linux-merge February 2016

linux-merge@lists.open-mesh.org

2 participants
121 discussions

[linux-next] LinuxNextTracking branch, master, updated. next-20160224

by batman＠open-mesh.org

The following commit has been merged in the master branch: commit b633353115e352d3c31c12d4c61978c810f05ea1 Merge: b1d95ae5c5bd3deba84d00c4f83d7d0836b5936f dea08e604408d0303e2332896c5fdd8c1f7d79a2 Author: David S. Miller <davem(a)davemloft.net> Date: Tue Feb 23 00:09:14 2016 -0500 Merge git://git.kernel.org/pub/scm/linux/kernel/git/davem/net Conflicts: drivers/net/phy/bcm7xxx.c drivers/net/phy/marvell.c drivers/net/vxlan.c All three conflicts were cases of simple overlapping changes. Signed-off-by: David S. Miller <davem(a)davemloft.net> diff --combined MAINTAINERS index 355e1c8,2fa3303..27393cf --- a/MAINTAINERS +++ b/MAINTAINERS @@@ -223,9 -223,7 +223,7 @@@ F: drivers/scsi/aacraid ABI/API L: linux-api(a)vger.kernel.org - F: Documentation/ABI/ F: include/linux/syscalls.h - F: include/uapi/ F: kernel/sys_ni.c ABIT UGURU 1,2 HARDWARE MONITOR DRIVER @@@ -686,13 -684,6 +684,6 @@@ M: Michael Hanselmann <linux-kernel@han S: Supported F: drivers/macintosh/ams/ - AMSO1100 RNIC DRIVER - M: Tom Tucker <tom(a)opengridcomputing.com> - M: Steve Wise <swise(a)opengridcomputing.com> - L: linux-rdma(a)vger.kernel.org - S: Maintained - F: drivers/infiniband/hw/amso1100/ - ANALOG DEVICES INC AD9389B DRIVER M: Hans Verkuil <hans.verkuil(a)cisco.com> L: linux-media(a)vger.kernel.org @@@ -967,6 -958,8 +958,8 @@@ M: Rob Herring <robh(a)kernel.org L: linux-arm-kernel(a)lists.infradead.org (moderated for non-subscribers) S: Maintained F: arch/arm/mach-highbank/ + F: arch/arm/boot/dts/highbank.dts + F: arch/arm/boot/dts/ecx-*.dts* ARM/CAVIUM NETWORKS CNS3XXX MACHINE SUPPORT M: Krzysztof Halasa <khalasa(a)piap.pl> @@@ -1042,6 -1035,7 +1035,7 @@@ M: Barry Song <baohua(a)kernel.org L: linux-arm-kernel(a)lists.infradead.org (moderated for non-subscribers) T: git git://git.kernel.org/pub/scm/linux/kernel/git/baohua/linux.git S: Maintained + F: arch/arm/boot/dts/prima2* F: arch/arm/mach-prima2/ F: drivers/clk/sirf/ F: drivers/clocksource/timer-prima2.c @@@ -1143,6 -1137,10 +1137,10 @@@ W: http://www.hisilicon.co S: Supported T: git git://github.com/hisilicon/linux-hisi.git F: arch/arm/mach-hisi/ + F: arch/arm/boot/dts/hi3* + F: arch/arm/boot/dts/hip* + F: arch/arm/boot/dts/hisi* + F: arch/arm64/boot/dts/hisilicon/ ARM/HP JORNADA 7XX MACHINE SUPPORT M: Kristoffer Ericson <kristoffer.ericson(a)gmail.com> @@@ -1219,6 -1217,7 +1217,7 @@@ M: Santosh Shilimkar <ssantosh(a)kernel.o L: linux-arm-kernel(a)lists.infradead.org (moderated for non-subscribers) S: Maintained F: arch/arm/mach-keystone/ + F: arch/arm/boot/dts/k2* T: git git://git.kernel.org/pub/scm/linux/kernel/git/ssantosh/linux-keystone.git ARM/TEXAS INSTRUMENT KEYSTONE CLOCK FRAMEWORK @@@ -1287,6 -1286,7 +1286,7 @@@ L: linux-arm-kernel(a)lists.infradead.or S: Maintained F: arch/arm/mach-berlin/ F: arch/arm/boot/dts/berlin* + F: arch/arm64/boot/dts/marvell/berlin* ARM/Marvell Dove/MV78xx0/Orion SOC support @@@ -1425,6 -1425,7 +1425,7 @@@ S: Maintaine F: arch/arm/boot/dts/qcom-*.dts F: arch/arm/boot/dts/qcom-*.dtsi F: arch/arm/mach-qcom/ + F: arch/arm64/boot/dts/qcom/* F: drivers/soc/qcom/ F: drivers/tty/serial/msm_serial.h F: drivers/tty/serial/msm_serial.c @@@ -1441,8 -1442,8 +1442,8 @@@ S: Maintaine ARM/RENESAS ARM64 ARCHITECTURE M: Simon Horman <horms(a)verge.net.au> M: Magnus Damm <magnus.damm(a)gmail.com> - L: linux-sh(a)vger.kernel.org - Q: http://patchwork.kernel.org/project/linux-sh/list/ + L: linux-renesas-soc(a)vger.kernel.org + Q: http://patchwork.kernel.org/project/linux-renesas-soc/list/ T: git git://git.kernel.org/pub/scm/linux/kernel/git/horms/renesas.git next S: Supported F: arch/arm64/boot/dts/renesas/ @@@ -1484,6 -1485,8 +1485,8 @@@ L: linux-arm-kernel(a)lists.infradead.or L: linux-samsung-soc(a)vger.kernel.org (moderated for non-subscribers) S: Maintained F: arch/arm/boot/dts/s3c* + F: arch/arm/boot/dts/s5p* + F: arch/arm/boot/dts/samsung* F: arch/arm/boot/dts/exynos* F: arch/arm64/boot/dts/exynos/ F: arch/arm/plat-samsung/ @@@ -1563,6 -1566,7 +1566,7 @@@ S: Maintaine F: arch/arm/mach-socfpga/ F: arch/arm/boot/dts/socfpga* F: arch/arm/configs/socfpga_defconfig + F: arch/arm64/boot/dts/altera/ W: http://www.rocketboards.org T: git git://git.kernel.org/pub/scm/linux/kernel/git/dinguyen/linux.git @@@ -1716,7 -1720,7 +1720,7 @@@ M: Lorenzo Pieralisi <lorenzo.pieralisi L: linux-arm-kernel(a)lists.infradead.org (moderated for non-subscribers) S: Maintained F: arch/arm/boot/dts/vexpress* - F: arch/arm64/boot/dts/arm/vexpress* + F: arch/arm64/boot/dts/arm/ F: arch/arm/mach-vexpress/ F: */*/vexpress* F: */*/*/vexpress* @@@ -2147,7 -2151,7 +2151,7 @@@ M: Marek Lindner <mareklindner@neomailb M: Simon Wunderlich <sw(a)simonwunderlich.de> M: Antonio Quartulli <a(a)unstable.cc> L: b.a.t.m.a.n(a)lists.open-mesh.org -W: http://www.open-mesh.org/ +W: https://www.open-mesh.org/ S: Maintained F: net/batman-adv/ @@@ -2343,6 -2347,7 +2347,7 @@@ F: arch/arm/mach-bcm F: arch/arm/boot/dts/bcm113* F: arch/arm/boot/dts/bcm216* F: arch/arm/boot/dts/bcm281* + F: arch/arm64/boot/dts/broadcom/ F: arch/arm/configs/bcm_defconfig F: drivers/mmc/host/sdhci-bcm-kona.c F: drivers/clocksource/bcm_kona_timer.c @@@ -2357,14 -2362,6 +2362,6 @@@ T: git git://git.kernel.org/pub/scm/lin S: Maintained N: bcm2835 - BROADCOM BCM33XX MIPS ARCHITECTURE - M: Kevin Cernekee <cernekee(a)gmail.com> - L: linux-mips(a)linux-mips.org - S: Maintained - F: arch/mips/bcm3384/* - F: arch/mips/include/asm/mach-bcm3384/* - F: arch/mips/kernel/*bmips* - BROADCOM BCM47XX MIPS ARCHITECTURE M: Hauke Mehrtens <hauke(a)hauke-m.de> M: Rafał Miłecki <zajec5(a)gmail.com> @@@ -2424,7 -2421,6 +2421,7 @@@ F: include/linux/bcm963xx_nvram. F: include/linux/bcm963xx_tag.h BROADCOM TG3 GIGABIT ETHERNET DRIVER +M: Siva Reddy Kallam <siva.kallam(a)broadcom.com> M: Prashant Sreedharan <prashant(a)broadcom.com> M: Michael Chan <mchan(a)broadcom.com> L: netdev(a)vger.kernel.org @@@ -3446,7 -3442,7 +3443,7 @@@ S: Maintaine F: drivers/usb/dwc2/ DESIGNWARE USB3 DRD IP DRIVER - M: Felipe Balbi <balbi(a)ti.com> + M: Felipe Balbi <balbi(a)kernel.org> L: linux-usb(a)vger.kernel.org L: linux-omap(a)vger.kernel.org T: git git://git.kernel.org/pub/scm/linux/kernel/git/balbi/usb.git @@@ -4185,13 -4181,6 +4182,6 @@@ W: http://aeschi.ch.eu.org/efs S: Orphan F: fs/efs/ - EHCA (IBM GX bus InfiniBand adapter) DRIVER - M: Hoang-Nam Nguyen <hnguyen(a)de.ibm.com> - M: Christoph Raisch <raisch(a)de.ibm.com> - L: linux-rdma(a)vger.kernel.org - S: Supported - F: drivers/infiniband/hw/ehca/ - EHEA (IBM pSeries eHEA 10Gb ethernet adapter) DRIVER M: Thadeu Lima de Souza Cascardo <cascardo(a)linux.vnet.ibm.com> L: netdev(a)vger.kernel.org @@@ -5810,12 -5799,6 +5800,6 @@@ M: Juanjo Ciarlante <jjciarla(a)raiz.uncu S: Maintained F: net/ipv4/netfilter/ipt_MASQUERADE.c - IPATH DRIVER - M: Mike Marciniszyn <infinipath(a)intel.com> - L: linux-rdma(a)vger.kernel.org - S: Maintained - F: drivers/staging/rdma/ipath/ - IPMI SUBSYSTEM M: Corey Minyard <minyard(a)acm.org> L: openipmi-developer(a)lists.sourceforge.net (moderated for non-subscribers) @@@ -6145,7 -6128,7 +6129,7 @@@ F: include/uapi/linux/sunrpc KERNEL SELFTEST FRAMEWORK M: Shuah Khan <shuahkh(a)osg.samsung.com> - L: linux-api(a)vger.kernel.org + L: linux-kselftest(a)vger.kernel.org T: git git://git.kernel.org/pub/scm/shuah/linux-kselftest S: Maintained F: tools/testing/selftests @@@ -7371,7 -7354,7 +7355,7 @@@ F: drivers/tty/isicom. F: include/linux/isicom.h MUSB MULTIPOINT HIGH SPEED DUAL-ROLE CONTROLLER - M: Felipe Balbi <balbi(a)ti.com> + M: Felipe Balbi <balbi(a)kernel.org> L: linux-usb(a)vger.kernel.org T: git git://git.kernel.org/pub/scm/linux/kernel/git/balbi/usb.git S: Maintained @@@ -7940,7 -7923,7 +7924,7 @@@ F: drivers/media/platform/omap3isp F: drivers/staging/media/omap4iss/ OMAP USB SUPPORT - M: Felipe Balbi <balbi(a)ti.com> + M: Felipe Balbi <balbi(a)kernel.org> L: linux-usb(a)vger.kernel.org L: linux-omap(a)vger.kernel.org T: git git://git.kernel.org/pub/scm/linux/kernel/git/balbi/usb.git @@@ -8819,6 -8802,7 +8803,7 @@@ L: linux-arm-kernel(a)lists.infradead.or T: git git://github.com/hzhuang1/linux.git T: git git://github.com/rjarzmik/linux.git S: Maintained + F: arch/arm/boot/dts/pxa* F: arch/arm/mach-pxa/ F: drivers/dma/pxa* F: drivers/pcmcia/pxa2xx* @@@ -8848,6 -8832,7 +8833,7 @@@ L: linux-arm-kernel(a)lists.infradead.or T: git git://github.com/hzhuang1/linux.git T: git git://git.linaro.org/people/ycmiao/pxa-linux.git S: Maintained + F: arch/arm/boot/dts/mmp* F: arch/arm/mach-mmp/ PXA MMCI DRIVER @@@ -9794,10 -9779,11 +9780,11 @@@ S: Supporte F: drivers/scsi/be2iscsi/ Emulex 10Gbps NIC BE2, BE3-R, Lancer, Skyhawk-R DRIVER - M: Sathya Perla <sathya.perla(a)avagotech.com> - M: Ajit Khaparde <ajit.khaparde(a)avagotech.com> - M: Padmanabh Ratnakar <padmanabh.ratnakar(a)avagotech.com> - M: Sriharsha Basavapatna <sriharsha.basavapatna(a)avagotech.com> + M: Sathya Perla <sathya.perla(a)broadcom.com> + M: Ajit Khaparde <ajit.khaparde(a)broadcom.com> + M: Padmanabh Ratnakar <padmanabh.ratnakar(a)broadcom.com> + M: Sriharsha Basavapatna <sriharsha.basavapatna(a)broadcom.com> + M: Somnath Kotur <somnath.kotur(a)broadcom.com> L: netdev(a)vger.kernel.org W: http://www.emulex.com S: Supported @@@ -10159,6 -10145,7 +10146,7 @@@ S: Supporte F: drivers/media/pci/solo6x10/ SOFTWARE RAID (Multiple Disks) SUPPORT + M: Shaohua Li <shli(a)kernel.org> L: linux-raid(a)vger.kernel.org T: git git://neil.brown.name/md S: Supported @@@ -10292,6 -10279,7 +10280,7 @@@ L: spear-devel(a)list.st.co L: linux-arm-kernel(a)lists.infradead.org (moderated for non-subscribers) W: http://www.st.com/spear S: Maintained + F: arch/arm/boot/dts/spear* F: arch/arm/mach-spear/ SPEAR CLOCK FRAMEWORK SUPPORT @@@ -11319,7 -11307,7 +11308,7 @@@ F: Documentation/usb/ehci.tx F: drivers/usb/host/ehci* USB GADGET/PERIPHERAL SUBSYSTEM - M: Felipe Balbi <balbi(a)ti.com> + M: Felipe Balbi <balbi(a)kernel.org> L: linux-usb(a)vger.kernel.org W: http://www.linux-usb.org/gadget T: git git://git.kernel.org/pub/scm/linux/kernel/git/balbi/usb.git @@@ -11395,7 -11383,7 +11384,7 @@@ S: Maintaine F: drivers/net/usb/pegasus.* USB PHY LAYER - M: Felipe Balbi <balbi(a)ti.com> + M: Felipe Balbi <balbi(a)kernel.org> L: linux-usb(a)vger.kernel.org T: git git://git.kernel.org/pub/scm/linux/kernel/git/balbi/usb.git S: Maintained @@@ -12025,7 -12013,6 +12014,6 @@@ F: arch/arm64/xen F: arch/arm64/include/asm/xen/ XEN NETWORK BACKEND DRIVER - M: Ian Campbell <ian.campbell(a)citrix.com> M: Wei Liu <wei.liu2(a)citrix.com> L: xen-devel(a)lists.xenproject.org (moderated for non-subscribers) L: netdev(a)vger.kernel.org @@@ -12134,7 -12121,7 +12122,7 @@@ F: drivers/net/hamradio/*scc. F: drivers/net/hamradio/z8530.h ZBUD COMPRESSED PAGE ALLOCATOR - M: Seth Jennings <sjennings(a)variantweb.net> + M: Seth Jennings <sjenning(a)redhat.com> L: linux-mm(a)kvack.org S: Maintained F: mm/zbud.c @@@ -12189,7 -12176,7 +12177,7 @@@ F: include/linux/zsmalloc. F: Documentation/vm/zsmalloc.txt ZSWAP COMPRESSED SWAP CACHING - M: Seth Jennings <sjennings(a)variantweb.net> + M: Seth Jennings <sjenning(a)redhat.com> L: linux-mm(a)kvack.org S: Maintained F: mm/zswap.c diff --combined drivers/net/bonding/bond_main.c index 45bdd87,b7f1a99..a6527d5 --- a/drivers/net/bonding/bond_main.c +++ b/drivers/net/bonding/bond_main.c @@@ -214,6 -214,8 +214,8 @@@ static void bond_uninit(struct net_devi static struct rtnl_link_stats64 *bond_get_stats(struct net_device *bond_dev, struct rtnl_link_stats64 *stats); static void bond_slave_arr_handler(struct work_struct *work); + static bool bond_time_in_interval(struct bonding *bond, unsigned long last_act, + int mod); /*---------------------------- General routines -----------------------------*/ @@@ -618,8 -620,8 +620,8 @@@ static void bond_hw_addr_swap(struct bo static void bond_set_dev_addr(struct net_device *bond_dev, struct net_device *slave_dev) { - netdev_dbg(bond_dev, "bond_dev=%p slave_dev=%p slave_dev->addr_len=%d\n", - bond_dev, slave_dev, slave_dev->addr_len); + netdev_dbg(bond_dev, "bond_dev=%p slave_dev=%p slave_dev->name=%s slave_dev->addr_len=%d\n", + bond_dev, slave_dev, slave_dev->name, slave_dev->addr_len); memcpy(bond_dev->dev_addr, slave_dev->dev_addr, slave_dev->addr_len); bond_dev->addr_assign_type = NET_ADDR_STOLEN; call_netdevice_notifiers(NETDEV_CHANGEADDR, bond_dev); @@@ -928,10 -930,11 +930,10 @@@ void bond_select_active_slave(struct bo if (!rv) return; - if (netif_carrier_ok(bond->dev)) { + if (netif_carrier_ok(bond->dev)) netdev_info(bond->dev, "first active interface up!\n"); - } else { + else netdev_info(bond->dev, "now running without any active interface!\n"); - } } } @@@ -1177,8 -1180,9 +1179,8 @@@ static rx_handler_result_t bond_handle_ } } - if (bond_should_deliver_exact_match(skb, slave, bond)) { + if (bond_should_deliver_exact_match(skb, slave, bond)) return RX_HANDLER_EXACT; - } skb->dev = bond->dev; @@@ -1239,7 -1243,7 +1241,7 @@@ static struct slave *bond_alloc_slave(s { struct slave *slave = NULL; - slave = kzalloc(sizeof(struct slave), GFP_KERNEL); + slave = kzalloc(sizeof(*slave), GFP_KERNEL); if (!slave) return NULL; @@@ -1379,7 -1383,8 +1381,7 @@@ int bond_enslave(struct net_device *bon if (slave_dev->flags & IFF_UP) { netdev_err(bond_dev, "%s is up - this may be due to an out of date ifenslave\n", slave_dev->name); - res = -EPERM; - goto err_undo_flags; + return -EPERM; } /* set bonding device ether type by slave - bonding netdevices are @@@ -1399,7 -1404,8 +1401,7 @@@ res = notifier_to_errno(res); if (res) { netdev_err(bond_dev, "refused to change device type\n"); - res = -EBUSY; - goto err_undo_flags; + return -EBUSY; } /* Flush unicast and multicast addresses */ @@@ -1419,7 -1425,8 +1421,7 @@@ } else if (bond_dev->type != slave_dev->type) { netdev_err(bond_dev, "%s ether type (%d) is different from other slaves (%d), can not enslave it\n", slave_dev->name, slave_dev->type, bond_dev->type); - res = -EINVAL; - goto err_undo_flags; + return -EINVAL; } if (slave_ops->ndo_set_mac_address == NULL) { @@@ -2122,6 -2129,7 +2124,7 @@@ static void bond_miimon_commit(struct b continue; case BOND_LINK_UP: + bond_update_speed_duplex(slave); bond_set_slave_link_state(slave, BOND_LINK_UP, BOND_SLAVE_NOTIFY_NOW); slave->last_link_up = jiffies; @@@ -2454,7 -2462,7 +2457,7 @@@ int bond_arp_rcv(const struct sk_buff * struct slave *slave) { struct arphdr *arp = (struct arphdr *)skb->data; - struct slave *curr_active_slave; + struct slave *curr_active_slave, *curr_arp_slave; unsigned char *arp_ptr; __be32 sip, tip; int alen, is_arp = skb->protocol == __cpu_to_be16(ETH_P_ARP); @@@ -2501,26 -2509,41 +2504,41 @@@ &sip, &tip); curr_active_slave = rcu_dereference(bond->curr_active_slave); + curr_arp_slave = rcu_dereference(bond->current_arp_slave); - /* Backup slaves won't see the ARP reply, but do come through - * here for each ARP probe (so we swap the sip/tip to validate - * the probe). In a "redundant switch, common router" type of - * configuration, the ARP probe will (hopefully) travel from - * the active, through one switch, the router, then the other - * switch before reaching the backup. + /* We 'trust' the received ARP enough to validate it if: * - * We 'trust' the arp requests if there is an active slave and - * it received valid arp reply(s) after it became active. This - * is done to avoid endless looping when we can't reach the + * (a) the slave receiving the ARP is active (which includes the + * current ARP slave, if any), or + * + * (b) the receiving slave isn't active, but there is a currently + * active slave and it received valid arp reply(s) after it became + * the currently active slave, or + * + * (c) there is an ARP slave that sent an ARP during the prior ARP + * interval, and we receive an ARP reply on any slave. We accept + * these because switch FDB update delays may deliver the ARP + * reply to a slave other than the sender of the ARP request. + * + * Note: for (b), backup slaves are receiving the broadcast ARP + * request, not a reply. This request passes from the sending + * slave through the L2 switch(es) to the receiving slave. Since + * this is checking the request, sip/tip are swapped for + * validation. + * + * This is done to avoid endless looping when we can't reach the * arp_ip_target and fool ourselves with our own arp requests. */ - if (bond_is_active_slave(slave)) bond_validate_arp(bond, slave, sip, tip); else if (curr_active_slave && time_after(slave_last_rx(bond, curr_active_slave), curr_active_slave->last_link_up)) bond_validate_arp(bond, slave, tip, sip); + else if (curr_arp_slave && (arp->ar_op == htons(ARPOP_REPLY)) && + bond_time_in_interval(bond, + dev_trans_start(curr_arp_slave->dev), 1)) + bond_validate_arp(bond, slave, sip, tip); out_unlock: if (arp != (struct arphdr *)skb->data) @@@ -3304,7 -3327,6 +3322,7 @@@ static struct rtnl_link_stats64 *bond_g stats->rx_bytes += sstats->rx_bytes - pstats->rx_bytes; stats->rx_errors += sstats->rx_errors - pstats->rx_errors; stats->rx_dropped += sstats->rx_dropped - pstats->rx_dropped; + stats->rx_nohandler += sstats->rx_nohandler - pstats->rx_nohandler; stats->tx_packets += sstats->tx_packets - pstats->tx_packets;; stats->tx_bytes += sstats->tx_bytes - pstats->tx_bytes; diff --combined drivers/net/ethernet/broadcom/bnxt/bnxt.c index 169920a,8ab000d..ff1507f --- a/drivers/net/ethernet/broadcom/bnxt/bnxt.c +++ b/drivers/net/ethernet/broadcom/bnxt/bnxt.c @@@ -69,7 -69,7 +69,7 @@@ MODULE_VERSION(DRV_MODULE_VERSION) #define BNXT_RX_DMA_OFFSET NET_SKB_PAD #define BNXT_RX_COPY_THRESH 256 - #define BNXT_TX_PUSH_THRESH 92 + #define BNXT_TX_PUSH_THRESH 164 enum board_idx { BCM57301, @@@ -223,11 -223,12 +223,12 @@@ static netdev_tx_t bnxt_start_xmit(stru } if (free_size == bp->tx_ring_size && length <= bp->tx_push_thresh) { - struct tx_push_bd *push = txr->tx_push; - struct tx_bd *tx_push = &push->txbd1; - struct tx_bd_ext *tx_push1 = &push->txbd2; - void *pdata = tx_push1 + 1; - int j; + struct tx_push_buffer *tx_push_buf = txr->tx_push; + struct tx_push_bd *tx_push = &tx_push_buf->push_bd; + struct tx_bd_ext *tx_push1 = &tx_push->txbd2; + void *pdata = tx_push_buf->data; + u64 *end; + int j, push_len; /* Set COAL_NOW to be ready quickly for the next push */ tx_push->tx_bd_len_flags_type = @@@ -247,6 -248,9 +248,9 @@@ tx_push1->tx_bd_cfa_meta = cpu_to_le32(vlan_tag_flags); tx_push1->tx_bd_cfa_action = cpu_to_le32(cfa_action); + end = PTR_ALIGN(pdata + length + 1, 8) - 1; + *end = 0; + skb_copy_from_linear_data(skb, pdata, len); pdata += len; for (j = 0; j < last_frag; j++) { @@@ -261,22 -265,29 +265,29 @@@ pdata += skb_frag_size(frag); } - memcpy(txbd, tx_push, sizeof(*txbd)); + txbd->tx_bd_len_flags_type = tx_push->tx_bd_len_flags_type; + txbd->tx_bd_haddr = txr->data_mapping; prod = NEXT_TX(prod); txbd = &txr->tx_desc_ring[TX_RING(prod)][TX_IDX(prod)]; memcpy(txbd, tx_push1, sizeof(*txbd)); prod = NEXT_TX(prod); - push->doorbell = + tx_push->doorbell = cpu_to_le32(DB_KEY_TX_PUSH | DB_LONG_TX_PUSH | prod); txr->tx_prod = prod; netdev_tx_sent_queue(txq, skb->len); - __iowrite64_copy(txr->tx_doorbell, push, - (length + sizeof(*push) + 8) / 8); + push_len = (length + sizeof(*tx_push) + 7) / 8; + if (push_len > 16) { + __iowrite64_copy(txr->tx_doorbell, tx_push_buf, 16); + __iowrite64_copy(txr->tx_doorbell + 4, tx_push_buf + 1, + push_len - 16); + } else { + __iowrite64_copy(txr->tx_doorbell, tx_push_buf, + push_len); + } tx_buf->is_push = 1; - goto tx_done; } @@@ -1753,7 -1764,7 +1764,7 @@@ static int bnxt_alloc_tx_rings(struct b push_size = L1_CACHE_ALIGN(sizeof(struct tx_push_bd) + bp->tx_push_thresh); - if (push_size > 128) { + if (push_size > 256) { push_size = 0; bp->tx_push_thresh = 0; } @@@ -1772,7 -1783,6 +1783,6 @@@ return rc; if (bp->tx_push_size) { - struct tx_bd *txbd; dma_addr_t mapping; /* One pre-allocated DMA buffer to backup @@@ -1786,13 -1796,11 +1796,11 @@@ if (!txr->tx_push) return -ENOMEM; - txbd = &txr->tx_push->txbd1; - mapping = txr->tx_push_mapping + sizeof(struct tx_push_bd); - txbd->tx_bd_haddr = cpu_to_le64(mapping); + txr->data_mapping = cpu_to_le64(mapping); - memset(txbd + 1, 0, sizeof(struct tx_bd_ext)); + memset(txr->tx_push, 0, sizeof(struct tx_push_bd)); } ring->queue_id = bp->q_info[j].queue_id; if (i % bp->tx_nr_rings_per_tc == (bp->tx_nr_rings_per_tc - 1)) @@@ -4546,20 -4554,18 +4554,18 @@@ static int bnxt_update_phy_setting(stru if (!(link_info->autoneg & BNXT_AUTONEG_FLOW_CTRL) && link_info->force_pause_setting != link_info->req_flow_ctrl) update_pause = true; - if (link_info->req_duplex != link_info->duplex_setting) - update_link = true; if (!(link_info->autoneg & BNXT_AUTONEG_SPEED)) { if (BNXT_AUTO_MODE(link_info->auto_mode)) update_link = true; if (link_info->req_link_speed != link_info->force_link_speed) update_link = true; + if (link_info->req_duplex != link_info->duplex_setting) + update_link = true; } else { if (link_info->auto_mode == BNXT_LINK_AUTO_NONE) update_link = true; if (link_info->advertising != link_info->auto_link_speeds) update_link = true; - if (link_info->req_link_speed != link_info->auto_link_speed) - update_link = true; } if (update_link) @@@ -4636,7 -4642,7 +4642,7 @@@ static int __bnxt_open_nic(struct bnxt if (link_re_init) { rc = bnxt_update_phy_setting(bp); if (rc) - goto open_err; + netdev_warn(bp->dev, "failed to update phy settings\n"); } if (irq_re_init) { @@@ -4654,6 -4660,7 +4660,7 @@@ /* Enable TX queues */ bnxt_tx_enable(bp); mod_timer(&bp->timer, jiffies + bp->current_interval); + bnxt_update_link(bp, true); return 0; @@@ -5370,16 -5377,9 +5377,16 @@@ static int bnxt_change_mtu(struct net_d return 0; } -static int bnxt_setup_tc(struct net_device *dev, u8 tc) +static int bnxt_setup_tc(struct net_device *dev, u32 handle, __be16 proto, + struct tc_to_netdev *ntc) { struct bnxt *bp = netdev_priv(dev); + u8 tc; + + if (handle != TC_H_ROOT || ntc->type != TC_SETUP_MQPRIO) + return -EINVAL; + + tc = ntc->tc; if (tc > bp->max_tc) { netdev_err(dev, "too many traffic classes requested: %d Max supported is %d\n", @@@ -5677,22 -5677,16 +5684,16 @@@ static int bnxt_probe_phy(struct bnxt * } /*initialize the ethool setting copy with NVM settings */ - if (BNXT_AUTO_MODE(link_info->auto_mode)) - link_info->autoneg |= BNXT_AUTONEG_SPEED; - - if (link_info->auto_pause_setting & BNXT_LINK_PAUSE_BOTH) { - if (link_info->auto_pause_setting == BNXT_LINK_PAUSE_BOTH) - link_info->autoneg |= BNXT_AUTONEG_FLOW_CTRL; + if (BNXT_AUTO_MODE(link_info->auto_mode)) { + link_info->autoneg = BNXT_AUTONEG_SPEED | + BNXT_AUTONEG_FLOW_CTRL; + link_info->advertising = link_info->auto_link_speeds; link_info->req_flow_ctrl = link_info->auto_pause_setting; - } else if (link_info->force_pause_setting & BNXT_LINK_PAUSE_BOTH) { + } else { + link_info->req_link_speed = link_info->force_link_speed; + link_info->req_duplex = link_info->duplex_setting; link_info->req_flow_ctrl = link_info->force_pause_setting; } - link_info->req_duplex = link_info->duplex_setting; - if (link_info->autoneg & BNXT_AUTONEG_SPEED) - link_info->req_link_speed = link_info->auto_link_speed; - else - link_info->req_link_speed = link_info->force_link_speed; - link_info->advertising = link_info->auto_link_speeds; snprintf(phy_ver, PHY_VER_STR_LEN, " ph %d.%d.%d", link_info->phy_ver[0], link_info->phy_ver[1], diff --combined drivers/net/ethernet/cavium/thunder/nicvf_main.c index 90ce93e,a009bc3..bfee298 --- a/drivers/net/ethernet/cavium/thunder/nicvf_main.c +++ b/drivers/net/ethernet/cavium/thunder/nicvf_main.c @@@ -574,8 -574,7 +574,7 @@@ static inline void nicvf_set_rxhash(str static void nicvf_rcv_pkt_handler(struct net_device *netdev, struct napi_struct *napi, - struct cmp_queue *cq, - struct cqe_rx_t *cqe_rx, int cqe_type) + struct cqe_rx_t *cqe_rx) { struct sk_buff *skb; struct nicvf *nic = netdev_priv(netdev); @@@ -591,7 -590,7 +590,7 @@@ } /* Check for errors */ - err = nicvf_check_cqe_rx_errs(nic, cq, cqe_rx); + err = nicvf_check_cqe_rx_errs(nic, cqe_rx); if (err && !cqe_rx->rb_cnt) return; @@@ -682,8 -681,7 +681,7 @@@ loop cq_idx, cq_desc->cqe_type); switch (cq_desc->cqe_type) { case CQE_TYPE_RX: - nicvf_rcv_pkt_handler(netdev, napi, cq, - cq_desc, CQE_TYPE_RX); + nicvf_rcv_pkt_handler(netdev, napi, cq_desc); work_done++; break; case CQE_TYPE_SEND: @@@ -828,7 -826,7 +826,7 @@@ static irqreturn_t nicvf_intr_handler(i nicvf_disable_intr(nic, NICVF_INTR_CQ, qidx); /* Schedule NAPI */ - napi_schedule(&cq_poll->napi); + napi_schedule_irqoff(&cq_poll->napi); /* Clear interrupt */ nicvf_clear_intr(nic, NICVF_INTR_CQ, qidx); @@@ -899,31 -897,6 +897,31 @@@ static void nicvf_disable_msix(struct n } } +static void nicvf_set_irq_affinity(struct nicvf *nic) +{ + int vec, cpu; + int irqnum; + + for (vec = 0; vec < nic->num_vec; vec++) { + if (!nic->irq_allocated[vec]) + continue; + + if (!zalloc_cpumask_var(&nic->affinity_mask[vec], GFP_KERNEL)) + return; + /* CQ interrupts */ + if (vec < NICVF_INTR_ID_SQ) + /* Leave CPU0 for RBDR and other interrupts */ + cpu = nicvf_netdev_qidx(nic, vec) + 1; + else + cpu = 0; + + cpumask_set_cpu(cpumask_local_spread(cpu, nic->node), + nic->affinity_mask[vec]); + irqnum = nic->msix_entries[vec].vector; + irq_set_affinity_hint(irqnum, nic->affinity_mask[vec]); + } +} + static int nicvf_register_interrupts(struct nicvf *nic) { int irq, ret = 0; @@@ -969,13 -942,8 +967,13 @@@ ret = request_irq(nic->msix_entries[irq].vector, nicvf_qs_err_intr_handler, 0, nic->irq_name[irq], nic); - if (!ret) - nic->irq_allocated[irq] = true; + if (ret) + goto err; + + nic->irq_allocated[irq] = true; + + /* Set IRQ affinities */ + nicvf_set_irq_affinity(nic); err: if (ret) @@@ -993,9 -961,6 +991,9 @@@ static void nicvf_unregister_interrupts if (!nic->irq_allocated[irq]) continue; + irq_set_affinity_hint(nic->msix_entries[irq].vector, NULL); + free_cpumask_var(nic->affinity_mask[irq]); + if (irq < NICVF_INTR_ID_SQ) free_irq(nic->msix_entries[irq].vector, nic->napi[irq]); else @@@ -1158,7 -1123,6 +1156,6 @@@ int nicvf_stop(struct net_device *netde /* Clear multiqset info */ nic->pnicvf = nic; - nic->sqs_count = 0; return 0; } @@@ -1387,6 -1351,9 +1384,9 @@@ void nicvf_update_stats(struct nicvf *n drv_stats->tx_frames_ok = stats->tx_ucast_frames_ok + stats->tx_bcast_frames_ok + stats->tx_mcast_frames_ok; + drv_stats->rx_frames_ok = stats->rx_ucast_frames + + stats->rx_bcast_frames + + stats->rx_mcast_frames; drv_stats->rx_drops = stats->rx_drop_red + stats->rx_drop_overrun; drv_stats->tx_drops = stats->tx_drops; @@@ -1427,7 -1394,6 +1427,7 @@@ static void nicvf_tx_timeout(struct net netdev_warn(dev, "%s: Transmit timed out, resetting\n", dev->name); + nic->drv_stats.tx_timeout++; schedule_work(&nic->reset_task); } @@@ -1572,6 -1538,9 +1572,9 @@@ static int nicvf_probe(struct pci_dev * nicvf_send_vf_struct(nic); + if (!pass1_silicon(nic->pdev)) + nic->hw_tso = true; + /* Check if this VF is in QS only mode */ if (nic->sqs_mode) return 0; @@@ -1591,9 -1560,6 +1594,6 @@@ netdev->vlan_features = NETIF_F_SG | NETIF_F_IP_CSUM | NETIF_F_TSO; - if (!pass1_silicon(nic->pdev)) - nic->hw_tso = true; - netdev->netdev_ops = &nicvf_netdev_ops; netdev->watchdog_timeo = NICVF_TX_TIMEOUT; diff --combined drivers/net/ethernet/cavium/thunder/nicvf_queues.c index 5adb208,767347b..0dd1abf --- a/drivers/net/ethernet/cavium/thunder/nicvf_queues.c +++ b/drivers/net/ethernet/cavium/thunder/nicvf_queues.c @@@ -78,7 -78,7 +78,7 @@@ static void nicvf_free_q_desc_mem(struc static inline int nicvf_alloc_rcv_buffer(struct nicvf *nic, gfp_t gfp, u32 buf_len, u64 **rbuf) { - int order = get_order(buf_len); + int order = (PAGE_SIZE <= 4096) ? PAGE_ALLOC_COSTLY_ORDER : 0; /* Check if request can be accomodated in previous allocated page */ if (nic->rb_page) { @@@ -96,7 -96,8 +96,7 @@@ nic->rb_page = alloc_pages(gfp | __GFP_COMP | __GFP_NOWARN, order); if (!nic->rb_page) { - netdev_err(nic->netdev, - "Failed to allocate new rcv buffer\n"); + nic->drv_stats.rcv_buffer_alloc_failures++; return -ENOMEM; } nic->rb_page_offset = 0; @@@ -1328,16 -1329,12 +1328,12 @@@ void nicvf_update_sq_stats(struct nicv } /* Check for errors in the receive cmp.queue entry */ - int nicvf_check_cqe_rx_errs(struct nicvf *nic, - struct cmp_queue *cq, struct cqe_rx_t *cqe_rx) + int nicvf_check_cqe_rx_errs(struct nicvf *nic, struct cqe_rx_t *cqe_rx) { struct nicvf_hw_stats *stats = &nic->hw_stats; - struct nicvf_drv_stats *drv_stats = &nic->drv_stats; - if (!cqe_rx->err_level && !cqe_rx->err_opcode) { - drv_stats->rx_frames_ok++; + if (!cqe_rx->err_level && !cqe_rx->err_opcode) return 0; - } if (netif_msg_rx_err(nic)) netdev_err(nic->netdev, diff --combined drivers/net/ethernet/mellanox/mlx4/en_netdev.c index 01d6a96,f191a16..96d95cb --- a/drivers/net/ethernet/mellanox/mlx4/en_netdev.c +++ b/drivers/net/ethernet/mellanox/mlx4/en_netdev.c @@@ -69,15 -69,6 +69,15 @@@ int mlx4_en_setup_tc(struct net_device return 0; } +static int __mlx4_en_setup_tc(struct net_device *dev, u32 handle, __be16 proto, + struct tc_to_netdev *tc) +{ + if (handle != TC_H_ROOT || tc->type != TC_SETUP_MQPRIO) + return -EINVAL; + + return mlx4_en_setup_tc(dev, tc->tc); +} + #ifdef CONFIG_RFS_ACCEL struct mlx4_en_filter { @@@ -2353,8 -2344,6 +2353,6 @@@ out /* set offloads */ priv->dev->hw_enc_features |= NETIF_F_IP_CSUM | NETIF_F_RXCSUM | NETIF_F_TSO | NETIF_F_GSO_UDP_TUNNEL; - priv->dev->hw_features |= NETIF_F_GSO_UDP_TUNNEL; - priv->dev->features |= NETIF_F_GSO_UDP_TUNNEL; } static void mlx4_en_del_vxlan_offloads(struct work_struct *work) @@@ -2365,8 -2354,6 +2363,6 @@@ /* unset offloads */ priv->dev->hw_enc_features &= ~(NETIF_F_IP_CSUM | NETIF_F_RXCSUM | NETIF_F_TSO | NETIF_F_GSO_UDP_TUNNEL); - priv->dev->hw_features &= ~NETIF_F_GSO_UDP_TUNNEL; - priv->dev->features &= ~NETIF_F_GSO_UDP_TUNNEL; ret = mlx4_SET_PORT_VXLAN(priv->mdev->dev, priv->port, VXLAN_STEER_BY_OUTER_MAC, 0); @@@ -2475,7 -2462,7 +2471,7 @@@ static const struct net_device_ops mlx4 #endif .ndo_set_features = mlx4_en_set_features, .ndo_fix_features = mlx4_en_fix_features, - .ndo_setup_tc = mlx4_en_setup_tc, + .ndo_setup_tc = __mlx4_en_setup_tc, #ifdef CONFIG_RFS_ACCEL .ndo_rx_flow_steer = mlx4_en_filter_rfs, #endif @@@ -2513,7 -2500,7 +2509,7 @@@ static const struct net_device_ops mlx4 #endif .ndo_set_features = mlx4_en_set_features, .ndo_fix_features = mlx4_en_fix_features, - .ndo_setup_tc = mlx4_en_setup_tc, + .ndo_setup_tc = __mlx4_en_setup_tc, #ifdef CONFIG_RFS_ACCEL .ndo_rx_flow_steer = mlx4_en_filter_rfs, #endif @@@ -2989,6 -2976,11 +2985,11 @@@ int mlx4_en_init_netdev(struct mlx4_en_ priv->rss_hash_fn = ETH_RSS_HASH_TOP; } + if (mdev->dev->caps.tunnel_offload_mode == MLX4_TUNNEL_OFFLOAD_MODE_VXLAN) { + dev->hw_features |= NETIF_F_GSO_UDP_TUNNEL; + dev->features |= NETIF_F_GSO_UDP_TUNNEL; + } + mdev->pndev[port] = dev; mdev->upper[port] = NULL; diff --combined drivers/net/ethernet/renesas/ravb_main.c index 331c596,744d780..88656ce --- a/drivers/net/ethernet/renesas/ravb_main.c +++ b/drivers/net/ethernet/renesas/ravb_main.c @@@ -2,7 -2,7 +2,7 @@@ * * Copyright (C) 2014-2015 Renesas Electronics Corporation * Copyright (C) 2015 Renesas Solutions Corp. - * Copyright (C) 2015 Cogent Embedded, Inc. <source(a)cogentembedded.com> + * Copyright (C) 2015-2016 Cogent Embedded, Inc. <source(a)cogentembedded.com> * * Based on the SuperH Ethernet driver * @@@ -42,12 -42,6 +42,12 @@@ NETIF_MSG_RX_ERR | \ NETIF_MSG_TX_ERR) +void ravb_modify(struct net_device *ndev, enum ravb_reg reg, u32 clear, + u32 set) +{ + ravb_write(ndev, (ravb_read(ndev, reg) & ~clear) | set, reg); +} + int ravb_wait(struct net_device *ndev, enum ravb_reg reg, u32 mask, u32 value) { int i; @@@ -65,7 -59,8 +65,7 @@@ static int ravb_config(struct net_devic int error; /* Set config mode */ - ravb_write(ndev, (ravb_read(ndev, CCC) & ~CCC_OPC) | CCC_OPC_CONFIG, - CCC); + ravb_modify(ndev, CCC, CCC_OPC, CCC_OPC_CONFIG); /* Check if the operating mode is changed to the config mode */ error = ravb_wait(ndev, CSR, CSR_OPS, CSR_OPS_CONFIG); if (error) @@@ -77,8 -72,13 +77,8 @@@ static void ravb_set_duplex(struct net_device *ndev) { struct ravb_private *priv = netdev_priv(ndev); - u32 ecmr = ravb_read(ndev, ECMR); - if (priv->duplex) /* Full */ - ecmr |= ECMR_DM; - else /* Half */ - ecmr &= ~ECMR_DM; - ravb_write(ndev, ecmr, ECMR); + ravb_modify(ndev, ECMR, ECMR_DM, priv->duplex ? ECMR_DM : 0); } static void ravb_set_rate(struct net_device *ndev) @@@ -92,6 -92,8 +92,6 @@@ case 1000: /* 1000BASE */ ravb_write(ndev, GECMR_SPEED_1000, GECMR); break; - default: - break; } } @@@ -129,8 -131,13 +129,8 @@@ static void ravb_mdio_ctrl(struct mdiob { struct ravb_private *priv = container_of(ctrl, struct ravb_private, mdiobb); - u32 pir = ravb_read(priv->ndev, PIR); - if (set) - pir |= mask; - else - pir &= ~mask; - ravb_write(priv->ndev, pir, PIR); + ravb_modify(priv->ndev, PIR, mask, set ? mask : 0); } /* MDC pin control */ @@@ -386,9 -393,9 +386,9 @@@ static int ravb_dmac_init(struct net_de ravb_ring_format(ndev, RAVB_NC); #if defined(__LITTLE_ENDIAN) - ravb_write(ndev, ravb_read(ndev, CCC) & ~CCC_BOC, CCC); + ravb_modify(ndev, CCC, CCC_BOC, 0); #else - ravb_write(ndev, ravb_read(ndev, CCC) | CCC_BOC, CCC); + ravb_modify(ndev, CCC, CCC_BOC, CCC_BOC); #endif /* Set AVB RX */ @@@ -411,7 -418,8 +411,7 @@@ ravb_write(ndev, TIC_FTE0 | TIC_FTE1 | TIC_TFUE, TIC); /* Setting the control will start the AVB-DMAC process. */ - ravb_write(ndev, (ravb_read(ndev, CCC) & ~CCC_OPC) | CCC_OPC_OPERATION, - CCC); + ravb_modify(ndev, CCC, CCC_OPC, CCC_OPC_OPERATION); return 0; } @@@ -485,7 -493,7 +485,7 @@@ static void ravb_get_tx_tstamp(struct n break; } } - ravb_write(ndev, ravb_read(ndev, TCCR) | TCCR_TFR, TCCR); + ravb_modify(ndev, TCCR, TCCR_TFR, TCCR_TFR); } } @@@ -605,13 -613,13 +605,13 @@@ static bool ravb_rx(struct net_device * static void ravb_rcv_snd_disable(struct net_device *ndev) { /* Disable TX and RX */ - ravb_write(ndev, ravb_read(ndev, ECMR) & ~(ECMR_RE | ECMR_TE), ECMR); + ravb_modify(ndev, ECMR, ECMR_RE | ECMR_TE, 0); } static void ravb_rcv_snd_enable(struct net_device *ndev) { /* Enable TX and RX */ - ravb_write(ndev, ravb_read(ndev, ECMR) | ECMR_RE | ECMR_TE, ECMR); + ravb_modify(ndev, ECMR, ECMR_RE | ECMR_TE, ECMR_RE | ECMR_TE); } /* function for waiting dma process finished */ @@@ -804,8 -812,8 +804,8 @@@ static int ravb_poll(struct napi_struc /* Re-enable RX/TX interrupts */ spin_lock_irqsave(&priv->lock, flags); - ravb_write(ndev, ravb_read(ndev, RIC0) | mask, RIC0); - ravb_write(ndev, ravb_read(ndev, TIC) | mask, TIC); + ravb_modify(ndev, RIC0, mask, mask); + ravb_modify(ndev, TIC, mask, mask); mmiowb(); spin_unlock_irqrestore(&priv->lock, flags); @@@ -844,7 -852,8 +844,7 @@@ static void ravb_adjust_link(struct net ravb_set_rate(ndev); } if (!priv->link) { - ravb_write(ndev, ravb_read(ndev, ECMR) & ~ECMR_TXF, - ECMR); + ravb_modify(ndev, ECMR, ECMR_TXF, 0); new_state = true; priv->link = phydev->link; if (priv->no_avb_link) @@@ -1130,7 -1139,8 +1130,8 @@@ static int ravb_set_ringparam(struct ne if (netif_running(ndev)) { netif_device_detach(ndev); /* Stop PTP Clock driver */ - ravb_ptp_stop(ndev); + if (priv->chip_id == RCAR_GEN2) + ravb_ptp_stop(ndev); /* Wait for DMA stopping */ error = ravb_stop_dma(ndev); if (error) { @@@ -1161,7 -1171,8 +1162,8 @@@ ravb_emac_init(ndev); /* Initialise PTP Clock driver */ - ravb_ptp_init(ndev, priv->pdev); + if (priv->chip_id == RCAR_GEN2) + ravb_ptp_init(ndev, priv->pdev); netif_device_attach(ndev); } @@@ -1289,7 -1300,8 +1291,8 @@@ static void ravb_tx_timeout_work(struc netif_tx_stop_all_queues(ndev); /* Stop PTP Clock driver */ - ravb_ptp_stop(ndev); + if (priv->chip_id == RCAR_GEN2) + ravb_ptp_stop(ndev); /* Wait for DMA stopping */ ravb_stop_dma(ndev); @@@ -1302,7 -1314,8 +1305,8 @@@ ravb_emac_init(ndev); /* Initialise PTP Clock driver */ - ravb_ptp_init(ndev, priv->pdev); + if (priv->chip_id == RCAR_GEN2) + ravb_ptp_init(ndev, priv->pdev); netif_tx_start_all_queues(ndev); } @@@ -1384,7 -1397,7 +1388,7 @@@ static netdev_tx_t ravb_start_xmit(stru desc--; desc->die_dt = DT_FSTART; - ravb_write(ndev, ravb_read(ndev, TCCR) | (TCCR_TSRQ0 << q), TCCR); + ravb_modify(ndev, TCCR, TCCR_TSRQ0 << q, TCCR_TSRQ0 << q); priv->cur_tx[q] += NUM_TX_DESC; if (priv->cur_tx[q] - priv->dirty_tx[q] > @@@ -1459,10 -1472,15 +1463,10 @@@ static void ravb_set_rx_mode(struct net { struct ravb_private *priv = netdev_priv(ndev); unsigned long flags; - u32 ecmr; spin_lock_irqsave(&priv->lock, flags); - ecmr = ravb_read(ndev, ECMR); - if (ndev->flags & IFF_PROMISC) - ecmr |= ECMR_PRM; - else - ecmr &= ~ECMR_PRM; - ravb_write(ndev, ecmr, ECMR); + ravb_modify(ndev, ECMR, ECMR_PRM, + ndev->flags & IFF_PROMISC ? ECMR_PRM : 0); mmiowb(); spin_unlock_irqrestore(&priv->lock, flags); } @@@ -1790,25 -1808,23 +1794,21 @@@ static int ravb_probe(struct platform_d /* Set AVB config mode */ if (chip_id == RCAR_GEN2) { - ravb_write(ndev, (ravb_read(ndev, CCC) & ~CCC_OPC) | - CCC_OPC_CONFIG, CCC); + ravb_modify(ndev, CCC, CCC_OPC, CCC_OPC_CONFIG); /* Set CSEL value */ - ravb_write(ndev, (ravb_read(ndev, CCC) & ~CCC_CSEL) | - CCC_CSEL_HPB, CCC); + ravb_modify(ndev, CCC, CCC_CSEL, CCC_CSEL_HPB); } else { - ravb_write(ndev, (ravb_read(ndev, CCC) & ~CCC_OPC) | - CCC_OPC_CONFIG | CCC_GAC | CCC_CSEL_HPB, CCC); + ravb_modify(ndev, CCC, CCC_OPC, CCC_OPC_CONFIG | + CCC_GAC | CCC_CSEL_HPB); } - /* Set CSEL value */ - ravb_write(ndev, (ravb_read(ndev, CCC) & ~CCC_CSEL) | CCC_CSEL_HPB, - CCC); - /* Set GTI value */ error = ravb_set_gti(ndev); if (error) goto out_release; /* Request GTI loading */ - ravb_write(ndev, ravb_read(ndev, GCCR) | GCCR_LTI, GCCR); + ravb_modify(ndev, GCCR, GCCR_LTI, GCCR_LTI); /* Allocate descriptor base address table */ priv->desc_bat_size = sizeof(struct ravb_desc) * DBAT_ENTRY_NUM; diff --combined drivers/net/ethernet/ti/netcp_core.c index 06a0a73,029841f..ed0c30f --- a/drivers/net/ethernet/ti/netcp_core.c +++ b/drivers/net/ethernet/ti/netcp_core.c @@@ -117,21 -117,17 +117,17 @@@ static void get_pkt_info(dma_addr_t *bu *ndesc = le32_to_cpu(desc->next_desc); } - static void get_pad_info(u32 *pad0, u32 *pad1, u32 *pad2, struct knav_dma_desc *desc) + static u32 get_sw_data(int index, struct knav_dma_desc *desc) { - *pad0 = le32_to_cpu(desc->pad[0]); - *pad1 = le32_to_cpu(desc->pad[1]); - *pad2 = le32_to_cpu(desc->pad[2]); + /* No Endian conversion needed as this data is untouched by hw */ + return desc->sw_data[index]; } - static void get_pad_ptr(void **padptr, struct knav_dma_desc *desc) - { - u64 pad64; - - pad64 = le32_to_cpu(desc->pad[0]) + - ((u64)le32_to_cpu(desc->pad[1]) << 32); - *padptr = (void *)(uintptr_t)pad64; - } + /* use these macros to get sw data */ + #define GET_SW_DATA0(desc) get_sw_data(0, desc) + #define GET_SW_DATA1(desc) get_sw_data(1, desc) + #define GET_SW_DATA2(desc) get_sw_data(2, desc) + #define GET_SW_DATA3(desc) get_sw_data(3, desc) static void get_org_pkt_info(dma_addr_t *buff, u32 *buff_len, struct knav_dma_desc *desc) @@@ -163,13 -159,18 +159,18 @@@ static void set_desc_info(u32 desc_info desc->packet_info = cpu_to_le32(pkt_info); } - static void set_pad_info(u32 pad0, u32 pad1, u32 pad2, struct knav_dma_desc *desc) + static void set_sw_data(int index, u32 data, struct knav_dma_desc *desc) { - desc->pad[0] = cpu_to_le32(pad0); - desc->pad[1] = cpu_to_le32(pad1); - desc->pad[2] = cpu_to_le32(pad1); + /* No Endian conversion needed as this data is untouched by hw */ + desc->sw_data[index] = data; } + /* use these macros to set sw data */ + #define SET_SW_DATA0(data, desc) set_sw_data(0, data, desc) + #define SET_SW_DATA1(data, desc) set_sw_data(1, data, desc) + #define SET_SW_DATA2(data, desc) set_sw_data(2, data, desc) + #define SET_SW_DATA3(data, desc) set_sw_data(3, data, desc) + static void set_org_pkt_info(dma_addr_t buff, u32 buff_len, struct knav_dma_desc *desc) { @@@ -581,7 -582,6 +582,6 @@@ static void netcp_free_rx_desc_chain(st dma_addr_t dma_desc, dma_buf; unsigned int buf_len, dma_sz = sizeof(*ndesc); void *buf_ptr; - u32 pad[2]; u32 tmp; get_words(&dma_desc, 1, &desc->next_desc); @@@ -593,14 -593,20 +593,20 @@@ break; } get_pkt_info(&dma_buf, &tmp, &dma_desc, ndesc); - get_pad_ptr(&buf_ptr, ndesc); + /* warning!!!! We are retrieving the virtual ptr in the sw_data + * field as a 32bit value. Will not work on 64bit machines + */ + buf_ptr = (void *)GET_SW_DATA0(ndesc); + buf_len = (int)GET_SW_DATA1(desc); dma_unmap_page(netcp->dev, dma_buf, PAGE_SIZE, DMA_FROM_DEVICE); __free_page(buf_ptr); knav_pool_desc_put(netcp->rx_pool, desc); } - - get_pad_info(&pad[0], &pad[1], &buf_len, desc); - buf_ptr = (void *)(uintptr_t)(pad[0] + ((u64)pad[1] << 32)); + /* warning!!!! We are retrieving the virtual ptr in the sw_data + * field as a 32bit value. Will not work on 64bit machines + */ + buf_ptr = (void *)GET_SW_DATA0(desc); + buf_len = (int)GET_SW_DATA1(desc); if (buf_ptr) netcp_frag_free(buf_len <= PAGE_SIZE, buf_ptr); @@@ -639,7 -645,6 +645,6 @@@ static int netcp_process_one_rx_packet( dma_addr_t dma_desc, dma_buff; struct netcp_packet p_info; struct sk_buff *skb; - u32 pad[2]; void *org_buf_ptr; dma_desc = knav_queue_pop(netcp->rx_queue, &dma_sz); @@@ -653,8 -658,11 +658,11 @@@ } get_pkt_info(&dma_buff, &buf_len, &dma_desc, desc); - get_pad_info(&pad[0], &pad[1], &org_buf_len, desc); - org_buf_ptr = (void *)(uintptr_t)(pad[0] + ((u64)pad[1] << 32)); + /* warning!!!! We are retrieving the virtual ptr in the sw_data + * field as a 32bit value. Will not work on 64bit machines + */ + org_buf_ptr = (void *)GET_SW_DATA0(desc); + org_buf_len = (int)GET_SW_DATA1(desc); if (unlikely(!org_buf_ptr)) { dev_err(netcp->ndev_dev, "NULL bufptr in desc\n"); @@@ -679,7 -687,6 +687,6 @@@ /* Fill in the page fragment list */ while (dma_desc) { struct page *page; - void *ptr; ndesc = knav_pool_desc_unmap(netcp->rx_pool, dma_desc, dma_sz); if (unlikely(!ndesc)) { @@@ -688,8 -695,10 +695,10 @@@ } get_pkt_info(&dma_buff, &buf_len, &dma_desc, ndesc); - get_pad_ptr(&ptr, ndesc); - page = ptr; + /* warning!!!! We are retrieving the virtual ptr in the sw_data + * field as a 32bit value. Will not work on 64bit machines + */ + page = (struct page *)GET_SW_DATA0(desc); if (likely(dma_buff && buf_len && page)) { dma_unmap_page(netcp->dev, dma_buff, PAGE_SIZE, @@@ -777,7 -786,10 +786,10 @@@ static void netcp_free_rx_buf(struct ne } get_org_pkt_info(&dma, &buf_len, desc); - get_pad_ptr(&buf_ptr, desc); + /* warning!!!! We are retrieving the virtual ptr in the sw_data + * field as a 32bit value. Will not work on 64bit machines + */ + buf_ptr = (void *)GET_SW_DATA0(desc); if (unlikely(!dma)) { dev_err(netcp->ndev_dev, "NULL orig_buff in desc\n"); @@@ -829,7 -841,7 +841,7 @@@ static int netcp_allocate_rx_buf(struc struct page *page; dma_addr_t dma; void *bufptr; - u32 pad[3]; + u32 sw_data[2]; /* Allocate descriptor */ hwdesc = knav_pool_desc_get(netcp->rx_pool); @@@ -846,7 -858,7 +858,7 @@@ SKB_DATA_ALIGN(sizeof(struct skb_shared_info)); bufptr = netdev_alloc_frag(primary_buf_len); - pad[2] = primary_buf_len; + sw_data[1] = primary_buf_len; if (unlikely(!bufptr)) { dev_warn_ratelimited(netcp->ndev_dev, @@@ -858,9 -870,10 +870,10 @@@ if (unlikely(dma_mapping_error(netcp->dev, dma))) goto fail; - pad[0] = lower_32_bits((uintptr_t)bufptr); - pad[1] = upper_32_bits((uintptr_t)bufptr); - + /* warning!!!! We are saving the virtual ptr in the sw_data + * field as a 32bit value. Will not work on 64bit machines + */ + sw_data[0] = (u32)bufptr; } else { /* Allocate a secondary receive queue entry */ page = alloc_page(GFP_ATOMIC | GFP_DMA | __GFP_COLD); @@@ -870,9 -883,11 +883,11 @@@ } buf_len = PAGE_SIZE; dma = dma_map_page(netcp->dev, page, 0, buf_len, DMA_TO_DEVICE); - pad[0] = lower_32_bits(dma); - pad[1] = upper_32_bits(dma); - pad[2] = 0; + /* warning!!!! We are saving the virtual ptr in the sw_data + * field as a 32bit value. Will not work on 64bit machines + */ + sw_data[0] = (u32)page; + sw_data[1] = 0; } desc_info = KNAV_DMA_DESC_PS_INFO_IN_DESC; @@@ -882,7 -897,8 +897,8 @@@ pkt_info |= (netcp->rx_queue_id & KNAV_DMA_DESC_RETQ_MASK) << KNAV_DMA_DESC_RETQ_SHIFT; set_org_pkt_info(dma, buf_len, hwdesc); - set_pad_info(pad[0], pad[1], pad[2], hwdesc); + SET_SW_DATA0(sw_data[0], hwdesc); + SET_SW_DATA1(sw_data[1], hwdesc); set_desc_info(desc_info, pkt_info, hwdesc); /* Push to FDQs */ @@@ -971,7 -987,6 +987,6 @@@ static int netcp_process_tx_compl_packe unsigned int budget) { struct knav_dma_desc *desc; - void *ptr; struct sk_buff *skb; unsigned int dma_sz; dma_addr_t dma; @@@ -988,8 -1003,10 +1003,10 @@@ continue; } - get_pad_ptr(&ptr, desc); - skb = ptr; + /* warning!!!! We are retrieving the virtual ptr in the sw_data + * field as a 32bit value. Will not work on 64bit machines + */ + skb = (struct sk_buff *)GET_SW_DATA0(desc); netcp_free_tx_desc_chain(netcp, desc, dma_sz); if (!skb) { dev_err(netcp->ndev_dev, "No skb in Tx desc\n"); @@@ -1194,10 -1211,10 +1211,10 @@@ static int netcp_tx_submit_skb(struct n } set_words(&tmp, 1, &desc->packet_info); - tmp = lower_32_bits((uintptr_t)&skb); - set_words(&tmp, 1, &desc->pad[0]); - tmp = upper_32_bits((uintptr_t)&skb); - set_words(&tmp, 1, &desc->pad[1]); + /* warning!!!! We are saving the virtual ptr in the sw_data + * field as a 32bit value. Will not work on 64bit machines + */ + SET_SW_DATA0((u32)skb, desc); if (tx_pipe->flags & SWITCH_TO_PORT_IN_TAGINFO) { tmp = tx_pipe->switch_to_port; @@@ -1835,26 -1852,22 +1852,26 @@@ static u16 netcp_select_queue(struct ne return 0; } -static int netcp_setup_tc(struct net_device *dev, u8 num_tc) +static int netcp_setup_tc(struct net_device *dev, u32 handle, __be16 proto, + struct tc_to_netdev *tc) { int i; /* setup tc must be called under rtnl lock */ ASSERT_RTNL(); + if (handle != TC_H_ROOT || tc->type != TC_SETUP_MQPRIO) + return -EINVAL; + /* Sanity-check the number of traffic classes requested */ if ((dev->real_num_tx_queues <= 1) || - (dev->real_num_tx_queues < num_tc)) + (dev->real_num_tx_queues < tc->tc)) return -EINVAL; /* Configure traffic class to queue mappings */ - if (num_tc) { - netdev_set_num_tc(dev, num_tc); - for (i = 0; i < num_tc; i++) + if (tc->tc) { + netdev_set_num_tc(dev, tc->tc); + for (i = 0; i < tc->tc; i++) netdev_set_tc_queue(dev, i, 1, i); } else { netdev_reset_tc(dev); diff --combined drivers/net/geneve.c index 8fa8388,0bf7edd..bc5da35 --- a/drivers/net/geneve.c +++ b/drivers/net/geneve.c @@@ -72,11 -72,10 +72,11 @@@ struct geneve_dev bool collect_md; struct gro_cells gro_cells; u32 flags; + struct dst_cache dst_cache; }; /* Geneve device flags */ -#define GENEVE_F_UDP_CSUM BIT(0) +#define GENEVE_F_UDP_ZERO_CSUM_TX BIT(0) #define GENEVE_F_UDP_ZERO_CSUM6_TX BIT(1) #define GENEVE_F_UDP_ZERO_CSUM6_RX BIT(2) @@@ -110,11 -109,6 +110,11 @@@ static __be64 vni_to_tunnel_id(const __ #endif } +static sa_family_t geneve_get_sk_family(struct geneve_sock *gs) +{ + return gs->sock->sk->sk_family; +} + static struct geneve_dev *geneve_lookup(struct geneve_sock *gs, __be32 addr, u8 vni[]) { @@@ -158,60 -152,58 +158,60 @@@ static inline struct genevehdr *geneve_ return (struct genevehdr *)(udp_hdr(skb) + 1); } -/* geneve receive/decap routine */ -static void geneve_rx(struct geneve_sock *gs, struct sk_buff *skb) +static struct geneve_dev *geneve_lookup_skb(struct geneve_sock *gs, + struct sk_buff *skb) { - struct genevehdr *gnvh = geneve_hdr(skb); - struct metadata_dst *tun_dst = NULL; - struct geneve_dev *geneve = NULL; - struct pcpu_sw_netstats *stats; - struct iphdr *iph = NULL; + u8 *vni; __be32 addr; static u8 zero_vni[3]; - u8 *vni; - int err = 0; - sa_family_t sa_family; #if IS_ENABLED(CONFIG_IPV6) - struct ipv6hdr *ip6h = NULL; - struct in6_addr addr6; static struct in6_addr zero_addr6; #endif - sa_family = gs->sock->sk->sk_family; + if (geneve_get_sk_family(gs) == AF_INET) { + struct iphdr *iph; - if (sa_family == AF_INET) { iph = ip_hdr(skb); /* outer IP header... */ if (gs->collect_md) { vni = zero_vni; addr = 0; } else { - vni = gnvh->vni; - + vni = geneve_hdr(skb)->vni; addr = iph->saddr; } - geneve = geneve_lookup(gs, addr, vni); + return geneve_lookup(gs, addr, vni); #if IS_ENABLED(CONFIG_IPV6) - } else if (sa_family == AF_INET6) { + } else if (geneve_get_sk_family(gs) == AF_INET6) { + struct ipv6hdr *ip6h; + struct in6_addr addr6; + ip6h = ipv6_hdr(skb); /* outer IPv6 header... */ if (gs->collect_md) { vni = zero_vni; addr6 = zero_addr6; } else { - vni = gnvh->vni; - + vni = geneve_hdr(skb)->vni; addr6 = ip6h->saddr; } - geneve = geneve6_lookup(gs, addr6, vni); + return geneve6_lookup(gs, addr6, vni); #endif } - if (!geneve) - goto drop; + return NULL; +} + +/* geneve receive/decap routine */ +static void geneve_rx(struct geneve_dev *geneve, struct geneve_sock *gs, + struct sk_buff *skb) +{ + struct genevehdr *gnvh = geneve_hdr(skb); + struct metadata_dst *tun_dst = NULL; + struct pcpu_sw_netstats *stats; + int err = 0; + void *oiph; if (ip_tunnel_collect_metadata() || gs->collect_md) { __be16 flags; @@@ -220,7 -212,7 +220,7 @@@ (gnvh->oam ? TUNNEL_OAM : 0) | (gnvh->critical ? TUNNEL_CRIT_OPT : 0); - tun_dst = udp_tun_rx_dst(skb, sa_family, flags, + tun_dst = udp_tun_rx_dst(skb, geneve_get_sk_family(gs), flags, vni_to_tunnel_id(gnvh->vni), gnvh->opt_len * 4); if (!tun_dst) @@@ -237,6 -229,7 +237,6 @@@ } skb_reset_mac_header(skb); - skb_scrub_packet(skb, !net_eq(geneve->net, dev_net(geneve->dev))); skb->protocol = eth_type_trans(skb, geneve->dev); skb_postpull_rcsum(skb, eth_hdr(skb), ETH_HLEN); @@@ -247,27 -240,25 +247,27 @@@ if (ether_addr_equal(eth_hdr(skb)->h_source, geneve->dev->dev_addr)) goto drop; + oiph = skb_network_header(skb); skb_reset_network_header(skb); - if (iph) - err = IP_ECN_decapsulate(iph, skb); + if (geneve_get_sk_family(gs) == AF_INET) + err = IP_ECN_decapsulate(oiph, skb); #if IS_ENABLED(CONFIG_IPV6) - if (ip6h) - err = IP6_ECN_decapsulate(ip6h, skb); + else + err = IP6_ECN_decapsulate(oiph, skb); #endif if (unlikely(err)) { if (log_ecn_error) { - if (iph) + if (geneve_get_sk_family(gs) == AF_INET) net_info_ratelimited("non-ECT from %pI4 " "with TOS=%#x\n", - &iph->saddr, iph->tos); + &((struct iphdr *)oiph)->saddr, + ((struct iphdr *)oiph)->tos); #if IS_ENABLED(CONFIG_IPV6) - if (ip6h) + else net_info_ratelimited("non-ECT from %pI6\n", - &ip6h->saddr); + &((struct ipv6hdr *)oiph)->saddr); #endif } if (err > 1) { @@@ -306,13 -297,6 +306,13 @@@ static int geneve_init(struct net_devic return err; } + err = dst_cache_init(&geneve->dst_cache, GFP_KERNEL); + if (err) { + free_percpu(dev->tstats); + gro_cells_destroy(&geneve->gro_cells); + return err; + } + return 0; } @@@ -320,7 -304,6 +320,7 @@@ static void geneve_uninit(struct net_de { struct geneve_dev *geneve = netdev_priv(dev); + dst_cache_destroy(&geneve->dst_cache); gro_cells_destroy(&geneve->gro_cells); free_percpu(dev->tstats); } @@@ -329,7 -312,6 +329,7 @@@ static int geneve_udp_encap_recv(struct sock *sk, struct sk_buff *skb) { struct genevehdr *geneveh; + struct geneve_dev *geneve; struct geneve_sock *gs; int opts_len; @@@ -345,21 -327,16 +345,21 @@@ if (unlikely(geneveh->proto_type != htons(ETH_P_TEB))) goto error; - opts_len = geneveh->opt_len * 4; - if (iptunnel_pull_header(skb, GENEVE_BASE_HLEN + opts_len, - htons(ETH_P_TEB))) - goto drop; - gs = rcu_dereference_sk_user_data(sk); if (!gs) goto drop; - geneve_rx(gs, skb); + geneve = geneve_lookup_skb(gs, skb); + if (!geneve) + goto drop; + + opts_len = geneveh->opt_len * 4; + if (iptunnel_pull_header(skb, GENEVE_BASE_HLEN + opts_len, + htons(ETH_P_TEB), + !net_eq(geneve->net, dev_net(geneve->dev)))) + goto drop; + + geneve_rx(geneve, gs, skb); return 0; drop: @@@ -406,7 -383,7 +406,7 @@@ static void geneve_notify_add_rx_port(s struct net_device *dev; struct sock *sk = gs->sock->sk; struct net *net = sock_net(sk); - sa_family_t sa_family = sk->sk_family; + sa_family_t sa_family = geneve_get_sk_family(gs); __be16 port = inet_sk(sk)->inet_sport; int err; @@@ -567,7 -544,7 +567,7 @@@ static void geneve_notify_del_rx_port(s struct net_device *dev; struct sock *sk = gs->sock->sk; struct net *net = sock_net(sk); - sa_family_t sa_family = sk->sk_family; + sa_family_t sa_family = geneve_get_sk_family(gs); __be16 port = inet_sk(sk)->inet_sport; rcu_read_lock(); @@@ -610,7 -587,7 +610,7 @@@ static struct geneve_sock *geneve_find_ list_for_each_entry(gs, &gn->sock_list, list) { if (inet_sk(gs->sock->sk)->inet_sport == dst_port && - inet_sk(gs->sock->sk)->sk.sk_family == family) { + geneve_get_sk_family(gs) == family) { return gs; } } @@@ -703,7 -680,7 +703,7 @@@ static int geneve_build_skb(struct rtab struct genevehdr *gnvh; int min_headroom; int err; - bool udp_sum = !!(flags & GENEVE_F_UDP_CSUM); + bool udp_sum = !(flags & GENEVE_F_UDP_ZERO_CSUM_TX); skb_scrub_packet(skb, xnet); @@@ -776,9 -753,7 +776,9 @@@ static struct rtable *geneve_get_v4_rt( struct ip_tunnel_info *info) { struct geneve_dev *geneve = netdev_priv(dev); + struct dst_cache *dst_cache; struct rtable *rt = NULL; + bool use_cache = true; __u8 tos; memset(fl4, 0, sizeof(*fl4)); @@@ -789,26 -764,16 +789,26 @@@ fl4->daddr = info->key.u.ipv4.dst; fl4->saddr = info->key.u.ipv4.src; fl4->flowi4_tos = RT_TOS(info->key.tos); + dst_cache = &info->dst_cache; } else { tos = geneve->tos; if (tos == 1) { const struct iphdr *iip = ip_hdr(skb); tos = ip_tunnel_get_dsfield(iip, skb); + use_cache = false; } fl4->flowi4_tos = RT_TOS(tos); fl4->daddr = geneve->remote.sin.sin_addr.s_addr; + dst_cache = &geneve->dst_cache; + } + + use_cache = use_cache && !skb->mark; + if (use_cache) { + rt = dst_cache_get_ip4(dst_cache, &fl4->saddr); + if (rt) + return rt; } rt = ip_route_output_key(geneve->net, fl4); @@@ -821,8 -786,6 +821,8 @@@ ip_rt_put(rt); return ERR_PTR(-ELOOP); } + if (use_cache) + dst_cache_set_ip4(dst_cache, &rt->dst, fl4->saddr); return rt; } @@@ -835,8 -798,6 +835,8 @@@ static struct dst_entry *geneve_get_v6_ struct geneve_dev *geneve = netdev_priv(dev); struct geneve_sock *gs6 = geneve->sock6; struct dst_entry *dst = NULL; + struct dst_cache *dst_cache; + bool use_cache = true; __u8 prio; memset(fl6, 0, sizeof(*fl6)); @@@ -847,26 -808,16 +847,26 @@@ fl6->daddr = info->key.u.ipv6.dst; fl6->saddr = info->key.u.ipv6.src; fl6->flowi6_tos = RT_TOS(info->key.tos); + dst_cache = &info->dst_cache; } else { prio = geneve->tos; if (prio == 1) { const struct iphdr *iip = ip_hdr(skb); prio = ip_tunnel_get_dsfield(iip, skb); + use_cache = false; } fl6->flowi6_tos = RT_TOS(prio); fl6->daddr = geneve->remote.sin6.sin6_addr; + dst_cache = &geneve->dst_cache; + } + + use_cache = use_cache && !skb->mark; + if (use_cache) { + dst = dst_cache_get_ip6(dst_cache, &fl6->saddr); + if (dst) + return dst; } if (ipv6_stub->ipv6_dst_lookup(geneve->net, gs6->sock->sk, &dst, fl6)) { @@@ -879,8 -830,6 +879,8 @@@ return ERR_PTR(-ELOOP); } + if (use_cache) + dst_cache_set_ip6(dst_cache, dst, &fl6->saddr); return dst; } #endif @@@ -944,9 -893,9 +944,9 @@@ static netdev_tx_t geneve_xmit_skb(stru opts = ip_tunnel_info_opts(info); if (key->tun_flags & TUNNEL_CSUM) - flags |= GENEVE_F_UDP_CSUM; + flags &= ~GENEVE_F_UDP_ZERO_CSUM_TX; else - flags &= ~GENEVE_F_UDP_CSUM; + flags |= GENEVE_F_UDP_ZERO_CSUM_TX; err = geneve_build_skb(rt, skb, key->tun_flags, vni, info->options_len, opts, flags, xnet); @@@ -972,7 -921,7 +972,7 @@@ udp_tunnel_xmit_skb(rt, gs4->sock->sk, skb, fl4.saddr, fl4.daddr, tos, ttl, df, sport, geneve->dst_port, !net_eq(geneve->net, dev_net(geneve->dev)), - !(flags & GENEVE_F_UDP_CSUM)); + !!(flags & GENEVE_F_UDP_ZERO_CSUM_TX)); return NETDEV_TX_OK; @@@ -1090,6 -1039,34 +1090,34 @@@ static netdev_tx_t geneve_xmit(struct s return geneve_xmit_skb(skb, dev, info); } + static int __geneve_change_mtu(struct net_device *dev, int new_mtu, bool strict) + { + /* The max_mtu calculation does not take account of GENEVE + * options, to avoid excluding potentially valid + * configurations. + */ + int max_mtu = IP_MAX_MTU - GENEVE_BASE_HLEN - sizeof(struct iphdr) + - dev->hard_header_len; + + if (new_mtu < 68) + return -EINVAL; + + if (new_mtu > max_mtu) { + if (strict) + return -EINVAL; + + new_mtu = max_mtu; + } + + dev->mtu = new_mtu; + return 0; + } + + static int geneve_change_mtu(struct net_device *dev, int new_mtu) + { + return __geneve_change_mtu(dev, new_mtu, true); + } + static int geneve_fill_metadata_dst(struct net_device *dev, struct sk_buff *skb) { struct ip_tunnel_info *info = skb_tunnel_info(skb); @@@ -1134,7 -1111,7 +1162,7 @@@ static const struct net_device_ops gene .ndo_stop = geneve_stop, .ndo_start_xmit = geneve_xmit, .ndo_get_stats64 = ip_tunnel_get_stats64, - .ndo_change_mtu = eth_change_mtu, + .ndo_change_mtu = geneve_change_mtu, .ndo_validate_addr = eth_validate_addr, .ndo_set_mac_address = eth_mac_addr, .ndo_fill_metadata_dst = geneve_fill_metadata_dst, @@@ -1201,6 -1178,7 +1229,7 @@@ static void geneve_setup(struct net_dev dev->hw_features |= NETIF_F_GSO_SOFTWARE; netif_keep_dst(dev); + dev->priv_flags &= ~IFF_TX_SKB_SHARING; dev->priv_flags |= IFF_LIVE_ADDR_CHANGE | IFF_NO_QUEUE; eth_hw_addr_random(dev); } @@@ -1323,8 -1301,6 +1352,8 @@@ static int geneve_configure(struct net return -EPERM; } + dst_cache_reset(&geneve->dst_cache); + err = register_netdevice(dev); if (err) return err; @@@ -1383,8 -1359,8 +1412,8 @@@ static int geneve_newlink(struct net *n metadata = true; if (data[IFLA_GENEVE_UDP_CSUM] && - nla_get_u8(data[IFLA_GENEVE_UDP_CSUM])) - flags |= GENEVE_F_UDP_CSUM; + !nla_get_u8(data[IFLA_GENEVE_UDP_CSUM])) + flags |= GENEVE_F_UDP_ZERO_CSUM_TX; if (data[IFLA_GENEVE_UDP_ZERO_CSUM6_TX] && nla_get_u8(data[IFLA_GENEVE_UDP_ZERO_CSUM6_TX])) @@@ -1454,7 -1430,7 +1483,7 @@@ static int geneve_fill_info(struct sk_b } if (nla_put_u8(skb, IFLA_GENEVE_UDP_CSUM, - !!(geneve->flags & GENEVE_F_UDP_CSUM)) || + !(geneve->flags & GENEVE_F_UDP_ZERO_CSUM_TX)) || nla_put_u8(skb, IFLA_GENEVE_UDP_ZERO_CSUM6_TX, !!(geneve->flags & GENEVE_F_UDP_ZERO_CSUM6_TX)) || nla_put_u8(skb, IFLA_GENEVE_UDP_ZERO_CSUM6_RX, @@@ -1494,12 -1470,23 +1523,23 @@@ struct net_device *geneve_dev_create_fb return dev; err = geneve_configure(net, dev, &geneve_remote_unspec, - 0, 0, 0, htons(dst_port), true, 0); - if (err) { - free_netdev(dev); - return ERR_PTR(err); - } + 0, 0, 0, htons(dst_port), true, + GENEVE_F_UDP_ZERO_CSUM6_RX); + if (err) + goto err; + + /* openvswitch users expect packet sizes to be unrestricted, + * so set the largest MTU we can. + */ + err = __geneve_change_mtu(dev, IP_MAX_MTU, false); + if (err) + goto err; + return dev; + + err: + free_netdev(dev); + return ERR_PTR(err); } EXPORT_SYMBOL_GPL(geneve_dev_create_fb); diff --combined drivers/net/hyperv/netvsc_drv.c index 202e2b1,98e34fe..2b6595e --- a/drivers/net/hyperv/netvsc_drv.c +++ b/drivers/net/hyperv/netvsc_drv.c @@@ -43,11 -43,6 +43,11 @@@ #define RING_SIZE_MIN 64 #define LINKCHANGE_INT (2 * HZ) +#define NETVSC_HW_FEATURES (NETIF_F_RXCSUM | \ + NETIF_F_SG | \ + NETIF_F_TSO | \ + NETIF_F_TSO6 | \ + NETIF_F_HW_CSUM) static int ring_size = 128; module_param(ring_size, int, S_IRUGO); MODULE_PARM_DESC(ring_size, "Ring buffer size (# of pages)"); @@@ -550,8 -545,6 +550,8 @@@ do_send packet->page_buf_cnt = init_page_array(rndis_msg, rndis_msg_size, skb, packet, &pb); + /* timestamp packet in software */ + skb_tx_timestamp(skb); ret = netvsc_send(net_device_ctx->device_ctx, packet, rndis_msg, &pb, skb); @@@ -922,7 -915,6 +922,7 @@@ static const struct ethtool_ops ethtool .get_link = ethtool_op_get_link, .get_channels = netvsc_get_channels, .set_channels = netvsc_set_channels, + .get_ts_info = ethtool_op_get_ts_info, }; static const struct net_device_ops device_ops = { @@@ -1089,12 -1081,17 +1089,15 @@@ static int netvsc_probe(struct hv_devic net->netdev_ops = &device_ops; - net->hw_features = NETIF_F_RXCSUM | NETIF_F_SG | NETIF_F_IP_CSUM | - NETIF_F_TSO; - net->features = NETIF_F_HW_VLAN_CTAG_TX | NETIF_F_SG | NETIF_F_RXCSUM | - NETIF_F_IP_CSUM | NETIF_F_TSO; + net->hw_features = NETVSC_HW_FEATURES; + net->features = NETVSC_HW_FEATURES | NETIF_F_HW_VLAN_CTAG_TX; net->ethtool_ops = &ethtool_ops; SET_NETDEV_DEV(net, &dev->device); + /* We always need headroom for rndis header */ + net->needed_headroom = RNDIS_AND_PPI_SIZE; + /* Notify the netvsc driver of the new device */ memset(&device_info, 0, sizeof(device_info)); device_info.ring_size = ring_size; diff --combined drivers/net/phy/bcm7xxx.c index 9b31104,db507e3..b881a7b1 --- a/drivers/net/phy/bcm7xxx.c +++ b/drivers/net/phy/bcm7xxx.c @@@ -24,7 -24,7 +24,7 @@@ #define MII_BCM7XXX_100TX_FALSE_CAR 0x13 #define MII_BCM7XXX_100TX_DISC 0x14 #define MII_BCM7XXX_AUX_MODE 0x1d -#define MII_BCM7XX_64CLK_MDIO BIT(12) +#define MII_BCM7XXX_64CLK_MDIO BIT(12) #define MII_BCM7XXX_TEST 0x1f #define MII_BCM7XXX_SHD_MODE_2 BIT(2) @@@ -247,13 -247,9 +247,9 @@@ static int bcm7xxx_config_init(struct p int ret; /* Enable 64 clock MDIO */ - phy_write(phydev, MII_BCM7XXX_AUX_MODE, MII_BCM7XX_64CLK_MDIO); + phy_write(phydev, MII_BCM7XXX_AUX_MODE, MII_BCM7XXX_64CLK_MDIO); phy_read(phydev, MII_BCM7XXX_AUX_MODE); - /* Workaround only required for 100Mbits/sec capable PHYs */ - if (phydev->supported & PHY_GBIT_FEATURES) - return 0; - /* set shadow mode 2 */ ret = phy_set_clr_bits(phydev, MII_BCM7XXX_TEST, MII_BCM7XXX_SHD_MODE_2, MII_BCM7XXX_SHD_MODE_2); @@@ -270,7 -266,7 +266,7 @@@ phy_write(phydev, MII_BCM7XXX_100TX_FALSE_CAR, 0x7555); /* reset shadow mode 2 */ - ret = phy_set_clr_bits(phydev, MII_BCM7XXX_TEST, MII_BCM7XXX_SHD_MODE_2, 0); + ret = phy_set_clr_bits(phydev, MII_BCM7XXX_TEST, 0, MII_BCM7XXX_SHD_MODE_2); if (ret < 0) return ret; @@@ -307,11 -303,6 +303,6 @@@ static int bcm7xxx_suspend(struct phy_d return 0; } - static int bcm7xxx_dummy_config_init(struct phy_device *phydev) - { - return 0; - } - #define BCM7XXX_28NM_GPHY(_oui, _name) \ { \ .phy_id = (_oui), \ @@@ -326,21 -317,6 +317,21 @@@ .resume = bcm7xxx_28nm_resume, \ } +#define BCM7XXX_40NM_EPHY(_oui, _name) \ +{ \ + .phy_id = (_oui), \ + .phy_id_mask = 0xfffffff0, \ + .name = _name, \ + .features = PHY_BASIC_FEATURES | \ + SUPPORTED_Pause | SUPPORTED_Asym_Pause, \ + .flags = PHY_IS_INTERNAL, \ + .config_init = bcm7xxx_config_init, \ + .config_aneg = genphy_config_aneg, \ + .read_status = genphy_read_status, \ + .suspend = bcm7xxx_suspend, \ + .resume = bcm7xxx_config_init, \ +} + static struct phy_driver bcm7xxx_driver[] = { BCM7XXX_28NM_GPHY(PHY_ID_BCM7250, "Broadcom BCM7250"), BCM7XXX_28NM_GPHY(PHY_ID_BCM7364, "Broadcom BCM7364"), @@@ -348,34 -324,43 +339,10 @@@ BCM7XXX_28NM_GPHY(PHY_ID_BCM7439, "Broadcom BCM7439"), BCM7XXX_28NM_GPHY(PHY_ID_BCM7439_2, "Broadcom BCM7439 (2)"), BCM7XXX_28NM_GPHY(PHY_ID_BCM7445, "Broadcom BCM7445"), -{ - .phy_id = PHY_ID_BCM7425, - .phy_id_mask = 0xfffffff0, - .name = "Broadcom BCM7425", - .features = PHY_BASIC_FEATURES | - SUPPORTED_Pause | SUPPORTED_Asym_Pause, - .flags = PHY_IS_INTERNAL, - .config_init = bcm7xxx_config_init, - .config_aneg = genphy_config_aneg, - .read_status = genphy_read_status, - .suspend = bcm7xxx_suspend, - .resume = bcm7xxx_config_init, -}, { - .phy_id = PHY_ID_BCM7429, - .phy_id_mask = 0xfffffff0, - .name = "Broadcom BCM7429", - .features = PHY_BASIC_FEATURES | - SUPPORTED_Pause | SUPPORTED_Asym_Pause, - .flags = PHY_IS_INTERNAL, - .config_init = bcm7xxx_config_init, - .config_aneg = genphy_config_aneg, - .read_status = genphy_read_status, - .suspend = bcm7xxx_suspend, - .resume = bcm7xxx_config_init, -}, { - .phy_id = PHY_ID_BCM7435, - .phy_id_mask = 0xfffffff0, - .name = "Broadcom BCM7435", - .features = PHY_BASIC_FEATURES | - SUPPORTED_Pause | SUPPORTED_Asym_Pause, - .flags = PHY_IS_INTERNAL, - .config_init = bcm7xxx_config_init, - .config_aneg = genphy_config_aneg, - .read_status = genphy_read_status, - .suspend = bcm7xxx_suspend, - .resume = bcm7xxx_config_init, -} }; + BCM7XXX_40NM_EPHY(PHY_ID_BCM7425, "Broadcom BCM7425"), + BCM7XXX_40NM_EPHY(PHY_ID_BCM7429, "Broadcom BCM7429"), + BCM7XXX_40NM_EPHY(PHY_ID_BCM7435, "Broadcom BCM7435"), - { - .phy_id = PHY_BCM_OUI_4, - .phy_id_mask = 0xffff0000, - .name = "Broadcom BCM7XXX 40nm", - .features = PHY_GBIT_FEATURES | - SUPPORTED_Pause | SUPPORTED_Asym_Pause, - .flags = PHY_IS_INTERNAL, - .config_init = bcm7xxx_config_init, - .config_aneg = genphy_config_aneg, - .read_status = genphy_read_status, - .suspend = bcm7xxx_suspend, - .resume = bcm7xxx_config_init, - }, { - .phy_id = PHY_BCM_OUI_5, - .phy_id_mask = 0xffffff00, - .name = "Broadcom BCM7XXX 65nm", - .features = PHY_BASIC_FEATURES | - SUPPORTED_Pause | SUPPORTED_Asym_Pause, - .flags = PHY_IS_INTERNAL, - .config_init = bcm7xxx_dummy_config_init, - .config_aneg = genphy_config_aneg, - .read_status = genphy_read_status, - .suspend = bcm7xxx_suspend, - .resume = bcm7xxx_config_init, - } }; ++}; static struct mdio_device_id __maybe_unused bcm7xxx_tbl[] = { { PHY_ID_BCM7250, 0xfffffff0, }, @@@ -386,8 -371,6 +353,6 @@@ { PHY_ID_BCM7439, 0xfffffff0, }, { PHY_ID_BCM7435, 0xfffffff0, }, { PHY_ID_BCM7445, 0xfffffff0, }, - { PHY_BCM_OUI_4, 0xffff0000 }, - { PHY_BCM_OUI_5, 0xffffff00 }, { } }; diff --combined drivers/net/phy/marvell.c index d0168f1,ab1d0fc..9fb9d80 --- a/drivers/net/phy/marvell.c +++ b/drivers/net/phy/marvell.c @@@ -133,11 -133,6 +133,11 @@@ #define MII_88E3016_DISABLE_SCRAMBLER 0x0200 #define MII_88E3016_AUTO_MDIX_CROSSOVER 0x0030 +#define MII_88E1510_GEN_CTRL_REG_1 0x14 +#define MII_88E1510_GEN_CTRL_REG_1_MODE_MASK 0x7 +#define MII_88E1510_GEN_CTRL_REG_1_MODE_SGMII 0x1 /* SGMII to copper */ +#define MII_88E1510_GEN_CTRL_REG_1_RESET 0x8000 /* Soft reset */ + MODULE_DESCRIPTION("Marvell PHY driver"); MODULE_AUTHOR("Andy Fleming"); MODULE_LICENSE("GPL"); @@@ -443,41 -438,6 +443,41 @@@ static int m88e1318_config_aneg(struct return m88e1121_config_aneg(phydev); } +static int m88e1510_config_init(struct phy_device *phydev) +{ + int err; + int temp; + + /* SGMII-to-Copper mode initialization */ + if (phydev->interface == PHY_INTERFACE_MODE_SGMII) { + /* Select page 18 */ + err = phy_write(phydev, MII_MARVELL_PHY_PAGE, 18); + if (err < 0) + return err; + + /* In reg 20, write MODE[2:0] = 0x1 (SGMII to Copper) */ + temp = phy_read(phydev, MII_88E1510_GEN_CTRL_REG_1); + temp &= ~MII_88E1510_GEN_CTRL_REG_1_MODE_MASK; + temp |= MII_88E1510_GEN_CTRL_REG_1_MODE_SGMII; + err = phy_write(phydev, MII_88E1510_GEN_CTRL_REG_1, temp); + if (err < 0) + return err; + + /* PHY reset is necessary after changing MODE[2:0] */ + temp |= MII_88E1510_GEN_CTRL_REG_1_RESET; + err = phy_write(phydev, MII_88E1510_GEN_CTRL_REG_1, temp); + if (err < 0) + return err; + + /* Reset page selection */ + err = phy_write(phydev, MII_MARVELL_PHY_PAGE, 0); + if (err < 0) + return err; + } + + return 0; +} + static int m88e1510_config_aneg(struct phy_device *phydev) { int err; @@@ -486,6 -446,12 +486,12 @@@ if (err < 0) return err; + return 0; + } + + static int marvell_config_init(struct phy_device *phydev) + { + /* Set registers from marvell,reg-init DT property */ return marvell_of_reg_init(phydev); } @@@ -535,7 -501,7 +541,7 @@@ static int m88e1116r_config_init(struc mdelay(500); - return 0; + return marvell_config_init(phydev); } static int m88e3016_config_init(struct phy_device *phydev) @@@ -554,7 -520,7 +560,7 @@@ if (reg < 0) return reg; - return 0; + return marvell_config_init(phydev); } static int m88e1111_config_init(struct phy_device *phydev) @@@ -1065,8 -1031,8 +1071,8 @@@ static u64 marvell_get_stat(struct phy_ { struct marvell_hw_stat stat = marvell_hw_stats[i]; struct marvell_priv *priv = phydev->priv; - int err, oldpage; - u64 val; + int err, oldpage, val; + u64 ret; oldpage = phy_read(phydev, MII_MARVELL_PHY_PAGE); err = phy_write(phydev, MII_MARVELL_PHY_PAGE, @@@ -1076,16 -1042,16 +1082,16 @@@ val = phy_read(phydev, stat.reg); if (val < 0) { - val = UINT64_MAX; + ret = UINT64_MAX; } else { val = val & ((1 << stat.bits) - 1); priv->stats[i] += val; - val = priv->stats[i]; + ret = priv->stats[i]; } phy_write(phydev, MII_MARVELL_PHY_PAGE, oldpage); - return val; + return ret; } static void marvell_get_stats(struct phy_device *phydev, @@@ -1118,6 -1084,7 +1124,7 @@@ static struct phy_driver marvell_driver .features = PHY_GBIT_FEATURES, .probe = marvell_probe, .flags = PHY_HAS_INTERRUPT, + .config_init = &marvell_config_init, .config_aneg = &marvell_config_aneg, .read_status = &genphy_read_status, .ack_interrupt = &marvell_ack_interrupt, @@@ -1189,6 -1156,7 +1196,7 @@@ .features = PHY_GBIT_FEATURES, .flags = PHY_HAS_INTERRUPT, .probe = marvell_probe, + .config_init = &marvell_config_init, .config_aneg = &m88e1121_config_aneg, .read_status = &marvell_read_status, .ack_interrupt = &marvell_ack_interrupt, @@@ -1207,6 -1175,7 +1215,7 @@@ .features = PHY_GBIT_FEATURES, .flags = PHY_HAS_INTERRUPT, .probe = marvell_probe, + .config_init = &marvell_config_init, .config_aneg = &m88e1318_config_aneg, .read_status = &marvell_read_status, .ack_interrupt = &marvell_ack_interrupt, @@@ -1299,7 -1268,7 +1308,7 @@@ .features = PHY_GBIT_FEATURES, .flags = PHY_HAS_INTERRUPT, .probe = marvell_probe, - .config_init = &marvell_config_init, + .config_init = &m88e1510_config_init, .config_aneg = &m88e1510_config_aneg, .read_status = &marvell_read_status, .ack_interrupt = &marvell_ack_interrupt, @@@ -1318,6 -1287,7 +1327,7 @@@ .features = PHY_GBIT_FEATURES, .flags = PHY_HAS_INTERRUPT, .probe = marvell_probe, + .config_init = &marvell_config_init, .config_aneg = &m88e1510_config_aneg, .read_status = &marvell_read_status, .ack_interrupt = &marvell_ack_interrupt, diff --combined drivers/net/vxlan.c index 2ddc642,e6944b2..909f793 --- a/drivers/net/vxlan.c +++ b/drivers/net/vxlan.c @@@ -42,7 -42,7 +42,7 @@@ #include <net/netns/generic.h> #include <net/vxlan.h> #include <net/protocol.h> -#include <net/udp_tunnel.h> + #if IS_ENABLED(CONFIG_IPV6) #include <net/ipv6.h> #include <net/addrconf.h> @@@ -197,9 -197,9 +197,9 @@@ static int vxlan_nla_put_addr(struct sk #endif /* Virtual Network hash table head */ -static inline struct hlist_head *vni_head(struct vxlan_sock *vs, u32 id) +static inline struct hlist_head *vni_head(struct vxlan_sock *vs, __be32 vni) { - return &vs->vni_list[hash_32(id, VNI_HASH_BITS)]; + return &vs->vni_list[hash_32((__force u32)vni, VNI_HASH_BITS)]; } /* Socket hash table head */ @@@ -242,16 -242,12 +242,16 @@@ static struct vxlan_sock *vxlan_find_so return NULL; } -static struct vxlan_dev *vxlan_vs_find_vni(struct vxlan_sock *vs, u32 id) +static struct vxlan_dev *vxlan_vs_find_vni(struct vxlan_sock *vs, __be32 vni) { struct vxlan_dev *vxlan; - hlist_for_each_entry_rcu(vxlan, vni_head(vs, id), hlist) { - if (vxlan->default_dst.remote_vni == id) + /* For flow based devices, map all packets to VNI 0 */ + if (vs->flags & VXLAN_F_COLLECT_METADATA) + vni = 0; + + hlist_for_each_entry_rcu(vxlan, vni_head(vs, vni), hlist) { + if (vxlan->default_dst.remote_vni == vni) return vxlan; } @@@ -259,7 -255,7 +259,7 @@@ } /* Look up VNI in a per net namespace table */ -static struct vxlan_dev *vxlan_find_vni(struct net *net, u32 id, +static struct vxlan_dev *vxlan_find_vni(struct net *net, __be32 vni, sa_family_t family, __be16 port, u32 flags) { @@@ -269,7 -265,7 +269,7 @@@ if (!vs) return NULL; - return vxlan_vs_find_vni(vs, id); + return vxlan_vs_find_vni(vs, vni); } /* Fill in neighbour message in skbuff. */ @@@ -319,7 -315,7 +319,7 @@@ static int vxlan_fdb_info(struct sk_buf nla_put_be16(skb, NDA_PORT, rdst->remote_port)) goto nla_put_failure; if (rdst->remote_vni != vxlan->default_dst.remote_vni && - nla_put_u32(skb, NDA_VNI, rdst->remote_vni)) + nla_put_u32(skb, NDA_VNI, be32_to_cpu(rdst->remote_vni))) goto nla_put_failure; if (rdst->remote_ifindex && nla_put_u32(skb, NDA_IFINDEX, rdst->remote_ifindex)) @@@ -387,7 -383,7 +387,7 @@@ static void vxlan_ip_miss(struct net_de }; struct vxlan_rdst remote = { .remote_ip = *ipa, /* goes to NDA_DST */ - .remote_vni = VXLAN_N_VID, + .remote_vni = cpu_to_be32(VXLAN_N_VID), }; vxlan_fdb_notify(vxlan, &f, &remote, RTM_GETNEIGH); @@@ -456,7 -452,7 +456,7 @@@ static struct vxlan_fdb *vxlan_find_mac /* caller should hold vxlan->hash_lock */ static struct vxlan_rdst *vxlan_fdb_find_rdst(struct vxlan_fdb *f, union vxlan_addr *ip, __be16 port, - __u32 vni, __u32 ifindex) + __be32 vni, __u32 ifindex) { struct vxlan_rdst *rd; @@@ -473,8 -469,7 +473,8 @@@ /* Replace destination of unicast mac */ static int vxlan_fdb_replace(struct vxlan_fdb *f, - union vxlan_addr *ip, __be16 port, __u32 vni, __u32 ifindex) + union vxlan_addr *ip, __be16 port, __be32 vni, + __u32 ifindex) { struct vxlan_rdst *rd; @@@ -485,8 -480,6 +485,8 @@@ rd = list_first_entry_or_null(&f->remotes, struct vxlan_rdst, list); if (!rd) return 0; + + dst_cache_reset(&rd->dst_cache); rd->remote_ip = *ip; rd->remote_port = port; rd->remote_vni = vni; @@@ -496,7 -489,7 +496,7 @@@ /* Add/update destinations for multicast */ static int vxlan_fdb_append(struct vxlan_fdb *f, - union vxlan_addr *ip, __be16 port, __u32 vni, + union vxlan_addr *ip, __be16 port, __be32 vni, __u32 ifindex, struct vxlan_rdst **rdp) { struct vxlan_rdst *rd; @@@ -508,12 -501,6 +508,12 @@@ rd = kmalloc(sizeof(*rd), GFP_ATOMIC); if (rd == NULL) return -ENOBUFS; + + if (dst_cache_init(&rd->dst_cache, GFP_ATOMIC)) { + kfree(rd); + return -ENOBUFS; + } + rd->remote_ip = *ip; rd->remote_port = port; rd->remote_vni = vni; @@@ -528,8 -515,7 +528,8 @@@ static struct vxlanhdr *vxlan_gro_remcsum(struct sk_buff *skb, unsigned int off, struct vxlanhdr *vh, size_t hdrlen, - u32 data, struct gro_remcsum *grc, + __be32 vni_field, + struct gro_remcsum *grc, bool nopartial) { size_t start, offset; @@@ -540,8 -526,10 +540,8 @@@ if (!NAPI_GRO_CB(skb)->csum_valid) return NULL; - start = (data & VXLAN_RCO_MASK) << VXLAN_RCO_SHIFT; - offset = start + ((data & VXLAN_RCO_UDP) ? - offsetof(struct udphdr, check) : - offsetof(struct tcphdr, check)); + start = vxlan_rco_start(vni_field); + offset = start + vxlan_rco_offset(vni_field); vh = skb_gro_remcsum_process(skb, (void *)vh, off, hdrlen, start, offset, grc, nopartial); @@@ -561,7 -549,7 +561,7 @@@ static struct sk_buff **vxlan_gro_recei int flush = 1; struct vxlan_sock *vs = container_of(uoff, struct vxlan_sock, udp_offloads); - u32 flags; + __be32 flags; struct gro_remcsum grc; skb_gro_remcsum_init(&grc); @@@ -577,11 -565,11 +577,11 @@@ skb_gro_postpull_rcsum(skb, vh, sizeof(struct vxlanhdr)); - flags = ntohl(vh->vx_flags); + flags = vh->vx_flags; if ((flags & VXLAN_HF_RCO) && (vs->flags & VXLAN_F_REMCSUM_RX)) { vh = vxlan_gro_remcsum(skb, off_vx, vh, sizeof(struct vxlanhdr), - ntohl(vh->vx_vni), &grc, + vh->vx_vni, &grc, !!(vs->flags & VXLAN_F_REMCSUM_NOPARTIAL)); @@@ -672,7 -660,7 +672,7 @@@ static void vxlan_notify_del_rx_port(st static int vxlan_fdb_create(struct vxlan_dev *vxlan, const u8 *mac, union vxlan_addr *ip, __u16 state, __u16 flags, - __be16 port, __u32 vni, __u32 ifindex, + __be16 port, __be32 vni, __u32 ifindex, __u8 ndm_flags) { struct vxlan_rdst *rd = NULL; @@@ -761,10 -749,8 +761,10 @@@ static void vxlan_fdb_free(struct rcu_h struct vxlan_fdb *f = container_of(head, struct vxlan_fdb, rcu); struct vxlan_rdst *rd, *nd; - list_for_each_entry_safe(rd, nd, &f->remotes, list) + list_for_each_entry_safe(rd, nd, &f->remotes, list) { + dst_cache_destroy(&rd->dst_cache); kfree(rd); + } kfree(f); } @@@ -781,8 -767,7 +781,8 @@@ static void vxlan_fdb_destroy(struct vx } static int vxlan_fdb_parse(struct nlattr *tb[], struct vxlan_dev *vxlan, - union vxlan_addr *ip, __be16 *port, u32 *vni, u32 *ifindex) + union vxlan_addr *ip, __be16 *port, __be32 *vni, + u32 *ifindex) { struct net *net = dev_net(vxlan->dev); int err; @@@ -815,7 -800,7 +815,7 @@@ if (tb[NDA_VNI]) { if (nla_len(tb[NDA_VNI]) != sizeof(u32)) return -EINVAL; - *vni = nla_get_u32(tb[NDA_VNI]); + *vni = cpu_to_be32(nla_get_u32(tb[NDA_VNI])); } else { *vni = vxlan->default_dst.remote_vni; } @@@ -845,8 -830,7 +845,8 @@@ static int vxlan_fdb_add(struct ndmsg * /* struct net *net = dev_net(vxlan->dev); */ union vxlan_addr ip; __be16 port; - u32 vni, ifindex; + __be32 vni; + u32 ifindex; int err; if (!(ndm->ndm_state & (NUD_PERMANENT|NUD_REACHABLE))) { @@@ -883,8 -867,7 +883,8 @@@ static int vxlan_fdb_delete(struct ndms struct vxlan_rdst *rd = NULL; union vxlan_addr ip; __be16 port; - u32 vni, ifindex; + __be32 vni; + u32 ifindex; int err; err = vxlan_fdb_parse(tb, vxlan, &ip, &port, &vni, &ifindex); @@@ -1139,65 -1122,54 +1139,65 @@@ static int vxlan_igmp_leave(struct vxla return ret; } -static struct vxlanhdr *vxlan_remcsum(struct sk_buff *skb, struct vxlanhdr *vh, - size_t hdrlen, u32 data, bool nopartial) +static bool vxlan_remcsum(struct vxlanhdr *unparsed, + struct sk_buff *skb, u32 vxflags) { size_t start, offset, plen; - if (skb->remcsum_offload) - return vh; + if (!(unparsed->vx_flags & VXLAN_HF_RCO) || skb->remcsum_offload) + goto out; - start = (data & VXLAN_RCO_MASK) << VXLAN_RCO_SHIFT; - offset = start + ((data & VXLAN_RCO_UDP) ? - offsetof(struct udphdr, check) : - offsetof(struct tcphdr, check)); + start = vxlan_rco_start(unparsed->vx_vni); + offset = start + vxlan_rco_offset(unparsed->vx_vni); - plen = hdrlen + offset + sizeof(u16); + plen = sizeof(struct vxlanhdr) + offset + sizeof(u16); if (!pskb_may_pull(skb, plen)) - return NULL; + return false; - vh = (struct vxlanhdr *)(udp_hdr(skb) + 1); + skb_remcsum_process(skb, (void *)(vxlan_hdr(skb) + 1), start, offset, + !!(vxflags & VXLAN_F_REMCSUM_NOPARTIAL)); +out: + unparsed->vx_flags &= ~VXLAN_HF_RCO; + unparsed->vx_vni &= VXLAN_VNI_MASK; + return true; +} - skb_remcsum_process(skb, (void *)vh + hdrlen, start, offset, - nopartial); +static void vxlan_parse_gbp_hdr(struct vxlanhdr *unparsed, + struct vxlan_metadata *md, + struct metadata_dst *tun_dst) +{ + struct vxlanhdr_gbp *gbp = (struct vxlanhdr_gbp *)unparsed; - return vh; + if (!(unparsed->vx_flags & VXLAN_HF_GBP)) + goto out; + + md->gbp = ntohs(gbp->policy_id); + + if (tun_dst) + tun_dst->u.tun_info.key.tun_flags |= TUNNEL_VXLAN_OPT; + + if (gbp->dont_learn) + md->gbp |= VXLAN_GBP_DONT_LEARN; + + if (gbp->policy_applied) + md->gbp |= VXLAN_GBP_POLICY_APPLIED; + +out: + unparsed->vx_flags &= ~VXLAN_GBP_USED_BITS; } -static void vxlan_rcv(struct vxlan_sock *vs, struct sk_buff *skb, - struct vxlan_metadata *md, u32 vni, +static void vxlan_rcv(struct vxlan_dev *vxlan, struct vxlan_sock *vs, + struct sk_buff *skb, struct vxlan_metadata *md, struct metadata_dst *tun_dst) { struct iphdr *oip = NULL; struct ipv6hdr *oip6 = NULL; - struct vxlan_dev *vxlan; struct pcpu_sw_netstats *stats; union vxlan_addr saddr; int err = 0; - /* For flow based devices, map all packets to VNI 0 */ - if (vs->flags & VXLAN_F_COLLECT_METADATA) - vni = 0; - - /* Is this VNI defined? */ - vxlan = vxlan_vs_find_vni(vs, vni); - if (!vxlan) - goto drop; - skb_reset_mac_header(skb); - skb_scrub_packet(skb, !net_eq(vxlan->net, dev_net(vxlan->dev))); skb->protocol = eth_type_trans(skb, vxlan->dev); skb_postpull_rcsum(skb, eth_hdr(skb), ETH_HLEN); @@@ -1274,45 -1246,48 +1274,45 @@@ drop static int vxlan_udp_encap_recv(struct sock *sk, struct sk_buff *skb) { struct metadata_dst *tun_dst = NULL; + struct vxlan_dev *vxlan; struct vxlan_sock *vs; - struct vxlanhdr *vxh; - u32 flags, vni; + struct vxlanhdr unparsed; struct vxlan_metadata _md; struct vxlan_metadata *md = &_md; /* Need Vxlan and inner Ethernet header to be present */ if (!pskb_may_pull(skb, VXLAN_HLEN)) - goto error; - - vxh = (struct vxlanhdr *)(udp_hdr(skb) + 1); - flags = ntohl(vxh->vx_flags); - vni = ntohl(vxh->vx_vni); + return 1; - if (flags & VXLAN_HF_VNI) { - flags &= ~VXLAN_HF_VNI; - } else { - /* VNI flag always required to be set */ - goto bad_flags; + unparsed = *vxlan_hdr(skb); + /* VNI flag always required to be set */ + if (!(unparsed.vx_flags & VXLAN_HF_VNI)) { + netdev_dbg(skb->dev, "invalid vxlan flags=%#x vni=%#x\n", + ntohl(vxlan_hdr(skb)->vx_flags), + ntohl(vxlan_hdr(skb)->vx_vni)); + /* Return non vxlan pkt */ + return 1; } - - if (iptunnel_pull_header(skb, VXLAN_HLEN, htons(ETH_P_TEB))) - goto drop; - vxh = (struct vxlanhdr *)(udp_hdr(skb) + 1); + unparsed.vx_flags &= ~VXLAN_HF_VNI; + unparsed.vx_vni &= ~VXLAN_VNI_MASK; vs = rcu_dereference_sk_user_data(sk); if (!vs) goto drop; - if ((flags & VXLAN_HF_RCO) && (vs->flags & VXLAN_F_REMCSUM_RX)) { - vxh = vxlan_remcsum(skb, vxh, sizeof(struct vxlanhdr), vni, - !!(vs->flags & VXLAN_F_REMCSUM_NOPARTIAL)); - if (!vxh) - goto drop; + vxlan = vxlan_vs_find_vni(vs, vxlan_vni(vxlan_hdr(skb)->vx_vni)); + if (!vxlan) + goto drop; - flags &= ~VXLAN_HF_RCO; - vni &= VXLAN_VNI_MASK; - } + if (iptunnel_pull_header(skb, VXLAN_HLEN, htons(ETH_P_TEB), + !net_eq(vxlan->net, dev_net(vxlan->dev)))) + goto drop; if (vxlan_collect_metadata(vs)) { + __be32 vni = vxlan_vni(vxlan_hdr(skb)->vx_vni); + tun_dst = udp_tun_rx_dst(skb, vxlan_get_sk_family(vs), TUNNEL_KEY, - cpu_to_be64(vni >> 8), sizeof(*md)); + vxlan_vni_to_tun_id(vni), sizeof(*md)); if (!tun_dst) goto drop; @@@ -1325,13 -1300,25 +1325,13 @@@ /* For backwards compatibility, only allow reserved fields to be * used by VXLAN extensions if explicitly requested. */ - if ((flags & VXLAN_HF_GBP) && (vs->flags & VXLAN_F_GBP)) { - struct vxlanhdr_gbp *gbp; - - gbp = (struct vxlanhdr_gbp *)vxh; - md->gbp = ntohs(gbp->policy_id); - - if (tun_dst) - tun_dst->u.tun_info.key.tun_flags |= TUNNEL_VXLAN_OPT; - - if (gbp->dont_learn) - md->gbp |= VXLAN_GBP_DONT_LEARN; - - if (gbp->policy_applied) - md->gbp |= VXLAN_GBP_POLICY_APPLIED; - - flags &= ~VXLAN_GBP_USED_BITS; - } + if (vs->flags & VXLAN_F_REMCSUM_RX) + if (!vxlan_remcsum(&unparsed, skb, vs->flags)) + goto drop; + if (vs->flags & VXLAN_F_GBP) + vxlan_parse_gbp_hdr(&unparsed, md, tun_dst); - if (flags || vni & ~VXLAN_VNI_MASK) { + if (unparsed.vx_flags || unparsed.vx_vni) { /* If there are any unprocessed flags remaining treat * this as a malformed packet. This behavior diverges from * VXLAN RFC (RFC7348) which stipulates that bits in reserved @@@ -1340,19 -1327,28 +1340,19 @@@ * is more robust and provides a little more security in * adding extensions to VXLAN. */ - - goto bad_flags; + goto drop; } - vxlan_rcv(vs, skb, md, vni >> 8, tun_dst); + vxlan_rcv(vxlan, vs, skb, md, tun_dst); return 0; drop: - /* Consume bad packet */ - kfree_skb(skb); - return 0; - -bad_flags: - netdev_dbg(skb->dev, "invalid vxlan flags=%#x vni=%#x\n", - ntohl(vxh->vx_flags), ntohl(vxh->vx_vni)); - -error: if (tun_dst) dst_release((struct dst_entry *)tun_dst); - /* Return non vxlan pkt */ - return 1; + /* Consume bad packet */ + kfree_skb(skb); + return 0; } static int arp_reduce(struct net_device *dev, struct sk_buff *skb) @@@ -1677,7 -1673,7 +1677,7 @@@ static void vxlan_build_gbp_hdr(struct return; gbp = (struct vxlanhdr_gbp *)vxh; - vxh->vx_flags |= htonl(VXLAN_HF_GBP); + vxh->vx_flags |= VXLAN_HF_GBP; if (md->gbp & VXLAN_GBP_DONT_LEARN) gbp->dont_learn = 1; @@@ -1688,15 -1684,20 +1688,15 @@@ gbp->policy_id = htons(md->gbp & VXLAN_GBP_ID_MASK); } -#if IS_ENABLED(CONFIG_IPV6) -static int vxlan6_xmit_skb(struct dst_entry *dst, struct sock *sk, - struct sk_buff *skb, - struct net_device *dev, struct in6_addr *saddr, - struct in6_addr *daddr, __u8 prio, __u8 ttl, - __be16 src_port, __be16 dst_port, __be32 vni, - struct vxlan_metadata *md, bool xnet, u32 vxflags) +static int vxlan_build_skb(struct sk_buff *skb, struct dst_entry *dst, + int iphdr_len, __be32 vni, + struct vxlan_metadata *md, u32 vxflags, + bool udp_sum) { struct vxlanhdr *vxh; int min_headroom; int err; - bool udp_sum = !(vxflags & VXLAN_F_UDP_ZERO_CSUM6_TX); int type = udp_sum ? SKB_GSO_UDP_TUNNEL_CSUM : SKB_GSO_UDP_TUNNEL; - u16 hdrlen = sizeof(struct vxlanhdr); if ((vxflags & VXLAN_F_REMCSUM_TX) && skb->ip_summed == CHECKSUM_PARTIAL) { @@@ -1705,39 -1706,50 +1705,39 @@@ if (csum_start <= VXLAN_MAX_REMCSUM_START && !(csum_start & VXLAN_RCO_SHIFT_MASK) && (skb->csum_offset == offsetof(struct udphdr, check) || - skb->csum_offset == offsetof(struct tcphdr, check))) { - udp_sum = false; + skb->csum_offset == offsetof(struct tcphdr, check))) type |= SKB_GSO_TUNNEL_REMCSUM; - } } - skb_scrub_packet(skb, xnet); - min_headroom = LL_RESERVED_SPACE(dst->dev) + dst->header_len - + VXLAN_HLEN + sizeof(struct ipv6hdr) + + VXLAN_HLEN + iphdr_len + (skb_vlan_tag_present(skb) ? VLAN_HLEN : 0); /* Need space for new headers (invalidates iph ptr) */ err = skb_cow_head(skb, min_headroom); if (unlikely(err)) { kfree_skb(skb); - goto err; + return err; } skb = vlan_hwaccel_push_inside(skb); - if (WARN_ON(!skb)) { - err = -ENOMEM; - goto err; - } + if (WARN_ON(!skb)) + return -ENOMEM; - skb = iptunnel_handle_offloads(skb, udp_sum, type); - if (IS_ERR(skb)) { - err = -EINVAL; - goto err; - } + skb = iptunnel_handle_offloads(skb, type); + if (IS_ERR(skb)) + return PTR_ERR(skb); vxh = (struct vxlanhdr *) __skb_push(skb, sizeof(*vxh)); - vxh->vx_flags = htonl(VXLAN_HF_VNI); - vxh->vx_vni = vni; + vxh->vx_flags = VXLAN_HF_VNI; + vxh->vx_vni = vxlan_vni_field(vni); if (type & SKB_GSO_TUNNEL_REMCSUM) { - u32 data = (skb_checksum_start_offset(skb) - hdrlen) >> - VXLAN_RCO_SHIFT; + unsigned int start; - if (skb->csum_offset == offsetof(struct udphdr, check)) - data |= VXLAN_RCO_UDP; - - vxh->vx_vni |= htonl(data); - vxh->vx_flags |= htonl(VXLAN_HF_RCO); + start = skb_checksum_start_offset(skb) - sizeof(struct vxlanhdr); + vxh->vx_vni |= vxlan_compute_rco(start, skb->csum_offset); + vxh->vx_flags |= VXLAN_HF_RCO; if (!skb_is_gso(skb)) { skb->ip_summed = CHECKSUM_NONE; @@@ -1749,63 -1761,102 +1749,63 @@@ vxlan_build_gbp_hdr(vxh, vxflags, md); skb_set_inner_protocol(skb, htons(ETH_P_TEB)); - - udp_tunnel6_xmit_skb(dst, sk, skb, dev, saddr, daddr, prio, - ttl, src_port, dst_port, - !!(vxflags & VXLAN_F_UDP_ZERO_CSUM6_TX)); return 0; -err: - dst_release(dst); - return err; } -#endif -static int vxlan_xmit_skb(struct rtable *rt, struct sock *sk, struct sk_buff *skb, - __be32 src, __be32 dst, __u8 tos, __u8 ttl, __be16 df, - __be16 src_port, __be16 dst_port, __be32 vni, - struct vxlan_metadata *md, bool xnet, u32 vxflags) +static struct rtable *vxlan_get_route(struct vxlan_dev *vxlan, + struct sk_buff *skb, int oif, u8 tos, + __be32 daddr, __be32 *saddr, + struct dst_cache *dst_cache, + struct ip_tunnel_info *info) { - struct vxlanhdr *vxh; - int min_headroom; - int err; - bool udp_sum = !!(vxflags & VXLAN_F_UDP_CSUM); - int type = udp_sum ? SKB_GSO_UDP_TUNNEL_CSUM : SKB_GSO_UDP_TUNNEL; - u16 hdrlen = sizeof(struct vxlanhdr); - - if ((vxflags & VXLAN_F_REMCSUM_TX) && - skb->ip_summed == CHECKSUM_PARTIAL) { - int csum_start = skb_checksum_start_offset(skb); - - if (csum_start <= VXLAN_MAX_REMCSUM_START && - !(csum_start & VXLAN_RCO_SHIFT_MASK) && - (skb->csum_offset == offsetof(struct udphdr, check) || - skb->csum_offset == offsetof(struct tcphdr, check))) { - udp_sum = false; - type |= SKB_GSO_TUNNEL_REMCSUM; - } - } - - min_headroom = LL_RESERVED_SPACE(rt->dst.dev) + rt->dst.header_len - + VXLAN_HLEN + sizeof(struct iphdr) - + (skb_vlan_tag_present(skb) ? VLAN_HLEN : 0); + struct rtable *rt = NULL; + bool use_cache = false; + struct flowi4 fl4; - /* Need space for new headers (invalidates iph ptr) */ - err = skb_cow_head(skb, min_headroom); - if (unlikely(err)) { - kfree_skb(skb); - return err; + /* when the ip_tunnel_info is availble, the tos used for lookup is + * packet independent, so we can use the cache + */ + if (!skb->mark && (!tos || info)) { + use_cache = true; + rt = dst_cache_get_ip4(dst_cache, saddr); + if (rt) + return rt; } - skb = vlan_hwaccel_push_inside(skb); - if (WARN_ON(!skb)) - return -ENOMEM; - - skb = iptunnel_handle_offloads(skb, udp_sum, type); - if (IS_ERR(skb)) - return PTR_ERR(skb); - - vxh = (struct vxlanhdr *) __skb_push(skb, sizeof(*vxh)); - vxh->vx_flags = htonl(VXLAN_HF_VNI); - vxh->vx_vni = vni; - - if (type & SKB_GSO_TUNNEL_REMCSUM) { - u32 data = (skb_checksum_start_offset(skb) - hdrlen) >> - VXLAN_RCO_SHIFT; - - if (skb->csum_offset == offsetof(struct udphdr, check)) - data |= VXLAN_RCO_UDP; - - vxh->vx_vni |= htonl(data); - vxh->vx_flags |= htonl(VXLAN_HF_RCO); + memset(&fl4, 0, sizeof(fl4)); + fl4.flowi4_oif = oif; + fl4.flowi4_tos = RT_TOS(tos); + fl4.flowi4_mark = skb->mark; + fl4.flowi4_proto = IPPROTO_UDP; + fl4.daddr = daddr; + fl4.saddr = vxlan->cfg.saddr.sin.sin_addr.s_addr; - if (!skb_is_gso(skb)) { - skb->ip_summed = CHECKSUM_NONE; - skb->encapsulation = 0; - } + rt = ip_route_output_key(vxlan->net, &fl4); + if (!IS_ERR(rt)) { + *saddr = fl4.saddr; + if (use_cache) + dst_cache_set_ip4(dst_cache, &rt->dst, fl4.saddr); } - - if (vxflags & VXLAN_F_GBP) - vxlan_build_gbp_hdr(vxh, vxflags, md); - - skb_set_inner_protocol(skb, htons(ETH_P_TEB)); - - udp_tunnel_xmit_skb(rt, sk, skb, src, dst, tos, ttl, df, - src_port, dst_port, xnet, - !(vxflags & VXLAN_F_UDP_CSUM)); - return 0; + return rt; } #if IS_ENABLED(CONFIG_IPV6) static struct dst_entry *vxlan6_get_route(struct vxlan_dev *vxlan, struct sk_buff *skb, int oif, const struct in6_addr *daddr, - struct in6_addr *saddr) + struct in6_addr *saddr, + struct dst_cache *dst_cache) { struct dst_entry *ndst; struct flowi6 fl6; int err; + if (!skb->mark) { + ndst = dst_cache_get_ip6(dst_cache, saddr); + if (ndst) + return ndst; + } + memset(&fl6, 0, sizeof(fl6)); fl6.flowi6_oif = oif; fl6.daddr = *daddr; @@@ -1820,8 -1871,6 +1820,8 @@@ return ERR_PTR(err); *saddr = fl6.saddr; + if (!skb->mark) + dst_cache_set_ip6(dst_cache, ndst, saddr); return ndst; } #endif @@@ -1874,24 -1923,22 +1874,24 @@@ static void vxlan_encap_bypass(struct s static void vxlan_xmit_one(struct sk_buff *skb, struct net_device *dev, struct vxlan_rdst *rdst, bool did_rsc) { + struct dst_cache *dst_cache; struct ip_tunnel_info *info; struct vxlan_dev *vxlan = netdev_priv(dev); struct sock *sk; struct rtable *rt = NULL; const struct iphdr *old_iph; - struct flowi4 fl4; union vxlan_addr *dst; union vxlan_addr remote_ip; struct vxlan_metadata _md; struct vxlan_metadata *md = &_md; __be16 src_port = 0, dst_port; - u32 vni; + __be32 vni; __be16 df = 0; __u8 tos, ttl; int err; u32 flags = vxlan->flags; + bool udp_sum = false; + bool xnet = !net_eq(vxlan->net, dev_net(vxlan->dev)); info = skb_tunnel_info(skb); @@@ -1899,7 -1946,6 +1899,7 @@@ dst_port = rdst->remote_port ? rdst->remote_port : vxlan->cfg.dst_port; vni = rdst->remote_vni; dst = &rdst->remote_ip; + dst_cache = &rdst->dst_cache; } else { if (!info) { WARN_ONCE(1, "%s: Missing encapsulation instructions\n", @@@ -1907,14 -1953,13 +1907,14 @@@ goto drop; } dst_port = info->key.tp_dst ? : vxlan->cfg.dst_port; - vni = be64_to_cpu(info->key.tun_id); + vni = vxlan_tun_id_to_vni(info->key.tun_id); remote_ip.sa.sa_family = ip_tunnel_info_af(info); if (remote_ip.sa.sa_family == AF_INET) remote_ip.sin.sin_addr.s_addr = info->key.u.ipv4.dst; else remote_ip.sin6.sin6_addr = info->key.u.ipv6.dst; dst = &remote_ip; + dst_cache = &info->dst_cache; } if (vxlan_addr_any(dst)) { @@@ -1942,7 -1987,6 +1942,7 @@@ if (info) { ttl = info->key.ttl; tos = info->key.tos; + udp_sum = !!(info->key.tun_flags & TUNNEL_CSUM); if (info->options_len) md = ip_tunnel_info_opts(info); @@@ -1951,16 -1995,29 +1951,16 @@@ } if (dst->sa.sa_family == AF_INET) { + __be32 saddr; + if (!vxlan->vn4_sock) goto drop; sk = vxlan->vn4_sock->sock->sk; - if (info) { - if (info->key.tun_flags & TUNNEL_DONT_FRAGMENT) - df = htons(IP_DF); - - if (info->key.tun_flags & TUNNEL_CSUM) - flags |= VXLAN_F_UDP_CSUM; - else - flags &= ~VXLAN_F_UDP_CSUM; - } - - memset(&fl4, 0, sizeof(fl4)); - fl4.flowi4_oif = rdst ? rdst->remote_ifindex : 0; - fl4.flowi4_tos = RT_TOS(tos); - fl4.flowi4_mark = skb->mark; - fl4.flowi4_proto = IPPROTO_UDP; - fl4.daddr = dst->sin.sin_addr.s_addr; - fl4.saddr = vxlan->cfg.saddr.sin.sin_addr.s_addr; - - rt = ip_route_output_key(vxlan->net, &fl4); + rt = vxlan_get_route(vxlan, skb, + rdst ? rdst->remote_ifindex : 0, tos, + dst->sin.sin_addr.s_addr, &saddr, + dst_cache, info); if (IS_ERR(rt)) { netdev_dbg(dev, "no route to %pI4\n", &dst->sin.sin_addr.s_addr); @@@ -1990,21 -2047,18 +1990,21 @@@ return; } + if (!info) + udp_sum = !(flags & VXLAN_F_UDP_ZERO_CSUM_TX); + else if (info->key.tun_flags & TUNNEL_DONT_FRAGMENT) + df = htons(IP_DF); + tos = ip_tunnel_ecn_encap(tos, old_iph, skb); ttl = ttl ? : ip4_dst_hoplimit(&rt->dst); - err = vxlan_xmit_skb(rt, sk, skb, fl4.saddr, - dst->sin.sin_addr.s_addr, tos, ttl, df, - src_port, dst_port, htonl(vni << 8), md, - !net_eq(vxlan->net, dev_net(vxlan->dev)), - flags); - if (err < 0) { - /* skb is already freed. */ - skb = NULL; - goto rt_tx_error; - } + err = vxlan_build_skb(skb, &rt->dst, sizeof(struct iphdr), + vni, md, flags, udp_sum); + if (err < 0) + goto xmit_tx_error; + + udp_tunnel_xmit_skb(rt, sk, skb, saddr, + dst->sin.sin_addr.s_addr, tos, ttl, df, + src_port, dst_port, xnet, !udp_sum); #if IS_ENABLED(CONFIG_IPV6) } else { struct dst_entry *ndst; @@@ -2017,8 -2071,7 +2017,8 @@@ ndst = vxlan6_get_route(vxlan, skb, rdst ? rdst->remote_ifindex : 0, - &dst->sin6.sin6_addr, &saddr); + &dst->sin6.sin6_addr, &saddr, + dst_cache); if (IS_ERR(ndst)) { netdev_dbg(dev, "no route to %pI6\n", &dst->sin6.sin6_addr); @@@ -2050,20 -2103,18 +2050,20 @@@ return; } - if (info) { - if (info->key.tun_flags & TUNNEL_CSUM) - flags &= ~VXLAN_F_UDP_ZERO_CSUM6_TX; - else - flags |= VXLAN_F_UDP_ZERO_CSUM6_TX; - } + if (!info) + udp_sum = !(flags & VXLAN_F_UDP_ZERO_CSUM6_TX); ttl = ttl ? : ip6_dst_hoplimit(ndst); - err = vxlan6_xmit_skb(ndst, sk, skb, dev, &saddr, &dst->sin6.sin6_addr, - 0, ttl, src_port, dst_port, htonl(vni << 8), md, - !net_eq(vxlan->net, dev_net(vxlan->dev)), - flags); + skb_scrub_packet(skb, xnet); + err = vxlan_build_skb(skb, ndst, sizeof(struct ipv6hdr), + vni, md, flags, udp_sum); + if (err < 0) { + dst_release(ndst); + return; + } + udp_tunnel6_xmit_skb(ndst, sk, skb, dev, + &saddr, &dst->sin6.sin6_addr, + 0, ttl, src_port, dst_port, !udp_sum); #endif } @@@ -2073,9 -2124,6 +2073,9 @@@ drop dev->stats.tx_dropped++; goto tx_free; +xmit_tx_error: + /* skb is already freed. */ + skb = NULL; rt_tx_error: ip_rt_put(rt); tx_error: @@@ -2123,9 -2171,11 +2123,11 @@@ static netdev_tx_t vxlan_xmit(struct sk #endif } - if (vxlan->flags & VXLAN_F_COLLECT_METADATA && - info && info->mode & IP_TUNNEL_INFO_TX) { - vxlan_xmit_one(skb, dev, NULL, false); + if (vxlan->flags & VXLAN_F_COLLECT_METADATA) { + if (info && info->mode & IP_TUNNEL_INFO_TX) + vxlan_xmit_one(skb, dev, NULL, false); + else + kfree_skb(skb); return NETDEV_TX_OK; } @@@ -2213,7 -2263,7 +2215,7 @@@ static void vxlan_cleanup(unsigned lon static void vxlan_vs_add_dev(struct vxlan_sock *vs, struct vxlan_dev *vxlan) { struct vxlan_net *vn = net_generic(vxlan->net, vxlan_net_id); - __u32 vni = vxlan->default_dst.remote_vni; + __be32 vni = vxlan->default_dst.remote_vni; spin_lock(&vn->sock_lock); hlist_add_head_rcu(&vxlan->hlist, vni_head(vs, vni)); @@@ -2319,29 -2369,68 +2321,43 @@@ static void vxlan_set_multicast_list(st { } - static int vxlan_change_mtu(struct net_device *dev, int new_mtu) + static int __vxlan_change_mtu(struct net_device *dev, + struct net_device *lowerdev, + struct vxlan_rdst *dst, int new_mtu, bool strict) { - struct vxlan_dev *vxlan = netdev_priv(dev); - struct vxlan_rdst *dst = &vxlan->default_dst; - struct net_device *lowerdev; - int max_mtu; + int max_mtu = IP_MAX_MTU; - lowerdev = __dev_get_by_index(vxlan->net, dst->remote_ifindex); - if (lowerdev == NULL) - return eth_change_mtu(dev, new_mtu); + if (lowerdev) + max_mtu = lowerdev->mtu; if (dst->remote_ip.sa.sa_family == AF_INET6) - max_mtu = lowerdev->mtu - VXLAN6_HEADROOM; + max_mtu -= VXLAN6_HEADROOM; else - max_mtu = lowerdev->mtu - VXLAN_HEADROOM; + max_mtu -= VXLAN_HEADROOM; - if (new_mtu < 68 || new_mtu > max_mtu) + if (new_mtu < 68) return -EINVAL; + if (new_mtu > max_mtu) { + if (strict) + return -EINVAL; + + new_mtu = max_mtu; + } + dev->mtu = new_mtu; return 0; } + static int vxlan_change_mtu(struct net_device *dev, int new_mtu) + { + struct vxlan_dev *vxlan = netdev_priv(dev); + struct vxlan_rdst *dst = &vxlan->default_dst; + struct net_device *lowerdev = __dev_get_by_index(vxlan->net, + dst->remote_ifindex); + return __vxlan_change_mtu(dev, lowerdev, dst, new_mtu, true); + } + -static int egress_ipv4_tun_info(struct net_device *dev, struct sk_buff *skb, - struct ip_tunnel_info *info, - __be16 sport, __be16 dport) -{ - struct vxlan_dev *vxlan = netdev_priv(dev); - struct rtable *rt; - struct flowi4 fl4; - - memset(&fl4, 0, sizeof(fl4)); - fl4.flowi4_tos = RT_TOS(info->key.tos); - fl4.flowi4_mark = skb->mark; - fl4.flowi4_proto = IPPROTO_UDP; - fl4.daddr = info->key.u.ipv4.dst; - - rt = ip_route_output_key(vxlan->net, &fl4); - if (IS_ERR(rt)) - return PTR_ERR(rt); - ip_rt_put(rt); - - info->key.u.ipv4.src = fl4.saddr; - info->key.tp_src = sport; - info->key.tp_dst = dport; - return 0; -} - static int vxlan_fill_metadata_dst(struct net_device *dev, struct sk_buff *skb) { struct vxlan_dev *vxlan = netdev_priv(dev); @@@ -2353,16 -2442,9 +2369,16 @@@ dport = info->key.tp_dst ? : vxlan->cfg.dst_port; if (ip_tunnel_info_af(info) == AF_INET) { + struct rtable *rt; + if (!vxlan->vn4_sock) return -EINVAL; - return egress_ipv4_tun_info(dev, skb, info, sport, dport); + rt = vxlan_get_route(vxlan, skb, 0, info->key.tos, + info->key.u.ipv4.dst, + &info->key.u.ipv4.src, NULL, info); + if (IS_ERR(rt)) + return PTR_ERR(rt); + ip_rt_put(rt); } else { #if IS_ENABLED(CONFIG_IPV6) struct dst_entry *ndst; @@@ -2371,16 -2453,17 +2387,16 @@@ return -EINVAL; ndst = vxlan6_get_route(vxlan, skb, 0, &info->key.u.ipv6.dst, - &info->key.u.ipv6.src); + &info->key.u.ipv6.src, NULL); if (IS_ERR(ndst)) return PTR_ERR(ndst); dst_release(ndst); - - info->key.tp_src = sport; - info->key.tp_dst = dport; #else /* !CONFIG_IPV6 */ return -EPFNOSUPPORT; #endif } + info->key.tp_src = sport; + info->key.tp_dst = dport; return 0; } @@@ -2456,6 -2539,7 +2472,7 @@@ static void vxlan_setup(struct net_devi dev->hw_features |= NETIF_F_GSO_SOFTWARE; dev->hw_features |= NETIF_F_HW_VLAN_CTAG_TX | NETIF_F_HW_VLAN_STAG_TX; netif_keep_dst(dev); + dev->priv_flags &= ~IFF_TX_SKB_SHARING; dev->priv_flags |= IFF_LIVE_ADDR_CHANGE | IFF_NO_QUEUE; INIT_LIST_HEAD(&vxlan->next); @@@ -2698,6 -2782,7 +2715,7 @@@ static int vxlan_dev_configure(struct n int err; bool use_ipv6 = false; __be16 default_port = vxlan->cfg.dst_port; + struct net_device *lowerdev = NULL; vxlan->net = src_net; @@@ -2718,9 -2803,7 +2736,7 @@@ } if (conf->remote_ifindex) { - struct net_device *lowerdev - = __dev_get_by_index(src_net, conf->remote_ifindex); - + lowerdev = __dev_get_by_index(src_net, conf->remote_ifindex); dst->remote_ifindex = conf->remote_ifindex; if (!lowerdev) { @@@ -2744,6 -2827,12 +2760,12 @@@ needed_headroom = lowerdev->hard_header_len; } + if (conf->mtu) { + err = __vxlan_change_mtu(dev, lowerdev, dst, conf->mtu, false); + if (err) + return err; + } + if (use_ipv6 || conf->flags & VXLAN_F_COLLECT_METADATA) needed_headroom += VXLAN6_HEADROOM; else @@@ -2828,7 -2917,7 +2850,7 @@@ static int vxlan_newlink(struct net *sr memset(&conf, 0, sizeof(conf)); if (data[IFLA_VXLAN_ID]) - conf.vni = nla_get_u32(data[IFLA_VXLAN_ID]); + conf.vni = cpu_to_be32(nla_get_u32(data[IFLA_VXLAN_ID])); if (data[IFLA_VXLAN_GROUP]) { conf.remote_ip.sin.sin_addr.s_addr = nla_get_in_addr(data[IFLA_VXLAN_GROUP]); @@@ -2896,9 -2985,8 +2918,9 @@@ if (data[IFLA_VXLAN_PORT]) conf.dst_port = nla_get_be16(data[IFLA_VXLAN_PORT]); - if (data[IFLA_VXLAN_UDP_CSUM] && nla_get_u8(data[IFLA_VXLAN_UDP_CSUM])) - conf.flags |= VXLAN_F_UDP_CSUM; + if (data[IFLA_VXLAN_UDP_CSUM] && + !nla_get_u8(data[IFLA_VXLAN_UDP_CSUM])) + conf.flags |= VXLAN_F_UDP_ZERO_CSUM_TX; if (data[IFLA_VXLAN_UDP_ZERO_CSUM6_TX] && nla_get_u8(data[IFLA_VXLAN_UDP_ZERO_CSUM6_TX])) @@@ -2933,7 -3021,7 +2955,7 @@@ break; case -EEXIST: - pr_info("duplicate VNI %u\n", conf.vni); + pr_info("duplicate VNI %u\n", be32_to_cpu(conf.vni)); break; } @@@ -2991,7 -3079,7 +3013,7 @@@ static int vxlan_fill_info(struct sk_bu .high = htons(vxlan->cfg.port_max), }; - if (nla_put_u32(skb, IFLA_VXLAN_ID, dst->remote_vni)) + if (nla_put_u32(skb, IFLA_VXLAN_ID, be32_to_cpu(dst->remote_vni))) goto nla_put_failure; if (!vxlan_addr_any(&dst->remote_ip)) { @@@ -3042,7 -3130,7 +3064,7 @@@ nla_put_u32(skb, IFLA_VXLAN_LIMIT, vxlan->cfg.addrmax) || nla_put_be16(skb, IFLA_VXLAN_PORT, vxlan->cfg.dst_port) || nla_put_u8(skb, IFLA_VXLAN_UDP_CSUM, - !!(vxlan->flags & VXLAN_F_UDP_CSUM)) || + !(vxlan->flags & VXLAN_F_UDP_ZERO_CSUM_TX)) || nla_put_u8(skb, IFLA_VXLAN_UDP_ZERO_CSUM6_TX, !!(vxlan->flags & VXLAN_F_UDP_ZERO_CSUM6_TX)) || nla_put_u8(skb, IFLA_VXLAN_UDP_ZERO_CSUM6_RX, diff --combined drivers/net/wireless/intel/iwlwifi/Kconfig index 11932d5,7438fbe..16c4f38 --- a/drivers/net/wireless/intel/iwlwifi/Kconfig +++ b/drivers/net/wireless/intel/iwlwifi/Kconfig @@@ -53,7 -53,6 +53,6 @@@ config IWLWIFI_LED config IWLDVM tristate "Intel Wireless WiFi DVM Firmware support" - depends on m help This is the driver that supports the DVM firmware. The list of the devices that use this firmware is available here: @@@ -99,18 -98,6 +98,18 @@@ config IWLWIFI_UAPS If unsure, say N. +config IWLWIFI_PCIE_RTPM + bool "Enable runtime power management mode for PCIe devices" + depends on IWLMVM && PM + default false + help + Say Y here to enable runtime power management for PCIe + devices. If enabled, the device will go into low power mode + when idle for a short period of time, allowing for improved + power saving during runtime. + + If unsure, say N. + menu "Debugging Options" config IWLWIFI_DEBUG diff --combined drivers/net/wireless/intel/iwlwifi/mvm/scan.c index 1e1ab9d,ea1e177..aa6d807 --- a/drivers/net/wireless/intel/iwlwifi/mvm/scan.c +++ b/drivers/net/wireless/intel/iwlwifi/mvm/scan.c @@@ -930,11 -930,8 +930,11 @@@ int iwl_mvm_config_scan(struct iwl_mvm if (WARN_ON(num_channels > mvm->fw->ucode_capa.n_scan_channels)) return -ENOBUFS; - if (type == mvm->scan_type) + if (type == mvm->scan_type) { + IWL_DEBUG_SCAN(mvm, + "Ignoring UMAC scan config of the same type\n"); return 0; + } cmd_size = sizeof(*scan_config) + mvm->fw->ucode_capa.n_scan_channels; @@@ -1112,7 -1109,7 +1112,7 @@@ static int iwl_mvm_scan_umac(struct iwl cmd->general_flags = cpu_to_le32(iwl_mvm_scan_umac_flags(mvm, params, vif)); - if (type == IWL_MVM_SCAN_SCHED) + if (type == IWL_MVM_SCAN_SCHED || type == IWL_MVM_SCAN_NETDETECT) cmd->flags = cpu_to_le32(IWL_UMAC_SCAN_FLAG_PREEMPTIVE); if (iwl_mvm_scan_use_ebs(mvm, vif)) @@@ -1301,6 -1298,10 +1301,10 @@@ int iwl_mvm_sched_scan_start(struct iwl return -EBUSY; } + /* we don't support "match all" in the firmware */ + if (!req->n_match_sets) + return -EOPNOTSUPP; + ret = iwl_mvm_check_running_scans(mvm, type); if (ret) return ret; @@@ -1354,7 -1355,7 +1358,7 @@@ if (fw_has_capa(&mvm->fw->ucode_capa, IWL_UCODE_TLV_CAPA_UMAC_SCAN)) { hcmd.id = iwl_cmd_id(SCAN_REQ_UMAC, IWL_ALWAYS_LONG_GROUP, 0); - ret = iwl_mvm_scan_umac(mvm, vif, &params, IWL_MVM_SCAN_SCHED); + ret = iwl_mvm_scan_umac(mvm, vif, &params, type); } else { hcmd.id = SCAN_OFFLOAD_REQUEST_CMD; ret = iwl_mvm_scan_lmac(mvm, vif, &params); diff --combined drivers/net/wireless/intel/iwlwifi/pcie/internal.h index 2f95916,73c9559..542bbc5 --- a/drivers/net/wireless/intel/iwlwifi/pcie/internal.h +++ b/drivers/net/wireless/intel/iwlwifi/pcie/internal.h @@@ -2,7 -2,6 +2,7 @@@ * * Copyright(c) 2003 - 2015 Intel Corporation. All rights reserved. * Copyright(c) 2013 - 2015 Intel Mobile Communications GmbH + * Copyright(c) 2016 Intel Deutschland GmbH * * Portions of this file are derived from the ipw3945 project, as well * as portions of the ieee80211 subsystem header files. @@@ -57,23 -56,17 +57,23 @@@ #define RX_NUM_QUEUES 1 #define RX_POST_REQ_ALLOC 2 #define RX_CLAIM_REQ_ALLOC 8 -#define RX_POOL_SIZE ((RX_CLAIM_REQ_ALLOC - RX_POST_REQ_ALLOC) * RX_NUM_QUEUES) -#define RX_LOW_WATERMARK 8 +#define RX_PENDING_WATERMARK 16 struct iwl_host_cmd; /*This file includes the declaration that are internal to the * trans_pcie layer */ +/** + * struct iwl_rx_mem_buffer + * @page_dma: bus address of rxb page + * @page: driver's pointer to the rxb page + * @vid: index of this rxb in the global table + */ struct iwl_rx_mem_buffer { dma_addr_t page_dma; struct page *page; + u16 vid; struct list_head list; }; @@@ -97,12 -90,8 +97,12 @@@ struct isr_statistics /** * struct iwl_rxq - Rx queue - * @bd: driver's pointer to buffer of receive buffer descriptors (rbd) + * @id: queue index + * @bd: driver's pointer to buffer of receive buffer descriptors (rbd). + * Address size is 32 bit in pre-9000 devices and 64 bit in 9000 devices. * @bd_dma: bus address of buffer of receive buffer descriptors (rbd) + * @ubd: driver's pointer to buffer of used receive buffer descriptors (rbd) + * @ubd_dma: physical address of buffer of used receive buffer descriptors (rbd) * @read: Shared index to newest available Rx buffer * @write: Shared index to oldest written Rx packet * @free_count: Number of pre-allocated buffers in rx_free @@@ -114,34 -103,32 +114,34 @@@ * @rb_stts: driver's pointer to receive buffer status * @rb_stts_dma: bus address of receive buffer status * @lock: - * @pool: initial pool of iwl_rx_mem_buffer for the queue - * @queue: actual rx queue + * @queue: actual rx queue. Not used for multi-rx queue. * * NOTE: rx_free and rx_used are used as a FIFO for iwl_rx_mem_buffers */ struct iwl_rxq { - __le32 *bd; + int id; + void *bd; dma_addr_t bd_dma; + __le32 *used_bd; + dma_addr_t used_bd_dma; u32 read; u32 write; u32 free_count; u32 used_count; u32 write_actual; + u32 queue_size; struct list_head rx_free; struct list_head rx_used; bool need_update; struct iwl_rb_status *rb_stts; dma_addr_t rb_stts_dma; spinlock_t lock; - struct iwl_rx_mem_buffer pool[RX_QUEUE_SIZE]; + struct napi_struct napi; struct iwl_rx_mem_buffer *queue[RX_QUEUE_SIZE]; }; /** * struct iwl_rb_allocator - Rx allocator - * @pool: initial pool of allocator * @req_pending: number of requests the allcator had not processed yet * @req_ready: number of requests honored and ready for claiming * @rbd_allocated: RBDs with pages allocated and ready to be handled to @@@ -153,6 -140,7 +153,6 @@@ * @rx_alloc: work struct for background calls */ struct iwl_rb_allocator { - struct iwl_rx_mem_buffer pool[RX_POOL_SIZE]; atomic_t req_pending; atomic_t req_ready; struct list_head rbd_allocated; @@@ -292,7 -280,6 +292,7 @@@ struct iwl_txq bool ampdu; bool block; unsigned long wd_timeout; + struct sk_buff_head overflow_q; }; static inline dma_addr_t @@@ -310,8 -297,6 +310,8 @@@ struct iwl_tso_hdr_page /** * struct iwl_trans_pcie - PCIe transport specific data * @rxq: all the RX queue data + * @rx_pool: initial pool of iwl_rx_mem_buffer for all the queues + * @global_table: table mapping received VID from hw to rxb * @rba: allocator for RX replenishing * @drv - pointer to iwl_drv * @trans: pointer to the generic transport area @@@ -338,14 -323,13 +338,14 @@@ * @fw_mon_size: size of the buffer for the firmware monitor */ struct iwl_trans_pcie { - struct iwl_rxq rxq; + struct iwl_rxq *rxq; + struct iwl_rx_mem_buffer rx_pool[MQ_RX_POOL_SIZE]; + struct iwl_rx_mem_buffer *global_table[MQ_RX_TABLE_SIZE]; struct iwl_rb_allocator rba; struct iwl_trans *trans; struct iwl_drv *drv; struct net_device napi_dev; - struct napi_struct napi; struct __percpu iwl_tso_hdr_page *tso_hdr_page; @@@ -375,7 -359,6 +375,7 @@@ bool ucode_write_complete; wait_queue_head_t ucode_write_waitq; wait_queue_head_t wait_command_queue; + wait_queue_head_t d0i3_waitq; u8 cmd_queue; u8 cmd_fifo; @@@ -507,6 -490,15 +507,15 @@@ static inline void iwl_enable_interrupt iwl_write32(trans, CSR_INT_MASK, trans_pcie->inta_mask); } + static inline void iwl_enable_fw_load_int(struct iwl_trans *trans) + { + struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans); + + IWL_DEBUG_ISR(trans, "Enabling FW load interrupt\n"); + trans_pcie->inta_mask = CSR_INT_BIT_FH_TX; + iwl_write32(trans, CSR_INT_MASK, trans_pcie->inta_mask); + } + static inline void iwl_enable_rfkill_int(struct iwl_trans *trans) { struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans); @@@ -596,7 -588,4 +605,7 @@@ static inline int iwl_trans_pcie_dbgfs_ } #endif +int iwl_pci_fw_exit_d0i3(struct iwl_trans *trans); +int iwl_pci_fw_enter_d0i3(struct iwl_trans *trans); + #endif /* __iwl_trans_int_pcie_h__ */ diff --combined drivers/net/wireless/intel/iwlwifi/pcie/rx.c index 51314e56,152cf9a..07973ef --- a/drivers/net/wireless/intel/iwlwifi/pcie/rx.c +++ b/drivers/net/wireless/intel/iwlwifi/pcie/rx.c @@@ -2,7 -2,6 +2,7 @@@ * * Copyright(c) 2003 - 2014 Intel Corporation. All rights reserved. * Copyright(c) 2013 - 2015 Intel Mobile Communications GmbH + * Copyright(c) 2016 Intel Deutschland GmbH * * Portions of this file are derived from the ipw3945 project, as well * as portions of the ieee80211 subsystem header files. @@@ -141,8 -140,8 +141,8 @@@ */ static int iwl_rxq_space(const struct iwl_rxq *rxq) { - /* Make sure RX_QUEUE_SIZE is a power of 2 */ - BUILD_BUG_ON(RX_QUEUE_SIZE & (RX_QUEUE_SIZE - 1)); + /* Make sure rx queue size is a power of 2 */ + WARN_ON(rxq->queue_size & (rxq->queue_size - 1)); /* * There can be up to (RX_QUEUE_SIZE - 1) free slots, to avoid ambiguity @@@ -150,7 -149,7 +150,7 @@@ * The following is equivalent to modulo by RX_QUEUE_SIZE and is well * defined for negative dividends. */ - return (rxq->read - rxq->write - 1) & (RX_QUEUE_SIZE - 1); + return (rxq->read - rxq->write - 1) & (rxq->queue_size - 1); } /* @@@ -161,12 -160,6 +161,12 @@@ static inline __le32 iwl_pcie_dma_addr2 return cpu_to_le32((u32)(dma_addr >> 8)); } +static void iwl_pcie_write_prph_64(struct iwl_trans *trans, u64 ofs, u64 val) +{ + iwl_write_prph(trans, ofs, val & 0xffffffff); + iwl_write_prph(trans, ofs + 4, val >> 32); +} + /* * iwl_pcie_rx_stop - stops the Rx DMA */ @@@ -180,9 -173,10 +180,9 @@@ int iwl_pcie_rx_stop(struct iwl_trans * /* * iwl_pcie_rxq_inc_wr_ptr - Update the write pointer for the RX queue */ -static void iwl_pcie_rxq_inc_wr_ptr(struct iwl_trans *trans) +static void iwl_pcie_rxq_inc_wr_ptr(struct iwl_trans *trans, + struct iwl_rxq *rxq) { - struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans); - struct iwl_rxq *rxq = &trans_pcie->rxq; u32 reg; lockdep_assert_held(&rxq->lock); @@@ -207,73 -201,24 +207,73 @@@ } rxq->write_actual = round_down(rxq->write, 8); - iwl_write32(trans, FH_RSCSR_CHNL0_WPTR, rxq->write_actual); + if (trans->cfg->mq_rx_supported) + iwl_write_prph(trans, RFH_Q_FRBDCB_WIDX(rxq->id), + rxq->write_actual); + else + iwl_write32(trans, FH_RSCSR_CHNL0_WPTR, rxq->write_actual); } static void iwl_pcie_rxq_check_wrptr(struct iwl_trans *trans) { struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans); - struct iwl_rxq *rxq = &trans_pcie->rxq; + int i; - spin_lock(&rxq->lock); + for (i = 0; i < trans->num_rx_queues; i++) { + struct iwl_rxq *rxq = &trans_pcie->rxq[i]; + + if (!rxq->need_update) + continue; + spin_lock(&rxq->lock); + iwl_pcie_rxq_inc_wr_ptr(trans, rxq); + rxq->need_update = false; + spin_unlock(&rxq->lock); + } +} + +static void iwl_pcie_rxq_mq_restock(struct iwl_trans *trans, + struct iwl_rxq *rxq) +{ + struct iwl_rx_mem_buffer *rxb; + + /* + * If the device isn't enabled - no need to try to add buffers... + * This can happen when we stop the device and still have an interrupt + * pending. We stop the APM before we sync the interrupts because we + * have to (see comment there). On the other hand, since the APM is + * stopped, we cannot access the HW (in particular not prph). + * So don't try to restock if the APM has been already stopped. + */ + if (!test_bit(STATUS_DEVICE_ENABLED, &trans->status)) + return; - if (!rxq->need_update) - goto exit_unlock; + spin_lock(&rxq->lock); + while (rxq->free_count) { + __le64 *bd = (__le64 *)rxq->bd; - iwl_pcie_rxq_inc_wr_ptr(trans); - rxq->need_update = false; + /* Get next free Rx buffer, remove from free list */ + rxb = list_first_entry(&rxq->rx_free, struct iwl_rx_mem_buffer, + list); + list_del(&rxb->list); - exit_unlock: + /* 12 first bits are expected to be empty */ + WARN_ON(rxb->page_dma & DMA_BIT_MASK(12)); + /* Point to Rx buffer via next RBD in circular buffer */ + bd[rxq->write] = cpu_to_le64(rxb->page_dma | rxb->vid); + rxq->write = (rxq->write + 1) & MQ_RX_TABLE_MASK; + rxq->free_count--; + } spin_unlock(&rxq->lock); + + /* + * If we've added more space for the firmware to place data, tell it. + * Increment device's write pointer in multiples of 8. + */ + if (rxq->write_actual != (rxq->write & ~0x7)) { + spin_lock(&rxq->lock); + iwl_pcie_rxq_inc_wr_ptr(trans, rxq); + spin_unlock(&rxq->lock); + } } /* @@@ -287,8 -232,10 +287,8 @@@ * also updates the memory address in the firmware to reference the new * target buffer. */ -static void iwl_pcie_rxq_restock(struct iwl_trans *trans) +static void iwl_pcie_rxq_restock(struct iwl_trans *trans, struct iwl_rxq *rxq) { - struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans); - struct iwl_rxq *rxq = &trans_pcie->rxq; struct iwl_rx_mem_buffer *rxb; /* @@@ -304,7 -251,6 +304,7 @@@ spin_lock(&rxq->lock); while ((iwl_rxq_space(rxq) > 0) && (rxq->free_count)) { + __le32 *bd = (__le32 *)rxq->bd; /* The overwritten rxb must be a used one */ rxb = rxq->queue[rxq->write]; BUG_ON(rxb && rxb->page); @@@ -315,7 -261,7 +315,7 @@@ list_del(&rxb->list); /* Point to Rx buffer via next RBD in circular buffer */ - rxq->bd[rxq->write] = iwl_pcie_dma_addr2rbd_ptr(rxb->page_dma); + bd[rxq->write] = iwl_pcie_dma_addr2rbd_ptr(rxb->page_dma); rxq->queue[rxq->write] = rxb; rxq->write = (rxq->write + 1) & RX_QUEUE_MASK; rxq->free_count--; @@@ -326,7 -272,7 +326,7 @@@ * Increment device's write pointer in multiples of 8. */ if (rxq->write_actual != (rxq->write & ~0x7)) { spin_lock(&rxq->lock); - iwl_pcie_rxq_inc_wr_ptr(trans); + iwl_pcie_rxq_inc_wr_ptr(trans, rxq); spin_unlock(&rxq->lock); } } @@@ -339,9 -285,13 +339,9 @@@ static struct page *iwl_pcie_rx_alloc_p gfp_t priority) { struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans); - struct iwl_rxq *rxq = &trans_pcie->rxq; struct page *page; gfp_t gfp_mask = priority; - if (rxq->free_count > RX_LOW_WATERMARK) - gfp_mask |= __GFP_NOWARN; - if (trans_pcie->rx_page_order > 0) gfp_mask |= __GFP_COMP; @@@ -351,13 -301,16 +351,13 @@@ if (net_ratelimit()) IWL_DEBUG_INFO(trans, "alloc_pages failed, order: %d\n", trans_pcie->rx_page_order); - /* Issue an error if the hardware has consumed more than half - * of its free buffer list and we don't have enough - * pre-allocated buffers. + /* + * Issue an error if we don't have enough pre-allocated + * buffers. ` */ - if (rxq->free_count <= RX_LOW_WATERMARK && - iwl_rxq_space(rxq) > (RX_QUEUE_SIZE / 2) && - net_ratelimit()) + if (!(gfp_mask & __GFP_NOWARN) && net_ratelimit()) IWL_CRIT(trans, - "Failed to alloc_pages with GFP_KERNEL. Only %u free buffers remaining.\n", - rxq->free_count); + "Failed to alloc_pages\n"); return NULL; } return page; @@@ -372,10 -325,10 +372,10 @@@ * iwl_pcie_rxq_restock. The latter function will update the HW to use the newly * allocated buffers. */ -static void iwl_pcie_rxq_alloc_rbs(struct iwl_trans *trans, gfp_t priority) +static void iwl_pcie_rxq_alloc_rbs(struct iwl_trans *trans, gfp_t priority, + struct iwl_rxq *rxq) { struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans); - struct iwl_rxq *rxq = &trans_pcie->rxq; struct iwl_rx_mem_buffer *rxb; struct page *page; @@@ -419,6 -372,10 +419,6 @@@ __free_pages(page, trans_pcie->rx_page_order); return; } - /* dma address must be no more than 36 bits */ - BUG_ON(rxb->page_dma & ~DMA_BIT_MASK(36)); - /* and also 256 byte aligned! */ - BUG_ON(rxb->page_dma & DMA_BIT_MASK(8)); spin_lock(&rxq->lock); @@@ -429,24 -386,41 +429,24 @@@ } } -static void iwl_pcie_rxq_free_rbs(struct iwl_trans *trans) +static void iwl_pcie_free_rbs_pool(struct iwl_trans *trans) { struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans); - struct iwl_rxq *rxq = &trans_pcie->rxq; int i; - lockdep_assert_held(&rxq->lock); - - for (i = 0; i < RX_QUEUE_SIZE; i++) { - if (!rxq->pool[i].page) + for (i = 0; i < MQ_RX_POOL_SIZE; i++) { + if (!trans_pcie->rx_pool[i].page) continue; - dma_unmap_page(trans->dev, rxq->pool[i].page_dma, + dma_unmap_page(trans->dev, trans_pcie->rx_pool[i].page_dma, PAGE_SIZE << trans_pcie->rx_page_order, DMA_FROM_DEVICE); - __free_pages(rxq->pool[i].page, trans_pcie->rx_page_order); - rxq->pool[i].page = NULL; + __free_pages(trans_pcie->rx_pool[i].page, + trans_pcie->rx_page_order); + trans_pcie->rx_pool[i].page = NULL; } } /* - * iwl_pcie_rx_replenish - Move all used buffers from rx_used to rx_free - * - * When moving to rx_free an page is allocated for the slot. - * - * Also restock the Rx queue via iwl_pcie_rxq_restock. - * This is called only during initialization - */ -static void iwl_pcie_rx_replenish(struct iwl_trans *trans) -{ - iwl_pcie_rxq_alloc_rbs(trans, GFP_KERNEL); - - iwl_pcie_rxq_restock(trans); -} - -/* * iwl_pcie_rx_allocator - Allocates pages in the background for RX queues * * Allocates for each received request 8 pages @@@ -470,11 -444,6 +470,11 @@@ static void iwl_pcie_rx_allocator(struc while (pending) { int i; struct list_head local_allocated; + gfp_t gfp_mask = GFP_KERNEL; + + /* Do not post a warning if there are only a few requests */ + if (pending < RX_PENDING_WATERMARK) + gfp_mask |= __GFP_NOWARN; INIT_LIST_HEAD(&local_allocated); @@@ -494,7 -463,7 +494,7 @@@ BUG_ON(rxb->page); /* Alloc a new receive buffer */ - page = iwl_pcie_rx_alloc_page(trans, GFP_KERNEL); + page = iwl_pcie_rx_alloc_page(trans, gfp_mask); if (!page) continue; rxb->page = page; @@@ -508,6 -477,10 +508,6 @@@ __free_pages(page, trans_pcie->rx_page_order); continue; } - /* dma address must be no more than 36 bits */ - BUG_ON(rxb->page_dma & ~DMA_BIT_MASK(36)); - /* and also 256 byte aligned! */ - BUG_ON(rxb->page_dma & DMA_BIT_MASK(8)); /* move the allocated entry to the out list */ list_move(&rxb->list, &local_allocated); @@@ -588,83 -561,38 +588,83 @@@ static void iwl_pcie_rx_allocator_work( static int iwl_pcie_rx_alloc(struct iwl_trans *trans) { struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans); - struct iwl_rxq *rxq = &trans_pcie->rxq; struct iwl_rb_allocator *rba = &trans_pcie->rba; struct device *dev = trans->dev; + int i; + int free_size = trans->cfg->mq_rx_supported ? sizeof(__le64) : + sizeof(__le32); + + if (WARN_ON(trans_pcie->rxq)) + return -EINVAL; - memset(&trans_pcie->rxq, 0, sizeof(trans_pcie->rxq)); + trans_pcie->rxq = kcalloc(trans->num_rx_queues, sizeof(struct iwl_rxq), + GFP_KERNEL); + if (!trans_pcie->rxq) + return -EINVAL; - spin_lock_init(&rxq->lock); spin_lock_init(&rba->lock); - if (WARN_ON(rxq->bd || rxq->rb_stts)) - return -EINVAL; + for (i = 0; i < trans->num_rx_queues; i++) { + struct iwl_rxq *rxq = &trans_pcie->rxq[i]; - /* Allocate the circular buffer of Read Buffer Descriptors (RBDs) */ - rxq->bd = dma_zalloc_coherent(dev, sizeof(__le32) * RX_QUEUE_SIZE, - &rxq->bd_dma, GFP_KERNEL); - if (!rxq->bd) - goto err_bd; + spin_lock_init(&rxq->lock); + if (trans->cfg->mq_rx_supported) + rxq->queue_size = MQ_RX_TABLE_SIZE; + else + rxq->queue_size = RX_QUEUE_SIZE; - /*Allocate the driver's pointer to receive buffer status */ - rxq->rb_stts = dma_zalloc_coherent(dev, sizeof(*rxq->rb_stts), - &rxq->rb_stts_dma, GFP_KERNEL); - if (!rxq->rb_stts) - goto err_rb_stts; + /* + * Allocate the circular buffer of Read Buffer Descriptors + * (RBDs) + */ + rxq->bd = dma_zalloc_coherent(dev, + free_size * rxq->queue_size, + &rxq->bd_dma, GFP_KERNEL); + if (!rxq->bd) + goto err; + + if (trans->cfg->mq_rx_supported) { + rxq->used_bd = dma_zalloc_coherent(dev, + sizeof(__le32) * + rxq->queue_size, + &rxq->used_bd_dma, + GFP_KERNEL); + if (!rxq->used_bd) + goto err; + } + /*Allocate the driver's pointer to receive buffer status */ + rxq->rb_stts = dma_zalloc_coherent(dev, sizeof(*rxq->rb_stts), + &rxq->rb_stts_dma, + GFP_KERNEL); + if (!rxq->rb_stts) + goto err; + } return 0; -err_rb_stts: - dma_free_coherent(dev, sizeof(__le32) * RX_QUEUE_SIZE, - rxq->bd, rxq->bd_dma); - rxq->bd_dma = 0; - rxq->bd = NULL; -err_bd: +err: + for (i = 0; i < trans->num_rx_queues; i++) { + struct iwl_rxq *rxq = &trans_pcie->rxq[i]; + + if (rxq->bd) + dma_free_coherent(dev, free_size * rxq->queue_size, + rxq->bd, rxq->bd_dma); + rxq->bd_dma = 0; + rxq->bd = NULL; + + if (rxq->rb_stts) + dma_free_coherent(trans->dev, + sizeof(struct iwl_rb_status), + rxq->rb_stts, rxq->rb_stts_dma); + + if (rxq->used_bd) + dma_free_coherent(dev, sizeof(__le32) * rxq->queue_size, + rxq->used_bd, rxq->used_bd_dma); + rxq->used_bd_dma = 0; + rxq->used_bd = NULL; + } + kfree(trans_pcie->rxq); + return -ENOMEM; } @@@ -731,103 -659,65 +731,103 @@@ static void iwl_pcie_rx_hw_init(struct iwl_set_bit(trans, CSR_INT_COALESCING, IWL_HOST_INT_OPER_MODE); } -static void iwl_pcie_rx_init_rxb_lists(struct iwl_rxq *rxq) +static void iwl_pcie_rx_mq_hw_init(struct iwl_trans *trans) { + struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans); + u32 rb_size, enabled = 0; int i; - lockdep_assert_held(&rxq->lock); - - INIT_LIST_HEAD(&rxq->rx_free); - INIT_LIST_HEAD(&rxq->rx_used); - rxq->free_count = 0; - rxq->used_count = 0; + switch (trans_pcie->rx_buf_size) { + case IWL_AMSDU_4K: + rb_size = RFH_RXF_DMA_RB_SIZE_4K; + break; + case IWL_AMSDU_8K: + rb_size = RFH_RXF_DMA_RB_SIZE_8K; + break; + case IWL_AMSDU_12K: + rb_size = RFH_RXF_DMA_RB_SIZE_12K; + break; + default: + WARN_ON(1); + rb_size = RFH_RXF_DMA_RB_SIZE_4K; + } - for (i = 0; i < RX_QUEUE_SIZE; i++) - list_add(&rxq->pool[i].list, &rxq->rx_used); -} + /* Stop Rx DMA */ + iwl_write_prph(trans, RFH_RXF_DMA_CFG, 0); + /* disable free amd used rx queue operation */ + iwl_write_prph(trans, RFH_RXF_RXQ_ACTIVE, 0); + + for (i = 0; i < trans->num_rx_queues; i++) { + /* Tell device where to find RBD free table in DRAM */ + iwl_pcie_write_prph_64(trans, RFH_Q_FRBDCB_BA_LSB(i), + (u64)(trans_pcie->rxq[i].bd_dma)); + /* Tell device where to find RBD used table in DRAM */ + iwl_pcie_write_prph_64(trans, RFH_Q_URBDCB_BA_LSB(i), + (u64)(trans_pcie->rxq[i].used_bd_dma)); + /* Tell device where in DRAM to update its Rx status */ + iwl_pcie_write_prph_64(trans, RFH_Q_URBD_STTS_WPTR_LSB(i), + trans_pcie->rxq[i].rb_stts_dma); + /* Reset device indice tables */ + iwl_write_prph(trans, RFH_Q_FRBDCB_WIDX(i), 0); + iwl_write_prph(trans, RFH_Q_FRBDCB_RIDX(i), 0); + iwl_write_prph(trans, RFH_Q_URBDCB_WIDX(i), 0); + + enabled |= BIT(i) | BIT(i + 16); + } -static void iwl_pcie_rx_init_rba(struct iwl_rb_allocator *rba) -{ - int i; + /* restock default queue */ + iwl_pcie_rxq_mq_restock(trans, &trans_pcie->rxq[0]); - lockdep_assert_held(&rba->lock); + /* + * Enable Rx DMA + * Single frame mode + * Rx buffer size 4 or 8k or 12k + * Min RB size 4 or 8 + * 512 RBDs + */ + iwl_write_prph(trans, RFH_RXF_DMA_CFG, + RFH_DMA_EN_ENABLE_VAL | + rb_size | RFH_RXF_DMA_SINGLE_FRAME_MASK | + RFH_RXF_DMA_MIN_RB_4_8 | + RFH_RXF_DMA_RBDCB_SIZE_512); - INIT_LIST_HEAD(&rba->rbd_allocated); - INIT_LIST_HEAD(&rba->rbd_empty); + iwl_write_prph(trans, RFH_GEN_CFG, RFH_GEN_CFG_RFH_DMA_SNOOP | + RFH_GEN_CFG_SERVICE_DMA_SNOOP); + iwl_write_prph(trans, RFH_RXF_RXQ_ACTIVE, enabled); - for (i = 0; i < RX_POOL_SIZE; i++) - list_add(&rba->pool[i].list, &rba->rbd_empty); + /* Set interrupt coalescing timer to default (2048 usecs) */ + iwl_write8(trans, CSR_INT_COALESCING, IWL_HOST_INT_TIMEOUT_DEF); } -static void iwl_pcie_rx_free_rba(struct iwl_trans *trans) +static void iwl_pcie_rx_init_rxb_lists(struct iwl_rxq *rxq) { - struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans); - struct iwl_rb_allocator *rba = &trans_pcie->rba; - int i; + lockdep_assert_held(&rxq->lock); - lockdep_assert_held(&rba->lock); + INIT_LIST_HEAD(&rxq->rx_free); + INIT_LIST_HEAD(&rxq->rx_used); + rxq->free_count = 0; + rxq->used_count = 0; +} - for (i = 0; i < RX_POOL_SIZE; i++) { - if (!rba->pool[i].page) - continue; - dma_unmap_page(trans->dev, rba->pool[i].page_dma, - PAGE_SIZE << trans_pcie->rx_page_order, - DMA_FROM_DEVICE); - __free_pages(rba->pool[i].page, trans_pcie->rx_page_order); - rba->pool[i].page = NULL; - } +static int iwl_pcie_dummy_napi_poll(struct napi_struct *napi, int budget) +{ + WARN_ON(1); + return 0; } int iwl_pcie_rx_init(struct iwl_trans *trans) { struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans); - struct iwl_rxq *rxq = &trans_pcie->rxq; + struct iwl_rxq *def_rxq; struct iwl_rb_allocator *rba = &trans_pcie->rba; - int i, err; + int i, err, num_rbds, allocator_pool_size; - if (!rxq->bd) { + if (!trans_pcie->rxq) { err = iwl_pcie_rx_alloc(trans); if (err) return err; } + def_rxq = trans_pcie->rxq; if (!rba->alloc_wq) rba->alloc_wq = alloc_workqueue("rb_allocator", WQ_HIGHPRI | WQ_UNBOUND, 1); @@@ -836,68 -726,34 +836,68 @@@ spin_lock(&rba->lock); atomic_set(&rba->req_pending, 0); atomic_set(&rba->req_ready, 0); - /* free all first - we might be reconfigured for a different size */ - iwl_pcie_rx_free_rba(trans); - iwl_pcie_rx_init_rba(rba); + INIT_LIST_HEAD(&rba->rbd_allocated); + INIT_LIST_HEAD(&rba->rbd_empty); spin_unlock(&rba->lock); - spin_lock(&rxq->lock); - /* free all first - we might be reconfigured for a different size */ - iwl_pcie_rxq_free_rbs(trans); - iwl_pcie_rx_init_rxb_lists(rxq); + iwl_pcie_free_rbs_pool(trans); for (i = 0; i < RX_QUEUE_SIZE; i++) - rxq->queue[i] = NULL; + def_rxq->queue[i] = NULL; - /* Set us so that we have processed and used all buffers, but have - * not restocked the Rx queue with fresh buffers */ - rxq->read = rxq->write = 0; - rxq->write_actual = 0; - memset(rxq->rb_stts, 0, sizeof(*rxq->rb_stts)); - spin_unlock(&rxq->lock); + for (i = 0; i < trans->num_rx_queues; i++) { + struct iwl_rxq *rxq = &trans_pcie->rxq[i]; - iwl_pcie_rx_replenish(trans); + rxq->id = i; - iwl_pcie_rx_hw_init(trans, rxq); + spin_lock(&rxq->lock); + /* + * Set read write pointer to reflect that we have processed + * and used all buffers, but have not restocked the Rx queue + * with fresh buffers + */ + rxq->read = 0; + rxq->write = 0; + rxq->write_actual = 0; + memset(rxq->rb_stts, 0, sizeof(*rxq->rb_stts)); - spin_lock(&rxq->lock); - iwl_pcie_rxq_inc_wr_ptr(trans); - spin_unlock(&rxq->lock); + iwl_pcie_rx_init_rxb_lists(rxq); + + if (!rxq->napi.poll) + netif_napi_add(&trans_pcie->napi_dev, &rxq->napi, + iwl_pcie_dummy_napi_poll, 64); + + spin_unlock(&rxq->lock); + } + + /* move the pool to the default queue and allocator ownerships */ + num_rbds = trans->cfg->mq_rx_supported ? + MQ_RX_POOL_SIZE : RX_QUEUE_SIZE; + allocator_pool_size = trans->num_rx_queues * + (RX_CLAIM_REQ_ALLOC - RX_POST_REQ_ALLOC); + for (i = 0; i < num_rbds; i++) { + struct iwl_rx_mem_buffer *rxb = &trans_pcie->rx_pool[i]; + + if (i < allocator_pool_size) + list_add(&rxb->list, &rba->rbd_empty); + else + list_add(&rxb->list, &def_rxq->rx_used); + trans_pcie->global_table[i] = rxb; + rxb->vid = (u16)i; + } + + iwl_pcie_rxq_alloc_rbs(trans, GFP_KERNEL, def_rxq); + if (trans->cfg->mq_rx_supported) { + iwl_pcie_rx_mq_hw_init(trans); + } else { + iwl_pcie_rxq_restock(trans, def_rxq); + iwl_pcie_rx_hw_init(trans, def_rxq); + } + + spin_lock(&def_rxq->lock); + iwl_pcie_rxq_inc_wr_ptr(trans, def_rxq); + spin_unlock(&def_rxq->lock); return 0; } @@@ -905,16 -761,12 +905,16 @@@ void iwl_pcie_rx_free(struct iwl_trans *trans) { struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans); - struct iwl_rxq *rxq = &trans_pcie->rxq; struct iwl_rb_allocator *rba = &trans_pcie->rba; + int free_size = trans->cfg->mq_rx_supported ? sizeof(__le64) : + sizeof(__le32); + int i; - /*if rxq->bd is NULL, it means that nothing has been allocated, - * exit now */ - if (!rxq->bd) { + /* + * if rxq is NULL, it means that nothing has been allocated, + * exit now + */ + if (!trans_pcie->rxq) { IWL_DEBUG_INFO(trans, "Free NULL rx context\n"); return; } @@@ -925,37 -777,27 +925,37 @@@ rba->alloc_wq = NULL; } - spin_lock(&rba->lock); - iwl_pcie_rx_free_rba(trans); - spin_unlock(&rba->lock); - - spin_lock(&rxq->lock); - iwl_pcie_rxq_free_rbs(trans); - spin_unlock(&rxq->lock); - - dma_free_coherent(trans->dev, sizeof(__le32) * RX_QUEUE_SIZE, - rxq->bd, rxq->bd_dma); - rxq->bd_dma = 0; - rxq->bd = NULL; - - if (rxq->rb_stts) - dma_free_coherent(trans->dev, - sizeof(struct iwl_rb_status), - rxq->rb_stts, rxq->rb_stts_dma); - else - IWL_DEBUG_INFO(trans, "Free rxq->rb_stts which is NULL\n"); - rxq->rb_stts_dma = 0; - rxq->rb_stts = NULL; + iwl_pcie_free_rbs_pool(trans); + + for (i = 0; i < trans->num_rx_queues; i++) { + struct iwl_rxq *rxq = &trans_pcie->rxq[i]; + + if (rxq->bd) + dma_free_coherent(trans->dev, + free_size * rxq->queue_size, + rxq->bd, rxq->bd_dma); + rxq->bd_dma = 0; + rxq->bd = NULL; + + if (rxq->rb_stts) + dma_free_coherent(trans->dev, + sizeof(struct iwl_rb_status), + rxq->rb_stts, rxq->rb_stts_dma); + else + IWL_DEBUG_INFO(trans, + "Free rxq->rb_stts which is NULL\n"); + + if (rxq->used_bd) + dma_free_coherent(trans->dev, + sizeof(__le32) * rxq->queue_size, + rxq->used_bd, rxq->used_bd_dma); + rxq->used_bd_dma = 0; + rxq->used_bd = NULL; + + if (rxq->napi.poll) + netif_napi_del(&rxq->napi); + } + kfree(trans_pcie->rxq); } /* @@@ -999,11 -841,11 +999,11 @@@ static void iwl_pcie_rx_reuse_rbd(struc } static void iwl_pcie_rx_handle_rb(struct iwl_trans *trans, + struct iwl_rxq *rxq, struct iwl_rx_mem_buffer *rxb, bool emergency) { struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans); - struct iwl_rxq *rxq = &trans_pcie->rxq; struct iwl_txq *txq = &trans_pcie->txq[trans_pcie->cmd_queue]; bool page_stolen = false; int max_len = PAGE_SIZE << trans_pcie->rx_page_order; @@@ -1069,12 -911,7 +1069,12 @@@ index = SEQ_TO_INDEX(sequence); cmd_index = get_cmd_index(&txq->q, index); - iwl_op_mode_rx(trans->op_mode, &trans_pcie->napi, &rxcb); + if (rxq->id == 0) + iwl_op_mode_rx(trans->op_mode, &rxq->napi, + &rxcb); + else + iwl_op_mode_rx_rss(trans->op_mode, &rxq->napi, + &rxcb, rxq->id); if (reclaim) { kzfree(txq->entries[cmd_index].free_buf); @@@ -1138,7 -975,7 +1138,7 @@@ static void iwl_pcie_rx_handle(struct iwl_trans *trans) { struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans); - struct iwl_rxq *rxq = &trans_pcie->rxq; + struct iwl_rxq *rxq = &trans_pcie->rxq[0]; u32 r, i, j, count = 0; bool emergency = false; @@@ -1156,26 -993,16 +1156,26 @@@ restart while (i != r) { struct iwl_rx_mem_buffer *rxb; - if (unlikely(rxq->used_count == RX_QUEUE_SIZE / 2)) + if (unlikely(rxq->used_count == rxq->queue_size / 2)) emergency = true; - rxb = rxq->queue[i]; - rxq->queue[i] = NULL; + if (trans->cfg->mq_rx_supported) { + /* + * used_bd is a 32 bit but only 12 are used to retrieve + * the vid + */ + u16 vid = (u16)le32_to_cpu(rxq->used_bd[i]); + + rxb = trans_pcie->global_table[vid]; + } else { + rxb = rxq->queue[i]; + rxq->queue[i] = NULL; + } IWL_DEBUG_RX(trans, "rxbuf: HW = %d, SW = %d\n", r, i); - iwl_pcie_rx_handle_rb(trans, rxb, emergency); + iwl_pcie_rx_handle_rb(trans, rxq, rxb, emergency); - i = (i + 1) & RX_QUEUE_MASK; + i = (i + 1) & (rxq->queue_size - 1); /* If we have RX_CLAIM_REQ_ALLOC released rx buffers - * try to claim the pre-allocated buffers from the allocator */ @@@ -1213,10 -1040,10 +1213,10 @@@ count++; if (count == 8) { count = 0; - if (rxq->used_count < RX_QUEUE_SIZE / 3) + if (rxq->used_count < rxq->queue_size / 3) emergency = false; spin_unlock(&rxq->lock); - iwl_pcie_rxq_alloc_rbs(trans, GFP_ATOMIC); + iwl_pcie_rxq_alloc_rbs(trans, GFP_ATOMIC, rxq); spin_lock(&rxq->lock); } } @@@ -1228,10 -1055,7 +1228,10 @@@ if (rxq->free_count >= RX_CLAIM_REQ_ALLOC) { rxq->read = i; spin_unlock(&rxq->lock); - iwl_pcie_rxq_restock(trans); + if (trans->cfg->mq_rx_supported) + iwl_pcie_rxq_mq_restock(trans, rxq); + else + iwl_pcie_rxq_restock(trans, rxq); goto restart; } } @@@ -1253,10 -1077,10 +1253,10 @@@ * will be restocked by the next call of iwl_pcie_rxq_restock. */ if (unlikely(emergency && count)) - iwl_pcie_rxq_alloc_rbs(trans, GFP_ATOMIC); + iwl_pcie_rxq_alloc_rbs(trans, GFP_ATOMIC, rxq); - if (trans_pcie->napi.poll) - napi_gro_flush(&trans_pcie->napi, false); + if (rxq->napi.poll) + napi_gro_flush(&rxq->napi, false); } /* @@@ -1614,9 -1438,11 +1614,11 @@@ irqreturn_t iwl_pcie_irq_handler(int ir inta & ~trans_pcie->inta_mask); } - /* Re-enable all interrupts */ - /* only Re-enable if disabled by irq */ - if (test_bit(STATUS_INT_ENABLED, &trans->status)) + /* we are loading the firmware, enable FH_TX interrupt only */ + if (handled & CSR_INT_BIT_FH_TX) + iwl_enable_fw_load_int(trans); + /* only Re-enable all interrupt if disabled by irq */ + else if (test_bit(STATUS_INT_ENABLED, &trans->status)) iwl_enable_interrupts(trans); /* Re-enable RF_KILL if it occurred */ else if (handled & CSR_INT_BIT_RF_KILL) diff --combined drivers/net/wireless/intel/iwlwifi/pcie/trans.c index b796952,5a854c6..58591ca --- a/drivers/net/wireless/intel/iwlwifi/pcie/trans.c +++ b/drivers/net/wireless/intel/iwlwifi/pcie/trans.c @@@ -72,7 -72,6 +72,7 @@@ #include <linux/bitops.h> #include <linux/gfp.h> #include <linux/vmalloc.h> +#include <linux/pm_runtime.h> #include "iwl-drv.h" #include "iwl-trans.h" @@@ -1022,82 -1021,6 +1022,6 @@@ static int iwl_pcie_load_given_ucode_80 &first_ucode_section); } - static int iwl_trans_pcie_start_fw(struct iwl_trans *trans, - const struct fw_img *fw, bool run_in_rfkill) - { - struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans); - bool hw_rfkill; - int ret; - - mutex_lock(&trans_pcie->mutex); - - /* Someone called stop_device, don't try to start_fw */ - if (trans_pcie->is_down) { - IWL_WARN(trans, - "Can't start_fw since the HW hasn't been started\n"); - ret = EIO; - goto out; - } - - /* This may fail if AMT took ownership of the device */ - if (iwl_pcie_prepare_card_hw(trans)) { - IWL_WARN(trans, "Exit HW not ready\n"); - ret = -EIO; - goto out; - } - - iwl_enable_rfkill_int(trans); - - /* If platform's RF_KILL switch is NOT set to KILL */ - hw_rfkill = iwl_is_rfkill_set(trans); - if (hw_rfkill) - set_bit(STATUS_RFKILL, &trans->status); - else - clear_bit(STATUS_RFKILL, &trans->status); - iwl_trans_pcie_rf_kill(trans, hw_rfkill); - if (hw_rfkill && !run_in_rfkill) { - ret = -ERFKILL; - goto out; - } - - iwl_write32(trans, CSR_INT, 0xFFFFFFFF); - - ret = iwl_pcie_nic_init(trans); - if (ret) { - IWL_ERR(trans, "Unable to init nic\n"); - goto out; - } - - /* make sure rfkill handshake bits are cleared */ - iwl_write32(trans, CSR_UCODE_DRV_GP1_CLR, CSR_UCODE_SW_BIT_RFKILL); - iwl_write32(trans, CSR_UCODE_DRV_GP1_CLR, - CSR_UCODE_DRV_GP1_BIT_CMD_BLOCKED); - - /* clear (again), then enable host interrupts */ - iwl_write32(trans, CSR_INT, 0xFFFFFFFF); - iwl_enable_interrupts(trans); - - /* really make sure rfkill handshake bits are cleared */ - iwl_write32(trans, CSR_UCODE_DRV_GP1_CLR, CSR_UCODE_SW_BIT_RFKILL); - iwl_write32(trans, CSR_UCODE_DRV_GP1_CLR, CSR_UCODE_SW_BIT_RFKILL); - - /* Load the given image to the HW */ - if (trans->cfg->device_family == IWL_DEVICE_FAMILY_8000) - ret = iwl_pcie_load_given_ucode_8000(trans, fw); - else - ret = iwl_pcie_load_given_ucode(trans, fw); - - out: - mutex_unlock(&trans_pcie->mutex); - return ret; - } - - static void iwl_trans_pcie_fw_alive(struct iwl_trans *trans, u32 scd_addr) - { - iwl_pcie_reset_ict(trans); - iwl_pcie_tx_start(trans, scd_addr); - } - static void _iwl_trans_pcie_stop_device(struct iwl_trans *trans, bool low_power) { struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans); @@@ -1128,7 -1051,8 +1052,8 @@@ * already dead. */ if (test_and_clear_bit(STATUS_DEVICE_ENABLED, &trans->status)) { - IWL_DEBUG_INFO(trans, "DEVICE_ENABLED bit was set and is now cleared\n"); + IWL_DEBUG_INFO(trans, + "DEVICE_ENABLED bit was set and is now cleared\n"); iwl_pcie_tx_stop(trans); iwl_pcie_rx_stop(trans); @@@ -1162,7 -1086,6 +1087,6 @@@ iwl_disable_interrupts(trans); spin_unlock(&trans_pcie->irq_lock); - /* clear all status bits */ clear_bit(STATUS_SYNC_HCMD_ACTIVE, &trans->status); clear_bit(STATUS_INT_ENABLED, &trans->status); @@@ -1195,10 -1118,116 +1119,116 @@@ if (hw_rfkill != was_hw_rfkill) iwl_trans_pcie_rf_kill(trans, hw_rfkill); - /* re-take ownership to prevent other users from stealing the deivce */ + /* re-take ownership to prevent other users from stealing the device */ iwl_pcie_prepare_card_hw(trans); } + static int iwl_trans_pcie_start_fw(struct iwl_trans *trans, + const struct fw_img *fw, bool run_in_rfkill) + { + struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans); + bool hw_rfkill; + int ret; + + /* This may fail if AMT took ownership of the device */ + if (iwl_pcie_prepare_card_hw(trans)) { + IWL_WARN(trans, "Exit HW not ready\n"); + ret = -EIO; + goto out; + } + + iwl_enable_rfkill_int(trans); + + iwl_write32(trans, CSR_INT, 0xFFFFFFFF); + + /* + * We enabled the RF-Kill interrupt and the handler may very + * well be running. Disable the interrupts to make sure no other + * interrupt can be fired. + */ + iwl_disable_interrupts(trans); + + /* Make sure it finished running */ + synchronize_irq(trans_pcie->pci_dev->irq); + + mutex_lock(&trans_pcie->mutex); + + /* If platform's RF_KILL switch is NOT set to KILL */ + hw_rfkill = iwl_is_rfkill_set(trans); + if (hw_rfkill) + set_bit(STATUS_RFKILL, &trans->status); + else + clear_bit(STATUS_RFKILL, &trans->status); + iwl_trans_pcie_rf_kill(trans, hw_rfkill); + if (hw_rfkill && !run_in_rfkill) { + ret = -ERFKILL; + goto out; + } + + /* Someone called stop_device, don't try to start_fw */ + if (trans_pcie->is_down) { + IWL_WARN(trans, + "Can't start_fw since the HW hasn't been started\n"); + ret = -EIO; + goto out; + } + + /* make sure rfkill handshake bits are cleared */ + iwl_write32(trans, CSR_UCODE_DRV_GP1_CLR, CSR_UCODE_SW_BIT_RFKILL); + iwl_write32(trans, CSR_UCODE_DRV_GP1_CLR, + CSR_UCODE_DRV_GP1_BIT_CMD_BLOCKED); + + /* clear (again), then enable host interrupts */ + iwl_write32(trans, CSR_INT, 0xFFFFFFFF); + + ret = iwl_pcie_nic_init(trans); + if (ret) { + IWL_ERR(trans, "Unable to init nic\n"); + goto out; + } + + /* + * Now, we load the firmware and don't want to be interrupted, even + * by the RF-Kill interrupt (hence mask all the interrupt besides the + * FH_TX interrupt which is needed to load the firmware). If the + * RF-Kill switch is toggled, we will find out after having loaded + * the firmware and return the proper value to the caller. + */ + iwl_enable_fw_load_int(trans); + + /* really make sure rfkill handshake bits are cleared */ + iwl_write32(trans, CSR_UCODE_DRV_GP1_CLR, CSR_UCODE_SW_BIT_RFKILL); + iwl_write32(trans, CSR_UCODE_DRV_GP1_CLR, CSR_UCODE_SW_BIT_RFKILL); + + /* Load the given image to the HW */ + if (trans->cfg->device_family == IWL_DEVICE_FAMILY_8000) + ret = iwl_pcie_load_given_ucode_8000(trans, fw); + else + ret = iwl_pcie_load_given_ucode(trans, fw); + iwl_enable_interrupts(trans); + + /* re-check RF-Kill state since we may have missed the interrupt */ + hw_rfkill = iwl_is_rfkill_set(trans); + if (hw_rfkill) + set_bit(STATUS_RFKILL, &trans->status); + else + clear_bit(STATUS_RFKILL, &trans->status); + + iwl_trans_pcie_rf_kill(trans, hw_rfkill); + if (hw_rfkill && !run_in_rfkill) + ret = -ERFKILL; + + out: + mutex_unlock(&trans_pcie->mutex); + return ret; + } + + static void iwl_trans_pcie_fw_alive(struct iwl_trans *trans, u32 scd_addr) + { + iwl_pcie_reset_ict(trans); + iwl_pcie_tx_start(trans, scd_addr); + } + static void iwl_trans_pcie_stop_device(struct iwl_trans *trans, bool low_power) { struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans); @@@ -1219,12 -1248,11 +1249,12 @@@ void iwl_trans_pcie_rf_kill(struct iwl_ _iwl_trans_pcie_stop_device(trans, true); } -static void iwl_trans_pcie_d3_suspend(struct iwl_trans *trans, bool test) +static void iwl_trans_pcie_d3_suspend(struct iwl_trans *trans, bool test, + bool reset) { struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans); - if (trans->system_pm_mode == IWL_PLAT_PM_MODE_D0I3) { + if (!reset) { /* Enable persistence mode to avoid reset */ iwl_set_bit(trans, CSR_HW_IF_CONFIG_REG, CSR_HW_IF_CONFIG_REG_PERSIST_MODE); @@@ -1248,7 -1276,7 +1278,7 @@@ iwl_clear_bit(trans, CSR_GP_CNTRL, CSR_GP_CNTRL_REG_FLAG_INIT_DONE); - if (trans->system_pm_mode == IWL_PLAT_PM_MODE_D3) { + if (reset) { /* * reset TX queues -- some of their registers reset during S3 * so if we don't reset everything here the D3 image would try @@@ -1262,7 -1290,7 +1292,7 @@@ static int iwl_trans_pcie_d3_resume(struct iwl_trans *trans, enum iwl_d3_status *status, - bool test) + bool test, bool reset) { u32 val; int ret; @@@ -1297,7 -1325,7 +1327,7 @@@ iwl_pcie_set_pwr(trans, false); - if (trans->system_pm_mode == IWL_PLAT_PM_MODE_D0I3) { + if (!reset) { iwl_clear_bit(trans, CSR_GP_CNTRL, CSR_GP_CNTRL_REG_FLAG_MAC_ACCESS_REQ); } else { @@@ -1355,10 -1383,6 +1385,10 @@@ static int _iwl_trans_pcie_start_hw(str /* ... rfkill can call stop_device and set it false if needed */ iwl_trans_pcie_rf_kill(trans, hw_rfkill); + /* Make sure we sync here, because we'll need full access later */ + if (low_power) + pm_runtime_resume(trans->dev); + return 0; } @@@ -1428,6 -1452,12 +1458,6 @@@ static void iwl_trans_pcie_write_prph(s iwl_trans_pcie_write32(trans, HBUS_TARG_PRPH_WDAT, val); } -static int iwl_pcie_dummy_napi_poll(struct napi_struct *napi, int budget) -{ - WARN_ON(1); - return 0; -} - static void iwl_trans_pcie_configure(struct iwl_trans *trans, const struct iwl_trans_config *trans_cfg) { @@@ -1464,8 -1494,11 +1494,8 @@@ * As this function may be called again in some corner cases don't * do anything if NAPI was already initialized. */ - if (!trans_pcie->napi.poll) { + if (trans_pcie->napi_dev.reg_state != NETREG_DUMMY) init_dummy_netdev(&trans_pcie->napi_dev); - netif_napi_add(&trans_pcie->napi_dev, &trans_pcie->napi, - iwl_pcie_dummy_napi_poll, 64); - } } void iwl_trans_pcie_free(struct iwl_trans *trans) @@@ -1473,9 -1506,6 +1503,9 @@@ struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans); int i; + /* TODO: check if this is really needed */ + pm_runtime_disable(trans->dev); + synchronize_irq(trans_pcie->pci_dev->irq); iwl_pcie_tx_free(trans); @@@ -1489,6 -1519,9 +1519,6 @@@ pci_release_regions(trans_pcie->pci_dev); pci_disable_device(trans_pcie->pci_dev); - if (trans_pcie->napi.poll) - netif_napi_del(&trans_pcie->napi); - iwl_pcie_free_fw_monitor(trans); for_each_possible_cpu(i) { @@@ -1828,7 -1861,6 +1858,7 @@@ void iwl_trans_pcie_ref(struct iwl_tran spin_lock_irqsave(&trans_pcie->ref_lock, flags); IWL_DEBUG_RPM(trans, "ref_counter: %d\n", trans_pcie->ref_count); trans_pcie->ref_count++; + pm_runtime_get(&trans_pcie->pci_dev->dev); spin_unlock_irqrestore(&trans_pcie->ref_lock, flags); } @@@ -1847,10 -1879,6 +1877,10 @@@ void iwl_trans_pcie_unref(struct iwl_tr return; } trans_pcie->ref_count--; + + pm_runtime_mark_last_busy(&trans_pcie->pci_dev->dev); + pm_runtime_put_autosuspend(&trans_pcie->pci_dev->dev); + spin_unlock_irqrestore(&trans_pcie->ref_lock, flags); } @@@ -2003,48 -2031,29 +2033,48 @@@ static ssize_t iwl_dbgfs_rx_queue_read( { struct iwl_trans *trans = file->private_data; struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans); - struct iwl_rxq *rxq = &trans_pcie->rxq; - char buf[256]; - int pos = 0; - const size_t bufsz = sizeof(buf); - - pos += scnprintf(buf + pos, bufsz - pos, "read: %u\n", - rxq->read); - pos += scnprintf(buf + pos, bufsz - pos, "write: %u\n", - rxq->write); - pos += scnprintf(buf + pos, bufsz - pos, "write_actual: %u\n", - rxq->write_actual); - pos += scnprintf(buf + pos, bufsz - pos, "need_update: %d\n", - rxq->need_update); - pos += scnprintf(buf + pos, bufsz - pos, "free_count: %u\n", - rxq->free_count); - if (rxq->rb_stts) { - pos += scnprintf(buf + pos, bufsz - pos, "closed_rb_num: %u\n", - le16_to_cpu(rxq->rb_stts->closed_rb_num) & 0x0FFF); - } else { - pos += scnprintf(buf + pos, bufsz - pos, - "closed_rb_num: Not Allocated\n"); + char *buf; + int pos = 0, i, ret; + size_t bufsz = sizeof(buf); + + bufsz = sizeof(char) * 121 * trans->num_rx_queues; + + if (!trans_pcie->rxq) + return -EAGAIN; + + buf = kzalloc(bufsz, GFP_KERNEL); + if (!buf) + return -ENOMEM; + + for (i = 0; i < trans->num_rx_queues && pos < bufsz; i++) { + struct iwl_rxq *rxq = &trans_pcie->rxq[i]; + + pos += scnprintf(buf + pos, bufsz - pos, "queue#: %2d\n", + i); + pos += scnprintf(buf + pos, bufsz - pos, "\tread: %u\n", + rxq->read); + pos += scnprintf(buf + pos, bufsz - pos, "\twrite: %u\n", + rxq->write); + pos += scnprintf(buf + pos, bufsz - pos, "\twrite_actual: %u\n", + rxq->write_actual); + pos += scnprintf(buf + pos, bufsz - pos, "\tneed_update: %2d\n", + rxq->need_update); + pos += scnprintf(buf + pos, bufsz - pos, "\tfree_count: %u\n", + rxq->free_count); + if (rxq->rb_stts) { + pos += scnprintf(buf + pos, bufsz - pos, + "\tclosed_rb_num: %u\n", + le16_to_cpu(rxq->rb_stts->closed_rb_num) & + 0x0FFF); + } else { + pos += scnprintf(buf + pos, bufsz - pos, + "\tclosed_rb_num: Not Allocated\n"); + } } - return simple_read_from_buffer(user_buf, count, ppos, buf, pos); + ret = simple_read_from_buffer(user_buf, count, ppos, buf, pos); + kfree(buf); + + return ret; } static ssize_t iwl_dbgfs_interrupt_read(struct file *file, @@@ -2209,8 -2218,7 +2239,8 @@@ static u32 iwl_trans_pcie_dump_rbs(stru { struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans); int max_len = PAGE_SIZE << trans_pcie->rx_page_order; - struct iwl_rxq *rxq = &trans_pcie->rxq; + /* Dump RBs is supported only for pre-9000 devices (1 queue) */ + struct iwl_rxq *rxq = &trans_pcie->rxq[0]; u32 i, r, j, rb_len = 0; spin_lock(&rxq->lock); @@@ -2405,8 -2413,7 +2435,8 @@@ static struct iwl_trans_dump_dat u32 len, num_rbs; u32 monitor_len; int i, ptr; - bool dump_rbs = test_bit(STATUS_FW_ERROR, &trans->status); + bool dump_rbs = test_bit(STATUS_FW_ERROR, &trans->status) && + !trans->cfg->mq_rx_supported; /* transport dump header */ len = sizeof(*dump_data); @@@ -2461,12 -2468,11 +2491,12 @@@ len += sizeof(*data) + (FH_MEM_UPPER_BOUND - FH_MEM_LOWER_BOUND); if (dump_rbs) { + /* Dump RBs is supported only for pre-9000 devices (1 queue) */ + struct iwl_rxq *rxq = &trans_pcie->rxq[0]; /* RBs */ - num_rbs = le16_to_cpu(ACCESS_ONCE( - trans_pcie->rxq.rb_stts->closed_rb_num)) + num_rbs = le16_to_cpu(ACCESS_ONCE(rxq->rb_stts->closed_rb_num)) & 0x0FFF; - num_rbs = (num_rbs - trans_pcie->rxq.read) & RX_QUEUE_MASK; + num_rbs = (num_rbs - rxq->read) & RX_QUEUE_MASK; len += num_rbs * (sizeof(*data) + sizeof(struct iwl_fw_error_dump_rb) + (PAGE_SIZE << trans_pcie->rx_page_order)); @@@ -2517,22 -2523,6 +2547,22 @@@ return dump_data; } +#ifdef CONFIG_PM_SLEEP +static int iwl_trans_pcie_suspend(struct iwl_trans *trans) +{ + if (trans->runtime_pm_mode == IWL_PLAT_PM_MODE_D0I3) + return iwl_pci_fw_enter_d0i3(trans); + + return 0; +} + +static void iwl_trans_pcie_resume(struct iwl_trans *trans) +{ + if (trans->runtime_pm_mode == IWL_PLAT_PM_MODE_D0I3) + iwl_pci_fw_exit_d0i3(trans); +} +#endif /* CONFIG_PM_SLEEP */ + static const struct iwl_trans_ops trans_ops_pcie = { .start_hw = iwl_trans_pcie_start_hw, .op_mode_leave = iwl_trans_pcie_op_mode_leave, @@@ -2543,11 -2533,6 +2573,11 @@@ .d3_suspend = iwl_trans_pcie_d3_suspend, .d3_resume = iwl_trans_pcie_d3_resume, +#ifdef CONFIG_PM_SLEEP + .suspend = iwl_trans_pcie_suspend, + .resume = iwl_trans_pcie_resume, +#endif /* CONFIG_PM_SLEEP */ + .send_cmd = iwl_trans_pcie_send_hcmd, .tx = iwl_trans_pcie_tx, @@@ -2586,7 -2571,7 +2616,7 @@@ struct iwl_trans *iwl_trans_pcie_alloc( struct iwl_trans_pcie *trans_pcie; struct iwl_trans *trans; u16 pci_cmd; - int ret; + int ret, addr_size; trans = iwl_trans_alloc(sizeof(struct iwl_trans_pcie), &pdev->dev, cfg, &trans_ops_pcie, 0); @@@ -2624,17 -2609,11 +2654,17 @@@ PCIE_LINK_STATE_CLKPM); } + if (cfg->mq_rx_supported) + addr_size = 64; + else + addr_size = 36; + pci_set_master(pdev); - ret = pci_set_dma_mask(pdev, DMA_BIT_MASK(36)); + ret = pci_set_dma_mask(pdev, DMA_BIT_MASK(addr_size)); if (!ret) - ret = pci_set_consistent_dma_mask(pdev, DMA_BIT_MASK(36)); + ret = pci_set_consistent_dma_mask(pdev, + DMA_BIT_MASK(addr_size)); if (ret) { ret = pci_set_dma_mask(pdev, DMA_BIT_MASK(32)); if (!ret) @@@ -2737,8 -2716,6 +2767,8 @@@ /* Initialize the wait queue for commands */ init_waitqueue_head(&trans_pcie->wait_command_queue); + init_waitqueue_head(&trans_pcie->d0i3_waitq); + ret = iwl_pcie_alloc_ict(trans); if (ret) goto out_pci_disable_msi; @@@ -2753,12 -2730,6 +2783,12 @@@ trans_pcie->inta_mask = CSR_INI_SET_MASK; +#ifdef CONFIG_IWLWIFI_PCIE_RTPM + trans->runtime_pm_mode = IWL_PLAT_PM_MODE_D0I3; +#else + trans->runtime_pm_mode = IWL_PLAT_PM_MODE_DISABLED; +#endif /* CONFIG_IWLWIFI_PCIE_RTPM */ + return trans; out_free_ict: diff --combined include/linux/netdevice.h index 47671ce0,5440b7b..e52077f --- a/include/linux/netdevice.h +++ b/include/linux/netdevice.h @@@ -51,7 -51,6 +51,7 @@@ #include <linux/neighbour.h> #include <uapi/linux/netdevice.h> #include <uapi/linux/if_bonding.h> +#include <uapi/linux/pkt_cls.h> struct netpoll_info; struct device; @@@ -779,25 -778,6 +779,25 @@@ static inline bool netdev_phys_item_id_ typedef u16 (*select_queue_fallback_t)(struct net_device *dev, struct sk_buff *skb); +/* These structures hold the attributes of qdisc and classifiers + * that are being passed to the netdevice through the setup_tc op. + */ +enum { + TC_SETUP_MQPRIO, + TC_SETUP_CLSU32, +}; + +struct tc_cls_u32_offload; + +struct tc_to_netdev { + unsigned int type; + union { + u8 tc; + struct tc_cls_u32_offload *cls_u32; + }; +}; + + /* * This structure defines the management hooks for network devices. * The following hooks can be defined; unless noted otherwise, they are @@@ -1170,10 -1150,7 +1170,10 @@@ struct net_device_ops int (*ndo_set_vf_rss_query_en)( struct net_device *dev, int vf, bool setting); - int (*ndo_setup_tc)(struct net_device *dev, u8 tc); + int (*ndo_setup_tc)(struct net_device *dev, + u32 handle, + __be16 protocol, + struct tc_to_netdev *tc); #if IS_ENABLED(CONFIG_FCOE) int (*ndo_fcoe_enable)(struct net_device *dev); int (*ndo_fcoe_disable)(struct net_device *dev); @@@ -1314,7 -1291,6 +1314,7 @@@ * @IFF_OPENVSWITCH: device is a Open vSwitch master * @IFF_L3MDEV_SLAVE: device is enslaved to an L3 master device * @IFF_TEAM: device is a team device + * @IFF_RXFH_CONFIGURED: device has had Rx Flow indirection table configured */ enum netdev_priv_flags { IFF_802_1Q_VLAN = 1<<0, @@@ -1342,7 -1318,6 +1342,7 @@@ IFF_OPENVSWITCH = 1<<22, IFF_L3MDEV_SLAVE = 1<<23, IFF_TEAM = 1<<24, + IFF_RXFH_CONFIGURED = 1<<25, }; #define IFF_802_1Q_VLAN IFF_802_1Q_VLAN @@@ -1370,7 -1345,6 +1370,7 @@@ #define IFF_OPENVSWITCH IFF_OPENVSWITCH #define IFF_L3MDEV_SLAVE IFF_L3MDEV_SLAVE #define IFF_TEAM IFF_TEAM +#define IFF_RXFH_CONFIGURED IFF_RXFH_CONFIGURED /** * struct net_device - The DEVICE structure. @@@ -1423,8 -1397,6 +1423,8 @@@ * do not use this in drivers * @tx_dropped: Dropped packets by core network, * do not use this in drivers + * @rx_nohandler: nohandler dropped packets by core network on + * inactive devices, do not use this in drivers * * @wireless_handlers: List of functions to handle Wireless Extensions, * instead of ioctl, @@@ -1639,7 -1611,6 +1639,7 @@@ struct net_device atomic_long_t rx_dropped; atomic_long_t tx_dropped; + atomic_long_t rx_nohandler; #ifdef CONFIG_WIRELESS_EXT const struct iw_handler_def * wireless_handlers; @@@ -3747,7 -3718,7 +3747,7 @@@ void *netdev_lower_get_next_private_rcu void *netdev_lower_get_next(struct net_device *dev, struct list_head **iter); #define netdev_for_each_lower_dev(dev, ldev, iter) \ - for (iter = &(dev)->adj_list.lower, \ + for (iter = (dev)->adj_list.lower.next, \ ldev = netdev_lower_get_next(dev, &(iter)); \ ldev; \ ldev = netdev_lower_get_next(dev, &(iter))) @@@ -3770,7 -3741,7 +3770,7 @@@ void netdev_lower_state_changed(struct /* RSS keys are 40 or 52 bytes long */ #define NETDEV_RSS_KEY_LEN 52 -extern u8 netdev_rss_key[NETDEV_RSS_KEY_LEN]; +extern u8 netdev_rss_key[NETDEV_RSS_KEY_LEN] __read_mostly; void netdev_rss_key_fill(void *buffer, size_t len); int dev_get_nest_level(struct net_device *dev, @@@ -4074,11 -4045,6 +4074,11 @@@ static inline bool netif_is_lag_port(co return netif_is_bond_slave(dev) || netif_is_team_port(dev); } +static inline bool netif_is_rxfh_configured(const struct net_device *dev) +{ + return dev->priv_flags & IFF_RXFH_CONFIGURED; +} + /* This device needs to keep skb dst for qdisc enqueue or ndo_start_xmit() */ static inline void netif_keep_dst(struct net_device *dev) { diff --combined include/linux/skbuff.h index 6a57757,4ce9ff7..eab4f8f --- a/include/linux/skbuff.h +++ b/include/linux/skbuff.h @@@ -299,6 -299,7 +299,7 @@@ struct sk_buff #else #define MAX_SKB_FRAGS (65536/PAGE_SIZE + 1) #endif + extern int sysctl_max_skb_frags; typedef struct skb_frag_struct skb_frag_t; @@@ -2161,11 -2162,6 +2162,11 @@@ static inline int skb_checksum_start_of return skb->csum_start - skb_headroom(skb); } +static inline unsigned char *skb_checksum_start(const struct sk_buff *skb) +{ + return skb->head + skb->csum_start; +} + static inline int skb_transport_offset(const struct sk_buff *skb) { return skb_transport_header(skb) - skb->data; @@@ -2404,10 -2400,6 +2405,10 @@@ static inline struct sk_buff *napi_allo { return __napi_alloc_skb(napi, length, GFP_ATOMIC); } +void napi_consume_skb(struct sk_buff *skb, int budget); + +void __kfree_skb_flush(void); +void __kfree_skb_defer(struct sk_buff *skb); /** * __dev_alloc_pages - allocate page for network Rx @@@ -2630,13 -2622,6 +2631,13 @@@ static inline int skb_clone_writable(co skb_headroom(skb) + len <= skb->hdr_len; } +static inline int skb_try_make_writable(struct sk_buff *skb, + unsigned int write_len) +{ + return skb_cloned(skb) && !skb_clone_writable(skb, write_len) && + pskb_expand_head(skb, 0, 0, GFP_ATOMIC); +} + static inline int __skb_cow(struct sk_buff *skb, unsigned int headroom, int cloned) { @@@ -3565,7 -3550,6 +3566,7 @@@ static inline struct sec_path *skb_sec_ struct skb_gso_cb { int mac_offset; int encap_level; + __wsum csum; __u16 csum_start; }; #define SKB_SGO_CB_OFFSET 32 @@@ -3592,16 -3576,6 +3593,16 @@@ static inline int gso_pskb_expand_head( return 0; } +static inline void gso_reset_checksum(struct sk_buff *skb, __wsum res) +{ + /* Do not update partial checksums if remote checksum is enabled. */ + if (skb->remcsum_offload) + return; + + SKB_GSO_CB(skb)->csum = res; + SKB_GSO_CB(skb)->csum_start = skb_checksum_start(skb) - skb->head; +} + /* Compute the checksum for a gso segment. First compute the checksum value * from the start of transport header to SKB_GSO_CB(skb)->csum_start, and * then add in skb->csum (checksum from csum_start to end of packet). @@@ -3612,14 -3586,15 +3613,14 @@@ */ static inline __sum16 gso_make_checksum(struct sk_buff *skb, __wsum res) { - int plen = SKB_GSO_CB(skb)->csum_start - skb_headroom(skb) - - skb_transport_offset(skb); - __wsum partial; + unsigned char *csum_start = skb_transport_header(skb); + int plen = (skb->head + SKB_GSO_CB(skb)->csum_start) - csum_start; + __wsum partial = SKB_GSO_CB(skb)->csum; - partial = csum_partial(skb_transport_header(skb), plen, skb->csum); - skb->csum = res; - SKB_GSO_CB(skb)->csum_start -= plen; + SKB_GSO_CB(skb)->csum = res; + SKB_GSO_CB(skb)->csum_start = csum_start - skb->head; - return csum_fold(partial); + return csum_fold(csum_partial(csum_start, plen, partial)); } static inline bool skb_is_gso(const struct sk_buff *skb) @@@ -3709,30 -3684,5 +3710,30 @@@ static inline unsigned int skb_gso_netw return hdr_len + skb_gso_transport_seglen(skb); } +/* Local Checksum Offload. + * Compute outer checksum based on the assumption that the + * inner checksum will be offloaded later. + * See Documentation/networking/checksum-offloads.txt for + * explanation of how this works. + * Fill in outer checksum adjustment (e.g. with sum of outer + * pseudo-header) before calling. + * Also ensure that inner checksum is in linear data area. + */ +static inline __wsum lco_csum(struct sk_buff *skb) +{ + unsigned char *csum_start = skb_checksum_start(skb); + unsigned char *l4_hdr = skb_transport_header(skb); + __wsum partial; + + /* Start with complement of inner checksum adjustment */ + partial = ~csum_unfold(*(__force __sum16 *)(csum_start + + skb->csum_offset)); + + /* Add in checksum of our headers (incl. outer checksum + * adjustment filled in by caller) and return result. + */ + return csum_partial(l4_hdr, csum_start - l4_hdr, partial); +} + #endif /* __KERNEL__ */ #endif /* _LINUX_SKBUFF_H */ diff --combined include/net/ip_tunnels.h index 4dd6163,dda9abf..5f28b60 --- a/include/net/ip_tunnels.h +++ b/include/net/ip_tunnels.h @@@ -13,7 -13,6 +13,7 @@@ #include <net/netns/generic.h> #include <net/rtnetlink.h> #include <net/lwtunnel.h> +#include <net/dst_cache.h> #if IS_ENABLED(CONFIG_IPV6) #include <net/ipv6.h> @@@ -58,9 -57,6 +58,9 @@@ struct ip_tunnel_key struct ip_tunnel_info { struct ip_tunnel_key key; +#ifdef CONFIG_DST_CACHE + struct dst_cache dst_cache; +#endif u8 options_len; u8 mode; }; @@@ -89,6 -85,11 +89,6 @@@ struct ip_tunnel_prl_entry struct rcu_head rcu_head; }; -struct ip_tunnel_dst { - struct dst_entry __rcu *dst; - __be32 saddr; -}; - struct metadata_dst; struct ip_tunnel { @@@ -107,7 -108,7 +107,7 @@@ int tun_hlen; /* Precalculated header length */ int mlink; - struct ip_tunnel_dst __percpu *dst_cache; + struct dst_cache dst_cache; struct ip_tunnel_parm parms; @@@ -229,6 -230,7 +229,7 @@@ void ip_tunnel_xmit(struct sk_buff *skb int ip_tunnel_ioctl(struct net_device *dev, struct ip_tunnel_parm *p, int cmd); int ip_tunnel_encap(struct sk_buff *skb, struct ip_tunnel *t, u8 *protocol, struct flowi4 *fl4); + int __ip_tunnel_change_mtu(struct net_device *dev, int new_mtu, bool strict); int ip_tunnel_change_mtu(struct net_device *dev, int new_mtu); struct rtnl_link_stats64 *ip_tunnel_get_stats64(struct net_device *dev, @@@ -246,6 -248,7 +247,6 @@@ int ip_tunnel_changelink(struct net_dev int ip_tunnel_newlink(struct net_device *dev, struct nlattr *tb[], struct ip_tunnel_parm *p); void ip_tunnel_setup(struct net_device *dev, int net_id); -void ip_tunnel_dst_reset_all(struct ip_tunnel *t); int ip_tunnel_encap_setup(struct ip_tunnel *t, struct ip_tunnel_encap *ipencap); @@@ -270,15 -273,15 +271,15 @@@ static inline u8 ip_tunnel_ecn_encap(u return INET_ECN_encapsulate(tos, inner); } -int iptunnel_pull_header(struct sk_buff *skb, int hdr_len, __be16 inner_proto); +int iptunnel_pull_header(struct sk_buff *skb, int hdr_len, __be16 inner_proto, + bool xnet); void iptunnel_xmit(struct sock *sk, struct rtable *rt, struct sk_buff *skb, __be32 src, __be32 dst, u8 proto, u8 tos, u8 ttl, __be16 df, bool xnet); struct metadata_dst *iptunnel_metadata_reply(struct metadata_dst *md, gfp_t flags); -struct sk_buff *iptunnel_handle_offloads(struct sk_buff *skb, bool gre_csum, - int gso_type_mask); +struct sk_buff *iptunnel_handle_offloads(struct sk_buff *skb, int gso_type_mask); static inline void iptunnel_xmit_stats(struct net_device *dev, int pkt_len) { diff --combined include/net/tcp.h index 9b2cb0c,ae6468f..e90db85 --- a/include/net/tcp.h +++ b/include/net/tcp.h @@@ -239,6 -239,13 +239,6 @@@ extern struct inet_timewait_death_row t extern int sysctl_tcp_timestamps; extern int sysctl_tcp_window_scaling; extern int sysctl_tcp_sack; -extern int sysctl_tcp_fin_timeout; -extern int sysctl_tcp_syn_retries; -extern int sysctl_tcp_synack_retries; -extern int sysctl_tcp_retries1; -extern int sysctl_tcp_retries2; -extern int sysctl_tcp_orphan_retries; -extern int sysctl_tcp_syncookies; extern int sysctl_tcp_fastopen; extern int sysctl_tcp_retrans_collapse; extern int sysctl_tcp_stdurg; @@@ -267,6 -274,7 +267,6 @@@ extern int sysctl_tcp_thin_dupack extern int sysctl_tcp_early_retrans; extern int sysctl_tcp_limit_output_bytes; extern int sysctl_tcp_challenge_ack_limit; -extern unsigned int sysctl_tcp_notsent_lowat; extern int sysctl_tcp_min_tso_segs; extern int sysctl_tcp_min_rtt_wlen; extern int sysctl_tcp_autocorking; @@@ -439,7 -447,7 +439,7 @@@ const u8 *tcp_parse_md5sig_option(cons void tcp_v4_send_check(struct sock *sk, struct sk_buff *skb); void tcp_v4_mtu_reduced(struct sock *sk); - void tcp_req_err(struct sock *sk, u32 seq); + void tcp_req_err(struct sock *sk, u32 seq, bool abort); int tcp_v4_conn_request(struct sock *sk, struct sk_buff *skb); struct sock *tcp_create_openreq_child(const struct sock *sk, struct request_sock *req, @@@ -560,7 -568,6 +560,7 @@@ void tcp_rearm_rto(struct sock *sk) void tcp_synack_rtt_meas(struct sock *sk, struct request_sock *req); void tcp_reset(struct sock *sk); void tcp_skb_mark_lost_uncond_verify(struct tcp_sock *tp, struct sk_buff *skb); +void tcp_fin(struct sock *sk); /* tcp_timer.c */ void tcp_init_xmit_timers(struct sock *); @@@ -956,11 -963,9 +956,11 @@@ static inline void tcp_enable_fack(stru */ static inline void tcp_enable_early_retrans(struct tcp_sock *tp) { + struct net *net = sock_net((struct sock *)tp); + tp->do_early_retrans = sysctl_tcp_early_retrans && sysctl_tcp_early_retrans < 4 && !sysctl_tcp_thin_dupack && - sysctl_tcp_reordering == 3; + net->ipv4.sysctl_tcp_reordering == 3; } static inline void tcp_disable_early_retrans(struct tcp_sock *tp) @@@ -1247,7 -1252,7 +1247,7 @@@ static inline u32 keepalive_time_elapse static inline int tcp_fin_time(const struct sock *sk) { - int fin_timeout = tcp_sk(sk)->linger2 ? : sysctl_tcp_fin_timeout; + int fin_timeout = tcp_sk(sk)->linger2 ? : sock_net(sk)->ipv4.sysctl_tcp_fin_timeout; const int rto = inet_csk(sk)->icsk_rto; if (fin_timeout < (rto << 2) - (rto >> 1)) @@@ -1432,7 -1437,6 +1432,7 @@@ void tcp_free_fastopen_req(struct tcp_s extern struct tcp_fastopen_context __rcu *tcp_fastopen_ctx; int tcp_fastopen_reset_cipher(void *key, unsigned int len); +void tcp_fastopen_add_skb(struct sock *sk, struct sk_buff *skb); struct sock *tcp_try_fastopen(struct sock *sk, struct sk_buff *skb, struct request_sock *req, struct tcp_fastopen_cookie *foc, @@@ -1681,8 -1685,7 +1681,8 @@@ void __tcp_v4_send_check(struct sk_buf static inline u32 tcp_notsent_lowat(const struct tcp_sock *tp) { - return tp->notsent_lowat ?: sysctl_tcp_notsent_lowat; + struct net *net = sock_net((struct sock *)tp); + return tp->notsent_lowat ?: net->ipv4.sysctl_tcp_notsent_lowat; } static inline bool tcp_stream_memory_free(const struct sock *sk) diff --combined kernel/bpf/verifier.c index 36dc497,2e7f7ab..2e08f8e --- a/kernel/bpf/verifier.c +++ b/kernel/bpf/verifier.c @@@ -246,7 -246,6 +246,7 @@@ static const struct {BPF_MAP_TYPE_PROG_ARRAY, BPF_FUNC_tail_call}, {BPF_MAP_TYPE_PERF_EVENT_ARRAY, BPF_FUNC_perf_event_read}, {BPF_MAP_TYPE_PERF_EVENT_ARRAY, BPF_FUNC_perf_event_output}, + {BPF_MAP_TYPE_STACK_TRACE, BPF_FUNC_get_stackid}, }; static void print_verifier_state(struct verifier_env *env) @@@ -779,24 -778,15 +779,24 @@@ static int check_xadd(struct verifier_e * bytes from that pointer, make sure that it's within stack boundary * and all elements of stack are initialized */ -static int check_stack_boundary(struct verifier_env *env, - int regno, int access_size) +static int check_stack_boundary(struct verifier_env *env, int regno, + int access_size, bool zero_size_allowed) { struct verifier_state *state = &env->cur_state; struct reg_state *regs = state->regs; int off, i; - if (regs[regno].type != PTR_TO_STACK) + if (regs[regno].type != PTR_TO_STACK) { + if (zero_size_allowed && access_size == 0 && + regs[regno].type == CONST_IMM && + regs[regno].imm == 0) + return 0; + + verbose("R%d type=%s expected=%s\n", regno, + reg_type_str[regs[regno].type], + reg_type_str[PTR_TO_STACK]); return -EACCES; + } off = regs[regno].imm; if (off >= 0 || off < -MAX_BPF_STACK || off + access_size > 0 || @@@ -839,24 -829,15 +839,24 @@@ static int check_func_arg(struct verifi return 0; } - if (arg_type == ARG_PTR_TO_STACK || arg_type == ARG_PTR_TO_MAP_KEY || + if (arg_type == ARG_PTR_TO_MAP_KEY || arg_type == ARG_PTR_TO_MAP_VALUE) { expected_type = PTR_TO_STACK; - } else if (arg_type == ARG_CONST_STACK_SIZE) { + } else if (arg_type == ARG_CONST_STACK_SIZE || + arg_type == ARG_CONST_STACK_SIZE_OR_ZERO) { expected_type = CONST_IMM; } else if (arg_type == ARG_CONST_MAP_PTR) { expected_type = CONST_PTR_TO_MAP; } else if (arg_type == ARG_PTR_TO_CTX) { expected_type = PTR_TO_CTX; + } else if (arg_type == ARG_PTR_TO_STACK) { + expected_type = PTR_TO_STACK; + /* One exception here. In case function allows for NULL to be + * passed in as argument, it's a CONST_IMM type. Final test + * happens during stack boundary checking. + */ + if (reg->type == CONST_IMM && reg->imm == 0) + expected_type = CONST_IMM; } else { verbose("unsupported arg_type %d\n", arg_type); return -EFAULT; @@@ -886,8 -867,8 +886,8 @@@ verbose("invalid map_ptr to access map->key\n"); return -EACCES; } - err = check_stack_boundary(env, regno, (*mapp)->key_size); - + err = check_stack_boundary(env, regno, (*mapp)->key_size, + false); } else if (arg_type == ARG_PTR_TO_MAP_VALUE) { /* bpf_map_xxx(..., map_ptr, ..., value) call: * check [value, value + map->value_size) validity @@@ -897,12 -878,9 +897,12 @@@ verbose("invalid map_ptr to access map->value\n"); return -EACCES; } - err = check_stack_boundary(env, regno, (*mapp)->value_size); + err = check_stack_boundary(env, regno, (*mapp)->value_size, + false); + } else if (arg_type == ARG_CONST_STACK_SIZE || + arg_type == ARG_CONST_STACK_SIZE_OR_ZERO) { + bool zero_size_allowed = (arg_type == ARG_CONST_STACK_SIZE_OR_ZERO); - } else if (arg_type == ARG_CONST_STACK_SIZE) { /* bpf_xxx(..., buf, len) call will access 'len' bytes * from stack pointer 'buf'. Check it * note: regno == len, regno - 1 == buf @@@ -912,8 -890,7 +912,8 @@@ verbose("ARG_CONST_STACK_SIZE cannot be first argument\n"); return -EACCES; } - err = check_stack_boundary(env, regno - 1, reg->imm); + err = check_stack_boundary(env, regno - 1, reg->imm, + zero_size_allowed); } return err; @@@ -934,11 -911,8 +934,11 @@@ static int check_map_func_compatibility * don't allow any other map type to be passed into * the special func; */ - if (bool_func && bool_map != bool_func) + if (bool_func && bool_map != bool_func) { + verbose("cannot pass map_type %d into func %d\n", + map->map_type, func_id); return -EINVAL; + } } return 0; @@@ -2108,7 -2082,7 +2108,7 @@@ static void adjust_branches(struct bpf_ /* adjust offset of jmps if necessary */ if (i < pos && i + insn->off + 1 > pos) insn->off += delta; - else if (i > pos && i + insn->off + 1 < pos) + else if (i > pos + delta && i + insn->off + 1 <= pos + delta) insn->off -= delta; } } diff --combined lib/Kconfig.debug index f890ee5,8bfd1ac..60d09e9 --- a/lib/Kconfig.debug +++ b/lib/Kconfig.debug @@@ -1400,6 -1400,21 +1400,21 @@@ config RCU_EQS_DEBU endmenu # "RCU Debugging" + config DEBUG_WQ_FORCE_RR_CPU + bool "Force round-robin CPU selection for unbound work items" + depends on DEBUG_KERNEL + default n + help + Workqueue used to implicitly guarantee that work items queued + without explicit CPU specified are put on the local CPU. This + guarantee is no longer true and while local CPU is still + preferred work items may be put on foreign CPUs. Kernel + parameter "workqueue.debug_force_rr_cpu" is added to force + round-robin CPU selection to flush out usages which depend on the + now broken guarantee. This config option enables the debug + feature by default. When enabled, memory and cache locality will + be impacted. + config DEBUG_BLOCK_EXT_DEVT bool "Force extended block device numbers and spread them" depends on DEBUG_KERNEL @@@ -1738,14 -1753,6 +1753,14 @@@ config TEST_KSTRTO config TEST_PRINTF tristate "Test printf() family of functions at runtime" +config TEST_BITMAP + tristate "Test bitmap_*() family of functions at runtime" + default n + help + Enable this option to test the bitmap functions at boot. + + If unsure, say N. + config TEST_RHASHTABLE tristate "Perform selftest on resizable hash table" default n diff --combined net/batman-adv/gateway_client.c index 261866e,ccf70be..4b598bd --- a/net/batman-adv/gateway_client.c +++ b/net/batman-adv/gateway_client.c @@@ -1,4 -1,4 +1,4 @@@ -/* Copyright (C) 2009-2015 B.A.T.M.A.N. contributors: +/* Copyright (C) 2009-2016 B.A.T.M.A.N. contributors: * * Marek Lindner * @@@ -28,7 -28,6 +28,7 @@@ #include <linux/ip.h> #include <linux/ipv6.h> #include <linux/kernel.h> +#include <linux/kref.h> #include <linux/list.h> #include <linux/netdevice.h> #include <linux/rculist.h> @@@ -60,29 -59,12 +60,29 @@@ */ #define BATADV_DHCP_CHADDR_OFFSET 28 +/** + * batadv_gw_node_release - release gw_node from lists and queue for free after + * rcu grace period + * @ref: kref pointer of the gw_node + */ +static void batadv_gw_node_release(struct kref *ref) +{ + struct batadv_gw_node *gw_node; + + gw_node = container_of(ref, struct batadv_gw_node, refcount); + + batadv_orig_node_free_ref(gw_node->orig_node); + kfree_rcu(gw_node, rcu); +} + +/** + * batadv_gw_node_free_ref - decrement the gw_node refcounter and possibly + * release it + * @gw_node: gateway node to free + */ static void batadv_gw_node_free_ref(struct batadv_gw_node *gw_node) { - if (atomic_dec_and_test(&gw_node->refcount)) { - batadv_orig_node_free_ref(gw_node->orig_node); - kfree_rcu(gw_node, rcu); - } + kref_put(&gw_node->refcount, batadv_gw_node_release); } static struct batadv_gw_node * @@@ -95,7 -77,7 +95,7 @@@ batadv_gw_get_selected_gw_node(struct b if (!gw_node) goto out; - if (!atomic_inc_not_zero(&gw_node->refcount)) + if (!kref_get_unless_zero(&gw_node->refcount)) gw_node = NULL; out: @@@ -118,7 -100,7 +118,7 @@@ batadv_gw_get_selected_orig(struct bata if (!orig_node) goto unlock; - if (!atomic_inc_not_zero(&orig_node->refcount)) + if (!kref_get_unless_zero(&orig_node->refcount)) orig_node = NULL; unlock: @@@ -136,7 -118,7 +136,7 @@@ static void batadv_gw_select(struct bat spin_lock_bh(&bat_priv->gw.list_lock); - if (new_gw_node && !atomic_inc_not_zero(&new_gw_node->refcount)) + if (new_gw_node && !kref_get_unless_zero(&new_gw_node->refcount)) new_gw_node = NULL; curr_gw_node = rcu_dereference_protected(bat_priv->gw.curr_gw, 1); @@@ -188,7 -170,7 +188,7 @@@ batadv_gw_get_best_gw_node(struct batad if (!router_ifinfo) goto next; - if (!atomic_inc_not_zero(&gw_node->refcount)) + if (!kref_get_unless_zero(&gw_node->refcount)) goto next; tq_avg = router_ifinfo->bat_iv.tq_avg; @@@ -206,7 -188,7 +206,7 @@@ if (curr_gw) batadv_gw_node_free_ref(curr_gw); curr_gw = gw_node; - atomic_inc(&curr_gw->refcount); + kref_get(&curr_gw->refcount); } break; @@@ -221,7 -203,7 +221,7 @@@ if (curr_gw) batadv_gw_node_free_ref(curr_gw); curr_gw = gw_node; - atomic_inc(&curr_gw->refcount); + kref_get(&curr_gw->refcount); } break; } @@@ -441,7 -423,7 +441,7 @@@ static void batadv_gw_node_add(struct b if (gateway->bandwidth_down == 0) return; - if (!atomic_inc_not_zero(&orig_node->refcount)) + if (!kref_get_unless_zero(&orig_node->refcount)) return; gw_node = kzalloc(sizeof(*gw_node), GFP_ATOMIC); @@@ -454,7 -436,7 +454,7 @@@ gw_node->orig_node = orig_node; gw_node->bandwidth_down = ntohl(gateway->bandwidth_down); gw_node->bandwidth_up = ntohl(gateway->bandwidth_up); - atomic_set(&gw_node->refcount, 1); + kref_init(&gw_node->refcount); spin_lock_bh(&bat_priv->gw.list_lock); hlist_add_head_rcu(&gw_node->list, &bat_priv->gw.list); @@@ -474,7 -456,7 +474,7 @@@ * @bat_priv: the bat priv with all the soft interface information * @orig_node: originator announcing gateway capabilities * - * Returns gateway node if found or NULL otherwise. + * Return: gateway node if found or NULL otherwise. */ static struct batadv_gw_node * batadv_gw_node_get(struct batadv_priv *bat_priv, @@@ -487,7 -469,7 +487,7 @@@ if (gw_node_tmp->orig_node != orig_node) continue; - if (!atomic_inc_not_zero(&gw_node_tmp->refcount)) + if (!kref_get_unless_zero(&gw_node_tmp->refcount)) continue; gw_node = gw_node_tmp; @@@ -545,11 -527,12 +545,12 @@@ void batadv_gw_node_update(struct batad * gets dereferenced. */ spin_lock_bh(&bat_priv->gw.list_lock); - hlist_del_init_rcu(&gw_node->list); + if (!hlist_unhashed(&gw_node->list)) { + hlist_del_init_rcu(&gw_node->list); + batadv_gw_node_free_ref(gw_node); + } spin_unlock_bh(&bat_priv->gw.list_lock); - batadv_gw_node_free_ref(gw_node); - curr_gw = batadv_gw_get_selected_gw_node(bat_priv); if (gw_node == curr_gw) batadv_gw_reselect(bat_priv); @@@ -673,13 -656,13 +674,13 @@@ out * @chaddr: buffer where the client address will be stored. Valid * only if the function returns BATADV_DHCP_TO_CLIENT * - * Returns: + * This function may re-allocate the data buffer of the skb passed as argument. + * + * Return: * - BATADV_DHCP_NO if the packet is not a dhcp message or if there was an error * while parsing it * - BATADV_DHCP_TO_SERVER if this is a message going to the DHCP server * - BATADV_DHCP_TO_CLIENT if this is a message going to a DHCP client - * - * This function may re-allocate the data buffer of the skb passed as argument. */ enum batadv_dhcp_recipient batadv_gw_dhcp_recipient_get(struct sk_buff *skb, unsigned int *header_len, @@@ -794,11 -777,11 +795,11 @@@ * server. Due to topology changes it may be the case that the GW server * previously selected is not the best one anymore. * - * Returns true if the packet destination is unicast and it is not the best gw, - * false otherwise. - * * This call might reallocate skb data. * Must be invoked only when the DHCP packet is going TO a DHCP SERVER. + * + * Return: true if the packet destination is unicast and it is not the best gw, + * false otherwise. */ bool batadv_gw_out_of_range(struct batadv_priv *bat_priv, struct sk_buff *skb) diff --combined net/batman-adv/hard-interface.c index fb2d9c0,57f71071..e2aaa4c --- a/net/batman-adv/hard-interface.c +++ b/net/batman-adv/hard-interface.c @@@ -1,4 -1,4 +1,4 @@@ -/* Copyright (C) 2007-2015 B.A.T.M.A.N. contributors: +/* Copyright (C) 2007-2016 B.A.T.M.A.N. contributors: * * Marek Lindner, Simon Wunderlich * @@@ -18,7 -18,6 +18,7 @@@ #include "hard-interface.h" #include "main.h" +#include <linux/atomic.h> #include <linux/bug.h> #include <linux/byteorder/generic.h> #include <linux/errno.h> @@@ -27,7 -26,6 +27,7 @@@ #include <linux/if_ether.h> #include <linux/if.h> #include <linux/kernel.h> +#include <linux/kref.h> #include <linux/list.h> #include <linux/netdevice.h> #include <linux/printk.h> @@@ -49,19 -47,13 +49,19 @@@ #include "sysfs.h" #include "translation-table.h" -void batadv_hardif_free_rcu(struct rcu_head *rcu) +/** + * batadv_hardif_release - release hard interface from lists and queue for + * free after rcu grace period + * @ref: kref pointer of the hard interface + */ +void batadv_hardif_release(struct kref *ref) { struct batadv_hard_iface *hard_iface; - hard_iface = container_of(rcu, struct batadv_hard_iface, rcu); + hard_iface = container_of(ref, struct batadv_hard_iface, refcount); dev_put(hard_iface->net_dev); - kfree(hard_iface); + + kfree_rcu(hard_iface, rcu); } struct batadv_hard_iface * @@@ -72,7 -64,7 +72,7 @@@ batadv_hardif_get_by_netdev(const struc rcu_read_lock(); list_for_each_entry_rcu(hard_iface, &batadv_hardif_list, list) { if (hard_iface->net_dev == net_dev && - atomic_inc_not_zero(&hard_iface->refcount)) + kref_get_unless_zero(&hard_iface->refcount)) goto out; } @@@ -84,6 -76,28 +84,28 @@@ out } /** + * batadv_mutual_parents - check if two devices are each others parent + * @dev1: 1st net_device + * @dev2: 2nd net_device + * + * veth devices come in pairs and each is the parent of the other! + * + * Return: true if the devices are each others parent, otherwise false + */ + static bool batadv_mutual_parents(const struct net_device *dev1, + const struct net_device *dev2) + { + int dev1_parent_iflink = dev_get_iflink(dev1); + int dev2_parent_iflink = dev_get_iflink(dev2); + + if (!dev1_parent_iflink || !dev2_parent_iflink) + return false; + + return (dev1_parent_iflink == dev2->ifindex) && + (dev2_parent_iflink == dev1->ifindex); + } + + /** * batadv_is_on_batman_iface - check if a device is a batman iface descendant * @net_dev: the device to check * @@@ -93,7 -107,7 +115,7 @@@ * This function recursively checks all the fathers of the device passed as * argument looking for a batman-adv soft interface. * - * Returns true if the device is descendant of a batman-adv mesh interface (or + * Return: true if the device is descendant of a batman-adv mesh interface (or * if it is a batman-adv interface itself), false otherwise */ static bool batadv_is_on_batman_iface(const struct net_device *net_dev) @@@ -116,6 -130,9 +138,9 @@@ if (WARN(!parent_dev, "Cannot find parent device")) return false; + if (batadv_mutual_parents(net_dev, parent_dev)) + return false; + ret = batadv_is_on_batman_iface(parent_dev); return ret; @@@ -144,7 -161,7 +169,7 @@@ static int batadv_is_valid_iface(const * interface * @net_device: the device to check * - * Returns true if the net device is a 802.11 wireless device, false otherwise. + * Return: true if the net device is a 802.11 wireless device, false otherwise. */ bool batadv_is_wifi_netdev(struct net_device *net_device) { @@@ -177,7 -194,7 +202,7 @@@ batadv_hardif_get_active(const struct n continue; if (hard_iface->if_status == BATADV_IF_ACTIVE && - atomic_inc_not_zero(&hard_iface->refcount)) + kref_get_unless_zero(&hard_iface->refcount)) goto out; } @@@ -211,7 -228,7 +236,7 @@@ static void batadv_primary_if_select(st ASSERT_RTNL(); - if (new_hard_iface && !atomic_inc_not_zero(&new_hard_iface->refcount)) + if (new_hard_iface && !kref_get_unless_zero(&new_hard_iface->refcount)) new_hard_iface = NULL; curr_hard_iface = rcu_dereference_protected(bat_priv->primary_if, 1); @@@ -409,8 -426,7 +434,8 @@@ batadv_hardif_deactivate_interface(stru * * Invoke ndo_del_slave on master passing slave as argument. In this way slave * is free'd and master can correctly change its internal state. - * Return 0 on success, a negative value representing the error otherwise + * + * Return: 0 on success, a negative value representing the error otherwise */ static int batadv_master_del_slave(struct batadv_hard_iface *slave, struct net_device *master) @@@ -439,7 -455,7 +464,7 @@@ int batadv_hardif_enable_interface(stru if (hard_iface->if_status != BATADV_IF_NOT_IN_USE) goto out; - if (!atomic_inc_not_zero(&hard_iface->refcount)) + if (!kref_get_unless_zero(&hard_iface->refcount)) goto out; soft_iface = dev_get_by_name(&init_net, iface_name); @@@ -660,8 -676,7 +685,8 @@@ batadv_hardif_add_interface(struct net_ hard_iface->num_bcasts = BATADV_NUM_BCASTS_WIRELESS; /* extra reference for return */ - atomic_set(&hard_iface->refcount, 2); + kref_init(&hard_iface->refcount); + kref_get(&hard_iface->refcount); batadv_check_known_mac_addr(hard_iface->net_dev); list_add_tail_rcu(&hard_iface->list, &batadv_hardif_list); diff --combined net/batman-adv/translation-table.c index 1188279,0e80fd1..5c7fa02 --- a/net/batman-adv/translation-table.c +++ b/net/batman-adv/translation-table.c @@@ -1,4 -1,4 +1,4 @@@ -/* Copyright (C) 2007-2015 B.A.T.M.A.N. contributors: +/* Copyright (C) 2007-2016 B.A.T.M.A.N. contributors: * * Marek Lindner, Simon Wunderlich, Antonio Quartulli * @@@ -31,7 -31,6 +31,7 @@@ #include <linux/jhash.h> #include <linux/jiffies.h> #include <linux/kernel.h> +#include <linux/kref.h> #include <linux/list.h> #include <linux/lockdep.h> #include <linux/netdevice.h> @@@ -69,15 -68,7 +69,15 @@@ static void batadv_tt_global_del(struc unsigned short vid, const char *message, bool roaming); -/* returns 1 if they are the same mac addr and vid */ +/** + * batadv_compare_tt - check if two TT entries are the same + * @node: the list element pointer of the first TT entry + * @data2: pointer to the tt_common_entry of the second TT entry + * + * Compare the MAC address and the VLAN ID of the two TT entries and check if + * they are the same TT client. + * Return: 1 if the two TT clients are the same, 0 otherwise + */ static int batadv_compare_tt(const struct hlist_node *node, const void *data2) { const void *data1 = container_of(node, struct batadv_tt_common_entry, @@@ -93,7 -84,7 +93,7 @@@ * @data: pointer to the tt_common_entry object to map * @size: the size of the hash table * - * Returns the hash index where the object represented by 'data' should be + * Return: the hash index where the object represented by 'data' should be * stored at. */ static inline u32 batadv_choose_tt(const void *data, u32 size) @@@ -114,7 -105,7 +114,7 @@@ * @addr: the mac address of the client to look for * @vid: VLAN identifier * - * Returns a pointer to the tt_common struct belonging to the searched client if + * Return: a pointer to the tt_common struct belonging to the searched client if * found, NULL otherwise. */ static struct batadv_tt_common_entry * @@@ -142,7 -133,7 +142,7 @@@ batadv_tt_hash_find(struct batadv_hasht if (tt->vid != vid) continue; - if (!atomic_inc_not_zero(&tt->refcount)) + if (!kref_get_unless_zero(&tt->refcount)) continue; tt_tmp = tt; @@@ -159,7 -150,7 +159,7 @@@ * @addr: the mac address of the client to look for * @vid: VLAN identifier * - * Returns a pointer to the corresponding tt_local_entry struct if the client is + * Return: a pointer to the corresponding tt_local_entry struct if the client is * found, NULL otherwise. */ static struct batadv_tt_local_entry * @@@ -184,7 -175,7 +184,7 @@@ batadv_tt_local_hash_find(struct batadv * @addr: the mac address of the client to look for * @vid: VLAN identifier * - * Returns a pointer to the corresponding tt_global_entry struct if the client + * Return: a pointer to the corresponding tt_global_entry struct if the client * is found, NULL otherwise. */ static struct batadv_tt_global_entry * @@@ -203,68 -194,34 +203,68 @@@ batadv_tt_global_hash_find(struct batad return tt_global_entry; } +/** + * batadv_tt_local_entry_release - release tt_local_entry from lists and queue + * for free after rcu grace period + * @ref: kref pointer of the nc_node + */ +static void batadv_tt_local_entry_release(struct kref *ref) +{ + struct batadv_tt_local_entry *tt_local_entry; + + tt_local_entry = container_of(ref, struct batadv_tt_local_entry, + common.refcount); + + kfree_rcu(tt_local_entry, common.rcu); +} + +/** + * batadv_tt_local_entry_free_ref - decrement the tt_local_entry refcounter and + * possibly release it + * @tt_local_entry: tt_local_entry to be free'd + */ static void batadv_tt_local_entry_free_ref(struct batadv_tt_local_entry *tt_local_entry) { - if (atomic_dec_and_test(&tt_local_entry->common.refcount)) - kfree_rcu(tt_local_entry, common.rcu); + kref_put(&tt_local_entry->common.refcount, + batadv_tt_local_entry_release); } /** - * batadv_tt_global_entry_free_ref - decrement the refcounter for a - * tt_global_entry and possibly free it - * @tt_global_entry: the object to free + * batadv_tt_global_entry_release - release tt_global_entry from lists and queue + * for free after rcu grace period + * @ref: kref pointer of the nc_node + */ +static void batadv_tt_global_entry_release(struct kref *ref) +{ + struct batadv_tt_global_entry *tt_global_entry; + + tt_global_entry = container_of(ref, struct batadv_tt_global_entry, + common.refcount); + + batadv_tt_global_del_orig_list(tt_global_entry); + kfree_rcu(tt_global_entry, common.rcu); +} + +/** + * batadv_tt_global_entry_free_ref - decrement the tt_global_entry refcounter + * and possibly release it + * @tt_global_entry: tt_global_entry to be free'd */ static void batadv_tt_global_entry_free_ref(struct batadv_tt_global_entry *tt_global_entry) { - if (atomic_dec_and_test(&tt_global_entry->common.refcount)) { - batadv_tt_global_del_orig_list(tt_global_entry); - kfree_rcu(tt_global_entry, common.rcu); - } + kref_put(&tt_global_entry->common.refcount, + batadv_tt_global_entry_release); } /** * batadv_tt_global_hash_count - count the number of orig entries - * @hash: hash table containing the tt entries + * @bat_priv: the bat priv with all the soft interface information * @addr: the mac address of the client to count entries for * @vid: VLAN identifier * - * Return the number of originators advertising the given address/data + * Return: the number of originators advertising the given address/data * (excluding ourself). */ int batadv_tt_global_hash_count(struct batadv_priv *bat_priv, @@@ -329,9 -286,9 +329,9 @@@ static void batadv_tt_local_size_dec(st } /** - * batadv_tt_global_size_mod - change the size by v of the local table - * identified by vid - * @bat_priv: the bat priv with all the soft interface information + * batadv_tt_global_size_mod - change the size by v of the global table + * for orig_node identified by vid + * @orig_node: the originator for which the table has to be modified * @vid: the VLAN identifier * @v: the amount to sum to the global table size */ @@@ -346,9 -303,11 +346,11 @@@ static void batadv_tt_global_size_mod(s if (atomic_add_return(v, &vlan->tt.num_entries) == 0) { spin_lock_bh(&orig_node->vlan_list_lock); - hlist_del_init_rcu(&vlan->list); + if (!hlist_unhashed(&vlan->list)) { + hlist_del_init_rcu(&vlan->list); + batadv_orig_node_vlan_free_ref(vlan); + } spin_unlock_bh(&orig_node->vlan_list_lock); - batadv_orig_node_vlan_free_ref(vlan); } batadv_orig_node_vlan_free_ref(vlan); @@@ -381,28 -340,22 +383,28 @@@ static void batadv_tt_global_size_dec(s /** * batadv_tt_orig_list_entry_release - release tt orig entry from lists and * queue for free after rcu grace period - * @orig_entry: tt orig entry to be free'd + * @ref: kref pointer of the tt orig entry */ -static void -batadv_tt_orig_list_entry_release(struct batadv_tt_orig_list_entry *orig_entry) +static void batadv_tt_orig_list_entry_release(struct kref *ref) { + struct batadv_tt_orig_list_entry *orig_entry; + + orig_entry = container_of(ref, struct batadv_tt_orig_list_entry, + refcount); + batadv_orig_node_free_ref(orig_entry->orig_node); kfree_rcu(orig_entry, rcu); } +/** + * batadv_tt_orig_list_entry_free_ref - decrement the tt orig entry refcounter + * and possibly release it + * @orig_entry: tt orig entry to be free'd + */ static void batadv_tt_orig_list_entry_free_ref(struct batadv_tt_orig_list_entry *orig_entry) { - if (!atomic_dec_and_test(&orig_entry->refcount)) - return; - - batadv_tt_orig_list_entry_release(orig_entry); + kref_put(&orig_entry->refcount, batadv_tt_orig_list_entry_release); } /** @@@ -484,7 -437,7 +486,7 @@@ unlock * batadv_tt_len - compute length in bytes of given number of tt changes * @changes_num: number of tt changes * - * Returns computed length in bytes. + * Return: computed length in bytes. */ static int batadv_tt_len(int changes_num) { @@@ -495,7 -448,7 +497,7 @@@ * batadv_tt_entries - compute the number of entries fitting in tt_len bytes * @tt_len: available space * - * Returns the number of entries. + * Return: the number of entries. */ static u16 batadv_tt_entries(u16 tt_len) { @@@ -507,7 -460,7 +509,7 @@@ * size when transmitted over the air * @bat_priv: the bat priv with all the soft interface information * - * Returns local translation table size in bytes. + * Return: local translation table size in bytes. */ static int batadv_tt_local_table_transmit_size(struct batadv_priv *bat_priv) { @@@ -573,7 -526,7 +575,7 @@@ static void batadv_tt_global_free(struc * @mark: the value contained in the skb->mark field of the received packet (if * any) * - * Returns true if the client was successfully added, false otherwise. + * Return: true if the client was successfully added, false otherwise. */ bool batadv_tt_local_add(struct net_device *soft_iface, const u8 *addr, unsigned short vid, int ifindex, u32 mark) @@@ -667,8 -620,7 +669,8 @@@ tt_local->common.vid = vid; if (batadv_is_wifi_netdev(in_dev)) tt_local->common.flags |= BATADV_TT_CLIENT_WIFI; - atomic_set(&tt_local->common.refcount, 2); + kref_init(&tt_local->common.refcount); + kref_get(&tt_local->common.refcount); tt_local->last_seen = jiffies; tt_local->common.added_at = tt_local->last_seen; @@@ -769,11 -721,12 +771,11 @@@ out * function reserves the amount of space needed to send the entire global TT * table. In case of success the value is updated with the real amount of * reserved bytes - * Allocate the needed amount of memory for the entire TT TVLV and write its * header made up by one tvlv_tt_data object and a series of tvlv_tt_vlan_data * objects, one per active VLAN served by the originator node. * - * Return the size of the allocated buffer or 0 in case of failure. + * Return: the size of the allocated buffer or 0 in case of failure. */ static u16 batadv_tt_prepare_tvlv_global_data(struct batadv_orig_node *orig_node, @@@ -847,7 -800,7 +849,7 @@@ out * header made up by one tvlv_tt_data object and a series of tvlv_tt_vlan_data * objects, one per active VLAN. * - * Return the size of the allocated buffer or 0 in case of failure. + * Return: the size of the allocated buffer or 0 in case of failure. */ static u16 batadv_tt_prepare_tvlv_local_data(struct batadv_priv *bat_priv, @@@ -1089,7 -1042,7 +1091,7 @@@ batadv_tt_local_set_pending(struct bata * @message: message to append to the log on deletion * @roaming: true if the deletion is due to a roaming event * - * Returns the flags assigned to the local entry before being deleted + * Return: the flags assigned to the local entry before being deleted */ u16 batadv_tt_local_remove(struct batadv_priv *bat_priv, const u8 *addr, unsigned short vid, const char *message, @@@ -1289,16 -1242,10 +1291,16 @@@ static void batadv_tt_changes_list_free spin_unlock_bh(&bat_priv->tt.changes_list_lock); } -/* retrieves the orig_tt_list_entry belonging to orig_node from the +/** + * batadv_tt_global_orig_entry_find - find a TT orig_list_entry + * @entry: the TT global entry where the orig_list_entry has to be + * extracted from + * @orig_node: the originator for which the orig_list_entry has to be found + * + * retrieve the orig_tt_list_entry belonging to orig_node from the * batadv_tt_global_entry list * - * returns it with an increased refcounter, NULL if not found + * Return: it with an increased refcounter, NULL if not found */ static struct batadv_tt_orig_list_entry * batadv_tt_global_orig_entry_find(const struct batadv_tt_global_entry *entry, @@@ -1312,7 -1259,7 +1314,7 @@@ hlist_for_each_entry_rcu(tmp_orig_entry, head, list) { if (tmp_orig_entry->orig_node != orig_node) continue; - if (!atomic_inc_not_zero(&tmp_orig_entry->refcount)) + if (!kref_get_unless_zero(&tmp_orig_entry->refcount)) continue; orig_entry = tmp_orig_entry; @@@ -1323,15 -1270,8 +1325,15 @@@ return orig_entry; } -/* find out if an orig_node is already in the list of a tt_global_entry. - * returns true if found, false otherwise +/** + * batadv_tt_global_entry_has_orig - check if a TT global entry is also handled + * by a given originator + * @entry: the TT global entry to check + * @orig_node: the originator to search in the list + * + * find out if an orig_node is already in the list of a tt_global_entry. + * + * Return: true if found, false otherwise */ static bool batadv_tt_global_entry_has_orig(const struct batadv_tt_global_entry *entry, @@@ -1369,12 -1309,11 +1371,12 @@@ batadv_tt_global_orig_entry_add(struct goto out; INIT_HLIST_NODE(&orig_entry->list); - atomic_inc(&orig_node->refcount); + kref_get(&orig_node->refcount); batadv_tt_global_size_inc(orig_node, tt_global->common.vid); orig_entry->orig_node = orig_node; orig_entry->ttvn = ttvn; - atomic_set(&orig_entry->refcount, 2); + kref_init(&orig_entry->refcount); + kref_get(&orig_entry->refcount); spin_lock_bh(&tt_global->list_lock); hlist_add_head_rcu(&orig_entry->list, @@@ -1404,7 -1343,7 +1406,7 @@@ out * * The caller must hold orig_node refcount. * - * Return true if the new entry has been added, false otherwise + * Return: true if the new entry has been added, false otherwise */ static bool batadv_tt_global_add(struct batadv_priv *bat_priv, struct batadv_orig_node *orig_node, @@@ -1450,8 -1389,7 +1452,8 @@@ */ if (flags & BATADV_TT_CLIENT_ROAM) tt_global_entry->roam_at = jiffies; - atomic_set(&common->refcount, 2); + kref_init(&common->refcount); + kref_get(&common->refcount); common->added_at = jiffies; INIT_HLIST_HEAD(&tt_global_entry->orig_list); @@@ -1563,7 -1501,7 +1565,7 @@@ out * @tt_global_entry: global translation table entry to be analyzed * * This functon assumes the caller holds rcu_read_lock(). - * Returns best originator list entry or NULL on errors. + * Return: best originator list entry or NULL on errors. */ static struct batadv_tt_orig_list_entry * batadv_transtable_best_orig(struct batadv_priv *bat_priv, @@@ -2093,7 -2031,7 +2095,7 @@@ _batadv_is_ap_isolated(struct batadv_tt * @addr: mac address of the destination client * @vid: VLAN identifier * - * Returns a pointer to the originator that was selected as destination in the + * Return: a pointer to the originator that was selected as destination in the * mesh for contacting the client 'addr', NULL otherwise. * In case of multiple originators serving the same client, the function returns * the best one (best in terms of metric towards the destination node). @@@ -2133,7 -2071,7 +2135,7 @@@ struct batadv_orig_node *batadv_transta /* found anything? */ if (best_entry) orig_node = best_entry->orig_node; - if (orig_node && !atomic_inc_not_zero(&orig_node->refcount)) + if (orig_node && !kref_get_unless_zero(&orig_node->refcount)) orig_node = NULL; rcu_read_unlock(); @@@ -2168,7 -2106,7 +2170,7 @@@ out * because the XOR operation can combine them all while trying to reduce the * noise as much as possible. * - * Returns the checksum of the global table of a given originator. + * Return: the checksum of the global table of a given originator. */ static u32 batadv_tt_global_crc(struct batadv_priv *bat_priv, struct batadv_orig_node *orig_node, @@@ -2245,7 -2183,7 +2247,7 @@@ * For details about the computation, please refer to the documentation for * batadv_tt_global_crc(). * - * Returns the checksum of the local table + * Return: the checksum of the local table */ static u32 batadv_tt_local_crc(struct batadv_priv *bat_priv, unsigned short vid) @@@ -2351,7 -2289,7 +2353,7 @@@ static void batadv_tt_req_purge(struct * @bat_priv: the bat priv with all the soft interface information * @orig_node: orig node this request is being issued for * - * Returns the pointer to the new tt_req_node struct if no request + * Return: the pointer to the new tt_req_node struct if no request * has already been issued for this orig_node, NULL otherwise. */ static struct batadv_tt_req_node * @@@ -2386,7 -2324,7 +2388,7 @@@ unlock * @entry_ptr: to be checked local tt entry * @data_ptr: not used but definition required to satisfy the callback prototype * - * Returns 1 if the entry is a valid, 0 otherwise. + * Return: 1 if the entry is a valid, 0 otherwise. */ static int batadv_tt_local_valid(const void *entry_ptr, const void *data_ptr) { @@@ -2470,8 -2408,9 +2472,8 @@@ static void batadv_tt_tvlv_generate(str * @orig_node: originator for which the CRCs have to be checked * @tt_vlan: pointer to the first tvlv VLAN entry * @num_vlan: number of tvlv VLAN entries - * @create: if true, create VLAN objects if not found * - * Return true if all the received CRCs match the locally stored ones, false + * Return: true if all the received CRCs match the locally stored ones, false * otherwise */ static bool batadv_tt_global_check_crc(struct batadv_orig_node *orig_node, @@@ -2574,8 -2513,6 +2576,8 @@@ static void batadv_tt_global_update_crc * @num_vlan: number of tvlv VLAN entries * @full_table: ask for the entire translation table if true, while only for the * last TT diff otherwise + * + * Return: true if the TT Request was sent, false otherwise */ static int batadv_send_tt_request(struct batadv_priv *bat_priv, struct batadv_orig_node *dst_orig_node, @@@ -2656,7 -2593,7 +2658,7 @@@ out * @req_src: mac address of tt request sender * @req_dst: mac address of tt request recipient * - * Returns true if tt request reply was sent, false otherwise. + * Return: true if tt request reply was sent, false otherwise. */ static bool batadv_send_other_tt_response(struct batadv_priv *bat_priv, struct batadv_tvlv_tt_data *tt_data, @@@ -2788,7 -2725,7 +2790,7 @@@ out * @tt_data: tt data containing the tt request information * @req_src: mac address of tt request sender * - * Returns true if tt request reply was sent, false otherwise. + * Return: true if tt request reply was sent, false otherwise. */ static bool batadv_send_my_tt_response(struct batadv_priv *bat_priv, struct batadv_tvlv_tt_data *tt_data, @@@ -2906,7 -2843,7 +2908,7 @@@ out * @req_src: mac address of tt request sender * @req_dst: mac address of tt request recipient * - * Returns true if tt request reply was sent, false otherwise. + * Return: true if tt request reply was sent, false otherwise. */ static bool batadv_send_tt_response(struct batadv_priv *bat_priv, struct batadv_tvlv_tt_data *tt_data, @@@ -3001,7 -2938,7 +3003,7 @@@ static void batadv_tt_update_changes(st * @addr: the mac address of the client to check * @vid: VLAN identifier * - * Returns true if the client is served by this node, false otherwise. + * Return: true if the client is served by this node, false otherwise. */ bool batadv_is_my_client(struct batadv_priv *bat_priv, const u8 *addr, unsigned short vid) @@@ -3118,16 -3055,11 +3120,16 @@@ static void batadv_tt_roam_purge(struc spin_unlock_bh(&bat_priv->tt.roam_list_lock); } -/* This function checks whether the client already reached the +/** + * batadv_tt_check_roam_count - check if a client has roamed too frequently + * @bat_priv: the bat priv with all the soft interface information + * @client: mac address of the roaming client + * + * This function checks whether the client already reached the * maximum number of possible roaming phases. In this case the ROAMING_ADV * will not be sent. * - * returns true if the ROAMING_ADV can be sent, false otherwise + * Return: true if the ROAMING_ADV can be sent, false otherwise */ static bool batadv_tt_check_roam_count(struct batadv_priv *bat_priv, u8 *client) { @@@ -3439,12 -3371,13 +3441,12 @@@ out * batadv_tt_update_orig - update global translation table with new tt * information received via ogms * @bat_priv: the bat priv with all the soft interface information - * @orig: the orig_node of the ogm - * @tt_vlan: pointer to the first tvlv VLAN entry + * @orig_node: the orig_node of the ogm + * @tt_buff: pointer to the first tvlv VLAN entry * @tt_num_vlan: number of tvlv VLAN entries * @tt_change: pointer to the first entry in the TT buffer * @tt_num_changes: number of tt changes inside the tt buffer * @ttvn: translation table version number of this changeset - * @tt_crc: crc32 checksum of orig node's translation table */ static void batadv_tt_update_orig(struct batadv_priv *bat_priv, struct batadv_orig_node *orig_node, @@@ -3526,7 -3459,7 +3528,7 @@@ request_table * @addr: the mac address of the client to check * @vid: VLAN identifier * - * Returns true if we know that the client has moved from its old originator + * Return: true if we know that the client has moved from its old originator * to another one. This entry is still kept for consistency purposes and will be * deleted later by a DEL or because of timeout */ @@@ -3552,7 -3485,7 +3554,7 @@@ out * @addr: the mac address of the local client to query * @vid: VLAN identifier * - * Returns true if the local client is known to be roaming (it is not served by + * Return: true if the local client is known to be roaming (it is not served by * this node anymore) or not. If yes, the client is still present in the table * to keep the latter consistent with the node TTVN */ @@@ -3681,7 -3614,7 +3683,7 @@@ static void batadv_tt_tvlv_ogm_handler_ * @tvlv_value: tvlv buffer containing the tt data * @tvlv_value_len: tvlv buffer length * - * Returns NET_RX_DROP if the tt tvlv is to be re-routed, NET_RX_SUCCESS + * Return: NET_RX_DROP if the tt tvlv is to be re-routed, NET_RX_SUCCESS * otherwise. */ static int batadv_tt_tvlv_unicast_handler_v1(struct batadv_priv *bat_priv, @@@ -3762,7 -3695,7 +3764,7 @@@ * @tvlv_value: tvlv buffer containing the tt data * @tvlv_value_len: tvlv buffer length * - * Returns NET_RX_DROP if the tt roam tvlv is to be re-routed, NET_RX_SUCCESS + * Return: NET_RX_DROP if the tt roam tvlv is to be re-routed, NET_RX_SUCCESS * otherwise. */ static int batadv_roam_tvlv_unicast_handler_v1(struct batadv_priv *bat_priv, @@@ -3808,7 -3741,7 +3810,7 @@@ out * batadv_tt_init - initialise the translation table internals * @bat_priv: the bat priv with all the soft interface information * - * Return 0 on success or negative error number in case of failure. + * Return: 0 on success or negative error number in case of failure. */ int batadv_tt_init(struct batadv_priv *bat_priv) { @@@ -3846,7 -3779,7 +3848,7 @@@ * @addr: the mac address of the client * @vid: the identifier of the VLAN where this client is connected * - * Returns true if the client is marked with the TT_CLIENT_ISOLA flag, false + * Return: true if the client is marked with the TT_CLIENT_ISOLA flag, false * otherwise */ bool batadv_tt_global_is_isolated(struct batadv_priv *bat_priv, diff --combined net/bridge/br_mdb.c index cf51b7b,74c278e..73786e2 --- a/net/bridge/br_mdb.c +++ b/net/bridge/br_mdb.c @@@ -41,14 -41,6 +41,14 @@@ fail return -EMSGSIZE; } +static void __mdb_entry_fill_flags(struct br_mdb_entry *e, unsigned char flags) +{ + e->state = flags & MDB_PG_FLAGS_PERMANENT; + e->flags = 0; + if (flags & MDB_PG_FLAGS_OFFLOAD) + e->flags |= MDB_FLAGS_OFFLOAD; +} + static int br_mdb_fill_info(struct sk_buff *skb, struct netlink_callback *cb, struct net_device *dev) { @@@ -88,41 -80,26 +88,41 @@@ for (pp = &mp->ports; (p = rcu_dereference(*pp)) != NULL; pp = &p->next) { + struct nlattr *nest_ent; + struct br_mdb_entry e; + port = p->port; - if (port) { - struct br_mdb_entry e; - memset(&e, 0, sizeof(e)); - e.ifindex = port->dev->ifindex; - e.state = p->state; - e.vid = p->addr.vid; - if (p->addr.proto == htons(ETH_P_IP)) - e.addr.u.ip4 = p->addr.u.ip4; + if (!port) + continue; + + memset(&e, 0, sizeof(e)); + e.ifindex = port->dev->ifindex; + e.vid = p->addr.vid; + __mdb_entry_fill_flags(&e, p->flags); + if (p->addr.proto == htons(ETH_P_IP)) + e.addr.u.ip4 = p->addr.u.ip4; #if IS_ENABLED(CONFIG_IPV6) - if (p->addr.proto == htons(ETH_P_IPV6)) - e.addr.u.ip6 = p->addr.u.ip6; + if (p->addr.proto == htons(ETH_P_IPV6)) + e.addr.u.ip6 = p->addr.u.ip6; #endif - e.addr.proto = p->addr.proto; - if (nla_put(skb, MDBA_MDB_ENTRY_INFO, sizeof(e), &e)) { - nla_nest_cancel(skb, nest2); - err = -EMSGSIZE; - goto out; - } + e.addr.proto = p->addr.proto; + nest_ent = nla_nest_start(skb, + MDBA_MDB_ENTRY_INFO); + if (!nest_ent) { + nla_nest_cancel(skb, nest2); + err = -EMSGSIZE; + goto out; } + if (nla_put_nohdr(skb, sizeof(e), &e) || + nla_put_u32(skb, + MDBA_MDB_EATTR_TIMER, + br_timer_value(&p->timer))) { + nla_nest_cancel(skb, nest_ent); + nla_nest_cancel(skb, nest2); + err = -EMSGSIZE; + goto out; + } + nla_nest_end(skb, nest_ent); } nla_nest_end(skb, nest2); skip: @@@ -232,7 -209,7 +232,7 @@@ static inline size_t rtnl_mdb_nlmsg_siz } static void __br_mdb_notify(struct net_device *dev, struct br_mdb_entry *entry, - int type) + int type, struct net_bridge_port_group *pg) { struct switchdev_obj_port_mdb mdb = { .obj = { @@@ -255,13 -232,10 +255,13 @@@ #endif mdb.obj.orig_dev = port_dev; - if (port_dev && type == RTM_NEWMDB) - switchdev_port_obj_add(port_dev, &mdb.obj); - else if (port_dev && type == RTM_DELMDB) + if (port_dev && type == RTM_NEWMDB) { + err = switchdev_port_obj_add(port_dev, &mdb.obj); + if (!err && pg) + pg->flags |= MDB_PG_FLAGS_OFFLOAD; + } else if (port_dev && type == RTM_DELMDB) { switchdev_port_obj_del(port_dev, &mdb.obj); + } skb = nlmsg_new(rtnl_mdb_nlmsg_size(), GFP_ATOMIC); if (!skb) @@@ -279,21 -253,21 +279,21 @@@ errout rtnl_set_sk_err(net, RTNLGRP_MDB, err); } -void br_mdb_notify(struct net_device *dev, struct net_bridge_port *port, - struct br_ip *group, int type, u8 state) +void br_mdb_notify(struct net_device *dev, struct net_bridge_port_group *pg, + int type) { struct br_mdb_entry entry; memset(&entry, 0, sizeof(entry)); - entry.ifindex = port->dev->ifindex; - entry.addr.proto = group->proto; - entry.addr.u.ip4 = group->u.ip4; + entry.ifindex = pg->port->dev->ifindex; + entry.addr.proto = pg->addr.proto; + entry.addr.u.ip4 = pg->addr.u.ip4; #if IS_ENABLED(CONFIG_IPV6) - entry.addr.u.ip6 = group->u.ip6; + entry.addr.u.ip6 = pg->addr.u.ip6; #endif - entry.state = state; - entry.vid = group->vid; - __br_mdb_notify(dev, &entry, type); + entry.vid = pg->addr.vid; + __mdb_entry_fill_flags(&entry, pg->flags); + __br_mdb_notify(dev, &entry, type, pg); } static int nlmsg_populate_rtr_fill(struct sk_buff *skb, @@@ -438,8 -412,7 +438,8 @@@ static int br_mdb_parse(struct sk_buff } static int br_mdb_add_group(struct net_bridge *br, struct net_bridge_port *port, - struct br_ip *group, unsigned char state) + struct br_ip *group, unsigned char state, + struct net_bridge_port_group **pg) { struct net_bridge_mdb_entry *mp; struct net_bridge_port_group *p; @@@ -452,8 -425,8 +452,8 @@@ mp = br_mdb_ip_get(mdb, group); if (!mp) { mp = br_multicast_new_group(br, port, group); - err = PTR_ERR(mp); - if (IS_ERR(mp)) + err = PTR_ERR_OR_ZERO(mp); + if (err) return err; } @@@ -470,7 -443,6 +470,7 @@@ if (unlikely(!p)) return -ENOMEM; rcu_assign_pointer(*pp, p); + *pg = p; if (state == MDB_TEMPORARY) mod_timer(&p->timer, now + br->multicast_membership_interval); @@@ -478,8 -450,7 +478,8 @@@ } static int __br_mdb_add(struct net *net, struct net_bridge *br, - struct br_mdb_entry *entry) + struct br_mdb_entry *entry, + struct net_bridge_port_group **pg) { struct br_ip ip; struct net_device *dev; @@@ -508,7 -479,7 +508,7 @@@ #endif spin_lock_bh(&br->multicast_lock); - ret = br_mdb_add_group(br, p, &ip, entry->state); + ret = br_mdb_add_group(br, p, &ip, entry->state, pg); spin_unlock_bh(&br->multicast_lock); return ret; } @@@ -516,7 -487,6 +516,7 @@@ static int br_mdb_add(struct sk_buff *skb, struct nlmsghdr *nlh) { struct net *net = sock_net(skb->sk); + struct net_bridge_port_group *pg; struct net_bridge_vlan_group *vg; struct net_device *dev, *pdev; struct br_mdb_entry *entry; @@@ -546,15 -516,15 +546,15 @@@ if (br_vlan_enabled(br) && vg && entry->vid == 0) { list_for_each_entry(v, &vg->vlan_list, vlist) { entry->vid = v->vid; - err = __br_mdb_add(net, br, entry); + err = __br_mdb_add(net, br, entry, &pg); if (err) break; - __br_mdb_notify(dev, entry, RTM_NEWMDB); + __br_mdb_notify(dev, entry, RTM_NEWMDB, pg); } } else { - err = __br_mdb_add(net, br, entry); + err = __br_mdb_add(net, br, entry, &pg); if (!err) - __br_mdb_notify(dev, entry, RTM_NEWMDB); + __br_mdb_notify(dev, entry, RTM_NEWMDB, pg); } return err; @@@ -598,7 -568,7 +598,7 @@@ static int __br_mdb_del(struct net_brid if (p->port->state == BR_STATE_DISABLED) goto unlock; - entry->state = p->state; + __mdb_entry_fill_flags(entry, p->flags); rcu_assign_pointer(*pp, p->next); hlist_del_init(&p->mglist); del_timer(&p->timer); @@@ -650,12 -620,12 +650,12 @@@ static int br_mdb_del(struct sk_buff *s entry->vid = v->vid; err = __br_mdb_del(br, entry); if (!err) - __br_mdb_notify(dev, entry, RTM_DELMDB); + __br_mdb_notify(dev, entry, RTM_DELMDB, NULL); } } else { err = __br_mdb_del(br, entry); if (!err) - __br_mdb_notify(dev, entry, RTM_DELMDB); + __br_mdb_notify(dev, entry, RTM_DELMDB, NULL); } return err; diff --combined net/core/dev.c index 3f4071a,0ef061b..edb7179 --- a/net/core/dev.c +++ b/net/core/dev.c @@@ -3829,14 -3829,8 +3829,14 @@@ static void net_tx_action(struct softir trace_consume_skb(skb); else trace_kfree_skb(skb, net_tx_action); - __kfree_skb(skb); + + if (skb->fclone != SKB_FCLONE_UNAVAILABLE) + __kfree_skb(skb); + else + __kfree_skb_defer(skb); } + + __kfree_skb_flush(); } if (sd->output_queue) { @@@ -4160,10 -4154,7 +4160,10 @@@ ncls ret = pt_prev->func(skb, skb->dev, pt_prev, orig_dev); } else { drop: - atomic_long_inc(&skb->dev->rx_dropped); + if (!deliver_exact) + atomic_long_inc(&skb->dev->rx_dropped); + else + atomic_long_inc(&skb->dev->rx_nohandler); kfree_skb(skb); /* Jamal, now you will not able to escape explaining * me how you were going to use this. :-) @@@ -5161,7 -5152,6 +5161,7 @@@ static void net_rx_action(struct softir } } + __kfree_skb_flush(); local_irq_disable(); list_splice_tail_init(&sd->poll_list, &list); @@@ -5389,12 -5379,12 +5389,12 @@@ void *netdev_lower_get_next(struct net_ { struct netdev_adjacent *lower; - lower = list_entry((*iter)->next, struct netdev_adjacent, list); + lower = list_entry(*iter, struct netdev_adjacent, list); if (&lower->list == &dev->adj_list.lower) return NULL; - *iter = &lower->list; + *iter = lower->list.next; return lower->dev; } @@@ -7263,31 -7253,24 +7263,31 @@@ void netdev_run_todo(void } } -/* Convert net_device_stats to rtnl_link_stats64. They have the same - * fields in the same order, with only the type differing. +/* Convert net_device_stats to rtnl_link_stats64. rtnl_link_stats64 has + * all the same fields in the same order as net_device_stats, with only + * the type differing, but rtnl_link_stats64 may have additional fields + * at the end for newer counters. */ void netdev_stats_to_stats64(struct rtnl_link_stats64 *stats64, const struct net_device_stats *netdev_stats) { #if BITS_PER_LONG == 64 - BUILD_BUG_ON(sizeof(*stats64) != sizeof(*netdev_stats)); + BUILD_BUG_ON(sizeof(*stats64) < sizeof(*netdev_stats)); memcpy(stats64, netdev_stats, sizeof(*stats64)); + /* zero out counters that only exist in rtnl_link_stats64 */ + memset((char *)stats64 + sizeof(*netdev_stats), 0, + sizeof(*stats64) - sizeof(*netdev_stats)); #else - size_t i, n = sizeof(*stats64) / sizeof(u64); + size_t i, n = sizeof(*netdev_stats) / sizeof(unsigned long); const unsigned long *src = (const unsigned long *)netdev_stats; u64 *dst = (u64 *)stats64; - BUILD_BUG_ON(sizeof(*netdev_stats) / sizeof(unsigned long) != - sizeof(*stats64) / sizeof(u64)); + BUILD_BUG_ON(n > sizeof(*stats64) / sizeof(u64)); for (i = 0; i < n; i++) dst[i] = src[i]; + /* zero out counters that only exist in rtnl_link_stats64 */ + memset((char *)stats64 + n * sizeof(u64), 0, + sizeof(*stats64) - n * sizeof(u64)); #endif } EXPORT_SYMBOL(netdev_stats_to_stats64); @@@ -7317,7 -7300,6 +7317,7 @@@ struct rtnl_link_stats64 *dev_get_stats } storage->rx_dropped += atomic_long_read(&dev->rx_dropped); storage->tx_dropped += atomic_long_read(&dev->tx_dropped); + storage->rx_nohandler += atomic_long_read(&dev->rx_nohandler); return storage; } EXPORT_SYMBOL(dev_get_stats); @@@ -7440,8 -7422,10 +7440,10 @@@ struct net_device *alloc_netdev_mqs(in dev->priv_flags = IFF_XMIT_DST_RELEASE | IFF_XMIT_DST_RELEASE_PERM; setup(dev); - if (!dev->tx_queue_len) + if (!dev->tx_queue_len) { dev->priv_flags |= IFF_NO_QUEUE; + dev->tx_queue_len = 1; + } dev->num_tx_queues = txqs; dev->real_num_tx_queues = txqs; diff --combined net/core/skbuff.c index 8bd4b79,5bf88f5..488566b --- a/net/core/skbuff.c +++ b/net/core/skbuff.c @@@ -79,6 -79,8 +79,8 @@@ struct kmem_cache *skbuff_head_cache __read_mostly; static struct kmem_cache *skbuff_fclone_cache __read_mostly; + int sysctl_max_skb_frags __read_mostly = MAX_SKB_FRAGS; + EXPORT_SYMBOL(sysctl_max_skb_frags); /** * skb_panic - private function for out-of-line support @@@ -347,16 -349,8 +349,16 @@@ struct sk_buff *build_skb(void *data, u } EXPORT_SYMBOL(build_skb); +#define NAPI_SKB_CACHE_SIZE 64 + +struct napi_alloc_cache { + struct page_frag_cache page; + size_t skb_count; + void *skb_cache[NAPI_SKB_CACHE_SIZE]; +}; + static DEFINE_PER_CPU(struct page_frag_cache, netdev_alloc_cache); -static DEFINE_PER_CPU(struct page_frag_cache, napi_alloc_cache); +static DEFINE_PER_CPU(struct napi_alloc_cache, napi_alloc_cache); static void *__netdev_alloc_frag(unsigned int fragsz, gfp_t gfp_mask) { @@@ -386,9 -380,9 +388,9 @@@ EXPORT_SYMBOL(netdev_alloc_frag) static void *__napi_alloc_frag(unsigned int fragsz, gfp_t gfp_mask) { - struct page_frag_cache *nc = this_cpu_ptr(&napi_alloc_cache); + struct napi_alloc_cache *nc = this_cpu_ptr(&napi_alloc_cache); - return __alloc_page_frag(nc, fragsz, gfp_mask); + return __alloc_page_frag(&nc->page, fragsz, gfp_mask); } void *napi_alloc_frag(unsigned int fragsz) @@@ -482,7 -476,7 +484,7 @@@ EXPORT_SYMBOL(__netdev_alloc_skb) struct sk_buff *__napi_alloc_skb(struct napi_struct *napi, unsigned int len, gfp_t gfp_mask) { - struct page_frag_cache *nc = this_cpu_ptr(&napi_alloc_cache); + struct napi_alloc_cache *nc = this_cpu_ptr(&napi_alloc_cache); struct sk_buff *skb; void *data; @@@ -502,7 -496,7 +504,7 @@@ if (sk_memalloc_socks()) gfp_mask |= __GFP_MEMALLOC; - data = __alloc_page_frag(nc, len, gfp_mask); + data = __alloc_page_frag(&nc->page, len, gfp_mask); if (unlikely(!data)) return NULL; @@@ -513,7 -507,7 +515,7 @@@ } /* use OR instead of assignment to avoid clearing of bits in mask */ - if (nc->pfmemalloc) + if (nc->page.pfmemalloc) skb->pfmemalloc = 1; skb->head_frag = 1; @@@ -755,73 -749,6 +757,73 @@@ void consume_skb(struct sk_buff *skb } EXPORT_SYMBOL(consume_skb); +void __kfree_skb_flush(void) +{ + struct napi_alloc_cache *nc = this_cpu_ptr(&napi_alloc_cache); + + /* flush skb_cache if containing objects */ + if (nc->skb_count) { + kmem_cache_free_bulk(skbuff_head_cache, nc->skb_count, + nc->skb_cache); + nc->skb_count = 0; + } +} + +static inline void _kfree_skb_defer(struct sk_buff *skb) +{ + struct napi_alloc_cache *nc = this_cpu_ptr(&napi_alloc_cache); + + /* drop skb->head and call any destructors for packet */ + skb_release_all(skb); + + /* record skb to CPU local list */ + nc->skb_cache[nc->skb_count++] = skb; + +#ifdef CONFIG_SLUB + /* SLUB writes into objects when freeing */ + prefetchw(skb); +#endif + + /* flush skb_cache if it is filled */ + if (unlikely(nc->skb_count == NAPI_SKB_CACHE_SIZE)) { + kmem_cache_free_bulk(skbuff_head_cache, NAPI_SKB_CACHE_SIZE, + nc->skb_cache); + nc->skb_count = 0; + } +} +void __kfree_skb_defer(struct sk_buff *skb) +{ + _kfree_skb_defer(skb); +} + +void napi_consume_skb(struct sk_buff *skb, int budget) +{ + if (unlikely(!skb)) + return; + + /* if budget is 0 assume netpoll w/ IRQs disabled */ + if (unlikely(!budget)) { + dev_consume_skb_irq(skb); + return; + } + + if (likely(atomic_read(&skb->users) == 1)) + smp_rmb(); + else if (likely(!atomic_dec_and_test(&skb->users))) + return; + /* if reaching here SKB is ready to free */ + trace_consume_skb(skb); + + /* if SKB is a clone, don't handle this case */ + if (unlikely(skb->fclone != SKB_FCLONE_UNAVAILABLE)) { + __kfree_skb(skb); + return; + } + + _kfree_skb_defer(skb); +} +EXPORT_SYMBOL(napi_consume_skb); + /* Make sure a field is enclosed inside headers_start/headers_end section */ #define CHECK_SKB_FIELD(field) \ BUILD_BUG_ON(offsetof(struct sk_buff, field) < \ @@@ -3079,7 -3006,8 +3081,7 @@@ struct sk_buff *skb_segment(struct sk_b if (unlikely(!proto)) return ERR_PTR(-EINVAL); - csum = !head_skb->encap_hdr_csum && - !!can_checksum_protocol(features, proto); + csum = !!can_checksum_protocol(features, proto); headroom = skb_headroom(head_skb); pos = skb_headlen(head_skb); @@@ -3172,15 -3100,13 +3174,15 @@@ if (nskb->len == len + doffset) goto perform_csum_check; - if (!sg && !nskb->remcsum_offload) { - nskb->ip_summed = CHECKSUM_NONE; - nskb->csum = skb_copy_and_csum_bits(head_skb, offset, - skb_put(nskb, len), - len, 0); + if (!sg) { + if (!nskb->remcsum_offload) + nskb->ip_summed = CHECKSUM_NONE; + SKB_GSO_CB(nskb)->csum = + skb_copy_and_csum_bits(head_skb, offset, + skb_put(nskb, len), + len, 0); SKB_GSO_CB(nskb)->csum_start = - skb_headroom(nskb) + doffset; + skb_headroom(nskb) + doffset; continue; } @@@ -3246,19 -3172,12 +3248,19 @@@ skip_fraglist nskb->truesize += nskb->data_len; perform_csum_check: - if (!csum && !nskb->remcsum_offload) { - nskb->csum = skb_checksum(nskb, doffset, - nskb->len - doffset, 0); - nskb->ip_summed = CHECKSUM_NONE; + if (!csum) { + if (skb_has_shared_frag(nskb)) { + err = __skb_linearize(nskb); + if (err) + goto err; + } + if (!nskb->remcsum_offload) + nskb->ip_summed = CHECKSUM_NONE; + SKB_GSO_CB(nskb)->csum = + skb_checksum(nskb, doffset, + nskb->len - doffset, 0); SKB_GSO_CB(nskb)->csum_start = - skb_headroom(nskb) + doffset; + skb_headroom(nskb) + doffset; } } while ((offset += len) < head_skb->len); @@@ -4496,7 -4415,9 +4498,7 @@@ int skb_vlan_push(struct sk_buff *skb, skb->mac_len += VLAN_HLEN; __skb_pull(skb, offset); - if (skb->ip_summed == CHECKSUM_COMPLETE) - skb->csum = csum_add(skb->csum, csum_partial(skb->data - + (2 * ETH_ALEN), VLAN_HLEN, 0)); + skb_postpush_rcsum(skb, skb->data + (2 * ETH_ALEN), VLAN_HLEN); } __vlan_hwaccel_put_tag(skb, vlan_proto, vlan_tci); return 0; diff --combined net/dccp/ipv4.c index 1e0c600,902d606..b5672e5 --- a/net/dccp/ipv4.c +++ b/net/dccp/ipv4.c @@@ -802,7 -802,7 +802,7 @@@ static int dccp_v4_rcv(struct sk_buff * } lookup: - sk = __inet_lookup_skb(&dccp_hashinfo, skb, + sk = __inet_lookup_skb(&dccp_hashinfo, skb, __dccp_hdr_len(dh), dh->dccph_sport, dh->dccph_dport); if (!sk) { dccp_pr_debug("failed to look up flow ID in table and " @@@ -824,26 -824,26 +824,26 @@@ if (sk->sk_state == DCCP_NEW_SYN_RECV) { struct request_sock *req = inet_reqsk(sk); - struct sock *nsk = NULL; + struct sock *nsk; sk = req->rsk_listener; - if (likely(sk->sk_state == DCCP_LISTEN)) { - nsk = dccp_check_req(sk, skb, req); - } else { + if (unlikely(sk->sk_state != DCCP_LISTEN)) { inet_csk_reqsk_queue_drop_and_put(sk, req); goto lookup; } + sock_hold(sk); + nsk = dccp_check_req(sk, skb, req); if (!nsk) { reqsk_put(req); - goto discard_it; + goto discard_and_relse; } if (nsk == sk) { - sock_hold(sk); reqsk_put(req); } else if (dccp_child_process(sk, nsk, skb)) { dccp_v4_ctl_send_reset(sk, skb); - goto discard_it; + goto discard_and_relse; } else { + sock_put(sk); return 0; } } diff --combined net/dccp/ipv6.c index 45cbe85,b8608b7..4663a01 --- a/net/dccp/ipv6.c +++ b/net/dccp/ipv6.c @@@ -668,7 -668,7 +668,7 @@@ static int dccp_v6_rcv(struct sk_buff * DCCP_SKB_CB(skb)->dccpd_ack_seq = dccp_hdr_ack_seq(skb); lookup: - sk = __inet6_lookup_skb(&dccp_hashinfo, skb, + sk = __inet6_lookup_skb(&dccp_hashinfo, skb, __dccp_hdr_len(dh), dh->dccph_sport, dh->dccph_dport, inet6_iif(skb)); if (!sk) { @@@ -691,26 -691,26 +691,26 @@@ if (sk->sk_state == DCCP_NEW_SYN_RECV) { struct request_sock *req = inet_reqsk(sk); - struct sock *nsk = NULL; + struct sock *nsk; sk = req->rsk_listener; - if (likely(sk->sk_state == DCCP_LISTEN)) { - nsk = dccp_check_req(sk, skb, req); - } else { + if (unlikely(sk->sk_state != DCCP_LISTEN)) { inet_csk_reqsk_queue_drop_and_put(sk, req); goto lookup; } + sock_hold(sk); + nsk = dccp_check_req(sk, skb, req); if (!nsk) { reqsk_put(req); - goto discard_it; + goto discard_and_relse; } if (nsk == sk) { - sock_hold(sk); reqsk_put(req); } else if (dccp_child_process(sk, nsk, skb)) { dccp_v6_ctl_send_reset(sk, skb); - goto discard_it; + goto discard_and_relse; } else { + sock_put(sk); return 0; } } @@@ -993,7 -993,7 +993,7 @@@ static struct proto dccp_v6_prot = .sendmsg = dccp_sendmsg, .recvmsg = dccp_recvmsg, .backlog_rcv = dccp_v6_do_rcv, - .hash = inet_hash, + .hash = inet6_hash, .unhash = inet_unhash, .accept = inet_csk_accept, .get_port = inet_csk_get_port, diff --combined net/ipv4/devinet.c index 3d83531,f6303b1..29b8d3a --- a/net/ipv4/devinet.c +++ b/net/ipv4/devinet.c @@@ -1847,7 -1847,7 +1847,7 @@@ static int inet_netconf_get_devconf(str if (err < 0) goto errout; - err = EINVAL; + err = -EINVAL; if (!tb[NETCONFA_IFINDEX]) goto errout; @@@ -2185,8 -2185,6 +2185,8 @@@ static struct devinet_sysctl_table "igmpv3_unsolicited_report_interval"), DEVINET_SYSCTL_RW_ENTRY(IGNORE_ROUTES_WITH_LINKDOWN, "ignore_routes_with_linkdown"), + DEVINET_SYSCTL_RW_ENTRY(DROP_GRATUITOUS_ARP, + "drop_gratuitous_arp"), DEVINET_SYSCTL_FLUSHING_ENTRY(NOXFRM, "disable_xfrm"), DEVINET_SYSCTL_FLUSHING_ENTRY(NOPOLICY, "disable_policy"), @@@ -2194,8 -2192,6 +2194,8 @@@ "promote_secondaries"), DEVINET_SYSCTL_FLUSHING_ENTRY(ROUTE_LOCALNET, "route_localnet"), + DEVINET_SYSCTL_FLUSHING_ENTRY(DROP_UNICAST_IN_L2_MULTICAST, + "drop_unicast_in_l2_multicast"), }, }; diff --combined net/ipv4/inet_connection_sock.c index 3d28c6d,6414891..d768230 --- a/net/ipv4/inet_connection_sock.c +++ b/net/ipv4/inet_connection_sock.c @@@ -24,7 -24,6 +24,7 @@@ #include <net/tcp_states.h> #include <net/xfrm.h> #include <net/tcp.h> +#include <net/sock_reuseport.h> #ifdef INET_CSK_DEBUG const char inet_csk_timer_bug_msg[] = "inet_csk BUG: unknown timer value\n"; @@@ -68,8 -67,7 +68,8 @@@ int inet_csk_bind_conflict(const struc if ((!reuse || !sk2->sk_reuse || sk2->sk_state == TCP_LISTEN) && (!reuseport || !sk2->sk_reuseport || - (sk2->sk_state != TCP_TIME_WAIT && + rcu_access_pointer(sk->sk_reuseport_cb) || + (sk2->sk_state != TCP_TIME_WAIT && !uid_eq(uid, sock_i_uid(sk2))))) { if (!sk2->sk_rcv_saddr || !sk->sk_rcv_saddr || @@@ -91,153 -89,161 +91,153 @@@ EXPORT_SYMBOL_GPL(inet_csk_bind_conflic /* Obtain a reference to a local port for the given sock, * if snum is zero it means select any available local port. + * We try to allocate an odd port (and leave even ports for connect()) */ int inet_csk_get_port(struct sock *sk, unsigned short snum) { - struct inet_hashinfo *hashinfo = sk->sk_prot->h.hashinfo; + bool reuse = sk->sk_reuse && sk->sk_state != TCP_LISTEN; + struct inet_hashinfo *hinfo = sk->sk_prot->h.hashinfo; + int ret = 1, attempts = 5, port = snum; + int smallest_size = -1, smallest_port; struct inet_bind_hashbucket *head; - struct inet_bind_bucket *tb; - int ret, attempts = 5; struct net *net = sock_net(sk); - int smallest_size = -1, smallest_rover; + int i, low, high, attempt_half; + struct inet_bind_bucket *tb; kuid_t uid = sock_i_uid(sk); - int attempt_half = (sk->sk_reuse == SK_CAN_REUSE) ? 1 : 0; + u32 remaining, offset; - local_bh_disable(); - if (!snum) { - int remaining, rover, low, high; + if (port) { +have_port: + head = &hinfo->bhash[inet_bhashfn(net, port, + hinfo->bhash_size)]; + spin_lock_bh(&head->lock); + inet_bind_bucket_for_each(tb, &head->chain) + if (net_eq(ib_net(tb), net) && tb->port == port) + goto tb_found; + goto tb_not_found; + } again: - inet_get_local_port_range(net, &low, &high); - if (attempt_half) { - int half = low + ((high - low) >> 1); - - if (attempt_half == 1) - high = half; - else - low = half; - } - remaining = (high - low) + 1; - smallest_rover = rover = prandom_u32() % remaining + low; - - smallest_size = -1; - do { - if (inet_is_local_reserved_port(net, rover)) - goto next_nolock; - head = &hashinfo->bhash[inet_bhashfn(net, rover, - hashinfo->bhash_size)]; - spin_lock(&head->lock); - inet_bind_bucket_for_each(tb, &head->chain) - if (net_eq(ib_net(tb), net) && tb->port == rover) { - if (((tb->fastreuse > 0 && - sk->sk_reuse && - sk->sk_state != TCP_LISTEN) || - (tb->fastreuseport > 0 && - sk->sk_reuseport && - uid_eq(tb->fastuid, uid))) && - (tb->num_owners < smallest_size || smallest_size == -1)) { - smallest_size = tb->num_owners; - smallest_rover = rover; - } - if (!inet_csk(sk)->icsk_af_ops->bind_conflict(sk, tb, false)) { - snum = rover; - goto tb_found; - } - goto next; + attempt_half = (sk->sk_reuse == SK_CAN_REUSE) ? 1 : 0; +other_half_scan: + inet_get_local_port_range(net, &low, &high); + high++; /* [32768, 60999] -> [32768, 61000[ */ + if (high - low < 4) + attempt_half = 0; + if (attempt_half) { + int half = low + (((high - low) >> 2) << 1); + + if (attempt_half == 1) + high = half; + else + low = half; + } + remaining = high - low; + if (likely(remaining > 1)) + remaining &= ~1U; + + offset = prandom_u32() % remaining; + /* __inet_hash_connect() favors ports having @low parity + * We do the opposite to not pollute connect() users. + */ + offset |= 1U; + smallest_size = -1; + smallest_port = low; /* avoid compiler warning */ + +other_parity_scan: + port = low + offset; + for (i = 0; i < remaining; i += 2, port += 2) { + if (unlikely(port >= high)) + port -= remaining; + if (inet_is_local_reserved_port(net, port)) + continue; + head = &hinfo->bhash[inet_bhashfn(net, port, + hinfo->bhash_size)]; + spin_lock_bh(&head->lock); + inet_bind_bucket_for_each(tb, &head->chain) + if (net_eq(ib_net(tb), net) && tb->port == port) { + if (((tb->fastreuse > 0 && reuse) || + (tb->fastreuseport > 0 && + sk->sk_reuseport && + !rcu_access_pointer(sk->sk_reuseport_cb) && + uid_eq(tb->fastuid, uid))) && + (tb->num_owners < smallest_size || smallest_size == -1)) { + smallest_size = tb->num_owners; + smallest_port = port; } - break; - next: - spin_unlock(&head->lock); - next_nolock: - if (++rover > high) - rover = low; - } while (--remaining > 0); - - /* Exhausted local port range during search? It is not - * possible for us to be holding one of the bind hash - * locks if this test triggers, because if 'remaining' - * drops to zero, we broke out of the do/while loop at - * the top level, not from the 'break;' statement. - */ - ret = 1; - if (remaining <= 0) { - if (smallest_size != -1) { - snum = smallest_rover; - goto have_snum; - } - if (attempt_half == 1) { - /* OK we now try the upper half of the range */ - attempt_half = 2; - goto again; + if (!inet_csk(sk)->icsk_af_ops->bind_conflict(sk, tb, false)) + goto tb_found; + goto next_port; } - goto fail; - } - /* OK, here is the one we will use. HEAD is - * non-NULL and we hold it's mutex. - */ - snum = rover; - } else { -have_snum: - head = &hashinfo->bhash[inet_bhashfn(net, snum, - hashinfo->bhash_size)]; - spin_lock(&head->lock); - inet_bind_bucket_for_each(tb, &head->chain) - if (net_eq(ib_net(tb), net) && tb->port == snum) - goto tb_found; + goto tb_not_found; +next_port: + spin_unlock_bh(&head->lock); + cond_resched(); + } + + if (smallest_size != -1) { + port = smallest_port; + goto have_port; } - tb = NULL; - goto tb_not_found; + offset--; + if (!(offset & 1)) + goto other_parity_scan; + + if (attempt_half == 1) { + /* OK we now try the upper half of the range */ + attempt_half = 2; + goto other_half_scan; + } + return ret; + +tb_not_found: + tb = inet_bind_bucket_create(hinfo->bind_bucket_cachep, + net, head, port); + if (!tb) + goto fail_unlock; tb_found: if (!hlist_empty(&tb->owners)) { if (sk->sk_reuse == SK_FORCE_REUSE) goto success; - if (((tb->fastreuse > 0 && - sk->sk_reuse && sk->sk_state != TCP_LISTEN) || + if (((tb->fastreuse > 0 && reuse) || (tb->fastreuseport > 0 && sk->sk_reuseport && uid_eq(tb->fastuid, uid))) && - smallest_size == -1) { + smallest_size == -1) goto success; - } else { - ret = 1; - if (inet_csk(sk)->icsk_af_ops->bind_conflict(sk, tb, true)) { - if (((sk->sk_reuse && sk->sk_state != TCP_LISTEN) || - (tb->fastreuseport > 0 && - sk->sk_reuseport && uid_eq(tb->fastuid, uid))) && - smallest_size != -1 && --attempts >= 0) { - spin_unlock(&head->lock); - goto again; - } - - goto fail_unlock; + if (inet_csk(sk)->icsk_af_ops->bind_conflict(sk, tb, true)) { + if ((reuse || + (tb->fastreuseport > 0 && + sk->sk_reuseport && + !rcu_access_pointer(sk->sk_reuseport_cb) && + uid_eq(tb->fastuid, uid))) && + smallest_size != -1 && --attempts >= 0) { + spin_unlock_bh(&head->lock); + goto again; } + goto fail_unlock; } - } -tb_not_found: - ret = 1; - if (!tb && (tb = inet_bind_bucket_create(hashinfo->bind_bucket_cachep, - net, head, snum)) == NULL) - goto fail_unlock; - if (hlist_empty(&tb->owners)) { - if (sk->sk_reuse && sk->sk_state != TCP_LISTEN) - tb->fastreuse = 1; - else + if (!reuse) tb->fastreuse = 0; + if (!sk->sk_reuseport || !uid_eq(tb->fastuid, uid)) + tb->fastreuseport = 0; + } else { + tb->fastreuse = reuse; if (sk->sk_reuseport) { tb->fastreuseport = 1; tb->fastuid = uid; - } else - tb->fastreuseport = 0; - } else { - if (tb->fastreuse && - (!sk->sk_reuse || sk->sk_state == TCP_LISTEN)) - tb->fastreuse = 0; - if (tb->fastreuseport && - (!sk->sk_reuseport || !uid_eq(tb->fastuid, uid))) + } else { tb->fastreuseport = 0; + } } success: if (!inet_csk(sk)->icsk_bind_hash) - inet_bind_hash(sk, tb, snum); + inet_bind_hash(sk, tb, port); WARN_ON(inet_csk(sk)->icsk_bind_hash != tb); ret = 0; fail_unlock: - spin_unlock(&head->lock); -fail: - local_bh_enable(); + spin_unlock_bh(&head->lock); return ret; } EXPORT_SYMBOL_GPL(inet_csk_get_port); @@@ -476,6 -482,10 +476,6 @@@ EXPORT_SYMBOL_GPL(inet_csk_route_child_ #define AF_INET_FAMILY(fam) true #endif -/* Only thing we need from tcp.h */ -extern int sysctl_tcp_synack_retries; - - /* Decide when to expire the request and when to resend SYN-ACK */ static inline void syn_ack_recalc(struct request_sock *req, const int thresh, const int max_retries, @@@ -547,7 -557,6 +547,7 @@@ static void reqsk_timer_handler(unsigne { struct request_sock *req = (struct request_sock *)data; struct sock *sk_listener = req->rsk_listener; + struct net *net = sock_net(sk_listener); struct inet_connection_sock *icsk = inet_csk(sk_listener); struct request_sock_queue *queue = &icsk->icsk_accept_queue; int qlen, expire = 0, resend = 0; @@@ -557,7 -566,7 +557,7 @@@ if (sk_state_load(sk_listener) != TCP_LISTEN) goto drop; - max_retries = icsk->icsk_syn_retries ? : sysctl_tcp_synack_retries; + max_retries = icsk->icsk_syn_retries ? : net->ipv4.sysctl_tcp_synack_retries; thresh = max_retries; /* Normally all the openreqs are young and become mature * (i.e. converted to established socket) for first timeout. @@@ -728,7 -737,6 +728,7 @@@ int inet_csk_listen_start(struct sock * { struct inet_connection_sock *icsk = inet_csk(sk); struct inet_sock *inet = inet_sk(sk); + int err = -EADDRINUSE; reqsk_queue_alloc(&icsk->icsk_accept_queue); @@@ -746,14 -754,13 +746,14 @@@ inet->inet_sport = htons(inet->inet_num); sk_dst_reset(sk); - sk->sk_prot->hash(sk); + err = sk->sk_prot->hash(sk); - return 0; + if (likely(!err)) + return 0; } sk->sk_state = TCP_CLOSE; - return -EADDRINUSE; + return err; } EXPORT_SYMBOL_GPL(inet_csk_listen_start); @@@ -782,14 -789,16 +782,16 @@@ static void inet_child_forget(struct so reqsk_put(req); } - void inet_csk_reqsk_queue_add(struct sock *sk, struct request_sock *req, - struct sock *child) + struct sock *inet_csk_reqsk_queue_add(struct sock *sk, + struct request_sock *req, + struct sock *child) { struct request_sock_queue *queue = &inet_csk(sk)->icsk_accept_queue; spin_lock(&queue->rskq_lock); if (unlikely(sk->sk_state != TCP_LISTEN)) { inet_child_forget(sk, req, child); + child = NULL; } else { req->sk = child; req->dl_next = NULL; @@@ -801,6 -810,7 +803,7 @@@ sk_acceptq_added(sk); } spin_unlock(&queue->rskq_lock); + return child; } EXPORT_SYMBOL(inet_csk_reqsk_queue_add); @@@ -810,11 -820,8 +813,8 @@@ struct sock *inet_csk_complete_hashdanc if (own_req) { inet_csk_reqsk_queue_drop(sk, req); reqsk_queue_removed(&inet_csk(sk)->icsk_accept_queue, req); - inet_csk_reqsk_queue_add(sk, req, child); - /* Warning: caller must not call reqsk_put(req); - * child stole last reference on it. - */ - return child; + if (inet_csk_reqsk_queue_add(sk, req, child)) + return child; } /* Too bad, another child took ownership of the request, undo. */ bh_unlock_sock(child); diff --combined net/ipv4/ip_gre.c index 12071e2,41ba68d..202437d --- a/net/ipv4/ip_gre.c +++ b/net/ipv4/ip_gre.c @@@ -238,7 -238,7 +238,7 @@@ static int parse_gre_header(struct sk_b return -EINVAL; } } - return iptunnel_pull_header(skb, hdr_len, tpi->proto); + return iptunnel_pull_header(skb, hdr_len, tpi->proto, false); } static void ipgre_err(struct sk_buff *skb, u32 info, @@@ -440,17 -440,6 +440,17 @@@ drop return 0; } +static __sum16 gre_checksum(struct sk_buff *skb) +{ + __wsum csum; + + if (skb->ip_summed == CHECKSUM_PARTIAL) + csum = lco_csum(skb); + else + csum = skb_checksum(skb, 0, skb->len, 0); + return csum_fold(csum); +} + static void build_header(struct sk_buff *skb, int hdr_len, __be16 flags, __be16 proto, __be32 key, __be32 seq) { @@@ -478,7 -467,8 +478,7 @@@ !(skb_shinfo(skb)->gso_type & (SKB_GSO_GRE | SKB_GSO_GRE_CSUM))) { *ptr = 0; - *(__sum16 *)ptr = csum_fold(skb_checksum(skb, 0, - skb->len, 0)); + *(__sum16 *)ptr = gre_checksum(skb); } } } @@@ -503,7 -493,8 +503,7 @@@ static void __gre_xmit(struct sk_buff * static struct sk_buff *gre_handle_offloads(struct sk_buff *skb, bool csum) { - return iptunnel_handle_offloads(skb, csum, - csum ? SKB_GSO_GRE_CSUM : SKB_GSO_GRE); + return iptunnel_handle_offloads(skb, csum ? SKB_GSO_GRE_CSUM : SKB_GSO_GRE); } static struct rtable *gre_get_rt(struct sk_buff *skb, @@@ -540,16 -531,9 +540,16 @@@ static void gre_fb_xmit(struct sk_buff goto err_free_skb; key = &tun_info->key; - rt = gre_get_rt(skb, dev, &fl, key); - if (IS_ERR(rt)) - goto err_free_skb; + rt = !skb->mark ? dst_cache_get_ip4(&tun_info->dst_cache, &fl.saddr) : + NULL; + if (!rt) { + rt = gre_get_rt(skb, dev, &fl, key); + if (IS_ERR(rt)) + goto err_free_skb; + if (!skb->mark) + dst_cache_set_ip4(&tun_info->dst_cache, &rt->dst, + fl.saddr); + } tunnel_hlen = ip_gre_calc_hlen(key->tun_flags); @@@ -1070,8 -1054,9 +1070,9 @@@ static const struct net_device_ops gre_ static void ipgre_tap_setup(struct net_device *dev) { ether_setup(dev); - dev->netdev_ops = &gre_tap_netdev_ops; - dev->priv_flags |= IFF_LIVE_ADDR_CHANGE; + dev->netdev_ops = &gre_tap_netdev_ops; + dev->priv_flags &= ~IFF_TX_SKB_SHARING; + dev->priv_flags |= IFF_LIVE_ADDR_CHANGE; ip_tunnel_setup(dev, gre_tap_net_id); } @@@ -1256,6 -1241,14 +1257,14 @@@ struct net_device *gretap_fb_dev_create err = ipgre_newlink(net, dev, tb, NULL); if (err < 0) goto out; + + /* openvswitch users expect packet sizes to be unrestricted, + * so set the largest MTU we can. + */ + err = __ip_tunnel_change_mtu(dev, IP_MAX_MTU, false); + if (err) + goto out; + return dev; out: free_netdev(dev); diff --combined net/ipv4/ip_sockglue.c index 3f1befc,a501242..035ad64 --- a/net/ipv4/ip_sockglue.c +++ b/net/ipv4/ip_sockglue.c @@@ -249,6 -249,8 +249,8 @@@ int ip_cmsg_send(struct net *net, struc switch (cmsg->cmsg_type) { case IP_RETOPTS: err = cmsg->cmsg_len - CMSG_ALIGN(sizeof(struct cmsghdr)); + + /* Our caller is responsible for freeing ipc->opt */ err = ip_options_get(net, &ipc->opt, CMSG_DATA(cmsg), err < 40 ? err : 40); if (err) @@@ -571,7 -573,6 +573,7 @@@ static int do_ip_setsockopt(struct soc int optname, char __user *optval, unsigned int optlen) { struct inet_sock *inet = inet_sk(sk); + struct net *net = sock_net(sk); int val = 0, err; bool needs_rtnl = setsockopt_needs_rtnl(optname); @@@ -911,7 -912,7 +913,7 @@@ } /* numsrc >= (1G-4) overflow in 32 bits */ if (msf->imsf_numsrc >= 0x3ffffffcU || - msf->imsf_numsrc > sysctl_igmp_max_msf) { + msf->imsf_numsrc > net->ipv4.sysctl_igmp_max_msf) { kfree(msf); err = -ENOBUFS; break; @@@ -1066,7 -1067,7 +1068,7 @@@ /* numsrc >= (4G-140)/128 overflow in 32 bits */ if (gsf->gf_numsrc >= 0x1ffffff || - gsf->gf_numsrc > sysctl_igmp_max_msf) { + gsf->gf_numsrc > net->ipv4.sysctl_igmp_max_msf) { err = -ENOBUFS; goto mc_msf_out; } @@@ -1341,13 -1342,10 +1343,13 @@@ static int do_ip_getsockopt(struct soc val = inet->tos; break; case IP_TTL: + { + struct net *net = sock_net(sk); val = (inet->uc_ttl == -1 ? - sysctl_ip_default_ttl : + net->ipv4.sysctl_ip_default_ttl : inet->uc_ttl); break; + } case IP_HDRINCL: val = inet->hdrincl; break; diff --combined net/ipv4/ip_tunnel.c index 4569da7,89e8861..dff8a05 --- a/net/ipv4/ip_tunnel.c +++ b/net/ipv4/ip_tunnel.c @@@ -68,6 -68,61 +68,6 @@@ static unsigned int ip_tunnel_hash(__be IP_TNL_HASH_BITS); } -static void __tunnel_dst_set(struct ip_tunnel_dst *idst, - struct dst_entry *dst, __be32 saddr) -{ - struct dst_entry *old_dst; - - dst_clone(dst); - old_dst = xchg((__force struct dst_entry **)&idst->dst, dst); - dst_release(old_dst); - idst->saddr = saddr; -} - -static noinline void tunnel_dst_set(struct ip_tunnel *t, - struct dst_entry *dst, __be32 saddr) -{ - __tunnel_dst_set(raw_cpu_ptr(t->dst_cache), dst, saddr); -} - -static void tunnel_dst_reset(struct ip_tunnel *t) -{ - tunnel_dst_set(t, NULL, 0); -} - -void ip_tunnel_dst_reset_all(struct ip_tunnel *t) -{ - int i; - - for_each_possible_cpu(i) - __tunnel_dst_set(per_cpu_ptr(t->dst_cache, i), NULL, 0); -} -EXPORT_SYMBOL(ip_tunnel_dst_reset_all); - -static struct rtable *tunnel_rtable_get(struct ip_tunnel *t, - u32 cookie, __be32 *saddr) -{ - struct ip_tunnel_dst *idst; - struct dst_entry *dst; - - rcu_read_lock(); - idst = raw_cpu_ptr(t->dst_cache); - dst = rcu_dereference(idst->dst); - if (dst && !atomic_inc_not_zero(&dst->__refcnt)) - dst = NULL; - if (dst) { - if (!dst->obsolete || dst->ops->check(dst, cookie)) { - *saddr = idst->saddr; - } else { - tunnel_dst_reset(t); - dst_release(dst); - dst = NULL; - } - } - rcu_read_unlock(); - return (struct rtable *)dst; -} - static bool ip_tunnel_key_match(const struct ip_tunnel_parm *p, __be16 flags, __be32 key) { @@@ -326,8 -381,7 +326,8 @@@ static int ip_tunnel_bind_dev(struct ne if (!IS_ERR(rt)) { tdev = rt->dst.dev; - tunnel_dst_set(tunnel, &rt->dst, fl4.saddr); + dst_cache_set_ip4(&tunnel->dst_cache, &rt->dst, + fl4.saddr); ip_rt_put(rt); } if (dev->type != ARPHRD_ETHER) @@@ -675,8 -729,7 +675,8 @@@ void ip_tunnel_xmit(struct sk_buff *skb if (ip_tunnel_encap(skb, tunnel, &protocol, &fl4) < 0) goto tx_error; - rt = connected ? tunnel_rtable_get(tunnel, 0, &fl4.saddr) : NULL; + rt = connected ? dst_cache_get_ip4(&tunnel->dst_cache, &fl4.saddr) : + NULL; if (!rt) { rt = ip_route_output_key(tunnel->net, &fl4); @@@ -686,8 -739,7 +686,8 @@@ goto tx_error; } if (connected) - tunnel_dst_set(tunnel, &rt->dst, fl4.saddr); + dst_cache_set_ip4(&tunnel->dst_cache, &rt->dst, + fl4.saddr); } if (rt->dst.dev == dev) { @@@ -784,7 -836,7 +784,7 @@@ static void ip_tunnel_update(struct ip_ if (set_mtu) dev->mtu = mtu; } - ip_tunnel_dst_reset_all(t); + dst_cache_reset(&t->dst_cache); netdev_state_change(dev); } @@@ -891,17 -943,31 +891,31 @@@ done } EXPORT_SYMBOL_GPL(ip_tunnel_ioctl); - int ip_tunnel_change_mtu(struct net_device *dev, int new_mtu) + int __ip_tunnel_change_mtu(struct net_device *dev, int new_mtu, bool strict) { struct ip_tunnel *tunnel = netdev_priv(dev); int t_hlen = tunnel->hlen + sizeof(struct iphdr); + int max_mtu = 0xFFF8 - dev->hard_header_len - t_hlen; - if (new_mtu < 68 || - new_mtu > 0xFFF8 - dev->hard_header_len - t_hlen) + if (new_mtu < 68) return -EINVAL; + + if (new_mtu > max_mtu) { + if (strict) + return -EINVAL; + + new_mtu = max_mtu; + } + dev->mtu = new_mtu; return 0; } + EXPORT_SYMBOL_GPL(__ip_tunnel_change_mtu); + + int ip_tunnel_change_mtu(struct net_device *dev, int new_mtu) + { + return __ip_tunnel_change_mtu(dev, new_mtu, true); + } EXPORT_SYMBOL_GPL(ip_tunnel_change_mtu); static void ip_tunnel_dev_free(struct net_device *dev) @@@ -909,7 -975,7 +923,7 @@@ struct ip_tunnel *tunnel = netdev_priv(dev); gro_cells_destroy(&tunnel->gro_cells); - free_percpu(tunnel->dst_cache); + dst_cache_destroy(&tunnel->dst_cache); free_percpu(dev->tstats); free_netdev(dev); } @@@ -1103,15 -1169,15 +1117,15 @@@ int ip_tunnel_init(struct net_device *d if (!dev->tstats) return -ENOMEM; - tunnel->dst_cache = alloc_percpu(struct ip_tunnel_dst); - if (!tunnel->dst_cache) { + err = dst_cache_init(&tunnel->dst_cache, GFP_KERNEL); + if (err) { free_percpu(dev->tstats); - return -ENOMEM; + return err; } err = gro_cells_init(&tunnel->gro_cells, dev); if (err) { - free_percpu(tunnel->dst_cache); + dst_cache_destroy(&tunnel->dst_cache); free_percpu(dev->tstats); return err; } @@@ -1141,7 -1207,7 +1155,7 @@@ void ip_tunnel_uninit(struct net_devic if (itn->fb_tunnel_dev != dev) ip_tunnel_del(itn, netdev_priv(dev)); - ip_tunnel_dst_reset_all(tunnel); + dst_cache_reset(&tunnel->dst_cache); } EXPORT_SYMBOL_GPL(ip_tunnel_uninit); diff --combined net/ipv4/ping.c index f6f93fc,d3a2716..76dce90 --- a/net/ipv4/ping.c +++ b/net/ipv4/ping.c @@@ -145,12 -145,10 +145,12 @@@ fail } EXPORT_SYMBOL_GPL(ping_get_port); -void ping_hash(struct sock *sk) +int ping_hash(struct sock *sk) { pr_debug("ping_hash(sk->port=%u)\n", inet_sk(sk)->inet_num); BUG(); /* "Please do not press this button again." */ + + return 0; } void ping_unhash(struct sock *sk) @@@ -748,8 -746,10 +748,10 @@@ static int ping_v4_sendmsg(struct sock if (msg->msg_controllen) { err = ip_cmsg_send(sock_net(sk), msg, &ipc, false); - if (err) + if (unlikely(err)) { + kfree(ipc.opt); return err; + } if (ipc.opt) free = 1; } diff --combined net/ipv4/raw.c index d635251,7113bae..8d22de7 --- a/net/ipv4/raw.c +++ b/net/ipv4/raw.c @@@ -93,7 -93,7 +93,7 @@@ static struct raw_hashinfo raw_v4_hashi .lock = __RW_LOCK_UNLOCKED(raw_v4_hashinfo.lock), }; -void raw_hash_sk(struct sock *sk) +int raw_hash_sk(struct sock *sk) { struct raw_hashinfo *h = sk->sk_prot->h.raw_hash; struct hlist_head *head; @@@ -104,8 -104,6 +104,8 @@@ sk_add_node(sk, head); sock_prot_inuse_add(sock_net(sk), sk->sk_prot, 1); write_unlock_bh(&h->lock); + + return 0; } EXPORT_SYMBOL_GPL(raw_hash_sk); @@@ -549,8 -547,10 +549,10 @@@ static int raw_sendmsg(struct sock *sk if (msg->msg_controllen) { err = ip_cmsg_send(net, msg, &ipc, false); - if (err) + if (unlikely(err)) { + kfree(ipc.opt); goto out; + } if (ipc.opt) free = 1; } diff --combined net/ipv4/tcp.c index f93150d,483ffdf..f9faadb --- a/net/ipv4/tcp.c +++ b/net/ipv4/tcp.c @@@ -282,6 -282,8 +282,6 @@@ #include <asm/unaligned.h> #include <net/busy_poll.h> -int sysctl_tcp_fin_timeout __read_mostly = TCP_FIN_TIMEOUT; - int sysctl_tcp_min_tso_segs __read_mostly = 2; int sysctl_tcp_autocorking __read_mostly = 1; @@@ -404,7 -406,7 +404,7 @@@ void tcp_init_sock(struct sock *sk tp->mss_cache = TCP_MSS_DEFAULT; u64_stats_init(&tp->syncp); - tp->reordering = sysctl_tcp_reordering; + tp->reordering = sock_net(sk)->ipv4.sysctl_tcp_reordering; tcp_enable_early_retrans(tp); tcp_assign_congestion_control(sk); @@@ -938,7 -940,7 +938,7 @@@ new_segment i = skb_shinfo(skb)->nr_frags; can_coalesce = skb_can_coalesce(skb, i, page, offset); - if (!can_coalesce && i >= MAX_SKB_FRAGS) { + if (!can_coalesce && i >= sysctl_max_skb_frags) { tcp_mark_push(tp, skb); goto new_segment; } @@@ -1211,7 -1213,7 +1211,7 @@@ new_segment if (!skb_can_coalesce(skb, i, pfrag->page, pfrag->offset)) { - if (i == MAX_SKB_FRAGS || !sg) { + if (i == sysctl_max_skb_frags || !sg) { tcp_mark_push(tp, skb); goto new_segment; } @@@ -1464,10 -1466,8 +1464,10 @@@ static struct sk_buff *tcp_recv_skb(str while ((skb = skb_peek(&sk->sk_receive_queue)) != NULL) { offset = seq - TCP_SKB_CB(skb)->seq; - if (TCP_SKB_CB(skb)->tcp_flags & TCPHDR_SYN) + if (unlikely(TCP_SKB_CB(skb)->tcp_flags & TCPHDR_SYN)) { + pr_err_once("%s: found a SYN, please report !\n", __func__); offset--; + } if (offset < skb->len || (TCP_SKB_CB(skb)->tcp_flags & TCPHDR_FIN)) { *off = offset; return skb; @@@ -1657,10 -1657,8 +1657,10 @@@ int tcp_recvmsg(struct sock *sk, struc break; offset = *seq - TCP_SKB_CB(skb)->seq; - if (TCP_SKB_CB(skb)->tcp_flags & TCPHDR_SYN) + if (unlikely(TCP_SKB_CB(skb)->tcp_flags & TCPHDR_SYN)) { + pr_err_once("%s: found a SYN, please report !\n", __func__); offset--; + } if (offset < skb->len) goto found_ok_skb; if (TCP_SKB_CB(skb)->tcp_flags & TCPHDR_FIN) @@@ -2328,7 -2326,6 +2328,7 @@@ static int do_tcp_setsockopt(struct soc { struct tcp_sock *tp = tcp_sk(sk); struct inet_connection_sock *icsk = inet_csk(sk); + struct net *net = sock_net(sk); int val; int err = 0; @@@ -2525,7 -2522,7 +2525,7 @@@ case TCP_LINGER2: if (val < 0) tp->linger2 = -1; - else if (val > sysctl_tcp_fin_timeout / HZ) + else if (val > net->ipv4.sysctl_tcp_fin_timeout / HZ) tp->linger2 = 0; else tp->linger2 = val * HZ; @@@ -2642,7 -2639,6 +2642,7 @@@ void tcp_get_info(struct sock *sk, stru const struct inet_connection_sock *icsk = inet_csk(sk); u32 now = tcp_time_stamp; unsigned int start; + int notsent_bytes; u64 rate64; u32 rate; @@@ -2723,11 -2719,6 +2723,11 @@@ } while (u64_stats_fetch_retry_irq(&tp->syncp, start)); info->tcpi_segs_out = tp->segs_out; info->tcpi_segs_in = tp->segs_in; + + notsent_bytes = READ_ONCE(tp->write_seq) - READ_ONCE(tp->snd_nxt); + info->tcpi_notsent_bytes = max(0, notsent_bytes); + + info->tcpi_min_rtt = tcp_min_rtt(tp); } EXPORT_SYMBOL_GPL(tcp_get_info); @@@ -2736,7 -2727,6 +2736,7 @@@ static int do_tcp_getsockopt(struct soc { struct inet_connection_sock *icsk = inet_csk(sk); struct tcp_sock *tp = tcp_sk(sk); + struct net *net = sock_net(sk); int val, len; if (get_user(len, optlen)) @@@ -2771,12 -2761,12 +2771,12 @@@ val = keepalive_probes(tp); break; case TCP_SYNCNT: - val = icsk->icsk_syn_retries ? : sysctl_tcp_syn_retries; + val = icsk->icsk_syn_retries ? : net->ipv4.sysctl_tcp_syn_retries; break; case TCP_LINGER2: val = tp->linger2; if (val >= 0) - val = (val ? : sysctl_tcp_fin_timeout) / HZ; + val = (val ? : net->ipv4.sysctl_tcp_fin_timeout) / HZ; break; case TCP_DEFER_ACCEPT: val = retrans_to_secs(icsk->icsk_accept_queue.rskq_defer_accept, @@@ -2960,7 -2950,7 +2960,7 @@@ static void __tcp_alloc_md5sig_pool(voi struct crypto_hash *hash; hash = crypto_alloc_hash("md5", 0, CRYPTO_ALG_ASYNC); - if (IS_ERR_OR_NULL(hash)) + if (IS_ERR(hash)) return; per_cpu(tcp_md5sig_pool, cpu).md5_desc.tfm = hash; } diff --combined net/ipv4/tcp_input.c index 5ee6fe0,3b2c8e9..e6e65f7 --- a/net/ipv4/tcp_input.c +++ b/net/ipv4/tcp_input.c @@@ -80,7 -80,9 +80,7 @@@ int sysctl_tcp_timestamps __read_mostl int sysctl_tcp_window_scaling __read_mostly = 1; int sysctl_tcp_sack __read_mostly = 1; int sysctl_tcp_fack __read_mostly = 1; -int sysctl_tcp_reordering __read_mostly = TCP_FASTRETRANS_THRESH; int sysctl_tcp_max_reordering __read_mostly = 300; -EXPORT_SYMBOL(sysctl_tcp_reordering); int sysctl_tcp_dsack __read_mostly = 1; int sysctl_tcp_app_win __read_mostly = 31; int sysctl_tcp_adv_win_scale __read_mostly = 1; @@@ -124,10 -126,6 +124,10 @@@ int sysctl_tcp_invalid_ratelimit __read #define TCP_REMNANT (TCP_FLAG_FIN|TCP_FLAG_URG|TCP_FLAG_SYN|TCP_FLAG_PSH) #define TCP_HP_BITS (~(TCP_RESERVED_BITS|TCP_FLAG_PSH)) +#define REXMIT_NONE 0 /* no loss recovery to do */ +#define REXMIT_LOST 1 /* retransmit packets marked lost */ +#define REXMIT_NEW 2 /* FRTO-style transmit of unsent/new packets */ + /* Adapt the MSS value used to make delayed ack decision to the * real world. */ @@@ -1212,7 -1210,6 +1212,7 @@@ static u8 tcp_sacktag_one(struct sock * sacked |= TCPCB_SACKED_ACKED; state->flag |= FLAG_DATA_SACKED; tp->sacked_out += pcount; + tp->delivered += pcount; /* Out-of-order packets delivered */ fack_count += pcount; @@@ -1824,12 -1821,8 +1824,12 @@@ static void tcp_check_reno_reordering(s static void tcp_add_reno_sack(struct sock *sk) { struct tcp_sock *tp = tcp_sk(sk); + u32 prior_sacked = tp->sacked_out; + tp->sacked_out++; tcp_check_reno_reordering(sk, 0); + if (tp->sacked_out > prior_sacked) + tp->delivered++; /* Some out-of-order packet is delivered */ tcp_verify_left_out(tp); } @@@ -1841,7 -1834,6 +1841,7 @@@ static void tcp_remove_reno_sacks(struc if (acked > 0) { /* One ACK acked hole. The rest eat duplicate ACKs. */ + tp->delivered += max_t(int, acked - tp->sacked_out, 1); if (acked - 1 >= tp->sacked_out) tp->sacked_out = 0; else @@@ -1881,7 -1873,6 +1881,7 @@@ void tcp_enter_loss(struct sock *sk { const struct inet_connection_sock *icsk = inet_csk(sk); struct tcp_sock *tp = tcp_sk(sk); + struct net *net = sock_net(sk); struct sk_buff *skb; bool new_recovery = icsk->icsk_ca_state < TCP_CA_Recovery; bool is_reneg; /* is receiver reneging on SACKs? */ @@@ -1932,9 -1923,9 +1932,9 @@@ * suggests that the degree of reordering is over-estimated. */ if (icsk->icsk_ca_state <= TCP_CA_Disorder && - tp->sacked_out >= sysctl_tcp_reordering) + tp->sacked_out >= net->ipv4.sysctl_tcp_reordering) tp->reordering = min_t(unsigned int, tp->reordering, - sysctl_tcp_reordering); + net->ipv4.sysctl_tcp_reordering); tcp_set_ca_state(sk, TCP_CA_Loss); tp->high_seq = tp->snd_nxt; tcp_ecn_queue_cwr(tp); @@@ -2118,7 -2109,6 +2118,7 @@@ static bool tcp_time_to_recover(struct { struct tcp_sock *tp = tcp_sk(sk); __u32 packets_out; + int tcp_reordering = sock_net(sk)->ipv4.sysctl_tcp_reordering; /* Trick#1: The loss is proven. */ if (tp->lost_out) @@@ -2133,7 -2123,7 +2133,7 @@@ */ packets_out = tp->packets_out; if (packets_out <= tp->reordering && - tp->sacked_out >= max_t(__u32, packets_out/2, sysctl_tcp_reordering) && + tp->sacked_out >= max_t(__u32, packets_out/2, tcp_reordering) && !tcp_may_send_now(sk)) { /* We have nothing to send. This connection is limited * either by receiver window or by application. @@@ -2477,12 -2467,14 +2477,12 @@@ static void tcp_init_cwnd_reduction(str tcp_ecn_queue_cwr(tp); } -static void tcp_cwnd_reduction(struct sock *sk, const int prior_unsacked, - int fast_rexmit, int flag) +static void tcp_cwnd_reduction(struct sock *sk, int newly_acked_sacked, + int flag) { struct tcp_sock *tp = tcp_sk(sk); int sndcnt = 0; int delta = tp->snd_ssthresh - tcp_packets_in_flight(tp); - int newly_acked_sacked = prior_unsacked - - (tp->packets_out - tp->sacked_out); if (newly_acked_sacked <= 0 || WARN_ON_ONCE(!tp->prior_cwnd)) return; @@@ -2500,8 -2492,7 +2500,8 @@@ } else { sndcnt = min(delta, newly_acked_sacked); } - sndcnt = max(sndcnt, (fast_rexmit ? 1 : 0)); + /* Force a fast retransmit upon entering fast recovery */ + sndcnt = max(sndcnt, (tp->prr_out ? 0 : 1)); tp->snd_cwnd = tcp_packets_in_flight(tp) + sndcnt; } @@@ -2546,7 -2537,7 +2546,7 @@@ static void tcp_try_keep_open(struct so } } -static void tcp_try_to_open(struct sock *sk, int flag, const int prior_unsacked) +static void tcp_try_to_open(struct sock *sk, int flag) { struct tcp_sock *tp = tcp_sk(sk); @@@ -2560,6 -2551,8 +2560,6 @@@ if (inet_csk(sk)->icsk_ca_state != TCP_CA_CWR) { tcp_try_keep_open(sk); - } else { - tcp_cwnd_reduction(sk, prior_unsacked, 0, flag); } } @@@ -2669,8 -2662,7 +2669,8 @@@ static void tcp_enter_recovery(struct s /* Process an ACK in CA_Loss state. Move to CA_Open if lost data are * recovered or spurious. Otherwise retransmits more on partial ACKs. */ -static void tcp_process_loss(struct sock *sk, int flag, bool is_dupack) +static void tcp_process_loss(struct sock *sk, int flag, bool is_dupack, + int *rexmit) { struct tcp_sock *tp = tcp_sk(sk); bool recovered = !before(tp->snd_una, tp->high_seq); @@@ -2692,15 -2684,10 +2692,15 @@@ tp->frto = 0; /* Step 3.a. loss was real */ } else if (flag & FLAG_SND_UNA_ADVANCED && !recovered) { tp->high_seq = tp->snd_nxt; - __tcp_push_pending_frames(sk, tcp_current_mss(sk), - TCP_NAGLE_OFF); - if (after(tp->snd_nxt, tp->high_seq)) - return; /* Step 2.b */ + /* Step 2.b. Try send new data (but deferred until cwnd + * is updated in tcp_ack()). Otherwise fall back to + * the conventional recovery. + */ + if (tcp_send_head(sk) && + after(tcp_wnd_end(tp), tp->snd_nxt)) { + *rexmit = REXMIT_NEW; + return; + } tp->frto = 0; } } @@@ -2719,11 -2706,12 +2719,11 @@@ else if (flag & FLAG_SND_UNA_ADVANCED) tcp_reset_reno_sack(tp); } - tcp_xmit_retransmit_queue(sk); + *rexmit = REXMIT_LOST; } /* Undo during fast recovery after partial ACK. */ -static bool tcp_try_undo_partial(struct sock *sk, const int acked, - const int prior_unsacked, int flag) +static bool tcp_try_undo_partial(struct sock *sk, const int acked) { struct tcp_sock *tp = tcp_sk(sk); @@@ -2738,8 -2726,10 +2738,8 @@@ * can undo. Otherwise we clock out new packets but do not * mark more packets lost or retransmit more. */ - if (tp->retrans_out) { - tcp_cwnd_reduction(sk, prior_unsacked, 0, flag); + if (tp->retrans_out) return true; - } if (!tcp_any_retrans_done(sk)) tp->retrans_stamp = 0; @@@ -2758,21 -2748,21 +2758,21 @@@ * taking into account both packets sitting in receiver's buffer and * packets lost by network. * - * Besides that it does CWND reduction, when packet loss is detected - * and changes state of machine. + * Besides that it updates the congestion state when packet loss or ECN + * is detected. But it does not reduce the cwnd, it is done by the + * congestion control later. * * It does _not_ decide what to send, it is made in function * tcp_xmit_retransmit_queue(). */ static void tcp_fastretrans_alert(struct sock *sk, const int acked, - const int prior_unsacked, - bool is_dupack, int flag) + bool is_dupack, int *ack_flag, int *rexmit) { struct inet_connection_sock *icsk = inet_csk(sk); struct tcp_sock *tp = tcp_sk(sk); + int fast_rexmit = 0, flag = *ack_flag; bool do_lost = is_dupack || ((flag & FLAG_DATA_SACKED) && (tcp_fackets_out(tp) > tp->reordering)); - int fast_rexmit = 0; if (WARN_ON(!tp->packets_out && tp->sacked_out)) tp->sacked_out = 0; @@@ -2819,10 -2809,8 +2819,10 @@@ /* Use RACK to detect loss */ if (sysctl_tcp_recovery & TCP_RACK_LOST_RETRANS && - tcp_rack_mark_lost(sk)) + tcp_rack_mark_lost(sk)) { flag |= FLAG_LOST_RETRANS; + *ack_flag |= FLAG_LOST_RETRANS; + } /* E. Process state. */ switch (icsk->icsk_ca_state) { @@@ -2831,7 -2819,7 +2831,7 @@@ if (tcp_is_reno(tp) && is_dupack) tcp_add_reno_sack(sk); } else { - if (tcp_try_undo_partial(sk, acked, prior_unsacked, flag)) + if (tcp_try_undo_partial(sk, acked)) return; /* Partial ACK arrived. Force fast retransmit. */ do_lost = tcp_is_reno(tp) || @@@ -2843,7 -2831,7 +2843,7 @@@ } break; case TCP_CA_Loss: - tcp_process_loss(sk, flag, is_dupack); + tcp_process_loss(sk, flag, is_dupack, rexmit); if (icsk->icsk_ca_state != TCP_CA_Open && !(flag & FLAG_LOST_RETRANS)) return; @@@ -2860,7 -2848,7 +2860,7 @@@ tcp_try_undo_dsack(sk); if (!tcp_time_to_recover(sk, flag)) { - tcp_try_to_open(sk, flag, prior_unsacked); + tcp_try_to_open(sk, flag); return; } @@@ -2882,7 -2870,8 +2882,7 @@@ if (do_lost) tcp_update_scoreboard(sk, fast_rexmit); - tcp_cwnd_reduction(sk, prior_unsacked, fast_rexmit, flag); - tcp_xmit_retransmit_queue(sk); + *rexmit = REXMIT_LOST; } /* Kathleen Nichols' algorithm for tracking the minimum value of @@@ -2907,7 -2896,10 +2907,10 @@@ static void tcp_update_rtt_min(struct s { const u32 now = tcp_time_stamp, wlen = sysctl_tcp_min_rtt_wlen * HZ; struct rtt_meas *m = tcp_sk(sk)->rtt_min; - struct rtt_meas rttm = { .rtt = (rtt_us ? : 1), .ts = now }; + struct rtt_meas rttm = { + .rtt = likely(rtt_us) ? rtt_us : jiffies_to_usecs(1), + .ts = now, + }; u32 elapsed; /* Check if the new measurement updates the 1st, 2nd, or 3rd choices */ @@@ -3104,7 -3096,7 +3107,7 @@@ static void tcp_ack_tstamp(struct sock * arrived at the other end. */ static int tcp_clean_rtx_queue(struct sock *sk, int prior_fackets, - u32 prior_snd_una, + u32 prior_snd_una, int *acked, struct tcp_sacktag_state *sack) { const struct inet_connection_sock *icsk = inet_csk(sk); @@@ -3162,13 -3154,10 +3165,13 @@@ flag |= FLAG_ORIG_SACK_ACKED; } - if (sacked & TCPCB_SACKED_ACKED) + if (sacked & TCPCB_SACKED_ACKED) { tp->sacked_out -= acked_pcount; - else if (tcp_is_sack(tp) && !tcp_skb_spurious_retrans(tp, skb)) - tcp_rack_advance(tp, &skb->skb_mstamp, sacked); + } else if (tcp_is_sack(tp)) { + tp->delivered += acked_pcount; + if (!tcp_skb_spurious_retrans(tp, skb)) + tcp_rack_advance(tp, &skb->skb_mstamp, sacked); + } if (sacked & TCPCB_LOST) tp->lost_out -= acked_pcount; @@@ -3277,7 -3266,6 +3280,7 @@@ } } #endif + *acked = pkts_acked; return flag; } @@@ -3311,36 -3299,21 +3314,36 @@@ static inline bool tcp_ack_is_dubious(c /* Decide wheather to run the increase function of congestion control. */ static inline bool tcp_may_raise_cwnd(const struct sock *sk, const int flag) { - if (tcp_in_cwnd_reduction(sk)) - return false; - /* If reordering is high then always grow cwnd whenever data is * delivered regardless of its ordering. Otherwise stay conservative * and only grow cwnd on in-order delivery (RFC5681). A stretched ACK w/ * new SACK or ECE mark may first advance cwnd here and later reduce * cwnd in tcp_fastretrans_alert() based on more states. */ - if (tcp_sk(sk)->reordering > sysctl_tcp_reordering) + if (tcp_sk(sk)->reordering > sock_net(sk)->ipv4.sysctl_tcp_reordering) return flag & FLAG_FORWARD_PROGRESS; return flag & FLAG_DATA_ACKED; } +/* The "ultimate" congestion control function that aims to replace the rigid + * cwnd increase and decrease control (tcp_cong_avoid,tcp_*cwnd_reduction). + * It's called toward the end of processing an ACK with precise rate + * information. All transmission or retransmission are delayed afterwards. + */ +static void tcp_cong_control(struct sock *sk, u32 ack, u32 acked_sacked, + int flag) +{ + if (tcp_in_cwnd_reduction(sk)) { + /* Reduce cwnd if state mandates */ + tcp_cwnd_reduction(sk, acked_sacked, flag); + } else if (tcp_may_raise_cwnd(sk, flag)) { + /* Advance cwnd if state allows */ + tcp_cong_avoid(sk, ack, acked_sacked); + } + tcp_update_pacing_rate(sk); +} + /* Check that window update is acceptable. * The function assumes that snd_una<=ack<=snd_next. */ @@@ -3536,27 -3509,6 +3539,27 @@@ static inline void tcp_in_ack_event(str icsk->icsk_ca_ops->in_ack_event(sk, flags); } +/* Congestion control has updated the cwnd already. So if we're in + * loss recovery then now we do any new sends (for FRTO) or + * retransmits (for CA_Loss or CA_recovery) that make sense. + */ +static void tcp_xmit_recovery(struct sock *sk, int rexmit) +{ + struct tcp_sock *tp = tcp_sk(sk); + + if (rexmit == REXMIT_NONE) + return; + + if (unlikely(rexmit == 2)) { + __tcp_push_pending_frames(sk, tcp_current_mss(sk), + TCP_NAGLE_OFF); + if (after(tp->snd_nxt, tp->high_seq)) + return; + tp->frto = 0; + } + tcp_xmit_retransmit_queue(sk); +} + /* This routine deals with incoming acks, but not outgoing ones. */ static int tcp_ack(struct sock *sk, const struct sk_buff *skb, int flag) { @@@ -3569,9 -3521,8 +3572,9 @@@ bool is_dupack = false; u32 prior_fackets; int prior_packets = tp->packets_out; - const int prior_unsacked = tp->packets_out - tp->sacked_out; + u32 prior_delivered = tp->delivered; int acked = 0; /* Number of packets newly acked */ + int rexmit = REXMIT_NONE; /* Flag to (re)transmit to recover losses */ sack_state.first_sackt.v64 = 0; @@@ -3660,16 -3611,23 +3663,16 @@@ goto no_queue; /* See if we can take anything off of the retransmit queue. */ - acked = tp->packets_out; - flag |= tcp_clean_rtx_queue(sk, prior_fackets, prior_snd_una, + flag |= tcp_clean_rtx_queue(sk, prior_fackets, prior_snd_una, &acked, &sack_state); - acked -= tp->packets_out; if (tcp_ack_is_dubious(sk, flag)) { is_dupack = !(flag & (FLAG_SND_UNA_ADVANCED | FLAG_NOT_DUP)); - tcp_fastretrans_alert(sk, acked, prior_unsacked, - is_dupack, flag); + tcp_fastretrans_alert(sk, acked, is_dupack, &flag, &rexmit); } if (tp->tlp_high_seq) tcp_process_tlp_ack(sk, ack, flag); - /* Advance cwnd if state allows */ - if (tcp_may_raise_cwnd(sk, flag)) - tcp_cong_avoid(sk, ack, acked); - if ((flag & FLAG_FORWARD_PROGRESS) || !(flag & FLAG_NOT_DUP)) { struct dst_entry *dst = __sk_dst_get(sk); if (dst) @@@ -3678,14 -3636,14 +3681,14 @@@ if (icsk->icsk_pending == ICSK_TIME_RETRANS) tcp_schedule_loss_probe(sk); - tcp_update_pacing_rate(sk); + tcp_cong_control(sk, ack, tp->delivered - prior_delivered, flag); + tcp_xmit_recovery(sk, rexmit); return 1; no_queue: /* If data was DSACKed, see if we can undo a cwnd reduction. */ if (flag & FLAG_DSACKING_ACK) - tcp_fastretrans_alert(sk, acked, prior_unsacked, - is_dupack, flag); + tcp_fastretrans_alert(sk, acked, is_dupack, &flag, &rexmit); /* If this ack opens up a zero window, clear backoff. It was * being used to time the probes, and is probably far higher than * it needs to be for normal retransmission. @@@ -3708,8 -3666,8 +3711,8 @@@ old_ack if (TCP_SKB_CB(skb)->sacked) { flag |= tcp_sacktag_write_queue(sk, skb, prior_snd_una, &sack_state); - tcp_fastretrans_alert(sk, acked, prior_unsacked, - is_dupack, flag); + tcp_fastretrans_alert(sk, acked, is_dupack, &flag, &rexmit); + tcp_xmit_recovery(sk, rexmit); } SOCK_DEBUG(sk, "Ack %u before %u:%u\n", ack, tp->snd_una, tp->snd_nxt); @@@ -4040,7 -3998,7 +4043,7 @@@ void tcp_reset(struct sock *sk * * If we are in FINWAIT-2, a received FIN moves us to TIME-WAIT. */ -static void tcp_fin(struct sock *sk) +void tcp_fin(struct sock *sk) { struct tcp_sock *tp = tcp_sk(sk); @@@ -5554,9 -5512,6 +5557,9 @@@ static bool tcp_rcv_fastopen_synack(str tp->syn_data_acked = tp->syn_data; if (tp->syn_data_acked) NET_INC_STATS_BH(sock_net(sk), LINUX_MIB_TCPFASTOPENACTIVE); + + tcp_fastopen_add_skb(sk, synack); + return false; } @@@ -6163,10 -6118,9 +6166,10 @@@ static bool tcp_syn_flood_action(const struct request_sock_queue *queue = &inet_csk(sk)->icsk_accept_queue; const char *msg = "Dropping request"; bool want_cookie = false; + struct net *net = sock_net(sk); #ifdef CONFIG_SYN_COOKIES - if (sysctl_tcp_syncookies) { + if (net->ipv4.sysctl_tcp_syncookies) { msg = "Sending cookies"; want_cookie = true; NET_INC_STATS_BH(sock_net(sk), LINUX_MIB_TCPREQQFULLDOCOOKIES); @@@ -6175,7 -6129,7 +6178,7 @@@ NET_INC_STATS_BH(sock_net(sk), LINUX_MIB_TCPREQQFULLDROP); if (!queue->synflood_warned && - sysctl_tcp_syncookies != 2 && + net->ipv4.sysctl_tcp_syncookies != 2 && xchg(&queue->synflood_warned, 1) == 0) pr_info("%s: Possible SYN flooding on port %d. %s. Check SNMP counters.\n", proto, ntohs(tcp_hdr(skb)->dest), msg); @@@ -6208,7 -6162,6 +6211,7 @@@ int tcp_conn_request(struct request_soc __u32 isn = TCP_SKB_CB(skb)->tcp_tw_isn; struct tcp_options_received tmp_opt; struct tcp_sock *tp = tcp_sk(sk); + struct net *net = sock_net(sk); struct sock *fastopen_sk = NULL; struct dst_entry *dst = NULL; struct request_sock *req; @@@ -6219,7 -6172,7 +6222,7 @@@ * limitations, they conserve resources and peer is * evidently real one. */ - if ((sysctl_tcp_syncookies == 2 || + if ((net->ipv4.sysctl_tcp_syncookies == 2 || inet_csk_reqsk_queue_is_full(sk)) && !isn) { want_cookie = tcp_syn_flood_action(sk, skb, rsk_ops->slab_name); if (!want_cookie) @@@ -6285,7 -6238,7 +6288,7 @@@ } } /* Kill the following clause, if you dislike this way. */ - else if (!sysctl_tcp_syncookies && + else if (!net->ipv4.sysctl_tcp_syncookies && (sysctl_max_syn_backlog - inet_csk_reqsk_queue_len(sk) < (sysctl_max_syn_backlog >> 2)) && !tcp_peer_is_proven(req, dst, false, diff --combined net/ipv4/tcp_ipv4.c index 3f872a6,487ac67..4c8d58d --- a/net/ipv4/tcp_ipv4.c +++ b/net/ipv4/tcp_ipv4.c @@@ -311,7 -311,7 +311,7 @@@ static void do_redirect(struct sk_buff /* handle ICMP messages on TCP_NEW_SYN_RECV request sockets */ - void tcp_req_err(struct sock *sk, u32 seq) + void tcp_req_err(struct sock *sk, u32 seq, bool abort) { struct request_sock *req = inet_reqsk(sk); struct net *net = sock_net(sk); @@@ -323,7 -323,7 +323,7 @@@ if (seq != tcp_rsk(req)->snt_isn) { NET_INC_STATS_BH(net, LINUX_MIB_OUTOFWINDOWICMPS); - } else { + } else if (abort) { /* * Still in SYN_RECV, just remove it silently. * There is no good way to pass the error to the newly @@@ -383,7 -383,12 +383,12 @@@ void tcp_v4_err(struct sk_buff *icmp_sk } seq = ntohl(th->seq); if (sk->sk_state == TCP_NEW_SYN_RECV) - return tcp_req_err(sk, seq); + return tcp_req_err(sk, seq, + type == ICMP_PARAMETERPROB || + type == ICMP_TIME_EXCEEDED || + (type == ICMP_DEST_UNREACH && + (code == ICMP_NET_UNREACH || + code == ICMP_HOST_UNREACH))); bh_lock_sock(sk); /* If too many ICMPs get dropped on busy @@@ -637,8 -642,8 +642,8 @@@ static void tcp_v4_send_reset(const str * Incoming packet is checked with md5 hash with finding key, * no RST generated if md5 hash doesn't match. */ - sk1 = __inet_lookup_listener(net, - &tcp_hashinfo, ip_hdr(skb)->saddr, + sk1 = __inet_lookup_listener(net, &tcp_hashinfo, NULL, 0, + ip_hdr(skb)->saddr, th->source, ip_hdr(skb)->daddr, ntohs(th->source), inet_iif(skb)); /* don't send rst if it can't find key */ @@@ -860,6 -865,7 +865,6 @@@ static void tcp_v4_reqsk_destructor(str kfree(inet_rsk(req)->opt); } - #ifdef CONFIG_TCP_MD5SIG /* * RFC2385 MD5 checksumming requires a mapping of @@@ -1581,8 -1587,7 +1586,8 @@@ int tcp_v4_rcv(struct sk_buff *skb TCP_SKB_CB(skb)->sacked = 0; lookup: - sk = __inet_lookup_skb(&tcp_hashinfo, skb, th->source, th->dest); + sk = __inet_lookup_skb(&tcp_hashinfo, skb, __tcp_hdrlen(th), th->source, + th->dest); if (!sk) goto no_tcp_socket; @@@ -1592,28 -1597,30 +1597,30 @@@ process if (sk->sk_state == TCP_NEW_SYN_RECV) { struct request_sock *req = inet_reqsk(sk); - struct sock *nsk = NULL; + struct sock *nsk; sk = req->rsk_listener; - if (tcp_v4_inbound_md5_hash(sk, skb)) - goto discard_and_relse; - if (likely(sk->sk_state == TCP_LISTEN)) { - nsk = tcp_check_req(sk, skb, req, false); - } else { + if (unlikely(tcp_v4_inbound_md5_hash(sk, skb))) { + reqsk_put(req); + goto discard_it; + } + if (unlikely(sk->sk_state != TCP_LISTEN)) { inet_csk_reqsk_queue_drop_and_put(sk, req); goto lookup; } + sock_hold(sk); + nsk = tcp_check_req(sk, skb, req, false); if (!nsk) { reqsk_put(req); - goto discard_it; + goto discard_and_relse; } if (nsk == sk) { - sock_hold(sk); reqsk_put(req); } else if (tcp_child_process(sk, nsk, skb)) { tcp_v4_send_reset(nsk, skb); - goto discard_it; + goto discard_and_relse; } else { + sock_put(sk); return 0; } } @@@ -1696,8 -1703,7 +1703,8 @@@ do_time_wait switch (tcp_timewait_state_process(inet_twsk(sk), skb, th)) { case TCP_TW_SYN: { struct sock *sk2 = inet_lookup_listener(dev_net(skb->dev), - &tcp_hashinfo, + &tcp_hashinfo, skb, + __tcp_hdrlen(th), iph->saddr, th->source, iph->daddr, th->dest, inet_iif(skb)); @@@ -2389,16 -2395,6 +2396,16 @@@ static int __net_init tcp_sk_init(struc net->ipv4.sysctl_tcp_keepalive_probes = TCP_KEEPALIVE_PROBES; net->ipv4.sysctl_tcp_keepalive_intvl = TCP_KEEPALIVE_INTVL; + net->ipv4.sysctl_tcp_syn_retries = TCP_SYN_RETRIES; + net->ipv4.sysctl_tcp_synack_retries = TCP_SYNACK_RETRIES; + net->ipv4.sysctl_tcp_syncookies = 1; + net->ipv4.sysctl_tcp_reordering = TCP_FASTRETRANS_THRESH; + net->ipv4.sysctl_tcp_retries1 = TCP_RETR1; + net->ipv4.sysctl_tcp_retries2 = TCP_RETR2; + net->ipv4.sysctl_tcp_orphan_retries = 0; + net->ipv4.sysctl_tcp_fin_timeout = TCP_FIN_TIMEOUT; + net->ipv4.sysctl_tcp_notsent_lowat = UINT_MAX; + return 0; fail: tcp_sk_exit(net); diff --combined net/ipv4/udp.c index 9fc4e9c,95d2f19..836abe5 --- a/net/ipv4/udp.c +++ b/net/ipv4/udp.c @@@ -356,8 -356,8 +356,8 @@@ EXPORT_SYMBOL(udp_lib_get_port) * match_wildcard == false: addresses must be exactly the same, i.e. * 0.0.0.0 only equals to 0.0.0.0 */ -static int ipv4_rcv_saddr_equal(const struct sock *sk1, const struct sock *sk2, - bool match_wildcard) +int ipv4_rcv_saddr_equal(const struct sock *sk1, const struct sock *sk2, + bool match_wildcard) { struct inet_sock *inet1 = inet_sk(sk1), *inet2 = inet_sk(sk2); @@@ -848,20 -848,32 +848,20 @@@ void udp_set_csum(bool nocheck, struct { struct udphdr *uh = udp_hdr(skb); - if (nocheck) + if (nocheck) { uh->check = 0; - else if (skb_is_gso(skb)) + } else if (skb_is_gso(skb)) { uh->check = ~udp_v4_check(len, saddr, daddr, 0); - else if (skb_dst(skb) && skb_dst(skb)->dev && - (skb_dst(skb)->dev->features & - (NETIF_F_IP_CSUM | NETIF_F_HW_CSUM))) { - - BUG_ON(skb->ip_summed == CHECKSUM_PARTIAL); - + } else if (skb->ip_summed == CHECKSUM_PARTIAL) { + uh->check = 0; + uh->check = udp_v4_check(len, saddr, daddr, lco_csum(skb)); + if (uh->check == 0) + uh->check = CSUM_MANGLED_0; + } else { skb->ip_summed = CHECKSUM_PARTIAL; skb->csum_start = skb_transport_header(skb) - skb->head; skb->csum_offset = offsetof(struct udphdr, check); uh->check = ~udp_v4_check(len, saddr, daddr, 0); - } else { - __wsum csum; - - BUG_ON(skb->ip_summed == CHECKSUM_PARTIAL); - - uh->check = 0; - csum = skb_checksum(skb, 0, len, 0); - uh->check = udp_v4_check(len, saddr, daddr, csum); - if (uh->check == 0) - uh->check = CSUM_MANGLED_0; - - skb->ip_summed = CHECKSUM_UNNECESSARY; } } EXPORT_SYMBOL(udp_set_csum); @@@ -1036,8 -1048,10 +1036,10 @@@ int udp_sendmsg(struct sock *sk, struc if (msg->msg_controllen) { err = ip_cmsg_send(sock_net(sk), msg, &ipc, sk->sk_family == AF_INET6); - if (err) + if (unlikely(err)) { + kfree(ipc.opt); return err; + } if (ipc.opt) free = 1; connected = 0; diff --combined net/ipv6/addrconf.c index ac0ba9e,bdd7eac..4751f89 --- a/net/ipv6/addrconf.c +++ b/net/ipv6/addrconf.c @@@ -583,7 -583,7 +583,7 @@@ static int inet6_netconf_get_devconf(st if (err < 0) goto errout; - err = EINVAL; + err = -EINVAL; if (!tb[NETCONFA_IFINDEX]) goto errout; @@@ -3538,6 -3538,7 +3538,7 @@@ static void addrconf_dad_begin(struct i { struct inet6_dev *idev = ifp->idev; struct net_device *dev = idev->dev; + bool notify = false; addrconf_join_solict(dev, &ifp->addr); @@@ -3583,7 -3584,7 +3584,7 @@@ /* Because optimistic nodes can use this address, * notify listeners. If DAD fails, RTM_DELADDR is sent. */ - ipv6_ifa_notify(RTM_NEWADDR, ifp); + notify = true; } } @@@ -3591,6 -3592,8 +3592,8 @@@ out: spin_unlock(&ifp->lock); read_unlock_bh(&idev->lock); + if (notify) + ipv6_ifa_notify(RTM_NEWADDR, ifp); } static void addrconf_dad_start(struct inet6_ifaddr *ifp) @@@ -4711,8 -4714,6 +4714,8 @@@ static inline void ipv6_store_devconf(s array[DEVCONF_IGNORE_ROUTES_WITH_LINKDOWN] = cnf->ignore_routes_with_linkdown; /* we omit DEVCONF_STABLE_SECRET for now */ array[DEVCONF_USE_OIF_ADDRS_ONLY] = cnf->use_oif_addrs_only; + array[DEVCONF_DROP_UNICAST_IN_L2_MULTICAST] = cnf->drop_unicast_in_l2_multicast; + array[DEVCONF_DROP_UNSOLICITED_NA] = cnf->drop_unsolicited_na; } static inline size_t inet6_ifla6_size(void) @@@ -5787,20 -5788,6 +5790,20 @@@ static struct addrconf_sysctl_tabl .proc_handler = addrconf_sysctl_ignore_routes_with_linkdown, }, { + .procname = "drop_unicast_in_l2_multicast", + .data = &ipv6_devconf.drop_unicast_in_l2_multicast, + .maxlen = sizeof(int), + .mode = 0644, + .proc_handler = proc_dointvec, + }, + { + .procname = "drop_unsolicited_na", + .data = &ipv6_devconf.drop_unsolicited_na, + .maxlen = sizeof(int), + .mode = 0644, + .proc_handler = proc_dointvec, + }, + { /* sentinel */ } }, diff --combined net/ipv6/ip6_gre.c index a94e506,a69aad1..f7c9560 --- a/net/ipv6/ip6_gre.c +++ b/net/ipv6/ip6_gre.c @@@ -360,7 -360,7 +360,7 @@@ static void ip6gre_tunnel_uninit(struc struct ip6gre_net *ign = net_generic(t->net, ip6gre_net_id); ip6gre_tunnel_unlink(ign, t); - ip6_tnl_dst_reset(t); + dst_cache_reset(&t->dst_cache); dev_put(dev); } @@@ -633,7 -633,7 +633,7 @@@ static netdev_tx_t ip6gre_xmit2(struct } if (!fl6->flowi6_mark) - dst = ip6_tnl_dst_get(tunnel); + dst = dst_cache_get(&tunnel->dst_cache); if (!dst) { dst = ip6_route_output(net, NULL, fl6); @@@ -702,7 -702,7 +702,7 @@@ } if (!fl6->flowi6_mark && ndst) - ip6_tnl_dst_set(tunnel, ndst); + dst_cache_set_ip6(&tunnel->dst_cache, ndst, &fl6->saddr); skb_dst_set(skb, dst); proto = NEXTHDR_GRE; @@@ -1009,7 -1009,7 +1009,7 @@@ static int ip6gre_tnl_change(struct ip6 t->parms.o_key = p->o_key; t->parms.i_flags = p->i_flags; t->parms.o_flags = p->o_flags; - ip6_tnl_dst_reset(t); + dst_cache_reset(&t->dst_cache); ip6gre_tnl_link_config(t, set_mtu); return 0; } @@@ -1219,7 -1219,7 +1219,7 @@@ static void ip6gre_dev_free(struct net_ { struct ip6_tnl *t = netdev_priv(dev); - ip6_tnl_dst_destroy(t); + dst_cache_destroy(&t->dst_cache); free_percpu(dev->tstats); free_netdev(dev); } @@@ -1257,7 -1257,7 +1257,7 @@@ static int ip6gre_tunnel_init_common(st if (!dev->tstats) return -ENOMEM; - ret = ip6_tnl_dst_init(tunnel); + ret = dst_cache_init(&tunnel->dst_cache, GFP_KERNEL); if (ret) { free_percpu(dev->tstats); dev->tstats = NULL; @@@ -1512,6 -1512,7 +1512,7 @@@ static void ip6gre_tap_setup(struct net dev->destructor = ip6gre_dev_free; dev->features |= NETIF_F_NETNS_LOCAL; + dev->priv_flags &= ~IFF_TX_SKB_SHARING; } static int ip6gre_newlink(struct net *src_net, struct net_device *dev, diff --combined net/ipv6/tcp_ipv6.c index 9977b6f,5c8c842..33f2820 --- a/net/ipv6/tcp_ipv6.c +++ b/net/ipv6/tcp_ipv6.c @@@ -327,6 -327,7 +327,7 @@@ static void tcp_v6_err(struct sk_buff * struct tcp_sock *tp; __u32 seq, snd_una; struct sock *sk; + bool fatal; int err; sk = __inet6_lookup_established(net, &tcp_hashinfo, @@@ -345,8 -346,9 +346,9 @@@ return; } seq = ntohl(th->seq); + fatal = icmpv6_err_convert(type, code, &err); if (sk->sk_state == TCP_NEW_SYN_RECV) - return tcp_req_err(sk, seq); + return tcp_req_err(sk, seq, fatal); bh_lock_sock(sk); if (sock_owned_by_user(sk) && type != ICMPV6_PKT_TOOBIG) @@@ -400,7 -402,6 +402,6 @@@ goto out; } - icmpv6_err_convert(type, code, &err); /* Might be for an request_sock */ switch (sk->sk_state) { @@@ -866,8 -867,7 +867,8 @@@ static void tcp_v6_send_reset(const str * no RST generated if md5 hash doesn't match. */ sk1 = inet6_lookup_listener(dev_net(skb_dst(skb)->dev), - &tcp_hashinfo, &ipv6h->saddr, + &tcp_hashinfo, NULL, 0, + &ipv6h->saddr, th->source, &ipv6h->daddr, ntohs(th->source), tcp_v6_iif(skb)); if (!sk1) @@@ -1376,8 -1376,8 +1377,8 @@@ static int tcp_v6_rcv(struct sk_buff *s hdr = ipv6_hdr(skb); lookup: - sk = __inet6_lookup_skb(&tcp_hashinfo, skb, th->source, th->dest, - inet6_iif(skb)); + sk = __inet6_lookup_skb(&tcp_hashinfo, skb, __tcp_hdrlen(th), + th->source, th->dest, inet6_iif(skb)); if (!sk) goto no_tcp_socket; @@@ -1387,7 -1387,7 +1388,7 @@@ process if (sk->sk_state == TCP_NEW_SYN_RECV) { struct request_sock *req = inet_reqsk(sk); - struct sock *nsk = NULL; + struct sock *nsk; sk = req->rsk_listener; tcp_v6_fill_cb(skb, hdr, th); @@@ -1395,24 -1395,24 +1396,24 @@@ reqsk_put(req); goto discard_it; } - if (likely(sk->sk_state == TCP_LISTEN)) { - nsk = tcp_check_req(sk, skb, req, false); - } else { + if (unlikely(sk->sk_state != TCP_LISTEN)) { inet_csk_reqsk_queue_drop_and_put(sk, req); goto lookup; } + sock_hold(sk); + nsk = tcp_check_req(sk, skb, req, false); if (!nsk) { reqsk_put(req); - goto discard_it; + goto discard_and_relse; } if (nsk == sk) { - sock_hold(sk); reqsk_put(req); tcp_v6_restore_cb(skb); } else if (tcp_child_process(sk, nsk, skb)) { tcp_v6_send_reset(nsk, skb); - goto discard_it; + goto discard_and_relse; } else { + sock_put(sk); return 0; } } @@@ -1501,7 -1501,6 +1502,7 @@@ do_time_wait struct sock *sk2; sk2 = inet6_lookup_listener(dev_net(skb->dev), &tcp_hashinfo, + skb, __tcp_hdrlen(th), &ipv6_hdr(skb)->saddr, th->source, &ipv6_hdr(skb)->daddr, ntohs(th->dest), tcp_v6_iif(skb)); @@@ -1867,7 -1866,7 +1868,7 @@@ struct proto tcpv6_prot = .sendpage = tcp_sendpage, .backlog_rcv = tcp_v6_do_rcv, .release_cb = tcp_release_cb, - .hash = inet_hash, + .hash = inet6_hash, .unhash = inet_unhash, .get_port = inet_csk_get_port, .enter_memory_pressure = tcp_enter_memory_pressure, diff --combined net/netfilter/nfnetlink.c index 9a99f68,857ae89..2278d9a --- a/net/netfilter/nfnetlink.c +++ b/net/netfilter/nfnetlink.c @@@ -127,6 -127,13 +127,6 @@@ int nfnetlink_has_listeners(struct net } EXPORT_SYMBOL_GPL(nfnetlink_has_listeners); -struct sk_buff *nfnetlink_alloc_skb(struct net *net, unsigned int size, - u32 dst_portid, gfp_t gfp_mask) -{ - return netlink_alloc_skb(net->nfnl, size, dst_portid, gfp_mask); -} -EXPORT_SYMBOL_GPL(nfnetlink_alloc_skb); - int nfnetlink_send(struct sk_buff *skb, struct net *net, u32 portid, unsigned int group, int echo, gfp_t flags) { @@@ -304,14 -311,14 +304,14 @@@ replay #endif { nfnl_unlock(subsys_id); - netlink_ack(skb, nlh, -EOPNOTSUPP); + netlink_ack(oskb, nlh, -EOPNOTSUPP); return kfree_skb(skb); } } if (!ss->commit || !ss->abort) { nfnl_unlock(subsys_id); - netlink_ack(skb, nlh, -EOPNOTSUPP); + netlink_ack(oskb, nlh, -EOPNOTSUPP); return kfree_skb(skb); } @@@ -321,10 -328,12 +321,12 @@@ nlh = nlmsg_hdr(skb); err = 0; - if (nlmsg_len(nlh) < sizeof(struct nfgenmsg) || - skb->len < nlh->nlmsg_len) { - err = -EINVAL; - goto ack; + if (nlh->nlmsg_len < NLMSG_HDRLEN || + skb->len < nlh->nlmsg_len || + nlmsg_len(nlh) < sizeof(struct nfgenmsg)) { + nfnl_err_reset(&err_list); + status |= NFNL_BATCH_FAILURE; + goto done; } /* Only requests are handled by the kernel */ @@@ -399,7 -408,7 +401,7 @@@ ack * pointing to the batch header. */ nfnl_err_reset(&err_list); - netlink_ack(skb, nlmsg_hdr(oskb), -ENOMEM); + netlink_ack(oskb, nlmsg_hdr(oskb), -ENOMEM); status |= NFNL_BATCH_FAILURE; goto done; } diff --combined net/sched/sch_api.c index c9673b5e,af1acf0..de1e176 --- a/net/sched/sch_api.c +++ b/net/sched/sch_api.c @@@ -1841,7 -1841,7 +1841,7 @@@ reclassify return err; } - return -1; + return TC_ACT_UNSPEC; /* signal: continue lookup */ #ifdef CONFIG_NET_CLS_ACT reset: if (unlikely(limit++ >= MAX_REC_LOOP)) { @@@ -1852,6 -1852,7 +1852,7 @@@ } tp = old_tp; + protocol = tc_skb_protocol(skb); goto reclassify; #endif } diff --combined net/sctp/socket.c index 6427b9d,e878da0..b89501e --- a/net/sctp/socket.c +++ b/net/sctp/socket.c @@@ -5538,6 -5538,7 +5538,7 @@@ static int sctp_getsockopt_hmac_ident(s struct sctp_hmac_algo_param *hmacs; __u16 data_len = 0; u32 num_idents; + int i; if (!ep->auth_enable) return -EACCES; @@@ -5555,8 -5556,12 +5556,12 @@@ return -EFAULT; if (put_user(num_idents, &p->shmac_num_idents)) return -EFAULT; - if (copy_to_user(p->shmac_idents, hmacs->hmac_ids, data_len)) - return -EFAULT; + for (i = 0; i < num_idents; i++) { + __u16 hmacid = ntohs(hmacs->hmac_ids[i]); + + if (copy_to_user(&p->shmac_idents[i], &hmacid, sizeof(__u16))) + return -EFAULT; + } return 0; } @@@ -6101,10 -6106,9 +6106,10 @@@ static int sctp_getsockopt(struct sock return retval; } -static void sctp_hash(struct sock *sk) +static int sctp_hash(struct sock *sk) { /* STUB */ + return 0; } static void sctp_unhash(struct sock *sk) diff --combined net/tipc/link.c index 3e513da,347cdc9..e31d92f --- a/net/tipc/link.c +++ b/net/tipc/link.c @@@ -123,6 -123,7 +123,6 @@@ struct tipc_stats struct tipc_link { u32 addr; char name[TIPC_MAX_LINK_NAME]; - struct tipc_media_addr *media_addr; struct net *net; /* Management and link supervision data */ @@@ -903,10 -904,8 +903,10 @@@ int tipc_link_xmit(struct tipc_link *l if (unlikely(l->backlog[i].len >= l->backlog[i].limit)) return link_schedule_user(l, list); } - if (unlikely(msg_size(hdr) > mtu)) + if (unlikely(msg_size(hdr) > mtu)) { + skb_queue_purge(list); return -EMSGSIZE; + } /* Prepare each packet for sending, and add to relevant queue: */ while (skb_queue_len(list)) { @@@ -918,10 -917,8 +918,10 @@@ if (likely(skb_queue_len(transmq) < maxwin)) { _skb = skb_clone(skb, GFP_ATOMIC); - if (!_skb) + if (!_skb) { + skb_queue_purge(list); return -ENOBUFS; + } __skb_dequeue(list); __skb_queue_tail(transmq, skb); __skb_queue_tail(xmitq, _skb); @@@ -1264,6 -1261,26 +1264,6 @@@ drop return rc; } -/* - * Send protocol message to the other endpoint. - */ -static void tipc_link_proto_xmit(struct tipc_link *l, u32 msg_typ, - int probe_msg, u32 gap, u32 tolerance, - u32 priority) -{ - struct sk_buff *skb = NULL; - struct sk_buff_head xmitq; - - __skb_queue_head_init(&xmitq); - tipc_link_build_proto_msg(l, msg_typ, probe_msg, gap, - tolerance, priority, &xmitq); - skb = __skb_dequeue(&xmitq); - if (!skb) - return; - tipc_bearer_xmit_skb(l->net, l->bearer_id, skb, l->media_addr); - l->rcv_unacked = 0; -} - static void tipc_link_build_proto_msg(struct tipc_link *l, int mtyp, bool probe, u16 rcvgap, int tolerance, int priority, struct sk_buff_head *xmitq) @@@ -1462,12 -1479,6 +1462,12 @@@ static int tipc_link_proto_rcv(struct t if (in_range(peers_tol, TIPC_MIN_LINK_TOL, TIPC_MAX_LINK_TOL)) l->tolerance = peers_tol; + if (peers_prio && in_range(peers_prio, TIPC_MIN_LINK_PRI, + TIPC_MAX_LINK_PRI)) { + l->priority = peers_prio; + rc = tipc_link_fsm_evt(l, LINK_FAILURE_EVT); + } + l->silent_intv_cnt = 0; l->stats.recv_states++; if (msg_probe(hdr)) @@@ -1962,8 -1973,10 +1962,10 @@@ int tipc_nl_add_bc_link(struct net *net hdr = genlmsg_put(msg->skb, msg->portid, msg->seq, &tipc_genl_family, NLM_F_MULTI, TIPC_NL_LINK_GET); - if (!hdr) + if (!hdr) { + tipc_bcast_unlock(net); return -EMSGSIZE; + } attrs = nla_nest_start(msg->skb, TIPC_NLA_LINK); if (!attrs) @@@ -2010,18 -2023,16 +2012,18 @@@ msg_full return -EMSGSIZE; } -void tipc_link_set_tolerance(struct tipc_link *l, u32 tol) +void tipc_link_set_tolerance(struct tipc_link *l, u32 tol, + struct sk_buff_head *xmitq) { l->tolerance = tol; - tipc_link_proto_xmit(l, STATE_MSG, 0, 0, tol, 0); + tipc_link_build_proto_msg(l, STATE_MSG, 0, 0, tol, 0, xmitq); } -void tipc_link_set_prio(struct tipc_link *l, u32 prio) +void tipc_link_set_prio(struct tipc_link *l, u32 prio, + struct sk_buff_head *xmitq) { l->priority = prio; - tipc_link_proto_xmit(l, STATE_MSG, 0, 0, 0, prio); + tipc_link_build_proto_msg(l, STATE_MSG, 0, 0, 0, prio, xmitq); } void tipc_link_set_abort_limit(struct tipc_link *l, u32 limit) diff --combined net/tipc/node.c index 10a1e87,9d7a16f..9fcc2fb --- a/net/tipc/node.c +++ b/net/tipc/node.c @@@ -346,12 -346,6 +346,6 @@@ struct tipc_node *tipc_node_create(stru skb_queue_head_init(&n->bc_entry.inputq2); for (i = 0; i < MAX_BEARERS; i++) spin_lock_init(&n->links[i].lock); - hlist_add_head_rcu(&n->hash, &tn->node_htable[tipc_hashfn(addr)]); - list_for_each_entry_rcu(temp_node, &tn->node_list, list) { - if (n->addr < temp_node->addr) - break; - } - list_add_tail_rcu(&n->list, &temp_node->list); n->state = SELF_DOWN_PEER_LEAVING; n->signature = INVALID_NODE_SIG; n->active_links[0] = INVALID_BEARER_ID; @@@ -372,6 -366,12 +366,12 @@@ tipc_node_get(n); setup_timer(&n->timer, tipc_node_timeout, (unsigned long)n); n->keepalive_intv = U32_MAX; + hlist_add_head_rcu(&n->hash, &tn->node_htable[tipc_hashfn(addr)]); + list_for_each_entry_rcu(temp_node, &tn->node_list, list) { + if (n->addr < temp_node->addr) + break; + } + list_add_tail_rcu(&n->list, &temp_node->list); exit: spin_unlock_bh(&tn->node_list_lock); return n; @@@ -1166,7 -1166,7 +1166,7 @@@ msg_full * @dnode: address of destination node * @selector: a number used for deterministic link selection * Consumes the buffer chain, except when returning -ELINKCONG - * Returns 0 if success, otherwise errno: -ELINKCONG,-EHOSTUNREACH,-EMSGSIZE + * Returns 0 if success, otherwise: -ELINKCONG,-EHOSTUNREACH,-EMSGSIZE,-ENOBUF */ int tipc_node_xmit(struct net *net, struct sk_buff_head *list, u32 dnode, int selector) @@@ -1174,43 -1174,33 +1174,43 @@@ struct tipc_link_entry *le = NULL; struct tipc_node *n; struct sk_buff_head xmitq; - int bearer_id = -1; - int rc = -EHOSTUNREACH; + int bearer_id; + int rc; + + if (in_own_node(net, dnode)) { + tipc_sk_rcv(net, list); + return 0; + } - __skb_queue_head_init(&xmitq); n = tipc_node_find(net, dnode); - if (likely(n)) { - tipc_node_read_lock(n); - bearer_id = n->active_links[selector & 1]; - if (bearer_id >= 0) { - le = &n->links[bearer_id]; - spin_lock_bh(&le->lock); - rc = tipc_link_xmit(le->link, list, &xmitq); - spin_unlock_bh(&le->lock); - } + if (unlikely(!n)) { + skb_queue_purge(list); + return -EHOSTUNREACH; + } + + tipc_node_read_lock(n); + bearer_id = n->active_links[selector & 1]; + if (unlikely(bearer_id == INVALID_BEARER_ID)) { tipc_node_read_unlock(n); - if (likely(!rc)) - tipc_bearer_xmit(net, bearer_id, &xmitq, &le->maddr); - else if (rc == -ENOBUFS) - tipc_node_link_down(n, bearer_id, false); tipc_node_put(n); - return rc; + skb_queue_purge(list); + return -EHOSTUNREACH; } - if (likely(in_own_node(net, dnode))) { - tipc_sk_rcv(net, list); - return 0; - } + __skb_queue_head_init(&xmitq); + le = &n->links[bearer_id]; + spin_lock_bh(&le->lock); + rc = tipc_link_xmit(le->link, list, &xmitq); + spin_unlock_bh(&le->lock); + tipc_node_read_unlock(n); + + if (likely(rc == 0)) + tipc_bearer_xmit(net, bearer_id, &xmitq, &le->maddr); + else if (rc == -ENOBUFS) + tipc_node_link_down(n, bearer_id, false); + + tipc_node_put(n); + return rc; } @@@ -1647,12 -1637,9 +1647,12 @@@ int tipc_nl_node_set_link(struct sk_buf char *name; struct tipc_link *link; struct tipc_node *node; + struct sk_buff_head xmitq; struct nlattr *attrs[TIPC_NLA_LINK_MAX + 1]; struct net *net = sock_net(skb->sk); + __skb_queue_head_init(&xmitq); + if (!info->attrs[TIPC_NLA_LINK]) return -EINVAL; @@@ -1696,13 -1683,13 +1696,13 @@@ u32 tol; tol = nla_get_u32(props[TIPC_NLA_PROP_TOL]); - tipc_link_set_tolerance(link, tol); + tipc_link_set_tolerance(link, tol, &xmitq); } if (props[TIPC_NLA_PROP_PRIO]) { u32 prio; prio = nla_get_u32(props[TIPC_NLA_PROP_PRIO]); - tipc_link_set_prio(link, prio); + tipc_link_set_prio(link, prio, &xmitq); } if (props[TIPC_NLA_PROP_WIN]) { u32 win; @@@ -1714,7 -1701,7 +1714,7 @@@ out: tipc_node_read_unlock(node); - + tipc_bearer_xmit(net, bearer_id, &xmitq, &node->links[bearer_id].maddr); return res; } diff --combined net/unix/af_unix.c index b374555,f75f847..8269da7 --- a/net/unix/af_unix.c +++ b/net/unix/af_unix.c @@@ -1496,7 -1496,7 +1496,7 @@@ static void unix_detach_fds(struct scm_ UNIXCB(skb).fp = NULL; for (i = scm->fp->count-1; i >= 0; i--) - unix_notinflight(scm->fp->fp[i]); + unix_notinflight(scm->fp->user, scm->fp->fp[i]); } static void unix_destruct_scm(struct sk_buff *skb) @@@ -1534,6 -1534,7 +1534,6 @@@ static int unix_attach_fds(struct scm_c { int i; unsigned char max_level = 0; - int unix_sock_count = 0; if (too_many_unix_fds(current)) return -ETOOMANYREFS; @@@ -1541,9 -1542,11 +1541,9 @@@ for (i = scm->fp->count - 1; i >= 0; i--) { struct sock *sk = unix_get_socket(scm->fp->fp[i]); - if (sk) { - unix_sock_count++; + if (sk) max_level = max(max_level, unix_sk(sk)->recursion_level); - } } if (unlikely(max_level > MAX_RECURSION_LEVEL)) return -ETOOMANYREFS; @@@ -1558,7 -1561,7 +1558,7 @@@ return -ENOMEM; for (i = scm->fp->count - 1; i >= 0; i--) - unix_inflight(scm->fp->fp[i]); + unix_inflight(scm->fp->user, scm->fp->fp[i]); return max_level; } @@@ -1778,7 -1781,12 +1778,12 @@@ restart_locked goto out_unlock; } - if (unlikely(unix_peer(other) != sk && unix_recvq_full(other))) { + /* other == sk && unix_peer(other) != sk if + * - unix_peer(sk) == NULL, destination address bound to sk + * - unix_peer(sk) == sk by time of get but disconnected before lock + */ + if (other != sk && + unlikely(unix_peer(other) != sk && unix_recvq_full(other))) { if (timeo) { timeo = unix_wait_for_peer(other, timeo); @@@ -2274,13 -2282,15 +2279,15 @@@ static int unix_stream_read_generic(str size_t size = state->size; unsigned int last_len; - err = -EINVAL; - if (sk->sk_state != TCP_ESTABLISHED) + if (unlikely(sk->sk_state != TCP_ESTABLISHED)) { + err = -EINVAL; goto out; + } - err = -EOPNOTSUPP; - if (flags & MSG_OOB) + if (unlikely(flags & MSG_OOB)) { + err = -EOPNOTSUPP; goto out; + } target = sock_rcvlowat(sk, flags & MSG_WAITALL, size); timeo = sock_rcvtimeo(sk, noblock); @@@ -2302,6 -2312,7 +2309,7 @@@ bool drop_skb; struct sk_buff *skb, *last; + redo: unix_state_lock(sk); if (sock_flag(sk, SOCK_DEAD)) { err = -ECONNRESET; @@@ -2326,9 -2337,11 +2334,11 @@@ again goto unlock; unix_state_unlock(sk); - err = -EAGAIN; - if (!timeo) + if (!timeo) { + err = -EAGAIN; break; + } + mutex_unlock(&u->readlock); timeo = unix_stream_data_wait(sk, timeo, last, @@@ -2341,7 -2354,7 +2351,7 @@@ } mutex_lock(&u->readlock); - continue; + goto redo; unlock: unix_state_unlock(sk); break; -- LinuxNextTracking

7 years

Noticed difference in packet.h in maint.diff

by postmaster＠open-mesh.org

--- batman-adv_maint.packet.h 2016-02-25 00:05:02.153098114 +0100 +++ batctl_maint.packet.h 2016-02-25 00:05:02.165098914 +0100 @@ -1,4 +1,4 @@ -/* Copyright (C) 2007-2015 B.A.T.M.A.N. contributors: +/* Copyright (C) 2007-2016 B.A.T.M.A.N. contributors: * * Marek Lindner, Simon Wunderlich *

7 years

Build check errors found: 2016-02-24

by postmaster＠open-mesh.org

Name of failed tests ==================== * difference between net-next and batadv next Output of different failed tests ================================ difference between net-next and batadv next ------------------------------------------- Documentation/ABI/testing/sysfs-class-net-batman-adv | 17 net/batman-adv/Kconfig | 14 net/batman-adv/Makefile | 3 net/batman-adv/bat_algo.h | 28 net/batman-adv/bat_iv_ogm.c | 2 net/batman-adv/bat_v.c | 347 ---- net/batman-adv/bat_v_elp.c | 515 ------ net/batman-adv/bat_v_elp.h | 33 net/batman-adv/bat_v_ogm.c | 833 ---------- net/batman-adv/bat_v_ogm.h | 36 net/batman-adv/distributed-arp-table.c | 4 net/batman-adv/fragmentation.c | 8 net/batman-adv/gateway_common.c | 4 net/batman-adv/gateway_common.h | 2 net/batman-adv/icmp_socket.c | 2 net/batman-adv/main.c | 7 net/batman-adv/main.h | 12 net/batman-adv/network-coding.c | 22 net/batman-adv/packet.h | 49 net/batman-adv/send.c | 55 net/batman-adv/send.h | 10 net/batman-adv/sysfs.c | 130 - net/batman-adv/types.h | 93 - 23 files changed, 37 insertions(+), 2189 deletions(-)

7 years, 1 month

[linux-next] LinuxNextTracking branch, master, updated. next-20160223

by batman＠open-mesh.org

The following commit has been merged in the master branch: commit 32fb1f1fccc4eac30d5cf8d573a0c89c05f3f684 Merge: 67dccb15f7f35215d6f93bd3d625ae283b6c5f51 b1d95ae5c5bd3deba84d00c4f83d7d0836b5936f Author: Stephen Rothwell <sfr(a)canb.auug.org.au> Date: Tue Feb 23 11:01:51 2016 +1100 Merge remote-tracking branch 'net-next/master' diff --combined MAINTAINERS index b2169ee,355e1c8..12cbb39 --- a/MAINTAINERS +++ b/MAINTAINERS @@@ -223,7 -223,9 +223,7 @@@ F: drivers/scsi/aacraid ABI/API L: linux-api(a)vger.kernel.org -F: Documentation/ABI/ F: include/linux/syscalls.h -F: include/uapi/ F: kernel/sys_ni.c ABIT UGURU 1,2 HARDWARE MONITOR DRIVER @@@ -673,25 -675,24 +673,25 @@@ F: drivers/gpu/drm/radeon/radeon_kfd. F: drivers/gpu/drm/radeon/radeon_kfd.h F: include/uapi/linux/kfd_ioctl.h +AMD SEATTLE DEVICE TREE SUPPORT +M: Brijesh Singh <brijeshkumar.singh(a)amd.com> +M: Suravee Suthikulpanit <suravee.suthikulpanit(a)amd.com> +M: Tom Lendacky <thomas.lendacky(a)amd.com> +S: Supported +F: arch/arm64/boot/dts/amd/ + AMD XGBE DRIVER M: Tom Lendacky <thomas.lendacky(a)amd.com> L: netdev(a)vger.kernel.org S: Supported F: drivers/net/ethernet/amd/xgbe/ +F: arch/arm64/boot/dts/amd/amd-seattle-xgbe*.dtsi AMS (Apple Motion Sensor) DRIVER M: Michael Hanselmann <linux-kernel(a)hansmi.ch> S: Supported F: drivers/macintosh/ams/ -AMSO1100 RNIC DRIVER -M: Tom Tucker <tom(a)opengridcomputing.com> -M: Steve Wise <swise(a)opengridcomputing.com> -L: linux-rdma(a)vger.kernel.org -S: Maintained -F: drivers/infiniband/hw/amso1100/ - ANALOG DEVICES INC AD9389B DRIVER M: Hans Verkuil <hans.verkuil(a)cisco.com> L: linux-media(a)vger.kernel.org @@@ -966,8 -967,6 +966,8 @@@ M: Rob Herring <robh(a)kernel.org L: linux-arm-kernel(a)lists.infradead.org (moderated for non-subscribers) S: Maintained F: arch/arm/mach-highbank/ +F: arch/arm/boot/dts/highbank.dts +F: arch/arm/boot/dts/ecx-*.dts* ARM/CAVIUM NETWORKS CNS3XXX MACHINE SUPPORT M: Krzysztof Halasa <khalasa(a)piap.pl> @@@ -1043,7 -1042,6 +1043,7 @@@ M: Barry Song <baohua(a)kernel.org L: linux-arm-kernel(a)lists.infradead.org (moderated for non-subscribers) T: git git://git.kernel.org/pub/scm/linux/kernel/git/baohua/linux.git S: Maintained +F: arch/arm/boot/dts/prima2* F: arch/arm/mach-prima2/ F: drivers/clk/sirf/ F: drivers/clocksource/timer-prima2.c @@@ -1145,10 -1143,6 +1145,10 @@@ W: http://www.hisilicon.co S: Supported T: git git://github.com/hisilicon/linux-hisi.git F: arch/arm/mach-hisi/ +F: arch/arm/boot/dts/hi3* +F: arch/arm/boot/dts/hip* +F: arch/arm/boot/dts/hisi* +F: arch/arm64/boot/dts/hisilicon/ ARM/HP JORNADA 7XX MACHINE SUPPORT M: Kristoffer Ericson <kristoffer.ericson(a)gmail.com> @@@ -1225,7 -1219,6 +1225,7 @@@ M: Santosh Shilimkar <ssantosh(a)kernel.o L: linux-arm-kernel(a)lists.infradead.org (moderated for non-subscribers) S: Maintained F: arch/arm/mach-keystone/ +F: arch/arm/boot/dts/k2* T: git git://git.kernel.org/pub/scm/linux/kernel/git/ssantosh/linux-keystone.git ARM/TEXAS INSTRUMENT KEYSTONE CLOCK FRAMEWORK @@@ -1286,7 -1279,6 +1286,7 @@@ F: arch/arm/mach-mvebu F: drivers/rtc/rtc-armada38x.c F: arch/arm/boot/dts/armada* F: arch/arm/boot/dts/kirkwood* +F: arch/arm64/boot/dts/marvell/armada* ARM/Marvell Berlin SoC support @@@ -1295,7 -1287,6 +1295,7 @@@ L: linux-arm-kernel(a)lists.infradead.or S: Maintained F: arch/arm/mach-berlin/ F: arch/arm/boot/dts/berlin* +F: arch/arm64/boot/dts/marvell/berlin* ARM/Marvell Dove/MV78xx0/Orion SOC support @@@ -1434,10 -1425,7 +1434,10 @@@ S: Maintaine F: arch/arm/boot/dts/qcom-*.dts F: arch/arm/boot/dts/qcom-*.dtsi F: arch/arm/mach-qcom/ +F: arch/arm64/boot/dts/qcom/* +F: drivers/i2c/busses/i2c-qup.c F: drivers/soc/qcom/ +F: drivers/spi/spi-qup.c F: drivers/tty/serial/msm_serial.h F: drivers/tty/serial/msm_serial.c F: drivers/*/pm8???-* @@@ -1453,8 -1441,8 +1453,8 @@@ S: Maintaine ARM/RENESAS ARM64 ARCHITECTURE M: Simon Horman <horms(a)verge.net.au> M: Magnus Damm <magnus.damm(a)gmail.com> -L: linux-sh(a)vger.kernel.org -Q: http://patchwork.kernel.org/project/linux-sh/list/ +L: linux-renesas-soc(a)vger.kernel.org +Q: http://patchwork.kernel.org/project/linux-renesas-soc/list/ T: git git://git.kernel.org/pub/scm/linux/kernel/git/horms/renesas.git next S: Supported F: arch/arm64/boot/dts/renesas/ @@@ -1496,8 -1484,6 +1496,8 @@@ L: linux-arm-kernel(a)lists.infradead.or L: linux-samsung-soc(a)vger.kernel.org (moderated for non-subscribers) S: Maintained F: arch/arm/boot/dts/s3c* +F: arch/arm/boot/dts/s5p* +F: arch/arm/boot/dts/samsung* F: arch/arm/boot/dts/exynos* F: arch/arm64/boot/dts/exynos/ F: arch/arm/plat-samsung/ @@@ -1507,7 -1493,6 +1507,7 @@@ F: arch/arm/mach-s5p* F: arch/arm/mach-exynos*/ F: drivers/*/*s3c2410* F: drivers/*/*/*s3c2410* +F: drivers/soc/samsung/* F: drivers/spi/spi-s3c* F: sound/soc/samsung/* F: Documentation/arm/Samsung/ @@@ -1578,7 -1563,6 +1578,7 @@@ S: Maintaine F: arch/arm/mach-socfpga/ F: arch/arm/boot/dts/socfpga* F: arch/arm/configs/socfpga_defconfig +F: arch/arm64/boot/dts/altera/ W: http://www.rocketboards.org T: git git://git.kernel.org/pub/scm/linux/kernel/git/dinguyen/linux.git @@@ -1732,7 -1716,7 +1732,7 @@@ M: Lorenzo Pieralisi <lorenzo.pieralisi L: linux-arm-kernel(a)lists.infradead.org (moderated for non-subscribers) S: Maintained F: arch/arm/boot/dts/vexpress* -F: arch/arm64/boot/dts/arm/vexpress* +F: arch/arm64/boot/dts/arm/ F: arch/arm/mach-vexpress/ F: */*/vexpress* F: */*/*/vexpress* @@@ -2163,7 -2147,7 +2163,7 @@@ M: Marek Lindner <mareklindner@neomailb M: Simon Wunderlich <sw(a)simonwunderlich.de> M: Antonio Quartulli <a(a)unstable.cc> L: b.a.t.m.a.n(a)lists.open-mesh.org - W: http://www.open-mesh.org/ + W: https://www.open-mesh.org/ S: Maintained F: net/batman-adv/ @@@ -2359,7 -2343,6 +2359,7 @@@ F: arch/arm/mach-bcm F: arch/arm/boot/dts/bcm113* F: arch/arm/boot/dts/bcm216* F: arch/arm/boot/dts/bcm281* +F: arch/arm64/boot/dts/broadcom/ F: arch/arm/configs/bcm_defconfig F: drivers/mmc/host/sdhci-bcm-kona.c F: drivers/clocksource/bcm_kona_timer.c @@@ -2374,6 -2357,14 +2374,6 @@@ T: git git://git.kernel.org/pub/scm/lin S: Maintained N: bcm2835 -BROADCOM BCM33XX MIPS ARCHITECTURE -M: Kevin Cernekee <cernekee(a)gmail.com> -L: linux-mips(a)linux-mips.org -S: Maintained -F: arch/mips/bcm3384/* -F: arch/mips/include/asm/mach-bcm3384/* -F: arch/mips/kernel/*bmips* - BROADCOM BCM47XX MIPS ARCHITECTURE M: Hauke Mehrtens <hauke(a)hauke-m.de> M: Rafał Miłecki <zajec5(a)gmail.com> @@@ -2433,6 -2424,7 +2433,7 @@@ F: include/linux/bcm963xx_nvram. F: include/linux/bcm963xx_tag.h BROADCOM TG3 GIGABIT ETHERNET DRIVER + M: Siva Reddy Kallam <siva.kallam(a)broadcom.com> M: Prashant Sreedharan <prashant(a)broadcom.com> M: Michael Chan <mchan(a)broadcom.com> L: netdev(a)vger.kernel.org @@@ -3454,8 -3446,9 +3455,8 @@@ S: Maintaine F: drivers/usb/dwc2/ DESIGNWARE USB3 DRD IP DRIVER -M: Felipe Balbi <balbi(a)ti.com> +M: Felipe Balbi <balbi(a)kernel.org> L: linux-usb(a)vger.kernel.org -L: linux-omap(a)vger.kernel.org T: git git://git.kernel.org/pub/scm/linux/kernel/git/balbi/usb.git S: Maintained F: drivers/usb/dwc3/ @@@ -4192,6 -4185,13 +4193,6 @@@ W: http://aeschi.ch.eu.org/efs S: Orphan F: fs/efs/ -EHCA (IBM GX bus InfiniBand adapter) DRIVER -M: Hoang-Nam Nguyen <hnguyen(a)de.ibm.com> -M: Christoph Raisch <raisch(a)de.ibm.com> -L: linux-rdma(a)vger.kernel.org -S: Supported -F: drivers/infiniband/hw/ehca/ - EHEA (IBM pSeries eHEA 10Gb ethernet adapter) DRIVER M: Thadeu Lima de Souza Cascardo <cascardo(a)linux.vnet.ibm.com> L: netdev(a)vger.kernel.org @@@ -4233,6 -4233,13 +4234,6 @@@ M: Maxim Levitsky <maximlevitsky@gmail. S: Maintained F: drivers/media/rc/ene_ir.* -ENHANCED ERROR HANDLING (EEH) -M: Gavin Shan <shangw(a)linux.vnet.ibm.com> -L: linuxppc-dev(a)lists.ozlabs.org -S: Supported -F: Documentation/powerpc/eeh-pci-error-recovery.txt -F: arch/powerpc/kernel/eeh*.c - EPSON S1D13XXX FRAMEBUFFER DRIVER M: Kristoffer Ericson <kristoffer.ericson(a)gmail.com> S: Maintained @@@ -5803,6 -5810,12 +5804,6 @@@ M: Juanjo Ciarlante <jjciarla(a)raiz.uncu S: Maintained F: net/ipv4/netfilter/ipt_MASQUERADE.c -IPATH DRIVER -M: Mike Marciniszyn <infinipath(a)intel.com> -L: linux-rdma(a)vger.kernel.org -S: Maintained -F: drivers/staging/rdma/ipath/ - IPMI SUBSYSTEM M: Corey Minyard <minyard(a)acm.org> L: openipmi-developer(a)lists.sourceforge.net (moderated for non-subscribers) @@@ -6132,7 -6145,7 +6133,7 @@@ F: include/uapi/linux/sunrpc KERNEL SELFTEST FRAMEWORK M: Shuah Khan <shuahkh(a)osg.samsung.com> -L: linux-api(a)vger.kernel.org +L: linux-kselftest(a)vger.kernel.org T: git git://git.kernel.org/pub/scm/shuah/linux-kselftest S: Maintained F: tools/testing/selftests @@@ -7358,7 -7371,7 +7359,7 @@@ F: drivers/tty/isicom. F: include/linux/isicom.h MUSB MULTIPOINT HIGH SPEED DUAL-ROLE CONTROLLER -M: Felipe Balbi <balbi(a)ti.com> +M: Bin Liu <b-liu(a)ti.com> L: linux-usb(a)vger.kernel.org T: git git://git.kernel.org/pub/scm/linux/kernel/git/balbi/usb.git S: Maintained @@@ -7927,9 -7940,11 +7928,9 @@@ F: drivers/media/platform/omap3isp F: drivers/staging/media/omap4iss/ OMAP USB SUPPORT -M: Felipe Balbi <balbi(a)ti.com> L: linux-usb(a)vger.kernel.org L: linux-omap(a)vger.kernel.org -T: git git://git.kernel.org/pub/scm/linux/kernel/git/balbi/usb.git -S: Maintained +S: Orphan F: drivers/usb/*/*omap* F: arch/arm/*omap*/usb* @@@ -8113,14 -8128,6 +8114,14 @@@ S: Supporte F: fs/overlayfs/ F: Documentation/filesystems/overlayfs.txt +ORANGEFS FILESYSTEM +M: Mike Marshall <hubcap(a)omnibond.com> +L: pvfs2-developers(a)beowulf-underground.org +T: git git://git.kernel.org/pub/scm/linux/kernel/git/hubcap/linux.git +S: Supported +F: fs/orangefs/ +F: Documentation/filesystems/orangefs.txt + P54 WIRELESS DRIVER M: Christian Lamparter <chunkeey(a)googlemail.com> L: linux-wireless(a)vger.kernel.org @@@ -8254,15 -8261,6 +8255,15 @@@ L: linux-pci(a)vger.kernel.or S: Supported F: Documentation/PCI/pci-error-recovery.txt +PCI ENHANCED ERROR HANDLING (EEH) FOR POWERPC +M: Russell Currey <ruscur(a)russell.cc> +L: linuxppc-dev(a)lists.ozlabs.org +S: Supported +F: Documentation/powerpc/eeh-pci-error-recovery.txt +F: arch/powerpc/kernel/eeh*.c +F: arch/powerpc/platforms/*/eeh*.c +F: arch/powerpc/include/*/eeh*.h + PCI SUBSYSTEM M: Bjorn Helgaas <bhelgaas(a)google.com> L: linux-pci(a)vger.kernel.org @@@ -8821,7 -8819,6 +8822,7 @@@ L: linux-arm-kernel(a)lists.infradead.or T: git git://github.com/hzhuang1/linux.git T: git git://github.com/rjarzmik/linux.git S: Maintained +F: arch/arm/boot/dts/pxa* F: arch/arm/mach-pxa/ F: drivers/dma/pxa* F: drivers/pcmcia/pxa2xx* @@@ -8851,7 -8848,6 +8852,7 @@@ L: linux-arm-kernel(a)lists.infradead.or T: git git://github.com/hzhuang1/linux.git T: git git://git.linaro.org/people/ycmiao/pxa-linux.git S: Maintained +F: arch/arm/boot/dts/mmp* F: arch/arm/mach-mmp/ PXA MMCI DRIVER @@@ -9798,11 -9794,10 +9799,11 @@@ S: Supporte F: drivers/scsi/be2iscsi/ Emulex 10Gbps NIC BE2, BE3-R, Lancer, Skyhawk-R DRIVER -M: Sathya Perla <sathya.perla(a)avagotech.com> -M: Ajit Khaparde <ajit.khaparde(a)avagotech.com> -M: Padmanabh Ratnakar <padmanabh.ratnakar(a)avagotech.com> -M: Sriharsha Basavapatna <sriharsha.basavapatna(a)avagotech.com> +M: Sathya Perla <sathya.perla(a)broadcom.com> +M: Ajit Khaparde <ajit.khaparde(a)broadcom.com> +M: Padmanabh Ratnakar <padmanabh.ratnakar(a)broadcom.com> +M: Sriharsha Basavapatna <sriharsha.basavapatna(a)broadcom.com> +M: Somnath Kotur <somnath.kotur(a)broadcom.com> L: netdev(a)vger.kernel.org W: http://www.emulex.com S: Supported @@@ -10164,7 -10159,6 +10165,7 @@@ S: Supporte F: drivers/media/pci/solo6x10/ SOFTWARE RAID (Multiple Disks) SUPPORT +M: Shaohua Li <shli(a)kernel.org> L: linux-raid(a)vger.kernel.org T: git git://neil.brown.name/md S: Supported @@@ -10298,7 -10292,6 +10299,7 @@@ L: spear-devel(a)list.st.co L: linux-arm-kernel(a)lists.infradead.org (moderated for non-subscribers) W: http://www.st.com/spear S: Maintained +F: arch/arm/boot/dts/spear* F: arch/arm/mach-spear/ SPEAR CLOCK FRAMEWORK SUPPORT @@@ -10861,14 -10854,6 +10862,14 @@@ L: linux-omap(a)vger.kernel.or S: Maintained F: drivers/thermal/ti-soc-thermal/ +TI VPE/CAL DRIVERS +M: Benoit Parrot <bparrot(a)ti.com> +L: linux-media(a)vger.kernel.org +W: http://linuxtv.org/ +Q: http://patchwork.linuxtv.org/project/linux-media/list/ +S: Maintained +F: drivers/media/platform/ti-vpe/ + TI CDCE706 CLOCK DRIVER M: Max Filippov <jcmvbkbc(a)gmail.com> S: Maintained @@@ -11334,7 -11319,7 +11335,7 @@@ F: Documentation/usb/ehci.tx F: drivers/usb/host/ehci* USB GADGET/PERIPHERAL SUBSYSTEM -M: Felipe Balbi <balbi(a)ti.com> +M: Felipe Balbi <balbi(a)kernel.org> L: linux-usb(a)vger.kernel.org W: http://www.linux-usb.org/gadget T: git git://git.kernel.org/pub/scm/linux/kernel/git/balbi/usb.git @@@ -11410,7 -11395,7 +11411,7 @@@ S: Maintaine F: drivers/net/usb/pegasus.* USB PHY LAYER -M: Felipe Balbi <balbi(a)ti.com> +M: Felipe Balbi <balbi(a)kernel.org> L: linux-usb(a)vger.kernel.org T: git git://git.kernel.org/pub/scm/linux/kernel/git/balbi/usb.git S: Maintained @@@ -12040,6 -12025,7 +12041,6 @@@ F: arch/arm64/xen F: arch/arm64/include/asm/xen/ XEN NETWORK BACKEND DRIVER -M: Ian Campbell <ian.campbell(a)citrix.com> M: Wei Liu <wei.liu2(a)citrix.com> L: xen-devel(a)lists.xenproject.org (moderated for non-subscribers) L: netdev(a)vger.kernel.org @@@ -12148,7 -12134,7 +12149,7 @@@ F: drivers/net/hamradio/*scc. F: drivers/net/hamradio/z8530.h ZBUD COMPRESSED PAGE ALLOCATOR -M: Seth Jennings <sjennings(a)variantweb.net> +M: Seth Jennings <sjenning(a)redhat.com> L: linux-mm(a)kvack.org S: Maintained F: mm/zbud.c @@@ -12203,7 -12189,7 +12204,7 @@@ F: include/linux/zsmalloc. F: Documentation/vm/zsmalloc.txt ZSWAP COMPRESSED SWAP CACHING -M: Seth Jennings <sjennings(a)variantweb.net> +M: Seth Jennings <sjenning(a)redhat.com> L: linux-mm(a)kvack.org S: Maintained F: mm/zswap.c diff --combined drivers/net/bonding/bond_main.c index b7f1a99,45bdd87..a6527d5 --- a/drivers/net/bonding/bond_main.c +++ b/drivers/net/bonding/bond_main.c @@@ -214,8 -214,6 +214,8 @@@ static void bond_uninit(struct net_devi static struct rtnl_link_stats64 *bond_get_stats(struct net_device *bond_dev, struct rtnl_link_stats64 *stats); static void bond_slave_arr_handler(struct work_struct *work); +static bool bond_time_in_interval(struct bonding *bond, unsigned long last_act, + int mod); /*---------------------------- General routines -----------------------------*/ @@@ -620,8 -618,8 +620,8 @@@ static void bond_hw_addr_swap(struct bo static void bond_set_dev_addr(struct net_device *bond_dev, struct net_device *slave_dev) { - netdev_dbg(bond_dev, "bond_dev=%p slave_dev=%p slave_dev->addr_len=%d\n", - bond_dev, slave_dev, slave_dev->addr_len); + netdev_dbg(bond_dev, "bond_dev=%p slave_dev=%p slave_dev->name=%s slave_dev->addr_len=%d\n", + bond_dev, slave_dev, slave_dev->name, slave_dev->addr_len); memcpy(bond_dev->dev_addr, slave_dev->dev_addr, slave_dev->addr_len); bond_dev->addr_assign_type = NET_ADDR_STOLEN; call_netdevice_notifiers(NETDEV_CHANGEADDR, bond_dev); @@@ -930,11 -928,10 +930,10 @@@ void bond_select_active_slave(struct bo if (!rv) return; - if (netif_carrier_ok(bond->dev)) { + if (netif_carrier_ok(bond->dev)) netdev_info(bond->dev, "first active interface up!\n"); - } else { + else netdev_info(bond->dev, "now running without any active interface!\n"); - } } } @@@ -1180,9 -1177,8 +1179,8 @@@ static rx_handler_result_t bond_handle_ } } - if (bond_should_deliver_exact_match(skb, slave, bond)) { + if (bond_should_deliver_exact_match(skb, slave, bond)) return RX_HANDLER_EXACT; - } skb->dev = bond->dev; @@@ -1243,7 -1239,7 +1241,7 @@@ static struct slave *bond_alloc_slave(s { struct slave *slave = NULL; - slave = kzalloc(sizeof(struct slave), GFP_KERNEL); + slave = kzalloc(sizeof(*slave), GFP_KERNEL); if (!slave) return NULL; @@@ -1383,8 -1379,7 +1381,7 @@@ int bond_enslave(struct net_device *bon if (slave_dev->flags & IFF_UP) { netdev_err(bond_dev, "%s is up - this may be due to an out of date ifenslave\n", slave_dev->name); - res = -EPERM; - goto err_undo_flags; + return -EPERM; } /* set bonding device ether type by slave - bonding netdevices are @@@ -1404,8 -1399,7 +1401,7 @@@ res = notifier_to_errno(res); if (res) { netdev_err(bond_dev, "refused to change device type\n"); - res = -EBUSY; - goto err_undo_flags; + return -EBUSY; } /* Flush unicast and multicast addresses */ @@@ -1425,8 -1419,7 +1421,7 @@@ } else if (bond_dev->type != slave_dev->type) { netdev_err(bond_dev, "%s ether type (%d) is different from other slaves (%d), can not enslave it\n", slave_dev->name, slave_dev->type, bond_dev->type); - res = -EINVAL; - goto err_undo_flags; + return -EINVAL; } if (slave_ops->ndo_set_mac_address == NULL) { @@@ -2129,7 -2122,6 +2124,7 @@@ static void bond_miimon_commit(struct b continue; case BOND_LINK_UP: + bond_update_speed_duplex(slave); bond_set_slave_link_state(slave, BOND_LINK_UP, BOND_SLAVE_NOTIFY_NOW); slave->last_link_up = jiffies; @@@ -2462,7 -2454,7 +2457,7 @@@ int bond_arp_rcv(const struct sk_buff * struct slave *slave) { struct arphdr *arp = (struct arphdr *)skb->data; - struct slave *curr_active_slave; + struct slave *curr_active_slave, *curr_arp_slave; unsigned char *arp_ptr; __be32 sip, tip; int alen, is_arp = skb->protocol == __cpu_to_be16(ETH_P_ARP); @@@ -2509,41 -2501,26 +2504,41 @@@ &sip, &tip); curr_active_slave = rcu_dereference(bond->curr_active_slave); + curr_arp_slave = rcu_dereference(bond->current_arp_slave); - /* Backup slaves won't see the ARP reply, but do come through - * here for each ARP probe (so we swap the sip/tip to validate - * the probe). In a "redundant switch, common router" type of - * configuration, the ARP probe will (hopefully) travel from - * the active, through one switch, the router, then the other - * switch before reaching the backup. + /* We 'trust' the received ARP enough to validate it if: * - * We 'trust' the arp requests if there is an active slave and - * it received valid arp reply(s) after it became active. This - * is done to avoid endless looping when we can't reach the + * (a) the slave receiving the ARP is active (which includes the + * current ARP slave, if any), or + * + * (b) the receiving slave isn't active, but there is a currently + * active slave and it received valid arp reply(s) after it became + * the currently active slave, or + * + * (c) there is an ARP slave that sent an ARP during the prior ARP + * interval, and we receive an ARP reply on any slave. We accept + * these because switch FDB update delays may deliver the ARP + * reply to a slave other than the sender of the ARP request. + * + * Note: for (b), backup slaves are receiving the broadcast ARP + * request, not a reply. This request passes from the sending + * slave through the L2 switch(es) to the receiving slave. Since + * this is checking the request, sip/tip are swapped for + * validation. + * + * This is done to avoid endless looping when we can't reach the * arp_ip_target and fool ourselves with our own arp requests. */ - if (bond_is_active_slave(slave)) bond_validate_arp(bond, slave, sip, tip); else if (curr_active_slave && time_after(slave_last_rx(bond, curr_active_slave), curr_active_slave->last_link_up)) bond_validate_arp(bond, slave, tip, sip); + else if (curr_arp_slave && (arp->ar_op == htons(ARPOP_REPLY)) && + bond_time_in_interval(bond, + dev_trans_start(curr_arp_slave->dev), 1)) + bond_validate_arp(bond, slave, sip, tip); out_unlock: if (arp != (struct arphdr *)skb->data) @@@ -3327,6 -3304,7 +3322,7 @@@ static struct rtnl_link_stats64 *bond_g stats->rx_bytes += sstats->rx_bytes - pstats->rx_bytes; stats->rx_errors += sstats->rx_errors - pstats->rx_errors; stats->rx_dropped += sstats->rx_dropped - pstats->rx_dropped; + stats->rx_nohandler += sstats->rx_nohandler - pstats->rx_nohandler; stats->tx_packets += sstats->tx_packets - pstats->tx_packets;; stats->tx_bytes += sstats->tx_bytes - pstats->tx_bytes; diff --combined drivers/net/ethernet/broadcom/bnxt/bnxt.c index 8ab000d,169920a..ff1507f --- a/drivers/net/ethernet/broadcom/bnxt/bnxt.c +++ b/drivers/net/ethernet/broadcom/bnxt/bnxt.c @@@ -69,7 -69,7 +69,7 @@@ MODULE_VERSION(DRV_MODULE_VERSION) #define BNXT_RX_DMA_OFFSET NET_SKB_PAD #define BNXT_RX_COPY_THRESH 256 -#define BNXT_TX_PUSH_THRESH 92 +#define BNXT_TX_PUSH_THRESH 164 enum board_idx { BCM57301, @@@ -223,12 -223,11 +223,12 @@@ static netdev_tx_t bnxt_start_xmit(stru } if (free_size == bp->tx_ring_size && length <= bp->tx_push_thresh) { - struct tx_push_bd *push = txr->tx_push; - struct tx_bd *tx_push = &push->txbd1; - struct tx_bd_ext *tx_push1 = &push->txbd2; - void *pdata = tx_push1 + 1; - int j; + struct tx_push_buffer *tx_push_buf = txr->tx_push; + struct tx_push_bd *tx_push = &tx_push_buf->push_bd; + struct tx_bd_ext *tx_push1 = &tx_push->txbd2; + void *pdata = tx_push_buf->data; + u64 *end; + int j, push_len; /* Set COAL_NOW to be ready quickly for the next push */ tx_push->tx_bd_len_flags_type = @@@ -248,9 -247,6 +248,9 @@@ tx_push1->tx_bd_cfa_meta = cpu_to_le32(vlan_tag_flags); tx_push1->tx_bd_cfa_action = cpu_to_le32(cfa_action); + end = PTR_ALIGN(pdata + length + 1, 8) - 1; + *end = 0; + skb_copy_from_linear_data(skb, pdata, len); pdata += len; for (j = 0; j < last_frag; j++) { @@@ -265,29 -261,22 +265,29 @@@ pdata += skb_frag_size(frag); } - memcpy(txbd, tx_push, sizeof(*txbd)); + txbd->tx_bd_len_flags_type = tx_push->tx_bd_len_flags_type; + txbd->tx_bd_haddr = txr->data_mapping; prod = NEXT_TX(prod); txbd = &txr->tx_desc_ring[TX_RING(prod)][TX_IDX(prod)]; memcpy(txbd, tx_push1, sizeof(*txbd)); prod = NEXT_TX(prod); - push->doorbell = + tx_push->doorbell = cpu_to_le32(DB_KEY_TX_PUSH | DB_LONG_TX_PUSH | prod); txr->tx_prod = prod; netdev_tx_sent_queue(txq, skb->len); - __iowrite64_copy(txr->tx_doorbell, push, - (length + sizeof(*push) + 8) / 8); + push_len = (length + sizeof(*tx_push) + 7) / 8; + if (push_len > 16) { + __iowrite64_copy(txr->tx_doorbell, tx_push_buf, 16); + __iowrite64_copy(txr->tx_doorbell + 4, tx_push_buf + 1, + push_len - 16); + } else { + __iowrite64_copy(txr->tx_doorbell, tx_push_buf, + push_len); + } tx_buf->is_push = 1; - goto tx_done; } @@@ -1764,7 -1753,7 +1764,7 @@@ static int bnxt_alloc_tx_rings(struct b push_size = L1_CACHE_ALIGN(sizeof(struct tx_push_bd) + bp->tx_push_thresh); - if (push_size > 128) { + if (push_size > 256) { push_size = 0; bp->tx_push_thresh = 0; } @@@ -1783,6 -1772,7 +1783,6 @@@ return rc; if (bp->tx_push_size) { - struct tx_bd *txbd; dma_addr_t mapping; /* One pre-allocated DMA buffer to backup @@@ -1796,11 -1786,13 +1796,11 @@@ if (!txr->tx_push) return -ENOMEM; - txbd = &txr->tx_push->txbd1; - mapping = txr->tx_push_mapping + sizeof(struct tx_push_bd); - txbd->tx_bd_haddr = cpu_to_le64(mapping); + txr->data_mapping = cpu_to_le64(mapping); - memset(txbd + 1, 0, sizeof(struct tx_bd_ext)); + memset(txr->tx_push, 0, sizeof(struct tx_push_bd)); } ring->queue_id = bp->q_info[j].queue_id; if (i % bp->tx_nr_rings_per_tc == (bp->tx_nr_rings_per_tc - 1)) @@@ -4554,18 -4546,20 +4554,18 @@@ static int bnxt_update_phy_setting(stru if (!(link_info->autoneg & BNXT_AUTONEG_FLOW_CTRL) && link_info->force_pause_setting != link_info->req_flow_ctrl) update_pause = true; - if (link_info->req_duplex != link_info->duplex_setting) - update_link = true; if (!(link_info->autoneg & BNXT_AUTONEG_SPEED)) { if (BNXT_AUTO_MODE(link_info->auto_mode)) update_link = true; if (link_info->req_link_speed != link_info->force_link_speed) update_link = true; + if (link_info->req_duplex != link_info->duplex_setting) + update_link = true; } else { if (link_info->auto_mode == BNXT_LINK_AUTO_NONE) update_link = true; if (link_info->advertising != link_info->auto_link_speeds) update_link = true; - if (link_info->req_link_speed != link_info->auto_link_speed) - update_link = true; } if (update_link) @@@ -4642,7 -4636,7 +4642,7 @@@ static int __bnxt_open_nic(struct bnxt if (link_re_init) { rc = bnxt_update_phy_setting(bp); if (rc) - goto open_err; + netdev_warn(bp->dev, "failed to update phy settings\n"); } if (irq_re_init) { @@@ -4660,7 -4654,6 +4660,7 @@@ /* Enable TX queues */ bnxt_tx_enable(bp); mod_timer(&bp->timer, jiffies + bp->current_interval); + bnxt_update_link(bp, true); return 0; @@@ -5377,9 -5370,16 +5377,16 @@@ static int bnxt_change_mtu(struct net_d return 0; } - static int bnxt_setup_tc(struct net_device *dev, u8 tc) + static int bnxt_setup_tc(struct net_device *dev, u32 handle, __be16 proto, + struct tc_to_netdev *ntc) { struct bnxt *bp = netdev_priv(dev); + u8 tc; + + if (handle != TC_H_ROOT || ntc->type != TC_SETUP_MQPRIO) + return -EINVAL; + + tc = ntc->tc; if (tc > bp->max_tc) { netdev_err(dev, "too many traffic classes requested: %d Max supported is %d\n", @@@ -5677,16 -5677,22 +5684,16 @@@ static int bnxt_probe_phy(struct bnxt * } /*initialize the ethool setting copy with NVM settings */ - if (BNXT_AUTO_MODE(link_info->auto_mode)) - link_info->autoneg |= BNXT_AUTONEG_SPEED; - - if (link_info->auto_pause_setting & BNXT_LINK_PAUSE_BOTH) { - if (link_info->auto_pause_setting == BNXT_LINK_PAUSE_BOTH) - link_info->autoneg |= BNXT_AUTONEG_FLOW_CTRL; + if (BNXT_AUTO_MODE(link_info->auto_mode)) { + link_info->autoneg = BNXT_AUTONEG_SPEED | + BNXT_AUTONEG_FLOW_CTRL; + link_info->advertising = link_info->auto_link_speeds; link_info->req_flow_ctrl = link_info->auto_pause_setting; - } else if (link_info->force_pause_setting & BNXT_LINK_PAUSE_BOTH) { + } else { + link_info->req_link_speed = link_info->force_link_speed; + link_info->req_duplex = link_info->duplex_setting; link_info->req_flow_ctrl = link_info->force_pause_setting; } - link_info->req_duplex = link_info->duplex_setting; - if (link_info->autoneg & BNXT_AUTONEG_SPEED) - link_info->req_link_speed = link_info->auto_link_speed; - else - link_info->req_link_speed = link_info->force_link_speed; - link_info->advertising = link_info->auto_link_speeds; snprintf(phy_ver, PHY_VER_STR_LEN, " ph %d.%d.%d", link_info->phy_ver[0], link_info->phy_ver[1], diff --combined drivers/net/ethernet/cavium/thunder/nicvf_main.c index a009bc3,90ce93e..bfee298 --- a/drivers/net/ethernet/cavium/thunder/nicvf_main.c +++ b/drivers/net/ethernet/cavium/thunder/nicvf_main.c @@@ -574,7 -574,8 +574,7 @@@ static inline void nicvf_set_rxhash(str static void nicvf_rcv_pkt_handler(struct net_device *netdev, struct napi_struct *napi, - struct cmp_queue *cq, - struct cqe_rx_t *cqe_rx, int cqe_type) + struct cqe_rx_t *cqe_rx) { struct sk_buff *skb; struct nicvf *nic = netdev_priv(netdev); @@@ -590,7 -591,7 +590,7 @@@ } /* Check for errors */ - err = nicvf_check_cqe_rx_errs(nic, cq, cqe_rx); + err = nicvf_check_cqe_rx_errs(nic, cqe_rx); if (err && !cqe_rx->rb_cnt) return; @@@ -681,7 -682,8 +681,7 @@@ loop cq_idx, cq_desc->cqe_type); switch (cq_desc->cqe_type) { case CQE_TYPE_RX: - nicvf_rcv_pkt_handler(netdev, napi, cq, - cq_desc, CQE_TYPE_RX); + nicvf_rcv_pkt_handler(netdev, napi, cq_desc); work_done++; break; case CQE_TYPE_SEND: @@@ -826,7 -828,7 +826,7 @@@ static irqreturn_t nicvf_intr_handler(i nicvf_disable_intr(nic, NICVF_INTR_CQ, qidx); /* Schedule NAPI */ - napi_schedule(&cq_poll->napi); + napi_schedule_irqoff(&cq_poll->napi); /* Clear interrupt */ nicvf_clear_intr(nic, NICVF_INTR_CQ, qidx); @@@ -897,6 -899,31 +897,31 @@@ static void nicvf_disable_msix(struct n } } + static void nicvf_set_irq_affinity(struct nicvf *nic) + { + int vec, cpu; + int irqnum; + + for (vec = 0; vec < nic->num_vec; vec++) { + if (!nic->irq_allocated[vec]) + continue; + + if (!zalloc_cpumask_var(&nic->affinity_mask[vec], GFP_KERNEL)) + return; + /* CQ interrupts */ + if (vec < NICVF_INTR_ID_SQ) + /* Leave CPU0 for RBDR and other interrupts */ + cpu = nicvf_netdev_qidx(nic, vec) + 1; + else + cpu = 0; + + cpumask_set_cpu(cpumask_local_spread(cpu, nic->node), + nic->affinity_mask[vec]); + irqnum = nic->msix_entries[vec].vector; + irq_set_affinity_hint(irqnum, nic->affinity_mask[vec]); + } + } + static int nicvf_register_interrupts(struct nicvf *nic) { int irq, ret = 0; @@@ -942,8 -969,13 +967,13 @@@ ret = request_irq(nic->msix_entries[irq].vector, nicvf_qs_err_intr_handler, 0, nic->irq_name[irq], nic); - if (!ret) - nic->irq_allocated[irq] = true; + if (ret) + goto err; + + nic->irq_allocated[irq] = true; + + /* Set IRQ affinities */ + nicvf_set_irq_affinity(nic); err: if (ret) @@@ -961,6 -993,9 +991,9 @@@ static void nicvf_unregister_interrupts if (!nic->irq_allocated[irq]) continue; + irq_set_affinity_hint(nic->msix_entries[irq].vector, NULL); + free_cpumask_var(nic->affinity_mask[irq]); + if (irq < NICVF_INTR_ID_SQ) free_irq(nic->msix_entries[irq].vector, nic->napi[irq]); else @@@ -1123,6 -1158,7 +1156,6 @@@ int nicvf_stop(struct net_device *netde /* Clear multiqset info */ nic->pnicvf = nic; - nic->sqs_count = 0; return 0; } @@@ -1351,9 -1387,6 +1384,9 @@@ void nicvf_update_stats(struct nicvf *n drv_stats->tx_frames_ok = stats->tx_ucast_frames_ok + stats->tx_bcast_frames_ok + stats->tx_mcast_frames_ok; + drv_stats->rx_frames_ok = stats->rx_ucast_frames + + stats->rx_bcast_frames + + stats->rx_mcast_frames; drv_stats->rx_drops = stats->rx_drop_red + stats->rx_drop_overrun; drv_stats->tx_drops = stats->tx_drops; @@@ -1394,6 -1427,7 +1427,7 @@@ static void nicvf_tx_timeout(struct net netdev_warn(dev, "%s: Transmit timed out, resetting\n", dev->name); + nic->drv_stats.tx_timeout++; schedule_work(&nic->reset_task); } @@@ -1538,9 -1572,6 +1572,9 @@@ static int nicvf_probe(struct pci_dev * nicvf_send_vf_struct(nic); + if (!pass1_silicon(nic->pdev)) + nic->hw_tso = true; + /* Check if this VF is in QS only mode */ if (nic->sqs_mode) return 0; @@@ -1560,6 -1591,9 +1594,6 @@@ netdev->vlan_features = NETIF_F_SG | NETIF_F_IP_CSUM | NETIF_F_TSO; - if (!pass1_silicon(nic->pdev)) - nic->hw_tso = true; - netdev->netdev_ops = &nicvf_netdev_ops; netdev->watchdog_timeo = NICVF_TX_TIMEOUT; diff --combined drivers/net/ethernet/cavium/thunder/nicvf_queues.c index 767347b,5adb208..0dd1abf --- a/drivers/net/ethernet/cavium/thunder/nicvf_queues.c +++ b/drivers/net/ethernet/cavium/thunder/nicvf_queues.c @@@ -78,7 -78,7 +78,7 @@@ static void nicvf_free_q_desc_mem(struc static inline int nicvf_alloc_rcv_buffer(struct nicvf *nic, gfp_t gfp, u32 buf_len, u64 **rbuf) { - int order = get_order(buf_len); + int order = (PAGE_SIZE <= 4096) ? PAGE_ALLOC_COSTLY_ORDER : 0; /* Check if request can be accomodated in previous allocated page */ if (nic->rb_page) { @@@ -96,8 -96,7 +96,7 @@@ nic->rb_page = alloc_pages(gfp | __GFP_COMP | __GFP_NOWARN, order); if (!nic->rb_page) { - netdev_err(nic->netdev, - "Failed to allocate new rcv buffer\n"); + nic->drv_stats.rcv_buffer_alloc_failures++; return -ENOMEM; } nic->rb_page_offset = 0; @@@ -1329,12 -1328,16 +1328,12 @@@ void nicvf_update_sq_stats(struct nicv } /* Check for errors in the receive cmp.queue entry */ -int nicvf_check_cqe_rx_errs(struct nicvf *nic, - struct cmp_queue *cq, struct cqe_rx_t *cqe_rx) +int nicvf_check_cqe_rx_errs(struct nicvf *nic, struct cqe_rx_t *cqe_rx) { struct nicvf_hw_stats *stats = &nic->hw_stats; - struct nicvf_drv_stats *drv_stats = &nic->drv_stats; - if (!cqe_rx->err_level && !cqe_rx->err_opcode) { - drv_stats->rx_frames_ok++; + if (!cqe_rx->err_level && !cqe_rx->err_opcode) return 0; - } if (netif_msg_rx_err(nic)) netdev_err(nic->netdev, diff --combined drivers/net/ethernet/mellanox/mlx4/en_netdev.c index f191a16,01d6a96..96d95cb --- a/drivers/net/ethernet/mellanox/mlx4/en_netdev.c +++ b/drivers/net/ethernet/mellanox/mlx4/en_netdev.c @@@ -69,6 -69,15 +69,15 @@@ int mlx4_en_setup_tc(struct net_device return 0; } + static int __mlx4_en_setup_tc(struct net_device *dev, u32 handle, __be16 proto, + struct tc_to_netdev *tc) + { + if (handle != TC_H_ROOT || tc->type != TC_SETUP_MQPRIO) + return -EINVAL; + + return mlx4_en_setup_tc(dev, tc->tc); + } + #ifdef CONFIG_RFS_ACCEL struct mlx4_en_filter { @@@ -2344,6 -2353,8 +2353,6 @@@ out /* set offloads */ priv->dev->hw_enc_features |= NETIF_F_IP_CSUM | NETIF_F_RXCSUM | NETIF_F_TSO | NETIF_F_GSO_UDP_TUNNEL; - priv->dev->hw_features |= NETIF_F_GSO_UDP_TUNNEL; - priv->dev->features |= NETIF_F_GSO_UDP_TUNNEL; } static void mlx4_en_del_vxlan_offloads(struct work_struct *work) @@@ -2354,6 -2365,8 +2363,6 @@@ /* unset offloads */ priv->dev->hw_enc_features &= ~(NETIF_F_IP_CSUM | NETIF_F_RXCSUM | NETIF_F_TSO | NETIF_F_GSO_UDP_TUNNEL); - priv->dev->hw_features &= ~NETIF_F_GSO_UDP_TUNNEL; - priv->dev->features &= ~NETIF_F_GSO_UDP_TUNNEL; ret = mlx4_SET_PORT_VXLAN(priv->mdev->dev, priv->port, VXLAN_STEER_BY_OUTER_MAC, 0); @@@ -2462,7 -2475,7 +2471,7 @@@ static const struct net_device_ops mlx4 #endif .ndo_set_features = mlx4_en_set_features, .ndo_fix_features = mlx4_en_fix_features, - .ndo_setup_tc = mlx4_en_setup_tc, + .ndo_setup_tc = __mlx4_en_setup_tc, #ifdef CONFIG_RFS_ACCEL .ndo_rx_flow_steer = mlx4_en_filter_rfs, #endif @@@ -2500,7 -2513,7 +2509,7 @@@ static const struct net_device_ops mlx4 #endif .ndo_set_features = mlx4_en_set_features, .ndo_fix_features = mlx4_en_fix_features, - .ndo_setup_tc = mlx4_en_setup_tc, + .ndo_setup_tc = __mlx4_en_setup_tc, #ifdef CONFIG_RFS_ACCEL .ndo_rx_flow_steer = mlx4_en_filter_rfs, #endif @@@ -2976,11 -2989,6 +2985,11 @@@ int mlx4_en_init_netdev(struct mlx4_en_ priv->rss_hash_fn = ETH_RSS_HASH_TOP; } + if (mdev->dev->caps.tunnel_offload_mode == MLX4_TUNNEL_OFFLOAD_MODE_VXLAN) { + dev->hw_features |= NETIF_F_GSO_UDP_TUNNEL; + dev->features |= NETIF_F_GSO_UDP_TUNNEL; + } + mdev->pndev[port] = dev; mdev->upper[port] = NULL; diff --combined drivers/net/ethernet/renesas/ravb_main.c index 744d780,331c596..88656ce --- a/drivers/net/ethernet/renesas/ravb_main.c +++ b/drivers/net/ethernet/renesas/ravb_main.c @@@ -2,7 -2,7 +2,7 @@@ * * Copyright (C) 2014-2015 Renesas Electronics Corporation * Copyright (C) 2015 Renesas Solutions Corp. - * Copyright (C) 2015 Cogent Embedded, Inc. <source(a)cogentembedded.com> + * Copyright (C) 2015-2016 Cogent Embedded, Inc. <source(a)cogentembedded.com> * * Based on the SuperH Ethernet driver * @@@ -42,6 -42,12 +42,12 @@@ NETIF_MSG_RX_ERR | \ NETIF_MSG_TX_ERR) + void ravb_modify(struct net_device *ndev, enum ravb_reg reg, u32 clear, + u32 set) + { + ravb_write(ndev, (ravb_read(ndev, reg) & ~clear) | set, reg); + } + int ravb_wait(struct net_device *ndev, enum ravb_reg reg, u32 mask, u32 value) { int i; @@@ -59,8 -65,7 +65,7 @@@ static int ravb_config(struct net_devic int error; /* Set config mode */ - ravb_write(ndev, (ravb_read(ndev, CCC) & ~CCC_OPC) | CCC_OPC_CONFIG, - CCC); + ravb_modify(ndev, CCC, CCC_OPC, CCC_OPC_CONFIG); /* Check if the operating mode is changed to the config mode */ error = ravb_wait(ndev, CSR, CSR_OPS, CSR_OPS_CONFIG); if (error) @@@ -72,13 -77,8 +77,8 @@@ static void ravb_set_duplex(struct net_device *ndev) { struct ravb_private *priv = netdev_priv(ndev); - u32 ecmr = ravb_read(ndev, ECMR); - if (priv->duplex) /* Full */ - ecmr |= ECMR_DM; - else /* Half */ - ecmr &= ~ECMR_DM; - ravb_write(ndev, ecmr, ECMR); + ravb_modify(ndev, ECMR, ECMR_DM, priv->duplex ? ECMR_DM : 0); } static void ravb_set_rate(struct net_device *ndev) @@@ -92,8 -92,6 +92,6 @@@ case 1000: /* 1000BASE */ ravb_write(ndev, GECMR_SPEED_1000, GECMR); break; - default: - break; } } @@@ -131,13 -129,8 +129,8 @@@ static void ravb_mdio_ctrl(struct mdiob { struct ravb_private *priv = container_of(ctrl, struct ravb_private, mdiobb); - u32 pir = ravb_read(priv->ndev, PIR); - if (set) - pir |= mask; - else - pir &= ~mask; - ravb_write(priv->ndev, pir, PIR); + ravb_modify(priv->ndev, PIR, mask, set ? mask : 0); } /* MDC pin control */ @@@ -393,9 -386,9 +386,9 @@@ static int ravb_dmac_init(struct net_de ravb_ring_format(ndev, RAVB_NC); #if defined(__LITTLE_ENDIAN) - ravb_write(ndev, ravb_read(ndev, CCC) & ~CCC_BOC, CCC); + ravb_modify(ndev, CCC, CCC_BOC, 0); #else - ravb_write(ndev, ravb_read(ndev, CCC) | CCC_BOC, CCC); + ravb_modify(ndev, CCC, CCC_BOC, CCC_BOC); #endif /* Set AVB RX */ @@@ -418,8 -411,7 +411,7 @@@ ravb_write(ndev, TIC_FTE0 | TIC_FTE1 | TIC_TFUE, TIC); /* Setting the control will start the AVB-DMAC process. */ - ravb_write(ndev, (ravb_read(ndev, CCC) & ~CCC_OPC) | CCC_OPC_OPERATION, - CCC); + ravb_modify(ndev, CCC, CCC_OPC, CCC_OPC_OPERATION); return 0; } @@@ -493,7 -485,7 +485,7 @@@ static void ravb_get_tx_tstamp(struct n break; } } - ravb_write(ndev, ravb_read(ndev, TCCR) | TCCR_TFR, TCCR); + ravb_modify(ndev, TCCR, TCCR_TFR, TCCR_TFR); } } @@@ -613,13 -605,13 +605,13 @@@ static bool ravb_rx(struct net_device * static void ravb_rcv_snd_disable(struct net_device *ndev) { /* Disable TX and RX */ - ravb_write(ndev, ravb_read(ndev, ECMR) & ~(ECMR_RE | ECMR_TE), ECMR); + ravb_modify(ndev, ECMR, ECMR_RE | ECMR_TE, 0); } static void ravb_rcv_snd_enable(struct net_device *ndev) { /* Enable TX and RX */ - ravb_write(ndev, ravb_read(ndev, ECMR) | ECMR_RE | ECMR_TE, ECMR); + ravb_modify(ndev, ECMR, ECMR_RE | ECMR_TE, ECMR_RE | ECMR_TE); } /* function for waiting dma process finished */ @@@ -812,8 -804,8 +804,8 @@@ static int ravb_poll(struct napi_struc /* Re-enable RX/TX interrupts */ spin_lock_irqsave(&priv->lock, flags); - ravb_write(ndev, ravb_read(ndev, RIC0) | mask, RIC0); - ravb_write(ndev, ravb_read(ndev, TIC) | mask, TIC); + ravb_modify(ndev, RIC0, mask, mask); + ravb_modify(ndev, TIC, mask, mask); mmiowb(); spin_unlock_irqrestore(&priv->lock, flags); @@@ -852,8 -844,7 +844,7 @@@ static void ravb_adjust_link(struct net ravb_set_rate(ndev); } if (!priv->link) { - ravb_write(ndev, ravb_read(ndev, ECMR) & ~ECMR_TXF, - ECMR); + ravb_modify(ndev, ECMR, ECMR_TXF, 0); new_state = true; priv->link = phydev->link; if (priv->no_avb_link) @@@ -1139,8 -1130,7 +1130,8 @@@ static int ravb_set_ringparam(struct ne if (netif_running(ndev)) { netif_device_detach(ndev); /* Stop PTP Clock driver */ - ravb_ptp_stop(ndev); + if (priv->chip_id == RCAR_GEN2) + ravb_ptp_stop(ndev); /* Wait for DMA stopping */ error = ravb_stop_dma(ndev); if (error) { @@@ -1171,8 -1161,7 +1162,8 @@@ ravb_emac_init(ndev); /* Initialise PTP Clock driver */ - ravb_ptp_init(ndev, priv->pdev); + if (priv->chip_id == RCAR_GEN2) + ravb_ptp_init(ndev, priv->pdev); netif_device_attach(ndev); } @@@ -1300,8 -1289,7 +1291,8 @@@ static void ravb_tx_timeout_work(struc netif_tx_stop_all_queues(ndev); /* Stop PTP Clock driver */ - ravb_ptp_stop(ndev); + if (priv->chip_id == RCAR_GEN2) + ravb_ptp_stop(ndev); /* Wait for DMA stopping */ ravb_stop_dma(ndev); @@@ -1314,8 -1302,7 +1305,8 @@@ ravb_emac_init(ndev); /* Initialise PTP Clock driver */ - ravb_ptp_init(ndev, priv->pdev); + if (priv->chip_id == RCAR_GEN2) + ravb_ptp_init(ndev, priv->pdev); netif_tx_start_all_queues(ndev); } @@@ -1397,7 -1384,7 +1388,7 @@@ static netdev_tx_t ravb_start_xmit(stru desc--; desc->die_dt = DT_FSTART; - ravb_write(ndev, ravb_read(ndev, TCCR) | (TCCR_TSRQ0 << q), TCCR); + ravb_modify(ndev, TCCR, TCCR_TSRQ0 << q, TCCR_TSRQ0 << q); priv->cur_tx[q] += NUM_TX_DESC; if (priv->cur_tx[q] - priv->dirty_tx[q] > @@@ -1472,15 -1459,10 +1463,10 @@@ static void ravb_set_rx_mode(struct net { struct ravb_private *priv = netdev_priv(ndev); unsigned long flags; - u32 ecmr; spin_lock_irqsave(&priv->lock, flags); - ecmr = ravb_read(ndev, ECMR); - if (ndev->flags & IFF_PROMISC) - ecmr |= ECMR_PRM; - else - ecmr &= ~ECMR_PRM; - ravb_write(ndev, ecmr, ECMR); + ravb_modify(ndev, ECMR, ECMR_PRM, + ndev->flags & IFF_PROMISC ? ECMR_PRM : 0); mmiowb(); spin_unlock_irqrestore(&priv->lock, flags); } @@@ -1808,23 -1790,25 +1794,21 @@@ static int ravb_probe(struct platform_d /* Set AVB config mode */ if (chip_id == RCAR_GEN2) { - ravb_write(ndev, (ravb_read(ndev, CCC) & ~CCC_OPC) | - CCC_OPC_CONFIG, CCC); + ravb_modify(ndev, CCC, CCC_OPC, CCC_OPC_CONFIG); /* Set CSEL value */ - ravb_write(ndev, (ravb_read(ndev, CCC) & ~CCC_CSEL) | - CCC_CSEL_HPB, CCC); + ravb_modify(ndev, CCC, CCC_CSEL, CCC_CSEL_HPB); } else { - ravb_write(ndev, (ravb_read(ndev, CCC) & ~CCC_OPC) | - CCC_OPC_CONFIG | CCC_GAC | CCC_CSEL_HPB, CCC); + ravb_modify(ndev, CCC, CCC_OPC, CCC_OPC_CONFIG | + CCC_GAC | CCC_CSEL_HPB); } - /* Set CSEL value */ - ravb_write(ndev, (ravb_read(ndev, CCC) & ~CCC_CSEL) | CCC_CSEL_HPB, - CCC); - /* Set GTI value */ error = ravb_set_gti(ndev); if (error) goto out_release; /* Request GTI loading */ - ravb_write(ndev, ravb_read(ndev, GCCR) | GCCR_LTI, GCCR); + ravb_modify(ndev, GCCR, GCCR_LTI, GCCR_LTI); /* Allocate descriptor base address table */ priv->desc_bat_size = sizeof(struct ravb_desc) * DBAT_ENTRY_NUM; diff --combined drivers/net/ethernet/ti/netcp_core.c index 029841f,06a0a73..ed0c30f --- a/drivers/net/ethernet/ti/netcp_core.c +++ b/drivers/net/ethernet/ti/netcp_core.c @@@ -117,17 -117,21 +117,17 @@@ static void get_pkt_info(dma_addr_t *bu *ndesc = le32_to_cpu(desc->next_desc); } -static void get_pad_info(u32 *pad0, u32 *pad1, u32 *pad2, struct knav_dma_desc *desc) +static u32 get_sw_data(int index, struct knav_dma_desc *desc) { - *pad0 = le32_to_cpu(desc->pad[0]); - *pad1 = le32_to_cpu(desc->pad[1]); - *pad2 = le32_to_cpu(desc->pad[2]); + /* No Endian conversion needed as this data is untouched by hw */ + return desc->sw_data[index]; } -static void get_pad_ptr(void **padptr, struct knav_dma_desc *desc) -{ - u64 pad64; - - pad64 = le32_to_cpu(desc->pad[0]) + - ((u64)le32_to_cpu(desc->pad[1]) << 32); - *padptr = (void *)(uintptr_t)pad64; -} +/* use these macros to get sw data */ +#define GET_SW_DATA0(desc) get_sw_data(0, desc) +#define GET_SW_DATA1(desc) get_sw_data(1, desc) +#define GET_SW_DATA2(desc) get_sw_data(2, desc) +#define GET_SW_DATA3(desc) get_sw_data(3, desc) static void get_org_pkt_info(dma_addr_t *buff, u32 *buff_len, struct knav_dma_desc *desc) @@@ -159,18 -163,13 +159,18 @@@ static void set_desc_info(u32 desc_info desc->packet_info = cpu_to_le32(pkt_info); } -static void set_pad_info(u32 pad0, u32 pad1, u32 pad2, struct knav_dma_desc *desc) +static void set_sw_data(int index, u32 data, struct knav_dma_desc *desc) { - desc->pad[0] = cpu_to_le32(pad0); - desc->pad[1] = cpu_to_le32(pad1); - desc->pad[2] = cpu_to_le32(pad1); + /* No Endian conversion needed as this data is untouched by hw */ + desc->sw_data[index] = data; } +/* use these macros to set sw data */ +#define SET_SW_DATA0(data, desc) set_sw_data(0, data, desc) +#define SET_SW_DATA1(data, desc) set_sw_data(1, data, desc) +#define SET_SW_DATA2(data, desc) set_sw_data(2, data, desc) +#define SET_SW_DATA3(data, desc) set_sw_data(3, data, desc) + static void set_org_pkt_info(dma_addr_t buff, u32 buff_len, struct knav_dma_desc *desc) { @@@ -582,6 -581,7 +582,6 @@@ static void netcp_free_rx_desc_chain(st dma_addr_t dma_desc, dma_buf; unsigned int buf_len, dma_sz = sizeof(*ndesc); void *buf_ptr; - u32 pad[2]; u32 tmp; get_words(&dma_desc, 1, &desc->next_desc); @@@ -593,20 -593,14 +593,20 @@@ break; } get_pkt_info(&dma_buf, &tmp, &dma_desc, ndesc); - get_pad_ptr(&buf_ptr, ndesc); + /* warning!!!! We are retrieving the virtual ptr in the sw_data + * field as a 32bit value. Will not work on 64bit machines + */ + buf_ptr = (void *)GET_SW_DATA0(ndesc); + buf_len = (int)GET_SW_DATA1(desc); dma_unmap_page(netcp->dev, dma_buf, PAGE_SIZE, DMA_FROM_DEVICE); __free_page(buf_ptr); knav_pool_desc_put(netcp->rx_pool, desc); } - - get_pad_info(&pad[0], &pad[1], &buf_len, desc); - buf_ptr = (void *)(uintptr_t)(pad[0] + ((u64)pad[1] << 32)); + /* warning!!!! We are retrieving the virtual ptr in the sw_data + * field as a 32bit value. Will not work on 64bit machines + */ + buf_ptr = (void *)GET_SW_DATA0(desc); + buf_len = (int)GET_SW_DATA1(desc); if (buf_ptr) netcp_frag_free(buf_len <= PAGE_SIZE, buf_ptr); @@@ -645,6 -639,7 +645,6 @@@ static int netcp_process_one_rx_packet( dma_addr_t dma_desc, dma_buff; struct netcp_packet p_info; struct sk_buff *skb; - u32 pad[2]; void *org_buf_ptr; dma_desc = knav_queue_pop(netcp->rx_queue, &dma_sz); @@@ -658,11 -653,8 +658,11 @@@ } get_pkt_info(&dma_buff, &buf_len, &dma_desc, desc); - get_pad_info(&pad[0], &pad[1], &org_buf_len, desc); - org_buf_ptr = (void *)(uintptr_t)(pad[0] + ((u64)pad[1] << 32)); + /* warning!!!! We are retrieving the virtual ptr in the sw_data + * field as a 32bit value. Will not work on 64bit machines + */ + org_buf_ptr = (void *)GET_SW_DATA0(desc); + org_buf_len = (int)GET_SW_DATA1(desc); if (unlikely(!org_buf_ptr)) { dev_err(netcp->ndev_dev, "NULL bufptr in desc\n"); @@@ -687,6 -679,7 +687,6 @@@ /* Fill in the page fragment list */ while (dma_desc) { struct page *page; - void *ptr; ndesc = knav_pool_desc_unmap(netcp->rx_pool, dma_desc, dma_sz); if (unlikely(!ndesc)) { @@@ -695,10 -688,8 +695,10 @@@ } get_pkt_info(&dma_buff, &buf_len, &dma_desc, ndesc); - get_pad_ptr(&ptr, ndesc); - page = ptr; + /* warning!!!! We are retrieving the virtual ptr in the sw_data + * field as a 32bit value. Will not work on 64bit machines + */ + page = (struct page *)GET_SW_DATA0(desc); if (likely(dma_buff && buf_len && page)) { dma_unmap_page(netcp->dev, dma_buff, PAGE_SIZE, @@@ -786,10 -777,7 +786,10 @@@ static void netcp_free_rx_buf(struct ne } get_org_pkt_info(&dma, &buf_len, desc); - get_pad_ptr(&buf_ptr, desc); + /* warning!!!! We are retrieving the virtual ptr in the sw_data + * field as a 32bit value. Will not work on 64bit machines + */ + buf_ptr = (void *)GET_SW_DATA0(desc); if (unlikely(!dma)) { dev_err(netcp->ndev_dev, "NULL orig_buff in desc\n"); @@@ -841,7 -829,7 +841,7 @@@ static int netcp_allocate_rx_buf(struc struct page *page; dma_addr_t dma; void *bufptr; - u32 pad[3]; + u32 sw_data[2]; /* Allocate descriptor */ hwdesc = knav_pool_desc_get(netcp->rx_pool); @@@ -858,7 -846,7 +858,7 @@@ SKB_DATA_ALIGN(sizeof(struct skb_shared_info)); bufptr = netdev_alloc_frag(primary_buf_len); - pad[2] = primary_buf_len; + sw_data[1] = primary_buf_len; if (unlikely(!bufptr)) { dev_warn_ratelimited(netcp->ndev_dev, @@@ -870,10 -858,9 +870,10 @@@ if (unlikely(dma_mapping_error(netcp->dev, dma))) goto fail; - pad[0] = lower_32_bits((uintptr_t)bufptr); - pad[1] = upper_32_bits((uintptr_t)bufptr); - + /* warning!!!! We are saving the virtual ptr in the sw_data + * field as a 32bit value. Will not work on 64bit machines + */ + sw_data[0] = (u32)bufptr; } else { /* Allocate a secondary receive queue entry */ page = alloc_page(GFP_ATOMIC | GFP_DMA | __GFP_COLD); @@@ -883,11 -870,9 +883,11 @@@ } buf_len = PAGE_SIZE; dma = dma_map_page(netcp->dev, page, 0, buf_len, DMA_TO_DEVICE); - pad[0] = lower_32_bits(dma); - pad[1] = upper_32_bits(dma); - pad[2] = 0; + /* warning!!!! We are saving the virtual ptr in the sw_data + * field as a 32bit value. Will not work on 64bit machines + */ + sw_data[0] = (u32)page; + sw_data[1] = 0; } desc_info = KNAV_DMA_DESC_PS_INFO_IN_DESC; @@@ -897,8 -882,7 +897,8 @@@ pkt_info |= (netcp->rx_queue_id & KNAV_DMA_DESC_RETQ_MASK) << KNAV_DMA_DESC_RETQ_SHIFT; set_org_pkt_info(dma, buf_len, hwdesc); - set_pad_info(pad[0], pad[1], pad[2], hwdesc); + SET_SW_DATA0(sw_data[0], hwdesc); + SET_SW_DATA1(sw_data[1], hwdesc); set_desc_info(desc_info, pkt_info, hwdesc); /* Push to FDQs */ @@@ -987,6 -971,7 +987,6 @@@ static int netcp_process_tx_compl_packe unsigned int budget) { struct knav_dma_desc *desc; - void *ptr; struct sk_buff *skb; unsigned int dma_sz; dma_addr_t dma; @@@ -1003,10 -988,8 +1003,10 @@@ continue; } - get_pad_ptr(&ptr, desc); - skb = ptr; + /* warning!!!! We are retrieving the virtual ptr in the sw_data + * field as a 32bit value. Will not work on 64bit machines + */ + skb = (struct sk_buff *)GET_SW_DATA0(desc); netcp_free_tx_desc_chain(netcp, desc, dma_sz); if (!skb) { dev_err(netcp->ndev_dev, "No skb in Tx desc\n"); @@@ -1211,10 -1194,10 +1211,10 @@@ static int netcp_tx_submit_skb(struct n } set_words(&tmp, 1, &desc->packet_info); - tmp = lower_32_bits((uintptr_t)&skb); - set_words(&tmp, 1, &desc->pad[0]); - tmp = upper_32_bits((uintptr_t)&skb); - set_words(&tmp, 1, &desc->pad[1]); + /* warning!!!! We are saving the virtual ptr in the sw_data + * field as a 32bit value. Will not work on 64bit machines + */ + SET_SW_DATA0((u32)skb, desc); if (tx_pipe->flags & SWITCH_TO_PORT_IN_TAGINFO) { tmp = tx_pipe->switch_to_port; @@@ -1852,22 -1835,26 +1852,26 @@@ static u16 netcp_select_queue(struct ne return 0; } - static int netcp_setup_tc(struct net_device *dev, u8 num_tc) + static int netcp_setup_tc(struct net_device *dev, u32 handle, __be16 proto, + struct tc_to_netdev *tc) { int i; /* setup tc must be called under rtnl lock */ ASSERT_RTNL(); + if (handle != TC_H_ROOT || tc->type != TC_SETUP_MQPRIO) + return -EINVAL; + /* Sanity-check the number of traffic classes requested */ if ((dev->real_num_tx_queues <= 1) || - (dev->real_num_tx_queues < num_tc)) + (dev->real_num_tx_queues < tc->tc)) return -EINVAL; /* Configure traffic class to queue mappings */ - if (num_tc) { - netdev_set_num_tc(dev, num_tc); - for (i = 0; i < num_tc; i++) + if (tc->tc) { + netdev_set_num_tc(dev, tc->tc); + for (i = 0; i < tc->tc; i++) netdev_set_tc_queue(dev, i, 1, i); } else { netdev_reset_tc(dev); diff --combined drivers/net/geneve.c index 0bf7edd,8fa8388..bc5da35 --- a/drivers/net/geneve.c +++ b/drivers/net/geneve.c @@@ -72,10 -72,11 +72,11 @@@ struct geneve_dev bool collect_md; struct gro_cells gro_cells; u32 flags; + struct dst_cache dst_cache; }; /* Geneve device flags */ - #define GENEVE_F_UDP_CSUM BIT(0) + #define GENEVE_F_UDP_ZERO_CSUM_TX BIT(0) #define GENEVE_F_UDP_ZERO_CSUM6_TX BIT(1) #define GENEVE_F_UDP_ZERO_CSUM6_RX BIT(2) @@@ -109,6 -110,11 +110,11 @@@ static __be64 vni_to_tunnel_id(const __ #endif } + static sa_family_t geneve_get_sk_family(struct geneve_sock *gs) + { + return gs->sock->sk->sk_family; + } + static struct geneve_dev *geneve_lookup(struct geneve_sock *gs, __be32 addr, u8 vni[]) { @@@ -152,58 -158,60 +158,60 @@@ static inline struct genevehdr *geneve_ return (struct genevehdr *)(udp_hdr(skb) + 1); } - /* geneve receive/decap routine */ - static void geneve_rx(struct geneve_sock *gs, struct sk_buff *skb) + static struct geneve_dev *geneve_lookup_skb(struct geneve_sock *gs, + struct sk_buff *skb) { - struct genevehdr *gnvh = geneve_hdr(skb); - struct metadata_dst *tun_dst = NULL; - struct geneve_dev *geneve = NULL; - struct pcpu_sw_netstats *stats; - struct iphdr *iph = NULL; + u8 *vni; __be32 addr; static u8 zero_vni[3]; - u8 *vni; - int err = 0; - sa_family_t sa_family; #if IS_ENABLED(CONFIG_IPV6) - struct ipv6hdr *ip6h = NULL; - struct in6_addr addr6; static struct in6_addr zero_addr6; #endif - sa_family = gs->sock->sk->sk_family; + if (geneve_get_sk_family(gs) == AF_INET) { + struct iphdr *iph; - if (sa_family == AF_INET) { iph = ip_hdr(skb); /* outer IP header... */ if (gs->collect_md) { vni = zero_vni; addr = 0; } else { - vni = gnvh->vni; - + vni = geneve_hdr(skb)->vni; addr = iph->saddr; } - geneve = geneve_lookup(gs, addr, vni); + return geneve_lookup(gs, addr, vni); #if IS_ENABLED(CONFIG_IPV6) - } else if (sa_family == AF_INET6) { + } else if (geneve_get_sk_family(gs) == AF_INET6) { + struct ipv6hdr *ip6h; + struct in6_addr addr6; + ip6h = ipv6_hdr(skb); /* outer IPv6 header... */ if (gs->collect_md) { vni = zero_vni; addr6 = zero_addr6; } else { - vni = gnvh->vni; - + vni = geneve_hdr(skb)->vni; addr6 = ip6h->saddr; } - geneve = geneve6_lookup(gs, addr6, vni); + return geneve6_lookup(gs, addr6, vni); #endif } - if (!geneve) - goto drop; + return NULL; + } + + /* geneve receive/decap routine */ + static void geneve_rx(struct geneve_dev *geneve, struct geneve_sock *gs, + struct sk_buff *skb) + { + struct genevehdr *gnvh = geneve_hdr(skb); + struct metadata_dst *tun_dst = NULL; + struct pcpu_sw_netstats *stats; + int err = 0; + void *oiph; if (ip_tunnel_collect_metadata() || gs->collect_md) { __be16 flags; @@@ -212,7 -220,7 +220,7 @@@ (gnvh->oam ? TUNNEL_OAM : 0) | (gnvh->critical ? TUNNEL_CRIT_OPT : 0); - tun_dst = udp_tun_rx_dst(skb, sa_family, flags, + tun_dst = udp_tun_rx_dst(skb, geneve_get_sk_family(gs), flags, vni_to_tunnel_id(gnvh->vni), gnvh->opt_len * 4); if (!tun_dst) @@@ -229,7 -237,6 +237,6 @@@ } skb_reset_mac_header(skb); - skb_scrub_packet(skb, !net_eq(geneve->net, dev_net(geneve->dev))); skb->protocol = eth_type_trans(skb, geneve->dev); skb_postpull_rcsum(skb, eth_hdr(skb), ETH_HLEN); @@@ -240,25 -247,27 +247,27 @@@ if (ether_addr_equal(eth_hdr(skb)->h_source, geneve->dev->dev_addr)) goto drop; + oiph = skb_network_header(skb); skb_reset_network_header(skb); - if (iph) - err = IP_ECN_decapsulate(iph, skb); + if (geneve_get_sk_family(gs) == AF_INET) + err = IP_ECN_decapsulate(oiph, skb); #if IS_ENABLED(CONFIG_IPV6) - if (ip6h) - err = IP6_ECN_decapsulate(ip6h, skb); + else + err = IP6_ECN_decapsulate(oiph, skb); #endif if (unlikely(err)) { if (log_ecn_error) { - if (iph) + if (geneve_get_sk_family(gs) == AF_INET) net_info_ratelimited("non-ECT from %pI4 " "with TOS=%#x\n", - &iph->saddr, iph->tos); + &((struct iphdr *)oiph)->saddr, + ((struct iphdr *)oiph)->tos); #if IS_ENABLED(CONFIG_IPV6) - if (ip6h) + else net_info_ratelimited("non-ECT from %pI6\n", - &ip6h->saddr); + &((struct ipv6hdr *)oiph)->saddr); #endif } if (err > 1) { @@@ -297,6 -306,13 +306,13 @@@ static int geneve_init(struct net_devic return err; } + err = dst_cache_init(&geneve->dst_cache, GFP_KERNEL); + if (err) { + free_percpu(dev->tstats); + gro_cells_destroy(&geneve->gro_cells); + return err; + } + return 0; } @@@ -304,6 -320,7 +320,7 @@@ static void geneve_uninit(struct net_de { struct geneve_dev *geneve = netdev_priv(dev); + dst_cache_destroy(&geneve->dst_cache); gro_cells_destroy(&geneve->gro_cells); free_percpu(dev->tstats); } @@@ -312,6 -329,7 +329,7 @@@ static int geneve_udp_encap_recv(struct sock *sk, struct sk_buff *skb) { struct genevehdr *geneveh; + struct geneve_dev *geneve; struct geneve_sock *gs; int opts_len; @@@ -327,16 -345,21 +345,21 @@@ if (unlikely(geneveh->proto_type != htons(ETH_P_TEB))) goto error; - opts_len = geneveh->opt_len * 4; - if (iptunnel_pull_header(skb, GENEVE_BASE_HLEN + opts_len, - htons(ETH_P_TEB))) - goto drop; - gs = rcu_dereference_sk_user_data(sk); if (!gs) goto drop; - geneve_rx(gs, skb); + geneve = geneve_lookup_skb(gs, skb); + if (!geneve) + goto drop; + + opts_len = geneveh->opt_len * 4; + if (iptunnel_pull_header(skb, GENEVE_BASE_HLEN + opts_len, + htons(ETH_P_TEB), + !net_eq(geneve->net, dev_net(geneve->dev)))) + goto drop; + + geneve_rx(geneve, gs, skb); return 0; drop: @@@ -383,7 -406,7 +406,7 @@@ static void geneve_notify_add_rx_port(s struct net_device *dev; struct sock *sk = gs->sock->sk; struct net *net = sock_net(sk); - sa_family_t sa_family = sk->sk_family; + sa_family_t sa_family = geneve_get_sk_family(gs); __be16 port = inet_sk(sk)->inet_sport; int err; @@@ -544,7 -567,7 +567,7 @@@ static void geneve_notify_del_rx_port(s struct net_device *dev; struct sock *sk = gs->sock->sk; struct net *net = sock_net(sk); - sa_family_t sa_family = sk->sk_family; + sa_family_t sa_family = geneve_get_sk_family(gs); __be16 port = inet_sk(sk)->inet_sport; rcu_read_lock(); @@@ -587,7 -610,7 +610,7 @@@ static struct geneve_sock *geneve_find_ list_for_each_entry(gs, &gn->sock_list, list) { if (inet_sk(gs->sock->sk)->inet_sport == dst_port && - inet_sk(gs->sock->sk)->sk.sk_family == family) { + geneve_get_sk_family(gs) == family) { return gs; } } @@@ -680,7 -703,7 +703,7 @@@ static int geneve_build_skb(struct rtab struct genevehdr *gnvh; int min_headroom; int err; - bool udp_sum = !!(flags & GENEVE_F_UDP_CSUM); + bool udp_sum = !(flags & GENEVE_F_UDP_ZERO_CSUM_TX); skb_scrub_packet(skb, xnet); @@@ -753,7 -776,9 +776,9 @@@ static struct rtable *geneve_get_v4_rt( struct ip_tunnel_info *info) { struct geneve_dev *geneve = netdev_priv(dev); + struct dst_cache *dst_cache; struct rtable *rt = NULL; + bool use_cache = true; __u8 tos; memset(fl4, 0, sizeof(*fl4)); @@@ -764,16 -789,26 +789,26 @@@ fl4->daddr = info->key.u.ipv4.dst; fl4->saddr = info->key.u.ipv4.src; fl4->flowi4_tos = RT_TOS(info->key.tos); + dst_cache = &info->dst_cache; } else { tos = geneve->tos; if (tos == 1) { const struct iphdr *iip = ip_hdr(skb); tos = ip_tunnel_get_dsfield(iip, skb); + use_cache = false; } fl4->flowi4_tos = RT_TOS(tos); fl4->daddr = geneve->remote.sin.sin_addr.s_addr; + dst_cache = &geneve->dst_cache; + } + + use_cache = use_cache && !skb->mark; + if (use_cache) { + rt = dst_cache_get_ip4(dst_cache, &fl4->saddr); + if (rt) + return rt; } rt = ip_route_output_key(geneve->net, fl4); @@@ -786,6 -821,8 +821,8 @@@ ip_rt_put(rt); return ERR_PTR(-ELOOP); } + if (use_cache) + dst_cache_set_ip4(dst_cache, &rt->dst, fl4->saddr); return rt; } @@@ -798,6 -835,8 +835,8 @@@ static struct dst_entry *geneve_get_v6_ struct geneve_dev *geneve = netdev_priv(dev); struct geneve_sock *gs6 = geneve->sock6; struct dst_entry *dst = NULL; + struct dst_cache *dst_cache; + bool use_cache = true; __u8 prio; memset(fl6, 0, sizeof(*fl6)); @@@ -808,16 -847,26 +847,26 @@@ fl6->daddr = info->key.u.ipv6.dst; fl6->saddr = info->key.u.ipv6.src; fl6->flowi6_tos = RT_TOS(info->key.tos); + dst_cache = &info->dst_cache; } else { prio = geneve->tos; if (prio == 1) { const struct iphdr *iip = ip_hdr(skb); prio = ip_tunnel_get_dsfield(iip, skb); + use_cache = false; } fl6->flowi6_tos = RT_TOS(prio); fl6->daddr = geneve->remote.sin6.sin6_addr; + dst_cache = &geneve->dst_cache; + } + + use_cache = use_cache && !skb->mark; + if (use_cache) { + dst = dst_cache_get_ip6(dst_cache, &fl6->saddr); + if (dst) + return dst; } if (ipv6_stub->ipv6_dst_lookup(geneve->net, gs6->sock->sk, &dst, fl6)) { @@@ -830,6 -879,8 +879,8 @@@ return ERR_PTR(-ELOOP); } + if (use_cache) + dst_cache_set_ip6(dst_cache, dst, &fl6->saddr); return dst; } #endif @@@ -893,9 -944,9 +944,9 @@@ static netdev_tx_t geneve_xmit_skb(stru opts = ip_tunnel_info_opts(info); if (key->tun_flags & TUNNEL_CSUM) - flags |= GENEVE_F_UDP_CSUM; + flags &= ~GENEVE_F_UDP_ZERO_CSUM_TX; else - flags &= ~GENEVE_F_UDP_CSUM; + flags |= GENEVE_F_UDP_ZERO_CSUM_TX; err = geneve_build_skb(rt, skb, key->tun_flags, vni, info->options_len, opts, flags, xnet); @@@ -921,7 -972,7 +972,7 @@@ udp_tunnel_xmit_skb(rt, gs4->sock->sk, skb, fl4.saddr, fl4.daddr, tos, ttl, df, sport, geneve->dst_port, !net_eq(geneve->net, dev_net(geneve->dev)), - !(flags & GENEVE_F_UDP_CSUM)); + !!(flags & GENEVE_F_UDP_ZERO_CSUM_TX)); return NETDEV_TX_OK; @@@ -1039,34 -1090,6 +1090,34 @@@ static netdev_tx_t geneve_xmit(struct s return geneve_xmit_skb(skb, dev, info); } +static int __geneve_change_mtu(struct net_device *dev, int new_mtu, bool strict) +{ + /* The max_mtu calculation does not take account of GENEVE + * options, to avoid excluding potentially valid + * configurations. + */ + int max_mtu = IP_MAX_MTU - GENEVE_BASE_HLEN - sizeof(struct iphdr) + - dev->hard_header_len; + + if (new_mtu < 68) + return -EINVAL; + + if (new_mtu > max_mtu) { + if (strict) + return -EINVAL; + + new_mtu = max_mtu; + } + + dev->mtu = new_mtu; + return 0; +} + +static int geneve_change_mtu(struct net_device *dev, int new_mtu) +{ + return __geneve_change_mtu(dev, new_mtu, true); +} + static int geneve_fill_metadata_dst(struct net_device *dev, struct sk_buff *skb) { struct ip_tunnel_info *info = skb_tunnel_info(skb); @@@ -1111,7 -1134,7 +1162,7 @@@ static const struct net_device_ops gene .ndo_stop = geneve_stop, .ndo_start_xmit = geneve_xmit, .ndo_get_stats64 = ip_tunnel_get_stats64, - .ndo_change_mtu = eth_change_mtu, + .ndo_change_mtu = geneve_change_mtu, .ndo_validate_addr = eth_validate_addr, .ndo_set_mac_address = eth_mac_addr, .ndo_fill_metadata_dst = geneve_fill_metadata_dst, @@@ -1178,7 -1201,6 +1229,7 @@@ static void geneve_setup(struct net_dev dev->hw_features |= NETIF_F_GSO_SOFTWARE; netif_keep_dst(dev); + dev->priv_flags &= ~IFF_TX_SKB_SHARING; dev->priv_flags |= IFF_LIVE_ADDR_CHANGE | IFF_NO_QUEUE; eth_hw_addr_random(dev); } @@@ -1301,6 -1323,8 +1352,8 @@@ static int geneve_configure(struct net return -EPERM; } + dst_cache_reset(&geneve->dst_cache); + err = register_netdevice(dev); if (err) return err; @@@ -1359,8 -1383,8 +1412,8 @@@ static int geneve_newlink(struct net *n metadata = true; if (data[IFLA_GENEVE_UDP_CSUM] && - nla_get_u8(data[IFLA_GENEVE_UDP_CSUM])) - flags |= GENEVE_F_UDP_CSUM; + !nla_get_u8(data[IFLA_GENEVE_UDP_CSUM])) + flags |= GENEVE_F_UDP_ZERO_CSUM_TX; if (data[IFLA_GENEVE_UDP_ZERO_CSUM6_TX] && nla_get_u8(data[IFLA_GENEVE_UDP_ZERO_CSUM6_TX])) @@@ -1430,7 -1454,7 +1483,7 @@@ static int geneve_fill_info(struct sk_b } if (nla_put_u8(skb, IFLA_GENEVE_UDP_CSUM, - !!(geneve->flags & GENEVE_F_UDP_CSUM)) || + !(geneve->flags & GENEVE_F_UDP_ZERO_CSUM_TX)) || nla_put_u8(skb, IFLA_GENEVE_UDP_ZERO_CSUM6_TX, !!(geneve->flags & GENEVE_F_UDP_ZERO_CSUM6_TX)) || nla_put_u8(skb, IFLA_GENEVE_UDP_ZERO_CSUM6_RX, @@@ -1470,23 -1494,12 +1523,23 @@@ struct net_device *geneve_dev_create_fb return dev; err = geneve_configure(net, dev, &geneve_remote_unspec, - 0, 0, 0, htons(dst_port), true, 0); - if (err) { - free_netdev(dev); - return ERR_PTR(err); - } + 0, 0, 0, htons(dst_port), true, + GENEVE_F_UDP_ZERO_CSUM6_RX); + if (err) + goto err; + + /* openvswitch users expect packet sizes to be unrestricted, + * so set the largest MTU we can. + */ + err = __geneve_change_mtu(dev, IP_MAX_MTU, false); + if (err) + goto err; + return dev; + + err: + free_netdev(dev); + return ERR_PTR(err); } EXPORT_SYMBOL_GPL(geneve_dev_create_fb); diff --combined drivers/net/hyperv/netvsc_drv.c index 98e34fe,202e2b1..2b6595e --- a/drivers/net/hyperv/netvsc_drv.c +++ b/drivers/net/hyperv/netvsc_drv.c @@@ -43,6 -43,11 +43,11 @@@ #define RING_SIZE_MIN 64 #define LINKCHANGE_INT (2 * HZ) + #define NETVSC_HW_FEATURES (NETIF_F_RXCSUM | \ + NETIF_F_SG | \ + NETIF_F_TSO | \ + NETIF_F_TSO6 | \ + NETIF_F_HW_CSUM) static int ring_size = 128; module_param(ring_size, int, S_IRUGO); MODULE_PARM_DESC(ring_size, "Ring buffer size (# of pages)"); @@@ -545,6 -550,8 +550,8 @@@ do_send packet->page_buf_cnt = init_page_array(rndis_msg, rndis_msg_size, skb, packet, &pb); + /* timestamp packet in software */ + skb_tx_timestamp(skb); ret = netvsc_send(net_device_ctx->device_ctx, packet, rndis_msg, &pb, skb); @@@ -915,6 -922,7 +922,7 @@@ static const struct ethtool_ops ethtool .get_link = ethtool_op_get_link, .get_channels = netvsc_get_channels, .set_channels = netvsc_set_channels, + .get_ts_info = ethtool_op_get_ts_info, }; static const struct net_device_ops device_ops = { @@@ -1081,17 -1089,12 +1089,15 @@@ static int netvsc_probe(struct hv_devic net->netdev_ops = &device_ops; - net->hw_features = NETIF_F_RXCSUM | NETIF_F_SG | NETIF_F_IP_CSUM | - NETIF_F_TSO; - net->features = NETIF_F_HW_VLAN_CTAG_TX | NETIF_F_SG | NETIF_F_RXCSUM | - NETIF_F_IP_CSUM | NETIF_F_TSO; + net->hw_features = NETVSC_HW_FEATURES; + net->features = NETVSC_HW_FEATURES | NETIF_F_HW_VLAN_CTAG_TX; net->ethtool_ops = &ethtool_ops; SET_NETDEV_DEV(net, &dev->device); + /* We always need headroom for rndis header */ + net->needed_headroom = RNDIS_AND_PPI_SIZE; + /* Notify the netvsc driver of the new device */ memset(&device_info, 0, sizeof(device_info)); device_info.ring_size = ring_size; diff --combined drivers/net/phy/bcm7xxx.c index db507e3,9b31104..b881a7b1 --- a/drivers/net/phy/bcm7xxx.c +++ b/drivers/net/phy/bcm7xxx.c @@@ -24,7 -24,7 +24,7 @@@ #define MII_BCM7XXX_100TX_FALSE_CAR 0x13 #define MII_BCM7XXX_100TX_DISC 0x14 #define MII_BCM7XXX_AUX_MODE 0x1d - #define MII_BCM7XX_64CLK_MDIO BIT(12) + #define MII_BCM7XXX_64CLK_MDIO BIT(12) #define MII_BCM7XXX_TEST 0x1f #define MII_BCM7XXX_SHD_MODE_2 BIT(2) @@@ -247,9 -247,13 +247,9 @@@ static int bcm7xxx_config_init(struct p int ret; /* Enable 64 clock MDIO */ - phy_write(phydev, MII_BCM7XXX_AUX_MODE, MII_BCM7XX_64CLK_MDIO); + phy_write(phydev, MII_BCM7XXX_AUX_MODE, MII_BCM7XXX_64CLK_MDIO); phy_read(phydev, MII_BCM7XXX_AUX_MODE); - /* Workaround only required for 100Mbits/sec capable PHYs */ - if (phydev->supported & PHY_GBIT_FEATURES) - return 0; - /* set shadow mode 2 */ ret = phy_set_clr_bits(phydev, MII_BCM7XXX_TEST, MII_BCM7XXX_SHD_MODE_2, MII_BCM7XXX_SHD_MODE_2); @@@ -266,7 -270,7 +266,7 @@@ phy_write(phydev, MII_BCM7XXX_100TX_FALSE_CAR, 0x7555); /* reset shadow mode 2 */ - ret = phy_set_clr_bits(phydev, MII_BCM7XXX_TEST, MII_BCM7XXX_SHD_MODE_2, 0); + ret = phy_set_clr_bits(phydev, MII_BCM7XXX_TEST, 0, MII_BCM7XXX_SHD_MODE_2); if (ret < 0) return ret; @@@ -303,6 -307,11 +303,6 @@@ static int bcm7xxx_suspend(struct phy_d return 0; } -static int bcm7xxx_dummy_config_init(struct phy_device *phydev) -{ - return 0; -} - #define BCM7XXX_28NM_GPHY(_oui, _name) \ { \ .phy_id = (_oui), \ @@@ -317,6 -326,21 +317,21 @@@ .resume = bcm7xxx_28nm_resume, \ } + #define BCM7XXX_40NM_EPHY(_oui, _name) \ + { \ + .phy_id = (_oui), \ + .phy_id_mask = 0xfffffff0, \ + .name = _name, \ + .features = PHY_BASIC_FEATURES | \ + SUPPORTED_Pause | SUPPORTED_Asym_Pause, \ + .flags = PHY_IS_INTERNAL, \ + .config_init = bcm7xxx_config_init, \ + .config_aneg = genphy_config_aneg, \ + .read_status = genphy_read_status, \ + .suspend = bcm7xxx_suspend, \ + .resume = bcm7xxx_config_init, \ + } + static struct phy_driver bcm7xxx_driver[] = { BCM7XXX_28NM_GPHY(PHY_ID_BCM7250, "Broadcom BCM7250"), BCM7XXX_28NM_GPHY(PHY_ID_BCM7364, "Broadcom BCM7364"), @@@ -324,43 -348,34 +339,10 @@@ BCM7XXX_28NM_GPHY(PHY_ID_BCM7439, "Broadcom BCM7439"), BCM7XXX_28NM_GPHY(PHY_ID_BCM7439_2, "Broadcom BCM7439 (2)"), BCM7XXX_28NM_GPHY(PHY_ID_BCM7445, "Broadcom BCM7445"), - { - .phy_id = PHY_ID_BCM7425, - .phy_id_mask = 0xfffffff0, - .name = "Broadcom BCM7425", - .features = PHY_BASIC_FEATURES | - SUPPORTED_Pause | SUPPORTED_Asym_Pause, - .flags = PHY_IS_INTERNAL, - .config_init = bcm7xxx_config_init, - .config_aneg = genphy_config_aneg, - .read_status = genphy_read_status, - .suspend = bcm7xxx_suspend, - .resume = bcm7xxx_config_init, - }, { - .phy_id = PHY_ID_BCM7429, - .phy_id_mask = 0xfffffff0, - .name = "Broadcom BCM7429", - .features = PHY_BASIC_FEATURES | - SUPPORTED_Pause | SUPPORTED_Asym_Pause, - .flags = PHY_IS_INTERNAL, - .config_init = bcm7xxx_config_init, - .config_aneg = genphy_config_aneg, - .read_status = genphy_read_status, - .suspend = bcm7xxx_suspend, - .resume = bcm7xxx_config_init, - }, { - .phy_id = PHY_ID_BCM7435, - .phy_id_mask = 0xfffffff0, - .name = "Broadcom BCM7435", - .features = PHY_BASIC_FEATURES | - SUPPORTED_Pause | SUPPORTED_Asym_Pause, - .flags = PHY_IS_INTERNAL, - .config_init = bcm7xxx_config_init, - .config_aneg = genphy_config_aneg, - .read_status = genphy_read_status, - .suspend = bcm7xxx_suspend, - .resume = bcm7xxx_config_init, - } }; + BCM7XXX_40NM_EPHY(PHY_ID_BCM7425, "Broadcom BCM7425"), + BCM7XXX_40NM_EPHY(PHY_ID_BCM7429, "Broadcom BCM7429"), + BCM7XXX_40NM_EPHY(PHY_ID_BCM7435, "Broadcom BCM7435"), -{ - .phy_id = PHY_BCM_OUI_4, - .phy_id_mask = 0xffff0000, - .name = "Broadcom BCM7XXX 40nm", - .features = PHY_GBIT_FEATURES | - SUPPORTED_Pause | SUPPORTED_Asym_Pause, - .flags = PHY_IS_INTERNAL, - .config_init = bcm7xxx_config_init, - .config_aneg = genphy_config_aneg, - .read_status = genphy_read_status, - .suspend = bcm7xxx_suspend, - .resume = bcm7xxx_config_init, -}, { - .phy_id = PHY_BCM_OUI_5, - .phy_id_mask = 0xffffff00, - .name = "Broadcom BCM7XXX 65nm", - .features = PHY_BASIC_FEATURES | - SUPPORTED_Pause | SUPPORTED_Asym_Pause, - .flags = PHY_IS_INTERNAL, - .config_init = bcm7xxx_dummy_config_init, - .config_aneg = genphy_config_aneg, - .read_status = genphy_read_status, - .suspend = bcm7xxx_suspend, - .resume = bcm7xxx_config_init, -} }; ++}; static struct mdio_device_id __maybe_unused bcm7xxx_tbl[] = { { PHY_ID_BCM7250, 0xfffffff0, }, @@@ -371,6 -386,8 +353,6 @@@ { PHY_ID_BCM7439, 0xfffffff0, }, { PHY_ID_BCM7435, 0xfffffff0, }, { PHY_ID_BCM7445, 0xfffffff0, }, - { PHY_BCM_OUI_4, 0xffff0000 }, - { PHY_BCM_OUI_5, 0xffffff00 }, { } }; diff --combined drivers/net/phy/marvell.c index ab1d0fc,d0168f1..9fb9d80 --- a/drivers/net/phy/marvell.c +++ b/drivers/net/phy/marvell.c @@@ -133,6 -133,11 +133,11 @@@ #define MII_88E3016_DISABLE_SCRAMBLER 0x0200 #define MII_88E3016_AUTO_MDIX_CROSSOVER 0x0030 + #define MII_88E1510_GEN_CTRL_REG_1 0x14 + #define MII_88E1510_GEN_CTRL_REG_1_MODE_MASK 0x7 + #define MII_88E1510_GEN_CTRL_REG_1_MODE_SGMII 0x1 /* SGMII to copper */ + #define MII_88E1510_GEN_CTRL_REG_1_RESET 0x8000 /* Soft reset */ + MODULE_DESCRIPTION("Marvell PHY driver"); MODULE_AUTHOR("Andy Fleming"); MODULE_LICENSE("GPL"); @@@ -438,6 -443,41 +443,41 @@@ static int m88e1318_config_aneg(struct return m88e1121_config_aneg(phydev); } + static int m88e1510_config_init(struct phy_device *phydev) + { + int err; + int temp; + + /* SGMII-to-Copper mode initialization */ + if (phydev->interface == PHY_INTERFACE_MODE_SGMII) { + /* Select page 18 */ + err = phy_write(phydev, MII_MARVELL_PHY_PAGE, 18); + if (err < 0) + return err; + + /* In reg 20, write MODE[2:0] = 0x1 (SGMII to Copper) */ + temp = phy_read(phydev, MII_88E1510_GEN_CTRL_REG_1); + temp &= ~MII_88E1510_GEN_CTRL_REG_1_MODE_MASK; + temp |= MII_88E1510_GEN_CTRL_REG_1_MODE_SGMII; + err = phy_write(phydev, MII_88E1510_GEN_CTRL_REG_1, temp); + if (err < 0) + return err; + + /* PHY reset is necessary after changing MODE[2:0] */ + temp |= MII_88E1510_GEN_CTRL_REG_1_RESET; + err = phy_write(phydev, MII_88E1510_GEN_CTRL_REG_1, temp); + if (err < 0) + return err; + + /* Reset page selection */ + err = phy_write(phydev, MII_MARVELL_PHY_PAGE, 0); + if (err < 0) + return err; + } + + return 0; + } + static int m88e1510_config_aneg(struct phy_device *phydev) { int err; @@@ -446,12 -486,6 +486,12 @@@ if (err < 0) return err; + return 0; +} + +static int marvell_config_init(struct phy_device *phydev) +{ + /* Set registers from marvell,reg-init DT property */ return marvell_of_reg_init(phydev); } @@@ -501,7 -535,7 +541,7 @@@ static int m88e1116r_config_init(struc mdelay(500); - return 0; + return marvell_config_init(phydev); } static int m88e3016_config_init(struct phy_device *phydev) @@@ -520,7 -554,7 +560,7 @@@ if (reg < 0) return reg; - return 0; + return marvell_config_init(phydev); } static int m88e1111_config_init(struct phy_device *phydev) @@@ -1031,8 -1065,8 +1071,8 @@@ static u64 marvell_get_stat(struct phy_ { struct marvell_hw_stat stat = marvell_hw_stats[i]; struct marvell_priv *priv = phydev->priv; - int err, oldpage; - u64 val; + int err, oldpage, val; + u64 ret; oldpage = phy_read(phydev, MII_MARVELL_PHY_PAGE); err = phy_write(phydev, MII_MARVELL_PHY_PAGE, @@@ -1042,16 -1076,16 +1082,16 @@@ val = phy_read(phydev, stat.reg); if (val < 0) { - val = UINT64_MAX; + ret = UINT64_MAX; } else { val = val & ((1 << stat.bits) - 1); priv->stats[i] += val; - val = priv->stats[i]; + ret = priv->stats[i]; } phy_write(phydev, MII_MARVELL_PHY_PAGE, oldpage); - return val; + return ret; } static void marvell_get_stats(struct phy_device *phydev, @@@ -1084,7 -1118,6 +1124,7 @@@ static struct phy_driver marvell_driver .features = PHY_GBIT_FEATURES, .probe = marvell_probe, .flags = PHY_HAS_INTERRUPT, + .config_init = &marvell_config_init, .config_aneg = &marvell_config_aneg, .read_status = &genphy_read_status, .ack_interrupt = &marvell_ack_interrupt, @@@ -1156,7 -1189,6 +1196,7 @@@ .features = PHY_GBIT_FEATURES, .flags = PHY_HAS_INTERRUPT, .probe = marvell_probe, + .config_init = &marvell_config_init, .config_aneg = &m88e1121_config_aneg, .read_status = &marvell_read_status, .ack_interrupt = &marvell_ack_interrupt, @@@ -1175,7 -1207,6 +1215,7 @@@ .features = PHY_GBIT_FEATURES, .flags = PHY_HAS_INTERRUPT, .probe = marvell_probe, + .config_init = &marvell_config_init, .config_aneg = &m88e1318_config_aneg, .read_status = &marvell_read_status, .ack_interrupt = &marvell_ack_interrupt, @@@ -1268,7 -1299,7 +1308,7 @@@ .features = PHY_GBIT_FEATURES, .flags = PHY_HAS_INTERRUPT, .probe = marvell_probe, - .config_init = &marvell_config_init, + .config_init = &m88e1510_config_init, .config_aneg = &m88e1510_config_aneg, .read_status = &marvell_read_status, .ack_interrupt = &marvell_ack_interrupt, @@@ -1287,7 -1318,6 +1327,7 @@@ .features = PHY_GBIT_FEATURES, .flags = PHY_HAS_INTERRUPT, .probe = marvell_probe, + .config_init = &marvell_config_init, .config_aneg = &m88e1510_config_aneg, .read_status = &marvell_read_status, .ack_interrupt = &marvell_ack_interrupt, diff --combined drivers/net/vxlan.c index e6944b2,2ddc642..909f793 --- a/drivers/net/vxlan.c +++ b/drivers/net/vxlan.c @@@ -42,7 -42,7 +42,7 @@@ #include <net/netns/generic.h> #include <net/vxlan.h> #include <net/protocol.h> - #include <net/udp_tunnel.h> + #if IS_ENABLED(CONFIG_IPV6) #include <net/ipv6.h> #include <net/addrconf.h> @@@ -197,9 -197,9 +197,9 @@@ static int vxlan_nla_put_addr(struct sk #endif /* Virtual Network hash table head */ - static inline struct hlist_head *vni_head(struct vxlan_sock *vs, u32 id) + static inline struct hlist_head *vni_head(struct vxlan_sock *vs, __be32 vni) { - return &vs->vni_list[hash_32(id, VNI_HASH_BITS)]; + return &vs->vni_list[hash_32((__force u32)vni, VNI_HASH_BITS)]; } /* Socket hash table head */ @@@ -242,12 -242,16 +242,16 @@@ static struct vxlan_sock *vxlan_find_so return NULL; } - static struct vxlan_dev *vxlan_vs_find_vni(struct vxlan_sock *vs, u32 id) + static struct vxlan_dev *vxlan_vs_find_vni(struct vxlan_sock *vs, __be32 vni) { struct vxlan_dev *vxlan; - hlist_for_each_entry_rcu(vxlan, vni_head(vs, id), hlist) { - if (vxlan->default_dst.remote_vni == id) + /* For flow based devices, map all packets to VNI 0 */ + if (vs->flags & VXLAN_F_COLLECT_METADATA) + vni = 0; + + hlist_for_each_entry_rcu(vxlan, vni_head(vs, vni), hlist) { + if (vxlan->default_dst.remote_vni == vni) return vxlan; } @@@ -255,7 -259,7 +259,7 @@@ } /* Look up VNI in a per net namespace table */ - static struct vxlan_dev *vxlan_find_vni(struct net *net, u32 id, + static struct vxlan_dev *vxlan_find_vni(struct net *net, __be32 vni, sa_family_t family, __be16 port, u32 flags) { @@@ -265,7 -269,7 +269,7 @@@ if (!vs) return NULL; - return vxlan_vs_find_vni(vs, id); + return vxlan_vs_find_vni(vs, vni); } /* Fill in neighbour message in skbuff. */ @@@ -315,7 -319,7 +319,7 @@@ static int vxlan_fdb_info(struct sk_buf nla_put_be16(skb, NDA_PORT, rdst->remote_port)) goto nla_put_failure; if (rdst->remote_vni != vxlan->default_dst.remote_vni && - nla_put_u32(skb, NDA_VNI, rdst->remote_vni)) + nla_put_u32(skb, NDA_VNI, be32_to_cpu(rdst->remote_vni))) goto nla_put_failure; if (rdst->remote_ifindex && nla_put_u32(skb, NDA_IFINDEX, rdst->remote_ifindex)) @@@ -383,7 -387,7 +387,7 @@@ static void vxlan_ip_miss(struct net_de }; struct vxlan_rdst remote = { .remote_ip = *ipa, /* goes to NDA_DST */ - .remote_vni = VXLAN_N_VID, + .remote_vni = cpu_to_be32(VXLAN_N_VID), }; vxlan_fdb_notify(vxlan, &f, &remote, RTM_GETNEIGH); @@@ -452,7 -456,7 +456,7 @@@ static struct vxlan_fdb *vxlan_find_mac /* caller should hold vxlan->hash_lock */ static struct vxlan_rdst *vxlan_fdb_find_rdst(struct vxlan_fdb *f, union vxlan_addr *ip, __be16 port, - __u32 vni, __u32 ifindex) + __be32 vni, __u32 ifindex) { struct vxlan_rdst *rd; @@@ -469,7 -473,8 +473,8 @@@ /* Replace destination of unicast mac */ static int vxlan_fdb_replace(struct vxlan_fdb *f, - union vxlan_addr *ip, __be16 port, __u32 vni, __u32 ifindex) + union vxlan_addr *ip, __be16 port, __be32 vni, + __u32 ifindex) { struct vxlan_rdst *rd; @@@ -480,6 -485,8 +485,8 @@@ rd = list_first_entry_or_null(&f->remotes, struct vxlan_rdst, list); if (!rd) return 0; + + dst_cache_reset(&rd->dst_cache); rd->remote_ip = *ip; rd->remote_port = port; rd->remote_vni = vni; @@@ -489,7 -496,7 +496,7 @@@ /* Add/update destinations for multicast */ static int vxlan_fdb_append(struct vxlan_fdb *f, - union vxlan_addr *ip, __be16 port, __u32 vni, + union vxlan_addr *ip, __be16 port, __be32 vni, __u32 ifindex, struct vxlan_rdst **rdp) { struct vxlan_rdst *rd; @@@ -501,6 -508,12 +508,12 @@@ rd = kmalloc(sizeof(*rd), GFP_ATOMIC); if (rd == NULL) return -ENOBUFS; + + if (dst_cache_init(&rd->dst_cache, GFP_ATOMIC)) { + kfree(rd); + return -ENOBUFS; + } + rd->remote_ip = *ip; rd->remote_port = port; rd->remote_vni = vni; @@@ -515,7 -528,8 +528,8 @@@ static struct vxlanhdr *vxlan_gro_remcsum(struct sk_buff *skb, unsigned int off, struct vxlanhdr *vh, size_t hdrlen, - u32 data, struct gro_remcsum *grc, + __be32 vni_field, + struct gro_remcsum *grc, bool nopartial) { size_t start, offset; @@@ -526,10 -540,8 +540,8 @@@ if (!NAPI_GRO_CB(skb)->csum_valid) return NULL; - start = (data & VXLAN_RCO_MASK) << VXLAN_RCO_SHIFT; - offset = start + ((data & VXLAN_RCO_UDP) ? - offsetof(struct udphdr, check) : - offsetof(struct tcphdr, check)); + start = vxlan_rco_start(vni_field); + offset = start + vxlan_rco_offset(vni_field); vh = skb_gro_remcsum_process(skb, (void *)vh, off, hdrlen, start, offset, grc, nopartial); @@@ -549,7 -561,7 +561,7 @@@ static struct sk_buff **vxlan_gro_recei int flush = 1; struct vxlan_sock *vs = container_of(uoff, struct vxlan_sock, udp_offloads); - u32 flags; + __be32 flags; struct gro_remcsum grc; skb_gro_remcsum_init(&grc); @@@ -565,11 -577,11 +577,11 @@@ skb_gro_postpull_rcsum(skb, vh, sizeof(struct vxlanhdr)); - flags = ntohl(vh->vx_flags); + flags = vh->vx_flags; if ((flags & VXLAN_HF_RCO) && (vs->flags & VXLAN_F_REMCSUM_RX)) { vh = vxlan_gro_remcsum(skb, off_vx, vh, sizeof(struct vxlanhdr), - ntohl(vh->vx_vni), &grc, + vh->vx_vni, &grc, !!(vs->flags & VXLAN_F_REMCSUM_NOPARTIAL)); @@@ -660,7 -672,7 +672,7 @@@ static void vxlan_notify_del_rx_port(st static int vxlan_fdb_create(struct vxlan_dev *vxlan, const u8 *mac, union vxlan_addr *ip, __u16 state, __u16 flags, - __be16 port, __u32 vni, __u32 ifindex, + __be16 port, __be32 vni, __u32 ifindex, __u8 ndm_flags) { struct vxlan_rdst *rd = NULL; @@@ -749,8 -761,10 +761,10 @@@ static void vxlan_fdb_free(struct rcu_h struct vxlan_fdb *f = container_of(head, struct vxlan_fdb, rcu); struct vxlan_rdst *rd, *nd; - list_for_each_entry_safe(rd, nd, &f->remotes, list) + list_for_each_entry_safe(rd, nd, &f->remotes, list) { + dst_cache_destroy(&rd->dst_cache); kfree(rd); + } kfree(f); } @@@ -767,7 -781,8 +781,8 @@@ static void vxlan_fdb_destroy(struct vx } static int vxlan_fdb_parse(struct nlattr *tb[], struct vxlan_dev *vxlan, - union vxlan_addr *ip, __be16 *port, u32 *vni, u32 *ifindex) + union vxlan_addr *ip, __be16 *port, __be32 *vni, + u32 *ifindex) { struct net *net = dev_net(vxlan->dev); int err; @@@ -800,7 -815,7 +815,7 @@@ if (tb[NDA_VNI]) { if (nla_len(tb[NDA_VNI]) != sizeof(u32)) return -EINVAL; - *vni = nla_get_u32(tb[NDA_VNI]); + *vni = cpu_to_be32(nla_get_u32(tb[NDA_VNI])); } else { *vni = vxlan->default_dst.remote_vni; } @@@ -830,7 -845,8 +845,8 @@@ static int vxlan_fdb_add(struct ndmsg * /* struct net *net = dev_net(vxlan->dev); */ union vxlan_addr ip; __be16 port; - u32 vni, ifindex; + __be32 vni; + u32 ifindex; int err; if (!(ndm->ndm_state & (NUD_PERMANENT|NUD_REACHABLE))) { @@@ -867,7 -883,8 +883,8 @@@ static int vxlan_fdb_delete(struct ndms struct vxlan_rdst *rd = NULL; union vxlan_addr ip; __be16 port; - u32 vni, ifindex; + __be32 vni; + u32 ifindex; int err; err = vxlan_fdb_parse(tb, vxlan, &ip, &port, &vni, &ifindex); @@@ -1122,54 -1139,65 +1139,65 @@@ static int vxlan_igmp_leave(struct vxla return ret; } - static struct vxlanhdr *vxlan_remcsum(struct sk_buff *skb, struct vxlanhdr *vh, - size_t hdrlen, u32 data, bool nopartial) + static bool vxlan_remcsum(struct vxlanhdr *unparsed, + struct sk_buff *skb, u32 vxflags) { size_t start, offset, plen; - if (skb->remcsum_offload) - return vh; + if (!(unparsed->vx_flags & VXLAN_HF_RCO) || skb->remcsum_offload) + goto out; - start = (data & VXLAN_RCO_MASK) << VXLAN_RCO_SHIFT; - offset = start + ((data & VXLAN_RCO_UDP) ? - offsetof(struct udphdr, check) : - offsetof(struct tcphdr, check)); + start = vxlan_rco_start(unparsed->vx_vni); + offset = start + vxlan_rco_offset(unparsed->vx_vni); - plen = hdrlen + offset + sizeof(u16); + plen = sizeof(struct vxlanhdr) + offset + sizeof(u16); if (!pskb_may_pull(skb, plen)) - return NULL; + return false; - vh = (struct vxlanhdr *)(udp_hdr(skb) + 1); + skb_remcsum_process(skb, (void *)(vxlan_hdr(skb) + 1), start, offset, + !!(vxflags & VXLAN_F_REMCSUM_NOPARTIAL)); + out: + unparsed->vx_flags &= ~VXLAN_HF_RCO; + unparsed->vx_vni &= VXLAN_VNI_MASK; + return true; + } - skb_remcsum_process(skb, (void *)vh + hdrlen, start, offset, - nopartial); + static void vxlan_parse_gbp_hdr(struct vxlanhdr *unparsed, + struct vxlan_metadata *md, + struct metadata_dst *tun_dst) + { + struct vxlanhdr_gbp *gbp = (struct vxlanhdr_gbp *)unparsed; - return vh; + if (!(unparsed->vx_flags & VXLAN_HF_GBP)) + goto out; + + md->gbp = ntohs(gbp->policy_id); + + if (tun_dst) + tun_dst->u.tun_info.key.tun_flags |= TUNNEL_VXLAN_OPT; + + if (gbp->dont_learn) + md->gbp |= VXLAN_GBP_DONT_LEARN; + + if (gbp->policy_applied) + md->gbp |= VXLAN_GBP_POLICY_APPLIED; + + out: + unparsed->vx_flags &= ~VXLAN_GBP_USED_BITS; } - static void vxlan_rcv(struct vxlan_sock *vs, struct sk_buff *skb, - struct vxlan_metadata *md, u32 vni, + static void vxlan_rcv(struct vxlan_dev *vxlan, struct vxlan_sock *vs, + struct sk_buff *skb, struct vxlan_metadata *md, struct metadata_dst *tun_dst) { struct iphdr *oip = NULL; struct ipv6hdr *oip6 = NULL; - struct vxlan_dev *vxlan; struct pcpu_sw_netstats *stats; union vxlan_addr saddr; int err = 0; - /* For flow based devices, map all packets to VNI 0 */ - if (vs->flags & VXLAN_F_COLLECT_METADATA) - vni = 0; - - /* Is this VNI defined? */ - vxlan = vxlan_vs_find_vni(vs, vni); - if (!vxlan) - goto drop; - skb_reset_mac_header(skb); - skb_scrub_packet(skb, !net_eq(vxlan->net, dev_net(vxlan->dev))); skb->protocol = eth_type_trans(skb, vxlan->dev); skb_postpull_rcsum(skb, eth_hdr(skb), ETH_HLEN); @@@ -1246,48 -1274,45 +1274,45 @@@ drop static int vxlan_udp_encap_recv(struct sock *sk, struct sk_buff *skb) { struct metadata_dst *tun_dst = NULL; + struct vxlan_dev *vxlan; struct vxlan_sock *vs; - struct vxlanhdr *vxh; - u32 flags, vni; + struct vxlanhdr unparsed; struct vxlan_metadata _md; struct vxlan_metadata *md = &_md; /* Need Vxlan and inner Ethernet header to be present */ if (!pskb_may_pull(skb, VXLAN_HLEN)) - goto error; - - vxh = (struct vxlanhdr *)(udp_hdr(skb) + 1); - flags = ntohl(vxh->vx_flags); - vni = ntohl(vxh->vx_vni); + return 1; - if (flags & VXLAN_HF_VNI) { - flags &= ~VXLAN_HF_VNI; - } else { - /* VNI flag always required to be set */ - goto bad_flags; + unparsed = *vxlan_hdr(skb); + /* VNI flag always required to be set */ + if (!(unparsed.vx_flags & VXLAN_HF_VNI)) { + netdev_dbg(skb->dev, "invalid vxlan flags=%#x vni=%#x\n", + ntohl(vxlan_hdr(skb)->vx_flags), + ntohl(vxlan_hdr(skb)->vx_vni)); + /* Return non vxlan pkt */ + return 1; } - - if (iptunnel_pull_header(skb, VXLAN_HLEN, htons(ETH_P_TEB))) - goto drop; - vxh = (struct vxlanhdr *)(udp_hdr(skb) + 1); + unparsed.vx_flags &= ~VXLAN_HF_VNI; + unparsed.vx_vni &= ~VXLAN_VNI_MASK; vs = rcu_dereference_sk_user_data(sk); if (!vs) goto drop; - if ((flags & VXLAN_HF_RCO) && (vs->flags & VXLAN_F_REMCSUM_RX)) { - vxh = vxlan_remcsum(skb, vxh, sizeof(struct vxlanhdr), vni, - !!(vs->flags & VXLAN_F_REMCSUM_NOPARTIAL)); - if (!vxh) - goto drop; + vxlan = vxlan_vs_find_vni(vs, vxlan_vni(vxlan_hdr(skb)->vx_vni)); + if (!vxlan) + goto drop; - flags &= ~VXLAN_HF_RCO; - vni &= VXLAN_VNI_MASK; - } + if (iptunnel_pull_header(skb, VXLAN_HLEN, htons(ETH_P_TEB), + !net_eq(vxlan->net, dev_net(vxlan->dev)))) + goto drop; if (vxlan_collect_metadata(vs)) { + __be32 vni = vxlan_vni(vxlan_hdr(skb)->vx_vni); + tun_dst = udp_tun_rx_dst(skb, vxlan_get_sk_family(vs), TUNNEL_KEY, - cpu_to_be64(vni >> 8), sizeof(*md)); + vxlan_vni_to_tun_id(vni), sizeof(*md)); if (!tun_dst) goto drop; @@@ -1300,25 -1325,13 +1325,13 @@@ /* For backwards compatibility, only allow reserved fields to be * used by VXLAN extensions if explicitly requested. */ - if ((flags & VXLAN_HF_GBP) && (vs->flags & VXLAN_F_GBP)) { - struct vxlanhdr_gbp *gbp; - - gbp = (struct vxlanhdr_gbp *)vxh; - md->gbp = ntohs(gbp->policy_id); - - if (tun_dst) - tun_dst->u.tun_info.key.tun_flags |= TUNNEL_VXLAN_OPT; - - if (gbp->dont_learn) - md->gbp |= VXLAN_GBP_DONT_LEARN; - - if (gbp->policy_applied) - md->gbp |= VXLAN_GBP_POLICY_APPLIED; - - flags &= ~VXLAN_GBP_USED_BITS; - } + if (vs->flags & VXLAN_F_REMCSUM_RX) + if (!vxlan_remcsum(&unparsed, skb, vs->flags)) + goto drop; + if (vs->flags & VXLAN_F_GBP) + vxlan_parse_gbp_hdr(&unparsed, md, tun_dst); - if (flags || vni & ~VXLAN_VNI_MASK) { + if (unparsed.vx_flags || unparsed.vx_vni) { /* If there are any unprocessed flags remaining treat * this as a malformed packet. This behavior diverges from * VXLAN RFC (RFC7348) which stipulates that bits in reserved @@@ -1327,28 -1340,19 +1340,19 @@@ * is more robust and provides a little more security in * adding extensions to VXLAN. */ - - goto bad_flags; + goto drop; } - vxlan_rcv(vs, skb, md, vni >> 8, tun_dst); + vxlan_rcv(vxlan, vs, skb, md, tun_dst); return 0; drop: - /* Consume bad packet */ - kfree_skb(skb); - return 0; - - bad_flags: - netdev_dbg(skb->dev, "invalid vxlan flags=%#x vni=%#x\n", - ntohl(vxh->vx_flags), ntohl(vxh->vx_vni)); - - error: if (tun_dst) dst_release((struct dst_entry *)tun_dst); - /* Return non vxlan pkt */ - return 1; + /* Consume bad packet */ + kfree_skb(skb); + return 0; } static int arp_reduce(struct net_device *dev, struct sk_buff *skb) @@@ -1673,7 -1677,7 +1677,7 @@@ static void vxlan_build_gbp_hdr(struct return; gbp = (struct vxlanhdr_gbp *)vxh; - vxh->vx_flags |= htonl(VXLAN_HF_GBP); + vxh->vx_flags |= VXLAN_HF_GBP; if (md->gbp & VXLAN_GBP_DONT_LEARN) gbp->dont_learn = 1; @@@ -1684,20 -1688,15 +1688,15 @@@ gbp->policy_id = htons(md->gbp & VXLAN_GBP_ID_MASK); } - #if IS_ENABLED(CONFIG_IPV6) - static int vxlan6_xmit_skb(struct dst_entry *dst, struct sock *sk, - struct sk_buff *skb, - struct net_device *dev, struct in6_addr *saddr, - struct in6_addr *daddr, __u8 prio, __u8 ttl, - __be16 src_port, __be16 dst_port, __be32 vni, - struct vxlan_metadata *md, bool xnet, u32 vxflags) + static int vxlan_build_skb(struct sk_buff *skb, struct dst_entry *dst, + int iphdr_len, __be32 vni, + struct vxlan_metadata *md, u32 vxflags, + bool udp_sum) { struct vxlanhdr *vxh; int min_headroom; int err; - bool udp_sum = !(vxflags & VXLAN_F_UDP_ZERO_CSUM6_TX); int type = udp_sum ? SKB_GSO_UDP_TUNNEL_CSUM : SKB_GSO_UDP_TUNNEL; - u16 hdrlen = sizeof(struct vxlanhdr); if ((vxflags & VXLAN_F_REMCSUM_TX) && skb->ip_summed == CHECKSUM_PARTIAL) { @@@ -1706,50 -1705,39 +1705,39 @@@ if (csum_start <= VXLAN_MAX_REMCSUM_START && !(csum_start & VXLAN_RCO_SHIFT_MASK) && (skb->csum_offset == offsetof(struct udphdr, check) || - skb->csum_offset == offsetof(struct tcphdr, check))) { - udp_sum = false; + skb->csum_offset == offsetof(struct tcphdr, check))) type |= SKB_GSO_TUNNEL_REMCSUM; - } } - skb_scrub_packet(skb, xnet); - min_headroom = LL_RESERVED_SPACE(dst->dev) + dst->header_len - + VXLAN_HLEN + sizeof(struct ipv6hdr) + + VXLAN_HLEN + iphdr_len + (skb_vlan_tag_present(skb) ? VLAN_HLEN : 0); /* Need space for new headers (invalidates iph ptr) */ err = skb_cow_head(skb, min_headroom); if (unlikely(err)) { kfree_skb(skb); - goto err; + return err; } skb = vlan_hwaccel_push_inside(skb); - if (WARN_ON(!skb)) { - err = -ENOMEM; - goto err; - } + if (WARN_ON(!skb)) + return -ENOMEM; - skb = iptunnel_handle_offloads(skb, udp_sum, type); - if (IS_ERR(skb)) { - err = -EINVAL; - goto err; - } + skb = iptunnel_handle_offloads(skb, type); + if (IS_ERR(skb)) + return PTR_ERR(skb); vxh = (struct vxlanhdr *) __skb_push(skb, sizeof(*vxh)); - vxh->vx_flags = htonl(VXLAN_HF_VNI); - vxh->vx_vni = vni; + vxh->vx_flags = VXLAN_HF_VNI; + vxh->vx_vni = vxlan_vni_field(vni); if (type & SKB_GSO_TUNNEL_REMCSUM) { - u32 data = (skb_checksum_start_offset(skb) - hdrlen) >> - VXLAN_RCO_SHIFT; + unsigned int start; - if (skb->csum_offset == offsetof(struct udphdr, check)) - data |= VXLAN_RCO_UDP; - - vxh->vx_vni |= htonl(data); - vxh->vx_flags |= htonl(VXLAN_HF_RCO); + start = skb_checksum_start_offset(skb) - sizeof(struct vxlanhdr); + vxh->vx_vni |= vxlan_compute_rco(start, skb->csum_offset); + vxh->vx_flags |= VXLAN_HF_RCO; if (!skb_is_gso(skb)) { skb->ip_summed = CHECKSUM_NONE; @@@ -1761,102 -1749,63 +1749,63 @@@ vxlan_build_gbp_hdr(vxh, vxflags, md); skb_set_inner_protocol(skb, htons(ETH_P_TEB)); - - udp_tunnel6_xmit_skb(dst, sk, skb, dev, saddr, daddr, prio, - ttl, src_port, dst_port, - !!(vxflags & VXLAN_F_UDP_ZERO_CSUM6_TX)); return 0; - err: - dst_release(dst); - return err; } - #endif - static int vxlan_xmit_skb(struct rtable *rt, struct sock *sk, struct sk_buff *skb, - __be32 src, __be32 dst, __u8 tos, __u8 ttl, __be16 df, - __be16 src_port, __be16 dst_port, __be32 vni, - struct vxlan_metadata *md, bool xnet, u32 vxflags) + static struct rtable *vxlan_get_route(struct vxlan_dev *vxlan, + struct sk_buff *skb, int oif, u8 tos, + __be32 daddr, __be32 *saddr, + struct dst_cache *dst_cache, + struct ip_tunnel_info *info) { - struct vxlanhdr *vxh; - int min_headroom; - int err; - bool udp_sum = !!(vxflags & VXLAN_F_UDP_CSUM); - int type = udp_sum ? SKB_GSO_UDP_TUNNEL_CSUM : SKB_GSO_UDP_TUNNEL; - u16 hdrlen = sizeof(struct vxlanhdr); - - if ((vxflags & VXLAN_F_REMCSUM_TX) && - skb->ip_summed == CHECKSUM_PARTIAL) { - int csum_start = skb_checksum_start_offset(skb); - - if (csum_start <= VXLAN_MAX_REMCSUM_START && - !(csum_start & VXLAN_RCO_SHIFT_MASK) && - (skb->csum_offset == offsetof(struct udphdr, check) || - skb->csum_offset == offsetof(struct tcphdr, check))) { - udp_sum = false; - type |= SKB_GSO_TUNNEL_REMCSUM; - } - } - - min_headroom = LL_RESERVED_SPACE(rt->dst.dev) + rt->dst.header_len - + VXLAN_HLEN + sizeof(struct iphdr) - + (skb_vlan_tag_present(skb) ? VLAN_HLEN : 0); + struct rtable *rt = NULL; + bool use_cache = false; + struct flowi4 fl4; - /* Need space for new headers (invalidates iph ptr) */ - err = skb_cow_head(skb, min_headroom); - if (unlikely(err)) { - kfree_skb(skb); - return err; + /* when the ip_tunnel_info is availble, the tos used for lookup is + * packet independent, so we can use the cache + */ + if (!skb->mark && (!tos || info)) { + use_cache = true; + rt = dst_cache_get_ip4(dst_cache, saddr); + if (rt) + return rt; } - skb = vlan_hwaccel_push_inside(skb); - if (WARN_ON(!skb)) - return -ENOMEM; - - skb = iptunnel_handle_offloads(skb, udp_sum, type); - if (IS_ERR(skb)) - return PTR_ERR(skb); - - vxh = (struct vxlanhdr *) __skb_push(skb, sizeof(*vxh)); - vxh->vx_flags = htonl(VXLAN_HF_VNI); - vxh->vx_vni = vni; - - if (type & SKB_GSO_TUNNEL_REMCSUM) { - u32 data = (skb_checksum_start_offset(skb) - hdrlen) >> - VXLAN_RCO_SHIFT; - - if (skb->csum_offset == offsetof(struct udphdr, check)) - data |= VXLAN_RCO_UDP; - - vxh->vx_vni |= htonl(data); - vxh->vx_flags |= htonl(VXLAN_HF_RCO); + memset(&fl4, 0, sizeof(fl4)); + fl4.flowi4_oif = oif; + fl4.flowi4_tos = RT_TOS(tos); + fl4.flowi4_mark = skb->mark; + fl4.flowi4_proto = IPPROTO_UDP; + fl4.daddr = daddr; + fl4.saddr = vxlan->cfg.saddr.sin.sin_addr.s_addr; - if (!skb_is_gso(skb)) { - skb->ip_summed = CHECKSUM_NONE; - skb->encapsulation = 0; - } + rt = ip_route_output_key(vxlan->net, &fl4); + if (!IS_ERR(rt)) { + *saddr = fl4.saddr; + if (use_cache) + dst_cache_set_ip4(dst_cache, &rt->dst, fl4.saddr); } - - if (vxflags & VXLAN_F_GBP) - vxlan_build_gbp_hdr(vxh, vxflags, md); - - skb_set_inner_protocol(skb, htons(ETH_P_TEB)); - - udp_tunnel_xmit_skb(rt, sk, skb, src, dst, tos, ttl, df, - src_port, dst_port, xnet, - !(vxflags & VXLAN_F_UDP_CSUM)); - return 0; + return rt; } #if IS_ENABLED(CONFIG_IPV6) static struct dst_entry *vxlan6_get_route(struct vxlan_dev *vxlan, struct sk_buff *skb, int oif, const struct in6_addr *daddr, - struct in6_addr *saddr) + struct in6_addr *saddr, + struct dst_cache *dst_cache) { struct dst_entry *ndst; struct flowi6 fl6; int err; + if (!skb->mark) { + ndst = dst_cache_get_ip6(dst_cache, saddr); + if (ndst) + return ndst; + } + memset(&fl6, 0, sizeof(fl6)); fl6.flowi6_oif = oif; fl6.daddr = *daddr; @@@ -1871,6 -1820,8 +1820,8 @@@ return ERR_PTR(err); *saddr = fl6.saddr; + if (!skb->mark) + dst_cache_set_ip6(dst_cache, ndst, saddr); return ndst; } #endif @@@ -1923,22 -1874,24 +1874,24 @@@ static void vxlan_encap_bypass(struct s static void vxlan_xmit_one(struct sk_buff *skb, struct net_device *dev, struct vxlan_rdst *rdst, bool did_rsc) { + struct dst_cache *dst_cache; struct ip_tunnel_info *info; struct vxlan_dev *vxlan = netdev_priv(dev); struct sock *sk; struct rtable *rt = NULL; const struct iphdr *old_iph; - struct flowi4 fl4; union vxlan_addr *dst; union vxlan_addr remote_ip; struct vxlan_metadata _md; struct vxlan_metadata *md = &_md; __be16 src_port = 0, dst_port; - u32 vni; + __be32 vni; __be16 df = 0; __u8 tos, ttl; int err; u32 flags = vxlan->flags; + bool udp_sum = false; + bool xnet = !net_eq(vxlan->net, dev_net(vxlan->dev)); info = skb_tunnel_info(skb); @@@ -1946,6 -1899,7 +1899,7 @@@ dst_port = rdst->remote_port ? rdst->remote_port : vxlan->cfg.dst_port; vni = rdst->remote_vni; dst = &rdst->remote_ip; + dst_cache = &rdst->dst_cache; } else { if (!info) { WARN_ONCE(1, "%s: Missing encapsulation instructions\n", @@@ -1953,13 -1907,14 +1907,14 @@@ goto drop; } dst_port = info->key.tp_dst ? : vxlan->cfg.dst_port; - vni = be64_to_cpu(info->key.tun_id); + vni = vxlan_tun_id_to_vni(info->key.tun_id); remote_ip.sa.sa_family = ip_tunnel_info_af(info); if (remote_ip.sa.sa_family == AF_INET) remote_ip.sin.sin_addr.s_addr = info->key.u.ipv4.dst; else remote_ip.sin6.sin6_addr = info->key.u.ipv6.dst; dst = &remote_ip; + dst_cache = &info->dst_cache; } if (vxlan_addr_any(dst)) { @@@ -1987,6 -1942,7 +1942,7 @@@ if (info) { ttl = info->key.ttl; tos = info->key.tos; + udp_sum = !!(info->key.tun_flags & TUNNEL_CSUM); if (info->options_len) md = ip_tunnel_info_opts(info); @@@ -1995,29 -1951,16 +1951,16 @@@ } if (dst->sa.sa_family == AF_INET) { + __be32 saddr; + if (!vxlan->vn4_sock) goto drop; sk = vxlan->vn4_sock->sock->sk; - if (info) { - if (info->key.tun_flags & TUNNEL_DONT_FRAGMENT) - df = htons(IP_DF); - - if (info->key.tun_flags & TUNNEL_CSUM) - flags |= VXLAN_F_UDP_CSUM; - else - flags &= ~VXLAN_F_UDP_CSUM; - } - - memset(&fl4, 0, sizeof(fl4)); - fl4.flowi4_oif = rdst ? rdst->remote_ifindex : 0; - fl4.flowi4_tos = RT_TOS(tos); - fl4.flowi4_mark = skb->mark; - fl4.flowi4_proto = IPPROTO_UDP; - fl4.daddr = dst->sin.sin_addr.s_addr; - fl4.saddr = vxlan->cfg.saddr.sin.sin_addr.s_addr; - - rt = ip_route_output_key(vxlan->net, &fl4); + rt = vxlan_get_route(vxlan, skb, + rdst ? rdst->remote_ifindex : 0, tos, + dst->sin.sin_addr.s_addr, &saddr, + dst_cache, info); if (IS_ERR(rt)) { netdev_dbg(dev, "no route to %pI4\n", &dst->sin.sin_addr.s_addr); @@@ -2047,18 -1990,21 +1990,21 @@@ return; } + if (!info) + udp_sum = !(flags & VXLAN_F_UDP_ZERO_CSUM_TX); + else if (info->key.tun_flags & TUNNEL_DONT_FRAGMENT) + df = htons(IP_DF); + tos = ip_tunnel_ecn_encap(tos, old_iph, skb); ttl = ttl ? : ip4_dst_hoplimit(&rt->dst); - err = vxlan_xmit_skb(rt, sk, skb, fl4.saddr, - dst->sin.sin_addr.s_addr, tos, ttl, df, - src_port, dst_port, htonl(vni << 8), md, - !net_eq(vxlan->net, dev_net(vxlan->dev)), - flags); - if (err < 0) { - /* skb is already freed. */ - skb = NULL; - goto rt_tx_error; - } + err = vxlan_build_skb(skb, &rt->dst, sizeof(struct iphdr), + vni, md, flags, udp_sum); + if (err < 0) + goto xmit_tx_error; + + udp_tunnel_xmit_skb(rt, sk, skb, saddr, + dst->sin.sin_addr.s_addr, tos, ttl, df, + src_port, dst_port, xnet, !udp_sum); #if IS_ENABLED(CONFIG_IPV6) } else { struct dst_entry *ndst; @@@ -2071,7 -2017,8 +2017,8 @@@ ndst = vxlan6_get_route(vxlan, skb, rdst ? rdst->remote_ifindex : 0, - &dst->sin6.sin6_addr, &saddr); + &dst->sin6.sin6_addr, &saddr, + dst_cache); if (IS_ERR(ndst)) { netdev_dbg(dev, "no route to %pI6\n", &dst->sin6.sin6_addr); @@@ -2103,18 -2050,20 +2050,20 @@@ return; } - if (info) { - if (info->key.tun_flags & TUNNEL_CSUM) - flags &= ~VXLAN_F_UDP_ZERO_CSUM6_TX; - else - flags |= VXLAN_F_UDP_ZERO_CSUM6_TX; - } + if (!info) + udp_sum = !(flags & VXLAN_F_UDP_ZERO_CSUM6_TX); ttl = ttl ? : ip6_dst_hoplimit(ndst); - err = vxlan6_xmit_skb(ndst, sk, skb, dev, &saddr, &dst->sin6.sin6_addr, - 0, ttl, src_port, dst_port, htonl(vni << 8), md, - !net_eq(vxlan->net, dev_net(vxlan->dev)), - flags); + skb_scrub_packet(skb, xnet); + err = vxlan_build_skb(skb, ndst, sizeof(struct ipv6hdr), + vni, md, flags, udp_sum); + if (err < 0) { + dst_release(ndst); + return; + } + udp_tunnel6_xmit_skb(ndst, sk, skb, dev, + &saddr, &dst->sin6.sin6_addr, + 0, ttl, src_port, dst_port, !udp_sum); #endif } @@@ -2124,6 -2073,9 +2073,9 @@@ drop dev->stats.tx_dropped++; goto tx_free; + xmit_tx_error: + /* skb is already freed. */ + skb = NULL; rt_tx_error: ip_rt_put(rt); tx_error: @@@ -2171,11 -2123,9 +2123,11 @@@ static netdev_tx_t vxlan_xmit(struct sk #endif } - if (vxlan->flags & VXLAN_F_COLLECT_METADATA && - info && info->mode & IP_TUNNEL_INFO_TX) { - vxlan_xmit_one(skb, dev, NULL, false); + if (vxlan->flags & VXLAN_F_COLLECT_METADATA) { + if (info && info->mode & IP_TUNNEL_INFO_TX) + vxlan_xmit_one(skb, dev, NULL, false); + else + kfree_skb(skb); return NETDEV_TX_OK; } @@@ -2263,7 -2213,7 +2215,7 @@@ static void vxlan_cleanup(unsigned lon static void vxlan_vs_add_dev(struct vxlan_sock *vs, struct vxlan_dev *vxlan) { struct vxlan_net *vn = net_generic(vxlan->net, vxlan_net_id); - __u32 vni = vxlan->default_dst.remote_vni; + __be32 vni = vxlan->default_dst.remote_vni; spin_lock(&vn->sock_lock); hlist_add_head_rcu(&vxlan->hlist, vni_head(vs, vni)); @@@ -2369,68 -2319,29 +2321,43 @@@ static void vxlan_set_multicast_list(st { } -static int vxlan_change_mtu(struct net_device *dev, int new_mtu) +static int __vxlan_change_mtu(struct net_device *dev, + struct net_device *lowerdev, + struct vxlan_rdst *dst, int new_mtu, bool strict) { - struct vxlan_dev *vxlan = netdev_priv(dev); - struct vxlan_rdst *dst = &vxlan->default_dst; - struct net_device *lowerdev; - int max_mtu; + int max_mtu = IP_MAX_MTU; - lowerdev = __dev_get_by_index(vxlan->net, dst->remote_ifindex); - if (lowerdev == NULL) - return eth_change_mtu(dev, new_mtu); + if (lowerdev) + max_mtu = lowerdev->mtu; if (dst->remote_ip.sa.sa_family == AF_INET6) - max_mtu = lowerdev->mtu - VXLAN6_HEADROOM; + max_mtu -= VXLAN6_HEADROOM; else - max_mtu = lowerdev->mtu - VXLAN_HEADROOM; + max_mtu -= VXLAN_HEADROOM; - if (new_mtu < 68 || new_mtu > max_mtu) + if (new_mtu < 68) return -EINVAL; + if (new_mtu > max_mtu) { + if (strict) + return -EINVAL; + + new_mtu = max_mtu; + } + dev->mtu = new_mtu; return 0; } +static int vxlan_change_mtu(struct net_device *dev, int new_mtu) +{ + struct vxlan_dev *vxlan = netdev_priv(dev); + struct vxlan_rdst *dst = &vxlan->default_dst; + struct net_device *lowerdev = __dev_get_by_index(vxlan->net, + dst->remote_ifindex); + return __vxlan_change_mtu(dev, lowerdev, dst, new_mtu, true); +} + - static int egress_ipv4_tun_info(struct net_device *dev, struct sk_buff *skb, - struct ip_tunnel_info *info, - __be16 sport, __be16 dport) - { - struct vxlan_dev *vxlan = netdev_priv(dev); - struct rtable *rt; - struct flowi4 fl4; - - memset(&fl4, 0, sizeof(fl4)); - fl4.flowi4_tos = RT_TOS(info->key.tos); - fl4.flowi4_mark = skb->mark; - fl4.flowi4_proto = IPPROTO_UDP; - fl4.daddr = info->key.u.ipv4.dst; - - rt = ip_route_output_key(vxlan->net, &fl4); - if (IS_ERR(rt)) - return PTR_ERR(rt); - ip_rt_put(rt); - - info->key.u.ipv4.src = fl4.saddr; - info->key.tp_src = sport; - info->key.tp_dst = dport; - return 0; - } - static int vxlan_fill_metadata_dst(struct net_device *dev, struct sk_buff *skb) { struct vxlan_dev *vxlan = netdev_priv(dev); @@@ -2442,9 -2353,16 +2369,16 @@@ dport = info->key.tp_dst ? : vxlan->cfg.dst_port; if (ip_tunnel_info_af(info) == AF_INET) { + struct rtable *rt; + if (!vxlan->vn4_sock) return -EINVAL; - return egress_ipv4_tun_info(dev, skb, info, sport, dport); + rt = vxlan_get_route(vxlan, skb, 0, info->key.tos, + info->key.u.ipv4.dst, + &info->key.u.ipv4.src, NULL, info); + if (IS_ERR(rt)) + return PTR_ERR(rt); + ip_rt_put(rt); } else { #if IS_ENABLED(CONFIG_IPV6) struct dst_entry *ndst; @@@ -2453,17 -2371,16 +2387,16 @@@ return -EINVAL; ndst = vxlan6_get_route(vxlan, skb, 0, &info->key.u.ipv6.dst, - &info->key.u.ipv6.src); + &info->key.u.ipv6.src, NULL); if (IS_ERR(ndst)) return PTR_ERR(ndst); dst_release(ndst); - - info->key.tp_src = sport; - info->key.tp_dst = dport; #else /* !CONFIG_IPV6 */ return -EPFNOSUPPORT; #endif } + info->key.tp_src = sport; + info->key.tp_dst = dport; return 0; } @@@ -2539,7 -2456,6 +2472,7 @@@ static void vxlan_setup(struct net_devi dev->hw_features |= NETIF_F_GSO_SOFTWARE; dev->hw_features |= NETIF_F_HW_VLAN_CTAG_TX | NETIF_F_HW_VLAN_STAG_TX; netif_keep_dst(dev); + dev->priv_flags &= ~IFF_TX_SKB_SHARING; dev->priv_flags |= IFF_LIVE_ADDR_CHANGE | IFF_NO_QUEUE; INIT_LIST_HEAD(&vxlan->next); @@@ -2782,7 -2698,6 +2715,7 @@@ static int vxlan_dev_configure(struct n int err; bool use_ipv6 = false; __be16 default_port = vxlan->cfg.dst_port; + struct net_device *lowerdev = NULL; vxlan->net = src_net; @@@ -2803,7 -2718,9 +2736,7 @@@ } if (conf->remote_ifindex) { - struct net_device *lowerdev - = __dev_get_by_index(src_net, conf->remote_ifindex); - + lowerdev = __dev_get_by_index(src_net, conf->remote_ifindex); dst->remote_ifindex = conf->remote_ifindex; if (!lowerdev) { @@@ -2827,12 -2744,6 +2760,12 @@@ needed_headroom = lowerdev->hard_header_len; } + if (conf->mtu) { + err = __vxlan_change_mtu(dev, lowerdev, dst, conf->mtu, false); + if (err) + return err; + } + if (use_ipv6 || conf->flags & VXLAN_F_COLLECT_METADATA) needed_headroom += VXLAN6_HEADROOM; else @@@ -2917,7 -2828,7 +2850,7 @@@ static int vxlan_newlink(struct net *sr memset(&conf, 0, sizeof(conf)); if (data[IFLA_VXLAN_ID]) - conf.vni = nla_get_u32(data[IFLA_VXLAN_ID]); + conf.vni = cpu_to_be32(nla_get_u32(data[IFLA_VXLAN_ID])); if (data[IFLA_VXLAN_GROUP]) { conf.remote_ip.sin.sin_addr.s_addr = nla_get_in_addr(data[IFLA_VXLAN_GROUP]); @@@ -2985,8 -2896,9 +2918,9 @@@ if (data[IFLA_VXLAN_PORT]) conf.dst_port = nla_get_be16(data[IFLA_VXLAN_PORT]); - if (data[IFLA_VXLAN_UDP_CSUM] && nla_get_u8(data[IFLA_VXLAN_UDP_CSUM])) - conf.flags |= VXLAN_F_UDP_CSUM; + if (data[IFLA_VXLAN_UDP_CSUM] && + !nla_get_u8(data[IFLA_VXLAN_UDP_CSUM])) + conf.flags |= VXLAN_F_UDP_ZERO_CSUM_TX; if (data[IFLA_VXLAN_UDP_ZERO_CSUM6_TX] && nla_get_u8(data[IFLA_VXLAN_UDP_ZERO_CSUM6_TX])) @@@ -3021,7 -2933,7 +2955,7 @@@ break; case -EEXIST: - pr_info("duplicate VNI %u\n", conf.vni); + pr_info("duplicate VNI %u\n", be32_to_cpu(conf.vni)); break; } @@@ -3079,7 -2991,7 +3013,7 @@@ static int vxlan_fill_info(struct sk_bu .high = htons(vxlan->cfg.port_max), }; - if (nla_put_u32(skb, IFLA_VXLAN_ID, dst->remote_vni)) + if (nla_put_u32(skb, IFLA_VXLAN_ID, be32_to_cpu(dst->remote_vni))) goto nla_put_failure; if (!vxlan_addr_any(&dst->remote_ip)) { @@@ -3130,7 -3042,7 +3064,7 @@@ nla_put_u32(skb, IFLA_VXLAN_LIMIT, vxlan->cfg.addrmax) || nla_put_be16(skb, IFLA_VXLAN_PORT, vxlan->cfg.dst_port) || nla_put_u8(skb, IFLA_VXLAN_UDP_CSUM, - !!(vxlan->flags & VXLAN_F_UDP_CSUM)) || + !(vxlan->flags & VXLAN_F_UDP_ZERO_CSUM_TX)) || nla_put_u8(skb, IFLA_VXLAN_UDP_ZERO_CSUM6_TX, !!(vxlan->flags & VXLAN_F_UDP_ZERO_CSUM6_TX)) || nla_put_u8(skb, IFLA_VXLAN_UDP_ZERO_CSUM6_RX, diff --combined drivers/net/wireless/intel/iwlwifi/Kconfig index 7438fbe,11932d5..16c4f38 --- a/drivers/net/wireless/intel/iwlwifi/Kconfig +++ b/drivers/net/wireless/intel/iwlwifi/Kconfig @@@ -53,6 -53,7 +53,6 @@@ config IWLWIFI_LED config IWLDVM tristate "Intel Wireless WiFi DVM Firmware support" - depends on m help This is the driver that supports the DVM firmware. The list of the devices that use this firmware is available here: @@@ -98,6 -99,18 +98,18 @@@ config IWLWIFI_UAPS If unsure, say N. + config IWLWIFI_PCIE_RTPM + bool "Enable runtime power management mode for PCIe devices" + depends on IWLMVM && PM + default false + help + Say Y here to enable runtime power management for PCIe + devices. If enabled, the device will go into low power mode + when idle for a short period of time, allowing for improved + power saving during runtime. + + If unsure, say N. + menu "Debugging Options" config IWLWIFI_DEBUG diff --combined drivers/net/wireless/intel/iwlwifi/mvm/scan.c index ea1e177,1e1ab9d..aa6d807 --- a/drivers/net/wireless/intel/iwlwifi/mvm/scan.c +++ b/drivers/net/wireless/intel/iwlwifi/mvm/scan.c @@@ -930,8 -930,11 +930,11 @@@ int iwl_mvm_config_scan(struct iwl_mvm if (WARN_ON(num_channels > mvm->fw->ucode_capa.n_scan_channels)) return -ENOBUFS; - if (type == mvm->scan_type) + if (type == mvm->scan_type) { + IWL_DEBUG_SCAN(mvm, + "Ignoring UMAC scan config of the same type\n"); return 0; + } cmd_size = sizeof(*scan_config) + mvm->fw->ucode_capa.n_scan_channels; @@@ -1109,7 -1112,7 +1112,7 @@@ static int iwl_mvm_scan_umac(struct iwl cmd->general_flags = cpu_to_le32(iwl_mvm_scan_umac_flags(mvm, params, vif)); - if (type == IWL_MVM_SCAN_SCHED) + if (type == IWL_MVM_SCAN_SCHED || type == IWL_MVM_SCAN_NETDETECT) cmd->flags = cpu_to_le32(IWL_UMAC_SCAN_FLAG_PREEMPTIVE); if (iwl_mvm_scan_use_ebs(mvm, vif)) @@@ -1298,10 -1301,6 +1301,10 @@@ int iwl_mvm_sched_scan_start(struct iwl return -EBUSY; } + /* we don't support "match all" in the firmware */ + if (!req->n_match_sets) + return -EOPNOTSUPP; + ret = iwl_mvm_check_running_scans(mvm, type); if (ret) return ret; @@@ -1355,7 -1354,7 +1358,7 @@@ if (fw_has_capa(&mvm->fw->ucode_capa, IWL_UCODE_TLV_CAPA_UMAC_SCAN)) { hcmd.id = iwl_cmd_id(SCAN_REQ_UMAC, IWL_ALWAYS_LONG_GROUP, 0); - ret = iwl_mvm_scan_umac(mvm, vif, &params, IWL_MVM_SCAN_SCHED); + ret = iwl_mvm_scan_umac(mvm, vif, &params, type); } else { hcmd.id = SCAN_OFFLOAD_REQUEST_CMD; ret = iwl_mvm_scan_lmac(mvm, vif, &params); diff --combined drivers/net/wireless/intel/iwlwifi/pcie/internal.h index 73c9559,2f95916..542bbc5 --- a/drivers/net/wireless/intel/iwlwifi/pcie/internal.h +++ b/drivers/net/wireless/intel/iwlwifi/pcie/internal.h @@@ -2,6 -2,7 +2,7 @@@ * * Copyright(c) 2003 - 2015 Intel Corporation. All rights reserved. * Copyright(c) 2013 - 2015 Intel Mobile Communications GmbH + * Copyright(c) 2016 Intel Deutschland GmbH * * Portions of this file are derived from the ipw3945 project, as well * as portions of the ieee80211 subsystem header files. @@@ -56,17 -57,23 +57,23 @@@ #define RX_NUM_QUEUES 1 #define RX_POST_REQ_ALLOC 2 #define RX_CLAIM_REQ_ALLOC 8 - #define RX_POOL_SIZE ((RX_CLAIM_REQ_ALLOC - RX_POST_REQ_ALLOC) * RX_NUM_QUEUES) - #define RX_LOW_WATERMARK 8 + #define RX_PENDING_WATERMARK 16 struct iwl_host_cmd; /*This file includes the declaration that are internal to the * trans_pcie layer */ + /** + * struct iwl_rx_mem_buffer + * @page_dma: bus address of rxb page + * @page: driver's pointer to the rxb page + * @vid: index of this rxb in the global table + */ struct iwl_rx_mem_buffer { dma_addr_t page_dma; struct page *page; + u16 vid; struct list_head list; }; @@@ -90,8 -97,12 +97,12 @@@ struct isr_statistics /** * struct iwl_rxq - Rx queue - * @bd: driver's pointer to buffer of receive buffer descriptors (rbd) + * @id: queue index + * @bd: driver's pointer to buffer of receive buffer descriptors (rbd). + * Address size is 32 bit in pre-9000 devices and 64 bit in 9000 devices. * @bd_dma: bus address of buffer of receive buffer descriptors (rbd) + * @ubd: driver's pointer to buffer of used receive buffer descriptors (rbd) + * @ubd_dma: physical address of buffer of used receive buffer descriptors (rbd) * @read: Shared index to newest available Rx buffer * @write: Shared index to oldest written Rx packet * @free_count: Number of pre-allocated buffers in rx_free @@@ -103,32 -114,34 +114,34 @@@ * @rb_stts: driver's pointer to receive buffer status * @rb_stts_dma: bus address of receive buffer status * @lock: - * @pool: initial pool of iwl_rx_mem_buffer for the queue - * @queue: actual rx queue + * @queue: actual rx queue. Not used for multi-rx queue. * * NOTE: rx_free and rx_used are used as a FIFO for iwl_rx_mem_buffers */ struct iwl_rxq { - __le32 *bd; + int id; + void *bd; dma_addr_t bd_dma; + __le32 *used_bd; + dma_addr_t used_bd_dma; u32 read; u32 write; u32 free_count; u32 used_count; u32 write_actual; + u32 queue_size; struct list_head rx_free; struct list_head rx_used; bool need_update; struct iwl_rb_status *rb_stts; dma_addr_t rb_stts_dma; spinlock_t lock; - struct iwl_rx_mem_buffer pool[RX_QUEUE_SIZE]; + struct napi_struct napi; struct iwl_rx_mem_buffer *queue[RX_QUEUE_SIZE]; }; /** * struct iwl_rb_allocator - Rx allocator - * @pool: initial pool of allocator * @req_pending: number of requests the allcator had not processed yet * @req_ready: number of requests honored and ready for claiming * @rbd_allocated: RBDs with pages allocated and ready to be handled to @@@ -140,7 -153,6 +153,6 @@@ * @rx_alloc: work struct for background calls */ struct iwl_rb_allocator { - struct iwl_rx_mem_buffer pool[RX_POOL_SIZE]; atomic_t req_pending; atomic_t req_ready; struct list_head rbd_allocated; @@@ -280,6 -292,7 +292,7 @@@ struct iwl_txq bool ampdu; bool block; unsigned long wd_timeout; + struct sk_buff_head overflow_q; }; static inline dma_addr_t @@@ -297,6 -310,8 +310,8 @@@ struct iwl_tso_hdr_page /** * struct iwl_trans_pcie - PCIe transport specific data * @rxq: all the RX queue data + * @rx_pool: initial pool of iwl_rx_mem_buffer for all the queues + * @global_table: table mapping received VID from hw to rxb * @rba: allocator for RX replenishing * @drv - pointer to iwl_drv * @trans: pointer to the generic transport area @@@ -323,13 -338,14 +338,14 @@@ * @fw_mon_size: size of the buffer for the firmware monitor */ struct iwl_trans_pcie { - struct iwl_rxq rxq; + struct iwl_rxq *rxq; + struct iwl_rx_mem_buffer rx_pool[MQ_RX_POOL_SIZE]; + struct iwl_rx_mem_buffer *global_table[MQ_RX_TABLE_SIZE]; struct iwl_rb_allocator rba; struct iwl_trans *trans; struct iwl_drv *drv; struct net_device napi_dev; - struct napi_struct napi; struct __percpu iwl_tso_hdr_page *tso_hdr_page; @@@ -359,6 -375,7 +375,7 @@@ bool ucode_write_complete; wait_queue_head_t ucode_write_waitq; wait_queue_head_t wait_command_queue; + wait_queue_head_t d0i3_waitq; u8 cmd_queue; u8 cmd_fifo; @@@ -490,15 -507,6 +507,15 @@@ static inline void iwl_enable_interrupt iwl_write32(trans, CSR_INT_MASK, trans_pcie->inta_mask); } +static inline void iwl_enable_fw_load_int(struct iwl_trans *trans) +{ + struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans); + + IWL_DEBUG_ISR(trans, "Enabling FW load interrupt\n"); + trans_pcie->inta_mask = CSR_INT_BIT_FH_TX; + iwl_write32(trans, CSR_INT_MASK, trans_pcie->inta_mask); +} + static inline void iwl_enable_rfkill_int(struct iwl_trans *trans) { struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans); @@@ -588,4 -596,7 +605,7 @@@ static inline int iwl_trans_pcie_dbgfs_ } #endif + int iwl_pci_fw_exit_d0i3(struct iwl_trans *trans); + int iwl_pci_fw_enter_d0i3(struct iwl_trans *trans); + #endif /* __iwl_trans_int_pcie_h__ */ diff --combined drivers/net/wireless/intel/iwlwifi/pcie/rx.c index 152cf9a,51314e56..07973ef --- a/drivers/net/wireless/intel/iwlwifi/pcie/rx.c +++ b/drivers/net/wireless/intel/iwlwifi/pcie/rx.c @@@ -2,6 -2,7 +2,7 @@@ * * Copyright(c) 2003 - 2014 Intel Corporation. All rights reserved. * Copyright(c) 2013 - 2015 Intel Mobile Communications GmbH + * Copyright(c) 2016 Intel Deutschland GmbH * * Portions of this file are derived from the ipw3945 project, as well * as portions of the ieee80211 subsystem header files. @@@ -140,8 -141,8 +141,8 @@@ */ static int iwl_rxq_space(const struct iwl_rxq *rxq) { - /* Make sure RX_QUEUE_SIZE is a power of 2 */ - BUILD_BUG_ON(RX_QUEUE_SIZE & (RX_QUEUE_SIZE - 1)); + /* Make sure rx queue size is a power of 2 */ + WARN_ON(rxq->queue_size & (rxq->queue_size - 1)); /* * There can be up to (RX_QUEUE_SIZE - 1) free slots, to avoid ambiguity @@@ -149,7 -150,7 +150,7 @@@ * The following is equivalent to modulo by RX_QUEUE_SIZE and is well * defined for negative dividends. */ - return (rxq->read - rxq->write - 1) & (RX_QUEUE_SIZE - 1); + return (rxq->read - rxq->write - 1) & (rxq->queue_size - 1); } /* @@@ -160,6 -161,12 +161,12 @@@ static inline __le32 iwl_pcie_dma_addr2 return cpu_to_le32((u32)(dma_addr >> 8)); } + static void iwl_pcie_write_prph_64(struct iwl_trans *trans, u64 ofs, u64 val) + { + iwl_write_prph(trans, ofs, val & 0xffffffff); + iwl_write_prph(trans, ofs + 4, val >> 32); + } + /* * iwl_pcie_rx_stop - stops the Rx DMA */ @@@ -173,10 -180,9 +180,9 @@@ int iwl_pcie_rx_stop(struct iwl_trans * /* * iwl_pcie_rxq_inc_wr_ptr - Update the write pointer for the RX queue */ - static void iwl_pcie_rxq_inc_wr_ptr(struct iwl_trans *trans) + static void iwl_pcie_rxq_inc_wr_ptr(struct iwl_trans *trans, + struct iwl_rxq *rxq) { - struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans); - struct iwl_rxq *rxq = &trans_pcie->rxq; u32 reg; lockdep_assert_held(&rxq->lock); @@@ -201,24 -207,73 +207,73 @@@ } rxq->write_actual = round_down(rxq->write, 8); - iwl_write32(trans, FH_RSCSR_CHNL0_WPTR, rxq->write_actual); + if (trans->cfg->mq_rx_supported) + iwl_write_prph(trans, RFH_Q_FRBDCB_WIDX(rxq->id), + rxq->write_actual); + else + iwl_write32(trans, FH_RSCSR_CHNL0_WPTR, rxq->write_actual); } static void iwl_pcie_rxq_check_wrptr(struct iwl_trans *trans) { struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans); - struct iwl_rxq *rxq = &trans_pcie->rxq; + int i; - spin_lock(&rxq->lock); + for (i = 0; i < trans->num_rx_queues; i++) { + struct iwl_rxq *rxq = &trans_pcie->rxq[i]; + + if (!rxq->need_update) + continue; + spin_lock(&rxq->lock); + iwl_pcie_rxq_inc_wr_ptr(trans, rxq); + rxq->need_update = false; + spin_unlock(&rxq->lock); + } + } + + static void iwl_pcie_rxq_mq_restock(struct iwl_trans *trans, + struct iwl_rxq *rxq) + { + struct iwl_rx_mem_buffer *rxb; + + /* + * If the device isn't enabled - no need to try to add buffers... + * This can happen when we stop the device and still have an interrupt + * pending. We stop the APM before we sync the interrupts because we + * have to (see comment there). On the other hand, since the APM is + * stopped, we cannot access the HW (in particular not prph). + * So don't try to restock if the APM has been already stopped. + */ + if (!test_bit(STATUS_DEVICE_ENABLED, &trans->status)) + return; - if (!rxq->need_update) - goto exit_unlock; + spin_lock(&rxq->lock); + while (rxq->free_count) { + __le64 *bd = (__le64 *)rxq->bd; - iwl_pcie_rxq_inc_wr_ptr(trans); - rxq->need_update = false; + /* Get next free Rx buffer, remove from free list */ + rxb = list_first_entry(&rxq->rx_free, struct iwl_rx_mem_buffer, + list); + list_del(&rxb->list); - exit_unlock: + /* 12 first bits are expected to be empty */ + WARN_ON(rxb->page_dma & DMA_BIT_MASK(12)); + /* Point to Rx buffer via next RBD in circular buffer */ + bd[rxq->write] = cpu_to_le64(rxb->page_dma | rxb->vid); + rxq->write = (rxq->write + 1) & MQ_RX_TABLE_MASK; + rxq->free_count--; + } spin_unlock(&rxq->lock); + + /* + * If we've added more space for the firmware to place data, tell it. + * Increment device's write pointer in multiples of 8. + */ + if (rxq->write_actual != (rxq->write & ~0x7)) { + spin_lock(&rxq->lock); + iwl_pcie_rxq_inc_wr_ptr(trans, rxq); + spin_unlock(&rxq->lock); + } } /* @@@ -232,10 -287,8 +287,8 @@@ * also updates the memory address in the firmware to reference the new * target buffer. */ - static void iwl_pcie_rxq_restock(struct iwl_trans *trans) + static void iwl_pcie_rxq_restock(struct iwl_trans *trans, struct iwl_rxq *rxq) { - struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans); - struct iwl_rxq *rxq = &trans_pcie->rxq; struct iwl_rx_mem_buffer *rxb; /* @@@ -251,6 -304,7 +304,7 @@@ spin_lock(&rxq->lock); while ((iwl_rxq_space(rxq) > 0) && (rxq->free_count)) { + __le32 *bd = (__le32 *)rxq->bd; /* The overwritten rxb must be a used one */ rxb = rxq->queue[rxq->write]; BUG_ON(rxb && rxb->page); @@@ -261,7 -315,7 +315,7 @@@ list_del(&rxb->list); /* Point to Rx buffer via next RBD in circular buffer */ - rxq->bd[rxq->write] = iwl_pcie_dma_addr2rbd_ptr(rxb->page_dma); + bd[rxq->write] = iwl_pcie_dma_addr2rbd_ptr(rxb->page_dma); rxq->queue[rxq->write] = rxb; rxq->write = (rxq->write + 1) & RX_QUEUE_MASK; rxq->free_count--; @@@ -272,7 -326,7 +326,7 @@@ * Increment device's write pointer in multiples of 8. */ if (rxq->write_actual != (rxq->write & ~0x7)) { spin_lock(&rxq->lock); - iwl_pcie_rxq_inc_wr_ptr(trans); + iwl_pcie_rxq_inc_wr_ptr(trans, rxq); spin_unlock(&rxq->lock); } } @@@ -285,13 -339,9 +339,9 @@@ static struct page *iwl_pcie_rx_alloc_p gfp_t priority) { struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans); - struct iwl_rxq *rxq = &trans_pcie->rxq; struct page *page; gfp_t gfp_mask = priority; - if (rxq->free_count > RX_LOW_WATERMARK) - gfp_mask |= __GFP_NOWARN; - if (trans_pcie->rx_page_order > 0) gfp_mask |= __GFP_COMP; @@@ -301,16 -351,13 +351,13 @@@ if (net_ratelimit()) IWL_DEBUG_INFO(trans, "alloc_pages failed, order: %d\n", trans_pcie->rx_page_order); - /* Issue an error if the hardware has consumed more than half - * of its free buffer list and we don't have enough - * pre-allocated buffers. + /* + * Issue an error if we don't have enough pre-allocated + * buffers. ` */ - if (rxq->free_count <= RX_LOW_WATERMARK && - iwl_rxq_space(rxq) > (RX_QUEUE_SIZE / 2) && - net_ratelimit()) + if (!(gfp_mask & __GFP_NOWARN) && net_ratelimit()) IWL_CRIT(trans, - "Failed to alloc_pages with GFP_KERNEL. Only %u free buffers remaining.\n", - rxq->free_count); + "Failed to alloc_pages\n"); return NULL; } return page; @@@ -325,10 -372,10 +372,10 @@@ * iwl_pcie_rxq_restock. The latter function will update the HW to use the newly * allocated buffers. */ - static void iwl_pcie_rxq_alloc_rbs(struct iwl_trans *trans, gfp_t priority) + static void iwl_pcie_rxq_alloc_rbs(struct iwl_trans *trans, gfp_t priority, + struct iwl_rxq *rxq) { struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans); - struct iwl_rxq *rxq = &trans_pcie->rxq; struct iwl_rx_mem_buffer *rxb; struct page *page; @@@ -372,10 -419,6 +419,6 @@@ __free_pages(page, trans_pcie->rx_page_order); return; } - /* dma address must be no more than 36 bits */ - BUG_ON(rxb->page_dma & ~DMA_BIT_MASK(36)); - /* and also 256 byte aligned! */ - BUG_ON(rxb->page_dma & DMA_BIT_MASK(8)); spin_lock(&rxq->lock); @@@ -386,41 -429,24 +429,24 @@@ } } - static void iwl_pcie_rxq_free_rbs(struct iwl_trans *trans) + static void iwl_pcie_free_rbs_pool(struct iwl_trans *trans) { struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans); - struct iwl_rxq *rxq = &trans_pcie->rxq; int i; - lockdep_assert_held(&rxq->lock); - - for (i = 0; i < RX_QUEUE_SIZE; i++) { - if (!rxq->pool[i].page) + for (i = 0; i < MQ_RX_POOL_SIZE; i++) { + if (!trans_pcie->rx_pool[i].page) continue; - dma_unmap_page(trans->dev, rxq->pool[i].page_dma, + dma_unmap_page(trans->dev, trans_pcie->rx_pool[i].page_dma, PAGE_SIZE << trans_pcie->rx_page_order, DMA_FROM_DEVICE); - __free_pages(rxq->pool[i].page, trans_pcie->rx_page_order); - rxq->pool[i].page = NULL; + __free_pages(trans_pcie->rx_pool[i].page, + trans_pcie->rx_page_order); + trans_pcie->rx_pool[i].page = NULL; } } /* - * iwl_pcie_rx_replenish - Move all used buffers from rx_used to rx_free - * - * When moving to rx_free an page is allocated for the slot. - * - * Also restock the Rx queue via iwl_pcie_rxq_restock. - * This is called only during initialization - */ - static void iwl_pcie_rx_replenish(struct iwl_trans *trans) - { - iwl_pcie_rxq_alloc_rbs(trans, GFP_KERNEL); - - iwl_pcie_rxq_restock(trans); - } - - /* * iwl_pcie_rx_allocator - Allocates pages in the background for RX queues * * Allocates for each received request 8 pages @@@ -444,6 -470,11 +470,11 @@@ static void iwl_pcie_rx_allocator(struc while (pending) { int i; struct list_head local_allocated; + gfp_t gfp_mask = GFP_KERNEL; + + /* Do not post a warning if there are only a few requests */ + if (pending < RX_PENDING_WATERMARK) + gfp_mask |= __GFP_NOWARN; INIT_LIST_HEAD(&local_allocated); @@@ -463,7 -494,7 +494,7 @@@ BUG_ON(rxb->page); /* Alloc a new receive buffer */ - page = iwl_pcie_rx_alloc_page(trans, GFP_KERNEL); + page = iwl_pcie_rx_alloc_page(trans, gfp_mask); if (!page) continue; rxb->page = page; @@@ -477,10 -508,6 +508,6 @@@ __free_pages(page, trans_pcie->rx_page_order); continue; } - /* dma address must be no more than 36 bits */ - BUG_ON(rxb->page_dma & ~DMA_BIT_MASK(36)); - /* and also 256 byte aligned! */ - BUG_ON(rxb->page_dma & DMA_BIT_MASK(8)); /* move the allocated entry to the out list */ list_move(&rxb->list, &local_allocated); @@@ -561,38 -588,83 +588,83 @@@ static void iwl_pcie_rx_allocator_work( static int iwl_pcie_rx_alloc(struct iwl_trans *trans) { struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans); - struct iwl_rxq *rxq = &trans_pcie->rxq; struct iwl_rb_allocator *rba = &trans_pcie->rba; struct device *dev = trans->dev; + int i; + int free_size = trans->cfg->mq_rx_supported ? sizeof(__le64) : + sizeof(__le32); + + if (WARN_ON(trans_pcie->rxq)) + return -EINVAL; - memset(&trans_pcie->rxq, 0, sizeof(trans_pcie->rxq)); + trans_pcie->rxq = kcalloc(trans->num_rx_queues, sizeof(struct iwl_rxq), + GFP_KERNEL); + if (!trans_pcie->rxq) + return -EINVAL; - spin_lock_init(&rxq->lock); spin_lock_init(&rba->lock); - if (WARN_ON(rxq->bd || rxq->rb_stts)) - return -EINVAL; + for (i = 0; i < trans->num_rx_queues; i++) { + struct iwl_rxq *rxq = &trans_pcie->rxq[i]; - /* Allocate the circular buffer of Read Buffer Descriptors (RBDs) */ - rxq->bd = dma_zalloc_coherent(dev, sizeof(__le32) * RX_QUEUE_SIZE, - &rxq->bd_dma, GFP_KERNEL); - if (!rxq->bd) - goto err_bd; + spin_lock_init(&rxq->lock); + if (trans->cfg->mq_rx_supported) + rxq->queue_size = MQ_RX_TABLE_SIZE; + else + rxq->queue_size = RX_QUEUE_SIZE; - /*Allocate the driver's pointer to receive buffer status */ - rxq->rb_stts = dma_zalloc_coherent(dev, sizeof(*rxq->rb_stts), - &rxq->rb_stts_dma, GFP_KERNEL); - if (!rxq->rb_stts) - goto err_rb_stts; + /* + * Allocate the circular buffer of Read Buffer Descriptors + * (RBDs) + */ + rxq->bd = dma_zalloc_coherent(dev, + free_size * rxq->queue_size, + &rxq->bd_dma, GFP_KERNEL); + if (!rxq->bd) + goto err; + + if (trans->cfg->mq_rx_supported) { + rxq->used_bd = dma_zalloc_coherent(dev, + sizeof(__le32) * + rxq->queue_size, + &rxq->used_bd_dma, + GFP_KERNEL); + if (!rxq->used_bd) + goto err; + } + /*Allocate the driver's pointer to receive buffer status */ + rxq->rb_stts = dma_zalloc_coherent(dev, sizeof(*rxq->rb_stts), + &rxq->rb_stts_dma, + GFP_KERNEL); + if (!rxq->rb_stts) + goto err; + } return 0; - err_rb_stts: - dma_free_coherent(dev, sizeof(__le32) * RX_QUEUE_SIZE, - rxq->bd, rxq->bd_dma); - rxq->bd_dma = 0; - rxq->bd = NULL; - err_bd: + err: + for (i = 0; i < trans->num_rx_queues; i++) { + struct iwl_rxq *rxq = &trans_pcie->rxq[i]; + + if (rxq->bd) + dma_free_coherent(dev, free_size * rxq->queue_size, + rxq->bd, rxq->bd_dma); + rxq->bd_dma = 0; + rxq->bd = NULL; + + if (rxq->rb_stts) + dma_free_coherent(trans->dev, + sizeof(struct iwl_rb_status), + rxq->rb_stts, rxq->rb_stts_dma); + + if (rxq->used_bd) + dma_free_coherent(dev, sizeof(__le32) * rxq->queue_size, + rxq->used_bd, rxq->used_bd_dma); + rxq->used_bd_dma = 0; + rxq->used_bd = NULL; + } + kfree(trans_pcie->rxq); + return -ENOMEM; } @@@ -659,65 -731,103 +731,103 @@@ static void iwl_pcie_rx_hw_init(struct iwl_set_bit(trans, CSR_INT_COALESCING, IWL_HOST_INT_OPER_MODE); } - static void iwl_pcie_rx_init_rxb_lists(struct iwl_rxq *rxq) + static void iwl_pcie_rx_mq_hw_init(struct iwl_trans *trans) { + struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans); + u32 rb_size, enabled = 0; int i; - lockdep_assert_held(&rxq->lock); - - INIT_LIST_HEAD(&rxq->rx_free); - INIT_LIST_HEAD(&rxq->rx_used); - rxq->free_count = 0; - rxq->used_count = 0; + switch (trans_pcie->rx_buf_size) { + case IWL_AMSDU_4K: + rb_size = RFH_RXF_DMA_RB_SIZE_4K; + break; + case IWL_AMSDU_8K: + rb_size = RFH_RXF_DMA_RB_SIZE_8K; + break; + case IWL_AMSDU_12K: + rb_size = RFH_RXF_DMA_RB_SIZE_12K; + break; + default: + WARN_ON(1); + rb_size = RFH_RXF_DMA_RB_SIZE_4K; + } - for (i = 0; i < RX_QUEUE_SIZE; i++) - list_add(&rxq->pool[i].list, &rxq->rx_used); - } + /* Stop Rx DMA */ + iwl_write_prph(trans, RFH_RXF_DMA_CFG, 0); + /* disable free amd used rx queue operation */ + iwl_write_prph(trans, RFH_RXF_RXQ_ACTIVE, 0); + + for (i = 0; i < trans->num_rx_queues; i++) { + /* Tell device where to find RBD free table in DRAM */ + iwl_pcie_write_prph_64(trans, RFH_Q_FRBDCB_BA_LSB(i), + (u64)(trans_pcie->rxq[i].bd_dma)); + /* Tell device where to find RBD used table in DRAM */ + iwl_pcie_write_prph_64(trans, RFH_Q_URBDCB_BA_LSB(i), + (u64)(trans_pcie->rxq[i].used_bd_dma)); + /* Tell device where in DRAM to update its Rx status */ + iwl_pcie_write_prph_64(trans, RFH_Q_URBD_STTS_WPTR_LSB(i), + trans_pcie->rxq[i].rb_stts_dma); + /* Reset device indice tables */ + iwl_write_prph(trans, RFH_Q_FRBDCB_WIDX(i), 0); + iwl_write_prph(trans, RFH_Q_FRBDCB_RIDX(i), 0); + iwl_write_prph(trans, RFH_Q_URBDCB_WIDX(i), 0); + + enabled |= BIT(i) | BIT(i + 16); + } - static void iwl_pcie_rx_init_rba(struct iwl_rb_allocator *rba) - { - int i; + /* restock default queue */ + iwl_pcie_rxq_mq_restock(trans, &trans_pcie->rxq[0]); - lockdep_assert_held(&rba->lock); + /* + * Enable Rx DMA + * Single frame mode + * Rx buffer size 4 or 8k or 12k + * Min RB size 4 or 8 + * 512 RBDs + */ + iwl_write_prph(trans, RFH_RXF_DMA_CFG, + RFH_DMA_EN_ENABLE_VAL | + rb_size | RFH_RXF_DMA_SINGLE_FRAME_MASK | + RFH_RXF_DMA_MIN_RB_4_8 | + RFH_RXF_DMA_RBDCB_SIZE_512); - INIT_LIST_HEAD(&rba->rbd_allocated); - INIT_LIST_HEAD(&rba->rbd_empty); + iwl_write_prph(trans, RFH_GEN_CFG, RFH_GEN_CFG_RFH_DMA_SNOOP | + RFH_GEN_CFG_SERVICE_DMA_SNOOP); + iwl_write_prph(trans, RFH_RXF_RXQ_ACTIVE, enabled); - for (i = 0; i < RX_POOL_SIZE; i++) - list_add(&rba->pool[i].list, &rba->rbd_empty); + /* Set interrupt coalescing timer to default (2048 usecs) */ + iwl_write8(trans, CSR_INT_COALESCING, IWL_HOST_INT_TIMEOUT_DEF); } - static void iwl_pcie_rx_free_rba(struct iwl_trans *trans) + static void iwl_pcie_rx_init_rxb_lists(struct iwl_rxq *rxq) { - struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans); - struct iwl_rb_allocator *rba = &trans_pcie->rba; - int i; + lockdep_assert_held(&rxq->lock); - lockdep_assert_held(&rba->lock); + INIT_LIST_HEAD(&rxq->rx_free); + INIT_LIST_HEAD(&rxq->rx_used); + rxq->free_count = 0; + rxq->used_count = 0; + } - for (i = 0; i < RX_POOL_SIZE; i++) { - if (!rba->pool[i].page) - continue; - dma_unmap_page(trans->dev, rba->pool[i].page_dma, - PAGE_SIZE << trans_pcie->rx_page_order, - DMA_FROM_DEVICE); - __free_pages(rba->pool[i].page, trans_pcie->rx_page_order); - rba->pool[i].page = NULL; - } + static int iwl_pcie_dummy_napi_poll(struct napi_struct *napi, int budget) + { + WARN_ON(1); + return 0; } int iwl_pcie_rx_init(struct iwl_trans *trans) { struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans); - struct iwl_rxq *rxq = &trans_pcie->rxq; + struct iwl_rxq *def_rxq; struct iwl_rb_allocator *rba = &trans_pcie->rba; - int i, err; + int i, err, num_rbds, allocator_pool_size; - if (!rxq->bd) { + if (!trans_pcie->rxq) { err = iwl_pcie_rx_alloc(trans); if (err) return err; } + def_rxq = trans_pcie->rxq; if (!rba->alloc_wq) rba->alloc_wq = alloc_workqueue("rb_allocator", WQ_HIGHPRI | WQ_UNBOUND, 1); @@@ -726,34 -836,68 +836,68 @@@ spin_lock(&rba->lock); atomic_set(&rba->req_pending, 0); atomic_set(&rba->req_ready, 0); - /* free all first - we might be reconfigured for a different size */ - iwl_pcie_rx_free_rba(trans); - iwl_pcie_rx_init_rba(rba); + INIT_LIST_HEAD(&rba->rbd_allocated); + INIT_LIST_HEAD(&rba->rbd_empty); spin_unlock(&rba->lock); - spin_lock(&rxq->lock); - /* free all first - we might be reconfigured for a different size */ - iwl_pcie_rxq_free_rbs(trans); - iwl_pcie_rx_init_rxb_lists(rxq); + iwl_pcie_free_rbs_pool(trans); for (i = 0; i < RX_QUEUE_SIZE; i++) - rxq->queue[i] = NULL; + def_rxq->queue[i] = NULL; - /* Set us so that we have processed and used all buffers, but have - * not restocked the Rx queue with fresh buffers */ - rxq->read = rxq->write = 0; - rxq->write_actual = 0; - memset(rxq->rb_stts, 0, sizeof(*rxq->rb_stts)); - spin_unlock(&rxq->lock); + for (i = 0; i < trans->num_rx_queues; i++) { + struct iwl_rxq *rxq = &trans_pcie->rxq[i]; - iwl_pcie_rx_replenish(trans); + rxq->id = i; - iwl_pcie_rx_hw_init(trans, rxq); + spin_lock(&rxq->lock); + /* + * Set read write pointer to reflect that we have processed + * and used all buffers, but have not restocked the Rx queue + * with fresh buffers + */ + rxq->read = 0; + rxq->write = 0; + rxq->write_actual = 0; + memset(rxq->rb_stts, 0, sizeof(*rxq->rb_stts)); - spin_lock(&rxq->lock); - iwl_pcie_rxq_inc_wr_ptr(trans); - spin_unlock(&rxq->lock); + iwl_pcie_rx_init_rxb_lists(rxq); + + if (!rxq->napi.poll) + netif_napi_add(&trans_pcie->napi_dev, &rxq->napi, + iwl_pcie_dummy_napi_poll, 64); + + spin_unlock(&rxq->lock); + } + + /* move the pool to the default queue and allocator ownerships */ + num_rbds = trans->cfg->mq_rx_supported ? + MQ_RX_POOL_SIZE : RX_QUEUE_SIZE; + allocator_pool_size = trans->num_rx_queues * + (RX_CLAIM_REQ_ALLOC - RX_POST_REQ_ALLOC); + for (i = 0; i < num_rbds; i++) { + struct iwl_rx_mem_buffer *rxb = &trans_pcie->rx_pool[i]; + + if (i < allocator_pool_size) + list_add(&rxb->list, &rba->rbd_empty); + else + list_add(&rxb->list, &def_rxq->rx_used); + trans_pcie->global_table[i] = rxb; + rxb->vid = (u16)i; + } + + iwl_pcie_rxq_alloc_rbs(trans, GFP_KERNEL, def_rxq); + if (trans->cfg->mq_rx_supported) { + iwl_pcie_rx_mq_hw_init(trans); + } else { + iwl_pcie_rxq_restock(trans, def_rxq); + iwl_pcie_rx_hw_init(trans, def_rxq); + } + + spin_lock(&def_rxq->lock); + iwl_pcie_rxq_inc_wr_ptr(trans, def_rxq); + spin_unlock(&def_rxq->lock); return 0; } @@@ -761,12 -905,16 +905,16 @@@ void iwl_pcie_rx_free(struct iwl_trans *trans) { struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans); - struct iwl_rxq *rxq = &trans_pcie->rxq; struct iwl_rb_allocator *rba = &trans_pcie->rba; + int free_size = trans->cfg->mq_rx_supported ? sizeof(__le64) : + sizeof(__le32); + int i; - /*if rxq->bd is NULL, it means that nothing has been allocated, - * exit now */ - if (!rxq->bd) { + /* + * if rxq is NULL, it means that nothing has been allocated, + * exit now + */ + if (!trans_pcie->rxq) { IWL_DEBUG_INFO(trans, "Free NULL rx context\n"); return; } @@@ -777,27 -925,37 +925,37 @@@ rba->alloc_wq = NULL; } - spin_lock(&rba->lock); - iwl_pcie_rx_free_rba(trans); - spin_unlock(&rba->lock); - - spin_lock(&rxq->lock); - iwl_pcie_rxq_free_rbs(trans); - spin_unlock(&rxq->lock); - - dma_free_coherent(trans->dev, sizeof(__le32) * RX_QUEUE_SIZE, - rxq->bd, rxq->bd_dma); - rxq->bd_dma = 0; - rxq->bd = NULL; - - if (rxq->rb_stts) - dma_free_coherent(trans->dev, - sizeof(struct iwl_rb_status), - rxq->rb_stts, rxq->rb_stts_dma); - else - IWL_DEBUG_INFO(trans, "Free rxq->rb_stts which is NULL\n"); - rxq->rb_stts_dma = 0; - rxq->rb_stts = NULL; + iwl_pcie_free_rbs_pool(trans); + + for (i = 0; i < trans->num_rx_queues; i++) { + struct iwl_rxq *rxq = &trans_pcie->rxq[i]; + + if (rxq->bd) + dma_free_coherent(trans->dev, + free_size * rxq->queue_size, + rxq->bd, rxq->bd_dma); + rxq->bd_dma = 0; + rxq->bd = NULL; + + if (rxq->rb_stts) + dma_free_coherent(trans->dev, + sizeof(struct iwl_rb_status), + rxq->rb_stts, rxq->rb_stts_dma); + else + IWL_DEBUG_INFO(trans, + "Free rxq->rb_stts which is NULL\n"); + + if (rxq->used_bd) + dma_free_coherent(trans->dev, + sizeof(__le32) * rxq->queue_size, + rxq->used_bd, rxq->used_bd_dma); + rxq->used_bd_dma = 0; + rxq->used_bd = NULL; + + if (rxq->napi.poll) + netif_napi_del(&rxq->napi); + } + kfree(trans_pcie->rxq); } /* @@@ -841,11 -999,11 +999,11 @@@ static void iwl_pcie_rx_reuse_rbd(struc } static void iwl_pcie_rx_handle_rb(struct iwl_trans *trans, + struct iwl_rxq *rxq, struct iwl_rx_mem_buffer *rxb, bool emergency) { struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans); - struct iwl_rxq *rxq = &trans_pcie->rxq; struct iwl_txq *txq = &trans_pcie->txq[trans_pcie->cmd_queue]; bool page_stolen = false; int max_len = PAGE_SIZE << trans_pcie->rx_page_order; @@@ -911,7 -1069,12 +1069,12 @@@ index = SEQ_TO_INDEX(sequence); cmd_index = get_cmd_index(&txq->q, index); - iwl_op_mode_rx(trans->op_mode, &trans_pcie->napi, &rxcb); + if (rxq->id == 0) + iwl_op_mode_rx(trans->op_mode, &rxq->napi, + &rxcb); + else + iwl_op_mode_rx_rss(trans->op_mode, &rxq->napi, + &rxcb, rxq->id); if (reclaim) { kzfree(txq->entries[cmd_index].free_buf); @@@ -975,7 -1138,7 +1138,7 @@@ static void iwl_pcie_rx_handle(struct iwl_trans *trans) { struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans); - struct iwl_rxq *rxq = &trans_pcie->rxq; + struct iwl_rxq *rxq = &trans_pcie->rxq[0]; u32 r, i, j, count = 0; bool emergency = false; @@@ -993,16 -1156,26 +1156,26 @@@ restart while (i != r) { struct iwl_rx_mem_buffer *rxb; - if (unlikely(rxq->used_count == RX_QUEUE_SIZE / 2)) + if (unlikely(rxq->used_count == rxq->queue_size / 2)) emergency = true; - rxb = rxq->queue[i]; - rxq->queue[i] = NULL; + if (trans->cfg->mq_rx_supported) { + /* + * used_bd is a 32 bit but only 12 are used to retrieve + * the vid + */ + u16 vid = (u16)le32_to_cpu(rxq->used_bd[i]); + + rxb = trans_pcie->global_table[vid]; + } else { + rxb = rxq->queue[i]; + rxq->queue[i] = NULL; + } IWL_DEBUG_RX(trans, "rxbuf: HW = %d, SW = %d\n", r, i); - iwl_pcie_rx_handle_rb(trans, rxb, emergency); + iwl_pcie_rx_handle_rb(trans, rxq, rxb, emergency); - i = (i + 1) & RX_QUEUE_MASK; + i = (i + 1) & (rxq->queue_size - 1); /* If we have RX_CLAIM_REQ_ALLOC released rx buffers - * try to claim the pre-allocated buffers from the allocator */ @@@ -1040,10 -1213,10 +1213,10 @@@ count++; if (count == 8) { count = 0; - if (rxq->used_count < RX_QUEUE_SIZE / 3) + if (rxq->used_count < rxq->queue_size / 3) emergency = false; spin_unlock(&rxq->lock); - iwl_pcie_rxq_alloc_rbs(trans, GFP_ATOMIC); + iwl_pcie_rxq_alloc_rbs(trans, GFP_ATOMIC, rxq); spin_lock(&rxq->lock); } } @@@ -1055,7 -1228,10 +1228,10 @@@ if (rxq->free_count >= RX_CLAIM_REQ_ALLOC) { rxq->read = i; spin_unlock(&rxq->lock); - iwl_pcie_rxq_restock(trans); + if (trans->cfg->mq_rx_supported) + iwl_pcie_rxq_mq_restock(trans, rxq); + else + iwl_pcie_rxq_restock(trans, rxq); goto restart; } } @@@ -1077,10 -1253,10 +1253,10 @@@ * will be restocked by the next call of iwl_pcie_rxq_restock. */ if (unlikely(emergency && count)) - iwl_pcie_rxq_alloc_rbs(trans, GFP_ATOMIC); + iwl_pcie_rxq_alloc_rbs(trans, GFP_ATOMIC, rxq); - if (trans_pcie->napi.poll) - napi_gro_flush(&trans_pcie->napi, false); + if (rxq->napi.poll) + napi_gro_flush(&rxq->napi, false); } /* @@@ -1438,11 -1614,9 +1614,11 @@@ irqreturn_t iwl_pcie_irq_handler(int ir inta & ~trans_pcie->inta_mask); } - /* Re-enable all interrupts */ - /* only Re-enable if disabled by irq */ - if (test_bit(STATUS_INT_ENABLED, &trans->status)) + /* we are loading the firmware, enable FH_TX interrupt only */ + if (handled & CSR_INT_BIT_FH_TX) + iwl_enable_fw_load_int(trans); + /* only Re-enable all interrupt if disabled by irq */ + else if (test_bit(STATUS_INT_ENABLED, &trans->status)) iwl_enable_interrupts(trans); /* Re-enable RF_KILL if it occurred */ else if (handled & CSR_INT_BIT_RF_KILL) diff --combined drivers/net/wireless/intel/iwlwifi/pcie/trans.c index 5a854c6,b796952..58591ca --- a/drivers/net/wireless/intel/iwlwifi/pcie/trans.c +++ b/drivers/net/wireless/intel/iwlwifi/pcie/trans.c @@@ -72,6 -72,7 +72,7 @@@ #include <linux/bitops.h> #include <linux/gfp.h> #include <linux/vmalloc.h> + #include <linux/pm_runtime.h> #include "iwl-drv.h" #include "iwl-trans.h" @@@ -1021,6 -1022,82 +1022,6 @@@ static int iwl_pcie_load_given_ucode_80 &first_ucode_section); } -static int iwl_trans_pcie_start_fw(struct iwl_trans *trans, - const struct fw_img *fw, bool run_in_rfkill) -{ - struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans); - bool hw_rfkill; - int ret; - - mutex_lock(&trans_pcie->mutex); - - /* Someone called stop_device, don't try to start_fw */ - if (trans_pcie->is_down) { - IWL_WARN(trans, - "Can't start_fw since the HW hasn't been started\n"); - ret = EIO; - goto out; - } - - /* This may fail if AMT took ownership of the device */ - if (iwl_pcie_prepare_card_hw(trans)) { - IWL_WARN(trans, "Exit HW not ready\n"); - ret = -EIO; - goto out; - } - - iwl_enable_rfkill_int(trans); - - /* If platform's RF_KILL switch is NOT set to KILL */ - hw_rfkill = iwl_is_rfkill_set(trans); - if (hw_rfkill) - set_bit(STATUS_RFKILL, &trans->status); - else - clear_bit(STATUS_RFKILL, &trans->status); - iwl_trans_pcie_rf_kill(trans, hw_rfkill); - if (hw_rfkill && !run_in_rfkill) { - ret = -ERFKILL; - goto out; - } - - iwl_write32(trans, CSR_INT, 0xFFFFFFFF); - - ret = iwl_pcie_nic_init(trans); - if (ret) { - IWL_ERR(trans, "Unable to init nic\n"); - goto out; - } - - /* make sure rfkill handshake bits are cleared */ - iwl_write32(trans, CSR_UCODE_DRV_GP1_CLR, CSR_UCODE_SW_BIT_RFKILL); - iwl_write32(trans, CSR_UCODE_DRV_GP1_CLR, - CSR_UCODE_DRV_GP1_BIT_CMD_BLOCKED); - - /* clear (again), then enable host interrupts */ - iwl_write32(trans, CSR_INT, 0xFFFFFFFF); - iwl_enable_interrupts(trans); - - /* really make sure rfkill handshake bits are cleared */ - iwl_write32(trans, CSR_UCODE_DRV_GP1_CLR, CSR_UCODE_SW_BIT_RFKILL); - iwl_write32(trans, CSR_UCODE_DRV_GP1_CLR, CSR_UCODE_SW_BIT_RFKILL); - - /* Load the given image to the HW */ - if (trans->cfg->device_family == IWL_DEVICE_FAMILY_8000) - ret = iwl_pcie_load_given_ucode_8000(trans, fw); - else - ret = iwl_pcie_load_given_ucode(trans, fw); - -out: - mutex_unlock(&trans_pcie->mutex); - return ret; -} - -static void iwl_trans_pcie_fw_alive(struct iwl_trans *trans, u32 scd_addr) -{ - iwl_pcie_reset_ict(trans); - iwl_pcie_tx_start(trans, scd_addr); -} - static void _iwl_trans_pcie_stop_device(struct iwl_trans *trans, bool low_power) { struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans); @@@ -1051,8 -1128,7 +1052,8 @@@ * already dead. */ if (test_and_clear_bit(STATUS_DEVICE_ENABLED, &trans->status)) { - IWL_DEBUG_INFO(trans, "DEVICE_ENABLED bit was set and is now cleared\n"); + IWL_DEBUG_INFO(trans, + "DEVICE_ENABLED bit was set and is now cleared\n"); iwl_pcie_tx_stop(trans); iwl_pcie_rx_stop(trans); @@@ -1086,6 -1162,7 +1087,6 @@@ iwl_disable_interrupts(trans); spin_unlock(&trans_pcie->irq_lock); - /* clear all status bits */ clear_bit(STATUS_SYNC_HCMD_ACTIVE, &trans->status); clear_bit(STATUS_INT_ENABLED, &trans->status); @@@ -1118,116 -1195,10 +1119,116 @@@ if (hw_rfkill != was_hw_rfkill) iwl_trans_pcie_rf_kill(trans, hw_rfkill); - /* re-take ownership to prevent other users from stealing the deivce */ + /* re-take ownership to prevent other users from stealing the device */ iwl_pcie_prepare_card_hw(trans); } +static int iwl_trans_pcie_start_fw(struct iwl_trans *trans, + const struct fw_img *fw, bool run_in_rfkill) +{ + struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans); + bool hw_rfkill; + int ret; + + /* This may fail if AMT took ownership of the device */ + if (iwl_pcie_prepare_card_hw(trans)) { + IWL_WARN(trans, "Exit HW not ready\n"); + ret = -EIO; + goto out; + } + + iwl_enable_rfkill_int(trans); + + iwl_write32(trans, CSR_INT, 0xFFFFFFFF); + + /* + * We enabled the RF-Kill interrupt and the handler may very + * well be running. Disable the interrupts to make sure no other + * interrupt can be fired. + */ + iwl_disable_interrupts(trans); + + /* Make sure it finished running */ + synchronize_irq(trans_pcie->pci_dev->irq); + + mutex_lock(&trans_pcie->mutex); + + /* If platform's RF_KILL switch is NOT set to KILL */ + hw_rfkill = iwl_is_rfkill_set(trans); + if (hw_rfkill) + set_bit(STATUS_RFKILL, &trans->status); + else + clear_bit(STATUS_RFKILL, &trans->status); + iwl_trans_pcie_rf_kill(trans, hw_rfkill); + if (hw_rfkill && !run_in_rfkill) { + ret = -ERFKILL; + goto out; + } + + /* Someone called stop_device, don't try to start_fw */ + if (trans_pcie->is_down) { + IWL_WARN(trans, + "Can't start_fw since the HW hasn't been started\n"); + ret = -EIO; + goto out; + } + + /* make sure rfkill handshake bits are cleared */ + iwl_write32(trans, CSR_UCODE_DRV_GP1_CLR, CSR_UCODE_SW_BIT_RFKILL); + iwl_write32(trans, CSR_UCODE_DRV_GP1_CLR, + CSR_UCODE_DRV_GP1_BIT_CMD_BLOCKED); + + /* clear (again), then enable host interrupts */ + iwl_write32(trans, CSR_INT, 0xFFFFFFFF); + + ret = iwl_pcie_nic_init(trans); + if (ret) { + IWL_ERR(trans, "Unable to init nic\n"); + goto out; + } + + /* + * Now, we load the firmware and don't want to be interrupted, even + * by the RF-Kill interrupt (hence mask all the interrupt besides the + * FH_TX interrupt which is needed to load the firmware). If the + * RF-Kill switch is toggled, we will find out after having loaded + * the firmware and return the proper value to the caller. + */ + iwl_enable_fw_load_int(trans); + + /* really make sure rfkill handshake bits are cleared */ + iwl_write32(trans, CSR_UCODE_DRV_GP1_CLR, CSR_UCODE_SW_BIT_RFKILL); + iwl_write32(trans, CSR_UCODE_DRV_GP1_CLR, CSR_UCODE_SW_BIT_RFKILL); + + /* Load the given image to the HW */ + if (trans->cfg->device_family == IWL_DEVICE_FAMILY_8000) + ret = iwl_pcie_load_given_ucode_8000(trans, fw); + else + ret = iwl_pcie_load_given_ucode(trans, fw); + iwl_enable_interrupts(trans); + + /* re-check RF-Kill state since we may have missed the interrupt */ + hw_rfkill = iwl_is_rfkill_set(trans); + if (hw_rfkill) + set_bit(STATUS_RFKILL, &trans->status); + else + clear_bit(STATUS_RFKILL, &trans->status); + + iwl_trans_pcie_rf_kill(trans, hw_rfkill); + if (hw_rfkill && !run_in_rfkill) + ret = -ERFKILL; + +out: + mutex_unlock(&trans_pcie->mutex); + return ret; +} + +static void iwl_trans_pcie_fw_alive(struct iwl_trans *trans, u32 scd_addr) +{ + iwl_pcie_reset_ict(trans); + iwl_pcie_tx_start(trans, scd_addr); +} + static void iwl_trans_pcie_stop_device(struct iwl_trans *trans, bool low_power) { struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans); @@@ -1248,11 -1219,12 +1249,12 @@@ void iwl_trans_pcie_rf_kill(struct iwl_ _iwl_trans_pcie_stop_device(trans, true); } - static void iwl_trans_pcie_d3_suspend(struct iwl_trans *trans, bool test) + static void iwl_trans_pcie_d3_suspend(struct iwl_trans *trans, bool test, + bool reset) { struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans); - if (trans->system_pm_mode == IWL_PLAT_PM_MODE_D0I3) { + if (!reset) { /* Enable persistence mode to avoid reset */ iwl_set_bit(trans, CSR_HW_IF_CONFIG_REG, CSR_HW_IF_CONFIG_REG_PERSIST_MODE); @@@ -1276,7 -1248,7 +1278,7 @@@ iwl_clear_bit(trans, CSR_GP_CNTRL, CSR_GP_CNTRL_REG_FLAG_INIT_DONE); - if (trans->system_pm_mode == IWL_PLAT_PM_MODE_D3) { + if (reset) { /* * reset TX queues -- some of their registers reset during S3 * so if we don't reset everything here the D3 image would try @@@ -1290,7 -1262,7 +1292,7 @@@ static int iwl_trans_pcie_d3_resume(struct iwl_trans *trans, enum iwl_d3_status *status, - bool test) + bool test, bool reset) { u32 val; int ret; @@@ -1325,7 -1297,7 +1327,7 @@@ iwl_pcie_set_pwr(trans, false); - if (trans->system_pm_mode == IWL_PLAT_PM_MODE_D0I3) { + if (!reset) { iwl_clear_bit(trans, CSR_GP_CNTRL, CSR_GP_CNTRL_REG_FLAG_MAC_ACCESS_REQ); } else { @@@ -1383,6 -1355,10 +1385,10 @@@ static int _iwl_trans_pcie_start_hw(str /* ... rfkill can call stop_device and set it false if needed */ iwl_trans_pcie_rf_kill(trans, hw_rfkill); + /* Make sure we sync here, because we'll need full access later */ + if (low_power) + pm_runtime_resume(trans->dev); + return 0; } @@@ -1452,12 -1428,6 +1458,6 @@@ static void iwl_trans_pcie_write_prph(s iwl_trans_pcie_write32(trans, HBUS_TARG_PRPH_WDAT, val); } - static int iwl_pcie_dummy_napi_poll(struct napi_struct *napi, int budget) - { - WARN_ON(1); - return 0; - } - static void iwl_trans_pcie_configure(struct iwl_trans *trans, const struct iwl_trans_config *trans_cfg) { @@@ -1494,11 -1464,8 +1494,8 @@@ * As this function may be called again in some corner cases don't * do anything if NAPI was already initialized. */ - if (!trans_pcie->napi.poll) { + if (trans_pcie->napi_dev.reg_state != NETREG_DUMMY) init_dummy_netdev(&trans_pcie->napi_dev); - netif_napi_add(&trans_pcie->napi_dev, &trans_pcie->napi, - iwl_pcie_dummy_napi_poll, 64); - } } void iwl_trans_pcie_free(struct iwl_trans *trans) @@@ -1506,6 -1473,9 +1503,9 @@@ struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans); int i; + /* TODO: check if this is really needed */ + pm_runtime_disable(trans->dev); + synchronize_irq(trans_pcie->pci_dev->irq); iwl_pcie_tx_free(trans); @@@ -1519,9 -1489,6 +1519,6 @@@ pci_release_regions(trans_pcie->pci_dev); pci_disable_device(trans_pcie->pci_dev); - if (trans_pcie->napi.poll) - netif_napi_del(&trans_pcie->napi); - iwl_pcie_free_fw_monitor(trans); for_each_possible_cpu(i) { @@@ -1861,6 -1828,7 +1858,7 @@@ void iwl_trans_pcie_ref(struct iwl_tran spin_lock_irqsave(&trans_pcie->ref_lock, flags); IWL_DEBUG_RPM(trans, "ref_counter: %d\n", trans_pcie->ref_count); trans_pcie->ref_count++; + pm_runtime_get(&trans_pcie->pci_dev->dev); spin_unlock_irqrestore(&trans_pcie->ref_lock, flags); } @@@ -1879,6 -1847,10 +1877,10 @@@ void iwl_trans_pcie_unref(struct iwl_tr return; } trans_pcie->ref_count--; + + pm_runtime_mark_last_busy(&trans_pcie->pci_dev->dev); + pm_runtime_put_autosuspend(&trans_pcie->pci_dev->dev); + spin_unlock_irqrestore(&trans_pcie->ref_lock, flags); } @@@ -2031,29 -2003,48 +2033,48 @@@ static ssize_t iwl_dbgfs_rx_queue_read( { struct iwl_trans *trans = file->private_data; struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans); - struct iwl_rxq *rxq = &trans_pcie->rxq; - char buf[256]; - int pos = 0; - const size_t bufsz = sizeof(buf); - - pos += scnprintf(buf + pos, bufsz - pos, "read: %u\n", - rxq->read); - pos += scnprintf(buf + pos, bufsz - pos, "write: %u\n", - rxq->write); - pos += scnprintf(buf + pos, bufsz - pos, "write_actual: %u\n", - rxq->write_actual); - pos += scnprintf(buf + pos, bufsz - pos, "need_update: %d\n", - rxq->need_update); - pos += scnprintf(buf + pos, bufsz - pos, "free_count: %u\n", - rxq->free_count); - if (rxq->rb_stts) { - pos += scnprintf(buf + pos, bufsz - pos, "closed_rb_num: %u\n", - le16_to_cpu(rxq->rb_stts->closed_rb_num) & 0x0FFF); - } else { - pos += scnprintf(buf + pos, bufsz - pos, - "closed_rb_num: Not Allocated\n"); + char *buf; + int pos = 0, i, ret; + size_t bufsz = sizeof(buf); + + bufsz = sizeof(char) * 121 * trans->num_rx_queues; + + if (!trans_pcie->rxq) + return -EAGAIN; + + buf = kzalloc(bufsz, GFP_KERNEL); + if (!buf) + return -ENOMEM; + + for (i = 0; i < trans->num_rx_queues && pos < bufsz; i++) { + struct iwl_rxq *rxq = &trans_pcie->rxq[i]; + + pos += scnprintf(buf + pos, bufsz - pos, "queue#: %2d\n", + i); + pos += scnprintf(buf + pos, bufsz - pos, "\tread: %u\n", + rxq->read); + pos += scnprintf(buf + pos, bufsz - pos, "\twrite: %u\n", + rxq->write); + pos += scnprintf(buf + pos, bufsz - pos, "\twrite_actual: %u\n", + rxq->write_actual); + pos += scnprintf(buf + pos, bufsz - pos, "\tneed_update: %2d\n", + rxq->need_update); + pos += scnprintf(buf + pos, bufsz - pos, "\tfree_count: %u\n", + rxq->free_count); + if (rxq->rb_stts) { + pos += scnprintf(buf + pos, bufsz - pos, + "\tclosed_rb_num: %u\n", + le16_to_cpu(rxq->rb_stts->closed_rb_num) & + 0x0FFF); + } else { + pos += scnprintf(buf + pos, bufsz - pos, + "\tclosed_rb_num: Not Allocated\n"); + } } - return simple_read_from_buffer(user_buf, count, ppos, buf, pos); + ret = simple_read_from_buffer(user_buf, count, ppos, buf, pos); + kfree(buf); + + return ret; } static ssize_t iwl_dbgfs_interrupt_read(struct file *file, @@@ -2218,7 -2209,8 +2239,8 @@@ static u32 iwl_trans_pcie_dump_rbs(stru { struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans); int max_len = PAGE_SIZE << trans_pcie->rx_page_order; - struct iwl_rxq *rxq = &trans_pcie->rxq; + /* Dump RBs is supported only for pre-9000 devices (1 queue) */ + struct iwl_rxq *rxq = &trans_pcie->rxq[0]; u32 i, r, j, rb_len = 0; spin_lock(&rxq->lock); @@@ -2413,7 -2405,8 +2435,8 @@@ static struct iwl_trans_dump_dat u32 len, num_rbs; u32 monitor_len; int i, ptr; - bool dump_rbs = test_bit(STATUS_FW_ERROR, &trans->status); + bool dump_rbs = test_bit(STATUS_FW_ERROR, &trans->status) && + !trans->cfg->mq_rx_supported; /* transport dump header */ len = sizeof(*dump_data); @@@ -2468,11 -2461,12 +2491,12 @@@ len += sizeof(*data) + (FH_MEM_UPPER_BOUND - FH_MEM_LOWER_BOUND); if (dump_rbs) { + /* Dump RBs is supported only for pre-9000 devices (1 queue) */ + struct iwl_rxq *rxq = &trans_pcie->rxq[0]; /* RBs */ - num_rbs = le16_to_cpu(ACCESS_ONCE( - trans_pcie->rxq.rb_stts->closed_rb_num)) + num_rbs = le16_to_cpu(ACCESS_ONCE(rxq->rb_stts->closed_rb_num)) & 0x0FFF; - num_rbs = (num_rbs - trans_pcie->rxq.read) & RX_QUEUE_MASK; + num_rbs = (num_rbs - rxq->read) & RX_QUEUE_MASK; len += num_rbs * (sizeof(*data) + sizeof(struct iwl_fw_error_dump_rb) + (PAGE_SIZE << trans_pcie->rx_page_order)); @@@ -2523,6 -2517,22 +2547,22 @@@ return dump_data; } + #ifdef CONFIG_PM_SLEEP + static int iwl_trans_pcie_suspend(struct iwl_trans *trans) + { + if (trans->runtime_pm_mode == IWL_PLAT_PM_MODE_D0I3) + return iwl_pci_fw_enter_d0i3(trans); + + return 0; + } + + static void iwl_trans_pcie_resume(struct iwl_trans *trans) + { + if (trans->runtime_pm_mode == IWL_PLAT_PM_MODE_D0I3) + iwl_pci_fw_exit_d0i3(trans); + } + #endif /* CONFIG_PM_SLEEP */ + static const struct iwl_trans_ops trans_ops_pcie = { .start_hw = iwl_trans_pcie_start_hw, .op_mode_leave = iwl_trans_pcie_op_mode_leave, @@@ -2533,6 -2543,11 +2573,11 @@@ .d3_suspend = iwl_trans_pcie_d3_suspend, .d3_resume = iwl_trans_pcie_d3_resume, + #ifdef CONFIG_PM_SLEEP + .suspend = iwl_trans_pcie_suspend, + .resume = iwl_trans_pcie_resume, + #endif /* CONFIG_PM_SLEEP */ + .send_cmd = iwl_trans_pcie_send_hcmd, .tx = iwl_trans_pcie_tx, @@@ -2571,7 -2586,7 +2616,7 @@@ struct iwl_trans *iwl_trans_pcie_alloc( struct iwl_trans_pcie *trans_pcie; struct iwl_trans *trans; u16 pci_cmd; - int ret; + int ret, addr_size; trans = iwl_trans_alloc(sizeof(struct iwl_trans_pcie), &pdev->dev, cfg, &trans_ops_pcie, 0); @@@ -2609,11 -2624,17 +2654,17 @@@ PCIE_LINK_STATE_CLKPM); } + if (cfg->mq_rx_supported) + addr_size = 64; + else + addr_size = 36; + pci_set_master(pdev); - ret = pci_set_dma_mask(pdev, DMA_BIT_MASK(36)); + ret = pci_set_dma_mask(pdev, DMA_BIT_MASK(addr_size)); if (!ret) - ret = pci_set_consistent_dma_mask(pdev, DMA_BIT_MASK(36)); + ret = pci_set_consistent_dma_mask(pdev, + DMA_BIT_MASK(addr_size)); if (ret) { ret = pci_set_dma_mask(pdev, DMA_BIT_MASK(32)); if (!ret) @@@ -2716,6 -2737,8 +2767,8 @@@ /* Initialize the wait queue for commands */ init_waitqueue_head(&trans_pcie->wait_command_queue); + init_waitqueue_head(&trans_pcie->d0i3_waitq); + ret = iwl_pcie_alloc_ict(trans); if (ret) goto out_pci_disable_msi; @@@ -2730,6 -2753,12 +2783,12 @@@ trans_pcie->inta_mask = CSR_INI_SET_MASK; + #ifdef CONFIG_IWLWIFI_PCIE_RTPM + trans->runtime_pm_mode = IWL_PLAT_PM_MODE_D0I3; + #else + trans->runtime_pm_mode = IWL_PLAT_PM_MODE_DISABLED; + #endif /* CONFIG_IWLWIFI_PCIE_RTPM */ + return trans; out_free_ict: diff --combined include/linux/netdevice.h index 5440b7b,47671ce0..e52077f --- a/include/linux/netdevice.h +++ b/include/linux/netdevice.h @@@ -51,6 -51,7 +51,7 @@@ #include <linux/neighbour.h> #include <uapi/linux/netdevice.h> #include <uapi/linux/if_bonding.h> + #include <uapi/linux/pkt_cls.h> struct netpoll_info; struct device; @@@ -778,6 -779,25 +779,25 @@@ static inline bool netdev_phys_item_id_ typedef u16 (*select_queue_fallback_t)(struct net_device *dev, struct sk_buff *skb); + /* These structures hold the attributes of qdisc and classifiers + * that are being passed to the netdevice through the setup_tc op. + */ + enum { + TC_SETUP_MQPRIO, + TC_SETUP_CLSU32, + }; + + struct tc_cls_u32_offload; + + struct tc_to_netdev { + unsigned int type; + union { + u8 tc; + struct tc_cls_u32_offload *cls_u32; + }; + }; + + /* * This structure defines the management hooks for network devices. * The following hooks can be defined; unless noted otherwise, they are @@@ -1150,7 -1170,10 +1170,10 @@@ struct net_device_ops int (*ndo_set_vf_rss_query_en)( struct net_device *dev, int vf, bool setting); - int (*ndo_setup_tc)(struct net_device *dev, u8 tc); + int (*ndo_setup_tc)(struct net_device *dev, + u32 handle, + __be16 protocol, + struct tc_to_netdev *tc); #if IS_ENABLED(CONFIG_FCOE) int (*ndo_fcoe_enable)(struct net_device *dev); int (*ndo_fcoe_disable)(struct net_device *dev); @@@ -1291,6 -1314,7 +1314,7 @@@ * @IFF_OPENVSWITCH: device is a Open vSwitch master * @IFF_L3MDEV_SLAVE: device is enslaved to an L3 master device * @IFF_TEAM: device is a team device + * @IFF_RXFH_CONFIGURED: device has had Rx Flow indirection table configured */ enum netdev_priv_flags { IFF_802_1Q_VLAN = 1<<0, @@@ -1318,6 -1342,7 +1342,7 @@@ IFF_OPENVSWITCH = 1<<22, IFF_L3MDEV_SLAVE = 1<<23, IFF_TEAM = 1<<24, + IFF_RXFH_CONFIGURED = 1<<25, }; #define IFF_802_1Q_VLAN IFF_802_1Q_VLAN @@@ -1345,6 -1370,7 +1370,7 @@@ #define IFF_OPENVSWITCH IFF_OPENVSWITCH #define IFF_L3MDEV_SLAVE IFF_L3MDEV_SLAVE #define IFF_TEAM IFF_TEAM + #define IFF_RXFH_CONFIGURED IFF_RXFH_CONFIGURED /** * struct net_device - The DEVICE structure. @@@ -1397,6 -1423,8 +1423,8 @@@ * do not use this in drivers * @tx_dropped: Dropped packets by core network, * do not use this in drivers + * @rx_nohandler: nohandler dropped packets by core network on + * inactive devices, do not use this in drivers * * @wireless_handlers: List of functions to handle Wireless Extensions, * instead of ioctl, @@@ -1611,6 -1639,7 +1639,7 @@@ struct net_device atomic_long_t rx_dropped; atomic_long_t tx_dropped; + atomic_long_t rx_nohandler; #ifdef CONFIG_WIRELESS_EXT const struct iw_handler_def * wireless_handlers; @@@ -3718,7 -3747,7 +3747,7 @@@ void *netdev_lower_get_next_private_rcu void *netdev_lower_get_next(struct net_device *dev, struct list_head **iter); #define netdev_for_each_lower_dev(dev, ldev, iter) \ - for (iter = &(dev)->adj_list.lower, \ + for (iter = (dev)->adj_list.lower.next, \ ldev = netdev_lower_get_next(dev, &(iter)); \ ldev; \ ldev = netdev_lower_get_next(dev, &(iter))) @@@ -3741,7 -3770,7 +3770,7 @@@ void netdev_lower_state_changed(struct /* RSS keys are 40 or 52 bytes long */ #define NETDEV_RSS_KEY_LEN 52 - extern u8 netdev_rss_key[NETDEV_RSS_KEY_LEN]; + extern u8 netdev_rss_key[NETDEV_RSS_KEY_LEN] __read_mostly; void netdev_rss_key_fill(void *buffer, size_t len); int dev_get_nest_level(struct net_device *dev, @@@ -4045,6 -4074,11 +4074,11 @@@ static inline bool netif_is_lag_port(co return netif_is_bond_slave(dev) || netif_is_team_port(dev); } + static inline bool netif_is_rxfh_configured(const struct net_device *dev) + { + return dev->priv_flags & IFF_RXFH_CONFIGURED; + } + /* This device needs to keep skb dst for qdisc enqueue or ndo_start_xmit() */ static inline void netif_keep_dst(struct net_device *dev) { diff --combined include/linux/skbuff.h index 4ce9ff7,6a57757..eab4f8f --- a/include/linux/skbuff.h +++ b/include/linux/skbuff.h @@@ -299,7 -299,6 +299,7 @@@ struct sk_buff #else #define MAX_SKB_FRAGS (65536/PAGE_SIZE + 1) #endif +extern int sysctl_max_skb_frags; typedef struct skb_frag_struct skb_frag_t; @@@ -2162,6 -2161,11 +2162,11 @@@ static inline int skb_checksum_start_of return skb->csum_start - skb_headroom(skb); } + static inline unsigned char *skb_checksum_start(const struct sk_buff *skb) + { + return skb->head + skb->csum_start; + } + static inline int skb_transport_offset(const struct sk_buff *skb) { return skb_transport_header(skb) - skb->data; @@@ -2400,6 -2404,10 +2405,10 @@@ static inline struct sk_buff *napi_allo { return __napi_alloc_skb(napi, length, GFP_ATOMIC); } + void napi_consume_skb(struct sk_buff *skb, int budget); + + void __kfree_skb_flush(void); + void __kfree_skb_defer(struct sk_buff *skb); /** * __dev_alloc_pages - allocate page for network Rx @@@ -2622,6 -2630,13 +2631,13 @@@ static inline int skb_clone_writable(co skb_headroom(skb) + len <= skb->hdr_len; } + static inline int skb_try_make_writable(struct sk_buff *skb, + unsigned int write_len) + { + return skb_cloned(skb) && !skb_clone_writable(skb, write_len) && + pskb_expand_head(skb, 0, 0, GFP_ATOMIC); + } + static inline int __skb_cow(struct sk_buff *skb, unsigned int headroom, int cloned) { @@@ -3550,6 -3565,7 +3566,7 @@@ static inline struct sec_path *skb_sec_ struct skb_gso_cb { int mac_offset; int encap_level; + __wsum csum; __u16 csum_start; }; #define SKB_SGO_CB_OFFSET 32 @@@ -3576,6 -3592,16 +3593,16 @@@ static inline int gso_pskb_expand_head( return 0; } + static inline void gso_reset_checksum(struct sk_buff *skb, __wsum res) + { + /* Do not update partial checksums if remote checksum is enabled. */ + if (skb->remcsum_offload) + return; + + SKB_GSO_CB(skb)->csum = res; + SKB_GSO_CB(skb)->csum_start = skb_checksum_start(skb) - skb->head; + } + /* Compute the checksum for a gso segment. First compute the checksum value * from the start of transport header to SKB_GSO_CB(skb)->csum_start, and * then add in skb->csum (checksum from csum_start to end of packet). @@@ -3586,15 -3612,14 +3613,14 @@@ */ static inline __sum16 gso_make_checksum(struct sk_buff *skb, __wsum res) { - int plen = SKB_GSO_CB(skb)->csum_start - skb_headroom(skb) - - skb_transport_offset(skb); - __wsum partial; + unsigned char *csum_start = skb_transport_header(skb); + int plen = (skb->head + SKB_GSO_CB(skb)->csum_start) - csum_start; + __wsum partial = SKB_GSO_CB(skb)->csum; - partial = csum_partial(skb_transport_header(skb), plen, skb->csum); - skb->csum = res; - SKB_GSO_CB(skb)->csum_start -= plen; + SKB_GSO_CB(skb)->csum = res; + SKB_GSO_CB(skb)->csum_start = csum_start - skb->head; - return csum_fold(partial); + return csum_fold(csum_partial(csum_start, plen, partial)); } static inline bool skb_is_gso(const struct sk_buff *skb) @@@ -3684,5 -3709,30 +3710,30 @@@ static inline unsigned int skb_gso_netw return hdr_len + skb_gso_transport_seglen(skb); } + /* Local Checksum Offload. + * Compute outer checksum based on the assumption that the + * inner checksum will be offloaded later. + * See Documentation/networking/checksum-offloads.txt for + * explanation of how this works. + * Fill in outer checksum adjustment (e.g. with sum of outer + * pseudo-header) before calling. + * Also ensure that inner checksum is in linear data area. + */ + static inline __wsum lco_csum(struct sk_buff *skb) + { + unsigned char *csum_start = skb_checksum_start(skb); + unsigned char *l4_hdr = skb_transport_header(skb); + __wsum partial; + + /* Start with complement of inner checksum adjustment */ + partial = ~csum_unfold(*(__force __sum16 *)(csum_start + + skb->csum_offset)); + + /* Add in checksum of our headers (incl. outer checksum + * adjustment filled in by caller) and return result. + */ + return csum_partial(l4_hdr, csum_start - l4_hdr, partial); + } + #endif /* __KERNEL__ */ #endif /* _LINUX_SKBUFF_H */ diff --combined include/net/ip_tunnels.h index dda9abf,4dd6163..5f28b60 --- a/include/net/ip_tunnels.h +++ b/include/net/ip_tunnels.h @@@ -13,6 -13,7 +13,7 @@@ #include <net/netns/generic.h> #include <net/rtnetlink.h> #include <net/lwtunnel.h> + #include <net/dst_cache.h> #if IS_ENABLED(CONFIG_IPV6) #include <net/ipv6.h> @@@ -57,6 -58,9 +58,9 @@@ struct ip_tunnel_key struct ip_tunnel_info { struct ip_tunnel_key key; + #ifdef CONFIG_DST_CACHE + struct dst_cache dst_cache; + #endif u8 options_len; u8 mode; }; @@@ -85,11 -89,6 +89,6 @@@ struct ip_tunnel_prl_entry struct rcu_head rcu_head; }; - struct ip_tunnel_dst { - struct dst_entry __rcu *dst; - __be32 saddr; - }; - struct metadata_dst; struct ip_tunnel { @@@ -108,7 -107,7 +107,7 @@@ int tun_hlen; /* Precalculated header length */ int mlink; - struct ip_tunnel_dst __percpu *dst_cache; + struct dst_cache dst_cache; struct ip_tunnel_parm parms; @@@ -230,7 -229,6 +229,7 @@@ void ip_tunnel_xmit(struct sk_buff *skb int ip_tunnel_ioctl(struct net_device *dev, struct ip_tunnel_parm *p, int cmd); int ip_tunnel_encap(struct sk_buff *skb, struct ip_tunnel *t, u8 *protocol, struct flowi4 *fl4); +int __ip_tunnel_change_mtu(struct net_device *dev, int new_mtu, bool strict); int ip_tunnel_change_mtu(struct net_device *dev, int new_mtu); struct rtnl_link_stats64 *ip_tunnel_get_stats64(struct net_device *dev, @@@ -248,7 -246,6 +247,6 @@@ int ip_tunnel_changelink(struct net_dev int ip_tunnel_newlink(struct net_device *dev, struct nlattr *tb[], struct ip_tunnel_parm *p); void ip_tunnel_setup(struct net_device *dev, int net_id); - void ip_tunnel_dst_reset_all(struct ip_tunnel *t); int ip_tunnel_encap_setup(struct ip_tunnel *t, struct ip_tunnel_encap *ipencap); @@@ -273,15 -270,15 +271,15 @@@ static inline u8 ip_tunnel_ecn_encap(u return INET_ECN_encapsulate(tos, inner); } - int iptunnel_pull_header(struct sk_buff *skb, int hdr_len, __be16 inner_proto); + int iptunnel_pull_header(struct sk_buff *skb, int hdr_len, __be16 inner_proto, + bool xnet); void iptunnel_xmit(struct sock *sk, struct rtable *rt, struct sk_buff *skb, __be32 src, __be32 dst, u8 proto, u8 tos, u8 ttl, __be16 df, bool xnet); struct metadata_dst *iptunnel_metadata_reply(struct metadata_dst *md, gfp_t flags); - struct sk_buff *iptunnel_handle_offloads(struct sk_buff *skb, bool gre_csum, - int gso_type_mask); + struct sk_buff *iptunnel_handle_offloads(struct sk_buff *skb, int gso_type_mask); static inline void iptunnel_xmit_stats(struct net_device *dev, int pkt_len) { diff --combined include/net/tcp.h index ae6468f,9b2cb0c..e90db85 --- a/include/net/tcp.h +++ b/include/net/tcp.h @@@ -239,13 -239,6 +239,6 @@@ extern struct inet_timewait_death_row t extern int sysctl_tcp_timestamps; extern int sysctl_tcp_window_scaling; extern int sysctl_tcp_sack; - extern int sysctl_tcp_fin_timeout; - extern int sysctl_tcp_syn_retries; - extern int sysctl_tcp_synack_retries; - extern int sysctl_tcp_retries1; - extern int sysctl_tcp_retries2; - extern int sysctl_tcp_orphan_retries; - extern int sysctl_tcp_syncookies; extern int sysctl_tcp_fastopen; extern int sysctl_tcp_retrans_collapse; extern int sysctl_tcp_stdurg; @@@ -274,7 -267,6 +267,6 @@@ extern int sysctl_tcp_thin_dupack extern int sysctl_tcp_early_retrans; extern int sysctl_tcp_limit_output_bytes; extern int sysctl_tcp_challenge_ack_limit; - extern unsigned int sysctl_tcp_notsent_lowat; extern int sysctl_tcp_min_tso_segs; extern int sysctl_tcp_min_rtt_wlen; extern int sysctl_tcp_autocorking; @@@ -447,7 -439,7 +439,7 @@@ const u8 *tcp_parse_md5sig_option(cons void tcp_v4_send_check(struct sock *sk, struct sk_buff *skb); void tcp_v4_mtu_reduced(struct sock *sk); -void tcp_req_err(struct sock *sk, u32 seq); +void tcp_req_err(struct sock *sk, u32 seq, bool abort); int tcp_v4_conn_request(struct sock *sk, struct sk_buff *skb); struct sock *tcp_create_openreq_child(const struct sock *sk, struct request_sock *req, @@@ -568,6 -560,7 +560,7 @@@ void tcp_rearm_rto(struct sock *sk) void tcp_synack_rtt_meas(struct sock *sk, struct request_sock *req); void tcp_reset(struct sock *sk); void tcp_skb_mark_lost_uncond_verify(struct tcp_sock *tp, struct sk_buff *skb); + void tcp_fin(struct sock *sk); /* tcp_timer.c */ void tcp_init_xmit_timers(struct sock *); @@@ -963,9 -956,11 +956,11 @@@ static inline void tcp_enable_fack(stru */ static inline void tcp_enable_early_retrans(struct tcp_sock *tp) { + struct net *net = sock_net((struct sock *)tp); + tp->do_early_retrans = sysctl_tcp_early_retrans && sysctl_tcp_early_retrans < 4 && !sysctl_tcp_thin_dupack && - sysctl_tcp_reordering == 3; + net->ipv4.sysctl_tcp_reordering == 3; } static inline void tcp_disable_early_retrans(struct tcp_sock *tp) @@@ -1252,7 -1247,7 +1247,7 @@@ static inline u32 keepalive_time_elapse static inline int tcp_fin_time(const struct sock *sk) { - int fin_timeout = tcp_sk(sk)->linger2 ? : sysctl_tcp_fin_timeout; + int fin_timeout = tcp_sk(sk)->linger2 ? : sock_net(sk)->ipv4.sysctl_tcp_fin_timeout; const int rto = inet_csk(sk)->icsk_rto; if (fin_timeout < (rto << 2) - (rto >> 1)) @@@ -1437,6 -1432,7 +1432,7 @@@ void tcp_free_fastopen_req(struct tcp_s extern struct tcp_fastopen_context __rcu *tcp_fastopen_ctx; int tcp_fastopen_reset_cipher(void *key, unsigned int len); + void tcp_fastopen_add_skb(struct sock *sk, struct sk_buff *skb); struct sock *tcp_try_fastopen(struct sock *sk, struct sk_buff *skb, struct request_sock *req, struct tcp_fastopen_cookie *foc, @@@ -1685,7 -1681,8 +1681,8 @@@ void __tcp_v4_send_check(struct sk_buf static inline u32 tcp_notsent_lowat(const struct tcp_sock *tp) { - return tp->notsent_lowat ?: sysctl_tcp_notsent_lowat; + struct net *net = sock_net((struct sock *)tp); + return tp->notsent_lowat ?: net->ipv4.sysctl_tcp_notsent_lowat; } static inline bool tcp_stream_memory_free(const struct sock *sk) diff --combined kernel/bpf/verifier.c index 2e7f7ab,36dc497..2e08f8e --- a/kernel/bpf/verifier.c +++ b/kernel/bpf/verifier.c @@@ -246,6 -246,7 +246,7 @@@ static const struct {BPF_MAP_TYPE_PROG_ARRAY, BPF_FUNC_tail_call}, {BPF_MAP_TYPE_PERF_EVENT_ARRAY, BPF_FUNC_perf_event_read}, {BPF_MAP_TYPE_PERF_EVENT_ARRAY, BPF_FUNC_perf_event_output}, + {BPF_MAP_TYPE_STACK_TRACE, BPF_FUNC_get_stackid}, }; static void print_verifier_state(struct verifier_env *env) @@@ -778,15 -779,24 +779,24 @@@ static int check_xadd(struct verifier_e * bytes from that pointer, make sure that it's within stack boundary * and all elements of stack are initialized */ - static int check_stack_boundary(struct verifier_env *env, - int regno, int access_size) + static int check_stack_boundary(struct verifier_env *env, int regno, + int access_size, bool zero_size_allowed) { struct verifier_state *state = &env->cur_state; struct reg_state *regs = state->regs; int off, i; - if (regs[regno].type != PTR_TO_STACK) + if (regs[regno].type != PTR_TO_STACK) { + if (zero_size_allowed && access_size == 0 && + regs[regno].type == CONST_IMM && + regs[regno].imm == 0) + return 0; + + verbose("R%d type=%s expected=%s\n", regno, + reg_type_str[regs[regno].type], + reg_type_str[PTR_TO_STACK]); return -EACCES; + } off = regs[regno].imm; if (off >= 0 || off < -MAX_BPF_STACK || off + access_size > 0 || @@@ -829,15 -839,24 +839,24 @@@ static int check_func_arg(struct verifi return 0; } - if (arg_type == ARG_PTR_TO_STACK || arg_type == ARG_PTR_TO_MAP_KEY || + if (arg_type == ARG_PTR_TO_MAP_KEY || arg_type == ARG_PTR_TO_MAP_VALUE) { expected_type = PTR_TO_STACK; - } else if (arg_type == ARG_CONST_STACK_SIZE) { + } else if (arg_type == ARG_CONST_STACK_SIZE || + arg_type == ARG_CONST_STACK_SIZE_OR_ZERO) { expected_type = CONST_IMM; } else if (arg_type == ARG_CONST_MAP_PTR) { expected_type = CONST_PTR_TO_MAP; } else if (arg_type == ARG_PTR_TO_CTX) { expected_type = PTR_TO_CTX; + } else if (arg_type == ARG_PTR_TO_STACK) { + expected_type = PTR_TO_STACK; + /* One exception here. In case function allows for NULL to be + * passed in as argument, it's a CONST_IMM type. Final test + * happens during stack boundary checking. + */ + if (reg->type == CONST_IMM && reg->imm == 0) + expected_type = CONST_IMM; } else { verbose("unsupported arg_type %d\n", arg_type); return -EFAULT; @@@ -867,8 -886,8 +886,8 @@@ verbose("invalid map_ptr to access map->key\n"); return -EACCES; } - err = check_stack_boundary(env, regno, (*mapp)->key_size); - + err = check_stack_boundary(env, regno, (*mapp)->key_size, + false); } else if (arg_type == ARG_PTR_TO_MAP_VALUE) { /* bpf_map_xxx(..., map_ptr, ..., value) call: * check [value, value + map->value_size) validity @@@ -878,9 -897,12 +897,12 @@@ verbose("invalid map_ptr to access map->value\n"); return -EACCES; } - err = check_stack_boundary(env, regno, (*mapp)->value_size); + err = check_stack_boundary(env, regno, (*mapp)->value_size, + false); + } else if (arg_type == ARG_CONST_STACK_SIZE || + arg_type == ARG_CONST_STACK_SIZE_OR_ZERO) { + bool zero_size_allowed = (arg_type == ARG_CONST_STACK_SIZE_OR_ZERO); - } else if (arg_type == ARG_CONST_STACK_SIZE) { /* bpf_xxx(..., buf, len) call will access 'len' bytes * from stack pointer 'buf'. Check it * note: regno == len, regno - 1 == buf @@@ -890,7 -912,8 +912,8 @@@ verbose("ARG_CONST_STACK_SIZE cannot be first argument\n"); return -EACCES; } - err = check_stack_boundary(env, regno - 1, reg->imm); + err = check_stack_boundary(env, regno - 1, reg->imm, + zero_size_allowed); } return err; @@@ -911,8 -934,11 +934,11 @@@ static int check_map_func_compatibility * don't allow any other map type to be passed into * the special func; */ - if (bool_func && bool_map != bool_func) + if (bool_func && bool_map != bool_func) { + verbose("cannot pass map_type %d into func %d\n", + map->map_type, func_id); return -EINVAL; + } } return 0; @@@ -2082,7 -2108,7 +2108,7 @@@ static void adjust_branches(struct bpf_ /* adjust offset of jmps if necessary */ if (i < pos && i + insn->off + 1 > pos) insn->off += delta; - else if (i > pos && i + insn->off + 1 < pos) + else if (i > pos + delta && i + insn->off + 1 <= pos + delta) insn->off -= delta; } } diff --combined lib/Kconfig.debug index 8bfd1ac,f890ee5..60d09e9 --- a/lib/Kconfig.debug +++ b/lib/Kconfig.debug @@@ -1400,21 -1400,6 +1400,21 @@@ config RCU_EQS_DEBU endmenu # "RCU Debugging" +config DEBUG_WQ_FORCE_RR_CPU + bool "Force round-robin CPU selection for unbound work items" + depends on DEBUG_KERNEL + default n + help + Workqueue used to implicitly guarantee that work items queued + without explicit CPU specified are put on the local CPU. This + guarantee is no longer true and while local CPU is still + preferred work items may be put on foreign CPUs. Kernel + parameter "workqueue.debug_force_rr_cpu" is added to force + round-robin CPU selection to flush out usages which depend on the + now broken guarantee. This config option enables the debug + feature by default. When enabled, memory and cache locality will + be impacted. + config DEBUG_BLOCK_EXT_DEVT bool "Force extended block device numbers and spread them" depends on DEBUG_KERNEL @@@ -1753,6 -1738,14 +1753,14 @@@ config TEST_KSTRTO config TEST_PRINTF tristate "Test printf() family of functions at runtime" + config TEST_BITMAP + tristate "Test bitmap_*() family of functions at runtime" + default n + help + Enable this option to test the bitmap functions at boot. + + If unsure, say N. + config TEST_RHASHTABLE tristate "Perform selftest on resizable hash table" default n diff --combined net/batman-adv/gateway_client.c index ccf70be,261866e..4b598bd --- a/net/batman-adv/gateway_client.c +++ b/net/batman-adv/gateway_client.c @@@ -1,4 -1,4 +1,4 @@@ - /* Copyright (C) 2009-2015 B.A.T.M.A.N. contributors: + /* Copyright (C) 2009-2016 B.A.T.M.A.N. contributors: * * Marek Lindner * @@@ -28,6 -28,7 +28,7 @@@ #include <linux/ip.h> #include <linux/ipv6.h> #include <linux/kernel.h> + #include <linux/kref.h> #include <linux/list.h> #include <linux/netdevice.h> #include <linux/rculist.h> @@@ -59,12 -60,29 +60,29 @@@ */ #define BATADV_DHCP_CHADDR_OFFSET 28 + /** + * batadv_gw_node_release - release gw_node from lists and queue for free after + * rcu grace period + * @ref: kref pointer of the gw_node + */ + static void batadv_gw_node_release(struct kref *ref) + { + struct batadv_gw_node *gw_node; + + gw_node = container_of(ref, struct batadv_gw_node, refcount); + + batadv_orig_node_free_ref(gw_node->orig_node); + kfree_rcu(gw_node, rcu); + } + + /** + * batadv_gw_node_free_ref - decrement the gw_node refcounter and possibly + * release it + * @gw_node: gateway node to free + */ static void batadv_gw_node_free_ref(struct batadv_gw_node *gw_node) { - if (atomic_dec_and_test(&gw_node->refcount)) { - batadv_orig_node_free_ref(gw_node->orig_node); - kfree_rcu(gw_node, rcu); - } + kref_put(&gw_node->refcount, batadv_gw_node_release); } static struct batadv_gw_node * @@@ -77,7 -95,7 +95,7 @@@ batadv_gw_get_selected_gw_node(struct b if (!gw_node) goto out; - if (!atomic_inc_not_zero(&gw_node->refcount)) + if (!kref_get_unless_zero(&gw_node->refcount)) gw_node = NULL; out: @@@ -100,7 -118,7 +118,7 @@@ batadv_gw_get_selected_orig(struct bata if (!orig_node) goto unlock; - if (!atomic_inc_not_zero(&orig_node->refcount)) + if (!kref_get_unless_zero(&orig_node->refcount)) orig_node = NULL; unlock: @@@ -118,7 -136,7 +136,7 @@@ static void batadv_gw_select(struct bat spin_lock_bh(&bat_priv->gw.list_lock); - if (new_gw_node && !atomic_inc_not_zero(&new_gw_node->refcount)) + if (new_gw_node && !kref_get_unless_zero(&new_gw_node->refcount)) new_gw_node = NULL; curr_gw_node = rcu_dereference_protected(bat_priv->gw.curr_gw, 1); @@@ -170,7 -188,7 +188,7 @@@ batadv_gw_get_best_gw_node(struct batad if (!router_ifinfo) goto next; - if (!atomic_inc_not_zero(&gw_node->refcount)) + if (!kref_get_unless_zero(&gw_node->refcount)) goto next; tq_avg = router_ifinfo->bat_iv.tq_avg; @@@ -188,7 -206,7 +206,7 @@@ if (curr_gw) batadv_gw_node_free_ref(curr_gw); curr_gw = gw_node; - atomic_inc(&curr_gw->refcount); + kref_get(&curr_gw->refcount); } break; @@@ -203,7 -221,7 +221,7 @@@ if (curr_gw) batadv_gw_node_free_ref(curr_gw); curr_gw = gw_node; - atomic_inc(&curr_gw->refcount); + kref_get(&curr_gw->refcount); } break; } @@@ -423,7 -441,7 +441,7 @@@ static void batadv_gw_node_add(struct b if (gateway->bandwidth_down == 0) return; - if (!atomic_inc_not_zero(&orig_node->refcount)) + if (!kref_get_unless_zero(&orig_node->refcount)) return; gw_node = kzalloc(sizeof(*gw_node), GFP_ATOMIC); @@@ -436,7 -454,7 +454,7 @@@ gw_node->orig_node = orig_node; gw_node->bandwidth_down = ntohl(gateway->bandwidth_down); gw_node->bandwidth_up = ntohl(gateway->bandwidth_up); - atomic_set(&gw_node->refcount, 1); + kref_init(&gw_node->refcount); spin_lock_bh(&bat_priv->gw.list_lock); hlist_add_head_rcu(&gw_node->list, &bat_priv->gw.list); @@@ -456,7 -474,7 +474,7 @@@ * @bat_priv: the bat priv with all the soft interface information * @orig_node: originator announcing gateway capabilities * - * Returns gateway node if found or NULL otherwise. + * Return: gateway node if found or NULL otherwise. */ static struct batadv_gw_node * batadv_gw_node_get(struct batadv_priv *bat_priv, @@@ -469,7 -487,7 +487,7 @@@ if (gw_node_tmp->orig_node != orig_node) continue; - if (!atomic_inc_not_zero(&gw_node_tmp->refcount)) + if (!kref_get_unless_zero(&gw_node_tmp->refcount)) continue; gw_node = gw_node_tmp; @@@ -527,12 -545,11 +545,12 @@@ void batadv_gw_node_update(struct batad * gets dereferenced. */ spin_lock_bh(&bat_priv->gw.list_lock); - hlist_del_init_rcu(&gw_node->list); + if (!hlist_unhashed(&gw_node->list)) { + hlist_del_init_rcu(&gw_node->list); + batadv_gw_node_free_ref(gw_node); + } spin_unlock_bh(&bat_priv->gw.list_lock); - batadv_gw_node_free_ref(gw_node); - curr_gw = batadv_gw_get_selected_gw_node(bat_priv); if (gw_node == curr_gw) batadv_gw_reselect(bat_priv); @@@ -656,13 -673,13 +674,13 @@@ out * @chaddr: buffer where the client address will be stored. Valid * only if the function returns BATADV_DHCP_TO_CLIENT * - * Returns: + * This function may re-allocate the data buffer of the skb passed as argument. + * + * Return: * - BATADV_DHCP_NO if the packet is not a dhcp message or if there was an error * while parsing it * - BATADV_DHCP_TO_SERVER if this is a message going to the DHCP server * - BATADV_DHCP_TO_CLIENT if this is a message going to a DHCP client - * - * This function may re-allocate the data buffer of the skb passed as argument. */ enum batadv_dhcp_recipient batadv_gw_dhcp_recipient_get(struct sk_buff *skb, unsigned int *header_len, @@@ -777,11 -794,11 +795,11 @@@ * server. Due to topology changes it may be the case that the GW server * previously selected is not the best one anymore. * - * Returns true if the packet destination is unicast and it is not the best gw, - * false otherwise. - * * This call might reallocate skb data. * Must be invoked only when the DHCP packet is going TO a DHCP SERVER. + * + * Return: true if the packet destination is unicast and it is not the best gw, + * false otherwise. */ bool batadv_gw_out_of_range(struct batadv_priv *bat_priv, struct sk_buff *skb) diff --combined net/batman-adv/hard-interface.c index 57f71071,fb2d9c0..e2aaa4c --- a/net/batman-adv/hard-interface.c +++ b/net/batman-adv/hard-interface.c @@@ -1,4 -1,4 +1,4 @@@ - /* Copyright (C) 2007-2015 B.A.T.M.A.N. contributors: + /* Copyright (C) 2007-2016 B.A.T.M.A.N. contributors: * * Marek Lindner, Simon Wunderlich * @@@ -18,6 -18,7 +18,7 @@@ #include "hard-interface.h" #include "main.h" + #include <linux/atomic.h> #include <linux/bug.h> #include <linux/byteorder/generic.h> #include <linux/errno.h> @@@ -26,6 -27,7 +27,7 @@@ #include <linux/if_ether.h> #include <linux/if.h> #include <linux/kernel.h> + #include <linux/kref.h> #include <linux/list.h> #include <linux/netdevice.h> #include <linux/printk.h> @@@ -47,13 -49,19 +49,19 @@@ #include "sysfs.h" #include "translation-table.h" - void batadv_hardif_free_rcu(struct rcu_head *rcu) + /** + * batadv_hardif_release - release hard interface from lists and queue for + * free after rcu grace period + * @ref: kref pointer of the hard interface + */ + void batadv_hardif_release(struct kref *ref) { struct batadv_hard_iface *hard_iface; - hard_iface = container_of(rcu, struct batadv_hard_iface, rcu); + hard_iface = container_of(ref, struct batadv_hard_iface, refcount); dev_put(hard_iface->net_dev); - kfree(hard_iface); + + kfree_rcu(hard_iface, rcu); } struct batadv_hard_iface * @@@ -64,7 -72,7 +72,7 @@@ batadv_hardif_get_by_netdev(const struc rcu_read_lock(); list_for_each_entry_rcu(hard_iface, &batadv_hardif_list, list) { if (hard_iface->net_dev == net_dev && - atomic_inc_not_zero(&hard_iface->refcount)) + kref_get_unless_zero(&hard_iface->refcount)) goto out; } @@@ -76,28 -84,6 +84,28 @@@ out } /** + * batadv_mutual_parents - check if two devices are each others parent + * @dev1: 1st net_device + * @dev2: 2nd net_device + * + * veth devices come in pairs and each is the parent of the other! + * + * Return: true if the devices are each others parent, otherwise false + */ +static bool batadv_mutual_parents(const struct net_device *dev1, + const struct net_device *dev2) +{ + int dev1_parent_iflink = dev_get_iflink(dev1); + int dev2_parent_iflink = dev_get_iflink(dev2); + + if (!dev1_parent_iflink || !dev2_parent_iflink) + return false; + + return (dev1_parent_iflink == dev2->ifindex) && + (dev2_parent_iflink == dev1->ifindex); +} + +/** * batadv_is_on_batman_iface - check if a device is a batman iface descendant * @net_dev: the device to check * @@@ -107,7 -93,7 +115,7 @@@ * This function recursively checks all the fathers of the device passed as * argument looking for a batman-adv soft interface. * - * Returns true if the device is descendant of a batman-adv mesh interface (or + * Return: true if the device is descendant of a batman-adv mesh interface (or * if it is a batman-adv interface itself), false otherwise */ static bool batadv_is_on_batman_iface(const struct net_device *net_dev) @@@ -130,9 -116,6 +138,9 @@@ if (WARN(!parent_dev, "Cannot find parent device")) return false; + if (batadv_mutual_parents(net_dev, parent_dev)) + return false; + ret = batadv_is_on_batman_iface(parent_dev); return ret; @@@ -161,7 -144,7 +169,7 @@@ static int batadv_is_valid_iface(const * interface * @net_device: the device to check * - * Returns true if the net device is a 802.11 wireless device, false otherwise. + * Return: true if the net device is a 802.11 wireless device, false otherwise. */ bool batadv_is_wifi_netdev(struct net_device *net_device) { @@@ -194,7 -177,7 +202,7 @@@ batadv_hardif_get_active(const struct n continue; if (hard_iface->if_status == BATADV_IF_ACTIVE && - atomic_inc_not_zero(&hard_iface->refcount)) + kref_get_unless_zero(&hard_iface->refcount)) goto out; } @@@ -228,7 -211,7 +236,7 @@@ static void batadv_primary_if_select(st ASSERT_RTNL(); - if (new_hard_iface && !atomic_inc_not_zero(&new_hard_iface->refcount)) + if (new_hard_iface && !kref_get_unless_zero(&new_hard_iface->refcount)) new_hard_iface = NULL; curr_hard_iface = rcu_dereference_protected(bat_priv->primary_if, 1); @@@ -426,7 -409,8 +434,8 @@@ batadv_hardif_deactivate_interface(stru * * Invoke ndo_del_slave on master passing slave as argument. In this way slave * is free'd and master can correctly change its internal state. - * Return 0 on success, a negative value representing the error otherwise + * + * Return: 0 on success, a negative value representing the error otherwise */ static int batadv_master_del_slave(struct batadv_hard_iface *slave, struct net_device *master) @@@ -455,7 -439,7 +464,7 @@@ int batadv_hardif_enable_interface(stru if (hard_iface->if_status != BATADV_IF_NOT_IN_USE) goto out; - if (!atomic_inc_not_zero(&hard_iface->refcount)) + if (!kref_get_unless_zero(&hard_iface->refcount)) goto out; soft_iface = dev_get_by_name(&init_net, iface_name); @@@ -676,7 -660,8 +685,8 @@@ batadv_hardif_add_interface(struct net_ hard_iface->num_bcasts = BATADV_NUM_BCASTS_WIRELESS; /* extra reference for return */ - atomic_set(&hard_iface->refcount, 2); + kref_init(&hard_iface->refcount); + kref_get(&hard_iface->refcount); batadv_check_known_mac_addr(hard_iface->net_dev); list_add_tail_rcu(&hard_iface->list, &batadv_hardif_list); diff --combined net/batman-adv/translation-table.c index 0e80fd1,1188279..5c7fa02 --- a/net/batman-adv/translation-table.c +++ b/net/batman-adv/translation-table.c @@@ -1,4 -1,4 +1,4 @@@ - /* Copyright (C) 2007-2015 B.A.T.M.A.N. contributors: + /* Copyright (C) 2007-2016 B.A.T.M.A.N. contributors: * * Marek Lindner, Simon Wunderlich, Antonio Quartulli * @@@ -31,6 -31,7 +31,7 @@@ #include <linux/jhash.h> #include <linux/jiffies.h> #include <linux/kernel.h> + #include <linux/kref.h> #include <linux/list.h> #include <linux/lockdep.h> #include <linux/netdevice.h> @@@ -68,7 -69,15 +69,15 @@@ static void batadv_tt_global_del(struc unsigned short vid, const char *message, bool roaming); - /* returns 1 if they are the same mac addr and vid */ + /** + * batadv_compare_tt - check if two TT entries are the same + * @node: the list element pointer of the first TT entry + * @data2: pointer to the tt_common_entry of the second TT entry + * + * Compare the MAC address and the VLAN ID of the two TT entries and check if + * they are the same TT client. + * Return: 1 if the two TT clients are the same, 0 otherwise + */ static int batadv_compare_tt(const struct hlist_node *node, const void *data2) { const void *data1 = container_of(node, struct batadv_tt_common_entry, @@@ -84,7 -93,7 +93,7 @@@ * @data: pointer to the tt_common_entry object to map * @size: the size of the hash table * - * Returns the hash index where the object represented by 'data' should be + * Return: the hash index where the object represented by 'data' should be * stored at. */ static inline u32 batadv_choose_tt(const void *data, u32 size) @@@ -105,7 -114,7 +114,7 @@@ * @addr: the mac address of the client to look for * @vid: VLAN identifier * - * Returns a pointer to the tt_common struct belonging to the searched client if + * Return: a pointer to the tt_common struct belonging to the searched client if * found, NULL otherwise. */ static struct batadv_tt_common_entry * @@@ -133,7 -142,7 +142,7 @@@ batadv_tt_hash_find(struct batadv_hasht if (tt->vid != vid) continue; - if (!atomic_inc_not_zero(&tt->refcount)) + if (!kref_get_unless_zero(&tt->refcount)) continue; tt_tmp = tt; @@@ -150,7 -159,7 +159,7 @@@ * @addr: the mac address of the client to look for * @vid: VLAN identifier * - * Returns a pointer to the corresponding tt_local_entry struct if the client is + * Return: a pointer to the corresponding tt_local_entry struct if the client is * found, NULL otherwise. */ static struct batadv_tt_local_entry * @@@ -175,7 -184,7 +184,7 @@@ batadv_tt_local_hash_find(struct batadv * @addr: the mac address of the client to look for * @vid: VLAN identifier * - * Returns a pointer to the corresponding tt_global_entry struct if the client + * Return: a pointer to the corresponding tt_global_entry struct if the client * is found, NULL otherwise. */ static struct batadv_tt_global_entry * @@@ -194,34 -203,68 +203,68 @@@ batadv_tt_global_hash_find(struct batad return tt_global_entry; } + /** + * batadv_tt_local_entry_release - release tt_local_entry from lists and queue + * for free after rcu grace period + * @ref: kref pointer of the nc_node + */ + static void batadv_tt_local_entry_release(struct kref *ref) + { + struct batadv_tt_local_entry *tt_local_entry; + + tt_local_entry = container_of(ref, struct batadv_tt_local_entry, + common.refcount); + + kfree_rcu(tt_local_entry, common.rcu); + } + + /** + * batadv_tt_local_entry_free_ref - decrement the tt_local_entry refcounter and + * possibly release it + * @tt_local_entry: tt_local_entry to be free'd + */ static void batadv_tt_local_entry_free_ref(struct batadv_tt_local_entry *tt_local_entry) { - if (atomic_dec_and_test(&tt_local_entry->common.refcount)) - kfree_rcu(tt_local_entry, common.rcu); + kref_put(&tt_local_entry->common.refcount, + batadv_tt_local_entry_release); } /** - * batadv_tt_global_entry_free_ref - decrement the refcounter for a - * tt_global_entry and possibly free it - * @tt_global_entry: the object to free + * batadv_tt_global_entry_release - release tt_global_entry from lists and queue + * for free after rcu grace period + * @ref: kref pointer of the nc_node + */ + static void batadv_tt_global_entry_release(struct kref *ref) + { + struct batadv_tt_global_entry *tt_global_entry; + + tt_global_entry = container_of(ref, struct batadv_tt_global_entry, + common.refcount); + + batadv_tt_global_del_orig_list(tt_global_entry); + kfree_rcu(tt_global_entry, common.rcu); + } + + /** + * batadv_tt_global_entry_free_ref - decrement the tt_global_entry refcounter + * and possibly release it + * @tt_global_entry: tt_global_entry to be free'd */ static void batadv_tt_global_entry_free_ref(struct batadv_tt_global_entry *tt_global_entry) { - if (atomic_dec_and_test(&tt_global_entry->common.refcount)) { - batadv_tt_global_del_orig_list(tt_global_entry); - kfree_rcu(tt_global_entry, common.rcu); - } + kref_put(&tt_global_entry->common.refcount, + batadv_tt_global_entry_release); } /** * batadv_tt_global_hash_count - count the number of orig entries - * @hash: hash table containing the tt entries + * @bat_priv: the bat priv with all the soft interface information * @addr: the mac address of the client to count entries for * @vid: VLAN identifier * - * Return the number of originators advertising the given address/data + * Return: the number of originators advertising the given address/data * (excluding ourself). */ int batadv_tt_global_hash_count(struct batadv_priv *bat_priv, @@@ -286,9 -329,9 +329,9 @@@ static void batadv_tt_local_size_dec(st } /** - * batadv_tt_global_size_mod - change the size by v of the local table - * identified by vid - * @bat_priv: the bat priv with all the soft interface information + * batadv_tt_global_size_mod - change the size by v of the global table + * for orig_node identified by vid + * @orig_node: the originator for which the table has to be modified * @vid: the VLAN identifier * @v: the amount to sum to the global table size */ @@@ -303,11 -346,9 +346,11 @@@ static void batadv_tt_global_size_mod(s if (atomic_add_return(v, &vlan->tt.num_entries) == 0) { spin_lock_bh(&orig_node->vlan_list_lock); - hlist_del_init_rcu(&vlan->list); + if (!hlist_unhashed(&vlan->list)) { + hlist_del_init_rcu(&vlan->list); + batadv_orig_node_vlan_free_ref(vlan); + } spin_unlock_bh(&orig_node->vlan_list_lock); - batadv_orig_node_vlan_free_ref(vlan); } batadv_orig_node_vlan_free_ref(vlan); @@@ -340,22 -381,28 +383,28 @@@ static void batadv_tt_global_size_dec(s /** * batadv_tt_orig_list_entry_release - release tt orig entry from lists and * queue for free after rcu grace period - * @orig_entry: tt orig entry to be free'd + * @ref: kref pointer of the tt orig entry */ - static void - batadv_tt_orig_list_entry_release(struct batadv_tt_orig_list_entry *orig_entry) + static void batadv_tt_orig_list_entry_release(struct kref *ref) { + struct batadv_tt_orig_list_entry *orig_entry; + + orig_entry = container_of(ref, struct batadv_tt_orig_list_entry, + refcount); + batadv_orig_node_free_ref(orig_entry->orig_node); kfree_rcu(orig_entry, rcu); } + /** + * batadv_tt_orig_list_entry_free_ref - decrement the tt orig entry refcounter + * and possibly release it + * @orig_entry: tt orig entry to be free'd + */ static void batadv_tt_orig_list_entry_free_ref(struct batadv_tt_orig_list_entry *orig_entry) { - if (!atomic_dec_and_test(&orig_entry->refcount)) - return; - - batadv_tt_orig_list_entry_release(orig_entry); + kref_put(&orig_entry->refcount, batadv_tt_orig_list_entry_release); } /** @@@ -437,7 -484,7 +486,7 @@@ unlock * batadv_tt_len - compute length in bytes of given number of tt changes * @changes_num: number of tt changes * - * Returns computed length in bytes. + * Return: computed length in bytes. */ static int batadv_tt_len(int changes_num) { @@@ -448,7 -495,7 +497,7 @@@ * batadv_tt_entries - compute the number of entries fitting in tt_len bytes * @tt_len: available space * - * Returns the number of entries. + * Return: the number of entries. */ static u16 batadv_tt_entries(u16 tt_len) { @@@ -460,7 -507,7 +509,7 @@@ * size when transmitted over the air * @bat_priv: the bat priv with all the soft interface information * - * Returns local translation table size in bytes. + * Return: local translation table size in bytes. */ static int batadv_tt_local_table_transmit_size(struct batadv_priv *bat_priv) { @@@ -526,7 -573,7 +575,7 @@@ static void batadv_tt_global_free(struc * @mark: the value contained in the skb->mark field of the received packet (if * any) * - * Returns true if the client was successfully added, false otherwise. + * Return: true if the client was successfully added, false otherwise. */ bool batadv_tt_local_add(struct net_device *soft_iface, const u8 *addr, unsigned short vid, int ifindex, u32 mark) @@@ -620,7 -667,8 +669,8 @@@ tt_local->common.vid = vid; if (batadv_is_wifi_netdev(in_dev)) tt_local->common.flags |= BATADV_TT_CLIENT_WIFI; - atomic_set(&tt_local->common.refcount, 2); + kref_init(&tt_local->common.refcount); + kref_get(&tt_local->common.refcount); tt_local->last_seen = jiffies; tt_local->common.added_at = tt_local->last_seen; @@@ -721,12 -769,11 +771,11 @@@ out * function reserves the amount of space needed to send the entire global TT * table. In case of success the value is updated with the real amount of * reserved bytes - * Allocate the needed amount of memory for the entire TT TVLV and write its * header made up by one tvlv_tt_data object and a series of tvlv_tt_vlan_data * objects, one per active VLAN served by the originator node. * - * Return the size of the allocated buffer or 0 in case of failure. + * Return: the size of the allocated buffer or 0 in case of failure. */ static u16 batadv_tt_prepare_tvlv_global_data(struct batadv_orig_node *orig_node, @@@ -800,7 -847,7 +849,7 @@@ out * header made up by one tvlv_tt_data object and a series of tvlv_tt_vlan_data * objects, one per active VLAN. * - * Return the size of the allocated buffer or 0 in case of failure. + * Return: the size of the allocated buffer or 0 in case of failure. */ static u16 batadv_tt_prepare_tvlv_local_data(struct batadv_priv *bat_priv, @@@ -1042,7 -1089,7 +1091,7 @@@ batadv_tt_local_set_pending(struct bata * @message: message to append to the log on deletion * @roaming: true if the deletion is due to a roaming event * - * Returns the flags assigned to the local entry before being deleted + * Return: the flags assigned to the local entry before being deleted */ u16 batadv_tt_local_remove(struct batadv_priv *bat_priv, const u8 *addr, unsigned short vid, const char *message, @@@ -1242,10 -1289,16 +1291,16 @@@ static void batadv_tt_changes_list_free spin_unlock_bh(&bat_priv->tt.changes_list_lock); } - /* retrieves the orig_tt_list_entry belonging to orig_node from the + /** + * batadv_tt_global_orig_entry_find - find a TT orig_list_entry + * @entry: the TT global entry where the orig_list_entry has to be + * extracted from + * @orig_node: the originator for which the orig_list_entry has to be found + * + * retrieve the orig_tt_list_entry belonging to orig_node from the * batadv_tt_global_entry list * - * returns it with an increased refcounter, NULL if not found + * Return: it with an increased refcounter, NULL if not found */ static struct batadv_tt_orig_list_entry * batadv_tt_global_orig_entry_find(const struct batadv_tt_global_entry *entry, @@@ -1259,7 -1312,7 +1314,7 @@@ hlist_for_each_entry_rcu(tmp_orig_entry, head, list) { if (tmp_orig_entry->orig_node != orig_node) continue; - if (!atomic_inc_not_zero(&tmp_orig_entry->refcount)) + if (!kref_get_unless_zero(&tmp_orig_entry->refcount)) continue; orig_entry = tmp_orig_entry; @@@ -1270,8 -1323,15 +1325,15 @@@ return orig_entry; } - /* find out if an orig_node is already in the list of a tt_global_entry. - * returns true if found, false otherwise + /** + * batadv_tt_global_entry_has_orig - check if a TT global entry is also handled + * by a given originator + * @entry: the TT global entry to check + * @orig_node: the originator to search in the list + * + * find out if an orig_node is already in the list of a tt_global_entry. + * + * Return: true if found, false otherwise */ static bool batadv_tt_global_entry_has_orig(const struct batadv_tt_global_entry *entry, @@@ -1309,11 -1369,12 +1371,12 @@@ batadv_tt_global_orig_entry_add(struct goto out; INIT_HLIST_NODE(&orig_entry->list); - atomic_inc(&orig_node->refcount); + kref_get(&orig_node->refcount); batadv_tt_global_size_inc(orig_node, tt_global->common.vid); orig_entry->orig_node = orig_node; orig_entry->ttvn = ttvn; - atomic_set(&orig_entry->refcount, 2); + kref_init(&orig_entry->refcount); + kref_get(&orig_entry->refcount); spin_lock_bh(&tt_global->list_lock); hlist_add_head_rcu(&orig_entry->list, @@@ -1343,7 -1404,7 +1406,7 @@@ out * * The caller must hold orig_node refcount. * - * Return true if the new entry has been added, false otherwise + * Return: true if the new entry has been added, false otherwise */ static bool batadv_tt_global_add(struct batadv_priv *bat_priv, struct batadv_orig_node *orig_node, @@@ -1389,7 -1450,8 +1452,8 @@@ */ if (flags & BATADV_TT_CLIENT_ROAM) tt_global_entry->roam_at = jiffies; - atomic_set(&common->refcount, 2); + kref_init(&common->refcount); + kref_get(&common->refcount); common->added_at = jiffies; INIT_HLIST_HEAD(&tt_global_entry->orig_list); @@@ -1501,7 -1563,7 +1565,7 @@@ out * @tt_global_entry: global translation table entry to be analyzed * * This functon assumes the caller holds rcu_read_lock(). - * Returns best originator list entry or NULL on errors. + * Return: best originator list entry or NULL on errors. */ static struct batadv_tt_orig_list_entry * batadv_transtable_best_orig(struct batadv_priv *bat_priv, @@@ -2031,7 -2093,7 +2095,7 @@@ _batadv_is_ap_isolated(struct batadv_tt * @addr: mac address of the destination client * @vid: VLAN identifier * - * Returns a pointer to the originator that was selected as destination in the + * Return: a pointer to the originator that was selected as destination in the * mesh for contacting the client 'addr', NULL otherwise. * In case of multiple originators serving the same client, the function returns * the best one (best in terms of metric towards the destination node). @@@ -2071,7 -2133,7 +2135,7 @@@ struct batadv_orig_node *batadv_transta /* found anything? */ if (best_entry) orig_node = best_entry->orig_node; - if (orig_node && !atomic_inc_not_zero(&orig_node->refcount)) + if (orig_node && !kref_get_unless_zero(&orig_node->refcount)) orig_node = NULL; rcu_read_unlock(); @@@ -2106,7 -2168,7 +2170,7 @@@ out * because the XOR operation can combine them all while trying to reduce the * noise as much as possible. * - * Returns the checksum of the global table of a given originator. + * Return: the checksum of the global table of a given originator. */ static u32 batadv_tt_global_crc(struct batadv_priv *bat_priv, struct batadv_orig_node *orig_node, @@@ -2183,7 -2245,7 +2247,7 @@@ * For details about the computation, please refer to the documentation for * batadv_tt_global_crc(). * - * Returns the checksum of the local table + * Return: the checksum of the local table */ static u32 batadv_tt_local_crc(struct batadv_priv *bat_priv, unsigned short vid) @@@ -2289,7 -2351,7 +2353,7 @@@ static void batadv_tt_req_purge(struct * @bat_priv: the bat priv with all the soft interface information * @orig_node: orig node this request is being issued for * - * Returns the pointer to the new tt_req_node struct if no request + * Return: the pointer to the new tt_req_node struct if no request * has already been issued for this orig_node, NULL otherwise. */ static struct batadv_tt_req_node * @@@ -2324,7 -2386,7 +2388,7 @@@ unlock * @entry_ptr: to be checked local tt entry * @data_ptr: not used but definition required to satisfy the callback prototype * - * Returns 1 if the entry is a valid, 0 otherwise. + * Return: 1 if the entry is a valid, 0 otherwise. */ static int batadv_tt_local_valid(const void *entry_ptr, const void *data_ptr) { @@@ -2408,9 -2470,8 +2472,8 @@@ static void batadv_tt_tvlv_generate(str * @orig_node: originator for which the CRCs have to be checked * @tt_vlan: pointer to the first tvlv VLAN entry * @num_vlan: number of tvlv VLAN entries - * @create: if true, create VLAN objects if not found * - * Return true if all the received CRCs match the locally stored ones, false + * Return: true if all the received CRCs match the locally stored ones, false * otherwise */ static bool batadv_tt_global_check_crc(struct batadv_orig_node *orig_node, @@@ -2513,6 -2574,8 +2576,8 @@@ static void batadv_tt_global_update_crc * @num_vlan: number of tvlv VLAN entries * @full_table: ask for the entire translation table if true, while only for the * last TT diff otherwise + * + * Return: true if the TT Request was sent, false otherwise */ static int batadv_send_tt_request(struct batadv_priv *bat_priv, struct batadv_orig_node *dst_orig_node, @@@ -2593,7 -2656,7 +2658,7 @@@ out * @req_src: mac address of tt request sender * @req_dst: mac address of tt request recipient * - * Returns true if tt request reply was sent, false otherwise. + * Return: true if tt request reply was sent, false otherwise. */ static bool batadv_send_other_tt_response(struct batadv_priv *bat_priv, struct batadv_tvlv_tt_data *tt_data, @@@ -2725,7 -2788,7 +2790,7 @@@ out * @tt_data: tt data containing the tt request information * @req_src: mac address of tt request sender * - * Returns true if tt request reply was sent, false otherwise. + * Return: true if tt request reply was sent, false otherwise. */ static bool batadv_send_my_tt_response(struct batadv_priv *bat_priv, struct batadv_tvlv_tt_data *tt_data, @@@ -2843,7 -2906,7 +2908,7 @@@ out * @req_src: mac address of tt request sender * @req_dst: mac address of tt request recipient * - * Returns true if tt request reply was sent, false otherwise. + * Return: true if tt request reply was sent, false otherwise. */ static bool batadv_send_tt_response(struct batadv_priv *bat_priv, struct batadv_tvlv_tt_data *tt_data, @@@ -2938,7 -3001,7 +3003,7 @@@ static void batadv_tt_update_changes(st * @addr: the mac address of the client to check * @vid: VLAN identifier * - * Returns true if the client is served by this node, false otherwise. + * Return: true if the client is served by this node, false otherwise. */ bool batadv_is_my_client(struct batadv_priv *bat_priv, const u8 *addr, unsigned short vid) @@@ -3055,11 -3118,16 +3120,16 @@@ static void batadv_tt_roam_purge(struc spin_unlock_bh(&bat_priv->tt.roam_list_lock); } - /* This function checks whether the client already reached the + /** + * batadv_tt_check_roam_count - check if a client has roamed too frequently + * @bat_priv: the bat priv with all the soft interface information + * @client: mac address of the roaming client + * + * This function checks whether the client already reached the * maximum number of possible roaming phases. In this case the ROAMING_ADV * will not be sent. * - * returns true if the ROAMING_ADV can be sent, false otherwise + * Return: true if the ROAMING_ADV can be sent, false otherwise */ static bool batadv_tt_check_roam_count(struct batadv_priv *bat_priv, u8 *client) { @@@ -3371,13 -3439,12 +3441,12 @@@ out * batadv_tt_update_orig - update global translation table with new tt * information received via ogms * @bat_priv: the bat priv with all the soft interface information - * @orig: the orig_node of the ogm - * @tt_vlan: pointer to the first tvlv VLAN entry + * @orig_node: the orig_node of the ogm + * @tt_buff: pointer to the first tvlv VLAN entry * @tt_num_vlan: number of tvlv VLAN entries * @tt_change: pointer to the first entry in the TT buffer * @tt_num_changes: number of tt changes inside the tt buffer * @ttvn: translation table version number of this changeset - * @tt_crc: crc32 checksum of orig node's translation table */ static void batadv_tt_update_orig(struct batadv_priv *bat_priv, struct batadv_orig_node *orig_node, @@@ -3459,7 -3526,7 +3528,7 @@@ request_table * @addr: the mac address of the client to check * @vid: VLAN identifier * - * Returns true if we know that the client has moved from its old originator + * Return: true if we know that the client has moved from its old originator * to another one. This entry is still kept for consistency purposes and will be * deleted later by a DEL or because of timeout */ @@@ -3485,7 -3552,7 +3554,7 @@@ out * @addr: the mac address of the local client to query * @vid: VLAN identifier * - * Returns true if the local client is known to be roaming (it is not served by + * Return: true if the local client is known to be roaming (it is not served by * this node anymore) or not. If yes, the client is still present in the table * to keep the latter consistent with the node TTVN */ @@@ -3614,7 -3681,7 +3683,7 @@@ static void batadv_tt_tvlv_ogm_handler_ * @tvlv_value: tvlv buffer containing the tt data * @tvlv_value_len: tvlv buffer length * - * Returns NET_RX_DROP if the tt tvlv is to be re-routed, NET_RX_SUCCESS + * Return: NET_RX_DROP if the tt tvlv is to be re-routed, NET_RX_SUCCESS * otherwise. */ static int batadv_tt_tvlv_unicast_handler_v1(struct batadv_priv *bat_priv, @@@ -3695,7 -3762,7 +3764,7 @@@ * @tvlv_value: tvlv buffer containing the tt data * @tvlv_value_len: tvlv buffer length * - * Returns NET_RX_DROP if the tt roam tvlv is to be re-routed, NET_RX_SUCCESS + * Return: NET_RX_DROP if the tt roam tvlv is to be re-routed, NET_RX_SUCCESS * otherwise. */ static int batadv_roam_tvlv_unicast_handler_v1(struct batadv_priv *bat_priv, @@@ -3741,7 -3808,7 +3810,7 @@@ out * batadv_tt_init - initialise the translation table internals * @bat_priv: the bat priv with all the soft interface information * - * Return 0 on success or negative error number in case of failure. + * Return: 0 on success or negative error number in case of failure. */ int batadv_tt_init(struct batadv_priv *bat_priv) { @@@ -3779,7 -3846,7 +3848,7 @@@ * @addr: the mac address of the client * @vid: the identifier of the VLAN where this client is connected * - * Returns true if the client is marked with the TT_CLIENT_ISOLA flag, false + * Return: true if the client is marked with the TT_CLIENT_ISOLA flag, false * otherwise */ bool batadv_tt_global_is_isolated(struct batadv_priv *bat_priv, diff --combined net/bridge/br_mdb.c index 74c278e,cf51b7b..73786e2 --- a/net/bridge/br_mdb.c +++ b/net/bridge/br_mdb.c @@@ -41,6 -41,14 +41,14 @@@ fail return -EMSGSIZE; } + static void __mdb_entry_fill_flags(struct br_mdb_entry *e, unsigned char flags) + { + e->state = flags & MDB_PG_FLAGS_PERMANENT; + e->flags = 0; + if (flags & MDB_PG_FLAGS_OFFLOAD) + e->flags |= MDB_FLAGS_OFFLOAD; + } + static int br_mdb_fill_info(struct sk_buff *skb, struct netlink_callback *cb, struct net_device *dev) { @@@ -80,26 -88,41 +88,41 @@@ for (pp = &mp->ports; (p = rcu_dereference(*pp)) != NULL; pp = &p->next) { + struct nlattr *nest_ent; + struct br_mdb_entry e; + port = p->port; - if (port) { - struct br_mdb_entry e; - memset(&e, 0, sizeof(e)); - e.ifindex = port->dev->ifindex; - e.state = p->state; - e.vid = p->addr.vid; - if (p->addr.proto == htons(ETH_P_IP)) - e.addr.u.ip4 = p->addr.u.ip4; + if (!port) + continue; + + memset(&e, 0, sizeof(e)); + e.ifindex = port->dev->ifindex; + e.vid = p->addr.vid; + __mdb_entry_fill_flags(&e, p->flags); + if (p->addr.proto == htons(ETH_P_IP)) + e.addr.u.ip4 = p->addr.u.ip4; #if IS_ENABLED(CONFIG_IPV6) - if (p->addr.proto == htons(ETH_P_IPV6)) - e.addr.u.ip6 = p->addr.u.ip6; + if (p->addr.proto == htons(ETH_P_IPV6)) + e.addr.u.ip6 = p->addr.u.ip6; #endif - e.addr.proto = p->addr.proto; - if (nla_put(skb, MDBA_MDB_ENTRY_INFO, sizeof(e), &e)) { - nla_nest_cancel(skb, nest2); - err = -EMSGSIZE; - goto out; - } + e.addr.proto = p->addr.proto; + nest_ent = nla_nest_start(skb, + MDBA_MDB_ENTRY_INFO); + if (!nest_ent) { + nla_nest_cancel(skb, nest2); + err = -EMSGSIZE; + goto out; } + if (nla_put_nohdr(skb, sizeof(e), &e) || + nla_put_u32(skb, + MDBA_MDB_EATTR_TIMER, + br_timer_value(&p->timer))) { + nla_nest_cancel(skb, nest_ent); + nla_nest_cancel(skb, nest2); + err = -EMSGSIZE; + goto out; + } + nla_nest_end(skb, nest_ent); } nla_nest_end(skb, nest2); skip: @@@ -209,7 -232,7 +232,7 @@@ static inline size_t rtnl_mdb_nlmsg_siz } static void __br_mdb_notify(struct net_device *dev, struct br_mdb_entry *entry, - int type) + int type, struct net_bridge_port_group *pg) { struct switchdev_obj_port_mdb mdb = { .obj = { @@@ -232,10 -255,13 +255,13 @@@ #endif mdb.obj.orig_dev = port_dev; - if (port_dev && type == RTM_NEWMDB) - switchdev_port_obj_add(port_dev, &mdb.obj); - else if (port_dev && type == RTM_DELMDB) + if (port_dev && type == RTM_NEWMDB) { + err = switchdev_port_obj_add(port_dev, &mdb.obj); + if (!err && pg) + pg->flags |= MDB_PG_FLAGS_OFFLOAD; + } else if (port_dev && type == RTM_DELMDB) { switchdev_port_obj_del(port_dev, &mdb.obj); + } skb = nlmsg_new(rtnl_mdb_nlmsg_size(), GFP_ATOMIC); if (!skb) @@@ -253,21 -279,21 +279,21 @@@ errout rtnl_set_sk_err(net, RTNLGRP_MDB, err); } - void br_mdb_notify(struct net_device *dev, struct net_bridge_port *port, - struct br_ip *group, int type, u8 state) + void br_mdb_notify(struct net_device *dev, struct net_bridge_port_group *pg, + int type) { struct br_mdb_entry entry; memset(&entry, 0, sizeof(entry)); - entry.ifindex = port->dev->ifindex; - entry.addr.proto = group->proto; - entry.addr.u.ip4 = group->u.ip4; + entry.ifindex = pg->port->dev->ifindex; + entry.addr.proto = pg->addr.proto; + entry.addr.u.ip4 = pg->addr.u.ip4; #if IS_ENABLED(CONFIG_IPV6) - entry.addr.u.ip6 = group->u.ip6; + entry.addr.u.ip6 = pg->addr.u.ip6; #endif - entry.state = state; - entry.vid = group->vid; - __br_mdb_notify(dev, &entry, type); + entry.vid = pg->addr.vid; + __mdb_entry_fill_flags(&entry, pg->flags); + __br_mdb_notify(dev, &entry, type, pg); } static int nlmsg_populate_rtr_fill(struct sk_buff *skb, @@@ -412,7 -438,8 +438,8 @@@ static int br_mdb_parse(struct sk_buff } static int br_mdb_add_group(struct net_bridge *br, struct net_bridge_port *port, - struct br_ip *group, unsigned char state) + struct br_ip *group, unsigned char state, + struct net_bridge_port_group **pg) { struct net_bridge_mdb_entry *mp; struct net_bridge_port_group *p; @@@ -425,8 -452,8 +452,8 @@@ mp = br_mdb_ip_get(mdb, group); if (!mp) { mp = br_multicast_new_group(br, port, group); - err = PTR_ERR(mp); - if (IS_ERR(mp)) + err = PTR_ERR_OR_ZERO(mp); + if (err) return err; } @@@ -443,6 -470,7 +470,7 @@@ if (unlikely(!p)) return -ENOMEM; rcu_assign_pointer(*pp, p); + *pg = p; if (state == MDB_TEMPORARY) mod_timer(&p->timer, now + br->multicast_membership_interval); @@@ -450,7 -478,8 +478,8 @@@ } static int __br_mdb_add(struct net *net, struct net_bridge *br, - struct br_mdb_entry *entry) + struct br_mdb_entry *entry, + struct net_bridge_port_group **pg) { struct br_ip ip; struct net_device *dev; @@@ -479,7 -508,7 +508,7 @@@ #endif spin_lock_bh(&br->multicast_lock); - ret = br_mdb_add_group(br, p, &ip, entry->state); + ret = br_mdb_add_group(br, p, &ip, entry->state, pg); spin_unlock_bh(&br->multicast_lock); return ret; } @@@ -487,6 -516,7 +516,7 @@@ static int br_mdb_add(struct sk_buff *skb, struct nlmsghdr *nlh) { struct net *net = sock_net(skb->sk); + struct net_bridge_port_group *pg; struct net_bridge_vlan_group *vg; struct net_device *dev, *pdev; struct br_mdb_entry *entry; @@@ -516,15 -546,15 +546,15 @@@ if (br_vlan_enabled(br) && vg && entry->vid == 0) { list_for_each_entry(v, &vg->vlan_list, vlist) { entry->vid = v->vid; - err = __br_mdb_add(net, br, entry); + err = __br_mdb_add(net, br, entry, &pg); if (err) break; - __br_mdb_notify(dev, entry, RTM_NEWMDB); + __br_mdb_notify(dev, entry, RTM_NEWMDB, pg); } } else { - err = __br_mdb_add(net, br, entry); + err = __br_mdb_add(net, br, entry, &pg); if (!err) - __br_mdb_notify(dev, entry, RTM_NEWMDB); + __br_mdb_notify(dev, entry, RTM_NEWMDB, pg); } return err; @@@ -568,7 -598,7 +598,7 @@@ static int __br_mdb_del(struct net_brid if (p->port->state == BR_STATE_DISABLED) goto unlock; - entry->state = p->state; + __mdb_entry_fill_flags(entry, p->flags); rcu_assign_pointer(*pp, p->next); hlist_del_init(&p->mglist); del_timer(&p->timer); @@@ -620,12 -650,12 +650,12 @@@ static int br_mdb_del(struct sk_buff *s entry->vid = v->vid; err = __br_mdb_del(br, entry); if (!err) - __br_mdb_notify(dev, entry, RTM_DELMDB); + __br_mdb_notify(dev, entry, RTM_DELMDB, NULL); } } else { err = __br_mdb_del(br, entry); if (!err) - __br_mdb_notify(dev, entry, RTM_DELMDB); + __br_mdb_notify(dev, entry, RTM_DELMDB, NULL); } return err; diff --combined net/core/dev.c index 0ef061b,3f4071a..edb7179 --- a/net/core/dev.c +++ b/net/core/dev.c @@@ -3829,8 -3829,14 +3829,14 @@@ static void net_tx_action(struct softir trace_consume_skb(skb); else trace_kfree_skb(skb, net_tx_action); - __kfree_skb(skb); + + if (skb->fclone != SKB_FCLONE_UNAVAILABLE) + __kfree_skb(skb); + else + __kfree_skb_defer(skb); } + + __kfree_skb_flush(); } if (sd->output_queue) { @@@ -4154,7 -4160,10 +4160,10 @@@ ncls ret = pt_prev->func(skb, skb->dev, pt_prev, orig_dev); } else { drop: - atomic_long_inc(&skb->dev->rx_dropped); + if (!deliver_exact) + atomic_long_inc(&skb->dev->rx_dropped); + else + atomic_long_inc(&skb->dev->rx_nohandler); kfree_skb(skb); /* Jamal, now you will not able to escape explaining * me how you were going to use this. :-) @@@ -5152,6 -5161,7 +5161,7 @@@ static void net_rx_action(struct softir } } + __kfree_skb_flush(); local_irq_disable(); list_splice_tail_init(&sd->poll_list, &list); @@@ -5379,12 -5389,12 +5389,12 @@@ void *netdev_lower_get_next(struct net_ { struct netdev_adjacent *lower; - lower = list_entry((*iter)->next, struct netdev_adjacent, list); + lower = list_entry(*iter, struct netdev_adjacent, list); if (&lower->list == &dev->adj_list.lower) return NULL; - *iter = &lower->list; + *iter = lower->list.next; return lower->dev; } @@@ -7253,24 -7263,31 +7263,31 @@@ void netdev_run_todo(void } } - /* Convert net_device_stats to rtnl_link_stats64. They have the same - * fields in the same order, with only the type differing. + /* Convert net_device_stats to rtnl_link_stats64. rtnl_link_stats64 has + * all the same fields in the same order as net_device_stats, with only + * the type differing, but rtnl_link_stats64 may have additional fields + * at the end for newer counters. */ void netdev_stats_to_stats64(struct rtnl_link_stats64 *stats64, const struct net_device_stats *netdev_stats) { #if BITS_PER_LONG == 64 - BUILD_BUG_ON(sizeof(*stats64) != sizeof(*netdev_stats)); + BUILD_BUG_ON(sizeof(*stats64) < sizeof(*netdev_stats)); memcpy(stats64, netdev_stats, sizeof(*stats64)); + /* zero out counters that only exist in rtnl_link_stats64 */ + memset((char *)stats64 + sizeof(*netdev_stats), 0, + sizeof(*stats64) - sizeof(*netdev_stats)); #else - size_t i, n = sizeof(*stats64) / sizeof(u64); + size_t i, n = sizeof(*netdev_stats) / sizeof(unsigned long); const unsigned long *src = (const unsigned long *)netdev_stats; u64 *dst = (u64 *)stats64; - BUILD_BUG_ON(sizeof(*netdev_stats) / sizeof(unsigned long) != - sizeof(*stats64) / sizeof(u64)); + BUILD_BUG_ON(n > sizeof(*stats64) / sizeof(u64)); for (i = 0; i < n; i++) dst[i] = src[i]; + /* zero out counters that only exist in rtnl_link_stats64 */ + memset((char *)stats64 + n * sizeof(u64), 0, + sizeof(*stats64) - n * sizeof(u64)); #endif } EXPORT_SYMBOL(netdev_stats_to_stats64); @@@ -7300,6 -7317,7 +7317,7 @@@ struct rtnl_link_stats64 *dev_get_stats } storage->rx_dropped += atomic_long_read(&dev->rx_dropped); storage->tx_dropped += atomic_long_read(&dev->tx_dropped); + storage->rx_nohandler += atomic_long_read(&dev->rx_nohandler); return storage; } EXPORT_SYMBOL(dev_get_stats); @@@ -7422,10 -7440,8 +7440,10 @@@ struct net_device *alloc_netdev_mqs(in dev->priv_flags = IFF_XMIT_DST_RELEASE | IFF_XMIT_DST_RELEASE_PERM; setup(dev); - if (!dev->tx_queue_len) + if (!dev->tx_queue_len) { dev->priv_flags |= IFF_NO_QUEUE; + dev->tx_queue_len = 1; + } dev->num_tx_queues = txqs; dev->real_num_tx_queues = txqs; diff --combined net/core/skbuff.c index 5bf88f5,8bd4b79..488566b --- a/net/core/skbuff.c +++ b/net/core/skbuff.c @@@ -79,8 -79,6 +79,8 @@@ struct kmem_cache *skbuff_head_cache __read_mostly; static struct kmem_cache *skbuff_fclone_cache __read_mostly; +int sysctl_max_skb_frags __read_mostly = MAX_SKB_FRAGS; +EXPORT_SYMBOL(sysctl_max_skb_frags); /** * skb_panic - private function for out-of-line support @@@ -349,8 -347,16 +349,16 @@@ struct sk_buff *build_skb(void *data, u } EXPORT_SYMBOL(build_skb); + #define NAPI_SKB_CACHE_SIZE 64 + + struct napi_alloc_cache { + struct page_frag_cache page; + size_t skb_count; + void *skb_cache[NAPI_SKB_CACHE_SIZE]; + }; + static DEFINE_PER_CPU(struct page_frag_cache, netdev_alloc_cache); - static DEFINE_PER_CPU(struct page_frag_cache, napi_alloc_cache); + static DEFINE_PER_CPU(struct napi_alloc_cache, napi_alloc_cache); static void *__netdev_alloc_frag(unsigned int fragsz, gfp_t gfp_mask) { @@@ -380,9 -386,9 +388,9 @@@ EXPORT_SYMBOL(netdev_alloc_frag) static void *__napi_alloc_frag(unsigned int fragsz, gfp_t gfp_mask) { - struct page_frag_cache *nc = this_cpu_ptr(&napi_alloc_cache); + struct napi_alloc_cache *nc = this_cpu_ptr(&napi_alloc_cache); - return __alloc_page_frag(nc, fragsz, gfp_mask); + return __alloc_page_frag(&nc->page, fragsz, gfp_mask); } void *napi_alloc_frag(unsigned int fragsz) @@@ -476,7 -482,7 +484,7 @@@ EXPORT_SYMBOL(__netdev_alloc_skb) struct sk_buff *__napi_alloc_skb(struct napi_struct *napi, unsigned int len, gfp_t gfp_mask) { - struct page_frag_cache *nc = this_cpu_ptr(&napi_alloc_cache); + struct napi_alloc_cache *nc = this_cpu_ptr(&napi_alloc_cache); struct sk_buff *skb; void *data; @@@ -496,7 -502,7 +504,7 @@@ if (sk_memalloc_socks()) gfp_mask |= __GFP_MEMALLOC; - data = __alloc_page_frag(nc, len, gfp_mask); + data = __alloc_page_frag(&nc->page, len, gfp_mask); if (unlikely(!data)) return NULL; @@@ -507,7 -513,7 +515,7 @@@ } /* use OR instead of assignment to avoid clearing of bits in mask */ - if (nc->pfmemalloc) + if (nc->page.pfmemalloc) skb->pfmemalloc = 1; skb->head_frag = 1; @@@ -749,6 -755,73 +757,73 @@@ void consume_skb(struct sk_buff *skb } EXPORT_SYMBOL(consume_skb); + void __kfree_skb_flush(void) + { + struct napi_alloc_cache *nc = this_cpu_ptr(&napi_alloc_cache); + + /* flush skb_cache if containing objects */ + if (nc->skb_count) { + kmem_cache_free_bulk(skbuff_head_cache, nc->skb_count, + nc->skb_cache); + nc->skb_count = 0; + } + } + + static inline void _kfree_skb_defer(struct sk_buff *skb) + { + struct napi_alloc_cache *nc = this_cpu_ptr(&napi_alloc_cache); + + /* drop skb->head and call any destructors for packet */ + skb_release_all(skb); + + /* record skb to CPU local list */ + nc->skb_cache[nc->skb_count++] = skb; + + #ifdef CONFIG_SLUB + /* SLUB writes into objects when freeing */ + prefetchw(skb); + #endif + + /* flush skb_cache if it is filled */ + if (unlikely(nc->skb_count == NAPI_SKB_CACHE_SIZE)) { + kmem_cache_free_bulk(skbuff_head_cache, NAPI_SKB_CACHE_SIZE, + nc->skb_cache); + nc->skb_count = 0; + } + } + void __kfree_skb_defer(struct sk_buff *skb) + { + _kfree_skb_defer(skb); + } + + void napi_consume_skb(struct sk_buff *skb, int budget) + { + if (unlikely(!skb)) + return; + + /* if budget is 0 assume netpoll w/ IRQs disabled */ + if (unlikely(!budget)) { + dev_consume_skb_irq(skb); + return; + } + + if (likely(atomic_read(&skb->users) == 1)) + smp_rmb(); + else if (likely(!atomic_dec_and_test(&skb->users))) + return; + /* if reaching here SKB is ready to free */ + trace_consume_skb(skb); + + /* if SKB is a clone, don't handle this case */ + if (unlikely(skb->fclone != SKB_FCLONE_UNAVAILABLE)) { + __kfree_skb(skb); + return; + } + + _kfree_skb_defer(skb); + } + EXPORT_SYMBOL(napi_consume_skb); + /* Make sure a field is enclosed inside headers_start/headers_end section */ #define CHECK_SKB_FIELD(field) \ BUILD_BUG_ON(offsetof(struct sk_buff, field) < \ @@@ -3006,8 -3079,7 +3081,7 @@@ struct sk_buff *skb_segment(struct sk_b if (unlikely(!proto)) return ERR_PTR(-EINVAL); - csum = !head_skb->encap_hdr_csum && - !!can_checksum_protocol(features, proto); + csum = !!can_checksum_protocol(features, proto); headroom = skb_headroom(head_skb); pos = skb_headlen(head_skb); @@@ -3100,13 -3172,15 +3174,15 @@@ if (nskb->len == len + doffset) goto perform_csum_check; - if (!sg && !nskb->remcsum_offload) { - nskb->ip_summed = CHECKSUM_NONE; - nskb->csum = skb_copy_and_csum_bits(head_skb, offset, - skb_put(nskb, len), - len, 0); + if (!sg) { + if (!nskb->remcsum_offload) + nskb->ip_summed = CHECKSUM_NONE; + SKB_GSO_CB(nskb)->csum = + skb_copy_and_csum_bits(head_skb, offset, + skb_put(nskb, len), + len, 0); SKB_GSO_CB(nskb)->csum_start = - skb_headroom(nskb) + doffset; + skb_headroom(nskb) + doffset; continue; } @@@ -3172,12 -3246,19 +3248,19 @@@ skip_fraglist nskb->truesize += nskb->data_len; perform_csum_check: - if (!csum && !nskb->remcsum_offload) { - nskb->csum = skb_checksum(nskb, doffset, - nskb->len - doffset, 0); - nskb->ip_summed = CHECKSUM_NONE; + if (!csum) { + if (skb_has_shared_frag(nskb)) { + err = __skb_linearize(nskb); + if (err) + goto err; + } + if (!nskb->remcsum_offload) + nskb->ip_summed = CHECKSUM_NONE; + SKB_GSO_CB(nskb)->csum = + skb_checksum(nskb, doffset, + nskb->len - doffset, 0); SKB_GSO_CB(nskb)->csum_start = - skb_headroom(nskb) + doffset; + skb_headroom(nskb) + doffset; } } while ((offset += len) < head_skb->len); @@@ -4415,9 -4496,7 +4498,7 @@@ int skb_vlan_push(struct sk_buff *skb, skb->mac_len += VLAN_HLEN; __skb_pull(skb, offset); - if (skb->ip_summed == CHECKSUM_COMPLETE) - skb->csum = csum_add(skb->csum, csum_partial(skb->data - + (2 * ETH_ALEN), VLAN_HLEN, 0)); + skb_postpush_rcsum(skb, skb->data + (2 * ETH_ALEN), VLAN_HLEN); } __vlan_hwaccel_put_tag(skb, vlan_proto, vlan_tci); return 0; diff --combined net/dccp/ipv4.c index 902d606,1e0c600..b5672e5 --- a/net/dccp/ipv4.c +++ b/net/dccp/ipv4.c @@@ -802,7 -802,7 +802,7 @@@ static int dccp_v4_rcv(struct sk_buff * } lookup: - sk = __inet_lookup_skb(&dccp_hashinfo, skb, + sk = __inet_lookup_skb(&dccp_hashinfo, skb, __dccp_hdr_len(dh), dh->dccph_sport, dh->dccph_dport); if (!sk) { dccp_pr_debug("failed to look up flow ID in table and " @@@ -824,26 -824,26 +824,26 @@@ if (sk->sk_state == DCCP_NEW_SYN_RECV) { struct request_sock *req = inet_reqsk(sk); - struct sock *nsk = NULL; + struct sock *nsk; sk = req->rsk_listener; - if (likely(sk->sk_state == DCCP_LISTEN)) { - nsk = dccp_check_req(sk, skb, req); - } else { + if (unlikely(sk->sk_state != DCCP_LISTEN)) { inet_csk_reqsk_queue_drop_and_put(sk, req); goto lookup; } + sock_hold(sk); + nsk = dccp_check_req(sk, skb, req); if (!nsk) { reqsk_put(req); - goto discard_it; + goto discard_and_relse; } if (nsk == sk) { - sock_hold(sk); reqsk_put(req); } else if (dccp_child_process(sk, nsk, skb)) { dccp_v4_ctl_send_reset(sk, skb); - goto discard_it; + goto discard_and_relse; } else { + sock_put(sk); return 0; } } diff --combined net/dccp/ipv6.c index b8608b7,45cbe85..4663a01 --- a/net/dccp/ipv6.c +++ b/net/dccp/ipv6.c @@@ -668,7 -668,7 +668,7 @@@ static int dccp_v6_rcv(struct sk_buff * DCCP_SKB_CB(skb)->dccpd_ack_seq = dccp_hdr_ack_seq(skb); lookup: - sk = __inet6_lookup_skb(&dccp_hashinfo, skb, + sk = __inet6_lookup_skb(&dccp_hashinfo, skb, __dccp_hdr_len(dh), dh->dccph_sport, dh->dccph_dport, inet6_iif(skb)); if (!sk) { @@@ -691,26 -691,26 +691,26 @@@ if (sk->sk_state == DCCP_NEW_SYN_RECV) { struct request_sock *req = inet_reqsk(sk); - struct sock *nsk = NULL; + struct sock *nsk; sk = req->rsk_listener; - if (likely(sk->sk_state == DCCP_LISTEN)) { - nsk = dccp_check_req(sk, skb, req); - } else { + if (unlikely(sk->sk_state != DCCP_LISTEN)) { inet_csk_reqsk_queue_drop_and_put(sk, req); goto lookup; } + sock_hold(sk); + nsk = dccp_check_req(sk, skb, req); if (!nsk) { reqsk_put(req); - goto discard_it; + goto discard_and_relse; } if (nsk == sk) { - sock_hold(sk); reqsk_put(req); } else if (dccp_child_process(sk, nsk, skb)) { dccp_v6_ctl_send_reset(sk, skb); - goto discard_it; + goto discard_and_relse; } else { + sock_put(sk); return 0; } } @@@ -993,7 -993,7 +993,7 @@@ static struct proto dccp_v6_prot = .sendmsg = dccp_sendmsg, .recvmsg = dccp_recvmsg, .backlog_rcv = dccp_v6_do_rcv, - .hash = inet_hash, + .hash = inet6_hash, .unhash = inet_unhash, .accept = inet_csk_accept, .get_port = inet_csk_get_port, diff --combined net/ipv4/devinet.c index f6303b1,3d83531..29b8d3a --- a/net/ipv4/devinet.c +++ b/net/ipv4/devinet.c @@@ -1847,7 -1847,7 +1847,7 @@@ static int inet_netconf_get_devconf(str if (err < 0) goto errout; - err = EINVAL; + err = -EINVAL; if (!tb[NETCONFA_IFINDEX]) goto errout; @@@ -2185,6 -2185,8 +2185,8 @@@ static struct devinet_sysctl_table "igmpv3_unsolicited_report_interval"), DEVINET_SYSCTL_RW_ENTRY(IGNORE_ROUTES_WITH_LINKDOWN, "ignore_routes_with_linkdown"), + DEVINET_SYSCTL_RW_ENTRY(DROP_GRATUITOUS_ARP, + "drop_gratuitous_arp"), DEVINET_SYSCTL_FLUSHING_ENTRY(NOXFRM, "disable_xfrm"), DEVINET_SYSCTL_FLUSHING_ENTRY(NOPOLICY, "disable_policy"), @@@ -2192,6 -2194,8 +2194,8 @@@ "promote_secondaries"), DEVINET_SYSCTL_FLUSHING_ENTRY(ROUTE_LOCALNET, "route_localnet"), + DEVINET_SYSCTL_FLUSHING_ENTRY(DROP_UNICAST_IN_L2_MULTICAST, + "drop_unicast_in_l2_multicast"), }, }; diff --combined net/ipv4/inet_connection_sock.c index 6414891,3d28c6d..d768230 --- a/net/ipv4/inet_connection_sock.c +++ b/net/ipv4/inet_connection_sock.c @@@ -24,6 -24,7 +24,7 @@@ #include <net/tcp_states.h> #include <net/xfrm.h> #include <net/tcp.h> + #include <net/sock_reuseport.h> #ifdef INET_CSK_DEBUG const char inet_csk_timer_bug_msg[] = "inet_csk BUG: unknown timer value\n"; @@@ -67,7 -68,8 +68,8 @@@ int inet_csk_bind_conflict(const struc if ((!reuse || !sk2->sk_reuse || sk2->sk_state == TCP_LISTEN) && (!reuseport || !sk2->sk_reuseport || - (sk2->sk_state != TCP_TIME_WAIT && + rcu_access_pointer(sk->sk_reuseport_cb) || + (sk2->sk_state != TCP_TIME_WAIT && !uid_eq(uid, sock_i_uid(sk2))))) { if (!sk2->sk_rcv_saddr || !sk->sk_rcv_saddr || @@@ -89,161 -91,153 +91,153 @@@ EXPORT_SYMBOL_GPL(inet_csk_bind_conflic /* Obtain a reference to a local port for the given sock, * if snum is zero it means select any available local port. + * We try to allocate an odd port (and leave even ports for connect()) */ int inet_csk_get_port(struct sock *sk, unsigned short snum) { - struct inet_hashinfo *hashinfo = sk->sk_prot->h.hashinfo; + bool reuse = sk->sk_reuse && sk->sk_state != TCP_LISTEN; + struct inet_hashinfo *hinfo = sk->sk_prot->h.hashinfo; + int ret = 1, attempts = 5, port = snum; + int smallest_size = -1, smallest_port; struct inet_bind_hashbucket *head; - struct inet_bind_bucket *tb; - int ret, attempts = 5; struct net *net = sock_net(sk); - int smallest_size = -1, smallest_rover; + int i, low, high, attempt_half; + struct inet_bind_bucket *tb; kuid_t uid = sock_i_uid(sk); - int attempt_half = (sk->sk_reuse == SK_CAN_REUSE) ? 1 : 0; + u32 remaining, offset; - local_bh_disable(); - if (!snum) { - int remaining, rover, low, high; + if (port) { + have_port: + head = &hinfo->bhash[inet_bhashfn(net, port, + hinfo->bhash_size)]; + spin_lock_bh(&head->lock); + inet_bind_bucket_for_each(tb, &head->chain) + if (net_eq(ib_net(tb), net) && tb->port == port) + goto tb_found; + goto tb_not_found; + } again: - inet_get_local_port_range(net, &low, &high); - if (attempt_half) { - int half = low + ((high - low) >> 1); - - if (attempt_half == 1) - high = half; - else - low = half; - } - remaining = (high - low) + 1; - smallest_rover = rover = prandom_u32() % remaining + low; - - smallest_size = -1; - do { - if (inet_is_local_reserved_port(net, rover)) - goto next_nolock; - head = &hashinfo->bhash[inet_bhashfn(net, rover, - hashinfo->bhash_size)]; - spin_lock(&head->lock); - inet_bind_bucket_for_each(tb, &head->chain) - if (net_eq(ib_net(tb), net) && tb->port == rover) { - if (((tb->fastreuse > 0 && - sk->sk_reuse && - sk->sk_state != TCP_LISTEN) || - (tb->fastreuseport > 0 && - sk->sk_reuseport && - uid_eq(tb->fastuid, uid))) && - (tb->num_owners < smallest_size || smallest_size == -1)) { - smallest_size = tb->num_owners; - smallest_rover = rover; - } - if (!inet_csk(sk)->icsk_af_ops->bind_conflict(sk, tb, false)) { - snum = rover; - goto tb_found; - } - goto next; + attempt_half = (sk->sk_reuse == SK_CAN_REUSE) ? 1 : 0; + other_half_scan: + inet_get_local_port_range(net, &low, &high); + high++; /* [32768, 60999] -> [32768, 61000[ */ + if (high - low < 4) + attempt_half = 0; + if (attempt_half) { + int half = low + (((high - low) >> 2) << 1); + + if (attempt_half == 1) + high = half; + else + low = half; + } + remaining = high - low; + if (likely(remaining > 1)) + remaining &= ~1U; + + offset = prandom_u32() % remaining; + /* __inet_hash_connect() favors ports having @low parity + * We do the opposite to not pollute connect() users. + */ + offset |= 1U; + smallest_size = -1; + smallest_port = low; /* avoid compiler warning */ + + other_parity_scan: + port = low + offset; + for (i = 0; i < remaining; i += 2, port += 2) { + if (unlikely(port >= high)) + port -= remaining; + if (inet_is_local_reserved_port(net, port)) + continue; + head = &hinfo->bhash[inet_bhashfn(net, port, + hinfo->bhash_size)]; + spin_lock_bh(&head->lock); + inet_bind_bucket_for_each(tb, &head->chain) + if (net_eq(ib_net(tb), net) && tb->port == port) { + if (((tb->fastreuse > 0 && reuse) || + (tb->fastreuseport > 0 && + sk->sk_reuseport && + !rcu_access_pointer(sk->sk_reuseport_cb) && + uid_eq(tb->fastuid, uid))) && + (tb->num_owners < smallest_size || smallest_size == -1)) { + smallest_size = tb->num_owners; + smallest_port = port; } - break; - next: - spin_unlock(&head->lock); - next_nolock: - if (++rover > high) - rover = low; - } while (--remaining > 0); - - /* Exhausted local port range during search? It is not - * possible for us to be holding one of the bind hash - * locks if this test triggers, because if 'remaining' - * drops to zero, we broke out of the do/while loop at - * the top level, not from the 'break;' statement. - */ - ret = 1; - if (remaining <= 0) { - if (smallest_size != -1) { - snum = smallest_rover; - goto have_snum; - } - if (attempt_half == 1) { - /* OK we now try the upper half of the range */ - attempt_half = 2; - goto again; + if (!inet_csk(sk)->icsk_af_ops->bind_conflict(sk, tb, false)) + goto tb_found; + goto next_port; } - goto fail; - } - /* OK, here is the one we will use. HEAD is - * non-NULL and we hold it's mutex. - */ - snum = rover; - } else { - have_snum: - head = &hashinfo->bhash[inet_bhashfn(net, snum, - hashinfo->bhash_size)]; - spin_lock(&head->lock); - inet_bind_bucket_for_each(tb, &head->chain) - if (net_eq(ib_net(tb), net) && tb->port == snum) - goto tb_found; + goto tb_not_found; + next_port: + spin_unlock_bh(&head->lock); + cond_resched(); + } + + if (smallest_size != -1) { + port = smallest_port; + goto have_port; } - tb = NULL; - goto tb_not_found; + offset--; + if (!(offset & 1)) + goto other_parity_scan; + + if (attempt_half == 1) { + /* OK we now try the upper half of the range */ + attempt_half = 2; + goto other_half_scan; + } + return ret; + + tb_not_found: + tb = inet_bind_bucket_create(hinfo->bind_bucket_cachep, + net, head, port); + if (!tb) + goto fail_unlock; tb_found: if (!hlist_empty(&tb->owners)) { if (sk->sk_reuse == SK_FORCE_REUSE) goto success; - if (((tb->fastreuse > 0 && - sk->sk_reuse && sk->sk_state != TCP_LISTEN) || + if (((tb->fastreuse > 0 && reuse) || (tb->fastreuseport > 0 && sk->sk_reuseport && uid_eq(tb->fastuid, uid))) && - smallest_size == -1) { + smallest_size == -1) goto success; - } else { - ret = 1; - if (inet_csk(sk)->icsk_af_ops->bind_conflict(sk, tb, true)) { - if (((sk->sk_reuse && sk->sk_state != TCP_LISTEN) || - (tb->fastreuseport > 0 && - sk->sk_reuseport && uid_eq(tb->fastuid, uid))) && - smallest_size != -1 && --attempts >= 0) { - spin_unlock(&head->lock); - goto again; - } - - goto fail_unlock; + if (inet_csk(sk)->icsk_af_ops->bind_conflict(sk, tb, true)) { + if ((reuse || + (tb->fastreuseport > 0 && + sk->sk_reuseport && + !rcu_access_pointer(sk->sk_reuseport_cb) && + uid_eq(tb->fastuid, uid))) && + smallest_size != -1 && --attempts >= 0) { + spin_unlock_bh(&head->lock); + goto again; } + goto fail_unlock; } - } - tb_not_found: - ret = 1; - if (!tb && (tb = inet_bind_bucket_create(hashinfo->bind_bucket_cachep, - net, head, snum)) == NULL) - goto fail_unlock; - if (hlist_empty(&tb->owners)) { - if (sk->sk_reuse && sk->sk_state != TCP_LISTEN) - tb->fastreuse = 1; - else + if (!reuse) tb->fastreuse = 0; + if (!sk->sk_reuseport || !uid_eq(tb->fastuid, uid)) + tb->fastreuseport = 0; + } else { + tb->fastreuse = reuse; if (sk->sk_reuseport) { tb->fastreuseport = 1; tb->fastuid = uid; - } else - tb->fastreuseport = 0; - } else { - if (tb->fastreuse && - (!sk->sk_reuse || sk->sk_state == TCP_LISTEN)) - tb->fastreuse = 0; - if (tb->fastreuseport && - (!sk->sk_reuseport || !uid_eq(tb->fastuid, uid))) + } else { tb->fastreuseport = 0; + } } success: if (!inet_csk(sk)->icsk_bind_hash) - inet_bind_hash(sk, tb, snum); + inet_bind_hash(sk, tb, port); WARN_ON(inet_csk(sk)->icsk_bind_hash != tb); ret = 0; fail_unlock: - spin_unlock(&head->lock); - fail: - local_bh_enable(); + spin_unlock_bh(&head->lock); return ret; } EXPORT_SYMBOL_GPL(inet_csk_get_port); @@@ -482,10 -476,6 +476,6 @@@ EXPORT_SYMBOL_GPL(inet_csk_route_child_ #define AF_INET_FAMILY(fam) true #endif - /* Only thing we need from tcp.h */ - extern int sysctl_tcp_synack_retries; - - /* Decide when to expire the request and when to resend SYN-ACK */ static inline void syn_ack_recalc(struct request_sock *req, const int thresh, const int max_retries, @@@ -557,6 -547,7 +547,7 @@@ static void reqsk_timer_handler(unsigne { struct request_sock *req = (struct request_sock *)data; struct sock *sk_listener = req->rsk_listener; + struct net *net = sock_net(sk_listener); struct inet_connection_sock *icsk = inet_csk(sk_listener); struct request_sock_queue *queue = &icsk->icsk_accept_queue; int qlen, expire = 0, resend = 0; @@@ -566,7 -557,7 +557,7 @@@ if (sk_state_load(sk_listener) != TCP_LISTEN) goto drop; - max_retries = icsk->icsk_syn_retries ? : sysctl_tcp_synack_retries; + max_retries = icsk->icsk_syn_retries ? : net->ipv4.sysctl_tcp_synack_retries; thresh = max_retries; /* Normally all the openreqs are young and become mature * (i.e. converted to established socket) for first timeout. @@@ -737,6 -728,7 +728,7 @@@ int inet_csk_listen_start(struct sock * { struct inet_connection_sock *icsk = inet_csk(sk); struct inet_sock *inet = inet_sk(sk); + int err = -EADDRINUSE; reqsk_queue_alloc(&icsk->icsk_accept_queue); @@@ -754,13 -746,14 +746,14 @@@ inet->inet_sport = htons(inet->inet_num); sk_dst_reset(sk); - sk->sk_prot->hash(sk); + err = sk->sk_prot->hash(sk); - return 0; + if (likely(!err)) + return 0; } sk->sk_state = TCP_CLOSE; - return -EADDRINUSE; + return err; } EXPORT_SYMBOL_GPL(inet_csk_listen_start); @@@ -789,16 -782,14 +782,16 @@@ static void inet_child_forget(struct so reqsk_put(req); } -void inet_csk_reqsk_queue_add(struct sock *sk, struct request_sock *req, - struct sock *child) +struct sock *inet_csk_reqsk_queue_add(struct sock *sk, + struct request_sock *req, + struct sock *child) { struct request_sock_queue *queue = &inet_csk(sk)->icsk_accept_queue; spin_lock(&queue->rskq_lock); if (unlikely(sk->sk_state != TCP_LISTEN)) { inet_child_forget(sk, req, child); + child = NULL; } else { req->sk = child; req->dl_next = NULL; @@@ -810,7 -801,6 +803,7 @@@ sk_acceptq_added(sk); } spin_unlock(&queue->rskq_lock); + return child; } EXPORT_SYMBOL(inet_csk_reqsk_queue_add); @@@ -820,8 -810,11 +813,8 @@@ struct sock *inet_csk_complete_hashdanc if (own_req) { inet_csk_reqsk_queue_drop(sk, req); reqsk_queue_removed(&inet_csk(sk)->icsk_accept_queue, req); - inet_csk_reqsk_queue_add(sk, req, child); - /* Warning: caller must not call reqsk_put(req); - * child stole last reference on it. - */ - return child; + if (inet_csk_reqsk_queue_add(sk, req, child)) + return child; } /* Too bad, another child took ownership of the request, undo. */ bh_unlock_sock(child); diff --combined net/ipv4/ip_gre.c index 41ba68d,12071e2..202437d --- a/net/ipv4/ip_gre.c +++ b/net/ipv4/ip_gre.c @@@ -238,7 -238,7 +238,7 @@@ static int parse_gre_header(struct sk_b return -EINVAL; } } - return iptunnel_pull_header(skb, hdr_len, tpi->proto); + return iptunnel_pull_header(skb, hdr_len, tpi->proto, false); } static void ipgre_err(struct sk_buff *skb, u32 info, @@@ -440,6 -440,17 +440,17 @@@ drop return 0; } + static __sum16 gre_checksum(struct sk_buff *skb) + { + __wsum csum; + + if (skb->ip_summed == CHECKSUM_PARTIAL) + csum = lco_csum(skb); + else + csum = skb_checksum(skb, 0, skb->len, 0); + return csum_fold(csum); + } + static void build_header(struct sk_buff *skb, int hdr_len, __be16 flags, __be16 proto, __be32 key, __be32 seq) { @@@ -467,8 -478,7 +478,7 @@@ !(skb_shinfo(skb)->gso_type & (SKB_GSO_GRE | SKB_GSO_GRE_CSUM))) { *ptr = 0; - *(__sum16 *)ptr = csum_fold(skb_checksum(skb, 0, - skb->len, 0)); + *(__sum16 *)ptr = gre_checksum(skb); } } } @@@ -493,8 -503,7 +503,7 @@@ static void __gre_xmit(struct sk_buff * static struct sk_buff *gre_handle_offloads(struct sk_buff *skb, bool csum) { - return iptunnel_handle_offloads(skb, csum, - csum ? SKB_GSO_GRE_CSUM : SKB_GSO_GRE); + return iptunnel_handle_offloads(skb, csum ? SKB_GSO_GRE_CSUM : SKB_GSO_GRE); } static struct rtable *gre_get_rt(struct sk_buff *skb, @@@ -531,9 -540,16 +540,16 @@@ static void gre_fb_xmit(struct sk_buff goto err_free_skb; key = &tun_info->key; - rt = gre_get_rt(skb, dev, &fl, key); - if (IS_ERR(rt)) - goto err_free_skb; + rt = !skb->mark ? dst_cache_get_ip4(&tun_info->dst_cache, &fl.saddr) : + NULL; + if (!rt) { + rt = gre_get_rt(skb, dev, &fl, key); + if (IS_ERR(rt)) + goto err_free_skb; + if (!skb->mark) + dst_cache_set_ip4(&tun_info->dst_cache, &rt->dst, + fl.saddr); + } tunnel_hlen = ip_gre_calc_hlen(key->tun_flags); @@@ -1054,9 -1070,8 +1070,9 @@@ static const struct net_device_ops gre_ static void ipgre_tap_setup(struct net_device *dev) { ether_setup(dev); - dev->netdev_ops = &gre_tap_netdev_ops; - dev->priv_flags |= IFF_LIVE_ADDR_CHANGE; + dev->netdev_ops = &gre_tap_netdev_ops; + dev->priv_flags &= ~IFF_TX_SKB_SHARING; + dev->priv_flags |= IFF_LIVE_ADDR_CHANGE; ip_tunnel_setup(dev, gre_tap_net_id); } @@@ -1241,14 -1256,6 +1257,14 @@@ struct net_device *gretap_fb_dev_create err = ipgre_newlink(net, dev, tb, NULL); if (err < 0) goto out; + + /* openvswitch users expect packet sizes to be unrestricted, + * so set the largest MTU we can. + */ + err = __ip_tunnel_change_mtu(dev, IP_MAX_MTU, false); + if (err) + goto out; + return dev; out: free_netdev(dev); diff --combined net/ipv4/ip_sockglue.c index a501242,3f1befc..035ad64 --- a/net/ipv4/ip_sockglue.c +++ b/net/ipv4/ip_sockglue.c @@@ -249,8 -249,6 +249,8 @@@ int ip_cmsg_send(struct net *net, struc switch (cmsg->cmsg_type) { case IP_RETOPTS: err = cmsg->cmsg_len - CMSG_ALIGN(sizeof(struct cmsghdr)); + + /* Our caller is responsible for freeing ipc->opt */ err = ip_options_get(net, &ipc->opt, CMSG_DATA(cmsg), err < 40 ? err : 40); if (err) @@@ -573,6 -571,7 +573,7 @@@ static int do_ip_setsockopt(struct soc int optname, char __user *optval, unsigned int optlen) { struct inet_sock *inet = inet_sk(sk); + struct net *net = sock_net(sk); int val = 0, err; bool needs_rtnl = setsockopt_needs_rtnl(optname); @@@ -912,7 -911,7 +913,7 @@@ } /* numsrc >= (1G-4) overflow in 32 bits */ if (msf->imsf_numsrc >= 0x3ffffffcU || - msf->imsf_numsrc > sysctl_igmp_max_msf) { + msf->imsf_numsrc > net->ipv4.sysctl_igmp_max_msf) { kfree(msf); err = -ENOBUFS; break; @@@ -1067,7 -1066,7 +1068,7 @@@ /* numsrc >= (4G-140)/128 overflow in 32 bits */ if (gsf->gf_numsrc >= 0x1ffffff || - gsf->gf_numsrc > sysctl_igmp_max_msf) { + gsf->gf_numsrc > net->ipv4.sysctl_igmp_max_msf) { err = -ENOBUFS; goto mc_msf_out; } @@@ -1342,10 -1341,13 +1343,13 @@@ static int do_ip_getsockopt(struct soc val = inet->tos; break; case IP_TTL: + { + struct net *net = sock_net(sk); val = (inet->uc_ttl == -1 ? - sysctl_ip_default_ttl : + net->ipv4.sysctl_ip_default_ttl : inet->uc_ttl); break; + } case IP_HDRINCL: val = inet->hdrincl; break; diff --combined net/ipv4/ip_tunnel.c index 89e8861,4569da7..dff8a05 --- a/net/ipv4/ip_tunnel.c +++ b/net/ipv4/ip_tunnel.c @@@ -68,61 -68,6 +68,6 @@@ static unsigned int ip_tunnel_hash(__be IP_TNL_HASH_BITS); } - static void __tunnel_dst_set(struct ip_tunnel_dst *idst, - struct dst_entry *dst, __be32 saddr) - { - struct dst_entry *old_dst; - - dst_clone(dst); - old_dst = xchg((__force struct dst_entry **)&idst->dst, dst); - dst_release(old_dst); - idst->saddr = saddr; - } - - static noinline void tunnel_dst_set(struct ip_tunnel *t, - struct dst_entry *dst, __be32 saddr) - { - __tunnel_dst_set(raw_cpu_ptr(t->dst_cache), dst, saddr); - } - - static void tunnel_dst_reset(struct ip_tunnel *t) - { - tunnel_dst_set(t, NULL, 0); - } - - void ip_tunnel_dst_reset_all(struct ip_tunnel *t) - { - int i; - - for_each_possible_cpu(i) - __tunnel_dst_set(per_cpu_ptr(t->dst_cache, i), NULL, 0); - } - EXPORT_SYMBOL(ip_tunnel_dst_reset_all); - - static struct rtable *tunnel_rtable_get(struct ip_tunnel *t, - u32 cookie, __be32 *saddr) - { - struct ip_tunnel_dst *idst; - struct dst_entry *dst; - - rcu_read_lock(); - idst = raw_cpu_ptr(t->dst_cache); - dst = rcu_dereference(idst->dst); - if (dst && !atomic_inc_not_zero(&dst->__refcnt)) - dst = NULL; - if (dst) { - if (!dst->obsolete || dst->ops->check(dst, cookie)) { - *saddr = idst->saddr; - } else { - tunnel_dst_reset(t); - dst_release(dst); - dst = NULL; - } - } - rcu_read_unlock(); - return (struct rtable *)dst; - } - static bool ip_tunnel_key_match(const struct ip_tunnel_parm *p, __be16 flags, __be32 key) { @@@ -381,7 -326,8 +326,8 @@@ static int ip_tunnel_bind_dev(struct ne if (!IS_ERR(rt)) { tdev = rt->dst.dev; - tunnel_dst_set(tunnel, &rt->dst, fl4.saddr); + dst_cache_set_ip4(&tunnel->dst_cache, &rt->dst, + fl4.saddr); ip_rt_put(rt); } if (dev->type != ARPHRD_ETHER) @@@ -729,7 -675,8 +675,8 @@@ void ip_tunnel_xmit(struct sk_buff *skb if (ip_tunnel_encap(skb, tunnel, &protocol, &fl4) < 0) goto tx_error; - rt = connected ? tunnel_rtable_get(tunnel, 0, &fl4.saddr) : NULL; + rt = connected ? dst_cache_get_ip4(&tunnel->dst_cache, &fl4.saddr) : + NULL; if (!rt) { rt = ip_route_output_key(tunnel->net, &fl4); @@@ -739,7 -686,8 +686,8 @@@ goto tx_error; } if (connected) - tunnel_dst_set(tunnel, &rt->dst, fl4.saddr); + dst_cache_set_ip4(&tunnel->dst_cache, &rt->dst, + fl4.saddr); } if (rt->dst.dev == dev) { @@@ -836,7 -784,7 +784,7 @@@ static void ip_tunnel_update(struct ip_ if (set_mtu) dev->mtu = mtu; } - ip_tunnel_dst_reset_all(t); + dst_cache_reset(&t->dst_cache); netdev_state_change(dev); } @@@ -943,31 -891,17 +891,31 @@@ done } EXPORT_SYMBOL_GPL(ip_tunnel_ioctl); -int ip_tunnel_change_mtu(struct net_device *dev, int new_mtu) +int __ip_tunnel_change_mtu(struct net_device *dev, int new_mtu, bool strict) { struct ip_tunnel *tunnel = netdev_priv(dev); int t_hlen = tunnel->hlen + sizeof(struct iphdr); + int max_mtu = 0xFFF8 - dev->hard_header_len - t_hlen; - if (new_mtu < 68 || - new_mtu > 0xFFF8 - dev->hard_header_len - t_hlen) + if (new_mtu < 68) return -EINVAL; + + if (new_mtu > max_mtu) { + if (strict) + return -EINVAL; + + new_mtu = max_mtu; + } + dev->mtu = new_mtu; return 0; } +EXPORT_SYMBOL_GPL(__ip_tunnel_change_mtu); + +int ip_tunnel_change_mtu(struct net_device *dev, int new_mtu) +{ + return __ip_tunnel_change_mtu(dev, new_mtu, true); +} EXPORT_SYMBOL_GPL(ip_tunnel_change_mtu); static void ip_tunnel_dev_free(struct net_device *dev) @@@ -975,7 -909,7 +923,7 @@@ struct ip_tunnel *tunnel = netdev_priv(dev); gro_cells_destroy(&tunnel->gro_cells); - free_percpu(tunnel->dst_cache); + dst_cache_destroy(&tunnel->dst_cache); free_percpu(dev->tstats); free_netdev(dev); } @@@ -1169,15 -1103,15 +1117,15 @@@ int ip_tunnel_init(struct net_device *d if (!dev->tstats) return -ENOMEM; - tunnel->dst_cache = alloc_percpu(struct ip_tunnel_dst); - if (!tunnel->dst_cache) { + err = dst_cache_init(&tunnel->dst_cache, GFP_KERNEL); + if (err) { free_percpu(dev->tstats); - return -ENOMEM; + return err; } err = gro_cells_init(&tunnel->gro_cells, dev); if (err) { - free_percpu(tunnel->dst_cache); + dst_cache_destroy(&tunnel->dst_cache); free_percpu(dev->tstats); return err; } @@@ -1207,7 -1141,7 +1155,7 @@@ void ip_tunnel_uninit(struct net_devic if (itn->fb_tunnel_dev != dev) ip_tunnel_del(itn, netdev_priv(dev)); - ip_tunnel_dst_reset_all(tunnel); + dst_cache_reset(&tunnel->dst_cache); } EXPORT_SYMBOL_GPL(ip_tunnel_uninit); diff --combined net/ipv4/ping.c index d3a2716,f6f93fc..76dce90 --- a/net/ipv4/ping.c +++ b/net/ipv4/ping.c @@@ -145,10 -145,12 +145,12 @@@ fail } EXPORT_SYMBOL_GPL(ping_get_port); - void ping_hash(struct sock *sk) + int ping_hash(struct sock *sk) { pr_debug("ping_hash(sk->port=%u)\n", inet_sk(sk)->inet_num); BUG(); /* "Please do not press this button again." */ + + return 0; } void ping_unhash(struct sock *sk) @@@ -746,10 -748,8 +748,10 @@@ static int ping_v4_sendmsg(struct sock if (msg->msg_controllen) { err = ip_cmsg_send(sock_net(sk), msg, &ipc, false); - if (err) + if (unlikely(err)) { + kfree(ipc.opt); return err; + } if (ipc.opt) free = 1; } diff --combined net/ipv4/raw.c index 7113bae,d635251..8d22de7 --- a/net/ipv4/raw.c +++ b/net/ipv4/raw.c @@@ -93,7 -93,7 +93,7 @@@ static struct raw_hashinfo raw_v4_hashi .lock = __RW_LOCK_UNLOCKED(raw_v4_hashinfo.lock), }; - void raw_hash_sk(struct sock *sk) + int raw_hash_sk(struct sock *sk) { struct raw_hashinfo *h = sk->sk_prot->h.raw_hash; struct hlist_head *head; @@@ -104,6 -104,8 +104,8 @@@ sk_add_node(sk, head); sock_prot_inuse_add(sock_net(sk), sk->sk_prot, 1); write_unlock_bh(&h->lock); + + return 0; } EXPORT_SYMBOL_GPL(raw_hash_sk); @@@ -547,10 -549,8 +549,10 @@@ static int raw_sendmsg(struct sock *sk if (msg->msg_controllen) { err = ip_cmsg_send(net, msg, &ipc, false); - if (err) + if (unlikely(err)) { + kfree(ipc.opt); goto out; + } if (ipc.opt) free = 1; } diff --combined net/ipv4/tcp.c index 483ffdf,f93150d..f9faadb --- a/net/ipv4/tcp.c +++ b/net/ipv4/tcp.c @@@ -282,8 -282,6 +282,6 @@@ #include <asm/unaligned.h> #include <net/busy_poll.h> - int sysctl_tcp_fin_timeout __read_mostly = TCP_FIN_TIMEOUT; - int sysctl_tcp_min_tso_segs __read_mostly = 2; int sysctl_tcp_autocorking __read_mostly = 1; @@@ -406,7 -404,7 +404,7 @@@ void tcp_init_sock(struct sock *sk tp->mss_cache = TCP_MSS_DEFAULT; u64_stats_init(&tp->syncp); - tp->reordering = sysctl_tcp_reordering; + tp->reordering = sock_net(sk)->ipv4.sysctl_tcp_reordering; tcp_enable_early_retrans(tp); tcp_assign_congestion_control(sk); @@@ -940,7 -938,7 +938,7 @@@ new_segment i = skb_shinfo(skb)->nr_frags; can_coalesce = skb_can_coalesce(skb, i, page, offset); - if (!can_coalesce && i >= MAX_SKB_FRAGS) { + if (!can_coalesce && i >= sysctl_max_skb_frags) { tcp_mark_push(tp, skb); goto new_segment; } @@@ -1213,7 -1211,7 +1211,7 @@@ new_segment if (!skb_can_coalesce(skb, i, pfrag->page, pfrag->offset)) { - if (i == MAX_SKB_FRAGS || !sg) { + if (i == sysctl_max_skb_frags || !sg) { tcp_mark_push(tp, skb); goto new_segment; } @@@ -1466,8 -1464,10 +1464,10 @@@ static struct sk_buff *tcp_recv_skb(str while ((skb = skb_peek(&sk->sk_receive_queue)) != NULL) { offset = seq - TCP_SKB_CB(skb)->seq; - if (TCP_SKB_CB(skb)->tcp_flags & TCPHDR_SYN) + if (unlikely(TCP_SKB_CB(skb)->tcp_flags & TCPHDR_SYN)) { + pr_err_once("%s: found a SYN, please report !\n", __func__); offset--; + } if (offset < skb->len || (TCP_SKB_CB(skb)->tcp_flags & TCPHDR_FIN)) { *off = offset; return skb; @@@ -1657,8 -1657,10 +1657,10 @@@ int tcp_recvmsg(struct sock *sk, struc break; offset = *seq - TCP_SKB_CB(skb)->seq; - if (TCP_SKB_CB(skb)->tcp_flags & TCPHDR_SYN) + if (unlikely(TCP_SKB_CB(skb)->tcp_flags & TCPHDR_SYN)) { + pr_err_once("%s: found a SYN, please report !\n", __func__); offset--; + } if (offset < skb->len) goto found_ok_skb; if (TCP_SKB_CB(skb)->tcp_flags & TCPHDR_FIN) @@@ -2326,6 -2328,7 +2328,7 @@@ static int do_tcp_setsockopt(struct soc { struct tcp_sock *tp = tcp_sk(sk); struct inet_connection_sock *icsk = inet_csk(sk); + struct net *net = sock_net(sk); int val; int err = 0; @@@ -2522,7 -2525,7 +2525,7 @@@ case TCP_LINGER2: if (val < 0) tp->linger2 = -1; - else if (val > sysctl_tcp_fin_timeout / HZ) + else if (val > net->ipv4.sysctl_tcp_fin_timeout / HZ) tp->linger2 = 0; else tp->linger2 = val * HZ; @@@ -2639,6 -2642,7 +2642,7 @@@ void tcp_get_info(struct sock *sk, stru const struct inet_connection_sock *icsk = inet_csk(sk); u32 now = tcp_time_stamp; unsigned int start; + int notsent_bytes; u64 rate64; u32 rate; @@@ -2719,6 -2723,11 +2723,11 @@@ } while (u64_stats_fetch_retry_irq(&tp->syncp, start)); info->tcpi_segs_out = tp->segs_out; info->tcpi_segs_in = tp->segs_in; + + notsent_bytes = READ_ONCE(tp->write_seq) - READ_ONCE(tp->snd_nxt); + info->tcpi_notsent_bytes = max(0, notsent_bytes); + + info->tcpi_min_rtt = tcp_min_rtt(tp); } EXPORT_SYMBOL_GPL(tcp_get_info); @@@ -2727,6 -2736,7 +2736,7 @@@ static int do_tcp_getsockopt(struct soc { struct inet_connection_sock *icsk = inet_csk(sk); struct tcp_sock *tp = tcp_sk(sk); + struct net *net = sock_net(sk); int val, len; if (get_user(len, optlen)) @@@ -2761,12 -2771,12 +2771,12 @@@ val = keepalive_probes(tp); break; case TCP_SYNCNT: - val = icsk->icsk_syn_retries ? : sysctl_tcp_syn_retries; + val = icsk->icsk_syn_retries ? : net->ipv4.sysctl_tcp_syn_retries; break; case TCP_LINGER2: val = tp->linger2; if (val >= 0) - val = (val ? : sysctl_tcp_fin_timeout) / HZ; + val = (val ? : net->ipv4.sysctl_tcp_fin_timeout) / HZ; break; case TCP_DEFER_ACCEPT: val = retrans_to_secs(icsk->icsk_accept_queue.rskq_defer_accept, @@@ -2950,7 -2960,7 +2960,7 @@@ static void __tcp_alloc_md5sig_pool(voi struct crypto_hash *hash; hash = crypto_alloc_hash("md5", 0, CRYPTO_ALG_ASYNC); - if (IS_ERR_OR_NULL(hash)) + if (IS_ERR(hash)) return; per_cpu(tcp_md5sig_pool, cpu).md5_desc.tfm = hash; } diff --combined net/ipv4/tcp_input.c index 3b2c8e9,5ee6fe0..e6e65f7 --- a/net/ipv4/tcp_input.c +++ b/net/ipv4/tcp_input.c @@@ -80,9 -80,7 +80,7 @@@ int sysctl_tcp_timestamps __read_mostl int sysctl_tcp_window_scaling __read_mostly = 1; int sysctl_tcp_sack __read_mostly = 1; int sysctl_tcp_fack __read_mostly = 1; - int sysctl_tcp_reordering __read_mostly = TCP_FASTRETRANS_THRESH; int sysctl_tcp_max_reordering __read_mostly = 300; - EXPORT_SYMBOL(sysctl_tcp_reordering); int sysctl_tcp_dsack __read_mostly = 1; int sysctl_tcp_app_win __read_mostly = 31; int sysctl_tcp_adv_win_scale __read_mostly = 1; @@@ -126,6 -124,10 +124,10 @@@ int sysctl_tcp_invalid_ratelimit __read #define TCP_REMNANT (TCP_FLAG_FIN|TCP_FLAG_URG|TCP_FLAG_SYN|TCP_FLAG_PSH) #define TCP_HP_BITS (~(TCP_RESERVED_BITS|TCP_FLAG_PSH)) + #define REXMIT_NONE 0 /* no loss recovery to do */ + #define REXMIT_LOST 1 /* retransmit packets marked lost */ + #define REXMIT_NEW 2 /* FRTO-style transmit of unsent/new packets */ + /* Adapt the MSS value used to make delayed ack decision to the * real world. */ @@@ -1210,6 -1212,7 +1212,7 @@@ static u8 tcp_sacktag_one(struct sock * sacked |= TCPCB_SACKED_ACKED; state->flag |= FLAG_DATA_SACKED; tp->sacked_out += pcount; + tp->delivered += pcount; /* Out-of-order packets delivered */ fack_count += pcount; @@@ -1821,8 -1824,12 +1824,12 @@@ static void tcp_check_reno_reordering(s static void tcp_add_reno_sack(struct sock *sk) { struct tcp_sock *tp = tcp_sk(sk); + u32 prior_sacked = tp->sacked_out; + tp->sacked_out++; tcp_check_reno_reordering(sk, 0); + if (tp->sacked_out > prior_sacked) + tp->delivered++; /* Some out-of-order packet is delivered */ tcp_verify_left_out(tp); } @@@ -1834,6 -1841,7 +1841,7 @@@ static void tcp_remove_reno_sacks(struc if (acked > 0) { /* One ACK acked hole. The rest eat duplicate ACKs. */ + tp->delivered += max_t(int, acked - tp->sacked_out, 1); if (acked - 1 >= tp->sacked_out) tp->sacked_out = 0; else @@@ -1873,6 -1881,7 +1881,7 @@@ void tcp_enter_loss(struct sock *sk { const struct inet_connection_sock *icsk = inet_csk(sk); struct tcp_sock *tp = tcp_sk(sk); + struct net *net = sock_net(sk); struct sk_buff *skb; bool new_recovery = icsk->icsk_ca_state < TCP_CA_Recovery; bool is_reneg; /* is receiver reneging on SACKs? */ @@@ -1923,9 -1932,9 +1932,9 @@@ * suggests that the degree of reordering is over-estimated. */ if (icsk->icsk_ca_state <= TCP_CA_Disorder && - tp->sacked_out >= sysctl_tcp_reordering) + tp->sacked_out >= net->ipv4.sysctl_tcp_reordering) tp->reordering = min_t(unsigned int, tp->reordering, - sysctl_tcp_reordering); + net->ipv4.sysctl_tcp_reordering); tcp_set_ca_state(sk, TCP_CA_Loss); tp->high_seq = tp->snd_nxt; tcp_ecn_queue_cwr(tp); @@@ -2109,6 -2118,7 +2118,7 @@@ static bool tcp_time_to_recover(struct { struct tcp_sock *tp = tcp_sk(sk); __u32 packets_out; + int tcp_reordering = sock_net(sk)->ipv4.sysctl_tcp_reordering; /* Trick#1: The loss is proven. */ if (tp->lost_out) @@@ -2123,7 -2133,7 +2133,7 @@@ */ packets_out = tp->packets_out; if (packets_out <= tp->reordering && - tp->sacked_out >= max_t(__u32, packets_out/2, sysctl_tcp_reordering) && + tp->sacked_out >= max_t(__u32, packets_out/2, tcp_reordering) && !tcp_may_send_now(sk)) { /* We have nothing to send. This connection is limited * either by receiver window or by application. @@@ -2467,14 -2477,12 +2477,12 @@@ static void tcp_init_cwnd_reduction(str tcp_ecn_queue_cwr(tp); } - static void tcp_cwnd_reduction(struct sock *sk, const int prior_unsacked, - int fast_rexmit, int flag) + static void tcp_cwnd_reduction(struct sock *sk, int newly_acked_sacked, + int flag) { struct tcp_sock *tp = tcp_sk(sk); int sndcnt = 0; int delta = tp->snd_ssthresh - tcp_packets_in_flight(tp); - int newly_acked_sacked = prior_unsacked - - (tp->packets_out - tp->sacked_out); if (newly_acked_sacked <= 0 || WARN_ON_ONCE(!tp->prior_cwnd)) return; @@@ -2492,7 -2500,8 +2500,8 @@@ } else { sndcnt = min(delta, newly_acked_sacked); } - sndcnt = max(sndcnt, (fast_rexmit ? 1 : 0)); + /* Force a fast retransmit upon entering fast recovery */ + sndcnt = max(sndcnt, (tp->prr_out ? 0 : 1)); tp->snd_cwnd = tcp_packets_in_flight(tp) + sndcnt; } @@@ -2537,7 -2546,7 +2546,7 @@@ static void tcp_try_keep_open(struct so } } - static void tcp_try_to_open(struct sock *sk, int flag, const int prior_unsacked) + static void tcp_try_to_open(struct sock *sk, int flag) { struct tcp_sock *tp = tcp_sk(sk); @@@ -2551,8 -2560,6 +2560,6 @@@ if (inet_csk(sk)->icsk_ca_state != TCP_CA_CWR) { tcp_try_keep_open(sk); - } else { - tcp_cwnd_reduction(sk, prior_unsacked, 0, flag); } } @@@ -2662,7 -2669,8 +2669,8 @@@ static void tcp_enter_recovery(struct s /* Process an ACK in CA_Loss state. Move to CA_Open if lost data are * recovered or spurious. Otherwise retransmits more on partial ACKs. */ - static void tcp_process_loss(struct sock *sk, int flag, bool is_dupack) + static void tcp_process_loss(struct sock *sk, int flag, bool is_dupack, + int *rexmit) { struct tcp_sock *tp = tcp_sk(sk); bool recovered = !before(tp->snd_una, tp->high_seq); @@@ -2684,10 -2692,15 +2692,15 @@@ tp->frto = 0; /* Step 3.a. loss was real */ } else if (flag & FLAG_SND_UNA_ADVANCED && !recovered) { tp->high_seq = tp->snd_nxt; - __tcp_push_pending_frames(sk, tcp_current_mss(sk), - TCP_NAGLE_OFF); - if (after(tp->snd_nxt, tp->high_seq)) - return; /* Step 2.b */ + /* Step 2.b. Try send new data (but deferred until cwnd + * is updated in tcp_ack()). Otherwise fall back to + * the conventional recovery. + */ + if (tcp_send_head(sk) && + after(tcp_wnd_end(tp), tp->snd_nxt)) { + *rexmit = REXMIT_NEW; + return; + } tp->frto = 0; } } @@@ -2706,12 -2719,11 +2719,11 @@@ else if (flag & FLAG_SND_UNA_ADVANCED) tcp_reset_reno_sack(tp); } - tcp_xmit_retransmit_queue(sk); + *rexmit = REXMIT_LOST; } /* Undo during fast recovery after partial ACK. */ - static bool tcp_try_undo_partial(struct sock *sk, const int acked, - const int prior_unsacked, int flag) + static bool tcp_try_undo_partial(struct sock *sk, const int acked) { struct tcp_sock *tp = tcp_sk(sk); @@@ -2726,10 -2738,8 +2738,8 @@@ * can undo. Otherwise we clock out new packets but do not * mark more packets lost or retransmit more. */ - if (tp->retrans_out) { - tcp_cwnd_reduction(sk, prior_unsacked, 0, flag); + if (tp->retrans_out) return true; - } if (!tcp_any_retrans_done(sk)) tp->retrans_stamp = 0; @@@ -2748,21 -2758,21 +2758,21 @@@ * taking into account both packets sitting in receiver's buffer and * packets lost by network. * - * Besides that it does CWND reduction, when packet loss is detected - * and changes state of machine. + * Besides that it updates the congestion state when packet loss or ECN + * is detected. But it does not reduce the cwnd, it is done by the + * congestion control later. * * It does _not_ decide what to send, it is made in function * tcp_xmit_retransmit_queue(). */ static void tcp_fastretrans_alert(struct sock *sk, const int acked, - const int prior_unsacked, - bool is_dupack, int flag) + bool is_dupack, int *ack_flag, int *rexmit) { struct inet_connection_sock *icsk = inet_csk(sk); struct tcp_sock *tp = tcp_sk(sk); + int fast_rexmit = 0, flag = *ack_flag; bool do_lost = is_dupack || ((flag & FLAG_DATA_SACKED) && (tcp_fackets_out(tp) > tp->reordering)); - int fast_rexmit = 0; if (WARN_ON(!tp->packets_out && tp->sacked_out)) tp->sacked_out = 0; @@@ -2809,8 -2819,10 +2819,10 @@@ /* Use RACK to detect loss */ if (sysctl_tcp_recovery & TCP_RACK_LOST_RETRANS && - tcp_rack_mark_lost(sk)) + tcp_rack_mark_lost(sk)) { flag |= FLAG_LOST_RETRANS; + *ack_flag |= FLAG_LOST_RETRANS; + } /* E. Process state. */ switch (icsk->icsk_ca_state) { @@@ -2819,7 -2831,7 +2831,7 @@@ if (tcp_is_reno(tp) && is_dupack) tcp_add_reno_sack(sk); } else { - if (tcp_try_undo_partial(sk, acked, prior_unsacked, flag)) + if (tcp_try_undo_partial(sk, acked)) return; /* Partial ACK arrived. Force fast retransmit. */ do_lost = tcp_is_reno(tp) || @@@ -2831,7 -2843,7 +2843,7 @@@ } break; case TCP_CA_Loss: - tcp_process_loss(sk, flag, is_dupack); + tcp_process_loss(sk, flag, is_dupack, rexmit); if (icsk->icsk_ca_state != TCP_CA_Open && !(flag & FLAG_LOST_RETRANS)) return; @@@ -2848,7 -2860,7 +2860,7 @@@ tcp_try_undo_dsack(sk); if (!tcp_time_to_recover(sk, flag)) { - tcp_try_to_open(sk, flag, prior_unsacked); + tcp_try_to_open(sk, flag); return; } @@@ -2870,8 -2882,7 +2882,7 @@@ if (do_lost) tcp_update_scoreboard(sk, fast_rexmit); - tcp_cwnd_reduction(sk, prior_unsacked, fast_rexmit, flag); - tcp_xmit_retransmit_queue(sk); + *rexmit = REXMIT_LOST; } /* Kathleen Nichols' algorithm for tracking the minimum value of @@@ -2896,10 -2907,7 +2907,10 @@@ static void tcp_update_rtt_min(struct s { const u32 now = tcp_time_stamp, wlen = sysctl_tcp_min_rtt_wlen * HZ; struct rtt_meas *m = tcp_sk(sk)->rtt_min; - struct rtt_meas rttm = { .rtt = (rtt_us ? : 1), .ts = now }; + struct rtt_meas rttm = { + .rtt = likely(rtt_us) ? rtt_us : jiffies_to_usecs(1), + .ts = now, + }; u32 elapsed; /* Check if the new measurement updates the 1st, 2nd, or 3rd choices */ @@@ -3096,7 -3104,7 +3107,7 @@@ static void tcp_ack_tstamp(struct sock * arrived at the other end. */ static int tcp_clean_rtx_queue(struct sock *sk, int prior_fackets, - u32 prior_snd_una, + u32 prior_snd_una, int *acked, struct tcp_sacktag_state *sack) { const struct inet_connection_sock *icsk = inet_csk(sk); @@@ -3154,10 -3162,13 +3165,13 @@@ flag |= FLAG_ORIG_SACK_ACKED; } - if (sacked & TCPCB_SACKED_ACKED) + if (sacked & TCPCB_SACKED_ACKED) { tp->sacked_out -= acked_pcount; - else if (tcp_is_sack(tp) && !tcp_skb_spurious_retrans(tp, skb)) - tcp_rack_advance(tp, &skb->skb_mstamp, sacked); + } else if (tcp_is_sack(tp)) { + tp->delivered += acked_pcount; + if (!tcp_skb_spurious_retrans(tp, skb)) + tcp_rack_advance(tp, &skb->skb_mstamp, sacked); + } if (sacked & TCPCB_LOST) tp->lost_out -= acked_pcount; @@@ -3266,6 -3277,7 +3280,7 @@@ } } #endif + *acked = pkts_acked; return flag; } @@@ -3299,21 -3311,36 +3314,36 @@@ static inline bool tcp_ack_is_dubious(c /* Decide wheather to run the increase function of congestion control. */ static inline bool tcp_may_raise_cwnd(const struct sock *sk, const int flag) { - if (tcp_in_cwnd_reduction(sk)) - return false; - /* If reordering is high then always grow cwnd whenever data is * delivered regardless of its ordering. Otherwise stay conservative * and only grow cwnd on in-order delivery (RFC5681). A stretched ACK w/ * new SACK or ECE mark may first advance cwnd here and later reduce * cwnd in tcp_fastretrans_alert() based on more states. */ - if (tcp_sk(sk)->reordering > sysctl_tcp_reordering) + if (tcp_sk(sk)->reordering > sock_net(sk)->ipv4.sysctl_tcp_reordering) return flag & FLAG_FORWARD_PROGRESS; return flag & FLAG_DATA_ACKED; } + /* The "ultimate" congestion control function that aims to replace the rigid + * cwnd increase and decrease control (tcp_cong_avoid,tcp_*cwnd_reduction). + * It's called toward the end of processing an ACK with precise rate + * information. All transmission or retransmission are delayed afterwards. + */ + static void tcp_cong_control(struct sock *sk, u32 ack, u32 acked_sacked, + int flag) + { + if (tcp_in_cwnd_reduction(sk)) { + /* Reduce cwnd if state mandates */ + tcp_cwnd_reduction(sk, acked_sacked, flag); + } else if (tcp_may_raise_cwnd(sk, flag)) { + /* Advance cwnd if state allows */ + tcp_cong_avoid(sk, ack, acked_sacked); + } + tcp_update_pacing_rate(sk); + } + /* Check that window update is acceptable. * The function assumes that snd_una<=ack<=snd_next. */ @@@ -3509,6 -3536,27 +3539,27 @@@ static inline void tcp_in_ack_event(str icsk->icsk_ca_ops->in_ack_event(sk, flags); } + /* Congestion control has updated the cwnd already. So if we're in + * loss recovery then now we do any new sends (for FRTO) or + * retransmits (for CA_Loss or CA_recovery) that make sense. + */ + static void tcp_xmit_recovery(struct sock *sk, int rexmit) + { + struct tcp_sock *tp = tcp_sk(sk); + + if (rexmit == REXMIT_NONE) + return; + + if (unlikely(rexmit == 2)) { + __tcp_push_pending_frames(sk, tcp_current_mss(sk), + TCP_NAGLE_OFF); + if (after(tp->snd_nxt, tp->high_seq)) + return; + tp->frto = 0; + } + tcp_xmit_retransmit_queue(sk); + } + /* This routine deals with incoming acks, but not outgoing ones. */ static int tcp_ack(struct sock *sk, const struct sk_buff *skb, int flag) { @@@ -3521,8 -3569,9 +3572,9 @@@ bool is_dupack = false; u32 prior_fackets; int prior_packets = tp->packets_out; - const int prior_unsacked = tp->packets_out - tp->sacked_out; + u32 prior_delivered = tp->delivered; int acked = 0; /* Number of packets newly acked */ + int rexmit = REXMIT_NONE; /* Flag to (re)transmit to recover losses */ sack_state.first_sackt.v64 = 0; @@@ -3611,23 -3660,16 +3663,16 @@@ goto no_queue; /* See if we can take anything off of the retransmit queue. */ - acked = tp->packets_out; - flag |= tcp_clean_rtx_queue(sk, prior_fackets, prior_snd_una, + flag |= tcp_clean_rtx_queue(sk, prior_fackets, prior_snd_una, &acked, &sack_state); - acked -= tp->packets_out; if (tcp_ack_is_dubious(sk, flag)) { is_dupack = !(flag & (FLAG_SND_UNA_ADVANCED | FLAG_NOT_DUP)); - tcp_fastretrans_alert(sk, acked, prior_unsacked, - is_dupack, flag); + tcp_fastretrans_alert(sk, acked, is_dupack, &flag, &rexmit); } if (tp->tlp_high_seq) tcp_process_tlp_ack(sk, ack, flag); - /* Advance cwnd if state allows */ - if (tcp_may_raise_cwnd(sk, flag)) - tcp_cong_avoid(sk, ack, acked); - if ((flag & FLAG_FORWARD_PROGRESS) || !(flag & FLAG_NOT_DUP)) { struct dst_entry *dst = __sk_dst_get(sk); if (dst) @@@ -3636,14 -3678,14 +3681,14 @@@ if (icsk->icsk_pending == ICSK_TIME_RETRANS) tcp_schedule_loss_probe(sk); - tcp_update_pacing_rate(sk); + tcp_cong_control(sk, ack, tp->delivered - prior_delivered, flag); + tcp_xmit_recovery(sk, rexmit); return 1; no_queue: /* If data was DSACKed, see if we can undo a cwnd reduction. */ if (flag & FLAG_DSACKING_ACK) - tcp_fastretrans_alert(sk, acked, prior_unsacked, - is_dupack, flag); + tcp_fastretrans_alert(sk, acked, is_dupack, &flag, &rexmit); /* If this ack opens up a zero window, clear backoff. It was * being used to time the probes, and is probably far higher than * it needs to be for normal retransmission. @@@ -3666,8 -3708,8 +3711,8 @@@ old_ack if (TCP_SKB_CB(skb)->sacked) { flag |= tcp_sacktag_write_queue(sk, skb, prior_snd_una, &sack_state); - tcp_fastretrans_alert(sk, acked, prior_unsacked, - is_dupack, flag); + tcp_fastretrans_alert(sk, acked, is_dupack, &flag, &rexmit); + tcp_xmit_recovery(sk, rexmit); } SOCK_DEBUG(sk, "Ack %u before %u:%u\n", ack, tp->snd_una, tp->snd_nxt); @@@ -3998,7 -4040,7 +4043,7 @@@ void tcp_reset(struct sock *sk * * If we are in FINWAIT-2, a received FIN moves us to TIME-WAIT. */ - static void tcp_fin(struct sock *sk) + void tcp_fin(struct sock *sk) { struct tcp_sock *tp = tcp_sk(sk); @@@ -5512,6 -5554,9 +5557,9 @@@ static bool tcp_rcv_fastopen_synack(str tp->syn_data_acked = tp->syn_data; if (tp->syn_data_acked) NET_INC_STATS_BH(sock_net(sk), LINUX_MIB_TCPFASTOPENACTIVE); + + tcp_fastopen_add_skb(sk, synack); + return false; } @@@ -6118,9 -6163,10 +6166,10 @@@ static bool tcp_syn_flood_action(const struct request_sock_queue *queue = &inet_csk(sk)->icsk_accept_queue; const char *msg = "Dropping request"; bool want_cookie = false; + struct net *net = sock_net(sk); #ifdef CONFIG_SYN_COOKIES - if (sysctl_tcp_syncookies) { + if (net->ipv4.sysctl_tcp_syncookies) { msg = "Sending cookies"; want_cookie = true; NET_INC_STATS_BH(sock_net(sk), LINUX_MIB_TCPREQQFULLDOCOOKIES); @@@ -6129,7 -6175,7 +6178,7 @@@ NET_INC_STATS_BH(sock_net(sk), LINUX_MIB_TCPREQQFULLDROP); if (!queue->synflood_warned && - sysctl_tcp_syncookies != 2 && + net->ipv4.sysctl_tcp_syncookies != 2 && xchg(&queue->synflood_warned, 1) == 0) pr_info("%s: Possible SYN flooding on port %d. %s. Check SNMP counters.\n", proto, ntohs(tcp_hdr(skb)->dest), msg); @@@ -6162,6 -6208,7 +6211,7 @@@ int tcp_conn_request(struct request_soc __u32 isn = TCP_SKB_CB(skb)->tcp_tw_isn; struct tcp_options_received tmp_opt; struct tcp_sock *tp = tcp_sk(sk); + struct net *net = sock_net(sk); struct sock *fastopen_sk = NULL; struct dst_entry *dst = NULL; struct request_sock *req; @@@ -6172,7 -6219,7 +6222,7 @@@ * limitations, they conserve resources and peer is * evidently real one. */ - if ((sysctl_tcp_syncookies == 2 || + if ((net->ipv4.sysctl_tcp_syncookies == 2 || inet_csk_reqsk_queue_is_full(sk)) && !isn) { want_cookie = tcp_syn_flood_action(sk, skb, rsk_ops->slab_name); if (!want_cookie) @@@ -6238,7 -6285,7 +6288,7 @@@ } } /* Kill the following clause, if you dislike this way. */ - else if (!sysctl_tcp_syncookies && + else if (!net->ipv4.sysctl_tcp_syncookies && (sysctl_max_syn_backlog - inet_csk_reqsk_queue_len(sk) < (sysctl_max_syn_backlog >> 2)) && !tcp_peer_is_proven(req, dst, false, diff --combined net/ipv4/tcp_ipv4.c index 487ac67,3f872a6..4c8d58d --- a/net/ipv4/tcp_ipv4.c +++ b/net/ipv4/tcp_ipv4.c @@@ -311,7 -311,7 +311,7 @@@ static void do_redirect(struct sk_buff /* handle ICMP messages on TCP_NEW_SYN_RECV request sockets */ -void tcp_req_err(struct sock *sk, u32 seq) +void tcp_req_err(struct sock *sk, u32 seq, bool abort) { struct request_sock *req = inet_reqsk(sk); struct net *net = sock_net(sk); @@@ -323,7 -323,7 +323,7 @@@ if (seq != tcp_rsk(req)->snt_isn) { NET_INC_STATS_BH(net, LINUX_MIB_OUTOFWINDOWICMPS); - } else { + } else if (abort) { /* * Still in SYN_RECV, just remove it silently. * There is no good way to pass the error to the newly @@@ -383,12 -383,7 +383,12 @@@ void tcp_v4_err(struct sk_buff *icmp_sk } seq = ntohl(th->seq); if (sk->sk_state == TCP_NEW_SYN_RECV) - return tcp_req_err(sk, seq); + return tcp_req_err(sk, seq, + type == ICMP_PARAMETERPROB || + type == ICMP_TIME_EXCEEDED || + (type == ICMP_DEST_UNREACH && + (code == ICMP_NET_UNREACH || + code == ICMP_HOST_UNREACH))); bh_lock_sock(sk); /* If too many ICMPs get dropped on busy @@@ -642,8 -637,8 +642,8 @@@ static void tcp_v4_send_reset(const str * Incoming packet is checked with md5 hash with finding key, * no RST generated if md5 hash doesn't match. */ - sk1 = __inet_lookup_listener(net, - &tcp_hashinfo, ip_hdr(skb)->saddr, + sk1 = __inet_lookup_listener(net, &tcp_hashinfo, NULL, 0, + ip_hdr(skb)->saddr, th->source, ip_hdr(skb)->daddr, ntohs(th->source), inet_iif(skb)); /* don't send rst if it can't find key */ @@@ -865,7 -860,6 +865,6 @@@ static void tcp_v4_reqsk_destructor(str kfree(inet_rsk(req)->opt); } - #ifdef CONFIG_TCP_MD5SIG /* * RFC2385 MD5 checksumming requires a mapping of @@@ -1587,7 -1581,8 +1586,8 @@@ int tcp_v4_rcv(struct sk_buff *skb TCP_SKB_CB(skb)->sacked = 0; lookup: - sk = __inet_lookup_skb(&tcp_hashinfo, skb, th->source, th->dest); + sk = __inet_lookup_skb(&tcp_hashinfo, skb, __tcp_hdrlen(th), th->source, + th->dest); if (!sk) goto no_tcp_socket; @@@ -1597,30 -1592,28 +1597,30 @@@ process if (sk->sk_state == TCP_NEW_SYN_RECV) { struct request_sock *req = inet_reqsk(sk); - struct sock *nsk = NULL; + struct sock *nsk; sk = req->rsk_listener; - if (tcp_v4_inbound_md5_hash(sk, skb)) - goto discard_and_relse; - if (likely(sk->sk_state == TCP_LISTEN)) { - nsk = tcp_check_req(sk, skb, req, false); - } else { + if (unlikely(tcp_v4_inbound_md5_hash(sk, skb))) { + reqsk_put(req); + goto discard_it; + } + if (unlikely(sk->sk_state != TCP_LISTEN)) { inet_csk_reqsk_queue_drop_and_put(sk, req); goto lookup; } + sock_hold(sk); + nsk = tcp_check_req(sk, skb, req, false); if (!nsk) { reqsk_put(req); - goto discard_it; + goto discard_and_relse; } if (nsk == sk) { - sock_hold(sk); reqsk_put(req); } else if (tcp_child_process(sk, nsk, skb)) { tcp_v4_send_reset(nsk, skb); - goto discard_it; + goto discard_and_relse; } else { + sock_put(sk); return 0; } } @@@ -1703,7 -1696,8 +1703,8 @@@ do_time_wait switch (tcp_timewait_state_process(inet_twsk(sk), skb, th)) { case TCP_TW_SYN: { struct sock *sk2 = inet_lookup_listener(dev_net(skb->dev), - &tcp_hashinfo, + &tcp_hashinfo, skb, + __tcp_hdrlen(th), iph->saddr, th->source, iph->daddr, th->dest, inet_iif(skb)); @@@ -2395,6 -2389,16 +2396,16 @@@ static int __net_init tcp_sk_init(struc net->ipv4.sysctl_tcp_keepalive_probes = TCP_KEEPALIVE_PROBES; net->ipv4.sysctl_tcp_keepalive_intvl = TCP_KEEPALIVE_INTVL; + net->ipv4.sysctl_tcp_syn_retries = TCP_SYN_RETRIES; + net->ipv4.sysctl_tcp_synack_retries = TCP_SYNACK_RETRIES; + net->ipv4.sysctl_tcp_syncookies = 1; + net->ipv4.sysctl_tcp_reordering = TCP_FASTRETRANS_THRESH; + net->ipv4.sysctl_tcp_retries1 = TCP_RETR1; + net->ipv4.sysctl_tcp_retries2 = TCP_RETR2; + net->ipv4.sysctl_tcp_orphan_retries = 0; + net->ipv4.sysctl_tcp_fin_timeout = TCP_FIN_TIMEOUT; + net->ipv4.sysctl_tcp_notsent_lowat = UINT_MAX; + return 0; fail: tcp_sk_exit(net); diff --combined net/ipv4/udp.c index 95d2f19,9fc4e9c..836abe5 --- a/net/ipv4/udp.c +++ b/net/ipv4/udp.c @@@ -356,8 -356,8 +356,8 @@@ EXPORT_SYMBOL(udp_lib_get_port) * match_wildcard == false: addresses must be exactly the same, i.e. * 0.0.0.0 only equals to 0.0.0.0 */ - static int ipv4_rcv_saddr_equal(const struct sock *sk1, const struct sock *sk2, - bool match_wildcard) + int ipv4_rcv_saddr_equal(const struct sock *sk1, const struct sock *sk2, + bool match_wildcard) { struct inet_sock *inet1 = inet_sk(sk1), *inet2 = inet_sk(sk2); @@@ -848,32 -848,20 +848,20 @@@ void udp_set_csum(bool nocheck, struct { struct udphdr *uh = udp_hdr(skb); - if (nocheck) + if (nocheck) { uh->check = 0; - else if (skb_is_gso(skb)) + } else if (skb_is_gso(skb)) { uh->check = ~udp_v4_check(len, saddr, daddr, 0); - else if (skb_dst(skb) && skb_dst(skb)->dev && - (skb_dst(skb)->dev->features & - (NETIF_F_IP_CSUM | NETIF_F_HW_CSUM))) { - - BUG_ON(skb->ip_summed == CHECKSUM_PARTIAL); - + } else if (skb->ip_summed == CHECKSUM_PARTIAL) { + uh->check = 0; + uh->check = udp_v4_check(len, saddr, daddr, lco_csum(skb)); + if (uh->check == 0) + uh->check = CSUM_MANGLED_0; + } else { skb->ip_summed = CHECKSUM_PARTIAL; skb->csum_start = skb_transport_header(skb) - skb->head; skb->csum_offset = offsetof(struct udphdr, check); uh->check = ~udp_v4_check(len, saddr, daddr, 0); - } else { - __wsum csum; - - BUG_ON(skb->ip_summed == CHECKSUM_PARTIAL); - - uh->check = 0; - csum = skb_checksum(skb, 0, len, 0); - uh->check = udp_v4_check(len, saddr, daddr, csum); - if (uh->check == 0) - uh->check = CSUM_MANGLED_0; - - skb->ip_summed = CHECKSUM_UNNECESSARY; } } EXPORT_SYMBOL(udp_set_csum); @@@ -1048,10 -1036,8 +1036,10 @@@ int udp_sendmsg(struct sock *sk, struc if (msg->msg_controllen) { err = ip_cmsg_send(sock_net(sk), msg, &ipc, sk->sk_family == AF_INET6); - if (err) + if (unlikely(err)) { + kfree(ipc.opt); return err; + } if (ipc.opt) free = 1; connected = 0; diff --combined net/ipv6/addrconf.c index bdd7eac,ac0ba9e..4751f89 --- a/net/ipv6/addrconf.c +++ b/net/ipv6/addrconf.c @@@ -583,7 -583,7 +583,7 @@@ static int inet6_netconf_get_devconf(st if (err < 0) goto errout; - err = EINVAL; + err = -EINVAL; if (!tb[NETCONFA_IFINDEX]) goto errout; @@@ -3538,7 -3538,6 +3538,7 @@@ static void addrconf_dad_begin(struct i { struct inet6_dev *idev = ifp->idev; struct net_device *dev = idev->dev; + bool notify = false; addrconf_join_solict(dev, &ifp->addr); @@@ -3584,7 -3583,7 +3584,7 @@@ /* Because optimistic nodes can use this address, * notify listeners. If DAD fails, RTM_DELADDR is sent. */ - ipv6_ifa_notify(RTM_NEWADDR, ifp); + notify = true; } } @@@ -3592,8 -3591,6 +3592,8 @@@ out: spin_unlock(&ifp->lock); read_unlock_bh(&idev->lock); + if (notify) + ipv6_ifa_notify(RTM_NEWADDR, ifp); } static void addrconf_dad_start(struct inet6_ifaddr *ifp) @@@ -4714,6 -4711,8 +4714,8 @@@ static inline void ipv6_store_devconf(s array[DEVCONF_IGNORE_ROUTES_WITH_LINKDOWN] = cnf->ignore_routes_with_linkdown; /* we omit DEVCONF_STABLE_SECRET for now */ array[DEVCONF_USE_OIF_ADDRS_ONLY] = cnf->use_oif_addrs_only; + array[DEVCONF_DROP_UNICAST_IN_L2_MULTICAST] = cnf->drop_unicast_in_l2_multicast; + array[DEVCONF_DROP_UNSOLICITED_NA] = cnf->drop_unsolicited_na; } static inline size_t inet6_ifla6_size(void) @@@ -5788,6 -5787,20 +5790,20 @@@ static struct addrconf_sysctl_tabl .proc_handler = addrconf_sysctl_ignore_routes_with_linkdown, }, { + .procname = "drop_unicast_in_l2_multicast", + .data = &ipv6_devconf.drop_unicast_in_l2_multicast, + .maxlen = sizeof(int), + .mode = 0644, + .proc_handler = proc_dointvec, + }, + { + .procname = "drop_unsolicited_na", + .data = &ipv6_devconf.drop_unsolicited_na, + .maxlen = sizeof(int), + .mode = 0644, + .proc_handler = proc_dointvec, + }, + { /* sentinel */ } }, diff --combined net/ipv6/ip6_gre.c index a69aad1,a94e506..f7c9560 --- a/net/ipv6/ip6_gre.c +++ b/net/ipv6/ip6_gre.c @@@ -360,7 -360,7 +360,7 @@@ static void ip6gre_tunnel_uninit(struc struct ip6gre_net *ign = net_generic(t->net, ip6gre_net_id); ip6gre_tunnel_unlink(ign, t); - ip6_tnl_dst_reset(t); + dst_cache_reset(&t->dst_cache); dev_put(dev); } @@@ -633,7 -633,7 +633,7 @@@ static netdev_tx_t ip6gre_xmit2(struct } if (!fl6->flowi6_mark) - dst = ip6_tnl_dst_get(tunnel); + dst = dst_cache_get(&tunnel->dst_cache); if (!dst) { dst = ip6_route_output(net, NULL, fl6); @@@ -702,7 -702,7 +702,7 @@@ } if (!fl6->flowi6_mark && ndst) - ip6_tnl_dst_set(tunnel, ndst); + dst_cache_set_ip6(&tunnel->dst_cache, ndst, &fl6->saddr); skb_dst_set(skb, dst); proto = NEXTHDR_GRE; @@@ -1009,7 -1009,7 +1009,7 @@@ static int ip6gre_tnl_change(struct ip6 t->parms.o_key = p->o_key; t->parms.i_flags = p->i_flags; t->parms.o_flags = p->o_flags; - ip6_tnl_dst_reset(t); + dst_cache_reset(&t->dst_cache); ip6gre_tnl_link_config(t, set_mtu); return 0; } @@@ -1219,7 -1219,7 +1219,7 @@@ static void ip6gre_dev_free(struct net_ { struct ip6_tnl *t = netdev_priv(dev); - ip6_tnl_dst_destroy(t); + dst_cache_destroy(&t->dst_cache); free_percpu(dev->tstats); free_netdev(dev); } @@@ -1257,7 -1257,7 +1257,7 @@@ static int ip6gre_tunnel_init_common(st if (!dev->tstats) return -ENOMEM; - ret = ip6_tnl_dst_init(tunnel); + ret = dst_cache_init(&tunnel->dst_cache, GFP_KERNEL); if (ret) { free_percpu(dev->tstats); dev->tstats = NULL; @@@ -1512,7 -1512,6 +1512,7 @@@ static void ip6gre_tap_setup(struct net dev->destructor = ip6gre_dev_free; dev->features |= NETIF_F_NETNS_LOCAL; + dev->priv_flags &= ~IFF_TX_SKB_SHARING; } static int ip6gre_newlink(struct net *src_net, struct net_device *dev, diff --combined net/ipv6/tcp_ipv6.c index 5c8c842,9977b6f..33f2820 --- a/net/ipv6/tcp_ipv6.c +++ b/net/ipv6/tcp_ipv6.c @@@ -327,7 -327,6 +327,7 @@@ static void tcp_v6_err(struct sk_buff * struct tcp_sock *tp; __u32 seq, snd_una; struct sock *sk; + bool fatal; int err; sk = __inet6_lookup_established(net, &tcp_hashinfo, @@@ -346,9 -345,8 +346,9 @@@ return; } seq = ntohl(th->seq); + fatal = icmpv6_err_convert(type, code, &err); if (sk->sk_state == TCP_NEW_SYN_RECV) - return tcp_req_err(sk, seq); + return tcp_req_err(sk, seq, fatal); bh_lock_sock(sk); if (sock_owned_by_user(sk) && type != ICMPV6_PKT_TOOBIG) @@@ -402,6 -400,7 +402,6 @@@ goto out; } - icmpv6_err_convert(type, code, &err); /* Might be for an request_sock */ switch (sk->sk_state) { @@@ -867,7 -866,8 +867,8 @@@ static void tcp_v6_send_reset(const str * no RST generated if md5 hash doesn't match. */ sk1 = inet6_lookup_listener(dev_net(skb_dst(skb)->dev), - &tcp_hashinfo, &ipv6h->saddr, + &tcp_hashinfo, NULL, 0, + &ipv6h->saddr, th->source, &ipv6h->daddr, ntohs(th->source), tcp_v6_iif(skb)); if (!sk1) @@@ -1376,8 -1376,8 +1377,8 @@@ static int tcp_v6_rcv(struct sk_buff *s hdr = ipv6_hdr(skb); lookup: - sk = __inet6_lookup_skb(&tcp_hashinfo, skb, th->source, th->dest, - inet6_iif(skb)); + sk = __inet6_lookup_skb(&tcp_hashinfo, skb, __tcp_hdrlen(th), + th->source, th->dest, inet6_iif(skb)); if (!sk) goto no_tcp_socket; @@@ -1387,7 -1387,7 +1388,7 @@@ process if (sk->sk_state == TCP_NEW_SYN_RECV) { struct request_sock *req = inet_reqsk(sk); - struct sock *nsk = NULL; + struct sock *nsk; sk = req->rsk_listener; tcp_v6_fill_cb(skb, hdr, th); @@@ -1395,24 -1395,24 +1396,24 @@@ reqsk_put(req); goto discard_it; } - if (likely(sk->sk_state == TCP_LISTEN)) { - nsk = tcp_check_req(sk, skb, req, false); - } else { + if (unlikely(sk->sk_state != TCP_LISTEN)) { inet_csk_reqsk_queue_drop_and_put(sk, req); goto lookup; } + sock_hold(sk); + nsk = tcp_check_req(sk, skb, req, false); if (!nsk) { reqsk_put(req); - goto discard_it; + goto discard_and_relse; } if (nsk == sk) { - sock_hold(sk); reqsk_put(req); tcp_v6_restore_cb(skb); } else if (tcp_child_process(sk, nsk, skb)) { tcp_v6_send_reset(nsk, skb); - goto discard_it; + goto discard_and_relse; } else { + sock_put(sk); return 0; } } @@@ -1501,6 -1501,7 +1502,7 @@@ do_time_wait struct sock *sk2; sk2 = inet6_lookup_listener(dev_net(skb->dev), &tcp_hashinfo, + skb, __tcp_hdrlen(th), &ipv6_hdr(skb)->saddr, th->source, &ipv6_hdr(skb)->daddr, ntohs(th->dest), tcp_v6_iif(skb)); @@@ -1866,7 -1867,7 +1868,7 @@@ struct proto tcpv6_prot = .sendpage = tcp_sendpage, .backlog_rcv = tcp_v6_do_rcv, .release_cb = tcp_release_cb, - .hash = inet_hash, + .hash = inet6_hash, .unhash = inet_unhash, .get_port = inet_csk_get_port, .enter_memory_pressure = tcp_enter_memory_pressure, diff --combined net/netfilter/nfnetlink.c index 857ae89,9a99f68..2278d9a --- a/net/netfilter/nfnetlink.c +++ b/net/netfilter/nfnetlink.c @@@ -127,13 -127,6 +127,6 @@@ int nfnetlink_has_listeners(struct net } EXPORT_SYMBOL_GPL(nfnetlink_has_listeners); - struct sk_buff *nfnetlink_alloc_skb(struct net *net, unsigned int size, - u32 dst_portid, gfp_t gfp_mask) - { - return netlink_alloc_skb(net->nfnl, size, dst_portid, gfp_mask); - } - EXPORT_SYMBOL_GPL(nfnetlink_alloc_skb); - int nfnetlink_send(struct sk_buff *skb, struct net *net, u32 portid, unsigned int group, int echo, gfp_t flags) { @@@ -311,14 -304,14 +304,14 @@@ replay #endif { nfnl_unlock(subsys_id); - netlink_ack(skb, nlh, -EOPNOTSUPP); + netlink_ack(oskb, nlh, -EOPNOTSUPP); return kfree_skb(skb); } } if (!ss->commit || !ss->abort) { nfnl_unlock(subsys_id); - netlink_ack(skb, nlh, -EOPNOTSUPP); + netlink_ack(oskb, nlh, -EOPNOTSUPP); return kfree_skb(skb); } @@@ -328,12 -321,10 +321,12 @@@ nlh = nlmsg_hdr(skb); err = 0; - if (nlmsg_len(nlh) < sizeof(struct nfgenmsg) || - skb->len < nlh->nlmsg_len) { - err = -EINVAL; - goto ack; + if (nlh->nlmsg_len < NLMSG_HDRLEN || + skb->len < nlh->nlmsg_len || + nlmsg_len(nlh) < sizeof(struct nfgenmsg)) { + nfnl_err_reset(&err_list); + status |= NFNL_BATCH_FAILURE; + goto done; } /* Only requests are handled by the kernel */ @@@ -408,7 -399,7 +401,7 @@@ ack * pointing to the batch header. */ nfnl_err_reset(&err_list); - netlink_ack(skb, nlmsg_hdr(oskb), -ENOMEM); + netlink_ack(oskb, nlmsg_hdr(oskb), -ENOMEM); status |= NFNL_BATCH_FAILURE; goto done; } diff --combined net/sched/sch_api.c index af1acf0,c9673b5e..de1e176 --- a/net/sched/sch_api.c +++ b/net/sched/sch_api.c @@@ -1841,7 -1841,7 +1841,7 @@@ reclassify return err; } - return -1; + return TC_ACT_UNSPEC; /* signal: continue lookup */ #ifdef CONFIG_NET_CLS_ACT reset: if (unlikely(limit++ >= MAX_REC_LOOP)) { @@@ -1852,7 -1852,6 +1852,7 @@@ } tp = old_tp; + protocol = tc_skb_protocol(skb); goto reclassify; #endif } diff --combined net/sctp/socket.c index e878da0,6427b9d..b89501e --- a/net/sctp/socket.c +++ b/net/sctp/socket.c @@@ -5538,7 -5538,6 +5538,7 @@@ static int sctp_getsockopt_hmac_ident(s struct sctp_hmac_algo_param *hmacs; __u16 data_len = 0; u32 num_idents; + int i; if (!ep->auth_enable) return -EACCES; @@@ -5556,12 -5555,8 +5556,12 @@@ return -EFAULT; if (put_user(num_idents, &p->shmac_num_idents)) return -EFAULT; - if (copy_to_user(p->shmac_idents, hmacs->hmac_ids, data_len)) - return -EFAULT; + for (i = 0; i < num_idents; i++) { + __u16 hmacid = ntohs(hmacs->hmac_ids[i]); + + if (copy_to_user(&p->shmac_idents[i], &hmacid, sizeof(__u16))) + return -EFAULT; + } return 0; } @@@ -6106,9 -6101,10 +6106,10 @@@ static int sctp_getsockopt(struct sock return retval; } - static void sctp_hash(struct sock *sk) + static int sctp_hash(struct sock *sk) { /* STUB */ + return 0; } static void sctp_unhash(struct sock *sk) diff --combined net/tipc/link.c index 347cdc9,3e513da..e31d92f --- a/net/tipc/link.c +++ b/net/tipc/link.c @@@ -123,7 -123,6 +123,6 @@@ struct tipc_stats struct tipc_link { u32 addr; char name[TIPC_MAX_LINK_NAME]; - struct tipc_media_addr *media_addr; struct net *net; /* Management and link supervision data */ @@@ -904,8 -903,10 +903,10 @@@ int tipc_link_xmit(struct tipc_link *l if (unlikely(l->backlog[i].len >= l->backlog[i].limit)) return link_schedule_user(l, list); } - if (unlikely(msg_size(hdr) > mtu)) + if (unlikely(msg_size(hdr) > mtu)) { + skb_queue_purge(list); return -EMSGSIZE; + } /* Prepare each packet for sending, and add to relevant queue: */ while (skb_queue_len(list)) { @@@ -917,8 -918,10 +918,10 @@@ if (likely(skb_queue_len(transmq) < maxwin)) { _skb = skb_clone(skb, GFP_ATOMIC); - if (!_skb) + if (!_skb) { + skb_queue_purge(list); return -ENOBUFS; + } __skb_dequeue(list); __skb_queue_tail(transmq, skb); __skb_queue_tail(xmitq, _skb); @@@ -1261,26 -1264,6 +1264,6 @@@ drop return rc; } - /* - * Send protocol message to the other endpoint. - */ - static void tipc_link_proto_xmit(struct tipc_link *l, u32 msg_typ, - int probe_msg, u32 gap, u32 tolerance, - u32 priority) - { - struct sk_buff *skb = NULL; - struct sk_buff_head xmitq; - - __skb_queue_head_init(&xmitq); - tipc_link_build_proto_msg(l, msg_typ, probe_msg, gap, - tolerance, priority, &xmitq); - skb = __skb_dequeue(&xmitq); - if (!skb) - return; - tipc_bearer_xmit_skb(l->net, l->bearer_id, skb, l->media_addr); - l->rcv_unacked = 0; - } - static void tipc_link_build_proto_msg(struct tipc_link *l, int mtyp, bool probe, u16 rcvgap, int tolerance, int priority, struct sk_buff_head *xmitq) @@@ -1479,6 -1462,12 +1462,12 @@@ static int tipc_link_proto_rcv(struct t if (in_range(peers_tol, TIPC_MIN_LINK_TOL, TIPC_MAX_LINK_TOL)) l->tolerance = peers_tol; + if (peers_prio && in_range(peers_prio, TIPC_MIN_LINK_PRI, + TIPC_MAX_LINK_PRI)) { + l->priority = peers_prio; + rc = tipc_link_fsm_evt(l, LINK_FAILURE_EVT); + } + l->silent_intv_cnt = 0; l->stats.recv_states++; if (msg_probe(hdr)) @@@ -1973,10 -1962,8 +1962,10 @@@ int tipc_nl_add_bc_link(struct net *net hdr = genlmsg_put(msg->skb, msg->portid, msg->seq, &tipc_genl_family, NLM_F_MULTI, TIPC_NL_LINK_GET); - if (!hdr) + if (!hdr) { + tipc_bcast_unlock(net); return -EMSGSIZE; + } attrs = nla_nest_start(msg->skb, TIPC_NLA_LINK); if (!attrs) @@@ -2023,16 -2010,18 +2012,18 @@@ msg_full return -EMSGSIZE; } - void tipc_link_set_tolerance(struct tipc_link *l, u32 tol) + void tipc_link_set_tolerance(struct tipc_link *l, u32 tol, + struct sk_buff_head *xmitq) { l->tolerance = tol; - tipc_link_proto_xmit(l, STATE_MSG, 0, 0, tol, 0); + tipc_link_build_proto_msg(l, STATE_MSG, 0, 0, tol, 0, xmitq); } - void tipc_link_set_prio(struct tipc_link *l, u32 prio) + void tipc_link_set_prio(struct tipc_link *l, u32 prio, + struct sk_buff_head *xmitq) { l->priority = prio; - tipc_link_proto_xmit(l, STATE_MSG, 0, 0, 0, prio); + tipc_link_build_proto_msg(l, STATE_MSG, 0, 0, 0, prio, xmitq); } void tipc_link_set_abort_limit(struct tipc_link *l, u32 limit) diff --combined net/tipc/node.c index 9d7a16f,10a1e87..9fcc2fb --- a/net/tipc/node.c +++ b/net/tipc/node.c @@@ -346,6 -346,12 +346,6 @@@ struct tipc_node *tipc_node_create(stru skb_queue_head_init(&n->bc_entry.inputq2); for (i = 0; i < MAX_BEARERS; i++) spin_lock_init(&n->links[i].lock); - hlist_add_head_rcu(&n->hash, &tn->node_htable[tipc_hashfn(addr)]); - list_for_each_entry_rcu(temp_node, &tn->node_list, list) { - if (n->addr < temp_node->addr) - break; - } - list_add_tail_rcu(&n->list, &temp_node->list); n->state = SELF_DOWN_PEER_LEAVING; n->signature = INVALID_NODE_SIG; n->active_links[0] = INVALID_BEARER_ID; @@@ -366,12 -372,6 +366,12 @@@ tipc_node_get(n); setup_timer(&n->timer, tipc_node_timeout, (unsigned long)n); n->keepalive_intv = U32_MAX; + hlist_add_head_rcu(&n->hash, &tn->node_htable[tipc_hashfn(addr)]); + list_for_each_entry_rcu(temp_node, &tn->node_list, list) { + if (n->addr < temp_node->addr) + break; + } + list_add_tail_rcu(&n->list, &temp_node->list); exit: spin_unlock_bh(&tn->node_list_lock); return n; @@@ -1166,7 -1166,7 +1166,7 @@@ msg_full * @dnode: address of destination node * @selector: a number used for deterministic link selection * Consumes the buffer chain, except when returning -ELINKCONG - * Returns 0 if success, otherwise errno: -ELINKCONG,-EHOSTUNREACH,-EMSGSIZE + * Returns 0 if success, otherwise: -ELINKCONG,-EHOSTUNREACH,-EMSGSIZE,-ENOBUF */ int tipc_node_xmit(struct net *net, struct sk_buff_head *list, u32 dnode, int selector) @@@ -1174,33 -1174,43 +1174,43 @@@ struct tipc_link_entry *le = NULL; struct tipc_node *n; struct sk_buff_head xmitq; - int bearer_id = -1; - int rc = -EHOSTUNREACH; + int bearer_id; + int rc; + + if (in_own_node(net, dnode)) { + tipc_sk_rcv(net, list); + return 0; + } - __skb_queue_head_init(&xmitq); n = tipc_node_find(net, dnode); - if (likely(n)) { - tipc_node_read_lock(n); - bearer_id = n->active_links[selector & 1]; - if (bearer_id >= 0) { - le = &n->links[bearer_id]; - spin_lock_bh(&le->lock); - rc = tipc_link_xmit(le->link, list, &xmitq); - spin_unlock_bh(&le->lock); - } + if (unlikely(!n)) { + skb_queue_purge(list); + return -EHOSTUNREACH; + } + + tipc_node_read_lock(n); + bearer_id = n->active_links[selector & 1]; + if (unlikely(bearer_id == INVALID_BEARER_ID)) { tipc_node_read_unlock(n); - if (likely(!rc)) - tipc_bearer_xmit(net, bearer_id, &xmitq, &le->maddr); - else if (rc == -ENOBUFS) - tipc_node_link_down(n, bearer_id, false); tipc_node_put(n); - return rc; + skb_queue_purge(list); + return -EHOSTUNREACH; } - if (likely(in_own_node(net, dnode))) { - tipc_sk_rcv(net, list); - return 0; - } + __skb_queue_head_init(&xmitq); + le = &n->links[bearer_id]; + spin_lock_bh(&le->lock); + rc = tipc_link_xmit(le->link, list, &xmitq); + spin_unlock_bh(&le->lock); + tipc_node_read_unlock(n); + + if (likely(rc == 0)) + tipc_bearer_xmit(net, bearer_id, &xmitq, &le->maddr); + else if (rc == -ENOBUFS) + tipc_node_link_down(n, bearer_id, false); + + tipc_node_put(n); + return rc; } @@@ -1637,9 -1647,12 +1647,12 @@@ int tipc_nl_node_set_link(struct sk_buf char *name; struct tipc_link *link; struct tipc_node *node; + struct sk_buff_head xmitq; struct nlattr *attrs[TIPC_NLA_LINK_MAX + 1]; struct net *net = sock_net(skb->sk); + __skb_queue_head_init(&xmitq); + if (!info->attrs[TIPC_NLA_LINK]) return -EINVAL; @@@ -1683,13 -1696,13 +1696,13 @@@ u32 tol; tol = nla_get_u32(props[TIPC_NLA_PROP_TOL]); - tipc_link_set_tolerance(link, tol); + tipc_link_set_tolerance(link, tol, &xmitq); } if (props[TIPC_NLA_PROP_PRIO]) { u32 prio; prio = nla_get_u32(props[TIPC_NLA_PROP_PRIO]); - tipc_link_set_prio(link, prio); + tipc_link_set_prio(link, prio, &xmitq); } if (props[TIPC_NLA_PROP_WIN]) { u32 win; @@@ -1701,7 -1714,7 +1714,7 @@@ out: tipc_node_read_unlock(node); - + tipc_bearer_xmit(net, bearer_id, &xmitq, &node->links[bearer_id].maddr); return res; } diff --combined net/unix/af_unix.c index f75f847,b374555..8269da7 --- a/net/unix/af_unix.c +++ b/net/unix/af_unix.c @@@ -1496,7 -1496,7 +1496,7 @@@ static void unix_detach_fds(struct scm_ UNIXCB(skb).fp = NULL; for (i = scm->fp->count-1; i >= 0; i--) - unix_notinflight(scm->fp->fp[i]); + unix_notinflight(scm->fp->user, scm->fp->fp[i]); } static void unix_destruct_scm(struct sk_buff *skb) @@@ -1534,7 -1534,6 +1534,6 @@@ static int unix_attach_fds(struct scm_c { int i; unsigned char max_level = 0; - int unix_sock_count = 0; if (too_many_unix_fds(current)) return -ETOOMANYREFS; @@@ -1542,11 -1541,9 +1541,9 @@@ for (i = scm->fp->count - 1; i >= 0; i--) { struct sock *sk = unix_get_socket(scm->fp->fp[i]); - if (sk) { - unix_sock_count++; + if (sk) max_level = max(max_level, unix_sk(sk)->recursion_level); - } } if (unlikely(max_level > MAX_RECURSION_LEVEL)) return -ETOOMANYREFS; @@@ -1561,7 -1558,7 +1558,7 @@@ return -ENOMEM; for (i = scm->fp->count - 1; i >= 0; i--) - unix_inflight(scm->fp->fp[i]); + unix_inflight(scm->fp->user, scm->fp->fp[i]); return max_level; } @@@ -1781,12 -1778,7 +1778,12 @@@ restart_locked goto out_unlock; } - if (unlikely(unix_peer(other) != sk && unix_recvq_full(other))) { + /* other == sk && unix_peer(other) != sk if + * - unix_peer(sk) == NULL, destination address bound to sk + * - unix_peer(sk) == sk by time of get but disconnected before lock + */ + if (other != sk && + unlikely(unix_peer(other) != sk && unix_recvq_full(other))) { if (timeo) { timeo = unix_wait_for_peer(other, timeo); @@@ -2282,15 -2274,13 +2279,15 @@@ static int unix_stream_read_generic(str size_t size = state->size; unsigned int last_len; - err = -EINVAL; - if (sk->sk_state != TCP_ESTABLISHED) + if (unlikely(sk->sk_state != TCP_ESTABLISHED)) { + err = -EINVAL; goto out; + } - err = -EOPNOTSUPP; - if (flags & MSG_OOB) + if (unlikely(flags & MSG_OOB)) { + err = -EOPNOTSUPP; goto out; + } target = sock_rcvlowat(sk, flags & MSG_WAITALL, size); timeo = sock_rcvtimeo(sk, noblock); @@@ -2312,7 -2302,6 +2309,7 @@@ bool drop_skb; struct sk_buff *skb, *last; +redo: unix_state_lock(sk); if (sock_flag(sk, SOCK_DEAD)) { err = -ECONNRESET; @@@ -2337,11 -2326,9 +2334,11 @@@ again goto unlock; unix_state_unlock(sk); - err = -EAGAIN; - if (!timeo) + if (!timeo) { + err = -EAGAIN; break; + } + mutex_unlock(&u->readlock); timeo = unix_stream_data_wait(sk, timeo, last, @@@ -2354,7 -2341,7 +2351,7 @@@ } mutex_lock(&u->readlock); - continue; + goto redo; unlock: unix_state_unlock(sk); break; -- LinuxNextTracking

7 years, 1 month

Noticed difference in packet.h in maint.diff

by postmaster＠open-mesh.org

--- batman-adv_maint.packet.h 2016-02-24 00:05:01.445977540 +0100 +++ batctl_maint.packet.h 2016-02-24 00:05:01.453978081 +0100 @@ -1,4 +1,4 @@ -/* Copyright (C) 2007-2015 B.A.T.M.A.N. contributors: +/* Copyright (C) 2007-2016 B.A.T.M.A.N. contributors: * * Marek Lindner, Simon Wunderlich *

7 years, 1 month

Build check errors found: 2016-02-23

by postmaster＠open-mesh.org

7 years, 1 month

[linux-next] LinuxNextTracking branch, master, updated. next-20160222

by batman＠open-mesh.org

The following commit has been merged in the master branch: commit c18bdd018e8912ca73ad6c12120b7283b5038875 Author: Sven Eckelmann <sven(a)narfation.org> Date: Sun Jan 31 13:27:59 2016 +0100 batman-adv: Only put gw_node list reference when removed The batadv_gw_node reference counter in batadv_gw_node_update can only be reduced when the list entry was actually removed. Otherwise the reference counter may reach zero when batadv_gw_node_update is called from two different contexts for the same gw_node but only one context is actually removing the entry from the list. The release function for this gw_node is not called inside the list_lock spinlock protected region because the function batadv_gw_node_update still holds a gw_node reference for the object pointer on the stack. Thus the actual release function (when required) will be called only at the end of the function. Fixes: bd3524c14bd0 ("batman-adv: remove obsolete deleted attribute for gateway node") Signed-off-by: Sven Eckelmann <sven(a)narfation.org> Signed-off-by: Marek Lindner <mareklindner(a)neomailbox.ch> Signed-off-by: Antonio Quartulli <a(a)unstable.cc> diff --git a/net/batman-adv/gateway_client.c b/net/batman-adv/gateway_client.c index e6c8382..ccf70be 100644 --- a/net/batman-adv/gateway_client.c +++ b/net/batman-adv/gateway_client.c @@ -527,11 +527,12 @@ void batadv_gw_node_update(struct batadv_priv *bat_priv, * gets dereferenced. */ spin_lock_bh(&bat_priv->gw.list_lock); - hlist_del_init_rcu(&gw_node->list); + if (!hlist_unhashed(&gw_node->list)) { + hlist_del_init_rcu(&gw_node->list); + batadv_gw_node_free_ref(gw_node); + } spin_unlock_bh(&bat_priv->gw.list_lock); - batadv_gw_node_free_ref(gw_node); - curr_gw = batadv_gw_get_selected_gw_node(bat_priv); if (gw_node == curr_gw) batadv_gw_reselect(bat_priv); -- LinuxNextTracking

7 years, 1 month

[linux-next] LinuxNextTracking branch, master, updated. next-20160222

by batman＠open-mesh.org

The following commit has been merged in the master branch: commit 3db152093efb750bc47fd4d69355b90b18113105 Author: Sven Eckelmann <sven(a)narfation.org> Date: Sun Jan 31 13:28:00 2016 +0100 batman-adv: Only put orig_node_vlan list reference when removed The batadv_orig_node_vlan reference counter in batadv_tt_global_size_mod can only be reduced when the list entry was actually removed. Otherwise the reference counter may reach zero when batadv_tt_global_size_mod is called from two different contexts for the same orig_node_vlan but only one context is actually removing the entry from the list. The release function for this orig_node_vlan is not called inside the vlan_list_lock spinlock protected region because the function batadv_tt_global_size_mod still holds a orig_node_vlan reference for the object pointer on the stack. Thus the actual release function (when required) will be called only at the end of the function. Fixes: 7ea7b4a14275 ("batman-adv: make the TT CRC logic VLAN specific") Signed-off-by: Sven Eckelmann <sven(a)narfation.org> Signed-off-by: Marek Lindner <mareklindner(a)neomailbox.ch> Signed-off-by: Antonio Quartulli <a(a)unstable.cc> diff --git a/net/batman-adv/translation-table.c b/net/batman-adv/translation-table.c index cdfc85f..0e80fd1 100644 --- a/net/batman-adv/translation-table.c +++ b/net/batman-adv/translation-table.c @@ -303,9 +303,11 @@ static void batadv_tt_global_size_mod(struct batadv_orig_node *orig_node, if (atomic_add_return(v, &vlan->tt.num_entries) == 0) { spin_lock_bh(&orig_node->vlan_list_lock); - hlist_del_init_rcu(&vlan->list); + if (!hlist_unhashed(&vlan->list)) { + hlist_del_init_rcu(&vlan->list); + batadv_orig_node_vlan_free_ref(vlan); + } spin_unlock_bh(&orig_node->vlan_list_lock); - batadv_orig_node_vlan_free_ref(vlan); } batadv_orig_node_vlan_free_ref(vlan); -- LinuxNextTracking

7 years, 1 month

[linux-next] LinuxNextTracking branch, master, updated. next-20160222

by batman＠open-mesh.org

The following commit has been merged in the master branch: commit 1bc4e2b000e7fa9773d6623bc8850561ce10a4fb Author: Andrew Lunn <andrew(a)lunn.ch> Date: Thu Feb 11 22:15:57 2016 +0100 batman-adv: Avoid endless loop in bat-on-bat netdevice check batman-adv checks in different situation if a new device is already on top of a different batman-adv device. This is done by getting the iflink of a device and all its parent. It assumes that this iflink is always a parent device in an acyclic graph. But this assumption is broken by devices like veth which are actually a pair of two devices linked to each other. The recursive check would therefore get veth0 when calling dev_get_iflink on veth1. And it gets veth0 when calling dev_get_iflink with veth1. Creating a veth pair and loading batman-adv freezes parts of the system ip link add veth0 type veth peer name veth1 modprobe batman-adv An RCU stall will be detected on the system which cannot be fixed. INFO: rcu_sched self-detected stall on CPU 1: (5264 ticks this GP) idle=3e9/140000000000001/0 softirq=144683/144686 fqs=5249 (t=5250 jiffies g=46 c=45 q=43) Task dump for CPU 1: insmod R running task 0 247 245 0x00000008 ffffffff8151f140 ffffffff8107888e ffff88000fd141c0 ffffffff8151f140 0000000000000000 ffffffff81552df0 ffffffff8107b420 0000000000000001 ffff88000e3fa700 ffffffff81540b00 ffffffff8107d667 0000000000000001 Call Trace: <IRQ> [<ffffffff8107888e>] ? rcu_dump_cpu_stacks+0x7e/0xd0 [<ffffffff8107b420>] ? rcu_check_callbacks+0x3f0/0x6b0 [<ffffffff8107d667>] ? hrtimer_run_queues+0x47/0x180 [<ffffffff8107cf9d>] ? update_process_times+0x2d/0x50 [<ffffffff810873fb>] ? tick_handle_periodic+0x1b/0x60 [<ffffffff810290ae>] ? smp_trace_apic_timer_interrupt+0x5e/0x90 [<ffffffff813bbae2>] ? apic_timer_interrupt+0x82/0x90 <EOI> [<ffffffff812c3fd7>] ? __dev_get_by_index+0x37/0x40 [<ffffffffa0031f3e>] ? batadv_hard_if_event+0xee/0x3a0 [batman_adv] [<ffffffff812c5801>] ? register_netdevice_notifier+0x81/0x1a0 [...] This can be avoided by checking if two devices are each others parent and stopping the check in this situation. Fixes: b7eddd0b3950 ("batman-adv: prevent using any virtual device created on batman-adv as hard-interface") Signed-off-by: Andrew Lunn <andrew(a)lunn.ch> [sven(a)narfation.org: rewritten description, extracted fix] Signed-off-by: Sven Eckelmann <sven(a)narfation.org> Signed-off-by: Marek Lindner <mareklindner(a)neomailbox.ch> Signed-off-by: Antonio Quartulli <a(a)unstable.cc> diff --git a/net/batman-adv/hard-interface.c b/net/batman-adv/hard-interface.c index 01acccc..57f71071 100644 --- a/net/batman-adv/hard-interface.c +++ b/net/batman-adv/hard-interface.c @@ -76,6 +76,28 @@ out: } /** + * batadv_mutual_parents - check if two devices are each others parent + * @dev1: 1st net_device + * @dev2: 2nd net_device + * + * veth devices come in pairs and each is the parent of the other! + * + * Return: true if the devices are each others parent, otherwise false + */ +static bool batadv_mutual_parents(const struct net_device *dev1, + const struct net_device *dev2) +{ + int dev1_parent_iflink = dev_get_iflink(dev1); + int dev2_parent_iflink = dev_get_iflink(dev2); + + if (!dev1_parent_iflink || !dev2_parent_iflink) + return false; + + return (dev1_parent_iflink == dev2->ifindex) && + (dev2_parent_iflink == dev1->ifindex); +} + +/** * batadv_is_on_batman_iface - check if a device is a batman iface descendant * @net_dev: the device to check * @@ -108,6 +130,9 @@ static bool batadv_is_on_batman_iface(const struct net_device *net_dev) if (WARN(!parent_dev, "Cannot find parent device")) return false; + if (batadv_mutual_parents(net_dev, parent_dev)) + return false; + ret = batadv_is_on_batman_iface(parent_dev); return ret; -- LinuxNextTracking

7 years, 1 month

[linux-next] LinuxNextTracking branch, master, updated. next-20160222

by batman＠open-mesh.org

The following commit has been merged in the master branch: commit fb54bdb6e7a1a576037398104c793d59581d9a91 Merge: 76daeaceb214b0af4f5f304aa6a665fffb85e9bf 80c804bfc487c6df783c258b9034b9d81c34f7a0 Author: Stephen Rothwell <sfr(a)canb.auug.org.au> Date: Mon Feb 22 10:35:25 2016 +1100 Merge remote-tracking branch 'net-next/master' diff --combined MAINTAINERS index d391f3a,355e1c8..576e7e0 --- a/MAINTAINERS +++ b/MAINTAINERS @@@ -223,7 -223,9 +223,7 @@@ F: drivers/scsi/aacraid ABI/API L: linux-api(a)vger.kernel.org -F: Documentation/ABI/ F: include/linux/syscalls.h -F: include/uapi/ F: kernel/sys_ni.c ABIT UGURU 1,2 HARDWARE MONITOR DRIVER @@@ -673,25 -675,24 +673,25 @@@ F: drivers/gpu/drm/radeon/radeon_kfd. F: drivers/gpu/drm/radeon/radeon_kfd.h F: include/uapi/linux/kfd_ioctl.h +AMD SEATTLE DEVICE TREE SUPPORT +M: Brijesh Singh <brijeshkumar.singh(a)amd.com> +M: Suravee Suthikulpanit <suravee.suthikulpanit(a)amd.com> +M: Tom Lendacky <thomas.lendacky(a)amd.com> +S: Supported +F: arch/arm64/boot/dts/amd/ + AMD XGBE DRIVER M: Tom Lendacky <thomas.lendacky(a)amd.com> L: netdev(a)vger.kernel.org S: Supported F: drivers/net/ethernet/amd/xgbe/ +F: arch/arm64/boot/dts/amd/amd-seattle-xgbe*.dtsi AMS (Apple Motion Sensor) DRIVER M: Michael Hanselmann <linux-kernel(a)hansmi.ch> S: Supported F: drivers/macintosh/ams/ -AMSO1100 RNIC DRIVER -M: Tom Tucker <tom(a)opengridcomputing.com> -M: Steve Wise <swise(a)opengridcomputing.com> -L: linux-rdma(a)vger.kernel.org -S: Maintained -F: drivers/infiniband/hw/amso1100/ - ANALOG DEVICES INC AD9389B DRIVER M: Hans Verkuil <hans.verkuil(a)cisco.com> L: linux-media(a)vger.kernel.org @@@ -966,8 -967,6 +966,8 @@@ M: Rob Herring <robh(a)kernel.org L: linux-arm-kernel(a)lists.infradead.org (moderated for non-subscribers) S: Maintained F: arch/arm/mach-highbank/ +F: arch/arm/boot/dts/highbank.dts +F: arch/arm/boot/dts/ecx-*.dts* ARM/CAVIUM NETWORKS CNS3XXX MACHINE SUPPORT M: Krzysztof Halasa <khalasa(a)piap.pl> @@@ -1043,7 -1042,6 +1043,7 @@@ M: Barry Song <baohua(a)kernel.org L: linux-arm-kernel(a)lists.infradead.org (moderated for non-subscribers) T: git git://git.kernel.org/pub/scm/linux/kernel/git/baohua/linux.git S: Maintained +F: arch/arm/boot/dts/prima2* F: arch/arm/mach-prima2/ F: drivers/clk/sirf/ F: drivers/clocksource/timer-prima2.c @@@ -1145,10 -1143,6 +1145,10 @@@ W: http://www.hisilicon.co S: Supported T: git git://github.com/hisilicon/linux-hisi.git F: arch/arm/mach-hisi/ +F: arch/arm/boot/dts/hi3* +F: arch/arm/boot/dts/hip* +F: arch/arm/boot/dts/hisi* +F: arch/arm64/boot/dts/hisilicon/ ARM/HP JORNADA 7XX MACHINE SUPPORT M: Kristoffer Ericson <kristoffer.ericson(a)gmail.com> @@@ -1225,7 -1219,6 +1225,7 @@@ M: Santosh Shilimkar <ssantosh(a)kernel.o L: linux-arm-kernel(a)lists.infradead.org (moderated for non-subscribers) S: Maintained F: arch/arm/mach-keystone/ +F: arch/arm/boot/dts/k2* T: git git://git.kernel.org/pub/scm/linux/kernel/git/ssantosh/linux-keystone.git ARM/TEXAS INSTRUMENT KEYSTONE CLOCK FRAMEWORK @@@ -1286,7 -1279,6 +1286,7 @@@ F: arch/arm/mach-mvebu F: drivers/rtc/rtc-armada38x.c F: arch/arm/boot/dts/armada* F: arch/arm/boot/dts/kirkwood* +F: arch/arm64/boot/dts/marvell/armada* ARM/Marvell Berlin SoC support @@@ -1295,7 -1287,6 +1295,7 @@@ L: linux-arm-kernel(a)lists.infradead.or S: Maintained F: arch/arm/mach-berlin/ F: arch/arm/boot/dts/berlin* +F: arch/arm64/boot/dts/marvell/berlin* ARM/Marvell Dove/MV78xx0/Orion SOC support @@@ -1434,10 -1425,7 +1434,10 @@@ S: Maintaine F: arch/arm/boot/dts/qcom-*.dts F: arch/arm/boot/dts/qcom-*.dtsi F: arch/arm/mach-qcom/ +F: arch/arm64/boot/dts/qcom/* +F: drivers/i2c/busses/i2c-qup.c F: drivers/soc/qcom/ +F: drivers/spi/spi-qup.c F: drivers/tty/serial/msm_serial.h F: drivers/tty/serial/msm_serial.c F: drivers/*/pm8???-* @@@ -1453,8 -1441,8 +1453,8 @@@ S: Maintaine ARM/RENESAS ARM64 ARCHITECTURE M: Simon Horman <horms(a)verge.net.au> M: Magnus Damm <magnus.damm(a)gmail.com> -L: linux-sh(a)vger.kernel.org -Q: http://patchwork.kernel.org/project/linux-sh/list/ +L: linux-renesas-soc(a)vger.kernel.org +Q: http://patchwork.kernel.org/project/linux-renesas-soc/list/ T: git git://git.kernel.org/pub/scm/linux/kernel/git/horms/renesas.git next S: Supported F: arch/arm64/boot/dts/renesas/ @@@ -1496,8 -1484,6 +1496,8 @@@ L: linux-arm-kernel(a)lists.infradead.or L: linux-samsung-soc(a)vger.kernel.org (moderated for non-subscribers) S: Maintained F: arch/arm/boot/dts/s3c* +F: arch/arm/boot/dts/s5p* +F: arch/arm/boot/dts/samsung* F: arch/arm/boot/dts/exynos* F: arch/arm64/boot/dts/exynos/ F: arch/arm/plat-samsung/ @@@ -1507,7 -1493,6 +1507,7 @@@ F: arch/arm/mach-s5p* F: arch/arm/mach-exynos*/ F: drivers/*/*s3c2410* F: drivers/*/*/*s3c2410* +F: drivers/soc/samsung/* F: drivers/spi/spi-s3c* F: sound/soc/samsung/* F: Documentation/arm/Samsung/ @@@ -1578,7 -1563,6 +1578,7 @@@ S: Maintaine F: arch/arm/mach-socfpga/ F: arch/arm/boot/dts/socfpga* F: arch/arm/configs/socfpga_defconfig +F: arch/arm64/boot/dts/altera/ W: http://www.rocketboards.org T: git git://git.kernel.org/pub/scm/linux/kernel/git/dinguyen/linux.git @@@ -1732,7 -1716,7 +1732,7 @@@ M: Lorenzo Pieralisi <lorenzo.pieralisi L: linux-arm-kernel(a)lists.infradead.org (moderated for non-subscribers) S: Maintained F: arch/arm/boot/dts/vexpress* -F: arch/arm64/boot/dts/arm/vexpress* +F: arch/arm64/boot/dts/arm/ F: arch/arm/mach-vexpress/ F: */*/vexpress* F: */*/*/vexpress* @@@ -2163,7 -2147,7 +2163,7 @@@ M: Marek Lindner <mareklindner@neomailb M: Simon Wunderlich <sw(a)simonwunderlich.de> M: Antonio Quartulli <a(a)unstable.cc> L: b.a.t.m.a.n(a)lists.open-mesh.org - W: http://www.open-mesh.org/ + W: https://www.open-mesh.org/ S: Maintained F: net/batman-adv/ @@@ -2359,7 -2343,6 +2359,7 @@@ F: arch/arm/mach-bcm F: arch/arm/boot/dts/bcm113* F: arch/arm/boot/dts/bcm216* F: arch/arm/boot/dts/bcm281* +F: arch/arm64/boot/dts/broadcom/ F: arch/arm/configs/bcm_defconfig F: drivers/mmc/host/sdhci-bcm-kona.c F: drivers/clocksource/bcm_kona_timer.c @@@ -2374,6 -2357,14 +2374,6 @@@ T: git git://git.kernel.org/pub/scm/lin S: Maintained N: bcm2835 -BROADCOM BCM33XX MIPS ARCHITECTURE -M: Kevin Cernekee <cernekee(a)gmail.com> -L: linux-mips(a)linux-mips.org -S: Maintained -F: arch/mips/bcm3384/* -F: arch/mips/include/asm/mach-bcm3384/* -F: arch/mips/kernel/*bmips* - BROADCOM BCM47XX MIPS ARCHITECTURE M: Hauke Mehrtens <hauke(a)hauke-m.de> M: Rafał Miłecki <zajec5(a)gmail.com> @@@ -2433,6 -2424,7 +2433,7 @@@ F: include/linux/bcm963xx_nvram. F: include/linux/bcm963xx_tag.h BROADCOM TG3 GIGABIT ETHERNET DRIVER + M: Siva Reddy Kallam <siva.kallam(a)broadcom.com> M: Prashant Sreedharan <prashant(a)broadcom.com> M: Michael Chan <mchan(a)broadcom.com> L: netdev(a)vger.kernel.org @@@ -3454,7 -3446,7 +3455,7 @@@ S: Maintaine F: drivers/usb/dwc2/ DESIGNWARE USB3 DRD IP DRIVER -M: Felipe Balbi <balbi(a)ti.com> +M: Felipe Balbi <balbi(a)kernel.org> L: linux-usb(a)vger.kernel.org L: linux-omap(a)vger.kernel.org T: git git://git.kernel.org/pub/scm/linux/kernel/git/balbi/usb.git @@@ -4193,6 -4185,13 +4194,6 @@@ W: http://aeschi.ch.eu.org/efs S: Orphan F: fs/efs/ -EHCA (IBM GX bus InfiniBand adapter) DRIVER -M: Hoang-Nam Nguyen <hnguyen(a)de.ibm.com> -M: Christoph Raisch <raisch(a)de.ibm.com> -L: linux-rdma(a)vger.kernel.org -S: Supported -F: drivers/infiniband/hw/ehca/ - EHEA (IBM pSeries eHEA 10Gb ethernet adapter) DRIVER M: Thadeu Lima de Souza Cascardo <cascardo(a)linux.vnet.ibm.com> L: netdev(a)vger.kernel.org @@@ -5811,6 -5810,12 +5812,6 @@@ M: Juanjo Ciarlante <jjciarla(a)raiz.uncu S: Maintained F: net/ipv4/netfilter/ipt_MASQUERADE.c -IPATH DRIVER -M: Mike Marciniszyn <infinipath(a)intel.com> -L: linux-rdma(a)vger.kernel.org -S: Maintained -F: drivers/staging/rdma/ipath/ - IPMI SUBSYSTEM M: Corey Minyard <minyard(a)acm.org> L: openipmi-developer(a)lists.sourceforge.net (moderated for non-subscribers) @@@ -6140,7 -6145,7 +6141,7 @@@ F: include/uapi/linux/sunrpc KERNEL SELFTEST FRAMEWORK M: Shuah Khan <shuahkh(a)osg.samsung.com> -L: linux-api(a)vger.kernel.org +L: linux-kselftest(a)vger.kernel.org T: git git://git.kernel.org/pub/scm/shuah/linux-kselftest S: Maintained F: tools/testing/selftests @@@ -7366,7 -7371,7 +7367,7 @@@ F: drivers/tty/isicom. F: include/linux/isicom.h MUSB MULTIPOINT HIGH SPEED DUAL-ROLE CONTROLLER -M: Felipe Balbi <balbi(a)ti.com> +M: Felipe Balbi <balbi(a)kernel.org> L: linux-usb(a)vger.kernel.org T: git git://git.kernel.org/pub/scm/linux/kernel/git/balbi/usb.git S: Maintained @@@ -7935,7 -7940,7 +7936,7 @@@ F: drivers/media/platform/omap3isp F: drivers/staging/media/omap4iss/ OMAP USB SUPPORT -M: Felipe Balbi <balbi(a)ti.com> +M: Felipe Balbi <balbi(a)kernel.org> L: linux-usb(a)vger.kernel.org L: linux-omap(a)vger.kernel.org T: git git://git.kernel.org/pub/scm/linux/kernel/git/balbi/usb.git @@@ -8123,14 -8128,6 +8124,14 @@@ S: Supporte F: fs/overlayfs/ F: Documentation/filesystems/overlayfs.txt +ORANGEFS FILESYSTEM +M: Mike Marshall <hubcap(a)omnibond.com> +L: pvfs2-developers(a)beowulf-underground.org +T: git git://git.kernel.org/pub/scm/linux/kernel/git/hubcap/linux.git +S: Supported +F: fs/orangefs/ +F: Documentation/filesystems/orangefs.txt + P54 WIRELESS DRIVER M: Christian Lamparter <chunkeey(a)googlemail.com> L: linux-wireless(a)vger.kernel.org @@@ -8822,7 -8819,6 +8823,7 @@@ L: linux-arm-kernel(a)lists.infradead.or T: git git://github.com/hzhuang1/linux.git T: git git://github.com/rjarzmik/linux.git S: Maintained +F: arch/arm/boot/dts/pxa* F: arch/arm/mach-pxa/ F: drivers/dma/pxa* F: drivers/pcmcia/pxa2xx* @@@ -8852,7 -8848,6 +8853,7 @@@ L: linux-arm-kernel(a)lists.infradead.or T: git git://github.com/hzhuang1/linux.git T: git git://git.linaro.org/people/ycmiao/pxa-linux.git S: Maintained +F: arch/arm/boot/dts/mmp* F: arch/arm/mach-mmp/ PXA MMCI DRIVER @@@ -9799,11 -9794,10 +9800,11 @@@ S: Supporte F: drivers/scsi/be2iscsi/ Emulex 10Gbps NIC BE2, BE3-R, Lancer, Skyhawk-R DRIVER -M: Sathya Perla <sathya.perla(a)avagotech.com> -M: Ajit Khaparde <ajit.khaparde(a)avagotech.com> -M: Padmanabh Ratnakar <padmanabh.ratnakar(a)avagotech.com> -M: Sriharsha Basavapatna <sriharsha.basavapatna(a)avagotech.com> +M: Sathya Perla <sathya.perla(a)broadcom.com> +M: Ajit Khaparde <ajit.khaparde(a)broadcom.com> +M: Padmanabh Ratnakar <padmanabh.ratnakar(a)broadcom.com> +M: Sriharsha Basavapatna <sriharsha.basavapatna(a)broadcom.com> +M: Somnath Kotur <somnath.kotur(a)broadcom.com> L: netdev(a)vger.kernel.org W: http://www.emulex.com S: Supported @@@ -10165,7 -10159,6 +10166,7 @@@ S: Supporte F: drivers/media/pci/solo6x10/ SOFTWARE RAID (Multiple Disks) SUPPORT +M: Shaohua Li <shli(a)kernel.org> L: linux-raid(a)vger.kernel.org T: git git://neil.brown.name/md S: Supported @@@ -10299,7 -10292,6 +10300,7 @@@ L: spear-devel(a)list.st.co L: linux-arm-kernel(a)lists.infradead.org (moderated for non-subscribers) W: http://www.st.com/spear S: Maintained +F: arch/arm/boot/dts/spear* F: arch/arm/mach-spear/ SPEAR CLOCK FRAMEWORK SUPPORT @@@ -10862,14 -10854,6 +10863,14 @@@ L: linux-omap(a)vger.kernel.or S: Maintained F: drivers/thermal/ti-soc-thermal/ +TI VPE/CAL DRIVERS +M: Benoit Parrot <bparrot(a)ti.com> +L: linux-media(a)vger.kernel.org +W: http://linuxtv.org/ +Q: http://patchwork.linuxtv.org/project/linux-media/list/ +S: Maintained +F: drivers/media/platform/ti-vpe/ + TI CDCE706 CLOCK DRIVER M: Max Filippov <jcmvbkbc(a)gmail.com> S: Maintained @@@ -11335,7 -11319,7 +11336,7 @@@ F: Documentation/usb/ehci.tx F: drivers/usb/host/ehci* USB GADGET/PERIPHERAL SUBSYSTEM -M: Felipe Balbi <balbi(a)ti.com> +M: Felipe Balbi <balbi(a)kernel.org> L: linux-usb(a)vger.kernel.org W: http://www.linux-usb.org/gadget T: git git://git.kernel.org/pub/scm/linux/kernel/git/balbi/usb.git @@@ -11411,7 -11395,7 +11412,7 @@@ S: Maintaine F: drivers/net/usb/pegasus.* USB PHY LAYER -M: Felipe Balbi <balbi(a)ti.com> +M: Felipe Balbi <balbi(a)kernel.org> L: linux-usb(a)vger.kernel.org T: git git://git.kernel.org/pub/scm/linux/kernel/git/balbi/usb.git S: Maintained @@@ -12150,7 -12134,7 +12151,7 @@@ F: drivers/net/hamradio/*scc. F: drivers/net/hamradio/z8530.h ZBUD COMPRESSED PAGE ALLOCATOR -M: Seth Jennings <sjennings(a)variantweb.net> +M: Seth Jennings <sjenning(a)redhat.com> L: linux-mm(a)kvack.org S: Maintained F: mm/zbud.c @@@ -12205,7 -12189,7 +12206,7 @@@ F: include/linux/zsmalloc. F: Documentation/vm/zsmalloc.txt ZSWAP COMPRESSED SWAP CACHING -M: Seth Jennings <sjennings(a)variantweb.net> +M: Seth Jennings <sjenning(a)redhat.com> L: linux-mm(a)kvack.org S: Maintained F: mm/zswap.c diff --combined drivers/net/bonding/bond_main.c index b7f1a99,45bdd87..a6527d5 --- a/drivers/net/bonding/bond_main.c +++ b/drivers/net/bonding/bond_main.c @@@ -214,8 -214,6 +214,8 @@@ static void bond_uninit(struct net_devi static struct rtnl_link_stats64 *bond_get_stats(struct net_device *bond_dev, struct rtnl_link_stats64 *stats); static void bond_slave_arr_handler(struct work_struct *work); +static bool bond_time_in_interval(struct bonding *bond, unsigned long last_act, + int mod); /*---------------------------- General routines -----------------------------*/ @@@ -620,8 -618,8 +620,8 @@@ static void bond_hw_addr_swap(struct bo static void bond_set_dev_addr(struct net_device *bond_dev, struct net_device *slave_dev) { - netdev_dbg(bond_dev, "bond_dev=%p slave_dev=%p slave_dev->addr_len=%d\n", - bond_dev, slave_dev, slave_dev->addr_len); + netdev_dbg(bond_dev, "bond_dev=%p slave_dev=%p slave_dev->name=%s slave_dev->addr_len=%d\n", + bond_dev, slave_dev, slave_dev->name, slave_dev->addr_len); memcpy(bond_dev->dev_addr, slave_dev->dev_addr, slave_dev->addr_len); bond_dev->addr_assign_type = NET_ADDR_STOLEN; call_netdevice_notifiers(NETDEV_CHANGEADDR, bond_dev); @@@ -930,11 -928,10 +930,10 @@@ void bond_select_active_slave(struct bo if (!rv) return; - if (netif_carrier_ok(bond->dev)) { + if (netif_carrier_ok(bond->dev)) netdev_info(bond->dev, "first active interface up!\n"); - } else { + else netdev_info(bond->dev, "now running without any active interface!\n"); - } } } @@@ -1180,9 -1177,8 +1179,8 @@@ static rx_handler_result_t bond_handle_ } } - if (bond_should_deliver_exact_match(skb, slave, bond)) { + if (bond_should_deliver_exact_match(skb, slave, bond)) return RX_HANDLER_EXACT; - } skb->dev = bond->dev; @@@ -1243,7 -1239,7 +1241,7 @@@ static struct slave *bond_alloc_slave(s { struct slave *slave = NULL; - slave = kzalloc(sizeof(struct slave), GFP_KERNEL); + slave = kzalloc(sizeof(*slave), GFP_KERNEL); if (!slave) return NULL; @@@ -1383,8 -1379,7 +1381,7 @@@ int bond_enslave(struct net_device *bon if (slave_dev->flags & IFF_UP) { netdev_err(bond_dev, "%s is up - this may be due to an out of date ifenslave\n", slave_dev->name); - res = -EPERM; - goto err_undo_flags; + return -EPERM; } /* set bonding device ether type by slave - bonding netdevices are @@@ -1404,8 -1399,7 +1401,7 @@@ res = notifier_to_errno(res); if (res) { netdev_err(bond_dev, "refused to change device type\n"); - res = -EBUSY; - goto err_undo_flags; + return -EBUSY; } /* Flush unicast and multicast addresses */ @@@ -1425,8 -1419,7 +1421,7 @@@ } else if (bond_dev->type != slave_dev->type) { netdev_err(bond_dev, "%s ether type (%d) is different from other slaves (%d), can not enslave it\n", slave_dev->name, slave_dev->type, bond_dev->type); - res = -EINVAL; - goto err_undo_flags; + return -EINVAL; } if (slave_ops->ndo_set_mac_address == NULL) { @@@ -2129,7 -2122,6 +2124,7 @@@ static void bond_miimon_commit(struct b continue; case BOND_LINK_UP: + bond_update_speed_duplex(slave); bond_set_slave_link_state(slave, BOND_LINK_UP, BOND_SLAVE_NOTIFY_NOW); slave->last_link_up = jiffies; @@@ -2462,7 -2454,7 +2457,7 @@@ int bond_arp_rcv(const struct sk_buff * struct slave *slave) { struct arphdr *arp = (struct arphdr *)skb->data; - struct slave *curr_active_slave; + struct slave *curr_active_slave, *curr_arp_slave; unsigned char *arp_ptr; __be32 sip, tip; int alen, is_arp = skb->protocol == __cpu_to_be16(ETH_P_ARP); @@@ -2509,41 -2501,26 +2504,41 @@@ &sip, &tip); curr_active_slave = rcu_dereference(bond->curr_active_slave); + curr_arp_slave = rcu_dereference(bond->current_arp_slave); - /* Backup slaves won't see the ARP reply, but do come through - * here for each ARP probe (so we swap the sip/tip to validate - * the probe). In a "redundant switch, common router" type of - * configuration, the ARP probe will (hopefully) travel from - * the active, through one switch, the router, then the other - * switch before reaching the backup. + /* We 'trust' the received ARP enough to validate it if: * - * We 'trust' the arp requests if there is an active slave and - * it received valid arp reply(s) after it became active. This - * is done to avoid endless looping when we can't reach the + * (a) the slave receiving the ARP is active (which includes the + * current ARP slave, if any), or + * + * (b) the receiving slave isn't active, but there is a currently + * active slave and it received valid arp reply(s) after it became + * the currently active slave, or + * + * (c) there is an ARP slave that sent an ARP during the prior ARP + * interval, and we receive an ARP reply on any slave. We accept + * these because switch FDB update delays may deliver the ARP + * reply to a slave other than the sender of the ARP request. + * + * Note: for (b), backup slaves are receiving the broadcast ARP + * request, not a reply. This request passes from the sending + * slave through the L2 switch(es) to the receiving slave. Since + * this is checking the request, sip/tip are swapped for + * validation. + * + * This is done to avoid endless looping when we can't reach the * arp_ip_target and fool ourselves with our own arp requests. */ - if (bond_is_active_slave(slave)) bond_validate_arp(bond, slave, sip, tip); else if (curr_active_slave && time_after(slave_last_rx(bond, curr_active_slave), curr_active_slave->last_link_up)) bond_validate_arp(bond, slave, tip, sip); + else if (curr_arp_slave && (arp->ar_op == htons(ARPOP_REPLY)) && + bond_time_in_interval(bond, + dev_trans_start(curr_arp_slave->dev), 1)) + bond_validate_arp(bond, slave, sip, tip); out_unlock: if (arp != (struct arphdr *)skb->data) @@@ -3327,6 -3304,7 +3322,7 @@@ static struct rtnl_link_stats64 *bond_g stats->rx_bytes += sstats->rx_bytes - pstats->rx_bytes; stats->rx_errors += sstats->rx_errors - pstats->rx_errors; stats->rx_dropped += sstats->rx_dropped - pstats->rx_dropped; + stats->rx_nohandler += sstats->rx_nohandler - pstats->rx_nohandler; stats->tx_packets += sstats->tx_packets - pstats->tx_packets;; stats->tx_bytes += sstats->tx_bytes - pstats->tx_bytes; diff --combined drivers/net/ethernet/broadcom/bnxt/bnxt.c index 8ab000d,169920a..ff1507f --- a/drivers/net/ethernet/broadcom/bnxt/bnxt.c +++ b/drivers/net/ethernet/broadcom/bnxt/bnxt.c @@@ -69,7 -69,7 +69,7 @@@ MODULE_VERSION(DRV_MODULE_VERSION) #define BNXT_RX_DMA_OFFSET NET_SKB_PAD #define BNXT_RX_COPY_THRESH 256 -#define BNXT_TX_PUSH_THRESH 92 +#define BNXT_TX_PUSH_THRESH 164 enum board_idx { BCM57301, @@@ -223,12 -223,11 +223,12 @@@ static netdev_tx_t bnxt_start_xmit(stru } if (free_size == bp->tx_ring_size && length <= bp->tx_push_thresh) { - struct tx_push_bd *push = txr->tx_push; - struct tx_bd *tx_push = &push->txbd1; - struct tx_bd_ext *tx_push1 = &push->txbd2; - void *pdata = tx_push1 + 1; - int j; + struct tx_push_buffer *tx_push_buf = txr->tx_push; + struct tx_push_bd *tx_push = &tx_push_buf->push_bd; + struct tx_bd_ext *tx_push1 = &tx_push->txbd2; + void *pdata = tx_push_buf->data; + u64 *end; + int j, push_len; /* Set COAL_NOW to be ready quickly for the next push */ tx_push->tx_bd_len_flags_type = @@@ -248,9 -247,6 +248,9 @@@ tx_push1->tx_bd_cfa_meta = cpu_to_le32(vlan_tag_flags); tx_push1->tx_bd_cfa_action = cpu_to_le32(cfa_action); + end = PTR_ALIGN(pdata + length + 1, 8) - 1; + *end = 0; + skb_copy_from_linear_data(skb, pdata, len); pdata += len; for (j = 0; j < last_frag; j++) { @@@ -265,29 -261,22 +265,29 @@@ pdata += skb_frag_size(frag); } - memcpy(txbd, tx_push, sizeof(*txbd)); + txbd->tx_bd_len_flags_type = tx_push->tx_bd_len_flags_type; + txbd->tx_bd_haddr = txr->data_mapping; prod = NEXT_TX(prod); txbd = &txr->tx_desc_ring[TX_RING(prod)][TX_IDX(prod)]; memcpy(txbd, tx_push1, sizeof(*txbd)); prod = NEXT_TX(prod); - push->doorbell = + tx_push->doorbell = cpu_to_le32(DB_KEY_TX_PUSH | DB_LONG_TX_PUSH | prod); txr->tx_prod = prod; netdev_tx_sent_queue(txq, skb->len); - __iowrite64_copy(txr->tx_doorbell, push, - (length + sizeof(*push) + 8) / 8); + push_len = (length + sizeof(*tx_push) + 7) / 8; + if (push_len > 16) { + __iowrite64_copy(txr->tx_doorbell, tx_push_buf, 16); + __iowrite64_copy(txr->tx_doorbell + 4, tx_push_buf + 1, + push_len - 16); + } else { + __iowrite64_copy(txr->tx_doorbell, tx_push_buf, + push_len); + } tx_buf->is_push = 1; - goto tx_done; } @@@ -1764,7 -1753,7 +1764,7 @@@ static int bnxt_alloc_tx_rings(struct b push_size = L1_CACHE_ALIGN(sizeof(struct tx_push_bd) + bp->tx_push_thresh); - if (push_size > 128) { + if (push_size > 256) { push_size = 0; bp->tx_push_thresh = 0; } @@@ -1783,6 -1772,7 +1783,6 @@@ return rc; if (bp->tx_push_size) { - struct tx_bd *txbd; dma_addr_t mapping; /* One pre-allocated DMA buffer to backup @@@ -1796,11 -1786,13 +1796,11 @@@ if (!txr->tx_push) return -ENOMEM; - txbd = &txr->tx_push->txbd1; - mapping = txr->tx_push_mapping + sizeof(struct tx_push_bd); - txbd->tx_bd_haddr = cpu_to_le64(mapping); + txr->data_mapping = cpu_to_le64(mapping); - memset(txbd + 1, 0, sizeof(struct tx_bd_ext)); + memset(txr->tx_push, 0, sizeof(struct tx_push_bd)); } ring->queue_id = bp->q_info[j].queue_id; if (i % bp->tx_nr_rings_per_tc == (bp->tx_nr_rings_per_tc - 1)) @@@ -4554,18 -4546,20 +4554,18 @@@ static int bnxt_update_phy_setting(stru if (!(link_info->autoneg & BNXT_AUTONEG_FLOW_CTRL) && link_info->force_pause_setting != link_info->req_flow_ctrl) update_pause = true; - if (link_info->req_duplex != link_info->duplex_setting) - update_link = true; if (!(link_info->autoneg & BNXT_AUTONEG_SPEED)) { if (BNXT_AUTO_MODE(link_info->auto_mode)) update_link = true; if (link_info->req_link_speed != link_info->force_link_speed) update_link = true; + if (link_info->req_duplex != link_info->duplex_setting) + update_link = true; } else { if (link_info->auto_mode == BNXT_LINK_AUTO_NONE) update_link = true; if (link_info->advertising != link_info->auto_link_speeds) update_link = true; - if (link_info->req_link_speed != link_info->auto_link_speed) - update_link = true; } if (update_link) @@@ -4642,7 -4636,7 +4642,7 @@@ static int __bnxt_open_nic(struct bnxt if (link_re_init) { rc = bnxt_update_phy_setting(bp); if (rc) - goto open_err; + netdev_warn(bp->dev, "failed to update phy settings\n"); } if (irq_re_init) { @@@ -4660,7 -4654,6 +4660,7 @@@ /* Enable TX queues */ bnxt_tx_enable(bp); mod_timer(&bp->timer, jiffies + bp->current_interval); + bnxt_update_link(bp, true); return 0; @@@ -5377,9 -5370,16 +5377,16 @@@ static int bnxt_change_mtu(struct net_d return 0; } - static int bnxt_setup_tc(struct net_device *dev, u8 tc) + static int bnxt_setup_tc(struct net_device *dev, u32 handle, __be16 proto, + struct tc_to_netdev *ntc) { struct bnxt *bp = netdev_priv(dev); + u8 tc; + + if (handle != TC_H_ROOT || ntc->type != TC_SETUP_MQPRIO) + return -EINVAL; + + tc = ntc->tc; if (tc > bp->max_tc) { netdev_err(dev, "too many traffic classes requested: %d Max supported is %d\n", @@@ -5677,16 -5677,22 +5684,16 @@@ static int bnxt_probe_phy(struct bnxt * } /*initialize the ethool setting copy with NVM settings */ - if (BNXT_AUTO_MODE(link_info->auto_mode)) - link_info->autoneg |= BNXT_AUTONEG_SPEED; - - if (link_info->auto_pause_setting & BNXT_LINK_PAUSE_BOTH) { - if (link_info->auto_pause_setting == BNXT_LINK_PAUSE_BOTH) - link_info->autoneg |= BNXT_AUTONEG_FLOW_CTRL; + if (BNXT_AUTO_MODE(link_info->auto_mode)) { + link_info->autoneg = BNXT_AUTONEG_SPEED | + BNXT_AUTONEG_FLOW_CTRL; + link_info->advertising = link_info->auto_link_speeds; link_info->req_flow_ctrl = link_info->auto_pause_setting; - } else if (link_info->force_pause_setting & BNXT_LINK_PAUSE_BOTH) { + } else { + link_info->req_link_speed = link_info->force_link_speed; + link_info->req_duplex = link_info->duplex_setting; link_info->req_flow_ctrl = link_info->force_pause_setting; } - link_info->req_duplex = link_info->duplex_setting; - if (link_info->autoneg & BNXT_AUTONEG_SPEED) - link_info->req_link_speed = link_info->auto_link_speed; - else - link_info->req_link_speed = link_info->force_link_speed; - link_info->advertising = link_info->auto_link_speeds; snprintf(phy_ver, PHY_VER_STR_LEN, " ph %d.%d.%d", link_info->phy_ver[0], link_info->phy_ver[1], diff --combined drivers/net/ethernet/cavium/thunder/nicvf_main.c index a009bc3,90ce93e..bfee298 --- a/drivers/net/ethernet/cavium/thunder/nicvf_main.c +++ b/drivers/net/ethernet/cavium/thunder/nicvf_main.c @@@ -574,7 -574,8 +574,7 @@@ static inline void nicvf_set_rxhash(str static void nicvf_rcv_pkt_handler(struct net_device *netdev, struct napi_struct *napi, - struct cmp_queue *cq, - struct cqe_rx_t *cqe_rx, int cqe_type) + struct cqe_rx_t *cqe_rx) { struct sk_buff *skb; struct nicvf *nic = netdev_priv(netdev); @@@ -590,7 -591,7 +590,7 @@@ } /* Check for errors */ - err = nicvf_check_cqe_rx_errs(nic, cq, cqe_rx); + err = nicvf_check_cqe_rx_errs(nic, cqe_rx); if (err && !cqe_rx->rb_cnt) return; @@@ -681,7 -682,8 +681,7 @@@ loop cq_idx, cq_desc->cqe_type); switch (cq_desc->cqe_type) { case CQE_TYPE_RX: - nicvf_rcv_pkt_handler(netdev, napi, cq, - cq_desc, CQE_TYPE_RX); + nicvf_rcv_pkt_handler(netdev, napi, cq_desc); work_done++; break; case CQE_TYPE_SEND: @@@ -826,7 -828,7 +826,7 @@@ static irqreturn_t nicvf_intr_handler(i nicvf_disable_intr(nic, NICVF_INTR_CQ, qidx); /* Schedule NAPI */ - napi_schedule(&cq_poll->napi); + napi_schedule_irqoff(&cq_poll->napi); /* Clear interrupt */ nicvf_clear_intr(nic, NICVF_INTR_CQ, qidx); @@@ -897,6 -899,31 +897,31 @@@ static void nicvf_disable_msix(struct n } } + static void nicvf_set_irq_affinity(struct nicvf *nic) + { + int vec, cpu; + int irqnum; + + for (vec = 0; vec < nic->num_vec; vec++) { + if (!nic->irq_allocated[vec]) + continue; + + if (!zalloc_cpumask_var(&nic->affinity_mask[vec], GFP_KERNEL)) + return; + /* CQ interrupts */ + if (vec < NICVF_INTR_ID_SQ) + /* Leave CPU0 for RBDR and other interrupts */ + cpu = nicvf_netdev_qidx(nic, vec) + 1; + else + cpu = 0; + + cpumask_set_cpu(cpumask_local_spread(cpu, nic->node), + nic->affinity_mask[vec]); + irqnum = nic->msix_entries[vec].vector; + irq_set_affinity_hint(irqnum, nic->affinity_mask[vec]); + } + } + static int nicvf_register_interrupts(struct nicvf *nic) { int irq, ret = 0; @@@ -942,8 -969,13 +967,13 @@@ ret = request_irq(nic->msix_entries[irq].vector, nicvf_qs_err_intr_handler, 0, nic->irq_name[irq], nic); - if (!ret) - nic->irq_allocated[irq] = true; + if (ret) + goto err; + + nic->irq_allocated[irq] = true; + + /* Set IRQ affinities */ + nicvf_set_irq_affinity(nic); err: if (ret) @@@ -961,6 -993,9 +991,9 @@@ static void nicvf_unregister_interrupts if (!nic->irq_allocated[irq]) continue; + irq_set_affinity_hint(nic->msix_entries[irq].vector, NULL); + free_cpumask_var(nic->affinity_mask[irq]); + if (irq < NICVF_INTR_ID_SQ) free_irq(nic->msix_entries[irq].vector, nic->napi[irq]); else @@@ -1123,6 -1158,7 +1156,6 @@@ int nicvf_stop(struct net_device *netde /* Clear multiqset info */ nic->pnicvf = nic; - nic->sqs_count = 0; return 0; } @@@ -1351,9 -1387,6 +1384,9 @@@ void nicvf_update_stats(struct nicvf *n drv_stats->tx_frames_ok = stats->tx_ucast_frames_ok + stats->tx_bcast_frames_ok + stats->tx_mcast_frames_ok; + drv_stats->rx_frames_ok = stats->rx_ucast_frames + + stats->rx_bcast_frames + + stats->rx_mcast_frames; drv_stats->rx_drops = stats->rx_drop_red + stats->rx_drop_overrun; drv_stats->tx_drops = stats->tx_drops; @@@ -1394,6 -1427,7 +1427,7 @@@ static void nicvf_tx_timeout(struct net netdev_warn(dev, "%s: Transmit timed out, resetting\n", dev->name); + nic->drv_stats.tx_timeout++; schedule_work(&nic->reset_task); } @@@ -1538,9 -1572,6 +1572,9 @@@ static int nicvf_probe(struct pci_dev * nicvf_send_vf_struct(nic); + if (!pass1_silicon(nic->pdev)) + nic->hw_tso = true; + /* Check if this VF is in QS only mode */ if (nic->sqs_mode) return 0; @@@ -1560,6 -1591,9 +1594,6 @@@ netdev->vlan_features = NETIF_F_SG | NETIF_F_IP_CSUM | NETIF_F_TSO; - if (!pass1_silicon(nic->pdev)) - nic->hw_tso = true; - netdev->netdev_ops = &nicvf_netdev_ops; netdev->watchdog_timeo = NICVF_TX_TIMEOUT; diff --combined drivers/net/ethernet/cavium/thunder/nicvf_queues.c index 767347b,5adb208..0dd1abf --- a/drivers/net/ethernet/cavium/thunder/nicvf_queues.c +++ b/drivers/net/ethernet/cavium/thunder/nicvf_queues.c @@@ -78,7 -78,7 +78,7 @@@ static void nicvf_free_q_desc_mem(struc static inline int nicvf_alloc_rcv_buffer(struct nicvf *nic, gfp_t gfp, u32 buf_len, u64 **rbuf) { - int order = get_order(buf_len); + int order = (PAGE_SIZE <= 4096) ? PAGE_ALLOC_COSTLY_ORDER : 0; /* Check if request can be accomodated in previous allocated page */ if (nic->rb_page) { @@@ -96,8 -96,7 +96,7 @@@ nic->rb_page = alloc_pages(gfp | __GFP_COMP | __GFP_NOWARN, order); if (!nic->rb_page) { - netdev_err(nic->netdev, - "Failed to allocate new rcv buffer\n"); + nic->drv_stats.rcv_buffer_alloc_failures++; return -ENOMEM; } nic->rb_page_offset = 0; @@@ -1329,12 -1328,16 +1328,12 @@@ void nicvf_update_sq_stats(struct nicv } /* Check for errors in the receive cmp.queue entry */ -int nicvf_check_cqe_rx_errs(struct nicvf *nic, - struct cmp_queue *cq, struct cqe_rx_t *cqe_rx) +int nicvf_check_cqe_rx_errs(struct nicvf *nic, struct cqe_rx_t *cqe_rx) { struct nicvf_hw_stats *stats = &nic->hw_stats; - struct nicvf_drv_stats *drv_stats = &nic->drv_stats; - if (!cqe_rx->err_level && !cqe_rx->err_opcode) { - drv_stats->rx_frames_ok++; + if (!cqe_rx->err_level && !cqe_rx->err_opcode) return 0; - } if (netif_msg_rx_err(nic)) netdev_err(nic->netdev, diff --combined drivers/net/ethernet/mellanox/mlx4/en_netdev.c index f191a16,01d6a96..96d95cb --- a/drivers/net/ethernet/mellanox/mlx4/en_netdev.c +++ b/drivers/net/ethernet/mellanox/mlx4/en_netdev.c @@@ -69,6 -69,15 +69,15 @@@ int mlx4_en_setup_tc(struct net_device return 0; } + static int __mlx4_en_setup_tc(struct net_device *dev, u32 handle, __be16 proto, + struct tc_to_netdev *tc) + { + if (handle != TC_H_ROOT || tc->type != TC_SETUP_MQPRIO) + return -EINVAL; + + return mlx4_en_setup_tc(dev, tc->tc); + } + #ifdef CONFIG_RFS_ACCEL struct mlx4_en_filter { @@@ -2344,6 -2353,8 +2353,6 @@@ out /* set offloads */ priv->dev->hw_enc_features |= NETIF_F_IP_CSUM | NETIF_F_RXCSUM | NETIF_F_TSO | NETIF_F_GSO_UDP_TUNNEL; - priv->dev->hw_features |= NETIF_F_GSO_UDP_TUNNEL; - priv->dev->features |= NETIF_F_GSO_UDP_TUNNEL; } static void mlx4_en_del_vxlan_offloads(struct work_struct *work) @@@ -2354,6 -2365,8 +2363,6 @@@ /* unset offloads */ priv->dev->hw_enc_features &= ~(NETIF_F_IP_CSUM | NETIF_F_RXCSUM | NETIF_F_TSO | NETIF_F_GSO_UDP_TUNNEL); - priv->dev->hw_features &= ~NETIF_F_GSO_UDP_TUNNEL; - priv->dev->features &= ~NETIF_F_GSO_UDP_TUNNEL; ret = mlx4_SET_PORT_VXLAN(priv->mdev->dev, priv->port, VXLAN_STEER_BY_OUTER_MAC, 0); @@@ -2462,7 -2475,7 +2471,7 @@@ static const struct net_device_ops mlx4 #endif .ndo_set_features = mlx4_en_set_features, .ndo_fix_features = mlx4_en_fix_features, - .ndo_setup_tc = mlx4_en_setup_tc, + .ndo_setup_tc = __mlx4_en_setup_tc, #ifdef CONFIG_RFS_ACCEL .ndo_rx_flow_steer = mlx4_en_filter_rfs, #endif @@@ -2500,7 -2513,7 +2509,7 @@@ static const struct net_device_ops mlx4 #endif .ndo_set_features = mlx4_en_set_features, .ndo_fix_features = mlx4_en_fix_features, - .ndo_setup_tc = mlx4_en_setup_tc, + .ndo_setup_tc = __mlx4_en_setup_tc, #ifdef CONFIG_RFS_ACCEL .ndo_rx_flow_steer = mlx4_en_filter_rfs, #endif @@@ -2976,11 -2989,6 +2985,11 @@@ int mlx4_en_init_netdev(struct mlx4_en_ priv->rss_hash_fn = ETH_RSS_HASH_TOP; } + if (mdev->dev->caps.tunnel_offload_mode == MLX4_TUNNEL_OFFLOAD_MODE_VXLAN) { + dev->hw_features |= NETIF_F_GSO_UDP_TUNNEL; + dev->features |= NETIF_F_GSO_UDP_TUNNEL; + } + mdev->pndev[port] = dev; mdev->upper[port] = NULL; diff --combined drivers/net/ethernet/renesas/ravb_main.c index 744d780,331c596..88656ce --- a/drivers/net/ethernet/renesas/ravb_main.c +++ b/drivers/net/ethernet/renesas/ravb_main.c @@@ -2,7 -2,7 +2,7 @@@ * * Copyright (C) 2014-2015 Renesas Electronics Corporation * Copyright (C) 2015 Renesas Solutions Corp. - * Copyright (C) 2015 Cogent Embedded, Inc. <source(a)cogentembedded.com> + * Copyright (C) 2015-2016 Cogent Embedded, Inc. <source(a)cogentembedded.com> * * Based on the SuperH Ethernet driver * @@@ -42,6 -42,12 +42,12 @@@ NETIF_MSG_RX_ERR | \ NETIF_MSG_TX_ERR) + void ravb_modify(struct net_device *ndev, enum ravb_reg reg, u32 clear, + u32 set) + { + ravb_write(ndev, (ravb_read(ndev, reg) & ~clear) | set, reg); + } + int ravb_wait(struct net_device *ndev, enum ravb_reg reg, u32 mask, u32 value) { int i; @@@ -59,8 -65,7 +65,7 @@@ static int ravb_config(struct net_devic int error; /* Set config mode */ - ravb_write(ndev, (ravb_read(ndev, CCC) & ~CCC_OPC) | CCC_OPC_CONFIG, - CCC); + ravb_modify(ndev, CCC, CCC_OPC, CCC_OPC_CONFIG); /* Check if the operating mode is changed to the config mode */ error = ravb_wait(ndev, CSR, CSR_OPS, CSR_OPS_CONFIG); if (error) @@@ -72,13 -77,8 +77,8 @@@ static void ravb_set_duplex(struct net_device *ndev) { struct ravb_private *priv = netdev_priv(ndev); - u32 ecmr = ravb_read(ndev, ECMR); - if (priv->duplex) /* Full */ - ecmr |= ECMR_DM; - else /* Half */ - ecmr &= ~ECMR_DM; - ravb_write(ndev, ecmr, ECMR); + ravb_modify(ndev, ECMR, ECMR_DM, priv->duplex ? ECMR_DM : 0); } static void ravb_set_rate(struct net_device *ndev) @@@ -92,8 -92,6 +92,6 @@@ case 1000: /* 1000BASE */ ravb_write(ndev, GECMR_SPEED_1000, GECMR); break; - default: - break; } } @@@ -131,13 -129,8 +129,8 @@@ static void ravb_mdio_ctrl(struct mdiob { struct ravb_private *priv = container_of(ctrl, struct ravb_private, mdiobb); - u32 pir = ravb_read(priv->ndev, PIR); - if (set) - pir |= mask; - else - pir &= ~mask; - ravb_write(priv->ndev, pir, PIR); + ravb_modify(priv->ndev, PIR, mask, set ? mask : 0); } /* MDC pin control */ @@@ -393,9 -386,9 +386,9 @@@ static int ravb_dmac_init(struct net_de ravb_ring_format(ndev, RAVB_NC); #if defined(__LITTLE_ENDIAN) - ravb_write(ndev, ravb_read(ndev, CCC) & ~CCC_BOC, CCC); + ravb_modify(ndev, CCC, CCC_BOC, 0); #else - ravb_write(ndev, ravb_read(ndev, CCC) | CCC_BOC, CCC); + ravb_modify(ndev, CCC, CCC_BOC, CCC_BOC); #endif /* Set AVB RX */ @@@ -418,8 -411,7 +411,7 @@@ ravb_write(ndev, TIC_FTE0 | TIC_FTE1 | TIC_TFUE, TIC); /* Setting the control will start the AVB-DMAC process. */ - ravb_write(ndev, (ravb_read(ndev, CCC) & ~CCC_OPC) | CCC_OPC_OPERATION, - CCC); + ravb_modify(ndev, CCC, CCC_OPC, CCC_OPC_OPERATION); return 0; } @@@ -493,7 -485,7 +485,7 @@@ static void ravb_get_tx_tstamp(struct n break; } } - ravb_write(ndev, ravb_read(ndev, TCCR) | TCCR_TFR, TCCR); + ravb_modify(ndev, TCCR, TCCR_TFR, TCCR_TFR); } } @@@ -613,13 -605,13 +605,13 @@@ static bool ravb_rx(struct net_device * static void ravb_rcv_snd_disable(struct net_device *ndev) { /* Disable TX and RX */ - ravb_write(ndev, ravb_read(ndev, ECMR) & ~(ECMR_RE | ECMR_TE), ECMR); + ravb_modify(ndev, ECMR, ECMR_RE | ECMR_TE, 0); } static void ravb_rcv_snd_enable(struct net_device *ndev) { /* Enable TX and RX */ - ravb_write(ndev, ravb_read(ndev, ECMR) | ECMR_RE | ECMR_TE, ECMR); + ravb_modify(ndev, ECMR, ECMR_RE | ECMR_TE, ECMR_RE | ECMR_TE); } /* function for waiting dma process finished */ @@@ -812,8 -804,8 +804,8 @@@ static int ravb_poll(struct napi_struc /* Re-enable RX/TX interrupts */ spin_lock_irqsave(&priv->lock, flags); - ravb_write(ndev, ravb_read(ndev, RIC0) | mask, RIC0); - ravb_write(ndev, ravb_read(ndev, TIC) | mask, TIC); + ravb_modify(ndev, RIC0, mask, mask); + ravb_modify(ndev, TIC, mask, mask); mmiowb(); spin_unlock_irqrestore(&priv->lock, flags); @@@ -852,8 -844,7 +844,7 @@@ static void ravb_adjust_link(struct net ravb_set_rate(ndev); } if (!priv->link) { - ravb_write(ndev, ravb_read(ndev, ECMR) & ~ECMR_TXF, - ECMR); + ravb_modify(ndev, ECMR, ECMR_TXF, 0); new_state = true; priv->link = phydev->link; if (priv->no_avb_link) @@@ -1139,8 -1130,7 +1130,8 @@@ static int ravb_set_ringparam(struct ne if (netif_running(ndev)) { netif_device_detach(ndev); /* Stop PTP Clock driver */ - ravb_ptp_stop(ndev); + if (priv->chip_id == RCAR_GEN2) + ravb_ptp_stop(ndev); /* Wait for DMA stopping */ error = ravb_stop_dma(ndev); if (error) { @@@ -1171,8 -1161,7 +1162,8 @@@ ravb_emac_init(ndev); /* Initialise PTP Clock driver */ - ravb_ptp_init(ndev, priv->pdev); + if (priv->chip_id == RCAR_GEN2) + ravb_ptp_init(ndev, priv->pdev); netif_device_attach(ndev); } @@@ -1300,8 -1289,7 +1291,8 @@@ static void ravb_tx_timeout_work(struc netif_tx_stop_all_queues(ndev); /* Stop PTP Clock driver */ - ravb_ptp_stop(ndev); + if (priv->chip_id == RCAR_GEN2) + ravb_ptp_stop(ndev); /* Wait for DMA stopping */ ravb_stop_dma(ndev); @@@ -1314,8 -1302,7 +1305,8 @@@ ravb_emac_init(ndev); /* Initialise PTP Clock driver */ - ravb_ptp_init(ndev, priv->pdev); + if (priv->chip_id == RCAR_GEN2) + ravb_ptp_init(ndev, priv->pdev); netif_tx_start_all_queues(ndev); } @@@ -1397,7 -1384,7 +1388,7 @@@ static netdev_tx_t ravb_start_xmit(stru desc--; desc->die_dt = DT_FSTART; - ravb_write(ndev, ravb_read(ndev, TCCR) | (TCCR_TSRQ0 << q), TCCR); + ravb_modify(ndev, TCCR, TCCR_TSRQ0 << q, TCCR_TSRQ0 << q); priv->cur_tx[q] += NUM_TX_DESC; if (priv->cur_tx[q] - priv->dirty_tx[q] > @@@ -1472,15 -1459,10 +1463,10 @@@ static void ravb_set_rx_mode(struct net { struct ravb_private *priv = netdev_priv(ndev); unsigned long flags; - u32 ecmr; spin_lock_irqsave(&priv->lock, flags); - ecmr = ravb_read(ndev, ECMR); - if (ndev->flags & IFF_PROMISC) - ecmr |= ECMR_PRM; - else - ecmr &= ~ECMR_PRM; - ravb_write(ndev, ecmr, ECMR); + ravb_modify(ndev, ECMR, ECMR_PRM, + ndev->flags & IFF_PROMISC ? ECMR_PRM : 0); mmiowb(); spin_unlock_irqrestore(&priv->lock, flags); } @@@ -1808,23 -1790,25 +1794,21 @@@ static int ravb_probe(struct platform_d /* Set AVB config mode */ if (chip_id == RCAR_GEN2) { - ravb_write(ndev, (ravb_read(ndev, CCC) & ~CCC_OPC) | - CCC_OPC_CONFIG, CCC); + ravb_modify(ndev, CCC, CCC_OPC, CCC_OPC_CONFIG); /* Set CSEL value */ - ravb_write(ndev, (ravb_read(ndev, CCC) & ~CCC_CSEL) | - CCC_CSEL_HPB, CCC); + ravb_modify(ndev, CCC, CCC_CSEL, CCC_CSEL_HPB); } else { - ravb_write(ndev, (ravb_read(ndev, CCC) & ~CCC_OPC) | - CCC_OPC_CONFIG | CCC_GAC | CCC_CSEL_HPB, CCC); + ravb_modify(ndev, CCC, CCC_OPC, CCC_OPC_CONFIG | + CCC_GAC | CCC_CSEL_HPB); } - /* Set CSEL value */ - ravb_write(ndev, (ravb_read(ndev, CCC) & ~CCC_CSEL) | CCC_CSEL_HPB, - CCC); - /* Set GTI value */ error = ravb_set_gti(ndev); if (error) goto out_release; /* Request GTI loading */ - ravb_write(ndev, ravb_read(ndev, GCCR) | GCCR_LTI, GCCR); + ravb_modify(ndev, GCCR, GCCR_LTI, GCCR_LTI); /* Allocate descriptor base address table */ priv->desc_bat_size = sizeof(struct ravb_desc) * DBAT_ENTRY_NUM; diff --combined drivers/net/geneve.c index 0bf7edd,dfbe3ca..dedb26a --- a/drivers/net/geneve.c +++ b/drivers/net/geneve.c @@@ -72,6 -72,7 +72,7 @@@ struct geneve_dev bool collect_md; struct gro_cells gro_cells; u32 flags; + struct dst_cache dst_cache; }; /* Geneve device flags */ @@@ -109,6 -110,11 +110,11 @@@ static __be64 vni_to_tunnel_id(const __ #endif } + static sa_family_t geneve_get_sk_family(struct geneve_sock *gs) + { + return gs->sock->sk->sk_family; + } + static struct geneve_dev *geneve_lookup(struct geneve_sock *gs, __be32 addr, u8 vni[]) { @@@ -152,58 -158,60 +158,60 @@@ static inline struct genevehdr *geneve_ return (struct genevehdr *)(udp_hdr(skb) + 1); } - /* geneve receive/decap routine */ - static void geneve_rx(struct geneve_sock *gs, struct sk_buff *skb) + static struct geneve_dev *geneve_lookup_skb(struct geneve_sock *gs, + struct sk_buff *skb) { - struct genevehdr *gnvh = geneve_hdr(skb); - struct metadata_dst *tun_dst = NULL; - struct geneve_dev *geneve = NULL; - struct pcpu_sw_netstats *stats; - struct iphdr *iph = NULL; + u8 *vni; __be32 addr; static u8 zero_vni[3]; - u8 *vni; - int err = 0; - sa_family_t sa_family; #if IS_ENABLED(CONFIG_IPV6) - struct ipv6hdr *ip6h = NULL; - struct in6_addr addr6; static struct in6_addr zero_addr6; #endif - sa_family = gs->sock->sk->sk_family; + if (geneve_get_sk_family(gs) == AF_INET) { + struct iphdr *iph; - if (sa_family == AF_INET) { iph = ip_hdr(skb); /* outer IP header... */ if (gs->collect_md) { vni = zero_vni; addr = 0; } else { - vni = gnvh->vni; - + vni = geneve_hdr(skb)->vni; addr = iph->saddr; } - geneve = geneve_lookup(gs, addr, vni); + return geneve_lookup(gs, addr, vni); #if IS_ENABLED(CONFIG_IPV6) - } else if (sa_family == AF_INET6) { + } else if (geneve_get_sk_family(gs) == AF_INET6) { + struct ipv6hdr *ip6h; + struct in6_addr addr6; + ip6h = ipv6_hdr(skb); /* outer IPv6 header... */ if (gs->collect_md) { vni = zero_vni; addr6 = zero_addr6; } else { - vni = gnvh->vni; - + vni = geneve_hdr(skb)->vni; addr6 = ip6h->saddr; } - geneve = geneve6_lookup(gs, addr6, vni); + return geneve6_lookup(gs, addr6, vni); #endif } - if (!geneve) - goto drop; + return NULL; + } + + /* geneve receive/decap routine */ + static void geneve_rx(struct geneve_dev *geneve, struct geneve_sock *gs, + struct sk_buff *skb) + { + struct genevehdr *gnvh = geneve_hdr(skb); + struct metadata_dst *tun_dst = NULL; + struct pcpu_sw_netstats *stats; + int err = 0; + void *oiph; if (ip_tunnel_collect_metadata() || gs->collect_md) { __be16 flags; @@@ -212,7 -220,7 +220,7 @@@ (gnvh->oam ? TUNNEL_OAM : 0) | (gnvh->critical ? TUNNEL_CRIT_OPT : 0); - tun_dst = udp_tun_rx_dst(skb, sa_family, flags, + tun_dst = udp_tun_rx_dst(skb, geneve_get_sk_family(gs), flags, vni_to_tunnel_id(gnvh->vni), gnvh->opt_len * 4); if (!tun_dst) @@@ -229,7 -237,6 +237,6 @@@ } skb_reset_mac_header(skb); - skb_scrub_packet(skb, !net_eq(geneve->net, dev_net(geneve->dev))); skb->protocol = eth_type_trans(skb, geneve->dev); skb_postpull_rcsum(skb, eth_hdr(skb), ETH_HLEN); @@@ -240,25 -247,27 +247,27 @@@ if (ether_addr_equal(eth_hdr(skb)->h_source, geneve->dev->dev_addr)) goto drop; + oiph = skb_network_header(skb); skb_reset_network_header(skb); - if (iph) - err = IP_ECN_decapsulate(iph, skb); + if (geneve_get_sk_family(gs) == AF_INET) + err = IP_ECN_decapsulate(oiph, skb); #if IS_ENABLED(CONFIG_IPV6) - if (ip6h) - err = IP6_ECN_decapsulate(ip6h, skb); + else + err = IP6_ECN_decapsulate(oiph, skb); #endif if (unlikely(err)) { if (log_ecn_error) { - if (iph) + if (geneve_get_sk_family(gs) == AF_INET) net_info_ratelimited("non-ECT from %pI4 " "with TOS=%#x\n", - &iph->saddr, iph->tos); + &((struct iphdr *)oiph)->saddr, + ((struct iphdr *)oiph)->tos); #if IS_ENABLED(CONFIG_IPV6) - if (ip6h) + else net_info_ratelimited("non-ECT from %pI6\n", - &ip6h->saddr); + &((struct ipv6hdr *)oiph)->saddr); #endif } if (err > 1) { @@@ -297,6 -306,13 +306,13 @@@ static int geneve_init(struct net_devic return err; } + err = dst_cache_init(&geneve->dst_cache, GFP_KERNEL); + if (err) { + free_percpu(dev->tstats); + gro_cells_destroy(&geneve->gro_cells); + return err; + } + return 0; } @@@ -304,6 -320,7 +320,7 @@@ static void geneve_uninit(struct net_de { struct geneve_dev *geneve = netdev_priv(dev); + dst_cache_destroy(&geneve->dst_cache); gro_cells_destroy(&geneve->gro_cells); free_percpu(dev->tstats); } @@@ -312,6 -329,7 +329,7 @@@ static int geneve_udp_encap_recv(struct sock *sk, struct sk_buff *skb) { struct genevehdr *geneveh; + struct geneve_dev *geneve; struct geneve_sock *gs; int opts_len; @@@ -327,16 -345,21 +345,21 @@@ if (unlikely(geneveh->proto_type != htons(ETH_P_TEB))) goto error; - opts_len = geneveh->opt_len * 4; - if (iptunnel_pull_header(skb, GENEVE_BASE_HLEN + opts_len, - htons(ETH_P_TEB))) - goto drop; - gs = rcu_dereference_sk_user_data(sk); if (!gs) goto drop; - geneve_rx(gs, skb); + geneve = geneve_lookup_skb(gs, skb); + if (!geneve) + goto drop; + + opts_len = geneveh->opt_len * 4; + if (iptunnel_pull_header(skb, GENEVE_BASE_HLEN + opts_len, + htons(ETH_P_TEB), + !net_eq(geneve->net, dev_net(geneve->dev)))) + goto drop; + + geneve_rx(geneve, gs, skb); return 0; drop: @@@ -383,7 -406,7 +406,7 @@@ static void geneve_notify_add_rx_port(s struct net_device *dev; struct sock *sk = gs->sock->sk; struct net *net = sock_net(sk); - sa_family_t sa_family = sk->sk_family; + sa_family_t sa_family = geneve_get_sk_family(gs); __be16 port = inet_sk(sk)->inet_sport; int err; @@@ -544,7 -567,7 +567,7 @@@ static void geneve_notify_del_rx_port(s struct net_device *dev; struct sock *sk = gs->sock->sk; struct net *net = sock_net(sk); - sa_family_t sa_family = sk->sk_family; + sa_family_t sa_family = geneve_get_sk_family(gs); __be16 port = inet_sk(sk)->inet_sport; rcu_read_lock(); @@@ -587,7 -610,7 +610,7 @@@ static struct geneve_sock *geneve_find_ list_for_each_entry(gs, &gn->sock_list, list) { if (inet_sk(gs->sock->sk)->inet_sport == dst_port && - inet_sk(gs->sock->sk)->sk.sk_family == family) { + geneve_get_sk_family(gs) == family) { return gs; } } @@@ -753,7 -776,9 +776,9 @@@ static struct rtable *geneve_get_v4_rt( struct ip_tunnel_info *info) { struct geneve_dev *geneve = netdev_priv(dev); + struct dst_cache *dst_cache; struct rtable *rt = NULL; + bool use_cache = true; __u8 tos; memset(fl4, 0, sizeof(*fl4)); @@@ -764,16 -789,26 +789,26 @@@ fl4->daddr = info->key.u.ipv4.dst; fl4->saddr = info->key.u.ipv4.src; fl4->flowi4_tos = RT_TOS(info->key.tos); + dst_cache = &info->dst_cache; } else { tos = geneve->tos; if (tos == 1) { const struct iphdr *iip = ip_hdr(skb); tos = ip_tunnel_get_dsfield(iip, skb); + use_cache = false; } fl4->flowi4_tos = RT_TOS(tos); fl4->daddr = geneve->remote.sin.sin_addr.s_addr; + dst_cache = &geneve->dst_cache; + } + + use_cache = use_cache && !skb->mark; + if (use_cache) { + rt = dst_cache_get_ip4(dst_cache, &fl4->saddr); + if (rt) + return rt; } rt = ip_route_output_key(geneve->net, fl4); @@@ -786,6 -821,8 +821,8 @@@ ip_rt_put(rt); return ERR_PTR(-ELOOP); } + if (use_cache) + dst_cache_set_ip4(dst_cache, &rt->dst, fl4->saddr); return rt; } @@@ -798,6 -835,8 +835,8 @@@ static struct dst_entry *geneve_get_v6_ struct geneve_dev *geneve = netdev_priv(dev); struct geneve_sock *gs6 = geneve->sock6; struct dst_entry *dst = NULL; + struct dst_cache *dst_cache; + bool use_cache = true; __u8 prio; memset(fl6, 0, sizeof(*fl6)); @@@ -808,16 -847,26 +847,26 @@@ fl6->daddr = info->key.u.ipv6.dst; fl6->saddr = info->key.u.ipv6.src; fl6->flowi6_tos = RT_TOS(info->key.tos); + dst_cache = &info->dst_cache; } else { prio = geneve->tos; if (prio == 1) { const struct iphdr *iip = ip_hdr(skb); prio = ip_tunnel_get_dsfield(iip, skb); + use_cache = false; } fl6->flowi6_tos = RT_TOS(prio); fl6->daddr = geneve->remote.sin6.sin6_addr; + dst_cache = &geneve->dst_cache; + } + + use_cache = use_cache && !skb->mark; + if (use_cache) { + dst = dst_cache_get_ip6(dst_cache, &fl6->saddr); + if (dst) + return dst; } if (ipv6_stub->ipv6_dst_lookup(geneve->net, gs6->sock->sk, &dst, fl6)) { @@@ -830,6 -879,8 +879,8 @@@ return ERR_PTR(-ELOOP); } + if (use_cache) + dst_cache_set_ip6(dst_cache, dst, &fl6->saddr); return dst; } #endif @@@ -1039,34 -1090,6 +1090,34 @@@ static netdev_tx_t geneve_xmit(struct s return geneve_xmit_skb(skb, dev, info); } +static int __geneve_change_mtu(struct net_device *dev, int new_mtu, bool strict) +{ + /* The max_mtu calculation does not take account of GENEVE + * options, to avoid excluding potentially valid + * configurations. + */ + int max_mtu = IP_MAX_MTU - GENEVE_BASE_HLEN - sizeof(struct iphdr) + - dev->hard_header_len; + + if (new_mtu < 68) + return -EINVAL; + + if (new_mtu > max_mtu) { + if (strict) + return -EINVAL; + + new_mtu = max_mtu; + } + + dev->mtu = new_mtu; + return 0; +} + +static int geneve_change_mtu(struct net_device *dev, int new_mtu) +{ + return __geneve_change_mtu(dev, new_mtu, true); +} + static int geneve_fill_metadata_dst(struct net_device *dev, struct sk_buff *skb) { struct ip_tunnel_info *info = skb_tunnel_info(skb); @@@ -1111,7 -1134,7 +1162,7 @@@ static const struct net_device_ops gene .ndo_stop = geneve_stop, .ndo_start_xmit = geneve_xmit, .ndo_get_stats64 = ip_tunnel_get_stats64, - .ndo_change_mtu = eth_change_mtu, + .ndo_change_mtu = geneve_change_mtu, .ndo_validate_addr = eth_validate_addr, .ndo_set_mac_address = eth_mac_addr, .ndo_fill_metadata_dst = geneve_fill_metadata_dst, @@@ -1178,7 -1201,6 +1229,7 @@@ static void geneve_setup(struct net_dev dev->hw_features |= NETIF_F_GSO_SOFTWARE; netif_keep_dst(dev); + dev->priv_flags &= ~IFF_TX_SKB_SHARING; dev->priv_flags |= IFF_LIVE_ADDR_CHANGE | IFF_NO_QUEUE; eth_hw_addr_random(dev); } @@@ -1301,6 -1323,8 +1352,8 @@@ static int geneve_configure(struct net return -EPERM; } + dst_cache_reset(&geneve->dst_cache); + err = register_netdevice(dev); if (err) return err; @@@ -1470,23 -1494,12 +1523,23 @@@ struct net_device *geneve_dev_create_fb return dev; err = geneve_configure(net, dev, &geneve_remote_unspec, - 0, 0, 0, htons(dst_port), true, 0); - if (err) { - free_netdev(dev); - return ERR_PTR(err); - } + 0, 0, 0, htons(dst_port), true, + GENEVE_F_UDP_ZERO_CSUM6_RX); + if (err) + goto err; + + /* openvswitch users expect packet sizes to be unrestricted, + * so set the largest MTU we can. + */ + err = __geneve_change_mtu(dev, IP_MAX_MTU, false); + if (err) + goto err; + return dev; + + err: + free_netdev(dev); + return ERR_PTR(err); } EXPORT_SYMBOL_GPL(geneve_dev_create_fb); diff --combined drivers/net/hyperv/netvsc_drv.c index 98e34fe,202e2b1..2b6595e --- a/drivers/net/hyperv/netvsc_drv.c +++ b/drivers/net/hyperv/netvsc_drv.c @@@ -43,6 -43,11 +43,11 @@@ #define RING_SIZE_MIN 64 #define LINKCHANGE_INT (2 * HZ) + #define NETVSC_HW_FEATURES (NETIF_F_RXCSUM | \ + NETIF_F_SG | \ + NETIF_F_TSO | \ + NETIF_F_TSO6 | \ + NETIF_F_HW_CSUM) static int ring_size = 128; module_param(ring_size, int, S_IRUGO); MODULE_PARM_DESC(ring_size, "Ring buffer size (# of pages)"); @@@ -545,6 -550,8 +550,8 @@@ do_send packet->page_buf_cnt = init_page_array(rndis_msg, rndis_msg_size, skb, packet, &pb); + /* timestamp packet in software */ + skb_tx_timestamp(skb); ret = netvsc_send(net_device_ctx->device_ctx, packet, rndis_msg, &pb, skb); @@@ -915,6 -922,7 +922,7 @@@ static const struct ethtool_ops ethtool .get_link = ethtool_op_get_link, .get_channels = netvsc_get_channels, .set_channels = netvsc_set_channels, + .get_ts_info = ethtool_op_get_ts_info, }; static const struct net_device_ops device_ops = { @@@ -1081,17 -1089,12 +1089,15 @@@ static int netvsc_probe(struct hv_devic net->netdev_ops = &device_ops; - net->hw_features = NETIF_F_RXCSUM | NETIF_F_SG | NETIF_F_IP_CSUM | - NETIF_F_TSO; - net->features = NETIF_F_HW_VLAN_CTAG_TX | NETIF_F_SG | NETIF_F_RXCSUM | - NETIF_F_IP_CSUM | NETIF_F_TSO; + net->hw_features = NETVSC_HW_FEATURES; + net->features = NETVSC_HW_FEATURES | NETIF_F_HW_VLAN_CTAG_TX; net->ethtool_ops = &ethtool_ops; SET_NETDEV_DEV(net, &dev->device); + /* We always need headroom for rndis header */ + net->needed_headroom = RNDIS_AND_PPI_SIZE; + /* Notify the netvsc driver of the new device */ memset(&device_info, 0, sizeof(device_info)); device_info.ring_size = ring_size; diff --combined drivers/net/phy/bcm7xxx.c index db507e3,9b31104..b881a7b1 --- a/drivers/net/phy/bcm7xxx.c +++ b/drivers/net/phy/bcm7xxx.c @@@ -24,7 -24,7 +24,7 @@@ #define MII_BCM7XXX_100TX_FALSE_CAR 0x13 #define MII_BCM7XXX_100TX_DISC 0x14 #define MII_BCM7XXX_AUX_MODE 0x1d - #define MII_BCM7XX_64CLK_MDIO BIT(12) + #define MII_BCM7XXX_64CLK_MDIO BIT(12) #define MII_BCM7XXX_TEST 0x1f #define MII_BCM7XXX_SHD_MODE_2 BIT(2) @@@ -247,9 -247,13 +247,9 @@@ static int bcm7xxx_config_init(struct p int ret; /* Enable 64 clock MDIO */ - phy_write(phydev, MII_BCM7XXX_AUX_MODE, MII_BCM7XX_64CLK_MDIO); + phy_write(phydev, MII_BCM7XXX_AUX_MODE, MII_BCM7XXX_64CLK_MDIO); phy_read(phydev, MII_BCM7XXX_AUX_MODE); - /* Workaround only required for 100Mbits/sec capable PHYs */ - if (phydev->supported & PHY_GBIT_FEATURES) - return 0; - /* set shadow mode 2 */ ret = phy_set_clr_bits(phydev, MII_BCM7XXX_TEST, MII_BCM7XXX_SHD_MODE_2, MII_BCM7XXX_SHD_MODE_2); @@@ -266,7 -270,7 +266,7 @@@ phy_write(phydev, MII_BCM7XXX_100TX_FALSE_CAR, 0x7555); /* reset shadow mode 2 */ - ret = phy_set_clr_bits(phydev, MII_BCM7XXX_TEST, MII_BCM7XXX_SHD_MODE_2, 0); + ret = phy_set_clr_bits(phydev, MII_BCM7XXX_TEST, 0, MII_BCM7XXX_SHD_MODE_2); if (ret < 0) return ret; @@@ -303,6 -307,11 +303,6 @@@ static int bcm7xxx_suspend(struct phy_d return 0; } -static int bcm7xxx_dummy_config_init(struct phy_device *phydev) -{ - return 0; -} - #define BCM7XXX_28NM_GPHY(_oui, _name) \ { \ .phy_id = (_oui), \ @@@ -317,6 -326,21 +317,21 @@@ .resume = bcm7xxx_28nm_resume, \ } + #define BCM7XXX_40NM_EPHY(_oui, _name) \ + { \ + .phy_id = (_oui), \ + .phy_id_mask = 0xfffffff0, \ + .name = _name, \ + .features = PHY_BASIC_FEATURES | \ + SUPPORTED_Pause | SUPPORTED_Asym_Pause, \ + .flags = PHY_IS_INTERNAL, \ + .config_init = bcm7xxx_config_init, \ + .config_aneg = genphy_config_aneg, \ + .read_status = genphy_read_status, \ + .suspend = bcm7xxx_suspend, \ + .resume = bcm7xxx_config_init, \ + } + static struct phy_driver bcm7xxx_driver[] = { BCM7XXX_28NM_GPHY(PHY_ID_BCM7250, "Broadcom BCM7250"), BCM7XXX_28NM_GPHY(PHY_ID_BCM7364, "Broadcom BCM7364"), @@@ -324,43 -348,34 +339,10 @@@ BCM7XXX_28NM_GPHY(PHY_ID_BCM7439, "Broadcom BCM7439"), BCM7XXX_28NM_GPHY(PHY_ID_BCM7439_2, "Broadcom BCM7439 (2)"), BCM7XXX_28NM_GPHY(PHY_ID_BCM7445, "Broadcom BCM7445"), - { - .phy_id = PHY_ID_BCM7425, - .phy_id_mask = 0xfffffff0, - .name = "Broadcom BCM7425", - .features = PHY_BASIC_FEATURES | - SUPPORTED_Pause | SUPPORTED_Asym_Pause, - .flags = PHY_IS_INTERNAL, - .config_init = bcm7xxx_config_init, - .config_aneg = genphy_config_aneg, - .read_status = genphy_read_status, - .suspend = bcm7xxx_suspend, - .resume = bcm7xxx_config_init, - }, { - .phy_id = PHY_ID_BCM7429, - .phy_id_mask = 0xfffffff0, - .name = "Broadcom BCM7429", - .features = PHY_BASIC_FEATURES | - SUPPORTED_Pause | SUPPORTED_Asym_Pause, - .flags = PHY_IS_INTERNAL, - .config_init = bcm7xxx_config_init, - .config_aneg = genphy_config_aneg, - .read_status = genphy_read_status, - .suspend = bcm7xxx_suspend, - .resume = bcm7xxx_config_init, - }, { - .phy_id = PHY_ID_BCM7435, - .phy_id_mask = 0xfffffff0, - .name = "Broadcom BCM7435", - .features = PHY_BASIC_FEATURES | - SUPPORTED_Pause | SUPPORTED_Asym_Pause, - .flags = PHY_IS_INTERNAL, - .config_init = bcm7xxx_config_init, - .config_aneg = genphy_config_aneg, - .read_status = genphy_read_status, - .suspend = bcm7xxx_suspend, - .resume = bcm7xxx_config_init, - } }; + BCM7XXX_40NM_EPHY(PHY_ID_BCM7425, "Broadcom BCM7425"), + BCM7XXX_40NM_EPHY(PHY_ID_BCM7429, "Broadcom BCM7429"), + BCM7XXX_40NM_EPHY(PHY_ID_BCM7435, "Broadcom BCM7435"), -{ - .phy_id = PHY_BCM_OUI_4, - .phy_id_mask = 0xffff0000, - .name = "Broadcom BCM7XXX 40nm", - .features = PHY_GBIT_FEATURES | - SUPPORTED_Pause | SUPPORTED_Asym_Pause, - .flags = PHY_IS_INTERNAL, - .config_init = bcm7xxx_config_init, - .config_aneg = genphy_config_aneg, - .read_status = genphy_read_status, - .suspend = bcm7xxx_suspend, - .resume = bcm7xxx_config_init, -}, { - .phy_id = PHY_BCM_OUI_5, - .phy_id_mask = 0xffffff00, - .name = "Broadcom BCM7XXX 65nm", - .features = PHY_BASIC_FEATURES | - SUPPORTED_Pause | SUPPORTED_Asym_Pause, - .flags = PHY_IS_INTERNAL, - .config_init = bcm7xxx_dummy_config_init, - .config_aneg = genphy_config_aneg, - .read_status = genphy_read_status, - .suspend = bcm7xxx_suspend, - .resume = bcm7xxx_config_init, -} }; ++}; static struct mdio_device_id __maybe_unused bcm7xxx_tbl[] = { { PHY_ID_BCM7250, 0xfffffff0, }, @@@ -371,6 -386,8 +353,6 @@@ { PHY_ID_BCM7439, 0xfffffff0, }, { PHY_ID_BCM7435, 0xfffffff0, }, { PHY_ID_BCM7445, 0xfffffff0, }, - { PHY_BCM_OUI_4, 0xffff0000 }, - { PHY_BCM_OUI_5, 0xffffff00 }, { } }; diff --combined drivers/net/phy/marvell.c index ab1d0fc,d0168f1..9fb9d80 --- a/drivers/net/phy/marvell.c +++ b/drivers/net/phy/marvell.c @@@ -133,6 -133,11 +133,11 @@@ #define MII_88E3016_DISABLE_SCRAMBLER 0x0200 #define MII_88E3016_AUTO_MDIX_CROSSOVER 0x0030 + #define MII_88E1510_GEN_CTRL_REG_1 0x14 + #define MII_88E1510_GEN_CTRL_REG_1_MODE_MASK 0x7 + #define MII_88E1510_GEN_CTRL_REG_1_MODE_SGMII 0x1 /* SGMII to copper */ + #define MII_88E1510_GEN_CTRL_REG_1_RESET 0x8000 /* Soft reset */ + MODULE_DESCRIPTION("Marvell PHY driver"); MODULE_AUTHOR("Andy Fleming"); MODULE_LICENSE("GPL"); @@@ -438,6 -443,41 +443,41 @@@ static int m88e1318_config_aneg(struct return m88e1121_config_aneg(phydev); } + static int m88e1510_config_init(struct phy_device *phydev) + { + int err; + int temp; + + /* SGMII-to-Copper mode initialization */ + if (phydev->interface == PHY_INTERFACE_MODE_SGMII) { + /* Select page 18 */ + err = phy_write(phydev, MII_MARVELL_PHY_PAGE, 18); + if (err < 0) + return err; + + /* In reg 20, write MODE[2:0] = 0x1 (SGMII to Copper) */ + temp = phy_read(phydev, MII_88E1510_GEN_CTRL_REG_1); + temp &= ~MII_88E1510_GEN_CTRL_REG_1_MODE_MASK; + temp |= MII_88E1510_GEN_CTRL_REG_1_MODE_SGMII; + err = phy_write(phydev, MII_88E1510_GEN_CTRL_REG_1, temp); + if (err < 0) + return err; + + /* PHY reset is necessary after changing MODE[2:0] */ + temp |= MII_88E1510_GEN_CTRL_REG_1_RESET; + err = phy_write(phydev, MII_88E1510_GEN_CTRL_REG_1, temp); + if (err < 0) + return err; + + /* Reset page selection */ + err = phy_write(phydev, MII_MARVELL_PHY_PAGE, 0); + if (err < 0) + return err; + } + + return 0; + } + static int m88e1510_config_aneg(struct phy_device *phydev) { int err; @@@ -446,12 -486,6 +486,12 @@@ if (err < 0) return err; + return 0; +} + +static int marvell_config_init(struct phy_device *phydev) +{ + /* Set registers from marvell,reg-init DT property */ return marvell_of_reg_init(phydev); } @@@ -501,7 -535,7 +541,7 @@@ static int m88e1116r_config_init(struc mdelay(500); - return 0; + return marvell_config_init(phydev); } static int m88e3016_config_init(struct phy_device *phydev) @@@ -520,7 -554,7 +560,7 @@@ if (reg < 0) return reg; - return 0; + return marvell_config_init(phydev); } static int m88e1111_config_init(struct phy_device *phydev) @@@ -1031,8 -1065,8 +1071,8 @@@ static u64 marvell_get_stat(struct phy_ { struct marvell_hw_stat stat = marvell_hw_stats[i]; struct marvell_priv *priv = phydev->priv; - int err, oldpage; - u64 val; + int err, oldpage, val; + u64 ret; oldpage = phy_read(phydev, MII_MARVELL_PHY_PAGE); err = phy_write(phydev, MII_MARVELL_PHY_PAGE, @@@ -1042,16 -1076,16 +1082,16 @@@ val = phy_read(phydev, stat.reg); if (val < 0) { - val = UINT64_MAX; + ret = UINT64_MAX; } else { val = val & ((1 << stat.bits) - 1); priv->stats[i] += val; - val = priv->stats[i]; + ret = priv->stats[i]; } phy_write(phydev, MII_MARVELL_PHY_PAGE, oldpage); - return val; + return ret; } static void marvell_get_stats(struct phy_device *phydev, @@@ -1084,7 -1118,6 +1124,7 @@@ static struct phy_driver marvell_driver .features = PHY_GBIT_FEATURES, .probe = marvell_probe, .flags = PHY_HAS_INTERRUPT, + .config_init = &marvell_config_init, .config_aneg = &marvell_config_aneg, .read_status = &genphy_read_status, .ack_interrupt = &marvell_ack_interrupt, @@@ -1156,7 -1189,6 +1196,7 @@@ .features = PHY_GBIT_FEATURES, .flags = PHY_HAS_INTERRUPT, .probe = marvell_probe, + .config_init = &marvell_config_init, .config_aneg = &m88e1121_config_aneg, .read_status = &marvell_read_status, .ack_interrupt = &marvell_ack_interrupt, @@@ -1175,7 -1207,6 +1215,7 @@@ .features = PHY_GBIT_FEATURES, .flags = PHY_HAS_INTERRUPT, .probe = marvell_probe, + .config_init = &marvell_config_init, .config_aneg = &m88e1318_config_aneg, .read_status = &marvell_read_status, .ack_interrupt = &marvell_ack_interrupt, @@@ -1268,7 -1299,7 +1308,7 @@@ .features = PHY_GBIT_FEATURES, .flags = PHY_HAS_INTERRUPT, .probe = marvell_probe, - .config_init = &marvell_config_init, + .config_init = &m88e1510_config_init, .config_aneg = &m88e1510_config_aneg, .read_status = &marvell_read_status, .ack_interrupt = &marvell_ack_interrupt, @@@ -1287,7 -1318,6 +1327,7 @@@ .features = PHY_GBIT_FEATURES, .flags = PHY_HAS_INTERRUPT, .probe = marvell_probe, + .config_init = &marvell_config_init, .config_aneg = &m88e1510_config_aneg, .read_status = &marvell_read_status, .ack_interrupt = &marvell_ack_interrupt, diff --combined drivers/net/vxlan.c index e6944b2,c963897..766e611 --- a/drivers/net/vxlan.c +++ b/drivers/net/vxlan.c @@@ -42,7 -42,7 +42,7 @@@ #include <net/netns/generic.h> #include <net/vxlan.h> #include <net/protocol.h> - #include <net/udp_tunnel.h> + #if IS_ENABLED(CONFIG_IPV6) #include <net/ipv6.h> #include <net/addrconf.h> @@@ -197,9 -197,9 +197,9 @@@ static int vxlan_nla_put_addr(struct sk #endif /* Virtual Network hash table head */ - static inline struct hlist_head *vni_head(struct vxlan_sock *vs, u32 id) + static inline struct hlist_head *vni_head(struct vxlan_sock *vs, __be32 vni) { - return &vs->vni_list[hash_32(id, VNI_HASH_BITS)]; + return &vs->vni_list[hash_32((__force u32)vni, VNI_HASH_BITS)]; } /* Socket hash table head */ @@@ -242,12 -242,16 +242,16 @@@ static struct vxlan_sock *vxlan_find_so return NULL; } - static struct vxlan_dev *vxlan_vs_find_vni(struct vxlan_sock *vs, u32 id) + static struct vxlan_dev *vxlan_vs_find_vni(struct vxlan_sock *vs, __be32 vni) { struct vxlan_dev *vxlan; - hlist_for_each_entry_rcu(vxlan, vni_head(vs, id), hlist) { - if (vxlan->default_dst.remote_vni == id) + /* For flow based devices, map all packets to VNI 0 */ + if (vs->flags & VXLAN_F_COLLECT_METADATA) + vni = 0; + + hlist_for_each_entry_rcu(vxlan, vni_head(vs, vni), hlist) { + if (vxlan->default_dst.remote_vni == vni) return vxlan; } @@@ -255,7 -259,7 +259,7 @@@ } /* Look up VNI in a per net namespace table */ - static struct vxlan_dev *vxlan_find_vni(struct net *net, u32 id, + static struct vxlan_dev *vxlan_find_vni(struct net *net, __be32 vni, sa_family_t family, __be16 port, u32 flags) { @@@ -265,7 -269,7 +269,7 @@@ if (!vs) return NULL; - return vxlan_vs_find_vni(vs, id); + return vxlan_vs_find_vni(vs, vni); } /* Fill in neighbour message in skbuff. */ @@@ -315,7 -319,7 +319,7 @@@ static int vxlan_fdb_info(struct sk_buf nla_put_be16(skb, NDA_PORT, rdst->remote_port)) goto nla_put_failure; if (rdst->remote_vni != vxlan->default_dst.remote_vni && - nla_put_u32(skb, NDA_VNI, rdst->remote_vni)) + nla_put_u32(skb, NDA_VNI, be32_to_cpu(rdst->remote_vni))) goto nla_put_failure; if (rdst->remote_ifindex && nla_put_u32(skb, NDA_IFINDEX, rdst->remote_ifindex)) @@@ -383,7 -387,7 +387,7 @@@ static void vxlan_ip_miss(struct net_de }; struct vxlan_rdst remote = { .remote_ip = *ipa, /* goes to NDA_DST */ - .remote_vni = VXLAN_N_VID, + .remote_vni = cpu_to_be32(VXLAN_N_VID), }; vxlan_fdb_notify(vxlan, &f, &remote, RTM_GETNEIGH); @@@ -452,7 -456,7 +456,7 @@@ static struct vxlan_fdb *vxlan_find_mac /* caller should hold vxlan->hash_lock */ static struct vxlan_rdst *vxlan_fdb_find_rdst(struct vxlan_fdb *f, union vxlan_addr *ip, __be16 port, - __u32 vni, __u32 ifindex) + __be32 vni, __u32 ifindex) { struct vxlan_rdst *rd; @@@ -469,7 -473,8 +473,8 @@@ /* Replace destination of unicast mac */ static int vxlan_fdb_replace(struct vxlan_fdb *f, - union vxlan_addr *ip, __be16 port, __u32 vni, __u32 ifindex) + union vxlan_addr *ip, __be16 port, __be32 vni, + __u32 ifindex) { struct vxlan_rdst *rd; @@@ -480,6 -485,8 +485,8 @@@ rd = list_first_entry_or_null(&f->remotes, struct vxlan_rdst, list); if (!rd) return 0; + + dst_cache_reset(&rd->dst_cache); rd->remote_ip = *ip; rd->remote_port = port; rd->remote_vni = vni; @@@ -489,7 -496,7 +496,7 @@@ /* Add/update destinations for multicast */ static int vxlan_fdb_append(struct vxlan_fdb *f, - union vxlan_addr *ip, __be16 port, __u32 vni, + union vxlan_addr *ip, __be16 port, __be32 vni, __u32 ifindex, struct vxlan_rdst **rdp) { struct vxlan_rdst *rd; @@@ -501,6 -508,12 +508,12 @@@ rd = kmalloc(sizeof(*rd), GFP_ATOMIC); if (rd == NULL) return -ENOBUFS; + + if (dst_cache_init(&rd->dst_cache, GFP_ATOMIC)) { + kfree(rd); + return -ENOBUFS; + } + rd->remote_ip = *ip; rd->remote_port = port; rd->remote_vni = vni; @@@ -515,7 -528,8 +528,8 @@@ static struct vxlanhdr *vxlan_gro_remcsum(struct sk_buff *skb, unsigned int off, struct vxlanhdr *vh, size_t hdrlen, - u32 data, struct gro_remcsum *grc, + __be32 vni_field, + struct gro_remcsum *grc, bool nopartial) { size_t start, offset; @@@ -526,10 -540,8 +540,8 @@@ if (!NAPI_GRO_CB(skb)->csum_valid) return NULL; - start = (data & VXLAN_RCO_MASK) << VXLAN_RCO_SHIFT; - offset = start + ((data & VXLAN_RCO_UDP) ? - offsetof(struct udphdr, check) : - offsetof(struct tcphdr, check)); + start = vxlan_rco_start(vni_field); + offset = start + vxlan_rco_offset(vni_field); vh = skb_gro_remcsum_process(skb, (void *)vh, off, hdrlen, start, offset, grc, nopartial); @@@ -549,7 -561,7 +561,7 @@@ static struct sk_buff **vxlan_gro_recei int flush = 1; struct vxlan_sock *vs = container_of(uoff, struct vxlan_sock, udp_offloads); - u32 flags; + __be32 flags; struct gro_remcsum grc; skb_gro_remcsum_init(&grc); @@@ -565,11 -577,11 +577,11 @@@ skb_gro_postpull_rcsum(skb, vh, sizeof(struct vxlanhdr)); - flags = ntohl(vh->vx_flags); + flags = vh->vx_flags; if ((flags & VXLAN_HF_RCO) && (vs->flags & VXLAN_F_REMCSUM_RX)) { vh = vxlan_gro_remcsum(skb, off_vx, vh, sizeof(struct vxlanhdr), - ntohl(vh->vx_vni), &grc, + vh->vx_vni, &grc, !!(vs->flags & VXLAN_F_REMCSUM_NOPARTIAL)); @@@ -660,7 -672,7 +672,7 @@@ static void vxlan_notify_del_rx_port(st static int vxlan_fdb_create(struct vxlan_dev *vxlan, const u8 *mac, union vxlan_addr *ip, __u16 state, __u16 flags, - __be16 port, __u32 vni, __u32 ifindex, + __be16 port, __be32 vni, __u32 ifindex, __u8 ndm_flags) { struct vxlan_rdst *rd = NULL; @@@ -749,8 -761,10 +761,10 @@@ static void vxlan_fdb_free(struct rcu_h struct vxlan_fdb *f = container_of(head, struct vxlan_fdb, rcu); struct vxlan_rdst *rd, *nd; - list_for_each_entry_safe(rd, nd, &f->remotes, list) + list_for_each_entry_safe(rd, nd, &f->remotes, list) { + dst_cache_destroy(&rd->dst_cache); kfree(rd); + } kfree(f); } @@@ -767,7 -781,8 +781,8 @@@ static void vxlan_fdb_destroy(struct vx } static int vxlan_fdb_parse(struct nlattr *tb[], struct vxlan_dev *vxlan, - union vxlan_addr *ip, __be16 *port, u32 *vni, u32 *ifindex) + union vxlan_addr *ip, __be16 *port, __be32 *vni, + u32 *ifindex) { struct net *net = dev_net(vxlan->dev); int err; @@@ -800,7 -815,7 +815,7 @@@ if (tb[NDA_VNI]) { if (nla_len(tb[NDA_VNI]) != sizeof(u32)) return -EINVAL; - *vni = nla_get_u32(tb[NDA_VNI]); + *vni = cpu_to_be32(nla_get_u32(tb[NDA_VNI])); } else { *vni = vxlan->default_dst.remote_vni; } @@@ -830,7 -845,8 +845,8 @@@ static int vxlan_fdb_add(struct ndmsg * /* struct net *net = dev_net(vxlan->dev); */ union vxlan_addr ip; __be16 port; - u32 vni, ifindex; + __be32 vni; + u32 ifindex; int err; if (!(ndm->ndm_state & (NUD_PERMANENT|NUD_REACHABLE))) { @@@ -867,7 -883,8 +883,8 @@@ static int vxlan_fdb_delete(struct ndms struct vxlan_rdst *rd = NULL; union vxlan_addr ip; __be16 port; - u32 vni, ifindex; + __be32 vni; + u32 ifindex; int err; err = vxlan_fdb_parse(tb, vxlan, &ip, &port, &vni, &ifindex); @@@ -1122,54 -1139,65 +1139,65 @@@ static int vxlan_igmp_leave(struct vxla return ret; } - static struct vxlanhdr *vxlan_remcsum(struct sk_buff *skb, struct vxlanhdr *vh, - size_t hdrlen, u32 data, bool nopartial) + static bool vxlan_remcsum(struct vxlanhdr *unparsed, + struct sk_buff *skb, u32 vxflags) { size_t start, offset, plen; - if (skb->remcsum_offload) - return vh; + if (!(unparsed->vx_flags & VXLAN_HF_RCO) || skb->remcsum_offload) + goto out; - start = (data & VXLAN_RCO_MASK) << VXLAN_RCO_SHIFT; - offset = start + ((data & VXLAN_RCO_UDP) ? - offsetof(struct udphdr, check) : - offsetof(struct tcphdr, check)); + start = vxlan_rco_start(unparsed->vx_vni); + offset = start + vxlan_rco_offset(unparsed->vx_vni); - plen = hdrlen + offset + sizeof(u16); + plen = sizeof(struct vxlanhdr) + offset + sizeof(u16); if (!pskb_may_pull(skb, plen)) - return NULL; + return false; - vh = (struct vxlanhdr *)(udp_hdr(skb) + 1); + skb_remcsum_process(skb, (void *)(vxlan_hdr(skb) + 1), start, offset, + !!(vxflags & VXLAN_F_REMCSUM_NOPARTIAL)); + out: + unparsed->vx_flags &= ~VXLAN_HF_RCO; + unparsed->vx_vni &= VXLAN_VNI_MASK; + return true; + } - skb_remcsum_process(skb, (void *)vh + hdrlen, start, offset, - nopartial); + static void vxlan_parse_gbp_hdr(struct vxlanhdr *unparsed, + struct vxlan_metadata *md, + struct metadata_dst *tun_dst) + { + struct vxlanhdr_gbp *gbp = (struct vxlanhdr_gbp *)unparsed; - return vh; + if (!(unparsed->vx_flags & VXLAN_HF_GBP)) + goto out; + + md->gbp = ntohs(gbp->policy_id); + + if (tun_dst) + tun_dst->u.tun_info.key.tun_flags |= TUNNEL_VXLAN_OPT; + + if (gbp->dont_learn) + md->gbp |= VXLAN_GBP_DONT_LEARN; + + if (gbp->policy_applied) + md->gbp |= VXLAN_GBP_POLICY_APPLIED; + + out: + unparsed->vx_flags &= ~VXLAN_GBP_USED_BITS; } - static void vxlan_rcv(struct vxlan_sock *vs, struct sk_buff *skb, - struct vxlan_metadata *md, u32 vni, + static void vxlan_rcv(struct vxlan_dev *vxlan, struct vxlan_sock *vs, + struct sk_buff *skb, struct vxlan_metadata *md, struct metadata_dst *tun_dst) { struct iphdr *oip = NULL; struct ipv6hdr *oip6 = NULL; - struct vxlan_dev *vxlan; struct pcpu_sw_netstats *stats; union vxlan_addr saddr; int err = 0; - /* For flow based devices, map all packets to VNI 0 */ - if (vs->flags & VXLAN_F_COLLECT_METADATA) - vni = 0; - - /* Is this VNI defined? */ - vxlan = vxlan_vs_find_vni(vs, vni); - if (!vxlan) - goto drop; - skb_reset_mac_header(skb); - skb_scrub_packet(skb, !net_eq(vxlan->net, dev_net(vxlan->dev))); skb->protocol = eth_type_trans(skb, vxlan->dev); skb_postpull_rcsum(skb, eth_hdr(skb), ETH_HLEN); @@@ -1246,48 -1274,45 +1274,45 @@@ drop static int vxlan_udp_encap_recv(struct sock *sk, struct sk_buff *skb) { struct metadata_dst *tun_dst = NULL; + struct vxlan_dev *vxlan; struct vxlan_sock *vs; - struct vxlanhdr *vxh; - u32 flags, vni; + struct vxlanhdr unparsed; struct vxlan_metadata _md; struct vxlan_metadata *md = &_md; /* Need Vxlan and inner Ethernet header to be present */ if (!pskb_may_pull(skb, VXLAN_HLEN)) - goto error; - - vxh = (struct vxlanhdr *)(udp_hdr(skb) + 1); - flags = ntohl(vxh->vx_flags); - vni = ntohl(vxh->vx_vni); + return 1; - if (flags & VXLAN_HF_VNI) { - flags &= ~VXLAN_HF_VNI; - } else { - /* VNI flag always required to be set */ - goto bad_flags; + unparsed = *vxlan_hdr(skb); + /* VNI flag always required to be set */ + if (!(unparsed.vx_flags & VXLAN_HF_VNI)) { + netdev_dbg(skb->dev, "invalid vxlan flags=%#x vni=%#x\n", + ntohl(vxlan_hdr(skb)->vx_flags), + ntohl(vxlan_hdr(skb)->vx_vni)); + /* Return non vxlan pkt */ + return 1; } - - if (iptunnel_pull_header(skb, VXLAN_HLEN, htons(ETH_P_TEB))) - goto drop; - vxh = (struct vxlanhdr *)(udp_hdr(skb) + 1); + unparsed.vx_flags &= ~VXLAN_HF_VNI; + unparsed.vx_vni &= ~VXLAN_VNI_MASK; vs = rcu_dereference_sk_user_data(sk); if (!vs) goto drop; - if ((flags & VXLAN_HF_RCO) && (vs->flags & VXLAN_F_REMCSUM_RX)) { - vxh = vxlan_remcsum(skb, vxh, sizeof(struct vxlanhdr), vni, - !!(vs->flags & VXLAN_F_REMCSUM_NOPARTIAL)); - if (!vxh) - goto drop; + vxlan = vxlan_vs_find_vni(vs, vxlan_vni(vxlan_hdr(skb)->vx_vni)); + if (!vxlan) + goto drop; - flags &= ~VXLAN_HF_RCO; - vni &= VXLAN_VNI_MASK; - } + if (iptunnel_pull_header(skb, VXLAN_HLEN, htons(ETH_P_TEB), + !net_eq(vxlan->net, dev_net(vxlan->dev)))) + goto drop; if (vxlan_collect_metadata(vs)) { + __be32 vni = vxlan_vni(vxlan_hdr(skb)->vx_vni); + tun_dst = udp_tun_rx_dst(skb, vxlan_get_sk_family(vs), TUNNEL_KEY, - cpu_to_be64(vni >> 8), sizeof(*md)); + vxlan_vni_to_tun_id(vni), sizeof(*md)); if (!tun_dst) goto drop; @@@ -1300,25 -1325,13 +1325,13 @@@ /* For backwards compatibility, only allow reserved fields to be * used by VXLAN extensions if explicitly requested. */ - if ((flags & VXLAN_HF_GBP) && (vs->flags & VXLAN_F_GBP)) { - struct vxlanhdr_gbp *gbp; - - gbp = (struct vxlanhdr_gbp *)vxh; - md->gbp = ntohs(gbp->policy_id); - - if (tun_dst) - tun_dst->u.tun_info.key.tun_flags |= TUNNEL_VXLAN_OPT; - - if (gbp->dont_learn) - md->gbp |= VXLAN_GBP_DONT_LEARN; - - if (gbp->policy_applied) - md->gbp |= VXLAN_GBP_POLICY_APPLIED; - - flags &= ~VXLAN_GBP_USED_BITS; - } + if (vs->flags & VXLAN_F_REMCSUM_RX) + if (!vxlan_remcsum(&unparsed, skb, vs->flags)) + goto drop; + if (vs->flags & VXLAN_F_GBP) + vxlan_parse_gbp_hdr(&unparsed, md, tun_dst); - if (flags || vni & ~VXLAN_VNI_MASK) { + if (unparsed.vx_flags || unparsed.vx_vni) { /* If there are any unprocessed flags remaining treat * this as a malformed packet. This behavior diverges from * VXLAN RFC (RFC7348) which stipulates that bits in reserved @@@ -1327,28 -1340,19 +1340,19 @@@ * is more robust and provides a little more security in * adding extensions to VXLAN. */ - - goto bad_flags; + goto drop; } - vxlan_rcv(vs, skb, md, vni >> 8, tun_dst); + vxlan_rcv(vxlan, vs, skb, md, tun_dst); return 0; drop: - /* Consume bad packet */ - kfree_skb(skb); - return 0; - - bad_flags: - netdev_dbg(skb->dev, "invalid vxlan flags=%#x vni=%#x\n", - ntohl(vxh->vx_flags), ntohl(vxh->vx_vni)); - - error: if (tun_dst) dst_release((struct dst_entry *)tun_dst); - /* Return non vxlan pkt */ - return 1; + /* Consume bad packet */ + kfree_skb(skb); + return 0; } static int arp_reduce(struct net_device *dev, struct sk_buff *skb) @@@ -1673,7 -1677,7 +1677,7 @@@ static void vxlan_build_gbp_hdr(struct return; gbp = (struct vxlanhdr_gbp *)vxh; - vxh->vx_flags |= htonl(VXLAN_HF_GBP); + vxh->vx_flags |= VXLAN_HF_GBP; if (md->gbp & VXLAN_GBP_DONT_LEARN) gbp->dont_learn = 1; @@@ -1684,20 -1688,15 +1688,15 @@@ gbp->policy_id = htons(md->gbp & VXLAN_GBP_ID_MASK); } - #if IS_ENABLED(CONFIG_IPV6) - static int vxlan6_xmit_skb(struct dst_entry *dst, struct sock *sk, - struct sk_buff *skb, - struct net_device *dev, struct in6_addr *saddr, - struct in6_addr *daddr, __u8 prio, __u8 ttl, - __be16 src_port, __be16 dst_port, __be32 vni, - struct vxlan_metadata *md, bool xnet, u32 vxflags) + static int vxlan_build_skb(struct sk_buff *skb, struct dst_entry *dst, + int iphdr_len, __be32 vni, + struct vxlan_metadata *md, u32 vxflags, + bool udp_sum) { struct vxlanhdr *vxh; int min_headroom; int err; - bool udp_sum = !(vxflags & VXLAN_F_UDP_ZERO_CSUM6_TX); int type = udp_sum ? SKB_GSO_UDP_TUNNEL_CSUM : SKB_GSO_UDP_TUNNEL; - u16 hdrlen = sizeof(struct vxlanhdr); if ((vxflags & VXLAN_F_REMCSUM_TX) && skb->ip_summed == CHECKSUM_PARTIAL) { @@@ -1706,50 -1705,39 +1705,39 @@@ if (csum_start <= VXLAN_MAX_REMCSUM_START && !(csum_start & VXLAN_RCO_SHIFT_MASK) && (skb->csum_offset == offsetof(struct udphdr, check) || - skb->csum_offset == offsetof(struct tcphdr, check))) { - udp_sum = false; + skb->csum_offset == offsetof(struct tcphdr, check))) type |= SKB_GSO_TUNNEL_REMCSUM; - } } - skb_scrub_packet(skb, xnet); - min_headroom = LL_RESERVED_SPACE(dst->dev) + dst->header_len - + VXLAN_HLEN + sizeof(struct ipv6hdr) + + VXLAN_HLEN + iphdr_len + (skb_vlan_tag_present(skb) ? VLAN_HLEN : 0); /* Need space for new headers (invalidates iph ptr) */ err = skb_cow_head(skb, min_headroom); if (unlikely(err)) { kfree_skb(skb); - goto err; + return err; } skb = vlan_hwaccel_push_inside(skb); - if (WARN_ON(!skb)) { - err = -ENOMEM; - goto err; - } + if (WARN_ON(!skb)) + return -ENOMEM; - skb = iptunnel_handle_offloads(skb, udp_sum, type); - if (IS_ERR(skb)) { - err = -EINVAL; - goto err; - } + skb = iptunnel_handle_offloads(skb, type); + if (IS_ERR(skb)) + return PTR_ERR(skb); vxh = (struct vxlanhdr *) __skb_push(skb, sizeof(*vxh)); - vxh->vx_flags = htonl(VXLAN_HF_VNI); - vxh->vx_vni = vni; + vxh->vx_flags = VXLAN_HF_VNI; + vxh->vx_vni = vxlan_vni_field(vni); if (type & SKB_GSO_TUNNEL_REMCSUM) { - u32 data = (skb_checksum_start_offset(skb) - hdrlen) >> - VXLAN_RCO_SHIFT; + unsigned int start; - if (skb->csum_offset == offsetof(struct udphdr, check)) - data |= VXLAN_RCO_UDP; - - vxh->vx_vni |= htonl(data); - vxh->vx_flags |= htonl(VXLAN_HF_RCO); + start = skb_checksum_start_offset(skb) - sizeof(struct vxlanhdr); + vxh->vx_vni |= vxlan_compute_rco(start, skb->csum_offset); + vxh->vx_flags |= VXLAN_HF_RCO; if (!skb_is_gso(skb)) { skb->ip_summed = CHECKSUM_NONE; @@@ -1761,102 -1749,63 +1749,63 @@@ vxlan_build_gbp_hdr(vxh, vxflags, md); skb_set_inner_protocol(skb, htons(ETH_P_TEB)); - - udp_tunnel6_xmit_skb(dst, sk, skb, dev, saddr, daddr, prio, - ttl, src_port, dst_port, - !!(vxflags & VXLAN_F_UDP_ZERO_CSUM6_TX)); return 0; - err: - dst_release(dst); - return err; } - #endif - static int vxlan_xmit_skb(struct rtable *rt, struct sock *sk, struct sk_buff *skb, - __be32 src, __be32 dst, __u8 tos, __u8 ttl, __be16 df, - __be16 src_port, __be16 dst_port, __be32 vni, - struct vxlan_metadata *md, bool xnet, u32 vxflags) + static struct rtable *vxlan_get_route(struct vxlan_dev *vxlan, + struct sk_buff *skb, int oif, u8 tos, + __be32 daddr, __be32 *saddr, + struct dst_cache *dst_cache, + struct ip_tunnel_info *info) { - struct vxlanhdr *vxh; - int min_headroom; - int err; - bool udp_sum = !!(vxflags & VXLAN_F_UDP_CSUM); - int type = udp_sum ? SKB_GSO_UDP_TUNNEL_CSUM : SKB_GSO_UDP_TUNNEL; - u16 hdrlen = sizeof(struct vxlanhdr); - - if ((vxflags & VXLAN_F_REMCSUM_TX) && - skb->ip_summed == CHECKSUM_PARTIAL) { - int csum_start = skb_checksum_start_offset(skb); - - if (csum_start <= VXLAN_MAX_REMCSUM_START && - !(csum_start & VXLAN_RCO_SHIFT_MASK) && - (skb->csum_offset == offsetof(struct udphdr, check) || - skb->csum_offset == offsetof(struct tcphdr, check))) { - udp_sum = false; - type |= SKB_GSO_TUNNEL_REMCSUM; - } - } - - min_headroom = LL_RESERVED_SPACE(rt->dst.dev) + rt->dst.header_len - + VXLAN_HLEN + sizeof(struct iphdr) - + (skb_vlan_tag_present(skb) ? VLAN_HLEN : 0); + struct rtable *rt = NULL; + bool use_cache = false; + struct flowi4 fl4; - /* Need space for new headers (invalidates iph ptr) */ - err = skb_cow_head(skb, min_headroom); - if (unlikely(err)) { - kfree_skb(skb); - return err; + /* when the ip_tunnel_info is availble, the tos used for lookup is + * packet independent, so we can use the cache + */ + if (!skb->mark && (!tos || info)) { + use_cache = true; + rt = dst_cache_get_ip4(dst_cache, saddr); + if (rt) + return rt; } - skb = vlan_hwaccel_push_inside(skb); - if (WARN_ON(!skb)) - return -ENOMEM; - - skb = iptunnel_handle_offloads(skb, udp_sum, type); - if (IS_ERR(skb)) - return PTR_ERR(skb); - - vxh = (struct vxlanhdr *) __skb_push(skb, sizeof(*vxh)); - vxh->vx_flags = htonl(VXLAN_HF_VNI); - vxh->vx_vni = vni; - - if (type & SKB_GSO_TUNNEL_REMCSUM) { - u32 data = (skb_checksum_start_offset(skb) - hdrlen) >> - VXLAN_RCO_SHIFT; - - if (skb->csum_offset == offsetof(struct udphdr, check)) - data |= VXLAN_RCO_UDP; - - vxh->vx_vni |= htonl(data); - vxh->vx_flags |= htonl(VXLAN_HF_RCO); + memset(&fl4, 0, sizeof(fl4)); + fl4.flowi4_oif = oif; + fl4.flowi4_tos = RT_TOS(tos); + fl4.flowi4_mark = skb->mark; + fl4.flowi4_proto = IPPROTO_UDP; + fl4.daddr = daddr; + fl4.saddr = vxlan->cfg.saddr.sin.sin_addr.s_addr; - if (!skb_is_gso(skb)) { - skb->ip_summed = CHECKSUM_NONE; - skb->encapsulation = 0; - } + rt = ip_route_output_key(vxlan->net, &fl4); + if (!IS_ERR(rt)) { + *saddr = fl4.saddr; + if (use_cache) + dst_cache_set_ip4(dst_cache, &rt->dst, fl4.saddr); } - - if (vxflags & VXLAN_F_GBP) - vxlan_build_gbp_hdr(vxh, vxflags, md); - - skb_set_inner_protocol(skb, htons(ETH_P_TEB)); - - udp_tunnel_xmit_skb(rt, sk, skb, src, dst, tos, ttl, df, - src_port, dst_port, xnet, - !(vxflags & VXLAN_F_UDP_CSUM)); - return 0; + return rt; } #if IS_ENABLED(CONFIG_IPV6) static struct dst_entry *vxlan6_get_route(struct vxlan_dev *vxlan, struct sk_buff *skb, int oif, const struct in6_addr *daddr, - struct in6_addr *saddr) + struct in6_addr *saddr, + struct dst_cache *dst_cache) { struct dst_entry *ndst; struct flowi6 fl6; int err; + if (!skb->mark) { + ndst = dst_cache_get_ip6(dst_cache, saddr); + if (ndst) + return ndst; + } + memset(&fl6, 0, sizeof(fl6)); fl6.flowi6_oif = oif; fl6.daddr = *daddr; @@@ -1871,6 -1820,8 +1820,8 @@@ return ERR_PTR(err); *saddr = fl6.saddr; + if (!skb->mark) + dst_cache_set_ip6(dst_cache, ndst, saddr); return ndst; } #endif @@@ -1923,22 -1874,24 +1874,24 @@@ static void vxlan_encap_bypass(struct s static void vxlan_xmit_one(struct sk_buff *skb, struct net_device *dev, struct vxlan_rdst *rdst, bool did_rsc) { + struct dst_cache *dst_cache; struct ip_tunnel_info *info; struct vxlan_dev *vxlan = netdev_priv(dev); struct sock *sk; struct rtable *rt = NULL; const struct iphdr *old_iph; - struct flowi4 fl4; union vxlan_addr *dst; union vxlan_addr remote_ip; struct vxlan_metadata _md; struct vxlan_metadata *md = &_md; __be16 src_port = 0, dst_port; - u32 vni; + __be32 vni; __be16 df = 0; __u8 tos, ttl; int err; u32 flags = vxlan->flags; + bool udp_sum = false; + bool xnet = !net_eq(vxlan->net, dev_net(vxlan->dev)); info = skb_tunnel_info(skb); @@@ -1946,6 -1899,7 +1899,7 @@@ dst_port = rdst->remote_port ? rdst->remote_port : vxlan->cfg.dst_port; vni = rdst->remote_vni; dst = &rdst->remote_ip; + dst_cache = &rdst->dst_cache; } else { if (!info) { WARN_ONCE(1, "%s: Missing encapsulation instructions\n", @@@ -1953,13 -1907,14 +1907,14 @@@ goto drop; } dst_port = info->key.tp_dst ? : vxlan->cfg.dst_port; - vni = be64_to_cpu(info->key.tun_id); + vni = vxlan_tun_id_to_vni(info->key.tun_id); remote_ip.sa.sa_family = ip_tunnel_info_af(info); if (remote_ip.sa.sa_family == AF_INET) remote_ip.sin.sin_addr.s_addr = info->key.u.ipv4.dst; else remote_ip.sin6.sin6_addr = info->key.u.ipv6.dst; dst = &remote_ip; + dst_cache = &info->dst_cache; } if (vxlan_addr_any(dst)) { @@@ -1987,6 -1942,7 +1942,7 @@@ if (info) { ttl = info->key.ttl; tos = info->key.tos; + udp_sum = !!(info->key.tun_flags & TUNNEL_CSUM); if (info->options_len) md = ip_tunnel_info_opts(info); @@@ -1995,6 -1951,8 +1951,8 @@@ } if (dst->sa.sa_family == AF_INET) { + __be32 saddr; + if (!vxlan->vn4_sock) goto drop; sk = vxlan->vn4_sock->sock->sk; @@@ -2002,22 -1960,14 +1960,14 @@@ if (info) { if (info->key.tun_flags & TUNNEL_DONT_FRAGMENT) df = htons(IP_DF); - - if (info->key.tun_flags & TUNNEL_CSUM) - flags |= VXLAN_F_UDP_CSUM; - else - flags &= ~VXLAN_F_UDP_CSUM; + } else { + udp_sum = !!(flags & VXLAN_F_UDP_CSUM); } - memset(&fl4, 0, sizeof(fl4)); - fl4.flowi4_oif = rdst ? rdst->remote_ifindex : 0; - fl4.flowi4_tos = RT_TOS(tos); - fl4.flowi4_mark = skb->mark; - fl4.flowi4_proto = IPPROTO_UDP; - fl4.daddr = dst->sin.sin_addr.s_addr; - fl4.saddr = vxlan->cfg.saddr.sin.sin_addr.s_addr; - - rt = ip_route_output_key(vxlan->net, &fl4); + rt = vxlan_get_route(vxlan, skb, + rdst ? rdst->remote_ifindex : 0, tos, + dst->sin.sin_addr.s_addr, &saddr, + dst_cache, info); if (IS_ERR(rt)) { netdev_dbg(dev, "no route to %pI4\n", &dst->sin.sin_addr.s_addr); @@@ -2049,16 -1999,14 +1999,14 @@@ tos = ip_tunnel_ecn_encap(tos, old_iph, skb); ttl = ttl ? : ip4_dst_hoplimit(&rt->dst); - err = vxlan_xmit_skb(rt, sk, skb, fl4.saddr, - dst->sin.sin_addr.s_addr, tos, ttl, df, - src_port, dst_port, htonl(vni << 8), md, - !net_eq(vxlan->net, dev_net(vxlan->dev)), - flags); - if (err < 0) { - /* skb is already freed. */ - skb = NULL; - goto rt_tx_error; - } + err = vxlan_build_skb(skb, &rt->dst, sizeof(struct iphdr), + vni, md, flags, udp_sum); + if (err < 0) + goto xmit_tx_error; + + udp_tunnel_xmit_skb(rt, sk, skb, saddr, + dst->sin.sin_addr.s_addr, tos, ttl, df, + src_port, dst_port, xnet, !udp_sum); #if IS_ENABLED(CONFIG_IPV6) } else { struct dst_entry *ndst; @@@ -2071,7 -2019,8 +2019,8 @@@ ndst = vxlan6_get_route(vxlan, skb, rdst ? rdst->remote_ifindex : 0, - &dst->sin6.sin6_addr, &saddr); + &dst->sin6.sin6_addr, &saddr, + dst_cache); if (IS_ERR(ndst)) { netdev_dbg(dev, "no route to %pI6\n", &dst->sin6.sin6_addr); @@@ -2103,18 -2052,20 +2052,20 @@@ return; } - if (info) { - if (info->key.tun_flags & TUNNEL_CSUM) - flags &= ~VXLAN_F_UDP_ZERO_CSUM6_TX; - else - flags |= VXLAN_F_UDP_ZERO_CSUM6_TX; - } + if (!info) + udp_sum = !(flags & VXLAN_F_UDP_ZERO_CSUM6_TX); ttl = ttl ? : ip6_dst_hoplimit(ndst); - err = vxlan6_xmit_skb(ndst, sk, skb, dev, &saddr, &dst->sin6.sin6_addr, - 0, ttl, src_port, dst_port, htonl(vni << 8), md, - !net_eq(vxlan->net, dev_net(vxlan->dev)), - flags); + skb_scrub_packet(skb, xnet); + err = vxlan_build_skb(skb, ndst, sizeof(struct ipv6hdr), + vni, md, flags, udp_sum); + if (err < 0) { + dst_release(ndst); + return; + } + udp_tunnel6_xmit_skb(ndst, sk, skb, dev, + &saddr, &dst->sin6.sin6_addr, + 0, ttl, src_port, dst_port, !udp_sum); #endif } @@@ -2124,6 -2075,9 +2075,9 @@@ drop dev->stats.tx_dropped++; goto tx_free; + xmit_tx_error: + /* skb is already freed. */ + skb = NULL; rt_tx_error: ip_rt_put(rt); tx_error: @@@ -2171,11 -2125,9 +2125,11 @@@ static netdev_tx_t vxlan_xmit(struct sk #endif } - if (vxlan->flags & VXLAN_F_COLLECT_METADATA && - info && info->mode & IP_TUNNEL_INFO_TX) { - vxlan_xmit_one(skb, dev, NULL, false); + if (vxlan->flags & VXLAN_F_COLLECT_METADATA) { + if (info && info->mode & IP_TUNNEL_INFO_TX) + vxlan_xmit_one(skb, dev, NULL, false); + else + kfree_skb(skb); return NETDEV_TX_OK; } @@@ -2263,7 -2215,7 +2217,7 @@@ static void vxlan_cleanup(unsigned lon static void vxlan_vs_add_dev(struct vxlan_sock *vs, struct vxlan_dev *vxlan) { struct vxlan_net *vn = net_generic(vxlan->net, vxlan_net_id); - __u32 vni = vxlan->default_dst.remote_vni; + __be32 vni = vxlan->default_dst.remote_vni; spin_lock(&vn->sock_lock); hlist_add_head_rcu(&vxlan->hlist, vni_head(vs, vni)); @@@ -2369,68 -2321,29 +2323,43 @@@ static void vxlan_set_multicast_list(st { } -static int vxlan_change_mtu(struct net_device *dev, int new_mtu) +static int __vxlan_change_mtu(struct net_device *dev, + struct net_device *lowerdev, + struct vxlan_rdst *dst, int new_mtu, bool strict) { - struct vxlan_dev *vxlan = netdev_priv(dev); - struct vxlan_rdst *dst = &vxlan->default_dst; - struct net_device *lowerdev; - int max_mtu; + int max_mtu = IP_MAX_MTU; - lowerdev = __dev_get_by_index(vxlan->net, dst->remote_ifindex); - if (lowerdev == NULL) - return eth_change_mtu(dev, new_mtu); + if (lowerdev) + max_mtu = lowerdev->mtu; if (dst->remote_ip.sa.sa_family == AF_INET6) - max_mtu = lowerdev->mtu - VXLAN6_HEADROOM; + max_mtu -= VXLAN6_HEADROOM; else - max_mtu = lowerdev->mtu - VXLAN_HEADROOM; + max_mtu -= VXLAN_HEADROOM; - if (new_mtu < 68 || new_mtu > max_mtu) + if (new_mtu < 68) return -EINVAL; + if (new_mtu > max_mtu) { + if (strict) + return -EINVAL; + + new_mtu = max_mtu; + } + dev->mtu = new_mtu; return 0; } +static int vxlan_change_mtu(struct net_device *dev, int new_mtu) +{ + struct vxlan_dev *vxlan = netdev_priv(dev); + struct vxlan_rdst *dst = &vxlan->default_dst; + struct net_device *lowerdev = __dev_get_by_index(vxlan->net, + dst->remote_ifindex); + return __vxlan_change_mtu(dev, lowerdev, dst, new_mtu, true); +} + - static int egress_ipv4_tun_info(struct net_device *dev, struct sk_buff *skb, - struct ip_tunnel_info *info, - __be16 sport, __be16 dport) - { - struct vxlan_dev *vxlan = netdev_priv(dev); - struct rtable *rt; - struct flowi4 fl4; - - memset(&fl4, 0, sizeof(fl4)); - fl4.flowi4_tos = RT_TOS(info->key.tos); - fl4.flowi4_mark = skb->mark; - fl4.flowi4_proto = IPPROTO_UDP; - fl4.daddr = info->key.u.ipv4.dst; - - rt = ip_route_output_key(vxlan->net, &fl4); - if (IS_ERR(rt)) - return PTR_ERR(rt); - ip_rt_put(rt); - - info->key.u.ipv4.src = fl4.saddr; - info->key.tp_src = sport; - info->key.tp_dst = dport; - return 0; - } - static int vxlan_fill_metadata_dst(struct net_device *dev, struct sk_buff *skb) { struct vxlan_dev *vxlan = netdev_priv(dev); @@@ -2442,9 -2355,16 +2371,16 @@@ dport = info->key.tp_dst ? : vxlan->cfg.dst_port; if (ip_tunnel_info_af(info) == AF_INET) { + struct rtable *rt; + if (!vxlan->vn4_sock) return -EINVAL; - return egress_ipv4_tun_info(dev, skb, info, sport, dport); + rt = vxlan_get_route(vxlan, skb, 0, info->key.tos, + info->key.u.ipv4.dst, + &info->key.u.ipv4.src, NULL, info); + if (IS_ERR(rt)) + return PTR_ERR(rt); + ip_rt_put(rt); } else { #if IS_ENABLED(CONFIG_IPV6) struct dst_entry *ndst; @@@ -2453,17 -2373,16 +2389,16 @@@ return -EINVAL; ndst = vxlan6_get_route(vxlan, skb, 0, &info->key.u.ipv6.dst, - &info->key.u.ipv6.src); + &info->key.u.ipv6.src, NULL); if (IS_ERR(ndst)) return PTR_ERR(ndst); dst_release(ndst); - - info->key.tp_src = sport; - info->key.tp_dst = dport; #else /* !CONFIG_IPV6 */ return -EPFNOSUPPORT; #endif } + info->key.tp_src = sport; + info->key.tp_dst = dport; return 0; } @@@ -2539,7 -2458,6 +2474,7 @@@ static void vxlan_setup(struct net_devi dev->hw_features |= NETIF_F_GSO_SOFTWARE; dev->hw_features |= NETIF_F_HW_VLAN_CTAG_TX | NETIF_F_HW_VLAN_STAG_TX; netif_keep_dst(dev); + dev->priv_flags &= ~IFF_TX_SKB_SHARING; dev->priv_flags |= IFF_LIVE_ADDR_CHANGE | IFF_NO_QUEUE; INIT_LIST_HEAD(&vxlan->next); @@@ -2782,7 -2700,6 +2717,7 @@@ static int vxlan_dev_configure(struct n int err; bool use_ipv6 = false; __be16 default_port = vxlan->cfg.dst_port; + struct net_device *lowerdev = NULL; vxlan->net = src_net; @@@ -2803,7 -2720,9 +2738,7 @@@ } if (conf->remote_ifindex) { - struct net_device *lowerdev - = __dev_get_by_index(src_net, conf->remote_ifindex); - + lowerdev = __dev_get_by_index(src_net, conf->remote_ifindex); dst->remote_ifindex = conf->remote_ifindex; if (!lowerdev) { @@@ -2827,12 -2746,6 +2762,12 @@@ needed_headroom = lowerdev->hard_header_len; } + if (conf->mtu) { + err = __vxlan_change_mtu(dev, lowerdev, dst, conf->mtu, false); + if (err) + return err; + } + if (use_ipv6 || conf->flags & VXLAN_F_COLLECT_METADATA) needed_headroom += VXLAN6_HEADROOM; else @@@ -2917,7 -2830,7 +2852,7 @@@ static int vxlan_newlink(struct net *sr memset(&conf, 0, sizeof(conf)); if (data[IFLA_VXLAN_ID]) - conf.vni = nla_get_u32(data[IFLA_VXLAN_ID]); + conf.vni = cpu_to_be32(nla_get_u32(data[IFLA_VXLAN_ID])); if (data[IFLA_VXLAN_GROUP]) { conf.remote_ip.sin.sin_addr.s_addr = nla_get_in_addr(data[IFLA_VXLAN_GROUP]); @@@ -3021,7 -2934,7 +2956,7 @@@ break; case -EEXIST: - pr_info("duplicate VNI %u\n", conf.vni); + pr_info("duplicate VNI %u\n", be32_to_cpu(conf.vni)); break; } @@@ -3079,7 -2992,7 +3014,7 @@@ static int vxlan_fill_info(struct sk_bu .high = htons(vxlan->cfg.port_max), }; - if (nla_put_u32(skb, IFLA_VXLAN_ID, dst->remote_vni)) + if (nla_put_u32(skb, IFLA_VXLAN_ID, be32_to_cpu(dst->remote_vni))) goto nla_put_failure; if (!vxlan_addr_any(&dst->remote_ip)) { diff --combined drivers/net/wireless/intel/iwlwifi/Kconfig index 7438fbe,11932d5..16c4f38 --- a/drivers/net/wireless/intel/iwlwifi/Kconfig +++ b/drivers/net/wireless/intel/iwlwifi/Kconfig @@@ -53,6 -53,7 +53,6 @@@ config IWLWIFI_LED config IWLDVM tristate "Intel Wireless WiFi DVM Firmware support" - depends on m help This is the driver that supports the DVM firmware. The list of the devices that use this firmware is available here: @@@ -98,6 -99,18 +98,18 @@@ config IWLWIFI_UAPS If unsure, say N. + config IWLWIFI_PCIE_RTPM + bool "Enable runtime power management mode for PCIe devices" + depends on IWLMVM && PM + default false + help + Say Y here to enable runtime power management for PCIe + devices. If enabled, the device will go into low power mode + when idle for a short period of time, allowing for improved + power saving during runtime. + + If unsure, say N. + menu "Debugging Options" config IWLWIFI_DEBUG diff --combined drivers/net/wireless/intel/iwlwifi/mvm/scan.c index ea1e177,1e1ab9d..aa6d807 --- a/drivers/net/wireless/intel/iwlwifi/mvm/scan.c +++ b/drivers/net/wireless/intel/iwlwifi/mvm/scan.c @@@ -930,8 -930,11 +930,11 @@@ int iwl_mvm_config_scan(struct iwl_mvm if (WARN_ON(num_channels > mvm->fw->ucode_capa.n_scan_channels)) return -ENOBUFS; - if (type == mvm->scan_type) + if (type == mvm->scan_type) { + IWL_DEBUG_SCAN(mvm, + "Ignoring UMAC scan config of the same type\n"); return 0; + } cmd_size = sizeof(*scan_config) + mvm->fw->ucode_capa.n_scan_channels; @@@ -1109,7 -1112,7 +1112,7 @@@ static int iwl_mvm_scan_umac(struct iwl cmd->general_flags = cpu_to_le32(iwl_mvm_scan_umac_flags(mvm, params, vif)); - if (type == IWL_MVM_SCAN_SCHED) + if (type == IWL_MVM_SCAN_SCHED || type == IWL_MVM_SCAN_NETDETECT) cmd->flags = cpu_to_le32(IWL_UMAC_SCAN_FLAG_PREEMPTIVE); if (iwl_mvm_scan_use_ebs(mvm, vif)) @@@ -1298,10 -1301,6 +1301,10 @@@ int iwl_mvm_sched_scan_start(struct iwl return -EBUSY; } + /* we don't support "match all" in the firmware */ + if (!req->n_match_sets) + return -EOPNOTSUPP; + ret = iwl_mvm_check_running_scans(mvm, type); if (ret) return ret; @@@ -1355,7 -1354,7 +1358,7 @@@ if (fw_has_capa(&mvm->fw->ucode_capa, IWL_UCODE_TLV_CAPA_UMAC_SCAN)) { hcmd.id = iwl_cmd_id(SCAN_REQ_UMAC, IWL_ALWAYS_LONG_GROUP, 0); - ret = iwl_mvm_scan_umac(mvm, vif, &params, IWL_MVM_SCAN_SCHED); + ret = iwl_mvm_scan_umac(mvm, vif, &params, type); } else { hcmd.id = SCAN_OFFLOAD_REQUEST_CMD; ret = iwl_mvm_scan_lmac(mvm, vif, &params); diff --combined drivers/net/wireless/intel/iwlwifi/pcie/internal.h index 73c9559,2f95916..542bbc5 --- a/drivers/net/wireless/intel/iwlwifi/pcie/internal.h +++ b/drivers/net/wireless/intel/iwlwifi/pcie/internal.h @@@ -2,6 -2,7 +2,7 @@@ * * Copyright(c) 2003 - 2015 Intel Corporation. All rights reserved. * Copyright(c) 2013 - 2015 Intel Mobile Communications GmbH + * Copyright(c) 2016 Intel Deutschland GmbH * * Portions of this file are derived from the ipw3945 project, as well * as portions of the ieee80211 subsystem header files. @@@ -56,17 -57,23 +57,23 @@@ #define RX_NUM_QUEUES 1 #define RX_POST_REQ_ALLOC 2 #define RX_CLAIM_REQ_ALLOC 8 - #define RX_POOL_SIZE ((RX_CLAIM_REQ_ALLOC - RX_POST_REQ_ALLOC) * RX_NUM_QUEUES) - #define RX_LOW_WATERMARK 8 + #define RX_PENDING_WATERMARK 16 struct iwl_host_cmd; /*This file includes the declaration that are internal to the * trans_pcie layer */ + /** + * struct iwl_rx_mem_buffer + * @page_dma: bus address of rxb page + * @page: driver's pointer to the rxb page + * @vid: index of this rxb in the global table + */ struct iwl_rx_mem_buffer { dma_addr_t page_dma; struct page *page; + u16 vid; struct list_head list; }; @@@ -90,8 -97,12 +97,12 @@@ struct isr_statistics /** * struct iwl_rxq - Rx queue - * @bd: driver's pointer to buffer of receive buffer descriptors (rbd) + * @id: queue index + * @bd: driver's pointer to buffer of receive buffer descriptors (rbd). + * Address size is 32 bit in pre-9000 devices and 64 bit in 9000 devices. * @bd_dma: bus address of buffer of receive buffer descriptors (rbd) + * @ubd: driver's pointer to buffer of used receive buffer descriptors (rbd) + * @ubd_dma: physical address of buffer of used receive buffer descriptors (rbd) * @read: Shared index to newest available Rx buffer * @write: Shared index to oldest written Rx packet * @free_count: Number of pre-allocated buffers in rx_free @@@ -103,32 -114,34 +114,34 @@@ * @rb_stts: driver's pointer to receive buffer status * @rb_stts_dma: bus address of receive buffer status * @lock: - * @pool: initial pool of iwl_rx_mem_buffer for the queue - * @queue: actual rx queue + * @queue: actual rx queue. Not used for multi-rx queue. * * NOTE: rx_free and rx_used are used as a FIFO for iwl_rx_mem_buffers */ struct iwl_rxq { - __le32 *bd; + int id; + void *bd; dma_addr_t bd_dma; + __le32 *used_bd; + dma_addr_t used_bd_dma; u32 read; u32 write; u32 free_count; u32 used_count; u32 write_actual; + u32 queue_size; struct list_head rx_free; struct list_head rx_used; bool need_update; struct iwl_rb_status *rb_stts; dma_addr_t rb_stts_dma; spinlock_t lock; - struct iwl_rx_mem_buffer pool[RX_QUEUE_SIZE]; + struct napi_struct napi; struct iwl_rx_mem_buffer *queue[RX_QUEUE_SIZE]; }; /** * struct iwl_rb_allocator - Rx allocator - * @pool: initial pool of allocator * @req_pending: number of requests the allcator had not processed yet * @req_ready: number of requests honored and ready for claiming * @rbd_allocated: RBDs with pages allocated and ready to be handled to @@@ -140,7 -153,6 +153,6 @@@ * @rx_alloc: work struct for background calls */ struct iwl_rb_allocator { - struct iwl_rx_mem_buffer pool[RX_POOL_SIZE]; atomic_t req_pending; atomic_t req_ready; struct list_head rbd_allocated; @@@ -280,6 -292,7 +292,7 @@@ struct iwl_txq bool ampdu; bool block; unsigned long wd_timeout; + struct sk_buff_head overflow_q; }; static inline dma_addr_t @@@ -297,6 -310,8 +310,8 @@@ struct iwl_tso_hdr_page /** * struct iwl_trans_pcie - PCIe transport specific data * @rxq: all the RX queue data + * @rx_pool: initial pool of iwl_rx_mem_buffer for all the queues + * @global_table: table mapping received VID from hw to rxb * @rba: allocator for RX replenishing * @drv - pointer to iwl_drv * @trans: pointer to the generic transport area @@@ -323,13 -338,14 +338,14 @@@ * @fw_mon_size: size of the buffer for the firmware monitor */ struct iwl_trans_pcie { - struct iwl_rxq rxq; + struct iwl_rxq *rxq; + struct iwl_rx_mem_buffer rx_pool[MQ_RX_POOL_SIZE]; + struct iwl_rx_mem_buffer *global_table[MQ_RX_TABLE_SIZE]; struct iwl_rb_allocator rba; struct iwl_trans *trans; struct iwl_drv *drv; struct net_device napi_dev; - struct napi_struct napi; struct __percpu iwl_tso_hdr_page *tso_hdr_page; @@@ -359,6 -375,7 +375,7 @@@ bool ucode_write_complete; wait_queue_head_t ucode_write_waitq; wait_queue_head_t wait_command_queue; + wait_queue_head_t d0i3_waitq; u8 cmd_queue; u8 cmd_fifo; @@@ -490,15 -507,6 +507,15 @@@ static inline void iwl_enable_interrupt iwl_write32(trans, CSR_INT_MASK, trans_pcie->inta_mask); } +static inline void iwl_enable_fw_load_int(struct iwl_trans *trans) +{ + struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans); + + IWL_DEBUG_ISR(trans, "Enabling FW load interrupt\n"); + trans_pcie->inta_mask = CSR_INT_BIT_FH_TX; + iwl_write32(trans, CSR_INT_MASK, trans_pcie->inta_mask); +} + static inline void iwl_enable_rfkill_int(struct iwl_trans *trans) { struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans); @@@ -588,4 -596,7 +605,7 @@@ static inline int iwl_trans_pcie_dbgfs_ } #endif + int iwl_pci_fw_exit_d0i3(struct iwl_trans *trans); + int iwl_pci_fw_enter_d0i3(struct iwl_trans *trans); + #endif /* __iwl_trans_int_pcie_h__ */ diff --combined drivers/net/wireless/intel/iwlwifi/pcie/rx.c index 152cf9a,51314e56..07973ef --- a/drivers/net/wireless/intel/iwlwifi/pcie/rx.c +++ b/drivers/net/wireless/intel/iwlwifi/pcie/rx.c @@@ -2,6 -2,7 +2,7 @@@ * * Copyright(c) 2003 - 2014 Intel Corporation. All rights reserved. * Copyright(c) 2013 - 2015 Intel Mobile Communications GmbH + * Copyright(c) 2016 Intel Deutschland GmbH * * Portions of this file are derived from the ipw3945 project, as well * as portions of the ieee80211 subsystem header files. @@@ -140,8 -141,8 +141,8 @@@ */ static int iwl_rxq_space(const struct iwl_rxq *rxq) { - /* Make sure RX_QUEUE_SIZE is a power of 2 */ - BUILD_BUG_ON(RX_QUEUE_SIZE & (RX_QUEUE_SIZE - 1)); + /* Make sure rx queue size is a power of 2 */ + WARN_ON(rxq->queue_size & (rxq->queue_size - 1)); /* * There can be up to (RX_QUEUE_SIZE - 1) free slots, to avoid ambiguity @@@ -149,7 -150,7 +150,7 @@@ * The following is equivalent to modulo by RX_QUEUE_SIZE and is well * defined for negative dividends. */ - return (rxq->read - rxq->write - 1) & (RX_QUEUE_SIZE - 1); + return (rxq->read - rxq->write - 1) & (rxq->queue_size - 1); } /* @@@ -160,6 -161,12 +161,12 @@@ static inline __le32 iwl_pcie_dma_addr2 return cpu_to_le32((u32)(dma_addr >> 8)); } + static void iwl_pcie_write_prph_64(struct iwl_trans *trans, u64 ofs, u64 val) + { + iwl_write_prph(trans, ofs, val & 0xffffffff); + iwl_write_prph(trans, ofs + 4, val >> 32); + } + /* * iwl_pcie_rx_stop - stops the Rx DMA */ @@@ -173,10 -180,9 +180,9 @@@ int iwl_pcie_rx_stop(struct iwl_trans * /* * iwl_pcie_rxq_inc_wr_ptr - Update the write pointer for the RX queue */ - static void iwl_pcie_rxq_inc_wr_ptr(struct iwl_trans *trans) + static void iwl_pcie_rxq_inc_wr_ptr(struct iwl_trans *trans, + struct iwl_rxq *rxq) { - struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans); - struct iwl_rxq *rxq = &trans_pcie->rxq; u32 reg; lockdep_assert_held(&rxq->lock); @@@ -201,24 -207,73 +207,73 @@@ } rxq->write_actual = round_down(rxq->write, 8); - iwl_write32(trans, FH_RSCSR_CHNL0_WPTR, rxq->write_actual); + if (trans->cfg->mq_rx_supported) + iwl_write_prph(trans, RFH_Q_FRBDCB_WIDX(rxq->id), + rxq->write_actual); + else + iwl_write32(trans, FH_RSCSR_CHNL0_WPTR, rxq->write_actual); } static void iwl_pcie_rxq_check_wrptr(struct iwl_trans *trans) { struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans); - struct iwl_rxq *rxq = &trans_pcie->rxq; + int i; - spin_lock(&rxq->lock); + for (i = 0; i < trans->num_rx_queues; i++) { + struct iwl_rxq *rxq = &trans_pcie->rxq[i]; + + if (!rxq->need_update) + continue; + spin_lock(&rxq->lock); + iwl_pcie_rxq_inc_wr_ptr(trans, rxq); + rxq->need_update = false; + spin_unlock(&rxq->lock); + } + } + + static void iwl_pcie_rxq_mq_restock(struct iwl_trans *trans, + struct iwl_rxq *rxq) + { + struct iwl_rx_mem_buffer *rxb; + + /* + * If the device isn't enabled - no need to try to add buffers... + * This can happen when we stop the device and still have an interrupt + * pending. We stop the APM before we sync the interrupts because we + * have to (see comment there). On the other hand, since the APM is + * stopped, we cannot access the HW (in particular not prph). + * So don't try to restock if the APM has been already stopped. + */ + if (!test_bit(STATUS_DEVICE_ENABLED, &trans->status)) + return; - if (!rxq->need_update) - goto exit_unlock; + spin_lock(&rxq->lock); + while (rxq->free_count) { + __le64 *bd = (__le64 *)rxq->bd; - iwl_pcie_rxq_inc_wr_ptr(trans); - rxq->need_update = false; + /* Get next free Rx buffer, remove from free list */ + rxb = list_first_entry(&rxq->rx_free, struct iwl_rx_mem_buffer, + list); + list_del(&rxb->list); - exit_unlock: + /* 12 first bits are expected to be empty */ + WARN_ON(rxb->page_dma & DMA_BIT_MASK(12)); + /* Point to Rx buffer via next RBD in circular buffer */ + bd[rxq->write] = cpu_to_le64(rxb->page_dma | rxb->vid); + rxq->write = (rxq->write + 1) & MQ_RX_TABLE_MASK; + rxq->free_count--; + } spin_unlock(&rxq->lock); + + /* + * If we've added more space for the firmware to place data, tell it. + * Increment device's write pointer in multiples of 8. + */ + if (rxq->write_actual != (rxq->write & ~0x7)) { + spin_lock(&rxq->lock); + iwl_pcie_rxq_inc_wr_ptr(trans, rxq); + spin_unlock(&rxq->lock); + } } /* @@@ -232,10 -287,8 +287,8 @@@ * also updates the memory address in the firmware to reference the new * target buffer. */ - static void iwl_pcie_rxq_restock(struct iwl_trans *trans) + static void iwl_pcie_rxq_restock(struct iwl_trans *trans, struct iwl_rxq *rxq) { - struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans); - struct iwl_rxq *rxq = &trans_pcie->rxq; struct iwl_rx_mem_buffer *rxb; /* @@@ -251,6 -304,7 +304,7 @@@ spin_lock(&rxq->lock); while ((iwl_rxq_space(rxq) > 0) && (rxq->free_count)) { + __le32 *bd = (__le32 *)rxq->bd; /* The overwritten rxb must be a used one */ rxb = rxq->queue[rxq->write]; BUG_ON(rxb && rxb->page); @@@ -261,7 -315,7 +315,7 @@@ list_del(&rxb->list); /* Point to Rx buffer via next RBD in circular buffer */ - rxq->bd[rxq->write] = iwl_pcie_dma_addr2rbd_ptr(rxb->page_dma); + bd[rxq->write] = iwl_pcie_dma_addr2rbd_ptr(rxb->page_dma); rxq->queue[rxq->write] = rxb; rxq->write = (rxq->write + 1) & RX_QUEUE_MASK; rxq->free_count--; @@@ -272,7 -326,7 +326,7 @@@ * Increment device's write pointer in multiples of 8. */ if (rxq->write_actual != (rxq->write & ~0x7)) { spin_lock(&rxq->lock); - iwl_pcie_rxq_inc_wr_ptr(trans); + iwl_pcie_rxq_inc_wr_ptr(trans, rxq); spin_unlock(&rxq->lock); } } @@@ -285,13 -339,9 +339,9 @@@ static struct page *iwl_pcie_rx_alloc_p gfp_t priority) { struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans); - struct iwl_rxq *rxq = &trans_pcie->rxq; struct page *page; gfp_t gfp_mask = priority; - if (rxq->free_count > RX_LOW_WATERMARK) - gfp_mask |= __GFP_NOWARN; - if (trans_pcie->rx_page_order > 0) gfp_mask |= __GFP_COMP; @@@ -301,16 -351,13 +351,13 @@@ if (net_ratelimit()) IWL_DEBUG_INFO(trans, "alloc_pages failed, order: %d\n", trans_pcie->rx_page_order); - /* Issue an error if the hardware has consumed more than half - * of its free buffer list and we don't have enough - * pre-allocated buffers. + /* + * Issue an error if we don't have enough pre-allocated + * buffers. ` */ - if (rxq->free_count <= RX_LOW_WATERMARK && - iwl_rxq_space(rxq) > (RX_QUEUE_SIZE / 2) && - net_ratelimit()) + if (!(gfp_mask & __GFP_NOWARN) && net_ratelimit()) IWL_CRIT(trans, - "Failed to alloc_pages with GFP_KERNEL. Only %u free buffers remaining.\n", - rxq->free_count); + "Failed to alloc_pages\n"); return NULL; } return page; @@@ -325,10 -372,10 +372,10 @@@ * iwl_pcie_rxq_restock. The latter function will update the HW to use the newly * allocated buffers. */ - static void iwl_pcie_rxq_alloc_rbs(struct iwl_trans *trans, gfp_t priority) + static void iwl_pcie_rxq_alloc_rbs(struct iwl_trans *trans, gfp_t priority, + struct iwl_rxq *rxq) { struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans); - struct iwl_rxq *rxq = &trans_pcie->rxq; struct iwl_rx_mem_buffer *rxb; struct page *page; @@@ -372,10 -419,6 +419,6 @@@ __free_pages(page, trans_pcie->rx_page_order); return; } - /* dma address must be no more than 36 bits */ - BUG_ON(rxb->page_dma & ~DMA_BIT_MASK(36)); - /* and also 256 byte aligned! */ - BUG_ON(rxb->page_dma & DMA_BIT_MASK(8)); spin_lock(&rxq->lock); @@@ -386,41 -429,24 +429,24 @@@ } } - static void iwl_pcie_rxq_free_rbs(struct iwl_trans *trans) + static void iwl_pcie_free_rbs_pool(struct iwl_trans *trans) { struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans); - struct iwl_rxq *rxq = &trans_pcie->rxq; int i; - lockdep_assert_held(&rxq->lock); - - for (i = 0; i < RX_QUEUE_SIZE; i++) { - if (!rxq->pool[i].page) + for (i = 0; i < MQ_RX_POOL_SIZE; i++) { + if (!trans_pcie->rx_pool[i].page) continue; - dma_unmap_page(trans->dev, rxq->pool[i].page_dma, + dma_unmap_page(trans->dev, trans_pcie->rx_pool[i].page_dma, PAGE_SIZE << trans_pcie->rx_page_order, DMA_FROM_DEVICE); - __free_pages(rxq->pool[i].page, trans_pcie->rx_page_order); - rxq->pool[i].page = NULL; + __free_pages(trans_pcie->rx_pool[i].page, + trans_pcie->rx_page_order); + trans_pcie->rx_pool[i].page = NULL; } } /* - * iwl_pcie_rx_replenish - Move all used buffers from rx_used to rx_free - * - * When moving to rx_free an page is allocated for the slot. - * - * Also restock the Rx queue via iwl_pcie_rxq_restock. - * This is called only during initialization - */ - static void iwl_pcie_rx_replenish(struct iwl_trans *trans) - { - iwl_pcie_rxq_alloc_rbs(trans, GFP_KERNEL); - - iwl_pcie_rxq_restock(trans); - } - - /* * iwl_pcie_rx_allocator - Allocates pages in the background for RX queues * * Allocates for each received request 8 pages @@@ -444,6 -470,11 +470,11 @@@ static void iwl_pcie_rx_allocator(struc while (pending) { int i; struct list_head local_allocated; + gfp_t gfp_mask = GFP_KERNEL; + + /* Do not post a warning if there are only a few requests */ + if (pending < RX_PENDING_WATERMARK) + gfp_mask |= __GFP_NOWARN; INIT_LIST_HEAD(&local_allocated); @@@ -463,7 -494,7 +494,7 @@@ BUG_ON(rxb->page); /* Alloc a new receive buffer */ - page = iwl_pcie_rx_alloc_page(trans, GFP_KERNEL); + page = iwl_pcie_rx_alloc_page(trans, gfp_mask); if (!page) continue; rxb->page = page; @@@ -477,10 -508,6 +508,6 @@@ __free_pages(page, trans_pcie->rx_page_order); continue; } - /* dma address must be no more than 36 bits */ - BUG_ON(rxb->page_dma & ~DMA_BIT_MASK(36)); - /* and also 256 byte aligned! */ - BUG_ON(rxb->page_dma & DMA_BIT_MASK(8)); /* move the allocated entry to the out list */ list_move(&rxb->list, &local_allocated); @@@ -561,38 -588,83 +588,83 @@@ static void iwl_pcie_rx_allocator_work( static int iwl_pcie_rx_alloc(struct iwl_trans *trans) { struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans); - struct iwl_rxq *rxq = &trans_pcie->rxq; struct iwl_rb_allocator *rba = &trans_pcie->rba; struct device *dev = trans->dev; + int i; + int free_size = trans->cfg->mq_rx_supported ? sizeof(__le64) : + sizeof(__le32); + + if (WARN_ON(trans_pcie->rxq)) + return -EINVAL; - memset(&trans_pcie->rxq, 0, sizeof(trans_pcie->rxq)); + trans_pcie->rxq = kcalloc(trans->num_rx_queues, sizeof(struct iwl_rxq), + GFP_KERNEL); + if (!trans_pcie->rxq) + return -EINVAL; - spin_lock_init(&rxq->lock); spin_lock_init(&rba->lock); - if (WARN_ON(rxq->bd || rxq->rb_stts)) - return -EINVAL; + for (i = 0; i < trans->num_rx_queues; i++) { + struct iwl_rxq *rxq = &trans_pcie->rxq[i]; - /* Allocate the circular buffer of Read Buffer Descriptors (RBDs) */ - rxq->bd = dma_zalloc_coherent(dev, sizeof(__le32) * RX_QUEUE_SIZE, - &rxq->bd_dma, GFP_KERNEL); - if (!rxq->bd) - goto err_bd; + spin_lock_init(&rxq->lock); + if (trans->cfg->mq_rx_supported) + rxq->queue_size = MQ_RX_TABLE_SIZE; + else + rxq->queue_size = RX_QUEUE_SIZE; - /*Allocate the driver's pointer to receive buffer status */ - rxq->rb_stts = dma_zalloc_coherent(dev, sizeof(*rxq->rb_stts), - &rxq->rb_stts_dma, GFP_KERNEL); - if (!rxq->rb_stts) - goto err_rb_stts; + /* + * Allocate the circular buffer of Read Buffer Descriptors + * (RBDs) + */ + rxq->bd = dma_zalloc_coherent(dev, + free_size * rxq->queue_size, + &rxq->bd_dma, GFP_KERNEL); + if (!rxq->bd) + goto err; + + if (trans->cfg->mq_rx_supported) { + rxq->used_bd = dma_zalloc_coherent(dev, + sizeof(__le32) * + rxq->queue_size, + &rxq->used_bd_dma, + GFP_KERNEL); + if (!rxq->used_bd) + goto err; + } + /*Allocate the driver's pointer to receive buffer status */ + rxq->rb_stts = dma_zalloc_coherent(dev, sizeof(*rxq->rb_stts), + &rxq->rb_stts_dma, + GFP_KERNEL); + if (!rxq->rb_stts) + goto err; + } return 0; - err_rb_stts: - dma_free_coherent(dev, sizeof(__le32) * RX_QUEUE_SIZE, - rxq->bd, rxq->bd_dma); - rxq->bd_dma = 0; - rxq->bd = NULL; - err_bd: + err: + for (i = 0; i < trans->num_rx_queues; i++) { + struct iwl_rxq *rxq = &trans_pcie->rxq[i]; + + if (rxq->bd) + dma_free_coherent(dev, free_size * rxq->queue_size, + rxq->bd, rxq->bd_dma); + rxq->bd_dma = 0; + rxq->bd = NULL; + + if (rxq->rb_stts) + dma_free_coherent(trans->dev, + sizeof(struct iwl_rb_status), + rxq->rb_stts, rxq->rb_stts_dma); + + if (rxq->used_bd) + dma_free_coherent(dev, sizeof(__le32) * rxq->queue_size, + rxq->used_bd, rxq->used_bd_dma); + rxq->used_bd_dma = 0; + rxq->used_bd = NULL; + } + kfree(trans_pcie->rxq); + return -ENOMEM; } @@@ -659,65 -731,103 +731,103 @@@ static void iwl_pcie_rx_hw_init(struct iwl_set_bit(trans, CSR_INT_COALESCING, IWL_HOST_INT_OPER_MODE); } - static void iwl_pcie_rx_init_rxb_lists(struct iwl_rxq *rxq) + static void iwl_pcie_rx_mq_hw_init(struct iwl_trans *trans) { + struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans); + u32 rb_size, enabled = 0; int i; - lockdep_assert_held(&rxq->lock); - - INIT_LIST_HEAD(&rxq->rx_free); - INIT_LIST_HEAD(&rxq->rx_used); - rxq->free_count = 0; - rxq->used_count = 0; + switch (trans_pcie->rx_buf_size) { + case IWL_AMSDU_4K: + rb_size = RFH_RXF_DMA_RB_SIZE_4K; + break; + case IWL_AMSDU_8K: + rb_size = RFH_RXF_DMA_RB_SIZE_8K; + break; + case IWL_AMSDU_12K: + rb_size = RFH_RXF_DMA_RB_SIZE_12K; + break; + default: + WARN_ON(1); + rb_size = RFH_RXF_DMA_RB_SIZE_4K; + } - for (i = 0; i < RX_QUEUE_SIZE; i++) - list_add(&rxq->pool[i].list, &rxq->rx_used); - } + /* Stop Rx DMA */ + iwl_write_prph(trans, RFH_RXF_DMA_CFG, 0); + /* disable free amd used rx queue operation */ + iwl_write_prph(trans, RFH_RXF_RXQ_ACTIVE, 0); + + for (i = 0; i < trans->num_rx_queues; i++) { + /* Tell device where to find RBD free table in DRAM */ + iwl_pcie_write_prph_64(trans, RFH_Q_FRBDCB_BA_LSB(i), + (u64)(trans_pcie->rxq[i].bd_dma)); + /* Tell device where to find RBD used table in DRAM */ + iwl_pcie_write_prph_64(trans, RFH_Q_URBDCB_BA_LSB(i), + (u64)(trans_pcie->rxq[i].used_bd_dma)); + /* Tell device where in DRAM to update its Rx status */ + iwl_pcie_write_prph_64(trans, RFH_Q_URBD_STTS_WPTR_LSB(i), + trans_pcie->rxq[i].rb_stts_dma); + /* Reset device indice tables */ + iwl_write_prph(trans, RFH_Q_FRBDCB_WIDX(i), 0); + iwl_write_prph(trans, RFH_Q_FRBDCB_RIDX(i), 0); + iwl_write_prph(trans, RFH_Q_URBDCB_WIDX(i), 0); + + enabled |= BIT(i) | BIT(i + 16); + } - static void iwl_pcie_rx_init_rba(struct iwl_rb_allocator *rba) - { - int i; + /* restock default queue */ + iwl_pcie_rxq_mq_restock(trans, &trans_pcie->rxq[0]); - lockdep_assert_held(&rba->lock); + /* + * Enable Rx DMA + * Single frame mode + * Rx buffer size 4 or 8k or 12k + * Min RB size 4 or 8 + * 512 RBDs + */ + iwl_write_prph(trans, RFH_RXF_DMA_CFG, + RFH_DMA_EN_ENABLE_VAL | + rb_size | RFH_RXF_DMA_SINGLE_FRAME_MASK | + RFH_RXF_DMA_MIN_RB_4_8 | + RFH_RXF_DMA_RBDCB_SIZE_512); - INIT_LIST_HEAD(&rba->rbd_allocated); - INIT_LIST_HEAD(&rba->rbd_empty); + iwl_write_prph(trans, RFH_GEN_CFG, RFH_GEN_CFG_RFH_DMA_SNOOP | + RFH_GEN_CFG_SERVICE_DMA_SNOOP); + iwl_write_prph(trans, RFH_RXF_RXQ_ACTIVE, enabled); - for (i = 0; i < RX_POOL_SIZE; i++) - list_add(&rba->pool[i].list, &rba->rbd_empty); + /* Set interrupt coalescing timer to default (2048 usecs) */ + iwl_write8(trans, CSR_INT_COALESCING, IWL_HOST_INT_TIMEOUT_DEF); } - static void iwl_pcie_rx_free_rba(struct iwl_trans *trans) + static void iwl_pcie_rx_init_rxb_lists(struct iwl_rxq *rxq) { - struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans); - struct iwl_rb_allocator *rba = &trans_pcie->rba; - int i; + lockdep_assert_held(&rxq->lock); - lockdep_assert_held(&rba->lock); + INIT_LIST_HEAD(&rxq->rx_free); + INIT_LIST_HEAD(&rxq->rx_used); + rxq->free_count = 0; + rxq->used_count = 0; + } - for (i = 0; i < RX_POOL_SIZE; i++) { - if (!rba->pool[i].page) - continue; - dma_unmap_page(trans->dev, rba->pool[i].page_dma, - PAGE_SIZE << trans_pcie->rx_page_order, - DMA_FROM_DEVICE); - __free_pages(rba->pool[i].page, trans_pcie->rx_page_order); - rba->pool[i].page = NULL; - } + static int iwl_pcie_dummy_napi_poll(struct napi_struct *napi, int budget) + { + WARN_ON(1); + return 0; } int iwl_pcie_rx_init(struct iwl_trans *trans) { struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans); - struct iwl_rxq *rxq = &trans_pcie->rxq; + struct iwl_rxq *def_rxq; struct iwl_rb_allocator *rba = &trans_pcie->rba; - int i, err; + int i, err, num_rbds, allocator_pool_size; - if (!rxq->bd) { + if (!trans_pcie->rxq) { err = iwl_pcie_rx_alloc(trans); if (err) return err; } + def_rxq = trans_pcie->rxq; if (!rba->alloc_wq) rba->alloc_wq = alloc_workqueue("rb_allocator", WQ_HIGHPRI | WQ_UNBOUND, 1); @@@ -726,34 -836,68 +836,68 @@@ spin_lock(&rba->lock); atomic_set(&rba->req_pending, 0); atomic_set(&rba->req_ready, 0); - /* free all first - we might be reconfigured for a different size */ - iwl_pcie_rx_free_rba(trans); - iwl_pcie_rx_init_rba(rba); + INIT_LIST_HEAD(&rba->rbd_allocated); + INIT_LIST_HEAD(&rba->rbd_empty); spin_unlock(&rba->lock); - spin_lock(&rxq->lock); - /* free all first - we might be reconfigured for a different size */ - iwl_pcie_rxq_free_rbs(trans); - iwl_pcie_rx_init_rxb_lists(rxq); + iwl_pcie_free_rbs_pool(trans); for (i = 0; i < RX_QUEUE_SIZE; i++) - rxq->queue[i] = NULL; + def_rxq->queue[i] = NULL; - /* Set us so that we have processed and used all buffers, but have - * not restocked the Rx queue with fresh buffers */ - rxq->read = rxq->write = 0; - rxq->write_actual = 0; - memset(rxq->rb_stts, 0, sizeof(*rxq->rb_stts)); - spin_unlock(&rxq->lock); + for (i = 0; i < trans->num_rx_queues; i++) { + struct iwl_rxq *rxq = &trans_pcie->rxq[i]; - iwl_pcie_rx_replenish(trans); + rxq->id = i; - iwl_pcie_rx_hw_init(trans, rxq); + spin_lock(&rxq->lock); + /* + * Set read write pointer to reflect that we have processed + * and used all buffers, but have not restocked the Rx queue + * with fresh buffers + */ + rxq->read = 0; + rxq->write = 0; + rxq->write_actual = 0; + memset(rxq->rb_stts, 0, sizeof(*rxq->rb_stts)); - spin_lock(&rxq->lock); - iwl_pcie_rxq_inc_wr_ptr(trans); - spin_unlock(&rxq->lock); + iwl_pcie_rx_init_rxb_lists(rxq); + + if (!rxq->napi.poll) + netif_napi_add(&trans_pcie->napi_dev, &rxq->napi, + iwl_pcie_dummy_napi_poll, 64); + + spin_unlock(&rxq->lock); + } + + /* move the pool to the default queue and allocator ownerships */ + num_rbds = trans->cfg->mq_rx_supported ? + MQ_RX_POOL_SIZE : RX_QUEUE_SIZE; + allocator_pool_size = trans->num_rx_queues * + (RX_CLAIM_REQ_ALLOC - RX_POST_REQ_ALLOC); + for (i = 0; i < num_rbds; i++) { + struct iwl_rx_mem_buffer *rxb = &trans_pcie->rx_pool[i]; + + if (i < allocator_pool_size) + list_add(&rxb->list, &rba->rbd_empty); + else + list_add(&rxb->list, &def_rxq->rx_used); + trans_pcie->global_table[i] = rxb; + rxb->vid = (u16)i; + } + + iwl_pcie_rxq_alloc_rbs(trans, GFP_KERNEL, def_rxq); + if (trans->cfg->mq_rx_supported) { + iwl_pcie_rx_mq_hw_init(trans); + } else { + iwl_pcie_rxq_restock(trans, def_rxq); + iwl_pcie_rx_hw_init(trans, def_rxq); + } + + spin_lock(&def_rxq->lock); + iwl_pcie_rxq_inc_wr_ptr(trans, def_rxq); + spin_unlock(&def_rxq->lock); return 0; } @@@ -761,12 -905,16 +905,16 @@@ void iwl_pcie_rx_free(struct iwl_trans *trans) { struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans); - struct iwl_rxq *rxq = &trans_pcie->rxq; struct iwl_rb_allocator *rba = &trans_pcie->rba; + int free_size = trans->cfg->mq_rx_supported ? sizeof(__le64) : + sizeof(__le32); + int i; - /*if rxq->bd is NULL, it means that nothing has been allocated, - * exit now */ - if (!rxq->bd) { + /* + * if rxq is NULL, it means that nothing has been allocated, + * exit now + */ + if (!trans_pcie->rxq) { IWL_DEBUG_INFO(trans, "Free NULL rx context\n"); return; } @@@ -777,27 -925,37 +925,37 @@@ rba->alloc_wq = NULL; } - spin_lock(&rba->lock); - iwl_pcie_rx_free_rba(trans); - spin_unlock(&rba->lock); - - spin_lock(&rxq->lock); - iwl_pcie_rxq_free_rbs(trans); - spin_unlock(&rxq->lock); - - dma_free_coherent(trans->dev, sizeof(__le32) * RX_QUEUE_SIZE, - rxq->bd, rxq->bd_dma); - rxq->bd_dma = 0; - rxq->bd = NULL; - - if (rxq->rb_stts) - dma_free_coherent(trans->dev, - sizeof(struct iwl_rb_status), - rxq->rb_stts, rxq->rb_stts_dma); - else - IWL_DEBUG_INFO(trans, "Free rxq->rb_stts which is NULL\n"); - rxq->rb_stts_dma = 0; - rxq->rb_stts = NULL; + iwl_pcie_free_rbs_pool(trans); + + for (i = 0; i < trans->num_rx_queues; i++) { + struct iwl_rxq *rxq = &trans_pcie->rxq[i]; + + if (rxq->bd) + dma_free_coherent(trans->dev, + free_size * rxq->queue_size, + rxq->bd, rxq->bd_dma); + rxq->bd_dma = 0; + rxq->bd = NULL; + + if (rxq->rb_stts) + dma_free_coherent(trans->dev, + sizeof(struct iwl_rb_status), + rxq->rb_stts, rxq->rb_stts_dma); + else + IWL_DEBUG_INFO(trans, + "Free rxq->rb_stts which is NULL\n"); + + if (rxq->used_bd) + dma_free_coherent(trans->dev, + sizeof(__le32) * rxq->queue_size, + rxq->used_bd, rxq->used_bd_dma); + rxq->used_bd_dma = 0; + rxq->used_bd = NULL; + + if (rxq->napi.poll) + netif_napi_del(&rxq->napi); + } + kfree(trans_pcie->rxq); } /* @@@ -841,11 -999,11 +999,11 @@@ static void iwl_pcie_rx_reuse_rbd(struc } static void iwl_pcie_rx_handle_rb(struct iwl_trans *trans, + struct iwl_rxq *rxq, struct iwl_rx_mem_buffer *rxb, bool emergency) { struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans); - struct iwl_rxq *rxq = &trans_pcie->rxq; struct iwl_txq *txq = &trans_pcie->txq[trans_pcie->cmd_queue]; bool page_stolen = false; int max_len = PAGE_SIZE << trans_pcie->rx_page_order; @@@ -911,7 -1069,12 +1069,12 @@@ index = SEQ_TO_INDEX(sequence); cmd_index = get_cmd_index(&txq->q, index); - iwl_op_mode_rx(trans->op_mode, &trans_pcie->napi, &rxcb); + if (rxq->id == 0) + iwl_op_mode_rx(trans->op_mode, &rxq->napi, + &rxcb); + else + iwl_op_mode_rx_rss(trans->op_mode, &rxq->napi, + &rxcb, rxq->id); if (reclaim) { kzfree(txq->entries[cmd_index].free_buf); @@@ -975,7 -1138,7 +1138,7 @@@ static void iwl_pcie_rx_handle(struct iwl_trans *trans) { struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans); - struct iwl_rxq *rxq = &trans_pcie->rxq; + struct iwl_rxq *rxq = &trans_pcie->rxq[0]; u32 r, i, j, count = 0; bool emergency = false; @@@ -993,16 -1156,26 +1156,26 @@@ restart while (i != r) { struct iwl_rx_mem_buffer *rxb; - if (unlikely(rxq->used_count == RX_QUEUE_SIZE / 2)) + if (unlikely(rxq->used_count == rxq->queue_size / 2)) emergency = true; - rxb = rxq->queue[i]; - rxq->queue[i] = NULL; + if (trans->cfg->mq_rx_supported) { + /* + * used_bd is a 32 bit but only 12 are used to retrieve + * the vid + */ + u16 vid = (u16)le32_to_cpu(rxq->used_bd[i]); + + rxb = trans_pcie->global_table[vid]; + } else { + rxb = rxq->queue[i]; + rxq->queue[i] = NULL; + } IWL_DEBUG_RX(trans, "rxbuf: HW = %d, SW = %d\n", r, i); - iwl_pcie_rx_handle_rb(trans, rxb, emergency); + iwl_pcie_rx_handle_rb(trans, rxq, rxb, emergency); - i = (i + 1) & RX_QUEUE_MASK; + i = (i + 1) & (rxq->queue_size - 1); /* If we have RX_CLAIM_REQ_ALLOC released rx buffers - * try to claim the pre-allocated buffers from the allocator */ @@@ -1040,10 -1213,10 +1213,10 @@@ count++; if (count == 8) { count = 0; - if (rxq->used_count < RX_QUEUE_SIZE / 3) + if (rxq->used_count < rxq->queue_size / 3) emergency = false; spin_unlock(&rxq->lock); - iwl_pcie_rxq_alloc_rbs(trans, GFP_ATOMIC); + iwl_pcie_rxq_alloc_rbs(trans, GFP_ATOMIC, rxq); spin_lock(&rxq->lock); } } @@@ -1055,7 -1228,10 +1228,10 @@@ if (rxq->free_count >= RX_CLAIM_REQ_ALLOC) { rxq->read = i; spin_unlock(&rxq->lock); - iwl_pcie_rxq_restock(trans); + if (trans->cfg->mq_rx_supported) + iwl_pcie_rxq_mq_restock(trans, rxq); + else + iwl_pcie_rxq_restock(trans, rxq); goto restart; } } @@@ -1077,10 -1253,10 +1253,10 @@@ * will be restocked by the next call of iwl_pcie_rxq_restock. */ if (unlikely(emergency && count)) - iwl_pcie_rxq_alloc_rbs(trans, GFP_ATOMIC); + iwl_pcie_rxq_alloc_rbs(trans, GFP_ATOMIC, rxq); - if (trans_pcie->napi.poll) - napi_gro_flush(&trans_pcie->napi, false); + if (rxq->napi.poll) + napi_gro_flush(&rxq->napi, false); } /* @@@ -1438,11 -1614,9 +1614,11 @@@ irqreturn_t iwl_pcie_irq_handler(int ir inta & ~trans_pcie->inta_mask); } - /* Re-enable all interrupts */ - /* only Re-enable if disabled by irq */ - if (test_bit(STATUS_INT_ENABLED, &trans->status)) + /* we are loading the firmware, enable FH_TX interrupt only */ + if (handled & CSR_INT_BIT_FH_TX) + iwl_enable_fw_load_int(trans); + /* only Re-enable all interrupt if disabled by irq */ + else if (test_bit(STATUS_INT_ENABLED, &trans->status)) iwl_enable_interrupts(trans); /* Re-enable RF_KILL if it occurred */ else if (handled & CSR_INT_BIT_RF_KILL) diff --combined drivers/net/wireless/intel/iwlwifi/pcie/trans.c index 5a854c6,b796952..58591ca --- a/drivers/net/wireless/intel/iwlwifi/pcie/trans.c +++ b/drivers/net/wireless/intel/iwlwifi/pcie/trans.c @@@ -72,6 -72,7 +72,7 @@@ #include <linux/bitops.h> #include <linux/gfp.h> #include <linux/vmalloc.h> + #include <linux/pm_runtime.h> #include "iwl-drv.h" #include "iwl-trans.h" @@@ -1021,6 -1022,82 +1022,6 @@@ static int iwl_pcie_load_given_ucode_80 &first_ucode_section); } -static int iwl_trans_pcie_start_fw(struct iwl_trans *trans, - const struct fw_img *fw, bool run_in_rfkill) -{ - struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans); - bool hw_rfkill; - int ret; - - mutex_lock(&trans_pcie->mutex); - - /* Someone called stop_device, don't try to start_fw */ - if (trans_pcie->is_down) { - IWL_WARN(trans, - "Can't start_fw since the HW hasn't been started\n"); - ret = EIO; - goto out; - } - - /* This may fail if AMT took ownership of the device */ - if (iwl_pcie_prepare_card_hw(trans)) { - IWL_WARN(trans, "Exit HW not ready\n"); - ret = -EIO; - goto out; - } - - iwl_enable_rfkill_int(trans); - - /* If platform's RF_KILL switch is NOT set to KILL */ - hw_rfkill = iwl_is_rfkill_set(trans); - if (hw_rfkill) - set_bit(STATUS_RFKILL, &trans->status); - else - clear_bit(STATUS_RFKILL, &trans->status); - iwl_trans_pcie_rf_kill(trans, hw_rfkill); - if (hw_rfkill && !run_in_rfkill) { - ret = -ERFKILL; - goto out; - } - - iwl_write32(trans, CSR_INT, 0xFFFFFFFF); - - ret = iwl_pcie_nic_init(trans); - if (ret) { - IWL_ERR(trans, "Unable to init nic\n"); - goto out; - } - - /* make sure rfkill handshake bits are cleared */ - iwl_write32(trans, CSR_UCODE_DRV_GP1_CLR, CSR_UCODE_SW_BIT_RFKILL); - iwl_write32(trans, CSR_UCODE_DRV_GP1_CLR, - CSR_UCODE_DRV_GP1_BIT_CMD_BLOCKED); - - /* clear (again), then enable host interrupts */ - iwl_write32(trans, CSR_INT, 0xFFFFFFFF); - iwl_enable_interrupts(trans); - - /* really make sure rfkill handshake bits are cleared */ - iwl_write32(trans, CSR_UCODE_DRV_GP1_CLR, CSR_UCODE_SW_BIT_RFKILL); - iwl_write32(trans, CSR_UCODE_DRV_GP1_CLR, CSR_UCODE_SW_BIT_RFKILL); - - /* Load the given image to the HW */ - if (trans->cfg->device_family == IWL_DEVICE_FAMILY_8000) - ret = iwl_pcie_load_given_ucode_8000(trans, fw); - else - ret = iwl_pcie_load_given_ucode(trans, fw); - -out: - mutex_unlock(&trans_pcie->mutex); - return ret; -} - -static void iwl_trans_pcie_fw_alive(struct iwl_trans *trans, u32 scd_addr) -{ - iwl_pcie_reset_ict(trans); - iwl_pcie_tx_start(trans, scd_addr); -} - static void _iwl_trans_pcie_stop_device(struct iwl_trans *trans, bool low_power) { struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans); @@@ -1051,8 -1128,7 +1052,8 @@@ * already dead. */ if (test_and_clear_bit(STATUS_DEVICE_ENABLED, &trans->status)) { - IWL_DEBUG_INFO(trans, "DEVICE_ENABLED bit was set and is now cleared\n"); + IWL_DEBUG_INFO(trans, + "DEVICE_ENABLED bit was set and is now cleared\n"); iwl_pcie_tx_stop(trans); iwl_pcie_rx_stop(trans); @@@ -1086,6 -1162,7 +1087,6 @@@ iwl_disable_interrupts(trans); spin_unlock(&trans_pcie->irq_lock); - /* clear all status bits */ clear_bit(STATUS_SYNC_HCMD_ACTIVE, &trans->status); clear_bit(STATUS_INT_ENABLED, &trans->status); @@@ -1118,116 -1195,10 +1119,116 @@@ if (hw_rfkill != was_hw_rfkill) iwl_trans_pcie_rf_kill(trans, hw_rfkill); - /* re-take ownership to prevent other users from stealing the deivce */ + /* re-take ownership to prevent other users from stealing the device */ iwl_pcie_prepare_card_hw(trans); } +static int iwl_trans_pcie_start_fw(struct iwl_trans *trans, + const struct fw_img *fw, bool run_in_rfkill) +{ + struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans); + bool hw_rfkill; + int ret; + + /* This may fail if AMT took ownership of the device */ + if (iwl_pcie_prepare_card_hw(trans)) { + IWL_WARN(trans, "Exit HW not ready\n"); + ret = -EIO; + goto out; + } + + iwl_enable_rfkill_int(trans); + + iwl_write32(trans, CSR_INT, 0xFFFFFFFF); + + /* + * We enabled the RF-Kill interrupt and the handler may very + * well be running. Disable the interrupts to make sure no other + * interrupt can be fired. + */ + iwl_disable_interrupts(trans); + + /* Make sure it finished running */ + synchronize_irq(trans_pcie->pci_dev->irq); + + mutex_lock(&trans_pcie->mutex); + + /* If platform's RF_KILL switch is NOT set to KILL */ + hw_rfkill = iwl_is_rfkill_set(trans); + if (hw_rfkill) + set_bit(STATUS_RFKILL, &trans->status); + else + clear_bit(STATUS_RFKILL, &trans->status); + iwl_trans_pcie_rf_kill(trans, hw_rfkill); + if (hw_rfkill && !run_in_rfkill) { + ret = -ERFKILL; + goto out; + } + + /* Someone called stop_device, don't try to start_fw */ + if (trans_pcie->is_down) { + IWL_WARN(trans, + "Can't start_fw since the HW hasn't been started\n"); + ret = -EIO; + goto out; + } + + /* make sure rfkill handshake bits are cleared */ + iwl_write32(trans, CSR_UCODE_DRV_GP1_CLR, CSR_UCODE_SW_BIT_RFKILL); + iwl_write32(trans, CSR_UCODE_DRV_GP1_CLR, + CSR_UCODE_DRV_GP1_BIT_CMD_BLOCKED); + + /* clear (again), then enable host interrupts */ + iwl_write32(trans, CSR_INT, 0xFFFFFFFF); + + ret = iwl_pcie_nic_init(trans); + if (ret) { + IWL_ERR(trans, "Unable to init nic\n"); + goto out; + } + + /* + * Now, we load the firmware and don't want to be interrupted, even + * by the RF-Kill interrupt (hence mask all the interrupt besides the + * FH_TX interrupt which is needed to load the firmware). If the + * RF-Kill switch is toggled, we will find out after having loaded + * the firmware and return the proper value to the caller. + */ + iwl_enable_fw_load_int(trans); + + /* really make sure rfkill handshake bits are cleared */ + iwl_write32(trans, CSR_UCODE_DRV_GP1_CLR, CSR_UCODE_SW_BIT_RFKILL); + iwl_write32(trans, CSR_UCODE_DRV_GP1_CLR, CSR_UCODE_SW_BIT_RFKILL); + + /* Load the given image to the HW */ + if (trans->cfg->device_family == IWL_DEVICE_FAMILY_8000) + ret = iwl_pcie_load_given_ucode_8000(trans, fw); + else + ret = iwl_pcie_load_given_ucode(trans, fw); + iwl_enable_interrupts(trans); + + /* re-check RF-Kill state since we may have missed the interrupt */ + hw_rfkill = iwl_is_rfkill_set(trans); + if (hw_rfkill) + set_bit(STATUS_RFKILL, &trans->status); + else + clear_bit(STATUS_RFKILL, &trans->status); + + iwl_trans_pcie_rf_kill(trans, hw_rfkill); + if (hw_rfkill && !run_in_rfkill) + ret = -ERFKILL; + +out: + mutex_unlock(&trans_pcie->mutex); + return ret; +} + +static void iwl_trans_pcie_fw_alive(struct iwl_trans *trans, u32 scd_addr) +{ + iwl_pcie_reset_ict(trans); + iwl_pcie_tx_start(trans, scd_addr); +} + static void iwl_trans_pcie_stop_device(struct iwl_trans *trans, bool low_power) { struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans); @@@ -1248,11 -1219,12 +1249,12 @@@ void iwl_trans_pcie_rf_kill(struct iwl_ _iwl_trans_pcie_stop_device(trans, true); } - static void iwl_trans_pcie_d3_suspend(struct iwl_trans *trans, bool test) + static void iwl_trans_pcie_d3_suspend(struct iwl_trans *trans, bool test, + bool reset) { struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans); - if (trans->system_pm_mode == IWL_PLAT_PM_MODE_D0I3) { + if (!reset) { /* Enable persistence mode to avoid reset */ iwl_set_bit(trans, CSR_HW_IF_CONFIG_REG, CSR_HW_IF_CONFIG_REG_PERSIST_MODE); @@@ -1276,7 -1248,7 +1278,7 @@@ iwl_clear_bit(trans, CSR_GP_CNTRL, CSR_GP_CNTRL_REG_FLAG_INIT_DONE); - if (trans->system_pm_mode == IWL_PLAT_PM_MODE_D3) { + if (reset) { /* * reset TX queues -- some of their registers reset during S3 * so if we don't reset everything here the D3 image would try @@@ -1290,7 -1262,7 +1292,7 @@@ static int iwl_trans_pcie_d3_resume(struct iwl_trans *trans, enum iwl_d3_status *status, - bool test) + bool test, bool reset) { u32 val; int ret; @@@ -1325,7 -1297,7 +1327,7 @@@ iwl_pcie_set_pwr(trans, false); - if (trans->system_pm_mode == IWL_PLAT_PM_MODE_D0I3) { + if (!reset) { iwl_clear_bit(trans, CSR_GP_CNTRL, CSR_GP_CNTRL_REG_FLAG_MAC_ACCESS_REQ); } else { @@@ -1383,6 -1355,10 +1385,10 @@@ static int _iwl_trans_pcie_start_hw(str /* ... rfkill can call stop_device and set it false if needed */ iwl_trans_pcie_rf_kill(trans, hw_rfkill); + /* Make sure we sync here, because we'll need full access later */ + if (low_power) + pm_runtime_resume(trans->dev); + return 0; } @@@ -1452,12 -1428,6 +1458,6 @@@ static void iwl_trans_pcie_write_prph(s iwl_trans_pcie_write32(trans, HBUS_TARG_PRPH_WDAT, val); } - static int iwl_pcie_dummy_napi_poll(struct napi_struct *napi, int budget) - { - WARN_ON(1); - return 0; - } - static void iwl_trans_pcie_configure(struct iwl_trans *trans, const struct iwl_trans_config *trans_cfg) { @@@ -1494,11 -1464,8 +1494,8 @@@ * As this function may be called again in some corner cases don't * do anything if NAPI was already initialized. */ - if (!trans_pcie->napi.poll) { + if (trans_pcie->napi_dev.reg_state != NETREG_DUMMY) init_dummy_netdev(&trans_pcie->napi_dev); - netif_napi_add(&trans_pcie->napi_dev, &trans_pcie->napi, - iwl_pcie_dummy_napi_poll, 64); - } } void iwl_trans_pcie_free(struct iwl_trans *trans) @@@ -1506,6 -1473,9 +1503,9 @@@ struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans); int i; + /* TODO: check if this is really needed */ + pm_runtime_disable(trans->dev); + synchronize_irq(trans_pcie->pci_dev->irq); iwl_pcie_tx_free(trans); @@@ -1519,9 -1489,6 +1519,6 @@@ pci_release_regions(trans_pcie->pci_dev); pci_disable_device(trans_pcie->pci_dev); - if (trans_pcie->napi.poll) - netif_napi_del(&trans_pcie->napi); - iwl_pcie_free_fw_monitor(trans); for_each_possible_cpu(i) { @@@ -1861,6 -1828,7 +1858,7 @@@ void iwl_trans_pcie_ref(struct iwl_tran spin_lock_irqsave(&trans_pcie->ref_lock, flags); IWL_DEBUG_RPM(trans, "ref_counter: %d\n", trans_pcie->ref_count); trans_pcie->ref_count++; + pm_runtime_get(&trans_pcie->pci_dev->dev); spin_unlock_irqrestore(&trans_pcie->ref_lock, flags); } @@@ -1879,6 -1847,10 +1877,10 @@@ void iwl_trans_pcie_unref(struct iwl_tr return; } trans_pcie->ref_count--; + + pm_runtime_mark_last_busy(&trans_pcie->pci_dev->dev); + pm_runtime_put_autosuspend(&trans_pcie->pci_dev->dev); + spin_unlock_irqrestore(&trans_pcie->ref_lock, flags); } @@@ -2031,29 -2003,48 +2033,48 @@@ static ssize_t iwl_dbgfs_rx_queue_read( { struct iwl_trans *trans = file->private_data; struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans); - struct iwl_rxq *rxq = &trans_pcie->rxq; - char buf[256]; - int pos = 0; - const size_t bufsz = sizeof(buf); - - pos += scnprintf(buf + pos, bufsz - pos, "read: %u\n", - rxq->read); - pos += scnprintf(buf + pos, bufsz - pos, "write: %u\n", - rxq->write); - pos += scnprintf(buf + pos, bufsz - pos, "write_actual: %u\n", - rxq->write_actual); - pos += scnprintf(buf + pos, bufsz - pos, "need_update: %d\n", - rxq->need_update); - pos += scnprintf(buf + pos, bufsz - pos, "free_count: %u\n", - rxq->free_count); - if (rxq->rb_stts) { - pos += scnprintf(buf + pos, bufsz - pos, "closed_rb_num: %u\n", - le16_to_cpu(rxq->rb_stts->closed_rb_num) & 0x0FFF); - } else { - pos += scnprintf(buf + pos, bufsz - pos, - "closed_rb_num: Not Allocated\n"); + char *buf; + int pos = 0, i, ret; + size_t bufsz = sizeof(buf); + + bufsz = sizeof(char) * 121 * trans->num_rx_queues; + + if (!trans_pcie->rxq) + return -EAGAIN; + + buf = kzalloc(bufsz, GFP_KERNEL); + if (!buf) + return -ENOMEM; + + for (i = 0; i < trans->num_rx_queues && pos < bufsz; i++) { + struct iwl_rxq *rxq = &trans_pcie->rxq[i]; + + pos += scnprintf(buf + pos, bufsz - pos, "queue#: %2d\n", + i); + pos += scnprintf(buf + pos, bufsz - pos, "\tread: %u\n", + rxq->read); + pos += scnprintf(buf + pos, bufsz - pos, "\twrite: %u\n", + rxq->write); + pos += scnprintf(buf + pos, bufsz - pos, "\twrite_actual: %u\n", + rxq->write_actual); + pos += scnprintf(buf + pos, bufsz - pos, "\tneed_update: %2d\n", + rxq->need_update); + pos += scnprintf(buf + pos, bufsz - pos, "\tfree_count: %u\n", + rxq->free_count); + if (rxq->rb_stts) { + pos += scnprintf(buf + pos, bufsz - pos, + "\tclosed_rb_num: %u\n", + le16_to_cpu(rxq->rb_stts->closed_rb_num) & + 0x0FFF); + } else { + pos += scnprintf(buf + pos, bufsz - pos, + "\tclosed_rb_num: Not Allocated\n"); + } } - return simple_read_from_buffer(user_buf, count, ppos, buf, pos); + ret = simple_read_from_buffer(user_buf, count, ppos, buf, pos); + kfree(buf); + + return ret; } static ssize_t iwl_dbgfs_interrupt_read(struct file *file, @@@ -2218,7 -2209,8 +2239,8 @@@ static u32 iwl_trans_pcie_dump_rbs(stru { struct iwl_trans_pcie *trans_pcie = IWL_TRANS_GET_PCIE_TRANS(trans); int max_len = PAGE_SIZE << trans_pcie->rx_page_order; - struct iwl_rxq *rxq = &trans_pcie->rxq; + /* Dump RBs is supported only for pre-9000 devices (1 queue) */ + struct iwl_rxq *rxq = &trans_pcie->rxq[0]; u32 i, r, j, rb_len = 0; spin_lock(&rxq->lock); @@@ -2413,7 -2405,8 +2435,8 @@@ static struct iwl_trans_dump_dat u32 len, num_rbs; u32 monitor_len; int i, ptr; - bool dump_rbs = test_bit(STATUS_FW_ERROR, &trans->status); + bool dump_rbs = test_bit(STATUS_FW_ERROR, &trans->status) && + !trans->cfg->mq_rx_supported; /* transport dump header */ len = sizeof(*dump_data); @@@ -2468,11 -2461,12 +2491,12 @@@ len += sizeof(*data) + (FH_MEM_UPPER_BOUND - FH_MEM_LOWER_BOUND); if (dump_rbs) { + /* Dump RBs is supported only for pre-9000 devices (1 queue) */ + struct iwl_rxq *rxq = &trans_pcie->rxq[0]; /* RBs */ - num_rbs = le16_to_cpu(ACCESS_ONCE( - trans_pcie->rxq.rb_stts->closed_rb_num)) + num_rbs = le16_to_cpu(ACCESS_ONCE(rxq->rb_stts->closed_rb_num)) & 0x0FFF; - num_rbs = (num_rbs - trans_pcie->rxq.read) & RX_QUEUE_MASK; + num_rbs = (num_rbs - rxq->read) & RX_QUEUE_MASK; len += num_rbs * (sizeof(*data) + sizeof(struct iwl_fw_error_dump_rb) + (PAGE_SIZE << trans_pcie->rx_page_order)); @@@ -2523,6 -2517,22 +2547,22 @@@ return dump_data; } + #ifdef CONFIG_PM_SLEEP + static int iwl_trans_pcie_suspend(struct iwl_trans *trans) + { + if (trans->runtime_pm_mode == IWL_PLAT_PM_MODE_D0I3) + return iwl_pci_fw_enter_d0i3(trans); + + return 0; + } + + static void iwl_trans_pcie_resume(struct iwl_trans *trans) + { + if (trans->runtime_pm_mode == IWL_PLAT_PM_MODE_D0I3) + iwl_pci_fw_exit_d0i3(trans); + } + #endif /* CONFIG_PM_SLEEP */ + static const struct iwl_trans_ops trans_ops_pcie = { .start_hw = iwl_trans_pcie_start_hw, .op_mode_leave = iwl_trans_pcie_op_mode_leave, @@@ -2533,6 -2543,11 +2573,11 @@@ .d3_suspend = iwl_trans_pcie_d3_suspend, .d3_resume = iwl_trans_pcie_d3_resume, + #ifdef CONFIG_PM_SLEEP + .suspend = iwl_trans_pcie_suspend, + .resume = iwl_trans_pcie_resume, + #endif /* CONFIG_PM_SLEEP */ + .send_cmd = iwl_trans_pcie_send_hcmd, .tx = iwl_trans_pcie_tx, @@@ -2571,7 -2586,7 +2616,7 @@@ struct iwl_trans *iwl_trans_pcie_alloc( struct iwl_trans_pcie *trans_pcie; struct iwl_trans *trans; u16 pci_cmd; - int ret; + int ret, addr_size; trans = iwl_trans_alloc(sizeof(struct iwl_trans_pcie), &pdev->dev, cfg, &trans_ops_pcie, 0); @@@ -2609,11 -2624,17 +2654,17 @@@ PCIE_LINK_STATE_CLKPM); } + if (cfg->mq_rx_supported) + addr_size = 64; + else + addr_size = 36; + pci_set_master(pdev); - ret = pci_set_dma_mask(pdev, DMA_BIT_MASK(36)); + ret = pci_set_dma_mask(pdev, DMA_BIT_MASK(addr_size)); if (!ret) - ret = pci_set_consistent_dma_mask(pdev, DMA_BIT_MASK(36)); + ret = pci_set_consistent_dma_mask(pdev, + DMA_BIT_MASK(addr_size)); if (ret) { ret = pci_set_dma_mask(pdev, DMA_BIT_MASK(32)); if (!ret) @@@ -2716,6 -2737,8 +2767,8 @@@ /* Initialize the wait queue for commands */ init_waitqueue_head(&trans_pcie->wait_command_queue); + init_waitqueue_head(&trans_pcie->d0i3_waitq); + ret = iwl_pcie_alloc_ict(trans); if (ret) goto out_pci_disable_msi; @@@ -2730,6 -2753,12 +2783,12 @@@ trans_pcie->inta_mask = CSR_INI_SET_MASK; + #ifdef CONFIG_IWLWIFI_PCIE_RTPM + trans->runtime_pm_mode = IWL_PLAT_PM_MODE_D0I3; + #else + trans->runtime_pm_mode = IWL_PLAT_PM_MODE_DISABLED; + #endif /* CONFIG_IWLWIFI_PCIE_RTPM */ + return trans; out_free_ict: diff --combined include/linux/netdevice.h index 5440b7b,47671ce0..e52077f --- a/include/linux/netdevice.h +++ b/include/linux/netdevice.h @@@ -51,6 -51,7 +51,7 @@@ #include <linux/neighbour.h> #include <uapi/linux/netdevice.h> #include <uapi/linux/if_bonding.h> + #include <uapi/linux/pkt_cls.h> struct netpoll_info; struct device; @@@ -778,6 -779,25 +779,25 @@@ static inline bool netdev_phys_item_id_ typedef u16 (*select_queue_fallback_t)(struct net_device *dev, struct sk_buff *skb); + /* These structures hold the attributes of qdisc and classifiers + * that are being passed to the netdevice through the setup_tc op. + */ + enum { + TC_SETUP_MQPRIO, + TC_SETUP_CLSU32, + }; + + struct tc_cls_u32_offload; + + struct tc_to_netdev { + unsigned int type; + union { + u8 tc; + struct tc_cls_u32_offload *cls_u32; + }; + }; + + /* * This structure defines the management hooks for network devices. * The following hooks can be defined; unless noted otherwise, they are @@@ -1150,7 -1170,10 +1170,10 @@@ struct net_device_ops int (*ndo_set_vf_rss_query_en)( struct net_device *dev, int vf, bool setting); - int (*ndo_setup_tc)(struct net_device *dev, u8 tc); + int (*ndo_setup_tc)(struct net_device *dev, + u32 handle, + __be16 protocol, + struct tc_to_netdev *tc); #if IS_ENABLED(CONFIG_FCOE) int (*ndo_fcoe_enable)(struct net_device *dev); int (*ndo_fcoe_disable)(struct net_device *dev); @@@ -1291,6 -1314,7 +1314,7 @@@ * @IFF_OPENVSWITCH: device is a Open vSwitch master * @IFF_L3MDEV_SLAVE: device is enslaved to an L3 master device * @IFF_TEAM: device is a team device + * @IFF_RXFH_CONFIGURED: device has had Rx Flow indirection table configured */ enum netdev_priv_flags { IFF_802_1Q_VLAN = 1<<0, @@@ -1318,6 -1342,7 +1342,7 @@@ IFF_OPENVSWITCH = 1<<22, IFF_L3MDEV_SLAVE = 1<<23, IFF_TEAM = 1<<24, + IFF_RXFH_CONFIGURED = 1<<25, }; #define IFF_802_1Q_VLAN IFF_802_1Q_VLAN @@@ -1345,6 -1370,7 +1370,7 @@@ #define IFF_OPENVSWITCH IFF_OPENVSWITCH #define IFF_L3MDEV_SLAVE IFF_L3MDEV_SLAVE #define IFF_TEAM IFF_TEAM + #define IFF_RXFH_CONFIGURED IFF_RXFH_CONFIGURED /** * struct net_device - The DEVICE structure. @@@ -1397,6 -1423,8 +1423,8 @@@ * do not use this in drivers * @tx_dropped: Dropped packets by core network, * do not use this in drivers + * @rx_nohandler: nohandler dropped packets by core network on + * inactive devices, do not use this in drivers * * @wireless_handlers: List of functions to handle Wireless Extensions, * instead of ioctl, @@@ -1611,6 -1639,7 +1639,7 @@@ struct net_device atomic_long_t rx_dropped; atomic_long_t tx_dropped; + atomic_long_t rx_nohandler; #ifdef CONFIG_WIRELESS_EXT const struct iw_handler_def * wireless_handlers; @@@ -3718,7 -3747,7 +3747,7 @@@ void *netdev_lower_get_next_private_rcu void *netdev_lower_get_next(struct net_device *dev, struct list_head **iter); #define netdev_for_each_lower_dev(dev, ldev, iter) \ - for (iter = &(dev)->adj_list.lower, \ + for (iter = (dev)->adj_list.lower.next, \ ldev = netdev_lower_get_next(dev, &(iter)); \ ldev; \ ldev = netdev_lower_get_next(dev, &(iter))) @@@ -3741,7 -3770,7 +3770,7 @@@ void netdev_lower_state_changed(struct /* RSS keys are 40 or 52 bytes long */ #define NETDEV_RSS_KEY_LEN 52 - extern u8 netdev_rss_key[NETDEV_RSS_KEY_LEN]; + extern u8 netdev_rss_key[NETDEV_RSS_KEY_LEN] __read_mostly; void netdev_rss_key_fill(void *buffer, size_t len); int dev_get_nest_level(struct net_device *dev, @@@ -4045,6 -4074,11 +4074,11 @@@ static inline bool netif_is_lag_port(co return netif_is_bond_slave(dev) || netif_is_team_port(dev); } + static inline bool netif_is_rxfh_configured(const struct net_device *dev) + { + return dev->priv_flags & IFF_RXFH_CONFIGURED; + } + /* This device needs to keep skb dst for qdisc enqueue or ndo_start_xmit() */ static inline void netif_keep_dst(struct net_device *dev) { diff --combined include/linux/skbuff.h index 4ce9ff7,89b5367..2c3dc08 --- a/include/linux/skbuff.h +++ b/include/linux/skbuff.h @@@ -299,7 -299,6 +299,7 @@@ struct sk_buff #else #define MAX_SKB_FRAGS (65536/PAGE_SIZE + 1) #endif +extern int sysctl_max_skb_frags; typedef struct skb_frag_struct skb_frag_t; @@@ -2162,6 -2161,11 +2162,11 @@@ static inline int skb_checksum_start_of return skb->csum_start - skb_headroom(skb); } + static inline unsigned char *skb_checksum_start(const struct sk_buff *skb) + { + return skb->head + skb->csum_start; + } + static inline int skb_transport_offset(const struct sk_buff *skb) { return skb_transport_header(skb) - skb->data; @@@ -2400,6 -2404,10 +2405,10 @@@ static inline struct sk_buff *napi_allo { return __napi_alloc_skb(napi, length, GFP_ATOMIC); } + void napi_consume_skb(struct sk_buff *skb, int budget); + + void __kfree_skb_flush(void); + void __kfree_skb_defer(struct sk_buff *skb); /** * __dev_alloc_pages - allocate page for network Rx @@@ -3550,6 -3558,7 +3559,7 @@@ static inline struct sec_path *skb_sec_ struct skb_gso_cb { int mac_offset; int encap_level; + __wsum csum; __u16 csum_start; }; #define SKB_SGO_CB_OFFSET 32 @@@ -3576,6 -3585,16 +3586,16 @@@ static inline int gso_pskb_expand_head( return 0; } + static inline void gso_reset_checksum(struct sk_buff *skb, __wsum res) + { + /* Do not update partial checksums if remote checksum is enabled. */ + if (skb->remcsum_offload) + return; + + SKB_GSO_CB(skb)->csum = res; + SKB_GSO_CB(skb)->csum_start = skb_checksum_start(skb) - skb->head; + } + /* Compute the checksum for a gso segment. First compute the checksum value * from the start of transport header to SKB_GSO_CB(skb)->csum_start, and * then add in skb->csum (checksum from csum_start to end of packet). @@@ -3586,15 -3605,14 +3606,14 @@@ */ static inline __sum16 gso_make_checksum(struct sk_buff *skb, __wsum res) { - int plen = SKB_GSO_CB(skb)->csum_start - skb_headroom(skb) - - skb_transport_offset(skb); - __wsum partial; + unsigned char *csum_start = skb_transport_header(skb); + int plen = (skb->head + SKB_GSO_CB(skb)->csum_start) - csum_start; + __wsum partial = SKB_GSO_CB(skb)->csum; - partial = csum_partial(skb_transport_header(skb), plen, skb->csum); - skb->csum = res; - SKB_GSO_CB(skb)->csum_start -= plen; + SKB_GSO_CB(skb)->csum = res; + SKB_GSO_CB(skb)->csum_start = csum_start - skb->head; - return csum_fold(partial); + return csum_fold(csum_partial(csum_start, plen, partial)); } static inline bool skb_is_gso(const struct sk_buff *skb) @@@ -3684,5 -3702,30 +3703,30 @@@ static inline unsigned int skb_gso_netw return hdr_len + skb_gso_transport_seglen(skb); } + /* Local Checksum Offload. + * Compute outer checksum based on the assumption that the + * inner checksum will be offloaded later. + * See Documentation/networking/checksum-offloads.txt for + * explanation of how this works. + * Fill in outer checksum adjustment (e.g. with sum of outer + * pseudo-header) before calling. + * Also ensure that inner checksum is in linear data area. + */ + static inline __wsum lco_csum(struct sk_buff *skb) + { + unsigned char *csum_start = skb_checksum_start(skb); + unsigned char *l4_hdr = skb_transport_header(skb); + __wsum partial; + + /* Start with complement of inner checksum adjustment */ + partial = ~csum_unfold(*(__force __sum16 *)(csum_start + + skb->csum_offset)); + + /* Add in checksum of our headers (incl. outer checksum + * adjustment filled in by caller) and return result. + */ + return csum_partial(l4_hdr, csum_start - l4_hdr, partial); + } + #endif /* __KERNEL__ */ #endif /* _LINUX_SKBUFF_H */ diff --combined include/net/ip_tunnels.h index dda9abf,4dd6163..5f28b60 --- a/include/net/ip_tunnels.h +++ b/include/net/ip_tunnels.h @@@ -13,6 -13,7 +13,7 @@@ #include <net/netns/generic.h> #include <net/rtnetlink.h> #include <net/lwtunnel.h> + #include <net/dst_cache.h> #if IS_ENABLED(CONFIG_IPV6) #include <net/ipv6.h> @@@ -57,6 -58,9 +58,9 @@@ struct ip_tunnel_key struct ip_tunnel_info { struct ip_tunnel_key key; + #ifdef CONFIG_DST_CACHE + struct dst_cache dst_cache; + #endif u8 options_len; u8 mode; }; @@@ -85,11 -89,6 +89,6 @@@ struct ip_tunnel_prl_entry struct rcu_head rcu_head; }; - struct ip_tunnel_dst { - struct dst_entry __rcu *dst; - __be32 saddr; - }; - struct metadata_dst; struct ip_tunnel { @@@ -108,7 -107,7 +107,7 @@@ int tun_hlen; /* Precalculated header length */ int mlink; - struct ip_tunnel_dst __percpu *dst_cache; + struct dst_cache dst_cache; struct ip_tunnel_parm parms; @@@ -230,7 -229,6 +229,7 @@@ void ip_tunnel_xmit(struct sk_buff *skb int ip_tunnel_ioctl(struct net_device *dev, struct ip_tunnel_parm *p, int cmd); int ip_tunnel_encap(struct sk_buff *skb, struct ip_tunnel *t, u8 *protocol, struct flowi4 *fl4); +int __ip_tunnel_change_mtu(struct net_device *dev, int new_mtu, bool strict); int ip_tunnel_change_mtu(struct net_device *dev, int new_mtu); struct rtnl_link_stats64 *ip_tunnel_get_stats64(struct net_device *dev, @@@ -248,7 -246,6 +247,6 @@@ int ip_tunnel_changelink(struct net_dev int ip_tunnel_newlink(struct net_device *dev, struct nlattr *tb[], struct ip_tunnel_parm *p); void ip_tunnel_setup(struct net_device *dev, int net_id); - void ip_tunnel_dst_reset_all(struct ip_tunnel *t); int ip_tunnel_encap_setup(struct ip_tunnel *t, struct ip_tunnel_encap *ipencap); @@@ -273,15 -270,15 +271,15 @@@ static inline u8 ip_tunnel_ecn_encap(u return INET_ECN_encapsulate(tos, inner); } - int iptunnel_pull_header(struct sk_buff *skb, int hdr_len, __be16 inner_proto); + int iptunnel_pull_header(struct sk_buff *skb, int hdr_len, __be16 inner_proto, + bool xnet); void iptunnel_xmit(struct sock *sk, struct rtable *rt, struct sk_buff *skb, __be32 src, __be32 dst, u8 proto, u8 tos, u8 ttl, __be16 df, bool xnet); struct metadata_dst *iptunnel_metadata_reply(struct metadata_dst *md, gfp_t flags); - struct sk_buff *iptunnel_handle_offloads(struct sk_buff *skb, bool gre_csum, - int gso_type_mask); + struct sk_buff *iptunnel_handle_offloads(struct sk_buff *skb, int gso_type_mask); static inline void iptunnel_xmit_stats(struct net_device *dev, int pkt_len) { diff --combined include/net/tcp.h index ae6468f,9b2cb0c..e90db85 --- a/include/net/tcp.h +++ b/include/net/tcp.h @@@ -239,13 -239,6 +239,6 @@@ extern struct inet_timewait_death_row t extern int sysctl_tcp_timestamps; extern int sysctl_tcp_window_scaling; extern int sysctl_tcp_sack; - extern int sysctl_tcp_fin_timeout; - extern int sysctl_tcp_syn_retries; - extern int sysctl_tcp_synack_retries; - extern int sysctl_tcp_retries1; - extern int sysctl_tcp_retries2; - extern int sysctl_tcp_orphan_retries; - extern int sysctl_tcp_syncookies; extern int sysctl_tcp_fastopen; extern int sysctl_tcp_retrans_collapse; extern int sysctl_tcp_stdurg; @@@ -274,7 -267,6 +267,6 @@@ extern int sysctl_tcp_thin_dupack extern int sysctl_tcp_early_retrans; extern int sysctl_tcp_limit_output_bytes; extern int sysctl_tcp_challenge_ack_limit; - extern unsigned int sysctl_tcp_notsent_lowat; extern int sysctl_tcp_min_tso_segs; extern int sysctl_tcp_min_rtt_wlen; extern int sysctl_tcp_autocorking; @@@ -447,7 -439,7 +439,7 @@@ const u8 *tcp_parse_md5sig_option(cons void tcp_v4_send_check(struct sock *sk, struct sk_buff *skb); void tcp_v4_mtu_reduced(struct sock *sk); -void tcp_req_err(struct sock *sk, u32 seq); +void tcp_req_err(struct sock *sk, u32 seq, bool abort); int tcp_v4_conn_request(struct sock *sk, struct sk_buff *skb); struct sock *tcp_create_openreq_child(const struct sock *sk, struct request_sock *req, @@@ -568,6 -560,7 +560,7 @@@ void tcp_rearm_rto(struct sock *sk) void tcp_synack_rtt_meas(struct sock *sk, struct request_sock *req); void tcp_reset(struct sock *sk); void tcp_skb_mark_lost_uncond_verify(struct tcp_sock *tp, struct sk_buff *skb); + void tcp_fin(struct sock *sk); /* tcp_timer.c */ void tcp_init_xmit_timers(struct sock *); @@@ -963,9 -956,11 +956,11 @@@ static inline void tcp_enable_fack(stru */ static inline void tcp_enable_early_retrans(struct tcp_sock *tp) { + struct net *net = sock_net((struct sock *)tp); + tp->do_early_retrans = sysctl_tcp_early_retrans && sysctl_tcp_early_retrans < 4 && !sysctl_tcp_thin_dupack && - sysctl_tcp_reordering == 3; + net->ipv4.sysctl_tcp_reordering == 3; } static inline void tcp_disable_early_retrans(struct tcp_sock *tp) @@@ -1252,7 -1247,7 +1247,7 @@@ static inline u32 keepalive_time_elapse static inline int tcp_fin_time(const struct sock *sk) { - int fin_timeout = tcp_sk(sk)->linger2 ? : sysctl_tcp_fin_timeout; + int fin_timeout = tcp_sk(sk)->linger2 ? : sock_net(sk)->ipv4.sysctl_tcp_fin_timeout; const int rto = inet_csk(sk)->icsk_rto; if (fin_timeout < (rto << 2) - (rto >> 1)) @@@ -1437,6 -1432,7 +1432,7 @@@ void tcp_free_fastopen_req(struct tcp_s extern struct tcp_fastopen_context __rcu *tcp_fastopen_ctx; int tcp_fastopen_reset_cipher(void *key, unsigned int len); + void tcp_fastopen_add_skb(struct sock *sk, struct sk_buff *skb); struct sock *tcp_try_fastopen(struct sock *sk, struct sk_buff *skb, struct request_sock *req, struct tcp_fastopen_cookie *foc, @@@ -1685,7 -1681,8 +1681,8 @@@ void __tcp_v4_send_check(struct sk_buf static inline u32 tcp_notsent_lowat(const struct tcp_sock *tp) { - return tp->notsent_lowat ?: sysctl_tcp_notsent_lowat; + struct net *net = sock_net((struct sock *)tp); + return tp->notsent_lowat ?: net->ipv4.sysctl_tcp_notsent_lowat; } static inline bool tcp_stream_memory_free(const struct sock *sk) diff --combined kernel/bpf/verifier.c index 2e7f7ab,42ba4cc..2fa2a7c --- a/kernel/bpf/verifier.c +++ b/kernel/bpf/verifier.c @@@ -246,6 -246,7 +246,7 @@@ static const struct {BPF_MAP_TYPE_PROG_ARRAY, BPF_FUNC_tail_call}, {BPF_MAP_TYPE_PERF_EVENT_ARRAY, BPF_FUNC_perf_event_read}, {BPF_MAP_TYPE_PERF_EVENT_ARRAY, BPF_FUNC_perf_event_output}, + {BPF_MAP_TYPE_STACK_TRACE, BPF_FUNC_get_stackid}, }; static void print_verifier_state(struct verifier_env *env) @@@ -911,8 -912,11 +912,11 @@@ static int check_map_func_compatibility * don't allow any other map type to be passed into * the special func; */ - if (bool_func && bool_map != bool_func) + if (bool_func && bool_map != bool_func) { + verbose("cannot pass map_type %d into func %d\n", + map->map_type, func_id); return -EINVAL; + } } return 0; @@@ -2082,7 -2086,7 +2086,7 @@@ static void adjust_branches(struct bpf_ /* adjust offset of jmps if necessary */ if (i < pos && i + insn->off + 1 > pos) insn->off += delta; - else if (i > pos && i + insn->off + 1 < pos) + else if (i > pos + delta && i + insn->off + 1 <= pos + delta) insn->off -= delta; } } diff --combined lib/Kconfig.debug index 8bfd1ac,f890ee5..60d09e9 --- a/lib/Kconfig.debug +++ b/lib/Kconfig.debug @@@ -1400,21 -1400,6 +1400,21 @@@ config RCU_EQS_DEBU endmenu # "RCU Debugging" +config DEBUG_WQ_FORCE_RR_CPU + bool "Force round-robin CPU selection for unbound work items" + depends on DEBUG_KERNEL + default n + help + Workqueue used to implicitly guarantee that work items queued + without explicit CPU specified are put on the local CPU. This + guarantee is no longer true and while local CPU is still + preferred work items may be put on foreign CPUs. Kernel + parameter "workqueue.debug_force_rr_cpu" is added to force + round-robin CPU selection to flush out usages which depend on the + now broken guarantee. This config option enables the debug + feature by default. When enabled, memory and cache locality will + be impacted. + config DEBUG_BLOCK_EXT_DEVT bool "Force extended block device numbers and spread them" depends on DEBUG_KERNEL @@@ -1753,6 -1738,14 +1753,14 @@@ config TEST_KSTRTO config TEST_PRINTF tristate "Test printf() family of functions at runtime" + config TEST_BITMAP + tristate "Test bitmap_*() family of functions at runtime" + default n + help + Enable this option to test the bitmap functions at boot. + + If unsure, say N. + config TEST_RHASHTABLE tristate "Perform selftest on resizable hash table" default n diff --combined net/batman-adv/gateway_client.c index ccf70be,261866e..4b598bd --- a/net/batman-adv/gateway_client.c +++ b/net/batman-adv/gateway_client.c @@@ -1,4 -1,4 +1,4 @@@ - /* Copyright (C) 2009-2015 B.A.T.M.A.N. contributors: + /* Copyright (C) 2009-2016 B.A.T.M.A.N. contributors: * * Marek Lindner * @@@ -28,6 -28,7 +28,7 @@@ #include <linux/ip.h> #include <linux/ipv6.h> #include <linux/kernel.h> + #include <linux/kref.h> #include <linux/list.h> #include <linux/netdevice.h> #include <linux/rculist.h> @@@ -59,12 -60,29 +60,29 @@@ */ #define BATADV_DHCP_CHADDR_OFFSET 28 + /** + * batadv_gw_node_release - release gw_node from lists and queue for free after + * rcu grace period + * @ref: kref pointer of the gw_node + */ + static void batadv_gw_node_release(struct kref *ref) + { + struct batadv_gw_node *gw_node; + + gw_node = container_of(ref, struct batadv_gw_node, refcount); + + batadv_orig_node_free_ref(gw_node->orig_node); + kfree_rcu(gw_node, rcu); + } + + /** + * batadv_gw_node_free_ref - decrement the gw_node refcounter and possibly + * release it + * @gw_node: gateway node to free + */ static void batadv_gw_node_free_ref(struct batadv_gw_node *gw_node) { - if (atomic_dec_and_test(&gw_node->refcount)) { - batadv_orig_node_free_ref(gw_node->orig_node); - kfree_rcu(gw_node, rcu); - } + kref_put(&gw_node->refcount, batadv_gw_node_release); } static struct batadv_gw_node * @@@ -77,7 -95,7 +95,7 @@@ batadv_gw_get_selected_gw_node(struct b if (!gw_node) goto out; - if (!atomic_inc_not_zero(&gw_node->refcount)) + if (!kref_get_unless_zero(&gw_node->refcount)) gw_node = NULL; out: @@@ -100,7 -118,7 +118,7 @@@ batadv_gw_get_selected_orig(struct bata if (!orig_node) goto unlock; - if (!atomic_inc_not_zero(&orig_node->refcount)) + if (!kref_get_unless_zero(&orig_node->refcount)) orig_node = NULL; unlock: @@@ -118,7 -136,7 +136,7 @@@ static void batadv_gw_select(struct bat spin_lock_bh(&bat_priv->gw.list_lock); - if (new_gw_node && !atomic_inc_not_zero(&new_gw_node->refcount)) + if (new_gw_node && !kref_get_unless_zero(&new_gw_node->refcount)) new_gw_node = NULL; curr_gw_node = rcu_dereference_protected(bat_priv->gw.curr_gw, 1); @@@ -170,7 -188,7 +188,7 @@@ batadv_gw_get_best_gw_node(struct batad if (!router_ifinfo) goto next; - if (!atomic_inc_not_zero(&gw_node->refcount)) + if (!kref_get_unless_zero(&gw_node->refcount)) goto next; tq_avg = router_ifinfo->bat_iv.tq_avg; @@@ -188,7 -206,7 +206,7 @@@ if (curr_gw) batadv_gw_node_free_ref(curr_gw); curr_gw = gw_node; - atomic_inc(&curr_gw->refcount); + kref_get(&curr_gw->refcount); } break; @@@ -203,7 -221,7 +221,7 @@@ if (curr_gw) batadv_gw_node_free_ref(curr_gw); curr_gw = gw_node; - atomic_inc(&curr_gw->refcount); + kref_get(&curr_gw->refcount); } break; } @@@ -423,7 -441,7 +441,7 @@@ static void batadv_gw_node_add(struct b if (gateway->bandwidth_down == 0) return; - if (!atomic_inc_not_zero(&orig_node->refcount)) + if (!kref_get_unless_zero(&orig_node->refcount)) return; gw_node = kzalloc(sizeof(*gw_node), GFP_ATOMIC); @@@ -436,7 -454,7 +454,7 @@@ gw_node->orig_node = orig_node; gw_node->bandwidth_down = ntohl(gateway->bandwidth_down); gw_node->bandwidth_up = ntohl(gateway->bandwidth_up); - atomic_set(&gw_node->refcount, 1); + kref_init(&gw_node->refcount); spin_lock_bh(&bat_priv->gw.list_lock); hlist_add_head_rcu(&gw_node->list, &bat_priv->gw.list); @@@ -456,7 -474,7 +474,7 @@@ * @bat_priv: the bat priv with all the soft interface information * @orig_node: originator announcing gateway capabilities * - * Returns gateway node if found or NULL otherwise. + * Return: gateway node if found or NULL otherwise. */ static struct batadv_gw_node * batadv_gw_node_get(struct batadv_priv *bat_priv, @@@ -469,7 -487,7 +487,7 @@@ if (gw_node_tmp->orig_node != orig_node) continue; - if (!atomic_inc_not_zero(&gw_node_tmp->refcount)) + if (!kref_get_unless_zero(&gw_node_tmp->refcount)) continue; gw_node = gw_node_tmp; @@@ -527,12 -545,11 +545,12 @@@ void batadv_gw_node_update(struct batad * gets dereferenced. */ spin_lock_bh(&bat_priv->gw.list_lock); - hlist_del_init_rcu(&gw_node->list); + if (!hlist_unhashed(&gw_node->list)) { + hlist_del_init_rcu(&gw_node->list); + batadv_gw_node_free_ref(gw_node); + } spin_unlock_bh(&bat_priv->gw.list_lock); - batadv_gw_node_free_ref(gw_node); - curr_gw = batadv_gw_get_selected_gw_node(bat_priv); if (gw_node == curr_gw) batadv_gw_reselect(bat_priv); @@@ -656,13 -673,13 +674,13 @@@ out * @chaddr: buffer where the client address will be stored. Valid * only if the function returns BATADV_DHCP_TO_CLIENT * - * Returns: + * This function may re-allocate the data buffer of the skb passed as argument. + * + * Return: * - BATADV_DHCP_NO if the packet is not a dhcp message or if there was an error * while parsing it * - BATADV_DHCP_TO_SERVER if this is a message going to the DHCP server * - BATADV_DHCP_TO_CLIENT if this is a message going to a DHCP client - * - * This function may re-allocate the data buffer of the skb passed as argument. */ enum batadv_dhcp_recipient batadv_gw_dhcp_recipient_get(struct sk_buff *skb, unsigned int *header_len, @@@ -777,11 -794,11 +795,11 @@@ * server. Due to topology changes it may be the case that the GW server * previously selected is not the best one anymore. * - * Returns true if the packet destination is unicast and it is not the best gw, - * false otherwise. - * * This call might reallocate skb data. * Must be invoked only when the DHCP packet is going TO a DHCP SERVER. + * + * Return: true if the packet destination is unicast and it is not the best gw, + * false otherwise. */ bool batadv_gw_out_of_range(struct batadv_priv *bat_priv, struct sk_buff *skb) diff --combined net/batman-adv/hard-interface.c index 57f71071,fb2d9c0..e2aaa4c --- a/net/batman-adv/hard-interface.c +++ b/net/batman-adv/hard-interface.c @@@ -1,4 -1,4 +1,4 @@@ - /* Copyright (C) 2007-2015 B.A.T.M.A.N. contributors: + /* Copyright (C) 2007-2016 B.A.T.M.A.N. contributors: * * Marek Lindner, Simon Wunderlich * @@@ -18,6 -18,7 +18,7 @@@ #include "hard-interface.h" #include "main.h" + #include <linux/atomic.h> #include <linux/bug.h> #include <linux/byteorder/generic.h> #include <linux/errno.h> @@@ -26,6 -27,7 +27,7 @@@ #include <linux/if_ether.h> #include <linux/if.h> #include <linux/kernel.h> + #include <linux/kref.h> #include <linux/list.h> #include <linux/netdevice.h> #include <linux/printk.h> @@@ -47,13 -49,19 +49,19 @@@ #include "sysfs.h" #include "translation-table.h" - void batadv_hardif_free_rcu(struct rcu_head *rcu) + /** + * batadv_hardif_release - release hard interface from lists and queue for + * free after rcu grace period + * @ref: kref pointer of the hard interface + */ + void batadv_hardif_release(struct kref *ref) { struct batadv_hard_iface *hard_iface; - hard_iface = container_of(rcu, struct batadv_hard_iface, rcu); + hard_iface = container_of(ref, struct batadv_hard_iface, refcount); dev_put(hard_iface->net_dev); - kfree(hard_iface); + + kfree_rcu(hard_iface, rcu); } struct batadv_hard_iface * @@@ -64,7 -72,7 +72,7 @@@ batadv_hardif_get_by_netdev(const struc rcu_read_lock(); list_for_each_entry_rcu(hard_iface, &batadv_hardif_list, list) { if (hard_iface->net_dev == net_dev && - atomic_inc_not_zero(&hard_iface->refcount)) + kref_get_unless_zero(&hard_iface->refcount)) goto out; } @@@ -76,28 -84,6 +84,28 @@@ out } /** + * batadv_mutual_parents - check if two devices are each others parent + * @dev1: 1st net_device + * @dev2: 2nd net_device + * + * veth devices come in pairs and each is the parent of the other! + * + * Return: true if the devices are each others parent, otherwise false + */ +static bool batadv_mutual_parents(const struct net_device *dev1, + const struct net_device *dev2) +{ + int dev1_parent_iflink = dev_get_iflink(dev1); + int dev2_parent_iflink = dev_get_iflink(dev2); + + if (!dev1_parent_iflink || !dev2_parent_iflink) + return false; + + return (dev1_parent_iflink == dev2->ifindex) && + (dev2_parent_iflink == dev1->ifindex); +} + +/** * batadv_is_on_batman_iface - check if a device is a batman iface descendant * @net_dev: the device to check * @@@ -107,7 -93,7 +115,7 @@@ * This function recursively checks all the fathers of the device passed as * argument looking for a batman-adv soft interface. * - * Returns true if the device is descendant of a batman-adv mesh interface (or + * Return: true if the device is descendant of a batman-adv mesh interface (or * if it is a batman-adv interface itself), false otherwise */ static bool batadv_is_on_batman_iface(const struct net_device *net_dev) @@@ -130,9 -116,6 +138,9 @@@ if (WARN(!parent_dev, "Cannot find parent device")) return false; + if (batadv_mutual_parents(net_dev, parent_dev)) + return false; + ret = batadv_is_on_batman_iface(parent_dev); return ret; @@@ -161,7 -144,7 +169,7 @@@ static int batadv_is_valid_iface(const * interface * @net_device: the device to check * - * Returns true if the net device is a 802.11 wireless device, false otherwise. + * Return: true if the net device is a 802.11 wireless device, false otherwise. */ bool batadv_is_wifi_netdev(struct net_device *net_device) { @@@ -194,7 -177,7 +202,7 @@@ batadv_hardif_get_active(const struct n continue; if (hard_iface->if_status == BATADV_IF_ACTIVE && - atomic_inc_not_zero(&hard_iface->refcount)) + kref_get_unless_zero(&hard_iface->refcount)) goto out; } @@@ -228,7 -211,7 +236,7 @@@ static void batadv_primary_if_select(st ASSERT_RTNL(); - if (new_hard_iface && !atomic_inc_not_zero(&new_hard_iface->refcount)) + if (new_hard_iface && !kref_get_unless_zero(&new_hard_iface->refcount)) new_hard_iface = NULL; curr_hard_iface = rcu_dereference_protected(bat_priv->primary_if, 1); @@@ -426,7 -409,8 +434,8 @@@ batadv_hardif_deactivate_interface(stru * * Invoke ndo_del_slave on master passing slave as argument. In this way slave * is free'd and master can correctly change its internal state. - * Return 0 on success, a negative value representing the error otherwise + * + * Return: 0 on success, a negative value representing the error otherwise */ static int batadv_master_del_slave(struct batadv_hard_iface *slave, struct net_device *master) @@@ -455,7 -439,7 +464,7 @@@ int batadv_hardif_enable_interface(stru if (hard_iface->if_status != BATADV_IF_NOT_IN_USE) goto out; - if (!atomic_inc_not_zero(&hard_iface->refcount)) + if (!kref_get_unless_zero(&hard_iface->refcount)) goto out; soft_iface = dev_get_by_name(&init_net, iface_name); @@@ -676,7 -660,8 +685,8 @@@ batadv_hardif_add_interface(struct net_ hard_iface->num_bcasts = BATADV_NUM_BCASTS_WIRELESS; /* extra reference for return */ - atomic_set(&hard_iface->refcount, 2); + kref_init(&hard_iface->refcount); + kref_get(&hard_iface->refcount); batadv_check_known_mac_addr(hard_iface->net_dev); list_add_tail_rcu(&hard_iface->list, &batadv_hardif_list); diff --combined net/batman-adv/translation-table.c index 0e80fd1,1188279..5c7fa02 --- a/net/batman-adv/translation-table.c +++ b/net/batman-adv/translation-table.c @@@ -1,4 -1,4 +1,4 @@@ - /* Copyright (C) 2007-2015 B.A.T.M.A.N. contributors: + /* Copyright (C) 2007-2016 B.A.T.M.A.N. contributors: * * Marek Lindner, Simon Wunderlich, Antonio Quartulli * @@@ -31,6 -31,7 +31,7 @@@ #include <linux/jhash.h> #include <linux/jiffies.h> #include <linux/kernel.h> + #include <linux/kref.h> #include <linux/list.h> #include <linux/lockdep.h> #include <linux/netdevice.h> @@@ -68,7 -69,15 +69,15 @@@ static void batadv_tt_global_del(struc unsigned short vid, const char *message, bool roaming); - /* returns 1 if they are the same mac addr and vid */ + /** + * batadv_compare_tt - check if two TT entries are the same + * @node: the list element pointer of the first TT entry + * @data2: pointer to the tt_common_entry of the second TT entry + * + * Compare the MAC address and the VLAN ID of the two TT entries and check if + * they are the same TT client. + * Return: 1 if the two TT clients are the same, 0 otherwise + */ static int batadv_compare_tt(const struct hlist_node *node, const void *data2) { const void *data1 = container_of(node, struct batadv_tt_common_entry, @@@ -84,7 -93,7 +93,7 @@@ * @data: pointer to the tt_common_entry object to map * @size: the size of the hash table * - * Returns the hash index where the object represented by 'data' should be + * Return: the hash index where the object represented by 'data' should be * stored at. */ static inline u32 batadv_choose_tt(const void *data, u32 size) @@@ -105,7 -114,7 +114,7 @@@ * @addr: the mac address of the client to look for * @vid: VLAN identifier * - * Returns a pointer to the tt_common struct belonging to the searched client if + * Return: a pointer to the tt_common struct belonging to the searched client if * found, NULL otherwise. */ static struct batadv_tt_common_entry * @@@ -133,7 -142,7 +142,7 @@@ batadv_tt_hash_find(struct batadv_hasht if (tt->vid != vid) continue; - if (!atomic_inc_not_zero(&tt->refcount)) + if (!kref_get_unless_zero(&tt->refcount)) continue; tt_tmp = tt; @@@ -150,7 -159,7 +159,7 @@@ * @addr: the mac address of the client to look for * @vid: VLAN identifier * - * Returns a pointer to the corresponding tt_local_entry struct if the client is + * Return: a pointer to the corresponding tt_local_entry struct if the client is * found, NULL otherwise. */ static struct batadv_tt_local_entry * @@@ -175,7 -184,7 +184,7 @@@ batadv_tt_local_hash_find(struct batadv * @addr: the mac address of the client to look for * @vid: VLAN identifier * - * Returns a pointer to the corresponding tt_global_entry struct if the client + * Return: a pointer to the corresponding tt_global_entry struct if the client * is found, NULL otherwise. */ static struct batadv_tt_global_entry * @@@ -194,34 -203,68 +203,68 @@@ batadv_tt_global_hash_find(struct batad return tt_global_entry; } + /** + * batadv_tt_local_entry_release - release tt_local_entry from lists and queue + * for free after rcu grace period + * @ref: kref pointer of the nc_node + */ + static void batadv_tt_local_entry_release(struct kref *ref) + { + struct batadv_tt_local_entry *tt_local_entry; + + tt_local_entry = container_of(ref, struct batadv_tt_local_entry, + common.refcount); + + kfree_rcu(tt_local_entry, common.rcu); + } + + /** + * batadv_tt_local_entry_free_ref - decrement the tt_local_entry refcounter and + * possibly release it + * @tt_local_entry: tt_local_entry to be free'd + */ static void batadv_tt_local_entry_free_ref(struct batadv_tt_local_entry *tt_local_entry) { - if (atomic_dec_and_test(&tt_local_entry->common.refcount)) - kfree_rcu(tt_local_entry, common.rcu); + kref_put(&tt_local_entry->common.refcount, + batadv_tt_local_entry_release); } /** - * batadv_tt_global_entry_free_ref - decrement the refcounter for a - * tt_global_entry and possibly free it - * @tt_global_entry: the object to free + * batadv_tt_global_entry_release - release tt_global_entry from lists and queue + * for free after rcu grace period + * @ref: kref pointer of the nc_node + */ + static void batadv_tt_global_entry_release(struct kref *ref) + { + struct batadv_tt_global_entry *tt_global_entry; + + tt_global_entry = container_of(ref, struct batadv_tt_global_entry, + common.refcount); + + batadv_tt_global_del_orig_list(tt_global_entry); + kfree_rcu(tt_global_entry, common.rcu); + } + + /** + * batadv_tt_global_entry_free_ref - decrement the tt_global_entry refcounter + * and possibly release it + * @tt_global_entry: tt_global_entry to be free'd */ static void batadv_tt_global_entry_free_ref(struct batadv_tt_global_entry *tt_global_entry) { - if (atomic_dec_and_test(&tt_global_entry->common.refcount)) { - batadv_tt_global_del_orig_list(tt_global_entry); - kfree_rcu(tt_global_entry, common.rcu); - } + kref_put(&tt_global_entry->common.refcount, + batadv_tt_global_entry_release); } /** * batadv_tt_global_hash_count - count the number of orig entries - * @hash: hash table containing the tt entries + * @bat_priv: the bat priv with all the soft interface information * @addr: the mac address of the client to count entries for * @vid: VLAN identifier * - * Return the number of originators advertising the given address/data + * Return: the number of originators advertising the given address/data * (excluding ourself). */ int batadv_tt_global_hash_count(struct batadv_priv *bat_priv, @@@ -286,9 -329,9 +329,9 @@@ static void batadv_tt_local_size_dec(st } /** - * batadv_tt_global_size_mod - change the size by v of the local table - * identified by vid - * @bat_priv: the bat priv with all the soft interface information + * batadv_tt_global_size_mod - change the size by v of the global table + * for orig_node identified by vid + * @orig_node: the originator for which the table has to be modified * @vid: the VLAN identifier * @v: the amount to sum to the global table size */ @@@ -303,11 -346,9 +346,11 @@@ static void batadv_tt_global_size_mod(s if (atomic_add_return(v, &vlan->tt.num_entries) == 0) { spin_lock_bh(&orig_node->vlan_list_lock); - hlist_del_init_rcu(&vlan->list); + if (!hlist_unhashed(&vlan->list)) { + hlist_del_init_rcu(&vlan->list); + batadv_orig_node_vlan_free_ref(vlan); + } spin_unlock_bh(&orig_node->vlan_list_lock); - batadv_orig_node_vlan_free_ref(vlan); } batadv_orig_node_vlan_free_ref(vlan); @@@ -340,22 -381,28 +383,28 @@@ static void batadv_tt_global_size_dec(s /** * batadv_tt_orig_list_entry_release - release tt orig entry from lists and * queue for free after rcu grace period - * @orig_entry: tt orig entry to be free'd + * @ref: kref pointer of the tt orig entry */ - static void - batadv_tt_orig_list_entry_release(struct batadv_tt_orig_list_entry *orig_entry) + static void batadv_tt_orig_list_entry_release(struct kref *ref) { + struct batadv_tt_orig_list_entry *orig_entry; + + orig_entry = container_of(ref, struct batadv_tt_orig_list_entry, + refcount); + batadv_orig_node_free_ref(orig_entry->orig_node); kfree_rcu(orig_entry, rcu); } + /** + * batadv_tt_orig_list_entry_free_ref - decrement the tt orig entry refcounter + * and possibly release it + * @orig_entry: tt orig entry to be free'd + */ static void batadv_tt_orig_list_entry_free_ref(struct batadv_tt_orig_list_entry *orig_entry) { - if (!atomic_dec_and_test(&orig_entry->refcount)) - return; - - batadv_tt_orig_list_entry_release(orig_entry); + kref_put(&orig_entry->refcount, batadv_tt_orig_list_entry_release); } /** @@@ -437,7 -484,7 +486,7 @@@ unlock * batadv_tt_len - compute length in bytes of given number of tt changes * @changes_num: number of tt changes * - * Returns computed length in bytes. + * Return: computed length in bytes. */ static int batadv_tt_len(int changes_num) { @@@ -448,7 -495,7 +497,7 @@@ * batadv_tt_entries - compute the number of entries fitting in tt_len bytes * @tt_len: available space * - * Returns the number of entries. + * Return: the number of entries. */ static u16 batadv_tt_entries(u16 tt_len) { @@@ -460,7 -507,7 +509,7 @@@ * size when transmitted over the air * @bat_priv: the bat priv with all the soft interface information * - * Returns local translation table size in bytes. + * Return: local translation table size in bytes. */ static int batadv_tt_local_table_transmit_size(struct batadv_priv *bat_priv) { @@@ -526,7 -573,7 +575,7 @@@ static void batadv_tt_global_free(struc * @mark: the value contained in the skb->mark field of the received packet (if * any) * - * Returns true if the client was successfully added, false otherwise. + * Return: true if the client was successfully added, false otherwise. */ bool batadv_tt_local_add(struct net_device *soft_iface, const u8 *addr, unsigned short vid, int ifindex, u32 mark) @@@ -620,7 -667,8 +669,8 @@@ tt_local->common.vid = vid; if (batadv_is_wifi_netdev(in_dev)) tt_local->common.flags |= BATADV_TT_CLIENT_WIFI; - atomic_set(&tt_local->common.refcount, 2); + kref_init(&tt_local->common.refcount); + kref_get(&tt_local->common.refcount); tt_local->last_seen = jiffies; tt_local->common.added_at = tt_local->last_seen; @@@ -721,12 -769,11 +771,11 @@@ out * function reserves the amount of space needed to send the entire global TT * table. In case of success the value is updated with the real amount of * reserved bytes - * Allocate the needed amount of memory for the entire TT TVLV and write its * header made up by one tvlv_tt_data object and a series of tvlv_tt_vlan_data * objects, one per active VLAN served by the originator node. * - * Return the size of the allocated buffer or 0 in case of failure. + * Return: the size of the allocated buffer or 0 in case of failure. */ static u16 batadv_tt_prepare_tvlv_global_data(struct batadv_orig_node *orig_node, @@@ -800,7 -847,7 +849,7 @@@ out * header made up by one tvlv_tt_data object and a series of tvlv_tt_vlan_data * objects, one per active VLAN. * - * Return the size of the allocated buffer or 0 in case of failure. + * Return: the size of the allocated buffer or 0 in case of failure. */ static u16 batadv_tt_prepare_tvlv_local_data(struct batadv_priv *bat_priv, @@@ -1042,7 -1089,7 +1091,7 @@@ batadv_tt_local_set_pending(struct bata * @message: message to append to the log on deletion * @roaming: true if the deletion is due to a roaming event * - * Returns the flags assigned to the local entry before being deleted + * Return: the flags assigned to the local entry before being deleted */ u16 batadv_tt_local_remove(struct batadv_priv *bat_priv, const u8 *addr, unsigned short vid, const char *message, @@@ -1242,10 -1289,16 +1291,16 @@@ static void batadv_tt_changes_list_free spin_unlock_bh(&bat_priv->tt.changes_list_lock); } - /* retrieves the orig_tt_list_entry belonging to orig_node from the + /** + * batadv_tt_global_orig_entry_find - find a TT orig_list_entry + * @entry: the TT global entry where the orig_list_entry has to be + * extracted from + * @orig_node: the originator for which the orig_list_entry has to be found + * + * retrieve the orig_tt_list_entry belonging to orig_node from the * batadv_tt_global_entry list * - * returns it with an increased refcounter, NULL if not found + * Return: it with an increased refcounter, NULL if not found */ static struct batadv_tt_orig_list_entry * batadv_tt_global_orig_entry_find(const struct batadv_tt_global_entry *entry, @@@ -1259,7 -1312,7 +1314,7 @@@ hlist_for_each_entry_rcu(tmp_orig_entry, head, list) { if (tmp_orig_entry->orig_node != orig_node) continue; - if (!atomic_inc_not_zero(&tmp_orig_entry->refcount)) + if (!kref_get_unless_zero(&tmp_orig_entry->refcount)) continue; orig_entry = tmp_orig_entry; @@@ -1270,8 -1323,15 +1325,15 @@@ return orig_entry; } - /* find out if an orig_node is already in the list of a tt_global_entry. - * returns true if found, false otherwise + /** + * batadv_tt_global_entry_has_orig - check if a TT global entry is also handled + * by a given originator + * @entry: the TT global entry to check + * @orig_node: the originator to search in the list + * + * find out if an orig_node is already in the list of a tt_global_entry. + * + * Return: true if found, false otherwise */ static bool batadv_tt_global_entry_has_orig(const struct batadv_tt_global_entry *entry, @@@ -1309,11 -1369,12 +1371,12 @@@ batadv_tt_global_orig_entry_add(struct goto out; INIT_HLIST_NODE(&orig_entry->list); - atomic_inc(&orig_node->refcount); + kref_get(&orig_node->refcount); batadv_tt_global_size_inc(orig_node, tt_global->common.vid); orig_entry->orig_node = orig_node; orig_entry->ttvn = ttvn; - atomic_set(&orig_entry->refcount, 2); + kref_init(&orig_entry->refcount); + kref_get(&orig_entry->refcount); spin_lock_bh(&tt_global->list_lock); hlist_add_head_rcu(&orig_entry->list, @@@ -1343,7 -1404,7 +1406,7 @@@ out * * The caller must hold orig_node refcount. * - * Return true if the new entry has been added, false otherwise + * Return: true if the new entry has been added, false otherwise */ static bool batadv_tt_global_add(struct batadv_priv *bat_priv, struct batadv_orig_node *orig_node, @@@ -1389,7 -1450,8 +1452,8 @@@ */ if (flags & BATADV_TT_CLIENT_ROAM) tt_global_entry->roam_at = jiffies; - atomic_set(&common->refcount, 2); + kref_init(&common->refcount); + kref_get(&common->refcount); common->added_at = jiffies; INIT_HLIST_HEAD(&tt_global_entry->orig_list); @@@ -1501,7 -1563,7 +1565,7 @@@ out * @tt_global_entry: global translation table entry to be analyzed * * This functon assumes the caller holds rcu_read_lock(). - * Returns best originator list entry or NULL on errors. + * Return: best originator list entry or NULL on errors. */ static struct batadv_tt_orig_list_entry * batadv_transtable_best_orig(struct batadv_priv *bat_priv, @@@ -2031,7 -2093,7 +2095,7 @@@ _batadv_is_ap_isolated(struct batadv_tt * @addr: mac address of the destination client * @vid: VLAN identifier * - * Returns a pointer to the originator that was selected as destination in the + * Return: a pointer to the originator that was selected as destination in the * mesh for contacting the client 'addr', NULL otherwise. * In case of multiple originators serving the same client, the function returns * the best one (best in terms of metric towards the destination node). @@@ -2071,7 -2133,7 +2135,7 @@@ struct batadv_orig_node *batadv_transta /* found anything? */ if (best_entry) orig_node = best_entry->orig_node; - if (orig_node && !atomic_inc_not_zero(&orig_node->refcount)) + if (orig_node && !kref_get_unless_zero(&orig_node->refcount)) orig_node = NULL; rcu_read_unlock(); @@@ -2106,7 -2168,7 +2170,7 @@@ out * because the XOR operation can combine them all while trying to reduce the * noise as much as possible. * - * Returns the checksum of the global table of a given originator. + * Return: the checksum of the global table of a given originator. */ static u32 batadv_tt_global_crc(struct batadv_priv *bat_priv, struct batadv_orig_node *orig_node, @@@ -2183,7 -2245,7 +2247,7 @@@ * For details about the computation, please refer to the documentation for * batadv_tt_global_crc(). * - * Returns the checksum of the local table + * Return: the checksum of the local table */ static u32 batadv_tt_local_crc(struct batadv_priv *bat_priv, unsigned short vid) @@@ -2289,7 -2351,7 +2353,7 @@@ static void batadv_tt_req_purge(struct * @bat_priv: the bat priv with all the soft interface information * @orig_node: orig node this request is being issued for * - * Returns the pointer to the new tt_req_node struct if no request + * Return: the pointer to the new tt_req_node struct if no request * has already been issued for this orig_node, NULL otherwise. */ static struct batadv_tt_req_node * @@@ -2324,7 -2386,7 +2388,7 @@@ unlock * @entry_ptr: to be checked local tt entry * @data_ptr: not used but definition required to satisfy the callback prototype * - * Returns 1 if the entry is a valid, 0 otherwise. + * Return: 1 if the entry is a valid, 0 otherwise. */ static int batadv_tt_local_valid(const void *entry_ptr, const void *data_ptr) { @@@ -2408,9 -2470,8 +2472,8 @@@ static void batadv_tt_tvlv_generate(str * @orig_node: originator for which the CRCs have to be checked * @tt_vlan: pointer to the first tvlv VLAN entry * @num_vlan: number of tvlv VLAN entries - * @create: if true, create VLAN objects if not found * - * Return true if all the received CRCs match the locally stored ones, false + * Return: true if all the received CRCs match the locally stored ones, false * otherwise */ static bool batadv_tt_global_check_crc(struct batadv_orig_node *orig_node, @@@ -2513,6 -2574,8 +2576,8 @@@ static void batadv_tt_global_update_crc * @num_vlan: number of tvlv VLAN entries * @full_table: ask for the entire translation table if true, while only for the * last TT diff otherwise + * + * Return: true if the TT Request was sent, false otherwise */ static int batadv_send_tt_request(struct batadv_priv *bat_priv, struct batadv_orig_node *dst_orig_node, @@@ -2593,7 -2656,7 +2658,7 @@@ out * @req_src: mac address of tt request sender * @req_dst: mac address of tt request recipient * - * Returns true if tt request reply was sent, false otherwise. + * Return: true if tt request reply was sent, false otherwise. */ static bool batadv_send_other_tt_response(struct batadv_priv *bat_priv, struct batadv_tvlv_tt_data *tt_data, @@@ -2725,7 -2788,7 +2790,7 @@@ out * @tt_data: tt data containing the tt request information * @req_src: mac address of tt request sender * - * Returns true if tt request reply was sent, false otherwise. + * Return: true if tt request reply was sent, false otherwise. */ static bool batadv_send_my_tt_response(struct batadv_priv *bat_priv, struct batadv_tvlv_tt_data *tt_data, @@@ -2843,7 -2906,7 +2908,7 @@@ out * @req_src: mac address of tt request sender * @req_dst: mac address of tt request recipient * - * Returns true if tt request reply was sent, false otherwise. + * Return: true if tt request reply was sent, false otherwise. */ static bool batadv_send_tt_response(struct batadv_priv *bat_priv, struct batadv_tvlv_tt_data *tt_data, @@@ -2938,7 -3001,7 +3003,7 @@@ static void batadv_tt_update_changes(st * @addr: the mac address of the client to check * @vid: VLAN identifier * - * Returns true if the client is served by this node, false otherwise. + * Return: true if the client is served by this node, false otherwise. */ bool batadv_is_my_client(struct batadv_priv *bat_priv, const u8 *addr, unsigned short vid) @@@ -3055,11 -3118,16 +3120,16 @@@ static void batadv_tt_roam_purge(struc spin_unlock_bh(&bat_priv->tt.roam_list_lock); } - /* This function checks whether the client already reached the + /** + * batadv_tt_check_roam_count - check if a client has roamed too frequently + * @bat_priv: the bat priv with all the soft interface information + * @client: mac address of the roaming client + * + * This function checks whether the client already reached the * maximum number of possible roaming phases. In this case the ROAMING_ADV * will not be sent. * - * returns true if the ROAMING_ADV can be sent, false otherwise + * Return: true if the ROAMING_ADV can be sent, false otherwise */ static bool batadv_tt_check_roam_count(struct batadv_priv *bat_priv, u8 *client) { @@@ -3371,13 -3439,12 +3441,12 @@@ out * batadv_tt_update_orig - update global translation table with new tt * information received via ogms * @bat_priv: the bat priv with all the soft interface information - * @orig: the orig_node of the ogm - * @tt_vlan: pointer to the first tvlv VLAN entry + * @orig_node: the orig_node of the ogm + * @tt_buff: pointer to the first tvlv VLAN entry * @tt_num_vlan: number of tvlv VLAN entries * @tt_change: pointer to the first entry in the TT buffer * @tt_num_changes: number of tt changes inside the tt buffer * @ttvn: translation table version number of this changeset - * @tt_crc: crc32 checksum of orig node's translation table */ static void batadv_tt_update_orig(struct batadv_priv *bat_priv, struct batadv_orig_node *orig_node, @@@ -3459,7 -3526,7 +3528,7 @@@ request_table * @addr: the mac address of the client to check * @vid: VLAN identifier * - * Returns true if we know that the client has moved from its old originator + * Return: true if we know that the client has moved from its old originator * to another one. This entry is still kept for consistency purposes and will be * deleted later by a DEL or because of timeout */ @@@ -3485,7 -3552,7 +3554,7 @@@ out * @addr: the mac address of the local client to query * @vid: VLAN identifier * - * Returns true if the local client is known to be roaming (it is not served by + * Return: true if the local client is known to be roaming (it is not served by * this node anymore) or not. If yes, the client is still present in the table * to keep the latter consistent with the node TTVN */ @@@ -3614,7 -3681,7 +3683,7 @@@ static void batadv_tt_tvlv_ogm_handler_ * @tvlv_value: tvlv buffer containing the tt data * @tvlv_value_len: tvlv buffer length * - * Returns NET_RX_DROP if the tt tvlv is to be re-routed, NET_RX_SUCCESS + * Return: NET_RX_DROP if the tt tvlv is to be re-routed, NET_RX_SUCCESS * otherwise. */ static int batadv_tt_tvlv_unicast_handler_v1(struct batadv_priv *bat_priv, @@@ -3695,7 -3762,7 +3764,7 @@@ * @tvlv_value: tvlv buffer containing the tt data * @tvlv_value_len: tvlv buffer length * - * Returns NET_RX_DROP if the tt roam tvlv is to be re-routed, NET_RX_SUCCESS + * Return: NET_RX_DROP if the tt roam tvlv is to be re-routed, NET_RX_SUCCESS * otherwise. */ static int batadv_roam_tvlv_unicast_handler_v1(struct batadv_priv *bat_priv, @@@ -3741,7 -3808,7 +3810,7 @@@ out * batadv_tt_init - initialise the translation table internals * @bat_priv: the bat priv with all the soft interface information * - * Return 0 on success or negative error number in case of failure. + * Return: 0 on success or negative error number in case of failure. */ int batadv_tt_init(struct batadv_priv *bat_priv) { @@@ -3779,7 -3846,7 +3848,7 @@@ * @addr: the mac address of the client * @vid: the identifier of the VLAN where this client is connected * - * Returns true if the client is marked with the TT_CLIENT_ISOLA flag, false + * Return: true if the client is marked with the TT_CLIENT_ISOLA flag, false * otherwise */ bool batadv_tt_global_is_isolated(struct batadv_priv *bat_priv, diff --combined net/bridge/br_mdb.c index 74c278e,cf51b7b..73786e2 --- a/net/bridge/br_mdb.c +++ b/net/bridge/br_mdb.c @@@ -41,6 -41,14 +41,14 @@@ fail return -EMSGSIZE; } + static void __mdb_entry_fill_flags(struct br_mdb_entry *e, unsigned char flags) + { + e->state = flags & MDB_PG_FLAGS_PERMANENT; + e->flags = 0; + if (flags & MDB_PG_FLAGS_OFFLOAD) + e->flags |= MDB_FLAGS_OFFLOAD; + } + static int br_mdb_fill_info(struct sk_buff *skb, struct netlink_callback *cb, struct net_device *dev) { @@@ -80,26 -88,41 +88,41 @@@ for (pp = &mp->ports; (p = rcu_dereference(*pp)) != NULL; pp = &p->next) { + struct nlattr *nest_ent; + struct br_mdb_entry e; + port = p->port; - if (port) { - struct br_mdb_entry e; - memset(&e, 0, sizeof(e)); - e.ifindex = port->dev->ifindex; - e.state = p->state; - e.vid = p->addr.vid; - if (p->addr.proto == htons(ETH_P_IP)) - e.addr.u.ip4 = p->addr.u.ip4; + if (!port) + continue; + + memset(&e, 0, sizeof(e)); + e.ifindex = port->dev->ifindex; + e.vid = p->addr.vid; + __mdb_entry_fill_flags(&e, p->flags); + if (p->addr.proto == htons(ETH_P_IP)) + e.addr.u.ip4 = p->addr.u.ip4; #if IS_ENABLED(CONFIG_IPV6) - if (p->addr.proto == htons(ETH_P_IPV6)) - e.addr.u.ip6 = p->addr.u.ip6; + if (p->addr.proto == htons(ETH_P_IPV6)) + e.addr.u.ip6 = p->addr.u.ip6; #endif - e.addr.proto = p->addr.proto; - if (nla_put(skb, MDBA_MDB_ENTRY_INFO, sizeof(e), &e)) { - nla_nest_cancel(skb, nest2); - err = -EMSGSIZE; - goto out; - } + e.addr.proto = p->addr.proto; + nest_ent = nla_nest_start(skb, + MDBA_MDB_ENTRY_INFO); + if (!nest_ent) { + nla_nest_cancel(skb, nest2); + err = -EMSGSIZE; + goto out; } + if (nla_put_nohdr(skb, sizeof(e), &e) || + nla_put_u32(skb, + MDBA_MDB_EATTR_TIMER, + br_timer_value(&p->timer))) { + nla_nest_cancel(skb, nest_ent); + nla_nest_cancel(skb, nest2); + err = -EMSGSIZE; + goto out; + } + nla_nest_end(skb, nest_ent); } nla_nest_end(skb, nest2); skip: @@@ -209,7 -232,7 +232,7 @@@ static inline size_t rtnl_mdb_nlmsg_siz } static void __br_mdb_notify(struct net_device *dev, struct br_mdb_entry *entry, - int type) + int type, struct net_bridge_port_group *pg) { struct switchdev_obj_port_mdb mdb = { .obj = { @@@ -232,10 -255,13 +255,13 @@@ #endif mdb.obj.orig_dev = port_dev; - if (port_dev && type == RTM_NEWMDB) - switchdev_port_obj_add(port_dev, &mdb.obj); - else if (port_dev && type == RTM_DELMDB) + if (port_dev && type == RTM_NEWMDB) { + err = switchdev_port_obj_add(port_dev, &mdb.obj); + if (!err && pg) + pg->flags |= MDB_PG_FLAGS_OFFLOAD; + } else if (port_dev && type == RTM_DELMDB) { switchdev_port_obj_del(port_dev, &mdb.obj); + } skb = nlmsg_new(rtnl_mdb_nlmsg_size(), GFP_ATOMIC); if (!skb) @@@ -253,21 -279,21 +279,21 @@@ errout rtnl_set_sk_err(net, RTNLGRP_MDB, err); } - void br_mdb_notify(struct net_device *dev, struct net_bridge_port *port, - struct br_ip *group, int type, u8 state) + void br_mdb_notify(struct net_device *dev, struct net_bridge_port_group *pg, + int type) { struct br_mdb_entry entry; memset(&entry, 0, sizeof(entry)); - entry.ifindex = port->dev->ifindex; - entry.addr.proto = group->proto; - entry.addr.u.ip4 = group->u.ip4; + entry.ifindex = pg->port->dev->ifindex; + entry.addr.proto = pg->addr.proto; + entry.addr.u.ip4 = pg->addr.u.ip4; #if IS_ENABLED(CONFIG_IPV6) - entry.addr.u.ip6 = group->u.ip6; + entry.addr.u.ip6 = pg->addr.u.ip6; #endif - entry.state = state; - entry.vid = group->vid; - __br_mdb_notify(dev, &entry, type); + entry.vid = pg->addr.vid; + __mdb_entry_fill_flags(&entry, pg->flags); + __br_mdb_notify(dev, &entry, type, pg); } static int nlmsg_populate_rtr_fill(struct sk_buff *skb, @@@ -412,7 -438,8 +438,8 @@@ static int br_mdb_parse(struct sk_buff } static int br_mdb_add_group(struct net_bridge *br, struct net_bridge_port *port, - struct br_ip *group, unsigned char state) + struct br_ip *group, unsigned char state, + struct net_bridge_port_group **pg) { struct net_bridge_mdb_entry *mp; struct net_bridge_port_group *p; @@@ -425,8 -452,8 +452,8 @@@ mp = br_mdb_ip_get(mdb, group); if (!mp) { mp = br_multicast_new_group(br, port, group); - err = PTR_ERR(mp); - if (IS_ERR(mp)) + err = PTR_ERR_OR_ZERO(mp); + if (err) return err; } @@@ -443,6 -470,7 +470,7 @@@ if (unlikely(!p)) return -ENOMEM; rcu_assign_pointer(*pp, p); + *pg = p; if (state == MDB_TEMPORARY) mod_timer(&p->timer, now + br->multicast_membership_interval); @@@ -450,7 -478,8 +478,8 @@@ } static int __br_mdb_add(struct net *net, struct net_bridge *br, - struct br_mdb_entry *entry) + struct br_mdb_entry *entry, + struct net_bridge_port_group **pg) { struct br_ip ip; struct net_device *dev; @@@ -479,7 -508,7 +508,7 @@@ #endif spin_lock_bh(&br->multicast_lock); - ret = br_mdb_add_group(br, p, &ip, entry->state); + ret = br_mdb_add_group(br, p, &ip, entry->state, pg); spin_unlock_bh(&br->multicast_lock); return ret; } @@@ -487,6 -516,7 +516,7 @@@ static int br_mdb_add(struct sk_buff *skb, struct nlmsghdr *nlh) { struct net *net = sock_net(skb->sk); + struct net_bridge_port_group *pg; struct net_bridge_vlan_group *vg; struct net_device *dev, *pdev; struct br_mdb_entry *entry; @@@ -516,15 -546,15 +546,15 @@@ if (br_vlan_enabled(br) && vg && entry->vid == 0) { list_for_each_entry(v, &vg->vlan_list, vlist) { entry->vid = v->vid; - err = __br_mdb_add(net, br, entry); + err = __br_mdb_add(net, br, entry, &pg); if (err) break; - __br_mdb_notify(dev, entry, RTM_NEWMDB); + __br_mdb_notify(dev, entry, RTM_NEWMDB, pg); } } else { - err = __br_mdb_add(net, br, entry); + err = __br_mdb_add(net, br, entry, &pg); if (!err) - __br_mdb_notify(dev, entry, RTM_NEWMDB); + __br_mdb_notify(dev, entry, RTM_NEWMDB, pg); } return err; @@@ -568,7 -598,7 +598,7 @@@ static int __br_mdb_del(struct net_brid if (p->port->state == BR_STATE_DISABLED) goto unlock; - entry->state = p->state; + __mdb_entry_fill_flags(entry, p->flags); rcu_assign_pointer(*pp, p->next); hlist_del_init(&p->mglist); del_timer(&p->timer); @@@ -620,12 -650,12 +650,12 @@@ static int br_mdb_del(struct sk_buff *s entry->vid = v->vid; err = __br_mdb_del(br, entry); if (!err) - __br_mdb_notify(dev, entry, RTM_DELMDB); + __br_mdb_notify(dev, entry, RTM_DELMDB, NULL); } } else { err = __br_mdb_del(br, entry); if (!err) - __br_mdb_notify(dev, entry, RTM_DELMDB); + __br_mdb_notify(dev, entry, RTM_DELMDB, NULL); } return err; diff --combined net/core/dev.c index 0ef061b,3f4071a..edb7179 --- a/net/core/dev.c +++ b/net/core/dev.c @@@ -3829,8 -3829,14 +3829,14 @@@ static void net_tx_action(struct softir trace_consume_skb(skb); else trace_kfree_skb(skb, net_tx_action); - __kfree_skb(skb); + + if (skb->fclone != SKB_FCLONE_UNAVAILABLE) + __kfree_skb(skb); + else + __kfree_skb_defer(skb); } + + __kfree_skb_flush(); } if (sd->output_queue) { @@@ -4154,7 -4160,10 +4160,10 @@@ ncls ret = pt_prev->func(skb, skb->dev, pt_prev, orig_dev); } else { drop: - atomic_long_inc(&skb->dev->rx_dropped); + if (!deliver_exact) + atomic_long_inc(&skb->dev->rx_dropped); + else + atomic_long_inc(&skb->dev->rx_nohandler); kfree_skb(skb); /* Jamal, now you will not able to escape explaining * me how you were going to use this. :-) @@@ -5152,6 -5161,7 +5161,7 @@@ static void net_rx_action(struct softir } } + __kfree_skb_flush(); local_irq_disable(); list_splice_tail_init(&sd->poll_list, &list); @@@ -5379,12 -5389,12 +5389,12 @@@ void *netdev_lower_get_next(struct net_ { struct netdev_adjacent *lower; - lower = list_entry((*iter)->next, struct netdev_adjacent, list); + lower = list_entry(*iter, struct netdev_adjacent, list); if (&lower->list == &dev->adj_list.lower) return NULL; - *iter = &lower->list; + *iter = lower->list.next; return lower->dev; } @@@ -7253,24 -7263,31 +7263,31 @@@ void netdev_run_todo(void } } - /* Convert net_device_stats to rtnl_link_stats64. They have the same - * fields in the same order, with only the type differing. + /* Convert net_device_stats to rtnl_link_stats64. rtnl_link_stats64 has + * all the same fields in the same order as net_device_stats, with only + * the type differing, but rtnl_link_stats64 may have additional fields + * at the end for newer counters. */ void netdev_stats_to_stats64(struct rtnl_link_stats64 *stats64, const struct net_device_stats *netdev_stats) { #if BITS_PER_LONG == 64 - BUILD_BUG_ON(sizeof(*stats64) != sizeof(*netdev_stats)); + BUILD_BUG_ON(sizeof(*stats64) < sizeof(*netdev_stats)); memcpy(stats64, netdev_stats, sizeof(*stats64)); + /* zero out counters that only exist in rtnl_link_stats64 */ + memset((char *)stats64 + sizeof(*netdev_stats), 0, + sizeof(*stats64) - sizeof(*netdev_stats)); #else - size_t i, n = sizeof(*stats64) / sizeof(u64); + size_t i, n = sizeof(*netdev_stats) / sizeof(unsigned long); const unsigned long *src = (const unsigned long *)netdev_stats; u64 *dst = (u64 *)stats64; - BUILD_BUG_ON(sizeof(*netdev_stats) / sizeof(unsigned long) != - sizeof(*stats64) / sizeof(u64)); + BUILD_BUG_ON(n > sizeof(*stats64) / sizeof(u64)); for (i = 0; i < n; i++) dst[i] = src[i]; + /* zero out counters that only exist in rtnl_link_stats64 */ + memset((char *)stats64 + n * sizeof(u64), 0, + sizeof(*stats64) - n * sizeof(u64)); #endif } EXPORT_SYMBOL(netdev_stats_to_stats64); @@@ -7300,6 -7317,7 +7317,7 @@@ struct rtnl_link_stats64 *dev_get_stats } storage->rx_dropped += atomic_long_read(&dev->rx_dropped); storage->tx_dropped += atomic_long_read(&dev->tx_dropped); + storage->rx_nohandler += atomic_long_read(&dev->rx_nohandler); return storage; } EXPORT_SYMBOL(dev_get_stats); @@@ -7422,10 -7440,8 +7440,10 @@@ struct net_device *alloc_netdev_mqs(in dev->priv_flags = IFF_XMIT_DST_RELEASE | IFF_XMIT_DST_RELEASE_PERM; setup(dev); - if (!dev->tx_queue_len) + if (!dev->tx_queue_len) { dev->priv_flags |= IFF_NO_QUEUE; + dev->tx_queue_len = 1; + } dev->num_tx_queues = txqs; dev->real_num_tx_queues = txqs; diff --combined net/core/skbuff.c index 5bf88f5,8bd4b79..488566b --- a/net/core/skbuff.c +++ b/net/core/skbuff.c @@@ -79,8 -79,6 +79,8 @@@ struct kmem_cache *skbuff_head_cache __read_mostly; static struct kmem_cache *skbuff_fclone_cache __read_mostly; +int sysctl_max_skb_frags __read_mostly = MAX_SKB_FRAGS; +EXPORT_SYMBOL(sysctl_max_skb_frags); /** * skb_panic - private function for out-of-line support @@@ -349,8 -347,16 +349,16 @@@ struct sk_buff *build_skb(void *data, u } EXPORT_SYMBOL(build_skb); + #define NAPI_SKB_CACHE_SIZE 64 + + struct napi_alloc_cache { + struct page_frag_cache page; + size_t skb_count; + void *skb_cache[NAPI_SKB_CACHE_SIZE]; + }; + static DEFINE_PER_CPU(struct page_frag_cache, netdev_alloc_cache); - static DEFINE_PER_CPU(struct page_frag_cache, napi_alloc_cache); + static DEFINE_PER_CPU(struct napi_alloc_cache, napi_alloc_cache); static void *__netdev_alloc_frag(unsigned int fragsz, gfp_t gfp_mask) { @@@ -380,9 -386,9 +388,9 @@@ EXPORT_SYMBOL(netdev_alloc_frag) static void *__napi_alloc_frag(unsigned int fragsz, gfp_t gfp_mask) { - struct page_frag_cache *nc = this_cpu_ptr(&napi_alloc_cache); + struct napi_alloc_cache *nc = this_cpu_ptr(&napi_alloc_cache); - return __alloc_page_frag(nc, fragsz, gfp_mask); + return __alloc_page_frag(&nc->page, fragsz, gfp_mask); } void *napi_alloc_frag(unsigned int fragsz) @@@ -476,7 -482,7 +484,7 @@@ EXPORT_SYMBOL(__netdev_alloc_skb) struct sk_buff *__napi_alloc_skb(struct napi_struct *napi, unsigned int len, gfp_t gfp_mask) { - struct page_frag_cache *nc = this_cpu_ptr(&napi_alloc_cache); + struct napi_alloc_cache *nc = this_cpu_ptr(&napi_alloc_cache); struct sk_buff *skb; void *data; @@@ -496,7 -502,7 +504,7 @@@ if (sk_memalloc_socks()) gfp_mask |= __GFP_MEMALLOC; - data = __alloc_page_frag(nc, len, gfp_mask); + data = __alloc_page_frag(&nc->page, len, gfp_mask); if (unlikely(!data)) return NULL; @@@ -507,7 -513,7 +515,7 @@@ } /* use OR instead of assignment to avoid clearing of bits in mask */ - if (nc->pfmemalloc) + if (nc->page.pfmemalloc) skb->pfmemalloc = 1; skb->head_frag = 1; @@@ -749,6 -755,73 +757,73 @@@ void consume_skb(struct sk_buff *skb } EXPORT_SYMBOL(consume_skb); + void __kfree_skb_flush(void) + { + struct napi_alloc_cache *nc = this_cpu_ptr(&napi_alloc_cache); + + /* flush skb_cache if containing objects */ + if (nc->skb_count) { + kmem_cache_free_bulk(skbuff_head_cache, nc->skb_count, + nc->skb_cache); + nc->skb_count = 0; + } + } + + static inline void _kfree_skb_defer(struct sk_buff *skb) + { + struct napi_alloc_cache *nc = this_cpu_ptr(&napi_alloc_cache); + + /* drop skb->head and call any destructors for packet */ + skb_release_all(skb); + + /* record skb to CPU local list */ + nc->skb_cache[nc->skb_count++] = skb; + + #ifdef CONFIG_SLUB + /* SLUB writes into objects when freeing */ + prefetchw(skb); + #endif + + /* flush skb_cache if it is filled */ + if (unlikely(nc->skb_count == NAPI_SKB_CACHE_SIZE)) { + kmem_cache_free_bulk(skbuff_head_cache, NAPI_SKB_CACHE_SIZE, + nc->skb_cache); + nc->skb_count = 0; + } + } + void __kfree_skb_defer(struct sk_buff *skb) + { + _kfree_skb_defer(skb); + } + + void napi_consume_skb(struct sk_buff *skb, int budget) + { + if (unlikely(!skb)) + return; + + /* if budget is 0 assume netpoll w/ IRQs disabled */ + if (unlikely(!budget)) { + dev_consume_skb_irq(skb); + return; + } + + if (likely(atomic_read(&skb->users) == 1)) + smp_rmb(); + else if (likely(!atomic_dec_and_test(&skb->users))) + return; + /* if reaching here SKB is ready to free */ + trace_consume_skb(skb); + + /* if SKB is a clone, don't handle this case */ + if (unlikely(skb->fclone != SKB_FCLONE_UNAVAILABLE)) { + __kfree_skb(skb); + return; + } + + _kfree_skb_defer(skb); + } + EXPORT_SYMBOL(napi_consume_skb); + /* Make sure a field is enclosed inside headers_start/headers_end section */ #define CHECK_SKB_FIELD(field) \ BUILD_BUG_ON(offsetof(struct sk_buff, field) < \ @@@ -3006,8 -3079,7 +3081,7 @@@ struct sk_buff *skb_segment(struct sk_b if (unlikely(!proto)) return ERR_PTR(-EINVAL); - csum = !head_skb->encap_hdr_csum && - !!can_checksum_protocol(features, proto); + csum = !!can_checksum_protocol(features, proto); headroom = skb_headroom(head_skb); pos = skb_headlen(head_skb); @@@ -3100,13 -3172,15 +3174,15 @@@ if (nskb->len == len + doffset) goto perform_csum_check; - if (!sg && !nskb->remcsum_offload) { - nskb->ip_summed = CHECKSUM_NONE; - nskb->csum = skb_copy_and_csum_bits(head_skb, offset, - skb_put(nskb, len), - len, 0); + if (!sg) { + if (!nskb->remcsum_offload) + nskb->ip_summed = CHECKSUM_NONE; + SKB_GSO_CB(nskb)->csum = + skb_copy_and_csum_bits(head_skb, offset, + skb_put(nskb, len), + len, 0); SKB_GSO_CB(nskb)->csum_start = - skb_headroom(nskb) + doffset; + skb_headroom(nskb) + doffset; continue; } @@@ -3172,12 -3246,19 +3248,19 @@@ skip_fraglist nskb->truesize += nskb->data_len; perform_csum_check: - if (!csum && !nskb->remcsum_offload) { - nskb->csum = skb_checksum(nskb, doffset, - nskb->len - doffset, 0); - nskb->ip_summed = CHECKSUM_NONE; + if (!csum) { + if (skb_has_shared_frag(nskb)) { + err = __skb_linearize(nskb); + if (err) + goto err; + } + if (!nskb->remcsum_offload) + nskb->ip_summed = CHECKSUM_NONE; + SKB_GSO_CB(nskb)->csum = + skb_checksum(nskb, doffset, + nskb->len - doffset, 0); SKB_GSO_CB(nskb)->csum_start = - skb_headroom(nskb) + doffset; + skb_headroom(nskb) + doffset; } } while ((offset += len) < head_skb->len); @@@ -4415,9 -4496,7 +4498,7 @@@ int skb_vlan_push(struct sk_buff *skb, skb->mac_len += VLAN_HLEN; __skb_pull(skb, offset); - if (skb->ip_summed == CHECKSUM_COMPLETE) - skb->csum = csum_add(skb->csum, csum_partial(skb->data - + (2 * ETH_ALEN), VLAN_HLEN, 0)); + skb_postpush_rcsum(skb, skb->data + (2 * ETH_ALEN), VLAN_HLEN); } __vlan_hwaccel_put_tag(skb, vlan_proto, vlan_tci); return 0; diff --combined net/dccp/ipv4.c index 902d606,1e0c600..b5672e5 --- a/net/dccp/ipv4.c +++ b/net/dccp/ipv4.c @@@ -802,7 -802,7 +802,7 @@@ static int dccp_v4_rcv(struct sk_buff * } lookup: - sk = __inet_lookup_skb(&dccp_hashinfo, skb, + sk = __inet_lookup_skb(&dccp_hashinfo, skb, __dccp_hdr_len(dh), dh->dccph_sport, dh->dccph_dport); if (!sk) { dccp_pr_debug("failed to look up flow ID in table and " @@@ -824,26 -824,26 +824,26 @@@ if (sk->sk_state == DCCP_NEW_SYN_RECV) { struct request_sock *req = inet_reqsk(sk); - struct sock *nsk = NULL; + struct sock *nsk; sk = req->rsk_listener; - if (likely(sk->sk_state == DCCP_LISTEN)) { - nsk = dccp_check_req(sk, skb, req); - } else { + if (unlikely(sk->sk_state != DCCP_LISTEN)) { inet_csk_reqsk_queue_drop_and_put(sk, req); goto lookup; } + sock_hold(sk); + nsk = dccp_check_req(sk, skb, req); if (!nsk) { reqsk_put(req); - goto discard_it; + goto discard_and_relse; } if (nsk == sk) { - sock_hold(sk); reqsk_put(req); } else if (dccp_child_process(sk, nsk, skb)) { dccp_v4_ctl_send_reset(sk, skb); - goto discard_it; + goto discard_and_relse; } else { + sock_put(sk); return 0; } } diff --combined net/dccp/ipv6.c index b8608b7,45cbe85..4663a01 --- a/net/dccp/ipv6.c +++ b/net/dccp/ipv6.c @@@ -668,7 -668,7 +668,7 @@@ static int dccp_v6_rcv(struct sk_buff * DCCP_SKB_CB(skb)->dccpd_ack_seq = dccp_hdr_ack_seq(skb); lookup: - sk = __inet6_lookup_skb(&dccp_hashinfo, skb, + sk = __inet6_lookup_skb(&dccp_hashinfo, skb, __dccp_hdr_len(dh), dh->dccph_sport, dh->dccph_dport, inet6_iif(skb)); if (!sk) { @@@ -691,26 -691,26 +691,26 @@@ if (sk->sk_state == DCCP_NEW_SYN_RECV) { struct request_sock *req = inet_reqsk(sk); - struct sock *nsk = NULL; + struct sock *nsk; sk = req->rsk_listener; - if (likely(sk->sk_state == DCCP_LISTEN)) { - nsk = dccp_check_req(sk, skb, req); - } else { + if (unlikely(sk->sk_state != DCCP_LISTEN)) { inet_csk_reqsk_queue_drop_and_put(sk, req); goto lookup; } + sock_hold(sk); + nsk = dccp_check_req(sk, skb, req); if (!nsk) { reqsk_put(req); - goto discard_it; + goto discard_and_relse; } if (nsk == sk) { - sock_hold(sk); reqsk_put(req); } else if (dccp_child_process(sk, nsk, skb)) { dccp_v6_ctl_send_reset(sk, skb); - goto discard_it; + goto discard_and_relse; } else { + sock_put(sk); return 0; } } @@@ -993,7 -993,7 +993,7 @@@ static struct proto dccp_v6_prot = .sendmsg = dccp_sendmsg, .recvmsg = dccp_recvmsg, .backlog_rcv = dccp_v6_do_rcv, - .hash = inet_hash, + .hash = inet6_hash, .unhash = inet_unhash, .accept = inet_csk_accept, .get_port = inet_csk_get_port, diff --combined net/ipv4/devinet.c index f6303b1,3d83531..29b8d3a --- a/net/ipv4/devinet.c +++ b/net/ipv4/devinet.c @@@ -1847,7 -1847,7 +1847,7 @@@ static int inet_netconf_get_devconf(str if (err < 0) goto errout; - err = EINVAL; + err = -EINVAL; if (!tb[NETCONFA_IFINDEX]) goto errout; @@@ -2185,6 -2185,8 +2185,8 @@@ static struct devinet_sysctl_table "igmpv3_unsolicited_report_interval"), DEVINET_SYSCTL_RW_ENTRY(IGNORE_ROUTES_WITH_LINKDOWN, "ignore_routes_with_linkdown"), + DEVINET_SYSCTL_RW_ENTRY(DROP_GRATUITOUS_ARP, + "drop_gratuitous_arp"), DEVINET_SYSCTL_FLUSHING_ENTRY(NOXFRM, "disable_xfrm"), DEVINET_SYSCTL_FLUSHING_ENTRY(NOPOLICY, "disable_policy"), @@@ -2192,6 -2194,8 +2194,8 @@@ "promote_secondaries"), DEVINET_SYSCTL_FLUSHING_ENTRY(ROUTE_LOCALNET, "route_localnet"), + DEVINET_SYSCTL_FLUSHING_ENTRY(DROP_UNICAST_IN_L2_MULTICAST, + "drop_unicast_in_l2_multicast"), }, }; diff --combined net/ipv4/inet_connection_sock.c index 6414891,3d28c6d..d768230 --- a/net/ipv4/inet_connection_sock.c +++ b/net/ipv4/inet_connection_sock.c @@@ -24,6 -24,7 +24,7 @@@ #include <net/tcp_states.h> #include <net/xfrm.h> #include <net/tcp.h> + #include <net/sock_reuseport.h> #ifdef INET_CSK_DEBUG const char inet_csk_timer_bug_msg[] = "inet_csk BUG: unknown timer value\n"; @@@ -67,7 -68,8 +68,8 @@@ int inet_csk_bind_conflict(const struc if ((!reuse || !sk2->sk_reuse || sk2->sk_state == TCP_LISTEN) && (!reuseport || !sk2->sk_reuseport || - (sk2->sk_state != TCP_TIME_WAIT && + rcu_access_pointer(sk->sk_reuseport_cb) || + (sk2->sk_state != TCP_TIME_WAIT && !uid_eq(uid, sock_i_uid(sk2))))) { if (!sk2->sk_rcv_saddr || !sk->sk_rcv_saddr || @@@ -89,161 -91,153 +91,153 @@@ EXPORT_SYMBOL_GPL(inet_csk_bind_conflic /* Obtain a reference to a local port for the given sock, * if snum is zero it means select any available local port. + * We try to allocate an odd port (and leave even ports for connect()) */ int inet_csk_get_port(struct sock *sk, unsigned short snum) { - struct inet_hashinfo *hashinfo = sk->sk_prot->h.hashinfo; + bool reuse = sk->sk_reuse && sk->sk_state != TCP_LISTEN; + struct inet_hashinfo *hinfo = sk->sk_prot->h.hashinfo; + int ret = 1, attempts = 5, port = snum; + int smallest_size = -1, smallest_port; struct inet_bind_hashbucket *head; - struct inet_bind_bucket *tb; - int ret, attempts = 5; struct net *net = sock_net(sk); - int smallest_size = -1, smallest_rover; + int i, low, high, attempt_half; + struct inet_bind_bucket *tb; kuid_t uid = sock_i_uid(sk); - int attempt_half = (sk->sk_reuse == SK_CAN_REUSE) ? 1 : 0; + u32 remaining, offset; - local_bh_disable(); - if (!snum) { - int remaining, rover, low, high; + if (port) { + have_port: + head = &hinfo->bhash[inet_bhashfn(net, port, + hinfo->bhash_size)]; + spin_lock_bh(&head->lock); + inet_bind_bucket_for_each(tb, &head->chain) + if (net_eq(ib_net(tb), net) && tb->port == port) + goto tb_found; + goto tb_not_found; + } again: - inet_get_local_port_range(net, &low, &high); - if (attempt_half) { - int half = low + ((high - low) >> 1); - - if (attempt_half == 1) - high = half; - else - low = half; - } - remaining = (high - low) + 1; - smallest_rover = rover = prandom_u32() % remaining + low; - - smallest_size = -1; - do { - if (inet_is_local_reserved_port(net, rover)) - goto next_nolock; - head = &hashinfo->bhash[inet_bhashfn(net, rover, - hashinfo->bhash_size)]; - spin_lock(&head->lock); - inet_bind_bucket_for_each(tb, &head->chain) - if (net_eq(ib_net(tb), net) && tb->port == rover) { - if (((tb->fastreuse > 0 && - sk->sk_reuse && - sk->sk_state != TCP_LISTEN) || - (tb->fastreuseport > 0 && - sk->sk_reuseport && - uid_eq(tb->fastuid, uid))) && - (tb->num_owners < smallest_size || smallest_size == -1)) { - smallest_size = tb->num_owners; - smallest_rover = rover; - } - if (!inet_csk(sk)->icsk_af_ops->bind_conflict(sk, tb, false)) { - snum = rover; - goto tb_found; - } - goto next; + attempt_half = (sk->sk_reuse == SK_CAN_REUSE) ? 1 : 0; + other_half_scan: + inet_get_local_port_range(net, &low, &high); + high++; /* [32768, 60999] -> [32768, 61000[ */ + if (high - low < 4) + attempt_half = 0; + if (attempt_half) { + int half = low + (((high - low) >> 2) << 1); + + if (attempt_half == 1) + high = half; + else + low = half; + } + remaining = high - low; + if (likely(remaining > 1)) + remaining &= ~1U; + + offset = prandom_u32() % remaining; + /* __inet_hash_connect() favors ports having @low parity + * We do the opposite to not pollute connect() users. + */ + offset |= 1U; + smallest_size = -1; + smallest_port = low; /* avoid compiler warning */ + + other_parity_scan: + port = low + offset; + for (i = 0; i < remaining; i += 2, port += 2) { + if (unlikely(port >= high)) + port -= remaining; + if (inet_is_local_reserved_port(net, port)) + continue; + head = &hinfo->bhash[inet_bhashfn(net, port, + hinfo->bhash_size)]; + spin_lock_bh(&head->lock); + inet_bind_bucket_for_each(tb, &head->chain) + if (net_eq(ib_net(tb), net) && tb->port == port) { + if (((tb->fastreuse > 0 && reuse) || + (tb->fastreuseport > 0 && + sk->sk_reuseport && + !rcu_access_pointer(sk->sk_reuseport_cb) && + uid_eq(tb->fastuid, uid))) && + (tb->num_owners < smallest_size || smallest_size == -1)) { + smallest_size = tb->num_owners; + smallest_port = port; } - break; - next: - spin_unlock(&head->lock); - next_nolock: - if (++rover > high) - rover = low; - } while (--remaining > 0); - - /* Exhausted local port range during search? It is not - * possible for us to be holding one of the bind hash - * locks if this test triggers, because if 'remaining' - * drops to zero, we broke out of the do/while loop at - * the top level, not from the 'break;' statement. - */ - ret = 1; - if (remaining <= 0) { - if (smallest_size != -1) { - snum = smallest_rover; - goto have_snum; - } - if (attempt_half == 1) { - /* OK we now try the upper half of the range */ - attempt_half = 2; - goto again; + if (!inet_csk(sk)->icsk_af_ops->bind_conflict(sk, tb, false)) + goto tb_found; + goto next_port; } - goto fail; - } - /* OK, here is the one we will use. HEAD is - * non-NULL and we hold it's mutex. - */ - snum = rover; - } else { - have_snum: - head = &hashinfo->bhash[inet_bhashfn(net, snum, - hashinfo->bhash_size)]; - spin_lock(&head->lock); - inet_bind_bucket_for_each(tb, &head->chain) - if (net_eq(ib_net(tb), net) && tb->port == snum) - goto tb_found; + goto tb_not_found; + next_port: + spin_unlock_bh(&head->lock); + cond_resched(); + } + + if (smallest_size != -1) { + port = smallest_port; + goto have_port; } - tb = NULL; - goto tb_not_found; + offset--; + if (!(offset & 1)) + goto other_parity_scan; + + if (attempt_half == 1) { + /* OK we now try the upper half of the range */ + attempt_half = 2; + goto other_half_scan; + } + return ret; + + tb_not_found: + tb = inet_bind_bucket_create(hinfo->bind_bucket_cachep, + net, head, port); + if (!tb) + goto fail_unlock; tb_found: if (!hlist_empty(&tb->owners)) { if (sk->sk_reuse == SK_FORCE_REUSE) goto success; - if (((tb->fastreuse > 0 && - sk->sk_reuse && sk->sk_state != TCP_LISTEN) || + if (((tb->fastreuse > 0 && reuse) || (tb->fastreuseport > 0 && sk->sk_reuseport && uid_eq(tb->fastuid, uid))) && - smallest_size == -1) { + smallest_size == -1) goto success; - } else { - ret = 1; - if (inet_csk(sk)->icsk_af_ops->bind_conflict(sk, tb, true)) { - if (((sk->sk_reuse && sk->sk_state != TCP_LISTEN) || - (tb->fastreuseport > 0 && - sk->sk_reuseport && uid_eq(tb->fastuid, uid))) && - smallest_size != -1 && --attempts >= 0) { - spin_unlock(&head->lock); - goto again; - } - - goto fail_unlock; + if (inet_csk(sk)->icsk_af_ops->bind_conflict(sk, tb, true)) { + if ((reuse || + (tb->fastreuseport > 0 && + sk->sk_reuseport && + !rcu_access_pointer(sk->sk_reuseport_cb) && + uid_eq(tb->fastuid, uid))) && + smallest_size != -1 && --attempts >= 0) { + spin_unlock_bh(&head->lock); + goto again; } + goto fail_unlock; } - } - tb_not_found: - ret = 1; - if (!tb && (tb = inet_bind_bucket_create(hashinfo->bind_bucket_cachep, - net, head, snum)) == NULL) - goto fail_unlock; - if (hlist_empty(&tb->owners)) { - if (sk->sk_reuse && sk->sk_state != TCP_LISTEN) - tb->fastreuse = 1; - else + if (!reuse) tb->fastreuse = 0; + if (!sk->sk_reuseport || !uid_eq(tb->fastuid, uid)) + tb->fastreuseport = 0; + } else { + tb->fastreuse = reuse; if (sk->sk_reuseport) { tb->fastreuseport = 1; tb->fastuid = uid; - } else - tb->fastreuseport = 0; - } else { - if (tb->fastreuse && - (!sk->sk_reuse || sk->sk_state == TCP_LISTEN)) - tb->fastreuse = 0; - if (tb->fastreuseport && - (!sk->sk_reuseport || !uid_eq(tb->fastuid, uid))) + } else { tb->fastreuseport = 0; + } } success: if (!inet_csk(sk)->icsk_bind_hash) - inet_bind_hash(sk, tb, snum); + inet_bind_hash(sk, tb, port); WARN_ON(inet_csk(sk)->icsk_bind_hash != tb); ret = 0; fail_unlock: - spin_unlock(&head->lock); - fail: - local_bh_enable(); + spin_unlock_bh(&head->lock); return ret; } EXPORT_SYMBOL_GPL(inet_csk_get_port); @@@ -482,10 -476,6 +476,6 @@@ EXPORT_SYMBOL_GPL(inet_csk_route_child_ #define AF_INET_FAMILY(fam) true #endif - /* Only thing we need from tcp.h */ - extern int sysctl_tcp_synack_retries; - - /* Decide when to expire the request and when to resend SYN-ACK */ static inline void syn_ack_recalc(struct request_sock *req, const int thresh, const int max_retries, @@@ -557,6 -547,7 +547,7 @@@ static void reqsk_timer_handler(unsigne { struct request_sock *req = (struct request_sock *)data; struct sock *sk_listener = req->rsk_listener; + struct net *net = sock_net(sk_listener); struct inet_connection_sock *icsk = inet_csk(sk_listener); struct request_sock_queue *queue = &icsk->icsk_accept_queue; int qlen, expire = 0, resend = 0; @@@ -566,7 -557,7 +557,7 @@@ if (sk_state_load(sk_listener) != TCP_LISTEN) goto drop; - max_retries = icsk->icsk_syn_retries ? : sysctl_tcp_synack_retries; + max_retries = icsk->icsk_syn_retries ? : net->ipv4.sysctl_tcp_synack_retries; thresh = max_retries; /* Normally all the openreqs are young and become mature * (i.e. converted to established socket) for first timeout. @@@ -737,6 -728,7 +728,7 @@@ int inet_csk_listen_start(struct sock * { struct inet_connection_sock *icsk = inet_csk(sk); struct inet_sock *inet = inet_sk(sk); + int err = -EADDRINUSE; reqsk_queue_alloc(&icsk->icsk_accept_queue); @@@ -754,13 -746,14 +746,14 @@@ inet->inet_sport = htons(inet->inet_num); sk_dst_reset(sk); - sk->sk_prot->hash(sk); + err = sk->sk_prot->hash(sk); - return 0; + if (likely(!err)) + return 0; } sk->sk_state = TCP_CLOSE; - return -EADDRINUSE; + return err; } EXPORT_SYMBOL_GPL(inet_csk_listen_start); @@@ -789,16 -782,14 +782,16 @@@ static void inet_child_forget(struct so reqsk_put(req); } -void inet_csk_reqsk_queue_add(struct sock *sk, struct request_sock *req, - struct sock *child) +struct sock *inet_csk_reqsk_queue_add(struct sock *sk, + struct request_sock *req, + struct sock *child) { struct request_sock_queue *queue = &inet_csk(sk)->icsk_accept_queue; spin_lock(&queue->rskq_lock); if (unlikely(sk->sk_state != TCP_LISTEN)) { inet_child_forget(sk, req, child); + child = NULL; } else { req->sk = child; req->dl_next = NULL; @@@ -810,7 -801,6 +803,7 @@@ sk_acceptq_added(sk); } spin_unlock(&queue->rskq_lock); + return child; } EXPORT_SYMBOL(inet_csk_reqsk_queue_add); @@@ -820,8 -810,11 +813,8 @@@ struct sock *inet_csk_complete_hashdanc if (own_req) { inet_csk_reqsk_queue_drop(sk, req); reqsk_queue_removed(&inet_csk(sk)->icsk_accept_queue, req); - inet_csk_reqsk_queue_add(sk, req, child); - /* Warning: caller must not call reqsk_put(req); - * child stole last reference on it. - */ - return child; + if (inet_csk_reqsk_queue_add(sk, req, child)) + return child; } /* Too bad, another child took ownership of the request, undo. */ bh_unlock_sock(child); diff --combined net/ipv4/ip_gre.c index 41ba68d,12071e2..202437d --- a/net/ipv4/ip_gre.c +++ b/net/ipv4/ip_gre.c @@@ -238,7 -238,7 +238,7 @@@ static int parse_gre_header(struct sk_b return -EINVAL; } } - return iptunnel_pull_header(skb, hdr_len, tpi->proto); + return iptunnel_pull_header(skb, hdr_len, tpi->proto, false); } static void ipgre_err(struct sk_buff *skb, u32 info, @@@ -440,6 -440,17 +440,17 @@@ drop return 0; } + static __sum16 gre_checksum(struct sk_buff *skb) + { + __wsum csum; + + if (skb->ip_summed == CHECKSUM_PARTIAL) + csum = lco_csum(skb); + else + csum = skb_checksum(skb, 0, skb->len, 0); + return csum_fold(csum); + } + static void build_header(struct sk_buff *skb, int hdr_len, __be16 flags, __be16 proto, __be32 key, __be32 seq) { @@@ -467,8 -478,7 +478,7 @@@ !(skb_shinfo(skb)->gso_type & (SKB_GSO_GRE | SKB_GSO_GRE_CSUM))) { *ptr = 0; - *(__sum16 *)ptr = csum_fold(skb_checksum(skb, 0, - skb->len, 0)); + *(__sum16 *)ptr = gre_checksum(skb); } } } @@@ -493,8 -503,7 +503,7 @@@ static void __gre_xmit(struct sk_buff * static struct sk_buff *gre_handle_offloads(struct sk_buff *skb, bool csum) { - return iptunnel_handle_offloads(skb, csum, - csum ? SKB_GSO_GRE_CSUM : SKB_GSO_GRE); + return iptunnel_handle_offloads(skb, csum ? SKB_GSO_GRE_CSUM : SKB_GSO_GRE); } static struct rtable *gre_get_rt(struct sk_buff *skb, @@@ -531,9 -540,16 +540,16 @@@ static void gre_fb_xmit(struct sk_buff goto err_free_skb; key = &tun_info->key; - rt = gre_get_rt(skb, dev, &fl, key); - if (IS_ERR(rt)) - goto err_free_skb; + rt = !skb->mark ? dst_cache_get_ip4(&tun_info->dst_cache, &fl.saddr) : + NULL; + if (!rt) { + rt = gre_get_rt(skb, dev, &fl, key); + if (IS_ERR(rt)) + goto err_free_skb; + if (!skb->mark) + dst_cache_set_ip4(&tun_info->dst_cache, &rt->dst, + fl.saddr); + } tunnel_hlen = ip_gre_calc_hlen(key->tun_flags); @@@ -1054,9 -1070,8 +1070,9 @@@ static const struct net_device_ops gre_ static void ipgre_tap_setup(struct net_device *dev) { ether_setup(dev); - dev->netdev_ops = &gre_tap_netdev_ops; - dev->priv_flags |= IFF_LIVE_ADDR_CHANGE; + dev->netdev_ops = &gre_tap_netdev_ops; + dev->priv_flags &= ~IFF_TX_SKB_SHARING; + dev->priv_flags |= IFF_LIVE_ADDR_CHANGE; ip_tunnel_setup(dev, gre_tap_net_id); } @@@ -1241,14 -1256,6 +1257,14 @@@ struct net_device *gretap_fb_dev_create err = ipgre_newlink(net, dev, tb, NULL); if (err < 0) goto out; + + /* openvswitch users expect packet sizes to be unrestricted, + * so set the largest MTU we can. + */ + err = __ip_tunnel_change_mtu(dev, IP_MAX_MTU, false); + if (err) + goto out; + return dev; out: free_netdev(dev); diff --combined net/ipv4/ip_sockglue.c index a501242,3f1befc..035ad64 --- a/net/ipv4/ip_sockglue.c +++ b/net/ipv4/ip_sockglue.c @@@ -249,8 -249,6 +249,8 @@@ int ip_cmsg_send(struct net *net, struc switch (cmsg->cmsg_type) { case IP_RETOPTS: err = cmsg->cmsg_len - CMSG_ALIGN(sizeof(struct cmsghdr)); + + /* Our caller is responsible for freeing ipc->opt */ err = ip_options_get(net, &ipc->opt, CMSG_DATA(cmsg), err < 40 ? err : 40); if (err) @@@ -573,6 -571,7 +573,7 @@@ static int do_ip_setsockopt(struct soc int optname, char __user *optval, unsigned int optlen) { struct inet_sock *inet = inet_sk(sk); + struct net *net = sock_net(sk); int val = 0, err; bool needs_rtnl = setsockopt_needs_rtnl(optname); @@@ -912,7 -911,7 +913,7 @@@ } /* numsrc >= (1G-4) overflow in 32 bits */ if (msf->imsf_numsrc >= 0x3ffffffcU || - msf->imsf_numsrc > sysctl_igmp_max_msf) { + msf->imsf_numsrc > net->ipv4.sysctl_igmp_max_msf) { kfree(msf); err = -ENOBUFS; break; @@@ -1067,7 -1066,7 +1068,7 @@@ /* numsrc >= (4G-140)/128 overflow in 32 bits */ if (gsf->gf_numsrc >= 0x1ffffff || - gsf->gf_numsrc > sysctl_igmp_max_msf) { + gsf->gf_numsrc > net->ipv4.sysctl_igmp_max_msf) { err = -ENOBUFS; goto mc_msf_out; } @@@ -1342,10 -1341,13 +1343,13 @@@ static int do_ip_getsockopt(struct soc val = inet->tos; break; case IP_TTL: + { + struct net *net = sock_net(sk); val = (inet->uc_ttl == -1 ? - sysctl_ip_default_ttl : + net->ipv4.sysctl_ip_default_ttl : inet->uc_ttl); break; + } case IP_HDRINCL: val = inet->hdrincl; break; diff --combined net/ipv4/ip_tunnel.c index 89e8861,4569da7..dff8a05 --- a/net/ipv4/ip_tunnel.c +++ b/net/ipv4/ip_tunnel.c @@@ -68,61 -68,6 +68,6 @@@ static unsigned int ip_tunnel_hash(__be IP_TNL_HASH_BITS); } - static void __tunnel_dst_set(struct ip_tunnel_dst *idst, - struct dst_entry *dst, __be32 saddr) - { - struct dst_entry *old_dst; - - dst_clone(dst); - old_dst = xchg((__force struct dst_entry **)&idst->dst, dst); - dst_release(old_dst); - idst->saddr = saddr; - } - - static noinline void tunnel_dst_set(struct ip_tunnel *t, - struct dst_entry *dst, __be32 saddr) - { - __tunnel_dst_set(raw_cpu_ptr(t->dst_cache), dst, saddr); - } - - static void tunnel_dst_reset(struct ip_tunnel *t) - { - tunnel_dst_set(t, NULL, 0); - } - - void ip_tunnel_dst_reset_all(struct ip_tunnel *t) - { - int i; - - for_each_possible_cpu(i) - __tunnel_dst_set(per_cpu_ptr(t->dst_cache, i), NULL, 0); - } - EXPORT_SYMBOL(ip_tunnel_dst_reset_all); - - static struct rtable *tunnel_rtable_get(struct ip_tunnel *t, - u32 cookie, __be32 *saddr) - { - struct ip_tunnel_dst *idst; - struct dst_entry *dst; - - rcu_read_lock(); - idst = raw_cpu_ptr(t->dst_cache); - dst = rcu_dereference(idst->dst); - if (dst && !atomic_inc_not_zero(&dst->__refcnt)) - dst = NULL; - if (dst) { - if (!dst->obsolete || dst->ops->check(dst, cookie)) { - *saddr = idst->saddr; - } else { - tunnel_dst_reset(t); - dst_release(dst); - dst = NULL; - } - } - rcu_read_unlock(); - return (struct rtable *)dst; - } - static bool ip_tunnel_key_match(const struct ip_tunnel_parm *p, __be16 flags, __be32 key) { @@@ -381,7 -326,8 +326,8 @@@ static int ip_tunnel_bind_dev(struct ne if (!IS_ERR(rt)) { tdev = rt->dst.dev; - tunnel_dst_set(tunnel, &rt->dst, fl4.saddr); + dst_cache_set_ip4(&tunnel->dst_cache, &rt->dst, + fl4.saddr); ip_rt_put(rt); } if (dev->type != ARPHRD_ETHER) @@@ -729,7 -675,8 +675,8 @@@ void ip_tunnel_xmit(struct sk_buff *skb if (ip_tunnel_encap(skb, tunnel, &protocol, &fl4) < 0) goto tx_error; - rt = connected ? tunnel_rtable_get(tunnel, 0, &fl4.saddr) : NULL; + rt = connected ? dst_cache_get_ip4(&tunnel->dst_cache, &fl4.saddr) : + NULL; if (!rt) { rt = ip_route_output_key(tunnel->net, &fl4); @@@ -739,7 -686,8 +686,8 @@@ goto tx_error; } if (connected) - tunnel_dst_set(tunnel, &rt->dst, fl4.saddr); + dst_cache_set_ip4(&tunnel->dst_cache, &rt->dst, + fl4.saddr); } if (rt->dst.dev == dev) { @@@ -836,7 -784,7 +784,7 @@@ static void ip_tunnel_update(struct ip_ if (set_mtu) dev->mtu = mtu; } - ip_tunnel_dst_reset_all(t); + dst_cache_reset(&t->dst_cache); netdev_state_change(dev); } @@@ -943,31 -891,17 +891,31 @@@ done } EXPORT_SYMBOL_GPL(ip_tunnel_ioctl); -int ip_tunnel_change_mtu(struct net_device *dev, int new_mtu) +int __ip_tunnel_change_mtu(struct net_device *dev, int new_mtu, bool strict) { struct ip_tunnel *tunnel = netdev_priv(dev); int t_hlen = tunnel->hlen + sizeof(struct iphdr); + int max_mtu = 0xFFF8 - dev->hard_header_len - t_hlen; - if (new_mtu < 68 || - new_mtu > 0xFFF8 - dev->hard_header_len - t_hlen) + if (new_mtu < 68) return -EINVAL; + + if (new_mtu > max_mtu) { + if (strict) + return -EINVAL; + + new_mtu = max_mtu; + } + dev->mtu = new_mtu; return 0; } +EXPORT_SYMBOL_GPL(__ip_tunnel_change_mtu); + +int ip_tunnel_change_mtu(struct net_device *dev, int new_mtu) +{ + return __ip_tunnel_change_mtu(dev, new_mtu, true); +} EXPORT_SYMBOL_GPL(ip_tunnel_change_mtu); static void ip_tunnel_dev_free(struct net_device *dev) @@@ -975,7 -909,7 +923,7 @@@ struct ip_tunnel *tunnel = netdev_priv(dev); gro_cells_destroy(&tunnel->gro_cells); - free_percpu(tunnel->dst_cache); + dst_cache_destroy(&tunnel->dst_cache); free_percpu(dev->tstats); free_netdev(dev); } @@@ -1169,15 -1103,15 +1117,15 @@@ int ip_tunnel_init(struct net_device *d if (!dev->tstats) return -ENOMEM; - tunnel->dst_cache = alloc_percpu(struct ip_tunnel_dst); - if (!tunnel->dst_cache) { + err = dst_cache_init(&tunnel->dst_cache, GFP_KERNEL); + if (err) { free_percpu(dev->tstats); - return -ENOMEM; + return err; } err = gro_cells_init(&tunnel->gro_cells, dev); if (err) { - free_percpu(tunnel->dst_cache); + dst_cache_destroy(&tunnel->dst_cache); free_percpu(dev->tstats); return err; } @@@ -1207,7 -1141,7 +1155,7 @@@ void ip_tunnel_uninit(struct net_devic if (itn->fb_tunnel_dev != dev) ip_tunnel_del(itn, netdev_priv(dev)); - ip_tunnel_dst_reset_all(tunnel); + dst_cache_reset(&tunnel->dst_cache); } EXPORT_SYMBOL_GPL(ip_tunnel_uninit); diff --combined net/ipv4/ping.c index d3a2716,f6f93fc..76dce90 --- a/net/ipv4/ping.c +++ b/net/ipv4/ping.c @@@ -145,10 -145,12 +145,12 @@@ fail } EXPORT_SYMBOL_GPL(ping_get_port); - void ping_hash(struct sock *sk) + int ping_hash(struct sock *sk) { pr_debug("ping_hash(sk->port=%u)\n", inet_sk(sk)->inet_num); BUG(); /* "Please do not press this button again." */ + + return 0; } void ping_unhash(struct sock *sk) @@@ -746,10 -748,8 +748,10 @@@ static int ping_v4_sendmsg(struct sock if (msg->msg_controllen) { err = ip_cmsg_send(sock_net(sk), msg, &ipc, false); - if (err) + if (unlikely(err)) { + kfree(ipc.opt); return err; + } if (ipc.opt) free = 1; } diff --combined net/ipv4/raw.c index 7113bae,d635251..8d22de7 --- a/net/ipv4/raw.c +++ b/net/ipv4/raw.c @@@ -93,7 -93,7 +93,7 @@@ static struct raw_hashinfo raw_v4_hashi .lock = __RW_LOCK_UNLOCKED(raw_v4_hashinfo.lock), }; - void raw_hash_sk(struct sock *sk) + int raw_hash_sk(struct sock *sk) { struct raw_hashinfo *h = sk->sk_prot->h.raw_hash; struct hlist_head *head; @@@ -104,6 -104,8 +104,8 @@@ sk_add_node(sk, head); sock_prot_inuse_add(sock_net(sk), sk->sk_prot, 1); write_unlock_bh(&h->lock); + + return 0; } EXPORT_SYMBOL_GPL(raw_hash_sk); @@@ -547,10 -549,8 +549,10 @@@ static int raw_sendmsg(struct sock *sk if (msg->msg_controllen) { err = ip_cmsg_send(net, msg, &ipc, false); - if (err) + if (unlikely(err)) { + kfree(ipc.opt); goto out; + } if (ipc.opt) free = 1; } diff --combined net/ipv4/tcp.c index 483ffdf,f93150d..f9faadb --- a/net/ipv4/tcp.c +++ b/net/ipv4/tcp.c @@@ -282,8 -282,6 +282,6 @@@ #include <asm/unaligned.h> #include <net/busy_poll.h> - int sysctl_tcp_fin_timeout __read_mostly = TCP_FIN_TIMEOUT; - int sysctl_tcp_min_tso_segs __read_mostly = 2; int sysctl_tcp_autocorking __read_mostly = 1; @@@ -406,7 -404,7 +404,7 @@@ void tcp_init_sock(struct sock *sk tp->mss_cache = TCP_MSS_DEFAULT; u64_stats_init(&tp->syncp); - tp->reordering = sysctl_tcp_reordering; + tp->reordering = sock_net(sk)->ipv4.sysctl_tcp_reordering; tcp_enable_early_retrans(tp); tcp_assign_congestion_control(sk); @@@ -940,7 -938,7 +938,7 @@@ new_segment i = skb_shinfo(skb)->nr_frags; can_coalesce = skb_can_coalesce(skb, i, page, offset); - if (!can_coalesce && i >= MAX_SKB_FRAGS) { + if (!can_coalesce && i >= sysctl_max_skb_frags) { tcp_mark_push(tp, skb); goto new_segment; } @@@ -1213,7 -1211,7 +1211,7 @@@ new_segment if (!skb_can_coalesce(skb, i, pfrag->page, pfrag->offset)) { - if (i == MAX_SKB_FRAGS || !sg) { + if (i == sysctl_max_skb_frags || !sg) { tcp_mark_push(tp, skb); goto new_segment; } @@@ -1466,8 -1464,10 +1464,10 @@@ static struct sk_buff *tcp_recv_skb(str while ((skb = skb_peek(&sk->sk_receive_queue)) != NULL) { offset = seq - TCP_SKB_CB(skb)->seq; - if (TCP_SKB_CB(skb)->tcp_flags & TCPHDR_SYN) + if (unlikely(TCP_SKB_CB(skb)->tcp_flags & TCPHDR_SYN)) { + pr_err_once("%s: found a SYN, please report !\n", __func__); offset--; + } if (offset < skb->len || (TCP_SKB_CB(skb)->tcp_flags & TCPHDR_FIN)) { *off = offset; return skb; @@@ -1657,8 -1657,10 +1657,10 @@@ int tcp_recvmsg(struct sock *sk, struc break; offset = *seq - TCP_SKB_CB(skb)->seq; - if (TCP_SKB_CB(skb)->tcp_flags & TCPHDR_SYN) + if (unlikely(TCP_SKB_CB(skb)->tcp_flags & TCPHDR_SYN)) { + pr_err_once("%s: found a SYN, please report !\n", __func__); offset--; + } if (offset < skb->len) goto found_ok_skb; if (TCP_SKB_CB(skb)->tcp_flags & TCPHDR_FIN) @@@ -2326,6 -2328,7 +2328,7 @@@ static int do_tcp_setsockopt(struct soc { struct tcp_sock *tp = tcp_sk(sk); struct inet_connection_sock *icsk = inet_csk(sk); + struct net *net = sock_net(sk); int val; int err = 0; @@@ -2522,7 -2525,7 +2525,7 @@@ case TCP_LINGER2: if (val < 0) tp->linger2 = -1; - else if (val > sysctl_tcp_fin_timeout / HZ) + else if (val > net->ipv4.sysctl_tcp_fin_timeout / HZ) tp->linger2 = 0; else tp->linger2 = val * HZ; @@@ -2639,6 -2642,7 +2642,7 @@@ void tcp_get_info(struct sock *sk, stru const struct inet_connection_sock *icsk = inet_csk(sk); u32 now = tcp_time_stamp; unsigned int start; + int notsent_bytes; u64 rate64; u32 rate; @@@ -2719,6 -2723,11 +2723,11 @@@ } while (u64_stats_fetch_retry_irq(&tp->syncp, start)); info->tcpi_segs_out = tp->segs_out; info->tcpi_segs_in = tp->segs_in; + + notsent_bytes = READ_ONCE(tp->write_seq) - READ_ONCE(tp->snd_nxt); + info->tcpi_notsent_bytes = max(0, notsent_bytes); + + info->tcpi_min_rtt = tcp_min_rtt(tp); } EXPORT_SYMBOL_GPL(tcp_get_info); @@@ -2727,6 -2736,7 +2736,7 @@@ static int do_tcp_getsockopt(struct soc { struct inet_connection_sock *icsk = inet_csk(sk); struct tcp_sock *tp = tcp_sk(sk); + struct net *net = sock_net(sk); int val, len; if (get_user(len, optlen)) @@@ -2761,12 -2771,12 +2771,12 @@@ val = keepalive_probes(tp); break; case TCP_SYNCNT: - val = icsk->icsk_syn_retries ? : sysctl_tcp_syn_retries; + val = icsk->icsk_syn_retries ? : net->ipv4.sysctl_tcp_syn_retries; break; case TCP_LINGER2: val = tp->linger2; if (val >= 0) - val = (val ? : sysctl_tcp_fin_timeout) / HZ; + val = (val ? : net->ipv4.sysctl_tcp_fin_timeout) / HZ; break; case TCP_DEFER_ACCEPT: val = retrans_to_secs(icsk->icsk_accept_queue.rskq_defer_accept, @@@ -2950,7 -2960,7 +2960,7 @@@ static void __tcp_alloc_md5sig_pool(voi struct crypto_hash *hash; hash = crypto_alloc_hash("md5", 0, CRYPTO_ALG_ASYNC); - if (IS_ERR_OR_NULL(hash)) + if (IS_ERR(hash)) return; per_cpu(tcp_md5sig_pool, cpu).md5_desc.tfm = hash; } diff --combined net/ipv4/tcp_input.c index 3b2c8e9,5ee6fe0..e6e65f7 --- a/net/ipv4/tcp_input.c +++ b/net/ipv4/tcp_input.c @@@ -80,9 -80,7 +80,7 @@@ int sysctl_tcp_timestamps __read_mostl int sysctl_tcp_window_scaling __read_mostly = 1; int sysctl_tcp_sack __read_mostly = 1; int sysctl_tcp_fack __read_mostly = 1; - int sysctl_tcp_reordering __read_mostly = TCP_FASTRETRANS_THRESH; int sysctl_tcp_max_reordering __read_mostly = 300; - EXPORT_SYMBOL(sysctl_tcp_reordering); int sysctl_tcp_dsack __read_mostly = 1; int sysctl_tcp_app_win __read_mostly = 31; int sysctl_tcp_adv_win_scale __read_mostly = 1; @@@ -126,6 -124,10 +124,10 @@@ int sysctl_tcp_invalid_ratelimit __read #define TCP_REMNANT (TCP_FLAG_FIN|TCP_FLAG_URG|TCP_FLAG_SYN|TCP_FLAG_PSH) #define TCP_HP_BITS (~(TCP_RESERVED_BITS|TCP_FLAG_PSH)) + #define REXMIT_NONE 0 /* no loss recovery to do */ + #define REXMIT_LOST 1 /* retransmit packets marked lost */ + #define REXMIT_NEW 2 /* FRTO-style transmit of unsent/new packets */ + /* Adapt the MSS value used to make delayed ack decision to the * real world. */ @@@ -1210,6 -1212,7 +1212,7 @@@ static u8 tcp_sacktag_one(struct sock * sacked |= TCPCB_SACKED_ACKED; state->flag |= FLAG_DATA_SACKED; tp->sacked_out += pcount; + tp->delivered += pcount; /* Out-of-order packets delivered */ fack_count += pcount; @@@ -1821,8 -1824,12 +1824,12 @@@ static void tcp_check_reno_reordering(s static void tcp_add_reno_sack(struct sock *sk) { struct tcp_sock *tp = tcp_sk(sk); + u32 prior_sacked = tp->sacked_out; + tp->sacked_out++; tcp_check_reno_reordering(sk, 0); + if (tp->sacked_out > prior_sacked) + tp->delivered++; /* Some out-of-order packet is delivered */ tcp_verify_left_out(tp); } @@@ -1834,6 -1841,7 +1841,7 @@@ static void tcp_remove_reno_sacks(struc if (acked > 0) { /* One ACK acked hole. The rest eat duplicate ACKs. */ + tp->delivered += max_t(int, acked - tp->sacked_out, 1); if (acked - 1 >= tp->sacked_out) tp->sacked_out = 0; else @@@ -1873,6 -1881,7 +1881,7 @@@ void tcp_enter_loss(struct sock *sk { const struct inet_connection_sock *icsk = inet_csk(sk); struct tcp_sock *tp = tcp_sk(sk); + struct net *net = sock_net(sk); struct sk_buff *skb; bool new_recovery = icsk->icsk_ca_state < TCP_CA_Recovery; bool is_reneg; /* is receiver reneging on SACKs? */ @@@ -1923,9 -1932,9 +1932,9 @@@ * suggests that the degree of reordering is over-estimated. */ if (icsk->icsk_ca_state <= TCP_CA_Disorder && - tp->sacked_out >= sysctl_tcp_reordering) + tp->sacked_out >= net->ipv4.sysctl_tcp_reordering) tp->reordering = min_t(unsigned int, tp->reordering, - sysctl_tcp_reordering); + net->ipv4.sysctl_tcp_reordering); tcp_set_ca_state(sk, TCP_CA_Loss); tp->high_seq = tp->snd_nxt; tcp_ecn_queue_cwr(tp); @@@ -2109,6 -2118,7 +2118,7 @@@ static bool tcp_time_to_recover(struct { struct tcp_sock *tp = tcp_sk(sk); __u32 packets_out; + int tcp_reordering = sock_net(sk)->ipv4.sysctl_tcp_reordering; /* Trick#1: The loss is proven. */ if (tp->lost_out) @@@ -2123,7 -2133,7 +2133,7 @@@ */ packets_out = tp->packets_out; if (packets_out <= tp->reordering && - tp->sacked_out >= max_t(__u32, packets_out/2, sysctl_tcp_reordering) && + tp->sacked_out >= max_t(__u32, packets_out/2, tcp_reordering) && !tcp_may_send_now(sk)) { /* We have nothing to send. This connection is limited * either by receiver window or by application. @@@ -2467,14 -2477,12 +2477,12 @@@ static void tcp_init_cwnd_reduction(str tcp_ecn_queue_cwr(tp); } - static void tcp_cwnd_reduction(struct sock *sk, const int prior_unsacked, - int fast_rexmit, int flag) + static void tcp_cwnd_reduction(struct sock *sk, int newly_acked_sacked, + int flag) { struct tcp_sock *tp = tcp_sk(sk); int sndcnt = 0; int delta = tp->snd_ssthresh - tcp_packets_in_flight(tp); - int newly_acked_sacked = prior_unsacked - - (tp->packets_out - tp->sacked_out); if (newly_acked_sacked <= 0 || WARN_ON_ONCE(!tp->prior_cwnd)) return; @@@ -2492,7 -2500,8 +2500,8 @@@ } else { sndcnt = min(delta, newly_acked_sacked); } - sndcnt = max(sndcnt, (fast_rexmit ? 1 : 0)); + /* Force a fast retransmit upon entering fast recovery */ + sndcnt = max(sndcnt, (tp->prr_out ? 0 : 1)); tp->snd_cwnd = tcp_packets_in_flight(tp) + sndcnt; } @@@ -2537,7 -2546,7 +2546,7 @@@ static void tcp_try_keep_open(struct so } } - static void tcp_try_to_open(struct sock *sk, int flag, const int prior_unsacked) + static void tcp_try_to_open(struct sock *sk, int flag) { struct tcp_sock *tp = tcp_sk(sk); @@@ -2551,8 -2560,6 +2560,6 @@@ if (inet_csk(sk)->icsk_ca_state != TCP_CA_CWR) { tcp_try_keep_open(sk); - } else { - tcp_cwnd_reduction(sk, prior_unsacked, 0, flag); } } @@@ -2662,7 -2669,8 +2669,8 @@@ static void tcp_enter_recovery(struct s /* Process an ACK in CA_Loss state. Move to CA_Open if lost data are * recovered or spurious. Otherwise retransmits more on partial ACKs. */ - static void tcp_process_loss(struct sock *sk, int flag, bool is_dupack) + static void tcp_process_loss(struct sock *sk, int flag, bool is_dupack, + int *rexmit) { struct tcp_sock *tp = tcp_sk(sk); bool recovered = !before(tp->snd_una, tp->high_seq); @@@ -2684,10 -2692,15 +2692,15 @@@ tp->frto = 0; /* Step 3.a. loss was real */ } else if (flag & FLAG_SND_UNA_ADVANCED && !recovered) { tp->high_seq = tp->snd_nxt; - __tcp_push_pending_frames(sk, tcp_current_mss(sk), - TCP_NAGLE_OFF); - if (after(tp->snd_nxt, tp->high_seq)) - return; /* Step 2.b */ + /* Step 2.b. Try send new data (but deferred until cwnd + * is updated in tcp_ack()). Otherwise fall back to + * the conventional recovery. + */ + if (tcp_send_head(sk) && + after(tcp_wnd_end(tp), tp->snd_nxt)) { + *rexmit = REXMIT_NEW; + return; + } tp->frto = 0; } } @@@ -2706,12 -2719,11 +2719,11 @@@ else if (flag & FLAG_SND_UNA_ADVANCED) tcp_reset_reno_sack(tp); } - tcp_xmit_retransmit_queue(sk); + *rexmit = REXMIT_LOST; } /* Undo during fast recovery after partial ACK. */ - static bool tcp_try_undo_partial(struct sock *sk, const int acked, - const int prior_unsacked, int flag) + static bool tcp_try_undo_partial(struct sock *sk, const int acked) { struct tcp_sock *tp = tcp_sk(sk); @@@ -2726,10 -2738,8 +2738,8 @@@ * can undo. Otherwise we clock out new packets but do not * mark more packets lost or retransmit more. */ - if (tp->retrans_out) { - tcp_cwnd_reduction(sk, prior_unsacked, 0, flag); + if (tp->retrans_out) return true; - } if (!tcp_any_retrans_done(sk)) tp->retrans_stamp = 0; @@@ -2748,21 -2758,21 +2758,21 @@@ * taking into account both packets sitting in receiver's buffer and * packets lost by network. * - * Besides that it does CWND reduction, when packet loss is detected - * and changes state of machine. + * Besides that it updates the congestion state when packet loss or ECN + * is detected. But it does not reduce the cwnd, it is done by the + * congestion control later. * * It does _not_ decide what to send, it is made in function * tcp_xmit_retransmit_queue(). */ static void tcp_fastretrans_alert(struct sock *sk, const int acked, - const int prior_unsacked, - bool is_dupack, int flag) + bool is_dupack, int *ack_flag, int *rexmit) { struct inet_connection_sock *icsk = inet_csk(sk); struct tcp_sock *tp = tcp_sk(sk); + int fast_rexmit = 0, flag = *ack_flag; bool do_lost = is_dupack || ((flag & FLAG_DATA_SACKED) && (tcp_fackets_out(tp) > tp->reordering)); - int fast_rexmit = 0; if (WARN_ON(!tp->packets_out && tp->sacked_out)) tp->sacked_out = 0; @@@ -2809,8 -2819,10 +2819,10 @@@ /* Use RACK to detect loss */ if (sysctl_tcp_recovery & TCP_RACK_LOST_RETRANS && - tcp_rack_mark_lost(sk)) + tcp_rack_mark_lost(sk)) { flag |= FLAG_LOST_RETRANS; + *ack_flag |= FLAG_LOST_RETRANS; + } /* E. Process state. */ switch (icsk->icsk_ca_state) { @@@ -2819,7 -2831,7 +2831,7 @@@ if (tcp_is_reno(tp) && is_dupack) tcp_add_reno_sack(sk); } else { - if (tcp_try_undo_partial(sk, acked, prior_unsacked, flag)) + if (tcp_try_undo_partial(sk, acked)) return; /* Partial ACK arrived. Force fast retransmit. */ do_lost = tcp_is_reno(tp) || @@@ -2831,7 -2843,7 +2843,7 @@@ } break; case TCP_CA_Loss: - tcp_process_loss(sk, flag, is_dupack); + tcp_process_loss(sk, flag, is_dupack, rexmit); if (icsk->icsk_ca_state != TCP_CA_Open && !(flag & FLAG_LOST_RETRANS)) return; @@@ -2848,7 -2860,7 +2860,7 @@@ tcp_try_undo_dsack(sk); if (!tcp_time_to_recover(sk, flag)) { - tcp_try_to_open(sk, flag, prior_unsacked); + tcp_try_to_open(sk, flag); return; } @@@ -2870,8 -2882,7 +2882,7 @@@ if (do_lost) tcp_update_scoreboard(sk, fast_rexmit); - tcp_cwnd_reduction(sk, prior_unsacked, fast_rexmit, flag); - tcp_xmit_retransmit_queue(sk); + *rexmit = REXMIT_LOST; } /* Kathleen Nichols' algorithm for tracking the minimum value of @@@ -2896,10 -2907,7 +2907,10 @@@ static void tcp_update_rtt_min(struct s { const u32 now = tcp_time_stamp, wlen = sysctl_tcp_min_rtt_wlen * HZ; struct rtt_meas *m = tcp_sk(sk)->rtt_min; - struct rtt_meas rttm = { .rtt = (rtt_us ? : 1), .ts = now }; + struct rtt_meas rttm = { + .rtt = likely(rtt_us) ? rtt_us : jiffies_to_usecs(1), + .ts = now, + }; u32 elapsed; /* Check if the new measurement updates the 1st, 2nd, or 3rd choices */ @@@ -3096,7 -3104,7 +3107,7 @@@ static void tcp_ack_tstamp(struct sock * arrived at the other end. */ static int tcp_clean_rtx_queue(struct sock *sk, int prior_fackets, - u32 prior_snd_una, + u32 prior_snd_una, int *acked, struct tcp_sacktag_state *sack) { const struct inet_connection_sock *icsk = inet_csk(sk); @@@ -3154,10 -3162,13 +3165,13 @@@ flag |= FLAG_ORIG_SACK_ACKED; } - if (sacked & TCPCB_SACKED_ACKED) + if (sacked & TCPCB_SACKED_ACKED) { tp->sacked_out -= acked_pcount; - else if (tcp_is_sack(tp) && !tcp_skb_spurious_retrans(tp, skb)) - tcp_rack_advance(tp, &skb->skb_mstamp, sacked); + } else if (tcp_is_sack(tp)) { + tp->delivered += acked_pcount; + if (!tcp_skb_spurious_retrans(tp, skb)) + tcp_rack_advance(tp, &skb->skb_mstamp, sacked); + } if (sacked & TCPCB_LOST) tp->lost_out -= acked_pcount; @@@ -3266,6 -3277,7 +3280,7 @@@ } } #endif + *acked = pkts_acked; return flag; } @@@ -3299,21 -3311,36 +3314,36 @@@ static inline bool tcp_ack_is_dubious(c /* Decide wheather to run the increase function of congestion control. */ static inline bool tcp_may_raise_cwnd(const struct sock *sk, const int flag) { - if (tcp_in_cwnd_reduction(sk)) - return false; - /* If reordering is high then always grow cwnd whenever data is * delivered regardless of its ordering. Otherwise stay conservative * and only grow cwnd on in-order delivery (RFC5681). A stretched ACK w/ * new SACK or ECE mark may first advance cwnd here and later reduce * cwnd in tcp_fastretrans_alert() based on more states. */ - if (tcp_sk(sk)->reordering > sysctl_tcp_reordering) + if (tcp_sk(sk)->reordering > sock_net(sk)->ipv4.sysctl_tcp_reordering) return flag & FLAG_FORWARD_PROGRESS; return flag & FLAG_DATA_ACKED; } + /* The "ultimate" congestion control function that aims to replace the rigid + * cwnd increase and decrease control (tcp_cong_avoid,tcp_*cwnd_reduction). + * It's called toward the end of processing an ACK with precise rate + * information. All transmission or retransmission are delayed afterwards. + */ + static void tcp_cong_control(struct sock *sk, u32 ack, u32 acked_sacked, + int flag) + { + if (tcp_in_cwnd_reduction(sk)) { + /* Reduce cwnd if state mandates */ + tcp_cwnd_reduction(sk, acked_sacked, flag); + } else if (tcp_may_raise_cwnd(sk, flag)) { + /* Advance cwnd if state allows */ + tcp_cong_avoid(sk, ack, acked_sacked); + } + tcp_update_pacing_rate(sk); + } + /* Check that window update is acceptable. * The function assumes that snd_una<=ack<=snd_next. */ @@@ -3509,6 -3536,27 +3539,27 @@@ static inline void tcp_in_ack_event(str icsk->icsk_ca_ops->in_ack_event(sk, flags); } + /* Congestion control has updated the cwnd already. So if we're in + * loss recovery then now we do any new sends (for FRTO) or + * retransmits (for CA_Loss or CA_recovery) that make sense. + */ + static void tcp_xmit_recovery(struct sock *sk, int rexmit) + { + struct tcp_sock *tp = tcp_sk(sk); + + if (rexmit == REXMIT_NONE) + return; + + if (unlikely(rexmit == 2)) { + __tcp_push_pending_frames(sk, tcp_current_mss(sk), + TCP_NAGLE_OFF); + if (after(tp->snd_nxt, tp->high_seq)) + return; + tp->frto = 0; + } + tcp_xmit_retransmit_queue(sk); + } + /* This routine deals with incoming acks, but not outgoing ones. */ static int tcp_ack(struct sock *sk, const struct sk_buff *skb, int flag) { @@@ -3521,8 -3569,9 +3572,9 @@@ bool is_dupack = false; u32 prior_fackets; int prior_packets = tp->packets_out; - const int prior_unsacked = tp->packets_out - tp->sacked_out; + u32 prior_delivered = tp->delivered; int acked = 0; /* Number of packets newly acked */ + int rexmit = REXMIT_NONE; /* Flag to (re)transmit to recover losses */ sack_state.first_sackt.v64 = 0; @@@ -3611,23 -3660,16 +3663,16 @@@ goto no_queue; /* See if we can take anything off of the retransmit queue. */ - acked = tp->packets_out; - flag |= tcp_clean_rtx_queue(sk, prior_fackets, prior_snd_una, + flag |= tcp_clean_rtx_queue(sk, prior_fackets, prior_snd_una, &acked, &sack_state); - acked -= tp->packets_out; if (tcp_ack_is_dubious(sk, flag)) { is_dupack = !(flag & (FLAG_SND_UNA_ADVANCED | FLAG_NOT_DUP)); - tcp_fastretrans_alert(sk, acked, prior_unsacked, - is_dupack, flag); + tcp_fastretrans_alert(sk, acked, is_dupack, &flag, &rexmit); } if (tp->tlp_high_seq) tcp_process_tlp_ack(sk, ack, flag); - /* Advance cwnd if state allows */ - if (tcp_may_raise_cwnd(sk, flag)) - tcp_cong_avoid(sk, ack, acked); - if ((flag & FLAG_FORWARD_PROGRESS) || !(flag & FLAG_NOT_DUP)) { struct dst_entry *dst = __sk_dst_get(sk); if (dst) @@@ -3636,14 -3678,14 +3681,14 @@@ if (icsk->icsk_pending == ICSK_TIME_RETRANS) tcp_schedule_loss_probe(sk); - tcp_update_pacing_rate(sk); + tcp_cong_control(sk, ack, tp->delivered - prior_delivered, flag); + tcp_xmit_recovery(sk, rexmit); return 1; no_queue: /* If data was DSACKed, see if we can undo a cwnd reduction. */ if (flag & FLAG_DSACKING_ACK) - tcp_fastretrans_alert(sk, acked, prior_unsacked, - is_dupack, flag); + tcp_fastretrans_alert(sk, acked, is_dupack, &flag, &rexmit); /* If this ack opens up a zero window, clear backoff. It was * being used to time the probes, and is probably far higher than * it needs to be for normal retransmission. @@@ -3666,8 -3708,8 +3711,8 @@@ old_ack if (TCP_SKB_CB(skb)->sacked) { flag |= tcp_sacktag_write_queue(sk, skb, prior_snd_una, &sack_state); - tcp_fastretrans_alert(sk, acked, prior_unsacked, - is_dupack, flag); + tcp_fastretrans_alert(sk, acked, is_dupack, &flag, &rexmit); + tcp_xmit_recovery(sk, rexmit); } SOCK_DEBUG(sk, "Ack %u before %u:%u\n", ack, tp->snd_una, tp->snd_nxt); @@@ -3998,7 -4040,7 +4043,7 @@@ void tcp_reset(struct sock *sk * * If we are in FINWAIT-2, a received FIN moves us to TIME-WAIT. */ - static void tcp_fin(struct sock *sk) + void tcp_fin(struct sock *sk) { struct tcp_sock *tp = tcp_sk(sk); @@@ -5512,6 -5554,9 +5557,9 @@@ static bool tcp_rcv_fastopen_synack(str tp->syn_data_acked = tp->syn_data; if (tp->syn_data_acked) NET_INC_STATS_BH(sock_net(sk), LINUX_MIB_TCPFASTOPENACTIVE); + + tcp_fastopen_add_skb(sk, synack); + return false; } @@@ -6118,9 -6163,10 +6166,10 @@@ static bool tcp_syn_flood_action(const struct request_sock_queue *queue = &inet_csk(sk)->icsk_accept_queue; const char *msg = "Dropping request"; bool want_cookie = false; + struct net *net = sock_net(sk); #ifdef CONFIG_SYN_COOKIES - if (sysctl_tcp_syncookies) { + if (net->ipv4.sysctl_tcp_syncookies) { msg = "Sending cookies"; want_cookie = true; NET_INC_STATS_BH(sock_net(sk), LINUX_MIB_TCPREQQFULLDOCOOKIES); @@@ -6129,7 -6175,7 +6178,7 @@@ NET_INC_STATS_BH(sock_net(sk), LINUX_MIB_TCPREQQFULLDROP); if (!queue->synflood_warned && - sysctl_tcp_syncookies != 2 && + net->ipv4.sysctl_tcp_syncookies != 2 && xchg(&queue->synflood_warned, 1) == 0) pr_info("%s: Possible SYN flooding on port %d. %s. Check SNMP counters.\n", proto, ntohs(tcp_hdr(skb)->dest), msg); @@@ -6162,6 -6208,7 +6211,7 @@@ int tcp_conn_request(struct request_soc __u32 isn = TCP_SKB_CB(skb)->tcp_tw_isn; struct tcp_options_received tmp_opt; struct tcp_sock *tp = tcp_sk(sk); + struct net *net = sock_net(sk); struct sock *fastopen_sk = NULL; struct dst_entry *dst = NULL; struct request_sock *req; @@@ -6172,7 -6219,7 +6222,7 @@@ * limitations, they conserve resources and peer is * evidently real one. */ - if ((sysctl_tcp_syncookies == 2 || + if ((net->ipv4.sysctl_tcp_syncookies == 2 || inet_csk_reqsk_queue_is_full(sk)) && !isn) { want_cookie = tcp_syn_flood_action(sk, skb, rsk_ops->slab_name); if (!want_cookie) @@@ -6238,7 -6285,7 +6288,7 @@@ } } /* Kill the following clause, if you dislike this way. */ - else if (!sysctl_tcp_syncookies && + else if (!net->ipv4.sysctl_tcp_syncookies && (sysctl_max_syn_backlog - inet_csk_reqsk_queue_len(sk) < (sysctl_max_syn_backlog >> 2)) && !tcp_peer_is_proven(req, dst, false, diff --combined net/ipv4/tcp_ipv4.c index 487ac67,3f872a6..4c8d58d --- a/net/ipv4/tcp_ipv4.c +++ b/net/ipv4/tcp_ipv4.c @@@ -311,7 -311,7 +311,7 @@@ static void do_redirect(struct sk_buff /* handle ICMP messages on TCP_NEW_SYN_RECV request sockets */ -void tcp_req_err(struct sock *sk, u32 seq) +void tcp_req_err(struct sock *sk, u32 seq, bool abort) { struct request_sock *req = inet_reqsk(sk); struct net *net = sock_net(sk); @@@ -323,7 -323,7 +323,7 @@@ if (seq != tcp_rsk(req)->snt_isn) { NET_INC_STATS_BH(net, LINUX_MIB_OUTOFWINDOWICMPS); - } else { + } else if (abort) { /* * Still in SYN_RECV, just remove it silently. * There is no good way to pass the error to the newly @@@ -383,12 -383,7 +383,12 @@@ void tcp_v4_err(struct sk_buff *icmp_sk } seq = ntohl(th->seq); if (sk->sk_state == TCP_NEW_SYN_RECV) - return tcp_req_err(sk, seq); + return tcp_req_err(sk, seq, + type == ICMP_PARAMETERPROB || + type == ICMP_TIME_EXCEEDED || + (type == ICMP_DEST_UNREACH && + (code == ICMP_NET_UNREACH || + code == ICMP_HOST_UNREACH))); bh_lock_sock(sk); /* If too many ICMPs get dropped on busy @@@ -642,8 -637,8 +642,8 @@@ static void tcp_v4_send_reset(const str * Incoming packet is checked with md5 hash with finding key, * no RST generated if md5 hash doesn't match. */ - sk1 = __inet_lookup_listener(net, - &tcp_hashinfo, ip_hdr(skb)->saddr, + sk1 = __inet_lookup_listener(net, &tcp_hashinfo, NULL, 0, + ip_hdr(skb)->saddr, th->source, ip_hdr(skb)->daddr, ntohs(th->source), inet_iif(skb)); /* don't send rst if it can't find key */ @@@ -865,7 -860,6 +865,6 @@@ static void tcp_v4_reqsk_destructor(str kfree(inet_rsk(req)->opt); } - #ifdef CONFIG_TCP_MD5SIG /* * RFC2385 MD5 checksumming requires a mapping of @@@ -1587,7 -1581,8 +1586,8 @@@ int tcp_v4_rcv(struct sk_buff *skb TCP_SKB_CB(skb)->sacked = 0; lookup: - sk = __inet_lookup_skb(&tcp_hashinfo, skb, th->source, th->dest); + sk = __inet_lookup_skb(&tcp_hashinfo, skb, __tcp_hdrlen(th), th->source, + th->dest); if (!sk) goto no_tcp_socket; @@@ -1597,30 -1592,28 +1597,30 @@@ process if (sk->sk_state == TCP_NEW_SYN_RECV) { struct request_sock *req = inet_reqsk(sk); - struct sock *nsk = NULL; + struct sock *nsk; sk = req->rsk_listener; - if (tcp_v4_inbound_md5_hash(sk, skb)) - goto discard_and_relse; - if (likely(sk->sk_state == TCP_LISTEN)) { - nsk = tcp_check_req(sk, skb, req, false); - } else { + if (unlikely(tcp_v4_inbound_md5_hash(sk, skb))) { + reqsk_put(req); + goto discard_it; + } + if (unlikely(sk->sk_state != TCP_LISTEN)) { inet_csk_reqsk_queue_drop_and_put(sk, req); goto lookup; } + sock_hold(sk); + nsk = tcp_check_req(sk, skb, req, false); if (!nsk) { reqsk_put(req); - goto discard_it; + goto discard_and_relse; } if (nsk == sk) { - sock_hold(sk); reqsk_put(req); } else if (tcp_child_process(sk, nsk, skb)) { tcp_v4_send_reset(nsk, skb); - goto discard_it; + goto discard_and_relse; } else { + sock_put(sk); return 0; } } @@@ -1703,7 -1696,8 +1703,8 @@@ do_time_wait switch (tcp_timewait_state_process(inet_twsk(sk), skb, th)) { case TCP_TW_SYN: { struct sock *sk2 = inet_lookup_listener(dev_net(skb->dev), - &tcp_hashinfo, + &tcp_hashinfo, skb, + __tcp_hdrlen(th), iph->saddr, th->source, iph->daddr, th->dest, inet_iif(skb)); @@@ -2395,6 -2389,16 +2396,16 @@@ static int __net_init tcp_sk_init(struc net->ipv4.sysctl_tcp_keepalive_probes = TCP_KEEPALIVE_PROBES; net->ipv4.sysctl_tcp_keepalive_intvl = TCP_KEEPALIVE_INTVL; + net->ipv4.sysctl_tcp_syn_retries = TCP_SYN_RETRIES; + net->ipv4.sysctl_tcp_synack_retries = TCP_SYNACK_RETRIES; + net->ipv4.sysctl_tcp_syncookies = 1; + net->ipv4.sysctl_tcp_reordering = TCP_FASTRETRANS_THRESH; + net->ipv4.sysctl_tcp_retries1 = TCP_RETR1; + net->ipv4.sysctl_tcp_retries2 = TCP_RETR2; + net->ipv4.sysctl_tcp_orphan_retries = 0; + net->ipv4.sysctl_tcp_fin_timeout = TCP_FIN_TIMEOUT; + net->ipv4.sysctl_tcp_notsent_lowat = UINT_MAX; + return 0; fail: tcp_sk_exit(net); diff --combined net/ipv4/udp.c index 95d2f19,9fc4e9c..836abe5 --- a/net/ipv4/udp.c +++ b/net/ipv4/udp.c @@@ -356,8 -356,8 +356,8 @@@ EXPORT_SYMBOL(udp_lib_get_port) * match_wildcard == false: addresses must be exactly the same, i.e. * 0.0.0.0 only equals to 0.0.0.0 */ - static int ipv4_rcv_saddr_equal(const struct sock *sk1, const struct sock *sk2, - bool match_wildcard) + int ipv4_rcv_saddr_equal(const struct sock *sk1, const struct sock *sk2, + bool match_wildcard) { struct inet_sock *inet1 = inet_sk(sk1), *inet2 = inet_sk(sk2); @@@ -848,32 -848,20 +848,20 @@@ void udp_set_csum(bool nocheck, struct { struct udphdr *uh = udp_hdr(skb); - if (nocheck) + if (nocheck) { uh->check = 0; - else if (skb_is_gso(skb)) + } else if (skb_is_gso(skb)) { uh->check = ~udp_v4_check(len, saddr, daddr, 0); - else if (skb_dst(skb) && skb_dst(skb)->dev && - (skb_dst(skb)->dev->features & - (NETIF_F_IP_CSUM | NETIF_F_HW_CSUM))) { - - BUG_ON(skb->ip_summed == CHECKSUM_PARTIAL); - + } else if (skb->ip_summed == CHECKSUM_PARTIAL) { + uh->check = 0; + uh->check = udp_v4_check(len, saddr, daddr, lco_csum(skb)); + if (uh->check == 0) + uh->check = CSUM_MANGLED_0; + } else { skb->ip_summed = CHECKSUM_PARTIAL; skb->csum_start = skb_transport_header(skb) - skb->head; skb->csum_offset = offsetof(struct udphdr, check); uh->check = ~udp_v4_check(len, saddr, daddr, 0); - } else { - __wsum csum; - - BUG_ON(skb->ip_summed == CHECKSUM_PARTIAL); - - uh->check = 0; - csum = skb_checksum(skb, 0, len, 0); - uh->check = udp_v4_check(len, saddr, daddr, csum); - if (uh->check == 0) - uh->check = CSUM_MANGLED_0; - - skb->ip_summed = CHECKSUM_UNNECESSARY; } } EXPORT_SYMBOL(udp_set_csum); @@@ -1048,10 -1036,8 +1036,10 @@@ int udp_sendmsg(struct sock *sk, struc if (msg->msg_controllen) { err = ip_cmsg_send(sock_net(sk), msg, &ipc, sk->sk_family == AF_INET6); - if (err) + if (unlikely(err)) { + kfree(ipc.opt); return err; + } if (ipc.opt) free = 1; connected = 0; diff --combined net/ipv6/addrconf.c index bdd7eac,ac0ba9e..4751f89 --- a/net/ipv6/addrconf.c +++ b/net/ipv6/addrconf.c @@@ -583,7 -583,7 +583,7 @@@ static int inet6_netconf_get_devconf(st if (err < 0) goto errout; - err = EINVAL; + err = -EINVAL; if (!tb[NETCONFA_IFINDEX]) goto errout; @@@ -3538,7 -3538,6 +3538,7 @@@ static void addrconf_dad_begin(struct i { struct inet6_dev *idev = ifp->idev; struct net_device *dev = idev->dev; + bool notify = false; addrconf_join_solict(dev, &ifp->addr); @@@ -3584,7 -3583,7 +3584,7 @@@ /* Because optimistic nodes can use this address, * notify listeners. If DAD fails, RTM_DELADDR is sent. */ - ipv6_ifa_notify(RTM_NEWADDR, ifp); + notify = true; } } @@@ -3592,8 -3591,6 +3592,8 @@@ out: spin_unlock(&ifp->lock); read_unlock_bh(&idev->lock); + if (notify) + ipv6_ifa_notify(RTM_NEWADDR, ifp); } static void addrconf_dad_start(struct inet6_ifaddr *ifp) @@@ -4714,6 -4711,8 +4714,8 @@@ static inline void ipv6_store_devconf(s array[DEVCONF_IGNORE_ROUTES_WITH_LINKDOWN] = cnf->ignore_routes_with_linkdown; /* we omit DEVCONF_STABLE_SECRET for now */ array[DEVCONF_USE_OIF_ADDRS_ONLY] = cnf->use_oif_addrs_only; + array[DEVCONF_DROP_UNICAST_IN_L2_MULTICAST] = cnf->drop_unicast_in_l2_multicast; + array[DEVCONF_DROP_UNSOLICITED_NA] = cnf->drop_unsolicited_na; } static inline size_t inet6_ifla6_size(void) @@@ -5788,6 -5787,20 +5790,20 @@@ static struct addrconf_sysctl_tabl .proc_handler = addrconf_sysctl_ignore_routes_with_linkdown, }, { + .procname = "drop_unicast_in_l2_multicast", + .data = &ipv6_devconf.drop_unicast_in_l2_multicast, + .maxlen = sizeof(int), + .mode = 0644, + .proc_handler = proc_dointvec, + }, + { + .procname = "drop_unsolicited_na", + .data = &ipv6_devconf.drop_unsolicited_na, + .maxlen = sizeof(int), + .mode = 0644, + .proc_handler = proc_dointvec, + }, + { /* sentinel */ } }, diff --combined net/ipv6/ip6_gre.c index a69aad1,a94e506..f7c9560 --- a/net/ipv6/ip6_gre.c +++ b/net/ipv6/ip6_gre.c @@@ -360,7 -360,7 +360,7 @@@ static void ip6gre_tunnel_uninit(struc struct ip6gre_net *ign = net_generic(t->net, ip6gre_net_id); ip6gre_tunnel_unlink(ign, t); - ip6_tnl_dst_reset(t); + dst_cache_reset(&t->dst_cache); dev_put(dev); } @@@ -633,7 -633,7 +633,7 @@@ static netdev_tx_t ip6gre_xmit2(struct } if (!fl6->flowi6_mark) - dst = ip6_tnl_dst_get(tunnel); + dst = dst_cache_get(&tunnel->dst_cache); if (!dst) { dst = ip6_route_output(net, NULL, fl6); @@@ -702,7 -702,7 +702,7 @@@ } if (!fl6->flowi6_mark && ndst) - ip6_tnl_dst_set(tunnel, ndst); + dst_cache_set_ip6(&tunnel->dst_cache, ndst, &fl6->saddr); skb_dst_set(skb, dst); proto = NEXTHDR_GRE; @@@ -1009,7 -1009,7 +1009,7 @@@ static int ip6gre_tnl_change(struct ip6 t->parms.o_key = p->o_key; t->parms.i_flags = p->i_flags; t->parms.o_flags = p->o_flags; - ip6_tnl_dst_reset(t); + dst_cache_reset(&t->dst_cache); ip6gre_tnl_link_config(t,

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

linux-merge February 2016