From 318892ac068397f40ff81d9155898da01493b1d2 Mon Sep 17 00:00:00 2001 From: Jakub Kicinski Date: Fri, 19 Jul 2019 10:29:14 -0700 Subject: net/tls: don't arm strparser immediately in tls_set_sw_offload() In tls_set_device_offload_rx() we prepare the software context for RX fallback and proceed to add the connection to the device. Unfortunately, software context prep includes arming strparser so in case of a later error we have to release the socket lock to call strp_done(). In preparation for not releasing the socket lock half way through callbacks move arming strparser into a separate function. Following patches will make use of that. Signed-off-by: Jakub Kicinski Reviewed-by: Dirk van der Merwe Signed-off-by: Daniel Borkmann --- include/net/tls.h | 1 + 1 file changed, 1 insertion(+) (limited to 'include') diff --git a/include/net/tls.h b/include/net/tls.h index 584609174fe0..43f551cd508b 100644 --- a/include/net/tls.h +++ b/include/net/tls.h @@ -355,6 +355,7 @@ int tls_sk_attach(struct sock *sk, int optname, char __user *optval, unsigned int optlen); int tls_set_sw_offload(struct sock *sk, struct tls_context *ctx, int tx); +void tls_sw_strparser_arm(struct sock *sk, struct tls_context *ctx); int tls_sw_sendmsg(struct sock *sk, struct msghdr *msg, size_t size); int tls_sw_sendpage(struct sock *sk, struct page *page, int offset, size_t size, int flags); -- cgit v1.2.3 From f87e62d45e51b12d48d2cb46b5cde8f83b866bc4 Mon Sep 17 00:00:00 2001 From: John Fastabend Date: Fri, 19 Jul 2019 10:29:16 -0700 Subject: net/tls: remove close callback sock unlock/lock around TX work flush The tls close() callback currently drops the sock lock, makes a cancel_delayed_work_sync() call, and then relocks the sock. By restructuring the code we can avoid droping lock and then reclaiming it. To simplify this we do the following, tls_sk_proto_close set_bit(CLOSING) set_bit(SCHEDULE) cancel_delay_work_sync() <- cancel workqueue lock_sock(sk) ... release_sock(sk) strp_done() Setting the CLOSING bit prevents the SCHEDULE bit from being cleared by any workqueue items e.g. if one happens to be scheduled and run between when we set SCHEDULE bit and cancel work. Then because SCHEDULE bit is set now no new work will be scheduled. Tested with net selftests and bpf selftests. Signed-off-by: John Fastabend Signed-off-by: Jakub Kicinski Reviewed-by: Dirk van der Merwe Signed-off-by: Daniel Borkmann --- include/net/tls.h | 2 ++ 1 file changed, 2 insertions(+) (limited to 'include') diff --git a/include/net/tls.h b/include/net/tls.h index 43f551cd508b..d4276cb6de53 100644 --- a/include/net/tls.h +++ b/include/net/tls.h @@ -162,6 +162,7 @@ struct tls_sw_context_tx { int async_capable; #define BIT_TX_SCHEDULED 0 +#define BIT_TX_CLOSING 1 unsigned long tx_bitmask; }; @@ -360,6 +361,7 @@ int tls_sw_sendmsg(struct sock *sk, struct msghdr *msg, size_t size); int tls_sw_sendpage(struct sock *sk, struct page *page, int offset, size_t size, int flags); void tls_sw_close(struct sock *sk, long timeout); +void tls_sw_cancel_work_tx(struct tls_context *tls_ctx); void tls_sw_free_resources_tx(struct sock *sk); void tls_sw_free_resources_rx(struct sock *sk); void tls_sw_release_resources_rx(struct sock *sk); -- cgit v1.2.3 From 313ab004805cf52a42673b15852b3842474ccd87 Mon Sep 17 00:00:00 2001 From: John Fastabend Date: Fri, 19 Jul 2019 10:29:17 -0700 Subject: net/tls: remove sock unlock/lock around strp_done() The tls close() callback currently drops the sock lock to call strp_done(). Split up the RX cleanup into stopping the strparser and releasing most resources, syncing strparser and finally freeing the context. To avoid the need for a strp_done() call on the cleanup path of device offload make sure we don't arm the strparser until we are sure init will be successful. Signed-off-by: John Fastabend Signed-off-by: Jakub Kicinski Reviewed-by: Dirk van der Merwe Signed-off-by: Daniel Borkmann --- include/net/tls.h | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) (limited to 'include') diff --git a/include/net/tls.h b/include/net/tls.h index d4276cb6de53..235508e35fd4 100644 --- a/include/net/tls.h +++ b/include/net/tls.h @@ -107,9 +107,7 @@ struct tls_device { enum { TLS_BASE, TLS_SW, -#ifdef CONFIG_TLS_DEVICE TLS_HW, -#endif TLS_HW_RECORD, TLS_NUM_CONFIG, }; @@ -357,14 +355,17 @@ int tls_sk_attach(struct sock *sk, int optname, char __user *optval, int tls_set_sw_offload(struct sock *sk, struct tls_context *ctx, int tx); void tls_sw_strparser_arm(struct sock *sk, struct tls_context *ctx); +void tls_sw_strparser_done(struct tls_context *tls_ctx); int tls_sw_sendmsg(struct sock *sk, struct msghdr *msg, size_t size); int tls_sw_sendpage(struct sock *sk, struct page *page, int offset, size_t size, int flags); void tls_sw_close(struct sock *sk, long timeout); void tls_sw_cancel_work_tx(struct tls_context *tls_ctx); -void tls_sw_free_resources_tx(struct sock *sk); +void tls_sw_release_resources_tx(struct sock *sk); +void tls_sw_free_ctx_tx(struct tls_context *tls_ctx); void tls_sw_free_resources_rx(struct sock *sk); void tls_sw_release_resources_rx(struct sock *sk); +void tls_sw_free_ctx_rx(struct tls_context *tls_ctx); int tls_sw_recvmsg(struct sock *sk, struct msghdr *msg, size_t len, int nonblock, int flags, int *addr_len); bool tls_sw_stream_read(const struct sock *sk); -- cgit v1.2.3 From 32857cf57f920cdc03b5095f08febec94cf9c36b Mon Sep 17 00:00:00 2001 From: John Fastabend Date: Fri, 19 Jul 2019 10:29:18 -0700 Subject: net/tls: fix transition through disconnect with close It is possible (via shutdown()) for TCP socks to go through TCP_CLOSE state via tcp_disconnect() without actually calling tcp_close which would then call the tls close callback. Because of this a user could disconnect a socket then put it in a LISTEN state which would break our assumptions about sockets always being ESTABLISHED state. More directly because close() can call unhash() and unhash is implemented by sockmap if a sockmap socket has TLS enabled we can incorrectly destroy the psock from unhash() and then call its close handler again. But because the psock (sockmap socket representation) is already destroyed we call close handler in sk->prot. However, in some cases (TLS BASE/BASE case) this will still point at the sockmap close handler resulting in a circular call and crash reported by syzbot. To fix both above issues implement the unhash() routine for TLS. v4: - add note about tls offload still needing the fix; - move sk_proto to the cold cache line; - split TX context free into "release" and "free", otherwise the GC work itself is in already freed memory; - more TX before RX for consistency; - reuse tls_ctx_free(); - schedule the GC work after we're done with context to avoid UAF; - don't set the unhash in all modes, all modes "inherit" TLS_BASE's callbacks anyway; - disable the unhash hook for TLS_HW. Fixes: 3c4d7559159bf ("tls: kernel TLS support") Reported-by: Eric Dumazet Signed-off-by: John Fastabend Signed-off-by: Jakub Kicinski Signed-off-by: Daniel Borkmann --- include/net/tls.h | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) (limited to 'include') diff --git a/include/net/tls.h b/include/net/tls.h index 235508e35fd4..9e425ac2de45 100644 --- a/include/net/tls.h +++ b/include/net/tls.h @@ -271,6 +271,8 @@ struct tls_context { unsigned long flags; /* cache cold stuff */ + struct proto *sk_proto; + void (*sk_destruct)(struct sock *sk); void (*sk_proto_close)(struct sock *sk, long timeout); @@ -288,6 +290,8 @@ struct tls_context { struct list_head list; refcount_t refcount; + + struct work_struct gc; }; enum tls_offload_ctx_dir { @@ -359,7 +363,6 @@ void tls_sw_strparser_done(struct tls_context *tls_ctx); int tls_sw_sendmsg(struct sock *sk, struct msghdr *msg, size_t size); int tls_sw_sendpage(struct sock *sk, struct page *page, int offset, size_t size, int flags); -void tls_sw_close(struct sock *sk, long timeout); void tls_sw_cancel_work_tx(struct tls_context *tls_ctx); void tls_sw_release_resources_tx(struct sock *sk); void tls_sw_free_ctx_tx(struct tls_context *tls_ctx); -- cgit v1.2.3 From 95fa145479fbc0a0c1fd3274ceb42ec03c042a4a Mon Sep 17 00:00:00 2001 From: John Fastabend Date: Fri, 19 Jul 2019 10:29:22 -0700 Subject: bpf: sockmap/tls, close can race with map free When a map free is called and in parallel a socket is closed we have two paths that can potentially reset the socket prot ops, the bpf close() path and the map free path. This creates a problem with which prot ops should be used from the socket closed side. If the map_free side completes first then we want to call the original lowest level ops. However, if the tls path runs first we want to call the sockmap ops. Additionally there was no locking around prot updates in TLS code paths so the prot ops could be changed multiple times once from TLS path and again from sockmap side potentially leaving ops pointed at either TLS or sockmap when psock and/or tls context have already been destroyed. To fix this race first only update ops inside callback lock so that TLS, sockmap and lowest level all agree on prot state. Second and a ULP callback update() so that lower layers can inform the upper layer when they are being removed allowing the upper layer to reset prot ops. This gets us close to allowing sockmap and tls to be stacked in arbitrary order but will save that patch for *next trees. v4: - make sure we don't free things for device; - remove the checks which swap the callbacks back only if TLS is at the top. Reported-by: syzbot+06537213db7ba2745c4a@syzkaller.appspotmail.com Fixes: 02c558b2d5d6 ("bpf: sockmap, support for msg_peek in sk_msg with redirect ingress") Signed-off-by: John Fastabend Signed-off-by: Jakub Kicinski Reviewed-by: Dirk van der Merwe Signed-off-by: Daniel Borkmann --- include/linux/skmsg.h | 8 +++++++- include/net/tcp.h | 3 +++ 2 files changed, 10 insertions(+), 1 deletion(-) (limited to 'include') diff --git a/include/linux/skmsg.h b/include/linux/skmsg.h index 50ced8aba9db..e4b3fb4bb77c 100644 --- a/include/linux/skmsg.h +++ b/include/linux/skmsg.h @@ -354,7 +354,13 @@ static inline void sk_psock_restore_proto(struct sock *sk, sk->sk_write_space = psock->saved_write_space; if (psock->sk_proto) { - sk->sk_prot = psock->sk_proto; + struct inet_connection_sock *icsk = inet_csk(sk); + bool has_ulp = !!icsk->icsk_ulp_data; + + if (has_ulp) + tcp_update_ulp(sk, psock->sk_proto); + else + sk->sk_prot = psock->sk_proto; psock->sk_proto = NULL; } } diff --git a/include/net/tcp.h b/include/net/tcp.h index f42d300f0cfa..c82a23470081 100644 --- a/include/net/tcp.h +++ b/include/net/tcp.h @@ -2103,6 +2103,8 @@ struct tcp_ulp_ops { /* initialize ulp */ int (*init)(struct sock *sk); + /* update ulp */ + void (*update)(struct sock *sk, struct proto *p); /* cleanup ulp */ void (*release)(struct sock *sk); @@ -2114,6 +2116,7 @@ void tcp_unregister_ulp(struct tcp_ulp_ops *type); int tcp_set_ulp(struct sock *sk, const char *name); void tcp_get_available_ulp(char *buf, size_t len); void tcp_cleanup_ulp(struct sock *sk); +void tcp_update_ulp(struct sock *sk, struct proto *p); #define MODULE_ALIAS_TCP_ULP(name) \ __MODULE_INFO(alias, alias_userspace, name); \ -- cgit v1.2.3 From d9b8aadaffa65809d146cf0f8632a22a946367d7 Mon Sep 17 00:00:00 2001 From: Ilya Leoshkevich Date: Fri, 19 Jul 2019 11:18:15 +0200 Subject: bpf: fix narrower loads on s390 The very first check in test_pkt_md_access is failing on s390, which happens because loading a part of a struct __sk_buff field produces an incorrect result. The preprocessed code of the check is: { __u8 tmp = *((volatile __u8 *)&skb->len + ((sizeof(skb->len) - sizeof(__u8)) / sizeof(__u8))); if (tmp != ((*(volatile __u32 *)&skb->len) & 0xFF)) return 2; }; clang generates the following code for it: 0: 71 21 00 03 00 00 00 00 r2 = *(u8 *)(r1 + 3) 1: 61 31 00 00 00 00 00 00 r3 = *(u32 *)(r1 + 0) 2: 57 30 00 00 00 00 00 ff r3 &= 255 3: 5d 23 00 1d 00 00 00 00 if r2 != r3 goto +29 Finally, verifier transforms it to: 0: (61) r2 = *(u32 *)(r1 +104) 1: (bc) w2 = w2 2: (74) w2 >>= 24 3: (bc) w2 = w2 4: (54) w2 &= 255 5: (bc) w2 = w2 The problem is that when verifier emits the code to replace a partial load of a struct __sk_buff field (*(u8 *)(r1 + 3)) with a full load of struct sk_buff field (*(u32 *)(r1 + 104)), an optional shift and a bitwise AND, it assumes that the machine is little endian and incorrectly decides to use a shift. Adjust shift count calculation to account for endianness. Fixes: 31fd85816dbe ("bpf: permits narrower load from bpf program context fields") Signed-off-by: Ilya Leoshkevich Signed-off-by: Alexei Starovoitov --- include/linux/filter.h | 13 +++++++++++++ 1 file changed, 13 insertions(+) (limited to 'include') diff --git a/include/linux/filter.h b/include/linux/filter.h index ff65d22cf336..92c6e31fb008 100644 --- a/include/linux/filter.h +++ b/include/linux/filter.h @@ -24,6 +24,7 @@ #include +#include #include #include @@ -747,6 +748,18 @@ bpf_ctx_narrow_access_ok(u32 off, u32 size, u32 size_default) return size <= size_default && (size & (size - 1)) == 0; } +static inline u8 +bpf_ctx_narrow_load_shift(u32 off, u32 size, u32 size_default) +{ + u8 load_off = off & (size_default - 1); + +#ifdef __LITTLE_ENDIAN + return load_off * 8; +#else + return (size_default - (load_off + size)) * 8; +#endif +} + #define bpf_ctx_wide_access_ok(off, size, type, field) \ (size == sizeof(__u64) && \ off >= offsetof(type, field) && \ -- cgit v1.2.3