virtio_ring.c 13.8 KB
Newer Older
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
/* Virtio ring implementation.
 *
 *  Copyright 2007 Rusty Russell IBM Corporation
 *
 *  This program is free software; you can redistribute it and/or modify
 *  it under the terms of the GNU General Public License as published by
 *  the Free Software Foundation; either version 2 of the License, or
 *  (at your option) any later version.
 *
 *  This program is distributed in the hope that it will be useful,
 *  but WITHOUT ANY WARRANTY; without even the implied warranty of
 *  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
 *  GNU General Public License for more details.
 *
 *  You should have received a copy of the GNU General Public License
 *  along with this program; if not, write to the Free Software
 *  Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA  02110-1301  USA
 */
#include <linux/virtio.h>
#include <linux/virtio_ring.h>
21
#include <linux/virtio_config.h>
22
#include <linux/device.h>
23
#include <linux/slab.h>
24
#include <linux/module.h>
25

26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43
/* virtio guest is communicating with a virtual "device" that actually runs on
 * a host processor.  Memory barriers are used to control SMP effects. */
#ifdef CONFIG_SMP
/* Where possible, use SMP barriers which are more lightweight than mandatory
 * barriers, because mandatory barriers control MMIO effects on accesses
 * through relaxed memory I/O windows (which virtio does not use). */
#define virtio_mb() smp_mb()
#define virtio_rmb() smp_rmb()
#define virtio_wmb() smp_wmb()
#else
/* We must force memory ordering even if guest is UP since host could be
 * running on another CPU, but SMP barriers are defined to barrier() in that
 * configuration. So fall back to mandatory barriers instead. */
#define virtio_mb() mb()
#define virtio_rmb() rmb()
#define virtio_wmb() wmb()
#endif

44 45
#ifdef DEBUG
/* For development, we want to crash whenever the ring is screwed. */
46 47 48 49 50 51
#define BAD_RING(_vq, fmt, args...)				\
	do {							\
		dev_err(&(_vq)->vq.vdev->dev,			\
			"%s:"fmt, (_vq)->vq.name, ##args);	\
		BUG();						\
	} while (0)
52 53 54 55
/* Caller is supposed to guarantee no reentry. */
#define START_USE(_vq)						\
	do {							\
		if ((_vq)->in_use)				\
56 57
			panic("%s:in_use = %i\n",		\
			      (_vq)->vq.name, (_vq)->in_use);	\
58
		(_vq)->in_use = __LINE__;			\
59
	} while (0)
60
#define END_USE(_vq) \
61
	do { BUG_ON(!(_vq)->in_use); (_vq)->in_use = 0; } while(0)
62
#else
63 64 65 66 67 68
#define BAD_RING(_vq, fmt, args...)				\
	do {							\
		dev_err(&_vq->vq.vdev->dev,			\
			"%s:"fmt, (_vq)->vq.name, ##args);	\
		(_vq)->broken = true;				\
	} while (0)
69 70 71 72 73 74 75 76 77 78 79 80 81 82
#define START_USE(vq)
#define END_USE(vq)
#endif

struct vring_virtqueue
{
	struct virtqueue vq;

	/* Actual memory layout for this queue */
	struct vring vring;

	/* Other side has made a mess, don't try any more. */
	bool broken;

83 84 85
	/* Host supports indirect buffers */
	bool indirect;

86 87 88
	/* Host publishes avail event idx */
	bool event;

89 90 91 92 93 94 95 96
	/* Number of free buffers */
	unsigned int num_free;
	/* Head of free buffer list. */
	unsigned int free_head;
	/* Number we've added since last sync. */
	unsigned int num_added;

	/* Last used index we've seen. */
A
Anthony Liguori 已提交
97
	u16 last_used_idx;
98 99 100 101 102 103 104 105 106 107 108 109 110 111 112

	/* How to notify other side. FIXME: commonalize hcalls! */
	void (*notify)(struct virtqueue *vq);

#ifdef DEBUG
	/* They're supposed to lock for us. */
	unsigned int in_use;
#endif

	/* Tokens for callbacks. */
	void *data[];
};

#define to_vvq(_vq) container_of(_vq, struct vring_virtqueue, vq)

113 114 115 116
/* Set up an indirect table of descriptors and add it to the queue. */
static int vring_add_indirect(struct vring_virtqueue *vq,
			      struct scatterlist sg[],
			      unsigned int out,
M
Michael S. Tsirkin 已提交
117 118
			      unsigned int in,
			      gfp_t gfp)
119 120 121 122 123
{
	struct vring_desc *desc;
	unsigned head;
	int i;

M
Michael S. Tsirkin 已提交
124
	desc = kmalloc((out + in) * sizeof(struct vring_desc), gfp);
125
	if (!desc)
126
		return -ENOMEM;
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

	/* Transfer entries from the sg list into the indirect page */
	for (i = 0; i < out; i++) {
		desc[i].flags = VRING_DESC_F_NEXT;
		desc[i].addr = sg_phys(sg);
		desc[i].len = sg->length;
		desc[i].next = i+1;
		sg++;
	}
	for (; i < (out + in); i++) {
		desc[i].flags = VRING_DESC_F_NEXT|VRING_DESC_F_WRITE;
		desc[i].addr = sg_phys(sg);
		desc[i].len = sg->length;
		desc[i].next = i+1;
		sg++;
	}

	/* Last one doesn't continue. */
	desc[i-1].flags &= ~VRING_DESC_F_NEXT;
	desc[i-1].next = 0;

	/* We're about to use a buffer */
	vq->num_free--;

	/* Use a single buffer which doesn't continue */
	head = vq->free_head;
	vq->vring.desc[head].flags = VRING_DESC_F_INDIRECT;
	vq->vring.desc[head].addr = virt_to_phys(desc);
	vq->vring.desc[head].len = i * sizeof(struct vring_desc);

	/* Update free pointer */
	vq->free_head = vq->vring.desc[head].next;

	return head;
}

M
Michael S. Tsirkin 已提交
163 164 165 166 167 168
int virtqueue_add_buf_gfp(struct virtqueue *_vq,
			  struct scatterlist sg[],
			  unsigned int out,
			  unsigned int in,
			  void *data,
			  gfp_t gfp)
169 170
{
	struct vring_virtqueue *vq = to_vvq(_vq);
M
Michael S. Tsirkin 已提交
171 172
	unsigned int i, avail, uninitialized_var(prev);
	int head;
173

174 175
	START_USE(vq);

176
	BUG_ON(data == NULL);
177 178 179 180

	/* If the host supports indirect descriptor tables, and we have multiple
	 * buffers, then go indirect. FIXME: tune this threshold */
	if (vq->indirect && (out + in) > 1 && vq->num_free) {
M
Michael S. Tsirkin 已提交
181
		head = vring_add_indirect(vq, sg, out, in, gfp);
M
Michael S. Tsirkin 已提交
182
		if (likely(head >= 0))
183 184 185
			goto add_head;
	}

186 187 188 189 190 191
	BUG_ON(out + in > vq->vring.num);
	BUG_ON(out + in == 0);

	if (vq->num_free < out + in) {
		pr_debug("Can't add buf len %i - avail = %i\n",
			 out + in, vq->num_free);
192 193 194 195 196
		/* FIXME: for historical reasons, we force a notify here if
		 * there are outgoing parts to the buffer.  Presumably the
		 * host should service the ring ASAP. */
		if (out)
			vq->notify(&vq->vq);
197 198 199 200 201 202 203 204 205 206
		END_USE(vq);
		return -ENOSPC;
	}

	/* We're about to use some buffers from the free list. */
	vq->num_free -= out + in;

	head = vq->free_head;
	for (i = vq->free_head; out; i = vq->vring.desc[i].next, out--) {
		vq->vring.desc[i].flags = VRING_DESC_F_NEXT;
207
		vq->vring.desc[i].addr = sg_phys(sg);
208 209 210 211 212 213
		vq->vring.desc[i].len = sg->length;
		prev = i;
		sg++;
	}
	for (; in; i = vq->vring.desc[i].next, in--) {
		vq->vring.desc[i].flags = VRING_DESC_F_NEXT|VRING_DESC_F_WRITE;
214
		vq->vring.desc[i].addr = sg_phys(sg);
215 216 217 218 219 220 221 222 223 224
		vq->vring.desc[i].len = sg->length;
		prev = i;
		sg++;
	}
	/* Last one doesn't continue. */
	vq->vring.desc[prev].flags &= ~VRING_DESC_F_NEXT;

	/* Update free pointer */
	vq->free_head = i;

225
add_head:
226 227 228 229 230 231 232 233 234 235
	/* Set token. */
	vq->data[head] = data;

	/* Put entry in available array (but don't update avail->idx until they
	 * do sync).  FIXME: avoid modulus here? */
	avail = (vq->vring.avail->idx + vq->num_added++) % vq->vring.num;
	vq->vring.avail->ring[avail] = head;

	pr_debug("Added buffer head %i to %p\n", head, vq);
	END_USE(vq);
236 237

	return vq->num_free;
238
}
M
Michael S. Tsirkin 已提交
239
EXPORT_SYMBOL_GPL(virtqueue_add_buf_gfp);
240

241
void virtqueue_kick(struct virtqueue *_vq)
242 243
{
	struct vring_virtqueue *vq = to_vvq(_vq);
244
	u16 new, old;
245 246 247
	START_USE(vq);
	/* Descriptors and available array need to be set before we expose the
	 * new available array entries. */
248
	virtio_wmb();
249

250 251
	old = vq->vring.avail->idx;
	new = vq->vring.avail->idx = old + vq->num_added;
252 253 254
	vq->num_added = 0;

	/* Need to update avail index before checking if we should notify */
255
	virtio_mb();
256

257 258 259
	if (vq->event ?
	    vring_need_event(vring_avail_event(&vq->vring), new, old) :
	    !(vq->vring.used->flags & VRING_USED_F_NO_NOTIFY))
260 261 262 263 264
		/* Prod other side to tell it about changes. */
		vq->notify(&vq->vq);

	END_USE(vq);
}
265
EXPORT_SYMBOL_GPL(virtqueue_kick);
266 267 268 269 270 271 272 273 274 275

static void detach_buf(struct vring_virtqueue *vq, unsigned int head)
{
	unsigned int i;

	/* Clear data ptr. */
	vq->data[head] = NULL;

	/* Put back on free list: find end */
	i = head;
276 277 278 279 280

	/* Free the indirect table */
	if (vq->vring.desc[i].flags & VRING_DESC_F_INDIRECT)
		kfree(phys_to_virt(vq->vring.desc[i].addr));

281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296
	while (vq->vring.desc[i].flags & VRING_DESC_F_NEXT) {
		i = vq->vring.desc[i].next;
		vq->num_free++;
	}

	vq->vring.desc[i].next = vq->free_head;
	vq->free_head = head;
	/* Plus final descriptor */
	vq->num_free++;
}

static inline bool more_used(const struct vring_virtqueue *vq)
{
	return vq->last_used_idx != vq->vring.used->idx;
}

297
void *virtqueue_get_buf(struct virtqueue *_vq, unsigned int *len)
298 299 300 301 302 303 304
{
	struct vring_virtqueue *vq = to_vvq(_vq);
	void *ret;
	unsigned int i;

	START_USE(vq);

305 306 307 308 309
	if (unlikely(vq->broken)) {
		END_USE(vq);
		return NULL;
	}

310 311 312 313 314 315
	if (!more_used(vq)) {
		pr_debug("No more buffers in queue\n");
		END_USE(vq);
		return NULL;
	}

316
	/* Only get used array entries after they have been exposed by host. */
317
	virtio_rmb();
318

319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334
	i = vq->vring.used->ring[vq->last_used_idx%vq->vring.num].id;
	*len = vq->vring.used->ring[vq->last_used_idx%vq->vring.num].len;

	if (unlikely(i >= vq->vring.num)) {
		BAD_RING(vq, "id %u out of range\n", i);
		return NULL;
	}
	if (unlikely(!vq->data[i])) {
		BAD_RING(vq, "id %u is not a head!\n", i);
		return NULL;
	}

	/* detach_buf clears data, so grab it now. */
	ret = vq->data[i];
	detach_buf(vq, i);
	vq->last_used_idx++;
335 336 337 338 339 340 341 342
	/* If we expect an interrupt for the next entry, tell host
	 * by writing event index and flush out the write before
	 * the read in the next get_buf call. */
	if (!(vq->vring.avail->flags & VRING_AVAIL_F_NO_INTERRUPT)) {
		vring_used_event(&vq->vring) = vq->last_used_idx;
		virtio_mb();
	}

343 344 345
	END_USE(vq);
	return ret;
}
346
EXPORT_SYMBOL_GPL(virtqueue_get_buf);
347

348
void virtqueue_disable_cb(struct virtqueue *_vq)
349 350 351 352 353
{
	struct vring_virtqueue *vq = to_vvq(_vq);

	vq->vring.avail->flags |= VRING_AVAIL_F_NO_INTERRUPT;
}
354
EXPORT_SYMBOL_GPL(virtqueue_disable_cb);
355

356
bool virtqueue_enable_cb(struct virtqueue *_vq)
357 358 359 360 361 362 363
{
	struct vring_virtqueue *vq = to_vvq(_vq);

	START_USE(vq);

	/* We optimistically turn back on interrupts, then check if there was
	 * more to do. */
364 365 366
	/* Depending on the VIRTIO_RING_F_EVENT_IDX feature, we need to
	 * either clear the flags bit or point the event index at the next
	 * entry. Always do both to keep code simple. */
367
	vq->vring.avail->flags &= ~VRING_AVAIL_F_NO_INTERRUPT;
368
	vring_used_event(&vq->vring) = vq->last_used_idx;
369
	virtio_mb();
370 371 372 373 374 375 376 377
	if (unlikely(more_used(vq))) {
		END_USE(vq);
		return false;
	}

	END_USE(vq);
	return true;
}
378
EXPORT_SYMBOL_GPL(virtqueue_enable_cb);
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
bool virtqueue_enable_cb_delayed(struct virtqueue *_vq)
{
	struct vring_virtqueue *vq = to_vvq(_vq);
	u16 bufs;

	START_USE(vq);

	/* We optimistically turn back on interrupts, then check if there was
	 * more to do. */
	/* Depending on the VIRTIO_RING_F_USED_EVENT_IDX feature, we need to
	 * either clear the flags bit or point the event index at the next
	 * entry. Always do both to keep code simple. */
	vq->vring.avail->flags &= ~VRING_AVAIL_F_NO_INTERRUPT;
	/* TODO: tune this threshold */
	bufs = (u16)(vq->vring.avail->idx - vq->last_used_idx) * 3 / 4;
	vring_used_event(&vq->vring) = vq->last_used_idx + bufs;
	virtio_mb();
	if (unlikely((u16)(vq->vring.used->idx - vq->last_used_idx) > bufs)) {
		END_USE(vq);
		return false;
	}

	END_USE(vq);
	return true;
}
EXPORT_SYMBOL_GPL(virtqueue_enable_cb_delayed);

407
void *virtqueue_detach_unused_buf(struct virtqueue *_vq)
408 409 410 411 412 413 414 415 416 417 418 419 420
{
	struct vring_virtqueue *vq = to_vvq(_vq);
	unsigned int i;
	void *buf;

	START_USE(vq);

	for (i = 0; i < vq->vring.num; i++) {
		if (!vq->data[i])
			continue;
		/* detach_buf clears data, so grab it now. */
		buf = vq->data[i];
		detach_buf(vq, i);
421
		vq->vring.avail->idx--;
422 423 424 425 426 427 428 429 430
		END_USE(vq);
		return buf;
	}
	/* That should have freed everything. */
	BUG_ON(vq->num_free != vq->vring.num);

	END_USE(vq);
	return NULL;
}
431
EXPORT_SYMBOL_GPL(virtqueue_detach_unused_buf);
432

433 434 435 436 437 438 439 440 441 442 443 444 445
irqreturn_t vring_interrupt(int irq, void *_vq)
{
	struct vring_virtqueue *vq = to_vvq(_vq);

	if (!more_used(vq)) {
		pr_debug("virtqueue interrupt with no work for %p\n", vq);
		return IRQ_NONE;
	}

	if (unlikely(vq->broken))
		return IRQ_HANDLED;

	pr_debug("virtqueue callback for %p (%p)\n", vq, vq->vq.callback);
446 447
	if (vq->vq.callback)
		vq->vq.callback(&vq->vq);
448 449 450

	return IRQ_HANDLED;
}
451
EXPORT_SYMBOL_GPL(vring_interrupt);
452 453

struct virtqueue *vring_new_virtqueue(unsigned int num,
454
				      unsigned int vring_align,
455 456 457
				      struct virtio_device *vdev,
				      void *pages,
				      void (*notify)(struct virtqueue *),
458 459
				      void (*callback)(struct virtqueue *),
				      const char *name)
460 461 462 463
{
	struct vring_virtqueue *vq;
	unsigned int i;

464 465 466 467 468 469
	/* We assume num is a power of 2. */
	if (num & (num - 1)) {
		dev_warn(&vdev->dev, "Bad virtqueue length %u\n", num);
		return NULL;
	}

470 471 472 473
	vq = kmalloc(sizeof(*vq) + sizeof(void *)*num, GFP_KERNEL);
	if (!vq)
		return NULL;

474
	vring_init(&vq->vring, num, pages, vring_align);
475 476
	vq->vq.callback = callback;
	vq->vq.vdev = vdev;
477
	vq->vq.name = name;
478 479 480 481
	vq->notify = notify;
	vq->broken = false;
	vq->last_used_idx = 0;
	vq->num_added = 0;
482
	list_add_tail(&vq->vq.list, &vdev->vqs);
483 484 485 486
#ifdef DEBUG
	vq->in_use = false;
#endif

487
	vq->indirect = virtio_has_feature(vdev, VIRTIO_RING_F_INDIRECT_DESC);
488
	vq->event = virtio_has_feature(vdev, VIRTIO_RING_F_EVENT_IDX);
489

490 491 492 493 494 495 496
	/* No callback?  Tell other side not to bother us. */
	if (!callback)
		vq->vring.avail->flags |= VRING_AVAIL_F_NO_INTERRUPT;

	/* Put everything in free lists. */
	vq->num_free = num;
	vq->free_head = 0;
497
	for (i = 0; i < num-1; i++) {
498
		vq->vring.desc[i].next = i+1;
499 500 501
		vq->data[i] = NULL;
	}
	vq->data[i] = NULL;
502 503 504

	return &vq->vq;
}
505
EXPORT_SYMBOL_GPL(vring_new_virtqueue);
506 507 508

void vring_del_virtqueue(struct virtqueue *vq)
{
509
	list_del(&vq->list);
510 511
	kfree(to_vvq(vq));
}
512
EXPORT_SYMBOL_GPL(vring_del_virtqueue);
513

514 515 516 517 518 519 520
/* Manipulates transport-specific feature bits. */
void vring_transport_features(struct virtio_device *vdev)
{
	unsigned int i;

	for (i = VIRTIO_TRANSPORT_F_START; i < VIRTIO_TRANSPORT_F_END; i++) {
		switch (i) {
521 522
		case VIRTIO_RING_F_INDIRECT_DESC:
			break;
523 524
		case VIRTIO_RING_F_EVENT_IDX:
			break;
525 526 527 528 529 530 531 532
		default:
			/* We don't understand this bit. */
			clear_bit(i, vdev->features);
		}
	}
}
EXPORT_SYMBOL_GPL(vring_transport_features);

R
Rick Jones 已提交
533 534 535 536 537 538 539 540 541 542
/* return the size of the vring within the virtqueue */
unsigned int virtqueue_get_vring_size(struct virtqueue *_vq)
{

	struct vring_virtqueue *vq = to_vvq(_vq);

	return vq->vring.num;
}
EXPORT_SYMBOL_GPL(virtqueue_get_vring_size);

543
MODULE_LICENSE("GPL");