History

Merge git://git.kernel.org/pub/scm/linux/kernel/git/davem/net

Pull networking fixes from David Miller: 1) Handle init flow failures properly in iwlwifi driver, from Shahar S Matityahu. 2) mac80211 TXQs need to be unscheduled on powersave start, from Felix Fietkau. 3) SKB memory accounting fix in A-MDSU aggregation, from Felix Fietkau. 4) Increase RCU lock hold time in mlx5 FPGA code, from Saeed Mahameed. 5) Avoid checksum complete with XDP in mlx5, also from Saeed. 6) Fix netdev feature clobbering in ibmvnic driver, from Thomas Falcon. 7) Partial sent TLS record leak fix from Jakub Kicinski. 8) Reject zero size iova range in vhost, from Jason Wang. 9) Allow pending work to complete before clcsock release from Karsten Graul. 10) Fix XDP handling max MTU in thunderx, from Matteo Croce. 11) A lot of protocols look at the sa_family field of a sockaddr before validating it's length is large enough, from Tetsuo Handa. 12) Don't write to free'd pointer in qede ptp error path, from Colin Ian King. 13) Have to recompile IP options in ipv4_link_failure because it can be invoked from ARP, from Stephen Suryaputra. 14) Doorbell handling fixes in qed from Denis Bolotin. 15) Revert net-sysfs kobject register leak fix, it causes new problems. From Wang Hai. 16) Spectre v1 fix in ATM code, from Gustavo A. R. Silva. 17) Fix put of BROPT_VLAN_STATS_PER_PORT in bridging code, from Nikolay Aleksandrov. * git://git.kernel.org/pub/scm/linux/kernel/git/davem/net: (111 commits) socket: fix compat SO_RCVTIMEO_NEW/SO_SNDTIMEO_NEW tcp: tcp_grow_window() needs to respect tcp_space() ocelot: Clean up stats update deferred work ocelot: Don't sleep in atomic context (irqs_disabled()) net: bridge: fix netlink export of vlan_stats_per_port option qed: fix spelling mistake "faspath" -> "fastpath" tipc: set sysctl_tipc_rmem and named_timeout right range tipc: fix link established but not in session net: Fix missing meta data in skb with vlan packet net: atm: Fix potential Spectre v1 vulnerabilities net/core: work around section mismatch warning for ptp_classifier net: bridge: fix per-port af_packet sockets bnx2x: fix spelling mistake "dicline" -> "decline" route: Avoid crash from dereferencing NULL rt->from MAINTAINERS: normalize Woojung Huh's email address bonding: fix event handling for stacked bonds Revert "net-sysfs: Fix memory leak in netdev_register_kobject" rtnetlink: fix rtnl_valid_stats_req() nlmsg_len check qed: Fix the DORQ's attentions handling qed: Fix missing DORQ attentions ...

Linus Torvalds committed 7 years ago

Commit 2a3a0282 parents 444fe99 e698642125 files · +1118 −741Browse files

Unified · Split

125 files
  1. Documentation/networking/rxrpc.txt +9 −7
  2. MAINTAINERS +1 −1
  3. drivers/isdn/mISDN/socket.c +2 −2
  4. drivers/net/bonding/bond_main.c +5 −1
  5. drivers/net/ethernet/broadcom/bnx2x/bnx2x_vfpf.c +1 −1
  6. drivers/net/ethernet/cavium/thunder/nicvf_main.c +20 −2
  7. drivers/net/ethernet/freescale/fec_main.c +21 −9
  8. drivers/net/ethernet/ibm/ibmvnic.c +25 −7
  9. drivers/net/ethernet/mellanox/mlx5/core/en.h +1 −0
  10. drivers/net/ethernet/mellanox/mlx5/core/en/reporter_tx.c +8 −3
  11. drivers/net/ethernet/mellanox/mlx5/core/en/tc_tun.c +4 −0
  12. drivers/net/ethernet/mellanox/mlx5/core/en_ethtool.c +2 −1
  13. drivers/net/ethernet/mellanox/mlx5/core/en_main.c +16 −5
  14. drivers/net/ethernet/mellanox/mlx5/core/en_rx.c +75 −19
  15. drivers/net/ethernet/mellanox/mlx5/core/en_stats.c +6 −0
  16. drivers/net/ethernet/mellanox/mlx5/core/en_stats.h +4 −0
  17. drivers/net/ethernet/mellanox/mlx5/core/fpga/tls.c +24 −37
  18. drivers/net/ethernet/mellanox/mlxsw/core.c +3 −3
  19. drivers/net/ethernet/mellanox/mlxsw/spectrum_buffers.c +11 −8
  20. drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c +1 −1
  21. drivers/net/ethernet/mellanox/mlxsw/spectrum_switchdev.c +1 −1
  22. drivers/net/ethernet/mscc/ocelot.c +15 −9
  23. drivers/net/ethernet/neterion/vxge/vxge-config.c +1 −0
  24. drivers/net/ethernet/qlogic/qed/qed.h +5 −2
  25. drivers/net/ethernet/qlogic/qed/qed_dev.c +34 −51
  26. drivers/net/ethernet/qlogic/qed/qed_int.c +63 −20
  27. drivers/net/ethernet/qlogic/qed/qed_int.h +2 −2
  28. drivers/net/ethernet/qlogic/qed/qed_main.c +1 −1
  29. drivers/net/ethernet/qlogic/qed/qed_sriov.c +1 −1
  30. drivers/net/ethernet/qlogic/qede/qede_ptp.c +3 −4
  31. drivers/net/team/team.c +26 −0
  32. drivers/net/wireless/ath/ath10k/htt_rx.c +1 −1
  33. drivers/net/wireless/ath/ath10k/mac.c +2 −2
  34. drivers/net/wireless/ath/ath9k/xmit.c +4 −1
  35. drivers/net/wireless/intel/iwlwifi/cfg/22000.c +22 −8
  36. drivers/net/wireless/intel/iwlwifi/fw/dbg.c +8 −26
  37. drivers/net/wireless/intel/iwlwifi/fw/init.c +0 −1
  38. drivers/net/wireless/intel/iwlwifi/iwl-config.h +2 −1
  39. drivers/net/wireless/intel/iwlwifi/iwl-csr.h +1 −0
  40. drivers/net/wireless/intel/iwlwifi/iwl-trans.h +6 −6
  41. drivers/net/wireless/intel/iwlwifi/mvm/mac80211.c +22 −49
  42. drivers/net/wireless/intel/iwlwifi/mvm/mvm.h +0 −1
  43. drivers/net/wireless/intel/iwlwifi/mvm/sta.c +4 −39
  44. drivers/net/wireless/intel/iwlwifi/mvm/sta.h +2 −5
  45. drivers/net/wireless/intel/iwlwifi/pcie/drv.c +7 −6
  46. drivers/net/wireless/intel/iwlwifi/pcie/internal.h +1 −1
  47. drivers/net/wireless/intel/iwlwifi/pcie/trans.c +8 −3
  48. drivers/net/wireless/intel/iwlwifi/pcie/tx-gen2.c +1 −1
  49. drivers/net/wireless/intel/iwlwifi/pcie/tx.c +1 −1
  50. drivers/net/wireless/mac80211_hwsim.c +15 −4
  51. drivers/net/wireless/mediatek/mt76/mt7603/init.c +2 −0
  52. drivers/net/wireless/mediatek/mt76/mt7603/mac.c +14 −39
  53. drivers/net/wireless/mediatek/mt76/mt7603/main.c +4 −4
  54. drivers/net/wireless/mediatek/mt76/mt7603/mt7603.h +1 −1
  55. drivers/net/wireless/mediatek/mt76/mt76x02_mac.c +8 −6
  56. drivers/net/wireless/ralink/rt2x00/rt2x00.h +0 −1
  57. drivers/net/wireless/ralink/rt2x00/rt2x00mac.c +0 −10
  58. drivers/net/wireless/ralink/rt2x00/rt2x00queue.c +9 −6
  59. drivers/vhost/vhost.c +5 −1
  60. fs/afs/rxrpc.c +17 −7
  61. include/linux/netdevice.h +3 −0
  62. include/net/af_rxrpc.h +3 −1
  63. include/net/cfg80211.h +5 −0
  64. include/net/mac80211.h +34 −29
  65. include/net/netrom.h +1 −1
  66. include/net/sock.h +0 −6
  67. include/net/tls.h +3 −1
  68. net/atm/lec.c +5 −1
  69. net/bluetooth/sco.c +2 −2
  70. net/bridge/br_input.c +14 −9
  71. net/bridge/br_multicast.c +3 −1
  72. net/bridge/br_netlink.c +1 −1
  73. net/core/dev.c +15 −1
  74. net/core/failover.c +3 −3
  75. net/core/filter.c +2 −0
  76. net/core/net-sysfs.c +5 −9
  77. net/core/ptp_classifier.c +4 −3
  78. net/core/rtnetlink.c +1 −1
  79. net/core/skbuff.c +9 −1
  80. net/core/sock.c +2 −2
  81. net/ipv4/fou.c +3 −1
  82. net/ipv4/route.c +15 −1
  83. net/ipv4/tcp_dctcp.c +17 −28
  84. net/ipv4/tcp_input.c +5 −5
  85. net/ipv6/route.c +4 −0
  86. net/ipv6/udp.c +2 −0
  87. net/llc/af_llc.c +1 −2
  88. net/mac80211/driver-ops.h +3 −0
  89. net/mac80211/key.c +4 −5
  90. net/mac80211/mesh_pathtbl.c +1 −1
  91. net/mac80211/rx.c +9 −1
  92. net/mac80211/trace_msg.h +6 −1
  93. net/mac80211/tx.c +23 −30
  94. net/netlink/af_netlink.c +2 −1
  95. net/netrom/af_netrom.c +54 −22
  96. net/netrom/nr_loopback.c +1 −1
  97. net/netrom/nr_route.c +1 −1
  98. net/netrom/sysctl_net_netrom.c +4 −1
  99. net/rds/af_rds.c +3 −0
  100. net/rds/bind.c +2 −0
  101. net/rxrpc/af_rxrpc.c +11 −6
  102. net/rxrpc/ar-internal.h +1 −0
  103. net/rxrpc/conn_event.c +7 −4
  104. net/rxrpc/input.c +12 −6
  105. net/rxrpc/peer_event.c +5 −0
  106. net/rxrpc/sendmsg.c +12 −9
  107. net/sctp/socket.c +2 −1
  108. net/smc/af_smc.c +39 −19
  109. net/smc/smc_close.c +21 −4
  110. net/smc/smc_close.h +1 −0
  111. net/smc/smc_ism.c +5 −0
  112. net/smc/smc_pnet.c +2 −1
  113. net/strparser/strparser.c +5 −7
  114. net/tipc/link.c +2 −0
  115. net/tipc/name_table.c +2 −1
  116. net/tipc/sysctl.c +6 −2
  117. net/tls/tls_device.c +11 −1
  118. net/tls/tls_main.c +24 −0
  119. net/tls/tls_sw.c +1 −14
  120. net/wireless/nl80211.c +12 −6
  121. net/wireless/reg.c +39 −0
  122. net/wireless/scan.c +1 −2
  123. net/wireless/util.c +4 −2
  124. tools/testing/selftests/drivers/net/mlxsw/rtnetlink.sh +20 −0
  125. tools/testing/selftests/net/fib_tests.sh +40 −54
Documentation/networking/rxrpc.txt+9−7
@@ -1009,16 +1009,18 @@ The kernel interface functions are as follows: · 1008 unmodified lines
10091009
10101010 (*) Check call still alive.
10111011
1012− u32 rxrpc_kernel_check_life(struct socket *sock,
1013− struct rxrpc_call *call);
1012+ bool rxrpc_kernel_check_life(struct socket *sock,
1013+ struct rxrpc_call *call,
1014+ u32 *_life);
10141015 void rxrpc_kernel_probe_life(struct socket *sock,
10151016 struct rxrpc_call *call);
10161017
1017− The first function returns a number that is updated when ACKs are received
1018− from the peer (notably including PING RESPONSE ACKs which we can elicit by
1019− sending PING ACKs to see if the call still exists on the server). The
1020− caller should compare the numbers of two calls to see if the call is still
1021− alive after waiting for a suitable interval.
1018+ The first function passes back in *_life a number that is updated when
1019+ ACKs are received from the peer (notably including PING RESPONSE ACKs
1020+ which we can elicit by sending PING ACKs to see if the call still exists
1021+ on the server). The caller should compare the numbers of two calls to see
1022+ if the call is still alive after waiting for a suitable interval. It also
1023+ returns true as long as the call hasn't yet reached the completed state.
10221024
10231025 This allows the caller to work out if the server is still contactable and
10241026 if the call is still alive on the server while waiting for the server to
MAINTAINERS+1−1
@@ -10145,7 +10145,7 @@ F: drivers/spi/spi-at91-usart.c · 10144 unmodified lines
1014510145 F: Documentation/devicetree/bindings/mfd/atmel-usart.txt
1014610146
1014710147 MICROCHIP KSZ SERIES ETHERNET SWITCH DRIVER
10148−M: Woojung Huh <Woojung.Huh@microchip.com>
10148+M: Woojung Huh <woojung.huh@microchip.com>
1014910149 M: Microchip Linux Driver Support <UNGLinuxDriver@microchip.com>
1015010150 L: netdev@vger.kernel.org
1015110151 S: Maintained
drivers/isdn/mISDN/socket.c+2−2
@@ -710,10 +710,10 @@ base_sock_bind(struct socket *sock, struct sockaddr *addr, int addr_len) · 709 unmodified lines
710710 struct sock *sk = sock->sk;
711711 int err = 0;
712712
713− if (!maddr || maddr->family != AF_ISDN)
713+ if (addr_len < sizeof(struct sockaddr_mISDN))
714714 return -EINVAL;
715715
716− if (addr_len < sizeof(struct sockaddr_mISDN))
716+ if (!maddr || maddr->family != AF_ISDN)
717717 return -EINVAL;
718718
719719 lock_sock(sk);
drivers/net/bonding/bond_main.c+5−1
@@ -3213,8 +3213,12 @@ static int bond_netdev_event(struct notifier_block *this, · 3212 unmodified lines
32133213 return NOTIFY_DONE;
32143214
32153215 if (event_dev->flags & IFF_MASTER) {
3216+ int ret;
3217+
32163218 netdev_dbg(event_dev, "IFF_MASTER\n");
3217− return bond_master_netdev_event(event, event_dev);
3219+ ret = bond_master_netdev_event(event, event_dev);
3220+ if (ret != NOTIFY_DONE)
3221+ return ret;
32183222 }
32193223
32203224 if (event_dev->flags & IFF_SLAVE) {
drivers/net/ethernet/broadcom/bnx2x/bnx2x_vfpf.c+1−1
@@ -957,7 +957,7 @@ int bnx2x_vfpf_update_vlan(struct bnx2x *bp, u16 vid, u8 vf_qid, bool add) · 956 unmodified lines
957957 bnx2x_sample_bulletin(bp);
958958
959959 if (bp->shadow_bulletin.content.valid_bitmap & 1 << VLAN_VALID) {
960− BNX2X_ERR("Hypervisor will dicline the request, avoiding\n");
960+ BNX2X_ERR("Hypervisor will decline the request, avoiding\n");
961961 rc = -EINVAL;
962962 goto out;
963963 }
drivers/net/ethernet/cavium/thunder/nicvf_main.c+20−2
@@ -32,6 +32,13 @@ · 31 unmodified lines
3232 #define DRV_NAME "nicvf"
3333 #define DRV_VERSION "1.0"
3434
35+/* NOTE: Packets bigger than 1530 are split across multiple pages and XDP needs
36+ * the buffer to be contiguous. Allow XDP to be set up only if we don't exceed
37+ * this value, keeping headroom for the 14 byte Ethernet header and two
38+ * VLAN tags (for QinQ)
39+ */
40+#define MAX_XDP_MTU (1530 - ETH_HLEN - VLAN_HLEN * 2)
41+
3542 /* Supported devices */
3643 static const struct pci_device_id nicvf_id_table[] = {
3744 { PCI_DEVICE_SUB(PCI_VENDOR_ID_CAVIUM,
@@ -1582,6 +1589,15 @@ static int nicvf_change_mtu(struct net_device *netdev, int new_mtu) · 1544 unmodified lines
15821589 struct nicvf *nic = netdev_priv(netdev);
15831590 int orig_mtu = netdev->mtu;
15841591
1592+ /* For now just support only the usual MTU sized frames,
1593+ * plus some headroom for VLAN, QinQ.
1594+ */
1595+ if (nic->xdp_prog && new_mtu > MAX_XDP_MTU) {
1596+ netdev_warn(netdev, "Jumbo frames not yet supported with XDP, current MTU %d.\n",
1597+ netdev->mtu);
1598+ return -EINVAL;
1599+ }
1600+
15851601 netdev->mtu = new_mtu;
15861602
15871603 if (!netif_running(netdev))
@@ -1830,8 +1846,10 @@ static int nicvf_xdp_setup(struct nicvf *nic, struct bpf_prog *prog) · 242 unmodified lines
18301846 bool bpf_attached = false;
18311847 int ret = 0;
18321848
1833− /* For now just support only the usual MTU sized frames */
1834− if (prog && (dev->mtu > 1500)) {
1849+ /* For now just support only the usual MTU sized frames,
1850+ * plus some headroom for VLAN, QinQ.
1851+ */
1852+ if (prog && dev->mtu > MAX_XDP_MTU) {
18351853 netdev_warn(dev, "Jumbo frames not yet supported with XDP, current MTU %d.\n",
18361854 dev->mtu);
18371855 return -EOPNOTSUPP;
drivers/net/ethernet/freescale/fec_main.c+21−9
@@ -1840,13 +1840,9 @@ static int fec_enet_clk_enable(struct net_device *ndev, bool enable) · 1839 unmodified lines
18401840 int ret;
18411841
18421842 if (enable) {
1843− ret = clk_prepare_enable(fep->clk_ahb);
1844− if (ret)
1845− return ret;
1846−
18471843 ret = clk_prepare_enable(fep->clk_enet_out);
18481844 if (ret)
1849− goto failed_clk_enet_out;
1845+ return ret;
18501846
18511847 if (fep->clk_ptp) {
18521848 mutex_lock(&fep->ptp_clk_mutex);
@@ -1866,7 +1862,6 @@ static int fec_enet_clk_enable(struct net_device *ndev, bool enable) · 13 unmodified lines
18661862
18671863 phy_reset_after_clk_enable(ndev->phydev);
18681864 } else {
1869− clk_disable_unprepare(fep->clk_ahb);
18701865 clk_disable_unprepare(fep->clk_enet_out);
18711866 if (fep->clk_ptp) {
18721867 mutex_lock(&fep->ptp_clk_mutex);
@@ -1885,8 +1880,6 @@ failed_clk_ref: · 12 unmodified lines
18851880 failed_clk_ptp:
18861881 if (fep->clk_enet_out)
18871882 clk_disable_unprepare(fep->clk_enet_out);
1888−failed_clk_enet_out:
1889− clk_disable_unprepare(fep->clk_ahb);
18901883
18911884 return ret;
18921885 }
@@ -3470,6 +3463,9 @@ fec_probe(struct platform_device *pdev) · 1577 unmodified lines
34703463 ret = clk_prepare_enable(fep->clk_ipg);
34713464 if (ret)
34723465 goto failed_clk_ipg;
3466+ ret = clk_prepare_enable(fep->clk_ahb);
3467+ if (ret)
3468+ goto failed_clk_ahb;
34733469
34743470 fep->reg_phy = devm_regulator_get_optional(&pdev->dev, "phy");
34753471 if (!IS_ERR(fep->reg_phy)) {
@@ -3563,6 +3559,9 @@ failed_reset: · 87 unmodified lines
35633559 pm_runtime_put(&pdev->dev);
35643560 pm_runtime_disable(&pdev->dev);
35653561 failed_regulator:
3562+ clk_disable_unprepare(fep->clk_ahb);
3563+failed_clk_ahb:
3564+ clk_disable_unprepare(fep->clk_ipg);
35663565 failed_clk_ipg:
35673566 fec_enet_clk_enable(ndev, false);
35683567 failed_clk:
@@ -3686,6 +3685,7 @@ static int __maybe_unused fec_runtime_suspend(struct device *dev) · 117 unmodified lines
36863685 struct net_device *ndev = dev_get_drvdata(dev);
36873686 struct fec_enet_private *fep = netdev_priv(ndev);
36883687
3688+ clk_disable_unprepare(fep->clk_ahb);
36893689 clk_disable_unprepare(fep->clk_ipg);
36903690
36913691 return 0;
@@ -3695,8 +3695,20 @@ static int __maybe_unused fec_runtime_resume(struct device *dev) · 3 unmodified lines
36953695 {
36963696 struct net_device *ndev = dev_get_drvdata(dev);
36973697 struct fec_enet_private *fep = netdev_priv(ndev);
3698+ int ret;
36983699
3699− return clk_prepare_enable(fep->clk_ipg);
3700+ ret = clk_prepare_enable(fep->clk_ahb);
3701+ if (ret)
3702+ return ret;
3703+ ret = clk_prepare_enable(fep->clk_ipg);
3704+ if (ret)
3705+ goto failed_clk_ipg;
3706+
3707+ return 0;
3708+
3709+failed_clk_ipg:
3710+ clk_disable_unprepare(fep->clk_ahb);
3711+ return ret;
37003712 }
37013713
37023714 static const struct dev_pm_ops fec_pm_ops = {
drivers/net/ethernet/ibm/ibmvnic.c+25−7
@@ -3762,6 +3762,7 @@ static void handle_query_ip_offload_rsp(struct ibmvnic_adapter *adapter) · 3761 unmodified lines
37623762 {
37633763 struct device *dev = &adapter->vdev->dev;
37643764 struct ibmvnic_query_ip_offload_buffer *buf = &adapter->ip_offload_buf;
3765+ netdev_features_t old_hw_features = 0;
37653766 union ibmvnic_crq crq;
37663767 int i;
37673768
@@ -3837,24 +3838,41 @@ static void handle_query_ip_offload_rsp(struct ibmvnic_adapter *adapter) · 69 unmodified lines
38373838 adapter->ip_offload_ctrl.large_rx_ipv4 = 0;
38383839 adapter->ip_offload_ctrl.large_rx_ipv6 = 0;
38393840
3840− adapter->netdev->features = NETIF_F_SG | NETIF_F_GSO;
3841+ if (adapter->state != VNIC_PROBING) {
3842+ old_hw_features = adapter->netdev->hw_features;
3843+ adapter->netdev->hw_features = 0;
3844+ }
3845+
3846+ adapter->netdev->hw_features = NETIF_F_SG | NETIF_F_GSO | NETIF_F_GRO;
38413847
38423848 if (buf->tcp_ipv4_chksum || buf->udp_ipv4_chksum)
3843− adapter->netdev->features |= NETIF_F_IP_CSUM;
3849+ adapter->netdev->hw_features |= NETIF_F_IP_CSUM;
38443850
38453851 if (buf->tcp_ipv6_chksum || buf->udp_ipv6_chksum)
3846− adapter->netdev->features |= NETIF_F_IPV6_CSUM;
3852+ adapter->netdev->hw_features |= NETIF_F_IPV6_CSUM;
38473853
38483854 if ((adapter->netdev->features &
38493855 (NETIF_F_IP_CSUM | NETIF_F_IPV6_CSUM)))
3850− adapter->netdev->features |= NETIF_F_RXCSUM;
3856+ adapter->netdev->hw_features |= NETIF_F_RXCSUM;
38513857
38523858 if (buf->large_tx_ipv4)
3853− adapter->netdev->features |= NETIF_F_TSO;
3859+ adapter->netdev->hw_features |= NETIF_F_TSO;
38543860 if (buf->large_tx_ipv6)
3855− adapter->netdev->features |= NETIF_F_TSO6;
3861+ adapter->netdev->hw_features |= NETIF_F_TSO6;
38563862
3857− adapter->netdev->hw_features |= adapter->netdev->features;
3863+ if (adapter->state == VNIC_PROBING) {
3864+ adapter->netdev->features |= adapter->netdev->hw_features;
3865+ } else if (old_hw_features != adapter->netdev->hw_features) {
3866+ netdev_features_t tmp = 0;
3867+
3868+ /* disable features no longer supported */
3869+ adapter->netdev->features &= adapter->netdev->hw_features;
3870+ /* turn on features now supported if previously enabled */
3871+ tmp = (old_hw_features ^ adapter->netdev->hw_features) &
3872+ adapter->netdev->hw_features;
3873+ adapter->netdev->features |=
3874+ tmp & adapter->netdev->wanted_features;
3875+ }
38583876
38593877 memset(&crq, 0, sizeof(crq));
38603878 crq.control_ip_offload.first = IBMVNIC_CRQ_CMD;
drivers/net/ethernet/mellanox/mlx5/core/en.h+1−0
@@ -858,6 +858,7 @@ void mlx5e_close_channels(struct mlx5e_channels *chs); · 857 unmodified lines
858858 * switching channels
859859 */
860860 typedef int (*mlx5e_fp_hw_modify)(struct mlx5e_priv *priv);
861+int mlx5e_safe_reopen_channels(struct mlx5e_priv *priv);
861862 int mlx5e_safe_switch_channels(struct mlx5e_priv *priv,
862863 struct mlx5e_channels *new_chs,
863864 mlx5e_fp_hw_modify hw_modify);
drivers/net/ethernet/mellanox/mlx5/core/en/reporter_tx.c+8−3
@@ -186,12 +186,17 @@ static int mlx5e_tx_reporter_recover_from_ctx(struct mlx5e_tx_err_ctx *err_ctx) · 185 unmodified lines
186186
187187 static int mlx5e_tx_reporter_recover_all(struct mlx5e_priv *priv)
188188 {
189− int err;
189+ int err = 0;
190190
191191 rtnl_lock();
192192 mutex_lock(&priv->state_lock);
193− mlx5e_close_locked(priv->netdev);
194− err = mlx5e_open_locked(priv->netdev);
193+
194+ if (!test_bit(MLX5E_STATE_OPENED, &priv->state))
195+ goto out;
196+
197+ err = mlx5e_safe_reopen_channels(priv);
198+
199+out:
195200 mutex_unlock(&priv->state_lock);
196201 rtnl_unlock();
197202
drivers/net/ethernet/mellanox/mlx5/core/en/tc_tun.c+4−0
@@ -39,6 +39,10 @@ static int get_route_and_out_devs(struct mlx5e_priv *priv, · 38 unmodified lines
3939 return -EOPNOTSUPP;
4040 }
4141
42+ if (!(mlx5e_eswitch_rep(*out_dev) &&
43+ mlx5e_is_uplink_rep(netdev_priv(*out_dev))))
44+ return -EOPNOTSUPP;
45+
4246 return 0;
4347 }
4448
drivers/net/ethernet/mellanox/mlx5/core/en_ethtool.c+2−1
@@ -1768,7 +1768,8 @@ static int set_pflag_rx_no_csum_complete(struct net_device *netdev, bool enable) · 1767 unmodified lines
17681768 struct mlx5e_channel *c;
17691769 int i;
17701770
1771− if (!test_bit(MLX5E_STATE_OPENED, &priv->state))
1771+ if (!test_bit(MLX5E_STATE_OPENED, &priv->state) ||
1772+ priv->channels.params.xdp_prog)
17721773 return 0;
17731774
17741775 for (i = 0; i < channels->num; i++) {
drivers/net/ethernet/mellanox/mlx5/core/en_main.c+16−5
@@ -951,7 +951,11 @@ static int mlx5e_open_rq(struct mlx5e_channel *c, · 950 unmodified lines
951951 if (params->rx_dim_enabled)
952952 __set_bit(MLX5E_RQ_STATE_AM, &c->rq.state);
953953
954− if (MLX5E_GET_PFLAG(params, MLX5E_PFLAG_RX_NO_CSUM_COMPLETE))
954+ /* We disable csum_complete when XDP is enabled since
955+ * XDP programs might manipulate packets which will render
956+ * skb->checksum incorrect.
957+ */
958+ if (MLX5E_GET_PFLAG(params, MLX5E_PFLAG_RX_NO_CSUM_COMPLETE) || c->xdp)
955959 __set_bit(MLX5E_RQ_STATE_NO_CSUM_COMPLETE, &c->rq.state);
956960
957961 return 0;
@@ -2937,6 +2941,14 @@ int mlx5e_safe_switch_channels(struct mlx5e_priv *priv, · 1979 unmodified lines
29372941 return 0;
29382942 }
29392943
2944+int mlx5e_safe_reopen_channels(struct mlx5e_priv *priv)
2945+{
2946+ struct mlx5e_channels new_channels = {};
2947+
2948+ new_channels.params = priv->channels.params;
2949+ return mlx5e_safe_switch_channels(priv, &new_channels, NULL);
2950+}
2951+
29402952 void mlx5e_timestamp_init(struct mlx5e_priv *priv)
29412953 {
29422954 priv->tstamp.tx_type = HWTSTAMP_TX_OFF;
@@ -4161,11 +4173,10 @@ static void mlx5e_tx_timeout_work(struct work_struct *work) · 1218 unmodified lines
41614173 if (!report_failed)
41624174 goto unlock;
41634175
4164− mlx5e_close_locked(priv->netdev);
4165− err = mlx5e_open_locked(priv->netdev);
4176+ err = mlx5e_safe_reopen_channels(priv);
41664177 if (err)
41674178 netdev_err(priv->netdev,
4168− "mlx5e_open_locked failed recovering from a tx_timeout, err(%d).\n",
4179+ "mlx5e_safe_reopen_channels failed recovering from a tx_timeout, err(%d).\n",
41694180 err);
41704181
41714182 unlock:
@@ -4553,7 +4564,7 @@ void mlx5e_build_rss_params(struct mlx5e_rss_params *rss_params, · 381 unmodified lines
45534564 {
45544565 enum mlx5e_traffic_types tt;
45554566
4556− rss_params->hfunc = ETH_RSS_HASH_XOR;
4567+ rss_params->hfunc = ETH_RSS_HASH_TOP;
45574568 netdev_rss_key_fill(rss_params->toeplitz_hash_key,
45584569 sizeof(rss_params->toeplitz_hash_key));
45594570 mlx5e_build_default_indir_rqt(rss_params->indirection_rqt,
drivers/net/ethernet/mellanox/mlx5/core/en_rx.c+75−19
@@ -692,7 +692,14 @@ static inline bool is_last_ethertype_ip(struct sk_buff *skb, int *network_depth, · 691 unmodified lines
692692 {
693693 *proto = ((struct ethhdr *)skb->data)->h_proto;
694694 *proto = __vlan_get_protocol(skb, *proto, network_depth);
695− return (*proto == htons(ETH_P_IP) || *proto == htons(ETH_P_IPV6));
695+
696+ if (*proto == htons(ETH_P_IP))
697+ return pskb_may_pull(skb, *network_depth + sizeof(struct iphdr));
698+
699+ if (*proto == htons(ETH_P_IPV6))
700+ return pskb_may_pull(skb, *network_depth + sizeof(struct ipv6hdr));
701+
702+ return false;
696703 }
697704
698705 static inline void mlx5e_enable_ecn(struct mlx5e_rq *rq, struct sk_buff *skb)
@@ -712,17 +719,6 @@ static inline void mlx5e_enable_ecn(struct mlx5e_rq *rq, struct sk_buff *skb) · 13 unmodified lines
712719 rq->stats->ecn_mark += !!rc;
713720 }
714721
715−static u32 mlx5e_get_fcs(const struct sk_buff *skb)
716−{
717− const void *fcs_bytes;
718− u32 _fcs_bytes;
719−
720− fcs_bytes = skb_header_pointer(skb, skb->len - ETH_FCS_LEN,
721− ETH_FCS_LEN, &_fcs_bytes);
722−
723− return __get_unaligned_cpu32(fcs_bytes);
724−}
725−
726722 static u8 get_ip_proto(struct sk_buff *skb, int network_depth, __be16 proto)
727723 {
728724 void *ip_p = skb->data + network_depth;
@@ -733,6 +729,68 @@ static u8 get_ip_proto(struct sk_buff *skb, int network_depth, __be16 proto) · 4 unmodified lines
733729
734730 #define short_frame(size) ((size) <= ETH_ZLEN + ETH_FCS_LEN)
735731
732+#define MAX_PADDING 8
733+
734+static void
735+tail_padding_csum_slow(struct sk_buff *skb, int offset, int len,
736+ struct mlx5e_rq_stats *stats)
737+{
738+ stats->csum_complete_tail_slow++;
739+ skb->csum = csum_block_add(skb->csum,
740+ skb_checksum(skb, offset, len, 0),
741+ offset);
742+}
743+
744+static void
745+tail_padding_csum(struct sk_buff *skb, int offset,
746+ struct mlx5e_rq_stats *stats)
747+{
748+ u8 tail_padding[MAX_PADDING];
749+ int len = skb->len - offset;
750+ void *tail;
751+
752+ if (unlikely(len > MAX_PADDING)) {
753+ tail_padding_csum_slow(skb, offset, len, stats);
754+ return;
755+ }
756+
757+ tail = skb_header_pointer(skb, offset, len, tail_padding);
758+ if (unlikely(!tail)) {
759+ tail_padding_csum_slow(skb, offset, len, stats);
760+ return;
761+ }
762+
763+ stats->csum_complete_tail++;
764+ skb->csum = csum_block_add(skb->csum, csum_partial(tail, len, 0), offset);
765+}
766+
767+static void
768+mlx5e_skb_padding_csum(struct sk_buff *skb, int network_depth, __be16 proto,
769+ struct mlx5e_rq_stats *stats)
770+{
771+ struct ipv6hdr *ip6;
772+ struct iphdr *ip4;
773+ int pkt_len;
774+
775+ switch (proto) {
776+ case htons(ETH_P_IP):
777+ ip4 = (struct iphdr *)(skb->data + network_depth);
778+ pkt_len = network_depth + ntohs(ip4->tot_len);
779+ break;
780+ case htons(ETH_P_IPV6):
781+ ip6 = (struct ipv6hdr *)(skb->data + network_depth);
782+ pkt_len = network_depth + sizeof(*ip6) + ntohs(ip6->payload_len);
783+ break;
784+ default:
785+ return;
786+ }
787+
788+ if (likely(pkt_len >= skb->len))
789+ return;
790+
791+ tail_padding_csum(skb, pkt_len, stats);
792+}
793+
736794 static inline void mlx5e_handle_csum(struct net_device *netdev,
737795 struct mlx5_cqe64 *cqe,
738796 struct mlx5e_rq *rq,
@@ -752,7 +810,8 @@ static inline void mlx5e_handle_csum(struct net_device *netdev, · 13 unmodified lines
752810 return;
753811 }
754812
755− if (unlikely(test_bit(MLX5E_RQ_STATE_NO_CSUM_COMPLETE, &rq->state)))
813+ /* True when explicitly set via priv flag, or XDP prog is loaded */
814+ if (test_bit(MLX5E_RQ_STATE_NO_CSUM_COMPLETE, &rq->state))
756815 goto csum_unnecessary;
757816
758817 /* CQE csum doesn't cover padding octets in short ethernet
@@ -780,18 +839,15 @@ static inline void mlx5e_handle_csum(struct net_device *netdev, · 21 unmodified lines
780839 skb->csum = csum_partial(skb->data + ETH_HLEN,
781840 network_depth - ETH_HLEN,
782841 skb->csum);
783− if (unlikely(netdev->features & NETIF_F_RXFCS))
784− skb->csum = csum_block_add(skb->csum,
785− (__force __wsum)mlx5e_get_fcs(skb),
786− skb->len - ETH_FCS_LEN);
842+
843+ mlx5e_skb_padding_csum(skb, network_depth, proto, stats);
787844 stats->csum_complete++;
788845 return;
789846 }
790847
791848 csum_unnecessary:
792849 if (likely((cqe->hds_ip_ext & CQE_L3_OK) &&
793− ((cqe->hds_ip_ext & CQE_L4_OK) ||
794− (get_cqe_l4_hdr_type(cqe) == CQE_L4_HDR_TYPE_NONE)))) {
850+ (cqe->hds_ip_ext & CQE_L4_OK))) {
795851 skb->ip_summed = CHECKSUM_UNNECESSARY;
796852 if (cqe_is_tunneled(cqe)) {
797853 skb->csum_level = 1;
drivers/net/ethernet/mellanox/mlx5/core/en_stats.c+6−0
@@ -59,6 +59,8 @@ static const struct counter_desc sw_stats_desc[] = { · 58 unmodified lines
5959 { MLX5E_DECLARE_STAT(struct mlx5e_sw_stats, rx_csum_unnecessary) },
6060 { MLX5E_DECLARE_STAT(struct mlx5e_sw_stats, rx_csum_none) },
6161 { MLX5E_DECLARE_STAT(struct mlx5e_sw_stats, rx_csum_complete) },
62+ { MLX5E_DECLARE_STAT(struct mlx5e_sw_stats, rx_csum_complete_tail) },
63+ { MLX5E_DECLARE_STAT(struct mlx5e_sw_stats, rx_csum_complete_tail_slow) },
6264 { MLX5E_DECLARE_STAT(struct mlx5e_sw_stats, rx_csum_unnecessary_inner) },
6365 { MLX5E_DECLARE_STAT(struct mlx5e_sw_stats, rx_xdp_drop) },
6466 { MLX5E_DECLARE_STAT(struct mlx5e_sw_stats, rx_xdp_redirect) },
@@ -151,6 +153,8 @@ static void mlx5e_grp_sw_update_stats(struct mlx5e_priv *priv) · 86 unmodified lines
151153 s->rx_removed_vlan_packets += rq_stats->removed_vlan_packets;
152154 s->rx_csum_none += rq_stats->csum_none;
153155 s->rx_csum_complete += rq_stats->csum_complete;
156+ s->rx_csum_complete_tail += rq_stats->csum_complete_tail;
157+ s->rx_csum_complete_tail_slow += rq_stats->csum_complete_tail_slow;
154158 s->rx_csum_unnecessary += rq_stats->csum_unnecessary;
155159 s->rx_csum_unnecessary_inner += rq_stats->csum_unnecessary_inner;
156160 s->rx_xdp_drop += rq_stats->xdp_drop;
@@ -1190,6 +1194,8 @@ static const struct counter_desc rq_stats_desc[] = { · 1033 unmodified lines
11901194 { MLX5E_DECLARE_RX_STAT(struct mlx5e_rq_stats, packets) },
11911195 { MLX5E_DECLARE_RX_STAT(struct mlx5e_rq_stats, bytes) },
11921196 { MLX5E_DECLARE_RX_STAT(struct mlx5e_rq_stats, csum_complete) },
1197+ { MLX5E_DECLARE_RX_STAT(struct mlx5e_rq_stats, csum_complete_tail) },
1198+ { MLX5E_DECLARE_RX_STAT(struct mlx5e_rq_stats, csum_complete_tail_slow) },
11931199 { MLX5E_DECLARE_RX_STAT(struct mlx5e_rq_stats, csum_unnecessary) },
11941200 { MLX5E_DECLARE_RX_STAT(struct mlx5e_rq_stats, csum_unnecessary_inner) },
11951201 { MLX5E_DECLARE_RX_STAT(struct mlx5e_rq_stats, csum_none) },
drivers/net/ethernet/mellanox/mlx5/core/en_stats.h+4−0
@@ -71,6 +71,8 @@ struct mlx5e_sw_stats { · 70 unmodified lines
7171 u64 rx_csum_unnecessary;
7272 u64 rx_csum_none;
7373 u64 rx_csum_complete;
74+ u64 rx_csum_complete_tail;
75+ u64 rx_csum_complete_tail_slow;
7476 u64 rx_csum_unnecessary_inner;
7577 u64 rx_xdp_drop;
7678 u64 rx_xdp_redirect;
@@ -181,6 +183,8 @@ struct mlx5e_rq_stats { · 104 unmodified lines
181183 u64 packets;
182184 u64 bytes;
183185 u64 csum_complete;
186+ u64 csum_complete_tail;
187+ u64 csum_complete_tail_slow;
184188 u64 csum_unnecessary;
185189 u64 csum_unnecessary_inner;
186190 u64 csum_none;
drivers/net/ethernet/mellanox/mlx5/core/fpga/tls.c+24−37
@@ -148,14 +148,16 @@ static int mlx5_fpga_tls_alloc_swid(struct idr *idr, spinlock_t *idr_spinlock, · 147 unmodified lines
148148 return ret;
149149 }
150150
151−static void mlx5_fpga_tls_release_swid(struct idr *idr,
152− spinlock_t *idr_spinlock, u32 swid)
151+static void *mlx5_fpga_tls_release_swid(struct idr *idr,
152+ spinlock_t *idr_spinlock, u32 swid)
153153 {
154154 unsigned long flags;
155+ void *ptr;
155156
156157 spin_lock_irqsave(idr_spinlock, flags);
157− idr_remove(idr, swid);
158+ ptr = idr_remove(idr, swid);
158159 spin_unlock_irqrestore(idr_spinlock, flags);
160+ return ptr;
159161 }
160162
161163 static void mlx_tls_kfree_complete(struct mlx5_fpga_conn *conn,
@@ -165,20 +167,12 @@ static void mlx_tls_kfree_complete(struct mlx5_fpga_conn *conn, · 3 unmodified lines
165167 kfree(buf);
166168 }
167169
168−struct mlx5_teardown_stream_context {
169− struct mlx5_fpga_tls_command_context cmd;
170− u32 swid;
171−};
172−
173170 static void
174171 mlx5_fpga_tls_teardown_completion(struct mlx5_fpga_conn *conn,
175172 struct mlx5_fpga_device *fdev,
176173 struct mlx5_fpga_tls_command_context *cmd,
177174 struct mlx5_fpga_dma_buf *resp)
178175 {
179− struct mlx5_teardown_stream_context *ctx =
180− container_of(cmd, struct mlx5_teardown_stream_context, cmd);
181−
182176 if (resp) {
183177 u32 syndrome = MLX5_GET(tls_resp, resp->sg[0].data, syndrome);
184178
@@ -186,14 +180,6 @@ mlx5_fpga_tls_teardown_completion(struct mlx5_fpga_conn *conn, · 1 unmodified line
186180 mlx5_fpga_err(fdev,
187181 "Teardown stream failed with syndrome = %d",
188182 syndrome);
189− else if (MLX5_GET(tls_cmd, cmd->buf.sg[0].data, direction_sx))
190− mlx5_fpga_tls_release_swid(&fdev->tls->tx_idr,
191− &fdev->tls->tx_idr_spinlock,
192− ctx->swid);
193− else
194− mlx5_fpga_tls_release_swid(&fdev->tls->rx_idr,
195− &fdev->tls->rx_idr_spinlock,
196− ctx->swid);
197183 }
198184 mlx5_fpga_tls_put_command_ctx(cmd);
199185 }
@@ -217,22 +203,22 @@ int mlx5_fpga_tls_resync_rx(struct mlx5_core_dev *mdev, u32 handle, u32 seq, · 17 unmodified lines
217203 void *cmd;
218204 int ret;
219205
220− rcu_read_lock();
221− flow = idr_find(&mdev->fpga->tls->rx_idr, ntohl(handle));
222− rcu_read_unlock();
223−
224− if (!flow) {
225− WARN_ONCE(1, "Received NULL pointer for handle\n");
226− return -EINVAL;
227− }
228−
229206 buf = kzalloc(size, GFP_ATOMIC);
230207 if (!buf)
231208 return -ENOMEM;
232209
233210 cmd = (buf + 1);
234211
212+ rcu_read_lock();
213+ flow = idr_find(&mdev->fpga->tls->rx_idr, ntohl(handle));
214+ if (unlikely(!flow)) {
215+ rcu_read_unlock();
216+ WARN_ONCE(1, "Received NULL pointer for handle\n");
217+ kfree(buf);
218+ return -EINVAL;
219+ }
235220 mlx5_fpga_tls_flow_to_cmd(flow, cmd);
221+ rcu_read_unlock();
236222
237223 MLX5_SET(tls_cmd, cmd, swid, ntohl(handle));
238224 MLX5_SET64(tls_cmd, cmd, tls_rcd_sn, be64_to_cpu(rcd_sn));
@@ -253,7 +239,7 @@ int mlx5_fpga_tls_resync_rx(struct mlx5_core_dev *mdev, u32 handle, u32 seq, · 14 unmodified lines
253239 static void mlx5_fpga_tls_send_teardown_cmd(struct mlx5_core_dev *mdev,
254240 void *flow, u32 swid, gfp_t flags)
255241 {
256− struct mlx5_teardown_stream_context *ctx;
242+ struct mlx5_fpga_tls_command_context *ctx;
257243 struct mlx5_fpga_dma_buf *buf;
258244 void *cmd;
259245
@@ -261,7 +247,7 @@ static void mlx5_fpga_tls_send_teardown_cmd(struct mlx5_core_dev *mdev, · 1 unmodified line
261247 if (!ctx)
262248 return;
263249
264− buf = &ctx->cmd.buf;
250+ buf = &ctx->buf;
265251 cmd = (ctx + 1);
266252 MLX5_SET(tls_cmd, cmd, command_type, CMD_TEARDOWN_STREAM);
267253 MLX5_SET(tls_cmd, cmd, swid, swid);
@@ -272,8 +258,7 @@ static void mlx5_fpga_tls_send_teardown_cmd(struct mlx5_core_dev *mdev, · 4 unmodified lines
272258 buf->sg[0].data = cmd;
273259 buf->sg[0].size = MLX5_TLS_COMMAND_SIZE;
274260
275− ctx->swid = swid;
276− mlx5_fpga_tls_cmd_send(mdev->fpga, &ctx->cmd,
261+ mlx5_fpga_tls_cmd_send(mdev->fpga, ctx,
277262 mlx5_fpga_tls_teardown_completion);
278263 }
279264
@@ -283,13 +268,14 @@ void mlx5_fpga_tls_del_flow(struct mlx5_core_dev *mdev, u32 swid, · 3 unmodified lines
283268 struct mlx5_fpga_tls *tls = mdev->fpga->tls;
284269 void *flow;
285270
286− rcu_read_lock();
287271 if (direction_sx)
288− flow = idr_find(&tls->tx_idr, swid);
272+ flow = mlx5_fpga_tls_release_swid(&tls->tx_idr,
273+ &tls->tx_idr_spinlock,
274+ swid);
289275 else
290− flow = idr_find(&tls->rx_idr, swid);
291−
292− rcu_read_unlock();
276+ flow = mlx5_fpga_tls_release_swid(&tls->rx_idr,
277+ &tls->rx_idr_spinlock,
278+ swid);
293279
294280 if (!flow) {
295281 mlx5_fpga_err(mdev->fpga, "No flow information for swid %u\n",
@@ -297,6 +283,7 @@ void mlx5_fpga_tls_del_flow(struct mlx5_core_dev *mdev, u32 swid, · 1 unmodified line
297283 return;
298284 }
299285
286+ synchronize_rcu(); /* before kfree(flow) */
300287 mlx5_fpga_tls_send_teardown_cmd(mdev, flow, swid, flags);
301288 }
302289
drivers/net/ethernet/mellanox/mlxsw/core.c+3−3
@@ -568,7 +568,7 @@ static int mlxsw_emad_init(struct mlxsw_core *mlxsw_core) · 567 unmodified lines
568568 if (!(mlxsw_core->bus->features & MLXSW_BUS_F_TXRX))
569569 return 0;
570570
571− emad_wq = alloc_workqueue("mlxsw_core_emad", WQ_MEM_RECLAIM, 0);
571+ emad_wq = alloc_workqueue("mlxsw_core_emad", 0, 0);
572572 if (!emad_wq)
573573 return -ENOMEM;
574574 mlxsw_core->emad_wq = emad_wq;
@@ -1958,10 +1958,10 @@ static int __init mlxsw_core_module_init(void) · 1383 unmodified lines
19581958 {
19591959 int err;
19601960
1961− mlxsw_wq = alloc_workqueue(mlxsw_core_driver_name, WQ_MEM_RECLAIM, 0);
1961+ mlxsw_wq = alloc_workqueue(mlxsw_core_driver_name, 0, 0);
19621962 if (!mlxsw_wq)
19631963 return -ENOMEM;
1964− mlxsw_owq = alloc_ordered_workqueue("%s_ordered", WQ_MEM_RECLAIM,
1964+ mlxsw_owq = alloc_ordered_workqueue("%s_ordered", 0,
19651965 mlxsw_core_driver_name);
19661966 if (!mlxsw_owq) {
19671967 err = -ENOMEM;
drivers/net/ethernet/mellanox/mlxsw/spectrum_buffers.c+11−8
@@ -70,6 +70,7 @@ static const struct mlxsw_sp_sb_pool_des mlxsw_sp2_sb_pool_dess[] = { · 69 unmodified lines
7070 {MLXSW_REG_SBXX_DIR_EGRESS, 1},
7171 {MLXSW_REG_SBXX_DIR_EGRESS, 2},
7272 {MLXSW_REG_SBXX_DIR_EGRESS, 3},
73+ {MLXSW_REG_SBXX_DIR_EGRESS, 15},
7374 };
7475
7576 #define MLXSW_SP_SB_ING_TC_COUNT 8
@@ -428,6 +429,7 @@ static const struct mlxsw_sp_sb_pr mlxsw_sp2_sb_prs[] = { · 352 unmodified lines
428429 MLXSW_SP_SB_PR(MLXSW_REG_SBPR_MODE_STATIC, 0),
429430 MLXSW_SP_SB_PR(MLXSW_REG_SBPR_MODE_STATIC, 0),
430431 MLXSW_SP_SB_PR(MLXSW_REG_SBPR_MODE_STATIC, 0),
432+ MLXSW_SP_SB_PR(MLXSW_REG_SBPR_MODE_STATIC, MLXSW_SP_SB_INFI),
431433 };
432434
433435 static int mlxsw_sp_sb_prs_init(struct mlxsw_sp *mlxsw_sp,
@@ -517,14 +519,14 @@ static const struct mlxsw_sp_sb_cm mlxsw_sp2_sb_cms_egress[] = { · 83 unmodified lines
517519 MLXSW_SP_SB_CM(0, 7, 4),
518520 MLXSW_SP_SB_CM(0, 7, 4),
519521 MLXSW_SP_SB_CM(0, 7, 4),
520− MLXSW_SP_SB_CM(0, 7, 4),
521− MLXSW_SP_SB_CM(0, 7, 4),
522− MLXSW_SP_SB_CM(0, 7, 4),
523− MLXSW_SP_SB_CM(0, 7, 4),
524− MLXSW_SP_SB_CM(0, 7, 4),
525− MLXSW_SP_SB_CM(0, 7, 4),
526− MLXSW_SP_SB_CM(0, 7, 4),
527− MLXSW_SP_SB_CM(0, 7, 4),
522+ MLXSW_SP_SB_CM(0, MLXSW_SP_SB_INFI, 8),
523+ MLXSW_SP_SB_CM(0, MLXSW_SP_SB_INFI, 8),
524+ MLXSW_SP_SB_CM(0, MLXSW_SP_SB_INFI, 8),
525+ MLXSW_SP_SB_CM(0, MLXSW_SP_SB_INFI, 8),
526+ MLXSW_SP_SB_CM(0, MLXSW_SP_SB_INFI, 8),
527+ MLXSW_SP_SB_CM(0, MLXSW_SP_SB_INFI, 8),
528+ MLXSW_SP_SB_CM(0, MLXSW_SP_SB_INFI, 8),
529+ MLXSW_SP_SB_CM(0, MLXSW_SP_SB_INFI, 8),
528530 MLXSW_SP_SB_CM(1, 0xff, 4),
529531 };
530532
@@ -671,6 +673,7 @@ static const struct mlxsw_sp_sb_pm mlxsw_sp2_sb_pms[] = { · 140 unmodified lines
671673 MLXSW_SP_SB_PM(0, 0),
672674 MLXSW_SP_SB_PM(0, 0),
673675 MLXSW_SP_SB_PM(0, 0),
676+ MLXSW_SP_SB_PM(10000, 90000),
674677 };
675678
676679 static int mlxsw_sp_port_sb_pms_init(struct mlxsw_sp_port *mlxsw_sp_port)
drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c+1−1
@@ -6781,7 +6781,7 @@ static int mlxsw_sp_router_port_check_rif_addr(struct mlxsw_sp *mlxsw_sp, · 6780 unmodified lines
67816781 /* A RIF is not created for macvlan netdevs. Their MAC is used to
67826782 * populate the FDB
67836783 */
6784− if (netif_is_macvlan(dev))
6784+ if (netif_is_macvlan(dev) || netif_is_l3_master(dev))
67856785 return 0;
67866786
67876787 for (i = 0; i < MLXSW_CORE_RES_GET(mlxsw_sp->core, MAX_RIFS); i++) {
drivers/net/ethernet/mellanox/mlxsw/spectrum_switchdev.c+1−1
@@ -1630,7 +1630,7 @@ static int mlxsw_sp_port_mdb_add(struct mlxsw_sp_port *mlxsw_sp_port, · 1629 unmodified lines
16301630 u16 fid_index;
16311631 int err = 0;
16321632
1633− if (switchdev_trans_ph_prepare(trans))
1633+ if (switchdev_trans_ph_commit(trans))
16341634 return 0;
16351635
16361636 bridge_port = mlxsw_sp_bridge_port_find(mlxsw_sp->bridge, orig_dev);
drivers/net/ethernet/mscc/ocelot.c+15−9
@@ -613,7 +613,7 @@ static int ocelot_mact_mc_add(struct ocelot_port *port, · 612 unmodified lines
613613 struct netdev_hw_addr *hw_addr)
614614 {
615615 struct ocelot *ocelot = port->ocelot;
616− struct netdev_hw_addr *ha = kzalloc(sizeof(*ha), GFP_KERNEL);
616+ struct netdev_hw_addr *ha = kzalloc(sizeof(*ha), GFP_ATOMIC);
617617
618618 if (!ha)
619619 return -ENOMEM;
@@ -959,10 +959,8 @@ static void ocelot_get_strings(struct net_device *netdev, u32 sset, u8 *data) · 339 unmodified lines
959959 ETH_GSTRING_LEN);
960960 }
961961
962−static void ocelot_check_stats(struct work_struct *work)
962+static void ocelot_update_stats(struct ocelot *ocelot)
963963 {
964− struct delayed_work *del_work = to_delayed_work(work);
965− struct ocelot *ocelot = container_of(del_work, struct ocelot, stats_work);
966964 int i, j;
967965
968966 mutex_lock(&ocelot->stats_lock);
@@ -986,11 +984,19 @@ static void ocelot_check_stats(struct work_struct *work) · 17 unmodified lines
986984 }
987985 }
988986
989− cancel_delayed_work(&ocelot->stats_work);
987+ mutex_unlock(&ocelot->stats_lock);
988+}
989+
990+static void ocelot_check_stats_work(struct work_struct *work)
991+{
992+ struct delayed_work *del_work = to_delayed_work(work);
993+ struct ocelot *ocelot = container_of(del_work, struct ocelot,
994+ stats_work);
995+
996+ ocelot_update_stats(ocelot);
997+
990998 queue_delayed_work(ocelot->stats_queue, &ocelot->stats_work,
991999 OCELOT_STATS_CHECK_DELAY);
992−
993− mutex_unlock(&ocelot->stats_lock);
9941000 }
9951001
9961002 static void ocelot_get_ethtool_stats(struct net_device *dev,
@@ -1001,7 +1007,7 @@ static void ocelot_get_ethtool_stats(struct net_device *dev, · 4 unmodified lines
10011007 int i;
10021008
10031009 /* check and update now */
1004− ocelot_check_stats(&ocelot->stats_work.work);
1010+ ocelot_update_stats(ocelot);
10051011
10061012 /* Copy all counters */
10071013 for (i = 0; i < ocelot->num_stats; i++)
@@ -1809,7 +1815,7 @@ int ocelot_init(struct ocelot *ocelot) · 801 unmodified lines
18091815 ANA_CPUQ_8021_CFG_CPUQ_BPDU_VAL(6),
18101816 ANA_CPUQ_8021_CFG, i);
18111817
1812− INIT_DELAYED_WORK(&ocelot->stats_work, ocelot_check_stats);
1818+ INIT_DELAYED_WORK(&ocelot->stats_work, ocelot_check_stats_work);
18131819 queue_delayed_work(ocelot->stats_queue, &ocelot->stats_work,
18141820 OCELOT_STATS_CHECK_DELAY);
18151821 return 0;
drivers/net/ethernet/neterion/vxge/vxge-config.c+1−0
@@ -2366,6 +2366,7 @@ static void *__vxge_hw_blockpool_malloc(struct __vxge_hw_device *devh, u32 size, · 2365 unmodified lines
23662366 dma_object->addr))) {
23672367 vxge_os_dma_free(devh->pdev, memblock,
23682368 &dma_object->acc_handle);
2369+ memblock = NULL;
23692370 goto exit;
23702371 }
23712372
drivers/net/ethernet/qlogic/qed/qed.h+5−2
@@ -431,12 +431,16 @@ struct qed_qm_info { · 430 unmodified lines
431431 u8 num_pf_rls;
432432 };
433433
434+#define QED_OVERFLOW_BIT 1
435+
434436 struct qed_db_recovery_info {
435437 struct list_head list;
436438
437439 /* Lock to protect the doorbell recovery mechanism list */
438440 spinlock_t lock;
441+ bool dorq_attn;
439442 u32 db_recovery_counter;
443+ unsigned long overflow;
440444 };
441445
442446 struct storm_stats {
@@ -920,8 +924,7 @@ u16 qed_get_cm_pq_idx_llt_mtc(struct qed_hwfn *p_hwfn, u8 tc); · 477 unmodified lines
920924
921925 /* doorbell recovery mechanism */
922926 void qed_db_recovery_dp(struct qed_hwfn *p_hwfn);
923−void qed_db_recovery_execute(struct qed_hwfn *p_hwfn,
924− enum qed_db_rec_exec db_exec);
927+void qed_db_recovery_execute(struct qed_hwfn *p_hwfn);
925928 bool qed_edpm_enabled(struct qed_hwfn *p_hwfn);
926929
927930 /* Other Linux specific common definitions */
drivers/net/ethernet/qlogic/qed/qed_dev.c+34−51
@@ -102,11 +102,15 @@ static void qed_db_recovery_dp_entry(struct qed_hwfn *p_hwfn, · 101 unmodified lines
102102
103103 /* Doorbell address sanity (address within doorbell bar range) */
104104 static bool qed_db_rec_sanity(struct qed_dev *cdev,
105− void __iomem *db_addr, void *db_data)
105+ void __iomem *db_addr,
106+ enum qed_db_rec_width db_width,
107+ void *db_data)
106108 {
109+ u32 width = (db_width == DB_REC_WIDTH_32B) ? 32 : 64;
110+
107111 /* Make sure doorbell address is within the doorbell bar */
108112 if (db_addr < cdev->doorbells ||
109− (u8 __iomem *)db_addr >
113+ (u8 __iomem *)db_addr + width >
110114 (u8 __iomem *)cdev->doorbells + cdev->db_size) {
111115 WARN(true,
112116 "Illegal doorbell address: %p. Legal range for doorbell addresses is [%p..%p]\n",
@@ -159,7 +163,7 @@ int qed_db_recovery_add(struct qed_dev *cdev, · 46 unmodified lines
159163 }
160164
161165 /* Sanitize doorbell address */
162− if (!qed_db_rec_sanity(cdev, db_addr, db_data))
166+ if (!qed_db_rec_sanity(cdev, db_addr, db_width, db_data))
163167 return -EINVAL;
164168
165169 /* Obtain hwfn from doorbell address */
@@ -205,10 +209,6 @@ int qed_db_recovery_del(struct qed_dev *cdev, · 39 unmodified lines
205209 return 0;
206210 }
207211
208− /* Sanitize doorbell address */
209− if (!qed_db_rec_sanity(cdev, db_addr, db_data))
210− return -EINVAL;
211−
212212 /* Obtain hwfn from doorbell address */
213213 p_hwfn = qed_db_rec_find_hwfn(cdev, db_addr);
214214
@@ -300,31 +300,24 @@ void qed_db_recovery_dp(struct qed_hwfn *p_hwfn) · 85 unmodified lines
300300
301301 /* Ring the doorbell of a single doorbell recovery entry */
302302 static void qed_db_recovery_ring(struct qed_hwfn *p_hwfn,
303− struct qed_db_recovery_entry *db_entry,
304− enum qed_db_rec_exec db_exec)
305−{
306− if (db_exec != DB_REC_ONCE) {
307− /* Print according to width */
308− if (db_entry->db_width == DB_REC_WIDTH_32B) {
309− DP_VERBOSE(p_hwfn, QED_MSG_SPQ,
310− "%s doorbell address %p data %x\n",
311− db_exec == DB_REC_DRY_RUN ?
312− "would have rung" : "ringing",
313− db_entry->db_addr,
314− *(u32 *)db_entry->db_data);
315− } else {
316− DP_VERBOSE(p_hwfn, QED_MSG_SPQ,
317− "%s doorbell address %p data %llx\n",
318− db_exec == DB_REC_DRY_RUN ?
319− "would have rung" : "ringing",
320− db_entry->db_addr,
321− *(u64 *)(db_entry->db_data));
322− }
303+ struct qed_db_recovery_entry *db_entry)
304+{
305+ /* Print according to width */
306+ if (db_entry->db_width == DB_REC_WIDTH_32B) {
307+ DP_VERBOSE(p_hwfn, QED_MSG_SPQ,
308+ "ringing doorbell address %p data %x\n",
309+ db_entry->db_addr,
310+ *(u32 *)db_entry->db_data);
311+ } else {
312+ DP_VERBOSE(p_hwfn, QED_MSG_SPQ,
313+ "ringing doorbell address %p data %llx\n",
314+ db_entry->db_addr,
315+ *(u64 *)(db_entry->db_data));
323316 }
324317
325318 /* Sanity */
326319 if (!qed_db_rec_sanity(p_hwfn->cdev, db_entry->db_addr,
327− db_entry->db_data))
320+ db_entry->db_width, db_entry->db_data))
328321 return;
329322
330323 /* Flush the write combined buffer. Since there are multiple doorbelling
@@ -334,14 +327,12 @@ static void qed_db_recovery_ring(struct qed_hwfn *p_hwfn, · 3 unmodified lines
334327 wmb();
335328
336329 /* Ring the doorbell */
337− if (db_exec == DB_REC_REAL_DEAL || db_exec == DB_REC_ONCE) {
338− if (db_entry->db_width == DB_REC_WIDTH_32B)
339− DIRECT_REG_WR(db_entry->db_addr,
340− *(u32 *)(db_entry->db_data));
341− else
342− DIRECT_REG_WR64(db_entry->db_addr,
343− *(u64 *)(db_entry->db_data));
344− }
330+ if (db_entry->db_width == DB_REC_WIDTH_32B)
331+ DIRECT_REG_WR(db_entry->db_addr,
332+ *(u32 *)(db_entry->db_data));
333+ else
334+ DIRECT_REG_WR64(db_entry->db_addr,
335+ *(u64 *)(db_entry->db_data));
345336
346337 /* Flush the write combined buffer. Next doorbell may come from a
347338 * different entity to the same address...
@@ -350,29 +341,21 @@ static void qed_db_recovery_ring(struct qed_hwfn *p_hwfn, · 2 unmodified lines
350341 }
351342
352343 /* Traverse the doorbell recovery entry list and ring all the doorbells */
353−void qed_db_recovery_execute(struct qed_hwfn *p_hwfn,
354− enum qed_db_rec_exec db_exec)
344+void qed_db_recovery_execute(struct qed_hwfn *p_hwfn)
355345 {
356346 struct qed_db_recovery_entry *db_entry = NULL;
357347
358− if (db_exec != DB_REC_ONCE) {
359− DP_NOTICE(p_hwfn,
360− "Executing doorbell recovery. Counter was %d\n",
361− p_hwfn->db_recovery_info.db_recovery_counter);
348+ DP_NOTICE(p_hwfn, "Executing doorbell recovery. Counter was %d\n",
349+ p_hwfn->db_recovery_info.db_recovery_counter);
362350
363− /* Track amount of times recovery was executed */
364− p_hwfn->db_recovery_info.db_recovery_counter++;
365− }
351+ /* Track amount of times recovery was executed */
352+ p_hwfn->db_recovery_info.db_recovery_counter++;
366353
367354 /* Protect the list */
368355 spin_lock_bh(&p_hwfn->db_recovery_info.lock);
369356 list_for_each_entry(db_entry,
370− &p_hwfn->db_recovery_info.list, list_entry) {
371− qed_db_recovery_ring(p_hwfn, db_entry, db_exec);
372− if (db_exec == DB_REC_ONCE)
373− break;
374− }
375−
357+ &p_hwfn->db_recovery_info.list, list_entry)
358+ qed_db_recovery_ring(p_hwfn, db_entry);
376359 spin_unlock_bh(&p_hwfn->db_recovery_info.lock);
377360 }
378361
drivers/net/ethernet/qlogic/qed/qed_int.c+63−20
@@ -378,6 +378,9 @@ static int qed_db_rec_flush_queue(struct qed_hwfn *p_hwfn, · 377 unmodified lines
378378 u32 count = QED_DB_REC_COUNT;
379379 u32 usage = 1;
380380
381+ /* Flush any pending (e)dpms as they may never arrive */
382+ qed_wr(p_hwfn, p_ptt, DORQ_REG_DPM_FORCE_ABORT, 0x1);
383+
381384 /* wait for usage to zero or count to run out. This is necessary since
382385 * EDPM doorbell transactions can take multiple 64b cycles, and as such
383386 * can "split" over the pci. Possibly, the doorbell drop can happen with
@@ -406,51 +409,74 @@ static int qed_db_rec_flush_queue(struct qed_hwfn *p_hwfn, · 22 unmodified lines
406409
407410 int qed_db_rec_handler(struct qed_hwfn *p_hwfn, struct qed_ptt *p_ptt)
408411 {
409− u32 overflow;
412+ u32 attn_ovfl, cur_ovfl;
410413 int rc;
411414
412− overflow = qed_rd(p_hwfn, p_ptt, DORQ_REG_PF_OVFL_STICKY);
413− DP_NOTICE(p_hwfn, "PF Overflow sticky 0x%x\n", overflow);
414− if (!overflow) {
415− qed_db_recovery_execute(p_hwfn, DB_REC_ONCE);
415+ attn_ovfl = test_and_clear_bit(QED_OVERFLOW_BIT,
416+ &p_hwfn->db_recovery_info.overflow);
417+ cur_ovfl = qed_rd(p_hwfn, p_ptt, DORQ_REG_PF_OVFL_STICKY);
418+ if (!cur_ovfl && !attn_ovfl)
416419 return 0;
417− }
418420
419− if (qed_edpm_enabled(p_hwfn)) {
421+ DP_NOTICE(p_hwfn, "PF Overflow sticky: attn %u current %u\n",
422+ attn_ovfl, cur_ovfl);
423+
424+ if (cur_ovfl && !p_hwfn->db_bar_no_edpm) {
420425 rc = qed_db_rec_flush_queue(p_hwfn, p_ptt);
421426 if (rc)
422427 return rc;
423428 }
424429
425− /* Flush any pending (e)dpm as they may never arrive */
426− qed_wr(p_hwfn, p_ptt, DORQ_REG_DPM_FORCE_ABORT, 0x1);
427−
428430 /* Release overflow sticky indication (stop silently dropping everything) */
429431 qed_wr(p_hwfn, p_ptt, DORQ_REG_PF_OVFL_STICKY, 0x0);
430432
431433 /* Repeat all last doorbells (doorbell drop recovery) */
432− qed_db_recovery_execute(p_hwfn, DB_REC_REAL_DEAL);
434+ qed_db_recovery_execute(p_hwfn);
433435
434436 return 0;
435437 }
436438
437−static int qed_dorq_attn_cb(struct qed_hwfn *p_hwfn)
439+static void qed_dorq_attn_overflow(struct qed_hwfn *p_hwfn)
438440 {
439− u32 int_sts, first_drop_reason, details, address, all_drops_reason;
440441 struct qed_ptt *p_ptt = p_hwfn->p_dpc_ptt;
442+ u32 overflow;
441443 int rc;
442444
443− int_sts = qed_rd(p_hwfn, p_ptt, DORQ_REG_INT_STS);
444− DP_NOTICE(p_hwfn->cdev, "DORQ attention. int_sts was %x\n", int_sts);
445+ overflow = qed_rd(p_hwfn, p_ptt, DORQ_REG_PF_OVFL_STICKY);
446+ if (!overflow)
447+ goto out;
448+
449+ /* Run PF doorbell recovery in next periodic handler */
450+ set_bit(QED_OVERFLOW_BIT, &p_hwfn->db_recovery_info.overflow);
451+
452+ if (!p_hwfn->db_bar_no_edpm) {
453+ rc = qed_db_rec_flush_queue(p_hwfn, p_ptt);
454+ if (rc)
455+ goto out;
456+ }
457+
458+ qed_wr(p_hwfn, p_ptt, DORQ_REG_PF_OVFL_STICKY, 0x0);
459+out:
460+ /* Schedule the handler even if overflow was not detected */
461+ qed_periodic_db_rec_start(p_hwfn);
462+}
463+
464+static int qed_dorq_attn_int_sts(struct qed_hwfn *p_hwfn)
465+{
466+ u32 int_sts, first_drop_reason, details, address, all_drops_reason;
467+ struct qed_ptt *p_ptt = p_hwfn->p_dpc_ptt;
445468
446469 /* int_sts may be zero since all PFs were interrupted for doorbell
447470 * overflow but another one already handled it. Can abort here. If
448471 * This PF also requires overflow recovery we will be interrupted again.
449472 * The masked almost full indication may also be set. Ignoring.
450473 */
474+ int_sts = qed_rd(p_hwfn, p_ptt, DORQ_REG_INT_STS);
451475 if (!(int_sts & ~DORQ_REG_INT_STS_DORQ_FIFO_AFULL))
452476 return 0;
453477
478+ DP_NOTICE(p_hwfn->cdev, "DORQ attention. int_sts was %x\n", int_sts);
479+
454480 /* check if db_drop or overflow happened */
455481 if (int_sts & (DORQ_REG_INT_STS_DB_DROP |
456482 DORQ_REG_INT_STS_DORQ_FIFO_OVFL_ERR)) {
@@ -477,11 +503,6 @@ static int qed_dorq_attn_cb(struct qed_hwfn *p_hwfn) · 20 unmodified lines
477503 GET_FIELD(details, QED_DORQ_ATTENTION_SIZE) * 4,
478504 first_drop_reason, all_drops_reason);
479505
480− rc = qed_db_rec_handler(p_hwfn, p_ptt);
481− qed_periodic_db_rec_start(p_hwfn);
482− if (rc)
483− return rc;
484−
485506 /* Clear the doorbell drop details and prepare for next drop */
486507 qed_wr(p_hwfn, p_ptt, DORQ_REG_DB_DROP_DETAILS_REL, 0);
487508
@@ -507,6 +528,25 @@ static int qed_dorq_attn_cb(struct qed_hwfn *p_hwfn) · 19 unmodified lines
507528 return -EINVAL;
508529 }
509530
531+static int qed_dorq_attn_cb(struct qed_hwfn *p_hwfn)
532+{
533+ p_hwfn->db_recovery_info.dorq_attn = true;
534+ qed_dorq_attn_overflow(p_hwfn);
535+
536+ return qed_dorq_attn_int_sts(p_hwfn);
537+}
538+
539+static void qed_dorq_attn_handler(struct qed_hwfn *p_hwfn)
540+{
541+ if (p_hwfn->db_recovery_info.dorq_attn)
542+ goto out;
543+
544+ /* Call DORQ callback if the attention was missed */
545+ qed_dorq_attn_cb(p_hwfn);
546+out:
547+ p_hwfn->db_recovery_info.dorq_attn = false;
548+}
549+
510550 /* Instead of major changes to the data-structure, we have a some 'special'
511551 * identifiers for sources that changed meaning between adapters.
512552 */
@@ -1080,6 +1120,9 @@ static int qed_int_deassertion(struct qed_hwfn *p_hwfn, · 567 unmodified lines
10801120 }
10811121 }
10821122
1123+ /* Handle missed DORQ attention */
1124+ qed_dorq_attn_handler(p_hwfn);
1125+
10831126 /* Clear IGU indication for the deasserted bits */
10841127 DIRECT_REG_WR((u8 __iomem *)p_hwfn->regview +
10851128 GTT_BAR0_MAP_REG_IGU_CMD +
drivers/net/ethernet/qlogic/qed/qed_int.h+2−2
@@ -192,8 +192,8 @@ void qed_int_disable_post_isr_release(struct qed_dev *cdev); · 191 unmodified lines
192192
193193 /**
194194 * @brief - Doorbell Recovery handler.
195− * Run DB_REAL_DEAL doorbell recovery in case of PF overflow
196− * (and flush DORQ if needed), otherwise run DB_REC_ONCE.
195+ * Run doorbell recovery in case of PF overflow (and flush DORQ if
196+ * needed).
197197 *
198198 * @param p_hwfn
199199 * @param p_ptt
drivers/net/ethernet/qlogic/qed/qed_main.c+1−1
@@ -970,7 +970,7 @@ static void qed_update_pf_params(struct qed_dev *cdev, · 969 unmodified lines
970970 }
971971 }
972972
973−#define QED_PERIODIC_DB_REC_COUNT 100
973+#define QED_PERIODIC_DB_REC_COUNT 10
974974 #define QED_PERIODIC_DB_REC_INTERVAL_MS 100
975975 #define QED_PERIODIC_DB_REC_INTERVAL \
976976 msecs_to_jiffies(QED_PERIODIC_DB_REC_INTERVAL_MS)
drivers/net/ethernet/qlogic/qed/qed_sriov.c+1−1
@@ -1591,7 +1591,7 @@ static void qed_iov_vf_mbx_acquire(struct qed_hwfn *p_hwfn, · 1590 unmodified lines
15911591 p_vfdev->eth_fp_hsi_minor = ETH_HSI_VER_NO_PKT_LEN_TUNN;
15921592 } else {
15931593 DP_INFO(p_hwfn,
1594− "VF[%d] needs fastpath HSI %02x.%02x, which is incompatible with loaded FW's faspath HSI %02x.%02x\n",
1594+ "VF[%d] needs fastpath HSI %02x.%02x, which is incompatible with loaded FW's fastpath HSI %02x.%02x\n",
15951595 vf->abs_vf_id,
15961596 req->vfdev_info.eth_fp_hsi_major,
15971597 req->vfdev_info.eth_fp_hsi_minor,
drivers/net/ethernet/qlogic/qede/qede_ptp.c+3−4
@@ -490,18 +490,17 @@ int qede_ptp_enable(struct qede_dev *edev, bool init_tc) · 489 unmodified lines
490490
491491 ptp->clock = ptp_clock_register(&ptp->clock_info, &edev->pdev->dev);
492492 if (IS_ERR(ptp->clock)) {
493− rc = -EINVAL;
494493 DP_ERR(edev, "PTP clock registration failed\n");
494+ qede_ptp_disable(edev);
495+ rc = -EINVAL;
495496 goto err2;
496497 }
497498
498499 return 0;
499500
500−err2:
501− qede_ptp_disable(edev);
502− ptp->clock = NULL;
503501 err1:
504502 kfree(ptp);
503+err2:
505504 edev->ptp = NULL;
506505
507506 return rc;
drivers/net/team/team.c+26−0
@@ -1246,6 +1246,23 @@ static int team_port_add(struct team *team, struct net_device *port_dev, · 1245 unmodified lines
12461246 goto err_option_port_add;
12471247 }
12481248
1249+ /* set promiscuity level to new slave */
1250+ if (dev->flags & IFF_PROMISC) {
1251+ err = dev_set_promiscuity(port_dev, 1);
1252+ if (err)
1253+ goto err_set_slave_promisc;
1254+ }
1255+
1256+ /* set allmulti level to new slave */
1257+ if (dev->flags & IFF_ALLMULTI) {
1258+ err = dev_set_allmulti(port_dev, 1);
1259+ if (err) {
1260+ if (dev->flags & IFF_PROMISC)
1261+ dev_set_promiscuity(port_dev, -1);
1262+ goto err_set_slave_promisc;
1263+ }
1264+ }
1265+
12491266 netif_addr_lock_bh(dev);
12501267 dev_uc_sync_multiple(port_dev, dev);
12511268 dev_mc_sync_multiple(port_dev, dev);
@@ -1262,6 +1279,9 @@ static int team_port_add(struct team *team, struct net_device *port_dev, · 10 unmodified lines
12621279
12631280 return 0;
12641281
1282+err_set_slave_promisc:
1283+ __team_option_inst_del_port(team, port);
1284+
12651285 err_option_port_add:
12661286 team_upper_dev_unlink(team, port);
12671287
@@ -1307,6 +1327,12 @@ static int team_port_del(struct team *team, struct net_device *port_dev) · 39 unmodified lines
13071327
13081328 team_port_disable(team, port);
13091329 list_del_rcu(&port->list);
1330+
1331+ if (dev->flags & IFF_PROMISC)
1332+ dev_set_promiscuity(port_dev, -1);
1333+ if (dev->flags & IFF_ALLMULTI)
1334+ dev_set_allmulti(port_dev, -1);
1335+
13101336 team_upper_dev_unlink(team, port);
13111337 netdev_rx_handler_unregister(port_dev);
13121338 team_port_disable_netpoll(port);
drivers/net/wireless/ath/ath10k/htt_rx.c+1−1
@@ -2728,7 +2728,7 @@ static void ath10k_htt_rx_tx_fetch_ind(struct ath10k *ar, struct sk_buff *skb) · 2727 unmodified lines
27282728 num_msdus++;
27292729 num_bytes += ret;
27302730 }
2731− ieee80211_return_txq(hw, txq);
2731+ ieee80211_return_txq(hw, txq, false);
27322732 ieee80211_txq_schedule_end(hw, txq->ac);
27332733
27342734 record->num_msdus = cpu_to_le16(num_msdus);
drivers/net/wireless/ath/ath10k/mac.c+2−2
@@ -4089,7 +4089,7 @@ static int ath10k_mac_schedule_txq(struct ieee80211_hw *hw, u32 ac) · 4088 unmodified lines
40894089 if (ret < 0)
40904090 break;
40914091 }
4092− ieee80211_return_txq(hw, txq);
4092+ ieee80211_return_txq(hw, txq, false);
40934093 ath10k_htt_tx_txq_update(hw, txq);
40944094 if (ret == -EBUSY)
40954095 break;
@@ -4374,7 +4374,7 @@ static void ath10k_mac_op_wake_tx_queue(struct ieee80211_hw *hw, · 278 unmodified lines
43744374 if (ret < 0)
43754375 break;
43764376 }
4377− ieee80211_return_txq(hw, txq);
4377+ ieee80211_return_txq(hw, txq, false);
43784378 ath10k_htt_tx_txq_update(hw, txq);
43794379 out:
43804380 ieee80211_txq_schedule_end(hw, ac);
drivers/net/wireless/ath/ath9k/xmit.c+4−1
@@ -1938,12 +1938,15 @@ void ath_txq_schedule(struct ath_softc *sc, struct ath_txq *txq) · 1937 unmodified lines
19381938 goto out;
19391939
19401940 while ((queue = ieee80211_next_txq(hw, txq->mac80211_qnum))) {
1941+ bool force;
1942+
19411943 tid = (struct ath_atx_tid *)queue->drv_priv;
19421944
19431945 ret = ath_tx_sched_aggr(sc, txq, tid);
19441946 ath_dbg(common, QUEUE, "ath_tx_sched_aggr returned %d\n", ret);
19451947
1946− ieee80211_return_txq(hw, queue);
1948+ force = !skb_queue_empty(&tid->retry_q);
1949+ ieee80211_return_txq(hw, queue, force);
19471950 }
19481951
19491952 out:
drivers/net/wireless/intel/iwlwifi/cfg/22000.c+22−8
@@ -82,6 +82,7 @@ · 81 unmodified lines
8282 #define IWL_22000_HR_A0_FW_PRE "iwlwifi-QuQnj-a0-hr-a0-"
8383 #define IWL_22000_SU_Z0_FW_PRE "iwlwifi-su-z0-"
8484 #define IWL_QU_B_JF_B_FW_PRE "iwlwifi-Qu-b0-jf-b0-"
85+#define IWL_QUZ_A_HR_B_FW_PRE "iwlwifi-QuZ-a0-hr-b0-"
8586 #define IWL_QNJ_B_JF_B_FW_PRE "iwlwifi-QuQnj-b0-jf-b0-"
8687 #define IWL_CC_A_FW_PRE "iwlwifi-cc-a0-"
8788 #define IWL_22000_SO_A_JF_B_FW_PRE "iwlwifi-so-a0-jf-b0-"
@@ -105,8 +106,8 @@ · 17 unmodified lines
105106 IWL_22000_HR_A0_FW_PRE __stringify(api) ".ucode"
106107 #define IWL_22000_SU_Z0_MODULE_FIRMWARE(api) \
107108 IWL_22000_SU_Z0_FW_PRE __stringify(api) ".ucode"
108−#define IWL_QU_B_JF_B_MODULE_FIRMWARE(api) \
109− IWL_QU_B_JF_B_FW_PRE __stringify(api) ".ucode"
109+#define IWL_QUZ_A_HR_B_MODULE_FIRMWARE(api) \
110+ IWL_QUZ_A_HR_B_FW_PRE __stringify(api) ".ucode"
110111 #define IWL_QU_B_JF_B_MODULE_FIRMWARE(api) \
111112 IWL_QU_B_JF_B_FW_PRE __stringify(api) ".ucode"
112113 #define IWL_QNJ_B_JF_B_MODULE_FIRMWARE(api) \
@@ -235,8 +236,20 @@ const struct iwl_cfg iwl_ax101_cfg_qu_hr = { · 122 unmodified lines
235236 .max_tx_agg_size = IEEE80211_MAX_AMPDU_BUF_HT,
236237 };
237238
238−const struct iwl_cfg iwl22260_2ax_cfg = {
239− .name = "Intel(R) Wireless-AX 22260",
239+const struct iwl_cfg iwl_ax101_cfg_quz_hr = {
240+ .name = "Intel(R) Wi-Fi 6 AX101",
241+ .fw_name_pre = IWL_QUZ_A_HR_B_FW_PRE,
242+ IWL_DEVICE_22500,
243+ /*
244+ * This device doesn't support receiving BlockAck with a large bitmap
245+ * so we need to restrict the size of transmitted aggregation to the
246+ * HT size; mac80211 would otherwise pick the HE max (256) by default.
247+ */
248+ .max_tx_agg_size = IEEE80211_MAX_AMPDU_BUF_HT,
249+};
250+
251+const struct iwl_cfg iwl_ax200_cfg_cc = {
252+ .name = "Intel(R) Wi-Fi 6 AX200 160MHz",
240253 .fw_name_pre = IWL_CC_A_FW_PRE,
241254 IWL_DEVICE_22500,
242255 /*
@@ -249,7 +262,7 @@ const struct iwl_cfg iwl22260_2ax_cfg = { · 6 unmodified lines
249262 };
250263
251264 const struct iwl_cfg killer1650x_2ax_cfg = {
252− .name = "Killer(R) Wireless-AX 1650x Wireless Network Adapter (200NGW)",
265+ .name = "Killer(R) Wi-Fi 6 AX1650x 160MHz Wireless Network Adapter (200NGW)",
253266 .fw_name_pre = IWL_CC_A_FW_PRE,
254267 IWL_DEVICE_22500,
255268 /*
@@ -262,7 +275,7 @@ const struct iwl_cfg killer1650x_2ax_cfg = { · 6 unmodified lines
262275 };
263276
264277 const struct iwl_cfg killer1650w_2ax_cfg = {
265− .name = "Killer(R) Wireless-AX 1650w Wireless Network Adapter (200D2W)",
278+ .name = "Killer(R) Wi-Fi 6 AX1650w 160MHz Wireless Network Adapter (200D2W)",
266279 .fw_name_pre = IWL_CC_A_FW_PRE,
267280 IWL_DEVICE_22500,
268281 /*
@@ -328,7 +341,7 @@ const struct iwl_cfg killer1550s_2ac_cfg_qu_b0_jf_b0 = { · 59 unmodified lines
328341 };
329342
330343 const struct iwl_cfg killer1650s_2ax_cfg_qu_b0_hr_b0 = {
331− .name = "Killer(R) Wireless-AX 1650i Wireless Network Adapter (22560NGW)",
344+ .name = "Killer(R) Wi-Fi 6 AX1650i 160MHz Wireless Network Adapter (201NGW)",
332345 .fw_name_pre = IWL_22000_QU_B_HR_B_FW_PRE,
333346 IWL_DEVICE_22500,
334347 /*
@@ -340,7 +353,7 @@ const struct iwl_cfg killer1650s_2ax_cfg_qu_b0_hr_b0 = { · 5 unmodified lines
340353 };
341354
342355 const struct iwl_cfg killer1650i_2ax_cfg_qu_b0_hr_b0 = {
343− .name = "Killer(R) Wireless-AX 1650s Wireless Network Adapter (22560D2W)",
356+ .name = "Killer(R) Wi-Fi 6 AX1650s 160MHz Wireless Network Adapter (201D2W)",
344357 .fw_name_pre = IWL_22000_QU_B_HR_B_FW_PRE,
345358 IWL_DEVICE_22500,
346359 /*
@@ -444,6 +457,7 @@ MODULE_FIRMWARE(IWL_22000_HR_B_QNJ_MODULE_FIRMWARE(IWL_22000_UCODE_API_MAX)); · 97 unmodified lines
444457 MODULE_FIRMWARE(IWL_22000_HR_A0_QNJ_MODULE_FIRMWARE(IWL_22000_UCODE_API_MAX));
445458 MODULE_FIRMWARE(IWL_22000_SU_Z0_MODULE_FIRMWARE(IWL_22000_UCODE_API_MAX));
446459 MODULE_FIRMWARE(IWL_QU_B_JF_B_MODULE_FIRMWARE(IWL_22000_UCODE_API_MAX));
460+MODULE_FIRMWARE(IWL_QUZ_A_HR_B_MODULE_FIRMWARE(IWL_22000_UCODE_API_MAX));
447461 MODULE_FIRMWARE(IWL_QNJ_B_JF_B_MODULE_FIRMWARE(IWL_22000_UCODE_API_MAX));
448462 MODULE_FIRMWARE(IWL_CC_A_MODULE_FIRMWARE(IWL_22000_UCODE_API_MAX));
449463 MODULE_FIRMWARE(IWL_22000_SO_A_JF_B_MODULE_FIRMWARE(IWL_22000_UCODE_API_MAX));
drivers/net/wireless/intel/iwlwifi/fw/dbg.c+8−26
@@ -1614,6 +1614,7 @@ iwl_dump_ini_mem(struct iwl_fw_runtime *fwrt, · 1613 unmodified lines
16141614 if (!range) {
16151615 IWL_ERR(fwrt, "Failed to fill region header: id=%d, type=%d\n",
16161616 le32_to_cpu(reg->region_id), type);
1617+ memset(*data, 0, le32_to_cpu((*data)->len));
16171618 return;
16181619 }
16191620
@@ -1623,6 +1624,7 @@ iwl_dump_ini_mem(struct iwl_fw_runtime *fwrt, · 3 unmodified lines
16231624 if (range_size < 0) {
16241625 IWL_ERR(fwrt, "Failed to dump region: id=%d, type=%d\n",
16251626 le32_to_cpu(reg->region_id), type);
1627+ memset(*data, 0, le32_to_cpu((*data)->len));
16261628 return;
16271629 }
16281630 range = range + range_size;
@@ -1807,12 +1809,12 @@ _iwl_fw_error_ini_dump(struct iwl_fw_runtime *fwrt, · 178 unmodified lines
18071809
18081810 trigger = fwrt->dump.active_trigs[id].trig;
18091811
1810− size = sizeof(*dump_file);
1811− size += iwl_fw_ini_get_trigger_len(fwrt, trigger);
1812−
1812+ size = iwl_fw_ini_get_trigger_len(fwrt, trigger);
18131813 if (!size)
18141814 return NULL;
18151815
1816+ size += sizeof(*dump_file);
1817+
18161818 dump_file = vzalloc(size);
18171819 if (!dump_file)
18181820 return NULL;
@@ -1942,14 +1944,10 @@ int iwl_fw_dbg_error_collect(struct iwl_fw_runtime *fwrt, · 123 unmodified lines
19421944 iwl_dump_error_desc->len = 0;
19431945
19441946 ret = iwl_fw_dbg_collect_desc(fwrt, iwl_dump_error_desc, false, 0);
1945− if (ret) {
1947+ if (ret)
19461948 kfree(iwl_dump_error_desc);
1947− } else {
1948− set_bit(STATUS_FW_WAIT_DUMP, &fwrt->trans->status);
1949−
1950− /* trigger nmi to halt the fw */
1951− iwl_force_nmi(fwrt->trans);
1952− }
1949+ else
1950+ iwl_trans_sync_nmi(fwrt->trans);
19531951
19541952 return ret;
19551953 }
@@ -2489,22 +2487,6 @@ IWL_EXPORT_SYMBOL(iwl_fw_dbg_apply_point); · 533 unmodified lines
24892487
24902488 void iwl_fwrt_stop_device(struct iwl_fw_runtime *fwrt)
24912489 {
2492− /* if the wait event timeout elapses instead of wake up then
2493− * the driver did not receive NMI interrupt and can not assume the FW
2494− * is halted
2495− */
2496− int ret = wait_event_timeout(fwrt->trans->fw_halt_waitq,
2497− !test_bit(STATUS_FW_WAIT_DUMP,
2498− &fwrt->trans->status),
2499− msecs_to_jiffies(2000));
2500− if (!ret) {
2501− /* failed to receive NMI interrupt, assuming the FW is stuck */
2502− set_bit(STATUS_FW_ERROR, &fwrt->trans->status);
2503−
2504− clear_bit(STATUS_FW_WAIT_DUMP, &fwrt->trans->status);
2505− }
2506−
2507− /* Assuming the op mode mutex is held at this point */
25082490 iwl_fw_dbg_collect_sync(fwrt);
25092491
25102492 iwl_trans_stop_device(fwrt->trans);
drivers/net/wireless/intel/iwlwifi/fw/init.c+0−1
@@ -76,7 +76,6 @@ void iwl_fw_runtime_init(struct iwl_fw_runtime *fwrt, struct iwl_trans *trans, · 75 unmodified lines
7676 fwrt->ops_ctx = ops_ctx;
7777 INIT_DELAYED_WORK(&fwrt->dump.wk, iwl_fw_error_dump_wk);
7878 iwl_fwrt_dbgfs_register(fwrt, dbgfs_dir);
79− init_waitqueue_head(&fwrt->trans->fw_halt_waitq);
8079 }
8180 IWL_EXPORT_SYMBOL(iwl_fw_runtime_init);
8281
drivers/net/wireless/intel/iwlwifi/iwl-config.h+2−1
@@ -549,8 +549,9 @@ extern const struct iwl_cfg iwl22000_2ac_cfg_hr; · 548 unmodified lines
549549 extern const struct iwl_cfg iwl22000_2ac_cfg_hr_cdb;
550550 extern const struct iwl_cfg iwl22000_2ac_cfg_jf;
551551 extern const struct iwl_cfg iwl_ax101_cfg_qu_hr;
552+extern const struct iwl_cfg iwl_ax101_cfg_quz_hr;
552553 extern const struct iwl_cfg iwl22000_2ax_cfg_hr;
553−extern const struct iwl_cfg iwl22260_2ax_cfg;
554+extern const struct iwl_cfg iwl_ax200_cfg_cc;
554555 extern const struct iwl_cfg killer1650s_2ax_cfg_qu_b0_hr_b0;
555556 extern const struct iwl_cfg killer1650i_2ax_cfg_qu_b0_hr_b0;
556557 extern const struct iwl_cfg killer1650x_2ax_cfg;
drivers/net/wireless/intel/iwlwifi/iwl-csr.h+1−0
@@ -327,6 +327,7 @@ enum { · 326 unmodified lines
327327 #define CSR_HW_REV_TYPE_NONE (0x00001F0)
328328 #define CSR_HW_REV_TYPE_QNJ (0x0000360)
329329 #define CSR_HW_REV_TYPE_QNJ_B0 (0x0000364)
330+#define CSR_HW_REV_TYPE_QUZ (0x0000354)
330331 #define CSR_HW_REV_TYPE_HR_CDB (0x0000340)
331332 #define CSR_HW_REV_TYPE_SO (0x0000370)
332333 #define CSR_HW_REV_TYPE_TY (0x0000420)
drivers/net/wireless/intel/iwlwifi/iwl-trans.h+6−6
@@ -338,7 +338,6 @@ enum iwl_d3_status { · 337 unmodified lines
338338 * are sent
339339 * @STATUS_TRANS_IDLE: the trans is idle - general commands are not to be sent
340340 * @STATUS_TRANS_DEAD: trans is dead - avoid any read/write operation
341− * @STATUS_FW_WAIT_DUMP: if set, wait until cleared before collecting dump
342341 */
343342 enum iwl_trans_status {
344343 STATUS_SYNC_HCMD_ACTIVE,
@@ -351,7 +350,6 @@ enum iwl_trans_status { · 6 unmodified lines
351350 STATUS_TRANS_GOING_IDLE,
352351 STATUS_TRANS_IDLE,
353352 STATUS_TRANS_DEAD,
354− STATUS_FW_WAIT_DUMP,
355353 };
356354
357355 static inline int
@@ -618,6 +616,7 @@ struct iwl_trans_ops { · 260 unmodified lines
618616 struct iwl_trans_dump_data *(*dump_data)(struct iwl_trans *trans,
619617 u32 dump_mask);
620618 void (*debugfs_cleanup)(struct iwl_trans *trans);
619+ void (*sync_nmi)(struct iwl_trans *trans);
621620 };
622621
623622 /**
@@ -831,7 +830,6 @@ struct iwl_trans { · 207 unmodified lines
831830 u32 lmac_error_event_table[2];
832831 u32 umac_error_event_table;
833832 unsigned int error_event_table_tlv_status;
834− wait_queue_head_t fw_halt_waitq;
835833
836834 /* pointer to trans specific struct */
837835 /*Ensure that this pointer will always be aligned to sizeof pointer */
@@ -1239,10 +1237,12 @@ static inline void iwl_trans_fw_error(struct iwl_trans *trans) · 401 unmodified lines
12391237 /* prevent double restarts due to the same erroneous FW */
12401238 if (!test_and_set_bit(STATUS_FW_ERROR, &trans->status))
12411239 iwl_op_mode_nic_error(trans->op_mode);
1240+}
12421241
1243− if (test_and_clear_bit(STATUS_FW_WAIT_DUMP, &trans->status))
1244− wake_up(&trans->fw_halt_waitq);
1245−
1242+static inline void iwl_trans_sync_nmi(struct iwl_trans *trans)
1243+{
1244+ if (trans->ops->sync_nmi)
1245+ trans->ops->sync_nmi(trans);
12461246 }
12471247
12481248 /*****************************************************
drivers/net/wireless/intel/iwlwifi/mvm/mac80211.c+22−49
@@ -2714,9 +2714,6 @@ static void iwl_mvm_stop_ap_ibss(struct ieee80211_hw *hw, · 2713 unmodified lines
27142714
27152715 iwl_mvm_mac_ctxt_remove(mvm, vif);
27162716
2717− kfree(mvmvif->ap_wep_key);
2718− mvmvif->ap_wep_key = NULL;
2719−
27202717 mutex_unlock(&mvm->mutex);
27212718 }
27222719
@@ -3183,24 +3180,7 @@ static int iwl_mvm_mac_sta_state(struct ieee80211_hw *hw, · 460 unmodified lines
31833180 ret = iwl_mvm_update_sta(mvm, vif, sta);
31843181 } else if (old_state == IEEE80211_STA_ASSOC &&
31853182 new_state == IEEE80211_STA_AUTHORIZED) {
3186− /* if wep is used, need to set the key for the station now */
3187− if (vif->type == NL80211_IFTYPE_AP && mvmvif->ap_wep_key) {
3188− mvm_sta->wep_key =
3189− kmemdup(mvmvif->ap_wep_key,
3190− sizeof(*mvmvif->ap_wep_key) +
3191− mvmvif->ap_wep_key->keylen,
3192− GFP_KERNEL);
3193− if (!mvm_sta->wep_key) {
3194− ret = -ENOMEM;
3195− goto out_unlock;
3196− }
3197−
3198− ret = iwl_mvm_set_sta_key(mvm, vif, sta,
3199− mvm_sta->wep_key,
3200− STA_KEY_IDX_INVALID);
3201− } else {
3202− ret = 0;
3203− }
3183+ ret = 0;
32043184
32053185 /* we don't support TDLS during DCM */
32063186 if (iwl_mvm_phy_ctx_count(mvm) > 1)
@@ -3242,17 +3222,6 @@ static int iwl_mvm_mac_sta_state(struct ieee80211_hw *hw, · 35 unmodified lines
32423222 NL80211_TDLS_DISABLE_LINK);
32433223 }
32443224
3245− /* Remove STA key if this is an AP using WEP */
3246− if (vif->type == NL80211_IFTYPE_AP && mvmvif->ap_wep_key) {
3247− int rm_ret = iwl_mvm_remove_sta_key(mvm, vif, sta,
3248− mvm_sta->wep_key);
3249−
3250− if (!ret)
3251− ret = rm_ret;
3252− kfree(mvm_sta->wep_key);
3253− mvm_sta->wep_key = NULL;
3254− }
3255−
32563225 if (unlikely(ret &&
32573226 test_bit(IWL_MVM_STATUS_HW_RESTART_REQUESTED,
32583227 &mvm->status)))
@@ -3289,6 +3258,13 @@ static void iwl_mvm_sta_rc_update(struct ieee80211_hw *hw, · 30 unmodified lines
32893258 struct ieee80211_sta *sta, u32 changed)
32903259 {
32913260 struct iwl_mvm *mvm = IWL_MAC80211_GET_MVM(hw);
3261+ struct iwl_mvm_vif *mvmvif = iwl_mvm_vif_from_mac80211(vif);
3262+
3263+ if (changed & (IEEE80211_RC_BW_CHANGED |
3264+ IEEE80211_RC_SUPP_RATES_CHANGED |
3265+ IEEE80211_RC_NSS_CHANGED))
3266+ iwl_mvm_rs_rate_init(mvm, sta, mvmvif->phy_ctxt->channel->band,
3267+ true);
32923268
32933269 if (vif->type == NL80211_IFTYPE_STATION &&
32943270 changed & IEEE80211_RC_NSS_CHANGED)
@@ -3439,20 +3415,12 @@ static int iwl_mvm_mac_set_key(struct ieee80211_hw *hw, · 144 unmodified lines
34393415 break;
34403416 case WLAN_CIPHER_SUITE_WEP40:
34413417 case WLAN_CIPHER_SUITE_WEP104:
3442− if (vif->type == NL80211_IFTYPE_AP) {
3443− struct iwl_mvm_vif *mvmvif =
3444− iwl_mvm_vif_from_mac80211(vif);
3445−
3446− mvmvif->ap_wep_key = kmemdup(key,
3447− sizeof(*key) + key->keylen,
3448− GFP_KERNEL);
3449− if (!mvmvif->ap_wep_key)
3450− return -ENOMEM;
3451− }
3452−
3453− if (vif->type != NL80211_IFTYPE_STATION)
3454− return 0;
3455− break;
3418+ if (vif->type == NL80211_IFTYPE_STATION)
3419+ break;
3420+ if (iwl_mvm_has_new_tx_api(mvm))
3421+ return -EOPNOTSUPP;
3422+ /* support HW crypto on TX */
3423+ return 0;
34563424 default:
34573425 /* currently FW supports only one optional cipher scheme */
34583426 if (hw->n_cipher_schemes &&
@@ -3540,12 +3508,17 @@ static int iwl_mvm_mac_set_key(struct ieee80211_hw *hw, · 81 unmodified lines
35403508 ret = iwl_mvm_set_sta_key(mvm, vif, sta, key, key_offset);
35413509 if (ret) {
35423510 IWL_WARN(mvm, "set key failed\n");
3511+ key->hw_key_idx = STA_KEY_IDX_INVALID;
35433512 /*
35443513 * can't add key for RX, but we don't need it
3545− * in the device for TX so still return 0
3514+ * in the device for TX so still return 0,
3515+ * unless we have new TX API where we cannot
3516+ * put key material into the TX_CMD
35463517 */
3547− key->hw_key_idx = STA_KEY_IDX_INVALID;
3548− ret = 0;
3518+ if (iwl_mvm_has_new_tx_api(mvm))
3519+ ret = -EOPNOTSUPP;
3520+ else
3521+ ret = 0;
35493522 }
35503523
35513524 break;
drivers/net/wireless/intel/iwlwifi/mvm/mvm.h+0−1
@@ -498,7 +498,6 @@ struct iwl_mvm_vif { · 497 unmodified lines
498498 netdev_features_t features;
499499
500500 struct iwl_probe_resp_data __rcu *probe_resp_data;
501− struct ieee80211_key_conf *ap_wep_key;
502501 };
503502
504503 static inline struct iwl_mvm_vif *
drivers/net/wireless/intel/iwlwifi/mvm/sta.c+4−39
@@ -8,7 +8,7 @@ · 7 unmodified lines
88 * Copyright(c) 2012 - 2015 Intel Corporation. All rights reserved.
99 * Copyright(c) 2013 - 2015 Intel Mobile Communications GmbH
1010 * Copyright(c) 2016 - 2017 Intel Deutschland GmbH
11− * Copyright(c) 2018 Intel Corporation
11+ * Copyright(c) 2018 - 2019 Intel Corporation
1212 *
1313 * This program is free software; you can redistribute it and/or modify
1414 * it under the terms of version 2 of the GNU General Public License as
@@ -31,7 +31,7 @@ · 16 unmodified lines
3131 * Copyright(c) 2012 - 2015 Intel Corporation. All rights reserved.
3232 * Copyright(c) 2013 - 2015 Intel Mobile Communications GmbH
3333 * Copyright(c) 2016 - 2017 Intel Deutschland GmbH
34− * Copyright(c) 2018 Intel Corporation
34+ * Copyright(c) 2018 - 2019 Intel Corporation
3535 * All rights reserved.
3636 *
3737 * Redistribution and use in source and binary forms, with or without
@@ -1399,7 +1399,9 @@ void iwl_mvm_add_new_dqa_stream_wk(struct work_struct *wk) · 1361 unmodified lines
13991399
14001400 iwl_mvm_sta_alloc_queue(mvm, txq->sta, txq->ac, tid);
14011401 list_del_init(&mvmtxq->list);
1402+ local_bh_disable();
14021403 iwl_mvm_mac_itxq_xmit(mvm->hw, txq);
1404+ local_bh_enable();
14031405 }
14041406
14051407 mutex_unlock(&mvm->mutex);
@@ -2333,21 +2335,6 @@ int iwl_mvm_add_mcast_sta(struct iwl_mvm *mvm, struct ieee80211_vif *vif) · 927 unmodified lines
23332335 iwl_mvm_enable_txq(mvm, NULL, mvmvif->cab_queue, 0, &cfg,
23342336 timeout);
23352337
2336− if (mvmvif->ap_wep_key) {
2337− u8 key_offset = iwl_mvm_set_fw_key_idx(mvm);
2338−
2339− __set_bit(key_offset, mvm->fw_key_table);
2340−
2341− if (key_offset == STA_KEY_IDX_INVALID)
2342− return -ENOSPC;
2343−
2344− ret = iwl_mvm_send_sta_key(mvm, mvmvif->mcast_sta.sta_id,
2345− mvmvif->ap_wep_key, true, 0, NULL, 0,
2346− key_offset, 0);
2347− if (ret)
2348− return ret;
2349− }
2350−
23512338 return 0;
23522339 }
23532340
@@ -2419,28 +2406,6 @@ int iwl_mvm_rm_mcast_sta(struct iwl_mvm *mvm, struct ieee80211_vif *vif) · 65 unmodified lines
24192406
24202407 iwl_mvm_disable_txq(mvm, NULL, mvmvif->cab_queue, 0, 0);
24212408
2422− if (mvmvif->ap_wep_key) {
2423− int i;
2424−
2425− if (!__test_and_clear_bit(mvmvif->ap_wep_key->hw_key_idx,
2426− mvm->fw_key_table)) {
2427− IWL_ERR(mvm, "offset %d not used in fw key table.\n",
2428− mvmvif->ap_wep_key->hw_key_idx);
2429− return -ENOENT;
2430− }
2431−
2432− /* track which key was deleted last */
2433− for (i = 0; i < STA_KEY_MAX_NUM; i++) {
2434− if (mvm->fw_key_deleted[i] < U8_MAX)
2435− mvm->fw_key_deleted[i]++;
2436− }
2437− mvm->fw_key_deleted[mvmvif->ap_wep_key->hw_key_idx] = 0;
2438− ret = __iwl_mvm_remove_sta_key(mvm, mvmvif->mcast_sta.sta_id,
2439− mvmvif->ap_wep_key, true);
2440− if (ret)
2441− return ret;
2442− }
2443−
24442409 ret = iwl_mvm_rm_sta_common(mvm, mvmvif->mcast_sta.sta_id);
24452410 if (ret)
24462411 IWL_WARN(mvm, "Failed sending remove station\n");
drivers/net/wireless/intel/iwlwifi/mvm/sta.h+2−5
@@ -8,7 +8,7 @@ · 7 unmodified lines
88 * Copyright(c) 2012 - 2014 Intel Corporation. All rights reserved.
99 * Copyright(c) 2013 - 2014 Intel Mobile Communications GmbH
1010 * Copyright(c) 2015 - 2016 Intel Deutschland GmbH
11− * Copyright(c) 2018 Intel Corporation
11+ * Copyright(c) 2018 - 2019 Intel Corporation
1212 *
1313 * This program is free software; you can redistribute it and/or modify
1414 * it under the terms of version 2 of the GNU General Public License as
@@ -31,7 +31,7 @@ · 16 unmodified lines
3131 * Copyright(c) 2012 - 2014 Intel Corporation. All rights reserved.
3232 * Copyright(c) 2013 - 2014 Intel Mobile Communications GmbH
3333 * Copyright(c) 2015 - 2016 Intel Deutschland GmbH
34− * Copyright(c) 2018 Intel Corporation
34+ * Copyright(c) 2018 - 2019 Intel Corporation
3535 * All rights reserved.
3636 *
3737 * Redistribution and use in source and binary forms, with or without
@@ -394,7 +394,6 @@ struct iwl_mvm_rxq_dup_data { · 356 unmodified lines
394394 * the BA window. To be used for UAPSD only.
395395 * @ptk_pn: per-queue PTK PN data structures
396396 * @dup_data: per queue duplicate packet detection data
397− * @wep_key: used in AP mode. Is a duplicate of the WEP key.
398397 * @deferred_traffic_tid_map: indication bitmap of deferred traffic per-TID
399398 * @tx_ant: the index of the antenna to use for data tx to this station. Only
400399 * used during connection establishment (e.g. for the 4 way handshake
@@ -426,8 +425,6 @@ struct iwl_mvm_sta { · 25 unmodified lines
426425 struct iwl_mvm_key_pn __rcu *ptk_pn[4];
427426 struct iwl_mvm_rxq_dup_data *dup_data;
428427
429− struct ieee80211_key_conf *wep_key;
430−
431428 u8 reserved_queue;
432429
433430 /* Temporary, until the new TLC will control the Tx protection */
drivers/net/wireless/intel/iwlwifi/pcie/drv.c+7−6
@@ -953,14 +953,15 @@ static const struct pci_device_id iwl_hw_card_ids[] = { · 952 unmodified lines
953953 {IWL_PCI_DEVICE(0xA0F0, 0x1652, killer1650i_2ax_cfg_qu_b0_hr_b0)},
954954 {IWL_PCI_DEVICE(0xA0F0, 0x4070, iwl_ax101_cfg_qu_hr)},
955955
956− {IWL_PCI_DEVICE(0x2723, 0x0080, iwl22260_2ax_cfg)},
957− {IWL_PCI_DEVICE(0x2723, 0x0084, iwl22260_2ax_cfg)},
958− {IWL_PCI_DEVICE(0x2723, 0x0088, iwl22260_2ax_cfg)},
959− {IWL_PCI_DEVICE(0x2723, 0x008C, iwl22260_2ax_cfg)},
956+ {IWL_PCI_DEVICE(0x2723, 0x0080, iwl_ax200_cfg_cc)},
957+ {IWL_PCI_DEVICE(0x2723, 0x0084, iwl_ax200_cfg_cc)},
958+ {IWL_PCI_DEVICE(0x2723, 0x0088, iwl_ax200_cfg_cc)},
959+ {IWL_PCI_DEVICE(0x2723, 0x008C, iwl_ax200_cfg_cc)},
960960 {IWL_PCI_DEVICE(0x2723, 0x1653, killer1650w_2ax_cfg)},
961961 {IWL_PCI_DEVICE(0x2723, 0x1654, killer1650x_2ax_cfg)},
962− {IWL_PCI_DEVICE(0x2723, 0x4080, iwl22260_2ax_cfg)},
963− {IWL_PCI_DEVICE(0x2723, 0x4088, iwl22260_2ax_cfg)},
962+ {IWL_PCI_DEVICE(0x2723, 0x2080, iwl_ax200_cfg_cc)},
963+ {IWL_PCI_DEVICE(0x2723, 0x4080, iwl_ax200_cfg_cc)},
964+ {IWL_PCI_DEVICE(0x2723, 0x4088, iwl_ax200_cfg_cc)},
964965
965966 {IWL_PCI_DEVICE(0x1a56, 0x1653, killer1650w_2ax_cfg)},
966967 {IWL_PCI_DEVICE(0x1a56, 0x1654, killer1650x_2ax_cfg)},
drivers/net/wireless/intel/iwlwifi/pcie/internal.h+1−1
@@ -1043,7 +1043,7 @@ static inline bool iwl_pcie_dbg_on(struct iwl_trans *trans) · 1042 unmodified lines
10431043
10441044 void iwl_trans_pcie_rf_kill(struct iwl_trans *trans, bool state);
10451045 void iwl_trans_pcie_dump_regs(struct iwl_trans *trans);
1046−void iwl_trans_sync_nmi(struct iwl_trans *trans);
1046+void iwl_trans_pcie_sync_nmi(struct iwl_trans *trans);
10471047
10481048 #ifdef CONFIG_IWLWIFI_DEBUGFS
10491049 int iwl_trans_pcie_dbgfs_register(struct iwl_trans *trans);
drivers/net/wireless/intel/iwlwifi/pcie/trans.c+8−3
@@ -3318,7 +3318,8 @@ static void iwl_trans_pcie_resume(struct iwl_trans *trans) · 3317 unmodified lines
33183318 .unref = iwl_trans_pcie_unref, \
33193319 .dump_data = iwl_trans_pcie_dump_data, \
33203320 .d3_suspend = iwl_trans_pcie_d3_suspend, \
3321− .d3_resume = iwl_trans_pcie_d3_resume
3321+ .d3_resume = iwl_trans_pcie_d3_resume, \
3322+ .sync_nmi = iwl_trans_pcie_sync_nmi
33223323
33233324 #ifdef CONFIG_PM_SLEEP
33243325 #define IWL_TRANS_PM_OPS \
@@ -3542,6 +3543,10 @@ struct iwl_trans *iwl_trans_pcie_alloc(struct pci_dev *pdev, · 217 unmodified lines
35423543 }
35433544 } else if (cfg == &iwl_ax101_cfg_qu_hr) {
35443545 if (CSR_HW_RF_ID_TYPE_CHIP_ID(trans->hw_rf_id) ==
3546+ CSR_HW_RF_ID_TYPE_CHIP_ID(CSR_HW_RF_ID_TYPE_HR) &&
3547+ trans->hw_rev == CSR_HW_REV_TYPE_QNJ_B0) {
3548+ trans->cfg = &iwl22000_2ax_cfg_qnj_hr_b0;
3549+ } else if (CSR_HW_RF_ID_TYPE_CHIP_ID(trans->hw_rf_id) ==
35453550 CSR_HW_RF_ID_TYPE_CHIP_ID(CSR_HW_RF_ID_TYPE_HR)) {
35463551 trans->cfg = &iwl_ax101_cfg_qu_hr;
35473552 } else if (CSR_HW_RF_ID_TYPE_CHIP_ID(trans->hw_rf_id) ==
@@ -3560,7 +3565,7 @@ struct iwl_trans *iwl_trans_pcie_alloc(struct pci_dev *pdev, · 12 unmodified lines
35603565 }
35613566 } else if (CSR_HW_RF_ID_TYPE_CHIP_ID(trans->hw_rf_id) ==
35623567 CSR_HW_RF_ID_TYPE_CHIP_ID(CSR_HW_RF_ID_TYPE_HR) &&
3563− (trans->cfg != &iwl22260_2ax_cfg ||
3568+ (trans->cfg != &iwl_ax200_cfg_cc ||
35643569 trans->hw_rev == CSR_HW_REV_TYPE_QNJ_B0)) {
35653570 u32 hw_status;
35663571
@@ -3637,7 +3642,7 @@ out_no_pci: · 70 unmodified lines
36373642 return ERR_PTR(ret);
36383643 }
36393644
3640−void iwl_trans_sync_nmi(struct iwl_trans *trans)
3645+void iwl_trans_pcie_sync_nmi(struct iwl_trans *trans)
36413646 {
36423647 unsigned long timeout = jiffies + IWL_TRANS_NMI_TIMEOUT;
36433648
drivers/net/wireless/intel/iwlwifi/pcie/tx-gen2.c+1−1
@@ -965,7 +965,7 @@ static int iwl_pcie_gen2_send_hcmd_sync(struct iwl_trans *trans, · 964 unmodified lines
965965 cmd_str);
966966 ret = -ETIMEDOUT;
967967
968− iwl_trans_sync_nmi(trans);
968+ iwl_trans_pcie_sync_nmi(trans);
969969 goto cancel;
970970 }
971971
drivers/net/wireless/intel/iwlwifi/pcie/tx.c+1−1
@@ -1960,7 +1960,7 @@ static int iwl_pcie_send_hcmd_sync(struct iwl_trans *trans, · 1959 unmodified lines
19601960 iwl_get_cmd_string(trans, cmd->id));
19611961 ret = -ETIMEDOUT;
19621962
1963− iwl_trans_sync_nmi(trans);
1963+ iwl_trans_pcie_sync_nmi(trans);
19641964 goto cancel;
19651965 }
19661966
drivers/net/wireless/mac80211_hwsim.c+15−4
@@ -2644,7 +2644,7 @@ static int mac80211_hwsim_new_radio(struct genl_info *info, · 2643 unmodified lines
26442644 enum nl80211_band band;
26452645 const struct ieee80211_ops *ops = &mac80211_hwsim_ops;
26462646 struct net *net;
2647− int idx;
2647+ int idx, i;
26482648 int n_limits = 0;
26492649
26502650 if (WARN_ON(param->channels > 1 && !param->use_chanctx))
@@ -2768,12 +2768,23 @@ static int mac80211_hwsim_new_radio(struct genl_info *info, · 117 unmodified lines
27682768 goto failed_hw;
27692769 }
27702770
2771+ data->if_combination.max_interfaces = 0;
2772+ for (i = 0; i < n_limits; i++)
2773+ data->if_combination.max_interfaces +=
2774+ data->if_limits[i].max;
2775+
27712776 data->if_combination.n_limits = n_limits;
2772− data->if_combination.max_interfaces = 2048;
27732777 data->if_combination.limits = data->if_limits;
27742778
2775− hw->wiphy->iface_combinations = &data->if_combination;
2776− hw->wiphy->n_iface_combinations = 1;
2779+ /*
2780+ * If we actually were asked to support combinations,
2781+ * advertise them - if there's only a single thing like
2782+ * only IBSS then don't advertise it as combinations.
2783+ */
2784+ if (data->if_combination.max_interfaces > 1) {
2785+ hw->wiphy->iface_combinations = &data->if_combination;
2786+ hw->wiphy->n_iface_combinations = 1;
2787+ }
27772788
27782789 if (param->ciphers) {
27792790 memcpy(data->ciphers, param->ciphers,
drivers/net/wireless/mediatek/mt76/mt7603/init.c+2−0
@@ -510,6 +510,8 @@ int mt7603_register_device(struct mt7603_dev *dev) · 509 unmodified lines
510510 bus_ops->rmw = mt7603_rmw;
511511 dev->mt76.bus = bus_ops;
512512
513+ spin_lock_init(&dev->ps_lock);
514+
513515 INIT_DELAYED_WORK(&dev->mac_work, mt7603_mac_work);
514516 tasklet_init(&dev->pre_tbtt_tasklet, mt7603_pre_tbtt_tasklet,
515517 (unsigned long)dev);
drivers/net/wireless/mediatek/mt76/mt7603/mac.c+14−39
@@ -343,7 +343,7 @@ void mt7603_mac_rx_ba_reset(struct mt7603_dev *dev, void *addr, u8 tid) · 342 unmodified lines
343343 MT_BA_CONTROL_1_RESET));
344344 }
345345
346−void mt7603_mac_tx_ba_reset(struct mt7603_dev *dev, int wcid, int tid, int ssn,
346+void mt7603_mac_tx_ba_reset(struct mt7603_dev *dev, int wcid, int tid,
347347 int ba_size)
348348 {
349349 u32 addr = mt7603_wtbl2_addr(wcid);
@@ -358,43 +358,6 @@ void mt7603_mac_tx_ba_reset(struct mt7603_dev *dev, int wcid, int tid, int ssn, · 8 unmodified lines
358358 mt76_clear(dev, addr + (15 * 4), tid_mask);
359359 return;
360360 }
361− mt76_poll(dev, MT_WTBL_UPDATE, MT_WTBL_UPDATE_BUSY, 0, 5000);
362−
363− mt7603_mac_stop(dev);
364− switch (tid) {
365− case 0:
366− mt76_rmw_field(dev, addr + (2 * 4), MT_WTBL2_W2_TID0_SN, ssn);
367− break;
368− case 1:
369− mt76_rmw_field(dev, addr + (2 * 4), MT_WTBL2_W2_TID1_SN, ssn);
370− break;
371− case 2:
372− mt76_rmw_field(dev, addr + (2 * 4), MT_WTBL2_W2_TID2_SN_LO,
373− ssn);
374− mt76_rmw_field(dev, addr + (3 * 4), MT_WTBL2_W3_TID2_SN_HI,
375− ssn >> 8);
376− break;
377− case 3:
378− mt76_rmw_field(dev, addr + (3 * 4), MT_WTBL2_W3_TID3_SN, ssn);
379− break;
380− case 4:
381− mt76_rmw_field(dev, addr + (3 * 4), MT_WTBL2_W3_TID4_SN, ssn);
382− break;
383− case 5:
384− mt76_rmw_field(dev, addr + (3 * 4), MT_WTBL2_W3_TID5_SN_LO,
385− ssn);
386− mt76_rmw_field(dev, addr + (4 * 4), MT_WTBL2_W4_TID5_SN_HI,
387− ssn >> 4);
388− break;
389− case 6:
390− mt76_rmw_field(dev, addr + (4 * 4), MT_WTBL2_W4_TID6_SN, ssn);
391− break;
392− case 7:
393− mt76_rmw_field(dev, addr + (4 * 4), MT_WTBL2_W4_TID7_SN, ssn);
394− break;
395− }
396− mt7603_wtbl_update(dev, wcid, MT_WTBL_UPDATE_WTBL2);
397− mt7603_mac_start(dev);
398361
399362 for (i = 7; i > 0; i--) {
400363 if (ba_size >= MT_AGG_SIZE_LIMIT(i))
@@ -827,6 +790,7 @@ mt7603_mac_write_txwi(struct mt7603_dev *dev, __le32 *txwi, · 426 unmodified lines
827790 struct ieee80211_tx_info *info = IEEE80211_SKB_CB(skb);
828791 struct ieee80211_tx_rate *rate = &info->control.rates[0];
829792 struct ieee80211_hdr *hdr = (struct ieee80211_hdr *)skb->data;
793+ struct ieee80211_bar *bar = (struct ieee80211_bar *)skb->data;
830794 struct ieee80211_vif *vif = info->control.vif;
831795 struct mt7603_vif *mvif;
832796 int wlan_idx;
@@ -834,6 +798,7 @@ mt7603_mac_write_txwi(struct mt7603_dev *dev, __le32 *txwi, · 1 unmodified line
834798 int tx_count = 8;
835799 u8 frame_type, frame_subtype;
836800 u16 fc = le16_to_cpu(hdr->frame_control);
801+ u16 seqno = 0;
837802 u8 vif_idx = 0;
838803 u32 val;
839804 u8 bw;
@@ -919,7 +884,17 @@ mt7603_mac_write_txwi(struct mt7603_dev *dev, __le32 *txwi, · 79 unmodified lines
919884 tx_count = 0x1f;
920885
921886 val = FIELD_PREP(MT_TXD3_REM_TX_COUNT, tx_count) |
922− FIELD_PREP(MT_TXD3_SEQ, le16_to_cpu(hdr->seq_ctrl));
887+ MT_TXD3_SN_VALID;
888+
889+ if (ieee80211_is_data_qos(hdr->frame_control))
890+ seqno = le16_to_cpu(hdr->seq_ctrl);
891+ else if (ieee80211_is_back_req(hdr->frame_control))
892+ seqno = le16_to_cpu(bar->start_seq_num);
893+ else
894+ val &= ~MT_TXD3_SN_VALID;
895+
896+ val |= FIELD_PREP(MT_TXD3_SEQ, seqno >> 4);
897+
923898 txwi[3] = cpu_to_le32(val);
924899
925900 if (key) {
drivers/net/wireless/mediatek/mt76/mt7603/main.c+4−4
@@ -372,7 +372,7 @@ mt7603_sta_ps(struct mt76_dev *mdev, struct ieee80211_sta *sta, bool ps) · 371 unmodified lines
372372 struct mt7603_sta *msta = (struct mt7603_sta *)sta->drv_priv;
373373 struct sk_buff_head list;
374374
375− mt76_stop_tx_queues(&dev->mt76, sta, false);
375+ mt76_stop_tx_queues(&dev->mt76, sta, true);
376376 mt7603_wtbl_set_ps(dev, msta, ps);
377377 if (ps)
378378 return;
@@ -584,13 +584,13 @@ mt7603_ampdu_action(struct ieee80211_hw *hw, struct ieee80211_vif *vif, · 205 unmodified lines
584584 case IEEE80211_AMPDU_TX_OPERATIONAL:
585585 mtxq->aggr = true;
586586 mtxq->send_bar = false;
587− mt7603_mac_tx_ba_reset(dev, msta->wcid.idx, tid, *ssn, ba_size);
587+ mt7603_mac_tx_ba_reset(dev, msta->wcid.idx, tid, ba_size);
588588 break;
589589 case IEEE80211_AMPDU_TX_STOP_FLUSH:
590590 case IEEE80211_AMPDU_TX_STOP_FLUSH_CONT:
591591 mtxq->aggr = false;
592592 ieee80211_send_bar(vif, sta->addr, tid, mtxq->agg_ssn);
593− mt7603_mac_tx_ba_reset(dev, msta->wcid.idx, tid, *ssn, -1);
593+ mt7603_mac_tx_ba_reset(dev, msta->wcid.idx, tid, -1);
594594 break;
595595 case IEEE80211_AMPDU_TX_START:
596596 mtxq->agg_ssn = *ssn << 4;
@@ -598,7 +598,7 @@ mt7603_ampdu_action(struct ieee80211_hw *hw, struct ieee80211_vif *vif, · 1 unmodified line
598598 break;
599599 case IEEE80211_AMPDU_TX_STOP_CONT:
600600 mtxq->aggr = false;
601− mt7603_mac_tx_ba_reset(dev, msta->wcid.idx, tid, *ssn, -1);
601+ mt7603_mac_tx_ba_reset(dev, msta->wcid.idx, tid, -1);
602602 ieee80211_stop_tx_ba_cb_irqsafe(vif, sta->addr, tid);
603603 break;
604604 }
drivers/net/wireless/mediatek/mt76/mt7603/mt7603.h+1−1
@@ -200,7 +200,7 @@ void mt7603_beacon_set_timer(struct mt7603_dev *dev, int idx, int intval); · 199 unmodified lines
200200 int mt7603_mac_fill_rx(struct mt7603_dev *dev, struct sk_buff *skb);
201201 void mt7603_mac_add_txs(struct mt7603_dev *dev, void *data);
202202 void mt7603_mac_rx_ba_reset(struct mt7603_dev *dev, void *addr, u8 tid);
203−void mt7603_mac_tx_ba_reset(struct mt7603_dev *dev, int wcid, int tid, int ssn,
203+void mt7603_mac_tx_ba_reset(struct mt7603_dev *dev, int wcid, int tid,
204204 int ba_size);
205205
206206 void mt7603_pse_client_reset(struct mt7603_dev *dev);
drivers/net/wireless/mediatek/mt76/mt76x02_mac.c+8−6
@@ -466,7 +466,6 @@ void mt76x02_send_tx_status(struct mt76x02_dev *dev, · 465 unmodified lines
466466 return;
467467
468468 rcu_read_lock();
469− mt76_tx_status_lock(mdev, &list);
470469
471470 if (stat->wcid < ARRAY_SIZE(dev->mt76.wcid))
472471 wcid = rcu_dereference(dev->mt76.wcid[stat->wcid]);
@@ -479,6 +478,8 @@ void mt76x02_send_tx_status(struct mt76x02_dev *dev, · 6 unmodified lines
479478 drv_priv);
480479 }
481480
481+ mt76_tx_status_lock(mdev, &list);
482+
482483 if (wcid) {
483484 if (stat->pktid >= MT_PACKET_ID_FIRST)
484485 status.skb = mt76_tx_status_skb_get(mdev, wcid,
@@ -498,7 +499,9 @@ void mt76x02_send_tx_status(struct mt76x02_dev *dev, · 13 unmodified lines
498499 if (*update == 0 && stat_val == stat_cache &&
499500 stat->wcid == msta->status.wcid && msta->n_frames < 32) {
500501 msta->n_frames++;
501− goto out;
502+ mt76_tx_status_unlock(mdev, &list);
503+ rcu_read_unlock();
504+ return;
502505 }
503506
504507 mt76x02_mac_fill_tx_status(dev, status.info, &msta->status,
@@ -514,11 +517,10 @@ void mt76x02_send_tx_status(struct mt76x02_dev *dev, · 9 unmodified lines
514517
515518 if (status.skb)
516519 mt76_tx_status_skb_done(mdev, status.skb, &list);
517− else
518− ieee80211_tx_status_ext(mt76_hw(dev), &status);
519−
520−out:
521520 mt76_tx_status_unlock(mdev, &list);
521+
522+ if (!status.skb)
523+ ieee80211_tx_status_ext(mt76_hw(dev), &status);
522524 rcu_read_unlock();
523525 }
524526
drivers/net/wireless/ralink/rt2x00/rt2x00.h+0−1
@@ -673,7 +673,6 @@ enum rt2x00_state_flags { · 672 unmodified lines
673673 CONFIG_CHANNEL_HT40,
674674 CONFIG_POWERSAVING,
675675 CONFIG_HT_DISABLED,
676− CONFIG_QOS_DISABLED,
677676 CONFIG_MONITORING,
678677
679678 /*
drivers/net/wireless/ralink/rt2x00/rt2x00mac.c+0−10
@@ -642,18 +642,8 @@ void rt2x00mac_bss_info_changed(struct ieee80211_hw *hw, · 641 unmodified lines
642642 rt2x00dev->intf_associated--;
643643
644644 rt2x00leds_led_assoc(rt2x00dev, !!rt2x00dev->intf_associated);
645−
646− clear_bit(CONFIG_QOS_DISABLED, &rt2x00dev->flags);
647645 }
648646
649− /*
650− * Check for access point which do not support 802.11e . We have to
651− * generate data frames sequence number in S/W for such AP, because
652− * of H/W bug.
653− */
654− if (changes & BSS_CHANGED_QOS && !bss_conf->qos)
655− set_bit(CONFIG_QOS_DISABLED, &rt2x00dev->flags);
656−
657647 /*
658648 * When the erp information has changed, we should perform
659649 * additional configuration steps. For all other changes we are done.
drivers/net/wireless/ralink/rt2x00/rt2x00queue.c+9−6
@@ -201,15 +201,18 @@ static void rt2x00queue_create_tx_descriptor_seq(struct rt2x00_dev *rt2x00dev, · 200 unmodified lines
201201 if (!rt2x00_has_cap_flag(rt2x00dev, REQUIRE_SW_SEQNO)) {
202202 /*
203203 * rt2800 has a H/W (or F/W) bug, device incorrectly increase
204− * seqno on retransmited data (non-QOS) frames. To workaround
205− * the problem let's generate seqno in software if QOS is
206− * disabled.
204+ * seqno on retransmitted data (non-QOS) and management frames.
205+ * To workaround the problem let's generate seqno in software.
206+ * Except for beacons which are transmitted periodically by H/W
207+ * hence hardware has to assign seqno for them.
207208 */
208− if (test_bit(CONFIG_QOS_DISABLED, &rt2x00dev->flags))
209− __clear_bit(ENTRY_TXD_GENERATE_SEQ, &txdesc->flags);
210− else
209+ if (ieee80211_is_beacon(hdr->frame_control)) {
210+ __set_bit(ENTRY_TXD_GENERATE_SEQ, &txdesc->flags);
211211 /* H/W will generate sequence number */
212212 return;
213+ }
214+
215+ __clear_bit(ENTRY_TXD_GENERATE_SEQ, &txdesc->flags);
213216 }
214217
215218 /*
drivers/vhost/vhost.c+5−1
@@ -911,8 +911,12 @@ static int vhost_new_umem_range(struct vhost_umem *umem, · 910 unmodified lines
911911 u64 start, u64 size, u64 end,
912912 u64 userspace_addr, int perm)
913913 {
914− struct vhost_umem_node *tmp, *node = kmalloc(sizeof(*node), GFP_ATOMIC);
914+ struct vhost_umem_node *tmp, *node;
915915
916+ if (!size)
917+ return -EFAULT;
918+
919+ node = kmalloc(sizeof(*node), GFP_ATOMIC);
916920 if (!node)
917921 return -ENOMEM;
918922
fs/afs/rxrpc.c+17−7
@@ -610,6 +610,7 @@ static long afs_wait_for_call_to_complete(struct afs_call *call, · 609 unmodified lines
610610 bool stalled = false;
611611 u64 rtt;
612612 u32 life, last_life;
613+ bool rxrpc_complete = false;
613614
614615 DECLARE_WAITQUEUE(myself, current);
615616
@@ -621,7 +622,7 @@ static long afs_wait_for_call_to_complete(struct afs_call *call, · 5 unmodified lines
621622 rtt2 = 2;
622623
623624 timeout = rtt2;
624− last_life = rxrpc_kernel_check_life(call->net->socket, call->rxcall);
625+ rxrpc_kernel_check_life(call->net->socket, call->rxcall, &last_life);
625626
626627 add_wait_queue(&call->waitq, &myself);
627628 for (;;) {
@@ -639,7 +640,12 @@ static long afs_wait_for_call_to_complete(struct afs_call *call, · 11 unmodified lines
639640 if (afs_check_call_state(call, AFS_CALL_COMPLETE))
640641 break;
641642
642− life = rxrpc_kernel_check_life(call->net->socket, call->rxcall);
643+ if (!rxrpc_kernel_check_life(call->net->socket, call->rxcall, &life)) {
644+ /* rxrpc terminated the call. */
645+ rxrpc_complete = true;
646+ break;
647+ }
648+
643649 if (timeout == 0 &&
644650 life == last_life && signal_pending(current)) {
645651 if (stalled)
@@ -663,12 +669,16 @@ static long afs_wait_for_call_to_complete(struct afs_call *call, · 17 unmodified lines
663669 remove_wait_queue(&call->waitq, &myself);
664670 __set_current_state(TASK_RUNNING);
665671
666− /* Kill off the call if it's still live. */
667672 if (!afs_check_call_state(call, AFS_CALL_COMPLETE)) {
668− _debug("call interrupted");
669− if (rxrpc_kernel_abort_call(call->net->socket, call->rxcall,
670− RX_USER_ABORT, -EINTR, "KWI"))
671− afs_set_call_complete(call, -EINTR, 0);
673+ if (rxrpc_complete) {
674+ afs_set_call_complete(call, call->error, call->abort_code);
675+ } else {
676+ /* Kill off the call if it's still live. */
677+ _debug("call interrupted");
678+ if (rxrpc_kernel_abort_call(call->net->socket, call->rxcall,
679+ RX_USER_ABORT, -EINTR, "KWI"))
680+ afs_set_call_complete(call, -EINTR, 0);
681+ }
672682 }
673683
674684 spin_lock_bh(&call->state_lock);
include/linux/netdevice.h+3−0
@@ -1500,6 +1500,7 @@ struct net_device_ops { · 1499 unmodified lines
15001500 * @IFF_FAILOVER: device is a failover master device
15011501 * @IFF_FAILOVER_SLAVE: device is lower dev of a failover master device
15021502 * @IFF_L3MDEV_RX_HANDLER: only invoke the rx handler of L3 master device
1503+ * @IFF_LIVE_RENAME_OK: rename is allowed while device is up and running
15031504 */
15041505 enum netdev_priv_flags {
15051506 IFF_802_1Q_VLAN = 1<<0,
@@ -1532,6 +1533,7 @@ enum netdev_priv_flags { · 26 unmodified lines
15321533 IFF_FAILOVER = 1<<27,
15331534 IFF_FAILOVER_SLAVE = 1<<28,
15341535 IFF_L3MDEV_RX_HANDLER = 1<<29,
1536+ IFF_LIVE_RENAME_OK = 1<<30,
15351537 };
15361538
15371539 #define IFF_802_1Q_VLAN IFF_802_1Q_VLAN
@@ -1563,6 +1565,7 @@ enum netdev_priv_flags { · 25 unmodified lines
15631565 #define IFF_FAILOVER IFF_FAILOVER
15641566 #define IFF_FAILOVER_SLAVE IFF_FAILOVER_SLAVE
15651567 #define IFF_L3MDEV_RX_HANDLER IFF_L3MDEV_RX_HANDLER
1568+#define IFF_LIVE_RENAME_OK IFF_LIVE_RENAME_OK
15661569
15671570 /**
15681571 * struct net_device - The DEVICE structure.
include/net/af_rxrpc.h+3−1
@@ -61,10 +61,12 @@ int rxrpc_kernel_charge_accept(struct socket *, rxrpc_notify_rx_t, · 60 unmodified lines
6161 rxrpc_user_attach_call_t, unsigned long, gfp_t,
6262 unsigned int);
6363 void rxrpc_kernel_set_tx_length(struct socket *, struct rxrpc_call *, s64);
64−u32 rxrpc_kernel_check_life(const struct socket *, const struct rxrpc_call *);
64+bool rxrpc_kernel_check_life(const struct socket *, const struct rxrpc_call *,
65+ u32 *);
6566 void rxrpc_kernel_probe_life(struct socket *, struct rxrpc_call *);
6667 u32 rxrpc_kernel_get_epoch(struct socket *, struct rxrpc_call *);
6768 bool rxrpc_kernel_get_reply_time(struct socket *, struct rxrpc_call *,
6869 ktime_t *);
70+bool rxrpc_kernel_call_is_complete(struct rxrpc_call *);
6971
7072 #endif /* _NET_RXRPC_H */
include/net/cfg80211.h+5−0
@@ -7183,6 +7183,11 @@ void cfg80211_pmsr_complete(struct wireless_dev *wdev, · 7182 unmodified lines
71837183 #define wiphy_info(wiphy, format, args...) \
71847184 dev_info(&(wiphy)->dev, format, ##args)
71857185
7186+#define wiphy_err_ratelimited(wiphy, format, args...) \
7187+ dev_err_ratelimited(&(wiphy)->dev, format, ##args)
7188+#define wiphy_warn_ratelimited(wiphy, format, args...) \
7189+ dev_warn_ratelimited(&(wiphy)->dev, format, ##args)
7190+
71867191 #define wiphy_debug(wiphy, format, args...) \
71877192 wiphy_printk(KERN_DEBUG, wiphy, format, ##args)
71887193
include/net/mac80211.h+34−29
@@ -6231,8 +6231,6 @@ struct sk_buff *ieee80211_tx_dequeue(struct ieee80211_hw *hw, · 6230 unmodified lines
62316231 * @hw: pointer as obtained from ieee80211_alloc_hw()
62326232 * @ac: AC number to return packets from.
62336233 *
6234− * Should only be called between calls to ieee80211_txq_schedule_start()
6235− * and ieee80211_txq_schedule_end().
62366234 * Returns the next txq if successful, %NULL if no queue is eligible. If a txq
62376235 * is returned, it should be returned with ieee80211_return_txq() after the
62386236 * driver has finished scheduling it.
@@ -6240,51 +6238,58 @@ struct sk_buff *ieee80211_tx_dequeue(struct ieee80211_hw *hw, · 1 unmodified line
62406238 struct ieee80211_txq *ieee80211_next_txq(struct ieee80211_hw *hw, u8 ac);
62416239
62426240 /**
6243− * ieee80211_return_txq - return a TXQ previously acquired by ieee80211_next_txq()
6244− *
6245− * @hw: pointer as obtained from ieee80211_alloc_hw()
6246− * @txq: pointer obtained from station or virtual interface
6247− *
6248− * Should only be called between calls to ieee80211_txq_schedule_start()
6249− * and ieee80211_txq_schedule_end().
6250− */
6251−void ieee80211_return_txq(struct ieee80211_hw *hw, struct ieee80211_txq *txq);
6252−
6253−/**
6254− * ieee80211_txq_schedule_start - acquire locks for safe scheduling of an AC
6241+ * ieee80211_txq_schedule_start - start new scheduling round for TXQs
62556242 *
62566243 * @hw: pointer as obtained from ieee80211_alloc_hw()
62576244 * @ac: AC number to acquire locks for
62586245 *
6259− * Acquire locks needed to schedule TXQs from the given AC. Should be called
6260− * before ieee80211_next_txq() or ieee80211_return_txq().
6246+ * Should be called before ieee80211_next_txq() or ieee80211_return_txq().
6247+ * The driver must not call multiple TXQ scheduling rounds concurrently.
62616248 */
6262−void ieee80211_txq_schedule_start(struct ieee80211_hw *hw, u8 ac)
6263− __acquires(txq_lock);
6249+void ieee80211_txq_schedule_start(struct ieee80211_hw *hw, u8 ac);
6250+
6251+/* (deprecated) */
6252+static inline void ieee80211_txq_schedule_end(struct ieee80211_hw *hw, u8 ac)
6253+{
6254+}
6255+
6256+void __ieee80211_schedule_txq(struct ieee80211_hw *hw,
6257+ struct ieee80211_txq *txq, bool force);
62646258
62656259 /**
6266− * ieee80211_txq_schedule_end - release locks for safe scheduling of an AC
6260+ * ieee80211_schedule_txq - schedule a TXQ for transmission
62676261 *
62686262 * @hw: pointer as obtained from ieee80211_alloc_hw()
6269− * @ac: AC number to acquire locks for
6263+ * @txq: pointer obtained from station or virtual interface
62706264 *
6271− * Release locks previously acquired by ieee80211_txq_schedule_end().
6265+ * Schedules a TXQ for transmission if it is not already scheduled,
6266+ * even if mac80211 does not have any packets buffered.
6267+ *
6268+ * The driver may call this function if it has buffered packets for
6269+ * this TXQ internally.
62726270 */
6273−void ieee80211_txq_schedule_end(struct ieee80211_hw *hw, u8 ac)
6274− __releases(txq_lock);
6271+static inline void
6272+ieee80211_schedule_txq(struct ieee80211_hw *hw, struct ieee80211_txq *txq)
6273+{
6274+ __ieee80211_schedule_txq(hw, txq, true);
6275+}
62756276
62766277 /**
6277− * ieee80211_schedule_txq - schedule a TXQ for transmission
6278+ * ieee80211_return_txq - return a TXQ previously acquired by ieee80211_next_txq()
62786279 *
62796280 * @hw: pointer as obtained from ieee80211_alloc_hw()
62806281 * @txq: pointer obtained from station or virtual interface
6282+ * @force: schedule txq even if mac80211 does not have any buffered packets.
62816283 *
6282− * Schedules a TXQ for transmission if it is not already scheduled. Takes a
6283− * lock, which means it must *not* be called between
6284− * ieee80211_txq_schedule_start() and ieee80211_txq_schedule_end()
6284+ * The driver may set force=true if it has buffered packets for this TXQ
6285+ * internally.
62856286 */
6286−void ieee80211_schedule_txq(struct ieee80211_hw *hw, struct ieee80211_txq *txq)
6287− __acquires(txq_lock) __releases(txq_lock);
6287+static inline void
6288+ieee80211_return_txq(struct ieee80211_hw *hw, struct ieee80211_txq *txq,
6289+ bool force)
6290+{
6291+ __ieee80211_schedule_txq(hw, txq, force);
6292+}
62886293
62896294 /**
62906295 * ieee80211_txq_may_transmit - check whether TXQ is allowed to transmit
include/net/netrom.h+1−1
@@ -266,7 +266,7 @@ void nr_stop_idletimer(struct sock *); · 265 unmodified lines
266266 int nr_t1timer_running(struct sock *);
267267
268268 /* sysctl_net_netrom.c */
269−void nr_register_sysctl(void);
269+int nr_register_sysctl(void);
270270 void nr_unregister_sysctl(void);
271271
272272 #endif
include/net/sock.h+0−6
@@ -2084,12 +2084,6 @@ static inline bool skwq_has_sleeper(struct socket_wq *wq) · 2083 unmodified lines
20842084 * @p: poll_table
20852085 *
20862086 * See the comments in the wq_has_sleeper function.
2087− *
2088− * Do not derive sock from filp->private_data here. An SMC socket establishes
2089− * an internal TCP socket that is used in the fallback case. All socket
2090− * operations on the SMC socket are then forwarded to the TCP socket. In case of
2091− * poll, the filp->private_data pointer references the SMC socket because the
2092− * TCP socket has no file assigned.
20932087 */
20942088 static inline void sock_poll_wait(struct file *filp, struct socket *sock,
20952089 poll_table *p)
include/net/tls.h+3−1
@@ -307,6 +307,7 @@ int tls_device_sendmsg(struct sock *sk, struct msghdr *msg, size_t size); · 306 unmodified lines
307307 int tls_device_sendpage(struct sock *sk, struct page *page,
308308 int offset, size_t size, int flags);
309309 void tls_device_sk_destruct(struct sock *sk);
310+void tls_device_free_resources_tx(struct sock *sk);
310311 void tls_device_init(void);
311312 void tls_device_cleanup(void);
312313 int tls_tx_records(struct sock *sk, int flags);
@@ -330,6 +331,7 @@ int tls_push_sg(struct sock *sk, struct tls_context *ctx, · 17 unmodified lines
330331 int flags);
331332 int tls_push_partial_record(struct sock *sk, struct tls_context *ctx,
332333 int flags);
334+bool tls_free_partial_record(struct sock *sk, struct tls_context *ctx);
333335
334336 static inline struct tls_msg *tls_msg(struct sk_buff *skb)
335337 {
@@ -379,7 +381,7 @@ tls_validate_xmit_skb(struct sock *sk, struct net_device *dev, · 43 unmodified lines
379381 static inline bool tls_is_sk_tx_device_offloaded(struct sock *sk)
380382 {
381383 #ifdef CONFIG_SOCK_VALIDATE_XMIT
382− return sk_fullsock(sk) &
384+ return sk_fullsock(sk) &&
383385 (smp_load_acquire(&sk->sk_validate_xmit_skb) ==
384386 &tls_validate_xmit_skb);
385387 #else
net/atm/lec.c+5−1
@@ -710,7 +710,10 @@ static int lec_vcc_attach(struct atm_vcc *vcc, void __user *arg) · 709 unmodified lines
710710
711711 static int lec_mcast_attach(struct atm_vcc *vcc, int arg)
712712 {
713− if (arg < 0 || arg >= MAX_LEC_ITF || !dev_lec[arg])
713+ if (arg < 0 || arg >= MAX_LEC_ITF)
714+ return -EINVAL;
715+ arg = array_index_nospec(arg, MAX_LEC_ITF);
716+ if (!dev_lec[arg])
714717 return -EINVAL;
715718 vcc->proto_data = dev_lec[arg];
716719 return lec_mcast_make(netdev_priv(dev_lec[arg]), vcc);
@@ -728,6 +731,7 @@ static int lecd_attach(struct atm_vcc *vcc, int arg) · 11 unmodified lines
728731 i = arg;
729732 if (arg >= MAX_LEC_ITF)
730733 return -EINVAL;
734+ i = array_index_nospec(arg, MAX_LEC_ITF);
731735 if (!dev_lec[i]) {
732736 int size;
733737
net/bluetooth/sco.c+2−2
@@ -523,12 +523,12 @@ static int sco_sock_bind(struct socket *sock, struct sockaddr *addr, · 522 unmodified lines
523523 struct sock *sk = sock->sk;
524524 int err = 0;
525525
526− BT_DBG("sk %p %pMR", sk, &sa->sco_bdaddr);
527−
528526 if (!addr || addr_len < sizeof(struct sockaddr_sco) ||
529527 addr->sa_family != AF_BLUETOOTH)
530528 return -EINVAL;
531529
530+ BT_DBG("sk %p %pMR", sk, &sa->sco_bdaddr);
531+
532532 lock_sock(sk);
533533
534534 if (sk->sk_state != BT_OPEN) {
net/bridge/br_input.c+14−9
@@ -197,13 +197,10 @@ static void __br_handle_local_finish(struct sk_buff *skb) · 196 unmodified lines
197197 /* note: already called with rcu_read_lock */
198198 static int br_handle_local_finish(struct net *net, struct sock *sk, struct sk_buff *skb)
199199 {
200− struct net_bridge_port *p = br_port_get_rcu(skb->dev);
201−
202200 __br_handle_local_finish(skb);
203201
204− BR_INPUT_SKB_CB(skb)->brdev = p->br->dev;
205− br_pass_frame_up(skb);
206− return 0;
202+ /* return 1 to signal the okfn() was called so it's ok to use the skb */
203+ return 1;
207204 }
208205
209206 /*
@@ -280,10 +277,18 @@ rx_handler_result_t br_handle_frame(struct sk_buff **pskb) · 70 unmodified lines
280277 goto forward;
281278 }
282279
283− /* Deliver packet to local host only */
284− NF_HOOK(NFPROTO_BRIDGE, NF_BR_LOCAL_IN, dev_net(skb->dev),
285− NULL, skb, skb->dev, NULL, br_handle_local_finish);
286− return RX_HANDLER_CONSUMED;
280+ /* The else clause should be hit when nf_hook():
281+ * - returns < 0 (drop/error)
282+ * - returns = 0 (stolen/nf_queue)
283+ * Thus return 1 from the okfn() to signal the skb is ok to pass
284+ */
285+ if (NF_HOOK(NFPROTO_BRIDGE, NF_BR_LOCAL_IN,
286+ dev_net(skb->dev), NULL, skb, skb->dev, NULL,
287+ br_handle_local_finish) == 1) {
288+ return RX_HANDLER_PASS;
289+ } else {
290+ return RX_HANDLER_CONSUMED;
291+ }
287292 }
288293
289294 forward:
net/bridge/br_multicast.c+3−1
@@ -2031,7 +2031,8 @@ static void br_multicast_start_querier(struct net_bridge *br, · 2030 unmodified lines
20312031
20322032 __br_multicast_open(br, query);
20332033
2034− list_for_each_entry(port, &br->port_list, list) {
2034+ rcu_read_lock();
2035+ list_for_each_entry_rcu(port, &br->port_list, list) {
20352036 if (port->state == BR_STATE_DISABLED ||
20362037 port->state == BR_STATE_BLOCKING)
20372038 continue;
@@ -2043,6 +2044,7 @@ static void br_multicast_start_querier(struct net_bridge *br, · 5 unmodified lines
20432044 br_multicast_enable(&port->ip6_own_query);
20442045 #endif
20452046 }
2047+ rcu_read_unlock();
20462048 }
20472049
20482050 int br_multicast_toggle(struct net_bridge *br, unsigned long val)
net/bridge/br_netlink.c+1−1
@@ -1441,7 +1441,7 @@ static int br_fill_info(struct sk_buff *skb, const struct net_device *brdev) · 1440 unmodified lines
14411441 nla_put_u8(skb, IFLA_BR_VLAN_STATS_ENABLED,
14421442 br_opt_get(br, BROPT_VLAN_STATS_ENABLED)) ||
14431443 nla_put_u8(skb, IFLA_BR_VLAN_STATS_PER_PORT,
1444− br_opt_get(br, IFLA_BR_VLAN_STATS_PER_PORT)))
1444+ br_opt_get(br, BROPT_VLAN_STATS_PER_PORT)))
14451445 return -EMSGSIZE;
14461446 #endif
14471447 #ifdef CONFIG_BRIDGE_IGMP_SNOOPING
net/core/dev.c+15−1
@@ -1184,7 +1184,21 @@ int dev_change_name(struct net_device *dev, const char *newname) · 1183 unmodified lines
11841184 BUG_ON(!dev_net(dev));
11851185
11861186 net = dev_net(dev);
1187− if (dev->flags & IFF_UP)
1187+
1188+ /* Some auto-enslaved devices e.g. failover slaves are
1189+ * special, as userspace might rename the device after
1190+ * the interface had been brought up and running since
1191+ * the point kernel initiated auto-enslavement. Allow
1192+ * live name change even when these slave devices are
1193+ * up and running.
1194+ *
1195+ * Typically, users of these auto-enslaving devices
1196+ * don't actually care about slave name change, as
1197+ * they are supposed to operate on master interface
1198+ * directly.
1199+ */
1200+ if (dev->flags & IFF_UP &&
1201+ likely(!(dev->priv_flags & IFF_LIVE_RENAME_OK)))
11881202 return -EBUSY;
11891203
11901204 write_seqcount_begin(&devnet_rename_seq);
net/core/failover.c+3−3
@@ -80,14 +80,14 @@ static int failover_slave_register(struct net_device *slave_dev) · 79 unmodified lines
8080 goto err_upper_link;
8181 }
8282
83− slave_dev->priv_flags |= IFF_FAILOVER_SLAVE;
83+ slave_dev->priv_flags |= (IFF_FAILOVER_SLAVE | IFF_LIVE_RENAME_OK);
8484
8585 if (fops && fops->slave_register &&
8686 !fops->slave_register(slave_dev, failover_dev))
8787 return NOTIFY_OK;
8888
8989 netdev_upper_dev_unlink(slave_dev, failover_dev);
90− slave_dev->priv_flags &= ~IFF_FAILOVER_SLAVE;
90+ slave_dev->priv_flags &= ~(IFF_FAILOVER_SLAVE | IFF_LIVE_RENAME_OK);
9191 err_upper_link:
9292 netdev_rx_handler_unregister(slave_dev);
9393 done:
@@ -121,7 +121,7 @@ int failover_slave_unregister(struct net_device *slave_dev) · 27 unmodified lines
121121
122122 netdev_rx_handler_unregister(slave_dev);
123123 netdev_upper_dev_unlink(slave_dev, failover_dev);
124− slave_dev->priv_flags &= ~IFF_FAILOVER_SLAVE;
124+ slave_dev->priv_flags &= ~(IFF_FAILOVER_SLAVE | IFF_LIVE_RENAME_OK);
125125
126126 if (fops && fops->slave_unregister &&
127127 !fops->slave_unregister(slave_dev, failover_dev))
net/core/filter.c+2−0
@@ -4383,6 +4383,8 @@ BPF_CALL_3(bpf_bind, struct bpf_sock_addr_kern *, ctx, struct sockaddr *, addr, · 4382 unmodified lines
43834383 * Only binding to IP is supported.
43844384 */
43854385 err = -EINVAL;
4386+ if (addr_len < offsetofend(struct sockaddr, sa_family))
4387+ return err;
43864388 if (addr->sa_family == AF_INET) {
43874389 if (addr_len < sizeof(struct sockaddr_in))
43884390 return err;
net/core/net-sysfs.c+5−9
@@ -1747,20 +1747,16 @@ int netdev_register_kobject(struct net_device *ndev) · 1746 unmodified lines
17471747
17481748 error = device_add(dev);
17491749 if (error)
1750− goto error_put_device;
1750+ return error;
17511751
17521752 error = register_queue_kobjects(ndev);
1753− if (error)
1754− goto error_device_del;
1753+ if (error) {
1754+ device_del(dev);
1755+ return error;
1756+ }
17551757
17561758 pm_runtime_set_memalloc_noio(dev, true);
17571759
1758− return 0;
1759−
1760−error_device_del:
1761− device_del(dev);
1762−error_put_device:
1763− put_device(dev);
17641760 return error;
17651761 }
17661762
net/core/ptp_classifier.c+4−3
@@ -185,9 +185,10 @@ void __init ptp_classifier_init(void) · 184 unmodified lines
185185 { 0x16, 0, 0, 0x00000000 },
186186 { 0x06, 0, 0, 0x00000000 },
187187 };
188− struct sock_fprog_kern ptp_prog = {
189− .len = ARRAY_SIZE(ptp_filter), .filter = ptp_filter,
190− };
188+ struct sock_fprog_kern ptp_prog;
189+
190+ ptp_prog.len = ARRAY_SIZE(ptp_filter);
191+ ptp_prog.filter = ptp_filter;
191192
192193 BUG_ON(bpf_prog_create(&ptp_insns, &ptp_prog));
193194 }
net/core/rtnetlink.c+1−1
@@ -4948,7 +4948,7 @@ static int rtnl_valid_stats_req(const struct nlmsghdr *nlh, bool strict_check, · 4947 unmodified lines
49484948 {
49494949 struct if_stats_msg *ifsm;
49504950
4951− if (nlh->nlmsg_len < sizeof(*ifsm)) {
4951+ if (nlh->nlmsg_len < nlmsg_msg_size(sizeof(*ifsm))) {
49524952 NL_SET_ERR_MSG(extack, "Invalid header for stats dump");
49534953 return -EINVAL;
49544954 }
net/core/skbuff.c+9−1
@@ -5083,7 +5083,8 @@ EXPORT_SYMBOL_GPL(skb_gso_validate_mac_len); · 5082 unmodified lines
50835083
50845084 static struct sk_buff *skb_reorder_vlan_header(struct sk_buff *skb)
50855085 {
5086− int mac_len;
5086+ int mac_len, meta_len;
5087+ void *meta;
50875088
50885089 if (skb_cow(skb, skb_headroom(skb)) < 0) {
50895090 kfree_skb(skb);
@@ -5095,6 +5096,13 @@ static struct sk_buff *skb_reorder_vlan_header(struct sk_buff *skb) · 5 unmodified lines
50955096 memmove(skb_mac_header(skb) + VLAN_HLEN, skb_mac_header(skb),
50965097 mac_len - VLAN_HLEN - ETH_TLEN);
50975098 }
5099+
5100+ meta_len = skb_metadata_len(skb);
5101+ if (meta_len) {
5102+ meta = skb_metadata_end(skb) - meta_len;
5103+ memmove(meta + VLAN_HLEN, meta, meta_len);
5104+ }
5105+
50985106 skb->mac_header += VLAN_HLEN;
50995107 return skb;
51005108 }
net/core/sock.c+2−2
@@ -348,7 +348,7 @@ static int sock_get_timeout(long timeo, void *optval, bool old_timeval) · 347 unmodified lines
348348 tv.tv_usec = ((timeo % HZ) * USEC_PER_SEC) / HZ;
349349 }
350350
351− if (in_compat_syscall() && !COMPAT_USE_64BIT_TIME) {
351+ if (old_timeval && in_compat_syscall() && !COMPAT_USE_64BIT_TIME) {
352352 struct old_timeval32 tv32 = { tv.tv_sec, tv.tv_usec };
353353 *(struct old_timeval32 *)optval = tv32;
354354 return sizeof(tv32);
@@ -372,7 +372,7 @@ static int sock_set_timeout(long *timeo_p, char __user *optval, int optlen, bool · 17 unmodified lines
372372 {
373373 struct __kernel_sock_timeval tv;
374374
375− if (in_compat_syscall() && !COMPAT_USE_64BIT_TIME) {
375+ if (old_timeval && in_compat_syscall() && !COMPAT_USE_64BIT_TIME) {
376376 struct old_timeval32 tv32;
377377
378378 if (optlen < sizeof(tv32))
net/ipv4/fou.c+3−1
@@ -121,6 +121,7 @@ static int gue_udp_recv(struct sock *sk, struct sk_buff *skb) · 120 unmodified lines
121121 struct guehdr *guehdr;
122122 void *data;
123123 u16 doffset = 0;
124+ u8 proto_ctype;
124125
125126 if (!fou)
126127 return 1;
@@ -212,13 +213,14 @@ static int gue_udp_recv(struct sock *sk, struct sk_buff *skb) · 85 unmodified lines
212213 if (unlikely(guehdr->control))
213214 return gue_control_message(skb, guehdr);
214215
216+ proto_ctype = guehdr->proto_ctype;
215217 __skb_pull(skb, sizeof(struct udphdr) + hdrlen);
216218 skb_reset_transport_header(skb);
217219
218220 if (iptunnel_pull_offloads(skb))
219221 goto drop;
220222
221− return -guehdr->proto_ctype;
223+ return -proto_ctype;
222224
223225 drop:
224226 kfree_skb(skb);
net/ipv4/route.c+15−1
@@ -1185,9 +1185,23 @@ static struct dst_entry *ipv4_dst_check(struct dst_entry *dst, u32 cookie) · 1184 unmodified lines
11851185
11861186 static void ipv4_link_failure(struct sk_buff *skb)
11871187 {
1188+ struct ip_options opt;
11881189 struct rtable *rt;
1190+ int res;
11891191
1190− icmp_send(skb, ICMP_DEST_UNREACH, ICMP_HOST_UNREACH, 0);
1192+ /* Recompile ip options since IPCB may not be valid anymore.
1193+ */
1194+ memset(&opt, 0, sizeof(opt));
1195+ opt.optlen = ip_hdr(skb)->ihl*4 - sizeof(struct iphdr);
1196+
1197+ rcu_read_lock();
1198+ res = __ip_options_compile(dev_net(skb->dev), &opt, skb, NULL);
1199+ rcu_read_unlock();
1200+
1201+ if (res)
1202+ return;
1203+
1204+ __icmp_send(skb, ICMP_DEST_UNREACH, ICMP_HOST_UNREACH, 0, &opt);
11911205
11921206 rt = skb_rtable(skb);
11931207 if (rt)
net/ipv4/tcp_dctcp.c+17−28
@@ -49,9 +49,8 @@ · 48 unmodified lines
4949 #define DCTCP_MAX_ALPHA 1024U
5050
5151 struct dctcp {
52− u32 acked_bytes_ecn;
53− u32 acked_bytes_total;
54− u32 prior_snd_una;
52+ u32 old_delivered;
53+ u32 old_delivered_ce;
5554 u32 prior_rcv_nxt;
5655 u32 dctcp_alpha;
5756 u32 next_seq;
@@ -73,8 +72,8 @@ static void dctcp_reset(const struct tcp_sock *tp, struct dctcp *ca) · 15 unmodified lines
7372 {
7473 ca->next_seq = tp->snd_nxt;
7574
76− ca->acked_bytes_ecn = 0;
77− ca->acked_bytes_total = 0;
75+ ca->old_delivered = tp->delivered;
76+ ca->old_delivered_ce = tp->delivered_ce;
7877 }
7978
8079 static void dctcp_init(struct sock *sk)
@@ -86,7 +85,6 @@ static void dctcp_init(struct sock *sk) · 5 unmodified lines
8685 sk->sk_state == TCP_CLOSE)) {
8786 struct dctcp *ca = inet_csk_ca(sk);
8887
89− ca->prior_snd_una = tp->snd_una;
9088 ca->prior_rcv_nxt = tp->rcv_nxt;
9189
9290 ca->dctcp_alpha = min(dctcp_alpha_on_init, DCTCP_MAX_ALPHA);
@@ -118,37 +116,25 @@ static void dctcp_update_alpha(struct sock *sk, u32 flags) · 25 unmodified lines
118116 {
119117 const struct tcp_sock *tp = tcp_sk(sk);
120118 struct dctcp *ca = inet_csk_ca(sk);
121− u32 acked_bytes = tp->snd_una - ca->prior_snd_una;
122−
123− /* If ack did not advance snd_una, count dupack as MSS size.
124− * If ack did update window, do not count it at all.
125− */
126− if (acked_bytes == 0 && !(flags & CA_ACK_WIN_UPDATE))
127− acked_bytes = inet_csk(sk)->icsk_ack.rcv_mss;
128− if (acked_bytes) {
129− ca->acked_bytes_total += acked_bytes;
130− ca->prior_snd_una = tp->snd_una;
131−
132− if (flags & CA_ACK_ECE)
133− ca->acked_bytes_ecn += acked_bytes;
134− }
135119
136120 /* Expired RTT */
137121 if (!before(tp->snd_una, ca->next_seq)) {
138− u64 bytes_ecn = ca->acked_bytes_ecn;
122+ u32 delivered_ce = tp->delivered_ce - ca->old_delivered_ce;
139123 u32 alpha = ca->dctcp_alpha;
140124
141125 /* alpha = (1 - g) * alpha + g * F */
142126
143127 alpha -= min_not_zero(alpha, alpha >> dctcp_shift_g);
144− if (bytes_ecn) {
128+ if (delivered_ce) {
129+ u32 delivered = tp->delivered - ca->old_delivered;
130+
145131 /* If dctcp_shift_g == 1, a 32bit value would overflow
146− * after 8 Mbytes.
132+ * after 8 M packets.
147133 */
148− bytes_ecn <<= (10 - dctcp_shift_g);
149− do_div(bytes_ecn, max(1U, ca->acked_bytes_total));
134+ delivered_ce <<= (10 - dctcp_shift_g);
135+ delivered_ce /= max(1U, delivered);
150136
151− alpha = min(alpha + (u32)bytes_ecn, DCTCP_MAX_ALPHA);
137+ alpha = min(alpha + delivered_ce, DCTCP_MAX_ALPHA);
152138 }
153139 /* dctcp_alpha can be read from dctcp_get_info() without
154140 * synchro, so we ask compiler to not use dctcp_alpha
@@ -200,6 +186,7 @@ static size_t dctcp_get_info(struct sock *sk, u32 ext, int *attr, · 45 unmodified lines
200186 union tcp_cc_info *info)
201187 {
202188 const struct dctcp *ca = inet_csk_ca(sk);
189+ const struct tcp_sock *tp = tcp_sk(sk);
203190
204191 /* Fill it also in case of VEGASINFO due to req struct limits.
205192 * We can still correctly retrieve it later.
@@ -211,8 +198,10 @@ static size_t dctcp_get_info(struct sock *sk, u32 ext, int *attr, · 5 unmodified lines
211198 info->dctcp.dctcp_enabled = 1;
212199 info->dctcp.dctcp_ce_state = (u16) ca->ce_state;
213200 info->dctcp.dctcp_alpha = ca->dctcp_alpha;
214− info->dctcp.dctcp_ab_ecn = ca->acked_bytes_ecn;
215− info->dctcp.dctcp_ab_tot = ca->acked_bytes_total;
201+ info->dctcp.dctcp_ab_ecn = tp->mss_cache *
202+ (tp->delivered_ce - ca->old_delivered_ce);
203+ info->dctcp.dctcp_ab_tot = tp->mss_cache *
204+ (tp->delivered - ca->old_delivered);
216205 }
217206
218207 *attr = INET_DIAG_DCTCPINFO;
net/ipv4/tcp_input.c+5−5
@@ -402,11 +402,12 @@ static int __tcp_grow_window(const struct sock *sk, const struct sk_buff *skb) · 401 unmodified lines
402402 static void tcp_grow_window(struct sock *sk, const struct sk_buff *skb)
403403 {
404404 struct tcp_sock *tp = tcp_sk(sk);
405+ int room;
406+
407+ room = min_t(int, tp->window_clamp, tcp_space(sk)) - tp->rcv_ssthresh;
405408
406409 /* Check #1 */
407− if (tp->rcv_ssthresh < tp->window_clamp &&
408− (int)tp->rcv_ssthresh < tcp_space(sk) &&
409− !tcp_under_memory_pressure(sk)) {
410+ if (room > 0 && !tcp_under_memory_pressure(sk)) {
410411 int incr;
411412
412413 /* Check #2. Increase window, if skb with such overhead
@@ -419,8 +420,7 @@ static void tcp_grow_window(struct sock *sk, const struct sk_buff *skb) · 6 unmodified lines
419420
420421 if (incr) {
421422 incr = max_t(int, incr, 2 * skb->len);
422− tp->rcv_ssthresh = min(tp->rcv_ssthresh + incr,
423− tp->window_clamp);
423+ tp->rcv_ssthresh += min(room, incr);
424424 inet_csk(sk)->icsk_ack.quick |= 1;
425425 }
426426 }
net/ipv6/route.c+4−0
@@ -2330,6 +2330,10 @@ static void __ip6_rt_update_pmtu(struct dst_entry *dst, const struct sock *sk, · 2329 unmodified lines
23302330
23312331 rcu_read_lock();
23322332 from = rcu_dereference(rt6->from);
2333+ if (!from) {
2334+ rcu_read_unlock();
2335+ return;
2336+ }
23332337 nrt6 = ip6_rt_cache_alloc(from, daddr, saddr);
23342338 if (nrt6) {
23352339 rt6_do_update_pmtu(nrt6, mtu);
net/ipv6/udp.c+2−0
@@ -1047,6 +1047,8 @@ static void udp_v6_flush_pending_frames(struct sock *sk) · 1046 unmodified lines
10471047 static int udpv6_pre_connect(struct sock *sk, struct sockaddr *uaddr,
10481048 int addr_len)
10491049 {
1050+ if (addr_len < offsetofend(struct sockaddr, sa_family))
1051+ return -EINVAL;
10501052 /* The following checks are replicated from __ip6_datagram_connect()
10511053 * and intended to prevent BPF program called below from accessing
10521054 * bytes that are out of the bound specified by user in addr_len.
net/llc/af_llc.c+1−2
@@ -320,14 +320,13 @@ static int llc_ui_bind(struct socket *sock, struct sockaddr *uaddr, int addrlen) · 319 unmodified lines
320320 struct llc_sap *sap;
321321 int rc = -EINVAL;
322322
323− dprintk("%s: binding %02X\n", __func__, addr->sllc_sap);
324−
325323 lock_sock(sk);
326324 if (unlikely(!sock_flag(sk, SOCK_ZAPPED) || addrlen != sizeof(*addr)))
327325 goto out;
328326 rc = -EAFNOSUPPORT;
329327 if (unlikely(addr->sllc_family != AF_LLC))
330328 goto out;
329+ dprintk("%s: binding %02X\n", __func__, addr->sllc_sap);
331330 rc = -ENODEV;
332331 rcu_read_lock();
333332 if (sk->sk_bound_dev_if) {
net/mac80211/driver-ops.h+3−0
@@ -1195,6 +1195,9 @@ static inline void drv_wake_tx_queue(struct ieee80211_local *local, · 1194 unmodified lines
11951195 {
11961196 struct ieee80211_sub_if_data *sdata = vif_to_sdata(txq->txq.vif);
11971197
1198+ if (local->in_reconfig)
1199+ return;
1200+
11981201 if (!check_sdata_in_driver(sdata))
11991202 return;
12001203
net/mac80211/key.c+4−5
@@ -167,8 +167,10 @@ static int ieee80211_key_enable_hw_accel(struct ieee80211_key *key) · 166 unmodified lines
167167 * The driver doesn't know anything about VLAN interfaces.
168168 * Hence, don't send GTKs for VLAN interfaces to the driver.
169169 */
170− if (!(key->conf.flags & IEEE80211_KEY_FLAG_PAIRWISE))
170+ if (!(key->conf.flags & IEEE80211_KEY_FLAG_PAIRWISE)) {
171+ ret = 1;
171172 goto out_unsupported;
173+ }
172174 }
173175
174176 ret = drv_set_key(key->local, SET_KEY, sdata,
@@ -213,11 +215,8 @@ static int ieee80211_key_enable_hw_accel(struct ieee80211_key *key) · 38 unmodified lines
213215 /* all of these we can do in software - if driver can */
214216 if (ret == 1)
215217 return 0;
216− if (ieee80211_hw_check(&key->local->hw, SW_CRYPTO_CONTROL)) {
217− if (sdata->vif.type == NL80211_IFTYPE_AP_VLAN)
218− return 0;
218+ if (ieee80211_hw_check(&key->local->hw, SW_CRYPTO_CONTROL))
219219 return -EINVAL;
220− }
221220 return 0;
222221 default:
223222 return -EINVAL;
net/mac80211/mesh_pathtbl.c+1−1
@@ -23,7 +23,7 @@ static void mesh_path_free_rcu(struct mesh_table *tbl, struct mesh_path *mpath); · 22 unmodified lines
2323 static u32 mesh_table_hash(const void *addr, u32 len, u32 seed)
2424 {
2525 /* Use last four bytes of hw addr as hash index */
26− return jhash_1word(*(u32 *)(addr+2), seed);
26+ return jhash_1word(__get_unaligned_cpu32((u8 *)addr + 2), seed);
2727 }
2828
2929 static const struct rhashtable_params mesh_rht_params = {
net/mac80211/rx.c+9−1
@@ -1568,7 +1568,15 @@ static void sta_ps_start(struct sta_info *sta) · 1567 unmodified lines
15681568 return;
15691569
15701570 for (tid = 0; tid < IEEE80211_NUM_TIDS; tid++) {
1571− if (txq_has_queue(sta->sta.txq[tid]))
1571+ struct ieee80211_txq *txq = sta->sta.txq[tid];
1572+ struct txq_info *txqi = to_txq_info(txq);
1573+
1574+ spin_lock(&local->active_txq_lock[txq->ac]);
1575+ if (!list_empty(&txqi->schedule_order))
1576+ list_del_init(&txqi->schedule_order);
1577+ spin_unlock(&local->active_txq_lock[txq->ac]);
1578+
1579+ if (txq_has_queue(txq))
15721580 set_bit(tid, &sta->txq_buffered_tids);
15731581 else
15741582 clear_bit(tid, &sta->txq_buffered_tids);
net/mac80211/trace_msg.h+6−1
@@ -1,4 +1,9 @@
11 /* SPDX-License-Identifier: GPL-2.0 */
2+/*
3+ * Portions of this file
4+ * Copyright (C) 2019 Intel Corporation
5+ */
6+
27 #ifdef CONFIG_MAC80211_MESSAGE_TRACING
38
49 #if !defined(__MAC80211_MSG_DRIVER_TRACE) || defined(TRACE_HEADER_MULTI_READ)
@@ -11,7 +16,7 @@ · 6 unmodified lines
1116 #undef TRACE_SYSTEM
1217 #define TRACE_SYSTEM mac80211_msg
1318
14−#define MAX_MSG_LEN 100
19+#define MAX_MSG_LEN 120
1520
1621 DECLARE_EVENT_CLASS(mac80211_msg_event,
1722 TP_PROTO(struct va_format *vaf),
net/mac80211/tx.c+23−30
@@ -3221,6 +3221,7 @@ static bool ieee80211_amsdu_aggregate(struct ieee80211_sub_if_data *sdata, · 3220 unmodified lines
32213221 u8 max_subframes = sta->sta.max_amsdu_subframes;
32223222 int max_frags = local->hw.max_tx_fragments;
32233223 int max_amsdu_len = sta->sta.max_amsdu_len;
3224+ int orig_truesize;
32243225 __be16 len;
32253226 void *data;
32263227 bool ret = false;
@@ -3261,6 +3262,7 @@ static bool ieee80211_amsdu_aggregate(struct ieee80211_sub_if_data *sdata, · 34 unmodified lines
32613262 if (!head || skb_is_gso(head))
32623263 goto out;
32633264
3265+ orig_truesize = head->truesize;
32643266 orig_len = head->len;
32653267
32663268 if (skb->len + head->len > max_amsdu_len)
@@ -3318,6 +3320,7 @@ static bool ieee80211_amsdu_aggregate(struct ieee80211_sub_if_data *sdata, · 51 unmodified lines
33183320 *frag_tail = skb;
33193321
33203322 out_recalc:
3323+ fq->memory_usage += head->truesize - orig_truesize;
33213324 if (head->len != orig_len) {
33223325 flow->backlog += head->len - orig_len;
33233326 tin->backlog_bytes += head->len - orig_len;
@@ -3646,16 +3649,17 @@ EXPORT_SYMBOL(ieee80211_tx_dequeue); · 322 unmodified lines
36463649 struct ieee80211_txq *ieee80211_next_txq(struct ieee80211_hw *hw, u8 ac)
36473650 {
36483651 struct ieee80211_local *local = hw_to_local(hw);
3652+ struct ieee80211_txq *ret = NULL;
36493653 struct txq_info *txqi = NULL;
36503654
3651− lockdep_assert_held(&local->active_txq_lock[ac]);
3655+ spin_lock_bh(&local->active_txq_lock[ac]);
36523656
36533657 begin:
36543658 txqi = list_first_entry_or_null(&local->active_txqs[ac],
36553659 struct txq_info,
36563660 schedule_order);
36573661 if (!txqi)
3658− return NULL;
3662+ goto out;
36593663
36603664 if (txqi->txq.sta) {
36613665 struct sta_info *sta = container_of(txqi->txq.sta,
@@ -3672,24 +3676,30 @@ struct ieee80211_txq *ieee80211_next_txq(struct ieee80211_hw *hw, u8 ac) · 10 unmodified lines
36723676
36733677
36743678 if (txqi->schedule_round == local->schedule_round[ac])
3675− return NULL;
3679+ goto out;
36763680
36773681 list_del_init(&txqi->schedule_order);
36783682 txqi->schedule_round = local->schedule_round[ac];
3679− return &txqi->txq;
3683+ ret = &txqi->txq;
3684+
3685+out:
3686+ spin_unlock_bh(&local->active_txq_lock[ac]);
3687+ return ret;
36803688 }
36813689 EXPORT_SYMBOL(ieee80211_next_txq);
36823690
3683−void ieee80211_return_txq(struct ieee80211_hw *hw,
3684− struct ieee80211_txq *txq)
3691+void __ieee80211_schedule_txq(struct ieee80211_hw *hw,
3692+ struct ieee80211_txq *txq,
3693+ bool force)
36853694 {
36863695 struct ieee80211_local *local = hw_to_local(hw);
36873696 struct txq_info *txqi = to_txq_info(txq);
36883697
3689− lockdep_assert_held(&local->active_txq_lock[txq->ac]);
3698+ spin_lock_bh(&local->active_txq_lock[txq->ac]);
36903699
36913700 if (list_empty(&txqi->schedule_order) &&
3692− (!skb_queue_empty(&txqi->frags) || txqi->tin.backlog_packets)) {
3701+ (force || !skb_queue_empty(&txqi->frags) ||
3702+ txqi->tin.backlog_packets)) {
36933703 /* If airtime accounting is active, always enqueue STAs at the
36943704 * head of the list to ensure that they only get moved to the
36953705 * back by the airtime DRR scheduler once they have a negative
@@ -3706,20 +3716,10 @@ void ieee80211_return_txq(struct ieee80211_hw *hw, · 10 unmodified lines
37063716 list_add_tail(&txqi->schedule_order,
37073717 &local->active_txqs[txq->ac]);
37083718 }
3709−}
3710−EXPORT_SYMBOL(ieee80211_return_txq);
37113719
3712−void ieee80211_schedule_txq(struct ieee80211_hw *hw,
3713− struct ieee80211_txq *txq)
3714− __acquires(txq_lock) __releases(txq_lock)
3715−{
3716− struct ieee80211_local *local = hw_to_local(hw);
3717−
3718− spin_lock_bh(&local->active_txq_lock[txq->ac]);
3719− ieee80211_return_txq(hw, txq);
37203720 spin_unlock_bh(&local->active_txq_lock[txq->ac]);
37213721 }
3722−EXPORT_SYMBOL(ieee80211_schedule_txq);
3722+EXPORT_SYMBOL(__ieee80211_schedule_txq);
37233723
37243724 bool ieee80211_txq_may_transmit(struct ieee80211_hw *hw,
37253725 struct ieee80211_txq *txq)
@@ -3729,7 +3729,7 @@ bool ieee80211_txq_may_transmit(struct ieee80211_hw *hw, · 3 unmodified lines
37293729 struct sta_info *sta;
37303730 u8 ac = txq->ac;
37313731
3732− lockdep_assert_held(&local->active_txq_lock[ac]);
3732+ spin_lock_bh(&local->active_txq_lock[ac]);
37333733
37343734 if (!txqi->txq.sta)
37353735 goto out;
@@ -3759,34 +3759,27 @@ bool ieee80211_txq_may_transmit(struct ieee80211_hw *hw, · 23 unmodified lines
37593759
37603760 sta->airtime[ac].deficit += sta->airtime_weight;
37613761 list_move_tail(&txqi->schedule_order, &local->active_txqs[ac]);
3762+ spin_unlock_bh(&local->active_txq_lock[ac]);
37623763
37633764 return false;
37643765 out:
37653766 if (!list_empty(&txqi->schedule_order))
37663767 list_del_init(&txqi->schedule_order);
3768+ spin_unlock_bh(&local->active_txq_lock[ac]);
37673769
37683770 return true;
37693771 }
37703772 EXPORT_SYMBOL(ieee80211_txq_may_transmit);
37713773
37723774 void ieee80211_txq_schedule_start(struct ieee80211_hw *hw, u8 ac)
3773− __acquires(txq_lock)
37743775 {
37753776 struct ieee80211_local *local = hw_to_local(hw);
37763777
37773778 spin_lock_bh(&local->active_txq_lock[ac]);
37783779 local->schedule_round[ac]++;
3779−}
3780−EXPORT_SYMBOL(ieee80211_txq_schedule_start);
3781−
3782−void ieee80211_txq_schedule_end(struct ieee80211_hw *hw, u8 ac)
3783− __releases(txq_lock)
3784−{
3785− struct ieee80211_local *local = hw_to_local(hw);
3786−
37873780 spin_unlock_bh(&local->active_txq_lock[ac]);
37883781 }
3789−EXPORT_SYMBOL(ieee80211_txq_schedule_end);
3782+EXPORT_SYMBOL(ieee80211_txq_schedule_start);
37903783
37913784 void __ieee80211_subif_start_xmit(struct sk_buff *skb,
37923785 struct net_device *dev,
net/netlink/af_netlink.c+2−1
@@ -988,7 +988,7 @@ static int netlink_bind(struct socket *sock, struct sockaddr *addr, · 987 unmodified lines
988988 struct netlink_sock *nlk = nlk_sk(sk);
989989 struct sockaddr_nl *nladdr = (struct sockaddr_nl *)addr;
990990 int err = 0;
991− unsigned long groups = nladdr->nl_groups;
991+ unsigned long groups;
992992 bool bound;
993993
994994 if (addr_len < sizeof(struct sockaddr_nl))
@@ -996,6 +996,7 @@ static int netlink_bind(struct socket *sock, struct sockaddr *addr, · 1 unmodified line
996996
997997 if (nladdr->nl_family != AF_NETLINK)
998998 return -EINVAL;
999+ groups = nladdr->nl_groups;
9991000
10001001 /* Only superuser is allowed to listen multicasts */
10011002 if (groups) {
net/netrom/af_netrom.c+54−22
@@ -1392,18 +1392,22 @@ static int __init nr_proto_init(void) · 1391 unmodified lines
13921392 int i;
13931393 int rc = proto_register(&nr_proto, 0);
13941394
1395− if (rc != 0)
1396− goto out;
1395+ if (rc)
1396+ return rc;
13971397
13981398 if (nr_ndevs > 0x7fffffff/sizeof(struct net_device *)) {
1399− printk(KERN_ERR "NET/ROM: nr_proto_init - nr_ndevs parameter to large\n");
1400− return -1;
1399+ pr_err("NET/ROM: %s - nr_ndevs parameter too large\n",
1400+ __func__);
1401+ rc = -EINVAL;
1402+ goto unregister_proto;
14011403 }
14021404
14031405 dev_nr = kcalloc(nr_ndevs, sizeof(struct net_device *), GFP_KERNEL);
1404− if (dev_nr == NULL) {
1405− printk(KERN_ERR "NET/ROM: nr_proto_init - unable to allocate device array\n");
1406− return -1;
1406+ if (!dev_nr) {
1407+ pr_err("NET/ROM: %s - unable to allocate device array\n",
1408+ __func__);
1409+ rc = -ENOMEM;
1410+ goto unregister_proto;
14071411 }
14081412
14091413 for (i = 0; i < nr_ndevs; i++) {
@@ -1413,13 +1417,13 @@ static int __init nr_proto_init(void) · 3 unmodified lines
14131417 sprintf(name, "nr%d", i);
14141418 dev = alloc_netdev(0, name, NET_NAME_UNKNOWN, nr_setup);
14151419 if (!dev) {
1416− printk(KERN_ERR "NET/ROM: nr_proto_init - unable to allocate device structure\n");
1420+ rc = -ENOMEM;
14171421 goto fail;
14181422 }
14191423
14201424 dev->base_addr = i;
1421− if (register_netdev(dev)) {
1422− printk(KERN_ERR "NET/ROM: nr_proto_init - unable to register network device\n");
1425+ rc = register_netdev(dev);
1426+ if (rc) {
14231427 free_netdev(dev);
14241428 goto fail;
14251429 }
@@ -1427,36 +1431,64 @@ static int __init nr_proto_init(void) · 1 unmodified line
14271431 dev_nr[i] = dev;
14281432 }
14291433
1430− if (sock_register(&nr_family_ops)) {
1431− printk(KERN_ERR "NET/ROM: nr_proto_init - unable to register socket family\n");
1434+ rc = sock_register(&nr_family_ops);
1435+ if (rc)
14321436 goto fail;
1433− }
14341437
1435− register_netdevice_notifier(&nr_dev_notifier);
1438+ rc = register_netdevice_notifier(&nr_dev_notifier);
1439+ if (rc)
1440+ goto out_sock;
14361441
14371442 ax25_register_pid(&nr_pid);
14381443 ax25_linkfail_register(&nr_linkfail_notifier);
14391444
14401445 #ifdef CONFIG_SYSCTL
1441− nr_register_sysctl();
1446+ rc = nr_register_sysctl();
1447+ if (rc)
1448+ goto out_sysctl;
14421449 #endif
14431450
14441451 nr_loopback_init();
14451452
1446− proc_create_seq("nr", 0444, init_net.proc_net, &nr_info_seqops);
1447− proc_create_seq("nr_neigh", 0444, init_net.proc_net, &nr_neigh_seqops);
1448− proc_create_seq("nr_nodes", 0444, init_net.proc_net, &nr_node_seqops);
1449−out:
1450− return rc;
1453+ rc = -ENOMEM;
1454+ if (!proc_create_seq("nr", 0444, init_net.proc_net, &nr_info_seqops))
1455+ goto proc_remove1;
1456+ if (!proc_create_seq("nr_neigh", 0444, init_net.proc_net,
1457+ &nr_neigh_seqops))
1458+ goto proc_remove2;
1459+ if (!proc_create_seq("nr_nodes", 0444, init_net.proc_net,
1460+ &nr_node_seqops))
1461+ goto proc_remove3;
1462+
1463+ return 0;
1464+
1465+proc_remove3:
1466+ remove_proc_entry("nr_neigh", init_net.proc_net);
1467+proc_remove2:
1468+ remove_proc_entry("nr", init_net.proc_net);
1469+proc_remove1:
1470+
1471+ nr_loopback_clear();
1472+ nr_rt_free();
1473+
1474+#ifdef CONFIG_SYSCTL
1475+ nr_unregister_sysctl();
1476+out_sysctl:
1477+#endif
1478+ ax25_linkfail_release(&nr_linkfail_notifier);
1479+ ax25_protocol_release(AX25_P_NETROM);
1480+ unregister_netdevice_notifier(&nr_dev_notifier);
1481+out_sock:
1482+ sock_unregister(PF_NETROM);
14511483 fail:
14521484 while (--i >= 0) {
14531485 unregister_netdev(dev_nr[i]);
14541486 free_netdev(dev_nr[i]);
14551487 }
14561488 kfree(dev_nr);
1489+unregister_proto:
14571490 proto_unregister(&nr_proto);
1458− rc = -1;
1459− goto out;
1491+ return rc;
14601492 }
14611493
14621494 module_init(nr_proto_init);
net/netrom/nr_loopback.c+1−1
@@ -70,7 +70,7 @@ static void nr_loopback_timer(struct timer_list *unused) · 69 unmodified lines
7070 }
7171 }
7272
73−void __exit nr_loopback_clear(void)
73+void nr_loopback_clear(void)
7474 {
7575 del_timer_sync(&loopback_timer);
7676 skb_queue_purge(&loopback_queue);
net/netrom/nr_route.c+1−1
@@ -953,7 +953,7 @@ const struct seq_operations nr_neigh_seqops = { · 952 unmodified lines
953953 /*
954954 * Free all memory associated with the nodes and routes lists.
955955 */
956−void __exit nr_rt_free(void)
956+void nr_rt_free(void)
957957 {
958958 struct nr_neigh *s = NULL;
959959 struct nr_node *t = NULL;
net/netrom/sysctl_net_netrom.c+4−1
@@ -146,9 +146,12 @@ static struct ctl_table nr_table[] = { · 145 unmodified lines
146146 { }
147147 };
148148
149−void __init nr_register_sysctl(void)
149+int __init nr_register_sysctl(void)
150150 {
151151 nr_table_header = register_net_sysctl(&init_net, "net/netrom", nr_table);
152+ if (!nr_table_header)
153+ return -ENOMEM;
154+ return 0;
152155 }
153156
154157 void nr_unregister_sysctl(void)
net/rds/af_rds.c+3−0
@@ -543,6 +543,9 @@ static int rds_connect(struct socket *sock, struct sockaddr *uaddr, · 542 unmodified lines
543543 struct rds_sock *rs = rds_sk_to_rs(sk);
544544 int ret = 0;
545545
546+ if (addr_len < offsetofend(struct sockaddr, sa_family))
547+ return -EINVAL;
548+
546549 lock_sock(sk);
547550
548551 switch (uaddr->sa_family) {
net/rds/bind.c+2−0
@@ -173,6 +173,8 @@ int rds_bind(struct socket *sock, struct sockaddr *uaddr, int addr_len) · 172 unmodified lines
173173 /* We allow an RDS socket to be bound to either IPv4 or IPv6
174174 * address.
175175 */
176+ if (addr_len < offsetofend(struct sockaddr, sa_family))
177+ return -EINVAL;
176178 if (uaddr->sa_family == AF_INET) {
177179 struct sockaddr_in *sin = (struct sockaddr_in *)uaddr;
178180
net/rxrpc/af_rxrpc.c+11−6
@@ -135,7 +135,7 @@ static int rxrpc_bind(struct socket *sock, struct sockaddr *saddr, int len) · 134 unmodified lines
135135 struct sockaddr_rxrpc *srx = (struct sockaddr_rxrpc *)saddr;
136136 struct rxrpc_local *local;
137137 struct rxrpc_sock *rx = rxrpc_sk(sock->sk);
138− u16 service_id = srx->srx_service;
138+ u16 service_id;
139139 int ret;
140140
141141 _enter("%p,%p,%d", rx, saddr, len);
@@ -143,6 +143,7 @@ static int rxrpc_bind(struct socket *sock, struct sockaddr *saddr, int len) · 1 unmodified line
143143 ret = rxrpc_validate_address(rx, srx, len);
144144 if (ret < 0)
145145 goto error;
146+ service_id = srx->srx_service;
146147
147148 lock_sock(&rx->sk);
148149
@@ -370,18 +371,22 @@ EXPORT_SYMBOL(rxrpc_kernel_end_call); · 221 unmodified lines
370371 * rxrpc_kernel_check_life - Check to see whether a call is still alive
371372 * @sock: The socket the call is on
372373 * @call: The call to check
374+ * @_life: Where to store the life value
373375 *
374376 * Allow a kernel service to find out whether a call is still alive - ie. we're
375− * getting ACKs from the server. Returns a number representing the life state
376− * which can be compared to that returned by a previous call.
377+ * getting ACKs from the server. Passes back in *_life a number representing
378+ * the life state which can be compared to that returned by a previous call and
379+ * return true if the call is still alive.
377380 *
378381 * If the life state stalls, rxrpc_kernel_probe_life() should be called and
379382 * then 2RTT waited.
380383 */
381−u32 rxrpc_kernel_check_life(const struct socket *sock,
382− const struct rxrpc_call *call)
384+bool rxrpc_kernel_check_life(const struct socket *sock,
385+ const struct rxrpc_call *call,
386+ u32 *_life)
383387 {
384− return call->acks_latest;
388+ *_life = call->acks_latest;
389+ return call->state != RXRPC_CALL_COMPLETE;
385390 }
386391 EXPORT_SYMBOL(rxrpc_kernel_check_life);
387392
net/rxrpc/ar-internal.h+1−0
@@ -654,6 +654,7 @@ struct rxrpc_call { · 653 unmodified lines
654654 u8 ackr_reason; /* reason to ACK */
655655 u16 ackr_skew; /* skew on packet being ACK'd */
656656 rxrpc_serial_t ackr_serial; /* serial of packet being ACK'd */
657+ rxrpc_serial_t ackr_first_seq; /* first sequence number received */
657658 rxrpc_seq_t ackr_prev_seq; /* previous sequence number received */
658659 rxrpc_seq_t ackr_consumed; /* Highest packet shown consumed */
659660 rxrpc_seq_t ackr_seen; /* Highest packet shown seen */
net/rxrpc/conn_event.c+7−4
@@ -153,7 +153,8 @@ static void rxrpc_conn_retransmit_call(struct rxrpc_connection *conn, · 152 unmodified lines
153153 * pass a connection-level abort onto all calls on that connection
154154 */
155155 static void rxrpc_abort_calls(struct rxrpc_connection *conn,
156− enum rxrpc_call_completion compl)
156+ enum rxrpc_call_completion compl,
157+ rxrpc_serial_t serial)
157158 {
158159 struct rxrpc_call *call;
159160 int i;
@@ -173,6 +174,9 @@ static void rxrpc_abort_calls(struct rxrpc_connection *conn, · 13 unmodified lines
173174 call->call_id, 0,
174175 conn->abort_code,
175176 conn->error);
177+ else
178+ trace_rxrpc_rx_abort(call, serial,
179+ conn->abort_code);
176180 if (rxrpc_set_call_completion(call, compl,
177181 conn->abort_code,
178182 conn->error))
@@ -213,8 +217,6 @@ static int rxrpc_abort_connection(struct rxrpc_connection *conn, · 34 unmodified lines
213217 conn->state = RXRPC_CONN_LOCALLY_ABORTED;
214218 spin_unlock_bh(&conn->state_lock);
215219
216− rxrpc_abort_calls(conn, RXRPC_CALL_LOCALLY_ABORTED);
217−
218220 msg.msg_name = &conn->params.peer->srx.transport;
219221 msg.msg_namelen = conn->params.peer->srx.transport_len;
220222 msg.msg_control = NULL;
@@ -242,6 +244,7 @@ static int rxrpc_abort_connection(struct rxrpc_connection *conn, · 21 unmodified lines
242244 len = iov[0].iov_len + iov[1].iov_len;
243245
244246 serial = atomic_inc_return(&conn->serial);
247+ rxrpc_abort_calls(conn, RXRPC_CALL_LOCALLY_ABORTED, serial);
245248 whdr.serial = htonl(serial);
246249 _proto("Tx CONN ABORT %%%u { %d }", serial, conn->abort_code);
247250
@@ -321,7 +324,7 @@ static int rxrpc_process_event(struct rxrpc_connection *conn, · 73 unmodified lines
321324 conn->error = -ECONNABORTED;
322325 conn->abort_code = abort_code;
323326 conn->state = RXRPC_CONN_REMOTELY_ABORTED;
324− rxrpc_abort_calls(conn, RXRPC_CALL_REMOTELY_ABORTED);
327+ rxrpc_abort_calls(conn, RXRPC_CALL_REMOTELY_ABORTED, sp->hdr.serial);
325328 return -ECONNABORTED;
326329
327330 case RXRPC_PACKET_TYPE_CHALLENGE:
net/rxrpc/input.c+12−6
@@ -837,7 +837,7 @@ static void rxrpc_input_ack(struct rxrpc_call *call, struct sk_buff *skb, · 836 unmodified lines
837837 u8 acks[RXRPC_MAXACKS];
838838 } buf;
839839 rxrpc_serial_t acked_serial;
840− rxrpc_seq_t first_soft_ack, hard_ack;
840+ rxrpc_seq_t first_soft_ack, hard_ack, prev_pkt;
841841 int nr_acks, offset, ioffset;
842842
843843 _enter("");
@@ -851,13 +851,14 @@ static void rxrpc_input_ack(struct rxrpc_call *call, struct sk_buff *skb, · 7 unmodified lines
851851
852852 acked_serial = ntohl(buf.ack.serial);
853853 first_soft_ack = ntohl(buf.ack.firstPacket);
854+ prev_pkt = ntohl(buf.ack.previousPacket);
854855 hard_ack = first_soft_ack - 1;
855856 nr_acks = buf.ack.nAcks;
856857 summary.ack_reason = (buf.ack.reason < RXRPC_ACK__INVALID ?
857858 buf.ack.reason : RXRPC_ACK__INVALID);
858859
859860 trace_rxrpc_rx_ack(call, sp->hdr.serial, acked_serial,
860− first_soft_ack, ntohl(buf.ack.previousPacket),
861+ first_soft_ack, prev_pkt,
861862 summary.ack_reason, nr_acks);
862863
863864 if (buf.ack.reason == RXRPC_ACK_PING_RESPONSE)
@@ -878,8 +879,9 @@ static void rxrpc_input_ack(struct rxrpc_call *call, struct sk_buff *skb, · 14 unmodified lines
878879 rxrpc_propose_ack_respond_to_ack);
879880 }
880881
881− /* Discard any out-of-order or duplicate ACKs. */
882− if (before_eq(sp->hdr.serial, call->acks_latest))
882+ /* Discard any out-of-order or duplicate ACKs (outside lock). */
883+ if (before(first_soft_ack, call->ackr_first_seq) ||
884+ before(prev_pkt, call->ackr_prev_seq))
883885 return;
884886
885887 buf.info.rxMTU = 0;
@@ -890,12 +892,16 @@ static void rxrpc_input_ack(struct rxrpc_call *call, struct sk_buff *skb, · 4 unmodified lines
890892
891893 spin_lock(&call->input_lock);
892894
893− /* Discard any out-of-order or duplicate ACKs. */
894− if (before_eq(sp->hdr.serial, call->acks_latest))
895+ /* Discard any out-of-order or duplicate ACKs (inside lock). */
896+ if (before(first_soft_ack, call->ackr_first_seq) ||
897+ before(prev_pkt, call->ackr_prev_seq))
895898 goto out;
896899 call->acks_latest_ts = skb->tstamp;
897900 call->acks_latest = sp->hdr.serial;
898901
902+ call->ackr_first_seq = first_soft_ack;
903+ call->ackr_prev_seq = prev_pkt;
904+
899905 /* Parse rwind and mtu sizes if provided. */
900906 if (buf.info.rxMTU)
901907 rxrpc_input_ackinfo(call, skb, &buf.info);
net/rxrpc/peer_event.c+5−0
@@ -157,6 +157,11 @@ void rxrpc_error_report(struct sock *sk) · 156 unmodified lines
157157
158158 _enter("%p{%d}", sk, local->debug_id);
159159
160+ /* Clear the outstanding error value on the socket so that it doesn't
161+ * cause kernel_sendmsg() to return it later.
162+ */
163+ sock_error(sk);
164+
160165 skb = sock_dequeue_err_skb(sk);
161166 if (!skb) {
162167 _leave("UDP socket errqueue empty");
net/rxrpc/sendmsg.c+12−9
@@ -152,12 +152,13 @@ static void rxrpc_notify_end_tx(struct rxrpc_sock *rx, struct rxrpc_call *call, · 151 unmodified lines
152152 }
153153
154154 /*
155− * Queue a DATA packet for transmission, set the resend timeout and send the
156− * packet immediately
155+ * Queue a DATA packet for transmission, set the resend timeout and send
156+ * the packet immediately. Returns the error from rxrpc_send_data_packet()
157+ * in case the caller wants to do something with it.
157158 */
158−static void rxrpc_queue_packet(struct rxrpc_sock *rx, struct rxrpc_call *call,
159− struct sk_buff *skb, bool last,
160− rxrpc_notify_end_tx_t notify_end_tx)
159+static int rxrpc_queue_packet(struct rxrpc_sock *rx, struct rxrpc_call *call,
160+ struct sk_buff *skb, bool last,
161+ rxrpc_notify_end_tx_t notify_end_tx)
161162 {
162163 struct rxrpc_skb_priv *sp = rxrpc_skb(skb);
163164 unsigned long now;
@@ -250,7 +251,8 @@ static void rxrpc_queue_packet(struct rxrpc_sock *rx, struct rxrpc_call *call, · 86 unmodified lines
250251
251252 out:
252253 rxrpc_free_skb(skb, rxrpc_skb_tx_freed);
253− _leave("");
254+ _leave(" = %d", ret);
255+ return ret;
254256 }
255257
256258 /*
@@ -423,9 +425,10 @@ static int rxrpc_send_data(struct rxrpc_sock *rx, · 166 unmodified lines
423425 if (ret < 0)
424426 goto out;
425427
426− rxrpc_queue_packet(rx, call, skb,
427− !msg_data_left(msg) && !more,
428− notify_end_tx);
428+ ret = rxrpc_queue_packet(rx, call, skb,
429+ !msg_data_left(msg) && !more,
430+ notify_end_tx);
431+ /* Should check for failure here */
429432 skb = NULL;
430433 }
431434 } while (msg_data_left(msg) > 0);
net/sctp/socket.c+2−1
@@ -4847,7 +4847,8 @@ static int sctp_connect(struct sock *sk, struct sockaddr *addr, · 4846 unmodified lines
48474847 }
48484848
48494849 /* Validate addr_len before calling common connect/connectx routine. */
4850− af = sctp_get_af_specific(addr->sa_family);
4850+ af = addr_len < offsetofend(struct sockaddr, sa_family) ? NULL :
4851+ sctp_get_af_specific(addr->sa_family);
48514852 if (!af || addr_len < af->sockaddr_len) {
48524853 err = -EINVAL;
48534854 } else {
net/smc/af_smc.c+39−19
@@ -167,10 +167,9 @@ static int smc_release(struct socket *sock) · 166 unmodified lines
167167
168168 if (sk->sk_state == SMC_CLOSED) {
169169 if (smc->clcsock) {
170− mutex_lock(&smc->clcsock_release_lock);
171− sock_release(smc->clcsock);
172− smc->clcsock = NULL;
173− mutex_unlock(&smc->clcsock_release_lock);
170+ release_sock(sk);
171+ smc_clcsock_release(smc);
172+ lock_sock(sk);
174173 }
175174 if (!smc->use_fallback)
176175 smc_conn_free(&smc->conn);
@@ -446,10 +445,19 @@ static void smc_link_save_peer_info(struct smc_link *link, · 269 unmodified lines
446445 link->peer_mtu = clc->qp_mtu;
447446 }
448447
448+static void smc_switch_to_fallback(struct smc_sock *smc)
449+{
450+ smc->use_fallback = true;
451+ if (smc->sk.sk_socket && smc->sk.sk_socket->file) {
452+ smc->clcsock->file = smc->sk.sk_socket->file;
453+ smc->clcsock->file->private_data = smc->clcsock;
454+ }
455+}
456+
449457 /* fall back during connect */
450458 static int smc_connect_fallback(struct smc_sock *smc, int reason_code)
451459 {
452− smc->use_fallback = true;
460+ smc_switch_to_fallback(smc);
453461 smc->fallback_rsn = reason_code;
454462 smc_copy_sock_settings_to_clc(smc);
455463 if (smc->sk.sk_state == SMC_INIT)
@@ -775,10 +783,14 @@ static void smc_connect_work(struct work_struct *work) · 319 unmodified lines
775783 smc->sk.sk_err = -rc;
776784
777785 out:
778− if (smc->sk.sk_err)
779− smc->sk.sk_state_change(&smc->sk);
780− else
781− smc->sk.sk_write_space(&smc->sk);
786+ if (!sock_flag(&smc->sk, SOCK_DEAD)) {
787+ if (smc->sk.sk_err) {
788+ smc->sk.sk_state_change(&smc->sk);
789+ } else { /* allow polling before and after fallback decision */
790+ smc->clcsock->sk->sk_write_space(smc->clcsock->sk);
791+ smc->sk.sk_write_space(&smc->sk);
792+ }
793+ }
782794 kfree(smc->connect_info);
783795 smc->connect_info = NULL;
784796 release_sock(&smc->sk);
@@ -872,11 +884,11 @@ static int smc_clcsock_accept(struct smc_sock *lsmc, struct smc_sock **new_smc) · 87 unmodified lines
872884 if (rc < 0)
873885 lsk->sk_err = -rc;
874886 if (rc < 0 || lsk->sk_state == SMC_CLOSED) {
887+ new_sk->sk_prot->unhash(new_sk);
875888 if (new_clcsock)
876889 sock_release(new_clcsock);
877890 new_sk->sk_state = SMC_CLOSED;
878891 sock_set_flag(new_sk, SOCK_DEAD);
879− new_sk->sk_prot->unhash(new_sk);
880892 sock_put(new_sk); /* final */
881893 *new_smc = NULL;
882894 goto out;
@@ -927,16 +939,21 @@ struct sock *smc_accept_dequeue(struct sock *parent, · 44 unmodified lines
927939
928940 smc_accept_unlink(new_sk);
929941 if (new_sk->sk_state == SMC_CLOSED) {
942+ new_sk->sk_prot->unhash(new_sk);
930943 if (isk->clcsock) {
931944 sock_release(isk->clcsock);
932945 isk->clcsock = NULL;
933946 }
934− new_sk->sk_prot->unhash(new_sk);
935947 sock_put(new_sk); /* final */
936948 continue;
937949 }
938− if (new_sock)
950+ if (new_sock) {
939951 sock_graft(new_sk, new_sock);
952+ if (isk->use_fallback) {
953+ smc_sk(new_sk)->clcsock->file = new_sock->file;
954+ isk->clcsock->file->private_data = isk->clcsock;
955+ }
956+ }
940957 return new_sk;
941958 }
942959 return NULL;
@@ -956,6 +973,7 @@ void smc_close_non_accepted(struct sock *sk) · 13 unmodified lines
956973 sock_set_flag(sk, SOCK_DEAD);
957974 sk->sk_shutdown |= SHUTDOWN_MASK;
958975 }
976+ sk->sk_prot->unhash(sk);
959977 if (smc->clcsock) {
960978 struct socket *tcp;
961979
@@ -971,7 +989,6 @@ void smc_close_non_accepted(struct sock *sk) · 9 unmodified lines
971989 smc_conn_free(&smc->conn);
972990 }
973991 release_sock(sk);
974− sk->sk_prot->unhash(sk);
975992 sock_put(sk); /* final sock_put */
976993 }
977994
@@ -1037,13 +1054,13 @@ static void smc_listen_out(struct smc_sock *new_smc) · 59 unmodified lines
10371054 struct smc_sock *lsmc = new_smc->listen_smc;
10381055 struct sock *newsmcsk = &new_smc->sk;
10391056
1040− lock_sock_nested(&lsmc->sk, SINGLE_DEPTH_NESTING);
10411057 if (lsmc->sk.sk_state == SMC_LISTEN) {
1058+ lock_sock_nested(&lsmc->sk, SINGLE_DEPTH_NESTING);
10421059 smc_accept_enqueue(&lsmc->sk, newsmcsk);
1060+ release_sock(&lsmc->sk);
10431061 } else { /* no longer listening */
10441062 smc_close_non_accepted(newsmcsk);
10451063 }
1046− release_sock(&lsmc->sk);
10471064
10481065 /* Wake up accept */
10491066 lsmc->sk.sk_data_ready(&lsmc->sk);
@@ -1087,7 +1104,7 @@ static void smc_listen_decline(struct smc_sock *new_smc, int reason_code, · 37 unmodified lines
10871104 return;
10881105 }
10891106 smc_conn_free(&new_smc->conn);
1090− new_smc->use_fallback = true;
1107+ smc_switch_to_fallback(new_smc);
10911108 new_smc->fallback_rsn = reason_code;
10921109 if (reason_code && reason_code != SMC_CLC_DECL_PEERDECL) {
10931110 if (smc_clc_send_decline(new_smc, reason_code) < 0) {
@@ -1237,6 +1254,9 @@ static void smc_listen_work(struct work_struct *work) · 143 unmodified lines
12371254 int rc = 0;
12381255 u8 ibport;
12391256
1257+ if (new_smc->listen_smc->sk.sk_state != SMC_LISTEN)
1258+ return smc_listen_out_err(new_smc);
1259+
12401260 if (new_smc->use_fallback) {
12411261 smc_listen_out_connected(new_smc);
12421262 return;
@@ -1244,7 +1264,7 @@ static void smc_listen_work(struct work_struct *work) · 1 unmodified line
12441264
12451265 /* check if peer is smc capable */
12461266 if (!tcp_sk(newclcsock->sk)->syn_smc) {
1247− new_smc->use_fallback = true;
1267+ smc_switch_to_fallback(new_smc);
12481268 new_smc->fallback_rsn = SMC_CLC_DECL_PEERNOSMC;
12491269 smc_listen_out_connected(new_smc);
12501270 return;
@@ -1501,7 +1521,7 @@ static int smc_sendmsg(struct socket *sock, struct msghdr *msg, size_t len) · 250 unmodified lines
15011521
15021522 if (msg->msg_flags & MSG_FASTOPEN) {
15031523 if (sk->sk_state == SMC_INIT) {
1504− smc->use_fallback = true;
1524+ smc_switch_to_fallback(smc);
15051525 smc->fallback_rsn = SMC_CLC_DECL_OPTUNSUPP;
15061526 } else {
15071527 rc = -EINVAL;
@@ -1703,7 +1723,7 @@ static int smc_setsockopt(struct socket *sock, int level, int optname, · 195 unmodified lines
17031723 case TCP_FASTOPEN_NO_COOKIE:
17041724 /* option not supported by SMC */
17051725 if (sk->sk_state == SMC_INIT) {
1706− smc->use_fallback = true;
1726+ smc_switch_to_fallback(smc);
17071727 smc->fallback_rsn = SMC_CLC_DECL_OPTUNSUPP;
17081728 } else {
17091729 if (!smc->use_fallback)
net/smc/smc_close.c+21−4
@@ -21,6 +21,22 @@ · 20 unmodified lines
2121
2222 #define SMC_CLOSE_WAIT_LISTEN_CLCSOCK_TIME (5 * HZ)
2323
24+/* release the clcsock that is assigned to the smc_sock */
25+void smc_clcsock_release(struct smc_sock *smc)
26+{
27+ struct socket *tcp;
28+
29+ if (smc->listen_smc && current_work() != &smc->smc_listen_work)
30+ cancel_work_sync(&smc->smc_listen_work);
31+ mutex_lock(&smc->clcsock_release_lock);
32+ if (smc->clcsock) {
33+ tcp = smc->clcsock;
34+ smc->clcsock = NULL;
35+ sock_release(tcp);
36+ }
37+ mutex_unlock(&smc->clcsock_release_lock);
38+}
39+
2440 static void smc_close_cleanup_listen(struct sock *parent)
2541 {
2642 struct sock *sk;
@@ -321,6 +337,7 @@ static void smc_close_passive_work(struct work_struct *work) · 294 unmodified lines
321337 close_work);
322338 struct smc_sock *smc = container_of(conn, struct smc_sock, conn);
323339 struct smc_cdc_conn_state_flags *rxflags;
340+ bool release_clcsock = false;
324341 struct sock *sk = &smc->sk;
325342 int old_state;
326343
@@ -400,13 +417,13 @@ wakeup: · 73 unmodified lines
400417 if ((sk->sk_state == SMC_CLOSED) &&
401418 (sock_flag(sk, SOCK_DEAD) || !sk->sk_socket)) {
402419 smc_conn_free(conn);
403− if (smc->clcsock) {
404− sock_release(smc->clcsock);
405− smc->clcsock = NULL;
406− }
420+ if (smc->clcsock)
421+ release_clcsock = true;
407422 }
408423 }
409424 release_sock(sk);
425+ if (release_clcsock)
426+ smc_clcsock_release(smc);
410427 sock_put(sk); /* sock_hold done by schedulers of close_work */
411428 }
412429
net/smc/smc_close.h+1−0
@@ -23,5 +23,6 @@ void smc_close_wake_tx_prepared(struct smc_sock *smc); · 22 unmodified lines
2323 int smc_close_active(struct smc_sock *smc);
2424 int smc_close_shutdown_write(struct smc_sock *smc);
2525 void smc_close_init(struct smc_sock *smc);
26+void smc_clcsock_release(struct smc_sock *smc);
2627
2728 #endif /* SMC_CLOSE_H */
net/smc/smc_ism.c+5−0
@@ -289,6 +289,11 @@ struct smcd_dev *smcd_alloc_dev(struct device *parent, const char *name, · 288 unmodified lines
289289 INIT_LIST_HEAD(&smcd->vlan);
290290 smcd->event_wq = alloc_ordered_workqueue("ism_evt_wq-%s)",
291291 WQ_MEM_RECLAIM, name);
292+ if (!smcd->event_wq) {
293+ kfree(smcd->conn);
294+ kfree(smcd);
295+ return NULL;
296+ }
292297 return smcd;
293298 }
294299 EXPORT_SYMBOL_GPL(smcd_alloc_dev);
net/smc/smc_pnet.c+2−1
@@ -603,7 +603,8 @@ static int smc_pnet_flush(struct sk_buff *skb, struct genl_info *info) · 602 unmodified lines
603603 {
604604 struct net *net = genl_info_net(info);
605605
606− return smc_pnet_remove_by_pnetid(net, NULL);
606+ smc_pnet_remove_by_pnetid(net, NULL);
607+ return 0;
607608 }
608609
609610 /* SMC_PNETID generic netlink operation definition */
net/strparser/strparser.c+5−7
@@ -140,13 +140,11 @@ static int __strp_recv(read_descriptor_t *desc, struct sk_buff *orig_skb, · 139 unmodified lines
140140 /* We are going to append to the frags_list of head.
141141 * Need to unshare the frag_list.
142142 */
143− if (skb_has_frag_list(head)) {
144− err = skb_unclone(head, GFP_ATOMIC);
145− if (err) {
146− STRP_STATS_INCR(strp->stats.mem_fail);
147− desc->error = err;
148− return 0;
149− }
143+ err = skb_unclone(head, GFP_ATOMIC);
144+ if (err) {
145+ STRP_STATS_INCR(strp->stats.mem_fail);
146+ desc->error = err;
147+ return 0;
150148 }
151149
152150 if (unlikely(skb_shinfo(head)->frag_list)) {
net/tipc/link.c+2−0
@@ -869,6 +869,8 @@ void tipc_link_reset(struct tipc_link *l) · 868 unmodified lines
869869 __skb_queue_head_init(&list);
870870
871871 l->in_session = false;
872+ /* Force re-synch of peer session number before establishing */
873+ l->peer_session--;
872874 l->session++;
873875 l->mtu = l->advertised_mtu;
874876
net/tipc/name_table.c+2−1
@@ -909,7 +909,8 @@ static int tipc_nl_service_list(struct net *net, struct tipc_nl_msg *msg, · 908 unmodified lines
909909 for (; i < TIPC_NAMETBL_SIZE; i++) {
910910 head = &tn->nametbl->services[i];
911911
912− if (*last_type) {
912+ if (*last_type ||
913+ (!i && *last_key && (*last_lower == *last_key))) {
913914 service = tipc_service_find(net, *last_type);
914915 if (!service)
915916 return -EPIPE;
net/tipc/sysctl.c+6−2
@@ -38,6 +38,8 @@ · 37 unmodified lines
3838
3939 #include <linux/sysctl.h>
4040
41+static int zero;
42+static int one = 1;
4143 static struct ctl_table_header *tipc_ctl_hdr;
4244
4345 static struct ctl_table tipc_table[] = {
@@ -46,14 +48,16 @@ static struct ctl_table tipc_table[] = { · 2 unmodified lines
4648 .data = &sysctl_tipc_rmem,
4749 .maxlen = sizeof(sysctl_tipc_rmem),
4850 .mode = 0644,
49− .proc_handler = proc_dointvec,
51+ .proc_handler = proc_dointvec_minmax,
52+ .extra1 = &one,
5053 },
5154 {
5255 .procname = "named_timeout",
5356 .data = &sysctl_tipc_named_timeout,
5457 .maxlen = sizeof(sysctl_tipc_named_timeout),
5558 .mode = 0644,
56− .proc_handler = proc_dointvec,
59+ .proc_handler = proc_dointvec_minmax,
60+ .extra1 = &zero,
5761 },
5862 {
5963 .procname = "sk_filter",
net/tls/tls_device.c+11−1
@@ -52,8 +52,11 @@ static DEFINE_SPINLOCK(tls_device_lock); · 51 unmodified lines
5252
5353 static void tls_device_free_ctx(struct tls_context *ctx)
5454 {
55− if (ctx->tx_conf == TLS_HW)
55+ if (ctx->tx_conf == TLS_HW) {
5656 kfree(tls_offload_ctx_tx(ctx));
57+ kfree(ctx->tx.rec_seq);
58+ kfree(ctx->tx.iv);
59+ }
5760
5861 if (ctx->rx_conf == TLS_HW)
5962 kfree(tls_offload_ctx_rx(ctx));
@@ -216,6 +219,13 @@ void tls_device_sk_destruct(struct sock *sk) · 156 unmodified lines
216219 }
217220 EXPORT_SYMBOL(tls_device_sk_destruct);
218221
222+void tls_device_free_resources_tx(struct sock *sk)
223+{
224+ struct tls_context *tls_ctx = tls_get_ctx(sk);
225+
226+ tls_free_partial_record(sk, tls_ctx);
227+}
228+
219229 static void tls_append_frag(struct tls_record_info *record,
220230 struct page_frag *pfrag,
221231 int size)
net/tls/tls_main.c+24−0
@@ -208,6 +208,26 @@ int tls_push_partial_record(struct sock *sk, struct tls_context *ctx, · 207 unmodified lines
208208 return tls_push_sg(sk, ctx, sg, offset, flags);
209209 }
210210
211+bool tls_free_partial_record(struct sock *sk, struct tls_context *ctx)
212+{
213+ struct scatterlist *sg;
214+
215+ sg = ctx->partially_sent_record;
216+ if (!sg)
217+ return false;
218+
219+ while (1) {
220+ put_page(sg_page(sg));
221+ sk_mem_uncharge(sk, sg->length);
222+
223+ if (sg_is_last(sg))
224+ break;
225+ sg++;
226+ }
227+ ctx->partially_sent_record = NULL;
228+ return true;
229+}
230+
211231 static void tls_write_space(struct sock *sk)
212232 {
213233 struct tls_context *ctx = tls_get_ctx(sk);
@@ -267,6 +287,10 @@ static void tls_sk_proto_close(struct sock *sk, long timeout) · 53 unmodified lines
267287 kfree(ctx->tx.rec_seq);
268288 kfree(ctx->tx.iv);
269289 tls_sw_free_resources_tx(sk);
290+#ifdef CONFIG_TLS_DEVICE
291+ } else if (ctx->tx_conf == TLS_HW) {
292+ tls_device_free_resources_tx(sk);
293+#endif
270294 }
271295
272296 if (ctx->rx_conf == TLS_SW) {
net/tls/tls_sw.c+1−14
@@ -2052,20 +2052,7 @@ void tls_sw_free_resources_tx(struct sock *sk) · 2051 unmodified lines
20522052 /* Free up un-sent records in tx_list. First, free
20532053 * the partially sent record if any at head of tx_list.
20542054 */
2055− if (tls_ctx->partially_sent_record) {
2056− struct scatterlist *sg = tls_ctx->partially_sent_record;
2057−
2058− while (1) {
2059− put_page(sg_page(sg));
2060− sk_mem_uncharge(sk, sg->length);
2061−
2062− if (sg_is_last(sg))
2063− break;
2064− sg++;
2065− }
2066−
2067− tls_ctx->partially_sent_record = NULL;
2068−
2055+ if (tls_free_partial_record(sk, tls_ctx)) {
20692056 rec = list_first_entry(&ctx->tx_list,
20702057 struct tls_rec, list);
20712058 list_del(&rec->list);
net/wireless/nl80211.c+12−6
@@ -13650,7 +13650,8 @@ static const struct genl_ops nl80211_ops[] = { · 13649 unmodified lines
1365013650 .policy = nl80211_policy,
1365113651 .flags = GENL_UNS_ADMIN_PERM,
1365213652 .internal_flags = NL80211_FLAG_NEED_NETDEV_UP |
13653− NL80211_FLAG_NEED_RTNL,
13653+ NL80211_FLAG_NEED_RTNL |
13654+ NL80211_FLAG_CLEAR_SKB,
1365413655 },
1365513656 {
1365613657 .cmd = NL80211_CMD_DEAUTHENTICATE,
@@ -13701,7 +13702,8 @@ static const struct genl_ops nl80211_ops[] = { · 44 unmodified lines
1370113702 .policy = nl80211_policy,
1370213703 .flags = GENL_UNS_ADMIN_PERM,
1370313704 .internal_flags = NL80211_FLAG_NEED_NETDEV_UP |
13704− NL80211_FLAG_NEED_RTNL,
13705+ NL80211_FLAG_NEED_RTNL |
13706+ NL80211_FLAG_CLEAR_SKB,
1370513707 },
1370613708 {
1370713709 .cmd = NL80211_CMD_UPDATE_CONNECT_PARAMS,
@@ -13709,7 +13711,8 @@ static const struct genl_ops nl80211_ops[] = { · 1 unmodified line
1370913711 .policy = nl80211_policy,
1371013712 .flags = GENL_ADMIN_PERM,
1371113713 .internal_flags = NL80211_FLAG_NEED_NETDEV_UP |
13712− NL80211_FLAG_NEED_RTNL,
13714+ NL80211_FLAG_NEED_RTNL |
13715+ NL80211_FLAG_CLEAR_SKB,
1371313716 },
1371413717 {
1371513718 .cmd = NL80211_CMD_DISCONNECT,
@@ -13738,7 +13741,8 @@ static const struct genl_ops nl80211_ops[] = { · 22 unmodified lines
1373813741 .policy = nl80211_policy,
1373913742 .flags = GENL_UNS_ADMIN_PERM,
1374013743 .internal_flags = NL80211_FLAG_NEED_NETDEV_UP |
13741− NL80211_FLAG_NEED_RTNL,
13744+ NL80211_FLAG_NEED_RTNL |
13745+ NL80211_FLAG_CLEAR_SKB,
1374213746 },
1374313747 {
1374413748 .cmd = NL80211_CMD_DEL_PMKSA,
@@ -14090,7 +14094,8 @@ static const struct genl_ops nl80211_ops[] = { · 345 unmodified lines
1409014094 .policy = nl80211_policy,
1409114095 .flags = GENL_UNS_ADMIN_PERM,
1409214096 .internal_flags = NL80211_FLAG_NEED_WIPHY |
14093− NL80211_FLAG_NEED_RTNL,
14097+ NL80211_FLAG_NEED_RTNL |
14098+ NL80211_FLAG_CLEAR_SKB,
1409414099 },
1409514100 {
1409614101 .cmd = NL80211_CMD_SET_QOS_MAP,
@@ -14145,7 +14150,8 @@ static const struct genl_ops nl80211_ops[] = { · 48 unmodified lines
1414514150 .doit = nl80211_set_pmk,
1414614151 .policy = nl80211_policy,
1414714152 .internal_flags = NL80211_FLAG_NEED_NETDEV_UP |
14148− NL80211_FLAG_NEED_RTNL,
14153+ NL80211_FLAG_NEED_RTNL |
14154+ NL80211_FLAG_CLEAR_SKB,
1414914155 },
1415014156 {
1415114157 .cmd = NL80211_CMD_DEL_PMK,
net/wireless/reg.c+39−0
@@ -1309,6 +1309,16 @@ reg_intersect_dfs_region(const enum nl80211_dfs_regions dfs_region1, · 1308 unmodified lines
13091309 return dfs_region1;
13101310 }
13111311
1312+static void reg_wmm_rules_intersect(const struct ieee80211_wmm_ac *wmm_ac1,
1313+ const struct ieee80211_wmm_ac *wmm_ac2,
1314+ struct ieee80211_wmm_ac *intersect)
1315+{
1316+ intersect->cw_min = max_t(u16, wmm_ac1->cw_min, wmm_ac2->cw_min);
1317+ intersect->cw_max = max_t(u16, wmm_ac1->cw_max, wmm_ac2->cw_max);
1318+ intersect->cot = min_t(u16, wmm_ac1->cot, wmm_ac2->cot);
1319+ intersect->aifsn = max_t(u8, wmm_ac1->aifsn, wmm_ac2->aifsn);
1320+}
1321+
13121322 /*
13131323 * Helper for regdom_intersect(), this does the real
13141324 * mathematical intersection fun
@@ -1323,6 +1333,8 @@ static int reg_rules_intersect(const struct ieee80211_regdomain *rd1, · 8 unmodified lines
13231333 struct ieee80211_freq_range *freq_range;
13241334 const struct ieee80211_power_rule *power_rule1, *power_rule2;
13251335 struct ieee80211_power_rule *power_rule;
1336+ const struct ieee80211_wmm_rule *wmm_rule1, *wmm_rule2;
1337+ struct ieee80211_wmm_rule *wmm_rule;
13261338 u32 freq_diff, max_bandwidth1, max_bandwidth2;
13271339
13281340 freq_range1 = &rule1->freq_range;
@@ -1333,6 +1345,10 @@ static int reg_rules_intersect(const struct ieee80211_regdomain *rd1, · 4 unmodified lines
13331345 power_rule2 = &rule2->power_rule;
13341346 power_rule = &intersected_rule->power_rule;
13351347
1348+ wmm_rule1 = &rule1->wmm_rule;
1349+ wmm_rule2 = &rule2->wmm_rule;
1350+ wmm_rule = &intersected_rule->wmm_rule;
1351+
13361352 freq_range->start_freq_khz = max(freq_range1->start_freq_khz,
13371353 freq_range2->start_freq_khz);
13381354 freq_range->end_freq_khz = min(freq_range1->end_freq_khz,
@@ -1376,6 +1392,29 @@ static int reg_rules_intersect(const struct ieee80211_regdomain *rd1, · 37 unmodified lines
13761392 intersected_rule->dfs_cac_ms = max(rule1->dfs_cac_ms,
13771393 rule2->dfs_cac_ms);
13781394
1395+ if (rule1->has_wmm && rule2->has_wmm) {
1396+ u8 ac;
1397+
1398+ for (ac = 0; ac < IEEE80211_NUM_ACS; ac++) {
1399+ reg_wmm_rules_intersect(&wmm_rule1->client[ac],
1400+ &wmm_rule2->client[ac],
1401+ &wmm_rule->client[ac]);
1402+ reg_wmm_rules_intersect(&wmm_rule1->ap[ac],
1403+ &wmm_rule2->ap[ac],
1404+ &wmm_rule->ap[ac]);
1405+ }
1406+
1407+ intersected_rule->has_wmm = true;
1408+ } else if (rule1->has_wmm) {
1409+ *wmm_rule = *wmm_rule1;
1410+ intersected_rule->has_wmm = true;
1411+ } else if (rule2->has_wmm) {
1412+ *wmm_rule = *wmm_rule2;
1413+ intersected_rule->has_wmm = true;
1414+ } else {
1415+ intersected_rule->has_wmm = false;
1416+ }
1417+
13791418 if (!is_valid_reg_rule(intersected_rule))
13801419 return -EINVAL;
13811420
net/wireless/scan.c+1−2
@@ -190,10 +190,9 @@ static size_t cfg80211_gen_new_ie(const u8 *ie, size_t ielen, · 189 unmodified lines
190190 /* copy subelement as we need to change its content to
191191 * mark an ie after it is processed.
192192 */
193− sub_copy = kmalloc(subie_len, gfp);
193+ sub_copy = kmemdup(subelement, subie_len, gfp);
194194 if (!sub_copy)
195195 return 0;
196− memcpy(sub_copy, subelement, subie_len);
197196
198197 pos = &new_ie[0];
199198
net/wireless/util.c+4−2
@@ -1220,9 +1220,11 @@ static u32 cfg80211_calculate_bitrate_he(struct rate_info *rate) · 1219 unmodified lines
12201220 else if (rate->bw == RATE_INFO_BW_HE_RU &&
12211221 rate->he_ru_alloc == NL80211_RATE_INFO_HE_RU_ALLOC_26)
12221222 result = rates_26[rate->he_gi];
1223− else if (WARN(1, "invalid HE MCS: bw:%d, ru:%d\n",
1224− rate->bw, rate->he_ru_alloc))
1223+ else {
1224+ WARN(1, "invalid HE MCS: bw:%d, ru:%d\n",
1225+ rate->bw, rate->he_ru_alloc);
12251226 return 0;
1227+ }
12261228
12271229 /* now scale to the appropriate MCS */
12281230 tmp = result;
tools/testing/selftests/drivers/net/mlxsw/rtnetlink.sh+20−0
@@ -11,6 +11,7 @@ lib_dir=$(dirname $0)/../../../net/forwarding · 10 unmodified lines
1111
1212 ALL_TESTS="
1313 rif_set_addr_test
14+ rif_vrf_set_addr_test
1415 rif_inherit_bridge_addr_test
1516 rif_non_inherit_bridge_addr_test
1617 vlan_interface_deletion_test
@@ -98,6 +99,25 @@ rif_set_addr_test() · 81 unmodified lines
9899 ip link set dev $swp1 addr $swp1_mac
99100 }
100101
102+rif_vrf_set_addr_test()
103+{
104+ # Test that it is possible to set an IP address on a VRF upper despite
105+ # its random MAC address.
106+ RET=0
107+
108+ ip link add name vrf-test type vrf table 10
109+ ip link set dev $swp1 master vrf-test
110+
111+ ip -4 address add 192.0.2.1/24 dev vrf-test
112+ check_err $? "failed to set IPv4 address on VRF"
113+ ip -6 address add 2001:db8:1::1/64 dev vrf-test
114+ check_err $? "failed to set IPv6 address on VRF"
115+
116+ log_test "RIF - setting IP address on VRF"
117+
118+ ip link del dev vrf-test
119+}
120+
101121 rif_inherit_bridge_addr_test()
102122 {
103123 RET=0
tools/testing/selftests/net/fib_tests.sh+40−54
@@ -605,6 +605,39 @@ run_cmd() · 604 unmodified lines
605605 return $rc
606606 }
607607
608+check_expected()
609+{
610+ local out="$1"
611+ local expected="$2"
612+ local rc=0
613+
614+ [ "${out}" = "${expected}" ] && return 0
615+
616+ if [ -z "${out}" ]; then
617+ if [ "$VERBOSE" = "1" ]; then
618+ printf "\nNo route entry found\n"
619+ printf "Expected:\n"
620+ printf " ${expected}\n"
621+ fi
622+ return 1
623+ fi
624+
625+ # tricky way to convert output to 1-line without ip's
626+ # messy '\'; this drops all extra white space
627+ out=$(echo ${out})
628+ if [ "${out}" != "${expected}" ]; then
629+ rc=1
630+ if [ "${VERBOSE}" = "1" ]; then
631+ printf " Unexpected route entry. Have:\n"
632+ printf " ${out}\n"
633+ printf " Expected:\n"
634+ printf " ${expected}\n\n"
635+ fi
636+ fi
637+
638+ return $rc
639+}
640+
608641 # add route for a prefix, flushing any existing routes first
609642 # expected to be the first step of a test
610643 add_route6()
@@ -652,31 +685,7 @@ check_route6() · 41 unmodified lines
652685 pfx=$1
653686
654687 out=$($IP -6 ro ls match ${pfx} | sed -e 's/ pref medium//')
655− [ "${out}" = "${expected}" ] && return 0
656−
657− if [ -z "${out}" ]; then
658− if [ "$VERBOSE" = "1" ]; then
659− printf "\nNo route entry found\n"
660− printf "Expected:\n"
661− printf " ${expected}\n"
662− fi
663− return 1
664− fi
665−
666− # tricky way to convert output to 1-line without ip's
667− # messy '\'; this drops all extra white space
668− out=$(echo ${out})
669− if [ "${out}" != "${expected}" ]; then
670− rc=1
671− if [ "${VERBOSE}" = "1" ]; then
672− printf " Unexpected route entry. Have:\n"
673− printf " ${out}\n"
674− printf " Expected:\n"
675− printf " ${expected}\n\n"
676− fi
677− fi
678−
679− return $rc
688+ check_expected "${out}" "${expected}"
680689 }
681690
682691 route_cleanup()
@@ -725,7 +734,7 @@ route_setup() · 42 unmodified lines
725734 ip -netns ns2 addr add 172.16.103.2/24 dev veth4
726735 ip -netns ns2 addr add 172.16.104.1/24 dev dummy1
727736
728− set +ex
737+ set +e
729738 }
730739
731740 # assumption is that basic add of a single path route works
@@ -960,7 +969,8 @@ ipv6_addr_metric_test() · 228 unmodified lines
960969 run_cmd "$IP li set dev dummy2 down"
961970 rc=$?
962971 if [ $rc -eq 0 ]; then
963− check_route6 ""
972+ out=$($IP -6 ro ls match 2001:db8:104::/64)
973+ check_expected "${out}" ""
964974 rc=$?
965975 fi
966976 log_test $rc 0 "Prefix route removed on link down"
@@ -1091,38 +1101,13 @@ check_route() · 124 unmodified lines
10911101 local pfx
10921102 local expected="$1"
10931103 local out
1094− local rc=0
10951104
10961105 set -- $expected
10971106 pfx=$1
10981107 [ "${pfx}" = "unreachable" ] && pfx=$2
10991108
11001109 out=$($IP ro ls match ${pfx})
1101− [ "${out}" = "${expected}" ] && return 0
1102−
1103− if [ -z "${out}" ]; then
1104− if [ "$VERBOSE" = "1" ]; then
1105− printf "\nNo route entry found\n"
1106− printf "Expected:\n"
1107− printf " ${expected}\n"
1108− fi
1109− return 1
1110− fi
1111−
1112− # tricky way to convert output to 1-line without ip's
1113− # messy '\'; this drops all extra white space
1114− out=$(echo ${out})
1115− if [ "${out}" != "${expected}" ]; then
1116− rc=1
1117− if [ "${VERBOSE}" = "1" ]; then
1118− printf " Unexpected route entry. Have:\n"
1119− printf " ${out}\n"
1120− printf " Expected:\n"
1121− printf " ${expected}\n\n"
1122− fi
1123− fi
1124−
1125− return $rc
1110+ check_expected "${out}" "${expected}"
11261111 }
11271112
11281113 # assumption is that basic add of a single path route works
@@ -1387,7 +1372,8 @@ ipv4_addr_metric_test() · 258 unmodified lines
13871372 run_cmd "$IP li set dev dummy2 down"
13881373 rc=$?
13891374 if [ $rc -eq 0 ]; then
1390− check_route ""
1375+ out=$($IP ro ls match 172.16.104.0/24)
1376+ check_expected "${out}" ""
13911377 rc=$?
13921378 fi
13931379 log_test $rc 0 "Prefix route removed on link down"

Keyboard shortcuts

General

Show keyboard shortcuts
?
Focus the current search
/

Issues and pull requests

Open filters
f

Issues, pull requests and Inbox

Select next item
j↓
Select previous item
k↑
Open selected item
Enter
Dismiss filters or clear selection
Esc

Repository code

Go to file
t