| Message ID | 1491979207-18686-5-git-send-email-niranjana.vishwanathapura@intel.com (mailing list archive) |
|---|---|
| State | Superseded |
| Headers | show |
On Tue, Apr 11, 2017 at 11:39:59PM -0700, Vishwanathapura, Niranjana wrote: > OPA VNIC netdev function supports Ethernet functionality over Omni-Path > fabric by encapsulating Ethernet packets inside Omni-Path packet header. > It allocates a rdma netdev device and interfaces with the network stack to > provide standard Ethernet network interfaces. It overrides HFI1 device's > netdev operations where it is required. > > Reviewed-by: Dennis Dalessandro <dennis.dalessandro@intel.com> > Reviewed-by: Ira Weiny <ira.weiny@intel.com> > Signed-off-by: Niranjana Vishwanathapura <niranjana.vishwanathapura@intel.com> > Signed-off-by: Sadanand Warrier <sadanand.warrier@intel.com> > Signed-off-by: Sudeep Dutt <sudeep.dutt@intel.com> > Signed-off-by: Andrzej Kacprowski <andrzej.kacprowski@intel.com> > --- > MAINTAINERS | 7 + > drivers/infiniband/Kconfig | 1 + > drivers/infiniband/ulp/Makefile | 1 + > drivers/infiniband/ulp/opa_vnic/Kconfig | 8 + > drivers/infiniband/ulp/opa_vnic/Makefile | 6 + > drivers/infiniband/ulp/opa_vnic/opa_vnic_encap.c | 239 +++++++++++++++++++++ > drivers/infiniband/ulp/opa_vnic/opa_vnic_encap.h | 62 ++++++ > drivers/infiniband/ulp/opa_vnic/opa_vnic_ethtool.c | 65 ++++++ > .../infiniband/ulp/opa_vnic/opa_vnic_internal.h | 186 ++++++++++++++++ > drivers/infiniband/ulp/opa_vnic/opa_vnic_netdev.c | 229 ++++++++++++++++++++ > 10 files changed, 804 insertions(+) > create mode 100644 drivers/infiniband/ulp/opa_vnic/Kconfig > create mode 100644 drivers/infiniband/ulp/opa_vnic/Makefile > create mode 100644 drivers/infiniband/ulp/opa_vnic/opa_vnic_encap.c > create mode 100644 drivers/infiniband/ulp/opa_vnic/opa_vnic_encap.h > create mode 100644 drivers/infiniband/ulp/opa_vnic/opa_vnic_ethtool.c > create mode 100644 drivers/infiniband/ulp/opa_vnic/opa_vnic_internal.h > create mode 100644 drivers/infiniband/ulp/opa_vnic/opa_vnic_netdev.c > > diff --git a/MAINTAINERS b/MAINTAINERS > index c776906..fc32256 100644 > --- a/MAINTAINERS > +++ b/MAINTAINERS > @@ -5843,6 +5843,13 @@ F: drivers/block/cciss* > F: include/linux/cciss_ioctl.h > F: include/uapi/linux/cciss_ioctl.h > > +OPA-VNIC DRIVER > +M: Dennis Dalessandro <dennis.dalessandro@intel.com> > +M: Niranjana Vishwanathapura <niranjana.vishwanathapura@intel.com> > +L: linux-rdma@vger.kernel.org > +S: Supported > +F: drivers/infiniband/ulp/opa_vnic > + > HFI1 DRIVER > M: Mike Marciniszyn <mike.marciniszyn@intel.com> > M: Dennis Dalessandro <dennis.dalessandro@intel.com> > diff --git a/drivers/infiniband/Kconfig b/drivers/infiniband/Kconfig > index 66f8602..234fe01 100644 > --- a/drivers/infiniband/Kconfig > +++ b/drivers/infiniband/Kconfig > @@ -85,6 +85,7 @@ source "drivers/infiniband/ulp/srpt/Kconfig" > source "drivers/infiniband/ulp/iser/Kconfig" > source "drivers/infiniband/ulp/isert/Kconfig" > > +source "drivers/infiniband/ulp/opa_vnic/Kconfig" > source "drivers/infiniband/sw/rdmavt/Kconfig" > source "drivers/infiniband/sw/rxe/Kconfig" > > diff --git a/drivers/infiniband/ulp/Makefile b/drivers/infiniband/ulp/Makefile > index f3c7dcf..c28af18 100644 > --- a/drivers/infiniband/ulp/Makefile > +++ b/drivers/infiniband/ulp/Makefile > @@ -3,3 +3,4 @@ obj-$(CONFIG_INFINIBAND_SRP) += srp/ > obj-$(CONFIG_INFINIBAND_SRPT) += srpt/ > obj-$(CONFIG_INFINIBAND_ISER) += iser/ > obj-$(CONFIG_INFINIBAND_ISERT) += isert/ > +obj-$(CONFIG_INFINIBAND_OPA_VNIC) += opa_vnic/ > diff --git a/drivers/infiniband/ulp/opa_vnic/Kconfig b/drivers/infiniband/ulp/opa_vnic/Kconfig > new file mode 100644 > index 0000000..48132ab > --- /dev/null > +++ b/drivers/infiniband/ulp/opa_vnic/Kconfig > @@ -0,0 +1,8 @@ > +config INFINIBAND_OPA_VNIC > + tristate "Intel OPA VNIC support" > + depends on X86_64 && INFINIBAND > + ---help--- > + This is Omni-Path (OPA) Virtual Network Interface Controller (VNIC) > + driver for Ethernet over Omni-Path feature. It implements the HW > + independent VNIC functionality. It interfaces with Linux stack for > + data path and IB MAD for the control path. > diff --git a/drivers/infiniband/ulp/opa_vnic/Makefile b/drivers/infiniband/ulp/opa_vnic/Makefile > new file mode 100644 > index 0000000..975c313 > --- /dev/null > +++ b/drivers/infiniband/ulp/opa_vnic/Makefile > @@ -0,0 +1,6 @@ > +# Makefile - Intel Omni-Path Virtual Network Controller driver > +# Copyright(c) 2017, Intel Corporation. > +# > +obj-$(CONFIG_INFINIBAND_OPA_VNIC) += opa_vnic.o > + > +opa_vnic-y := opa_vnic_netdev.o opa_vnic_encap.o opa_vnic_ethtool.o > diff --git a/drivers/infiniband/ulp/opa_vnic/opa_vnic_encap.c b/drivers/infiniband/ulp/opa_vnic/opa_vnic_encap.c > new file mode 100644 > index 0000000..c74d02a > --- /dev/null > +++ b/drivers/infiniband/ulp/opa_vnic/opa_vnic_encap.c > @@ -0,0 +1,239 @@ > +/* > + * Copyright(c) 2017 Intel Corporation. > + * > + * This file is provided under a dual BSD/GPLv2 license. When using or > + * redistributing this file, you may do so under either license. > + * > + * GPL LICENSE SUMMARY > + * > + * This program is free software; you can redistribute it and/or modify > + * it under the terms of version 2 of the GNU General Public License as > + * published by the Free Software Foundation. > + * > + * This program is distributed in the hope that it will be useful, but > + * WITHOUT ANY WARRANTY; without even the implied warranty of > + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU > + * General Public License for more details. > + * > + * BSD LICENSE > + * > + * Redistribution and use in source and binary forms, with or without > + * modification, are permitted provided that the following conditions > + * are met: > + * > + * - Redistributions of source code must retain the above copyright > + * notice, this list of conditions and the following disclaimer. > + * - Redistributions in binary form must reproduce the above copyright > + * notice, this list of conditions and the following disclaimer in > + * the documentation and/or other materials provided with the > + * distribution. > + * - Neither the name of Intel Corporation nor the names of its > + * contributors may be used to endorse or promote products derived > + * from this software without specific prior written permission. > + * > + * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS > + * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT > + * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR > + * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT > + * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, > + * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT > + * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, > + * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY > + * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT > + * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE > + * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. > + * > + */ > + > +/* > + * This file contains OPA VNIC encapsulation/decapsulation function. > + */ > + > +#include <linux/if_ether.h> > +#include <linux/if_vlan.h> > + > +#include "opa_vnic_internal.h" > + > +/* OPA 16B Header fields */ > +#define OPA_16B_LID_MASK 0xFFFFFull > +#define OPA_16B_SLID_HIGH_SHFT 8 > +#define OPA_16B_SLID_MASK 0xF00ull > +#define OPA_16B_DLID_MASK 0xF000ull > +#define OPA_16B_DLID_HIGH_SHFT 12 > +#define OPA_16B_LEN_SHFT 20 > +#define OPA_16B_SC_SHFT 20 > +#define OPA_16B_RC_SHFT 25 > +#define OPA_16B_PKEY_SHFT 16 > + > +#define OPA_VNIC_L4_HDR_SHFT 16 > + > +/* L2+L4 hdr len is 20 bytes (5 quad words) */ > +#define OPA_VNIC_HDR_QW_LEN 5 > + > +static inline void opa_vnic_make_header(u8 *hdr, u32 slid, u32 dlid, u16 len, > + u16 pkey, u16 entropy, u8 sc, u8 rc, > + u8 l4_type, u16 l4_hdr) > +{ > + /* h[1]: LT=1, 16B L2=10 */ > + u32 h[OPA_VNIC_HDR_QW_LEN] = {0, 0xc0000000, 0, 0, 0}; > + > + h[2] = l4_type; > + h[3] = entropy; > + h[4] = l4_hdr << OPA_VNIC_L4_HDR_SHFT; > + > + /* Extract and set 4 upper bits and 20 lower bits of the lids */ > + h[0] |= (slid & OPA_16B_LID_MASK); > + h[2] |= ((slid >> (20 - OPA_16B_SLID_HIGH_SHFT)) & OPA_16B_SLID_MASK); > + > + h[1] |= (dlid & OPA_16B_LID_MASK); > + h[2] |= ((dlid >> (20 - OPA_16B_DLID_HIGH_SHFT)) & OPA_16B_DLID_MASK); > + > + h[0] |= (len << OPA_16B_LEN_SHFT); > + h[1] |= (rc << OPA_16B_RC_SHFT); > + h[1] |= (sc << OPA_16B_SC_SHFT); > + h[2] |= ((u32)pkey << OPA_16B_PKEY_SHFT); > + > + memcpy(hdr, h, OPA_VNIC_HDR_LEN); > +} > + > +/* opa_vnic_get_dlid - find and return the DLID */ > +static uint32_t opa_vnic_get_dlid(struct opa_vnic_adapter *adapter, > + struct sk_buff *skb, u8 def_port) > +{ > + struct __opa_veswport_info *info = &adapter->info; > + struct ethhdr *mac_hdr = (struct ethhdr *)skb_mac_header(skb); > + u32 dlid; > + > + if (is_multicast_ether_addr(mac_hdr->h_dest)) { > + dlid = info->vesw.u_mcast_dlid; > + } else { > + if (is_local_ether_addr(mac_hdr->h_dest)) { > + dlid = ((uint32_t)mac_hdr->h_dest[5] << 16) | > + ((uint32_t)mac_hdr->h_dest[4] << 8) | > + mac_hdr->h_dest[3]; > + if (unlikely(!dlid)) > + v_warn("Null dlid in MAC address\n"); > + } else if (def_port != OPA_VNIC_INVALID_PORT) { > + dlid = info->vesw.u_ucast_dlid[def_port]; > + } > + } > + > + return dlid; > +} > + > +/* opa_vnic_get_sc - return the service class */ > +static u8 opa_vnic_get_sc(struct __opa_veswport_info *info, > + struct sk_buff *skb) > +{ > + struct ethhdr *mac_hdr = (struct ethhdr *)skb_mac_header(skb); > + u16 vlan_tci; > + u8 sc; > + > + if (!__vlan_get_tag(skb, &vlan_tci)) { > + u8 pcp = OPA_VNIC_VLAN_PCP(vlan_tci); > + > + if (is_multicast_ether_addr(mac_hdr->h_dest)) > + sc = info->vport.pcp_to_sc_mc[pcp]; > + else > + sc = info->vport.pcp_to_sc_uc[pcp]; > + } else { > + if (is_multicast_ether_addr(mac_hdr->h_dest)) > + sc = info->vport.non_vlan_sc_mc; > + else > + sc = info->vport.non_vlan_sc_uc; > + } > + > + return sc; > +} > + > +u8 opa_vnic_get_vl(struct opa_vnic_adapter *adapter, struct sk_buff *skb) > +{ > + struct ethhdr *mac_hdr = (struct ethhdr *)skb_mac_header(skb); > + struct __opa_veswport_info *info = &adapter->info; > + u8 vl; > + > + if (skb_vlan_tag_present(skb)) { > + u8 pcp = skb_vlan_tag_get(skb) >> VLAN_PRIO_SHIFT; > + > + if (is_multicast_ether_addr(mac_hdr->h_dest)) > + vl = info->vport.pcp_to_vl_mc[pcp]; > + else > + vl = info->vport.pcp_to_vl_uc[pcp]; > + } else { > + if (is_multicast_ether_addr(mac_hdr->h_dest)) > + vl = info->vport.non_vlan_vl_mc; > + else > + vl = info->vport.non_vlan_vl_uc; > + } > + > + return vl; > +} > + > +/* opa_vnic_calc_entropy - calculate the packet entropy */ > +u8 opa_vnic_calc_entropy(struct opa_vnic_adapter *adapter, struct sk_buff *skb) > +{ > + u16 hash16; > + > + /* > + * Get flow based 16-bit hash and then XOR the upper and lower bytes > + * to get the entropy. > + * __skb_tx_hash limits qcount to 16 bits. Hence, get 15-bit hash. > + */ > + hash16 = __skb_tx_hash(adapter->netdev, skb, BIT(15)); > + return (u8)((hash16 >> 8) ^ (hash16 & 0xff)); > +} > + > +/* opa_vnic_get_def_port - get default port based on entropy */ > +static inline u8 opa_vnic_get_def_port(struct opa_vnic_adapter *adapter, > + u8 entropy) > +{ > + u8 flow_id; > + > + /* Add the upper and lower 4-bits of entropy to get the flow id */ > + flow_id = ((entropy & 0xf) + (entropy >> 4)); > + return adapter->flow_tbl[flow_id & (OPA_VNIC_FLOW_TBL_SIZE - 1)]; > +} > + > +/* Calculate packet length including OPA header, crc and padding */ > +static inline int opa_vnic_wire_length(struct sk_buff *skb) > +{ > + u32 pad_len; > + > + /* padding for 8 bytes size alignment */ > + pad_len = -(skb->len + OPA_VNIC_ICRC_TAIL_LEN) & 0x7; > + pad_len += OPA_VNIC_ICRC_TAIL_LEN; > + > + return (skb->len + pad_len) >> 3; > +} > + > +/* opa_vnic_encap_skb - encapsulate skb packet with OPA header and meta data */ > +void opa_vnic_encap_skb(struct opa_vnic_adapter *adapter, struct sk_buff *skb) > +{ > + struct __opa_veswport_info *info = &adapter->info; > + struct opa_vnic_skb_mdata *mdata; > + u8 def_port, sc, entropy, *hdr; > + u16 len, l4_hdr; > + u32 dlid; > + > + hdr = skb_push(skb, OPA_VNIC_HDR_LEN); > + > + entropy = opa_vnic_calc_entropy(adapter, skb); > + def_port = opa_vnic_get_def_port(adapter, entropy); > + len = opa_vnic_wire_length(skb); > + dlid = opa_vnic_get_dlid(adapter, skb, def_port); > + sc = opa_vnic_get_sc(info, skb); > + l4_hdr = info->vesw.vesw_id; > + > + mdata = (struct opa_vnic_skb_mdata *)skb_push(skb, sizeof(*mdata)); > + mdata->vl = opa_vnic_get_vl(adapter, skb); > + mdata->entropy = entropy; > + mdata->flags = 0; > + if (unlikely(!dlid)) { > + mdata->flags = OPA_VNIC_SKB_MDATA_ENCAP_ERR; > + return; > + } > + > + opa_vnic_make_header(hdr, info->vport.encap_slid, dlid, len, > + info->vesw.pkey, entropy, sc, 0, > + OPA_VNIC_L4_ETHR, l4_hdr); > +} > diff --git a/drivers/infiniband/ulp/opa_vnic/opa_vnic_encap.h b/drivers/infiniband/ulp/opa_vnic/opa_vnic_encap.h > new file mode 100644 > index 0000000..176fca9 > --- /dev/null > +++ b/drivers/infiniband/ulp/opa_vnic/opa_vnic_encap.h > @@ -0,0 +1,62 @@ > +#ifndef _OPA_VNIC_ENCAP_H > +#define _OPA_VNIC_ENCAP_H > +/* > + * Copyright(c) 2017 Intel Corporation. > + * > + * This file is provided under a dual BSD/GPLv2 license. When using or > + * redistributing this file, you may do so under either license. > + * > + * GPL LICENSE SUMMARY > + * > + * This program is free software; you can redistribute it and/or modify > + * it under the terms of version 2 of the GNU General Public License as > + * published by the Free Software Foundation. > + * > + * This program is distributed in the hope that it will be useful, but > + * WITHOUT ANY WARRANTY; without even the implied warranty of > + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU > + * General Public License for more details. > + * > + * BSD LICENSE > + * > + * Redistribution and use in source and binary forms, with or without > + * modification, are permitted provided that the following conditions > + * are met: > + * > + * - Redistributions of source code must retain the above copyright > + * notice, this list of conditions and the following disclaimer. > + * - Redistributions in binary form must reproduce the above copyright > + * notice, this list of conditions and the following disclaimer in > + * the documentation and/or other materials provided with the > + * distribution. > + * - Neither the name of Intel Corporation nor the names of its > + * contributors may be used to endorse or promote products derived > + * from this software without specific prior written permission. > + * > + * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS > + * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT > + * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR > + * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT > + * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, > + * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT > + * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, > + * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY > + * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT > + * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE > + * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. > + * > + */ > + > +/* > + * This file contains all OPA VNIC declaration required for encapsulation > + * and decapsulation of Ethernet packets > + */ > + > +/* VNIC configured and operational state values */ > +#define OPA_VNIC_STATE_DROP_ALL 0x1 > +#define OPA_VNIC_STATE_FORWARDING 0x3 > + > +#define OPA_VESW_MAX_NUM_DEF_PORT 16 > +#define OPA_VNIC_MAX_NUM_PCP 8 > + > +#endif /* _OPA_VNIC_ENCAP_H */ > diff --git a/drivers/infiniband/ulp/opa_vnic/opa_vnic_ethtool.c b/drivers/infiniband/ulp/opa_vnic/opa_vnic_ethtool.c > new file mode 100644 > index 0000000..b74f6ad > --- /dev/null > +++ b/drivers/infiniband/ulp/opa_vnic/opa_vnic_ethtool.c > @@ -0,0 +1,65 @@ > +/* > + * Copyright(c) 2017 Intel Corporation. > + * > + * This file is provided under a dual BSD/GPLv2 license. When using or > + * redistributing this file, you may do so under either license. > + * > + * GPL LICENSE SUMMARY > + * > + * This program is free software; you can redistribute it and/or modify > + * it under the terms of version 2 of the GNU General Public License as > + * published by the Free Software Foundation. > + * > + * This program is distributed in the hope that it will be useful, but > + * WITHOUT ANY WARRANTY; without even the implied warranty of > + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU > + * General Public License for more details. > + * > + * BSD LICENSE > + * > + * Redistribution and use in source and binary forms, with or without > + * modification, are permitted provided that the following conditions > + * are met: > + * > + * - Redistributions of source code must retain the above copyright > + * notice, this list of conditions and the following disclaimer. > + * - Redistributions in binary form must reproduce the above copyright > + * notice, this list of conditions and the following disclaimer in > + * the documentation and/or other materials provided with the > + * distribution. > + * - Neither the name of Intel Corporation nor the names of its > + * contributors may be used to endorse or promote products derived > + * from this software without specific prior written permission. > + * > + * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS > + * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT > + * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR > + * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT > + * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, > + * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT > + * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, > + * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY > + * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT > + * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE > + * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. > + * > + */ > + > +/* > + * This file contains OPA VNIC ethtool functions > + */ > + > +#include <linux/ethtool.h> > + > +#include "opa_vnic_internal.h" > + > +/* ethtool ops */ > +static const struct ethtool_ops opa_vnic_ethtool_ops = { > + .get_link = ethtool_op_get_link, > +}; > + > +/* opa_vnic_set_ethtool_ops - set ethtool ops */ > +void opa_vnic_set_ethtool_ops(struct net_device *netdev) > +{ > + netdev->ethtool_ops = &opa_vnic_ethtool_ops; > +} > diff --git a/drivers/infiniband/ulp/opa_vnic/opa_vnic_internal.h b/drivers/infiniband/ulp/opa_vnic/opa_vnic_internal.h > new file mode 100644 > index 0000000..83ffa91 > --- /dev/null > +++ b/drivers/infiniband/ulp/opa_vnic/opa_vnic_internal.h > @@ -0,0 +1,186 @@ > +#ifndef _OPA_VNIC_INTERNAL_H > +#define _OPA_VNIC_INTERNAL_H > +/* > + * Copyright(c) 2017 Intel Corporation. > + * > + * This file is provided under a dual BSD/GPLv2 license. When using or > + * redistributing this file, you may do so under either license. > + * > + * GPL LICENSE SUMMARY > + * > + * This program is free software; you can redistribute it and/or modify > + * it under the terms of version 2 of the GNU General Public License as > + * published by the Free Software Foundation. > + * > + * This program is distributed in the hope that it will be useful, but > + * WITHOUT ANY WARRANTY; without even the implied warranty of > + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU > + * General Public License for more details. > + * > + * BSD LICENSE > + * > + * Redistribution and use in source and binary forms, with or without > + * modification, are permitted provided that the following conditions > + * are met: > + * > + * - Redistributions of source code must retain the above copyright > + * notice, this list of conditions and the following disclaimer. > + * - Redistributions in binary form must reproduce the above copyright > + * notice, this list of conditions and the following disclaimer in > + * the documentation and/or other materials provided with the > + * distribution. > + * - Neither the name of Intel Corporation nor the names of its > + * contributors may be used to endorse or promote products derived > + * from this software without specific prior written permission. > + * > + * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS > + * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT > + * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR > + * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT > + * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, > + * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT > + * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, > + * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY > + * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT > + * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE > + * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. > + * > + */ > + > +/* > + * This file contains OPA VNIC driver internal declarations > + */ > + > +#include <linux/bitops.h> > +#include <linux/etherdevice.h> > +#include <linux/hashtable.h> > +#include <linux/sizes.h> > +#include <rdma/opa_vnic.h> > + > +#include "opa_vnic_encap.h" > + > +#define OPA_VNIC_VLAN_PCP(vlan_tci) \ > + (((vlan_tci) & VLAN_PRIO_MASK) >> VLAN_PRIO_SHIFT) > + > +/* Flow to default port redirection table size */ > +#define OPA_VNIC_FLOW_TBL_SIZE 32 > + > +/* Invalid port number */ > +#define OPA_VNIC_INVALID_PORT 0xff > + > +struct opa_vnic_adapter; > + > +/** > + * struct __opa_vesw_info - OPA vnic virtual switch info > + */ > +struct __opa_vesw_info { > + u16 fabric_id; > + u16 vesw_id; > + > + u8 rsvd0[6]; > + u16 def_port_mask; > + > + u8 rsvd1[2]; > + u16 pkey; > + > + u8 rsvd2[4]; > + u32 u_mcast_dlid; > + u32 u_ucast_dlid[OPA_VESW_MAX_NUM_DEF_PORT]; > + > + u8 rsvd3[44]; > + u16 eth_mtu[OPA_VNIC_MAX_NUM_PCP]; > + u16 eth_mtu_non_vlan; > + u8 rsvd4[2]; > +} __packed; > + > +/** > + * struct __opa_per_veswport_info - OPA vnic per port info > + */ > +struct __opa_per_veswport_info { > + u32 port_num; > + > + u8 eth_link_status; > + u8 rsvd0[3]; > + > + u8 base_mac_addr[ETH_ALEN]; > + u8 config_state; > + u8 oper_state; > + > + u16 max_mac_tbl_ent; > + u16 max_smac_ent; > + u32 mac_tbl_digest; > + u8 rsvd1[4]; > + > + u32 encap_slid; > + > + u8 pcp_to_sc_uc[OPA_VNIC_MAX_NUM_PCP]; > + u8 pcp_to_vl_uc[OPA_VNIC_MAX_NUM_PCP]; > + u8 pcp_to_sc_mc[OPA_VNIC_MAX_NUM_PCP]; > + u8 pcp_to_vl_mc[OPA_VNIC_MAX_NUM_PCP]; > + > + u8 non_vlan_sc_uc; > + u8 non_vlan_vl_uc; > + u8 non_vlan_sc_mc; > + u8 non_vlan_vl_mc; > + > + u8 rsvd2[48]; > + > + u16 uc_macs_gen_count; > + u16 mc_macs_gen_count; > + > + u8 rsvd3[8]; > +} __packed; > + > +/** > + * struct __opa_veswport_info - OPA vnic port info > + */ > +struct __opa_veswport_info { > + struct __opa_vesw_info vesw; > + struct __opa_per_veswport_info vport; > +}; > + > +/** > + * struct opa_vnic_adapter - OPA VNIC netdev private data structure > + * @netdev: pointer to associated netdev > + * @ibdev: ib device > + * @rn_ops: rdma netdev's net_device_ops > + * @port_num: OPA port number > + * @vport_num: vesw port number > + * @lock: adapter lock > + * @info: virtual ethernet switch port information > + * @flow_tbl: flow to default port redirection table > + */ > +struct opa_vnic_adapter { > + struct net_device *netdev; > + struct ib_device *ibdev; > + const struct net_device_ops *rn_ops; > + > + u8 port_num; > + u8 vport_num; > + > + /* Lock used around concurrent updates to netdev */ > + struct mutex lock; > + > + struct __opa_veswport_info info; > + > + u8 flow_tbl[OPA_VNIC_FLOW_TBL_SIZE]; > +}; > + > +#define v_dbg(format, arg...) \ > + netdev_dbg(adapter->netdev, format, ## arg) > +#define v_err(format, arg...) \ > + netdev_err(adapter->netdev, format, ## arg) > +#define v_info(format, arg...) \ > + netdev_info(adapter->netdev, format, ## arg) > +#define v_warn(format, arg...) \ > + netdev_warn(adapter->netdev, format, ## arg) > + IMHO, these wrappers are redundant. > +struct opa_vnic_adapter *opa_vnic_add_netdev(struct ib_device *ibdev, > + u8 port_num, u8 vport_num); > +void opa_vnic_rem_netdev(struct opa_vnic_adapter *adapter); > +void opa_vnic_encap_skb(struct opa_vnic_adapter *adapter, struct sk_buff *skb); > +u8 opa_vnic_get_vl(struct opa_vnic_adapter *adapter, struct sk_buff *skb); > +u8 opa_vnic_calc_entropy(struct opa_vnic_adapter *adapter, struct sk_buff *skb); > +void opa_vnic_set_ethtool_ops(struct net_device *netdev); > + > +#endif /* _OPA_VNIC_INTERNAL_H */ > diff --git a/drivers/infiniband/ulp/opa_vnic/opa_vnic_netdev.c b/drivers/infiniband/ulp/opa_vnic/opa_vnic_netdev.c > new file mode 100644 > index 0000000..17e8d36 > --- /dev/null > +++ b/drivers/infiniband/ulp/opa_vnic/opa_vnic_netdev.c > @@ -0,0 +1,229 @@ > +/* > + * Copyright(c) 2017 Intel Corporation. > + * > + * This file is provided under a dual BSD/GPLv2 license. When using or > + * redistributing this file, you may do so under either license. > + * > + * GPL LICENSE SUMMARY > + * > + * This program is free software; you can redistribute it and/or modify > + * it under the terms of version 2 of the GNU General Public License as > + * published by the Free Software Foundation. > + * > + * This program is distributed in the hope that it will be useful, but > + * WITHOUT ANY WARRANTY; without even the implied warranty of > + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU > + * General Public License for more details. > + * > + * BSD LICENSE > + * > + * Redistribution and use in source and binary forms, with or without > + * modification, are permitted provided that the following conditions > + * are met: > + * > + * - Redistributions of source code must retain the above copyright > + * notice, this list of conditions and the following disclaimer. > + * - Redistributions in binary form must reproduce the above copyright > + * notice, this list of conditions and the following disclaimer in > + * the documentation and/or other materials provided with the > + * distribution. > + * - Neither the name of Intel Corporation nor the names of its > + * contributors may be used to endorse or promote products derived > + * from this software without specific prior written permission. > + * > + * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS > + * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT > + * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR > + * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT > + * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, > + * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT > + * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, > + * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY > + * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT > + * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE > + * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. > + * > + */ > + > +/* > + * This file contains OPA Virtual Network Interface Controller (VNIC) driver > + * netdev functionality. > + */ > + > +#include <linux/module.h> > +#include <linux/if_vlan.h> > + > +#include "opa_vnic_internal.h" > + > +#define OPA_TX_TIMEOUT_MS 1000 > + > +#define OPA_VNIC_SKB_HEADROOM \ > + ALIGN((OPA_VNIC_HDR_LEN + OPA_VNIC_SKB_MDATA_LEN), 8) > + > +/* opa_netdev_start_xmit - transmit function */ > +static netdev_tx_t opa_netdev_start_xmit(struct sk_buff *skb, > + struct net_device *netdev) > +{ > + struct opa_vnic_adapter *adapter = opa_vnic_priv(netdev); > + > + v_dbg("xmit: queue %d skb len %d\n", skb->queue_mapping, skb->len); > + /* pad to ensure mininum ethernet packet length */ > + if (unlikely(skb->len < ETH_ZLEN)) { > + if (skb_padto(skb, ETH_ZLEN)) > + return NETDEV_TX_OK; > + > + skb_put(skb, ETH_ZLEN - skb->len); > + } > + > + opa_vnic_encap_skb(adapter, skb); > + return adapter->rn_ops->ndo_start_xmit(skb, netdev); > +} > + > +static u16 opa_vnic_select_queue(struct net_device *netdev, struct sk_buff *skb, > + void *accel_priv, > + select_queue_fallback_t fallback) > +{ > + struct opa_vnic_adapter *adapter = opa_vnic_priv(netdev); > + struct opa_vnic_skb_mdata *mdata; > + int rc; > + > + /* pass entropy and vl as metadata in skb */ > + mdata = (struct opa_vnic_skb_mdata *)skb_push(skb, sizeof(*mdata)); > + mdata->entropy = opa_vnic_calc_entropy(adapter, skb); > + mdata->vl = opa_vnic_get_vl(adapter, skb); > + rc = adapter->rn_ops->ndo_select_queue(netdev, skb, > + accel_priv, fallback); > + skb_pull(skb, sizeof(*mdata)); > + return rc; > +} > + > +/* opa_vnic_set_mac_addr - change mac address */ > +static int opa_vnic_set_mac_addr(struct net_device *netdev, void *addr) > +{ > + struct opa_vnic_adapter *adapter = opa_vnic_priv(netdev); > + struct sockaddr *sa = addr; > + int rc; > + > + if (!memcmp(netdev->dev_addr, sa->sa_data, ETH_ALEN)) > + return 0; > + > + mutex_lock(&adapter->lock); > + rc = eth_mac_addr(netdev, addr); > + mutex_unlock(&adapter->lock); > + > + return rc; > +} > + > +/* opa_netdev_open - activate network interface */ > +static int opa_netdev_open(struct net_device *netdev) > +{ > + struct opa_vnic_adapter *adapter = opa_vnic_priv(netdev); > + int rc; > + > + rc = adapter->rn_ops->ndo_open(adapter->netdev); > + if (rc) { > + v_dbg("open failed %d\n", rc); > + return rc; > + } > + > + v_info("opened\n"); All these v_info are achieved by tracepoints (function tracer). > + return 0; > +} > + > +/* opa_netdev_close - disable network interface */ > +static int opa_netdev_close(struct net_device *netdev) > +{ > + struct opa_vnic_adapter *adapter = opa_vnic_priv(netdev); > + int rc; > + > + rc = adapter->rn_ops->ndo_stop(adapter->netdev); > + if (rc) { > + v_dbg("close failed %d\n", rc); > + return rc; > + } > + > + v_info("closed\n"); > + return 0; > +} > + > +/* netdev ops */ > +static const struct net_device_ops opa_netdev_ops = { > + .ndo_open = opa_netdev_open, > + .ndo_stop = opa_netdev_close, > + .ndo_start_xmit = opa_netdev_start_xmit, > + .ndo_select_queue = opa_vnic_select_queue, > + .ndo_set_mac_address = opa_vnic_set_mac_addr, > +}; > + > +/* opa_vnic_add_netdev - create vnic netdev interface */ > +struct opa_vnic_adapter *opa_vnic_add_netdev(struct ib_device *ibdev, > + u8 port_num, u8 vport_num) > +{ > + struct opa_vnic_adapter *adapter; > + struct net_device *netdev; > + struct rdma_netdev *rn; > + int rc; > + > + netdev = ibdev->alloc_rdma_netdev(ibdev, port_num, > + RDMA_NETDEV_OPA_VNIC, > + "veth%d", NET_NAME_UNKNOWN, > + ether_setup); > + if (!netdev) > + return ERR_PTR(-ENOMEM); > + else if (IS_ERR(netdev)) > + return ERR_CAST(netdev); > + Erez and Jason came to this code for IPoIB, it is better to have same error handling for all alloc_rdma_netdev callers. + if (hca->alloc_rdma_netdev) { + dev = hca->alloc_rdma_netdev(hca, port, + RDMA_NETDEV_IPOIB, name, + NET_NAME_UNKNOWN, + ipoib_setup_common); + if (IS_ERR_OR_NULL(dev) && PTR_ERR(dev) != -EOPNOTSUPP) + return NULL; + } > + adapter = kzalloc(sizeof(*adapter), GFP_KERNEL); > + if (!adapter) { > + rc = -ENOMEM; > + goto adapter_err; > + } > + > + rn = netdev_priv(netdev); > + rn->clnt_priv = adapter; > + rn->hca = ibdev; > + rn->port_num = port_num; > + adapter->netdev = netdev; > + adapter->ibdev = ibdev; > + adapter->port_num = port_num; > + adapter->vport_num = vport_num; > + adapter->rn_ops = netdev->netdev_ops; > + > + netdev->netdev_ops = &opa_netdev_ops; > + netdev->priv_flags |= IFF_LIVE_ADDR_CHANGE; > + netdev->hard_header_len += OPA_VNIC_SKB_HEADROOM; > + mutex_init(&adapter->lock); > + > + SET_NETDEV_DEV(netdev, ibdev->dev.parent); > + > + opa_vnic_set_ethtool_ops(netdev); > + rc = register_netdev(netdev); > + if (rc) > + goto netdev_err; > + > + netif_carrier_off(netdev); > + netif_dormant_on(netdev); > + v_info("initialized\n"); > + > + return adapter; > +netdev_err: > + mutex_destroy(&adapter->lock); > + kfree(adapter); > +adapter_err: > + ibdev->free_rdma_netdev(netdev); > + > + return ERR_PTR(rc); > +} > + > +/* opa_vnic_rem_netdev - remove vnic netdev interface */ > +void opa_vnic_rem_netdev(struct opa_vnic_adapter *adapter) > +{ > + struct net_device *netdev = adapter->netdev; > + struct ib_device *ibdev = adapter->ibdev; > + > + v_info("removing\n"); > + unregister_netdev(netdev); > + mutex_destroy(&adapter->lock); > + kfree(adapter); > + ibdev->free_rdma_netdev(netdev); > +} > -- > 1.8.3.1 > > -- > To unsubscribe from this list: send the line "unsubscribe linux-rdma" in > the body of a message to majordomo@vger.kernel.org > More majordomo info at http://vger.kernel.org/majordomo-info.html
On Wed, Apr 12, 2017 at 10:08:30AM +0300, Leon Romanovsky wrote: >> +#define v_dbg(format, arg...) \ >> + netdev_dbg(adapter->netdev, format, ## arg) >> +#define v_err(format, arg...) \ >> + netdev_err(adapter->netdev, format, ## arg) >> +#define v_info(format, arg...) \ >> + netdev_info(adapter->netdev, format, ## arg) >> +#define v_warn(format, arg...) \ >> + netdev_warn(adapter->netdev, format, ## arg) >> + > >IMHO, these wrappers are redundant. > Using same constructs as some Intel standard ethernet drivers. >> +/* opa_netdev_open - activate network interface */ >> +static int opa_netdev_open(struct net_device *netdev) >> +{ >> + struct opa_vnic_adapter *adapter = opa_vnic_priv(netdev); >> + int rc; >> + >> + rc = adapter->rn_ops->ndo_open(adapter->netdev); >> + if (rc) { >> + v_dbg("open failed %d\n", rc); >> + return rc; >> + } >> + >> + v_info("opened\n"); > >All these v_info are achieved by tracepoints (function tracer). > Some of these messages are useful for analysing reported logs. Let me change these opened/closed messges to debug level. >> + >> + netdev = ibdev->alloc_rdma_netdev(ibdev, port_num, >> + RDMA_NETDEV_OPA_VNIC, >> + "veth%d", NET_NAME_UNKNOWN, >> + ether_setup); >> + if (!netdev) >> + return ERR_PTR(-ENOMEM); >> + else if (IS_ERR(netdev)) >> + return ERR_CAST(netdev); >> + > >Erez and Jason came to this code for IPoIB, it is better to have same >error handling for all alloc_rdma_netdev callers. >+ if (hca->alloc_rdma_netdev) { >+ dev = hca->alloc_rdma_netdev(hca, port, >+ RDMA_NETDEV_IPOIB, name, >+ NET_NAME_UNKNOWN, >+ ipoib_setup_common); >+ if (IS_ERR_OR_NULL(dev) && PTR_ERR(dev) != -EOPNOTSUPP) >+ return NULL; >+ } > > IPoIB handles EOPNOTSUPP differently (by assigning default operations). It is not applicable to OPA VNIC, hence it just returns the error code. I just noticed that IPoIB is using EOPNOTSUPP, however OPA VNIC is using ENOTSUPP. EOPNOTSUPP seesm to be widely used, so I will change OPA VNIC to use the same. Will also document this requirement in ib_verbs.h where this function is defined. Niranjana -- To unsubscribe from this list: send the line "unsubscribe linux-rdma" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
On Wed, Apr 12, 2017 at 11:31:37AM -0700, Vishwanathapura, Niranjana wrote: > On Wed, Apr 12, 2017 at 10:08:30AM +0300, Leon Romanovsky wrote: > > > +#define v_dbg(format, arg...) \ > > > + netdev_dbg(adapter->netdev, format, ## arg) > > > +#define v_err(format, arg...) \ > > > + netdev_err(adapter->netdev, format, ## arg) > > > +#define v_info(format, arg...) \ > > > + netdev_info(adapter->netdev, format, ## arg) > > > +#define v_warn(format, arg...) \ > > > + netdev_warn(adapter->netdev, format, ## arg) > > > + > > > > IMHO, these wrappers are redundant. > > > > Using same constructs as some Intel standard ethernet drivers. I'll leave this decision to Doug. IMHO open coded variant is preferable. > > > > +/* opa_netdev_open - activate network interface */ > > > +static int opa_netdev_open(struct net_device *netdev) > > > +{ > > > + struct opa_vnic_adapter *adapter = opa_vnic_priv(netdev); > > > + int rc; > > > + > > > + rc = adapter->rn_ops->ndo_open(adapter->netdev); > > > + if (rc) { > > > + v_dbg("open failed %d\n", rc); > > > + return rc; > > > + } > > > + > > > + v_info("opened\n"); > > > > All these v_info are achieved by tracepoints (function tracer). > > > > Some of these messages are useful for analysing reported logs. > Let me change these opened/closed messges to debug level. I agree with you regarding error messages, I disagree regarding flow/info messages. > > > > + > > > + netdev = ibdev->alloc_rdma_netdev(ibdev, port_num, > > > + RDMA_NETDEV_OPA_VNIC, > > > + "veth%d", NET_NAME_UNKNOWN, > > > + ether_setup); > > > + if (!netdev) > > > + return ERR_PTR(-ENOMEM); > > > + else if (IS_ERR(netdev)) > > > + return ERR_CAST(netdev); > > > + > > > > > Erez and Jason came to this code for IPoIB, it is better to have same > > error handling for all alloc_rdma_netdev callers. > > + if (hca->alloc_rdma_netdev) { > > + dev = hca->alloc_rdma_netdev(hca, port, > > + RDMA_NETDEV_IPOIB, name, > > + NET_NAME_UNKNOWN, > > + ipoib_setup_common); > > + if (IS_ERR_OR_NULL(dev) && PTR_ERR(dev) != -EOPNOTSUPP) > > + return NULL; > > + } > > > > > > IPoIB handles EOPNOTSUPP differently (by assigning default operations). > It is not applicable to OPA VNIC, hence it just returns the error code. > > I just noticed that IPoIB is using EOPNOTSUPP, however OPA VNIC is using > ENOTSUPP. EOPNOTSUPP seesm to be widely used, so I will change OPA VNIC to > use the same. Will also document this requirement in ib_verbs.h where this > function is defined. The ENOTSUPP is for NFS and it can't be returned to user space. The proper error is EOPNOTSUPP. Thanks > > Niranjana > > -- > To unsubscribe from this list: send the line "unsubscribe linux-rdma" in > the body of a message to majordomo@vger.kernel.org > More majordomo info at http://vger.kernel.org/majordomo-info.html
diff --git a/MAINTAINERS b/MAINTAINERS index c776906..fc32256 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -5843,6 +5843,13 @@ F: drivers/block/cciss* F: include/linux/cciss_ioctl.h F: include/uapi/linux/cciss_ioctl.h +OPA-VNIC DRIVER +M: Dennis Dalessandro <dennis.dalessandro@intel.com> +M: Niranjana Vishwanathapura <niranjana.vishwanathapura@intel.com> +L: linux-rdma@vger.kernel.org +S: Supported +F: drivers/infiniband/ulp/opa_vnic + HFI1 DRIVER M: Mike Marciniszyn <mike.marciniszyn@intel.com> M: Dennis Dalessandro <dennis.dalessandro@intel.com> diff --git a/drivers/infiniband/Kconfig b/drivers/infiniband/Kconfig index 66f8602..234fe01 100644 --- a/drivers/infiniband/Kconfig +++ b/drivers/infiniband/Kconfig @@ -85,6 +85,7 @@ source "drivers/infiniband/ulp/srpt/Kconfig" source "drivers/infiniband/ulp/iser/Kconfig" source "drivers/infiniband/ulp/isert/Kconfig" +source "drivers/infiniband/ulp/opa_vnic/Kconfig" source "drivers/infiniband/sw/rdmavt/Kconfig" source "drivers/infiniband/sw/rxe/Kconfig" diff --git a/drivers/infiniband/ulp/Makefile b/drivers/infiniband/ulp/Makefile index f3c7dcf..c28af18 100644 --- a/drivers/infiniband/ulp/Makefile +++ b/drivers/infiniband/ulp/Makefile @@ -3,3 +3,4 @@ obj-$(CONFIG_INFINIBAND_SRP) += srp/ obj-$(CONFIG_INFINIBAND_SRPT) += srpt/ obj-$(CONFIG_INFINIBAND_ISER) += iser/ obj-$(CONFIG_INFINIBAND_ISERT) += isert/ +obj-$(CONFIG_INFINIBAND_OPA_VNIC) += opa_vnic/ diff --git a/drivers/infiniband/ulp/opa_vnic/Kconfig b/drivers/infiniband/ulp/opa_vnic/Kconfig new file mode 100644 index 0000000..48132ab --- /dev/null +++ b/drivers/infiniband/ulp/opa_vnic/Kconfig @@ -0,0 +1,8 @@ +config INFINIBAND_OPA_VNIC + tristate "Intel OPA VNIC support" + depends on X86_64 && INFINIBAND + ---help--- + This is Omni-Path (OPA) Virtual Network Interface Controller (VNIC) + driver for Ethernet over Omni-Path feature. It implements the HW + independent VNIC functionality. It interfaces with Linux stack for + data path and IB MAD for the control path. diff --git a/drivers/infiniband/ulp/opa_vnic/Makefile b/drivers/infiniband/ulp/opa_vnic/Makefile new file mode 100644 index 0000000..975c313 --- /dev/null +++ b/drivers/infiniband/ulp/opa_vnic/Makefile @@ -0,0 +1,6 @@ +# Makefile - Intel Omni-Path Virtual Network Controller driver +# Copyright(c) 2017, Intel Corporation. +# +obj-$(CONFIG_INFINIBAND_OPA_VNIC) += opa_vnic.o + +opa_vnic-y := opa_vnic_netdev.o opa_vnic_encap.o opa_vnic_ethtool.o diff --git a/drivers/infiniband/ulp/opa_vnic/opa_vnic_encap.c b/drivers/infiniband/ulp/opa_vnic/opa_vnic_encap.c new file mode 100644 index 0000000..c74d02a --- /dev/null +++ b/drivers/infiniband/ulp/opa_vnic/opa_vnic_encap.c @@ -0,0 +1,239 @@ +/* + * Copyright(c) 2017 Intel Corporation. + * + * This file is provided under a dual BSD/GPLv2 license. When using or + * redistributing this file, you may do so under either license. + * + * GPL LICENSE SUMMARY + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of version 2 of the GNU General Public License as + * published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, but + * WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + * General Public License for more details. + * + * BSD LICENSE + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * + * - Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * - Redistributions in binary form must reproduce the above copyright + * notice, this list of conditions and the following disclaimer in + * the documentation and/or other materials provided with the + * distribution. + * - Neither the name of Intel Corporation nor the names of its + * contributors may be used to endorse or promote products derived + * from this software without specific prior written permission. + * + * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS + * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT + * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR + * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT + * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, + * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT + * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, + * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY + * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT + * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE + * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + * + */ + +/* + * This file contains OPA VNIC encapsulation/decapsulation function. + */ + +#include <linux/if_ether.h> +#include <linux/if_vlan.h> + +#include "opa_vnic_internal.h" + +/* OPA 16B Header fields */ +#define OPA_16B_LID_MASK 0xFFFFFull +#define OPA_16B_SLID_HIGH_SHFT 8 +#define OPA_16B_SLID_MASK 0xF00ull +#define OPA_16B_DLID_MASK 0xF000ull +#define OPA_16B_DLID_HIGH_SHFT 12 +#define OPA_16B_LEN_SHFT 20 +#define OPA_16B_SC_SHFT 20 +#define OPA_16B_RC_SHFT 25 +#define OPA_16B_PKEY_SHFT 16 + +#define OPA_VNIC_L4_HDR_SHFT 16 + +/* L2+L4 hdr len is 20 bytes (5 quad words) */ +#define OPA_VNIC_HDR_QW_LEN 5 + +static inline void opa_vnic_make_header(u8 *hdr, u32 slid, u32 dlid, u16 len, + u16 pkey, u16 entropy, u8 sc, u8 rc, + u8 l4_type, u16 l4_hdr) +{ + /* h[1]: LT=1, 16B L2=10 */ + u32 h[OPA_VNIC_HDR_QW_LEN] = {0, 0xc0000000, 0, 0, 0}; + + h[2] = l4_type; + h[3] = entropy; + h[4] = l4_hdr << OPA_VNIC_L4_HDR_SHFT; + + /* Extract and set 4 upper bits and 20 lower bits of the lids */ + h[0] |= (slid & OPA_16B_LID_MASK); + h[2] |= ((slid >> (20 - OPA_16B_SLID_HIGH_SHFT)) & OPA_16B_SLID_MASK); + + h[1] |= (dlid & OPA_16B_LID_MASK); + h[2] |= ((dlid >> (20 - OPA_16B_DLID_HIGH_SHFT)) & OPA_16B_DLID_MASK); + + h[0] |= (len << OPA_16B_LEN_SHFT); + h[1] |= (rc << OPA_16B_RC_SHFT); + h[1] |= (sc << OPA_16B_SC_SHFT); + h[2] |= ((u32)pkey << OPA_16B_PKEY_SHFT); + + memcpy(hdr, h, OPA_VNIC_HDR_LEN); +} + +/* opa_vnic_get_dlid - find and return the DLID */ +static uint32_t opa_vnic_get_dlid(struct opa_vnic_adapter *adapter, + struct sk_buff *skb, u8 def_port) +{ + struct __opa_veswport_info *info = &adapter->info; + struct ethhdr *mac_hdr = (struct ethhdr *)skb_mac_header(skb); + u32 dlid; + + if (is_multicast_ether_addr(mac_hdr->h_dest)) { + dlid = info->vesw.u_mcast_dlid; + } else { + if (is_local_ether_addr(mac_hdr->h_dest)) { + dlid = ((uint32_t)mac_hdr->h_dest[5] << 16) | + ((uint32_t)mac_hdr->h_dest[4] << 8) | + mac_hdr->h_dest[3]; + if (unlikely(!dlid)) + v_warn("Null dlid in MAC address\n"); + } else if (def_port != OPA_VNIC_INVALID_PORT) { + dlid = info->vesw.u_ucast_dlid[def_port]; + } + } + + return dlid; +} + +/* opa_vnic_get_sc - return the service class */ +static u8 opa_vnic_get_sc(struct __opa_veswport_info *info, + struct sk_buff *skb) +{ + struct ethhdr *mac_hdr = (struct ethhdr *)skb_mac_header(skb); + u16 vlan_tci; + u8 sc; + + if (!__vlan_get_tag(skb, &vlan_tci)) { + u8 pcp = OPA_VNIC_VLAN_PCP(vlan_tci); + + if (is_multicast_ether_addr(mac_hdr->h_dest)) + sc = info->vport.pcp_to_sc_mc[pcp]; + else + sc = info->vport.pcp_to_sc_uc[pcp]; + } else { + if (is_multicast_ether_addr(mac_hdr->h_dest)) + sc = info->vport.non_vlan_sc_mc; + else + sc = info->vport.non_vlan_sc_uc; + } + + return sc; +} + +u8 opa_vnic_get_vl(struct opa_vnic_adapter *adapter, struct sk_buff *skb) +{ + struct ethhdr *mac_hdr = (struct ethhdr *)skb_mac_header(skb); + struct __opa_veswport_info *info = &adapter->info; + u8 vl; + + if (skb_vlan_tag_present(skb)) { + u8 pcp = skb_vlan_tag_get(skb) >> VLAN_PRIO_SHIFT; + + if (is_multicast_ether_addr(mac_hdr->h_dest)) + vl = info->vport.pcp_to_vl_mc[pcp]; + else + vl = info->vport.pcp_to_vl_uc[pcp]; + } else { + if (is_multicast_ether_addr(mac_hdr->h_dest)) + vl = info->vport.non_vlan_vl_mc; + else + vl = info->vport.non_vlan_vl_uc; + } + + return vl; +} + +/* opa_vnic_calc_entropy - calculate the packet entropy */ +u8 opa_vnic_calc_entropy(struct opa_vnic_adapter *adapter, struct sk_buff *skb) +{ + u16 hash16; + + /* + * Get flow based 16-bit hash and then XOR the upper and lower bytes + * to get the entropy. + * __skb_tx_hash limits qcount to 16 bits. Hence, get 15-bit hash. + */ + hash16 = __skb_tx_hash(adapter->netdev, skb, BIT(15)); + return (u8)((hash16 >> 8) ^ (hash16 & 0xff)); +} + +/* opa_vnic_get_def_port - get default port based on entropy */ +static inline u8 opa_vnic_get_def_port(struct opa_vnic_adapter *adapter, + u8 entropy) +{ + u8 flow_id; + + /* Add the upper and lower 4-bits of entropy to get the flow id */ + flow_id = ((entropy & 0xf) + (entropy >> 4)); + return adapter->flow_tbl[flow_id & (OPA_VNIC_FLOW_TBL_SIZE - 1)]; +} + +/* Calculate packet length including OPA header, crc and padding */ +static inline int opa_vnic_wire_length(struct sk_buff *skb) +{ + u32 pad_len; + + /* padding for 8 bytes size alignment */ + pad_len = -(skb->len + OPA_VNIC_ICRC_TAIL_LEN) & 0x7; + pad_len += OPA_VNIC_ICRC_TAIL_LEN; + + return (skb->len + pad_len) >> 3; +} + +/* opa_vnic_encap_skb - encapsulate skb packet with OPA header and meta data */ +void opa_vnic_encap_skb(struct opa_vnic_adapter *adapter, struct sk_buff *skb) +{ + struct __opa_veswport_info *info = &adapter->info; + struct opa_vnic_skb_mdata *mdata; + u8 def_port, sc, entropy, *hdr; + u16 len, l4_hdr; + u32 dlid; + + hdr = skb_push(skb, OPA_VNIC_HDR_LEN); + + entropy = opa_vnic_calc_entropy(adapter, skb); + def_port = opa_vnic_get_def_port(adapter, entropy); + len = opa_vnic_wire_length(skb); + dlid = opa_vnic_get_dlid(adapter, skb, def_port); + sc = opa_vnic_get_sc(info, skb); + l4_hdr = info->vesw.vesw_id; + + mdata = (struct opa_vnic_skb_mdata *)skb_push(skb, sizeof(*mdata)); + mdata->vl = opa_vnic_get_vl(adapter, skb); + mdata->entropy = entropy; + mdata->flags = 0; + if (unlikely(!dlid)) { + mdata->flags = OPA_VNIC_SKB_MDATA_ENCAP_ERR; + return; + } + + opa_vnic_make_header(hdr, info->vport.encap_slid, dlid, len, + info->vesw.pkey, entropy, sc, 0, + OPA_VNIC_L4_ETHR, l4_hdr); +} diff --git a/drivers/infiniband/ulp/opa_vnic/opa_vnic_encap.h b/drivers/infiniband/ulp/opa_vnic/opa_vnic_encap.h new file mode 100644 index 0000000..176fca9 --- /dev/null +++ b/drivers/infiniband/ulp/opa_vnic/opa_vnic_encap.h @@ -0,0 +1,62 @@ +#ifndef _OPA_VNIC_ENCAP_H +#define _OPA_VNIC_ENCAP_H +/* + * Copyright(c) 2017 Intel Corporation. + * + * This file is provided under a dual BSD/GPLv2 license. When using or + * redistributing this file, you may do so under either license. + * + * GPL LICENSE SUMMARY + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of version 2 of the GNU General Public License as + * published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, but + * WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + * General Public License for more details. + * + * BSD LICENSE + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * + * - Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * - Redistributions in binary form must reproduce the above copyright + * notice, this list of conditions and the following disclaimer in + * the documentation and/or other materials provided with the + * distribution. + * - Neither the name of Intel Corporation nor the names of its + * contributors may be used to endorse or promote products derived + * from this software without specific prior written permission. + * + * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS + * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT + * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR + * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT + * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, + * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT + * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, + * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY + * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT + * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE + * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + * + */ + +/* + * This file contains all OPA VNIC declaration required for encapsulation + * and decapsulation of Ethernet packets + */ + +/* VNIC configured and operational state values */ +#define OPA_VNIC_STATE_DROP_ALL 0x1 +#define OPA_VNIC_STATE_FORWARDING 0x3 + +#define OPA_VESW_MAX_NUM_DEF_PORT 16 +#define OPA_VNIC_MAX_NUM_PCP 8 + +#endif /* _OPA_VNIC_ENCAP_H */ diff --git a/drivers/infiniband/ulp/opa_vnic/opa_vnic_ethtool.c b/drivers/infiniband/ulp/opa_vnic/opa_vnic_ethtool.c new file mode 100644 index 0000000..b74f6ad --- /dev/null +++ b/drivers/infiniband/ulp/opa_vnic/opa_vnic_ethtool.c @@ -0,0 +1,65 @@ +/* + * Copyright(c) 2017 Intel Corporation. + * + * This file is provided under a dual BSD/GPLv2 license. When using or + * redistributing this file, you may do so under either license. + * + * GPL LICENSE SUMMARY + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of version 2 of the GNU General Public License as + * published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, but + * WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + * General Public License for more details. + * + * BSD LICENSE + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * + * - Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * - Redistributions in binary form must reproduce the above copyright + * notice, this list of conditions and the following disclaimer in + * the documentation and/or other materials provided with the + * distribution. + * - Neither the name of Intel Corporation nor the names of its + * contributors may be used to endorse or promote products derived + * from this software without specific prior written permission. + * + * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS + * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT + * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR + * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT + * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, + * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT + * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, + * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY + * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT + * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE + * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + * + */ + +/* + * This file contains OPA VNIC ethtool functions + */ + +#include <linux/ethtool.h> + +#include "opa_vnic_internal.h" + +/* ethtool ops */ +static const struct ethtool_ops opa_vnic_ethtool_ops = { + .get_link = ethtool_op_get_link, +}; + +/* opa_vnic_set_ethtool_ops - set ethtool ops */ +void opa_vnic_set_ethtool_ops(struct net_device *netdev) +{ + netdev->ethtool_ops = &opa_vnic_ethtool_ops; +} diff --git a/drivers/infiniband/ulp/opa_vnic/opa_vnic_internal.h b/drivers/infiniband/ulp/opa_vnic/opa_vnic_internal.h new file mode 100644 index 0000000..83ffa91 --- /dev/null +++ b/drivers/infiniband/ulp/opa_vnic/opa_vnic_internal.h @@ -0,0 +1,186 @@ +#ifndef _OPA_VNIC_INTERNAL_H +#define _OPA_VNIC_INTERNAL_H +/* + * Copyright(c) 2017 Intel Corporation. + * + * This file is provided under a dual BSD/GPLv2 license. When using or + * redistributing this file, you may do so under either license. + * + * GPL LICENSE SUMMARY + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of version 2 of the GNU General Public License as + * published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, but + * WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + * General Public License for more details. + * + * BSD LICENSE + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * + * - Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * - Redistributions in binary form must reproduce the above copyright + * notice, this list of conditions and the following disclaimer in + * the documentation and/or other materials provided with the + * distribution. + * - Neither the name of Intel Corporation nor the names of its + * contributors may be used to endorse or promote products derived + * from this software without specific prior written permission. + * + * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS + * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT + * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR + * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT + * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, + * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT + * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, + * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY + * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT + * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE + * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + * + */ + +/* + * This file contains OPA VNIC driver internal declarations + */ + +#include <linux/bitops.h> +#include <linux/etherdevice.h> +#include <linux/hashtable.h> +#include <linux/sizes.h> +#include <rdma/opa_vnic.h> + +#include "opa_vnic_encap.h" + +#define OPA_VNIC_VLAN_PCP(vlan_tci) \ + (((vlan_tci) & VLAN_PRIO_MASK) >> VLAN_PRIO_SHIFT) + +/* Flow to default port redirection table size */ +#define OPA_VNIC_FLOW_TBL_SIZE 32 + +/* Invalid port number */ +#define OPA_VNIC_INVALID_PORT 0xff + +struct opa_vnic_adapter; + +/** + * struct __opa_vesw_info - OPA vnic virtual switch info + */ +struct __opa_vesw_info { + u16 fabric_id; + u16 vesw_id; + + u8 rsvd0[6]; + u16 def_port_mask; + + u8 rsvd1[2]; + u16 pkey; + + u8 rsvd2[4]; + u32 u_mcast_dlid; + u32 u_ucast_dlid[OPA_VESW_MAX_NUM_DEF_PORT]; + + u8 rsvd3[44]; + u16 eth_mtu[OPA_VNIC_MAX_NUM_PCP]; + u16 eth_mtu_non_vlan; + u8 rsvd4[2]; +} __packed; + +/** + * struct __opa_per_veswport_info - OPA vnic per port info + */ +struct __opa_per_veswport_info { + u32 port_num; + + u8 eth_link_status; + u8 rsvd0[3]; + + u8 base_mac_addr[ETH_ALEN]; + u8 config_state; + u8 oper_state; + + u16 max_mac_tbl_ent; + u16 max_smac_ent; + u32 mac_tbl_digest; + u8 rsvd1[4]; + + u32 encap_slid; + + u8 pcp_to_sc_uc[OPA_VNIC_MAX_NUM_PCP]; + u8 pcp_to_vl_uc[OPA_VNIC_MAX_NUM_PCP]; + u8 pcp_to_sc_mc[OPA_VNIC_MAX_NUM_PCP]; + u8 pcp_to_vl_mc[OPA_VNIC_MAX_NUM_PCP]; + + u8 non_vlan_sc_uc; + u8 non_vlan_vl_uc; + u8 non_vlan_sc_mc; + u8 non_vlan_vl_mc; + + u8 rsvd2[48]; + + u16 uc_macs_gen_count; + u16 mc_macs_gen_count; + + u8 rsvd3[8]; +} __packed; + +/** + * struct __opa_veswport_info - OPA vnic port info + */ +struct __opa_veswport_info { + struct __opa_vesw_info vesw; + struct __opa_per_veswport_info vport; +}; + +/** + * struct opa_vnic_adapter - OPA VNIC netdev private data structure + * @netdev: pointer to associated netdev + * @ibdev: ib device + * @rn_ops: rdma netdev's net_device_ops + * @port_num: OPA port number + * @vport_num: vesw port number + * @lock: adapter lock + * @info: virtual ethernet switch port information + * @flow_tbl: flow to default port redirection table + */ +struct opa_vnic_adapter { + struct net_device *netdev; + struct ib_device *ibdev; + const struct net_device_ops *rn_ops; + + u8 port_num; + u8 vport_num; + + /* Lock used around concurrent updates to netdev */ + struct mutex lock; + + struct __opa_veswport_info info; + + u8 flow_tbl[OPA_VNIC_FLOW_TBL_SIZE]; +}; + +#define v_dbg(format, arg...) \ + netdev_dbg(adapter->netdev, format, ## arg) +#define v_err(format, arg...) \ + netdev_err(adapter->netdev, format, ## arg) +#define v_info(format, arg...) \ + netdev_info(adapter->netdev, format, ## arg) +#define v_warn(format, arg...) \ + netdev_warn(adapter->netdev, format, ## arg) + +struct opa_vnic_adapter *opa_vnic_add_netdev(struct ib_device *ibdev, + u8 port_num, u8 vport_num); +void opa_vnic_rem_netdev(struct opa_vnic_adapter *adapter); +void opa_vnic_encap_skb(struct opa_vnic_adapter *adapter, struct sk_buff *skb); +u8 opa_vnic_get_vl(struct opa_vnic_adapter *adapter, struct sk_buff *skb); +u8 opa_vnic_calc_entropy(struct opa_vnic_adapter *adapter, struct sk_buff *skb); +void opa_vnic_set_ethtool_ops(struct net_device *netdev); + +#endif /* _OPA_VNIC_INTERNAL_H */ diff --git a/drivers/infiniband/ulp/opa_vnic/opa_vnic_netdev.c b/drivers/infiniband/ulp/opa_vnic/opa_vnic_netdev.c new file mode 100644 index 0000000..17e8d36 --- /dev/null +++ b/drivers/infiniband/ulp/opa_vnic/opa_vnic_netdev.c @@ -0,0 +1,229 @@ +/* + * Copyright(c) 2017 Intel Corporation. + * + * This file is provided under a dual BSD/GPLv2 license. When using or + * redistributing this file, you may do so under either license. + * + * GPL LICENSE SUMMARY + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of version 2 of the GNU General Public License as + * published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, but + * WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + * General Public License for more details. + * + * BSD LICENSE + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * + * - Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * - Redistributions in binary form must reproduce the above copyright + * notice, this list of conditions and the following disclaimer in + * the documentation and/or other materials provided with the + * distribution. + * - Neither the name of Intel Corporation nor the names of its + * contributors may be used to endorse or promote products derived + * from this software without specific prior written permission. + * + * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS + * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT + * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR + * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT + * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, + * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT + * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, + * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY + * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT + * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE + * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + * + */ + +/* + * This file contains OPA Virtual Network Interface Controller (VNIC) driver + * netdev functionality. + */ + +#include <linux/module.h> +#include <linux/if_vlan.h> + +#include "opa_vnic_internal.h" + +#define OPA_TX_TIMEOUT_MS 1000 + +#define OPA_VNIC_SKB_HEADROOM \ + ALIGN((OPA_VNIC_HDR_LEN + OPA_VNIC_SKB_MDATA_LEN), 8) + +/* opa_netdev_start_xmit - transmit function */ +static netdev_tx_t opa_netdev_start_xmit(struct sk_buff *skb, + struct net_device *netdev) +{ + struct opa_vnic_adapter *adapter = opa_vnic_priv(netdev); + + v_dbg("xmit: queue %d skb len %d\n", skb->queue_mapping, skb->len); + /* pad to ensure mininum ethernet packet length */ + if (unlikely(skb->len < ETH_ZLEN)) { + if (skb_padto(skb, ETH_ZLEN)) + return NETDEV_TX_OK; + + skb_put(skb, ETH_ZLEN - skb->len); + } + + opa_vnic_encap_skb(adapter, skb); + return adapter->rn_ops->ndo_start_xmit(skb, netdev); +} + +static u16 opa_vnic_select_queue(struct net_device *netdev, struct sk_buff *skb, + void *accel_priv, + select_queue_fallback_t fallback) +{ + struct opa_vnic_adapter *adapter = opa_vnic_priv(netdev); + struct opa_vnic_skb_mdata *mdata; + int rc; + + /* pass entropy and vl as metadata in skb */ + mdata = (struct opa_vnic_skb_mdata *)skb_push(skb, sizeof(*mdata)); + mdata->entropy = opa_vnic_calc_entropy(adapter, skb); + mdata->vl = opa_vnic_get_vl(adapter, skb); + rc = adapter->rn_ops->ndo_select_queue(netdev, skb, + accel_priv, fallback); + skb_pull(skb, sizeof(*mdata)); + return rc; +} + +/* opa_vnic_set_mac_addr - change mac address */ +static int opa_vnic_set_mac_addr(struct net_device *netdev, void *addr) +{ + struct opa_vnic_adapter *adapter = opa_vnic_priv(netdev); + struct sockaddr *sa = addr; + int rc; + + if (!memcmp(netdev->dev_addr, sa->sa_data, ETH_ALEN)) + return 0; + + mutex_lock(&adapter->lock); + rc = eth_mac_addr(netdev, addr); + mutex_unlock(&adapter->lock); + + return rc; +} + +/* opa_netdev_open - activate network interface */ +static int opa_netdev_open(struct net_device *netdev) +{ + struct opa_vnic_adapter *adapter = opa_vnic_priv(netdev); + int rc; + + rc = adapter->rn_ops->ndo_open(adapter->netdev); + if (rc) { + v_dbg("open failed %d\n", rc); + return rc; + } + + v_info("opened\n"); + return 0; +} + +/* opa_netdev_close - disable network interface */ +static int opa_netdev_close(struct net_device *netdev) +{ + struct opa_vnic_adapter *adapter = opa_vnic_priv(netdev); + int rc; + + rc = adapter->rn_ops->ndo_stop(adapter->netdev); + if (rc) { + v_dbg("close failed %d\n", rc); + return rc; + } + + v_info("closed\n"); + return 0; +} + +/* netdev ops */ +static const struct net_device_ops opa_netdev_ops = { + .ndo_open = opa_netdev_open, + .ndo_stop = opa_netdev_close, + .ndo_start_xmit = opa_netdev_start_xmit, + .ndo_select_queue = opa_vnic_select_queue, + .ndo_set_mac_address = opa_vnic_set_mac_addr, +}; + +/* opa_vnic_add_netdev - create vnic netdev interface */ +struct opa_vnic_adapter *opa_vnic_add_netdev(struct ib_device *ibdev, + u8 port_num, u8 vport_num) +{ + struct opa_vnic_adapter *adapter; + struct net_device *netdev; + struct rdma_netdev *rn; + int rc; + + netdev = ibdev->alloc_rdma_netdev(ibdev, port_num, + RDMA_NETDEV_OPA_VNIC, + "veth%d", NET_NAME_UNKNOWN, + ether_setup); + if (!netdev) + return ERR_PTR(-ENOMEM); + else if (IS_ERR(netdev)) + return ERR_CAST(netdev); + + adapter = kzalloc(sizeof(*adapter), GFP_KERNEL); + if (!adapter) { + rc = -ENOMEM; + goto adapter_err; + } + + rn = netdev_priv(netdev); + rn->clnt_priv = adapter; + rn->hca = ibdev; + rn->port_num = port_num; + adapter->netdev = netdev; + adapter->ibdev = ibdev; + adapter->port_num = port_num; + adapter->vport_num = vport_num; + adapter->rn_ops = netdev->netdev_ops; + + netdev->netdev_ops = &opa_netdev_ops; + netdev->priv_flags |= IFF_LIVE_ADDR_CHANGE; + netdev->hard_header_len += OPA_VNIC_SKB_HEADROOM; + mutex_init(&adapter->lock); + + SET_NETDEV_DEV(netdev, ibdev->dev.parent); + + opa_vnic_set_ethtool_ops(netdev); + rc = register_netdev(netdev); + if (rc) + goto netdev_err; + + netif_carrier_off(netdev); + netif_dormant_on(netdev); + v_info("initialized\n"); + + return adapter; +netdev_err: + mutex_destroy(&adapter->lock); + kfree(adapter); +adapter_err: + ibdev->free_rdma_netdev(netdev); + + return ERR_PTR(rc); +} + +/* opa_vnic_rem_netdev - remove vnic netdev interface */ +void opa_vnic_rem_netdev(struct opa_vnic_adapter *adapter) +{ + struct net_device *netdev = adapter->netdev; + struct ib_device *ibdev = adapter->ibdev; + + v_info("removing\n"); + unregister_netdev(netdev); + mutex_destroy(&adapter->lock); + kfree(adapter); + ibdev->free_rdma_netdev(netdev); +}