main.c 67.7 KB
Newer Older
1 2
/*
 * Copyright (c) 2006, 2007 Cisco Systems, Inc. All rights reserved.
3
 * Copyright (c) 2007, 2008 Mellanox Technologies. All rights reserved.
4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35
 *
 * This software is available to you under a choice of one of two
 * licenses.  You may choose to be licensed under the terms of the GNU
 * General Public License (GPL) Version 2, available from the file
 * COPYING in the main directory of this source tree, or the
 * OpenIB.org BSD license below:
 *
 *     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.
 *
 * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
 * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
 * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
 * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
 * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
 * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
 * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
 * SOFTWARE.
 */

#include <linux/module.h>
#include <linux/init.h>
36
#include <linux/slab.h>
37
#include <linux/errno.h>
E
Eli Cohen 已提交
38 39 40
#include <linux/netdevice.h>
#include <linux/inetdevice.h>
#include <linux/rtnetlink.h>
E
Eli Cohen 已提交
41
#include <linux/if_vlan.h>
42 43
#include <net/ipv6.h>
#include <net/addrconf.h>
44 45 46

#include <rdma/ib_smi.h>
#include <rdma/ib_user_verbs.h>
E
Eli Cohen 已提交
47
#include <rdma/ib_addr.h>
48 49 50

#include <linux/mlx4/driver.h>
#include <linux/mlx4/cmd.h>
51
#include <linux/mlx4/qp.h>
52 53 54 55

#include "mlx4_ib.h"
#include "user.h"

J
Jack Morgenstein 已提交
56
#define DRV_NAME	MLX4_IB_DRV_NAME
57 58
#define DRV_VERSION	"2.2-1"
#define DRV_RELDATE	"Feb 2014"
59

60
#define MLX4_IB_FLOW_MAX_PRIO 0xFFF
61
#define MLX4_IB_FLOW_QPN_MASK 0xFFFFFF
62

63 64 65 66 67
MODULE_AUTHOR("Roland Dreier");
MODULE_DESCRIPTION("Mellanox ConnectX HCA InfiniBand driver");
MODULE_LICENSE("Dual BSD/GPL");
MODULE_VERSION(DRV_VERSION);

68 69 70 71
int mlx4_ib_sm_guid_assign = 1;
module_param_named(sm_guid_assign, mlx4_ib_sm_guid_assign, int, 0444);
MODULE_PARM_DESC(sm_guid_assign, "Enable SM alias_GUID assignment if sm_guid_assign > 0 (Default: 1)");

72
static const char mlx4_ib_version[] =
73 74 75
	DRV_NAME ": Mellanox ConnectX InfiniBand driver v"
	DRV_VERSION " (" DRV_RELDATE ")\n";

E
Eli Cohen 已提交
76 77 78 79 80 81 82
struct update_gid_work {
	struct work_struct	work;
	union ib_gid		gids[128];
	struct mlx4_ib_dev     *dev;
	int			port;
};

83 84
static void do_slave_init(struct mlx4_ib_dev *ibdev, int slave, int do_init);

E
Eli Cohen 已提交
85 86
static struct workqueue_struct *wq;

87 88 89 90 91 92 93 94
static void init_query_mad(struct ib_smp *mad)
{
	mad->base_version  = 1;
	mad->mgmt_class    = IB_MGMT_CLASS_SUBN_LID_ROUTED;
	mad->class_version = 1;
	mad->method	   = IB_MGMT_METHOD_GET;
}

E
Eli Cohen 已提交
95 96
static union ib_gid zgid;

97 98
static int check_flow_steering_support(struct mlx4_dev *dev)
{
99
	int eth_num_ports = 0;
100 101
	int ib_num_ports = 0;

102 103 104 105 106 107 108 109 110 111 112 113 114 115 116
	int dmfs = dev->caps.steering_mode == MLX4_STEERING_MODE_DEVICE_MANAGED;

	if (dmfs) {
		int i;
		mlx4_foreach_port(i, dev, MLX4_PORT_TYPE_ETH)
			eth_num_ports++;
		mlx4_foreach_port(i, dev, MLX4_PORT_TYPE_IB)
			ib_num_ports++;
		dmfs &= (!ib_num_ports ||
			 (dev->caps.flags2 & MLX4_DEV_CAP_FLAG2_DMFS_IPOIB)) &&
			(!eth_num_ports ||
			 (dev->caps.flags2 & MLX4_DEV_CAP_FLAG2_FS_EN));
		if (ib_num_ports && mlx4_is_mfunc(dev)) {
			pr_warn("Device managed flow steering is unavailable for IB port in multifunction env.\n");
			dmfs = 0;
117 118
		}
	}
119
	return dmfs;
120 121
}

122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137
static int mlx4_ib_query_device(struct ib_device *ibdev,
				struct ib_device_attr *props)
{
	struct mlx4_ib_dev *dev = to_mdev(ibdev);
	struct ib_smp *in_mad  = NULL;
	struct ib_smp *out_mad = NULL;
	int err = -ENOMEM;

	in_mad  = kzalloc(sizeof *in_mad, GFP_KERNEL);
	out_mad = kmalloc(sizeof *out_mad, GFP_KERNEL);
	if (!in_mad || !out_mad)
		goto out;

	init_query_mad(in_mad);
	in_mad->attr_id = IB_SMP_ATTR_NODE_INFO;

138 139
	err = mlx4_MAD_IFC(to_mdev(ibdev), MLX4_MAD_IFC_IGNORE_KEYS,
			   1, NULL, NULL, in_mad, out_mad);
140 141 142 143 144 145 146 147 148
	if (err)
		goto out;

	memset(props, 0, sizeof *props);

	props->fw_ver = dev->dev->caps.fw_ver;
	props->device_cap_flags    = IB_DEVICE_CHANGE_PHY_PORT |
		IB_DEVICE_PORT_ACTIVE_EVENT		|
		IB_DEVICE_SYS_IMAGE_GUID		|
149 150
		IB_DEVICE_RC_RNR_NAK_GEN		|
		IB_DEVICE_BLOCK_MULTICAST_LOOPBACK;
151 152 153 154 155 156 157 158
	if (dev->dev->caps.flags & MLX4_DEV_CAP_FLAG_BAD_PKEY_CNTR)
		props->device_cap_flags |= IB_DEVICE_BAD_PKEY_CNTR;
	if (dev->dev->caps.flags & MLX4_DEV_CAP_FLAG_BAD_QKEY_CNTR)
		props->device_cap_flags |= IB_DEVICE_BAD_QKEY_CNTR;
	if (dev->dev->caps.flags & MLX4_DEV_CAP_FLAG_APM)
		props->device_cap_flags |= IB_DEVICE_AUTO_PATH_MIG;
	if (dev->dev->caps.flags & MLX4_DEV_CAP_FLAG_UD_AV_PORT)
		props->device_cap_flags |= IB_DEVICE_UD_AV_PORT_ENFORCE;
159 160
	if (dev->dev->caps.flags & MLX4_DEV_CAP_FLAG_IPOIB_CSUM)
		props->device_cap_flags |= IB_DEVICE_UD_IP_CSUM;
161
	if (dev->dev->caps.max_gso_sz && dev->dev->caps.flags & MLX4_DEV_CAP_FLAG_BLH)
E
Eli Cohen 已提交
162
		props->device_cap_flags |= IB_DEVICE_UD_TSO;
163 164 165 166 167 168
	if (dev->dev->caps.bmme_flags & MLX4_BMME_FLAG_RESERVED_LKEY)
		props->device_cap_flags |= IB_DEVICE_LOCAL_DMA_LKEY;
	if ((dev->dev->caps.bmme_flags & MLX4_BMME_FLAG_LOCAL_INV) &&
	    (dev->dev->caps.bmme_flags & MLX4_BMME_FLAG_REMOTE_INV) &&
	    (dev->dev->caps.bmme_flags & MLX4_BMME_FLAG_FAST_REG_WR))
		props->device_cap_flags |= IB_DEVICE_MEM_MGT_EXTENSIONS;
S
Sean Hefty 已提交
169 170
	if (dev->dev->caps.flags & MLX4_DEV_CAP_FLAG_XRC)
		props->device_cap_flags |= IB_DEVICE_XRC;
S
Shani Michaeli 已提交
171 172 173 174 175 176 177
	if (dev->dev->caps.flags & MLX4_DEV_CAP_FLAG_MEM_WINDOW)
		props->device_cap_flags |= IB_DEVICE_MEM_WINDOW;
	if (dev->dev->caps.bmme_flags & MLX4_BMME_FLAG_TYPE_2_WIN) {
		if (dev->dev->caps.bmme_flags & MLX4_BMME_FLAG_WIN_TYPE_2B)
			props->device_cap_flags |= IB_DEVICE_MEM_WINDOW_TYPE_2B;
		else
			props->device_cap_flags |= IB_DEVICE_MEM_WINDOW_TYPE_2A;
178
	if (dev->steering_support ==  MLX4_STEERING_MODE_DEVICE_MANAGED)
179
		props->device_cap_flags |= IB_DEVICE_MANAGED_FLOW_STEERING;
S
Shani Michaeli 已提交
180
	}
181 182 183

	props->vendor_id	   = be32_to_cpup((__be32 *) (out_mad->data + 36)) &
		0xffffff;
184
	props->vendor_part_id	   = dev->dev->pdev->device;
185 186 187 188 189
	props->hw_ver		   = be32_to_cpup((__be32 *) (out_mad->data + 32));
	memcpy(&props->sys_image_guid, out_mad->data +	4, 8);

	props->max_mr_size	   = ~0ull;
	props->page_size_cap	   = dev->dev->caps.page_size_cap;
190
	props->max_qp		   = dev->dev->quotas.qp;
191
	props->max_qp_wr	   = dev->dev->caps.max_wqes - MLX4_IB_SQ_MAX_SPARE;
192 193
	props->max_sge		   = min(dev->dev->caps.max_sq_sg,
					 dev->dev->caps.max_rq_sg);
194
	props->max_cq		   = dev->dev->quotas.cq;
195
	props->max_cqe		   = dev->dev->caps.max_cqes;
196
	props->max_mr		   = dev->dev->quotas.mpt;
197 198 199 200
	props->max_pd		   = dev->dev->caps.num_pds - dev->dev->caps.reserved_pds;
	props->max_qp_rd_atom	   = dev->dev->caps.max_qp_dest_rdma;
	props->max_qp_init_rd_atom = dev->dev->caps.max_qp_init_rdma;
	props->max_res_rd_atom	   = props->max_qp_rd_atom * props->max_qp;
201
	props->max_srq		   = dev->dev->quotas.srq;
202
	props->max_srq_wr	   = dev->dev->caps.max_srq_wqes - 1;
203
	props->max_srq_sge	   = dev->dev->caps.max_srq_sge;
204
	props->max_fast_reg_page_list_len = MLX4_MAX_FAST_REG_PAGES;
205 206 207
	props->local_ca_ack_delay  = dev->dev->caps.local_ca_ack_delay;
	props->atomic_cap	   = dev->dev->caps.flags & MLX4_DEV_CAP_FLAG_ATOMIC ?
		IB_ATOMIC_HCA : IB_ATOMIC_NONE;
208
	props->masked_atomic_cap   = props->atomic_cap;
209
	props->max_pkeys	   = dev->dev->caps.pkey_table_len[1];
210 211 212 213
	props->max_mcast_grp	   = dev->dev->caps.num_mgms + dev->dev->caps.num_amgms;
	props->max_mcast_qp_attach = dev->dev->caps.num_qp_per_mgm;
	props->max_total_mcast_qp_attach = props->max_mcast_qp_attach *
					   props->max_mcast_grp;
214
	props->max_map_per_fmr = dev->dev->caps.max_fmr_maps;
215 216 217 218 219 220 221 222

out:
	kfree(in_mad);
	kfree(out_mad);

	return err;
}

E
Eli Cohen 已提交
223 224
static enum rdma_link_layer
mlx4_ib_port_link_layer(struct ib_device *device, u8 port_num)
225
{
E
Eli Cohen 已提交
226
	struct mlx4_dev *dev = to_mdev(device)->dev;
227

228
	return dev->caps.port_mask[port_num] == MLX4_PORT_TYPE_IB ?
E
Eli Cohen 已提交
229 230
		IB_LINK_LAYER_INFINIBAND : IB_LINK_LAYER_ETHERNET;
}
231

E
Eli Cohen 已提交
232
static int ib_link_query_port(struct ib_device *ibdev, u8 port,
233
			      struct ib_port_attr *props, int netw_view)
E
Eli Cohen 已提交
234
{
235 236
	struct ib_smp *in_mad  = NULL;
	struct ib_smp *out_mad = NULL;
237
	int ext_active_speed;
238
	int mad_ifc_flags = MLX4_MAD_IFC_IGNORE_KEYS;
239 240 241 242 243 244 245 246 247 248 249
	int err = -ENOMEM;

	in_mad  = kzalloc(sizeof *in_mad, GFP_KERNEL);
	out_mad = kmalloc(sizeof *out_mad, GFP_KERNEL);
	if (!in_mad || !out_mad)
		goto out;

	init_query_mad(in_mad);
	in_mad->attr_id  = IB_SMP_ATTR_PORT_INFO;
	in_mad->attr_mod = cpu_to_be32(port);

250 251 252 253
	if (mlx4_is_mfunc(to_mdev(ibdev)->dev) && netw_view)
		mad_ifc_flags |= MLX4_MAD_IFC_NET_VIEW;

	err = mlx4_MAD_IFC(to_mdev(ibdev), mad_ifc_flags, port, NULL, NULL,
254 255 256 257
				in_mad, out_mad);
	if (err)
		goto out;

258

259 260 261 262 263 264 265
	props->lid		= be16_to_cpup((__be16 *) (out_mad->data + 16));
	props->lmc		= out_mad->data[34] & 0x7;
	props->sm_lid		= be16_to_cpup((__be16 *) (out_mad->data + 18));
	props->sm_sl		= out_mad->data[36] & 0xf;
	props->state		= out_mad->data[32] & 0xf;
	props->phys_state	= out_mad->data[33] >> 4;
	props->port_cap_flags	= be32_to_cpup((__be32 *) (out_mad->data + 20));
266 267 268 269
	if (netw_view)
		props->gid_tbl_len = out_mad->data[50];
	else
		props->gid_tbl_len = to_mdev(ibdev)->dev->caps.gid_table_len[port];
270
	props->max_msg_sz	= to_mdev(ibdev)->dev->caps.max_msg_sz;
271
	props->pkey_tbl_len	= to_mdev(ibdev)->dev->caps.pkey_table_len[port];
272 273 274 275 276 277 278 279 280 281
	props->bad_pkey_cntr	= be16_to_cpup((__be16 *) (out_mad->data + 46));
	props->qkey_viol_cntr	= be16_to_cpup((__be16 *) (out_mad->data + 48));
	props->active_width	= out_mad->data[31] & 0xf;
	props->active_speed	= out_mad->data[35] >> 4;
	props->max_mtu		= out_mad->data[41] & 0xf;
	props->active_mtu	= out_mad->data[36] >> 4;
	props->subnet_timeout	= out_mad->data[51] & 0x1f;
	props->max_vl_num	= out_mad->data[37] >> 4;
	props->init_type_reply	= out_mad->data[41] >> 4;

282 283 284 285 286 287
	/* Check if extended speeds (EDR/FDR/...) are supported */
	if (props->port_cap_flags & IB_PORT_EXTENDED_SPEEDS_SUP) {
		ext_active_speed = out_mad->data[62] >> 4;

		switch (ext_active_speed) {
		case 1:
288
			props->active_speed = IB_SPEED_FDR;
289 290
			break;
		case 2:
291
			props->active_speed = IB_SPEED_EDR;
292 293 294 295 296
			break;
		}
	}

	/* If reported active speed is QDR, check if is FDR-10 */
297
	if (props->active_speed == IB_SPEED_QDR) {
298 299 300 301
		init_query_mad(in_mad);
		in_mad->attr_id = MLX4_ATTR_EXTENDED_PORT_INFO;
		in_mad->attr_mod = cpu_to_be32(port);

302
		err = mlx4_MAD_IFC(to_mdev(ibdev), mad_ifc_flags, port,
303 304
				   NULL, NULL, in_mad, out_mad);
		if (err)
305
			goto out;
306 307 308 309

		/* Checking LinkSpeedActive for FDR-10 */
		if (out_mad->data[15] & 0x1)
			props->active_speed = IB_SPEED_FDR10;
310
	}
311 312 313 314 315

	/* Avoid wrong speed value returned by FW if the IB link is down. */
	if (props->state == IB_PORT_DOWN)
		 props->active_speed = IB_SPEED_SDR;

316 317 318 319
out:
	kfree(in_mad);
	kfree(out_mad);
	return err;
E
Eli Cohen 已提交
320 321 322 323 324 325 326 327
}

static u8 state_to_phys_state(enum ib_port_state state)
{
	return state == IB_PORT_ACTIVE ? 5 : 3;
}

static int eth_link_query_port(struct ib_device *ibdev, u8 port,
328
			       struct ib_port_attr *props, int netw_view)
E
Eli Cohen 已提交
329
{
330 331 332

	struct mlx4_ib_dev *mdev = to_mdev(ibdev);
	struct mlx4_ib_iboe *iboe = &mdev->iboe;
E
Eli Cohen 已提交
333 334
	struct net_device *ndev;
	enum ib_mtu tmp;
335 336 337 338 339 340
	struct mlx4_cmd_mailbox *mailbox;
	int err = 0;

	mailbox = mlx4_alloc_cmd_mailbox(mdev->dev);
	if (IS_ERR(mailbox))
		return PTR_ERR(mailbox);
E
Eli Cohen 已提交
341

342 343 344 345 346 347 348 349
	err = mlx4_cmd_box(mdev->dev, 0, mailbox->dma, port, 0,
			   MLX4_CMD_QUERY_PORT, MLX4_CMD_TIME_CLASS_B,
			   MLX4_CMD_WRAPPED);
	if (err)
		goto out;

	props->active_width	=  (((u8 *)mailbox->buf)[5] == 0x40) ?
						IB_WIDTH_4X : IB_WIDTH_1X;
350
	props->active_speed	= IB_SPEED_QDR;
351
	props->port_cap_flags	= IB_PORT_CM_SUP | IB_PORT_IP_BASED_GIDS;
352 353
	props->gid_tbl_len	= mdev->dev->caps.gid_table_len[port];
	props->max_msg_sz	= mdev->dev->caps.max_msg_sz;
E
Eli Cohen 已提交
354
	props->pkey_tbl_len	= 1;
O
Or Gerlitz 已提交
355
	props->max_mtu		= IB_MTU_4096;
356
	props->max_vl_num	= 2;
E
Eli Cohen 已提交
357 358 359 360 361 362
	props->state		= IB_PORT_DOWN;
	props->phys_state	= state_to_phys_state(props->state);
	props->active_mtu	= IB_MTU_256;
	spin_lock(&iboe->lock);
	ndev = iboe->netdevs[port - 1];
	if (!ndev)
363
		goto out_unlock;
E
Eli Cohen 已提交
364 365 366 367

	tmp = iboe_get_mtu(ndev->mtu);
	props->active_mtu = tmp ? min(props->max_mtu, tmp) : IB_MTU_256;

E
Eli Cohen 已提交
368
	props->state		= (netif_running(ndev) && netif_carrier_ok(ndev)) ?
E
Eli Cohen 已提交
369 370
					IB_PORT_ACTIVE : IB_PORT_DOWN;
	props->phys_state	= state_to_phys_state(props->state);
371
out_unlock:
E
Eli Cohen 已提交
372
	spin_unlock(&iboe->lock);
373 374 375
out:
	mlx4_free_cmd_mailbox(mdev->dev, mailbox);
	return err;
E
Eli Cohen 已提交
376 377
}

378 379
int __mlx4_ib_query_port(struct ib_device *ibdev, u8 port,
			 struct ib_port_attr *props, int netw_view)
E
Eli Cohen 已提交
380
{
381
	int err;
E
Eli Cohen 已提交
382 383 384 385

	memset(props, 0, sizeof *props);

	err = mlx4_ib_port_link_layer(ibdev, port) == IB_LINK_LAYER_INFINIBAND ?
386 387
		ib_link_query_port(ibdev, port, props, netw_view) :
				eth_link_query_port(ibdev, port, props, netw_view);
388 389 390 391

	return err;
}

392 393 394 395 396 397 398
static int mlx4_ib_query_port(struct ib_device *ibdev, u8 port,
			      struct ib_port_attr *props)
{
	/* returns host view */
	return __mlx4_ib_query_port(ibdev, port, props, 0);
}

399 400
int __mlx4_ib_query_gid(struct ib_device *ibdev, u8 port, int index,
			union ib_gid *gid, int netw_view)
401 402 403 404
{
	struct ib_smp *in_mad  = NULL;
	struct ib_smp *out_mad = NULL;
	int err = -ENOMEM;
405 406 407
	struct mlx4_ib_dev *dev = to_mdev(ibdev);
	int clear = 0;
	int mad_ifc_flags = MLX4_MAD_IFC_IGNORE_KEYS;
408 409 410 411 412 413 414 415 416 417

	in_mad  = kzalloc(sizeof *in_mad, GFP_KERNEL);
	out_mad = kmalloc(sizeof *out_mad, GFP_KERNEL);
	if (!in_mad || !out_mad)
		goto out;

	init_query_mad(in_mad);
	in_mad->attr_id  = IB_SMP_ATTR_PORT_INFO;
	in_mad->attr_mod = cpu_to_be32(port);

418 419 420 421
	if (mlx4_is_mfunc(dev->dev) && netw_view)
		mad_ifc_flags |= MLX4_MAD_IFC_NET_VIEW;

	err = mlx4_MAD_IFC(dev, mad_ifc_flags, port, NULL, NULL, in_mad, out_mad);
422 423 424 425 426
	if (err)
		goto out;

	memcpy(gid->raw, out_mad->data + 8, 8);

427 428 429 430 431 432 433 434 435
	if (mlx4_is_mfunc(dev->dev) && !netw_view) {
		if (index) {
			/* For any index > 0, return the null guid */
			err = 0;
			clear = 1;
			goto out;
		}
	}

436 437 438 439
	init_query_mad(in_mad);
	in_mad->attr_id  = IB_SMP_ATTR_GUID_INFO;
	in_mad->attr_mod = cpu_to_be32(index / 8);

440
	err = mlx4_MAD_IFC(dev, mad_ifc_flags, port,
441
			   NULL, NULL, in_mad, out_mad);
442 443 444 445 446 447
	if (err)
		goto out;

	memcpy(gid->raw + 8, out_mad->data + (index % 8) * 8, 8);

out:
448 449
	if (clear)
		memset(gid->raw + 8, 0, 8);
450 451 452 453 454
	kfree(in_mad);
	kfree(out_mad);
	return err;
}

E
Eli Cohen 已提交
455 456 457 458 459 460 461 462 463 464 465 466 467 468
static int iboe_query_gid(struct ib_device *ibdev, u8 port, int index,
			  union ib_gid *gid)
{
	struct mlx4_ib_dev *dev = to_mdev(ibdev);

	*gid = dev->iboe.gid_table[port - 1][index];

	return 0;
}

static int mlx4_ib_query_gid(struct ib_device *ibdev, u8 port, int index,
			     union ib_gid *gid)
{
	if (rdma_port_get_link_layer(ibdev, port) == IB_LINK_LAYER_INFINIBAND)
469
		return __mlx4_ib_query_gid(ibdev, port, index, gid, 0);
E
Eli Cohen 已提交
470 471 472 473
	else
		return iboe_query_gid(ibdev, port, index, gid);
}

474 475
int __mlx4_ib_query_pkey(struct ib_device *ibdev, u8 port, u16 index,
			 u16 *pkey, int netw_view)
476 477 478
{
	struct ib_smp *in_mad  = NULL;
	struct ib_smp *out_mad = NULL;
479
	int mad_ifc_flags = MLX4_MAD_IFC_IGNORE_KEYS;
480 481 482 483 484 485 486 487 488 489 490
	int err = -ENOMEM;

	in_mad  = kzalloc(sizeof *in_mad, GFP_KERNEL);
	out_mad = kmalloc(sizeof *out_mad, GFP_KERNEL);
	if (!in_mad || !out_mad)
		goto out;

	init_query_mad(in_mad);
	in_mad->attr_id  = IB_SMP_ATTR_PKEY_TABLE;
	in_mad->attr_mod = cpu_to_be32(index / 32);

491 492 493 494 495
	if (mlx4_is_mfunc(to_mdev(ibdev)->dev) && netw_view)
		mad_ifc_flags |= MLX4_MAD_IFC_NET_VIEW;

	err = mlx4_MAD_IFC(to_mdev(ibdev), mad_ifc_flags, port, NULL, NULL,
			   in_mad, out_mad);
496 497 498 499 500 501 502 503 504 505 506
	if (err)
		goto out;

	*pkey = be16_to_cpu(((__be16 *) out_mad->data)[index % 32]);

out:
	kfree(in_mad);
	kfree(out_mad);
	return err;
}

507 508 509 510 511
static int mlx4_ib_query_pkey(struct ib_device *ibdev, u8 port, u16 index, u16 *pkey)
{
	return __mlx4_ib_query_pkey(ibdev, port, index, pkey, 0);
}

512 513 514
static int mlx4_ib_modify_device(struct ib_device *ibdev, int mask,
				 struct ib_device_modify *props)
{
515
	struct mlx4_cmd_mailbox *mailbox;
516
	unsigned long flags;
517

518 519 520
	if (mask & ~IB_DEVICE_MODIFY_NODE_DESC)
		return -EOPNOTSUPP;

521 522 523
	if (!(mask & IB_DEVICE_MODIFY_NODE_DESC))
		return 0;

524 525 526
	if (mlx4_is_slave(to_mdev(ibdev)->dev))
		return -EOPNOTSUPP;

527
	spin_lock_irqsave(&to_mdev(ibdev)->sm_lock, flags);
528
	memcpy(ibdev->node_desc, props->node_desc, 64);
529
	spin_unlock_irqrestore(&to_mdev(ibdev)->sm_lock, flags);
530 531 532 533 534 535 536 537 538 539 540

	/*
	 * If possible, pass node desc to FW, so it can generate
	 * a 144 trap.  If cmd fails, just ignore.
	 */
	mailbox = mlx4_alloc_cmd_mailbox(to_mdev(ibdev)->dev);
	if (IS_ERR(mailbox))
		return 0;

	memcpy(mailbox->buf, props->node_desc, 64);
	mlx4_cmd(to_mdev(ibdev)->dev, mailbox->dma, 1, 0,
541
		 MLX4_CMD_SET_NODE, MLX4_CMD_TIME_CLASS_A, MLX4_CMD_NATIVE);
542 543

	mlx4_free_cmd_mailbox(to_mdev(ibdev)->dev, mailbox);
544 545 546 547

	return 0;
}

548 549
static int mlx4_ib_SET_PORT(struct mlx4_ib_dev *dev, u8 port, int reset_qkey_viols,
			    u32 cap_mask)
550 551 552 553 554 555 556 557
{
	struct mlx4_cmd_mailbox *mailbox;
	int err;

	mailbox = mlx4_alloc_cmd_mailbox(dev->dev);
	if (IS_ERR(mailbox))
		return PTR_ERR(mailbox);

558 559 560 561 562 563 564
	if (dev->dev->flags & MLX4_FLAG_OLD_PORT_CMDS) {
		*(u8 *) mailbox->buf	     = !!reset_qkey_viols << 6;
		((__be32 *) mailbox->buf)[2] = cpu_to_be32(cap_mask);
	} else {
		((u8 *) mailbox->buf)[3]     = !!reset_qkey_viols;
		((__be32 *) mailbox->buf)[1] = cpu_to_be32(cap_mask);
	}
565

566 567
	err = mlx4_cmd(dev->dev, mailbox->dma, port, 0, MLX4_CMD_SET_PORT,
		       MLX4_CMD_TIME_CLASS_B, MLX4_CMD_WRAPPED);
568 569 570 571 572 573 574 575

	mlx4_free_cmd_mailbox(dev->dev, mailbox);
	return err;
}

static int mlx4_ib_modify_port(struct ib_device *ibdev, u8 port, int mask,
			       struct ib_port_modify *props)
{
576 577
	struct mlx4_ib_dev *mdev = to_mdev(ibdev);
	u8 is_eth = mdev->dev->caps.port_type[port] == MLX4_PORT_TYPE_ETH;
578 579 580 581
	struct ib_port_attr attr;
	u32 cap_mask;
	int err;

582 583 584 585 586 587 588 589
	/* return OK if this is RoCE. CM calls ib_modify_port() regardless
	 * of whether port link layer is ETH or IB. For ETH ports, qkey
	 * violations and port capabilities are not meaningful.
	 */
	if (is_eth)
		return 0;

	mutex_lock(&mdev->cap_mask_mutex);
590 591 592 593 594 595 596 597

	err = mlx4_ib_query_port(ibdev, port, &attr);
	if (err)
		goto out;

	cap_mask = (attr.port_cap_flags | props->set_port_cap_mask) &
		~props->clr_port_cap_mask;

598 599 600
	err = mlx4_ib_SET_PORT(mdev, port,
			       !!(mask & IB_PORT_RESET_QKEY_CNTR),
			       cap_mask);
601 602 603 604 605 606 607 608 609 610 611

out:
	mutex_unlock(&to_mdev(ibdev)->cap_mask_mutex);
	return err;
}

static struct ib_ucontext *mlx4_ib_alloc_ucontext(struct ib_device *ibdev,
						  struct ib_udata *udata)
{
	struct mlx4_ib_dev *dev = to_mdev(ibdev);
	struct mlx4_ib_ucontext *context;
O
Or Gerlitz 已提交
612
	struct mlx4_ib_alloc_ucontext_resp_v3 resp_v3;
613 614 615
	struct mlx4_ib_alloc_ucontext_resp resp;
	int err;

616 617 618
	if (!dev->ib_active)
		return ERR_PTR(-EAGAIN);

O
Or Gerlitz 已提交
619 620 621 622 623 624 625 626 627 628 629
	if (ibdev->uverbs_abi_ver == MLX4_IB_UVERBS_NO_DEV_CAPS_ABI_VERSION) {
		resp_v3.qp_tab_size      = dev->dev->caps.num_qps;
		resp_v3.bf_reg_size      = dev->dev->caps.bf_reg_size;
		resp_v3.bf_regs_per_page = dev->dev->caps.bf_regs_per_page;
	} else {
		resp.dev_caps	      = dev->dev->caps.userspace_caps;
		resp.qp_tab_size      = dev->dev->caps.num_qps;
		resp.bf_reg_size      = dev->dev->caps.bf_reg_size;
		resp.bf_regs_per_page = dev->dev->caps.bf_regs_per_page;
		resp.cqe_size	      = dev->dev->caps.cqe_size;
	}
630 631 632 633 634 635 636 637 638 639 640 641 642 643

	context = kmalloc(sizeof *context, GFP_KERNEL);
	if (!context)
		return ERR_PTR(-ENOMEM);

	err = mlx4_uar_alloc(to_mdev(ibdev)->dev, &context->uar);
	if (err) {
		kfree(context);
		return ERR_PTR(err);
	}

	INIT_LIST_HEAD(&context->db_page_list);
	mutex_init(&context->db_page_mutex);

O
Or Gerlitz 已提交
644 645 646 647 648
	if (ibdev->uverbs_abi_ver == MLX4_IB_UVERBS_NO_DEV_CAPS_ABI_VERSION)
		err = ib_copy_to_udata(udata, &resp_v3, sizeof(resp_v3));
	else
		err = ib_copy_to_udata(udata, &resp, sizeof(resp));

649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682
	if (err) {
		mlx4_uar_free(to_mdev(ibdev)->dev, &context->uar);
		kfree(context);
		return ERR_PTR(-EFAULT);
	}

	return &context->ibucontext;
}

static int mlx4_ib_dealloc_ucontext(struct ib_ucontext *ibcontext)
{
	struct mlx4_ib_ucontext *context = to_mucontext(ibcontext);

	mlx4_uar_free(to_mdev(ibcontext->device)->dev, &context->uar);
	kfree(context);

	return 0;
}

static int mlx4_ib_mmap(struct ib_ucontext *context, struct vm_area_struct *vma)
{
	struct mlx4_ib_dev *dev = to_mdev(context->device);

	if (vma->vm_end - vma->vm_start != PAGE_SIZE)
		return -EINVAL;

	if (vma->vm_pgoff == 0) {
		vma->vm_page_prot = pgprot_noncached(vma->vm_page_prot);

		if (io_remap_pfn_range(vma, vma->vm_start,
				       to_mucontext(context)->uar.pfn,
				       PAGE_SIZE, vma->vm_page_prot))
			return -EAGAIN;
	} else if (vma->vm_pgoff == 1 && dev->dev->caps.bf_reg_size != 0) {
683
		vma->vm_page_prot = pgprot_writecombine(vma->vm_page_prot);
684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730

		if (io_remap_pfn_range(vma, vma->vm_start,
				       to_mucontext(context)->uar.pfn +
				       dev->dev->caps.num_uars,
				       PAGE_SIZE, vma->vm_page_prot))
			return -EAGAIN;
	} else
		return -EINVAL;

	return 0;
}

static struct ib_pd *mlx4_ib_alloc_pd(struct ib_device *ibdev,
				      struct ib_ucontext *context,
				      struct ib_udata *udata)
{
	struct mlx4_ib_pd *pd;
	int err;

	pd = kmalloc(sizeof *pd, GFP_KERNEL);
	if (!pd)
		return ERR_PTR(-ENOMEM);

	err = mlx4_pd_alloc(to_mdev(ibdev)->dev, &pd->pdn);
	if (err) {
		kfree(pd);
		return ERR_PTR(err);
	}

	if (context)
		if (ib_copy_to_udata(udata, &pd->pdn, sizeof (__u32))) {
			mlx4_pd_free(to_mdev(ibdev)->dev, pd->pdn);
			kfree(pd);
			return ERR_PTR(-EFAULT);
		}

	return &pd->ibpd;
}

static int mlx4_ib_dealloc_pd(struct ib_pd *pd)
{
	mlx4_pd_free(to_mdev(pd->device)->dev, to_mpd(pd)->pdn);
	kfree(pd);

	return 0;
}

S
Sean Hefty 已提交
731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781
static struct ib_xrcd *mlx4_ib_alloc_xrcd(struct ib_device *ibdev,
					  struct ib_ucontext *context,
					  struct ib_udata *udata)
{
	struct mlx4_ib_xrcd *xrcd;
	int err;

	if (!(to_mdev(ibdev)->dev->caps.flags & MLX4_DEV_CAP_FLAG_XRC))
		return ERR_PTR(-ENOSYS);

	xrcd = kmalloc(sizeof *xrcd, GFP_KERNEL);
	if (!xrcd)
		return ERR_PTR(-ENOMEM);

	err = mlx4_xrcd_alloc(to_mdev(ibdev)->dev, &xrcd->xrcdn);
	if (err)
		goto err1;

	xrcd->pd = ib_alloc_pd(ibdev);
	if (IS_ERR(xrcd->pd)) {
		err = PTR_ERR(xrcd->pd);
		goto err2;
	}

	xrcd->cq = ib_create_cq(ibdev, NULL, NULL, xrcd, 1, 0);
	if (IS_ERR(xrcd->cq)) {
		err = PTR_ERR(xrcd->cq);
		goto err3;
	}

	return &xrcd->ibxrcd;

err3:
	ib_dealloc_pd(xrcd->pd);
err2:
	mlx4_xrcd_free(to_mdev(ibdev)->dev, xrcd->xrcdn);
err1:
	kfree(xrcd);
	return ERR_PTR(err);
}

static int mlx4_ib_dealloc_xrcd(struct ib_xrcd *xrcd)
{
	ib_destroy_cq(to_mxrcd(xrcd)->cq);
	ib_dealloc_pd(to_mxrcd(xrcd)->pd);
	mlx4_xrcd_free(to_mdev(xrcd->device)->dev, to_mxrcd(xrcd)->xrcdn);
	kfree(xrcd);

	return 0;
}

E
Eli Cohen 已提交
782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827
static int add_gid_entry(struct ib_qp *ibqp, union ib_gid *gid)
{
	struct mlx4_ib_qp *mqp = to_mqp(ibqp);
	struct mlx4_ib_dev *mdev = to_mdev(ibqp->device);
	struct mlx4_ib_gid_entry *ge;

	ge = kzalloc(sizeof *ge, GFP_KERNEL);
	if (!ge)
		return -ENOMEM;

	ge->gid = *gid;
	if (mlx4_ib_add_mc(mdev, mqp, gid)) {
		ge->port = mqp->port;
		ge->added = 1;
	}

	mutex_lock(&mqp->mutex);
	list_add_tail(&ge->list, &mqp->gid_list);
	mutex_unlock(&mqp->mutex);

	return 0;
}

int mlx4_ib_add_mc(struct mlx4_ib_dev *mdev, struct mlx4_ib_qp *mqp,
		   union ib_gid *gid)
{
	struct net_device *ndev;
	int ret = 0;

	if (!mqp->port)
		return 0;

	spin_lock(&mdev->iboe.lock);
	ndev = mdev->iboe.netdevs[mqp->port - 1];
	if (ndev)
		dev_hold(ndev);
	spin_unlock(&mdev->iboe.lock);

	if (ndev) {
		ret = 1;
		dev_put(ndev);
	}

	return ret;
}

828 829 830 831 832 833
struct mlx4_ib_steering {
	struct list_head list;
	u64 reg_id;
	union ib_gid gid;
};

834
static int parse_flow_attr(struct mlx4_dev *dev,
835
			   u32 qp_num,
836 837 838 839 840 841 842 843 844 845 846 847 848 849 850
			   union ib_flow_spec *ib_spec,
			   struct _rule_hw *mlx4_spec)
{
	enum mlx4_net_trans_rule_id type;

	switch (ib_spec->type) {
	case IB_FLOW_SPEC_ETH:
		type = MLX4_NET_TRANS_RULE_ID_ETH;
		memcpy(mlx4_spec->eth.dst_mac, ib_spec->eth.val.dst_mac,
		       ETH_ALEN);
		memcpy(mlx4_spec->eth.dst_mac_msk, ib_spec->eth.mask.dst_mac,
		       ETH_ALEN);
		mlx4_spec->eth.vlan_tag = ib_spec->eth.val.vlan_tag;
		mlx4_spec->eth.vlan_tag_msk = ib_spec->eth.mask.vlan_tag;
		break;
851 852 853 854 855 856 857 858
	case IB_FLOW_SPEC_IB:
		type = MLX4_NET_TRANS_RULE_ID_IB;
		mlx4_spec->ib.l3_qpn =
			cpu_to_be32(qp_num);
		mlx4_spec->ib.qpn_mask =
			cpu_to_be32(MLX4_IB_FLOW_QPN_MASK);
		break;

859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889

	case IB_FLOW_SPEC_IPV4:
		type = MLX4_NET_TRANS_RULE_ID_IPV4;
		mlx4_spec->ipv4.src_ip = ib_spec->ipv4.val.src_ip;
		mlx4_spec->ipv4.src_ip_msk = ib_spec->ipv4.mask.src_ip;
		mlx4_spec->ipv4.dst_ip = ib_spec->ipv4.val.dst_ip;
		mlx4_spec->ipv4.dst_ip_msk = ib_spec->ipv4.mask.dst_ip;
		break;

	case IB_FLOW_SPEC_TCP:
	case IB_FLOW_SPEC_UDP:
		type = ib_spec->type == IB_FLOW_SPEC_TCP ?
					MLX4_NET_TRANS_RULE_ID_TCP :
					MLX4_NET_TRANS_RULE_ID_UDP;
		mlx4_spec->tcp_udp.dst_port = ib_spec->tcp_udp.val.dst_port;
		mlx4_spec->tcp_udp.dst_port_msk = ib_spec->tcp_udp.mask.dst_port;
		mlx4_spec->tcp_udp.src_port = ib_spec->tcp_udp.val.src_port;
		mlx4_spec->tcp_udp.src_port_msk = ib_spec->tcp_udp.mask.src_port;
		break;

	default:
		return -EINVAL;
	}
	if (mlx4_map_sw_to_hw_steering_id(dev, type) < 0 ||
	    mlx4_hw_rule_sz(dev, type) < 0)
		return -EINVAL;
	mlx4_spec->id = cpu_to_be16(mlx4_map_sw_to_hw_steering_id(dev, type));
	mlx4_spec->size = mlx4_hw_rule_sz(dev, type) >> 2;
	return mlx4_hw_rule_sz(dev, type);
}

890 891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912
struct default_rules {
	__u32 mandatory_fields[IB_FLOW_SPEC_SUPPORT_LAYERS];
	__u32 mandatory_not_fields[IB_FLOW_SPEC_SUPPORT_LAYERS];
	__u32 rules_create_list[IB_FLOW_SPEC_SUPPORT_LAYERS];
	__u8  link_layer;
};
static const struct default_rules default_table[] = {
	{
		.mandatory_fields = {IB_FLOW_SPEC_IPV4},
		.mandatory_not_fields = {IB_FLOW_SPEC_ETH},
		.rules_create_list = {IB_FLOW_SPEC_IB},
		.link_layer = IB_LINK_LAYER_INFINIBAND
	}
};

static int __mlx4_ib_default_rules_match(struct ib_qp *qp,
					 struct ib_flow_attr *flow_attr)
{
	int i, j, k;
	void *ib_flow;
	const struct default_rules *pdefault_rules = default_table;
	u8 link_layer = rdma_port_get_link_layer(qp->device, flow_attr->port);

913
	for (i = 0; i < ARRAY_SIZE(default_table); i++, pdefault_rules++) {
914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966
		__u32 field_types[IB_FLOW_SPEC_SUPPORT_LAYERS];
		memset(&field_types, 0, sizeof(field_types));

		if (link_layer != pdefault_rules->link_layer)
			continue;

		ib_flow = flow_attr + 1;
		/* we assume the specs are sorted */
		for (j = 0, k = 0; k < IB_FLOW_SPEC_SUPPORT_LAYERS &&
		     j < flow_attr->num_of_specs; k++) {
			union ib_flow_spec *current_flow =
				(union ib_flow_spec *)ib_flow;

			/* same layer but different type */
			if (((current_flow->type & IB_FLOW_SPEC_LAYER_MASK) ==
			     (pdefault_rules->mandatory_fields[k] &
			      IB_FLOW_SPEC_LAYER_MASK)) &&
			    (current_flow->type !=
			     pdefault_rules->mandatory_fields[k]))
				goto out;

			/* same layer, try match next one */
			if (current_flow->type ==
			    pdefault_rules->mandatory_fields[k]) {
				j++;
				ib_flow +=
					((union ib_flow_spec *)ib_flow)->size;
			}
		}

		ib_flow = flow_attr + 1;
		for (j = 0; j < flow_attr->num_of_specs;
		     j++, ib_flow += ((union ib_flow_spec *)ib_flow)->size)
			for (k = 0; k < IB_FLOW_SPEC_SUPPORT_LAYERS; k++)
				/* same layer and same type */
				if (((union ib_flow_spec *)ib_flow)->type ==
				    pdefault_rules->mandatory_not_fields[k])
					goto out;

		return i;
	}
out:
	return -1;
}

static int __mlx4_ib_create_default_rules(
		struct mlx4_ib_dev *mdev,
		struct ib_qp *qp,
		const struct default_rules *pdefault_rules,
		struct _rule_hw *mlx4_spec) {
	int size = 0;
	int i;

967
	for (i = 0; i < ARRAY_SIZE(pdefault_rules->rules_create_list); i++) {
968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 992 993 994 995 996
		int ret;
		union ib_flow_spec ib_spec;
		switch (pdefault_rules->rules_create_list[i]) {
		case 0:
			/* no rule */
			continue;
		case IB_FLOW_SPEC_IB:
			ib_spec.type = IB_FLOW_SPEC_IB;
			ib_spec.size = sizeof(struct ib_flow_spec_ib);

			break;
		default:
			/* invalid rule */
			return -EINVAL;
		}
		/* We must put empty rule, qpn is being ignored */
		ret = parse_flow_attr(mdev->dev, 0, &ib_spec,
				      mlx4_spec);
		if (ret < 0) {
			pr_info("invalid parsing\n");
			return -EINVAL;
		}

		mlx4_spec = (void *)mlx4_spec + ret;
		size += ret;
	}
	return size;
}

997 998 999 1000 1001 1002 1003 1004 1005 1006 1007
static int __mlx4_ib_create_flow(struct ib_qp *qp, struct ib_flow_attr *flow_attr,
			  int domain,
			  enum mlx4_net_trans_promisc_mode flow_type,
			  u64 *reg_id)
{
	int ret, i;
	int size = 0;
	void *ib_flow;
	struct mlx4_ib_dev *mdev = to_mdev(qp->device);
	struct mlx4_cmd_mailbox *mailbox;
	struct mlx4_net_trans_rule_hw_ctrl *ctrl;
1008
	int default_flow;
1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042

	static const u16 __mlx4_domain[] = {
		[IB_FLOW_DOMAIN_USER] = MLX4_DOMAIN_UVERBS,
		[IB_FLOW_DOMAIN_ETHTOOL] = MLX4_DOMAIN_ETHTOOL,
		[IB_FLOW_DOMAIN_RFS] = MLX4_DOMAIN_RFS,
		[IB_FLOW_DOMAIN_NIC] = MLX4_DOMAIN_NIC,
	};

	if (flow_attr->priority > MLX4_IB_FLOW_MAX_PRIO) {
		pr_err("Invalid priority value %d\n", flow_attr->priority);
		return -EINVAL;
	}

	if (domain >= IB_FLOW_DOMAIN_NUM) {
		pr_err("Invalid domain value %d\n", domain);
		return -EINVAL;
	}

	if (mlx4_map_sw_to_hw_steering_mode(mdev->dev, flow_type) < 0)
		return -EINVAL;

	mailbox = mlx4_alloc_cmd_mailbox(mdev->dev);
	if (IS_ERR(mailbox))
		return PTR_ERR(mailbox);
	ctrl = mailbox->buf;

	ctrl->prio = cpu_to_be16(__mlx4_domain[domain] |
				 flow_attr->priority);
	ctrl->type = mlx4_map_sw_to_hw_steering_mode(mdev->dev, flow_type);
	ctrl->port = flow_attr->port;
	ctrl->qpn = cpu_to_be32(qp->qp_num);

	ib_flow = flow_attr + 1;
	size += sizeof(struct mlx4_net_trans_rule_hw_ctrl);
1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054
	/* Add default flows */
	default_flow = __mlx4_ib_default_rules_match(qp, flow_attr);
	if (default_flow >= 0) {
		ret = __mlx4_ib_create_default_rules(
				mdev, qp, default_table + default_flow,
				mailbox->buf + size);
		if (ret < 0) {
			mlx4_free_cmd_mailbox(mdev->dev, mailbox);
			return -EINVAL;
		}
		size += ret;
	}
1055
	for (i = 0; i < flow_attr->num_of_specs; i++) {
1056 1057
		ret = parse_flow_attr(mdev->dev, qp->qp_num, ib_flow,
				      mailbox->buf + size);
1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097 1098 1099 1100 1101 1102 1103 1104 1105 1106 1107 1108 1109 1110 1111 1112 1113 1114 1115 1116 1117 1118 1119 1120 1121 1122 1123 1124 1125 1126 1127 1128 1129 1130 1131 1132 1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 1143 1144 1145 1146 1147 1148 1149 1150 1151 1152 1153 1154 1155 1156 1157 1158 1159 1160 1161 1162 1163
		if (ret < 0) {
			mlx4_free_cmd_mailbox(mdev->dev, mailbox);
			return -EINVAL;
		}
		ib_flow += ((union ib_flow_spec *) ib_flow)->size;
		size += ret;
	}

	ret = mlx4_cmd_imm(mdev->dev, mailbox->dma, reg_id, size >> 2, 0,
			   MLX4_QP_FLOW_STEERING_ATTACH, MLX4_CMD_TIME_CLASS_A,
			   MLX4_CMD_NATIVE);
	if (ret == -ENOMEM)
		pr_err("mcg table is full. Fail to register network rule.\n");
	else if (ret == -ENXIO)
		pr_err("Device managed flow steering is disabled. Fail to register network rule.\n");
	else if (ret)
		pr_err("Invalid argumant. Fail to register network rule.\n");

	mlx4_free_cmd_mailbox(mdev->dev, mailbox);
	return ret;
}

static int __mlx4_ib_destroy_flow(struct mlx4_dev *dev, u64 reg_id)
{
	int err;
	err = mlx4_cmd(dev, reg_id, 0, 0,
		       MLX4_QP_FLOW_STEERING_DETACH, MLX4_CMD_TIME_CLASS_A,
		       MLX4_CMD_NATIVE);
	if (err)
		pr_err("Fail to detach network rule. registration id = 0x%llx\n",
		       reg_id);
	return err;
}

static struct ib_flow *mlx4_ib_create_flow(struct ib_qp *qp,
				    struct ib_flow_attr *flow_attr,
				    int domain)
{
	int err = 0, i = 0;
	struct mlx4_ib_flow *mflow;
	enum mlx4_net_trans_promisc_mode type[2];

	memset(type, 0, sizeof(type));

	mflow = kzalloc(sizeof(*mflow), GFP_KERNEL);
	if (!mflow) {
		err = -ENOMEM;
		goto err_free;
	}

	switch (flow_attr->type) {
	case IB_FLOW_ATTR_NORMAL:
		type[0] = MLX4_FS_REGULAR;
		break;

	case IB_FLOW_ATTR_ALL_DEFAULT:
		type[0] = MLX4_FS_ALL_DEFAULT;
		break;

	case IB_FLOW_ATTR_MC_DEFAULT:
		type[0] = MLX4_FS_MC_DEFAULT;
		break;

	case IB_FLOW_ATTR_SNIFFER:
		type[0] = MLX4_FS_UC_SNIFFER;
		type[1] = MLX4_FS_MC_SNIFFER;
		break;

	default:
		err = -EINVAL;
		goto err_free;
	}

	while (i < ARRAY_SIZE(type) && type[i]) {
		err = __mlx4_ib_create_flow(qp, flow_attr, domain, type[i],
					    &mflow->reg_id[i]);
		if (err)
			goto err_free;
		i++;
	}

	return &mflow->ibflow;

err_free:
	kfree(mflow);
	return ERR_PTR(err);
}

static int mlx4_ib_destroy_flow(struct ib_flow *flow_id)
{
	int err, ret = 0;
	int i = 0;
	struct mlx4_ib_dev *mdev = to_mdev(flow_id->qp->device);
	struct mlx4_ib_flow *mflow = to_mflow(flow_id);

	while (i < ARRAY_SIZE(mflow->reg_id) && mflow->reg_id[i]) {
		err = __mlx4_ib_destroy_flow(mdev->dev, mflow->reg_id[i]);
		if (err)
			ret = err;
		i++;
	}

	kfree(mflow);
	return ret;
}

1164 1165
static int mlx4_ib_mcg_attach(struct ib_qp *ibqp, union ib_gid *gid, u16 lid)
{
E
Eli Cohen 已提交
1166 1167 1168
	int err;
	struct mlx4_ib_dev *mdev = to_mdev(ibqp->device);
	struct mlx4_ib_qp *mqp = to_mqp(ibqp);
1169 1170
	u64 reg_id;
	struct mlx4_ib_steering *ib_steering = NULL;
1171 1172
	enum mlx4_protocol prot = (gid->raw[1] == 0x0e) ?
		MLX4_PROT_IB_IPV4 : MLX4_PROT_IB_IPV6;
1173 1174 1175 1176 1177 1178 1179

	if (mdev->dev->caps.steering_mode ==
	    MLX4_STEERING_MODE_DEVICE_MANAGED) {
		ib_steering = kmalloc(sizeof(*ib_steering), GFP_KERNEL);
		if (!ib_steering)
			return -ENOMEM;
	}
E
Eli Cohen 已提交
1180

1181 1182 1183
	err = mlx4_multicast_attach(mdev->dev, &mqp->mqp, gid->raw, mqp->port,
				    !!(mqp->flags &
				       MLX4_IB_QP_BLOCK_MULTICAST_LOOPBACK),
1184
				    prot, &reg_id);
E
Eli Cohen 已提交
1185
	if (err)
1186
		goto err_malloc;
E
Eli Cohen 已提交
1187 1188 1189 1190 1191

	err = add_gid_entry(ibqp, gid);
	if (err)
		goto err_add;

1192 1193 1194 1195 1196 1197 1198
	if (ib_steering) {
		memcpy(ib_steering->gid.raw, gid->raw, 16);
		ib_steering->reg_id = reg_id;
		mutex_lock(&mqp->mutex);
		list_add(&ib_steering->list, &mqp->steering_rules);
		mutex_unlock(&mqp->mutex);
	}
E
Eli Cohen 已提交
1199 1200 1201
	return 0;

err_add:
1202
	mlx4_multicast_detach(mdev->dev, &mqp->mqp, gid->raw,
1203
			      prot, reg_id);
1204 1205 1206
err_malloc:
	kfree(ib_steering);

E
Eli Cohen 已提交
1207 1208 1209 1210 1211 1212 1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 1223
	return err;
}

static struct mlx4_ib_gid_entry *find_gid_entry(struct mlx4_ib_qp *qp, u8 *raw)
{
	struct mlx4_ib_gid_entry *ge;
	struct mlx4_ib_gid_entry *tmp;
	struct mlx4_ib_gid_entry *ret = NULL;

	list_for_each_entry_safe(ge, tmp, &qp->gid_list, list) {
		if (!memcmp(raw, ge->gid.raw, 16)) {
			ret = ge;
			break;
		}
	}

	return ret;
1224 1225 1226 1227
}

static int mlx4_ib_mcg_detach(struct ib_qp *ibqp, union ib_gid *gid, u16 lid)
{
E
Eli Cohen 已提交
1228 1229 1230 1231 1232
	int err;
	struct mlx4_ib_dev *mdev = to_mdev(ibqp->device);
	struct mlx4_ib_qp *mqp = to_mqp(ibqp);
	struct net_device *ndev;
	struct mlx4_ib_gid_entry *ge;
1233
	u64 reg_id = 0;
1234 1235
	enum mlx4_protocol prot = (gid->raw[1] == 0x0e) ?
		MLX4_PROT_IB_IPV4 : MLX4_PROT_IB_IPV6;
1236 1237 1238 1239 1240 1241 1242 1243 1244 1245 1246 1247 1248 1249 1250 1251 1252 1253 1254 1255

	if (mdev->dev->caps.steering_mode ==
	    MLX4_STEERING_MODE_DEVICE_MANAGED) {
		struct mlx4_ib_steering *ib_steering;

		mutex_lock(&mqp->mutex);
		list_for_each_entry(ib_steering, &mqp->steering_rules, list) {
			if (!memcmp(ib_steering->gid.raw, gid->raw, 16)) {
				list_del(&ib_steering->list);
				break;
			}
		}
		mutex_unlock(&mqp->mutex);
		if (&ib_steering->list == &mqp->steering_rules) {
			pr_err("Couldn't find reg_id for mgid. Steering rule is left attached\n");
			return -EINVAL;
		}
		reg_id = ib_steering->reg_id;
		kfree(ib_steering);
	}
E
Eli Cohen 已提交
1256

1257
	err = mlx4_multicast_detach(mdev->dev, &mqp->mqp, gid->raw,
1258
				    prot, reg_id);
E
Eli Cohen 已提交
1259 1260 1261 1262 1263 1264 1265 1266 1267 1268 1269
	if (err)
		return err;

	mutex_lock(&mqp->mutex);
	ge = find_gid_entry(mqp, gid->raw);
	if (ge) {
		spin_lock(&mdev->iboe.lock);
		ndev = ge->added ? mdev->iboe.netdevs[ge->port - 1] : NULL;
		if (ndev)
			dev_hold(ndev);
		spin_unlock(&mdev->iboe.lock);
1270
		if (ndev)
E
Eli Cohen 已提交
1271 1272 1273 1274
			dev_put(ndev);
		list_del(&ge->list);
		kfree(ge);
	} else
1275
		pr_warn("could not find mgid entry\n");
E
Eli Cohen 已提交
1276 1277 1278 1279

	mutex_unlock(&mqp->mutex);

	return 0;
1280 1281 1282 1283 1284 1285
}

static int init_node_data(struct mlx4_ib_dev *dev)
{
	struct ib_smp *in_mad  = NULL;
	struct ib_smp *out_mad = NULL;
1286
	int mad_ifc_flags = MLX4_MAD_IFC_IGNORE_KEYS;
1287 1288 1289 1290 1291 1292 1293 1294 1295
	int err = -ENOMEM;

	in_mad  = kzalloc(sizeof *in_mad, GFP_KERNEL);
	out_mad = kmalloc(sizeof *out_mad, GFP_KERNEL);
	if (!in_mad || !out_mad)
		goto out;

	init_query_mad(in_mad);
	in_mad->attr_id = IB_SMP_ATTR_NODE_DESC;
1296 1297
	if (mlx4_is_master(dev->dev))
		mad_ifc_flags |= MLX4_MAD_IFC_NET_VIEW;
1298

1299
	err = mlx4_MAD_IFC(dev, mad_ifc_flags, 1, NULL, NULL, in_mad, out_mad);
1300 1301 1302 1303 1304 1305 1306
	if (err)
		goto out;

	memcpy(dev->ib_dev.node_desc, out_mad->data, 64);

	in_mad->attr_id = IB_SMP_ATTR_NODE_INFO;

1307
	err = mlx4_MAD_IFC(dev, mad_ifc_flags, 1, NULL, NULL, in_mad, out_mad);
1308 1309 1310
	if (err)
		goto out;

1311
	dev->dev->rev_id = be32_to_cpup((__be32 *) (out_mad->data + 32));
1312 1313 1314 1315 1316 1317 1318 1319
	memcpy(&dev->ib_dev.node_guid, out_mad->data + 12, 8);

out:
	kfree(in_mad);
	kfree(out_mad);
	return err;
}

1320 1321
static ssize_t show_hca(struct device *device, struct device_attribute *attr,
			char *buf)
1322
{
1323 1324
	struct mlx4_ib_dev *dev =
		container_of(device, struct mlx4_ib_dev, ib_dev.dev);
1325 1326 1327
	return sprintf(buf, "MT%d\n", dev->dev->pdev->device);
}

1328 1329
static ssize_t show_fw_ver(struct device *device, struct device_attribute *attr,
			   char *buf)
1330
{
1331 1332
	struct mlx4_ib_dev *dev =
		container_of(device, struct mlx4_ib_dev, ib_dev.dev);
1333 1334 1335 1336 1337
	return sprintf(buf, "%d.%d.%d\n", (int) (dev->dev->caps.fw_ver >> 32),
		       (int) (dev->dev->caps.fw_ver >> 16) & 0xffff,
		       (int) dev->dev->caps.fw_ver & 0xffff);
}

1338 1339
static ssize_t show_rev(struct device *device, struct device_attribute *attr,
			char *buf)
1340
{
1341 1342
	struct mlx4_ib_dev *dev =
		container_of(device, struct mlx4_ib_dev, ib_dev.dev);
1343 1344 1345
	return sprintf(buf, "%x\n", dev->dev->rev_id);
}

1346 1347
static ssize_t show_board(struct device *device, struct device_attribute *attr,
			  char *buf)
1348
{
1349 1350 1351 1352
	struct mlx4_ib_dev *dev =
		container_of(device, struct mlx4_ib_dev, ib_dev.dev);
	return sprintf(buf, "%.*s\n", MLX4_BOARD_ID_LEN,
		       dev->dev->board_id);
1353 1354
}

1355 1356 1357 1358
static DEVICE_ATTR(hw_rev,   S_IRUGO, show_rev,    NULL);
static DEVICE_ATTR(fw_ver,   S_IRUGO, show_fw_ver, NULL);
static DEVICE_ATTR(hca_type, S_IRUGO, show_hca,    NULL);
static DEVICE_ATTR(board_id, S_IRUGO, show_board,  NULL);
1359

1360 1361 1362 1363 1364
static struct device_attribute *mlx4_class_attributes[] = {
	&dev_attr_hw_rev,
	&dev_attr_fw_ver,
	&dev_attr_hca_type,
	&dev_attr_board_id
1365 1366
};

1367 1368 1369 1370 1371 1372 1373 1374 1375 1376 1377 1378 1379 1380 1381
static void mlx4_addrconf_ifid_eui48(u8 *eui, u16 vlan_id,
				     struct net_device *dev)
{
	memcpy(eui, dev->dev_addr, 3);
	memcpy(eui + 5, dev->dev_addr + 3, 3);
	if (vlan_id < 0x1000) {
		eui[3] = vlan_id >> 8;
		eui[4] = vlan_id & 0xff;
	} else {
		eui[3] = 0xff;
		eui[4] = 0xfe;
	}
	eui[0] ^= 2;
}

E
Eli Cohen 已提交
1382 1383 1384 1385 1386 1387 1388 1389 1390 1391
static void update_gids_task(struct work_struct *work)
{
	struct update_gid_work *gw = container_of(work, struct update_gid_work, work);
	struct mlx4_cmd_mailbox *mailbox;
	union ib_gid *gids;
	int err;
	struct mlx4_dev	*dev = gw->dev->dev;

	mailbox = mlx4_alloc_cmd_mailbox(dev);
	if (IS_ERR(mailbox)) {
1392
		pr_warn("update gid table failed %ld\n", PTR_ERR(mailbox));
E
Eli Cohen 已提交
1393 1394 1395 1396 1397 1398 1399
		return;
	}

	gids = mailbox->buf;
	memcpy(gids, gw->gids, sizeof gw->gids);

	err = mlx4_cmd(dev, mailbox->dma, MLX4_SET_PORT_GID_TABLE << 8 | gw->port,
1400
		       1, MLX4_CMD_SET_PORT, MLX4_CMD_TIME_CLASS_B,
1401
		       MLX4_CMD_WRAPPED);
E
Eli Cohen 已提交
1402
	if (err)
1403
		pr_warn("set port command failed\n");
1404
	else
1405
		mlx4_ib_dispatch_event(gw->dev, gw->port, IB_EVENT_GID_CHANGE);
E
Eli Cohen 已提交
1406 1407 1408 1409 1410

	mlx4_free_cmd_mailbox(dev, mailbox);
	kfree(gw);
}

1411
static void reset_gids_task(struct work_struct *work)
E
Eli Cohen 已提交
1412
{
1413 1414 1415 1416 1417 1418
	struct update_gid_work *gw =
			container_of(work, struct update_gid_work, work);
	struct mlx4_cmd_mailbox *mailbox;
	union ib_gid *gids;
	int err;
	struct mlx4_dev	*dev = gw->dev->dev;
E
Eli Cohen 已提交
1419

1420 1421 1422 1423 1424
	mailbox = mlx4_alloc_cmd_mailbox(dev);
	if (IS_ERR(mailbox)) {
		pr_warn("reset gid table failed\n");
		goto free;
	}
E
Eli Cohen 已提交
1425

1426 1427 1428
	gids = mailbox->buf;
	memcpy(gids, gw->gids, sizeof(gw->gids));

1429 1430 1431 1432 1433 1434 1435 1436 1437 1438
	if (mlx4_ib_port_link_layer(&gw->dev->ib_dev, gw->port) ==
				    IB_LINK_LAYER_ETHERNET) {
		err = mlx4_cmd(dev, mailbox->dma,
			       MLX4_SET_PORT_GID_TABLE << 8 | gw->port,
			       1, MLX4_CMD_SET_PORT,
			       MLX4_CMD_TIME_CLASS_B,
			       MLX4_CMD_WRAPPED);
		if (err)
			pr_warn(KERN_WARNING
				"set port %d command failed\n", gw->port);
E
Eli Cohen 已提交
1439 1440
	}

1441 1442 1443 1444
	mlx4_free_cmd_mailbox(dev, mailbox);
free:
	kfree(gw);
}
E
Eli Cohen 已提交
1445

1446
static int update_gid_table(struct mlx4_ib_dev *dev, int port,
1447 1448
			    union ib_gid *gid, int clear,
			    int default_gid)
1449 1450 1451 1452 1453 1454 1455 1456
{
	struct update_gid_work *work;
	int i;
	int need_update = 0;
	int free = -1;
	int found = -1;
	int max_gids;

1457 1458 1459 1460 1461 1462
	if (default_gid) {
		free = 0;
	} else {
		max_gids = dev->dev->caps.gid_table_len[port];
		for (i = 1; i < max_gids; ++i) {
			if (!memcmp(&dev->iboe.gid_table[port - 1][i], gid,
1463
				    sizeof(*gid)))
1464 1465 1466 1467 1468 1469 1470 1471 1472 1473 1474 1475 1476 1477 1478 1479 1480 1481
				found = i;

			if (clear) {
				if (found >= 0) {
					need_update = 1;
					dev->iboe.gid_table[port - 1][found] =
						zgid;
					break;
				}
			} else {
				if (found >= 0)
					break;

				if (free < 0 &&
				    !memcmp(&dev->iboe.gid_table[port - 1][i],
					    &zgid, sizeof(*gid)))
					free = i;
			}
E
Eli Cohen 已提交
1482
		}
E
Eli Cohen 已提交
1483
	}
E
Eli Cohen 已提交
1484

1485 1486 1487 1488
	if (found == -1 && !clear && free >= 0) {
		dev->iboe.gid_table[port - 1][free] = *gid;
		need_update = 1;
	}
E
Eli Cohen 已提交
1489

1490 1491 1492 1493 1494 1495 1496 1497 1498 1499 1500 1501
	if (!need_update)
		return 0;

	work = kzalloc(sizeof(*work), GFP_ATOMIC);
	if (!work)
		return -ENOMEM;

	memcpy(work->gids, dev->iboe.gid_table[port - 1], sizeof(work->gids));
	INIT_WORK(&work->work, update_gids_task);
	work->port = port;
	work->dev = dev;
	queue_work(wq, &work->work);
E
Eli Cohen 已提交
1502 1503

	return 0;
1504
}
E
Eli Cohen 已提交
1505

1506
static void mlx4_make_default_gid(struct  net_device *dev, union ib_gid *gid)
1507
{
1508 1509 1510 1511
	gid->global.subnet_prefix = cpu_to_be64(0xfe80000000000000LL);
	mlx4_addrconf_ifid_eui48(&gid->raw[8], 0xffff, dev);
}

1512

1513
static int reset_gid_table(struct mlx4_ib_dev *dev, u8 port)
1514 1515 1516 1517 1518 1519
{
	struct update_gid_work *work;

	work = kzalloc(sizeof(*work), GFP_ATOMIC);
	if (!work)
		return -ENOMEM;
1520 1521

	memset(dev->iboe.gid_table[port - 1], 0, sizeof(work->gids));
1522 1523 1524
	memset(work->gids, 0, sizeof(work->gids));
	INIT_WORK(&work->work, reset_gids_task);
	work->dev = dev;
1525
	work->port = port;
1526 1527
	queue_work(wq, &work->work);
	return 0;
E
Eli Cohen 已提交
1528 1529
}

1530 1531
static int mlx4_ib_addr_event(int event, struct net_device *event_netdev,
			      struct mlx4_ib_dev *ibdev, union ib_gid *gid)
E
Eli Cohen 已提交
1532
{
1533 1534 1535 1536 1537
	struct mlx4_ib_iboe *iboe;
	int port = 0;
	struct net_device *real_dev = rdma_vlan_dev_real_dev(event_netdev) ?
				rdma_vlan_dev_real_dev(event_netdev) :
				event_netdev;
1538 1539 1540 1541 1542 1543
	union ib_gid default_gid;

	mlx4_make_default_gid(real_dev, &default_gid);

	if (!memcmp(gid, &default_gid, sizeof(*gid)))
		return 0;
1544 1545 1546 1547 1548 1549 1550 1551 1552 1553 1554 1555

	if (event != NETDEV_DOWN && event != NETDEV_UP)
		return 0;

	if ((real_dev != event_netdev) &&
	    (event == NETDEV_DOWN) &&
	    rdma_link_local_addr((struct in6_addr *)gid))
		return 0;

	iboe = &ibdev->iboe;
	spin_lock(&iboe->lock);

1556
	for (port = 1; port <= ibdev->dev->caps.num_ports; ++port)
1557 1558 1559 1560 1561
		if ((netif_is_bond_master(real_dev) &&
		     (real_dev == iboe->masters[port - 1])) ||
		     (!netif_is_bond_master(real_dev) &&
		     (real_dev == iboe->netdevs[port - 1])))
			update_gid_table(ibdev, port, gid,
1562
					 event == NETDEV_DOWN, 0);
1563 1564 1565

	spin_unlock(&iboe->lock);
	return 0;
E
Eli Cohen 已提交
1566 1567 1568

}

1569 1570
static u8 mlx4_ib_get_dev_port(struct net_device *dev,
			       struct mlx4_ib_dev *ibdev)
E
Eli Cohen 已提交
1571
{
1572 1573 1574 1575 1576 1577 1578
	u8 port = 0;
	struct mlx4_ib_iboe *iboe;
	struct net_device *real_dev = rdma_vlan_dev_real_dev(dev) ?
				rdma_vlan_dev_real_dev(dev) : dev;

	iboe = &ibdev->iboe;

1579
	for (port = 1; port <= ibdev->dev->caps.num_ports; ++port)
1580 1581 1582 1583 1584 1585
		if ((netif_is_bond_master(real_dev) &&
		     (real_dev == iboe->masters[port - 1])) ||
		     (!netif_is_bond_master(real_dev) &&
		     (real_dev == iboe->netdevs[port - 1])))
			break;

1586
	if ((port == 0) || (port > ibdev->dev->caps.num_ports))
1587 1588 1589
		return 0;
	else
		return port;
E
Eli Cohen 已提交
1590 1591
}

1592 1593
static int mlx4_ib_inet_event(struct notifier_block *this, unsigned long event,
				void *ptr)
E
Eli Cohen 已提交
1594
{
1595 1596 1597 1598 1599 1600 1601 1602 1603 1604 1605
	struct mlx4_ib_dev *ibdev;
	struct in_ifaddr *ifa = ptr;
	union ib_gid gid;
	struct net_device *event_netdev = ifa->ifa_dev->dev;

	ipv6_addr_set_v4mapped(ifa->ifa_address, (struct in6_addr *)&gid);

	ibdev = container_of(this, struct mlx4_ib_dev, iboe.nb_inet);

	mlx4_ib_addr_event(event, event_netdev, ibdev, &gid);
	return NOTIFY_DONE;
E
Eli Cohen 已提交
1606 1607
}

1608
#if IS_ENABLED(CONFIG_IPV6)
1609
static int mlx4_ib_inet6_event(struct notifier_block *this, unsigned long event,
E
Eli Cohen 已提交
1610 1611 1612
				void *ptr)
{
	struct mlx4_ib_dev *ibdev;
1613 1614 1615 1616 1617 1618 1619 1620 1621 1622 1623
	struct inet6_ifaddr *ifa = ptr;
	union  ib_gid *gid = (union ib_gid *)&ifa->addr;
	struct net_device *event_netdev = ifa->idev->dev;

	ibdev = container_of(this, struct mlx4_ib_dev, iboe.nb_inet6);

	mlx4_ib_addr_event(event, event_netdev, ibdev, gid);
	return NOTIFY_DONE;
}
#endif

1624 1625 1626 1627 1628 1629 1630 1631 1632 1633 1634 1635 1636 1637 1638 1639 1640 1641 1642 1643 1644 1645 1646 1647 1648 1649 1650 1651 1652 1653 1654 1655 1656 1657 1658 1659 1660 1661 1662 1663 1664 1665 1666 1667 1668 1669 1670
#define MLX4_IB_INVALID_MAC	((u64)-1)
static void mlx4_ib_update_qps(struct mlx4_ib_dev *ibdev,
			       struct net_device *dev,
			       int port)
{
	u64 new_smac = 0;
	u64 release_mac = MLX4_IB_INVALID_MAC;
	struct mlx4_ib_qp *qp;

	read_lock(&dev_base_lock);
	new_smac = mlx4_mac_to_u64(dev->dev_addr);
	read_unlock(&dev_base_lock);

	mutex_lock(&ibdev->qp1_proxy_lock[port - 1]);
	qp = ibdev->qp1_proxy[port - 1];
	if (qp) {
		int new_smac_index;
		u64 old_smac = qp->pri.smac;
		struct mlx4_update_qp_params update_params;

		if (new_smac == old_smac)
			goto unlock;

		new_smac_index = mlx4_register_mac(ibdev->dev, port, new_smac);

		if (new_smac_index < 0)
			goto unlock;

		update_params.smac_index = new_smac_index;
		if (mlx4_update_qp(ibdev->dev, &qp->mqp, MLX4_UPDATE_QP_SMAC,
				   &update_params)) {
			release_mac = new_smac;
			goto unlock;
		}

		qp->pri.smac = new_smac;
		qp->pri.smac_index = new_smac_index;

		release_mac = old_smac;
	}

unlock:
	mutex_unlock(&ibdev->qp1_proxy_lock[port - 1]);
	if (release_mac != MLX4_IB_INVALID_MAC)
		mlx4_unregister_mac(ibdev->dev, port, release_mac);
}

1671 1672 1673 1674
static void mlx4_ib_get_dev_addr(struct net_device *dev,
				 struct mlx4_ib_dev *ibdev, u8 port)
{
	struct in_device *in_dev;
1675
#if IS_ENABLED(CONFIG_IPV6)
1676 1677 1678 1679 1680 1681 1682
	struct inet6_dev *in6_dev;
	union ib_gid  *pgid;
	struct inet6_ifaddr *ifp;
#endif
	union ib_gid gid;


1683
	if ((port == 0) || (port > ibdev->dev->caps.num_ports))
1684 1685 1686 1687 1688 1689 1690 1691 1692
		return;

	/* IPv4 gids */
	in_dev = in_dev_get(dev);
	if (in_dev) {
		for_ifa(in_dev) {
			/*ifa->ifa_address;*/
			ipv6_addr_set_v4mapped(ifa->ifa_address,
					       (struct in6_addr *)&gid);
1693
			update_gid_table(ibdev, port, &gid, 0, 0);
1694 1695 1696 1697
		}
		endfor_ifa(in_dev);
		in_dev_put(in_dev);
	}
1698
#if IS_ENABLED(CONFIG_IPV6)
1699 1700 1701 1702 1703 1704
	/* IPv6 gids */
	in6_dev = in6_dev_get(dev);
	if (in6_dev) {
		read_lock_bh(&in6_dev->lock);
		list_for_each_entry(ifp, &in6_dev->addr_list, if_list) {
			pgid = (union ib_gid *)&ifp->addr;
1705
			update_gid_table(ibdev, port, pgid, 0, 0);
1706 1707 1708 1709 1710 1711 1712
		}
		read_unlock_bh(&in6_dev->lock);
		in6_dev_put(in6_dev);
	}
#endif
}

1713 1714 1715 1716 1717 1718 1719 1720
static void mlx4_ib_set_default_gid(struct mlx4_ib_dev *ibdev,
				 struct  net_device *dev, u8 port)
{
	union ib_gid gid;
	mlx4_make_default_gid(dev, &gid);
	update_gid_table(ibdev, port, &gid, 0, 1);
}

1721 1722 1723
static int mlx4_ib_init_gid_table(struct mlx4_ib_dev *ibdev)
{
	struct	net_device *dev;
1724
	struct mlx4_ib_iboe *iboe = &ibdev->iboe;
1725
	int i;
1726

1727 1728 1729
	for (i = 1; i <= ibdev->num_ports; ++i)
		if (reset_gid_table(ibdev, i))
			return -1;
1730 1731

	read_lock(&dev_base_lock);
1732
	spin_lock(&iboe->lock);
1733 1734 1735 1736 1737 1738 1739

	for_each_netdev(&init_net, dev) {
		u8 port = mlx4_ib_get_dev_port(dev, ibdev);
		if (port)
			mlx4_ib_get_dev_addr(dev, ibdev, port);
	}

1740
	spin_unlock(&iboe->lock);
1741 1742 1743 1744 1745
	read_unlock(&dev_base_lock);

	return 0;
}

1746 1747 1748 1749
static void mlx4_ib_scan_netdevs(struct mlx4_ib_dev *ibdev,
				 struct net_device *dev,
				 unsigned long event)

1750
{
E
Eli Cohen 已提交
1751
	struct mlx4_ib_iboe *iboe;
1752
	int update_qps_port = -1;
E
Eli Cohen 已提交
1753 1754 1755 1756 1757 1758
	int port;

	iboe = &ibdev->iboe;

	spin_lock(&iboe->lock);
	mlx4_foreach_ib_transport_port(port, ibdev->dev) {
1759
		enum ib_port_state	port_state = IB_PORT_NOP;
1760
		struct net_device *old_master = iboe->masters[port - 1];
1761
		struct net_device *curr_netdev;
1762
		struct net_device *curr_master;
1763

E
Eli Cohen 已提交
1764
		iboe->netdevs[port - 1] =
1765
			mlx4_get_protocol_dev(ibdev->dev, MLX4_PROT_ETH, port);
1766 1767 1768
		if (iboe->netdevs[port - 1])
			mlx4_ib_set_default_gid(ibdev,
						iboe->netdevs[port - 1], port);
1769
		curr_netdev = iboe->netdevs[port - 1];
1770 1771 1772 1773 1774

		if (iboe->netdevs[port - 1] &&
		    netif_is_bond_slave(iboe->netdevs[port - 1])) {
			iboe->masters[port - 1] = netdev_master_upper_dev_get(
				iboe->netdevs[port - 1]);
1775 1776
		} else {
			iboe->masters[port - 1] = NULL;
E
Eli Cohen 已提交
1777
		}
1778
		curr_master = iboe->masters[port - 1];
E
Eli Cohen 已提交
1779

1780 1781 1782 1783 1784
		if (dev == iboe->netdevs[port - 1] &&
		    (event == NETDEV_CHANGEADDR || event == NETDEV_REGISTER ||
		     event == NETDEV_UP || event == NETDEV_CHANGE))
			update_qps_port = port;

1785 1786 1787 1788 1789 1790 1791 1792 1793 1794 1795 1796 1797 1798 1799
		if (curr_netdev) {
			port_state = (netif_running(curr_netdev) && netif_carrier_ok(curr_netdev)) ?
						IB_PORT_ACTIVE : IB_PORT_DOWN;
			mlx4_ib_set_default_gid(ibdev, curr_netdev, port);
		} else {
			reset_gid_table(ibdev, port);
		}
		/* if using bonding/team and a slave port is down, we don't the bond IP
		 * based gids in the table since flows that select port by gid may get
		 * the down port.
		 */
		if (curr_master && (port_state == IB_PORT_DOWN)) {
			reset_gid_table(ibdev, port);
			mlx4_ib_set_default_gid(ibdev, curr_netdev, port);
		}
1800
		/* if bonding is used it is possible that we add it to masters
1801 1802 1803 1804 1805 1806
		 * only after IP address is assigned to the net bonding
		 * interface.
		*/
		if (curr_master && (old_master != curr_master)) {
			reset_gid_table(ibdev, port);
			mlx4_ib_set_default_gid(ibdev, curr_netdev, port);
1807
			mlx4_ib_get_dev_addr(curr_master, ibdev, port);
1808 1809 1810 1811 1812 1813 1814
		}

		if (!curr_master && (old_master != curr_master)) {
			reset_gid_table(ibdev, port);
			mlx4_ib_set_default_gid(ibdev, curr_netdev, port);
			mlx4_ib_get_dev_addr(curr_netdev, ibdev, port);
		}
1815
	}
E
Eli Cohen 已提交
1816 1817

	spin_unlock(&iboe->lock);
1818 1819 1820

	if (update_qps_port > 0)
		mlx4_ib_update_qps(ibdev, dev, update_qps_port);
1821 1822 1823 1824 1825 1826 1827 1828 1829 1830 1831 1832
}

static int mlx4_ib_netdev_event(struct notifier_block *this,
				unsigned long event, void *ptr)
{
	struct net_device *dev = netdev_notifier_info_to_dev(ptr);
	struct mlx4_ib_dev *ibdev;

	if (!net_eq(dev_net(dev), &init_net))
		return NOTIFY_DONE;

	ibdev = container_of(this, struct mlx4_ib_dev, iboe.nb);
1833
	mlx4_ib_scan_netdevs(ibdev, dev, event);
E
Eli Cohen 已提交
1834 1835 1836 1837

	return NOTIFY_DONE;
}

1838 1839 1840 1841 1842 1843 1844 1845 1846 1847 1848 1849 1850 1851 1852 1853 1854 1855 1856 1857 1858 1859 1860 1861 1862 1863 1864 1865 1866 1867 1868 1869
static void init_pkeys(struct mlx4_ib_dev *ibdev)
{
	int port;
	int slave;
	int i;

	if (mlx4_is_master(ibdev->dev)) {
		for (slave = 0; slave <= ibdev->dev->num_vfs; ++slave) {
			for (port = 1; port <= ibdev->dev->caps.num_ports; ++port) {
				for (i = 0;
				     i < ibdev->dev->phys_caps.pkey_phys_table_len[port];
				     ++i) {
					ibdev->pkeys.virt2phys_pkey[slave][port - 1][i] =
					/* master has the identity virt2phys pkey mapping */
						(slave == mlx4_master_func_num(ibdev->dev) || !i) ? i :
							ibdev->dev->phys_caps.pkey_phys_table_len[port] - 1;
					mlx4_sync_pkey_table(ibdev->dev, slave, port, i,
							     ibdev->pkeys.virt2phys_pkey[slave][port - 1][i]);
				}
			}
		}
		/* initialize pkey cache */
		for (port = 1; port <= ibdev->dev->caps.num_ports; ++port) {
			for (i = 0;
			     i < ibdev->dev->phys_caps.pkey_phys_table_len[port];
			     ++i)
				ibdev->pkeys.phys_pkey_cache[port-1][i] =
					(i) ? 0 : 0xFFFF;
		}
	}
}

1870 1871
static void mlx4_ib_alloc_eqs(struct mlx4_dev *dev, struct mlx4_ib_dev *ibdev)
{
1872
	char name[80];
1873 1874 1875 1876 1877
	int eq_per_port = 0;
	int added_eqs = 0;
	int total_eqs = 0;
	int i, j, eq;

1878 1879 1880
	/* Legacy mode or comp_pool is not large enough */
	if (dev->caps.comp_pool == 0 ||
	    dev->caps.num_ports > dev->caps.comp_pool)
1881 1882 1883 1884 1885 1886 1887 1888 1889 1890 1891 1892 1893 1894 1895 1896 1897 1898 1899 1900 1901
		return;

	eq_per_port = rounddown_pow_of_two(dev->caps.comp_pool/
					dev->caps.num_ports);

	/* Init eq table */
	added_eqs = 0;
	mlx4_foreach_port(i, dev, MLX4_PORT_TYPE_IB)
		added_eqs += eq_per_port;

	total_eqs = dev->caps.num_comp_vectors + added_eqs;

	ibdev->eq_table = kzalloc(total_eqs * sizeof(int), GFP_KERNEL);
	if (!ibdev->eq_table)
		return;

	ibdev->eq_added = added_eqs;

	eq = 0;
	mlx4_foreach_port(i, dev, MLX4_PORT_TYPE_IB) {
		for (j = 0; j < eq_per_port; j++) {
1902 1903
			snprintf(name, sizeof(name), "mlx4-ib-%d-%d@%s",
				 i, j, dev->pdev->bus->name);
1904
			/* Set IRQ for specific name (per ring) */
1905 1906
			if (mlx4_assign_eq(dev, name, NULL,
					   &ibdev->eq_table[eq])) {
1907 1908 1909 1910 1911 1912 1913 1914 1915 1916 1917 1918 1919 1920 1921 1922 1923 1924 1925 1926
				/* Use legacy (same as mlx4_en driver) */
				pr_warn("Can't allocate EQ %d; reverting to legacy\n", eq);
				ibdev->eq_table[eq] =
					(eq % dev->caps.num_comp_vectors);
			}
			eq++;
		}
	}

	/* Fill the reset of the vector with legacy EQ */
	for (i = 0, eq = added_eqs; i < dev->caps.num_comp_vectors; i++)
		ibdev->eq_table[eq++] = i;

	/* Advertise the new number of EQs to clients */
	ibdev->ib_dev.num_comp_vectors = total_eqs;
}

static void mlx4_ib_free_eqs(struct mlx4_dev *dev, struct mlx4_ib_dev *ibdev)
{
	int i;
1927 1928 1929 1930

	/* no additional eqs were added */
	if (!ibdev->eq_table)
		return;
1931 1932 1933 1934 1935 1936 1937 1938 1939 1940 1941 1942 1943 1944 1945

	/* Reset the advertised EQ number */
	ibdev->ib_dev.num_comp_vectors = dev->caps.num_comp_vectors;

	/* Free only the added eqs */
	for (i = 0; i < ibdev->eq_added; i++) {
		/* Don't free legacy eqs if used */
		if (ibdev->eq_table[i] <= dev->caps.num_comp_vectors)
			continue;
		mlx4_release_eq(dev, ibdev->eq_table[i]);
	}

	kfree(ibdev->eq_table);
}

1946 1947 1948
static void *mlx4_ib_add(struct mlx4_dev *dev)
{
	struct mlx4_ib_dev *ibdev;
1949
	int num_ports = 0;
1950
	int i, j;
E
Eli Cohen 已提交
1951 1952
	int err;
	struct mlx4_ib_iboe *iboe;
1953
	int ib_num_ports = 0;
1954

1955
	pr_info_once("%s", mlx4_ib_version);
1956

1957
	num_ports = 0;
E
Eli Cohen 已提交
1958
	mlx4_foreach_ib_transport_port(i, dev)
1959 1960 1961 1962 1963 1964
		num_ports++;

	/* No point in registering a device with no ports... */
	if (num_ports == 0)
		return NULL;

1965 1966 1967 1968 1969 1970
	ibdev = (struct mlx4_ib_dev *) ib_alloc_device(sizeof *ibdev);
	if (!ibdev) {
		dev_err(&dev->pdev->dev, "Device struct alloc failed\n");
		return NULL;
	}

E
Eli Cohen 已提交
1971 1972
	iboe = &ibdev->iboe;

1973 1974 1975 1976 1977 1978
	if (mlx4_pd_alloc(dev, &ibdev->priv_pdn))
		goto err_dealloc;

	if (mlx4_uar_alloc(dev, &ibdev->priv_uar))
		goto err_pd;

1979 1980
	ibdev->uar_map = ioremap((phys_addr_t) ibdev->priv_uar.pfn << PAGE_SHIFT,
				 PAGE_SIZE);
1981 1982
	if (!ibdev->uar_map)
		goto err_uar;
1983
	MLX4_INIT_DOORBELL_LOCK(&ibdev->uar_lock);
1984 1985 1986 1987 1988 1989

	ibdev->dev = dev;

	strlcpy(ibdev->ib_dev.name, "mlx4_%d", IB_DEVICE_NAME_MAX);
	ibdev->ib_dev.owner		= THIS_MODULE;
	ibdev->ib_dev.node_type		= RDMA_NODE_IB_CA;
1990
	ibdev->ib_dev.local_dma_lkey	= dev->caps.reserved_lkey;
1991
	ibdev->num_ports		= num_ports;
1992
	ibdev->ib_dev.phys_port_cnt     = ibdev->num_ports;
1993
	ibdev->ib_dev.num_comp_vectors	= dev->caps.num_comp_vectors;
1994 1995
	ibdev->ib_dev.dma_device	= &dev->pdev->dev;

O
Or Gerlitz 已提交
1996 1997 1998 1999 2000
	if (dev->caps.userspace_caps)
		ibdev->ib_dev.uverbs_abi_ver = MLX4_IB_UVERBS_ABI_VERSION;
	else
		ibdev->ib_dev.uverbs_abi_ver = MLX4_IB_UVERBS_NO_DEV_CAPS_ABI_VERSION;

2001 2002 2003 2004 2005 2006 2007
	ibdev->ib_dev.uverbs_cmd_mask	=
		(1ull << IB_USER_VERBS_CMD_GET_CONTEXT)		|
		(1ull << IB_USER_VERBS_CMD_QUERY_DEVICE)	|
		(1ull << IB_USER_VERBS_CMD_QUERY_PORT)		|
		(1ull << IB_USER_VERBS_CMD_ALLOC_PD)		|
		(1ull << IB_USER_VERBS_CMD_DEALLOC_PD)		|
		(1ull << IB_USER_VERBS_CMD_REG_MR)		|
2008
		(1ull << IB_USER_VERBS_CMD_REREG_MR)		|
2009 2010 2011
		(1ull << IB_USER_VERBS_CMD_DEREG_MR)		|
		(1ull << IB_USER_VERBS_CMD_CREATE_COMP_CHANNEL)	|
		(1ull << IB_USER_VERBS_CMD_CREATE_CQ)		|
2012
		(1ull << IB_USER_VERBS_CMD_RESIZE_CQ)		|
2013 2014 2015
		(1ull << IB_USER_VERBS_CMD_DESTROY_CQ)		|
		(1ull << IB_USER_VERBS_CMD_CREATE_QP)		|
		(1ull << IB_USER_VERBS_CMD_MODIFY_QP)		|
J
Jack Morgenstein 已提交
2016
		(1ull << IB_USER_VERBS_CMD_QUERY_QP)		|
2017 2018 2019 2020 2021
		(1ull << IB_USER_VERBS_CMD_DESTROY_QP)		|
		(1ull << IB_USER_VERBS_CMD_ATTACH_MCAST)	|
		(1ull << IB_USER_VERBS_CMD_DETACH_MCAST)	|
		(1ull << IB_USER_VERBS_CMD_CREATE_SRQ)		|
		(1ull << IB_USER_VERBS_CMD_MODIFY_SRQ)		|
J
Jack Morgenstein 已提交
2022
		(1ull << IB_USER_VERBS_CMD_QUERY_SRQ)		|
S
Sean Hefty 已提交
2023
		(1ull << IB_USER_VERBS_CMD_DESTROY_SRQ)		|
2024 2025
		(1ull << IB_USER_VERBS_CMD_CREATE_XSRQ)		|
		(1ull << IB_USER_VERBS_CMD_OPEN_QP);
2026 2027 2028

	ibdev->ib_dev.query_device	= mlx4_ib_query_device;
	ibdev->ib_dev.query_port	= mlx4_ib_query_port;
E
Eli Cohen 已提交
2029
	ibdev->ib_dev.get_link_layer	= mlx4_ib_port_link_layer;
2030 2031 2032 2033 2034 2035 2036 2037 2038 2039 2040 2041 2042 2043
	ibdev->ib_dev.query_gid		= mlx4_ib_query_gid;
	ibdev->ib_dev.query_pkey	= mlx4_ib_query_pkey;
	ibdev->ib_dev.modify_device	= mlx4_ib_modify_device;
	ibdev->ib_dev.modify_port	= mlx4_ib_modify_port;
	ibdev->ib_dev.alloc_ucontext	= mlx4_ib_alloc_ucontext;
	ibdev->ib_dev.dealloc_ucontext	= mlx4_ib_dealloc_ucontext;
	ibdev->ib_dev.mmap		= mlx4_ib_mmap;
	ibdev->ib_dev.alloc_pd		= mlx4_ib_alloc_pd;
	ibdev->ib_dev.dealloc_pd	= mlx4_ib_dealloc_pd;
	ibdev->ib_dev.create_ah		= mlx4_ib_create_ah;
	ibdev->ib_dev.query_ah		= mlx4_ib_query_ah;
	ibdev->ib_dev.destroy_ah	= mlx4_ib_destroy_ah;
	ibdev->ib_dev.create_srq	= mlx4_ib_create_srq;
	ibdev->ib_dev.modify_srq	= mlx4_ib_modify_srq;
J
Jack Morgenstein 已提交
2044
	ibdev->ib_dev.query_srq		= mlx4_ib_query_srq;
2045 2046 2047 2048
	ibdev->ib_dev.destroy_srq	= mlx4_ib_destroy_srq;
	ibdev->ib_dev.post_srq_recv	= mlx4_ib_post_srq_recv;
	ibdev->ib_dev.create_qp		= mlx4_ib_create_qp;
	ibdev->ib_dev.modify_qp		= mlx4_ib_modify_qp;
J
Jack Morgenstein 已提交
2049
	ibdev->ib_dev.query_qp		= mlx4_ib_query_qp;
2050 2051 2052 2053
	ibdev->ib_dev.destroy_qp	= mlx4_ib_destroy_qp;
	ibdev->ib_dev.post_send		= mlx4_ib_post_send;
	ibdev->ib_dev.post_recv		= mlx4_ib_post_recv;
	ibdev->ib_dev.create_cq		= mlx4_ib_create_cq;
2054
	ibdev->ib_dev.modify_cq		= mlx4_ib_modify_cq;
2055
	ibdev->ib_dev.resize_cq		= mlx4_ib_resize_cq;
2056 2057 2058 2059 2060
	ibdev->ib_dev.destroy_cq	= mlx4_ib_destroy_cq;
	ibdev->ib_dev.poll_cq		= mlx4_ib_poll_cq;
	ibdev->ib_dev.req_notify_cq	= mlx4_ib_arm_cq;
	ibdev->ib_dev.get_dma_mr	= mlx4_ib_get_dma_mr;
	ibdev->ib_dev.reg_user_mr	= mlx4_ib_reg_user_mr;
2061
	ibdev->ib_dev.rereg_user_mr	= mlx4_ib_rereg_user_mr;
2062
	ibdev->ib_dev.dereg_mr		= mlx4_ib_dereg_mr;
2063 2064 2065
	ibdev->ib_dev.alloc_fast_reg_mr = mlx4_ib_alloc_fast_reg_mr;
	ibdev->ib_dev.alloc_fast_reg_page_list = mlx4_ib_alloc_fast_reg_page_list;
	ibdev->ib_dev.free_fast_reg_page_list  = mlx4_ib_free_fast_reg_page_list;
2066 2067 2068 2069
	ibdev->ib_dev.attach_mcast	= mlx4_ib_mcg_attach;
	ibdev->ib_dev.detach_mcast	= mlx4_ib_mcg_detach;
	ibdev->ib_dev.process_mad	= mlx4_ib_process_mad;

2070 2071 2072 2073 2074 2075
	if (!mlx4_is_slave(ibdev->dev)) {
		ibdev->ib_dev.alloc_fmr		= mlx4_ib_fmr_alloc;
		ibdev->ib_dev.map_phys_fmr	= mlx4_ib_map_phys_fmr;
		ibdev->ib_dev.unmap_fmr		= mlx4_ib_unmap_fmr;
		ibdev->ib_dev.dealloc_fmr	= mlx4_ib_fmr_dealloc;
	}
J
Jack Morgenstein 已提交
2076

S
Shani Michaeli 已提交
2077 2078 2079 2080 2081 2082 2083 2084 2085 2086 2087
	if (dev->caps.flags & MLX4_DEV_CAP_FLAG_MEM_WINDOW ||
	    dev->caps.bmme_flags & MLX4_BMME_FLAG_TYPE_2_WIN) {
		ibdev->ib_dev.alloc_mw = mlx4_ib_alloc_mw;
		ibdev->ib_dev.bind_mw = mlx4_ib_bind_mw;
		ibdev->ib_dev.dealloc_mw = mlx4_ib_dealloc_mw;

		ibdev->ib_dev.uverbs_cmd_mask |=
			(1ull << IB_USER_VERBS_CMD_ALLOC_MW) |
			(1ull << IB_USER_VERBS_CMD_DEALLOC_MW);
	}

S
Sean Hefty 已提交
2088 2089 2090 2091 2092 2093 2094 2095
	if (dev->caps.flags & MLX4_DEV_CAP_FLAG_XRC) {
		ibdev->ib_dev.alloc_xrcd = mlx4_ib_alloc_xrcd;
		ibdev->ib_dev.dealloc_xrcd = mlx4_ib_dealloc_xrcd;
		ibdev->ib_dev.uverbs_cmd_mask |=
			(1ull << IB_USER_VERBS_CMD_OPEN_XRCD) |
			(1ull << IB_USER_VERBS_CMD_CLOSE_XRCD);
	}

2096
	if (check_flow_steering_support(dev)) {
2097
		ibdev->steering_support = MLX4_STEERING_MODE_DEVICE_MANAGED;
2098 2099 2100
		ibdev->ib_dev.create_flow	= mlx4_ib_create_flow;
		ibdev->ib_dev.destroy_flow	= mlx4_ib_destroy_flow;

2101 2102 2103
		ibdev->ib_dev.uverbs_ex_cmd_mask	|=
			(1ull << IB_USER_VERBS_EX_CMD_CREATE_FLOW) |
			(1ull << IB_USER_VERBS_EX_CMD_DESTROY_FLOW);
2104 2105
	}

2106 2107
	mlx4_ib_alloc_eqs(dev, ibdev);

E
Eli Cohen 已提交
2108 2109
	spin_lock_init(&iboe->lock);

2110 2111 2112
	if (init_node_data(ibdev))
		goto err_map;

2113
	for (i = 0; i < ibdev->num_ports; ++i) {
2114
		mutex_init(&ibdev->qp1_proxy_lock[i]);
2115 2116 2117 2118 2119
		if (mlx4_ib_port_link_layer(&ibdev->ib_dev, i + 1) ==
						IB_LINK_LAYER_ETHERNET) {
			err = mlx4_counter_alloc(ibdev->dev, &ibdev->counters[i]);
			if (err)
				ibdev->counters[i] = -1;
2120 2121 2122
		} else {
			ibdev->counters[i] = -1;
		}
2123 2124
	}

2125 2126 2127
	mlx4_foreach_port(i, dev, MLX4_PORT_TYPE_IB)
		ib_num_ports++;

2128 2129 2130
	spin_lock_init(&ibdev->sm_lock);
	mutex_init(&ibdev->cap_mask_mutex);

2131 2132
	if (ibdev->steering_support == MLX4_STEERING_MODE_DEVICE_MANAGED &&
	    ib_num_ports) {
2133 2134 2135 2136 2137 2138 2139 2140 2141 2142 2143 2144 2145 2146 2147 2148 2149 2150 2151 2152 2153 2154 2155 2156 2157 2158
		ibdev->steer_qpn_count = MLX4_IB_UC_MAX_NUM_QPS;
		err = mlx4_qp_reserve_range(dev, ibdev->steer_qpn_count,
					    MLX4_IB_UC_STEER_QPN_ALIGN,
					    &ibdev->steer_qpn_base);
		if (err)
			goto err_counter;

		ibdev->ib_uc_qpns_bitmap =
			kmalloc(BITS_TO_LONGS(ibdev->steer_qpn_count) *
				sizeof(long),
				GFP_KERNEL);
		if (!ibdev->ib_uc_qpns_bitmap) {
			dev_err(&dev->pdev->dev, "bit map alloc failed\n");
			goto err_steer_qp_release;
		}

		bitmap_zero(ibdev->ib_uc_qpns_bitmap, ibdev->steer_qpn_count);

		err = mlx4_FLOW_STEERING_IB_UC_QP_RANGE(
				dev, ibdev->steer_qpn_base,
				ibdev->steer_qpn_base +
				ibdev->steer_qpn_count - 1);
		if (err)
			goto err_steer_free_bitmap;
	}

2159
	if (ib_register_device(&ibdev->ib_dev, NULL))
2160
		goto err_steer_free_bitmap;
2161 2162 2163 2164

	if (mlx4_ib_mad_init(ibdev))
		goto err_reg;

2165 2166 2167
	if (mlx4_ib_init_sriov(ibdev))
		goto err_mad;

2168 2169 2170 2171 2172 2173 2174 2175 2176 2177 2178 2179 2180 2181 2182 2183 2184
	if (dev->caps.flags & MLX4_DEV_CAP_FLAG_IBOE) {
		if (!iboe->nb.notifier_call) {
			iboe->nb.notifier_call = mlx4_ib_netdev_event;
			err = register_netdevice_notifier(&iboe->nb);
			if (err) {
				iboe->nb.notifier_call = NULL;
				goto err_notif;
			}
		}
		if (!iboe->nb_inet.notifier_call) {
			iboe->nb_inet.notifier_call = mlx4_ib_inet_event;
			err = register_inetaddr_notifier(&iboe->nb_inet);
			if (err) {
				iboe->nb_inet.notifier_call = NULL;
				goto err_notif;
			}
		}
2185
#if IS_ENABLED(CONFIG_IPV6)
2186 2187 2188 2189 2190 2191 2192 2193 2194
		if (!iboe->nb_inet6.notifier_call) {
			iboe->nb_inet6.notifier_call = mlx4_ib_inet6_event;
			err = register_inet6addr_notifier(&iboe->nb_inet6);
			if (err) {
				iboe->nb_inet6.notifier_call = NULL;
				goto err_notif;
			}
		}
#endif
2195 2196
		for (i = 1 ; i <= ibdev->num_ports ; ++i)
			reset_gid_table(ibdev, i);
2197
		rtnl_lock();
2198
		mlx4_ib_scan_netdevs(ibdev, NULL, 0);
2199
		rtnl_unlock();
2200
		mlx4_ib_init_gid_table(ibdev);
E
Eli Cohen 已提交
2201 2202
	}

2203
	for (j = 0; j < ARRAY_SIZE(mlx4_class_attributes); ++j) {
2204
		if (device_create_file(&ibdev->ib_dev.dev,
2205
				       mlx4_class_attributes[j]))
E
Eli Cohen 已提交
2206
			goto err_notif;
2207 2208
	}

2209 2210
	ibdev->ib_active = true;

2211 2212 2213
	if (mlx4_is_mfunc(ibdev->dev))
		init_pkeys(ibdev);

2214 2215 2216 2217 2218 2219 2220 2221 2222
	/* create paravirt contexts for any VFs which are active */
	if (mlx4_is_master(ibdev->dev)) {
		for (j = 0; j < MLX4_MFUNC_MAX; j++) {
			if (j == mlx4_master_func_num(ibdev->dev))
				continue;
			if (mlx4_is_slave_active(ibdev->dev, j))
				do_slave_init(ibdev, j, 1);
		}
	}
2223 2224
	return ibdev;

E
Eli Cohen 已提交
2225
err_notif:
2226 2227 2228 2229 2230 2231 2232 2233 2234 2235
	if (ibdev->iboe.nb.notifier_call) {
		if (unregister_netdevice_notifier(&ibdev->iboe.nb))
			pr_warn("failure unregistering notifier\n");
		ibdev->iboe.nb.notifier_call = NULL;
	}
	if (ibdev->iboe.nb_inet.notifier_call) {
		if (unregister_inetaddr_notifier(&ibdev->iboe.nb_inet))
			pr_warn("failure unregistering notifier\n");
		ibdev->iboe.nb_inet.notifier_call = NULL;
	}
2236
#if IS_ENABLED(CONFIG_IPV6)
2237 2238 2239 2240 2241 2242
	if (ibdev->iboe.nb_inet6.notifier_call) {
		if (unregister_inet6addr_notifier(&ibdev->iboe.nb_inet6))
			pr_warn("failure unregistering notifier\n");
		ibdev->iboe.nb_inet6.notifier_call = NULL;
	}
#endif
E
Eli Cohen 已提交
2243 2244
	flush_workqueue(wq);

2245 2246 2247 2248 2249
	mlx4_ib_close_sriov(ibdev);

err_mad:
	mlx4_ib_mad_cleanup(ibdev);

2250 2251 2252
err_reg:
	ib_unregister_device(&ibdev->ib_dev);

2253 2254 2255 2256 2257 2258 2259
err_steer_free_bitmap:
	kfree(ibdev->ib_uc_qpns_bitmap);

err_steer_qp_release:
	if (ibdev->steering_support == MLX4_STEERING_MODE_DEVICE_MANAGED)
		mlx4_qp_release_range(dev, ibdev->steer_qpn_base,
				      ibdev->steer_qpn_count);
2260 2261
err_counter:
	for (; i; --i)
2262 2263
		if (ibdev->counters[i - 1] != -1)
			mlx4_counter_free(ibdev->dev, ibdev->counters[i - 1]);
2264

2265 2266 2267 2268 2269 2270 2271 2272 2273 2274 2275 2276 2277 2278 2279
err_map:
	iounmap(ibdev->uar_map);

err_uar:
	mlx4_uar_free(dev, &ibdev->priv_uar);

err_pd:
	mlx4_pd_free(dev, ibdev->priv_pdn);

err_dealloc:
	ib_dealloc_device(&ibdev->ib_dev);

	return NULL;
}

2280 2281 2282 2283 2284 2285 2286 2287 2288 2289 2290 2291 2292 2293 2294 2295 2296 2297 2298 2299 2300 2301 2302 2303 2304 2305 2306 2307 2308 2309 2310 2311 2312 2313 2314 2315 2316 2317 2318 2319 2320 2321 2322 2323 2324 2325 2326 2327 2328 2329 2330 2331 2332 2333 2334 2335 2336 2337 2338 2339 2340 2341 2342
int mlx4_ib_steer_qp_alloc(struct mlx4_ib_dev *dev, int count, int *qpn)
{
	int offset;

	WARN_ON(!dev->ib_uc_qpns_bitmap);

	offset = bitmap_find_free_region(dev->ib_uc_qpns_bitmap,
					 dev->steer_qpn_count,
					 get_count_order(count));
	if (offset < 0)
		return offset;

	*qpn = dev->steer_qpn_base + offset;
	return 0;
}

void mlx4_ib_steer_qp_free(struct mlx4_ib_dev *dev, u32 qpn, int count)
{
	if (!qpn ||
	    dev->steering_support != MLX4_STEERING_MODE_DEVICE_MANAGED)
		return;

	BUG_ON(qpn < dev->steer_qpn_base);

	bitmap_release_region(dev->ib_uc_qpns_bitmap,
			      qpn - dev->steer_qpn_base,
			      get_count_order(count));
}

int mlx4_ib_steer_qp_reg(struct mlx4_ib_dev *mdev, struct mlx4_ib_qp *mqp,
			 int is_attach)
{
	int err;
	size_t flow_size;
	struct ib_flow_attr *flow = NULL;
	struct ib_flow_spec_ib *ib_spec;

	if (is_attach) {
		flow_size = sizeof(struct ib_flow_attr) +
			    sizeof(struct ib_flow_spec_ib);
		flow = kzalloc(flow_size, GFP_KERNEL);
		if (!flow)
			return -ENOMEM;
		flow->port = mqp->port;
		flow->num_of_specs = 1;
		flow->size = flow_size;
		ib_spec = (struct ib_flow_spec_ib *)(flow + 1);
		ib_spec->type = IB_FLOW_SPEC_IB;
		ib_spec->size = sizeof(struct ib_flow_spec_ib);
		/* Add an empty rule for IB L2 */
		memset(&ib_spec->mask, 0, sizeof(ib_spec->mask));

		err = __mlx4_ib_create_flow(&mqp->ibqp, flow,
					    IB_FLOW_DOMAIN_NIC,
					    MLX4_FS_REGULAR,
					    &mqp->reg_id);
	} else {
		err = __mlx4_ib_destroy_flow(mdev->dev, mqp->reg_id);
	}
	kfree(flow);
	return err;
}

2343 2344 2345 2346 2347
static void mlx4_ib_remove(struct mlx4_dev *dev, void *ibdev_ptr)
{
	struct mlx4_ib_dev *ibdev = ibdev_ptr;
	int p;

2348
	mlx4_ib_close_sriov(ibdev);
2349 2350
	mlx4_ib_mad_cleanup(ibdev);
	ib_unregister_device(&ibdev->ib_dev);
E
Eli Cohen 已提交
2351 2352
	if (ibdev->iboe.nb.notifier_call) {
		if (unregister_netdevice_notifier(&ibdev->iboe.nb))
2353
			pr_warn("failure unregistering notifier\n");
E
Eli Cohen 已提交
2354 2355
		ibdev->iboe.nb.notifier_call = NULL;
	}
2356 2357 2358 2359 2360 2361 2362

	if (ibdev->steering_support == MLX4_STEERING_MODE_DEVICE_MANAGED) {
		mlx4_qp_release_range(dev, ibdev->steer_qpn_base,
				      ibdev->steer_qpn_count);
		kfree(ibdev->ib_uc_qpns_bitmap);
	}

2363 2364 2365 2366 2367
	if (ibdev->iboe.nb_inet.notifier_call) {
		if (unregister_inetaddr_notifier(&ibdev->iboe.nb_inet))
			pr_warn("failure unregistering notifier\n");
		ibdev->iboe.nb_inet.notifier_call = NULL;
	}
2368
#if IS_ENABLED(CONFIG_IPV6)
2369 2370 2371 2372 2373 2374
	if (ibdev->iboe.nb_inet6.notifier_call) {
		if (unregister_inet6addr_notifier(&ibdev->iboe.nb_inet6))
			pr_warn("failure unregistering notifier\n");
		ibdev->iboe.nb_inet6.notifier_call = NULL;
	}
#endif
2375

E
Eli Cohen 已提交
2376
	iounmap(ibdev->uar_map);
2377
	for (p = 0; p < ibdev->num_ports; ++p)
2378 2379
		if (ibdev->counters[p] != -1)
			mlx4_counter_free(ibdev->dev, ibdev->counters[p]);
E
Eli Cohen 已提交
2380
	mlx4_foreach_port(p, dev, MLX4_PORT_TYPE_IB)
2381 2382
		mlx4_CLOSE_PORT(dev, p);

2383 2384
	mlx4_ib_free_eqs(dev, ibdev);

2385 2386 2387 2388 2389
	mlx4_uar_free(dev, &ibdev->priv_uar);
	mlx4_pd_free(dev, ibdev->priv_pdn);
	ib_dealloc_device(&ibdev->ib_dev);
}

2390 2391 2392 2393 2394 2395
static void do_slave_init(struct mlx4_ib_dev *ibdev, int slave, int do_init)
{
	struct mlx4_ib_demux_work **dm = NULL;
	struct mlx4_dev *dev = ibdev->dev;
	int i;
	unsigned long flags;
M
Matan Barak 已提交
2396 2397 2398
	struct mlx4_active_ports actv_ports;
	unsigned int ports;
	unsigned int first_port;
2399 2400 2401 2402

	if (!mlx4_is_master(dev))
		return;

M
Matan Barak 已提交
2403 2404 2405 2406 2407
	actv_ports = mlx4_get_active_ports(dev, slave);
	ports = bitmap_weight(actv_ports.ports, dev->caps.num_ports);
	first_port = find_first_bit(actv_ports.ports, dev->caps.num_ports);

	dm = kcalloc(ports, sizeof(*dm), GFP_ATOMIC);
2408 2409 2410 2411 2412
	if (!dm) {
		pr_err("failed to allocate memory for tunneling qp update\n");
		goto out;
	}

M
Matan Barak 已提交
2413
	for (i = 0; i < ports; i++) {
2414 2415 2416 2417 2418 2419 2420 2421 2422 2423 2424
		dm[i] = kmalloc(sizeof (struct mlx4_ib_demux_work), GFP_ATOMIC);
		if (!dm[i]) {
			pr_err("failed to allocate memory for tunneling qp update work struct\n");
			for (i = 0; i < dev->caps.num_ports; i++) {
				if (dm[i])
					kfree(dm[i]);
			}
			goto out;
		}
	}
	/* initialize or tear down tunnel QPs for the slave */
M
Matan Barak 已提交
2425
	for (i = 0; i < ports; i++) {
2426
		INIT_WORK(&dm[i]->work, mlx4_ib_tunnels_update_work);
M
Matan Barak 已提交
2427
		dm[i]->port = first_port + i + 1;
2428 2429 2430 2431 2432 2433 2434 2435 2436
		dm[i]->slave = slave;
		dm[i]->do_init = do_init;
		dm[i]->dev = ibdev;
		spin_lock_irqsave(&ibdev->sriov.going_down_lock, flags);
		if (!ibdev->sriov.is_going_down)
			queue_work(ibdev->sriov.demux[i].ud_wq, &dm[i]->work);
		spin_unlock_irqrestore(&ibdev->sriov.going_down_lock, flags);
	}
out:
2437
	kfree(dm);
2438 2439 2440
	return;
}

2441
static void mlx4_ib_event(struct mlx4_dev *dev, void *ibdev_ptr,
2442
			  enum mlx4_dev_event event, unsigned long param)
2443 2444
{
	struct ib_event ibev;
2445
	struct mlx4_ib_dev *ibdev = to_mdev((struct ib_device *) ibdev_ptr);
2446 2447
	struct mlx4_eqe *eqe = NULL;
	struct ib_event_work *ew;
2448
	int p = 0;
2449 2450 2451 2452

	if (event == MLX4_DEV_EVENT_PORT_MGMT_CHANGE)
		eqe = (struct mlx4_eqe *)param;
	else
2453
		p = (int) param;
2454 2455

	switch (event) {
2456
	case MLX4_DEV_EVENT_PORT_UP:
2457 2458
		if (p > ibdev->num_ports)
			return;
2459 2460 2461 2462 2463
		if (mlx4_is_master(dev) &&
		    rdma_port_get_link_layer(&ibdev->ib_dev, p) ==
			IB_LINK_LAYER_INFINIBAND) {
			mlx4_ib_invalidate_all_guid_record(ibdev, p);
		}
2464
		ibev.event = IB_EVENT_PORT_ACTIVE;
2465 2466
		break;

2467
	case MLX4_DEV_EVENT_PORT_DOWN:
2468 2469
		if (p > ibdev->num_ports)
			return;
2470 2471 2472 2473
		ibev.event = IB_EVENT_PORT_ERR;
		break;

	case MLX4_DEV_EVENT_CATASTROPHIC_ERROR:
2474
		ibdev->ib_active = false;
2475 2476 2477
		ibev.event = IB_EVENT_DEVICE_FATAL;
		break;

2478 2479 2480 2481 2482 2483 2484 2485 2486 2487
	case MLX4_DEV_EVENT_PORT_MGMT_CHANGE:
		ew = kmalloc(sizeof *ew, GFP_ATOMIC);
		if (!ew) {
			pr_err("failed to allocate memory for events work\n");
			break;
		}

		INIT_WORK(&ew->work, handle_port_mgmt_change_event);
		memcpy(&ew->ib_eqe, eqe, sizeof *eqe);
		ew->ib_dev = ibdev;
2488 2489 2490 2491 2492
		/* need to queue only for port owner, which uses GEN_EQE */
		if (mlx4_is_master(dev))
			queue_work(wq, &ew->work);
		else
			handle_port_mgmt_change_event(&ew->work);
2493 2494
		return;

2495 2496 2497 2498 2499 2500 2501 2502 2503 2504
	case MLX4_DEV_EVENT_SLAVE_INIT:
		/* here, p is the slave id */
		do_slave_init(ibdev, p, 1);
		return;

	case MLX4_DEV_EVENT_SLAVE_SHUTDOWN:
		/* here, p is the slave id */
		do_slave_init(ibdev, p, 0);
		return;

2505 2506 2507 2508 2509
	default:
		return;
	}

	ibev.device	      = ibdev_ptr;
2510
	ibev.element.port_num = (u8) p;
2511 2512 2513 2514 2515

	ib_dispatch_event(&ibev);
}

static struct mlx4_interface mlx4_ib_interface = {
E
Eli Cohen 已提交
2516 2517 2518
	.add		= mlx4_ib_add,
	.remove		= mlx4_ib_remove,
	.event		= mlx4_ib_event,
2519
	.protocol	= MLX4_PROT_IB_IPV6
2520 2521 2522 2523
};

static int __init mlx4_ib_init(void)
{
E
Eli Cohen 已提交
2524 2525 2526 2527 2528 2529
	int err;

	wq = create_singlethread_workqueue("mlx4_ib");
	if (!wq)
		return -ENOMEM;

2530 2531 2532 2533
	err = mlx4_ib_mcg_init();
	if (err)
		goto clean_wq;

E
Eli Cohen 已提交
2534
	err = mlx4_register_interface(&mlx4_ib_interface);
2535 2536
	if (err)
		goto clean_mcg;
E
Eli Cohen 已提交
2537 2538

	return 0;
2539 2540 2541 2542 2543 2544 2545

clean_mcg:
	mlx4_ib_mcg_destroy();

clean_wq:
	destroy_workqueue(wq);
	return err;
2546 2547 2548 2549 2550
}

static void __exit mlx4_ib_cleanup(void)
{
	mlx4_unregister_interface(&mlx4_ib_interface);
2551
	mlx4_ib_mcg_destroy();
E
Eli Cohen 已提交
2552
	destroy_workqueue(wq);
2553 2554 2555 2556
}

module_init(mlx4_ib_init);
module_exit(mlx4_ib_cleanup);