From c0d358e705424d19443f06178c6f87b44c0b814f Mon Sep 17 00:00:00 2001 From: Greg Kroah-Hartman Date: Wed, 16 May 2018 17:44:18 +0200 Subject: [PATCH] 4.9-stable patches added patches: 8139too-use-disable_irq_nosync-in-rtl8139_poll_controller.patch bonding-do-not-allow-rlb-updates-to-invalid-mac.patch bonding-send-learning-packets-for-vlans-on-slave.patch bridge-check-iface-upper-dev-when-setting-master-via-ioctl.patch dccp-fix-tasklet-usage.patch ipv4-fix-memory-leaks-in-udp_sendmsg-ping_v4_sendmsg.patch llc-better-deal-with-too-small-mtu.patch net-ethernet-sun-niu-set-correct-packet-size-in-skb.patch net-ethernet-ti-cpsw-fix-packet-leaking-in-dual_mac-mode.patch net-mlx4_en-verify-coalescing-parameters-are-in-range.patch net-mlx5-avoid-cleaning-flow-steering-table-twice-during-error-flow.patch net-mlx5-e-switch-include-vf-rdma-stats-in-vport-statistics.patch net-support-compat-64-bit-time-in-s-g-etsockopt.patch net_sched-fq-take-care-of-throttled-flows-before-reuse.patch openvswitch-don-t-swap-table-in-nlattr_set-after-ovs_attr_nested-is-found.patch qmi_wwan-do-not-steal-interfaces-from-class-drivers.patch r8169-fix-powering-up-rtl8168h.patch sctp-delay-the-authentication-for-the-duplicated-cookie-echo-chunk.patch sctp-fix-the-issue-that-the-cookie-ack-with-auth-can-t-get-processed.patch sctp-handle-two-v4-addrs-comparison-in-sctp_inet6_cmp_addr.patch sctp-remove-sctp_chunk_put-from-fail_mark-err-path-in-sctp_ulpevent_make_rcvmsg.patch sctp-use-the-old-asoc-when-making-the-cookie-ack-chunk-in-dupcook_d.patch tcp-ignore-fast-open-on-repair-mode.patch tcp_bbr-fix-to-zero-idle_restart-only-upon-s-acked-data.patch tg3-fix-vunmap-bug_on-triggered-from-tg3_free_consistent.patch --- queue-4.4/series | 20 ++ ...rq_nosync-in-rtl8139_poll_controller.patch | 32 +++ ...not-allow-rlb-updates-to-invalid-mac.patch | 31 +++ ...-learning-packets-for-vlans-on-slave.patch | 87 ++++++ ...er-dev-when-setting-master-via-ioctl.patch | 42 +++ queue-4.9/dccp-fix-tasklet-usage.patch | 114 ++++++++ ...leaks-in-udp_sendmsg-ping_v4_sendmsg.patch | 77 ++++++ .../llc-better-deal-with-too-small-mtu.patch | 80 ++++++ ...n-niu-set-correct-packet-size-in-skb.patch | 58 ++++ ...-fix-packet-leaking-in-dual_mac-mode.patch | 46 ++++ ...y-coalescing-parameters-are-in-range.patch | 79 ++++++ ...eering-table-twice-during-error-flow.patch | 72 +++++ ...de-vf-rdma-stats-in-vport-statistics.patch | 61 +++++ ...-compat-64-bit-time-in-s-g-etsockopt.patch | 140 ++++++++++ ...care-of-throttled-flows-before-reuse.patch | 96 +++++++ ...r_set-after-ovs_attr_nested-is-found.patch | 110 ++++++++ ...-steal-interfaces-from-class-drivers.patch | 49 ++++ .../r8169-fix-powering-up-rtl8168h.patch | 49 ++++ ...for-the-duplicated-cookie-echo-chunk.patch | 249 ++++++++++++++++++ ...ie-ack-with-auth-can-t-get-processed.patch | 42 +++ ...rs-comparison-in-sctp_inet6_cmp_addr.patch | 47 ++++ ...rr-path-in-sctp_ulpevent_make_rcvmsg.patch | 40 +++ ...ng-the-cookie-ack-chunk-in-dupcook_d.patch | 40 +++ queue-4.9/series | 25 ++ .../tcp-ignore-fast-open-on-repair-mode.patch | 45 ++++ ...-idle_restart-only-upon-s-acked-data.patch | 48 ++++ ...n-triggered-from-tg3_free_consistent.patch | 57 ++++ 27 files changed, 1836 insertions(+) create mode 100644 queue-4.4/series create mode 100644 queue-4.9/8139too-use-disable_irq_nosync-in-rtl8139_poll_controller.patch create mode 100644 queue-4.9/bonding-do-not-allow-rlb-updates-to-invalid-mac.patch create mode 100644 queue-4.9/bonding-send-learning-packets-for-vlans-on-slave.patch create mode 100644 queue-4.9/bridge-check-iface-upper-dev-when-setting-master-via-ioctl.patch create mode 100644 queue-4.9/dccp-fix-tasklet-usage.patch create mode 100644 queue-4.9/ipv4-fix-memory-leaks-in-udp_sendmsg-ping_v4_sendmsg.patch create mode 100644 queue-4.9/llc-better-deal-with-too-small-mtu.patch create mode 100644 queue-4.9/net-ethernet-sun-niu-set-correct-packet-size-in-skb.patch create mode 100644 queue-4.9/net-ethernet-ti-cpsw-fix-packet-leaking-in-dual_mac-mode.patch create mode 100644 queue-4.9/net-mlx4_en-verify-coalescing-parameters-are-in-range.patch create mode 100644 queue-4.9/net-mlx5-avoid-cleaning-flow-steering-table-twice-during-error-flow.patch create mode 100644 queue-4.9/net-mlx5-e-switch-include-vf-rdma-stats-in-vport-statistics.patch create mode 100644 queue-4.9/net-support-compat-64-bit-time-in-s-g-etsockopt.patch create mode 100644 queue-4.9/net_sched-fq-take-care-of-throttled-flows-before-reuse.patch create mode 100644 queue-4.9/openvswitch-don-t-swap-table-in-nlattr_set-after-ovs_attr_nested-is-found.patch create mode 100644 queue-4.9/qmi_wwan-do-not-steal-interfaces-from-class-drivers.patch create mode 100644 queue-4.9/r8169-fix-powering-up-rtl8168h.patch create mode 100644 queue-4.9/sctp-delay-the-authentication-for-the-duplicated-cookie-echo-chunk.patch create mode 100644 queue-4.9/sctp-fix-the-issue-that-the-cookie-ack-with-auth-can-t-get-processed.patch create mode 100644 queue-4.9/sctp-handle-two-v4-addrs-comparison-in-sctp_inet6_cmp_addr.patch create mode 100644 queue-4.9/sctp-remove-sctp_chunk_put-from-fail_mark-err-path-in-sctp_ulpevent_make_rcvmsg.patch create mode 100644 queue-4.9/sctp-use-the-old-asoc-when-making-the-cookie-ack-chunk-in-dupcook_d.patch create mode 100644 queue-4.9/series create mode 100644 queue-4.9/tcp-ignore-fast-open-on-repair-mode.patch create mode 100644 queue-4.9/tcp_bbr-fix-to-zero-idle_restart-only-upon-s-acked-data.patch create mode 100644 queue-4.9/tg3-fix-vunmap-bug_on-triggered-from-tg3_free_consistent.patch diff --git a/queue-4.4/series b/queue-4.4/series new file mode 100644 index 00000000000..3131a27faf7 --- /dev/null +++ b/queue-4.4/series @@ -0,0 +1,20 @@ +8139too-use-disable_irq_nosync-in-rtl8139_poll_controller.patch +bridge-check-iface-upper-dev-when-setting-master-via-ioctl.patch +dccp-fix-tasklet-usage.patch +ipv4-fix-memory-leaks-in-udp_sendmsg-ping_v4_sendmsg.patch +llc-better-deal-with-too-small-mtu.patch +net-ethernet-sun-niu-set-correct-packet-size-in-skb.patch +net-mlx4_en-verify-coalescing-parameters-are-in-range.patch +net_sched-fq-take-care-of-throttled-flows-before-reuse.patch +net-support-compat-64-bit-time-in-s-g-etsockopt.patch +openvswitch-don-t-swap-table-in-nlattr_set-after-ovs_attr_nested-is-found.patch +qmi_wwan-do-not-steal-interfaces-from-class-drivers.patch +r8169-fix-powering-up-rtl8168h.patch +sctp-handle-two-v4-addrs-comparison-in-sctp_inet6_cmp_addr.patch +sctp-use-the-old-asoc-when-making-the-cookie-ack-chunk-in-dupcook_d.patch +tg3-fix-vunmap-bug_on-triggered-from-tg3_free_consistent.patch +bonding-do-not-allow-rlb-updates-to-invalid-mac.patch +tcp-ignore-fast-open-on-repair-mode.patch +sctp-fix-the-issue-that-the-cookie-ack-with-auth-can-t-get-processed.patch +sctp-delay-the-authentication-for-the-duplicated-cookie-echo-chunk.patch +net-ethernet-ti-cpsw-fix-packet-leaking-in-dual_mac-mode.patch diff --git a/queue-4.9/8139too-use-disable_irq_nosync-in-rtl8139_poll_controller.patch b/queue-4.9/8139too-use-disable_irq_nosync-in-rtl8139_poll_controller.patch new file mode 100644 index 00000000000..ef246a48ee7 --- /dev/null +++ b/queue-4.9/8139too-use-disable_irq_nosync-in-rtl8139_poll_controller.patch @@ -0,0 +1,32 @@ +From foo@baz Wed May 16 12:08:47 CEST 2018 +From: Ingo Molnar +Date: Wed, 2 May 2018 13:30:57 +0200 +Subject: 8139too: Use disable_irq_nosync() in rtl8139_poll_controller() + +From: Ingo Molnar + +[ Upstream commit af3e0fcf78879f718c5f73df0814951bd7057d34 ] + +Use disable_irq_nosync() instead of disable_irq() as this might be +called in atomic context with netpoll. + +Signed-off-by: Ingo Molnar +Signed-off-by: Thomas Gleixner +Signed-off-by: Sebastian Andrzej Siewior +Signed-off-by: David S. Miller +Signed-off-by: Greg Kroah-Hartman +--- + drivers/net/ethernet/realtek/8139too.c | 2 +- + 1 file changed, 1 insertion(+), 1 deletion(-) + +--- a/drivers/net/ethernet/realtek/8139too.c ++++ b/drivers/net/ethernet/realtek/8139too.c +@@ -2233,7 +2233,7 @@ static void rtl8139_poll_controller(stru + struct rtl8139_private *tp = netdev_priv(dev); + const int irq = tp->pci_dev->irq; + +- disable_irq(irq); ++ disable_irq_nosync(irq); + rtl8139_interrupt(irq, dev); + enable_irq(irq); + } diff --git a/queue-4.9/bonding-do-not-allow-rlb-updates-to-invalid-mac.patch b/queue-4.9/bonding-do-not-allow-rlb-updates-to-invalid-mac.patch new file mode 100644 index 00000000000..aa606a7a31c --- /dev/null +++ b/queue-4.9/bonding-do-not-allow-rlb-updates-to-invalid-mac.patch @@ -0,0 +1,31 @@ +From foo@baz Wed May 16 12:08:47 CEST 2018 +From: Debabrata Banerjee +Date: Wed, 9 May 2018 19:32:10 -0400 +Subject: bonding: do not allow rlb updates to invalid mac + +From: Debabrata Banerjee + +[ Upstream commit 4fa8667ca3989ce14cf66301fa251544fbddbdd0 ] + +Make sure multicast, broadcast, and zero mac's cannot be the output of rlb +updates, which should all be directed arps. Receive load balancing will be +collapsed if any of these happen, as the switch will broadcast. + +Signed-off-by: Debabrata Banerjee +Signed-off-by: David S. Miller +Signed-off-by: Greg Kroah-Hartman +--- + drivers/net/bonding/bond_alb.c | 2 +- + 1 file changed, 1 insertion(+), 1 deletion(-) + +--- a/drivers/net/bonding/bond_alb.c ++++ b/drivers/net/bonding/bond_alb.c +@@ -450,7 +450,7 @@ static void rlb_update_client(struct rlb + { + int i; + +- if (!client_info->slave) ++ if (!client_info->slave || !is_valid_ether_addr(client_info->mac_dst)) + return; + + for (i = 0; i < RLB_ARP_BURST_SIZE; i++) { diff --git a/queue-4.9/bonding-send-learning-packets-for-vlans-on-slave.patch b/queue-4.9/bonding-send-learning-packets-for-vlans-on-slave.patch new file mode 100644 index 00000000000..4aa66975459 --- /dev/null +++ b/queue-4.9/bonding-send-learning-packets-for-vlans-on-slave.patch @@ -0,0 +1,87 @@ +From foo@baz Wed May 16 12:08:47 CEST 2018 +From: Debabrata Banerjee +Date: Wed, 9 May 2018 19:32:11 -0400 +Subject: bonding: send learning packets for vlans on slave + +From: Debabrata Banerjee + +[ Upstream commit 21706ee8a47d3ede7fdae0be6d7c0a0e31a83229 ] + +There was a regression at some point from the intended functionality of +commit f60c3704e87d ("bonding: Fix alb mode to only use first level +vlans.") + +Given the return value vlan_get_encap_level() we need to store the nest +level of the bond device, and then compare the vlan's encap level to +this. Without this, this check always fails and learning packets are +never sent. + +In addition, this same commit caused a regression in the behavior of +balance_alb, which requires learning packets be sent for all interfaces +using the slave's mac in order to load balance properly. For vlan's +that have not set a user mac, we can send after checking one bit. +Otherwise we need send the set mac, albeit defeating rx load balancing +for that vlan. + +Signed-off-by: Debabrata Banerjee +Signed-off-by: David S. Miller +Signed-off-by: Greg Kroah-Hartman +--- + drivers/net/bonding/bond_alb.c | 13 ++++++++----- + drivers/net/bonding/bond_main.c | 2 ++ + include/net/bonding.h | 1 + + 3 files changed, 11 insertions(+), 5 deletions(-) + +--- a/drivers/net/bonding/bond_alb.c ++++ b/drivers/net/bonding/bond_alb.c +@@ -944,6 +944,10 @@ static void alb_send_lp_vid(struct slave + skb->priority = TC_PRIO_CONTROL; + skb->dev = slave->dev; + ++ netdev_dbg(slave->bond->dev, ++ "Send learning packet: dev %s mac %pM vlan %d\n", ++ slave->dev->name, mac_addr, vid); ++ + if (vid) + __vlan_hwaccel_put_tag(skb, vlan_proto, vid); + +@@ -966,14 +970,13 @@ static void alb_send_learning_packets(st + */ + rcu_read_lock(); + netdev_for_each_all_upper_dev_rcu(bond->dev, upper, iter) { +- if (is_vlan_dev(upper) && vlan_get_encap_level(upper) == 0) { +- if (strict_match && +- ether_addr_equal_64bits(mac_addr, +- upper->dev_addr)) { ++ if (is_vlan_dev(upper) && ++ bond->nest_level == vlan_get_encap_level(upper) - 1) { ++ if (upper->addr_assign_type == NET_ADDR_STOLEN) { + alb_send_lp_vid(slave, mac_addr, + vlan_dev_vlan_proto(upper), + vlan_dev_vlan_id(upper)); +- } else if (!strict_match) { ++ } else { + alb_send_lp_vid(slave, upper->dev_addr, + vlan_dev_vlan_proto(upper), + vlan_dev_vlan_id(upper)); +--- a/drivers/net/bonding/bond_main.c ++++ b/drivers/net/bonding/bond_main.c +@@ -1732,6 +1732,8 @@ int bond_enslave(struct net_device *bond + if (bond_mode_uses_xmit_hash(bond)) + bond_update_slave_arr(bond, NULL); + ++ bond->nest_level = dev_get_nest_level(bond_dev); ++ + netdev_info(bond_dev, "Enslaving %s as %s interface with %s link\n", + slave_dev->name, + bond_is_active_slave(new_slave) ? "an active" : "a backup", +--- a/include/net/bonding.h ++++ b/include/net/bonding.h +@@ -197,6 +197,7 @@ struct bonding { + struct slave __rcu *primary_slave; + struct bond_up_slave __rcu *slave_arr; /* Array of usable slaves */ + bool force_primary; ++ u32 nest_level; + s32 slave_cnt; /* never change this value outside the attach/detach wrappers */ + int (*recv_probe)(const struct sk_buff *, struct bonding *, + struct slave *); diff --git a/queue-4.9/bridge-check-iface-upper-dev-when-setting-master-via-ioctl.patch b/queue-4.9/bridge-check-iface-upper-dev-when-setting-master-via-ioctl.patch new file mode 100644 index 00000000000..3469396efd3 --- /dev/null +++ b/queue-4.9/bridge-check-iface-upper-dev-when-setting-master-via-ioctl.patch @@ -0,0 +1,42 @@ +From foo@baz Wed May 16 12:08:47 CEST 2018 +From: Hangbin Liu +Date: Fri, 27 Apr 2018 20:59:24 +0800 +Subject: bridge: check iface upper dev when setting master via ioctl + +From: Hangbin Liu + +[ Upstream commit e8238fc2bd7b4c3c7554fa2df067e796610212fc ] + +When we set a bond slave's master to bridge via ioctl, we only check +the IFF_BRIDGE_PORT flag. Although we will find the slave's real master +at netdev_master_upper_dev_link() later, it already does some settings +and allocates some resources. It would be better to return as early +as possible. + +v1 -> v2: +use netdev_master_upper_dev_get() instead of netdev_has_any_upper_dev() +to check if we have a master, because not all upper devs are masters, +e.g. vlan device. + +Reported-by: syzbot+de73361ee4971b6e6f75@syzkaller.appspotmail.com +Signed-off-by: Hangbin Liu +Acked-by: Nikolay Aleksandrov +Signed-off-by: David S. Miller +Signed-off-by: Greg Kroah-Hartman +--- + net/bridge/br_if.c | 4 ++-- + 1 file changed, 2 insertions(+), 2 deletions(-) + +--- a/net/bridge/br_if.c ++++ b/net/bridge/br_if.c +@@ -504,8 +504,8 @@ int br_add_if(struct net_bridge *br, str + if (dev->netdev_ops->ndo_start_xmit == br_dev_xmit) + return -ELOOP; + +- /* Device is already being bridged */ +- if (br_port_exists(dev)) ++ /* Device has master upper dev */ ++ if (netdev_master_upper_dev_get(dev)) + return -EBUSY; + + /* No bridging devices that dislike that (e.g. wireless) */ diff --git a/queue-4.9/dccp-fix-tasklet-usage.patch b/queue-4.9/dccp-fix-tasklet-usage.patch new file mode 100644 index 00000000000..735ecada310 --- /dev/null +++ b/queue-4.9/dccp-fix-tasklet-usage.patch @@ -0,0 +1,114 @@ +From foo@baz Wed May 16 12:08:47 CEST 2018 +From: Eric Dumazet +Date: Thu, 3 May 2018 09:39:20 -0700 +Subject: dccp: fix tasklet usage + +From: Eric Dumazet + +[ Upstream commit a8d7aa17bbc970971ccdf71988ea19230ab368b1 ] + +syzbot reported a crash in tasklet_action_common() caused by dccp. + +dccp needs to make sure socket wont disappear before tasklet handler +has completed. + +This patch takes a reference on the socket when arming the tasklet, +and moves the sock_put() from dccp_write_xmit_timer() to dccp_write_xmitlet() + +kernel BUG at kernel/softirq.c:514! +invalid opcode: 0000 [#1] SMP KASAN +Dumping ftrace buffer: + (ftrace buffer empty) +Modules linked in: +CPU: 1 PID: 17 Comm: ksoftirqd/1 Not tainted 4.17.0-rc3+ #30 +Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011 +RIP: 0010:tasklet_action_common.isra.19+0x6db/0x700 kernel/softirq.c:515 +RSP: 0018:ffff8801d9b3faf8 EFLAGS: 00010246 +dccp_close: ABORT with 65423 bytes unread +RAX: 1ffff1003b367f6b RBX: ffff8801daf1f3f0 RCX: 0000000000000000 +RDX: ffff8801cf895498 RSI: 0000000000000004 RDI: 0000000000000000 +RBP: ffff8801d9b3fc40 R08: ffffed0039f12a95 R09: ffffed0039f12a94 +dccp_close: ABORT with 65423 bytes unread +R10: ffffed0039f12a94 R11: ffff8801cf8954a3 R12: 0000000000000000 +R13: ffff8801d9b3fc18 R14: dffffc0000000000 R15: ffff8801cf895490 +FS: 0000000000000000(0000) GS:ffff8801daf00000(0000) knlGS:0000000000000000 +CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 +CR2: 0000001b2bc28000 CR3: 00000001a08a9000 CR4: 00000000001406e0 +DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 +DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400 +Call Trace: + tasklet_action+0x1d/0x20 kernel/softirq.c:533 + __do_softirq+0x2e0/0xaf5 kernel/softirq.c:285 +dccp_close: ABORT with 65423 bytes unread + run_ksoftirqd+0x86/0x100 kernel/softirq.c:646 + smpboot_thread_fn+0x417/0x870 kernel/smpboot.c:164 + kthread+0x345/0x410 kernel/kthread.c:238 + ret_from_fork+0x3a/0x50 arch/x86/entry/entry_64.S:412 +Code: 48 8b 85 e8 fe ff ff 48 8b 95 f0 fe ff ff e9 94 fb ff ff 48 89 95 f0 fe ff ff e8 81 53 6e 00 48 8b 95 f0 fe ff ff e9 62 fb ff ff <0f> 0b 48 89 cf 48 89 8d e8 fe ff ff e8 64 53 6e 00 48 8b 8d e8 +RIP: tasklet_action_common.isra.19+0x6db/0x700 kernel/softirq.c:515 RSP: ffff8801d9b3faf8 + +Fixes: dc841e30eaea ("dccp: Extend CCID packet dequeueing interface") +Signed-off-by: Eric Dumazet +Reported-by: syzbot +Cc: Gerrit Renker +Cc: dccp@vger.kernel.org +Signed-off-by: David S. Miller +Signed-off-by: Greg Kroah-Hartman +--- + net/dccp/ccids/ccid2.c | 14 ++++++++++++-- + net/dccp/timer.c | 2 +- + 2 files changed, 13 insertions(+), 3 deletions(-) + +--- a/net/dccp/ccids/ccid2.c ++++ b/net/dccp/ccids/ccid2.c +@@ -126,6 +126,16 @@ static void ccid2_change_l_seq_window(st + DCCPF_SEQ_WMAX)); + } + ++static void dccp_tasklet_schedule(struct sock *sk) ++{ ++ struct tasklet_struct *t = &dccp_sk(sk)->dccps_xmitlet; ++ ++ if (!test_and_set_bit(TASKLET_STATE_SCHED, &t->state)) { ++ sock_hold(sk); ++ __tasklet_schedule(t); ++ } ++} ++ + static void ccid2_hc_tx_rto_expire(unsigned long data) + { + struct sock *sk = (struct sock *)data; +@@ -166,7 +176,7 @@ static void ccid2_hc_tx_rto_expire(unsig + + /* if we were blocked before, we may now send cwnd=1 packet */ + if (sender_was_blocked) +- tasklet_schedule(&dccp_sk(sk)->dccps_xmitlet); ++ dccp_tasklet_schedule(sk); + /* restart backed-off timer */ + sk_reset_timer(sk, &hc->tx_rtotimer, jiffies + hc->tx_rto); + out: +@@ -706,7 +716,7 @@ static void ccid2_hc_tx_packet_recv(stru + done: + /* check if incoming Acks allow pending packets to be sent */ + if (sender_was_blocked && !ccid2_cwnd_network_limited(hc)) +- tasklet_schedule(&dccp_sk(sk)->dccps_xmitlet); ++ dccp_tasklet_schedule(sk); + dccp_ackvec_parsed_cleanup(&hc->tx_av_chunks); + } + +--- a/net/dccp/timer.c ++++ b/net/dccp/timer.c +@@ -230,12 +230,12 @@ static void dccp_write_xmitlet(unsigned + else + dccp_write_xmit(sk); + bh_unlock_sock(sk); ++ sock_put(sk); + } + + static void dccp_write_xmit_timer(unsigned long data) + { + dccp_write_xmitlet(data); +- sock_put((struct sock *)data); + } + + void dccp_init_xmit_timers(struct sock *sk) diff --git a/queue-4.9/ipv4-fix-memory-leaks-in-udp_sendmsg-ping_v4_sendmsg.patch b/queue-4.9/ipv4-fix-memory-leaks-in-udp_sendmsg-ping_v4_sendmsg.patch new file mode 100644 index 00000000000..818ca195a50 --- /dev/null +++ b/queue-4.9/ipv4-fix-memory-leaks-in-udp_sendmsg-ping_v4_sendmsg.patch @@ -0,0 +1,77 @@ +From foo@baz Wed May 16 12:08:47 CEST 2018 +From: Andrey Ignatov +Date: Thu, 10 May 2018 10:59:34 -0700 +Subject: ipv4: fix memory leaks in udp_sendmsg, ping_v4_sendmsg + +From: Andrey Ignatov + +[ Upstream commit 1b97013bfb11d66f041de691de6f0fec748ce016 ] + +Fix more memory leaks in ip_cmsg_send() callers. Part of them were fixed +earlier in 919483096bfe. + +* udp_sendmsg one was there since the beginning when linux sources were + first added to git; +* ping_v4_sendmsg one was copy/pasted in c319b4d76b9e. + +Whenever return happens in udp_sendmsg() or ping_v4_sendmsg() IP options +have to be freed if they were allocated previously. + +Add label so that future callers (if any) can use it instead of kfree() +before return that is easy to forget. + +Fixes: c319b4d76b9e (net: ipv4: add IPPROTO_ICMP socket kind) +Signed-off-by: Andrey Ignatov +Signed-off-by: David S. Miller +Signed-off-by: Greg Kroah-Hartman +--- + net/ipv4/ping.c | 7 +++++-- + net/ipv4/udp.c | 7 +++++-- + 2 files changed, 10 insertions(+), 4 deletions(-) + +--- a/net/ipv4/ping.c ++++ b/net/ipv4/ping.c +@@ -775,8 +775,10 @@ static int ping_v4_sendmsg(struct sock * + ipc.addr = faddr = daddr; + + if (ipc.opt && ipc.opt->opt.srr) { +- if (!daddr) +- return -EINVAL; ++ if (!daddr) { ++ err = -EINVAL; ++ goto out_free; ++ } + faddr = ipc.opt->opt.faddr; + } + tos = get_rttos(&ipc, inet); +@@ -841,6 +843,7 @@ back_from_confirm: + + out: + ip_rt_put(rt); ++out_free: + if (free) + kfree(ipc.opt); + if (!err) { +--- a/net/ipv4/udp.c ++++ b/net/ipv4/udp.c +@@ -982,8 +982,10 @@ int udp_sendmsg(struct sock *sk, struct + sock_tx_timestamp(sk, ipc.sockc.tsflags, &ipc.tx_flags); + + if (ipc.opt && ipc.opt->opt.srr) { +- if (!daddr) +- return -EINVAL; ++ if (!daddr) { ++ err = -EINVAL; ++ goto out_free; ++ } + faddr = ipc.opt->opt.faddr; + connected = 0; + } +@@ -1090,6 +1092,7 @@ do_append_data: + + out: + ip_rt_put(rt); ++out_free: + if (free) + kfree(ipc.opt); + if (!err) diff --git a/queue-4.9/llc-better-deal-with-too-small-mtu.patch b/queue-4.9/llc-better-deal-with-too-small-mtu.patch new file mode 100644 index 00000000000..d54a1e4e4d2 --- /dev/null +++ b/queue-4.9/llc-better-deal-with-too-small-mtu.patch @@ -0,0 +1,80 @@ +From foo@baz Wed May 16 12:08:47 CEST 2018 +From: Eric Dumazet +Date: Mon, 7 May 2018 09:02:25 -0700 +Subject: llc: better deal with too small mtu + +From: Eric Dumazet + +[ Upstream commit 2c5d5b13c6eb79f5677e206b8aad59b3a2097f60 ] + +syzbot loves to set very small mtu on devices, since it brings joy. +We must make llc_ui_sendmsg() fool proof. + +usercopy: Kernel memory overwrite attempt detected to wrapped address (offset 0, size 18446612139802320068)! + +kernel BUG at mm/usercopy.c:100! +invalid opcode: 0000 [#1] SMP KASAN +Dumping ftrace buffer: + (ftrace buffer empty) +Modules linked in: +CPU: 0 PID: 17464 Comm: syz-executor1 Not tainted 4.17.0-rc3+ #36 +Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011 +RIP: 0010:usercopy_abort+0xbb/0xbd mm/usercopy.c:88 +RSP: 0018:ffff8801868bf800 EFLAGS: 00010282 +RAX: 000000000000006c RBX: ffffffff87d2fb00 RCX: 0000000000000000 +RDX: 000000000000006c RSI: ffffffff81610731 RDI: ffffed0030d17ef6 +RBP: ffff8801868bf858 R08: ffff88018daa4200 R09: ffffed003b5c4fb0 +R10: ffffed003b5c4fb0 R11: ffff8801dae27d87 R12: ffffffff87d2f8e0 +R13: ffffffff87d2f7a0 R14: ffffffff87d2f7a0 R15: ffffffff87d2f7a0 +FS: 00007f56a14ac700(0000) GS:ffff8801dae00000(0000) knlGS:0000000000000000 +CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 +CR2: 0000001b2bc21000 CR3: 00000001abeb1000 CR4: 00000000001426f0 +DR0: 0000000020000000 DR1: 0000000000000000 DR2: 0000000000000000 +DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000030602 +Call Trace: + check_bogus_address mm/usercopy.c:153 [inline] + __check_object_size+0x5d9/0x5d9 mm/usercopy.c:256 + check_object_size include/linux/thread_info.h:108 [inline] + check_copy_size include/linux/thread_info.h:139 [inline] + copy_from_iter_full include/linux/uio.h:121 [inline] + memcpy_from_msg include/linux/skbuff.h:3305 [inline] + llc_ui_sendmsg+0x4b1/0x1530 net/llc/af_llc.c:941 + sock_sendmsg_nosec net/socket.c:629 [inline] + sock_sendmsg+0xd5/0x120 net/socket.c:639 + __sys_sendto+0x3d7/0x670 net/socket.c:1789 + __do_sys_sendto net/socket.c:1801 [inline] + __se_sys_sendto net/socket.c:1797 [inline] + __x64_sys_sendto+0xe1/0x1a0 net/socket.c:1797 + do_syscall_64+0x1b1/0x800 arch/x86/entry/common.c:287 + entry_SYSCALL_64_after_hwframe+0x49/0xbe +RIP: 0033:0x455979 +RSP: 002b:00007f56a14abc68 EFLAGS: 00000246 ORIG_RAX: 000000000000002c +RAX: ffffffffffffffda RBX: 00007f56a14ac6d4 RCX: 0000000000455979 +RDX: 0000000000000000 RSI: 0000000020000000 RDI: 0000000000000018 +RBP: 000000000072bea0 R08: 00000000200012c0 R09: 0000000000000010 +R10: 0000000000000000 R11: 0000000000000246 R12: 00000000ffffffff +R13: 0000000000000548 R14: 00000000006fbf60 R15: 0000000000000000 +Code: 55 c0 e8 c0 55 bb ff ff 75 c8 48 8b 55 c0 4d 89 f9 ff 75 d0 4d 89 e8 48 89 d9 4c 89 e6 41 56 48 c7 c7 80 fa d2 87 e8 a0 0b a3 ff <0f> 0b e8 95 55 bb ff e8 c0 a8 f7 ff 8b 95 14 ff ff ff 4d 89 e8 +RIP: usercopy_abort+0xbb/0xbd mm/usercopy.c:88 RSP: ffff8801868bf800 + +Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2") +Signed-off-by: Eric Dumazet +Reported-by: syzbot +Signed-off-by: David S. Miller +Signed-off-by: Greg Kroah-Hartman +--- + net/llc/af_llc.c | 3 +++ + 1 file changed, 3 insertions(+) + +--- a/net/llc/af_llc.c ++++ b/net/llc/af_llc.c +@@ -926,6 +926,9 @@ static int llc_ui_sendmsg(struct socket + if (size > llc->dev->mtu) + size = llc->dev->mtu; + copied = size - hdrlen; ++ rc = -EINVAL; ++ if (copied < 0) ++ goto release; + release_sock(sk); + skb = sock_alloc_send_skb(sk, size, noblock, &rc); + lock_sock(sk); diff --git a/queue-4.9/net-ethernet-sun-niu-set-correct-packet-size-in-skb.patch b/queue-4.9/net-ethernet-sun-niu-set-correct-packet-size-in-skb.patch new file mode 100644 index 00000000000..5a091ab2ce8 --- /dev/null +++ b/queue-4.9/net-ethernet-sun-niu-set-correct-packet-size-in-skb.patch @@ -0,0 +1,58 @@ +From foo@baz Wed May 16 12:08:47 CEST 2018 +From: Rob Taglang +Date: Thu, 3 May 2018 17:13:06 -0400 +Subject: net: ethernet: sun: niu set correct packet size in skb + +From: Rob Taglang + +[ Upstream commit 14224923c3600bae2ac4dcae3bf0c3d4dc2812be ] + +Currently, skb->len and skb->data_len are set to the page size, not +the packet size. This causes the frame check sequence to not be +located at the "end" of the packet resulting in ethernet frame check +errors. The driver does work currently, but stricter kernel facing +networking solutions like OpenVSwitch will drop these packets as +invalid. + +These changes set the packet size correctly so that these errors no +longer occur. The length does not include the frame check sequence, so +that subtraction was removed. + +Tested on Oracle/SUN Multithreaded 10-Gigabit Ethernet Network +Controller [108e:abcd] and validated in wireshark. + +Signed-off-by: Rob Taglang +Signed-off-by: David S. Miller +Signed-off-by: Greg Kroah-Hartman +--- + drivers/net/ethernet/sun/niu.c | 5 ++--- + 1 file changed, 2 insertions(+), 3 deletions(-) + +--- a/drivers/net/ethernet/sun/niu.c ++++ b/drivers/net/ethernet/sun/niu.c +@@ -3442,7 +3442,7 @@ static int niu_process_rx_pkt(struct nap + + len = (val & RCR_ENTRY_L2_LEN) >> + RCR_ENTRY_L2_LEN_SHIFT; +- len -= ETH_FCS_LEN; ++ append_size = len + ETH_HLEN + ETH_FCS_LEN; + + addr = (val & RCR_ENTRY_PKT_BUF_ADDR) << + RCR_ENTRY_PKT_BUF_ADDR_SHIFT; +@@ -3452,7 +3452,6 @@ static int niu_process_rx_pkt(struct nap + RCR_ENTRY_PKTBUFSZ_SHIFT]; + + off = addr & ~PAGE_MASK; +- append_size = rcr_size; + if (num_rcr == 1) { + int ptype; + +@@ -3465,7 +3464,7 @@ static int niu_process_rx_pkt(struct nap + else + skb_checksum_none_assert(skb); + } else if (!(val & RCR_ENTRY_MULTI)) +- append_size = len - skb->len; ++ append_size = append_size - skb->len; + + niu_rx_skb_append(skb, page, off, append_size, rcr_size); + if ((page->index + rp->rbr_block_size) - rcr_size == addr) { diff --git a/queue-4.9/net-ethernet-ti-cpsw-fix-packet-leaking-in-dual_mac-mode.patch b/queue-4.9/net-ethernet-ti-cpsw-fix-packet-leaking-in-dual_mac-mode.patch new file mode 100644 index 00000000000..ebf1afc1f57 --- /dev/null +++ b/queue-4.9/net-ethernet-ti-cpsw-fix-packet-leaking-in-dual_mac-mode.patch @@ -0,0 +1,46 @@ +From foo@baz Wed May 16 12:08:47 CEST 2018 +From: Grygorii Strashko +Date: Tue, 1 May 2018 12:41:22 -0500 +Subject: net: ethernet: ti: cpsw: fix packet leaking in dual_mac mode + +From: Grygorii Strashko + +[ Upstream commit 5e5add172ea81152d518b161ec5706503ad3d799 ] + +In dual_mac mode packets arrived on one port should not be forwarded by +switch hw to another port. Only Linux Host can forward packets between +ports. The below test case (reported in [1]) shows that packet arrived on +one port can be leaked to anoter (reproducible with dual port evms): + - connect port 1 (eth0) to linux Host 0 and run tcpdump or Wireshark + - connect port 2 (eth1) to linux Host 1 with vlan 1 configured + - ping from Host 1 through vlan 1 interface. +ARP packets will be seen on Host 0. + +Issue happens because dual_mac mode is implemnted using two vlans: 1 (Port +1+Port 0) and 2 (Port 2+Port 0), so there are vlan records created for for +each vlan. By default, the ALE will find valid vlan record in its table +when vlan 1 tagged packet arrived on Port 2 and so forwards packet to all +ports which are vlan 1 members (like Port. + +To avoid such behaviorr the ALE VLAN ID Ingress Check need to be enabled +for each external CPSW port (ALE_PORTCTLn.VID_INGRESS_CHECK) so ALE will +drop ingress packets if Rx port is not VLAN member. + +Signed-off-by: Grygorii Strashko +Signed-off-by: David S. Miller +Signed-off-by: Greg Kroah-Hartman +--- + drivers/net/ethernet/ti/cpsw.c | 2 ++ + 1 file changed, 2 insertions(+) + +--- a/drivers/net/ethernet/ti/cpsw.c ++++ b/drivers/net/ethernet/ti/cpsw.c +@@ -1141,6 +1141,8 @@ static inline void cpsw_add_dual_emac_de + cpsw_ale_add_ucast(cpsw->ale, priv->mac_addr, + HOST_PORT_NUM, ALE_VLAN | + ALE_SECURE, slave->port_vlan); ++ cpsw_ale_control_set(cpsw->ale, slave_port, ++ ALE_PORT_DROP_UNKNOWN_VLAN, 1); + } + + static void soft_reset_slave(struct cpsw_slave *slave) diff --git a/queue-4.9/net-mlx4_en-verify-coalescing-parameters-are-in-range.patch b/queue-4.9/net-mlx4_en-verify-coalescing-parameters-are-in-range.patch new file mode 100644 index 00000000000..6278b127d60 --- /dev/null +++ b/queue-4.9/net-mlx4_en-verify-coalescing-parameters-are-in-range.patch @@ -0,0 +1,79 @@ +From foo@baz Wed May 16 12:08:47 CEST 2018 +From: Moshe Shemesh +Date: Wed, 9 May 2018 18:35:13 +0300 +Subject: net/mlx4_en: Verify coalescing parameters are in range + +From: Moshe Shemesh + +[ Upstream commit 6ad4e91c6d796b38a7f0e724db1de28eeb122bad ] + +Add check of coalescing parameters received through ethtool are within +range of values supported by the HW. +Driver gets the coalescing rx/tx-usecs and rx/tx-frames as set by the +users through ethtool. The ethtool support up to 32 bit value for each. +However, mlx4 modify cq limits the coalescing time parameter and +coalescing frames parameters to 16 bits. +Return out of range error if user tries to set these parameters to +higher values. +Change type of sample-interval and adaptive_rx_coal parameters in mlx4 +driver to u32 as the ethtool holds them as u32 and these parameters are +not limited due to mlx4 HW. + +Fixes: c27a02cd94d6 ('mlx4_en: Add driver for Mellanox ConnectX 10GbE NIC') +Signed-off-by: Moshe Shemesh +Signed-off-by: Tariq Toukan +Signed-off-by: David S. Miller +Signed-off-by: Greg Kroah-Hartman +--- + drivers/net/ethernet/mellanox/mlx4/en_ethtool.c | 16 ++++++++++++++++ + drivers/net/ethernet/mellanox/mlx4/mlx4_en.h | 7 +++++-- + 2 files changed, 21 insertions(+), 2 deletions(-) + +--- a/drivers/net/ethernet/mellanox/mlx4/en_ethtool.c ++++ b/drivers/net/ethernet/mellanox/mlx4/en_ethtool.c +@@ -970,6 +970,22 @@ static int mlx4_en_set_coalesce(struct n + if (!coal->tx_max_coalesced_frames_irq) + return -EINVAL; + ++ if (coal->tx_coalesce_usecs > MLX4_EN_MAX_COAL_TIME || ++ coal->rx_coalesce_usecs > MLX4_EN_MAX_COAL_TIME || ++ coal->rx_coalesce_usecs_low > MLX4_EN_MAX_COAL_TIME || ++ coal->rx_coalesce_usecs_high > MLX4_EN_MAX_COAL_TIME) { ++ netdev_info(dev, "%s: maximum coalesce time supported is %d usecs\n", ++ __func__, MLX4_EN_MAX_COAL_TIME); ++ return -ERANGE; ++ } ++ ++ if (coal->tx_max_coalesced_frames > MLX4_EN_MAX_COAL_PKTS || ++ coal->rx_max_coalesced_frames > MLX4_EN_MAX_COAL_PKTS) { ++ netdev_info(dev, "%s: maximum coalesced frames supported is %d\n", ++ __func__, MLX4_EN_MAX_COAL_PKTS); ++ return -ERANGE; ++ } ++ + priv->rx_frames = (coal->rx_max_coalesced_frames == + MLX4_EN_AUTO_CONF) ? + MLX4_EN_RX_COAL_TARGET : +--- a/drivers/net/ethernet/mellanox/mlx4/mlx4_en.h ++++ b/drivers/net/ethernet/mellanox/mlx4/mlx4_en.h +@@ -141,6 +141,9 @@ enum { + #define MLX4_EN_TX_COAL_PKTS 16 + #define MLX4_EN_TX_COAL_TIME 0x10 + ++#define MLX4_EN_MAX_COAL_PKTS U16_MAX ++#define MLX4_EN_MAX_COAL_TIME U16_MAX ++ + #define MLX4_EN_RX_RATE_LOW 400000 + #define MLX4_EN_RX_COAL_TIME_LOW 0 + #define MLX4_EN_RX_RATE_HIGH 450000 +@@ -543,8 +546,8 @@ struct mlx4_en_priv { + u16 rx_usecs_low; + u32 pkt_rate_high; + u16 rx_usecs_high; +- u16 sample_interval; +- u16 adaptive_rx_coal; ++ u32 sample_interval; ++ u32 adaptive_rx_coal; + u32 msg_enable; + u32 loopback_ok; + u32 validate_loopback; diff --git a/queue-4.9/net-mlx5-avoid-cleaning-flow-steering-table-twice-during-error-flow.patch b/queue-4.9/net-mlx5-avoid-cleaning-flow-steering-table-twice-during-error-flow.patch new file mode 100644 index 00000000000..732f892f778 --- /dev/null +++ b/queue-4.9/net-mlx5-avoid-cleaning-flow-steering-table-twice-during-error-flow.patch @@ -0,0 +1,72 @@ +From foo@baz Wed May 16 12:08:47 CEST 2018 +From: Talat Batheesh +Date: Sun, 15 Apr 2018 11:26:19 +0300 +Subject: net/mlx5: Avoid cleaning flow steering table twice during error flow + +From: Talat Batheesh + +[ Upstream commit 9c26f5f89d01ca21560c6b8a8e4054c271cc3a9c ] + +When we fail to initialize the RX root namespace, we need +to clean only that and not the entire flow steering. + +Currently the code may try to clean the flow steering twice +on error witch leads to null pointer deference. +Make sure we clean correctly. + +Fixes: fba53f7b5719 ("net/mlx5: Introduce mlx5_flow_steering structure") +Signed-off-by: Talat Batheesh +Reviewed-by: Mark Bloch +Signed-off-by: Saeed Mahameed +Signed-off-by: Greg Kroah-Hartman +--- + drivers/net/ethernet/mellanox/mlx5/core/fs_core.c | 21 +++++++++++++-------- + 1 file changed, 13 insertions(+), 8 deletions(-) + +--- a/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c ++++ b/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c +@@ -153,6 +153,7 @@ static void del_rule(struct fs_node *nod + static void del_flow_table(struct fs_node *node); + static void del_flow_group(struct fs_node *node); + static void del_fte(struct fs_node *node); ++static void cleanup_root_ns(struct mlx5_flow_root_namespace *root_ns); + + static void tree_init_node(struct fs_node *node, + unsigned int refcount, +@@ -1690,24 +1691,28 @@ static int create_anchor_flow_table(stru + + static int init_root_ns(struct mlx5_flow_steering *steering) + { ++ int err; + + steering->root_ns = create_root_ns(steering, FS_FT_NIC_RX); + if (!steering->root_ns) +- goto cleanup; ++ return -ENOMEM; + +- if (init_root_tree(steering, &root_fs, &steering->root_ns->ns.node)) +- goto cleanup; ++ err = init_root_tree(steering, &root_fs, &steering->root_ns->ns.node); ++ if (err) ++ goto out_err; + + set_prio_attrs(steering->root_ns); + +- if (create_anchor_flow_table(steering)) +- goto cleanup; ++ err = create_anchor_flow_table(steering); ++ if (err) ++ goto out_err; + + return 0; + +-cleanup: +- mlx5_cleanup_fs(steering->dev); +- return -ENOMEM; ++out_err: ++ cleanup_root_ns(steering->root_ns); ++ steering->root_ns = NULL; ++ return err; + } + + static void clean_tree(struct fs_node *node) diff --git a/queue-4.9/net-mlx5-e-switch-include-vf-rdma-stats-in-vport-statistics.patch b/queue-4.9/net-mlx5-e-switch-include-vf-rdma-stats-in-vport-statistics.patch new file mode 100644 index 00000000000..490080a358a --- /dev/null +++ b/queue-4.9/net-mlx5-e-switch-include-vf-rdma-stats-in-vport-statistics.patch @@ -0,0 +1,61 @@ +From foo@baz Wed May 16 12:08:47 CEST 2018 +From: Adi Nissim +Date: Wed, 25 Apr 2018 11:21:32 +0300 +Subject: net/mlx5: E-Switch, Include VF RDMA stats in vport statistics + +From: Adi Nissim + +[ Upstream commit 88d725bbb43cd63a40c8ef70dd373f1d38ead2e3 ] + +The host side reporting of VF vport statistics didn't include the VF +RDMA traffic. + +Fixes: 3b751a2a418a ("net/mlx5: E-Switch, Introduce get vf statistics") +Signed-off-by: Adi Nissim +Reported-by: Ariel Almog +Reviewed-by: Or Gerlitz +Signed-off-by: Saeed Mahameed +Signed-off-by: Greg Kroah-Hartman +--- + drivers/net/ethernet/mellanox/mlx5/core/eswitch.c | 11 ++++++++++- + 1 file changed, 10 insertions(+), 1 deletion(-) + +--- a/drivers/net/ethernet/mellanox/mlx5/core/eswitch.c ++++ b/drivers/net/ethernet/mellanox/mlx5/core/eswitch.c +@@ -1924,26 +1924,35 @@ int mlx5_eswitch_get_vport_stats(struct + memset(vf_stats, 0, sizeof(*vf_stats)); + vf_stats->rx_packets = + MLX5_GET_CTR(out, received_eth_unicast.packets) + ++ MLX5_GET_CTR(out, received_ib_unicast.packets) + + MLX5_GET_CTR(out, received_eth_multicast.packets) + ++ MLX5_GET_CTR(out, received_ib_multicast.packets) + + MLX5_GET_CTR(out, received_eth_broadcast.packets); + + vf_stats->rx_bytes = + MLX5_GET_CTR(out, received_eth_unicast.octets) + ++ MLX5_GET_CTR(out, received_ib_unicast.octets) + + MLX5_GET_CTR(out, received_eth_multicast.octets) + ++ MLX5_GET_CTR(out, received_ib_multicast.octets) + + MLX5_GET_CTR(out, received_eth_broadcast.octets); + + vf_stats->tx_packets = + MLX5_GET_CTR(out, transmitted_eth_unicast.packets) + ++ MLX5_GET_CTR(out, transmitted_ib_unicast.packets) + + MLX5_GET_CTR(out, transmitted_eth_multicast.packets) + ++ MLX5_GET_CTR(out, transmitted_ib_multicast.packets) + + MLX5_GET_CTR(out, transmitted_eth_broadcast.packets); + + vf_stats->tx_bytes = + MLX5_GET_CTR(out, transmitted_eth_unicast.octets) + ++ MLX5_GET_CTR(out, transmitted_ib_unicast.octets) + + MLX5_GET_CTR(out, transmitted_eth_multicast.octets) + ++ MLX5_GET_CTR(out, transmitted_ib_multicast.octets) + + MLX5_GET_CTR(out, transmitted_eth_broadcast.octets); + + vf_stats->multicast = +- MLX5_GET_CTR(out, received_eth_multicast.packets); ++ MLX5_GET_CTR(out, received_eth_multicast.packets) + ++ MLX5_GET_CTR(out, received_ib_multicast.packets); + + vf_stats->broadcast = + MLX5_GET_CTR(out, received_eth_broadcast.packets); diff --git a/queue-4.9/net-support-compat-64-bit-time-in-s-g-etsockopt.patch b/queue-4.9/net-support-compat-64-bit-time-in-s-g-etsockopt.patch new file mode 100644 index 00000000000..e5653037f54 --- /dev/null +++ b/queue-4.9/net-support-compat-64-bit-time-in-s-g-etsockopt.patch @@ -0,0 +1,140 @@ +From foo@baz Wed May 16 12:08:47 CEST 2018 +From: Lance Richardson +Date: Wed, 25 Apr 2018 10:21:54 -0400 +Subject: net: support compat 64-bit time in {s,g}etsockopt + +From: Lance Richardson + +[ Upstream commit 988bf7243e03ef69238381594e0334a79cef74a6 ] + +For the x32 ABI, struct timeval has two 64-bit fields. However +the kernel currently interprets the user-space values used for +the SO_RCVTIMEO and SO_SNDTIMEO socket options as having a pair +of 32-bit fields. + +When the seconds portion of the requested timeout is less than 2**32, +the seconds portion of the effective timeout is correct but the +microseconds portion is zero. When the seconds portion of the +requested timeout is zero and the microseconds portion is non-zero, +the kernel interprets the timeout as zero (never timeout). + +Fix by using 64-bit time for SO_RCVTIMEO/SO_SNDTIMEO as required +for the ABI. + +The code included below demonstrates the problem. + +Results before patch: + $ gcc -m64 -Wall -O2 -o socktmo socktmo.c && ./socktmo + recv time: 2.008181 seconds + send time: 2.015985 seconds + + $ gcc -m32 -Wall -O2 -o socktmo socktmo.c && ./socktmo + recv time: 2.016763 seconds + send time: 2.016062 seconds + + $ gcc -mx32 -Wall -O2 -o socktmo socktmo.c && ./socktmo + recv time: 1.007239 seconds + send time: 1.023890 seconds + +Results after patch: + $ gcc -m64 -O2 -Wall -o socktmo socktmo.c && ./socktmo + recv time: 2.010062 seconds + send time: 2.015836 seconds + + $ gcc -m32 -O2 -Wall -o socktmo socktmo.c && ./socktmo + recv time: 2.013974 seconds + send time: 2.015981 seconds + + $ gcc -mx32 -O2 -Wall -o socktmo socktmo.c && ./socktmo + recv time: 2.030257 seconds + send time: 2.013383 seconds + + #include + #include + #include + #include + #include + + void checkrc(char *str, int rc) + { + if (rc >= 0) + return; + + perror(str); + exit(1); + } + + static char buf[1024]; + int main(int argc, char **argv) + { + int rc; + int socks[2]; + struct timeval tv; + struct timeval start, end, delta; + + rc = socketpair(AF_UNIX, SOCK_STREAM, 0, socks); + checkrc("socketpair", rc); + + /* set timeout to 1.999999 seconds */ + tv.tv_sec = 1; + tv.tv_usec = 999999; + rc = setsockopt(socks[0], SOL_SOCKET, SO_RCVTIMEO, &tv, sizeof tv); + rc = setsockopt(socks[0], SOL_SOCKET, SO_SNDTIMEO, &tv, sizeof tv); + checkrc("setsockopt", rc); + + /* measure actual receive timeout */ + gettimeofday(&start, NULL); + rc = recv(socks[0], buf, sizeof buf, 0); + gettimeofday(&end, NULL); + timersub(&end, &start, &delta); + + printf("recv time: %ld.%06ld seconds\n", + (long)delta.tv_sec, (long)delta.tv_usec); + + /* fill send buffer */ + do { + rc = send(socks[0], buf, sizeof buf, 0); + } while (rc > 0); + + /* measure actual send timeout */ + gettimeofday(&start, NULL); + rc = send(socks[0], buf, sizeof buf, 0); + gettimeofday(&end, NULL); + timersub(&end, &start, &delta); + + printf("send time: %ld.%06ld seconds\n", + (long)delta.tv_sec, (long)delta.tv_usec); + exit(0); + } + +Fixes: 515c7af85ed9 ("x32: Use compat shims for {g,s}etsockopt") +Reported-by: Gopal RajagopalSai +Signed-off-by: Lance Richardson +Signed-off-by: David S. Miller +Signed-off-by: Greg Kroah-Hartman +--- + net/compat.c | 6 ++++-- + 1 file changed, 4 insertions(+), 2 deletions(-) + +--- a/net/compat.c ++++ b/net/compat.c +@@ -372,7 +372,8 @@ static int compat_sock_setsockopt(struct + optname == SO_ATTACH_REUSEPORT_CBPF) + return do_set_attach_filter(sock, level, optname, + optval, optlen); +- if (optname == SO_RCVTIMEO || optname == SO_SNDTIMEO) ++ if (!COMPAT_USE_64BIT_TIME && ++ (optname == SO_RCVTIMEO || optname == SO_SNDTIMEO)) + return do_set_sock_timeout(sock, level, optname, optval, optlen); + + return sock_setsockopt(sock, level, optname, optval, optlen); +@@ -437,7 +438,8 @@ static int do_get_sock_timeout(struct so + static int compat_sock_getsockopt(struct socket *sock, int level, int optname, + char __user *optval, int __user *optlen) + { +- if (optname == SO_RCVTIMEO || optname == SO_SNDTIMEO) ++ if (!COMPAT_USE_64BIT_TIME && ++ (optname == SO_RCVTIMEO || optname == SO_SNDTIMEO)) + return do_get_sock_timeout(sock, level, optname, optval, optlen); + return sock_getsockopt(sock, level, optname, optval, optlen); + } diff --git a/queue-4.9/net_sched-fq-take-care-of-throttled-flows-before-reuse.patch b/queue-4.9/net_sched-fq-take-care-of-throttled-flows-before-reuse.patch new file mode 100644 index 00000000000..701693c6b9d --- /dev/null +++ b/queue-4.9/net_sched-fq-take-care-of-throttled-flows-before-reuse.patch @@ -0,0 +1,96 @@ +From foo@baz Wed May 16 12:08:47 CEST 2018 +From: Eric Dumazet +Date: Wed, 2 May 2018 10:03:30 -0700 +Subject: net_sched: fq: take care of throttled flows before reuse + +From: Eric Dumazet + +[ Upstream commit 7df40c2673a1307c3260aab6f9d4b9bf97ca8fd7 ] + +Normally, a socket can not be freed/reused unless all its TX packets +left qdisc and were TX-completed. However connect(AF_UNSPEC) allows +this to happen. + +With commit fc59d5bdf1e3 ("pkt_sched: fq: clear time_next_packet for +reused flows") we cleared f->time_next_packet but took no special +action if the flow was still in the throttled rb-tree. + +Since f->time_next_packet is the key used in the rb-tree searches, +blindly clearing it might break rb-tree integrity. We need to make +sure the flow is no longer in the rb-tree to avoid this problem. + +Fixes: fc59d5bdf1e3 ("pkt_sched: fq: clear time_next_packet for reused flows") +Signed-off-by: Eric Dumazet +Signed-off-by: David S. Miller +Signed-off-by: Greg Kroah-Hartman +--- + net/sched/sch_fq.c | 37 +++++++++++++++++++++++++------------ + 1 file changed, 25 insertions(+), 12 deletions(-) + +--- a/net/sched/sch_fq.c ++++ b/net/sched/sch_fq.c +@@ -128,6 +128,28 @@ static bool fq_flow_is_detached(const st + return f->next == &detached; + } + ++static bool fq_flow_is_throttled(const struct fq_flow *f) ++{ ++ return f->next == &throttled; ++} ++ ++static void fq_flow_add_tail(struct fq_flow_head *head, struct fq_flow *flow) ++{ ++ if (head->first) ++ head->last->next = flow; ++ else ++ head->first = flow; ++ head->last = flow; ++ flow->next = NULL; ++} ++ ++static void fq_flow_unset_throttled(struct fq_sched_data *q, struct fq_flow *f) ++{ ++ rb_erase(&f->rate_node, &q->delayed); ++ q->throttled_flows--; ++ fq_flow_add_tail(&q->old_flows, f); ++} ++ + static void fq_flow_set_throttled(struct fq_sched_data *q, struct fq_flow *f) + { + struct rb_node **p = &q->delayed.rb_node, *parent = NULL; +@@ -155,15 +177,6 @@ static void fq_flow_set_throttled(struct + + static struct kmem_cache *fq_flow_cachep __read_mostly; + +-static void fq_flow_add_tail(struct fq_flow_head *head, struct fq_flow *flow) +-{ +- if (head->first) +- head->last->next = flow; +- else +- head->first = flow; +- head->last = flow; +- flow->next = NULL; +-} + + /* limit number of collected flows per round */ + #define FQ_GC_MAX 8 +@@ -267,6 +280,8 @@ static struct fq_flow *fq_classify(struc + f->socket_hash != sk->sk_hash)) { + f->credit = q->initial_quantum; + f->socket_hash = sk->sk_hash; ++ if (fq_flow_is_throttled(f)) ++ fq_flow_unset_throttled(q, f); + f->time_next_packet = 0ULL; + } + return f; +@@ -430,9 +445,7 @@ static void fq_check_throttled(struct fq + q->time_next_delayed_flow = f->time_next_packet; + break; + } +- rb_erase(p, &q->delayed); +- q->throttled_flows--; +- fq_flow_add_tail(&q->old_flows, f); ++ fq_flow_unset_throttled(q, f); + } + } + diff --git a/queue-4.9/openvswitch-don-t-swap-table-in-nlattr_set-after-ovs_attr_nested-is-found.patch b/queue-4.9/openvswitch-don-t-swap-table-in-nlattr_set-after-ovs_attr_nested-is-found.patch new file mode 100644 index 00000000000..f194c22eaf0 --- /dev/null +++ b/queue-4.9/openvswitch-don-t-swap-table-in-nlattr_set-after-ovs_attr_nested-is-found.patch @@ -0,0 +1,110 @@ +From foo@baz Wed May 16 12:08:47 CEST 2018 +From: Stefano Brivio +Date: Thu, 3 May 2018 18:13:25 +0200 +Subject: openvswitch: Don't swap table in nlattr_set() after OVS_ATTR_NESTED is found + +From: Stefano Brivio + +[ Upstream commit 72f17baf2352ded6a1d3f4bb2d15da8c678cd2cb ] + +If an OVS_ATTR_NESTED attribute type is found while walking +through netlink attributes, we call nlattr_set() recursively +passing the length table for the following nested attributes, if +different from the current one. + +However, once we're done with those sub-nested attributes, we +should continue walking through attributes using the current +table, instead of using the one related to the sub-nested +attributes. + +For example, given this sequence: + +1 OVS_KEY_ATTR_PRIORITY +2 OVS_KEY_ATTR_TUNNEL +3 OVS_TUNNEL_KEY_ATTR_ID +4 OVS_TUNNEL_KEY_ATTR_IPV4_SRC +5 OVS_TUNNEL_KEY_ATTR_IPV4_DST +6 OVS_TUNNEL_KEY_ATTR_TTL +7 OVS_TUNNEL_KEY_ATTR_TP_SRC +8 OVS_TUNNEL_KEY_ATTR_TP_DST +9 OVS_KEY_ATTR_IN_PORT +10 OVS_KEY_ATTR_SKB_MARK +11 OVS_KEY_ATTR_MPLS + +we switch to the 'ovs_tunnel_key_lens' table on attribute #3, +and we don't switch back to 'ovs_key_lens' while setting +attributes #9 to #11 in the sequence. As OVS_KEY_ATTR_MPLS +evaluates to 21, and the array size of 'ovs_tunnel_key_lens' is +15, we also get this kind of KASan splat while accessing the +wrong table: + +[ 7654.586496] ================================================================== +[ 7654.594573] BUG: KASAN: global-out-of-bounds in nlattr_set+0x164/0xde9 [openvswitch] +[ 7654.603214] Read of size 4 at addr ffffffffc169ecf0 by task handler29/87430 +[ 7654.610983] +[ 7654.612644] CPU: 21 PID: 87430 Comm: handler29 Kdump: loaded Not tainted 3.10.0-866.el7.test.x86_64 #1 +[ 7654.623030] Hardware name: Dell Inc. PowerEdge R730/072T6D, BIOS 2.1.7 06/16/2016 +[ 7654.631379] Call Trace: +[ 7654.634108] [] dump_stack+0x19/0x1b +[ 7654.639843] [] print_address_description+0x33/0x290 +[ 7654.647129] [] ? nlattr_set+0x164/0xde9 [openvswitch] +[ 7654.654607] [] kasan_report.part.3+0x242/0x330 +[ 7654.661406] [] __asan_report_load4_noabort+0x34/0x40 +[ 7654.668789] [] nlattr_set+0x164/0xde9 [openvswitch] +[ 7654.676076] [] ovs_nla_get_match+0x10c8/0x1900 [openvswitch] +[ 7654.684234] [] ? genl_rcv+0x28/0x40 +[ 7654.689968] [] ? netlink_unicast+0x3f3/0x590 +[ 7654.696574] [] ? ovs_nla_put_tunnel_info+0xb0/0xb0 [openvswitch] +[ 7654.705122] [] ? unwind_get_return_address+0xb0/0xb0 +[ 7654.712503] [] ? system_call_fastpath+0x1c/0x21 +[ 7654.719401] [] ? update_stack_state+0x229/0x370 +[ 7654.726298] [] ? update_stack_state+0x229/0x370 +[ 7654.733195] [] ? kasan_unpoison_shadow+0x35/0x50 +[ 7654.740187] [] ? kasan_kmalloc+0xaa/0xe0 +[ 7654.746406] [] ? kasan_slab_alloc+0x12/0x20 +[ 7654.752914] [] ? memset+0x31/0x40 +[ 7654.758456] [] ovs_flow_cmd_new+0x2b2/0xf00 [openvswitch] + +[snip] + +[ 7655.132484] The buggy address belongs to the variable: +[ 7655.138226] ovs_tunnel_key_lens+0xf0/0xffffffffffffd400 [openvswitch] +[ 7655.145507] +[ 7655.147166] Memory state around the buggy address: +[ 7655.152514] ffffffffc169eb80: 00 00 00 00 00 00 00 00 00 00 fa fa fa fa fa fa +[ 7655.160585] ffffffffc169ec00: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 +[ 7655.168644] >ffffffffc169ec80: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 fa fa +[ 7655.176701] ^ +[ 7655.184372] ffffffffc169ed00: fa fa fa fa 00 00 00 00 fa fa fa fa 00 00 00 05 +[ 7655.192431] ffffffffc169ed80: fa fa fa fa 00 00 00 00 00 00 00 00 00 00 00 00 +[ 7655.200490] ================================================================== + +Reported-by: Hangbin Liu +Fixes: 982b52700482 ("openvswitch: Fix mask generation for nested attributes.") +Signed-off-by: Stefano Brivio +Reviewed-by: Sabrina Dubroca +Signed-off-by: David S. Miller +Signed-off-by: Greg Kroah-Hartman +--- + net/openvswitch/flow_netlink.c | 9 +++------ + 1 file changed, 3 insertions(+), 6 deletions(-) + +--- a/net/openvswitch/flow_netlink.c ++++ b/net/openvswitch/flow_netlink.c +@@ -1296,13 +1296,10 @@ static void nlattr_set(struct nlattr *at + + /* The nlattr stream should already have been validated */ + nla_for_each_nested(nla, attr, rem) { +- if (tbl[nla_type(nla)].len == OVS_ATTR_NESTED) { +- if (tbl[nla_type(nla)].next) +- tbl = tbl[nla_type(nla)].next; +- nlattr_set(nla, val, tbl); +- } else { ++ if (tbl[nla_type(nla)].len == OVS_ATTR_NESTED) ++ nlattr_set(nla, val, tbl[nla_type(nla)].next ? : tbl); ++ else + memset(nla_data(nla), val, nla_len(nla)); +- } + + if (nla_type(nla) == OVS_KEY_ATTR_CT_STATE) + *(u32 *)nla_data(nla) &= CT_SUPPORTED_MASK; diff --git a/queue-4.9/qmi_wwan-do-not-steal-interfaces-from-class-drivers.patch b/queue-4.9/qmi_wwan-do-not-steal-interfaces-from-class-drivers.patch new file mode 100644 index 00000000000..f01fe4ac696 --- /dev/null +++ b/queue-4.9/qmi_wwan-do-not-steal-interfaces-from-class-drivers.patch @@ -0,0 +1,49 @@ +From foo@baz Wed May 16 12:08:47 CEST 2018 +From: "Bjørn Mork" +Date: Wed, 2 May 2018 22:22:54 +0200 +Subject: qmi_wwan: do not steal interfaces from class drivers + +From: "Bjørn Mork" + +[ Upstream commit 5697db4a696c41601a1d15c1922150b4dbf5726c ] + +The USB_DEVICE_INTERFACE_NUMBER matching macro assumes that +the { vendorid, productid, interfacenumber } set uniquely +identifies one specific function. This has proven to fail +for some configurable devices. One example is the Quectel +EM06/EP06 where the same interface number can be either +QMI or MBIM, without the device ID changing either. + +Fix by requiring the vendor-specific class for interface number +based matching. Functions of other classes can and should use +class based matching instead. + +Fixes: 03304bcb5ec4 ("net: qmi_wwan: use fixed interface number matching") +Signed-off-by: Bjørn Mork +Signed-off-by: David S. Miller +Signed-off-by: Greg Kroah-Hartman +--- + drivers/net/usb/qmi_wwan.c | 12 ++++++++++++ + 1 file changed, 12 insertions(+) + +--- a/drivers/net/usb/qmi_wwan.c ++++ b/drivers/net/usb/qmi_wwan.c +@@ -1039,6 +1039,18 @@ static int qmi_wwan_probe(struct usb_int + id->driver_info = (unsigned long)&qmi_wwan_info; + } + ++ /* There are devices where the same interface number can be ++ * configured as different functions. We should only bind to ++ * vendor specific functions when matching on interface number ++ */ ++ if (id->match_flags & USB_DEVICE_ID_MATCH_INT_NUMBER && ++ desc->bInterfaceClass != USB_CLASS_VENDOR_SPEC) { ++ dev_dbg(&intf->dev, ++ "Rejecting interface number match for class %02x\n", ++ desc->bInterfaceClass); ++ return -ENODEV; ++ } ++ + /* Quectel EC20 quirk where we've QMI on interface 4 instead of 0 */ + if (quectel_ec20_detected(intf) && desc->bInterfaceNumber == 0) { + dev_dbg(&intf->dev, "Quectel EC20 quirk, skipping interface 0\n"); diff --git a/queue-4.9/r8169-fix-powering-up-rtl8168h.patch b/queue-4.9/r8169-fix-powering-up-rtl8168h.patch new file mode 100644 index 00000000000..82f267bb3af --- /dev/null +++ b/queue-4.9/r8169-fix-powering-up-rtl8168h.patch @@ -0,0 +1,49 @@ +From foo@baz Wed May 16 12:08:47 CEST 2018 +From: Heiner Kallweit +Date: Mon, 7 May 2018 21:11:21 +0200 +Subject: r8169: fix powering up RTL8168h + +From: Heiner Kallweit + +[ Upstream commit 3148dedfe79e422f448a10250d3e2cdf8b7ee617 ] + +Since commit a92a08499b1f "r8169: improve runtime pm in general and +suspend unused ports" interfaces w/o link are runtime-suspended after +10s. On systems where drivers take longer to load this can lead to the +situation that the interface is runtime-suspended already when it's +initially brought up. +This shouldn't be a problem because rtl_open() resumes MAC/PHY. +However with at least one chip version the interface doesn't properly +come up, as reported here: +https://bugzilla.kernel.org/show_bug.cgi?id=199549 + +The vendor driver uses a delay to give certain chip versions some +time to resume before starting the PHY configuration. So let's do +the same. I don't know which chip versions may be affected, +therefore apply this delay always. + +This patch was reported to fix the issue for RTL8168h. +I was able to reproduce the issue on an Asus H310I-Plus which also +uses a RTL8168h. Also in my case the patch fixed the issue. + +Reported-by: Slava Kardakov +Tested-by: Slava Kardakov +Signed-off-by: Heiner Kallweit +Signed-off-by: David S. Miller +Signed-off-by: Greg Kroah-Hartman +--- + drivers/net/ethernet/realtek/r8169.c | 3 +++ + 1 file changed, 3 insertions(+) + +--- a/drivers/net/ethernet/realtek/r8169.c ++++ b/drivers/net/ethernet/realtek/r8169.c +@@ -4861,6 +4861,9 @@ static void rtl_pll_power_down(struct rt + static void rtl_pll_power_up(struct rtl8169_private *tp) + { + rtl_generic_op(tp, tp->pll_power_ops.up); ++ ++ /* give MAC/PHY some time to resume */ ++ msleep(20); + } + + static void rtl_init_pll_power_ops(struct rtl8169_private *tp) diff --git a/queue-4.9/sctp-delay-the-authentication-for-the-duplicated-cookie-echo-chunk.patch b/queue-4.9/sctp-delay-the-authentication-for-the-duplicated-cookie-echo-chunk.patch new file mode 100644 index 00000000000..711f1ef431b --- /dev/null +++ b/queue-4.9/sctp-delay-the-authentication-for-the-duplicated-cookie-echo-chunk.patch @@ -0,0 +1,249 @@ +From foo@baz Wed May 16 12:08:47 CEST 2018 +From: Xin Long +Date: Sat, 5 May 2018 14:59:47 +0800 +Subject: sctp: delay the authentication for the duplicated cookie-echo chunk + +From: Xin Long + +[ Upstream commit 59d8d4434f429b4fa8a346fd889058bda427a837 ] + +Now sctp only delays the authentication for the normal cookie-echo +chunk by setting chunk->auth_chunk in sctp_endpoint_bh_rcv(). But +for the duplicated one with auth, in sctp_assoc_bh_rcv(), it does +authentication first based on the old asoc, which will definitely +fail due to the different auth info in the old asoc. + +The duplicated cookie-echo chunk will create a new asoc with the +auth info from this chunk, and the authentication should also be +done with the new asoc's auth info for all of the collision 'A', +'B' and 'D'. Otherwise, the duplicated cookie-echo chunk with auth +will never pass the authentication and create the new connection. + +This issue exists since very beginning, and this fix is to make +sctp_assoc_bh_rcv() follow the way sctp_endpoint_bh_rcv() does +for the normal cookie-echo chunk to delay the authentication. + +While at it, remove the unused params from sctp_sf_authenticate() +and define sctp_auth_chunk_verify() used for all the places that +do the delayed authentication. + +v1->v2: + fix the typo in changelog as Marcelo noticed. + +Acked-by: Marcelo Ricardo Leitner +Signed-off-by: Xin Long +Acked-by: Neil Horman +Signed-off-by: David S. Miller +Signed-off-by: Greg Kroah-Hartman +--- + net/sctp/associola.c | 30 ++++++++++++++++ + net/sctp/sm_statefuns.c | 87 ++++++++++++++++++++++++++---------------------- + 2 files changed, 77 insertions(+), 40 deletions(-) + +--- a/net/sctp/associola.c ++++ b/net/sctp/associola.c +@@ -1006,9 +1006,10 @@ static void sctp_assoc_bh_rcv(struct wor + struct sctp_endpoint *ep; + struct sctp_chunk *chunk; + struct sctp_inq *inqueue; +- int state; + sctp_subtype_t subtype; ++ int first_time = 1; /* is this the first time through the loop */ + int error = 0; ++ int state; + + /* The association should be held so we should be safe. */ + ep = asoc->ep; +@@ -1019,6 +1020,30 @@ static void sctp_assoc_bh_rcv(struct wor + state = asoc->state; + subtype = SCTP_ST_CHUNK(chunk->chunk_hdr->type); + ++ /* If the first chunk in the packet is AUTH, do special ++ * processing specified in Section 6.3 of SCTP-AUTH spec ++ */ ++ if (first_time && subtype.chunk == SCTP_CID_AUTH) { ++ struct sctp_chunkhdr *next_hdr; ++ ++ next_hdr = sctp_inq_peek(inqueue); ++ if (!next_hdr) ++ goto normal; ++ ++ /* If the next chunk is COOKIE-ECHO, skip the AUTH ++ * chunk while saving a pointer to it so we can do ++ * Authentication later (during cookie-echo ++ * processing). ++ */ ++ if (next_hdr->type == SCTP_CID_COOKIE_ECHO) { ++ chunk->auth_chunk = skb_clone(chunk->skb, ++ GFP_ATOMIC); ++ chunk->auth = 1; ++ continue; ++ } ++ } ++ ++normal: + /* SCTP-AUTH, Section 6.3: + * The receiver has a list of chunk types which it expects + * to be received only after an AUTH-chunk. This list has +@@ -1057,6 +1082,9 @@ static void sctp_assoc_bh_rcv(struct wor + /* If there is an error on chunk, discard this packet. */ + if (error && chunk) + chunk->pdiscard = 1; ++ ++ if (first_time) ++ first_time = 0; + } + sctp_association_put(asoc); + } +--- a/net/sctp/sm_statefuns.c ++++ b/net/sctp/sm_statefuns.c +@@ -144,10 +144,8 @@ static sctp_disposition_t sctp_sf_violat + void *arg, + sctp_cmd_seq_t *commands); + +-static sctp_ierror_t sctp_sf_authenticate(struct net *net, +- const struct sctp_endpoint *ep, ++static sctp_ierror_t sctp_sf_authenticate( + const struct sctp_association *asoc, +- const sctp_subtype_t type, + struct sctp_chunk *chunk); + + static sctp_disposition_t __sctp_sf_do_9_1_abort(struct net *net, +@@ -615,6 +613,38 @@ sctp_disposition_t sctp_sf_do_5_1C_ack(s + return SCTP_DISPOSITION_CONSUME; + } + ++static bool sctp_auth_chunk_verify(struct net *net, struct sctp_chunk *chunk, ++ const struct sctp_association *asoc) ++{ ++ struct sctp_chunk auth; ++ ++ if (!chunk->auth_chunk) ++ return true; ++ ++ /* SCTP-AUTH: auth_chunk pointer is only set when the cookie-echo ++ * is supposed to be authenticated and we have to do delayed ++ * authentication. We've just recreated the association using ++ * the information in the cookie and now it's much easier to ++ * do the authentication. ++ */ ++ ++ /* Make sure that we and the peer are AUTH capable */ ++ if (!net->sctp.auth_enable || !asoc->peer.auth_capable) ++ return false; ++ ++ /* set-up our fake chunk so that we can process it */ ++ auth.skb = chunk->auth_chunk; ++ auth.asoc = chunk->asoc; ++ auth.sctp_hdr = chunk->sctp_hdr; ++ auth.chunk_hdr = (struct sctp_chunkhdr *) ++ skb_push(chunk->auth_chunk, ++ sizeof(struct sctp_chunkhdr)); ++ skb_pull(chunk->auth_chunk, sizeof(struct sctp_chunkhdr)); ++ auth.transport = chunk->transport; ++ ++ return sctp_sf_authenticate(asoc, &auth) == SCTP_IERROR_NO_ERROR; ++} ++ + /* + * Respond to a normal COOKIE ECHO chunk. + * We are the side that is being asked for an association. +@@ -751,36 +781,9 @@ sctp_disposition_t sctp_sf_do_5_1D_ce(st + if (error) + goto nomem_init; + +- /* SCTP-AUTH: auth_chunk pointer is only set when the cookie-echo +- * is supposed to be authenticated and we have to do delayed +- * authentication. We've just recreated the association using +- * the information in the cookie and now it's much easier to +- * do the authentication. +- */ +- if (chunk->auth_chunk) { +- struct sctp_chunk auth; +- sctp_ierror_t ret; +- +- /* Make sure that we and the peer are AUTH capable */ +- if (!net->sctp.auth_enable || !new_asoc->peer.auth_capable) { +- sctp_association_free(new_asoc); +- return sctp_sf_pdiscard(net, ep, asoc, type, arg, commands); +- } +- +- /* set-up our fake chunk so that we can process it */ +- auth.skb = chunk->auth_chunk; +- auth.asoc = chunk->asoc; +- auth.sctp_hdr = chunk->sctp_hdr; +- auth.chunk_hdr = (sctp_chunkhdr_t *)skb_push(chunk->auth_chunk, +- sizeof(sctp_chunkhdr_t)); +- skb_pull(chunk->auth_chunk, sizeof(sctp_chunkhdr_t)); +- auth.transport = chunk->transport; +- +- ret = sctp_sf_authenticate(net, ep, new_asoc, type, &auth); +- if (ret != SCTP_IERROR_NO_ERROR) { +- sctp_association_free(new_asoc); +- return sctp_sf_pdiscard(net, ep, asoc, type, arg, commands); +- } ++ if (!sctp_auth_chunk_verify(net, chunk, new_asoc)) { ++ sctp_association_free(new_asoc); ++ return sctp_sf_pdiscard(net, ep, asoc, type, arg, commands); + } + + repl = sctp_make_cookie_ack(new_asoc, chunk); +@@ -1717,13 +1720,15 @@ static sctp_disposition_t sctp_sf_do_dup + GFP_ATOMIC)) + goto nomem; + ++ if (!sctp_auth_chunk_verify(net, chunk, new_asoc)) ++ return SCTP_DISPOSITION_DISCARD; ++ + /* Make sure no new addresses are being added during the + * restart. Though this is a pretty complicated attack + * since you'd have to get inside the cookie. + */ +- if (!sctp_sf_check_restart_addrs(new_asoc, asoc, chunk, commands)) { ++ if (!sctp_sf_check_restart_addrs(new_asoc, asoc, chunk, commands)) + return SCTP_DISPOSITION_CONSUME; +- } + + /* If the endpoint is in the SHUTDOWN-ACK-SENT state and recognizes + * the peer has restarted (Action A), it MUST NOT setup a new +@@ -1828,6 +1833,9 @@ static sctp_disposition_t sctp_sf_do_dup + GFP_ATOMIC)) + goto nomem; + ++ if (!sctp_auth_chunk_verify(net, chunk, new_asoc)) ++ return SCTP_DISPOSITION_DISCARD; ++ + /* Update the content of current association. */ + sctp_add_cmd_sf(commands, SCTP_CMD_UPDATE_ASSOC, SCTP_ASOC(new_asoc)); + sctp_add_cmd_sf(commands, SCTP_CMD_NEW_STATE, +@@ -1920,6 +1928,9 @@ static sctp_disposition_t sctp_sf_do_dup + * a COOKIE ACK. + */ + ++ if (!sctp_auth_chunk_verify(net, chunk, asoc)) ++ return SCTP_DISPOSITION_DISCARD; ++ + /* Don't accidentally move back into established state. */ + if (asoc->state < SCTP_STATE_ESTABLISHED) { + sctp_add_cmd_sf(commands, SCTP_CMD_TIMER_STOP, +@@ -3981,10 +3992,8 @@ gen_shutdown: + * + * The return value is the disposition of the chunk. + */ +-static sctp_ierror_t sctp_sf_authenticate(struct net *net, +- const struct sctp_endpoint *ep, ++static sctp_ierror_t sctp_sf_authenticate( + const struct sctp_association *asoc, +- const sctp_subtype_t type, + struct sctp_chunk *chunk) + { + struct sctp_authhdr *auth_hdr; +@@ -4083,7 +4092,7 @@ sctp_disposition_t sctp_sf_eat_auth(stru + commands); + + auth_hdr = (struct sctp_authhdr *)chunk->skb->data; +- error = sctp_sf_authenticate(net, ep, asoc, type, chunk); ++ error = sctp_sf_authenticate(asoc, chunk); + switch (error) { + case SCTP_IERROR_AUTH_BAD_HMAC: + /* Generate the ERROR chunk and discard the rest diff --git a/queue-4.9/sctp-fix-the-issue-that-the-cookie-ack-with-auth-can-t-get-processed.patch b/queue-4.9/sctp-fix-the-issue-that-the-cookie-ack-with-auth-can-t-get-processed.patch new file mode 100644 index 00000000000..f6c7c4a9922 --- /dev/null +++ b/queue-4.9/sctp-fix-the-issue-that-the-cookie-ack-with-auth-can-t-get-processed.patch @@ -0,0 +1,42 @@ +From foo@baz Wed May 16 12:08:47 CEST 2018 +From: Xin Long +Date: Wed, 2 May 2018 13:45:12 +0800 +Subject: sctp: fix the issue that the cookie-ack with auth can't get processed + +From: Xin Long + +[ Upstream commit ce402f044e4e432c296f90eaabb8dbe8f3624391 ] + +When auth is enabled for cookie-ack chunk, in sctp_inq_pop, sctp +processes auth chunk first, then continues to the next chunk in +this packet if chunk_end + chunk_hdr size < skb_tail_pointer(). +Otherwise, it will go to the next packet or discard this chunk. + +However, it missed the fact that cookie-ack chunk's size is equal +to chunk_hdr size, which couldn't match that check, and thus this +chunk would not get processed. + +This patch fixes it by changing the check to chunk_end + chunk_hdr +size <= skb_tail_pointer(). + +Fixes: 26b87c788100 ("net: sctp: fix remote memory pressure from excessive queueing") +Signed-off-by: Xin Long +Acked-by: Neil Horman +Acked-by: Marcelo Ricardo Leitner +Signed-off-by: David S. Miller +Signed-off-by: Greg Kroah-Hartman +--- + net/sctp/inqueue.c | 2 +- + 1 file changed, 1 insertion(+), 1 deletion(-) + +--- a/net/sctp/inqueue.c ++++ b/net/sctp/inqueue.c +@@ -217,7 +217,7 @@ new_skb: + skb_pull(chunk->skb, sizeof(sctp_chunkhdr_t)); + chunk->subh.v = NULL; /* Subheader is no longer valid. */ + +- if (chunk->chunk_end + sizeof(sctp_chunkhdr_t) < ++ if (chunk->chunk_end + sizeof(sctp_chunkhdr_t) <= + skb_tail_pointer(chunk->skb)) { + /* This is not a singleton */ + chunk->singleton = 0; diff --git a/queue-4.9/sctp-handle-two-v4-addrs-comparison-in-sctp_inet6_cmp_addr.patch b/queue-4.9/sctp-handle-two-v4-addrs-comparison-in-sctp_inet6_cmp_addr.patch new file mode 100644 index 00000000000..ce508c14ae5 --- /dev/null +++ b/queue-4.9/sctp-handle-two-v4-addrs-comparison-in-sctp_inet6_cmp_addr.patch @@ -0,0 +1,47 @@ +From foo@baz Wed May 16 12:08:47 CEST 2018 +From: Xin Long +Date: Thu, 26 Apr 2018 14:13:57 +0800 +Subject: sctp: handle two v4 addrs comparison in sctp_inet6_cmp_addr + +From: Xin Long + +[ Upstream commit d625329b06e46bd20baf9ee40847d11982569204 ] + +Since sctp ipv6 socket also supports v4 addrs, it's possible to +compare two v4 addrs in pf v6 .cmp_addr, sctp_inet6_cmp_addr. + +However after Commit 1071ec9d453a ("sctp: do not check port in +sctp_inet6_cmp_addr"), it no longer calls af1->cmp_addr, which +in this case is sctp_v4_cmp_addr, but calls __sctp_v6_cmp_addr +where it handles them as two v6 addrs. It would cause a out of +bounds crash. + +syzbot found this crash when trying to bind two v4 addrs to a +v6 socket. + +This patch fixes it by adding the process for two v4 addrs in +sctp_inet6_cmp_addr. + +Fixes: 1071ec9d453a ("sctp: do not check port in sctp_inet6_cmp_addr") +Reported-by: syzbot+cd494c1dd681d4d93ebb@syzkaller.appspotmail.com +Signed-off-by: Xin Long +Acked-by: Neil Horman +Acked-by: Marcelo Ricardo Leitner +Signed-off-by: David S. Miller +Signed-off-by: Greg Kroah-Hartman +--- + net/sctp/ipv6.c | 3 +++ + 1 file changed, 3 insertions(+) + +--- a/net/sctp/ipv6.c ++++ b/net/sctp/ipv6.c +@@ -864,6 +864,9 @@ static int sctp_inet6_cmp_addr(const uni + if (sctp_is_any(sk, addr1) || sctp_is_any(sk, addr2)) + return 1; + ++ if (addr1->sa.sa_family == AF_INET && addr2->sa.sa_family == AF_INET) ++ return addr1->v4.sin_addr.s_addr == addr2->v4.sin_addr.s_addr; ++ + return __sctp_v6_cmp_addr(addr1, addr2); + } + diff --git a/queue-4.9/sctp-remove-sctp_chunk_put-from-fail_mark-err-path-in-sctp_ulpevent_make_rcvmsg.patch b/queue-4.9/sctp-remove-sctp_chunk_put-from-fail_mark-err-path-in-sctp_ulpevent_make_rcvmsg.patch new file mode 100644 index 00000000000..480a72c8c5c --- /dev/null +++ b/queue-4.9/sctp-remove-sctp_chunk_put-from-fail_mark-err-path-in-sctp_ulpevent_make_rcvmsg.patch @@ -0,0 +1,40 @@ +From foo@baz Wed May 16 12:08:47 CEST 2018 +From: Xin Long +Date: Thu, 10 May 2018 17:34:13 +0800 +Subject: sctp: remove sctp_chunk_put from fail_mark err path in sctp_ulpevent_make_rcvmsg + +From: Xin Long + +[ Upstream commit 6910e25de2257e2c82c7a2d126e3463cd8e50810 ] + +In Commit 1f45f78f8e51 ("sctp: allow GSO frags to access the chunk too"), +it held the chunk in sctp_ulpevent_make_rcvmsg to access it safely later +in recvmsg. However, it also added sctp_chunk_put in fail_mark err path, +which is only triggered before holding the chunk. + +syzbot reported a use-after-free crash happened on this err path, where +it shouldn't call sctp_chunk_put. + +This patch simply removes this call. + +Fixes: 1f45f78f8e51 ("sctp: allow GSO frags to access the chunk too") +Reported-by: syzbot+141d898c5f24489db4aa@syzkaller.appspotmail.com +Signed-off-by: Xin Long +Acked-by: Neil Horman +Acked-by: Marcelo Ricardo Leitner +Signed-off-by: David S. Miller +Signed-off-by: Greg Kroah-Hartman +--- + net/sctp/ulpevent.c | 1 - + 1 file changed, 1 deletion(-) + +--- a/net/sctp/ulpevent.c ++++ b/net/sctp/ulpevent.c +@@ -723,7 +723,6 @@ struct sctp_ulpevent *sctp_ulpevent_make + return event; + + fail_mark: +- sctp_chunk_put(chunk); + kfree_skb(skb); + fail: + return NULL; diff --git a/queue-4.9/sctp-use-the-old-asoc-when-making-the-cookie-ack-chunk-in-dupcook_d.patch b/queue-4.9/sctp-use-the-old-asoc-when-making-the-cookie-ack-chunk-in-dupcook_d.patch new file mode 100644 index 00000000000..fe124c5ba58 --- /dev/null +++ b/queue-4.9/sctp-use-the-old-asoc-when-making-the-cookie-ack-chunk-in-dupcook_d.patch @@ -0,0 +1,40 @@ +From foo@baz Wed May 16 12:08:47 CEST 2018 +From: Xin Long +Date: Wed, 2 May 2018 13:39:46 +0800 +Subject: sctp: use the old asoc when making the cookie-ack chunk in dupcook_d + +From: Xin Long + +[ Upstream commit 46e16d4b956867013e0bbd7f2bad206f4aa55752 ] + +When processing a duplicate cookie-echo chunk, for case 'D', sctp will +not process the param from this chunk. It means old asoc has nothing +to be updated, and the new temp asoc doesn't have the complete info. + +So there's no reason to use the new asoc when creating the cookie-ack +chunk. Otherwise, like when auth is enabled for cookie-ack, the chunk +can not be set with auth, and it will definitely be dropped by peer. + +This issue is there since very beginning, and we fix it by using the +old asoc instead. + +Signed-off-by: Xin Long +Acked-by: Neil Horman +Acked-by: Marcelo Ricardo Leitner +Signed-off-by: David S. Miller +Signed-off-by: Greg Kroah-Hartman +--- + net/sctp/sm_statefuns.c | 2 +- + 1 file changed, 1 insertion(+), 1 deletion(-) + +--- a/net/sctp/sm_statefuns.c ++++ b/net/sctp/sm_statefuns.c +@@ -1959,7 +1959,7 @@ static sctp_disposition_t sctp_sf_do_dup + } + } + +- repl = sctp_make_cookie_ack(new_asoc, chunk); ++ repl = sctp_make_cookie_ack(asoc, chunk); + if (!repl) + goto nomem; + diff --git a/queue-4.9/series b/queue-4.9/series new file mode 100644 index 00000000000..cc8e702163b --- /dev/null +++ b/queue-4.9/series @@ -0,0 +1,25 @@ +8139too-use-disable_irq_nosync-in-rtl8139_poll_controller.patch +bridge-check-iface-upper-dev-when-setting-master-via-ioctl.patch +dccp-fix-tasklet-usage.patch +ipv4-fix-memory-leaks-in-udp_sendmsg-ping_v4_sendmsg.patch +llc-better-deal-with-too-small-mtu.patch +net-ethernet-sun-niu-set-correct-packet-size-in-skb.patch +net-ethernet-ti-cpsw-fix-packet-leaking-in-dual_mac-mode.patch +net-mlx4_en-verify-coalescing-parameters-are-in-range.patch +net-mlx5-e-switch-include-vf-rdma-stats-in-vport-statistics.patch +net_sched-fq-take-care-of-throttled-flows-before-reuse.patch +net-support-compat-64-bit-time-in-s-g-etsockopt.patch +openvswitch-don-t-swap-table-in-nlattr_set-after-ovs_attr_nested-is-found.patch +qmi_wwan-do-not-steal-interfaces-from-class-drivers.patch +r8169-fix-powering-up-rtl8168h.patch +sctp-handle-two-v4-addrs-comparison-in-sctp_inet6_cmp_addr.patch +sctp-remove-sctp_chunk_put-from-fail_mark-err-path-in-sctp_ulpevent_make_rcvmsg.patch +sctp-use-the-old-asoc-when-making-the-cookie-ack-chunk-in-dupcook_d.patch +tcp_bbr-fix-to-zero-idle_restart-only-upon-s-acked-data.patch +tg3-fix-vunmap-bug_on-triggered-from-tg3_free_consistent.patch +bonding-do-not-allow-rlb-updates-to-invalid-mac.patch +net-mlx5-avoid-cleaning-flow-steering-table-twice-during-error-flow.patch +bonding-send-learning-packets-for-vlans-on-slave.patch +tcp-ignore-fast-open-on-repair-mode.patch +sctp-fix-the-issue-that-the-cookie-ack-with-auth-can-t-get-processed.patch +sctp-delay-the-authentication-for-the-duplicated-cookie-echo-chunk.patch diff --git a/queue-4.9/tcp-ignore-fast-open-on-repair-mode.patch b/queue-4.9/tcp-ignore-fast-open-on-repair-mode.patch new file mode 100644 index 00000000000..c58779f0ad1 --- /dev/null +++ b/queue-4.9/tcp-ignore-fast-open-on-repair-mode.patch @@ -0,0 +1,45 @@ +From foo@baz Wed May 16 12:08:47 CEST 2018 +From: Yuchung Cheng +Date: Wed, 25 Apr 2018 11:33:08 -0700 +Subject: tcp: ignore Fast Open on repair mode + +From: Yuchung Cheng + +[ Upstream commit 16ae6aa1705299789f71fdea59bfb119c1fbd9c0 ] + +The TCP repair sequence of operation is to first set the socket in +repair mode, then inject the TCP stats into the socket with repair +socket options, then call connect() to re-activate the socket. The +connect syscall simply returns and set state to ESTABLISHED +mode. As a result Fast Open is meaningless for TCP repair. + +However allowing sendto() system call with MSG_FASTOPEN flag half-way +during the repair operation could unexpectedly cause data to be +sent, before the operation finishes changing the internal TCP stats +(e.g. MSS). This in turn triggers TCP warnings on inconsistent +packet accounting. + +The fix is to simply disallow Fast Open operation once the socket +is in the repair mode. + +Reported-by: syzbot +Signed-off-by: Yuchung Cheng +Reviewed-by: Neal Cardwell +Reviewed-by: Eric Dumazet +Signed-off-by: David S. Miller +Signed-off-by: Greg Kroah-Hartman +--- + net/ipv4/tcp.c | 2 +- + 1 file changed, 1 insertion(+), 1 deletion(-) + +--- a/net/ipv4/tcp.c ++++ b/net/ipv4/tcp.c +@@ -1118,7 +1118,7 @@ int tcp_sendmsg(struct sock *sk, struct + lock_sock(sk); + + flags = msg->msg_flags; +- if (flags & MSG_FASTOPEN) { ++ if ((flags & MSG_FASTOPEN) && !tp->repair) { + err = tcp_sendmsg_fastopen(sk, msg, &copied_syn, size); + if (err == -EINPROGRESS && copied_syn > 0) + goto out; diff --git a/queue-4.9/tcp_bbr-fix-to-zero-idle_restart-only-upon-s-acked-data.patch b/queue-4.9/tcp_bbr-fix-to-zero-idle_restart-only-upon-s-acked-data.patch new file mode 100644 index 00000000000..89ed9b0235e --- /dev/null +++ b/queue-4.9/tcp_bbr-fix-to-zero-idle_restart-only-upon-s-acked-data.patch @@ -0,0 +1,48 @@ +From foo@baz Wed May 16 12:08:47 CEST 2018 +From: Neal Cardwell +Date: Tue, 1 May 2018 21:45:41 -0400 +Subject: tcp_bbr: fix to zero idle_restart only upon S/ACKed data + +From: Neal Cardwell + +[ Upstream commit e6e6a278b1eaffa19d42186bfacd1ffc15a50b3f ] + +Previously the bbr->idle_restart tracking was zeroing out the +bbr->idle_restart bit upon ACKs that did not SACK or ACK anything, +e.g. receiving incoming data or receiver window updates. In such +situations BBR would forget that this was a restart-from-idle +situation, and if the min_rtt had expired it would unnecessarily enter +PROBE_RTT (even though we were actually restarting from idle but had +merely forgotten that fact). + +The fix is simple: we need to remember we are restarting from idle +until we receive a S/ACK for some data (a S/ACK for the first flight +of data we send as we are restarting). + +This commit is a stable candidate for kernels back as far as 4.9. + +Fixes: 0f8782ea1497 ("tcp_bbr: add BBR congestion control") +Signed-off-by: Neal Cardwell +Signed-off-by: Yuchung Cheng +Signed-off-by: Soheil Hassas Yeganeh +Signed-off-by: Priyaranjan Jha +Signed-off-by: Yousuk Seung +Signed-off-by: David S. Miller +Signed-off-by: Greg Kroah-Hartman +--- + net/ipv4/tcp_bbr.c | 4 +++- + 1 file changed, 3 insertions(+), 1 deletion(-) + +--- a/net/ipv4/tcp_bbr.c ++++ b/net/ipv4/tcp_bbr.c +@@ -773,7 +773,9 @@ static void bbr_update_min_rtt(struct so + } + } + } +- bbr->idle_restart = 0; ++ /* Restart after idle ends only once we process a new S/ACK for data */ ++ if (rs->delivered > 0) ++ bbr->idle_restart = 0; + } + + static void bbr_update_model(struct sock *sk, const struct rate_sample *rs) diff --git a/queue-4.9/tg3-fix-vunmap-bug_on-triggered-from-tg3_free_consistent.patch b/queue-4.9/tg3-fix-vunmap-bug_on-triggered-from-tg3_free_consistent.patch new file mode 100644 index 00000000000..9d3555d1101 --- /dev/null +++ b/queue-4.9/tg3-fix-vunmap-bug_on-triggered-from-tg3_free_consistent.patch @@ -0,0 +1,57 @@ +From foo@baz Wed May 16 12:08:47 CEST 2018 +From: Michael Chan +Date: Thu, 3 May 2018 20:04:27 -0400 +Subject: tg3: Fix vunmap() BUG_ON() triggered from tg3_free_consistent(). + +From: Michael Chan + +[ Upstream commit d89a2adb8bfe6f8949ff389acdb9fa298b6e8e12 ] + +tg3_free_consistent() calls dma_free_coherent() to free tp->hw_stats +under spinlock and can trigger BUG_ON() in vunmap() because vunmap() +may sleep. Fix it by removing the spinlock and relying on the +TG3_FLAG_INIT_COMPLETE flag to prevent race conditions between +tg3_get_stats64() and tg3_free_consistent(). TG3_FLAG_INIT_COMPLETE +is always cleared under tp->lock before tg3_free_consistent() +and therefore tg3_get_stats64() can safely access tp->hw_stats +under tp->lock if TG3_FLAG_INIT_COMPLETE is set. + +Fixes: f5992b72ebe0 ("tg3: Fix race condition in tg3_get_stats64().") +Reported-by: Zumeng Chen +Signed-off-by: Michael Chan +Signed-off-by: David S. Miller +Signed-off-by: Greg Kroah-Hartman +--- + drivers/net/ethernet/broadcom/tg3.c | 9 +++++---- + 1 file changed, 5 insertions(+), 4 deletions(-) + +--- a/drivers/net/ethernet/broadcom/tg3.c ++++ b/drivers/net/ethernet/broadcom/tg3.c +@@ -8720,14 +8720,15 @@ static void tg3_free_consistent(struct t + tg3_mem_rx_release(tp); + tg3_mem_tx_release(tp); + +- /* Protect tg3_get_stats64() from reading freed tp->hw_stats. */ +- tg3_full_lock(tp, 0); ++ /* tp->hw_stats can be referenced safely: ++ * 1. under rtnl_lock ++ * 2. or under tp->lock if TG3_FLAG_INIT_COMPLETE is set. ++ */ + if (tp->hw_stats) { + dma_free_coherent(&tp->pdev->dev, sizeof(struct tg3_hw_stats), + tp->hw_stats, tp->stats_mapping); + tp->hw_stats = NULL; + } +- tg3_full_unlock(tp); + } + + /* +@@ -14161,7 +14162,7 @@ static struct rtnl_link_stats64 *tg3_get + struct tg3 *tp = netdev_priv(dev); + + spin_lock_bh(&tp->lock); +- if (!tp->hw_stats) { ++ if (!tp->hw_stats || !tg3_flag(tp, INIT_COMPLETE)) { + *stats = tp->net_stats_prev; + spin_unlock_bh(&tp->lock); + return stats; -- 2.47.3