From b357a364c57c940ddb932224542494363df37378 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Thu, 23 Apr 2015 18:03:44 -0700 Subject: inet: fix possible panic in reqsk_queue_unlink() [ 3897.923145] BUG: unable to handle kernel NULL pointer dereference at 0000000000000080 [ 3897.931025] IP: [] reqsk_timer_handler+0x1a6/0x243 There is a race when reqsk_timer_handler() and tcp_check_req() call inet_csk_reqsk_queue_unlink() on the same req at the same time. Before commit fa76ce7328b2 ("inet: get rid of central tcp/dccp listener timer"), listener spinlock was held and race could not happen. To solve this bug, we change reqsk_queue_unlink() to not assume req must be found, and we return a status, to conditionally release a refcount on the request sock. This also means tcp_check_req() in non fastopen case might or not consume req refcount, so tcp_v6_hnd_req() & tcp_v4_hnd_req() have to properly handle this. (Same remark for dccp_check_req() and its callers) inet_csk_reqsk_queue_drop() is now too big to be inlined, as it is called 4 times in tcp and 3 times in dccp. Fixes: fa76ce7328b2 ("inet: get rid of central tcp/dccp listener timer") Signed-off-by: Eric Dumazet Reported-by: Yuchung Cheng Signed-off-by: David S. Miller --- include/net/inet_connection_sock.h | 20 +------------------- include/net/request_sock.h | 18 ------------------ 2 files changed, 1 insertion(+), 37 deletions(-) (limited to 'include/net') diff --git a/include/net/inet_connection_sock.h b/include/net/inet_connection_sock.h index 7b5887cd1172..48a815823587 100644 --- a/include/net/inet_connection_sock.h +++ b/include/net/inet_connection_sock.h @@ -279,12 +279,6 @@ static inline void inet_csk_reqsk_queue_add(struct sock *sk, void inet_csk_reqsk_queue_hash_add(struct sock *sk, struct request_sock *req, unsigned long timeout); -static inline void inet_csk_reqsk_queue_removed(struct sock *sk, - struct request_sock *req) -{ - reqsk_queue_removed(&inet_csk(sk)->icsk_accept_queue, req); -} - static inline void inet_csk_reqsk_queue_added(struct sock *sk, const unsigned long timeout) { @@ -306,19 +300,7 @@ static inline int inet_csk_reqsk_queue_is_full(const struct sock *sk) return reqsk_queue_is_full(&inet_csk(sk)->icsk_accept_queue); } -static inline void inet_csk_reqsk_queue_unlink(struct sock *sk, - struct request_sock *req) -{ - reqsk_queue_unlink(&inet_csk(sk)->icsk_accept_queue, req); -} - -static inline void inet_csk_reqsk_queue_drop(struct sock *sk, - struct request_sock *req) -{ - inet_csk_reqsk_queue_unlink(sk, req); - inet_csk_reqsk_queue_removed(sk, req); - reqsk_put(req); -} +void inet_csk_reqsk_queue_drop(struct sock *sk, struct request_sock *req); void inet_csk_destroy_sock(struct sock *sk); void inet_csk_prepare_forced_close(struct sock *sk); diff --git a/include/net/request_sock.h b/include/net/request_sock.h index fe41f3ceb008..9f4265ce8892 100644 --- a/include/net/request_sock.h +++ b/include/net/request_sock.h @@ -212,24 +212,6 @@ static inline int reqsk_queue_empty(struct request_sock_queue *queue) return queue->rskq_accept_head == NULL; } -static inline void reqsk_queue_unlink(struct request_sock_queue *queue, - struct request_sock *req) -{ - struct listen_sock *lopt = queue->listen_opt; - struct request_sock **prev; - - spin_lock(&queue->syn_wait_lock); - - prev = &lopt->syn_table[req->rsk_hash]; - while (*prev != req) - prev = &(*prev)->dl_next; - *prev = req->dl_next; - - spin_unlock(&queue->syn_wait_lock); - if (del_timer(&req->rsk_timer)) - reqsk_put(req); -} - static inline void reqsk_queue_add(struct request_sock_queue *queue, struct request_sock *req, struct sock *parent, -- cgit v1.2.3 From 73b5a6f2a7a1cb78ccdec3900afc8657e11bc6bf Mon Sep 17 00:00:00 2001 From: Matan Barak Date: Sun, 26 Apr 2015 15:55:57 +0300 Subject: net/bonding: Make DRV macros private The bonding modules currently defines four macros with general names that pollute the global namespace: DRV_VERSION DRV_RELDATE DRV_NAME DRV_DESCRIPTION Fixing that by defining a private bonding_priv.h header files which includes those defines. Signed-off-by: Matan Barak Signed-off-by: Or Gerlitz Signed-off-by: David S. Miller --- include/net/bonding.h | 7 ------- 1 file changed, 7 deletions(-) (limited to 'include/net') diff --git a/include/net/bonding.h b/include/net/bonding.h index fda6feeb6c1f..78ed135e9dea 100644 --- a/include/net/bonding.h +++ b/include/net/bonding.h @@ -30,13 +30,6 @@ #include #include -#define DRV_VERSION "3.7.1" -#define DRV_RELDATE "April 27, 2011" -#define DRV_NAME "bonding" -#define DRV_DESCRIPTION "Ethernet Channel Bonding Driver" - -#define bond_version DRV_DESCRIPTION ": v" DRV_VERSION " (" DRV_RELDATE ")\n" - #define BOND_MAX_ARP_TARGETS 16 #define BOND_DEFAULT_MIIMON 100 -- cgit v1.2.3 From 0df48c26d8418c5c9fba63fac15b660d70ca2f1c Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Tue, 28 Apr 2015 15:28:17 -0700 Subject: tcp: add tcpi_bytes_acked to tcp_info This patch tracks total number of bytes acked for a TCP socket. This is the sum of all changes done to tp->snd_una, and allows for precise tracking of delivered data. RFC4898 named this : tcpEStatsAppHCThruOctetsAcked This is a 64bit field, and can be fetched both from TCP_INFO getsockopt() if one has a handle on a TCP socket, or from inet_diag netlink facility (iproute2/ss patch will follow) Note that tp->bytes_acked was placed near tp->snd_una for best data locality and minimal performance impact. Signed-off-by: Eric Dumazet Acked-by: Yuchung Cheng Cc: Matt Mathis Cc: Eric Salo Cc: Martin Lau Cc: Chris Rapier Signed-off-by: David S. Miller --- include/net/tcp.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'include/net') diff --git a/include/net/tcp.h b/include/net/tcp.h index 051dc5c2802d..dd7b4ea6a10c 100644 --- a/include/net/tcp.h +++ b/include/net/tcp.h @@ -576,7 +576,7 @@ static inline int tcp_bound_to_half_wnd(struct tcp_sock *tp, int pktsize) } /* tcp.c */ -void tcp_get_info(const struct sock *, struct tcp_info *); +void tcp_get_info(struct sock *, struct tcp_info *); /* Read 'sendfile()'-style from a TCP socket */ typedef int (*sk_read_actor_t)(read_descriptor_t *, struct sk_buff *, -- cgit v1.2.3 From 64f40ff5bbdb1b679fb3c4dbc8230d6517d2b8dc Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Tue, 28 Apr 2015 16:23:48 -0700 Subject: tcp: prepare CC get_info() access from getsockopt() We would like that optional info provided by Congestion Control modules using netlink can also be read using getsockopt() This patch changes get_info() to put this information in a buffer, instead of skb, like tcp_get_info(), so that following patch can reuse this common infrastructure. Signed-off-by: Eric Dumazet Cc: Yuchung Cheng Cc: Neal Cardwell Acked-by: Neal Cardwell Acked-by: Daniel Borkmann Acked-by: Yuchung Cheng Signed-off-by: David S. Miller --- include/net/tcp.h | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) (limited to 'include/net') diff --git a/include/net/tcp.h b/include/net/tcp.h index dd7b4ea6a10c..6d204f3f9df8 100644 --- a/include/net/tcp.h +++ b/include/net/tcp.h @@ -804,6 +804,8 @@ enum tcp_ca_ack_event_flags { /* Requires ECN/ECT set on all packets */ #define TCP_CONG_NEEDS_ECN 0x2 +union tcp_cc_info; + struct tcp_congestion_ops { struct list_head list; u32 key; @@ -829,7 +831,8 @@ struct tcp_congestion_ops { /* hook for packet ack accounting (optional) */ void (*pkts_acked)(struct sock *sk, u32 num_acked, s32 rtt_us); /* get info for inet_diag (optional) */ - int (*get_info)(struct sock *sk, u32 ext, struct sk_buff *skb); + size_t (*get_info)(struct sock *sk, u32 ext, int *attr, + union tcp_cc_info *info); char name[TCP_CA_NAME_MAX]; struct module *owner; -- cgit v1.2.3 From 42fb23e2f57accbed69804e6d72ea3a88b8a84c4 Mon Sep 17 00:00:00 2001 From: Varka Bhadram Date: Thu, 30 Apr 2015 17:44:52 +0200 Subject: mac802154: add description to mac802154 APIs This patch adds the proper description to the mac802154 core APIs. Signed-off-by: Varka Bhadram Acked-by: Alexander Aring Signed-off-by: Alexander Aring Signed-off-by: Marcel Holtmann --- include/net/mac802154.h | 94 +++++++++++++++++++++++++++++++++++++++++++++++-- 1 file changed, 92 insertions(+), 2 deletions(-) (limited to 'include/net') diff --git a/include/net/mac802154.h b/include/net/mac802154.h index e18e7fd43f47..7df28a4c23f9 100644 --- a/include/net/mac802154.h +++ b/include/net/mac802154.h @@ -247,19 +247,109 @@ static inline void ieee802154_le64_to_be64(void *be64_dst, const void *le64_src) __put_unaligned_memmove64(swab64p(le64_src), be64_dst); } -/* Basic interface to register ieee802154 device */ +/** + * ieee802154_alloc_hw - Allocate a new hardware device + * + * This must be called once for each hardware device. The returned pointer + * must be used to refer to this device when calling other functions. + * mac802154 allocates a private data area for the driver pointed to by + * @priv in &struct ieee802154_hw, the size of this area is given as + * @priv_data_len. + * + * @priv_data_len: length of private data + * @ops: callbacks for this device + * + * Return: A pointer to the new hardware device, or %NULL on error. + */ struct ieee802154_hw * ieee802154_alloc_hw(size_t priv_data_len, const struct ieee802154_ops *ops); + +/** + * ieee802154_free_hw - free hardware descriptor + * + * This function frees everything that was allocated, including the + * private data for the driver. You must call ieee802154_unregister_hw() + * before calling this function. + * + * @hw: the hardware to free + */ void ieee802154_free_hw(struct ieee802154_hw *hw); + +/** + * ieee802154_register_hw - Register hardware device + * + * You must call this function before any other functions in + * mac802154. Note that before a hardware can be registered, you + * need to fill the contained wpan_phy's information. + * + * @hw: the device to register as returned by ieee802154_alloc_hw() + * + * Return: 0 on success. An error code otherwise. + */ int ieee802154_register_hw(struct ieee802154_hw *hw); + +/** + * ieee802154_unregister_hw - Unregister a hardware device + * + * This function instructs mac802154 to free allocated resources + * and unregister netdevices from the networking subsystem. + * + * @hw: the hardware to unregister + */ void ieee802154_unregister_hw(struct ieee802154_hw *hw); +/** + * ieee802154_rx - receive frame + * + * Use this function to hand received frames to mac802154. The receive + * buffer in @skb must start with an IEEE 802.15.4 header. In case of a + * paged @skb is used, the driver is recommended to put the ieee802154 + * header of the frame on the linear part of the @skb to avoid memory + * allocation and/or memcpy by the stack. + * + * This function may not be called in IRQ context. Calls to this function + * for a single hardware must be synchronized against each other. + * + * @hw: the hardware this frame came in on + * @skb: the buffer to receive, owned by mac802154 after this call + */ void ieee802154_rx(struct ieee802154_hw *hw, struct sk_buff *skb); + +/** + * ieee802154_rx_irqsafe - receive frame + * + * Like ieee802154_rx() but can be called in IRQ context + * (internally defers to a tasklet.) + * + * @hw: the hardware this frame came in on + * @skb: the buffer to receive, owned by mac802154 after this call + * @lqi: link quality indicator + */ void ieee802154_rx_irqsafe(struct ieee802154_hw *hw, struct sk_buff *skb, u8 lqi); - +/** + * ieee802154_wake_queue - wake ieee802154 queue + * @hw: pointer as obtained from ieee802154_alloc_hw(). + * + * Drivers should use this function instead of netif_wake_queue. + */ void ieee802154_wake_queue(struct ieee802154_hw *hw); + +/** + * ieee802154_stop_queue - stop ieee802154 queue + * @hw: pointer as obtained from ieee802154_alloc_hw(). + * + * Drivers should use this function instead of netif_stop_queue. + */ void ieee802154_stop_queue(struct ieee802154_hw *hw); + +/** + * ieee802154_xmit_complete - frame transmission complete + * + * @hw: pointer as obtained from ieee802154_alloc_hw(). + * @skb: buffer for transmission + * @ifs_handling: indicate interframe space handling + */ void ieee802154_xmit_complete(struct ieee802154_hw *hw, struct sk_buff *skb, bool ifs_handling); -- cgit v1.2.3 From 5b4a10390460cccf17a9fac739e153d68cf25ef5 Mon Sep 17 00:00:00 2001 From: Varka Bhadram Date: Thu, 30 Apr 2015 17:44:57 +0200 Subject: cfg802154: pass name_assign_type to rdev_add_virtual_intf() This code is based on commit 6bab2e19c5ffd ("cfg80211: pass name_assign_type to rdev_add_virtual_intf()") This will expose in sysfs whether the ifname of a IEEE-802.15.4 device is set by userspace or generated by the kernel. We are using two types of name_assign_types o NET_NAME_ENUM: Default interface name provided by kernel o NET_NAME_USER: Interface name provided by user. Signed-off-by: Varka Bhadram Signed-off-by: Alexander Aring Signed-off-by: Marcel Holtmann --- include/net/cfg802154.h | 2 ++ 1 file changed, 2 insertions(+) (limited to 'include/net') diff --git a/include/net/cfg802154.h b/include/net/cfg802154.h index eeda67652766..6ea16c84293b 100644 --- a/include/net/cfg802154.h +++ b/include/net/cfg802154.h @@ -30,11 +30,13 @@ struct wpan_phy_cca; struct cfg802154_ops { struct net_device * (*add_virtual_intf_deprecated)(struct wpan_phy *wpan_phy, const char *name, + unsigned char name_assign_type, int type); void (*del_virtual_intf_deprecated)(struct wpan_phy *wpan_phy, struct net_device *dev); int (*add_virtual_intf)(struct wpan_phy *wpan_phy, const char *name, + unsigned char name_assign_type, enum nl802154_iftype type, __le64 extended_addr); int (*del_virtual_intf)(struct wpan_phy *wpan_phy, -- cgit v1.2.3 From a5d28090405038ca1f40c13f38d6d4285456efee Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Thu, 30 Apr 2015 09:40:40 -0700 Subject: codel: fix maxpacket/mtu confusion Under presence of TSO/GSO/GRO packets, codel at low rates can be quite useless. In following example, not a single packet was ever dropped, while average delay in codel queue is ~100 ms ! qdisc codel 0: parent 1:12 limit 16000p target 5.0ms interval 100.0ms Sent 134376498 bytes 88797 pkt (dropped 0, overlimits 0 requeues 0) backlog 13626b 3p requeues 0 count 0 lastcount 0 ldelay 96.9ms drop_next 0us maxpacket 9084 ecn_mark 0 drop_overlimit 0 This comes from a confusion of what should be the minimal backlog. It is pretty clear it is not 64KB or whatever max GSO packet ever reached the qdisc. codel intent was to use MTU of the device. After the fix, we finally drop some packets, and rtt/cwnd of my single TCP flow are meeting our expectations. qdisc codel 0: parent 1:12 limit 16000p target 5.0ms interval 100.0ms Sent 102798497 bytes 67912 pkt (dropped 1365, overlimits 0 requeues 0) backlog 6056b 3p requeues 0 count 1 lastcount 1 ldelay 36.3ms drop_next 0us maxpacket 10598 ecn_mark 0 drop_overlimit 0 Signed-off-by: Eric Dumazet Cc: Kathleen Nichols Cc: Dave Taht Cc: Van Jacobson Signed-off-by: David S. Miller --- include/net/codel.h | 10 +++++++--- 1 file changed, 7 insertions(+), 3 deletions(-) (limited to 'include/net') diff --git a/include/net/codel.h b/include/net/codel.h index aeee28081245..1e18005f7f65 100644 --- a/include/net/codel.h +++ b/include/net/codel.h @@ -120,11 +120,13 @@ static inline u32 codel_time_to_us(codel_time_t val) * struct codel_params - contains codel parameters * @target: target queue size (in time units) * @interval: width of moving time window + * @mtu: device mtu, or minimal queue backlog in bytes. * @ecn: is Explicit Congestion Notification enabled */ struct codel_params { codel_time_t target; codel_time_t interval; + u32 mtu; bool ecn; }; @@ -166,10 +168,12 @@ struct codel_stats { u32 ecn_mark; }; -static void codel_params_init(struct codel_params *params) +static void codel_params_init(struct codel_params *params, + const struct Qdisc *sch) { params->interval = MS2TIME(100); params->target = MS2TIME(5); + params->mtu = psched_mtu(qdisc_dev(sch)); params->ecn = false; } @@ -180,7 +184,7 @@ static void codel_vars_init(struct codel_vars *vars) static void codel_stats_init(struct codel_stats *stats) { - stats->maxpacket = 256; + stats->maxpacket = 0; } /* @@ -234,7 +238,7 @@ static bool codel_should_drop(const struct sk_buff *skb, stats->maxpacket = qdisc_pkt_len(skb); if (codel_time_before(vars->ldelay, params->target) || - sch->qstats.backlog <= stats->maxpacket) { + sch->qstats.backlog <= params->mtu) { /* went below - stay below for at least interval */ vars->first_above_time = 0; return false; -- cgit v1.2.3 From ff419b3f95ab7a97c5f72876b53f12a249dacc2a Mon Sep 17 00:00:00 2001 From: Randy Dunlap Date: Sun, 26 Apr 2015 20:13:34 -0700 Subject: mac80211: fix 90 kernel-doc warnings Eliminate 90 of these warnings: Warning(..//include/net/mac80211.h:1682): No description found for parameter 'drv_priv[0]' Signed-off-by: Randy Dunlap Signed-off-by: Johannes Berg --- include/net/mac80211.h | 2 ++ 1 file changed, 2 insertions(+) (limited to 'include/net') diff --git a/include/net/mac80211.h b/include/net/mac80211.h index b4bef1152c05..8e3668b44c29 100644 --- a/include/net/mac80211.h +++ b/include/net/mac80211.h @@ -1666,6 +1666,8 @@ struct ieee80211_tx_control { * @sta: station table entry, %NULL for per-vif queue * @tid: the TID for this queue (unused for per-vif queue) * @ac: the AC for this queue + * @drv_priv: data area for driver use, will always be aligned to + * sizeof(void *). * * The driver can obtain packets from this queue by calling * ieee80211_tx_dequeue(). -- cgit v1.2.3 From b5d721d761fccf491f92eefc611e318561683d0a Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Mon, 18 May 2015 17:06:14 -0700 Subject: inet: properly align icsk_ca_priv tcp_illinois and upcoming tcp_cdg require 64bit alignment of icsk_ca_priv x86 does not care, but other architectures might. Fixes: 05cbc0db03e82 ("ipv4: Create probe timer for tcp PMTU as per RFC4821") Signed-off-by: Eric Dumazet Cc: Fan Du Acked-by: Fan Du Signed-off-by: David S. Miller --- include/net/inet_connection_sock.h | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) (limited to 'include/net') diff --git a/include/net/inet_connection_sock.h b/include/net/inet_connection_sock.h index 48a815823587..497bc14cdb85 100644 --- a/include/net/inet_connection_sock.h +++ b/include/net/inet_connection_sock.h @@ -129,9 +129,10 @@ struct inet_connection_sock { u32 probe_timestamp; } icsk_mtup; - u32 icsk_ca_priv[16]; u32 icsk_user_timeout; -#define ICSK_CA_PRIV_SIZE (16 * sizeof(u32)) + + u64 icsk_ca_priv[64 / sizeof(u64)]; +#define ICSK_CA_PRIV_SIZE (8 * sizeof(u64)) }; #define ICSK_TIME_RETRANS 1 /* Retransmit timer */ -- cgit v1.2.3 From 9302d7bb0c5cd46be5706859301f18c137b2439f Mon Sep 17 00:00:00 2001 From: Jason Gunthorpe Date: Tue, 26 May 2015 17:30:17 -0600 Subject: sctp: Fix mangled IPv4 addresses on a IPv6 listening socket sctp_v4_map_v6 was subtly writing and reading from members of a union in a way the clobbered data it needed to read before it read it. Zeroing the v6 flowinfo overwrites the v4 sin_addr with 0, meaning that every place that calls sctp_v4_map_v6 gets ::ffff:0.0.0.0 as the result. Reorder things to guarantee correct behaviour no matter what the union layout is. This impacts user space clients that open an IPv6 SCTP socket and receive IPv4 connections. Prior to 299ee user space would see a sockaddr with AF_INET and a correct address, after 299ee the sockaddr is AF_INET6, but the address is wrong. Fixes: 299ee123e198 (sctp: Fixup v4mapped behaviour to comply with Sock API) Signed-off-by: Jason Gunthorpe Acked-by: Daniel Borkmann Acked-by: Neil Horman Signed-off-by: David S. Miller --- include/net/sctp/sctp.h | 7 +++++-- 1 file changed, 5 insertions(+), 2 deletions(-) (limited to 'include/net') diff --git a/include/net/sctp/sctp.h b/include/net/sctp/sctp.h index c56a438c3a1e..ce13cf20f625 100644 --- a/include/net/sctp/sctp.h +++ b/include/net/sctp/sctp.h @@ -574,11 +574,14 @@ static inline void sctp_v6_map_v4(union sctp_addr *addr) /* Map v4 address to v4-mapped v6 address */ static inline void sctp_v4_map_v6(union sctp_addr *addr) { + __be16 port; + + port = addr->v4.sin_port; + addr->v6.sin6_addr.s6_addr32[3] = addr->v4.sin_addr.s_addr; + addr->v6.sin6_port = port; addr->v6.sin6_family = AF_INET6; addr->v6.sin6_flowinfo = 0; addr->v6.sin6_scope_id = 0; - addr->v6.sin6_port = addr->v4.sin_port; - addr->v6.sin6_addr.s6_addr32[3] = addr->v4.sin_addr.s_addr; addr->v6.sin6_addr.s6_addr32[0] = 0; addr->v6.sin6_addr.s6_addr32[1] = 0; addr->v6.sin6_addr.s6_addr32[2] = htonl(0x0000ffff); -- cgit v1.2.3