nldev.c 33.3 KB
Newer Older
1 2 3 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
/*
 * Copyright (c) 2017 Mellanox Technologies. All rights reserved.
 *
 * Redistribution and use in source and binary forms, with or without
 * modification, are permitted provided that the following conditions are met:
 *
 * 1. Redistributions of source code must retain the above copyright
 *    notice, this list of conditions and the following disclaimer.
 * 2. 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.
 * 3. Neither the names of the copyright holders nor the names of its
 *    contributors may be used to endorse or promote products derived from
 *    this software without specific prior written permission.
 *
 * Alternatively, this software may be distributed under the terms of the
 * GNU General Public License ("GPL") version 2 as published by the Free
 * Software Foundation.
 *
 * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
 * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
 * ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE
 * LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
 * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
 * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
 * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
 * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
 * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
 * POSSIBILITY OF SUCH DAMAGE.
 */

33
#include <linux/module.h>
34 35
#include <linux/pid.h>
#include <linux/pid_namespace.h>
36
#include <net/netlink.h>
37
#include <rdma/rdma_cm.h>
38 39 40
#include <rdma/rdma_netlink.h>

#include "core_priv.h"
41
#include "cma_priv.h"
42
#include "restrack.h"
43

44 45 46 47 48
static const struct nla_policy nldev_policy[RDMA_NLDEV_ATTR_MAX] = {
	[RDMA_NLDEV_ATTR_DEV_INDEX]     = { .type = NLA_U32 },
	[RDMA_NLDEV_ATTR_DEV_NAME]	= { .type = NLA_NUL_STRING,
					    .len = IB_DEVICE_NAME_MAX - 1},
	[RDMA_NLDEV_ATTR_PORT_INDEX]	= { .type = NLA_U32 },
49 50
	[RDMA_NLDEV_ATTR_FW_VERSION]	= { .type = NLA_NUL_STRING,
					    .len = IB_FW_VERSION_NAME_MAX - 1},
51 52
	[RDMA_NLDEV_ATTR_NODE_GUID]	= { .type = NLA_U64 },
	[RDMA_NLDEV_ATTR_SYS_IMAGE_GUID] = { .type = NLA_U64 },
53
	[RDMA_NLDEV_ATTR_SUBNET_PREFIX]	= { .type = NLA_U64 },
54 55
	[RDMA_NLDEV_ATTR_LID]		= { .type = NLA_U32 },
	[RDMA_NLDEV_ATTR_SM_LID]	= { .type = NLA_U32 },
56
	[RDMA_NLDEV_ATTR_LMC]		= { .type = NLA_U8 },
57 58
	[RDMA_NLDEV_ATTR_PORT_STATE]	= { .type = NLA_U8 },
	[RDMA_NLDEV_ATTR_PORT_PHYS_STATE] = { .type = NLA_U8 },
L
Leon Romanovsky 已提交
59
	[RDMA_NLDEV_ATTR_DEV_NODE_TYPE] = { .type = NLA_U8 },
60 61 62 63 64
	[RDMA_NLDEV_ATTR_RES_SUMMARY]	= { .type = NLA_NESTED },
	[RDMA_NLDEV_ATTR_RES_SUMMARY_ENTRY]	= { .type = NLA_NESTED },
	[RDMA_NLDEV_ATTR_RES_SUMMARY_ENTRY_NAME] = { .type = NLA_NUL_STRING,
					     .len = 16 },
	[RDMA_NLDEV_ATTR_RES_SUMMARY_ENTRY_CURR] = { .type = NLA_U64 },
65 66 67 68 69 70 71 72 73 74 75 76
	[RDMA_NLDEV_ATTR_RES_QP]		= { .type = NLA_NESTED },
	[RDMA_NLDEV_ATTR_RES_QP_ENTRY]		= { .type = NLA_NESTED },
	[RDMA_NLDEV_ATTR_RES_LQPN]		= { .type = NLA_U32 },
	[RDMA_NLDEV_ATTR_RES_RQPN]		= { .type = NLA_U32 },
	[RDMA_NLDEV_ATTR_RES_RQ_PSN]		= { .type = NLA_U32 },
	[RDMA_NLDEV_ATTR_RES_SQ_PSN]		= { .type = NLA_U32 },
	[RDMA_NLDEV_ATTR_RES_PATH_MIG_STATE] = { .type = NLA_U8 },
	[RDMA_NLDEV_ATTR_RES_TYPE]		= { .type = NLA_U8 },
	[RDMA_NLDEV_ATTR_RES_STATE]		= { .type = NLA_U8 },
	[RDMA_NLDEV_ATTR_RES_PID]		= { .type = NLA_U32 },
	[RDMA_NLDEV_ATTR_RES_KERN_NAME]		= { .type = NLA_NUL_STRING,
						    .len = TASK_COMM_LEN },
77 78 79 80 81 82 83
	[RDMA_NLDEV_ATTR_RES_CM_ID]		= { .type = NLA_NESTED },
	[RDMA_NLDEV_ATTR_RES_CM_ID_ENTRY]	= { .type = NLA_NESTED },
	[RDMA_NLDEV_ATTR_RES_PS]		= { .type = NLA_U32 },
	[RDMA_NLDEV_ATTR_RES_SRC_ADDR]	= {
			.len = sizeof(struct __kernel_sockaddr_storage) },
	[RDMA_NLDEV_ATTR_RES_DST_ADDR]	= {
			.len = sizeof(struct __kernel_sockaddr_storage) },
84 85 86 87 88
	[RDMA_NLDEV_ATTR_RES_CQ]		= { .type = NLA_NESTED },
	[RDMA_NLDEV_ATTR_RES_CQ_ENTRY]		= { .type = NLA_NESTED },
	[RDMA_NLDEV_ATTR_RES_CQE]		= { .type = NLA_U32 },
	[RDMA_NLDEV_ATTR_RES_USECNT]		= { .type = NLA_U64 },
	[RDMA_NLDEV_ATTR_RES_POLL_CTX]		= { .type = NLA_U8 },
89 90 91 92 93 94
	[RDMA_NLDEV_ATTR_RES_MR]		= { .type = NLA_NESTED },
	[RDMA_NLDEV_ATTR_RES_MR_ENTRY]		= { .type = NLA_NESTED },
	[RDMA_NLDEV_ATTR_RES_RKEY]		= { .type = NLA_U32 },
	[RDMA_NLDEV_ATTR_RES_LKEY]		= { .type = NLA_U32 },
	[RDMA_NLDEV_ATTR_RES_IOVA]		= { .type = NLA_U64 },
	[RDMA_NLDEV_ATTR_RES_MRLEN]		= { .type = NLA_U64 },
95 96 97 98
	[RDMA_NLDEV_ATTR_RES_PD]		= { .type = NLA_NESTED },
	[RDMA_NLDEV_ATTR_RES_PD_ENTRY]		= { .type = NLA_NESTED },
	[RDMA_NLDEV_ATTR_RES_LOCAL_DMA_LKEY]	= { .type = NLA_U32 },
	[RDMA_NLDEV_ATTR_RES_UNSAFE_GLOBAL_RKEY] = { .type = NLA_U32 },
99 100 101
	[RDMA_NLDEV_ATTR_NDEV_INDEX]		= { .type = NLA_U32 },
	[RDMA_NLDEV_ATTR_NDEV_NAME]		= { .type = NLA_NUL_STRING,
						    .len = IFNAMSIZ },
102 103 104 105 106 107 108 109 110
	[RDMA_NLDEV_ATTR_DRIVER]		= { .type = NLA_NESTED },
	[RDMA_NLDEV_ATTR_DRIVER_ENTRY]		= { .type = NLA_NESTED },
	[RDMA_NLDEV_ATTR_DRIVER_STRING]		= { .type = NLA_NUL_STRING,
				    .len = RDMA_NLDEV_ATTR_ENTRY_STRLEN },
	[RDMA_NLDEV_ATTR_DRIVER_PRINT_TYPE]	= { .type = NLA_U8 },
	[RDMA_NLDEV_ATTR_DRIVER_S32]		= { .type = NLA_S32 },
	[RDMA_NLDEV_ATTR_DRIVER_U32]		= { .type = NLA_U32 },
	[RDMA_NLDEV_ATTR_DRIVER_S64]		= { .type = NLA_S64 },
	[RDMA_NLDEV_ATTR_DRIVER_U64]		= { .type = NLA_U64 },
111 112 113 114
	[RDMA_NLDEV_ATTR_RES_PDN]		= { .type = NLA_U32 },
	[RDMA_NLDEV_ATTR_RES_CQN]               = { .type = NLA_U32 },
	[RDMA_NLDEV_ATTR_RES_MRN]               = { .type = NLA_U32 },
	[RDMA_NLDEV_ATTR_RES_CM_IDN]            = { .type = NLA_U32 },
115
	[RDMA_NLDEV_ATTR_RES_CTXN]              = { .type = NLA_U32 },
116 117
};

118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183
static int put_driver_name_print_type(struct sk_buff *msg, const char *name,
				      enum rdma_nldev_print_type print_type)
{
	if (nla_put_string(msg, RDMA_NLDEV_ATTR_DRIVER_STRING, name))
		return -EMSGSIZE;
	if (print_type != RDMA_NLDEV_PRINT_TYPE_UNSPEC &&
	    nla_put_u8(msg, RDMA_NLDEV_ATTR_DRIVER_PRINT_TYPE, print_type))
		return -EMSGSIZE;

	return 0;
}

static int _rdma_nl_put_driver_u32(struct sk_buff *msg, const char *name,
				   enum rdma_nldev_print_type print_type,
				   u32 value)
{
	if (put_driver_name_print_type(msg, name, print_type))
		return -EMSGSIZE;
	if (nla_put_u32(msg, RDMA_NLDEV_ATTR_DRIVER_U32, value))
		return -EMSGSIZE;

	return 0;
}

static int _rdma_nl_put_driver_u64(struct sk_buff *msg, const char *name,
				   enum rdma_nldev_print_type print_type,
				   u64 value)
{
	if (put_driver_name_print_type(msg, name, print_type))
		return -EMSGSIZE;
	if (nla_put_u64_64bit(msg, RDMA_NLDEV_ATTR_DRIVER_U64, value,
			      RDMA_NLDEV_ATTR_PAD))
		return -EMSGSIZE;

	return 0;
}

int rdma_nl_put_driver_u32(struct sk_buff *msg, const char *name, u32 value)
{
	return _rdma_nl_put_driver_u32(msg, name, RDMA_NLDEV_PRINT_TYPE_UNSPEC,
				       value);
}
EXPORT_SYMBOL(rdma_nl_put_driver_u32);

int rdma_nl_put_driver_u32_hex(struct sk_buff *msg, const char *name,
			       u32 value)
{
	return _rdma_nl_put_driver_u32(msg, name, RDMA_NLDEV_PRINT_TYPE_HEX,
				       value);
}
EXPORT_SYMBOL(rdma_nl_put_driver_u32_hex);

int rdma_nl_put_driver_u64(struct sk_buff *msg, const char *name, u64 value)
{
	return _rdma_nl_put_driver_u64(msg, name, RDMA_NLDEV_PRINT_TYPE_UNSPEC,
				       value);
}
EXPORT_SYMBOL(rdma_nl_put_driver_u64);

int rdma_nl_put_driver_u64_hex(struct sk_buff *msg, const char *name, u64 value)
{
	return _rdma_nl_put_driver_u64(msg, name, RDMA_NLDEV_PRINT_TYPE_HEX,
				       value);
}
EXPORT_SYMBOL(rdma_nl_put_driver_u64_hex);

184
static int fill_nldev_handle(struct sk_buff *msg, struct ib_device *device)
185 186 187
{
	if (nla_put_u32(msg, RDMA_NLDEV_ATTR_DEV_INDEX, device->index))
		return -EMSGSIZE;
188 189
	if (nla_put_string(msg, RDMA_NLDEV_ATTR_DEV_NAME,
			   dev_name(&device->dev)))
190
		return -EMSGSIZE;
191 192 193 194 195 196 197 198 199 200 201

	return 0;
}

static int fill_dev_info(struct sk_buff *msg, struct ib_device *device)
{
	char fw[IB_FW_VERSION_NAME_MAX];

	if (fill_nldev_handle(msg, device))
		return -EMSGSIZE;

202 203
	if (nla_put_u32(msg, RDMA_NLDEV_ATTR_PORT_INDEX, rdma_end_port(device)))
		return -EMSGSIZE;
204 205 206

	BUILD_BUG_ON(sizeof(device->attrs.device_cap_flags) != sizeof(u64));
	if (nla_put_u64_64bit(msg, RDMA_NLDEV_ATTR_CAP_FLAGS,
207 208
			      device->attrs.device_cap_flags,
			      RDMA_NLDEV_ATTR_PAD))
209 210
		return -EMSGSIZE;

211
	ib_get_device_fw_str(device, fw);
212
	/* Device without FW has strlen(fw) = 0 */
213 214 215
	if (strlen(fw) && nla_put_string(msg, RDMA_NLDEV_ATTR_FW_VERSION, fw))
		return -EMSGSIZE;

216
	if (nla_put_u64_64bit(msg, RDMA_NLDEV_ATTR_NODE_GUID,
217 218
			      be64_to_cpu(device->node_guid),
			      RDMA_NLDEV_ATTR_PAD))
219 220
		return -EMSGSIZE;
	if (nla_put_u64_64bit(msg, RDMA_NLDEV_ATTR_SYS_IMAGE_GUID,
221 222
			      be64_to_cpu(device->attrs.sys_image_guid),
			      RDMA_NLDEV_ATTR_PAD))
223
		return -EMSGSIZE;
L
Leon Romanovsky 已提交
224 225
	if (nla_put_u8(msg, RDMA_NLDEV_ATTR_DEV_NODE_TYPE, device->node_type))
		return -EMSGSIZE;
226 227 228
	return 0;
}

229
static int fill_port_info(struct sk_buff *msg,
230 231
			  struct ib_device *device, u32 port,
			  const struct net *net)
232
{
233
	struct net_device *netdev = NULL;
234 235
	struct ib_port_attr attr;
	int ret;
236
	u64 cap_flags = 0;
237

238
	if (fill_nldev_handle(msg, device))
239
		return -EMSGSIZE;
240

241 242
	if (nla_put_u32(msg, RDMA_NLDEV_ATTR_PORT_INDEX, port))
		return -EMSGSIZE;
243 244 245 246 247

	ret = ib_query_port(device, port, &attr);
	if (ret)
		return ret;

248
	if (rdma_protocol_ib(device, port)) {
249 250 251 252
		BUILD_BUG_ON((sizeof(attr.port_cap_flags) +
				sizeof(attr.port_cap_flags2)) > sizeof(u64));
		cap_flags = attr.port_cap_flags |
			((u64)attr.port_cap_flags2 << 32);
253
		if (nla_put_u64_64bit(msg, RDMA_NLDEV_ATTR_CAP_FLAGS,
254
				      cap_flags, RDMA_NLDEV_ATTR_PAD))
255 256 257 258
			return -EMSGSIZE;
		if (nla_put_u64_64bit(msg, RDMA_NLDEV_ATTR_SUBNET_PREFIX,
				      attr.subnet_prefix, RDMA_NLDEV_ATTR_PAD))
			return -EMSGSIZE;
259 260 261 262
		if (nla_put_u32(msg, RDMA_NLDEV_ATTR_LID, attr.lid))
			return -EMSGSIZE;
		if (nla_put_u32(msg, RDMA_NLDEV_ATTR_SM_LID, attr.sm_lid))
			return -EMSGSIZE;
263 264
		if (nla_put_u8(msg, RDMA_NLDEV_ATTR_LMC, attr.lmc))
			return -EMSGSIZE;
265
	}
266 267 268 269
	if (nla_put_u8(msg, RDMA_NLDEV_ATTR_PORT_STATE, attr.state))
		return -EMSGSIZE;
	if (nla_put_u8(msg, RDMA_NLDEV_ATTR_PORT_PHYS_STATE, attr.phys_state))
		return -EMSGSIZE;
270

K
Kamal Heib 已提交
271 272
	if (device->ops.get_netdev)
		netdev = device->ops.get_netdev(device, port);
273 274 275 276 277 278 279 280 281 282 283 284 285 286

	if (netdev && net_eq(dev_net(netdev), net)) {
		ret = nla_put_u32(msg,
				  RDMA_NLDEV_ATTR_NDEV_INDEX, netdev->ifindex);
		if (ret)
			goto out;
		ret = nla_put_string(msg,
				     RDMA_NLDEV_ATTR_NDEV_NAME, netdev->name);
	}

out:
	if (netdev)
		dev_put(netdev);
	return ret;
287 288
}

289 290 291 292 293 294 295 296 297 298 299
static int fill_res_info_entry(struct sk_buff *msg,
			       const char *name, u64 curr)
{
	struct nlattr *entry_attr;

	entry_attr = nla_nest_start(msg, RDMA_NLDEV_ATTR_RES_SUMMARY_ENTRY);
	if (!entry_attr)
		return -EMSGSIZE;

	if (nla_put_string(msg, RDMA_NLDEV_ATTR_RES_SUMMARY_ENTRY_NAME, name))
		goto err;
300 301
	if (nla_put_u64_64bit(msg, RDMA_NLDEV_ATTR_RES_SUMMARY_ENTRY_CURR, curr,
			      RDMA_NLDEV_ATTR_PAD))
302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317
		goto err;

	nla_nest_end(msg, entry_attr);
	return 0;

err:
	nla_nest_cancel(msg, entry_attr);
	return -EMSGSIZE;
}

static int fill_res_info(struct sk_buff *msg, struct ib_device *device)
{
	static const char * const names[RDMA_RESTRACK_MAX] = {
		[RDMA_RESTRACK_PD] = "pd",
		[RDMA_RESTRACK_CQ] = "cq",
		[RDMA_RESTRACK_QP] = "qp",
318
		[RDMA_RESTRACK_CM_ID] = "cm_id",
319
		[RDMA_RESTRACK_MR] = "mr",
320
		[RDMA_RESTRACK_CTX] = "ctx",
321 322 323 324 325 326 327 328 329 330 331 332 333 334 335
	};

	struct nlattr *table_attr;
	int ret, i, curr;

	if (fill_nldev_handle(msg, device))
		return -EMSGSIZE;

	table_attr = nla_nest_start(msg, RDMA_NLDEV_ATTR_RES_SUMMARY);
	if (!table_attr)
		return -EMSGSIZE;

	for (i = 0; i < RDMA_RESTRACK_MAX; i++) {
		if (!names[i])
			continue;
336 337
		curr = rdma_restrack_count(device, i,
					   task_active_pid_ns(current));
338 339 340 341 342 343 344 345 346 347 348 349 350
		ret = fill_res_info_entry(msg, names[i], curr);
		if (ret)
			goto err;
	}

	nla_nest_end(msg, table_attr);
	return 0;

err:
	nla_nest_cancel(msg, table_attr);
	return ret;
}

351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369
static int fill_res_name_pid(struct sk_buff *msg,
			     struct rdma_restrack_entry *res)
{
	/*
	 * For user resources, user is should read /proc/PID/comm to get the
	 * name of the task file.
	 */
	if (rdma_is_kernel_res(res)) {
		if (nla_put_string(msg, RDMA_NLDEV_ATTR_RES_KERN_NAME,
		    res->kern_name))
			return -EMSGSIZE;
	} else {
		if (nla_put_u32(msg, RDMA_NLDEV_ATTR_RES_PID,
		    task_pid_vnr(res->task)))
			return -EMSGSIZE;
	}
	return 0;
}

370 371 372 373 374 375 376 377
static bool fill_res_entry(struct ib_device *dev, struct sk_buff *msg,
			   struct rdma_restrack_entry *res)
{
	if (!dev->ops.fill_res_entry)
		return false;
	return dev->ops.fill_res_entry(msg, res);
}

378
static int fill_res_qp_entry(struct sk_buff *msg, bool has_cap_net_admin,
379
			     struct rdma_restrack_entry *res, uint32_t port)
380
{
381
	struct ib_qp *qp = container_of(res, struct ib_qp, res);
382
	struct ib_device *dev = qp->device;
383 384 385 386 387 388 389 390 391
	struct ib_qp_init_attr qp_init_attr;
	struct ib_qp_attr qp_attr;
	int ret;

	ret = ib_query_qp(qp, &qp_attr, 0, &qp_init_attr);
	if (ret)
		return ret;

	if (port && port != qp_attr.port_num)
392
		return -EAGAIN;
393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423

	/* In create_qp() port is not set yet */
	if (qp_attr.port_num &&
	    nla_put_u32(msg, RDMA_NLDEV_ATTR_PORT_INDEX, qp_attr.port_num))
		goto err;

	if (nla_put_u32(msg, RDMA_NLDEV_ATTR_RES_LQPN, qp->qp_num))
		goto err;
	if (qp->qp_type == IB_QPT_RC || qp->qp_type == IB_QPT_UC) {
		if (nla_put_u32(msg, RDMA_NLDEV_ATTR_RES_RQPN,
				qp_attr.dest_qp_num))
			goto err;
		if (nla_put_u32(msg, RDMA_NLDEV_ATTR_RES_RQ_PSN,
				qp_attr.rq_psn))
			goto err;
	}

	if (nla_put_u32(msg, RDMA_NLDEV_ATTR_RES_SQ_PSN, qp_attr.sq_psn))
		goto err;

	if (qp->qp_type == IB_QPT_RC || qp->qp_type == IB_QPT_UC ||
	    qp->qp_type == IB_QPT_XRC_INI || qp->qp_type == IB_QPT_XRC_TGT) {
		if (nla_put_u8(msg, RDMA_NLDEV_ATTR_RES_PATH_MIG_STATE,
			       qp_attr.path_mig_state))
			goto err;
	}
	if (nla_put_u8(msg, RDMA_NLDEV_ATTR_RES_TYPE, qp->qp_type))
		goto err;
	if (nla_put_u8(msg, RDMA_NLDEV_ATTR_RES_STATE, qp_attr.qp_state))
		goto err;

424 425 426 427
	if (!rdma_is_kernel_res(res) &&
	    nla_put_u32(msg, RDMA_NLDEV_ATTR_RES_PDN, qp->pd->res.id))
		goto err;

428 429 430
	if (fill_res_name_pid(msg, res))
		goto err;

431
	if (fill_res_entry(dev, msg, res))
432 433
		goto err;

434 435
	return 0;

436
err:	return -EMSGSIZE;
437 438
}

439
static int fill_res_cm_id_entry(struct sk_buff *msg, bool has_cap_net_admin,
440 441 442 443
				struct rdma_restrack_entry *res, uint32_t port)
{
	struct rdma_id_private *id_priv =
				container_of(res, struct rdma_id_private, res);
444
	struct ib_device *dev = id_priv->id.device;
445 446 447 448 449 450 451 452 453 454 455
	struct rdma_cm_id *cm_id = &id_priv->id;

	if (port && port != cm_id->port_num)
		return 0;

	if (cm_id->port_num &&
	    nla_put_u32(msg, RDMA_NLDEV_ATTR_PORT_INDEX, cm_id->port_num))
		goto err;

	if (id_priv->qp_num) {
		if (nla_put_u32(msg, RDMA_NLDEV_ATTR_RES_LQPN, id_priv->qp_num))
456
			goto err;
457
		if (nla_put_u8(msg, RDMA_NLDEV_ATTR_RES_TYPE, cm_id->qp_type))
458 459 460
			goto err;
	}

461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477
	if (nla_put_u32(msg, RDMA_NLDEV_ATTR_RES_PS, cm_id->ps))
		goto err;

	if (nla_put_u8(msg, RDMA_NLDEV_ATTR_RES_STATE, id_priv->state))
		goto err;

	if (cm_id->route.addr.src_addr.ss_family &&
	    nla_put(msg, RDMA_NLDEV_ATTR_RES_SRC_ADDR,
		    sizeof(cm_id->route.addr.src_addr),
		    &cm_id->route.addr.src_addr))
		goto err;
	if (cm_id->route.addr.dst_addr.ss_family &&
	    nla_put(msg, RDMA_NLDEV_ATTR_RES_DST_ADDR,
		    sizeof(cm_id->route.addr.dst_addr),
		    &cm_id->route.addr.dst_addr))
		goto err;

478 479 480
	if (nla_put_u32(msg, RDMA_NLDEV_ATTR_RES_CM_IDN, res->id))
		goto err;

481 482 483
	if (fill_res_name_pid(msg, res))
		goto err;

484
	if (fill_res_entry(dev, msg, res))
485 486
		goto err;

487 488
	return 0;

489
err: return -EMSGSIZE;
490 491
}

492
static int fill_res_cq_entry(struct sk_buff *msg, bool has_cap_net_admin,
493 494 495
			     struct rdma_restrack_entry *res, uint32_t port)
{
	struct ib_cq *cq = container_of(res, struct ib_cq, res);
496
	struct ib_device *dev = cq->device;
497 498 499 500

	if (nla_put_u32(msg, RDMA_NLDEV_ATTR_RES_CQE, cq->cqe))
		goto err;
	if (nla_put_u64_64bit(msg, RDMA_NLDEV_ATTR_RES_USECNT,
501
			      atomic_read(&cq->usecnt), RDMA_NLDEV_ATTR_PAD))
502 503 504 505 506 507 508
		goto err;

	/* Poll context is only valid for kernel CQs */
	if (rdma_is_kernel_res(res) &&
	    nla_put_u8(msg, RDMA_NLDEV_ATTR_RES_POLL_CTX, cq->poll_ctx))
		goto err;

509 510
	if (nla_put_u32(msg, RDMA_NLDEV_ATTR_RES_CQN, res->id))
		goto err;
511 512 513 514
	if (!rdma_is_kernel_res(res) &&
	    nla_put_u32(msg, RDMA_NLDEV_ATTR_RES_CTXN,
			cq->uobject->context->res.id))
		goto err;
515

516 517 518
	if (fill_res_name_pid(msg, res))
		goto err;

519
	if (fill_res_entry(dev, msg, res))
520 521
		goto err;

522 523
	return 0;

524
err:	return -EMSGSIZE;
525 526
}

527
static int fill_res_mr_entry(struct sk_buff *msg, bool has_cap_net_admin,
528 529 530
			     struct rdma_restrack_entry *res, uint32_t port)
{
	struct ib_mr *mr = container_of(res, struct ib_mr, res);
531
	struct ib_device *dev = mr->pd->device;
532

533
	if (has_cap_net_admin) {
534 535 536 537 538 539
		if (nla_put_u32(msg, RDMA_NLDEV_ATTR_RES_RKEY, mr->rkey))
			goto err;
		if (nla_put_u32(msg, RDMA_NLDEV_ATTR_RES_LKEY, mr->lkey))
			goto err;
	}

540 541
	if (nla_put_u64_64bit(msg, RDMA_NLDEV_ATTR_RES_MRLEN, mr->length,
			      RDMA_NLDEV_ATTR_PAD))
542 543
		goto err;

544 545 546
	if (nla_put_u32(msg, RDMA_NLDEV_ATTR_RES_MRN, res->id))
		goto err;

547 548 549 550
	if (!rdma_is_kernel_res(res) &&
	    nla_put_u32(msg, RDMA_NLDEV_ATTR_RES_PDN, mr->pd->res.id))
		goto err;

551 552 553
	if (fill_res_name_pid(msg, res))
		goto err;

554
	if (fill_res_entry(dev, msg, res))
555 556
		goto err;

557 558
	return 0;

559
err:	return -EMSGSIZE;
560 561
}

562
static int fill_res_pd_entry(struct sk_buff *msg, bool has_cap_net_admin,
563 564 565
			     struct rdma_restrack_entry *res, uint32_t port)
{
	struct ib_pd *pd = container_of(res, struct ib_pd, res);
566
	struct ib_device *dev = pd->device;
567

568
	if (has_cap_net_admin) {
569 570 571 572 573 574 575 576 577
		if (nla_put_u32(msg, RDMA_NLDEV_ATTR_RES_LOCAL_DMA_LKEY,
				pd->local_dma_lkey))
			goto err;
		if ((pd->flags & IB_PD_UNSAFE_GLOBAL_RKEY) &&
		    nla_put_u32(msg, RDMA_NLDEV_ATTR_RES_UNSAFE_GLOBAL_RKEY,
				pd->unsafe_global_rkey))
			goto err;
	}
	if (nla_put_u64_64bit(msg, RDMA_NLDEV_ATTR_RES_USECNT,
578
			      atomic_read(&pd->usecnt), RDMA_NLDEV_ATTR_PAD))
579 580
		goto err;

581 582 583
	if (nla_put_u32(msg, RDMA_NLDEV_ATTR_RES_PDN, res->id))
		goto err;

584 585 586 587 588
	if (!rdma_is_kernel_res(res) &&
	    nla_put_u32(msg, RDMA_NLDEV_ATTR_RES_CTXN,
			pd->uobject->context->res.id))
		goto err;

589 590 591
	if (fill_res_name_pid(msg, res))
		goto err;

592
	if (fill_res_entry(dev, msg, res))
593 594
		goto err;

595 596
	return 0;

597
err:	return -EMSGSIZE;
598 599
}

600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615
static int nldev_get_doit(struct sk_buff *skb, struct nlmsghdr *nlh,
			  struct netlink_ext_ack *extack)
{
	struct nlattr *tb[RDMA_NLDEV_ATTR_MAX];
	struct ib_device *device;
	struct sk_buff *msg;
	u32 index;
	int err;

	err = nlmsg_parse(nlh, 0, tb, RDMA_NLDEV_ATTR_MAX - 1,
			  nldev_policy, extack);
	if (err || !tb[RDMA_NLDEV_ATTR_DEV_INDEX])
		return -EINVAL;

	index = nla_get_u32(tb[RDMA_NLDEV_ATTR_DEV_INDEX]);

616
	device = ib_device_get_by_index(index);
617 618 619 620
	if (!device)
		return -EINVAL;

	msg = nlmsg_new(NLMSG_DEFAULT_SIZE, GFP_KERNEL);
621 622 623 624
	if (!msg) {
		err = -ENOMEM;
		goto err;
	}
625 626 627 628 629 630

	nlh = nlmsg_put(msg, NETLINK_CB(skb).portid, nlh->nlmsg_seq,
			RDMA_NL_GET_TYPE(RDMA_NL_NLDEV, RDMA_NLDEV_CMD_GET),
			0, 0);

	err = fill_dev_info(msg, device);
631 632
	if (err)
		goto err_free;
633 634 635

	nlmsg_end(msg, nlh);

636
	ib_device_put(device);
637
	return rdma_nl_unicast(msg, NETLINK_CB(skb).portid);
638 639 640 641

err_free:
	nlmsg_free(msg);
err:
642
	ib_device_put(device);
643
	return err;
644 645
}

646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671
static int nldev_set_doit(struct sk_buff *skb, struct nlmsghdr *nlh,
			  struct netlink_ext_ack *extack)
{
	struct nlattr *tb[RDMA_NLDEV_ATTR_MAX];
	struct ib_device *device;
	u32 index;
	int err;

	err = nlmsg_parse(nlh, 0, tb, RDMA_NLDEV_ATTR_MAX - 1, nldev_policy,
			  extack);
	if (err || !tb[RDMA_NLDEV_ATTR_DEV_INDEX])
		return -EINVAL;

	index = nla_get_u32(tb[RDMA_NLDEV_ATTR_DEV_INDEX]);
	device = ib_device_get_by_index(index);
	if (!device)
		return -EINVAL;

	if (tb[RDMA_NLDEV_ATTR_DEV_NAME]) {
		char name[IB_DEVICE_NAME_MAX] = {};

		nla_strlcpy(name, tb[RDMA_NLDEV_ATTR_DEV_NAME],
			    IB_DEVICE_NAME_MAX);
		err = ib_device_rename(device, name);
	}

672
	ib_device_put(device);
673 674 675
	return err;
}

676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712
static int _nldev_get_dumpit(struct ib_device *device,
			     struct sk_buff *skb,
			     struct netlink_callback *cb,
			     unsigned int idx)
{
	int start = cb->args[0];
	struct nlmsghdr *nlh;

	if (idx < start)
		return 0;

	nlh = nlmsg_put(skb, NETLINK_CB(cb->skb).portid, cb->nlh->nlmsg_seq,
			RDMA_NL_GET_TYPE(RDMA_NL_NLDEV, RDMA_NLDEV_CMD_GET),
			0, NLM_F_MULTI);

	if (fill_dev_info(skb, device)) {
		nlmsg_cancel(skb, nlh);
		goto out;
	}

	nlmsg_end(skb, nlh);

	idx++;

out:	cb->args[0] = idx;
	return skb->len;
}

static int nldev_get_dumpit(struct sk_buff *skb, struct netlink_callback *cb)
{
	/*
	 * There is no need to take lock, because
	 * we are relying on ib_core's lists_rwsem
	 */
	return ib_enum_all_devs(_nldev_get_dumpit, skb, cb);
}

713 714 715 716 717 718 719 720 721 722 723 724
static int nldev_port_get_doit(struct sk_buff *skb, struct nlmsghdr *nlh,
			       struct netlink_ext_ack *extack)
{
	struct nlattr *tb[RDMA_NLDEV_ATTR_MAX];
	struct ib_device *device;
	struct sk_buff *msg;
	u32 index;
	u32 port;
	int err;

	err = nlmsg_parse(nlh, 0, tb, RDMA_NLDEV_ATTR_MAX - 1,
			  nldev_policy, extack);
725 726 727
	if (err ||
	    !tb[RDMA_NLDEV_ATTR_DEV_INDEX] ||
	    !tb[RDMA_NLDEV_ATTR_PORT_INDEX])
728 729 730
		return -EINVAL;

	index = nla_get_u32(tb[RDMA_NLDEV_ATTR_DEV_INDEX]);
731
	device = ib_device_get_by_index(index);
732 733 734 735
	if (!device)
		return -EINVAL;

	port = nla_get_u32(tb[RDMA_NLDEV_ATTR_PORT_INDEX]);
736 737 738 739
	if (!rdma_is_port_valid(device, port)) {
		err = -EINVAL;
		goto err;
	}
740 741

	msg = nlmsg_new(NLMSG_DEFAULT_SIZE, GFP_KERNEL);
742 743 744 745
	if (!msg) {
		err = -ENOMEM;
		goto err;
	}
746 747 748 749 750

	nlh = nlmsg_put(msg, NETLINK_CB(skb).portid, nlh->nlmsg_seq,
			RDMA_NL_GET_TYPE(RDMA_NL_NLDEV, RDMA_NLDEV_CMD_GET),
			0, 0);

751
	err = fill_port_info(msg, device, port, sock_net(skb->sk));
752 753
	if (err)
		goto err_free;
754 755

	nlmsg_end(msg, nlh);
756
	ib_device_put(device);
757 758

	return rdma_nl_unicast(msg, NETLINK_CB(skb).portid);
759 760 761 762

err_free:
	nlmsg_free(msg);
err:
763
	ib_device_put(device);
764
	return err;
765 766
}

767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784
static int nldev_port_get_dumpit(struct sk_buff *skb,
				 struct netlink_callback *cb)
{
	struct nlattr *tb[RDMA_NLDEV_ATTR_MAX];
	struct ib_device *device;
	int start = cb->args[0];
	struct nlmsghdr *nlh;
	u32 idx = 0;
	u32 ifindex;
	int err;
	u32 p;

	err = nlmsg_parse(cb->nlh, 0, tb, RDMA_NLDEV_ATTR_MAX - 1,
			  nldev_policy, NULL);
	if (err || !tb[RDMA_NLDEV_ATTR_DEV_INDEX])
		return -EINVAL;

	ifindex = nla_get_u32(tb[RDMA_NLDEV_ATTR_DEV_INDEX]);
785
	device = ib_device_get_by_index(ifindex);
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
	if (!device)
		return -EINVAL;

	for (p = rdma_start_port(device); p <= rdma_end_port(device); ++p) {
		/*
		 * The dumpit function returns all information from specific
		 * index. This specific index is taken from the netlink
		 * messages request sent by user and it is available
		 * in cb->args[0].
		 *
		 * Usually, the user doesn't fill this field and it causes
		 * to return everything.
		 *
		 */
		if (idx < start) {
			idx++;
			continue;
		}

		nlh = nlmsg_put(skb, NETLINK_CB(cb->skb).portid,
				cb->nlh->nlmsg_seq,
				RDMA_NL_GET_TYPE(RDMA_NL_NLDEV,
						 RDMA_NLDEV_CMD_PORT_GET),
				0, NLM_F_MULTI);

811
		if (fill_port_info(skb, device, p, sock_net(skb->sk))) {
812 813 814 815 816 817 818
			nlmsg_cancel(skb, nlh);
			goto out;
		}
		idx++;
		nlmsg_end(skb, nlh);
	}

819
out:
820
	ib_device_put(device);
821
	cb->args[0] = idx;
822 823 824
	return skb->len;
}

825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844
static int nldev_res_get_doit(struct sk_buff *skb, struct nlmsghdr *nlh,
			      struct netlink_ext_ack *extack)
{
	struct nlattr *tb[RDMA_NLDEV_ATTR_MAX];
	struct ib_device *device;
	struct sk_buff *msg;
	u32 index;
	int ret;

	ret = nlmsg_parse(nlh, 0, tb, RDMA_NLDEV_ATTR_MAX - 1,
			  nldev_policy, extack);
	if (ret || !tb[RDMA_NLDEV_ATTR_DEV_INDEX])
		return -EINVAL;

	index = nla_get_u32(tb[RDMA_NLDEV_ATTR_DEV_INDEX]);
	device = ib_device_get_by_index(index);
	if (!device)
		return -EINVAL;

	msg = nlmsg_new(NLMSG_DEFAULT_SIZE, GFP_KERNEL);
845 846
	if (!msg) {
		ret = -ENOMEM;
847
		goto err;
848
	}
849 850 851 852 853 854 855 856 857 858

	nlh = nlmsg_put(msg, NETLINK_CB(skb).portid, nlh->nlmsg_seq,
			RDMA_NL_GET_TYPE(RDMA_NL_NLDEV, RDMA_NLDEV_CMD_RES_GET),
			0, 0);

	ret = fill_res_info(msg, device);
	if (ret)
		goto err_free;

	nlmsg_end(msg, nlh);
859
	ib_device_put(device);
860 861 862 863 864
	return rdma_nl_unicast(msg, NETLINK_CB(skb).portid);

err_free:
	nlmsg_free(msg);
err:
865
	ib_device_put(device);
866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 901 902 903
	return ret;
}

static int _nldev_res_get_dumpit(struct ib_device *device,
				 struct sk_buff *skb,
				 struct netlink_callback *cb,
				 unsigned int idx)
{
	int start = cb->args[0];
	struct nlmsghdr *nlh;

	if (idx < start)
		return 0;

	nlh = nlmsg_put(skb, NETLINK_CB(cb->skb).portid, cb->nlh->nlmsg_seq,
			RDMA_NL_GET_TYPE(RDMA_NL_NLDEV, RDMA_NLDEV_CMD_RES_GET),
			0, NLM_F_MULTI);

	if (fill_res_info(skb, device)) {
		nlmsg_cancel(skb, nlh);
		goto out;
	}

	nlmsg_end(skb, nlh);

	idx++;

out:
	cb->args[0] = idx;
	return skb->len;
}

static int nldev_res_get_dumpit(struct sk_buff *skb,
				struct netlink_callback *cb)
{
	return ib_enum_all_devs(_nldev_res_get_dumpit, skb, cb);
}

904
struct nldev_fill_res_entry {
905
	int (*fill_res_func)(struct sk_buff *msg, bool has_cap_net_admin,
906 907 908
			     struct rdma_restrack_entry *res, u32 port);
	enum rdma_nldev_attr nldev_attr;
	enum rdma_nldev_command nldev_cmd;
909 910 911 912 913 914 915
	u8 flags;
	u32 entry;
	u32 id;
};

enum nldev_res_flags {
	NLDEV_PER_DEV = 1 << 0,
916 917 918 919 920 921 922
};

static const struct nldev_fill_res_entry fill_entries[RDMA_RESTRACK_MAX] = {
	[RDMA_RESTRACK_QP] = {
		.fill_res_func = fill_res_qp_entry,
		.nldev_cmd = RDMA_NLDEV_CMD_RES_QP_GET,
		.nldev_attr = RDMA_NLDEV_ATTR_RES_QP,
923
		.entry = RDMA_NLDEV_ATTR_RES_QP_ENTRY,
924
		.id = RDMA_NLDEV_ATTR_RES_LQPN,
925
	},
926 927 928 929
	[RDMA_RESTRACK_CM_ID] = {
		.fill_res_func = fill_res_cm_id_entry,
		.nldev_cmd = RDMA_NLDEV_CMD_RES_CM_ID_GET,
		.nldev_attr = RDMA_NLDEV_ATTR_RES_CM_ID,
930
		.entry = RDMA_NLDEV_ATTR_RES_CM_ID_ENTRY,
931
		.id = RDMA_NLDEV_ATTR_RES_CM_IDN,
932
	},
933 934 935 936
	[RDMA_RESTRACK_CQ] = {
		.fill_res_func = fill_res_cq_entry,
		.nldev_cmd = RDMA_NLDEV_CMD_RES_CQ_GET,
		.nldev_attr = RDMA_NLDEV_ATTR_RES_CQ,
937 938
		.flags = NLDEV_PER_DEV,
		.entry = RDMA_NLDEV_ATTR_RES_CQ_ENTRY,
939
		.id = RDMA_NLDEV_ATTR_RES_CQN,
940
	},
941 942 943 944
	[RDMA_RESTRACK_MR] = {
		.fill_res_func = fill_res_mr_entry,
		.nldev_cmd = RDMA_NLDEV_CMD_RES_MR_GET,
		.nldev_attr = RDMA_NLDEV_ATTR_RES_MR,
945 946
		.flags = NLDEV_PER_DEV,
		.entry = RDMA_NLDEV_ATTR_RES_MR_ENTRY,
947
		.id = RDMA_NLDEV_ATTR_RES_MRN,
948
	},
949 950 951 952
	[RDMA_RESTRACK_PD] = {
		.fill_res_func = fill_res_pd_entry,
		.nldev_cmd = RDMA_NLDEV_CMD_RES_PD_GET,
		.nldev_attr = RDMA_NLDEV_ATTR_RES_PD,
953 954
		.flags = NLDEV_PER_DEV,
		.entry = RDMA_NLDEV_ATTR_RES_PD_ENTRY,
955
		.id = RDMA_NLDEV_ATTR_RES_PDN,
956
	},
957 958
};

959 960 961 962 963 964 965 966 967 968 969
static bool is_visible_in_pid_ns(struct rdma_restrack_entry *res)
{
	/*
	 * 1. Kern resources should be visible in init name space only
	 * 2. Present only resources visible in the current namespace
	 */
	if (rdma_is_kernel_res(res))
		return task_active_pid_ns(current) == &init_pid_ns;
	return task_active_pid_ns(current) == task_active_pid_ns(res->task);
}

970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 992 993 994 995 996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 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 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052
static int res_get_common_doit(struct sk_buff *skb, struct nlmsghdr *nlh,
			       struct netlink_ext_ack *extack,
			       enum rdma_restrack_type res_type)
{
	const struct nldev_fill_res_entry *fe = &fill_entries[res_type];
	struct nlattr *tb[RDMA_NLDEV_ATTR_MAX];
	struct rdma_restrack_entry *res;
	struct ib_device *device;
	u32 index, id, port = 0;
	bool has_cap_net_admin;
	struct sk_buff *msg;
	int ret;

	ret = nlmsg_parse(nlh, 0, tb, RDMA_NLDEV_ATTR_MAX - 1,
			  nldev_policy, extack);
	if (ret || !tb[RDMA_NLDEV_ATTR_DEV_INDEX] || !fe->id || !tb[fe->id])
		return -EINVAL;

	index = nla_get_u32(tb[RDMA_NLDEV_ATTR_DEV_INDEX]);
	device = ib_device_get_by_index(index);
	if (!device)
		return -EINVAL;

	if (tb[RDMA_NLDEV_ATTR_PORT_INDEX]) {
		port = nla_get_u32(tb[RDMA_NLDEV_ATTR_PORT_INDEX]);
		if (!rdma_is_port_valid(device, port)) {
			ret = -EINVAL;
			goto err;
		}
	}

	if ((port && fe->flags & NLDEV_PER_DEV) ||
	    (!port && ~fe->flags & NLDEV_PER_DEV)) {
		ret = -EINVAL;
		goto err;
	}

	id = nla_get_u32(tb[fe->id]);
	res = rdma_restrack_get_byid(device, res_type, id);
	if (IS_ERR(res)) {
		ret = PTR_ERR(res);
		goto err;
	}

	if (!is_visible_in_pid_ns(res)) {
		ret = -ENOENT;
		goto err_get;
	}

	msg = nlmsg_new(NLMSG_DEFAULT_SIZE, GFP_KERNEL);
	if (!msg) {
		ret = -ENOMEM;
		goto err;
	}

	nlh = nlmsg_put(msg, NETLINK_CB(skb).portid, nlh->nlmsg_seq,
			RDMA_NL_GET_TYPE(RDMA_NL_NLDEV, fe->nldev_cmd),
			0, 0);

	if (fill_nldev_handle(msg, device)) {
		ret = -EMSGSIZE;
		goto err_free;
	}

	has_cap_net_admin = netlink_capable(skb, CAP_NET_ADMIN);
	ret = fe->fill_res_func(msg, has_cap_net_admin, res, port);
	rdma_restrack_put(res);
	if (ret)
		goto err_free;

	nlmsg_end(msg, nlh);
	ib_device_put(device);
	return rdma_nl_unicast(msg, NETLINK_CB(skb).portid);

err_free:
	nlmsg_free(msg);
err_get:
	rdma_restrack_put(res);
err:
	ib_device_put(device);
	return ret;
}

1053 1054 1055
static int res_get_common_dumpit(struct sk_buff *skb,
				 struct netlink_callback *cb,
				 enum rdma_restrack_type res_type)
1056
{
1057
	const struct nldev_fill_res_entry *fe = &fill_entries[res_type];
1058 1059
	struct nlattr *tb[RDMA_NLDEV_ATTR_MAX];
	struct rdma_restrack_entry *res;
1060
	struct rdma_restrack_root *rt;
1061 1062
	int err, ret = 0, idx = 0;
	struct nlattr *table_attr;
1063
	struct nlattr *entry_attr;
1064 1065
	struct ib_device *device;
	int start = cb->args[0];
1066
	bool has_cap_net_admin;
1067
	struct nlmsghdr *nlh;
1068
	unsigned long id;
1069
	u32 index, port = 0;
1070
	bool filled = false;
1071 1072 1073 1074

	err = nlmsg_parse(cb->nlh, 0, tb, RDMA_NLDEV_ATTR_MAX - 1,
			  nldev_policy, NULL);
	/*
1075
	 * Right now, we are expecting the device index to get res information,
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
	 * but it is possible to extend this code to return all devices in
	 * one shot by checking the existence of RDMA_NLDEV_ATTR_DEV_INDEX.
	 * if it doesn't exist, we will iterate over all devices.
	 *
	 * But it is not needed for now.
	 */
	if (err || !tb[RDMA_NLDEV_ATTR_DEV_INDEX])
		return -EINVAL;

	index = nla_get_u32(tb[RDMA_NLDEV_ATTR_DEV_INDEX]);
	device = ib_device_get_by_index(index);
	if (!device)
		return -EINVAL;

	/*
	 * If no PORT_INDEX is supplied, we will return all QPs from that device
	 */
	if (tb[RDMA_NLDEV_ATTR_PORT_INDEX]) {
		port = nla_get_u32(tb[RDMA_NLDEV_ATTR_PORT_INDEX]);
		if (!rdma_is_port_valid(device, port)) {
			ret = -EINVAL;
			goto err_index;
		}
	}

	nlh = nlmsg_put(skb, NETLINK_CB(cb->skb).portid, cb->nlh->nlmsg_seq,
1102
			RDMA_NL_GET_TYPE(RDMA_NL_NLDEV, fe->nldev_cmd),
1103 1104 1105 1106 1107 1108 1109
			0, NLM_F_MULTI);

	if (fill_nldev_handle(skb, device)) {
		ret = -EMSGSIZE;
		goto err;
	}

1110
	table_attr = nla_nest_start(skb, fe->nldev_attr);
1111 1112 1113 1114 1115
	if (!table_attr) {
		ret = -EMSGSIZE;
		goto err;
	}

1116 1117
	has_cap_net_admin = netlink_capable(cb->skb, CAP_NET_ADMIN);

1118 1119
	rt = &device->res[res_type];
	xa_lock(&rt->xa);
1120 1121 1122 1123 1124
	/*
	 * FIXME: if the skip ahead is something common this loop should
	 * use xas_for_each & xas_pause to optimize, we can have a lot of
	 * objects.
	 */
1125
	xa_for_each(&rt->xa, id, res) {
1126 1127 1128
		if (idx < start)
			goto next;

1129
		if (!is_visible_in_pid_ns(res))
1130 1131 1132 1133 1134
			goto next;

		if (!rdma_restrack_get(res))
			goto next;

1135 1136
		xa_unlock(&rt->xa);

1137
		filled = true;
1138

1139 1140 1141 1142
		entry_attr = nla_nest_start(skb, fe->entry);
		if (!entry_attr) {
			ret = -EMSGSIZE;
			rdma_restrack_put(res);
1143
			goto msg_full;
1144 1145
		}

1146
		ret = fe->fill_res_func(skb, has_cap_net_admin, res, port);
1147 1148
		rdma_restrack_put(res);

1149
		if (ret) {
1150
			nla_nest_cancel(skb, entry_attr);
1151 1152 1153 1154
			if (ret == -EMSGSIZE)
				goto msg_full;
			if (ret == -EAGAIN)
				goto again;
1155
			goto res_err;
1156
		}
1157
		nla_nest_end(skb, entry_attr);
1158
again:		xa_lock(&rt->xa);
1159 1160
next:		idx++;
	}
1161
	xa_unlock(&rt->xa);
1162

1163
msg_full:
1164 1165 1166 1167 1168
	nla_nest_end(skb, table_attr);
	nlmsg_end(skb, nlh);
	cb->args[0] = idx;

	/*
1169
	 * No more entries to fill, cancel the message and
1170 1171
	 * return 0 to mark end of dumpit.
	 */
1172
	if (!filled)
1173 1174
		goto err;

1175
	ib_device_put(device);
1176 1177 1178 1179 1180 1181 1182 1183 1184
	return skb->len;

res_err:
	nla_nest_cancel(skb, table_attr);

err:
	nlmsg_cancel(skb, nlh);

err_index:
1185
	ib_device_put(device);
1186 1187 1188
	return ret;
}

1189 1190 1191 1192 1193
#define RES_GET_FUNCS(name, type)                                              \
	static int nldev_res_get_##name##_dumpit(struct sk_buff *skb,          \
						 struct netlink_callback *cb)  \
	{                                                                      \
		return res_get_common_dumpit(skb, cb, type);                   \
1194 1195 1196 1197 1198 1199
	}                                                                      \
	static int nldev_res_get_##name##_doit(struct sk_buff *skb,            \
					       struct nlmsghdr *nlh,           \
					       struct netlink_ext_ack *extack) \
	{                                                                      \
		return res_get_common_doit(skb, nlh, extack, type);            \
1200
	}
1201

1202 1203 1204 1205 1206
RES_GET_FUNCS(qp, RDMA_RESTRACK_QP);
RES_GET_FUNCS(cm_id, RDMA_RESTRACK_CM_ID);
RES_GET_FUNCS(cq, RDMA_RESTRACK_CQ);
RES_GET_FUNCS(pd, RDMA_RESTRACK_PD);
RES_GET_FUNCS(mr, RDMA_RESTRACK_MR);
1207

1208
static const struct rdma_nl_cbs nldev_cb_table[RDMA_NLDEV_NUM_OPS] = {
1209
	[RDMA_NLDEV_CMD_GET] = {
1210
		.doit = nldev_get_doit,
1211 1212
		.dump = nldev_get_dumpit,
	},
1213 1214 1215 1216
	[RDMA_NLDEV_CMD_SET] = {
		.doit = nldev_set_doit,
		.flags = RDMA_NL_ADMIN_PERM,
	},
1217
	[RDMA_NLDEV_CMD_PORT_GET] = {
1218
		.doit = nldev_port_get_doit,
1219 1220
		.dump = nldev_port_get_dumpit,
	},
1221 1222 1223 1224
	[RDMA_NLDEV_CMD_RES_GET] = {
		.doit = nldev_res_get_doit,
		.dump = nldev_res_get_dumpit,
	},
1225
	[RDMA_NLDEV_CMD_RES_QP_GET] = {
1226
		.doit = nldev_res_get_qp_doit,
1227 1228
		.dump = nldev_res_get_qp_dumpit,
	},
1229
	[RDMA_NLDEV_CMD_RES_CM_ID_GET] = {
1230
		.doit = nldev_res_get_cm_id_doit,
1231 1232
		.dump = nldev_res_get_cm_id_dumpit,
	},
1233
	[RDMA_NLDEV_CMD_RES_CQ_GET] = {
1234
		.doit = nldev_res_get_cq_doit,
1235 1236
		.dump = nldev_res_get_cq_dumpit,
	},
1237
	[RDMA_NLDEV_CMD_RES_MR_GET] = {
1238
		.doit = nldev_res_get_mr_doit,
1239 1240
		.dump = nldev_res_get_mr_dumpit,
	},
1241
	[RDMA_NLDEV_CMD_RES_PD_GET] = {
1242
		.doit = nldev_res_get_pd_doit,
1243 1244
		.dump = nldev_res_get_pd_dumpit,
	},
1245 1246
};

1247 1248
void __init nldev_init(void)
{
1249
	rdma_nl_register(RDMA_NL_NLDEV, nldev_cb_table);
1250 1251 1252 1253 1254 1255
}

void __exit nldev_exit(void)
{
	rdma_nl_unregister(RDMA_NL_NLDEV);
}
1256 1257

MODULE_ALIAS_RDMA_NETLINK(RDMA_NL_NLDEV, 5);