| lists.openwall.net | lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC | |
|
Open Source and information security mailing list archives
| ||
|
Date: Sun, 9 Mar 2014 16:54:02 +0200
From: Or Gerlitz <ogerlitz@...lanox.com>
To: davem@...emloft.net
Cc: netdev@...r.kernel.org, roland@...nel.org,
jackm@....mellanox.co.il, amirv@...lanox.com,
Moni Shoua <monis@...lanox.co.il>,
Or Gerlitz <ogerlitz@...lanox.com>
Subject: [PATCH net-next 5/7] mlx4: Implement IP based gids support for RoCE/SRIOV
From: Jack Morgenstein <jackm@....mellanox.co.il>
Since there is no connection between the MAC/VLAN and the GID
when using IP-based addressing, the proxy QP1 (running on the
slave) must pass the source-mac, destination-mac, and vlan_id
information separately from the GID. Additionally, the Host
must pass the remote source-mac and vlan_id back to the slave,
This is achieved as follows:
Outgoing MADs:
1. Source MAC: obtained from the CQ completion structure
(struct ib_wc, smac field).
2. Destination MAC: obtained from the tunnel header
3. vlan_id: obtained from the tunnel header.
Incoming MADs
1. The source (i.e., remote) MAC and vlan_id are passed in
the tunnel header to the proxy QP1.
VST mode support:
For outgoing MADs, the vlan_id obtained from the header is
discarded, and the vlan_id specified by the Hypervisor is used
instead.
For incoming MADs, the incoming vlan_id (in the wc) is discarded, and the
"invalid" vlan (0xffff) is substituted when forwarding to the slave.
Signed-off-by: Moni Shoua <monis@...lanox.co.il>
Signed-off-by: Jack Morgenstein <jackm@....mellanox.co.il>
Signed-off-by: Or Gerlitz <ogerlitz@...lanox.com>
---
drivers/infiniband/hw/mlx4/cq.c | 42 +++++++++++++++++++---------
drivers/infiniband/hw/mlx4/mad.c | 44 ++++++++++++++++++++++++++---
drivers/infiniband/hw/mlx4/mcg.c | 2 +-
drivers/infiniband/hw/mlx4/mlx4_ib.h | 2 +-
drivers/infiniband/hw/mlx4/qp.c | 11 ++++++-
drivers/net/ethernet/mellanox/mlx4/cmd.c | 23 +++++++++++++++
include/linux/mlx4/cmd.h | 6 ++++
include/linux/mlx4/device.h | 3 +-
8 files changed, 109 insertions(+), 24 deletions(-)
diff --git a/drivers/infiniband/hw/mlx4/cq.c b/drivers/infiniband/hw/mlx4/cq.c
index cc40f08..5f64081 100644
--- a/drivers/infiniband/hw/mlx4/cq.c
+++ b/drivers/infiniband/hw/mlx4/cq.c
@@ -564,7 +564,7 @@ static int mlx4_ib_ipoib_csum_ok(__be16 status, __be16 checksum)
}
static int use_tunnel_data(struct mlx4_ib_qp *qp, struct mlx4_ib_cq *cq, struct ib_wc *wc,
- unsigned tail, struct mlx4_cqe *cqe)
+ unsigned tail, struct mlx4_cqe *cqe, int is_eth)
{
struct mlx4_ib_proxy_sqp_hdr *hdr;
@@ -574,12 +574,20 @@ static int use_tunnel_data(struct mlx4_ib_qp *qp, struct mlx4_ib_cq *cq, struct
DMA_FROM_DEVICE);
hdr = (struct mlx4_ib_proxy_sqp_hdr *) (qp->sqp_proxy_rcv[tail].addr);
wc->pkey_index = be16_to_cpu(hdr->tun.pkey_index);
- wc->slid = be16_to_cpu(hdr->tun.slid_mac_47_32);
- wc->sl = (u8) (be16_to_cpu(hdr->tun.sl_vid) >> 12);
wc->src_qp = be32_to_cpu(hdr->tun.flags_src_qp) & 0xFFFFFF;
wc->wc_flags |= (hdr->tun.g_ml_path & 0x80) ? (IB_WC_GRH) : 0;
wc->dlid_path_bits = 0;
+ if (is_eth) {
+ wc->vlan_id = be16_to_cpu(hdr->tun.sl_vid);
+ memcpy(&(wc->smac[0]), (char *)&hdr->tun.mac_31_0, 4);
+ memcpy(&(wc->smac[4]), (char *)&hdr->tun.slid_mac_47_32, 2);
+ wc->wc_flags |= (IB_WC_WITH_VLAN | IB_WC_WITH_SMAC);
+ } else {
+ wc->slid = be16_to_cpu(hdr->tun.slid_mac_47_32);
+ wc->sl = (u8) (be16_to_cpu(hdr->tun.sl_vid) >> 12);
+ }
+
return 0;
}
@@ -594,6 +602,7 @@ static int mlx4_ib_poll_one(struct mlx4_ib_cq *cq,
struct mlx4_srq *msrq = NULL;
int is_send;
int is_error;
+ int is_eth;
u32 g_mlpath_rqpn;
u16 wqe_ctr;
unsigned tail = 0;
@@ -778,11 +787,15 @@ repoll:
break;
}
+ is_eth = (rdma_port_get_link_layer(wc->qp->device,
+ (*cur_qp)->port) ==
+ IB_LINK_LAYER_ETHERNET);
if (mlx4_is_mfunc(to_mdev(cq->ibcq.device)->dev)) {
if ((*cur_qp)->mlx4_ib_qp_type &
(MLX4_IB_QPT_PROXY_SMI_OWNER |
MLX4_IB_QPT_PROXY_SMI | MLX4_IB_QPT_PROXY_GSI))
- return use_tunnel_data(*cur_qp, cq, wc, tail, cqe);
+ return use_tunnel_data(*cur_qp, cq, wc, tail,
+ cqe, is_eth);
}
wc->slid = be16_to_cpu(cqe->rlid);
@@ -793,20 +806,21 @@ repoll:
wc->pkey_index = be32_to_cpu(cqe->immed_rss_invalid) & 0x7f;
wc->wc_flags |= mlx4_ib_ipoib_csum_ok(cqe->status,
cqe->checksum) ? IB_WC_IP_CSUM_OK : 0;
- if (rdma_port_get_link_layer(wc->qp->device,
- (*cur_qp)->port) == IB_LINK_LAYER_ETHERNET)
+ if (is_eth) {
wc->sl = be16_to_cpu(cqe->sl_vid) >> 13;
- else
- wc->sl = be16_to_cpu(cqe->sl_vid) >> 12;
- if (be32_to_cpu(cqe->vlan_my_qpn) & MLX4_CQE_VLAN_PRESENT_MASK) {
- wc->vlan_id = be16_to_cpu(cqe->sl_vid) &
- MLX4_CQE_VID_MASK;
+ if (be32_to_cpu(cqe->vlan_my_qpn) &
+ MLX4_CQE_VLAN_PRESENT_MASK) {
+ wc->vlan_id = be16_to_cpu(cqe->sl_vid) &
+ MLX4_CQE_VID_MASK;
+ } else {
+ wc->vlan_id = 0xffff;
+ }
+ memcpy(wc->smac, cqe->smac, ETH_ALEN);
+ wc->wc_flags |= (IB_WC_WITH_VLAN | IB_WC_WITH_SMAC);
} else {
+ wc->sl = be16_to_cpu(cqe->sl_vid) >> 12;
wc->vlan_id = 0xffff;
}
- wc->wc_flags |= IB_WC_WITH_VLAN;
- memcpy(wc->smac, cqe->smac, ETH_ALEN);
- wc->wc_flags |= IB_WC_WITH_SMAC;
}
return 0;
diff --git a/drivers/infiniband/hw/mlx4/mad.c b/drivers/infiniband/hw/mlx4/mad.c
index 37a3fc7..7aeec4a 100644
--- a/drivers/infiniband/hw/mlx4/mad.c
+++ b/drivers/infiniband/hw/mlx4/mad.c
@@ -545,11 +545,36 @@ int mlx4_ib_send_to_slave(struct mlx4_ib_dev *dev, int slave, u8 port,
/* adjust tunnel data */
tun_mad->hdr.pkey_index = cpu_to_be16(tun_pkey_ix);
- tun_mad->hdr.sl_vid = cpu_to_be16(((u16)(wc->sl)) << 12);
- tun_mad->hdr.slid_mac_47_32 = cpu_to_be16(wc->slid);
tun_mad->hdr.flags_src_qp = cpu_to_be32(wc->src_qp & 0xFFFFFF);
tun_mad->hdr.g_ml_path = (grh && (wc->wc_flags & IB_WC_GRH)) ? 0x80 : 0;
+ if (is_eth) {
+ u16 vlan = 0;
+ if (mlx4_get_slave_default_vlan(dev->dev, port, slave, &vlan,
+ NULL)) {
+ /* VST mode */
+ if (vlan != wc->vlan_id)
+ /* Packet vlan is not the VST-assigned vlan.
+ * Drop the packet.
+ */
+ goto out;
+ else
+ /* Remove the vlan tag before forwarding
+ * the packet to the VF.
+ */
+ vlan = 0xffff;
+ } else {
+ vlan = wc->vlan_id;
+ }
+
+ tun_mad->hdr.sl_vid = cpu_to_be16(vlan);
+ memcpy((char *)&tun_mad->hdr.mac_31_0, &(wc->smac[0]), 4);
+ memcpy((char *)&tun_mad->hdr.slid_mac_47_32, &(wc->smac[4]), 2);
+ } else {
+ tun_mad->hdr.sl_vid = cpu_to_be16(((u16)(wc->sl)) << 12);
+ tun_mad->hdr.slid_mac_47_32 = cpu_to_be16(wc->slid);
+ }
+
ib_dma_sync_single_for_device(&dev->ib_dev,
tun_qp->tx_ring[tun_tx_ix].buf.map,
sizeof (struct mlx4_rcv_tunnel_mad),
@@ -1116,8 +1141,9 @@ static int is_proxy_qp0(struct mlx4_ib_dev *dev, int qpn, int slave)
int mlx4_ib_send_to_wire(struct mlx4_ib_dev *dev, int slave, u8 port,
- enum ib_qp_type dest_qpt, u16 pkey_index, u32 remote_qpn,
- u32 qkey, struct ib_ah_attr *attr, struct ib_mad *mad)
+ enum ib_qp_type dest_qpt, u16 pkey_index,
+ u32 remote_qpn, u32 qkey, struct ib_ah_attr *attr,
+ u8 *s_mac, struct ib_mad *mad)
{
struct ib_sge list;
struct ib_send_wr wr, *bad_wr;
@@ -1206,6 +1232,9 @@ int mlx4_ib_send_to_wire(struct mlx4_ib_dev *dev, int slave, u8 port,
wr.num_sge = 1;
wr.opcode = IB_WR_SEND;
wr.send_flags = IB_SEND_SIGNALED;
+ if (s_mac)
+ memcpy(to_mah(ah)->av.eth.s_mac, s_mac, 6);
+
ret = ib_post_send(send_qp, &wr, &bad_wr);
out:
@@ -1333,6 +1362,11 @@ static void mlx4_ib_multiplex_mad(struct mlx4_ib_demux_pv_ctx *ctx, struct ib_wc
if (ah_attr.ah_flags & IB_AH_GRH)
if (get_real_sgid_index(dev, slave, ctx->port, &ah_attr))
return;
+ memcpy(ah_attr.dmac, tunnel->hdr.mac, 6);
+ ah_attr.vlan_id = be16_to_cpu(tunnel->hdr.vlan);
+ /* if slave have default vlan use it */
+ mlx4_get_slave_default_vlan(dev->dev, ctx->port, slave,
+ &ah_attr.vlan_id, &ah_attr.sl);
mlx4_ib_send_to_wire(dev, slave, ctx->port,
is_proxy_qp0(dev, wc->src_qp, slave) ?
@@ -1340,7 +1374,7 @@ static void mlx4_ib_multiplex_mad(struct mlx4_ib_demux_pv_ctx *ctx, struct ib_wc
be16_to_cpu(tunnel->hdr.pkey_index),
be32_to_cpu(tunnel->hdr.remote_qpn),
be32_to_cpu(tunnel->hdr.qkey),
- &ah_attr, &tunnel->mad);
+ &ah_attr, wc->smac, &tunnel->mad);
}
static int mlx4_ib_alloc_pv_bufs(struct mlx4_ib_demux_pv_ctx *ctx,
diff --git a/drivers/infiniband/hw/mlx4/mcg.c b/drivers/infiniband/hw/mlx4/mcg.c
index 25b2cdf..e192b3c 100644
--- a/drivers/infiniband/hw/mlx4/mcg.c
+++ b/drivers/infiniband/hw/mlx4/mcg.c
@@ -216,7 +216,7 @@ static int send_mad_to_wire(struct mlx4_ib_demux_ctx *ctx, struct ib_mad *mad)
mlx4_ib_query_ah(dev->sm_ah[ctx->port - 1], &ah_attr);
spin_unlock(&dev->sm_lock);
return mlx4_ib_send_to_wire(dev, mlx4_master_func_num(dev->dev), ctx->port,
- IB_QPT_GSI, 0, 1, IB_QP1_QKEY, &ah_attr, mad);
+ IB_QPT_GSI, 0, 1, IB_QP1_QKEY, &ah_attr, NULL, mad);
}
static int send_mad_to_slave(int slave, struct mlx4_ib_demux_ctx *ctx,
diff --git a/drivers/infiniband/hw/mlx4/mlx4_ib.h b/drivers/infiniband/hw/mlx4/mlx4_ib.h
index febc8f9..8216345 100644
--- a/drivers/infiniband/hw/mlx4/mlx4_ib.h
+++ b/drivers/infiniband/hw/mlx4/mlx4_ib.h
@@ -739,7 +739,7 @@ int mlx4_ib_send_to_slave(struct mlx4_ib_dev *dev, int slave, u8 port,
struct ib_grh *grh, struct ib_mad *mad);
int mlx4_ib_send_to_wire(struct mlx4_ib_dev *dev, int slave, u8 port,
enum ib_qp_type dest_qpt, u16 pkey_index, u32 remote_qpn,
- u32 qkey, struct ib_ah_attr *attr, struct ib_mad *mad);
+ u32 qkey, struct ib_ah_attr *attr, u8 *s_mac, struct ib_mad *mad);
__be64 mlx4_ib_get_new_demux_tid(struct mlx4_ib_demux_ctx *ctx);
int mlx4_ib_demux_cm_handler(struct ib_device *ibdev, int port, int *slave,
diff --git a/drivers/infiniband/hw/mlx4/qp.c b/drivers/infiniband/hw/mlx4/qp.c
index 11332f0..e1dc9c1 100644
--- a/drivers/infiniband/hw/mlx4/qp.c
+++ b/drivers/infiniband/hw/mlx4/qp.c
@@ -2152,7 +2152,7 @@ static int build_mlx_header(struct mlx4_ib_sqp *sqp, struct ib_send_wr *wr,
}
if (is_eth) {
- u8 smac[6];
+ u8 *smac;
struct in6_addr in6;
u16 pcp = (be32_to_cpu(ah->av.ib.sl_tclass_flowlabel) >> 29) << 13;
@@ -2164,7 +2164,12 @@ static int build_mlx_header(struct mlx4_ib_sqp *sqp, struct ib_send_wr *wr,
memcpy(&ctrl->srcrb_flags16[0], ah->av.eth.mac, 2);
memcpy(&ctrl->imm, ah->av.eth.mac + 2, 4);
memcpy(&in6, sgid.raw, sizeof(in6));
- rdma_get_ll_mac(&in6, smac);
+
+ if (!mlx4_is_mfunc(to_mdev(ib_dev)->dev))
+ smac = to_mdev(sqp->qp.ibqp.device)->
+ iboe.netdevs[sqp->qp.port - 1]->dev_addr;
+ else
+ smac = ah->av.eth.s_mac; /* use the src mac of the tunnel */
memcpy(sqp->ud_header.eth.smac_h, smac, 6);
if (!memcmp(sqp->ud_header.eth.smac_h, sqp->ud_header.eth.dmac_h, 6))
mlx->flags |= cpu_to_be32(MLX4_WQE_CTRL_FORCE_LOOPBACK);
@@ -2396,6 +2401,8 @@ static void build_tunnel_header(struct ib_send_wr *wr, void *wqe, unsigned *mlx_
hdr.remote_qpn = cpu_to_be32(wr->wr.ud.remote_qpn);
hdr.pkey_index = cpu_to_be16(wr->wr.ud.pkey_index);
hdr.qkey = cpu_to_be32(wr->wr.ud.remote_qkey);
+ memcpy(hdr.mac, ah->av.eth.mac, 6);
+ hdr.vlan = ah->av.eth.vlan;
spc = MLX4_INLINE_ALIGN -
((unsigned long) (inl + 1) & (MLX4_INLINE_ALIGN - 1));
diff --git a/drivers/net/ethernet/mellanox/mlx4/cmd.c b/drivers/net/ethernet/mellanox/mlx4/cmd.c
index 0d02fba..e0c1ea4 100644
--- a/drivers/net/ethernet/mellanox/mlx4/cmd.c
+++ b/drivers/net/ethernet/mellanox/mlx4/cmd.c
@@ -2289,6 +2289,29 @@ int mlx4_set_vf_vlan(struct mlx4_dev *dev, int port, int vf, u16 vlan, u8 qos)
}
EXPORT_SYMBOL_GPL(mlx4_set_vf_vlan);
+ /* mlx4_get_slave_default_vlan -
+ * return true if VST ( default vlan)
+ * if VST, will return vlan & qos (if not NULL)
+ */
+bool mlx4_get_slave_default_vlan(struct mlx4_dev *dev, int port, int slave, u16 *vlan, u8 *qos)
+{
+ struct mlx4_vport_oper_state *vp_oper;
+ struct mlx4_priv *priv;
+
+ priv = mlx4_priv(dev);
+ vp_oper = &priv->mfunc.master.vf_oper[slave].vport[port];
+
+ if (MLX4_VGT != vp_oper->state.default_vlan) {
+ if (vlan)
+ *vlan = vp_oper->state.default_vlan;
+ if (qos)
+ *qos = vp_oper->state.default_qos;
+ return true;
+ }
+ return false;
+}
+EXPORT_SYMBOL_GPL(mlx4_get_slave_default_vlan);
+
int mlx4_set_vf_spoofchk(struct mlx4_dev *dev, int port, int vf, bool setting)
{
struct mlx4_priv *priv = mlx4_priv(dev);
diff --git a/include/linux/mlx4/cmd.h b/include/linux/mlx4/cmd.h
index 79a3472..bae5f58 100644
--- a/include/linux/mlx4/cmd.h
+++ b/include/linux/mlx4/cmd.h
@@ -240,6 +240,12 @@ int mlx4_set_vf_vlan(struct mlx4_dev *dev, int port, int vf, u16 vlan, u8 qos);
int mlx4_set_vf_spoofchk(struct mlx4_dev *dev, int port, int vf, bool setting);
int mlx4_get_vf_config(struct mlx4_dev *dev, int port, int vf, struct ifla_vf_info *ivf);
int mlx4_set_vf_link_state(struct mlx4_dev *dev, int port, int vf, int link_state);
+/*
+ * mlx4_get_slave_default_vlan -
+ * return true if VST ( default vlan)
+ * if VST, will return vlan & qos (if not NULL)
+ */
+bool mlx4_get_slave_default_vlan(struct mlx4_dev *dev, int port, int slave, u16 *vlan, u8 *qos);
#define MLX4_COMM_GET_IF_REV(cmd_chan_ver) (u8)((cmd_chan_ver) >> 8)
diff --git a/include/linux/mlx4/device.h b/include/linux/mlx4/device.h
index 5cb7cee..5fe57bb 100644
--- a/include/linux/mlx4/device.h
+++ b/include/linux/mlx4/device.h
@@ -632,7 +632,8 @@ struct mlx4_eth_av {
u8 hop_limit;
__be32 sl_tclass_flowlabel;
u8 dgid[16];
- u32 reserved4[2];
+ u8 s_mac[6];
+ u8 reserved4[2];
__be16 vlan;
u8 mac[ETH_ALEN];
};
--
1.7.1
--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Powered by blists - more mailing lists