)]}' { "log": [ { "commit": "a9517d0f43832d787a4a0348163d659641bfd83c", "tree": "c3b56b1e29d2dda70888fc69feb704ce557cc0ae", "parents": [ "3bad540ed8285fb53f6365420bba0320d8cd2066" ], "author": { "name": "Jiri Pirko", "email": "jiri@resnulli.us", "time": "Wed Jan 22 09:05:57 2014 +0100" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Wed Jan 22 21:57:17 2014 -0800" }, "message": "rtnetlink: remove ndo_get_slave\n\nNo longer used API bond-specific can be removed now. This is now handled\nin a generic way in rtnl_link_ops.\n\nSigned-off-by: Jiri Pirko \u003cjiri@resnulli.us\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "b582ef0990d457f7ce8ccf827af51a575ca0b4a6", "tree": "2893cba0f3c386795a7324c71851d165a68d891e", "parents": [ "2618abb73c8953f0848511fc13f68da4d8337574" ], "author": { "name": "Or Gerlitz", "email": "ogerlitz@mellanox.com", "time": "Mon Jan 20 13:59:19 2014 +0200" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Tue Jan 21 18:05:04 2014 -0800" }, "message": "net: Add GRO support for UDP encapsulating protocols\n\nAdd GRO handlers for protocols that do UDP encapsulation, with the intent of\nbeing able to coalesce packets which encapsulate packets belonging to\nthe same TCP session.\n\nFor GRO purposes, the destination UDP port takes the role of the ether type\nfield in the ethernet header or the next protocol in the IP header.\n\nThe UDP GRO handler will only attempt to coalesce packets whose destination\nport is registered to have gro handler.\n\nUse a mark on the skb GRO CB data to disallow (flush) running the udp gro receive\ncode twice on a packet. This solves the problem of udp encapsulated packets whose\ninner VM packet is udp and happen to carry a port which has registered offloads.\n\nSigned-off-by: Shlomo Pongratz \u003cshlomop@mellanox.com\u003e\nSigned-off-by: Or Gerlitz \u003cogerlitz@mellanox.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "1d3ee88ae0d605629bf369ab0b868dae8ca62a48", "tree": "7d547bb8ba12b9417a8f0055a08f1026d6392cf2", "parents": [ "07699f9a7c8d1002e07011d5aa382cd63241eea8" ], "author": { "name": "sfeldma@cumulusnetworks.com", "email": "sfeldma@cumulusnetworks.com", "time": "Thu Jan 16 22:57:56 2014 -0800" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Fri Jan 17 18:51:58 2014 -0800" }, "message": "bonding: add netlink attributes to slave link dev\n\nIf link is IFF_SLAVE, extend link dev netlink attributes to include\nslave attributes with new IFLA_SLAVE nest. Add netlink notification\n(RTM_NEWLINK) when slave status changes from backup to active, or\nvisa-versa.\n\nAdds new ndo_get_slave op to net_device_ops to fill skb with IFLA_SLAVE\nattributes. Currently only used by bonding driver, but could be\nused by other aggregating devices with slaves.\n\nSigned-off-by: Scott Feldman \u003csfeldma@cumulusnetworks.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "a953be53ce40440acb4740edb48577b9468d4c3d", "tree": "c9bef1d8ce80a7f2dc852f87bcd2a969ef34e1f8", "parents": [ "ab7db91705e95ed1bba1304388936fccfa58c992" ], "author": { "name": "Michael Dalton", "email": "mwdalton@google.com", "time": "Thu Jan 16 22:23:28 2014 -0800" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Jan 16 23:46:06 2014 -0800" }, "message": "net-sysfs: add support for device-specific rx queue sysfs attributes\n\nExtend existing support for netdevice receive queue sysfs attributes to\npermit a device-specific attribute group. Initial use case for this\nsupport will be to allow the virtio-net device to export per-receive\nqueue mergeable receive buffer size.\n\nSigned-off-by: Michael Dalton \u003cmwdalton@google.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "1d486bfb66971ebacc2a46a23431ace9af70dc66", "tree": "b280bfff1e11a8857f9a2e967e92e6442a9d043b", "parents": [ "31cf344caf444ca7411d89c8ac907d886eeab1a7" ], "author": { "name": "Veaceslav Falico", "email": "vfalico@redhat.com", "time": "Thu Jan 16 00:02:18 2014 +0100" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Jan 16 17:15:41 2014 -0800" }, "message": "net: add NETDEV_PRECHANGEMTU to notify before mtu change happens\n\nCurrently, if a device changes its mtu, first the change happens (invloving\nall the side effects), and after that the NETDEV_CHANGEMTU is sent so that\nother devices can catch up with the new mtu. However, if they return\nNOTIFY_BAD, then the change is reverted and error returned.\n\nThis is a really long and costy operation (sometimes). To fix this, add\nNETDEV_PRECHANGEMTU notification which is called prior to any change\nactually happening, and if any callee returns NOTIFY_BAD - the change is\naborted. This way we\u0027re skipping all the playing with apply/revert the mtu.\n\nCC: \"David S. Miller\" \u003cdavem@davemloft.net\u003e\nCC: Jiri Pirko \u003cjiri@resnulli.us\u003e\nCC: Eric Dumazet \u003cedumazet@google.com\u003e\nCC: Nicolas Dichtel \u003cnicolas.dichtel@6wind.com\u003e\nCC: Cong Wang \u003camwang@redhat.com\u003e\nSigned-off-by: Veaceslav Falico \u003cvfalico@redhat.com\u003e\nAcked-by: Jiri Pirko \u003cjiri@resnulli.us\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "5bb025fae53889cc99a21058c5dd369bf8cce820", "tree": "cc80601cdb7096b9dc4d9d5f14eb477c4dd82322", "parents": [ "3ee32707560955e92d30f7f6e5138cb92a3b1a7e" ], "author": { "name": "Veaceslav Falico", "email": "vfalico@redhat.com", "time": "Tue Jan 14 21:58:51 2014 +0100" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Wed Jan 15 15:16:20 2014 -0800" }, "message": "net: rename sysfs symlinks on device name change\n\nCurrently, we don\u0027t rename the upper/lower_ifc symlinks in\n/sys/class/net/*/ , which might result stale/duplicate links/names.\n\nFix this by adding netdev_adjacent_rename_links(dev, oldname) which renames\nall the upper/lower interface\u0027s links to dev from the upper/lower_oldname\nto the new name.\n\nWe don\u0027t need a rollback because only we control these symlinks and if we\nfail to rename them - sysfs will anyway complain.\n\nReported-by: Ding Tianhong \u003cdingtianhong@huawei.com\u003e\nCC: Ding Tianhong \u003cdingtianhong@huawei.com\u003e\nCC: \"David S. Miller\" \u003cdavem@davemloft.net\u003e\nCC: Eric Dumazet \u003cedumazet@google.com\u003e\nCC: Nicolas Dichtel \u003cnicolas.dichtel@6wind.com\u003e\nCC: Cong Wang \u003camwang@redhat.com\u003e\nSigned-off-by: Veaceslav Falico \u003cvfalico@redhat.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "0a379e21c503b2ff66b44d588df9f231e9b0b9ca", "tree": "22b875fcf4b67fcd007726f00c5fc1748ce985d0", "parents": [ "a49da8811e71c5355b52c65ee32976741d5834cd", "fdc3452cd2c7b2bfe0f378f92123f4f9a98fa2bd" ], "author": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Tue Jan 14 14:37:09 2014 -0800" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Tue Jan 14 14:42:42 2014 -0800" }, "message": "Merge git://git.kernel.org/pub/scm/linux/kernel/git/davem/net\n" }, { "commit": "f663dd9aaf9ed124f25f0f8452edf238f087ad50", "tree": "2aee7dfcfd373c6905de5b2a9f810f9e209156b6", "parents": [ "b13ba1b83f524732523db1079e56478b32c85c96" ], "author": { "name": "Jason Wang", "email": "jasowang@redhat.com", "time": "Fri Jan 10 16:18:26 2014 +0800" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Fri Jan 10 13:23:08 2014 -0500" }, "message": "net: core: explicitly select a txq before doing l2 forwarding\n\nCurrently, the tx queue were selected implicitly in ndo_dfwd_start_xmit(). The\nwill cause several issues:\n\n- NETIF_F_LLTX were removed for macvlan, so txq lock were done for macvlan\n instead of lower device which misses the necessary txq synchronization for\n lower device such as txq stopping or frozen required by dev watchdog or\n control path.\n- dev_hard_start_xmit() was called with NULL txq which bypasses the net device\n watchdog.\n- dev_hard_start_xmit() does not check txq everywhere which will lead a crash\n when tso is disabled for lower device.\n\nFix this by explicitly introducing a new param for .ndo_select_queue() for just\nselecting queues in the case of l2 forwarding offload. netdev_pick_tx() was also\nextended to accept this parameter and dev_queue_xmit_accel() was used to do l2\nforwarding transmission.\n\nWith this fixes, NETIF_F_LLTX could be preserved for macvlan and there\u0027s no need\nto check txq against NULL in dev_hard_start_xmit(). Also there\u0027s no need to keep\na dedicated ndo_dfwd_start_xmit() and we can just reuse the code of\ndev_queue_xmit() to do the transmission.\n\nIn the future, it was also required for macvtap l2 forwarding support since it\nprovides a necessary synchronization method.\n\nCc: John Fastabend \u003cjohn.r.fastabend@intel.com\u003e\nCc: Neil Horman \u003cnhorman@tuxdriver.com\u003e\nCc: e1000-devel@lists.sourceforge.net\nSigned-off-by: Jason Wang \u003cjasowang@redhat.com\u003e\nAcked-by: Neil Horman \u003cnhorman@tuxdriver.com\u003e\nAcked-by: John Fastabend \u003cjohn.r.fastabend@intel.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "bf5a755f5e9186406bbf50f4087100af5bd68e40", "tree": "c971c1aafbcb999a65b5f088bf2627c48006072a", "parents": [ "cdb3f4a31b64c3a1c6eef40bc01ebc9594c58a8c" ], "author": { "name": "Jerry Chu", "email": "hkchu@google.com", "time": "Tue Jan 07 10:23:19 2014 -0800" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Tue Jan 07 16:21:31 2014 -0500" }, "message": "net-gre-gro: Add GRE support to the GRO stack\n\nThis patch built on top of Commit 299603e8370a93dd5d8e8d800f0dff1ce2c53d36\n(\"net-gro: Prepare GRO stack for the upcoming tunneling support\") to add\nthe support of the standard GRE (RFC1701/RFC2784/RFC2890) to the GRO\nstack. It also serves as an example for supporting other encapsulation\nprotocols in the GRO stack in the future.\n\nThe patch supports version 0 and all the flags (key, csum, seq#) but\nwill flush any pkt with the S (seq#) flag. This is because the S flag\nis not support by GSO, and a GRO pkt may end up in the forwarding path,\nthus requiring GSO support to break it up correctly.\n\nCurrently the \"packet_offload\" structure only contains L3 (ETH_P_IP/\nETH_P_IPV6) GRO offload support so the encapped pkts are limited to\nIP pkts (i.e., w/o L2 hdr). But support for other protocol type can\nbe easily added, so is the support for GRE variations like NVGRE.\n\nThe patch also support csum offload. Specifically if the csum flag is on\nand the h/w is capable of checksumming the payload (CHECKSUM_COMPLETE),\nthe code will take advantage of the csum computed by the h/w when\nvalidating the GRE csum.\n\nNote that commit 60769a5dcd8755715c7143b4571d5c44f01796f1 \"ipv4: gre:\nadd GRO capability\" already introduces GRO capability to IPv4 GRE\ntunnels, using the gro_cells infrastructure. But GRO is done after\nGRE hdr has been removed (i.e., decapped). The following patch applies\nGRO when pkts first come in (before hitting the GRE tunnel code). There\nis some performance advantage for applying GRO as early as possible.\nAlso this approach is transparent to other subsystem like Open vSwitch\nwhere GRE decap is handled outside of the IP stack hence making it\nharder for the gro_cells stuff to apply. On the other hand, some NICs\nare still not capable of hashing on the inner hdr of a GRE pkt (RSS).\nIn that case the GRO processing of pkts from the same remote host will\nall happen on the same CPU and the performance may be suboptimal.\n\nI\u0027m including some rough preliminary performance numbers below. Note\nthat the performance will be highly dependent on traffic load, mix as\nusual. Moreover it also depends on NIC offload features hence the\nfollowing is by no means a comprehesive study. Local testing and tuning\nwill be needed to decide the best setting.\n\nAll tests spawned 50 copies of netperf TCP_STREAM and ran for 30 secs.\n(super_netperf 50 -H 192.168.1.18 -l 30)\n\nAn IP GRE tunnel with only the key flag on (e.g., ip tunnel add gre1\nmode gre local 10.246.17.18 remote 10.246.17.17 ttl 255 key 123)\nis configured.\n\nThe GRO support for pkts AFTER decap are controlled through the device\nfeature of the GRE device (e.g., ethtool -K gre1 gro on/off).\n\n1.1 ethtool -K gre1 gro off; ethtool -K eth0 gro off\nthruput: 9.16Gbps\nCPU utilization: 19%\n\n1.2 ethtool -K gre1 gro on; ethtool -K eth0 gro off\nthruput: 5.9Gbps\nCPU utilization: 15%\n\n1.3 ethtool -K gre1 gro off; ethtool -K eth0 gro on\nthruput: 9.26Gbps\nCPU utilization: 12-13%\n\n1.4 ethtool -K gre1 gro on; ethtool -K eth0 gro on\nthruput: 9.26Gbps\nCPU utilization: 10%\n\nThe following tests were performed on a different NIC that is capable of\ncsum offload. I.e., the h/w is capable of computing IP payload csum\n(CHECKSUM_COMPLETE).\n\n2.1 ethtool -K gre1 gro on (hence will use gro_cells)\n\n2.1.1 ethtool -K eth0 gro off; csum offload disabled\nthruput: 8.53Gbps\nCPU utilization: 9%\n\n2.1.2 ethtool -K eth0 gro off; csum offload enabled\nthruput: 8.97Gbps\nCPU utilization: 7-8%\n\n2.1.3 ethtool -K eth0 gro on; csum offload disabled\nthruput: 8.83Gbps\nCPU utilization: 5-6%\n\n2.1.4 ethtool -K eth0 gro on; csum offload enabled\nthruput: 8.98Gbps\nCPU utilization: 5%\n\n2.2 ethtool -K gre1 gro off\n\n2.2.1 ethtool -K eth0 gro off; csum offload disabled\nthruput: 5.93Gbps\nCPU utilization: 9%\n\n2.2.2 ethtool -K eth0 gro off; csum offload enabled\nthruput: 5.62Gbps\nCPU utilization: 8%\n\n2.2.3 ethtool -K eth0 gro on; csum offload disabled\nthruput: 7.69Gbps\nCPU utilization: 8%\n\n2.2.4 ethtool -K eth0 gro on; csum offload enabled\nthruput: 8.96Gbps\nCPU utilization: 5-6%\n\nSigned-off-by: H.K. Jerry Chu \u003chkchu@google.com\u003e\nReviewed-by: Eric Dumazet \u003cedumazet@google.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "56a4342dfe3145cd66f766adccb28fd9b571606d", "tree": "d1593764488ff8cbb0b83cb9ae35fd968bf81760", "parents": [ "805c1f4aedaba1bc8d839e7c27b128083dd5c2f0", "fe0d692bbc645786bce1a98439e548ae619269f5" ], "author": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Mon Jan 06 17:37:45 2014 -0500" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Mon Jan 06 17:37:45 2014 -0500" }, "message": "Merge branch \u0027master\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/davem/net\n\nConflicts:\n\tdrivers/net/ethernet/qlogic/qlcnic/qlcnic_sriov_pf.c\n\tnet/ipv6/ip6_tunnel.c\n\tnet/ipv6/ip6_vti.c\n\nipv6 tunnel statistic bug fixes conflicting with consolidation into\ngeneric sw per-cpu net stats.\n\nqlogic conflict between queue counting bug fix and the addition\nof multiple MAC address support.\n\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "855404efae0d449cc491978d54ea5d117a3cb271", "tree": "3c44948365a77058d8b1f2ed6e6683bfc52ef256", "parents": [ "a1d4b03a076d95edc88d070f7627a73ab80abddc", "82a37132f300ea53bdcd812917af5a6329ec80c3" ], "author": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Sun Jan 05 20:18:50 2014 -0500" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Sun Jan 05 20:18:50 2014 -0500" }, "message": "Merge git://git.kernel.org/pub/scm/linux/kernel/git/pablo/nf-next\n\nPablo Neira Ayuso says:\n\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\nnetfilter/IPVS updates for net-next\n\nThe following patchset contains Netfilter updates for your net-next tree,\nthey are:\n\n* Add full port randomization support. Some crazy researchers found a way\n to reconstruct the secure ephemeral ports that are allocated in random mode\n by sending off-path bursts of UDP packets to overrun the socket buffer of\n the DNS resolver to trigger retransmissions, then if the timing for the\n DNS resolution done by a client is larger than usual, then they conclude\n that the port that received the burst of UDP packets is the one that was\n opened. It seems a bit aggressive method to me but it seems to work for\n them. As a result, Daniel Borkmann and Hannes Frederic Sowa came up with a\n new NAT mode to fully randomize ports using prandom.\n\n* Add a new classifier to x_tables based on the socket net_cls set via\n cgroups. These includes two patches to prepare the field as requested by\n Zefan Li. Also from Daniel Borkmann.\n\n* Use prandom instead of get_random_bytes in several locations of the\n netfilter code, from Florian Westphal.\n\n* Allow to use the CTA_MARK_MASK in ctnetlink when mangling the conntrack\n mark, also from Florian Westphal.\n\n* Fix compilation warning due to unused variable in IPVS, from Geert\n Uytterhoeven.\n\n* Add support for UID/GID via nfnetlink_queue, from Valentina Giusti.\n\n* Add IPComp extension to x_tables, from Fan Du.\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\n\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "8f84985fec10de64a6b4cdfea45f2b0ab8f07c78", "tree": "03d58e9d72384e8066a6232c793b3ee4242668f6", "parents": [ "653864d9dd6ae26d884abfd53420e61e7383b1da" ], "author": { "name": "Li RongQing", "email": "roy.qing.li@gmail.com", "time": "Sat Jan 04 13:57:59 2014 +0800" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Sat Jan 04 20:10:24 2014 -0500" }, "message": "net: unify the pcpu_tstats and br_cpu_netstats as one\n\nThey are same, so unify them as one, pcpu_sw_netstats.\n\nDefine pcpu_sw_netstat in netdevice.h, remove pcpu_tstats\nfrom if_tunnel and remove br_cpu_netstats from br_private.h\n\nCc: Cong Wang \u003cxiyou.wangcong@gmail.com\u003e\nCc: Stephen Hemminger \u003cstephen@networkplumber.org\u003e\nSigned-off-by: Li RongQing \u003croy.qing.li@gmail.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "86f8515f9721fa171483f0fe0391968fbb949cc9", "tree": "a21f4df024bc16044cb174948494180769cbcbbe", "parents": [ "fe1217c4f3f7d7cbf8efdd8dd5fdc7204a1d65a8" ], "author": { "name": "Daniel Borkmann", "email": "dborkman@redhat.com", "time": "Sun Dec 29 17:27:11 2013 +0100" }, "committer": { "name": "Pablo Neira Ayuso", "email": "pablo@netfilter.org", "time": "Fri Jan 03 23:41:42 2014 +0100" }, "message": "net: netprio: rename config to be more consistent with cgroup configs\n\nWhile we\u0027re at it and introduced CGROUP_NET_CLASSID, lets also make\nNETPRIO_CGROUP more consistent with the rest of cgroups and rename it\ninto CONFIG_CGROUP_NET_PRIO so that for networking, we now have\nCONFIG_CGROUP_NET_{PRIO,CLASSID}. This not only makes the CONFIG\noption consistent among networking cgroups, but also among cgroups\nCONFIG conventions in general as the vast majority has a prefix of\nCONFIG_CGROUP_\u003cSUBSYS\u003e.\n\nSigned-off-by: Daniel Borkmann \u003cdborkman@redhat.com\u003e\nCc: Zefan Li \u003clizefan@huawei.com\u003e\nCc: cgroups@vger.kernel.org\nAcked-by: Li Zefan \u003clizefan@huawei.com\u003e\nSigned-off-by: Pablo Neira Ayuso \u003cpablo@netfilter.org\u003e\n" }, { "commit": "7a7ffbabf99445704be01bff5d7e360da908cf8e", "tree": "3b4f71699307ac2853a8fdffde60fb7055cb1b04", "parents": [ "619a60ee04be33238721a15c1f9704a2a515a33e" ], "author": { "name": "Wei-Chun Chao", "email": "weichunc@plumgrid.com", "time": "Thu Dec 26 13:10:22 2013 -0800" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Jan 02 19:06:47 2014 -0500" }, "message": "ipv4: fix tunneled VM traffic over hw VXLAN/GRE GSO NIC\n\nVM to VM GSO traffic is broken if it goes through VXLAN or GRE\ntunnel and the physical NIC on the host supports hardware VXLAN/GRE\nGSO offload (e.g. bnx2x and next-gen mlx4).\n\nTwo issues -\n(VXLAN) VM traffic has SKB_GSO_DODGY and SKB_GSO_UDP_TUNNEL with\nSKB_GSO_TCP/UDP set depending on the inner protocol. GSO header\nintegrity check fails in udp4_ufo_fragment if inner protocol is\nTCP. Also gso_segs is calculated incorrectly using skb-\u003elen that\nincludes tunnel header. Fix: robust check should only be applied\nto the inner packet.\n\n(VXLAN \u0026 GRE) Once GSO header integrity check passes, NULL segs\nis returned and the original skb is sent to hardware. However the\ntunnel header is already pulled. Fix: tunnel header needs to be\nrestored so that hardware can perform GSO properly on the original\npacket.\n\nSigned-off-by: Wei-Chun Chao \u003cweichunc@plumgrid.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "1d143d9f0c833fcf38cc737eb0a8698fa2dd144c", "tree": "bb777ea9004c6828671f7ef284006bfeffcdc6a8", "parents": [ "2173f8d953e76026bf6ede8ec81ad35ec158dae9" ], "author": { "name": "stephen hemminger", "email": "stephen@networkplumber.org", "time": "Sun Dec 29 14:01:29 2013 -0800" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Wed Jan 01 23:46:09 2014 -0500" }, "message": "net: core functions cleanup\n\nThe following functions are not used outside of net/core/dev.c\nand should be declared static.\n\n call_netdevice_notifiers_info\n __dev_remove_offload\n netdev_has_any_upper_dev\n __netdev_adjacent_dev_remove\n __netdev_adjacent_dev_link_lists\n __netdev_adjacent_dev_unlink_lists\n __netdev_adjacent_dev_unlink\n __netdev_adjacent_dev_link_neighbour\n __netdev_adjacent_dev_unlink_neighbour\n\nAnd the following are never used and should be deleted\n netdev_lower_dev_get_private_rcu\n __netdev_find_adj_rcu\n\nSigned-off-by: Stephen Hemminger \u003cstephen@networkplumber.org\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "2205369a314e12fcec4781cc73ac9c08fc2b47de", "tree": "5daa8338d99f62013631fc9c2ab1a6e6896ef0cd", "parents": [ "71ce176ee6ed1735b9a1160a5704a915d13849b1" ], "author": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Tue Dec 31 16:23:35 2013 -0500" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Tue Dec 31 16:23:35 2013 -0500" }, "message": "vlan: Fix header ops passthru when doing TX VLAN offload.\n\nWhen the vlan code detects that the real device can do TX VLAN offloads\nin hardware, it tries to arrange for the real device\u0027s header_ops to\nbe invoked directly.\n\nBut it does so illegally, by simply hooking the real device\u0027s\nheader_ops up to the VLAN device.\n\nThis doesn\u0027t work because we will end up invoking a set of header_ops\nroutines which expect a device type which matches the real device, but\nwill see a VLAN device instead.\n\nFix this by providing a pass-thru set of header_ops which will arrange\nto pass the proper real device instead.\n\nTo facilitate this add a dev_rebuild_header(). There are\nimplementations which provide a -\u003ecache and -\u003ecreate but not a\n-\u003erebuild (f.e. PLIP). So we need a helper function just like\ndev_hard_header() to avoid crashes.\n\nUse this helper in the one existing place where the\nheader_ops-\u003erebuild was being invoked, the neighbour code.\n\nWith lots of help from Florian Westphal.\n\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "143c9054949436cb05e468439dc5e46231f33d09", "tree": "c2e972d8188fb1b36368e9acb5b6b59466c9d903", "parents": [ "0b6807034791160d5e584138943d2daea765436d", "35eecf052250f663f07a4cded7d3503fd1b50729" ], "author": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Wed Dec 18 16:42:06 2013 -0500" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Wed Dec 18 16:42:06 2013 -0500" }, "message": "Merge git://git.kernel.org/pub/scm/linux/kernel/git/davem/net\n\nConflicts:\n\tdrivers/net/ethernet/intel/i40e/i40e_main.c\n\tdrivers/net/macvtap.c\n\nBoth minor merge hassles, simple overlapping changes.\n\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "477bb93320cec7ae74d5ccfad4f2bfa0b28fbe90", "tree": "c9aad9574baba6c3988d496f766ed743ea32a5cf", "parents": [ "6ea09d8a096487b6418364c3b501c228f1889121" ], "author": { "name": "stephen hemminger", "email": "shemming@brocade.com", "time": "Fri Dec 13 12:35:56 2013 -0800" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Tue Dec 17 15:14:04 2013 -0500" }, "message": "net: remove dead code for add/del multiple\n\nThese function to manipulate multiple addresses are not used anywhere\nin current net-next tree. Some out of tree code maybe using these but\ntoo bad; they should submit their code upstream..\n\nAlso, make __hw_addr_flush local since only used by dev_addr_lists.c\n\nSigned-off-by: Stephen Hemminger \u003cstephen@networkplumber.org\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "e001bfad913bf119fb67c1e8dd2d4ec1f5d392fa", "tree": "fe08e42121a1b5e9bf173af18c04e78c884cc5bf", "parents": [ "2e52f4fe3655c7a2311070c6713f7feabc75486c" ], "author": { "name": "dingtianhong", "email": "dingtianhong@huawei.com", "time": "Fri Dec 13 10:19:55 2013 +0800" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Sat Dec 14 01:58:02 2013 -0500" }, "message": "bonding: create bond_first_slave_rcu()\n\nThe bond_first_slave_rcu() will be used to instead of bond_first_slave()\nin rcu_read_lock().\n\nAccording to the Jay Vosburgh\u0027s suggestion, the struct netdev_adjacent\nshould hide from users who wanted to use it directly. so I package a\nnew function to get the first slave of the bond.\n\nSuggested-by: Nikolay Aleksandrov \u003cnikolay@redhat.com\u003e\nSuggested-by: Jay Vosburgh \u003cfubar@us.ibm.com\u003e\nSuggested-by: Veaceslav Falico \u003cvfalico@redhat.com\u003e\nSigned-off-by: Ding Tianhong \u003cdingtianhong@huawei.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "a0a9663dd2146d54339237764e1bfc60e8a39328", "tree": "ba691d621e527466f5a2a3e47135f35948b167a1", "parents": [ "f280e89ad6a29d9969cb6b216123c798e1689bc4" ], "author": { "name": "Sebastian Siewior", "email": "bigeasy@linutronix.de", "time": "Thu Dec 12 10:15:59 2013 +0100" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Dec 12 15:35:09 2013 -0500" }, "message": "net: make neigh_priv_len in struct net_device 16bit instead of 8bit\n\nneigh_priv_len is defined as u8. With all debug enabled struct\nipoib_neigh has 200 bytes. The largest part is sk_buff_head with 96\nbytes and here the spinlock with 72 bytes.\nThe size value still fits in this u8 leaving some room for more.\n\nOn -RT struct ipoib_neigh put on weight and has 392 bytes. The main\nreason is sk_buff_head with 288 and the fatty here is spinlock with 192\nbytes. This does no longer fit into into neigh_priv_len and gcc\ncomplains.\n\nThis patch changes neigh_priv_len from being 8bit to 16bit. Since the\nfollowing element (dev_id) is 16bit followed by a spinlock which is\naligned, the struct remains with a total size of 3200 (allmodconfig) /\n2048 (with as much debug off as possible) bytes on x86-64.\nOn x86-32 the struct is 1856 (allmodconfig) / 1216 (with as much debug\noff as possible) bytes long. The numbers were gained with and without\nthe patch to prove that this change does not increase the size of the\nstruct.\n\nCc: Steven Rostedt \u003crostedt@goodmis.org\u003e\nCc: Thomas Gleixner \u003ctglx@linutronix.de\u003e\nSigned-off-by: Sebastian Andrzej Siewior \u003cbigeasy@linutronix.de\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "299603e8370a93dd5d8e8d800f0dff1ce2c53d36", "tree": "2a10106aabe88c278a0cd02b93af1add04f5ffcc", "parents": [ "a46dc748caea185d4d0978280a1af0112bf6a8f8" ], "author": { "name": "Jerry Chu", "email": "hkchu@google.com", "time": "Wed Dec 11 20:53:45 2013 -0800" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Dec 12 13:47:53 2013 -0500" }, "message": "net-gro: Prepare GRO stack for the upcoming tunneling support\n\nThis patch modifies the GRO stack to avoid the use of \"network_header\"\nand associated macros like ip_hdr() and ipv6_hdr() in order to allow\nan arbitary number of IP hdrs (v4 or v6) to be used in the\nencapsulation chain. This lays the foundation for various IP\ntunneling support (IP-in-IP, GRE, VXLAN, SIT,...) to be added later.\n\nWith this patch, the GRO stack traversing now is mostly based on\nskb_gro_offset rather than special hdr offsets saved in skb (e.g.,\nskb-\u003enetwork_header). As a result all but the top layer (i.e., the\nthe transport layer) must have hdrs of the same length in order for\na pkt to be considered for aggregation. Therefore when adding a new\nencap layer (e.g., for tunneling), one must check and skip flows\n(e.g., by setting NAPI_GRO_CB(p)-\u003esame_flow to 0) that have a\ndifferent hdr length.\n\nNote that unlike the network header, the transport header can and\nwill continue to be set by the GRO code since there will be at\nmost one \"transport layer\" in the encap chain.\n\nSigned-off-by: H.K. Jerry Chu \u003chkchu@google.com\u003e\nSuggested-by: Eric Dumazet \u003cedumazet@google.com\u003e\nReviewed-by: Eric Dumazet \u003cedumazet@google.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "37cb0620073cb64101d9307931c135c70b2e3f04", "tree": "302cb1bf079ddf1b09dd79cb965d55566d2b75a5", "parents": [ "ef72a7e02a28adfd9d5d0d1de81c0b75f3823aa5" ], "author": { "name": "Ying Xue", "email": "ying.xue@windriver.com", "time": "Tue Dec 10 20:45:41 2013 -0800" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Wed Dec 11 00:17:42 2013 -0500" }, "message": "tipc: remove TIPC usage of field af_packet_priv in struct net_device\n\nTIPC is currently using the field \u0027af_packet_priv\u0027 in struct net_device\nas a handle to find the bearer instance associated to the given network\ndevice. But, by doing so it is blocking other networking cleanups, such\nas the one discussed here:\n\nhttp://patchwork.ozlabs.org/patch/178044/\n\nThis commit removes this usage from TIPC. Instead, we introduce a new\nfield, \u0027tipc_ptr\u0027, to the net_device structure, to serve this purpose.\nWhen TIPC bearer is enabled, the bearer object is associated to\n\u0027tipc_ptr\u0027. When a TIPC packet arrives in the recv_msg() upcall\nfrom a networking device, the bearer object can now be obtained from\n\u0027tipc_ptr\u0027. When a bearer is disabled, the bearer object is detached\nfrom its underlying network device by setting \u0027tipc_ptr\u0027 to NULL.\n\nAdditionally, an RCU lock is used to protect the new pointer.\nHenceforth, the existing tipc_net_lock is used in write mode to\nserialize write accesses to this pointer, while the new RCU lock is\napplied on the read side to ensure that the pointer is 100% valid\nwithin its wrapped area for all readers.\n\nSigned-off-by: Ying Xue \u003cying.xue@windriver.com\u003e\nCc: Patrick McHardy \u003ckaber@trash.net\u003e\nReviewed-by: Paul Gortmaker \u003cpaul.gortmaker@windriver.com\u003e\nSigned-off-by: Jon Maloy \u003cjon.maloy@ericsson.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "e6247027e5173c00efb2084d688d06ff835bc3b0", "tree": "44c1fc4da5358f778fc82a1056ef9a2a6e7eea0e", "parents": [ "f96eb74c84ebb09a03031f7f1e51789a928f9de0" ], "author": { "name": "Eric Dumazet", "email": "edumazet@google.com", "time": "Thu Dec 05 04:45:08 2013 -0800" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Fri Dec 06 15:24:02 2013 -0500" }, "message": "net: introduce dev_consume_skb_any()\n\nSome network drivers use dev_kfree_skb_any() and dev_kfree_skb_irq()\nhelpers to free skbs, both for dropped packets and TX completed ones.\n\nWe need to separate the two causes to get better diagnostics\ngiven by dropwatch or \"perf record -e skb:kfree_skb\"\n\nThis patch provides two new helpers, dev_consume_skb_any() and\ndev_consume_skb_irq() to be used for consumed skbs.\n\n__dev_kfree_skb_irq() is slightly optimized to remove one\natomic_dec_and_test() in fast path, and use this_cpu_{r|w} accessors.\n\nSigned-off-by: Eric Dumazet \u003cedumazet@google.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "9073e1a804c3096eda84ee7cbf11d1f174236c75", "tree": "2f1b488b606bc03f8c5ce133f9158196fef07cd2", "parents": [ "4937e2a6f939a41bf811378e80d71f68aa0950c6", "2bb9936beac2eed0c1ed907f83694bc752e159d9" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Nov 15 16:47:22 2013 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Nov 15 16:47:22 2013 -0800" }, "message": "Merge branch \u0027for-linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/jikos/trivial\n\nPull trivial tree updates from Jiri Kosina:\n \"Usual earth-shaking, news-breaking, rocket science pile from\n trivial.git\"\n\n* \u0027for-linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/jikos/trivial: (23 commits)\n doc: usb: Fix typo in Documentation/usb/gadget_configs.txt\n doc: add missing files to timers/00-INDEX\n timekeeping: Fix some trivial typos in comments\n mm: Fix some trivial typos in comments\n irq: Fix some trivial typos in comments\n NUMA: fix typos in Kconfig help text\n mm: update 00-INDEX\n doc: Documentation/DMA-attributes.txt fix typo\n DRM: comment: `halve\u0027 -\u003e `half\u0027\n Docs: Kconfig: `devlopers\u0027 -\u003e `developers\u0027\n doc: typo on word accounting in kprobes.c in mutliple architectures\n treewide: fix \"usefull\" typo\n treewide: fix \"distingush\" typo\n mm/Kconfig: Grammar s/an/a/\n kexec: Typo s/the/then/\n Documentation/kvm: Update cpuid documentation for steal time and pv eoi\n treewide: Fix common typo in \"identify\"\n __page_to_pfn: Fix typo in comment\n Correct some typos for word frequency\n clk: fixed-factor: Fix a trivial typo\n ...\n" }, { "commit": "42a2d923cc349583ebf6fdd52a7d35e1c2f7e6bd", "tree": "2b2b0c03b5389c1301800119333967efafd994ca", "parents": [ "5cbb3d216e2041700231bcfc383ee5f8b7fc8b74", "75ecab1df14d90e86cebef9ec5c76befde46e65f" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed Nov 13 17:40:34 2013 +0900" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed Nov 13 17:40:34 2013 +0900" }, "message": "Merge git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-next\n\nPull networking updates from David Miller:\n\n 1) The addition of nftables. No longer will we need protocol aware\n firewall filtering modules, it can all live in userspace.\n\n At the core of nftables is a, for lack of a better term, virtual\n machine that executes byte codes to inspect packet or metadata\n (arriving interface index, etc.) and make verdict decisions.\n\n Besides support for loading packet contents and comparing them, the\n interpreter supports lookups in various datastructures as\n fundamental operations. For example sets are supports, and\n therefore one could create a set of whitelist IP address entries\n which have ACCEPT verdicts attached to them, and use the appropriate\n byte codes to do such lookups.\n\n Since the interpreted code is composed in userspace, userspace can\n do things like optimize things before giving it to the kernel.\n\n Another major improvement is the capability of atomically updating\n portions of the ruleset. In the existing netfilter implementation,\n one has to update the entire rule set in order to make a change and\n this is very expensive.\n\n Userspace tools exist to create nftables rules using existing\n netfilter rule sets, but both kernel implementations will need to\n co-exist for quite some time as we transition from the old to the\n new stuff.\n\n Kudos to Patrick McHardy, Pablo Neira Ayuso, and others who have\n worked so hard on this.\n\n 2) Daniel Borkmann and Hannes Frederic Sowa made several improvements\n to our pseudo-random number generator, mostly used for things like\n UDP port randomization and netfitler, amongst other things.\n\n In particular the taus88 generater is updated to taus113, and test\n cases are added.\n\n 3) Support 64-bit rates in HTB and TBF schedulers, from Eric Dumazet\n and Yang Yingliang.\n\n 4) Add support for new 577xx tigon3 chips to tg3 driver, from Nithin\n Sujir.\n\n 5) Fix two fatal flaws in TCP dynamic right sizing, from Eric Dumazet,\n Neal Cardwell, and Yuchung Cheng.\n\n 6) Allow IP_TOS and IP_TTL to be specified in sendmsg() ancillary\n control message data, much like other socket option attributes.\n From Francesco Fusco.\n\n 7) Allow applications to specify a cap on the rate computed\n automatically by the kernel for pacing flows, via a new\n SO_MAX_PACING_RATE socket option. From Eric Dumazet.\n\n 8) Make the initial autotuned send buffer sizing in TCP more closely\n reflect actual needs, from Eric Dumazet.\n\n 9) Currently early socket demux only happens for TCP sockets, but we\n can do it for connected UDP sockets too. Implementation from Shawn\n Bohrer.\n\n10) Refactor inet socket demux with the goal of improving hash demux\n performance for listening sockets. With the main goals being able\n to use RCU lookups on even request sockets, and eliminating the\n listening lock contention. From Eric Dumazet.\n\n11) The bonding layer has many demuxes in it\u0027s fast path, and an RCU\n conversion was started back in 3.11, several changes here extend the\n RCU usage to even more locations. From Ding Tianhong and Wang\n Yufen, based upon suggestions by Nikolay Aleksandrov and Veaceslav\n Falico.\n\n12) Allow stackability of segmentation offloads to, in particular, allow\n segmentation offloading over tunnels. From Eric Dumazet.\n\n13) Significantly improve the handling of secret keys we input into the\n various hash functions in the inet hashtables, TCP fast open, as\n well as syncookies. From Hannes Frederic Sowa. The key fundamental\n operation is \"net_get_random_once()\" which uses static keys.\n\n Hannes even extended this to ipv4/ipv6 fragmentation handling and\n our generic flow dissector.\n\n14) The generic driver layer takes care now to set the driver data to\n NULL on device removal, so it\u0027s no longer necessary for drivers to\n explicitly set it to NULL any more. Many drivers have been cleaned\n up in this way, from Jingoo Han.\n\n15) Add a BPF based packet scheduler classifier, from Daniel Borkmann.\n\n16) Improve CRC32 interfaces and generic SKB checksum iterators so that\n SCTP\u0027s checksumming can more cleanly be handled. Also from Daniel\n Borkmann.\n\n17) Add a new PMTU discovery mode, IP_PMTUDISC_INTERFACE, which forces\n using the interface MTU value. This helps avoid PMTU attacks,\n particularly on DNS servers. From Hannes Frederic Sowa.\n\n18) Use generic XPS for transmit queue steering rather than internal\n (re-)implementation in virtio-net. From Jason Wang.\n\n* git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-next: (1622 commits)\n random32: add test cases for taus113 implementation\n random32: upgrade taus88 generator to taus113 from errata paper\n random32: move rnd_state to linux/random.h\n random32: add prandom_reseed_late() and call when nonblocking pool becomes initialized\n random32: add periodic reseeding\n random32: fix off-by-one in seeding requirement\n PHY: Add RTL8201CP phy_driver to realtek\n xtsonic: add missing platform_set_drvdata() in xtsonic_probe()\n macmace: add missing platform_set_drvdata() in mace_probe()\n ethernet/arc/arc_emac: add missing platform_set_drvdata() in arc_emac_probe()\n ipv6: protect for_each_sk_fl_rcu in mem_check with rcu_read_lock_bh\n vlan: Implement vlan_dev_get_egress_qos_mask as an inline.\n ixgbe: add warning when max_vfs is out of range.\n igb: Update link modes display in ethtool\n netfilter: push reasm skb through instead of original frag skbs\n ip6_output: fragment outgoing reassembled skb properly\n MAINTAINERS: mv643xx_eth: take over maintainership from Lennart\n net_sched: tbf: support of 64bit rates\n ixgbe: deleting dfwd stations out of order can cause null ptr deref\n ixgbe: fix build err, num_rx_queues is only available with CONFIG_RPS\n ...\n" }, { "commit": "a6cc0cfa72e0b6d9f2c8fd858aacc32313c4f272", "tree": "310baaaa6df9798431aeb75ab6df1f13feee81c2", "parents": [ "1ec4864b10171b0691ee196d7006ae56d2c153f2" ], "author": { "name": "John Fastabend", "email": "john.r.fastabend@intel.com", "time": "Wed Nov 06 09:54:46 2013 -0800" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Nov 07 19:11:41 2013 -0500" }, "message": "net: Add layer 2 hardware acceleration operations for macvlan devices\n\nAdd a operations structure that allows a network interface to export\nthe fact that it supports package forwarding in hardware between\nphysical interfaces and other mac layer devices assigned to it (such\nas macvlans). This operaions structure can be used by virtual mac\ndevices to bypass software switching so that forwarding can be done\nin hardware more efficiently.\n\nSigned-off-by: John Fastabend \u003cjohn.r.fastabend@intel.com\u003e\nSigned-off-by: Neil Horman \u003cnhorman@tuxdriver.com\u003e\nCC: Andy Gospodarek \u003candy@greyhouse.net\u003e\nCC: \"David S. Miller\" \u003cdavem@davemloft.net\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "df42153c59a38a65c6f7440d5c70d87d1c24438d", "tree": "3d928a77844d6c81e429e4cb5bc50097587ab805", "parents": [ "78032f9b3eae3a43da9468d8ce30caa6be84f89d" ], "author": { "name": "Dan Carpenter", "email": "dan.carpenter@oracle.com", "time": "Thu Nov 07 10:48:49 2013 +0300" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Nov 07 19:11:13 2013 -0500" }, "message": "net: make ndev-\u003eirq signed for error handling\n\nThere is a bug in cpsw_probe() where we do:\n\n\tndev-\u003eirq \u003d platform_get_irq(pdev, 0);\n\tif (ndev-\u003eirq \u003c 0) {\n\nThe problem is that \"ndev-\u003eirq\" is unsigned so the error handling\ndoesn\u0027t work. I have changed it to a regular int.\n\nSigned-off-by: Dan Carpenter \u003cdan.carpenter@oracle.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "0324e74534241f3f00910ec04ef67de1fe1542f4", "tree": "88f0b2a40a47dc0a22c0ce5e9b75a58470a56204", "parents": [ "1071ec7bc2dabd0a9d12a1ae5570f4fd3ba944ca", "0cae60f91494e34a0c5391f1455f825d5849b05f" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Nov 07 10:03:42 2013 +0900" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Nov 07 11:42:15 2013 +0900" }, "message": "Merge tag \u0027driver-core-3.13-rc1\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/driver-core\n\nPull driver core / sysfs patches from Greg KH:\n \"Here\u0027s the big driver core / sysfs update for 3.13-rc1.\n\n There\u0027s lots of dev_groups updates for different subsystems, as they\n all get slowly migrated over to the safe versions of the attribute\n groups (removing userspace races with the creation of the sysfs\n files.) Also in here are some kobject updates, devres expansions, and\n the first round of Tejun\u0027s sysfs reworking to enable it to be used by\n other subsystems as a backend for an in-kernel filesystem.\n\n All of these have been in linux-next for a while with no reported\n issues\"\n\n* tag \u0027driver-core-3.13-rc1\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/driver-core: (83 commits)\n sysfs: rename sysfs_assoc_lock and explain what it\u0027s about\n sysfs: use generic_file_llseek() for sysfs_file_operations\n sysfs: return correct error code on unimplemented mmap()\n mdio_bus: convert bus code to use dev_groups\n device: Make dev_WARN/dev_WARN_ONCE print device as well as driver name\n sysfs: separate out dup filename warning into a separate function\n sysfs: move sysfs_hash_and_remove() to fs/sysfs/dir.c\n sysfs: remove unused sysfs_get_dentry() prototype\n sysfs: honor bin_attr.attr.ignore_lockdep\n sysfs: merge sysfs_elem_bin_attr into sysfs_elem_attr\n devres: restore zeroing behavior of devres_alloc()\n sysfs: fix sysfs_write_file for bin file\n input: gameport: convert bus code to use dev_groups\n input: serio: remove bus usage of dev_attrs\n input: serio: use DEVICE_ATTR_RO()\n i2o: convert bus code to use dev_groups\n memstick: convert bus code to use dev_groups\n tifm: convert bus code to use dev_groups\n virtio: convert bus code to use dev_groups\n ipack: convert bus code to use dev_groups\n ...\n" }, { "commit": "74d332c13b2148ae934ea94dac1745ae92efe8e5", "tree": "4f8f287774deaba28fa4c85d085d587ae7515aaf", "parents": [ "b397f99921827e114d7f5600447e172a99c50165" ], "author": { "name": "Eric Dumazet", "email": "edumazet@google.com", "time": "Wed Oct 30 13:10:44 2013 -0700" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Sun Nov 03 23:19:00 2013 -0500" }, "message": "net: extend net_device allocation to vmalloc()\n\nJoby Poriyath provided a xen-netback patch to reduce the size of\nxenvif structure as some netdev allocation could fail under\nmemory pressure/fragmentation.\n\nThis patch is handling the problem at the core level, allowing\nany netdev structures to use vmalloc() if kmalloc() failed.\n\nAs vmalloc() adds overhead on a critical network path, add __GFP_REPEAT\nto kzalloc() flags to do this fallback only when really needed.\n\nSigned-off-by: Eric Dumazet \u003cedumazet@google.com\u003e\nReported-by: Joby Poriyath \u003cjoby.poriyath@citrix.com\u003e\nCc: Ben Hutchings \u003cbhutchings@solarflare.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "80c33ddd31d0e801953e02b7b003f395c1920e4e", "tree": "4f80ae0bdf146705c932aa039fcb48d47626478e", "parents": [ "e6cd988c27bb5918630db67b3526c9e78e786818" ], "author": { "name": "Jacob Keller", "email": "jacob.e.keller@intel.com", "time": "Sat Sep 21 05:05:39 2013 +0000" }, "committer": { "name": "Jeff Kirsher", "email": "jeffrey.t.kirsher@intel.com", "time": "Tue Oct 29 02:40:21 2013 -0700" }, "message": "net: add might_sleep() call to napi_disable\n\nnapi_disable uses an msleep() call to wait for outstanding napi work to be\nfinished after setting the disable bit. It does not always sleep incase there\nwas no outstanding work. This resulted in a rare bug in ixgbe_down operation\nwhere a napi_disable call took place inside of a local_bh_disable()d context.\nIn order to enable easier detection of future sleep while atomic BUGs, this\npatch adds a might_sleep() call, so that every use of napi_disable during\natomic context will be visible.\n\nSigned-off-by: Jacob Keller \u003cjacob.e.keller@intel.com\u003e\nCc: Eliezer Tamir \u003celiezer.tamir@linux.intel.com\u003e\nCc: Alexander Duyck \u003calexander.duyck@intel.com\u003e\nCc: Hyong-Youb Kim \u003chykim@myri.com\u003e\nCc: Amir Vadai \u003camirv@mellanox.com\u003e\nCc: Dmitry Kravkov \u003cdmitry@broadcom.com\u003e\nTested-by: Phil Schmitt \u003cphillip.j.schmitt@intel.com\u003e\nSigned-off-by: Jeff Kirsher \u003cjeffrey.t.kirsher@intel.com\u003e\n" }, { "commit": "7cc7c5e54b7128195a1403747a63971c3c3f8e25", "tree": "59d2d33955806c0bdca4d7752c24fa8548c8d8f4", "parents": [ "400dfd3ae899849b27d398ca7894e1b44430887f" ], "author": { "name": "Ben Hutchings", "email": "bhutchings@solarflare.com", "time": "Mon Oct 14 21:49:21 2013 +0100" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Fri Oct 18 00:13:21 2013 -0400" }, "message": "net: Delete trailing semi-colon from definition of netdev_WARN()\n\nMacro definitions should not normally end with a semi-colon, as this\nmakes it dangerous to use them an if...else statement. Happily this\nhas not happened yet.\n\nSigned-off-by: Ben Hutchings \u003cbhutchings@solarflare.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "3f79410c7c9c8ef33ccff60c61e1f1166f5ed64a", "tree": "ccfb1e1e68c140af1c2440b9b547506617d0d247", "parents": [ "1a49123b34346154ffd70f0b60dce007f1605676" ], "author": { "name": "Maxime Jayat", "email": "maxime@artisandeveloppeur.fr", "time": "Sat Oct 12 01:29:46 2013 +0200" }, "committer": { "name": "Jiri Kosina", "email": "jkosina@suse.cz", "time": "Mon Oct 14 15:31:06 2013 +0200" }, "message": "treewide: Fix common typo in \"identify\"\n\nCorrect common misspelling of \"identify\" as \"indentify\" throughout\nthe kernel\n\nSigned-off-by: Maxime Jayat \u003cmaxime@artisandeveloppeur.fr\u003e\nSigned-off-by: Jiri Kosina \u003cjkosina@suse.cz\u003e\n" }, { "commit": "53af53ae83fe960ceb9ef74cac7915e9088f4266", "tree": "0cd5309f2a931d3f210aa3d2025aae64f2276e8c", "parents": [ "b343ca84b4e3ba65508503333c923a797801a588", "9684d7b0dab3cf3a897edd85dca501d413888d56" ], "author": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Tue Oct 08 23:07:53 2013 -0400" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Tue Oct 08 23:07:53 2013 -0400" }, "message": "Merge git://git.kernel.org/pub/scm/linux/kernel/git/davem/net\n\nConflicts:\n\tinclude/linux/netdevice.h\n\tnet/core/sock.c\n\nTrivial merge issues.\n\nRemoval of \"extern\" for functions declaration in netdevice.h\nat the same time \"const\" was added to an argument.\n\nTwo parallel line additions in net/core/sock.c\n\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "5cde282938915f36a2e6769b51c24c4159654859", "tree": "9e364e2988fb3556313eda68eea6fb5655b6df1e", "parents": [ "d639feaaf3f40cd90b75a2fec5b7d5c3f96c2c88" ], "author": { "name": "Eric W. Biederman", "email": "ebiederm@xmission.com", "time": "Sat Oct 05 19:26:05 2013 -0700" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Mon Oct 07 15:23:14 2013 -0400" }, "message": "net: Separate the close_list and the unreg_list v2\n\nSeparate the unreg_list and the close_list in dev_close_many preventing\ndev_close_many from permuting the unreg_list. The permutations of the\nunreg_list have resulted in cases where the loopback device is accessed\nit has been freed in code such as dst_ifdown. Resulting in subtle memory\ncorruption.\n\nThis is the second bug from sharing the storage between the close_list\nand the unreg_list. The issues that crop up with sharing are\napparently too subtle to show up in normal testing or usage, so let\u0027s\nforget about being clever and use two separate lists.\n\nv2: Make all callers pass in a close_list to dev_close_many\n\nSigned-off-by: \"Eric W. Biederman\" \u003cebiederm@xmission.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "3573540cafa4296dd60f8be02f2aecaa31047525", "tree": "d1df5c4648f5cead63251abc03409cce50c40928", "parents": [ "5e8a402f831dbe7ee831340a91439e46f0d38acd" ], "author": { "name": "Michael S. Tsirkin", "email": "mst@redhat.com", "time": "Wed Oct 02 09:14:06 2013 +0300" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Mon Oct 07 12:29:26 2013 -0400" }, "message": "netif_set_xps_queue: make cpu mask const\n\nvirtio wants to pass in cpumask_of(cpu), make parameter\nconst to avoid build warnings.\n\nSigned-off-by: Michael S. Tsirkin \u003cmst@redhat.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "cb178190ab7d9c9b9fb90e36488ebd26ca004d32", "tree": "b2c7b9d6dda5600df21bb109eb7616b89039629d", "parents": [ "7b77d161ce7e480a660692cab079712fc0268ba2", "f629d208d27a22f495b7734eede585b5d207e912" ], "author": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Mon Sep 30 15:36:45 2013 -0400" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Mon Sep 30 15:36:45 2013 -0400" }, "message": "Merge branch \u002720130926_include_linux_networking_externs\u0027 of git://repo.or.cz/linux-2.6/trivial-mods\n\nConflicts:\n\tinclude/linux/netdevice.h\n\nMore extern removals from Joe Perches.\n\nMinor conflict with the dev_notify_flags changes which added a new\nargument to __dev_notify_flags().\n\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "a528c219df2e865e178c538c7178961dfed5a13c", "tree": "bea412291c7c60b6651cc16752556aa07f94f4e3", "parents": [ "e05849512662e789232dafed71ba65729f101e70" ], "author": { "name": "Nicolas Dichtel", "email": "nicolas.dichtel@6wind.com", "time": "Wed Sep 25 12:02:44 2013 +0200" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Mon Sep 30 15:08:12 2013 -0400" }, "message": "dev: update __dev_notify_flags() to send rtnl msg\n\nThis patch only prepares the next one, there is no functional change.\nNow, __dev_notify_flags() can also be used to notify flags changes via\nrtnetlink.\n\nSigned-off-by: Nicolas Dichtel \u003cnicolas.dichtel@6wind.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "f629d208d27a22f495b7734eede585b5d207e912", "tree": "1fd20f80eb03de507641f4069b39a07df3d57000", "parents": [ "7965bd4d71ef7cf1db00afb9e406ddfc13443c13" ], "author": { "name": "Joe Perches", "email": "joe@perches.com", "time": "Thu Sep 26 14:48:15 2013 -0700" }, "committer": { "name": "Joe Perches", "email": "joe@perches.com", "time": "Thu Sep 26 15:06:58 2013 -0700" }, "message": "[networking]device.h: Remove extern from function prototypes\n\nThere are a mix of function prototypes with and without extern\nin the kernel sources. Standardize on not using extern for\nfunction prototypes.\n\nFunction prototypes don\u0027t need to be written with extern.\nextern is assumed by the compiler. Its use is as unnecessary as\nusing auto to declare automatic/local variables in a block.\n\nSigned-off-by: Joe Perches \u003cjoe@perches.com\u003e\n" }, { "commit": "58292cbe6669d74498a5f08db13e57cb3bcfb81d", "tree": "be6ed2b906a9796c58be9ff21251e124bf01ddcb", "parents": [ "bcac3769ca6d6278f93afb6cc2b234d260ee5951" ], "author": { "name": "Tejun Heo", "email": "tj@kernel.org", "time": "Wed Sep 11 22:29:04 2013 -0400" }, "committer": { "name": "Greg Kroah-Hartman", "email": "gregkh@linuxfoundation.org", "time": "Thu Sep 26 14:50:01 2013 -0700" }, "message": "sysfs: make attr namespace interface less convoluted\n\nsysfs ns (namespace) implementation became more convoluted than\nnecessary while trying to hide ns information from visible interface.\nThe relatively recent attr ns support is a good example.\n\n* attr ns tag is determined by sysfs_ops-\u003enamespace() callback while\n dir tag is determined by kobj_type-\u003enamespace(). The placement is\n arbitrary.\n\n* Instead of performing operations with explicit ns tag, the namespace\n callback is routed through sysfs_attr_ns(), sysfs_ops-\u003enamespace(),\n class_attr_namespace(), class_attr-\u003enamespace(). It\u0027s not simpler\n in any sense. The only thing this convolution does is traversing\n the whole stack backwards.\n\nThe namespace callbacks are unncessary because the operations involved\nare inherently synchronous. The information can be provided in in\nstraight-forward top-down direction and reversing that direction is\nunnecessary and against basic design principles.\n\nThis backward interface is unnecessarily convoluted and hinders\nproperly separating out sysfs from driver model / kobject for proper\nlayering. This patch updates attr ns support such that\n\n* sysfs_ops-\u003enamespace() and class_attr-\u003enamespace() are dropped.\n\n* sysfs_{create|remove}_file_ns(), which take explicit @ns param, are\n added and sysfs_{create|remove}_file() are now simple wrappers\n around the ns aware functions.\n\n* ns handling is dropped from sysfs_chmod_file(). Nobody uses it at\n this point. sysfs_chmod_file_ns() can be added later if necessary.\n\n* Explicit @ns is propagated through class_{create|remove}_file_ns()\n and netdev_class_{create|remove}_file_ns().\n\n* driver/net/bonding which is currently the only user of attr\n namespace is updated to use netdev_class_{create|remove}_file_ns()\n with @bh-\u003enet as the ns tag instead of using the namespace callback.\n\nThis patch should be an equivalent conversion without any functional\ndifference. It makes the code easier to follow, reduces lines of code\na bit and helps proper separation and layering.\n\nSigned-off-by: Tejun Heo \u003ctj@kernel.org\u003e\nCc: Eric W. Biederman \u003cebiederm@xmission.com\u003e\nCc: Kay Sievers \u003ckay@vrfy.org\u003e\nAcked-by: David S. Miller \u003cdavem@davemloft.net\u003e\nSigned-off-by: Greg Kroah-Hartman \u003cgregkh@linuxfoundation.org\u003e\n" }, { "commit": "b6ccba4c681fdaf0070e580bf951badf7edc860b", "tree": "69182002ffd5eba43619b6f7a1b8fe1273f59296", "parents": [ "70039aa7c6c182c488ec23a9669d9f6b21aebe16" ], "author": { "name": "Veaceslav Falico", "email": "vfalico@redhat.com", "time": "Wed Sep 25 09:20:23 2013 +0200" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Sep 26 16:02:06 2013 -0400" }, "message": "net: add a possibility to get private from netdev_adjacent-\u003elist\n\nIt will be useful to get first/last element.\n\nCC: \"David S. Miller\" \u003cdavem@davemloft.net\u003e\nCC: Eric Dumazet \u003cedumazet@google.com\u003e\nCC: Jiri Pirko \u003cjiri@resnulli.us\u003e\nCC: Alexander Duyck \u003calexander.h.duyck@intel.com\u003e\nSigned-off-by: Veaceslav Falico \u003cvfalico@redhat.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "31088a113c2a948856ed2047d8c21c217b13e85d", "tree": "15847ff3e7b82f578f4a06cb801a5dccc0fd961e", "parents": [ "46bb4807b5d95a049b008efd639ff8942970d815" ], "author": { "name": "Veaceslav Falico", "email": "vfalico@redhat.com", "time": "Wed Sep 25 09:20:12 2013 +0200" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Sep 26 16:02:04 2013 -0400" }, "message": "net: add for_each iterators through neighbour lower link\u0027s private\n\nAdd a possibility to iterate through netdev_adjacent\u0027s private, currently\nonly for lower neighbours.\n\nAdd both RCU and RTNL/other locking variants of iterators, and make the\nnon-rcu variant to be safe from removal.\n\nCC: \"David S. Miller\" \u003cdavem@davemloft.net\u003e\nCC: Eric Dumazet \u003cedumazet@google.com\u003e\nCC: Jiri Pirko \u003cjiri@resnulli.us\u003e\nCC: Alexander Duyck \u003calexander.h.duyck@intel.com\u003e\nSigned-off-by: Veaceslav Falico \u003cvfalico@redhat.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "402dae9614557296e84543008a8e582c28fb1db3", "tree": "80d4da7981cb8f5f68340e11d2f6449117a15794", "parents": [ "5249dec7380cb64928d2ae6201028b4da1dceb1e" ], "author": { "name": "Veaceslav Falico", "email": "vfalico@redhat.com", "time": "Wed Sep 25 09:20:09 2013 +0200" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Sep 26 16:02:04 2013 -0400" }, "message": "net: add netdev_adjacent-\u003eprivate and allow to use it\n\nCurrently, even though we can access any linked device, we can\u0027t attach\nanything to it, which is vital to properly manage them.\n\nTo fix this, add a new void *private to netdev_adjacent and functions\nsetting/getting it (per link), so that we can save, per example, bonding\u0027s\nslave structures there, per slave device.\n\nnetdev_master_upper_dev_link_private(dev, upper_dev, private) links dev to\nupper dev and populates the neighbour link only with private.\n\nnetdev_lower_dev_get_private{,_rcu}() returns the private, if found.\n\nCC: \"David S. Miller\" \u003cdavem@davemloft.net\u003e\nCC: Eric Dumazet \u003cedumazet@google.com\u003e\nCC: Jiri Pirko \u003cjiri@resnulli.us\u003e\nCC: Alexander Duyck \u003calexander.h.duyck@intel.com\u003e\nSigned-off-by: Veaceslav Falico \u003cvfalico@redhat.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "2f268f129c2d1a05d297fe3ee34d393f862d2b22", "tree": "5373446c2933311d0aeeeaa85a6823cb621710d4", "parents": [ "7863c054d1b4fd35f76c13e2e918f7f483fe48f4" ], "author": { "name": "Veaceslav Falico", "email": "vfalico@redhat.com", "time": "Wed Sep 25 09:20:07 2013 +0200" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Sep 26 16:02:04 2013 -0400" }, "message": "net: add adj_list to save only neighbours\n\nCurrently, we distinguish neighbours (first-level linked devices) from\nnon-neighbours by the neighbour bool in the netdev_adjacent. This could be\nquite time-consuming in case we would like to traverse *only* through\nneighbours - cause we\u0027d have to traverse through all devices and check for\nthis flag, and in a (quite common) scenario where we have lots of vlans on\ntop of bridge, which is on top of a bond - the bonding would have to go\nthrough all those vlans to get its upper neighbour linked devices.\n\nThis situation is really unpleasant, cause there are already a lot of cases\nwhen a device with slaves needs to go through them in hot path.\n\nTo fix this, introduce a new upper/lower device lists structure -\nadj_list, which contains only the neighbours. It works always in\npair with the all_adj_list structure (renamed from upper/lower_dev_list),\ni.e. both of them contain the same links, only that all_adj_list contains\nalso non-neighbour device links. It\u0027s really a small change visible,\ncurrently, only for __netdev_adjacent_dev_insert/remove(), and doesn\u0027t\nchange the main linked logic at all.\n\nAlso, add some comments a fix a name collision in\nnetdev_for_each_upper_dev_rcu() and rework the naming by the following\nrules:\n\nnetdev_(all_)(upper|lower)_*\n\nIf \"all_\" is present, then we work with the whole list of upper/lower\ndevices, otherwise - only with direct neighbours. Uninline functions - to\nget better stack traces.\n\nCC: \"David S. Miller\" \u003cdavem@davemloft.net\u003e\nCC: Eric Dumazet \u003cedumazet@google.com\u003e\nCC: Jiri Pirko \u003cjiri@resnulli.us\u003e\nCC: Alexander Duyck \u003calexander.h.duyck@intel.com\u003e\nCC: Cong Wang \u003camwang@redhat.com\u003e\nSigned-off-by: Veaceslav Falico \u003cvfalico@redhat.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "35e4237973665c8a1ad4e3f7a7cb87573deaa24a", "tree": "ce2529c137ec6b1da4f81aa77c2882706bd0c80c", "parents": [ "5c44bbdab5fd7831859cf0d38626317ffcf39558" ], "author": { "name": "Joseph Gasparakis", "email": "joseph.gasparakis@intel.com", "time": "Fri Sep 13 07:34:13 2013 -0700" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Sun Sep 15 22:18:13 2013 -0400" }, "message": "vxlan: Fix sparse warnings\n\nThis patch fixes sparse warnings when incorrectly handling the port number\nand using int instead of unsigned int iterating through \u0026vn-\u003esock_list[].\nKeeping the port as __be16 also makes things clearer wrt endianess.\nAlso, it was pointed out that vxlan_get_rx_port() had unnecessary checks\nwhich got removed.\n\nSigned-off-by: Joseph Gasparakis \u003cjoseph.gasparakis@intel.com\u003e\nSigned-off-by: Jeff Kirsher \u003cjeffrey.t.kirsher@intel.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "0042d0c840c616186a5b09207a0e77fab7581db3", "tree": "7046db91faa3d51529ba71484d7f1594ea73be52", "parents": [ "635ad31002fad736af85adfc96bb9408fb813a7b" ], "author": { "name": "Florian Fainelli", "email": "f.fainelli@gmail.com", "time": "Fri Sep 06 16:58:00 2013 +0100" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Fri Sep 06 14:43:49 2013 -0400" }, "message": "net: add documentation for BQL helpers\n\nProvide a kernel-doc comment documentation for the BQL helpers:\n- netdev_sent_queue\n- netdev_completed_queue\n- netdev_reset_queue\n\nSimilarly to how it is done for the other functions, the documentation\nonly covers the function operating on struct net_device and not struct\nnetdev_queue.\n\nSigned-off-by: Florian Fainelli \u003cf.fainelli@gmail.com\u003e\nAcked-by: Eric Dumazet \u003cedumazet@google.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "53cf527513eed6e7170e9dceacd198f9267171b0", "tree": "a3492c0373949048084213dc4bd02efd130aa5cf", "parents": [ "eef23b53985f8e08256f24d76e576d69fb0d44d3" ], "author": { "name": "Joseph Gasparakis", "email": "joseph.gasparakis@intel.com", "time": "Wed Sep 04 02:13:38 2013 -0700" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Sep 05 12:44:30 2013 -0400" }, "message": "vxlan: Notify drivers for listening UDP port changes\n\nThis patch adds two more ndo ops: ndo_add_rx_vxlan_port() and\nndo_del_rx_vxlan_port().\n\nDrivers can get notifications through the above functions about changes\nof the UDP listening port of VXLAN. Also, when physical ports come up,\nnow they can call vxlan_get_rx_port() in order to obtain the port number(s)\nof the existing VXLAN interface in case they already up before them.\n\nThis information about the listening UDP port would be used for VXLAN\nrelated offloads.\n\nA big thank you to John Fastabend (john.r.fastabend@intel.com) for his\ninput and his suggestions on this patch set.\n\nCC: John Fastabend \u003cjohn.r.fastabend@intel.com\u003e\nCC: Stephen Hemminger \u003cstephen@networkplumber.org\u003e\nSigned-off-by: Joseph Gasparakis \u003cjoseph.gasparakis@intel.com\u003e\nSigned-off-by: Jeff Kirsher \u003cjeffrey.t.kirsher@intel.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "8b5be8561b804edf6b58fc27edbccf1d45863e08", "tree": "969e67314edbc253f3bcb0460b7a16d93de619c5", "parents": [ "48311f46853c0361f9fba7e0e6bb1652d633c049" ], "author": { "name": "Veaceslav Falico", "email": "vfalico@redhat.com", "time": "Wed Aug 28 23:25:08 2013 +0200" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Aug 29 16:19:42 2013 -0400" }, "message": "net: add netdev_for_each_upper_dev_rcu()\n\nThe new macro netdev_for_each_upper_dev_rcu(dev, upper, iter) iterates\nthrough the dev-\u003eupper_dev_list starting from the first element, using\nthe netdev_upper_get_next_dev_rcu(dev, \u0026iter).\n\nMust be called under RCU read lock.\n\nCC: \"David S. Miller\" \u003cdavem@davemloft.net\u003e\nCC: Eric Dumazet \u003cedumazet@google.com\u003e\nCC: Jiri Pirko \u003cjiri@resnulli.us\u003e\nCC: Alexander Duyck \u003calexander.h.duyck@intel.com\u003e\nCC: Cong Wang \u003camwang@redhat.com\u003e\nSigned-off-by: Veaceslav Falico \u003cvfalico@redhat.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "5d261913ca3daf6c2d21d38924235667b3d07c40", "tree": "f15af2dea3316393cea047c637f434aa04db7726", "parents": [ "aa9d85605f5ab070b64842b3eba797cf81698ae1" ], "author": { "name": "Veaceslav Falico", "email": "vfalico@redhat.com", "time": "Wed Aug 28 23:25:05 2013 +0200" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Aug 29 16:19:42 2013 -0400" }, "message": "net: add lower_dev_list to net_device and make a full mesh\n\nThis patch adds lower_dev_list list_head to net_device, which is the same\nas upper_dev_list, only for lower devices, and begins to use it in the same\nway as the upper list.\n\nIt also changes the way the whole adjacent device lists work - now they\ncontain *all* of upper/lower devices, not only the first level. The first\nlevel devices are distinguished by the bool neighbour field in\nnetdev_adjacent, also added by this patch.\n\nThere are cases when a device can be added several times to the adjacent\nlist, the simplest would be:\n\n /---- eth0.10 ---\\\neth0-\t\t --- bond0\n \\---- eth0.20 ---/\n\nwhere both bond0 and eth0 \u0027see\u0027 each other in the adjacent lists two times.\nTo avoid duplication of netdev_adjacent structures ref_nr is being kept as\nthe number of times the device was added to the list.\n\nThe \u0027full view\u0027 is achieved by adding, on link creation, all of the\nupper_dev\u0027s upper_dev_list devices as upper devices to all of the\nlower_dev\u0027s lower_dev_list devices (and to the lower_dev itself), and vice\nversa. On unlink they are removed using the same logic.\n\nI\u0027ve tested it with thousands vlans/bonds/bridges, everything works ok and\nno observable lags even on a huge number of interfaces.\n\nMemory footprint for 128 devices interconnected with each other via both\nupper and lower (which is impossible, but for the comparison) lists would be:\n\n128*128*2*sizeof(netdev_adjacent) \u003d 1.5MB\n\nbut in the real world we usualy have at most several devices with slaves\nand a lot of vlans, so the footprint will be much lower.\n\nCC: \"David S. Miller\" \u003cdavem@davemloft.net\u003e\nCC: Eric Dumazet \u003cedumazet@google.com\u003e\nCC: Jiri Pirko \u003cjiri@resnulli.us\u003e\nCC: Alexander Duyck \u003calexander.h.duyck@intel.com\u003e\nCC: Cong Wang \u003camwang@redhat.com\u003e\nSigned-off-by: Veaceslav Falico \u003cvfalico@redhat.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "0e76a3a587fc7abda2badf249053b427baad255e", "tree": "646a1da67dc25654552028225ca4b7a0a8043cc9", "parents": [ "fba3679d34511c42bf452e89dda457a1219eb43a", "72a67a94bcba71a5fddd6b3596a20604d2b5dcd6" ], "author": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Sat Aug 03 21:36:46 2013 -0700" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Sat Aug 03 21:36:46 2013 -0700" }, "message": "Merge git://git.kernel.org/pub/scm/linux/kernel/git/davem/net\n\nMerge net into net-next to setup some infrastructure Eric\nDumazet needs for usbnet changes.\n\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "e0d1095ae3405404d247afb00233ef837d58da83", "tree": "1f83aed50dbe9298ca0cdc5a0d51a3750f0fd720", "parents": [ "dfcefb0be1231982784df2152213103ad33c1cfd" ], "author": { "name": "Cong Wang", "email": "amwang@redhat.com", "time": "Thu Aug 01 11:10:25 2013 +0800" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Aug 01 15:11:17 2013 -0700" }, "message": "net: rename CONFIG_NET_LL_RX_POLL to CONFIG_NET_RX_BUSY_POLL\n\nEliezer renames several *ll_poll to *busy_poll, but forgets\nCONFIG_NET_LL_RX_POLL, so in case of confusion, rename it too.\n\nCc: Eliezer Tamir \u003celiezer.tamir@linux.intel.com\u003e\nCc: David S. Miller \u003cdavem@davemloft.net\u003e\nSigned-off-by: Cong Wang \u003camwang@redhat.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "66b52b0dc82c5c88d769dc1c7d44cf45d0deb07c", "tree": "2fa19024bd1afefcc0fc4a09ea7ae1b6c095edcd", "parents": [ "9ad8fef6c61a9dfcaa8444e526020de88c540fa9" ], "author": { "name": "Jiri Pirko", "email": "jiri@resnulli.us", "time": "Mon Jul 29 18:16:49 2013 +0200" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Tue Jul 30 17:31:24 2013 -0700" }, "message": "net: add ndo to get id of physical port of the device\n\nThis patch adds a ndo for getting physical port of the device. Driver\nwhich is aware of being virtual function of some physical port should\nimplement this ndo. This is applicable not only for IOV, but for other\nsolutions (NPAR, multichannel) as well. Basically if there is possible\nto have multiple netdevs on the single hw port.\n\nSigned-off-by: Jiri Pirko \u003cjiri@resnulli.us\u003e\nAcked-by: Ben Hutchings \u003cbhutchings@solarflare.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "18afa4b028b46f8b45ca64f94aefe717c297b07d", "tree": "9ef226f810b232cbfc1166340abf006b29c959f2", "parents": [ "c9bee3b7fdecb0c1d070c7b54113b3bdfb9a3d36" ], "author": { "name": "Thomas Gleixner", "email": "tglx@linutronix.de", "time": "Tue Jul 23 16:13:17 2013 +0200" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Wed Jul 24 17:57:26 2013 -0700" }, "message": "net: Make devnet_rename_seq static\n\nNo users outside net/core/dev.c.\n\nSigned-off-by: Thomas Gleixner \u003ctglx@linutronix.de\u003e\nAcked-by: Eric Dumazet \u003cedumazet@google.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "4aa5dee4d9997879adff858514844efab5a15a01", "tree": "f3f3bf83f34dd832c8e163095c03e6aff29f20df", "parents": [ "fc423ff00df3a19554414eed80aef9de9b50313e" ], "author": { "name": "Jiri Pirko", "email": "jiri@resnulli.us", "time": "Sat Jul 20 12:13:53 2013 +0200" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Tue Jul 23 16:52:47 2013 -0700" }, "message": "net: convert resend IGMP to notifier event\n\nUntil now, bond_resend_igmp_join_requests() looks for vlans attached to\nbonding device, bridge where bonding act as port manually. It does not\ncare of other scenarios, like stacked bonds or team device above. Make\nthis more generic and use netdev notifier to propagate the event to\nupper devices and to actually call ip_mc_rejoin_groups().\n\nSigned-off-by: Jiri Pirko \u003cjiri@resnulli.us\u003e\nAcked-by: Veaceslav Falico \u003cvfalico@redhat.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "8b80cda536ea9bceec0364e897868a30ee13b992", "tree": "75edd97f04e379e6afedd67beef81eb80ba0adaf", "parents": [ "076bb0c82a44fbe46fe2c8527a5b5b64b69f679d" ], "author": { "name": "Eliezer Tamir", "email": "eliezer.tamir@linux.intel.com", "time": "Wed Jul 10 17:13:26 2013 +0300" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Wed Jul 10 17:08:27 2013 -0700" }, "message": "net: rename ll methods to busy-poll\n\nRename ndo_ll_poll to ndo_busy_poll.\nRename sk_mark_ll to sk_mark_napi_id.\nRename skb_mark_ll to skb_mark_napi_id.\nCorrect all useres of these functions.\nUpdate comments and defines in include/net/busy_poll.h\n\nSigned-off-by: Eliezer Tamir \u003celiezer.tamir@linux.intel.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "0c1072ae0242fbdffd9a0bba36e7a7033d287f9c", "tree": "e0f4dbdbf5078d4a707911177e7bdc17a70bdce5", "parents": [ "c50cd357887acf9fd7af3a5d492911bd825555a2", "8bb495e3f02401ee6f76d1b1d77f3ac9f079e376" ], "author": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Wed Jul 03 14:50:41 2013 -0700" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Wed Jul 03 14:55:13 2013 -0700" }, "message": "Merge git://git.kernel.org/pub/scm/linux/kernel/git/davem/net\n\nConflicts:\n\tdrivers/net/ethernet/freescale/fec_main.c\n\tdrivers/net/ethernet/renesas/sh_eth.c\n\tnet/ipv4/gre.c\n\nThe GRE conflict is between a bug fix (kfree_skb --\u003e kfree_skb_list)\nand the splitting of the gre.c code into seperate files.\n\nThe FEC conflict was two sets of changes adding ethtool support code\nin an \"!CONFIG_M5272\" CPP protected block.\n\nFinally the sh_eth.c conflict was between one commit add bits set\nin the .eesr_err_check mask whilst another commit removed the\n.tx_error_check member and assignments.\n\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "5dbe7c178d3f0a4634f088d9e729f1909b9ddcd1", "tree": "8945b6c5125b57cee5f36e903fc995e58664a639", "parents": [ "6d446ec32f169c6a5d9bc90684a8082a6cbe90f6" ], "author": { "name": "Nicolas Schichan", "email": "nschichan@freebox.fr", "time": "Wed Jun 26 17:23:42 2013 +0200" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Wed Jun 26 13:42:54 2013 -0700" }, "message": "net: fix kernel deadlock with interface rename and netdev name retrieval.\n\nWhen the kernel (compiled with CONFIG_PREEMPT\u003dn) is performing the\nrename of a network interface, it can end up waiting for a workqueue\nto complete. If userland is able to invoke a SIOCGIFNAME ioctl or a\nSO_BINDTODEVICE getsockopt in between, the kernel will deadlock due to\nthe fact that read_secklock_begin() will spin forever waiting for the\nwriter process (the one doing the interface rename) to update the\ndevnet_rename_seq sequence.\n\nThis patch fixes the problem by adding a helper (netdev_get_name())\nand using it in the code handling the SIOCGIFNAME ioctl and\nSO_BINDTODEVICE setsockopt.\n\nThe netdev_get_name() helper uses raw_seqcount_begin() to avoid\nspinning forever, waiting for devnet_rename_seq-\u003esequence to become\neven. cond_resched() is used in the contended case, before retrying\nthe access to give the writer process a chance to finish.\n\nThe use of raw_seqcount_begin() will incur some unneeded work in the\nreader process in the contended case, but this is better than\ndeadlocking the system.\n\nSigned-off-by: Nicolas Schichan \u003cnschichan@freebox.fr\u003e\nAcked-by: Eric Dumazet \u003cedumazet@google.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "1d8faf48c74b8329a0322dc4b2a2030ae5003c86", "tree": "63471c7ee9379f70e30b27b82cf6f155683b4468", "parents": [ "3dc6475c0c9e55ac7f053ad6b8b398e779954545" ], "author": { "name": "Rony Efraim", "email": "ronye@mellanox.com", "time": "Thu Jun 13 13:19:10 2013 +0300" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Jun 13 17:51:04 2013 -0700" }, "message": "net/core: Add VF link state control\n\nAdd netlink directives and ndo entry to allow for controling\nVF link, which can be in one of three states:\n\nAuto - VF link state reflects the PF link state (default)\n\nUp - VF link state is up, traffic from VF to VF works even if\nthe actual PF link is down\n\nDown - VF link state is down, no traffic from/to this VF, can be of\nuse while configuring the VF\n\nSigned-off-by: Rony Efraim \u003cronye@mellanox.com\u003e\nSigned-off-by: Or Gerlitz \u003cogerlitz@mellanox.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "5f121b9a83b499a61ed44e5ba619c7de8f7271ad", "tree": "6765d95f468e5c8c884af26f5fbfd2c6428d255d", "parents": [ "85f16525a2eb66e6092cbd8dcf42371df8334ed0" ], "author": { "name": "Willem de Bruijn", "email": "willemb@google.com", "time": "Thu Jun 13 15:29:38 2013 -0400" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Jun 13 17:13:05 2013 -0700" }, "message": "net-rps: fixes for rps flow limit\n\nCaught by sparse:\n- __rcu: missing annotation to sd-\u003eflow_limit\n- __user: direct access in cpumask_scnprintf\n\nAlso\n- add endline character when printing bitmap if room in buffer\n- avoid bucket overflow by reducing FLOW_LIMIT_HISTORY\n\nThe last item warrants some explanation. The hashtable buckets are\nsubject to overflow if FLOW_LIMIT_HISTORY is larger than or equal\nto bucket size, since all packets may end up in a single bucket. The\ncurrent (rather arbitrary) history value of 256 happens to match the\nbuffer size (u8).\n\nAs a result, with a single flow, the first 128 packets are accepted\n(correct), the second 128 packets dropped (correct) and then the\nhistory[] array has filled, so that each subsequent new packet\ncauses an increment in the bucket for new_flow plus a decrement\nfor old_flow: a steady state.\n\nThis is fine if packets are dropped, as the steady state goes away\nas soon as a mix of traffic reappears. But, because the 256th packet\noverflowed the bucket to 0: no packets are dropped.\n\nInstead of explicitly adding an overflow check, this patch changes\nFLOW_LIMIT_HISTORY to never be able to overflow a single bucket.\n\nReported-by: Fengguang Wu \u003cfengguang.wu@intel.com\u003e\n(first item)\n\nSigned-off-by: Willem de Bruijn \u003cwillemb@google.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "dffebd2c5cd528a136b276a2a75c56222312d7a4", "tree": "05364403f2245264ac2eafa4b26ad1b81f4a1e43", "parents": [ "32766fff24a761079636bb96c389411d24a334c6" ], "author": { "name": "Narendra K", "email": "narendra_k@dell.com", "time": "Mon Jun 10 19:34:03 2013 +0530" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Wed Jun 12 03:03:47 2013 -0700" }, "message": "doc:networking: Update comment for dev_id field in netdevice.h\n\nThis patch updates the comment for \u0027dev_id\u0027 field in\n\u0027include/linux/netdevice.h\u0027 to reflect the intended\nusage of \u0027dev_id\u0027.\n\nReferences: http://marc.info/?l\u003dlinux-netdev\u0026m\u003d136992115300526\u0026w\u003d2\nReferences: http://marc.info/?l\u003dlinux-netdev\u0026m\u003d137062569014612\u0026w\u003d2\n\nSigned-off-by: Narendra K \u003cnarendra_k@dell.com\u003e\nReviewed-by: Ben Hutchings \u003cbhutchings@solarflare.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "060212928670593fb89243640bf05cf89560b023", "tree": "32ae5c7be094983fb06430dbe09d36e2b317ca9c", "parents": [ "af12fa6e46aa651e7b86a4c4117b562518fef184" ], "author": { "name": "Eliezer Tamir", "email": "eliezer.tamir@linux.intel.com", "time": "Mon Jun 10 11:39:50 2013 +0300" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Mon Jun 10 21:22:35 2013 -0700" }, "message": "net: add low latency socket poll\n\nAdds an ndo_ll_poll method and the code that supports it.\nThis method can be used by low latency applications to busy-poll\nEthernet device queues directly from the socket code.\nsysctl_net_ll_poll controls how many microseconds to poll.\nDefault is zero (disabled).\nIndividual protocol support will be added by subsequent patches.\n\nSigned-off-by: Alexander Duyck \u003calexander.h.duyck@intel.com\u003e\nSigned-off-by: Jesse Brandeburg \u003cjesse.brandeburg@intel.com\u003e\nSigned-off-by: Eliezer Tamir \u003celiezer.tamir@linux.intel.com\u003e\nAcked-by: Eric Dumazet \u003cedumazet@google.com\u003e\nTested-by: Willem de Bruijn \u003cwillemb@google.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "af12fa6e46aa651e7b86a4c4117b562518fef184", "tree": "1425ef11083e66af192c33abb3b1fa3b04b26f4b", "parents": [ "6f00a0229627ca189529cad3f9154ac2f9e5c7db" ], "author": { "name": "Eliezer Tamir", "email": "eliezer.tamir@linux.intel.com", "time": "Mon Jun 10 11:39:41 2013 +0300" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Mon Jun 10 21:22:35 2013 -0700" }, "message": "net: add napi_id and hash\n\nAdds a napi_id and a hashing mechanism to lookup a napi by id.\nThis will be used by subsequent patches to implement low latency\nEthernet device polling.\nBased on a code sample by Eric Dumazet.\n\nSigned-off-by: Eliezer Tamir \u003celiezer.tamir@linux.intel.com\u003e\nSigned-off-by: Eric Dumazet \u003cedumazet@google.com\u003e\nTested-by: Willem de Bruijn \u003cwillemb@google.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "75538c2b85cf22eb9af6adfaf26ed7219025adeb", "tree": "f6791d0d285ff7e716fe8f225c4648e62a34421e", "parents": [ "06ecf24bdf2b7afc6c8fd13de6dba2a96dd331b6" ], "author": { "name": "Cong Wang", "email": "amwang@redhat.com", "time": "Wed May 29 11:30:50 2013 +0800" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Tue May 28 21:58:54 2013 -0700" }, "message": "net: always pass struct netdev_notifier_info to netdevice notifiers\n\ncommit 351638e7deeed2ec8ce451b53d3 (net: pass info struct via netdevice notifier)\nbreaks booting of my KVM guest, this is due to we still forget to pass\nstruct netdev_notifier_info in several places. This patch completes it.\n\nCc: Jiri Pirko \u003cjiri@resnulli.us\u003e\nCc: David S. Miller \u003cdavem@davemloft.net\u003e\nSigned-off-by: Cong Wang \u003camwang@redhat.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "be9efd3653284f2827fd82861e8e9db9a8f726e1", "tree": "601a02b498c24c40a2712feef0c0cfca544fc2c1", "parents": [ "351638e7deeed2ec8ce451b53d33921b3da68f83" ], "author": { "name": "Jiri Pirko", "email": "jiri@resnulli.us", "time": "Tue May 28 01:30:22 2013 +0000" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Tue May 28 13:11:01 2013 -0700" }, "message": "net: pass changed flags along with NETDEV_CHANGE event\n\nUse new netdevice notifier infrastructure to pass along changed flags.\n\nSigned-off-by: Timo Teräs \u003ctimo.teras@iki.fi\u003e\nSigned-off-by: Jiri Pirko \u003cjiri@resnulli.us\u003e\n\nv2-\u003ev3: shortened notifier_info struct name\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "351638e7deeed2ec8ce451b53d33921b3da68f83", "tree": "175dfff289b5e3baecffbc7e97d1884e9a18345c", "parents": [ "b1098bbe1b24d5d90cff92fbd716d2ef4bed2cff" ], "author": { "name": "Jiri Pirko", "email": "jiri@resnulli.us", "time": "Tue May 28 01:30:21 2013 +0000" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Tue May 28 13:11:01 2013 -0700" }, "message": "net: pass info struct via netdevice notifier\n\nSo far, only net_device * could be passed along with netdevice notifier\nevent. This patch provides a possibility to pass custom structure\nable to provide info that event listener needs to know.\n\nSigned-off-by: Jiri Pirko \u003cjiri@resnulli.us\u003e\n\nv2-\u003ev3: fix typo on simeth\n\tshortened dev_getter\n\tshortened notifier_info struct name\nv1-\u003ev2: fix notifier_call parameter in call_netdevice_notifier()\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "0d89d2035fe063461a5ddb609b2c12e7fb006e44", "tree": "cc7fd6947941ba8e7cb63d5e190e14f48503ca37", "parents": [ "1a37e412a0225fcba5587f24c0dfc7636efc8b69" ], "author": { "name": "Simon Horman", "email": "horms@verge.net.au", "time": "Thu May 23 21:02:52 2013 +0000" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Mon May 27 22:50:59 2013 -0700" }, "message": "MPLS: Add limited GSO support\n\nIn the case where a non-MPLS packet is received and an MPLS stack is\nadded it may well be the case that the original skb is GSO but the\nNIC used for transmit does not support GSO of MPLS packets.\n\nThe aim of this code is to provide GSO in software for MPLS packets\nwhose skbs are GSO.\n\nSKB Usage:\n\nWhen an implementation adds an MPLS stack to a non-MPLS packet it should do\nthe following to skb metadata:\n\n* Set skb-\u003einner_protocol to the old non-MPLS ethertype of the packet.\n skb-\u003einner_protocol is added by this patch.\n\n* Set skb-\u003eprotocol to the new MPLS ethertype of the packet.\n\n* Set skb-\u003enetwork_header to correspond to the\n end of the L3 header, including the MPLS label stack.\n\nI have posted a patch, \"[PATCH v3.29] datapath: Add basic MPLS support to\nkernel\" which adds MPLS support to the kernel datapath of Open vSwtich.\nThat patch sets the above requirements in datapath/actions.c:push_mpls()\nand was used to exercise this code. The datapath patch is against the Open\nvSwtich tree but it is intended that it be added to the Open vSwtich code\npresent in the mainline Linux kernel at some point.\n\nFeatures:\n\nI believe that the approach that I have taken is at least partially\nconsistent with the handling of other protocols. Jesse, I understand that\nyou have some ideas here. I am more than happy to change my implementation.\n\nThis patch adds dev-\u003empls_features which may be used by devices\nto advertise features supported for MPLS packets.\n\nA new NETIF_F_MPLS_GSO feature is added for devices which support\nhardware MPLS GSO offload. Currently no devices support this\nand MPLS GSO always falls back to software.\n\nAlternate Implementation:\n\nOne possible alternate implementation is to teach netif_skb_features()\nand skb_network_protocol() about MPLS, in a similar way to their\nunderstanding of VLANs. I believe this would avoid the need\nfor net/mpls/mpls_gso.c and in particular the calls to\n__skb_push() and __skb_push() in mpls_gso_segment().\n\nI have decided on the implementation in this patch as it should\nnot introduce any overhead in the case where mpls_gso is not compiled\ninto the kernel or inserted as a module.\n\nMPLS GSO suggested by Jesse Gross.\nBased in part on \"v4 GRE: Add TCP segmentation offload for GRE\"\nby Pravin B Shelar.\n\nCc: Jesse Gross \u003cjesse@nicira.com\u003e\nCc: Pravin B Shelar \u003cpshelar@nicira.com\u003e\nSigned-off-by: Simon Horman \u003chorms@verge.net.au\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "42e52bf9e3ae80fd44b21ddfcd64c54e6db2ff76", "tree": "4d395ed55ed5e36d0b54ea49329e797c06783f9f", "parents": [ "8513fbd880093f00a47e85a552f14ca2de8d84d6" ], "author": { "name": "Jiri Pirko", "email": "jiri@resnulli.us", "time": "Sat May 25 04:12:10 2013 +0000" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Sat May 25 23:12:19 2013 -0700" }, "message": "net: add netnotifier event for upper device change\n\nNow when upper device is changed, event is not propagated via RT Netlink\nto userspace. Userspace might never now about the change. Fix this by\nadding upper-device-change notifier event.\n\nSigned-off-by: Jiri Pirko \u003cjiri@resnulli.us\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "e6ff4c75f9095f61b3a66c2a78e47b62864022dd", "tree": "425ea9463cbec0b1975b8a33d9a56817143055d0", "parents": [ "ee9c799c231324de681eb21e06d8bf4842768b75", "0e255f1c0c9add2f0c920240ac4cadc28ae274c3" ], "author": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Fri May 24 16:48:28 2013 -0700" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Fri May 24 16:48:28 2013 -0700" }, "message": "Merge git://git.kernel.org/pub/scm/linux/kernel/git/davem/net\n\nMerge net into net-next because some upcoming net-next changes\nbuild on top of bug fixes that went into net.\n\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "99bbc70741903c063b3ccad90a3e06fc55df9245", "tree": "a3377d2461242bf1134464ce3fe6d69f82c907c2", "parents": [ "4a5bddf7ea6b6c5916eccbc2fa1950555073ff48" ], "author": { "name": "Willem de Bruijn", "email": "willemb@google.com", "time": "Mon May 20 04:02:32 2013 +0000" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Mon May 20 13:48:04 2013 -0700" }, "message": "rps: selective flow shedding during softnet overflow\n\nA cpu executing the network receive path sheds packets when its input\nqueue grows to netdev_max_backlog. A single high rate flow (such as a\nspoofed source DoS) can exceed a single cpu processing rate and will\ndegrade throughput of other flows hashed onto the same cpu.\n\nThis patch adds a more fine grained hashtable. If the netdev backlog\nis above a threshold, IRQ cpus track the ratio of total traffic of\neach flow (using 4096 buckets, configurable). The ratio is measured\nby counting the number of packets per flow over the last 256 packets\nfrom the source cpu. Any flow that occupies a large fraction of this\n(set at 50%) will see packet drop while above the threshold.\n\nTested:\nSetup is a muli-threaded UDP echo server with network rx IRQ on cpu0,\nkernel receive (RPS) on cpu0 and application threads on cpus 2--7\neach handling 20k req/s. Throughput halves when hit with a 400 kpps\nantagonist storm. With this patch applied, antagonist overload is\ndropped and the server processes its complete load.\n\nThe patch is effective when kernel receive processing is the\nbottleneck. The above RPS scenario is a extreme, but the same is\nreached with RFS and sufficient kernel processing (iptables, packet\nsocket tap, ..).\n\nSigned-off-by: Willem de Bruijn \u003cwillemb@google.com\u003e\nAcked-by: Eric Dumazet \u003cedumazet@google.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "b0ce3508b25ea6fa10ae3ca254de1d695b521702", "tree": "d5158ba0fe71d3807c563148c6cddb3757e7f348", "parents": [ "3aefe2b4a8003517d6e15112f806fd4069785389" ], "author": { "name": "Eric Dumazet", "email": "edumazet@google.com", "time": "Thu May 16 07:34:53 2013 +0000" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu May 16 15:02:01 2013 -0700" }, "message": "bonding: allow TSO being set on bonding master\n\nIn some situations, we need to disable TSO on bonding slaves.\n\nbonding device automatically unset TSO in bond_fix_features(), and\nperformance is not good because :\n\n1) We consume more cpu cycles.\n\n2) GSO segmentation has some bugs leading to out of order TCP packets\nif this segmentation is done before virtual device. This particular\nproblem will be addressed in a separate patch.\n\nThis patch allows TSO being set/unset on the bonding master,\nso that GSO segmentation is done after bonding layer.\n\nSigned-off-by: Eric Dumazet \u003cedumazet@google.com\u003e\nCc: Michał Mirosław \u003cmirqus@gmail.com\u003e\nCc: Jay Vosburgh \u003cfubar@us.ibm.com\u003e\nCc: Andy Gospodarek \u003candy@greyhouse.net\u003e\nCc: Maciej Żenczykowski \u003cmaze@google.com\u003e\nCc: Tom Herbert \u003ctherbert@google.com\u003e\nCc: Neal Cardwell \u003cncardwell@google.com\u003e\nCc: Yuchung Cheng \u003cycheng@google.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "243198d09f535f5cd74114f1b779c3da25bc70c8", "tree": "c8bc152937a6a5de879b4d8f958a8f21a8131ae4", "parents": [ "0020356355192cbaf6d315515e6c95bd09618c3b" ], "author": { "name": "Al Viro", "email": "viro@ZenIV.linux.org.uk", "time": "Sun May 05 16:05:55 2013 +0000" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Mon May 06 11:06:51 2013 -0400" }, "message": "rps_dev_flow_table_release(): no need to delay vfree()\n\nThe same story as with fib_trie patch - vfree() from RCU callbacks\nis legitimate now.\n\nSigned-off-by: Al Viro \u003cviro@zeniv.linux.org.uk\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "80d5c3689b886308247da295a228a54df49a44f6", "tree": "ce7b1e6898c3f9fea945e67fd06b125dfcc61fe7", "parents": [ "f646968f8f7c624587de729115d802372b9063dd" ], "author": { "name": "Patrick McHardy", "email": "kaber@trash.net", "time": "Fri Apr 19 02:04:28 2013 +0000" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Fri Apr 19 14:45:27 2013 -0400" }, "message": "net: vlan: prepare for 802.1ad VLAN filtering offload\n\nChange the rx_{add,kill}_vid callbacks to take a protocol argument in\npreparation of 802.1ad support. The protocol argument used so far is\nalways htons(ETH_P_8021Q).\n\nSigned-off-by: Patrick McHardy \u003ckaber@trash.net\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "f646968f8f7c624587de729115d802372b9063dd", "tree": "2b8c6604306f5e74af9e16c17e2b611610982b65", "parents": [ "c2962897c94605bc8f158a37dee8d867dda9f116" ], "author": { "name": "Patrick McHardy", "email": "kaber@trash.net", "time": "Fri Apr 19 02:04:27 2013 +0000" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Fri Apr 19 14:45:26 2013 -0400" }, "message": "net: vlan: rename NETIF_F_HW_VLAN_* feature flags to NETIF_F_HW_VLAN_CTAG_*\n\nRename the hardware VLAN acceleration features to include \"CTAG\" to indicate\nthat they only support CTAGs. Follow up patches will introduce 802.1ad\nserver provider tagging (STAGs) and require the distinction for hardware not\nsupporting acclerating both.\n\nSigned-off-by: Patrick McHardy \u003ckaber@trash.net\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "4cd729b04285b7330edaf5a7080aa795d6d15ff3", "tree": "274beba6b944edcd71879b779f2e674dede3cd91", "parents": [ "0022d2dd4d76e0e7d5c241c343a5016fdfa2ad4f" ], "author": { "name": "Vlad Yasevich", "email": "vyasevic@redhat.com", "time": "Mon Apr 15 09:54:25 2013 +0000" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Mon Apr 15 16:10:47 2013 -0400" }, "message": "net: add dev_uc_sync_multiple() and dev_mc_sync_multiple() api\n\nThe current implementation of dev_uc_sync/unsync() assumes that there is\na strict 1-to-1 relationship between the source and destination of the sync.\nIn other words, once an address has been synced to a destination device, it\nwill not be synced to any other device through the sync API.\nHowever, there are some virtual devices that aggreate a number of lower\ndevices and need to sync addresses to all of them. The current\nAPI falls short there.\n\nThis patch introduces a new dev_uc_sync_multiple() api that can be called\nin the above circumstances and allows sync to work for every invocation.\n\nCC: Jiri Pirko \u003cjiri@resnulli.us\u003e\nSigned-off-by: Vlad Yasevich \u003cvyasevic@redhat.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "d978a6361ad13f1f9694fcb7b5852d253a544d92", "tree": "8e8a8c62286fab2c044c4b53563222c1b66d7cb0", "parents": [ "8303e699f7089a1cd1421750fb33f289e5f3e1b9", "cb28ea3b13b86fb23448525f34720e659bda7aa8" ], "author": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Sun Apr 07 18:37:01 2013 -0400" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Sun Apr 07 18:37:01 2013 -0400" }, "message": "Merge git://git.kernel.org/pub/scm/linux/kernel/git/davem/net\n\nConflicts:\n\tdrivers/nfc/microread/mei.c\n\tnet/netfilter/nfnetlink_queue_core.c\n\nPull in \u0027net\u0027 to get Eric Biederman\u0027s AF_UNIX fix, upon which\nsome cleanups are going to go on-top.\n\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "4543fbefe6e06a9e40d9f2b28d688393a299f079", "tree": "be52f567c2f791b4328f75bee9cecaf022ed6a35", "parents": [ "e2409d83434d77874b461b78af6a19cd6e6a1280" ], "author": { "name": "Vlad Yasevich", "email": "vyasevic@redhat.com", "time": "Tue Apr 02 17:10:07 2013 -0400" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Fri Apr 05 00:18:46 2013 -0400" }, "message": "net: count hw_addr syncs so that unsync works properly.\n\nA few drivers use dev_uc_sync/unsync to synchronize the\naddress lists from master down to slave/lower devices. In\nsome cases (bond/team) a single address list is synched down\nto multiple devices. At the time of unsync, we have a leak\nin these lower devices, because \"synced\" is treated as a\nboolean and the address will not be unsynced for anything after\nthe first device/call.\n\nTreat \"synced\" as a count (same as refcount) and allow all\nunsync calls to work.\n\nSigned-off-by: Vlad Yasevich \u003cvyasevic@redhat.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "a210576cf891e9e6d2c238eabcf5c1286b1e7526", "tree": "0fa81a901cf628b25e6ee79057700cf39e59818a", "parents": [ "7d4c04fc170087119727119074e72445f2bb192b", "3658f3604066d5500ebd73a04084f127dc779441" ], "author": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Mon Apr 01 13:36:50 2013 -0400" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Mon Apr 01 13:36:50 2013 -0400" }, "message": "Merge git://git.kernel.org/pub/scm/linux/kernel/git/davem/net\n\nConflicts:\n\tnet/mac80211/sta_info.c\n\tnet/wireless/core.h\n\nTwo minor conflicts in wireless. Overlapping additions of extern\ndeclarations in net/wireless/core.h and a bug fix overlapping with\nthe addition of a boolean parameter to __ieee80211_key_free().\n\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "4c3d5e7b41dda1b1372bfc2545ef092a1bc5ad33", "tree": "462060a03b25ae2ab4768c692804de1bca366389", "parents": [ "8f0923c17d34ac838b7f57709f87fe33f76ee219" ], "author": { "name": "Eric Dumazet", "email": "edumazet@google.com", "time": "Sat Mar 30 06:31:03 2013 +0000" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Sat Mar 30 17:31:24 2013 -0400" }, "message": "net: reorder some fields of net_device\n\nAs time passed, some fields were added in net_device, and not\nat sensible offsets.\n\nLets reorder some fields to reduce number of cache lines in RX path.\nFields not used in data path should be moved out of this critical cache\nline.\n\nIn particular, move broadcast[] to the end of the rx section,\nas it is less used, and ethernet uses only the beginning of the 32bytes\nfield.\n\nBefore patch :\n\noffsetof(struct net_device,dev_addr)\u003d0x258\noffsetof(struct net_device,rx_handler)\u003d0x2b8\noffsetof(struct net_device,ingress_queue)\u003d0x2c8\noffsetof(struct net_device,broadcast)\u003d0x278\n\nAfter :\n\noffsetof(struct net_device,dev_addr)\u003d0x280\noffsetof(struct net_device,rx_handler)\u003d0x298\noffsetof(struct net_device,ingress_queue)\u003d0x2a8\noffsetof(struct net_device,broadcast)\u003d0x2b0\n\nSigned-off-by: Eric Dumazet \u003cedumazet@google.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "f3d4039242af92a9d93dee2fd9ae47066b20ca29", "tree": "c6d841b45268cb820e263af6794db6a0477e0308", "parents": [ "fbbdb8f096e0e5d8244e1ffa46e364146ab9a440" ], "author": { "name": "Paul Bolle", "email": "pebolle@tiscali.nl", "time": "Wed Mar 27 10:52:28 2013 +0000" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Wed Mar 27 23:24:51 2013 -0400" }, "message": "tokenring: delete last holdout of CONFIG_TR\n\nTokenring support was deleted in v3.5. One last holdout of the macro\nCONFIG_TR escaped that fate. Until now.\n\nSigned-off-by: Paul Bolle \u003cpebolle@tiscali.nl\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "24f11a5cf538a8fcc10d61f457e9a2094fb965b4", "tree": "950cfd826b3be166972cda88adbed56a638eaa33", "parents": [ "715230a44310a8cf66fbfb5a46f9a62a9b2de424" ], "author": { "name": "Dmitry Kravkov", "email": "dmitry@broadcom.com", "time": "Wed Mar 27 06:54:00 2013 +0000" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Wed Mar 27 14:06:41 2013 -0400" }, "message": "line up comment for ndo_bridge_getlink\n\nSigned-off-by: Dmitry Kravkov \u003cdmitry@broadcom.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "764444f5a324ad5a272773f078192819084388ce", "tree": "5917634cfb1330d2188de52316ecee9bc7c190cc", "parents": [ "2a89f9247ae43a2cf36fbc07b21b5792fc7b9efe" ], "author": { "name": "Fernando Luis Vazquez Cao", "email": "fernando_b1@lab.ntt.co.jp", "time": "Wed Mar 13 16:57:25 2013 +0000" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Fri Mar 15 08:30:46 2013 -0400" }, "message": "net: clean leftover of COMPAT_NET_DEV_OPS removal\n\nCOMPAT_NET_DEV_OPS was removed a while back and with it the definition of\nnetdev_resync_ops() went away. Let\u0027s finish the clean-up.\n\nSigned-off-by: Fernando Luis Vazquez Cao \u003cfernando@oss.ntt.co.jp\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "8a7fbfab4be39b8690543f3d29b26860d2f6c576", "tree": "6618fe6878b31f43d094be20643df7e7186083f1", "parents": [ "1e32b0c65c9e3f1a09cb5d1bcc0c5becf849d85f" ], "author": { "name": "nikolay@redhat.com", "email": "nikolay@redhat.com", "time": "Tue Mar 12 02:49:01 2013 +0000" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Fri Mar 15 08:22:20 2013 -0400" }, "message": "netxen: write IP address to firmware when using bonding\n\nThis patch allows LRO aggregation on bonded devices that contain an\nNX3031 device. It also adds a for_each_netdev_in_bond_rcu(bond, slave)\nmacro which executes for each slave that has bond as master.\n\nV3: After testing and discussing this with Rajesh, I decided to keep the\n vlan ip cache and just rename it to ip_cache since it will store bond\n ip addresses too. A new master flag has been added to the ip cache to\n denote that the address has been added because of a master device.\n I\u0027ve taken care of the enslave/release cases by checking for various\n combinations of events and flags (e.g. netxen has a master, it\u0027s a\n bond master and it\u0027s not marked as a slave means it is being enslaved\n and is dev_open()ed in bond_enslave).\n I\u0027ve changed netxen_free_ip_list() to have a \"master\" parameter which\n causes all IP addresses marked as master to be deleted (used when a\n netxen is being released). I\u0027ve made the patch use the new upper\n device API as well. The following cases were tested:\n - bond -\u003e netxen\n - vlan -\u003e netxen\n - vlan -\u003e bond -\u003e netxen\n\nV2: Remove local ip caching, retrieve addresses dynamically and\n restore them if necessary.\n\nNote: Tested with NX3031 adapter.\n\nTested-by: Rajesh Borundia \u003crajesh.borundia@qlogic.com\u003e\nSigned-off-by: Andy Gospodarek \u003cagospoda@redhat.com\u003e\nSigned-off-by: Nikolay Aleksandrov \u003cnikolay@redhat.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "ec5f061564238892005257c83565a0b58ec79295", "tree": "1ea3987010c27985a3edb7af66e1d143b2cefc77", "parents": [ "3868b7aa76310d1d723d4db05c3526c908fb1e8b" ], "author": { "name": "Pravin B Shelar", "email": "pshelar@nicira.com", "time": "Thu Mar 07 09:28:01 2013 +0000" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Sat Mar 09 16:08:57 2013 -0500" }, "message": "net: Kill link between CSUM and SG features.\n\nEarlier SG was unset if CSUM was not available for given device to\nforce skb copy to avoid sending inconsistent csum.\nCommit c9af6db4c11c (net: Fix possible wrong checksum generation)\nadded explicit flag to force copy to fix this issue. Therefore\nthere is no need to link SG and CSUM, following patch kills this\nlink between there two features.\n\nThis patch is also required following patch in series.\n\nSigned-off-by: Pravin B Shelar \u003cpshelar@nicira.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "82dc3c63c692b1e1d59378ecee948ac88e034aad", "tree": "d324df2801b34974e8735ac79f28ca43298bea53", "parents": [ "dd9f319d94c99b96fc9b34ccde7389a91059fe31" ], "author": { "name": "Eric Dumazet", "email": "edumazet@google.com", "time": "Tue Mar 05 15:57:22 2013 +0000" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Tue Mar 05 23:40:01 2013 -0500" }, "message": "net: introduce NAPI_POLL_WEIGHT\n\nSome drivers use a too big NAPI poll weight.\n\nThis patch adds a NAPI_POLL_WEIGHT default value\nand issues an error message if a driver attempts\nto use a bigger weight.\n\nSigned-off-by: Eric Dumazet \u003cedumazet@google.com\u003e\nCc: Eilon Greenstein \u003ceilong@broadcom.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "cd0615746ba0f6643fb984345ae6ee0b73404ca6", "tree": "f55557971fec403884c37cf28326c28c7670b4cd", "parents": [ "082c7ca42b0bf33aa3fb5eeb619dfbcfaa20d216" ], "author": { "name": "Cong Wang", "email": "xiyou.wangcong@gmail.com", "time": "Tue Feb 19 02:47:05 2013 +0000" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Tue Feb 19 13:18:13 2013 -0500" }, "message": "net: fix a build failure when !CONFIG_PROC_FS\n\nWhen !CONFIG_PROC_FS dev_mcast_init() is not defined,\nactually we can just merge dev_mcast_init() into\ndev_proc_init().\n\nReported-by: Gao feng \u003cgaofeng@cn.fujitsu.com\u003e\nCc: Gao feng \u003cgaofeng@cn.fujitsu.com\u003e\nCc: \"David S. Miller\" \u003cdavem@davemloft.net\u003e\nSigned-off-by: Cong Wang \u003cxiyou.wangcong@gmail.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "900ff8c6321418dafa03c22e215cb9646a2541b9", "tree": "da94d724e34cda801d23aea533a5575084d31a8d", "parents": [ "45d213f5f8733ab4b5b2424b21c0034d54c5a866" ], "author": { "name": "Cong Wang", "email": "xiyou.wangcong@gmail.com", "time": "Mon Feb 18 19:20:33 2013 +0000" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Tue Feb 19 00:51:10 2013 -0500" }, "message": "net: move procfs code to net/core/net-procfs.c\n\nSimilar to net/core/net-sysfs.c, group procfs code to\na single unit.\n\nCc: \"David S. Miller\" \u003cdavem@davemloft.net\u003e\nSigned-off-by: Cong Wang \u003cxiyou.wangcong@gmail.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "05e8ef4ab2d8087d360e814d14da20b9f7fb2283", "tree": "f5161c863bff859fd1c52e36d88fe4a10b84af0f", "parents": [ "14bbd6a565e1bcdc240d44687edb93f721cfdf99" ], "author": { "name": "Pravin B Shelar", "email": "pshelar@nicira.com", "time": "Thu Feb 14 09:44:55 2013 +0000" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Fri Feb 15 15:16:03 2013 -0500" }, "message": "net: factor out skb_mac_gso_segment() from skb_gso_segment()\n\nThis function will be used in next GRE_GSO patch. This patch does\nnot change any functionality. It only exports skb_mac_gso_segment()\nfunction.\n\n[ Use skb_reset_mac_len() -DaveM ]\n\nSigned-off-by: Pravin B Shelar \u003cpshelar@nicira.com\u003e\nAcked-by: Eric Dumazet \u003cedumazet@google.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "1690be63a27b20ae65c792729a44f5970561ffa4", "tree": "c0c86bc471e24207ee61b544c8683d84fc43d105", "parents": [ "b0e9a30dd669a844bb4f74515f8bcd307018ffd0" ], "author": { "name": "Vlad Yasevich", "email": "vyasevic@redhat.com", "time": "Wed Feb 13 12:00:18 2013 +0000" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Wed Feb 13 19:42:16 2013 -0500" }, "message": "bridge: Add vlan support to static neighbors\n\nWhen a user adds bridge neighbors, allow him to specify VLAN id.\nIf the VLAN id is not specified, the neighbor will be added\nfor VLANs currently in the ports filter list. If no VLANs are\nconfigured on the port, we use vlan 0 and only add 1 entry.\n\nSigned-off-by: Vlad Yasevich \u003cvyasevic@redhat.com\u003e\nAcked-by: Jitendra Kalsaria \u003cjitendra.kalsaria@qlogic.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "6cbdceeb1cb12c7d620161925a8c3e81daadb2e4", "tree": "81852fee23a380456cce4d2c1bb5ec52c7ff0013", "parents": [ "407af3299ef1ac7e87ce3fb530e32a009d1a9efd" ], "author": { "name": "Vlad Yasevich", "email": "vyasevic@redhat.com", "time": "Wed Feb 13 12:00:13 2013 +0000" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Wed Feb 13 19:41:46 2013 -0500" }, "message": "bridge: Dump vlan information from a bridge port\n\nUsing the RTM_GETLINK dump the vlan filter list of a given\nbridge port. The information depends on setting the filter\nflag similar to how nic VF info is dumped.\n\nSigned-off-by: Vlad Yasevich \u003cvyasevic@redhat.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "407af3299ef1ac7e87ce3fb530e32a009d1a9efd", "tree": "9572e0b02eb9f6c2952b490f3a79f4b790fcea3c", "parents": [ "85f46c6baef1486ce20e13dd7cdea5dd15be2a90" ], "author": { "name": "Vlad Yasevich", "email": "vyasevic@redhat.com", "time": "Wed Feb 13 12:00:12 2013 +0000" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Wed Feb 13 19:41:46 2013 -0500" }, "message": "bridge: Add netlink interface to configure vlans on bridge ports\n\nAdd a netlink interface to add and remove vlan configuration on bridge port.\nThe interface uses the RTM_SETLINK message and encodes the vlan\nconfiguration inside the IFLA_AF_SPEC. It is possble to include multiple\nvlans to either add or remove in a single message.\n\nSigned-off-by: Vlad Yasevich \u003cvyasevic@redhat.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "febf018d22347b5df94066bca05d0c11a84e839d", "tree": "726ff7c7ee289736d6c36996d4340928f9594298", "parents": [ "5b815b52f63c8f5dcd03964d69c335ee47851878" ], "author": { "name": "David Ward", "email": "david.ward@ll.mit.edu", "time": "Fri Feb 08 17:17:06 2013 +0000" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Sun Feb 10 20:37:22 2013 -0500" }, "message": "net/802: Implement Multiple Registration Protocol (MRP)\n\nInitial implementation of the Multiple Registration Protocol (MRP)\nfrom IEEE 802.1Q-2011, based on the existing implementation of the\nGeneric Attribute Registration Protocol (GARP).\n\nSigned-off-by: David Ward \u003cdavid.ward@ll.mit.edu\u003e\nAcked-by: Patrick McHardy \u003ckaber@trash.net\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "12b0004d1d1e2a9aa667412d479041e403bcafae", "tree": "132aecb3e4a78707552a0ddf4d4618734af47efa", "parents": [ "25060d8f3f2c21daadb4fc1fb0e37ce2c992e30b" ], "author": { "name": "Cong Wang", "email": "amwang@redhat.com", "time": "Tue Feb 05 16:36:38 2013 +0000" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Wed Feb 06 15:58:00 2013 -0500" }, "message": "net: adjust skb_gso_segment() for calling in rx path\n\nskb_gso_segment() is almost always called in tx path,\nexcept for openvswitch. It calls this function when\nit receives the packet and tries to queue it to user-space.\nIn this special case, the -\u003eip_summed check inside\nskb_gso_segment() is no longer true, as -\u003eip_summed value\nhas different meanings on rx path.\n\nThis patch adjusts skb_gso_segment() so that we can at least\navoid such warnings on checksum.\n\nCc: Jesse Gross \u003cjesse@nicira.com\u003e\nCc: David S. Miller \u003cdavem@davemloft.net\u003e\nSigned-off-by: Cong Wang \u003camwang@redhat.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "5fbee843c32e5de2d8af68ba0bdd113bb0af9d86", "tree": "1b5af9ab4177611b00412abe3ee3217030740307", "parents": [ "ce4a600e477a4da600c8056897e71e2f4a8c5ac0" ], "author": { "name": "Cong Wang", "email": "amwang@redhat.com", "time": "Tue Jan 22 21:29:39 2013 +0000" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Mon Jan 28 18:32:54 2013 -0500" }, "message": "netpoll: add RCU annotation to npinfo field\n\ndev-\u003enpinfo is protected by RCU.\n\nThis fixes the following sparse warnings:\n\nnet/core/netpoll.c:177:48: error: incompatible types in comparison expression (different address spaces)\nnet/core/netpoll.c:200:35: error: incompatible types in comparison expression (different address spaces)\nnet/core/netpoll.c:221:35: error: incompatible types in comparison expression (different address spaces)\nnet/core/netpoll.c:327:18: error: incompatible types in comparison expression (different address spaces)\n\nCc: Eric Dumazet \u003ceric.dumazet@gmail.com\u003e\nCc: David S. Miller \u003cdavem@davemloft.net\u003e\nSigned-off-by: Cong Wang \u003camwang@redhat.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "4b87f922598acf91eee18f71688a33f54f57bcde", "tree": "9cdfe30c6b96c47093da5392ed82d147290cd64c", "parents": [ "55eb555d9674e2ebe9d4de0146602f96ff18e7d6", "daf3ec688e057f6060fb9bb0819feac7a8bbf45c" ], "author": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Tue Jan 15 15:05:59 2013 -0500" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Tue Jan 15 15:05:59 2013 -0500" }, "message": "Merge git://git.kernel.org/pub/scm/linux/kernel/git/davem/net\n\nConflicts:\n\tDocumentation/networking/ip-sysctl.txt\n\tdrivers/net/ethernet/broadcom/bnx2x/bnx2x_cmn.c\n\nBoth conflicts were simply overlapping context.\n\nA build fix for qlcnic is in here too, simply removing the added\ndevinit annotations which no longer exist.\n\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "d07d7507bfb4e23735c9b83e397c43e1e8a173e8", "tree": "1422f8bc9cfa0340a7b1cdec03456a6bf5e3a7d5", "parents": [ "f7e9e230f41f991a6e3e7cb6096424fdfb626081" ], "author": { "name": "Stanislaw Gruszka", "email": "sgruszka@redhat.com", "time": "Thu Jan 10 23:19:10 2013 +0000" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Fri Jan 11 15:55:48 2013 -0800" }, "message": "net, wireless: overwrite default_ethtool_ops\n\nSince:\n\ncommit 2c60db037034d27f8c636403355d52872da92f81\nAuthor: Eric Dumazet \u003cedumazet@google.com\u003e\nDate: Sun Sep 16 09:17:26 2012 +0000\n\n net: provide a default dev-\u003eethtool_ops\n\nwireless core does not correctly assign ethtool_ops.\n\nAfter alloc_netdev*() call, some cfg80211 drivers provide they own\nethtool_ops, but some do not. For them, wireless core provide generic\ncfg80211_ethtool_ops, which is assigned in NETDEV_REGISTER notify call:\n\n if (!dev-\u003eethtool_ops)\n dev-\u003eethtool_ops \u003d \u0026cfg80211_ethtool_ops;\n\nBut after Eric\u0027s commit, dev-\u003eethtool_ops is no longer NULL (on cfg80211\ndrivers without custom ethtool_ops), but points to \u0026default_ethtool_ops.\n\nIn order to fix the problem, provide function which will overwrite\ndefault_ethtool_ops and use it by wireless core.\n\nSigned-off-by: Stanislaw Gruszka \u003csgruszka@redhat.com\u003e\nAcked-by: Johannes Berg \u003cjohannes@sipsolutions.net\u003e\nAcked-by: Ben Hutchings \u003cbhutchings@solarflare.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "024e9679a2daaa67642693366fb63a6b8c61b9f3", "tree": "2661a1d8f50699e508f7bb0fbe83c90bbd2de3d2", "parents": [ "01c5f864e62b1c1ee72fb492c6c5e47a162a507b" ], "author": { "name": "Alexander Duyck", "email": "alexander.h.duyck@intel.com", "time": "Thu Jan 10 08:57:46 2013 +0000" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Jan 10 22:47:04 2013 -0800" }, "message": "net: Add support for XPS without sysfs being defined\n\nThis patch makes it so that we can support transmit packet steering without\nsysfs needing to be enabled. The reason for making this change is to make\nit so that a driver can make use of the XPS even while the sysfs portion of\nthe interface is not present.\n\nSigned-off-by: Alexander Duyck \u003calexander.h.duyck@intel.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "537c00de1c9ba9876b91d869e84caceefe2b8bf9", "tree": "f808a5221aec07b3a0a01513e5de9d5cdc2a6b54", "parents": [ "416186fbf8c5b4e4465a10c6ac7a45b6c47144b2" ], "author": { "name": "Alexander Duyck", "email": "alexander.h.duyck@intel.com", "time": "Thu Jan 10 08:57:02 2013 +0000" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Jan 10 22:47:03 2013 -0800" }, "message": "net: Add functions netif_reset_xps_queue and netif_set_xps_queue\n\nThis patch adds two functions, netif_reset_xps_queue and\nnetif_set_xps_queue. The main idea behind these two functions is to\nprovide a mechanism through which drivers can update their defaults in\nregards to XPS.\n\nCurrently no such mechanism exists and as a result we cannot use XPS for\nthings such as ATR which would require a basic configuration to start in\nwhich the Tx queues are mapped to CPUs via a 1:1 mapping. With this change\nI am making it possible for drivers such as ixgbe to be able to use the XPS\nfeature by controlling the default configuration.\n\nSigned-off-by: Alexander Duyck \u003calexander.h.duyck@intel.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "416186fbf8c5b4e4465a10c6ac7a45b6c47144b2", "tree": "8670e015c8e6ba3e41c7373671bcfe1b2703b653", "parents": [ "c10d73671ad30f54692f7f69f0e09e75d3a8926a" ], "author": { "name": "Alexander Duyck", "email": "alexander.h.duyck@intel.com", "time": "Thu Jan 10 08:56:51 2013 +0000" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Jan 10 22:47:03 2013 -0800" }, "message": "net: Split core bits of netdev_pick_tx into __netdev_pick_tx\n\nThis change splits the core bits of netdev_pick_tx into a separate function.\nThe main idea behind this is to make this code accessible to select queue\nfunctions when they decide to process the standard path instead of their\nown custom path in their select queue routine.\n\nSigned-off-by: Alexander Duyck \u003calexander.h.duyck@intel.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "85464ef271a0f5496f404c6a2f2dfbf1d76e1a49", "tree": "f14aed911d388b41d4abde15605186a5f8e9509f", "parents": [ "8b98a70c28a607a02b3c3d41bc9a4c141f421052" ], "author": { "name": "Jiri Pirko", "email": "jiri@resnulli.us", "time": "Thu Jan 03 22:49:03 2013 +0000" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Fri Jan 04 13:31:51 2013 -0800" }, "message": "net: kill dev-\u003emaster\n\nNobody uses this now. Remove it.\n\nSigned-off-by: Jiri Pirko \u003cjiri@resnulli.us\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "8b98a70c28a607a02b3c3d41bc9a4c141f421052", "tree": "7f3ad0f2476c993f7b4b84f2796f73c012e9fb41", "parents": [ "471cb5a33dcbd7c529684a2ac7ba4451414ee4a7" ], "author": { "name": "Jiri Pirko", "email": "jiri@resnulli.us", "time": "Thu Jan 03 22:49:02 2013 +0000" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Fri Jan 04 13:31:50 2013 -0800" }, "message": "net: remove no longer used netdev_set_bond_master() and netdev_set_master()\n\nSigned-off-by: Jiri Pirko \u003cjiri@resnulli.us\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "9ff162a8b96c96238773972e26288a366e403b0c", "tree": "ae74065e46364c807a1996e5268aec349d9d86d4", "parents": [ "04e406dcc54cfd84d333ea673fa986fb5a1840e7" ], "author": { "name": "Jiri Pirko", "email": "jiri@resnulli.us", "time": "Thu Jan 03 22:48:49 2013 +0000" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Fri Jan 04 13:31:49 2013 -0800" }, "message": "net: introduce upper device lists\n\nThis lists are supposed to serve for storing pointers to all upper devices.\nEventually it will replace dev-\u003emaster pointer which is used for\nbonding, bridge, team but it cannot be used for vlan, macvlan where\nthere might be multiple upper present. In case the upper link is\nreplacement for dev-\u003emaster, it is marked with \"master\" flag.\n\nNew upper device list resolves this limitation. Also, the information\nstored in lists is used for preventing looping setups like\n\"bond-\u003esomethingelse-\u003esamebond\"\n\nSigned-off-by: Jiri Pirko \u003cjiri@resnulli.us\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" } ], "next": "fbdeca2d7753aa1ab929aeb77ccc46489eed02b9" }