diff options
Diffstat (limited to 'kernel')
77 files changed, 1269 insertions, 941 deletions
diff --git a/kernel/bpf/hashtab.c b/kernel/bpf/hashtab.c index 04b8eda94e7d..03cc59ee9c95 100644 --- a/kernel/bpf/hashtab.c +++ b/kernel/bpf/hashtab.c @@ -15,6 +15,7 @@ #include <linux/jhash.h> #include <linux/filter.h> #include <linux/rculist_nulls.h> +#include <linux/random.h> #include <uapi/linux/btf.h> #include "percpu_freelist.h" #include "bpf_lru_list.h" @@ -41,6 +42,7 @@ struct bpf_htab { atomic_t count; /* number of elements in this hashtable */ u32 n_buckets; /* number of hash buckets */ u32 elem_size; /* size of each element in bytes */ + u32 hashrnd; }; /* each htab element is struct htab_elem + key + value */ @@ -371,6 +373,7 @@ static struct bpf_map *htab_map_alloc(union bpf_attr *attr) if (!htab->buckets) goto free_htab; + htab->hashrnd = get_random_int(); for (i = 0; i < htab->n_buckets; i++) { INIT_HLIST_NULLS_HEAD(&htab->buckets[i].head, i); raw_spin_lock_init(&htab->buckets[i].lock); @@ -402,9 +405,9 @@ free_htab: return ERR_PTR(err); } -static inline u32 htab_map_hash(const void *key, u32 key_len) +static inline u32 htab_map_hash(const void *key, u32 key_len, u32 hashrnd) { - return jhash(key, key_len, 0); + return jhash(key, key_len, hashrnd); } static inline struct bucket *__select_bucket(struct bpf_htab *htab, u32 hash) @@ -470,7 +473,7 @@ static void *__htab_map_lookup_elem(struct bpf_map *map, void *key) key_size = map->key_size; - hash = htab_map_hash(key, key_size); + hash = htab_map_hash(key, key_size, htab->hashrnd); head = select_bucket(htab, hash); @@ -597,7 +600,7 @@ static int htab_map_get_next_key(struct bpf_map *map, void *key, void *next_key) if (!key) goto find_first_elem; - hash = htab_map_hash(key, key_size); + hash = htab_map_hash(key, key_size, htab->hashrnd); head = select_bucket(htab, hash); @@ -824,7 +827,7 @@ static int htab_map_update_elem(struct bpf_map *map, void *key, void *value, key_size = map->key_size; - hash = htab_map_hash(key, key_size); + hash = htab_map_hash(key, key_size, htab->hashrnd); b = __select_bucket(htab, hash); head = &b->head; @@ -880,7 +883,7 @@ static int htab_lru_map_update_elem(struct bpf_map *map, void *key, void *value, key_size = map->key_size; - hash = htab_map_hash(key, key_size); + hash = htab_map_hash(key, key_size, htab->hashrnd); b = __select_bucket(htab, hash); head = &b->head; @@ -945,7 +948,7 @@ static int __htab_percpu_map_update_elem(struct bpf_map *map, void *key, key_size = map->key_size; - hash = htab_map_hash(key, key_size); + hash = htab_map_hash(key, key_size, htab->hashrnd); b = __select_bucket(htab, hash); head = &b->head; @@ -998,7 +1001,7 @@ static int __htab_lru_percpu_map_update_elem(struct bpf_map *map, void *key, key_size = map->key_size; - hash = htab_map_hash(key, key_size); + hash = htab_map_hash(key, key_size, htab->hashrnd); b = __select_bucket(htab, hash); head = &b->head; @@ -1071,7 +1074,7 @@ static int htab_map_delete_elem(struct bpf_map *map, void *key) key_size = map->key_size; - hash = htab_map_hash(key, key_size); + hash = htab_map_hash(key, key_size, htab->hashrnd); b = __select_bucket(htab, hash); head = &b->head; @@ -1103,7 +1106,7 @@ static int htab_lru_map_delete_elem(struct bpf_map *map, void *key) key_size = map->key_size; - hash = htab_map_hash(key, key_size); + hash = htab_map_hash(key, key_size, htab->hashrnd); b = __select_bucket(htab, hash); head = &b->head; diff --git a/kernel/bpf/sockmap.c b/kernel/bpf/sockmap.c index 98e621a29e8e..488ef9663c01 100644 --- a/kernel/bpf/sockmap.c +++ b/kernel/bpf/sockmap.c @@ -236,7 +236,7 @@ static int bpf_tcp_init(struct sock *sk) } static void smap_release_sock(struct smap_psock *psock, struct sock *sock); -static int free_start_sg(struct sock *sk, struct sk_msg_buff *md); +static int free_start_sg(struct sock *sk, struct sk_msg_buff *md, bool charge); static void bpf_tcp_release(struct sock *sk) { @@ -248,7 +248,7 @@ static void bpf_tcp_release(struct sock *sk) goto out; if (psock->cork) { - free_start_sg(psock->sock, psock->cork); + free_start_sg(psock->sock, psock->cork, true); kfree(psock->cork); psock->cork = NULL; } @@ -330,14 +330,14 @@ static void bpf_tcp_close(struct sock *sk, long timeout) close_fun = psock->save_close; if (psock->cork) { - free_start_sg(psock->sock, psock->cork); + free_start_sg(psock->sock, psock->cork, true); kfree(psock->cork); psock->cork = NULL; } list_for_each_entry_safe(md, mtmp, &psock->ingress, list) { list_del(&md->list); - free_start_sg(psock->sock, md); + free_start_sg(psock->sock, md, true); kfree(md); } @@ -369,7 +369,7 @@ static void bpf_tcp_close(struct sock *sk, long timeout) /* If another thread deleted this object skip deletion. * The refcnt on psock may or may not be zero. */ - if (l) { + if (l && l == link) { hlist_del_rcu(&link->hash_node); smap_release_sock(psock, link->sk); free_htab_elem(htab, link); @@ -570,14 +570,16 @@ static void free_bytes_sg(struct sock *sk, int bytes, md->sg_start = i; } -static int free_sg(struct sock *sk, int start, struct sk_msg_buff *md) +static int free_sg(struct sock *sk, int start, + struct sk_msg_buff *md, bool charge) { struct scatterlist *sg = md->sg_data; int i = start, free = 0; while (sg[i].length) { free += sg[i].length; - sk_mem_uncharge(sk, sg[i].length); + if (charge) + sk_mem_uncharge(sk, sg[i].length); if (!md->skb) put_page(sg_page(&sg[i])); sg[i].length = 0; @@ -594,9 +596,9 @@ static int free_sg(struct sock *sk, int start, struct sk_msg_buff *md) return free; } -static int free_start_sg(struct sock *sk, struct sk_msg_buff *md) +static int free_start_sg(struct sock *sk, struct sk_msg_buff *md, bool charge) { - int free = free_sg(sk, md->sg_start, md); + int free = free_sg(sk, md->sg_start, md, charge); md->sg_start = md->sg_end; return free; @@ -604,7 +606,7 @@ static int free_start_sg(struct sock *sk, struct sk_msg_buff *md) static int free_curr_sg(struct sock *sk, struct sk_msg_buff *md) { - return free_sg(sk, md->sg_curr, md); + return free_sg(sk, md->sg_curr, md, true); } static int bpf_map_msg_verdict(int _rc, struct sk_msg_buff *md) @@ -718,7 +720,7 @@ static int bpf_tcp_ingress(struct sock *sk, int apply_bytes, list_add_tail(&r->list, &psock->ingress); sk->sk_data_ready(sk); } else { - free_start_sg(sk, r); + free_start_sg(sk, r, true); kfree(r); } @@ -752,14 +754,10 @@ static int bpf_tcp_sendmsg_do_redirect(struct sock *sk, int send, release_sock(sk); } smap_release_sock(psock, sk); - if (unlikely(err)) - goto out; - return 0; + return err; out_rcu: rcu_read_unlock(); -out: - free_bytes_sg(NULL, send, md, false); - return err; + return 0; } static inline void bpf_md_init(struct smap_psock *psock) @@ -822,7 +820,7 @@ more_data: case __SK_PASS: err = bpf_tcp_push(sk, send, m, flags, true); if (unlikely(err)) { - *copied -= free_start_sg(sk, m); + *copied -= free_start_sg(sk, m, true); break; } @@ -845,16 +843,17 @@ more_data: lock_sock(sk); if (unlikely(err < 0)) { - free_start_sg(sk, m); + int free = free_start_sg(sk, m, false); + psock->sg_size = 0; if (!cork) - *copied -= send; + *copied -= free; } else { psock->sg_size -= send; } if (cork) { - free_start_sg(sk, m); + free_start_sg(sk, m, true); psock->sg_size = 0; kfree(m); m = NULL; @@ -912,6 +911,8 @@ static int bpf_tcp_recvmsg(struct sock *sk, struct msghdr *msg, size_t len, if (unlikely(flags & MSG_ERRQUEUE)) return inet_recv_error(sk, msg, len, addr_len); + if (!skb_queue_empty(&sk->sk_receive_queue)) + return tcp_recvmsg(sk, msg, len, nonblock, flags, addr_len); rcu_read_lock(); psock = smap_psock_sk(sk); @@ -922,9 +923,6 @@ static int bpf_tcp_recvmsg(struct sock *sk, struct msghdr *msg, size_t len, goto out; rcu_read_unlock(); - if (!skb_queue_empty(&sk->sk_receive_queue)) - return tcp_recvmsg(sk, msg, len, nonblock, flags, addr_len); - lock_sock(sk); bytes_ready: while (copied != len) { @@ -1122,7 +1120,7 @@ wait_for_memory: err = sk_stream_wait_memory(sk, &timeo); if (err) { if (m && m != psock->cork) - free_start_sg(sk, m); + free_start_sg(sk, m, true); goto out_err; } } @@ -1427,12 +1425,15 @@ out: static void smap_write_space(struct sock *sk) { struct smap_psock *psock; + void (*write_space)(struct sock *sk); rcu_read_lock(); psock = smap_psock_sk(sk); if (likely(psock && test_bit(SMAP_TX_RUNNING, &psock->state))) schedule_work(&psock->tx_work); + write_space = psock->save_write_space; rcu_read_unlock(); + write_space(sk); } static void smap_stop_sock(struct smap_psock *psock, struct sock *sk) @@ -1461,10 +1462,16 @@ static void smap_destroy_psock(struct rcu_head *rcu) schedule_work(&psock->gc_work); } +static bool psock_is_smap_sk(struct sock *sk) +{ + return inet_csk(sk)->icsk_ulp_ops == &bpf_tcp_ulp_ops; +} + static void smap_release_sock(struct smap_psock *psock, struct sock *sock) { if (refcount_dec_and_test(&psock->refcnt)) { - tcp_cleanup_ulp(sock); + if (psock_is_smap_sk(sock)) + tcp_cleanup_ulp(sock); write_lock_bh(&sock->sk_callback_lock); smap_stop_sock(psock, sock); write_unlock_bh(&sock->sk_callback_lock); @@ -1578,13 +1585,13 @@ static void smap_gc_work(struct work_struct *w) bpf_prog_put(psock->bpf_tx_msg); if (psock->cork) { - free_start_sg(psock->sock, psock->cork); + free_start_sg(psock->sock, psock->cork, true); kfree(psock->cork); } list_for_each_entry_safe(md, mtmp, &psock->ingress, list) { list_del(&md->list); - free_start_sg(psock->sock, md); + free_start_sg(psock->sock, md, true); kfree(md); } @@ -1891,6 +1898,10 @@ static int __sock_map_ctx_update_elem(struct bpf_map *map, * doesn't update user data. */ if (psock) { + if (!psock_is_smap_sk(sock)) { + err = -EBUSY; + goto out_progs; + } if (READ_ONCE(psock->bpf_parse) && parse) { err = -EBUSY; goto out_progs; @@ -2140,7 +2151,9 @@ static struct bpf_map *sock_hash_alloc(union bpf_attr *attr) return ERR_PTR(-EPERM); /* check sanity of attributes */ - if (attr->max_entries == 0 || attr->value_size != 4 || + if (attr->max_entries == 0 || + attr->key_size == 0 || + attr->value_size != 4 || attr->map_flags & ~SOCK_CREATE_FLAG_MASK) return ERR_PTR(-EINVAL); @@ -2267,8 +2280,10 @@ static struct htab_elem *alloc_sock_hash_elem(struct bpf_htab *htab, } l_new = kmalloc_node(htab->elem_size, GFP_ATOMIC | __GFP_NOWARN, htab->map.numa_node); - if (!l_new) + if (!l_new) { + atomic_dec(&htab->count); return ERR_PTR(-ENOMEM); + } memcpy(l_new->key, key, key_size); l_new->sk = sk; diff --git a/kernel/cgroup/cgroup-internal.h b/kernel/cgroup/cgroup-internal.h index 77ff1cd6a252..75568fcf2180 100644 --- a/kernel/cgroup/cgroup-internal.h +++ b/kernel/cgroup/cgroup-internal.h @@ -8,6 +8,32 @@ #include <linux/list.h> #include <linux/refcount.h> +#define TRACE_CGROUP_PATH_LEN 1024 +extern spinlock_t trace_cgroup_path_lock; +extern char trace_cgroup_path[TRACE_CGROUP_PATH_LEN]; + +/* + * cgroup_path() takes a spin lock. It is good practice not to take + * spin locks within trace point handlers, as they are mostly hidden + * from normal view. As cgroup_path() can take the kernfs_rename_lock + * spin lock, it is best to not call that function from the trace event + * handler. + * + * Note: trace_cgroup_##type##_enabled() is a static branch that will only + * be set when the trace event is enabled. + */ +#define TRACE_CGROUP_PATH(type, cgrp, ...) \ + do { \ + if (trace_cgroup_##type##_enabled()) { \ + spin_lock(&trace_cgroup_path_lock); \ + cgroup_path(cgrp, trace_cgroup_path, \ + TRACE_CGROUP_PATH_LEN); \ + trace_cgroup_##type(cgrp, trace_cgroup_path, \ + ##__VA_ARGS__); \ + spin_unlock(&trace_cgroup_path_lock); \ + } \ + } while (0) + /* * A cgroup can be associated with multiple css_sets as different tasks may * belong to different cgroups on different hierarchies. In the other diff --git a/kernel/cgroup/cgroup-v1.c b/kernel/cgroup/cgroup-v1.c index 8b4f0768efd6..51063e7a93c2 100644 --- a/kernel/cgroup/cgroup-v1.c +++ b/kernel/cgroup/cgroup-v1.c @@ -135,7 +135,7 @@ int cgroup_transfer_tasks(struct cgroup *to, struct cgroup *from) if (task) { ret = cgroup_migrate(task, false, &mgctx); if (!ret) - trace_cgroup_transfer_tasks(to, task, false); + TRACE_CGROUP_PATH(transfer_tasks, to, task, false); put_task_struct(task); } } while (task && !ret); @@ -865,7 +865,7 @@ static int cgroup1_rename(struct kernfs_node *kn, struct kernfs_node *new_parent ret = kernfs_rename(kn, new_parent, new_name_str); if (!ret) - trace_cgroup_rename(cgrp); + TRACE_CGROUP_PATH(rename, cgrp); mutex_unlock(&cgroup_mutex); diff --git a/kernel/cgroup/cgroup.c b/kernel/cgroup/cgroup.c index 35cf3d71f8aa..aae10baf1902 100644 --- a/kernel/cgroup/cgroup.c +++ b/kernel/cgroup/cgroup.c @@ -83,6 +83,9 @@ EXPORT_SYMBOL_GPL(cgroup_mutex); EXPORT_SYMBOL_GPL(css_set_lock); #endif +DEFINE_SPINLOCK(trace_cgroup_path_lock); +char trace_cgroup_path[TRACE_CGROUP_PATH_LEN]; + /* * Protects cgroup_idr and css_idr so that IDs can be released without * grabbing cgroup_mutex. @@ -2638,7 +2641,7 @@ int cgroup_attach_task(struct cgroup *dst_cgrp, struct task_struct *leader, cgroup_migrate_finish(&mgctx); if (!ret) - trace_cgroup_attach_task(dst_cgrp, leader, threadgroup); + TRACE_CGROUP_PATH(attach_task, dst_cgrp, leader, threadgroup); return ret; } @@ -4636,7 +4639,7 @@ static void css_release_work_fn(struct work_struct *work) struct cgroup *tcgrp; /* cgroup release path */ - trace_cgroup_release(cgrp); + TRACE_CGROUP_PATH(release, cgrp); if (cgroup_on_dfl(cgrp)) cgroup_rstat_flush(cgrp); @@ -4979,7 +4982,7 @@ int cgroup_mkdir(struct kernfs_node *parent_kn, const char *name, umode_t mode) if (ret) goto out_destroy; |