)]}' { "log": [ { "commit": "4ba9920e5e9c0e16b5ed24292d45322907bb9035", "tree": "7d023baea59ed0886ded1f0b6d1c6385690b88f7", "parents": [ "82c477669a4665eb4e52030792051e0559ee2a36", "8b662fe70c68282f78482dc272df0c4f355e49f5" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Jan 25 11:17:34 2014 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Jan 25 11:17:34 2014 -0800" }, "message": "Merge git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-next\n\nPull networking updates from David Miller:\n\n 1) BPF debugger and asm tool by Daniel Borkmann.\n\n 2) Speed up create/bind in AF_PACKET, also from Daniel Borkmann.\n\n 3) Correct reciprocal_divide and update users, from Hannes Frederic\n Sowa and Daniel Borkmann.\n\n 4) Currently we only have a \"set\" operation for the hw timestamp socket\n ioctl, add a \"get\" operation to match. From Ben Hutchings.\n\n 5) Add better trace events for debugging driver datapath problems, also\n from Ben Hutchings.\n\n 6) Implement auto corking in TCP, from Eric Dumazet. Basically, if we\n have a small send and a previous packet is already in the qdisc or\n device queue, defer until TX completion or we get more data.\n\n 7) Allow userspace to manage ipv6 temporary addresses, from Jiri Pirko.\n\n 8) Add a qdisc bypass option for AF_PACKET sockets, from Daniel\n Borkmann.\n\n 9) Share IP header compression code between Bluetooth and IEEE802154\n layers, from Jukka Rissanen.\n\n10) Fix ipv6 router reachability probing, from Jiri Benc.\n\n11) Allow packets to be captured on macvtap devices, from Vlad Yasevich.\n\n12) Support tunneling in GRO layer, from Jerry Chu.\n\n13) Allow bonding to be configured fully using netlink, from Scott\n Feldman.\n\n14) Allow AF_PACKET users to obtain the VLAN TPID, just like they can\n already get the TCI. From Atzm Watanabe.\n\n15) New \"Heavy Hitter\" qdisc, from Terry Lam.\n\n16) Significantly improve the IPSEC support in pktgen, from Fan Du.\n\n17) Allow ipv4 tunnels to cache routes, just like sockets. From Tom\n Herbert.\n\n18) Add Proportional Integral Enhanced packet scheduler, from Vijay\n Subramanian.\n\n19) Allow openvswitch to mmap\u0027d netlink, from Thomas Graf.\n\n20) Key TCP metrics blobs also by source address, not just destination\n address. From Christoph Paasch.\n\n21) Support 10G in generic phylib. From Andy Fleming.\n\n22) Try to short-circuit GRO flow compares using device provided RX\n hash, if provided. From Tom Herbert.\n\nThe wireless and netfilter folks have been busy little bees too.\n\n* git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-next: (2064 commits)\n net/cxgb4: Fix referencing freed adapter\n ipv6: reallocate addrconf router for ipv6 address when lo device up\n fib_frontend: fix possible NULL pointer dereference\n rtnetlink: remove IFLA_BOND_SLAVE definition\n rtnetlink: remove check for fill_slave_info in rtnl_have_link_slave_info\n qlcnic: update version to 5.3.55\n qlcnic: Enhance logic to calculate msix vectors.\n qlcnic: Refactor interrupt coalescing code for all adapters.\n qlcnic: Update poll controller code path\n qlcnic: Interrupt code cleanup\n qlcnic: Enhance Tx timeout debugging.\n qlcnic: Use bool for rx_mac_learn.\n bonding: fix u64 division\n rtnetlink: add missing IFLA_BOND_AD_INFO_UNSPEC\n sfc: Use the correct maximum TX DMA ring size for SFC9100\n Add Shradha Shah as the sfc driver maintainer.\n net/vxlan: Share RX skb de-marking and checksum checks with ovs\n tulip: cleanup by using ARRAY_SIZE()\n ip_tunnel: clear IPCB in ip_tunnel_xmit() in case dst_link_failure() is called\n net/cxgb4: Don\u0027t retrieve stats during recovery\n ...\n" }, { "commit": "fb1b5034e4987b158179a62732fb6dfb8f7ec88e", "tree": "93d02ff7f0f530286fd54d03b632c6eaabc9dcc9", "parents": [ "8f399921ea9a562bc8221258c4b8a7bd69577939", "27cdef637c25705b433d5c4deeef4cf8dcb75d6a" ], "author": { "name": "Roland Dreier", "email": "roland@purestorage.com", "time": "Wed Jan 22 23:24:21 2014 -0800" }, "committer": { "name": "Roland Dreier", "email": "roland@purestorage.com", "time": "Wed Jan 22 23:24:21 2014 -0800" }, "message": "Merge branch \u0027ip-roce\u0027 into for-next\n\nConflicts:\n\tdrivers/infiniband/hw/mlx4/main.c\n" }, { "commit": "297e0dad720664dad44baa2cdd13f871979fb58c", "tree": "2c14a33b5fcabd51e41335884896b5209db29de2", "parents": [ "d487ee77740ccf79d7dc1935d4daa77887283028" ], "author": { "name": "Moni Shoua", "email": "monis@mellanox.com", "time": "Thu Dec 12 18:03:14 2013 +0200" }, "committer": { "name": "Roland Dreier", "email": "roland@purestorage.com", "time": "Sat Jan 18 14:12:53 2014 -0800" }, "message": "IB/mlx4: Handle Ethernet L2 parameters for IP based GID addressing\n\nIP based RoCE gids don\u0027t store Ethernet L2 parameters, MAC and VLAN.\n\nTherefore, we need to extract them from the CQE and place them in\nstruct ib_wc (to be used for cases were they were taken from the gid).\n\nAlso, when modifying a QP or building address handle, instead of\nparsing the dgid to get the MAC and VLAN, take them from the address\nhandle attributes.\n\nSigned-off-by: Moni Shoua \u003cmonis@mellanox.com\u003e\nSigned-off-by: Or Gerlitz \u003cogerlitz@mellanox.com\u003e\nSigned-off-by: Roland Dreier \u003croland@purestorage.com\u003e\n" }, { "commit": "dd5f03beb4f76ae65d76d8c22a8815e424fc607c", "tree": "162cb1bece8602841de5cff1d37e781de8528c0f", "parents": [ "7e22e91102c6b9df7c4ae2168910e19d2bb14cd6" ], "author": { "name": "Matan Barak", "email": "matanb@mellanox.com", "time": "Thu Dec 12 18:03:11 2013 +0200" }, "committer": { "name": "Roland Dreier", "email": "roland@purestorage.com", "time": "Tue Jan 14 14:20:54 2014 -0800" }, "message": "IB/core: Ethernet L2 attributes in verbs/cm structures\n\nThis patch add the support for Ethernet L2 attributes in the\nverbs/cm/cma structures.\n\nWhen dealing with L2 Ethernet, we should use smac, dmac, vlan ID and priority\nin a similar manner that the IB L2 (and the L4 PKEY) attributes are used.\n\nThus, those attributes were added to the following structures:\n\n* ib_ah_attr - added dmac\n* ib_qp_attr - added smac and vlan_id, (sl remains vlan priority)\n* ib_wc - added smac, vlan_id\n* ib_sa_path_rec - added smac, dmac, vlan_id\n* cm_av - added smac and vlan_id\n\nFor the path record structure, extra care was taken to avoid the new\nfields when packing it into wire format, so we don\u0027t break the IB CM\nand SA wire protocol.\n\nOn the active side, the CM fills. its internal structures from the\npath provided by the ULP. We add there taking the ETH L2 attributes\nand placing them into the CM Address Handle (struct cm_av).\n\nOn the passive side, the CM fills its internal structures from the WC\nassociated with the REQ message. We add there taking the ETH L2\nattributes from the WC.\n\nWhen the HW driver provides the required ETH L2 attributes in the WC,\nthey set the IB_WC_WITH_SMAC and IB_WC_WITH_VLAN flags. The IB core\ncode checks for the presence of these flags, and in their absence does\naddress resolution from the ib_init_ah_from_wc() helper function.\n\nib_modify_qp_is_ok is also updated to consider the link layer. Some\nparameters are mandatory for Ethernet link layer, while they are\nirrelevant for IB. Vendor drivers are modified to support the new\nfunction signature.\n\nSigned-off-by: Matan Barak \u003cmatanb@mellanox.com\u003e\nSigned-off-by: Or Gerlitz \u003cogerlitz@mellanox.com\u003e\nSigned-off-by: Roland Dreier \u003croland@purestorage.com\u003e\n" }, { "commit": "4de6580360867d44adecb2d05febed1c8d186c82", "tree": "4645ce69a405ff0ed519da0ad7d8ba12c76e05a8", "parents": [ "240ae00e4d834e387f4f09e236130f520e357a70" ], "author": { "name": "Matan Barak", "email": "matanb@mellanox.com", "time": "Thu Nov 07 15:25:14 2013 +0200" }, "committer": { "name": "Roland Dreier", "email": "roland@purestorage.com", "time": "Tue Jan 14 14:06:50 2014 -0800" }, "message": "mlx4_core: Add support for steerable IB UD QPs\n\nThis patch adds support for allocating IB UD QPs that we can steer\ntraffic from. We introduce a new firmware command FLOW_STEERING_IB_UC_QP_RANGE\nand a capability bit.\n\nThis command isn\u0027t supported for VFs.\n\nSigned-off-by: Matan Barak \u003cmatanb@mellanox.com\u003e\nSigned-off-by: Or Gerlitz \u003cogerlitz@mellanox.com\u003e\nSigned-off-by: Roland Dreier \u003croland@purestorage.com\u003e\n" }, { "commit": "7ffdf726cfe0d188907bdbb0e7729fb35a69c219", "tree": "ead297c29f54fbb91741822dd6df13dbc3193fd4", "parents": [ "604d13c97f0d863e41da3f5835c62e3cf899962b" ], "author": { "name": "Or Gerlitz", "email": "ogerlitz@mellanox.com", "time": "Mon Dec 23 16:09:43 2013 +0200" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Tue Dec 31 14:31:43 2013 -0500" }, "message": "net/mlx4_core: Add basic support for TCP/IP offloads under tunneling\n\nAdd the low-level device commands and definitions used for TCP/IP HW offloads\nof tunneled/vxlan traffic which are supported by the ConnectX3-pro NIC.\n\nThis is done through the following elements:\n\n - read tunneling device caps in QUERY_DEV_CAP\n - add helper function to do SET_PORT for tunneling\n - add DMFS VXLAN steering rule definitions\n - add CQE and WQE checksum offload field definitions\n\nSigned-off-by: Or Gerlitz \u003cogerlitz@mellanox.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "8e1a28e8e6797449dfdfa4739002d1e5939355a8", "tree": "10bf6bb908f2a5e7632280eb0f362112c27cfc29", "parents": [ "eb17711bc1d6611e934af5b6dabc225936084128" ], "author": { "name": "Hadar Hen Zion", "email": "hadarh@mellanox.com", "time": "Thu Dec 19 21:20:12 2013 +0200" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Dec 19 19:04:43 2013 -0500" }, "message": "net/mlx4_core: Expose physical port id as PF/VF capability\n\nAdd the infrastructure needed to support ndo_get_phys_port_id which\nallows users to identify to which physical port a net-device is connected\nto by reading a unique port id.\nThis will work for VFs and PFs.\nThe driver uses a new device capability - phys_port_id, The PF driver\nreads the port phys_port_id from Firmware and stores it. The VF driver\nreads the port phys_port_id from the PF using QUERY_FUNC_CAP command.\n\nSigned-off-by: Hadar Hen Zion \u003chadarh@mellanox.com\u003e\nSigned-off-by: Amir Vadai \u003camirv@mellanox.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "163561a4e2f8af44e96453bc10c7a4f9bcc736e1", "tree": "8d328edce3828c60e80989267eb7594d2d7e24d6", "parents": [ "6e7136ed7793fa4948b0192dcd6862d12a50d67c" ], "author": { "name": "Eugenia Emantayev", "email": "eugenia@mellanox.com", "time": "Thu Nov 07 12:19:54 2013 +0200" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Nov 07 19:22:48 2013 -0500" }, "message": "net/mlx4_en: Datapath structures are allocated per NUMA node\n\nFor each RX/TX ring and its CQ, allocation is done on a NUMA node that\ncorresponds to the core that the data structure should operate on.\nThe assumption is that the core number is reflected by the ring index.\nThe affected allocations are the ring/CQ data structures,\nthe TX/RX info and the shared HW/SW buffer.\nFor TX rings, each core has rings of all UPs.\n\nSigned-off-by: Yevgeny Petrilin \u003cyevgenyp@mellanox.com\u003e\nSigned-off-by: Eugenia Emantayev \u003ceugenia@mellanox.com\u003e\nReviewed-by: Hadar Hen Zion \u003chadarh@mellanox.com\u003e\nSigned-off-by: Amir Vadai \u003camirv@mellanox.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "6e7136ed7793fa4948b0192dcd6862d12a50d67c", "tree": "fe23b4a06d76512b39a97010324f0a7bce489059", "parents": [ "41d942d56cfd21058fba465804e14ba349541442" ], "author": { "name": "Eugenia Emantayev", "email": "eugenia@mellanox.com", "time": "Thu Nov 07 12:19:53 2013 +0200" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Nov 07 19:22:48 2013 -0500" }, "message": "net/mlx4_core: ICM pages are allocated on device NUMA node\n\nThis is done to optimize FW/HW access to host memory.\n\nSigned-off-by: Yevgeny Petrilin \u003cyevgenyp@mellanox.com\u003e\nSigned-off-by: Eugenia Emantayev \u003ceugenia@mellanox.com\u003e\nReviewed-by: Hadar Hen Zion \u003chadarh@mellanox.com\u003e\nSigned-off-by: Amir Vadai \u003camirv@mellanox.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "5a0d0a6161aecbbc76e4c1d2b82e4c7cef88bb29", "tree": "ab181f2d85ad79385f739c3c1c682b1db835b9aa", "parents": [ "a30f1bc5c0254cee94bc22304e57fe030fcf68f5" ], "author": { "name": "Jack Morgenstein", "email": "jackm@dev.mellanox.co.il", "time": "Sun Nov 03 10:03:23 2013 +0200" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Mon Nov 04 16:19:07 2013 -0500" }, "message": "mlx4: Structures and init/teardown for VF resource quotas\n\nThis is step #1 for implementing SRIOV resource quotas for VFs.\n\nQuotas are implemented per resource type for VFs and the PF, to prevent\nany entity from simply grabbing all the resources for itself and leaving\nthe other entities unable to obtain such resources.\n\nResources which are allocated using quotas: QPs, CQs, SRQs, MPTs, MTTs, MAC,\n VLAN, and Counters.\n\nThe quota system works as follows:\nEach entity (VF or PF) is given a max number of a given resource (its quota),\nand a guaranteed minimum number for each resource (starvation prevention).\n\nFor QPs, CQs, SRQs, MPTs and MTTs:\n50% of the available quantity for the resource is divided equally among\nthe PF and all the active VFs (i.e., the number of VFs in the mlx4_core module\nparameter \"num_vfs\"). This 50% represents the \"guaranteed minimum\" pool.\nThe other 50% is the \"free pool\", allocated on a first-come-first-serve basis.\nFor each VF/PF, resources are first allocated from its \"guaranteed-minimum\"\npool. When that pool is exhausted, the driver attempts to allocate from\nthe resource \"free-pool\".\n\nThe quota (i.e., max) for the VFs and the PF is:\n The free-pool amount (50% of the real max) + the guaranteed minimum\n\nFor MACs:\n Guarantee 2 MACs per VF/PF per port. As a result, since we have only\n 128 MACs per port, reduce the allowable number of VFs from 64 to 63.\n Any remaining MACs are put into a free pool.\n\nFor VLANs:\n For the PF, the per-port quota is 128 and guarantee is 64\n (to allow the PF to register at least a VLAN per VF in VST mode).\n For the VFs, the per-port quota is 64 and the guarantee is 0.\n We assume that VGT VFs are trusted not to abuse the VLAN resource.\n\nFor Counters:\n For all functions (PF and VFs), the quota is 128 and the guarantee is 0.\n\nIn this patch, we define the needed structures, which are added to the\nresource-tracker struct. In addition, we do initialization\nfor the resource quota, and adjust the query_device response to use quotas\nrather than resource maxima.\n\nAs part of the implementation, we introduce a new field in\nmlx4_dev: quotas. This field holds the resource quotas used\nto report maxima to the upper layers (ib_core, via query_device).\n\nThe HCA maxima of these values are passed to the VFs (via\nQUERY_HCA) so that they may continue to use these in handling\nQPs, CQs, SRQs and MPTs.\n\nSigned-off-by: Jack Morgenstein \u003cjackm@dev.mellanox.co.il\u003e\nSigned-off-by: Or Gerlitz \u003cogerlitz@mellanox.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "2009d0059c084288f060b1ffe3d14229588acb67", "tree": "066b709ab7cdb70f26b8ab7fdd6a60eb9dee1158", "parents": [ "acddd5dd44d4fd9b45dd5ee69cd8b183052b1cdc" ], "author": { "name": "Jack Morgenstein", "email": "jackm@dev.mellanox.co.il", "time": "Sun Nov 03 10:03:19 2013 +0200" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Mon Nov 04 16:19:07 2013 -0500" }, "message": "net/mlx4_en: Use vlan id instead of vlan index for unregistration\n\nUse of vlan_index created problems unregistering vlans on guests.\n\nIn addition, tools delete vlan by tag, not by index, lets follow that.\n\nSigned-off-by: Jack Morgenstein \u003cjackm@dev.mellanox.co.il\u003e\nSigned-off-by: Or Gerlitz \u003cogerlitz@mellanox.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "acddd5dd44d4fd9b45dd5ee69cd8b183052b1cdc", "tree": "d3c71be4ecbd48fc72164195d216d17d487bfcce", "parents": [ "162226a1dceaadc4fd61b03fb4ae8be8a1f5863e" ], "author": { "name": "Jack Morgenstein", "email": "jackm@dev.mellanox.co.il", "time": "Sun Nov 03 10:03:18 2013 +0200" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Mon Nov 04 16:19:06 2013 -0500" }, "message": "net/mlx4_core: Fix reg/unreg vlan/mac to conform to the firmware spec\n\nThe functions mlx4_register_vlan, mlx4_unregister_vlan, mlx4_register_mac,\nmlx4_unregister_mac all made illegal use of the out_param in multifunc mode\nto pass the port number. The firmware spec specifies that the port number\nshould be passed in bits 8..15 of the input-modifier field for ALLOC_RES and\nFREE_RES (sections 20.15.1 and 20.15.2).\n\nFor MAC register/unregister, this patch contains workarounds so that guests\nrunning previous kernels continue to work on a new Hypervisor, and guests\nrunning the new kernel will continue to work on old hypervisors.\n\nVlan registeration capability is still not operational in multifunction mode,\nsince the vlan wrapper functions are not implemented in this patch.\n\nSigned-off-by: Jack Morgenstein \u003cjackm@dev.mellanox.co.il\u003e\nSigned-off-by: Or Gerlitz \u003cogerlitz@mellanox.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "5930e8d0ab3689f1e239566443ca8f53e45e01cc", "tree": "e803509c76df87434c44e2eabc1f1092268e756a", "parents": [ "fe66bb2db51c9847c23682ef9c140bab6e14b0fa" ], "author": { "name": "Or Gerlitz", "email": "ogerlitz@mellanox.com", "time": "Tue Oct 15 16:55:22 2013 +0200" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Oct 17 15:10:50 2013 -0400" }, "message": "net/mlx4: Fix typo, move similar defs to same location\n\nSmall code cleanup:\n\n1. change MLX4_DEV_CAP_FLAGS2_REASSIGN_MAC_EN to MLX4_DEV_CAP_FLAG2_REASSIGN_MAC_EN\n\n2. put MLX4_SET_PORT_PRIO2TC and MLX4_SET_PORT_SCHEDULER in the same union with the\n other MLX4_SET_PORT_yyy\n\nSigned-off-by: Or Gerlitz \u003cogerlitz@mellanox.com\u003e\nSigned-off-by: Amir Vadai \u003camirv@mellanox.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "cc998ff8811530be521f6b316f37ab7676a07938", "tree": "a054b3bf4b2ef406bf756a6cfc9be2f9115f17ae", "parents": [ "57d730924d5cc2c3e280af16a9306587c3a511db", "0d40f75bdab241868c0eb6f97aef9f8b3a66f7b3" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Sep 05 14:54:29 2013 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Sep 05 14:54:29 2013 -0700" }, "message": "Merge git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-next\n\nPull networking changes from David Miller:\n \"Noteworthy changes this time around:\n\n 1) Multicast rejoin support for team driver, from Jiri Pirko.\n\n 2) Centralize and simplify TCP RTT measurement handling in order to\n reduce the impact of bad RTO seeding from SYN/ACKs. Also, when\n both timestamps and local RTT measurements are available prefer\n the later because there are broken middleware devices which\n scramble the timestamp.\n\n From Yuchung Cheng.\n\n 3) Add TCP_NOTSENT_LOWAT socket option to limit the amount of kernel\n memory consumed to queue up unsend user data. From Eric Dumazet.\n\n 4) Add a \"physical port ID\" abstraction for network devices, from\n Jiri Pirko.\n\n 5) Add a \"suppress\" operation to influence fib_rules lookups, from\n Stefan Tomanek.\n\n 6) Add a networking development FAQ, from Paul Gortmaker.\n\n 7) Extend the information provided by tcp_probe and add ipv6 support,\n from Daniel Borkmann.\n\n 8) Use RCU locking more extensively in openvswitch data paths, from\n Pravin B Shelar.\n\n 9) Add SCTP support to openvswitch, from Joe Stringer.\n\n 10) Add EF10 chip support to SFC driver, from Ben Hutchings.\n\n 11) Add new SYNPROXY netfilter target, from Patrick McHardy.\n\n 12) Compute a rate approximation for sending in TCP sockets, and use\n this to more intelligently coalesce TSO frames. Furthermore, add\n a new packet scheduler which takes advantage of this estimate when\n available. From Eric Dumazet.\n\n 13) Allow AF_PACKET fanouts with random selection, from Daniel\n Borkmann.\n\n 14) Add ipv6 support to vxlan driver, from Cong Wang\"\n\nResolved conflicts as per discussion.\n\n* git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-next: (1218 commits)\n openvswitch: Fix alignment of struct sw_flow_key.\n netfilter: Fix build errors with xt_socket.c\n tcp: Add missing braces to do_tcp_setsockopt\n caif: Add missing braces to multiline if in cfctrl_linkup_request\n bnx2x: Add missing braces in bnx2x:bnx2x_link_initialize\n vxlan: Fix kernel panic on device delete.\n net: mvneta: implement -\u003endo_do_ioctl() to support PHY ioctls\n net: mvneta: properly disable HW PHY polling and ensure adjust_link() works\n icplus: Use netif_running to determine device state\n ethernet/arc/arc_emac: Fix huge delays in large file copies\n tuntap: orphan frags before trying to set tx timestamp\n tuntap: purge socket error queue on detach\n qlcnic: use standard NAPI weights\n ipv6:introduce function to find route for redirect\n bnx2x: VF RSS support - VF side\n bnx2x: VF RSS support - PF side\n vxlan: Notify drivers for listening UDP port changes\n net: usbnet: update addr_assign_type if appropriate\n driver/net: enic: update enic maintainers and driver\n driver/net: enic: Exposing symbols for Cisco\u0027s low latency driver\n ...\n" }, { "commit": "f77c0162a339400ad16f657603fdc3bf11654fd3", "tree": "cf7bd72a5ac499549a697e305fc0cec46be79a6f", "parents": [ "436f2ad05a0b65b1467ddf51bc68171c381bf844" ], "author": { "name": "Hadar Hen Zion", "email": "hadarh@mellanox.com", "time": "Wed Aug 14 13:58:31 2013 +0300" }, "committer": { "name": "Roland Dreier", "email": "roland@purestorage.com", "time": "Wed Aug 28 09:53:56 2013 -0700" }, "message": "IB/mlx4: Add receive flow steering support\n\nImplement ib_create_flow() and ib_destroy_flow().\n\nTranslate the verbs structures provided by the user to HW structures\nand call the MLX4_QP_FLOW_STEERING_ATTACH/DETACH firmware commands.\n\nOn the ATTACH command completion, the firmware provides a 64-bit\nregistration ID, which is placed into struct mlx4_ib_flow that wraps\nthe instance of struct ib_flow which is retuned to caller. Later,\nthis reg ID is used for detaching that flow from the firmware.\n\nSigned-off-by: Hadar Hen Zion \u003chadarh@mellanox.com\u003e\nSigned-off-by: Or Gerlitz \u003cogerlitz@mellanox.com\u003e\nSigned-off-by: Roland Dreier \u003croland@purestorage.com\u003e\n" }, { "commit": "574e2af7c0af3273836def5e66f236521bb433c9", "tree": "5cfa1b92c82a86426a906ee9203a1548f0e3d993", "parents": [ "e216975ad97cfcfc436789aa66d59a0e93f337f7" ], "author": { "name": "Joe Perches", "email": "joe@perches.com", "time": "Thu Aug 01 16:17:48 2013 -0700" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Fri Aug 02 12:33:54 2013 -0700" }, "message": "include: Convert ethernet mac address declarations to use ETH_ALEN\n\nIt\u0027s convenient to have ethernet mac addresses use\nETH_ALEN to be able to grep for them a bit easier and\nalso to ensure that the addresses are __aligned(2).\n\nAdd #include \u003clinux/if_ether.h\u003e as necessary.\n\nSigned-off-by: Joe Perches \u003cjoe@perches.com\u003e\nAcked-by: Mauro Carvalho Chehab \u003cm.chehab@samsung.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "fe6f700d6cbb7e8a61711e325f53d9c9e0a42a4c", "tree": "1a3d5b95434519950cb10ed9e51b50c2583e53fd", "parents": [ "2d4b646613d6b12175b017aca18113945af1faf3" ], "author": { "name": "Yevgeny Petrilin", "email": "yevgenyp@mellanox.com", "time": "Sun Jul 28 18:54:21 2013 +0300" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Mon Jul 29 01:12:40 2013 -0700" }, "message": "net/mlx4_core: Respond to operation request by firmware\n\nThis commit adds new firmware command and new firmware event. The firmware\nraises the MLX4_EVENT_TYPE_OP_REQUIRED event in order to signal the driver it\nneeds to perform an administrative operation throughout the MLX4_CMD_GET_OP_REQ\ncommand. At the moment the supported operation is adding/removing multicast\nentries which are used by the firmware for handling NCSI traffic in B0\nsteering mode.\n\nAlso, had to swap the order of mlx4_init_mcg_table() and\nmlx4_init_eq_table() to make sure that driver will get events only after\nresources are initialized to handle it.\n\nSigned-off-by: Yevgeny Petrilin \u003cyevgenyp@mellanox.com\u003e\nSigned-off-by: Jack Morgenstein \u003cjackm@dev.mellanox.com\u003e\nSigned-off-by: Eugenia Emantayev \u003ceugenia@mellanox.com\u003e\nSigned-off-by: Amir Vadai \u003camirv@mellanox.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "0a6eac24583848490e9a9c02daef5e33c997431f", "tree": "c314d372854be62218ccaccfc93ebcd9cae29efb", "parents": [ "b01978cacfd7e3a4ca703b0e48f2e18de8865df5" ], "author": { "name": "Rony Efraim", "email": "ronye@mellanox.com", "time": "Thu Jun 27 19:05:22 2013 +0300" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Mon Jul 01 13:10:57 2013 -0700" }, "message": "net/mlx4_core: Add HW enforcement to VF link state\n\nWhen the firmware supports the UPDATE_QP command, if the VF link is disabled,\nblock all QPs opened by the VF, by programming the UPDATE_QP command to drop\nall RX \u0026 TX traffic to/from these QPs. Operates only in VST mode.\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": "b01978cacfd7e3a4ca703b0e48f2e18de8865df5", "tree": "73beae91335dbcf4a4129b3970be68f1082aff0b", "parents": [ "4e144d3a807d6d2aa03d2cb234d88ef1a140e8c3" ], "author": { "name": "Jack Morgenstein", "email": "jackm@dev.mellanox.co.il", "time": "Thu Jun 27 19:05:21 2013 +0300" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Mon Jul 01 13:10:22 2013 -0700" }, "message": "net/mlx4_core: Dynamic VST to VST vlan/qos changes\n\nWithin VST mode, enable modifying the vlan and/or qos\nfor a VF without requiring unbind/rebind.\n\nThis requires firmware which supports the UPDATE_QP command.\n(If the command is not available, we fall back to requiring\nunbind/bind to activate these changes).\n\nTo avoid race conditions with modify-qp on QPs that are affected\nby update-qp, this operation is performed on the comm_wq.\n\nIf the update operation succeeds for all the necessary QPs, a\nvlan_unregister is performed for the abandoned vlan id.\n\nSigned-off-by: Jack Morgenstein \u003cjackm@dev.mellanox.co.il\u003e\nSigned-off-by: Or Gerlitz \u003cogerlitz@mellanox.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "948e306d7d645af80ea331b60495710fe4fe12bb", "tree": "c66b8872cbc566df779b454d78c9c181c94440c8", "parents": [ "1d8faf48c74b8329a0322dc4b2a2030ae5003c86" ], "author": { "name": "Rony Efraim", "email": "ronye@mellanox.com", "time": "Thu Jun 13 13:19:11 2013 +0300" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Jun 13 17:51:04 2013 -0700" }, "message": "net/mlx4: Add VF link state support\n\nAdd support to change the link state of VF (vPort)\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": "7677fc965fba41d1386fa3b76a1f00303f02bb2d", "tree": "1324b8af989563dea62f9bc7f952c411f8d8a2dd", "parents": [ "4e8cf5b8a1da8dbe1ece5f084b99f5c902ea709b" ], "author": { "name": "Rony Efraim", "email": "ronye@mellanox.com", "time": "Wed May 08 22:22:35 2013 +0000" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Sat May 11 16:12:44 2013 -0700" }, "message": "net/mlx4: Strengthen VLAN tags/priorities enforcement in VST mode\n\nMake sure that the following steps are taken:\n\n- drop packets sent by the VF with vlan tag\n- block packets with vlan tag which are steered to the VF\n- drop/block tagged packets when the policy is priority-tagged\n- make sure VLAN stripping for received packets is set\n- make sure force UP bit for the VF QP is set\n\nUse enum values for all the above instead of numerical bit offsets.\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": "e0fd9affeb64088eff407dfc98bbd3a5c17ea479", "tree": "94ee7e3410fffb305aa6901053b85245686444a2", "parents": [ "3d15b798eafd3b6b3cc25f20747008ab9401a57f", "ea9627c800e99a902e2668ac8e6377f02d6f720a" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed May 08 15:29:48 2013 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed May 08 15:29:48 2013 -0700" }, "message": "Merge tag \u0027rdma-for-linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/roland/infiniband\n\nPull InfiniBand/RDMA changes from Roland Dreier:\n - XRC transport fixes\n - Fix DHCP on IPoIB\n - mlx4 preparations for flow steering\n - iSER fixes\n - miscellaneous other fixes\n\n* tag \u0027rdma-for-linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/roland/infiniband: (23 commits)\n IB/iser: Add support for iser CM REQ additional info\n IB/iser: Return error to upper layers on EAGAIN registration failures\n IB/iser: Move informational messages from error to info level\n IB/iser: Add module version\n mlx4_core: Expose a few helpers to fill DMFS HW strucutures\n mlx4_core: Directly expose fields of DMFS HW rule control segment\n mlx4_core: Change a few DMFS fields names to match firmare spec\n mlx4: Match DMFS promiscuous field names to firmware spec\n mlx4_core: Move DMFS HW structs to common header file\n IB/mlx4: Set link type for RAW PACKET QPs in the QP context\n IB/mlx4: Disable VLAN stripping for RAW PACKET QPs\n mlx4_core: Reduce warning message for SRQ_LIMIT event to debug level\n RDMA/iwcm: Don\u0027t touch cmid after dropping reference\n IB/qib: Correct qib_verbs_register_sysfs() error handling\n IB/ipath: Correct ipath_verbs_register_sysfs() error handling\n RDMA/cxgb4: Fix SQ allocation when on-chip SQ is disabled\n SRPT: Fix odd use of WARN_ON()\n IPoIB: Fix ipoib_hard_header() return value\n RDMA: Rename random32() to prandom_u32()\n RDMA/cxgb3: Fix uninitialized variable\n ...\n" }, { "commit": "2cccb9e4f3476da916146c2ec571c4f3eff738b1", "tree": "79524a1d906c0ab14331dea45e364f600c42b355", "parents": [ "e6b6a2316379feebebacec0979b3ebc5743e7502" ], "author": { "name": "Rony Efraim", "email": "ronye@mellanox.com", "time": "Thu Apr 25 05:22:30 2013 +0000" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Fri Apr 26 23:29:14 2013 -0400" }, "message": "net/mlx4: Add support to get VF config\n\nSupport getting VF config.\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": "e6b6a2316379feebebacec0979b3ebc5743e7502", "tree": "2b9c6722ba3191afad8e7d3eb9c937202ed53cee", "parents": [ "3f7fb021d081c8aaac1d0cf69a288d21625e872e" ], "author": { "name": "Rony Efraim", "email": "ronye@mellanox.com", "time": "Thu Apr 25 05:22:29 2013 +0000" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Fri Apr 26 23:29:14 2013 -0400" }, "message": "net/mlx4: Add VF MAC spoof checking support\n\nAdd ndo_set_vf_spoofchk support\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": "3f7fb021d081c8aaac1d0cf69a288d21625e872e", "tree": "6a3fd7ec98398d8715f026fecf8413c9124bb21b", "parents": [ "8f7ba3ca12f6f16526fa4a8aaf2cae91563eee69" ], "author": { "name": "Rony Efraim", "email": "ronye@mellanox.com", "time": "Thu Apr 25 05:22:28 2013 +0000" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Fri Apr 26 23:29:13 2013 -0400" }, "message": "net/mlx4: Add set VF default vlan ID and priority support\n\nAdd support to ndo_set_vf_vlan in the driver. Once this call is used the vport\nis considered to be in VST mode. In this mode, the PPF driver configures\nEthernet QPs created by this VF to use this vlan id and priority. Currently\nRoCE isn\u0027t supported on that mode.\n\nThe special values of VID\u003d4095 or VID\u003d0,UP\u003d0 are considered as VGT.\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": "8f7ba3ca12f6f16526fa4a8aaf2cae91563eee69", "tree": "4bf50f09e298b484003cd545b9ffa5d656915a98", "parents": [ "0eb62b93cbe0e8dea4cfe5ee761755e982663727" ], "author": { "name": "Rony Efraim", "email": "ronye@mellanox.com", "time": "Thu Apr 25 05:22:27 2013 +0000" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Fri Apr 26 23:29:13 2013 -0400" }, "message": "net/mlx4: Add set VF mac address support\n\nAdd ndo_set_vf_mac support which allows to set the MAC address\nfor mlx4 VF Ethernet NICs from the host.\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": "c2c19dc3c9a1585e58804041e5a328cde425403a", "tree": "9d869a0721ea8bae67f604aab06038eb5363e0e9", "parents": [ "bcf372971d471f6cb4070adb7bfc987d8b3d21f2" ], "author": { "name": "Hadar Hen Zion", "email": "hadarh@mellanox.com", "time": "Wed Apr 24 13:58:48 2013 +0000" }, "committer": { "name": "Roland Dreier", "email": "roland@purestorage.com", "time": "Wed Apr 24 17:51:30 2013 -0700" }, "message": "mlx4_core: Expose a few helpers to fill DMFS HW strucutures\n\nRe-arrange some of code which fills DMFS HW structures so we can use\nit from within the core driver and from the IB driver too, e.g when\nverbs DMFS structures are transformed into mlx4 hardware structs.\n\nAlso, add struct mlx4_flow_handle struct which will be of use by the\nDMFS verbs flow in the IB driver.\n\nSigned-off-by: Hadar Hen Zion \u003chadarh@mellanox.com\u003e\nSigned-off-by: Or Gerlitz \u003cogerlitz@mellanox.com\u003e\nSigned-off-by: Roland Dreier \u003croland@purestorage.com\u003e\n" }, { "commit": "bcf372971d471f6cb4070adb7bfc987d8b3d21f2", "tree": "d99f6faa007d10a1246a0f7a4853dff6e8b98d44", "parents": [ "ba60a3560ccf8e4253da9c0e8d1292375fd51796" ], "author": { "name": "Hadar Hen Zion", "email": "hadarh@mellanox.com", "time": "Wed Apr 24 13:58:47 2013 +0000" }, "committer": { "name": "Roland Dreier", "email": "roland@purestorage.com", "time": "Wed Apr 24 17:51:29 2013 -0700" }, "message": "mlx4_core: Directly expose fields of DMFS HW rule control segment\n\nSome of struct mlx4_net_trans_rule_hw_ctrl fields were packed into u32\nand accessed through bit field operations. Expose and access them\ndirectly as u8 or u16.\n\nSigned-off-by: Hadar Hen Zion \u003chadarh@mellanox.com\u003e\nSigned-off-by: Or Gerlitz \u003cogerlitz@mellanox.com\u003e\nSigned-off-by: Roland Dreier \u003croland@purestorage.com\u003e\n" }, { "commit": "ba60a3560ccf8e4253da9c0e8d1292375fd51796", "tree": "2ffc9ce3c20f3a3159eda08be16d8ac4d1a71f9e", "parents": [ "f91625398a2e6e03f0155861b630021ceddb42e7" ], "author": { "name": "Hadar Hen Zion", "email": "hadarh@mellanox.com", "time": "Wed Apr 24 13:58:46 2013 +0000" }, "committer": { "name": "Roland Dreier", "email": "roland@purestorage.com", "time": "Wed Apr 24 17:51:29 2013 -0700" }, "message": "mlx4_core: Change a few DMFS fields names to match firmare spec\n\nChange struct mlx4_net_trans_rule_hw_eth :: vlan_id name to vlan_tag\n\nChange struct mlx4_net_trans_rule_hw_ib :: r_u_qpn name to l3_qpn\n\nThe patch doesn\u0027t introduce any functional change or API change\ntowards the firmware.\n\nSigned-off-by: Hadar Hen Zion \u003chadarh@mellanox.com\u003e\nSigned-off-by: Or Gerlitz \u003cogerlitz@mellanox.com\u003e\nSigned-off-by: Roland Dreier \u003croland@purestorage.com\u003e\n" }, { "commit": "f91625398a2e6e03f0155861b630021ceddb42e7", "tree": "1c48ebeed68c41b8b11f1ac6a42ae6e369a1c0f6", "parents": [ "3cd0e1789ad39c3b7ed006ce53a83328bdadbee8" ], "author": { "name": "Hadar Hen Zion", "email": "hadarh@mellanox.com", "time": "Wed Apr 24 13:58:45 2013 +0000" }, "committer": { "name": "Roland Dreier", "email": "roland@purestorage.com", "time": "Wed Apr 24 17:51:29 2013 -0700" }, "message": "mlx4: Match DMFS promiscuous field names to firmware spec\n\nAlign the names used by enum mlx4_net_trans_promisc_mode with the\nactual firmware specification. The patch doesn\u0027t introduce any\nfunctional change or API change towards the firmware.\n\nRemove MLX4_FS_PROMISC_FUNCTION_PORT which isn\u0027t of use. Add new\nenums MLX4_FS_{UC/MC}_SNIFFER as a preparation step for sniffer\nsupport.\n\nSigned-off-by: Hadar Hen Zion \u003chadarh@mellanox.com\u003e\nSigned-off-by: Or Gerlitz \u003cogerlitz@mellanox.com\u003e\nSigned-off-by: Roland Dreier \u003croland@purestorage.com\u003e\n" }, { "commit": "3cd0e1789ad39c3b7ed006ce53a83328bdadbee8", "tree": "33ec91ea2c011974aca81e4c21b7c728aa9ccf60", "parents": [ "3528f69637723183f997573c7ec50ca869f219ba" ], "author": { "name": "Hadar Hen Zion", "email": "hadarh@mellanox.com", "time": "Wed Apr 24 13:58:44 2013 +0000" }, "committer": { "name": "Roland Dreier", "email": "roland@purestorage.com", "time": "Wed Apr 24 17:51:28 2013 -0700" }, "message": "mlx4_core: Move DMFS HW structs to common header file\n\nMove flow steering HW structures to be on the public mlx4 include\ndirectory, as a pre-step for the mlx4 IB driver to use them too.\n\nSigned-off-by: Hadar Hen Zion \u003chadarh@mellanox.com\u003e\nSigned-off-by: Or Gerlitz \u003cogerlitz@mellanox.com\u003e\nSigned-off-by: Roland Dreier \u003croland@purestorage.com\u003e\n" }, { "commit": "ec693d47010e8302e61e0bdf3f47496c5610641a", "tree": "8fbcfe632cb618318dcb1c9a9050afd025462b0a", "parents": [ "ddd8a6c12d7e494902a9435a9a7a543ef730b2d8" ], "author": { "name": "Amir Vadai", "email": "amirv@mellanox.com", "time": "Tue Apr 23 06:06:49 2013 +0000" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Wed Apr 24 16:30:14 2013 -0400" }, "message": "net/mlx4_en: Add HW timestamping (TS) support\n\nThe patch allows to enable/disable HW timestamping for incoming and/or\noutgoing packets. It adds and initializes all structs and callbacks\nneeded by kernel TS API.\nTo enable/disable HW timestamping appropriate ioctl should be used.\nCurrently HWTSTAMP_FILTER_ALL/NONE and HWTSAMP_TX_ON/OFF only are\nsupported.\nWhen enabling TS on receive flow - VLAN stripping will be disabled.\nAlso were made all relevant changes in RX/TX flows to consider TS request\nand plant HW timestamps into relevant structures.\nmlx4_ib was fixed to compile with new mlx4_cq_alloc() signature.\n\nSigned-off-by: Eugenia Emantayev \u003ceugenia@mellanox.com\u003e\nSigned-off-by: Amir Vadai \u003camirv@mellanox.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "ddd8a6c12d7e494902a9435a9a7a543ef730b2d8", "tree": "5dd526d9fef754d1de7fce210ee002e71cb1818d", "parents": [ "d998735f443427c1530cac5eeda0a45c8cb60a57" ], "author": { "name": "Eugenia Emantayev", "email": "eugenia@mellanox.com", "time": "Tue Apr 23 06:06:48 2013 +0000" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Wed Apr 24 16:30:13 2013 -0400" }, "message": "net/mlx4_core: Read HCA frequency and map internal clock\n\nRead HCA frequency, read PCI clock bar and offset, map internal clock to\nPCI bar.\n\nSigned-off-by: Eugenia Emantayev \u003ceugenia@mellanox.com\u003e\nSigned-off-by: Amir Vadai \u003camirv@mellanox.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "d998735f443427c1530cac5eeda0a45c8cb60a57", "tree": "79eadec3dff362d05e5f26e60dc3259ff335c8d0", "parents": [ "8bfadc32b7ef56ac6245b7ae0a06f5ece193004b" ], "author": { "name": "Eugenia Emantayev", "email": "eugenia@mellanox.com", "time": "Tue Apr 23 06:06:47 2013 +0000" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Wed Apr 24 16:30:13 2013 -0400" }, "message": "net/mlx4_core: Add timestamping device capability\n\nAdd new device capability for timestamping support and query FW to retrieve it.\n\nSigned-off-by: Eugenia Emantayev \u003ceugenia@mellanox.com\u003e\nSigned-off-by: Or Gerlitz \u003cogerlitz@mellanox.com\u003e\nSigned-off-by: Amir Vadai \u003camirv@mellanox.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "9f550553a470d3b05fc8bdced3a738d8ed5b8d8a", "tree": "8fd3f2332afb1827cc701192cca1153e63962185", "parents": [ "eec9e29fc5e9b417649830ab76a3aa10b90d2e9f" ], "author": { "name": "Shlomo Pongratz", "email": "shlomop@mellanox.com", "time": "Wed Apr 10 14:26:47 2013 +0000" }, "committer": { "name": "Roland Dreier", "email": "roland@purestorage.com", "time": "Tue Apr 16 22:42:55 2013 -0700" }, "message": "mlx4_core: Implement SRQ object lookup from srqn\n\nExpose a new API mlx4_srq_lookup() to retrive a SRQ based on its\nnumber. This API is needed in the mlx4_ib driver CQ polling logic,\nwhen a work completion is associated with a XRC TGT QP. Since a\ntarget QP may redirect to more than one XRC SRQ, the srq field in the\nQP has no usage and the real XRC SRQ need to be retrived using the\ninformation from the XRCETH IB header which is placed in the HW CQE.\n\nSigned-off-by: Shlomo Pongratz \u003cshlomop@mellanox.com\u003e\nSigned-off-by: Or Gerlitz \u003cogerlitz@mellanox.com\u003e\nSigned-off-by: Roland Dreier \u003croland@purestorage.com\u003e\n" }, { "commit": "540b3a39eea0056d305f17dda47eb185c4d56ddc", "tree": "4e3bbd87a657d525eca9601b2bae91aabb49b87a", "parents": [ "4d531aa8ab44983561857098592f860c8b3f5ec4" ], "author": { "name": "Or Gerlitz", "email": "ogerlitz@mellanox.com", "time": "Sun Apr 07 03:44:07 2013 +0000" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Sun Apr 07 16:55:46 2013 -0400" }, "message": "net/mlx4_en: Enable DCB ETS ops only when supported by the firmware\n\nEnable the DCB ETS ops only when supported by the firmware. For older firmware/cards\nwhich don\u0027t support ETS, advertize only PFC DCB ops.\n\nSigned-off-by: Eugenia Emantayev \u003ceugenia@mellanox.co.il\u003e\nSigned-off-by: Or Gerlitz \u003cogerlitz@mellanox.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "70a3a06d01ed9ca887316a881813cdefb8a20170", "tree": "fbdb7982040ba77818e4b738d76eef8bb06fb47f", "parents": [ "f6c0ffa8f0b0781f4954cb06f0a81d6c10c1b434", "ef4e359d9b9e2dc022f79840fd207796b524a893" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Tue Feb 26 11:41:08 2013 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Tue Feb 26 11:41:08 2013 -0800" }, "message": "Merge tag \u0027rdma-for-linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/roland/infiniband\n\nPull infiniband update from Roland Dreier:\n \"Main batch of InfiniBand/RDMA changes for 3.9:\n\n - SRP error handling fixes from Bart Van Assche\n\n - Implementation of memory windows for mlx4 from Shani Michaeli\n\n - Lots of cxgb4 HW driver fixes from Vipul Pandya\n\n - Make iSER work for virtual functions, other fixes from Or Gerlitz\n\n - Fix for bug in qib HW driver from Mike Marciniszyn\n\n - IPoIB fixes from me, Itai Garbi, Shlomo Pongratz, Yan Burman\n\n - Various cleanups and warning fixes from Julia Lawall, Paul Bolle,\n Wei Yongjun\"\n\n* tag \u0027rdma-for-linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/roland/infiniband: (41 commits)\n IB/mlx4: Advertise MW support\n IB/mlx4: Support memory window binding\n mlx4: Implement memory windows allocation and deallocation\n mlx4_core: Enable memory windows in {INIT, QUERY}_HCA\n mlx4_core: Disable memory windows for virtual functions\n IPoIB: Free ipoib neigh on path record failure so path rec queries are retried\n IB/srp: Fail I/O requests if the transport is offline\n IB/srp: Avoid endless SCSI error handling loop\n IB/srp: Avoid sending a task management function needlessly\n IB/srp: Track connection state properly\n IB/mlx4: Remove redundant NULL check before kfree\n IB/mlx4: Fix compiler warning about uninitialized \u0027vlan\u0027 variable\n IB/mlx4: Convert is_xxx variables in build_mlx_header() to bool\n IB/iser: Enable iser when FMRs are not supported\n IB/iser: Avoid error prints on EAGAIN registration failures\n IB/iser: Use proper define for the commands per LUN value advertised to SCSI ML\n IB/uverbs: Implement memory windows support in uverbs\n IB/core: Add \"type 2\" memory windows support\n mlx4_core: Propagate MR deregistration failures to caller\n mlx4_core: Rename MPT-related functions to have mpt_ prefix\n ...\n" }, { "commit": "6ff63e194066a1f14aee805366a1d79c541fddae", "tree": "d619416f696cf6859587fa5875845609a9ba6a4e", "parents": [ "804d6a89a5c0b076317966bcbcd7a63d42241831" ], "author": { "name": "Shani Michaeli", "email": "shanim@mellanox.com", "time": "Wed Feb 06 16:19:15 2013 +0000" }, "committer": { "name": "Roland Dreier", "email": "roland@purestorage.com", "time": "Mon Feb 25 10:44:32 2013 -0800" }, "message": "IB/mlx4: Support memory window binding\n\n* Implement memory windows binding in mlx4_ib_post_send.\n\n* Implement mlx4_ib_bind_mw by deferring to mlx4_ib_post_send.\n\n* Rename MLX4_WQE_FMR_PERM_* flags to MLX4_WQE_FMR_AND_BIND_PERM_*,\n indicating that they are used both for fast registration work\n requests, and for memory window bind work requests.\n\nSigned-off-by: Haggai Eran \u003chaggaie@mellanox.com\u003e\nSigned-off-by: Shani Michaeli \u003cshanim@mellanox.com\u003e\nSigned-off-by: Or Gerlitz \u003cogerlitz@mellanox.com\u003e\nSigned-off-by: Roland Dreier \u003croland@purestorage.com\u003e\n" }, { "commit": "804d6a89a5c0b076317966bcbcd7a63d42241831", "tree": "07360d8b8c6f099bfc5007b951844a133d4c43bd", "parents": [ "e448834e3545e02789897ab68905220aea39cd40" ], "author": { "name": "Shani Michaeli", "email": "shanim@mellanox.com", "time": "Wed Feb 06 16:19:14 2013 +0000" }, "committer": { "name": "Roland Dreier", "email": "roland@purestorage.com", "time": "Mon Feb 25 10:44:32 2013 -0800" }, "message": "mlx4: Implement memory windows allocation and deallocation\n\nImplement MW allocation and deallocation in mlx4_core and mlx4_ib.\nPass down the enable bind flag when registering memory regions.\n\nSigned-off-by: Haggai Eran \u003chaggaie@mellanox.com\u003e\nSigned-off-by: Shani Michaeli \u003cshanim@mellanox.com\u003e\nSigned-off-by: Or Gerlitz \u003cogerlitz@mellanox.com\u003e\nSigned-off-by: Roland Dreier \u003croland@purestorage.com\u003e\n" }, { "commit": "61083720702a329ed5952e32bda384e3bbc9093c", "tree": "e31fb437442c23fdab04f94b69d0c612b870478d", "parents": [ "b20e519a81d648aebfbb90811743cc86cd469a48" ], "author": { "name": "Shani Michaeli", "email": "shanim@mellanox.com", "time": "Wed Feb 06 16:19:09 2013 +0000" }, "committer": { "name": "Roland Dreier", "email": "roland@purestorage.com", "time": "Thu Feb 21 11:38:43 2013 -0800" }, "message": "mlx4_core: Propagate MR deregistration failures to caller\n\nMR deregistration fails when memory windows are bound to the MR.\nHandle such failures by propagating them to the caller ULP.\n\nSigned-off-by: Haggai Eran \u003chaggaie@mellanox.com\u003e\nSigned-off-by: Shani Michaeli \u003cshanim@mellanox.com\u003e\nSigned-off-by: Or Gerlitz \u003cogerlitz@mellanox.com\u003e\nSigned-off-by: Roland Dreier \u003croland@purestorage.com\u003e\n" }, { "commit": "aee38fadd25989c3e6d99fc08752e2d87601ffc1", "tree": "dd7306b19a5cc74dc4d563c6a9eadc71072c269b", "parents": [ "6950a235b86cf4e73d2b8e5476e7d0eb8f61af63" ], "author": { "name": "Shani Michaeli", "email": "shanim@mellanox.com", "time": "Wed Feb 06 16:19:07 2013 +0000" }, "committer": { "name": "Roland Dreier", "email": "roland@purestorage.com", "time": "Thu Feb 21 11:36:29 2013 -0800" }, "message": "IB/mlx4_ib: Remove local invalidate segment unused fields\n\nRemove unused fields from the local invalidate WQE segment structure.\n\nSigned-off-by: Haggai Eran \u003chaggaie@mellanox.com\u003e\nSigned-off-by: Shani Michaeli \u003cshanim@mellanox.com\u003e\nSigned-off-by: Or Gerlitz \u003cogerlitz@mellanox.com\u003e\nSigned-off-by: Roland Dreier \u003croland@purestorage.com\u003e\n" }, { "commit": "16a10ffd20a13215243bdba64c8e57ef277a55b9", "tree": "3454a2b1371b65db0134cc73c5ed920ab5bc5afb", "parents": [ "48e551ff3da5084371bbca93380368ee18050c9e" ], "author": { "name": "Yan Burman", "email": "yanb@mellanox.com", "time": "Thu Feb 07 02:25:22 2013 +0000" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Feb 07 23:26:12 2013 -0500" }, "message": "net/mlx4: Move Ethernet related functionality from mlx4_core to mlx4_en\n\nMove low level code that deals with management of Ethernet MACs and QPs from mlx4_core to mlx4_en.\nAlso convert the new functions to deal with MACs in form of char array instead of u64.\n\nActual functions moved:\nmlx4_replace_mac\nmlx4_get_eth_qp\nmlx4_put_eth_qp\n\nTo conduct this change, some functionality had to be exported from the core,\nthe following functions were added:\nmlx4_get_base_qp\n__mlx4_replace_mac (low level function for CX1/A0 compatibility)\n\nSigned-off-by: Yan Burman \u003cyanb@mellanox.com\u003e\nSigned-off-by: Amir Vadai \u003camirv@mellanox.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "955154fa33df2b74f0fea8e7c84df6dfd954dab2", "tree": "c53021a647f2587d29f451b9e8b4b24d645a70d8", "parents": [ "23537b732f5dd9c917767419272aeb65cc4b8cfd" ], "author": { "name": "Matan Barak", "email": "matanb@mellanox.com", "time": "Wed Jan 30 23:07:10 2013 +0000" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Jan 31 12:48:47 2013 -0500" }, "message": "net/mlx4_en: Don\u0027t reassign port mac address on firmware that supports it\n\nMac reassignments should only be done when not supported by the firmware. To\naccomplish that, checking firmware capability bit to know whether we should\nreassign macs in the driver.\n\nSigned-off-by: Matan Barak \u003cmatanb@mellanox.com\u003e\nSigned-off-by: Amir Vadai \u003camirv@mellanox.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "3c439b5586e9200f7e6287ee77c175c4d5b0eeed", "tree": "8194cef0f549fbb81aa70985401220c91f228ee7", "parents": [ "7b8157bedce2a7b5f07f7ec98c96cd6bd64a2a84" ], "author": { "name": "Jack Morgenstein", "email": "jackm@dev.mellanox.co.il", "time": "Thu Dec 06 17:12:00 2012 +0000" }, "committer": { "name": "Roland Dreier", "email": "roland@purestorage.com", "time": "Wed Dec 19 11:47:22 2012 -0800" }, "message": "mlx4_core: Allow choosing flow steering mode\n\nDevice managed flow steering will be enabled only under administrator\ndirective provided through setting the existing module parameter\nlog_num_mgm_entry_size to -1 (if the device actually supports flow\nsteering). If flow steering isn\u0027t requested or not available, the\ndriver will use the value of log_num_mgm_entry_size and B0 steering.\n\nSigned-off-by: Jack Morgenstein \u003cjackm@dev.mellanox.co.il\u003e\nSigned-off-by: Or Gerlitz \u003cogerlitz@mellanox.com\u003e\nSigned-off-by: Roland Dreier \u003croland@purestorage.com\u003e\n" }, { "commit": "08ff32352d6ff7083533dc1c25618d42f92ec28e", "tree": "00cc99b260b03c7c3e66a017f9dafcfa56b53364", "parents": [ "f4a75d2eb7b1e2206094b901be09adb31ba63681" ], "author": { "name": "Or Gerlitz", "email": "ogerlitz@mellanox.com", "time": "Sun Oct 21 14:59:24 2012 +0000" }, "committer": { "name": "Roland Dreier", "email": "roland@purestorage.com", "time": "Mon Nov 26 10:19:17 2012 -0800" }, "message": "mlx4: 64-byte CQE/EQE support\n\nConnectX-3 devices can use either 64- or 32-byte completion queue\nentries (CQEs) and event queue entries (EQEs). Using 64-byte\nEQEs/CQEs performs better because each entry is aligned to a complete\ncacheline. This patch queries the HCA\u0027s capabilities, and if it\nsupports 64-byte CQEs and EQES the driver will configure the HW to\nwork in 64-byte mode.\n\nThe 32-byte vs 64-byte mode is global per HCA and not per CQ or EQ.\n\nSince this mode is global, userspace (libmlx4) must be updated to work\nwith the configured CQE size, and guests using SR-IOV virtual\nfunctions need to know both EQE and CQE size.\n\nIn case one of the 64-byte CQE/EQE capabilities is activated, the\npatch makes sure that older guest drivers that use the QUERY_DEV_FUNC\ncommand (e.g as done in mlx4_core of Linux 3.3..3.6) will notice that\nthey need an update to be able to work with the PPF. This is done by\nchanging the returned pf_context_behaviour not to be zero any more. In\ncase none of these capabilities is activated that value remains zero\nand older guest drivers can run OK.\n\nThe SRIOV related flow is as follows\n\n1. the PPF does the detection of the new capabilities using\n QUERY_DEV_CAP command.\n\n2. the PPF activates the new capabilities using INIT_HCA.\n\n3. the VF detects if the PPF activated the capabilities using\n QUERY_HCA, and if this is the case activates them for itself too.\n\nNote that the VF detects that it must be aware to the new PF behaviour\nusing QUERY_FUNC_CAP. Steps 1 and 2 apply also for native mode.\n\nUser space notification is done through a new field introduced in\nstruct mlx4_ib_ucontext which holds device capabilities for which user\nspace must take action. This changes the binary interface so the ABI\ntowards libmlx4 exposed through uverbs is bumped from 3 to 4 but only\nwhen **needed** i.e. only when the driver does use 64-byte CQEs or\nfuture device capabilities which must be in sync by user space. This\npractice allows to work with unmodified libmlx4 on older devices (e.g\nA0, B0) which don\u0027t support 64-byte CQEs.\n\nIn order to keep existing systems functional when they update to a\nnewer kernel that contains these changes in VF and userspace ABI, a\nmodule parameter enable_64b_cqe_eqe must be set to enable 64-byte\nmode; the default is currently false.\n\nSigned-off-by: Eli Cohen \u003celi@mellanox.com\u003e\nSigned-off-by: Or Gerlitz \u003cogerlitz@mellanox.com\u003e\nSigned-off-by: Roland Dreier \u003croland@purestorage.com\u003e\n" }, { "commit": "47605df953985c2b792ac9f3ddf70d270b89adb8", "tree": "b95cef28771ea4982a5bca6130dab4ace79d622c", "parents": [ "afa8fd1db9f295a0c4130bc6d87bf8b05bdd0523" ], "author": { "name": "Jack Morgenstein", "email": "jackm@dev.mellanox.co.il", "time": "Fri Aug 03 08:40:57 2012 +0000" }, "committer": { "name": "Roland Dreier", "email": "roland@purestorage.com", "time": "Sun Sep 30 20:33:43 2012 -0700" }, "message": "mlx4: Modify proxy/tunnel QP mechanism so that guests do no calculations\n\nPreviously, the structure of a guest\u0027s proxy QPs followed the\nstructure of the PPF special qps (qp0 port 1, qp0 port 2, qp1 port 1,\nqp1 port 2, ...). The guest then did offset calculations on the\nsqp_base qp number that the PPF passed to it in QUERY_FUNC_CAP().\n\nThis is now changed so that the guest does no offset calculations\nregarding proxy or tunnel QPs to use. This change frees the PPF from\nneeding to adhere to a specific order in allocating proxy and tunnel\nQPs.\n\nNow QUERY_FUNC_CAP provides each port individually with its proxy\nqp0, proxy qp1, tunnel qp0, and tunnel qp1 QP numbers, and these are\nused directly where required (with no offset calculations).\n\nTo accomplish this change, several fields were added to the phys_caps\nstructure for use by the PPF and by non-SR-IOV mode:\n\n base_sqpn -- in non-sriov mode, this was formerly sqp_start.\n base_proxy_sqpn -- the first physical proxy qp number -- used by PPF\n base_tunnel_sqpn -- the first physical tunnel qp number -- used by PPF.\n\nThe current code in the PPF still adheres to the previous layout of\nsqps, proxy-sqps and tunnel-sqps. However, the PPF can change this\nlayout without affecting VF or (paravirtualized) PF code.\n\nSigned-off-by: Jack Morgenstein \u003cjackm@dev.mellanox.co.il\u003e\nSigned-off-by: Roland Dreier \u003croland@purestorage.com\u003e\n" }, { "commit": "afa8fd1db9f295a0c4130bc6d87bf8b05bdd0523", "tree": "725c4a2c6a9044ff9ac79e4f6a863393756953f0", "parents": [ "026149cbaada391d98f1cbec47c488cb548f753a" ], "author": { "name": "Jack Morgenstein", "email": "jackm@dev.mellanox.co.il", "time": "Fri Aug 03 08:40:56 2012 +0000" }, "committer": { "name": "Roland Dreier", "email": "roland@purestorage.com", "time": "Sun Sep 30 20:33:43 2012 -0700" }, "message": "mlx4: Paravirtualize Node Guids for slaves\n\nThis is necessary in order to support \u003e 1 VF/PF in a VM for software\nthat uses the node guid as a discriminator, such as librdmacm.\n\nSigned-off-by: Jack Morgenstein \u003cjackm@dev.mellanox.co.il\u003e\nSigned-off-by: Roland Dreier \u003croland@purestorage.com\u003e\n" }, { "commit": "026149cbaada391d98f1cbec47c488cb548f753a", "tree": "7996c6907496c5af0770aaaaa78a5fc851443789", "parents": [ "992e8e6e8781b71fd475bd1fd0555da7dba59966" ], "author": { "name": "Jack Morgenstein", "email": "jackm@dev.mellanox.co.il", "time": "Fri Aug 03 08:40:55 2012 +0000" }, "committer": { "name": "Roland Dreier", "email": "roland@purestorage.com", "time": "Sun Sep 30 20:33:42 2012 -0700" }, "message": "mlx4: Activate SR-IOV mode for IB\n\nRemove the error returns for IB ports from mlx4_ib_add,\nmlx4_INIT_PORT_wrapper, and mlx4_CLOSE_PORT_wrapper.\n\nCurrently, SRIOV is supported only for devices for which the\nlink layer is IB on all ports; RoCE support will be added later.\n\nSigned-off-by: Jack Morgenstein \u003cjackm@dev.mellanox.co.il\u003e\nSigned-off-by: Roland Dreier \u003croland@purestorage.com\u003e\n" }, { "commit": "efcd235d736ab05ef2b29d7fe1493a2f52b07b66", "tree": "7111fa75eb2290a2c55962e6e1cd25007bdf0be0", "parents": [ "c1e7e466120b80ce49e91af0c9da1ce6dee4844a" ], "author": { "name": "Jack Morgenstein", "email": "jackm@dev.mellanox.co.il", "time": "Fri Aug 03 08:40:52 2012 +0000" }, "committer": { "name": "Roland Dreier", "email": "roland@purestorage.com", "time": "Sun Sep 30 20:33:40 2012 -0700" }, "message": "net/mlx4_core: Adjustments to SET_PORT for IB SR-IOV\n\n1. Slaves may not set the IS_SM capability for the port.\n2. DEV_MGMT may not be set in multifunction mode.\n\nSigned-off-by: Jack Morgenstein \u003cjackm@dev.mellanox.co.il\u003e\nSigned-off-by: Roland Dreier \u003croland@purestorage.com\u003e\n" }, { "commit": "993c401e207946fa56f69c51e39f015e7108497b", "tree": "17ef6798c9723d6cbaa3c48b6d24cda765d7057d", "parents": [ "3cf69cc8dbebf15b99deb342ea422105ae9c2774" ], "author": { "name": "Jack Morgenstein", "email": "jackm@dev.mellanox.co.il", "time": "Fri Aug 03 08:40:48 2012 +0000" }, "committer": { "name": "Roland Dreier", "email": "roland@purestorage.com", "time": "Sun Sep 30 20:33:37 2012 -0700" }, "message": "mlx4_core: Add IB port-state machine and port mgmt event propagation\n\nFor an IB port, a slave should not show port active until that slave\nhas a valid alias-guid (provided by the subnet manager). Therefore\nthe port-up event should be passed to a slave only after both the port\nis up, and the slave\u0027s alias-guid has been set.\n\nAlso, provide the infrastructure for propagating port-management\nevents (client-reregister, etc) to slaves.\n\nSigned-off-by: Jack Morgenstein \u003cjackm@dev.mellanox.co.il\u003e\nSigned-off-by: Roland Dreier \u003croland@purestorage.com\u003e\n" }, { "commit": "54679e148287f0ca1bdd09264c908bacb9f19b3f", "tree": "8ef261b3d7a51f8c29c2f625273ae9b399cb0402", "parents": [ "fc06573dfaf8a33bc0533bb70c49de13fa5232a4" ], "author": { "name": "Jack Morgenstein", "email": "jackm@dev.mellanox.co.il", "time": "Fri Aug 03 08:40:43 2012 +0000" }, "committer": { "name": "Roland Dreier", "email": "roland@purestorage.com", "time": "Sun Sep 30 20:33:33 2012 -0700" }, "message": "mlx4: Implement QP paravirtualization and maintain phys_pkey_cache for smp_snoop\n\nThis requires:\n\n1. Replacing the paravirtualized P_Key index (inserted by the guest)\n with the real P_Key index.\n\n2. For UD QPs, placing the guest\u0027s true source GID index in the\n address path structure mgid field, and setting the ud_force_mgid\n bit so that the mgid is taken from the QP context and not from the\n WQE when posting sends.\n\n3. For UC and RC QPs, placing the guest\u0027s true source GID index in the\n address path structure mgid field.\n\n4. For tunnel and proxy QPs, setting the Q_Key value reserved for that\n proxy/tunnel pair.\n\nSince not all the above adjustments occur in all the QP transitions,\nthe QP transitions require separate wrapper functions.\n\nSecondly, initialize the P_Key virtualization table to its default\nvalues: Master virtualized table is 1-1 with the real P_Key table,\nguest virtualized table has P_Key index 0 mapped to the real P_Key\nindex 0, and all the other P_Key indices mapped to the reserved\n(invalid) P_Key at index 127.\n\nFinally, add logic in smp_snoop for maintaining the phys_P_Key_cache.\nand generating events on the master only if a P_Key actually changed.\n\nSigned-off-by: Jack Morgenstein \u003cjackm@dev.mellanox.co.il\u003e\nSigned-off-by: Roland Dreier \u003croland@purestorage.com\u003e\n" }, { "commit": "fc06573dfaf8a33bc0533bb70c49de13fa5232a4", "tree": "034d40d1b3f21a53ff52be78b8fcfea0222aa891", "parents": [ "e2c76824ca16a3e8443cc7b26abcb21af7c27b10" ], "author": { "name": "Jack Morgenstein", "email": "jackm@dev.mellanox.co.il", "time": "Fri Aug 03 08:40:42 2012 +0000" }, "committer": { "name": "Roland Dreier", "email": "roland@purestorage.com", "time": "Sun Sep 30 20:33:32 2012 -0700" }, "message": "IB/mlx4: Initialize SR-IOV IB support for slaves in master context\n\nAllocate SR-IOV paravirtualization resources and MAD demuxing contexts\non the master.\n\nThis has two parts. The first part is to initialize the structures to\ncontain the contexts. This is done at master startup time in\nmlx4_ib_init_sriov().\n\nThe second part is to actually create the tunneling resources required\non the master to support a slave. This is performed the master\ndetects that a slave has started up (MLX4_DEV_EVENT_SLAVE_INIT event\ngenerated when a slave initializes its comm channel).\n\nFor the master, there is no such startup event, so it creates its own\ntunneling resources when it starts up. In addition, the master also\ncreates the real special QPs. The ib_core layer on the master causes\ncreation of proxy special QPs, since the master is also\nparavirtualized at the ib_core layer.\n\nSigned-off-by: Jack Morgenstein \u003cjackm@dev.mellanox.co.il\u003e\nSigned-off-by: Roland Dreier \u003croland@purestorage.com\u003e\n" }, { "commit": "e2c76824ca16a3e8443cc7b26abcb21af7c27b10", "tree": "41d39111862e2e6faecb9ab356198e985c9bfe7d", "parents": [ "1ffeb2eb8be9936e9dc1f9af2d5f4c14d69a0d36" ], "author": { "name": "Jack Morgenstein", "email": "jackm@dev.mellanox.co.il", "time": "Fri Aug 03 08:40:41 2012 +0000" }, "committer": { "name": "Roland Dreier", "email": "roland@purestorage.com", "time": "Sun Sep 30 20:33:31 2012 -0700" }, "message": "mlx4_core: Add proxy and tunnel QPs to the reserved QP area\n\nIn addition, pass the proxy and tunnel QP numbers to slaves so the\ndriver can perform special QP paravirtualization.\n\nSigned-off-by: Jack Morgenstein \u003cjackm@dev.mellanox.co.il\u003e\nSigned-off-by: Roland Dreier \u003croland@purestorage.com\u003e\n" }, { "commit": "1ffeb2eb8be9936e9dc1f9af2d5f4c14d69a0d36", "tree": "1e79a4a6c3955a21c7802618d58b3bd7605a937e", "parents": [ "73aaa7418f8069103ca56fc620b3cd16c5a37d6e" ], "author": { "name": "Jack Morgenstein", "email": "jackm@dev.mellanox.co.il", "time": "Fri Aug 03 08:40:40 2012 +0000" }, "committer": { "name": "Roland Dreier", "email": "roland@purestorage.com", "time": "Sun Sep 30 20:33:30 2012 -0700" }, "message": "IB/mlx4: SR-IOV IB context objects and proxy/tunnel SQP support\n\n1. Introduce the basic SR-IOV parvirtualization context objects for\n multiplexing and demultiplexing MADs.\n2. Introduce support for the new proxy and tunnel QP types.\n\nThis patch introduces the objects required by the master for managing\nQP paravirtualization for guests.\n\nstruct mlx4_ib_sriov is created by the master only.\nIt is a container for the following:\n\n1. All the info required by the PPF to multiplex and de-multiplex MADs\n (including those from the PF). (struct mlx4_ib_demux_ctx demux)\n2. All the info required to manage alias GUIDs (i.e., the GUID at\n index 0 that each guest perceives. In fact, this is not the GUID\n which is actually at index 0, but is, in fact, the GUID which is at\n index[\u003cVF number\u003e] in the physical table.\n3. structures which are used to manage CM paravirtualization\n4. structures for managing the real special QPs when running in SR-IOV\n mode. The real SQPs are controlled by the PPF in this case. All\n SQPs created and controlled by the ib core layer are proxy SQP.\n\nstruct mlx4_ib_demux_ctx contains the information per port needed\nto manage paravirtualization:\n\n1. All multicast paravirt info\n2. All tunnel-qp paravirt info for the port.\n3. GUID-table and GUID-prefix for the port\n4. work queues.\n\nstruct mlx4_ib_demux_pv_ctx contains all the info for managing the\nparavirtualized QPs for one slave/port.\n\nstruct mlx4_ib_demux_pv_qp contains the info need to run an individual\nQP (either tunnel qp or real SQP).\n\nNote: We made use of the 2 most significant bits in enum\nmlx4_ib_qp_flags (based on enum ib_qp_create_flags in ib_verbs.h).\nWe need these bits in the low-level driver for internal purposes.\n\nSigned-off-by: Jack Morgenstein \u003cjackm@dev.mellanox.co.il\u003e\nSigned-off-by: Roland Dreier \u003croland@purestorage.com\u003e\n" }, { "commit": "7fb40f87c4195ec1728527f30bc744c47a45b366", "tree": "4a7356fc76c4db03731d7354c66f7b158682d295", "parents": [ "a8edc3bf05a3465726afdf635a820761fae0d50b" ], "author": { "name": "Hadar Hen Zion", "email": "hadarh@mellanox.com", "time": "Wed Sep 05 22:50:49 2012 +0000" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Fri Sep 07 12:55:59 2012 -0400" }, "message": "net/mlx4_core: Add security check / enforcement for flow steering rules set for VMs\n\nSince VFs may be mapped to VMs which aren\u0027t trusted entities, flow\nsteering rules attached through the wrapper on behalf of VFs must be\nchecked to make sure that their L2 specification relate to MAC address\nassigned to that VF, and add L2 specification if its missing.\n\nSigned-off-by: Hadar Hen Zion \u003chadarh@mellanox.com\u003e\nSigned-off-by: Or Gerlitz \u003cogerlitz@mellanox.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "a8edc3bf05a3465726afdf635a820761fae0d50b", "tree": "345e02de5037f0b36685d90717aadae4ef19ed69", "parents": [ "d90c92fee89ccd75ef2646f3bde0b4c0450666c3" ], "author": { "name": "Hadar Hen Zion", "email": "hadarh@mellanox.com", "time": "Wed Sep 05 22:50:48 2012 +0000" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Fri Sep 07 12:55:59 2012 -0400" }, "message": "net/mlx4_core: Put Firmware flow steering structures in common header files\n\nTo allow for usage of the flow steering Firmware structures in more locations over the driver,\nsuch as the resource tracker, move them from mcg.c to common header files.\n\nSigned-off-by: Hadar Hen Zion \u003chadarh@mellanox.com\u003e\nSigned-off-by: Or Gerlitz \u003cogerlitz@mellanox.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "5dedb9f3bd5bcb186313ea0c0cff8f2c525d4122", "tree": "88514547a6e95176e7a9dc2fbdc7fa7c1bec3107", "parents": [ "ddb03448274b95bff6df2a2f1a74d7eb4be529d3", "089117e1ad265625b523a4168f77f2521b18fd32" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Tue Jul 24 13:56:26 2012 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Tue Jul 24 13:56:26 2012 -0700" }, "message": "Merge tag \u0027rdma-for-3.6\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/roland/infiniband\n\nPull InfiniBand/RDMA changes from Roland Dreier:\n - Updates to the qib low-level driver\n - First chunk of changes for SR-IOV support for mlx4 IB\n - RDMA CM support for IPv6-only binding\n - Other misc cleanups and fixes\n\nFix up some add-add conflicts in include/linux/mlx4/device.h and\ndrivers/net/ethernet/mellanox/mlx4/main.c\n\n* tag \u0027rdma-for-3.6\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/roland/infiniband: (30 commits)\n IB/qib: checkpatch fixes\n IB/qib: Add congestion control agent implementation\n IB/qib: Reduce sdma_lock contention\n IB/qib: Fix an incorrect log message\n IB/qib: Fix QP RCU sparse warnings\n mlx4: Put physical GID and P_Key table sizes in mlx4_phys_caps struct and paravirtualize them\n mlx4_core: Allow guests to have IB ports\n mlx4_core: Implement mechanism for reserved Q_Keys\n net/mlx4_core: Free ICM table in case of error\n IB/cm: Destroy idr as part of the module init error flow\n mlx4_core: Remove double function declarations\n IB/mlx4: Fill the masked_atomic_cap attribute in query device\n IB/mthca: Fill in sq_sig_type in query QP\n IB/mthca: Warning about event for non-existent QPs should show event type\n IB/qib: Fix sparse RCU warnings in qib_keys.c\n net/mlx4_core: Initialize IB port capabilities for all slaves\n mlx4: Use port management change event instead of smp_snoop\n IB/qib: RCU locking for MR validation\n IB/qib: Avoid returning EBUSY from MR deregister\n IB/qib: Fix UC MR refs for immediate operations\n ...\n" }, { "commit": "d9236c3f10490cd0b3fd4516af12ba62dcbae0b0", "tree": "9fbfa70657455d1e22eecc2ef71f5e1dfa58c5be", "parents": [ "122733a189d687087364d6dc3ecc7c682554f6a0" ], "author": { "name": "Amir Vadai", "email": "amirv@mellanox.com", "time": "Wed Jul 18 22:33:51 2012 +0000" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Jul 19 08:34:37 2012 -0700" }, "message": "{NET,IB}/mlx4: Add rmap support to mlx4_assign_eq\n\nEnable callers of mlx4_assign_eq to supply a pointer to cpu_rmap.\nIf supplied, the assigned IRQ is tracked using rmap infrastructure.\n\nSigned-off-by: Amir Vadai \u003camirv@mellanox.com\u003e\nSigned-off-by: Or Gerlitz \u003cogerlitz@mellanox.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "af22d9de45caf8b2a99f2b27a927169c029528b4", "tree": "62299712c84e193a737d5d9e5ce474a01485a9a8", "parents": [ "0cc535a29916c6a0e6e6af0f3d42c2fe3b0b145d" ], "author": { "name": "Amir Vadai", "email": "amirv@mellanox.com", "time": "Wed Jul 18 22:33:49 2012 +0000" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Jul 19 08:34:37 2012 -0700" }, "message": "net/mlx4: Move MAC_MASK to a common place\n\nDefine this macro is one common place instead of duplicating it over the code\n\nSigned-off-by: Amir Vadai \u003camirv@mellanox.com\u003e\nSigned-off-by: Or Gerlitz \u003cogerlitz@mellanox.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "6634961c14d38ef64ec284c07aecb03d3dd03b4a", "tree": "a91bc4fa44848799c3035a9bfea1b288c2edac82", "parents": [ "105c320f6ac37af30252577d419e47b39edb5843" ], "author": { "name": "Jack Morgenstein", "email": "jackm@dev.mellanox.co.il", "time": "Tue Jun 19 11:21:44 2012 +0300" }, "committer": { "name": "Roland Dreier", "email": "roland@purestorage.com", "time": "Wed Jul 11 11:52:23 2012 -0700" }, "message": "mlx4: Put physical GID and P_Key table sizes in mlx4_phys_caps struct and paravirtualize them\n\nTo allow easy paravirtualization of P_Key and GID table sizes, keep\nparavirtualized sizes in mlx4_dev-\u003ecaps, but save the actual physical\nsizes from FW in struct: mlx4_dev-\u003ephys_cap.\n\nIn addition, in SR-IOV mode, do the following:\n\n1. Reduce reported P_Key table size by 1.\n This is done to reserve the highest P_Key index for internal use,\n for declaring an invalid P_Key in P_Key paravirtualization.\n We require a P_Key index which always contain an invalid P_Key\n value for this purpose (i.e., one which cannot be modified by\n the subnet manager). The way to do this is to reduce the\n P_Key table size reported to the subnet manager by 1, so that\n it will not attempt to access the P_Key at index #127.\n\n2. Paravirtualize the GID table size to 1. Thus, each guest sees\n only a single GID (at its paravirtualized index 0).\n\nIn addition, since we are paravirtualizing the GID table size to 1, we\nadd paravirtualization of the master GID event here (i.e., we do not\ndo ib_dispatch_event() for the GUID change event on the master, since\nits (only) GUID never changes).\n\nSigned-off-by: Jack Morgenstein \u003cjackm@dev.mellanox.co.il\u003e\nSigned-off-by: Roland Dreier \u003croland@purestorage.com\u003e\n" }, { "commit": "396f2feb05d7cc5549c611c05abfb4108cd1c6d6", "tree": "797eabb6ea6b2e85480ed9375849c7648d72ed7e", "parents": [ "2aca1172c2f5b27fbc37297574f716c1c15f4153" ], "author": { "name": "Jack Morgenstein", "email": "jackm@dev.mellanox.co.il", "time": "Tue Jun 19 11:21:42 2012 +0300" }, "committer": { "name": "Roland Dreier", "email": "roland@purestorage.com", "time": "Wed Jul 11 11:35:55 2012 -0700" }, "message": "mlx4_core: Implement mechanism for reserved Q_Keys\n\nThe SR-IOV special QP tunneling mechanism uses proxy special QPs\n(instead of the real special QPs) for MADs on guests. These proxy QPs\nsend their packets to a \"tunnel\" QP owned by the master. The master\nthen forwards the MAD (after any required paravirtualization) to the\nreal special QP, which sends out the MAD.\n\nFor security reasons (i.e., to prevent guests from sending MADs to\ntunnel QPs belonging to other guests), each proxy-tunnel QP pair is\nassigned a unique, reserved, Q_Key. These Q_Keys are available only\nfor proxy and tunnel QPs -- if the guest tries to use these Q_Keys\nwith other QPs, it will fail.\n\nThis patch introduces a mechanism for reserving a block of 64K Q_Keys\nfor proxy/tunneling use.\n\nThe patch introduces also two new fields into mlx4_dev: base_sqpn and\nbase_tunnel_sqpn.\n\nIn SR-IOV mode, the QP numbers for the \"real,\" proxy, and tunnel sqps\nare added to the reserved QPN area (so that they will not change).\nThere are 8 special QPs per port in the HCA, and each of them is\nassigned both a proxy and a tunnel QP, for each VF and for the PF as\nwell in SR-IOV mode.\n\nThe QPNs for these QPs are arranged as follows:\n 1. The real SQP numbers (8)\n 2. The proxy SQPs (8 * (max number of VFs + max number of PFs)\n 3. The tunnel SQPs (8 * (max number of VFs + max number of PFs)\n\nTo support these QPs, two new fields are added to struct mlx4_dev:\n\n base_sqp: this is the QP number of the first of the real SQPs\n base_tunnel_sqp: this is the qp number of the first qp in the tunnel\n sqp region. (On guests, this is the first tunnel\n sqp of the 8 which are assigned to that guest).\n\nIn addition, in SR-IOV mode, sqp_start is the number of the first\nproxy SQP in the proxy SQP region. (In guests, this is the first\nproxy SQP of the 8 which are assigned to that guest)\n\nNote that in non-SR-IOV mode, there are no proxies and no tunnels.\nIn this case, sqp_start is set to sqp_base -- which minimizes code\nchanges.\n\nSigned-off-by: Jack Morgenstein \u003cjackm@dev.mellanox.co.il\u003e\nSigned-off-by: Or Gerlitz \u003cogerlitz@mellanox.com\u003e\nSigned-off-by: Roland Dreier \u003croland@purestorage.com\u003e\n" }, { "commit": "2aca1172c2f5b27fbc37297574f716c1c15f4153", "tree": "8f6eb799bb0fa1c172ea45773d9d1c41ca934c87", "parents": [ "00f5ce99dc6ee46c3113393cc8fa12173f9bbcd7" ], "author": { "name": "Jack Morgenstein", "email": "jackm@dev.mellanox.co.il", "time": "Tue Jun 19 11:21:41 2012 +0300" }, "committer": { "name": "Roland Dreier", "email": "roland@purestorage.com", "time": "Tue Jul 10 09:57:06 2012 -0700" }, "message": "net/mlx4_core: Initialize IB port capabilities for all slaves\n\nWith IB SR-IOV, each slave has its own separate copy of the port\ncapabilities flags. For example, the master can run a subnet manager\n(which causes the IsSM bit to be set in the master\u0027s port\ncapabilities) without affecting the port capabilities seen by the\nslaves (the IsSM bit will be seen as cleared in the slaves).\n\nAlso add a static inline mlx4_master_func_num() to enhance readability\nof the code.\n\nSigned-off-by: Jack Morgenstein \u003cjackm@dev.mellanox.co.il\u003e\nSigned-off-by: Or Gerlitz \u003cogerlitz@mellanox.com\u003e\nSigned-off-by: Roland Dreier \u003croland@purestorage.com\u003e\n" }, { "commit": "00f5ce99dc6ee46c3113393cc8fa12173f9bbcd7", "tree": "cbcd8e2afcd492b3677bf3eedec29b34f249129c", "parents": [ "3045f0920367e625bbec7d66fadb444e673515af" ], "author": { "name": "Jack Morgenstein", "email": "jackm@dev.mellanox.co.il", "time": "Tue Jun 19 11:21:40 2012 +0300" }, "committer": { "name": "Roland Dreier", "email": "roland@purestorage.com", "time": "Tue Jul 10 09:47:10 2012 -0700" }, "message": "mlx4: Use port management change event instead of smp_snoop\n\nThe port management change event can replace smp_snoop. If the\ncapability bit for this event is set in dev-caps, the event is used\n(by the driver setting the PORT_MNG_CHG_EVENT bit in the async event\nmask in the MAP_EQ fw command). In this case, when the driver passes\nincoming SMP PORT_INFO SET mads to the FW, the FW generates port\nmanagement change events to signal any changes to the driver.\n\nIf the FW generates these events, smp_snoop shouldn\u0027t be invoked in\nib_process_mad(), or duplicate events will occur (once from the\nFW-generated event, and once from smp_snoop).\n\nIn the case where the FW does not generate port management change\nevents smp_snoop needs to be invoked to create these events. The flow\nin smp_snoop has been modified to make use of the same procedures as\nin the fw-generated-event event case to generate the port management\nevents (LID change, Client-rereg, Pkey change, and/or GID change).\n\nPort management change event handling required changing the\nmlx4_ib_event and mlx4_dispatch_event prototypes; the \"param\" argument\n(last argument) had to be changed to unsigned long in order to\naccomodate passing the EQE pointer.\n\nWe also needed to move the definition of struct mlx4_eqe from\nnet/mlx4.h to file device.h -- to make it available to the IB driver,\nto handle port management change events.\n\nSigned-off-by: Jack Morgenstein \u003cjackm@dev.mellanox.co.il\u003e\nSigned-off-by: Or Gerlitz \u003cogerlitz@mellanox.com\u003e\nSigned-off-by: Roland Dreier \u003croland@purestorage.com\u003e\n" }, { "commit": "752a50cab600c6d46c5a1921c6a6d2fb116c8a4b", "tree": "b6b4fac3ff771523c737eeb0ad922903977b2ec8", "parents": [ "6887a4131da3adaab011613776d865f4bcfb5678" ], "author": { "name": "Jack Morgenstein", "email": "jackm@dev.mellanox.co.il", "time": "Tue Jun 19 11:21:33 2012 +0300" }, "committer": { "name": "Roland Dreier", "email": "roland@purestorage.com", "time": "Sun Jul 08 18:05:05 2012 -0700" }, "message": "mlx4_core: Pass an invalid PCI id number to VFs\n\nCurrently, VFs have 0 in their dev-\u003ecaps.function field. This is a\nvalid pci id (usually of the PF). Instead, pass an invalid PCI id to\nthe VF via QUERY_FW, so that if the value gets accessed in the VF\ndriver, we\u0027ll catch the problem.\n\nSigned-off-by: Jack Morgenstein \u003cjackm@dev.mellanox.co.il\u003e\nSigned-off-by: Roland Dreier \u003croland@purestorage.com\u003e\n" }, { "commit": "592e49dda8122ab621cdc59cc429bdb968ee6364", "tree": "285ab049c09b7fd6a1614e869f2ceac3cfb73961", "parents": [ "1b9c6b064ef4fbc6f55485a8b5aab7ce889592c2" ], "author": { "name": "Hadar Hen Zion", "email": "hadarh@mellanox.co.il", "time": "Thu Jul 05 04:03:48 2012 +0000" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Sat Jul 07 16:23:06 2012 -0700" }, "message": "net/mlx4: Implement promiscuous mode with device managed flow-steering\n\nThe device managed flow steering API has three promiscuous modes:\n\n1. Uplink - captures all the packets that arrive to the port.\n2. Allmulti - captures all multicast packets arriving to the port.\n3. Function port - for future use, this mode is not implemented yet.\n\nUse these modes with the flow_attach and flow_detach firmware commands\naccording to the promiscuous state of the netdevice.\n\nSigned-off-by: Hadar Hen Zion \u003chadarh@mellanox.co.il\u003e\nSigned-off-by: Or Gerlitz \u003cogerlitz@mellanox.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "0ff1fb654bec0cff62ddf81a8a8edec4263604a0", "tree": "1556f95adf6802cfa0cce5efb4f1e1707888ce49", "parents": [ "8fcfb4db74352d3d447b7a559ad54f7577074d19" ], "author": { "name": "Hadar Hen Zion", "email": "hadarh@mellanox.co.il", "time": "Thu Jul 05 04:03:46 2012 +0000" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Sat Jul 07 16:23:05 2012 -0700" }, "message": "{NET, IB}/mlx4: Add device managed flow steering firmware API\n\nThe driver is modified to support three operation modes.\n\nIf supported by firmware use the device managed flow steering\nAPI, that which we call device managed steering mode. Else, if\nthe firmware supports the B0 steering mode use it, and finally,\nif none of the above, use the A0 steering mode.\n\nWhen the steering mode is device managed, the code is modified\nsuch that L2 based rules set by the mlx4_en driver for Ethernet\nunicast and multicast, and the IB stack multicast attach calls\ndone through the mlx4_ib driver are all routed to use the device\nmanaged API.\n\nWhen attaching rule using device managed flow steering API,\nthe firmware returns a 64 bit registration id, which is to be\nprovided during detach.\n\nCurrently the firmware is always programmed during HCA initialization\nto use standard L2 hashing. Future work should be done to allow\nconfiguring the flow-steering hash function with common, non\nproprietary means.\n\nSigned-off-by: Hadar Hen Zion \u003chadarh@mellanox.co.il\u003e\nSigned-off-by: Or Gerlitz \u003cogerlitz@mellanox.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "8fcfb4db74352d3d447b7a559ad54f7577074d19", "tree": "3d60738ca8b0323189041305fb03128c839104bf", "parents": [ "c96d97f4d127b61def87b3ee056bec20cfc265d1" ], "author": { "name": "Hadar Hen Zion", "email": "hadarh@mellanox.co.il", "time": "Thu Jul 05 04:03:45 2012 +0000" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Sat Jul 07 16:23:05 2012 -0700" }, "message": "net/mlx4_core: Add firmware commands to support device managed flow steering\n\nAdd support for firmware commands to attach/detach a new device managed\nsteering mode. Such network steering rules allow the user to provide an\nL2/L3/L4 flow specification to the firmware and have the device to steer\ntraffic that matches that specification to the provided QP.\n\nSigned-off-by: Hadar Hen Zion \u003chadarh@mellanox.co.il\u003e\nSigned-off-by: Or Gerlitz \u003cogerlitz@mellanox.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "c96d97f4d127b61def87b3ee056bec20cfc265d1", "tree": "84c89ead71dd4f7a0795aa703b60bb5c5eb2d5a4", "parents": [ "6d19993788e080edb557178cc6aba2d963edce4e" ], "author": { "name": "Hadar Hen Zion", "email": "hadarh@mellanox.co.il", "time": "Thu Jul 05 04:03:44 2012 +0000" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Sat Jul 07 16:23:05 2012 -0700" }, "message": "net/mlx4: Set steering mode according to device capabilities\n\nInstead of checking the firmware supported steering mode in various\nplaces in the code, add a dedicated field in the mlx4 device capabilities\nstructure which is written once during the initialization flow and read\nacross the code.\n\nThis also set the grounds for add new steering modes. Currently two modes\nare supported, and are named after the ConnectX HW versions A0 and B0.\n\nA0 steering uses mac_index, vlan_index and priority to steer traffic\ninto pre-defined range of QPs.\n\nB0 steering uses Ethernet L2 hashing rules and is enabled only\nif the firmware supports both unicast and multicast B0 steering,\n\nThe current steering modes are relevant for Ethernet traffic only,\nsuch that Infiniband steering remains untouched.\n\nSigned-off-by: Hadar Hen Zion \u003chadarh@mellanox.co.il\u003e\nSigned-off-by: Or Gerlitz \u003cogerlitz@mellanox.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "3fc929e2d693185aac2686e5e64e24eae10642a4", "tree": "c13786131ec3695af7caf97d55b92197cc6fdd6a", "parents": [ "30f7c73bed749814f4985b1ce4566fe64b9c25f0" ], "author": { "name": "Marcel Apfelbaum", "email": "marcela@dev.mellanox.co.il", "time": "Wed May 30 09:14:51 2012 +0000" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu May 31 18:18:16 2012 -0400" }, "message": "net/mlx4_core: Fix number of EQs used in ICM initialisation\n\nIn SRIOV mode, the number of EQs used when computing the total ICM size\nwas incorrect.\n\nTo fix this, we do the following:\n1. We add a new structure to mlx4_dev, mlx4_phys_caps, to contain physical HCA\n capabilities. The PPF uses the phys capabilities when it computes things\n like ICM size.\n\n The dev_caps structure will then contain the paravirtualized values, making\n bookkeeping much easier in SRIOV mode. We add a structure rather than a\n single parameter because there will be other fields in the phys_caps.\n\n The first field we add to the mlx4_phys_caps structure is num_phys_eqs.\n\n2. In INIT_HCA, when running in SRIOV mode, the \"log_num_eqs\" parameter\n passed to the FW is the number of EQs per VF/PF; each function (PF or VF)\n has this number of EQs available.\n\n However, the total number of EQs which must be allowed for in the ICM is\n (1 \u003c\u003c log_num_eqs) * (#VFs + #PFs). Rather than compute this quantity,\n we allocate ICM space for 1024 EQs (which is the device maximum\n number of EQs, and which is the value we place in the mlx4_phys_caps structure).\n\n For INIT_HCA, however, we use the per-function number of EQs as described\n above.\n\nSigned-off-by: Marcel Apfelbaum \u003cmarcela@dev.mellanox.co.il\u003e\nSigned-off-by: Jack Morgenstein \u003cjackm@dev.mellanox.co.il\u003e\nReviewed-by: Or Gerlitz \u003cogerlitz@mellanox.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "c23ddf7857bdb2e8001b0a058603497c765a580d", "tree": "f1d826612114a17d6ab543b7095adf04b5ba614a", "parents": [ "da4f58ffa08a7b7012fab9c205fa0f6ba40fec42", "cc169165c82e14ea43e313f937a0a475ca97e588" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Mon May 21 17:54:55 2012 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Mon May 21 17:54:55 2012 -0700" }, "message": "Merge tag \u0027rdma-for-3.5\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/roland/infiniband\n\nPull InfiniBand/RDMA changes from Roland Dreier:\n - Add ocrdma hardware driver for Emulex IB-over-Ethernet adapters\n - Add generic and mlx4 support for \"raw\" QPs: allow suitably privileged\n applications to send and receive arbitrary packets directly to/from\n the hardware\n - Add \"doorbell drop\" handling to the cxgb4 driver\n - A fairly large batch of qib hardware driver changes\n - A few fixes for lockdep-detected issues\n - A few other miscellaneous fixes and cleanups\n\nFix up trivial conflict in drivers/net/ethernet/emulex/benet/be.h.\n\n* tag \u0027rdma-for-3.5\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/roland/infiniband: (53 commits)\n RDMA/cxgb4: Include vmalloc.h for vmalloc and vfree\n IB/mlx4: Fix mlx4_ib_add() error flow\n IB/core: Fix IB_SA_COMP_MASK macro\n IB/iser: Fix error flow in iser ep connection establishment\n IB/mlx4: Increase the number of vectors (EQs) available for ULPs\n RDMA/cxgb4: Add query_qp support\n RDMA/cxgb4: Remove kfifo usage\n RDMA/cxgb4: Use vmalloc() for debugfs QP dump\n RDMA/cxgb4: DB Drop Recovery for RDMA and LLD queues\n RDMA/cxgb4: Disable interrupts in c4iw_ev_dispatch()\n RDMA/cxgb4: Add DB Overflow Avoidance\n RDMA/cxgb4: Add debugfs RDMA memory stats\n cxgb4: DB Drop Recovery for RDMA and LLD queues\n cxgb4: Common platform specific changes for DB Drop Recovery\n cxgb4: Detect DB FULL events and notify RDMA ULD\n RDMA/cxgb4: Drop peer_abort when no endpoint found\n RDMA/cxgb4: Always wake up waiters in c4iw_peer_abort_intr()\n mlx4_core: Change bitmap allocator to work in round-robin fashion\n RDMA/nes: Don\u0027t call event handler if pointer is NULL\n RDMA/nes: Fix for the ORD value of the connecting peer\n ...\n" }, { "commit": "b3416f4476f39f76fd5ca238e8332ba9b8f059d9", "tree": "abe8502067df86eba56d32d37f75f6e434acda27", "parents": [ "c0c1d3d761c32ece93b1c99d200591415e3cbf7c" ], "author": { "name": "Shlomo Pongratz", "email": "shlomop@mellanox.com", "time": "Sun Apr 29 17:04:25 2012 +0300" }, "committer": { "name": "Roland Dreier", "email": "roland@purestorage.com", "time": "Tue May 08 11:54:32 2012 -0700" }, "message": "mlx4_core: Add second capabilities flags field\n\nThis patch adds a 64-bit flags2 features member to struct mlx4_dev to\nexport further features of the hardware. The original flags field\ntracks features whose support bits are advertised by the firmware in\noffsets 0x40 and 0x44 of the query device capabilities command.\nflags2 will track features whose support bits are scattered at various\noffsets.\n\nRSS support is the first feature to be exported through flags2. RSS\ncapabilities are located at offset 0x2e. The size of the RSS\nindirection table is also given in this offset.\n\nSigned-off-by: Shlomo Pongratz \u003cshlomop@mellanox.com\u003e\nSigned-off-by: Roland Dreier \u003croland@purestorage.com\u003e\n" }, { "commit": "c0c1d3d761c32ece93b1c99d200591415e3cbf7c", "tree": "bd1099c23f115a3720f002e9fe27b832d34cf9e4", "parents": [ "d48b97b403d23f6df0b990cee652bdf9a52337a3" ], "author": { "name": "Oren Duer", "email": "oren@mellanox.co.il", "time": "Sun Apr 29 17:04:24 2012 +0300" }, "committer": { "name": "Roland Dreier", "email": "roland@purestorage.com", "time": "Tue May 08 11:48:09 2012 -0700" }, "message": "IB/mlx4: Put priority bits in WQE of IBoE MLX QP\n\nOtherwise CM packets going over MLX QP1 get fixed scheduling priority 0.\nWe want CM packets to get the same scheduling priority, and therefore\nmap to the same SQ (Schedule Queue) and eventually TC (Traffic Class),\nas the application requested for the actual QP used for the connection.\n\nSigned-off-by: Oren Duer \u003coren@mellanox.co.il\u003e\nSigned-off-by: Or Gerlitz \u003cogerlitz@mellanox.com\u003e\nSigned-off-by: Roland Dreier \u003croland@purestorage.com\u003e\n" }, { "commit": "e5395e92a470769d67369c002a41e59619f5214b", "tree": "2fd6a4d09c6e4514856f8e4762fe8ed301b8c9b5", "parents": [ "0e98b523c4a4119cbd17e58dff385cc329064694" ], "author": { "name": "Amir Vadai", "email": "amirv@mellanox.com", "time": "Wed Apr 04 21:33:25 2012 +0000" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Apr 05 05:08:03 2012 -0400" }, "message": "net/mlx4_core: set port QoS attributes\n\nAdding QoS firmware commands:\n- mlx4_en_SET_PORT_PRIO2TC - set UP \u003c\u003d\u003e TC\n- mlx4_en_SET_PORT_SCHEDULER - set promised BW, max BW and PG number\n\nSigned-off-by: Amir Vadai \u003camirv@mellanox.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "0e98b523c4a4119cbd17e58dff385cc329064694", "tree": "4c16b13b5f118cfcf607195d1683ed01ba508b77", "parents": [ "73a0d907301ece200d32b4e8ba2da2ca296b507f" ], "author": { "name": "Amir Vadai", "email": "amirv@mellanox.com", "time": "Wed Apr 04 21:33:24 2012 +0000" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Thu Apr 05 05:08:03 2012 -0400" }, "message": "net/mlx4_en: Force user priority by QP attribute\n\nInstead of relying on HW to change schedule queue by UP, schedule\nqueue is fixed for a tx_ring, and UP in WQE is ignored in this aspect. This\nresolves two issues with untagged traffic:\n1. untagged traffic has no UP in packet which is needed for QoS. The change\n above allows setting the schedule queue (and by that the UP) of such a stream.\n2. BlueFlame uses the same field used by vlan tag. So forcing UP from QPC\n allows using BF for untagged but prioritized traffic.\n\nIn old firmware that force UP is not supported, untagged traffic will not subject to\nQoS.\n\nBecause UP is set by QP, need to always have a tx ring per UP, even if pfcrx\nmodule paramter is false.\n\nSigned-off-by: Amir Vadai \u003camirv@mellanox.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "250f6715a4112d6686670c5a62ceb9305da94616", "tree": "ee1c9b41ed1fed8174efb312421902f19c877e8c", "parents": [ "11bcb32848ddb5ab28f09f142b625e2ba4d55c4c", "313162d0b83836e2f57e51b9b8650fb4b9c396ea" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Mar 24 10:41:37 2012 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Mar 24 10:41:37 2012 -0700" }, "message": "Merge tag \u0027device-for-3.4\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/paulg/linux\n\nPull \u003clinux/device.h\u003e avoidance patches from Paul Gortmaker:\n \"Nearly every subsystem has some kind of header with a proto like:\n\n\tvoid foo(struct device *dev);\n\n and yet there is no reason for most of these guys to care about the\n sub fields within the device struct. This allows us to significantly\n reduce the scope of headers including headers. For this instance, a\n reduction of about 40% is achieved by replacing the include with the\n simple fact that the device is some kind of a struct.\n\n Unlike the much larger module.h cleanup, this one is simply two\n commits. One to fix the implicit \u003clinux/device.h\u003e users, and then one\n to delete the device.h includes from the linux/include/ dir wherever\n possible.\"\n\n* tag \u0027device-for-3.4\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/paulg/linux:\n device.h: audit and cleanup users in main include dir\n device.h: cleanup users outside of linux/include (C files)\n" }, { "commit": "0c2fe82a9b106f1c03719783134360586d718a69", "tree": "ec1a18ca49f1e6724ef3a93727f5f73b7df61931", "parents": [ "5f0e685f316a1de6d3af8b23eaf46651faca32ab", "f0e88aeb19dac00ed2e09fd4d39ee65f32d5e968" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed Mar 21 10:33:42 2012 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed Mar 21 10:33:42 2012 -0700" }, "message": "Merge tag \u0027rdma-for-linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/roland/infiniband\n\nPull InfiniBand/RDMA changes for the 3.4 merge window from Roland Dreier:\n \"Nothing big really stands out; by patch count lots of fixes to the\n mlx4 driver plus some cleanups and fixes to the core and other\n drivers.\"\n\n* tag \u0027rdma-for-linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/roland/infiniband: (28 commits)\n mlx4_core: Scale size of MTT table with system RAM\n mlx4_core: Allow dynamic MTU configuration for IB ports\n IB/mlx4: Fix info returned when querying IBoE ports\n IB/mlx4: Fix possible missed completion event\n mlx4_core: Report thermal error events\n mlx4_core: Fix one more static exported function\n IB: Change CQE \"csum_ok\" field to a bit flag\n RDMA/iwcm: Reject connect requests if cmid is not in LISTEN state\n RDMA/cxgb3: Don\u0027t pass irq flags to flush_qp()\n mlx4_core: Get rid of redundant ext_port_cap flags\n RDMA/ucma: Fix AB-BA deadlock\n IB/ehca: Fix ilog2() compile failure\n IB: Use central enum for speed instead of hard-coded values\n IB/iser: Post initial receive buffers before sending the final login request\n IB/iser: Free IB connection resources in the proper place\n IB/srp: Consolidate repetitive sysfs code\n IB/srp: Use pr_fmt() and pr_err()/pr_warn()\n IB/core: Fix SDR rates in sysfs\n mlx4: Enforce device max FMR maps in FMR alloc\n IB/mlx4: Set bad_wr for invalid send opcode\n ...\n" }, { "commit": "313162d0b83836e2f57e51b9b8650fb4b9c396ea", "tree": "8f39f20515a924d392cf9a72db0918fda07a34c7", "parents": [ "51990e825431089747f8896244b5c17d3a6423f1" ], "author": { "name": "Paul Gortmaker", "email": "paul.gortmaker@windriver.com", "time": "Mon Jan 30 11:46:54 2012 -0500" }, "committer": { "name": "Paul Gortmaker", "email": "paul.gortmaker@windriver.com", "time": "Fri Mar 16 10:38:24 2012 -0400" }, "message": "device.h: audit and cleanup users in main include dir\n\nThe \u003clinux/device.h\u003e header includes a lot of stuff, and\nit in turn gets a lot of use just for the basic \"struct device\"\nwhich appears so often.\n\nClean up the users as follows:\n\n1) For those headers only needing \"struct device\" as a pointer\nin fcn args, replace the include with exactly that.\n\n2) For headers not really using anything from device.h, simply\ndelete the include altogether.\n\n3) For headers relying on getting device.h implicitly before\nbeing included themselves, now explicitly include device.h\n\n4) For files in which doing #1 or #2 uncovers an implicit\ndependency on some other header, fix by explicitly adding\nthe required header(s).\n\nAny C files that were implicitly relying on device.h to be\npresent have already been dealt with in advance.\n\nTotal removals from #1 and #2: 51. Total additions coming\nfrom #3: 9. Total other implicit dependencies from #4: 7.\n\nAs of 3.3-rc1, there were 110, so a net removal of 42 gives\nabout a 38% reduction in device.h presence in include/*\n\nSigned-off-by: Paul Gortmaker \u003cpaul.gortmaker@windriver.com\u003e\n" }, { "commit": "42872c7a5ed8d3ed49f51cb783978ca50369c564", "tree": "0deea7b61c163e6997a4634ab4ee310014984688", "parents": [ "d927d505c59a0c7353343174e6225c43c61fba6d", "db5a7a65c05867cb6ff5cb6d556a0edfce631d2d" ], "author": { "name": "Roland Dreier", "email": "roland@purestorage.com", "time": "Mon Mar 12 16:25:28 2012 -0700" }, "committer": { "name": "Roland Dreier", "email": "roland@purestorage.com", "time": "Mon Mar 12 16:25:28 2012 -0700" }, "message": "Merge branches \u0027misc\u0027 and \u0027mlx4\u0027 into for-next\n\nConflicts:\n\tdrivers/infiniband/hw/mlx4/main.c\n\tdrivers/net/ethernet/mellanox/mlx4/main.c\n\tinclude/linux/mlx4/device.h\n" }, { "commit": "096335b3f9830b90d13aee77252cf6f5f12a258c", "tree": "08d90d43ff9b4db64bc977210b67f2b05a3d15d9", "parents": [ "a9c766bb75ee2caad2735e41784387784ffd87db" ], "author": { "name": "Or Gerlitz", "email": "ogerlitz@mellanox.com", "time": "Wed Jan 11 19:02:17 2012 +0200" }, "committer": { "name": "Roland Dreier", "email": "roland@purestorage.com", "time": "Mon Mar 12 16:24:59 2012 -0700" }, "message": "mlx4_core: Allow dynamic MTU configuration for IB ports\n\nSet the MTU for IB ports in the driver instead of using the firmware\ndefault of 2KB (the driver defaults to 4KB). Allow for dynamic mtu\nconfiguration through a new, per-port sysfs entry.\n\nSince there\u0027s a dependency between the port MTU and the max number of\nHW VLs the port can support, apply a mim/max approach, using a loop\nthat goes down from the highest possible number of VLs to the lowest,\nusing the firmware return status to know whether the requested number\nof VLs is possible with a given MTU.\n\nFor now, as with the dynamic link type change / VPI support, the sysfs\nentry to change the mtu is exposed only when NOT running in SR-IOV\nmode. To allow changing the MTU for the master in SR-IOV mode,\nprimary-function-initiated FLR (Function Level Reset) needs to be\nimplemented.\n\nSigned-off-by: Or Gerlitz \u003cogerlitz@mellanox.com\u003e\nSigned-off-by: Roland Dreier \u003croland@purestorage.com\u003e\n" }, { "commit": "5984be90046fa978d94a5ec08bbf8f760cff2b30", "tree": "428b0735fd2fdd696e7d1601b554179d225fe67c", "parents": [ "e10903b087e425298fb86c6ad4b1a88735480db7" ], "author": { "name": "Jack Morgenstein", "email": "jackm@dev.mellanox.co.il", "time": "Tue Mar 06 15:50:49 2012 +0200" }, "committer": { "name": "Roland Dreier", "email": "roland@purestorage.com", "time": "Mon Mar 12 16:24:59 2012 -0700" }, "message": "mlx4_core: Report thermal error events\n\nPrint an error message when a thermal error async event is reported by the HW.\n\nSigned-off-by: Jack Morgenstein \u003cjackm@dev.mellanox.co.il\u003e\nSigned-off-by: Dotan Barak \u003cdotanb@mellanox.com\u003e\nSigned-off-by: Roland Dreier \u003croland@purestorage.com\u003e\n" }, { "commit": "8154c07fe14e387c5a7c7f2eb70534813634e45e", "tree": "04a62c867cb09976b415e8d6e04d41f0c2e793e1", "parents": [ "2e96691c31ecf749f48aa94ea837b95dd656f5c2" ], "author": { "name": "Or Gerlitz", "email": "ogerlitz@mellanox.com", "time": "Tue Mar 06 15:50:50 2012 +0200" }, "committer": { "name": "Roland Dreier", "email": "roland@purestorage.com", "time": "Tue Mar 06 17:25:18 2012 -0800" }, "message": "mlx4_core: Get rid of redundant ext_port_cap flags\n\nWhile doing the work for commit a6f7feae6d (\"IB/mlx4: pass SMP\nvendor-specific attribute MADs to firmware\") we realized that the\nfirmware would respond on all sorts of vendor-specific MADs.\nTherefore commit 97285b7817 (\"mlx4_core: Add extended port\ncapabilities support\") adds redundant code into the driver, since\nthere\u0027s no real reaon to maintain the extended capabilities of the\nport, as they can be queried on demand (e.g the FDR10 capability).\n\nThis patch reverts commit 97285b7817 and removes the check for\nextended caps from the mlx4_ib driver port query flow.\n\nSigned-off-by: Or Gerlitz \u003cogerlitz@mellanox.com\u003e\nSigned-off-by: Roland Dreier \u003croland@purestorage.com\u003e\n" }, { "commit": "9a9a232a9295deb3b6b5f4ce4290a7d05ff061fa", "tree": "2e63b1c1ecfe665579222829446358f58f980650", "parents": [ "2ab573c586446ebab9d7a3e340bd4ff8fc2c70d4" ], "author": { "name": "Yevgeny Petrilin", "email": "yevgenyp@mellanox.co.il", "time": "Tue Mar 06 04:04:47 2012 +0000" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Tue Mar 06 15:19:18 2012 -0500" }, "message": "net/mlx4: fixing sparse warnings for not declared, functions\n\nThe SET_PORT functions are implemented in port.c, which is part\nof mlx4_core, these functions are exported. The functions are in use by\nthe mlx4_en module (were originally part of mlx4_en).\nTheir declaration remained in mlx4_en module, moving the declaration to the right location.\n\nSigned-off-by: Yevgeny Petrilin \u003cyevgenyp@mellanox.co.il\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "622121719934f60378279eb440d3cec2fc3176d2", "tree": "ed5285b379efd994078cd2010968361bcc5759b3", "parents": [ "0f298a285f2e365cb34f69d1f79bb9fc996f683d" ], "author": { "name": "Eric Dumazet", "email": "eric.dumazet@gmail.com", "time": "Sat Feb 25 00:51:05 2012 +0000" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Sun Feb 26 14:22:05 2012 -0500" }, "message": "mlx4_en: dont change mac_header on xmit\n\nA driver xmit function is not allowed to change skb without special\ncare.\n\nmlx4_en_xmit() should not call skb_reset_mac_header() and instead should\nuse skb-\u003edata to access ethernet header.\n\nThis removes a dumb test : if (ethh \u0026\u0026 ethh-\u003eh_dest)\n\nAlso remove this slow mlx4_en_mac_to_u64() call, we can use\nget_unaligned() to get faster code.\n\nSigned-off-by: Eric Dumazet \u003ceric.dumazet@gmail.com\u003e\nCc: Yevgeny Petrilin \u003cyevgenyp@mellanox.co.il\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "a5bbe892da9441835cb6fece26d9bbd95fc820be", "tree": "2642ee58996918ae4a09be9d82d97a6516eb215b", "parents": [ "4ba6b8eaa9d67d03fb653cb8a13afca3236d4d70" ], "author": { "name": "Eli Cohen", "email": "eli@dev.mellanox.co.il", "time": "Thu Feb 09 18:10:06 2012 +0200" }, "committer": { "name": "Roland Dreier", "email": "roland@purestorage.com", "time": "Sun Feb 26 01:43:37 2012 -0800" }, "message": "mlx4: Enforce device max FMR maps in FMR alloc\n\nConnectX devices have a limit on the number of mappings that can be\ndone on an FMR before having to call sync_tpt. The current\nmlx4_ib driver reports the limit correctly in max_map_per_fmr in\n.query_device(), but mlx4_core doesn\u0027t check it when actually\nallocating FMRs.\n\nAdd a max_fmr_maps field to struct mlx4_caps and enforce this maximum\nvalue on FMR allocations.\n\nSigned-off-by: Eli Cohen \u003celi@mellanox.com\u003e\nSigned-off-by: Roland Dreier \u003croland@purestorage.com\u003e\n" }, { "commit": "93ece0c1a7ace88f10411dbb5643d2aa2fe00ebf", "tree": "c6a00edd4504b9c026bc555e3d74fe155cacadbf", "parents": [ "35fb9afbdeef9d5859d9a878d0372907baf119e1" ], "author": { "name": "Eugenia Emantayev", "email": "eugenia@mellanox.co.il", "time": "Thu Jan 19 09:45:05 2012 +0000" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Sun Jan 22 15:08:43 2012 -0500" }, "message": "mlx4_en: eth statistics modification\n\nIn native mode display all available staticstics.\nIn SRIOV mode on VF display only SW counters statistics,\nin SRIOV mode on hypervisor display SW counters and errors (got from FW)\nstatistics.\n\nSigned-off-by: Eugenia Emantayev \u003ceugenia@mellanox.co.il\u003e\nReviewed-by: Yevgeny Petrilin \u003cyevgenyp@mellanox.co.il\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "8d0fc7b61191c9433a4f738987b89e1d962eb637", "tree": "f538c085fefdeddd321a66ed0fb54a0cd86cbf34", "parents": [ "58a60168d12c4e5be21c29420a3de4a41ef3470f" ], "author": { "name": "Yevgeny Petrilin", "email": "yevgenyp@mellanox.co.il", "time": "Mon Dec 19 04:00:34 2011 +0000" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Mon Dec 19 14:57:07 2011 -0500" }, "message": "mlx4_core: Changing link sensing logic\n\nNew FW can give clues to driver regarding default port type\nand whether or not we should default to link sensing on the port.\n\n2 bits are added to QUERY_PORT command:\n1. suggested_type: This bit gives a hint whether the default port type should be\n IB or Ethernet.\n The driver will use this hint in case the user didn\u0027t specify explicitly the link layer\n type he wants to set.\n2. default_sense: If this bit is set, we would sense the port type on start-up\n and default the port to link sensing\n\nSigned-off-by: Yevgeny Petrilin \u003cyevgenyp@mellanox.co.il\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "58a60168d12c4e5be21c29420a3de4a41ef3470f", "tree": "407a7f5c7ef86a4346577ce1860bf686d40d4b93", "parents": [ "9f6e20cee6253c4ca5faacba8dbd09ebe70132ed" ], "author": { "name": "Yevgeny Petrilin", "email": "yevgenyp@mellanox.co.il", "time": "Mon Dec 19 04:00:26 2011 +0000" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Mon Dec 19 14:57:06 2011 -0500" }, "message": "mlx4: capability for link sensing\n\nFor ConnectX3 devices, we allow link sensing only if FW explicitly\nreported it supports the feature.\nFor older versions (ConnectX1 and 2), if the card supports both link layer types\n(Ethenet and Infiniband), link sensing is supported.\n\nSigned-off-by: Yevgeny Petrilin \u003cyevgenyp@mellanox.co.il\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "ab9c17a009ee8eb8c667f22dc0be0709effceab9", "tree": "b85acd5d8cf28a42ef5f7361ef43b4195b89912d", "parents": [ "d81c7186aa16a0da9e39961af6bad0c855a5d684" ], "author": { "name": "Jack Morgenstein", "email": "jackm@dev.mellanox.co.il", "time": "Tue Dec 13 04:18:30 2011 +0000" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Tue Dec 13 13:56:08 2011 -0500" }, "message": "mlx4_core: Modify driver initialization flow to accommodate SRIOV for Ethernet\n\n1. Added module parameters sr_iov and probe_vf for controlling enablement of\n SRIOV mode.\n2. Increased default max num-qps, num-mpts and log_num_macs to accomodate\n SRIOV mode\n3. Added port_type_array as a module parameter to allow driver startup with\n ports configured as desired.\n In SRIOV mode, only ETH is supported, and this array is ignored; otherwise,\n for the case where the FW supports both port types (ETH and IB), the\n port_type_array parameter is used.\n By default, the port_type_array is set to configure both ports as IB.\n4. When running in sriov mode, the master needs to initialize the ICM eq table\n to hold the eq\u0027s for itself and also for all the slaves.\n5. mlx4_set_port_mask() now invoked from mlx4_init_hca, instead of in mlx4_dev_cap.\n6. Introduced sriov VF (slave) device startup/teardown logic (mainly procedures\n mlx4_init_slave, mlx4_slave_exit, mlx4_slave_cap, mlx4_slave_exit and flow\n modifications in __mlx4_init_one, mlx4_init_hca, and mlx4_setup_hca).\n VFs obtain their startup information from the PF (master) device via the\n comm channel.\n7. In SRIOV mode (both PF and VF), MSI_X must be enabled, or the driver\n aborts loading the device.\n8. Do not allow setting port type via sysfs when running in SRIOV mode.\n9. mlx4_get_ownership: Currently, only one PF is supported by the driver.\n If the HCA is burned with FW which enables more than one PF, only one\n of the PFs is allowed to run. The first one up grabs a FW ownership\n semaphone -- all other PFs will find that semaphore taken, and the\n driver will not allow them to run.\n\nSigned-off-by: Jack Morgenstein \u003cjackm@dev.mellanox.co.il\u003e\nSigned-off-by: Yevgeny Petrilin \u003cyevgenyp@mellanox.co.il\u003e\nSigned-off-by: Liran Liss \u003cliranl@mellanox.co.il\u003e\nSigned-off-by: Marcel Apfelbaum \u003cmarcela@mellanox.co.il\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "2b8fb2867ca2736a715a88067fd0ec2904777cbe", "tree": "507a7645aaeda05151045157756e1a828e60fe11", "parents": [ "5b4c4d36860ef1c411d0669ffc15090417a33389" ], "author": { "name": "Marcel Apfelbaum", "email": "marcela@dev.mellanox.co.il", "time": "Tue Dec 13 04:16:56 2011 +0000" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Tue Dec 13 13:56:07 2011 -0500" }, "message": "mlx4_core: mtts resources units changed to offset\n\nIn the previous implementation mtts are managed by:\n1. order - log(mtt segments), \u0027mtt segment\u0027 groups several mtts together.\n2. first_seg - segment location relative to mtt table.\nIn the current implementation:\n1. order - log(mtts) rather than segments\n2. offset - mtt index in mtt table\n\nNote: The actual mtt allocation is made in segments but it is\n transparent to callers.\n\nRational: The mtt resource holders are not interested on how the allocation\n of mtt is done, but rather on how they will use it.\n\nSigned-off-by: Marcel Apfelbaum \u003cmarcela@dev.mellanox.co.il\u003e\nReviewed-by: Jack Morgenstein \u003cjackm@dev.mellanox.co.il\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "ffe455ad04681f3fc48eef595fe526a795f809a3", "tree": "ba21abb4371d780357dd1a91810dd171ce3b05b2", "parents": [ "0ec2c0f86d31ab36547307f133b0016006bdc6b5" ], "author": { "name": "Eugenia Emantayev", "email": "eugenia@mellanox.co.il", "time": "Tue Dec 13 04:16:21 2011 +0000" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Tue Dec 13 13:56:07 2011 -0500" }, "message": "mlx4: Ethernet port management modifications\n\nThe physical port is now common to the PF and VFs.\nThe port resources and configuration is managed by the PF, VFs can\nonly influence the MTU of the port, it is set as max among all functions,\nEach function allocates RX buffers of required size to meet it\u0027s MTU enforcement.\nPort management code was moved to mlx4_core, as the mlx4_en module is\nvirtualization unaware\n\nMove handling qp functionality to mlx4_get_eth_qp/mlx4_put_eth_qp\nincluding reserve/release range and add/release unicast steering.\nLet mlx4_register/unregister_mac deal only with MAC (un)registration.\n\nSigned-off-by: Eugenia Emantayev \u003ceugenia@mellanox.co.il\u003e\nSigned-off-by: Yevgeny Petrilin \u003cyevgenyp@mellanox.co.il\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "f5311ac109b21c9b47118655a5b6d887bcc686f8", "tree": "08b36eb29661485c60f101bb9f3f50be8a0eff2c", "parents": [ "f9baff509f8a05a79626defdbdf4f4aa4efd373b" ], "author": { "name": "Jack Morgenstein", "email": "jackm@dev.mellanox.co.il", "time": "Tue Dec 13 04:12:13 2011 +0000" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Tue Dec 13 13:56:05 2011 -0500" }, "message": "mlx4_core: Reduce number of PD bits to 17\n\nWhen SRIOV is enabled on the chip (at FW burning time),\nthe HCA uses only 17 bits for the PD. The remaining 7 high-order bits\nare ignored.\n\nChange the allocator to return only 17 bits for the PD. The MSB 7\nbits will be used to encode the slave number for consistency\nchecking later on in the resource tracker.\n\nSigned-off-by: Jack Morgenstein \u003cjackm@dev.mellanox.co.il\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "f9baff509f8a05a79626defdbdf4f4aa4efd373b", "tree": "d4f0e425cd8c8999775f0f135c9825e3bbdc180c", "parents": [ "65dab25deb8da7dba4b6dd0145a9143be7f8369f" ], "author": { "name": "Jack Morgenstein", "email": "jackm@dev.mellanox.co.il", "time": "Tue Dec 13 04:10:51 2011 +0000" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Tue Dec 13 13:56:05 2011 -0500" }, "message": "mlx4_core: Add \"native\" argument to mlx4_cmd and its callers (where needed)\n\nFor SRIOV, some Hypervisor commands can be executed directly (native \u003d 1).\nOthers should go through the command wrapper flow (for tracking resource\nusage, for example, or for changing some HCA configurations that slaves\nneed to be notified of).\n\nThis patch sets the groundwork for this capability -- adding the correct\nvalue of \"native\" in each case.\n\nNote that if SRIOV is not activated, this parameter has no effect.\n\nSigned-off-by: Jack Morgenstein \u003cjackm@dev.mellanox.co.il\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "65dab25deb8da7dba4b6dd0145a9143be7f8369f", "tree": "a494864c9c97eef341f30529d71feceb08dcdb64", "parents": [ "623ed84b1f9553bc962c2aca92f488aa6f27ecd1" ], "author": { "name": "Jack Morgenstein", "email": "jackm@dev.mellanox.co.il", "time": "Tue Dec 13 04:10:41 2011 +0000" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Tue Dec 13 13:56:05 2011 -0500" }, "message": "mlx4: Extanding port_mask functionality\n\nPort mask now has additional state.\nPort can be set as \"none\". In this case neither the mlx4_en or mlx4_ib\ndrivers take ownership of the port.\nIn multifunction mode there is an option to set the vfs as single ported devices.\n(in single function mode, both physical ports belong to same function)\n\nSigned-off-by: Jack Morgenstein \u003cjackm@dev.mellanox.co.il\u003e\nSigned-off-by: Yevgeny Petrilin \u003cyevgenyp@mellanox.co.il\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "623ed84b1f9553bc962c2aca92f488aa6f27ecd1", "tree": "e83f5c25e071e3dcef12122166a1996c0027c612", "parents": [ "9f048bfba15a22d1d1ce0c1f44567fa16bed4d25" ], "author": { "name": "Jack Morgenstein", "email": "jackm@dev.mellanox.co.il", "time": "Tue Dec 13 04:10:33 2011 +0000" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Tue Dec 13 13:56:05 2011 -0500" }, "message": "mlx4_core: initial header-file changes for SRIOV support\n\nThese changes will not affect module operation as yet. They\nare only to get some structs and enums in place for use by\nsubsequent patches (making those smaller).\n\nAdded here:\n* sriov state structs and inlines (mlx4_is_master/slave/mfunc)\n* comm-channel and vhcr support structures\n* enum values for new FW and comm-channel virtual commands\n (i.e., commands, passed via the comm channel to the PF-driver).\n* prototypes for many command wrapper functions (used by the\n PF context for processing FW commands passed to it by the VFs).\n* struct mlx4_eqe is moved from eq.c to mlx4.h (it will be used\n by other mlx4_core source files).\n\nSigned-off-by: Jack Morgenstein \u003cjackm@dev.mellanox.co.il\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "60d6fe99e4a507f77b63c090eb8aacb67e21687a", "tree": "52473d54554d44fbaf2b020fd203965b934ccaeb", "parents": [ "559a9f1d354b577af28f84181751820ff7d29feb" ], "author": { "name": "Amir Vadai", "email": "amirv@mellanox.co.il", "time": "Sat Nov 26 19:55:19 2011 +0000" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Sun Nov 27 17:17:04 2011 -0500" }, "message": "net/mlx4_en: adding loopback support\n\nDevice must be in promiscuous mode or DMAC must be same as the host MAC, or\nelse packet will be dropped by the HW rx filtering.\n\nSigned-off-by: Amir Vadai \u003camirv@mellanox.co.il\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "559a9f1d354b577af28f84181751820ff7d29feb", "tree": "e148a9161da5a03d8778e05d1492955f8d38e0ac", "parents": [ "f0ab34f011d805ce5b1a341409c9c26f0fc8252b" ], "author": { "name": "Oren Duer", "email": "oren@mellanox.co.il", "time": "Sat Nov 26 19:55:15 2011 +0000" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Sun Nov 27 17:17:04 2011 -0500" }, "message": "net/mlx4_en: fix WOL handlers were always looking at port2 capability bit\n\nThere are 2 capability bits for WOL, one for each port.\nWOL handlers were looking only on the second bit, regardless of the port.\n\nSigned-off-by: Oren Duer \u003coren@mellanox.co.il\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "876f6e67d1c617c098c67934a8d00b065bb9688b", "tree": "1d66f498806971c4369f6d152046bcb163ee528f", "parents": [ "fd4f862717ece652a76b4bb6c6d2c9656dbf0f7e" ], "author": { "name": "Or Gerlitz", "email": "ogerlitz@mellanox.com", "time": "Sat Nov 26 19:54:58 2011 +0000" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Sun Nov 27 17:17:03 2011 -0500" }, "message": "net/mlx4: move RSS related definitions to be global\n\nTowards adding RSS support for IB drivers/application who use\nthe mlx4 HW, make the RSS related definitions global and change\nthe mlx4_en driver to use them.\n\nSigned-off-by: Or Gerlitz \u003cogerlitz@mellanox.com\u003e\nSigned-off-by: Shlomo Pongratz \u003cshlomop@mellanox.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "f470f8d4e702593ee1d0852871ad80373bce707b", "tree": "85a67e65c5e5b9777639bd8f4c763a4cf8787e0e", "parents": [ "dc47d3810cdcb4f32bfa31d50f26af97aced0638", "504255f8d0480cf293962adf4bc3aecac645ae71" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Tue Nov 01 10:51:38 2011 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Tue Nov 01 10:51:38 2011 -0700" }, "message": "Merge branch \u0027for-linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/roland/infiniband\n\n* \u0027for-linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/roland/infiniband: (62 commits)\n mlx4_core: Deprecate log_num_vlan module param\n IB/mlx4: Don\u0027t set VLAN in IBoE WQEs\u0027 control segment\n IB/mlx4: Enable 4K mtu for IBoE\n RDMA/cxgb4: Mark QP in error before disabling the queue in firmware\n RDMA/cxgb4: Serialize calls to CQ\u0027s comp_handler\n RDMA/cxgb3: Serialize calls to CQ\u0027s comp_handler\n IB/qib: Fix issue with link states and QSFP cables\n IB/mlx4: Configure extended active speeds\n mlx4_core: Add extended port capabilities support\n IB/qib: Hold links until tuning data is available\n IB/qib: Clean up checkpatch issue\n IB/qib: Remove s_lock around header validation\n IB/qib: Precompute timeout jiffies to optimize latency\n IB/qib: Use RCU for qpn lookup\n IB/qib: Eliminate divide/mod in converting idx to egr buf pointer\n IB/qib: Decode path MTU optimization\n IB/qib: Optimize RC/UC code by IB operation\n IPoIB: Use the right function to do DMA unmap pages\n RDMA/cxgb4: Use correct QID in insert_recv_cqe()\n RDMA/cxgb4: Make sure flush CQ entries are collected on connection close\n ...\n" }, { "commit": "504255f8d0480cf293962adf4bc3aecac645ae71", "tree": "5ae8663d2bc2eb9e796bfb689f3e462a14742e06", "parents": [ "2be6053318aa4f2787b7bc5cf5160017d7d1586b", "615eb715ae10cbaa8079ab8cacf8f4596be4087a", "f7cc25d018f1e9af6767ee7774bbe83452e9fdf4", "d32ae393dbf0daf778f9e33b0bc6591cd102391e", "a5e12dff757b562bbecd6a2359fdc4c43d4d97de", "3e60a77ea218d1915bdc9127f88ab866fd2ac643", "787adb9d6ad9afb498a1580a7d8ad05f779c488a", "caf6e3f221ddc12ccabd1cd0ba149561db0090d4", "cb29688aaa4caa4d54df2976118fe99a839bb433", "0f0bee8bbc2b3e49baa703118041f99db9ef41a1", "16d99812d58b8af2df29cd337a74cd965b53da04", "42849b2697c36abdafa6aef64186b15055392046" ], "author": { "name": "Roland Dreier", "email": "roland@purestorage.com", "time": "Tue Nov 01 09:37:08 2011 -0700" }, "committer": { "name": "Roland Dreier", "email": "roland@purestorage.com", "time": "Tue Nov 01 09:37:08 2011 -0700" }, "message": "Merge branches \u0027amso1100\u0027, \u0027cma\u0027, \u0027cxgb3\u0027, \u0027cxgb4\u0027, \u0027fdr\u0027, \u0027ipath\u0027, \u0027ipoib\u0027, \u0027misc\u0027, \u0027mlx4\u0027, \u0027misc\u0027, \u0027nes\u0027, \u0027qib\u0027 and \u0027xrc\u0027 into for-next\n" }, { "commit": "97285b78174423e5576b2e06aa45f64df009da5b", "tree": "d6f44a503b98f385e22fc9243d1abaf98d68947d", "parents": [ "e36fb88a9a0fb8ac4b87c8ac709214a408de6d97" ], "author": { "name": "Marcel Apfelbaum", "email": "marcela@dev.mellanox.co.il", "time": "Mon Oct 24 11:02:34 2011 +0200" }, "committer": { "name": "Roland Dreier", "email": "roland@purestorage.com", "time": "Fri Oct 28 11:33:38 2011 -0700" }, "message": "mlx4_core: Add extended port capabilities support\n\nAn Extended Port Info packet is sent to each hw port during HCA init.\nIf it returns without error, we assume the port supports extended port\ncapabilities.\n\nSigned-off-by: Marcel Apfelbaum \u003cmarcela@dev.mellanox.co.il\u003e\nReviewed-by: Jack Morgenstein \u003cjackm@dev.mellanox.com\u003e\nSigned-off-by: Roland Dreier \u003croland@purestorage.com\u003e\n" } ], "next": "f3a9d1f25dfeadf22c775880633a587cc6778872" }