From 53615ad26e9789bfcdf3a4dccbcecb15294ea024 Mon Sep 17 00:00:00 2001 From: Byungchul Park Date: Mon, 13 Oct 2025 13:41:33 +0900 Subject: netmem: replace __netmem_clear_lsb() with netmem_to_nmdesc() Now that we have struct netmem_desc, it'd better access the pp fields via struct netmem_desc rather than struct net_iov. Introduce netmem_to_nmdesc() for safely converting netmem_ref to netmem_desc regardless of the type underneath e.i. netmem_desc, net_iov. While at it, remove __netmem_clear_lsb() and make netmem_to_nmdesc() used instead. Suggested-by: Pavel Begunkov Signed-off-by: Byungchul Park Reviewed-by: Mina Almasry Link: https://patch.msgid.link/20251013044133.69472-1-byungchul@sk.com Signed-off-by: Paolo Abeni --- include/net/netmem.h | 66 ++++++++++++++++++++++++++-------------------------- 1 file changed, 33 insertions(+), 33 deletions(-) (limited to 'include/net') diff --git a/include/net/netmem.h b/include/net/netmem.h index f7dacc9e75fd..651e2c62d1dd 100644 --- a/include/net/netmem.h +++ b/include/net/netmem.h @@ -247,6 +247,23 @@ static inline unsigned long netmem_pfn_trace(netmem_ref netmem) return page_to_pfn(netmem_to_page(netmem)); } +/* XXX: How to extract netmem_desc from page must be changed, once + * netmem_desc no longer overlays on page and will be allocated through + * slab. + */ +#define __pp_page_to_nmdesc(p) (_Generic((p), \ + const struct page * : (const struct netmem_desc *)(p), \ + struct page * : (struct netmem_desc *)(p))) + +/* CAUTION: Check if the page is a pp page before calling this helper or + * know it's a pp page. + */ +#define pp_page_to_nmdesc(p) \ +({ \ + DEBUG_NET_WARN_ON_ONCE(!page_pool_page_is_pp(p)); \ + __pp_page_to_nmdesc(p); \ +}) + /** * __netmem_to_nmdesc - unsafely get pointer to the &netmem_desc backing * @netmem @@ -265,42 +282,25 @@ static inline struct netmem_desc *__netmem_to_nmdesc(netmem_ref netmem) return (__force struct netmem_desc *)netmem; } -/* __netmem_clear_lsb - convert netmem_ref to struct net_iov * for access to - * common fields. - * @netmem: netmem reference to extract as net_iov. - * - * All the sub types of netmem_ref (page, net_iov) have the same pp, pp_magic, - * dma_addr, and pp_ref_count fields at the same offsets. Thus, we can access - * these fields without a type check to make sure that the underlying mem is - * net_iov or page. +/* netmem_to_nmdesc - convert netmem_ref to struct netmem_desc * for + * access to common fields. + * @netmem: netmem reference to get netmem_desc. * - * The resulting value of this function can only be used to access the fields - * that are NET_IOV_ASSERT_OFFSET'd. Accessing any other fields will result in - * undefined behavior. + * All the sub types of netmem_ref (netmem_desc, net_iov) have the same + * pp, pp_magic, dma_addr, and pp_ref_count fields via netmem_desc. * - * Return: the netmem_ref cast to net_iov* regardless of its underlying type. + * Return: the pointer to struct netmem_desc * regardless of its + * underlying type. */ -static inline struct net_iov *__netmem_clear_lsb(netmem_ref netmem) +static inline struct netmem_desc *netmem_to_nmdesc(netmem_ref netmem) { - return (struct net_iov *)((__force unsigned long)netmem & ~NET_IOV); -} + void *p = (void *)((__force unsigned long)netmem & ~NET_IOV); -/* XXX: How to extract netmem_desc from page must be changed, once - * netmem_desc no longer overlays on page and will be allocated through - * slab. - */ -#define __pp_page_to_nmdesc(p) (_Generic((p), \ - const struct page * : (const struct netmem_desc *)(p), \ - struct page * : (struct netmem_desc *)(p))) + if (netmem_is_net_iov(netmem)) + return &((struct net_iov *)p)->desc; -/* CAUTION: Check if the page is a pp page before calling this helper or - * know it's a pp page. - */ -#define pp_page_to_nmdesc(p) \ -({ \ - DEBUG_NET_WARN_ON_ONCE(!page_pool_page_is_pp(p)); \ - __pp_page_to_nmdesc(p); \ -}) + return __pp_page_to_nmdesc((struct page *)p); +} /** * __netmem_get_pp - unsafely get pointer to the &page_pool backing @netmem @@ -320,12 +320,12 @@ static inline struct page_pool *__netmem_get_pp(netmem_ref netmem) static inline struct page_pool *netmem_get_pp(netmem_ref netmem) { - return __netmem_clear_lsb(netmem)->pp; + return netmem_to_nmdesc(netmem)->pp; } static inline atomic_long_t *netmem_get_pp_ref_count_ref(netmem_ref netmem) { - return &__netmem_clear_lsb(netmem)->pp_ref_count; + return &netmem_to_nmdesc(netmem)->pp_ref_count; } static inline bool netmem_is_pref_nid(netmem_ref netmem, int pref_nid) @@ -390,7 +390,7 @@ static inline bool netmem_is_pfmemalloc(netmem_ref netmem) static inline unsigned long netmem_get_dma_addr(netmem_ref netmem) { - return __netmem_clear_lsb(netmem)->dma_addr; + return netmem_to_nmdesc(netmem)->dma_addr; } void get_netmem(netmem_ref netmem); -- cgit v1.2.3 From 1c51450f1afff1e7419797720df3fbd9ccbf610c Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Mon, 13 Oct 2025 14:59:26 +0000 Subject: tcp: better handle TCP_TX_DELAY on established flows Some applications uses TCP_TX_DELAY socket option after TCP flow is established. Some metrics need to be updated, otherwise TCP might take time to adapt to the new (emulated) RTT. This patch adjusts tp->srtt_us, tp->rtt_min, icsk_rto and sk->sk_pacing_rate. This is best effort, and for instance icsk_rto is reset without taking backoff into account. Signed-off-by: Eric Dumazet Link: https://patch.msgid.link/20251013145926.833198-1-edumazet@google.com Signed-off-by: Jakub Kicinski --- include/net/tcp.h | 2 ++ net/ipv4/tcp.c | 31 +++++++++++++++++++++++++++---- net/ipv4/tcp_input.c | 4 ++-- 3 files changed, 31 insertions(+), 6 deletions(-) (limited to 'include/net') diff --git a/include/net/tcp.h b/include/net/tcp.h index 5ca230ed526a..1e547138f4fb 100644 --- a/include/net/tcp.h +++ b/include/net/tcp.h @@ -461,6 +461,8 @@ enum skb_drop_reason tcp_child_process(struct sock *parent, struct sock *child, void tcp_enter_loss(struct sock *sk); void tcp_cwnd_reduction(struct sock *sk, int newly_acked_sacked, int newly_lost, int flag); void tcp_clear_retrans(struct tcp_sock *tp); +void tcp_update_pacing_rate(struct sock *sk); +void tcp_set_rto(struct sock *sk); void tcp_update_metrics(struct sock *sk); void tcp_init_metrics(struct sock *sk); void tcp_metrics_init(void); diff --git a/net/ipv4/tcp.c b/net/ipv4/tcp.c index 8a18aeca7ab0..4d720aa09a4c 100644 --- a/net/ipv4/tcp.c +++ b/net/ipv4/tcp.c @@ -3583,9 +3583,12 @@ static int tcp_repair_options_est(struct sock *sk, sockptr_t optbuf, DEFINE_STATIC_KEY_FALSE(tcp_tx_delay_enabled); EXPORT_IPV6_MOD(tcp_tx_delay_enabled); -static void tcp_enable_tx_delay(void) +static void tcp_enable_tx_delay(struct sock *sk, int val) { - if (!static_branch_unlikely(&tcp_tx_delay_enabled)) { + struct tcp_sock *tp = tcp_sk(sk); + s32 delta = (val - tp->tcp_tx_delay) << 3; + + if (val && !static_branch_unlikely(&tcp_tx_delay_enabled)) { static int __tcp_tx_delay_enabled = 0; if (cmpxchg(&__tcp_tx_delay_enabled, 0, 1) == 0) { @@ -3593,6 +3596,22 @@ static void tcp_enable_tx_delay(void) pr_info("TCP_TX_DELAY enabled\n"); } } + /* If we change tcp_tx_delay on a live flow, adjust tp->srtt_us, + * tp->rtt_min, icsk_rto and sk->sk_pacing_rate. + * This is best effort. + */ + if (delta && sk->sk_state == TCP_ESTABLISHED) { + s64 srtt = (s64)tp->srtt_us + delta; + + tp->srtt_us = clamp_t(s64, srtt, 1, ~0U); + + /* Note: does not deal with non zero icsk_backoff */ + tcp_set_rto(sk); + + minmax_reset(&tp->rtt_min, tcp_jiffies32, ~0U); + + tcp_update_pacing_rate(sk); + } } /* When set indicates to always queue non-full frames. Later the user clears @@ -4119,8 +4138,12 @@ ao_parse: tp->recvmsg_inq = val; break; case TCP_TX_DELAY: - if (val) - tcp_enable_tx_delay(); + /* tp->srtt_us is u32, and is shifted by 3 */ + if (val < 0 || val >= (1U << (31 - 3))) { + err = -EINVAL; + break; + } + tcp_enable_tx_delay(sk, val); WRITE_ONCE(tp->tcp_tx_delay, val); break; default: diff --git a/net/ipv4/tcp_input.c b/net/ipv4/tcp_input.c index 31ea5af49f2d..8fc97f4d8a6b 100644 --- a/net/ipv4/tcp_input.c +++ b/net/ipv4/tcp_input.c @@ -1095,7 +1095,7 @@ static void tcp_rtt_estimator(struct sock *sk, long mrtt_us) tp->srtt_us = max(1U, srtt); } -static void tcp_update_pacing_rate(struct sock *sk) +void tcp_update_pacing_rate(struct sock *sk) { const struct tcp_sock *tp = tcp_sk(sk); u64 rate; @@ -1132,7 +1132,7 @@ static void tcp_update_pacing_rate(struct sock *sk) /* Calculate rto without backoff. This is the second half of Van Jacobson's * routine referred to above. */ -static void tcp_set_rto(struct sock *sk) +void tcp_set_rto(struct sock *sk) { const struct tcp_sock *tp = tcp_sk(sk); /* Old crap is replaced with new one. 8) -- cgit v1.2.3 From 6ddb811a579f87b8506344020002d396f814f7c8 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Mon, 13 Oct 2025 15:22:31 +0000 Subject: net: add SK_WMEM_ALLOC_BIAS constant sk->sk_wmem_alloc is initialized to 1, and sk_wmem_alloc_get() takes care of this initial value. Add SK_WMEM_ALLOC_BIAS define to not spread this magic value. Signed-off-by: Eric Dumazet Reviewed-by: Neal Cardwell Reviewed-by: Kuniyuki Iwashima Link: https://patch.msgid.link/20251013152234.842065-2-edumazet@google.com Signed-off-by: Jakub Kicinski --- include/net/sock.h | 3 ++- net/atm/common.c | 2 +- net/core/sock.c | 5 ++--- 3 files changed, 5 insertions(+), 5 deletions(-) (limited to 'include/net') diff --git a/include/net/sock.h b/include/net/sock.h index 60bcb13f045c..2794bc5c5654 100644 --- a/include/net/sock.h +++ b/include/net/sock.h @@ -2303,6 +2303,7 @@ static inline int skb_copy_to_page_nocache(struct sock *sk, struct iov_iter *fro return 0; } +#define SK_WMEM_ALLOC_BIAS 1 /** * sk_wmem_alloc_get - returns write allocations * @sk: socket @@ -2311,7 +2312,7 @@ static inline int skb_copy_to_page_nocache(struct sock *sk, struct iov_iter *fro */ static inline int sk_wmem_alloc_get(const struct sock *sk) { - return refcount_read(&sk->sk_wmem_alloc) - 1; + return refcount_read(&sk->sk_wmem_alloc) - SK_WMEM_ALLOC_BIAS; } /** diff --git a/net/atm/common.c b/net/atm/common.c index 881c7f259dbd..cecc71a8bee1 100644 --- a/net/atm/common.c +++ b/net/atm/common.c @@ -157,7 +157,7 @@ int vcc_create(struct net *net, struct socket *sock, int protocol, int family, i memset(&vcc->local, 0, sizeof(struct sockaddr_atmsvc)); memset(&vcc->remote, 0, sizeof(struct sockaddr_atmsvc)); vcc->qos.txtp.max_sdu = 1 << 16; /* for meta VCs */ - refcount_set(&sk->sk_wmem_alloc, 1); + refcount_set(&sk->sk_wmem_alloc, SK_WMEM_ALLOC_BIAS); atomic_set(&sk->sk_rmem_alloc, 0); vcc->push = NULL; vcc->pop = NULL; diff --git a/net/core/sock.c b/net/core/sock.c index dc03d4b5909a..542cfa16ee12 100644 --- a/net/core/sock.c +++ b/net/core/sock.c @@ -2313,7 +2313,7 @@ struct sock *sk_alloc(struct net *net, int family, gfp_t priority, } sock_net_set(sk, net); - refcount_set(&sk->sk_wmem_alloc, 1); + refcount_set(&sk->sk_wmem_alloc, SK_WMEM_ALLOC_BIAS); mem_cgroup_sk_alloc(sk); cgroup_sk_alloc(&sk->sk_cgrp_data); @@ -2494,8 +2494,7 @@ struct sock *sk_clone_lock(const struct sock *sk, const gfp_t priority) atomic_set(&newsk->sk_rmem_alloc, 0); - /* sk_wmem_alloc set to one (see sk_free() and sock_wfree()) */ - refcount_set(&newsk->sk_wmem_alloc, 1); + refcount_set(&newsk->sk_wmem_alloc, SK_WMEM_ALLOC_BIAS); atomic_set(&newsk->sk_omem_alloc, 0); sk_init_common(newsk); -- cgit v1.2.3 From 2ddef3462b3a5d62e5485e22ce128a5c02276438 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Mon, 13 Oct 2025 15:22:33 +0000 Subject: net: add /proc/sys/net/core/txq_reselection_ms control Add a new sysctl to control how often a queue reselection can happen even if a flow has a persistent queue of skbs in a Qdisc or NIC queue. A value of zero means the feature is disabled. Default is 1000 (1 second). This sysctl is used in the following patch. Signed-off-by: Eric Dumazet Reviewed-by: Neal Cardwell Reviewed-by: Kuniyuki Iwashima Link: https://patch.msgid.link/20251013152234.842065-4-edumazet@google.com Signed-off-by: Jakub Kicinski --- Documentation/admin-guide/sysctl/net.rst | 17 +++++++++++++++++ include/net/netns/core.h | 1 + net/core/net_namespace.c | 1 + net/core/sysctl_net_core.c | 7 +++++++ 4 files changed, 26 insertions(+) (limited to 'include/net') diff --git a/Documentation/admin-guide/sysctl/net.rst b/Documentation/admin-guide/sysctl/net.rst index 2ef50828aff1..40749b3cd356 100644 --- a/Documentation/admin-guide/sysctl/net.rst +++ b/Documentation/admin-guide/sysctl/net.rst @@ -406,6 +406,23 @@ to SOCK_TXREHASH_DEFAULT (i. e. not overridden by setsockopt). If set to 1 (default), hash rethink is performed on listening socket. If set to 0, hash rethink is not performed. +txq_reselection_ms +------------------ + +Controls how often (in ms) a busy connected flow can select another tx queue. + +A resection is desirable when/if user thread has migrated and XPS +would select a different queue. Same can occur without XPS +if the flow hash has changed. + +But switching txq can introduce reorders, especially if the +old queue is under high pressure. Modern TCP stacks deal +well with reorders if they happen not too often. + +To disable this feature, set the value to 0. + +Default : 1000 + gro_normal_batch ---------------- diff --git a/include/net/netns/core.h b/include/net/netns/core.h index 9b36f0ff0c20..cb9c3e4cd738 100644 --- a/include/net/netns/core.h +++ b/include/net/netns/core.h @@ -13,6 +13,7 @@ struct netns_core { struct ctl_table_header *sysctl_hdr; int sysctl_somaxconn; + int sysctl_txq_reselection; int sysctl_optmem_max; u8 sysctl_txrehash; u8 sysctl_tstamp_allow_data; diff --git a/net/core/net_namespace.c b/net/core/net_namespace.c index b0e0f22d7b21..adcfef55a66f 100644 --- a/net/core/net_namespace.c +++ b/net/core/net_namespace.c @@ -395,6 +395,7 @@ static __net_init void preinit_net_sysctl(struct net *net) net->core.sysctl_optmem_max = 128 * 1024; net->core.sysctl_txrehash = SOCK_TXREHASH_ENABLED; net->core.sysctl_tstamp_allow_data = 1; + net->core.sysctl_txq_reselection = msecs_to_jiffies(1000); } /* init code that must occur even if setup_net() is not called. */ diff --git a/net/core/sysctl_net_core.c b/net/core/sysctl_net_core.c index 8cf04b57ade1..f79137826d7f 100644 --- a/net/core/sysctl_net_core.c +++ b/net/core/sysctl_net_core.c @@ -667,6 +667,13 @@ static struct ctl_table netns_core_table[] = { .extra2 = SYSCTL_ONE, .proc_handler = proc_dou8vec_minmax, }, + { + .procname = "txq_reselection_ms", + .data = &init_net.core.sysctl_txq_reselection, + .maxlen = sizeof(int), + .mode = 0644, + .proc_handler = proc_dointvec_ms_jiffies, + }, { .procname = "tstamp_allow_data", .data = &init_net.core.sysctl_tstamp_allow_data, -- cgit v1.2.3 From 4a7708443dec13b074bc43855f494358fedbd3c0 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Mon, 13 Oct 2025 15:22:34 +0000 Subject: net: allow busy connected flows to switch tx queues This is a followup of commit 726e9e8b94b9 ("tcp: refine skb->ooo_okay setting") and of prior commit in this series ("net: control skb->ooo_okay from skb_set_owner_w()") skb->ooo_okay might never be set for bulk flows that always have at least one skb in a qdisc queue of NIC queue, especially if TX completion is delayed because of a stressed cpu. The so-called "strange attractors" has caused many performance issues (see for instance 9b462d02d6dd ("tcp: TCP Small Queues and strange attractors")), we need to do better. We have tried very hard to avoid reorders because TCP was not dealing with them nicely a decade ago. Use the new net.core.txq_reselection_ms sysctl to let flows follow XPS and select a more efficient queue. After this patch, we no longer have to make sure threads are pinned to cpus, they now can be migrated without adding too much spinlock/qdisc/TX completion pressure anymore. TX completion part was problematic, because it added false sharing on various socket fields, but also added false sharing and spinlock contention in mm layers. Calling skb_orphan() from ndo_start_xmit() is not an option unfortunately. Note for later: 1) move sk->sk_tx_queue_mapping closer to sk_tx_queue_mapping_jiffies for better cache locality. 2) Study if 9b462d02d6dd ("tcp: TCP Small Queues and strange attractors") could be revised. Tested: Used a host with 32 TX queues, shared by groups of 8 cores. XPS setup : echo ff >/sys/class/net/eth1/queue/tx-0/xps_cpus echo ff00 >/sys/class/net/eth1/queue/tx-1/xps_cpus echo ff0000 >/sys/class/net/eth1/queue/tx-2/xps_cpus echo ff000000 >/sys/class/net/eth1/queue/tx-3/xps_cpus echo ff,00000000 >/sys/class/net/eth1/queue/tx-4/xps_cpus echo ff00,00000000 >/sys/class/net/eth1/queue/tx-5/xps_cpus echo ff0000,00000000 >/sys/class/net/eth1/queue/tx-6/xps_cpus echo ff000000,00000000 >/sys/class/net/eth1/queue/tx-7/xps_cpus ... Launched a tcp_stream with 15 threads and 1000 flows, initially affined to core 0-15 taskset -c 0-15 tcp_stream -T15 -F1000 -l1000 -c -H target_host Checked that only queues 0 and 1 are used as instructed by XPS : tc -s qdisc show dev eth1|grep backlog|grep -v "backlog 0b 0p" backlog 123489410b 1890p backlog 69809026b 1064p backlog 52401054b 805p Then force each thread to run on cpu 1,9,17,25,33,41,49,57,65,73,81,89,97,105,113,121 C=1;PID=`pidof tcp_stream`;for P in `ls /proc/$PID/task`; do taskset -pc $C $P; C=$(($C + 8));done Set txq_reselection_ms to 1000 echo 1000 > /proc/sys/net/core/txq_reselection_ms Check that the flows have migrated nicely: tc -s qdisc show dev eth1|grep backlog|grep -v "backlog 0b 0p" backlog 130508314b 1916p backlog 8584380b 126p backlog 8584380b 126p backlog 8379990b 123p backlog 8584380b 126p backlog 8487484b 125p backlog 8584380b 126p backlog 8448120b 124p backlog 8584380b 126p backlog 8720640b 128p backlog 8856900b 130p backlog 8584380b 126p backlog 8652510b 127p backlog 8448120b 124p backlog 8516250b 125p backlog 7834950b 115p Signed-off-by: Eric Dumazet Reviewed-by: Kuniyuki Iwashima Link: https://patch.msgid.link/20251013152234.842065-5-edumazet@google.com Signed-off-by: Jakub Kicinski --- include/net/sock.h | 26 ++++++++++++-------------- net/core/dev.c | 29 +++++++++++++++++++++++++++-- 2 files changed, 39 insertions(+), 16 deletions(-) (limited to 'include/net') diff --git a/include/net/sock.h b/include/net/sock.h index 2794bc5c5654..f0d00928db9e 100644 --- a/include/net/sock.h +++ b/include/net/sock.h @@ -313,6 +313,7 @@ struct sk_filter; * @sk_bind_phc: SO_TIMESTAMPING bind PHC index of PTP virtual clock * for timestamping * @sk_tskey: counter to disambiguate concurrent tstamp requests + * @sk_tx_queue_mapping_jiffies: time in jiffies of last @sk_tx_queue_mapping refresh. * @sk_zckey: counter to order MSG_ZEROCOPY notifications * @sk_socket: Identd and reporting IO signals * @sk_user_data: RPC layer private data. Write-protected by @sk_callback_lock. @@ -485,6 +486,7 @@ struct sock { unsigned long sk_pacing_rate; /* bytes per second */ atomic_t sk_zckey; atomic_t sk_tskey; + unsigned long sk_tx_queue_mapping_jiffies; __cacheline_group_end(sock_write_tx); __cacheline_group_begin(sock_read_tx); @@ -1992,7 +1994,15 @@ static inline void sk_tx_queue_set(struct sock *sk, int tx_queue) /* Paired with READ_ONCE() in sk_tx_queue_get() and * other WRITE_ONCE() because socket lock might be not held. */ - WRITE_ONCE(sk->sk_tx_queue_mapping, tx_queue); + if (READ_ONCE(sk->sk_tx_queue_mapping) != tx_queue) { + WRITE_ONCE(sk->sk_tx_queue_mapping, tx_queue); + WRITE_ONCE(sk->sk_tx_queue_mapping_jiffies, jiffies); + return; + } + + /* Refresh sk_tx_queue_mapping_jiffies if too old. */ + if (time_is_before_jiffies(READ_ONCE(sk->sk_tx_queue_mapping_jiffies) + HZ)) + WRITE_ONCE(sk->sk_tx_queue_mapping_jiffies, jiffies); } #define NO_QUEUE_MAPPING USHRT_MAX @@ -2005,19 +2015,7 @@ static inline void sk_tx_queue_clear(struct sock *sk) WRITE_ONCE(sk->sk_tx_queue_mapping, NO_QUEUE_MAPPING); } -static inline int sk_tx_queue_get(const struct sock *sk) -{ - if (sk) { - /* Paired with WRITE_ONCE() in sk_tx_queue_clear() - * and sk_tx_queue_set(). - */ - int val = READ_ONCE(sk->sk_tx_queue_mapping); - - if (val != NO_QUEUE_MAPPING) - return val; - } - return -1; -} +int sk_tx_queue_get(const struct sock *sk); static inline void __sk_rx_queue_set(struct sock *sk, const struct sk_buff *skb, diff --git a/net/core/dev.c b/net/core/dev.c index a64cef2c537e..33e6101dbc45 100644 --- a/net/core/dev.c +++ b/net/core/dev.c @@ -4591,6 +4591,32 @@ u16 dev_pick_tx_zero(struct net_device *dev, struct sk_buff *skb, } EXPORT_SYMBOL(dev_pick_tx_zero); +int sk_tx_queue_get(const struct sock *sk) +{ + int resel, val; + + if (!sk) + return -1; + /* Paired with WRITE_ONCE() in sk_tx_queue_clear() + * and sk_tx_queue_set(). + */ + val = READ_ONCE(sk->sk_tx_queue_mapping); + + if (val == NO_QUEUE_MAPPING) + return -1; + + if (!sk_fullsock(sk)) + return val; + + resel = READ_ONCE(sock_net(sk)->core.sysctl_txq_reselection); + if (resel && time_is_before_jiffies( + READ_ONCE(sk->sk_tx_queue_mapping_jiffies) + resel)) + return -1; + + return val; +} +EXPORT_SYMBOL(sk_tx_queue_get); + u16 netdev_pick_tx(struct net_device *dev, struct sk_buff *skb, struct net_device *sb_dev) { @@ -4606,8 +4632,7 @@ u16 netdev_pick_tx(struct net_device *dev, struct sk_buff *skb, if (new_index < 0) new_index = skb_tx_hash(dev, sb_dev, skb); - if (queue_index != new_index && sk && - sk_fullsock(sk) && + if (sk && sk_fullsock(sk) && rcu_access_pointer(sk->sk_dst_cache)) sk_tx_queue_set(sk, new_index); -- cgit v1.2.3 From e5b670e5439bda09ea7e3dd3dd32edb2f367c0d3 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Tue, 14 Oct 2025 14:06:05 +0000 Subject: net: remove obsolete WARN_ON(refcount_read(&sk->sk_refcnt) == 1) sk->sk_refcnt has been converted to refcount_t in 2017. __sock_put(sk) being refcount_dec(&sk->sk_refcnt), it will complain loudly if the current refcnt is 1 (or less) in a non racy way. We can remove four WARN_ON() in favor of the generic refcount_dec() check. Signed-off-by: Eric Dumazet Reviewed-by: Kuniyuki Iwashima Reviewed-by: Xuanqiang Luo Link: https://patch.msgid.link/20251014140605.2982703-1-edumazet@google.com Signed-off-by: Jakub Kicinski --- include/net/sock.h | 12 ++++-------- net/netlink/af_netlink.c | 4 +--- net/tipc/socket.c | 4 +--- 3 files changed, 6 insertions(+), 14 deletions(-) (limited to 'include/net') diff --git a/include/net/sock.h b/include/net/sock.h index f0d00928db9e..30ac2eb4ef9b 100644 --- a/include/net/sock.h +++ b/include/net/sock.h @@ -830,11 +830,9 @@ static inline bool sk_del_node_init(struct sock *sk) { bool rc = __sk_del_node_init(sk); - if (rc) { - /* paranoid for a while -acme */ - WARN_ON(refcount_read(&sk->sk_refcnt) == 1); + if (rc) __sock_put(sk); - } + return rc; } #define sk_del_node_init_rcu(sk) sk_del_node_init(sk) @@ -852,11 +850,9 @@ static inline bool sk_nulls_del_node_init_rcu(struct sock *sk) { bool rc = __sk_nulls_del_node_init_rcu(sk); - if (rc) { - /* paranoid for a while -acme */ - WARN_ON(refcount_read(&sk->sk_refcnt) == 1); + if (rc) __sock_put(sk); - } + return rc; } diff --git a/net/netlink/af_netlink.c b/net/netlink/af_netlink.c index 2b46c0cd752a..687a84c48882 100644 --- a/net/netlink/af_netlink.c +++ b/net/netlink/af_netlink.c @@ -596,10 +596,8 @@ static void netlink_remove(struct sock *sk) table = &nl_table[sk->sk_protocol]; if (!rhashtable_remove_fast(&table->hash, &nlk_sk(sk)->node, - netlink_rhashtable_params)) { - WARN_ON(refcount_read(&sk->sk_refcnt) == 1); + netlink_rhashtable_params)) __sock_put(sk); - } netlink_table_grab(); if (nlk_sk(sk)->subscriptions) { diff --git a/net/tipc/socket.c b/net/tipc/socket.c index 1574a83384f8..bc614a1f019c 100644 --- a/net/tipc/socket.c +++ b/net/tipc/socket.c @@ -3031,10 +3031,8 @@ static void tipc_sk_remove(struct tipc_sock *tsk) struct sock *sk = &tsk->sk; struct tipc_net *tn = net_generic(sock_net(sk), tipc_net_id); - if (!rhashtable_remove_fast(&tn->sk_rht, &tsk->node, tsk_rht_params)) { - WARN_ON(refcount_read(&sk->sk_refcnt) == 1); + if (!rhashtable_remove_fast(&tn->sk_rht, &tsk->node, tsk_rht_params)) __sock_put(sk); - } } static const struct rhashtable_params tsk_rht_params = { -- cgit v1.2.3 From 7c268eaeec6388b7bee36aef3fb5e62c9222ad3b Mon Sep 17 00:00:00 2001 From: Kuniyuki Iwashima Date: Tue, 14 Oct 2025 23:54:55 +0000 Subject: net: Allow opt-out from global protocol memory accounting. Some protocols (e.g., TCP, UDP) implement memory accounting for socket buffers and charge memory to per-protocol global counters pointed to by sk->sk_proto->memory_allocated. Sometimes, system processes do not want that limitation. For a similar purpose, there is SO_RESERVE_MEM for sockets under memcg. Also, by opting out of the per-protocol accounting, sockets under memcg can avoid paying costs for two orthogonal memory accounting mechanisms. A microbenchmark result is in the subsequent bpf patch. Let's allow opt-out from the per-protocol memory accounting if sk->sk_bypass_prot_mem is true. sk->sk_bypass_prot_mem and sk->sk_prot are placed in the same cache line, and sk_has_account() always fetches sk->sk_prot before accessing sk->sk_bypass_prot_mem, so there is no extra cache miss for this patch. The following patches will set sk->sk_bypass_prot_mem to true, and then, the per-protocol memory accounting will be skipped. Note that this does NOT disable memcg, but rather the per-protocol one. Another option not to use the hole in struct sock_common is create sk_prot variants like tcp_prot_bypass, but this would complicate SOCKMAP logic, tcp_bpf_prots etc. Signed-off-by: Kuniyuki Iwashima Signed-off-by: Martin KaFai Lau Reviewed-by: Shakeel Butt Reviewed-by: Eric Dumazet Acked-by: Roman Gushchin Link: https://patch.msgid.link/20251014235604.3057003-3-kuniyu@google.com --- include/net/proto_memory.h | 3 +++ include/net/sock.h | 3 +++ include/net/tcp.h | 3 +++ net/core/sock.c | 32 +++++++++++++++++++++++++------- net/ipv4/tcp.c | 3 ++- net/ipv4/tcp_output.c | 7 ++++++- net/mptcp/protocol.c | 7 ++++--- net/tls/tls_device.c | 3 ++- 8 files changed, 48 insertions(+), 13 deletions(-) (limited to 'include/net') diff --git a/include/net/proto_memory.h b/include/net/proto_memory.h index 8e91a8fa31b5..ad6d703ce6fe 100644 --- a/include/net/proto_memory.h +++ b/include/net/proto_memory.h @@ -35,6 +35,9 @@ static inline bool sk_under_memory_pressure(const struct sock *sk) mem_cgroup_sk_under_memory_pressure(sk)) return true; + if (sk->sk_bypass_prot_mem) + return false; + return !!READ_ONCE(*sk->sk_prot->memory_pressure); } diff --git a/include/net/sock.h b/include/net/sock.h index 30ac2eb4ef9b..415e7381aa50 100644 --- a/include/net/sock.h +++ b/include/net/sock.h @@ -118,6 +118,7 @@ typedef __u64 __bitwise __addrpair; * @skc_reuseport: %SO_REUSEPORT setting * @skc_ipv6only: socket is IPV6 only * @skc_net_refcnt: socket is using net ref counting + * @skc_bypass_prot_mem: bypass the per-protocol memory accounting for skb * @skc_bound_dev_if: bound device index if != 0 * @skc_bind_node: bind hash linkage for various protocol lookup tables * @skc_portaddr_node: second hash linkage for UDP/UDP-Lite protocol @@ -174,6 +175,7 @@ struct sock_common { unsigned char skc_reuseport:1; unsigned char skc_ipv6only:1; unsigned char skc_net_refcnt:1; + unsigned char skc_bypass_prot_mem:1; int skc_bound_dev_if; union { struct hlist_node skc_bind_node; @@ -381,6 +383,7 @@ struct sock { #define sk_reuseport __sk_common.skc_reuseport #define sk_ipv6only __sk_common.skc_ipv6only #define sk_net_refcnt __sk_common.skc_net_refcnt +#define sk_bypass_prot_mem __sk_common.skc_bypass_prot_mem #define sk_bound_dev_if __sk_common.skc_bound_dev_if #define sk_bind_node __sk_common.skc_bind_node #define sk_prot __sk_common.skc_prot diff --git a/include/net/tcp.h b/include/net/tcp.h index 1e547138f4fb..439e327fdbfa 100644 --- a/include/net/tcp.h +++ b/include/net/tcp.h @@ -303,6 +303,9 @@ static inline bool tcp_under_memory_pressure(const struct sock *sk) mem_cgroup_sk_under_memory_pressure(sk)) return true; + if (sk->sk_bypass_prot_mem) + return false; + return READ_ONCE(tcp_memory_pressure); } /* diff --git a/net/core/sock.c b/net/core/sock.c index 08ae20069b6d..5bf208579c02 100644 --- a/net/core/sock.c +++ b/net/core/sock.c @@ -1046,9 +1046,13 @@ static int sock_reserve_memory(struct sock *sk, int bytes) if (!charged) return -ENOMEM; + if (sk->sk_bypass_prot_mem) + goto success; + /* pre-charge to forward_alloc */ sk_memory_allocated_add(sk, pages); allocated = sk_memory_allocated(sk); + /* If the system goes into memory pressure with this * precharge, give up and return error. */ @@ -1057,6 +1061,8 @@ static int sock_reserve_memory(struct sock *sk, int bytes) mem_cgroup_sk_uncharge(sk, pages); return -ENOMEM; } + +success: sk_forward_alloc_add(sk, pages << PAGE_SHIFT); WRITE_ONCE(sk->sk_reserved_mem, @@ -3145,8 +3151,11 @@ bool sk_page_frag_refill(struct sock *sk, struct page_frag *pfrag) if (likely(skb_page_frag_refill(32U, pfrag, sk->sk_allocation))) return true; - sk_enter_memory_pressure(sk); + if (!sk->sk_bypass_prot_mem) + sk_enter_memory_pressure(sk); + sk_stream_moderate_sndbuf(sk); + return false; } EXPORT_SYMBOL(sk_page_frag_refill); @@ -3263,10 +3272,12 @@ int __sk_mem_raise_allocated(struct sock *sk, int size, int amt, int kind) { bool memcg_enabled = false, charged = false; struct proto *prot = sk->sk_prot; - long allocated; + long allocated = 0; - sk_memory_allocated_add(sk, amt); - allocated = sk_memory_allocated(sk); + if (!sk->sk_bypass_prot_mem) { + sk_memory_allocated_add(sk, amt); + allocated = sk_memory_allocated(sk); + } if (mem_cgroup_sk_enabled(sk)) { memcg_enabled = true; @@ -3275,6 +3286,9 @@ int __sk_mem_raise_allocated(struct sock *sk, int size, int amt, int kind) goto suppress_allocation; } + if (!allocated) + return 1; + /* Under limit. */ if (allocated <= sk_prot_mem_limits(sk, 0)) { sk_leave_memory_pressure(sk); @@ -3353,7 +3367,8 @@ suppress_allocation: trace_sock_exceed_buf_limit(sk, prot, allocated, kind); - sk_memory_allocated_sub(sk, amt); + if (allocated) + sk_memory_allocated_sub(sk, amt); if (charged) mem_cgroup_sk_uncharge(sk, amt); @@ -3392,11 +3407,14 @@ EXPORT_SYMBOL(__sk_mem_schedule); */ void __sk_mem_reduce_allocated(struct sock *sk, int amount) { - sk_memory_allocated_sub(sk, amount); - if (mem_cgroup_sk_enabled(sk)) mem_cgroup_sk_uncharge(sk, amount); + if (sk->sk_bypass_prot_mem) + return; + + sk_memory_allocated_sub(sk, amount); + if (sk_under_global_memory_pressure(sk) && (sk_memory_allocated(sk) < sk_prot_mem_limits(sk, 0))) sk_leave_memory_pressure(sk); diff --git a/net/ipv4/tcp.c b/net/ipv4/tcp.c index 4d720aa09a4c..54def27326f1 100644 --- a/net/ipv4/tcp.c +++ b/net/ipv4/tcp.c @@ -928,7 +928,8 @@ struct sk_buff *tcp_stream_alloc_skb(struct sock *sk, gfp_t gfp, } __kfree_skb(skb); } else { - sk->sk_prot->enter_memory_pressure(sk); + if (!sk->sk_bypass_prot_mem) + tcp_enter_memory_pressure(sk); sk_stream_moderate_sndbuf(sk); } return NULL; diff --git a/net/ipv4/tcp_output.c b/net/ipv4/tcp_output.c index b94efb3050d2..7f5df7a71f62 100644 --- a/net/ipv4/tcp_output.c +++ b/net/ipv4/tcp_output.c @@ -3743,12 +3743,17 @@ void sk_forced_mem_schedule(struct sock *sk, int size) delta = size - sk->sk_forward_alloc; if (delta <= 0) return; + amt = sk_mem_pages(delta); sk_forward_alloc_add(sk, amt << PAGE_SHIFT); - sk_memory_allocated_add(sk, amt); if (mem_cgroup_sk_enabled(sk)) mem_cgroup_sk_charge(sk, amt, gfp_memcg_charge() | __GFP_NOFAIL); + + if (sk->sk_bypass_prot_mem) + return; + + sk_memory_allocated_add(sk, amt); } /* Send a FIN. The caller locks the socket for us. diff --git a/net/mptcp/protocol.c b/net/mptcp/protocol.c index 0292162a14ee..94a5f6dcc577 100644 --- a/net/mptcp/protocol.c +++ b/net/mptcp/protocol.c @@ -1065,11 +1065,12 @@ static void mptcp_enter_memory_pressure(struct sock *sk) mptcp_for_each_subflow(msk, subflow) { struct sock *ssk = mptcp_subflow_tcp_sock(subflow); - if (first) + if (first && !ssk->sk_bypass_prot_mem) { tcp_enter_memory_pressure(ssk); - sk_stream_moderate_sndbuf(ssk); + first = false; + } - first = false; + sk_stream_moderate_sndbuf(ssk); } __mptcp_sync_sndbuf(sk); } diff --git a/net/tls/tls_device.c b/net/tls/tls_device.c index a64ae15b1a60..caa2b5d24622 100644 --- a/net/tls/tls_device.c +++ b/net/tls/tls_device.c @@ -373,7 +373,8 @@ static int tls_do_allocation(struct sock *sk, if (!offload_ctx->open_record) { if (unlikely(!skb_page_frag_refill(prepend_size, pfrag, sk->sk_allocation))) { - READ_ONCE(sk->sk_prot)->enter_memory_pressure(sk); + if (!sk->sk_bypass_prot_mem) + READ_ONCE(sk->sk_prot)->enter_memory_pressure(sk); sk_stream_moderate_sndbuf(sk); return -ENOMEM; } -- cgit v1.2.3 From b46ab63181ff973ddce44ebc9ac24b269d42f481 Mon Sep 17 00:00:00 2001 From: Kuniyuki Iwashima Date: Tue, 14 Oct 2025 23:54:56 +0000 Subject: net: Introduce net.core.bypass_prot_mem sysctl. If a socket has sk->sk_bypass_prot_mem flagged, the socket opts out of the global protocol memory accounting. Let's control the flag by a new sysctl knob. The flag is written once during socket(2) and is inherited to child sockets. Tested with a script that creates local socket pairs and send()s a bunch of data without recv()ing. Setup: # mkdir /sys/fs/cgroup/test # echo $$ >> /sys/fs/cgroup/test/cgroup.procs # sysctl -q net.ipv4.tcp_mem="1000 1000 1000" # ulimit -n 524288 Without net.core.bypass_prot_mem, charged to tcp_mem & memcg # python3 pressure.py & # cat /sys/fs/cgroup/test/memory.stat | grep sock sock 22642688 <-------------------------------------- charged to memcg # cat /proc/net/sockstat| grep TCP TCP: inuse 2006 orphan 0 tw 0 alloc 2008 mem 5376 <-- charged to tcp_mem # ss -tn | head -n 5 State Recv-Q Send-Q Local Address:Port Peer Address:Port ESTAB 2000 0 127.0.0.1:34479 127.0.0.1:53188 ESTAB 2000 0 127.0.0.1:34479 127.0.0.1:49972 ESTAB 2000 0 127.0.0.1:34479 127.0.0.1:53868 ESTAB 2000 0 127.0.0.1:34479 127.0.0.1:53554 # nstat | grep Pressure || echo no pressure TcpExtTCPMemoryPressures 1 0.0 With net.core.bypass_prot_mem=1, charged to memcg only: # sysctl -q net.core.bypass_prot_mem=1 # python3 pressure.py & # cat /sys/fs/cgroup/test/memory.stat | grep sock sock 2757468160 <------------------------------------ charged to memcg # cat /proc/net/sockstat | grep TCP TCP: inuse 2006 orphan 0 tw 0 alloc 2008 mem 0 <- NOT charged to tcp_mem # ss -tn | head -n 5 State Recv-Q Send-Q Local Address:Port Peer Address:Port ESTAB 111000 0 127.0.0.1:36019 127.0.0.1:49026 ESTAB 110000 0 127.0.0.1:36019 127.0.0.1:45630 ESTAB 110000 0 127.0.0.1:36019 127.0.0.1:44870 ESTAB 111000 0 127.0.0.1:36019 127.0.0.1:45274 # nstat | grep Pressure || echo no pressure no pressure Signed-off-by: Kuniyuki Iwashima Signed-off-by: Martin KaFai Lau Reviewed-by: Shakeel Butt Reviewed-by: Eric Dumazet Acked-by: Roman Gushchin Link: https://patch.msgid.link/20251014235604.3057003-4-kuniyu@google.com --- Documentation/admin-guide/sysctl/net.rst | 8 ++++++++ include/net/netns/core.h | 1 + net/core/sock.c | 5 +++++ net/core/sysctl_net_core.c | 9 +++++++++ 4 files changed, 23 insertions(+) (limited to 'include/net') diff --git a/Documentation/admin-guide/sysctl/net.rst b/Documentation/admin-guide/sysctl/net.rst index 40749b3cd356..991773dcb9cf 100644 --- a/Documentation/admin-guide/sysctl/net.rst +++ b/Documentation/admin-guide/sysctl/net.rst @@ -212,6 +212,14 @@ mem_pcpu_rsv Per-cpu reserved forward alloc cache size in page units. Default 1MB per CPU. +bypass_prot_mem +--------------- + +Skip charging socket buffers to the global per-protocol memory +accounting controlled by net.ipv4.tcp_mem, net.ipv4.udp_mem, etc. + +Default: 0 (off) + rmem_default ------------ diff --git a/include/net/netns/core.h b/include/net/netns/core.h index cb9c3e4cd738..9ef3d70e5e9c 100644 --- a/include/net/netns/core.h +++ b/include/net/netns/core.h @@ -17,6 +17,7 @@ struct netns_core { int sysctl_optmem_max; u8 sysctl_txrehash; u8 sysctl_tstamp_allow_data; + u8 sysctl_bypass_prot_mem; #ifdef CONFIG_PROC_FS struct prot_inuse __percpu *prot_inuse; diff --git a/net/core/sock.c b/net/core/sock.c index 5bf208579c02..b78533fb9268 100644 --- a/net/core/sock.c +++ b/net/core/sock.c @@ -2306,8 +2306,13 @@ struct sock *sk_alloc(struct net *net, int family, gfp_t priority, * why we need sk_prot_creator -acme */ sk->sk_prot = sk->sk_prot_creator = prot; + + if (READ_ONCE(net->core.sysctl_bypass_prot_mem)) + sk->sk_bypass_prot_mem = 1; + sk->sk_kern_sock = kern; sock_lock_init(sk); + sk->sk_net_refcnt = kern ? 0 : 1; if (likely(sk->sk_net_refcnt)) { get_net_track(net, &sk->ns_tracker, priority); diff --git a/net/core/sysctl_net_core.c b/net/core/sysctl_net_core.c index f79137826d7f..8d4decb2606f 100644 --- a/net/core/sysctl_net_core.c +++ b/net/core/sysctl_net_core.c @@ -683,6 +683,15 @@ static struct ctl_table netns_core_table[] = { .extra1 = SYSCTL_ZERO, .extra2 = SYSCTL_ONE }, + { + .procname = "bypass_prot_mem", + .data = &init_net.core.sysctl_bypass_prot_mem, + .maxlen = sizeof(u8), + .mode = 0644, + .proc_handler = proc_dou8vec_minmax, + .extra1 = SYSCTL_ZERO, + .extra2 = SYSCTL_ONE + }, /* sysctl_core_net_init() will set the values after this * to readonly in network namespaces */ -- cgit v1.2.3 From 178ca30889a13b555dddab7689fd2cc58c8e5dac Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Tue, 14 Oct 2025 17:19:05 +0000 Subject: Revert "net/sched: Fix mirred deadlock on device recursion" MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit This reverts commits 0f022d32c3eca477fbf79a205243a6123ed0fe11 and 44180feaccf266d9b0b28cc4ceaac019817deb5c. Prior patch in this series implemented loop detection in act_mirred, we can remove q->owner to save some cycles in the fast path. Signed-off-by: Eric Dumazet Reviewed-by: Kuniyuki Iwashima Reviewed-by: Toke Høiland-Jørgensen Reviewed-by: Victor Nogueira Tested-by: Jamal Hadi Salim Acked-by: Jamal Hadi Salim Link: https://patch.msgid.link/20251014171907.3554413-5-edumazet@google.com Signed-off-by: Jakub Kicinski --- include/net/sch_generic.h | 1 - net/core/dev.c | 6 ------ net/sched/sch_generic.c | 2 -- 3 files changed, 9 deletions(-) (limited to 'include/net') diff --git a/include/net/sch_generic.h b/include/net/sch_generic.h index 738cd5b13c62..32e9961570b4 100644 --- a/include/net/sch_generic.h +++ b/include/net/sch_generic.h @@ -117,7 +117,6 @@ struct Qdisc { struct qdisc_skb_head q; struct gnet_stats_basic_sync bstats; struct gnet_stats_queue qstats; - int owner; unsigned long state; unsigned long state2; /* must be written under qdisc spinlock */ struct Qdisc *next_sched; diff --git a/net/core/dev.c b/net/core/dev.c index 35010faf0b78..1d8e7a76d83b 100644 --- a/net/core/dev.c +++ b/net/core/dev.c @@ -4167,10 +4167,6 @@ no_lock_out: return rc; } - if (unlikely(READ_ONCE(q->owner) == smp_processor_id())) { - kfree_skb_reason(skb, SKB_DROP_REASON_TC_RECLASSIFY_LOOP); - return NET_XMIT_DROP; - } /* * Heuristic to force contended enqueues to serialize on a * separate lock before trying to get qdisc main lock. @@ -4210,9 +4206,7 @@ no_lock_out: qdisc_run_end(q); rc = NET_XMIT_SUCCESS; } else { - WRITE_ONCE(q->owner, smp_processor_id()); rc = dev_qdisc_enqueue(skb, q, &to_free, txq); - WRITE_ONCE(q->owner, -1); if (qdisc_run_begin(q)) { if (unlikely(contended)) { spin_unlock(&q->busylock); diff --git a/net/sched/sch_generic.c b/net/sched/sch_generic.c index 1e008a228ebd..dfa8e8e667d2 100644 --- a/net/sched/sch_generic.c +++ b/net/sched/sch_generic.c @@ -679,7 +679,6 @@ struct Qdisc noop_qdisc = { .qlen = 0, .lock = __SPIN_LOCK_UNLOCKED(noop_qdisc.skb_bad_txq.lock), }, - .owner = -1, }; EXPORT_SYMBOL(noop_qdisc); @@ -985,7 +984,6 @@ struct Qdisc *qdisc_alloc(struct netdev_queue *dev_queue, sch->enqueue = ops->enqueue; sch->dequeue = ops->dequeue; sch->dev_queue = dev_queue; - sch->owner = -1; netdev_hold(dev, &sch->dev_tracker, GFP_KERNEL); refcount_set(&sch->refcnt, 1); -- cgit v1.2.3 From 526f5fb112f7c89c5a9b8b2f9870c8cb76ca4e42 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Tue, 14 Oct 2025 17:19:06 +0000 Subject: net: sched: claim one cache line in Qdisc MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Replace state2 field with a boolean. Move it to a hole between qstats and state so that we shrink Qdisc by a full cache line. Signed-off-by: Eric Dumazet Reviewed-by: Kuniyuki Iwashima Reviewed-by: Toke Høiland-Jørgensen Tested-by: Jamal Hadi Salim Acked-by: Jamal Hadi Salim Link: https://patch.msgid.link/20251014171907.3554413-6-edumazet@google.com Signed-off-by: Jakub Kicinski --- include/net/sch_generic.h | 18 +++++++----------- 1 file changed, 7 insertions(+), 11 deletions(-) (limited to 'include/net') diff --git a/include/net/sch_generic.h b/include/net/sch_generic.h index 32e9961570b4..31561291bc92 100644 --- a/include/net/sch_generic.h +++ b/include/net/sch_generic.h @@ -41,13 +41,6 @@ enum qdisc_state_t { __QDISC_STATE_DRAINING, }; -enum qdisc_state2_t { - /* Only for !TCQ_F_NOLOCK qdisc. Never access it directly. - * Use qdisc_run_begin/end() or qdisc_is_running() instead. - */ - __QDISC_STATE2_RUNNING, -}; - #define QDISC_STATE_MISSED BIT(__QDISC_STATE_MISSED) #define QDISC_STATE_DRAINING BIT(__QDISC_STATE_DRAINING) @@ -117,8 +110,8 @@ struct Qdisc { struct qdisc_skb_head q; struct gnet_stats_basic_sync bstats; struct gnet_stats_queue qstats; + bool running; /* must be written under qdisc spinlock */ unsigned long state; - unsigned long state2; /* must be written under qdisc spinlock */ struct Qdisc *next_sched; struct sk_buff_head skb_bad_txq; @@ -167,7 +160,7 @@ static inline bool qdisc_is_running(struct Qdisc *qdisc) { if (qdisc->flags & TCQ_F_NOLOCK) return spin_is_locked(&qdisc->seqlock); - return test_bit(__QDISC_STATE2_RUNNING, &qdisc->state2); + return READ_ONCE(qdisc->running); } static inline bool nolock_qdisc_is_empty(const struct Qdisc *qdisc) @@ -210,7 +203,10 @@ static inline bool qdisc_run_begin(struct Qdisc *qdisc) */ return spin_trylock(&qdisc->seqlock); } - return !__test_and_set_bit(__QDISC_STATE2_RUNNING, &qdisc->state2); + if (READ_ONCE(qdisc->running)) + return false; + WRITE_ONCE(qdisc->running, true); + return true; } static inline void qdisc_run_end(struct Qdisc *qdisc) @@ -228,7 +224,7 @@ static inline void qdisc_run_end(struct Qdisc *qdisc) &qdisc->state))) __netif_schedule(qdisc); } else { - __clear_bit(__QDISC_STATE2_RUNNING, &qdisc->state2); + WRITE_ONCE(qdisc->running, false); } } -- cgit v1.2.3 From 100dfa74cad9d4665cdcf0cc8e673b123a3ea910 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Tue, 14 Oct 2025 17:19:07 +0000 Subject: net: dev_queue_xmit() llist adoption MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Remove busylock spinlock and use a lockless list (llist) to reduce spinlock contention to the minimum. Idea is that only one cpu might spin on the qdisc spinlock, while others simply add their skb in the llist. After this patch, we get a 300 % improvement on heavy TX workloads. - Sending twice the number of packets per second. - While consuming 50 % less cycles. Note that this also allows in the future to submit batches to various qdisc->enqueue() methods. Tested: - Dual Intel(R) Xeon(R) 6985P-C (480 hyper threads). - 100Gbit NIC, 30 TX queues with FQ packet scheduler. - echo 64 >/sys/kernel/slab/skbuff_small_head/cpu_partial (avoid contention in mm) - 240 concurrent "netperf -t UDP_STREAM -- -m 120 -n" Before: 16 Mpps (41 Mpps if each thread is pinned to a different cpu) vmstat 2 5 procs -----------memory---------- ---swap-- -----io---- -system-- ------cpu----- r b swpd free buff cache si so bi bo in cs us sy id wa st 243 0 0 2368988672 51036 1100852 0 0 146 1 242 60 0 9 91 0 0 244 0 0 2368988672 51036 1100852 0 0 536 10 487745 14718 0 52 48 0 0 244 0 0 2368988672 51036 1100852 0 0 512 0 503067 46033 0 52 48 0 0 244 0 0 2368988672 51036 1100852 0 0 512 0 494807 12107 0 52 48 0 0 244 0 0 2368988672 51036 1100852 0 0 702 26 492845 10110 0 52 48 0 0 Lock contention (1 second sample taken on 8 cores) perf lock record -C0-7 sleep 1; perf lock contention contended total wait max wait avg wait type caller 442111 6.79 s 162.47 ms 15.35 us spinlock dev_hard_start_xmit+0xcd 5961 9.57 ms 8.12 us 1.60 us spinlock __dev_queue_xmit+0x3a0 244 560.63 us 7.63 us 2.30 us spinlock do_softirq+0x5b 13 25.09 us 3.21 us 1.93 us spinlock net_tx_action+0xf8 If netperf threads are pinned, spinlock stress is very high. perf lock record -C0-7 sleep 1; perf lock contention contended total wait max wait avg wait type caller 964508 7.10 s 147.25 ms 7.36 us spinlock dev_hard_start_xmit+0xcd 201 268.05 us 4.65 us 1.33 us spinlock __dev_queue_xmit+0x3a0 12 26.05 us 3.84 us 2.17 us spinlock do_softirq+0x5b @__dev_queue_xmit_ns: [256, 512) 21 | | [512, 1K) 631 | | [1K, 2K) 27328 |@ | [2K, 4K) 265392 |@@@@@@@@@@@@@@@@ | [4K, 8K) 417543 |@@@@@@@@@@@@@@@@@@@@@@@@@@ | [8K, 16K) 826292 |@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@| [16K, 32K) 733822 |@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@ | [32K, 64K) 19055 |@ | [64K, 128K) 17240 |@ | [128K, 256K) 25633 |@ | [256K, 512K) 4 | | After: 29 Mpps (57 Mpps if each thread is pinned to a different cpu) vmstat 2 5 procs -----------memory---------- ---swap-- -----io---- -system-- ------cpu----- r b swpd free buff cache si so bi bo in cs us sy id wa st 78 0 0 2369573632 32896 1350988 0 0 22 0 331 254 0 8 92 0 0 75 0 0 2369573632 32896 1350988 0 0 22 50 425713 280199 0 23 76 0 0 104 0 0 2369573632 32896 1350988 0 0 290 0 430238 298247 0 23 76 0 0 86 0 0 2369573632 32896 1350988 0 0 132 0 428019 291865 0 24 76 0 0 90 0 0 2369573632 32896 1350988 0 0 502 0 422498 278672 0 23 76 0 0 perf lock record -C0-7 sleep 1; perf lock contention contended total wait max wait avg wait type caller 2524 116.15 ms 486.61 us 46.02 us spinlock __dev_queue_xmit+0x55b 5821 107.18 ms 371.67 us 18.41 us spinlock dev_hard_start_xmit+0xcd 2377 9.73 ms 35.86 us 4.09 us spinlock ___slab_alloc+0x4e0 923 5.74 ms 20.91 us 6.22 us spinlock ___slab_alloc+0x5c9 121 3.42 ms 193.05 us 28.24 us spinlock net_tx_action+0xf8 6 564.33 us 167.60 us 94.05 us spinlock do_softirq+0x5b If netperf threads are pinned (~54 Mpps) perf lock record -C0-7 sleep 1; perf lock contention 32907 316.98 ms 195.98 us 9.63 us spinlock dev_hard_start_xmit+0xcd 4507 61.83 ms 212.73 us 13.72 us spinlock __dev_queue_xmit+0x554 2781 23.53 ms 40.03 us 8.46 us spinlock ___slab_alloc+0x5c9 3554 18.94 ms 34.69 us 5.33 us spinlock ___slab_alloc+0x4e0 233 9.09 ms 215.70 us 38.99 us spinlock do_softirq+0x5b 153 930.66 us 48.67 us 6.08 us spinlock net_tx_action+0xfd 84 331.10 us 14.22 us 3.94 us spinlock ___slab_alloc+0x5c9 140 323.71 us 9.94 us 2.31 us spinlock ___slab_alloc+0x4e0 @__dev_queue_xmit_ns: [128, 256) 1539830 |@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@ | [256, 512) 2299558 |@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@| [512, 1K) 483936 |@@@@@@@@@@ | [1K, 2K) 265345 |@@@@@@ | [2K, 4K) 145463 |@@@ | [4K, 8K) 54571 |@ | [8K, 16K) 10270 | | [16K, 32K) 9385 | | [32K, 64K) 7749 | | [64K, 128K) 26799 | | [128K, 256K) 2665 | | [256K, 512K) 665 | | Signed-off-by: Eric Dumazet Reviewed-by: Toke Høiland-Jørgensen Reviewed-by: Kuniyuki Iwashima Tested-by: Jamal Hadi Salim Acked-by: Jamal Hadi Salim Link: https://patch.msgid.link/20251014171907.3554413-7-edumazet@google.com Signed-off-by: Jakub Kicinski --- include/net/sch_generic.h | 4 ++- net/core/dev.c | 91 +++++++++++++++++++++++++++++------------------ net/sched/sch_generic.c | 5 --- 3 files changed, 59 insertions(+), 41 deletions(-) (limited to 'include/net') diff --git a/include/net/sch_generic.h b/include/net/sch_generic.h index 31561291bc92..94966692ccdf 100644 --- a/include/net/sch_generic.h +++ b/include/net/sch_generic.h @@ -115,7 +115,9 @@ struct Qdisc { struct Qdisc *next_sched; struct sk_buff_head skb_bad_txq; - spinlock_t busylock ____cacheline_aligned_in_smp; + atomic_long_t defer_count ____cacheline_aligned_in_smp; + struct llist_head defer_list; + spinlock_t seqlock; struct rcu_head rcu; diff --git a/net/core/dev.c b/net/core/dev.c index 1d8e7a76d83b..821e7c718924 100644 --- a/net/core/dev.c +++ b/net/core/dev.c @@ -4125,9 +4125,10 @@ static inline int __dev_xmit_skb(struct sk_buff *skb, struct Qdisc *q, struct net_device *dev, struct netdev_queue *txq) { + struct sk_buff *next, *to_free = NULL; spinlock_t *root_lock = qdisc_lock(q); - struct sk_buff *to_free = NULL; - bool contended; + struct llist_node *ll_list, *first_n; + unsigned long defer_count = 0; int rc; qdisc_calculate_pkt_len(skb, q); @@ -4167,61 +4168,81 @@ no_lock_out: return rc; } - /* - * Heuristic to force contended enqueues to serialize on a - * separate lock before trying to get qdisc main lock. - * This permits qdisc->running owner to get the lock more - * often and dequeue packets faster. - * On PREEMPT_RT it is possible to preempt the qdisc owner during xmit - * and then other tasks will only enqueue packets. The packets will be - * sent after the qdisc owner is scheduled again. To prevent this - * scenario the task always serialize on the lock. + /* Open code llist_add(&skb->ll_node, &q->defer_list) + queue limit. + * In the try_cmpxchg() loop, we want to increment q->defer_count + * at most once to limit the number of skbs in defer_list. + * We perform the defer_count increment only if the list is not empty, + * because some arches have slow atomic_long_inc_return(). + */ + first_n = READ_ONCE(q->defer_list.first); + do { + if (first_n && !defer_count) { + defer_count = atomic_long_inc_return(&q->defer_count); + if (unlikely(defer_count > q->limit)) { + kfree_skb_reason(skb, SKB_DROP_REASON_QDISC_DROP); + return NET_XMIT_DROP; + } + } + skb->ll_node.next = first_n; + } while (!try_cmpxchg(&q->defer_list.first, &first_n, &skb->ll_node)); + + /* If defer_list was not empty, we know the cpu which queued + * the first skb will process the whole list for us. */ - contended = qdisc_is_running(q) || IS_ENABLED(CONFIG_PREEMPT_RT); - if (unlikely(contended)) - spin_lock(&q->busylock); + if (first_n) + return NET_XMIT_SUCCESS; spin_lock(root_lock); + + ll_list = llist_del_all(&q->defer_list); + /* There is a small race because we clear defer_count not atomically + * with the prior llist_del_all(). This means defer_list could grow + * over q->limit. + */ + atomic_long_set(&q->defer_count, 0); + + ll_list = llist_reverse_order(ll_list); + if (unlikely(test_bit(__QDISC_STATE_DEACTIVATED, &q->state))) { - __qdisc_drop(skb, &to_free); + llist_for_each_entry_safe(skb, next, ll_list, ll_node) + __qdisc_drop(skb, &to_free); rc = NET_XMIT_DROP; - } else if ((q->flags & TCQ_F_CAN_BYPASS) && !qdisc_qlen(q) && - qdisc_run_begin(q)) { + goto unlock; + } + if ((q->flags & TCQ_F_CAN_BYPASS) && !qdisc_qlen(q) && + !llist_next(ll_list) && qdisc_run_begin(q)) { /* * This is a work-conserving queue; there are no old skbs * waiting to be sent out; and the qdisc is not running - * xmit the skb directly. */ + DEBUG_NET_WARN_ON_ONCE(skb != llist_entry(ll_list, + struct sk_buff, + ll_node)); qdisc_bstats_update(q, skb); - - if (sch_direct_xmit(skb, q, dev, txq, root_lock, true)) { - if (unlikely(contended)) { - spin_unlock(&q->busylock); - contended = false; - } + if (sch_direct_xmit(skb, q, dev, txq, root_lock, true)) __qdisc_run(q); - } - qdisc_run_end(q); rc = NET_XMIT_SUCCESS; } else { - rc = dev_qdisc_enqueue(skb, q, &to_free, txq); - if (qdisc_run_begin(q)) { - if (unlikely(contended)) { - spin_unlock(&q->busylock); - contended = false; - } - __qdisc_run(q); - qdisc_run_end(q); + int count = 0; + + llist_for_each_entry_safe(skb, next, ll_list, ll_node) { + prefetch(next); + skb_mark_not_on_list(skb); + rc = dev_qdisc_enqueue(skb, q, &to_free, txq); + count++; } + qdisc_run(q); + if (count != 1) + rc = NET_XMIT_SUCCESS; } +unlock: spin_unlock(root_lock); if (unlikely(to_free)) kfree_skb_list_reason(to_free, tcf_get_drop_reason(to_free)); - if (unlikely(contended)) - spin_unlock(&q->busylock); return rc; } diff --git a/net/sched/sch_generic.c b/net/sched/sch_generic.c index dfa8e8e667d2..d9a98d02a55f 100644 --- a/net/sched/sch_generic.c +++ b/net/sched/sch_generic.c @@ -666,7 +666,6 @@ struct Qdisc noop_qdisc = { .ops = &noop_qdisc_ops, .q.lock = __SPIN_LOCK_UNLOCKED(noop_qdisc.q.lock), .dev_queue = &noop_netdev_queue, - .busylock = __SPIN_LOCK_UNLOCKED(noop_qdisc.busylock), .gso_skb = { .next = (struct sk_buff *)&noop_qdisc.gso_skb, .prev = (struct sk_buff *)&noop_qdisc.gso_skb, @@ -970,10 +969,6 @@ struct Qdisc *qdisc_alloc(struct netdev_queue *dev_queue, } } - spin_lock_init(&sch->busylock); - lockdep_set_class(&sch->busylock, - dev->qdisc_tx_busylock ?: &qdisc_tx_busylock); - /* seqlock has the same scope of busylock, for NOLOCK qdisc */ spin_lock_init(&sch->seqlock); lockdep_set_class(&sch->seqlock, -- cgit v1.2.3 From 1c17f4373d4db1e1f0ebd3ddcd8e7a642927a826 Mon Sep 17 00:00:00 2001 From: Kuniyuki Iwashima Date: Tue, 14 Oct 2025 22:42:07 +0000 Subject: ipv6: Move ipv6_fl_list from ipv6_pinfo to inet_sock. In {tcp6,udp6,raw6}_sock, struct ipv6_pinfo is always placed at the beginning of a new cache line because 1. __alignof__(struct tcp_sock) is 64 due to ____cacheline_aligned of __cacheline_group_begin(tcp_sock_write_tx) 2. __alignof__(struct udp_sock) is 64 due to ____cacheline_aligned of struct numa_drop_counters 3. in raw6_sock, struct numa_drop_counters is placed before struct ipv6_pinfo . struct ipv6_pinfo is 136 bytes, but the last cache line is only used by ipv6_fl_list: $ pahole -C ipv6_pinfo vmlinux struct ipv6_pinfo { ... /* --- cacheline 2 boundary (128 bytes) --- */ struct ipv6_fl_socklist * ipv6_fl_list; /* 128 8 */ /* size: 136, cachelines: 3, members: 23 */ Let's move ipv6_fl_list from struct ipv6_pinfo to struct inet_sock to save a full cache line for {tcp6,udp6,raw6}_sock. Now, struct ipv6_pinfo is 128 bytes, and {tcp6,udp6,raw6}_sock have 64 bytes less, while {tcp,udp,raw}_sock retain the same size. Before: # grep -E "^(RAW|UDP[^L\-]|TCP)" /proc/slabinfo | awk '{print $1, "\t", $4}' RAWv6 1408 UDPv6 1472 TCPv6 2560 RAW 1152 UDP 1280 TCP 2368 After: # grep -E "^(RAW|UDP[^L\-]|TCP)" /proc/slabinfo | awk '{print $1, "\t", $4}' RAWv6 1344 UDPv6 1408 TCPv6 2496 RAW 1152 UDP 1280 TCP 2368 Also, ipv6_fl_list and inet_flags (SNDFLOW bit) are placed in the same cache line. $ pahole -C inet_sock vmlinux ... /* --- cacheline 11 boundary (704 bytes) was 56 bytes ago --- */ struct ipv6_pinfo * pinet6; /* 760 8 */ /* --- cacheline 12 boundary (768 bytes) --- */ struct ipv6_fl_socklist * ipv6_fl_list; /* 768 8 */ unsigned long inet_flags; /* 776 8 */ Doc churn is due to the insufficient Type column (only 1 space short). Suggested-by: Eric Dumazet Signed-off-by: Kuniyuki Iwashima Link: https://patch.msgid.link/20251014224210.2964778-1-kuniyu@google.com Signed-off-by: Jakub Kicinski --- .../networking/net_cachelines/inet_sock.rst | 79 +++++++++++----------- .../chelsio/inline_crypto/chtls/chtls_cm.c | 4 +- include/linux/ipv6.h | 1 - include/net/inet_sock.h | 1 + net/ipv6/ip6_flowlabel.c | 44 ++++++------ net/ipv6/tcp_ipv6.c | 13 ++-- net/sctp/ipv6.c | 8 ++- 7 files changed, 76 insertions(+), 74 deletions(-) (limited to 'include/net') diff --git a/Documentation/networking/net_cachelines/inet_sock.rst b/Documentation/networking/net_cachelines/inet_sock.rst index b11bf48fa2b3..4c72a28a7012 100644 --- a/Documentation/networking/net_cachelines/inet_sock.rst +++ b/Documentation/networking/net_cachelines/inet_sock.rst @@ -5,42 +5,43 @@ inet_sock struct fast path usage breakdown ========================================== -======================= ===================== =================== =================== ====================================================================================================== -Type Name fastpath_tx_access fastpath_rx_access comment -======================= ===================== =================== =================== ====================================================================================================== -struct sock sk read_mostly read_mostly tcp_init_buffer_space,tcp_init_transfer,tcp_finish_connect,tcp_connect,tcp_send_rcvq,tcp_send_syn_data -struct ipv6_pinfo* pinet6 -be16 inet_sport read_mostly __tcp_transmit_skb -be32 inet_daddr read_mostly ip_select_ident_segs -be32 inet_rcv_saddr -be16 inet_dport read_mostly __tcp_transmit_skb -u16 inet_num -be32 inet_saddr -s16 uc_ttl read_mostly __ip_queue_xmit/ip_select_ttl -u16 cmsg_flags -struct ip_options_rcu* inet_opt read_mostly __ip_queue_xmit -u16 inet_id read_mostly ip_select_ident_segs -u8 tos read_mostly ip_queue_xmit -u8 min_ttl -u8 mc_ttl -u8 pmtudisc -u8:1 recverr -u8:1 is_icsk -u8:1 freebind -u8:1 hdrincl -u8:1 mc_loop -u8:1 transparent -u8:1 mc_all -u8:1 nodefrag -u8:1 bind_address_no_port -u8:1 recverr_rfc4884 -u8:1 defer_connect read_mostly tcp_sendmsg_fastopen -u8 rcv_tos -u8 convert_csum -int uc_index -int mc_index -be32 mc_addr -struct ip_mc_socklist* mc_list -struct inet_cork_full cork read_mostly __tcp_transmit_skb -struct local_port_range -======================= ===================== =================== =================== ====================================================================================================== +======================== ===================== =================== =================== ====================================================================================================== +Type Name fastpath_tx_access fastpath_rx_access comment +======================== ===================== =================== =================== ====================================================================================================== +struct sock sk read_mostly read_mostly tcp_init_buffer_space,tcp_init_transfer,tcp_finish_connect,tcp_connect,tcp_send_rcvq,tcp_send_syn_data +struct ipv6_pinfo* pinet6 +struct ipv6_fl_socklist* ipv6_fl_list read_mostly tcp_v6_connect,__ip6_datagram_connect,udpv6_sendmsg,rawv6_sendmsg +be16 inet_sport read_mostly __tcp_transmit_skb +be32 inet_daddr read_mostly ip_select_ident_segs +be32 inet_rcv_saddr +be16 inet_dport read_mostly __tcp_transmit_skb +u16 inet_num +be32 inet_saddr +s16 uc_ttl read_mostly __ip_queue_xmit/ip_select_ttl +u16 cmsg_flags +struct ip_options_rcu* inet_opt read_mostly __ip_queue_xmit +u16 inet_id read_mostly ip_select_ident_segs +u8 tos read_mostly ip_queue_xmit +u8 min_ttl +u8 mc_ttl +u8 pmtudisc +u8:1 recverr +u8:1 is_icsk +u8:1 freebind +u8:1 hdrincl +u8:1 mc_loop +u8:1 transparent +u8:1 mc_all +u8:1 nodefrag +u8:1 bind_address_no_port +u8:1 recverr_rfc4884 +u8:1 defer_connect read_mostly tcp_sendmsg_fastopen +u8 rcv_tos +u8 convert_csum +int uc_index +int mc_index +be32 mc_addr +struct ip_mc_socklist* mc_list +struct inet_cork_full cork read_mostly __tcp_transmit_skb +struct local_port_range +======================== ===================== =================== =================== ====================================================================================================== diff --git a/drivers/net/ethernet/chelsio/inline_crypto/chtls/chtls_cm.c b/drivers/net/ethernet/chelsio/inline_crypto/chtls/chtls_cm.c index 4ee970f3bad6..ee0154337a9c 100644 --- a/drivers/net/ethernet/chelsio/inline_crypto/chtls/chtls_cm.c +++ b/drivers/net/ethernet/chelsio/inline_crypto/chtls/chtls_cm.c @@ -1199,12 +1199,12 @@ static struct sock *chtls_recv_sock(struct sock *lsk, struct ipv6_pinfo *newnp = inet6_sk(newsk); struct ipv6_pinfo *np = inet6_sk(lsk); - inet_sk(newsk)->pinet6 = &newtcp6sk->inet6; + newinet->pinet6 = &newtcp6sk->inet6; + newinet->ipv6_fl_list = NULL; memcpy(newnp, np, sizeof(struct ipv6_pinfo)); newsk->sk_v6_daddr = treq->ir_v6_rmt_addr; newsk->sk_v6_rcv_saddr = treq->ir_v6_loc_addr; inet6_sk(newsk)->saddr = treq->ir_v6_loc_addr; - newnp->ipv6_fl_list = NULL; newnp->pktoptions = NULL; newsk->sk_bound_dev_if = treq->ir_iif; newinet->inet_opt = NULL; diff --git a/include/linux/ipv6.h b/include/linux/ipv6.h index 43b7bb828738..7294e4e89b79 100644 --- a/include/linux/ipv6.h +++ b/include/linux/ipv6.h @@ -271,7 +271,6 @@ struct ipv6_pinfo { struct ipv6_mc_socklist __rcu *ipv6_mc_list; struct ipv6_ac_socklist *ipv6_ac_list; - struct ipv6_fl_socklist __rcu *ipv6_fl_list; }; /* We currently use available bits from inet_sk(sk)->inet_flags, diff --git a/include/net/inet_sock.h b/include/net/inet_sock.h index 1086256549fa..b6ec08072533 100644 --- a/include/net/inet_sock.h +++ b/include/net/inet_sock.h @@ -214,6 +214,7 @@ struct inet_sock { struct sock sk; #if IS_ENABLED(CONFIG_IPV6) struct ipv6_pinfo *pinet6; + struct ipv6_fl_socklist __rcu *ipv6_fl_list; #endif /* Socket demultiplex comparisons on incoming packets. */ #define inet_daddr sk.__sk_common.skc_daddr diff --git a/net/ipv6/ip6_flowlabel.c b/net/ipv6/ip6_flowlabel.c index a3ff575798dd..60d0be47a9f3 100644 --- a/net/ipv6/ip6_flowlabel.c +++ b/net/ipv6/ip6_flowlabel.c @@ -66,8 +66,8 @@ EXPORT_SYMBOL(ipv6_flowlabel_exclusive); fl != NULL; \ fl = rcu_dereference(fl->next)) -#define for_each_sk_fl_rcu(np, sfl) \ - for (sfl = rcu_dereference(np->ipv6_fl_list); \ +#define for_each_sk_fl_rcu(sk, sfl) \ + for (sfl = rcu_dereference(inet_sk(sk)->ipv6_fl_list); \ sfl != NULL; \ sfl = rcu_dereference(sfl->next)) @@ -262,12 +262,11 @@ static struct ip6_flowlabel *fl_intern(struct net *net, struct ip6_flowlabel *__fl6_sock_lookup(struct sock *sk, __be32 label) { struct ipv6_fl_socklist *sfl; - struct ipv6_pinfo *np = inet6_sk(sk); label &= IPV6_FLOWLABEL_MASK; rcu_read_lock(); - for_each_sk_fl_rcu(np, sfl) { + for_each_sk_fl_rcu(sk, sfl) { struct ip6_flowlabel *fl = sfl->fl; if (fl->label == label && atomic_inc_not_zero(&fl->users)) { @@ -283,16 +282,16 @@ EXPORT_SYMBOL_GPL(__fl6_sock_lookup); void fl6_free_socklist(struct sock *sk) { - struct ipv6_pinfo *np = inet6_sk(sk); + struct inet_sock *inet = inet_sk(sk); struct ipv6_fl_socklist *sfl; - if (!rcu_access_pointer(np->ipv6_fl_list)) + if (!rcu_access_pointer(inet->ipv6_fl_list)) return; spin_lock_bh(&ip6_sk_fl_lock); - while ((sfl = rcu_dereference_protected(np->ipv6_fl_list, + while ((sfl = rcu_dereference_protected(inet->ipv6_fl_list, lockdep_is_held(&ip6_sk_fl_lock))) != NULL) { - np->ipv6_fl_list = sfl->next; + inet->ipv6_fl_list = sfl->next; spin_unlock_bh(&ip6_sk_fl_lock); fl_release(sfl->fl); @@ -470,16 +469,15 @@ done: static int mem_check(struct sock *sk) { - struct ipv6_pinfo *np = inet6_sk(sk); - struct ipv6_fl_socklist *sfl; int room = FL_MAX_SIZE - atomic_read(&fl_size); + struct ipv6_fl_socklist *sfl; int count = 0; if (room > FL_MAX_SIZE - FL_MAX_PER_SOCK) return 0; rcu_read_lock(); - for_each_sk_fl_rcu(np, sfl) + for_each_sk_fl_rcu(sk, sfl) count++; rcu_read_unlock(); @@ -492,13 +490,15 @@ static int mem_check(struct sock *sk) return 0; } -static inline void fl_link(struct ipv6_pinfo *np, struct ipv6_fl_socklist *sfl, - struct ip6_flowlabel *fl) +static inline void fl_link(struct sock *sk, struct ipv6_fl_socklist *sfl, + struct ip6_flowlabel *fl) { + struct inet_sock *inet = inet_sk(sk); + spin_lock_bh(&ip6_sk_fl_lock); sfl->fl = fl; - sfl->next = np->ipv6_fl_list; - rcu_assign_pointer(np->ipv6_fl_list, sfl); + sfl->next = inet->ipv6_fl_list; + rcu_assign_pointer(inet->ipv6_fl_list, sfl); spin_unlock_bh(&ip6_sk_fl_lock); } @@ -520,7 +520,7 @@ int ipv6_flowlabel_opt_get(struct sock *sk, struct in6_flowlabel_req *freq, rcu_read_lock(); - for_each_sk_fl_rcu(np, sfl) { + for_each_sk_fl_rcu(sk, sfl) { if (sfl->fl->label == (np->flow_label & IPV6_FLOWLABEL_MASK)) { spin_lock_bh(&ip6_fl_lock); freq->flr_label = sfl->fl->label; @@ -559,7 +559,7 @@ static int ipv6_flowlabel_put(struct sock *sk, struct in6_flowlabel_req *freq) } spin_lock_bh(&ip6_sk_fl_lock); - for (sflp = &np->ipv6_fl_list; + for (sflp = &inet_sk(sk)->ipv6_fl_list; (sfl = socklist_dereference(*sflp)) != NULL; sflp = &sfl->next) { if (sfl->fl->label == freq->flr_label) @@ -579,13 +579,12 @@ found: static int ipv6_flowlabel_renew(struct sock *sk, struct in6_flowlabel_req *freq) { - struct ipv6_pinfo *np = inet6_sk(sk); struct net *net = sock_net(sk); struct ipv6_fl_socklist *sfl; int err; rcu_read_lock(); - for_each_sk_fl_rcu(np, sfl) { + for_each_sk_fl_rcu(sk, sfl) { if (sfl->fl->label == freq->flr_label) { err = fl6_renew(sfl->fl, freq->flr_linger, freq->flr_expires); @@ -614,7 +613,6 @@ static int ipv6_flowlabel_get(struct sock *sk, struct in6_flowlabel_req *freq, { struct ipv6_fl_socklist *sfl, *sfl1 = NULL; struct ip6_flowlabel *fl, *fl1 = NULL; - struct ipv6_pinfo *np = inet6_sk(sk); struct net *net = sock_net(sk); int err; @@ -645,7 +643,7 @@ static int ipv6_flowlabel_get(struct sock *sk, struct in6_flowlabel_req *freq, if (freq->flr_label) { err = -EEXIST; rcu_read_lock(); - for_each_sk_fl_rcu(np, sfl) { + for_each_sk_fl_rcu(sk, sfl) { if (sfl->fl->label == freq->flr_label) { if (freq->flr_flags & IPV6_FL_F_EXCL) { rcu_read_unlock(); @@ -682,7 +680,7 @@ recheck: fl1->linger = fl->linger; if ((long)(fl->expires - fl1->expires) > 0) fl1->expires = fl->expires; - fl_link(np, sfl1, fl1); + fl_link(sk, sfl1, fl1); fl_free(fl); return 0; @@ -716,7 +714,7 @@ release: } } - fl_link(np, sfl1, fl); + fl_link(sk, sfl1, fl); return 0; done: fl_free(fl); diff --git a/net/ipv6/tcp_ipv6.c b/net/ipv6/tcp_ipv6.c index 59c4977a811a..6197dd4e6261 100644 --- a/net/ipv6/tcp_ipv6.c +++ b/net/ipv6/tcp_ipv6.c @@ -1386,7 +1386,9 @@ static struct sock *tcp_v6_syn_recv_sock(const struct sock *sk, struct sk_buff * if (!newsk) return NULL; - inet_sk(newsk)->pinet6 = tcp_inet6_sk(newsk); + newinet = inet_sk(newsk); + newinet->pinet6 = tcp_inet6_sk(newsk); + newinet->ipv6_fl_list = NULL; newnp = tcp_inet6_sk(newsk); newtp = tcp_sk(newsk); @@ -1405,7 +1407,6 @@ static struct sock *tcp_v6_syn_recv_sock(const struct sock *sk, struct sk_buff * newnp->ipv6_mc_list = NULL; newnp->ipv6_ac_list = NULL; - newnp->ipv6_fl_list = NULL; newnp->pktoptions = NULL; newnp->opt = NULL; newnp->mcast_oif = inet_iif(skb); @@ -1453,10 +1454,12 @@ static struct sock *tcp_v6_syn_recv_sock(const struct sock *sk, struct sk_buff * newsk->sk_gso_type = SKB_GSO_TCPV6; inet6_sk_rx_dst_set(newsk, skb); - inet_sk(newsk)->pinet6 = tcp_inet6_sk(newsk); + newinet = inet_sk(newsk); + newinet->pinet6 = tcp_inet6_sk(newsk); + newinet->ipv6_fl_list = NULL; + newinet->inet_opt = NULL; newtp = tcp_sk(newsk); - newinet = inet_sk(newsk); newnp = tcp_inet6_sk(newsk); memcpy(newnp, np, sizeof(struct ipv6_pinfo)); @@ -1469,10 +1472,8 @@ static struct sock *tcp_v6_syn_recv_sock(const struct sock *sk, struct sk_buff * First: no IPv4 options. */ - newinet->inet_opt = NULL; newnp->ipv6_mc_list = NULL; newnp->ipv6_ac_list = NULL; - newnp->ipv6_fl_list = NULL; /* Clone RX bits */ newnp->rxopt.all = np->rxopt.all; diff --git a/net/sctp/ipv6.c b/net/sctp/ipv6.c index 568ff8797c39..d725b2158758 100644 --- a/net/sctp/ipv6.c +++ b/net/sctp/ipv6.c @@ -782,9 +782,10 @@ static struct sock *sctp_v6_create_accept_sk(struct sock *sk, struct sctp_association *asoc, bool kern) { - struct sock *newsk; struct ipv6_pinfo *newnp, *np = inet6_sk(sk); struct sctp6_sock *newsctp6sk; + struct inet_sock *newinet; + struct sock *newsk; newsk = sk_alloc(sock_net(sk), PF_INET6, GFP_KERNEL, sk->sk_prot, kern); if (!newsk) @@ -796,7 +797,9 @@ static struct sock *sctp_v6_create_accept_sk(struct sock *sk, sock_reset_flag(sk, SOCK_ZAPPED); newsctp6sk = (struct sctp6_sock *)newsk; - inet_sk(newsk)->pinet6 = &newsctp6sk->inet6; + newinet = inet_sk(newsk); + newinet->pinet6 = &newsctp6sk->inet6; + newinet->ipv6_fl_list = NULL; sctp_sk(newsk)->v4mapped = sctp_sk(sk)->v4mapped; @@ -805,7 +808,6 @@ static struct sock *sctp_v6_create_accept_sk(struct sock *sk, memcpy(newnp, np, sizeof(struct ipv6_pinfo)); newnp->ipv6_mc_list = NULL; newnp->ipv6_ac_list = NULL; - newnp->ipv6_fl_list = NULL; sctp_v6_copy_ip_options(sk, newsk); -- cgit v1.2.3 From 1532ed0d0753c83e72595f785f82b48c28bbe5dc Mon Sep 17 00:00:00 2001 From: Xuanqiang Luo Date: Wed, 15 Oct 2025 10:02:35 +0800 Subject: inet: Avoid ehash lookup race in inet_ehash_insert() Since ehash lookups are lockless, if one CPU performs a lookup while another concurrently deletes and inserts (removing reqsk and inserting sk), the lookup may fail to find the socket, an RST may be sent. The call trace map is drawn as follows: CPU 0 CPU 1 ----- ----- inet_ehash_insert() spin_lock() sk_nulls_del_node_init_rcu(osk) __inet_lookup_established() (lookup failed) __sk_nulls_add_node_rcu(sk, list) spin_unlock() As both deletion and insertion operate on the same ehash chain, this patch introduces a new sk_nulls_replace_node_init_rcu() helper functions to implement atomic replacement. Fixes: 5e0724d027f0 ("tcp/dccp: fix hashdance race for passive sessions") Reviewed-by: Kuniyuki Iwashima Reviewed-by: Jiayuan Chen Signed-off-by: Xuanqiang Luo Reviewed-by: Eric Dumazet Link: https://patch.msgid.link/20251015020236.431822-3-xuanqiang.luo@linux.dev Signed-off-by: Jakub Kicinski --- include/net/sock.h | 13 +++++++++++++ net/ipv4/inet_hashtables.c | 8 ++++++-- 2 files changed, 19 insertions(+), 2 deletions(-) (limited to 'include/net') diff --git a/include/net/sock.h b/include/net/sock.h index 30ac2eb4ef9b..335d0da82d79 100644 --- a/include/net/sock.h +++ b/include/net/sock.h @@ -856,6 +856,19 @@ static inline bool sk_nulls_del_node_init_rcu(struct sock *sk) return rc; } +static inline bool sk_nulls_replace_node_init_rcu(struct sock *old, + struct sock *new) +{ + if (sk_hashed(old)) { + hlist_nulls_replace_init_rcu(&old->sk_nulls_node, + &new->sk_nulls_node); + __sock_put(old); + return true; + } + + return false; +} + static inline void __sk_add_node(struct sock *sk, struct hlist_head *list) { hlist_add_head(&sk->sk_node, list); diff --git a/net/ipv4/inet_hashtables.c b/net/ipv4/inet_hashtables.c index b7024e3d9ac3..f5826ec4bcaa 100644 --- a/net/ipv4/inet_hashtables.c +++ b/net/ipv4/inet_hashtables.c @@ -720,8 +720,11 @@ bool inet_ehash_insert(struct sock *sk, struct sock *osk, bool *found_dup_sk) spin_lock(lock); if (osk) { WARN_ON_ONCE(sk->sk_hash != osk->sk_hash); - ret = sk_nulls_del_node_init_rcu(osk); - } else if (found_dup_sk) { + ret = sk_nulls_replace_node_init_rcu(osk, sk); + goto unlock; + } + + if (found_dup_sk) { *found_dup_sk = inet_ehash_lookup_by_sk(sk, list); if (*found_dup_sk) ret = false; @@ -730,6 +733,7 @@ bool inet_ehash_insert(struct sock *sk, struct sock *osk, bool *found_dup_sk) if (ret) __sk_nulls_add_node_rcu(sk, list); +unlock: spin_unlock(lock); return ret; -- cgit v1.2.3 From 37a183d3b7cdb873e7f5f9daef1ad6d8f7c95fb7 Mon Sep 17 00:00:00 2001 From: Eric Biggers Date: Tue, 14 Oct 2025 14:58:36 -0700 Subject: tcp: Convert tcp-md5 to use MD5 library instead of crypto_ahash Make tcp-md5 use the MD5 library API (added in 6.18) instead of the crypto_ahash API. This is much simpler and also more efficient: - The library API just operates on struct md5_ctx. Just allocate this struct on the stack instead of using a pool of pre-allocated crypto_ahash and ahash_request objects. - The library API accepts standard pointers and doesn't require scatterlists. So, for hashing the headers just use an on-stack buffer instead of a pool of pre-allocated kmalloc'ed scratch buffers. - The library API never fails. Therefore, checking for MD5 hashing errors is no longer necessary. Update tcp_v4_md5_hash_skb(), tcp_v6_md5_hash_skb(), tcp_v4_md5_hash_hdr(), tcp_v6_md5_hash_hdr(), tcp_md5_hash_key(), tcp_sock_af_ops::calc_md5_hash, and tcp_request_sock_ops::calc_md5_hash to return void instead of int. - The library API provides direct access to the MD5 code, eliminating unnecessary overhead such as indirect function calls and scatterlist management. Microbenchmarks of tcp_v4_md5_hash_skb() on x86_64 show a speedup from 7518 to 7041 cycles (6% fewer) with skb->len == 1440, or from 1020 to 678 cycles (33% fewer) with skb->len == 140. Since tcp_sigpool_hash_skb_data() can no longer be used, add a function tcp_md5_hash_skb_data() which is specialized to MD5. Of course, to the extent that this duplicates any code, it's well worth it. To preserve the existing behavior of TCP-MD5 support being disabled when the kernel is booted with "fips=1", make tcp_md5_do_add() check fips_enabled itself. Previously it relied on the error from crypto_alloc_ahash("md5") being bubbled up. I don't know for sure that this is actually needed, but this preserves the existing behavior. Tested with bidirectional TCP-MD5, both IPv4 and IPv6, between a kernel that includes this commit and a kernel that doesn't include this commit. (Side note: please don't use TCP-MD5! It's cryptographically weak. But as long as Linux supports it, it might as well be implemented properly.) Signed-off-by: Eric Biggers Link: https://patch.msgid.link/20251014215836.115616-1-ebiggers@kernel.org Signed-off-by: Jakub Kicinski --- include/net/tcp.h | 26 +++------ net/ipv4/Kconfig | 4 +- net/ipv4/tcp.c | 73 +++++++++++-------------- net/ipv4/tcp_ipv4.c | 137 +++++++++++++++-------------------------------- net/ipv4/tcp_minisocks.c | 2 - net/ipv6/tcp_ipv6.c | 119 +++++++++++++--------------------------- 6 files changed, 121 insertions(+), 240 deletions(-) (limited to 'include/net') diff --git a/include/net/tcp.h b/include/net/tcp.h index 1e547138f4fb..67fdd2523d92 100644 --- a/include/net/tcp.h +++ b/include/net/tcp.h @@ -1898,13 +1898,6 @@ struct tcp6_pseudohdr { __be32 protocol; /* including padding */ }; -union tcp_md5sum_block { - struct tcp4_pseudohdr ip4; -#if IS_ENABLED(CONFIG_IPV6) - struct tcp6_pseudohdr ip6; -#endif -}; - /* * struct tcp_sigpool - per-CPU pool of ahash_requests * @scratch: per-CPU temporary area, that can be used between @@ -1939,8 +1932,8 @@ int tcp_sigpool_start(unsigned int id, struct tcp_sigpool *c); void tcp_sigpool_end(struct tcp_sigpool *c); size_t tcp_sigpool_algo(unsigned int id, char *buf, size_t buf_len); /* - functions */ -int tcp_v4_md5_hash_skb(char *md5_hash, const struct tcp_md5sig_key *key, - const struct sock *sk, const struct sk_buff *skb); +void tcp_v4_md5_hash_skb(char *md5_hash, const struct tcp_md5sig_key *key, + const struct sock *sk, const struct sk_buff *skb); int tcp_md5_do_add(struct sock *sk, const union tcp_md5_addr *addr, int family, u8 prefixlen, int l3index, u8 flags, const u8 *newkey, u8 newkeylen); @@ -1999,13 +1992,10 @@ static inline void tcp_md5_destruct_sock(struct sock *sk) } #endif -int tcp_md5_alloc_sigpool(void); -void tcp_md5_release_sigpool(void); -void tcp_md5_add_sigpool(void); -extern int tcp_md5_sigpool_id; - -int tcp_md5_hash_key(struct tcp_sigpool *hp, - const struct tcp_md5sig_key *key); +struct md5_ctx; +void tcp_md5_hash_skb_data(struct md5_ctx *ctx, const struct sk_buff *skb, + unsigned int header_len); +void tcp_md5_hash_key(struct md5_ctx *ctx, const struct tcp_md5sig_key *key); /* From tcp_fastopen.c */ void tcp_fastopen_cache_get(struct sock *sk, u16 *mss, @@ -2355,7 +2345,7 @@ struct tcp_sock_af_ops { #ifdef CONFIG_TCP_MD5SIG struct tcp_md5sig_key *(*md5_lookup) (const struct sock *sk, const struct sock *addr_sk); - int (*calc_md5_hash)(char *location, + void (*calc_md5_hash)(char *location, const struct tcp_md5sig_key *md5, const struct sock *sk, const struct sk_buff *skb); @@ -2383,7 +2373,7 @@ struct tcp_request_sock_ops { #ifdef CONFIG_TCP_MD5SIG struct tcp_md5sig_key *(*req_md5_lookup)(const struct sock *sk, const struct sock *addr_sk); - int (*calc_md5_hash) (char *location, + void (*calc_md5_hash) (char *location, const struct tcp_md5sig_key *md5, const struct sock *sk, const struct sk_buff *skb); diff --git a/net/ipv4/Kconfig b/net/ipv4/Kconfig index 12850a277251..b71c22475c51 100644 --- a/net/ipv4/Kconfig +++ b/net/ipv4/Kconfig @@ -760,9 +760,7 @@ config TCP_AO config TCP_MD5SIG bool "TCP: MD5 Signature Option support (RFC2385)" - select CRYPTO - select CRYPTO_MD5 - select TCP_SIGPOOL + select CRYPTO_LIB_MD5 help RFC2385 specifies a method of giving MD5 protection to TCP sessions. Its main (only?) use is to protect BGP sessions between core routers diff --git a/net/ipv4/tcp.c b/net/ipv4/tcp.c index 4d720aa09a4c..0ccc5405e740 100644 --- a/net/ipv4/tcp.c +++ b/net/ipv4/tcp.c @@ -243,7 +243,7 @@ #define pr_fmt(fmt) "TCP: " fmt -#include +#include #include #include #include @@ -253,7 +253,6 @@ #include #include #include -#include #include #include #include @@ -425,7 +424,6 @@ void tcp_md5_destruct_sock(struct sock *sk) tcp_clear_md5_list(sk); kfree(rcu_replace_pointer(tp->md5sig_info, NULL, 1)); static_branch_slow_dec_deferred(&tcp_md5_needed); - tcp_md5_release_sigpool(); } } EXPORT_IPV6_MOD_GPL(tcp_md5_destruct_sock); @@ -4838,52 +4836,45 @@ int tcp_getsockopt(struct sock *sk, int level, int optname, char __user *optval, EXPORT_IPV6_MOD(tcp_getsockopt); #ifdef CONFIG_TCP_MD5SIG -int tcp_md5_sigpool_id = -1; -EXPORT_IPV6_MOD_GPL(tcp_md5_sigpool_id); - -int tcp_md5_alloc_sigpool(void) +void tcp_md5_hash_skb_data(struct md5_ctx *ctx, const struct sk_buff *skb, + unsigned int header_len) { - size_t scratch_size; - int ret; + const unsigned int head_data_len = skb_headlen(skb) > header_len ? + skb_headlen(skb) - header_len : 0; + const struct skb_shared_info *shi = skb_shinfo(skb); + struct sk_buff *frag_iter; + unsigned int i; - scratch_size = sizeof(union tcp_md5sum_block) + sizeof(struct tcphdr); - ret = tcp_sigpool_alloc_ahash("md5", scratch_size); - if (ret >= 0) { - /* As long as any md5 sigpool was allocated, the return - * id would stay the same. Re-write the id only for the case - * when previously all MD5 keys were deleted and this call - * allocates the first MD5 key, which may return a different - * sigpool id than was used previously. - */ - WRITE_ONCE(tcp_md5_sigpool_id, ret); /* Avoids the compiler potentially being smart here */ - return 0; - } - return ret; -} + md5_update(ctx, (const u8 *)tcp_hdr(skb) + header_len, head_data_len); -void tcp_md5_release_sigpool(void) -{ - tcp_sigpool_release(READ_ONCE(tcp_md5_sigpool_id)); -} + for (i = 0; i < shi->nr_frags; ++i) { + const skb_frag_t *f = &shi->frags[i]; + u32 p_off, p_len, copied; + const void *vaddr; + struct page *p; -void tcp_md5_add_sigpool(void) -{ - tcp_sigpool_get(READ_ONCE(tcp_md5_sigpool_id)); + skb_frag_foreach_page(f, skb_frag_off(f), skb_frag_size(f), + p, p_off, p_len, copied) { + vaddr = kmap_local_page(p); + md5_update(ctx, vaddr + p_off, p_len); + kunmap_local(vaddr); + } + } + + skb_walk_frags(skb, frag_iter) + tcp_md5_hash_skb_data(ctx, frag_iter, 0); } +EXPORT_IPV6_MOD(tcp_md5_hash_skb_data); -int tcp_md5_hash_key(struct tcp_sigpool *hp, - const struct tcp_md5sig_key *key) +void tcp_md5_hash_key(struct md5_ctx *ctx, + const struct tcp_md5sig_key *key) { u8 keylen = READ_ONCE(key->keylen); /* paired with WRITE_ONCE() in tcp_md5_do_add */ - struct scatterlist sg; - - sg_init_one(&sg, key->key, keylen); - ahash_request_set_crypt(hp->req, &sg, NULL, keylen); /* We use data_race() because tcp_md5_do_add() might change * key->key under us */ - return data_race(crypto_ahash_update(hp->req)); + data_race(({ md5_update(ctx, key->key, keylen), 0; })); } EXPORT_IPV6_MOD(tcp_md5_hash_key); @@ -4902,7 +4893,6 @@ tcp_inbound_md5_hash(const struct sock *sk, const struct sk_buff *skb, const struct tcp_sock *tp = tcp_sk(sk); struct tcp_md5sig_key *key; u8 newhash[16]; - int genhash; key = tcp_md5_do_lookup(sk, l3index, saddr, family); @@ -4917,11 +4907,10 @@ tcp_inbound_md5_hash(const struct sock *sk, const struct sk_buff *skb, * IPv4-mapped case. */ if (family == AF_INET) - genhash = tcp_v4_md5_hash_skb(newhash, key, NULL, skb); + tcp_v4_md5_hash_skb(newhash, key, NULL, skb); else - genhash = tp->af_specific->calc_md5_hash(newhash, key, - NULL, skb); - if (genhash || memcmp(hash_location, newhash, 16) != 0) { + tp->af_specific->calc_md5_hash(newhash, key, NULL, skb); + if (memcmp(hash_location, newhash, 16) != 0) { NET_INC_STATS(sock_net(sk), LINUX_MIB_TCPMD5FAILURE); trace_tcp_hash_md5_mismatch(sk, skb); return SKB_DROP_REASON_TCP_MD5FAILURE; diff --git a/net/ipv4/tcp_ipv4.c b/net/ipv4/tcp_ipv4.c index b1fcf3e4e1ce..40a76da5364a 100644 --- a/net/ipv4/tcp_ipv4.c +++ b/net/ipv4/tcp_ipv4.c @@ -53,6 +53,7 @@ #include #include #include +#include #include #include #include @@ -86,14 +87,13 @@ #include #include -#include -#include +#include #include #ifdef CONFIG_TCP_MD5SIG -static int tcp_v4_md5_hash_hdr(char *md5_hash, const struct tcp_md5sig_key *key, - __be32 daddr, __be32 saddr, const struct tcphdr *th); +static void tcp_v4_md5_hash_hdr(char *md5_hash, const struct tcp_md5sig_key *key, + __be32 daddr, __be32 saddr, const struct tcphdr *th); #endif struct inet_hashinfo tcp_hashinfo; @@ -754,7 +754,6 @@ static void tcp_v4_send_reset(const struct sock *sk, struct sk_buff *skb, struct tcp_md5sig_key *key = NULL; unsigned char newhash[16]; struct sock *sk1 = NULL; - int genhash; #endif u64 transmit_time = 0; struct sock *ctl_sk; @@ -840,11 +839,9 @@ static void tcp_v4_send_reset(const struct sock *sk, struct sk_buff *skb, if (!key) goto out; - - genhash = tcp_v4_md5_hash_skb(newhash, key, NULL, skb); - if (genhash || memcmp(md5_hash_location, newhash, 16) != 0) + tcp_v4_md5_hash_skb(newhash, key, NULL, skb); + if (memcmp(md5_hash_location, newhash, 16) != 0) goto out; - } if (key) { @@ -1425,13 +1422,13 @@ int tcp_md5_do_add(struct sock *sk, const union tcp_md5_addr *addr, struct tcp_sock *tp = tcp_sk(sk); if (!rcu_dereference_protected(tp->md5sig_info, lockdep_sock_is_held(sk))) { - if (tcp_md5_alloc_sigpool()) - return -ENOMEM; + if (fips_enabled) { + pr_warn_once("TCP-MD5 support is disabled due to FIPS\n"); + return -EOPNOTSUPP; + } - if (tcp_md5sig_info_add(sk, GFP_KERNEL)) { - tcp_md5_release_sigpool(); + if (tcp_md5sig_info_add(sk, GFP_KERNEL)) return -ENOMEM; - } if (!static_branch_inc(&tcp_md5_needed.key)) { struct tcp_md5sig_info *md5sig; @@ -1439,7 +1436,6 @@ int tcp_md5_do_add(struct sock *sk, const union tcp_md5_addr *addr, md5sig = rcu_dereference_protected(tp->md5sig_info, lockdep_sock_is_held(sk)); rcu_assign_pointer(tp->md5sig_info, NULL); kfree_rcu(md5sig, rcu); - tcp_md5_release_sigpool(); return -EUSERS; } } @@ -1456,12 +1452,9 @@ int tcp_md5_key_copy(struct sock *sk, const union tcp_md5_addr *addr, struct tcp_sock *tp = tcp_sk(sk); if (!rcu_dereference_protected(tp->md5sig_info, lockdep_sock_is_held(sk))) { - tcp_md5_add_sigpool(); - if (tcp_md5sig_info_add(sk, sk_gfp_mask(sk, GFP_ATOMIC))) { - tcp_md5_release_sigpool(); + if (tcp_md5sig_info_add(sk, sk_gfp_mask(sk, GFP_ATOMIC))) return -ENOMEM; - } if (!static_key_fast_inc_not_disabled(&tcp_md5_needed.key.key)) { struct tcp_md5sig_info *md5sig; @@ -1470,7 +1463,6 @@ int tcp_md5_key_copy(struct sock *sk, const union tcp_md5_addr *addr, net_warn_ratelimited("Too many TCP-MD5 keys in the system\n"); rcu_assign_pointer(tp->md5sig_info, NULL); kfree_rcu(md5sig, rcu); - tcp_md5_release_sigpool(); return -EUSERS; } } @@ -1578,66 +1570,44 @@ static int tcp_v4_parse_md5_keys(struct sock *sk, int optname, cmd.tcpm_key, cmd.tcpm_keylen); } -static int tcp_v4_md5_hash_headers(struct tcp_sigpool *hp, - __be32 daddr, __be32 saddr, - const struct tcphdr *th, int nbytes) +static void tcp_v4_md5_hash_headers(struct md5_ctx *ctx, + __be32 daddr, __be32 saddr, + const struct tcphdr *th, int nbytes) { - struct tcp4_pseudohdr *bp; - struct scatterlist sg; - struct tcphdr *_th; - - bp = hp->scratch; - bp->saddr = saddr; - bp->daddr = daddr; - bp->pad = 0; - bp->protocol = IPPROTO_TCP; - bp->len = cpu_to_be16(nbytes); - - _th = (struct tcphdr *)(bp + 1); - memcpy(_th, th, sizeof(*th)); - _th->check = 0; + struct { + struct tcp4_pseudohdr ip; + struct tcphdr tcp; + } h; - sg_init_one(&sg, bp, sizeof(*bp) + sizeof(*th)); - ahash_request_set_crypt(hp->req, &sg, NULL, - sizeof(*bp) + sizeof(*th)); - return crypto_ahash_update(hp->req); + h.ip.saddr = saddr; + h.ip.daddr = daddr; + h.ip.pad = 0; + h.ip.protocol = IPPROTO_TCP; + h.ip.len = cpu_to_be16(nbytes); + h.tcp = *th; + h.tcp.check = 0; + md5_update(ctx, (const u8 *)&h, sizeof(h.ip) + sizeof(h.tcp)); } -static int tcp_v4_md5_hash_hdr(char *md5_hash, const struct tcp_md5sig_key *key, - __be32 daddr, __be32 saddr, const struct tcphdr *th) +static noinline_for_stack void +tcp_v4_md5_hash_hdr(char *md5_hash, const struct tcp_md5sig_key *key, + __be32 daddr, __be32 saddr, const struct tcphdr *th) { - struct tcp_sigpool hp; + struct md5_ctx ctx; - if (tcp_sigpool_start(tcp_md5_sigpool_id, &hp)) - goto clear_hash_nostart; - - if (crypto_ahash_init(hp.req)) - goto clear_hash; - if (tcp_v4_md5_hash_headers(&hp, daddr, saddr, th, th->doff << 2)) - goto clear_hash; - if (tcp_md5_hash_key(&hp, key)) - goto clear_hash; - ahash_request_set_crypt(hp.req, NULL, md5_hash, 0); - if (crypto_ahash_final(hp.req)) - goto clear_hash; - - tcp_sigpool_end(&hp); - return 0; - -clear_hash: - tcp_sigpool_end(&hp); -clear_hash_nostart: - memset(md5_hash, 0, 16); - return 1; + md5_init(&ctx); + tcp_v4_md5_hash_headers(&ctx, daddr, saddr, th, th->doff << 2); + tcp_md5_hash_key(&ctx, key); + md5_final(&ctx, md5_hash); } -int tcp_v4_md5_hash_skb(char *md5_hash, const struct tcp_md5sig_key *key, - const struct sock *sk, - const struct sk_buff *skb) +noinline_for_stack void +tcp_v4_md5_hash_skb(char *md5_hash, const struct tcp_md5sig_key *key, + const struct sock *sk, const struct sk_buff *skb) { const struct tcphdr *th = tcp_hdr(skb); - struct tcp_sigpool hp; __be32 saddr, daddr; + struct md5_ctx ctx; if (sk) { /* valid for establish/request sockets */ saddr = sk->sk_rcv_saddr; @@ -1648,30 +1618,11 @@ int tcp_v4_md5_hash_skb(char *md5_hash, const struct tcp_md5sig_key *key, daddr = iph->daddr; } - if (tcp_sigpool_start(tcp_md5_sigpool_id, &hp)) - goto clear_hash_nostart; - - if (crypto_ahash_init(hp.req)) - goto clear_hash; - - if (tcp_v4_md5_hash_headers(&hp, daddr, saddr, th, skb->len)) - goto clear_hash; - if (tcp_sigpool_hash_skb_data(&hp, skb, th->doff << 2)) - goto clear_hash; - if (tcp_md5_hash_key(&hp, key)) - goto clear_hash; - ahash_request_set_crypt(hp.req, NULL, md5_hash, 0); - if (crypto_ahash_final(hp.req)) - goto clear_hash; - - tcp_sigpool_end(&hp); - return 0; - -clear_hash: - tcp_sigpool_end(&hp); -clear_hash_nostart: - memset(md5_hash, 0, 16); - return 1; + md5_init(&ctx); + tcp_v4_md5_hash_headers(&ctx, daddr, saddr, th, skb->len); + tcp_md5_hash_skb_data(&ctx, skb, th->doff << 2); + tcp_md5_hash_key(&ctx, key); + md5_final(&ctx, md5_hash); } EXPORT_IPV6_MOD(tcp_v4_md5_hash_skb); diff --git a/net/ipv4/tcp_minisocks.c b/net/ipv4/tcp_minisocks.c index 2ec8c6f1cdcc..ded2cf1f6006 100644 --- a/net/ipv4/tcp_minisocks.c +++ b/net/ipv4/tcp_minisocks.c @@ -312,7 +312,6 @@ static void tcp_time_wait_init(struct sock *sk, struct tcp_timewait_sock *tcptw) return; if (!static_key_fast_inc_not_disabled(&tcp_md5_needed.key.key)) goto out_free; - tcp_md5_add_sigpool(); } return; out_free: @@ -406,7 +405,6 @@ void tcp_twsk_destructor(struct sock *sk) if (twsk->tw_md5_key) { kfree(twsk->tw_md5_key); static_branch_slow_dec_deferred(&tcp_md5_needed); - tcp_md5_release_sigpool(); } } #endif diff --git a/net/ipv6/tcp_ipv6.c b/net/ipv6/tcp_ipv6.c index 6197dd4e6261..06eb90e4078e 100644 --- a/net/ipv6/tcp_ipv6.c +++ b/net/ipv6/tcp_ipv6.c @@ -67,8 +67,7 @@ #include #include -#include -#include +#include #include @@ -691,69 +690,45 @@ static int tcp_v6_parse_md5_keys(struct sock *sk, int optname, cmd.tcpm_key, cmd.tcpm_keylen); } -static int tcp_v6_md5_hash_headers(struct tcp_sigpool *hp, - const struct in6_addr *daddr, - const struct in6_addr *saddr, - const struct tcphdr *th, int nbytes) +static void tcp_v6_md5_hash_headers(struct md5_ctx *ctx, + const struct in6_addr *daddr, + const struct in6_addr *saddr, + const struct tcphdr *th, int nbytes) { - struct tcp6_pseudohdr *bp; - struct scatterlist sg; - struct tcphdr *_th; - - bp = hp->scratch; - /* 1. TCP pseudo-header (RFC2460) */ - bp->saddr = *saddr; - bp->daddr = *daddr; - bp->protocol = cpu_to_be32(IPPROTO_TCP); - bp->len = cpu_to_be32(nbytes); - - _th = (struct tcphdr *)(bp + 1); - memcpy(_th, th, sizeof(*th)); - _th->check = 0; - - sg_init_one(&sg, bp, sizeof(*bp) + sizeof(*th)); - ahash_request_set_crypt(hp->req, &sg, NULL, - sizeof(*bp) + sizeof(*th)); - return crypto_ahash_update(hp->req); + struct { + struct tcp6_pseudohdr ip; /* TCP pseudo-header (RFC2460) */ + struct tcphdr tcp; + } h; + + h.ip.saddr = *saddr; + h.ip.daddr = *daddr; + h.ip.protocol = cpu_to_be32(IPPROTO_TCP); + h.ip.len = cpu_to_be32(nbytes); + h.tcp = *th; + h.tcp.check = 0; + md5_update(ctx, (const u8 *)&h, sizeof(h.ip) + sizeof(h.tcp)); } -static int tcp_v6_md5_hash_hdr(char *md5_hash, const struct tcp_md5sig_key *key, - const struct in6_addr *daddr, struct in6_addr *saddr, - const struct tcphdr *th) +static noinline_for_stack void +tcp_v6_md5_hash_hdr(char *md5_hash, const struct tcp_md5sig_key *key, + const struct in6_addr *daddr, struct in6_addr *saddr, + const struct tcphdr *th) { - struct tcp_sigpool hp; - - if (tcp_sigpool_start(tcp_md5_sigpool_id, &hp)) - goto clear_hash_nostart; - - if (crypto_ahash_init(hp.req)) - goto clear_hash; - if (tcp_v6_md5_hash_headers(&hp, daddr, saddr, th, th->doff << 2)) - goto clear_hash; - if (tcp_md5_hash_key(&hp, key)) - goto clear_hash; - ahash_request_set_crypt(hp.req, NULL, md5_hash, 0); - if (crypto_ahash_final(hp.req)) - goto clear_hash; - - tcp_sigpool_end(&hp); - return 0; + struct md5_ctx ctx; -clear_hash: - tcp_sigpool_end(&hp); -clear_hash_nostart: - memset(md5_hash, 0, 16); - return 1; + md5_init(&ctx); + tcp_v6_md5_hash_headers(&ctx, daddr, saddr, th, th->doff << 2); + tcp_md5_hash_key(&ctx, key); + md5_final(&ctx, md5_hash); } -static int tcp_v6_md5_hash_skb(char *md5_hash, - const struct tcp_md5sig_key *key, - const struct sock *sk, - const struct sk_buff *skb) +static noinline_for_stack void +tcp_v6_md5_hash_skb(char *md5_hash, const struct tcp_md5sig_key *key, + const struct sock *sk, const struct sk_buff *skb) { const struct tcphdr *th = tcp_hdr(skb); const struct in6_addr *saddr, *daddr; - struct tcp_sigpool hp; + struct md5_ctx ctx; if (sk) { /* valid for establish/request sockets */ saddr = &sk->sk_v6_rcv_saddr; @@ -764,30 +739,11 @@ static int tcp_v6_md5_hash_skb(char *md5_hash, daddr = &ip6h->daddr; } - if (tcp_sigpool_start(tcp_md5_sigpool_id, &hp)) - goto clear_hash_nostart; - - if (crypto_ahash_init(hp.req)) - goto clear_hash; - - if (tcp_v6_md5_hash_headers(&hp, daddr, saddr, th, skb->len)) - goto clear_hash; - if (tcp_sigpool_hash_skb_data(&hp, skb, th->doff << 2)) - goto clear_hash; - if (tcp_md5_hash_key(&hp, key)) - goto clear_hash; - ahash_request_set_crypt(hp.req, NULL, md5_hash, 0); - if (crypto_ahash_final(hp.req)) - goto clear_hash; - - tcp_sigpool_end(&hp); - return 0; - -clear_hash: - tcp_sigpool_end(&hp); -clear_hash_nostart: - memset(md5_hash, 0, 16); - return 1; + md5_init(&ctx); + tcp_v6_md5_hash_headers(&ctx, daddr, saddr, th, skb->len); + tcp_md5_hash_skb_data(&ctx, skb, th->doff << 2); + tcp_md5_hash_key(&ctx, key); + md5_final(&ctx, md5_hash); } #endif @@ -1032,7 +988,6 @@ static void tcp_v6_send_reset(const struct sock *sk, struct sk_buff *skb, int oif = 0; #ifdef CONFIG_TCP_MD5SIG unsigned char newhash[16]; - int genhash; struct sock *sk1 = NULL; #endif @@ -1091,8 +1046,8 @@ static void tcp_v6_send_reset(const struct sock *sk, struct sk_buff *skb, goto out; key.type = TCP_KEY_MD5; - genhash = tcp_v6_md5_hash_skb(newhash, key.md5_key, NULL, skb); - if (genhash || memcmp(md5_hash_location, newhash, 16) != 0) + tcp_v6_md5_hash_skb(newhash, key.md5_key, NULL, skb); + if (memcmp(md5_hash_location, newhash, 16) != 0) goto out; } #endif -- cgit v1.2.3 From 370157293175a702036203faec3e0495b081f135 Mon Sep 17 00:00:00 2001 From: Randy Dunlap Date: Wed, 15 Oct 2025 20:59:17 -0700 Subject: nl802154: fix some kernel-doc warnings Correct multiple kernel-doc warnings in nl802154.h: - Fix a typo on one enum name to avoid a kernel-doc warning. - Drop 2 enum descriptions that are no longer needed. - Mark 2 internal enums as "private:" so that kernel-doc is not needed for them. Warning: nl802154.h:239 Enum value 'NL802154_CAP_ATTR_MAX_MAXBE' not described in enum 'nl802154_wpan_phy_capability_attr' Warning: nl802154.h:239 Excess enum value '%NL802154_CAP_ATTR_MIN_CCA_ED_LEVEL' description in 'nl802154_wpan_phy_capability_attr' Warning: nl802154.h:239 Excess enum value '%NL802154_CAP_ATTR_MAX_CCA_ED_LEVEL' description in 'nl802154_wpan_phy_capability_attr' Warning: nl802154.h:369 Enum value '__NL802154_CCA_OPT_ATTR_AFTER_LAST' not described in enum 'nl802154_cca_opts' Warning: nl802154.h:369 Enum value 'NL802154_CCA_OPT_ATTR_MAX' not described in enum 'nl802154_cca_opts' Signed-off-by: Randy Dunlap Reviewed-by: Simon Horman Link: https://patch.msgid.link/20251016035917.1148012-1-rdunlap@infradead.org Signed-off-by: Jakub Kicinski --- include/net/nl802154.h | 5 ++--- 1 file changed, 2 insertions(+), 3 deletions(-) (limited to 'include/net') diff --git a/include/net/nl802154.h b/include/net/nl802154.h index a994dea74596..442822746e92 100644 --- a/include/net/nl802154.h +++ b/include/net/nl802154.h @@ -191,14 +191,12 @@ enum nl802154_iftype { * @NL802154_CAP_ATTR_CHANNELS: a nested attribute for nl802154_channel_attr * @NL802154_CAP_ATTR_TX_POWERS: a nested attribute for * nl802154_wpan_phy_tx_power - * @NL802154_CAP_ATTR_MIN_CCA_ED_LEVEL: minimum value for cca_ed_level - * @NL802154_CAP_ATTR_MAX_CCA_ED_LEVEL: maximum value for cca_ed_level * @NL802154_CAP_ATTR_CCA_MODES: nl802154_cca_modes flags * @NL802154_CAP_ATTR_CCA_OPTS: nl802154_cca_opts flags * @NL802154_CAP_ATTR_MIN_MINBE: minimum of minbe value * @NL802154_CAP_ATTR_MAX_MINBE: maximum of minbe value * @NL802154_CAP_ATTR_MIN_MAXBE: minimum of maxbe value - * @NL802154_CAP_ATTR_MAX_MINBE: maximum of maxbe value + * @NL802154_CAP_ATTR_MAX_MAXBE: maximum of maxbe value * @NL802154_CAP_ATTR_MIN_CSMA_BACKOFFS: minimum of csma backoff value * @NL802154_CAP_ATTR_MAX_CSMA_BACKOFFS: maximum of csma backoffs value * @NL802154_CAP_ATTR_MIN_FRAME_RETRIES: minimum of frame retries value @@ -364,6 +362,7 @@ enum nl802154_cca_opts { NL802154_CCA_OPT_ENERGY_CARRIER_AND, NL802154_CCA_OPT_ENERGY_CARRIER_OR, + /* private: */ /* keep last */ __NL802154_CCA_OPT_ATTR_AFTER_LAST, NL802154_CCA_OPT_ATTR_MAX = __NL802154_CCA_OPT_ATTR_AFTER_LAST - 1 -- cgit v1.2.3 From 3ff9bcecce83f12169ab3e42671bd76554ca521a Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Fri, 17 Oct 2025 13:37:12 +0000 Subject: net: avoid extra access to sk->sk_wmem_alloc in sock_wfree() UDP TX packets destructor is sock_wfree(). It suffers from a cache line bouncing in sock_def_write_space_wfree(). Instead of reading sk->sk_wmem_alloc after we just did an atomic RMW on it, use __refcount_sub_and_test() to get the old value for free, and pass the new value to sock_def_write_space_wfree(). Add __sock_writeable() helper. Signed-off-by: Eric Dumazet Reviewed-by: Kuniyuki Iwashima Link: https://patch.msgid.link/20251017133712.2842665-1-edumazet@google.com Signed-off-by: Paolo Abeni --- include/net/sock.h | 6 +++++- net/core/sock.c | 14 ++++++++------ 2 files changed, 13 insertions(+), 7 deletions(-) (limited to 'include/net') diff --git a/include/net/sock.h b/include/net/sock.h index 5c564f114ae9..01ce231603db 100644 --- a/include/net/sock.h +++ b/include/net/sock.h @@ -2607,12 +2607,16 @@ static inline struct page_frag *sk_page_frag(struct sock *sk) bool sk_page_frag_refill(struct sock *sk, struct page_frag *pfrag); +static inline bool __sock_writeable(const struct sock *sk, int wmem_alloc) +{ + return wmem_alloc < (READ_ONCE(sk->sk_sndbuf) >> 1); +} /* * Default write policy as shown to user space via poll/select/SIGIO */ static inline bool sock_writeable(const struct sock *sk) { - return refcount_read(&sk->sk_wmem_alloc) < (READ_ONCE(sk->sk_sndbuf) >> 1); + return __sock_writeable(sk, refcount_read(&sk->sk_wmem_alloc)); } static inline gfp_t gfp_any(void) diff --git a/net/core/sock.c b/net/core/sock.c index b78533fb9268..a99132cc0965 100644 --- a/net/core/sock.c +++ b/net/core/sock.c @@ -155,7 +155,7 @@ static DEFINE_MUTEX(proto_list_mutex); static LIST_HEAD(proto_list); -static void sock_def_write_space_wfree(struct sock *sk); +static void sock_def_write_space_wfree(struct sock *sk, int wmem_alloc); static void sock_def_write_space(struct sock *sk); /** @@ -2659,16 +2659,18 @@ EXPORT_SYMBOL_GPL(sk_setup_caps); */ void sock_wfree(struct sk_buff *skb) { - struct sock *sk = skb->sk; unsigned int len = skb->truesize; + struct sock *sk = skb->sk; bool free; + int old; if (!sock_flag(sk, SOCK_USE_WRITE_QUEUE)) { if (sock_flag(sk, SOCK_RCU_FREE) && sk->sk_write_space == sock_def_write_space) { rcu_read_lock(); - free = refcount_sub_and_test(len, &sk->sk_wmem_alloc); - sock_def_write_space_wfree(sk); + free = __refcount_sub_and_test(len, &sk->sk_wmem_alloc, + &old); + sock_def_write_space_wfree(sk, old - len); rcu_read_unlock(); if (unlikely(free)) __sk_free(sk); @@ -3612,12 +3614,12 @@ static void sock_def_write_space(struct sock *sk) * for SOCK_RCU_FREE sockets under RCU read section and after putting * ->sk_wmem_alloc. */ -static void sock_def_write_space_wfree(struct sock *sk) +static void sock_def_write_space_wfree(struct sock *sk, int wmem_alloc) { /* Do not wake up a writer until he can make "significant" * progress. --DaveM */ - if (sock_writeable(sk)) { + if (__sock_writeable(sk, wmem_alloc)) { struct socket_wq *wq = rcu_dereference(sk->sk_wq); /* rely on refcount_sub from sock_wfree() */ -- cgit v1.2.3 From ca4709843b7e72f96976cd6b35bca148a4071673 Mon Sep 17 00:00:00 2001 From: David Yang Date: Fri, 17 Oct 2025 14:08:54 +0800 Subject: net: dsa: tag_yt921x: add support for Motorcomm YT921x tags Add support for Motorcomm YT921x tags, which includes a proper configurable ethertype field (default to 0x9988). Signed-off-by: David Yang Reviewed-by: Andrew Lunn Link: https://patch.msgid.link/20251017060859.326450-3-mmyangfl@gmail.com Signed-off-by: Jakub Kicinski --- include/net/dsa.h | 2 + include/uapi/linux/if_ether.h | 1 + net/dsa/Kconfig | 6 ++ net/dsa/Makefile | 1 + net/dsa/tag_yt921x.c | 141 ++++++++++++++++++++++++++++++++++++++++++ 5 files changed, 151 insertions(+) create mode 100644 net/dsa/tag_yt921x.c (limited to 'include/net') diff --git a/include/net/dsa.h b/include/net/dsa.h index d73ea0880066..67762fdaf3c7 100644 --- a/include/net/dsa.h +++ b/include/net/dsa.h @@ -55,6 +55,7 @@ struct tc_action; #define DSA_TAG_PROTO_LAN937X_VALUE 27 #define DSA_TAG_PROTO_VSC73XX_8021Q_VALUE 28 #define DSA_TAG_PROTO_BRCM_LEGACY_FCS_VALUE 29 +#define DSA_TAG_PROTO_YT921X_VALUE 30 enum dsa_tag_protocol { DSA_TAG_PROTO_NONE = DSA_TAG_PROTO_NONE_VALUE, @@ -87,6 +88,7 @@ enum dsa_tag_protocol { DSA_TAG_PROTO_RZN1_A5PSW = DSA_TAG_PROTO_RZN1_A5PSW_VALUE, DSA_TAG_PROTO_LAN937X = DSA_TAG_PROTO_LAN937X_VALUE, DSA_TAG_PROTO_VSC73XX_8021Q = DSA_TAG_PROTO_VSC73XX_8021Q_VALUE, + DSA_TAG_PROTO_YT921X = DSA_TAG_PROTO_YT921X_VALUE, }; struct dsa_switch; diff --git a/include/uapi/linux/if_ether.h b/include/uapi/linux/if_ether.h index 69e0457eb200..cfd200c87e5e 100644 --- a/include/uapi/linux/if_ether.h +++ b/include/uapi/linux/if_ether.h @@ -114,6 +114,7 @@ #define ETH_P_QINQ1 0x9100 /* deprecated QinQ VLAN [ NOT AN OFFICIALLY REGISTERED ID ] */ #define ETH_P_QINQ2 0x9200 /* deprecated QinQ VLAN [ NOT AN OFFICIALLY REGISTERED ID ] */ #define ETH_P_QINQ3 0x9300 /* deprecated QinQ VLAN [ NOT AN OFFICIALLY REGISTERED ID ] */ +#define ETH_P_YT921X 0x9988 /* Motorcomm YT921x DSA [ NOT AN OFFICIALLY REGISTERED ID ] */ #define ETH_P_EDSA 0xDADA /* Ethertype DSA [ NOT AN OFFICIALLY REGISTERED ID ] */ #define ETH_P_DSA_8021Q 0xDADB /* Fake VLAN Header for DSA [ NOT AN OFFICIALLY REGISTERED ID ] */ #define ETH_P_DSA_A5PSW 0xE001 /* A5PSW Tag Value [ NOT AN OFFICIALLY REGISTERED ID ] */ diff --git a/net/dsa/Kconfig b/net/dsa/Kconfig index 869cbe57162f..6b94028b1fcc 100644 --- a/net/dsa/Kconfig +++ b/net/dsa/Kconfig @@ -190,4 +190,10 @@ config NET_DSA_TAG_XRS700X Say Y or M if you want to enable support for tagging frames for Arrow SpeedChips XRS700x switches that use a single byte tag trailer. +config NET_DSA_TAG_YT921X + tristate "Tag driver for Motorcomm YT921x switches" + help + Say Y or M if you want to enable support for tagging frames for + Motorcomm YT921x switches. + endif diff --git a/net/dsa/Makefile b/net/dsa/Makefile index 555c07cfeb71..4b011a1d5c87 100644 --- a/net/dsa/Makefile +++ b/net/dsa/Makefile @@ -39,6 +39,7 @@ obj-$(CONFIG_NET_DSA_TAG_SJA1105) += tag_sja1105.o obj-$(CONFIG_NET_DSA_TAG_TRAILER) += tag_trailer.o obj-$(CONFIG_NET_DSA_TAG_VSC73XX_8021Q) += tag_vsc73xx_8021q.o obj-$(CONFIG_NET_DSA_TAG_XRS700X) += tag_xrs700x.o +obj-$(CONFIG_NET_DSA_TAG_YT921X) += tag_yt921x.o # for tracing framework to find trace.h CFLAGS_trace.o := -I$(src) diff --git a/net/dsa/tag_yt921x.c b/net/dsa/tag_yt921x.c new file mode 100644 index 000000000000..995da44f0a2a --- /dev/null +++ b/net/dsa/tag_yt921x.c @@ -0,0 +1,141 @@ +// SPDX-License-Identifier: GPL-2.0-or-later +/* + * Motorcomm YT921x Switch Extended CPU Port Tagging + * + * Copyright (c) 2025 David Yang + * + * +----+----+-------+-----+----+--------- + * | DA | SA | TagET | Tag | ET | Payload ... + * +----+----+-------+-----+----+--------- + * 6 6 2 6 2 N + * + * Tag Ethertype: CPU_TAG_TPID_TPID (default: ETH_P_YT921X = 0x9988) + * * Hardcoded for the moment, but still configurable. Discuss it if there + * are conflicts somewhere and/or you want to change it for some reason. + * Tag: + * 2: VLAN Tag + * 2: Rx Port + * 15b: Rx Port Valid + * 14b-11b: Rx Port + * 10b-0b: Cmd? + * 2: Tx Port(s) + * 15b: Tx Port(s) Valid + * 10b-0b: Tx Port(s) Mask + */ + +#include + +#include "tag.h" + +#define YT921X_TAG_NAME "yt921x" + +#define YT921X_TAG_LEN 8 + +#define YT921X_TAG_PORT_EN BIT(15) +#define YT921X_TAG_RX_PORT_M GENMASK(14, 11) +#define YT921X_TAG_RX_CMD_M GENMASK(10, 0) +#define YT921X_TAG_RX_CMD(x) FIELD_PREP(YT921X_TAG_RX_CMD_M, (x)) +#define YT921X_TAG_RX_CMD_FORWARDED 0x80 +#define YT921X_TAG_RX_CMD_UNK_UCAST 0xb2 +#define YT921X_TAG_RX_CMD_UNK_MCAST 0xb4 +#define YT921X_TAG_TX_PORTS_M GENMASK(10, 0) +#define YT921X_TAG_TX_PORTn(port) BIT(port) + +static struct sk_buff * +yt921x_tag_xmit(struct sk_buff *skb, struct net_device *netdev) +{ + struct dsa_port *dp = dsa_user_to_port(netdev); + unsigned int port = dp->index; + __be16 *tag; + u16 tx; + + skb_push(skb, YT921X_TAG_LEN); + dsa_alloc_etype_header(skb, YT921X_TAG_LEN); + + tag = dsa_etype_header_pos_tx(skb); + + tag[0] = htons(ETH_P_YT921X); + /* VLAN tag unrelated when TX */ + tag[1] = 0; + tag[2] = 0; + tx = YT921X_TAG_PORT_EN | YT921X_TAG_TX_PORTn(port); + tag[3] = htons(tx); + + return skb; +} + +static struct sk_buff * +yt921x_tag_rcv(struct sk_buff *skb, struct net_device *netdev) +{ + unsigned int port; + __be16 *tag; + u16 cmd; + u16 rx; + + if (unlikely(!pskb_may_pull(skb, YT921X_TAG_LEN))) + return NULL; + + tag = dsa_etype_header_pos_rx(skb); + + if (unlikely(tag[0] != htons(ETH_P_YT921X))) { + dev_warn_ratelimited(&netdev->dev, + "Unexpected EtherType 0x%04x\n", + ntohs(tag[0])); + return NULL; + } + + /* Locate which port this is coming from */ + rx = ntohs(tag[2]); + if (unlikely((rx & YT921X_TAG_PORT_EN) == 0)) { + dev_warn_ratelimited(&netdev->dev, + "Unexpected rx tag 0x%04x\n", rx); + return NULL; + } + + port = FIELD_GET(YT921X_TAG_RX_PORT_M, rx); + skb->dev = dsa_conduit_find_user(netdev, 0, port); + if (unlikely(!skb->dev)) { + dev_warn_ratelimited(&netdev->dev, + "Couldn't decode source port %u\n", port); + return NULL; + } + + cmd = FIELD_GET(YT921X_TAG_RX_CMD_M, rx); + switch (cmd) { + case YT921X_TAG_RX_CMD_FORWARDED: + /* Already forwarded by hardware */ + dsa_default_offload_fwd_mark(skb); + break; + case YT921X_TAG_RX_CMD_UNK_UCAST: + case YT921X_TAG_RX_CMD_UNK_MCAST: + /* NOTE: hardware doesn't distinguish between TRAP (copy to CPU + * only) and COPY (forward and copy to CPU). In order to perform + * a soft switch, NEVER use COPY action in the switch driver. + */ + break; + default: + dev_warn_ratelimited(&netdev->dev, + "Unexpected rx cmd 0x%02x\n", cmd); + break; + } + + /* Remove YT921x tag and update checksum */ + skb_pull_rcsum(skb, YT921X_TAG_LEN); + dsa_strip_etype_header(skb, YT921X_TAG_LEN); + + return skb; +} + +static const struct dsa_device_ops yt921x_netdev_ops = { + .name = YT921X_TAG_NAME, + .proto = DSA_TAG_PROTO_YT921X, + .xmit = yt921x_tag_xmit, + .rcv = yt921x_tag_rcv, + .needed_headroom = YT921X_TAG_LEN, +}; + +MODULE_DESCRIPTION("DSA tag driver for Motorcomm YT921x switches"); +MODULE_LICENSE("GPL"); +MODULE_ALIAS_DSA_TAG_DRIVER(DSA_TAG_PROTO_YT921X, YT921X_TAG_NAME); + +module_dsa_tag_driver(yt921x_netdev_ops); -- cgit v1.2.3 From 114573962a68a527835f2f1433a89bc2f9feac1b Mon Sep 17 00:00:00 2001 From: Yue Haibing Date: Tue, 21 Oct 2025 19:46:26 +0800 Subject: net/sched: Remove unused inline helper qdisc_from_priv() Since commit fb38306ceb9e ("net/sched: Retire ATM qdisc"), this is not used and can be removed. Signed-off-by: Yue Haibing Link: https://patch.msgid.link/20251021114626.3148894-1-yuehaibing@huawei.com Signed-off-by: Jakub Kicinski --- include/net/pkt_sched.h | 5 ----- 1 file changed, 5 deletions(-) (limited to 'include/net') diff --git a/include/net/pkt_sched.h b/include/net/pkt_sched.h index 8a75c73fc555..c660ac871083 100644 --- a/include/net/pkt_sched.h +++ b/include/net/pkt_sched.h @@ -25,11 +25,6 @@ struct qdisc_walker { const struct Qdisc * : (const void *)&q->privdata, \ struct Qdisc * : (void *)&q->privdata) -static inline struct Qdisc *qdisc_from_priv(void *priv) -{ - return container_of(priv, struct Qdisc, privdata); -} - /* Timer resolution MUST BE < 10% of min_schedulable_packet_size/bandwidth -- cgit v1.2.3 From 35d7c70870338aa6a367b9e4ed528914320b0be0 Mon Sep 17 00:00:00 2001 From: Kuniyuki Iwashima Date: Wed, 22 Oct 2025 05:39:46 +0000 Subject: neighbour: Annotate access to neigh_parms fields. NEIGH_VAR() is read locklessly in the fast path, and IPv6 ndisc uses NEIGH_VAR_SET() locklessly. The next patch will convert neightbl_dump_info() to RCU. Let's annotate accesses to neigh_param with READ_ONCE() and WRITE_ONCE(). Note that ndisc_ifinfo_sysctl_change() uses &NEIGH_VAR() and we cannot use '&' with READ_ONCE(), so NEIGH_VAR_PTR() is introduced. Note also that NEIGH_VAR_INIT() does not need WRITE_ONCE() as it is before parms is published. Also, the only user hippi_neigh_setup_dev() is no longer called since commit e3804cbebb67 ("net: remove COMPAT_NET_DEV_OPS"), which looks wrong, but probably no one uses HIPPI and RoadRunner. Signed-off-by: Kuniyuki Iwashima Reviewed-by: Eric Dumazet Link: https://patch.msgid.link/20251022054004.2514876-3-kuniyu@google.com Signed-off-by: Jakub Kicinski --- include/net/neighbour.h | 15 ++++++++++++--- net/core/neighbour.c | 17 ++++++----------- net/ipv6/ndisc.c | 8 ++++---- 3 files changed, 22 insertions(+), 18 deletions(-) (limited to 'include/net') diff --git a/include/net/neighbour.h b/include/net/neighbour.h index 4a30bd458c5a..998ff9eccebb 100644 --- a/include/net/neighbour.h +++ b/include/net/neighbour.h @@ -92,15 +92,17 @@ struct neigh_parms { static inline void neigh_var_set(struct neigh_parms *p, int index, int val) { set_bit(index, p->data_state); - p->data[index] = val; + WRITE_ONCE(p->data[index], val); } -#define NEIGH_VAR(p, attr) ((p)->data[NEIGH_VAR_ ## attr]) +#define __NEIGH_VAR(p, attr) ((p)->data[NEIGH_VAR_ ## attr]) +#define NEIGH_VAR(p, attr) READ_ONCE(__NEIGH_VAR(p, attr)) +#define NEIGH_VAR_PTR(p, attr) (&(__NEIGH_VAR(p, attr))) /* In ndo_neigh_setup, NEIGH_VAR_INIT should be used. * In other cases, NEIGH_VAR_SET should be used. */ -#define NEIGH_VAR_INIT(p, attr, val) (NEIGH_VAR(p, attr) = val) +#define NEIGH_VAR_INIT(p, attr, val) (__NEIGH_VAR(p, attr) = val) #define NEIGH_VAR_SET(p, attr, val) neigh_var_set(p, NEIGH_VAR_ ## attr, val) static inline void neigh_parms_data_state_setall(struct neigh_parms *p) @@ -378,6 +380,13 @@ struct net *neigh_parms_net(const struct neigh_parms *parms) unsigned long neigh_rand_reach_time(unsigned long base); +static inline void neigh_set_reach_time(struct neigh_parms *p) +{ + unsigned long base = NEIGH_VAR(p, BASE_REACHABLE_TIME); + + WRITE_ONCE(p->reachable_time, neigh_rand_reach_time(base)); +} + void pneigh_enqueue(struct neigh_table *tbl, struct neigh_parms *p, struct sk_buff *skb); struct pneigh_entry *pneigh_lookup(struct neigh_table *tbl, struct net *net, diff --git a/net/core/neighbour.c b/net/core/neighbour.c index 98428f60731b..5bbebbfcba43 100644 --- a/net/core/neighbour.c +++ b/net/core/neighbour.c @@ -995,8 +995,7 @@ static void neigh_periodic_work(struct work_struct *work) WRITE_ONCE(tbl->last_rand, jiffies); list_for_each_entry(p, &tbl->parms_list, list) - p->reachable_time = - neigh_rand_reach_time(NEIGH_VAR(p, BASE_REACHABLE_TIME)); + neigh_set_reach_time(p); } if (atomic_read(&tbl->entries) < READ_ONCE(tbl->gc_thresh1)) @@ -1749,8 +1748,7 @@ struct neigh_parms *neigh_parms_alloc(struct net_device *dev, if (p) { p->tbl = tbl; refcount_set(&p->refcnt, 1); - p->reachable_time = - neigh_rand_reach_time(NEIGH_VAR(p, BASE_REACHABLE_TIME)); + neigh_set_reach_time(p); p->qlen = 0; netdev_hold(dev, &p->dev_tracker, GFP_KERNEL); p->dev = dev; @@ -1810,8 +1808,7 @@ void neigh_table_init(int index, struct neigh_table *tbl) list_add(&tbl->parms.list, &tbl->parms_list); write_pnet(&tbl->parms.net, &init_net); refcount_set(&tbl->parms.refcnt, 1); - tbl->parms.reachable_time = - neigh_rand_reach_time(NEIGH_VAR(&tbl->parms, BASE_REACHABLE_TIME)); + neigh_set_reach_time(&tbl->parms); tbl->parms.qlen = 0; tbl->stats = alloc_percpu(struct neigh_statistics); @@ -2194,7 +2191,7 @@ static int neightbl_fill_parms(struct sk_buff *skb, struct neigh_parms *parms) NEIGH_VAR(parms, MCAST_PROBES)) || nla_put_u32(skb, NDTPA_MCAST_REPROBES, NEIGH_VAR(parms, MCAST_REPROBES)) || - nla_put_msecs(skb, NDTPA_REACHABLE_TIME, parms->reachable_time, + nla_put_msecs(skb, NDTPA_REACHABLE_TIME, READ_ONCE(parms->reachable_time), NDTPA_PAD) || nla_put_msecs(skb, NDTPA_BASE_REACHABLE_TIME, NEIGH_VAR(parms, BASE_REACHABLE_TIME), NDTPA_PAD) || @@ -2475,8 +2472,7 @@ static int neightbl_set(struct sk_buff *skb, struct nlmsghdr *nlh, * only be effective after the next time neigh_periodic_work * decides to recompute it (can be multiple minutes) */ - p->reachable_time = - neigh_rand_reach_time(NEIGH_VAR(p, BASE_REACHABLE_TIME)); + neigh_set_reach_time(p); break; case NDTPA_GC_STALETIME: NEIGH_VAR_SET(p, GC_STALETIME, @@ -3721,8 +3717,7 @@ static int neigh_proc_base_reachable_time(const struct ctl_table *ctl, int write * only be effective after the next time neigh_periodic_work * decides to recompute it */ - p->reachable_time = - neigh_rand_reach_time(NEIGH_VAR(p, BASE_REACHABLE_TIME)); + neigh_set_reach_time(p); } return ret; } diff --git a/net/ipv6/ndisc.c b/net/ipv6/ndisc.c index f427e41e9c49..59d17b6f06bf 100644 --- a/net/ipv6/ndisc.c +++ b/net/ipv6/ndisc.c @@ -1449,7 +1449,7 @@ skip_defrtr: BASE_REACHABLE_TIME, rtime); NEIGH_VAR_SET(in6_dev->nd_parms, GC_STALETIME, 3 * rtime); - in6_dev->nd_parms->reachable_time = neigh_rand_reach_time(rtime); + neigh_set_reach_time(in6_dev->nd_parms); in6_dev->tstamp = jiffies; send_ifinfo_notify = true; } @@ -1948,9 +1948,9 @@ int ndisc_ifinfo_sysctl_change(const struct ctl_table *ctl, int write, void *buf ret = -1; if (write && ret == 0 && dev && (idev = in6_dev_get(dev)) != NULL) { - if (ctl->data == &NEIGH_VAR(idev->nd_parms, BASE_REACHABLE_TIME)) - idev->nd_parms->reachable_time = - neigh_rand_reach_time(NEIGH_VAR(idev->nd_parms, BASE_REACHABLE_TIME)); + if (ctl->data == NEIGH_VAR_PTR(idev->nd_parms, BASE_REACHABLE_TIME)) + neigh_set_reach_time(idev->nd_parms); + WRITE_ONCE(idev->tstamp, jiffies); inet6_ifinfo_notify(RTM_NEWLINK, idev); in6_dev_put(idev); -- cgit v1.2.3 From 3064d0fe02af23a3956d2b690461abb44da88cf4 Mon Sep 17 00:00:00 2001 From: Kuniyuki Iwashima Date: Wed, 22 Oct 2025 05:39:49 +0000 Subject: neighbour: Convert rwlock of struct neigh_table to spinlock. Only neigh_for_each() and neigh_seq_start/stop() are on the reader side of neigh_table.lock. Let's convert rwlock to the plain spinlock. Signed-off-by: Kuniyuki Iwashima Reviewed-by: Eric Dumazet Link: https://patch.msgid.link/20251022054004.2514876-6-kuniyu@google.com Signed-off-by: Jakub Kicinski --- include/net/neighbour.h | 2 +- net/atm/clip.c | 4 +-- net/core/neighbour.c | 68 +++++++++++++++++++++++++------------------------ net/ipv4/arp.c | 4 +-- 4 files changed, 40 insertions(+), 38 deletions(-) (limited to 'include/net') diff --git a/include/net/neighbour.h b/include/net/neighbour.h index 998ff9eccebb..2dfee6d4258a 100644 --- a/include/net/neighbour.h +++ b/include/net/neighbour.h @@ -238,7 +238,7 @@ struct neigh_table { atomic_t gc_entries; struct list_head gc_list; struct list_head managed_list; - rwlock_t lock; + spinlock_t lock; unsigned long last_rand; struct neigh_statistics __percpu *stats; struct neigh_hash_table __rcu *nht; diff --git a/net/atm/clip.c b/net/atm/clip.c index f7a5565e794e..8f152e5fa659 100644 --- a/net/atm/clip.c +++ b/net/atm/clip.c @@ -168,10 +168,10 @@ static int neigh_check_cb(struct neighbour *n) static void idle_timer_check(struct timer_list *unused) { - write_lock(&arp_tbl.lock); + spin_lock(&arp_tbl.lock); __neigh_for_each_release(&arp_tbl, neigh_check_cb); mod_timer(&idle_timer, jiffies + CLIP_CHECK_INTERVAL * HZ); - write_unlock(&arp_tbl.lock); + spin_unlock(&arp_tbl.lock); } static int clip_arp_rcv(struct sk_buff *skb) diff --git a/net/core/neighbour.c b/net/core/neighbour.c index 6d2164b4d999..96a3b1a93252 100644 --- a/net/core/neighbour.c +++ b/net/core/neighbour.c @@ -81,7 +81,7 @@ static struct hlist_head *neigh_get_dev_table(struct net_device *dev, int family } /* - Neighbour hash table buckets are protected with rwlock tbl->lock. + Neighbour hash table buckets are protected with tbl->lock. - All the scans/updates to hash buckets MUST be made under this lock. - NOTHING clever should be made under this lock: no callbacks @@ -149,7 +149,7 @@ static void neigh_update_gc_list(struct neighbour *n) { bool on_gc_list, exempt_from_gc; - write_lock_bh(&n->tbl->lock); + spin_lock_bh(&n->tbl->lock); write_lock(&n->lock); if (n->dead) goto out; @@ -172,14 +172,14 @@ static void neigh_update_gc_list(struct neighbour *n) } out: write_unlock(&n->lock); - write_unlock_bh(&n->tbl->lock); + spin_unlock_bh(&n->tbl->lock); } static void neigh_update_managed_list(struct neighbour *n) { bool on_managed_list, add_to_managed; - write_lock_bh(&n->tbl->lock); + spin_lock_bh(&n->tbl->lock); write_lock(&n->lock); if (n->dead) goto out; @@ -193,7 +193,7 @@ static void neigh_update_managed_list(struct neighbour *n) list_add_tail(&n->managed_list, &n->tbl->managed_list); out: write_unlock(&n->lock); - write_unlock_bh(&n->tbl->lock); + spin_unlock_bh(&n->tbl->lock); } static void neigh_update_flags(struct neighbour *neigh, u32 flags, int *notify, @@ -263,7 +263,7 @@ static int neigh_forced_gc(struct neigh_table *tbl) NEIGH_CACHE_STAT_INC(tbl, forced_gc_runs); - write_lock_bh(&tbl->lock); + spin_lock_bh(&tbl->lock); list_for_each_entry_safe(n, tmp, &tbl->gc_list, gc_list) { if (refcount_read(&n->refcnt) == 1) { @@ -292,7 +292,7 @@ static int neigh_forced_gc(struct neigh_table *tbl) WRITE_ONCE(tbl->last_flush, jiffies); unlock: - write_unlock_bh(&tbl->lock); + spin_unlock_bh(&tbl->lock); return shrunk; } @@ -454,23 +454,23 @@ static void neigh_flush_table(struct neigh_table *tbl) void neigh_changeaddr(struct neigh_table *tbl, struct net_device *dev) { - write_lock_bh(&tbl->lock); + spin_lock_bh(&tbl->lock); neigh_flush_dev(tbl, dev, false); - write_unlock_bh(&tbl->lock); + spin_unlock_bh(&tbl->lock); } EXPORT_SYMBOL(neigh_changeaddr); static int __neigh_ifdown(struct neigh_table *tbl, struct net_device *dev, bool skip_perm) { - write_lock_bh(&tbl->lock); + spin_lock_bh(&tbl->lock); if (likely(dev)) { neigh_flush_dev(tbl, dev, skip_perm); } else { DEBUG_NET_WARN_ON_ONCE(skip_perm); neigh_flush_table(tbl); } - write_unlock_bh(&tbl->lock); + spin_unlock_bh(&tbl->lock); pneigh_ifdown(tbl, dev, skip_perm); pneigh_queue_purge(&tbl->proxy_queue, dev ? dev_net(dev) : NULL, @@ -687,7 +687,7 @@ ___neigh_create(struct neigh_table *tbl, const void *pkey, n->confirmed = jiffies - (NEIGH_VAR(n->parms, BASE_REACHABLE_TIME) << 1); - write_lock_bh(&tbl->lock); + spin_lock_bh(&tbl->lock); nht = rcu_dereference_protected(tbl->nht, lockdep_is_held(&tbl->lock)); @@ -722,13 +722,13 @@ ___neigh_create(struct neigh_table *tbl, const void *pkey, hlist_add_head_rcu(&n->dev_list, neigh_get_dev_table(dev, tbl->family)); - write_unlock_bh(&tbl->lock); + spin_unlock_bh(&tbl->lock); neigh_dbg(2, "neigh %p is created\n", n); rc = n; out: return rc; out_tbl_unlock: - write_unlock_bh(&tbl->lock); + spin_unlock_bh(&tbl->lock); out_neigh_release: if (!exempt_from_gc) atomic_dec(&tbl->gc_entries); @@ -982,7 +982,7 @@ static void neigh_periodic_work(struct work_struct *work) NEIGH_CACHE_STAT_INC(tbl, periodic_gc_runs); - write_lock_bh(&tbl->lock); + spin_lock_bh(&tbl->lock); nht = rcu_dereference_protected(tbl->nht, lockdep_is_held(&tbl->lock)); @@ -1036,9 +1036,9 @@ static void neigh_periodic_work(struct work_struct *work) * It's fine to release lock here, even if hash table * grows while we are preempted. */ - write_unlock_bh(&tbl->lock); + spin_unlock_bh(&tbl->lock); cond_resched(); - write_lock_bh(&tbl->lock); + spin_lock_bh(&tbl->lock); nht = rcu_dereference_protected(tbl->nht, lockdep_is_held(&tbl->lock)); } @@ -1049,7 +1049,7 @@ out: */ queue_delayed_work(system_power_efficient_wq, &tbl->gc_work, NEIGH_VAR(&tbl->parms, BASE_REACHABLE_TIME) >> 1); - write_unlock_bh(&tbl->lock); + spin_unlock_bh(&tbl->lock); } static __inline__ int neigh_max_probes(struct neighbour *n) @@ -1641,12 +1641,12 @@ static void neigh_managed_work(struct work_struct *work) managed_work.work); struct neighbour *neigh; - write_lock_bh(&tbl->lock); + spin_lock_bh(&tbl->lock); list_for_each_entry(neigh, &tbl->managed_list, managed_list) neigh_event_send_probe(neigh, NULL, false); queue_delayed_work(system_power_efficient_wq, &tbl->managed_work, NEIGH_VAR(&tbl->parms, INTERVAL_PROBE_TIME_MS)); - write_unlock_bh(&tbl->lock); + spin_unlock_bh(&tbl->lock); } static void neigh_proxy_process(struct timer_list *t) @@ -1761,9 +1761,9 @@ struct neigh_parms *neigh_parms_alloc(struct net_device *dev, return NULL; } - write_lock_bh(&tbl->lock); + spin_lock_bh(&tbl->lock); list_add_rcu(&p->list, &tbl->parms.list); - write_unlock_bh(&tbl->lock); + spin_unlock_bh(&tbl->lock); neigh_parms_data_state_cleanall(p); } @@ -1783,10 +1783,12 @@ void neigh_parms_release(struct neigh_table *tbl, struct neigh_parms *parms) { if (!parms || parms == &tbl->parms) return; - write_lock_bh(&tbl->lock); + + spin_lock_bh(&tbl->lock); list_del_rcu(&parms->list); parms->dead = 1; - write_unlock_bh(&tbl->lock); + spin_unlock_bh(&tbl->lock); + netdev_put(parms->dev, &parms->dev_tracker); call_rcu(&parms->rcu_head, neigh_rcu_free_parms); } @@ -1835,7 +1837,7 @@ void neigh_table_init(int index, struct neigh_table *tbl) else WARN_ON(tbl->entry_size % NEIGH_PRIV_ALIGN); - rwlock_init(&tbl->lock); + spin_lock_init(&tbl->lock); mutex_init(&tbl->phash_lock); INIT_DEFERRABLE_WORK(&tbl->gc_work, neigh_periodic_work); @@ -1978,10 +1980,10 @@ static int neigh_delete(struct sk_buff *skb, struct nlmsghdr *nlh, err = __neigh_update(neigh, NULL, NUD_FAILED, NEIGH_UPDATE_F_OVERRIDE | NEIGH_UPDATE_F_ADMIN, NETLINK_CB(skb).portid, extack); - write_lock_bh(&tbl->lock); + spin_lock_bh(&tbl->lock); neigh_release(neigh); neigh_remove_one(neigh); - write_unlock_bh(&tbl->lock); + spin_unlock_bh(&tbl->lock); out: return err; @@ -2406,7 +2408,7 @@ static int neightbl_set(struct sk_buff *skb, struct nlmsghdr *nlh, * We acquire tbl->lock to be nice to the periodic timers and * make sure they always see a consistent set of values. */ - write_lock_bh(&tbl->lock); + spin_lock_bh(&tbl->lock); if (tb[NDTA_PARMS]) { struct nlattr *tbp[NDTPA_MAX+1]; @@ -2525,7 +2527,7 @@ static int neightbl_set(struct sk_buff *skb, struct nlmsghdr *nlh, err = 0; errout_tbl_lock: - write_unlock_bh(&tbl->lock); + spin_unlock_bh(&tbl->lock); rcu_read_unlock(); errout: return err; @@ -3125,14 +3127,14 @@ void neigh_for_each(struct neigh_table *tbl, void (*cb)(struct neighbour *, void rcu_read_lock(); nht = rcu_dereference(tbl->nht); - read_lock_bh(&tbl->lock); /* avoid resizes */ + spin_lock_bh(&tbl->lock); /* avoid resizes */ for (chain = 0; chain < (1 << nht->hash_shift); chain++) { struct neighbour *n; neigh_for_each_in_bucket(n, &nht->hash_heads[chain]) cb(n, cookie); } - read_unlock_bh(&tbl->lock); + spin_unlock_bh(&tbl->lock); rcu_read_unlock(); } EXPORT_SYMBOL(neigh_for_each); @@ -3402,7 +3404,7 @@ void *neigh_seq_start(struct seq_file *seq, loff_t *pos, struct neigh_table *tbl rcu_read_lock(); state->nht = rcu_dereference(tbl->nht); - read_lock_bh(&tbl->lock); + spin_lock_bh(&tbl->lock); return *pos ? neigh_get_idx_any(seq, pos) : SEQ_START_TOKEN; } @@ -3442,7 +3444,7 @@ void neigh_seq_stop(struct seq_file *seq, void *v) struct neigh_seq_state *state = seq->private; struct neigh_table *tbl = state->tbl; - read_unlock_bh(&tbl->lock); + spin_unlock_bh(&tbl->lock); rcu_read_unlock(); } EXPORT_SYMBOL(neigh_seq_stop); diff --git a/net/ipv4/arp.c b/net/ipv4/arp.c index 833f2cf97178..f3bfecf8a234 100644 --- a/net/ipv4/arp.c +++ b/net/ipv4/arp.c @@ -1217,10 +1217,10 @@ int arp_invalidate(struct net_device *dev, __be32 ip, bool force) err = neigh_update(neigh, NULL, NUD_FAILED, NEIGH_UPDATE_F_OVERRIDE| NEIGH_UPDATE_F_ADMIN, 0); - write_lock_bh(&tbl->lock); + spin_lock_bh(&tbl->lock); neigh_release(neigh); neigh_remove_one(neigh); - write_unlock_bh(&tbl->lock); + spin_unlock_bh(&tbl->lock); } return err; -- cgit v1.2.3 From a392cde88d19af917740d27e13115447d3b21a06 Mon Sep 17 00:00:00 2001 From: Ryder Lee Date: Tue, 23 Sep 2025 17:23:22 +0000 Subject: wifi: cfg80211/mac80211: validate radio frequency range for monitor mode In multi-radio devices, it is possible to have an MLD AP and a monitor interface active at the same time. In such cases, monitor mode may not be able to specify a fixed channel and could end up capturing frames from all radios, including those outside the intended frequency bands. This patch adds frequency validation for monitor mode. Received frames are now only processed if their frequency fall within the allowed ranges of the radios specified by the interface's radio_mask. This prevents monitor mode from capturing frames outside the supported radio. Signed-off-by: Ryder Lee Link: https://patch.msgid.link/700b8284e845d96654eb98431f8eeb5a81503862.1758647858.git.ryder.lee@mediatek.com Signed-off-by: Johannes Berg --- include/net/cfg80211.h | 14 ++++++++++++++ net/mac80211/rx.c | 49 +++++++++++++++++++++++++++++++++++++++++++++++++ net/wireless/util.c | 6 +++--- 3 files changed, 66 insertions(+), 3 deletions(-) (limited to 'include/net') diff --git a/include/net/cfg80211.h b/include/net/cfg80211.h index 781624f5913a..3b6f48a783bb 100644 --- a/include/net/cfg80211.h +++ b/include/net/cfg80211.h @@ -1015,6 +1015,7 @@ const struct cfg80211_chan_def * cfg80211_chandef_compatible(const struct cfg80211_chan_def *chandef1, const struct cfg80211_chan_def *chandef2); + /** * nl80211_chan_width_to_mhz - get the channel width in MHz * @chan_width: the channel width from &enum nl80211_chan_width @@ -6882,6 +6883,19 @@ static inline bool cfg80211_channel_is_psc(struct ieee80211_channel *chan) return ieee80211_frequency_to_channel(chan->center_freq) % 16 == 5; } +/** + * ieee80211_radio_freq_range_valid - Check if the radio supports the + * specified frequency range + * + * @radio: wiphy radio + * @freq: the frequency (in KHz) to be queried + * @width: the bandwidth (in KHz) to be queried + * + * Return: whether or not the given frequency range is valid for the given radio + */ +bool ieee80211_radio_freq_range_valid(const struct wiphy_radio *radio, + u32 freq, u32 width); + /** * cfg80211_radio_chandef_valid - Check if the radio supports the chandef * diff --git a/net/mac80211/rx.c b/net/mac80211/rx.c index 6af43dfefdd6..29175a0c9f68 100644 --- a/net/mac80211/rx.c +++ b/net/mac80211/rx.c @@ -763,6 +763,51 @@ ieee80211_make_monitor_skb(struct ieee80211_local *local, return skb; } +static bool +ieee80211_validate_monitor_radio(struct ieee80211_sub_if_data *sdata, + struct ieee80211_local *local, + struct ieee80211_rx_status *status) +{ + struct wiphy *wiphy = local->hw.wiphy; + int i, freq, bw; + + if (!wiphy->n_radio) + return true; + + switch (status->bw) { + case RATE_INFO_BW_20: + bw = 20000; + break; + case RATE_INFO_BW_40: + bw = 40000; + break; + case RATE_INFO_BW_80: + bw = 80000; + break; + case RATE_INFO_BW_160: + bw = 160000; + break; + case RATE_INFO_BW_320: + bw = 320000; + break; + default: + return false; + } + + freq = MHZ_TO_KHZ(status->freq); + + for (i = 0; i < wiphy->n_radio; i++) { + if (!(sdata->wdev.radio_mask & BIT(i))) + continue; + + if (!ieee80211_radio_freq_range_valid(&wiphy->radio[i], freq, bw)) + continue; + + return true; + } + return false; +} + /* * This function copies a received frame to all monitor interfaces and * returns a cleaned-up SKB that no longer includes the FCS nor the @@ -855,6 +900,10 @@ ieee80211_rx_monitor(struct ieee80211_local *local, struct sk_buff *origskb, chandef->chan->center_freq != status->freq) continue; + if (ieee80211_hw_check(&local->hw, NO_VIRTUAL_MONITOR) && + !ieee80211_validate_monitor_radio(sdata, local, status)) + continue; + if (!prev_sdata) { prev_sdata = sdata; continue; diff --git a/net/wireless/util.c b/net/wireless/util.c index 56724b33af04..97f40c6d1e9d 100644 --- a/net/wireless/util.c +++ b/net/wireless/util.c @@ -2942,9 +2942,8 @@ cfg80211_get_iftype_ext_capa(struct wiphy *wiphy, enum nl80211_iftype type) } EXPORT_SYMBOL(cfg80211_get_iftype_ext_capa); -static bool -ieee80211_radio_freq_range_valid(const struct wiphy_radio *radio, - u32 freq, u32 width) +bool ieee80211_radio_freq_range_valid(const struct wiphy_radio *radio, + u32 freq, u32 width) { const struct wiphy_radio_freq_range *r; int i; @@ -2958,6 +2957,7 @@ ieee80211_radio_freq_range_valid(const struct wiphy_radio *radio, return false; } +EXPORT_SYMBOL(ieee80211_radio_freq_range_valid); bool cfg80211_radio_chandef_valid(const struct wiphy_radio *radio, const struct cfg80211_chan_def *chandef) -- cgit v1.2.3 From 7cc986c04a9b07d91684f7e326fa5b960215bc97 Mon Sep 17 00:00:00 2001 From: Roopni Devanathan Date: Fri, 24 Oct 2025 10:16:48 +0530 Subject: wifi: cfg80211: Add debugfs support for multi-radio wiphy In multi-radio wiphy architecture, where a single wiphy can have multiple radios tied to it, radio specific configuration parameters and global wiphy parameters are maintained for the entire physical device and common to all radios. But, each radio in a wiphy can have different values for each radio configuration parameter, like RTS threshold. With the current debugfs directory structure, the values of global wiphy configuration parameters can be viewed, but, values of individual radio configuration parameters cannot be viewed, as radio specific configuration parameters are not maintained, separately. To address this, in addition to maintaining global wiphy configuration parameters common to all radios, create separate debugfs directories for each radio in a wiphy to maintain parameters corresponding to that radio in this directory. In implementation, maintain a dentry structure in wiphy_radio_cfg, a structure containing radio configurations of a wiphy. This struct is maintained to denote per-radio configurations of a wiphy. Create separate directories representing each radio within phy#X directory in debugfs during wiphy registration. Sample directory structure with this change: ls /sys/kernel/debug/ieee80211/phy0/radio radio0/ radio1/ radio2/ Signed-off-by: Roopni Devanathan Link: https://patch.msgid.link/20251024044649.483557-2-quic_rdevanat@quicinc.com Signed-off-by: Johannes Berg --- include/net/cfg80211.h | 4 ++++ net/wireless/core.c | 15 +++++++++++++++ 2 files changed, 19 insertions(+) (limited to 'include/net') diff --git a/include/net/cfg80211.h b/include/net/cfg80211.h index 3b6f48a783bb..53490eb04e87 100644 --- a/include/net/cfg80211.h +++ b/include/net/cfg80211.h @@ -5684,9 +5684,13 @@ struct wiphy_iftype_akm_suites { * * @rts_threshold: RTS threshold (dot11RTSThreshold); * -1 (default) = RTS/CTS disabled + * @radio_debugfsdir: Pointer to debugfs directory containing the radio- + * specific parameters. + * NULL (default) = Debugfs directory not created */ struct wiphy_radio_cfg { u32 rts_threshold; + struct dentry *radio_debugfsdir; }; /** diff --git a/net/wireless/core.c b/net/wireless/core.c index 797f9f2004a6..f3568eb5e592 100644 --- a/net/wireless/core.c +++ b/net/wireless/core.c @@ -34,6 +34,9 @@ /* name for sysfs, %d is appended */ #define PHY_NAME "phy" +/* maximum length of radio debugfs directory name */ +#define RADIO_DEBUGFSDIR_MAX_LEN 8 + MODULE_AUTHOR("Johannes Berg"); MODULE_LICENSE("GPL"); MODULE_DESCRIPTION("wireless configuration support"); @@ -1042,6 +1045,18 @@ int wiphy_register(struct wiphy *wiphy) /* add to debugfs */ rdev->wiphy.debugfsdir = debugfs_create_dir(wiphy_name(&rdev->wiphy), ieee80211_debugfs_dir); + if (wiphy->n_radio > 0) { + int idx; + char radio_name[RADIO_DEBUGFSDIR_MAX_LEN]; + + for (idx = 0; idx < wiphy->n_radio; idx++) { + scnprintf(radio_name, sizeof(radio_name), "radio%d", + idx); + wiphy->radio_cfg[idx].radio_debugfsdir = + debugfs_create_dir(radio_name, + rdev->wiphy.debugfsdir); + } + } cfg80211_debugfs_rdev_add(rdev); nl80211_notify_wiphy(rdev, NL80211_CMD_NEW_WIPHY); -- cgit v1.2.3 From 82cb5be6ad64198a3a028aeb49dcc7f6224d558a Mon Sep 17 00:00:00 2001 From: Wilfred Mallawa Date: Wed, 22 Oct 2025 10:19:36 +1000 Subject: net/tls: support setting the maximum payload size During a handshake, an endpoint may specify a maximum record size limit. Currently, the kernel defaults to TLS_MAX_PAYLOAD_SIZE (16KB) for the maximum record size. Meaning that, the outgoing records from the kernel can exceed a lower size negotiated during the handshake. In such a case, the TLS endpoint must send a fatal "record_overflow" alert [1], and thus the record is discarded. Upcoming Western Digital NVMe-TCP hardware controllers implement TLS support. For these devices, supporting TLS record size negotiation is necessary because the maximum TLS record size supported by the controller is less than the default 16KB currently used by the kernel. Currently, there is no way to inform the kernel of such a limit. This patch adds support to a new setsockopt() option `TLS_TX_MAX_PAYLOAD_LEN` that allows for setting the maximum plaintext fragment size. Once set, outgoing records are no larger than the size specified. This option can be used to specify the record size limit. [1] https://www.rfc-editor.org/rfc/rfc8449 Signed-off-by: Wilfred Mallawa Reviewed-by: Sabrina Dubroca Link: https://patch.msgid.link/20251022001937.20155-1-wilfred.opensource@gmail.com Signed-off-by: Jakub Kicinski --- Documentation/networking/tls.rst | 20 +++++++++++++ include/net/tls.h | 3 ++ include/uapi/linux/tls.h | 2 ++ net/tls/tls_device.c | 2 +- net/tls/tls_main.c | 64 ++++++++++++++++++++++++++++++++++++++++ net/tls/tls_sw.c | 2 +- 6 files changed, 91 insertions(+), 2 deletions(-) (limited to 'include/net') diff --git a/Documentation/networking/tls.rst b/Documentation/networking/tls.rst index 36cc7afc2527..980c442d7161 100644 --- a/Documentation/networking/tls.rst +++ b/Documentation/networking/tls.rst @@ -280,6 +280,26 @@ If the record decrypted turns out to had been padded or is not a data record it will be decrypted again into a kernel buffer without zero copy. Such events are counted in the ``TlsDecryptRetry`` statistic. +TLS_TX_MAX_PAYLOAD_LEN +~~~~~~~~~~~~~~~~~~~~~~ + +Specifies the maximum size of the plaintext payload for transmitted TLS records. + +When this option is set, the kernel enforces the specified limit on all outgoing +TLS records. No plaintext fragment will exceed this size. This option can be used +to implement the TLS Record Size Limit extension [1]. + +* For TLS 1.2, the value corresponds directly to the record size limit. +* For TLS 1.3, the value should be set to record_size_limit - 1, since + the record size limit includes one additional byte for the ContentType + field. + +The valid range for this option is 64 to 16384 bytes for TLS 1.2, and 63 to +16384 bytes for TLS 1.3. The lower minimum for TLS 1.3 accounts for the +extra byte used by the ContentType field. + +[1] https://datatracker.ietf.org/doc/html/rfc8449 + Statistics ========== diff --git a/include/net/tls.h b/include/net/tls.h index 857340338b69..f2af113728aa 100644 --- a/include/net/tls.h +++ b/include/net/tls.h @@ -53,6 +53,8 @@ struct tls_rec; /* Maximum data size carried in a TLS record */ #define TLS_MAX_PAYLOAD_SIZE ((size_t)1 << 14) +/* Minimum record size limit as per RFC8449 */ +#define TLS_MIN_RECORD_SIZE_LIM ((size_t)1 << 6) #define TLS_HEADER_SIZE 5 #define TLS_NONCE_OFFSET TLS_HEADER_SIZE @@ -226,6 +228,7 @@ struct tls_context { u8 rx_conf:3; u8 zerocopy_sendfile:1; u8 rx_no_pad:1; + u16 tx_max_payload_len; int (*push_pending_record)(struct sock *sk, int flags); void (*sk_write_space)(struct sock *sk); diff --git a/include/uapi/linux/tls.h b/include/uapi/linux/tls.h index b66a800389cc..b8b9c42f848c 100644 --- a/include/uapi/linux/tls.h +++ b/include/uapi/linux/tls.h @@ -41,6 +41,7 @@ #define TLS_RX 2 /* Set receive parameters */ #define TLS_TX_ZEROCOPY_RO 3 /* TX zerocopy (only sendfile now) */ #define TLS_RX_EXPECT_NO_PAD 4 /* Attempt opportunistic zero-copy */ +#define TLS_TX_MAX_PAYLOAD_LEN 5 /* Maximum plaintext size */ /* Supported versions */ #define TLS_VERSION_MINOR(ver) ((ver) & 0xFF) @@ -194,6 +195,7 @@ enum { TLS_INFO_RXCONF, TLS_INFO_ZC_RO_TX, TLS_INFO_RX_NO_PAD, + TLS_INFO_TX_MAX_PAYLOAD_LEN, __TLS_INFO_MAX, }; #define TLS_INFO_MAX (__TLS_INFO_MAX - 1) diff --git a/net/tls/tls_device.c b/net/tls/tls_device.c index caa2b5d24622..4d29b390aed9 100644 --- a/net/tls/tls_device.c +++ b/net/tls/tls_device.c @@ -462,7 +462,7 @@ static int tls_push_data(struct sock *sk, /* TLS_HEADER_SIZE is not counted as part of the TLS record, and * we need to leave room for an authentication tag. */ - max_open_record_len = TLS_MAX_PAYLOAD_SIZE + + max_open_record_len = tls_ctx->tx_max_payload_len + prot->prepend_size; do { rc = tls_do_allocation(sk, ctx, pfrag, prot->prepend_size); diff --git a/net/tls/tls_main.c b/net/tls/tls_main.c index 39a2ab47fe72..56ce0bc8317b 100644 --- a/net/tls/tls_main.c +++ b/net/tls/tls_main.c @@ -541,6 +541,28 @@ static int do_tls_getsockopt_no_pad(struct sock *sk, char __user *optval, return 0; } +static int do_tls_getsockopt_tx_payload_len(struct sock *sk, char __user *optval, + int __user *optlen) +{ + struct tls_context *ctx = tls_get_ctx(sk); + u16 payload_len = ctx->tx_max_payload_len; + int len; + + if (get_user(len, optlen)) + return -EFAULT; + + if (len < sizeof(payload_len)) + return -EINVAL; + + if (put_user(sizeof(payload_len), optlen)) + return -EFAULT; + + if (copy_to_user(optval, &payload_len, sizeof(payload_len))) + return -EFAULT; + + return 0; +} + static int do_tls_getsockopt(struct sock *sk, int optname, char __user *optval, int __user *optlen) { @@ -560,6 +582,9 @@ static int do_tls_getsockopt(struct sock *sk, int optname, case TLS_RX_EXPECT_NO_PAD: rc = do_tls_getsockopt_no_pad(sk, optval, optlen); break; + case TLS_TX_MAX_PAYLOAD_LEN: + rc = do_tls_getsockopt_tx_payload_len(sk, optval, optlen); + break; default: rc = -ENOPROTOOPT; break; @@ -809,6 +834,32 @@ static int do_tls_setsockopt_no_pad(struct sock *sk, sockptr_t optval, return rc; } +static int do_tls_setsockopt_tx_payload_len(struct sock *sk, sockptr_t optval, + unsigned int optlen) +{ + struct tls_context *ctx = tls_get_ctx(sk); + struct tls_sw_context_tx *sw_ctx = tls_sw_ctx_tx(ctx); + u16 value; + bool tls_13 = ctx->prot_info.version == TLS_1_3_VERSION; + + if (sw_ctx && sw_ctx->open_rec) + return -EBUSY; + + if (sockptr_is_null(optval) || optlen != sizeof(value)) + return -EINVAL; + + if (copy_from_sockptr(&value, optval, sizeof(value))) + return -EFAULT; + + if (value < TLS_MIN_RECORD_SIZE_LIM - (tls_13 ? 1 : 0) || + value > TLS_MAX_PAYLOAD_SIZE) + return -EINVAL; + + ctx->tx_max_payload_len = value; + + return 0; +} + static int do_tls_setsockopt(struct sock *sk, int optname, sockptr_t optval, unsigned int optlen) { @@ -830,6 +881,11 @@ static int do_tls_setsockopt(struct sock *sk, int optname, sockptr_t optval, case TLS_RX_EXPECT_NO_PAD: rc = do_tls_setsockopt_no_pad(sk, optval, optlen); break; + case TLS_TX_MAX_PAYLOAD_LEN: + lock_sock(sk); + rc = do_tls_setsockopt_tx_payload_len(sk, optval, optlen); + release_sock(sk); + break; default: rc = -ENOPROTOOPT; break; @@ -1019,6 +1075,7 @@ static int tls_init(struct sock *sk) ctx->tx_conf = TLS_BASE; ctx->rx_conf = TLS_BASE; + ctx->tx_max_payload_len = TLS_MAX_PAYLOAD_SIZE; update_sk_prot(sk, ctx); out: write_unlock_bh(&sk->sk_callback_lock); @@ -1108,6 +1165,12 @@ static int tls_get_info(struct sock *sk, struct sk_buff *skb, bool net_admin) goto nla_failure; } + err = nla_put_u16(skb, TLS_INFO_TX_MAX_PAYLOAD_LEN, + ctx->tx_max_payload_len); + + if (err) + goto nla_failure; + rcu_read_unlock(); nla_nest_end(skb, start); return 0; @@ -1129,6 +1192,7 @@ static size_t tls_get_info_size(const struct sock *sk, bool net_admin) nla_total_size(sizeof(u16)) + /* TLS_INFO_TXCONF */ nla_total_size(0) + /* TLS_INFO_ZC_RO_TX */ nla_total_size(0) + /* TLS_INFO_RX_NO_PAD */ + nla_total_size(sizeof(u16)) + /* TLS_INFO_TX_MAX_PAYLOAD_LEN */ 0; return size; diff --git a/net/tls/tls_sw.c b/net/tls/tls_sw.c index d17135369980..9937d4c810f2 100644 --- a/net/tls/tls_sw.c +++ b/net/tls/tls_sw.c @@ -1079,7 +1079,7 @@ static int tls_sw_sendmsg_locked(struct sock *sk, struct msghdr *msg, orig_size = msg_pl->sg.size; full_record = false; try_to_copy = msg_data_left(msg); - record_room = TLS_MAX_PAYLOAD_SIZE - msg_pl->sg.size; + record_room = tls_ctx->tx_max_payload_len - msg_pl->sg.size; if (try_to_copy >= record_room) { try_to_copy = record_room; full_record = true; -- cgit v1.2.3 From 151b98d10ef7c3174465e409b99d8762e7e8de60 Mon Sep 17 00:00:00 2001 From: Kuniyuki Iwashima Date: Thu, 23 Oct 2025 23:16:53 +0000 Subject: net: Add sk_clone(). sctp_accept() will use sk_clone_lock(), but it will be called with the parent socket locked, and sctp_migrate() acquires the child lock later. Let's add no lock version of sk_clone_lock(). Note that lockdep complains if we simply use bh_lock_sock_nested(). Signed-off-by: Kuniyuki Iwashima Reviewed-by: Xin Long Link: https://patch.msgid.link/20251023231751.4168390-5-kuniyu@google.com Signed-off-by: Jakub Kicinski --- include/net/sock.h | 7 ++++++- net/core/sock.c | 24 ++++++++++++++++-------- 2 files changed, 22 insertions(+), 9 deletions(-) (limited to 'include/net') diff --git a/include/net/sock.h b/include/net/sock.h index 01ce231603db..c7e58b8e8a90 100644 --- a/include/net/sock.h +++ b/include/net/sock.h @@ -1822,7 +1822,12 @@ struct sock *sk_alloc(struct net *net, int family, gfp_t priority, void sk_free(struct sock *sk); void sk_net_refcnt_upgrade(struct sock *sk); void sk_destruct(struct sock *sk); -struct sock *sk_clone_lock(const struct sock *sk, const gfp_t priority); +struct sock *sk_clone(const struct sock *sk, const gfp_t priority, bool lock); + +static inline struct sock *sk_clone_lock(const struct sock *sk, const gfp_t priority) +{ + return sk_clone(sk, priority, true); +} struct sk_buff *sock_wmalloc(struct sock *sk, unsigned long size, int force, gfp_t priority); diff --git a/net/core/sock.c b/net/core/sock.c index a99132cc0965..7a9bbc2afcf0 100644 --- a/net/core/sock.c +++ b/net/core/sock.c @@ -2462,13 +2462,16 @@ static void sk_init_common(struct sock *sk) } /** - * sk_clone_lock - clone a socket, and lock its clone - * @sk: the socket to clone - * @priority: for allocation (%GFP_KERNEL, %GFP_ATOMIC, etc) + * sk_clone - clone a socket + * @sk: the socket to clone + * @priority: for allocation (%GFP_KERNEL, %GFP_ATOMIC, etc) + * @lock: if true, lock the cloned sk * - * Caller must unlock socket even in error path (bh_unlock_sock(newsk)) + * If @lock is true, the clone is locked by bh_lock_sock(), and + * caller must unlock socket even in error path by bh_unlock_sock(). */ -struct sock *sk_clone_lock(const struct sock *sk, const gfp_t priority) +struct sock *sk_clone(const struct sock *sk, const gfp_t priority, + bool lock) { struct proto *prot = READ_ONCE(sk->sk_prot); struct sk_filter *filter; @@ -2497,9 +2500,13 @@ struct sock *sk_clone_lock(const struct sock *sk, const gfp_t priority) __netns_tracker_alloc(sock_net(newsk), &newsk->ns_tracker, false, priority); } + sk_node_init(&newsk->sk_node); sock_lock_init(newsk); - bh_lock_sock(newsk); + + if (lock) + bh_lock_sock(newsk); + newsk->sk_backlog.head = newsk->sk_backlog.tail = NULL; newsk->sk_backlog.len = 0; @@ -2590,12 +2597,13 @@ free: * destructor and make plain sk_free() */ newsk->sk_destruct = NULL; - bh_unlock_sock(newsk); + if (lock) + bh_unlock_sock(newsk); sk_free(newsk); newsk = NULL; goto out; } -EXPORT_SYMBOL_GPL(sk_clone_lock); +EXPORT_SYMBOL_GPL(sk_clone); static u32 sk_dst_gso_max_size(struct sock *sk, const struct net_device *dev) { -- cgit v1.2.3 From c49ed521f1772ca9203d22a1e5950f337fd5f930 Mon Sep 17 00:00:00 2001 From: Kuniyuki Iwashima Date: Thu, 23 Oct 2025 23:16:55 +0000 Subject: sctp: Remove sctp_pf.create_accept_sk(). sctp_v[46]_create_accept_sk() are no longer used. Let's remove sctp_pf.create_accept_sk(). Signed-off-by: Kuniyuki Iwashima Acked-by: Xin Long Link: https://patch.msgid.link/20251023231751.4168390-7-kuniyu@google.com Signed-off-by: Jakub Kicinski --- include/net/sctp/structs.h | 3 --- net/sctp/ipv6.c | 45 --------------------------------------------- net/sctp/protocol.c | 27 --------------------------- 3 files changed, 75 deletions(-) (limited to 'include/net') diff --git a/include/net/sctp/structs.h b/include/net/sctp/structs.h index 2ae390219efd..3dd304e411d0 100644 --- a/include/net/sctp/structs.h +++ b/include/net/sctp/structs.h @@ -497,9 +497,6 @@ struct sctp_pf { int (*bind_verify) (struct sctp_sock *, union sctp_addr *); int (*send_verify) (struct sctp_sock *, union sctp_addr *); int (*supported_addrs)(const struct sctp_sock *, __be16 *); - struct sock *(*create_accept_sk) (struct sock *sk, - struct sctp_association *asoc, - bool kern); int (*addr_to_user)(struct sctp_sock *sk, union sctp_addr *addr); void (*to_sk_saddr)(union sctp_addr *, struct sock *sk); void (*to_sk_daddr)(union sctp_addr *, struct sock *sk); diff --git a/net/sctp/ipv6.c b/net/sctp/ipv6.c index c0762424a854..069b7e45d8bd 100644 --- a/net/sctp/ipv6.c +++ b/net/sctp/ipv6.c @@ -777,50 +777,6 @@ static enum sctp_scope sctp_v6_scope(union sctp_addr *addr) return retval; } -/* Create and initialize a new sk for the socket to be returned by accept(). */ -static struct sock *sctp_v6_create_accept_sk(struct sock *sk, - struct sctp_association *asoc, - bool kern) -{ - struct ipv6_pinfo *newnp, *np = inet6_sk(sk); - struct sctp6_sock *newsctp6sk; - struct inet_sock *newinet; - struct sock *newsk; - - newsk = sk_alloc(sock_net(sk), PF_INET6, GFP_KERNEL, sk->sk_prot, kern); - if (!newsk) - return NULL; - - sock_init_data(NULL, newsk); - - sctp_copy_sock(newsk, sk, asoc); - sock_reset_flag(sk, SOCK_ZAPPED); - - newsctp6sk = (struct sctp6_sock *)newsk; - newinet = inet_sk(newsk); - newinet->pinet6 = &newsctp6sk->inet6; - newinet->ipv6_fl_list = NULL; - - sctp_sk(newsk)->v4mapped = sctp_sk(sk)->v4mapped; - - newnp = inet6_sk(newsk); - - memcpy(newnp, np, sizeof(struct ipv6_pinfo)); - newnp->ipv6_mc_list = NULL; - newnp->ipv6_ac_list = NULL; - - sctp_v6_copy_ip_options(sk, newsk); - - /* Initialize sk's sport, dport, rcv_saddr and daddr for getsockname() - * and getpeername(). - */ - sctp_v6_to_sk_daddr(&asoc->peer.primary_addr, newsk); - - newsk->sk_v6_rcv_saddr = sk->sk_v6_rcv_saddr; - - return newsk; -} - /* Format a sockaddr for return to user space. This makes sure the return is * AF_INET or AF_INET6 depending on the SCTP_I_WANT_MAPPED_V4_ADDR option. */ @@ -1167,7 +1123,6 @@ static struct sctp_pf sctp_pf_inet6 = { .bind_verify = sctp_inet6_bind_verify, .send_verify = sctp_inet6_send_verify, .supported_addrs = sctp_inet6_supported_addrs, - .create_accept_sk = sctp_v6_create_accept_sk, .addr_to_user = sctp_v6_addr_to_user, .to_sk_saddr = sctp_v6_to_sk_saddr, .to_sk_daddr = sctp_v6_to_sk_daddr, diff --git a/net/sctp/protocol.c b/net/sctp/protocol.c index ad2722d1ec15..2c3398f75d76 100644 --- a/net/sctp/protocol.c +++ b/net/sctp/protocol.c @@ -580,32 +580,6 @@ static int sctp_v4_is_ce(const struct sk_buff *skb) return INET_ECN_is_ce(ip_hdr(skb)->tos); } -/* Create and initialize a new sk for the socket returned by accept(). */ -static struct sock *sctp_v4_create_accept_sk(struct sock *sk, - struct sctp_association *asoc, - bool kern) -{ - struct sock *newsk = sk_alloc(sock_net(sk), PF_INET, GFP_KERNEL, - sk->sk_prot, kern); - struct inet_sock *newinet; - - if (!newsk) - return NULL; - - sock_init_data(NULL, newsk); - - sctp_copy_sock(newsk, sk, asoc); - sock_reset_flag(newsk, SOCK_ZAPPED); - - sctp_v4_copy_ip_options(sk, newsk); - - newinet = inet_sk(newsk); - - newinet->inet_daddr = asoc->peer.primary_addr.v4.sin_addr.s_addr; - - return newsk; -} - static int sctp_v4_addr_to_user(struct sctp_sock *sp, union sctp_addr *addr) { /* No address mapping for V4 sockets */ @@ -1113,7 +1087,6 @@ static struct sctp_pf sctp_pf_inet = { .bind_verify = sctp_inet_bind_verify, .send_verify = sctp_inet_send_verify, .supported_addrs = sctp_inet_supported_addrs, - .create_accept_sk = sctp_v4_create_accept_sk, .addr_to_user = sctp_v4_addr_to_user, .to_sk_saddr = sctp_v4_to_sk_saddr, .to_sk_daddr = sctp_v4_to_sk_daddr, -- cgit v1.2.3 From 71068e2e1b6bd78f5599e5bc89e125a75149884b Mon Sep 17 00:00:00 2001 From: Kuniyuki Iwashima Date: Thu, 23 Oct 2025 23:16:57 +0000 Subject: sctp: Remove sctp_copy_sock() and sctp_copy_descendant(). Now, sctp_accept() and sctp_do_peeloff() use sk_clone(), and we no longer need sctp_copy_sock() and sctp_copy_descendant(). Let's remove them. Signed-off-by: Kuniyuki Iwashima Acked-by: Xin Long Link: https://patch.msgid.link/20251023231751.4168390-9-kuniyu@google.com Signed-off-by: Jakub Kicinski --- include/net/inet_sock.h | 8 ------ include/net/sctp/sctp.h | 3 +-- net/sctp/socket.c | 71 ------------------------------------------------- 3 files changed, 1 insertion(+), 81 deletions(-) (limited to 'include/net') diff --git a/include/net/inet_sock.h b/include/net/inet_sock.h index b6ec08072533..ac1c75975908 100644 --- a/include/net/inet_sock.h +++ b/include/net/inet_sock.h @@ -355,14 +355,6 @@ static inline struct sock *skb_to_full_sk(const struct sk_buff *skb) #define inet_sk(ptr) container_of_const(ptr, struct inet_sock, sk) -static inline void __inet_sk_copy_descendant(struct sock *sk_to, - const struct sock *sk_from, - const int ancestor_size) -{ - memcpy(inet_sk(sk_to) + 1, inet_sk(sk_from) + 1, - sk_from->sk_prot->obj_size - ancestor_size); -} - int inet_sk_rebuild_header(struct sock *sk); /** diff --git a/include/net/sctp/sctp.h b/include/net/sctp/sctp.h index e96d1bd087f6..bb4b80c12541 100644 --- a/include/net/sctp/sctp.h +++ b/include/net/sctp/sctp.h @@ -94,8 +94,7 @@ void sctp_data_ready(struct sock *sk); __poll_t sctp_poll(struct file *file, struct socket *sock, poll_table *wait); void sctp_sock_rfree(struct sk_buff *skb); -void sctp_copy_sock(struct sock *newsk, struct sock *sk, - struct sctp_association *asoc); + extern struct percpu_counter sctp_sockets_allocated; int sctp_asconf_mgmt(struct sctp_sock *, struct sctp_sockaddr_entry *); struct sk_buff *sctp_skb_recv_datagram(struct sock *, int, int *); diff --git a/net/sctp/socket.c b/net/sctp/socket.c index 60d3e340dfed..ac737e60829b 100644 --- a/net/sctp/socket.c +++ b/net/sctp/socket.c @@ -9491,72 +9491,6 @@ done: sctp_skb_set_owner_r(skb, sk); } -void sctp_copy_sock(struct sock *newsk, struct sock *sk, - struct sctp_association *asoc) -{ - struct inet_sock *inet = inet_sk(sk); - struct inet_sock *newinet; - struct sctp_sock *sp = sctp_sk(sk); - - newsk->sk_type = sk->sk_type; - newsk->sk_bound_dev_if = sk->sk_bound_dev_if; - newsk->sk_flags = sk->sk_flags; - newsk->sk_tsflags = sk->sk_tsflags; - newsk->sk_no_check_tx = sk->sk_no_check_tx; - newsk->sk_no_check_rx = sk->sk_no_check_rx; - newsk->sk_reuse = sk->sk_reuse; - sctp_sk(newsk)->reuse = sp->reuse; - - newsk->sk_shutdown = sk->sk_shutdown; - newsk->sk_destruct = sk->sk_destruct; - newsk->sk_family = sk->sk_family; - newsk->sk_protocol = IPPROTO_SCTP; - newsk->sk_backlog_rcv = sk->sk_prot->backlog_rcv; - newsk->sk_sndbuf = sk->sk_sndbuf; - newsk->sk_rcvbuf = sk->sk_rcvbuf; - newsk->sk_lingertime = sk->sk_lingertime; - newsk->sk_rcvtimeo = READ_ONCE(sk->sk_rcvtimeo); - newsk->sk_sndtimeo = READ_ONCE(sk->sk_sndtimeo); - newsk->sk_rxhash = sk->sk_rxhash; - newsk->sk_gso_type = sk->sk_gso_type; - - newinet = inet_sk(newsk); - - /* Initialize sk's sport, dport, rcv_saddr and daddr for - * getsockname() and getpeername() - */ - newinet->inet_sport = inet->inet_sport; - newinet->inet_saddr = inet->inet_saddr; - newinet->inet_rcv_saddr = inet->inet_rcv_saddr; - newinet->inet_dport = htons(asoc->peer.port); - newinet->pmtudisc = inet->pmtudisc; - atomic_set(&newinet->inet_id, get_random_u16()); - - newinet->uc_ttl = inet->uc_ttl; - inet_set_bit(MC_LOOP, newsk); - newinet->mc_ttl = 1; - newinet->mc_index = 0; - newinet->mc_list = NULL; - - if (newsk->sk_flags & SK_FLAGS_TIMESTAMP) - net_enable_timestamp(); - - /* Set newsk security attributes from original sk and connection - * security attribute from asoc. - */ - security_sctp_sk_clone(asoc, sk, newsk); -} - -static inline void sctp_copy_descendant(struct sock *sk_to, - const struct sock *sk_from) -{ - size_t ancestor_size = sizeof(struct inet_sock); - - ancestor_size += sk_from->sk_prot->obj_size; - ancestor_size -= offsetof(struct sctp_sock, pd_lobby); - __inet_sk_copy_descendant(sk_to, sk_from, ancestor_size); -} - /* Populate the fields of the newsk from the oldsk and migrate the assoc * and its messages to the newsk. */ @@ -9573,11 +9507,6 @@ static int sctp_sock_migrate(struct sock *oldsk, struct sock *newsk, struct sctp_bind_hashbucket *head; int err; - /* Migrate all the socket level options to the new socket. - * Brute force copy old sctp opt. - */ - sctp_copy_descendant(newsk, oldsk); - /* Restore the ep value that was overwritten with the above structure * copy. */ -- cgit v1.2.3 From 6f147c8328e045de3a35155ca7c883d88da9e916 Mon Sep 17 00:00:00 2001 From: Yue Haibing Date: Fri, 24 Oct 2025 10:51:45 +0800 Subject: net/sched: Remove unused typedef psched_tdiff_t Since commit 051d44209842 ("net/sched: Retire CBQ qdisc") this is not used anymore. Signed-off-by: Yue Haibing Link: https://patch.msgid.link/20251024025145.4069583-1-yuehaibing@huawei.com Signed-off-by: Jakub Kicinski --- include/net/pkt_sched.h | 1 - 1 file changed, 1 deletion(-) (limited to 'include/net') diff --git a/include/net/pkt_sched.h b/include/net/pkt_sched.h index c660ac871083..4678db45832a 100644 --- a/include/net/pkt_sched.h +++ b/include/net/pkt_sched.h @@ -43,7 +43,6 @@ struct qdisc_walker { */ typedef u64 psched_time_t; -typedef long psched_tdiff_t; /* Avoid doing 64 bit divide */ #define PSCHED_SHIFT 6 -- cgit v1.2.3 From 8443c3160858b860bfc2db6a8397c72c9f6b513e Mon Sep 17 00:00:00 2001 From: Bobby Eshleman Date: Fri, 24 Oct 2025 11:02:56 -0700 Subject: net: netmem: remove NET_IOV_MAX from net_iov_type enum Remove the NET_IOV_MAX workaround from the net_iov_type enum. This entry was previously added to force the enum size to unsigned long to satisfy the NET_IOV_ASSERT_OFFSET static assertions. After commit f3d85c9ee510 ("netmem: introduce struct netmem_desc mirroring struct page") this approach became unnecessary by placing the net_iov_type after the netmem_desc. Placing the net_iov_type after netmem_desc results in the net_iov_type size having no effect on the position or layout of the fields that mirror the struct page. The layout before this patch: struct net_iov { union { struct netmem_desc desc; /* 0 48 */ struct { long unsigned int _flags; /* 0 8 */ long unsigned int pp_magic; /* 8 8 */ struct page_pool * pp; /* 16 8 */ long unsigned int _pp_mapping_pad; /* 24 8 */ long unsigned int dma_addr; /* 32 8 */ atomic_long_t pp_ref_count; /* 40 8 */ }; /* 0 48 */ }; /* 0 48 */ struct net_iov_area * owner; /* 48 8 */ enum net_iov_type type; /* 56 8 */ /* size: 64, cachelines: 1, members: 3 */ }; The layout after this patch: struct net_iov { union { struct netmem_desc desc; /* 0 48 */ struct { long unsigned int _flags; /* 0 8 */ long unsigned int pp_magic; /* 8 8 */ struct page_pool * pp; /* 16 8 */ long unsigned int _pp_mapping_pad; /* 24 8 */ long unsigned int dma_addr; /* 32 8 */ atomic_long_t pp_ref_count; /* 40 8 */ }; /* 0 48 */ }; /* 0 48 */ struct net_iov_area * owner; /* 48 8 */ enum net_iov_type type; /* 56 4 */ /* size: 64, cachelines: 1, members: 3 */ /* padding: 4 */ }; Signed-off-by: Bobby Eshleman Reviewed-by: Mina Almasry Link: https://patch.msgid.link/20251024-b4-devmem-remove-niov-max-v1-1-ba72c68bc869@meta.com Signed-off-by: Jakub Kicinski --- include/net/netmem.h | 4 ---- 1 file changed, 4 deletions(-) (limited to 'include/net') diff --git a/include/net/netmem.h b/include/net/netmem.h index 651e2c62d1dd..9e10f4ac50c3 100644 --- a/include/net/netmem.h +++ b/include/net/netmem.h @@ -68,10 +68,6 @@ DECLARE_STATIC_KEY_FALSE(page_pool_mem_providers); enum net_iov_type { NET_IOV_DMABUF, NET_IOV_IOURING, - - /* Force size to unsigned long to make the NET_IOV_ASSERTS below pass. - */ - NET_IOV_MAX = ULONG_MAX }; /* A memory descriptor representing abstract networking I/O vectors, -- cgit v1.2.3 From 294bfe0343da3b59db040c3a4dac05b4c91ce013 Mon Sep 17 00:00:00 2001 From: Christophe JAILLET Date: Sat, 25 Oct 2025 09:40:59 +0200 Subject: sctp: Constify struct sctp_sched_ops 'struct sctp_sched_ops' is not modified in these drivers. Constifying this structure moves some data to a read-only section, so increases overall security, especially when the structure holds some function pointers. On a x86_64, with allmodconfig, as an example: Before: ====== text data bss dec hex filename 8019 568 0 8587 218b net/sctp/stream_sched_fc.o After: ===== text data bss dec hex filename 8275 312 0 8587 218b net/sctp/stream_sched_fc.o Signed-off-by: Christophe JAILLET Link: https://patch.msgid.link/dce03527eb7b7cc8a3c26d5cdac12bafe3350135.1761377890.git.christophe.jaillet@wanadoo.fr Signed-off-by: Jakub Kicinski --- include/net/sctp/stream_sched.h | 4 ++-- include/net/sctp/structs.h | 2 +- net/sctp/stream.c | 8 ++++---- net/sctp/stream_sched.c | 16 ++++++++-------- net/sctp/stream_sched_fc.c | 4 ++-- net/sctp/stream_sched_prio.c | 2 +- net/sctp/stream_sched_rr.c | 2 +- 7 files changed, 19 insertions(+), 19 deletions(-) (limited to 'include/net') diff --git a/include/net/sctp/stream_sched.h b/include/net/sctp/stream_sched.h index 8034bf5febbe..77806ef1cb70 100644 --- a/include/net/sctp/stream_sched.h +++ b/include/net/sctp/stream_sched.h @@ -52,10 +52,10 @@ void sctp_sched_dequeue_done(struct sctp_outq *q, struct sctp_chunk *ch); void sctp_sched_dequeue_common(struct sctp_outq *q, struct sctp_chunk *ch); int sctp_sched_init_sid(struct sctp_stream *stream, __u16 sid, gfp_t gfp); -struct sctp_sched_ops *sctp_sched_ops_from_stream(struct sctp_stream *stream); +const struct sctp_sched_ops *sctp_sched_ops_from_stream(struct sctp_stream *stream); void sctp_sched_ops_register(enum sctp_sched_type sched, - struct sctp_sched_ops *sched_ops); + const struct sctp_sched_ops *sched_ops); void sctp_sched_ops_prio_init(void); void sctp_sched_ops_rr_init(void); void sctp_sched_ops_fc_init(void); diff --git a/include/net/sctp/structs.h b/include/net/sctp/structs.h index 3dd304e411d0..5900196d65fd 100644 --- a/include/net/sctp/structs.h +++ b/include/net/sctp/structs.h @@ -1073,7 +1073,7 @@ struct sctp_outq { struct list_head out_chunk_list; /* Stream scheduler being used */ - struct sctp_sched_ops *sched; + const struct sctp_sched_ops *sched; unsigned int out_qlen; /* Total length of queued data chunks. */ diff --git a/net/sctp/stream.c b/net/sctp/stream.c index f205556c5b24..0615e4426341 100644 --- a/net/sctp/stream.c +++ b/net/sctp/stream.c @@ -54,7 +54,7 @@ static void sctp_stream_shrink_out(struct sctp_stream *stream, __u16 outcnt) static void sctp_stream_free_ext(struct sctp_stream *stream, __u16 sid) { - struct sctp_sched_ops *sched; + const struct sctp_sched_ops *sched; if (!SCTP_SO(stream, sid)->ext) return; @@ -130,7 +130,7 @@ out: int sctp_stream_init(struct sctp_stream *stream, __u16 outcnt, __u16 incnt, gfp_t gfp) { - struct sctp_sched_ops *sched = sctp_sched_ops_from_stream(stream); + const struct sctp_sched_ops *sched = sctp_sched_ops_from_stream(stream); int i, ret = 0; gfp |= __GFP_NOWARN; @@ -182,7 +182,7 @@ int sctp_stream_init_ext(struct sctp_stream *stream, __u16 sid) void sctp_stream_free(struct sctp_stream *stream) { - struct sctp_sched_ops *sched = sctp_sched_ops_from_stream(stream); + const struct sctp_sched_ops *sched = sctp_sched_ops_from_stream(stream); int i; sched->unsched_all(stream); @@ -207,7 +207,7 @@ void sctp_stream_clear(struct sctp_stream *stream) void sctp_stream_update(struct sctp_stream *stream, struct sctp_stream *new) { - struct sctp_sched_ops *sched = sctp_sched_ops_from_stream(stream); + const struct sctp_sched_ops *sched = sctp_sched_ops_from_stream(stream); sched->unsched_all(stream); sctp_stream_outq_migrate(stream, new, new->outcnt); diff --git a/net/sctp/stream_sched.c b/net/sctp/stream_sched.c index 54afbe4fb087..50f8b5240359 100644 --- a/net/sctp/stream_sched.c +++ b/net/sctp/stream_sched.c @@ -91,7 +91,7 @@ static void sctp_sched_fcfs_unsched_all(struct sctp_stream *stream) { } -static struct sctp_sched_ops sctp_sched_fcfs = { +static const struct sctp_sched_ops sctp_sched_fcfs = { .set = sctp_sched_fcfs_set, .get = sctp_sched_fcfs_get, .init = sctp_sched_fcfs_init, @@ -111,10 +111,10 @@ static void sctp_sched_ops_fcfs_init(void) /* API to other parts of the stack */ -static struct sctp_sched_ops *sctp_sched_ops[SCTP_SS_MAX + 1]; +static const struct sctp_sched_ops *sctp_sched_ops[SCTP_SS_MAX + 1]; void sctp_sched_ops_register(enum sctp_sched_type sched, - struct sctp_sched_ops *sched_ops) + const struct sctp_sched_ops *sched_ops) { sctp_sched_ops[sched] = sched_ops; } @@ -130,7 +130,7 @@ void sctp_sched_ops_init(void) static void sctp_sched_free_sched(struct sctp_stream *stream) { - struct sctp_sched_ops *sched = sctp_sched_ops_from_stream(stream); + const struct sctp_sched_ops *sched = sctp_sched_ops_from_stream(stream); struct sctp_stream_out_ext *soute; int i; @@ -148,9 +148,9 @@ static void sctp_sched_free_sched(struct sctp_stream *stream) int sctp_sched_set_sched(struct sctp_association *asoc, enum sctp_sched_type sched) { - struct sctp_sched_ops *old = asoc->outqueue.sched; + const struct sctp_sched_ops *old = asoc->outqueue.sched; struct sctp_datamsg *msg = NULL; - struct sctp_sched_ops *n; + const struct sctp_sched_ops *n; struct sctp_chunk *ch; int i, ret = 0; @@ -263,14 +263,14 @@ void sctp_sched_dequeue_common(struct sctp_outq *q, struct sctp_chunk *ch) int sctp_sched_init_sid(struct sctp_stream *stream, __u16 sid, gfp_t gfp) { - struct sctp_sched_ops *sched = sctp_sched_ops_from_stream(stream); + const struct sctp_sched_ops *sched = sctp_sched_ops_from_stream(stream); struct sctp_stream_out_ext *ext = SCTP_SO(stream, sid)->ext; INIT_LIST_HEAD(&ext->outq); return sched->init_sid(stream, sid, gfp); } -struct sctp_sched_ops *sctp_sched_ops_from_stream(struct sctp_stream *stream) +const struct sctp_sched_ops *sctp_sched_ops_from_stream(struct sctp_stream *stream) { struct sctp_association *asoc; diff --git a/net/sctp/stream_sched_fc.c b/net/sctp/stream_sched_fc.c index 4bd18a497a6d..776c6de46c22 100644 --- a/net/sctp/stream_sched_fc.c +++ b/net/sctp/stream_sched_fc.c @@ -188,7 +188,7 @@ static void sctp_sched_fc_unsched_all(struct sctp_stream *stream) list_del_init(&soute->fc_list); } -static struct sctp_sched_ops sctp_sched_fc = { +static const struct sctp_sched_ops sctp_sched_fc = { .set = sctp_sched_fc_set, .get = sctp_sched_fc_get, .init = sctp_sched_fc_init, @@ -206,7 +206,7 @@ void sctp_sched_ops_fc_init(void) sctp_sched_ops_register(SCTP_SS_FC, &sctp_sched_fc); } -static struct sctp_sched_ops sctp_sched_wfq = { +static const struct sctp_sched_ops sctp_sched_wfq = { .set = sctp_sched_wfq_set, .get = sctp_sched_wfq_get, .init = sctp_sched_fc_init, diff --git a/net/sctp/stream_sched_prio.c b/net/sctp/stream_sched_prio.c index 4d4d9da331f4..fb6c55e5615d 100644 --- a/net/sctp/stream_sched_prio.c +++ b/net/sctp/stream_sched_prio.c @@ -300,7 +300,7 @@ static void sctp_sched_prio_unsched_all(struct sctp_stream *stream) sctp_sched_prio_unsched(soute); } -static struct sctp_sched_ops sctp_sched_prio = { +static const struct sctp_sched_ops sctp_sched_prio = { .set = sctp_sched_prio_set, .get = sctp_sched_prio_get, .init = sctp_sched_prio_init, diff --git a/net/sctp/stream_sched_rr.c b/net/sctp/stream_sched_rr.c index 1f235e7f643a..9157b653f196 100644 --- a/net/sctp/stream_sched_rr.c +++ b/net/sctp/stream_sched_rr.c @@ -171,7 +171,7 @@ static void sctp_sched_rr_unsched_all(struct sctp_stream *stream) sctp_sched_rr_unsched(stream, soute); } -static struct sctp_sched_ops sctp_sched_rr = { +static const struct sctp_sched_ops sctp_sched_rr = { .set = sctp_sched_rr_set, .get = sctp_sched_rr_get, .init = sctp_sched_rr_init, -- cgit v1.2.3 From f0e7036fc9cb08bdfb27d64eee7fc003ba0bc2e5 Mon Sep 17 00:00:00 2001 From: Ido Schimmel Date: Mon, 27 Oct 2025 10:22:30 +0200 Subject: ipv4: icmp: Add RFC 5837 support Add the ability to append the incoming IP interface information to ICMPv4 error messages in accordance with RFC 5837 and RFC 4884. This is required for more meaningful traceroute results in unnumbered networks. The feature is disabled by default and controlled via a new sysctl ("net.ipv4.icmp_errors_extension_mask") which accepts a bitmask of ICMP extensions to append to ICMP error messages. Currently, only a single value is supported, but the interface and the implementation should be able to support more extensions, if needed. Clone the skb and copy the relevant data portions before modifying the skb as the caller of __icmp_send() still owns the skb after the function returns. This should be fine since by default ICMP error messages are rate limited to 1000 per second and no more than 1 per second per specific host. Trim or pad the packet to 128 bytes before appending the ICMP extension structure in order to be compatible with legacy applications that assume that the ICMP extension structure always starts at this offset (the minimum length specified by RFC 4884). Reviewed-by: Petr Machata Reviewed-by: David Ahern Reviewed-by: Willem de Bruijn Signed-off-by: Ido Schimmel Reviewed-by: Eric Dumazet Link: https://patch.msgid.link/20251027082232.232571-2-idosch@nvidia.com Signed-off-by: Jakub Kicinski --- Documentation/networking/ip-sysctl.rst | 17 +++ include/linux/icmp.h | 32 ++++++ include/net/netns/ipv4.h | 1 + net/ipv4/icmp.c | 191 ++++++++++++++++++++++++++++++++- net/ipv4/sysctl_net_ipv4.c | 11 ++ 5 files changed, 251 insertions(+), 1 deletion(-) (limited to 'include/net') diff --git a/Documentation/networking/ip-sysctl.rst b/Documentation/networking/ip-sysctl.rst index a06cb99d66dc..ece1187ba0f1 100644 --- a/Documentation/networking/ip-sysctl.rst +++ b/Documentation/networking/ip-sysctl.rst @@ -1796,6 +1796,23 @@ icmp_errors_use_inbound_ifaddr - BOOLEAN Default: 0 (disabled) +icmp_errors_extension_mask - UNSIGNED INTEGER + Bitmask of ICMP extensions to append to ICMPv4 error messages + ("Destination Unreachable", "Time Exceeded" and "Parameter Problem"). + The original datagram is trimmed / padded to 128 bytes in order to be + compatible with applications that do not comply with RFC 4884. + + Possible extensions are: + + ==== ============================================================== + 0x01 Incoming IP interface information according to RFC 5837. + Extension will include the index, IPv4 address (if present), + name and MTU of the IP interface that received the datagram + which elicited the ICMP error. + ==== ============================================================== + + Default: 0x00 (no extensions) + igmp_max_memberships - INTEGER Change the maximum number of multicast groups we can subscribe to. Default: 20 diff --git a/include/linux/icmp.h b/include/linux/icmp.h index 0af4d210ee31..043ec5d9c882 100644 --- a/include/linux/icmp.h +++ b/include/linux/icmp.h @@ -40,4 +40,36 @@ void ip_icmp_error_rfc4884(const struct sk_buff *skb, struct sock_ee_data_rfc4884 *out, int thlen, int off); +/* RFC 4884 */ +#define ICMP_EXT_ORIG_DGRAM_MIN_LEN 128 +#define ICMP_EXT_VERSION_2 2 + +/* ICMP Extension Object Classes */ +#define ICMP_EXT_OBJ_CLASS_IIO 2 /* RFC 5837 */ + +/* Interface Information Object - RFC 5837 */ +enum { + ICMP_EXT_CTYPE_IIO_ROLE_IIF, +}; + +#define ICMP_EXT_CTYPE_IIO_ROLE(ROLE) ((ROLE) << 6) +#define ICMP_EXT_CTYPE_IIO_MTU BIT(0) +#define ICMP_EXT_CTYPE_IIO_NAME BIT(1) +#define ICMP_EXT_CTYPE_IIO_IPADDR BIT(2) +#define ICMP_EXT_CTYPE_IIO_IFINDEX BIT(3) + +struct icmp_ext_iio_name_subobj { + u8 len; + char name[IFNAMSIZ]; +}; + +enum { + /* RFC 5837 - Incoming IP Interface Role */ + ICMP_ERR_EXT_IIO_IIF, + /* Add new constants above. Used by "icmp_errors_extension_mask" + * sysctl. + */ + ICMP_ERR_EXT_COUNT, +}; + #endif /* _LINUX_ICMP_H */ diff --git a/include/net/netns/ipv4.h b/include/net/netns/ipv4.h index 34eb3aecb3f2..0e96c90e56c6 100644 --- a/include/net/netns/ipv4.h +++ b/include/net/netns/ipv4.h @@ -135,6 +135,7 @@ struct netns_ipv4 { u8 sysctl_icmp_echo_ignore_broadcasts; u8 sysctl_icmp_ignore_bogus_error_responses; u8 sysctl_icmp_errors_use_inbound_ifaddr; + u8 sysctl_icmp_errors_extension_mask; int sysctl_icmp_ratelimit; int sysctl_icmp_ratemask; int sysctl_icmp_msgs_per_sec; diff --git a/net/ipv4/icmp.c b/net/ipv4/icmp.c index 1b7fb5d935ed..4abbec2f47ef 100644 --- a/net/ipv4/icmp.c +++ b/net/ipv4/icmp.c @@ -582,6 +582,185 @@ relookup_failed: return ERR_PTR(err); } +struct icmp_ext_iio_addr4_subobj { + __be16 afi; + __be16 reserved; + __be32 addr4; +}; + +static unsigned int icmp_ext_iio_len(void) +{ + return sizeof(struct icmp_extobj_hdr) + + /* ifIndex */ + sizeof(__be32) + + /* Interface Address Sub-Object */ + sizeof(struct icmp_ext_iio_addr4_subobj) + + /* Interface Name Sub-Object. Length must be a multiple of 4 + * bytes. + */ + ALIGN(sizeof(struct icmp_ext_iio_name_subobj), 4) + + /* MTU */ + sizeof(__be32); +} + +static unsigned int icmp_ext_max_len(u8 ext_objs) +{ + unsigned int ext_max_len; + + ext_max_len = sizeof(struct icmp_ext_hdr); + + if (ext_objs & BIT(ICMP_ERR_EXT_IIO_IIF)) + ext_max_len += icmp_ext_iio_len(); + + return ext_max_len; +} + +static __be32 icmp_ext_iio_addr4_find(const struct net_device *dev) +{ + struct in_device *in_dev; + struct in_ifaddr *ifa; + + in_dev = __in_dev_get_rcu(dev); + if (!in_dev) + return 0; + + /* It is unclear from RFC 5837 which IP address should be chosen, but + * it makes sense to choose a global unicast address. + */ + in_dev_for_each_ifa_rcu(ifa, in_dev) { + if (READ_ONCE(ifa->ifa_flags) & IFA_F_SECONDARY) + continue; + if (ifa->ifa_scope != RT_SCOPE_UNIVERSE || + ipv4_is_multicast(ifa->ifa_address)) + continue; + return ifa->ifa_address; + } + + return 0; +} + +static void icmp_ext_iio_iif_append(struct net *net, struct sk_buff *skb, + int iif) +{ + struct icmp_ext_iio_name_subobj *name_subobj; + struct icmp_extobj_hdr *objh; + struct net_device *dev; + __be32 data; + + if (!iif) + return; + + /* Add the fields in the order specified by RFC 5837. */ + objh = skb_put(skb, sizeof(*objh)); + objh->class_num = ICMP_EXT_OBJ_CLASS_IIO; + objh->class_type = ICMP_EXT_CTYPE_IIO_ROLE(ICMP_EXT_CTYPE_IIO_ROLE_IIF); + + data = htonl(iif); + skb_put_data(skb, &data, sizeof(__be32)); + objh->class_type |= ICMP_EXT_CTYPE_IIO_IFINDEX; + + rcu_read_lock(); + + dev = dev_get_by_index_rcu(net, iif); + if (!dev) + goto out; + + data = icmp_ext_iio_addr4_find(dev); + if (data) { + struct icmp_ext_iio_addr4_subobj *addr4_subobj; + + addr4_subobj = skb_put_zero(skb, sizeof(*addr4_subobj)); + addr4_subobj->afi = htons(ICMP_AFI_IP); + addr4_subobj->addr4 = data; + objh->class_type |= ICMP_EXT_CTYPE_IIO_IPADDR; + } + + name_subobj = skb_put_zero(skb, ALIGN(sizeof(*name_subobj), 4)); + name_subobj->len = ALIGN(sizeof(*name_subobj), 4); + netdev_copy_name(dev, name_subobj->name); + objh->class_type |= ICMP_EXT_CTYPE_IIO_NAME; + + data = htonl(READ_ONCE(dev->mtu)); + skb_put_data(skb, &data, sizeof(__be32)); + objh->class_type |= ICMP_EXT_CTYPE_IIO_MTU; + +out: + rcu_read_unlock(); + objh->length = htons(skb_tail_pointer(skb) - (unsigned char *)objh); +} + +static void icmp_ext_objs_append(struct net *net, struct sk_buff *skb, + u8 ext_objs, int iif) +{ + if (ext_objs & BIT(ICMP_ERR_EXT_IIO_IIF)) + icmp_ext_iio_iif_append(net, skb, iif); +} + +static struct sk_buff * +icmp_ext_append(struct net *net, struct sk_buff *skb_in, struct icmphdr *icmph, + unsigned int room, int iif) +{ + unsigned int payload_len, ext_max_len, ext_len; + struct icmp_ext_hdr *ext_hdr; + struct sk_buff *skb; + u8 ext_objs; + int nhoff; + + switch (icmph->type) { + case ICMP_DEST_UNREACH: + case ICMP_TIME_EXCEEDED: + case ICMP_PARAMETERPROB: + break; + default: + return NULL; + } + + ext_objs = READ_ONCE(net->ipv4.sysctl_icmp_errors_extension_mask); + if (!ext_objs) + return NULL; + + ext_max_len = icmp_ext_max_len(ext_objs); + if (ICMP_EXT_ORIG_DGRAM_MIN_LEN + ext_max_len > room) + return NULL; + + skb = skb_clone(skb_in, GFP_ATOMIC); + if (!skb) + return NULL; + + nhoff = skb_network_offset(skb); + payload_len = min(skb->len - nhoff, ICMP_EXT_ORIG_DGRAM_MIN_LEN); + + if (!pskb_network_may_pull(skb, payload_len)) + goto free_skb; + + if (pskb_trim(skb, nhoff + ICMP_EXT_ORIG_DGRAM_MIN_LEN) || + __skb_put_padto(skb, nhoff + ICMP_EXT_ORIG_DGRAM_MIN_LEN, false)) + goto free_skb; + + if (pskb_expand_head(skb, 0, ext_max_len, GFP_ATOMIC)) + goto free_skb; + + ext_hdr = skb_put_zero(skb, sizeof(*ext_hdr)); + ext_hdr->version = ICMP_EXT_VERSION_2; + + icmp_ext_objs_append(net, skb, ext_objs, iif); + + /* Do not send an empty extension structure. */ + ext_len = skb_tail_pointer(skb) - (unsigned char *)ext_hdr; + if (ext_len == sizeof(*ext_hdr)) + goto free_skb; + + ext_hdr->checksum = ip_compute_csum(ext_hdr, ext_len); + /* The length of the original datagram in 32-bit words (RFC 4884). */ + icmph->un.reserved[1] = ICMP_EXT_ORIG_DGRAM_MIN_LEN / sizeof(u32); + + return skb; + +free_skb: + consume_skb(skb); + return NULL; +} + /* * Send an ICMP message in response to a situation * @@ -601,6 +780,7 @@ void __icmp_send(struct sk_buff *skb_in, int type, int code, __be32 info, struct icmp_bxm icmp_param; struct rtable *rt = skb_rtable(skb_in); bool apply_ratelimit = false; + struct sk_buff *ext_skb; struct ipcm_cookie ipc; struct flowi4 fl4; __be32 saddr; @@ -770,7 +950,12 @@ void __icmp_send(struct sk_buff *skb_in, int type, int code, __be32 info, if (room <= (int)sizeof(struct iphdr)) goto ende; - icmp_param.data_len = skb_in->len - icmp_param.offset; + ext_skb = icmp_ext_append(net, skb_in, &icmp_param.data.icmph, room, + parm->iif); + if (ext_skb) + icmp_param.skb = ext_skb; + + icmp_param.data_len = icmp_param.skb->len - icmp_param.offset; if (icmp_param.data_len > room) icmp_param.data_len = room; icmp_param.head_len = sizeof(struct icmphdr); @@ -785,6 +970,9 @@ void __icmp_send(struct sk_buff *skb_in, int type, int code, __be32 info, trace_icmp_send(skb_in, type, code); icmp_push_reply(sk, &icmp_param, &fl4, &ipc, &rt); + + if (ext_skb) + consume_skb(ext_skb); ende: ip_rt_put(rt); out_unlock: @@ -1502,6 +1690,7 @@ static int __net_init icmp_sk_init(struct net *net) net->ipv4.sysctl_icmp_ratelimit = 1 * HZ; net->ipv4.sysctl_icmp_ratemask = 0x1818; net->ipv4.sysctl_icmp_errors_use_inbound_ifaddr = 0; + net->ipv4.sysctl_icmp_errors_extension_mask = 0; net->ipv4.sysctl_icmp_msgs_per_sec = 1000; net->ipv4.sysctl_icmp_msgs_burst = 50; diff --git a/net/ipv4/sysctl_net_ipv4.c b/net/ipv4/sysctl_net_ipv4.c index 24dbc603cc44..0c7c8f9041cb 100644 --- a/net/ipv4/sysctl_net_ipv4.c +++ b/net/ipv4/sysctl_net_ipv4.c @@ -48,6 +48,8 @@ static int tcp_plb_max_rounds = 31; static int tcp_plb_max_cong_thresh = 256; static unsigned int tcp_tw_reuse_delay_max = TCP_PAWS_MSL * MSEC_PER_SEC; static int tcp_ecn_mode_max = 2; +static u32 icmp_errors_extension_mask_all = + GENMASK_U8(ICMP_ERR_EXT_COUNT - 1, 0); /* obsolete */ static int sysctl_tcp_low_latency __read_mostly; @@ -674,6 +676,15 @@ static struct ctl_table ipv4_net_table[] = { .extra1 = SYSCTL_ZERO, .extra2 = SYSCTL_ONE }, + { + .procname = "icmp_errors_extension_mask", + .data = &init_net.ipv4.sysctl_icmp_errors_extension_mask, + .maxlen = sizeof(u8), + .mode = 0644, + .proc_handler = proc_dou8vec_minmax, + .extra1 = SYSCTL_ZERO, + .extra2 = &icmp_errors_extension_mask_all, + }, { .procname = "icmp_ratelimit", .data = &init_net.ipv4.sysctl_icmp_ratelimit, -- cgit v1.2.3 From d12d04d221f8d928a27a66236228e7501cd4cad5 Mon Sep 17 00:00:00 2001 From: Ido Schimmel Date: Mon, 27 Oct 2025 10:22:31 +0200 Subject: ipv6: icmp: Add RFC 5837 support Add the ability to append the incoming IP interface information to ICMPv6 error messages in accordance with RFC 5837 and RFC 4884. This is required for more meaningful traceroute results in unnumbered networks. The feature is disabled by default and controlled via a new sysctl ("net.ipv6.icmp.errors_extension_mask") which accepts a bitmask of ICMP extensions to append to ICMP error messages. Currently, only a single value is supported, but the interface and the implementation should be able to support more extensions, if needed. Clone the skb and copy the relevant data portions before modifying the skb as the caller of icmp6_send() still owns the skb after the function returns. This should be fine since by default ICMP error messages are rate limited to 1000 per second and no more than 1 per second per specific host. Trim or pad the packet to 128 bytes before appending the ICMP extension structure in order to be compatible with legacy applications that assume that the ICMP extension structure always starts at this offset (the minimum length specified by RFC 4884). Since commit 20e1954fe238 ("ipv6: RFC 4884 partial support for SIT/GRE tunnels") it is possible for icmp6_send() to be called with an skb that already contains ICMP extensions. This can happen when we receive an ICMPv4 message with extensions from a tunnel and translate it to an ICMPv6 message towards an IPv6 host in the overlay network. I could not find an RFC that supports this behavior, but it makes sense to not overwrite the original extensions that were appended to the packet. Therefore, avoid appending extensions if the length field in the provided ICMPv6 header is already filled. Export netdev_copy_name() using EXPORT_IPV6_MOD_GPL() to make it available to IPv6 when it is built as a module. Reviewed-by: Petr Machata Reviewed-by: David Ahern Reviewed-by: Willem de Bruijn Signed-off-by: Ido Schimmel Reviewed-by: Eric Dumazet Link: https://patch.msgid.link/20251027082232.232571-3-idosch@nvidia.com Signed-off-by: Jakub Kicinski --- Documentation/networking/ip-sysctl.rst | 17 +++ include/net/netns/ipv6.h | 1 + net/core/dev.c | 1 + net/ipv6/af_inet6.c | 1 + net/ipv6/icmp.c | 214 ++++++++++++++++++++++++++++++++- 5 files changed, 232 insertions(+), 2 deletions(-) (limited to 'include/net') diff --git a/Documentation/networking/ip-sysctl.rst b/Documentation/networking/ip-sysctl.rst index ece1187ba0f1..7cd35bfd39e6 100644 --- a/Documentation/networking/ip-sysctl.rst +++ b/Documentation/networking/ip-sysctl.rst @@ -3279,6 +3279,23 @@ error_anycast_as_unicast - BOOLEAN Default: 0 (disabled) +errors_extension_mask - UNSIGNED INTEGER + Bitmask of ICMP extensions to append to ICMPv6 error messages + ("Destination Unreachable" and "Time Exceeded"). The original datagram + is trimmed / padded to 128 bytes in order to be compatible with + applications that do not comply with RFC 4884. + + Possible extensions are: + + ==== ============================================================== + 0x01 Incoming IP interface information according to RFC 5837. + Extension will include the index, IPv6 address (if present), + name and MTU of the IP interface that received the datagram + which elicited the ICMP error. + ==== ============================================================== + + Default: 0x00 (no extensions) + xfrm6_gc_thresh - INTEGER (Obsolete since linux-4.14) The threshold at which we will start garbage collecting for IPv6 diff --git a/include/net/netns/ipv6.h b/include/net/netns/ipv6.h index 47dc70d8100a..08d2ecc96e2b 100644 --- a/include/net/netns/ipv6.h +++ b/include/net/netns/ipv6.h @@ -56,6 +56,7 @@ struct netns_sysctl_ipv6 { u8 skip_notify_on_dev_down; u8 fib_notify_on_flag_change; u8 icmpv6_error_anycast_as_unicast; + u8 icmpv6_errors_extension_mask; }; struct netns_ipv6 { diff --git a/net/core/dev.c b/net/core/dev.c index d32f0b0c03bb..dccc1176f3c6 100644 --- a/net/core/dev.c +++ b/net/core/dev.c @@ -1163,6 +1163,7 @@ void netdev_copy_name(struct net_device *dev, char *name) strscpy(name, dev->name, IFNAMSIZ); } while (read_seqretry(&netdev_rename_lock, seq)); } +EXPORT_IPV6_MOD_GPL(netdev_copy_name); /** * netdev_get_name - get a netdevice name, knowing its ifindex. diff --git a/net/ipv6/af_inet6.c b/net/ipv6/af_inet6.c index 1b0314644e0c..44d7de1eec4f 100644 --- a/net/ipv6/af_inet6.c +++ b/net/ipv6/af_inet6.c @@ -960,6 +960,7 @@ static int __net_init inet6_net_init(struct net *net) net->ipv6.sysctl.icmpv6_echo_ignore_multicast = 0; net->ipv6.sysctl.icmpv6_echo_ignore_anycast = 0; net->ipv6.sysctl.icmpv6_error_anycast_as_unicast = 0; + net->ipv6.sysctl.icmpv6_errors_extension_mask = 0; /* By default, rate limit error messages. * Except for pmtu discovery, it would break it. diff --git a/net/ipv6/icmp.c b/net/ipv6/icmp.c index 56c974cf75d1..5d2f90babaa5 100644 --- a/net/ipv6/icmp.c +++ b/net/ipv6/icmp.c @@ -444,6 +444,193 @@ static int icmp6_iif(const struct sk_buff *skb) return icmp6_dev(skb)->ifindex; } +struct icmp6_ext_iio_addr6_subobj { + __be16 afi; + __be16 reserved; + struct in6_addr addr6; +}; + +static unsigned int icmp6_ext_iio_len(void) +{ + return sizeof(struct icmp_extobj_hdr) + + /* ifIndex */ + sizeof(__be32) + + /* Interface Address Sub-Object */ + sizeof(struct icmp6_ext_iio_addr6_subobj) + + /* Interface Name Sub-Object. Length must be a multiple of 4 + * bytes. + */ + ALIGN(sizeof(struct icmp_ext_iio_name_subobj), 4) + + /* MTU */ + sizeof(__be32); +} + +static unsigned int icmp6_ext_max_len(u8 ext_objs) +{ + unsigned int ext_max_len; + + ext_max_len = sizeof(struct icmp_ext_hdr); + + if (ext_objs & BIT(ICMP_ERR_EXT_IIO_IIF)) + ext_max_len += icmp6_ext_iio_len(); + + return ext_max_len; +} + +static struct in6_addr *icmp6_ext_iio_addr6_find(const struct net_device *dev) +{ + struct inet6_dev *in6_dev; + struct inet6_ifaddr *ifa; + + in6_dev = __in6_dev_get(dev); + if (!in6_dev) + return NULL; + + /* It is unclear from RFC 5837 which IP address should be chosen, but + * it makes sense to choose a global unicast address. + */ + list_for_each_entry_rcu(ifa, &in6_dev->addr_list, if_list) { + if (ifa->flags & (IFA_F_TENTATIVE | IFA_F_DADFAILED)) + continue; + if (ipv6_addr_type(&ifa->addr) != IPV6_ADDR_UNICAST || + ipv6_addr_src_scope(&ifa->addr) != IPV6_ADDR_SCOPE_GLOBAL) + continue; + return &ifa->addr; + } + + return NULL; +} + +static void icmp6_ext_iio_iif_append(struct net *net, struct sk_buff *skb, + int iif) +{ + struct icmp_ext_iio_name_subobj *name_subobj; + struct icmp_extobj_hdr *objh; + struct net_device *dev; + struct in6_addr *addr6; + __be32 data; + + if (!iif) + return; + + /* Add the fields in the order specified by RFC 5837. */ + objh = skb_put(skb, sizeof(*objh)); + objh->class_num = ICMP_EXT_OBJ_CLASS_IIO; + objh->class_type = ICMP_EXT_CTYPE_IIO_ROLE(ICMP_EXT_CTYPE_IIO_ROLE_IIF); + + data = htonl(iif); + skb_put_data(skb, &data, sizeof(__be32)); + objh->class_type |= ICMP_EXT_CTYPE_IIO_IFINDEX; + + rcu_read_lock(); + + dev = dev_get_by_index_rcu(net, iif); + if (!dev) + goto out; + + addr6 = icmp6_ext_iio_addr6_find(dev); + if (addr6) { + struct icmp6_ext_iio_addr6_subobj *addr6_subobj; + + addr6_subobj = skb_put_zero(skb, sizeof(*addr6_subobj)); + addr6_subobj->afi = htons(ICMP_AFI_IP6); + addr6_subobj->addr6 = *addr6; + objh->class_type |= ICMP_EXT_CTYPE_IIO_IPADDR; + } + + name_subobj = skb_put_zero(skb, ALIGN(sizeof(*name_subobj), 4)); + name_subobj->len = ALIGN(sizeof(*name_subobj), 4); + netdev_copy_name(dev, name_subobj->name); + objh->class_type |= ICMP_EXT_CTYPE_IIO_NAME; + + data = htonl(READ_ONCE(dev->mtu)); + skb_put_data(skb, &data, sizeof(__be32)); + objh->class_type |= ICMP_EXT_CTYPE_IIO_MTU; + +out: + rcu_read_unlock(); + objh->length = htons(skb_tail_pointer(skb) - (unsigned char *)objh); +} + +static void icmp6_ext_objs_append(struct net *net, struct sk_buff *skb, + u8 ext_objs, int iif) +{ + if (ext_objs & BIT(ICMP_ERR_EXT_IIO_IIF)) + icmp6_ext_iio_iif_append(net, skb, iif); +} + +static struct sk_buff * +icmp6_ext_append(struct net *net, struct sk_buff *skb_in, + struct icmp6hdr *icmp6h, unsigned int room, int iif) +{ + unsigned int payload_len, ext_max_len, ext_len; + struct icmp_ext_hdr *ext_hdr; + struct sk_buff *skb; + u8 ext_objs; + int nhoff; + + switch (icmp6h->icmp6_type) { + case ICMPV6_DEST_UNREACH: + case ICMPV6_TIME_EXCEED: + break; + default: + return NULL; + } + + /* Do not overwrite existing extensions. This can happen when we + * receive an ICMPv4 message with extensions from a tunnel and + * translate it to an ICMPv6 message towards an IPv6 host in the + * overlay network. + */ + if (icmp6h->icmp6_datagram_len) + return NULL; + + ext_objs = READ_ONCE(net->ipv6.sysctl.icmpv6_errors_extension_mask); + if (!ext_objs) + return NULL; + + ext_max_len = icmp6_ext_max_len(ext_objs); + if (ICMP_EXT_ORIG_DGRAM_MIN_LEN + ext_max_len > room) + return NULL; + + skb = skb_clone(skb_in, GFP_ATOMIC); + if (!skb) + return NULL; + + nhoff = skb_network_offset(skb); + payload_len = min(skb->len - nhoff, ICMP_EXT_ORIG_DGRAM_MIN_LEN); + + if (!pskb_network_may_pull(skb, payload_len)) + goto free_skb; + + if (pskb_trim(skb, nhoff + ICMP_EXT_ORIG_DGRAM_MIN_LEN) || + __skb_put_padto(skb, nhoff + ICMP_EXT_ORIG_DGRAM_MIN_LEN, false)) + goto free_skb; + + if (pskb_expand_head(skb, 0, ext_max_len, GFP_ATOMIC)) + goto free_skb; + + ext_hdr = skb_put_zero(skb, sizeof(*ext_hdr)); + ext_hdr->version = ICMP_EXT_VERSION_2; + + icmp6_ext_objs_append(net, skb, ext_objs, iif); + + /* Do not send an empty extension structure. */ + ext_len = skb_tail_pointer(skb) - (unsigned char *)ext_hdr; + if (ext_len == sizeof(*ext_hdr)) + goto free_skb; + + ext_hdr->checksum = ip_compute_csum(ext_hdr, ext_len); + /* The length of the original datagram in 64-bit words (RFC 4884). */ + icmp6h->icmp6_datagram_len = ICMP_EXT_ORIG_DGRAM_MIN_LEN / sizeof(u64); + + return skb; + +free_skb: + consume_skb(skb); + return NULL; +} + /* * Send an ICMP message in response to a packet in error */ @@ -458,7 +645,9 @@ void icmp6_send(struct sk_buff *skb, u8 type, u8 code, __u32 info, struct ipv6_pinfo *np; const struct in6_addr *saddr = NULL; bool apply_ratelimit = false; + struct sk_buff *ext_skb; struct dst_entry *dst; + unsigned int room; struct icmp6hdr tmp_hdr; struct flowi6 fl6; struct icmpv6_msg msg; @@ -612,8 +801,13 @@ void icmp6_send(struct sk_buff *skb, u8 type, u8 code, __u32 info, msg.offset = skb_network_offset(skb); msg.type = type; - len = skb->len - msg.offset; - len = min_t(unsigned int, len, IPV6_MIN_MTU - sizeof(struct ipv6hdr) - sizeof(struct icmp6hdr)); + room = IPV6_MIN_MTU - sizeof(struct ipv6hdr) - sizeof(struct icmp6hdr); + ext_skb = icmp6_ext_append(net, skb, &tmp_hdr, room, parm->iif); + if (ext_skb) + msg.skb = ext_skb; + + len = msg.skb->len - msg.offset; + len = min_t(unsigned int, len, room); if (len < 0) { net_dbg_ratelimited("icmp: len problem [%pI6c > %pI6c]\n", &hdr->saddr, &hdr->daddr); @@ -635,6 +829,8 @@ void icmp6_send(struct sk_buff *skb, u8 type, u8 code, __u32 info, } out_dst_release: + if (ext_skb) + consume_skb(ext_skb); dst_release(dst); out_unlock: icmpv6_xmit_unlock(sk); @@ -1171,6 +1367,10 @@ int icmpv6_err_convert(u8 type, u8 code, int *err) EXPORT_SYMBOL(icmpv6_err_convert); #ifdef CONFIG_SYSCTL + +static u32 icmpv6_errors_extension_mask_all = + GENMASK_U8(ICMP_ERR_EXT_COUNT - 1, 0); + static struct ctl_table ipv6_icmp_table_template[] = { { .procname = "ratelimit", @@ -1216,6 +1416,15 @@ static struct ctl_table ipv6_icmp_table_template[] = { .extra1 = SYSCTL_ZERO, .extra2 = SYSCTL_ONE, }, + { + .procname = "errors_extension_mask", + .data = &init_net.ipv6.sysctl.icmpv6_errors_extension_mask, + .maxlen = sizeof(u8), + .mode = 0644, + .proc_handler = proc_dou8vec_minmax, + .extra1 = SYSCTL_ZERO, + .extra2 = &icmpv6_errors_extension_mask_all, + }, }; struct ctl_table * __net_init ipv6_icmp_sysctl_init(struct net *net) @@ -1233,6 +1442,7 @@ struct ctl_table * __net_init ipv6_icmp_sysctl_init(struct net *net) table[3].data = &net->ipv6.sysctl.icmpv6_echo_ignore_anycast; table[4].data = &net->ipv6.sysctl.icmpv6_ratemask_ptr; table[5].data = &net->ipv6.sysctl.icmpv6_error_anycast_as_unicast; + table[6].data = &net->ipv6.sysctl.icmpv6_errors_extension_mask; } return table; } -- cgit v1.2.3 From db82ddeaf42b93799a52df347284062893ea2ad6 Mon Sep 17 00:00:00 2001 From: Benjamin Berg Date: Mon, 27 Oct 2025 14:22:14 +0200 Subject: wifi: mac80211: add RX flag to report radiotap VHT information mac80211 already reports some basic information in the radiotap header with the known fields declared by the driver. However, drivers may want to report more accurate information and in that case the full VHT radiotap structure needs to be provided. Add a new RX_FLAG_RADIOTAP_VHT which is set when the VHT information should be pulled from the skb. Update the code to fill in the VHT fields to only do so when requested by the driver or if the information has not yet been set. This way the driver can fully control the information if it chooses so. Signed-off-by: Benjamin Berg Reviewed-by: Johannes Berg Signed-off-by: Miri Korenblit Link: https://patch.msgid.link/20251027142118.0bad1c307a21.I2cf285c20a822698039603f2af00ed9c548f2ee0@changeid Signed-off-by: Johannes Berg --- include/net/ieee80211_radiotap.h | 20 +++++++- include/net/mac80211.h | 2 + net/mac80211/rx.c | 104 +++++++++++++++++++++++++-------------- 3 files changed, 89 insertions(+), 37 deletions(-) (limited to 'include/net') diff --git a/include/net/ieee80211_radiotap.h b/include/net/ieee80211_radiotap.h index 813e163ce27c..c60867e7e43c 100644 --- a/include/net/ieee80211_radiotap.h +++ b/include/net/ieee80211_radiotap.h @@ -1,6 +1,6 @@ /* * Copyright (c) 2017 Intel Deutschland GmbH - * Copyright (c) 2018-2019, 2021-2022 Intel Corporation + * Copyright (c) 2018-2019, 2021-2022, 2025 Intel Corporation * * Permission to use, copy, modify, and/or distribute this software for any * purpose with or without fee is hereby granted, provided that the above @@ -202,6 +202,24 @@ enum ieee80211_radiotap_vht_coding { IEEE80211_RADIOTAP_CODING_LDPC_USER3 = 0x08, }; +enum ieee80211_radiotap_vht_bandwidth { + /* Note: more values are defined but can't really be used */ + IEEE80211_RADIOTAP_VHT_BW_20 = 0, + IEEE80211_RADIOTAP_VHT_BW_40 = 1, + IEEE80211_RADIOTAP_VHT_BW_80 = 4, + IEEE80211_RADIOTAP_VHT_BW_160 = 11, +}; + +struct ieee80211_radiotap_vht { + __le16 known; + u8 flags; + u8 bandwidth; + u8 mcs_nss[4]; + u8 coding; + u8 group_id; + __le16 partial_aid; +} __packed; + /* for IEEE80211_RADIOTAP_TIMESTAMP */ enum ieee80211_radiotap_timestamp_unit_spos { IEEE80211_RADIOTAP_TIMESTAMP_UNIT_MASK = 0x000F, diff --git a/include/net/mac80211.h b/include/net/mac80211.h index a55085cf4ec4..c326243e1f01 100644 --- a/include/net/mac80211.h +++ b/include/net/mac80211.h @@ -1529,6 +1529,7 @@ ieee80211_tx_info_clear_status(struct ieee80211_tx_info *info) * known the frame shouldn't be reported. * @RX_FLAG_8023: the frame has an 802.3 header (decap offload performed by * hardware or driver) + * @RX_FLAG_RADIOTAP_VHT: VHT radiotap data is present */ enum mac80211_rx_flags { RX_FLAG_MMIC_ERROR = BIT(0), @@ -1564,6 +1565,7 @@ enum mac80211_rx_flags { RX_FLAG_RADIOTAP_LSIG = BIT(28), RX_FLAG_NO_PSDU = BIT(29), RX_FLAG_8023 = BIT(30), + RX_FLAG_RADIOTAP_VHT = BIT(31), }; /** diff --git a/net/mac80211/rx.c b/net/mac80211/rx.c index 4641a2a80856..b59aeed340b3 100644 --- a/net/mac80211/rx.c +++ b/net/mac80211/rx.c @@ -59,7 +59,8 @@ static struct sk_buff *ieee80211_clean_skb(struct sk_buff *skb, status->flag &= ~(RX_FLAG_RADIOTAP_TLV_AT_END | RX_FLAG_RADIOTAP_LSIG | RX_FLAG_RADIOTAP_HE_MU | - RX_FLAG_RADIOTAP_HE); + RX_FLAG_RADIOTAP_HE | + RX_FLAG_RADIOTAP_VHT); hdr = (void *)skb->data; fc = hdr->frame_control; @@ -151,8 +152,10 @@ ieee80211_rx_radiotap_hdrlen(struct ieee80211_local *local, } if (status->encoding == RX_ENC_VHT) { + /* Included even if RX_FLAG_RADIOTAP_VHT is not set */ len = ALIGN(len, 2); len += 12; + BUILD_BUG_ON(sizeof(struct ieee80211_radiotap_vht) != 12); } if (local->hw.radiotap_timestamp.units_pos >= 0) { @@ -195,6 +198,9 @@ ieee80211_rx_radiotap_hdrlen(struct ieee80211_local *local, * The position to look at depends on the existence (or non- * existence) of other elements, so take that into account... */ + if (status->flag & RX_FLAG_RADIOTAP_VHT) + tlv_offset += + sizeof(struct ieee80211_radiotap_vht); if (status->flag & RX_FLAG_RADIOTAP_HE) tlv_offset += sizeof(struct ieee80211_radiotap_he); @@ -319,10 +325,17 @@ ieee80211_add_rx_radiotap_header(struct ieee80211_local *local, u32 tlvs_len = 0; int mpdulen, chain; unsigned long chains = status->chains; + struct ieee80211_radiotap_vht vht = {}; struct ieee80211_radiotap_he he = {}; struct ieee80211_radiotap_he_mu he_mu = {}; struct ieee80211_radiotap_lsig lsig = {}; + if (status->flag & RX_FLAG_RADIOTAP_VHT) { + vht = *(struct ieee80211_radiotap_vht *)skb->data; + skb_pull(skb, sizeof(vht)); + WARN_ON_ONCE(status->encoding != RX_ENC_VHT); + } + if (status->flag & RX_FLAG_RADIOTAP_HE) { he = *(struct ieee80211_radiotap_he *)skb->data; skb_pull(skb, sizeof(he)); @@ -530,45 +543,61 @@ ieee80211_add_rx_radiotap_header(struct ieee80211_local *local, } if (status->encoding == RX_ENC_VHT) { - u16 known = local->hw.radiotap_vht_details; + u16 fill = local->hw.radiotap_vht_details; - rthdr->it_present |= cpu_to_le32(BIT(IEEE80211_RADIOTAP_VHT)); - put_unaligned_le16(known, pos); - pos += 2; - /* flags */ - if (status->enc_flags & RX_ENC_FLAG_SHORT_GI) - *pos |= IEEE80211_RADIOTAP_VHT_FLAG_SGI; + /* Leave driver filled fields alone */ + fill &= ~le16_to_cpu(vht.known); + vht.known |= cpu_to_le16(fill); + + if (fill & IEEE80211_RADIOTAP_VHT_KNOWN_GI && + status->enc_flags & RX_ENC_FLAG_SHORT_GI) + vht.flags |= IEEE80211_RADIOTAP_VHT_FLAG_SGI; /* in VHT, STBC is binary */ - if (status->enc_flags & RX_ENC_FLAG_STBC_MASK) - *pos |= IEEE80211_RADIOTAP_VHT_FLAG_STBC; - if (status->enc_flags & RX_ENC_FLAG_BF) + if (fill & IEEE80211_RADIOTAP_VHT_KNOWN_STBC && + status->enc_flags & RX_ENC_FLAG_STBC_MASK) + vht.flags |= IEEE80211_RADIOTAP_VHT_FLAG_STBC; + if (fill & IEEE80211_RADIOTAP_VHT_KNOWN_BEAMFORMED && + status->enc_flags & RX_ENC_FLAG_BF) *pos |= IEEE80211_RADIOTAP_VHT_FLAG_BEAMFORMED; - pos++; - /* bandwidth */ - switch (status->bw) { - case RATE_INFO_BW_80: - *pos++ = 4; - break; - case RATE_INFO_BW_160: - *pos++ = 11; - break; - case RATE_INFO_BW_40: - *pos++ = 1; - break; - default: - *pos++ = 0; + + if (fill & IEEE80211_RADIOTAP_VHT_KNOWN_BANDWIDTH) { + switch (status->bw) { + case RATE_INFO_BW_40: + vht.bandwidth = IEEE80211_RADIOTAP_VHT_BW_40; + break; + case RATE_INFO_BW_80: + vht.bandwidth = IEEE80211_RADIOTAP_VHT_BW_80; + break; + case RATE_INFO_BW_160: + vht.bandwidth = IEEE80211_RADIOTAP_VHT_BW_160; + break; + default: + vht.bandwidth = IEEE80211_RADIOTAP_VHT_BW_20; + break; + } } - /* MCS/NSS */ - *pos = (status->rate_idx << 4) | status->nss; - pos += 4; - /* coding field */ - if (status->enc_flags & RX_ENC_FLAG_LDPC) - *pos |= IEEE80211_RADIOTAP_CODING_LDPC_USER0; - pos++; - /* group ID */ - pos++; - /* partial_aid */ - pos += 2; + + /* + * If the driver filled in mcs_nss[0], then do not touch it. + * + * Otherwise, put some information about MCS/NSS into the + * user 0 field. Note that this is not technically correct for + * an MU frame as we might have decoded a different user. + */ + if (!vht.mcs_nss[0]) { + vht.mcs_nss[0] = (status->rate_idx << 4) | status->nss; + + /* coding field */ + if (status->enc_flags & RX_ENC_FLAG_LDPC) + vht.coding |= IEEE80211_RADIOTAP_CODING_LDPC_USER0; + } + + /* ensure 2 byte alignment */ + while ((pos - (u8 *)rthdr) & 1) + pos++; + rthdr->it_present |= cpu_to_le32(BIT(IEEE80211_RADIOTAP_VHT)); + memcpy(pos, &vht, sizeof(vht)); + pos += sizeof(vht); } if (local->hw.radiotap_timestamp.units_pos >= 0) { @@ -834,6 +863,9 @@ ieee80211_rx_monitor(struct ieee80211_local *local, struct sk_buff *origskb, return NULL; } + if (status->flag & RX_FLAG_RADIOTAP_VHT) + rtap_space += sizeof(struct ieee80211_radiotap_vht); + if (status->flag & RX_FLAG_RADIOTAP_HE) rtap_space += sizeof(struct ieee80211_radiotap_he); -- cgit v1.2.3 From 61fafbee6cfed283c02a320896089f658fa67e56 Mon Sep 17 00:00:00 2001 From: Jianbo Liu Date: Tue, 28 Oct 2025 04:22:48 +0200 Subject: xfrm: Determine inner GSO type from packet inner protocol The GSO segmentation functions for ESP tunnel mode (xfrm4_tunnel_gso_segment and xfrm6_tunnel_gso_segment) were determining the inner packet's L2 protocol type by checking the static x->inner_mode.family field from the xfrm state. This is unreliable. In tunnel mode, the state's actual inner family could be defined by x->inner_mode.family or by x->inner_mode_iaf.family. Checking only the former can lead to a mismatch with the actual packet being processed, causing GSO to create segments with the wrong L2 header type. This patch fixes the bug by deriving the inner mode directly from the packet's inner protocol stored in XFRM_MODE_SKB_CB(skb)->protocol. Instead of replicating the code, this patch modifies the xfrm_ip2inner_mode helper function. It now correctly returns &x->inner_mode if the selector family (x->sel.family) is already specified, thereby handling both specific and AF_UNSPEC cases appropriately. With this change, ESP GSO can use xfrm_ip2inner_mode to get the correct inner mode. It doesn't affect existing callers, as the updated logic now mirrors the checks they were already performing externally. Fixes: 26dbd66eab80 ("esp: choose the correct inner protocol for GSO on inter address family tunnels") Signed-off-by: Jianbo Liu Reviewed-by: Cosmin Ratiu Reviewed-by: Sabrina Dubroca Signed-off-by: Steffen Klassert --- include/net/xfrm.h | 3 ++- net/ipv4/esp4_offload.c | 6 ++++-- net/ipv6/esp6_offload.c | 6 ++++-- 3 files changed, 10 insertions(+), 5 deletions(-) (limited to 'include/net') diff --git a/include/net/xfrm.h b/include/net/xfrm.h index f3014e4f54fc..0a14daaa5dd4 100644 --- a/include/net/xfrm.h +++ b/include/net/xfrm.h @@ -536,7 +536,8 @@ static inline int xfrm_af2proto(unsigned int family) static inline const struct xfrm_mode *xfrm_ip2inner_mode(struct xfrm_state *x, int ipproto) { - if ((ipproto == IPPROTO_IPIP && x->props.family == AF_INET) || + if ((x->sel.family != AF_UNSPEC) || + (ipproto == IPPROTO_IPIP && x->props.family == AF_INET) || (ipproto == IPPROTO_IPV6 && x->props.family == AF_INET6)) return &x->inner_mode; else diff --git a/net/ipv4/esp4_offload.c b/net/ipv4/esp4_offload.c index e0d94270da28..05828d4cb6cd 100644 --- a/net/ipv4/esp4_offload.c +++ b/net/ipv4/esp4_offload.c @@ -122,8 +122,10 @@ static struct sk_buff *xfrm4_tunnel_gso_segment(struct xfrm_state *x, struct sk_buff *skb, netdev_features_t features) { - __be16 type = x->inner_mode.family == AF_INET6 ? htons(ETH_P_IPV6) - : htons(ETH_P_IP); + const struct xfrm_mode *inner_mode = xfrm_ip2inner_mode(x, + XFRM_MODE_SKB_CB(skb)->protocol); + __be16 type = inner_mode->family == AF_INET6 ? htons(ETH_P_IPV6) + : htons(ETH_P_IP); return skb_eth_gso_segment(skb, features, type); } diff --git a/net/ipv6/esp6_offload.c b/net/ipv6/esp6_offload.c index 7b41fb4f00b5..22410243ebe8 100644 --- a/net/ipv6/esp6_offload.c +++ b/net/ipv6/esp6_offload.c @@ -158,8 +158,10 @@ static struct sk_buff *xfrm6_tunnel_gso_segment(struct xfrm_state *x, struct sk_buff *skb, netdev_features_t features) { - __be16 type = x->inner_mode.family == AF_INET ? htons(ETH_P_IP) - : htons(ETH_P_IPV6); + const struct xfrm_mode *inner_mode = xfrm_ip2inner_mode(x, + XFRM_MODE_SKB_CB(skb)->protocol); + __be16 type = inner_mode->family == AF_INET ? htons(ETH_P_IP) + : htons(ETH_P_IPV6); return skb_eth_gso_segment(skb, features, type); } -- cgit v1.2.3 From 57347d58a4011551e7d0e030f2f12e4d1a28feb6 Mon Sep 17 00:00:00 2001 From: "caivive (Weibiao Tu)" Date: Thu, 28 Nov 2024 20:52:04 +0800 Subject: netfilter: fix typo in nf_conntrack_l4proto.h comment In the comment for nf_conntrack_l4proto.h, the word "nfnetink" was incorrectly spelled. It has been corrected to "nfnetlink". Fixes a typo to enhance readability and ensure consistency. Signed-off-by: caivive (Weibiao Tu) Signed-off-by: Florian Westphal --- include/net/netfilter/nf_conntrack_l4proto.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'include/net') diff --git a/include/net/netfilter/nf_conntrack_l4proto.h b/include/net/netfilter/nf_conntrack_l4proto.h index 6929f8daf1ed..cd5020835a6d 100644 --- a/include/net/netfilter/nf_conntrack_l4proto.h +++ b/include/net/netfilter/nf_conntrack_l4proto.h @@ -30,7 +30,7 @@ struct nf_conntrack_l4proto { /* called by gc worker if table is full */ bool (*can_early_drop)(const struct nf_conn *ct); - /* convert protoinfo to nfnetink attributes */ + /* convert protoinfo to nfnetlink attributes */ int (*to_nlattr)(struct sk_buff *skb, struct nlattr *nla, struct nf_conn *ct, bool destroy); -- cgit v1.2.3 From aef3cdb47bbbef9fea9512ed6c02d64394449d53 Mon Sep 17 00:00:00 2001 From: Halil Pasic Date: Mon, 27 Oct 2025 23:48:55 +0100 Subject: net/smc: make wr buffer count configurable Think SMC_WR_BUF_CNT_SEND := SMC_WR_BUF_CNT used in send context and SMC_WR_BUF_CNT_RECV := 3 * SMC_WR_BUF_CNT used in recv context. Those get replaced with lgr->max_send_wr and lgr->max_recv_wr respective. Please note that although with the default sysctl values qp_attr.cap.max_send_wr == qp_attr.cap.max_recv_wr is maintained but can not be assumed to be generally true any more. I see no downside to that, but my confidence level is rather modest. Signed-off-by: Halil Pasic Reviewed-by: Sidraya Jayagond Reviewed-by: Dust Li Tested-by: Mahanta Jambigi Link: https://patch.msgid.link/20251027224856.2970019-2-pasic@linux.ibm.com Signed-off-by: Paolo Abeni --- Documentation/networking/smc-sysctl.rst | 36 +++++++++++++++++++++++++++++++++ include/net/netns/smc.h | 2 ++ net/smc/smc_core.h | 6 ++++++ net/smc/smc_ib.c | 10 ++++----- net/smc/smc_llc.c | 2 ++ net/smc/smc_sysctl.c | 22 ++++++++++++++++++++ net/smc/smc_sysctl.h | 2 ++ net/smc/smc_wr.c | 31 ++++++++++++++-------------- net/smc/smc_wr.h | 2 -- 9 files changed, 91 insertions(+), 22 deletions(-) (limited to 'include/net') diff --git a/Documentation/networking/smc-sysctl.rst b/Documentation/networking/smc-sysctl.rst index a874d007f2db..337ac2be167e 100644 --- a/Documentation/networking/smc-sysctl.rst +++ b/Documentation/networking/smc-sysctl.rst @@ -71,3 +71,39 @@ smcr_max_conns_per_lgr - INTEGER acceptable value ranges from 16 to 255. Only for SMC-R v2.1 and later. Default: 255 + +smcr_max_send_wr - INTEGER + So-called work request buffers are SMCR link (and RDMA queue pair) level + resources necessary for performing RDMA operations. Since up to 255 + connections can share a link group and thus also a link and the number + of the work request buffers is decided when the link is allocated, + depending on the workload it can be a bottleneck in a sense that threads + have to wait for work request buffers to become available. Before the + introduction of this control the maximal number of work request buffers + available on the send path used to be hard coded to 16. With this control + it becomes configurable. The acceptable range is between 2 and 2048. + + Please be aware that all the buffers need to be allocated as a physically + continuous array in which each element is a single buffer and has the size + of SMC_WR_BUF_SIZE (48) bytes. If the allocation fails we give up much + like before having this control. + + Default: 16 + +smcr_max_recv_wr - INTEGER + So-called work request buffers are SMCR link (and RDMA queue pair) level + resources necessary for performing RDMA operations. Since up to 255 + connections can share a link group and thus also a link and the number + of the work request buffers is decided when the link is allocated, + depending on the workload it can be a bottleneck in a sense that threads + have to wait for work request buffers to become available. Before the + introduction of this control the maximal number of work request buffers + available on the receive path used to be hard coded to 16. With this control + it becomes configurable. The acceptable range is between 2 and 2048. + + Please be aware that all the buffers need to be allocated as a physically + continuous array in which each element is a single buffer and has the size + of SMC_WR_BUF_SIZE (48) bytes. If the allocation fails we give up much + like before having this control. + + Default: 48 diff --git a/include/net/netns/smc.h b/include/net/netns/smc.h index fc752a50f91b..6ceb12baec24 100644 --- a/include/net/netns/smc.h +++ b/include/net/netns/smc.h @@ -24,5 +24,7 @@ struct netns_smc { int sysctl_rmem; int sysctl_max_links_per_lgr; int sysctl_max_conns_per_lgr; + unsigned int sysctl_smcr_max_send_wr; + unsigned int sysctl_smcr_max_recv_wr; }; #endif diff --git a/net/smc/smc_core.h b/net/smc/smc_core.h index a5a78cbff341..8d06c8bb14e9 100644 --- a/net/smc/smc_core.h +++ b/net/smc/smc_core.h @@ -34,6 +34,8 @@ * distributions may modify it to a value between * 16-255 as needed. */ +#define SMCR_MAX_SEND_WR_DEF 16 /* Default number of work requests per send queue */ +#define SMCR_MAX_RECV_WR_DEF 48 /* Default number of work requests per recv queue */ struct smc_lgr_list { /* list of link group definition */ struct list_head list; @@ -366,6 +368,10 @@ struct smc_link_group { /* max conn can be assigned to lgr */ u8 max_links; /* max links can be added in lgr */ + u16 max_send_wr; + /* number of WR buffers on send */ + u16 max_recv_wr; + /* number of WR buffers on recv */ }; struct { /* SMC-D */ struct smcd_gid peer_gid; diff --git a/net/smc/smc_ib.c b/net/smc/smc_ib.c index 0052f02756eb..1154907c5c05 100644 --- a/net/smc/smc_ib.c +++ b/net/smc/smc_ib.c @@ -669,11 +669,6 @@ int smc_ib_create_queue_pair(struct smc_link *lnk) .recv_cq = lnk->smcibdev->roce_cq_recv, .srq = NULL, .cap = { - /* include unsolicited rdma_writes as well, - * there are max. 2 RDMA_WRITE per 1 WR_SEND - */ - .max_send_wr = SMC_WR_BUF_CNT * 3, - .max_recv_wr = SMC_WR_BUF_CNT * 3, .max_send_sge = SMC_IB_MAX_SEND_SGE, .max_recv_sge = lnk->wr_rx_sge_cnt, .max_inline_data = 0, @@ -683,6 +678,11 @@ int smc_ib_create_queue_pair(struct smc_link *lnk) }; int rc; + /* include unsolicited rdma_writes as well, + * there are max. 2 RDMA_WRITE per 1 WR_SEND + */ + qp_attr.cap.max_send_wr = 3 * lnk->lgr->max_send_wr; + qp_attr.cap.max_recv_wr = lnk->lgr->max_recv_wr; lnk->roce_qp = ib_create_qp(lnk->roce_pd, &qp_attr); rc = PTR_ERR_OR_ZERO(lnk->roce_qp); if (IS_ERR(lnk->roce_qp)) diff --git a/net/smc/smc_llc.c b/net/smc/smc_llc.c index f865c58c3aa7..f5d5eb617526 100644 --- a/net/smc/smc_llc.c +++ b/net/smc/smc_llc.c @@ -2157,6 +2157,8 @@ void smc_llc_lgr_init(struct smc_link_group *lgr, struct smc_sock *smc) init_waitqueue_head(&lgr->llc_msg_waiter); init_rwsem(&lgr->llc_conf_mutex); lgr->llc_testlink_time = READ_ONCE(net->smc.sysctl_smcr_testlink_time); + lgr->max_send_wr = (u16)(READ_ONCE(net->smc.sysctl_smcr_max_send_wr)); + lgr->max_recv_wr = (u16)(READ_ONCE(net->smc.sysctl_smcr_max_recv_wr)); } /* called after lgr was removed from lgr_list */ diff --git a/net/smc/smc_sysctl.c b/net/smc/smc_sysctl.c index 2fab6456f765..7b2471904d04 100644 --- a/net/smc/smc_sysctl.c +++ b/net/smc/smc_sysctl.c @@ -29,6 +29,8 @@ static int links_per_lgr_min = SMC_LINKS_ADD_LNK_MIN; static int links_per_lgr_max = SMC_LINKS_ADD_LNK_MAX; static int conns_per_lgr_min = SMC_CONN_PER_LGR_MIN; static int conns_per_lgr_max = SMC_CONN_PER_LGR_MAX; +static unsigned int smcr_max_wr_min = 2; +static unsigned int smcr_max_wr_max = 2048; static struct ctl_table smc_table[] = { { @@ -99,6 +101,24 @@ static struct ctl_table smc_table[] = { .extra1 = SYSCTL_ZERO, .extra2 = SYSCTL_ONE, }, + { + .procname = "smcr_max_send_wr", + .data = &init_net.smc.sysctl_smcr_max_send_wr, + .maxlen = sizeof(int), + .mode = 0644, + .proc_handler = proc_dointvec_minmax, + .extra1 = &smcr_max_wr_min, + .extra2 = &smcr_max_wr_max, + }, + { + .procname = "smcr_max_recv_wr", + .data = &init_net.smc.sysctl_smcr_max_recv_wr, + .maxlen = sizeof(int), + .mode = 0644, + .proc_handler = proc_dointvec_minmax, + .extra1 = &smcr_max_wr_min, + .extra2 = &smcr_max_wr_max, + }, }; int __net_init smc_sysctl_net_init(struct net *net) @@ -130,6 +150,8 @@ int __net_init smc_sysctl_net_init(struct net *net) WRITE_ONCE(net->smc.sysctl_rmem, net_smc_rmem_init); net->smc.sysctl_max_links_per_lgr = SMC_LINKS_PER_LGR_MAX_PREFER; net->smc.sysctl_max_conns_per_lgr = SMC_CONN_PER_LGR_PREFER; + net->smc.sysctl_smcr_max_send_wr = SMCR_MAX_SEND_WR_DEF; + net->smc.sysctl_smcr_max_recv_wr = SMCR_MAX_RECV_WR_DEF; /* disable handshake limitation by default */ net->smc.limit_smc_hs = 0; diff --git a/net/smc/smc_sysctl.h b/net/smc/smc_sysctl.h index eb2465ae1e15..8538915af7af 100644 --- a/net/smc/smc_sysctl.h +++ b/net/smc/smc_sysctl.h @@ -25,6 +25,8 @@ static inline int smc_sysctl_net_init(struct net *net) net->smc.sysctl_autocorking_size = SMC_AUTOCORKING_DEFAULT_SIZE; net->smc.sysctl_max_links_per_lgr = SMC_LINKS_PER_LGR_MAX_PREFER; net->smc.sysctl_max_conns_per_lgr = SMC_CONN_PER_LGR_PREFER; + net->smc.sysctl_smcr_max_send_wr = SMCR_MAX_SEND_WR_DEF; + net->smc.sysctl_smcr_max_recv_wr = SMCR_MAX_RECV_WR_DEF; return 0; } diff --git a/net/smc/smc_wr.c b/net/smc/smc_wr.c index b04a21b8c511..883fb0f1ce43 100644 --- a/net/smc/smc_wr.c +++ b/net/smc/smc_wr.c @@ -547,9 +547,9 @@ void smc_wr_remember_qp_attr(struct smc_link *lnk) IB_QP_DEST_QPN, &init_attr); - lnk->wr_tx_cnt = min_t(size_t, SMC_WR_BUF_CNT, + lnk->wr_tx_cnt = min_t(size_t, lnk->lgr->max_send_wr, lnk->qp_attr.cap.max_send_wr); - lnk->wr_rx_cnt = min_t(size_t, SMC_WR_BUF_CNT * 3, + lnk->wr_rx_cnt = min_t(size_t, lnk->lgr->max_recv_wr, lnk->qp_attr.cap.max_recv_wr); } @@ -741,50 +741,51 @@ int smc_wr_alloc_lgr_mem(struct smc_link_group *lgr) int smc_wr_alloc_link_mem(struct smc_link *link) { /* allocate link related memory */ - link->wr_tx_bufs = kcalloc(SMC_WR_BUF_CNT, SMC_WR_BUF_SIZE, GFP_KERNEL); + link->wr_tx_bufs = kcalloc(link->lgr->max_send_wr, + SMC_WR_BUF_SIZE, GFP_KERNEL); if (!link->wr_tx_bufs) goto no_mem; - link->wr_rx_bufs = kcalloc(SMC_WR_BUF_CNT * 3, link->wr_rx_buflen, + link->wr_rx_bufs = kcalloc(link->lgr->max_recv_wr, link->wr_rx_buflen, GFP_KERNEL); if (!link->wr_rx_bufs) goto no_mem_wr_tx_bufs; - link->wr_tx_ibs = kcalloc(SMC_WR_BUF_CNT, sizeof(link->wr_tx_ibs[0]), - GFP_KERNEL); + link->wr_tx_ibs = kcalloc(link->lgr->max_send_wr, + sizeof(link->wr_tx_ibs[0]), GFP_KERNEL); if (!link->wr_tx_ibs) goto no_mem_wr_rx_bufs; - link->wr_rx_ibs = kcalloc(SMC_WR_BUF_CNT * 3, + link->wr_rx_ibs = kcalloc(link->lgr->max_recv_wr, sizeof(link->wr_rx_ibs[0]), GFP_KERNEL); if (!link->wr_rx_ibs) goto no_mem_wr_tx_ibs; - link->wr_tx_rdmas = kcalloc(SMC_WR_BUF_CNT, + link->wr_tx_rdmas = kcalloc(link->lgr->max_send_wr, sizeof(link->wr_tx_rdmas[0]), GFP_KERNEL); if (!link->wr_tx_rdmas) goto no_mem_wr_rx_ibs; - link->wr_tx_rdma_sges = kcalloc(SMC_WR_BUF_CNT, + link->wr_tx_rdma_sges = kcalloc(link->lgr->max_send_wr, sizeof(link->wr_tx_rdma_sges[0]), GFP_KERNEL); if (!link->wr_tx_rdma_sges) goto no_mem_wr_tx_rdmas; - link->wr_tx_sges = kcalloc(SMC_WR_BUF_CNT, sizeof(link->wr_tx_sges[0]), + link->wr_tx_sges = kcalloc(link->lgr->max_send_wr, sizeof(link->wr_tx_sges[0]), GFP_KERNEL); if (!link->wr_tx_sges) goto no_mem_wr_tx_rdma_sges; - link->wr_rx_sges = kcalloc(SMC_WR_BUF_CNT * 3, + link->wr_rx_sges = kcalloc(link->lgr->max_recv_wr, sizeof(link->wr_rx_sges[0]) * link->wr_rx_sge_cnt, GFP_KERNEL); if (!link->wr_rx_sges) goto no_mem_wr_tx_sges; - link->wr_tx_mask = bitmap_zalloc(SMC_WR_BUF_CNT, GFP_KERNEL); + link->wr_tx_mask = bitmap_zalloc(link->lgr->max_send_wr, GFP_KERNEL); if (!link->wr_tx_mask) goto no_mem_wr_rx_sges; - link->wr_tx_pends = kcalloc(SMC_WR_BUF_CNT, + link->wr_tx_pends = kcalloc(link->lgr->max_send_wr, sizeof(link->wr_tx_pends[0]), GFP_KERNEL); if (!link->wr_tx_pends) goto no_mem_wr_tx_mask; - link->wr_tx_compl = kcalloc(SMC_WR_BUF_CNT, + link->wr_tx_compl = kcalloc(link->lgr->max_send_wr, sizeof(link->wr_tx_compl[0]), GFP_KERNEL); if (!link->wr_tx_compl) @@ -905,7 +906,7 @@ int smc_wr_create_link(struct smc_link *lnk) goto dma_unmap; } smc_wr_init_sge(lnk); - bitmap_zero(lnk->wr_tx_mask, SMC_WR_BUF_CNT); + bitmap_zero(lnk->wr_tx_mask, lnk->lgr->max_send_wr); init_waitqueue_head(&lnk->wr_tx_wait); rc = percpu_ref_init(&lnk->wr_tx_refs, smcr_wr_tx_refs_free, 0, GFP_KERNEL); if (rc) diff --git a/net/smc/smc_wr.h b/net/smc/smc_wr.h index f3008dda222a..aa4533af9122 100644 --- a/net/smc/smc_wr.h +++ b/net/smc/smc_wr.h @@ -19,8 +19,6 @@ #include "smc.h" #include "smc_core.h" -#define SMC_WR_BUF_CNT 16 /* # of ctrl buffers per link */ - #define SMC_WR_TX_WAIT_FREE_SLOT_TIME (10 * HZ) #define SMC_WR_TX_SIZE 44 /* actual size of wr_send data (<=SMC_WR_BUF_SIZE) */ -- cgit v1.2.3 From 54133f9b4b53ffa2204eb27cfc9d50072c9a52d2 Mon Sep 17 00:00:00 2001 From: Haiyang Zhang Date: Wed, 29 Oct 2025 13:43:10 -0700 Subject: net: mana: Support HW link state events Handle the NIC hardware link state events received from the HW channel, then set the proper link state accordingly. And, add a feature bit, GDMA_DRV_CAP_FLAG_1_HW_VPORT_LINK_AWARE, to inform the NIC hardware this handler exists. Our MANA NIC only sends out the link state down/up messages when we need to let the VM rerun DHCP client and change IP address. So, add netif_carrier_on() in the probe(), let the NIC show the right initial state in /sys/class/net/ethX/operstate. Signed-off-by: Haiyang Zhang Link: https://patch.msgid.link/1761770601-16920-1-git-send-email-haiyangz@linux.microsoft.com Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/microsoft/mana/gdma_main.c | 1 + drivers/net/ethernet/microsoft/mana/hw_channel.c | 12 ++++++ drivers/net/ethernet/microsoft/mana/mana_en.c | 54 +++++++++++++++++++++--- include/net/mana/gdma.h | 4 +- include/net/mana/hw_channel.h | 2 + include/net/mana/mana.h | 4 ++ 6 files changed, 71 insertions(+), 6 deletions(-) (limited to 'include/net') diff --git a/drivers/net/ethernet/microsoft/mana/gdma_main.c b/drivers/net/ethernet/microsoft/mana/gdma_main.c index 43f034e180c4..effe0a2f207a 100644 --- a/drivers/net/ethernet/microsoft/mana/gdma_main.c +++ b/drivers/net/ethernet/microsoft/mana/gdma_main.c @@ -528,6 +528,7 @@ static void mana_gd_process_eqe(struct gdma_queue *eq) case GDMA_EQE_HWC_INIT_DONE: case GDMA_EQE_HWC_SOC_SERVICE: case GDMA_EQE_RNIC_QP_FATAL: + case GDMA_EQE_HWC_SOC_RECONFIG_DATA: if (!eq->eq.callback) break; diff --git a/drivers/net/ethernet/microsoft/mana/hw_channel.c b/drivers/net/ethernet/microsoft/mana/hw_channel.c index ada6c78a2bef..aa4e2731e2ba 100644 --- a/drivers/net/ethernet/microsoft/mana/hw_channel.c +++ b/drivers/net/ethernet/microsoft/mana/hw_channel.c @@ -118,6 +118,7 @@ static void mana_hwc_init_event_handler(void *ctx, struct gdma_queue *q_self, struct gdma_dev *gd = hwc->gdma_dev; union hwc_init_type_data type_data; union hwc_init_eq_id_db eq_db; + struct mana_context *ac; u32 type, val; int ret; @@ -196,6 +197,17 @@ static void mana_hwc_init_event_handler(void *ctx, struct gdma_queue *q_self, hwc->hwc_timeout = val; break; + case HWC_DATA_HW_LINK_CONNECT: + case HWC_DATA_HW_LINK_DISCONNECT: + ac = gd->gdma_context->mana.driver_data; + if (!ac) + break; + + WRITE_ONCE(ac->link_event, type); + schedule_work(&ac->link_change_work); + + break; + default: dev_warn(hwc->dev, "Received unknown reconfig type %u\n", type); break; diff --git a/drivers/net/ethernet/microsoft/mana/mana_en.c b/drivers/net/ethernet/microsoft/mana/mana_en.c index 0142fd98392c..739087081dfd 100644 --- a/drivers/net/ethernet/microsoft/mana/mana_en.c +++ b/drivers/net/ethernet/microsoft/mana/mana_en.c @@ -20,6 +20,7 @@ #include #include +#include static DEFINE_IDA(mana_adev_ida); @@ -84,7 +85,6 @@ static int mana_open(struct net_device *ndev) /* Ensure port state updated before txq state */ smp_wmb(); - netif_carrier_on(ndev); netif_tx_wake_all_queues(ndev); netdev_dbg(ndev, "%s successful\n", __func__); return 0; @@ -100,6 +100,46 @@ static int mana_close(struct net_device *ndev) return mana_detach(ndev, true); } +static void mana_link_state_handle(struct work_struct *w) +{ + struct mana_context *ac; + struct net_device *ndev; + u32 link_event; + bool link_up; + int i; + + ac = container_of(w, struct mana_context, link_change_work); + + rtnl_lock(); + + link_event = READ_ONCE(ac->link_event); + + if (link_event == HWC_DATA_HW_LINK_CONNECT) + link_up = true; + else if (link_event == HWC_DATA_HW_LINK_DISCONNECT) + link_up = false; + else + goto out; + + /* Process all ports */ + for (i = 0; i < ac->num_ports; i++) { + ndev = ac->ports[i]; + if (!ndev) + continue; + + if (link_up) { + netif_carrier_on(ndev); + + __netdev_notify_peers(ndev); + } else { + netif_carrier_off(ndev); + } + } + +out: + rtnl_unlock(); +} + static bool mana_can_tx(struct gdma_queue *wq) { return mana_gd_wq_avail_space(wq) >= MAX_TX_WQE_SIZE; @@ -3059,9 +3099,6 @@ int mana_attach(struct net_device *ndev) /* Ensure port state updated before txq state */ smp_wmb(); - if (apc->port_is_up) - netif_carrier_on(ndev); - netif_device_attach(ndev); return 0; @@ -3154,7 +3191,6 @@ int mana_detach(struct net_device *ndev, bool from_close) smp_wmb(); netif_tx_disable(ndev); - netif_carrier_off(ndev); if (apc->port_st_save) { err = mana_dealloc_queues(ndev); @@ -3243,6 +3279,8 @@ static int mana_probe_port(struct mana_context *ac, int port_idx, goto free_indir; } + netif_carrier_on(ndev); + debugfs_create_u32("current_speed", 0400, apc->mana_port_debugfs, &apc->speed); return 0; @@ -3431,6 +3469,8 @@ int mana_probe(struct gdma_dev *gd, bool resuming) if (!resuming) { ac->num_ports = num_ports; + + INIT_WORK(&ac->link_change_work, mana_link_state_handle); } else { if (ac->num_ports != num_ports) { dev_err(dev, "The number of vPorts changed: %d->%d\n", @@ -3438,6 +3478,8 @@ int mana_probe(struct gdma_dev *gd, bool resuming) err = -EPROTO; goto out; } + + enable_work(&ac->link_change_work); } if (ac->num_ports == 0) @@ -3500,6 +3542,8 @@ void mana_remove(struct gdma_dev *gd, bool suspending) int err; int i; + disable_work_sync(&ac->link_change_work); + /* adev currently doesn't support suspending, always remove it */ if (gd->adev) remove_adev(gd); diff --git a/include/net/mana/gdma.h b/include/net/mana/gdma.h index 57df78cfbf82..637f42485dba 100644 --- a/include/net/mana/gdma.h +++ b/include/net/mana/gdma.h @@ -590,6 +590,7 @@ enum { /* Driver can self reset on FPGA Reconfig EQE notification */ #define GDMA_DRV_CAP_FLAG_1_HANDLE_RECONFIG_EQE BIT(17) +#define GDMA_DRV_CAP_FLAG_1_HW_VPORT_LINK_AWARE BIT(6) #define GDMA_DRV_CAP_FLAGS1 \ (GDMA_DRV_CAP_FLAG_1_EQ_SHARING_MULTI_VPORT | \ @@ -599,7 +600,8 @@ enum { GDMA_DRV_CAP_FLAG_1_DEV_LIST_HOLES_SUP | \ GDMA_DRV_CAP_FLAG_1_DYNAMIC_IRQ_ALLOC_SUPPORT | \ GDMA_DRV_CAP_FLAG_1_SELF_RESET_ON_EQE | \ - GDMA_DRV_CAP_FLAG_1_HANDLE_RECONFIG_EQE) + GDMA_DRV_CAP_FLAG_1_HANDLE_RECONFIG_EQE | \ + GDMA_DRV_CAP_FLAG_1_HW_VPORT_LINK_AWARE) #define GDMA_DRV_CAP_FLAGS2 0 diff --git a/include/net/mana/hw_channel.h b/include/net/mana/hw_channel.h index 83cf93338eb3..16feb39616c1 100644 --- a/include/net/mana/hw_channel.h +++ b/include/net/mana/hw_channel.h @@ -24,6 +24,8 @@ #define HWC_INIT_DATA_PF_DEST_CQ_ID 11 #define HWC_DATA_CFG_HWC_TIMEOUT 1 +#define HWC_DATA_HW_LINK_CONNECT 2 +#define HWC_DATA_HW_LINK_DISCONNECT 3 #define HW_CHANNEL_WAIT_RESOURCE_TIMEOUT_MS 30000 diff --git a/include/net/mana/mana.h b/include/net/mana/mana.h index 0921485565c0..8906901535f5 100644 --- a/include/net/mana/mana.h +++ b/include/net/mana/mana.h @@ -477,6 +477,10 @@ struct mana_context { struct dentry *mana_eqs_debugfs; struct net_device *ports[MAX_PORTS_IN_MANA_DEV]; + + /* Link state change work */ + struct work_struct link_change_work; + u32 link_event; }; struct mana_port_context { -- cgit v1.2.3 From e21d451a82f39e91b7635c4fc3ff5ac082873ec3 Mon Sep 17 00:00:00 2001 From: Pierre Barre Date: Thu, 16 Oct 2025 15:58:36 +0200 Subject: 9p: Use kvmalloc for message buffers on supported transports While developing a 9P server (https://github.com/Barre/ZeroFS) and testing it under high-load, I was running into allocation failures. The failures occur even with plenty of free memory available because kmalloc requires contiguous physical memory. This results in errors like: ls: page allocation failure: order:7, mode:0x40c40(GFP_NOFS|__GFP_COMP) This patch introduces a transport capability flag (supports_vmalloc) that indicates whether a transport can work with vmalloc'd buffers (non-physically contiguous memory). Transports requiring DMA should leave this flag as false. The fd-based transports (tcp, unix, fd) set this flag to true, and p9_fcall_init will use kvmalloc instead of kmalloc for these transports. This allows the allocator to fall back to vmalloc when contiguous physical memory is not available. Additionally, if kmem_cache_alloc fails, the code falls back to kvmalloc for transports that support it. Signed-off-by: Pierre Barre Reviewed-by: Christian Schoenebeck Message-ID: Signed-off-by: Dominique Martinet --- include/net/9p/transport.h | 4 ++++ net/9p/client.c | 11 +++++++++-- net/9p/trans_fd.c | 3 +++ net/9p/trans_rdma.c | 1 + net/9p/trans_usbg.c | 1 + net/9p/trans_virtio.c | 1 + net/9p/trans_xen.c | 1 + 7 files changed, 20 insertions(+), 2 deletions(-) (limited to 'include/net') diff --git a/include/net/9p/transport.h b/include/net/9p/transport.h index 766ec07c9599..f0981515148d 100644 --- a/include/net/9p/transport.h +++ b/include/net/9p/transport.h @@ -24,6 +24,9 @@ * we're less flexible when choosing the response message * size in this case * @def: set if this transport should be considered the default + * @supports_vmalloc: set if this transport can work with vmalloc'd buffers + * (non-physically contiguous memory). Transports requiring + * DMA should leave this as false. * @create: member function to create a new connection on this transport * @close: member function to discard a connection on this transport * @request: member function to issue a request to the transport @@ -44,6 +47,7 @@ struct p9_trans_module { int maxsize; /* max message size of transport */ bool pooled_rbuffers; int def; /* this transport should be default */ + bool supports_vmalloc; /* can work with vmalloc'd buffers */ struct module *owner; int (*create)(struct p9_client *client, const char *devname, char *args); diff --git a/net/9p/client.c b/net/9p/client.c index 5c1ca57ccd28..2a4884c880c1 100644 --- a/net/9p/client.c +++ b/net/9p/client.c @@ -229,8 +229,15 @@ static int p9_fcall_init(struct p9_client *c, struct p9_fcall *fc, if (likely(c->fcall_cache) && alloc_msize == c->msize) { fc->sdata = kmem_cache_alloc(c->fcall_cache, GFP_NOFS); fc->cache = c->fcall_cache; + if (!fc->sdata && c->trans_mod->supports_vmalloc) { + fc->sdata = kvmalloc(alloc_msize, GFP_NOFS); + fc->cache = NULL; + } } else { - fc->sdata = kmalloc(alloc_msize, GFP_NOFS); + if (c->trans_mod->supports_vmalloc) + fc->sdata = kvmalloc(alloc_msize, GFP_NOFS); + else + fc->sdata = kmalloc(alloc_msize, GFP_NOFS); fc->cache = NULL; } if (!fc->sdata) @@ -252,7 +259,7 @@ void p9_fcall_fini(struct p9_fcall *fc) if (fc->cache) kmem_cache_free(fc->cache, fc->sdata); else - kfree(fc->sdata); + kvfree(fc->sdata); } EXPORT_SYMBOL(p9_fcall_fini); diff --git a/net/9p/trans_fd.c b/net/9p/trans_fd.c index a516745f732f..e7334033eba5 100644 --- a/net/9p/trans_fd.c +++ b/net/9p/trans_fd.c @@ -1101,6 +1101,7 @@ static struct p9_trans_module p9_tcp_trans = { .maxsize = MAX_SOCK_BUF, .pooled_rbuffers = false, .def = 0, + .supports_vmalloc = true, .create = p9_fd_create_tcp, .close = p9_fd_close, .request = p9_fd_request, @@ -1115,6 +1116,7 @@ static struct p9_trans_module p9_unix_trans = { .name = "unix", .maxsize = MAX_SOCK_BUF, .def = 0, + .supports_vmalloc = true, .create = p9_fd_create_unix, .close = p9_fd_close, .request = p9_fd_request, @@ -1129,6 +1131,7 @@ static struct p9_trans_module p9_fd_trans = { .name = "fd", .maxsize = MAX_SOCK_BUF, .def = 0, + .supports_vmalloc = true, .create = p9_fd_create, .close = p9_fd_close, .request = p9_fd_request, diff --git a/net/9p/trans_rdma.c b/net/9p/trans_rdma.c index b84748baf9cb..6c5ad232c194 100644 --- a/net/9p/trans_rdma.c +++ b/net/9p/trans_rdma.c @@ -749,6 +749,7 @@ static struct p9_trans_module p9_rdma_trans = { .maxsize = P9_RDMA_MAXSIZE, .pooled_rbuffers = true, .def = 0, + .supports_vmalloc = false, .owner = THIS_MODULE, .create = rdma_create_trans, .close = rdma_close, diff --git a/net/9p/trans_usbg.c b/net/9p/trans_usbg.c index 468f7e8f0277..2542ef099233 100644 --- a/net/9p/trans_usbg.c +++ b/net/9p/trans_usbg.c @@ -514,6 +514,7 @@ static struct p9_trans_module p9_usbg_trans = { .close = p9_usbg_close, .request = p9_usbg_request, .cancel = p9_usbg_cancel, + .supports_vmalloc = false, .owner = THIS_MODULE, }; diff --git a/net/9p/trans_virtio.c b/net/9p/trans_virtio.c index 0b8086f58ad5..12806207f4f0 100644 --- a/net/9p/trans_virtio.c +++ b/net/9p/trans_virtio.c @@ -803,6 +803,7 @@ static struct p9_trans_module p9_virtio_trans = { .maxsize = PAGE_SIZE * (VIRTQUEUE_NUM - 3), .pooled_rbuffers = false, .def = 1, + .supports_vmalloc = false, .owner = THIS_MODULE, }; diff --git a/net/9p/trans_xen.c b/net/9p/trans_xen.c index b9ff69c7522a..4b1cec0ab829 100644 --- a/net/9p/trans_xen.c +++ b/net/9p/trans_xen.c @@ -258,6 +258,7 @@ static struct p9_trans_module p9_xen_trans = { .maxsize = 1 << (XEN_9PFS_RING_ORDER + XEN_PAGE_SHIFT - 2), .pooled_rbuffers = false, .def = 1, + .supports_vmalloc = false, .create = p9_xen_create, .close = p9_xen_close, .request = p9_xen_request, -- cgit v1.2.3 From eeaf38a798aff6384983e5a0ac464d146de7ff55 Mon Sep 17 00:00:00 2001 From: Dominique Martinet Date: Fri, 31 Oct 2025 16:40:07 +0900 Subject: net/9p: cleanup: change p9_trans_module->def to bool '->def' is only ever used as a true/false flag Reported-by: Christophe JAILLET Message-ID: <20251103-v9fs_trans_def_bool-v1-1-f33dc7ed9e81@codewreck.org> Signed-off-by: Dominique Martinet --- include/net/9p/transport.h | 2 +- net/9p/trans_fd.c | 6 +++--- net/9p/trans_rdma.c | 2 +- net/9p/trans_virtio.c | 2 +- net/9p/trans_xen.c | 2 +- 5 files changed, 7 insertions(+), 7 deletions(-) (limited to 'include/net') diff --git a/include/net/9p/transport.h b/include/net/9p/transport.h index f0981515148d..0aedabc9b7eb 100644 --- a/include/net/9p/transport.h +++ b/include/net/9p/transport.h @@ -46,7 +46,7 @@ struct p9_trans_module { char *name; /* name of transport */ int maxsize; /* max message size of transport */ bool pooled_rbuffers; - int def; /* this transport should be default */ + bool def; /* this transport should be default */ bool supports_vmalloc; /* can work with vmalloc'd buffers */ struct module *owner; int (*create)(struct p9_client *client, diff --git a/net/9p/trans_fd.c b/net/9p/trans_fd.c index e7334033eba5..bd4903d64827 100644 --- a/net/9p/trans_fd.c +++ b/net/9p/trans_fd.c @@ -1100,7 +1100,7 @@ static struct p9_trans_module p9_tcp_trans = { .name = "tcp", .maxsize = MAX_SOCK_BUF, .pooled_rbuffers = false, - .def = 0, + .def = false, .supports_vmalloc = true, .create = p9_fd_create_tcp, .close = p9_fd_close, @@ -1115,7 +1115,7 @@ MODULE_ALIAS_9P("tcp"); static struct p9_trans_module p9_unix_trans = { .name = "unix", .maxsize = MAX_SOCK_BUF, - .def = 0, + .def = false, .supports_vmalloc = true, .create = p9_fd_create_unix, .close = p9_fd_close, @@ -1130,7 +1130,7 @@ MODULE_ALIAS_9P("unix"); static struct p9_trans_module p9_fd_trans = { .name = "fd", .maxsize = MAX_SOCK_BUF, - .def = 0, + .def = false, .supports_vmalloc = true, .create = p9_fd_create, .close = p9_fd_close, diff --git a/net/9p/trans_rdma.c b/net/9p/trans_rdma.c index 6c5ad232c194..a0bc766199da 100644 --- a/net/9p/trans_rdma.c +++ b/net/9p/trans_rdma.c @@ -748,7 +748,7 @@ static struct p9_trans_module p9_rdma_trans = { .name = "rdma", .maxsize = P9_RDMA_MAXSIZE, .pooled_rbuffers = true, - .def = 0, + .def = false, .supports_vmalloc = false, .owner = THIS_MODULE, .create = rdma_create_trans, diff --git a/net/9p/trans_virtio.c b/net/9p/trans_virtio.c index 12806207f4f0..b58f50b00c72 100644 --- a/net/9p/trans_virtio.c +++ b/net/9p/trans_virtio.c @@ -802,7 +802,7 @@ static struct p9_trans_module p9_virtio_trans = { */ .maxsize = PAGE_SIZE * (VIRTQUEUE_NUM - 3), .pooled_rbuffers = false, - .def = 1, + .def = true, .supports_vmalloc = false, .owner = THIS_MODULE, }; diff --git a/net/9p/trans_xen.c b/net/9p/trans_xen.c index 4b1cec0ab829..9389c1247001 100644 --- a/net/9p/trans_xen.c +++ b/net/9p/trans_xen.c @@ -257,7 +257,7 @@ static struct p9_trans_module p9_xen_trans = { .name = "xen", .maxsize = 1 << (XEN_9PFS_RING_ORDER + XEN_PAGE_SHIFT - 2), .pooled_rbuffers = false, - .def = 1, + .def = true, .supports_vmalloc = false, .create = p9_xen_create, .close = p9_xen_close, -- cgit v1.2.3 From c44393d84149d6fc91d94fa39321c9657e91b388 Mon Sep 17 00:00:00 2001 From: Eric Sandeen Date: Fri, 10 Oct 2025 16:36:17 -0500 Subject: net/9p: move structures and macros to header files With the new mount API all option parsing will need to happen in fs/v9fs.c, so move some existing data structures and macros to header files to facilitate this. Rename some to reflect the transport they are used for (rdma, fd, etc), for clarity. Signed-off-by: Eric Sandeen Message-ID: <20251010214222.1347785-3-sandeen@redhat.com> Signed-off-by: Dominique Martinet --- include/net/9p/client.h | 6 ++++++ include/net/9p/transport.h | 39 +++++++++++++++++++++++++++++++++++++++ net/9p/client.c | 6 ------ net/9p/trans_fd.c | 20 ++------------------ net/9p/trans_rdma.c | 25 ++----------------------- 5 files changed, 49 insertions(+), 47 deletions(-) (limited to 'include/net') diff --git a/include/net/9p/client.h b/include/net/9p/client.h index 4f785098c67a..2d46f8017bd5 100644 --- a/include/net/9p/client.h +++ b/include/net/9p/client.h @@ -16,6 +16,12 @@ /* Number of requests per row */ #define P9_ROW_MAXTAG 255 +/* DEFAULT MSIZE = 32 pages worth of payload + P9_HDRSZ + + * room for write (16 extra) or read (11 extra) operands. + */ + +#define DEFAULT_MSIZE ((128 * 1024) + P9_IOHDRSZ) + /** enum p9_proto_versions - 9P protocol versions * @p9_proto_legacy: 9P Legacy mode, pre-9P2000.u * @p9_proto_2000u: 9P2000.u extension diff --git a/include/net/9p/transport.h b/include/net/9p/transport.h index 0aedabc9b7eb..db6ad369a171 100644 --- a/include/net/9p/transport.h +++ b/include/net/9p/transport.h @@ -14,6 +14,45 @@ #define P9_DEF_MIN_RESVPORT (665U) #define P9_DEF_MAX_RESVPORT (1023U) +#define P9_FD_PORT 564 + +#define P9_RDMA_PORT 5640 +#define P9_RDMA_SQ_DEPTH 32 +#define P9_RDMA_RQ_DEPTH 32 +#define P9_RDMA_TIMEOUT 30000 /* 30 seconds */ + +/** + * struct p9_fd_opts - per-transport options for fd transport + * @rfd: file descriptor for reading (trans=fd) + * @wfd: file descriptor for writing (trans=fd) + * @port: port to connect to (trans=tcp) + * @privport: port is privileged + */ + +struct p9_fd_opts { + int rfd; + int wfd; + u16 port; + bool privport; +}; + +/** + * struct p9_rdma_opts - Collection of mount options for rdma transport + * @port: port of connection + * @privport: Whether a privileged port may be used + * @sq_depth: The requested depth of the SQ. This really doesn't need + * to be any deeper than the number of threads used in the client + * @rq_depth: The depth of the RQ. Should be greater than or equal to SQ depth + * @timeout: Time to wait in msecs for CM events + */ +struct p9_rdma_opts { + short port; + bool privport; + int sq_depth; + int rq_depth; + long timeout; +}; + /** * struct p9_trans_module - transport module interface * @list: used to maintain a list of currently available transports diff --git a/net/9p/client.c b/net/9p/client.c index 2a4884c880c1..802f548332a5 100644 --- a/net/9p/client.c +++ b/net/9p/client.c @@ -29,12 +29,6 @@ #define CREATE_TRACE_POINTS #include -/* DEFAULT MSIZE = 32 pages worth of payload + P9_HDRSZ + - * room for write (16 extra) or read (11 extra) operands. - */ - -#define DEFAULT_MSIZE ((128 * 1024) + P9_IOHDRSZ) - /* Client Option Parsing (code inspired by NFS code) * - a little lazy - parse all client options */ diff --git a/net/9p/trans_fd.c b/net/9p/trans_fd.c index bd4903d64827..b7e5933c4617 100644 --- a/net/9p/trans_fd.c +++ b/net/9p/trans_fd.c @@ -31,28 +31,12 @@ #include /* killme */ -#define P9_PORT 564 #define MAX_SOCK_BUF (1024*1024) #define MAXPOLLWADDR 2 static struct p9_trans_module p9_tcp_trans; static struct p9_trans_module p9_fd_trans; -/** - * struct p9_fd_opts - per-transport options - * @rfd: file descriptor for reading (trans=fd) - * @wfd: file descriptor for writing (trans=fd) - * @port: port to connect to (trans=tcp) - * @privport: port is privileged - */ - -struct p9_fd_opts { - int rfd; - int wfd; - u16 port; - bool privport; -}; - /* * Option Parsing (code inspired by NFS code) * - a little lazy - parse all fd-transport options @@ -742,7 +726,7 @@ static int p9_fd_cancelled(struct p9_client *client, struct p9_req_t *req) static int p9_fd_show_options(struct seq_file *m, struct p9_client *clnt) { if (clnt->trans_mod == &p9_tcp_trans) { - if (clnt->trans_opts.tcp.port != P9_PORT) + if (clnt->trans_opts.tcp.port != P9_FD_PORT) seq_printf(m, ",port=%u", clnt->trans_opts.tcp.port); } else if (clnt->trans_mod == &p9_fd_trans) { if (clnt->trans_opts.fd.rfd != ~0) @@ -768,7 +752,7 @@ static int parse_opts(char *params, struct p9_fd_opts *opts) int option; char *options, *tmp_options; - opts->port = P9_PORT; + opts->port = P9_FD_PORT; opts->rfd = ~0; opts->wfd = ~0; opts->privport = false; diff --git a/net/9p/trans_rdma.c b/net/9p/trans_rdma.c index a0bc766199da..87246463a954 100644 --- a/net/9p/trans_rdma.c +++ b/net/9p/trans_rdma.c @@ -32,14 +32,10 @@ #include #include -#define P9_PORT 5640 -#define P9_RDMA_SQ_DEPTH 32 -#define P9_RDMA_RQ_DEPTH 32 #define P9_RDMA_SEND_SGE 4 #define P9_RDMA_RECV_SGE 4 #define P9_RDMA_IRD 0 #define P9_RDMA_ORD 0 -#define P9_RDMA_TIMEOUT 30000 /* 30 seconds */ #define P9_RDMA_MAXSIZE (1024*1024) /* 1MB */ /** @@ -110,23 +106,6 @@ struct p9_rdma_context { }; }; -/** - * struct p9_rdma_opts - Collection of mount options - * @port: port of connection - * @privport: Whether a privileged port may be used - * @sq_depth: The requested depth of the SQ. This really doesn't need - * to be any deeper than the number of threads used in the client - * @rq_depth: The depth of the RQ. Should be greater than or equal to SQ depth - * @timeout: Time to wait in msecs for CM events - */ -struct p9_rdma_opts { - short port; - bool privport; - int sq_depth; - int rq_depth; - long timeout; -}; - /* * Option Parsing (code inspired by NFS code) */ @@ -151,7 +130,7 @@ static int p9_rdma_show_options(struct seq_file *m, struct p9_client *clnt) { struct p9_trans_rdma *rdma = clnt->trans; - if (rdma->port != P9_PORT) + if (rdma->port != P9_RDMA_PORT) seq_printf(m, ",port=%u", rdma->port); if (rdma->sq_depth != P9_RDMA_SQ_DEPTH) seq_printf(m, ",sq=%u", rdma->sq_depth); @@ -178,7 +157,7 @@ static int parse_opts(char *params, struct p9_rdma_opts *opts) int option; char *options, *tmp_options; - opts->port = P9_PORT; + opts->port = P9_RDMA_PORT; opts->sq_depth = P9_RDMA_SQ_DEPTH; opts->rq_depth = P9_RDMA_RQ_DEPTH; opts->timeout = P9_RDMA_TIMEOUT; -- cgit v1.2.3 From 075e8bd4127f007910fc302ad5c3c471d0be4799 Mon Sep 17 00:00:00 2001 From: Eric Sandeen Date: Fri, 10 Oct 2025 16:36:18 -0500 Subject: 9p: create a v9fs_context structure to hold parsed options This patch creates a new v9fs_context structure which includes new p9_session_opts and p9_client_opts structures, as well as re-using the existing p9_fd_opts and p9_rdma_opts to store options during parsing. The new structure will be used in the next commit to pass all parsed options to the appropriate transports. Signed-off-by: Eric Sandeen Message-ID: <20251010214222.1347785-4-sandeen@redhat.com> Signed-off-by: Dominique Martinet --- include/net/9p/client.h | 90 ++++++++++++++++++++++++++++++++++++++++++++++ include/net/9p/transport.h | 32 ----------------- 2 files changed, 90 insertions(+), 32 deletions(-) (limited to 'include/net') diff --git a/include/net/9p/client.h b/include/net/9p/client.h index 2d46f8017bd5..cc18443f7d51 100644 --- a/include/net/9p/client.h +++ b/include/net/9p/client.h @@ -132,6 +132,96 @@ struct p9_client { char name[__NEW_UTS_LEN + 1]; }; +/** + * struct p9_fd_opts - holds client options during parsing + * @msize: maximum data size negotiated by protocol + * @prot-Oversion: 9P protocol version to use + * @trans_mod: module API instantiated with this client + * + * These parsed options get transferred into client in + * apply_client_options() + */ +struct p9_client_opts { + unsigned int msize; + unsigned char proto_version; + struct p9_trans_module *trans_mod; +}; + +/** + * struct p9_fd_opts - per-transport options for fd transport + * @rfd: file descriptor for reading (trans=fd) + * @wfd: file descriptor for writing (trans=fd) + * @port: port to connect to (trans=tcp) + * @privport: port is privileged + */ +struct p9_fd_opts { + int rfd; + int wfd; + u16 port; + bool privport; +}; + +/** + * struct p9_rdma_opts - Collection of mount options for rdma transport + * @port: port of connection + * @privport: Whether a privileged port may be used + * @sq_depth: The requested depth of the SQ. This really doesn't need + * to be any deeper than the number of threads used in the client + * @rq_depth: The depth of the RQ. Should be greater than or equal to SQ depth + * @timeout: Time to wait in msecs for CM events + */ +struct p9_rdma_opts { + short port; + bool privport; + int sq_depth; + int rq_depth; + long timeout; +}; + +/** + * struct p9_session_opts - holds parsed options for v9fs_session_info + * @flags: session options of type &p9_session_flags + * @nodev: set to 1 to disable device mapping + * @debug: debug level + * @afid: authentication handle + * @cache: cache mode of type &p9_cache_bits + * @cachetag: the tag of the cache associated with this session + * @uname: string user name to mount hierarchy as + * @aname: mount specifier for remote hierarchy + * @dfltuid: default numeric userid to mount hierarchy as + * @dfltgid: default numeric groupid to mount hierarchy as + * @uid: if %V9FS_ACCESS_SINGLE, the numeric uid which mounted the hierarchy + * @session_lock_timeout: retry interval for blocking locks + * + * This strucure holds options which are parsed and will be transferred + * to the v9fs_session_info structure when mounted, and therefore largely + * duplicates struct v9fs_session_info. + */ +struct p9_session_opts { + unsigned int flags; + unsigned char nodev; + unsigned short debug; + unsigned int afid; + unsigned int cache; +#ifdef CONFIG_9P_FSCACHE + char *cachetag; +#endif + char *uname; + char *aname; + kuid_t dfltuid; + kgid_t dfltgid; + kuid_t uid; + long session_lock_timeout; +}; + +/* Used by mount API to store parsed mount options */ +struct v9fs_context { + struct p9_client_opts client_opts; + struct p9_fd_opts fd_opts; + struct p9_rdma_opts rdma_opts; + struct p9_session_opts session_opts; +}; + /** * struct p9_fid - file system entity handle * @clnt: back pointer to instantiating &p9_client diff --git a/include/net/9p/transport.h b/include/net/9p/transport.h index db6ad369a171..898a432a8063 100644 --- a/include/net/9p/transport.h +++ b/include/net/9p/transport.h @@ -21,38 +21,6 @@ #define P9_RDMA_RQ_DEPTH 32 #define P9_RDMA_TIMEOUT 30000 /* 30 seconds */ -/** - * struct p9_fd_opts - per-transport options for fd transport - * @rfd: file descriptor for reading (trans=fd) - * @wfd: file descriptor for writing (trans=fd) - * @port: port to connect to (trans=tcp) - * @privport: port is privileged - */ - -struct p9_fd_opts { - int rfd; - int wfd; - u16 port; - bool privport; -}; - -/** - * struct p9_rdma_opts - Collection of mount options for rdma transport - * @port: port of connection - * @privport: Whether a privileged port may be used - * @sq_depth: The requested depth of the SQ. This really doesn't need - * to be any deeper than the number of threads used in the client - * @rq_depth: The depth of the RQ. Should be greater than or equal to SQ depth - * @timeout: Time to wait in msecs for CM events - */ -struct p9_rdma_opts { - short port; - bool privport; - int sq_depth; - int rq_depth; - long timeout; -}; - /** * struct p9_trans_module - transport module interface * @list: used to maintain a list of currently available transports -- cgit v1.2.3 From 1f3e4142c0eb178089ea0cbc97506a061470ad27 Mon Sep 17 00:00:00 2001 From: Eric Sandeen Date: Fri, 10 Oct 2025 16:36:19 -0500 Subject: 9p: convert to the new mount API Convert 9p to the new mount API. This patch consolidates all parsing into fs/9p/v9fs.c, which stores all results into a filesystem context which can be passed to the various transports as needed. Some of the parsing helper functions such as get_cache_mode() have been eliminated in favor of using the new mount API's enum param type, for simplicity. Signed-off-by: Eric Sandeen Message-ID: <20251010214222.1347785-5-sandeen@redhat.com> [ Dominique: handled source explicitly as per follow-up discussion ] Signed-off-by: Dominique Martinet --- fs/9p/v9fs.c | 552 +++++++++++++++++++++++---------------------- fs/9p/v9fs.h | 7 +- fs/9p/vfs_super.c | 130 +++++++---- include/net/9p/client.h | 2 +- include/net/9p/transport.h | 2 +- net/9p/client.c | 148 +----------- net/9p/mod.c | 2 +- net/9p/trans_fd.c | 109 +-------- net/9p/trans_rdma.c | 108 +-------- net/9p/trans_usbg.c | 4 +- net/9p/trans_virtio.c | 8 +- net/9p/trans_xen.c | 4 +- 12 files changed, 424 insertions(+), 652 deletions(-) (limited to 'include/net') diff --git a/fs/9p/v9fs.c b/fs/9p/v9fs.c index a020a8f00a1a..05fc2ba3c5d4 100644 --- a/fs/9p/v9fs.c +++ b/fs/9p/v9fs.c @@ -13,7 +13,8 @@ #include #include #include -#include +#include +#include #include #include #include @@ -33,6 +34,10 @@ struct kmem_cache *v9fs_inode_cache; */ enum { + /* Mount-point source, we need to handle this explicitly because + * the code below accepts unknown args and the vfs layer only handles + * source if we rejected it as EINVAL */ + Opt_source, /* Options that take integer arguments */ Opt_debug, Opt_dfltuid, Opt_dfltgid, Opt_afid, /* String options */ @@ -43,55 +48,81 @@ enum { Opt_access, Opt_posixacl, /* Lock timeout option */ Opt_locktimeout, - /* Error token */ - Opt_err + + /* Client options */ + Opt_msize, Opt_trans, Opt_legacy, Opt_version, + + /* fd transport options */ + /* Options that take integer arguments */ + Opt_rfdno, Opt_wfdno, + /* Options that take no arguments */ + + /* rdma transport options */ + /* Options that take integer arguments */ + Opt_rq_depth, Opt_sq_depth, Opt_timeout, + + /* Options for both fd and rdma transports */ + Opt_port, Opt_privport, }; -static const match_table_t tokens = { - {Opt_debug, "debug=%x"}, - {Opt_dfltuid, "dfltuid=%u"}, - {Opt_dfltgid, "dfltgid=%u"}, - {Opt_afid, "afid=%u"}, - {Opt_uname, "uname=%s"}, - {Opt_remotename, "aname=%s"}, - {Opt_nodevmap, "nodevmap"}, - {Opt_noxattr, "noxattr"}, - {Opt_directio, "directio"}, - {Opt_ignoreqv, "ignoreqv"}, - {Opt_cache, "cache=%s"}, - {Opt_cachetag, "cachetag=%s"}, - {Opt_access, "access=%s"}, - {Opt_posixacl, "posixacl"}, - {Opt_locktimeout, "locktimeout=%u"}, - {Opt_err, NULL} +static const struct constant_table p9_versions[] = { + { "9p2000", p9_proto_legacy }, + { "9p2000.u", p9_proto_2000u }, + { "9p2000.L", p9_proto_2000L }, + {} }; -/* Interpret mount options for cache mode */ -static int get_cache_mode(char *s) -{ - int version = -EINVAL; - - if (!strcmp(s, "loose")) { - version = CACHE_SC_LOOSE; - p9_debug(P9_DEBUG_9P, "Cache mode: loose\n"); - } else if (!strcmp(s, "fscache")) { - version = CACHE_SC_FSCACHE; - p9_debug(P9_DEBUG_9P, "Cache mode: fscache\n"); - } else if (!strcmp(s, "mmap")) { - version = CACHE_SC_MMAP; - p9_debug(P9_DEBUG_9P, "Cache mode: mmap\n"); - } else if (!strcmp(s, "readahead")) { - version = CACHE_SC_READAHEAD; - p9_debug(P9_DEBUG_9P, "Cache mode: readahead\n"); - } else if (!strcmp(s, "none")) { - version = CACHE_SC_NONE; - p9_debug(P9_DEBUG_9P, "Cache mode: none\n"); - } else if (kstrtoint(s, 0, &version) != 0) { - version = -EINVAL; - pr_info("Unknown Cache mode or invalid value %s\n", s); - } - return version; -} +static const struct constant_table p9_cache_mode[] = { + { "loose", CACHE_SC_LOOSE }, + { "fscache", CACHE_SC_FSCACHE }, + { "mmap", CACHE_SC_MMAP }, + { "readahead", CACHE_SC_READAHEAD }, + { "none", CACHE_SC_NONE }, + {} +}; + +/* + * This structure contains all parameters used for the core code, + * the client, and all the transports. + */ +const struct fs_parameter_spec v9fs_param_spec[] = { + fsparam_string ("source", Opt_source), + fsparam_u32hex ("debug", Opt_debug), + fsparam_uid ("dfltuid", Opt_dfltuid), + fsparam_gid ("dfltgid", Opt_dfltgid), + fsparam_u32 ("afid", Opt_afid), + fsparam_string ("uname", Opt_uname), + fsparam_string ("aname", Opt_remotename), + fsparam_flag ("nodevmap", Opt_nodevmap), + fsparam_flag ("noxattr", Opt_noxattr), + fsparam_flag ("directio", Opt_directio), + fsparam_flag ("ignoreqv", Opt_ignoreqv), + fsparam_enum ("cache", Opt_cache, p9_cache_mode), + fsparam_string ("cachetag", Opt_cachetag), + fsparam_string ("access", Opt_access), + fsparam_flag ("posixacl", Opt_posixacl), + fsparam_u32 ("locktimeout", Opt_locktimeout), + + /* client options */ + fsparam_u32 ("msize", Opt_msize), + fsparam_flag ("noextend", Opt_legacy), + fsparam_string ("trans", Opt_trans), + fsparam_enum ("version", Opt_version, p9_versions), + + /* fd transport options */ + fsparam_u32 ("rfdno", Opt_rfdno), + fsparam_u32 ("wfdno", Opt_wfdno), + + /* rdma transport options */ + fsparam_u32 ("sq", Opt_sq_depth), + fsparam_u32 ("rq", Opt_rq_depth), + fsparam_u32 ("timeout", Opt_timeout), + + /* fd and rdma transprt options */ + fsparam_u32 ("port", Opt_port), + fsparam_flag ("privport", Opt_privport), + {} +}; /* * Display the mount options in /proc/mounts. @@ -153,267 +184,252 @@ int v9fs_show_options(struct seq_file *m, struct dentry *root) } /** - * v9fs_parse_options - parse mount options into session structure - * @v9ses: existing v9fs session information - * @opts: The mount option string + * v9fs_parse_param - parse a mount option into the filesystem context + * @fc: the filesystem context + * @param: the parameter to parse * * Return 0 upon success, -ERRNO upon failure. */ - -static int v9fs_parse_options(struct v9fs_session_info *v9ses, char *opts) +int v9fs_parse_param(struct fs_context *fc, struct fs_parameter *param) { - char *options, *tmp_options; - substring_t args[MAX_OPT_ARGS]; - char *p; - int option = 0; + struct v9fs_context *ctx = fc->fs_private; + struct fs_parse_result result; char *s; - int ret = 0; - - /* setup defaults */ - v9ses->afid = ~0; - v9ses->debug = 0; - v9ses->cache = CACHE_NONE; -#ifdef CONFIG_9P_FSCACHE - v9ses->cachetag = NULL; -#endif - v9ses->session_lock_timeout = P9_LOCK_TIMEOUT; - - if (!opts) - return 0; + int r; + int opt; + struct p9_client_opts *clnt = &ctx->client_opts; + struct p9_fd_opts *fd_opts = &ctx->fd_opts; + struct p9_rdma_opts *rdma_opts = &ctx->rdma_opts; + struct p9_session_opts *session_opts = &ctx->session_opts; + + opt = fs_parse(fc, v9fs_param_spec, param, &result); + if (opt < 0) { + /* + * We might like to report bad mount options here, but + * traditionally 9p has ignored unknown mount options + */ + if (opt == -ENOPARAM) + return 0; - tmp_options = kstrdup(opts, GFP_KERNEL); - if (!tmp_options) { - ret = -ENOMEM; - goto fail_option_alloc; + return opt; } - options = tmp_options; - - while ((p = strsep(&options, ",")) != NULL) { - int token, r; - - if (!*p) - continue; - - token = match_token(p, tokens, args); - switch (token) { - case Opt_debug: - r = match_int(&args[0], &option); - if (r < 0) { - p9_debug(P9_DEBUG_ERROR, - "integer field, but no integer?\n"); - ret = r; - } else { - v9ses->debug = option; + + switch (opt) { + case Opt_source: + if (fc->source) { + pr_info("p9: multiple sources not supported\n"); + return -EINVAL; + } + fc->source = param->string; + param->string = NULL; + break; + case Opt_debug: + session_opts->debug = result.uint_32; #ifdef CONFIG_NET_9P_DEBUG - p9_debug_level = option; + p9_debug_level = result.uint_32; #endif - } - break; - - case Opt_dfltuid: - r = match_int(&args[0], &option); - if (r < 0) { - p9_debug(P9_DEBUG_ERROR, - "integer field, but no integer?\n"); - ret = r; - continue; - } - v9ses->dfltuid = make_kuid(current_user_ns(), option); - if (!uid_valid(v9ses->dfltuid)) { - p9_debug(P9_DEBUG_ERROR, - "uid field, but not a uid?\n"); - ret = -EINVAL; - } - break; - case Opt_dfltgid: - r = match_int(&args[0], &option); - if (r < 0) { - p9_debug(P9_DEBUG_ERROR, - "integer field, but no integer?\n"); - ret = r; - continue; - } - v9ses->dfltgid = make_kgid(current_user_ns(), option); - if (!gid_valid(v9ses->dfltgid)) { - p9_debug(P9_DEBUG_ERROR, - "gid field, but not a gid?\n"); - ret = -EINVAL; - } - break; - case Opt_afid: - r = match_int(&args[0], &option); - if (r < 0) { - p9_debug(P9_DEBUG_ERROR, - "integer field, but no integer?\n"); - ret = r; - } else { - v9ses->afid = option; - } - break; - case Opt_uname: - kfree(v9ses->uname); - v9ses->uname = match_strdup(&args[0]); - if (!v9ses->uname) { - ret = -ENOMEM; - goto free_and_return; - } - break; - case Opt_remotename: - kfree(v9ses->aname); - v9ses->aname = match_strdup(&args[0]); - if (!v9ses->aname) { - ret = -ENOMEM; - goto free_and_return; - } - break; - case Opt_nodevmap: - v9ses->nodev = 1; - break; - case Opt_noxattr: - v9ses->flags |= V9FS_NO_XATTR; - break; - case Opt_directio: - v9ses->flags |= V9FS_DIRECT_IO; - break; - case Opt_ignoreqv: - v9ses->flags |= V9FS_IGNORE_QV; - break; - case Opt_cachetag: + break; + + case Opt_dfltuid: + session_opts->dfltuid = result.uid; + break; + case Opt_dfltgid: + session_opts->dfltgid = result.gid; + break; + case Opt_afid: + session_opts->afid = result.uint_32; + break; + case Opt_uname: + kfree(session_opts->uname); + session_opts->uname = param->string; + param->string = NULL; + break; + case Opt_remotename: + kfree(session_opts->aname); + session_opts->aname = param->string; + param->string = NULL; + break; + case Opt_nodevmap: + session_opts->nodev = 1; + break; + case Opt_noxattr: + session_opts->flags |= V9FS_NO_XATTR; + break; + case Opt_directio: + session_opts->flags |= V9FS_DIRECT_IO; + break; + case Opt_ignoreqv: + session_opts->flags |= V9FS_IGNORE_QV; + break; + case Opt_cachetag: #ifdef CONFIG_9P_FSCACHE - kfree(v9ses->cachetag); - v9ses->cachetag = match_strdup(&args[0]); - if (!v9ses->cachetag) { - ret = -ENOMEM; - goto free_and_return; - } + kfree(session_opts->cachetag); + session_opts->cachetag = param->string; + param->string = NULL; #endif - break; - case Opt_cache: - s = match_strdup(&args[0]); - if (!s) { - ret = -ENOMEM; - p9_debug(P9_DEBUG_ERROR, - "problem allocating copy of cache arg\n"); - goto free_and_return; - } - r = get_cache_mode(s); - if (r < 0) - ret = r; - else - v9ses->cache = r; - - kfree(s); - break; - - case Opt_access: - s = match_strdup(&args[0]); - if (!s) { - ret = -ENOMEM; - p9_debug(P9_DEBUG_ERROR, - "problem allocating copy of access arg\n"); - goto free_and_return; + break; + case Opt_cache: + session_opts->cache = result.uint_32; + p9_debug(P9_DEBUG_9P, "Cache mode: %s\n", param->string); + break; + case Opt_access: + s = param->string; + session_opts->flags &= ~V9FS_ACCESS_MASK; + if (strcmp(s, "user") == 0) { + session_opts->flags |= V9FS_ACCESS_USER; + } else if (strcmp(s, "any") == 0) { + session_opts->flags |= V9FS_ACCESS_ANY; + } else if (strcmp(s, "client") == 0) { + session_opts->flags |= V9FS_ACCESS_CLIENT; + } else { + uid_t uid; + + session_opts->flags |= V9FS_ACCESS_SINGLE; + r = kstrtouint(s, 10, &uid); + if (r) { + pr_info("Unknown access argument %s: %d\n", + param->string, r); + return r; } - - v9ses->flags &= ~V9FS_ACCESS_MASK; - if (strcmp(s, "user") == 0) - v9ses->flags |= V9FS_ACCESS_USER; - else if (strcmp(s, "any") == 0) - v9ses->flags |= V9FS_ACCESS_ANY; - else if (strcmp(s, "client") == 0) { - v9ses->flags |= V9FS_ACCESS_CLIENT; - } else { - uid_t uid; - - v9ses->flags |= V9FS_ACCESS_SINGLE; - r = kstrtouint(s, 10, &uid); - if (r) { - ret = r; - pr_info("Unknown access argument %s: %d\n", - s, r); - kfree(s); - continue; - } - v9ses->uid = make_kuid(current_user_ns(), uid); - if (!uid_valid(v9ses->uid)) { - ret = -EINVAL; - pr_info("Unknown uid %s\n", s); - } + session_opts->uid = make_kuid(current_user_ns(), uid); + if (!uid_valid(session_opts->uid)) { + pr_info("Unknown uid %s\n", s); + return -EINVAL; } + } + break; - kfree(s); - break; - - case Opt_posixacl: + case Opt_posixacl: #ifdef CONFIG_9P_FS_POSIX_ACL - v9ses->flags |= V9FS_POSIX_ACL; + session_opts->flags |= V9FS_POSIX_ACL; #else - p9_debug(P9_DEBUG_ERROR, - "Not defined CONFIG_9P_FS_POSIX_ACL. Ignoring posixacl option\n"); + p9_debug(P9_DEBUG_ERROR, + "Not defined CONFIG_9P_FS_POSIX_ACL. Ignoring posixacl option\n"); #endif - break; - - case Opt_locktimeout: - r = match_int(&args[0], &option); - if (r < 0) { - p9_debug(P9_DEBUG_ERROR, - "integer field, but no integer?\n"); - ret = r; - continue; - } - if (option < 1) { - p9_debug(P9_DEBUG_ERROR, - "locktimeout must be a greater than zero integer.\n"); - ret = -EINVAL; - continue; - } - v9ses->session_lock_timeout = (long)option * HZ; - break; + break; + + case Opt_locktimeout: + if (result.uint_32 < 1) { + p9_debug(P9_DEBUG_ERROR, + "locktimeout must be a greater than zero integer.\n"); + return -EINVAL; + } + session_opts->session_lock_timeout = (long)result.uint_32 * HZ; + break; - default: - continue; + /* Options for client */ + case Opt_msize: + if (result.uint_32 < 4096) { + p9_debug(P9_DEBUG_ERROR, "msize should be at least 4k\n"); + return -EINVAL; + } + if (result.uint_32 > INT_MAX) { + p9_debug(P9_DEBUG_ERROR, "msize too big\n"); + return -EINVAL; } + clnt->msize = result.uint_32; + break; + case Opt_trans: + v9fs_put_trans(clnt->trans_mod); + clnt->trans_mod = v9fs_get_trans_by_name(param->string); + if (!clnt->trans_mod) { + pr_info("Could not find request transport: %s\n", + param->string); + return -EINVAL; + } + break; + case Opt_legacy: + clnt->proto_version = p9_proto_legacy; + break; + case Opt_version: + clnt->proto_version = result.uint_32; + p9_debug(P9_DEBUG_9P, "Protocol version: %s\n", param->string); + break; + /* Options for fd transport */ + case Opt_rfdno: + fd_opts->rfd = result.uint_32; + break; + case Opt_wfdno: + fd_opts->wfd = result.uint_32; + break; + /* Options for rdma transport */ + case Opt_sq_depth: + rdma_opts->sq_depth = result.uint_32; + break; + case Opt_rq_depth: + rdma_opts->rq_depth = result.uint_32; + break; + case Opt_timeout: + rdma_opts->timeout = result.uint_32; + break; + /* Options for both fd and rdma transports */ + case Opt_port: + fd_opts->port = result.uint_32; + rdma_opts->port = result.uint_32; + break; + case Opt_privport: + fd_opts->privport = true; + rdma_opts->port = true; + break; } -free_and_return: - kfree(tmp_options); -fail_option_alloc: - return ret; + return 0; +} + +static void v9fs_apply_options(struct v9fs_session_info *v9ses, + struct fs_context *fc) +{ + struct v9fs_context *ctx = fc->fs_private; + + v9ses->debug = ctx->session_opts.debug; + v9ses->dfltuid = ctx->session_opts.dfltuid; + v9ses->dfltgid = ctx->session_opts.dfltgid; + v9ses->afid = ctx->session_opts.afid; + v9ses->uname = ctx->session_opts.uname; + ctx->session_opts.uname = NULL; + v9ses->aname = ctx->session_opts.aname; + ctx->session_opts.aname = NULL; + v9ses->nodev = ctx->session_opts.nodev; + /* + * Note that we must |= flags here as session_init already + * set basic flags. This adds in flags from parsed options. + */ + v9ses->flags |= ctx->session_opts.flags; +#ifdef CONFIG_9P_FSCACHE + v9ses->cachetag = ctx->session_opts.cachetag; + ctx->session_opts.cachetag = NULL; +#endif + v9ses->cache = ctx->session_opts.cache; + v9ses->uid = ctx->session_opts.uid; + v9ses->session_lock_timeout = ctx->session_opts.session_lock_timeout; } /** * v9fs_session_init - initialize session * @v9ses: session information structure - * @dev_name: device being mounted - * @data: options + * @fc: the filesystem mount context * */ struct p9_fid *v9fs_session_init(struct v9fs_session_info *v9ses, - const char *dev_name, char *data) + struct fs_context *fc) { struct p9_fid *fid; int rc = -ENOMEM; - v9ses->uname = kstrdup(V9FS_DEFUSER, GFP_KERNEL); - if (!v9ses->uname) - goto err_names; - - v9ses->aname = kstrdup(V9FS_DEFANAME, GFP_KERNEL); - if (!v9ses->aname) - goto err_names; init_rwsem(&v9ses->rename_sem); - v9ses->uid = INVALID_UID; - v9ses->dfltuid = V9FS_DEFUID; - v9ses->dfltgid = V9FS_DEFGID; - - v9ses->clnt = p9_client_create(dev_name, data); + v9ses->clnt = p9_client_create(fc); if (IS_ERR(v9ses->clnt)) { rc = PTR_ERR(v9ses->clnt); p9_debug(P9_DEBUG_ERROR, "problem initializing 9p client\n"); goto err_names; } + /* + * Initialize flags on the real v9ses. v9fs_apply_options below + * will |= the additional flags from parsed options. + */ v9ses->flags = V9FS_ACCESS_USER; if (p9_is_proto_dotl(v9ses->clnt)) { @@ -423,9 +439,7 @@ struct p9_fid *v9fs_session_init(struct v9fs_session_info *v9ses, v9ses->flags |= V9FS_PROTO_2000U; } - rc = v9fs_parse_options(v9ses, data); - if (rc < 0) - goto err_clnt; + v9fs_apply_options(v9ses, fc); v9ses->maxdata = v9ses->clnt->msize - P9_IOHDRSZ; @@ -471,7 +485,7 @@ struct p9_fid *v9fs_session_init(struct v9fs_session_info *v9ses, #ifdef CONFIG_9P_FSCACHE /* register the session for caching */ if (v9ses->cache & CACHE_FSCACHE) { - rc = v9fs_cache_session_get_cookie(v9ses, dev_name); + rc = v9fs_cache_session_get_cookie(v9ses, fc->source); if (rc < 0) goto err_clnt; } diff --git a/fs/9p/v9fs.h b/fs/9p/v9fs.h index f28bc763847a..6a12445d3858 100644 --- a/fs/9p/v9fs.h +++ b/fs/9p/v9fs.h @@ -10,6 +10,9 @@ #include #include +#include +#include +#include /** * enum p9_session_flags - option flags for each 9P session @@ -163,11 +166,13 @@ static inline struct fscache_volume *v9fs_session_cache(struct v9fs_session_info #endif } +extern const struct fs_parameter_spec v9fs_param_spec[]; +extern int v9fs_parse_param(struct fs_context *fc, struct fs_parameter *param); extern int v9fs_show_options(struct seq_file *m, struct dentry *root); struct p9_fid *v9fs_session_init(struct v9fs_session_info *v9ses, - const char *dev_name, char *data); + struct fs_context *fc); extern void v9fs_session_close(struct v9fs_session_info *v9ses); extern void v9fs_session_cancel(struct v9fs_session_info *v9ses); extern void v9fs_session_begin_cancel(struct v9fs_session_info *v9ses); diff --git a/fs/9p/vfs_super.c b/fs/9p/vfs_super.c index 1581ebac5bb4..315336de6f02 100644 --- a/fs/9p/vfs_super.c +++ b/fs/9p/vfs_super.c @@ -19,6 +19,7 @@ #include #include #include +#include #include #include @@ -30,32 +31,10 @@ static const struct super_operations v9fs_super_ops, v9fs_super_ops_dotl; -/** - * v9fs_set_super - set the superblock - * @s: super block - * @data: file system specific data - * - */ - -static int v9fs_set_super(struct super_block *s, void *data) -{ - s->s_fs_info = data; - return set_anon_super(s, data); -} - -/** - * v9fs_fill_super - populate superblock with info - * @sb: superblock - * @v9ses: session information - * @flags: flags propagated from v9fs_mount() - * - */ - -static int -v9fs_fill_super(struct super_block *sb, struct v9fs_session_info *v9ses, - int flags) +static int v9fs_fill_super(struct super_block *sb) { int ret; + struct v9fs_session_info *v9ses = v9ses = sb->s_fs_info; sb->s_maxbytes = MAX_LFS_FILESIZE; sb->s_blocksize_bits = fls(v9ses->maxdata - 1); @@ -95,16 +74,12 @@ v9fs_fill_super(struct super_block *sb, struct v9fs_session_info *v9ses, } /** - * v9fs_mount - mount a superblock - * @fs_type: file system type - * @flags: mount flags - * @dev_name: device name that was mounted - * @data: mount options + * v9fs_get_tree - create the mountable root and superblock + * @fc: the filesystem context * */ -static struct dentry *v9fs_mount(struct file_system_type *fs_type, int flags, - const char *dev_name, void *data) +static int v9fs_get_tree(struct fs_context *fc) { struct super_block *sb = NULL; struct inode *inode = NULL; @@ -117,20 +92,21 @@ static struct dentry *v9fs_mount(struct file_system_type *fs_type, int flags, v9ses = kzalloc(sizeof(struct v9fs_session_info), GFP_KERNEL); if (!v9ses) - return ERR_PTR(-ENOMEM); + return -ENOMEM; - fid = v9fs_session_init(v9ses, dev_name, data); + fid = v9fs_session_init(v9ses, fc); if (IS_ERR(fid)) { retval = PTR_ERR(fid); goto free_session; } - sb = sget(fs_type, NULL, v9fs_set_super, flags, v9ses); + fc->s_fs_info = v9ses; + sb = sget_fc(fc, NULL, set_anon_super_fc); if (IS_ERR(sb)) { retval = PTR_ERR(sb); goto clunk_fid; } - retval = v9fs_fill_super(sb, v9ses, flags); + retval = v9fs_fill_super(sb); if (retval) goto release_sb; @@ -159,14 +135,15 @@ static struct dentry *v9fs_mount(struct file_system_type *fs_type, int flags, v9fs_fid_add(root, &fid); p9_debug(P9_DEBUG_VFS, " simple set mount, return 0\n"); - return dget(sb->s_root); + fc->root = dget(sb->s_root); + return 0; clunk_fid: p9_fid_put(fid); v9fs_session_close(v9ses); free_session: kfree(v9ses); - return ERR_PTR(retval); + return retval; release_sb: /* @@ -177,7 +154,7 @@ release_sb: */ p9_fid_put(fid); deactivate_locked_super(sb); - return ERR_PTR(retval); + return retval; } /** @@ -303,11 +280,86 @@ static const struct super_operations v9fs_super_ops_dotl = { .write_inode = v9fs_write_inode_dotl, }; +static void v9fs_free_fc(struct fs_context *fc) +{ + struct v9fs_context *ctx = fc->fs_private; + + if (!ctx) + return; + + /* These should be NULL by now but guard against leaks */ + kfree(ctx->session_opts.uname); + kfree(ctx->session_opts.aname); +#ifdef CONFIG_9P_FSCACHE + kfree(ctx->session_opts.cachetag); +#endif + if (ctx->client_opts.trans_mod) + v9fs_put_trans(ctx->client_opts.trans_mod); + kfree(ctx); +} + +static const struct fs_context_operations v9fs_context_ops = { + .parse_param = v9fs_parse_param, + .get_tree = v9fs_get_tree, + .free = v9fs_free_fc, +}; + +static int v9fs_init_fs_context(struct fs_context *fc) +{ + struct v9fs_context *ctx; + + ctx = kzalloc(sizeof(*ctx), GFP_KERNEL); + if (!ctx) + return -ENOMEM; + + /* initialize core options */ + ctx->session_opts.afid = ~0; + ctx->session_opts.cache = CACHE_NONE; + ctx->session_opts.session_lock_timeout = P9_LOCK_TIMEOUT; + ctx->session_opts.uname = kstrdup(V9FS_DEFUSER, GFP_KERNEL); + if (!ctx->session_opts.uname) + goto error; + + ctx->session_opts.aname = kstrdup(V9FS_DEFANAME, GFP_KERNEL); + if (!ctx->session_opts.aname) + goto error; + + ctx->session_opts.uid = INVALID_UID; + ctx->session_opts.dfltuid = V9FS_DEFUID; + ctx->session_opts.dfltgid = V9FS_DEFGID; + + /* initialize client options */ + ctx->client_opts.proto_version = p9_proto_2000L; + ctx->client_opts.msize = DEFAULT_MSIZE; + + /* initialize fd transport options */ + ctx->fd_opts.port = P9_FD_PORT; + ctx->fd_opts.rfd = ~0; + ctx->fd_opts.wfd = ~0; + ctx->fd_opts.privport = false; + + /* initialize rdma transport options */ + ctx->rdma_opts.port = P9_RDMA_PORT; + ctx->rdma_opts.sq_depth = P9_RDMA_SQ_DEPTH; + ctx->rdma_opts.rq_depth = P9_RDMA_RQ_DEPTH; + ctx->rdma_opts.timeout = P9_RDMA_TIMEOUT; + ctx->rdma_opts.privport = false; + + fc->ops = &v9fs_context_ops; + fc->fs_private = ctx; + + return 0; +error: + fc->need_free = 1; + return -ENOMEM; +} + struct file_system_type v9fs_fs_type = { .name = "9p", - .mount = v9fs_mount, .kill_sb = v9fs_kill_super, .owner = THIS_MODULE, .fs_flags = FS_RENAME_DOES_D_MOVE, + .init_fs_context = v9fs_init_fs_context, + .parameters = v9fs_param_spec, }; MODULE_ALIAS_FS("9p"); diff --git a/include/net/9p/client.h b/include/net/9p/client.h index cc18443f7d51..838a94218b59 100644 --- a/include/net/9p/client.h +++ b/include/net/9p/client.h @@ -279,7 +279,7 @@ int p9_client_rename(struct p9_fid *fid, struct p9_fid *newdirfid, const char *name); int p9_client_renameat(struct p9_fid *olddirfid, const char *old_name, struct p9_fid *newdirfid, const char *new_name); -struct p9_client *p9_client_create(const char *dev_name, char *options); +struct p9_client *p9_client_create(struct fs_context *fc); void p9_client_destroy(struct p9_client *clnt); void p9_client_disconnect(struct p9_client *clnt); void p9_client_begin_disconnect(struct p9_client *clnt); diff --git a/include/net/9p/transport.h b/include/net/9p/transport.h index 898a432a8063..a912bbaa862f 100644 --- a/include/net/9p/transport.h +++ b/include/net/9p/transport.h @@ -57,7 +57,7 @@ struct p9_trans_module { bool supports_vmalloc; /* can work with vmalloc'd buffers */ struct module *owner; int (*create)(struct p9_client *client, - const char *devname, char *args); + struct fs_context *fc); void (*close)(struct p9_client *client); int (*request)(struct p9_client *client, struct p9_req_t *req); int (*cancel)(struct p9_client *client, struct p9_req_t *req); diff --git a/net/9p/client.c b/net/9p/client.c index 802f548332a5..f60d1d041adb 100644 --- a/net/9p/client.c +++ b/net/9p/client.c @@ -20,8 +20,8 @@ #include #include #include -#include #include +#include #include #include #include "protocol.h" @@ -33,22 +33,6 @@ * - a little lazy - parse all client options */ -enum { - Opt_msize, - Opt_trans, - Opt_legacy, - Opt_version, - Opt_err, -}; - -static const match_table_t tokens = { - {Opt_msize, "msize=%u"}, - {Opt_legacy, "noextend"}, - {Opt_trans, "trans=%s"}, - {Opt_version, "version=%s"}, - {Opt_err, NULL}, -}; - inline int p9_is_proto_dotl(struct p9_client *clnt) { return clnt->proto_version == p9_proto_2000L; @@ -97,124 +81,16 @@ static int safe_errno(int err) return err; } -/* Interpret mount option for protocol version */ -static int get_protocol_version(char *s) +static int apply_client_options(struct p9_client *clnt, struct fs_context *fc) { - int version = -EINVAL; - - if (!strcmp(s, "9p2000")) { - version = p9_proto_legacy; - p9_debug(P9_DEBUG_9P, "Protocol version: Legacy\n"); - } else if (!strcmp(s, "9p2000.u")) { - version = p9_proto_2000u; - p9_debug(P9_DEBUG_9P, "Protocol version: 9P2000.u\n"); - } else if (!strcmp(s, "9p2000.L")) { - version = p9_proto_2000L; - p9_debug(P9_DEBUG_9P, "Protocol version: 9P2000.L\n"); - } else { - pr_info("Unknown protocol version %s\n", s); - } + struct v9fs_context *ctx = fc->fs_private; - return version; -} - -/** - * parse_opts - parse mount options into client structure - * @opts: options string passed from mount - * @clnt: existing v9fs client information - * - * Return 0 upon success, -ERRNO upon failure - */ - -static int parse_opts(char *opts, struct p9_client *clnt) -{ - char *options, *tmp_options; - char *p; - substring_t args[MAX_OPT_ARGS]; - int option; - char *s; - int ret = 0; - - clnt->proto_version = p9_proto_2000L; - clnt->msize = DEFAULT_MSIZE; - - if (!opts) - return 0; + clnt->msize = ctx->client_opts.msize; + clnt->trans_mod = ctx->client_opts.trans_mod; + ctx->client_opts.trans_mod = NULL; + clnt->proto_version = ctx->client_opts.proto_version; - tmp_options = kstrdup(opts, GFP_KERNEL); - if (!tmp_options) - return -ENOMEM; - options = tmp_options; - - while ((p = strsep(&options, ",")) != NULL) { - int token, r; - - if (!*p) - continue; - token = match_token(p, tokens, args); - switch (token) { - case Opt_msize: - r = match_int(&args[0], &option); - if (r < 0) { - p9_debug(P9_DEBUG_ERROR, - "integer field, but no integer?\n"); - ret = r; - continue; - } - if (option < 4096) { - p9_debug(P9_DEBUG_ERROR, - "msize should be at least 4k\n"); - ret = -EINVAL; - continue; - } - clnt->msize = option; - break; - case Opt_trans: - s = match_strdup(&args[0]); - if (!s) { - ret = -ENOMEM; - p9_debug(P9_DEBUG_ERROR, - "problem allocating copy of trans arg\n"); - goto free_and_return; - } - - v9fs_put_trans(clnt->trans_mod); - clnt->trans_mod = v9fs_get_trans_by_name(s); - if (!clnt->trans_mod) { - pr_info("Could not find request transport: %s\n", - s); - ret = -EINVAL; - } - kfree(s); - break; - case Opt_legacy: - clnt->proto_version = p9_proto_legacy; - break; - case Opt_version: - s = match_strdup(&args[0]); - if (!s) { - ret = -ENOMEM; - p9_debug(P9_DEBUG_ERROR, - "problem allocating copy of version arg\n"); - goto free_and_return; - } - r = get_protocol_version(s); - if (r < 0) - ret = r; - else - clnt->proto_version = r; - kfree(s); - break; - default: - continue; - } - } - -free_and_return: - if (ret) - v9fs_put_trans(clnt->trans_mod); - kfree(tmp_options); - return ret; + return 0; } static int p9_fcall_init(struct p9_client *c, struct p9_fcall *fc, @@ -975,7 +851,7 @@ error: return err; } -struct p9_client *p9_client_create(const char *dev_name, char *options) +struct p9_client *p9_client_create(struct fs_context *fc) { int err; static atomic_t seqno = ATOMIC_INIT(0); @@ -998,8 +874,8 @@ struct p9_client *p9_client_create(const char *dev_name, char *options) idr_init(&clnt->fids); idr_init(&clnt->reqs); - err = parse_opts(options, clnt); - if (err < 0) + err = apply_client_options(clnt, fc); + if (err) goto free_client; if (!clnt->trans_mod) @@ -1015,7 +891,7 @@ struct p9_client *p9_client_create(const char *dev_name, char *options) p9_debug(P9_DEBUG_MUX, "clnt %p trans %p msize %d protocol %d\n", clnt, clnt->trans_mod, clnt->msize, clnt->proto_version); - err = clnt->trans_mod->create(clnt, dev_name, options); + err = clnt->trans_mod->create(clnt, fc); if (err) goto put_trans; diff --git a/net/9p/mod.c b/net/9p/mod.c index 55576c1866fa..85160b52da55 100644 --- a/net/9p/mod.c +++ b/net/9p/mod.c @@ -16,7 +16,6 @@ #include #include #include -#include #include #include #include @@ -171,6 +170,7 @@ void v9fs_put_trans(struct p9_trans_module *m) if (m) module_put(m->owner); } +EXPORT_SYMBOL(v9fs_put_trans); /** * init_p9 - Initialize module diff --git a/net/9p/trans_fd.c b/net/9p/trans_fd.c index b7e5933c4617..4cec4bba222d 100644 --- a/net/9p/trans_fd.c +++ b/net/9p/trans_fd.c @@ -22,7 +22,7 @@ #include #include #include -#include +#include #include #include #include @@ -37,26 +37,6 @@ static struct p9_trans_module p9_tcp_trans; static struct p9_trans_module p9_fd_trans; -/* - * Option Parsing (code inspired by NFS code) - * - a little lazy - parse all fd-transport options - */ - -enum { - /* Options that take integer arguments */ - Opt_port, Opt_rfdno, Opt_wfdno, Opt_err, - /* Options that take no arguments */ - Opt_privport, -}; - -static const match_table_t tokens = { - {Opt_port, "port=%u"}, - {Opt_rfdno, "rfdno=%u"}, - {Opt_wfdno, "wfdno=%u"}, - {Opt_privport, "privport"}, - {Opt_err, NULL}, -}; - enum { Rworksched = 1, /* read work scheduled or running */ Rpending = 2, /* can read */ @@ -737,73 +717,6 @@ static int p9_fd_show_options(struct seq_file *m, struct p9_client *clnt) return 0; } -/** - * parse_opts - parse mount options into p9_fd_opts structure - * @params: options string passed from mount - * @opts: fd transport-specific structure to parse options into - * - * Returns 0 upon success, -ERRNO upon failure - */ - -static int parse_opts(char *params, struct p9_fd_opts *opts) -{ - char *p; - substring_t args[MAX_OPT_ARGS]; - int option; - char *options, *tmp_options; - - opts->port = P9_FD_PORT; - opts->rfd = ~0; - opts->wfd = ~0; - opts->privport = false; - - if (!params) - return 0; - - tmp_options = kstrdup(params, GFP_KERNEL); - if (!tmp_options) { - p9_debug(P9_DEBUG_ERROR, - "failed to allocate copy of option string\n"); - return -ENOMEM; - } - options = tmp_options; - - while ((p = strsep(&options, ",")) != NULL) { - int token; - int r; - if (!*p) - continue; - token = match_token(p, tokens, args); - if ((token != Opt_err) && (token != Opt_privport)) { - r = match_int(&args[0], &option); - if (r < 0) { - p9_debug(P9_DEBUG_ERROR, - "integer field, but no integer?\n"); - continue; - } - } - switch (token) { - case Opt_port: - opts->port = option; - break; - case Opt_rfdno: - opts->rfd = option; - break; - case Opt_wfdno: - opts->wfd = option; - break; - case Opt_privport: - opts->privport = true; - break; - default: - continue; - } - } - - kfree(tmp_options); - return 0; -} - static int p9_fd_open(struct p9_client *client, int rfd, int wfd) { struct p9_trans_fd *ts = kzalloc(sizeof(struct p9_trans_fd), @@ -958,17 +871,18 @@ static int p9_bind_privport(struct socket *sock) } static int -p9_fd_create_tcp(struct p9_client *client, const char *addr, char *args) +p9_fd_create_tcp(struct p9_client *client, struct fs_context *fc) { + const char *addr = fc->source; + struct v9fs_context *ctx = fc->fs_private; int err; char port_str[6]; struct socket *csocket; struct sockaddr_storage stor = { 0 }; struct p9_fd_opts opts; - err = parse_opts(args, &opts); - if (err < 0) - return err; + /* opts are already parsed in context */ + opts = ctx->fd_opts; if (!addr) return -EINVAL; @@ -1015,8 +929,9 @@ p9_fd_create_tcp(struct p9_client *client, const char *addr, char *args) } static int -p9_fd_create_unix(struct p9_client *client, const char *addr, char *args) +p9_fd_create_unix(struct p9_client *client, struct fs_context *fc) { + const char *addr = fc->source; int err; struct socket *csocket; struct sockaddr_un sun_server; @@ -1055,14 +970,12 @@ p9_fd_create_unix(struct p9_client *client, const char *addr, char *args) } static int -p9_fd_create(struct p9_client *client, const char *addr, char *args) +p9_fd_create(struct p9_client *client, struct fs_context *fc) { + struct v9fs_context *ctx = fc->fs_private; + struct p9_fd_opts opts = ctx->fd_opts; int err; - struct p9_fd_opts opts; - err = parse_opts(args, &opts); - if (err < 0) - return err; client->trans_opts.fd.rfd = opts.rfd; client->trans_opts.fd.wfd = opts.wfd; diff --git a/net/9p/trans_rdma.c b/net/9p/trans_rdma.c index 87246463a954..4d406479f83b 100644 --- a/net/9p/trans_rdma.c +++ b/net/9p/trans_rdma.c @@ -22,7 +22,7 @@ #include #include #include -#include +#include #include #include #include @@ -106,26 +106,6 @@ struct p9_rdma_context { }; }; -/* - * Option Parsing (code inspired by NFS code) - */ -enum { - /* Options that take integer arguments */ - Opt_port, Opt_rq_depth, Opt_sq_depth, Opt_timeout, - /* Options that take no argument */ - Opt_privport, - Opt_err, -}; - -static match_table_t tokens = { - {Opt_port, "port=%u"}, - {Opt_sq_depth, "sq=%u"}, - {Opt_rq_depth, "rq=%u"}, - {Opt_timeout, "timeout=%u"}, - {Opt_privport, "privport"}, - {Opt_err, NULL}, -}; - static int p9_rdma_show_options(struct seq_file *m, struct p9_client *clnt) { struct p9_trans_rdma *rdma = clnt->trans; @@ -143,77 +123,6 @@ static int p9_rdma_show_options(struct seq_file *m, struct p9_client *clnt) return 0; } -/** - * parse_opts - parse mount options into rdma options structure - * @params: options string passed from mount - * @opts: rdma transport-specific structure to parse options into - * - * Returns 0 upon success, -ERRNO upon failure - */ -static int parse_opts(char *params, struct p9_rdma_opts *opts) -{ - char *p; - substring_t args[MAX_OPT_ARGS]; - int option; - char *options, *tmp_options; - - opts->port = P9_RDMA_PORT; - opts->sq_depth = P9_RDMA_SQ_DEPTH; - opts->rq_depth = P9_RDMA_RQ_DEPTH; - opts->timeout = P9_RDMA_TIMEOUT; - opts->privport = false; - - if (!params) - return 0; - - tmp_options = kstrdup(params, GFP_KERNEL); - if (!tmp_options) { - p9_debug(P9_DEBUG_ERROR, - "failed to allocate copy of option string\n"); - return -ENOMEM; - } - options = tmp_options; - - while ((p = strsep(&options, ",")) != NULL) { - int token; - int r; - if (!*p) - continue; - token = match_token(p, tokens, args); - if ((token != Opt_err) && (token != Opt_privport)) { - r = match_int(&args[0], &option); - if (r < 0) { - p9_debug(P9_DEBUG_ERROR, - "integer field, but no integer?\n"); - continue; - } - } - switch (token) { - case Opt_port: - opts->port = option; - break; - case Opt_sq_depth: - opts->sq_depth = option; - break; - case Opt_rq_depth: - opts->rq_depth = option; - break; - case Opt_timeout: - opts->timeout = option; - break; - case Opt_privport: - opts->privport = true; - break; - default: - continue; - } - } - /* RQ must be at least as large as the SQ */ - opts->rq_depth = max(opts->rq_depth, opts->sq_depth); - kfree(tmp_options); - return 0; -} - static int p9_cm_event_handler(struct rdma_cm_id *id, struct rdma_cm_event *event) { @@ -607,14 +516,15 @@ static int p9_rdma_bind_privport(struct p9_trans_rdma *rdma) /** * rdma_create_trans - Transport method for creating a transport instance * @client: client instance - * @addr: IP address string - * @args: Mount options string + * @fc: The filesystem context */ static int -rdma_create_trans(struct p9_client *client, const char *addr, char *args) +rdma_create_trans(struct p9_client *client, struct fs_context *fc) { + const char *addr = fc->source; + struct v9fs_context *ctx = fc->fs_private; + struct p9_rdma_opts opts = ctx->rdma_opts; int err; - struct p9_rdma_opts opts; struct p9_trans_rdma *rdma; struct rdma_conn_param conn_param; struct ib_qp_init_attr qp_attr; @@ -622,10 +532,8 @@ rdma_create_trans(struct p9_client *client, const char *addr, char *args) if (addr == NULL) return -EINVAL; - /* Parse the transport specific mount options */ - err = parse_opts(args, &opts); - if (err < 0) - return err; + /* options are already parsed, in the fs context */ + opts = ctx->rdma_opts; /* Create and initialize the RDMA transport structure */ rdma = alloc_rdma(&opts); diff --git a/net/9p/trans_usbg.c b/net/9p/trans_usbg.c index 2542ef099233..93547637deae 100644 --- a/net/9p/trans_usbg.c +++ b/net/9p/trans_usbg.c @@ -27,6 +27,7 @@ #include #include #include +#include #include #include @@ -376,8 +377,9 @@ out: return ret; } -static int p9_usbg_create(struct p9_client *client, const char *devname, char *args) +static int p9_usbg_create(struct p9_client *client, struct fs_context *fc) { + const char *devname = fc->source; struct f_usb9pfs_dev *dev; struct f_usb9pfs *usb9pfs; int ret = -ENOENT; diff --git a/net/9p/trans_virtio.c b/net/9p/trans_virtio.c index b58f50b00c72..10c2dd486438 100644 --- a/net/9p/trans_virtio.c +++ b/net/9p/trans_virtio.c @@ -26,7 +26,7 @@ #include #include #include -#include +#include #include #include #include @@ -679,8 +679,7 @@ fail: /** * p9_virtio_create - allocate a new virtio channel * @client: client instance invoking this transport - * @devname: string identifying the channel to connect to (unused) - * @args: args passed from sys_mount() for per-transport options (unused) + * @fc: the filesystem context * * This sets up a transport channel for 9p communication. Right now * we only match the first available channel, but eventually we could look up @@ -691,8 +690,9 @@ fail: */ static int -p9_virtio_create(struct p9_client *client, const char *devname, char *args) +p9_virtio_create(struct p9_client *client, struct fs_context *fc) { + const char *devname = fc->source; struct virtio_chan *chan; int ret = -ENOENT; int found = 0; diff --git a/net/9p/trans_xen.c b/net/9p/trans_xen.c index 9389c1247001..12f752a92332 100644 --- a/net/9p/trans_xen.c +++ b/net/9p/trans_xen.c @@ -15,6 +15,7 @@ #include #include +#include #include #include #include @@ -66,8 +67,9 @@ static int p9_xen_cancel(struct p9_client *client, struct p9_req_t *req) return 1; } -static int p9_xen_create(struct p9_client *client, const char *addr, char *args) +static int p9_xen_create(struct p9_client *client, struct fs_context *fc) { + const char *addr = fc->source; struct xen_9pfs_front_priv *priv; if (addr == NULL) -- cgit v1.2.3 From d8f9581e1b7f1fe2e1ac985f4ea508d044c90733 Mon Sep 17 00:00:00 2001 From: Kuniyuki Iwashima Date: Wed, 29 Oct 2025 17:32:56 +0000 Subject: ipv6: Add in6_dev_rcu(). rcu_dereference_rtnl() does not clearly tell whether the caller is under RCU or RTNL. Let's add in6_dev_rcu() to make it easy to remove __in6_dev_get() in the future. Signed-off-by: Kuniyuki Iwashima Reviewed-by: Guillaume Nault Link: https://patch.msgid.link/20251029173344.2934622-5-kuniyu@google.com Signed-off-by: Jakub Kicinski --- include/net/addrconf.h | 5 +++++ 1 file changed, 5 insertions(+) (limited to 'include/net') diff --git a/include/net/addrconf.h b/include/net/addrconf.h index 9e5e95988b9e..78e8b877fb25 100644 --- a/include/net/addrconf.h +++ b/include/net/addrconf.h @@ -347,6 +347,11 @@ static inline struct inet6_dev *__in6_dev_get(const struct net_device *dev) return rcu_dereference_rtnl(dev->ip6_ptr); } +static inline struct inet6_dev *in6_dev_rcu(const struct net_device *dev) +{ + return rcu_dereference(dev->ip6_ptr); +} + static inline struct inet6_dev *__in6_dev_get_rtnl_net(const struct net_device *dev) { return rtnl_net_dereference(dev_net(dev), dev->ip6_ptr); -- cgit v1.2.3 From e833eb25161aae6cd0caf14782f405d0ed5765ed Mon Sep 17 00:00:00 2001 From: Kuniyuki Iwashima Date: Wed, 29 Oct 2025 17:33:04 +0000 Subject: mpls: Protect net->mpls.platform_label with a per-netns mutex. MPLS (re)uses RTNL to protect net->mpls.platform_label, but the lock does not need to be RTNL at all. Let's protect net->mpls.platform_label with a dedicated per-netns mutex. Signed-off-by: Kuniyuki Iwashima Reviewed-by: Guillaume Nault Link: https://patch.msgid.link/20251029173344.2934622-13-kuniyu@google.com Signed-off-by: Jakub Kicinski --- include/net/netns/mpls.h | 1 + net/mpls/af_mpls.c | 55 +++++++++++++++++++++++++++++++----------------- net/mpls/internal.h | 7 +++++- 3 files changed, 43 insertions(+), 20 deletions(-) (limited to 'include/net') diff --git a/include/net/netns/mpls.h b/include/net/netns/mpls.h index 19ad2574b267..6682e51513ef 100644 --- a/include/net/netns/mpls.h +++ b/include/net/netns/mpls.h @@ -16,6 +16,7 @@ struct netns_mpls { int default_ttl; size_t platform_labels; struct mpls_route __rcu * __rcu *platform_label; + struct mutex platform_mutex; struct ctl_table_header *ctl; }; diff --git a/net/mpls/af_mpls.c b/net/mpls/af_mpls.c index 49fd15232dbe..d0d047dd2245 100644 --- a/net/mpls/af_mpls.c +++ b/net/mpls/af_mpls.c @@ -79,8 +79,8 @@ static struct mpls_route *mpls_route_input(struct net *net, unsigned int index) { struct mpls_route __rcu **platform_label; - platform_label = rtnl_dereference(net->mpls.platform_label); - return rtnl_dereference(platform_label[index]); + platform_label = mpls_dereference(net, net->mpls.platform_label); + return mpls_dereference(net, platform_label[index]); } static struct mpls_route *mpls_route_input_rcu(struct net *net, unsigned int index) @@ -578,10 +578,8 @@ static void mpls_route_update(struct net *net, unsigned index, struct mpls_route __rcu **platform_label; struct mpls_route *rt; - ASSERT_RTNL(); - - platform_label = rtnl_dereference(net->mpls.platform_label); - rt = rtnl_dereference(platform_label[index]); + platform_label = mpls_dereference(net, net->mpls.platform_label); + rt = mpls_dereference(net, platform_label[index]); rcu_assign_pointer(platform_label[index], new); mpls_notify_route(net, index, rt, new, info); @@ -1472,8 +1470,6 @@ static struct mpls_dev *mpls_add_dev(struct net_device *dev) int err = -ENOMEM; int i; - ASSERT_RTNL(); - mdev = kzalloc(sizeof(*mdev), GFP_KERNEL); if (!mdev) return ERR_PTR(err); @@ -1633,6 +1629,8 @@ static int mpls_dev_notify(struct notifier_block *this, unsigned long event, unsigned int flags; int err; + mutex_lock(&net->mpls.platform_mutex); + if (event == NETDEV_REGISTER) { mdev = mpls_add_dev(dev); if (IS_ERR(mdev)) { @@ -1695,9 +1693,11 @@ static int mpls_dev_notify(struct notifier_block *this, unsigned long event, } out: + mutex_unlock(&net->mpls.platform_mutex); return NOTIFY_OK; err: + mutex_unlock(&net->mpls.platform_mutex); return notifier_from_errno(err); } @@ -1973,6 +1973,7 @@ errout: static int mpls_rtm_delroute(struct sk_buff *skb, struct nlmsghdr *nlh, struct netlink_ext_ack *extack) { + struct net *net = sock_net(skb->sk); struct mpls_route_config *cfg; int err; @@ -1984,7 +1985,9 @@ static int mpls_rtm_delroute(struct sk_buff *skb, struct nlmsghdr *nlh, if (err < 0) goto out; + mutex_lock(&net->mpls.platform_mutex); err = mpls_route_del(cfg, extack); + mutex_unlock(&net->mpls.platform_mutex); out: kfree(cfg); @@ -1995,6 +1998,7 @@ out: static int mpls_rtm_newroute(struct sk_buff *skb, struct nlmsghdr *nlh, struct netlink_ext_ack *extack) { + struct net *net = sock_net(skb->sk); struct mpls_route_config *cfg; int err; @@ -2006,7 +2010,9 @@ static int mpls_rtm_newroute(struct sk_buff *skb, struct nlmsghdr *nlh, if (err < 0) goto out; + mutex_lock(&net->mpls.platform_mutex); err = mpls_route_add(cfg, extack); + mutex_unlock(&net->mpls.platform_mutex); out: kfree(cfg); @@ -2407,6 +2413,8 @@ static int mpls_getroute(struct sk_buff *in_skb, struct nlmsghdr *in_nlh, u8 n_labels; int err; + mutex_lock(&net->mpls.platform_mutex); + err = mpls_valid_getroute_req(in_skb, in_nlh, tb, extack); if (err < 0) goto errout; @@ -2450,7 +2458,8 @@ static int mpls_getroute(struct sk_buff *in_skb, struct nlmsghdr *in_nlh, goto errout_free; } - return rtnl_unicast(skb, net, portid); + err = rtnl_unicast(skb, net, portid); + goto errout; } if (tb[RTA_NEWDST]) { @@ -2542,12 +2551,14 @@ static int mpls_getroute(struct sk_buff *in_skb, struct nlmsghdr *in_nlh, err = rtnl_unicast(skb, net, portid); errout: + mutex_unlock(&net->mpls.platform_mutex); return err; nla_put_failure: nlmsg_cancel(skb, nlh); err = -EMSGSIZE; errout_free: + mutex_unlock(&net->mpls.platform_mutex); kfree_skb(skb); return err; } @@ -2603,9 +2614,10 @@ static int resize_platform_label_table(struct net *net, size_t limit) lo->addr_len); } - rtnl_lock(); + mutex_lock(&net->mpls.platform_mutex); + /* Remember the original table */ - old = rtnl_dereference(net->mpls.platform_label); + old = mpls_dereference(net, net->mpls.platform_label); old_limit = net->mpls.platform_labels; /* Free any labels beyond the new table */ @@ -2636,7 +2648,7 @@ static int resize_platform_label_table(struct net *net, size_t limit) net->mpls.platform_labels = limit; rcu_assign_pointer(net->mpls.platform_label, labels); - rtnl_unlock(); + mutex_unlock(&net->mpls.platform_mutex); mpls_rt_free(rt2); mpls_rt_free(rt0); @@ -2709,12 +2721,13 @@ static const struct ctl_table mpls_table[] = { }, }; -static int mpls_net_init(struct net *net) +static __net_init int mpls_net_init(struct net *net) { size_t table_size = ARRAY_SIZE(mpls_table); struct ctl_table *table; int i; + mutex_init(&net->mpls.platform_mutex); net->mpls.platform_labels = 0; net->mpls.platform_label = NULL; net->mpls.ip_ttl_propagate = 1; @@ -2740,7 +2753,7 @@ static int mpls_net_init(struct net *net) return 0; } -static void mpls_net_exit(struct net *net) +static __net_exit void mpls_net_exit(struct net *net) { struct mpls_route __rcu **platform_label; size_t platform_labels; @@ -2760,16 +2773,20 @@ static void mpls_net_exit(struct net *net) * As such no additional rcu synchronization is necessary when * freeing the platform_label table. */ - rtnl_lock(); - platform_label = rtnl_dereference(net->mpls.platform_label); + mutex_lock(&net->mpls.platform_mutex); + + platform_label = mpls_dereference(net, net->mpls.platform_label); platform_labels = net->mpls.platform_labels; + for (index = 0; index < platform_labels; index++) { - struct mpls_route *rt = rtnl_dereference(platform_label[index]); - RCU_INIT_POINTER(platform_label[index], NULL); + struct mpls_route *rt; + + rt = mpls_dereference(net, platform_label[index]); mpls_notify_route(net, index, rt, NULL, NULL); mpls_rt_free(rt); } - rtnl_unlock(); + + mutex_unlock(&net->mpls.platform_mutex); kvfree(platform_label); } diff --git a/net/mpls/internal.h b/net/mpls/internal.h index 0df01a5395ee..80cb5bbcd946 100644 --- a/net/mpls/internal.h +++ b/net/mpls/internal.h @@ -185,6 +185,11 @@ static inline struct mpls_entry_decoded mpls_entry_decode(struct mpls_shim_hdr * return result; } +#define mpls_dereference(net, p) \ + rcu_dereference_protected( \ + (p), \ + lockdep_is_held(&(net)->mpls.platform_mutex)) + static inline struct mpls_dev *mpls_dev_rcu(const struct net_device *dev) { return rcu_dereference(dev->mpls_ptr); @@ -193,7 +198,7 @@ static inline struct mpls_dev *mpls_dev_rcu(const struct net_device *dev) static inline struct mpls_dev *mpls_dev_get(const struct net *net, const struct net_device *dev) { - return rcu_dereference_rtnl(dev->mpls_ptr); + return mpls_dereference(net, dev->mpls_ptr); } int nla_put_labels(struct sk_buff *skb, int attrtype, u8 labels, -- cgit v1.2.3 From 30ed05adca4a05c50594384cff18910858dd1d35 Mon Sep 17 00:00:00 2001 From: Jason Xing Date: Thu, 30 Oct 2025 08:06:46 +0800 Subject: xsk: use a smaller new lock for shared pool case - Split cq_lock into two smaller locks: cq_prod_lock and cq_cached_prod_lock - Avoid disabling/enabling interrupts in the hot xmit path In either xsk_cq_cancel_locked() or xsk_cq_reserve_locked() function, the race condition is only between multiple xsks sharing the same pool. They are all in the process context rather than interrupt context, so now the small lock named cq_cached_prod_lock can be used without handling interrupts. While cq_cached_prod_lock ensures the exclusive modification of @cached_prod, cq_prod_lock in xsk_cq_submit_addr_locked() only cares about @producer and corresponding @desc. Both of them don't necessarily be consistent with @cached_prod protected by cq_cached_prod_lock. That's the reason why the previous big lock can be split into two smaller ones. Please note that SPSC rule is all about the global state of producer and consumer that can affect both layers instead of local or cached ones. Frequently disabling and enabling interrupt are very time consuming in some cases, especially in a per-descriptor granularity, which now can be avoided after this optimization, even when the pool is shared by multiple xsks. With this patch, the performance number[1] could go from 1,872,565 pps to 1,961,009 pps. It's a minor rise of around 5%. [1]: taskset -c 1 ./xdpsock -i enp2s0f1 -q 0 -t -S -s 64 Signed-off-by: Jason Xing Acked-by: Maciej Fijalkowski Link: https://patch.msgid.link/20251030000646.18859-3-kerneljasonxing@gmail.com Signed-off-by: Paolo Abeni --- include/net/xsk_buff_pool.h | 13 +++++++++---- net/xdp/xsk.c | 15 ++++++--------- net/xdp/xsk_buff_pool.c | 3 ++- 3 files changed, 17 insertions(+), 14 deletions(-) (limited to 'include/net') diff --git a/include/net/xsk_buff_pool.h b/include/net/xsk_buff_pool.h index cac56e6b0869..92a2358c6ce3 100644 --- a/include/net/xsk_buff_pool.h +++ b/include/net/xsk_buff_pool.h @@ -85,11 +85,16 @@ struct xsk_buff_pool { bool unaligned; bool tx_sw_csum; void *addrs; - /* Mutual exclusion of the completion ring in the SKB mode. Two cases to protect: - * NAPI TX thread and sendmsg error paths in the SKB destructor callback and when - * sockets share a single cq when the same netdev and queue id is shared. + /* Mutual exclusion of the completion ring in the SKB mode. + * Protect: NAPI TX thread and sendmsg error paths in the SKB + * destructor callback. */ - spinlock_t cq_lock; + spinlock_t cq_prod_lock; + /* Mutual exclusion of the completion ring in the SKB mode. + * Protect: when sockets share a single cq when the same netdev + * and queue id is shared. + */ + spinlock_t cq_cached_prod_lock; struct xdp_buff_xsk *free_heads[]; }; diff --git a/net/xdp/xsk.c b/net/xdp/xsk.c index 7b0c68a70888..2f26c918d448 100644 --- a/net/xdp/xsk.c +++ b/net/xdp/xsk.c @@ -548,12 +548,11 @@ static int xsk_wakeup(struct xdp_sock *xs, u8 flags) static int xsk_cq_reserve_locked(struct xsk_buff_pool *pool) { - unsigned long flags; int ret; - spin_lock_irqsave(&pool->cq_lock, flags); + spin_lock(&pool->cq_cached_prod_lock); ret = xskq_prod_reserve(pool->cq); - spin_unlock_irqrestore(&pool->cq_lock, flags); + spin_unlock(&pool->cq_cached_prod_lock); return ret; } @@ -566,7 +565,7 @@ static void xsk_cq_submit_addr_locked(struct xsk_buff_pool *pool, unsigned long flags; u32 idx; - spin_lock_irqsave(&pool->cq_lock, flags); + spin_lock_irqsave(&pool->cq_prod_lock, flags); idx = xskq_get_prod(pool->cq); xskq_prod_write_addr(pool->cq, idx, @@ -583,16 +582,14 @@ static void xsk_cq_submit_addr_locked(struct xsk_buff_pool *pool, } } xskq_prod_submit_n(pool->cq, descs_processed); - spin_unlock_irqrestore(&pool->cq_lock, flags); + spin_unlock_irqrestore(&pool->cq_prod_lock, flags); } static void xsk_cq_cancel_locked(struct xsk_buff_pool *pool, u32 n) { - unsigned long flags; - - spin_lock_irqsave(&pool->cq_lock, flags); + spin_lock(&pool->cq_cached_prod_lock); xskq_prod_cancel_n(pool->cq, n); - spin_unlock_irqrestore(&pool->cq_lock, flags); + spin_unlock(&pool->cq_cached_prod_lock); } static void xsk_inc_num_desc(struct sk_buff *skb) diff --git a/net/xdp/xsk_buff_pool.c b/net/xdp/xsk_buff_pool.c index 309075050b2a..00a4eddaa0cd 100644 --- a/net/xdp/xsk_buff_pool.c +++ b/net/xdp/xsk_buff_pool.c @@ -90,7 +90,8 @@ struct xsk_buff_pool *xp_create_and_assign_umem(struct xdp_sock *xs, INIT_LIST_HEAD(&pool->xskb_list); INIT_LIST_HEAD(&pool->xsk_tx_list); spin_lock_init(&pool->xsk_tx_list_lock); - spin_lock_init(&pool->cq_lock); + spin_lock_init(&pool->cq_prod_lock); + spin_lock_init(&pool->cq_cached_prod_lock); refcount_set(&pool->users, 1); pool->fq = xs->fq_tmp; -- cgit v1.2.3 From 0e50474fa514822e9d990874e554bf8043a201d7 Mon Sep 17 00:00:00 2001 From: Kees Cook Date: Mon, 3 Nov 2025 16:26:10 -0800 Subject: net: Convert proto_ops bind() callbacks to use sockaddr_unsized Update all struct proto_ops bind() callback function prototypes from "struct sockaddr *" to "struct sockaddr_unsized *" to avoid lying to the compiler about object sizes. Calls into struct proto handlers gain casts that will be removed in the struct proto conversion patch. No binary changes expected. Signed-off-by: Kees Cook Link: https://patch.msgid.link/20251104002617.2752303-2-kees@kernel.org Signed-off-by: Jakub Kicinski --- crypto/af_alg.c | 2 +- drivers/block/drbd/drbd_receiver.c | 4 ++-- drivers/infiniband/hw/erdma/erdma_cm.c | 4 ++-- drivers/infiniband/sw/siw/siw_cm.c | 6 +++--- drivers/isdn/mISDN/l1oip_core.c | 2 +- drivers/isdn/mISDN/socket.c | 4 ++-- drivers/net/ppp/pptp.c | 4 ++-- drivers/nvme/host/tcp.c | 2 +- drivers/nvme/target/tcp.c | 2 +- drivers/target/iscsi/iscsi_target_login.c | 2 +- drivers/xen/pvcalls-back.c | 2 +- fs/afs/rxrpc.c | 6 +++--- fs/dlm/lowcomms.c | 6 +++--- fs/ocfs2/cluster/tcp.c | 4 ++-- fs/smb/client/connect.c | 2 +- fs/smb/server/transport_tcp.c | 4 ++-- include/linux/net.h | 4 ++-- include/net/inet_common.h | 2 +- include/net/ipv6.h | 2 +- include/net/sock.h | 2 +- net/9p/trans_fd.c | 2 +- net/appletalk/ddp.c | 2 +- net/atm/pvc.c | 4 ++-- net/atm/svc.c | 2 +- net/ax25/af_ax25.c | 2 +- net/bluetooth/hci_sock.c | 2 +- net/bluetooth/iso.c | 4 ++-- net/bluetooth/l2cap_sock.c | 2 +- net/bluetooth/rfcomm/core.c | 4 ++-- net/bluetooth/rfcomm/sock.c | 2 +- net/bluetooth/sco.c | 2 +- net/can/isotp.c | 2 +- net/can/j1939/socket.c | 2 +- net/can/raw.c | 2 +- net/core/sock.c | 2 +- net/ieee802154/socket.c | 4 ++-- net/ipv4/af_inet.c | 4 ++-- net/ipv4/udp_tunnel_core.c | 2 +- net/ipv6/af_inet6.c | 4 ++-- net/ipv6/ip6_udp_tunnel.c | 2 +- net/iucv/af_iucv.c | 2 +- net/l2tp/l2tp_core.c | 4 ++-- net/llc/af_llc.c | 2 +- net/mctp/af_mctp.c | 2 +- net/mctp/test/route-test.c | 2 +- net/mptcp/protocol.c | 6 +++--- net/mptcp/subflow.c | 2 +- net/netfilter/ipvs/ip_vs_sync.c | 4 ++-- net/netlink/af_netlink.c | 2 +- net/netrom/af_netrom.c | 2 +- net/nfc/llcp_sock.c | 4 ++-- net/packet/af_packet.c | 11 ++++++----- net/phonet/socket.c | 8 ++++---- net/qrtr/af_qrtr.c | 2 +- net/qrtr/ns.c | 2 +- net/rds/bind.c | 2 +- net/rds/rds.h | 2 +- net/rds/tcp_connect.c | 2 +- net/rds/tcp_listen.c | 2 +- net/rose/af_rose.c | 2 +- net/rxrpc/af_rxrpc.c | 2 +- net/rxrpc/rxperf.c | 2 +- net/smc/af_smc.c | 2 +- net/smc/smc.h | 2 +- net/socket.c | 6 +++--- net/sunrpc/clnt.c | 4 ++-- net/sunrpc/svcsock.c | 2 +- net/sunrpc/xprtsock.c | 4 ++-- net/tipc/socket.c | 4 ++-- net/unix/af_unix.c | 4 ++-- net/vmw_vsock/af_vsock.c | 4 ++-- net/x25/af_x25.c | 2 +- net/xdp/xsk.c | 2 +- tools/testing/selftests/bpf/test_kmods/bpf_testmod.c | 2 +- 74 files changed, 113 insertions(+), 112 deletions(-) (limited to 'include/net') diff --git a/crypto/af_alg.c b/crypto/af_alg.c index ca6fdcc6c54a..5e760ab62618 100644 --- a/crypto/af_alg.c +++ b/crypto/af_alg.c @@ -145,7 +145,7 @@ void af_alg_release_parent(struct sock *sk) } EXPORT_SYMBOL_GPL(af_alg_release_parent); -static int alg_bind(struct socket *sock, struct sockaddr *uaddr, int addr_len) +static int alg_bind(struct socket *sock, struct sockaddr_unsized *uaddr, int addr_len) { const u32 allowed = CRYPTO_ALG_KERN_DRIVER_ONLY; struct sock *sk = sock->sk; diff --git a/drivers/block/drbd/drbd_receiver.c b/drivers/block/drbd/drbd_receiver.c index caaf2781136d..d9296f74f902 100644 --- a/drivers/block/drbd/drbd_receiver.c +++ b/drivers/block/drbd/drbd_receiver.c @@ -450,7 +450,7 @@ static struct socket *drbd_try_connect(struct drbd_connection *connection) * a free one dynamically. */ what = "bind before connect"; - err = sock->ops->bind(sock, (struct sockaddr *) &src_in6, my_addr_len); + err = sock->ops->bind(sock, (struct sockaddr_unsized *) &src_in6, my_addr_len); if (err < 0) goto out; @@ -537,7 +537,7 @@ static int prepare_listen_socket(struct drbd_connection *connection, struct acce drbd_setbufsize(s_listen, sndbuf_size, rcvbuf_size); what = "bind before listen"; - err = s_listen->ops->bind(s_listen, (struct sockaddr *)&my_addr, my_addr_len); + err = s_listen->ops->bind(s_listen, (struct sockaddr_unsized *)&my_addr, my_addr_len); if (err < 0) goto out; diff --git a/drivers/infiniband/hw/erdma/erdma_cm.c b/drivers/infiniband/hw/erdma/erdma_cm.c index e0acc185e719..ef66a6359eb9 100644 --- a/drivers/infiniband/hw/erdma/erdma_cm.c +++ b/drivers/infiniband/hw/erdma/erdma_cm.c @@ -993,7 +993,7 @@ static int kernel_bindconnect(struct socket *s, struct sockaddr *laddr, int ret; sock_set_reuseaddr(s->sk); - ret = s->ops->bind(s, laddr, laddrlen); + ret = s->ops->bind(s, (struct sockaddr_unsized *)laddr, laddrlen); if (ret) return ret; ret = s->ops->connect(s, raddr, raddrlen, flags); @@ -1315,7 +1315,7 @@ int erdma_create_listen(struct iw_cm_id *id, int backlog) if (ipv4_is_zeronet(laddr->sin_addr.s_addr)) s->sk->sk_bound_dev_if = dev->netdev->ifindex; - ret = s->ops->bind(s, (struct sockaddr *)laddr, + ret = s->ops->bind(s, (struct sockaddr_unsized *)laddr, sizeof(struct sockaddr_in)); if (ret) goto error; diff --git a/drivers/infiniband/sw/siw/siw_cm.c b/drivers/infiniband/sw/siw/siw_cm.c index 708b13993fdf..7fe118cacb3f 100644 --- a/drivers/infiniband/sw/siw/siw_cm.c +++ b/drivers/infiniband/sw/siw/siw_cm.c @@ -1340,7 +1340,7 @@ static int kernel_bindconnect(struct socket *s, struct sockaddr *laddr, return rv; } - rv = s->ops->bind(s, laddr, size); + rv = s->ops->bind(s, (struct sockaddr_unsized *)laddr, size); if (rv < 0) return rv; @@ -1789,7 +1789,7 @@ int siw_create_listen(struct iw_cm_id *id, int backlog) goto error; } } - rv = s->ops->bind(s, (struct sockaddr *)laddr, + rv = s->ops->bind(s, (struct sockaddr_unsized *)laddr, sizeof(struct sockaddr_in)); } else { struct sockaddr_in6 *laddr = &to_sockaddr_in6(id->local_addr); @@ -1813,7 +1813,7 @@ int siw_create_listen(struct iw_cm_id *id, int backlog) goto error; } } - rv = s->ops->bind(s, (struct sockaddr *)laddr, + rv = s->ops->bind(s, (struct sockaddr_unsized *)laddr, sizeof(struct sockaddr_in6)); } if (rv) { diff --git a/drivers/isdn/mISDN/l1oip_core.c b/drivers/isdn/mISDN/l1oip_core.c index f732f6614d37..6ab036e4a35f 100644 --- a/drivers/isdn/mISDN/l1oip_core.c +++ b/drivers/isdn/mISDN/l1oip_core.c @@ -676,7 +676,7 @@ l1oip_socket_thread(void *data) hc->sin_remote.sin_port = htons((unsigned short)hc->remoteport); /* bind to incoming port */ - if (socket->ops->bind(socket, (struct sockaddr *)&hc->sin_local, + if (socket->ops->bind(socket, (struct sockaddr_unsized *)&hc->sin_local, sizeof(hc->sin_local))) { printk(KERN_ERR "%s: Failed to bind socket to port %d.\n", __func__, hc->localport); diff --git a/drivers/isdn/mISDN/socket.c b/drivers/isdn/mISDN/socket.c index b215b28cad7b..77b900db1cac 100644 --- a/drivers/isdn/mISDN/socket.c +++ b/drivers/isdn/mISDN/socket.c @@ -462,7 +462,7 @@ static int data_sock_getsockopt(struct socket *sock, int level, int optname, } static int -data_sock_bind(struct socket *sock, struct sockaddr *addr, int addr_len) +data_sock_bind(struct socket *sock, struct sockaddr_unsized *addr, int addr_len) { struct sockaddr_mISDN *maddr = (struct sockaddr_mISDN *) addr; struct sock *sk = sock->sk; @@ -696,7 +696,7 @@ base_sock_ioctl(struct socket *sock, unsigned int cmd, unsigned long arg) } static int -base_sock_bind(struct socket *sock, struct sockaddr *addr, int addr_len) +base_sock_bind(struct socket *sock, struct sockaddr_unsized *addr, int addr_len) { struct sockaddr_mISDN *maddr = (struct sockaddr_mISDN *) addr; struct sock *sk = sock->sk; diff --git a/drivers/net/ppp/pptp.c b/drivers/net/ppp/pptp.c index 90737cb71892..d07e87a0974c 100644 --- a/drivers/net/ppp/pptp.c +++ b/drivers/net/ppp/pptp.c @@ -382,8 +382,8 @@ drop: return NET_RX_DROP; } -static int pptp_bind(struct socket *sock, struct sockaddr *uservaddr, - int sockaddr_len) +static int pptp_bind(struct socket *sock, struct sockaddr_unsized *uservaddr, + int sockaddr_len) { struct sock *sk = sock->sk; struct sockaddr_pppox *sp = (struct sockaddr_pppox *) uservaddr; diff --git a/drivers/nvme/host/tcp.c b/drivers/nvme/host/tcp.c index 9a96df1a511c..35d0bd91f6fd 100644 --- a/drivers/nvme/host/tcp.c +++ b/drivers/nvme/host/tcp.c @@ -1834,7 +1834,7 @@ static int nvme_tcp_alloc_queue(struct nvme_ctrl *nctrl, int qid, sk_set_memalloc(queue->sock->sk); if (nctrl->opts->mask & NVMF_OPT_HOST_TRADDR) { - ret = kernel_bind(queue->sock, (struct sockaddr *)&ctrl->src_addr, + ret = kernel_bind(queue->sock, (struct sockaddr_unsized *)&ctrl->src_addr, sizeof(ctrl->src_addr)); if (ret) { dev_err(nctrl->device, diff --git a/drivers/nvme/target/tcp.c b/drivers/nvme/target/tcp.c index 470bf37e5a63..d543da09ef8e 100644 --- a/drivers/nvme/target/tcp.c +++ b/drivers/nvme/target/tcp.c @@ -2055,7 +2055,7 @@ static int nvmet_tcp_add_port(struct nvmet_port *nport) if (so_priority > 0) sock_set_priority(port->sock->sk, so_priority); - ret = kernel_bind(port->sock, (struct sockaddr *)&port->addr, + ret = kernel_bind(port->sock, (struct sockaddr_unsized *)&port->addr, sizeof(port->addr)); if (ret) { pr_err("failed to bind port socket %d\n", ret); diff --git a/drivers/target/iscsi/iscsi_target_login.c b/drivers/target/iscsi/iscsi_target_login.c index c2ac9a99ebbb..53aca059dc16 100644 --- a/drivers/target/iscsi/iscsi_target_login.c +++ b/drivers/target/iscsi/iscsi_target_login.c @@ -822,7 +822,7 @@ int iscsit_setup_np( sock_set_reuseaddr(sock->sk); ip_sock_set_freebind(sock->sk); - ret = kernel_bind(sock, (struct sockaddr *)&np->np_sockaddr, len); + ret = kernel_bind(sock, (struct sockaddr_unsized *)&np->np_sockaddr, len); if (ret < 0) { pr_err("kernel_bind() failed: %d\n", ret); goto fail; diff --git a/drivers/xen/pvcalls-back.c b/drivers/xen/pvcalls-back.c index fd7ed65e0197..da1b516b9cfd 100644 --- a/drivers/xen/pvcalls-back.c +++ b/drivers/xen/pvcalls-back.c @@ -650,7 +650,7 @@ static int pvcalls_back_bind(struct xenbus_device *dev, if (ret < 0) goto out; - ret = inet_bind(map->sock, (struct sockaddr *)&req->u.bind.addr, + ret = inet_bind(map->sock, (struct sockaddr_unsized *)&req->u.bind.addr, req->u.bind.len); if (ret < 0) goto out; diff --git a/fs/afs/rxrpc.c b/fs/afs/rxrpc.c index c1cadf8fb346..bf0e4ea0aafd 100644 --- a/fs/afs/rxrpc.c +++ b/fs/afs/rxrpc.c @@ -82,16 +82,16 @@ int afs_open_socket(struct afs_net *net) if (ret < 0) pr_err("Couldn't create RxGK CM key: %d\n", ret); - ret = kernel_bind(socket, (struct sockaddr *) &srx, sizeof(srx)); + ret = kernel_bind(socket, (struct sockaddr_unsized *) &srx, sizeof(srx)); if (ret == -EADDRINUSE) { srx.transport.sin6.sin6_port = 0; - ret = kernel_bind(socket, (struct sockaddr *) &srx, sizeof(srx)); + ret = kernel_bind(socket, (struct sockaddr_unsized *) &srx, sizeof(srx)); } if (ret < 0) goto error_2; srx.srx_service = YFS_CM_SERVICE; - ret = kernel_bind(socket, (struct sockaddr *) &srx, sizeof(srx)); + ret = kernel_bind(socket, (struct sockaddr_unsized *) &srx, sizeof(srx)); if (ret < 0) goto error_2; diff --git a/fs/dlm/lowcomms.c b/fs/dlm/lowcomms.c index 9a0b6c2b6b01..0500421b6e3b 100644 --- a/fs/dlm/lowcomms.c +++ b/fs/dlm/lowcomms.c @@ -1134,7 +1134,7 @@ static int sctp_bind_addrs(struct socket *sock, __be16 port) make_sockaddr(&localaddr, port, &addr_len); if (!i) - result = kernel_bind(sock, addr, addr_len); + result = kernel_bind(sock, (struct sockaddr_unsized *)addr, addr_len); else result = sock_bind_add(sock->sk, addr, addr_len); @@ -1813,7 +1813,7 @@ static int dlm_tcp_bind(struct socket *sock) memcpy(&src_addr, &dlm_local_addr[0], sizeof(src_addr)); make_sockaddr(&src_addr, 0, &addr_len); - result = kernel_bind(sock, (struct sockaddr *)&src_addr, + result = kernel_bind(sock, (struct sockaddr_unsized *)&src_addr, addr_len); if (result < 0) { /* This *may* not indicate a critical error */ @@ -1852,7 +1852,7 @@ static int dlm_tcp_listen_bind(struct socket *sock) /* Bind to our port */ make_sockaddr(&dlm_local_addr[0], dlm_config.ci_tcp_port, &addr_len); - return kernel_bind(sock, (struct sockaddr *)&dlm_local_addr[0], + return kernel_bind(sock, (struct sockaddr_unsized *)&dlm_local_addr[0], addr_len); } diff --git a/fs/ocfs2/cluster/tcp.c b/fs/ocfs2/cluster/tcp.c index b05d4e9d13b2..c7734193d8d7 100644 --- a/fs/ocfs2/cluster/tcp.c +++ b/fs/ocfs2/cluster/tcp.c @@ -1615,7 +1615,7 @@ static void o2net_start_connect(struct work_struct *work) myaddr.sin_addr.s_addr = mynode->nd_ipv4_address; myaddr.sin_port = htons(0); /* any port */ - ret = sock->ops->bind(sock, (struct sockaddr *)&myaddr, + ret = sock->ops->bind(sock, (struct sockaddr_unsized *)&myaddr, sizeof(myaddr)); if (ret) { mlog(ML_ERROR, "bind failed with %d at address %pI4\n", @@ -2002,7 +2002,7 @@ static int o2net_open_listening_sock(__be32 addr, __be16 port) INIT_WORK(&o2net_listen_work, o2net_accept_many); sock->sk->sk_reuse = SK_CAN_REUSE; - ret = sock->ops->bind(sock, (struct sockaddr *)&sin, sizeof(sin)); + ret = sock->ops->bind(sock, (struct sockaddr_unsized *)&sin, sizeof(sin)); if (ret < 0) { printk(KERN_ERR "o2net: Error %d while binding socket at " "%pI4:%u\n", ret, &addr, ntohs(port)); diff --git a/fs/smb/client/connect.c b/fs/smb/client/connect.c index dd12f3eb61dc..96d972263020 100644 --- a/fs/smb/client/connect.c +++ b/fs/smb/client/connect.c @@ -3112,7 +3112,7 @@ bind_socket(struct TCP_Server_Info *server) struct socket *socket = server->ssocket; rc = kernel_bind(socket, - (struct sockaddr *) &server->srcaddr, + (struct sockaddr_unsized *) &server->srcaddr, sizeof(server->srcaddr)); if (rc < 0) { struct sockaddr_in *saddr4; diff --git a/fs/smb/server/transport_tcp.c b/fs/smb/server/transport_tcp.c index 7a1e3dcc2cde..bf694bc78c65 100644 --- a/fs/smb/server/transport_tcp.c +++ b/fs/smb/server/transport_tcp.c @@ -519,10 +519,10 @@ static int create_socket(struct interface *iface) } if (ipv4) - ret = kernel_bind(ksmbd_socket, (struct sockaddr *)&sin, + ret = kernel_bind(ksmbd_socket, (struct sockaddr_unsized *)&sin, sizeof(sin)); else - ret = kernel_bind(ksmbd_socket, (struct sockaddr *)&sin6, + ret = kernel_bind(ksmbd_socket, (struct sockaddr_unsized *)&sin6, sizeof(sin6)); if (ret) { pr_err("Failed to bind socket: %d\n", ret); diff --git a/include/linux/net.h b/include/linux/net.h index ec09620f40f7..0e316f063113 100644 --- a/include/linux/net.h +++ b/include/linux/net.h @@ -163,7 +163,7 @@ struct proto_ops { struct module *owner; int (*release) (struct socket *sock); int (*bind) (struct socket *sock, - struct sockaddr *myaddr, + struct sockaddr_unsized *myaddr, int sockaddr_len); int (*connect) (struct socket *sock, struct sockaddr *vaddr, @@ -345,7 +345,7 @@ int kernel_sendmsg(struct socket *sock, struct msghdr *msg, struct kvec *vec, int kernel_recvmsg(struct socket *sock, struct msghdr *msg, struct kvec *vec, size_t num, size_t len, int flags); -int kernel_bind(struct socket *sock, struct sockaddr *addr, int addrlen); +int kernel_bind(struct socket *sock, struct sockaddr_unsized *addr, int addrlen); int kernel_listen(struct socket *sock, int backlog); int kernel_accept(struct socket *sock, struct socket **newsock, int flags); int kernel_connect(struct socket *sock, struct sockaddr *addr, int addrlen, diff --git a/include/net/inet_common.h b/include/net/inet_common.h index c17a6585d0b0..1666cf6f539e 100644 --- a/include/net/inet_common.h +++ b/include/net/inet_common.h @@ -42,7 +42,7 @@ int inet_shutdown(struct socket *sock, int how); int inet_listen(struct socket *sock, int backlog); int __inet_listen_sk(struct sock *sk, int backlog); void inet_sock_destruct(struct sock *sk); -int inet_bind(struct socket *sock, struct sockaddr *uaddr, int addr_len); +int inet_bind(struct socket *sock, struct sockaddr_unsized *uaddr, int addr_len); int inet_bind_sk(struct sock *sk, struct sockaddr *uaddr, int addr_len); /* Don't allocate port at this moment, defer to connect. */ #define BIND_FORCE_ADDRESS_NO_PORT (1 << 0) diff --git a/include/net/ipv6.h b/include/net/ipv6.h index 2ccdf85f34f1..2188bad9a687 100644 --- a/include/net/ipv6.h +++ b/include/net/ipv6.h @@ -1208,7 +1208,7 @@ void ipv6_local_rxpmtu(struct sock *sk, struct flowi6 *fl6, u32 mtu); void inet6_cleanup_sock(struct sock *sk); void inet6_sock_destruct(struct sock *sk); int inet6_release(struct socket *sock); -int inet6_bind(struct socket *sock, struct sockaddr *uaddr, int addr_len); +int inet6_bind(struct socket *sock, struct sockaddr_unsized *uaddr, int addr_len); int inet6_bind_sk(struct sock *sk, struct sockaddr *uaddr, int addr_len); int inet6_getname(struct socket *sock, struct sockaddr *uaddr, int peer); diff --git a/include/net/sock.h b/include/net/sock.h index c7e58b8e8a90..acbb78c96d69 100644 --- a/include/net/sock.h +++ b/include/net/sock.h @@ -1920,7 +1920,7 @@ int sock_cmsg_send(struct sock *sk, struct msghdr *msg, * Functions to fill in entries in struct proto_ops when a protocol * does not implement a particular function. */ -int sock_no_bind(struct socket *, struct sockaddr *, int); +int sock_no_bind(struct socket *sock, struct sockaddr_unsized *saddr, int len); int sock_no_connect(struct socket *, struct sockaddr *, int, int); int sock_no_socketpair(struct socket *, struct socket *); int sock_no_accept(struct socket *, struct socket *, struct proto_accept_arg *); diff --git a/net/9p/trans_fd.c b/net/9p/trans_fd.c index a516745f732f..ef517bb307e2 100644 --- a/net/9p/trans_fd.c +++ b/net/9p/trans_fd.c @@ -966,7 +966,7 @@ static int p9_bind_privport(struct socket *sock) ((struct sockaddr_in *)&stor)->sin_port = htons((ushort)port); else ((struct sockaddr_in6 *)&stor)->sin6_port = htons((ushort)port); - err = kernel_bind(sock, (struct sockaddr *)&stor, sizeof(stor)); + err = kernel_bind(sock, (struct sockaddr_unsized *)&stor, sizeof(stor)); if (err != -EADDRINUSE) break; } diff --git a/net/appletalk/ddp.c b/net/appletalk/ddp.c index 30242fe10341..45db43cde67f 100644 --- a/net/appletalk/ddp.c +++ b/net/appletalk/ddp.c @@ -1149,7 +1149,7 @@ out: } /* Set the address 'our end' of the connection */ -static int atalk_bind(struct socket *sock, struct sockaddr *uaddr, int addr_len) +static int atalk_bind(struct socket *sock, struct sockaddr_unsized *uaddr, int addr_len) { struct sockaddr_at *addr = (struct sockaddr_at *)uaddr; struct sock *sk = sock->sk; diff --git a/net/atm/pvc.c b/net/atm/pvc.c index 66d9a9bd5896..62fdf07c53de 100644 --- a/net/atm/pvc.c +++ b/net/atm/pvc.c @@ -24,7 +24,7 @@ static int pvc_shutdown(struct socket *sock, int how) return 0; } -static int pvc_bind(struct socket *sock, struct sockaddr *sockaddr, +static int pvc_bind(struct socket *sock, struct sockaddr_unsized *sockaddr, int sockaddr_len) { struct sock *sk = sock->sk; @@ -59,7 +59,7 @@ out: static int pvc_connect(struct socket *sock, struct sockaddr *sockaddr, int sockaddr_len, int flags) { - return pvc_bind(sock, sockaddr, sockaddr_len); + return pvc_bind(sock, (struct sockaddr_unsized *)sockaddr, sockaddr_len); } static int pvc_setsockopt(struct socket *sock, int level, int optname, diff --git a/net/atm/svc.c b/net/atm/svc.c index f8137ae693b0..1906a493c8aa 100644 --- a/net/atm/svc.c +++ b/net/atm/svc.c @@ -97,7 +97,7 @@ static int svc_release(struct socket *sock) return 0; } -static int svc_bind(struct socket *sock, struct sockaddr *sockaddr, +static int svc_bind(struct socket *sock, struct sockaddr_unsized *sockaddr, int sockaddr_len) { DEFINE_WAIT(wait); diff --git a/net/ax25/af_ax25.c b/net/ax25/af_ax25.c index 6ef8b2a57a9b..23c558ff9682 100644 --- a/net/ax25/af_ax25.c +++ b/net/ax25/af_ax25.c @@ -1094,7 +1094,7 @@ static int ax25_release(struct socket *sock) * that we've implemented support for SO_BINDTODEVICE. It is however small * and trivially backward compatible. */ -static int ax25_bind(struct socket *sock, struct sockaddr *uaddr, int addr_len) +static int ax25_bind(struct socket *sock, struct sockaddr_unsized *uaddr, int addr_len) { struct sock *sk = sock->sk; struct full_sockaddr_ax25 *addr = (struct full_sockaddr_ax25 *)uaddr; diff --git a/net/bluetooth/hci_sock.c b/net/bluetooth/hci_sock.c index fc866759910d..ba9f48771e11 100644 --- a/net/bluetooth/hci_sock.c +++ b/net/bluetooth/hci_sock.c @@ -1185,7 +1185,7 @@ static int hci_sock_compat_ioctl(struct socket *sock, unsigned int cmd, } #endif -static int hci_sock_bind(struct socket *sock, struct sockaddr *addr, +static int hci_sock_bind(struct socket *sock, struct sockaddr_unsized *addr, int addr_len) { struct sockaddr_hci haddr; diff --git a/net/bluetooth/iso.c b/net/bluetooth/iso.c index 3d98cb6291da..6a7e1b4a8701 100644 --- a/net/bluetooth/iso.c +++ b/net/bluetooth/iso.c @@ -944,7 +944,7 @@ static int iso_sock_create(struct net *net, struct socket *sock, int protocol, return 0; } -static int iso_sock_bind_bc(struct socket *sock, struct sockaddr *addr, +static int iso_sock_bind_bc(struct socket *sock, struct sockaddr_unsized *addr, int addr_len) { struct sockaddr_iso *sa = (struct sockaddr_iso *)addr; @@ -1022,7 +1022,7 @@ done: return err; } -static int iso_sock_bind(struct socket *sock, struct sockaddr *addr, +static int iso_sock_bind(struct socket *sock, struct sockaddr_unsized *addr, int addr_len) { struct sockaddr_iso *sa = (struct sockaddr_iso *)addr; diff --git a/net/bluetooth/l2cap_sock.c b/net/bluetooth/l2cap_sock.c index 814fb8610ac4..ca7394d8fa4e 100644 --- a/net/bluetooth/l2cap_sock.c +++ b/net/bluetooth/l2cap_sock.c @@ -80,7 +80,7 @@ static int l2cap_validate_le_psm(u16 psm) return 0; } -static int l2cap_sock_bind(struct socket *sock, struct sockaddr *addr, int alen) +static int l2cap_sock_bind(struct socket *sock, struct sockaddr_unsized *addr, int alen) { struct sock *sk = sock->sk; struct l2cap_chan *chan = l2cap_pi(sk)->chan; diff --git a/net/bluetooth/rfcomm/core.c b/net/bluetooth/rfcomm/core.c index 96250807b32b..d62fd6c57617 100644 --- a/net/bluetooth/rfcomm/core.c +++ b/net/bluetooth/rfcomm/core.c @@ -781,7 +781,7 @@ static struct rfcomm_session *rfcomm_session_create(bdaddr_t *src, addr.l2_psm = 0; addr.l2_cid = 0; addr.l2_bdaddr_type = BDADDR_BREDR; - *err = kernel_bind(sock, (struct sockaddr *) &addr, sizeof(addr)); + *err = kernel_bind(sock, (struct sockaddr_unsized *)&addr, sizeof(addr)); if (*err < 0) goto failed; @@ -2068,7 +2068,7 @@ static int rfcomm_add_listener(bdaddr_t *ba) addr.l2_psm = cpu_to_le16(L2CAP_PSM_RFCOMM); addr.l2_cid = 0; addr.l2_bdaddr_type = BDADDR_BREDR; - err = kernel_bind(sock, (struct sockaddr *) &addr, sizeof(addr)); + err = kernel_bind(sock, (struct sockaddr_unsized *)&addr, sizeof(addr)); if (err < 0) { BT_ERR("Bind failed %d", err); goto failed; diff --git a/net/bluetooth/rfcomm/sock.c b/net/bluetooth/rfcomm/sock.c index 913402806fa0..8c8762bbc6de 100644 --- a/net/bluetooth/rfcomm/sock.c +++ b/net/bluetooth/rfcomm/sock.c @@ -324,7 +324,7 @@ static int rfcomm_sock_create(struct net *net, struct socket *sock, return 0; } -static int rfcomm_sock_bind(struct socket *sock, struct sockaddr *addr, int addr_len) +static int rfcomm_sock_bind(struct socket *sock, struct sockaddr_unsized *addr, int addr_len) { struct sockaddr_rc sa; struct sock *sk = sock->sk; diff --git a/net/bluetooth/sco.c b/net/bluetooth/sco.c index ab0cf442d57b..01d878205e58 100644 --- a/net/bluetooth/sco.c +++ b/net/bluetooth/sco.c @@ -605,7 +605,7 @@ static int sco_sock_create(struct net *net, struct socket *sock, int protocol, return 0; } -static int sco_sock_bind(struct socket *sock, struct sockaddr *addr, +static int sco_sock_bind(struct socket *sock, struct sockaddr_unsized *addr, int addr_len) { struct sockaddr_sco *sa = (struct sockaddr_sco *) addr; diff --git a/net/can/isotp.c b/net/can/isotp.c index 74ee1e52249b..ce588b85665a 100644 --- a/net/can/isotp.c +++ b/net/can/isotp.c @@ -1246,7 +1246,7 @@ static int isotp_release(struct socket *sock) return 0; } -static int isotp_bind(struct socket *sock, struct sockaddr *uaddr, int len) +static int isotp_bind(struct socket *sock, struct sockaddr_unsized *uaddr, int len) { struct sockaddr_can *addr = (struct sockaddr_can *)uaddr; struct sock *sk = sock->sk; diff --git a/net/can/j1939/socket.c b/net/can/j1939/socket.c index 88e7160d4248..a2abedc757d0 100644 --- a/net/can/j1939/socket.c +++ b/net/can/j1939/socket.c @@ -440,7 +440,7 @@ static int j1939_sk_sanity_check(struct sockaddr_can *addr, int len) return 0; } -static int j1939_sk_bind(struct socket *sock, struct sockaddr *uaddr, int len) +static int j1939_sk_bind(struct socket *sock, struct sockaddr_unsized *uaddr, int len) { struct sockaddr_can *addr = (struct sockaddr_can *)uaddr; struct j1939_sock *jsk = j1939_sk(sock->sk); diff --git a/net/can/raw.c b/net/can/raw.c index a53853f5e9af..f36a83d3447c 100644 --- a/net/can/raw.c +++ b/net/can/raw.c @@ -449,7 +449,7 @@ static int raw_release(struct socket *sock) return 0; } -static int raw_bind(struct socket *sock, struct sockaddr *uaddr, int len) +static int raw_bind(struct socket *sock, struct sockaddr_unsized *uaddr, int len) { struct sockaddr_can *addr = (struct sockaddr_can *)uaddr; struct sock *sk = sock->sk; diff --git a/net/core/sock.c b/net/core/sock.c index 7a9bbc2afcf0..1e1ce18bba16 100644 --- a/net/core/sock.c +++ b/net/core/sock.c @@ -3462,7 +3462,7 @@ EXPORT_SYMBOL_GPL(sk_set_peek_off); * function, some default processing is provided. */ -int sock_no_bind(struct socket *sock, struct sockaddr *saddr, int len) +int sock_no_bind(struct socket *sock, struct sockaddr_unsized *saddr, int len) { return -EOPNOTSUPP; } diff --git a/net/ieee802154/socket.c b/net/ieee802154/socket.c index 18d267921bb5..99ddfad9bb88 100644 --- a/net/ieee802154/socket.c +++ b/net/ieee802154/socket.c @@ -96,13 +96,13 @@ static int ieee802154_sock_sendmsg(struct socket *sock, struct msghdr *msg, return sk->sk_prot->sendmsg(sk, msg, len); } -static int ieee802154_sock_bind(struct socket *sock, struct sockaddr *uaddr, +static int ieee802154_sock_bind(struct socket *sock, struct sockaddr_unsized *uaddr, int addr_len) { struct sock *sk = sock->sk; if (sk->sk_prot->bind) - return sk->sk_prot->bind(sk, uaddr, addr_len); + return sk->sk_prot->bind(sk, (struct sockaddr *)uaddr, addr_len); return sock_no_bind(sock, uaddr, addr_len); } diff --git a/net/ipv4/af_inet.c b/net/ipv4/af_inet.c index 0784e2a873a1..aa43d16e48ff 100644 --- a/net/ipv4/af_inet.c +++ b/net/ipv4/af_inet.c @@ -464,9 +464,9 @@ int inet_bind_sk(struct sock *sk, struct sockaddr *uaddr, int addr_len) return __inet_bind(sk, uaddr, addr_len, flags); } -int inet_bind(struct socket *sock, struct sockaddr *uaddr, int addr_len) +int inet_bind(struct socket *sock, struct sockaddr_unsized *uaddr, int addr_len) { - return inet_bind_sk(sock->sk, uaddr, addr_len); + return inet_bind_sk(sock->sk, (struct sockaddr *)uaddr, addr_len); } EXPORT_SYMBOL(inet_bind); diff --git a/net/ipv4/udp_tunnel_core.c b/net/ipv4/udp_tunnel_core.c index 54386e06a813..11e5a88c923d 100644 --- a/net/ipv4/udp_tunnel_core.c +++ b/net/ipv4/udp_tunnel_core.c @@ -29,7 +29,7 @@ int udp_sock_create4(struct net *net, struct udp_port_cfg *cfg, udp_addr.sin_family = AF_INET; udp_addr.sin_addr = cfg->local_ip; udp_addr.sin_port = cfg->local_udp_port; - err = kernel_bind(sock, (struct sockaddr *)&udp_addr, + err = kernel_bind(sock, (struct sockaddr_unsized *)&udp_addr, sizeof(udp_addr)); if (err < 0) goto error; diff --git a/net/ipv6/af_inet6.c b/net/ipv6/af_inet6.c index 44d7de1eec4f..c92d27e35fbc 100644 --- a/net/ipv6/af_inet6.c +++ b/net/ipv6/af_inet6.c @@ -465,9 +465,9 @@ int inet6_bind_sk(struct sock *sk, struct sockaddr *uaddr, int addr_len) } /* bind for INET6 API */ -int inet6_bind(struct socket *sock, struct sockaddr *uaddr, int addr_len) +int inet6_bind(struct socket *sock, struct sockaddr_unsized *uaddr, int addr_len) { - return inet6_bind_sk(sock->sk, uaddr, addr_len); + return inet6_bind_sk(sock->sk, (struct sockaddr *)uaddr, addr_len); } EXPORT_SYMBOL(inet6_bind); diff --git a/net/ipv6/ip6_udp_tunnel.c b/net/ipv6/ip6_udp_tunnel.c index 0ff547a4bff7..b0d9286b33c8 100644 --- a/net/ipv6/ip6_udp_tunnel.c +++ b/net/ipv6/ip6_udp_tunnel.c @@ -40,7 +40,7 @@ int udp_sock_create6(struct net *net, struct udp_port_cfg *cfg, memcpy(&udp6_addr.sin6_addr, &cfg->local_ip6, sizeof(udp6_addr.sin6_addr)); udp6_addr.sin6_port = cfg->local_udp_port; - err = kernel_bind(sock, (struct sockaddr *)&udp6_addr, + err = kernel_bind(sock, (struct sockaddr_unsized *)&udp6_addr, sizeof(udp6_addr)); if (err < 0) goto error; diff --git a/net/iucv/af_iucv.c b/net/iucv/af_iucv.c index 4ddfc633d30c..3941e32cda69 100644 --- a/net/iucv/af_iucv.c +++ b/net/iucv/af_iucv.c @@ -563,7 +563,7 @@ static void __iucv_auto_name(struct iucv_sock *iucv) } /* Bind an unbound socket */ -static int iucv_sock_bind(struct socket *sock, struct sockaddr *addr, +static int iucv_sock_bind(struct socket *sock, struct sockaddr_unsized *addr, int addr_len) { DECLARE_SOCKADDR(struct sockaddr_iucv *, sa, addr); diff --git a/net/l2tp/l2tp_core.c b/net/l2tp/l2tp_core.c index 369a2f2e459c..4b5e372a5cd4 100644 --- a/net/l2tp/l2tp_core.c +++ b/net/l2tp/l2tp_core.c @@ -1503,7 +1503,7 @@ static int l2tp_tunnel_sock_create(struct net *net, memcpy(&ip6_addr.l2tp_addr, cfg->local_ip6, sizeof(ip6_addr.l2tp_addr)); ip6_addr.l2tp_conn_id = tunnel_id; - err = kernel_bind(sock, (struct sockaddr *)&ip6_addr, + err = kernel_bind(sock, (struct sockaddr_unsized *)&ip6_addr, sizeof(ip6_addr)); if (err < 0) goto out; @@ -1530,7 +1530,7 @@ static int l2tp_tunnel_sock_create(struct net *net, ip_addr.l2tp_family = AF_INET; ip_addr.l2tp_addr = cfg->local_ip; ip_addr.l2tp_conn_id = tunnel_id; - err = kernel_bind(sock, (struct sockaddr *)&ip_addr, + err = kernel_bind(sock, (struct sockaddr_unsized *)&ip_addr, sizeof(ip_addr)); if (err < 0) goto out; diff --git a/net/llc/af_llc.c b/net/llc/af_llc.c index 5958a80fe14c..e5bb0c0d708c 100644 --- a/net/llc/af_llc.c +++ b/net/llc/af_llc.c @@ -337,7 +337,7 @@ out: * otherwise all hell will break loose. * Returns: 0 upon success, negative otherwise. */ -static int llc_ui_bind(struct socket *sock, struct sockaddr *uaddr, int addrlen) +static int llc_ui_bind(struct socket *sock, struct sockaddr_unsized *uaddr, int addrlen) { struct sockaddr_llc *addr = (struct sockaddr_llc *)uaddr; struct sock *sk = sock->sk; diff --git a/net/mctp/af_mctp.c b/net/mctp/af_mctp.c index b99ba14f39d2..5b1ef50637b7 100644 --- a/net/mctp/af_mctp.c +++ b/net/mctp/af_mctp.c @@ -49,7 +49,7 @@ static bool mctp_sockaddr_ext_is_ok(const struct sockaddr_mctp_ext *addr) !addr->__smctp_pad0[2]; } -static int mctp_bind(struct socket *sock, struct sockaddr *addr, int addrlen) +static int mctp_bind(struct socket *sock, struct sockaddr_unsized *addr, int addrlen) { struct sock *sk = sock->sk; struct mctp_sock *msk = container_of(sk, struct mctp_sock, sk); diff --git a/net/mctp/test/route-test.c b/net/mctp/test/route-test.c index 69a3ccfc6310..be9149ac79dd 100644 --- a/net/mctp/test/route-test.c +++ b/net/mctp/test/route-test.c @@ -205,7 +205,7 @@ static void __mctp_route_test_init(struct kunit *test, addr.smctp_network = netid; addr.smctp_addr.s_addr = 8; addr.smctp_type = 0; - rc = kernel_bind(sock, (struct sockaddr *)&addr, sizeof(addr)); + rc = kernel_bind(sock, (struct sockaddr_unsized *)&addr, sizeof(addr)); KUNIT_ASSERT_EQ(test, rc, 0); *devp = dev; diff --git a/net/mptcp/protocol.c b/net/mptcp/protocol.c index d568575cdcb5..53e2b095dfb1 100644 --- a/net/mptcp/protocol.c +++ b/net/mptcp/protocol.c @@ -3856,7 +3856,7 @@ static struct proto mptcp_prot = { .no_autobind = true, }; -static int mptcp_bind(struct socket *sock, struct sockaddr *uaddr, int addr_len) +static int mptcp_bind(struct socket *sock, struct sockaddr_unsized *uaddr, int addr_len) { struct mptcp_sock *msk = mptcp_sk(sock->sk); struct sock *ssk, *sk = sock->sk; @@ -3870,10 +3870,10 @@ static int mptcp_bind(struct socket *sock, struct sockaddr *uaddr, int addr_len) } if (sk->sk_family == AF_INET) - err = inet_bind_sk(ssk, uaddr, addr_len); + err = inet_bind_sk(ssk, (struct sockaddr *)uaddr, addr_len); #if IS_ENABLED(CONFIG_MPTCP_IPV6) else if (sk->sk_family == AF_INET6) - err = inet6_bind_sk(ssk, uaddr, addr_len); + err = inet6_bind_sk(ssk, (struct sockaddr *)uaddr, addr_len); #endif if (!err) mptcp_copy_inaddrs(sk, ssk); diff --git a/net/mptcp/subflow.c b/net/mptcp/subflow.c index e8325890a322..d90237bf433c 100644 --- a/net/mptcp/subflow.c +++ b/net/mptcp/subflow.c @@ -1660,7 +1660,7 @@ int __mptcp_subflow_connect(struct sock *sk, const struct mptcp_pm_local *local, addrlen = sizeof(struct sockaddr_in6); #endif ssk->sk_bound_dev_if = local->ifindex; - err = kernel_bind(sf, (struct sockaddr *)&addr, addrlen); + err = kernel_bind(sf, (struct sockaddr_unsized *)&addr, addrlen); if (err) { MPTCP_INC_STATS(sock_net(sk), MPTCP_MIB_JOINSYNTXBINDERR); pr_debug("msk=%p local=%d remote=%d bind error: %d\n", diff --git a/net/netfilter/ipvs/ip_vs_sync.c b/net/netfilter/ipvs/ip_vs_sync.c index 3402675bf521..d8c089ef387c 100644 --- a/net/netfilter/ipvs/ip_vs_sync.c +++ b/net/netfilter/ipvs/ip_vs_sync.c @@ -1435,7 +1435,7 @@ static int bind_mcastif_addr(struct socket *sock, struct net_device *dev) sin.sin_addr.s_addr = addr; sin.sin_port = 0; - return kernel_bind(sock, (struct sockaddr *)&sin, sizeof(sin)); + return kernel_bind(sock, (struct sockaddr_unsized *)&sin, sizeof(sin)); } static void get_mcast_sockaddr(union ipvs_sockaddr *sa, int *salen, @@ -1542,7 +1542,7 @@ static int make_receive_sock(struct netns_ipvs *ipvs, int id, get_mcast_sockaddr(&mcast_addr, &salen, &ipvs->bcfg, id); sock->sk->sk_bound_dev_if = dev->ifindex; - result = kernel_bind(sock, (struct sockaddr *)&mcast_addr, salen); + result = kernel_bind(sock, (struct sockaddr_unsized *)&mcast_addr, salen); if (result < 0) { pr_err("Error binding to the multicast addr\n"); goto error; diff --git a/net/netlink/af_netlink.c b/net/netlink/af_netlink.c index 687a84c48882..18490a56edd0 100644 --- a/net/netlink/af_netlink.c +++ b/net/netlink/af_netlink.c @@ -966,7 +966,7 @@ static void netlink_undo_bind(int group, long unsigned int groups, nlk->netlink_unbind(sock_net(sk), undo + 1); } -static int netlink_bind(struct socket *sock, struct sockaddr *addr, +static int netlink_bind(struct socket *sock, struct sockaddr_unsized *addr, int addr_len) { struct sock *sk = sock->sk; diff --git a/net/netrom/af_netrom.c b/net/netrom/af_netrom.c index 3331669d8e33..33468124d53d 100644 --- a/net/netrom/af_netrom.c +++ b/net/netrom/af_netrom.c @@ -561,7 +561,7 @@ static int nr_release(struct socket *sock) return 0; } -static int nr_bind(struct socket *sock, struct sockaddr *uaddr, int addr_len) +static int nr_bind(struct socket *sock, struct sockaddr_unsized *uaddr, int addr_len) { struct sock *sk = sock->sk; struct nr_sock *nr = nr_sk(sk); diff --git a/net/nfc/llcp_sock.c b/net/nfc/llcp_sock.c index 57a2f97004e1..26e6ceb48a82 100644 --- a/net/nfc/llcp_sock.c +++ b/net/nfc/llcp_sock.c @@ -56,7 +56,7 @@ static struct proto llcp_sock_proto = { .obj_size = sizeof(struct nfc_llcp_sock), }; -static int llcp_sock_bind(struct socket *sock, struct sockaddr *addr, int alen) +static int llcp_sock_bind(struct socket *sock, struct sockaddr_unsized *addr, int alen) { struct sock *sk = sock->sk; struct nfc_llcp_sock *llcp_sock = nfc_llcp_sock(sk); @@ -146,7 +146,7 @@ error: return ret; } -static int llcp_raw_sock_bind(struct socket *sock, struct sockaddr *addr, +static int llcp_raw_sock_bind(struct socket *sock, struct sockaddr_unsized *addr, int alen) { struct sock *sk = sock->sk; diff --git a/net/packet/af_packet.c b/net/packet/af_packet.c index 173e6edda08f..fccad2a529cc 100644 --- a/net/packet/af_packet.c +++ b/net/packet/af_packet.c @@ -3279,11 +3279,12 @@ out_unlock: * Bind a packet socket to a device */ -static int packet_bind_spkt(struct socket *sock, struct sockaddr *uaddr, +static int packet_bind_spkt(struct socket *sock, struct sockaddr_unsized *uaddr, int addr_len) { struct sock *sk = sock->sk; - char name[sizeof(uaddr->sa_data_min) + 1]; + struct sockaddr *sa = (struct sockaddr *)uaddr; + char name[sizeof(sa->sa_data_min) + 1]; /* * Check legality @@ -3294,13 +3295,13 @@ static int packet_bind_spkt(struct socket *sock, struct sockaddr *uaddr, /* uaddr->sa_data comes from the userspace, it's not guaranteed to be * zero-terminated. */ - memcpy(name, uaddr->sa_data, sizeof(uaddr->sa_data_min)); - name[sizeof(uaddr->sa_data_min)] = 0; + memcpy(name, sa->sa_data, sizeof(sa->sa_data_min)); + name[sizeof(sa->sa_data_min)] = 0; return packet_do_bind(sk, name, 0, 0); } -static int packet_bind(struct socket *sock, struct sockaddr *uaddr, int addr_len) +static int packet_bind(struct socket *sock, struct sockaddr_unsized *uaddr, int addr_len) { struct sockaddr_ll *sll = (struct sockaddr_ll *)uaddr; struct sock *sk = sock->sk; diff --git a/net/phonet/socket.c b/net/phonet/socket.c index db2d552e9b32..478b02647733 100644 --- a/net/phonet/socket.c +++ b/net/phonet/socket.c @@ -153,7 +153,7 @@ EXPORT_SYMBOL(pn_sock_unhash); static DEFINE_MUTEX(port_mutex); -static int pn_socket_bind(struct socket *sock, struct sockaddr *addr, int len) +static int pn_socket_bind(struct socket *sock, struct sockaddr_unsized *addr, int len) { struct sock *sk = sock->sk; struct pn_sock *pn = pn_sk(sk); @@ -163,7 +163,7 @@ static int pn_socket_bind(struct socket *sock, struct sockaddr *addr, int len) u8 saddr; if (sk->sk_prot->bind) - return sk->sk_prot->bind(sk, addr, len); + return sk->sk_prot->bind(sk, (struct sockaddr *)addr, len); if (len < sizeof(struct sockaddr_pn)) return -EINVAL; @@ -206,8 +206,8 @@ static int pn_socket_autobind(struct socket *sock) memset(&sa, 0, sizeof(sa)); sa.spn_family = AF_PHONET; - err = pn_socket_bind(sock, (struct sockaddr *)&sa, - sizeof(struct sockaddr_pn)); + err = pn_socket_bind(sock, (struct sockaddr_unsized *)&sa, + sizeof(struct sockaddr_pn)); if (err != -EINVAL) return err; BUG_ON(!pn_port(pn_sk(sock->sk)->sobject)); diff --git a/net/qrtr/af_qrtr.c b/net/qrtr/af_qrtr.c index 00c51cf693f3..00bd3dd9f0f9 100644 --- a/net/qrtr/af_qrtr.c +++ b/net/qrtr/af_qrtr.c @@ -824,7 +824,7 @@ static int qrtr_autobind(struct socket *sock) } /* Bind socket to specified sockaddr. */ -static int qrtr_bind(struct socket *sock, struct sockaddr *saddr, int len) +static int qrtr_bind(struct socket *sock, struct sockaddr_unsized *saddr, int len) { DECLARE_SOCKADDR(struct sockaddr_qrtr *, addr, saddr); struct qrtr_sock *ipc = qrtr_sk(sock->sk); diff --git a/net/qrtr/ns.c b/net/qrtr/ns.c index 3de9350cbf30..bfcc1a453f23 100644 --- a/net/qrtr/ns.c +++ b/net/qrtr/ns.c @@ -714,7 +714,7 @@ int qrtr_ns_init(void) sq.sq_port = QRTR_PORT_CTRL; qrtr_ns.local_node = sq.sq_node; - ret = kernel_bind(qrtr_ns.sock, (struct sockaddr *)&sq, sizeof(sq)); + ret = kernel_bind(qrtr_ns.sock, (struct sockaddr_unsized *)&sq, sizeof(sq)); if (ret < 0) { pr_err("failed to bind to socket\n"); goto err_wq; diff --git a/net/rds/bind.c b/net/rds/bind.c index 97a29172a8ee..f800d920d969 100644 --- a/net/rds/bind.c +++ b/net/rds/bind.c @@ -160,7 +160,7 @@ void rds_remove_bound(struct rds_sock *rs) rs->rs_bound_addr = in6addr_any; } -int rds_bind(struct socket *sock, struct sockaddr *uaddr, int addr_len) +int rds_bind(struct socket *sock, struct sockaddr_unsized *uaddr, int addr_len) { struct sock *sk = sock->sk; struct rds_sock *rs = rds_sk_to_rs(sk); diff --git a/net/rds/rds.h b/net/rds/rds.h index 5b1c072e2e7f..a029e5fcdea7 100644 --- a/net/rds/rds.h +++ b/net/rds/rds.h @@ -735,7 +735,7 @@ extern wait_queue_head_t rds_poll_waitq; /* bind.c */ -int rds_bind(struct socket *sock, struct sockaddr *uaddr, int addr_len); +int rds_bind(struct socket *sock, struct sockaddr_unsized *uaddr, int addr_len); void rds_remove_bound(struct rds_sock *rs); struct rds_sock *rds_find_bound(const struct in6_addr *addr, __be16 port, __u32 scope_id); diff --git a/net/rds/tcp_connect.c b/net/rds/tcp_connect.c index a0046e99d6df..1eff3b03ab77 100644 --- a/net/rds/tcp_connect.c +++ b/net/rds/tcp_connect.c @@ -145,7 +145,7 @@ int rds_tcp_conn_path_connect(struct rds_conn_path *cp) addrlen = sizeof(sin); } - ret = kernel_bind(sock, addr, addrlen); + ret = kernel_bind(sock, (struct sockaddr_unsized *)addr, addrlen); if (ret) { rdsdebug("bind failed with %d at address %pI6c\n", ret, &conn->c_laddr); diff --git a/net/rds/tcp_listen.c b/net/rds/tcp_listen.c index 91e34af3fe5d..820d3e20de19 100644 --- a/net/rds/tcp_listen.c +++ b/net/rds/tcp_listen.c @@ -290,7 +290,7 @@ struct socket *rds_tcp_listen_init(struct net *net, bool isv6) addr_len = sizeof(*sin); } - ret = kernel_bind(sock, (struct sockaddr *)&ss, addr_len); + ret = kernel_bind(sock, (struct sockaddr_unsized *)&ss, addr_len); if (ret < 0) { rdsdebug("could not bind %s listener socket: %d\n", isv6 ? "IPv6" : "IPv4", ret); diff --git a/net/rose/af_rose.c b/net/rose/af_rose.c index 543f9e8ebb69..47369eab5aec 100644 --- a/net/rose/af_rose.c +++ b/net/rose/af_rose.c @@ -693,7 +693,7 @@ static int rose_release(struct socket *sock) return 0; } -static int rose_bind(struct socket *sock, struct sockaddr *uaddr, int addr_len) +static int rose_bind(struct socket *sock, struct sockaddr_unsized *uaddr, int addr_len) { struct sock *sk = sock->sk; struct rose_sock *rose = rose_sk(sk); diff --git a/net/rxrpc/af_rxrpc.c b/net/rxrpc/af_rxrpc.c index 36df0274d7b7..245f37a74394 100644 --- a/net/rxrpc/af_rxrpc.c +++ b/net/rxrpc/af_rxrpc.c @@ -127,7 +127,7 @@ static int rxrpc_validate_address(struct rxrpc_sock *rx, /* * bind a local address to an RxRPC socket */ -static int rxrpc_bind(struct socket *sock, struct sockaddr *saddr, int len) +static int rxrpc_bind(struct socket *sock, struct sockaddr_unsized *saddr, int len) { struct sockaddr_rxrpc *srx = (struct sockaddr_rxrpc *)saddr; struct rxrpc_local *local; diff --git a/net/rxrpc/rxperf.c b/net/rxrpc/rxperf.c index 2ea71e3831f7..98ea76fae70f 100644 --- a/net/rxrpc/rxperf.c +++ b/net/rxrpc/rxperf.c @@ -211,7 +211,7 @@ static int rxperf_open_socket(void) ret = rxrpc_sock_set_security_keyring(socket->sk, rxperf_sec_keyring); - ret = kernel_bind(socket, (struct sockaddr *)&srx, sizeof(srx)); + ret = kernel_bind(socket, (struct sockaddr_unsized *)&srx, sizeof(srx)); if (ret < 0) goto error_2; diff --git a/net/smc/af_smc.c b/net/smc/af_smc.c index e9d0e62e0b1b..be18ab08f15d 100644 --- a/net/smc/af_smc.c +++ b/net/smc/af_smc.c @@ -421,7 +421,7 @@ static struct sock *smc_sock_alloc(struct net *net, struct socket *sock, return sk; } -int smc_bind(struct socket *sock, struct sockaddr *uaddr, +int smc_bind(struct socket *sock, struct sockaddr_unsized *uaddr, int addr_len) { struct sockaddr_in *addr = (struct sockaddr_in *)uaddr; diff --git a/net/smc/smc.h b/net/smc/smc.h index 2c9084963739..a008dbe6d6f6 100644 --- a/net/smc/smc.h +++ b/net/smc/smc.h @@ -42,7 +42,7 @@ void smc_unhash_sk(struct sock *sk); void smc_release_cb(struct sock *sk); int smc_release(struct socket *sock); -int smc_bind(struct socket *sock, struct sockaddr *uaddr, +int smc_bind(struct socket *sock, struct sockaddr_unsized *uaddr, int addr_len); int smc_connect(struct socket *sock, struct sockaddr *addr, int alen, int flags); diff --git a/net/socket.c b/net/socket.c index e8892b218708..aaefb2e519a7 100644 --- a/net/socket.c +++ b/net/socket.c @@ -1872,7 +1872,7 @@ int __sys_bind_socket(struct socket *sock, struct sockaddr_storage *address, addrlen); if (!err) err = READ_ONCE(sock->ops)->bind(sock, - (struct sockaddr *)address, + (struct sockaddr_unsized *)address, addrlen); return err; } @@ -3583,13 +3583,13 @@ static long compat_sock_ioctl(struct file *file, unsigned int cmd, * Returns 0 or an error. */ -int kernel_bind(struct socket *sock, struct sockaddr *addr, int addrlen) +int kernel_bind(struct socket *sock, struct sockaddr_unsized *addr, int addrlen) { struct sockaddr_storage address; memcpy(&address, addr, addrlen); - return READ_ONCE(sock->ops)->bind(sock, (struct sockaddr *)&address, + return READ_ONCE(sock->ops)->bind(sock, (struct sockaddr_unsized *)&address, addrlen); } EXPORT_SYMBOL(kernel_bind); diff --git a/net/sunrpc/clnt.c b/net/sunrpc/clnt.c index 8ca354ecfd02..318ee24ad900 100644 --- a/net/sunrpc/clnt.c +++ b/net/sunrpc/clnt.c @@ -1457,12 +1457,12 @@ static int rpc_sockname(struct net *net, struct sockaddr *sap, size_t salen, switch (sap->sa_family) { case AF_INET: err = kernel_bind(sock, - (struct sockaddr *)&rpc_inaddr_loopback, + (struct sockaddr_unsized *)&rpc_inaddr_loopback, sizeof(rpc_inaddr_loopback)); break; case AF_INET6: err = kernel_bind(sock, - (struct sockaddr *)&rpc_in6addr_loopback, + (struct sockaddr_unsized *)&rpc_in6addr_loopback, sizeof(rpc_in6addr_loopback)); break; default: diff --git a/net/sunrpc/svcsock.c b/net/sunrpc/svcsock.c index 7b90abc5cf0e..16ff6c100821 100644 --- a/net/sunrpc/svcsock.c +++ b/net/sunrpc/svcsock.c @@ -1557,7 +1557,7 @@ static struct svc_xprt *svc_create_socket(struct svc_serv *serv, ip6_sock_set_v6only(sock->sk); if (type == SOCK_STREAM) sock->sk->sk_reuse = SK_CAN_REUSE; /* allow address reuse */ - error = kernel_bind(sock, sin, len); + error = kernel_bind(sock, (struct sockaddr_unsized *)sin, len); if (error < 0) goto bummer; diff --git a/net/sunrpc/xprtsock.c b/net/sunrpc/xprtsock.c index 3aa987e7f072..95732a45b059 100644 --- a/net/sunrpc/xprtsock.c +++ b/net/sunrpc/xprtsock.c @@ -1845,8 +1845,8 @@ static int xs_bind(struct sock_xprt *transport, struct socket *sock) memcpy(&myaddr, &transport->srcaddr, transport->xprt.addrlen); do { rpc_set_port((struct sockaddr *)&myaddr, port); - err = kernel_bind(sock, (struct sockaddr *)&myaddr, - transport->xprt.addrlen); + err = kernel_bind(sock, (struct sockaddr_unsized *)&myaddr, + transport->xprt.addrlen); if (err == 0) { if (transport->xprt.reuseport) transport->srcport = port; diff --git a/net/tipc/socket.c b/net/tipc/socket.c index bc614a1f019c..3903a97ada7d 100644 --- a/net/tipc/socket.c +++ b/net/tipc/socket.c @@ -710,7 +710,7 @@ int tipc_sk_bind(struct socket *sock, struct sockaddr *skaddr, int alen) return res; } -static int tipc_bind(struct socket *sock, struct sockaddr *skaddr, int alen) +static int tipc_bind(struct socket *sock, struct sockaddr_unsized *skaddr, int alen) { struct tipc_uaddr *ua = (struct tipc_uaddr *)skaddr; u32 atype = ua->addrtype; @@ -726,7 +726,7 @@ static int tipc_bind(struct socket *sock, struct sockaddr *skaddr, int alen) return -EACCES; } } - return tipc_sk_bind(sock, skaddr, alen); + return tipc_sk_bind(sock, (struct sockaddr *)skaddr, alen); } /** diff --git a/net/unix/af_unix.c b/net/unix/af_unix.c index 54177caa9c12..788775f0eea7 100644 --- a/net/unix/af_unix.c +++ b/net/unix/af_unix.c @@ -843,7 +843,7 @@ out: } static int unix_release(struct socket *); -static int unix_bind(struct socket *, struct sockaddr *, int); +static int unix_bind(struct socket *, struct sockaddr_unsized *, int); static int unix_stream_connect(struct socket *, struct sockaddr *, int addr_len, int flags); static int unix_socketpair(struct socket *, struct socket *); @@ -1466,7 +1466,7 @@ out: return err; } -static int unix_bind(struct socket *sock, struct sockaddr *uaddr, int addr_len) +static int unix_bind(struct socket *sock, struct sockaddr_unsized *uaddr, int addr_len) { struct sockaddr_un *sunaddr = (struct sockaddr_un *)uaddr; struct sock *sk = sock->sk; diff --git a/net/vmw_vsock/af_vsock.c b/net/vmw_vsock/af_vsock.c index 76763247a377..0e5609e7284b 100644 --- a/net/vmw_vsock/af_vsock.c +++ b/net/vmw_vsock/af_vsock.c @@ -987,7 +987,7 @@ static int vsock_release(struct socket *sock) } static int -vsock_bind(struct socket *sock, struct sockaddr *addr, int addr_len) +vsock_bind(struct socket *sock, struct sockaddr_unsized *addr, int addr_len) { int err; struct sock *sk; @@ -995,7 +995,7 @@ vsock_bind(struct socket *sock, struct sockaddr *addr, int addr_len) sk = sock->sk; - if (vsock_addr_cast(addr, addr_len, &vm_addr) != 0) + if (vsock_addr_cast((struct sockaddr *)addr, addr_len, &vm_addr) != 0) return -EINVAL; lock_sock(sk); diff --git a/net/x25/af_x25.c b/net/x25/af_x25.c index 655d1e0ae25f..ca8006d8f792 100644 --- a/net/x25/af_x25.c +++ b/net/x25/af_x25.c @@ -670,7 +670,7 @@ out: return 0; } -static int x25_bind(struct socket *sock, struct sockaddr *uaddr, int addr_len) +static int x25_bind(struct socket *sock, struct sockaddr_unsized *uaddr, int addr_len) { struct sock *sk = sock->sk; struct sockaddr_x25 *addr = (struct sockaddr_x25 *)uaddr; diff --git a/net/xdp/xsk.c b/net/xdp/xsk.c index 2f26c918d448..ed8b612ec29d 100644 --- a/net/xdp/xsk.c +++ b/net/xdp/xsk.c @@ -1238,7 +1238,7 @@ static bool xsk_validate_queues(struct xdp_sock *xs) return xs->fq_tmp && xs->cq_tmp; } -static int xsk_bind(struct socket *sock, struct sockaddr *addr, int addr_len) +static int xsk_bind(struct socket *sock, struct sockaddr_unsized *addr, int addr_len) { struct sockaddr_xdp *sxdp = (struct sockaddr_xdp *)addr; struct sock *sk = sock->sk; diff --git a/tools/testing/selftests/bpf/test_kmods/bpf_testmod.c b/tools/testing/selftests/bpf/test_kmods/bpf_testmod.c index 8074bc5f6f20..0497b5dea25c 100644 --- a/tools/testing/selftests/bpf/test_kmods/bpf_testmod.c +++ b/tools/testing/selftests/bpf/test_kmods/bpf_testmod.c @@ -923,7 +923,7 @@ __bpf_kfunc int bpf_kfunc_call_kernel_bind(struct addr_args *args) goto out; } - err = kernel_bind(sock, (struct sockaddr *)&args->addr, args->addrlen); + err = kernel_bind(sock, (struct sockaddr_unsized *)&args->addr, args->addrlen); out: mutex_unlock(&sock_lock); -- cgit v1.2.3 From 85cb0757d7e1f9370a8b52a8b8144c37941cba0a Mon Sep 17 00:00:00 2001 From: Kees Cook Date: Mon, 3 Nov 2025 16:26:11 -0800 Subject: net: Convert proto_ops connect() callbacks to use sockaddr_unsized Update all struct proto_ops connect() callback function prototypes from "struct sockaddr *" to "struct sockaddr_unsized *" to avoid lying to the compiler about object sizes. Calls into struct proto handlers gain casts that will be removed in the struct proto conversion patch. No binary changes expected. Signed-off-by: Kees Cook Link: https://patch.msgid.link/20251104002617.2752303-3-kees@kernel.org Signed-off-by: Jakub Kicinski --- drivers/block/drbd/drbd_receiver.c | 2 +- drivers/infiniband/hw/erdma/erdma_cm.c | 2 +- drivers/infiniband/sw/siw/siw_cm.c | 2 +- drivers/net/ppp/pppoe.c | 4 ++-- drivers/net/ppp/pptp.c | 4 ++-- drivers/net/wireless/ath/ath10k/qmi.c | 2 +- drivers/net/wireless/ath/ath11k/qmi.c | 2 +- drivers/net/wireless/ath/ath12k/qmi.c | 2 +- drivers/nvme/host/tcp.c | 2 +- drivers/slimbus/qcom-ngd-ctrl.c | 2 +- drivers/xen/pvcalls-back.c | 2 +- fs/coredump.c | 2 +- fs/dlm/lowcomms.c | 2 +- fs/ocfs2/cluster/tcp.c | 2 +- fs/smb/client/connect.c | 2 +- include/linux/bpf-cgroup.h | 6 +++--- include/linux/net.h | 4 ++-- include/net/inet_common.h | 6 +++--- include/net/sctp/sctp.h | 2 +- include/net/sock.h | 2 +- include/net/vsock_addr.h | 2 +- net/9p/trans_fd.c | 6 +++--- net/appletalk/ddp.c | 2 +- net/atm/pvc.c | 4 ++-- net/atm/svc.c | 2 +- net/ax25/af_ax25.c | 2 +- net/bluetooth/iso.c | 2 +- net/bluetooth/l2cap_sock.c | 2 +- net/bluetooth/rfcomm/core.c | 2 +- net/bluetooth/rfcomm/sock.c | 3 ++- net/bluetooth/sco.c | 2 +- net/caif/caif_socket.c | 2 +- net/can/bcm.c | 2 +- net/can/j1939/socket.c | 2 +- net/ceph/messenger.c | 2 +- net/core/sock.c | 2 +- net/ieee802154/socket.c | 4 ++-- net/ipv4/af_inet.c | 14 +++++++------- net/ipv4/tcp.c | 2 +- net/ipv4/udp_tunnel_core.c | 2 +- net/ipv6/ip6_udp_tunnel.c | 2 +- net/iucv/af_iucv.c | 4 ++-- net/l2tp/l2tp_core.c | 4 ++-- net/l2tp/l2tp_ppp.c | 2 +- net/llc/af_llc.c | 2 +- net/mctp/af_mctp.c | 2 +- net/mctp/test/utils.c | 5 +++-- net/mptcp/subflow.c | 2 +- net/netfilter/ipvs/ip_vs_sync.c | 2 +- net/netlink/af_netlink.c | 2 +- net/netrom/af_netrom.c | 4 ++-- net/nfc/llcp_sock.c | 2 +- net/nfc/rawsock.c | 2 +- net/phonet/socket.c | 6 +++--- net/qrtr/af_qrtr.c | 2 +- net/rds/af_rds.c | 2 +- net/rds/tcp_connect.c | 2 +- net/rose/af_rose.c | 3 ++- net/rxrpc/af_rxrpc.c | 2 +- net/sctp/socket.c | 4 ++-- net/smc/af_smc.c | 4 ++-- net/smc/smc.h | 2 +- net/socket.c | 8 ++++---- net/sunrpc/clnt.c | 2 +- net/sunrpc/xprtsock.c | 5 +++-- net/tipc/socket.c | 2 +- net/unix/af_unix.c | 8 ++++---- net/vmw_vsock/af_vsock.c | 6 +++--- net/vmw_vsock/vsock_addr.c | 2 +- net/x25/af_x25.c | 2 +- samples/qmi/qmi_sample_client.c | 2 +- tools/testing/selftests/bpf/test_kmods/bpf_testmod.c | 2 +- 72 files changed, 110 insertions(+), 106 deletions(-) (limited to 'include/net') diff --git a/drivers/block/drbd/drbd_receiver.c b/drivers/block/drbd/drbd_receiver.c index d9296f74f902..33bc91665fe8 100644 --- a/drivers/block/drbd/drbd_receiver.c +++ b/drivers/block/drbd/drbd_receiver.c @@ -458,7 +458,7 @@ static struct socket *drbd_try_connect(struct drbd_connection *connection) * stay C_WF_CONNECTION, don't go Disconnecting! */ disconnect_on_error = 0; what = "connect"; - err = sock->ops->connect(sock, (struct sockaddr *) &peer_in6, peer_addr_len, 0); + err = sock->ops->connect(sock, (struct sockaddr_unsized *) &peer_in6, peer_addr_len, 0); out: if (err < 0) { diff --git a/drivers/infiniband/hw/erdma/erdma_cm.c b/drivers/infiniband/hw/erdma/erdma_cm.c index ef66a6359eb9..ed21ba0037a4 100644 --- a/drivers/infiniband/hw/erdma/erdma_cm.c +++ b/drivers/infiniband/hw/erdma/erdma_cm.c @@ -996,7 +996,7 @@ static int kernel_bindconnect(struct socket *s, struct sockaddr *laddr, ret = s->ops->bind(s, (struct sockaddr_unsized *)laddr, laddrlen); if (ret) return ret; - ret = s->ops->connect(s, raddr, raddrlen, flags); + ret = s->ops->connect(s, (struct sockaddr_unsized *)raddr, raddrlen, flags); return ret < 0 ? ret : 0; } diff --git a/drivers/infiniband/sw/siw/siw_cm.c b/drivers/infiniband/sw/siw/siw_cm.c index 7fe118cacb3f..eb0bd4f79a85 100644 --- a/drivers/infiniband/sw/siw/siw_cm.c +++ b/drivers/infiniband/sw/siw/siw_cm.c @@ -1344,7 +1344,7 @@ static int kernel_bindconnect(struct socket *s, struct sockaddr *laddr, if (rv < 0) return rv; - rv = s->ops->connect(s, raddr, size, flags); + rv = s->ops->connect(s, (struct sockaddr_unsized *)raddr, size, flags); return rv < 0 ? rv : 0; } diff --git a/drivers/net/ppp/pppoe.c b/drivers/net/ppp/pppoe.c index 4ac6afce267b..4275b393a454 100644 --- a/drivers/net/ppp/pppoe.c +++ b/drivers/net/ppp/pppoe.c @@ -608,8 +608,8 @@ static int pppoe_release(struct socket *sock) return 0; } -static int pppoe_connect(struct socket *sock, struct sockaddr *uservaddr, - int sockaddr_len, int flags) +static int pppoe_connect(struct socket *sock, struct sockaddr_unsized *uservaddr, + int sockaddr_len, int flags) { struct sock *sk = sock->sk; struct sockaddr_pppox *sp = (struct sockaddr_pppox *)uservaddr; diff --git a/drivers/net/ppp/pptp.c b/drivers/net/ppp/pptp.c index d07e87a0974c..b18acd810561 100644 --- a/drivers/net/ppp/pptp.c +++ b/drivers/net/ppp/pptp.c @@ -415,8 +415,8 @@ out: return error; } -static int pptp_connect(struct socket *sock, struct sockaddr *uservaddr, - int sockaddr_len, int flags) +static int pptp_connect(struct socket *sock, struct sockaddr_unsized *uservaddr, + int sockaddr_len, int flags) { struct sock *sk = sock->sk; struct sockaddr_pppox *sp = (struct sockaddr_pppox *) uservaddr; diff --git a/drivers/net/wireless/ath/ath10k/qmi.c b/drivers/net/wireless/ath/ath10k/qmi.c index f1f33af0170a..8275345631a0 100644 --- a/drivers/net/wireless/ath/ath10k/qmi.c +++ b/drivers/net/wireless/ath/ath10k/qmi.c @@ -986,7 +986,7 @@ static int ath10k_qmi_new_server(struct qmi_handle *qmi_hdl, ath10k_dbg(ar, ATH10K_DBG_QMI, "wifi fw qmi service found\n"); - ret = kernel_connect(qmi_hdl->sock, (struct sockaddr *)&qmi->sq, + ret = kernel_connect(qmi_hdl->sock, (struct sockaddr_unsized *)&qmi->sq, sizeof(qmi->sq), 0); if (ret) { ath10k_err(ar, "failed to connect to a remote QMI service port\n"); diff --git a/drivers/net/wireless/ath/ath11k/qmi.c b/drivers/net/wireless/ath/ath11k/qmi.c index aea56c38bf8f..ff6a97e328b8 100644 --- a/drivers/net/wireless/ath/ath11k/qmi.c +++ b/drivers/net/wireless/ath/ath11k/qmi.c @@ -3177,7 +3177,7 @@ static int ath11k_qmi_ops_new_server(struct qmi_handle *qmi_hdl, sq->sq_node = service->node; sq->sq_port = service->port; - ret = kernel_connect(qmi_hdl->sock, (struct sockaddr *)sq, + ret = kernel_connect(qmi_hdl->sock, (struct sockaddr_unsized *)sq, sizeof(*sq), 0); if (ret) { ath11k_warn(ab, "failed to connect to qmi remote service: %d\n", ret); diff --git a/drivers/net/wireless/ath/ath12k/qmi.c b/drivers/net/wireless/ath/ath12k/qmi.c index 36325e62aa24..cf9c25df3ffd 100644 --- a/drivers/net/wireless/ath/ath12k/qmi.c +++ b/drivers/net/wireless/ath/ath12k/qmi.c @@ -3740,7 +3740,7 @@ static int ath12k_qmi_ops_new_server(struct qmi_handle *qmi_hdl, sq->sq_node = service->node; sq->sq_port = service->port; - ret = kernel_connect(qmi_hdl->sock, (struct sockaddr *)sq, + ret = kernel_connect(qmi_hdl->sock, (struct sockaddr_unsized *)sq, sizeof(*sq), 0); if (ret) { ath12k_warn(ab, "qmi failed to connect to remote service %d\n", ret); diff --git a/drivers/nvme/host/tcp.c b/drivers/nvme/host/tcp.c index 35d0bd91f6fd..6795b8286c35 100644 --- a/drivers/nvme/host/tcp.c +++ b/drivers/nvme/host/tcp.c @@ -1872,7 +1872,7 @@ static int nvme_tcp_alloc_queue(struct nvme_ctrl *nctrl, int qid, dev_dbg(nctrl->device, "connecting queue %d\n", nvme_tcp_queue_id(queue)); - ret = kernel_connect(queue->sock, (struct sockaddr *)&ctrl->addr, + ret = kernel_connect(queue->sock, (struct sockaddr_unsized *)&ctrl->addr, sizeof(ctrl->addr), 0); if (ret) { dev_err(nctrl->device, diff --git a/drivers/slimbus/qcom-ngd-ctrl.c b/drivers/slimbus/qcom-ngd-ctrl.c index 4fb66986cc22..fdb94dc4a730 100644 --- a/drivers/slimbus/qcom-ngd-ctrl.c +++ b/drivers/slimbus/qcom-ngd-ctrl.c @@ -463,7 +463,7 @@ static int qcom_slim_qmi_init(struct qcom_slim_ngd_ctrl *ctrl, } rc = kernel_connect(handle->sock, - (struct sockaddr *)&ctrl->qmi.svc_info, + (struct sockaddr_unsized *)&ctrl->qmi.svc_info, sizeof(ctrl->qmi.svc_info), 0); if (rc < 0) { dev_err(ctrl->dev, "Remote Service connect failed: %d\n", rc); diff --git a/drivers/xen/pvcalls-back.c b/drivers/xen/pvcalls-back.c index da1b516b9cfd..c5b6f6fa11eb 100644 --- a/drivers/xen/pvcalls-back.c +++ b/drivers/xen/pvcalls-back.c @@ -409,7 +409,7 @@ static int pvcalls_back_connect(struct xenbus_device *dev, ret = sock_create(AF_INET, SOCK_STREAM, 0, &sock); if (ret < 0) goto out; - ret = inet_stream_connect(sock, sa, req->u.connect.len, 0); + ret = inet_stream_connect(sock, (struct sockaddr_unsized *)sa, req->u.connect.len, 0); if (ret < 0) { sock_release(sock); goto out; diff --git a/fs/coredump.c b/fs/coredump.c index 5c1c381ee380..14837d9e2abb 100644 --- a/fs/coredump.c +++ b/fs/coredump.c @@ -708,7 +708,7 @@ static bool coredump_sock_connect(struct core_name *cn, struct coredump_params * */ pidfs_coredump(cprm); - retval = kernel_connect(socket, (struct sockaddr *)(&addr), addr_len, + retval = kernel_connect(socket, (struct sockaddr_unsized *)(&addr), addr_len, O_NONBLOCK | SOCK_COREDUMP); if (retval) { diff --git a/fs/dlm/lowcomms.c b/fs/dlm/lowcomms.c index 0500421b6e3b..f832dafdaca8 100644 --- a/fs/dlm/lowcomms.c +++ b/fs/dlm/lowcomms.c @@ -1599,7 +1599,7 @@ static int dlm_connect(struct connection *con) log_print_ratelimited("connecting to %d", con->nodeid); make_sockaddr(&addr, dlm_config.ci_tcp_port, &addr_len); - result = kernel_connect(sock, (struct sockaddr *)&addr, addr_len, 0); + result = kernel_connect(sock, (struct sockaddr_unsized *)&addr, addr_len, 0); switch (result) { case -EINPROGRESS: /* not an error */ diff --git a/fs/ocfs2/cluster/tcp.c b/fs/ocfs2/cluster/tcp.c index c7734193d8d7..79b281e32f4c 100644 --- a/fs/ocfs2/cluster/tcp.c +++ b/fs/ocfs2/cluster/tcp.c @@ -1638,7 +1638,7 @@ static void o2net_start_connect(struct work_struct *work) remoteaddr.sin_port = node->nd_ipv4_port; ret = sc->sc_sock->ops->connect(sc->sc_sock, - (struct sockaddr *)&remoteaddr, + (struct sockaddr_unsized *)&remoteaddr, sizeof(remoteaddr), O_NONBLOCK); if (ret == -EINPROGRESS) diff --git a/fs/smb/client/connect.c b/fs/smb/client/connect.c index 96d972263020..73120988661a 100644 --- a/fs/smb/client/connect.c +++ b/fs/smb/client/connect.c @@ -3411,7 +3411,7 @@ generic_ip_connect(struct TCP_Server_Info *server) socket->sk->sk_sndbuf, socket->sk->sk_rcvbuf, socket->sk->sk_rcvtimeo); - rc = kernel_connect(socket, saddr, slen, + rc = kernel_connect(socket, (struct sockaddr_unsized *)saddr, slen, server->noblockcnt ? O_NONBLOCK : 0); /* * When mounting SMB root file systems, we do not want to block in diff --git a/include/linux/bpf-cgroup.h b/include/linux/bpf-cgroup.h index aedf573bdb42..a7fb4f46974f 100644 --- a/include/linux/bpf-cgroup.h +++ b/include/linux/bpf-cgroup.h @@ -238,7 +238,7 @@ static inline bool cgroup_bpf_sock_enabled(struct sock *sk, ({ \ int __ret = 0; \ if (cgroup_bpf_enabled(atype)) \ - __ret = __cgroup_bpf_run_filter_sock_addr(sk, uaddr, uaddrlen, \ + __ret = __cgroup_bpf_run_filter_sock_addr(sk, (struct sockaddr *)uaddr, uaddrlen, \ atype, NULL, NULL); \ __ret; \ }) @@ -248,7 +248,7 @@ static inline bool cgroup_bpf_sock_enabled(struct sock *sk, int __ret = 0; \ if (cgroup_bpf_enabled(atype)) { \ lock_sock(sk); \ - __ret = __cgroup_bpf_run_filter_sock_addr(sk, uaddr, uaddrlen, \ + __ret = __cgroup_bpf_run_filter_sock_addr(sk, (struct sockaddr *)uaddr, uaddrlen, \ atype, t_ctx, NULL); \ release_sock(sk); \ } \ @@ -266,7 +266,7 @@ static inline bool cgroup_bpf_sock_enabled(struct sock *sk, int __ret = 0; \ if (cgroup_bpf_enabled(atype)) { \ lock_sock(sk); \ - __ret = __cgroup_bpf_run_filter_sock_addr(sk, uaddr, uaddrlen, \ + __ret = __cgroup_bpf_run_filter_sock_addr(sk, (struct sockaddr *)uaddr, uaddrlen, \ atype, NULL, &__flags); \ release_sock(sk); \ if (__flags & BPF_RET_BIND_NO_CAP_NET_BIND_SERVICE) \ diff --git a/include/linux/net.h b/include/linux/net.h index 0e316f063113..db6bc997ca5b 100644 --- a/include/linux/net.h +++ b/include/linux/net.h @@ -166,7 +166,7 @@ struct proto_ops { struct sockaddr_unsized *myaddr, int sockaddr_len); int (*connect) (struct socket *sock, - struct sockaddr *vaddr, + struct sockaddr_unsized *vaddr, int sockaddr_len, int flags); int (*socketpair)(struct socket *sock1, struct socket *sock2); @@ -348,7 +348,7 @@ int kernel_recvmsg(struct socket *sock, struct msghdr *msg, struct kvec *vec, int kernel_bind(struct socket *sock, struct sockaddr_unsized *addr, int addrlen); int kernel_listen(struct socket *sock, int backlog); int kernel_accept(struct socket *sock, struct socket **newsock, int flags); -int kernel_connect(struct socket *sock, struct sockaddr *addr, int addrlen, +int kernel_connect(struct socket *sock, struct sockaddr_unsized *addr, int addrlen, int flags); int kernel_getsockname(struct socket *sock, struct sockaddr *addr); int kernel_getpeername(struct socket *sock, struct sockaddr *addr); diff --git a/include/net/inet_common.h b/include/net/inet_common.h index 1666cf6f539e..ebafd96912bb 100644 --- a/include/net/inet_common.h +++ b/include/net/inet_common.h @@ -23,11 +23,11 @@ struct sockaddr; struct socket; int inet_release(struct socket *sock); -int inet_stream_connect(struct socket *sock, struct sockaddr *uaddr, +int inet_stream_connect(struct socket *sock, struct sockaddr_unsized *uaddr, int addr_len, int flags); -int __inet_stream_connect(struct socket *sock, struct sockaddr *uaddr, +int __inet_stream_connect(struct socket *sock, struct sockaddr_unsized *uaddr, int addr_len, int flags, int is_sendmsg); -int inet_dgram_connect(struct socket *sock, struct sockaddr *uaddr, +int inet_dgram_connect(struct socket *sock, struct sockaddr_unsized *uaddr, int addr_len, int flags); int inet_accept(struct socket *sock, struct socket *newsock, struct proto_accept_arg *arg); diff --git a/include/net/sctp/sctp.h b/include/net/sctp/sctp.h index bb4b80c12541..58242b37b47a 100644 --- a/include/net/sctp/sctp.h +++ b/include/net/sctp/sctp.h @@ -85,7 +85,7 @@ void sctp_udp_sock_stop(struct net *net); /* * sctp/socket.c */ -int sctp_inet_connect(struct socket *sock, struct sockaddr *uaddr, +int sctp_inet_connect(struct socket *sock, struct sockaddr_unsized *uaddr, int addr_len, int flags); int sctp_backlog_rcv(struct sock *sk, struct sk_buff *skb); int sctp_inet_listen(struct socket *sock, int backlog); diff --git a/include/net/sock.h b/include/net/sock.h index acbb78c96d69..589fbce77217 100644 --- a/include/net/sock.h +++ b/include/net/sock.h @@ -1921,7 +1921,7 @@ int sock_cmsg_send(struct sock *sk, struct msghdr *msg, * does not implement a particular function. */ int sock_no_bind(struct socket *sock, struct sockaddr_unsized *saddr, int len); -int sock_no_connect(struct socket *, struct sockaddr *, int, int); +int sock_no_connect(struct socket *sock, struct sockaddr_unsized *saddr, int len, int flags); int sock_no_socketpair(struct socket *, struct socket *); int sock_no_accept(struct socket *, struct socket *, struct proto_accept_arg *); int sock_no_getname(struct socket *, struct sockaddr *, int); diff --git a/include/net/vsock_addr.h b/include/net/vsock_addr.h index cf8cc140d68d..c3f4cc206198 100644 --- a/include/net/vsock_addr.h +++ b/include/net/vsock_addr.h @@ -16,7 +16,7 @@ bool vsock_addr_bound(const struct sockaddr_vm *addr); void vsock_addr_unbind(struct sockaddr_vm *addr); bool vsock_addr_equals_addr(const struct sockaddr_vm *addr, const struct sockaddr_vm *other); -int vsock_addr_cast(const struct sockaddr *addr, size_t len, +int vsock_addr_cast(const struct sockaddr_unsized *addr, size_t len, struct sockaddr_vm **out_addr); #endif diff --git a/net/9p/trans_fd.c b/net/9p/trans_fd.c index ef517bb307e2..49d674f5e73a 100644 --- a/net/9p/trans_fd.c +++ b/net/9p/trans_fd.c @@ -1018,7 +1018,7 @@ p9_fd_create_tcp(struct p9_client *client, const char *addr, char *args) } err = READ_ONCE(csocket->ops)->connect(csocket, - (struct sockaddr *)&stor, + (struct sockaddr_unsized *)&stor, sizeof(stor), 0); if (err < 0) { pr_err("%s (%d): problem connecting socket to %s\n", @@ -1058,8 +1058,8 @@ p9_fd_create_unix(struct p9_client *client, const char *addr, char *args) return err; } - err = READ_ONCE(csocket->ops)->connect(csocket, (struct sockaddr *)&sun_server, - sizeof(struct sockaddr_un) - 1, 0); + err = READ_ONCE(csocket->ops)->connect(csocket, (struct sockaddr_unsized *)&sun_server, + sizeof(struct sockaddr_un) - 1, 0); if (err < 0) { pr_err("%s (%d): problem connecting socket: %s: %d\n", __func__, task_pid_nr(current), addr, err); diff --git a/net/appletalk/ddp.c b/net/appletalk/ddp.c index 45db43cde67f..2a01fff46c9d 100644 --- a/net/appletalk/ddp.c +++ b/net/appletalk/ddp.c @@ -1204,7 +1204,7 @@ out: } /* Set the address we talk to */ -static int atalk_connect(struct socket *sock, struct sockaddr *uaddr, +static int atalk_connect(struct socket *sock, struct sockaddr_unsized *uaddr, int addr_len, int flags) { struct sock *sk = sock->sk; diff --git a/net/atm/pvc.c b/net/atm/pvc.c index 62fdf07c53de..8f5e76f5dd9e 100644 --- a/net/atm/pvc.c +++ b/net/atm/pvc.c @@ -56,10 +56,10 @@ out: return error; } -static int pvc_connect(struct socket *sock, struct sockaddr *sockaddr, +static int pvc_connect(struct socket *sock, struct sockaddr_unsized *sockaddr, int sockaddr_len, int flags) { - return pvc_bind(sock, (struct sockaddr_unsized *)sockaddr, sockaddr_len); + return pvc_bind(sock, sockaddr, sockaddr_len); } static int pvc_setsockopt(struct socket *sock, int level, int optname, diff --git a/net/atm/svc.c b/net/atm/svc.c index 1906a493c8aa..005964250ecd 100644 --- a/net/atm/svc.c +++ b/net/atm/svc.c @@ -153,7 +153,7 @@ out: return error; } -static int svc_connect(struct socket *sock, struct sockaddr *sockaddr, +static int svc_connect(struct socket *sock, struct sockaddr_unsized *sockaddr, int sockaddr_len, int flags) { DEFINE_WAIT(wait); diff --git a/net/ax25/af_ax25.c b/net/ax25/af_ax25.c index 23c558ff9682..7ebbff2f0020 100644 --- a/net/ax25/af_ax25.c +++ b/net/ax25/af_ax25.c @@ -1175,7 +1175,7 @@ out: * FIXME: nonblock behaviour looks like it may have a bug. */ static int __must_check ax25_connect(struct socket *sock, - struct sockaddr *uaddr, int addr_len, int flags) + struct sockaddr_unsized *uaddr, int addr_len, int flags) { struct sock *sk = sock->sk; ax25_cb *ax25 = sk_to_ax25(sk), *ax25t; diff --git a/net/bluetooth/iso.c b/net/bluetooth/iso.c index 6a7e1b4a8701..243505b89733 100644 --- a/net/bluetooth/iso.c +++ b/net/bluetooth/iso.c @@ -1080,7 +1080,7 @@ done: return err; } -static int iso_sock_connect(struct socket *sock, struct sockaddr *addr, +static int iso_sock_connect(struct socket *sock, struct sockaddr_unsized *addr, int alen, int flags) { struct sockaddr_iso *sa = (struct sockaddr_iso *)addr; diff --git a/net/bluetooth/l2cap_sock.c b/net/bluetooth/l2cap_sock.c index ca7394d8fa4e..9ee189c815d4 100644 --- a/net/bluetooth/l2cap_sock.c +++ b/net/bluetooth/l2cap_sock.c @@ -178,7 +178,7 @@ done: return err; } -static int l2cap_sock_connect(struct socket *sock, struct sockaddr *addr, +static int l2cap_sock_connect(struct socket *sock, struct sockaddr_unsized *addr, int alen, int flags) { struct sock *sk = sock->sk; diff --git a/net/bluetooth/rfcomm/core.c b/net/bluetooth/rfcomm/core.c index d62fd6c57617..57b1dca8141f 100644 --- a/net/bluetooth/rfcomm/core.c +++ b/net/bluetooth/rfcomm/core.c @@ -808,7 +808,7 @@ static struct rfcomm_session *rfcomm_session_create(bdaddr_t *src, addr.l2_psm = cpu_to_le16(L2CAP_PSM_RFCOMM); addr.l2_cid = 0; addr.l2_bdaddr_type = BDADDR_BREDR; - *err = kernel_connect(sock, (struct sockaddr *) &addr, sizeof(addr), O_NONBLOCK); + *err = kernel_connect(sock, (struct sockaddr_unsized *)&addr, sizeof(addr), O_NONBLOCK); if (*err == 0 || *err == -EINPROGRESS) return s; diff --git a/net/bluetooth/rfcomm/sock.c b/net/bluetooth/rfcomm/sock.c index 8c8762bbc6de..be6639cd6f59 100644 --- a/net/bluetooth/rfcomm/sock.c +++ b/net/bluetooth/rfcomm/sock.c @@ -371,7 +371,8 @@ done: return err; } -static int rfcomm_sock_connect(struct socket *sock, struct sockaddr *addr, int alen, int flags) +static int rfcomm_sock_connect(struct socket *sock, struct sockaddr_unsized *addr, + int alen, int flags) { struct sockaddr_rc *sa = (struct sockaddr_rc *) addr; struct sock *sk = sock->sk; diff --git a/net/bluetooth/sco.c b/net/bluetooth/sco.c index 01d878205e58..7afe65e7ff37 100644 --- a/net/bluetooth/sco.c +++ b/net/bluetooth/sco.c @@ -639,7 +639,7 @@ done: return err; } -static int sco_sock_connect(struct socket *sock, struct sockaddr *addr, int alen, int flags) +static int sco_sock_connect(struct socket *sock, struct sockaddr_unsized *addr, int alen, int flags) { struct sockaddr_sco *sa = (struct sockaddr_sco *) addr; struct sock *sk = sock->sk; diff --git a/net/caif/caif_socket.c b/net/caif/caif_socket.c index 039dfbd367c9..af218742af5a 100644 --- a/net/caif/caif_socket.c +++ b/net/caif/caif_socket.c @@ -734,7 +734,7 @@ bad_sol: * o sock->state: holds the SS_* socket state and is updated by connect and * disconnect. */ -static int caif_connect(struct socket *sock, struct sockaddr *uaddr, +static int caif_connect(struct socket *sock, struct sockaddr_unsized *uaddr, int addr_len, int flags) { struct sock *sk = sock->sk; diff --git a/net/can/bcm.c b/net/can/bcm.c index 5e690a2377e4..7eba8ae01a5b 100644 --- a/net/can/bcm.c +++ b/net/can/bcm.c @@ -1657,7 +1657,7 @@ static int bcm_release(struct socket *sock) return 0; } -static int bcm_connect(struct socket *sock, struct sockaddr *uaddr, int len, +static int bcm_connect(struct socket *sock, struct sockaddr_unsized *uaddr, int len, int flags) { struct sockaddr_can *addr = (struct sockaddr_can *)uaddr; diff --git a/net/can/j1939/socket.c b/net/can/j1939/socket.c index a2abedc757d0..6272326dd614 100644 --- a/net/can/j1939/socket.c +++ b/net/can/j1939/socket.c @@ -535,7 +535,7 @@ static int j1939_sk_bind(struct socket *sock, struct sockaddr_unsized *uaddr, in return ret; } -static int j1939_sk_connect(struct socket *sock, struct sockaddr *uaddr, +static int j1939_sk_connect(struct socket *sock, struct sockaddr_unsized *uaddr, int len, int flags) { struct sockaddr_can *addr = (struct sockaddr_can *)uaddr; diff --git a/net/ceph/messenger.c b/net/ceph/messenger.c index f8181acaf870..70b25f4ecba6 100644 --- a/net/ceph/messenger.c +++ b/net/ceph/messenger.c @@ -460,7 +460,7 @@ int ceph_tcp_connect(struct ceph_connection *con) set_sock_callbacks(sock, con); con_sock_state_connecting(con); - ret = kernel_connect(sock, (struct sockaddr *)&ss, sizeof(ss), + ret = kernel_connect(sock, (struct sockaddr_unsized *)&ss, sizeof(ss), O_NONBLOCK); if (ret == -EINPROGRESS) { dout("connect %s EINPROGRESS sk_state = %u\n", diff --git a/net/core/sock.c b/net/core/sock.c index 1e1ce18bba16..f97a0e958991 100644 --- a/net/core/sock.c +++ b/net/core/sock.c @@ -3468,7 +3468,7 @@ int sock_no_bind(struct socket *sock, struct sockaddr_unsized *saddr, int len) } EXPORT_SYMBOL(sock_no_bind); -int sock_no_connect(struct socket *sock, struct sockaddr *saddr, +int sock_no_connect(struct socket *sock, struct sockaddr_unsized *saddr, int len, int flags) { return -EOPNOTSUPP; diff --git a/net/ieee802154/socket.c b/net/ieee802154/socket.c index 99ddfad9bb88..b93fd85f248a 100644 --- a/net/ieee802154/socket.c +++ b/net/ieee802154/socket.c @@ -107,7 +107,7 @@ static int ieee802154_sock_bind(struct socket *sock, struct sockaddr_unsized *ua return sock_no_bind(sock, uaddr, addr_len); } -static int ieee802154_sock_connect(struct socket *sock, struct sockaddr *uaddr, +static int ieee802154_sock_connect(struct socket *sock, struct sockaddr_unsized *uaddr, int addr_len, int flags) { struct sock *sk = sock->sk; @@ -118,7 +118,7 @@ static int ieee802154_sock_connect(struct socket *sock, struct sockaddr *uaddr, if (uaddr->sa_family == AF_UNSPEC) return sk->sk_prot->disconnect(sk, flags); - return sk->sk_prot->connect(sk, uaddr, addr_len); + return sk->sk_prot->connect(sk, (struct sockaddr *)uaddr, addr_len); } static int ieee802154_dev_ioctl(struct sock *sk, struct ifreq __user *arg, diff --git a/net/ipv4/af_inet.c b/net/ipv4/af_inet.c index aa43d16e48ff..0844de9ac6a4 100644 --- a/net/ipv4/af_inet.c +++ b/net/ipv4/af_inet.c @@ -567,7 +567,7 @@ out: return err; } -int inet_dgram_connect(struct socket *sock, struct sockaddr *uaddr, +int inet_dgram_connect(struct socket *sock, struct sockaddr_unsized *uaddr, int addr_len, int flags) { struct sock *sk = sock->sk; @@ -584,14 +584,14 @@ int inet_dgram_connect(struct socket *sock, struct sockaddr *uaddr, return prot->disconnect(sk, flags); if (BPF_CGROUP_PRE_CONNECT_ENABLED(sk)) { - err = prot->pre_connect(sk, uaddr, addr_len); + err = prot->pre_connect(sk, (struct sockaddr *)uaddr, addr_len); if (err) return err; } if (data_race(!inet_sk(sk)->inet_num) && inet_autobind(sk)) return -EAGAIN; - return prot->connect(sk, uaddr, addr_len); + return prot->connect(sk, (struct sockaddr *)uaddr, addr_len); } EXPORT_SYMBOL(inet_dgram_connect); @@ -623,7 +623,7 @@ static long inet_wait_for_connect(struct sock *sk, long timeo, int writebias) * Connect to a remote host. There is regrettably still a little * TCP 'magic' in here. */ -int __inet_stream_connect(struct socket *sock, struct sockaddr *uaddr, +int __inet_stream_connect(struct socket *sock, struct sockaddr_unsized *uaddr, int addr_len, int flags, int is_sendmsg) { struct sock *sk = sock->sk; @@ -671,12 +671,12 @@ int __inet_stream_connect(struct socket *sock, struct sockaddr *uaddr, goto out; if (BPF_CGROUP_PRE_CONNECT_ENABLED(sk)) { - err = sk->sk_prot->pre_connect(sk, uaddr, addr_len); + err = sk->sk_prot->pre_connect(sk, (struct sockaddr *)uaddr, addr_len); if (err) goto out; } - err = sk->sk_prot->connect(sk, uaddr, addr_len); + err = sk->sk_prot->connect(sk, (struct sockaddr *)uaddr, addr_len); if (err < 0) goto out; @@ -741,7 +741,7 @@ sock_error: } EXPORT_SYMBOL(__inet_stream_connect); -int inet_stream_connect(struct socket *sock, struct sockaddr *uaddr, +int inet_stream_connect(struct socket *sock, struct sockaddr_unsized *uaddr, int addr_len, int flags) { int err; diff --git a/net/ipv4/tcp.c b/net/ipv4/tcp.c index a9345aa5a2e5..dee578aad690 100644 --- a/net/ipv4/tcp.c +++ b/net/ipv4/tcp.c @@ -1061,7 +1061,7 @@ int tcp_sendmsg_fastopen(struct sock *sk, struct msghdr *msg, int *copied, } } flags = (msg->msg_flags & MSG_DONTWAIT) ? O_NONBLOCK : 0; - err = __inet_stream_connect(sk->sk_socket, uaddr, + err = __inet_stream_connect(sk->sk_socket, (struct sockaddr_unsized *)uaddr, msg->msg_namelen, flags, 1); /* fastopen_req could already be freed in __inet_stream_connect * if the connection times out or gets rst diff --git a/net/ipv4/udp_tunnel_core.c b/net/ipv4/udp_tunnel_core.c index 11e5a88c923d..b1f667c52cb2 100644 --- a/net/ipv4/udp_tunnel_core.c +++ b/net/ipv4/udp_tunnel_core.c @@ -38,7 +38,7 @@ int udp_sock_create4(struct net *net, struct udp_port_cfg *cfg, udp_addr.sin_family = AF_INET; udp_addr.sin_addr = cfg->peer_ip; udp_addr.sin_port = cfg->peer_udp_port; - err = kernel_connect(sock, (struct sockaddr *)&udp_addr, + err = kernel_connect(sock, (struct sockaddr_unsized *)&udp_addr, sizeof(udp_addr), 0); if (err < 0) goto error; diff --git a/net/ipv6/ip6_udp_tunnel.c b/net/ipv6/ip6_udp_tunnel.c index b0d9286b33c8..cef3e0210744 100644 --- a/net/ipv6/ip6_udp_tunnel.c +++ b/net/ipv6/ip6_udp_tunnel.c @@ -52,7 +52,7 @@ int udp_sock_create6(struct net *net, struct udp_port_cfg *cfg, sizeof(udp6_addr.sin6_addr)); udp6_addr.sin6_port = cfg->peer_udp_port; err = kernel_connect(sock, - (struct sockaddr *)&udp6_addr, + (struct sockaddr_unsized *)&udp6_addr, sizeof(udp6_addr), 0); } if (err < 0) diff --git a/net/iucv/af_iucv.c b/net/iucv/af_iucv.c index 3941e32cda69..a4f1df92417d 100644 --- a/net/iucv/af_iucv.c +++ b/net/iucv/af_iucv.c @@ -668,7 +668,7 @@ static int iucv_sock_autobind(struct sock *sk) return err; } -static int afiucv_path_connect(struct socket *sock, struct sockaddr *addr) +static int afiucv_path_connect(struct socket *sock, struct sockaddr_unsized *addr) { DECLARE_SOCKADDR(struct sockaddr_iucv *, sa, addr); struct sock *sk = sock->sk; @@ -714,7 +714,7 @@ done: } /* Connect an unconnected socket */ -static int iucv_sock_connect(struct socket *sock, struct sockaddr *addr, +static int iucv_sock_connect(struct socket *sock, struct sockaddr_unsized *addr, int alen, int flags) { DECLARE_SOCKADDR(struct sockaddr_iucv *, sa, addr); diff --git a/net/l2tp/l2tp_core.c b/net/l2tp/l2tp_core.c index 4b5e372a5cd4..c4f4a57cd67c 100644 --- a/net/l2tp/l2tp_core.c +++ b/net/l2tp/l2tp_core.c @@ -1513,7 +1513,7 @@ static int l2tp_tunnel_sock_create(struct net *net, sizeof(ip6_addr.l2tp_addr)); ip6_addr.l2tp_conn_id = peer_tunnel_id; err = kernel_connect(sock, - (struct sockaddr *)&ip6_addr, + (struct sockaddr_unsized *)&ip6_addr, sizeof(ip6_addr), 0); if (err < 0) goto out; @@ -1538,7 +1538,7 @@ static int l2tp_tunnel_sock_create(struct net *net, ip_addr.l2tp_family = AF_INET; ip_addr.l2tp_addr = cfg->peer_ip; ip_addr.l2tp_conn_id = peer_tunnel_id; - err = kernel_connect(sock, (struct sockaddr *)&ip_addr, + err = kernel_connect(sock, (struct sockaddr_unsized *)&ip_addr, sizeof(ip_addr), 0); if (err < 0) goto out; diff --git a/net/l2tp/l2tp_ppp.c b/net/l2tp/l2tp_ppp.c index 5e12e7ce17d8..ae4543d5597b 100644 --- a/net/l2tp/l2tp_ppp.c +++ b/net/l2tp/l2tp_ppp.c @@ -684,7 +684,7 @@ static struct l2tp_tunnel *pppol2tp_tunnel_get(struct net *net, /* connect() handler. Attach a PPPoX socket to a tunnel UDP socket */ -static int pppol2tp_connect(struct socket *sock, struct sockaddr *uservaddr, +static int pppol2tp_connect(struct socket *sock, struct sockaddr_unsized *uservaddr, int sockaddr_len, int flags) { struct sock *sk = sock->sk; diff --git a/net/llc/af_llc.c b/net/llc/af_llc.c index e5bb0c0d708c..59d593bb5d18 100644 --- a/net/llc/af_llc.c +++ b/net/llc/af_llc.c @@ -477,7 +477,7 @@ out: * This function will autobind if user did not previously call bind. * Returns: 0 upon success, negative otherwise. */ -static int llc_ui_connect(struct socket *sock, struct sockaddr *uaddr, +static int llc_ui_connect(struct socket *sock, struct sockaddr_unsized *uaddr, int addrlen, int flags) { struct sock *sk = sock->sk; diff --git a/net/mctp/af_mctp.c b/net/mctp/af_mctp.c index 5b1ef50637b7..209a963112e3 100644 --- a/net/mctp/af_mctp.c +++ b/net/mctp/af_mctp.c @@ -128,7 +128,7 @@ out_release: /* Used to set a specific peer prior to bind. Not used for outbound * connections (Tag Owner set) since MCTP is a datagram protocol. */ -static int mctp_connect(struct socket *sock, struct sockaddr *addr, +static int mctp_connect(struct socket *sock, struct sockaddr_unsized *addr, int addrlen, int flags) { struct sock *sk = sock->sk; diff --git a/net/mctp/test/utils.c b/net/mctp/test/utils.c index 953d41902771..35f6be814567 100644 --- a/net/mctp/test/utils.c +++ b/net/mctp/test/utils.c @@ -279,7 +279,7 @@ void mctp_test_bind_run(struct kunit *test, addr.smctp_addr.s_addr = setup->peer_addr; /* connect() type must match bind() type */ addr.smctp_type = setup->bind_type; - rc = kernel_connect(*sock, (struct sockaddr *)&addr, + rc = kernel_connect(*sock, (struct sockaddr_unsized *)&addr, sizeof(addr), 0); KUNIT_EXPECT_EQ(test, rc, 0); } @@ -292,5 +292,6 @@ void mctp_test_bind_run(struct kunit *test, addr.smctp_type = setup->bind_type; *ret_bind_errno = - kernel_bind(*sock, (struct sockaddr *)&addr, sizeof(addr)); + kernel_bind(*sock, (struct sockaddr_unsized *)&addr, + sizeof(addr)); } diff --git a/net/mptcp/subflow.c b/net/mptcp/subflow.c index d90237bf433c..30961b3d1702 100644 --- a/net/mptcp/subflow.c +++ b/net/mptcp/subflow.c @@ -1680,7 +1680,7 @@ int __mptcp_subflow_connect(struct sock *sk, const struct mptcp_pm_local *local, sock_hold(ssk); list_add_tail(&subflow->node, &msk->conn_list); - err = kernel_connect(sf, (struct sockaddr *)&addr, addrlen, O_NONBLOCK); + err = kernel_connect(sf, (struct sockaddr_unsized *)&addr, addrlen, O_NONBLOCK); if (err && err != -EINPROGRESS) { MPTCP_INC_STATS(sock_net(sk), MPTCP_MIB_JOINSYNTXCONNECTERR); pr_debug("msk=%p local=%d remote=%d connect error: %d\n", diff --git a/net/netfilter/ipvs/ip_vs_sync.c b/net/netfilter/ipvs/ip_vs_sync.c index d8c089ef387c..5a0c6f42bd8f 100644 --- a/net/netfilter/ipvs/ip_vs_sync.c +++ b/net/netfilter/ipvs/ip_vs_sync.c @@ -1501,7 +1501,7 @@ static int make_send_sock(struct netns_ipvs *ipvs, int id, } get_mcast_sockaddr(&mcast_addr, &salen, &ipvs->mcfg, id); - result = kernel_connect(sock, (struct sockaddr *)&mcast_addr, + result = kernel_connect(sock, (struct sockaddr_unsized *)&mcast_addr, salen, 0); if (result < 0) { pr_err("Error connecting to the multicast addr\n"); diff --git a/net/netlink/af_netlink.c b/net/netlink/af_netlink.c index 18490a56edd0..8e5151f0c6e4 100644 --- a/net/netlink/af_netlink.c +++ b/net/netlink/af_netlink.c @@ -1054,7 +1054,7 @@ unlock: return err; } -static int netlink_connect(struct socket *sock, struct sockaddr *addr, +static int netlink_connect(struct socket *sock, struct sockaddr_unsized *addr, int alen, int flags) { int err = 0; diff --git a/net/netrom/af_netrom.c b/net/netrom/af_netrom.c index 33468124d53d..5ed1a71ceec1 100644 --- a/net/netrom/af_netrom.c +++ b/net/netrom/af_netrom.c @@ -632,8 +632,8 @@ static int nr_bind(struct socket *sock, struct sockaddr_unsized *uaddr, int addr return 0; } -static int nr_connect(struct socket *sock, struct sockaddr *uaddr, - int addr_len, int flags) +static int nr_connect(struct socket *sock, struct sockaddr_unsized *uaddr, + int addr_len, int flags) { struct sock *sk = sock->sk; struct nr_sock *nr = nr_sk(sk); diff --git a/net/nfc/llcp_sock.c b/net/nfc/llcp_sock.c index 26e6ceb48a82..f1be1e84f665 100644 --- a/net/nfc/llcp_sock.c +++ b/net/nfc/llcp_sock.c @@ -648,7 +648,7 @@ out: return err; } -static int llcp_sock_connect(struct socket *sock, struct sockaddr *_addr, +static int llcp_sock_connect(struct socket *sock, struct sockaddr_unsized *_addr, int len, int flags) { struct sock *sk = sock->sk; diff --git a/net/nfc/rawsock.c b/net/nfc/rawsock.c index 5125392bb68e..b049022399ae 100644 --- a/net/nfc/rawsock.c +++ b/net/nfc/rawsock.c @@ -73,7 +73,7 @@ static int rawsock_release(struct socket *sock) return 0; } -static int rawsock_connect(struct socket *sock, struct sockaddr *_addr, +static int rawsock_connect(struct socket *sock, struct sockaddr_unsized *_addr, int len, int flags) { struct sock *sk = sock->sk; diff --git a/net/phonet/socket.c b/net/phonet/socket.c index 478b02647733..9391378083a4 100644 --- a/net/phonet/socket.c +++ b/net/phonet/socket.c @@ -214,8 +214,8 @@ static int pn_socket_autobind(struct socket *sock) return 0; /* socket was already bound */ } -static int pn_socket_connect(struct socket *sock, struct sockaddr *addr, - int len, int flags) +static int pn_socket_connect(struct socket *sock, struct sockaddr_unsized *addr, + int len, int flags) { struct sock *sk = sock->sk; struct pn_sock *pn = pn_sk(sk); @@ -252,7 +252,7 @@ static int pn_socket_connect(struct socket *sock, struct sockaddr *addr, pn->resource = pn_sockaddr_get_resource(spn); sock->state = SS_CONNECTING; - err = sk->sk_prot->connect(sk, addr, len); + err = sk->sk_prot->connect(sk, (struct sockaddr *)addr, len); if (err) { sock->state = SS_UNCONNECTED; pn->dobject = 0; diff --git a/net/qrtr/af_qrtr.c b/net/qrtr/af_qrtr.c index 00bd3dd9f0f9..dab839f61ee9 100644 --- a/net/qrtr/af_qrtr.c +++ b/net/qrtr/af_qrtr.c @@ -1084,7 +1084,7 @@ out: return rc; } -static int qrtr_connect(struct socket *sock, struct sockaddr *saddr, +static int qrtr_connect(struct socket *sock, struct sockaddr_unsized *saddr, int len, int flags) { DECLARE_SOCKADDR(struct sockaddr_qrtr *, addr, saddr); diff --git a/net/rds/af_rds.c b/net/rds/af_rds.c index 4a7217fbeab6..b396c673dfaf 100644 --- a/net/rds/af_rds.c +++ b/net/rds/af_rds.c @@ -533,7 +533,7 @@ out: } -static int rds_connect(struct socket *sock, struct sockaddr *uaddr, +static int rds_connect(struct socket *sock, struct sockaddr_unsized *uaddr, int addr_len, int flags) { struct sock *sk = sock->sk; diff --git a/net/rds/tcp_connect.c b/net/rds/tcp_connect.c index 1eff3b03ab77..92891b0d224d 100644 --- a/net/rds/tcp_connect.c +++ b/net/rds/tcp_connect.c @@ -173,7 +173,7 @@ int rds_tcp_conn_path_connect(struct rds_conn_path *cp) * own the socket */ rds_tcp_set_callbacks(sock, cp); - ret = kernel_connect(sock, addr, addrlen, O_NONBLOCK); + ret = kernel_connect(sock, (struct sockaddr_unsized *)addr, addrlen, O_NONBLOCK); rdsdebug("connect to address %pI6c returned %d\n", &conn->c_faddr, ret); if (ret == -EINPROGRESS) diff --git a/net/rose/af_rose.c b/net/rose/af_rose.c index 47369eab5aec..fd67494f2815 100644 --- a/net/rose/af_rose.c +++ b/net/rose/af_rose.c @@ -765,7 +765,8 @@ out_release: return err; } -static int rose_connect(struct socket *sock, struct sockaddr *uaddr, int addr_len, int flags) +static int rose_connect(struct socket *sock, struct sockaddr_unsized *uaddr, int addr_len, + int flags) { struct sock *sk = sock->sk; struct rose_sock *rose = rose_sk(sk); diff --git a/net/rxrpc/af_rxrpc.c b/net/rxrpc/af_rxrpc.c index 245f37a74394..0c2c68c4b07e 100644 --- a/net/rxrpc/af_rxrpc.c +++ b/net/rxrpc/af_rxrpc.c @@ -481,7 +481,7 @@ EXPORT_SYMBOL(rxrpc_kernel_set_notifications); * - this just targets it at a specific destination; no actual connection * negotiation takes place */ -static int rxrpc_connect(struct socket *sock, struct sockaddr *addr, +static int rxrpc_connect(struct socket *sock, struct sockaddr_unsized *addr, int addr_len, int flags) { struct sockaddr_rxrpc *srx = (struct sockaddr_rxrpc *)addr; diff --git a/net/sctp/socket.c b/net/sctp/socket.c index ac737e60829b..940abbced191 100644 --- a/net/sctp/socket.c +++ b/net/sctp/socket.c @@ -4820,7 +4820,7 @@ static int sctp_connect(struct sock *sk, struct sockaddr *addr, return err; } -int sctp_inet_connect(struct socket *sock, struct sockaddr *uaddr, +int sctp_inet_connect(struct socket *sock, struct sockaddr_unsized *uaddr, int addr_len, int flags) { if (addr_len < sizeof(uaddr->sa_family)) @@ -4829,7 +4829,7 @@ int sctp_inet_connect(struct socket *sock, struct sockaddr *uaddr, if (uaddr->sa_family == AF_UNSPEC) return -EOPNOTSUPP; - return sctp_connect(sock->sk, uaddr, addr_len, flags); + return sctp_connect(sock->sk, (struct sockaddr *)uaddr, addr_len, flags); } /* Only called when shutdown a listening SCTP socket. */ diff --git a/net/smc/af_smc.c b/net/smc/af_smc.c index be18ab08f15d..0ef3e16a8517 100644 --- a/net/smc/af_smc.c +++ b/net/smc/af_smc.c @@ -1642,7 +1642,7 @@ out: release_sock(&smc->sk); } -int smc_connect(struct socket *sock, struct sockaddr *addr, +int smc_connect(struct socket *sock, struct sockaddr_unsized *addr, int alen, int flags) { struct sock *sk = sock->sk; @@ -1694,7 +1694,7 @@ int smc_connect(struct socket *sock, struct sockaddr *addr, rc = -EALREADY; goto out; } - rc = kernel_connect(smc->clcsock, addr, alen, flags); + rc = kernel_connect(smc->clcsock, (struct sockaddr_unsized *)addr, alen, flags); if (rc && rc != -EINPROGRESS) goto out; diff --git a/net/smc/smc.h b/net/smc/smc.h index a008dbe6d6f6..9e6af72784ba 100644 --- a/net/smc/smc.h +++ b/net/smc/smc.h @@ -44,7 +44,7 @@ void smc_release_cb(struct sock *sk); int smc_release(struct socket *sock); int smc_bind(struct socket *sock, struct sockaddr_unsized *uaddr, int addr_len); -int smc_connect(struct socket *sock, struct sockaddr *addr, +int smc_connect(struct socket *sock, struct sockaddr_unsized *addr, int alen, int flags); int smc_accept(struct socket *sock, struct socket *new_sock, struct proto_accept_arg *arg); diff --git a/net/socket.c b/net/socket.c index aaefb2e519a7..101a7ed574e7 100644 --- a/net/socket.c +++ b/net/socket.c @@ -2099,8 +2099,8 @@ int __sys_connect_file(struct file *file, struct sockaddr_storage *address, if (err) goto out; - err = READ_ONCE(sock->ops)->connect(sock, (struct sockaddr *)address, - addrlen, sock->file->f_flags | file_flags); + err = READ_ONCE(sock->ops)->connect(sock, (struct sockaddr_unsized *)address, + addrlen, sock->file->f_flags | file_flags); out: return err; } @@ -3662,14 +3662,14 @@ EXPORT_SYMBOL(kernel_accept); * Returns 0 or an error code. */ -int kernel_connect(struct socket *sock, struct sockaddr *addr, int addrlen, +int kernel_connect(struct socket *sock, struct sockaddr_unsized *addr, int addrlen, int flags) { struct sockaddr_storage address; memcpy(&address, addr, addrlen); - return READ_ONCE(sock->ops)->connect(sock, (struct sockaddr *)&address, + return READ_ONCE(sock->ops)->connect(sock, (struct sockaddr_unsized *)&address, addrlen, flags); } EXPORT_SYMBOL(kernel_connect); diff --git a/net/sunrpc/clnt.c b/net/sunrpc/clnt.c index 318ee24ad900..58442ae1c2da 100644 --- a/net/sunrpc/clnt.c +++ b/net/sunrpc/clnt.c @@ -1474,7 +1474,7 @@ static int rpc_sockname(struct net *net, struct sockaddr *sap, size_t salen, goto out_release; } - err = kernel_connect(sock, sap, salen, 0); + err = kernel_connect(sock, (struct sockaddr_unsized *)sap, salen, 0); if (err < 0) { dprintk("RPC: can't connect UDP socket (%d)\n", err); goto out_release; diff --git a/net/sunrpc/xprtsock.c b/net/sunrpc/xprtsock.c index 95732a45b059..2e1fe6013361 100644 --- a/net/sunrpc/xprtsock.c +++ b/net/sunrpc/xprtsock.c @@ -2005,7 +2005,7 @@ static int xs_local_finish_connecting(struct rpc_xprt *xprt, xs_stream_start_connect(transport); - return kernel_connect(sock, xs_addr(xprt), xprt->addrlen, 0); + return kernel_connect(sock, (struct sockaddr_unsized *)xs_addr(xprt), xprt->addrlen, 0); } /** @@ -2405,7 +2405,8 @@ static int xs_tcp_finish_connecting(struct rpc_xprt *xprt, struct socket *sock) /* Tell the socket layer to start connecting... */ set_bit(XPRT_SOCK_CONNECTING, &transport->sock_state); - return kernel_connect(sock, xs_addr(xprt), xprt->addrlen, O_NONBLOCK); + return kernel_connect(sock, (struct sockaddr_unsized *)xs_addr(xprt), + xprt->addrlen, O_NONBLOCK); } /** diff --git a/net/tipc/socket.c b/net/tipc/socket.c index 3903a97ada7d..817b07d95a91 100644 --- a/net/tipc/socket.c +++ b/net/tipc/socket.c @@ -2565,7 +2565,7 @@ static bool tipc_sockaddr_is_sane(struct sockaddr_tipc *addr) * * Return: 0 on success, errno otherwise */ -static int tipc_connect(struct socket *sock, struct sockaddr *dest, +static int tipc_connect(struct socket *sock, struct sockaddr_unsized *dest, int destlen, int flags) { struct sock *sk = sock->sk; diff --git a/net/unix/af_unix.c b/net/unix/af_unix.c index 788775f0eea7..3b44cadaed96 100644 --- a/net/unix/af_unix.c +++ b/net/unix/af_unix.c @@ -844,7 +844,7 @@ out: static int unix_release(struct socket *); static int unix_bind(struct socket *, struct sockaddr_unsized *, int); -static int unix_stream_connect(struct socket *, struct sockaddr *, +static int unix_stream_connect(struct socket *, struct sockaddr_unsized *, int addr_len, int flags); static int unix_socketpair(struct socket *, struct socket *); static int unix_accept(struct socket *, struct socket *, struct proto_accept_arg *arg); @@ -866,7 +866,7 @@ static int unix_dgram_sendmsg(struct socket *, struct msghdr *, size_t); static int unix_dgram_recvmsg(struct socket *, struct msghdr *, size_t, int); static int unix_read_skb(struct sock *sk, skb_read_actor_t recv_actor); static int unix_stream_read_skb(struct sock *sk, skb_read_actor_t recv_actor); -static int unix_dgram_connect(struct socket *, struct sockaddr *, +static int unix_dgram_connect(struct socket *, struct sockaddr_unsized *, int, int); static int unix_seqpacket_sendmsg(struct socket *, struct msghdr *, size_t); static int unix_seqpacket_recvmsg(struct socket *, struct msghdr *, size_t, @@ -1512,7 +1512,7 @@ static void unix_state_double_unlock(struct sock *sk1, struct sock *sk2) unix_state_unlock(sk2); } -static int unix_dgram_connect(struct socket *sock, struct sockaddr *addr, +static int unix_dgram_connect(struct socket *sock, struct sockaddr_unsized *addr, int alen, int flags) { struct sockaddr_un *sunaddr = (struct sockaddr_un *)addr; @@ -1631,7 +1631,7 @@ static long unix_wait_for_peer(struct sock *other, long timeo) return timeo; } -static int unix_stream_connect(struct socket *sock, struct sockaddr *uaddr, +static int unix_stream_connect(struct socket *sock, struct sockaddr_unsized *uaddr, int addr_len, int flags) { struct sockaddr_un *sunaddr = (struct sockaddr_un *)uaddr; diff --git a/net/vmw_vsock/af_vsock.c b/net/vmw_vsock/af_vsock.c index 0e5609e7284b..72bb6b7ed386 100644 --- a/net/vmw_vsock/af_vsock.c +++ b/net/vmw_vsock/af_vsock.c @@ -995,7 +995,7 @@ vsock_bind(struct socket *sock, struct sockaddr_unsized *addr, int addr_len) sk = sock->sk; - if (vsock_addr_cast((struct sockaddr *)addr, addr_len, &vm_addr) != 0) + if (vsock_addr_cast(addr, addr_len, &vm_addr) != 0) return -EINVAL; lock_sock(sk); @@ -1328,7 +1328,7 @@ out: } static int vsock_dgram_connect(struct socket *sock, - struct sockaddr *addr, int addr_len, int flags) + struct sockaddr_unsized *addr, int addr_len, int flags) { int err; struct sock *sk; @@ -1528,7 +1528,7 @@ static void vsock_connect_timeout(struct work_struct *work) sock_put(sk); } -static int vsock_connect(struct socket *sock, struct sockaddr *addr, +static int vsock_connect(struct socket *sock, struct sockaddr_unsized *addr, int addr_len, int flags) { int err; diff --git a/net/vmw_vsock/vsock_addr.c b/net/vmw_vsock/vsock_addr.c index 223b9660a759..a986aa6fff9b 100644 --- a/net/vmw_vsock/vsock_addr.c +++ b/net/vmw_vsock/vsock_addr.c @@ -57,7 +57,7 @@ bool vsock_addr_equals_addr(const struct sockaddr_vm *addr, } EXPORT_SYMBOL_GPL(vsock_addr_equals_addr); -int vsock_addr_cast(const struct sockaddr *addr, +int vsock_addr_cast(const struct sockaddr_unsized *addr, size_t len, struct sockaddr_vm **out_addr) { if (len < sizeof(**out_addr)) diff --git a/net/x25/af_x25.c b/net/x25/af_x25.c index ca8006d8f792..af8762b24039 100644 --- a/net/x25/af_x25.c +++ b/net/x25/af_x25.c @@ -743,7 +743,7 @@ static int x25_wait_for_connection_establishment(struct sock *sk) return rc; } -static int x25_connect(struct socket *sock, struct sockaddr *uaddr, +static int x25_connect(struct socket *sock, struct sockaddr_unsized *uaddr, int addr_len, int flags) { struct sock *sk = sock->sk; diff --git a/samples/qmi/qmi_sample_client.c b/samples/qmi/qmi_sample_client.c index b27d861f354f..d1814582319b 100644 --- a/samples/qmi/qmi_sample_client.c +++ b/samples/qmi/qmi_sample_client.c @@ -468,7 +468,7 @@ static int qmi_sample_probe(struct platform_device *pdev) return ret; sq = dev_get_platdata(&pdev->dev); - ret = kernel_connect(sample->qmi.sock, (struct sockaddr *)sq, + ret = kernel_connect(sample->qmi.sock, (struct sockaddr_unsized *)sq, sizeof(*sq), 0); if (ret < 0) { pr_err("failed to connect to remote service port\n"); diff --git a/tools/testing/selftests/bpf/test_kmods/bpf_testmod.c b/tools/testing/selftests/bpf/test_kmods/bpf_testmod.c index 0497b5dea25c..8eeebaa951f0 100644 --- a/tools/testing/selftests/bpf/test_kmods/bpf_testmod.c +++ b/tools/testing/selftests/bpf/test_kmods/bpf_testmod.c @@ -900,7 +900,7 @@ __bpf_kfunc int bpf_kfunc_call_kernel_connect(struct addr_args *args) goto out; } - err = kernel_connect(sock, (struct sockaddr *)&args->addr, + err = kernel_connect(sock, (struct sockaddr_unsized *)&args->addr, args->addrlen, 0); out: mutex_unlock(&sock_lock); -- cgit v1.2.3 From 449f68f8fffa2c41fc265730bd05a3c4947916c1 Mon Sep 17 00:00:00 2001 From: Kees Cook Date: Mon, 3 Nov 2025 16:26:13 -0800 Subject: net: Convert proto callbacks from sockaddr to sockaddr_unsized Convert struct proto pre_connect(), connect(), bind(), and bind_add() callback function prototypes from struct sockaddr to struct sockaddr_unsized. This does not change per-implementation use of sockaddr for passing around an arbitrarily sized sockaddr struct. Those will be addressed in future patches. Additionally removes the no longer referenced struct sockaddr from include/net/inet_common.h. No binary changes expected. Signed-off-by: Kees Cook Link: https://patch.msgid.link/20251104002617.2752303-5-kees@kernel.org Signed-off-by: Jakub Kicinski --- fs/dlm/lowcomms.c | 4 ++-- include/net/inet_common.h | 5 ++--- include/net/ip.h | 4 ++-- include/net/ipv6.h | 8 ++++---- include/net/ipv6_stubs.h | 2 +- include/net/ping.h | 2 +- include/net/sock.h | 10 +++++----- include/net/tcp.h | 2 +- include/net/udp.h | 2 +- net/core/filter.c | 5 +++-- net/core/sock.c | 2 +- net/ieee802154/socket.c | 12 ++++++------ net/ipv4/af_inet.c | 14 +++++++------- net/ipv4/datagram.c | 4 ++-- net/ipv4/ping.c | 8 ++++---- net/ipv4/raw.c | 3 ++- net/ipv4/tcp_ipv4.c | 4 ++-- net/ipv4/udp.c | 6 ++++-- net/ipv6/af_inet6.c | 6 +++--- net/ipv6/datagram.c | 8 ++++---- net/ipv6/ping.c | 2 +- net/ipv6/raw.c | 3 ++- net/ipv6/tcp_ipv6.c | 6 +++--- net/ipv6/udp.c | 5 +++-- net/l2tp/l2tp_ip.c | 6 ++++-- net/l2tp/l2tp_ip6.c | 5 +++-- net/mptcp/pm_kernel.c | 4 ++-- net/mptcp/protocol.c | 7 ++++--- net/phonet/pep.c | 3 ++- net/phonet/socket.c | 4 ++-- net/sctp/socket.c | 9 +++++---- 31 files changed, 88 insertions(+), 77 deletions(-) (limited to 'include/net') diff --git a/fs/dlm/lowcomms.c b/fs/dlm/lowcomms.c index f832dafdaca8..b3958008ba3f 100644 --- a/fs/dlm/lowcomms.c +++ b/fs/dlm/lowcomms.c @@ -1126,7 +1126,7 @@ static void writequeue_entry_complete(struct writequeue_entry *e, int completed) static int sctp_bind_addrs(struct socket *sock, __be16 port) { struct sockaddr_storage localaddr; - struct sockaddr *addr = (struct sockaddr *)&localaddr; + struct sockaddr_unsized *addr = (struct sockaddr_unsized *)&localaddr; int i, addr_len, result = 0; for (i = 0; i < dlm_local_count; i++) { @@ -1134,7 +1134,7 @@ static int sctp_bind_addrs(struct socket *sock, __be16 port) make_sockaddr(&localaddr, port, &addr_len); if (!i) - result = kernel_bind(sock, (struct sockaddr_unsized *)addr, addr_len); + result = kernel_bind(sock, addr, addr_len); else result = sock_bind_add(sock->sk, addr, addr_len); diff --git a/include/net/inet_common.h b/include/net/inet_common.h index ebafd96912bb..5dd2bf24449e 100644 --- a/include/net/inet_common.h +++ b/include/net/inet_common.h @@ -19,7 +19,6 @@ struct msghdr; struct net; struct page; struct sock; -struct sockaddr; struct socket; int inet_release(struct socket *sock); @@ -43,7 +42,7 @@ int inet_listen(struct socket *sock, int backlog); int __inet_listen_sk(struct sock *sk, int backlog); void inet_sock_destruct(struct sock *sk); int inet_bind(struct socket *sock, struct sockaddr_unsized *uaddr, int addr_len); -int inet_bind_sk(struct sock *sk, struct sockaddr *uaddr, int addr_len); +int inet_bind_sk(struct sock *sk, struct sockaddr_unsized *uaddr, int addr_len); /* Don't allocate port at this moment, defer to connect. */ #define BIND_FORCE_ADDRESS_NO_PORT (1 << 0) /* Grab and release socket lock. */ @@ -52,7 +51,7 @@ int inet_bind_sk(struct sock *sk, struct sockaddr *uaddr, int addr_len); #define BIND_FROM_BPF (1 << 2) /* Skip CAP_NET_BIND_SERVICE check. */ #define BIND_NO_CAP_NET_BIND_SERVICE (1 << 3) -int __inet_bind(struct sock *sk, struct sockaddr *uaddr, int addr_len, +int __inet_bind(struct sock *sk, struct sockaddr_unsized *uaddr, int addr_len, u32 flags); int inet_getname(struct socket *sock, struct sockaddr *uaddr, int peer); diff --git a/include/net/ip.h b/include/net/ip.h index 380afb691c41..69d5cef46004 100644 --- a/include/net/ip.h +++ b/include/net/ip.h @@ -261,8 +261,8 @@ static inline u8 ip_sendmsg_scope(const struct inet_sock *inet, } /* datagram.c */ -int __ip4_datagram_connect(struct sock *sk, struct sockaddr *uaddr, int addr_len); -int ip4_datagram_connect(struct sock *sk, struct sockaddr *uaddr, int addr_len); +int __ip4_datagram_connect(struct sock *sk, struct sockaddr_unsized *uaddr, int addr_len); +int ip4_datagram_connect(struct sock *sk, struct sockaddr_unsized *uaddr, int addr_len); void ip4_datagram_release_cb(struct sock *sk); diff --git a/include/net/ipv6.h b/include/net/ipv6.h index 2188bad9a687..74fbf1ad8065 100644 --- a/include/net/ipv6.h +++ b/include/net/ipv6.h @@ -1188,10 +1188,10 @@ int do_ipv6_getsockopt(struct sock *sk, int level, int optname, int ipv6_getsockopt(struct sock *sk, int level, int optname, char __user *optval, int __user *optlen); -int __ip6_datagram_connect(struct sock *sk, struct sockaddr *addr, +int __ip6_datagram_connect(struct sock *sk, struct sockaddr_unsized *addr, int addr_len); -int ip6_datagram_connect(struct sock *sk, struct sockaddr *addr, int addr_len); -int ip6_datagram_connect_v6_only(struct sock *sk, struct sockaddr *addr, +int ip6_datagram_connect(struct sock *sk, struct sockaddr_unsized *addr, int addr_len); +int ip6_datagram_connect_v6_only(struct sock *sk, struct sockaddr_unsized *addr, int addr_len); int ip6_datagram_dst_update(struct sock *sk, bool fix_sk_saddr); void ip6_datagram_release_cb(struct sock *sk); @@ -1209,7 +1209,7 @@ void inet6_cleanup_sock(struct sock *sk); void inet6_sock_destruct(struct sock *sk); int inet6_release(struct socket *sock); int inet6_bind(struct socket *sock, struct sockaddr_unsized *uaddr, int addr_len); -int inet6_bind_sk(struct sock *sk, struct sockaddr *uaddr, int addr_len); +int inet6_bind_sk(struct sock *sk, struct sockaddr_unsized *uaddr, int addr_len); int inet6_getname(struct socket *sock, struct sockaddr *uaddr, int peer); int inet6_ioctl(struct socket *sock, unsigned int cmd, unsigned long arg); diff --git a/include/net/ipv6_stubs.h b/include/net/ipv6_stubs.h index 8a3465c8c2c5..d3013e721b14 100644 --- a/include/net/ipv6_stubs.h +++ b/include/net/ipv6_stubs.h @@ -80,7 +80,7 @@ extern const struct ipv6_stub *ipv6_stub __read_mostly; /* A stub used by bpf helpers. Similarly ugly as ipv6_stub */ struct ipv6_bpf_stub { - int (*inet6_bind)(struct sock *sk, struct sockaddr *uaddr, int addr_len, + int (*inet6_bind)(struct sock *sk, struct sockaddr_unsized *uaddr, int addr_len, u32 flags); struct sock *(*udp6_lib_lookup)(const struct net *net, const struct in6_addr *saddr, __be16 sport, diff --git a/include/net/ping.h b/include/net/ping.h index 9634b8800814..05bfd594a64c 100644 --- a/include/net/ping.h +++ b/include/net/ping.h @@ -58,7 +58,7 @@ void ping_unhash(struct sock *sk); int ping_init_sock(struct sock *sk); void ping_close(struct sock *sk, long timeout); -int ping_bind(struct sock *sk, struct sockaddr *uaddr, int addr_len); +int ping_bind(struct sock *sk, struct sockaddr_unsized *uaddr, int addr_len); void ping_err(struct sk_buff *skb, int offset, u32 info); int ping_getfrag(void *from, char *to, int offset, int fraglen, int odd, struct sk_buff *); diff --git a/include/net/sock.h b/include/net/sock.h index 589fbce77217..a5f36ea9d46f 100644 --- a/include/net/sock.h +++ b/include/net/sock.h @@ -1274,10 +1274,10 @@ struct proto { void (*close)(struct sock *sk, long timeout); int (*pre_connect)(struct sock *sk, - struct sockaddr *uaddr, + struct sockaddr_unsized *uaddr, int addr_len); int (*connect)(struct sock *sk, - struct sockaddr *uaddr, + struct sockaddr_unsized *uaddr, int addr_len); int (*disconnect)(struct sock *sk, int flags); @@ -1306,9 +1306,9 @@ struct proto { size_t len, int flags, int *addr_len); void (*splice_eof)(struct socket *sock); int (*bind)(struct sock *sk, - struct sockaddr *addr, int addr_len); + struct sockaddr_unsized *addr, int addr_len); int (*bind_add)(struct sock *sk, - struct sockaddr *addr, int addr_len); + struct sockaddr_unsized *addr, int addr_len); int (*backlog_rcv) (struct sock *sk, struct sk_buff *skb); @@ -3105,7 +3105,7 @@ void sock_set_reuseaddr(struct sock *sk); void sock_set_reuseport(struct sock *sk); void sock_set_sndtimeo(struct sock *sk, s64 secs); -int sock_bind_add(struct sock *sk, struct sockaddr *addr, int addr_len); +int sock_bind_add(struct sock *sk, struct sockaddr_unsized *addr, int addr_len); int sock_get_timeout(long timeo, void *optval, bool old_timeval); int sock_copy_user_timeval(struct __kernel_sock_timeval *tv, diff --git a/include/net/tcp.h b/include/net/tcp.h index 4fd6d8d1230d..0aa1f07d036a 100644 --- a/include/net/tcp.h +++ b/include/net/tcp.h @@ -535,7 +535,7 @@ struct sock *tcp_v4_syn_recv_sock(const struct sock *sk, struct sk_buff *skb, struct request_sock *req_unhash, bool *own_req); int tcp_v4_do_rcv(struct sock *sk, struct sk_buff *skb); -int tcp_v4_connect(struct sock *sk, struct sockaddr *uaddr, int addr_len); +int tcp_v4_connect(struct sock *sk, struct sockaddr_unsized *uaddr, int addr_len); int tcp_connect(struct sock *sk); enum tcp_synack_type { TCP_SYNACK_NORMAL, diff --git a/include/net/udp.h b/include/net/udp.h index cffedb3e40f2..a061d1b22ddc 100644 --- a/include/net/udp.h +++ b/include/net/udp.h @@ -424,7 +424,7 @@ void udp4_hwcsum(struct sk_buff *skb, __be32 src, __be32 dst); int udp_rcv(struct sk_buff *skb); int udp_ioctl(struct sock *sk, int cmd, int *karg); int udp_init_sock(struct sock *sk); -int udp_pre_connect(struct sock *sk, struct sockaddr *uaddr, int addr_len); +int udp_pre_connect(struct sock *sk, struct sockaddr_unsized *uaddr, int addr_len); int __udp_disconnect(struct sock *sk, int flags); int udp_disconnect(struct sock *sk, int flags); __poll_t udp_poll(struct file *file, struct socket *sock, poll_table *wait); diff --git a/net/core/filter.c b/net/core/filter.c index 16105f52927d..90273da74807 100644 --- a/net/core/filter.c +++ b/net/core/filter.c @@ -5978,7 +5978,7 @@ BPF_CALL_3(bpf_bind, struct bpf_sock_addr_kern *, ctx, struct sockaddr *, addr, return err; if (((struct sockaddr_in *)addr)->sin_port == htons(0)) flags |= BIND_FORCE_ADDRESS_NO_PORT; - return __inet_bind(sk, addr, addr_len, flags); + return __inet_bind(sk, (struct sockaddr_unsized *)addr, addr_len, flags); #if IS_ENABLED(CONFIG_IPV6) } else if (addr->sa_family == AF_INET6) { if (addr_len < SIN6_LEN_RFC2133) @@ -5988,7 +5988,8 @@ BPF_CALL_3(bpf_bind, struct bpf_sock_addr_kern *, ctx, struct sockaddr *, addr, /* ipv6_bpf_stub cannot be NULL, since it's called from * bpf_cgroup_inet6_connect hook and ipv6 is already loaded */ - return ipv6_bpf_stub->inet6_bind(sk, addr, addr_len, flags); + return ipv6_bpf_stub->inet6_bind(sk, (struct sockaddr_unsized *)addr, + addr_len, flags); #endif /* CONFIG_IPV6 */ } #endif /* CONFIG_INET */ diff --git a/net/core/sock.c b/net/core/sock.c index f97a0e958991..3b74fc71f51c 100644 --- a/net/core/sock.c +++ b/net/core/sock.c @@ -4395,7 +4395,7 @@ bool sk_busy_loop_end(void *p, unsigned long start_time) EXPORT_SYMBOL(sk_busy_loop_end); #endif /* CONFIG_NET_RX_BUSY_POLL */ -int sock_bind_add(struct sock *sk, struct sockaddr *addr, int addr_len) +int sock_bind_add(struct sock *sk, struct sockaddr_unsized *addr, int addr_len) { if (!sk->sk_prot->bind_add) return -EOPNOTSUPP; diff --git a/net/ieee802154/socket.c b/net/ieee802154/socket.c index b93fd85f248a..e542fbe113e7 100644 --- a/net/ieee802154/socket.c +++ b/net/ieee802154/socket.c @@ -102,7 +102,7 @@ static int ieee802154_sock_bind(struct socket *sock, struct sockaddr_unsized *ua struct sock *sk = sock->sk; if (sk->sk_prot->bind) - return sk->sk_prot->bind(sk, (struct sockaddr *)uaddr, addr_len); + return sk->sk_prot->bind(sk, uaddr, addr_len); return sock_no_bind(sock, uaddr, addr_len); } @@ -118,7 +118,7 @@ static int ieee802154_sock_connect(struct socket *sock, struct sockaddr_unsized if (uaddr->sa_family == AF_UNSPEC) return sk->sk_prot->disconnect(sk, flags); - return sk->sk_prot->connect(sk, (struct sockaddr *)uaddr, addr_len); + return sk->sk_prot->connect(sk, uaddr, addr_len); } static int ieee802154_dev_ioctl(struct sock *sk, struct ifreq __user *arg, @@ -193,7 +193,7 @@ static void raw_close(struct sock *sk, long timeout) sk_common_release(sk); } -static int raw_bind(struct sock *sk, struct sockaddr *_uaddr, int len) +static int raw_bind(struct sock *sk, struct sockaddr_unsized *_uaddr, int len) { struct ieee802154_addr addr; struct sockaddr_ieee802154 *uaddr = (struct sockaddr_ieee802154 *)_uaddr; @@ -227,7 +227,7 @@ out: return err; } -static int raw_connect(struct sock *sk, struct sockaddr *uaddr, +static int raw_connect(struct sock *sk, struct sockaddr_unsized *uaddr, int addr_len) { return -ENOTSUPP; @@ -485,7 +485,7 @@ static void dgram_close(struct sock *sk, long timeout) sk_common_release(sk); } -static int dgram_bind(struct sock *sk, struct sockaddr *uaddr, int len) +static int dgram_bind(struct sock *sk, struct sockaddr_unsized *uaddr, int len) { struct sockaddr_ieee802154 *addr = (struct sockaddr_ieee802154 *)uaddr; struct ieee802154_addr haddr; @@ -563,7 +563,7 @@ static int dgram_ioctl(struct sock *sk, int cmd, int *karg) } /* FIXME: autobind */ -static int dgram_connect(struct sock *sk, struct sockaddr *uaddr, +static int dgram_connect(struct sock *sk, struct sockaddr_unsized *uaddr, int len) { struct sockaddr_ieee802154 *addr = (struct sockaddr_ieee802154 *)uaddr; diff --git a/net/ipv4/af_inet.c b/net/ipv4/af_inet.c index 0844de9ac6a4..d5ac089356eb 100644 --- a/net/ipv4/af_inet.c +++ b/net/ipv4/af_inet.c @@ -441,7 +441,7 @@ int inet_release(struct socket *sock) } EXPORT_SYMBOL(inet_release); -int inet_bind_sk(struct sock *sk, struct sockaddr *uaddr, int addr_len) +int inet_bind_sk(struct sock *sk, struct sockaddr_unsized *uaddr, int addr_len) { u32 flags = BIND_WITH_LOCK; int err; @@ -466,11 +466,11 @@ int inet_bind_sk(struct sock *sk, struct sockaddr *uaddr, int addr_len) int inet_bind(struct socket *sock, struct sockaddr_unsized *uaddr, int addr_len) { - return inet_bind_sk(sock->sk, (struct sockaddr *)uaddr, addr_len); + return inet_bind_sk(sock->sk, uaddr, addr_len); } EXPORT_SYMBOL(inet_bind); -int __inet_bind(struct sock *sk, struct sockaddr *uaddr, int addr_len, +int __inet_bind(struct sock *sk, struct sockaddr_unsized *uaddr, int addr_len, u32 flags) { struct sockaddr_in *addr = (struct sockaddr_in *)uaddr; @@ -584,14 +584,14 @@ int inet_dgram_connect(struct socket *sock, struct sockaddr_unsized *uaddr, return prot->disconnect(sk, flags); if (BPF_CGROUP_PRE_CONNECT_ENABLED(sk)) { - err = prot->pre_connect(sk, (struct sockaddr *)uaddr, addr_len); + err = prot->pre_connect(sk, uaddr, addr_len); if (err) return err; } if (data_race(!inet_sk(sk)->inet_num) && inet_autobind(sk)) return -EAGAIN; - return prot->connect(sk, (struct sockaddr *)uaddr, addr_len); + return prot->connect(sk, uaddr, addr_len); } EXPORT_SYMBOL(inet_dgram_connect); @@ -671,12 +671,12 @@ int __inet_stream_connect(struct socket *sock, struct sockaddr_unsized *uaddr, goto out; if (BPF_CGROUP_PRE_CONNECT_ENABLED(sk)) { - err = sk->sk_prot->pre_connect(sk, (struct sockaddr *)uaddr, addr_len); + err = sk->sk_prot->pre_connect(sk, uaddr, addr_len); if (err) goto out; } - err = sk->sk_prot->connect(sk, (struct sockaddr *)uaddr, addr_len); + err = sk->sk_prot->connect(sk, uaddr, addr_len); if (err < 0) goto out; diff --git a/net/ipv4/datagram.c b/net/ipv4/datagram.c index c2b2cda1a7e5..1614593b6d72 100644 --- a/net/ipv4/datagram.c +++ b/net/ipv4/datagram.c @@ -16,7 +16,7 @@ #include #include -int __ip4_datagram_connect(struct sock *sk, struct sockaddr *uaddr, int addr_len) +int __ip4_datagram_connect(struct sock *sk, struct sockaddr_unsized *uaddr, int addr_len) { struct inet_sock *inet = inet_sk(sk); struct sockaddr_in *usin = (struct sockaddr_in *) uaddr; @@ -84,7 +84,7 @@ out: } EXPORT_SYMBOL(__ip4_datagram_connect); -int ip4_datagram_connect(struct sock *sk, struct sockaddr *uaddr, int addr_len) +int ip4_datagram_connect(struct sock *sk, struct sockaddr_unsized *uaddr, int addr_len) { int res; diff --git a/net/ipv4/ping.c b/net/ipv4/ping.c index 5321c5801c64..ad56588107cc 100644 --- a/net/ipv4/ping.c +++ b/net/ipv4/ping.c @@ -286,7 +286,7 @@ void ping_close(struct sock *sk, long timeout) } EXPORT_IPV6_MOD_GPL(ping_close); -static int ping_pre_connect(struct sock *sk, struct sockaddr *uaddr, +static int ping_pre_connect(struct sock *sk, struct sockaddr_unsized *uaddr, int addr_len) { /* This check is replicated from __ip4_datagram_connect() and @@ -301,7 +301,7 @@ static int ping_pre_connect(struct sock *sk, struct sockaddr *uaddr, /* Checks the bind address and possibly modifies sk->sk_bound_dev_if. */ static int ping_check_bind_addr(struct sock *sk, struct inet_sock *isk, - struct sockaddr *uaddr, int addr_len) + struct sockaddr_unsized *uaddr, int addr_len) { struct net *net = sock_net(sk); if (sk->sk_family == AF_INET) { @@ -387,7 +387,7 @@ static int ping_check_bind_addr(struct sock *sk, struct inet_sock *isk, return 0; } -static void ping_set_saddr(struct sock *sk, struct sockaddr *saddr) +static void ping_set_saddr(struct sock *sk, struct sockaddr_unsized *saddr) { if (saddr->sa_family == AF_INET) { struct inet_sock *isk = inet_sk(sk); @@ -407,7 +407,7 @@ static void ping_set_saddr(struct sock *sk, struct sockaddr *saddr) * Moreover, we don't allow binding to multi- and broadcast addresses. */ -int ping_bind(struct sock *sk, struct sockaddr *uaddr, int addr_len) +int ping_bind(struct sock *sk, struct sockaddr_unsized *uaddr, int addr_len) { struct inet_sock *isk = inet_sk(sk); unsigned short snum; diff --git a/net/ipv4/raw.c b/net/ipv4/raw.c index d54ebb7df966..5998c4cc6f47 100644 --- a/net/ipv4/raw.c +++ b/net/ipv4/raw.c @@ -697,7 +697,8 @@ static void raw_destroy(struct sock *sk) } /* This gets rid of all the nasties in af_inet. -DaveM */ -static int raw_bind(struct sock *sk, struct sockaddr *uaddr, int addr_len) +static int raw_bind(struct sock *sk, struct sockaddr_unsized *uaddr, + int addr_len) { struct inet_sock *inet = inet_sk(sk); struct sockaddr_in *addr = (struct sockaddr_in *) uaddr; diff --git a/net/ipv4/tcp_ipv4.c b/net/ipv4/tcp_ipv4.c index 40a76da5364a..b7526a7888cb 100644 --- a/net/ipv4/tcp_ipv4.c +++ b/net/ipv4/tcp_ipv4.c @@ -205,7 +205,7 @@ int tcp_twsk_unique(struct sock *sk, struct sock *sktw, void *twp) } EXPORT_IPV6_MOD_GPL(tcp_twsk_unique); -static int tcp_v4_pre_connect(struct sock *sk, struct sockaddr *uaddr, +static int tcp_v4_pre_connect(struct sock *sk, struct sockaddr_unsized *uaddr, int addr_len) { /* This check is replicated from tcp_v4_connect() and intended to @@ -221,7 +221,7 @@ static int tcp_v4_pre_connect(struct sock *sk, struct sockaddr *uaddr, } /* This will initiate an outgoing connection. */ -int tcp_v4_connect(struct sock *sk, struct sockaddr *uaddr, int addr_len) +int tcp_v4_connect(struct sock *sk, struct sockaddr_unsized *uaddr, int addr_len) { struct sockaddr_in *usin = (struct sockaddr_in *)uaddr; struct inet_timewait_death_row *tcp_death_row; diff --git a/net/ipv4/udp.c b/net/ipv4/udp.c index 30dfbf73729d..ffe074cb5865 100644 --- a/net/ipv4/udp.c +++ b/net/ipv4/udp.c @@ -2159,7 +2159,8 @@ csum_copy_err: goto try_again; } -int udp_pre_connect(struct sock *sk, struct sockaddr *uaddr, int addr_len) +int udp_pre_connect(struct sock *sk, struct sockaddr_unsized *uaddr, + int addr_len) { /* This check is replicated from __ip4_datagram_connect() and * intended to prevent BPF program called below from accessing bytes @@ -2172,7 +2173,8 @@ int udp_pre_connect(struct sock *sk, struct sockaddr *uaddr, int addr_len) } EXPORT_IPV6_MOD(udp_pre_connect); -static int udp_connect(struct sock *sk, struct sockaddr *uaddr, int addr_len) +static int udp_connect(struct sock *sk, struct sockaddr_unsized *uaddr, + int addr_len) { int res; diff --git a/net/ipv6/af_inet6.c b/net/ipv6/af_inet6.c index c92d27e35fbc..b705751eb73c 100644 --- a/net/ipv6/af_inet6.c +++ b/net/ipv6/af_inet6.c @@ -277,7 +277,7 @@ out_sk_release: goto out; } -static int __inet6_bind(struct sock *sk, struct sockaddr *uaddr, int addr_len, +static int __inet6_bind(struct sock *sk, struct sockaddr_unsized *uaddr, int addr_len, u32 flags) { struct sockaddr_in6 *addr = (struct sockaddr_in6 *)uaddr; @@ -438,7 +438,7 @@ out_unlock: goto out; } -int inet6_bind_sk(struct sock *sk, struct sockaddr *uaddr, int addr_len) +int inet6_bind_sk(struct sock *sk, struct sockaddr_unsized *uaddr, int addr_len) { u32 flags = BIND_WITH_LOCK; const struct proto *prot; @@ -467,7 +467,7 @@ int inet6_bind_sk(struct sock *sk, struct sockaddr *uaddr, int addr_len) /* bind for INET6 API */ int inet6_bind(struct socket *sock, struct sockaddr_unsized *uaddr, int addr_len) { - return inet6_bind_sk(sock->sk, (struct sockaddr *)uaddr, addr_len); + return inet6_bind_sk(sock->sk, uaddr, addr_len); } EXPORT_SYMBOL(inet6_bind); diff --git a/net/ipv6/datagram.c b/net/ipv6/datagram.c index 33ebe93d80e3..83e03176819c 100644 --- a/net/ipv6/datagram.c +++ b/net/ipv6/datagram.c @@ -138,7 +138,7 @@ void ip6_datagram_release_cb(struct sock *sk) } EXPORT_SYMBOL_GPL(ip6_datagram_release_cb); -int __ip6_datagram_connect(struct sock *sk, struct sockaddr *uaddr, +int __ip6_datagram_connect(struct sock *sk, struct sockaddr_unsized *uaddr, int addr_len) { struct sockaddr_in6 *usin = (struct sockaddr_in6 *) uaddr; @@ -194,7 +194,7 @@ int __ip6_datagram_connect(struct sock *sk, struct sockaddr *uaddr, sin.sin_port = usin->sin6_port; err = __ip4_datagram_connect(sk, - (struct sockaddr *) &sin, + (struct sockaddr_unsized *)&sin, sizeof(sin)); ipv4_connected: @@ -271,7 +271,7 @@ out: } EXPORT_SYMBOL_GPL(__ip6_datagram_connect); -int ip6_datagram_connect(struct sock *sk, struct sockaddr *uaddr, int addr_len) +int ip6_datagram_connect(struct sock *sk, struct sockaddr_unsized *uaddr, int addr_len) { int res; @@ -282,7 +282,7 @@ int ip6_datagram_connect(struct sock *sk, struct sockaddr *uaddr, int addr_len) } EXPORT_SYMBOL_GPL(ip6_datagram_connect); -int ip6_datagram_connect_v6_only(struct sock *sk, struct sockaddr *uaddr, +int ip6_datagram_connect_v6_only(struct sock *sk, struct sockaddr_unsized *uaddr, int addr_len) { DECLARE_SOCKADDR(struct sockaddr_in6 *, sin6, uaddr); diff --git a/net/ipv6/ping.c b/net/ipv6/ping.c index d7a2cdaa2631..e4afc651731a 100644 --- a/net/ipv6/ping.c +++ b/net/ipv6/ping.c @@ -45,7 +45,7 @@ static int dummy_ipv6_chk_addr(struct net *net, const struct in6_addr *addr, return 0; } -static int ping_v6_pre_connect(struct sock *sk, struct sockaddr *uaddr, +static int ping_v6_pre_connect(struct sock *sk, struct sockaddr_unsized *uaddr, int addr_len) { /* This check is replicated from __ip6_datagram_connect() and diff --git a/net/ipv6/raw.c b/net/ipv6/raw.c index e369f54844dd..b4cd05dba9b6 100644 --- a/net/ipv6/raw.c +++ b/net/ipv6/raw.c @@ -214,7 +214,8 @@ bool raw6_local_deliver(struct sk_buff *skb, int nexthdr) } /* This cleans up af_inet6 a bit. -DaveM */ -static int rawv6_bind(struct sock *sk, struct sockaddr *uaddr, int addr_len) +static int rawv6_bind(struct sock *sk, struct sockaddr_unsized *uaddr, + int addr_len) { struct inet_sock *inet = inet_sk(sk); struct ipv6_pinfo *np = inet6_sk(sk); diff --git a/net/ipv6/tcp_ipv6.c b/net/ipv6/tcp_ipv6.c index 06eb90e4078e..7df21c1cba21 100644 --- a/net/ipv6/tcp_ipv6.c +++ b/net/ipv6/tcp_ipv6.c @@ -118,7 +118,7 @@ static u32 tcp_v6_init_ts_off(const struct net *net, const struct sk_buff *skb) ipv6_hdr(skb)->saddr.s6_addr32); } -static int tcp_v6_pre_connect(struct sock *sk, struct sockaddr *uaddr, +static int tcp_v6_pre_connect(struct sock *sk, struct sockaddr_unsized *uaddr, int addr_len) { /* This check is replicated from tcp_v6_connect() and intended to @@ -133,7 +133,7 @@ static int tcp_v6_pre_connect(struct sock *sk, struct sockaddr *uaddr, return BPF_CGROUP_RUN_PROG_INET6_CONNECT(sk, uaddr, &addr_len); } -static int tcp_v6_connect(struct sock *sk, struct sockaddr *uaddr, +static int tcp_v6_connect(struct sock *sk, struct sockaddr_unsized *uaddr, int addr_len) { struct sockaddr_in6 *usin = (struct sockaddr_in6 *) uaddr; @@ -238,7 +238,7 @@ static int tcp_v6_connect(struct sock *sk, struct sockaddr *uaddr, tp->af_specific = &tcp_sock_ipv6_mapped_specific; #endif - err = tcp_v4_connect(sk, (struct sockaddr *)&sin, sizeof(sin)); + err = tcp_v4_connect(sk, (struct sockaddr_unsized *)&sin, sizeof(sin)); if (err) { icsk->icsk_ext_hdr_len = exthdrlen; diff --git a/net/ipv6/udp.c b/net/ipv6/udp.c index 813a2ba75824..794c13674e8a 100644 --- a/net/ipv6/udp.c +++ b/net/ipv6/udp.c @@ -1282,7 +1282,7 @@ static void udp_v6_flush_pending_frames(struct sock *sk) } } -static int udpv6_pre_connect(struct sock *sk, struct sockaddr *uaddr, +static int udpv6_pre_connect(struct sock *sk, struct sockaddr_unsized *uaddr, int addr_len) { if (addr_len < offsetofend(struct sockaddr, sa_family)) @@ -1303,7 +1303,8 @@ static int udpv6_pre_connect(struct sock *sk, struct sockaddr *uaddr, return BPF_CGROUP_RUN_PROG_INET6_CONNECT_LOCK(sk, uaddr, &addr_len); } -static int udpv6_connect(struct sock *sk, struct sockaddr *uaddr, int addr_len) +static int udpv6_connect(struct sock *sk, struct sockaddr_unsized *uaddr, + int addr_len) { int res; diff --git a/net/l2tp/l2tp_ip.c b/net/l2tp/l2tp_ip.c index 29795d2839e8..cac1ff59cb83 100644 --- a/net/l2tp/l2tp_ip.c +++ b/net/l2tp/l2tp_ip.c @@ -267,7 +267,8 @@ static void l2tp_ip_destroy_sock(struct sock *sk) } } -static int l2tp_ip_bind(struct sock *sk, struct sockaddr *uaddr, int addr_len) +static int l2tp_ip_bind(struct sock *sk, struct sockaddr_unsized *uaddr, + int addr_len) { struct inet_sock *inet = inet_sk(sk); struct sockaddr_l2tpip *addr = (struct sockaddr_l2tpip *)uaddr; @@ -328,7 +329,8 @@ out: return ret; } -static int l2tp_ip_connect(struct sock *sk, struct sockaddr *uaddr, int addr_len) +static int l2tp_ip_connect(struct sock *sk, struct sockaddr_unsized *uaddr, + int addr_len) { struct sockaddr_l2tpip *lsa = (struct sockaddr_l2tpip *)uaddr; struct l2tp_ip_net *pn = l2tp_ip_pernet(sock_net(sk)); diff --git a/net/l2tp/l2tp_ip6.c b/net/l2tp/l2tp_ip6.c index ea232f338dcb..05a396ba6a3e 100644 --- a/net/l2tp/l2tp_ip6.c +++ b/net/l2tp/l2tp_ip6.c @@ -280,7 +280,8 @@ static void l2tp_ip6_destroy_sock(struct sock *sk) } } -static int l2tp_ip6_bind(struct sock *sk, struct sockaddr *uaddr, int addr_len) +static int l2tp_ip6_bind(struct sock *sk, struct sockaddr_unsized *uaddr, + int addr_len) { struct inet_sock *inet = inet_sk(sk); struct ipv6_pinfo *np = inet6_sk(sk); @@ -383,7 +384,7 @@ out_unlock: return err; } -static int l2tp_ip6_connect(struct sock *sk, struct sockaddr *uaddr, +static int l2tp_ip6_connect(struct sock *sk, struct sockaddr_unsized *uaddr, int addr_len) { struct sockaddr_l2tpip6 *lsa = (struct sockaddr_l2tpip6 *)uaddr; diff --git a/net/mptcp/pm_kernel.c b/net/mptcp/pm_kernel.c index e50721c670d0..598f01a573c1 100644 --- a/net/mptcp/pm_kernel.c +++ b/net/mptcp/pm_kernel.c @@ -867,10 +867,10 @@ static int mptcp_pm_nl_create_listen_socket(struct sock *sk, addrlen = sizeof(struct sockaddr_in6); #endif if (ssk->sk_family == AF_INET) - err = inet_bind_sk(ssk, (struct sockaddr *)&addr, addrlen); + err = inet_bind_sk(ssk, (struct sockaddr_unsized *)&addr, addrlen); #if IS_ENABLED(CONFIG_MPTCP_IPV6) else if (ssk->sk_family == AF_INET6) - err = inet6_bind_sk(ssk, (struct sockaddr *)&addr, addrlen); + err = inet6_bind_sk(ssk, (struct sockaddr_unsized *)&addr, addrlen); #endif if (err) return err; diff --git a/net/mptcp/protocol.c b/net/mptcp/protocol.c index 53e2b095dfb1..4cd5df01446e 100644 --- a/net/mptcp/protocol.c +++ b/net/mptcp/protocol.c @@ -3746,7 +3746,8 @@ static int mptcp_ioctl(struct sock *sk, int cmd, int *karg) return 0; } -static int mptcp_connect(struct sock *sk, struct sockaddr *uaddr, int addr_len) +static int mptcp_connect(struct sock *sk, struct sockaddr_unsized *uaddr, + int addr_len) { struct mptcp_subflow_context *subflow; struct mptcp_sock *msk = mptcp_sk(sk); @@ -3870,10 +3871,10 @@ static int mptcp_bind(struct socket *sock, struct sockaddr_unsized *uaddr, int a } if (sk->sk_family == AF_INET) - err = inet_bind_sk(ssk, (struct sockaddr *)uaddr, addr_len); + err = inet_bind_sk(ssk, uaddr, addr_len); #if IS_ENABLED(CONFIG_MPTCP_IPV6) else if (sk->sk_family == AF_INET6) - err = inet6_bind_sk(ssk, (struct sockaddr *)uaddr, addr_len); + err = inet6_bind_sk(ssk, uaddr, addr_len); #endif if (!err) mptcp_copy_inaddrs(sk, ssk); diff --git a/net/phonet/pep.c b/net/phonet/pep.c index 4db564d9d522..120e711ea78c 100644 --- a/net/phonet/pep.c +++ b/net/phonet/pep.c @@ -882,7 +882,8 @@ drop: return newsk; } -static int pep_sock_connect(struct sock *sk, struct sockaddr *addr, int len) +static int pep_sock_connect(struct sock *sk, struct sockaddr_unsized *addr, + int len) { struct pep_sock *pn = pep_sk(sk); int err; diff --git a/net/phonet/socket.c b/net/phonet/socket.c index 9391378083a4..4423d483c630 100644 --- a/net/phonet/socket.c +++ b/net/phonet/socket.c @@ -163,7 +163,7 @@ static int pn_socket_bind(struct socket *sock, struct sockaddr_unsized *addr, in u8 saddr; if (sk->sk_prot->bind) - return sk->sk_prot->bind(sk, (struct sockaddr *)addr, len); + return sk->sk_prot->bind(sk, addr, len); if (len < sizeof(struct sockaddr_pn)) return -EINVAL; @@ -252,7 +252,7 @@ static int pn_socket_connect(struct socket *sock, struct sockaddr_unsized *addr, pn->resource = pn_sockaddr_get_resource(spn); sock->state = SS_CONNECTING; - err = sk->sk_prot->connect(sk, (struct sockaddr *)addr, len); + err = sk->sk_prot->connect(sk, addr, len); if (err) { sock->state = SS_UNCONNECTED; pn->dobject = 0; diff --git a/net/sctp/socket.c b/net/sctp/socket.c index 940abbced191..38d2932acebf 100644 --- a/net/sctp/socket.c +++ b/net/sctp/socket.c @@ -306,7 +306,8 @@ static struct sctp_transport *sctp_addr_id2transport(struct sock *sk, * sockaddr_in6 [RFC 2553]), * addr_len - the size of the address structure. */ -static int sctp_bind(struct sock *sk, struct sockaddr *addr, int addr_len) +static int sctp_bind(struct sock *sk, struct sockaddr_unsized *addr, + int addr_len) { int retval = 0; @@ -1053,13 +1054,13 @@ static int sctp_setsockopt_bindx(struct sock *sk, struct sockaddr *addrs, } } -static int sctp_bind_add(struct sock *sk, struct sockaddr *addrs, - int addrlen) +static int sctp_bind_add(struct sock *sk, struct sockaddr_unsized *addrs, + int addrlen) { int err; lock_sock(sk); - err = sctp_setsockopt_bindx(sk, addrs, addrlen, SCTP_BINDX_ADD_ADDR); + err = sctp_setsockopt_bindx(sk, (struct sockaddr *)addrs, addrlen, SCTP_BINDX_ADD_ADDR); release_sock(sk); return err; } -- cgit v1.2.3 From 6b47af35a6dded074ff583361f6d6668dd7a401d Mon Sep 17 00:00:00 2001 From: Raju Rangoju Date: Fri, 31 Oct 2025 16:48:11 +0530 Subject: net: selftests: export packet creation helpers for driver use Export the network selftest packet creation infrastructure to allow network drivers to reuse the existing selftest framework instead of duplicating packet creation code. Signed-off-by: Raju Rangoju Reviewed-by: Maxime Chevallier Link: https://patch.msgid.link/20251031111811.775434-1-Raju.Rangoju@amd.com Signed-off-by: Paolo Abeni --- include/net/selftests.h | 45 +++++++++++++++++++++++++++++++++++++++++++++ net/core/selftests.c | 48 +++++++----------------------------------------- 2 files changed, 52 insertions(+), 41 deletions(-) (limited to 'include/net') diff --git a/include/net/selftests.h b/include/net/selftests.h index e65e8d230d33..c36e07406ad4 100644 --- a/include/net/selftests.h +++ b/include/net/selftests.h @@ -3,9 +3,48 @@ #define _NET_SELFTESTS #include +#include + +struct net_packet_attrs { + const unsigned char *src; + const unsigned char *dst; + u32 ip_src; + u32 ip_dst; + bool tcp; + u16 sport; + u16 dport; + int timeout; + int size; + int max_size; + u8 id; + u16 queue_mapping; + bool bad_csum; +}; + +struct net_test_priv { + struct net_packet_attrs *packet; + struct packet_type pt; + struct completion comp; + int double_vlan; + int vlan_id; + int ok; +}; + +struct netsfhdr { + __be32 version; + __be64 magic; + u8 id; +} __packed; + +#define NET_TEST_PKT_SIZE (sizeof(struct ethhdr) + sizeof(struct iphdr) + \ + sizeof(struct netsfhdr)) +#define NET_TEST_PKT_MAGIC 0xdeadcafecafedeadULL +#define NET_LB_TIMEOUT msecs_to_jiffies(200) #if IS_ENABLED(CONFIG_NET_SELFTESTS) +struct sk_buff *net_test_get_skb(struct net_device *ndev, u8 id, + struct net_packet_attrs *attr); void net_selftest(struct net_device *ndev, struct ethtool_test *etest, u64 *buf); int net_selftest_get_count(void); @@ -13,6 +52,12 @@ void net_selftest_get_strings(u8 *data); #else +static inline struct sk_buff *net_test_get_skb(struct net_device *ndev, u8 id, + struct net_packet_attrs *attr) +{ + return NULL; +} + static inline void net_selftest(struct net_device *ndev, struct ethtool_test *etest, u64 *buf) { diff --git a/net/core/selftests.c b/net/core/selftests.c index 3d79133a91a6..8b81feb82c4a 100644 --- a/net/core/selftests.c +++ b/net/core/selftests.c @@ -14,46 +14,10 @@ #include #include -struct net_packet_attrs { - const unsigned char *src; - const unsigned char *dst; - u32 ip_src; - u32 ip_dst; - bool tcp; - u16 sport; - u16 dport; - int timeout; - int size; - int max_size; - u8 id; - u16 queue_mapping; - bool bad_csum; -}; - -struct net_test_priv { - struct net_packet_attrs *packet; - struct packet_type pt; - struct completion comp; - int double_vlan; - int vlan_id; - int ok; -}; - -struct netsfhdr { - __be32 version; - __be64 magic; - u8 id; -} __packed; - static u8 net_test_next_id; -#define NET_TEST_PKT_SIZE (sizeof(struct ethhdr) + sizeof(struct iphdr) + \ - sizeof(struct netsfhdr)) -#define NET_TEST_PKT_MAGIC 0xdeadcafecafedeadULL -#define NET_LB_TIMEOUT msecs_to_jiffies(200) - -static struct sk_buff *net_test_get_skb(struct net_device *ndev, - struct net_packet_attrs *attr) +struct sk_buff *net_test_get_skb(struct net_device *ndev, u8 id, + struct net_packet_attrs *attr) { struct sk_buff *skb = NULL; struct udphdr *uhdr = NULL; @@ -142,8 +106,8 @@ static struct sk_buff *net_test_get_skb(struct net_device *ndev, shdr = skb_put(skb, sizeof(*shdr)); shdr->version = 0; shdr->magic = cpu_to_be64(NET_TEST_PKT_MAGIC); - attr->id = net_test_next_id; - shdr->id = net_test_next_id++; + attr->id = id; + shdr->id = id; if (attr->size) { void *payload = skb_put(skb, attr->size); @@ -190,6 +154,7 @@ static struct sk_buff *net_test_get_skb(struct net_device *ndev, return skb; } +EXPORT_SYMBOL_GPL(net_test_get_skb); static int net_test_loopback_validate(struct sk_buff *skb, struct net_device *ndev, @@ -286,12 +251,13 @@ static int __net_test_loopback(struct net_device *ndev, tpriv->packet = attr; dev_add_pack(&tpriv->pt); - skb = net_test_get_skb(ndev, attr); + skb = net_test_get_skb(ndev, net_test_next_id, attr); if (!skb) { ret = -ENOMEM; goto cleanup; } + net_test_next_id++; ret = dev_direct_xmit(skb, attr->queue_mapping); if (ret < 0) { goto cleanup; -- cgit v1.2.3 From 9352d40c8bcd2ef29366d2c38b163c0b115039ed Mon Sep 17 00:00:00 2001 From: Mohammad Heib Date: Sat, 25 Oct 2025 16:08:58 +0300 Subject: devlink: Add new "max_mac_per_vf" generic device param Add a new device generic parameter to controls the maximum number of MAC filters allowed per VF. For example, to limit a VF to 3 MAC addresses: $ devlink dev param set pci/0000:3b:00.0 name max_mac_per_vf \ value 3 \ cmode runtime Signed-off-by: Mohammad Heib Reviewed-by: Simon Horman Signed-off-by: Jacob Keller Signed-off-by: Tony Nguyen --- Documentation/networking/devlink/devlink-params.rst | 4 ++++ include/net/devlink.h | 4 ++++ net/devlink/param.c | 5 +++++ 3 files changed, 13 insertions(+) (limited to 'include/net') diff --git a/Documentation/networking/devlink/devlink-params.rst b/Documentation/networking/devlink/devlink-params.rst index 0a9c20d70122..c0597d456641 100644 --- a/Documentation/networking/devlink/devlink-params.rst +++ b/Documentation/networking/devlink/devlink-params.rst @@ -151,3 +151,7 @@ own name. * - ``num_doorbells`` - u32 - Controls the number of doorbells used by the device. + * - ``max_mac_per_vf`` + - u32 + - Controls the maximum number of MAC address filters that can be assigned + to a Virtual Function (VF). diff --git a/include/net/devlink.h b/include/net/devlink.h index 9e824f61e40f..d01046ef0577 100644 --- a/include/net/devlink.h +++ b/include/net/devlink.h @@ -532,6 +532,7 @@ enum devlink_param_generic_id { DEVLINK_PARAM_GENERIC_ID_CLOCK_ID, DEVLINK_PARAM_GENERIC_ID_TOTAL_VFS, DEVLINK_PARAM_GENERIC_ID_NUM_DOORBELLS, + DEVLINK_PARAM_GENERIC_ID_MAX_MAC_PER_VF, /* add new param generic ids above here*/ __DEVLINK_PARAM_GENERIC_ID_MAX, @@ -602,6 +603,9 @@ enum devlink_param_generic_id { #define DEVLINK_PARAM_GENERIC_NUM_DOORBELLS_NAME "num_doorbells" #define DEVLINK_PARAM_GENERIC_NUM_DOORBELLS_TYPE DEVLINK_PARAM_TYPE_U32 +#define DEVLINK_PARAM_GENERIC_MAX_MAC_PER_VF_NAME "max_mac_per_vf" +#define DEVLINK_PARAM_GENERIC_MAX_MAC_PER_VF_TYPE DEVLINK_PARAM_TYPE_U32 + #define DEVLINK_PARAM_GENERIC(_id, _cmodes, _get, _set, _validate) \ { \ .id = DEVLINK_PARAM_GENERIC_ID_##_id, \ diff --git a/net/devlink/param.c b/net/devlink/param.c index 70e69523412c..6b233b13b69a 100644 --- a/net/devlink/param.c +++ b/net/devlink/param.c @@ -112,6 +112,11 @@ static const struct devlink_param devlink_param_generic[] = { .name = DEVLINK_PARAM_GENERIC_NUM_DOORBELLS_NAME, .type = DEVLINK_PARAM_GENERIC_NUM_DOORBELLS_TYPE, }, + { + .id = DEVLINK_PARAM_GENERIC_ID_MAX_MAC_PER_VF, + .name = DEVLINK_PARAM_GENERIC_MAX_MAC_PER_VF_NAME, + .type = DEVLINK_PARAM_GENERIC_MAX_MAC_PER_VF_TYPE, + }, }; static int devlink_param_generic_verify(const struct devlink_param *param) -- cgit v1.2.3 From c6230446b1a6f3c91effafd99f604de455da52e5 Mon Sep 17 00:00:00 2001 From: Daniel Golle Date: Mon, 3 Nov 2025 12:20:20 +0000 Subject: net: dsa: add tagging driver for MaxLinear GSW1xx switch family Add support for a new DSA tagging protocol driver for the MaxLinear GSW1xx switch family. The GSW1xx switches use a proprietary 8-byte special tag inserted between the source MAC address and the EtherType field to indicate the source and destination ports for frames traversing the CPU port. Implement the tag handling logic to insert the special tag on transmit and parse it on receive. Signed-off-by: Daniel Golle Reviewed-by: Alexander Sverdlin Tested-by: Alexander Sverdlin Link: https://patch.msgid.link/0e973ebfd9433c30c96f50670da9e9449a0d98f2.1762170107.git.daniel@makrotopia.org Signed-off-by: Jakub Kicinski --- MAINTAINERS | 3 +- include/net/dsa.h | 2 + include/uapi/linux/if_ether.h | 1 + net/dsa/Kconfig | 8 +++ net/dsa/Makefile | 1 + net/dsa/tag_mxl-gsw1xx.c | 116 ++++++++++++++++++++++++++++++++++++++++++ 6 files changed, 130 insertions(+), 1 deletion(-) create mode 100644 net/dsa/tag_mxl-gsw1xx.c (limited to 'include/net') diff --git a/MAINTAINERS b/MAINTAINERS index 12cd8a5ab274..0dc4aa37d903 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -14053,7 +14053,7 @@ F: tools/testing/selftests/landlock/ K: landlock K: LANDLOCK -LANTIQ / INTEL Ethernet drivers +LANTIQ / MAXLINEAR / INTEL Ethernet DSA drivers M: Hauke Mehrtens L: netdev@vger.kernel.org S: Maintained @@ -14061,6 +14061,7 @@ F: Documentation/devicetree/bindings/net/dsa/lantiq,gswip.yaml F: drivers/net/dsa/lantiq/* F: drivers/net/ethernet/lantiq_xrx200.c F: net/dsa/tag_gswip.c +F: net/dsa/tag_mxl-gsw1xx.c LANTIQ MIPS ARCHITECTURE M: John Crispin diff --git a/include/net/dsa.h b/include/net/dsa.h index 67762fdaf3c7..2df2e2ead9a8 100644 --- a/include/net/dsa.h +++ b/include/net/dsa.h @@ -56,6 +56,7 @@ struct tc_action; #define DSA_TAG_PROTO_VSC73XX_8021Q_VALUE 28 #define DSA_TAG_PROTO_BRCM_LEGACY_FCS_VALUE 29 #define DSA_TAG_PROTO_YT921X_VALUE 30 +#define DSA_TAG_PROTO_MXL_GSW1XX_VALUE 31 enum dsa_tag_protocol { DSA_TAG_PROTO_NONE = DSA_TAG_PROTO_NONE_VALUE, @@ -89,6 +90,7 @@ enum dsa_tag_protocol { DSA_TAG_PROTO_LAN937X = DSA_TAG_PROTO_LAN937X_VALUE, DSA_TAG_PROTO_VSC73XX_8021Q = DSA_TAG_PROTO_VSC73XX_8021Q_VALUE, DSA_TAG_PROTO_YT921X = DSA_TAG_PROTO_YT921X_VALUE, + DSA_TAG_PROTO_MXL_GSW1XX = DSA_TAG_PROTO_MXL_GSW1XX_VALUE, }; struct dsa_switch; diff --git a/include/uapi/linux/if_ether.h b/include/uapi/linux/if_ether.h index cfd200c87e5e..2c93b7b731c8 100644 --- a/include/uapi/linux/if_ether.h +++ b/include/uapi/linux/if_ether.h @@ -92,6 +92,7 @@ #define ETH_P_ETHERCAT 0x88A4 /* EtherCAT */ #define ETH_P_8021AD 0x88A8 /* 802.1ad Service VLAN */ #define ETH_P_802_EX1 0x88B5 /* 802.1 Local Experimental 1. */ +#define ETH_P_MXLGSW 0x88C3 /* MaxLinear GSW DSA [ NOT AN OFFICIALLY REGISTERED ID ] */ #define ETH_P_PREAUTH 0x88C7 /* 802.11 Preauthentication */ #define ETH_P_TIPC 0x88CA /* TIPC */ #define ETH_P_LLDP 0x88CC /* Link Layer Discovery Protocol */ diff --git a/net/dsa/Kconfig b/net/dsa/Kconfig index 6b94028b1fcc..f86b30742122 100644 --- a/net/dsa/Kconfig +++ b/net/dsa/Kconfig @@ -104,6 +104,14 @@ config NET_DSA_TAG_MTK Say Y or M if you want to enable support for tagging frames for Mediatek switches. +config NET_DSA_TAG_MXL_GSW1XX + tristate "Tag driver for MaxLinear GSW1xx switches" + help + The GSW1xx family of switches supports an 8-byte special tag which + can be used on the CPU port of the switch. + Say Y or M if you want to enable support for tagging frames for + MaxLinear GSW1xx switches. + config NET_DSA_TAG_KSZ tristate "Tag driver for Microchip 8795/937x/9477/9893 families of switches" help diff --git a/net/dsa/Makefile b/net/dsa/Makefile index 4b011a1d5c87..42d173f5a701 100644 --- a/net/dsa/Makefile +++ b/net/dsa/Makefile @@ -28,6 +28,7 @@ obj-$(CONFIG_NET_DSA_TAG_HELLCREEK) += tag_hellcreek.o obj-$(CONFIG_NET_DSA_TAG_KSZ) += tag_ksz.o obj-$(CONFIG_NET_DSA_TAG_LAN9303) += tag_lan9303.o obj-$(CONFIG_NET_DSA_TAG_MTK) += tag_mtk.o +obj-$(CONFIG_NET_DSA_TAG_MXL_GSW1XX) += tag_mxl-gsw1xx.o obj-$(CONFIG_NET_DSA_TAG_NONE) += tag_none.o obj-$(CONFIG_NET_DSA_TAG_OCELOT) += tag_ocelot.o obj-$(CONFIG_NET_DSA_TAG_OCELOT_8021Q) += tag_ocelot_8021q.o diff --git a/net/dsa/tag_mxl-gsw1xx.c b/net/dsa/tag_mxl-gsw1xx.c new file mode 100644 index 000000000000..701a079955f2 --- /dev/null +++ b/net/dsa/tag_mxl-gsw1xx.c @@ -0,0 +1,116 @@ +// SPDX-License-Identifier: GPL-2.0+ +/* + * DSA driver Special Tag support for MaxLinear GSW1xx switch chips + * + * Copyright (C) 2025 Daniel Golle + * Copyright (C) 2023 - 2024 MaxLinear Inc. + */ + +#include +#include +#include +#include + +#include "tag.h" + +/* To define the outgoing port and to discover the incoming port a special + * tag is used by the GSW1xx. + * + * Dest MAC Src MAC special TAG EtherType + * ...| 1 2 3 4 5 6 | 1 2 3 4 5 6 | 1 2 3 4 5 6 7 8 | 1 2 |... + * |<--------------->| + */ + +#define GSW1XX_TAG_NAME "gsw1xx" + +/* special tag header length (RX and TX) */ +#define GSW1XX_HEADER_LEN 8 + +/* Word 0 = Ethertype -> 0x88C3 */ + +/* Word 1 */ +#define GSW1XX_TX_PORT_MAP GENMASK(7, 0) +#define GSW1XX_TX_PORT_MAP_EN BIT(15) +#define GSW1XX_TX_CLASS_EN BIT(14) +#define GSW1XX_TX_TIME_STAMP_EN BIT(13) +#define GSW1XX_TX_LRN_DIS BIT(12) +#define GSW1XX_TX_CLASS GENMASK(11, 8) + +/* special tag in RX path header */ +/* Word 2 */ +#define GSW1XX_RX_PORT_MAP GENMASK(15, 8) + +static struct sk_buff *gsw1xx_tag_xmit(struct sk_buff *skb, + struct net_device *dev) +{ + struct dsa_port *dp = dsa_user_to_port(dev); + __be16 *gsw1xx_tag; + + /* provide additional space 'GSW1XX_HEADER_LEN' bytes */ + skb_push(skb, GSW1XX_HEADER_LEN); + + /* add space between MAC address and Ethertype */ + dsa_alloc_etype_header(skb, GSW1XX_HEADER_LEN); + + /* special tag ingress */ + gsw1xx_tag = dsa_etype_header_pos_tx(skb); + gsw1xx_tag[0] = htons(ETH_P_MXLGSW); + gsw1xx_tag[1] = htons(GSW1XX_TX_PORT_MAP_EN | GSW1XX_TX_LRN_DIS | + FIELD_PREP(GSW1XX_TX_PORT_MAP, BIT(dp->index))); + + gsw1xx_tag[2] = 0; + gsw1xx_tag[3] = 0; + + return skb; +} + +static struct sk_buff *gsw1xx_tag_rcv(struct sk_buff *skb, + struct net_device *dev) +{ + int port; + __be16 *gsw1xx_tag; + + if (unlikely(!pskb_may_pull(skb, GSW1XX_HEADER_LEN))) { + dev_warn_ratelimited(&dev->dev, "Dropping packet, cannot pull SKB\n"); + return NULL; + } + + gsw1xx_tag = dsa_etype_header_pos_rx(skb); + + if (unlikely(ntohs(gsw1xx_tag[0]) != ETH_P_MXLGSW)) { + dev_warn_ratelimited(&dev->dev, "Dropping packet due to invalid special tag\n"); + dev_warn_ratelimited(&dev->dev, "Tag: %8ph\n", gsw1xx_tag); + return NULL; + } + + /* Get source port information */ + port = FIELD_GET(GSW1XX_RX_PORT_MAP, ntohs(gsw1xx_tag[1])); + skb->dev = dsa_conduit_find_user(dev, 0, port); + if (!skb->dev) { + dev_warn_ratelimited(&dev->dev, "Dropping packet due to invalid source port\n"); + dev_warn_ratelimited(&dev->dev, "Tag: %8ph\n", gsw1xx_tag); + return NULL; + } + + /* remove the GSW1xx special tag between MAC addresses and the current + * ethertype field. + */ + skb_pull_rcsum(skb, GSW1XX_HEADER_LEN); + dsa_strip_etype_header(skb, GSW1XX_HEADER_LEN); + + return skb; +} + +static const struct dsa_device_ops gsw1xx_netdev_ops = { + .name = GSW1XX_TAG_NAME, + .proto = DSA_TAG_PROTO_MXL_GSW1XX, + .xmit = gsw1xx_tag_xmit, + .rcv = gsw1xx_tag_rcv, + .needed_headroom = GSW1XX_HEADER_LEN, +}; + +MODULE_DESCRIPTION("DSA tag driver for MaxLinear GSW1xx 8 byte protocol"); +MODULE_LICENSE("GPL"); +MODULE_ALIAS_DSA_TAG_DRIVER(DSA_TAG_PROTO_MXL_GSW1XX, GSW1XX_TAG_NAME); + +module_dsa_tag_driver(gsw1xx_netdev_ops); -- cgit v1.2.3 From 24ab8efb9aea77764dd99d2bad41fd8991223013 Mon Sep 17 00:00:00 2001 From: Daniel Borkmann Date: Fri, 31 Oct 2025 22:20:55 +0100 Subject: xsk: Move NETDEV_XDP_ACT_ZC into generic header Move NETDEV_XDP_ACT_ZC into xdp_sock_drv.h header such that external code can reuse it, and rename it into more generic NETDEV_XDP_ACT_XSK. Signed-off-by: Daniel Borkmann Co-developed-by: David Wei Signed-off-by: David Wei Reviewed-by: Maciej Fijalkowski Reviewed-by: Nikolay Aleksandrov Link: https://patch.msgid.link/20251031212103.310683-7-daniel@iogearbox.net Signed-off-by: Jakub Kicinski --- include/net/xdp_sock_drv.h | 4 ++++ net/xdp/xsk_buff_pool.c | 6 +----- 2 files changed, 5 insertions(+), 5 deletions(-) (limited to 'include/net') diff --git a/include/net/xdp_sock_drv.h b/include/net/xdp_sock_drv.h index 4f2d3268a676..242e34f771cc 100644 --- a/include/net/xdp_sock_drv.h +++ b/include/net/xdp_sock_drv.h @@ -12,6 +12,10 @@ #define XDP_UMEM_MIN_CHUNK_SHIFT 11 #define XDP_UMEM_MIN_CHUNK_SIZE (1 << XDP_UMEM_MIN_CHUNK_SHIFT) +#define NETDEV_XDP_ACT_XSK (NETDEV_XDP_ACT_BASIC | \ + NETDEV_XDP_ACT_REDIRECT | \ + NETDEV_XDP_ACT_XSK_ZEROCOPY) + struct xsk_cb_desc { void *src; u8 off; diff --git a/net/xdp/xsk_buff_pool.c b/net/xdp/xsk_buff_pool.c index 00a4eddaa0cd..51526034c42a 100644 --- a/net/xdp/xsk_buff_pool.c +++ b/net/xdp/xsk_buff_pool.c @@ -155,10 +155,6 @@ static void xp_disable_drv_zc(struct xsk_buff_pool *pool) } } -#define NETDEV_XDP_ACT_ZC (NETDEV_XDP_ACT_BASIC | \ - NETDEV_XDP_ACT_REDIRECT | \ - NETDEV_XDP_ACT_XSK_ZEROCOPY) - int xp_assign_dev(struct xsk_buff_pool *pool, struct net_device *netdev, u16 queue_id, u16 flags) { @@ -200,7 +196,7 @@ int xp_assign_dev(struct xsk_buff_pool *pool, /* For copy-mode, we are done. */ return 0; - if ((netdev->xdp_features & NETDEV_XDP_ACT_ZC) != NETDEV_XDP_ACT_ZC) { + if ((netdev->xdp_features & NETDEV_XDP_ACT_XSK) != NETDEV_XDP_ACT_XSK) { err = -EOPNOTSUPP; goto err_unreg_pool; } -- cgit v1.2.3 From be88c549e9d78828a2e06126ed7e17fc2e030f1f Mon Sep 17 00:00:00 2001 From: Kuniyuki Iwashima Date: Thu, 6 Nov 2025 00:32:40 +0000 Subject: tcp: Call tcp_syn_ack_timeout() directly. Since DCCP has been removed, we do not need to use request_sock_ops.syn_ack_timeout(). Let's call tcp_syn_ack_timeout() directly. Now other function pointers of request_sock_ops are protocol-dependent. Signed-off-by: Kuniyuki Iwashima Reviewed-by: Eric Dumazet Link: https://patch.msgid.link/20251106003357.273403-2-kuniyu@google.com Signed-off-by: Jakub Kicinski --- include/net/request_sock.h | 1 - net/ipv4/inet_connection_sock.c | 4 +++- net/ipv4/tcp_ipv4.c | 1 - net/ipv4/tcp_timer.c | 3 +-- net/ipv6/tcp_ipv6.c | 1 - 5 files changed, 4 insertions(+), 6 deletions(-) (limited to 'include/net') diff --git a/include/net/request_sock.h b/include/net/request_sock.h index cd4d4cf71d0d..9b9e04f6bb89 100644 --- a/include/net/request_sock.h +++ b/include/net/request_sock.h @@ -36,7 +36,6 @@ struct request_sock_ops { struct sk_buff *skb, enum sk_rst_reason reason); void (*destructor)(struct request_sock *req); - void (*syn_ack_timeout)(const struct request_sock *req); }; struct saved_syn { diff --git a/net/ipv4/inet_connection_sock.c b/net/ipv4/inet_connection_sock.c index 3b83b66b2284..6a86c1ac3011 100644 --- a/net/ipv4/inet_connection_sock.c +++ b/net/ipv4/inet_connection_sock.c @@ -1096,9 +1096,11 @@ static void reqsk_timer_handler(struct timer_list *t) young <<= 1; } } + syn_ack_recalc(req, max_syn_ack_retries, READ_ONCE(queue->rskq_defer_accept), &expire, &resend); - req->rsk_ops->syn_ack_timeout(req); + tcp_syn_ack_timeout(req); + if (!expire && (!resend || !tcp_rtx_synack(sk_listener, req) || diff --git a/net/ipv4/tcp_ipv4.c b/net/ipv4/tcp_ipv4.c index b7526a7888cb..0cfebac33a91 100644 --- a/net/ipv4/tcp_ipv4.c +++ b/net/ipv4/tcp_ipv4.c @@ -1660,7 +1660,6 @@ struct request_sock_ops tcp_request_sock_ops __read_mostly = { .send_ack = tcp_v4_reqsk_send_ack, .destructor = tcp_v4_reqsk_destructor, .send_reset = tcp_v4_send_reset, - .syn_ack_timeout = tcp_syn_ack_timeout, }; const struct tcp_request_sock_ops tcp_request_sock_ipv4_ops = { diff --git a/net/ipv4/tcp_timer.c b/net/ipv4/tcp_timer.c index 2dd73a4e8e51..0672c3d8f4f1 100644 --- a/net/ipv4/tcp_timer.c +++ b/net/ipv4/tcp_timer.c @@ -458,7 +458,7 @@ static void tcp_fastopen_synack_timer(struct sock *sk, struct request_sock *req) struct tcp_sock *tp = tcp_sk(sk); int max_retries; - req->rsk_ops->syn_ack_timeout(req); + tcp_syn_ack_timeout(req); /* Add one more retry for fastopen. * Paired with WRITE_ONCE() in tcp_sock_set_syncnt() @@ -752,7 +752,6 @@ void tcp_syn_ack_timeout(const struct request_sock *req) __NET_INC_STATS(net, LINUX_MIB_TCPTIMEOUTS); } -EXPORT_IPV6_MOD(tcp_syn_ack_timeout); void tcp_reset_keepalive_timer(struct sock *sk, unsigned long len) { diff --git a/net/ipv6/tcp_ipv6.c b/net/ipv6/tcp_ipv6.c index 7df21c1cba21..08113f430124 100644 --- a/net/ipv6/tcp_ipv6.c +++ b/net/ipv6/tcp_ipv6.c @@ -796,7 +796,6 @@ struct request_sock_ops tcp6_request_sock_ops __read_mostly = { .send_ack = tcp_v6_reqsk_send_ack, .destructor = tcp_v6_reqsk_destructor, .send_reset = tcp_v6_send_reset, - .syn_ack_timeout = tcp_syn_ack_timeout, }; const struct tcp_request_sock_ops tcp_request_sock_ipv6_ops = { -- cgit v1.2.3 From 3ce5dd8161ecdf12ffe0af99ff8980f1432f64a5 Mon Sep 17 00:00:00 2001 From: Kuniyuki Iwashima Date: Thu, 6 Nov 2025 00:32:41 +0000 Subject: tcp: Remove timeout arg from reqsk_queue_hash_req(). inet_csk_reqsk_queue_hash_add() is no longer shared by DCCP. We do not need to pass req->timeout down to reqsk_queue_hash_req(). Let's move tcp_timeout_init() from tcp_conn_request() to reqsk_queue_hash_req(). Signed-off-by: Kuniyuki Iwashima Reviewed-by: Eric Dumazet Link: https://patch.msgid.link/20251106003357.273403-3-kuniyu@google.com Signed-off-by: Jakub Kicinski --- include/net/inet_connection_sock.h | 3 +-- net/ipv4/inet_connection_sock.c | 11 +++++------ net/ipv4/tcp_input.c | 14 +++++--------- 3 files changed, 11 insertions(+), 17 deletions(-) (limited to 'include/net') diff --git a/include/net/inet_connection_sock.h b/include/net/inet_connection_sock.h index b4b886647607..90a99a2fc804 100644 --- a/include/net/inet_connection_sock.h +++ b/include/net/inet_connection_sock.h @@ -267,8 +267,7 @@ struct dst_entry *inet_csk_route_child_sock(const struct sock *sk, struct sock *inet_csk_reqsk_queue_add(struct sock *sk, struct request_sock *req, struct sock *child); -bool inet_csk_reqsk_queue_hash_add(struct sock *sk, struct request_sock *req, - unsigned long timeout); +bool inet_csk_reqsk_queue_hash_add(struct sock *sk, struct request_sock *req); struct sock *inet_csk_complete_hashdance(struct sock *sk, struct sock *child, struct request_sock *req, bool own_req); diff --git a/net/ipv4/inet_connection_sock.c b/net/ipv4/inet_connection_sock.c index 6a86c1ac3011..d9c674403eb0 100644 --- a/net/ipv4/inet_connection_sock.c +++ b/net/ipv4/inet_connection_sock.c @@ -1144,8 +1144,7 @@ drop: reqsk_put(oreq); } -static bool reqsk_queue_hash_req(struct request_sock *req, - unsigned long timeout) +static bool reqsk_queue_hash_req(struct request_sock *req) { bool found_dup_sk = false; @@ -1153,8 +1152,9 @@ static bool reqsk_queue_hash_req(struct request_sock *req, return false; /* The timer needs to be setup after a successful insertion. */ + req->timeout = tcp_timeout_init((struct sock *)req); timer_setup(&req->rsk_timer, reqsk_timer_handler, TIMER_PINNED); - mod_timer(&req->rsk_timer, jiffies + timeout); + mod_timer(&req->rsk_timer, jiffies + req->timeout); /* before letting lookups find us, make sure all req fields * are committed to memory and refcnt initialized. @@ -1164,10 +1164,9 @@ static bool reqsk_queue_hash_req(struct request_sock *req, return true; } -bool inet_csk_reqsk_queue_hash_add(struct sock *sk, struct request_sock *req, - unsigned long timeout) +bool inet_csk_reqsk_queue_hash_add(struct sock *sk, struct request_sock *req) { - if (!reqsk_queue_hash_req(req, timeout)) + if (!reqsk_queue_hash_req(req)) return false; inet_csk_reqsk_queue_added(sk); diff --git a/net/ipv4/tcp_input.c b/net/ipv4/tcp_input.c index 6db1d4c36a88..804ec56bdd24 100644 --- a/net/ipv4/tcp_input.c +++ b/net/ipv4/tcp_input.c @@ -7531,15 +7531,11 @@ int tcp_conn_request(struct request_sock_ops *rsk_ops, sock_put(fastopen_sk); } else { tcp_rsk(req)->tfo_listener = false; - if (!want_cookie) { - req->timeout = tcp_timeout_init((struct sock *)req); - if (unlikely(!inet_csk_reqsk_queue_hash_add(sk, req, - req->timeout))) { - reqsk_free(req); - dst_release(dst); - return 0; - } - + if (!want_cookie && + unlikely(!inet_csk_reqsk_queue_hash_add(sk, req))) { + reqsk_free(req); + dst_release(dst); + return 0; } af_ops->send_synack(sk, dst, &fl, req, &foc, !want_cookie ? TCP_SYNACK_NORMAL : -- cgit v1.2.3 From 207ce0f6bc131812c96cf4f6db328af5396cebac Mon Sep 17 00:00:00 2001 From: Kuniyuki Iwashima Date: Thu, 6 Nov 2025 00:32:43 +0000 Subject: tcp: Remove timeout arg from reqsk_timeout(). reqsk_timeout() is always called with @timeout being TCP_RTO_MAX. Let's remove the arg. As a prep for the next patch, reqsk_timeout() is moved to tcp.h and renamed to tcp_reqsk_timeout(). Signed-off-by: Kuniyuki Iwashima Reviewed-by: Eric Dumazet Link: https://patch.msgid.link/20251106003357.273403-5-kuniyu@google.com Signed-off-by: Jakub Kicinski --- include/net/inet_connection_sock.h | 8 -------- include/net/tcp.h | 7 +++++++ net/ipv4/inet_connection_sock.c | 2 +- net/ipv4/tcp_minisocks.c | 5 +++-- 4 files changed, 11 insertions(+), 11 deletions(-) (limited to 'include/net') diff --git a/include/net/inet_connection_sock.h b/include/net/inet_connection_sock.h index 90a99a2fc804..fd40af2221b9 100644 --- a/include/net/inet_connection_sock.h +++ b/include/net/inet_connection_sock.h @@ -290,14 +290,6 @@ static inline int inet_csk_reqsk_queue_is_full(const struct sock *sk) bool inet_csk_reqsk_queue_drop(struct sock *sk, struct request_sock *req); void inet_csk_reqsk_queue_drop_and_put(struct sock *sk, struct request_sock *req); -static inline unsigned long -reqsk_timeout(struct request_sock *req, unsigned long max_timeout) -{ - u64 timeout = (u64)req->timeout << req->num_timeout; - - return (unsigned long)min_t(u64, timeout, max_timeout); -} - void inet_csk_destroy_sock(struct sock *sk); void inet_csk_prepare_for_destroy_sock(struct sock *sk); void inet_csk_prepare_forced_close(struct sock *sk); diff --git a/include/net/tcp.h b/include/net/tcp.h index 0aa1f07d036a..0c7274ac7ed5 100644 --- a/include/net/tcp.h +++ b/include/net/tcp.h @@ -841,6 +841,13 @@ static inline u32 __tcp_set_rto(const struct tcp_sock *tp) return usecs_to_jiffies((tp->srtt_us >> 3) + tp->rttvar_us); } +static inline unsigned long tcp_reqsk_timeout(struct request_sock *req) +{ + u64 timeout = (u64)req->timeout << req->num_timeout; + + return (unsigned long)min_t(u64, timeout, TCP_RTO_MAX); +} + u32 tcp_delack_max(const struct sock *sk); /* Compute the actual rto_min value */ diff --git a/net/ipv4/inet_connection_sock.c b/net/ipv4/inet_connection_sock.c index 2bfe7af51bbb..b4eae731c9ba 100644 --- a/net/ipv4/inet_connection_sock.c +++ b/net/ipv4/inet_connection_sock.c @@ -1105,7 +1105,7 @@ static void reqsk_timer_handler(struct timer_list *t) inet_rsk(req)->acked)) { if (req->num_timeout++ == 0) atomic_dec(&queue->young); - mod_timer(&req->rsk_timer, jiffies + reqsk_timeout(req, TCP_RTO_MAX)); + mod_timer(&req->rsk_timer, jiffies + tcp_reqsk_timeout(req)); if (!nreq) return; diff --git a/net/ipv4/tcp_minisocks.c b/net/ipv4/tcp_minisocks.c index ded2cf1f6006..d8f4d813e8dd 100644 --- a/net/ipv4/tcp_minisocks.c +++ b/net/ipv4/tcp_minisocks.c @@ -714,7 +714,8 @@ struct sock *tcp_check_req(struct sock *sk, struct sk_buff *skb, * it can be estimated (approximately) * from another data. */ - tmp_opt.ts_recent_stamp = ktime_get_seconds() - reqsk_timeout(req, TCP_RTO_MAX) / HZ; + tmp_opt.ts_recent_stamp = ktime_get_seconds() - + tcp_reqsk_timeout(req) / HZ; paws_reject = tcp_paws_reject(&tmp_opt, th->rst); } } @@ -753,7 +754,7 @@ struct sock *tcp_check_req(struct sock *sk, struct sk_buff *skb, !tcp_rtx_synack(sk, req)) { unsigned long expires = jiffies; - expires += reqsk_timeout(req, TCP_RTO_MAX); + expires += tcp_reqsk_timeout(req); if (!fastopen) mod_timer_pending(&req->rsk_timer, expires); else -- cgit v1.2.3 From 1e9d3005e02cba82047d49f859982fc73b9a100b Mon Sep 17 00:00:00 2001 From: Kuniyuki Iwashima Date: Thu, 6 Nov 2025 00:32:44 +0000 Subject: tcp: Apply max RTO to non-TFO SYN+ACK. Since commit 54a378f43425 ("tcp: add the ability to control max RTO"), TFO SYN+ACK RTO is capped by the TFO full sk's inet_csk(sk)->icsk_rto_max. The value is inherited from the parent listener. Let's apply the same cap to non-TFO SYN+ACK. Note that req->rsk_listener is always non-NULL when we call tcp_reqsk_timeout() in reqsk_timer_handler() or tcp_check_req(). It could be NULL for SYN cookie req, but we do not use req->timeout then. Signed-off-by: Kuniyuki Iwashima Reviewed-by: Eric Dumazet Link: https://patch.msgid.link/20251106003357.273403-6-kuniyu@google.com Signed-off-by: Jakub Kicinski --- include/net/tcp.h | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) (limited to 'include/net') diff --git a/include/net/tcp.h b/include/net/tcp.h index 0c7274ac7ed5..4833ec7903ec 100644 --- a/include/net/tcp.h +++ b/include/net/tcp.h @@ -845,7 +845,8 @@ static inline unsigned long tcp_reqsk_timeout(struct request_sock *req) { u64 timeout = (u64)req->timeout << req->num_timeout; - return (unsigned long)min_t(u64, timeout, TCP_RTO_MAX); + return (unsigned long)min_t(u64, timeout, + tcp_rto_max(req->rsk_listener)); } u32 tcp_delack_max(const struct sock *sk); -- cgit v1.2.3 From 416dd649f3aa3774907c668167a29c668dbc634b Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Thu, 6 Nov 2025 11:52:36 +0000 Subject: tcp: add net.ipv4.tcp_comp_sack_rtt_percent TCP SACK compression has been added in 2018 in commit 5d9f4262b7ea ("tcp: add SACK compression"). It is working great for WAN flows (with large RTT). Wifi in particular gets a significant boost _when_ ACK are suppressed. Add a new sysctl so that we can tune the very conservative 5 % value that has been used so far in this formula, so that small RTT flows can benefit from this feature. delay = min ( 5 % of RTT, 1 ms) This patch adds new tcp_comp_sack_rtt_percent sysctl to ease experiments and tuning. Given that we cap the delay to 1ms (tcp_comp_sack_delay_ns sysctl), set the default value to 33 %. Quoting Neal Cardwell ( https://lore.kernel.org/netdev/CADVnQymZ1tFnEA1Q=vtECs0=Db7zHQ8=+WCQtnhHFVbEOzjVnQ@mail.gmail.com/ ) The rationale for 33% is basically to try to facilitate pipelining, where there are always at least 3 ACKs and 3 GSO/TSO skbs per SRTT, so that the path can maintain a budget for 3 full-sized GSO/TSO skbs "in flight" at all times: + 1 skb in the qdisc waiting to be sent by the NIC next + 1 skb being sent by the NIC (being serialized by the NIC out onto the wire) + 1 skb being received and aggregated by the receiver machine's aggregation mechanism (some combination of LRO, GRO, and sack compression) Note that this is basically the same magic number (3) and the same rationales as: (a) tcp_tso_should_defer() ensuring that we defer sending data for no longer than cwnd/tcp_tso_win_divisor (where tcp_tso_win_divisor = 3), and (b) bbr_quantization_budget() ensuring that cwnd is at least 3 GSO/TSO skbs to maintain pipelining and full throughput at low RTTs Signed-off-by: Eric Dumazet Reviewed-by: Neal Cardwell Link: https://patch.msgid.link/20251106115236.3450026-1-edumazet@google.com Signed-off-by: Jakub Kicinski --- Documentation/networking/ip-sysctl.rst | 13 +++++++++++-- include/net/netns/ipv4.h | 1 + net/ipv4/sysctl_net_ipv4.c | 9 +++++++++ net/ipv4/tcp_input.c | 26 ++++++++++++++++++-------- net/ipv4/tcp_ipv4.c | 1 + 5 files changed, 40 insertions(+), 10 deletions(-) (limited to 'include/net') diff --git a/Documentation/networking/ip-sysctl.rst b/Documentation/networking/ip-sysctl.rst index 7cd35bfd39e6..2bae61be1859 100644 --- a/Documentation/networking/ip-sysctl.rst +++ b/Documentation/networking/ip-sysctl.rst @@ -854,9 +854,18 @@ tcp_sack - BOOLEAN Default: 1 (enabled) +tcp_comp_sack_rtt_percent - INTEGER + Percentage of SRTT used for the compressed SACK feature. + See tcp_comp_sack_nr, tcp_comp_sack_delay_ns, tcp_comp_sack_slack_ns. + + Possible values : 1 - 1000 + + Default : 33 % + tcp_comp_sack_delay_ns - LONG INTEGER - TCP tries to reduce number of SACK sent, using a timer - based on 5% of SRTT, capped by this sysctl, in nano seconds. + TCP tries to reduce number of SACK sent, using a timer based + on tcp_comp_sack_rtt_percent of SRTT, capped by this sysctl + in nano seconds. The default is 1ms, based on TSO autosizing period. Default : 1,000,000 ns (1 ms) diff --git a/include/net/netns/ipv4.h b/include/net/netns/ipv4.h index 0e96c90e56c6..de9d36acc8e2 100644 --- a/include/net/netns/ipv4.h +++ b/include/net/netns/ipv4.h @@ -221,6 +221,7 @@ struct netns_ipv4 { int sysctl_tcp_pacing_ss_ratio; int sysctl_tcp_pacing_ca_ratio; unsigned int sysctl_tcp_child_ehash_entries; + int sysctl_tcp_comp_sack_rtt_percent; unsigned long sysctl_tcp_comp_sack_delay_ns; unsigned long sysctl_tcp_comp_sack_slack_ns; int sysctl_max_syn_backlog; diff --git a/net/ipv4/sysctl_net_ipv4.c b/net/ipv4/sysctl_net_ipv4.c index 0c7c8f9041cb..35367f8e2da3 100644 --- a/net/ipv4/sysctl_net_ipv4.c +++ b/net/ipv4/sysctl_net_ipv4.c @@ -1451,6 +1451,15 @@ static struct ctl_table ipv4_net_table[] = { .mode = 0644, .proc_handler = proc_doulongvec_minmax, }, + { + .procname = "tcp_comp_sack_rtt_percent", + .data = &init_net.ipv4.sysctl_tcp_comp_sack_rtt_percent, + .maxlen = sizeof(int), + .mode = 0644, + .proc_handler = proc_dointvec_minmax, + .extra1 = SYSCTL_ONE, + .extra2 = SYSCTL_ONE_THOUSAND, + }, { .procname = "tcp_comp_sack_slack_ns", .data = &init_net.ipv4.sysctl_tcp_comp_sack_slack_ns, diff --git a/net/ipv4/tcp_input.c b/net/ipv4/tcp_input.c index 804ec56bdd24..9df5d7515605 100644 --- a/net/ipv4/tcp_input.c +++ b/net/ipv4/tcp_input.c @@ -5893,7 +5893,9 @@ static inline void tcp_data_snd_check(struct sock *sk) static void __tcp_ack_snd_check(struct sock *sk, int ofo_possible) { struct tcp_sock *tp = tcp_sk(sk); - unsigned long rtt, delay; + struct net *net = sock_net(sk); + unsigned long rtt; + u64 delay; /* More than one full frame received... */ if (((tp->rcv_nxt - tp->rcv_wup) > inet_csk(sk)->icsk_ack.rcv_mss && @@ -5912,7 +5914,7 @@ static void __tcp_ack_snd_check(struct sock *sk, int ofo_possible) * Defer the ack until tcp_release_cb(). */ if (sock_owned_by_user_nocheck(sk) && - READ_ONCE(sock_net(sk)->ipv4.sysctl_tcp_backlog_ack_defer)) { + READ_ONCE(net->ipv4.sysctl_tcp_backlog_ack_defer)) { set_bit(TCP_ACK_DEFERRED, &sk->sk_tsq_flags); return; } @@ -5927,7 +5929,7 @@ send_now: } if (!tcp_is_sack(tp) || - tp->compressed_ack >= READ_ONCE(sock_net(sk)->ipv4.sysctl_tcp_comp_sack_nr)) + tp->compressed_ack >= READ_ONCE(net->ipv4.sysctl_tcp_comp_sack_nr)) goto send_now; if (tp->compressed_ack_rcv_nxt != tp->rcv_nxt) { @@ -5942,18 +5944,26 @@ send_now: if (hrtimer_is_queued(&tp->compressed_ack_timer)) return; - /* compress ack timer : 5 % of rtt, but no more than tcp_comp_sack_delay_ns */ + /* compress ack timer : comp_sack_rtt_percent of rtt, + * but no more than tcp_comp_sack_delay_ns. + */ rtt = tp->rcv_rtt_est.rtt_us; if (tp->srtt_us && tp->srtt_us < rtt) rtt = tp->srtt_us; - delay = min_t(unsigned long, - READ_ONCE(sock_net(sk)->ipv4.sysctl_tcp_comp_sack_delay_ns), - rtt * (NSEC_PER_USEC >> 3)/20); + /* delay = (rtt >> 3) * NSEC_PER_USEC * comp_sack_rtt_percent / 100 + * -> + * delay = rtt * 1.25 * comp_sack_rtt_percent + */ + delay = (u64)(rtt + (rtt >> 2)) * + READ_ONCE(net->ipv4.sysctl_tcp_comp_sack_rtt_percent); + + delay = min(delay, READ_ONCE(net->ipv4.sysctl_tcp_comp_sack_delay_ns)); + sock_hold(sk); hrtimer_start_range_ns(&tp->compressed_ack_timer, ns_to_ktime(delay), - READ_ONCE(sock_net(sk)->ipv4.sysctl_tcp_comp_sack_slack_ns), + READ_ONCE(net->ipv4.sysctl_tcp_comp_sack_slack_ns), HRTIMER_MODE_REL_PINNED_SOFT); } diff --git a/net/ipv4/tcp_ipv4.c b/net/ipv4/tcp_ipv4.c index 0cfebac33a91..a7d9fec2950b 100644 --- a/net/ipv4/tcp_ipv4.c +++ b/net/ipv4/tcp_ipv4.c @@ -3595,6 +3595,7 @@ static int __net_init tcp_sk_init(struct net *net) net->ipv4.sysctl_tcp_comp_sack_delay_ns = NSEC_PER_MSEC; net->ipv4.sysctl_tcp_comp_sack_slack_ns = 100 * NSEC_PER_USEC; net->ipv4.sysctl_tcp_comp_sack_nr = 44; + net->ipv4.sysctl_tcp_comp_sack_rtt_percent = 33; net->ipv4.sysctl_tcp_backlog_ack_defer = 1; net->ipv4.sysctl_tcp_fastopen = TFO_CLIENT_ENABLE; net->ipv4.sysctl_tcp_fastopen_blackhole_timeout = 0; -- cgit v1.2.3 From dae4a92399fa8d68aa917db6bb3245f83021e762 Mon Sep 17 00:00:00 2001 From: Jakub Kicinski Date: Wed, 5 Nov 2025 16:26:02 -0800 Subject: psp: report basic stats from the core Track and report stats common to all psp devices from the core. A 'stale-event' is when the core marks the rx state of an active psp_assoc as incapable of authenticating psp encapsulated data. Signed-off-by: Daniel Zahka Link: https://patch.msgid.link/20251106002608.1578518-2-daniel.zahka@gmail.com Signed-off-by: Jakub Kicinski --- Documentation/netlink/specs/psp.yaml | 40 +++++++++++++++++++ include/net/psp/types.h | 9 +++++ include/uapi/linux/psp.h | 10 +++++ net/psp/psp-nl-gen.c | 19 +++++++++ net/psp/psp-nl-gen.h | 2 + net/psp/psp_nl.c | 74 ++++++++++++++++++++++++++++++++++++ net/psp/psp_sock.c | 4 +- 7 files changed, 157 insertions(+), 1 deletion(-) (limited to 'include/net') diff --git a/Documentation/netlink/specs/psp.yaml b/Documentation/netlink/specs/psp.yaml index 944429e5c9a8..914148221384 100644 --- a/Documentation/netlink/specs/psp.yaml +++ b/Documentation/netlink/specs/psp.yaml @@ -76,6 +76,28 @@ attribute-sets: name: spi doc: Security Parameters Index (SPI) of the association. type: u32 + - + name: stats + attributes: + - + name: dev-id + doc: PSP device ID. + type: u32 + checks: + min: 1 + - + name: key-rotations + type: uint + doc: | + Number of key rotations during the lifetime of the device. + Kernel statistic. + - + name: stale-events + type: uint + doc: | + Number of times a socket's Rx got shut down due to using + a key which went stale (fully rotated out). + Kernel statistic. operations: list: @@ -177,6 +199,24 @@ operations: pre: psp-assoc-device-get-locked post: psp-device-unlock + - + name: get-stats + doc: Get device statistics. + attribute-set: stats + do: + request: + attributes: + - dev-id + reply: &stats-all + attributes: + - dev-id + - key-rotations + - stale-events + pre: psp-device-get-locked + post: psp-device-unlock + dump: + reply: *stats-all + mcast-groups: list: - diff --git a/include/net/psp/types.h b/include/net/psp/types.h index 31cee64b7c86..5b0ccaac3882 100644 --- a/include/net/psp/types.h +++ b/include/net/psp/types.h @@ -59,6 +59,10 @@ struct psp_dev_config { * device key * @stale_assocs: associations which use a rotated out key * + * @stats: statistics maintained by the core + * @stats.rotations: See stats attr key-rotations + * @stats.stales: See stats attr stale-events + * * @rcu: RCU head for freeing the structure */ struct psp_dev { @@ -81,6 +85,11 @@ struct psp_dev { struct list_head prev_assocs; struct list_head stale_assocs; + struct { + unsigned long rotations; + unsigned long stales; + } stats; + struct rcu_head rcu; }; diff --git a/include/uapi/linux/psp.h b/include/uapi/linux/psp.h index 607c42c39ba5..31592760ad79 100644 --- a/include/uapi/linux/psp.h +++ b/include/uapi/linux/psp.h @@ -45,6 +45,15 @@ enum { PSP_A_KEYS_MAX = (__PSP_A_KEYS_MAX - 1) }; +enum { + PSP_A_STATS_DEV_ID = 1, + PSP_A_STATS_KEY_ROTATIONS, + PSP_A_STATS_STALE_EVENTS, + + __PSP_A_STATS_MAX, + PSP_A_STATS_MAX = (__PSP_A_STATS_MAX - 1) +}; + enum { PSP_CMD_DEV_GET = 1, PSP_CMD_DEV_ADD_NTF, @@ -55,6 +64,7 @@ enum { PSP_CMD_KEY_ROTATE_NTF, PSP_CMD_RX_ASSOC, PSP_CMD_TX_ASSOC, + PSP_CMD_GET_STATS, __PSP_CMD_MAX, PSP_CMD_MAX = (__PSP_CMD_MAX - 1) diff --git a/net/psp/psp-nl-gen.c b/net/psp/psp-nl-gen.c index 9fdd6f831803..73f8b06d66f0 100644 --- a/net/psp/psp-nl-gen.c +++ b/net/psp/psp-nl-gen.c @@ -47,6 +47,11 @@ static const struct nla_policy psp_tx_assoc_nl_policy[PSP_A_ASSOC_SOCK_FD + 1] = [PSP_A_ASSOC_SOCK_FD] = { .type = NLA_U32, }, }; +/* PSP_CMD_GET_STATS - do */ +static const struct nla_policy psp_get_stats_nl_policy[PSP_A_STATS_DEV_ID + 1] = { + [PSP_A_STATS_DEV_ID] = NLA_POLICY_MIN(NLA_U32, 1), +}; + /* Ops table for psp */ static const struct genl_split_ops psp_nl_ops[] = { { @@ -99,6 +104,20 @@ static const struct genl_split_ops psp_nl_ops[] = { .maxattr = PSP_A_ASSOC_SOCK_FD, .flags = GENL_CMD_CAP_DO, }, + { + .cmd = PSP_CMD_GET_STATS, + .pre_doit = psp_device_get_locked, + .doit = psp_nl_get_stats_doit, + .post_doit = psp_device_unlock, + .policy = psp_get_stats_nl_policy, + .maxattr = PSP_A_STATS_DEV_ID, + .flags = GENL_CMD_CAP_DO, + }, + { + .cmd = PSP_CMD_GET_STATS, + .dumpit = psp_nl_get_stats_dumpit, + .flags = GENL_CMD_CAP_DUMP, + }, }; static const struct genl_multicast_group psp_nl_mcgrps[] = { diff --git a/net/psp/psp-nl-gen.h b/net/psp/psp-nl-gen.h index 25268ed11fb5..5bc3b5d5a53e 100644 --- a/net/psp/psp-nl-gen.h +++ b/net/psp/psp-nl-gen.h @@ -28,6 +28,8 @@ int psp_nl_dev_set_doit(struct sk_buff *skb, struct genl_info *info); int psp_nl_key_rotate_doit(struct sk_buff *skb, struct genl_info *info); int psp_nl_rx_assoc_doit(struct sk_buff *skb, struct genl_info *info); int psp_nl_tx_assoc_doit(struct sk_buff *skb, struct genl_info *info); +int psp_nl_get_stats_doit(struct sk_buff *skb, struct genl_info *info); +int psp_nl_get_stats_dumpit(struct sk_buff *skb, struct netlink_callback *cb); enum { PSP_NLGRP_MGMT, diff --git a/net/psp/psp_nl.c b/net/psp/psp_nl.c index 8aaca62744c3..f990cccbe99c 100644 --- a/net/psp/psp_nl.c +++ b/net/psp/psp_nl.c @@ -262,6 +262,7 @@ int psp_nl_key_rotate_doit(struct sk_buff *skb, struct genl_info *info) psd->generation & ~PSP_GEN_VALID_MASK); psp_assocs_key_rotated(psd); + psd->stats.rotations++; nlmsg_end(ntf, (struct nlmsghdr *)ntf->data); genlmsg_multicast_netns(&psp_nl_family, dev_net(psd->main_netdev), ntf, @@ -503,3 +504,76 @@ err_free_msg: nlmsg_free(rsp); return err; } + +static int +psp_nl_stats_fill(struct psp_dev *psd, struct sk_buff *rsp, + const struct genl_info *info) +{ + void *hdr; + + hdr = genlmsg_iput(rsp, info); + if (!hdr) + return -EMSGSIZE; + + if (nla_put_u32(rsp, PSP_A_STATS_DEV_ID, psd->id) || + nla_put_uint(rsp, PSP_A_STATS_KEY_ROTATIONS, + psd->stats.rotations) || + nla_put_uint(rsp, PSP_A_STATS_STALE_EVENTS, psd->stats.stales)) + goto err_cancel_msg; + + genlmsg_end(rsp, hdr); + return 0; + +err_cancel_msg: + genlmsg_cancel(rsp, hdr); + return -EMSGSIZE; +} + +int psp_nl_get_stats_doit(struct sk_buff *skb, struct genl_info *info) +{ + struct psp_dev *psd = info->user_ptr[0]; + struct sk_buff *rsp; + int err; + + rsp = genlmsg_new(GENLMSG_DEFAULT_SIZE, GFP_KERNEL); + if (!rsp) + return -ENOMEM; + + err = psp_nl_stats_fill(psd, rsp, info); + if (err) + goto err_free_msg; + + return genlmsg_reply(rsp, info); + +err_free_msg: + nlmsg_free(rsp); + return err; +} + +static int +psp_nl_stats_get_dumpit_one(struct sk_buff *rsp, struct netlink_callback *cb, + struct psp_dev *psd) +{ + if (psp_dev_check_access(psd, sock_net(rsp->sk))) + return 0; + + return psp_nl_stats_fill(psd, rsp, genl_info_dump(cb)); +} + +int psp_nl_get_stats_dumpit(struct sk_buff *rsp, struct netlink_callback *cb) +{ + struct psp_dev *psd; + int err = 0; + + mutex_lock(&psp_devs_lock); + xa_for_each_start(&psp_devs, cb->args[0], psd, cb->args[0]) { + mutex_lock(&psd->lock); + err = psp_nl_stats_get_dumpit_one(rsp, cb, psd); + mutex_unlock(&psd->lock); + if (err) + break; + } + mutex_unlock(&psp_devs_lock); + + return err; +} diff --git a/net/psp/psp_sock.c b/net/psp/psp_sock.c index a931d825d1cc..f785672b7df6 100644 --- a/net/psp/psp_sock.c +++ b/net/psp/psp_sock.c @@ -253,8 +253,10 @@ void psp_assocs_key_rotated(struct psp_dev *psd) /* Mark the stale associations as invalid, they will no longer * be able to Rx any traffic. */ - list_for_each_entry_safe(pas, next, &psd->prev_assocs, assocs_list) + list_for_each_entry_safe(pas, next, &psd->prev_assocs, assocs_list) { pas->generation |= ~PSP_GEN_VALID_MASK; + psd->stats.stales++; + } list_splice_init(&psd->prev_assocs, &psd->stale_assocs); list_splice_init(&psd->active_assocs, &psd->prev_assocs); -- cgit v1.2.3 From f05d26198cf2c71f25f6bbe62ca4481c15543922 Mon Sep 17 00:00:00 2001 From: Jakub Kicinski Date: Wed, 5 Nov 2025 16:26:04 -0800 Subject: psp: add stats from psp spec to driver facing api Provide a driver api for reporting device statistics required by the "Implementation Requirements" section of the PSP Architecture Specification. Use a warning to ensure drivers report stats required by the spec. Signed-off-by: Daniel Zahka Link: https://patch.msgid.link/20251106002608.1578518-4-daniel.zahka@gmail.com Signed-off-by: Jakub Kicinski --- Documentation/netlink/specs/psp.yaml | 55 ++++++++++++++++++++++++++++++++++++ include/net/psp/types.h | 23 +++++++++++++++ include/uapi/linux/psp.h | 8 ++++++ net/psp/psp_main.c | 3 +- net/psp/psp_nl.c | 21 +++++++++++++- 5 files changed, 108 insertions(+), 2 deletions(-) (limited to 'include/net') diff --git a/Documentation/netlink/specs/psp.yaml b/Documentation/netlink/specs/psp.yaml index 914148221384..f3a57782d2cf 100644 --- a/Documentation/netlink/specs/psp.yaml +++ b/Documentation/netlink/specs/psp.yaml @@ -98,6 +98,61 @@ attribute-sets: Number of times a socket's Rx got shut down due to using a key which went stale (fully rotated out). Kernel statistic. + - + name: rx-packets + type: uint + doc: | + Number of successfully processed and authenticated PSP packets. + Device statistic (from the PSP spec). + - + name: rx-bytes + type: uint + doc: | + Number of successfully authenticated PSP bytes received, counting from + the first byte after the IV through the last byte of payload. + The fixed initial portion of the PSP header (16 bytes) + and the PSP trailer/ICV (16 bytes) are not included in this count. + Device statistic (from the PSP spec). + - + name: rx-auth-fail + type: uint + doc: | + Number of received PSP packets with unsuccessful authentication. + Device statistic (from the PSP spec). + - + name: rx-error + type: uint + doc: | + Number of received PSP packets with length/framing errors. + Device statistic (from the PSP spec). + - + name: rx-bad + type: uint + doc: | + Number of received PSP packets with miscellaneous errors + (invalid master key indicated by SPI, unsupported version, etc.) + Device statistic (from the PSP spec). + - + name: tx-packets + type: uint + doc: | + Number of successfully processed PSP packets for transmission. + Device statistic (from the PSP spec). + - + name: tx-bytes + type: uint + doc: | + Number of successfully processed PSP bytes for transmit, counting from + the first byte after the IV through the last byte of payload. + The fixed initial portion of the PSP header (16 bytes) + and the PSP trailer/ICV (16 bytes) are not included in this count. + Device statistic (from the PSP spec). + - + name: tx-error + type: uint + doc: | + Number of PSP packets for transmission with errors. + Device statistic (from the PSP spec). operations: list: diff --git a/include/net/psp/types.h b/include/net/psp/types.h index 5b0ccaac3882..25a9096d4e7d 100644 --- a/include/net/psp/types.h +++ b/include/net/psp/types.h @@ -150,6 +150,22 @@ struct psp_assoc { u8 drv_data[] __aligned(8); }; +struct psp_dev_stats { + union { + struct { + u64 rx_packets; + u64 rx_bytes; + u64 rx_auth_fail; + u64 rx_error; + u64 rx_bad; + u64 tx_packets; + u64 tx_bytes; + u64 tx_error; + }; + DECLARE_FLEX_ARRAY(u64, required); + }; +}; + /** * struct psp_dev_ops - netdev driver facing PSP callbacks */ @@ -188,6 +204,13 @@ struct psp_dev_ops { * Remove an association from the device. */ void (*tx_key_del)(struct psp_dev *psd, struct psp_assoc *pas); + + /** + * @get_stats: get statistics from the device + * Stats required by the spec must be maintained and filled in. + * Stats must be filled in member-by-member, never memset the struct. + */ + void (*get_stats)(struct psp_dev *psd, struct psp_dev_stats *stats); }; #endif /* __NET_PSP_H */ diff --git a/include/uapi/linux/psp.h b/include/uapi/linux/psp.h index 31592760ad79..d8449c043ba1 100644 --- a/include/uapi/linux/psp.h +++ b/include/uapi/linux/psp.h @@ -49,6 +49,14 @@ enum { PSP_A_STATS_DEV_ID = 1, PSP_A_STATS_KEY_ROTATIONS, PSP_A_STATS_STALE_EVENTS, + PSP_A_STATS_RX_PACKETS, + PSP_A_STATS_RX_BYTES, + PSP_A_STATS_RX_AUTH_FAIL, + PSP_A_STATS_RX_ERROR, + PSP_A_STATS_RX_BAD, + PSP_A_STATS_TX_PACKETS, + PSP_A_STATS_TX_BYTES, + PSP_A_STATS_TX_ERROR, __PSP_A_STATS_MAX, PSP_A_STATS_MAX = (__PSP_A_STATS_MAX - 1) diff --git a/net/psp/psp_main.c b/net/psp/psp_main.c index 481aaf0fc9fc..a8534124f626 100644 --- a/net/psp/psp_main.c +++ b/net/psp/psp_main.c @@ -60,7 +60,8 @@ psp_dev_create(struct net_device *netdev, !psd_ops->key_rotate || !psd_ops->rx_spi_alloc || !psd_ops->tx_key_add || - !psd_ops->tx_key_del)) + !psd_ops->tx_key_del || + !psd_ops->get_stats)) return ERR_PTR(-EINVAL); psd = kzalloc(sizeof(*psd), GFP_KERNEL); diff --git a/net/psp/psp_nl.c b/net/psp/psp_nl.c index f990cccbe99c..6afd7707ec12 100644 --- a/net/psp/psp_nl.c +++ b/net/psp/psp_nl.c @@ -1,5 +1,6 @@ // SPDX-License-Identifier: GPL-2.0-only +#include #include #include #include @@ -509,7 +510,17 @@ static int psp_nl_stats_fill(struct psp_dev *psd, struct sk_buff *rsp, const struct genl_info *info) { + unsigned int required_cnt = sizeof(struct psp_dev_stats) / sizeof(u64); + struct psp_dev_stats stats; void *hdr; + int i; + + memset(&stats, 0xff, sizeof(stats)); + psd->ops->get_stats(psd, &stats); + + for (i = 0; i < required_cnt; i++) + if (WARN_ON_ONCE(stats.required[i] == ETHTOOL_STAT_NOT_SET)) + return -EOPNOTSUPP; hdr = genlmsg_iput(rsp, info); if (!hdr) @@ -518,7 +529,15 @@ psp_nl_stats_fill(struct psp_dev *psd, struct sk_buff *rsp, if (nla_put_u32(rsp, PSP_A_STATS_DEV_ID, psd->id) || nla_put_uint(rsp, PSP_A_STATS_KEY_ROTATIONS, psd->stats.rotations) || - nla_put_uint(rsp, PSP_A_STATS_STALE_EVENTS, psd->stats.stales)) + nla_put_uint(rsp, PSP_A_STATS_STALE_EVENTS, psd->stats.stales) || + nla_put_uint(rsp, PSP_A_STATS_RX_PACKETS, stats.rx_packets) || + nla_put_uint(rsp, PSP_A_STATS_RX_BYTES, stats.rx_bytes) || + nla_put_uint(rsp, PSP_A_STATS_RX_AUTH_FAIL, stats.rx_auth_fail) || + nla_put_uint(rsp, PSP_A_STATS_RX_ERROR, stats.rx_error) || + nla_put_uint(rsp, PSP_A_STATS_RX_BAD, stats.rx_bad) || + nla_put_uint(rsp, PSP_A_STATS_TX_PACKETS, stats.tx_packets) || + nla_put_uint(rsp, PSP_A_STATS_TX_BYTES, stats.tx_bytes) || + nla_put_uint(rsp, PSP_A_STATS_TX_ERROR, stats.tx_error)) goto err_cancel_msg; genlmsg_end(rsp, hdr); -- cgit v1.2.3 From 30b6089aad35500e683025dddc029ac28705385d Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Wed, 5 Nov 2025 15:39:57 +0100 Subject: wifi: cfg80211: fix EHT typo This is clearly EHT, not ETH, fix the typo. Link: https://patch.msgid.link/20251105153958.e9d4af3b768e.I5f3378326837e3f62928a2f1fd3403f29cea069b@changeid Signed-off-by: Johannes Berg --- include/net/cfg80211.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'include/net') diff --git a/include/net/cfg80211.h b/include/net/cfg80211.h index f2e8963cfaac..84be0cdd1da0 100644 --- a/include/net/cfg80211.h +++ b/include/net/cfg80211.h @@ -685,7 +685,7 @@ ieee80211_get_he_6ghz_capa(const struct ieee80211_supported_band *sband, } /** - * ieee80211_get_eht_iftype_cap - return ETH capabilities for an sband's iftype + * ieee80211_get_eht_iftype_cap - return EHT capabilities for an sband's iftype * @sband: the sband to search for the iftype on * @iftype: enum nl80211_iftype * -- cgit v1.2.3 From 1a1cad924e8a60252132446fbba1284035010b4f Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Wed, 5 Nov 2025 15:39:58 +0100 Subject: wifi: mac80211: fix EHT typo This is clearly EHT, not ETH, fix the typo. Link: https://patch.msgid.link/20251105153958.12a04517f7ec.Idcf800817fa30605b1002c3d2287cad016e7aea7@changeid Signed-off-by: Johannes Berg --- include/net/mac80211.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'include/net') diff --git a/include/net/mac80211.h b/include/net/mac80211.h index c326243e1f01..c2e49542626c 100644 --- a/include/net/mac80211.h +++ b/include/net/mac80211.h @@ -7223,7 +7223,7 @@ ieee80211_get_he_6ghz_capa_vif(const struct ieee80211_supported_band *sband, } /** - * ieee80211_get_eht_iftype_cap_vif - return ETH capabilities for sband/vif + * ieee80211_get_eht_iftype_cap_vif - return EHT capabilities for sband/vif * @sband: the sband to search for the iftype on * @vif: the vif to get the iftype from * -- cgit v1.2.3 From 473235677af46ecb167917887586646e9d70d9ff Mon Sep 17 00:00:00 2001 From: Chien Wong Date: Fri, 7 Nov 2025 22:23:32 +0800 Subject: wifi: cfg80211: fix doc of struct key_params The seq in struct key_params is for many ciphers, including CCMP, GCMP, CMAC, GMAC. In addition to get_key(), it is also used when setting keys. Signed-off-by: Chien Wong Link: https://patch.msgid.link/20251107142332.181308-1-m@xv97.com Signed-off-by: Johannes Berg --- include/net/cfg80211.h | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) (limited to 'include/net') diff --git a/include/net/cfg80211.h b/include/net/cfg80211.h index 84be0cdd1da0..d87c18e1b133 100644 --- a/include/net/cfg80211.h +++ b/include/net/cfg80211.h @@ -786,8 +786,7 @@ struct vif_params { * @key: key material * @key_len: length of key material * @cipher: cipher suite selector - * @seq: sequence counter (IV/PN) for TKIP and CCMP keys, only used - * with the get_key() callback, must be in little endian, + * @seq: sequence counter (IV/PN), must be in little endian, * length given by @seq_len. * @seq_len: length of @seq. * @vlan_id: vlan_id for VLAN group key (if nonzero) -- cgit v1.2.3 From 15f295f55656658e65bdbc9b901d6b2e49d68d72 Mon Sep 17 00:00:00 2001 From: "D. Wythe" Date: Fri, 7 Nov 2025 11:56:31 +0800 Subject: net/smc: bpf: Introduce generic hook for handshake flow The introduction of IPPROTO_SMC enables eBPF programs to determine whether to use SMC based on the context of socket creation, such as network namespaces, PID and comm name, etc. As a subsequent enhancement, to introduce a new generic hook that allows decisions on whether to use SMC or not at runtime, including but not limited to local/remote IP address or ports. User can write their own implememtion via bpf_struct_ops now to choose whether to use SMC or not before TCP 3rd handshake to be comleted. Signed-off-by: D. Wythe Signed-off-by: Martin KaFai Lau Reviewed-by: Dust Li Link: https://patch.msgid.link/20251107035632.115950-3-alibuda@linux.alibaba.com --- include/net/netns/smc.h | 3 ++ include/net/smc.h | 53 ++++++++++++++++++ net/ipv4/tcp_output.c | 31 ++++++----- net/smc/Kconfig | 10 ++++ net/smc/Makefile | 1 + net/smc/af_smc.c | 9 ++++ net/smc/smc_hs_bpf.c | 140 ++++++++++++++++++++++++++++++++++++++++++++++++ net/smc/smc_hs_bpf.h | 31 +++++++++++ net/smc/smc_sysctl.c | 91 +++++++++++++++++++++++++++++++ 9 files changed, 355 insertions(+), 14 deletions(-) create mode 100644 net/smc/smc_hs_bpf.c create mode 100644 net/smc/smc_hs_bpf.h (limited to 'include/net') diff --git a/include/net/netns/smc.h b/include/net/netns/smc.h index 6ceb12baec24..ed24c9f638ee 100644 --- a/include/net/netns/smc.h +++ b/include/net/netns/smc.h @@ -17,6 +17,9 @@ struct netns_smc { #ifdef CONFIG_SYSCTL struct ctl_table_header *smc_hdr; #endif +#if IS_ENABLED(CONFIG_SMC_HS_CTRL_BPF) + struct smc_hs_ctrl __rcu *hs_ctrl; +#endif /* CONFIG_SMC_HS_CTRL_BPF */ unsigned int sysctl_autocorking_size; unsigned int sysctl_smcr_buf_type; int sysctl_smcr_testlink_time; diff --git a/include/net/smc.h b/include/net/smc.h index 08bee529ed8d..bfdc4c41f019 100644 --- a/include/net/smc.h +++ b/include/net/smc.h @@ -17,6 +17,8 @@ #include #include +struct tcp_sock; +struct inet_request_sock; struct sock; #define SMC_MAX_PNETID_LEN 16 /* Max. length of PNET id */ @@ -50,4 +52,55 @@ struct smcd_dev { u8 going_away : 1; }; +#define SMC_HS_CTRL_NAME_MAX 16 + +enum { + /* ops can be inherit from init_net */ + SMC_HS_CTRL_FLAG_INHERITABLE = 0x1, + + SMC_HS_CTRL_ALL_FLAGS = SMC_HS_CTRL_FLAG_INHERITABLE, +}; + +struct smc_hs_ctrl { + /* private */ + + struct list_head list; + struct module *owner; + + /* public */ + + /* unique name */ + char name[SMC_HS_CTRL_NAME_MAX]; + int flags; + + /* Invoked before computing SMC option for SYN packets. + * We can control whether to set SMC options by returning various value. + * Return 0 to disable SMC, or return any other value to enable it. + */ + int (*syn_option)(struct tcp_sock *tp); + + /* Invoked before Set up SMC options for SYN-ACK packets + * We can control whether to respond SMC options by returning various + * value. Return 0 to disable SMC, or return any other value to enable + * it. + */ + int (*synack_option)(const struct tcp_sock *tp, + struct inet_request_sock *ireq); +}; + +#if IS_ENABLED(CONFIG_SMC_HS_CTRL_BPF) +#define smc_call_hsbpf(init_val, tp, func, ...) ({ \ + typeof(init_val) __ret = (init_val); \ + struct smc_hs_ctrl *ctrl; \ + rcu_read_lock(); \ + ctrl = rcu_dereference(sock_net((struct sock *)(tp))->smc.hs_ctrl); \ + if (ctrl && ctrl->func) \ + __ret = ctrl->func(tp, ##__VA_ARGS__); \ + rcu_read_unlock(); \ + __ret; \ +}) +#else +#define smc_call_hsbpf(init_val, tp, ...) ({ (void)(tp); (init_val); }) +#endif /* CONFIG_SMC_HS_CTRL_BPF */ + #endif /* _SMC_H */ diff --git a/net/ipv4/tcp_output.c b/net/ipv4/tcp_output.c index 7f5df7a71f62..479afb714bdf 100644 --- a/net/ipv4/tcp_output.c +++ b/net/ipv4/tcp_output.c @@ -40,6 +40,7 @@ #include #include #include +#include #include #include @@ -802,34 +803,36 @@ static void tcp_options_write(struct tcphdr *th, struct tcp_sock *tp, mptcp_options_write(th, ptr, tp, opts); } -static void smc_set_option(const struct tcp_sock *tp, +static void smc_set_option(struct tcp_sock *tp, struct tcp_out_options *opts, unsigned int *remaining) { #if IS_ENABLED(CONFIG_SMC) - if (static_branch_unlikely(&tcp_have_smc)) { - if (tp->syn_smc) { - if (*remaining >= TCPOLEN_EXP_SMC_BASE_ALIGNED) { - opts->options |= OPTION_SMC; - *remaining -= TCPOLEN_EXP_SMC_BASE_ALIGNED; - } + if (static_branch_unlikely(&tcp_have_smc) && tp->syn_smc) { + tp->syn_smc = !!smc_call_hsbpf(1, tp, syn_option); + /* re-check syn_smc */ + if (tp->syn_smc && + *remaining >= TCPOLEN_EXP_SMC_BASE_ALIGNED) { + opts->options |= OPTION_SMC; + *remaining -= TCPOLEN_EXP_SMC_BASE_ALIGNED; } } #endif } static void smc_set_option_cond(const struct tcp_sock *tp, - const struct inet_request_sock *ireq, + struct inet_request_sock *ireq, struct tcp_out_options *opts, unsigned int *remaining) { #if IS_ENABLED(CONFIG_SMC) - if (static_branch_unlikely(&tcp_have_smc)) { - if (tp->syn_smc && ireq->smc_ok) { - if (*remaining >= TCPOLEN_EXP_SMC_BASE_ALIGNED) { - opts->options |= OPTION_SMC; - *remaining -= TCPOLEN_EXP_SMC_BASE_ALIGNED; - } + if (static_branch_unlikely(&tcp_have_smc) && tp->syn_smc && ireq->smc_ok) { + ireq->smc_ok = !!smc_call_hsbpf(1, tp, synack_option, ireq); + /* re-check smc_ok */ + if (ireq->smc_ok && + *remaining >= TCPOLEN_EXP_SMC_BASE_ALIGNED) { + opts->options |= OPTION_SMC; + *remaining -= TCPOLEN_EXP_SMC_BASE_ALIGNED; } } #endif diff --git a/net/smc/Kconfig b/net/smc/Kconfig index 99ecd59d1f4b..325addf83cc6 100644 --- a/net/smc/Kconfig +++ b/net/smc/Kconfig @@ -19,3 +19,13 @@ config SMC_DIAG smcss. if unsure, say Y. + +config SMC_HS_CTRL_BPF + bool "Generic eBPF hook for SMC handshake flow" + depends on SMC && BPF_SYSCALL + default y + help + SMC_HS_CTRL_BPF enables support to register generic eBPF hook for SMC + handshake flow, which offer much greater flexibility in modifying the behavior + of the SMC protocol stack compared to a complete kernel-based approach. Select + this option if you want filtring the handshake process via eBPF programs. \ No newline at end of file diff --git a/net/smc/Makefile b/net/smc/Makefile index 0e754cbc38f9..5368634c5dd6 100644 --- a/net/smc/Makefile +++ b/net/smc/Makefile @@ -6,3 +6,4 @@ smc-y := af_smc.o smc_pnet.o smc_ib.o smc_clc.o smc_core.o smc_wr.o smc_llc.o smc-y += smc_cdc.o smc_tx.o smc_rx.o smc_close.o smc_ism.o smc_netlink.o smc_stats.o smc-y += smc_tracepoint.o smc_inet.o smc-$(CONFIG_SYSCTL) += smc_sysctl.o +smc-$(CONFIG_SMC_HS_CTRL_BPF) += smc_hs_bpf.o diff --git a/net/smc/af_smc.c b/net/smc/af_smc.c index 0ef3e16a8517..e388de8dca09 100644 --- a/net/smc/af_smc.c +++ b/net/smc/af_smc.c @@ -58,6 +58,7 @@ #include "smc_tracepoint.h" #include "smc_sysctl.h" #include "smc_inet.h" +#include "smc_hs_bpf.h" static DEFINE_MUTEX(smc_server_lgr_pending); /* serialize link group * creation on server @@ -3600,8 +3601,16 @@ static int __init smc_init(void) pr_err("%s: smc_inet_init fails with %d\n", __func__, rc); goto out_ulp; } + rc = bpf_smc_hs_ctrl_init(); + if (rc) { + pr_err("%s: bpf_smc_hs_ctrl_init fails with %d\n", __func__, + rc); + goto out_inet; + } static_branch_enable(&tcp_have_smc); return 0; +out_inet: + smc_inet_exit(); out_ulp: tcp_unregister_ulp(&smc_ulp_ops); out_ib: diff --git a/net/smc/smc_hs_bpf.c b/net/smc/smc_hs_bpf.c new file mode 100644 index 000000000000..063d23d85850 --- /dev/null +++ b/net/smc/smc_hs_bpf.c @@ -0,0 +1,140 @@ +// SPDX-License-Identifier: GPL-2.0-only +/* + * Shared Memory Communications over RDMA (SMC-R) and RoCE + * + * Generic hook for SMC handshake flow. + * + * Copyright IBM Corp. 2016 + * Copyright (c) 2025, Alibaba Inc. + * + * Author: D. Wythe + */ + +#include +#include +#include +#include + +#include "smc_hs_bpf.h" + +static DEFINE_SPINLOCK(smc_hs_ctrl_list_lock); +static LIST_HEAD(smc_hs_ctrl_list); + +static int smc_hs_ctrl_reg(struct smc_hs_ctrl *ctrl) +{ + int ret = 0; + + spin_lock(&smc_hs_ctrl_list_lock); + /* already exist or duplicate name */ + if (smc_hs_ctrl_find_by_name(ctrl->name)) + ret = -EEXIST; + else + list_add_tail_rcu(&ctrl->list, &smc_hs_ctrl_list); + spin_unlock(&smc_hs_ctrl_list_lock); + return ret; +} + +static void smc_hs_ctrl_unreg(struct smc_hs_ctrl *ctrl) +{ + spin_lock(&smc_hs_ctrl_list_lock); + list_del_rcu(&ctrl->list); + spin_unlock(&smc_hs_ctrl_list_lock); + + /* Ensure that all readers to complete */ + synchronize_rcu(); +} + +struct smc_hs_ctrl *smc_hs_ctrl_find_by_name(const char *name) +{ + struct smc_hs_ctrl *ctrl; + + list_for_each_entry_rcu(ctrl, &smc_hs_ctrl_list, list) { + if (strcmp(ctrl->name, name) == 0) + return ctrl; + } + return NULL; +} + +static int __smc_bpf_stub_set_tcp_option(struct tcp_sock *tp) { return 1; } +static int __smc_bpf_stub_set_tcp_option_cond(const struct tcp_sock *tp, + struct inet_request_sock *ireq) +{ + return 1; +} + +static struct smc_hs_ctrl __smc_bpf_hs_ctrl = { + .syn_option = __smc_bpf_stub_set_tcp_option, + .synack_option = __smc_bpf_stub_set_tcp_option_cond, +}; + +static int smc_bpf_hs_ctrl_init(struct btf *btf) { return 0; } + +static int smc_bpf_hs_ctrl_reg(void *kdata, struct bpf_link *link) +{ + if (link) + return -EOPNOTSUPP; + + return smc_hs_ctrl_reg(kdata); +} + +static void smc_bpf_hs_ctrl_unreg(void *kdata, struct bpf_link *link) +{ + smc_hs_ctrl_unreg(kdata); +} + +static int smc_bpf_hs_ctrl_init_member(const struct btf_type *t, + const struct btf_member *member, + void *kdata, const void *udata) +{ + const struct smc_hs_ctrl *u_ctrl; + struct smc_hs_ctrl *k_ctrl; + u32 moff; + + u_ctrl = (const struct smc_hs_ctrl *)udata; + k_ctrl = (struct smc_hs_ctrl *)kdata; + + moff = __btf_member_bit_offset(t, member) / 8; + switch (moff) { + case offsetof(struct smc_hs_ctrl, name): + if (bpf_obj_name_cpy(k_ctrl->name, u_ctrl->name, + sizeof(u_ctrl->name)) <= 0) + return -EINVAL; + return 1; + case offsetof(struct smc_hs_ctrl, flags): + if (u_ctrl->flags & ~SMC_HS_CTRL_ALL_FLAGS) + return -EINVAL; + k_ctrl->flags = u_ctrl->flags; + return 1; + default: + break; + } + + return 0; +} + +static const struct bpf_func_proto * +bpf_smc_hs_func_proto(enum bpf_func_id func_id, const struct bpf_prog *prog) +{ + return bpf_base_func_proto(func_id, prog); +} + +static const struct bpf_verifier_ops smc_bpf_verifier_ops = { + .get_func_proto = bpf_smc_hs_func_proto, + .is_valid_access = bpf_tracing_btf_ctx_access, +}; + +static struct bpf_struct_ops bpf_smc_hs_ctrl_ops = { + .name = "smc_hs_ctrl", + .init = smc_bpf_hs_ctrl_init, + .reg = smc_bpf_hs_ctrl_reg, + .unreg = smc_bpf_hs_ctrl_unreg, + .cfi_stubs = &__smc_bpf_hs_ctrl, + .verifier_ops = &smc_bpf_verifier_ops, + .init_member = smc_bpf_hs_ctrl_init_member, + .owner = THIS_MODULE, +}; + +int bpf_smc_hs_ctrl_init(void) +{ + return register_bpf_struct_ops(&bpf_smc_hs_ctrl_ops, smc_hs_ctrl); +} diff --git a/net/smc/smc_hs_bpf.h b/net/smc/smc_hs_bpf.h new file mode 100644 index 000000000000..f5f1807c079e --- /dev/null +++ b/net/smc/smc_hs_bpf.h @@ -0,0 +1,31 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +/* + * Shared Memory Communications over RDMA (SMC-R) and RoCE + * + * Generic hook for SMC handshake flow. + * + * Copyright IBM Corp. 2016 + * Copyright (c) 2025, Alibaba Inc. + * + * Author: D. Wythe + */ + +#ifndef __SMC_HS_CTRL +#define __SMC_HS_CTRL + +#include + +/* Find hs_ctrl by the target name, which required to be a c-string. + * Return NULL if no such ctrl was found,otherwise, return a valid ctrl. + * + * Note: Caller MUST ensure it's was invoked under rcu_read_lock. + */ +struct smc_hs_ctrl *smc_hs_ctrl_find_by_name(const char *name); + +#if IS_ENABLED(CONFIG_SMC_HS_CTRL_BPF) +int bpf_smc_hs_ctrl_init(void); +#else +static inline int bpf_smc_hs_ctrl_init(void) { return 0; } +#endif /* CONFIG_SMC_HS_CTRL_BPF */ + +#endif /* __SMC_HS_CTRL */ diff --git a/net/smc/smc_sysctl.c b/net/smc/smc_sysctl.c index 7b2471904d04..b1efed546243 100644 --- a/net/smc/smc_sysctl.c +++ b/net/smc/smc_sysctl.c @@ -12,12 +12,14 @@ #include #include +#include #include #include "smc.h" #include "smc_core.h" #include "smc_llc.h" #include "smc_sysctl.h" +#include "smc_hs_bpf.h" static int min_sndbuf = SMC_BUF_MIN_SIZE; static int min_rcvbuf = SMC_BUF_MIN_SIZE; @@ -32,6 +34,69 @@ static int conns_per_lgr_max = SMC_CONN_PER_LGR_MAX; static unsigned int smcr_max_wr_min = 2; static unsigned int smcr_max_wr_max = 2048; +#if IS_ENABLED(CONFIG_SMC_HS_CTRL_BPF) +static int smc_net_replace_smc_hs_ctrl(struct net *net, const char *name) +{ + struct smc_hs_ctrl *ctrl = NULL; + + rcu_read_lock(); + /* null or empty name ask to clear current ctrl */ + if (name && name[0]) { + ctrl = smc_hs_ctrl_find_by_name(name); + if (!ctrl) { + rcu_read_unlock(); + return -EINVAL; + } + /* no change, just return */ + if (ctrl == rcu_dereference(net->smc.hs_ctrl)) { + rcu_read_unlock(); + return 0; + } + if (!bpf_try_module_get(ctrl, ctrl->owner)) { + rcu_read_unlock(); + return -EBUSY; + } + } + /* xhcg old ctrl with the new one atomically */ + ctrl = unrcu_pointer(xchg(&net->smc.hs_ctrl, RCU_INITIALIZER(ctrl))); + /* release old ctrl */ + if (ctrl) + bpf_module_put(ctrl, ctrl->owner); + + rcu_read_unlock(); + return 0; +} + +static int proc_smc_hs_ctrl(const struct ctl_table *ctl, int write, + void *buffer, size_t *lenp, loff_t *ppos) +{ + struct net *net = container_of(ctl->data, struct net, smc.hs_ctrl); + char val[SMC_HS_CTRL_NAME_MAX]; + const struct ctl_table tbl = { + .data = val, + .maxlen = SMC_HS_CTRL_NAME_MAX, + }; + struct smc_hs_ctrl *ctrl; + int ret; + + rcu_read_lock(); + ctrl = rcu_dereference(net->smc.hs_ctrl); + if (ctrl) + memcpy(val, ctrl->name, sizeof(ctrl->name)); + else + val[0] = '\0'; + rcu_read_unlock(); + + ret = proc_dostring(&tbl, write, buffer, lenp, ppos); + if (ret) + return ret; + + if (write) + ret = smc_net_replace_smc_hs_ctrl(net, val); + return ret; +} +#endif /* CONFIG_SMC_HS_CTRL_BPF */ + static struct ctl_table smc_table[] = { { .procname = "autocorking_size", @@ -119,6 +184,15 @@ static struct ctl_table smc_table[] = { .extra1 = &smcr_max_wr_min, .extra2 = &smcr_max_wr_max, }, +#if IS_ENABLED(CONFIG_SMC_HS_CTRL_BPF) + { + .procname = "hs_ctrl", + .data = &init_net.smc.hs_ctrl, + .mode = 0644, + .maxlen = SMC_HS_CTRL_NAME_MAX, + .proc_handler = proc_smc_hs_ctrl, + }, +#endif /* CONFIG_SMC_HS_CTRL_BPF */ }; int __net_init smc_sysctl_net_init(struct net *net) @@ -129,6 +203,16 @@ int __net_init smc_sysctl_net_init(struct net *net) table = smc_table; if (!net_eq(net, &init_net)) { int i; +#if IS_ENABLED(CONFIG_SMC_HS_CTRL_BPF) + struct smc_hs_ctrl *ctrl; + + rcu_read_lock(); + ctrl = rcu_dereference(init_net.smc.hs_ctrl); + if (ctrl && ctrl->flags & SMC_HS_CTRL_FLAG_INHERITABLE && + bpf_try_module_get(ctrl, ctrl->owner)) + rcu_assign_pointer(net->smc.hs_ctrl, ctrl); + rcu_read_unlock(); +#endif /* CONFIG_SMC_HS_CTRL_BPF */ table = kmemdup(table, sizeof(smc_table), GFP_KERNEL); if (!table) @@ -161,6 +245,9 @@ err_reg: if (!net_eq(net, &init_net)) kfree(table); err_alloc: +#if IS_ENABLED(CONFIG_SMC_HS_CTRL_BPF) + smc_net_replace_smc_hs_ctrl(net, NULL); +#endif /* CONFIG_SMC_HS_CTRL_BPF */ return -ENOMEM; } @@ -170,6 +257,10 @@ void __net_exit smc_sysctl_net_exit(struct net *net) table = net->smc.smc_hdr->ctl_table_arg; unregister_net_sysctl_table(net->smc.smc_hdr); +#if IS_ENABLED(CONFIG_SMC_HS_CTRL_BPF) + smc_net_replace_smc_hs_ctrl(net, NULL); +#endif /* CONFIG_SMC_HS_CTRL_BPF */ + if (!net_eq(net, &init_net)) kfree(table); } -- cgit v1.2.3 From 73edb26b06ea0eb5bd8c6cae5b2df212ae3c7ab5 Mon Sep 17 00:00:00 2001 From: Kuniyuki Iwashima Date: Thu, 6 Nov 2025 22:34:06 +0000 Subject: sctp: Don't inherit do_auto_asconf in sctp_clone_sock(). syzbot reported list_del(&sp->auto_asconf_list) corruption in sctp_destroy_sock(). The repro calls setsockopt(SCTP_AUTO_ASCONF, 1) to a SCTP listener, calls accept(), and close()s the child socket. setsockopt(SCTP_AUTO_ASCONF, 1) sets sp->do_auto_asconf to 1 and links sp->auto_asconf_list to a per-netns list. Both fields are placed after sp->pd_lobby in struct sctp_sock, and sctp_copy_descendant() did not copy the fields before the cited commit. Also, sctp_clone_sock() did not set them explicitly. In addition, sctp_auto_asconf_init() is called from sctp_sock_migrate(), but it initialises the fields only conditionally. The two fields relied on __GFP_ZERO added in sk_alloc(), but sk_clone() does not use it. Let's clear newsp->do_auto_asconf in sctp_clone_sock(). [0]: list_del corruption. prev->next should be ffff8880799e9148, but was ffff8880799e8808. (prev=ffff88803347d9f8) kernel BUG at lib/list_debug.c:64! Oops: invalid opcode: 0000 [#1] SMP KASAN PTI CPU: 0 UID: 0 PID: 6008 Comm: syz.0.17 Not tainted syzkaller #0 PREEMPT(full) Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 10/02/2025 RIP: 0010:__list_del_entry_valid_or_report+0x15a/0x190 lib/list_debug.c:62 Code: e8 7b 26 71 fd 43 80 3c 2c 00 74 08 4c 89 ff e8 7c ee 92 fd 49 8b 17 48 c7 c7 80 0a bf 8b 48 89 de 4c 89 f9 e8 07 c6 94 fc 90 <0f> 0b 4c 89 f7 e8 4c 26 71 fd 43 80 3c 2c 00 74 08 4c 89 ff e8 4d RSP: 0018:ffffc90003067ad8 EFLAGS: 00010246 RAX: 000000000000006d RBX: ffff8880799e9148 RCX: b056988859ee6e00 RDX: 0000000000000000 RSI: 0000000000000202 RDI: 0000000000000000 RBP: dffffc0000000000 R08: ffffc90003067807 R09: 1ffff9200060cf00 R10: dffffc0000000000 R11: fffff5200060cf01 R12: 1ffff1100668fb3f R13: dffffc0000000000 R14: ffff88803347d9f8 R15: ffff88803347d9f8 FS: 00005555823e5500(0000) GS:ffff88812613e000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 0000200000000480 CR3: 00000000741ce000 CR4: 00000000003526f0 Call Trace: __list_del_entry_valid include/linux/list.h:132 [inline] __list_del_entry include/linux/list.h:223 [inline] list_del include/linux/list.h:237 [inline] sctp_destroy_sock+0xb4/0x370 net/sctp/socket.c:5163 sk_common_release+0x75/0x310 net/core/sock.c:3961 sctp_close+0x77e/0x900 net/sctp/socket.c:1550 inet_release+0x144/0x190 net/ipv4/af_inet.c:437 __sock_release net/socket.c:662 [inline] sock_close+0xc3/0x240 net/socket.c:1455 __fput+0x44c/0xa70 fs/file_table.c:468 task_work_run+0x1d4/0x260 kernel/task_work.c:227 resume_user_mode_work include/linux/resume_user_mode.h:50 [inline] exit_to_user_mode_loop+0xe9/0x130 kernel/entry/common.c:43 exit_to_user_mode_prepare include/linux/irq-entry-common.h:225 [inline] syscall_exit_to_user_mode_work include/linux/entry-common.h:175 [inline] syscall_exit_to_user_mode include/linux/entry-common.h:210 [inline] do_syscall_64+0x2bd/0xfa0 arch/x86/entry/syscall_64.c:100 entry_SYSCALL_64_after_hwframe+0x77/0x7f Fixes: 16942cf4d3e3 ("sctp: Use sk_clone() in sctp_accept().") Reported-by: syzbot+ba535cb417f106327741@syzkaller.appspotmail.com Closes: https://lore.kernel.org/netdev/690d2185.a70a0220.22f260.000e.GAE@google.com/ Signed-off-by: Kuniyuki Iwashima Acked-by: Xin Long Link: https://patch.msgid.link/20251106223418.1455510-1-kuniyu@google.com Signed-off-by: Jakub Kicinski --- include/net/sctp/structs.h | 4 ---- net/sctp/socket.c | 1 + 2 files changed, 1 insertion(+), 4 deletions(-) (limited to 'include/net') diff --git a/include/net/sctp/structs.h b/include/net/sctp/structs.h index 5900196d65fd..affee44bd38e 100644 --- a/include/net/sctp/structs.h +++ b/include/net/sctp/structs.h @@ -228,10 +228,6 @@ struct sctp_sock { atomic_t pd_mode; - /* Fields after this point will be skipped on copies, like on accept - * and peeloff operations - */ - /* Receive to here while partial delivery is in effect. */ struct sk_buff_head pd_lobby; diff --git a/net/sctp/socket.c b/net/sctp/socket.c index 38d2932acebf..d808096f5ab1 100644 --- a/net/sctp/socket.c +++ b/net/sctp/socket.c @@ -4885,6 +4885,7 @@ static struct sock *sctp_clone_sock(struct sock *sk, } #endif + newsp->do_auto_asconf = 0; skb_queue_head_init(&newsp->pd_lobby); newsp->ep = sctp_endpoint_new(newsk, GFP_KERNEL); -- cgit v1.2.3 From 8da7bea7db692e786165b71729fb68b7ff65ee56 Mon Sep 17 00:00:00 2001 From: Jason Xing Date: Fri, 31 Oct 2025 18:33:28 +0800 Subject: xsk: add indirect call for xsk_destruct_skb Since Eric proposed an idea about adding indirect call wrappers for UDP and managed to see a huge improvement[1], the same situation can also be applied in xsk scenario. This patch adds an indirect call for xsk and helps current copy mode improve the performance by around 1% stably which was observed with IXGBE at 10Gb/sec loaded. If the throughput grows, the positive effect will be magnified. I applied this patch on top of batch xmit series[2], and was able to see <5% improvement from our internal application which is a little bit unstable though. Use INDIRECT wrappers to keep xsk_destruct_skb static as it used to be when the mitigation config is off. Be aware of the freeing path that can be very hot since the frequency can reach around 2,000,000 times per second with the xdpsock test. [1]: https://lore.kernel.org/netdev/20251006193103.2684156-2-edumazet@google.com/ [2]: https://lore.kernel.org/all/20251021131209.41491-1-kerneljasonxing@gmail.com/ Suggested-by: Alexander Lobakin Signed-off-by: Jason Xing Reviewed-by: Alexander Lobakin Link: https://patch.msgid.link/20251031103328.95468-1-kerneljasonxing@gmail.com Signed-off-by: Paolo Abeni --- include/net/xdp_sock.h | 7 +++++++ net/core/skbuff.c | 8 +++++--- net/xdp/xsk.c | 3 ++- 3 files changed, 14 insertions(+), 4 deletions(-) (limited to 'include/net') diff --git a/include/net/xdp_sock.h b/include/net/xdp_sock.h index ce587a225661..23e8861e8b25 100644 --- a/include/net/xdp_sock.h +++ b/include/net/xdp_sock.h @@ -125,6 +125,7 @@ struct xsk_tx_metadata_ops { int xsk_generic_rcv(struct xdp_sock *xs, struct xdp_buff *xdp); int __xsk_map_redirect(struct xdp_sock *xs, struct xdp_buff *xdp); void __xsk_map_flush(struct list_head *flush_list); +INDIRECT_CALLABLE_DECLARE(void xsk_destruct_skb(struct sk_buff *)); /** * xsk_tx_metadata_to_compl - Save enough relevant metadata information @@ -218,6 +219,12 @@ static inline void __xsk_map_flush(struct list_head *flush_list) { } +#ifdef CONFIG_MITIGATION_RETPOLINE +static inline void xsk_destruct_skb(struct sk_buff *skb) +{ +} +#endif + static inline void xsk_tx_metadata_to_compl(struct xsk_tx_metadata *meta, struct xsk_tx_metadata_compl *compl) { diff --git a/net/core/skbuff.c b/net/core/skbuff.c index d95658b738d1..4f4d7ab7057f 100644 --- a/net/core/skbuff.c +++ b/net/core/skbuff.c @@ -81,6 +81,7 @@ #include #include #include +#include #include #include @@ -1140,12 +1141,13 @@ void skb_release_head_state(struct sk_buff *skb) if (skb->destructor) { DEBUG_NET_WARN_ON_ONCE(in_hardirq()); #ifdef CONFIG_INET - INDIRECT_CALL_3(skb->destructor, + INDIRECT_CALL_4(skb->destructor, tcp_wfree, __sock_wfree, sock_wfree, + xsk_destruct_skb, skb); #else - INDIRECT_CALL_1(skb->destructor, - sock_wfree, + INDIRECT_CALL_2(skb->destructor, + sock_wfree, xsk_destruct_skb, skb); #endif diff --git a/net/xdp/xsk.c b/net/xdp/xsk.c index ed8b612ec29d..bcfd400e9cf8 100644 --- a/net/xdp/xsk.c +++ b/net/xdp/xsk.c @@ -602,7 +602,8 @@ static u32 xsk_get_num_desc(struct sk_buff *skb) return XSKCB(skb)->num_descs; } -static void xsk_destruct_skb(struct sk_buff *skb) +INDIRECT_CALLABLE_SCOPE +void xsk_destruct_skb(struct sk_buff *skb) { struct xsk_tx_metadata_compl *compl = &skb_shinfo(skb)->xsk_meta; -- cgit v1.2.3 From e18efacc9c2f17b12c6e019cabad70a2989bd3a9 Mon Sep 17 00:00:00 2001 From: Pagadala Yesu Anjaneyulu Date: Mon, 10 Nov 2025 14:10:29 +0200 Subject: wifi: cfg80211/mac80211: clean up duplicate ap_power handling Move duplicated ap_power type handling code to an inline function in cfg80211. Signed-off-by: Pagadala Yesu Anjaneyulu Reviewed-by: Johannes Berg Signed-off-by: Miri Korenblit Link: https://patch.msgid.link/20251110140806.959948da1cb5.I893b5168329fb3232f249c182a35c99804112da6@changeid Signed-off-by: Johannes Berg --- include/net/cfg80211.h | 24 ++++++++++++++++++++++++ net/mac80211/mlme.c | 20 +------------------- net/wireless/scan.c | 15 +-------------- 3 files changed, 26 insertions(+), 33 deletions(-) (limited to 'include/net') diff --git a/include/net/cfg80211.h b/include/net/cfg80211.h index d87c18e1b133..1b257eaf8de5 100644 --- a/include/net/cfg80211.h +++ b/include/net/cfg80211.h @@ -10134,6 +10134,30 @@ static inline int cfg80211_color_change_notify(struct net_device *dev, 0, 0, link_id); } +/** + * cfg80211_6ghz_power_type - determine AP regulatory power type + * @control: control flags + * + * Return: regulatory power type from &enum ieee80211_ap_reg_power + */ +static inline enum ieee80211_ap_reg_power +cfg80211_6ghz_power_type(u8 control) +{ + switch (u8_get_bits(control, IEEE80211_HE_6GHZ_OPER_CTRL_REG_INFO)) { + case IEEE80211_6GHZ_CTRL_REG_LPI_AP: + case IEEE80211_6GHZ_CTRL_REG_INDOOR_LPI_AP: + return IEEE80211_REG_LPI_AP; + case IEEE80211_6GHZ_CTRL_REG_SP_AP: + case IEEE80211_6GHZ_CTRL_REG_INDOOR_SP_AP: + case IEEE80211_6GHZ_CTRL_REG_INDOOR_SP_AP_OLD: + return IEEE80211_REG_SP_AP; + case IEEE80211_6GHZ_CTRL_REG_VLP_AP: + return IEEE80211_REG_VLP_AP; + default: + return IEEE80211_REG_UNSET_AP; + } +} + /** * cfg80211_links_removed - Notify about removed STA MLD setup links. * @dev: network device. diff --git a/net/mac80211/mlme.c b/net/mac80211/mlme.c index 804c3a95b7c6..64230696f3e4 100644 --- a/net/mac80211/mlme.c +++ b/net/mac80211/mlme.c @@ -6034,24 +6034,6 @@ ieee80211_determine_our_sta_mode_assoc(struct ieee80211_sub_if_data *sdata, conn->bw_limit, tmp.bw_limit); } -static enum ieee80211_ap_reg_power -ieee80211_ap_power_type(u8 control) -{ - switch (u8_get_bits(control, IEEE80211_HE_6GHZ_OPER_CTRL_REG_INFO)) { - case IEEE80211_6GHZ_CTRL_REG_LPI_AP: - case IEEE80211_6GHZ_CTRL_REG_INDOOR_LPI_AP: - return IEEE80211_REG_LPI_AP; - case IEEE80211_6GHZ_CTRL_REG_SP_AP: - case IEEE80211_6GHZ_CTRL_REG_INDOOR_SP_AP: - case IEEE80211_6GHZ_CTRL_REG_INDOOR_SP_AP_OLD: - return IEEE80211_REG_SP_AP; - case IEEE80211_6GHZ_CTRL_REG_VLP_AP: - return IEEE80211_REG_VLP_AP; - default: - return IEEE80211_REG_UNSET_AP; - } -} - static int ieee80211_prep_channel(struct ieee80211_sub_if_data *sdata, struct ieee80211_link_data *link, int link_id, @@ -6094,7 +6076,7 @@ static int ieee80211_prep_channel(struct ieee80211_sub_if_data *sdata, he_6ghz_oper = ieee80211_he_6ghz_oper(elems->he_operation); if (he_6ghz_oper) link->conf->power_type = - ieee80211_ap_power_type(he_6ghz_oper->control); + cfg80211_6ghz_power_type(he_6ghz_oper->control); else link_info(link, "HE 6 GHz operation missing (on %d MHz), expect issues\n", diff --git a/net/wireless/scan.c b/net/wireless/scan.c index 90a9187a6b13..68c4130d602f 100644 --- a/net/wireless/scan.c +++ b/net/wireless/scan.c @@ -2230,20 +2230,7 @@ cfg80211_get_6ghz_power_type(const u8 *elems, size_t elems_len) if (!he_6ghz_oper) return IEEE80211_REG_UNSET_AP; - switch (u8_get_bits(he_6ghz_oper->control, - IEEE80211_HE_6GHZ_OPER_CTRL_REG_INFO)) { - case IEEE80211_6GHZ_CTRL_REG_LPI_AP: - case IEEE80211_6GHZ_CTRL_REG_INDOOR_LPI_AP: - return IEEE80211_REG_LPI_AP; - case IEEE80211_6GHZ_CTRL_REG_SP_AP: - case IEEE80211_6GHZ_CTRL_REG_INDOOR_SP_AP: - case IEEE80211_6GHZ_CTRL_REG_INDOOR_SP_AP_OLD: - return IEEE80211_REG_SP_AP; - case IEEE80211_6GHZ_CTRL_REG_VLP_AP: - return IEEE80211_REG_VLP_AP; - default: - return IEEE80211_REG_UNSET_AP; - } + return cfg80211_6ghz_power_type(he_6ghz_oper->control); } static bool cfg80211_6ghz_power_type_valid(const u8 *elems, size_t elems_len, -- cgit v1.2.3 From b54cf0f4495a8f3fa94245cdda7716792400299e Mon Sep 17 00:00:00 2001 From: Pagadala Yesu Anjaneyulu Date: Mon, 10 Nov 2025 14:10:30 +0200 Subject: wifi: cfg80211/mac80211: Add fallback mechanism for INDOOR_SP connection Implement fallback to LPI mode when SP mode is not permitted by regulatory constraints for INDOOR_SP connections. Limit fallback mechanism to client mode. Signed-off-by: Pagadala Yesu Anjaneyulu Reviewed-by: Johannes Berg Signed-off-by: Miri Korenblit Link: https://patch.msgid.link/20251110140806.8b43201a34ae.I37fc7bb5892eb9d044d619802e8f2095fde6b296@changeid Signed-off-by: Johannes Berg --- include/net/cfg80211.h | 9 +++++++-- net/mac80211/mlme.c | 3 ++- net/wireless/core.h | 3 ++- net/wireless/nl80211.c | 4 ++-- net/wireless/scan.c | 7 ++++--- 5 files changed, 17 insertions(+), 9 deletions(-) (limited to 'include/net') diff --git a/include/net/cfg80211.h b/include/net/cfg80211.h index 1b257eaf8de5..625cb2c78361 100644 --- a/include/net/cfg80211.h +++ b/include/net/cfg80211.h @@ -10137,22 +10137,27 @@ static inline int cfg80211_color_change_notify(struct net_device *dev, /** * cfg80211_6ghz_power_type - determine AP regulatory power type * @control: control flags + * @client_flags: &enum ieee80211_channel_flags for station mode to enable + * SP to LPI fallback, zero otherwise. * * Return: regulatory power type from &enum ieee80211_ap_reg_power */ static inline enum ieee80211_ap_reg_power -cfg80211_6ghz_power_type(u8 control) +cfg80211_6ghz_power_type(u8 control, u32 client_flags) { switch (u8_get_bits(control, IEEE80211_HE_6GHZ_OPER_CTRL_REG_INFO)) { case IEEE80211_6GHZ_CTRL_REG_LPI_AP: case IEEE80211_6GHZ_CTRL_REG_INDOOR_LPI_AP: return IEEE80211_REG_LPI_AP; case IEEE80211_6GHZ_CTRL_REG_SP_AP: - case IEEE80211_6GHZ_CTRL_REG_INDOOR_SP_AP: case IEEE80211_6GHZ_CTRL_REG_INDOOR_SP_AP_OLD: return IEEE80211_REG_SP_AP; case IEEE80211_6GHZ_CTRL_REG_VLP_AP: return IEEE80211_REG_VLP_AP; + case IEEE80211_6GHZ_CTRL_REG_INDOOR_SP_AP: + if (client_flags & IEEE80211_CHAN_NO_6GHZ_AFC_CLIENT) + return IEEE80211_REG_LPI_AP; + return IEEE80211_REG_SP_AP; default: return IEEE80211_REG_UNSET_AP; } diff --git a/net/mac80211/mlme.c b/net/mac80211/mlme.c index 64230696f3e4..c705d3f45aff 100644 --- a/net/mac80211/mlme.c +++ b/net/mac80211/mlme.c @@ -6076,7 +6076,8 @@ static int ieee80211_prep_channel(struct ieee80211_sub_if_data *sdata, he_6ghz_oper = ieee80211_he_6ghz_oper(elems->he_operation); if (he_6ghz_oper) link->conf->power_type = - cfg80211_6ghz_power_type(he_6ghz_oper->control); + cfg80211_6ghz_power_type(he_6ghz_oper->control, + cbss->channel->flags); else link_info(link, "HE 6 GHz operation missing (on %d MHz), expect issues\n", diff --git a/net/wireless/core.h b/net/wireless/core.h index b6bd7f4d6385..82f343663e8f 100644 --- a/net/wireless/core.h +++ b/net/wireless/core.h @@ -550,7 +550,8 @@ cfg80211_bss_update(struct cfg80211_registered_device *rdev, bool signal_valid, unsigned long ts); enum ieee80211_ap_reg_power -cfg80211_get_6ghz_power_type(const u8 *elems, size_t elems_len); +cfg80211_get_6ghz_power_type(const u8 *elems, size_t elems_len, + u32 client_flags); #ifdef CONFIG_CFG80211_DEVELOPER_WARNINGS #define CFG80211_DEV_WARN_ON(cond) WARN_ON(cond) diff --git a/net/wireless/nl80211.c b/net/wireless/nl80211.c index 2187e148389d..29c92bc8291b 100644 --- a/net/wireless/nl80211.c +++ b/net/wireless/nl80211.c @@ -6748,7 +6748,7 @@ static int nl80211_start_ap(struct sk_buff *skb, struct genl_info *info) beacon_check.relax = true; beacon_check.reg_power = cfg80211_get_6ghz_power_type(params->beacon.tail, - params->beacon.tail_len); + params->beacon.tail_len, 0); if (!cfg80211_reg_check_beaconing(&rdev->wiphy, ¶ms->chandef, &beacon_check)) { err = -EINVAL; @@ -6927,7 +6927,7 @@ static int nl80211_set_beacon(struct sk_buff *skb, struct genl_info *info) beacon_check.relax = true; beacon_check.reg_power = cfg80211_get_6ghz_power_type(params->beacon.tail, - params->beacon.tail_len); + params->beacon.tail_len, 0); if (!cfg80211_reg_check_beaconing(&rdev->wiphy, &wdev->links[link_id].ap.chandef, &beacon_check)) { diff --git a/net/wireless/scan.c b/net/wireless/scan.c index 68c4130d602f..7546647752fd 100644 --- a/net/wireless/scan.c +++ b/net/wireless/scan.c @@ -2212,7 +2212,8 @@ struct cfg80211_inform_single_bss_data { }; enum ieee80211_ap_reg_power -cfg80211_get_6ghz_power_type(const u8 *elems, size_t elems_len) +cfg80211_get_6ghz_power_type(const u8 *elems, size_t elems_len, + u32 client_flags) { const struct ieee80211_he_6ghz_oper *he_6ghz_oper; struct ieee80211_he_operation *he_oper; @@ -2230,13 +2231,13 @@ cfg80211_get_6ghz_power_type(const u8 *elems, size_t elems_len) if (!he_6ghz_oper) return IEEE80211_REG_UNSET_AP; - return cfg80211_6ghz_power_type(he_6ghz_oper->control); + return cfg80211_6ghz_power_type(he_6ghz_oper->control, client_flags); } static bool cfg80211_6ghz_power_type_valid(const u8 *elems, size_t elems_len, const u32 flags) { - switch (cfg80211_get_6ghz_power_type(elems, elems_len)) { + switch (cfg80211_get_6ghz_power_type(elems, elems_len, flags)) { case IEEE80211_REG_LPI_AP: return true; case IEEE80211_REG_SP_AP: -- cgit v1.2.3 From 4aa73c6051cb65046e6fa601b7877b5c1e6edc85 Mon Sep 17 00:00:00 2001 From: Heiner Kallweit Date: Wed, 12 Nov 2025 21:46:24 +0100 Subject: net: dsa: remove definition of struct dsa_switch_driver Since 93e86b3bc842 ("net: dsa: Remove legacy probing support") this struct has no user any longer. Signed-off-by: Heiner Kallweit Reviewed-by: Vladimir Oltean Link: https://patch.msgid.link/4053a98f-052f-4dc1-a3d4-ed9b3d3cc7cb@gmail.com Signed-off-by: Jakub Kicinski --- include/net/dsa.h | 5 ----- 1 file changed, 5 deletions(-) (limited to 'include/net') diff --git a/include/net/dsa.h b/include/net/dsa.h index 2df2e2ead9a8..97d5f401cfcf 100644 --- a/include/net/dsa.h +++ b/include/net/dsa.h @@ -1314,11 +1314,6 @@ static inline int dsa_devlink_port_to_port(struct devlink_port *port) return port->index; } -struct dsa_switch_driver { - struct list_head list; - const struct dsa_switch_ops *ops; -}; - bool dsa_fdb_present_in_other_db(struct dsa_switch *ds, int port, const unsigned char *addr, u16 vid, struct dsa_db db); -- cgit v1.2.3 From 6d650ae9282bcec1e76205b44cb8f17e2265052e Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Thu, 13 Nov 2025 14:03:57 +0000 Subject: tcp: gro: inline tcp_gro_pull_header() tcp_gro_pull_header() is used in GRO fast path, inline it. Signed-off-by: Eric Dumazet Link: https://patch.msgid.link/20251113140358.58242-1-edumazet@google.com Signed-off-by: Jakub Kicinski --- include/net/gro.h | 27 +++++++++++++++++++++++++++ include/net/tcp.h | 1 - net/ipv4/tcp_offload.c | 27 --------------------------- 3 files changed, 27 insertions(+), 28 deletions(-) (limited to 'include/net') diff --git a/include/net/gro.h b/include/net/gro.h index e3affb2e2ca8..b65f631c521d 100644 --- a/include/net/gro.h +++ b/include/net/gro.h @@ -593,4 +593,31 @@ static inline void inet6_get_iif_sdif(const struct sk_buff *skb, int *iif, int * struct packet_offload *gro_find_receive_by_type(__be16 type); struct packet_offload *gro_find_complete_by_type(__be16 type); +static inline struct tcphdr *tcp_gro_pull_header(struct sk_buff *skb) +{ + unsigned int thlen, hlen, off; + struct tcphdr *th; + + off = skb_gro_offset(skb); + hlen = off + sizeof(*th); + th = skb_gro_header(skb, hlen, off); + if (unlikely(!th)) + return NULL; + + thlen = th->doff * 4; + if (unlikely(thlen < sizeof(*th))) + return NULL; + + hlen = off + thlen; + if (!skb_gro_may_pull(skb, hlen)) { + th = skb_gro_header_slow(skb, hlen, off); + if (unlikely(!th)) + return NULL; + } + + skb_gro_pull(skb, thlen); + + return th; +} + #endif /* _NET_GRO_H */ diff --git a/include/net/tcp.h b/include/net/tcp.h index 4833ec7903ec..0deb5e9dd911 100644 --- a/include/net/tcp.h +++ b/include/net/tcp.h @@ -2313,7 +2313,6 @@ void tcp_v4_destroy_sock(struct sock *sk); struct sk_buff *tcp_gso_segment(struct sk_buff *skb, netdev_features_t features); -struct tcphdr *tcp_gro_pull_header(struct sk_buff *skb); struct sk_buff *tcp_gro_lookup(struct list_head *head, struct tcphdr *th); struct sk_buff *tcp_gro_receive(struct list_head *head, struct sk_buff *skb, struct tcphdr *th); diff --git a/net/ipv4/tcp_offload.c b/net/ipv4/tcp_offload.c index 2cb93da93abc..fdda18b1abda 100644 --- a/net/ipv4/tcp_offload.c +++ b/net/ipv4/tcp_offload.c @@ -282,33 +282,6 @@ struct sk_buff *tcp_gro_lookup(struct list_head *head, struct tcphdr *th) return NULL; } -struct tcphdr *tcp_gro_pull_header(struct sk_buff *skb) -{ - unsigned int thlen, hlen, off; - struct tcphdr *th; - - off = skb_gro_offset(skb); - hlen = off + sizeof(*th); - th = skb_gro_header(skb, hlen, off); - if (unlikely(!th)) - return NULL; - - thlen = th->doff * 4; - if (thlen < sizeof(*th)) - return NULL; - - hlen = off + thlen; - if (!skb_gro_may_pull(skb, hlen)) { - th = skb_gro_header_slow(skb, hlen, off); - if (unlikely(!th)) - return NULL; - } - - skb_gro_pull(skb, thlen); - - return th; -} - struct sk_buff *tcp_gro_receive(struct list_head *head, struct sk_buff *skb, struct tcphdr *th) { -- cgit v1.2.3 From 06ac470658190e97518f131df01c9c530c293320 Mon Sep 17 00:00:00 2001 From: Yue Haibing Date: Thu, 13 Nov 2025 19:45:01 +0800 Subject: sctp: Remove unused declaration sctp_auth_init_hmacs() Commit bf40785fa437 ("sctp: Use HMAC-SHA1 and HMAC-SHA256 library for chunk authentication") removed the implementation but leave declaration. Signed-off-by: Yue Haibing Reviewed-by: Eric Biggers Link: https://patch.msgid.link/20251113114501.32905-1-yuehaibing@huawei.com Signed-off-by: Jakub Kicinski --- include/net/sctp/auth.h | 1 - 1 file changed, 1 deletion(-) (limited to 'include/net') diff --git a/include/net/sctp/auth.h b/include/net/sctp/auth.h index 3d5879e08e78..6f2cd562b1de 100644 --- a/include/net/sctp/auth.h +++ b/include/net/sctp/auth.h @@ -72,7 +72,6 @@ struct sctp_shared_key *sctp_auth_get_shkey( int sctp_auth_asoc_copy_shkeys(const struct sctp_endpoint *ep, struct sctp_association *asoc, gfp_t gfp); -int sctp_auth_init_hmacs(struct sctp_endpoint *ep, gfp_t gfp); const struct sctp_hmac *sctp_auth_get_hmac(__u16 hmac_id); const struct sctp_hmac * sctp_auth_asoc_get_hmac(const struct sctp_association *asoc); -- cgit v1.2.3 From d929525c2e30abee621bf71f143ba6104c81ff2b Mon Sep 17 00:00:00 2001 From: Shakeel Butt Date: Thu, 16 Oct 2025 09:10:35 -0700 Subject: memcg: net: track network throttling due to memcg memory pressure The kernel can throttle network sockets if the memory cgroup associated with the corresponding socket is under memory pressure. The throttling actions include clamping the transmit window, failing to expand receive or send buffers, aggressively prune out-of-order receive queue, FIN deferred to a retransmitted packet and more. Let's add memcg metric to track such throttling actions. At the moment memcg memory pressure is defined through vmpressure and in future it may be defined using PSI or we may add more flexible way for the users to define memory pressure, maybe through ebpf. However the potential throttling actions will remain the same, so this newly introduced metric will continue to track throttling actions irrespective of how memcg memory pressure is defined. Link: https://lkml.kernel.org/r/20251016161035.86161-1-shakeel.butt@linux.dev Signed-off-by: Shakeel Butt Acked-by: Roman Gushchin Reviewed-by: Kuniyuki Iwashima Reviewed-by: Daniel Sedlak Cc: David S. Miller Cc: Eric Dumazet Cc: Jakub Kacinski Cc: Johannes Weiner Cc: Michal Hocko Cc: Muchun Song Cc: Neal Cardwell Cc: Paolo Abeni Cc: Simon Horman Cc: Tejun Heo Cc: Willem de Bruijn Signed-off-by: Andrew Morton --- Documentation/admin-guide/cgroup-v2.rst | 4 ++++ include/linux/memcontrol.h | 1 + include/net/sock.h | 6 +++++- kernel/cgroup/cgroup.c | 1 + mm/memcontrol.c | 3 +++ 5 files changed, 14 insertions(+), 1 deletion(-) (limited to 'include/net') diff --git a/Documentation/admin-guide/cgroup-v2.rst b/Documentation/admin-guide/cgroup-v2.rst index 0e6c67ac585a..3345961c30ac 100644 --- a/Documentation/admin-guide/cgroup-v2.rst +++ b/Documentation/admin-guide/cgroup-v2.rst @@ -1515,6 +1515,10 @@ The following nested keys are defined. oom_group_kill The number of times a group OOM has occurred. + sock_throttled + The number of times network sockets associated with + this cgroup are throttled. + memory.events.local Similar to memory.events but the fields in the file are local to the cgroup i.e. not hierarchical. The file modified event diff --git a/include/linux/memcontrol.h b/include/linux/memcontrol.h index 873e510d6f8d..5fe254813123 100644 --- a/include/linux/memcontrol.h +++ b/include/linux/memcontrol.h @@ -52,6 +52,7 @@ enum memcg_memory_event { MEMCG_SWAP_HIGH, MEMCG_SWAP_MAX, MEMCG_SWAP_FAIL, + MEMCG_SOCK_THROTTLED, MEMCG_NR_MEMORY_EVENTS, }; diff --git a/include/net/sock.h b/include/net/sock.h index 60bcb13f045c..ff7d49af1619 100644 --- a/include/net/sock.h +++ b/include/net/sock.h @@ -2635,8 +2635,12 @@ static inline bool mem_cgroup_sk_under_memory_pressure(const struct sock *sk) #endif /* CONFIG_MEMCG_V1 */ do { - if (time_before64(get_jiffies_64(), mem_cgroup_get_socket_pressure(memcg))) + if (time_before64(get_jiffies_64(), + mem_cgroup_get_socket_pressure(memcg))) { + memcg_memory_event(mem_cgroup_from_sk(sk), + MEMCG_SOCK_THROTTLED); return true; + } } while ((memcg = parent_mem_cgroup(memcg))); return false; diff --git a/kernel/cgroup/cgroup.c b/kernel/cgroup/cgroup.c index fdee387f0d6b..8df671c59987 100644 --- a/kernel/cgroup/cgroup.c +++ b/kernel/cgroup/cgroup.c @@ -4704,6 +4704,7 @@ void cgroup_file_notify(struct cgroup_file *cfile) } spin_unlock_irqrestore(&cgroup_file_kn_lock, flags); } +EXPORT_SYMBOL_GPL(cgroup_file_notify); /** * cgroup_file_show - show or hide a hidden cgroup file diff --git a/mm/memcontrol.c b/mm/memcontrol.c index 3ae5cbcaed75..976412c8196e 100644 --- a/mm/memcontrol.c +++ b/mm/memcontrol.c @@ -81,6 +81,7 @@ struct cgroup_subsys memory_cgrp_subsys __read_mostly; EXPORT_SYMBOL(memory_cgrp_subsys); struct mem_cgroup *root_mem_cgroup __read_mostly; +EXPORT_SYMBOL(root_mem_cgroup); /* Active memory cgroup to use from an interrupt context */ DEFINE_PER_CPU(struct mem_cgroup *, int_active_memcg); @@ -4463,6 +4464,8 @@ static void __memory_events_show(struct seq_file *m, atomic_long_t *events) atomic_long_read(&events[MEMCG_OOM_KILL])); seq_printf(m, "oom_group_kill %lu\n", atomic_long_read(&events[MEMCG_OOM_GROUP_KILL])); + seq_printf(m, "sock_throttled %lu\n", + atomic_long_read(&events[MEMCG_SOCK_THROTTLED])); } static int memory_events_show(struct seq_file *m, void *v) -- cgit v1.2.3 From e275d9091c01b3b46f3ec534ce4ac77cffc9e3ae Mon Sep 17 00:00:00 2001 From: Erni Sri Satya Vennela Date: Fri, 14 Nov 2025 03:43:18 -0800 Subject: net: mana: Move hardware counter stats from per-port to per-VF context Move hardware counter (HC) statistics from mana_port_context to mana_context to enable sharing stats across multiple network ports on the same MANA VF. Previously, each network port queried hardware counters independently using MANA_QUERY_GF_STAT command (GF = Generic Function stats from GDMA hardware), resulting in redundant queries when multiple ports existed on the same device. Isolate hardware counter stats by introducing mana_ethtool_hc_stats in mana_context and update the code to ensure all stats are properly reported via ethtool -S , maintaining consistency with previous behavior. Signed-off-by: Erni Sri Satya Vennela Reviewed-by: Haiyang Zhang Link: https://patch.msgid.link/1763120599-6331-2-git-send-email-ernis@linux.microsoft.com Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/microsoft/mana/mana_en.c | 67 ++++++++--------- drivers/net/ethernet/microsoft/mana/mana_ethtool.c | 85 ++++++++++++---------- include/net/mana/mana.h | 14 ++-- 3 files changed, 90 insertions(+), 76 deletions(-) (limited to 'include/net') diff --git a/drivers/net/ethernet/microsoft/mana/mana_en.c b/drivers/net/ethernet/microsoft/mana/mana_en.c index cccd5b63cee6..d8ce4402c696 100644 --- a/drivers/net/ethernet/microsoft/mana/mana_en.c +++ b/drivers/net/ethernet/microsoft/mana/mana_en.c @@ -2809,11 +2809,12 @@ int mana_config_rss(struct mana_port_context *apc, enum TRI_STATE rx, return 0; } -void mana_query_gf_stats(struct mana_port_context *apc) +void mana_query_gf_stats(struct mana_context *ac) { + struct gdma_context *gc = ac->gdma_dev->gdma_context; struct mana_query_gf_stat_resp resp = {}; struct mana_query_gf_stat_req req = {}; - struct net_device *ndev = apc->ndev; + struct device *dev = gc->dev; int err; mana_gd_init_req_hdr(&req.hdr, MANA_QUERY_GF_STAT, @@ -2847,52 +2848,52 @@ void mana_query_gf_stats(struct mana_port_context *apc) STATISTICS_FLAGS_HC_TX_BCAST_BYTES | STATISTICS_FLAGS_TX_ERRORS_GDMA_ERROR; - err = mana_send_request(apc->ac, &req, sizeof(req), &resp, + err = mana_send_request(ac, &req, sizeof(req), &resp, sizeof(resp)); if (err) { - netdev_err(ndev, "Failed to query GF stats: %d\n", err); + dev_err(dev, "Failed to query GF stats: %d\n", err); return; } err = mana_verify_resp_hdr(&resp.hdr, MANA_QUERY_GF_STAT, sizeof(resp)); if (err || resp.hdr.status) { - netdev_err(ndev, "Failed to query GF stats: %d, 0x%x\n", err, - resp.hdr.status); + dev_err(dev, "Failed to query GF stats: %d, 0x%x\n", err, + resp.hdr.status); return; } - apc->eth_stats.hc_rx_discards_no_wqe = resp.rx_discards_nowqe; - apc->eth_stats.hc_rx_err_vport_disabled = resp.rx_err_vport_disabled; - apc->eth_stats.hc_rx_bytes = resp.hc_rx_bytes; - apc->eth_stats.hc_rx_ucast_pkts = resp.hc_rx_ucast_pkts; - apc->eth_stats.hc_rx_ucast_bytes = resp.hc_rx_ucast_bytes; - apc->eth_stats.hc_rx_bcast_pkts = resp.hc_rx_bcast_pkts; - apc->eth_stats.hc_rx_bcast_bytes = resp.hc_rx_bcast_bytes; - apc->eth_stats.hc_rx_mcast_pkts = resp.hc_rx_mcast_pkts; - apc->eth_stats.hc_rx_mcast_bytes = resp.hc_rx_mcast_bytes; - apc->eth_stats.hc_tx_err_gf_disabled = resp.tx_err_gf_disabled; - apc->eth_stats.hc_tx_err_vport_disabled = resp.tx_err_vport_disabled; - apc->eth_stats.hc_tx_err_inval_vportoffset_pkt = + ac->hc_stats.hc_rx_discards_no_wqe = resp.rx_discards_nowqe; + ac->hc_stats.hc_rx_err_vport_disabled = resp.rx_err_vport_disabled; + ac->hc_stats.hc_rx_bytes = resp.hc_rx_bytes; + ac->hc_stats.hc_rx_ucast_pkts = resp.hc_rx_ucast_pkts; + ac->hc_stats.hc_rx_ucast_bytes = resp.hc_rx_ucast_bytes; + ac->hc_stats.hc_rx_bcast_pkts = resp.hc_rx_bcast_pkts; + ac->hc_stats.hc_rx_bcast_bytes = resp.hc_rx_bcast_bytes; + ac->hc_stats.hc_rx_mcast_pkts = resp.hc_rx_mcast_pkts; + ac->hc_stats.hc_rx_mcast_bytes = resp.hc_rx_mcast_bytes; + ac->hc_stats.hc_tx_err_gf_disabled = resp.tx_err_gf_disabled; + ac->hc_stats.hc_tx_err_vport_disabled = resp.tx_err_vport_disabled; + ac->hc_stats.hc_tx_err_inval_vportoffset_pkt = resp.tx_err_inval_vport_offset_pkt; - apc->eth_stats.hc_tx_err_vlan_enforcement = + ac->hc_stats.hc_tx_err_vlan_enforcement = resp.tx_err_vlan_enforcement; - apc->eth_stats.hc_tx_err_eth_type_enforcement = + ac->hc_stats.hc_tx_err_eth_type_enforcement = resp.tx_err_ethtype_enforcement; - apc->eth_stats.hc_tx_err_sa_enforcement = resp.tx_err_SA_enforcement; - apc->eth_stats.hc_tx_err_sqpdid_enforcement = + ac->hc_stats.hc_tx_err_sa_enforcement = resp.tx_err_SA_enforcement; + ac->hc_stats.hc_tx_err_sqpdid_enforcement = resp.tx_err_SQPDID_enforcement; - apc->eth_stats.hc_tx_err_cqpdid_enforcement = + ac->hc_stats.hc_tx_err_cqpdid_enforcement = resp.tx_err_CQPDID_enforcement; - apc->eth_stats.hc_tx_err_mtu_violation = resp.tx_err_mtu_violation; - apc->eth_stats.hc_tx_err_inval_oob = resp.tx_err_inval_oob; - apc->eth_stats.hc_tx_bytes = resp.hc_tx_bytes; - apc->eth_stats.hc_tx_ucast_pkts = resp.hc_tx_ucast_pkts; - apc->eth_stats.hc_tx_ucast_bytes = resp.hc_tx_ucast_bytes; - apc->eth_stats.hc_tx_bcast_pkts = resp.hc_tx_bcast_pkts; - apc->eth_stats.hc_tx_bcast_bytes = resp.hc_tx_bcast_bytes; - apc->eth_stats.hc_tx_mcast_pkts = resp.hc_tx_mcast_pkts; - apc->eth_stats.hc_tx_mcast_bytes = resp.hc_tx_mcast_bytes; - apc->eth_stats.hc_tx_err_gdma = resp.tx_err_gdma; + ac->hc_stats.hc_tx_err_mtu_violation = resp.tx_err_mtu_violation; + ac->hc_stats.hc_tx_err_inval_oob = resp.tx_err_inval_oob; + ac->hc_stats.hc_tx_bytes = resp.hc_tx_bytes; + ac->hc_stats.hc_tx_ucast_pkts = resp.hc_tx_ucast_pkts; + ac->hc_stats.hc_tx_ucast_bytes = resp.hc_tx_ucast_bytes; + ac->hc_stats.hc_tx_bcast_pkts = resp.hc_tx_bcast_pkts; + ac->hc_stats.hc_tx_bcast_bytes = resp.hc_tx_bcast_bytes; + ac->hc_stats.hc_tx_mcast_pkts = resp.hc_tx_mcast_pkts; + ac->hc_stats.hc_tx_mcast_bytes = resp.hc_tx_mcast_bytes; + ac->hc_stats.hc_tx_err_gdma = resp.tx_err_gdma; } void mana_query_phy_stats(struct mana_port_context *apc) diff --git a/drivers/net/ethernet/microsoft/mana/mana_ethtool.c b/drivers/net/ethernet/microsoft/mana/mana_ethtool.c index a1afa75a9463..3dfd96146424 100644 --- a/drivers/net/ethernet/microsoft/mana/mana_ethtool.c +++ b/drivers/net/ethernet/microsoft/mana/mana_ethtool.c @@ -15,66 +15,69 @@ struct mana_stats_desc { static const struct mana_stats_desc mana_eth_stats[] = { {"stop_queue", offsetof(struct mana_ethtool_stats, stop_queue)}, {"wake_queue", offsetof(struct mana_ethtool_stats, wake_queue)}, - {"hc_rx_discards_no_wqe", offsetof(struct mana_ethtool_stats, + {"tx_cq_err", offsetof(struct mana_ethtool_stats, tx_cqe_err)}, + {"tx_cqe_unknown_type", offsetof(struct mana_ethtool_stats, + tx_cqe_unknown_type)}, + {"rx_coalesced_err", offsetof(struct mana_ethtool_stats, + rx_coalesced_err)}, + {"rx_cqe_unknown_type", offsetof(struct mana_ethtool_stats, + rx_cqe_unknown_type)}, +}; + +static const struct mana_stats_desc mana_hc_stats[] = { + {"hc_rx_discards_no_wqe", offsetof(struct mana_ethtool_hc_stats, hc_rx_discards_no_wqe)}, - {"hc_rx_err_vport_disabled", offsetof(struct mana_ethtool_stats, + {"hc_rx_err_vport_disabled", offsetof(struct mana_ethtool_hc_stats, hc_rx_err_vport_disabled)}, - {"hc_rx_bytes", offsetof(struct mana_ethtool_stats, hc_rx_bytes)}, - {"hc_rx_ucast_pkts", offsetof(struct mana_ethtool_stats, + {"hc_rx_bytes", offsetof(struct mana_ethtool_hc_stats, hc_rx_bytes)}, + {"hc_rx_ucast_pkts", offsetof(struct mana_ethtool_hc_stats, hc_rx_ucast_pkts)}, - {"hc_rx_ucast_bytes", offsetof(struct mana_ethtool_stats, + {"hc_rx_ucast_bytes", offsetof(struct mana_ethtool_hc_stats, hc_rx_ucast_bytes)}, - {"hc_rx_bcast_pkts", offsetof(struct mana_ethtool_stats, + {"hc_rx_bcast_pkts", offsetof(struct mana_ethtool_hc_stats, hc_rx_bcast_pkts)}, - {"hc_rx_bcast_bytes", offsetof(struct mana_ethtool_stats, + {"hc_rx_bcast_bytes", offsetof(struct mana_ethtool_hc_stats, hc_rx_bcast_bytes)}, - {"hc_rx_mcast_pkts", offsetof(struct mana_ethtool_stats, - hc_rx_mcast_pkts)}, - {"hc_rx_mcast_bytes", offsetof(struct mana_ethtool_stats, + {"hc_rx_mcast_pkts", offsetof(struct mana_ethtool_hc_stats, + hc_rx_mcast_pkts)}, + {"hc_rx_mcast_bytes", offsetof(struct mana_ethtool_hc_stats, hc_rx_mcast_bytes)}, - {"hc_tx_err_gf_disabled", offsetof(struct mana_ethtool_stats, + {"hc_tx_err_gf_disabled", offsetof(struct mana_ethtool_hc_stats, hc_tx_err_gf_disabled)}, - {"hc_tx_err_vport_disabled", offsetof(struct mana_ethtool_stats, + {"hc_tx_err_vport_disabled", offsetof(struct mana_ethtool_hc_stats, hc_tx_err_vport_disabled)}, {"hc_tx_err_inval_vportoffset_pkt", - offsetof(struct mana_ethtool_stats, + offsetof(struct mana_ethtool_hc_stats, hc_tx_err_inval_vportoffset_pkt)}, - {"hc_tx_err_vlan_enforcement", offsetof(struct mana_ethtool_stats, + {"hc_tx_err_vlan_enforcement", offsetof(struct mana_ethtool_hc_stats, hc_tx_err_vlan_enforcement)}, {"hc_tx_err_eth_type_enforcement", - offsetof(struct mana_ethtool_stats, hc_tx_err_eth_type_enforcement)}, - {"hc_tx_err_sa_enforcement", offsetof(struct mana_ethtool_stats, + offsetof(struct mana_ethtool_hc_stats, hc_tx_err_eth_type_enforcement)}, + {"hc_tx_err_sa_enforcement", offsetof(struct mana_ethtool_hc_stats, hc_tx_err_sa_enforcement)}, {"hc_tx_err_sqpdid_enforcement", - offsetof(struct mana_ethtool_stats, hc_tx_err_sqpdid_enforcement)}, + offsetof(struct mana_ethtool_hc_stats, hc_tx_err_sqpdid_enforcement)}, {"hc_tx_err_cqpdid_enforcement", - offsetof(struct mana_ethtool_stats, hc_tx_err_cqpdid_enforcement)}, - {"hc_tx_err_mtu_violation", offsetof(struct mana_ethtool_stats, + offsetof(struct mana_ethtool_hc_stats, hc_tx_err_cqpdid_enforcement)}, + {"hc_tx_err_mtu_violation", offsetof(struct mana_ethtool_hc_stats, hc_tx_err_mtu_violation)}, - {"hc_tx_err_inval_oob", offsetof(struct mana_ethtool_stats, + {"hc_tx_err_inval_oob", offsetof(struct mana_ethtool_hc_stats, hc_tx_err_inval_oob)}, - {"hc_tx_err_gdma", offsetof(struct mana_ethtool_stats, + {"hc_tx_err_gdma", offsetof(struct mana_ethtool_hc_stats, hc_tx_err_gdma)}, - {"hc_tx_bytes", offsetof(struct mana_ethtool_stats, hc_tx_bytes)}, - {"hc_tx_ucast_pkts", offsetof(struct mana_ethtool_stats, + {"hc_tx_bytes", offsetof(struct mana_ethtool_hc_stats, hc_tx_bytes)}, + {"hc_tx_ucast_pkts", offsetof(struct mana_ethtool_hc_stats, hc_tx_ucast_pkts)}, - {"hc_tx_ucast_bytes", offsetof(struct mana_ethtool_stats, + {"hc_tx_ucast_bytes", offsetof(struct mana_ethtool_hc_stats, hc_tx_ucast_bytes)}, - {"hc_tx_bcast_pkts", offsetof(struct mana_ethtool_stats, + {"hc_tx_bcast_pkts", offsetof(struct mana_ethtool_hc_stats, hc_tx_bcast_pkts)}, - {"hc_tx_bcast_bytes", offsetof(struct mana_ethtool_stats, + {"hc_tx_bcast_bytes", offsetof(struct mana_ethtool_hc_stats, hc_tx_bcast_bytes)}, - {"hc_tx_mcast_pkts", offsetof(struct mana_ethtool_stats, + {"hc_tx_mcast_pkts", offsetof(struct mana_ethtool_hc_stats, hc_tx_mcast_pkts)}, - {"hc_tx_mcast_bytes", offsetof(struct mana_ethtool_stats, + {"hc_tx_mcast_bytes", offsetof(struct mana_ethtool_hc_stats, hc_tx_mcast_bytes)}, - {"tx_cq_err", offsetof(struct mana_ethtool_stats, tx_cqe_err)}, - {"tx_cqe_unknown_type", offsetof(struct mana_ethtool_stats, - tx_cqe_unknown_type)}, - {"rx_coalesced_err", offsetof(struct mana_ethtool_stats, - rx_coalesced_err)}, - {"rx_cqe_unknown_type", offsetof(struct mana_ethtool_stats, - rx_cqe_unknown_type)}, }; static const struct mana_stats_desc mana_phy_stats[] = { @@ -138,7 +141,7 @@ static int mana_get_sset_count(struct net_device *ndev, int stringset) if (stringset != ETH_SS_STATS) return -EINVAL; - return ARRAY_SIZE(mana_eth_stats) + ARRAY_SIZE(mana_phy_stats) + + return ARRAY_SIZE(mana_eth_stats) + ARRAY_SIZE(mana_phy_stats) + ARRAY_SIZE(mana_hc_stats) + num_queues * (MANA_STATS_RX_COUNT + MANA_STATS_TX_COUNT); } @@ -150,10 +153,12 @@ static void mana_get_strings(struct net_device *ndev, u32 stringset, u8 *data) if (stringset != ETH_SS_STATS) return; - for (i = 0; i < ARRAY_SIZE(mana_eth_stats); i++) ethtool_puts(&data, mana_eth_stats[i].name); + for (i = 0; i < ARRAY_SIZE(mana_hc_stats); i++) + ethtool_puts(&data, mana_hc_stats[i].name); + for (i = 0; i < ARRAY_SIZE(mana_phy_stats); i++) ethtool_puts(&data, mana_phy_stats[i].name); @@ -186,6 +191,7 @@ static void mana_get_ethtool_stats(struct net_device *ndev, struct mana_port_context *apc = netdev_priv(ndev); unsigned int num_queues = apc->num_queues; void *eth_stats = &apc->eth_stats; + void *hc_stats = &apc->ac->hc_stats; void *phy_stats = &apc->phy_stats; struct mana_stats_rx *rx_stats; struct mana_stats_tx *tx_stats; @@ -208,7 +214,7 @@ static void mana_get_ethtool_stats(struct net_device *ndev, if (!apc->port_is_up) return; /* we call mana function to update stats from GDMA */ - mana_query_gf_stats(apc); + mana_query_gf_stats(apc->ac); /* We call this mana function to get the phy stats from GDMA and includes * aggregate tx/rx drop counters, Per-TC(Traffic Channel) tx/rx and pause @@ -219,6 +225,9 @@ static void mana_get_ethtool_stats(struct net_device *ndev, for (q = 0; q < ARRAY_SIZE(mana_eth_stats); q++) data[i++] = *(u64 *)(eth_stats + mana_eth_stats[q].offset); + for (q = 0; q < ARRAY_SIZE(mana_hc_stats); q++) + data[i++] = *(u64 *)(hc_stats + mana_hc_stats[q].offset); + for (q = 0; q < ARRAY_SIZE(mana_phy_stats); q++) data[i++] = *(u64 *)(phy_stats + mana_phy_stats[q].offset); diff --git a/include/net/mana/mana.h b/include/net/mana/mana.h index 8906901535f5..3484f42803e3 100644 --- a/include/net/mana/mana.h +++ b/include/net/mana/mana.h @@ -375,6 +375,13 @@ struct mana_tx_qp { struct mana_ethtool_stats { u64 stop_queue; u64 wake_queue; + u64 tx_cqe_err; + u64 tx_cqe_unknown_type; + u64 rx_coalesced_err; + u64 rx_cqe_unknown_type; +}; + +struct mana_ethtool_hc_stats { u64 hc_rx_discards_no_wqe; u64 hc_rx_err_vport_disabled; u64 hc_rx_bytes; @@ -402,10 +409,6 @@ struct mana_ethtool_stats { u64 hc_tx_mcast_pkts; u64 hc_tx_mcast_bytes; u64 hc_tx_err_gdma; - u64 tx_cqe_err; - u64 tx_cqe_unknown_type; - u64 rx_coalesced_err; - u64 rx_cqe_unknown_type; }; struct mana_ethtool_phy_stats { @@ -473,6 +476,7 @@ struct mana_context { u16 num_ports; u8 bm_hostmode; + struct mana_ethtool_hc_stats hc_stats; struct mana_eq *eqs; struct dentry *mana_eqs_debugfs; @@ -577,7 +581,7 @@ u32 mana_run_xdp(struct net_device *ndev, struct mana_rxq *rxq, struct bpf_prog *mana_xdp_get(struct mana_port_context *apc); void mana_chn_setxdp(struct mana_port_context *apc, struct bpf_prog *prog); int mana_bpf(struct net_device *ndev, struct netdev_bpf *bpf); -void mana_query_gf_stats(struct mana_port_context *apc); +void mana_query_gf_stats(struct mana_context *ac); int mana_query_link_cfg(struct mana_port_context *apc); int mana_set_bw_clamp(struct mana_port_context *apc, u32 speed, int enable_clamping); -- cgit v1.2.3 From be4f1d67ec56f23f37714ac73c01094e63c7ff28 Mon Sep 17 00:00:00 2001 From: Erni Sri Satya Vennela Date: Fri, 14 Nov 2025 03:43:19 -0800 Subject: net: mana: Add standard counter rx_missed_errors Report standard counter stats->rx_missed_errors using hc_rx_discards_no_wqe from the hardware. Add a global workqueue to periodically run mana_query_gf_stats every 2 seconds to get the latest info in eth_stats and define a driver capability flag to notify hardware of the periodic queries. To avoid repeated failures and log flooding, the workqueue is not rescheduled if mana_query_gf_stats fails on HWC timeout error and the stats are reset to 0. Other errors are transient which will not need a VF reset for recovery. Signed-off-by: Erni Sri Satya Vennela Reviewed-by: Haiyang Zhang Link: https://patch.msgid.link/1763120599-6331-3-git-send-email-ernis@linux.microsoft.com Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/microsoft/mana/mana_en.c | 36 ++++++++++++++++++++-- drivers/net/ethernet/microsoft/mana/mana_ethtool.c | 2 -- include/net/mana/gdma.h | 6 +++- include/net/mana/mana.h | 6 +++- 4 files changed, 43 insertions(+), 7 deletions(-) (limited to 'include/net') diff --git a/drivers/net/ethernet/microsoft/mana/mana_en.c b/drivers/net/ethernet/microsoft/mana/mana_en.c index d8ce4402c696..13f47be7aca6 100644 --- a/drivers/net/ethernet/microsoft/mana/mana_en.c +++ b/drivers/net/ethernet/microsoft/mana/mana_en.c @@ -534,6 +534,11 @@ static void mana_get_stats64(struct net_device *ndev, netdev_stats_to_stats64(st, &ndev->stats); + if (apc->ac->hwc_timeout_occurred) + netdev_warn_once(ndev, "HWC timeout occurred\n"); + + st->rx_missed_errors = apc->ac->hc_stats.hc_rx_discards_no_wqe; + for (q = 0; q < num_queues; q++) { rx_stats = &apc->rxqs[q]->stats; @@ -2809,7 +2814,7 @@ int mana_config_rss(struct mana_port_context *apc, enum TRI_STATE rx, return 0; } -void mana_query_gf_stats(struct mana_context *ac) +int mana_query_gf_stats(struct mana_context *ac) { struct gdma_context *gc = ac->gdma_dev->gdma_context; struct mana_query_gf_stat_resp resp = {}; @@ -2852,14 +2857,14 @@ void mana_query_gf_stats(struct mana_context *ac) sizeof(resp)); if (err) { dev_err(dev, "Failed to query GF stats: %d\n", err); - return; + return err; } err = mana_verify_resp_hdr(&resp.hdr, MANA_QUERY_GF_STAT, sizeof(resp)); if (err || resp.hdr.status) { dev_err(dev, "Failed to query GF stats: %d, 0x%x\n", err, resp.hdr.status); - return; + return err; } ac->hc_stats.hc_rx_discards_no_wqe = resp.rx_discards_nowqe; @@ -2894,6 +2899,8 @@ void mana_query_gf_stats(struct mana_context *ac) ac->hc_stats.hc_tx_mcast_pkts = resp.hc_tx_mcast_pkts; ac->hc_stats.hc_tx_mcast_bytes = resp.hc_tx_mcast_bytes; ac->hc_stats.hc_tx_err_gdma = resp.tx_err_gdma; + + return 0; } void mana_query_phy_stats(struct mana_port_context *apc) @@ -3428,6 +3435,24 @@ int mana_rdma_service_event(struct gdma_context *gc, enum gdma_service_type even return 0; } +#define MANA_GF_STATS_PERIOD (2 * HZ) + +static void mana_gf_stats_work_handler(struct work_struct *work) +{ + struct mana_context *ac = + container_of(to_delayed_work(work), struct mana_context, gf_stats_work); + int err; + + err = mana_query_gf_stats(ac); + if (err == -ETIMEDOUT) { + /* HWC timeout detected - reset stats and stop rescheduling */ + ac->hwc_timeout_occurred = true; + memset(&ac->hc_stats, 0, sizeof(ac->hc_stats)); + return; + } + schedule_delayed_work(&ac->gf_stats_work, MANA_GF_STATS_PERIOD); +} + int mana_probe(struct gdma_dev *gd, bool resuming) { struct gdma_context *gc = gd->gdma_context; @@ -3520,6 +3545,10 @@ int mana_probe(struct gdma_dev *gd, bool resuming) } err = add_adev(gd, "eth"); + + INIT_DELAYED_WORK(&ac->gf_stats_work, mana_gf_stats_work_handler); + schedule_delayed_work(&ac->gf_stats_work, MANA_GF_STATS_PERIOD); + out: if (err) { mana_remove(gd, false); @@ -3544,6 +3573,7 @@ void mana_remove(struct gdma_dev *gd, bool suspending) int i; disable_work_sync(&ac->link_change_work); + cancel_delayed_work_sync(&ac->gf_stats_work); /* adev currently doesn't support suspending, always remove it */ if (gd->adev) diff --git a/drivers/net/ethernet/microsoft/mana/mana_ethtool.c b/drivers/net/ethernet/microsoft/mana/mana_ethtool.c index 3dfd96146424..99e811208683 100644 --- a/drivers/net/ethernet/microsoft/mana/mana_ethtool.c +++ b/drivers/net/ethernet/microsoft/mana/mana_ethtool.c @@ -213,8 +213,6 @@ static void mana_get_ethtool_stats(struct net_device *ndev, if (!apc->port_is_up) return; - /* we call mana function to update stats from GDMA */ - mana_query_gf_stats(apc->ac); /* We call this mana function to get the phy stats from GDMA and includes * aggregate tx/rx drop counters, Per-TC(Traffic Channel) tx/rx and pause diff --git a/include/net/mana/gdma.h b/include/net/mana/gdma.h index 637f42485dba..2e4f2f3175e5 100644 --- a/include/net/mana/gdma.h +++ b/include/net/mana/gdma.h @@ -592,6 +592,9 @@ enum { #define GDMA_DRV_CAP_FLAG_1_HANDLE_RECONFIG_EQE BIT(17) #define GDMA_DRV_CAP_FLAG_1_HW_VPORT_LINK_AWARE BIT(6) +/* Driver can send HWC periodically to query stats */ +#define GDMA_DRV_CAP_FLAG_1_PERIODIC_STATS_QUERY BIT(21) + #define GDMA_DRV_CAP_FLAGS1 \ (GDMA_DRV_CAP_FLAG_1_EQ_SHARING_MULTI_VPORT | \ GDMA_DRV_CAP_FLAG_1_NAPI_WKDONE_FIX | \ @@ -601,7 +604,8 @@ enum { GDMA_DRV_CAP_FLAG_1_DYNAMIC_IRQ_ALLOC_SUPPORT | \ GDMA_DRV_CAP_FLAG_1_SELF_RESET_ON_EQE | \ GDMA_DRV_CAP_FLAG_1_HANDLE_RECONFIG_EQE | \ - GDMA_DRV_CAP_FLAG_1_HW_VPORT_LINK_AWARE) + GDMA_DRV_CAP_FLAG_1_HW_VPORT_LINK_AWARE | \ + GDMA_DRV_CAP_FLAG_1_PERIODIC_STATS_QUERY) #define GDMA_DRV_CAP_FLAGS2 0 diff --git a/include/net/mana/mana.h b/include/net/mana/mana.h index 3484f42803e3..d37f4cea0ac3 100644 --- a/include/net/mana/mana.h +++ b/include/net/mana/mana.h @@ -480,6 +480,10 @@ struct mana_context { struct mana_eq *eqs; struct dentry *mana_eqs_debugfs; + /* Workqueue for querying hardware stats */ + struct delayed_work gf_stats_work; + bool hwc_timeout_occurred; + struct net_device *ports[MAX_PORTS_IN_MANA_DEV]; /* Link state change work */ @@ -581,7 +585,7 @@ u32 mana_run_xdp(struct net_device *ndev, struct mana_rxq *rxq, struct bpf_prog *mana_xdp_get(struct mana_port_context *apc); void mana_chn_setxdp(struct mana_port_context *apc, struct bpf_prog *prog); int mana_bpf(struct net_device *ndev, struct netdev_bpf *bpf); -void mana_query_gf_stats(struct mana_context *ac); +int mana_query_gf_stats(struct mana_context *ac); int mana_query_link_cfg(struct mana_port_context *apc); int mana_set_bw_clamp(struct mana_port_context *apc, u32 speed, int enable_clamping); -- cgit v1.2.3 From 934fa943b53795339486cc0026b3ab7ad39dc600 Mon Sep 17 00:00:00 2001 From: Aditya Garg Date: Tue, 18 Nov 2025 03:11:08 -0800 Subject: net: mana: Handle SKB if TX SGEs exceed hardware limit The MANA hardware supports a maximum of 30 scatter-gather entries (SGEs) per TX WQE. Exceeding this limit can cause TX failures. Add ndo_features_check() callback to validate SKB layout before transmission. For GSO SKBs that would exceed the hardware SGE limit, clear NETIF_F_GSO_MASK to enforce software segmentation in the stack. Add a fallback in mana_start_xmit() to linearize non-GSO SKBs that still exceed the SGE limit. Also, Add ethtool counter for SKBs linearized Co-developed-by: Dipayaan Roy Signed-off-by: Dipayaan Roy Signed-off-by: Aditya Garg Reviewed-by: Eric Dumazet Reviewed-by: Haiyang Zhang Link: https://patch.msgid.link/1763464269-10431-2-git-send-email-gargaditya@linux.microsoft.com Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/microsoft/mana/mana_en.c | 40 ++++++++++++++++++++-- drivers/net/ethernet/microsoft/mana/mana_ethtool.c | 2 ++ include/net/mana/gdma.h | 8 ++++- include/net/mana/mana.h | 1 + 4 files changed, 48 insertions(+), 3 deletions(-) (limited to 'include/net') diff --git a/drivers/net/ethernet/microsoft/mana/mana_en.c b/drivers/net/ethernet/microsoft/mana/mana_en.c index 13f47be7aca6..7b49ab005e2d 100644 --- a/drivers/net/ethernet/microsoft/mana/mana_en.c +++ b/drivers/net/ethernet/microsoft/mana/mana_en.c @@ -11,6 +11,7 @@ #include #include #include +#include #include #include @@ -329,6 +330,21 @@ netdev_tx_t mana_start_xmit(struct sk_buff *skb, struct net_device *ndev) cq = &apc->tx_qp[txq_idx].tx_cq; tx_stats = &txq->stats; + BUILD_BUG_ON(MAX_TX_WQE_SGL_ENTRIES != MANA_MAX_TX_WQE_SGL_ENTRIES); + if (MAX_SKB_FRAGS + 2 > MAX_TX_WQE_SGL_ENTRIES && + skb_shinfo(skb)->nr_frags + 2 > MAX_TX_WQE_SGL_ENTRIES) { + /* GSO skb with Hardware SGE limit exceeded is not expected here + * as they are handled in mana_features_check() callback + */ + if (skb_linearize(skb)) { + netdev_warn_once(ndev, "Failed to linearize skb with nr_frags=%d and is_gso=%d\n", + skb_shinfo(skb)->nr_frags, + skb_is_gso(skb)); + goto tx_drop_count; + } + apc->eth_stats.tx_linear_pkt_cnt++; + } + pkg.tx_oob.s_oob.vcq_num = cq->gdma_id; pkg.tx_oob.s_oob.vsq_frame = txq->vsq_frame; @@ -442,8 +458,6 @@ netdev_tx_t mana_start_xmit(struct sk_buff *skb, struct net_device *ndev) } } - WARN_ON_ONCE(pkg.wqe_req.num_sge > MAX_TX_WQE_SGL_ENTRIES); - if (pkg.wqe_req.num_sge <= ARRAY_SIZE(pkg.sgl_array)) { pkg.wqe_req.sgl = pkg.sgl_array; } else { @@ -518,6 +532,25 @@ tx_drop: return NETDEV_TX_OK; } +#if (MAX_SKB_FRAGS + 2 > MANA_MAX_TX_WQE_SGL_ENTRIES) +static netdev_features_t mana_features_check(struct sk_buff *skb, + struct net_device *ndev, + netdev_features_t features) +{ + if (skb_shinfo(skb)->nr_frags + 2 > MAX_TX_WQE_SGL_ENTRIES) { + /* Exceeds HW SGE limit. + * GSO case: + * Disable GSO so the stack will software-segment the skb + * into smaller skbs that fit the SGE budget. + * Non-GSO case: + * The xmit path will attempt skb_linearize() as a fallback. + */ + features &= ~NETIF_F_GSO_MASK; + } + return features; +} +#endif + static void mana_get_stats64(struct net_device *ndev, struct rtnl_link_stats64 *st) { @@ -883,6 +916,9 @@ static const struct net_device_ops mana_devops = { .ndo_open = mana_open, .ndo_stop = mana_close, .ndo_select_queue = mana_select_queue, +#if (MAX_SKB_FRAGS + 2 > MANA_MAX_TX_WQE_SGL_ENTRIES) + .ndo_features_check = mana_features_check, +#endif .ndo_start_xmit = mana_start_xmit, .ndo_validate_addr = eth_validate_addr, .ndo_get_stats64 = mana_get_stats64, diff --git a/drivers/net/ethernet/microsoft/mana/mana_ethtool.c b/drivers/net/ethernet/microsoft/mana/mana_ethtool.c index 99e811208683..0e2f4343ac67 100644 --- a/drivers/net/ethernet/microsoft/mana/mana_ethtool.c +++ b/drivers/net/ethernet/microsoft/mana/mana_ethtool.c @@ -18,6 +18,8 @@ static const struct mana_stats_desc mana_eth_stats[] = { {"tx_cq_err", offsetof(struct mana_ethtool_stats, tx_cqe_err)}, {"tx_cqe_unknown_type", offsetof(struct mana_ethtool_stats, tx_cqe_unknown_type)}, + {"tx_linear_pkt_cnt", offsetof(struct mana_ethtool_stats, + tx_linear_pkt_cnt)}, {"rx_coalesced_err", offsetof(struct mana_ethtool_stats, rx_coalesced_err)}, {"rx_cqe_unknown_type", offsetof(struct mana_ethtool_stats, diff --git a/include/net/mana/gdma.h b/include/net/mana/gdma.h index 2e4f2f3175e5..a4cf307859f8 100644 --- a/include/net/mana/gdma.h +++ b/include/net/mana/gdma.h @@ -486,6 +486,8 @@ struct gdma_wqe { #define INLINE_OOB_SMALL_SIZE 8 #define INLINE_OOB_LARGE_SIZE 24 +#define MANA_MAX_TX_WQE_SGL_ENTRIES 30 + #define MAX_TX_WQE_SIZE 512 #define MAX_RX_WQE_SIZE 256 @@ -592,6 +594,9 @@ enum { #define GDMA_DRV_CAP_FLAG_1_HANDLE_RECONFIG_EQE BIT(17) #define GDMA_DRV_CAP_FLAG_1_HW_VPORT_LINK_AWARE BIT(6) +/* Driver supports linearizing the skb when num_sge exceeds hardware limit */ +#define GDMA_DRV_CAP_FLAG_1_SKB_LINEARIZE BIT(20) + /* Driver can send HWC periodically to query stats */ #define GDMA_DRV_CAP_FLAG_1_PERIODIC_STATS_QUERY BIT(21) @@ -605,7 +610,8 @@ enum { GDMA_DRV_CAP_FLAG_1_SELF_RESET_ON_EQE | \ GDMA_DRV_CAP_FLAG_1_HANDLE_RECONFIG_EQE | \ GDMA_DRV_CAP_FLAG_1_HW_VPORT_LINK_AWARE | \ - GDMA_DRV_CAP_FLAG_1_PERIODIC_STATS_QUERY) + GDMA_DRV_CAP_FLAG_1_PERIODIC_STATS_QUERY | \ + GDMA_DRV_CAP_FLAG_1_SKB_LINEARIZE) #define GDMA_DRV_CAP_FLAGS2 0 diff --git a/include/net/mana/mana.h b/include/net/mana/mana.h index d37f4cea0ac3..fb28b3cac067 100644 --- a/include/net/mana/mana.h +++ b/include/net/mana/mana.h @@ -377,6 +377,7 @@ struct mana_ethtool_stats { u64 wake_queue; u64 tx_cqe_err; u64 tx_cqe_unknown_type; + u64 tx_linear_pkt_cnt; u64 rx_coalesced_err; u64 rx_cqe_unknown_type; }; -- cgit v1.2.3 From 45120304e84171fd215c1b57b15b285446d15106 Mon Sep 17 00:00:00 2001 From: Aditya Garg Date: Tue, 18 Nov 2025 03:11:09 -0800 Subject: net: mana: Drop TX skb on post_work_request failure and unmap resources Drop TX packets when posting the work request fails and ensure DMA mappings are always cleaned up. Signed-off-by: Aditya Garg Reviewed-by: Haiyang Zhang Link: https://patch.msgid.link/1763464269-10431-3-git-send-email-gargaditya@linux.microsoft.com Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/microsoft/mana/gdma_main.c | 6 +----- drivers/net/ethernet/microsoft/mana/mana_en.c | 7 +++---- include/net/mana/mana.h | 1 + 3 files changed, 5 insertions(+), 9 deletions(-) (limited to 'include/net') diff --git a/drivers/net/ethernet/microsoft/mana/gdma_main.c b/drivers/net/ethernet/microsoft/mana/gdma_main.c index effe0a2f207a..8fd70b34807a 100644 --- a/drivers/net/ethernet/microsoft/mana/gdma_main.c +++ b/drivers/net/ethernet/microsoft/mana/gdma_main.c @@ -1300,7 +1300,6 @@ int mana_gd_post_work_request(struct gdma_queue *wq, struct gdma_posted_wqe_info *wqe_info) { u32 client_oob_size = wqe_req->inline_oob_size; - struct gdma_context *gc; u32 sgl_data_size; u32 max_wqe_size; u32 wqe_size; @@ -1330,11 +1329,8 @@ int mana_gd_post_work_request(struct gdma_queue *wq, if (wqe_size > max_wqe_size) return -EINVAL; - if (wq->monitor_avl_buf && wqe_size > mana_gd_wq_avail_space(wq)) { - gc = wq->gdma_dev->gdma_context; - dev_err(gc->dev, "unsuccessful flow control!\n"); + if (wq->monitor_avl_buf && wqe_size > mana_gd_wq_avail_space(wq)) return -ENOSPC; - } if (wqe_info) wqe_info->wqe_size_in_bu = wqe_size / GDMA_WQE_BU_SIZE; diff --git a/drivers/net/ethernet/microsoft/mana/mana_en.c b/drivers/net/ethernet/microsoft/mana/mana_en.c index 7b49ab005e2d..1ad154f9db1a 100644 --- a/drivers/net/ethernet/microsoft/mana/mana_en.c +++ b/drivers/net/ethernet/microsoft/mana/mana_en.c @@ -492,9 +492,9 @@ netdev_tx_t mana_start_xmit(struct sk_buff *skb, struct net_device *ndev) if (err) { (void)skb_dequeue_tail(&txq->pending_skbs); + mana_unmap_skb(skb, apc); netdev_warn(ndev, "Failed to post TX OOB: %d\n", err); - err = NETDEV_TX_BUSY; - goto tx_busy; + goto free_sgl_ptr; } err = NETDEV_TX_OK; @@ -514,7 +514,6 @@ netdev_tx_t mana_start_xmit(struct sk_buff *skb, struct net_device *ndev) tx_stats->bytes += len + ((num_gso_seg - 1) * gso_hs); u64_stats_update_end(&tx_stats->syncp); -tx_busy: if (netif_tx_queue_stopped(net_txq) && mana_can_tx(gdma_sq)) { netif_tx_wake_queue(net_txq); apc->eth_stats.wake_queue++; @@ -1687,7 +1686,7 @@ static int mana_move_wq_tail(struct gdma_queue *wq, u32 num_units) return 0; } -static void mana_unmap_skb(struct sk_buff *skb, struct mana_port_context *apc) +void mana_unmap_skb(struct sk_buff *skb, struct mana_port_context *apc) { struct mana_skb_head *ash = (struct mana_skb_head *)skb->head; struct gdma_context *gc = apc->ac->gdma_dev->gdma_context; diff --git a/include/net/mana/mana.h b/include/net/mana/mana.h index fb28b3cac067..d7e089c6b694 100644 --- a/include/net/mana/mana.h +++ b/include/net/mana/mana.h @@ -593,6 +593,7 @@ int mana_set_bw_clamp(struct mana_port_context *apc, u32 speed, void mana_query_phy_stats(struct mana_port_context *apc); int mana_pre_alloc_rxbufs(struct mana_port_context *apc, int mtu, int num_queues); void mana_pre_dealloc_rxbufs(struct mana_port_context *apc); +void mana_unmap_skb(struct sk_buff *skb, struct mana_port_context *apc); extern const struct ethtool_ops mana_ethtool_ops; extern struct dentry *mana_debugfs_root; -- cgit v1.2.3 From a77f0ad44fde89874654ba48f461209fb0382107 Mon Sep 17 00:00:00 2001 From: Pagadala Yesu Anjaneyulu Date: Wed, 12 Nov 2025 11:10:23 +0200 Subject: wifi: cfg80211: Add support for 6GHz AP role not relevant AP type Add IEEE80211_6GHZ_CTRL_REG_AP_ROLE_NOT_RELEVANT and map it to IEEE80211_REG_LPI_AP for safe regulatory compliance when AP role classification is not applicable. Use LPI as safe fallback to prevent power limit violations. Signed-off-by: Pagadala Yesu Anjaneyulu Reviewed-by: Johannes Berg Signed-off-by: Miri Korenblit Link: https://patch.msgid.link/20251112110828.856283677cc7.I36138a34847c3b4e680974bf347dde844448f3bc@changeid Signed-off-by: Johannes Berg --- include/linux/ieee80211-he.h | 1 + include/net/cfg80211.h | 1 + 2 files changed, 2 insertions(+) (limited to 'include/net') diff --git a/include/linux/ieee80211-he.h b/include/linux/ieee80211-he.h index 904d50db5bb8..a08c446fbb04 100644 --- a/include/linux/ieee80211-he.h +++ b/include/linux/ieee80211-he.h @@ -548,6 +548,7 @@ static inline bool ieee80211_he_capa_size_ok(const u8 *data, u8 len) #define IEEE80211_6GHZ_CTRL_REG_VLP_AP 2 #define IEEE80211_6GHZ_CTRL_REG_INDOOR_LPI_AP 3 #define IEEE80211_6GHZ_CTRL_REG_INDOOR_SP_AP_OLD 4 +#define IEEE80211_6GHZ_CTRL_REG_AP_ROLE_NOT_RELEVANT 7 #define IEEE80211_6GHZ_CTRL_REG_INDOOR_SP_AP 8 /** diff --git a/include/net/cfg80211.h b/include/net/cfg80211.h index 625cb2c78361..3d3ed1932262 100644 --- a/include/net/cfg80211.h +++ b/include/net/cfg80211.h @@ -10148,6 +10148,7 @@ cfg80211_6ghz_power_type(u8 control, u32 client_flags) switch (u8_get_bits(control, IEEE80211_HE_6GHZ_OPER_CTRL_REG_INFO)) { case IEEE80211_6GHZ_CTRL_REG_LPI_AP: case IEEE80211_6GHZ_CTRL_REG_INDOOR_LPI_AP: + case IEEE80211_6GHZ_CTRL_REG_AP_ROLE_NOT_RELEVANT: return IEEE80211_REG_LPI_AP; case IEEE80211_6GHZ_CTRL_REG_SP_AP: case IEEE80211_6GHZ_CTRL_REG_INDOOR_SP_AP_OLD: -- cgit v1.2.3 From 4015b979767125cf8a2233a145a3b3af78bfd8fb Mon Sep 17 00:00:00 2001 From: Chris Lu Date: Wed, 12 Nov 2025 15:53:34 +0800 Subject: Bluetooth: btusb: mediatek: Fix kernel crash when releasing mtk iso interface When performing reset tests and encountering abnormal card drop issues that lead to a kernel crash, it is necessary to perform a null check before releasing resources to avoid attempting to release a null pointer. <4>[ 29.158070] Hardware name: Google Quigon sku196612/196613 board (DT) <4>[ 29.158076] Workqueue: hci0 hci_cmd_sync_work [bluetooth] <4>[ 29.158154] pstate: 20400009 (nzCv daif +PAN -UAO -TCO -DIT -SSBS BTYPE=--) <4>[ 29.158162] pc : klist_remove+0x90/0x158 <4>[ 29.158174] lr : klist_remove+0x88/0x158 <4>[ 29.158180] sp : ffffffc0846b3c00 <4>[ 29.158185] pmr_save: 000000e0 <4>[ 29.158188] x29: ffffffc0846b3c30 x28: ffffff80cd31f880 x27: ffffff80c1bdc058 <4>[ 29.158199] x26: dead000000000100 x25: ffffffdbdc624ea3 x24: ffffff80c1bdc4c0 <4>[ 29.158209] x23: ffffffdbdc62a3e6 x22: ffffff80c6c07000 x21: ffffffdbdc829290 <4>[ 29.158219] x20: 0000000000000000 x19: ffffff80cd3e0648 x18: 000000031ec97781 <4>[ 29.158229] x17: ffffff80c1bdc4a8 x16: ffffffdc10576548 x15: ffffff80c1180428 <4>[ 29.158238] x14: 0000000000000000 x13: 000000000000e380 x12: 0000000000000018 <4>[ 29.158248] x11: ffffff80c2a7fd10 x10: 0000000000000000 x9 : 0000000100000000 <4>[ 29.158257] x8 : 0000000000000000 x7 : 7f7f7f7f7f7f7f7f x6 : 2d7223ff6364626d <4>[ 29.158266] x5 : 0000008000000000 x4 : 0000000000000020 x3 : 2e7325006465636e <4>[ 29.158275] x2 : ffffffdc11afeff8 x1 : 0000000000000000 x0 : ffffffdc11be4d0c <4>[ 29.158285] Call trace: <4>[ 29.158290] klist_remove+0x90/0x158 <4>[ 29.158298] device_release_driver_internal+0x20c/0x268 <4>[ 29.158308] device_release_driver+0x1c/0x30 <4>[ 29.158316] usb_driver_release_interface+0x70/0x88 <4>[ 29.158325] btusb_mtk_release_iso_intf+0x68/0xd8 [btusb (HASH:e8b6 5)] <4>[ 29.158347] btusb_mtk_reset+0x5c/0x480 [btusb (HASH:e8b6 5)] <4>[ 29.158361] hci_cmd_sync_work+0x10c/0x188 [bluetooth (HASH:a4fa 6)] <4>[ 29.158430] process_scheduled_works+0x258/0x4e8 <4>[ 29.158441] worker_thread+0x300/0x428 <4>[ 29.158448] kthread+0x108/0x1d0 <4>[ 29.158455] ret_from_fork+0x10/0x20 <0>[ 29.158467] Code: 91343000 940139d1 f9400268 927ff914 (f9401297) <4>[ 29.158474] ---[ end trace 0000000000000000 ]--- <0>[ 29.167129] Kernel panic - not syncing: Oops: Fatal exception <2>[ 29.167144] SMP: stopping secondary CPUs <4>[ 29.167158] ------------[ cut here ]------------ Fixes: ceac1cb0259d ("Bluetooth: btusb: mediatek: add ISO data transmission functions") Signed-off-by: Chris Lu Signed-off-by: Luiz Augusto von Dentz --- drivers/bluetooth/btusb.c | 34 +++++++++++++++++++++++++++------- include/net/bluetooth/hci_core.h | 1 - 2 files changed, 27 insertions(+), 8 deletions(-) (limited to 'include/net') diff --git a/drivers/bluetooth/btusb.c b/drivers/bluetooth/btusb.c index a722446ec73d..202a845e0236 100644 --- a/drivers/bluetooth/btusb.c +++ b/drivers/bluetooth/btusb.c @@ -2711,9 +2711,16 @@ static int btusb_recv_event_realtek(struct hci_dev *hdev, struct sk_buff *skb) static void btusb_mtk_claim_iso_intf(struct btusb_data *data) { - struct btmtk_data *btmtk_data = hci_get_priv(data->hdev); + struct btmtk_data *btmtk_data; int err; + if (!data->hdev) + return; + + btmtk_data = hci_get_priv(data->hdev); + if (!btmtk_data) + return; + /* * The function usb_driver_claim_interface() is documented to need * locks held if it's not called from a probe routine. The code here @@ -2735,17 +2742,30 @@ static void btusb_mtk_claim_iso_intf(struct btusb_data *data) static void btusb_mtk_release_iso_intf(struct hci_dev *hdev) { - struct btmtk_data *btmtk_data = hci_get_priv(hdev); + struct btmtk_data *btmtk_data; + + if (!hdev) + return; + + btmtk_data = hci_get_priv(hdev); + if (!btmtk_data) + return; if (test_bit(BTMTK_ISOPKT_OVER_INTR, &btmtk_data->flags)) { usb_kill_anchored_urbs(&btmtk_data->isopkt_anchor); clear_bit(BTMTK_ISOPKT_RUNNING, &btmtk_data->flags); - dev_kfree_skb_irq(btmtk_data->isopkt_skb); - btmtk_data->isopkt_skb = NULL; - usb_set_intfdata(btmtk_data->isopkt_intf, NULL); - usb_driver_release_interface(&btusb_driver, - btmtk_data->isopkt_intf); + if (btmtk_data->isopkt_skb) { + dev_kfree_skb_irq(btmtk_data->isopkt_skb); + btmtk_data->isopkt_skb = NULL; + } + + if (btmtk_data->isopkt_intf) { + usb_set_intfdata(btmtk_data->isopkt_intf, NULL); + usb_driver_release_interface(&btusb_driver, + btmtk_data->isopkt_intf); + btmtk_data->isopkt_intf = NULL; + } } clear_bit(BTMTK_ISOPKT_OVER_INTR, &btmtk_data->flags); diff --git a/include/net/bluetooth/hci_core.h b/include/net/bluetooth/hci_core.h index b8100dbfe5d7..32b1c08c8bba 100644 --- a/include/net/bluetooth/hci_core.h +++ b/include/net/bluetooth/hci_core.h @@ -749,7 +749,6 @@ struct hci_conn { __u8 remote_cap; __u8 remote_auth; - __u8 remote_id; unsigned int sent; -- cgit v1.2.3 From 79a2d4678ba90bdba577dc3af88cc900d6dcd5ee Mon Sep 17 00:00:00 2001 From: Pauli Virtanen Date: Sat, 15 Nov 2025 18:43:55 +0200 Subject: Bluetooth: hci_core: lookup hci_conn on RX path on protocol side The hdev lock/lookup/unlock/use pattern in the packet RX path doesn't ensure hci_conn* is not concurrently modified/deleted. This locking appears to be leftover from before conn_hash started using RCU commit bf4c63252490b ("Bluetooth: convert conn hash to RCU") and not clear if it had purpose since then. Currently, there are code paths that delete hci_conn* from elsewhere than the ordered hdev->workqueue where the RX work runs in. E.g. commit 5af1f84ed13a ("Bluetooth: hci_sync: Fix UAF on hci_abort_conn_sync") introduced some of these, and there probably were a few others before it. It's better to do the locking so that even if these run concurrently no UAF is possible. Move the lookup of hci_conn and associated socket-specific conn to protocol recv handlers, and do them within a single critical section to cover hci_conn* usage and lookup. syzkaller has reported a crash that appears to be this issue: [Task hdev->workqueue] [Task 2] hci_disconnect_all_sync l2cap_recv_acldata(hcon) hci_conn_get(hcon) hci_abort_conn_sync(hcon) hci_dev_lock hci_dev_lock hci_conn_del(hcon) v-------------------------------- hci_dev_unlock hci_conn_put(hcon) conn = hcon->l2cap_data (UAF) Fixes: 5af1f84ed13a ("Bluetooth: hci_sync: Fix UAF on hci_abort_conn_sync") Reported-by: syzbot+d32d77220b92eddd89ad@syzkaller.appspotmail.com Closes: https://syzkaller.appspot.com/bug?extid=d32d77220b92eddd89ad Signed-off-by: Pauli Virtanen Signed-off-by: Luiz Augusto von Dentz --- include/net/bluetooth/hci_core.h | 20 +++++++---- net/bluetooth/hci_core.c | 73 ++++++++++++++-------------------------- net/bluetooth/iso.c | 30 ++++++++++++++--- net/bluetooth/l2cap_core.c | 23 ++++++++++--- net/bluetooth/sco.c | 35 ++++++++++++++----- 5 files changed, 108 insertions(+), 73 deletions(-) (limited to 'include/net') diff --git a/include/net/bluetooth/hci_core.h b/include/net/bluetooth/hci_core.h index 32b1c08c8bba..0cb87687837f 100644 --- a/include/net/bluetooth/hci_core.h +++ b/include/net/bluetooth/hci_core.h @@ -856,11 +856,12 @@ extern struct mutex hci_cb_list_lock; /* ----- HCI interface to upper protocols ----- */ int l2cap_connect_ind(struct hci_dev *hdev, bdaddr_t *bdaddr); int l2cap_disconn_ind(struct hci_conn *hcon); -void l2cap_recv_acldata(struct hci_conn *hcon, struct sk_buff *skb, u16 flags); +int l2cap_recv_acldata(struct hci_dev *hdev, u16 handle, struct sk_buff *skb, + u16 flags); #if IS_ENABLED(CONFIG_BT_BREDR) int sco_connect_ind(struct hci_dev *hdev, bdaddr_t *bdaddr, __u8 *flags); -void sco_recv_scodata(struct hci_conn *hcon, struct sk_buff *skb); +int sco_recv_scodata(struct hci_dev *hdev, u16 handle, struct sk_buff *skb); #else static inline int sco_connect_ind(struct hci_dev *hdev, bdaddr_t *bdaddr, __u8 *flags) @@ -868,23 +869,30 @@ static inline int sco_connect_ind(struct hci_dev *hdev, bdaddr_t *bdaddr, return 0; } -static inline void sco_recv_scodata(struct hci_conn *hcon, struct sk_buff *skb) +static inline int sco_recv_scodata(struct hci_dev *hdev, u16 handle, + struct sk_buff *skb) { + kfree_skb(skb); + return -ENOENT; } #endif #if IS_ENABLED(CONFIG_BT_LE) int iso_connect_ind(struct hci_dev *hdev, bdaddr_t *bdaddr, __u8 *flags); -void iso_recv(struct hci_conn *hcon, struct sk_buff *skb, u16 flags); +int iso_recv(struct hci_dev *hdev, u16 handle, struct sk_buff *skb, + u16 flags); #else static inline int iso_connect_ind(struct hci_dev *hdev, bdaddr_t *bdaddr, __u8 *flags) { return 0; } -static inline void iso_recv(struct hci_conn *hcon, struct sk_buff *skb, - u16 flags) + +static inline int iso_recv(struct hci_dev *hdev, u16 handle, + struct sk_buff *skb, u16 flags) { + kfree_skb(skb); + return -ENOENT; } #endif diff --git a/net/bluetooth/hci_core.c b/net/bluetooth/hci_core.c index 1920e3d62bda..8ccec73dce45 100644 --- a/net/bluetooth/hci_core.c +++ b/net/bluetooth/hci_core.c @@ -3832,13 +3832,14 @@ static void hci_tx_work(struct work_struct *work) static void hci_acldata_packet(struct hci_dev *hdev, struct sk_buff *skb) { struct hci_acl_hdr *hdr; - struct hci_conn *conn; __u16 handle, flags; + int err; hdr = skb_pull_data(skb, sizeof(*hdr)); if (!hdr) { bt_dev_err(hdev, "ACL packet too small"); - goto drop; + kfree_skb(skb); + return; } handle = __le16_to_cpu(hdr->handle); @@ -3850,36 +3851,27 @@ static void hci_acldata_packet(struct hci_dev *hdev, struct sk_buff *skb) hdev->stat.acl_rx++; - hci_dev_lock(hdev); - conn = hci_conn_hash_lookup_handle(hdev, handle); - hci_dev_unlock(hdev); - - if (conn) { - hci_conn_enter_active_mode(conn, BT_POWER_FORCE_ACTIVE_OFF); - - /* Send to upper protocol */ - l2cap_recv_acldata(conn, skb, flags); - return; - } else { + err = l2cap_recv_acldata(hdev, handle, skb, flags); + if (err == -ENOENT) bt_dev_err(hdev, "ACL packet for unknown connection handle %d", handle); - } - -drop: - kfree_skb(skb); + else if (err) + bt_dev_dbg(hdev, "ACL packet recv for handle %d failed: %d", + handle, err); } /* SCO data packet */ static void hci_scodata_packet(struct hci_dev *hdev, struct sk_buff *skb) { struct hci_sco_hdr *hdr; - struct hci_conn *conn; __u16 handle, flags; + int err; hdr = skb_pull_data(skb, sizeof(*hdr)); if (!hdr) { bt_dev_err(hdev, "SCO packet too small"); - goto drop; + kfree_skb(skb); + return; } handle = __le16_to_cpu(hdr->handle); @@ -3891,34 +3883,28 @@ static void hci_scodata_packet(struct hci_dev *hdev, struct sk_buff *skb) hdev->stat.sco_rx++; - hci_dev_lock(hdev); - conn = hci_conn_hash_lookup_handle(hdev, handle); - hci_dev_unlock(hdev); + hci_skb_pkt_status(skb) = flags & 0x03; - if (conn) { - /* Send to upper protocol */ - hci_skb_pkt_status(skb) = flags & 0x03; - sco_recv_scodata(conn, skb); - return; - } else { + err = sco_recv_scodata(hdev, handle, skb); + if (err == -ENOENT) bt_dev_err_ratelimited(hdev, "SCO packet for unknown connection handle %d", handle); - } - -drop: - kfree_skb(skb); + else if (err) + bt_dev_dbg(hdev, "SCO packet recv for handle %d failed: %d", + handle, err); } static void hci_isodata_packet(struct hci_dev *hdev, struct sk_buff *skb) { struct hci_iso_hdr *hdr; - struct hci_conn *conn; __u16 handle, flags; + int err; hdr = skb_pull_data(skb, sizeof(*hdr)); if (!hdr) { bt_dev_err(hdev, "ISO packet too small"); - goto drop; + kfree_skb(skb); + return; } handle = __le16_to_cpu(hdr->handle); @@ -3928,22 +3914,13 @@ static void hci_isodata_packet(struct hci_dev *hdev, struct sk_buff *skb) bt_dev_dbg(hdev, "len %d handle 0x%4.4x flags 0x%4.4x", skb->len, handle, flags); - hci_dev_lock(hdev); - conn = hci_conn_hash_lookup_handle(hdev, handle); - hci_dev_unlock(hdev); - - if (!conn) { + err = iso_recv(hdev, handle, skb, flags); + if (err == -ENOENT) bt_dev_err(hdev, "ISO packet for unknown connection handle %d", handle); - goto drop; - } - - /* Send to upper protocol */ - iso_recv(conn, skb, flags); - return; - -drop: - kfree_skb(skb); + else if (err) + bt_dev_dbg(hdev, "ISO packet recv for handle %d failed: %d", + handle, err); } static bool hci_req_is_complete(struct hci_dev *hdev) diff --git a/net/bluetooth/iso.c b/net/bluetooth/iso.c index 3d98cb6291da..616c2fef91d2 100644 --- a/net/bluetooth/iso.c +++ b/net/bluetooth/iso.c @@ -2314,14 +2314,31 @@ static void iso_disconn_cfm(struct hci_conn *hcon, __u8 reason) iso_conn_del(hcon, bt_to_errno(reason)); } -void iso_recv(struct hci_conn *hcon, struct sk_buff *skb, u16 flags) +int iso_recv(struct hci_dev *hdev, u16 handle, struct sk_buff *skb, u16 flags) { - struct iso_conn *conn = hcon->iso_data; + struct hci_conn *hcon; + struct iso_conn *conn; struct skb_shared_hwtstamps *hwts; __u16 pb, ts, len, sn; - if (!conn) - goto drop; + hci_dev_lock(hdev); + + hcon = hci_conn_hash_lookup_handle(hdev, handle); + if (!hcon) { + hci_dev_unlock(hdev); + kfree_skb(skb); + return -ENOENT; + } + + conn = iso_conn_hold_unless_zero(hcon->iso_data); + hcon = NULL; + + hci_dev_unlock(hdev); + + if (!conn) { + kfree_skb(skb); + return -EINVAL; + } pb = hci_iso_flags_pb(flags); ts = hci_iso_flags_ts(flags); @@ -2377,7 +2394,7 @@ void iso_recv(struct hci_conn *hcon, struct sk_buff *skb, u16 flags) hci_skb_pkt_status(skb) = flags & 0x03; hci_skb_pkt_seqnum(skb) = sn; iso_recv_frame(conn, skb); - return; + goto done; } if (pb == ISO_SINGLE) { @@ -2455,6 +2472,9 @@ void iso_recv(struct hci_conn *hcon, struct sk_buff *skb, u16 flags) drop: kfree_skb(skb); +done: + iso_conn_put(conn); + return 0; } static struct hci_cb iso_cb = { diff --git a/net/bluetooth/l2cap_core.c b/net/bluetooth/l2cap_core.c index 35c57657bcf4..07b493331fd7 100644 --- a/net/bluetooth/l2cap_core.c +++ b/net/bluetooth/l2cap_core.c @@ -7510,13 +7510,24 @@ struct l2cap_conn *l2cap_conn_hold_unless_zero(struct l2cap_conn *c) return c; } -void l2cap_recv_acldata(struct hci_conn *hcon, struct sk_buff *skb, u16 flags) +int l2cap_recv_acldata(struct hci_dev *hdev, u16 handle, + struct sk_buff *skb, u16 flags) { + struct hci_conn *hcon; struct l2cap_conn *conn; int len; - /* Lock hdev to access l2cap_data to avoid race with l2cap_conn_del */ - hci_dev_lock(hcon->hdev); + /* Lock hdev for hci_conn, and race on l2cap_data vs. l2cap_conn_del */ + hci_dev_lock(hdev); + + hcon = hci_conn_hash_lookup_handle(hdev, handle); + if (!hcon) { + hci_dev_unlock(hdev); + kfree_skb(skb); + return -ENOENT; + } + + hci_conn_enter_active_mode(hcon, BT_POWER_FORCE_ACTIVE_OFF); conn = hcon->l2cap_data; @@ -7524,12 +7535,13 @@ void l2cap_recv_acldata(struct hci_conn *hcon, struct sk_buff *skb, u16 flags) conn = l2cap_conn_add(hcon); conn = l2cap_conn_hold_unless_zero(conn); + hcon = NULL; - hci_dev_unlock(hcon->hdev); + hci_dev_unlock(hdev); if (!conn) { kfree_skb(skb); - return; + return -EINVAL; } BT_DBG("conn %p len %u flags 0x%x", conn, skb->len, flags); @@ -7643,6 +7655,7 @@ drop: unlock: mutex_unlock(&conn->lock); l2cap_conn_put(conn); + return 0; } static struct hci_cb l2cap_cb = { diff --git a/net/bluetooth/sco.c b/net/bluetooth/sco.c index ab0cf442d57b..298c2a9ab4df 100644 --- a/net/bluetooth/sco.c +++ b/net/bluetooth/sco.c @@ -1458,22 +1458,39 @@ static void sco_disconn_cfm(struct hci_conn *hcon, __u8 reason) sco_conn_del(hcon, bt_to_errno(reason)); } -void sco_recv_scodata(struct hci_conn *hcon, struct sk_buff *skb) +int sco_recv_scodata(struct hci_dev *hdev, u16 handle, struct sk_buff *skb) { - struct sco_conn *conn = hcon->sco_data; + struct hci_conn *hcon; + struct sco_conn *conn; - if (!conn) - goto drop; + hci_dev_lock(hdev); + + hcon = hci_conn_hash_lookup_handle(hdev, handle); + if (!hcon) { + hci_dev_unlock(hdev); + kfree_skb(skb); + return -ENOENT; + } + + conn = sco_conn_hold_unless_zero(hcon->sco_data); + hcon = NULL; + + hci_dev_unlock(hdev); + + if (!conn) { + kfree_skb(skb); + return -EINVAL; + } BT_DBG("conn %p len %u", conn, skb->len); - if (skb->len) { + if (skb->len) sco_recv_frame(conn, skb); - return; - } + else + kfree_skb(skb); -drop: - kfree_skb(skb); + sco_conn_put(conn); + return 0; } static struct hci_cb sco_cb = { -- cgit v1.2.3 From 6d5dea68246ecb190a50a7fecbaf7f8c1ddb15e4 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Wed, 19 Nov 2025 08:48:12 +0000 Subject: tcp: tcp_moderate_rcvbuf is only used in rx path sysctl_tcp_moderate_rcvbuf is only used from tcp_rcvbuf_grow(). Move it to netns_ipv4_read_rx group. Remove various CACHELINE_ASSERT_GROUP_SIZE() from netns_ipv4_struct_check(), as they have no real benefit but cause pain for all changes. Signed-off-by: Eric Dumazet Reviewed-by: Kuniyuki Iwashima Link: https://patch.msgid.link/20251119084813.3684576-2-edumazet@google.com Signed-off-by: Jakub Kicinski --- Documentation/networking/net_cachelines/netns_ipv4_sysctl.rst | 2 +- include/net/netns/ipv4.h | 2 +- net/core/net_namespace.c | 9 ++------- 3 files changed, 4 insertions(+), 9 deletions(-) (limited to 'include/net') diff --git a/Documentation/networking/net_cachelines/netns_ipv4_sysctl.rst b/Documentation/networking/net_cachelines/netns_ipv4_sysctl.rst index 6e7b20afd2d4..5d5d54fb6ab1 100644 --- a/Documentation/networking/net_cachelines/netns_ipv4_sysctl.rst +++ b/Documentation/networking/net_cachelines/netns_ipv4_sysctl.rst @@ -102,7 +102,7 @@ u8 sysctl_tcp_app_win u8 sysctl_tcp_frto tcp_enter_loss u8 sysctl_tcp_nometrics_save TCP_LAST_ACK/tcp_update_metrics u8 sysctl_tcp_no_ssthresh_metrics_save TCP_LAST_ACK/tcp_(update/init)_metrics -u8 sysctl_tcp_moderate_rcvbuf read_mostly read_mostly tcp_tso_should_defer(tx);tcp_rcv_space_adjust(rx) +u8 sysctl_tcp_moderate_rcvbuf read_mostly tcp_rcvbuf_grow() u8 sysctl_tcp_tso_win_divisor read_mostly tcp_tso_should_defer(tcp_write_xmit) u8 sysctl_tcp_workaround_signed_windows tcp_select_window int sysctl_tcp_limit_output_bytes read_mostly tcp_small_queue_check(tcp_write_xmit) diff --git a/include/net/netns/ipv4.h b/include/net/netns/ipv4.h index de9d36acc8e2..11837d3ccc0a 100644 --- a/include/net/netns/ipv4.h +++ b/include/net/netns/ipv4.h @@ -74,11 +74,11 @@ struct netns_ipv4 { /* TXRX readonly hotpath cache lines */ __cacheline_group_begin(netns_ipv4_read_txrx); - u8 sysctl_tcp_moderate_rcvbuf; __cacheline_group_end(netns_ipv4_read_txrx); /* RX readonly hotpath cache line */ __cacheline_group_begin(netns_ipv4_read_rx); + u8 sysctl_tcp_moderate_rcvbuf; u8 sysctl_ip_early_demux; u8 sysctl_tcp_early_demux; u8 sysctl_tcp_l3mdev_accept; diff --git a/net/core/net_namespace.c b/net/core/net_namespace.c index adcfef55a66f..c8adbbe01451 100644 --- a/net/core/net_namespace.c +++ b/net/core/net_namespace.c @@ -1223,14 +1223,10 @@ static void __init netns_ipv4_struct_check(void) sysctl_tcp_wmem); CACHELINE_ASSERT_GROUP_MEMBER(struct netns_ipv4, netns_ipv4_read_tx, sysctl_ip_fwd_use_pmtu); - CACHELINE_ASSERT_GROUP_SIZE(struct netns_ipv4, netns_ipv4_read_tx, 33); - - /* TXRX readonly hotpath cache lines */ - CACHELINE_ASSERT_GROUP_MEMBER(struct netns_ipv4, netns_ipv4_read_txrx, - sysctl_tcp_moderate_rcvbuf); - CACHELINE_ASSERT_GROUP_SIZE(struct netns_ipv4, netns_ipv4_read_txrx, 1); /* RX readonly hotpath cache line */ + CACHELINE_ASSERT_GROUP_MEMBER(struct netns_ipv4, netns_ipv4_read_rx, + sysctl_tcp_moderate_rcvbuf); CACHELINE_ASSERT_GROUP_MEMBER(struct netns_ipv4, netns_ipv4_read_rx, sysctl_ip_early_demux); CACHELINE_ASSERT_GROUP_MEMBER(struct netns_ipv4, netns_ipv4_read_rx, @@ -1241,7 +1237,6 @@ static void __init netns_ipv4_struct_check(void) sysctl_tcp_reordering); CACHELINE_ASSERT_GROUP_MEMBER(struct netns_ipv4, netns_ipv4_read_rx, sysctl_tcp_rmem); - CACHELINE_ASSERT_GROUP_SIZE(struct netns_ipv4, netns_ipv4_read_rx, 22); } #endif -- cgit v1.2.3 From ecfea98b7d0d56c5bf2df3fc02c5501afa5cef6f Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Wed, 19 Nov 2025 08:48:13 +0000 Subject: tcp: add net.ipv4.tcp_rcvbuf_low_rtt This is a follow up of commit aa251c84636c ("tcp: fix too slow tcp_rcvbuf_grow() action") which brought again the issue that I tried to fix in commit 65c5287892e9 ("tcp: fix sk_rcvbuf overshoot") We also recently increased tcp_rmem[2] to 32 MB in commit 572be9bf9d0d ("tcp: increase tcp_rmem[2] to 32 MB") Idea of this patch is to not let tcp_rcvbuf_grow() grow sk->sk_rcvbuf too fast for small RTT flows. If sk->sk_rcvbuf is too big, this can force NIC driver to not recycle pages from their page pool, and also can cause cache evictions for DDIO enabled cpus/NIC, as receivers are usually slower than senders. Add net.ipv4.tcp_rcvbuf_low_rtt sysctl, set by default to 1000 usec (1 ms) If RTT if smaller than the sysctl value, use the RTT/tcp_rcvbuf_low_rtt ratio to control sk_rcvbuf inflation. Tested: Pair of hosts with a 200Gbit IDPF NIC. Using netperf/netserver Client initiates 8 TCP bulk flows, asking netserver to use CPU #10 only. super_netperf 8 -H server -T,10 -l 30 On server, use perf -e tcp:tcp_rcvbuf_grow while test is running. Before: sysctl -w net.ipv4.tcp_rcvbuf_low_rtt=1 perf record -a -e tcp:tcp_rcvbuf_grow sleep 30 ; perf script|tail -20|cut -c30-230 1153.051201: tcp:tcp_rcvbuf_grow: time=398 rtt_us=382 copied=6905856 inq=180224 space=6115328 ooo=0 scaling_ratio=240 rcvbuf=27666235 rcv_ssthresh=25878235 window_clamp=25937095 rcv_wnd=25600000 famil 1153.138752: tcp:tcp_rcvbuf_grow: time=446 rtt_us=413 copied=5529600 inq=180224 space=4505600 ooo=0 scaling_ratio=240 rcvbuf=23068672 rcv_ssthresh=21571860 window_clamp=21626880 rcv_wnd=21286912 famil 1153.361484: tcp:tcp_rcvbuf_grow: time=415 rtt_us=380 copied=7061504 inq=204800 space=6725632 ooo=0 scaling_ratio=240 rcvbuf=27666235 rcv_ssthresh=25878235 window_clamp=25937095 rcv_wnd=25600000 famil 1153.457642: tcp:tcp_rcvbuf_grow: time=483 rtt_us=421 copied=5885952 inq=720896 space=4407296 ooo=0 scaling_ratio=240 rcvbuf=23763511 rcv_ssthresh=22223271 window_clamp=22278291 rcv_wnd=21430272 famil 1153.466002: tcp:tcp_rcvbuf_grow: time=308 rtt_us=281 copied=3244032 inq=180224 space=2883584 ooo=0 scaling_ratio=240 rcvbuf=44854314 rcv_ssthresh=41992059 window_clamp=42050919 rcv_wnd=41713664 famil 1153.747792: tcp:tcp_rcvbuf_grow: time=394 rtt_us=332 copied=4460544 inq=585728 space=3063808 ooo=0 scaling_ratio=240 rcvbuf=44854314 rcv_ssthresh=41992059 window_clamp=42050919 rcv_wnd=41373696 famil 1154.260747: tcp:tcp_rcvbuf_grow: time=652 rtt_us=226 copied=10977280 inq=737280 space=9486336 ooo=0 scaling_ratio=240 rcvbuf=31165538 rcv_ssthresh=29197743 window_clamp=29217691 rcv_wnd=28368896 fami 1154.375019: tcp:tcp_rcvbuf_grow: time=461 rtt_us=443 copied=7573504 inq=507904 space=6856704 ooo=0 scaling_ratio=240 rcvbuf=27666235 rcv_ssthresh=25878235 window_clamp=25937095 rcv_wnd=25288704 famil 1154.463072: tcp:tcp_rcvbuf_grow: time=494 rtt_us=408 copied=7983104 inq=200704 space=7065600 ooo=0 scaling_ratio=240 rcvbuf=27666235 rcv_ssthresh=25878235 window_clamp=25937095 rcv_wnd=25579520 famil 1154.474658: tcp:tcp_rcvbuf_grow: time=507 rtt_us=459 copied=5586944 inq=540672 space=4718592 ooo=0 scaling_ratio=240 rcvbuf=17852266 rcv_ssthresh=16692999 window_clamp=16736499 rcv_wnd=16056320 famil 1154.584657: tcp:tcp_rcvbuf_grow: time=494 rtt_us=427 copied=8126464 inq=204800 space=7782400 ooo=0 scaling_ratio=240 rcvbuf=27666235 rcv_ssthresh=25878235 window_clamp=25937095 rcv_wnd=25600000 famil 1154.702117: tcp:tcp_rcvbuf_grow: time=480 rtt_us=406 copied=5734400 inq=180224 space=5349376 ooo=0 scaling_ratio=240 rcvbuf=23068672 rcv_ssthresh=21571860 window_clamp=21626880 rcv_wnd=21286912 famil 1155.941595: tcp:tcp_rcvbuf_grow: time=717 rtt_us=670 copied=11042816 inq=3784704 space=7159808 ooo=0 scaling_ratio=240 rcvbuf=19581357 rcv_ssthresh=18333222 window_clamp=18357522 rcv_wnd=14614528 fam 1156.384735: tcp:tcp_rcvbuf_grow: time=529 rtt_us=473 copied=9011200 inq=180224 space=7258112 ooo=0 scaling_ratio=240 rcvbuf=19581357 rcv_ssthresh=18333222 window_clamp=18357522 rcv_wnd=18018304 famil 1157.821676: tcp:tcp_rcvbuf_grow: time=529 rtt_us=272 copied=8224768 inq=602112 space=6545408 ooo=0 scaling_ratio=240 rcvbuf=67000000 rcv_ssthresh=62793576 window_clamp=62812500 rcv_wnd=62115840 famil 1158.906379: tcp:tcp_rcvbuf_grow: time=710 rtt_us=445 copied=11845632 inq=540672 space=10240000 ooo=0 scaling_ratio=240 rcvbuf=31165538 rcv_ssthresh=29205935 window_clamp=29217691 rcv_wnd=28536832 fam 1164.600160: tcp:tcp_rcvbuf_grow: time=841 rtt_us=430 copied=12976128 inq=1290240 space=11304960 ooo=0 scaling_ratio=240 rcvbuf=31165538 rcv_ssthresh=29212591 window_clamp=29217691 rcv_wnd=27856896 fa 1165.163572: tcp:tcp_rcvbuf_grow: time=845 rtt_us=800 copied=12632064 inq=540672 space=7921664 ooo=0 scaling_ratio=240 rcvbuf=27666235 rcv_ssthresh=25912795 window_clamp=25937095 rcv_wnd=25260032 fami 1165.653464: tcp:tcp_rcvbuf_grow: time=388 rtt_us=309 copied=4493312 inq=180224 space=3874816 ooo=0 scaling_ratio=240 rcvbuf=44854314 rcv_ssthresh=41995899 window_clamp=42050919 rcv_wnd=41713664 famil 1166.651211: tcp:tcp_rcvbuf_grow: time=556 rtt_us=553 copied=6328320 inq=540672 space=5554176 ooo=0 scaling_ratio=240 rcvbuf=23068672 rcv_ssthresh=21571860 window_clamp=21626880 rcv_wnd=20946944 famil After: sysctl -w net.ipv4.tcp_rcvbuf_low_rtt=1000 perf record -a -e tcp:tcp_rcvbuf_grow sleep 30 ; perf script|tail -20|cut -c30-230 1457.053149: tcp:tcp_rcvbuf_grow: time=128 rtt_us=24 copied=1441792 inq=40960 space=1269760 ooo=0 scaling_ratio=240 rcvbuf=2960741 rcv_ssthresh=2605474 window_clamp=2775694 rcv_wnd=2568192 family=AF_I 1458.000778: tcp:tcp_rcvbuf_grow: time=128 rtt_us=31 copied=1441792 inq=24576 space=1400832 ooo=0 scaling_ratio=240 rcvbuf=3060163 rcv_ssthresh=2810042 window_clamp=2868902 rcv_wnd=2674688 family=AF_I 1458.088059: tcp:tcp_rcvbuf_grow: time=190 rtt_us=110 copied=3227648 inq=385024 space=2781184 ooo=0 scaling_ratio=240 rcvbuf=6728240 rcv_ssthresh=6252705 window_clamp=6307725 rcv_wnd=5799936 family=AF 1458.148549: tcp:tcp_rcvbuf_grow: time=232 rtt_us=129 copied=3956736 inq=237568 space=2842624 ooo=0 scaling_ratio=240 rcvbuf=6731333 rcv_ssthresh=6252705 window_clamp=6310624 rcv_wnd=5918720 family=AF 1458.466861: tcp:tcp_rcvbuf_grow: time=193 rtt_us=83 copied=2949120 inq=180224 space=2457600 ooo=0 scaling_ratio=240 rcvbuf=5751438 rcv_ssthresh=5357689 window_clamp=5391973 rcv_wnd=5054464 family=AF_ 1458.775476: tcp:tcp_rcvbuf_grow: time=257 rtt_us=127 copied=4304896 inq=352256 space=3346432 ooo=0 scaling_ratio=240 rcvbuf=8067131 rcv_ssthresh=7523275 window_clamp=7562935 rcv_wnd=7061504 family=AF 1458.776631: tcp:tcp_rcvbuf_grow: time=200 rtt_us=96 copied=3260416 inq=143360 space=2768896 ooo=0 scaling_ratio=240 rcvbuf=6397256 rcv_ssthresh=5938567 window_clamp=5997427 rcv_wnd=5828608 family=AF_ 1459.707973: tcp:tcp_rcvbuf_grow: time=215 rtt_us=96 copied=2506752 inq=163840 space=1388544 ooo=0 scaling_ratio=240 rcvbuf=3068867 rcv_ssthresh=2768282 window_clamp=2877062 rcv_wnd=2555904 family=AF_ 1460.246494: tcp:tcp_rcvbuf_grow: time=231 rtt_us=80 copied=3756032 inq=204800 space=3117056 ooo=0 scaling_ratio=240 rcvbuf=7288091 rcv_ssthresh=6773725 window_clamp=6832585 rcv_wnd=6471680 family=AF_ 1460.714596: tcp:tcp_rcvbuf_grow: time=270 rtt_us=110 copied=4714496 inq=311296 space=3719168 ooo=0 scaling_ratio=240 rcvbuf=8957739 rcv_ssthresh=8339020 window_clamp=8397880 rcv_wnd=7933952 family=AF 1462.029977: tcp:tcp_rcvbuf_grow: time=101 rtt_us=19 copied=1105920 inq=40960 space=1036288 ooo=0 scaling_ratio=240 rcvbuf=2338970 rcv_ssthresh=2091684 window_clamp=2192784 rcv_wnd=1986560 family=AF_I 1462.802385: tcp:tcp_rcvbuf_grow: time=89 rtt_us=45 copied=1069056 inq=0 space=1064960 ooo=0 scaling_ratio=240 rcvbuf=2338970 rcv_ssthresh=2091684 window_clamp=2192784 rcv_wnd=2035712 family=AF_INET6 1462.918648: tcp:tcp_rcvbuf_grow: time=105 rtt_us=33 copied=1441792 inq=180224 space=1069056 ooo=0 scaling_ratio=240 rcvbuf=2383282 rcv_ssthresh=2091684 window_clamp=2234326 rcv_wnd=1896448 family=AF_ 1463.222533: tcp:tcp_rcvbuf_grow: time=273 rtt_us=144 copied=4603904 inq=385024 space=3469312 ooo=0 scaling_ratio=240 rcvbuf=8422564 rcv_ssthresh=7891053 window_clamp=7896153 rcv_wnd=7409664 family=AF 1466.519312: tcp:tcp_rcvbuf_grow: time=130 rtt_us=23 copied=1343488 inq=0 space=1261568 ooo=0 scaling_ratio=240 rcvbuf=2780158 rcv_ssthresh=2493778 window_clamp=2606398 rcv_wnd=2494464 family=AF_INET6 1466.681003: tcp:tcp_rcvbuf_grow: time=128 rtt_us=21 copied=1441792 inq=12288 space=1343488 ooo=0 scaling_ratio=240 rcvbuf=2932027 rcv_ssthresh=2578555 window_clamp=2748775 rcv_wnd=2568192 family=AF_I 1470.689959: tcp:tcp_rcvbuf_grow: time=255 rtt_us=122 copied=3932160 inq=204800 space=3551232 ooo=0 scaling_ratio=240 rcvbuf=8182038 rcv_ssthresh=7647384 window_clamp=7670660 rcv_wnd=7442432 family=AF 1471.754154: tcp:tcp_rcvbuf_grow: time=188 rtt_us=95 copied=2138112 inq=577536 space=1429504 ooo=0 scaling_ratio=240 rcvbuf=3113650 rcv_ssthresh=2806426 window_clamp=2919046 rcv_wnd=2248704 family=AF_ 1476.813542: tcp:tcp_rcvbuf_grow: time=269 rtt_us=99 copied=3088384 inq=180224 space=2564096 ooo=0 scaling_ratio=240 rcvbuf=6219470 rcv_ssthresh=5771893 window_clamp=5830753 rcv_wnd=5509120 family=AF_ 1477.738309: tcp:tcp_rcvbuf_grow: time=166 rtt_us=54 copied=1777664 inq=180224 space=1417216 ooo=0 scaling_ratio=240 rcvbuf=3117118 rcv_ssthresh=2874958 window_clamp=2922298 rcv_wnd=2613248 family=AF_ We can see sk_rcvbuf values are much smaller, and that rtt_us (estimation of rtt from a receiver point of view) is kept small, instead of being bloated. No difference in throughput. Signed-off-by: Eric Dumazet Reviewed-by: Kuniyuki Iwashima Tested-by: Paolo Abeni Link: https://patch.msgid.link/20251119084813.3684576-3-edumazet@google.com Signed-off-by: Jakub Kicinski --- Documentation/networking/ip-sysctl.rst | 10 ++++++++++ .../networking/net_cachelines/netns_ipv4_sysctl.rst | 1 + include/net/netns/ipv4.h | 1 + net/core/net_namespace.c | 2 ++ net/ipv4/sysctl_net_ipv4.c | 9 +++++++++ net/ipv4/tcp_input.c | 18 ++++++++++++++---- net/ipv4/tcp_ipv4.c | 1 + 7 files changed, 38 insertions(+), 4 deletions(-) (limited to 'include/net') diff --git a/Documentation/networking/ip-sysctl.rst b/Documentation/networking/ip-sysctl.rst index f4ad739a6b53..bc9a01606daf 100644 --- a/Documentation/networking/ip-sysctl.rst +++ b/Documentation/networking/ip-sysctl.rst @@ -673,6 +673,16 @@ tcp_moderate_rcvbuf - BOOLEAN Default: 1 (enabled) +tcp_rcvbuf_low_rtt - INTEGER + rcvbuf autotuning can over estimate final socket rcvbuf, which + can lead to cache trashing for high throughput flows. + + For small RTT flows (below tcp_rcvbuf_low_rtt usecs), we can relax + rcvbuf growth: Few additional ms to reach the final (and smaller) + rcvbuf is a good tradeoff. + + Default : 1000 (1 ms) + tcp_mtu_probing - INTEGER Controls TCP Packetization-Layer Path MTU Discovery. Takes three values: diff --git a/Documentation/networking/net_cachelines/netns_ipv4_sysctl.rst b/Documentation/networking/net_cachelines/netns_ipv4_sysctl.rst index 5d5d54fb6ab1..beaf1880a19b 100644 --- a/Documentation/networking/net_cachelines/netns_ipv4_sysctl.rst +++ b/Documentation/networking/net_cachelines/netns_ipv4_sysctl.rst @@ -103,6 +103,7 @@ u8 sysctl_tcp_frto u8 sysctl_tcp_nometrics_save TCP_LAST_ACK/tcp_update_metrics u8 sysctl_tcp_no_ssthresh_metrics_save TCP_LAST_ACK/tcp_(update/init)_metrics u8 sysctl_tcp_moderate_rcvbuf read_mostly tcp_rcvbuf_grow() +u32 sysctl_tcp_rcvbuf_low_rtt read_mostly tcp_rcvbuf_grow() u8 sysctl_tcp_tso_win_divisor read_mostly tcp_tso_should_defer(tcp_write_xmit) u8 sysctl_tcp_workaround_signed_windows tcp_select_window int sysctl_tcp_limit_output_bytes read_mostly tcp_small_queue_check(tcp_write_xmit) diff --git a/include/net/netns/ipv4.h b/include/net/netns/ipv4.h index 11837d3ccc0a..2dbd46fc4734 100644 --- a/include/net/netns/ipv4.h +++ b/include/net/netns/ipv4.h @@ -85,6 +85,7 @@ struct netns_ipv4 { /* 3 bytes hole, try to pack */ int sysctl_tcp_reordering; int sysctl_tcp_rmem[3]; + int sysctl_tcp_rcvbuf_low_rtt; __cacheline_group_end(netns_ipv4_read_rx); struct inet_timewait_death_row tcp_death_row; diff --git a/net/core/net_namespace.c b/net/core/net_namespace.c index c8adbbe01451..dfad7c03b809 100644 --- a/net/core/net_namespace.c +++ b/net/core/net_namespace.c @@ -1227,6 +1227,8 @@ static void __init netns_ipv4_struct_check(void) /* RX readonly hotpath cache line */ CACHELINE_ASSERT_GROUP_MEMBER(struct netns_ipv4, netns_ipv4_read_rx, sysctl_tcp_moderate_rcvbuf); + CACHELINE_ASSERT_GROUP_MEMBER(struct netns_ipv4, netns_ipv4_read_rx, + sysctl_tcp_rcvbuf_low_rtt); CACHELINE_ASSERT_GROUP_MEMBER(struct netns_ipv4, netns_ipv4_read_rx, sysctl_ip_early_demux); CACHELINE_ASSERT_GROUP_MEMBER(struct netns_ipv4, netns_ipv4_read_rx, diff --git a/net/ipv4/sysctl_net_ipv4.c b/net/ipv4/sysctl_net_ipv4.c index 35367f8e2da3..a1a50a5c80dc 100644 --- a/net/ipv4/sysctl_net_ipv4.c +++ b/net/ipv4/sysctl_net_ipv4.c @@ -1342,6 +1342,15 @@ static struct ctl_table ipv4_net_table[] = { .mode = 0644, .proc_handler = proc_dou8vec_minmax, }, + { + .procname = "tcp_rcvbuf_low_rtt", + .data = &init_net.ipv4.sysctl_tcp_rcvbuf_low_rtt, + .maxlen = sizeof(int), + .mode = 0644, + .proc_handler = proc_dointvec_minmax, + .extra1 = SYSCTL_ZERO, + .extra2 = SYSCTL_INT_MAX, + }, { .procname = "tcp_tso_win_divisor", .data = &init_net.ipv4.sysctl_tcp_tso_win_divisor, diff --git a/net/ipv4/tcp_input.c b/net/ipv4/tcp_input.c index 9df5d7515605..198f8a0d37be 100644 --- a/net/ipv4/tcp_input.c +++ b/net/ipv4/tcp_input.c @@ -896,6 +896,7 @@ void tcp_rcvbuf_grow(struct sock *sk, u32 newval) const struct net *net = sock_net(sk); struct tcp_sock *tp = tcp_sk(sk); u32 rcvwin, rcvbuf, cap, oldval; + u32 rtt_threshold, rtt_us; u64 grow; oldval = tp->rcvq_space.space; @@ -908,10 +909,19 @@ void tcp_rcvbuf_grow(struct sock *sk, u32 newval) /* DRS is always one RTT late. */ rcvwin = newval << 1; - /* slow start: allow the sender to double its rate. */ - grow = (u64)rcvwin * (newval - oldval); - do_div(grow, oldval); - rcvwin += grow << 1; + rtt_us = tp->rcv_rtt_est.rtt_us >> 3; + rtt_threshold = READ_ONCE(net->ipv4.sysctl_tcp_rcvbuf_low_rtt); + if (rtt_us < rtt_threshold) { + /* For small RTT, we set @grow to rcvwin * rtt_us/rtt_threshold. + * It might take few additional ms to reach 'line rate', + * but will avoid sk_rcvbuf inflation and poor cache use. + */ + grow = div_u64((u64)rcvwin * rtt_us, rtt_threshold); + } else { + /* slow start: allow the sender to double its rate. */ + grow = div_u64(((u64)rcvwin << 1) * (newval - oldval), oldval); + } + rcvwin += grow; if (!RB_EMPTY_ROOT(&tp->out_of_order_queue)) rcvwin += TCP_SKB_CB(tp->ooo_last_skb)->end_seq - tp->rcv_nxt; diff --git a/net/ipv4/tcp_ipv4.c b/net/ipv4/tcp_ipv4.c index 6fcaecb67284..e0bb8d9e2d9c 100644 --- a/net/ipv4/tcp_ipv4.c +++ b/net/ipv4/tcp_ipv4.c @@ -3566,6 +3566,7 @@ static int __net_init tcp_sk_init(struct net *net) net->ipv4.sysctl_tcp_adv_win_scale = 1; net->ipv4.sysctl_tcp_frto = 2; net->ipv4.sysctl_tcp_moderate_rcvbuf = 1; + net->ipv4.sysctl_tcp_rcvbuf_low_rtt = USEC_PER_MSEC; /* This limits the percentage of the congestion window which we * will allow a single TSO frame to consume. Building TSO frames * which are too large can cause TCP streams to be bursty. -- cgit v1.2.3 From 011d133bb988f80d597a9cbdab659414ba7ff72b Mon Sep 17 00:00:00 2001 From: Daniel Zahka Date: Tue, 18 Nov 2025 18:50:31 -0800 Subject: devlink: pass extack through to devlink_param::get() Allow devlink_param::get() handlers to report error messages via extack. This function is called in a few different contexts, but not all of them will have an valid extack to use. When devlink_param::get() is called from param_get_doit or param_get_dumpit contexts, pass the extack through so that drivers can report errors when retrieving param values. devlink_param::get() is called from the context of devlink_param_notify(), pass NULL in for the extack. Reviewed-by: Saeed Mahameed Reviewed-by: Aleksandr Loktionov Signed-off-by: Daniel Zahka Link: https://patch.msgid.link/20251119025038.651131-2-daniel.zahka@gmail.com Signed-off-by: Jakub Kicinski --- drivers/crypto/marvell/octeontx2/otx2_cpt_devlink.c | 6 ++++-- drivers/net/ethernet/amd/pds_core/core.h | 3 ++- drivers/net/ethernet/amd/pds_core/devlink.c | 3 ++- drivers/net/ethernet/broadcom/bnxt/bnxt_devlink.c | 6 ++++-- drivers/net/ethernet/intel/i40e/i40e_devlink.c | 3 ++- drivers/net/ethernet/intel/ice/devlink/devlink.c | 14 ++++++++++---- .../net/ethernet/marvell/octeontx2/af/rvu_devlink.c | 15 ++++++++++----- .../net/ethernet/marvell/octeontx2/nic/otx2_devlink.c | 6 ++++-- drivers/net/ethernet/mellanox/mlx4/main.c | 6 ++++-- drivers/net/ethernet/mellanox/mlx5/core/eswitch.c | 3 ++- .../ethernet/mellanox/mlx5/core/eswitch_offloads.c | 3 ++- drivers/net/ethernet/mellanox/mlx5/core/fs_core.c | 3 ++- drivers/net/ethernet/mellanox/mlx5/core/fw_reset.c | 3 ++- .../net/ethernet/mellanox/mlx5/core/lib/nv_param.c | 9 ++++++--- .../net/ethernet/mellanox/mlxsw/spectrum_acl_tcam.c | 3 ++- drivers/net/ethernet/netronome/nfp/devlink_param.c | 3 ++- drivers/net/ethernet/qlogic/qed/qed_devlink.c | 3 ++- drivers/net/ethernet/stmicro/stmmac/stmmac_main.c | 3 ++- drivers/net/ethernet/ti/am65-cpsw-nuss.c | 3 ++- drivers/net/ethernet/ti/cpsw_new.c | 6 ++++-- drivers/net/wwan/iosm/iosm_ipc_devlink.c | 3 ++- include/net/devlink.h | 3 ++- include/net/dsa.h | 3 ++- net/devlink/param.c | 19 +++++++++++-------- net/dsa/devlink.c | 3 ++- 25 files changed, 89 insertions(+), 46 deletions(-) (limited to 'include/net') diff --git a/drivers/crypto/marvell/octeontx2/otx2_cpt_devlink.c b/drivers/crypto/marvell/octeontx2/otx2_cpt_devlink.c index 215a1a8ba7e9..07a74f702c3a 100644 --- a/drivers/crypto/marvell/octeontx2/otx2_cpt_devlink.c +++ b/drivers/crypto/marvell/octeontx2/otx2_cpt_devlink.c @@ -24,7 +24,8 @@ static int otx2_cpt_dl_egrp_delete(struct devlink *dl, u32 id, } static int otx2_cpt_dl_uc_info(struct devlink *dl, u32 id, - struct devlink_param_gset_ctx *ctx) + struct devlink_param_gset_ctx *ctx, + struct netlink_ext_ack *extack) { ctx->val.vstr[0] = '\0'; @@ -32,7 +33,8 @@ static int otx2_cpt_dl_uc_info(struct devlink *dl, u32 id, } static int otx2_cpt_dl_t106_mode_get(struct devlink *dl, u32 id, - struct devlink_param_gset_ctx *ctx) + struct devlink_param_gset_ctx *ctx, + struct netlink_ext_ack *extack) { struct otx2_cpt_devlink *cpt_dl = devlink_priv(dl); struct otx2_cptpf_dev *cptpf = cpt_dl->cptpf; diff --git a/drivers/net/ethernet/amd/pds_core/core.h b/drivers/net/ethernet/amd/pds_core/core.h index 0b53a1fab46d..4a6b35c84dab 100644 --- a/drivers/net/ethernet/amd/pds_core/core.h +++ b/drivers/net/ethernet/amd/pds_core/core.h @@ -255,7 +255,8 @@ int pdsc_dl_flash_update(struct devlink *dl, struct devlink_flash_update_params *params, struct netlink_ext_ack *extack); int pdsc_dl_enable_get(struct devlink *dl, u32 id, - struct devlink_param_gset_ctx *ctx); + struct devlink_param_gset_ctx *ctx, + struct netlink_ext_ack *extack); int pdsc_dl_enable_set(struct devlink *dl, u32 id, struct devlink_param_gset_ctx *ctx, struct netlink_ext_ack *extack); diff --git a/drivers/net/ethernet/amd/pds_core/devlink.c b/drivers/net/ethernet/amd/pds_core/devlink.c index d8dc39da4161..b576be626a29 100644 --- a/drivers/net/ethernet/amd/pds_core/devlink.c +++ b/drivers/net/ethernet/amd/pds_core/devlink.c @@ -22,7 +22,8 @@ pdsc_viftype *pdsc_dl_find_viftype_by_id(struct pdsc *pdsc, } int pdsc_dl_enable_get(struct devlink *dl, u32 id, - struct devlink_param_gset_ctx *ctx) + struct devlink_param_gset_ctx *ctx, + struct netlink_ext_ack *extack) { struct pdsc *pdsc = devlink_priv(dl); struct pdsc_viftype *vt_entry; diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt_devlink.c b/drivers/net/ethernet/broadcom/bnxt/bnxt_devlink.c index 67ca02d84c97..15de802bbac4 100644 --- a/drivers/net/ethernet/broadcom/bnxt/bnxt_devlink.c +++ b/drivers/net/ethernet/broadcom/bnxt/bnxt_devlink.c @@ -1086,7 +1086,8 @@ static int bnxt_hwrm_nvm_req(struct bnxt *bp, u32 param_id, void *msg, } static int bnxt_dl_nvm_param_get(struct devlink *dl, u32 id, - struct devlink_param_gset_ctx *ctx) + struct devlink_param_gset_ctx *ctx, + struct netlink_ext_ack *extack) { struct bnxt *bp = bnxt_get_bp_from_dl(dl); struct hwrm_nvm_get_variable_input *req; @@ -1168,7 +1169,8 @@ static int bnxt_dl_msix_validate(struct devlink *dl, u32 id, } static int bnxt_remote_dev_reset_get(struct devlink *dl, u32 id, - struct devlink_param_gset_ctx *ctx) + struct devlink_param_gset_ctx *ctx, + struct netlink_ext_ack *extack) { struct bnxt *bp = bnxt_get_bp_from_dl(dl); diff --git a/drivers/net/ethernet/intel/i40e/i40e_devlink.c b/drivers/net/ethernet/intel/i40e/i40e_devlink.c index bc205e3077c7..229179ccc131 100644 --- a/drivers/net/ethernet/intel/i40e/i40e_devlink.c +++ b/drivers/net/ethernet/intel/i40e/i40e_devlink.c @@ -24,7 +24,8 @@ static int i40e_max_mac_per_vf_set(struct devlink *devlink, static int i40e_max_mac_per_vf_get(struct devlink *devlink, u32 id, - struct devlink_param_gset_ctx *ctx) + struct devlink_param_gset_ctx *ctx, + struct netlink_ext_ack *extack) { struct i40e_pf *pf = devlink_priv(devlink); diff --git a/drivers/net/ethernet/intel/ice/devlink/devlink.c b/drivers/net/ethernet/intel/ice/devlink/devlink.c index 938914abbe06..d88b7f3fd1f9 100644 --- a/drivers/net/ethernet/intel/ice/devlink/devlink.c +++ b/drivers/net/ethernet/intel/ice/devlink/devlink.c @@ -610,11 +610,13 @@ exit_release_res: * @devlink: pointer to the devlink instance * @id: the parameter ID to set * @ctx: context to store the parameter value + * @extack: netlink extended ACK structure * * Return: zero on success and negative value on failure. */ static int ice_devlink_tx_sched_layers_get(struct devlink *devlink, u32 id, - struct devlink_param_gset_ctx *ctx) + struct devlink_param_gset_ctx *ctx, + struct netlink_ext_ack *extack) { struct ice_pf *pf = devlink_priv(devlink); int err; @@ -1349,7 +1351,8 @@ static const struct devlink_ops ice_sf_devlink_ops; static int ice_devlink_enable_roce_get(struct devlink *devlink, u32 id, - struct devlink_param_gset_ctx *ctx) + struct devlink_param_gset_ctx *ctx, + struct netlink_ext_ack *extack) { struct ice_pf *pf = devlink_priv(devlink); struct iidc_rdma_core_dev_info *cdev; @@ -1415,7 +1418,8 @@ ice_devlink_enable_roce_validate(struct devlink *devlink, u32 id, static int ice_devlink_enable_iw_get(struct devlink *devlink, u32 id, - struct devlink_param_gset_ctx *ctx) + struct devlink_param_gset_ctx *ctx, + struct netlink_ext_ack *extack) { struct ice_pf *pf = devlink_priv(devlink); struct iidc_rdma_core_dev_info *cdev; @@ -1522,11 +1526,13 @@ static int ice_devlink_local_fwd_str_to_mode(const char *mode_str) * @devlink: Pointer to the devlink instance. * @id: The parameter ID to set. * @ctx: Context to store the parameter value. + * @extack: netlink extended ACK structure * * Return: Zero. */ static int ice_devlink_local_fwd_get(struct devlink *devlink, u32 id, - struct devlink_param_gset_ctx *ctx) + struct devlink_param_gset_ctx *ctx, + struct netlink_ext_ack *extack) { struct ice_pf *pf = devlink_priv(devlink); struct ice_port_info *pi; diff --git a/drivers/net/ethernet/marvell/octeontx2/af/rvu_devlink.c b/drivers/net/ethernet/marvell/octeontx2/af/rvu_devlink.c index 3735372539bd..0f9953eaf1b0 100644 --- a/drivers/net/ethernet/marvell/octeontx2/af/rvu_devlink.c +++ b/drivers/net/ethernet/marvell/octeontx2/af/rvu_devlink.c @@ -1233,7 +1233,8 @@ static int rvu_af_dl_dwrr_mtu_set(struct devlink *devlink, u32 id, } static int rvu_af_dl_dwrr_mtu_get(struct devlink *devlink, u32 id, - struct devlink_param_gset_ctx *ctx) + struct devlink_param_gset_ctx *ctx, + struct netlink_ext_ack *extack) { struct rvu_devlink *rvu_dl = devlink_priv(devlink); struct rvu *rvu = rvu_dl->rvu; @@ -1259,7 +1260,8 @@ enum rvu_af_dl_param_id { }; static int rvu_af_npc_exact_feature_get(struct devlink *devlink, u32 id, - struct devlink_param_gset_ctx *ctx) + struct devlink_param_gset_ctx *ctx, + struct netlink_ext_ack *extack) { struct rvu_devlink *rvu_dl = devlink_priv(devlink); struct rvu *rvu = rvu_dl->rvu; @@ -1314,7 +1316,8 @@ static int rvu_af_npc_exact_feature_validate(struct devlink *devlink, u32 id, } static int rvu_af_dl_npc_mcam_high_zone_percent_get(struct devlink *devlink, u32 id, - struct devlink_param_gset_ctx *ctx) + struct devlink_param_gset_ctx *ctx, + struct netlink_ext_ack *extack) { struct rvu_devlink *rvu_dl = devlink_priv(devlink); struct rvu *rvu = rvu_dl->rvu; @@ -1376,7 +1379,8 @@ static int rvu_af_dl_npc_mcam_high_zone_percent_validate(struct devlink *devlink } static int rvu_af_dl_npc_def_rule_cntr_get(struct devlink *devlink, u32 id, - struct devlink_param_gset_ctx *ctx) + struct devlink_param_gset_ctx *ctx, + struct netlink_ext_ack *extack) { struct rvu_devlink *rvu_dl = devlink_priv(devlink); struct rvu *rvu = rvu_dl->rvu; @@ -1402,7 +1406,8 @@ static int rvu_af_dl_npc_def_rule_cntr_set(struct devlink *devlink, u32 id, } static int rvu_af_dl_nix_maxlf_get(struct devlink *devlink, u32 id, - struct devlink_param_gset_ctx *ctx) + struct devlink_param_gset_ctx *ctx, + struct netlink_ext_ack *extack) { struct rvu_devlink *rvu_dl = devlink_priv(devlink); struct rvu *rvu = rvu_dl->rvu; diff --git a/drivers/net/ethernet/marvell/octeontx2/nic/otx2_devlink.c b/drivers/net/ethernet/marvell/octeontx2/nic/otx2_devlink.c index e13ae5484c19..a72694219df4 100644 --- a/drivers/net/ethernet/marvell/octeontx2/nic/otx2_devlink.c +++ b/drivers/net/ethernet/marvell/octeontx2/nic/otx2_devlink.c @@ -48,7 +48,8 @@ static int otx2_dl_mcam_count_set(struct devlink *devlink, u32 id, } static int otx2_dl_mcam_count_get(struct devlink *devlink, u32 id, - struct devlink_param_gset_ctx *ctx) + struct devlink_param_gset_ctx *ctx, + struct netlink_ext_ack *extack) { struct otx2_devlink *otx2_dl = devlink_priv(devlink); struct otx2_nic *pfvf = otx2_dl->pfvf; @@ -84,7 +85,8 @@ static int otx2_dl_ucast_flt_cnt_set(struct devlink *devlink, u32 id, } static int otx2_dl_ucast_flt_cnt_get(struct devlink *devlink, u32 id, - struct devlink_param_gset_ctx *ctx) + struct devlink_param_gset_ctx *ctx, + struct netlink_ext_ack *extack) { struct otx2_devlink *otx2_dl = devlink_priv(devlink); struct otx2_nic *pfvf = otx2_dl->pfvf; diff --git a/drivers/net/ethernet/mellanox/mlx4/main.c b/drivers/net/ethernet/mellanox/mlx4/main.c index 03d2fc7d9b09..2de226951e19 100644 --- a/drivers/net/ethernet/mellanox/mlx4/main.c +++ b/drivers/net/ethernet/mellanox/mlx4/main.c @@ -174,7 +174,8 @@ MODULE_PARM_DESC(port_type_array, "Array of port types: HW_DEFAULT (0) is defaul static atomic_t pf_loading = ATOMIC_INIT(0); static int mlx4_devlink_ierr_reset_get(struct devlink *devlink, u32 id, - struct devlink_param_gset_ctx *ctx) + struct devlink_param_gset_ctx *ctx, + struct netlink_ext_ack *extack) { ctx->val.vbool = !!mlx4_internal_err_reset; return 0; @@ -189,7 +190,8 @@ static int mlx4_devlink_ierr_reset_set(struct devlink *devlink, u32 id, } static int mlx4_devlink_crdump_snapshot_get(struct devlink *devlink, u32 id, - struct devlink_param_gset_ctx *ctx) + struct devlink_param_gset_ctx *ctx, + struct netlink_ext_ack *extack) { struct mlx4_priv *priv = devlink_priv(devlink); struct mlx4_dev *dev = &priv->dev; diff --git a/drivers/net/ethernet/mellanox/mlx5/core/eswitch.c b/drivers/net/ethernet/mellanox/mlx5/core/eswitch.c index 3adf2b1cd26a..4b7a1ce7f406 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/eswitch.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/eswitch.c @@ -1969,7 +1969,8 @@ static int mlx5_devlink_esw_multiport_set(struct devlink *devlink, u32 id, } static int mlx5_devlink_esw_multiport_get(struct devlink *devlink, u32 id, - struct devlink_param_gset_ctx *ctx) + struct devlink_param_gset_ctx *ctx, + struct netlink_ext_ack *extack) { struct mlx5_core_dev *dev = devlink_priv(devlink); diff --git a/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c b/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c index 8ebca0d17f65..8de6c7f6c294 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c @@ -2618,7 +2618,8 @@ done: } static int esw_port_metadata_get(struct devlink *devlink, u32 id, - struct devlink_param_gset_ctx *ctx) + struct devlink_param_gset_ctx *ctx, + struct netlink_ext_ack *extack) { struct mlx5_core_dev *dev = devlink_priv(devlink); diff --git a/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c b/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c index 2b755a0035ce..0a6031a64c6f 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c @@ -3833,7 +3833,8 @@ static int mlx5_fs_mode_set(struct devlink *devlink, u32 id, } static int mlx5_fs_mode_get(struct devlink *devlink, u32 id, - struct devlink_param_gset_ctx *ctx) + struct devlink_param_gset_ctx *ctx, + struct netlink_ext_ack *extack) { struct mlx5_core_dev *dev = devlink_priv(devlink); diff --git a/drivers/net/ethernet/mellanox/mlx5/core/fw_reset.c b/drivers/net/ethernet/mellanox/mlx5/core/fw_reset.c index 89e399606877..2bceb42c98cc 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/fw_reset.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/fw_reset.c @@ -73,7 +73,8 @@ static int mlx5_fw_reset_enable_remote_dev_reset_set(struct devlink *devlink, u3 } static int mlx5_fw_reset_enable_remote_dev_reset_get(struct devlink *devlink, u32 id, - struct devlink_param_gset_ctx *ctx) + struct devlink_param_gset_ctx *ctx, + struct netlink_ext_ack *extack) { struct mlx5_core_dev *dev = devlink_priv(devlink); struct mlx5_fw_reset *fw_reset; diff --git a/drivers/net/ethernet/mellanox/mlx5/core/lib/nv_param.c b/drivers/net/ethernet/mellanox/mlx5/core/lib/nv_param.c index 459a0b4d08e6..70cb22fa96db 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/lib/nv_param.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/lib/nv_param.c @@ -200,7 +200,8 @@ static const char *const static int mlx5_nv_param_devlink_cqe_compress_get(struct devlink *devlink, u32 id, - struct devlink_param_gset_ctx *ctx) + struct devlink_param_gset_ctx *ctx, + struct netlink_ext_ack *extack) { struct mlx5_core_dev *dev = devlink_priv(devlink); u32 mnvda[MLX5_ST_SZ_DW(mnvda_reg)] = {}; @@ -302,7 +303,8 @@ static int mlx5_nv_param_read_per_host_pf_conf(struct mlx5_core_dev *dev, } static int mlx5_devlink_enable_sriov_get(struct devlink *devlink, u32 id, - struct devlink_param_gset_ctx *ctx) + struct devlink_param_gset_ctx *ctx, + struct netlink_ext_ack *extack) { struct mlx5_core_dev *dev = devlink_priv(devlink); u32 mnvda[MLX5_ST_SZ_DW(mnvda_reg)] = {}; @@ -413,7 +415,8 @@ static int mlx5_devlink_enable_sriov_set(struct devlink *devlink, u32 id, } static int mlx5_devlink_total_vfs_get(struct devlink *devlink, u32 id, - struct devlink_param_gset_ctx *ctx) + struct devlink_param_gset_ctx *ctx, + struct netlink_ext_ack *extack) { struct mlx5_core_dev *dev = devlink_priv(devlink); u32 mnvda[MLX5_ST_SZ_DW(mnvda_reg)] = {}; diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_acl_tcam.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_acl_tcam.c index b1d08e958bf9..69f9da9fb305 100644 --- a/drivers/net/ethernet/mellanox/mlxsw/spectrum_acl_tcam.c +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_acl_tcam.c @@ -1489,7 +1489,8 @@ mlxsw_sp_acl_tcam_vregion_rehash(struct mlxsw_sp *mlxsw_sp, static int mlxsw_sp_acl_tcam_region_rehash_intrvl_get(struct devlink *devlink, u32 id, - struct devlink_param_gset_ctx *ctx) + struct devlink_param_gset_ctx *ctx, + struct netlink_ext_ack *extack) { struct mlxsw_core *mlxsw_core = devlink_priv(devlink); struct mlxsw_sp_acl_tcam *tcam; diff --git a/drivers/net/ethernet/netronome/nfp/devlink_param.c b/drivers/net/ethernet/netronome/nfp/devlink_param.c index 0e1a3800f371..85e3b19e6165 100644 --- a/drivers/net/ethernet/netronome/nfp/devlink_param.c +++ b/drivers/net/ethernet/netronome/nfp/devlink_param.c @@ -81,7 +81,8 @@ static const struct nfp_devlink_param_u8_arg nfp_devlink_u8_args[] = { static int nfp_devlink_param_u8_get(struct devlink *devlink, u32 id, - struct devlink_param_gset_ctx *ctx) + struct devlink_param_gset_ctx *ctx, + struct netlink_ext_ack *extack) { const struct nfp_devlink_param_u8_arg *arg; struct nfp_pf *pf = devlink_priv(devlink); diff --git a/drivers/net/ethernet/qlogic/qed/qed_devlink.c b/drivers/net/ethernet/qlogic/qed/qed_devlink.c index 94c5689b5abd..0c5278c0598c 100644 --- a/drivers/net/ethernet/qlogic/qed/qed_devlink.c +++ b/drivers/net/ethernet/qlogic/qed/qed_devlink.c @@ -121,7 +121,8 @@ void qed_fw_reporters_destroy(struct devlink *devlink) } static int qed_dl_param_get(struct devlink *dl, u32 id, - struct devlink_param_gset_ctx *ctx) + struct devlink_param_gset_ctx *ctx, + struct netlink_ext_ack *extack) { struct qed_devlink *qed_dl = devlink_priv(dl); struct qed_dev *cdev; diff --git a/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c b/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c index b1aa236d8051..6cacedb2c9b3 100644 --- a/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c +++ b/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c @@ -7531,7 +7531,8 @@ static int stmmac_dl_ts_coarse_set(struct devlink *dl, u32 id, } static int stmmac_dl_ts_coarse_get(struct devlink *dl, u32 id, - struct devlink_param_gset_ctx *ctx) + struct devlink_param_gset_ctx *ctx, + struct netlink_ext_ack *extack) { struct stmmac_devlink_priv *dl_priv = devlink_priv(dl); struct stmmac_priv *priv = dl_priv->stmmac_priv; diff --git a/drivers/net/ethernet/ti/am65-cpsw-nuss.c b/drivers/net/ethernet/ti/am65-cpsw-nuss.c index d5f358ec9820..5924db6be3fe 100644 --- a/drivers/net/ethernet/ti/am65-cpsw-nuss.c +++ b/drivers/net/ethernet/ti/am65-cpsw-nuss.c @@ -3068,7 +3068,8 @@ static void am65_cpsw_init_host_port_emac(struct am65_cpsw_common *common) } static int am65_cpsw_dl_switch_mode_get(struct devlink *dl, u32 id, - struct devlink_param_gset_ctx *ctx) + struct devlink_param_gset_ctx *ctx, + struct netlink_ext_ack *extack) { struct am65_cpsw_devlink *dl_priv = devlink_priv(dl); struct am65_cpsw_common *common = dl_priv->common; diff --git a/drivers/net/ethernet/ti/cpsw_new.c b/drivers/net/ethernet/ti/cpsw_new.c index 8b9e2078c602..ab88d4c02cbd 100644 --- a/drivers/net/ethernet/ti/cpsw_new.c +++ b/drivers/net/ethernet/ti/cpsw_new.c @@ -1618,7 +1618,8 @@ static const struct devlink_ops cpsw_devlink_ops = { }; static int cpsw_dl_switch_mode_get(struct devlink *dl, u32 id, - struct devlink_param_gset_ctx *ctx) + struct devlink_param_gset_ctx *ctx, + struct netlink_ext_ack *extack) { struct cpsw_devlink *dl_priv = devlink_priv(dl); struct cpsw_common *cpsw = dl_priv->cpsw; @@ -1753,7 +1754,8 @@ exit: } static int cpsw_dl_ale_ctrl_get(struct devlink *dl, u32 id, - struct devlink_param_gset_ctx *ctx) + struct devlink_param_gset_ctx *ctx, + struct netlink_ext_ack *extack) { struct cpsw_devlink *dl_priv = devlink_priv(dl); struct cpsw_common *cpsw = dl_priv->cpsw; diff --git a/drivers/net/wwan/iosm/iosm_ipc_devlink.c b/drivers/net/wwan/iosm/iosm_ipc_devlink.c index 33d6342124bc..301a9d294d30 100644 --- a/drivers/net/wwan/iosm/iosm_ipc_devlink.c +++ b/drivers/net/wwan/iosm/iosm_ipc_devlink.c @@ -21,7 +21,8 @@ static struct iosm_coredump_file_info list[IOSM_NOF_CD_REGION] = { /* Get the param values for the specific param ID's */ static int ipc_devlink_get_param(struct devlink *dl, u32 id, - struct devlink_param_gset_ctx *ctx) + struct devlink_param_gset_ctx *ctx, + struct netlink_ext_ack *extack) { struct iosm_devlink *ipc_devlink = devlink_priv(dl); diff --git a/include/net/devlink.h b/include/net/devlink.h index d01046ef0577..5f479227144d 100644 --- a/include/net/devlink.h +++ b/include/net/devlink.h @@ -490,7 +490,8 @@ struct devlink_param { enum devlink_param_type type; unsigned long supported_cmodes; int (*get)(struct devlink *devlink, u32 id, - struct devlink_param_gset_ctx *ctx); + struct devlink_param_gset_ctx *ctx, + struct netlink_ext_ack *extack); int (*set)(struct devlink *devlink, u32 id, struct devlink_param_gset_ctx *ctx, struct netlink_ext_ack *extack); diff --git a/include/net/dsa.h b/include/net/dsa.h index 97d5f401cfcf..e40cdc12f7f3 100644 --- a/include/net/dsa.h +++ b/include/net/dsa.h @@ -1251,7 +1251,8 @@ struct dsa_switch_ops { dsa_devlink_param_get, dsa_devlink_param_set, NULL) int dsa_devlink_param_get(struct devlink *dl, u32 id, - struct devlink_param_gset_ctx *ctx); + struct devlink_param_gset_ctx *ctx, + struct netlink_ext_ack *extack); int dsa_devlink_param_set(struct devlink *dl, u32 id, struct devlink_param_gset_ctx *ctx, struct netlink_ext_ack *extack); diff --git a/net/devlink/param.c b/net/devlink/param.c index 6b233b13b69a..3dbd023e4c36 100644 --- a/net/devlink/param.c +++ b/net/devlink/param.c @@ -174,11 +174,12 @@ devlink_param_cmode_is_supported(const struct devlink_param *param, static int devlink_param_get(struct devlink *devlink, const struct devlink_param *param, - struct devlink_param_gset_ctx *ctx) + struct devlink_param_gset_ctx *ctx, + struct netlink_ext_ack *extack) { if (!param->get) return -EOPNOTSUPP; - return param->get(devlink, param->id, ctx); + return param->get(devlink, param->id, ctx, extack); } static int devlink_param_set(struct devlink *devlink, @@ -250,7 +251,8 @@ static int devlink_nl_param_fill(struct sk_buff *msg, struct devlink *devlink, unsigned int port_index, struct devlink_param_item *param_item, enum devlink_command cmd, - u32 portid, u32 seq, int flags) + u32 portid, u32 seq, int flags, + struct netlink_ext_ack *extack) { union devlink_param_value param_value[DEVLINK_PARAM_CMODE_MAX + 1]; bool param_value_set[DEVLINK_PARAM_CMODE_MAX + 1] = {}; @@ -275,7 +277,7 @@ static int devlink_nl_param_fill(struct sk_buff *msg, struct devlink *devlink, return -EOPNOTSUPP; } else { ctx.cmode = i; - err = devlink_param_get(devlink, param, &ctx); + err = devlink_param_get(devlink, param, &ctx, extack); if (err) return err; param_value[i] = ctx.val; @@ -357,7 +359,7 @@ static void devlink_param_notify(struct devlink *devlink, if (!msg) return; err = devlink_nl_param_fill(msg, devlink, port_index, param_item, cmd, - 0, 0, 0); + 0, 0, 0, NULL); if (err) { nlmsg_free(msg); return; @@ -400,7 +402,8 @@ static int devlink_nl_param_get_dump_one(struct sk_buff *msg, err = devlink_nl_param_fill(msg, devlink, 0, param_item, DEVLINK_CMD_PARAM_GET, NETLINK_CB(cb->skb).portid, - cb->nlh->nlmsg_seq, flags); + cb->nlh->nlmsg_seq, flags, + cb->extack); if (err == -EOPNOTSUPP) { err = 0; } else if (err) { @@ -509,8 +512,8 @@ int devlink_nl_param_get_doit(struct sk_buff *skb, return -ENOMEM; err = devlink_nl_param_fill(msg, devlink, 0, param_item, - DEVLINK_CMD_PARAM_GET, - info->snd_portid, info->snd_seq, 0); + DEVLINK_CMD_PARAM_GET, info->snd_portid, + info->snd_seq, 0, info->extack); if (err) { nlmsg_free(msg); return err; diff --git a/net/dsa/devlink.c b/net/dsa/devlink.c index f41f9fc2194e..ed342f345692 100644 --- a/net/dsa/devlink.c +++ b/net/dsa/devlink.c @@ -182,7 +182,8 @@ static const struct devlink_ops dsa_devlink_ops = { }; int dsa_devlink_param_get(struct devlink *dl, u32 id, - struct devlink_param_gset_ctx *ctx) + struct devlink_param_gset_ctx *ctx, + struct netlink_ext_ack *extack) { struct dsa_switch *ds = dsa_devlink_to_ds(dl); -- cgit v1.2.3 From 2a367002ed321e884276c3d7232a362ddd1bf7d6 Mon Sep 17 00:00:00 2001 From: Daniel Zahka Date: Tue, 18 Nov 2025 18:50:33 -0800 Subject: devlink: support default values for param-get and param-set Support querying and resetting to default param values. Introduce two new devlink netlink attrs: DEVLINK_ATTR_PARAM_VALUE_DEFAULT and DEVLINK_ATTR_PARAM_RESET_DEFAULT. The former is used to contain an optional parameter value inside of the param_value nested attribute. The latter is used in param-set requests from userspace to indicate that the driver should reset the param to its default value. To implement this, two new functions are added to the devlink driver api: devlink_param::get_default() and devlink_param::reset_default(). These callbacks allow drivers to implement default param actions for runtime and permanent cmodes. For driverinit params, the core latches the last value set by a driver via devl_param_driverinit_value_set(), and uses that as the default value for a param. Because default parameter values are optional, it would be impossible to discern whether or not a param of type bool has default value of false or not provided if the default value is encoded using a netlink flag type. For this reason, when a DEVLINK_PARAM_TYPE_BOOL has an associated default value, the default value is encoded using a u8 type. Signed-off-by: Daniel Zahka Link: https://patch.msgid.link/20251119025038.651131-4-daniel.zahka@gmail.com Signed-off-by: Jakub Kicinski --- Documentation/netlink/specs/devlink.yaml | 9 ++ .../networking/devlink/devlink-params.rst | 10 ++ include/net/devlink.h | 42 +++++++++ include/uapi/linux/devlink.h | 3 + net/devlink/netlink_gen.c | 5 +- net/devlink/param.c | 105 ++++++++++++++++++--- 6 files changed, 160 insertions(+), 14 deletions(-) (limited to 'include/net') diff --git a/Documentation/netlink/specs/devlink.yaml b/Documentation/netlink/specs/devlink.yaml index 426d5aa7d955..837112da6738 100644 --- a/Documentation/netlink/specs/devlink.yaml +++ b/Documentation/netlink/specs/devlink.yaml @@ -859,6 +859,14 @@ attribute-sets: name: health-reporter-burst-period type: u64 doc: Time (in msec) for recoveries before starting the grace period. + + # TODO: fill in the attributes in between + + - + name: param-reset-default + type: flag + doc: Request restoring parameter to its default value. + value: 183 - name: dl-dev-stats subset-of: devlink @@ -1793,6 +1801,7 @@ operations: - param-type # param-value-data is missing here as the type is variable - param-value-cmode + - param-reset-default - name: region-get diff --git a/Documentation/networking/devlink/devlink-params.rst b/Documentation/networking/devlink/devlink-params.rst index c0597d456641..ea17756dcda6 100644 --- a/Documentation/networking/devlink/devlink-params.rst +++ b/Documentation/networking/devlink/devlink-params.rst @@ -41,6 +41,16 @@ In order for ``driverinit`` parameters to take effect, the driver must support reloading via the ``devlink-reload`` command. This command will request a reload of the device driver. +Default parameter values +========================= + +Drivers may optionally export default values for parameters of cmode +``runtime`` and ``permanent``. For ``driverinit`` parameters, the last +value set by the driver will be used as the default value. Drivers can +also support resetting params with cmode ``runtime`` and ``permanent`` +to their default values. Resetting ``driverinit`` params is supported +by devlink core without additional driver support needed. + .. _devlink_params_generic: Generic configuration parameters diff --git a/include/net/devlink.h b/include/net/devlink.h index 5f479227144d..cb839e0435a1 100644 --- a/include/net/devlink.h +++ b/include/net/devlink.h @@ -479,6 +479,10 @@ struct devlink_flash_notify { * @set: set parameter value, used for runtime and permanent * configuration modes * @validate: validate input value is applicable (within value range, etc.) + * @get_default: get parameter default value, used for runtime and permanent + * configuration modes + * @reset_default: reset parameter to default value, used for runtime and permanent + * configuration modes * * This struct should be used by the driver to fill the data for * a parameter it registers. @@ -498,6 +502,12 @@ struct devlink_param { int (*validate)(struct devlink *devlink, u32 id, union devlink_param_value val, struct netlink_ext_ack *extack); + int (*get_default)(struct devlink *devlink, u32 id, + struct devlink_param_gset_ctx *ctx, + struct netlink_ext_ack *extack); + int (*reset_default)(struct devlink *devlink, u32 id, + enum devlink_param_cmode cmode, + struct netlink_ext_ack *extack); }; struct devlink_param_item { @@ -509,6 +519,7 @@ struct devlink_param_item { * until reload. */ bool driverinit_value_new_valid; + union devlink_param_value driverinit_default; }; enum devlink_param_generic_id { @@ -630,6 +641,37 @@ enum devlink_param_generic_id { .validate = _validate, \ } +#define DEVLINK_PARAM_GENERIC_WITH_DEFAULTS(_id, _cmodes, _get, _set, \ + _validate, _get_default, \ + _reset_default) \ +{ \ + .id = DEVLINK_PARAM_GENERIC_ID_##_id, \ + .name = DEVLINK_PARAM_GENERIC_##_id##_NAME, \ + .type = DEVLINK_PARAM_GENERIC_##_id##_TYPE, \ + .generic = true, \ + .supported_cmodes = _cmodes, \ + .get = _get, \ + .set = _set, \ + .validate = _validate, \ + .get_default = _get_default, \ + .reset_default = _reset_default, \ +} + +#define DEVLINK_PARAM_DRIVER_WITH_DEFAULTS(_id, _name, _type, _cmodes, \ + _get, _set, _validate, \ + _get_default, _reset_default) \ +{ \ + .id = _id, \ + .name = _name, \ + .type = _type, \ + .supported_cmodes = _cmodes, \ + .get = _get, \ + .set = _set, \ + .validate = _validate, \ + .get_default = _get_default, \ + .reset_default = _reset_default, \ +} + /* Identifier of board design */ #define DEVLINK_INFO_VERSION_GENERIC_BOARD_ID "board.id" /* Revision of board design */ diff --git a/include/uapi/linux/devlink.h b/include/uapi/linux/devlink.h index 157f11d3fb72..e7d6b6d13470 100644 --- a/include/uapi/linux/devlink.h +++ b/include/uapi/linux/devlink.h @@ -639,6 +639,9 @@ enum devlink_attr { DEVLINK_ATTR_HEALTH_REPORTER_BURST_PERIOD, /* u64 */ + DEVLINK_ATTR_PARAM_VALUE_DEFAULT, /* dynamic */ + DEVLINK_ATTR_PARAM_RESET_DEFAULT, /* flag */ + /* Add new attributes above here, update the spec in * Documentation/netlink/specs/devlink.yaml and re-generate * net/devlink/netlink_gen.c. diff --git a/net/devlink/netlink_gen.c b/net/devlink/netlink_gen.c index 5ad435aee29d..580985025f49 100644 --- a/net/devlink/netlink_gen.c +++ b/net/devlink/netlink_gen.c @@ -301,12 +301,13 @@ static const struct nla_policy devlink_param_get_dump_nl_policy[DEVLINK_ATTR_DEV }; /* DEVLINK_CMD_PARAM_SET - do */ -static const struct nla_policy devlink_param_set_nl_policy[DEVLINK_ATTR_PARAM_VALUE_CMODE + 1] = { +static const struct nla_policy devlink_param_set_nl_policy[DEVLINK_ATTR_PARAM_RESET_DEFAULT + 1] = { [DEVLINK_ATTR_BUS_NAME] = { .type = NLA_NUL_STRING, }, [DEVLINK_ATTR_DEV_NAME] = { .type = NLA_NUL_STRING, }, [DEVLINK_ATTR_PARAM_NAME] = { .type = NLA_NUL_STRING, }, [DEVLINK_ATTR_PARAM_TYPE] = NLA_POLICY_VALIDATE_FN(NLA_U8, &devlink_attr_param_type_validate), [DEVLINK_ATTR_PARAM_VALUE_CMODE] = NLA_POLICY_MAX(NLA_U8, 2), + [DEVLINK_ATTR_PARAM_RESET_DEFAULT] = { .type = NLA_FLAG, }, }; /* DEVLINK_CMD_REGION_GET - do */ @@ -919,7 +920,7 @@ const struct genl_split_ops devlink_nl_ops[74] = { .doit = devlink_nl_param_set_doit, .post_doit = devlink_nl_post_doit, .policy = devlink_param_set_nl_policy, - .maxattr = DEVLINK_ATTR_PARAM_VALUE_CMODE, + .maxattr = DEVLINK_ATTR_PARAM_RESET_DEFAULT, .flags = GENL_ADMIN_PERM | GENL_CMD_CAP_DO, }, { diff --git a/net/devlink/param.c b/net/devlink/param.c index 3aa14ef345f0..e0ea93eded43 100644 --- a/net/devlink/param.c +++ b/net/devlink/param.c @@ -192,9 +192,32 @@ static int devlink_param_set(struct devlink *devlink, return param->set(devlink, param->id, ctx, extack); } +static int devlink_param_get_default(struct devlink *devlink, + const struct devlink_param *param, + struct devlink_param_gset_ctx *ctx, + struct netlink_ext_ack *extack) +{ + if (!param->get_default) + return -EOPNOTSUPP; + + return param->get_default(devlink, param->id, ctx, extack); +} + +static int devlink_param_reset_default(struct devlink *devlink, + const struct devlink_param *param, + enum devlink_param_cmode cmode, + struct netlink_ext_ack *extack) +{ + if (!param->reset_default) + return -EOPNOTSUPP; + + return param->reset_default(devlink, param->id, cmode, extack); +} + static int devlink_nl_param_value_put(struct sk_buff *msg, enum devlink_param_type type, - int nla_type, union devlink_param_value val) + int nla_type, union devlink_param_value val, + bool flag_as_u8) { switch (type) { case DEVLINK_PARAM_TYPE_U8: @@ -218,8 +241,16 @@ devlink_nl_param_value_put(struct sk_buff *msg, enum devlink_param_type type, return -EMSGSIZE; break; case DEVLINK_PARAM_TYPE_BOOL: - if (val.vbool && nla_put_flag(msg, nla_type)) - return -EMSGSIZE; + /* default values of type bool are encoded with u8, so that + * false can be distinguished from not present + */ + if (flag_as_u8) { + if (nla_put_u8(msg, nla_type, val.vbool)) + return -EMSGSIZE; + } else { + if (val.vbool && nla_put_flag(msg, nla_type)) + return -EMSGSIZE; + } break; } return 0; @@ -229,7 +260,9 @@ static int devlink_nl_param_value_fill_one(struct sk_buff *msg, enum devlink_param_type type, enum devlink_param_cmode cmode, - union devlink_param_value val) + union devlink_param_value val, + union devlink_param_value default_val, + bool has_default) { struct nlattr *param_value_attr; int err = -EMSGSIZE; @@ -243,10 +276,19 @@ devlink_nl_param_value_fill_one(struct sk_buff *msg, goto value_nest_cancel; err = devlink_nl_param_value_put(msg, type, - DEVLINK_ATTR_PARAM_VALUE_DATA, val); + DEVLINK_ATTR_PARAM_VALUE_DATA, + val, false); if (err) goto value_nest_cancel; + if (has_default) { + err = devlink_nl_param_value_put(msg, type, + DEVLINK_ATTR_PARAM_VALUE_DEFAULT, + default_val, true); + if (err) + goto value_nest_cancel; + } + nla_nest_end(msg, param_value_attr); return 0; @@ -262,7 +304,9 @@ static int devlink_nl_param_fill(struct sk_buff *msg, struct devlink *devlink, u32 portid, u32 seq, int flags, struct netlink_ext_ack *extack) { + union devlink_param_value default_value[DEVLINK_PARAM_CMODE_MAX + 1]; union devlink_param_value param_value[DEVLINK_PARAM_CMODE_MAX + 1]; + bool default_value_set[DEVLINK_PARAM_CMODE_MAX + 1] = {}; bool param_value_set[DEVLINK_PARAM_CMODE_MAX + 1] = {}; const struct devlink_param *param = param_item->param; struct devlink_param_gset_ctx ctx; @@ -283,12 +327,26 @@ static int devlink_nl_param_fill(struct sk_buff *msg, struct devlink *devlink, param_value[i] = param_item->driverinit_value; else return -EOPNOTSUPP; + + if (param_item->driverinit_value_valid) { + default_value[i] = param_item->driverinit_default; + default_value_set[i] = true; + } } else { ctx.cmode = i; err = devlink_param_get(devlink, param, &ctx, extack); if (err) return err; param_value[i] = ctx.val; + + err = devlink_param_get_default(devlink, param, &ctx, + extack); + if (!err) { + default_value[i] = ctx.val; + default_value_set[i] = true; + } else if (err != -EOPNOTSUPP) { + return err; + } } param_value_set[i] = true; } @@ -325,7 +383,9 @@ static int devlink_nl_param_fill(struct sk_buff *msg, struct devlink *devlink, if (!param_value_set[i]) continue; err = devlink_nl_param_value_fill_one(msg, param->type, - i, param_value[i]); + i, param_value[i], + default_value[i], + default_value_set[i]); if (err) goto values_list_nest_cancel; } @@ -542,6 +602,7 @@ static int __devlink_nl_cmd_param_set_doit(struct devlink *devlink, struct devlink_param_item *param_item; const struct devlink_param *param; union devlink_param_value value; + bool reset_default; int err = 0; param_item = devlink_param_get_from_info(params, info); @@ -553,13 +614,18 @@ static int __devlink_nl_cmd_param_set_doit(struct devlink *devlink, return err; if (param_type != param->type) return -EINVAL; - err = devlink_param_value_get_from_info(param, info, &value); - if (err) - return err; - if (param->validate) { - err = param->validate(devlink, param->id, value, info->extack); + + reset_default = info->attrs[DEVLINK_ATTR_PARAM_RESET_DEFAULT]; + if (!reset_default) { + err = devlink_param_value_get_from_info(param, info, &value); if (err) return err; + if (param->validate) { + err = param->validate(devlink, param->id, value, + info->extack); + if (err) + return err; + } } if (GENL_REQ_ATTR_CHECK(info, DEVLINK_ATTR_PARAM_VALUE_CMODE)) @@ -569,6 +635,15 @@ static int __devlink_nl_cmd_param_set_doit(struct devlink *devlink, return -EOPNOTSUPP; if (cmode == DEVLINK_PARAM_CMODE_DRIVERINIT) { + if (reset_default) { + if (!param_item->driverinit_value_valid) { + NL_SET_ERR_MSG(info->extack, + "Default value not available"); + return -EOPNOTSUPP; + } + value = param_item->driverinit_default; + } + param_item->driverinit_value_new = value; param_item->driverinit_value_new_valid = true; } else { @@ -576,7 +651,12 @@ static int __devlink_nl_cmd_param_set_doit(struct devlink *devlink, return -EOPNOTSUPP; ctx.val = value; ctx.cmode = cmode; - err = devlink_param_set(devlink, param, &ctx, info->extack); + if (reset_default) + err = devlink_param_reset_default(devlink, param, cmode, + info->extack); + else + err = devlink_param_set(devlink, param, &ctx, + info->extack); if (err) return err; } @@ -824,6 +904,7 @@ void devl_param_driverinit_value_set(struct devlink *devlink, u32 param_id, param_item->driverinit_value = init_val; param_item->driverinit_value_valid = true; + param_item->driverinit_default = init_val; devlink_param_notify(devlink, 0, param_item, DEVLINK_CMD_PARAM_NEW); } -- cgit v1.2.3 From 4fe5a00ec70717a7f1002d8913ec6143582b3c8e Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Fri, 21 Nov 2025 15:41:00 +0000 Subject: net: sched: fix TCF_LAYER_TRANSPORT handling in tcf_get_base_ptr() syzbot reported that tcf_get_base_ptr() can be called while transport header is not set [1]. Instead of returning a dangling pointer, return NULL. Fix tcf_get_base_ptr() callers to handle this NULL value. [1] WARNING: CPU: 1 PID: 6019 at ./include/linux/skbuff.h:3071 skb_transport_header include/linux/skbuff.h:3071 [inline] WARNING: CPU: 1 PID: 6019 at ./include/linux/skbuff.h:3071 tcf_get_base_ptr include/net/pkt_cls.h:539 [inline] WARNING: CPU: 1 PID: 6019 at ./include/linux/skbuff.h:3071 em_nbyte_match+0x2d8/0x3f0 net/sched/em_nbyte.c:43 Modules linked in: CPU: 1 UID: 0 PID: 6019 Comm: syz.0.17 Not tainted syzkaller #0 PREEMPT(full) Call Trace: tcf_em_match net/sched/ematch.c:494 [inline] __tcf_em_tree_match+0x1ac/0x770 net/sched/ematch.c:520 tcf_em_tree_match include/net/pkt_cls.h:512 [inline] basic_classify+0x115/0x2d0 net/sched/cls_basic.c:50 tc_classify include/net/tc_wrapper.h:197 [inline] __tcf_classify net/sched/cls_api.c:1764 [inline] tcf_classify+0x4cf/0x1140 net/sched/cls_api.c:1860 multiq_classify net/sched/sch_multiq.c:39 [inline] multiq_enqueue+0xfd/0x4c0 net/sched/sch_multiq.c:66 dev_qdisc_enqueue+0x4e/0x260 net/core/dev.c:4118 __dev_xmit_skb net/core/dev.c:4214 [inline] __dev_queue_xmit+0xe83/0x3b50 net/core/dev.c:4729 packet_snd net/packet/af_packet.c:3076 [inline] packet_sendmsg+0x3e33/0x5080 net/packet/af_packet.c:3108 sock_sendmsg_nosec net/socket.c:727 [inline] __sock_sendmsg+0x21c/0x270 net/socket.c:742 ____sys_sendmsg+0x505/0x830 net/socket.c:2630 Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2") Reported-by: syzbot+f3a497f02c389d86ef16@syzkaller.appspotmail.com Closes: https://lore.kernel.org/netdev/6920855a.a70a0220.2ea503.0058.GAE@google.com/T/#u Signed-off-by: Eric Dumazet Reviewed-by: Jamal Hadi Salim Link: https://patch.msgid.link/20251121154100.1616228-1-edumazet@google.com Signed-off-by: Jakub Kicinski --- include/net/pkt_cls.h | 2 ++ net/sched/em_cmp.c | 5 ++++- net/sched/em_nbyte.c | 2 ++ net/sched/em_text.c | 11 +++++++++-- 4 files changed, 17 insertions(+), 3 deletions(-) (limited to 'include/net') diff --git a/include/net/pkt_cls.h b/include/net/pkt_cls.h index c64fd896b1f9..99ac747b7906 100644 --- a/include/net/pkt_cls.h +++ b/include/net/pkt_cls.h @@ -536,6 +536,8 @@ static inline unsigned char * tcf_get_base_ptr(struct sk_buff *skb, int layer) case TCF_LAYER_NETWORK: return skb_network_header(skb); case TCF_LAYER_TRANSPORT: + if (!skb_transport_header_was_set(skb)) + break; return skb_transport_header(skb); } diff --git a/net/sched/em_cmp.c b/net/sched/em_cmp.c index 64b637f18bc7..48c1bce74f49 100644 --- a/net/sched/em_cmp.c +++ b/net/sched/em_cmp.c @@ -22,9 +22,12 @@ static int em_cmp_match(struct sk_buff *skb, struct tcf_ematch *em, struct tcf_pkt_info *info) { struct tcf_em_cmp *cmp = (struct tcf_em_cmp *) em->data; - unsigned char *ptr = tcf_get_base_ptr(skb, cmp->layer) + cmp->off; + unsigned char *ptr = tcf_get_base_ptr(skb, cmp->layer); u32 val = 0; + if (!ptr) + return 0; + ptr += cmp->off; if (!tcf_valid_offset(skb, ptr, cmp->align)) return 0; diff --git a/net/sched/em_nbyte.c b/net/sched/em_nbyte.c index 4f9f21a05d5e..c65ffa5fff94 100644 --- a/net/sched/em_nbyte.c +++ b/net/sched/em_nbyte.c @@ -42,6 +42,8 @@ static int em_nbyte_match(struct sk_buff *skb, struct tcf_ematch *em, struct nbyte_data *nbyte = (struct nbyte_data *) em->data; unsigned char *ptr = tcf_get_base_ptr(skb, nbyte->hdr.layer); + if (!ptr) + return 0; ptr += nbyte->hdr.off; if (!tcf_valid_offset(skb, ptr, nbyte->hdr.len)) diff --git a/net/sched/em_text.c b/net/sched/em_text.c index 6b3d0af72c39..692e2be1793e 100644 --- a/net/sched/em_text.c +++ b/net/sched/em_text.c @@ -29,12 +29,19 @@ static int em_text_match(struct sk_buff *skb, struct tcf_ematch *m, struct tcf_pkt_info *info) { struct text_match *tm = EM_TEXT_PRIV(m); + unsigned char *ptr; int from, to; - from = tcf_get_base_ptr(skb, tm->from_layer) - skb->data; + ptr = tcf_get_base_ptr(skb, tm->from_layer); + if (!ptr) + return 0; + from = ptr - skb->data; from += tm->from_offset; - to = tcf_get_base_ptr(skb, tm->to_layer) - skb->data; + ptr = tcf_get_base_ptr(skb, tm->to_layer); + if (!ptr) + return 0; + to = ptr - skb->data; to += tm->to_offset; return skb_find_text(skb, from, to, tm->config) != UINT_MAX; -- cgit v1.2.3 From 075b19c211dfeea5f27075293ddf8795b78c9bd9 Mon Sep 17 00:00:00 2001 From: Paolo Abeni Date: Fri, 21 Nov 2025 18:02:00 +0100 Subject: net: factor-out _sk_charge() helper Move out of __inet_accept() the code dealing charging newly accepted socket to memcg. MPTCP will soon use it to on a per subflow basis, in different contexts. No functional changes intended. Signed-off-by: Paolo Abeni Acked-by: Geliang Tang Acked-by: Matthieu Baerts (NGI0) Signed-off-by: Matthieu Baerts (NGI0) Link: https://patch.msgid.link/20251121-net-next-mptcp-memcg-backlog-imp-v1-1-1f34b6c1e0b1@kernel.org Signed-off-by: Jakub Kicinski --- include/net/sock.h | 2 ++ net/core/sock.c | 18 ++++++++++++++++++ net/ipv4/af_inet.c | 17 +---------------- 3 files changed, 21 insertions(+), 16 deletions(-) (limited to 'include/net') diff --git a/include/net/sock.h b/include/net/sock.h index a5f36ea9d46f..38d48cfe0741 100644 --- a/include/net/sock.h +++ b/include/net/sock.h @@ -1631,6 +1631,8 @@ static inline void sk_mem_uncharge(struct sock *sk, int size) sk_mem_reclaim(sk); } +void __sk_charge(struct sock *sk, gfp_t gfp); + #if IS_ENABLED(CONFIG_PROVE_LOCKING) && IS_ENABLED(CONFIG_MODULES) static inline void sk_owner_set(struct sock *sk, struct module *owner) { diff --git a/net/core/sock.c b/net/core/sock.c index 3b74fc71f51c..b26a6cdc9bcd 100644 --- a/net/core/sock.c +++ b/net/core/sock.c @@ -3448,6 +3448,24 @@ void __sk_mem_reclaim(struct sock *sk, int amount) } EXPORT_SYMBOL(__sk_mem_reclaim); +void __sk_charge(struct sock *sk, gfp_t gfp) +{ + int amt; + + gfp |= __GFP_NOFAIL; + if (mem_cgroup_from_sk(sk)) { + /* The socket has not been accepted yet, no need + * to look at newsk->sk_wmem_queued. + */ + amt = sk_mem_pages(sk->sk_forward_alloc + + atomic_read(&sk->sk_rmem_alloc)); + if (amt) + mem_cgroup_sk_charge(sk, amt, gfp); + } + + kmem_cache_charge(sk, gfp); +} + int sk_set_peek_off(struct sock *sk, int val) { WRITE_ONCE(sk->sk_peek_off, val); diff --git a/net/ipv4/af_inet.c b/net/ipv4/af_inet.c index a31b94ce8968..08d811f11896 100644 --- a/net/ipv4/af_inet.c +++ b/net/ipv4/af_inet.c @@ -756,23 +756,8 @@ EXPORT_SYMBOL(inet_stream_connect); void __inet_accept(struct socket *sock, struct socket *newsock, struct sock *newsk) { if (mem_cgroup_sockets_enabled) { - gfp_t gfp = GFP_KERNEL | __GFP_NOFAIL; - mem_cgroup_sk_alloc(newsk); - - if (mem_cgroup_from_sk(newsk)) { - int amt; - - /* The socket has not been accepted yet, no need - * to look at newsk->sk_wmem_queued. - */ - amt = sk_mem_pages(newsk->sk_forward_alloc + - atomic_read(&newsk->sk_rmem_alloc)); - if (amt) - mem_cgroup_sk_charge(newsk, amt, gfp); - } - - kmem_cache_charge(newsk, gfp); + __sk_charge(newsk, GFP_KERNEL); } sock_rps_record_flow(newsk); -- cgit v1.2.3 From cba1ba11c1bae87de9c2e13d342bfbd6a3c1cf63 Mon Sep 17 00:00:00 2001 From: Lachlan Hodges Date: Tue, 25 Nov 2025 13:59:26 +1100 Subject: wifi: cfg80211: include s1g_primary_2mhz when comparing chandefs When comparing chandefs, ensure we include s1g_primary_2mhz. Signed-off-by: Lachlan Hodges Link: https://patch.msgid.link/20251125025927.245280-3-lachlan.hodges@morsemicro.com Signed-off-by: Johannes Berg --- include/net/cfg80211.h | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) (limited to 'include/net') diff --git a/include/net/cfg80211.h b/include/net/cfg80211.h index 3d3ed1932262..899f267b7cf9 100644 --- a/include/net/cfg80211.h +++ b/include/net/cfg80211.h @@ -974,7 +974,8 @@ cfg80211_chandef_identical(const struct cfg80211_chan_def *chandef1, chandef1->center_freq1 == chandef2->center_freq1 && chandef1->freq1_offset == chandef2->freq1_offset && chandef1->center_freq2 == chandef2->center_freq2 && - chandef1->punctured == chandef2->punctured); + chandef1->punctured == chandef2->punctured && + chandef1->s1g_primary_2mhz == chandef2->s1g_primary_2mhz); } /** -- cgit v1.2.3 From b2a38f6df9dab0b05858746edcbe2403f8f4e4ec Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Fri, 21 Nov 2025 08:32:43 +0000 Subject: net_sched: make room for (struct qdisc_skb_cb)->pkt_segs Add a new u16 field, next to pkt_len : pkt_segs This will cache shinfo->gso_segs to speed up qdisc deqeue(). Move slave_dev_queue_mapping at the end of qdisc_skb_cb, and move three bits from tc_skb_cb : - post_ct - post_ct_snat - post_ct_dnat Signed-off-by: Eric Dumazet Link: https://patch.msgid.link/20251121083256.674562-2-edumazet@google.com Signed-off-by: Paolo Abeni --- include/net/sch_generic.h | 18 +++++++++--------- net/core/dev.c | 2 +- net/sched/act_ct.c | 8 ++++---- net/sched/cls_api.c | 6 +++--- net/sched/cls_flower.c | 2 +- 5 files changed, 18 insertions(+), 18 deletions(-) (limited to 'include/net') diff --git a/include/net/sch_generic.h b/include/net/sch_generic.h index 94966692ccdf..9cd8b5d4b236 100644 --- a/include/net/sch_generic.h +++ b/include/net/sch_generic.h @@ -429,13 +429,16 @@ struct tcf_proto { }; struct qdisc_skb_cb { - struct { - unsigned int pkt_len; - u16 slave_dev_queue_mapping; - u16 tc_classid; - }; + unsigned int pkt_len; + u16 pkt_segs; + u16 tc_classid; #define QDISC_CB_PRIV_LEN 20 unsigned char data[QDISC_CB_PRIV_LEN]; + + u16 slave_dev_queue_mapping; + u8 post_ct:1; + u8 post_ct_snat:1; + u8 post_ct_dnat:1; }; typedef void tcf_chain_head_change_t(struct tcf_proto *tp_head, void *priv); @@ -1064,11 +1067,8 @@ struct tc_skb_cb { struct qdisc_skb_cb qdisc_cb; u32 drop_reason; - u16 zone; /* Only valid if post_ct = true */ + u16 zone; /* Only valid if qdisc_skb_cb(skb)->post_ct = true */ u16 mru; - u8 post_ct:1; - u8 post_ct_snat:1; - u8 post_ct_dnat:1; }; static inline struct tc_skb_cb *tc_skb_cb(const struct sk_buff *skb) diff --git a/net/core/dev.c b/net/core/dev.c index 69515edd17bc..46ce6c610780 100644 --- a/net/core/dev.c +++ b/net/core/dev.c @@ -4355,7 +4355,7 @@ static int tc_run(struct tcx_entry *entry, struct sk_buff *skb, return ret; tc_skb_cb(skb)->mru = 0; - tc_skb_cb(skb)->post_ct = false; + qdisc_skb_cb(skb)->post_ct = false; tcf_set_drop_reason(skb, *drop_reason); mini_qdisc_bstats_cpu_update(miniq, skb); diff --git a/net/sched/act_ct.c b/net/sched/act_ct.c index 6749a4a9a9cd..2b6ac7069dc1 100644 --- a/net/sched/act_ct.c +++ b/net/sched/act_ct.c @@ -948,9 +948,9 @@ static int tcf_ct_act_nat(struct sk_buff *skb, return err & NF_VERDICT_MASK; if (action & BIT(NF_NAT_MANIP_SRC)) - tc_skb_cb(skb)->post_ct_snat = 1; + qdisc_skb_cb(skb)->post_ct_snat = 1; if (action & BIT(NF_NAT_MANIP_DST)) - tc_skb_cb(skb)->post_ct_dnat = 1; + qdisc_skb_cb(skb)->post_ct_dnat = 1; return err; #else @@ -986,7 +986,7 @@ TC_INDIRECT_SCOPE int tcf_ct_act(struct sk_buff *skb, const struct tc_action *a, tcf_action_update_bstats(&c->common, skb); if (clear) { - tc_skb_cb(skb)->post_ct = false; + qdisc_skb_cb(skb)->post_ct = false; ct = nf_ct_get(skb, &ctinfo); if (ct) { nf_ct_put(ct); @@ -1097,7 +1097,7 @@ do_nat: out_push: skb_push_rcsum(skb, nh_ofs); - tc_skb_cb(skb)->post_ct = true; + qdisc_skb_cb(skb)->post_ct = true; tc_skb_cb(skb)->zone = p->zone; out_clear: if (defrag) diff --git a/net/sched/cls_api.c b/net/sched/cls_api.c index f751cd5eeac8..ebca4b926dcf 100644 --- a/net/sched/cls_api.c +++ b/net/sched/cls_api.c @@ -1872,9 +1872,9 @@ int tcf_classify(struct sk_buff *skb, } ext->chain = last_executed_chain; ext->mru = cb->mru; - ext->post_ct = cb->post_ct; - ext->post_ct_snat = cb->post_ct_snat; - ext->post_ct_dnat = cb->post_ct_dnat; + ext->post_ct = qdisc_skb_cb(skb)->post_ct; + ext->post_ct_snat = qdisc_skb_cb(skb)->post_ct_snat; + ext->post_ct_dnat = qdisc_skb_cb(skb)->post_ct_dnat; ext->zone = cb->zone; } } diff --git a/net/sched/cls_flower.c b/net/sched/cls_flower.c index 099ff6a3e1f5..7669371c1354 100644 --- a/net/sched/cls_flower.c +++ b/net/sched/cls_flower.c @@ -326,7 +326,7 @@ TC_INDIRECT_SCOPE int fl_classify(struct sk_buff *skb, struct tcf_result *res) { struct cls_fl_head *head = rcu_dereference_bh(tp->root); - bool post_ct = tc_skb_cb(skb)->post_ct; + bool post_ct = qdisc_skb_cb(skb)->post_ct; u16 zone = tc_skb_cb(skb)->zone; struct fl_flow_key skb_key; struct fl_flow_mask *mask; -- cgit v1.2.3 From 2773cb0b3120eb5c4b66d949eb99853d5bae1221 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Fri, 21 Nov 2025 08:32:47 +0000 Subject: net_sched: use qdisc_skb_cb(skb)->pkt_segs in bstats_update() Avoid up to two cache line misses in qdisc dequeue() to fetch skb_shinfo(skb)->gso_segs/gso_size while qdisc spinlock is held. This gives a 5 % improvement in a TX intensive workload. Signed-off-by: Eric Dumazet Link: https://patch.msgid.link/20251121083256.674562-6-edumazet@google.com Signed-off-by: Paolo Abeni --- include/net/sch_generic.h | 13 ++++++++++--- net/sched/sch_cake.c | 1 + net/sched/sch_dualpi2.c | 1 + net/sched/sch_netem.c | 1 + net/sched/sch_qfq.c | 2 +- net/sched/sch_taprio.c | 1 + net/sched/sch_tbf.c | 1 + 7 files changed, 16 insertions(+), 4 deletions(-) (limited to 'include/net') diff --git a/include/net/sch_generic.h b/include/net/sch_generic.h index 9cd8b5d4b236..cdf7a58ebcf5 100644 --- a/include/net/sch_generic.h +++ b/include/net/sch_generic.h @@ -829,6 +829,15 @@ static inline unsigned int qdisc_pkt_len(const struct sk_buff *skb) return qdisc_skb_cb(skb)->pkt_len; } +static inline unsigned int qdisc_pkt_segs(const struct sk_buff *skb) +{ + u32 pkt_segs = qdisc_skb_cb(skb)->pkt_segs; + + DEBUG_NET_WARN_ON_ONCE(pkt_segs != + (skb_is_gso(skb) ? skb_shinfo(skb)->gso_segs : 1)); + return pkt_segs; +} + /* additional qdisc xmit flags (NET_XMIT_MASK in linux/netdevice.h) */ enum net_xmit_qdisc_t { __NET_XMIT_STOLEN = 0x00010000, @@ -870,9 +879,7 @@ static inline void _bstats_update(struct gnet_stats_basic_sync *bstats, static inline void bstats_update(struct gnet_stats_basic_sync *bstats, const struct sk_buff *skb) { - _bstats_update(bstats, - qdisc_pkt_len(skb), - skb_is_gso(skb) ? skb_shinfo(skb)->gso_segs : 1); + _bstats_update(bstats, qdisc_pkt_len(skb), qdisc_pkt_segs(skb)); } static inline void qdisc_bstats_cpu_update(struct Qdisc *sch, diff --git a/net/sched/sch_cake.c b/net/sched/sch_cake.c index 9213129f0de1..a20880034aa5 100644 --- a/net/sched/sch_cake.c +++ b/net/sched/sch_cake.c @@ -1800,6 +1800,7 @@ static s32 cake_enqueue(struct sk_buff *skb, struct Qdisc *sch, skb_list_walk_safe(segs, segs, nskb) { skb_mark_not_on_list(segs); qdisc_skb_cb(segs)->pkt_len = segs->len; + qdisc_skb_cb(segs)->pkt_segs = 1; cobalt_set_enqueue_time(segs, now); get_cobalt_cb(segs)->adjusted_len = cake_overhead(q, segs); diff --git a/net/sched/sch_dualpi2.c b/net/sched/sch_dualpi2.c index 4b975feb52b1..6d7e6389758d 100644 --- a/net/sched/sch_dualpi2.c +++ b/net/sched/sch_dualpi2.c @@ -475,6 +475,7 @@ static int dualpi2_qdisc_enqueue(struct sk_buff *skb, struct Qdisc *sch, * (3) Enqueue fragment & set ts in dualpi2_enqueue_skb */ qdisc_skb_cb(nskb)->pkt_len = nskb->len; + qdisc_skb_cb(nskb)->pkt_segs = 1; dualpi2_skb_cb(nskb)->classified = dualpi2_skb_cb(skb)->classified; dualpi2_skb_cb(nskb)->ect = dualpi2_skb_cb(skb)->ect; diff --git a/net/sched/sch_netem.c b/net/sched/sch_netem.c index eafc316ae319..32a5f3304046 100644 --- a/net/sched/sch_netem.c +++ b/net/sched/sch_netem.c @@ -429,6 +429,7 @@ static struct sk_buff *netem_segment(struct sk_buff *skb, struct Qdisc *sch, struct sk_buff *segs; netdev_features_t features = netif_skb_features(skb); + qdisc_skb_cb(skb)->pkt_segs = 1; segs = skb_gso_segment(skb, features & ~NETIF_F_GSO_MASK); if (IS_ERR_OR_NULL(segs)) { diff --git a/net/sched/sch_qfq.c b/net/sched/sch_qfq.c index 2255355e51d3..d920f57dc6d7 100644 --- a/net/sched/sch_qfq.c +++ b/net/sched/sch_qfq.c @@ -1250,7 +1250,7 @@ static int qfq_enqueue(struct sk_buff *skb, struct Qdisc *sch, } } - gso_segs = skb_is_gso(skb) ? skb_shinfo(skb)->gso_segs : 1; + gso_segs = qdisc_pkt_segs(skb); err = qdisc_enqueue(skb, cl->qdisc, to_free); if (unlikely(err != NET_XMIT_SUCCESS)) { pr_debug("qfq_enqueue: enqueue failed %d\n", err); diff --git a/net/sched/sch_taprio.c b/net/sched/sch_taprio.c index 39b735386996..300d577b3286 100644 --- a/net/sched/sch_taprio.c +++ b/net/sched/sch_taprio.c @@ -595,6 +595,7 @@ static int taprio_enqueue_segmented(struct sk_buff *skb, struct Qdisc *sch, skb_list_walk_safe(segs, segs, nskb) { skb_mark_not_on_list(segs); qdisc_skb_cb(segs)->pkt_len = segs->len; + qdisc_skb_cb(segs)->pkt_segs = 1; slen += segs->len; /* FIXME: we should be segmenting to a smaller size diff --git a/net/sched/sch_tbf.c b/net/sched/sch_tbf.c index 4c977f049670..f2340164f579 100644 --- a/net/sched/sch_tbf.c +++ b/net/sched/sch_tbf.c @@ -221,6 +221,7 @@ static int tbf_segment(struct sk_buff *skb, struct Qdisc *sch, skb_mark_not_on_list(segs); seg_len = segs->len; qdisc_skb_cb(segs)->pkt_len = seg_len; + qdisc_skb_cb(segs)->pkt_segs = 1; ret = qdisc_enqueue(segs, q->qdisc, to_free); if (ret != NET_XMIT_SUCCESS) { if (net_xmit_drop_count(ret)) -- cgit v1.2.3 From ad50d5a3fc20327e133e2db849c6e67fc49650e6 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Fri, 21 Nov 2025 08:32:49 +0000 Subject: net_sched: add Qdisc_read_mostly and Qdisc_write groups It is possible to reorg Qdisc to avoid always dirtying 2 cache lines in fast path by reducing this to a single dirtied cache line. In current layout, we change only four/six fields in the first cache line: - q.spinlock - q.qlen - bstats.bytes - bstats.packets - some Qdisc also change q.next/q.prev In the second cache line we change in the fast path: - running - state - qstats.backlog /* --- cacheline 2 boundary (128 bytes) --- */ struct sk_buff_head gso_skb __attribute__((__aligned__(64))); /* 0x80 0x18 */ struct qdisc_skb_head q; /* 0x98 0x18 */ struct gnet_stats_basic_sync bstats __attribute__((__aligned__(16))); /* 0xb0 0x10 */ /* --- cacheline 3 boundary (192 bytes) --- */ struct gnet_stats_queue qstats; /* 0xc0 0x14 */ bool running; /* 0xd4 0x1 */ /* XXX 3 bytes hole, try to pack */ unsigned long state; /* 0xd8 0x8 */ struct Qdisc * next_sched; /* 0xe0 0x8 */ struct sk_buff_head skb_bad_txq; /* 0xe8 0x18 */ /* --- cacheline 4 boundary (256 bytes) --- */ Reorganize things to have a first cache line mostly read, then a mostly written one. This gives a ~3% increase of performance under tx stress. Note that there is an additional hole because @qstats now spans over a third cache line. /* --- cacheline 2 boundary (128 bytes) --- */ __u8 __cacheline_group_begin__Qdisc_read_mostly[0] __attribute__((__aligned__(64))); /* 0x80 0 */ struct sk_buff_head gso_skb; /* 0x80 0x18 */ struct Qdisc * next_sched; /* 0x98 0x8 */ struct sk_buff_head skb_bad_txq; /* 0xa0 0x18 */ __u8 __cacheline_group_end__Qdisc_read_mostly[0]; /* 0xb8 0 */ /* XXX 8 bytes hole, try to pack */ /* --- cacheline 3 boundary (192 bytes) --- */ __u8 __cacheline_group_begin__Qdisc_write[0] __attribute__((__aligned__(64))); /* 0xc0 0 */ struct qdisc_skb_head q; /* 0xc0 0x18 */ unsigned long state; /* 0xd8 0x8 */ struct gnet_stats_basic_sync bstats __attribute__((__aligned__(16))); /* 0xe0 0x10 */ bool running; /* 0xf0 0x1 */ /* XXX 3 bytes hole, try to pack */ struct gnet_stats_queue qstats; /* 0xf4 0x14 */ /* --- cacheline 4 boundary (256 bytes) was 8 bytes ago --- */ __u8 __cacheline_group_end__Qdisc_write[0]; /* 0x108 0 */ /* XXX 56 bytes hole, try to pack */ Signed-off-by: Eric Dumazet Link: https://patch.msgid.link/20251121083256.674562-8-edumazet@google.com Signed-off-by: Paolo Abeni --- include/net/sch_generic.h | 29 ++++++++++++++++++----------- 1 file changed, 18 insertions(+), 11 deletions(-) (limited to 'include/net') diff --git a/include/net/sch_generic.h b/include/net/sch_generic.h index cdf7a58ebcf5..79501499dafb 100644 --- a/include/net/sch_generic.h +++ b/include/net/sch_generic.h @@ -103,17 +103,24 @@ struct Qdisc { int pad; refcount_t refcnt; - /* - * For performance sake on SMP, we put highly modified fields at the end - */ - struct sk_buff_head gso_skb ____cacheline_aligned_in_smp; - struct qdisc_skb_head q; - struct gnet_stats_basic_sync bstats; - struct gnet_stats_queue qstats; - bool running; /* must be written under qdisc spinlock */ - unsigned long state; - struct Qdisc *next_sched; - struct sk_buff_head skb_bad_txq; + /* Cache line potentially dirtied in dequeue() or __netif_reschedule(). */ + __cacheline_group_begin(Qdisc_read_mostly) ____cacheline_aligned; + struct sk_buff_head gso_skb; + struct Qdisc *next_sched; + struct sk_buff_head skb_bad_txq; + __cacheline_group_end(Qdisc_read_mostly); + + /* Fields dirtied in dequeue() fast path. */ + __cacheline_group_begin(Qdisc_write) ____cacheline_aligned; + struct qdisc_skb_head q; + unsigned long state; + struct gnet_stats_basic_sync bstats; + bool running; /* must be written under qdisc spinlock */ + + /* Note : we only change qstats.backlog in fast path. */ + struct gnet_stats_queue qstats; + __cacheline_group_end(Qdisc_write); + atomic_long_t defer_count ____cacheline_aligned_in_smp; struct llist_head defer_list; -- cgit v1.2.3 From 0170d7f47c8bb0311bc802bad52245c045f151fe Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Fri, 21 Nov 2025 08:32:54 +0000 Subject: net_sched: add tcf_kfree_skb_list() helper Using kfree_skb_list_reason() to free list of skbs from qdisc operations seems wrong as each skb might have a different drop reason. Cleanup __dev_xmit_skb() to call tcf_kfree_skb_list() once in preparation of the following patch. Signed-off-by: Eric Dumazet Link: https://patch.msgid.link/20251121083256.674562-13-edumazet@google.com Signed-off-by: Paolo Abeni --- include/net/sch_generic.h | 11 +++++++++++ net/core/dev.c | 15 +++++---------- 2 files changed, 16 insertions(+), 10 deletions(-) (limited to 'include/net') diff --git a/include/net/sch_generic.h b/include/net/sch_generic.h index 79501499dafb..b8092d0378a0 100644 --- a/include/net/sch_generic.h +++ b/include/net/sch_generic.h @@ -1105,6 +1105,17 @@ static inline void tcf_set_drop_reason(const struct sk_buff *skb, tc_skb_cb(skb)->drop_reason = reason; } +static inline void tcf_kfree_skb_list(struct sk_buff *skb) +{ + while (unlikely(skb)) { + struct sk_buff *next = skb->next; + + prefetch(next); + kfree_skb_reason(skb, tcf_get_drop_reason(skb)); + skb = next; + } +} + /* Instead of calling kfree_skb() while root qdisc lock is held, * queue the skb for future freeing at end of __dev_xmit_skb() */ diff --git a/net/core/dev.c b/net/core/dev.c index 10042139dbb0..e865cdb9b696 100644 --- a/net/core/dev.c +++ b/net/core/dev.c @@ -4162,7 +4162,7 @@ static inline int __dev_xmit_skb(struct sk_buff *skb, struct Qdisc *q, __qdisc_run(q); qdisc_run_end(q); - goto no_lock_out; + goto free_skbs; } qdisc_bstats_cpu_update(q, skb); @@ -4176,12 +4176,7 @@ static inline int __dev_xmit_skb(struct sk_buff *skb, struct Qdisc *q, rc = dev_qdisc_enqueue(skb, q, &to_free, txq); qdisc_run(q); - -no_lock_out: - if (unlikely(to_free)) - kfree_skb_list_reason(to_free, - tcf_get_drop_reason(to_free)); - return rc; + goto free_skbs; } /* Open code llist_add(&skb->ll_node, &q->defer_list) + queue limit. @@ -4257,9 +4252,9 @@ no_lock_out: } unlock: spin_unlock(root_lock); - if (unlikely(to_free)) - kfree_skb_list_reason(to_free, - tcf_get_drop_reason(to_free)); + +free_skbs: + tcf_kfree_skb_list(to_free); return rc; } -- cgit v1.2.3 From 191ff13e42a7b7824fec5b2ed84fd6481356754d Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Fri, 21 Nov 2025 08:32:55 +0000 Subject: net_sched: add qdisc_dequeue_drop() helper Some qdisc like cake, codel, fq_codel might drop packets in their dequeue() method. This is currently problematic because dequeue() runs with the qdisc spinlock held. Freeing skbs can be extremely expensive. Add qdisc_dequeue_drop() method and a new TCQ_F_DEQUEUE_DROPS so that these qdiscs can opt-in to defer the skb frees after the socket spinlock is released. TCQ_F_DEQUEUE_DROPS is an attempt to not penalize other qdiscs with an extra cache line miss. Signed-off-by: Eric Dumazet Link: https://patch.msgid.link/20251121083256.674562-14-edumazet@google.com Signed-off-by: Paolo Abeni --- include/net/pkt_sched.h | 5 +++-- include/net/sch_generic.h | 30 +++++++++++++++++++++++++++--- net/core/dev.c | 22 +++++++++++++--------- 3 files changed, 43 insertions(+), 14 deletions(-) (limited to 'include/net') diff --git a/include/net/pkt_sched.h b/include/net/pkt_sched.h index 4678db45832a..e703c507d0da 100644 --- a/include/net/pkt_sched.h +++ b/include/net/pkt_sched.h @@ -114,12 +114,13 @@ bool sch_direct_xmit(struct sk_buff *skb, struct Qdisc *q, void __qdisc_run(struct Qdisc *q); -static inline void qdisc_run(struct Qdisc *q) +static inline struct sk_buff *qdisc_run(struct Qdisc *q) { if (qdisc_run_begin(q)) { __qdisc_run(q); - qdisc_run_end(q); + return qdisc_run_end(q); } + return NULL; } extern const struct nla_policy rtm_tca_policy[TCA_MAX + 1]; diff --git a/include/net/sch_generic.h b/include/net/sch_generic.h index b8092d0378a0..c3a7268b567e 100644 --- a/include/net/sch_generic.h +++ b/include/net/sch_generic.h @@ -88,6 +88,8 @@ struct Qdisc { #define TCQ_F_INVISIBLE 0x80 /* invisible by default in dump */ #define TCQ_F_NOLOCK 0x100 /* qdisc does not require locking */ #define TCQ_F_OFFLOADED 0x200 /* qdisc is offloaded to HW */ +#define TCQ_F_DEQUEUE_DROPS 0x400 /* ->dequeue() can drop packets in q->to_free */ + u32 limit; const struct Qdisc_ops *ops; struct qdisc_size_table __rcu *stab; @@ -119,6 +121,8 @@ struct Qdisc { /* Note : we only change qstats.backlog in fast path. */ struct gnet_stats_queue qstats; + + struct sk_buff *to_free; __cacheline_group_end(Qdisc_write); @@ -218,8 +222,10 @@ static inline bool qdisc_run_begin(struct Qdisc *qdisc) return true; } -static inline void qdisc_run_end(struct Qdisc *qdisc) +static inline struct sk_buff *qdisc_run_end(struct Qdisc *qdisc) { + struct sk_buff *to_free = NULL; + if (qdisc->flags & TCQ_F_NOLOCK) { spin_unlock(&qdisc->seqlock); @@ -232,9 +238,16 @@ static inline void qdisc_run_end(struct Qdisc *qdisc) if (unlikely(test_bit(__QDISC_STATE_MISSED, &qdisc->state))) __netif_schedule(qdisc); - } else { - WRITE_ONCE(qdisc->running, false); + return NULL; + } + + if (qdisc->flags & TCQ_F_DEQUEUE_DROPS) { + to_free = qdisc->to_free; + if (to_free) + qdisc->to_free = NULL; } + WRITE_ONCE(qdisc->running, false); + return to_free; } static inline bool qdisc_may_bulk(const struct Qdisc *qdisc) @@ -1116,6 +1129,17 @@ static inline void tcf_kfree_skb_list(struct sk_buff *skb) } } +static inline void qdisc_dequeue_drop(struct Qdisc *q, struct sk_buff *skb, + enum skb_drop_reason reason) +{ + DEBUG_NET_WARN_ON_ONCE(!(q->flags & TCQ_F_DEQUEUE_DROPS)); + DEBUG_NET_WARN_ON_ONCE(q->flags & TCQ_F_NOLOCK); + + tcf_set_drop_reason(skb, reason); + skb->next = q->to_free; + q->to_free = skb; +} + /* Instead of calling kfree_skb() while root qdisc lock is held, * queue the skb for future freeing at end of __dev_xmit_skb() */ diff --git a/net/core/dev.c b/net/core/dev.c index e865cdb9b696..9094c0fb8c68 100644 --- a/net/core/dev.c +++ b/net/core/dev.c @@ -4141,7 +4141,7 @@ static inline int __dev_xmit_skb(struct sk_buff *skb, struct Qdisc *q, struct net_device *dev, struct netdev_queue *txq) { - struct sk_buff *next, *to_free = NULL; + struct sk_buff *next, *to_free = NULL, *to_free2 = NULL; spinlock_t *root_lock = qdisc_lock(q); struct llist_node *ll_list, *first_n; unsigned long defer_count = 0; @@ -4160,7 +4160,7 @@ static inline int __dev_xmit_skb(struct sk_buff *skb, struct Qdisc *q, if (unlikely(!nolock_qdisc_is_empty(q))) { rc = dev_qdisc_enqueue(skb, q, &to_free, txq); __qdisc_run(q); - qdisc_run_end(q); + to_free2 = qdisc_run_end(q); goto free_skbs; } @@ -4170,12 +4170,13 @@ static inline int __dev_xmit_skb(struct sk_buff *skb, struct Qdisc *q, !nolock_qdisc_is_empty(q)) __qdisc_run(q); - qdisc_run_end(q); - return NET_XMIT_SUCCESS; + to_free2 = qdisc_run_end(q); + rc = NET_XMIT_SUCCESS; + goto free_skbs; } rc = dev_qdisc_enqueue(skb, q, &to_free, txq); - qdisc_run(q); + to_free2 = qdisc_run(q); goto free_skbs; } @@ -4234,7 +4235,7 @@ static inline int __dev_xmit_skb(struct sk_buff *skb, struct Qdisc *q, qdisc_bstats_update(q, skb); if (sch_direct_xmit(skb, q, dev, txq, root_lock, true)) __qdisc_run(q); - qdisc_run_end(q); + to_free2 = qdisc_run_end(q); rc = NET_XMIT_SUCCESS; } else { int count = 0; @@ -4246,7 +4247,7 @@ static inline int __dev_xmit_skb(struct sk_buff *skb, struct Qdisc *q, rc = dev_qdisc_enqueue(skb, q, &to_free, txq); count++; } - qdisc_run(q); + to_free2 = qdisc_run(q); if (count != 1) rc = NET_XMIT_SUCCESS; } @@ -4255,6 +4256,7 @@ unlock: free_skbs: tcf_kfree_skb_list(to_free); + tcf_kfree_skb_list(to_free2); return rc; } @@ -5747,8 +5749,9 @@ static __latent_entropy void net_tx_action(void) rcu_read_lock(); while (head) { - struct Qdisc *q = head; spinlock_t *root_lock = NULL; + struct sk_buff *to_free; + struct Qdisc *q = head; head = head->next_sched; @@ -5775,9 +5778,10 @@ static __latent_entropy void net_tx_action(void) } clear_bit(__QDISC_STATE_SCHED, &q->state); - qdisc_run(q); + to_free = qdisc_run(q); if (root_lock) spin_unlock(root_lock); + tcf_kfree_skb_list(to_free); } rcu_read_unlock(); -- cgit v1.2.3 From 3a6e8fd0bf4042c572dc52e634878b9aca02970d Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Mon, 24 Nov 2025 17:50:10 +0000 Subject: tcp: rename icsk_timeout() to tcp_timeout_expires() In preparation of sk->tcp_timeout_timer introduction, rename icsk_timeout() helper and change its argument to plain 'const struct sock *sk'. Signed-off-by: Eric Dumazet Reviewed-by: Kuniyuki Iwashima Link: https://patch.msgid.link/20251124175013.1473655-2-edumazet@google.com Signed-off-by: Jakub Kicinski --- include/net/inet_connection_sock.h | 5 ++--- net/ipv4/inet_diag.c | 4 ++-- net/ipv4/tcp_ipv4.c | 4 ++-- net/ipv4/tcp_timer.c | 6 +++--- net/ipv6/tcp_ipv6.c | 4 ++-- net/mptcp/protocol.c | 2 +- 6 files changed, 12 insertions(+), 13 deletions(-) (limited to 'include/net') diff --git a/include/net/inet_connection_sock.h b/include/net/inet_connection_sock.h index fd40af2221b9..765c2149d678 100644 --- a/include/net/inet_connection_sock.h +++ b/include/net/inet_connection_sock.h @@ -184,10 +184,9 @@ static inline void inet_csk_delack_init(struct sock *sk) memset(&inet_csk(sk)->icsk_ack, 0, sizeof(inet_csk(sk)->icsk_ack)); } -static inline unsigned long -icsk_timeout(const struct inet_connection_sock *icsk) +static inline unsigned long tcp_timeout_expires(const struct sock *sk) { - return READ_ONCE(icsk->icsk_retransmit_timer.expires); + return READ_ONCE(inet_csk(sk)->icsk_retransmit_timer.expires); } static inline unsigned long diff --git a/net/ipv4/inet_diag.c b/net/ipv4/inet_diag.c index f0b6c5a411a2..9f63c09439a0 100644 --- a/net/ipv4/inet_diag.c +++ b/net/ipv4/inet_diag.c @@ -287,12 +287,12 @@ int inet_sk_diag_fill(struct sock *sk, struct inet_connection_sock *icsk, r->idiag_timer = 1; r->idiag_retrans = READ_ONCE(icsk->icsk_retransmits); r->idiag_expires = - jiffies_delta_to_msecs(icsk_timeout(icsk) - jiffies); + jiffies_delta_to_msecs(tcp_timeout_expires(sk) - jiffies); } else if (icsk_pending == ICSK_TIME_PROBE0) { r->idiag_timer = 4; r->idiag_retrans = READ_ONCE(icsk->icsk_probes_out); r->idiag_expires = - jiffies_delta_to_msecs(icsk_timeout(icsk) - jiffies); + jiffies_delta_to_msecs(tcp_timeout_expires(sk) - jiffies); } else if (timer_pending(&sk->sk_timer)) { r->idiag_timer = 2; r->idiag_retrans = READ_ONCE(icsk->icsk_probes_out); diff --git a/net/ipv4/tcp_ipv4.c b/net/ipv4/tcp_ipv4.c index e0bb8d9e2d9c..7b8af2c8d03a 100644 --- a/net/ipv4/tcp_ipv4.c +++ b/net/ipv4/tcp_ipv4.c @@ -2869,10 +2869,10 @@ static void get_tcp4_sock(struct sock *sk, struct seq_file *f, int i) icsk_pending == ICSK_TIME_REO_TIMEOUT || icsk_pending == ICSK_TIME_LOSS_PROBE) { timer_active = 1; - timer_expires = icsk_timeout(icsk); + timer_expires = tcp_timeout_expires(sk); } else if (icsk_pending == ICSK_TIME_PROBE0) { timer_active = 4; - timer_expires = icsk_timeout(icsk); + timer_expires = tcp_timeout_expires(sk); } else if (timer_pending(&sk->sk_timer)) { timer_active = 2; timer_expires = sk->sk_timer.expires; diff --git a/net/ipv4/tcp_timer.c b/net/ipv4/tcp_timer.c index 0672c3d8f4f1..afbd901e610e 100644 --- a/net/ipv4/tcp_timer.c +++ b/net/ipv4/tcp_timer.c @@ -510,7 +510,7 @@ static bool tcp_rtx_probe0_timed_out(const struct sock *sk, * and tp->rcv_tstamp might very well have been written recently. * rcv_delta can thus be negative. */ - rcv_delta = icsk_timeout(icsk) - tp->rcv_tstamp; + rcv_delta = tcp_timeout_expires(sk) - tp->rcv_tstamp; if (rcv_delta <= timeout) return false; @@ -697,9 +697,9 @@ void tcp_write_timer_handler(struct sock *sk) !icsk->icsk_pending) return; - if (time_after(icsk_timeout(icsk), jiffies)) { + if (time_after(tcp_timeout_expires(sk), jiffies)) { sk_reset_timer(sk, &icsk->icsk_retransmit_timer, - icsk_timeout(icsk)); + tcp_timeout_expires(sk)); return; } tcp_mstamp_refresh(tcp_sk(sk)); diff --git a/net/ipv6/tcp_ipv6.c b/net/ipv6/tcp_ipv6.c index 08113f430124..33c76c3a6da7 100644 --- a/net/ipv6/tcp_ipv6.c +++ b/net/ipv6/tcp_ipv6.c @@ -2163,10 +2163,10 @@ static void get_tcp6_sock(struct seq_file *seq, struct sock *sp, int i) icsk_pending == ICSK_TIME_REO_TIMEOUT || icsk_pending == ICSK_TIME_LOSS_PROBE) { timer_active = 1; - timer_expires = icsk_timeout(icsk); + timer_expires = tcp_timeout_expires(sp); } else if (icsk_pending == ICSK_TIME_PROBE0) { timer_active = 4; - timer_expires = icsk_timeout(icsk); + timer_expires = tcp_timeout_expires(sp); } else if (timer_pending(&sp->sk_timer)) { timer_active = 2; timer_expires = sp->sk_timer.expires; diff --git a/net/mptcp/protocol.c b/net/mptcp/protocol.c index e4ccc57b6f57..4288b6a53b6e 100644 --- a/net/mptcp/protocol.c +++ b/net/mptcp/protocol.c @@ -519,7 +519,7 @@ static long mptcp_timeout_from_subflow(const struct mptcp_subflow_context *subfl const struct sock *ssk = mptcp_subflow_tcp_sock(subflow); return inet_csk(ssk)->icsk_pending && !subflow->stale_count ? - icsk_timeout(inet_csk(ssk)) - jiffies : 0; + tcp_timeout_expires(ssk) - jiffies : 0; } static void mptcp_set_timeout(struct sock *sk) -- cgit v1.2.3 From 27e8257a86516682e2ec5d7543a8909c37ae8b00 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Mon, 24 Nov 2025 17:50:11 +0000 Subject: net: move sk_dst_pending_confirm and sk_pacing_status to sock_read_tx group These two fields are mostly read in TCP tx path, move them in an more appropriate group for better cache locality. Signed-off-by: Eric Dumazet Reviewed-by: Kuniyuki Iwashima Link: https://patch.msgid.link/20251124175013.1473655-3-edumazet@google.com Signed-off-by: Jakub Kicinski --- include/net/sock.h | 4 ++-- net/core/sock.c | 4 ++-- 2 files changed, 4 insertions(+), 4 deletions(-) (limited to 'include/net') diff --git a/include/net/sock.h b/include/net/sock.h index 38d48cfe0741..a89aa97151f5 100644 --- a/include/net/sock.h +++ b/include/net/sock.h @@ -481,8 +481,6 @@ struct sock { struct rb_root tcp_rtx_queue; }; struct sk_buff_head sk_write_queue; - u32 sk_dst_pending_confirm; - u32 sk_pacing_status; /* see enum sk_pacing */ struct page_frag sk_frag; struct timer_list sk_timer; @@ -493,6 +491,8 @@ struct sock { __cacheline_group_end(sock_write_tx); __cacheline_group_begin(sock_read_tx); + u32 sk_dst_pending_confirm; + u32 sk_pacing_status; /* see enum sk_pacing */ unsigned long sk_max_pacing_rate; long sk_sndtimeo; u32 sk_priority; diff --git a/net/core/sock.c b/net/core/sock.c index b26a6cdc9bcd..45c98bf524b2 100644 --- a/net/core/sock.c +++ b/net/core/sock.c @@ -4519,14 +4519,14 @@ static int __init sock_struct_check(void) CACHELINE_ASSERT_GROUP_MEMBER(struct sock, sock_write_tx, sk_send_head); CACHELINE_ASSERT_GROUP_MEMBER(struct sock, sock_write_tx, sk_write_queue); CACHELINE_ASSERT_GROUP_MEMBER(struct sock, sock_write_tx, sk_write_pending); - CACHELINE_ASSERT_GROUP_MEMBER(struct sock, sock_write_tx, sk_dst_pending_confirm); - CACHELINE_ASSERT_GROUP_MEMBER(struct sock, sock_write_tx, sk_pacing_status); CACHELINE_ASSERT_GROUP_MEMBER(struct sock, sock_write_tx, sk_frag); CACHELINE_ASSERT_GROUP_MEMBER(struct sock, sock_write_tx, sk_timer); CACHELINE_ASSERT_GROUP_MEMBER(struct sock, sock_write_tx, sk_pacing_rate); CACHELINE_ASSERT_GROUP_MEMBER(struct sock, sock_write_tx, sk_zckey); CACHELINE_ASSERT_GROUP_MEMBER(struct sock, sock_write_tx, sk_tskey); + CACHELINE_ASSERT_GROUP_MEMBER(struct sock, sock_read_tx, sk_dst_pending_confirm); + CACHELINE_ASSERT_GROUP_MEMBER(struct sock, sock_read_tx, sk_pacing_status); CACHELINE_ASSERT_GROUP_MEMBER(struct sock, sock_read_tx, sk_max_pacing_rate); CACHELINE_ASSERT_GROUP_MEMBER(struct sock, sock_read_tx, sk_sndtimeo); CACHELINE_ASSERT_GROUP_MEMBER(struct sock, sock_read_tx, sk_priority); -- cgit v1.2.3 From 08dfe370239e53494453cee1e2ded2cdaa1efd12 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Mon, 24 Nov 2025 17:50:12 +0000 Subject: tcp: introduce icsk->icsk_keepalive_timer sk->sk_timer has been used for TCP keepalives. Keepalive timers are not in fast path, we want to use sk->sk_timer storage for retransmit timers, for better cache locality. Create icsk->icsk_keepalive_timer and change keepalive code to no longer use sk->sk_timer. Added space is reclaimed in the following patch. This includes changes to MPTCP, which was also using sk_timer. Alias icsk->mptcp_tout_timer and icsk->icsk_keepalive_timer for inet_sk_diag_fill() sake. Signed-off-by: Eric Dumazet Reviewed-by: Kuniyuki Iwashima Link: https://patch.msgid.link/20251124175013.1473655-4-edumazet@google.com Signed-off-by: Jakub Kicinski --- .../networking/net_cachelines/inet_connection_sock.rst | 1 + include/net/inet_connection_sock.h | 11 +++++++++-- net/ipv4/inet_connection_sock.c | 6 +++--- net/ipv4/inet_diag.c | 4 ++-- net/ipv4/tcp_ipv4.c | 4 ++-- net/ipv4/tcp_timer.c | 9 +++++---- net/ipv6/tcp_ipv6.c | 4 ++-- net/mptcp/protocol.c | 10 ++++++---- net/mptcp/protocol.h | 2 +- tools/testing/selftests/bpf/progs/bpf_iter_tcp4.c | 4 ++-- tools/testing/selftests/bpf/progs/bpf_iter_tcp6.c | 4 ++-- 11 files changed, 35 insertions(+), 24 deletions(-) (limited to 'include/net') diff --git a/Documentation/networking/net_cachelines/inet_connection_sock.rst b/Documentation/networking/net_cachelines/inet_connection_sock.rst index 8fae85ebb773..4f65de2def8c 100644 --- a/Documentation/networking/net_cachelines/inet_connection_sock.rst +++ b/Documentation/networking/net_cachelines/inet_connection_sock.rst @@ -14,6 +14,7 @@ struct inet_bind_bucket icsk_bind_hash read_mostly struct inet_bind2_bucket icsk_bind2_hash read_mostly tcp_set_state,inet_put_port struct timer_list icsk_retransmit_timer read_write inet_csk_reset_xmit_timer,tcp_connect struct timer_list icsk_delack_timer read_mostly inet_csk_reset_xmit_timer,tcp_connect +struct timer_list icsk_keepalive_timer u32 icsk_rto read_write tcp_cwnd_validate,tcp_schedule_loss_probe,tcp_connect_init,tcp_connect,tcp_write_xmit,tcp_push_one u32 icsk_rto_min u32 icsk_rto_max read_mostly tcp_reset_xmit_timer diff --git a/include/net/inet_connection_sock.h b/include/net/inet_connection_sock.h index 765c2149d678..e0d90b996348 100644 --- a/include/net/inet_connection_sock.h +++ b/include/net/inet_connection_sock.h @@ -57,6 +57,9 @@ struct inet_connection_sock_af_ops { * @icsk_bind_hash: Bind node * @icsk_bind2_hash: Bind node in the bhash2 table * @icsk_retransmit_timer: Resend (no ack) + * @icsk_delack_timer: Delayed ACK timer + * @icsk_keepalive_timer: Keepalive timer + * @mptcp_tout_timer: mptcp timer * @icsk_rto: Retransmit timeout * @icsk_pmtu_cookie Last pmtu seen by socket * @icsk_ca_ops Pluggable congestion control hook @@ -81,8 +84,12 @@ struct inet_connection_sock { struct request_sock_queue icsk_accept_queue; struct inet_bind_bucket *icsk_bind_hash; struct inet_bind2_bucket *icsk_bind2_hash; - struct timer_list icsk_retransmit_timer; - struct timer_list icsk_delack_timer; + struct timer_list icsk_retransmit_timer; + struct timer_list icsk_delack_timer; + union { + struct timer_list icsk_keepalive_timer; + struct timer_list mptcp_tout_timer; + }; __u32 icsk_rto; __u32 icsk_rto_min; u32 icsk_rto_max; diff --git a/net/ipv4/inet_connection_sock.c b/net/ipv4/inet_connection_sock.c index b4eae731c9ba..4fc09f9bf25d 100644 --- a/net/ipv4/inet_connection_sock.c +++ b/net/ipv4/inet_connection_sock.c @@ -739,7 +739,7 @@ void inet_csk_init_xmit_timers(struct sock *sk, timer_setup(&icsk->icsk_retransmit_timer, retransmit_handler, 0); timer_setup(&icsk->icsk_delack_timer, delack_handler, 0); - timer_setup(&sk->sk_timer, keepalive_handler, 0); + timer_setup(&icsk->icsk_keepalive_timer, keepalive_handler, 0); icsk->icsk_pending = icsk->icsk_ack.pending = 0; } @@ -752,7 +752,7 @@ void inet_csk_clear_xmit_timers(struct sock *sk) sk_stop_timer(sk, &icsk->icsk_retransmit_timer); sk_stop_timer(sk, &icsk->icsk_delack_timer); - sk_stop_timer(sk, &sk->sk_timer); + sk_stop_timer(sk, &icsk->icsk_keepalive_timer); } void inet_csk_clear_xmit_timers_sync(struct sock *sk) @@ -767,7 +767,7 @@ void inet_csk_clear_xmit_timers_sync(struct sock *sk) sk_stop_timer_sync(sk, &icsk->icsk_retransmit_timer); sk_stop_timer_sync(sk, &icsk->icsk_delack_timer); - sk_stop_timer_sync(sk, &sk->sk_timer); + sk_stop_timer_sync(sk, &icsk->icsk_keepalive_timer); } struct dst_entry *inet_csk_route_req(const struct sock *sk, diff --git a/net/ipv4/inet_diag.c b/net/ipv4/inet_diag.c index 9f63c09439a0..3f5b1418a610 100644 --- a/net/ipv4/inet_diag.c +++ b/net/ipv4/inet_diag.c @@ -293,11 +293,11 @@ int inet_sk_diag_fill(struct sock *sk, struct inet_connection_sock *icsk, r->idiag_retrans = READ_ONCE(icsk->icsk_probes_out); r->idiag_expires = jiffies_delta_to_msecs(tcp_timeout_expires(sk) - jiffies); - } else if (timer_pending(&sk->sk_timer)) { + } else if (timer_pending(&icsk->icsk_keepalive_timer)) { r->idiag_timer = 2; r->idiag_retrans = READ_ONCE(icsk->icsk_probes_out); r->idiag_expires = - jiffies_delta_to_msecs(sk->sk_timer.expires - jiffies); + jiffies_delta_to_msecs(icsk->icsk_keepalive_timer.expires - jiffies); } if ((ext & (1 << (INET_DIAG_INFO - 1))) && handler->idiag_info_size) { diff --git a/net/ipv4/tcp_ipv4.c b/net/ipv4/tcp_ipv4.c index 7b8af2c8d03a..f8a9596e8f4d 100644 --- a/net/ipv4/tcp_ipv4.c +++ b/net/ipv4/tcp_ipv4.c @@ -2873,9 +2873,9 @@ static void get_tcp4_sock(struct sock *sk, struct seq_file *f, int i) } else if (icsk_pending == ICSK_TIME_PROBE0) { timer_active = 4; timer_expires = tcp_timeout_expires(sk); - } else if (timer_pending(&sk->sk_timer)) { + } else if (timer_pending(&icsk->icsk_keepalive_timer)) { timer_active = 2; - timer_expires = sk->sk_timer.expires; + timer_expires = icsk->icsk_keepalive_timer.expires; } else { timer_active = 0; timer_expires = jiffies; diff --git a/net/ipv4/tcp_timer.c b/net/ipv4/tcp_timer.c index afbd901e610e..d2678dfd8118 100644 --- a/net/ipv4/tcp_timer.c +++ b/net/ipv4/tcp_timer.c @@ -755,12 +755,12 @@ void tcp_syn_ack_timeout(const struct request_sock *req) void tcp_reset_keepalive_timer(struct sock *sk, unsigned long len) { - sk_reset_timer(sk, &sk->sk_timer, jiffies + len); + sk_reset_timer(sk, &inet_csk(sk)->icsk_keepalive_timer, jiffies + len); } static void tcp_delete_keepalive_timer(struct sock *sk) { - sk_stop_timer(sk, &sk->sk_timer); + sk_stop_timer(sk, &inet_csk(sk)->icsk_keepalive_timer); } void tcp_set_keepalive(struct sock *sk, int val) @@ -777,8 +777,9 @@ EXPORT_IPV6_MOD_GPL(tcp_set_keepalive); static void tcp_keepalive_timer(struct timer_list *t) { - struct sock *sk = timer_container_of(sk, t, sk_timer); - struct inet_connection_sock *icsk = inet_csk(sk); + struct inet_connection_sock *icsk = + timer_container_of(icsk, t, icsk_keepalive_timer); + struct sock *sk = &icsk->icsk_inet.sk; struct tcp_sock *tp = tcp_sk(sk); u32 elapsed; diff --git a/net/ipv6/tcp_ipv6.c b/net/ipv6/tcp_ipv6.c index 33c76c3a6da7..280fe5978559 100644 --- a/net/ipv6/tcp_ipv6.c +++ b/net/ipv6/tcp_ipv6.c @@ -2167,9 +2167,9 @@ static void get_tcp6_sock(struct seq_file *seq, struct sock *sp, int i) } else if (icsk_pending == ICSK_TIME_PROBE0) { timer_active = 4; timer_expires = tcp_timeout_expires(sp); - } else if (timer_pending(&sp->sk_timer)) { + } else if (timer_pending(&icsk->icsk_keepalive_timer)) { timer_active = 2; - timer_expires = sp->sk_timer.expires; + timer_expires = icsk->icsk_keepalive_timer.expires; } else { timer_active = 0; timer_expires = jiffies; diff --git a/net/mptcp/protocol.c b/net/mptcp/protocol.c index 4288b6a53b6e..89a5f63921e6 100644 --- a/net/mptcp/protocol.c +++ b/net/mptcp/protocol.c @@ -2374,7 +2374,9 @@ static void mptcp_retransmit_timer(struct timer_list *t) static void mptcp_tout_timer(struct timer_list *t) { - struct sock *sk = timer_container_of(sk, t, sk_timer); + struct inet_connection_sock *icsk = + timer_container_of(icsk, t, mptcp_tout_timer); + struct sock *sk = &icsk->icsk_inet.sk; mptcp_schedule_work(sk); sock_put(sk); @@ -2828,7 +2830,7 @@ void mptcp_reset_tout_timer(struct mptcp_sock *msk, unsigned long fail_tout) */ timeout = inet_csk(sk)->icsk_mtup.probe_timestamp ? close_timeout : fail_tout; - sk_reset_timer(sk, &sk->sk_timer, timeout); + sk_reset_timer(sk, &inet_csk(sk)->mptcp_tout_timer, timeout); } static void mptcp_mp_fail_no_response(struct mptcp_sock *msk) @@ -2974,7 +2976,7 @@ static void __mptcp_init_sock(struct sock *sk) /* re-use the csk retrans timer for MPTCP-level retrans */ timer_setup(&msk->sk.icsk_retransmit_timer, mptcp_retransmit_timer, 0); - timer_setup(&sk->sk_timer, mptcp_tout_timer, 0); + timer_setup(&msk->sk.mptcp_tout_timer, mptcp_tout_timer, 0); } static void mptcp_ca_reset(struct sock *sk) @@ -3176,7 +3178,7 @@ static void __mptcp_destroy_sock(struct sock *sk) might_sleep(); mptcp_stop_rtx_timer(sk); - sk_stop_timer(sk, &sk->sk_timer); + sk_stop_timer(sk, &inet_csk(sk)->mptcp_tout_timer); msk->pm.status = 0; mptcp_release_sched(msk); diff --git a/net/mptcp/protocol.h b/net/mptcp/protocol.h index 313da78e2b75..9c0d17876b22 100644 --- a/net/mptcp/protocol.h +++ b/net/mptcp/protocol.h @@ -892,7 +892,7 @@ static inline void mptcp_stop_tout_timer(struct sock *sk) if (!inet_csk(sk)->icsk_mtup.probe_timestamp) return; - sk_stop_timer(sk, &sk->sk_timer); + sk_stop_timer(sk, &inet_csk(sk)->mptcp_tout_timer); inet_csk(sk)->icsk_mtup.probe_timestamp = 0; } diff --git a/tools/testing/selftests/bpf/progs/bpf_iter_tcp4.c b/tools/testing/selftests/bpf/progs/bpf_iter_tcp4.c index 164640db3a29..685811326a04 100644 --- a/tools/testing/selftests/bpf/progs/bpf_iter_tcp4.c +++ b/tools/testing/selftests/bpf/progs/bpf_iter_tcp4.c @@ -103,9 +103,9 @@ static int dump_tcp_sock(struct seq_file *seq, struct tcp_sock *tp, } else if (icsk->icsk_pending == ICSK_TIME_PROBE0) { timer_active = 4; timer_expires = icsk->icsk_retransmit_timer.expires; - } else if (timer_pending(&sp->sk_timer)) { + } else if (timer_pending(&icsk->icsk_keepalive_timer)) { timer_active = 2; - timer_expires = sp->sk_timer.expires; + timer_expires = icsk->icsk_keepalive_timer.expires; } else { timer_active = 0; timer_expires = bpf_jiffies64(); diff --git a/tools/testing/selftests/bpf/progs/bpf_iter_tcp6.c b/tools/testing/selftests/bpf/progs/bpf_iter_tcp6.c index 591c703f5032..0f4a92712751 100644 --- a/tools/testing/selftests/bpf/progs/bpf_iter_tcp6.c +++ b/tools/testing/selftests/bpf/progs/bpf_iter_tcp6.c @@ -103,9 +103,9 @@ static int dump_tcp6_sock(struct seq_file *seq, struct tcp6_sock *tp, } else if (icsk->icsk_pending == ICSK_TIME_PROBE0) { timer_active = 4; timer_expires = icsk->icsk_retransmit_timer.expires; - } else if (timer_pending(&sp->sk_timer)) { + } else if (timer_pending(&icsk->icsk_keepalive_timer)) { timer_active = 2; - timer_expires = sp->sk_timer.expires; + timer_expires = icsk->icsk_keepalive_timer.expires; } else { timer_active = 0; timer_expires = bpf_jiffies64(); -- cgit v1.2.3 From 9a5e5334adc039fa652aa071ea95b18db0bc1f43 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Mon, 24 Nov 2025 17:50:13 +0000 Subject: tcp: remove icsk->icsk_retransmit_timer Now sk->sk_timer is no longer used by TCP keepalive, we can use its storage for TCP and MPTCP retransmit timers for better cache locality. Signed-off-by: Eric Dumazet Reviewed-by: Kuniyuki Iwashima Link: https://patch.msgid.link/20251124175013.1473655-5-edumazet@google.com Signed-off-by: Jakub Kicinski --- .../networking/net_cachelines/inet_connection_sock.rst | 1 - include/net/inet_connection_sock.h | 8 +++----- include/net/sock.h | 9 +++++++-- net/ipv4/inet_connection_sock.c | 6 +++--- net/ipv4/tcp_timer.c | 8 +++----- net/mptcp/protocol.c | 15 +++++---------- tools/testing/selftests/bpf/progs/bpf_iter_tcp4.c | 4 ++-- tools/testing/selftests/bpf/progs/bpf_iter_tcp6.c | 4 ++-- 8 files changed, 25 insertions(+), 30 deletions(-) (limited to 'include/net') diff --git a/Documentation/networking/net_cachelines/inet_connection_sock.rst b/Documentation/networking/net_cachelines/inet_connection_sock.rst index 4f65de2def8c..cc2000f55c29 100644 --- a/Documentation/networking/net_cachelines/inet_connection_sock.rst +++ b/Documentation/networking/net_cachelines/inet_connection_sock.rst @@ -12,7 +12,6 @@ struct inet_sock icsk_inet read_mostly r struct request_sock_queue icsk_accept_queue struct inet_bind_bucket icsk_bind_hash read_mostly tcp_set_state struct inet_bind2_bucket icsk_bind2_hash read_mostly tcp_set_state,inet_put_port -struct timer_list icsk_retransmit_timer read_write inet_csk_reset_xmit_timer,tcp_connect struct timer_list icsk_delack_timer read_mostly inet_csk_reset_xmit_timer,tcp_connect struct timer_list icsk_keepalive_timer u32 icsk_rto read_write tcp_cwnd_validate,tcp_schedule_loss_probe,tcp_connect_init,tcp_connect,tcp_write_xmit,tcp_push_one diff --git a/include/net/inet_connection_sock.h b/include/net/inet_connection_sock.h index e0d90b996348..ecb362025c4e 100644 --- a/include/net/inet_connection_sock.h +++ b/include/net/inet_connection_sock.h @@ -56,7 +56,6 @@ struct inet_connection_sock_af_ops { * @icsk_accept_queue: FIFO of established children * @icsk_bind_hash: Bind node * @icsk_bind2_hash: Bind node in the bhash2 table - * @icsk_retransmit_timer: Resend (no ack) * @icsk_delack_timer: Delayed ACK timer * @icsk_keepalive_timer: Keepalive timer * @mptcp_tout_timer: mptcp timer @@ -84,7 +83,6 @@ struct inet_connection_sock { struct request_sock_queue icsk_accept_queue; struct inet_bind_bucket *icsk_bind_hash; struct inet_bind2_bucket *icsk_bind2_hash; - struct timer_list icsk_retransmit_timer; struct timer_list icsk_delack_timer; union { struct timer_list icsk_keepalive_timer; @@ -193,7 +191,7 @@ static inline void inet_csk_delack_init(struct sock *sk) static inline unsigned long tcp_timeout_expires(const struct sock *sk) { - return READ_ONCE(inet_csk(sk)->icsk_retransmit_timer.expires); + return READ_ONCE(sk->tcp_retransmit_timer.expires); } static inline unsigned long @@ -209,7 +207,7 @@ static inline void inet_csk_clear_xmit_timer(struct sock *sk, const int what) if (what == ICSK_TIME_RETRANS || what == ICSK_TIME_PROBE0) { smp_store_release(&icsk->icsk_pending, 0); #ifdef INET_CSK_CLEAR_TIMERS - sk_stop_timer(sk, &icsk->icsk_retransmit_timer); + sk_stop_timer(sk, &sk->tcp_retransmit_timer); #endif } else if (what == ICSK_TIME_DACK) { smp_store_release(&icsk->icsk_ack.pending, 0); @@ -241,7 +239,7 @@ static inline void inet_csk_reset_xmit_timer(struct sock *sk, const int what, if (what == ICSK_TIME_RETRANS || what == ICSK_TIME_PROBE0 || what == ICSK_TIME_LOSS_PROBE || what == ICSK_TIME_REO_TIMEOUT) { smp_store_release(&icsk->icsk_pending, what); - sk_reset_timer(sk, &icsk->icsk_retransmit_timer, when); + sk_reset_timer(sk, &sk->tcp_retransmit_timer, when); } else if (what == ICSK_TIME_DACK) { smp_store_release(&icsk->icsk_ack.pending, icsk->icsk_ack.pending | ICSK_ACK_TIMER); diff --git a/include/net/sock.h b/include/net/sock.h index a89aa97151f5..02253c6a578b 100644 --- a/include/net/sock.h +++ b/include/net/sock.h @@ -305,6 +305,8 @@ struct sk_filter; * @sk_txrehash: enable TX hash rethink * @sk_filter: socket filtering instructions * @sk_timer: sock cleanup timer + * @tcp_retransmit_timer: tcp retransmit timer + * @mptcp_retransmit_timer: mptcp retransmit timer * @sk_stamp: time stamp of last packet received * @sk_stamp_seq: lock for accessing sk_stamp on 32 bit architectures only * @sk_tsflags: SO_TIMESTAMPING flags @@ -482,8 +484,11 @@ struct sock { }; struct sk_buff_head sk_write_queue; struct page_frag sk_frag; - struct timer_list sk_timer; - + union { + struct timer_list sk_timer; + struct timer_list tcp_retransmit_timer; + struct timer_list mptcp_retransmit_timer; + }; unsigned long sk_pacing_rate; /* bytes per second */ atomic_t sk_zckey; atomic_t sk_tskey; diff --git a/net/ipv4/inet_connection_sock.c b/net/ipv4/inet_connection_sock.c index 4fc09f9bf25d..97d57c52b9ad 100644 --- a/net/ipv4/inet_connection_sock.c +++ b/net/ipv4/inet_connection_sock.c @@ -737,7 +737,7 @@ void inet_csk_init_xmit_timers(struct sock *sk, { struct inet_connection_sock *icsk = inet_csk(sk); - timer_setup(&icsk->icsk_retransmit_timer, retransmit_handler, 0); + timer_setup(&sk->tcp_retransmit_timer, retransmit_handler, 0); timer_setup(&icsk->icsk_delack_timer, delack_handler, 0); timer_setup(&icsk->icsk_keepalive_timer, keepalive_handler, 0); icsk->icsk_pending = icsk->icsk_ack.pending = 0; @@ -750,7 +750,7 @@ void inet_csk_clear_xmit_timers(struct sock *sk) smp_store_release(&icsk->icsk_pending, 0); smp_store_release(&icsk->icsk_ack.pending, 0); - sk_stop_timer(sk, &icsk->icsk_retransmit_timer); + sk_stop_timer(sk, &sk->tcp_retransmit_timer); sk_stop_timer(sk, &icsk->icsk_delack_timer); sk_stop_timer(sk, &icsk->icsk_keepalive_timer); } @@ -765,7 +765,7 @@ void inet_csk_clear_xmit_timers_sync(struct sock *sk) smp_store_release(&icsk->icsk_pending, 0); smp_store_release(&icsk->icsk_ack.pending, 0); - sk_stop_timer_sync(sk, &icsk->icsk_retransmit_timer); + sk_stop_timer_sync(sk, &sk->tcp_retransmit_timer); sk_stop_timer_sync(sk, &icsk->icsk_delack_timer); sk_stop_timer_sync(sk, &icsk->icsk_keepalive_timer); } diff --git a/net/ipv4/tcp_timer.c b/net/ipv4/tcp_timer.c index d2678dfd8118..160080c9021d 100644 --- a/net/ipv4/tcp_timer.c +++ b/net/ipv4/tcp_timer.c @@ -698,7 +698,7 @@ void tcp_write_timer_handler(struct sock *sk) return; if (time_after(tcp_timeout_expires(sk), jiffies)) { - sk_reset_timer(sk, &icsk->icsk_retransmit_timer, + sk_reset_timer(sk, &sk->tcp_retransmit_timer, tcp_timeout_expires(sk)); return; } @@ -725,12 +725,10 @@ void tcp_write_timer_handler(struct sock *sk) static void tcp_write_timer(struct timer_list *t) { - struct inet_connection_sock *icsk = - timer_container_of(icsk, t, icsk_retransmit_timer); - struct sock *sk = &icsk->icsk_inet.sk; + struct sock *sk = timer_container_of(sk, t, tcp_retransmit_timer); /* Avoid locking the socket when there is no pending event. */ - if (!smp_load_acquire(&icsk->icsk_pending)) + if (!smp_load_acquire(&inet_csk(sk)->icsk_pending)) goto out; bh_lock_sock(sk); diff --git a/net/mptcp/protocol.c b/net/mptcp/protocol.c index 89a5f63921e6..bb7d634cf312 100644 --- a/net/mptcp/protocol.c +++ b/net/mptcp/protocol.c @@ -411,9 +411,7 @@ static bool __mptcp_move_skb(struct sock *sk, struct sk_buff *skb) static void mptcp_stop_rtx_timer(struct sock *sk) { - struct inet_connection_sock *icsk = inet_csk(sk); - - sk_stop_timer(sk, &icsk->icsk_retransmit_timer); + sk_stop_timer(sk, &sk->mptcp_retransmit_timer); mptcp_sk(sk)->timer_ival = 0; } @@ -954,12 +952,11 @@ static void __mptcp_flush_join_list(struct sock *sk, struct list_head *join_list static bool mptcp_rtx_timer_pending(struct sock *sk) { - return timer_pending(&inet_csk(sk)->icsk_retransmit_timer); + return timer_pending(&sk->mptcp_retransmit_timer); } static void mptcp_reset_rtx_timer(struct sock *sk) { - struct inet_connection_sock *icsk = inet_csk(sk); unsigned long tout; /* prevent rescheduling on close */ @@ -967,7 +964,7 @@ static void mptcp_reset_rtx_timer(struct sock *sk) return; tout = mptcp_sk(sk)->timer_ival; - sk_reset_timer(sk, &icsk->icsk_retransmit_timer, jiffies + tout); + sk_reset_timer(sk, &sk->mptcp_retransmit_timer, jiffies + tout); } bool mptcp_schedule_work(struct sock *sk) @@ -2354,9 +2351,7 @@ out_err: static void mptcp_retransmit_timer(struct timer_list *t) { - struct inet_connection_sock *icsk = timer_container_of(icsk, t, - icsk_retransmit_timer); - struct sock *sk = &icsk->icsk_inet.sk; + struct sock *sk = timer_container_of(sk, t, mptcp_retransmit_timer); struct mptcp_sock *msk = mptcp_sk(sk); bh_lock_sock(sk); @@ -2975,7 +2970,7 @@ static void __mptcp_init_sock(struct sock *sk) spin_lock_init(&msk->fallback_lock); /* re-use the csk retrans timer for MPTCP-level retrans */ - timer_setup(&msk->sk.icsk_retransmit_timer, mptcp_retransmit_timer, 0); + timer_setup(&sk->mptcp_retransmit_timer, mptcp_retransmit_timer, 0); timer_setup(&msk->sk.mptcp_tout_timer, mptcp_tout_timer, 0); } diff --git a/tools/testing/selftests/bpf/progs/bpf_iter_tcp4.c b/tools/testing/selftests/bpf/progs/bpf_iter_tcp4.c index 685811326a04..b1e509b231cd 100644 --- a/tools/testing/selftests/bpf/progs/bpf_iter_tcp4.c +++ b/tools/testing/selftests/bpf/progs/bpf_iter_tcp4.c @@ -99,10 +99,10 @@ static int dump_tcp_sock(struct seq_file *seq, struct tcp_sock *tp, icsk->icsk_pending == ICSK_TIME_REO_TIMEOUT || icsk->icsk_pending == ICSK_TIME_LOSS_PROBE) { timer_active = 1; - timer_expires = icsk->icsk_retransmit_timer.expires; + timer_expires = sp->tcp_retransmit_timer.expires; } else if (icsk->icsk_pending == ICSK_TIME_PROBE0) { timer_active = 4; - timer_expires = icsk->icsk_retransmit_timer.expires; + timer_expires = sp->tcp_retransmit_timer.expires; } else if (timer_pending(&icsk->icsk_keepalive_timer)) { timer_active = 2; timer_expires = icsk->icsk_keepalive_timer.expires; diff --git a/tools/testing/selftests/bpf/progs/bpf_iter_tcp6.c b/tools/testing/selftests/bpf/progs/bpf_iter_tcp6.c index 0f4a92712751..dbc7166aee91 100644 --- a/tools/testing/selftests/bpf/progs/bpf_iter_tcp6.c +++ b/tools/testing/selftests/bpf/progs/bpf_iter_tcp6.c @@ -99,10 +99,10 @@ static int dump_tcp6_sock(struct seq_file *seq, struct tcp6_sock *tp, icsk->icsk_pending == ICSK_TIME_REO_TIMEOUT || icsk->icsk_pending == ICSK_TIME_LOSS_PROBE) { timer_active = 1; - timer_expires = icsk->icsk_retransmit_timer.expires; + timer_expires = sp->tcp_retransmit_timer.expires; } else if (icsk->icsk_pending == ICSK_TIME_PROBE0) { timer_active = 4; - timer_expires = icsk->icsk_retransmit_timer.expires; + timer_expires = sp->tcp_retransmit_timer.expires; } else if (timer_pending(&icsk->icsk_keepalive_timer)) { timer_active = 2; timer_expires = icsk->icsk_keepalive_timer.expires; -- cgit v1.2.3 From 93d7a7ed07342f5e3da2d250cfd67f899d0b5318 Mon Sep 17 00:00:00 2001 From: Pablo Neira Ayuso Date: Fri, 10 Oct 2025 12:32:23 +0200 Subject: netfilter: flowtable: move path discovery infrastructure to its own file This file contains the path discovery that is run from the forward chain for the packet offloading the flow into the flowtable. This consists of a series of calls to dev_fill_forward_path() for each device stack. More topologies may be supported in the future, so move this code to its own file to separate it from the nftables flow_offload expression. Signed-off-by: Pablo Neira Ayuso --- include/net/netfilter/nf_flow_table.h | 6 + net/netfilter/Makefile | 1 + net/netfilter/nf_flow_table_path.c | 274 ++++++++++++++++++++++++++++++++++ net/netfilter/nft_flow_offload.c | 259 -------------------------------- 4 files changed, 281 insertions(+), 259 deletions(-) create mode 100644 net/netfilter/nf_flow_table_path.c (limited to 'include/net') diff --git a/include/net/netfilter/nf_flow_table.h b/include/net/netfilter/nf_flow_table.h index c003cd194fa2..e9f72d2558e9 100644 --- a/include/net/netfilter/nf_flow_table.h +++ b/include/net/netfilter/nf_flow_table.h @@ -222,6 +222,12 @@ struct nf_flow_route { struct flow_offload *flow_offload_alloc(struct nf_conn *ct); void flow_offload_free(struct flow_offload *flow); +struct nft_flowtable; +struct nft_pktinfo; +int nft_flow_route(const struct nft_pktinfo *pkt, const struct nf_conn *ct, + struct nf_flow_route *route, enum ip_conntrack_dir dir, + struct nft_flowtable *ft); + static inline int nf_flow_table_offload_add_cb(struct nf_flowtable *flow_table, flow_setup_cb_t *cb, void *cb_priv) diff --git a/net/netfilter/Makefile b/net/netfilter/Makefile index e43e20f529f8..6bfc250e474f 100644 --- a/net/netfilter/Makefile +++ b/net/netfilter/Makefile @@ -141,6 +141,7 @@ obj-$(CONFIG_NFT_FWD_NETDEV) += nft_fwd_netdev.o # flow table infrastructure obj-$(CONFIG_NF_FLOW_TABLE) += nf_flow_table.o nf_flow_table-objs := nf_flow_table_core.o nf_flow_table_ip.o \ + nf_flow_table_path.o \ nf_flow_table_offload.o nf_flow_table_xdp.o nf_flow_table-$(CONFIG_NF_FLOW_TABLE_PROCFS) += nf_flow_table_procfs.o ifeq ($(CONFIG_NF_FLOW_TABLE),m) diff --git a/net/netfilter/nf_flow_table_path.c b/net/netfilter/nf_flow_table_path.c new file mode 100644 index 000000000000..e525e3745651 --- /dev/null +++ b/net/netfilter/nf_flow_table_path.c @@ -0,0 +1,274 @@ +// SPDX-License-Identifier: GPL-2.0-only +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +static enum flow_offload_xmit_type nft_xmit_type(struct dst_entry *dst) +{ + if (dst_xfrm(dst)) + return FLOW_OFFLOAD_XMIT_XFRM; + + return FLOW_OFFLOAD_XMIT_NEIGH; +} + +static void nft_default_forward_path(struct nf_flow_route *route, + struct dst_entry *dst_cache, + enum ip_conntrack_dir dir) +{ + route->tuple[!dir].in.ifindex = dst_cache->dev->ifindex; + route->tuple[dir].dst = dst_cache; + route->tuple[dir].xmit_type = nft_xmit_type(dst_cache); +} + +static bool nft_is_valid_ether_device(const struct net_device *dev) +{ + if (!dev || (dev->flags & IFF_LOOPBACK) || dev->type != ARPHRD_ETHER || + dev->addr_len != ETH_ALEN || !is_valid_ether_addr(dev->dev_addr)) + return false; + + return true; +} + +static int nft_dev_fill_forward_path(const struct nf_flow_route *route, + const struct dst_entry *dst_cache, + const struct nf_conn *ct, + enum ip_conntrack_dir dir, u8 *ha, + struct net_device_path_stack *stack) +{ + const void *daddr = &ct->tuplehash[!dir].tuple.src.u3; + struct net_device *dev = dst_cache->dev; + struct neighbour *n; + u8 nud_state; + + if (!nft_is_valid_ether_device(dev)) + goto out; + + n = dst_neigh_lookup(dst_cache, daddr); + if (!n) + return -1; + + read_lock_bh(&n->lock); + nud_state = n->nud_state; + ether_addr_copy(ha, n->ha); + read_unlock_bh(&n->lock); + neigh_release(n); + + if (!(nud_state & NUD_VALID)) + return -1; + +out: + return dev_fill_forward_path(dev, ha, stack); +} + +struct nft_forward_info { + const struct net_device *indev; + const struct net_device *outdev; + const struct net_device *hw_outdev; + struct id { + __u16 id; + __be16 proto; + } encap[NF_FLOW_TABLE_ENCAP_MAX]; + u8 num_encaps; + u8 ingress_vlans; + u8 h_source[ETH_ALEN]; + u8 h_dest[ETH_ALEN]; + enum flow_offload_xmit_type xmit_type; +}; + +static void nft_dev_path_info(const struct net_device_path_stack *stack, + struct nft_forward_info *info, + unsigned char *ha, struct nf_flowtable *flowtable) +{ + const struct net_device_path *path; + int i; + + memcpy(info->h_dest, ha, ETH_ALEN); + + for (i = 0; i < stack->num_paths; i++) { + path = &stack->path[i]; + switch (path->type) { + case DEV_PATH_ETHERNET: + case DEV_PATH_DSA: + case DEV_PATH_VLAN: + case DEV_PATH_PPPOE: + info->indev = path->dev; + if (is_zero_ether_addr(info->h_source)) + memcpy(info->h_source, path->dev->dev_addr, ETH_ALEN); + + if (path->type == DEV_PATH_ETHERNET) + break; + if (path->type == DEV_PATH_DSA) { + i = stack->num_paths; + break; + } + + /* DEV_PATH_VLAN and DEV_PATH_PPPOE */ + if (info->num_encaps >= NF_FLOW_TABLE_ENCAP_MAX) { + info->indev = NULL; + break; + } + if (!info->outdev) + info->outdev = path->dev; + info->encap[info->num_encaps].id = path->encap.id; + info->encap[info->num_encaps].proto = path->encap.proto; + info->num_encaps++; + if (path->type == DEV_PATH_PPPOE) + memcpy(info->h_dest, path->encap.h_dest, ETH_ALEN); + break; + case DEV_PATH_BRIDGE: + if (is_zero_ether_addr(info->h_source)) + memcpy(info->h_source, path->dev->dev_addr, ETH_ALEN); + + switch (path->bridge.vlan_mode) { + case DEV_PATH_BR_VLAN_UNTAG_HW: + info->ingress_vlans |= BIT(info->num_encaps - 1); + break; + case DEV_PATH_BR_VLAN_TAG: + if (info->num_encaps >= NF_FLOW_TABLE_ENCAP_MAX) { + info->indev = NULL; + break; + } + info->encap[info->num_encaps].id = path->bridge.vlan_id; + info->encap[info->num_encaps].proto = path->bridge.vlan_proto; + info->num_encaps++; + break; + case DEV_PATH_BR_VLAN_UNTAG: + if (WARN_ON_ONCE(info->num_encaps-- == 0)) { + info->indev = NULL; + break; + } + break; + case DEV_PATH_BR_VLAN_KEEP: + break; + } + info->xmit_type = FLOW_OFFLOAD_XMIT_DIRECT; + break; + default: + info->indev = NULL; + break; + } + } + if (!info->outdev) + info->outdev = info->indev; + + info->hw_outdev = info->indev; + + if (nf_flowtable_hw_offload(flowtable) && + nft_is_valid_ether_device(info->indev)) + info->xmit_type = FLOW_OFFLOAD_XMIT_DIRECT; +} + +static bool nft_flowtable_find_dev(const struct net_device *dev, + struct nft_flowtable *ft) +{ + struct nft_hook *hook; + bool found = false; + + list_for_each_entry_rcu(hook, &ft->hook_list, list) { + if (!nft_hook_find_ops_rcu(hook, dev)) + continue; + + found = true; + break; + } + + return found; +} + +static void nft_dev_forward_path(struct nf_flow_route *route, + const struct nf_conn *ct, + enum ip_conntrack_dir dir, + struct nft_flowtable *ft) +{ + const struct dst_entry *dst = route->tuple[dir].dst; + struct net_device_path_stack stack; + struct nft_forward_info info = {}; + unsigned char ha[ETH_ALEN]; + int i; + + if (nft_dev_fill_forward_path(route, dst, ct, dir, ha, &stack) >= 0) + nft_dev_path_info(&stack, &info, ha, &ft->data); + + if (!info.indev || !nft_flowtable_find_dev(info.indev, ft)) + return; + + route->tuple[!dir].in.ifindex = info.indev->ifindex; + for (i = 0; i < info.num_encaps; i++) { + route->tuple[!dir].in.encap[i].id = info.encap[i].id; + route->tuple[!dir].in.encap[i].proto = info.encap[i].proto; + } + route->tuple[!dir].in.num_encaps = info.num_encaps; + route->tuple[!dir].in.ingress_vlans = info.ingress_vlans; + + if (info.xmit_type == FLOW_OFFLOAD_XMIT_DIRECT) { + memcpy(route->tuple[dir].out.h_source, info.h_source, ETH_ALEN); + memcpy(route->tuple[dir].out.h_dest, info.h_dest, ETH_ALEN); + route->tuple[dir].out.ifindex = info.outdev->ifindex; + route->tuple[dir].out.hw_ifindex = info.hw_outdev->ifindex; + route->tuple[dir].xmit_type = info.xmit_type; + } +} + +int nft_flow_route(const struct nft_pktinfo *pkt, const struct nf_conn *ct, + struct nf_flow_route *route, enum ip_conntrack_dir dir, + struct nft_flowtable *ft) +{ + struct dst_entry *this_dst = skb_dst(pkt->skb); + struct dst_entry *other_dst = NULL; + struct flowi fl; + + memset(&fl, 0, sizeof(fl)); + switch (nft_pf(pkt)) { + case NFPROTO_IPV4: + fl.u.ip4.daddr = ct->tuplehash[dir].tuple.src.u3.ip; + fl.u.ip4.saddr = ct->tuplehash[!dir].tuple.src.u3.ip; + fl.u.ip4.flowi4_oif = nft_in(pkt)->ifindex; + fl.u.ip4.flowi4_iif = this_dst->dev->ifindex; + fl.u.ip4.flowi4_dscp = ip4h_dscp(ip_hdr(pkt->skb)); + fl.u.ip4.flowi4_mark = pkt->skb->mark; + fl.u.ip4.flowi4_flags = FLOWI_FLAG_ANYSRC; + break; + case NFPROTO_IPV6: + fl.u.ip6.daddr = ct->tuplehash[dir].tuple.src.u3.in6; + fl.u.ip6.saddr = ct->tuplehash[!dir].tuple.src.u3.in6; + fl.u.ip6.flowi6_oif = nft_in(pkt)->ifindex; + fl.u.ip6.flowi6_iif = this_dst->dev->ifindex; + fl.u.ip6.flowlabel = ip6_flowinfo(ipv6_hdr(pkt->skb)); + fl.u.ip6.flowi6_mark = pkt->skb->mark; + fl.u.ip6.flowi6_flags = FLOWI_FLAG_ANYSRC; + break; + } + + if (!dst_hold_safe(this_dst)) + return -ENOENT; + + nf_route(nft_net(pkt), &other_dst, &fl, false, nft_pf(pkt)); + if (!other_dst) { + dst_release(this_dst); + return -ENOENT; + } + + nft_default_forward_path(route, this_dst, dir); + nft_default_forward_path(route, other_dst, !dir); + + if (route->tuple[dir].xmit_type == FLOW_OFFLOAD_XMIT_NEIGH && + route->tuple[!dir].xmit_type == FLOW_OFFLOAD_XMIT_NEIGH) { + nft_dev_forward_path(route, ct, dir, ft); + nft_dev_forward_path(route, ct, !dir, ft); + } + + return 0; +} +EXPORT_SYMBOL_GPL(nft_flow_route); diff --git a/net/netfilter/nft_flow_offload.c b/net/netfilter/nft_flow_offload.c index e95e5f59a3d6..b8f76c9057fd 100644 --- a/net/netfilter/nft_flow_offload.c +++ b/net/netfilter/nft_flow_offload.c @@ -20,265 +20,6 @@ struct nft_flow_offload { struct nft_flowtable *flowtable; }; -static enum flow_offload_xmit_type nft_xmit_type(struct dst_entry *dst) -{ - if (dst_xfrm(dst)) - return FLOW_OFFLOAD_XMIT_XFRM; - - return FLOW_OFFLOAD_XMIT_NEIGH; -} - -static void nft_default_forward_path(struct nf_flow_route *route, - struct dst_entry *dst_cache, - enum ip_conntrack_dir dir) -{ - route->tuple[!dir].in.ifindex = dst_cache->dev->ifindex; - route->tuple[dir].dst = dst_cache; - route->tuple[dir].xmit_type = nft_xmit_type(dst_cache); -} - -static bool nft_is_valid_ether_device(const struct net_device *dev) -{ - if (!dev || (dev->flags & IFF_LOOPBACK) || dev->type != ARPHRD_ETHER || - dev->addr_len != ETH_ALEN || !is_valid_ether_addr(dev->dev_addr)) - return false; - - return true; -} - -static int nft_dev_fill_forward_path(const struct nf_flow_route *route, - const struct dst_entry *dst_cache, - const struct nf_conn *ct, - enum ip_conntrack_dir dir, u8 *ha, - struct net_device_path_stack *stack) -{ - const void *daddr = &ct->tuplehash[!dir].tuple.src.u3; - struct net_device *dev = dst_cache->dev; - struct neighbour *n; - u8 nud_state; - - if (!nft_is_valid_ether_device(dev)) - goto out; - - n = dst_neigh_lookup(dst_cache, daddr); - if (!n) - return -1; - - read_lock_bh(&n->lock); - nud_state = n->nud_state; - ether_addr_copy(ha, n->ha); - read_unlock_bh(&n->lock); - neigh_release(n); - - if (!(nud_state & NUD_VALID)) - return -1; - -out: - return dev_fill_forward_path(dev, ha, stack); -} - -struct nft_forward_info { - const struct net_device *indev; - const struct net_device *outdev; - const struct net_device *hw_outdev; - struct id { - __u16 id; - __be16 proto; - } encap[NF_FLOW_TABLE_ENCAP_MAX]; - u8 num_encaps; - u8 ingress_vlans; - u8 h_source[ETH_ALEN]; - u8 h_dest[ETH_ALEN]; - enum flow_offload_xmit_type xmit_type; -}; - -static void nft_dev_path_info(const struct net_device_path_stack *stack, - struct nft_forward_info *info, - unsigned char *ha, struct nf_flowtable *flowtable) -{ - const struct net_device_path *path; - int i; - - memcpy(info->h_dest, ha, ETH_ALEN); - - for (i = 0; i < stack->num_paths; i++) { - path = &stack->path[i]; - switch (path->type) { - case DEV_PATH_ETHERNET: - case DEV_PATH_DSA: - case DEV_PATH_VLAN: - case DEV_PATH_PPPOE: - info->indev = path->dev; - if (is_zero_ether_addr(info->h_source)) - memcpy(info->h_source, path->dev->dev_addr, ETH_ALEN); - - if (path->type == DEV_PATH_ETHERNET) - break; - if (path->type == DEV_PATH_DSA) { - i = stack->num_paths; - break; - } - - /* DEV_PATH_VLAN and DEV_PATH_PPPOE */ - if (info->num_encaps >= NF_FLOW_TABLE_ENCAP_MAX) { - info->indev = NULL; - break; - } - if (!info->outdev) - info->outdev = path->dev; - info->encap[info->num_encaps].id = path->encap.id; - info->encap[info->num_encaps].proto = path->encap.proto; - info->num_encaps++; - if (path->type == DEV_PATH_PPPOE) - memcpy(info->h_dest, path->encap.h_dest, ETH_ALEN); - break; - case DEV_PATH_BRIDGE: - if (is_zero_ether_addr(info->h_source)) - memcpy(info->h_source, path->dev->dev_addr, ETH_ALEN); - - switch (path->bridge.vlan_mode) { - case DEV_PATH_BR_VLAN_UNTAG_HW: - info->ingress_vlans |= BIT(info->num_encaps - 1); - break; - case DEV_PATH_BR_VLAN_TAG: - if (info->num_encaps >= NF_FLOW_TABLE_ENCAP_MAX) { - info->indev = NULL; - break; - } - info->encap[info->num_encaps].id = path->bridge.vlan_id; - info->encap[info->num_encaps].proto = path->bridge.vlan_proto; - info->num_encaps++; - break; - case DEV_PATH_BR_VLAN_UNTAG: - if (WARN_ON_ONCE(info->num_encaps-- == 0)) { - info->indev = NULL; - break; - } - break; - case DEV_PATH_BR_VLAN_KEEP: - break; - } - info->xmit_type = FLOW_OFFLOAD_XMIT_DIRECT; - break; - default: - info->indev = NULL; - break; - } - } - if (!info->outdev) - info->outdev = info->indev; - - info->hw_outdev = info->indev; - - if (nf_flowtable_hw_offload(flowtable) && - nft_is_valid_ether_device(info->indev)) - info->xmit_type = FLOW_OFFLOAD_XMIT_DIRECT; -} - -static bool nft_flowtable_find_dev(const struct net_device *dev, - struct nft_flowtable *ft) -{ - struct nft_hook *hook; - bool found = false; - - list_for_each_entry_rcu(hook, &ft->hook_list, list) { - if (!nft_hook_find_ops_rcu(hook, dev)) - continue; - - found = true; - break; - } - - return found; -} - -static void nft_dev_forward_path(struct nf_flow_route *route, - const struct nf_conn *ct, - enum ip_conntrack_dir dir, - struct nft_flowtable *ft) -{ - const struct dst_entry *dst = route->tuple[dir].dst; - struct net_device_path_stack stack; - struct nft_forward_info info = {}; - unsigned char ha[ETH_ALEN]; - int i; - - if (nft_dev_fill_forward_path(route, dst, ct, dir, ha, &stack) >= 0) - nft_dev_path_info(&stack, &info, ha, &ft->data); - - if (!info.indev || !nft_flowtable_find_dev(info.indev, ft)) - return; - - route->tuple[!dir].in.ifindex = info.indev->ifindex; - for (i = 0; i < info.num_encaps; i++) { - route->tuple[!dir].in.encap[i].id = info.encap[i].id; - route->tuple[!dir].in.encap[i].proto = info.encap[i].proto; - } - route->tuple[!dir].in.num_encaps = info.num_encaps; - route->tuple[!dir].in.ingress_vlans = info.ingress_vlans; - - if (info.xmit_type == FLOW_OFFLOAD_XMIT_DIRECT) { - memcpy(route->tuple[dir].out.h_source, info.h_source, ETH_ALEN); - memcpy(route->tuple[dir].out.h_dest, info.h_dest, ETH_ALEN); - route->tuple[dir].out.ifindex = info.outdev->ifindex; - route->tuple[dir].out.hw_ifindex = info.hw_outdev->ifindex; - route->tuple[dir].xmit_type = info.xmit_type; - } -} - -static int nft_flow_route(const struct nft_pktinfo *pkt, - const struct nf_conn *ct, - struct nf_flow_route *route, - enum ip_conntrack_dir dir, - struct nft_flowtable *ft) -{ - struct dst_entry *this_dst = skb_dst(pkt->skb); - struct dst_entry *other_dst = NULL; - struct flowi fl; - - memset(&fl, 0, sizeof(fl)); - switch (nft_pf(pkt)) { - case NFPROTO_IPV4: - fl.u.ip4.daddr = ct->tuplehash[dir].tuple.src.u3.ip; - fl.u.ip4.saddr = ct->tuplehash[!dir].tuple.src.u3.ip; - fl.u.ip4.flowi4_oif = nft_in(pkt)->ifindex; - fl.u.ip4.flowi4_iif = this_dst->dev->ifindex; - fl.u.ip4.flowi4_dscp = ip4h_dscp(ip_hdr(pkt->skb)); - fl.u.ip4.flowi4_mark = pkt->skb->mark; - fl.u.ip4.flowi4_flags = FLOWI_FLAG_ANYSRC; - break; - case NFPROTO_IPV6: - fl.u.ip6.daddr = ct->tuplehash[dir].tuple.src.u3.in6; - fl.u.ip6.saddr = ct->tuplehash[!dir].tuple.src.u3.in6; - fl.u.ip6.flowi6_oif = nft_in(pkt)->ifindex; - fl.u.ip6.flowi6_iif = this_dst->dev->ifindex; - fl.u.ip6.flowlabel = ip6_flowinfo(ipv6_hdr(pkt->skb)); - fl.u.ip6.flowi6_mark = pkt->skb->mark; - fl.u.ip6.flowi6_flags = FLOWI_FLAG_ANYSRC; - break; - } - - if (!dst_hold_safe(this_dst)) - return -ENOENT; - - nf_route(nft_net(pkt), &other_dst, &fl, false, nft_pf(pkt)); - if (!other_dst) { - dst_release(this_dst); - return -ENOENT; - } - - nft_default_forward_path(route, this_dst, dir); - nft_default_forward_path(route, other_dst, !dir); - - if (route->tuple[dir].xmit_type == FLOW_OFFLOAD_XMIT_NEIGH && - route->tuple[!dir].xmit_type == FLOW_OFFLOAD_XMIT_NEIGH) { - nft_dev_forward_path(route, ct, dir, ft); - nft_dev_forward_path(route, ct, !dir, ft); - } - - return 0; -} - static bool nft_flow_offload_skip(struct sk_buff *skb, int family) { if (skb_sec_path(skb)) -- cgit v1.2.3 From b5964aac51e0c286a50e68225e0dfcf11fb554cb Mon Sep 17 00:00:00 2001 From: Pablo Neira Ayuso Date: Fri, 10 Oct 2025 12:32:35 +0200 Subject: netfilter: flowtable: consolidate xmit path Use dev_queue_xmit() for the XMIT_NEIGH case. Store the interface index of the real device behind the vlan/pppoe device, this introduces an extra lookup for the real device in the xmit path because rt->dst.dev provides the vlan/pppoe device. XMIT_NEIGH now looks more similar to XMIT_DIRECT but the check for stale dst and the neighbour lookup still remain in place which is convenient to deal with network topology changes. Note that nft_flow_route() needs to relax the check for _XMIT_NEIGH so the existing basic xfrm offload (which only works in one direction) does not break. Signed-off-by: Pablo Neira Ayuso --- include/net/netfilter/nf_flow_table.h | 1 + net/netfilter/nf_flow_table_core.c | 1 + net/netfilter/nf_flow_table_ip.c | 87 +++++++++++++++++++++-------------- net/netfilter/nf_flow_table_path.c | 7 ++- 4 files changed, 57 insertions(+), 39 deletions(-) (limited to 'include/net') diff --git a/include/net/netfilter/nf_flow_table.h b/include/net/netfilter/nf_flow_table.h index e9f72d2558e9..7c330caae52b 100644 --- a/include/net/netfilter/nf_flow_table.h +++ b/include/net/netfilter/nf_flow_table.h @@ -141,6 +141,7 @@ struct flow_offload_tuple { union { struct { struct dst_entry *dst_cache; + u32 ifidx; u32 dst_cookie; }; struct { diff --git a/net/netfilter/nf_flow_table_core.c b/net/netfilter/nf_flow_table_core.c index 9441ac3d8c1a..98d7b3708602 100644 --- a/net/netfilter/nf_flow_table_core.c +++ b/net/netfilter/nf_flow_table_core.c @@ -132,6 +132,7 @@ static int flow_offload_fill_route(struct flow_offload *flow, break; case FLOW_OFFLOAD_XMIT_XFRM: case FLOW_OFFLOAD_XMIT_NEIGH: + flow_tuple->ifidx = route->tuple[dir].out.ifindex; flow_tuple->dst_cache = dst; flow_tuple->dst_cookie = flow_offload_dst_cookie(flow_tuple); break; diff --git a/net/netfilter/nf_flow_table_ip.c b/net/netfilter/nf_flow_table_ip.c index 8cd4cf7ae211..eb4f6a11e779 100644 --- a/net/netfilter/nf_flow_table_ip.c +++ b/net/netfilter/nf_flow_table_ip.c @@ -333,19 +333,18 @@ static void nf_flow_encap_pop(struct sk_buff *skb, } } +struct nf_flow_xmit { + const void *dest; + const void *source; + struct net_device *outdev; +}; + static unsigned int nf_flow_queue_xmit(struct net *net, struct sk_buff *skb, - const struct flow_offload_tuple_rhash *tuplehash, - unsigned short type) + struct nf_flow_xmit *xmit) { - struct net_device *outdev; - - outdev = dev_get_by_index_rcu(net, tuplehash->tuple.out.ifidx); - if (!outdev) - return NF_DROP; - - skb->dev = outdev; - dev_hard_header(skb, skb->dev, type, tuplehash->tuple.out.h_dest, - tuplehash->tuple.out.h_source, skb->len); + skb->dev = xmit->outdev; + dev_hard_header(skb, skb->dev, ntohs(skb->protocol), + xmit->dest, xmit->source, skb->len); dev_queue_xmit(skb); return NF_STOLEN; @@ -424,10 +423,10 @@ nf_flow_offload_ip_hook(void *priv, struct sk_buff *skb, struct nf_flowtable_ctx ctx = { .in = state->in, }; + struct nf_flow_xmit xmit = {}; struct flow_offload *flow; - struct net_device *outdev; + struct neighbour *neigh; struct rtable *rt; - __be32 nexthop; int ret; tuplehash = nf_flow_offload_lookup(&ctx, flow_table, skb); @@ -454,25 +453,34 @@ nf_flow_offload_ip_hook(void *priv, struct sk_buff *skb, switch (tuplehash->tuple.xmit_type) { case FLOW_OFFLOAD_XMIT_NEIGH: rt = dst_rtable(tuplehash->tuple.dst_cache); - outdev = rt->dst.dev; - skb->dev = outdev; - nexthop = rt_nexthop(rt, flow->tuplehash[!dir].tuple.src_v4.s_addr); + xmit.outdev = dev_get_by_index_rcu(state->net, tuplehash->tuple.ifidx); + if (!xmit.outdev) { + flow_offload_teardown(flow); + return NF_DROP; + } + neigh = ip_neigh_gw4(rt->dst.dev, rt_nexthop(rt, flow->tuplehash[!dir].tuple.src_v4.s_addr)); + if (IS_ERR(neigh)) { + flow_offload_teardown(flow); + return NF_DROP; + } + xmit.dest = neigh->ha; skb_dst_set_noref(skb, &rt->dst); - neigh_xmit(NEIGH_ARP_TABLE, outdev, &nexthop, skb); - ret = NF_STOLEN; break; case FLOW_OFFLOAD_XMIT_DIRECT: - ret = nf_flow_queue_xmit(state->net, skb, tuplehash, ETH_P_IP); - if (ret == NF_DROP) + xmit.outdev = dev_get_by_index_rcu(state->net, tuplehash->tuple.out.ifidx); + if (!xmit.outdev) { flow_offload_teardown(flow); + return NF_DROP; + } + xmit.dest = tuplehash->tuple.out.h_dest; + xmit.source = tuplehash->tuple.out.h_source; break; default: WARN_ON_ONCE(1); - ret = NF_DROP; - break; + return NF_DROP; } - return ret; + return nf_flow_queue_xmit(state->net, skb, &xmit); } EXPORT_SYMBOL_GPL(nf_flow_offload_ip_hook); @@ -719,9 +727,9 @@ nf_flow_offload_ipv6_hook(void *priv, struct sk_buff *skb, struct nf_flowtable_ctx ctx = { .in = state->in, }; - const struct in6_addr *nexthop; + struct nf_flow_xmit xmit = {}; struct flow_offload *flow; - struct net_device *outdev; + struct neighbour *neigh; struct rt6_info *rt; int ret; @@ -749,24 +757,33 @@ nf_flow_offload_ipv6_hook(void *priv, struct sk_buff *skb, switch (tuplehash->tuple.xmit_type) { case FLOW_OFFLOAD_XMIT_NEIGH: rt = dst_rt6_info(tuplehash->tuple.dst_cache); - outdev = rt->dst.dev; - skb->dev = outdev; - nexthop = rt6_nexthop(rt, &flow->tuplehash[!dir].tuple.src_v6); + xmit.outdev = dev_get_by_index_rcu(state->net, tuplehash->tuple.ifidx); + if (!xmit.outdev) { + flow_offload_teardown(flow); + return NF_DROP; + } + neigh = ip_neigh_gw6(rt->dst.dev, rt6_nexthop(rt, &flow->tuplehash[!dir].tuple.src_v6)); + if (IS_ERR(neigh)) { + flow_offload_teardown(flow); + return NF_DROP; + } + xmit.dest = neigh->ha; skb_dst_set_noref(skb, &rt->dst); - neigh_xmit(NEIGH_ND_TABLE, outdev, nexthop, skb); - ret = NF_STOLEN; break; case FLOW_OFFLOAD_XMIT_DIRECT: - ret = nf_flow_queue_xmit(state->net, skb, tuplehash, ETH_P_IPV6); - if (ret == NF_DROP) + xmit.outdev = dev_get_by_index_rcu(state->net, tuplehash->tuple.out.ifidx); + if (!xmit.outdev) { flow_offload_teardown(flow); + return NF_DROP; + } + xmit.dest = tuplehash->tuple.out.h_dest; + xmit.source = tuplehash->tuple.out.h_source; break; default: WARN_ON_ONCE(1); - ret = NF_DROP; - break; + return NF_DROP; } - return ret; + return nf_flow_queue_xmit(state->net, skb, &xmit); } EXPORT_SYMBOL_GPL(nf_flow_offload_ipv6_hook); diff --git a/net/netfilter/nf_flow_table_path.c b/net/netfilter/nf_flow_table_path.c index e525e3745651..e0c69fea2e0c 100644 --- a/net/netfilter/nf_flow_table_path.c +++ b/net/netfilter/nf_flow_table_path.c @@ -211,11 +211,11 @@ static void nft_dev_forward_path(struct nf_flow_route *route, } route->tuple[!dir].in.num_encaps = info.num_encaps; route->tuple[!dir].in.ingress_vlans = info.ingress_vlans; + route->tuple[dir].out.ifindex = info.outdev->ifindex; if (info.xmit_type == FLOW_OFFLOAD_XMIT_DIRECT) { memcpy(route->tuple[dir].out.h_source, info.h_source, ETH_ALEN); memcpy(route->tuple[dir].out.h_dest, info.h_dest, ETH_ALEN); - route->tuple[dir].out.ifindex = info.outdev->ifindex; route->tuple[dir].out.hw_ifindex = info.hw_outdev->ifindex; route->tuple[dir].xmit_type = info.xmit_type; } @@ -263,11 +263,10 @@ int nft_flow_route(const struct nft_pktinfo *pkt, const struct nf_conn *ct, nft_default_forward_path(route, this_dst, dir); nft_default_forward_path(route, other_dst, !dir); - if (route->tuple[dir].xmit_type == FLOW_OFFLOAD_XMIT_NEIGH && - route->tuple[!dir].xmit_type == FLOW_OFFLOAD_XMIT_NEIGH) { + if (route->tuple[dir].xmit_type == FLOW_OFFLOAD_XMIT_NEIGH) nft_dev_forward_path(route, ct, dir, ft); + if (route->tuple[!dir].xmit_type == FLOW_OFFLOAD_XMIT_NEIGH) nft_dev_forward_path(route, ct, !dir, ft); - } return 0; } -- cgit v1.2.3 From 030feea3097c41ed268c81240e5c334d9977b1c4 Mon Sep 17 00:00:00 2001 From: Pablo Neira Ayuso Date: Fri, 10 Oct 2025 12:50:07 +0200 Subject: netfilter: flowtable: remove hw_ifidx hw_ifidx was originally introduced to store the real netdevice as a requirement for the hardware offload support in: 73f97025a972 ("netfilter: nft_flow_offload: use direct xmit if hardware offload is enabled") Since ("netfilter: flowtable: consolidate xmit path"), ifidx and hw_ifidx points to the real device in the xmit path, remove it. Signed-off-by: Pablo Neira Ayuso --- include/net/netfilter/nf_flow_table.h | 1 - net/netfilter/nf_flow_table_core.c | 1 - net/netfilter/nf_flow_table_offload.c | 2 +- net/netfilter/nf_flow_table_path.c | 3 --- 4 files changed, 1 insertion(+), 6 deletions(-) (limited to 'include/net') diff --git a/include/net/netfilter/nf_flow_table.h b/include/net/netfilter/nf_flow_table.h index 7c330caae52b..f7306276ece7 100644 --- a/include/net/netfilter/nf_flow_table.h +++ b/include/net/netfilter/nf_flow_table.h @@ -146,7 +146,6 @@ struct flow_offload_tuple { }; struct { u32 ifidx; - u32 hw_ifidx; u8 h_source[ETH_ALEN]; u8 h_dest[ETH_ALEN]; } out; diff --git a/net/netfilter/nf_flow_table_core.c b/net/netfilter/nf_flow_table_core.c index 98d7b3708602..6c6a5165f993 100644 --- a/net/netfilter/nf_flow_table_core.c +++ b/net/netfilter/nf_flow_table_core.c @@ -127,7 +127,6 @@ static int flow_offload_fill_route(struct flow_offload *flow, memcpy(flow_tuple->out.h_source, route->tuple[dir].out.h_source, ETH_ALEN); flow_tuple->out.ifidx = route->tuple[dir].out.ifindex; - flow_tuple->out.hw_ifidx = route->tuple[dir].out.hw_ifindex; dst_release(dst); break; case FLOW_OFFLOAD_XMIT_XFRM: diff --git a/net/netfilter/nf_flow_table_offload.c b/net/netfilter/nf_flow_table_offload.c index e06bc36f49fe..d8f7bfd60ac6 100644 --- a/net/netfilter/nf_flow_table_offload.c +++ b/net/netfilter/nf_flow_table_offload.c @@ -555,7 +555,7 @@ static void flow_offload_redirect(struct net *net, switch (this_tuple->xmit_type) { case FLOW_OFFLOAD_XMIT_DIRECT: this_tuple = &flow->tuplehash[dir].tuple; - ifindex = this_tuple->out.hw_ifidx; + ifindex = this_tuple->out.ifidx; break; case FLOW_OFFLOAD_XMIT_NEIGH: other_tuple = &flow->tuplehash[!dir].tuple; diff --git a/net/netfilter/nf_flow_table_path.c b/net/netfilter/nf_flow_table_path.c index c51e310bb2ab..eb9b33a1873a 100644 --- a/net/netfilter/nf_flow_table_path.c +++ b/net/netfilter/nf_flow_table_path.c @@ -75,7 +75,6 @@ out: struct nft_forward_info { const struct net_device *indev; const struct net_device *outdev; - const struct net_device *hw_outdev; struct id { __u16 id; __be16 proto; @@ -159,7 +158,6 @@ static void nft_dev_path_info(const struct net_device_path_stack *stack, } } info->outdev = info->indev; - info->hw_outdev = info->indev; if (nf_flowtable_hw_offload(flowtable) && nft_is_valid_ether_device(info->indev)) @@ -212,7 +210,6 @@ static void nft_dev_forward_path(struct nf_flow_route *route, if (info.xmit_type == FLOW_OFFLOAD_XMIT_DIRECT) { memcpy(route->tuple[dir].out.h_source, info.h_source, ETH_ALEN); memcpy(route->tuple[dir].out.h_dest, info.h_dest, ETH_ALEN); - route->tuple[dir].out.hw_ifindex = info.hw_outdev->ifindex; route->tuple[dir].xmit_type = info.xmit_type; } } -- cgit v1.2.3 From ab427db17885814069bae891834f20842f0ac3a4 Mon Sep 17 00:00:00 2001 From: Lorenzo Bianconi Date: Fri, 7 Nov 2025 12:14:46 +0100 Subject: netfilter: flowtable: Add IPIP rx sw acceleration Introduce sw acceleration for rx path of IPIP tunnels relying on the netfilter flowtable infrastructure. Subsequent patches will add sw acceleration for IPIP tunnels tx path. This series introduces basic infrastructure to accelerate other tunnel types (e.g. IP6IP6). IPIP rx sw acceleration can be tested running the following scenario where the traffic is forwarded between two NICs (eth0 and eth1) and an IPIP tunnel is used to access a remote site (using eth1 as the underlay device): ETH0 -- TUN0 <==> ETH1 -- [IP network] -- TUN1 (192.168.100.2) $ip addr show 6: eth0: mtu 1500 qdisc noqueue state UP group default qlen 1000 link/ether 00:00:22:33:11:55 brd ff:ff:ff:ff:ff:ff inet 192.168.0.2/24 scope global eth0 valid_lft forever preferred_lft forever 7: eth1: mtu 1500 qdisc noqueue state UP group default qlen 1000 link/ether 00:11:22:33:11:55 brd ff:ff:ff:ff:ff:ff inet 192.168.1.1/24 scope global eth1 valid_lft forever preferred_lft forever 8: tun0@NONE: mtu 1480 qdisc noqueue state UNKNOWN group default qlen 1000 link/ipip 192.168.1.1 peer 192.168.1.2 inet 192.168.100.1/24 scope global tun0 valid_lft forever preferred_lft forever $ip route show default via 192.168.100.2 dev tun0 192.168.0.0/24 dev eth0 proto kernel scope link src 192.168.0.2 192.168.1.0/24 dev eth1 proto kernel scope link src 192.168.1.1 192.168.100.0/24 dev tun0 proto kernel scope link src 192.168.100.1 $nft list ruleset table inet filter { flowtable ft { hook ingress priority filter devices = { eth0, eth1 } } chain forward { type filter hook forward priority filter; policy accept; meta l4proto { tcp, udp } flow add @ft } } Reproducing the scenario described above using veths I got the following results: - TCP stream received from the IPIP tunnel: - net-next: (baseline) ~ 71Gbps - net-next + IPIP flowtbale support: ~101Gbps Signed-off-by: Lorenzo Bianconi Signed-off-by: Pablo Neira Ayuso --- include/linux/netdevice.h | 13 +++++++ include/net/netfilter/nf_flow_table.h | 18 +++++++++ net/ipv4/ipip.c | 25 +++++++++++++ net/netfilter/nf_flow_table_core.c | 3 ++ net/netfilter/nf_flow_table_ip.c | 69 ++++++++++++++++++++++++++++++++--- net/netfilter/nf_flow_table_path.c | 38 +++++++++++++++---- 6 files changed, 153 insertions(+), 13 deletions(-) (limited to 'include/net') diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h index e808071dbb7d..bf99fe8622da 100644 --- a/include/linux/netdevice.h +++ b/include/linux/netdevice.h @@ -877,6 +877,7 @@ enum net_device_path_type { DEV_PATH_PPPOE, DEV_PATH_DSA, DEV_PATH_MTK_WDMA, + DEV_PATH_TUN, }; struct net_device_path { @@ -888,6 +889,18 @@ struct net_device_path { __be16 proto; u8 h_dest[ETH_ALEN]; } encap; + struct { + union { + struct in_addr src_v4; + struct in6_addr src_v6; + }; + union { + struct in_addr dst_v4; + struct in6_addr dst_v6; + }; + + u8 l3_proto; + } tun; struct { enum { DEV_PATH_BR_VLAN_KEEP, diff --git a/include/net/netfilter/nf_flow_table.h b/include/net/netfilter/nf_flow_table.h index f7306276ece7..b09c11c048d5 100644 --- a/include/net/netfilter/nf_flow_table.h +++ b/include/net/netfilter/nf_flow_table.h @@ -107,6 +107,19 @@ enum flow_offload_xmit_type { #define NF_FLOW_TABLE_ENCAP_MAX 2 +struct flow_offload_tunnel { + union { + struct in_addr src_v4; + struct in6_addr src_v6; + }; + union { + struct in_addr dst_v4; + struct in6_addr dst_v6; + }; + + u8 l3_proto; +}; + struct flow_offload_tuple { union { struct in_addr src_v4; @@ -130,12 +143,15 @@ struct flow_offload_tuple { __be16 proto; } encap[NF_FLOW_TABLE_ENCAP_MAX]; + struct flow_offload_tunnel tun; + /* All members above are keys for lookups, see flow_offload_hash(). */ struct { } __hash; u8 dir:2, xmit_type:3, encap_num:2, + tun_num:2, in_vlan_ingress:2; u16 mtu; union { @@ -206,7 +222,9 @@ struct nf_flow_route { u16 id; __be16 proto; } encap[NF_FLOW_TABLE_ENCAP_MAX]; + struct flow_offload_tunnel tun; u8 num_encaps:2, + num_tuns:2, ingress_vlans:2; } in; struct { diff --git a/net/ipv4/ipip.c b/net/ipv4/ipip.c index 3e03af073a1c..ff95b1b9908e 100644 --- a/net/ipv4/ipip.c +++ b/net/ipv4/ipip.c @@ -353,6 +353,30 @@ ipip_tunnel_ctl(struct net_device *dev, struct ip_tunnel_parm_kern *p, int cmd) return ip_tunnel_ctl(dev, p, cmd); } +static int ipip_fill_forward_path(struct net_device_path_ctx *ctx, + struct net_device_path *path) +{ + struct ip_tunnel *tunnel = netdev_priv(ctx->dev); + const struct iphdr *tiph = &tunnel->parms.iph; + struct rtable *rt; + + rt = ip_route_output(dev_net(ctx->dev), tiph->daddr, 0, 0, 0, + RT_SCOPE_UNIVERSE); + if (IS_ERR(rt)) + return PTR_ERR(rt); + + path->type = DEV_PATH_TUN; + path->tun.src_v4.s_addr = tiph->saddr; + path->tun.dst_v4.s_addr = tiph->daddr; + path->tun.l3_proto = IPPROTO_IPIP; + path->dev = ctx->dev; + + ctx->dev = rt->dst.dev; + ip_rt_put(rt); + + return 0; +} + static const struct net_device_ops ipip_netdev_ops = { .ndo_init = ipip_tunnel_init, .ndo_uninit = ip_tunnel_uninit, @@ -362,6 +386,7 @@ static const struct net_device_ops ipip_netdev_ops = { .ndo_get_stats64 = dev_get_tstats64, .ndo_get_iflink = ip_tunnel_get_iflink, .ndo_tunnel_ctl = ipip_tunnel_ctl, + .ndo_fill_forward_path = ipip_fill_forward_path, }; #define IPIP_FEATURES (NETIF_F_SG | \ diff --git a/net/netfilter/nf_flow_table_core.c b/net/netfilter/nf_flow_table_core.c index 6c6a5165f993..06e8251a6644 100644 --- a/net/netfilter/nf_flow_table_core.c +++ b/net/netfilter/nf_flow_table_core.c @@ -118,7 +118,10 @@ static int flow_offload_fill_route(struct flow_offload *flow, flow_tuple->in_vlan_ingress |= BIT(j); j++; } + + flow_tuple->tun = route->tuple[dir].in.tun; flow_tuple->encap_num = route->tuple[dir].in.num_encaps; + flow_tuple->tun_num = route->tuple[dir].in.num_tuns; switch (route->tuple[dir].xmit_type) { case FLOW_OFFLOAD_XMIT_DIRECT: diff --git a/net/netfilter/nf_flow_table_ip.c b/net/netfilter/nf_flow_table_ip.c index 083ceb64ac17..d6a1f0cda189 100644 --- a/net/netfilter/nf_flow_table_ip.c +++ b/net/netfilter/nf_flow_table_ip.c @@ -145,8 +145,11 @@ static bool ip_has_options(unsigned int thoff) static void nf_flow_tuple_encap(struct sk_buff *skb, struct flow_offload_tuple *tuple) { + __be16 inner_proto = skb->protocol; struct vlan_ethhdr *veth; struct pppoe_hdr *phdr; + struct iphdr *iph; + u16 offset = 0; int i = 0; if (skb_vlan_tag_present(skb)) { @@ -159,13 +162,26 @@ static void nf_flow_tuple_encap(struct sk_buff *skb, veth = (struct vlan_ethhdr *)skb_mac_header(skb); tuple->encap[i].id = ntohs(veth->h_vlan_TCI); tuple->encap[i].proto = skb->protocol; + inner_proto = veth->h_vlan_encapsulated_proto; + offset += VLAN_HLEN; break; case htons(ETH_P_PPP_SES): phdr = (struct pppoe_hdr *)skb_network_header(skb); tuple->encap[i].id = ntohs(phdr->sid); tuple->encap[i].proto = skb->protocol; + inner_proto = *((__be16 *)(phdr + 1)); + offset += PPPOE_SES_HLEN; break; } + + if (inner_proto == htons(ETH_P_IP)) { + iph = (struct iphdr *)(skb_network_header(skb) + offset); + if (iph->protocol == IPPROTO_IPIP) { + tuple->tun.dst_v4.s_addr = iph->daddr; + tuple->tun.src_v4.s_addr = iph->saddr; + tuple->tun.l3_proto = IPPROTO_IPIP; + } + } } struct nf_flowtable_ctx { @@ -277,11 +293,46 @@ static unsigned int nf_flow_xmit_xfrm(struct sk_buff *skb, return NF_STOLEN; } +static bool nf_flow_ip4_tunnel_proto(struct sk_buff *skb, u32 *psize) +{ + struct iphdr *iph; + u16 size; + + if (!pskb_may_pull(skb, sizeof(*iph) + *psize)) + return false; + + iph = (struct iphdr *)(skb_network_header(skb) + *psize); + size = iph->ihl << 2; + + if (ip_is_fragment(iph) || unlikely(ip_has_options(size))) + return false; + + if (iph->ttl <= 1) + return false; + + if (iph->protocol == IPPROTO_IPIP) + *psize += size; + + return true; +} + +static void nf_flow_ip4_tunnel_pop(struct sk_buff *skb) +{ + struct iphdr *iph = (struct iphdr *)skb_network_header(skb); + + if (iph->protocol != IPPROTO_IPIP) + return; + + skb_pull(skb, iph->ihl << 2); + skb_reset_network_header(skb); +} + static bool nf_flow_skb_encap_protocol(struct sk_buff *skb, __be16 proto, u32 *offset) { + __be16 inner_proto = skb->protocol; struct vlan_ethhdr *veth; - __be16 inner_proto; + bool ret = false; switch (skb->protocol) { case htons(ETH_P_8021Q): @@ -291,19 +342,23 @@ static bool nf_flow_skb_encap_protocol(struct sk_buff *skb, __be16 proto, veth = (struct vlan_ethhdr *)skb_mac_header(skb); if (veth->h_vlan_encapsulated_proto == proto) { *offset += VLAN_HLEN; - return true; + inner_proto = proto; + ret = true; } break; case htons(ETH_P_PPP_SES): if (nf_flow_pppoe_proto(skb, &inner_proto) && inner_proto == proto) { *offset += PPPOE_SES_HLEN; - return true; + ret = true; } break; } - return false; + if (inner_proto == htons(ETH_P_IP)) + ret = nf_flow_ip4_tunnel_proto(skb, offset); + + return ret; } static void nf_flow_encap_pop(struct sk_buff *skb, @@ -331,6 +386,9 @@ static void nf_flow_encap_pop(struct sk_buff *skb, break; } } + + if (skb->protocol == htons(ETH_P_IP)) + nf_flow_ip4_tunnel_pop(skb); } struct nf_flow_xmit { @@ -356,8 +414,7 @@ nf_flow_offload_lookup(struct nf_flowtable_ctx *ctx, { struct flow_offload_tuple tuple = {}; - if (skb->protocol != htons(ETH_P_IP) && - !nf_flow_skb_encap_protocol(skb, htons(ETH_P_IP), &ctx->offset)) + if (!nf_flow_skb_encap_protocol(skb, htons(ETH_P_IP), &ctx->offset)) return NULL; if (nf_flow_tuple_ip(ctx, skb, &tuple) < 0) diff --git a/net/netfilter/nf_flow_table_path.c b/net/netfilter/nf_flow_table_path.c index eb9b33a1873a..73717cc32df5 100644 --- a/net/netfilter/nf_flow_table_path.c +++ b/net/netfilter/nf_flow_table_path.c @@ -80,6 +80,8 @@ struct nft_forward_info { __be16 proto; } encap[NF_FLOW_TABLE_ENCAP_MAX]; u8 num_encaps; + struct flow_offload_tunnel tun; + u8 num_tuns; u8 ingress_vlans; u8 h_source[ETH_ALEN]; u8 h_dest[ETH_ALEN]; @@ -102,6 +104,7 @@ static void nft_dev_path_info(const struct net_device_path_stack *stack, case DEV_PATH_DSA: case DEV_PATH_VLAN: case DEV_PATH_PPPOE: + case DEV_PATH_TUN: info->indev = path->dev; if (is_zero_ether_addr(info->h_source)) memcpy(info->h_source, path->dev->dev_addr, ETH_ALEN); @@ -113,14 +116,27 @@ static void nft_dev_path_info(const struct net_device_path_stack *stack, break; } - /* DEV_PATH_VLAN and DEV_PATH_PPPOE */ - if (info->num_encaps >= NF_FLOW_TABLE_ENCAP_MAX) { - info->indev = NULL; - break; + /* DEV_PATH_VLAN, DEV_PATH_PPPOE and DEV_PATH_TUN */ + if (path->type == DEV_PATH_TUN) { + if (info->num_tuns) { + info->indev = NULL; + break; + } + info->tun.src_v6 = path->tun.src_v6; + info->tun.dst_v6 = path->tun.dst_v6; + info->tun.l3_proto = path->tun.l3_proto; + info->num_tuns++; + } else { + if (info->num_encaps >= NF_FLOW_TABLE_ENCAP_MAX) { + info->indev = NULL; + break; + } + info->encap[info->num_encaps].id = + path->encap.id; + info->encap[info->num_encaps].proto = + path->encap.proto; + info->num_encaps++; } - info->encap[info->num_encaps].id = path->encap.id; - info->encap[info->num_encaps].proto = path->encap.proto; - info->num_encaps++; if (path->type == DEV_PATH_PPPOE) memcpy(info->h_dest, path->encap.h_dest, ETH_ALEN); break; @@ -203,6 +219,14 @@ static void nft_dev_forward_path(struct nf_flow_route *route, route->tuple[!dir].in.encap[i].id = info.encap[i].id; route->tuple[!dir].in.encap[i].proto = info.encap[i].proto; } + + if (info.num_tuns) { + route->tuple[!dir].in.tun.src_v6 = info.tun.dst_v6; + route->tuple[!dir].in.tun.dst_v6 = info.tun.src_v6; + route->tuple[!dir].in.tun.l3_proto = info.tun.l3_proto; + route->tuple[!dir].in.num_tuns = info.num_tuns; + } + route->tuple[!dir].in.num_encaps = info.num_encaps; route->tuple[!dir].in.ingress_vlans = info.ingress_vlans; route->tuple[dir].out.ifindex = info.outdev->ifindex; -- cgit v1.2.3 From be102eb6a0e7c03db00e50540622f4e43b2d2844 Mon Sep 17 00:00:00 2001 From: Fernando Fernandez Mancera Date: Fri, 21 Nov 2025 01:14:30 +0100 Subject: netfilter: nf_conncount: rework API to use sk_buff directly When using nf_conncount infrastructure for non-confirmed connections a duplicated track is possible due to an optimization introduced since commit d265929930e2 ("netfilter: nf_conncount: reduce unnecessary GC"). In order to fix this introduce a new conncount API that receives directly an sk_buff struct. It fetches the tuple and zone and the corresponding ct from it. It comes with both existing conncount variants nf_conncount_count_skb() and nf_conncount_add_skb(). In addition remove the old API and adjust all the users to use the new one. This way, for each sk_buff struct it is possible to check if there is a ct present and already confirmed. If so, skip the add operation. Fixes: d265929930e2 ("netfilter: nf_conncount: reduce unnecessary GC") Signed-off-by: Fernando Fernandez Mancera Signed-off-by: Pablo Neira Ayuso --- include/net/netfilter/nf_conntrack_count.h | 17 ++- net/netfilter/nf_conncount.c | 177 +++++++++++++++++++---------- net/netfilter/nft_connlimit.c | 21 +--- net/netfilter/xt_connlimit.c | 14 +-- net/openvswitch/conntrack.c | 16 +-- 5 files changed, 142 insertions(+), 103 deletions(-) (limited to 'include/net') diff --git a/include/net/netfilter/nf_conntrack_count.h b/include/net/netfilter/nf_conntrack_count.h index 1b58b5b91ff6..52a06de41aa0 100644 --- a/include/net/netfilter/nf_conntrack_count.h +++ b/include/net/netfilter/nf_conntrack_count.h @@ -18,15 +18,14 @@ struct nf_conncount_list { struct nf_conncount_data *nf_conncount_init(struct net *net, unsigned int keylen); void nf_conncount_destroy(struct net *net, struct nf_conncount_data *data); -unsigned int nf_conncount_count(struct net *net, - struct nf_conncount_data *data, - const u32 *key, - const struct nf_conntrack_tuple *tuple, - const struct nf_conntrack_zone *zone); - -int nf_conncount_add(struct net *net, struct nf_conncount_list *list, - const struct nf_conntrack_tuple *tuple, - const struct nf_conntrack_zone *zone); +unsigned int nf_conncount_count_skb(struct net *net, + const struct sk_buff *skb, + u16 l3num, + struct nf_conncount_data *data, + const u32 *key); + +int nf_conncount_add_skb(struct net *net, const struct sk_buff *skb, + u16 l3num, struct nf_conncount_list *list); void nf_conncount_list_init(struct nf_conncount_list *list); diff --git a/net/netfilter/nf_conncount.c b/net/netfilter/nf_conncount.c index 913ede2f57f9..0ffc5ff78a71 100644 --- a/net/netfilter/nf_conncount.c +++ b/net/netfilter/nf_conncount.c @@ -122,15 +122,65 @@ find_or_evict(struct net *net, struct nf_conncount_list *list, return ERR_PTR(-EAGAIN); } +static bool get_ct_or_tuple_from_skb(struct net *net, + const struct sk_buff *skb, + u16 l3num, + struct nf_conn **ct, + struct nf_conntrack_tuple *tuple, + const struct nf_conntrack_zone **zone, + bool *refcounted) +{ + const struct nf_conntrack_tuple_hash *h; + enum ip_conntrack_info ctinfo; + struct nf_conn *found_ct; + + found_ct = nf_ct_get(skb, &ctinfo); + if (found_ct && !nf_ct_is_template(found_ct)) { + *tuple = found_ct->tuplehash[IP_CT_DIR_ORIGINAL].tuple; + *zone = nf_ct_zone(found_ct); + *ct = found_ct; + return true; + } + + if (!nf_ct_get_tuplepr(skb, skb_network_offset(skb), l3num, net, tuple)) + return false; + + if (found_ct) + *zone = nf_ct_zone(found_ct); + + h = nf_conntrack_find_get(net, *zone, tuple); + if (!h) + return true; + + found_ct = nf_ct_tuplehash_to_ctrack(h); + *refcounted = true; + *ct = found_ct; + + return true; +} + static int __nf_conncount_add(struct net *net, - struct nf_conncount_list *list, - const struct nf_conntrack_tuple *tuple, - const struct nf_conntrack_zone *zone) + const struct sk_buff *skb, + u16 l3num, + struct nf_conncount_list *list) { + const struct nf_conntrack_zone *zone = &nf_ct_zone_dflt; const struct nf_conntrack_tuple_hash *found; struct nf_conncount_tuple *conn, *conn_n; + struct nf_conntrack_tuple tuple; + struct nf_conn *ct = NULL; struct nf_conn *found_ct; unsigned int collect = 0; + bool refcounted = false; + + if (!get_ct_or_tuple_from_skb(net, skb, l3num, &ct, &tuple, &zone, &refcounted)) + return -ENOENT; + + if (ct && nf_ct_is_confirmed(ct)) { + if (refcounted) + nf_ct_put(ct); + return 0; + } if ((u32)jiffies == list->last_gc) goto add_new_node; @@ -144,10 +194,10 @@ static int __nf_conncount_add(struct net *net, if (IS_ERR(found)) { /* Not found, but might be about to be confirmed */ if (PTR_ERR(found) == -EAGAIN) { - if (nf_ct_tuple_equal(&conn->tuple, tuple) && + if (nf_ct_tuple_equal(&conn->tuple, &tuple) && nf_ct_zone_id(&conn->zone, conn->zone.dir) == nf_ct_zone_id(zone, zone->dir)) - return 0; /* already exists */ + goto out_put; /* already exists */ } else { collect++; } @@ -156,7 +206,7 @@ static int __nf_conncount_add(struct net *net, found_ct = nf_ct_tuplehash_to_ctrack(found); - if (nf_ct_tuple_equal(&conn->tuple, tuple) && + if (nf_ct_tuple_equal(&conn->tuple, &tuple) && nf_ct_zone_equal(found_ct, zone, zone->dir)) { /* * We should not see tuples twice unless someone hooks @@ -165,7 +215,7 @@ static int __nf_conncount_add(struct net *net, * Attempt to avoid a re-add in this case. */ nf_ct_put(found_ct); - return 0; + goto out_put; } else if (already_closed(found_ct)) { /* * we do not care about connections which are @@ -188,31 +238,35 @@ add_new_node: if (conn == NULL) return -ENOMEM; - conn->tuple = *tuple; + conn->tuple = tuple; conn->zone = *zone; conn->cpu = raw_smp_processor_id(); conn->jiffies32 = (u32)jiffies; list_add_tail(&conn->node, &list->head); list->count++; list->last_gc = (u32)jiffies; + +out_put: + if (refcounted) + nf_ct_put(ct); return 0; } -int nf_conncount_add(struct net *net, - struct nf_conncount_list *list, - const struct nf_conntrack_tuple *tuple, - const struct nf_conntrack_zone *zone) +int nf_conncount_add_skb(struct net *net, + const struct sk_buff *skb, + u16 l3num, + struct nf_conncount_list *list) { int ret; /* check the saved connections */ spin_lock_bh(&list->list_lock); - ret = __nf_conncount_add(net, list, tuple, zone); + ret = __nf_conncount_add(net, skb, l3num, list); spin_unlock_bh(&list->list_lock); return ret; } -EXPORT_SYMBOL_GPL(nf_conncount_add); +EXPORT_SYMBOL_GPL(nf_conncount_add_skb); void nf_conncount_list_init(struct nf_conncount_list *list) { @@ -309,19 +363,22 @@ static void schedule_gc_worker(struct nf_conncount_data *data, int tree) static unsigned int insert_tree(struct net *net, + const struct sk_buff *skb, + u16 l3num, struct nf_conncount_data *data, struct rb_root *root, unsigned int hash, - const u32 *key, - const struct nf_conntrack_tuple *tuple, - const struct nf_conntrack_zone *zone) + const u32 *key) { struct nf_conncount_rb *gc_nodes[CONNCOUNT_GC_MAX_NODES]; + const struct nf_conntrack_zone *zone = &nf_ct_zone_dflt; + bool do_gc = true, refcounted = false; + unsigned int count = 0, gc_count = 0; struct rb_node **rbnode, *parent; - struct nf_conncount_rb *rbconn; + struct nf_conntrack_tuple tuple; struct nf_conncount_tuple *conn; - unsigned int count = 0, gc_count = 0; - bool do_gc = true; + struct nf_conncount_rb *rbconn; + struct nf_conn *ct = NULL; spin_lock_bh(&nf_conncount_locks[hash]); restart: @@ -340,7 +397,7 @@ restart: } else { int ret; - ret = nf_conncount_add(net, &rbconn->list, tuple, zone); + ret = nf_conncount_add_skb(net, skb, l3num, &rbconn->list); if (ret) count = 0; /* hotdrop */ else @@ -364,30 +421,35 @@ restart: goto restart; } - /* expected case: match, insert new node */ - rbconn = kmem_cache_alloc(conncount_rb_cachep, GFP_ATOMIC); - if (rbconn == NULL) - goto out_unlock; + if (get_ct_or_tuple_from_skb(net, skb, l3num, &ct, &tuple, &zone, &refcounted)) { + /* expected case: match, insert new node */ + rbconn = kmem_cache_alloc(conncount_rb_cachep, GFP_ATOMIC); + if (rbconn == NULL) + goto out_unlock; - conn = kmem_cache_alloc(conncount_conn_cachep, GFP_ATOMIC); - if (conn == NULL) { - kmem_cache_free(conncount_rb_cachep, rbconn); - goto out_unlock; - } + conn = kmem_cache_alloc(conncount_conn_cachep, GFP_ATOMIC); + if (conn == NULL) { + kmem_cache_free(conncount_rb_cachep, rbconn); + goto out_unlock; + } - conn->tuple = *tuple; - conn->zone = *zone; - conn->cpu = raw_smp_processor_id(); - conn->jiffies32 = (u32)jiffies; - memcpy(rbconn->key, key, sizeof(u32) * data->keylen); + conn->tuple = tuple; + conn->zone = *zone; + conn->cpu = raw_smp_processor_id(); + conn->jiffies32 = (u32)jiffies; + memcpy(rbconn->key, key, sizeof(u32) * data->keylen); + + nf_conncount_list_init(&rbconn->list); + list_add(&conn->node, &rbconn->list.head); + count = 1; + rbconn->list.count = count; - nf_conncount_list_init(&rbconn->list); - list_add(&conn->node, &rbconn->list.head); - count = 1; - rbconn->list.count = count; + rb_link_node_rcu(&rbconn->node, parent, rbnode); + rb_insert_color(&rbconn->node, root); - rb_link_node_rcu(&rbconn->node, parent, rbnode); - rb_insert_color(&rbconn->node, root); + if (refcounted) + nf_ct_put(ct); + } out_unlock: spin_unlock_bh(&nf_conncount_locks[hash]); return count; @@ -395,10 +457,10 @@ out_unlock: static unsigned int count_tree(struct net *net, + const struct sk_buff *skb, + u16 l3num, struct nf_conncount_data *data, - const u32 *key, - const struct nf_conntrack_tuple *tuple, - const struct nf_conntrack_zone *zone) + const u32 *key) { struct rb_root *root; struct rb_node *parent; @@ -422,7 +484,7 @@ count_tree(struct net *net, } else { int ret; - if (!tuple) { + if (!skb) { nf_conncount_gc_list(net, &rbconn->list); return rbconn->list.count; } @@ -437,7 +499,7 @@ count_tree(struct net *net, } /* same source network -> be counted! */ - ret = __nf_conncount_add(net, &rbconn->list, tuple, zone); + ret = __nf_conncount_add(net, skb, l3num, &rbconn->list); spin_unlock_bh(&rbconn->list.list_lock); if (ret) return 0; /* hotdrop */ @@ -446,10 +508,10 @@ count_tree(struct net *net, } } - if (!tuple) + if (!skb) return 0; - return insert_tree(net, data, root, hash, key, tuple, zone); + return insert_tree(net, skb, l3num, data, root, hash, key); } static void tree_gc_worker(struct work_struct *work) @@ -511,18 +573,19 @@ next: } /* Count and return number of conntrack entries in 'net' with particular 'key'. - * If 'tuple' is not null, insert it into the accounting data structure. - * Call with RCU read lock. + * If 'skb' is not null, insert the corresponding tuple into the accounting + * data structure. Call with RCU read lock. */ -unsigned int nf_conncount_count(struct net *net, - struct nf_conncount_data *data, - const u32 *key, - const struct nf_conntrack_tuple *tuple, - const struct nf_conntrack_zone *zone) +unsigned int nf_conncount_count_skb(struct net *net, + const struct sk_buff *skb, + u16 l3num, + struct nf_conncount_data *data, + const u32 *key) { - return count_tree(net, data, key, tuple, zone); + return count_tree(net, skb, l3num, data, key); + } -EXPORT_SYMBOL_GPL(nf_conncount_count); +EXPORT_SYMBOL_GPL(nf_conncount_count_skb); struct nf_conncount_data *nf_conncount_init(struct net *net, unsigned int keylen) { diff --git a/net/netfilter/nft_connlimit.c b/net/netfilter/nft_connlimit.c index fc35a11cdca2..5df7134131d2 100644 --- a/net/netfilter/nft_connlimit.c +++ b/net/netfilter/nft_connlimit.c @@ -24,26 +24,11 @@ static inline void nft_connlimit_do_eval(struct nft_connlimit *priv, const struct nft_pktinfo *pkt, const struct nft_set_ext *ext) { - const struct nf_conntrack_zone *zone = &nf_ct_zone_dflt; - const struct nf_conntrack_tuple *tuple_ptr; - struct nf_conntrack_tuple tuple; - enum ip_conntrack_info ctinfo; - const struct nf_conn *ct; unsigned int count; + int err; - tuple_ptr = &tuple; - - ct = nf_ct_get(pkt->skb, &ctinfo); - if (ct != NULL) { - tuple_ptr = &ct->tuplehash[IP_CT_DIR_ORIGINAL].tuple; - zone = nf_ct_zone(ct); - } else if (!nf_ct_get_tuplepr(pkt->skb, skb_network_offset(pkt->skb), - nft_pf(pkt), nft_net(pkt), &tuple)) { - regs->verdict.code = NF_DROP; - return; - } - - if (nf_conncount_add(nft_net(pkt), priv->list, tuple_ptr, zone)) { + err = nf_conncount_add_skb(nft_net(pkt), pkt->skb, nft_pf(pkt), priv->list); + if (err) { regs->verdict.code = NF_DROP; return; } diff --git a/net/netfilter/xt_connlimit.c b/net/netfilter/xt_connlimit.c index 0189f8b6b0bd..848287ab79cf 100644 --- a/net/netfilter/xt_connlimit.c +++ b/net/netfilter/xt_connlimit.c @@ -31,8 +31,6 @@ connlimit_mt(const struct sk_buff *skb, struct xt_action_param *par) { struct net *net = xt_net(par); const struct xt_connlimit_info *info = par->matchinfo; - struct nf_conntrack_tuple tuple; - const struct nf_conntrack_tuple *tuple_ptr = &tuple; const struct nf_conntrack_zone *zone = &nf_ct_zone_dflt; enum ip_conntrack_info ctinfo; const struct nf_conn *ct; @@ -40,13 +38,8 @@ connlimit_mt(const struct sk_buff *skb, struct xt_action_param *par) u32 key[5]; ct = nf_ct_get(skb, &ctinfo); - if (ct != NULL) { - tuple_ptr = &ct->tuplehash[IP_CT_DIR_ORIGINAL].tuple; + if (ct) zone = nf_ct_zone(ct); - } else if (!nf_ct_get_tuplepr(skb, skb_network_offset(skb), - xt_family(par), net, &tuple)) { - goto hotdrop; - } if (xt_family(par) == NFPROTO_IPV6) { const struct ipv6hdr *iph = ipv6_hdr(skb); @@ -69,10 +62,9 @@ connlimit_mt(const struct sk_buff *skb, struct xt_action_param *par) key[1] = zone->id; } - connections = nf_conncount_count(net, info->data, key, tuple_ptr, - zone); + connections = nf_conncount_count_skb(net, skb, xt_family(par), info->data, key); if (connections == 0) - /* kmalloc failed, drop it entirely */ + /* kmalloc failed or tuple couldn't be found, drop it entirely */ goto hotdrop; return (connections > info->limit) ^ !!(info->flags & XT_CONNLIMIT_INVERT); diff --git a/net/openvswitch/conntrack.c b/net/openvswitch/conntrack.c index e573e9221302..a0811e1fba65 100644 --- a/net/openvswitch/conntrack.c +++ b/net/openvswitch/conntrack.c @@ -928,8 +928,8 @@ static u32 ct_limit_get(const struct ovs_ct_limit_info *info, u16 zone) } static int ovs_ct_check_limit(struct net *net, - const struct ovs_conntrack_info *info, - const struct nf_conntrack_tuple *tuple) + const struct sk_buff *skb, + const struct ovs_conntrack_info *info) { struct ovs_net *ovs_net = net_generic(net, ovs_net_id); const struct ovs_ct_limit_info *ct_limit_info = ovs_net->ct_limit_info; @@ -942,8 +942,9 @@ static int ovs_ct_check_limit(struct net *net, if (per_zone_limit == OVS_CT_LIMIT_UNLIMITED) return 0; - connections = nf_conncount_count(net, ct_limit_info->data, - &conncount_key, tuple, &info->zone); + connections = nf_conncount_count_skb(net, skb, info->family, + ct_limit_info->data, + &conncount_key); if (connections > per_zone_limit) return -ENOMEM; @@ -972,8 +973,7 @@ static int ovs_ct_commit(struct net *net, struct sw_flow_key *key, #if IS_ENABLED(CONFIG_NETFILTER_CONNCOUNT) if (static_branch_unlikely(&ovs_ct_limit_enabled)) { if (!nf_ct_is_confirmed(ct)) { - err = ovs_ct_check_limit(net, info, - &ct->tuplehash[IP_CT_DIR_ORIGINAL].tuple); + err = ovs_ct_check_limit(net, skb, info); if (err) { net_warn_ratelimited("openvswitch: zone: %u " "exceeds conntrack limit\n", @@ -1770,8 +1770,8 @@ static int __ovs_ct_limit_get_zone_limit(struct net *net, zone_limit.limit = limit; nf_ct_zone_init(&ct_zone, zone_id, NF_CT_DEFAULT_ZONE_DIR, 0); - zone_limit.count = nf_conncount_count(net, data, &conncount_key, NULL, - &ct_zone); + zone_limit.count = nf_conncount_count_skb(net, NULL, 0, data, + &conncount_key); return nla_put_nohdr(reply, sizeof(zone_limit), &zone_limit); } -- cgit v1.2.3 From 33b2835f0b7e2a458473b0e3a23b54b92108b6b0 Mon Sep 17 00:00:00 2001 From: Luiz Augusto von Dentz Date: Tue, 2 Sep 2025 11:11:40 -0400 Subject: Bluetooth: HCI: Add initial support for PAST This adds PAST related commands (HCI_OP_LE_PAST, HCI_OP_LE_PAST_SET_INFO and HCI_OP_LE_PAST_PARAMS) and events (HCI_EV_LE_PAST_RECEIVED) along with handling of PAST sender and receiver features bits including new MGMG settings ( HCI_EV_LE_PAST_RECEIVED and MGMT_SETTING_PAST_RECEIVER) which userspace can use to determine if PAST is supported by the controller. Signed-off-by: Luiz Augusto von Dentz --- include/net/bluetooth/hci.h | 54 +++++++++++++++++++++++++++ include/net/bluetooth/hci_core.h | 12 ++++++ include/net/bluetooth/mgmt.h | 2 + net/bluetooth/hci_event.c | 79 +++++++++++++++++++++++++++++++++++----- net/bluetooth/hci_sync.c | 3 ++ net/bluetooth/iso.c | 25 +++++++++++++ net/bluetooth/mgmt.c | 12 ++++++ 7 files changed, 177 insertions(+), 10 deletions(-) (limited to 'include/net') diff --git a/include/net/bluetooth/hci.h b/include/net/bluetooth/hci.h index cb4c02d00759..d883ad233ebc 100644 --- a/include/net/bluetooth/hci.h +++ b/include/net/bluetooth/hci.h @@ -647,6 +647,8 @@ enum { #define HCI_LE_EXT_ADV 0x10 #define HCI_LE_PERIODIC_ADV 0x20 #define HCI_LE_CHAN_SEL_ALG2 0x40 +#define HCI_LE_PAST_SENDER 0x01 +#define HCI_LE_PAST_RECEIVER 0x02 #define HCI_LE_CIS_CENTRAL 0x10 #define HCI_LE_CIS_PERIPHERAL 0x20 #define HCI_LE_ISO_BROADCASTER 0x40 @@ -2068,6 +2070,44 @@ struct hci_cp_le_set_privacy_mode { __u8 mode; } __packed; +#define HCI_OP_LE_PAST 0x205a +struct hci_cp_le_past { + __le16 handle; + __le16 service_data; + __le16 sync_handle; +} __packed; + +struct hci_rp_le_past { + __u8 status; + __le16 handle; +} __packed; + +#define HCI_OP_LE_PAST_SET_INFO 0x205b +struct hci_cp_le_past_set_info { + __le16 handle; + __le16 service_data; + __u8 adv_handle; +} __packed; + +struct hci_rp_le_past_set_info { + __u8 status; + __le16 handle; +} __packed; + +#define HCI_OP_LE_PAST_PARAMS 0x205c +struct hci_cp_le_past_params { + __le16 handle; + __u8 mode; + __le16 skip; + __le16 sync_timeout; + __u8 cte_type; +} __packed; + +struct hci_rp_le_past_params { + __u8 status; + __le16 handle; +} __packed; + #define HCI_OP_LE_READ_BUFFER_SIZE_V2 0x2060 struct hci_rp_le_read_buffer_size_v2 { __u8 status; @@ -2800,6 +2840,20 @@ struct hci_evt_le_ext_adv_set_term { __u8 num_evts; } __packed; +#define HCI_EV_LE_PAST_RECEIVED 0x18 +struct hci_ev_le_past_received { + __u8 status; + __le16 handle; + __le16 service_data; + __le16 sync_handle; + __u8 sid; + __u8 bdaddr_type; + bdaddr_t bdaddr; + __u8 phy; + __le16 interval; + __u8 clock_accuracy; +} __packed; + #define HCI_EVT_LE_CIS_ESTABLISHED 0x19 struct hci_evt_le_cis_established { __u8 status; diff --git a/include/net/bluetooth/hci_core.h b/include/net/bluetooth/hci_core.h index 0cb87687837f..1bd12c303e25 100644 --- a/include/net/bluetooth/hci_core.h +++ b/include/net/bluetooth/hci_core.h @@ -2053,6 +2053,18 @@ void hci_conn_del_sysfs(struct hci_conn *conn); #define sync_recv_capable(dev) \ ((dev)->le_features[3] & HCI_LE_ISO_SYNC_RECEIVER) #define sync_recv_enabled(dev) (le_enabled(dev) && sync_recv_capable(dev)) +#define past_sender_capable(dev) \ + ((dev)->le_features[3] & HCI_LE_PAST_SENDER) +#define past_receiver_capable(dev) \ + ((dev)->le_features[3] & HCI_LE_PAST_RECEIVER) +#define past_capable(dev) \ + (past_sender_capable(dev) || past_receiver_capable(dev)) +#define past_sender_enabled(dev) \ + (le_enabled(dev) && past_sender_capable(dev)) +#define past_receiver_enabled(dev) \ + (le_enabled(dev) && past_receiver_capable(dev)) +#define past_enabled(dev) \ + (past_sender_enabled(dev) || past_receiver_enabled(dev)) #define mws_transport_config_capable(dev) (((dev)->commands[30] & 0x08) && \ (!hci_test_quirk((dev), HCI_QUIRK_BROKEN_MWS_TRANSPORT_CONFIG))) diff --git a/include/net/bluetooth/mgmt.h b/include/net/bluetooth/mgmt.h index f5be96f08b9d..8234915854b6 100644 --- a/include/net/bluetooth/mgmt.h +++ b/include/net/bluetooth/mgmt.h @@ -119,6 +119,8 @@ struct mgmt_rp_read_index_list { #define MGMT_SETTING_ISO_BROADCASTER BIT(20) #define MGMT_SETTING_ISO_SYNC_RECEIVER BIT(21) #define MGMT_SETTING_LL_PRIVACY BIT(22) +#define MGMT_SETTING_PAST_SENDER BIT(23) +#define MGMT_SETTING_PAST_RECEIVER BIT(24) #define MGMT_OP_READ_INFO 0x0004 #define MGMT_READ_INFO_SIZE 0 diff --git a/net/bluetooth/hci_event.c b/net/bluetooth/hci_event.c index 3838b90343d9..af34c9938509 100644 --- a/net/bluetooth/hci_event.c +++ b/net/bluetooth/hci_event.c @@ -5936,6 +5936,71 @@ unlock: hci_dev_unlock(hdev); } +static int hci_le_pa_term_sync(struct hci_dev *hdev, __le16 handle) +{ + struct hci_cp_le_pa_term_sync cp; + + memset(&cp, 0, sizeof(cp)); + cp.handle = handle; + + return hci_send_cmd(hdev, HCI_OP_LE_PA_TERM_SYNC, sizeof(cp), &cp); +} + +static void hci_le_past_received_evt(struct hci_dev *hdev, void *data, + struct sk_buff *skb) +{ + struct hci_ev_le_past_received *ev = data; + int mask = hdev->link_mode; + __u8 flags = 0; + struct hci_conn *pa_sync, *conn; + + bt_dev_dbg(hdev, "status 0x%2.2x", ev->status); + + hci_dev_lock(hdev); + + hci_dev_clear_flag(hdev, HCI_PA_SYNC); + + conn = hci_conn_hash_lookup_create_pa_sync(hdev); + if (!conn) { + bt_dev_err(hdev, + "Unable to find connection for dst %pMR sid 0x%2.2x", + &ev->bdaddr, ev->sid); + goto unlock; + } + + conn->sync_handle = le16_to_cpu(ev->sync_handle); + conn->sid = HCI_SID_INVALID; + + mask |= hci_proto_connect_ind(hdev, &ev->bdaddr, PA_LINK, + &flags); + if (!(mask & HCI_LM_ACCEPT)) { + hci_le_pa_term_sync(hdev, ev->sync_handle); + goto unlock; + } + + if (!(flags & HCI_PROTO_DEFER)) + goto unlock; + + /* Add connection to indicate PA sync event */ + pa_sync = hci_conn_add_unset(hdev, PA_LINK, BDADDR_ANY, + HCI_ROLE_SLAVE); + + if (IS_ERR(pa_sync)) + goto unlock; + + pa_sync->sync_handle = le16_to_cpu(ev->sync_handle); + + if (ev->status) { + set_bit(HCI_CONN_PA_SYNC_FAILED, &pa_sync->flags); + + /* Notify iso layer */ + hci_connect_cfm(pa_sync, ev->status); + } + +unlock: + hci_dev_unlock(hdev); +} + static void hci_le_conn_update_complete_evt(struct hci_dev *hdev, void *data, struct sk_buff *skb) { @@ -6412,16 +6477,6 @@ static void hci_le_ext_adv_report_evt(struct hci_dev *hdev, void *data, hci_dev_unlock(hdev); } -static int hci_le_pa_term_sync(struct hci_dev *hdev, __le16 handle) -{ - struct hci_cp_le_pa_term_sync cp; - - memset(&cp, 0, sizeof(cp)); - cp.handle = handle; - - return hci_send_cmd(hdev, HCI_OP_LE_PA_TERM_SYNC, sizeof(cp), &cp); -} - static void hci_le_pa_sync_established_evt(struct hci_dev *hdev, void *data, struct sk_buff *skb) { @@ -7206,6 +7261,10 @@ static const struct hci_le_ev { /* [0x12 = HCI_EV_LE_EXT_ADV_SET_TERM] */ HCI_LE_EV(HCI_EV_LE_EXT_ADV_SET_TERM, hci_le_ext_adv_term_evt, sizeof(struct hci_evt_le_ext_adv_set_term)), + /* [0x18 = HCI_EVT_LE_PAST_RECEIVED] */ + HCI_LE_EV(HCI_EV_LE_PAST_RECEIVED, + hci_le_past_received_evt, + sizeof(struct hci_ev_le_past_received)), /* [0x19 = HCI_EVT_LE_CIS_ESTABLISHED] */ HCI_LE_EV(HCI_EVT_LE_CIS_ESTABLISHED, hci_le_cis_established_evt, sizeof(struct hci_evt_le_cis_established)), diff --git a/net/bluetooth/hci_sync.c b/net/bluetooth/hci_sync.c index 6e76798ec786..54ce93236551 100644 --- a/net/bluetooth/hci_sync.c +++ b/net/bluetooth/hci_sync.c @@ -4393,6 +4393,9 @@ static int hci_le_set_event_mask_sync(struct hci_dev *hdev) if (ext_adv_capable(hdev)) events[2] |= 0x02; /* LE Advertising Set Terminated */ + if (past_receiver_capable(hdev)) + events[2] |= 0x80; /* LE PAST Received */ + if (cis_capable(hdev)) { events[3] |= 0x01; /* LE CIS Established */ if (cis_peripheral_capable(hdev)) diff --git a/net/bluetooth/iso.c b/net/bluetooth/iso.c index 5859ec1c04dd..d0a79f601e8d 100644 --- a/net/bluetooth/iso.c +++ b/net/bluetooth/iso.c @@ -80,6 +80,7 @@ static struct bt_iso_qos default_qos; static bool check_ucast_qos(struct bt_iso_qos *qos); static bool check_bcast_qos(struct bt_iso_qos *qos); static bool iso_match_sid(struct sock *sk, void *data); +static bool iso_match_sid_past(struct sock *sk, void *data); static bool iso_match_sync_handle(struct sock *sk, void *data); static bool iso_match_sync_handle_pa_report(struct sock *sk, void *data); static void iso_sock_disconn(struct sock *sk); @@ -2090,6 +2091,16 @@ static bool iso_match_sid(struct sock *sk, void *data) return ev->sid == iso_pi(sk)->bc_sid; } +static bool iso_match_sid_past(struct sock *sk, void *data) +{ + struct hci_ev_le_past_received *ev = data; + + if (iso_pi(sk)->bc_sid == HCI_SID_INVALID) + return true; + + return ev->sid == iso_pi(sk)->bc_sid; +} + static bool iso_match_sync_handle(struct sock *sk, void *data) { struct hci_evt_le_big_info_adv_report *ev = data; @@ -2109,6 +2120,7 @@ static bool iso_match_sync_handle_pa_report(struct sock *sk, void *data) int iso_connect_ind(struct hci_dev *hdev, bdaddr_t *bdaddr, __u8 *flags) { struct hci_ev_le_pa_sync_established *ev1; + struct hci_ev_le_past_received *ev1a; struct hci_evt_le_big_info_adv_report *ev2; struct hci_ev_le_per_adv_report *ev3; struct sock *sk; @@ -2122,6 +2134,7 @@ int iso_connect_ind(struct hci_dev *hdev, bdaddr_t *bdaddr, __u8 *flags) * SID to listen to and once sync is established its handle needs to * be stored in iso_pi(sk)->sync_handle so it can be matched once * receiving the BIG Info. + * 1a. HCI_EV_LE_PAST_RECEIVED: alternative to 1. * 2. HCI_EVT_LE_BIG_INFO_ADV_REPORT: When connect_ind is triggered by a * a BIG Info it attempts to check if there any listening socket with * the same sync_handle and if it does then attempt to create a sync. @@ -2141,6 +2154,18 @@ int iso_connect_ind(struct hci_dev *hdev, bdaddr_t *bdaddr, __u8 *flags) goto done; } + ev1a = hci_recv_event_data(hdev, HCI_EV_LE_PAST_RECEIVED); + if (ev1a) { + sk = iso_get_sock(&hdev->bdaddr, bdaddr, BT_LISTEN, + iso_match_sid_past, ev1a); + if (sk && !ev1a->status) { + iso_pi(sk)->sync_handle = le16_to_cpu(ev1a->sync_handle); + iso_pi(sk)->bc_sid = ev1a->sid; + } + + goto done; + } + ev2 = hci_recv_event_data(hdev, HCI_EVT_LE_BIG_INFO_ADV_REPORT); if (ev2) { /* Check if BIGInfo report has already been handled */ diff --git a/net/bluetooth/mgmt.c b/net/bluetooth/mgmt.c index 262bf984d2aa..7dbd3d57e66c 100644 --- a/net/bluetooth/mgmt.c +++ b/net/bluetooth/mgmt.c @@ -852,6 +852,12 @@ static u32 get_supported_settings(struct hci_dev *hdev) if (ll_privacy_capable(hdev)) settings |= MGMT_SETTING_LL_PRIVACY; + if (past_sender_capable(hdev)) + settings |= MGMT_SETTING_PAST_SENDER; + + if (past_receiver_capable(hdev)) + settings |= MGMT_SETTING_PAST_RECEIVER; + settings |= MGMT_SETTING_PHY_CONFIGURATION; return settings; @@ -937,6 +943,12 @@ static u32 get_current_settings(struct hci_dev *hdev) if (ll_privacy_enabled(hdev)) settings |= MGMT_SETTING_LL_PRIVACY; + if (past_sender_enabled(hdev)) + settings |= MGMT_SETTING_PAST_SENDER; + + if (past_receiver_enabled(hdev)) + settings |= MGMT_SETTING_PAST_RECEIVER; + return settings; } -- cgit v1.2.3 From c530569adc19b5f0c62955de41f067bad34e3fe0 Mon Sep 17 00:00:00 2001 From: Luiz Augusto von Dentz Date: Tue, 2 Sep 2025 11:14:28 -0400 Subject: Bluetooth: hci_core: Introduce HCI_CONN_FLAG_PAST This introduces a new device flag so userspace can indicate if it wants to enable PAST Receiver for a specific device. Signed-off-by: Luiz Augusto von Dentz --- include/net/bluetooth/hci_core.h | 1 + net/bluetooth/hci_sync.c | 55 +++++++++++++++++++++++++++++++++++++++- 2 files changed, 55 insertions(+), 1 deletion(-) (limited to 'include/net') diff --git a/include/net/bluetooth/hci_core.h b/include/net/bluetooth/hci_core.h index 1bd12c303e25..8c2235444808 100644 --- a/include/net/bluetooth/hci_core.h +++ b/include/net/bluetooth/hci_core.h @@ -166,6 +166,7 @@ enum hci_conn_flags { HCI_CONN_FLAG_REMOTE_WAKEUP = BIT(0), HCI_CONN_FLAG_DEVICE_PRIVACY = BIT(1), HCI_CONN_FLAG_ADDRESS_RESOLUTION = BIT(2), + HCI_CONN_FLAG_PAST = BIT(3), }; typedef u8 hci_conn_flags_t; diff --git a/net/bluetooth/hci_sync.c b/net/bluetooth/hci_sync.c index 54ce93236551..ba6f13e9235c 100644 --- a/net/bluetooth/hci_sync.c +++ b/net/bluetooth/hci_sync.c @@ -4324,6 +4324,10 @@ static int hci_le_set_event_mask_sync(struct hci_dev *hdev) if (ll_privacy_capable(hdev)) hdev->conn_flags |= HCI_CONN_FLAG_ADDRESS_RESOLUTION; + /* Mark PAST if supported */ + if (past_capable(hdev)) + hdev->conn_flags |= HCI_CONN_FLAG_PAST; + /* If the controller supports Extended Scanner Filter * Policies, enable the corresponding event. */ @@ -7024,10 +7028,41 @@ unlock: hci_dev_unlock(hdev); } +static int hci_le_past_params_sync(struct hci_dev *hdev, struct hci_conn *conn, + struct hci_conn *acl, struct bt_iso_qos *qos) +{ + struct hci_cp_le_past_params cp; + int err; + + memset(&cp, 0, sizeof(cp)); + cp.handle = cpu_to_le16(acl->handle); + /* An HCI_LE_Periodic_Advertising_Sync_Transfer_Received event is sent + * to the Host. HCI_LE_Periodic_Advertising_Report events will be + * enabled with duplicate filtering enabled. + */ + cp.mode = 0x03; + cp.skip = cpu_to_le16(qos->bcast.skip); + cp.sync_timeout = cpu_to_le16(qos->bcast.sync_timeout); + cp.cte_type = qos->bcast.sync_cte_type; + + /* HCI_LE_PAST_PARAMS command returns a command complete event so it + * cannot wait for HCI_EV_LE_PAST_RECEIVED. + */ + err = __hci_cmd_sync_status(hdev, HCI_OP_LE_PAST_PARAMS, + sizeof(cp), &cp, HCI_CMD_TIMEOUT); + if (err) + return err; + + /* Wait for HCI_EV_LE_PAST_RECEIVED event */ + return __hci_cmd_sync_status_sk(hdev, HCI_OP_NOP, 0, NULL, + HCI_EV_LE_PAST_RECEIVED, + conn->conn_timeout, NULL); +} + static int hci_le_pa_create_sync(struct hci_dev *hdev, void *data) { struct hci_cp_le_pa_create_sync cp; - struct hci_conn *conn = data; + struct hci_conn *conn = data, *le; struct bt_iso_qos *qos = &conn->iso_qos; int err; @@ -7059,6 +7094,24 @@ static int hci_le_pa_create_sync(struct hci_dev *hdev, void *data) hci_update_passive_scan_sync(hdev); + /* Check if PAST is possible: + * + * 1. Check if an ACL connection with the destination address exists + * 2. Check if that HCI_CONN_FLAG_PAST has been set which indicates that + * user really intended to use PAST. + */ + le = hci_conn_hash_lookup_le(hdev, &conn->dst, conn->dst_type); + if (le) { + struct hci_conn_params *params; + + params = hci_conn_params_lookup(hdev, &le->dst, le->dst_type); + if (params && params->flags & HCI_CONN_FLAG_PAST) { + err = hci_le_past_params_sync(hdev, conn, le, qos); + if (!err) + goto done; + } + } + /* SID has not been set listen for HCI_EV_LE_EXT_ADV_REPORT to update * it. */ -- cgit v1.2.3 From d3413703d5f8b7d1e6f514f9440ed5da1bc30796 Mon Sep 17 00:00:00 2001 From: Luiz Augusto von Dentz Date: Fri, 5 Sep 2025 11:34:44 -0400 Subject: Bluetooth: ISO: Add support to bind to trigger PAST This makes it possible to bind to a different destination address after being connected (BT_CONNECTED, BT_CONNECT2) which then triggers PAST Sender proceedure to transfer the PA Sync to the destination address. Signed-off-by: Luiz Augusto von Dentz --- include/net/bluetooth/hci_core.h | 1 + include/net/bluetooth/hci_sync.h | 1 + net/bluetooth/hci_conn.c | 12 +++++ net/bluetooth/hci_sync.c | 92 +++++++++++++++++++++++++++++++++++++ net/bluetooth/iso.c | 98 +++++++++++++++++++++++++++++++++------- 5 files changed, 187 insertions(+), 17 deletions(-) (limited to 'include/net') diff --git a/include/net/bluetooth/hci_core.h b/include/net/bluetooth/hci_core.h index 8c2235444808..1f74722f3f4d 100644 --- a/include/net/bluetooth/hci_core.h +++ b/include/net/bluetooth/hci_core.h @@ -1602,6 +1602,7 @@ struct hci_conn *hci_bind_cis(struct hci_dev *hdev, bdaddr_t *dst, struct hci_conn *hci_bind_bis(struct hci_dev *hdev, bdaddr_t *dst, __u8 sid, struct bt_iso_qos *qos, __u8 base_len, __u8 *base, u16 timeout); +int hci_past_bis(struct hci_conn *conn, bdaddr_t *dst, __u8 dst_type); struct hci_conn *hci_connect_cis(struct hci_dev *hdev, bdaddr_t *dst, __u8 dst_type, struct bt_iso_qos *qos, u16 timeout); diff --git a/include/net/bluetooth/hci_sync.h b/include/net/bluetooth/hci_sync.h index e352a4e0ef8d..3133f40fa9f9 100644 --- a/include/net/bluetooth/hci_sync.h +++ b/include/net/bluetooth/hci_sync.h @@ -188,3 +188,4 @@ int hci_le_conn_update_sync(struct hci_dev *hdev, struct hci_conn *conn, int hci_connect_pa_sync(struct hci_dev *hdev, struct hci_conn *conn); int hci_connect_big_sync(struct hci_dev *hdev, struct hci_conn *conn); +int hci_past_sync(struct hci_conn *conn, struct hci_conn *le); diff --git a/net/bluetooth/hci_conn.c b/net/bluetooth/hci_conn.c index 6fc0692abf05..4f9dc1435ccc 100644 --- a/net/bluetooth/hci_conn.c +++ b/net/bluetooth/hci_conn.c @@ -2245,6 +2245,18 @@ struct hci_conn *hci_bind_bis(struct hci_dev *hdev, bdaddr_t *dst, __u8 sid, return conn; } +int hci_past_bis(struct hci_conn *conn, bdaddr_t *dst, __u8 dst_type) +{ + struct hci_conn *le; + + /* Lookup existing LE connection to rebind to */ + le = hci_conn_hash_lookup_le(conn->hdev, dst, dst_type); + if (!le) + return -EINVAL; + + return hci_past_sync(conn, le); +} + static void bis_mark_per_adv(struct hci_conn *conn, void *data) { struct iso_list_data *d = data; diff --git a/net/bluetooth/hci_sync.c b/net/bluetooth/hci_sync.c index ba6f13e9235c..65f2701beb49 100644 --- a/net/bluetooth/hci_sync.c +++ b/net/bluetooth/hci_sync.c @@ -7228,3 +7228,95 @@ int hci_connect_big_sync(struct hci_dev *hdev, struct hci_conn *conn) return hci_cmd_sync_queue_once(hdev, hci_le_big_create_sync, conn, create_big_complete); } + +struct past_data { + struct hci_conn *conn; + struct hci_conn *le; +}; + +static void past_complete(struct hci_dev *hdev, void *data, int err) +{ + struct past_data *past = data; + + bt_dev_dbg(hdev, "err %d", err); + + kfree(past); +} + +static int hci_le_past_set_info_sync(struct hci_dev *hdev, void *data) +{ + struct past_data *past = data; + struct hci_cp_le_past_set_info cp; + + hci_dev_lock(hdev); + + if (!hci_conn_valid(hdev, past->conn) || + !hci_conn_valid(hdev, past->le)) { + hci_dev_unlock(hdev); + return -ECANCELED; + } + + memset(&cp, 0, sizeof(cp)); + cp.handle = cpu_to_le16(past->le->handle); + cp.adv_handle = past->conn->iso_qos.bcast.bis; + + hci_dev_unlock(hdev); + + return __hci_cmd_sync_status(hdev, HCI_OP_LE_PAST_SET_INFO, + sizeof(cp), &cp, HCI_CMD_TIMEOUT); +} + +static int hci_le_past_sync(struct hci_dev *hdev, void *data) +{ + struct past_data *past = data; + struct hci_cp_le_past cp; + + hci_dev_lock(hdev); + + if (!hci_conn_valid(hdev, past->conn) || + !hci_conn_valid(hdev, past->le)) { + hci_dev_unlock(hdev); + return -ECANCELED; + } + + memset(&cp, 0, sizeof(cp)); + cp.handle = cpu_to_le16(past->le->handle); + cp.sync_handle = cpu_to_le16(past->conn->sync_handle); + + hci_dev_unlock(hdev); + + return __hci_cmd_sync_status(hdev, HCI_OP_LE_PAST, + sizeof(cp), &cp, HCI_CMD_TIMEOUT); +} + +int hci_past_sync(struct hci_conn *conn, struct hci_conn *le) +{ + struct past_data *data; + int err; + + if (conn->type != BIS_LINK && conn->type != PA_LINK) + return -EINVAL; + + if (!past_sender_capable(conn->hdev)) + return -EOPNOTSUPP; + + data = kmalloc(sizeof(*data), GFP_KERNEL); + if (!data) + return -ENOMEM; + + data->conn = conn; + data->le = le; + + if (conn->role == HCI_ROLE_MASTER) + err = hci_cmd_sync_queue_once(conn->hdev, + hci_le_past_set_info_sync, data, + past_complete); + else + err = hci_cmd_sync_queue_once(conn->hdev, hci_le_past_sync, + data, past_complete); + + if (err) + kfree(data); + + return err; +} diff --git a/net/bluetooth/iso.c b/net/bluetooth/iso.c index d0a79f601e8d..85fa9363b897 100644 --- a/net/bluetooth/iso.c +++ b/net/bluetooth/iso.c @@ -987,20 +987,14 @@ static int iso_sock_bind_bc(struct socket *sock, struct sockaddr_unsized *addr, return 0; } -static int iso_sock_bind_pa_sk(struct sock *sk, struct sockaddr_iso *sa, +/* Must be called on the locked socket. */ +static int iso_sock_rebind_bis(struct sock *sk, struct sockaddr_iso *sa, int addr_len) { int err = 0; - if (sk->sk_type != SOCK_SEQPACKET) { - err = -EINVAL; - goto done; - } - - if (addr_len != sizeof(*sa) + sizeof(*sa->iso_bc)) { - err = -EINVAL; - goto done; - } + if (!test_bit(BT_SK_PA_SYNC, &iso_pi(sk)->flags)) + return -EBADFD; if (sa->iso_bc->bc_num_bis > ISO_MAX_NUM_BIS) { err = -EINVAL; @@ -1023,6 +1017,77 @@ done: return err; } +static struct hci_dev *iso_conn_get_hdev(struct iso_conn *conn) +{ + struct hci_dev *hdev = NULL; + + iso_conn_lock(conn); + if (conn->hcon) + hdev = hci_dev_hold(conn->hcon->hdev); + iso_conn_unlock(conn); + + return hdev; +} + +/* Must be called on the locked socket. */ +static int iso_sock_rebind_bc(struct sock *sk, struct sockaddr_iso *sa, + int addr_len) +{ + struct hci_dev *hdev; + struct hci_conn *bis; + int err; + + if (sk->sk_type != SOCK_SEQPACKET || !iso_pi(sk)->conn) + return -EINVAL; + + /* Check if it is really a Broadcast address being requested */ + if (addr_len != sizeof(*sa) + sizeof(*sa->iso_bc)) + return -EINVAL; + + /* Check if the address hasn't changed then perhaps only the number of + * bis has changed. + */ + if (!bacmp(&iso_pi(sk)->dst, &sa->iso_bc->bc_bdaddr) || + !bacmp(&sa->iso_bc->bc_bdaddr, BDADDR_ANY)) + return iso_sock_rebind_bis(sk, sa, addr_len); + + /* Check if the address type is of LE type */ + if (!bdaddr_type_is_le(sa->iso_bc->bc_bdaddr_type)) + return -EINVAL; + + hdev = iso_conn_get_hdev(iso_pi(sk)->conn); + if (!hdev) + return -EINVAL; + + bis = iso_pi(sk)->conn->hcon; + + /* Release the socket before lookups since that requires hci_dev_lock + * which shall not be acquired while holding sock_lock for proper + * ordering. + */ + release_sock(sk); + hci_dev_lock(bis->hdev); + lock_sock(sk); + + if (!iso_pi(sk)->conn || iso_pi(sk)->conn->hcon != bis) { + /* raced with iso_conn_del() or iso_disconn_sock() */ + err = -ENOTCONN; + goto unlock; + } + + BT_DBG("sk %p %pMR type %u", sk, &sa->iso_bc->bc_bdaddr, + sa->iso_bc->bc_bdaddr_type); + + err = hci_past_bis(bis, &sa->iso_bc->bc_bdaddr, + le_addr_type(sa->iso_bc->bc_bdaddr_type)); + +unlock: + hci_dev_unlock(hdev); + hci_dev_put(hdev); + + return err; +} + static int iso_sock_bind(struct socket *sock, struct sockaddr_unsized *addr, int addr_len) { @@ -1038,13 +1103,12 @@ static int iso_sock_bind(struct socket *sock, struct sockaddr_unsized *addr, lock_sock(sk); - /* Allow the user to bind a PA sync socket to a number - * of BISes to sync to. - */ - if ((sk->sk_state == BT_CONNECT2 || - sk->sk_state == BT_CONNECTED) && - test_bit(BT_SK_PA_SYNC, &iso_pi(sk)->flags)) { - err = iso_sock_bind_pa_sk(sk, sa, addr_len); + if ((sk->sk_state == BT_CONNECT2 || sk->sk_state == BT_CONNECTED) && + addr_len > sizeof(*sa)) { + /* Allow the user to rebind to a different address using + * PAST procedures. + */ + err = iso_sock_rebind_bc(sk, sa, addr_len); goto done; } -- cgit v1.2.3 From 14b06c3a88f7031d64fbce197fad1d400e507663 Mon Sep 17 00:00:00 2001 From: Luiz Augusto von Dentz Date: Fri, 26 Sep 2025 15:56:43 -0400 Subject: Bluetooth: HCI: Always use the identity address when initializing a connection This makes sure hci_conn is initialized with the identity address if a matching IRK exists which avoids the trouble of having to do it at multiple places which seems to be missing (e.g. CIS, BIS and PA). Signed-off-by: Luiz Augusto von Dentz --- include/net/bluetooth/hci_core.h | 4 ++-- net/bluetooth/hci_conn.c | 43 +++++++++++++++++++++++++--------------- net/bluetooth/hci_event.c | 20 +++++++++---------- net/bluetooth/hci_sync.c | 2 +- 4 files changed, 40 insertions(+), 29 deletions(-) (limited to 'include/net') diff --git a/include/net/bluetooth/hci_core.h b/include/net/bluetooth/hci_core.h index 1f74722f3f4d..858b58206e80 100644 --- a/include/net/bluetooth/hci_core.h +++ b/include/net/bluetooth/hci_core.h @@ -1571,9 +1571,9 @@ int hci_le_create_cis_pending(struct hci_dev *hdev); int hci_conn_check_create_cis(struct hci_conn *conn); struct hci_conn *hci_conn_add(struct hci_dev *hdev, int type, bdaddr_t *dst, - u8 role, u16 handle); + u8 dst_type, u8 role, u16 handle); struct hci_conn *hci_conn_add_unset(struct hci_dev *hdev, int type, - bdaddr_t *dst, u8 role); + bdaddr_t *dst, u8 dst_type, u8 role); void hci_conn_del(struct hci_conn *conn); void hci_conn_hash_flush(struct hci_dev *hdev); diff --git a/net/bluetooth/hci_conn.c b/net/bluetooth/hci_conn.c index 4f9dc1435ccc..c3f7828bf9d5 100644 --- a/net/bluetooth/hci_conn.c +++ b/net/bluetooth/hci_conn.c @@ -922,10 +922,12 @@ static int hci_conn_hash_alloc_unset(struct hci_dev *hdev) U16_MAX, GFP_ATOMIC); } -static struct hci_conn *__hci_conn_add(struct hci_dev *hdev, int type, bdaddr_t *dst, +static struct hci_conn *__hci_conn_add(struct hci_dev *hdev, int type, + bdaddr_t *dst, u8 dst_type, u8 role, u16 handle) { struct hci_conn *conn; + struct smp_irk *irk = NULL; switch (type) { case ACL_LINK: @@ -937,12 +939,14 @@ static struct hci_conn *__hci_conn_add(struct hci_dev *hdev, int type, bdaddr_t case PA_LINK: if (!hdev->iso_mtu) return ERR_PTR(-ECONNREFUSED); + irk = hci_get_irk(hdev, dst, dst_type); break; case LE_LINK: if (hdev->le_mtu && hdev->le_mtu < HCI_MIN_LE_MTU) return ERR_PTR(-ECONNREFUSED); if (!hdev->le_mtu && hdev->acl_mtu < HCI_MIN_LE_MTU) return ERR_PTR(-ECONNREFUSED); + irk = hci_get_irk(hdev, dst, dst_type); break; case SCO_LINK: case ESCO_LINK: @@ -960,7 +964,15 @@ static struct hci_conn *__hci_conn_add(struct hci_dev *hdev, int type, bdaddr_t if (!conn) return ERR_PTR(-ENOMEM); - bacpy(&conn->dst, dst); + /* If and IRK exists use its identity address */ + if (!irk) { + bacpy(&conn->dst, dst); + conn->dst_type = dst_type; + } else { + bacpy(&conn->dst, &irk->bdaddr); + conn->dst_type = irk->addr_type; + } + bacpy(&conn->src, &hdev->bdaddr); conn->handle = handle; conn->hdev = hdev; @@ -1059,7 +1071,7 @@ static struct hci_conn *__hci_conn_add(struct hci_dev *hdev, int type, bdaddr_t } struct hci_conn *hci_conn_add_unset(struct hci_dev *hdev, int type, - bdaddr_t *dst, u8 role) + bdaddr_t *dst, u8 dst_type, u8 role) { int handle; @@ -1069,16 +1081,16 @@ struct hci_conn *hci_conn_add_unset(struct hci_dev *hdev, int type, if (unlikely(handle < 0)) return ERR_PTR(-ECONNREFUSED); - return __hci_conn_add(hdev, type, dst, role, handle); + return __hci_conn_add(hdev, type, dst, dst_type, role, handle); } struct hci_conn *hci_conn_add(struct hci_dev *hdev, int type, bdaddr_t *dst, - u8 role, u16 handle) + u8 dst_type, u8 role, u16 handle) { if (handle > HCI_CONN_HANDLE_MAX) return ERR_PTR(-EINVAL); - return __hci_conn_add(hdev, type, dst, role, handle); + return __hci_conn_add(hdev, type, dst, dst_type, role, handle); } static void hci_conn_cleanup_child(struct hci_conn *conn, u8 reason) @@ -1410,14 +1422,13 @@ struct hci_conn *hci_connect_le(struct hci_dev *hdev, bdaddr_t *dst, if (conn) { bacpy(&conn->dst, dst); } else { - conn = hci_conn_add_unset(hdev, LE_LINK, dst, role); + conn = hci_conn_add_unset(hdev, LE_LINK, dst, dst_type, role); if (IS_ERR(conn)) return conn; hci_conn_hold(conn); conn->pending_sec_level = sec_level; } - conn->dst_type = dst_type; conn->sec_level = BT_SECURITY_LOW; conn->conn_timeout = conn_timeout; conn->le_adv_phy = phy; @@ -1587,7 +1598,7 @@ static struct hci_conn *hci_add_bis(struct hci_dev *hdev, bdaddr_t *dst, memcmp(conn->le_per_adv_data, base, base_len))) return ERR_PTR(-EADDRINUSE); - conn = hci_conn_add_unset(hdev, BIS_LINK, dst, HCI_ROLE_MASTER); + conn = hci_conn_add_unset(hdev, BIS_LINK, dst, 0, HCI_ROLE_MASTER); if (IS_ERR(conn)) return conn; @@ -1633,7 +1644,8 @@ struct hci_conn *hci_connect_le_scan(struct hci_dev *hdev, bdaddr_t *dst, BT_DBG("requesting refresh of dst_addr"); - conn = hci_conn_add_unset(hdev, LE_LINK, dst, HCI_ROLE_MASTER); + conn = hci_conn_add_unset(hdev, LE_LINK, dst, dst_type, + HCI_ROLE_MASTER); if (IS_ERR(conn)) return conn; @@ -1644,7 +1656,6 @@ struct hci_conn *hci_connect_le_scan(struct hci_dev *hdev, bdaddr_t *dst, conn->state = BT_CONNECT; set_bit(HCI_CONN_SCANNING, &conn->flags); - conn->dst_type = dst_type; conn->sec_level = BT_SECURITY_LOW; conn->pending_sec_level = sec_level; conn->conn_timeout = conn_timeout; @@ -1681,7 +1692,8 @@ struct hci_conn *hci_connect_acl(struct hci_dev *hdev, bdaddr_t *dst, acl = hci_conn_hash_lookup_ba(hdev, ACL_LINK, dst); if (!acl) { - acl = hci_conn_add_unset(hdev, ACL_LINK, dst, HCI_ROLE_MASTER); + acl = hci_conn_add_unset(hdev, ACL_LINK, dst, 0, + HCI_ROLE_MASTER); if (IS_ERR(acl)) return acl; } @@ -1750,7 +1762,7 @@ struct hci_conn *hci_connect_sco(struct hci_dev *hdev, int type, bdaddr_t *dst, sco = hci_conn_hash_lookup_ba(hdev, type, dst); if (!sco) { - sco = hci_conn_add_unset(hdev, type, dst, HCI_ROLE_MASTER); + sco = hci_conn_add_unset(hdev, type, dst, 0, HCI_ROLE_MASTER); if (IS_ERR(sco)) { hci_conn_drop(acl); return sco; @@ -1942,7 +1954,7 @@ struct hci_conn *hci_bind_cis(struct hci_dev *hdev, bdaddr_t *dst, cis = hci_conn_hash_lookup_cis(hdev, dst, dst_type, qos->ucast.cig, qos->ucast.cis); if (!cis) { - cis = hci_conn_add_unset(hdev, CIS_LINK, dst, + cis = hci_conn_add_unset(hdev, CIS_LINK, dst, dst_type, HCI_ROLE_MASTER); if (IS_ERR(cis)) return cis; @@ -2133,12 +2145,11 @@ struct hci_conn *hci_pa_create_sync(struct hci_dev *hdev, bdaddr_t *dst, bt_dev_dbg(hdev, "dst %pMR type %d sid %d", dst, dst_type, sid); - conn = hci_conn_add_unset(hdev, PA_LINK, dst, HCI_ROLE_SLAVE); + conn = hci_conn_add_unset(hdev, PA_LINK, dst, dst_type, HCI_ROLE_SLAVE); if (IS_ERR(conn)) return conn; conn->iso_qos = *qos; - conn->dst_type = dst_type; conn->sid = sid; conn->state = BT_LISTEN; conn->conn_timeout = msecs_to_jiffies(qos->bcast.sync_timeout * 10); diff --git a/net/bluetooth/hci_event.c b/net/bluetooth/hci_event.c index af34c9938509..7c4ca14f13e5 100644 --- a/net/bluetooth/hci_event.c +++ b/net/bluetooth/hci_event.c @@ -2267,7 +2267,7 @@ static void hci_cs_create_conn(struct hci_dev *hdev, __u8 status) } else { if (!conn) { conn = hci_conn_add_unset(hdev, ACL_LINK, &cp->bdaddr, - HCI_ROLE_MASTER); + 0, HCI_ROLE_MASTER); if (IS_ERR(conn)) bt_dev_err(hdev, "connection err: %ld", PTR_ERR(conn)); } @@ -3123,7 +3123,8 @@ static void hci_conn_complete_evt(struct hci_dev *hdev, void *data, &ev->bdaddr, BDADDR_BREDR)) { conn = hci_conn_add_unset(hdev, ev->link_type, - &ev->bdaddr, HCI_ROLE_SLAVE); + &ev->bdaddr, 0, + HCI_ROLE_SLAVE); if (IS_ERR(conn)) { bt_dev_err(hdev, "connection err: %ld", PTR_ERR(conn)); goto unlock; @@ -3299,7 +3300,7 @@ static void hci_conn_request_evt(struct hci_dev *hdev, void *data, conn = hci_conn_hash_lookup_ba(hdev, ev->link_type, &ev->bdaddr); if (!conn) { - conn = hci_conn_add_unset(hdev, ev->link_type, &ev->bdaddr, + conn = hci_conn_add_unset(hdev, ev->link_type, &ev->bdaddr, 0, HCI_ROLE_SLAVE); if (IS_ERR(conn)) { bt_dev_err(hdev, "connection err: %ld", PTR_ERR(conn)); @@ -5670,14 +5671,13 @@ static void le_conn_complete_evt(struct hci_dev *hdev, u8 status, if (status) goto unlock; - conn = hci_conn_add_unset(hdev, LE_LINK, bdaddr, role); + conn = hci_conn_add_unset(hdev, LE_LINK, bdaddr, bdaddr_type, + role); if (IS_ERR(conn)) { bt_dev_err(hdev, "connection err: %ld", PTR_ERR(conn)); goto unlock; } - conn->dst_type = bdaddr_type; - /* If we didn't have a hci_conn object previously * but we're in central role this must be something * initiated using an accept list. Since accept list based @@ -5982,7 +5982,7 @@ static void hci_le_past_received_evt(struct hci_dev *hdev, void *data, goto unlock; /* Add connection to indicate PA sync event */ - pa_sync = hci_conn_add_unset(hdev, PA_LINK, BDADDR_ANY, + pa_sync = hci_conn_add_unset(hdev, PA_LINK, BDADDR_ANY, 0, HCI_ROLE_SLAVE); if (IS_ERR(pa_sync)) @@ -6515,7 +6515,7 @@ static void hci_le_pa_sync_established_evt(struct hci_dev *hdev, void *data, goto unlock; /* Add connection to indicate PA sync event */ - pa_sync = hci_conn_add_unset(hdev, PA_LINK, BDADDR_ANY, + pa_sync = hci_conn_add_unset(hdev, PA_LINK, BDADDR_ANY, 0, HCI_ROLE_SLAVE); if (IS_ERR(pa_sync)) @@ -6956,7 +6956,7 @@ static void hci_le_cis_req_evt(struct hci_dev *hdev, void *data, cis = hci_conn_hash_lookup_handle(hdev, cis_handle); if (!cis) { - cis = hci_conn_add(hdev, CIS_LINK, &acl->dst, + cis = hci_conn_add(hdev, CIS_LINK, &acl->dst, acl->dst_type, HCI_ROLE_SLAVE, cis_handle); if (IS_ERR(cis)) { hci_le_reject_cis(hdev, ev->cis_handle); @@ -7073,7 +7073,7 @@ static void hci_le_big_sync_established_evt(struct hci_dev *hdev, void *data, bt_dev_dbg(hdev, "ignore too large handle %u", handle); continue; } - bis = hci_conn_add(hdev, BIS_LINK, BDADDR_ANY, + bis = hci_conn_add(hdev, BIS_LINK, BDADDR_ANY, 0, HCI_ROLE_SLAVE, handle); if (IS_ERR(bis)) continue; diff --git a/net/bluetooth/hci_sync.c b/net/bluetooth/hci_sync.c index 65f2701beb49..a36d2414a3ca 100644 --- a/net/bluetooth/hci_sync.c +++ b/net/bluetooth/hci_sync.c @@ -7013,7 +7013,7 @@ static void create_pa_complete(struct hci_dev *hdev, void *data, int err) goto unlock; /* Add connection to indicate PA sync error */ - pa_sync = hci_conn_add_unset(hdev, PA_LINK, BDADDR_ANY, + pa_sync = hci_conn_add_unset(hdev, PA_LINK, BDADDR_ANY, 0, HCI_ROLE_SLAVE); if (IS_ERR(pa_sync)) -- cgit v1.2.3 From a106e50be74b0896583f4d010a69f9806e4194f4 Mon Sep 17 00:00:00 2001 From: Luiz Augusto von Dentz Date: Fri, 14 Nov 2025 09:29:28 -0500 Subject: Bluetooth: HCI: Add support for LL Extended Feature Set MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit This adds support for emulating LL Extended Feature Set introduced in 6.0 that adds the following: Commands: - HCI_LE_Read_All_Local_Supported_­Features(0x2087)(Feature:47,1) - HCI_LE_Read_All_Remote_Features(0x2088)(Feature:47,2) Events: - HCI_LE_Read_All_Remote_Features_Complete(0x2b)(Mask bit:42) Signed-off-by: Luiz Augusto von Dentz --- include/net/bluetooth/hci.h | 23 +++++++ include/net/bluetooth/hci_core.h | 5 +- include/net/bluetooth/hci_sync.h | 2 + net/bluetooth/hci_event.c | 125 ++++++++++++++++++++++++++++++--------- net/bluetooth/hci_sync.c | 102 +++++++++++++++++++++++++++++++- 5 files changed, 227 insertions(+), 30 deletions(-) (limited to 'include/net') diff --git a/include/net/bluetooth/hci.h b/include/net/bluetooth/hci.h index d883ad233ebc..a27cd3626b87 100644 --- a/include/net/bluetooth/hci.h +++ b/include/net/bluetooth/hci.h @@ -653,6 +653,7 @@ enum { #define HCI_LE_CIS_PERIPHERAL 0x20 #define HCI_LE_ISO_BROADCASTER 0x40 #define HCI_LE_ISO_SYNC_RECEIVER 0x80 +#define HCI_LE_LL_EXT_FEATURE 0x80 /* Connection modes */ #define HCI_CM_ACTIVE 0x0000 @@ -2255,6 +2256,19 @@ struct hci_cp_le_set_host_feature { __u8 bit_value; } __packed; +#define HCI_OP_LE_READ_ALL_LOCAL_FEATURES 0x2087 +struct hci_rp_le_read_all_local_features { + __u8 status; + __u8 page; + __u8 features[248]; +} __packed; + +#define HCI_OP_LE_READ_ALL_REMOTE_FEATURES 0x2088 +struct hci_cp_le_read_all_remote_features { + __le16 handle; + __u8 pages; +} __packed; + /* ---- HCI Events ---- */ struct hci_ev_status { __u8 status; @@ -2937,6 +2951,15 @@ struct hci_evt_le_big_info_adv_report { __u8 encryption; } __packed; +#define HCI_EVT_LE_ALL_REMOTE_FEATURES_COMPLETE 0x2b +struct hci_evt_le_read_all_remote_features_complete { + __u8 status; + __le16 handle; + __u8 max_pages; + __u8 valid_pages; + __u8 features[248]; +} __packed; + #define HCI_EV_VENDOR 0xff /* Internal events generated by Bluetooth stack */ diff --git a/include/net/bluetooth/hci_core.h b/include/net/bluetooth/hci_core.h index 858b58206e80..4263e71a23ef 100644 --- a/include/net/bluetooth/hci_core.h +++ b/include/net/bluetooth/hci_core.h @@ -378,7 +378,7 @@ struct hci_dev { __u8 minor_class; __u8 max_page; __u8 features[HCI_MAX_PAGES][8]; - __u8 le_features[8]; + __u8 le_features[248]; __u8 le_accept_list_size; __u8 le_resolv_list_size; __u8 le_num_of_adv_sets; @@ -702,6 +702,7 @@ struct hci_conn { __u8 attempt; __u8 dev_class[3]; __u8 features[HCI_MAX_PAGES][8]; + __u8 le_features[248]; __u16 pkt_type; __u16 link_policy; __u8 key_type; @@ -2067,6 +2068,8 @@ void hci_conn_del_sysfs(struct hci_conn *conn); (le_enabled(dev) && past_receiver_capable(dev)) #define past_enabled(dev) \ (past_sender_enabled(dev) || past_receiver_enabled(dev)) +#define ll_ext_feature_capable(dev) \ + ((dev)->le_features[7] & HCI_LE_LL_EXT_FEATURE) #define mws_transport_config_capable(dev) (((dev)->commands[30] & 0x08) && \ (!hci_test_quirk((dev), HCI_QUIRK_BROKEN_MWS_TRANSPORT_CONFIG))) diff --git a/include/net/bluetooth/hci_sync.h b/include/net/bluetooth/hci_sync.h index 3133f40fa9f9..56076bbc981d 100644 --- a/include/net/bluetooth/hci_sync.h +++ b/include/net/bluetooth/hci_sync.h @@ -189,3 +189,5 @@ int hci_le_conn_update_sync(struct hci_dev *hdev, struct hci_conn *conn, int hci_connect_pa_sync(struct hci_dev *hdev, struct hci_conn *conn); int hci_connect_big_sync(struct hci_dev *hdev, struct hci_conn *conn); int hci_past_sync(struct hci_conn *conn, struct hci_conn *le); + +int hci_le_read_remote_features(struct hci_conn *conn); diff --git a/net/bluetooth/hci_event.c b/net/bluetooth/hci_event.c index 7c4ca14f13e5..a9868f17ef40 100644 --- a/net/bluetooth/hci_event.c +++ b/net/bluetooth/hci_event.c @@ -2886,12 +2886,8 @@ static void hci_cs_le_read_remote_features(struct hci_dev *hdev, u8 status) hci_dev_lock(hdev); conn = hci_conn_hash_lookup_handle(hdev, __le16_to_cpu(cp->handle)); - if (conn) { - if (conn->state == BT_CONFIG) { - hci_connect_cfm(conn, status); - hci_conn_drop(conn); - } - } + if (conn && conn->state == BT_CONFIG) + hci_connect_cfm(conn, status); hci_dev_unlock(hdev); } @@ -3915,11 +3911,49 @@ unlock: return rp->status; } +static u8 hci_cc_le_read_all_local_features(struct hci_dev *hdev, void *data, + struct sk_buff *skb) +{ + struct hci_rp_le_read_all_local_features *rp = data; + + bt_dev_dbg(hdev, "status 0x%2.2x", rp->status); + + if (rp->status) + return rp->status; + + memcpy(hdev->le_features, rp->features, 248); + + return rp->status; +} + static void hci_cs_le_create_big(struct hci_dev *hdev, u8 status) { bt_dev_dbg(hdev, "status 0x%2.2x", status); } +static void hci_cs_le_read_all_remote_features(struct hci_dev *hdev, u8 status) +{ + struct hci_cp_le_read_remote_features *cp; + struct hci_conn *conn; + + bt_dev_dbg(hdev, "status 0x%2.2x", status); + + if (!status) + return; + + cp = hci_sent_cmd_data(hdev, HCI_OP_LE_READ_ALL_REMOTE_FEATURES); + if (!cp) + return; + + hci_dev_lock(hdev); + + conn = hci_conn_hash_lookup_handle(hdev, __le16_to_cpu(cp->handle)); + if (conn && conn->state == BT_CONFIG) + hci_connect_cfm(conn, status); + + hci_dev_unlock(hdev); +} + static u8 hci_cc_set_per_adv_param(struct hci_dev *hdev, void *data, struct sk_buff *skb) { @@ -4171,6 +4205,9 @@ static const struct hci_cc { sizeof(struct hci_rp_le_set_cig_params), HCI_MAX_EVENT_SIZE), HCI_CC(HCI_OP_LE_SETUP_ISO_PATH, hci_cc_le_setup_iso_path, sizeof(struct hci_rp_le_setup_iso_path)), + HCI_CC(HCI_OP_LE_READ_ALL_LOCAL_FEATURES, + hci_cc_le_read_all_local_features, + sizeof(struct hci_rp_le_read_all_local_features)), }; static u8 hci_cc_func(struct hci_dev *hdev, const struct hci_cc *cc, @@ -4325,6 +4362,8 @@ static const struct hci_cs { HCI_CS(HCI_OP_LE_EXT_CREATE_CONN, hci_cs_le_ext_create_conn), HCI_CS(HCI_OP_LE_CREATE_CIS, hci_cs_le_create_cis), HCI_CS(HCI_OP_LE_CREATE_BIG, hci_cs_le_create_big), + HCI_CS(HCI_OP_LE_READ_ALL_REMOTE_FEATURES, + hci_cs_le_read_all_remote_features), }; static void hci_cmd_status_evt(struct hci_dev *hdev, void *data, @@ -5645,6 +5684,7 @@ static void le_conn_complete_evt(struct hci_dev *hdev, u8 status, struct hci_conn *conn; struct smp_irk *irk; u8 addr_type; + int err; hci_dev_lock(hdev); @@ -5775,26 +5815,8 @@ static void le_conn_complete_evt(struct hci_dev *hdev, u8 status, hci_debugfs_create_conn(conn); hci_conn_add_sysfs(conn); - /* The remote features procedure is defined for central - * role only. So only in case of an initiated connection - * request the remote features. - * - * If the local controller supports peripheral-initiated features - * exchange, then requesting the remote features in peripheral - * role is possible. Otherwise just transition into the - * connected state without requesting the remote features. - */ - if (conn->out || - (hdev->le_features[0] & HCI_LE_PERIPHERAL_FEATURES)) { - struct hci_cp_le_read_remote_features cp; - - cp.handle = __cpu_to_le16(conn->handle); - - hci_send_cmd(hdev, HCI_OP_LE_READ_REMOTE_FEATURES, - sizeof(cp), &cp); - - hci_conn_hold(conn); - } else { + err = hci_le_read_remote_features(conn); + if (err) { conn->state = BT_CONNECTED; hci_connect_cfm(conn, status); } @@ -6608,7 +6630,6 @@ static void hci_le_remote_feat_complete_evt(struct hci_dev *hdev, void *data, conn->state = BT_CONNECTED; hci_connect_cfm(conn, status); - hci_conn_drop(conn); } } @@ -7186,6 +7207,50 @@ unlock: hci_dev_unlock(hdev); } +static void hci_le_read_all_remote_features_evt(struct hci_dev *hdev, + void *data, struct sk_buff *skb) +{ + struct hci_evt_le_read_all_remote_features_complete *ev = data; + struct hci_conn *conn; + + bt_dev_dbg(hdev, "status 0x%2.2x", ev->status); + + hci_dev_lock(hdev); + + conn = hci_conn_hash_lookup_handle(hdev, __le16_to_cpu(ev->handle)); + if (!conn) + goto unlock; + + if (!ev->status) + memcpy(conn->le_features, ev->features, 248); + + if (conn->state == BT_CONFIG) { + __u8 status; + + /* If the local controller supports peripheral-initiated + * features exchange, but the remote controller does + * not, then it is possible that the error code 0x1a + * for unsupported remote feature gets returned. + * + * In this specific case, allow the connection to + * transition into connected state and mark it as + * successful. + */ + if (!conn->out && + ev->status == HCI_ERROR_UNSUPPORTED_REMOTE_FEATURE && + (hdev->le_features[0] & HCI_LE_PERIPHERAL_FEATURES)) + status = 0x00; + else + status = ev->status; + + conn->state = BT_CONNECTED; + hci_connect_cfm(conn, status); + } + +unlock: + hci_dev_unlock(hdev); +} + #define HCI_LE_EV_VL(_op, _func, _min_len, _max_len) \ [_op] = { \ .func = _func, \ @@ -7291,6 +7356,12 @@ static const struct hci_le_ev { hci_le_big_info_adv_report_evt, sizeof(struct hci_evt_le_big_info_adv_report), HCI_MAX_EVENT_SIZE), + /* [0x2b = HCI_EVT_LE_ALL_REMOTE_FEATURES_COMPLETE] */ + HCI_LE_EV_VL(HCI_EVT_LE_ALL_REMOTE_FEATURES_COMPLETE, + hci_le_read_all_remote_features_evt, + sizeof(struct + hci_evt_le_read_all_remote_features_complete), + HCI_MAX_EVENT_SIZE), }; static void hci_le_meta_evt(struct hci_dev *hdev, void *data, diff --git a/net/bluetooth/hci_sync.c b/net/bluetooth/hci_sync.c index a36d2414a3ca..a9f5b1a68356 100644 --- a/net/bluetooth/hci_sync.c +++ b/net/bluetooth/hci_sync.c @@ -4011,8 +4011,19 @@ static int hci_le_read_buffer_size_sync(struct hci_dev *hdev) /* Read LE Local Supported Features */ static int hci_le_read_local_features_sync(struct hci_dev *hdev) { - return __hci_cmd_sync_status(hdev, HCI_OP_LE_READ_LOCAL_FEATURES, - 0, NULL, HCI_CMD_TIMEOUT); + int err; + + err = __hci_cmd_sync_status(hdev, HCI_OP_LE_READ_LOCAL_FEATURES, + 0, NULL, HCI_CMD_TIMEOUT); + if (err) + return err; + + if (ll_ext_feature_capable(hdev) && hdev->commands[47] & BIT(2)) + return __hci_cmd_sync_status(hdev, + HCI_OP_LE_READ_ALL_LOCAL_FEATURES, + 0, NULL, HCI_CMD_TIMEOUT); + + return err; } /* Read LE Supported States */ @@ -7320,3 +7331,90 @@ int hci_past_sync(struct hci_conn *conn, struct hci_conn *le) return err; } + +static void le_read_features_complete(struct hci_dev *hdev, void *data, int err) +{ + struct hci_conn *conn = data; + + bt_dev_dbg(hdev, "err %d", err); + + if (err == -ECANCELED) + return; + + hci_conn_drop(conn); +} + +static int hci_le_read_all_remote_features_sync(struct hci_dev *hdev, + void *data) +{ + struct hci_conn *conn = data; + struct hci_cp_le_read_all_remote_features cp; + + memset(&cp, 0, sizeof(cp)); + cp.handle = cpu_to_le16(conn->handle); + cp.pages = 10; /* Attempt to read all pages */ + + /* Wait for HCI_EVT_LE_ALL_REMOTE_FEATURES_COMPLETE event otherwise + * hci_conn_drop may run prematurely causing a disconnection. + */ + return __hci_cmd_sync_status_sk(hdev, + HCI_OP_LE_READ_ALL_REMOTE_FEATURES, + sizeof(cp), &cp, + HCI_EVT_LE_ALL_REMOTE_FEATURES_COMPLETE, + HCI_CMD_TIMEOUT, NULL); + + return __hci_cmd_sync_status(hdev, HCI_OP_LE_READ_ALL_REMOTE_FEATURES, + sizeof(cp), &cp, HCI_CMD_TIMEOUT); +} + +static int hci_le_read_remote_features_sync(struct hci_dev *hdev, void *data) +{ + struct hci_conn *conn = data; + struct hci_cp_le_read_remote_features cp; + + if (!hci_conn_valid(hdev, conn)) + return -ECANCELED; + + /* Check if LL Extended Feature Set is supported and + * HCI_OP_LE_READ_ALL_REMOTE_FEATURES is supported then use that to read + * all features. + */ + if (ll_ext_feature_capable(hdev) && hdev->commands[47] & BIT(3)) + return hci_le_read_all_remote_features_sync(hdev, data); + + memset(&cp, 0, sizeof(cp)); + cp.handle = cpu_to_le16(conn->handle); + + /* Wait for HCI_EV_LE_REMOTE_FEAT_COMPLETE event otherwise + * hci_conn_drop may run prematurely causing a disconnection. + */ + return __hci_cmd_sync_status_sk(hdev, HCI_OP_LE_READ_REMOTE_FEATURES, + sizeof(cp), &cp, + HCI_EV_LE_REMOTE_FEAT_COMPLETE, + HCI_CMD_TIMEOUT, NULL); +} + +int hci_le_read_remote_features(struct hci_conn *conn) +{ + struct hci_dev *hdev = conn->hdev; + int err; + + /* The remote features procedure is defined for central + * role only. So only in case of an initiated connection + * request the remote features. + * + * If the local controller supports peripheral-initiated features + * exchange, then requesting the remote features in peripheral + * role is possible. Otherwise just transition into the + * connected state without requesting the remote features. + */ + if (conn->out || (hdev->le_features[0] & HCI_LE_PERIPHERAL_FEATURES)) + err = hci_cmd_sync_queue_once(hdev, + hci_le_read_remote_features_sync, + hci_conn_hold(conn), + le_read_features_complete); + else + err = -EOPNOTSUPP; + + return err; +} -- cgit v1.2.3 From 9bf66036d686b9a67000ba22bd94be13a4ea79ac Mon Sep 17 00:00:00 2001 From: Long Li Date: Wed, 26 Nov 2025 13:45:52 -0800 Subject: net: mana: Handle hardware recovery events when probing the device When MANA is being probed, it's possible that hardware is in recovery mode and the device may get GDMA_EQE_HWC_RESET_REQUEST over HWC in the middle of the probe. Detect such condition and go through the recovery service procedure. Signed-off-by: Long Li Reviewed-by: Haiyang Zhang Link: https://patch.msgid.link/1764193552-9712-1-git-send-email-longli@linux.microsoft.com Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/microsoft/mana/gdma_main.c | 176 +++++++++++++++++++++--- include/net/mana/gdma.h | 12 +- 2 files changed, 170 insertions(+), 18 deletions(-) (limited to 'include/net') diff --git a/drivers/net/ethernet/microsoft/mana/gdma_main.c b/drivers/net/ethernet/microsoft/mana/gdma_main.c index 8fd70b34807a..efb4e412ec7e 100644 --- a/drivers/net/ethernet/microsoft/mana/gdma_main.c +++ b/drivers/net/ethernet/microsoft/mana/gdma_main.c @@ -15,6 +15,20 @@ struct dentry *mana_debugfs_root; +struct mana_dev_recovery { + struct list_head list; + struct pci_dev *pdev; + enum gdma_eqe_type type; +}; + +static struct mana_dev_recovery_work { + struct list_head dev_list; + struct delayed_work work; + + /* Lock for dev_list above */ + spinlock_t lock; +} mana_dev_recovery_work; + static u32 mana_gd_r32(struct gdma_context *g, u64 offset) { return readl(g->bar0_va + offset); @@ -387,6 +401,25 @@ EXPORT_SYMBOL_NS(mana_gd_ring_cq, "NET_MANA"); #define MANA_SERVICE_PERIOD 10 +static void mana_serv_rescan(struct pci_dev *pdev) +{ + struct pci_bus *parent; + + pci_lock_rescan_remove(); + + parent = pdev->bus; + if (!parent) { + dev_err(&pdev->dev, "MANA service: no parent bus\n"); + goto out; + } + + pci_stop_and_remove_bus_device(pdev); + pci_rescan_bus(parent); + +out: + pci_unlock_rescan_remove(); +} + static void mana_serv_fpga(struct pci_dev *pdev) { struct pci_bus *bus, *parent; @@ -419,9 +452,12 @@ static void mana_serv_reset(struct pci_dev *pdev) { struct gdma_context *gc = pci_get_drvdata(pdev); struct hw_channel_context *hwc; + int ret; if (!gc) { - dev_err(&pdev->dev, "MANA service: no GC\n"); + /* Perform PCI rescan on device if GC is not set up */ + dev_err(&pdev->dev, "MANA service: GC not setup, rescanning\n"); + mana_serv_rescan(pdev); return; } @@ -440,9 +476,18 @@ static void mana_serv_reset(struct pci_dev *pdev) msleep(MANA_SERVICE_PERIOD * 1000); - mana_gd_resume(pdev); + ret = mana_gd_resume(pdev); + if (ret == -ETIMEDOUT || ret == -EPROTO) { + /* Perform PCI rescan on device if we failed on HWC */ + dev_err(&pdev->dev, "MANA service: resume failed, rescanning\n"); + mana_serv_rescan(pdev); + goto out; + } - dev_info(&pdev->dev, "MANA reset cycle completed\n"); + if (ret) + dev_info(&pdev->dev, "MANA reset cycle failed err %d\n", ret); + else + dev_info(&pdev->dev, "MANA reset cycle completed\n"); out: gc->in_service = false; @@ -454,18 +499,9 @@ struct mana_serv_work { enum gdma_eqe_type type; }; -static void mana_serv_func(struct work_struct *w) +static void mana_do_service(enum gdma_eqe_type type, struct pci_dev *pdev) { - struct mana_serv_work *mns_wk; - struct pci_dev *pdev; - - mns_wk = container_of(w, struct mana_serv_work, serv_work); - pdev = mns_wk->pdev; - - if (!pdev) - goto out; - - switch (mns_wk->type) { + switch (type) { case GDMA_EQE_HWC_FPGA_RECONFIG: mana_serv_fpga(pdev); break; @@ -475,12 +511,48 @@ static void mana_serv_func(struct work_struct *w) break; default: - dev_err(&pdev->dev, "MANA service: unknown type %d\n", - mns_wk->type); + dev_err(&pdev->dev, "MANA service: unknown type %d\n", type); break; } +} + +static void mana_recovery_delayed_func(struct work_struct *w) +{ + struct mana_dev_recovery_work *work; + struct mana_dev_recovery *dev; + unsigned long flags; + + work = container_of(w, struct mana_dev_recovery_work, work.work); + + spin_lock_irqsave(&work->lock, flags); + + while (!list_empty(&work->dev_list)) { + dev = list_first_entry(&work->dev_list, + struct mana_dev_recovery, list); + list_del(&dev->list); + spin_unlock_irqrestore(&work->lock, flags); + + mana_do_service(dev->type, dev->pdev); + pci_dev_put(dev->pdev); + kfree(dev); + + spin_lock_irqsave(&work->lock, flags); + } + + spin_unlock_irqrestore(&work->lock, flags); +} + +static void mana_serv_func(struct work_struct *w) +{ + struct mana_serv_work *mns_wk; + struct pci_dev *pdev; + + mns_wk = container_of(w, struct mana_serv_work, serv_work); + pdev = mns_wk->pdev; + + if (pdev) + mana_do_service(mns_wk->type, pdev); -out: pci_dev_put(pdev); kfree(mns_wk); module_put(THIS_MODULE); @@ -541,6 +613,17 @@ static void mana_gd_process_eqe(struct gdma_queue *eq) case GDMA_EQE_HWC_RESET_REQUEST: dev_info(gc->dev, "Recv MANA service type:%d\n", type); + if (!test_and_set_bit(GC_PROBE_SUCCEEDED, &gc->flags)) { + /* + * Device is in probe and we received a hardware reset + * event, the probe function will detect that the flag + * has changed and perform service procedure. + */ + dev_info(gc->dev, + "Service is to be processed in probe\n"); + break; + } + if (gc->in_service) { dev_info(gc->dev, "Already in service\n"); break; @@ -1938,8 +2021,19 @@ static int mana_gd_probe(struct pci_dev *pdev, const struct pci_device_id *ent) if (err) goto cleanup_mana; + /* + * If a hardware reset event has occurred over HWC during probe, + * rollback and perform hardware reset procedure. + */ + if (test_and_set_bit(GC_PROBE_SUCCEEDED, &gc->flags)) { + err = -EPROTO; + goto cleanup_mana_rdma; + } + return 0; +cleanup_mana_rdma: + mana_rdma_remove(&gc->mana_ib); cleanup_mana: mana_remove(&gc->mana, false); cleanup_gd: @@ -1963,6 +2057,35 @@ release_region: disable_dev: pci_disable_device(pdev); dev_err(&pdev->dev, "gdma probe failed: err = %d\n", err); + + /* + * Hardware could be in recovery mode and the HWC returns TIMEDOUT or + * EPROTO from mana_gd_setup(), mana_probe() or mana_rdma_probe(), or + * we received a hardware reset event over HWC interrupt. In this case, + * perform the device recovery procedure after MANA_SERVICE_PERIOD + * seconds. + */ + if (err == -ETIMEDOUT || err == -EPROTO) { + struct mana_dev_recovery *dev; + unsigned long flags; + + dev_info(&pdev->dev, "Start MANA recovery mode\n"); + + dev = kzalloc(sizeof(*dev), GFP_KERNEL); + if (!dev) + return err; + + dev->pdev = pci_dev_get(pdev); + dev->type = GDMA_EQE_HWC_RESET_REQUEST; + + spin_lock_irqsave(&mana_dev_recovery_work.lock, flags); + list_add_tail(&dev->list, &mana_dev_recovery_work.dev_list); + spin_unlock_irqrestore(&mana_dev_recovery_work.lock, flags); + + schedule_delayed_work(&mana_dev_recovery_work.work, + secs_to_jiffies(MANA_SERVICE_PERIOD)); + } + return err; } @@ -2067,6 +2190,10 @@ static int __init mana_driver_init(void) { int err; + INIT_LIST_HEAD(&mana_dev_recovery_work.dev_list); + spin_lock_init(&mana_dev_recovery_work.lock); + INIT_DELAYED_WORK(&mana_dev_recovery_work.work, mana_recovery_delayed_func); + mana_debugfs_root = debugfs_create_dir("mana", NULL); err = pci_register_driver(&mana_driver); @@ -2080,6 +2207,21 @@ static int __init mana_driver_init(void) static void __exit mana_driver_exit(void) { + struct mana_dev_recovery *dev; + unsigned long flags; + + disable_delayed_work_sync(&mana_dev_recovery_work.work); + + spin_lock_irqsave(&mana_dev_recovery_work.lock, flags); + while (!list_empty(&mana_dev_recovery_work.dev_list)) { + dev = list_first_entry(&mana_dev_recovery_work.dev_list, + struct mana_dev_recovery, list); + list_del(&dev->list); + pci_dev_put(dev->pdev); + kfree(dev); + } + spin_unlock_irqrestore(&mana_dev_recovery_work.lock, flags); + pci_unregister_driver(&mana_driver); debugfs_remove(mana_debugfs_root); diff --git a/include/net/mana/gdma.h b/include/net/mana/gdma.h index a4cf307859f8..eaa27483f99b 100644 --- a/include/net/mana/gdma.h +++ b/include/net/mana/gdma.h @@ -382,6 +382,10 @@ struct gdma_irq_context { char name[MANA_IRQ_NAME_SZ]; }; +enum gdma_context_flags { + GC_PROBE_SUCCEEDED = 0, +}; + struct gdma_context { struct device *dev; struct dentry *mana_pci_debugfs; @@ -430,6 +434,8 @@ struct gdma_context { u64 pf_cap_flags1; struct workqueue_struct *service_wq; + + unsigned long flags; }; static inline bool mana_gd_is_mana(struct gdma_dev *gd) @@ -600,6 +606,9 @@ enum { /* Driver can send HWC periodically to query stats */ #define GDMA_DRV_CAP_FLAG_1_PERIODIC_STATS_QUERY BIT(21) +/* Driver can handle hardware recovery events during probe */ +#define GDMA_DRV_CAP_FLAG_1_PROBE_RECOVERY BIT(22) + #define GDMA_DRV_CAP_FLAGS1 \ (GDMA_DRV_CAP_FLAG_1_EQ_SHARING_MULTI_VPORT | \ GDMA_DRV_CAP_FLAG_1_NAPI_WKDONE_FIX | \ @@ -611,7 +620,8 @@ enum { GDMA_DRV_CAP_FLAG_1_HANDLE_RECONFIG_EQE | \ GDMA_DRV_CAP_FLAG_1_HW_VPORT_LINK_AWARE | \ GDMA_DRV_CAP_FLAG_1_PERIODIC_STATS_QUERY | \ - GDMA_DRV_CAP_FLAG_1_SKB_LINEARIZE) + GDMA_DRV_CAP_FLAG_1_SKB_LINEARIZE | \ + GDMA_DRV_CAP_FLAG_1_PROBE_RECOVERY) #define GDMA_DRV_CAP_FLAGS2 0 -- cgit v1.2.3 From 0e75bfe340bf05d1586eaf02942438573bda69e3 Mon Sep 17 00:00:00 2001 From: Vladimir Oltean Date: Sun, 30 Nov 2025 15:16:47 +0200 Subject: net: dsa: add simple HSR offload helpers MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit It turns out that HSR offloads are so fine-grained that many DSA switches can do a small part even though they weren't specifically designed for the protocols supported by that driver (HSR and PRP). Specifically NETIF_F_HW_HSR_DUP - it is simple packet duplication on transmit, towards all (aka 2) ports members of the HSR device. For many DSA switches, we know how to duplicate a packet, even though we never typically use that feature. The transmit port mask from the tagging protocol can have multiple bits set, and the switch should send the packet once to every port with a bit set from that mask. Nonetheless, not all tagging protocols are like this, and sometimes the port is a single numeric value rather than a bit mask. For that reason, and also because switches can sometimes change tagging protocols for different ones, we need to make HSR offload helpers opt-in. For devices that can do nothing else HSR-specific, we introduce dsa_port_simple_hsr_join() and dsa_port_simple_hsr_leave(). These functions monitor when two user ports of the same switch are part of the same HSR device, and when that condition is true, they toggle the NETIF_F_HW_HSR_DUP feature flag of both net devices. Normally only dsa_port_simple_hsr_join() and dsa_port_simple_hsr_leave() are needed. The dsa_port_simple_hsr_validate() helper is just to see what kind of configuration could be offloadable using the generic helpers. This is used by switch drivers which are not currently using the right tagging protocol to offload this HSR ring, but could in principle offload it after changing the tagger. Suggested-by: David Yang Cc: "Alvin Šipraga" Cc: Chester A. Unal" Cc: "Clément Léger" Cc: Daniel Golle Cc: DENG Qingfang Cc: Florian Fainelli Cc: George McCollister Cc: Hauke Mehrtens Cc: Jonas Gorski Cc: Kurt Kanzenbach Cc: Linus Walleij Cc: Sean Wang Cc: UNGLinuxDriver@microchip.com Cc: Woojung Huh Signed-off-by: Vladimir Oltean Link: https://patch.msgid.link/20251130131657.65080-6-vladimir.oltean@nxp.com Signed-off-by: Jakub Kicinski --- include/net/dsa.h | 9 ++++++++ net/dsa/dsa.c | 65 +++++++++++++++++++++++++++++++++++++++++++++++++++++++ 2 files changed, 74 insertions(+) (limited to 'include/net') diff --git a/include/net/dsa.h b/include/net/dsa.h index e40cdc12f7f3..cced1a866757 100644 --- a/include/net/dsa.h +++ b/include/net/dsa.h @@ -1322,6 +1322,15 @@ bool dsa_mdb_present_in_other_db(struct dsa_switch *ds, int port, const struct switchdev_obj_port_mdb *mdb, struct dsa_db db); +int dsa_port_simple_hsr_validate(struct dsa_switch *ds, int port, + struct net_device *hsr, + struct netlink_ext_ack *extack); +int dsa_port_simple_hsr_join(struct dsa_switch *ds, int port, + struct net_device *hsr, + struct netlink_ext_ack *extack); +int dsa_port_simple_hsr_leave(struct dsa_switch *ds, int port, + struct net_device *hsr); + /* Keep inline for faster access in hot path */ static inline bool netdev_uses_dsa(const struct net_device *dev) { diff --git a/net/dsa/dsa.c b/net/dsa/dsa.c index 5b01a0e43ebe..a20efabe778f 100644 --- a/net/dsa/dsa.c +++ b/net/dsa/dsa.c @@ -9,6 +9,7 @@ #include #include +#include #include #include #include @@ -1766,6 +1767,70 @@ bool dsa_mdb_present_in_other_db(struct dsa_switch *ds, int port, } EXPORT_SYMBOL_GPL(dsa_mdb_present_in_other_db); +/* Helpers for switches without specific HSR offloads, but which can implement + * NETIF_F_HW_HSR_DUP because their tagger uses dsa_xmit_port_mask() + */ +int dsa_port_simple_hsr_validate(struct dsa_switch *ds, int port, + struct net_device *hsr, + struct netlink_ext_ack *extack) +{ + enum hsr_port_type type; + int err; + + err = hsr_get_port_type(hsr, dsa_to_port(ds, port)->user, &type); + if (err) + return err; + + if (type != HSR_PT_SLAVE_A && type != HSR_PT_SLAVE_B) { + NL_SET_ERR_MSG_MOD(extack, + "Only HSR slave ports can be offloaded"); + return -EOPNOTSUPP; + } + + return 0; +} +EXPORT_SYMBOL_GPL(dsa_port_simple_hsr_validate); + +int dsa_port_simple_hsr_join(struct dsa_switch *ds, int port, + struct net_device *hsr, + struct netlink_ext_ack *extack) +{ + struct dsa_port *dp = dsa_to_port(ds, port), *other_dp; + int err; + + err = dsa_port_simple_hsr_validate(ds, port, hsr, extack); + if (err) + return err; + + dsa_hsr_foreach_port(other_dp, ds, hsr) { + if (other_dp != dp) { + dp->user->features |= NETIF_F_HW_HSR_DUP; + other_dp->user->features |= NETIF_F_HW_HSR_DUP; + break; + } + } + + return 0; +} +EXPORT_SYMBOL_GPL(dsa_port_simple_hsr_join); + +int dsa_port_simple_hsr_leave(struct dsa_switch *ds, int port, + struct net_device *hsr) +{ + struct dsa_port *dp = dsa_to_port(ds, port), *other_dp; + + dsa_hsr_foreach_port(other_dp, ds, hsr) { + if (other_dp != dp) { + dp->user->features &= ~NETIF_F_HW_HSR_DUP; + other_dp->user->features &= ~NETIF_F_HW_HSR_DUP; + break; + } + } + + return 0; +} +EXPORT_SYMBOL_GPL(dsa_port_simple_hsr_leave); + static const struct dsa_stubs __dsa_stubs = { .conduit_hwtstamp_validate = __dsa_conduit_hwtstamp_validate, }; -- cgit v1.2.3