DPDK logo

Elixir Cross Referencer

  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
 33
 34
 35
 36
 37
 38
 39
 40
 41
 42
 43
 44
 45
 46
 47
 48
 49
 50
 51
 52
 53
 54
 55
 56
 57
 58
 59
 60
 61
 62
 63
 64
 65
 66
 67
 68
 69
 70
 71
 72
 73
 74
 75
 76
 77
 78
 79
 80
 81
 82
 83
 84
 85
 86
 87
 88
 89
 90
 91
 92
 93
 94
 95
 96
 97
 98
 99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
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
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
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
424
425
426
427
428
429
/*-
 *   BSD LICENSE
 *
 *   Copyright(c) 2010-2014 Intel Corporation. All rights reserved.
 *   All rights reserved.
 *
 *   Redistribution and use in source and binary forms, with or without
 *   modification, are permitted provided that the following conditions
 *   are met:
 *
 *     * Redistributions of source code must retain the above copyright
 *       notice, this list of conditions and the following disclaimer.
 *     * Redistributions in binary form must reproduce the above copyright
 *       notice, this list of conditions and the following disclaimer in
 *       the documentation and/or other materials provided with the
 *       distribution.
 *     * Neither the name of Intel Corporation nor the names of its
 *       contributors may be used to endorse or promote products derived
 *       from this software without specific prior written permission.
 *
 *   THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
 *   "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
 *   LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
 *   A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
 *   OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
 *   SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
 *   LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
 *   DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
 *   THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
 *   (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
 *   OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
 */

#include <errno.h>
#include <fuse/cuse_lowlevel.h>
#include <linux/limits.h>
#include <linux/vhost.h>
#include <stdint.h>
#include <string.h>
#include <unistd.h>

#include <rte_ethdev.h>
#include <rte_log.h>
#include <rte_string_fns.h>
#include <rte_virtio_net.h>

#include "virtio-net-cdev.h"
#include "vhost-net.h"
#include "eventfd_copy.h"

#define FUSE_OPT_DUMMY "\0\0"
#define FUSE_OPT_FORE  "-f\0\0"
#define FUSE_OPT_NOMULTI "-s\0\0"

static const uint32_t default_major = 231;
static const uint32_t default_minor = 1;
static const char cuse_device_name[] = "/dev/cuse";
static const char default_cdev[] = "vhost-net";

static struct fuse_session *session;
struct vhost_net_device_ops const *ops;

/*
 * Returns vhost_device_ctx from given fuse_req_t. The index is populated later
 * when the device is added to the device linked list.
 */
static struct vhost_device_ctx
fuse_req_to_vhost_ctx(fuse_req_t req, struct fuse_file_info *fi)
{
	struct vhost_device_ctx ctx;
	struct fuse_ctx const *const req_ctx = fuse_req_ctx(req);

	ctx.pid = req_ctx->pid;
	ctx.fh = fi->fh;

	return ctx;
}

/*
 * When the device is created in QEMU it gets initialised here and
 * added to the device linked list.
 */
static void
vhost_net_open(fuse_req_t req, struct fuse_file_info *fi)
{
	struct vhost_device_ctx ctx = fuse_req_to_vhost_ctx(req, fi);
	int err = 0;

	err = ops->new_device(ctx);
	if (err == -1) {
		fuse_reply_err(req, EPERM);
		return;
	}

	fi->fh = err;

	RTE_LOG(INFO, VHOST_CONFIG,
		"(%"PRIu64") Device configuration started\n", fi->fh);
	fuse_reply_open(req, fi);
}

/*
 * When QEMU is shutdown or killed the device gets released.
 */
static void
vhost_net_release(fuse_req_t req, struct fuse_file_info *fi)
{
	int err = 0;
	struct vhost_device_ctx ctx = fuse_req_to_vhost_ctx(req, fi);

	ops->destroy_device(ctx);
	RTE_LOG(INFO, VHOST_CONFIG, "(%"PRIu64") Device released\n", ctx.fh);
	fuse_reply_err(req, err);
}

/*
 * Boilerplate code for CUSE IOCTL
 * Implicit arguments: ctx, req, result.
 */
#define VHOST_IOCTL(func) do {	\
	result = (func)(ctx);	\
	fuse_reply_ioctl(req, result, NULL, 0);	\
} while (0)

/*
 * Boilerplate IOCTL RETRY
 * Implicit arguments: req.
 */
#define VHOST_IOCTL_RETRY(size_r, size_w) do {	\
	struct iovec iov_r = { arg, (size_r) };	\
	struct iovec iov_w = { arg, (size_w) };	\
	fuse_reply_ioctl_retry(req, &iov_r,	\
		(size_r) ? 1 : 0, &iov_w, (size_w) ? 1 : 0);\
} while (0)

/*
 * Boilerplate code for CUSE Read IOCTL
 * Implicit arguments: ctx, req, result, in_bufsz, in_buf.
 */
#define VHOST_IOCTL_R(type, var, func) do {	\
	if (!in_bufsz) {	\
		VHOST_IOCTL_RETRY(sizeof(type), 0);\
	} else {	\
		(var) = *(const type*)in_buf;	\
		result = func(ctx, &(var));	\
		fuse_reply_ioctl(req, result, NULL, 0);\
	}	\
} while (0)

/*
 * Boilerplate code for CUSE Write IOCTL
 * Implicit arguments: ctx, req, result, out_bufsz.
 */
#define VHOST_IOCTL_W(type, var, func) do {	\
	if (!out_bufsz) {	\
		VHOST_IOCTL_RETRY(0, sizeof(type));\
	} else {	\
		result = (func)(ctx, &(var));\
		fuse_reply_ioctl(req, result, &(var), sizeof(type));\
	} \
} while (0)

/*
 * Boilerplate code for CUSE Read/Write IOCTL
 * Implicit arguments: ctx, req, result, in_bufsz, in_buf.
 */
#define VHOST_IOCTL_RW(type1, var1, type2, var2, func) do {	\
	if (!in_bufsz) {	\
		VHOST_IOCTL_RETRY(sizeof(type1), sizeof(type2));\
	} else {	\
		(var1) = *(const type1*) (in_buf);	\
		result = (func)(ctx, (var1), &(var2));	\
		fuse_reply_ioctl(req, result, &(var2), sizeof(type2));\
	}	\
} while (0)

/*
 * The IOCTLs are handled using CUSE/FUSE in userspace. Depending on the type
 * of IOCTL a buffer is requested to read or to write. This request is handled
 * by FUSE and the buffer is then given to CUSE.
 */
static void
vhost_net_ioctl(fuse_req_t req, int cmd, void *arg,
		struct fuse_file_info *fi, __rte_unused unsigned flags,
		const void *in_buf, size_t in_bufsz, size_t out_bufsz)
{
	struct vhost_device_ctx ctx = fuse_req_to_vhost_ctx(req, fi);
	struct vhost_vring_file file;
	struct vhost_vring_state state;
	struct vhost_vring_addr addr;
	uint64_t features;
	uint32_t index;
	int result = 0;

	switch (cmd) {
	case VHOST_NET_SET_BACKEND:
		LOG_DEBUG(VHOST_CONFIG,
			"(%"PRIu64") IOCTL: VHOST_NET_SET_BACKEND\n", ctx.fh);
		if (!in_buf) {
			VHOST_IOCTL_RETRY(sizeof(file), 0);
			break;
		}
		file = *(const struct vhost_vring_file *)in_buf;
		result = cuse_set_backend(ctx, &file);
		fuse_reply_ioctl(req, result, NULL, 0);
		break;

	case VHOST_GET_FEATURES:
		LOG_DEBUG(VHOST_CONFIG,
			"(%"PRIu64") IOCTL: VHOST_GET_FEATURES\n", ctx.fh);
		VHOST_IOCTL_W(uint64_t, features, ops->get_features);
		break;

	case VHOST_SET_FEATURES:
		LOG_DEBUG(VHOST_CONFIG,
			"(%"PRIu64") IOCTL: VHOST_SET_FEATURES\n", ctx.fh);
		VHOST_IOCTL_R(uint64_t, features, ops->set_features);
		break;

	case VHOST_RESET_OWNER:
		LOG_DEBUG(VHOST_CONFIG,
			"(%"PRIu64") IOCTL: VHOST_RESET_OWNER\n", ctx.fh);
		VHOST_IOCTL(ops->reset_owner);
		break;

	case VHOST_SET_OWNER:
		LOG_DEBUG(VHOST_CONFIG,
			"(%"PRIu64") IOCTL: VHOST_SET_OWNER\n", ctx.fh);
		VHOST_IOCTL(ops->set_owner);
		break;

	case VHOST_SET_MEM_TABLE:
		/*TODO fix race condition.*/
		LOG_DEBUG(VHOST_CONFIG,
			"(%"PRIu64") IOCTL: VHOST_SET_MEM_TABLE\n", ctx.fh);
		static struct vhost_memory mem_temp;

		switch (in_bufsz) {
		case 0:
			VHOST_IOCTL_RETRY(sizeof(struct vhost_memory), 0);
			break;

		case sizeof(struct vhost_memory):
			mem_temp = *(const struct vhost_memory *) in_buf;

			if (mem_temp.nregions > 0) {
				VHOST_IOCTL_RETRY(sizeof(struct vhost_memory) +
					(sizeof(struct vhost_memory_region) *
						mem_temp.nregions), 0);
			} else {
				result = -1;
				fuse_reply_ioctl(req, result, NULL, 0);
			}
			break;

		default:
			result = cuse_set_mem_table(ctx, in_buf,
				mem_temp.nregions);
			if (result)
				fuse_reply_err(req, EINVAL);
			else
				fuse_reply_ioctl(req, result, NULL, 0);
		}
		break;

	case VHOST_SET_VRING_NUM:
		LOG_DEBUG(VHOST_CONFIG,
			"(%"PRIu64") IOCTL: VHOST_SET_VRING_NUM\n", ctx.fh);
		VHOST_IOCTL_R(struct vhost_vring_state, state,
			ops->set_vring_num);
		break;

	case VHOST_SET_VRING_BASE:
		LOG_DEBUG(VHOST_CONFIG,
			"(%"PRIu64") IOCTL: VHOST_SET_VRING_BASE\n", ctx.fh);
		VHOST_IOCTL_R(struct vhost_vring_state, state,
			ops->set_vring_base);
		break;

	case VHOST_GET_VRING_BASE:
		LOG_DEBUG(VHOST_CONFIG,
			"(%"PRIu64") IOCTL: VHOST_GET_VRING_BASE\n", ctx.fh);
		VHOST_IOCTL_RW(uint32_t, index,
			struct vhost_vring_state, state, ops->get_vring_base);
		break;

	case VHOST_SET_VRING_ADDR:
		LOG_DEBUG(VHOST_CONFIG,
			"(%"PRIu64") IOCTL: VHOST_SET_VRING_ADDR\n", ctx.fh);
		VHOST_IOCTL_R(struct vhost_vring_addr, addr,
			ops->set_vring_addr);
		break;

	case VHOST_SET_VRING_KICK:
	case VHOST_SET_VRING_CALL:
		if (cmd == VHOST_SET_VRING_KICK)
			LOG_DEBUG(VHOST_CONFIG,
				"(%"PRIu64") IOCTL: VHOST_SET_VRING_KICK\n",
			ctx.fh);
		else
			LOG_DEBUG(VHOST_CONFIG,
				"(%"PRIu64") IOCTL: VHOST_SET_VRING_CALL\n",
			ctx.fh);
		if (!in_buf)
			VHOST_IOCTL_RETRY(sizeof(struct vhost_vring_file), 0);
		else {
			int fd;
			file = *(const struct vhost_vring_file *)in_buf;
			LOG_DEBUG(VHOST_CONFIG,
				"idx:%d fd:%d\n", file.index, file.fd);
			fd = eventfd_copy(file.fd, ctx.pid);
			if (fd < 0) {
				fuse_reply_ioctl(req, -1, NULL, 0);
				result = -1;
				break;
			}
			file.fd = fd;
			if (cmd == VHOST_SET_VRING_KICK) {
				result = ops->set_vring_kick(ctx, &file);
				fuse_reply_ioctl(req, result, NULL, 0);
			} else {
				result = ops->set_vring_call(ctx, &file);
				fuse_reply_ioctl(req, result, NULL, 0);
			}
		}
		break;

	default:
		RTE_LOG(ERR, VHOST_CONFIG,
			"(%"PRIu64") IOCTL: DOESN NOT EXIST\n", ctx.fh);
		result = -1;
		fuse_reply_ioctl(req, result, NULL, 0);
	}

	if (result < 0)
		LOG_DEBUG(VHOST_CONFIG,
			"(%"PRIu64") IOCTL: FAIL\n", ctx.fh);
	else
		LOG_DEBUG(VHOST_CONFIG,
			"(%"PRIu64") IOCTL: SUCCESS\n", ctx.fh);
}

/*
 * Structure handling open, release and ioctl function pointers is populated.
 */
static const struct cuse_lowlevel_ops vhost_net_ops = {
	.open		= vhost_net_open,
	.release	= vhost_net_release,
	.ioctl		= vhost_net_ioctl,
};

/*
 * cuse_info is populated and used to register the cuse device.
 * vhost_net_device_ops are also passed when the device is registered in app.
 */
int
rte_vhost_driver_register(const char *dev_name)
{
	struct cuse_info cuse_info;
	char device_name[PATH_MAX] = "";
	char char_device_name[PATH_MAX] = "";
	const char *device_argv[] = { device_name };

	char fuse_opt_dummy[] = FUSE_OPT_DUMMY;
	char fuse_opt_fore[] = FUSE_OPT_FORE;
	char fuse_opt_nomulti[] = FUSE_OPT_NOMULTI;
	char *fuse_argv[] = {fuse_opt_dummy, fuse_opt_fore, fuse_opt_nomulti};

	if (access(cuse_device_name, R_OK | W_OK) < 0) {
		RTE_LOG(ERR, VHOST_CONFIG,
			"char device %s can't be accessed, maybe not exist\n",
			cuse_device_name);
		return -1;
	}

	if (eventfd_init() < 0)
		return -1;

	/*
	 * The device name is created. This is passed to QEMU so that it can
	 * register the device with our application.
	 */
	snprintf(device_name, PATH_MAX, "DEVNAME=%s", dev_name);
	snprintf(char_device_name, PATH_MAX, "/dev/%s", dev_name);

	/* Check if device already exists. */
	if (access(char_device_name, F_OK) != -1) {
		RTE_LOG(ERR, VHOST_CONFIG,
			"char device %s already exists\n", char_device_name);
		return -1;
	}

	memset(&cuse_info, 0, sizeof(cuse_info));
	cuse_info.dev_major = default_major;
	cuse_info.dev_minor = default_minor;
	cuse_info.dev_info_argc = 1;
	cuse_info.dev_info_argv = device_argv;
	cuse_info.flags = CUSE_UNRESTRICTED_IOCTL;

	ops = get_virtio_net_callbacks();

	session = cuse_lowlevel_setup(3, fuse_argv,
			&cuse_info, &vhost_net_ops, 0, NULL);
	if (session == NULL)
		return -1;

	return 0;
}

/**
 * An empty function for unregister
 */
int
rte_vhost_driver_unregister(const char *dev_name __rte_unused)
{
	return 0;
}

/**
 * The CUSE session is launched allowing the application to receive open,
 * release and ioctl calls.
 */
int
rte_vhost_driver_session_start(void)
{
	fuse_session_loop(session);

	return 0;
}