucket {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 82) struct hlist_nulls_head head;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 83) union {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 84) raw_spinlock_t raw_lock;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 85) spinlock_t lock;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 86) };
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 87) };
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 88)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 89) struct bpf_htab {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 90) struct bpf_map map;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 91) struct bucket *buckets;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 92) void *elems;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 93) union {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 94) struct pcpu_freelist freelist;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 95) struct bpf_lru lru;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 96) };
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 97) struct htab_elem *__percpu *extra_elems;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 98) atomic_t count; /* number of elements in this hashtable */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 99) u32 n_buckets; /* number of hash buckets */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 100) u32 elem_size; /* size of each element in bytes */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 101) u32 hashrnd;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 102) };
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 103)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 104) /* each htab element is struct htab_elem + key + value */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 105) struct htab_elem {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 106) union {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 107) struct hlist_nulls_node hash_node;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 108) struct {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 109) void *padding;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 110) union {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 111) struct bpf_htab *htab;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 112) struct pcpu_freelist_node fnode;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 113) struct htab_elem *batch_flink;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 114) };
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 115) };
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 116) };
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 117) union {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 118) struct rcu_head rcu;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 119) struct bpf_lru_node lru_node;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 120) };
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 121) u32 hash;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 122) char key[] __aligned(8);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 123) };
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 124)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 125) static inline bool htab_is_prealloc(const struct bpf_htab *htab)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 126) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 127) return !(htab->map.map_flags & BPF_F_NO_PREALLOC);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 128) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 129)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 130) static inline bool htab_use_raw_lock(const struct bpf_htab *htab)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 131) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 132) return (!IS_ENABLED(CONFIG_PREEMPT_RT) || htab_is_prealloc(htab));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 133) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 134)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 135) static void htab_init_buckets(struct bpf_htab *htab)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 136) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 137) unsigned i;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 138)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 139) for (i = 0; i < htab->n_buckets; i++) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 140) INIT_HLIST_NULLS_HEAD(&htab->buckets[i].head, i);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 141) if (htab_use_raw_lock(htab))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 142) raw_spin_lock_init(&htab->buckets[i].raw_lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 143) else
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 144) spin_lock_init(&htab->buckets[i].lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 145) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 146) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 147)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 148) static inline unsigned long htab_lock_bucket(const struct bpf_htab *htab,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 149) struct bucket *b)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 150) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 151) unsigned long flags;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 152)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 153) if (htab_use_raw_lock(htab))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 154) raw_spin_lock_irqsave(&b->raw_lock, flags);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 155) else
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 156) spin_lock_irqsave(&b->lock, flags);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 157) return flags;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 158) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 159)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 160) static inline void htab_unlock_bucket(const struct bpf_htab *htab,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 161) struct bucket *b,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 162) unsigned long flags)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 163) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 164) if (htab_use_raw_lock(htab))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 165) raw_spin_unlock_irqrestore(&b->raw_lock, flags);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 166) else
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 167) spin_unlock_irqrestore(&b->lock, flags);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 168) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 169)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 170) static bool htab_lru_map_delete_node(void *arg, struct bpf_lru_node *node);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 171)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 172) static bool htab_is_lru(const struct bpf_htab *htab)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 173) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 174) return htab->map.map_type == BPF_MAP_TYPE_LRU_HASH ||
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 175) htab->map.map_type == BPF_MAP_TYPE_LRU_PERCPU_HASH;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 176) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 177)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 178) static bool htab_is_percpu(const struct bpf_htab *htab)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 179) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 180) return htab->map.map_type == BPF_MAP_TYPE_PERCPU_HASH ||
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 181) htab->map.map_type == BPF_MAP_TYPE_LRU_PERCPU_HASH;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 182) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 183)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 184) static inline void htab_elem_set_ptr(struct htab_elem *l, u32 key_size,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 185) void __percpu *pptr)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 186) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 187) *(void __percpu **)(l->key + key_size) = pptr;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 188) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 189)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 190) static inline void __percpu *htab_elem_get_ptr(struct htab_elem *l, u32 key_size)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 191) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 192) return *(void __percpu **)(l->key + key_size);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 193) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 194)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 195) static void *fd_htab_map_get_ptr(const struct bpf_map *map, struct htab_elem *l)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 196) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 197) return *(void **)(l->key + roundup(map->key_size, 8));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 198) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 199)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 200) static struct htab_elem *get_htab_elem(struct bpf_htab *htab, int i)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 201) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 202) return (struct htab_elem *) (htab->elems + i * htab->elem_size);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 203) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 204)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 205) static void htab_free_elems(struct bpf_htab *htab)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 206) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 207) int i;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 208)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 209) if (!htab_is_percpu(htab))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 210) goto free_elems;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 211)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 212) for (i = 0; i < htab->map.max_entries; i++) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 213) void __percpu *pptr;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 214)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 215) pptr = htab_elem_get_ptr(get_htab_elem(htab, i),
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 216) htab->map.key_size);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 217) free_percpu(pptr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 218) cond_resched();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 219) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 220) free_elems:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 221) bpf_map_area_free(htab->elems);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 222) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 223)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 224) /* The LRU list has a lock (lru_lock). Each htab bucket has a lock
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 225) * (bucket_lock). If both locks need to be acquired together, the lock
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 226) * order is always lru_lock -> bucket_lock and this only happens in
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 227) * bpf_lru_list.c logic. For example, certain code path of
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 228) * bpf_lru_pop_free(), which is called by function prealloc_lru_pop(),
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 229) * will acquire lru_lock first followed by acquiring bucket_lock.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 230) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 231) * In hashtab.c, to avoid deadlock, lock acquisition of
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 232) * bucket_lock followed by lru_lock is not allowed. In such cases,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 233) * bucket_lock needs to be released first before acquiring lru_lock.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 234) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 235) static struct htab_elem *prealloc_lru_pop(struct bpf_htab *htab, void *key,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 236) u32 hash)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 237) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 238) struct bpf_lru_node *node = bpf_lru_pop_free(&htab->lru, hash);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 239) struct htab_elem *l;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 240)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 241) if (node) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 242) l = container_of(node, struct htab_elem, lru_node);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 243) memcpy(l->key, key, htab->map.key_size);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 244) return l;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 245) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 246)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 247) return NULL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 248) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 249)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 250) static int prealloc_init(struct bpf_htab *htab)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 251) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 252) u32 num_entries = htab->map.max_entries;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 253) int err = -ENOMEM, i;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 254)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 255) if (!htab_is_percpu(htab) && !htab_is_lru(htab))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 256) num_entries += num_possible_cpus();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 257)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 258) htab->elems = bpf_map_area_alloc(htab->elem_size * num_entries,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 259) htab->map.numa_node);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 260) if (!htab->elems)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 261) return -ENOMEM;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 262)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 263) if (!htab_is_percpu(htab))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 264) goto skip_percpu_elems;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 265)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 266) for (i = 0; i < num_entries; i++) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 267) u32 size = round_up(htab->map.value_size, 8);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 268) void __percpu *pptr;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 269)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 270) pptr = __alloc_percpu_gfp(size, 8, GFP_USER | __GFP_NOWARN);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 271) if (!pptr)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 272) goto free_elems;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 273) htab_elem_set_ptr(get_htab_elem(htab, i), htab->map.key_size,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 274) pptr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 275) cond_resched();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 276) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 277)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 278) skip_percpu_elems:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 279) if (htab_is_lru(htab))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 280) err = bpf_lru_init(&htab->lru,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 281) htab->map.map_flags & BPF_F_NO_COMMON_LRU,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 282) offsetof(struct htab_elem, hash) -
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 283) offsetof(struct htab_elem, lru_node),
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 284) htab_lru_map_delete_node,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 285) htab);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 286) else
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 287) err = pcpu_freelist_init(&htab->freelist);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 288)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 289) if (err)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 290) goto free_elems;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 291)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 292) if (htab_is_lru(htab))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 293) bpf_lru_populate(&htab->lru, htab->elems,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 294) offsetof(struct htab_elem, lru_node),
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 295) htab->elem_size, num_entries);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 296) else
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 297) pcpu_freelist_populate(&htab->freelist,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 298) htab->elems + offsetof(struct htab_elem, fnode),
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 299) htab->elem_size, num_entries);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 300)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 301) return 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 302)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 303) free_elems:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 304) htab_free_elems(htab);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 305) return err;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 306) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 307)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 308) static void prealloc_destroy(struct bpf_htab *htab)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 309) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 310) htab_free_elems(htab);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 311)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 312) if (htab_is_lru(htab))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 313) bpf_lru_destroy(&htab->lru);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 314) else
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 315) pcpu_freelist_destroy(&htab->freelist);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 316) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 317)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 318) static int alloc_extra_elems(struct bpf_htab *htab)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 319) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 320) struct htab_elem *__percpu *pptr, *l_new;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 321) struct pcpu_freelist_node *l;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 322) int cpu;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 323)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 324) pptr = __alloc_percpu_gfp(sizeof(struct htab_elem *), 8,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 325) GFP_USER | __GFP_NOWARN);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 326) if (!pptr)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 327) return -ENOMEM;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 328)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 329) for_each_possible_cpu(cpu) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 330) l = pcpu_freelist_pop(&htab->freelist);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 331) /* pop will succeed, since prealloc_init()
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 332) * preallocated extra num_possible_cpus elements
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 333) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 334) l_new = container_of(l, struct htab_elem, fnode);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 335) *per_cpu_ptr(pptr, cpu) = l_new;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 336) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 337) htab->extra_elems = pptr;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 338) return 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 339) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 340)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 341) /* Called from syscall */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 342) static int htab_map_alloc_check(union bpf_attr *attr)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 343) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 344) bool percpu = (attr->map_type == BPF_MAP_TYPE_PERCPU_HASH ||
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 345) attr->map_type == BPF_MAP_TYPE_LRU_PERCPU_HASH);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 346) bool lru = (attr->map_type == BPF_MAP_TYPE_LRU_HASH ||
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 347) attr->map_type == BPF_MAP_TYPE_LRU_PERCPU_HASH);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 348) /* percpu_lru means each cpu has its own LRU list.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 349) * it is different from BPF_MAP_TYPE_PERCPU_HASH where
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 350) * the map's value itself is percpu. percpu_lru has
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 351) * nothing to do with the map's value.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 352) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 353) bool percpu_lru = (attr->map_flags & BPF_F_NO_COMMON_LRU);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 354) bool prealloc = !(attr->map_flags & BPF_F_NO_PREALLOC);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 355) bool zero_seed = (attr->map_flags & BPF_F_ZERO_SEED);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 356) int numa_node = bpf_map_attr_numa_node(attr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 357)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 358) BUILD_BUG_ON(offsetof(struct htab_elem, htab) !=
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 359) offsetof(struct htab_elem, hash_node.pprev));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 360) BUILD_BUG_ON(offsetof(struct htab_elem, fnode.next) !=
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 361) offsetof(struct htab_elem, hash_node.pprev));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 362)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 363) if (lru && !bpf_capable())
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 364) /* LRU implementation is much complicated than other
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 365) * maps. Hence, limit to CAP_BPF.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 366) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 367) return -EPERM;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 368)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 369) if (zero_seed && !capable(CAP_SYS_ADMIN))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 370) /* Guard against local DoS, and discourage production use. */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 371) return -EPERM;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 372)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 373) if (attr->map_flags & ~HTAB_CREATE_FLAG_MASK ||
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 374) !bpf_map_flags_access_ok(attr->map_flags))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 375) return -EINVAL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 376)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 377) if (!lru && percpu_lru)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 378) return -EINVAL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 379)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 380) if (lru && !prealloc)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 381) return -ENOTSUPP;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 382)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 383) if (numa_node != NUMA_NO_NODE && (percpu || percpu_lru))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 384) return -EINVAL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 385)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 386) /* check sanity of attributes.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 387) * value_size == 0 may be allowed in the future to use map as a set
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 388) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 389) if (attr->max_entries == 0 || attr->key_size == 0 ||
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 390) attr->value_size == 0)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 391) return -EINVAL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 392)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 393) if (attr->key_size > MAX_BPF_STACK)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 394) /* eBPF programs initialize keys on stack, so they cannot be
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 395) * larger than max stack size
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 396) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 397) return -E2BIG;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 398)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 399) if (attr->value_size >= KMALLOC_MAX_SIZE -
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 400) MAX_BPF_STACK - sizeof(struct htab_elem))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 401) /* if value_size is bigger, the user space won't be able to
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 402) * access the elements via bpf syscall. This check also makes
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 403) * sure that the elem_size doesn't overflow and it's
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 404) * kmalloc-able later in htab_map_update_elem()
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 405) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 406) return -E2BIG;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 407)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 408) return 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 409) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 410)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 411) static struct bpf_map *htab_map_alloc(union bpf_attr *attr)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 412) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 413) bool percpu = (attr->map_type == BPF_MAP_TYPE_PERCPU_HASH ||
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 414) attr->map_type == BPF_MAP_TYPE_LRU_PERCPU_HASH);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 415) bool lru = (attr->map_type == BPF_MAP_TYPE_LRU_HASH ||
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 416) attr->map_type == BPF_MAP_TYPE_LRU_PERCPU_HASH);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 417) /* percpu_lru means each cpu has its own LRU list.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 418) * it is different from BPF_MAP_TYPE_PERCPU_HASH where
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 419) * the map's value itself is percpu. percpu_lru has
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 420) * nothing to do with the map's value.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 421) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 422) bool percpu_lru = (attr->map_flags & BPF_F_NO_COMMON_LRU);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 423) bool prealloc = !(attr->map_flags & BPF_F_NO_PREALLOC);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 424) struct bpf_htab *htab;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 425) u64 cost;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 426) int err;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 427)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 428) htab = kzalloc(sizeof(*htab), GFP_USER);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 429) if (!htab)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 430) return ERR_PTR(-ENOMEM);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 431)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 432) bpf_map_init_from_attr(&htab->map, attr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 433)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 434) if (percpu_lru) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 435) /* ensure each CPU's lru list has >=1 elements.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 436) * since we are at it, make each lru list has the same
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 437) * number of elements.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 438) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 439) htab->map.max_entries = roundup(attr->max_entries,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 440) num_possible_cpus());
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 441) if (htab->map.max_entries < attr->max_entries)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 442) htab->map.max_entries = rounddown(attr->max_entries,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 443) num_possible_cpus());
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 444) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 445)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 446) /* hash table size must be power of 2 */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 447) htab->n_buckets = roundup_pow_of_two(htab->map.max_entries);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 448)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 449) htab->elem_size = sizeof(struct htab_elem) +
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 450) round_up(htab->map.key_size, 8);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 451) if (percpu)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 452) htab->elem_size += sizeof(void *);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 453) else
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 454) htab->elem_size += round_up(htab->map.value_size, 8);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 455)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 456) err = -E2BIG;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 457) /* prevent zero size kmalloc and check for u32 overflow */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 458) if (htab->n_buckets == 0 ||
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 459) htab->n_buckets > U32_MAX / sizeof(struct bucket))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 460) goto free_htab;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 461)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 462) cost = (u64) htab->n_buckets * sizeof(struct bucket) +
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 463) (u64) htab->elem_size * htab->map.max_entries;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 464)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 465) if (percpu)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 466) cost += (u64) round_up(htab->map.value_size, 8) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 467) num_possible_cpus() * htab->map.max_entries;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 468) else
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 469) cost += (u64) htab->elem_size * num_possible_cpus();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 470)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 471) /* if map size is larger than memlock limit, reject it */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 472) err = bpf_map_charge_init(&htab->map.memory, cost);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 473) if (err)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 474) goto free_htab;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 475)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 476) err = -ENOMEM;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 477) htab->buckets = bpf_map_area_alloc(htab->n_buckets *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 478) sizeof(struct bucket),
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 479) htab->map.numa_node);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 480) if (!htab->buckets)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 481) goto free_charge;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 482)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 483) if (htab->map.map_flags & BPF_F_ZERO_SEED)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 484) htab->hashrnd = 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 485) else
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 486) htab->hashrnd = get_random_int();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 487)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 488) htab_init_buckets(htab);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 489)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 490) if (prealloc) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 491) err = prealloc_init(htab);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 492) if (err)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 493) goto free_buckets;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 494)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 495) if (!percpu && !lru) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 496) /* lru itself can remove the least used element, so
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 497) * there is no need for an extra elem during map_update.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 498) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 499) err = alloc_extra_elems(htab);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 500) if (err)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 501) goto free_prealloc;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 502) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 503) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 504)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 505) return &htab->map;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 506)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 507) free_prealloc:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 508) prealloc_destroy(htab);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 509) free_buckets:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 510) bpf_map_area_free(htab->buckets);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 511) free_charge:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 512) bpf_map_charge_finish(&htab->map.memory);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 513) free_htab:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 514) kfree(htab);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 515) return ERR_PTR(err);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 516) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 517)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 518) static inline u32 htab_map_hash(const void *key, u32 key_len, u32 hashrnd)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 519) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 520) return jhash(key, key_len, hashrnd);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 521) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 522)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 523) static inline struct bucket *__select_bucket(struct bpf_htab *htab, u32 hash)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 524) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 525) return &htab->buckets[hash & (htab->n_buckets - 1)];
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 526) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 527)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 528) static inline struct hlist_nulls_head *select_bucket(struct bpf_htab *htab, u32 hash)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 529) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 530) return &__select_bucket(htab, hash)->head;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 531) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 532)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 533) /* this lookup function can only be called with bucket lock taken */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 534) static struct htab_elem *lookup_elem_raw(struct hlist_nulls_head *head, u32 hash,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 535) void *key, u32 key_size)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 536) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 537) struct hlist_nulls_node *n;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 538) struct htab_elem *l;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 539)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 540) hlist_nulls_for_each_entry_rcu(l, n, head, hash_node)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 541) if (l->hash == hash && !memcmp(&l->key, key, key_size))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 542) return l;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 543)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 544) return NULL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 545) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 546)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 547) /* can be called without bucket lock. it will repeat the loop in
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 548) * the unlikely event when elements moved from one bucket into another
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 549) * while link list is being walked
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 550) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 551) static struct htab_elem *lookup_nulls_elem_raw(struct hlist_nulls_head *head,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 552) u32 hash, void *key,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 553) u32 key_size, u32 n_buckets)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 554) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 555) struct hlist_nulls_node *n;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 556) struct htab_elem *l;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 557)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 558) again:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 559) hlist_nulls_for_each_entry_rcu(l, n, head, hash_node)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 560) if (l->hash == hash && !memcmp(&l->key, key, key_size))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 561) return l;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 562)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 563) if (unlikely(get_nulls_value(n) != (hash & (n_buckets - 1))))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 564) goto again;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 565)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 566) return NULL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 567) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 568)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 569) /* Called from syscall or from eBPF program directly, so
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 570) * arguments have to match bpf_map_lookup_elem() exactly.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 571) * The return value is adjusted by BPF instructions
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 572) * in htab_map_gen_lookup().
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 573) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 574) static void *__htab_map_lookup_elem(struct bpf_map *map, void *key)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 575) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 576) struct bpf_htab *htab = container_of(map, struct bpf_htab, map);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 577) struct hlist_nulls_head *head;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 578) struct htab_elem *l;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 579) u32 hash, key_size;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 580)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 581) WARN_ON_ONCE(!rcu_read_lock_held() && !rcu_read_lock_trace_held());
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 582)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 583) key_size = map->key_size;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 584)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 585) hash = htab_map_hash(key, key_size, htab->hashrnd);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 586)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 587) head = select_bucket(htab, hash);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 588)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 589) l = lookup_nulls_elem_raw(head, hash, key, key_size, htab->n_buckets);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 590)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 591) return l;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 592) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 593)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 594) static void *htab_map_lookup_elem(struct bpf_map *map, void *key)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 595) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 596) struct htab_elem *l = __htab_map_lookup_elem(map, key);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 597)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 598) if (l)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 599) return l->key + round_up(map->key_size, 8);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 600)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 601) return NULL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 602) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 603)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 604) /* inline bpf_map_lookup_elem() call.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 605) * Instead of:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 606) * bpf_prog
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 607) * bpf_map_lookup_elem
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 608) * map->ops->map_lookup_elem
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 609) * htab_map_lookup_elem
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 610) * __htab_map_lookup_elem
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 611) * do:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 612) * bpf_prog
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 613) * __htab_map_lookup_elem
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 614) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 615) static int htab_map_gen_lookup(struct bpf_map *map, struct bpf_insn *insn_buf)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 616) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 617) struct bpf_insn *insn = insn_buf;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 618) const int ret = BPF_REG_0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 619)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 620) BUILD_BUG_ON(!__same_type(&__htab_map_lookup_elem,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 621) (void *(*)(struct bpf_map *map, void *key))NULL));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 622) *insn++ = BPF_EMIT_CALL(BPF_CAST_CALL(__htab_map_lookup_elem));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 623) *insn++ = BPF_JMP_IMM(BPF_JEQ, ret, 0, 1);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 624) *insn++ = BPF_ALU64_IMM(BPF_ADD, ret,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 625) offsetof(struct htab_elem, key) +
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 626) round_up(map->key_size, 8));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 627) return insn - insn_buf;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 628) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 629)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 630) static __always_inline void *__htab_lru_map_lookup_elem(struct bpf_map *map,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 631) void *key, const bool mark)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 632) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 633) struct htab_elem *l = __htab_map_lookup_elem(map, key);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 634)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 635) if (l) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 636) if (mark)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 637) bpf_lru_node_set_ref(&l->lru_node);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 638) return l->key + round_up(map->key_size, 8);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 639) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 640)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 641) return NULL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 642) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 643)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 644) static void *htab_lru_map_lookup_elem(struct bpf_map *map, void *key)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 645) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 646) return __htab_lru_map_lookup_elem(map, key, true);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 647) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 648)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 649) static void *htab_lru_map_lookup_elem_sys(struct bpf_map *map, void *key)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 650) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 651) return __htab_lru_map_lookup_elem(map, key, false);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 652) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 653)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 654) static int htab_lru_map_gen_lookup(struct bpf_map *map,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 655) struct bpf_insn *insn_buf)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 656) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 657) struct bpf_insn *insn = insn_buf;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 658) const int ret = BPF_REG_0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 659) const int ref_reg = BPF_REG_1;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 660)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 661) BUILD_BUG_ON(!__same_type(&__htab_map_lookup_elem,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 662) (void *(*)(struct bpf_map *map, void *key))NULL));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 663) *insn++ = BPF_EMIT_CALL(BPF_CAST_CALL(__htab_map_lookup_elem));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 664) *insn++ = BPF_JMP_IMM(BPF_JEQ, ret, 0, 4);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 665) *insn++ = BPF_LDX_MEM(BPF_B, ref_reg, ret,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 666) offsetof(struct htab_elem, lru_node) +
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 667) offsetof(struct bpf_lru_node, ref));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 668) *insn++ = BPF_JMP_IMM(BPF_JNE, ref_reg, 0, 1);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 669) *insn++ = BPF_ST_MEM(BPF_B, ret,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 670) offsetof(struct htab_elem, lru_node) +
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 671) offsetof(struct bpf_lru_node, ref),
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 672) 1);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 673) *insn++ = BPF_ALU64_IMM(BPF_ADD, ret,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 674) offsetof(struct htab_elem, key) +
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 675) round_up(map->key_size, 8));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 676) return insn - insn_buf;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 677) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 678)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 679) /* It is called from the bpf_lru_list when the LRU needs to delete
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 680) * older elements from the htab.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 681) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 682) static bool htab_lru_map_delete_node(void *arg, struct bpf_lru_node *node)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 683) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 684) struct bpf_htab *htab = (struct bpf_htab *)arg;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 685) struct htab_elem *l = NULL, *tgt_l;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 686) struct hlist_nulls_head *head;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 687) struct hlist_nulls_node *n;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 688) unsigned long flags;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 689) struct bucket *b;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 690)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 691) tgt_l = container_of(node, struct htab_elem, lru_node);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 692) b = __select_bucket(htab, tgt_l->hash);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 693) head = &b->head;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 694)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 695) flags = htab_lock_bucket(htab, b);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 696)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 697) hlist_nulls_for_each_entry_rcu(l, n, head, hash_node)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 698) if (l == tgt_l) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 699) hlist_nulls_del_rcu(&l->hash_node);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 700) break;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 701) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 702)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 703) htab_unlock_bucket(htab, b, flags);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 704)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 705) return l == tgt_l;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 706) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 707)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 708) /* Called from syscall */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 709) static int htab_map_get_next_key(struct bpf_map *map, void *key, void *next_key)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 710) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 711) struct bpf_htab *htab = container_of(map, struct bpf_htab, map);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 712) struct hlist_nulls_head *head;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 713) struct htab_elem *l, *next_l;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 714) u32 hash, key_size;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 715) int i = 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 716)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 717) WARN_ON_ONCE(!rcu_read_lock_held());
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 718)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 719) key_size = map->key_size;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 720)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 721) if (!key)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 722) goto find_first_elem;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 723)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 724) hash = htab_map_hash(key, key_size, htab->hashrnd);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 725)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 726) head = select_bucket(htab, hash);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 727)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 728) /* lookup the key */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 729) l = lookup_nulls_elem_raw(head, hash, key, key_size, htab->n_buckets);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 730)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 731) if (!l)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 732) goto find_first_elem;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 733)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 734) /* key was found, get next key in the same bucket */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 735) next_l = hlist_nulls_entry_safe(rcu_dereference_raw(hlist_nulls_next_rcu(&l->hash_node)),
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 736) struct htab_elem, hash_node);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 737)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 738) if (next_l) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 739) /* if next elem in this hash list is non-zero, just return it */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 740) memcpy(next_key, next_l->key, key_size);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 741) return 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 742) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 743)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 744) /* no more elements in this hash list, go to the next bucket */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 745) i = hash & (htab->n_buckets - 1);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 746) i++;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 747)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 748) find_first_elem:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 749) /* iterate over buckets */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 750) for (; i < htab->n_buckets; i++) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 751) head = select_bucket(htab, i);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 752)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 753) /* pick first element in the bucket */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 754) next_l = hlist_nulls_entry_safe(rcu_dereference_raw(hlist_nulls_first_rcu(head)),
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 755) struct htab_elem, hash_node);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 756) if (next_l) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 757) /* if it's not empty, just return it */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 758) memcpy(next_key, next_l->key, key_size);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 759) return 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 760) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 761) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 762)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 763) /* iterated over all buckets and all elements */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 764) return -ENOENT;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 765) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 766)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 767) static void htab_elem_free(struct bpf_htab *htab, struct htab_elem *l)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 768) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 769) if (htab->map.map_type == BPF_MAP_TYPE_PERCPU_HASH)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 770) free_percpu(htab_elem_get_ptr(l, htab->map.key_size));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 771) kfree(l);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 772) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 773)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 774) static void htab_elem_free_rcu(struct rcu_head *head)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 775) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 776) struct htab_elem *l = container_of(head, struct htab_elem, rcu);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 777) struct bpf_htab *htab = l->htab;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 778)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 779) htab_elem_free(htab, l);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 780) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 781)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 782) static void htab_put_fd_value(struct bpf_htab *htab, struct htab_elem *l)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 783) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 784) struct bpf_map *map = &htab->map;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 785) void *ptr;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 786)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 787) if (map->ops->map_fd_put_ptr) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 788) ptr = fd_htab_map_get_ptr(map, l);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 789) map->ops->map_fd_put_ptr(ptr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 790) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 791) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 792)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 793) static void free_htab_elem(struct bpf_htab *htab, struct htab_elem *l)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 794) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 795) htab_put_fd_value(htab, l);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 796)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 797) if (htab_is_prealloc(htab)) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 798) __pcpu_freelist_push(&htab->freelist, &l->fnode);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 799) } else {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 800) atomic_dec(&htab->count);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 801) l->htab = htab;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 802) call_rcu(&l->rcu, htab_elem_free_rcu);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 803) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 804) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 805)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 806) static void pcpu_copy_value(struct bpf_htab *htab, void __percpu *pptr,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 807) void *value, bool onallcpus)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 808) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 809) if (!onallcpus) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 810) /* copy true value_size bytes */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 811) memcpy(this_cpu_ptr(pptr), value, htab->map.value_size);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 812) } else {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 813) u32 size = round_up(htab->map.value_size, 8);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 814) int off = 0, cpu;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 815)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 816) for_each_possible_cpu(cpu) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 817) bpf_long_memcpy(per_cpu_ptr(pptr, cpu),
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 818) value + off, size);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 819) off += size;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 820) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 821) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 822) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 823)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 824) static void pcpu_init_value(struct bpf_htab *htab, void __percpu *pptr,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 825) void *value, bool onallcpus)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 826) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 827) /* When using prealloc and not setting the initial value on all cpus,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 828) * zero-fill element values for other cpus (just as what happens when
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 829) * not using prealloc). Otherwise, bpf program has no way to ensure
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 830) * known initial values for cpus other than current one
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 831) * (onallcpus=false always when coming from bpf prog).
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 832) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 833) if (htab_is_prealloc(htab) && !onallcpus) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 834) u32 size = round_up(htab->map.value_size, 8);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 835) int current_cpu = raw_smp_processor_id();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 836) int cpu;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 837)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 838) for_each_possible_cpu(cpu) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 839) if (cpu == current_cpu)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 840) bpf_long_memcpy(per_cpu_ptr(pptr, cpu), value,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 841) size);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 842) else
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 843) memset(per_cpu_ptr(pptr, cpu), 0, size);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 844) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 845) } else {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 846) pcpu_copy_value(htab, pptr, value, onallcpus);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 847) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 848) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 849)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 850) static bool fd_htab_map_needs_adjust(const struct bpf_htab *htab)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 851) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 852) return htab->map.map_type == BPF_MAP_TYPE_HASH_OF_MAPS &&
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 853) BITS_PER_LONG == 64;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 854) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 855)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 856) static struct htab_elem *alloc_htab_elem(struct bpf_htab *htab, void *key,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 857) void *value, u32 key_size, u32 hash,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 858) bool percpu, bool onallcpus,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 859) struct htab_elem *old_elem)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 860) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 861) u32 size = htab->map.value_size;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 862) bool prealloc = htab_is_prealloc(htab);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 863) struct htab_elem *l_new, **pl_new;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 864) void __percpu *pptr;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 865)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 866) if (prealloc) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 867) if (old_elem) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 868) /* if we're updating the existing element,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 869) * use per-cpu extra elems to avoid freelist_pop/push
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 870) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 871) pl_new = this_cpu_ptr(htab->extra_elems);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 872) l_new = *pl_new;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 873) htab_put_fd_value(htab, old_elem);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 874) *pl_new = old_elem;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 875) } else {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 876) struct pcpu_freelist_node *l;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 877)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 878) l = __pcpu_freelist_pop(&htab->freelist);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 879) if (!l)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 880) return ERR_PTR(-E2BIG);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 881) l_new = container_of(l, struct htab_elem, fnode);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 882) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 883) } else {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 884) if (atomic_inc_return(&htab->count) > htab->map.max_entries)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 885) if (!old_elem) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 886) /* when map is full and update() is replacing
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 887) * old element, it's ok to allocate, since
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 888) * old element will be freed immediately.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 889) * Otherwise return an error
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 890) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 891) l_new = ERR_PTR(-E2BIG);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 892) goto dec_count;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 893) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 894) l_new = kmalloc_node(htab->elem_size, GFP_ATOMIC | __GFP_NOWARN,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 895) htab->map.numa_node);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 896) if (!l_new) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 897) l_new = ERR_PTR(-ENOMEM);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 898) goto dec_count;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 899) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 900) check_and_init_map_lock(&htab->map,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 901) l_new->key + round_up(key_size, 8));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 902) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 903)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 904) memcpy(l_new->key, key, key_size);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 905) if (percpu) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 906) size = round_up(size, 8);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 907) if (prealloc) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 908) pptr = htab_elem_get_ptr(l_new, key_size);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 909) } else {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 910) /* alloc_percpu zero-fills */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 911) pptr = __alloc_percpu_gfp(size, 8,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 912) GFP_ATOMIC | __GFP_NOWARN);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 913) if (!pptr) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 914) kfree(l_new);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 915) l_new = ERR_PTR(-ENOMEM);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 916) goto dec_count;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 917) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 918) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 919)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 920) pcpu_init_value(htab, pptr, value, onallcpus);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 921)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 922) if (!prealloc)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 923) htab_elem_set_ptr(l_new, key_size, pptr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 924) } else if (fd_htab_map_needs_adjust(htab)) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 925) size = round_up(size, 8);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 926) memcpy(l_new->key + round_up(key_size, 8), value, size);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 927) } else {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 928) copy_map_value(&htab->map,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 929) l_new->key + round_up(key_size, 8),
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 930) value);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 931) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 932)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 933) l_new->hash = hash;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 934) return l_new;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 935) dec_count:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 936) atomic_dec(&htab->count);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 937) return l_new;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 938) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 939)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 940) static int check_flags(struct bpf_htab *htab, struct htab_elem *l_old,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 941) u64 map_flags)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 942) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 943) if (l_old && (map_flags & ~BPF_F_LOCK) == BPF_NOEXIST)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 944) /* elem already exists */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 945) return -EEXIST;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 946)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 947) if (!l_old && (map_flags & ~BPF_F_LOCK) == BPF_EXIST)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 948) /* elem doesn't exist, cannot update it */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 949) return -ENOENT;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 950)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 951) return 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 952) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 953)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 954) /* Called from syscall or from eBPF program */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 955) static int htab_map_update_elem(struct bpf_map *map, void *key, void *value,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 956) u64 map_flags)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 957) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 958) struct bpf_htab *htab = container_of(map, struct bpf_htab, map);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 959) struct htab_elem *l_new = NULL, *l_old;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 960) struct hlist_nulls_head *head;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 961) unsigned long flags;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 962) struct bucket *b;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 963) u32 key_size, hash;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 964) int ret;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 965)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 966) if (unlikely((map_flags & ~BPF_F_LOCK) > BPF_EXIST))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 967) /* unknown flags */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 968) return -EINVAL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 969)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 970) WARN_ON_ONCE(!rcu_read_lock_held() && !rcu_read_lock_trace_held());
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 971)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 972) key_size = map->key_size;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 973)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 974) hash = htab_map_hash(key, key_size, htab->hashrnd);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 975)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 976) b = __select_bucket(htab, hash);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 977) head = &b->head;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 978)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 979) if (unlikely(map_flags & BPF_F_LOCK)) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 980) if (unlikely(!map_value_has_spin_lock(map)))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 981) return -EINVAL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 982) /* find an element without taking the bucket lock */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 983) l_old = lookup_nulls_elem_raw(head, hash, key, key_size,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 984) htab->n_buckets);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 985) ret = check_flags(htab, l_old, map_flags);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 986) if (ret)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 987) return ret;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 988) if (l_old) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 989) /* grab the element lock and update value in place */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 990) copy_map_value_locked(map,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 991) l_old->key + round_up(key_size, 8),
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 992) value, false);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 993) return 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 994) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 995) /* fall through, grab the bucket lock and lookup again.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 996) * 99.9% chance that the element won't be found,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 997) * but second lookup under lock has to be done.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 998) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 999) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1000)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1001) flags = htab_lock_bucket(htab, b);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1002)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1003) l_old = lookup_elem_raw(head, hash, key, key_size);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1004)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1005) ret = check_flags(htab, l_old, map_flags);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1006) if (ret)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1007) goto err;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1008)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1009) if (unlikely(l_old && (map_flags & BPF_F_LOCK))) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1010) /* first lookup without the bucket lock didn't find the element,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1011) * but second lookup with the bucket lock found it.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1012) * This case is highly unlikely, but has to be dealt with:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1013) * grab the element lock in addition to the bucket lock
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1014) * and update element in place
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1015) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1016) copy_map_value_locked(map,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1017) l_old->key + round_up(key_size, 8),
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1018) value, false);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1019) ret = 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1020) goto err;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1021) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1022)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1023) l_new = alloc_htab_elem(htab, key, value, key_size, hash, false, false,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1024) l_old);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1025) if (IS_ERR(l_new)) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1026) /* all pre-allocated elements are in use or memory exhausted */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1027) ret = PTR_ERR(l_new);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1028) goto err;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1029) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1030)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1031) /* add new element to the head of the list, so that
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1032) * concurrent search will find it before old elem
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1033) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1034) hlist_nulls_add_head_rcu(&l_new->hash_node, head);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1035) if (l_old) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1036) hlist_nulls_del_rcu(&l_old->hash_node);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1037) if (!htab_is_prealloc(htab))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1038) free_htab_elem(htab, l_old);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1039) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1040) ret = 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1041) err:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1042) htab_unlock_bucket(htab, b, flags);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1043) return ret;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1044) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1045)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1046) static int htab_lru_map_update_elem(struct bpf_map *map, void *key, void *value,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1047) u64 map_flags)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1048) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1049) struct bpf_htab *htab = container_of(map, struct bpf_htab, map);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1050) struct htab_elem *l_new, *l_old = NULL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1051) struct hlist_nulls_head *head;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1052) unsigned long flags;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1053) struct bucket *b;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1054) u32 key_size, hash;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1055) int ret;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1056)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1057) if (unlikely(map_flags > BPF_EXIST))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1058) /* unknown flags */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1059) return -EINVAL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1060)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1061) WARN_ON_ONCE(!rcu_read_lock_held() && !rcu_read_lock_trace_held());
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1062)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1063) key_size = map->key_size;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1064)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1065) hash = htab_map_hash(key, key_size, htab->hashrnd);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1066)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1067) b = __select_bucket(htab, hash);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1068) head = &b->head;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1069)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1070) /* For LRU, we need to alloc before taking bucket's
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1071) * spinlock because getting free nodes from LRU may need
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1072) * to remove older elements from htab and this removal
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1073) * operation will need a bucket lock.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1074) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1075) l_new = prealloc_lru_pop(htab, key, hash);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1076) if (!l_new)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1077) return -ENOMEM;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1078) memcpy(l_new->key + round_up(map->key_size, 8), value, map->value_size);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1079)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1080) flags = htab_lock_bucket(htab, b);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1081)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1082) l_old = lookup_elem_raw(head, hash, key, key_size);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1083)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1084) ret = check_flags(htab, l_old, map_flags);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1085) if (ret)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1086) goto err;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1087)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1088) /* add new element to the head of the list, so that
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1089) * concurrent search will find it before old elem
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1090) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1091) hlist_nulls_add_head_rcu(&l_new->hash_node, head);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1092) if (l_old) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1093) bpf_lru_node_set_ref(&l_new->lru_node);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1094) hlist_nulls_del_rcu(&l_old->hash_node);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1095) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1096) ret = 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1097)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1098) err:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1099) htab_unlock_bucket(htab, b, flags);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1100)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1101) if (ret)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1102) bpf_lru_push_free(&htab->lru, &l_new->lru_node);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1103) else if (l_old)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1104) bpf_lru_push_free(&htab->lru, &l_old->lru_node);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1105)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1106) return ret;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1107) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1108)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1109) static int __htab_percpu_map_update_elem(struct bpf_map *map, void *key,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1110) void *value, u64 map_flags,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1111) bool onallcpus)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1112) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1113) struct bpf_htab *htab = container_of(map, struct bpf_htab, map);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1114) struct htab_elem *l_new = NULL, *l_old;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1115) struct hlist_nulls_head *head;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1116) unsigned long flags;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1117) struct bucket *b;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1118) u32 key_size, hash;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1119) int ret;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1120)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1121) if (unlikely(map_flags > BPF_EXIST))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1122) /* unknown flags */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1123) return -EINVAL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1124)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1125) WARN_ON_ONCE(!rcu_read_lock_held());
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1126)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1127) key_size = map->key_size;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1128)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1129) hash = htab_map_hash(key, key_size, htab->hashrnd);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1130)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1131) b = __select_bucket(htab, hash);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1132) head = &b->head;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1133)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1134) flags = htab_lock_bucket(htab, b);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1135)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1136) l_old = lookup_elem_raw(head, hash, key, key_size);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1137)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1138) ret = check_flags(htab, l_old, map_flags);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1139) if (ret)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1140) goto err;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1141)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1142) if (l_old) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1143) /* per-cpu hash map can update value in-place */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1144) pcpu_copy_value(htab, htab_elem_get_ptr(l_old, key_size),
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1145) value, onallcpus);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1146) } else {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1147) l_new = alloc_htab_elem(htab, key, value, key_size,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1148) hash, true, onallcpus, NULL);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1149) if (IS_ERR(l_new)) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1150) ret = PTR_ERR(l_new);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1151) goto err;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1152) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1153) hlist_nulls_add_head_rcu(&l_new->hash_node, head);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1154) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1155) ret = 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1156) err:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1157) htab_unlock_bucket(htab, b, flags);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1158) return ret;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1159) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1160)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1161) static int __htab_lru_percpu_map_update_elem(struct bpf_map *map, void *key,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1162) void *value, u64 map_flags,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1163) bool onallcpus)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1164) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1165) struct bpf_htab *htab = container_of(map, struct bpf_htab, map);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1166) struct htab_elem *l_new = NULL, *l_old;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1167) struct hlist_nulls_head *head;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1168) unsigned long flags;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1169) struct bucket *b;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1170) u32 key_size, hash;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1171) int ret;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1172)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1173) if (unlikely(map_flags > BPF_EXIST))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1174) /* unknown flags */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1175) return -EINVAL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1176)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1177) WARN_ON_ONCE(!rcu_read_lock_held());
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1178)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1179) key_size = map->key_size;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1180)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1181) hash = htab_map_hash(key, key_size, htab->hashrnd);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1182)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1183) b = __select_bucket(htab, hash);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1184) head = &b->head;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1185)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1186) /* For LRU, we need to alloc before taking bucket's
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1187) * spinlock because LRU's elem alloc may need
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1188) * to remove older elem from htab and this removal
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1189) * operation will need a bucket lock.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1190) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1191) if (map_flags != BPF_EXIST) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1192) l_new = prealloc_lru_pop(htab, key, hash);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1193) if (!l_new)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1194) return -ENOMEM;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1195) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1196)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1197) flags = htab_lock_bucket(htab, b);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1198)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1199) l_old = lookup_elem_raw(head, hash, key, key_size);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1200)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1201) ret = check_flags(htab, l_old, map_flags);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1202) if (ret)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1203) goto err;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1204)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1205) if (l_old) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1206) bpf_lru_node_set_ref(&l_old->lru_node);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1207)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1208) /* per-cpu hash map can update value in-place */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1209) pcpu_copy_value(htab, htab_elem_get_ptr(l_old, key_size),
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1210) value, onallcpus);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1211) } else {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1212) pcpu_init_value(htab, htab_elem_get_ptr(l_new, key_size),
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1213) value, onallcpus);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1214) hlist_nulls_add_head_rcu(&l_new->hash_node, head);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1215) l_new = NULL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1216) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1217) ret = 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1218) err:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1219) htab_unlock_bucket(htab, b, flags);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1220) if (l_new)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1221) bpf_lru_push_free(&htab->lru, &l_new->lru_node);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1222) return ret;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1223) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1224)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1225) static int htab_percpu_map_update_elem(struct bpf_map *map, void *key,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1226) void *value, u64 map_flags)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1227) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1228) return __htab_percpu_map_update_elem(map, key, value, map_flags, false);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1229) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1230)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1231) static int htab_lru_percpu_map_update_elem(struct bpf_map *map, void *key,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1232) void *value, u64 map_flags)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1233) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1234) return __htab_lru_percpu_map_update_elem(map, key, value, map_flags,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1235) false);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1236) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1237)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1238) /* Called from syscall or from eBPF program */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1239) static int htab_map_delete_elem(struct bpf_map *map, void *key)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1240) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1241) struct bpf_htab *htab = container_of(map, struct bpf_htab, map);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1242) struct hlist_nulls_head *head;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1243) struct bucket *b;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1244) struct htab_elem *l;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1245) unsigned long flags;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1246) u32 hash, key_size;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1247) int ret = -ENOENT;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1248)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1249) WARN_ON_ONCE(!rcu_read_lock_held() && !rcu_read_lock_trace_held());
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1250)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1251) key_size = map->key_size;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1252)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1253) hash = htab_map_hash(key, key_size, htab->hashrnd);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1254) b = __select_bucket(htab, hash);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1255) head = &b->head;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1256)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1257) flags = htab_lock_bucket(htab, b);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1258)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1259) l = lookup_elem_raw(head, hash, key, key_size);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1260)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1261) if (l) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1262) hlist_nulls_del_rcu(&l->hash_node);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1263) free_htab_elem(htab, l);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1264) ret = 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1265) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1266)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1267) htab_unlock_bucket(htab, b, flags);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1268) return ret;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1269) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1270)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1271) static int htab_lru_map_delete_elem(struct bpf_map *map, void *key)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1272) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1273) struct bpf_htab *htab = container_of(map, struct bpf_htab, map);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1274) struct hlist_nulls_head *head;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1275) struct bucket *b;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1276) struct htab_elem *l;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1277) unsigned long flags;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1278) u32 hash, key_size;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1279) int ret = -ENOENT;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1280)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1281) WARN_ON_ONCE(!rcu_read_lock_held() && !rcu_read_lock_trace_held());
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1282)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1283) key_size = map->key_size;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1284)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1285) hash = htab_map_hash(key, key_size, htab->hashrnd);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1286) b = __select_bucket(htab, hash);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1287) head = &b->head;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1288)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1289) flags = htab_lock_bucket(htab, b);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1290)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1291) l = lookup_elem_raw(head, hash, key, key_size);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1292)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1293) if (l) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1294) hlist_nulls_del_rcu(&l->hash_node);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1295) ret = 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1296) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1297)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1298) htab_unlock_bucket(htab, b, flags);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1299) if (l)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1300) bpf_lru_push_free(&htab->lru, &l->lru_node);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1301) return ret;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1302) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1303)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1304) static void delete_all_elements(struct bpf_htab *htab)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1305) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1306) int i;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1307)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1308) for (i = 0; i < htab->n_buckets; i++) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1309) struct hlist_nulls_head *head = select_bucket(htab, i);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1310) struct hlist_nulls_node *n;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1311) struct htab_elem *l;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1312)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1313) hlist_nulls_for_each_entry_safe(l, n, head, hash_node) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1314) hlist_nulls_del_rcu(&l->hash_node);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1315) htab_elem_free(htab, l);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1316) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1317) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1318) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1319)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1320) /* Called when map->refcnt goes to zero, either from workqueue or from syscall */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1321) static void htab_map_free(struct bpf_map *map)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1322) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1323) struct bpf_htab *htab = container_of(map, struct bpf_htab, map);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1324)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1325) /* bpf_free_used_maps() or close(map_fd) will trigger this map_free callback.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1326) * bpf_free_used_maps() is called after bpf prog is no longer executing.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1327) * There is no need to synchronize_rcu() here to protect map elements.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1328) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1329)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1330) /* some of free_htab_elem() callbacks for elements of this map may
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1331) * not have executed. Wait for them.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1332) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1333) rcu_barrier();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1334) if (!htab_is_prealloc(htab))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1335) delete_all_elements(htab);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1336) else
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1337) prealloc_destroy(htab);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1338)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1339) free_percpu(htab->extra_elems);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1340) bpf_map_area_free(htab->buckets);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1341) kfree(htab);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1342) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1343)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1344) static void htab_map_seq_show_elem(struct bpf_map *map, void *key,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1345) struct seq_file *m)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1346) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1347) void *value;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1348)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1349) rcu_read_lock();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1350)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1351) value = htab_map_lookup_elem(map, key);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1352) if (!value) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1353) rcu_read_unlock();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1354) return;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1355) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1356)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1357) btf_type_seq_show(map->btf, map->btf_key_type_id, key, m);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1358) seq_puts(m, ": ");
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1359) btf_type_seq_show(map->btf, map->btf_value_type_id, value, m);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1360) seq_puts(m, "\n");
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1361)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1362) rcu_read_unlock();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1363) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1364)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1365) static int
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1366) __htab_map_lookup_and_delete_batch(struct bpf_map *map,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1367) const union bpf_attr *attr,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1368) union bpf_attr __user *uattr,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1369) bool do_delete, bool is_lru_map,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1370) bool is_percpu)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1371) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1372) struct bpf_htab *htab = container_of(map, struct bpf_htab, map);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1373) u32 bucket_cnt, total, key_size, value_size, roundup_key_size;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1374) void *keys = NULL, *values = NULL, *value, *dst_key, *dst_val;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1375) void __user *uvalues = u64_to_user_ptr(attr->batch.values);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1376) void __user *ukeys = u64_to_user_ptr(attr->batch.keys);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1377) void *ubatch = u64_to_user_ptr(attr->batch.in_batch);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1378) u32 batch, max_count, size, bucket_size;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1379) struct htab_elem *node_to_free = NULL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1380) u64 elem_map_flags, map_flags;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1381) struct hlist_nulls_head *head;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1382) struct hlist_nulls_node *n;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1383) unsigned long flags = 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1384) bool locked = false;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1385) struct htab_elem *l;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1386) struct bucket *b;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1387) int ret = 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1388)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1389) elem_map_flags = attr->batch.elem_flags;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1390) if ((elem_map_flags & ~BPF_F_LOCK) ||
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1391) ((elem_map_flags & BPF_F_LOCK) && !map_value_has_spin_lock(map)))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1392) return -EINVAL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1393)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1394) map_flags = attr->batch.flags;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1395) if (map_flags)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1396) return -EINVAL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1397)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1398) max_count = attr->batch.count;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1399) if (!max_count)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1400) return 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1401)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1402) if (put_user(0, &uattr->batch.count))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1403) return -EFAULT;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1404)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1405) batch = 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1406) if (ubatch && copy_from_user(&batch, ubatch, sizeof(batch)))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1407) return -EFAULT;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1408)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1409) if (batch >= htab->n_buckets)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1410) return -ENOENT;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1411)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1412) key_size = htab->map.key_size;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1413) roundup_key_size = round_up(htab->map.key_size, 8);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1414) value_size = htab->map.value_size;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1415) size = round_up(value_size, 8);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1416) if (is_percpu)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1417) value_size = size * num_possible_cpus();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1418) total = 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1419) /* while experimenting with hash tables with sizes ranging from 10 to
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1420) * 1000, it was observed that a bucket can have upto 5 entries.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1421) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1422) bucket_size = 5;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1423)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1424) alloc:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1425) /* We cannot do copy_from_user or copy_to_user inside
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1426) * the rcu_read_lock. Allocate enough space here.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1427) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1428) keys = kvmalloc_array(key_size, bucket_size, GFP_USER | __GFP_NOWARN);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1429) values = kvmalloc_array(value_size, bucket_size, GFP_USER | __GFP_NOWARN);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1430) if (!keys || !values) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1431) ret = -ENOMEM;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1432) goto after_loop;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1433) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1434)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1435) again:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1436) bpf_disable_instrumentation();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1437) rcu_read_lock();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1438) again_nocopy:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1439) dst_key = keys;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1440) dst_val = values;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1441) b = &htab->buckets[batch];
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1442) head = &b->head;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1443) /* do not grab the lock unless need it (bucket_cnt > 0). */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1444) if (locked)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1445) flags = htab_lock_bucket(htab, b);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1446)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1447) bucket_cnt = 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1448) hlist_nulls_for_each_entry_rcu(l, n, head, hash_node)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1449) bucket_cnt++;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1450)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1451) if (bucket_cnt && !locked) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1452) locked = true;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1453) goto again_nocopy;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1454) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1455)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1456) if (bucket_cnt > (max_count - total)) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1457) if (total == 0)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1458) ret = -ENOSPC;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1459) /* Note that since bucket_cnt > 0 here, it is implicit
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1460) * that the locked was grabbed, so release it.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1461) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1462) htab_unlock_bucket(htab, b, flags);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1463) rcu_read_unlock();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1464) bpf_enable_instrumentation();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1465) goto after_loop;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1466) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1467)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1468) if (bucket_cnt > bucket_size) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1469) bucket_size = bucket_cnt;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1470) /* Note that since bucket_cnt > 0 here, it is implicit
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1471) * that the locked was grabbed, so release it.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1472) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1473) htab_unlock_bucket(htab, b, flags);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1474) rcu_read_unlock();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1475) bpf_enable_instrumentation();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1476) kvfree(keys);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1477) kvfree(values);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1478) goto alloc;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1479) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1480)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1481) /* Next block is only safe to run if you have grabbed the lock */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1482) if (!locked)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1483) goto next_batch;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1484)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1485) hlist_nulls_for_each_entry_safe(l, n, head, hash_node) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1486) memcpy(dst_key, l->key, key_size);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1487)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1488) if (is_percpu) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1489) int off = 0, cpu;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1490) void __percpu *pptr;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1491)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1492) pptr = htab_elem_get_ptr(l, map->key_size);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1493) for_each_possible_cpu(cpu) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1494) bpf_long_memcpy(dst_val + off,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1495) per_cpu_ptr(pptr, cpu), size);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1496) off += size;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1497) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1498) } else {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1499) value = l->key + roundup_key_size;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1500) if (elem_map_flags & BPF_F_LOCK)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1501) copy_map_value_locked(map, dst_val, value,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1502) true);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1503) else
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1504) copy_map_value(map, dst_val, value);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1505) check_and_init_map_lock(map, dst_val);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1506) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1507) if (do_delete) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1508) hlist_nulls_del_rcu(&l->hash_node);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1509)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1510) /* bpf_lru_push_free() will acquire lru_lock, which
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1511) * may cause deadlock. See comments in function
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1512) * prealloc_lru_pop(). Let us do bpf_lru_push_free()
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1513) * after releasing the bucket lock.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1514) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1515) if (is_lru_map) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1516) l->batch_flink = node_to_free;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1517) node_to_free = l;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1518) } else {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1519) free_htab_elem(htab, l);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1520) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1521) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1522) dst_key += key_size;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1523) dst_val += value_size;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1524) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1525)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1526) htab_unlock_bucket(htab, b, flags);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1527) locked = false;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1528)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1529) while (node_to_free) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1530) l = node_to_free;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1531) node_to_free = node_to_free->batch_flink;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1532) bpf_lru_push_free(&htab->lru, &l->lru_node);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1533) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1534)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1535) next_batch:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1536) /* If we are not copying data, we can go to next bucket and avoid
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1537) * unlocking the rcu.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1538) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1539) if (!bucket_cnt && (batch + 1 < htab->n_buckets)) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1540) batch++;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1541) goto again_nocopy;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1542) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1543)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1544) rcu_read_unlock();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1545) bpf_enable_instrumentation();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1546) if (bucket_cnt && (copy_to_user(ukeys + total * key_size, keys,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1547) key_size * bucket_cnt) ||
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1548) copy_to_user(uvalues + total * value_size, values,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1549) value_size * bucket_cnt))) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1550) ret = -EFAULT;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1551) goto after_loop;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1552) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1553)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1554) total += bucket_cnt;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1555) batch++;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1556) if (batch >= htab->n_buckets) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1557) ret = -ENOENT;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1558) goto after_loop;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1559) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1560) goto again;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1561)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1562) after_loop:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1563) if (ret == -EFAULT)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1564) goto out;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1565)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1566) /* copy # of entries and next batch */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1567) ubatch = u64_to_user_ptr(attr->batch.out_batch);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1568) if (copy_to_user(ubatch, &batch, sizeof(batch)) ||
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1569) put_user(total, &uattr->batch.count))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1570) ret = -EFAULT;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1571)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1572) out:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1573) kvfree(keys);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1574) kvfree(values);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1575) return ret;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1576) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1577)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1578) static int
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1579) htab_percpu_map_lookup_batch(struct bpf_map *map, const union bpf_attr *attr,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1580) union bpf_attr __user *uattr)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1581) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1582) return __htab_map_lookup_and_delete_batch(map, attr, uattr, false,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1583) false, true);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1584) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1585)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1586) static int
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1587) htab_percpu_map_lookup_and_delete_batch(struct bpf_map *map,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1588) const union bpf_attr *attr,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1589) union bpf_attr __user *uattr)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1590) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1591) return __htab_map_lookup_and_delete_batch(map, attr, uattr, true,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1592) false, true);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1593) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1594)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1595) static int
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1596) htab_map_lookup_batch(struct bpf_map *map, const union bpf_attr *attr,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1597) union bpf_attr __user *uattr)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1598) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1599) return __htab_map_lookup_and_delete_batch(map, attr, uattr, false,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1600) false, false);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1601) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1602)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1603) static int
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1604) htab_map_lookup_and_delete_batch(struct bpf_map *map,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1605) const union bpf_attr *attr,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1606) union bpf_attr __user *uattr)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1607) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1608) return __htab_map_lookup_and_delete_batch(map, attr, uattr, true,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1609) false, false);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1610) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1611)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1612) static int
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1613) htab_lru_percpu_map_lookup_batch(struct bpf_map *map,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1614) const union bpf_attr *attr,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1615) union bpf_attr __user *uattr)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1616) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1617) return __htab_map_lookup_and_delete_batch(map, attr, uattr, false,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1618) true, true);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1619) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1620)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1621) static int
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1622) htab_lru_percpu_map_lookup_and_delete_batch(struct bpf_map *map,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1623) const union bpf_attr *attr,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1624) union bpf_attr __user *uattr)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1625) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1626) return __htab_map_lookup_and_delete_batch(map, attr, uattr, true,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1627) true, true);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1628) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1629)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1630) static int
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1631) htab_lru_map_lookup_batch(struct bpf_map *map, const union bpf_attr *attr,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1632) union bpf_attr __user *uattr)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1633) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1634) return __htab_map_lookup_and_delete_batch(map, attr, uattr, false,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1635) true, false);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1636) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1637)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1638) static int
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1639) htab_lru_map_lookup_and_delete_batch(struct bpf_map *map,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1640) const union bpf_attr *attr,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1641) union bpf_attr __user *uattr)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1642) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1643) return __htab_map_lookup_and_delete_batch(map, attr, uattr, true,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1644) true, false);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1645) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1646)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1647) struct bpf_iter_seq_hash_map_info {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1648) struct bpf_map *map;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1649) struct bpf_htab *htab;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1650) void *percpu_value_buf; // non-zero means percpu hash
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1651) u32 bucket_id;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1652) u32 skip_elems;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1653) };
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1654)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1655) static struct htab_elem *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1656) bpf_hash_map_seq_find_next(struct bpf_iter_seq_hash_map_info *info,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1657) struct htab_elem *prev_elem)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1658) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1659) const struct bpf_htab *htab = info->htab;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1660) u32 skip_elems = info->skip_elems;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1661) u32 bucket_id = info->bucket_id;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1662) struct hlist_nulls_head *head;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1663) struct hlist_nulls_node *n;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1664) struct htab_elem *elem;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1665) struct bucket *b;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1666) u32 i, count;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1667)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1668) if (bucket_id >= htab->n_buckets)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1669) return NULL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1670)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1671) /* try to find next elem in the same bucket */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1672) if (prev_elem) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1673) /* no update/deletion on this bucket, prev_elem should be still valid
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1674) * and we won't skip elements.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1675) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1676) n = rcu_dereference_raw(hlist_nulls_next_rcu(&prev_elem->hash_node));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1677) elem = hlist_nulls_entry_safe(n, struct htab_elem, hash_node);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1678) if (elem)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1679) return elem;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1680)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1681) /* not found, unlock and go to the next bucket */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1682) b = &htab->buckets[bucket_id++];
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1683) rcu_read_unlock();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1684) skip_elems = 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1685) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1686)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1687) for (i = bucket_id; i < htab->n_buckets; i++) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1688) b = &htab->buckets[i];
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1689) rcu_read_lock();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1690)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1691) count = 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1692) head = &b->head;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1693) hlist_nulls_for_each_entry_rcu(elem, n, head, hash_node) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1694) if (count >= skip_elems) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1695) info->bucket_id = i;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1696) info->skip_elems = count;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1697) return elem;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1698) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1699) count++;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1700) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1701)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1702) rcu_read_unlock();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1703) skip_elems = 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1704) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1705)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1706) info->bucket_id = i;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1707) info->skip_elems = 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1708) return NULL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1709) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1710)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1711) static void *bpf_hash_map_seq_start(struct seq_file *seq, loff_t *pos)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1712) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1713) struct bpf_iter_seq_hash_map_info *info = seq->private;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1714) struct htab_elem *elem;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1715)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1716) elem = bpf_hash_map_seq_find_next(info, NULL);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1717) if (!elem)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1718) return NULL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1719)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1720) if (*pos == 0)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1721) ++*pos;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1722) return elem;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1723) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1724)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1725) static void *bpf_hash_map_seq_next(struct seq_file *seq, void *v, loff_t *pos)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1726) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1727) struct bpf_iter_seq_hash_map_info *info = seq->private;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1728)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1729) ++*pos;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1730) ++info->skip_elems;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1731) return bpf_hash_map_seq_find_next(info, v);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1732) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1733)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1734) static int __bpf_hash_map_seq_show(struct seq_file *seq, struct htab_elem *elem)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1735) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1736) struct bpf_iter_seq_hash_map_info *info = seq->private;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1737) u32 roundup_key_size, roundup_value_size;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1738) struct bpf_iter__bpf_map_elem ctx = {};
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1739) struct bpf_map *map = info->map;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1740) struct bpf_iter_meta meta;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1741) int ret = 0, off = 0, cpu;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1742) struct bpf_prog *prog;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1743) void __percpu *pptr;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1744)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1745) meta.seq = seq;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1746) prog = bpf_iter_get_info(&meta, elem == NULL);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1747) if (prog) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1748) ctx.meta = &meta;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1749) ctx.map = info->map;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1750) if (elem) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1751) roundup_key_size = round_up(map->key_size, 8);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1752) ctx.key = elem->key;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1753) if (!info->percpu_value_buf) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1754) ctx.value = elem->key + roundup_key_size;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1755) } else {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1756) roundup_value_size = round_up(map->value_size, 8);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1757) pptr = htab_elem_get_ptr(elem, map->key_size);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1758) for_each_possible_cpu(cpu) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1759) bpf_long_memcpy(info->percpu_value_buf + off,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1760) per_cpu_ptr(pptr, cpu),
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1761) roundup_value_size);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1762) off += roundup_value_size;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1763) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1764) ctx.value = info->percpu_value_buf;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1765) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1766) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1767) ret = bpf_iter_run_prog(prog, &ctx);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1768) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1769)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1770) return ret;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1771) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1772)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1773) static int bpf_hash_map_seq_show(struct seq_file *seq, void *v)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1774) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1775) return __bpf_hash_map_seq_show(seq, v);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1776) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1777)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1778) static void bpf_hash_map_seq_stop(struct seq_file *seq, void *v)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1779) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1780) if (!v)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1781) (void)__bpf_hash_map_seq_show(seq, NULL);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1782) else
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1783) rcu_read_unlock();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1784) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1785)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1786) static int bpf_iter_init_hash_map(void *priv_data,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1787) struct bpf_iter_aux_info *aux)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1788) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1789) struct bpf_iter_seq_hash_map_info *seq_info = priv_data;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1790) struct bpf_map *map = aux->map;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1791) void *value_buf;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1792) u32 buf_size;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1793)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1794) if (map->map_type == BPF_MAP_TYPE_PERCPU_HASH ||
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1795) map->map_type == BPF_MAP_TYPE_LRU_PERCPU_HASH) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1796) buf_size = round_up(map->value_size, 8) * num_possible_cpus();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1797) value_buf = kmalloc(buf_size, GFP_USER | __GFP_NOWARN);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1798) if (!value_buf)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1799) return -ENOMEM;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1800)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1801) seq_info->percpu_value_buf = value_buf;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1802) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1803)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1804) seq_info->map = map;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1805) seq_info->htab = container_of(map, struct bpf_htab, map);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1806) return 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1807) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1808)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1809) static void bpf_iter_fini_hash_map(void *priv_data)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1810) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1811) struct bpf_iter_seq_hash_map_info *seq_info = priv_data;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1812)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1813) kfree(seq_info->percpu_value_buf);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1814) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1815)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1816) static const struct seq_operations bpf_hash_map_seq_ops = {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1817) .start = bpf_hash_map_seq_start,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1818) .next = bpf_hash_map_seq_next,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1819) .stop = bpf_hash_map_seq_stop,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1820) .show = bpf_hash_map_seq_show,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1821) };
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1822)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1823) static const struct bpf_iter_seq_info iter_seq_info = {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1824) .seq_ops = &bpf_hash_map_seq_ops,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1825) .init_seq_private = bpf_iter_init_hash_map,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1826) .fini_seq_private = bpf_iter_fini_hash_map,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1827) .seq_priv_size = sizeof(struct bpf_iter_seq_hash_map_info),
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1828) };
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1829)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1830) static int htab_map_btf_id;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1831) const struct bpf_map_ops htab_map_ops = {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1832) .map_meta_equal = bpf_map_meta_equal,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1833) .map_alloc_check = htab_map_alloc_check,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1834) .map_alloc = htab_map_alloc,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1835) .map_free = htab_map_free,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1836) .map_get_next_key = htab_map_get_next_key,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1837) .map_lookup_elem = htab_map_lookup_elem,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1838) .map_update_elem = htab_map_update_elem,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1839) .map_delete_elem = htab_map_delete_elem,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1840) .map_gen_lookup = htab_map_gen_lookup,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1841) .map_seq_show_elem = htab_map_seq_show_elem,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1842) BATCH_OPS(htab),
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1843) .map_btf_name = "bpf_htab",
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1844) .map_btf_id = &htab_map_btf_id,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1845) .iter_seq_info = &iter_seq_info,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1846) };
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1847)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1848) static int htab_lru_map_btf_id;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1849) const struct bpf_map_ops htab_lru_map_ops = {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1850) .map_meta_equal = bpf_map_meta_equal,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1851) .map_alloc_check = htab_map_alloc_check,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1852) .map_alloc = htab_map_alloc,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1853) .map_free = htab_map_free,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1854) .map_get_next_key = htab_map_get_next_key,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1855) .map_lookup_elem = htab_lru_map_lookup_elem,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1856) .map_lookup_elem_sys_only = htab_lru_map_lookup_elem_sys,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1857) .map_update_elem = htab_lru_map_update_elem,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1858) .map_delete_elem = htab_lru_map_delete_elem,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1859) .map_gen_lookup = htab_lru_map_gen_lookup,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1860) .map_seq_show_elem = htab_map_seq_show_elem,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1861) BATCH_OPS(htab_lru),
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1862) .map_btf_name = "bpf_htab",
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1863) .map_btf_id = &htab_lru_map_btf_id,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1864) .iter_seq_info = &iter_seq_info,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1865) };
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1866)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1867) /* Called from eBPF program */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1868) static void *htab_percpu_map_lookup_elem(struct bpf_map *map, void *key)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1869) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1870) struct htab_elem *l = __htab_map_lookup_elem(map, key);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1871)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1872) if (l)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1873) return this_cpu_ptr(htab_elem_get_ptr(l, map->key_size));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1874) else
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1875) return NULL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1876) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1877)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1878) static void *htab_lru_percpu_map_lookup_elem(struct bpf_map *map, void *key)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1879) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1880) struct htab_elem *l = __htab_map_lookup_elem(map, key);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1881)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1882) if (l) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1883) bpf_lru_node_set_ref(&l->lru_node);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1884) return this_cpu_ptr(htab_elem_get_ptr(l, map->key_size));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1885) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1886)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1887) return NULL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1888) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1889)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1890) int bpf_percpu_hash_copy(struct bpf_map *map, void *key, void *value)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1891) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1892) struct htab_elem *l;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1893) void __percpu *pptr;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1894) int ret = -ENOENT;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1895) int cpu, off = 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1896) u32 size;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1897)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1898) /* per_cpu areas are zero-filled and bpf programs can only
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1899) * access 'value_size' of them, so copying rounded areas
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1900) * will not leak any kernel data
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1901) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1902) size = round_up(map->value_size, 8);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1903) rcu_read_lock();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1904) l = __htab_map_lookup_elem(map, key);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1905) if (!l)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1906) goto out;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1907) /* We do not mark LRU map element here in order to not mess up
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1908) * eviction heuristics when user space does a map walk.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1909) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1910) pptr = htab_elem_get_ptr(l, map->key_size);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1911) for_each_possible_cpu(cpu) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1912) bpf_long_memcpy(value + off,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1913) per_cpu_ptr(pptr, cpu), size);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1914) off += size;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1915) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1916) ret = 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1917) out:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1918) rcu_read_unlock();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1919) return ret;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1920) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1921)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1922) int bpf_percpu_hash_update(struct bpf_map *map, void *key, void *value,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1923) u64 map_flags)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1924) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1925) struct bpf_htab *htab = container_of(map, struct bpf_htab, map);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1926) int ret;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1927)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1928) rcu_read_lock();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1929) if (htab_is_lru(htab))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1930) ret = __htab_lru_percpu_map_update_elem(map, key, value,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1931) map_flags, true);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1932) else
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1933) ret = __htab_percpu_map_update_elem(map, key, value, map_flags,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1934) true);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1935) rcu_read_unlock();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1936)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1937) return ret;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1938) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1939)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1940) static void htab_percpu_map_seq_show_elem(struct bpf_map *map, void *key,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1941) struct seq_file *m)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1942) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1943) struct htab_elem *l;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1944) void __percpu *pptr;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1945) int cpu;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1946)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1947) rcu_read_lock();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1948)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1949) l = __htab_map_lookup_elem(map, key);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1950) if (!l) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1951) rcu_read_unlock();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1952) return;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1953) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1954)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1955) btf_type_seq_show(map->btf, map->btf_key_type_id, key, m);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1956) seq_puts(m, ": {\n");
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1957) pptr = htab_elem_get_ptr(l, map->key_size);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1958) for_each_possible_cpu(cpu) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1959) seq_printf(m, "\tcpu%d: ", cpu);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1960) btf_type_seq_show(map->btf, map->btf_value_type_id,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1961) per_cpu_ptr(pptr, cpu), m);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1962) seq_puts(m, "\n");
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1963) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1964) seq_puts(m, "}\n");
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1965)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1966) rcu_read_unlock();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1967) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1968)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1969) static int htab_percpu_map_btf_id;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1970) const struct bpf_map_ops htab_percpu_map_ops = {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1971) .map_meta_equal = bpf_map_meta_equal,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1972) .map_alloc_check = htab_map_alloc_check,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1973) .map_alloc = htab_map_alloc,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1974) .map_free = htab_map_free,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1975) .map_get_next_key = htab_map_get_next_key,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1976) .map_lookup_elem = htab_percpu_map_lookup_elem,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1977) .map_update_elem = htab_percpu_map_update_elem,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1978) .map_delete_elem = htab_map_delete_elem,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1979) .map_seq_show_elem = htab_percpu_map_seq_show_elem,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1980) BATCH_OPS(htab_percpu),
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1981) .map_btf_name = "bpf_htab",
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1982) .map_btf_id = &htab_percpu_map_btf_id,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1983) .iter_seq_info = &iter_seq_info,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1984) };
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1985)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1986) static int htab_lru_percpu_map_btf_id;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1987) const struct bpf_map_ops htab_lru_percpu_map_ops = {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1988) .map_meta_equal = bpf_map_meta_equal,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1989) .map_alloc_check = htab_map_alloc_check,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1990) .map_alloc = htab_map_alloc,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1991) .map_free = htab_map_free,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1992) .map_get_next_key = htab_map_get_next_key,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1993) .map_lookup_elem = htab_lru_percpu_map_lookup_elem,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1994) .map_update_elem = htab_lru_percpu_map_update_elem,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1995) .map_delete_elem = htab_lru_map_delete_elem,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1996) .map_seq_show_elem = htab_percpu_map_seq_show_elem,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1997) BATCH_OPS(htab_lru_percpu),
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1998) .map_btf_name = "bpf_htab",
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1999) .map_btf_id = &htab_lru_percpu_map_btf_id,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2000) .iter_seq_info = &iter_seq_info,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2001) };
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2002)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2003) static int fd_htab_map_alloc_check(union bpf_attr *attr)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2004) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2005) if (attr->value_size != sizeof(u32))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2006) return -EINVAL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2007) return htab_map_alloc_check(attr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2008) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2009)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2010) static void fd_htab_map_free(struct bpf_map *map)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2011) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2012) struct bpf_htab *htab = container_of(map, struct bpf_htab, map);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2013) struct hlist_nulls_node *n;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2014) struct hlist_nulls_head *head;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2015) struct htab_elem *l;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2016) int i;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2017)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2018) for (i = 0; i < htab->n_buckets; i++) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2019) head = select_bucket(htab, i);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2020)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2021) hlist_nulls_for_each_entry_safe(l, n, head, hash_node) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2022) void *ptr = fd_htab_map_get_ptr(map, l);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2023)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2024) map->ops->map_fd_put_ptr(ptr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2025) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2026) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2027)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2028) htab_map_free(map);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2029) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2030)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2031) /* only called from syscall */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2032) int bpf_fd_htab_map_lookup_elem(struct bpf_map *map, void *key, u32 *value)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2033) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2034) void **ptr;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2035) int ret = 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2036)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2037) if (!map->ops->map_fd_sys_lookup_elem)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2038) return -ENOTSUPP;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2039)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2040) rcu_read_lock();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2041) ptr = htab_map_lookup_elem(map, key);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2042) if (ptr)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2043) *value = map->ops->map_fd_sys_lookup_elem(READ_ONCE(*ptr));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2044) else
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2045) ret = -ENOENT;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2046) rcu_read_unlock();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2047)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2048) return ret;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2049) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2050)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2051) /* only called from syscall */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2052) int bpf_fd_htab_map_update_elem(struct bpf_map *map, struct file *map_file,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2053) void *key, void *value, u64 map_flags)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2054) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2055) void *ptr;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2056) int ret;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2057) u32 ufd = *(u32 *)value;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2058)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2059) ptr = map->ops->map_fd_get_ptr(map, map_file, ufd);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2060) if (IS_ERR(ptr))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2061) return PTR_ERR(ptr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2062)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2063) ret = htab_map_update_elem(map, key, &ptr, map_flags);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2064) if (ret)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2065) map->ops->map_fd_put_ptr(ptr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2066)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2067) return ret;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2068) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2069)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2070) static struct bpf_map *htab_of_map_alloc(union bpf_attr *attr)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2071) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2072) struct bpf_map *map, *inner_map_meta;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2073)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2074) inner_map_meta = bpf_map_meta_alloc(attr->inner_map_fd);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2075) if (IS_ERR(inner_map_meta))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2076) return inner_map_meta;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2077)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2078) map = htab_map_alloc(attr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2079) if (IS_ERR(map)) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2080) bpf_map_meta_free(inner_map_meta);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2081) return map;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2082) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2083)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2084) map->inner_map_meta = inner_map_meta;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2085)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2086) return map;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2087) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2088)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2089) static void *htab_of_map_lookup_elem(struct bpf_map *map, void *key)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2090) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2091) struct bpf_map **inner_map = htab_map_lookup_elem(map, key);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2092)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2093) if (!inner_map)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2094) return NULL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2095)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2096) return READ_ONCE(*inner_map);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2097) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2098)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2099) static int htab_of_map_gen_lookup(struct bpf_map *map,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2100) struct bpf_insn *insn_buf)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2101) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2102) struct bpf_insn *insn = insn_buf;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2103) const int ret = BPF_REG_0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2104)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2105) BUILD_BUG_ON(!__same_type(&__htab_map_lookup_elem,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2106) (void *(*)(struct bpf_map *map, void *key))NULL));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2107) *insn++ = BPF_EMIT_CALL(BPF_CAST_CALL(__htab_map_lookup_elem));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2108) *insn++ = BPF_JMP_IMM(BPF_JEQ, ret, 0, 2);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2109) *insn++ = BPF_ALU64_IMM(BPF_ADD, ret,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2110) offsetof(struct htab_elem, key) +
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2111) round_up(map->key_size, 8));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2112) *insn++ = BPF_LDX_MEM(BPF_DW, ret, ret, 0);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2113)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2114) return insn - insn_buf;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2115) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2116)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2117) static void htab_of_map_free(struct bpf_map *map)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2118) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2119) bpf_map_meta_free(map->inner_map_meta);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2120) fd_htab_map_free(map);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2121) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2122)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2123) static int htab_of_maps_map_btf_id;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2124) const struct bpf_map_ops htab_of_maps_map_ops = {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2125) .map_alloc_check = fd_htab_map_alloc_check,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2126) .map_alloc = htab_of_map_alloc,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2127) .map_free = htab_of_map_free,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2128) .map_get_next_key = htab_map_get_next_key,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2129) .map_lookup_elem = htab_of_map_lookup_elem,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2130) .map_delete_elem = htab_map_delete_elem,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2131) .map_fd_get_ptr = bpf_map_fd_get_ptr,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2132) .map_fd_put_ptr = bpf_map_fd_put_ptr,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2133) .map_fd_sys_lookup_elem = bpf_map_fd_sys_lookup_elem,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2134) .map_gen_lookup = htab_of_map_gen_lookup,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2135) .map_check_btf = map_check_no_btf,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2136) .map_btf_name = "bpf_htab",
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2137) .map_btf_id = &htab_of_maps_map_btf_id,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2138) };
Orange Pi5 kernel
Deprecated Linux kernel 5.10.110 for OrangePi 5/5B/5+ boards
3 Commits
0 Branches
0 Tags