Linux kernel mirror (for testing)
git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git
kernel
os
linux
1// SPDX-License-Identifier: GPL-2.0
2/*
3 * Copyright (C) 2018-2020 Intel Corporation.
4 * Copyright (C) 2020 Red Hat, Inc.
5 *
6 * Author: Tiwei Bie <tiwei.bie@intel.com>
7 * Jason Wang <jasowang@redhat.com>
8 *
9 * Thanks Michael S. Tsirkin for the valuable comments and
10 * suggestions. And thanks to Cunming Liang and Zhihong Wang for all
11 * their supports.
12 */
13
14#include <linux/kernel.h>
15#include <linux/module.h>
16#include <linux/cdev.h>
17#include <linux/device.h>
18#include <linux/mm.h>
19#include <linux/slab.h>
20#include <linux/iommu.h>
21#include <linux/uuid.h>
22#include <linux/vdpa.h>
23#include <linux/nospec.h>
24#include <linux/vhost.h>
25
26#include "vhost.h"
27
28enum {
29 VHOST_VDPA_BACKEND_FEATURES =
30 (1ULL << VHOST_BACKEND_F_IOTLB_MSG_V2) |
31 (1ULL << VHOST_BACKEND_F_IOTLB_BATCH) |
32 (1ULL << VHOST_BACKEND_F_IOTLB_ASID),
33};
34
35#define VHOST_VDPA_DEV_MAX (1U << MINORBITS)
36
37#define VHOST_VDPA_IOTLB_BUCKETS 16
38
39struct vhost_vdpa_as {
40 struct hlist_node hash_link;
41 struct vhost_iotlb iotlb;
42 u32 id;
43};
44
45struct vhost_vdpa {
46 struct vhost_dev vdev;
47 struct iommu_domain *domain;
48 struct vhost_virtqueue *vqs;
49 struct completion completion;
50 struct vdpa_device *vdpa;
51 struct hlist_head as[VHOST_VDPA_IOTLB_BUCKETS];
52 struct device dev;
53 struct cdev cdev;
54 atomic_t opened;
55 u32 nvqs;
56 int virtio_id;
57 int minor;
58 struct eventfd_ctx *config_ctx;
59 int in_batch;
60 struct vdpa_iova_range range;
61 u32 batch_asid;
62};
63
64static DEFINE_IDA(vhost_vdpa_ida);
65
66static dev_t vhost_vdpa_major;
67
68static void vhost_vdpa_iotlb_unmap(struct vhost_vdpa *v,
69 struct vhost_iotlb *iotlb, u64 start,
70 u64 last, u32 asid);
71
72static inline u32 iotlb_to_asid(struct vhost_iotlb *iotlb)
73{
74 struct vhost_vdpa_as *as = container_of(iotlb, struct
75 vhost_vdpa_as, iotlb);
76 return as->id;
77}
78
79static struct vhost_vdpa_as *asid_to_as(struct vhost_vdpa *v, u32 asid)
80{
81 struct hlist_head *head = &v->as[asid % VHOST_VDPA_IOTLB_BUCKETS];
82 struct vhost_vdpa_as *as;
83
84 hlist_for_each_entry(as, head, hash_link)
85 if (as->id == asid)
86 return as;
87
88 return NULL;
89}
90
91static struct vhost_iotlb *asid_to_iotlb(struct vhost_vdpa *v, u32 asid)
92{
93 struct vhost_vdpa_as *as = asid_to_as(v, asid);
94
95 if (!as)
96 return NULL;
97
98 return &as->iotlb;
99}
100
101static struct vhost_vdpa_as *vhost_vdpa_alloc_as(struct vhost_vdpa *v, u32 asid)
102{
103 struct hlist_head *head = &v->as[asid % VHOST_VDPA_IOTLB_BUCKETS];
104 struct vhost_vdpa_as *as;
105
106 if (asid_to_as(v, asid))
107 return NULL;
108
109 if (asid >= v->vdpa->nas)
110 return NULL;
111
112 as = kmalloc(sizeof(*as), GFP_KERNEL);
113 if (!as)
114 return NULL;
115
116 vhost_iotlb_init(&as->iotlb, 0, 0);
117 as->id = asid;
118 hlist_add_head(&as->hash_link, head);
119
120 return as;
121}
122
123static struct vhost_vdpa_as *vhost_vdpa_find_alloc_as(struct vhost_vdpa *v,
124 u32 asid)
125{
126 struct vhost_vdpa_as *as = asid_to_as(v, asid);
127
128 if (as)
129 return as;
130
131 return vhost_vdpa_alloc_as(v, asid);
132}
133
134static int vhost_vdpa_remove_as(struct vhost_vdpa *v, u32 asid)
135{
136 struct vhost_vdpa_as *as = asid_to_as(v, asid);
137
138 if (!as)
139 return -EINVAL;
140
141 hlist_del(&as->hash_link);
142 vhost_vdpa_iotlb_unmap(v, &as->iotlb, 0ULL, 0ULL - 1, asid);
143 kfree(as);
144
145 return 0;
146}
147
148static void handle_vq_kick(struct vhost_work *work)
149{
150 struct vhost_virtqueue *vq = container_of(work, struct vhost_virtqueue,
151 poll.work);
152 struct vhost_vdpa *v = container_of(vq->dev, struct vhost_vdpa, vdev);
153 const struct vdpa_config_ops *ops = v->vdpa->config;
154
155 ops->kick_vq(v->vdpa, vq - v->vqs);
156}
157
158static irqreturn_t vhost_vdpa_virtqueue_cb(void *private)
159{
160 struct vhost_virtqueue *vq = private;
161 struct eventfd_ctx *call_ctx = vq->call_ctx.ctx;
162
163 if (call_ctx)
164 eventfd_signal(call_ctx, 1);
165
166 return IRQ_HANDLED;
167}
168
169static irqreturn_t vhost_vdpa_config_cb(void *private)
170{
171 struct vhost_vdpa *v = private;
172 struct eventfd_ctx *config_ctx = v->config_ctx;
173
174 if (config_ctx)
175 eventfd_signal(config_ctx, 1);
176
177 return IRQ_HANDLED;
178}
179
180static void vhost_vdpa_setup_vq_irq(struct vhost_vdpa *v, u16 qid)
181{
182 struct vhost_virtqueue *vq = &v->vqs[qid];
183 const struct vdpa_config_ops *ops = v->vdpa->config;
184 struct vdpa_device *vdpa = v->vdpa;
185 int ret, irq;
186
187 if (!ops->get_vq_irq)
188 return;
189
190 irq = ops->get_vq_irq(vdpa, qid);
191 if (irq < 0)
192 return;
193
194 irq_bypass_unregister_producer(&vq->call_ctx.producer);
195 if (!vq->call_ctx.ctx)
196 return;
197
198 vq->call_ctx.producer.token = vq->call_ctx.ctx;
199 vq->call_ctx.producer.irq = irq;
200 ret = irq_bypass_register_producer(&vq->call_ctx.producer);
201 if (unlikely(ret))
202 dev_info(&v->dev, "vq %u, irq bypass producer (token %p) registration fails, ret = %d\n",
203 qid, vq->call_ctx.producer.token, ret);
204}
205
206static void vhost_vdpa_unsetup_vq_irq(struct vhost_vdpa *v, u16 qid)
207{
208 struct vhost_virtqueue *vq = &v->vqs[qid];
209
210 irq_bypass_unregister_producer(&vq->call_ctx.producer);
211}
212
213static int vhost_vdpa_reset(struct vhost_vdpa *v)
214{
215 struct vdpa_device *vdpa = v->vdpa;
216
217 v->in_batch = 0;
218
219 return vdpa_reset(vdpa);
220}
221
222static long vhost_vdpa_bind_mm(struct vhost_vdpa *v)
223{
224 struct vdpa_device *vdpa = v->vdpa;
225 const struct vdpa_config_ops *ops = vdpa->config;
226
227 if (!vdpa->use_va || !ops->bind_mm)
228 return 0;
229
230 return ops->bind_mm(vdpa, v->vdev.mm);
231}
232
233static void vhost_vdpa_unbind_mm(struct vhost_vdpa *v)
234{
235 struct vdpa_device *vdpa = v->vdpa;
236 const struct vdpa_config_ops *ops = vdpa->config;
237
238 if (!vdpa->use_va || !ops->unbind_mm)
239 return;
240
241 ops->unbind_mm(vdpa);
242}
243
244static long vhost_vdpa_get_device_id(struct vhost_vdpa *v, u8 __user *argp)
245{
246 struct vdpa_device *vdpa = v->vdpa;
247 const struct vdpa_config_ops *ops = vdpa->config;
248 u32 device_id;
249
250 device_id = ops->get_device_id(vdpa);
251
252 if (copy_to_user(argp, &device_id, sizeof(device_id)))
253 return -EFAULT;
254
255 return 0;
256}
257
258static long vhost_vdpa_get_status(struct vhost_vdpa *v, u8 __user *statusp)
259{
260 struct vdpa_device *vdpa = v->vdpa;
261 const struct vdpa_config_ops *ops = vdpa->config;
262 u8 status;
263
264 status = ops->get_status(vdpa);
265
266 if (copy_to_user(statusp, &status, sizeof(status)))
267 return -EFAULT;
268
269 return 0;
270}
271
272static long vhost_vdpa_set_status(struct vhost_vdpa *v, u8 __user *statusp)
273{
274 struct vdpa_device *vdpa = v->vdpa;
275 const struct vdpa_config_ops *ops = vdpa->config;
276 u8 status, status_old;
277 u32 nvqs = v->nvqs;
278 int ret;
279 u16 i;
280
281 if (copy_from_user(&status, statusp, sizeof(status)))
282 return -EFAULT;
283
284 status_old = ops->get_status(vdpa);
285
286 /*
287 * Userspace shouldn't remove status bits unless reset the
288 * status to 0.
289 */
290 if (status != 0 && (status_old & ~status) != 0)
291 return -EINVAL;
292
293 if ((status_old & VIRTIO_CONFIG_S_DRIVER_OK) && !(status & VIRTIO_CONFIG_S_DRIVER_OK))
294 for (i = 0; i < nvqs; i++)
295 vhost_vdpa_unsetup_vq_irq(v, i);
296
297 if (status == 0) {
298 ret = vdpa_reset(vdpa);
299 if (ret)
300 return ret;
301 } else
302 vdpa_set_status(vdpa, status);
303
304 if ((status & VIRTIO_CONFIG_S_DRIVER_OK) && !(status_old & VIRTIO_CONFIG_S_DRIVER_OK))
305 for (i = 0; i < nvqs; i++)
306 vhost_vdpa_setup_vq_irq(v, i);
307
308 return 0;
309}
310
311static int vhost_vdpa_config_validate(struct vhost_vdpa *v,
312 struct vhost_vdpa_config *c)
313{
314 struct vdpa_device *vdpa = v->vdpa;
315 size_t size = vdpa->config->get_config_size(vdpa);
316
317 if (c->len == 0 || c->off > size)
318 return -EINVAL;
319
320 if (c->len > size - c->off)
321 return -E2BIG;
322
323 return 0;
324}
325
326static long vhost_vdpa_get_config(struct vhost_vdpa *v,
327 struct vhost_vdpa_config __user *c)
328{
329 struct vdpa_device *vdpa = v->vdpa;
330 struct vhost_vdpa_config config;
331 unsigned long size = offsetof(struct vhost_vdpa_config, buf);
332 u8 *buf;
333
334 if (copy_from_user(&config, c, size))
335 return -EFAULT;
336 if (vhost_vdpa_config_validate(v, &config))
337 return -EINVAL;
338 buf = kvzalloc(config.len, GFP_KERNEL);
339 if (!buf)
340 return -ENOMEM;
341
342 vdpa_get_config(vdpa, config.off, buf, config.len);
343
344 if (copy_to_user(c->buf, buf, config.len)) {
345 kvfree(buf);
346 return -EFAULT;
347 }
348
349 kvfree(buf);
350 return 0;
351}
352
353static long vhost_vdpa_set_config(struct vhost_vdpa *v,
354 struct vhost_vdpa_config __user *c)
355{
356 struct vdpa_device *vdpa = v->vdpa;
357 struct vhost_vdpa_config config;
358 unsigned long size = offsetof(struct vhost_vdpa_config, buf);
359 u8 *buf;
360
361 if (copy_from_user(&config, c, size))
362 return -EFAULT;
363 if (vhost_vdpa_config_validate(v, &config))
364 return -EINVAL;
365
366 buf = vmemdup_user(c->buf, config.len);
367 if (IS_ERR(buf))
368 return PTR_ERR(buf);
369
370 vdpa_set_config(vdpa, config.off, buf, config.len);
371
372 kvfree(buf);
373 return 0;
374}
375
376static bool vhost_vdpa_can_suspend(const struct vhost_vdpa *v)
377{
378 struct vdpa_device *vdpa = v->vdpa;
379 const struct vdpa_config_ops *ops = vdpa->config;
380
381 return ops->suspend;
382}
383
384static bool vhost_vdpa_can_resume(const struct vhost_vdpa *v)
385{
386 struct vdpa_device *vdpa = v->vdpa;
387 const struct vdpa_config_ops *ops = vdpa->config;
388
389 return ops->resume;
390}
391
392static long vhost_vdpa_get_features(struct vhost_vdpa *v, u64 __user *featurep)
393{
394 struct vdpa_device *vdpa = v->vdpa;
395 const struct vdpa_config_ops *ops = vdpa->config;
396 u64 features;
397
398 features = ops->get_device_features(vdpa);
399
400 if (copy_to_user(featurep, &features, sizeof(features)))
401 return -EFAULT;
402
403 return 0;
404}
405
406static long vhost_vdpa_set_features(struct vhost_vdpa *v, u64 __user *featurep)
407{
408 struct vdpa_device *vdpa = v->vdpa;
409 const struct vdpa_config_ops *ops = vdpa->config;
410 struct vhost_dev *d = &v->vdev;
411 u64 actual_features;
412 u64 features;
413 int i;
414
415 /*
416 * It's not allowed to change the features after they have
417 * been negotiated.
418 */
419 if (ops->get_status(vdpa) & VIRTIO_CONFIG_S_FEATURES_OK)
420 return -EBUSY;
421
422 if (copy_from_user(&features, featurep, sizeof(features)))
423 return -EFAULT;
424
425 if (vdpa_set_features(vdpa, features))
426 return -EINVAL;
427
428 /* let the vqs know what has been configured */
429 actual_features = ops->get_driver_features(vdpa);
430 for (i = 0; i < d->nvqs; ++i) {
431 struct vhost_virtqueue *vq = d->vqs[i];
432
433 mutex_lock(&vq->mutex);
434 vq->acked_features = actual_features;
435 mutex_unlock(&vq->mutex);
436 }
437
438 return 0;
439}
440
441static long vhost_vdpa_get_vring_num(struct vhost_vdpa *v, u16 __user *argp)
442{
443 struct vdpa_device *vdpa = v->vdpa;
444 const struct vdpa_config_ops *ops = vdpa->config;
445 u16 num;
446
447 num = ops->get_vq_num_max(vdpa);
448
449 if (copy_to_user(argp, &num, sizeof(num)))
450 return -EFAULT;
451
452 return 0;
453}
454
455static void vhost_vdpa_config_put(struct vhost_vdpa *v)
456{
457 if (v->config_ctx) {
458 eventfd_ctx_put(v->config_ctx);
459 v->config_ctx = NULL;
460 }
461}
462
463static long vhost_vdpa_set_config_call(struct vhost_vdpa *v, u32 __user *argp)
464{
465 struct vdpa_callback cb;
466 int fd;
467 struct eventfd_ctx *ctx;
468
469 cb.callback = vhost_vdpa_config_cb;
470 cb.private = v;
471 if (copy_from_user(&fd, argp, sizeof(fd)))
472 return -EFAULT;
473
474 ctx = fd == VHOST_FILE_UNBIND ? NULL : eventfd_ctx_fdget(fd);
475 swap(ctx, v->config_ctx);
476
477 if (!IS_ERR_OR_NULL(ctx))
478 eventfd_ctx_put(ctx);
479
480 if (IS_ERR(v->config_ctx)) {
481 long ret = PTR_ERR(v->config_ctx);
482
483 v->config_ctx = NULL;
484 return ret;
485 }
486
487 v->vdpa->config->set_config_cb(v->vdpa, &cb);
488
489 return 0;
490}
491
492static long vhost_vdpa_get_iova_range(struct vhost_vdpa *v, u32 __user *argp)
493{
494 struct vhost_vdpa_iova_range range = {
495 .first = v->range.first,
496 .last = v->range.last,
497 };
498
499 if (copy_to_user(argp, &range, sizeof(range)))
500 return -EFAULT;
501 return 0;
502}
503
504static long vhost_vdpa_get_config_size(struct vhost_vdpa *v, u32 __user *argp)
505{
506 struct vdpa_device *vdpa = v->vdpa;
507 const struct vdpa_config_ops *ops = vdpa->config;
508 u32 size;
509
510 size = ops->get_config_size(vdpa);
511
512 if (copy_to_user(argp, &size, sizeof(size)))
513 return -EFAULT;
514
515 return 0;
516}
517
518static long vhost_vdpa_get_vqs_count(struct vhost_vdpa *v, u32 __user *argp)
519{
520 struct vdpa_device *vdpa = v->vdpa;
521
522 if (copy_to_user(argp, &vdpa->nvqs, sizeof(vdpa->nvqs)))
523 return -EFAULT;
524
525 return 0;
526}
527
528/* After a successful return of ioctl the device must not process more
529 * virtqueue descriptors. The device can answer to read or writes of config
530 * fields as if it were not suspended. In particular, writing to "queue_enable"
531 * with a value of 1 will not make the device start processing buffers.
532 */
533static long vhost_vdpa_suspend(struct vhost_vdpa *v)
534{
535 struct vdpa_device *vdpa = v->vdpa;
536 const struct vdpa_config_ops *ops = vdpa->config;
537
538 if (!ops->suspend)
539 return -EOPNOTSUPP;
540
541 return ops->suspend(vdpa);
542}
543
544/* After a successful return of this ioctl the device resumes processing
545 * virtqueue descriptors. The device becomes fully operational the same way it
546 * was before it was suspended.
547 */
548static long vhost_vdpa_resume(struct vhost_vdpa *v)
549{
550 struct vdpa_device *vdpa = v->vdpa;
551 const struct vdpa_config_ops *ops = vdpa->config;
552
553 if (!ops->resume)
554 return -EOPNOTSUPP;
555
556 return ops->resume(vdpa);
557}
558
559static long vhost_vdpa_vring_ioctl(struct vhost_vdpa *v, unsigned int cmd,
560 void __user *argp)
561{
562 struct vdpa_device *vdpa = v->vdpa;
563 const struct vdpa_config_ops *ops = vdpa->config;
564 struct vdpa_vq_state vq_state;
565 struct vdpa_callback cb;
566 struct vhost_virtqueue *vq;
567 struct vhost_vring_state s;
568 u32 idx;
569 long r;
570
571 r = get_user(idx, (u32 __user *)argp);
572 if (r < 0)
573 return r;
574
575 if (idx >= v->nvqs)
576 return -ENOBUFS;
577
578 idx = array_index_nospec(idx, v->nvqs);
579 vq = &v->vqs[idx];
580
581 switch (cmd) {
582 case VHOST_VDPA_SET_VRING_ENABLE:
583 if (copy_from_user(&s, argp, sizeof(s)))
584 return -EFAULT;
585 ops->set_vq_ready(vdpa, idx, s.num);
586 return 0;
587 case VHOST_VDPA_GET_VRING_GROUP:
588 if (!ops->get_vq_group)
589 return -EOPNOTSUPP;
590 s.index = idx;
591 s.num = ops->get_vq_group(vdpa, idx);
592 if (s.num >= vdpa->ngroups)
593 return -EIO;
594 else if (copy_to_user(argp, &s, sizeof(s)))
595 return -EFAULT;
596 return 0;
597 case VHOST_VDPA_SET_GROUP_ASID:
598 if (copy_from_user(&s, argp, sizeof(s)))
599 return -EFAULT;
600 if (s.num >= vdpa->nas)
601 return -EINVAL;
602 if (!ops->set_group_asid)
603 return -EOPNOTSUPP;
604 return ops->set_group_asid(vdpa, idx, s.num);
605 case VHOST_GET_VRING_BASE:
606 r = ops->get_vq_state(v->vdpa, idx, &vq_state);
607 if (r)
608 return r;
609
610 if (vhost_has_feature(vq, VIRTIO_F_RING_PACKED)) {
611 vq->last_avail_idx = vq_state.packed.last_avail_idx |
612 (vq_state.packed.last_avail_counter << 15);
613 vq->last_used_idx = vq_state.packed.last_used_idx |
614 (vq_state.packed.last_used_counter << 15);
615 } else {
616 vq->last_avail_idx = vq_state.split.avail_index;
617 }
618 break;
619 }
620
621 r = vhost_vring_ioctl(&v->vdev, cmd, argp);
622 if (r)
623 return r;
624
625 switch (cmd) {
626 case VHOST_SET_VRING_ADDR:
627 if (ops->set_vq_address(vdpa, idx,
628 (u64)(uintptr_t)vq->desc,
629 (u64)(uintptr_t)vq->avail,
630 (u64)(uintptr_t)vq->used))
631 r = -EINVAL;
632 break;
633
634 case VHOST_SET_VRING_BASE:
635 if (vhost_has_feature(vq, VIRTIO_F_RING_PACKED)) {
636 vq_state.packed.last_avail_idx = vq->last_avail_idx & 0x7fff;
637 vq_state.packed.last_avail_counter = !!(vq->last_avail_idx & 0x8000);
638 vq_state.packed.last_used_idx = vq->last_used_idx & 0x7fff;
639 vq_state.packed.last_used_counter = !!(vq->last_used_idx & 0x8000);
640 } else {
641 vq_state.split.avail_index = vq->last_avail_idx;
642 }
643 r = ops->set_vq_state(vdpa, idx, &vq_state);
644 break;
645
646 case VHOST_SET_VRING_CALL:
647 if (vq->call_ctx.ctx) {
648 cb.callback = vhost_vdpa_virtqueue_cb;
649 cb.private = vq;
650 cb.trigger = vq->call_ctx.ctx;
651 } else {
652 cb.callback = NULL;
653 cb.private = NULL;
654 cb.trigger = NULL;
655 }
656 ops->set_vq_cb(vdpa, idx, &cb);
657 vhost_vdpa_setup_vq_irq(v, idx);
658 break;
659
660 case VHOST_SET_VRING_NUM:
661 ops->set_vq_num(vdpa, idx, vq->num);
662 break;
663 }
664
665 return r;
666}
667
668static long vhost_vdpa_unlocked_ioctl(struct file *filep,
669 unsigned int cmd, unsigned long arg)
670{
671 struct vhost_vdpa *v = filep->private_data;
672 struct vhost_dev *d = &v->vdev;
673 void __user *argp = (void __user *)arg;
674 u64 __user *featurep = argp;
675 u64 features;
676 long r = 0;
677
678 if (cmd == VHOST_SET_BACKEND_FEATURES) {
679 if (copy_from_user(&features, featurep, sizeof(features)))
680 return -EFAULT;
681 if (features & ~(VHOST_VDPA_BACKEND_FEATURES |
682 BIT_ULL(VHOST_BACKEND_F_SUSPEND) |
683 BIT_ULL(VHOST_BACKEND_F_RESUME)))
684 return -EOPNOTSUPP;
685 if ((features & BIT_ULL(VHOST_BACKEND_F_SUSPEND)) &&
686 !vhost_vdpa_can_suspend(v))
687 return -EOPNOTSUPP;
688 if ((features & BIT_ULL(VHOST_BACKEND_F_RESUME)) &&
689 !vhost_vdpa_can_resume(v))
690 return -EOPNOTSUPP;
691 vhost_set_backend_features(&v->vdev, features);
692 return 0;
693 }
694
695 mutex_lock(&d->mutex);
696
697 switch (cmd) {
698 case VHOST_VDPA_GET_DEVICE_ID:
699 r = vhost_vdpa_get_device_id(v, argp);
700 break;
701 case VHOST_VDPA_GET_STATUS:
702 r = vhost_vdpa_get_status(v, argp);
703 break;
704 case VHOST_VDPA_SET_STATUS:
705 r = vhost_vdpa_set_status(v, argp);
706 break;
707 case VHOST_VDPA_GET_CONFIG:
708 r = vhost_vdpa_get_config(v, argp);
709 break;
710 case VHOST_VDPA_SET_CONFIG:
711 r = vhost_vdpa_set_config(v, argp);
712 break;
713 case VHOST_GET_FEATURES:
714 r = vhost_vdpa_get_features(v, argp);
715 break;
716 case VHOST_SET_FEATURES:
717 r = vhost_vdpa_set_features(v, argp);
718 break;
719 case VHOST_VDPA_GET_VRING_NUM:
720 r = vhost_vdpa_get_vring_num(v, argp);
721 break;
722 case VHOST_VDPA_GET_GROUP_NUM:
723 if (copy_to_user(argp, &v->vdpa->ngroups,
724 sizeof(v->vdpa->ngroups)))
725 r = -EFAULT;
726 break;
727 case VHOST_VDPA_GET_AS_NUM:
728 if (copy_to_user(argp, &v->vdpa->nas, sizeof(v->vdpa->nas)))
729 r = -EFAULT;
730 break;
731 case VHOST_SET_LOG_BASE:
732 case VHOST_SET_LOG_FD:
733 r = -ENOIOCTLCMD;
734 break;
735 case VHOST_VDPA_SET_CONFIG_CALL:
736 r = vhost_vdpa_set_config_call(v, argp);
737 break;
738 case VHOST_GET_BACKEND_FEATURES:
739 features = VHOST_VDPA_BACKEND_FEATURES;
740 if (vhost_vdpa_can_suspend(v))
741 features |= BIT_ULL(VHOST_BACKEND_F_SUSPEND);
742 if (vhost_vdpa_can_resume(v))
743 features |= BIT_ULL(VHOST_BACKEND_F_RESUME);
744 if (copy_to_user(featurep, &features, sizeof(features)))
745 r = -EFAULT;
746 break;
747 case VHOST_VDPA_GET_IOVA_RANGE:
748 r = vhost_vdpa_get_iova_range(v, argp);
749 break;
750 case VHOST_VDPA_GET_CONFIG_SIZE:
751 r = vhost_vdpa_get_config_size(v, argp);
752 break;
753 case VHOST_VDPA_GET_VQS_COUNT:
754 r = vhost_vdpa_get_vqs_count(v, argp);
755 break;
756 case VHOST_VDPA_SUSPEND:
757 r = vhost_vdpa_suspend(v);
758 break;
759 case VHOST_VDPA_RESUME:
760 r = vhost_vdpa_resume(v);
761 break;
762 default:
763 r = vhost_dev_ioctl(&v->vdev, cmd, argp);
764 if (r == -ENOIOCTLCMD)
765 r = vhost_vdpa_vring_ioctl(v, cmd, argp);
766 break;
767 }
768
769 if (r)
770 goto out;
771
772 switch (cmd) {
773 case VHOST_SET_OWNER:
774 r = vhost_vdpa_bind_mm(v);
775 if (r)
776 vhost_dev_reset_owner(d, NULL);
777 break;
778 }
779out:
780 mutex_unlock(&d->mutex);
781 return r;
782}
783static void vhost_vdpa_general_unmap(struct vhost_vdpa *v,
784 struct vhost_iotlb_map *map, u32 asid)
785{
786 struct vdpa_device *vdpa = v->vdpa;
787 const struct vdpa_config_ops *ops = vdpa->config;
788 if (ops->dma_map) {
789 ops->dma_unmap(vdpa, asid, map->start, map->size);
790 } else if (ops->set_map == NULL) {
791 iommu_unmap(v->domain, map->start, map->size);
792 }
793}
794
795static void vhost_vdpa_pa_unmap(struct vhost_vdpa *v, struct vhost_iotlb *iotlb,
796 u64 start, u64 last, u32 asid)
797{
798 struct vhost_dev *dev = &v->vdev;
799 struct vhost_iotlb_map *map;
800 struct page *page;
801 unsigned long pfn, pinned;
802
803 while ((map = vhost_iotlb_itree_first(iotlb, start, last)) != NULL) {
804 pinned = PFN_DOWN(map->size);
805 for (pfn = PFN_DOWN(map->addr);
806 pinned > 0; pfn++, pinned--) {
807 page = pfn_to_page(pfn);
808 if (map->perm & VHOST_ACCESS_WO)
809 set_page_dirty_lock(page);
810 unpin_user_page(page);
811 }
812 atomic64_sub(PFN_DOWN(map->size), &dev->mm->pinned_vm);
813 vhost_vdpa_general_unmap(v, map, asid);
814 vhost_iotlb_map_free(iotlb, map);
815 }
816}
817
818static void vhost_vdpa_va_unmap(struct vhost_vdpa *v, struct vhost_iotlb *iotlb,
819 u64 start, u64 last, u32 asid)
820{
821 struct vhost_iotlb_map *map;
822 struct vdpa_map_file *map_file;
823
824 while ((map = vhost_iotlb_itree_first(iotlb, start, last)) != NULL) {
825 map_file = (struct vdpa_map_file *)map->opaque;
826 fput(map_file->file);
827 kfree(map_file);
828 vhost_vdpa_general_unmap(v, map, asid);
829 vhost_iotlb_map_free(iotlb, map);
830 }
831}
832
833static void vhost_vdpa_iotlb_unmap(struct vhost_vdpa *v,
834 struct vhost_iotlb *iotlb, u64 start,
835 u64 last, u32 asid)
836{
837 struct vdpa_device *vdpa = v->vdpa;
838
839 if (vdpa->use_va)
840 return vhost_vdpa_va_unmap(v, iotlb, start, last, asid);
841
842 return vhost_vdpa_pa_unmap(v, iotlb, start, last, asid);
843}
844
845static int perm_to_iommu_flags(u32 perm)
846{
847 int flags = 0;
848
849 switch (perm) {
850 case VHOST_ACCESS_WO:
851 flags |= IOMMU_WRITE;
852 break;
853 case VHOST_ACCESS_RO:
854 flags |= IOMMU_READ;
855 break;
856 case VHOST_ACCESS_RW:
857 flags |= (IOMMU_WRITE | IOMMU_READ);
858 break;
859 default:
860 WARN(1, "invalidate vhost IOTLB permission\n");
861 break;
862 }
863
864 return flags | IOMMU_CACHE;
865}
866
867static int vhost_vdpa_map(struct vhost_vdpa *v, struct vhost_iotlb *iotlb,
868 u64 iova, u64 size, u64 pa, u32 perm, void *opaque)
869{
870 struct vhost_dev *dev = &v->vdev;
871 struct vdpa_device *vdpa = v->vdpa;
872 const struct vdpa_config_ops *ops = vdpa->config;
873 u32 asid = iotlb_to_asid(iotlb);
874 int r = 0;
875
876 r = vhost_iotlb_add_range_ctx(iotlb, iova, iova + size - 1,
877 pa, perm, opaque);
878 if (r)
879 return r;
880
881 if (ops->dma_map) {
882 r = ops->dma_map(vdpa, asid, iova, size, pa, perm, opaque);
883 } else if (ops->set_map) {
884 if (!v->in_batch)
885 r = ops->set_map(vdpa, asid, iotlb);
886 } else {
887 r = iommu_map(v->domain, iova, pa, size,
888 perm_to_iommu_flags(perm), GFP_KERNEL);
889 }
890 if (r) {
891 vhost_iotlb_del_range(iotlb, iova, iova + size - 1);
892 return r;
893 }
894
895 if (!vdpa->use_va)
896 atomic64_add(PFN_DOWN(size), &dev->mm->pinned_vm);
897
898 return 0;
899}
900
901static void vhost_vdpa_unmap(struct vhost_vdpa *v,
902 struct vhost_iotlb *iotlb,
903 u64 iova, u64 size)
904{
905 struct vdpa_device *vdpa = v->vdpa;
906 const struct vdpa_config_ops *ops = vdpa->config;
907 u32 asid = iotlb_to_asid(iotlb);
908
909 vhost_vdpa_iotlb_unmap(v, iotlb, iova, iova + size - 1, asid);
910
911 if (ops->set_map) {
912 if (!v->in_batch)
913 ops->set_map(vdpa, asid, iotlb);
914 }
915
916}
917
918static int vhost_vdpa_va_map(struct vhost_vdpa *v,
919 struct vhost_iotlb *iotlb,
920 u64 iova, u64 size, u64 uaddr, u32 perm)
921{
922 struct vhost_dev *dev = &v->vdev;
923 u64 offset, map_size, map_iova = iova;
924 struct vdpa_map_file *map_file;
925 struct vm_area_struct *vma;
926 int ret = 0;
927
928 mmap_read_lock(dev->mm);
929
930 while (size) {
931 vma = find_vma(dev->mm, uaddr);
932 if (!vma) {
933 ret = -EINVAL;
934 break;
935 }
936 map_size = min(size, vma->vm_end - uaddr);
937 if (!(vma->vm_file && (vma->vm_flags & VM_SHARED) &&
938 !(vma->vm_flags & (VM_IO | VM_PFNMAP))))
939 goto next;
940
941 map_file = kzalloc(sizeof(*map_file), GFP_KERNEL);
942 if (!map_file) {
943 ret = -ENOMEM;
944 break;
945 }
946 offset = (vma->vm_pgoff << PAGE_SHIFT) + uaddr - vma->vm_start;
947 map_file->offset = offset;
948 map_file->file = get_file(vma->vm_file);
949 ret = vhost_vdpa_map(v, iotlb, map_iova, map_size, uaddr,
950 perm, map_file);
951 if (ret) {
952 fput(map_file->file);
953 kfree(map_file);
954 break;
955 }
956next:
957 size -= map_size;
958 uaddr += map_size;
959 map_iova += map_size;
960 }
961 if (ret)
962 vhost_vdpa_unmap(v, iotlb, iova, map_iova - iova);
963
964 mmap_read_unlock(dev->mm);
965
966 return ret;
967}
968
969static int vhost_vdpa_pa_map(struct vhost_vdpa *v,
970 struct vhost_iotlb *iotlb,
971 u64 iova, u64 size, u64 uaddr, u32 perm)
972{
973 struct vhost_dev *dev = &v->vdev;
974 struct page **page_list;
975 unsigned long list_size = PAGE_SIZE / sizeof(struct page *);
976 unsigned int gup_flags = FOLL_LONGTERM;
977 unsigned long npages, cur_base, map_pfn, last_pfn = 0;
978 unsigned long lock_limit, sz2pin, nchunks, i;
979 u64 start = iova;
980 long pinned;
981 int ret = 0;
982
983 /* Limit the use of memory for bookkeeping */
984 page_list = (struct page **) __get_free_page(GFP_KERNEL);
985 if (!page_list)
986 return -ENOMEM;
987
988 if (perm & VHOST_ACCESS_WO)
989 gup_flags |= FOLL_WRITE;
990
991 npages = PFN_UP(size + (iova & ~PAGE_MASK));
992 if (!npages) {
993 ret = -EINVAL;
994 goto free;
995 }
996
997 mmap_read_lock(dev->mm);
998
999 lock_limit = PFN_DOWN(rlimit(RLIMIT_MEMLOCK));
1000 if (npages + atomic64_read(&dev->mm->pinned_vm) > lock_limit) {
1001 ret = -ENOMEM;
1002 goto unlock;
1003 }
1004
1005 cur_base = uaddr & PAGE_MASK;
1006 iova &= PAGE_MASK;
1007 nchunks = 0;
1008
1009 while (npages) {
1010 sz2pin = min_t(unsigned long, npages, list_size);
1011 pinned = pin_user_pages(cur_base, sz2pin,
1012 gup_flags, page_list, NULL);
1013 if (sz2pin != pinned) {
1014 if (pinned < 0) {
1015 ret = pinned;
1016 } else {
1017 unpin_user_pages(page_list, pinned);
1018 ret = -ENOMEM;
1019 }
1020 goto out;
1021 }
1022 nchunks++;
1023
1024 if (!last_pfn)
1025 map_pfn = page_to_pfn(page_list[0]);
1026
1027 for (i = 0; i < pinned; i++) {
1028 unsigned long this_pfn = page_to_pfn(page_list[i]);
1029 u64 csize;
1030
1031 if (last_pfn && (this_pfn != last_pfn + 1)) {
1032 /* Pin a contiguous chunk of memory */
1033 csize = PFN_PHYS(last_pfn - map_pfn + 1);
1034 ret = vhost_vdpa_map(v, iotlb, iova, csize,
1035 PFN_PHYS(map_pfn),
1036 perm, NULL);
1037 if (ret) {
1038 /*
1039 * Unpin the pages that are left unmapped
1040 * from this point on in the current
1041 * page_list. The remaining outstanding
1042 * ones which may stride across several
1043 * chunks will be covered in the common
1044 * error path subsequently.
1045 */
1046 unpin_user_pages(&page_list[i],
1047 pinned - i);
1048 goto out;
1049 }
1050
1051 map_pfn = this_pfn;
1052 iova += csize;
1053 nchunks = 0;
1054 }
1055
1056 last_pfn = this_pfn;
1057 }
1058
1059 cur_base += PFN_PHYS(pinned);
1060 npages -= pinned;
1061 }
1062
1063 /* Pin the rest chunk */
1064 ret = vhost_vdpa_map(v, iotlb, iova, PFN_PHYS(last_pfn - map_pfn + 1),
1065 PFN_PHYS(map_pfn), perm, NULL);
1066out:
1067 if (ret) {
1068 if (nchunks) {
1069 unsigned long pfn;
1070
1071 /*
1072 * Unpin the outstanding pages which are yet to be
1073 * mapped but haven't due to vdpa_map() or
1074 * pin_user_pages() failure.
1075 *
1076 * Mapped pages are accounted in vdpa_map(), hence
1077 * the corresponding unpinning will be handled by
1078 * vdpa_unmap().
1079 */
1080 WARN_ON(!last_pfn);
1081 for (pfn = map_pfn; pfn <= last_pfn; pfn++)
1082 unpin_user_page(pfn_to_page(pfn));
1083 }
1084 vhost_vdpa_unmap(v, iotlb, start, size);
1085 }
1086unlock:
1087 mmap_read_unlock(dev->mm);
1088free:
1089 free_page((unsigned long)page_list);
1090 return ret;
1091
1092}
1093
1094static int vhost_vdpa_process_iotlb_update(struct vhost_vdpa *v,
1095 struct vhost_iotlb *iotlb,
1096 struct vhost_iotlb_msg *msg)
1097{
1098 struct vdpa_device *vdpa = v->vdpa;
1099
1100 if (msg->iova < v->range.first || !msg->size ||
1101 msg->iova > U64_MAX - msg->size + 1 ||
1102 msg->iova + msg->size - 1 > v->range.last)
1103 return -EINVAL;
1104
1105 if (vhost_iotlb_itree_first(iotlb, msg->iova,
1106 msg->iova + msg->size - 1))
1107 return -EEXIST;
1108
1109 if (vdpa->use_va)
1110 return vhost_vdpa_va_map(v, iotlb, msg->iova, msg->size,
1111 msg->uaddr, msg->perm);
1112
1113 return vhost_vdpa_pa_map(v, iotlb, msg->iova, msg->size, msg->uaddr,
1114 msg->perm);
1115}
1116
1117static int vhost_vdpa_process_iotlb_msg(struct vhost_dev *dev, u32 asid,
1118 struct vhost_iotlb_msg *msg)
1119{
1120 struct vhost_vdpa *v = container_of(dev, struct vhost_vdpa, vdev);
1121 struct vdpa_device *vdpa = v->vdpa;
1122 const struct vdpa_config_ops *ops = vdpa->config;
1123 struct vhost_iotlb *iotlb = NULL;
1124 struct vhost_vdpa_as *as = NULL;
1125 int r = 0;
1126
1127 mutex_lock(&dev->mutex);
1128
1129 r = vhost_dev_check_owner(dev);
1130 if (r)
1131 goto unlock;
1132
1133 if (msg->type == VHOST_IOTLB_UPDATE ||
1134 msg->type == VHOST_IOTLB_BATCH_BEGIN) {
1135 as = vhost_vdpa_find_alloc_as(v, asid);
1136 if (!as) {
1137 dev_err(&v->dev, "can't find and alloc asid %d\n",
1138 asid);
1139 r = -EINVAL;
1140 goto unlock;
1141 }
1142 iotlb = &as->iotlb;
1143 } else
1144 iotlb = asid_to_iotlb(v, asid);
1145
1146 if ((v->in_batch && v->batch_asid != asid) || !iotlb) {
1147 if (v->in_batch && v->batch_asid != asid) {
1148 dev_info(&v->dev, "batch id %d asid %d\n",
1149 v->batch_asid, asid);
1150 }
1151 if (!iotlb)
1152 dev_err(&v->dev, "no iotlb for asid %d\n", asid);
1153 r = -EINVAL;
1154 goto unlock;
1155 }
1156
1157 switch (msg->type) {
1158 case VHOST_IOTLB_UPDATE:
1159 r = vhost_vdpa_process_iotlb_update(v, iotlb, msg);
1160 break;
1161 case VHOST_IOTLB_INVALIDATE:
1162 vhost_vdpa_unmap(v, iotlb, msg->iova, msg->size);
1163 break;
1164 case VHOST_IOTLB_BATCH_BEGIN:
1165 v->batch_asid = asid;
1166 v->in_batch = true;
1167 break;
1168 case VHOST_IOTLB_BATCH_END:
1169 if (v->in_batch && ops->set_map)
1170 ops->set_map(vdpa, asid, iotlb);
1171 v->in_batch = false;
1172 break;
1173 default:
1174 r = -EINVAL;
1175 break;
1176 }
1177unlock:
1178 mutex_unlock(&dev->mutex);
1179
1180 return r;
1181}
1182
1183static ssize_t vhost_vdpa_chr_write_iter(struct kiocb *iocb,
1184 struct iov_iter *from)
1185{
1186 struct file *file = iocb->ki_filp;
1187 struct vhost_vdpa *v = file->private_data;
1188 struct vhost_dev *dev = &v->vdev;
1189
1190 return vhost_chr_write_iter(dev, from);
1191}
1192
1193static int vhost_vdpa_alloc_domain(struct vhost_vdpa *v)
1194{
1195 struct vdpa_device *vdpa = v->vdpa;
1196 const struct vdpa_config_ops *ops = vdpa->config;
1197 struct device *dma_dev = vdpa_get_dma_dev(vdpa);
1198 const struct bus_type *bus;
1199 int ret;
1200
1201 /* Device want to do DMA by itself */
1202 if (ops->set_map || ops->dma_map)
1203 return 0;
1204
1205 bus = dma_dev->bus;
1206 if (!bus)
1207 return -EFAULT;
1208
1209 if (!device_iommu_capable(dma_dev, IOMMU_CAP_CACHE_COHERENCY)) {
1210 dev_warn_once(&v->dev,
1211 "Failed to allocate domain, device is not IOMMU cache coherent capable\n");
1212 return -ENOTSUPP;
1213 }
1214
1215 v->domain = iommu_domain_alloc(bus);
1216 if (!v->domain)
1217 return -EIO;
1218
1219 ret = iommu_attach_device(v->domain, dma_dev);
1220 if (ret)
1221 goto err_attach;
1222
1223 return 0;
1224
1225err_attach:
1226 iommu_domain_free(v->domain);
1227 v->domain = NULL;
1228 return ret;
1229}
1230
1231static void vhost_vdpa_free_domain(struct vhost_vdpa *v)
1232{
1233 struct vdpa_device *vdpa = v->vdpa;
1234 struct device *dma_dev = vdpa_get_dma_dev(vdpa);
1235
1236 if (v->domain) {
1237 iommu_detach_device(v->domain, dma_dev);
1238 iommu_domain_free(v->domain);
1239 }
1240
1241 v->domain = NULL;
1242}
1243
1244static void vhost_vdpa_set_iova_range(struct vhost_vdpa *v)
1245{
1246 struct vdpa_iova_range *range = &v->range;
1247 struct vdpa_device *vdpa = v->vdpa;
1248 const struct vdpa_config_ops *ops = vdpa->config;
1249
1250 if (ops->get_iova_range) {
1251 *range = ops->get_iova_range(vdpa);
1252 } else if (v->domain && v->domain->geometry.force_aperture) {
1253 range->first = v->domain->geometry.aperture_start;
1254 range->last = v->domain->geometry.aperture_end;
1255 } else {
1256 range->first = 0;
1257 range->last = ULLONG_MAX;
1258 }
1259}
1260
1261static void vhost_vdpa_cleanup(struct vhost_vdpa *v)
1262{
1263 struct vhost_vdpa_as *as;
1264 u32 asid;
1265
1266 for (asid = 0; asid < v->vdpa->nas; asid++) {
1267 as = asid_to_as(v, asid);
1268 if (as)
1269 vhost_vdpa_remove_as(v, asid);
1270 }
1271
1272 vhost_vdpa_free_domain(v);
1273 vhost_dev_cleanup(&v->vdev);
1274 kfree(v->vdev.vqs);
1275}
1276
1277static int vhost_vdpa_open(struct inode *inode, struct file *filep)
1278{
1279 struct vhost_vdpa *v;
1280 struct vhost_dev *dev;
1281 struct vhost_virtqueue **vqs;
1282 int r, opened;
1283 u32 i, nvqs;
1284
1285 v = container_of(inode->i_cdev, struct vhost_vdpa, cdev);
1286
1287 opened = atomic_cmpxchg(&v->opened, 0, 1);
1288 if (opened)
1289 return -EBUSY;
1290
1291 nvqs = v->nvqs;
1292 r = vhost_vdpa_reset(v);
1293 if (r)
1294 goto err;
1295
1296 vqs = kmalloc_array(nvqs, sizeof(*vqs), GFP_KERNEL);
1297 if (!vqs) {
1298 r = -ENOMEM;
1299 goto err;
1300 }
1301
1302 dev = &v->vdev;
1303 for (i = 0; i < nvqs; i++) {
1304 vqs[i] = &v->vqs[i];
1305 vqs[i]->handle_kick = handle_vq_kick;
1306 }
1307 vhost_dev_init(dev, vqs, nvqs, 0, 0, 0, false,
1308 vhost_vdpa_process_iotlb_msg);
1309
1310 r = vhost_vdpa_alloc_domain(v);
1311 if (r)
1312 goto err_alloc_domain;
1313
1314 vhost_vdpa_set_iova_range(v);
1315
1316 filep->private_data = v;
1317
1318 return 0;
1319
1320err_alloc_domain:
1321 vhost_vdpa_cleanup(v);
1322err:
1323 atomic_dec(&v->opened);
1324 return r;
1325}
1326
1327static void vhost_vdpa_clean_irq(struct vhost_vdpa *v)
1328{
1329 u32 i;
1330
1331 for (i = 0; i < v->nvqs; i++)
1332 vhost_vdpa_unsetup_vq_irq(v, i);
1333}
1334
1335static int vhost_vdpa_release(struct inode *inode, struct file *filep)
1336{
1337 struct vhost_vdpa *v = filep->private_data;
1338 struct vhost_dev *d = &v->vdev;
1339
1340 mutex_lock(&d->mutex);
1341 filep->private_data = NULL;
1342 vhost_vdpa_clean_irq(v);
1343 vhost_vdpa_reset(v);
1344 vhost_dev_stop(&v->vdev);
1345 vhost_vdpa_unbind_mm(v);
1346 vhost_vdpa_config_put(v);
1347 vhost_vdpa_cleanup(v);
1348 mutex_unlock(&d->mutex);
1349
1350 atomic_dec(&v->opened);
1351 complete(&v->completion);
1352
1353 return 0;
1354}
1355
1356#ifdef CONFIG_MMU
1357static vm_fault_t vhost_vdpa_fault(struct vm_fault *vmf)
1358{
1359 struct vhost_vdpa *v = vmf->vma->vm_file->private_data;
1360 struct vdpa_device *vdpa = v->vdpa;
1361 const struct vdpa_config_ops *ops = vdpa->config;
1362 struct vdpa_notification_area notify;
1363 struct vm_area_struct *vma = vmf->vma;
1364 u16 index = vma->vm_pgoff;
1365
1366 notify = ops->get_vq_notification(vdpa, index);
1367
1368 vma->vm_page_prot = pgprot_noncached(vma->vm_page_prot);
1369 if (remap_pfn_range(vma, vmf->address & PAGE_MASK,
1370 PFN_DOWN(notify.addr), PAGE_SIZE,
1371 vma->vm_page_prot))
1372 return VM_FAULT_SIGBUS;
1373
1374 return VM_FAULT_NOPAGE;
1375}
1376
1377static const struct vm_operations_struct vhost_vdpa_vm_ops = {
1378 .fault = vhost_vdpa_fault,
1379};
1380
1381static int vhost_vdpa_mmap(struct file *file, struct vm_area_struct *vma)
1382{
1383 struct vhost_vdpa *v = vma->vm_file->private_data;
1384 struct vdpa_device *vdpa = v->vdpa;
1385 const struct vdpa_config_ops *ops = vdpa->config;
1386 struct vdpa_notification_area notify;
1387 unsigned long index = vma->vm_pgoff;
1388
1389 if (vma->vm_end - vma->vm_start != PAGE_SIZE)
1390 return -EINVAL;
1391 if ((vma->vm_flags & VM_SHARED) == 0)
1392 return -EINVAL;
1393 if (vma->vm_flags & VM_READ)
1394 return -EINVAL;
1395 if (index > 65535)
1396 return -EINVAL;
1397 if (!ops->get_vq_notification)
1398 return -ENOTSUPP;
1399
1400 /* To be safe and easily modelled by userspace, We only
1401 * support the doorbell which sits on the page boundary and
1402 * does not share the page with other registers.
1403 */
1404 notify = ops->get_vq_notification(vdpa, index);
1405 if (notify.addr & (PAGE_SIZE - 1))
1406 return -EINVAL;
1407 if (vma->vm_end - vma->vm_start != notify.size)
1408 return -ENOTSUPP;
1409
1410 vm_flags_set(vma, VM_IO | VM_PFNMAP | VM_DONTEXPAND | VM_DONTDUMP);
1411 vma->vm_ops = &vhost_vdpa_vm_ops;
1412 return 0;
1413}
1414#endif /* CONFIG_MMU */
1415
1416static const struct file_operations vhost_vdpa_fops = {
1417 .owner = THIS_MODULE,
1418 .open = vhost_vdpa_open,
1419 .release = vhost_vdpa_release,
1420 .write_iter = vhost_vdpa_chr_write_iter,
1421 .unlocked_ioctl = vhost_vdpa_unlocked_ioctl,
1422#ifdef CONFIG_MMU
1423 .mmap = vhost_vdpa_mmap,
1424#endif /* CONFIG_MMU */
1425 .compat_ioctl = compat_ptr_ioctl,
1426};
1427
1428static void vhost_vdpa_release_dev(struct device *device)
1429{
1430 struct vhost_vdpa *v =
1431 container_of(device, struct vhost_vdpa, dev);
1432
1433 ida_simple_remove(&vhost_vdpa_ida, v->minor);
1434 kfree(v->vqs);
1435 kfree(v);
1436}
1437
1438static int vhost_vdpa_probe(struct vdpa_device *vdpa)
1439{
1440 const struct vdpa_config_ops *ops = vdpa->config;
1441 struct vhost_vdpa *v;
1442 int minor;
1443 int i, r;
1444
1445 /* We can't support platform IOMMU device with more than 1
1446 * group or as
1447 */
1448 if (!ops->set_map && !ops->dma_map &&
1449 (vdpa->ngroups > 1 || vdpa->nas > 1))
1450 return -EOPNOTSUPP;
1451
1452 v = kzalloc(sizeof(*v), GFP_KERNEL | __GFP_RETRY_MAYFAIL);
1453 if (!v)
1454 return -ENOMEM;
1455
1456 minor = ida_simple_get(&vhost_vdpa_ida, 0,
1457 VHOST_VDPA_DEV_MAX, GFP_KERNEL);
1458 if (minor < 0) {
1459 kfree(v);
1460 return minor;
1461 }
1462
1463 atomic_set(&v->opened, 0);
1464 v->minor = minor;
1465 v->vdpa = vdpa;
1466 v->nvqs = vdpa->nvqs;
1467 v->virtio_id = ops->get_device_id(vdpa);
1468
1469 device_initialize(&v->dev);
1470 v->dev.release = vhost_vdpa_release_dev;
1471 v->dev.parent = &vdpa->dev;
1472 v->dev.devt = MKDEV(MAJOR(vhost_vdpa_major), minor);
1473 v->vqs = kmalloc_array(v->nvqs, sizeof(struct vhost_virtqueue),
1474 GFP_KERNEL);
1475 if (!v->vqs) {
1476 r = -ENOMEM;
1477 goto err;
1478 }
1479
1480 r = dev_set_name(&v->dev, "vhost-vdpa-%u", minor);
1481 if (r)
1482 goto err;
1483
1484 cdev_init(&v->cdev, &vhost_vdpa_fops);
1485 v->cdev.owner = THIS_MODULE;
1486
1487 r = cdev_device_add(&v->cdev, &v->dev);
1488 if (r)
1489 goto err;
1490
1491 init_completion(&v->completion);
1492 vdpa_set_drvdata(vdpa, v);
1493
1494 for (i = 0; i < VHOST_VDPA_IOTLB_BUCKETS; i++)
1495 INIT_HLIST_HEAD(&v->as[i]);
1496
1497 return 0;
1498
1499err:
1500 put_device(&v->dev);
1501 ida_simple_remove(&vhost_vdpa_ida, v->minor);
1502 return r;
1503}
1504
1505static void vhost_vdpa_remove(struct vdpa_device *vdpa)
1506{
1507 struct vhost_vdpa *v = vdpa_get_drvdata(vdpa);
1508 int opened;
1509
1510 cdev_device_del(&v->cdev, &v->dev);
1511
1512 do {
1513 opened = atomic_cmpxchg(&v->opened, 0, 1);
1514 if (!opened)
1515 break;
1516 wait_for_completion(&v->completion);
1517 } while (1);
1518
1519 put_device(&v->dev);
1520}
1521
1522static struct vdpa_driver vhost_vdpa_driver = {
1523 .driver = {
1524 .name = "vhost_vdpa",
1525 },
1526 .probe = vhost_vdpa_probe,
1527 .remove = vhost_vdpa_remove,
1528};
1529
1530static int __init vhost_vdpa_init(void)
1531{
1532 int r;
1533
1534 r = alloc_chrdev_region(&vhost_vdpa_major, 0, VHOST_VDPA_DEV_MAX,
1535 "vhost-vdpa");
1536 if (r)
1537 goto err_alloc_chrdev;
1538
1539 r = vdpa_register_driver(&vhost_vdpa_driver);
1540 if (r)
1541 goto err_vdpa_register_driver;
1542
1543 return 0;
1544
1545err_vdpa_register_driver:
1546 unregister_chrdev_region(vhost_vdpa_major, VHOST_VDPA_DEV_MAX);
1547err_alloc_chrdev:
1548 return r;
1549}
1550module_init(vhost_vdpa_init);
1551
1552static void __exit vhost_vdpa_exit(void)
1553{
1554 vdpa_unregister_driver(&vhost_vdpa_driver);
1555 unregister_chrdev_region(vhost_vdpa_major, VHOST_VDPA_DEV_MAX);
1556}
1557module_exit(vhost_vdpa_exit);
1558
1559MODULE_VERSION("0.0.1");
1560MODULE_LICENSE("GPL v2");
1561MODULE_AUTHOR("Intel Corporation");
1562MODULE_DESCRIPTION("vDPA-based vhost backend for virtio");