Loading...
1/* Copyright (C) 2009 Red Hat, Inc.
2 * Copyright (C) 2006 Rusty Russell IBM Corporation
3 *
4 * Author: Michael S. Tsirkin <mst@redhat.com>
5 *
6 * Inspiration, some code, and most witty comments come from
7 * Documentation/virtual/lguest/lguest.c, by Rusty Russell
8 *
9 * This work is licensed under the terms of the GNU GPL, version 2.
10 *
11 * Generic code for virtio server in host kernel.
12 */
13
14#include <linux/eventfd.h>
15#include <linux/vhost.h>
16#include <linux/virtio_net.h>
17#include <linux/mm.h>
18#include <linux/mmu_context.h>
19#include <linux/miscdevice.h>
20#include <linux/mutex.h>
21#include <linux/rcupdate.h>
22#include <linux/poll.h>
23#include <linux/file.h>
24#include <linux/highmem.h>
25#include <linux/slab.h>
26#include <linux/kthread.h>
27#include <linux/cgroup.h>
28
29#include <linux/net.h>
30#include <linux/if_packet.h>
31#include <linux/if_arp.h>
32
33#include "vhost.h"
34
35enum {
36 VHOST_MEMORY_MAX_NREGIONS = 64,
37 VHOST_MEMORY_F_LOG = 0x1,
38};
39
40static unsigned vhost_zcopy_mask __read_mostly;
41
42#define vhost_used_event(vq) ((u16 __user *)&vq->avail->ring[vq->num])
43#define vhost_avail_event(vq) ((u16 __user *)&vq->used->ring[vq->num])
44
45static void vhost_poll_func(struct file *file, wait_queue_head_t *wqh,
46 poll_table *pt)
47{
48 struct vhost_poll *poll;
49
50 poll = container_of(pt, struct vhost_poll, table);
51 poll->wqh = wqh;
52 add_wait_queue(wqh, &poll->wait);
53}
54
55static int vhost_poll_wakeup(wait_queue_t *wait, unsigned mode, int sync,
56 void *key)
57{
58 struct vhost_poll *poll = container_of(wait, struct vhost_poll, wait);
59
60 if (!((unsigned long)key & poll->mask))
61 return 0;
62
63 vhost_poll_queue(poll);
64 return 0;
65}
66
67static void vhost_work_init(struct vhost_work *work, vhost_work_fn_t fn)
68{
69 INIT_LIST_HEAD(&work->node);
70 work->fn = fn;
71 init_waitqueue_head(&work->done);
72 work->flushing = 0;
73 work->queue_seq = work->done_seq = 0;
74}
75
76/* Init poll structure */
77void vhost_poll_init(struct vhost_poll *poll, vhost_work_fn_t fn,
78 unsigned long mask, struct vhost_dev *dev)
79{
80 init_waitqueue_func_entry(&poll->wait, vhost_poll_wakeup);
81 init_poll_funcptr(&poll->table, vhost_poll_func);
82 poll->mask = mask;
83 poll->dev = dev;
84
85 vhost_work_init(&poll->work, fn);
86}
87
88/* Start polling a file. We add ourselves to file's wait queue. The caller must
89 * keep a reference to a file until after vhost_poll_stop is called. */
90void vhost_poll_start(struct vhost_poll *poll, struct file *file)
91{
92 unsigned long mask;
93
94 mask = file->f_op->poll(file, &poll->table);
95 if (mask)
96 vhost_poll_wakeup(&poll->wait, 0, 0, (void *)mask);
97}
98
99/* Stop polling a file. After this function returns, it becomes safe to drop the
100 * file reference. You must also flush afterwards. */
101void vhost_poll_stop(struct vhost_poll *poll)
102{
103 remove_wait_queue(poll->wqh, &poll->wait);
104}
105
106static bool vhost_work_seq_done(struct vhost_dev *dev, struct vhost_work *work,
107 unsigned seq)
108{
109 int left;
110
111 spin_lock_irq(&dev->work_lock);
112 left = seq - work->done_seq;
113 spin_unlock_irq(&dev->work_lock);
114 return left <= 0;
115}
116
117static void vhost_work_flush(struct vhost_dev *dev, struct vhost_work *work)
118{
119 unsigned seq;
120 int flushing;
121
122 spin_lock_irq(&dev->work_lock);
123 seq = work->queue_seq;
124 work->flushing++;
125 spin_unlock_irq(&dev->work_lock);
126 wait_event(work->done, vhost_work_seq_done(dev, work, seq));
127 spin_lock_irq(&dev->work_lock);
128 flushing = --work->flushing;
129 spin_unlock_irq(&dev->work_lock);
130 BUG_ON(flushing < 0);
131}
132
133/* Flush any work that has been scheduled. When calling this, don't hold any
134 * locks that are also used by the callback. */
135void vhost_poll_flush(struct vhost_poll *poll)
136{
137 vhost_work_flush(poll->dev, &poll->work);
138}
139
140static inline void vhost_work_queue(struct vhost_dev *dev,
141 struct vhost_work *work)
142{
143 unsigned long flags;
144
145 spin_lock_irqsave(&dev->work_lock, flags);
146 if (list_empty(&work->node)) {
147 list_add_tail(&work->node, &dev->work_list);
148 work->queue_seq++;
149 wake_up_process(dev->worker);
150 }
151 spin_unlock_irqrestore(&dev->work_lock, flags);
152}
153
154void vhost_poll_queue(struct vhost_poll *poll)
155{
156 vhost_work_queue(poll->dev, &poll->work);
157}
158
159static void vhost_vq_reset(struct vhost_dev *dev,
160 struct vhost_virtqueue *vq)
161{
162 vq->num = 1;
163 vq->desc = NULL;
164 vq->avail = NULL;
165 vq->used = NULL;
166 vq->last_avail_idx = 0;
167 vq->avail_idx = 0;
168 vq->last_used_idx = 0;
169 vq->signalled_used = 0;
170 vq->signalled_used_valid = false;
171 vq->used_flags = 0;
172 vq->log_used = false;
173 vq->log_addr = -1ull;
174 vq->vhost_hlen = 0;
175 vq->sock_hlen = 0;
176 vq->private_data = NULL;
177 vq->log_base = NULL;
178 vq->error_ctx = NULL;
179 vq->error = NULL;
180 vq->kick = NULL;
181 vq->call_ctx = NULL;
182 vq->call = NULL;
183 vq->log_ctx = NULL;
184 vq->upend_idx = 0;
185 vq->done_idx = 0;
186 vq->ubufs = NULL;
187}
188
189static int vhost_worker(void *data)
190{
191 struct vhost_dev *dev = data;
192 struct vhost_work *work = NULL;
193 unsigned uninitialized_var(seq);
194
195 use_mm(dev->mm);
196
197 for (;;) {
198 /* mb paired w/ kthread_stop */
199 set_current_state(TASK_INTERRUPTIBLE);
200
201 spin_lock_irq(&dev->work_lock);
202 if (work) {
203 work->done_seq = seq;
204 if (work->flushing)
205 wake_up_all(&work->done);
206 }
207
208 if (kthread_should_stop()) {
209 spin_unlock_irq(&dev->work_lock);
210 __set_current_state(TASK_RUNNING);
211 break;
212 }
213 if (!list_empty(&dev->work_list)) {
214 work = list_first_entry(&dev->work_list,
215 struct vhost_work, node);
216 list_del_init(&work->node);
217 seq = work->queue_seq;
218 } else
219 work = NULL;
220 spin_unlock_irq(&dev->work_lock);
221
222 if (work) {
223 __set_current_state(TASK_RUNNING);
224 work->fn(work);
225 } else
226 schedule();
227
228 }
229 unuse_mm(dev->mm);
230 return 0;
231}
232
233static void vhost_vq_free_iovecs(struct vhost_virtqueue *vq)
234{
235 kfree(vq->indirect);
236 vq->indirect = NULL;
237 kfree(vq->log);
238 vq->log = NULL;
239 kfree(vq->heads);
240 vq->heads = NULL;
241 kfree(vq->ubuf_info);
242 vq->ubuf_info = NULL;
243}
244
245void vhost_enable_zcopy(int vq)
246{
247 vhost_zcopy_mask |= 0x1 << vq;
248}
249
250/* Helper to allocate iovec buffers for all vqs. */
251static long vhost_dev_alloc_iovecs(struct vhost_dev *dev)
252{
253 int i;
254 bool zcopy;
255
256 for (i = 0; i < dev->nvqs; ++i) {
257 dev->vqs[i].indirect = kmalloc(sizeof *dev->vqs[i].indirect *
258 UIO_MAXIOV, GFP_KERNEL);
259 dev->vqs[i].log = kmalloc(sizeof *dev->vqs[i].log * UIO_MAXIOV,
260 GFP_KERNEL);
261 dev->vqs[i].heads = kmalloc(sizeof *dev->vqs[i].heads *
262 UIO_MAXIOV, GFP_KERNEL);
263 zcopy = vhost_zcopy_mask & (0x1 << i);
264 if (zcopy)
265 dev->vqs[i].ubuf_info =
266 kmalloc(sizeof *dev->vqs[i].ubuf_info *
267 UIO_MAXIOV, GFP_KERNEL);
268 if (!dev->vqs[i].indirect || !dev->vqs[i].log ||
269 !dev->vqs[i].heads ||
270 (zcopy && !dev->vqs[i].ubuf_info))
271 goto err_nomem;
272 }
273 return 0;
274
275err_nomem:
276 for (; i >= 0; --i)
277 vhost_vq_free_iovecs(&dev->vqs[i]);
278 return -ENOMEM;
279}
280
281static void vhost_dev_free_iovecs(struct vhost_dev *dev)
282{
283 int i;
284
285 for (i = 0; i < dev->nvqs; ++i)
286 vhost_vq_free_iovecs(&dev->vqs[i]);
287}
288
289long vhost_dev_init(struct vhost_dev *dev,
290 struct vhost_virtqueue *vqs, int nvqs)
291{
292 int i;
293
294 dev->vqs = vqs;
295 dev->nvqs = nvqs;
296 mutex_init(&dev->mutex);
297 dev->log_ctx = NULL;
298 dev->log_file = NULL;
299 dev->memory = NULL;
300 dev->mm = NULL;
301 spin_lock_init(&dev->work_lock);
302 INIT_LIST_HEAD(&dev->work_list);
303 dev->worker = NULL;
304
305 for (i = 0; i < dev->nvqs; ++i) {
306 dev->vqs[i].log = NULL;
307 dev->vqs[i].indirect = NULL;
308 dev->vqs[i].heads = NULL;
309 dev->vqs[i].ubuf_info = NULL;
310 dev->vqs[i].dev = dev;
311 mutex_init(&dev->vqs[i].mutex);
312 vhost_vq_reset(dev, dev->vqs + i);
313 if (dev->vqs[i].handle_kick)
314 vhost_poll_init(&dev->vqs[i].poll,
315 dev->vqs[i].handle_kick, POLLIN, dev);
316 }
317
318 return 0;
319}
320
321/* Caller should have device mutex */
322long vhost_dev_check_owner(struct vhost_dev *dev)
323{
324 /* Are you the owner? If not, I don't think you mean to do that */
325 return dev->mm == current->mm ? 0 : -EPERM;
326}
327
328struct vhost_attach_cgroups_struct {
329 struct vhost_work work;
330 struct task_struct *owner;
331 int ret;
332};
333
334static void vhost_attach_cgroups_work(struct vhost_work *work)
335{
336 struct vhost_attach_cgroups_struct *s;
337
338 s = container_of(work, struct vhost_attach_cgroups_struct, work);
339 s->ret = cgroup_attach_task_all(s->owner, current);
340}
341
342static int vhost_attach_cgroups(struct vhost_dev *dev)
343{
344 struct vhost_attach_cgroups_struct attach;
345
346 attach.owner = current;
347 vhost_work_init(&attach.work, vhost_attach_cgroups_work);
348 vhost_work_queue(dev, &attach.work);
349 vhost_work_flush(dev, &attach.work);
350 return attach.ret;
351}
352
353/* Caller should have device mutex */
354static long vhost_dev_set_owner(struct vhost_dev *dev)
355{
356 struct task_struct *worker;
357 int err;
358
359 /* Is there an owner already? */
360 if (dev->mm) {
361 err = -EBUSY;
362 goto err_mm;
363 }
364
365 /* No owner, become one */
366 dev->mm = get_task_mm(current);
367 worker = kthread_create(vhost_worker, dev, "vhost-%d", current->pid);
368 if (IS_ERR(worker)) {
369 err = PTR_ERR(worker);
370 goto err_worker;
371 }
372
373 dev->worker = worker;
374 wake_up_process(worker); /* avoid contributing to loadavg */
375
376 err = vhost_attach_cgroups(dev);
377 if (err)
378 goto err_cgroup;
379
380 err = vhost_dev_alloc_iovecs(dev);
381 if (err)
382 goto err_cgroup;
383
384 return 0;
385err_cgroup:
386 kthread_stop(worker);
387 dev->worker = NULL;
388err_worker:
389 if (dev->mm)
390 mmput(dev->mm);
391 dev->mm = NULL;
392err_mm:
393 return err;
394}
395
396/* Caller should have device mutex */
397long vhost_dev_reset_owner(struct vhost_dev *dev)
398{
399 struct vhost_memory *memory;
400
401 /* Restore memory to default empty mapping. */
402 memory = kmalloc(offsetof(struct vhost_memory, regions), GFP_KERNEL);
403 if (!memory)
404 return -ENOMEM;
405
406 vhost_dev_cleanup(dev);
407
408 memory->nregions = 0;
409 RCU_INIT_POINTER(dev->memory, memory);
410 return 0;
411}
412
413/* In case of DMA done not in order in lower device driver for some reason.
414 * upend_idx is used to track end of used idx, done_idx is used to track head
415 * of used idx. Once lower device DMA done contiguously, we will signal KVM
416 * guest used idx.
417 */
418int vhost_zerocopy_signal_used(struct vhost_virtqueue *vq)
419{
420 int i;
421 int j = 0;
422
423 for (i = vq->done_idx; i != vq->upend_idx; i = (i + 1) % UIO_MAXIOV) {
424 if ((vq->heads[i].len == VHOST_DMA_DONE_LEN)) {
425 vq->heads[i].len = VHOST_DMA_CLEAR_LEN;
426 vhost_add_used_and_signal(vq->dev, vq,
427 vq->heads[i].id, 0);
428 ++j;
429 } else
430 break;
431 }
432 if (j)
433 vq->done_idx = i;
434 return j;
435}
436
437/* Caller should have device mutex */
438void vhost_dev_cleanup(struct vhost_dev *dev)
439{
440 int i;
441
442 for (i = 0; i < dev->nvqs; ++i) {
443 if (dev->vqs[i].kick && dev->vqs[i].handle_kick) {
444 vhost_poll_stop(&dev->vqs[i].poll);
445 vhost_poll_flush(&dev->vqs[i].poll);
446 }
447 /* Wait for all lower device DMAs done. */
448 if (dev->vqs[i].ubufs)
449 vhost_ubuf_put_and_wait(dev->vqs[i].ubufs);
450
451 /* Signal guest as appropriate. */
452 vhost_zerocopy_signal_used(&dev->vqs[i]);
453
454 if (dev->vqs[i].error_ctx)
455 eventfd_ctx_put(dev->vqs[i].error_ctx);
456 if (dev->vqs[i].error)
457 fput(dev->vqs[i].error);
458 if (dev->vqs[i].kick)
459 fput(dev->vqs[i].kick);
460 if (dev->vqs[i].call_ctx)
461 eventfd_ctx_put(dev->vqs[i].call_ctx);
462 if (dev->vqs[i].call)
463 fput(dev->vqs[i].call);
464 vhost_vq_reset(dev, dev->vqs + i);
465 }
466 vhost_dev_free_iovecs(dev);
467 if (dev->log_ctx)
468 eventfd_ctx_put(dev->log_ctx);
469 dev->log_ctx = NULL;
470 if (dev->log_file)
471 fput(dev->log_file);
472 dev->log_file = NULL;
473 /* No one will access memory at this point */
474 kfree(rcu_dereference_protected(dev->memory,
475 lockdep_is_held(&dev->mutex)));
476 RCU_INIT_POINTER(dev->memory, NULL);
477 WARN_ON(!list_empty(&dev->work_list));
478 if (dev->worker) {
479 kthread_stop(dev->worker);
480 dev->worker = NULL;
481 }
482 if (dev->mm)
483 mmput(dev->mm);
484 dev->mm = NULL;
485}
486
487static int log_access_ok(void __user *log_base, u64 addr, unsigned long sz)
488{
489 u64 a = addr / VHOST_PAGE_SIZE / 8;
490
491 /* Make sure 64 bit math will not overflow. */
492 if (a > ULONG_MAX - (unsigned long)log_base ||
493 a + (unsigned long)log_base > ULONG_MAX)
494 return 0;
495
496 return access_ok(VERIFY_WRITE, log_base + a,
497 (sz + VHOST_PAGE_SIZE * 8 - 1) / VHOST_PAGE_SIZE / 8);
498}
499
500/* Caller should have vq mutex and device mutex. */
501static int vq_memory_access_ok(void __user *log_base, struct vhost_memory *mem,
502 int log_all)
503{
504 int i;
505
506 if (!mem)
507 return 0;
508
509 for (i = 0; i < mem->nregions; ++i) {
510 struct vhost_memory_region *m = mem->regions + i;
511 unsigned long a = m->userspace_addr;
512 if (m->memory_size > ULONG_MAX)
513 return 0;
514 else if (!access_ok(VERIFY_WRITE, (void __user *)a,
515 m->memory_size))
516 return 0;
517 else if (log_all && !log_access_ok(log_base,
518 m->guest_phys_addr,
519 m->memory_size))
520 return 0;
521 }
522 return 1;
523}
524
525/* Can we switch to this memory table? */
526/* Caller should have device mutex but not vq mutex */
527static int memory_access_ok(struct vhost_dev *d, struct vhost_memory *mem,
528 int log_all)
529{
530 int i;
531
532 for (i = 0; i < d->nvqs; ++i) {
533 int ok;
534 mutex_lock(&d->vqs[i].mutex);
535 /* If ring is inactive, will check when it's enabled. */
536 if (d->vqs[i].private_data)
537 ok = vq_memory_access_ok(d->vqs[i].log_base, mem,
538 log_all);
539 else
540 ok = 1;
541 mutex_unlock(&d->vqs[i].mutex);
542 if (!ok)
543 return 0;
544 }
545 return 1;
546}
547
548static int vq_access_ok(struct vhost_dev *d, unsigned int num,
549 struct vring_desc __user *desc,
550 struct vring_avail __user *avail,
551 struct vring_used __user *used)
552{
553 size_t s = vhost_has_feature(d, VIRTIO_RING_F_EVENT_IDX) ? 2 : 0;
554 return access_ok(VERIFY_READ, desc, num * sizeof *desc) &&
555 access_ok(VERIFY_READ, avail,
556 sizeof *avail + num * sizeof *avail->ring + s) &&
557 access_ok(VERIFY_WRITE, used,
558 sizeof *used + num * sizeof *used->ring + s);
559}
560
561/* Can we log writes? */
562/* Caller should have device mutex but not vq mutex */
563int vhost_log_access_ok(struct vhost_dev *dev)
564{
565 struct vhost_memory *mp;
566
567 mp = rcu_dereference_protected(dev->memory,
568 lockdep_is_held(&dev->mutex));
569 return memory_access_ok(dev, mp, 1);
570}
571
572/* Verify access for write logging. */
573/* Caller should have vq mutex and device mutex */
574static int vq_log_access_ok(struct vhost_dev *d, struct vhost_virtqueue *vq,
575 void __user *log_base)
576{
577 struct vhost_memory *mp;
578 size_t s = vhost_has_feature(d, VIRTIO_RING_F_EVENT_IDX) ? 2 : 0;
579
580 mp = rcu_dereference_protected(vq->dev->memory,
581 lockdep_is_held(&vq->mutex));
582 return vq_memory_access_ok(log_base, mp,
583 vhost_has_feature(vq->dev, VHOST_F_LOG_ALL)) &&
584 (!vq->log_used || log_access_ok(log_base, vq->log_addr,
585 sizeof *vq->used +
586 vq->num * sizeof *vq->used->ring + s));
587}
588
589/* Can we start vq? */
590/* Caller should have vq mutex and device mutex */
591int vhost_vq_access_ok(struct vhost_virtqueue *vq)
592{
593 return vq_access_ok(vq->dev, vq->num, vq->desc, vq->avail, vq->used) &&
594 vq_log_access_ok(vq->dev, vq, vq->log_base);
595}
596
597static long vhost_set_memory(struct vhost_dev *d, struct vhost_memory __user *m)
598{
599 struct vhost_memory mem, *newmem, *oldmem;
600 unsigned long size = offsetof(struct vhost_memory, regions);
601
602 if (copy_from_user(&mem, m, size))
603 return -EFAULT;
604 if (mem.padding)
605 return -EOPNOTSUPP;
606 if (mem.nregions > VHOST_MEMORY_MAX_NREGIONS)
607 return -E2BIG;
608 newmem = kmalloc(size + mem.nregions * sizeof *m->regions, GFP_KERNEL);
609 if (!newmem)
610 return -ENOMEM;
611
612 memcpy(newmem, &mem, size);
613 if (copy_from_user(newmem->regions, m->regions,
614 mem.nregions * sizeof *m->regions)) {
615 kfree(newmem);
616 return -EFAULT;
617 }
618
619 if (!memory_access_ok(d, newmem,
620 vhost_has_feature(d, VHOST_F_LOG_ALL))) {
621 kfree(newmem);
622 return -EFAULT;
623 }
624 oldmem = rcu_dereference_protected(d->memory,
625 lockdep_is_held(&d->mutex));
626 rcu_assign_pointer(d->memory, newmem);
627 synchronize_rcu();
628 kfree(oldmem);
629 return 0;
630}
631
632static long vhost_set_vring(struct vhost_dev *d, int ioctl, void __user *argp)
633{
634 struct file *eventfp, *filep = NULL,
635 *pollstart = NULL, *pollstop = NULL;
636 struct eventfd_ctx *ctx = NULL;
637 u32 __user *idxp = argp;
638 struct vhost_virtqueue *vq;
639 struct vhost_vring_state s;
640 struct vhost_vring_file f;
641 struct vhost_vring_addr a;
642 u32 idx;
643 long r;
644
645 r = get_user(idx, idxp);
646 if (r < 0)
647 return r;
648 if (idx >= d->nvqs)
649 return -ENOBUFS;
650
651 vq = d->vqs + idx;
652
653 mutex_lock(&vq->mutex);
654
655 switch (ioctl) {
656 case VHOST_SET_VRING_NUM:
657 /* Resizing ring with an active backend?
658 * You don't want to do that. */
659 if (vq->private_data) {
660 r = -EBUSY;
661 break;
662 }
663 if (copy_from_user(&s, argp, sizeof s)) {
664 r = -EFAULT;
665 break;
666 }
667 if (!s.num || s.num > 0xffff || (s.num & (s.num - 1))) {
668 r = -EINVAL;
669 break;
670 }
671 vq->num = s.num;
672 break;
673 case VHOST_SET_VRING_BASE:
674 /* Moving base with an active backend?
675 * You don't want to do that. */
676 if (vq->private_data) {
677 r = -EBUSY;
678 break;
679 }
680 if (copy_from_user(&s, argp, sizeof s)) {
681 r = -EFAULT;
682 break;
683 }
684 if (s.num > 0xffff) {
685 r = -EINVAL;
686 break;
687 }
688 vq->last_avail_idx = s.num;
689 /* Forget the cached index value. */
690 vq->avail_idx = vq->last_avail_idx;
691 break;
692 case VHOST_GET_VRING_BASE:
693 s.index = idx;
694 s.num = vq->last_avail_idx;
695 if (copy_to_user(argp, &s, sizeof s))
696 r = -EFAULT;
697 break;
698 case VHOST_SET_VRING_ADDR:
699 if (copy_from_user(&a, argp, sizeof a)) {
700 r = -EFAULT;
701 break;
702 }
703 if (a.flags & ~(0x1 << VHOST_VRING_F_LOG)) {
704 r = -EOPNOTSUPP;
705 break;
706 }
707 /* For 32bit, verify that the top 32bits of the user
708 data are set to zero. */
709 if ((u64)(unsigned long)a.desc_user_addr != a.desc_user_addr ||
710 (u64)(unsigned long)a.used_user_addr != a.used_user_addr ||
711 (u64)(unsigned long)a.avail_user_addr != a.avail_user_addr) {
712 r = -EFAULT;
713 break;
714 }
715 if ((a.avail_user_addr & (sizeof *vq->avail->ring - 1)) ||
716 (a.used_user_addr & (sizeof *vq->used->ring - 1)) ||
717 (a.log_guest_addr & (sizeof *vq->used->ring - 1))) {
718 r = -EINVAL;
719 break;
720 }
721
722 /* We only verify access here if backend is configured.
723 * If it is not, we don't as size might not have been setup.
724 * We will verify when backend is configured. */
725 if (vq->private_data) {
726 if (!vq_access_ok(d, vq->num,
727 (void __user *)(unsigned long)a.desc_user_addr,
728 (void __user *)(unsigned long)a.avail_user_addr,
729 (void __user *)(unsigned long)a.used_user_addr)) {
730 r = -EINVAL;
731 break;
732 }
733
734 /* Also validate log access for used ring if enabled. */
735 if ((a.flags & (0x1 << VHOST_VRING_F_LOG)) &&
736 !log_access_ok(vq->log_base, a.log_guest_addr,
737 sizeof *vq->used +
738 vq->num * sizeof *vq->used->ring)) {
739 r = -EINVAL;
740 break;
741 }
742 }
743
744 vq->log_used = !!(a.flags & (0x1 << VHOST_VRING_F_LOG));
745 vq->desc = (void __user *)(unsigned long)a.desc_user_addr;
746 vq->avail = (void __user *)(unsigned long)a.avail_user_addr;
747 vq->log_addr = a.log_guest_addr;
748 vq->used = (void __user *)(unsigned long)a.used_user_addr;
749 break;
750 case VHOST_SET_VRING_KICK:
751 if (copy_from_user(&f, argp, sizeof f)) {
752 r = -EFAULT;
753 break;
754 }
755 eventfp = f.fd == -1 ? NULL : eventfd_fget(f.fd);
756 if (IS_ERR(eventfp)) {
757 r = PTR_ERR(eventfp);
758 break;
759 }
760 if (eventfp != vq->kick) {
761 pollstop = filep = vq->kick;
762 pollstart = vq->kick = eventfp;
763 } else
764 filep = eventfp;
765 break;
766 case VHOST_SET_VRING_CALL:
767 if (copy_from_user(&f, argp, sizeof f)) {
768 r = -EFAULT;
769 break;
770 }
771 eventfp = f.fd == -1 ? NULL : eventfd_fget(f.fd);
772 if (IS_ERR(eventfp)) {
773 r = PTR_ERR(eventfp);
774 break;
775 }
776 if (eventfp != vq->call) {
777 filep = vq->call;
778 ctx = vq->call_ctx;
779 vq->call = eventfp;
780 vq->call_ctx = eventfp ?
781 eventfd_ctx_fileget(eventfp) : NULL;
782 } else
783 filep = eventfp;
784 break;
785 case VHOST_SET_VRING_ERR:
786 if (copy_from_user(&f, argp, sizeof f)) {
787 r = -EFAULT;
788 break;
789 }
790 eventfp = f.fd == -1 ? NULL : eventfd_fget(f.fd);
791 if (IS_ERR(eventfp)) {
792 r = PTR_ERR(eventfp);
793 break;
794 }
795 if (eventfp != vq->error) {
796 filep = vq->error;
797 vq->error = eventfp;
798 ctx = vq->error_ctx;
799 vq->error_ctx = eventfp ?
800 eventfd_ctx_fileget(eventfp) : NULL;
801 } else
802 filep = eventfp;
803 break;
804 default:
805 r = -ENOIOCTLCMD;
806 }
807
808 if (pollstop && vq->handle_kick)
809 vhost_poll_stop(&vq->poll);
810
811 if (ctx)
812 eventfd_ctx_put(ctx);
813 if (filep)
814 fput(filep);
815
816 if (pollstart && vq->handle_kick)
817 vhost_poll_start(&vq->poll, vq->kick);
818
819 mutex_unlock(&vq->mutex);
820
821 if (pollstop && vq->handle_kick)
822 vhost_poll_flush(&vq->poll);
823 return r;
824}
825
826/* Caller must have device mutex */
827long vhost_dev_ioctl(struct vhost_dev *d, unsigned int ioctl, unsigned long arg)
828{
829 void __user *argp = (void __user *)arg;
830 struct file *eventfp, *filep = NULL;
831 struct eventfd_ctx *ctx = NULL;
832 u64 p;
833 long r;
834 int i, fd;
835
836 /* If you are not the owner, you can become one */
837 if (ioctl == VHOST_SET_OWNER) {
838 r = vhost_dev_set_owner(d);
839 goto done;
840 }
841
842 /* You must be the owner to do anything else */
843 r = vhost_dev_check_owner(d);
844 if (r)
845 goto done;
846
847 switch (ioctl) {
848 case VHOST_SET_MEM_TABLE:
849 r = vhost_set_memory(d, argp);
850 break;
851 case VHOST_SET_LOG_BASE:
852 if (copy_from_user(&p, argp, sizeof p)) {
853 r = -EFAULT;
854 break;
855 }
856 if ((u64)(unsigned long)p != p) {
857 r = -EFAULT;
858 break;
859 }
860 for (i = 0; i < d->nvqs; ++i) {
861 struct vhost_virtqueue *vq;
862 void __user *base = (void __user *)(unsigned long)p;
863 vq = d->vqs + i;
864 mutex_lock(&vq->mutex);
865 /* If ring is inactive, will check when it's enabled. */
866 if (vq->private_data && !vq_log_access_ok(d, vq, base))
867 r = -EFAULT;
868 else
869 vq->log_base = base;
870 mutex_unlock(&vq->mutex);
871 }
872 break;
873 case VHOST_SET_LOG_FD:
874 r = get_user(fd, (int __user *)argp);
875 if (r < 0)
876 break;
877 eventfp = fd == -1 ? NULL : eventfd_fget(fd);
878 if (IS_ERR(eventfp)) {
879 r = PTR_ERR(eventfp);
880 break;
881 }
882 if (eventfp != d->log_file) {
883 filep = d->log_file;
884 ctx = d->log_ctx;
885 d->log_ctx = eventfp ?
886 eventfd_ctx_fileget(eventfp) : NULL;
887 } else
888 filep = eventfp;
889 for (i = 0; i < d->nvqs; ++i) {
890 mutex_lock(&d->vqs[i].mutex);
891 d->vqs[i].log_ctx = d->log_ctx;
892 mutex_unlock(&d->vqs[i].mutex);
893 }
894 if (ctx)
895 eventfd_ctx_put(ctx);
896 if (filep)
897 fput(filep);
898 break;
899 default:
900 r = vhost_set_vring(d, ioctl, argp);
901 break;
902 }
903done:
904 return r;
905}
906
907static const struct vhost_memory_region *find_region(struct vhost_memory *mem,
908 __u64 addr, __u32 len)
909{
910 struct vhost_memory_region *reg;
911 int i;
912
913 /* linear search is not brilliant, but we really have on the order of 6
914 * regions in practice */
915 for (i = 0; i < mem->nregions; ++i) {
916 reg = mem->regions + i;
917 if (reg->guest_phys_addr <= addr &&
918 reg->guest_phys_addr + reg->memory_size - 1 >= addr)
919 return reg;
920 }
921 return NULL;
922}
923
924/* TODO: This is really inefficient. We need something like get_user()
925 * (instruction directly accesses the data, with an exception table entry
926 * returning -EFAULT). See Documentation/x86/exception-tables.txt.
927 */
928static int set_bit_to_user(int nr, void __user *addr)
929{
930 unsigned long log = (unsigned long)addr;
931 struct page *page;
932 void *base;
933 int bit = nr + (log % PAGE_SIZE) * 8;
934 int r;
935
936 r = get_user_pages_fast(log, 1, 1, &page);
937 if (r < 0)
938 return r;
939 BUG_ON(r != 1);
940 base = kmap_atomic(page, KM_USER0);
941 set_bit(bit, base);
942 kunmap_atomic(base, KM_USER0);
943 set_page_dirty_lock(page);
944 put_page(page);
945 return 0;
946}
947
948static int log_write(void __user *log_base,
949 u64 write_address, u64 write_length)
950{
951 u64 write_page = write_address / VHOST_PAGE_SIZE;
952 int r;
953
954 if (!write_length)
955 return 0;
956 write_length += write_address % VHOST_PAGE_SIZE;
957 for (;;) {
958 u64 base = (u64)(unsigned long)log_base;
959 u64 log = base + write_page / 8;
960 int bit = write_page % 8;
961 if ((u64)(unsigned long)log != log)
962 return -EFAULT;
963 r = set_bit_to_user(bit, (void __user *)(unsigned long)log);
964 if (r < 0)
965 return r;
966 if (write_length <= VHOST_PAGE_SIZE)
967 break;
968 write_length -= VHOST_PAGE_SIZE;
969 write_page += 1;
970 }
971 return r;
972}
973
974int vhost_log_write(struct vhost_virtqueue *vq, struct vhost_log *log,
975 unsigned int log_num, u64 len)
976{
977 int i, r;
978
979 /* Make sure data written is seen before log. */
980 smp_wmb();
981 for (i = 0; i < log_num; ++i) {
982 u64 l = min(log[i].len, len);
983 r = log_write(vq->log_base, log[i].addr, l);
984 if (r < 0)
985 return r;
986 len -= l;
987 if (!len) {
988 if (vq->log_ctx)
989 eventfd_signal(vq->log_ctx, 1);
990 return 0;
991 }
992 }
993 /* Length written exceeds what we have stored. This is a bug. */
994 BUG();
995 return 0;
996}
997
998static int vhost_update_used_flags(struct vhost_virtqueue *vq)
999{
1000 void __user *used;
1001 if (__put_user(vq->used_flags, &vq->used->flags) < 0)
1002 return -EFAULT;
1003 if (unlikely(vq->log_used)) {
1004 /* Make sure the flag is seen before log. */
1005 smp_wmb();
1006 /* Log used flag write. */
1007 used = &vq->used->flags;
1008 log_write(vq->log_base, vq->log_addr +
1009 (used - (void __user *)vq->used),
1010 sizeof vq->used->flags);
1011 if (vq->log_ctx)
1012 eventfd_signal(vq->log_ctx, 1);
1013 }
1014 return 0;
1015}
1016
1017static int vhost_update_avail_event(struct vhost_virtqueue *vq, u16 avail_event)
1018{
1019 if (__put_user(vq->avail_idx, vhost_avail_event(vq)))
1020 return -EFAULT;
1021 if (unlikely(vq->log_used)) {
1022 void __user *used;
1023 /* Make sure the event is seen before log. */
1024 smp_wmb();
1025 /* Log avail event write */
1026 used = vhost_avail_event(vq);
1027 log_write(vq->log_base, vq->log_addr +
1028 (used - (void __user *)vq->used),
1029 sizeof *vhost_avail_event(vq));
1030 if (vq->log_ctx)
1031 eventfd_signal(vq->log_ctx, 1);
1032 }
1033 return 0;
1034}
1035
1036int vhost_init_used(struct vhost_virtqueue *vq)
1037{
1038 int r;
1039 if (!vq->private_data)
1040 return 0;
1041
1042 r = vhost_update_used_flags(vq);
1043 if (r)
1044 return r;
1045 vq->signalled_used_valid = false;
1046 return get_user(vq->last_used_idx, &vq->used->idx);
1047}
1048
1049static int translate_desc(struct vhost_dev *dev, u64 addr, u32 len,
1050 struct iovec iov[], int iov_size)
1051{
1052 const struct vhost_memory_region *reg;
1053 struct vhost_memory *mem;
1054 struct iovec *_iov;
1055 u64 s = 0;
1056 int ret = 0;
1057
1058 rcu_read_lock();
1059
1060 mem = rcu_dereference(dev->memory);
1061 while ((u64)len > s) {
1062 u64 size;
1063 if (unlikely(ret >= iov_size)) {
1064 ret = -ENOBUFS;
1065 break;
1066 }
1067 reg = find_region(mem, addr, len);
1068 if (unlikely(!reg)) {
1069 ret = -EFAULT;
1070 break;
1071 }
1072 _iov = iov + ret;
1073 size = reg->memory_size - addr + reg->guest_phys_addr;
1074 _iov->iov_len = min((u64)len, size);
1075 _iov->iov_base = (void __user *)(unsigned long)
1076 (reg->userspace_addr + addr - reg->guest_phys_addr);
1077 s += size;
1078 addr += size;
1079 ++ret;
1080 }
1081
1082 rcu_read_unlock();
1083 return ret;
1084}
1085
1086/* Each buffer in the virtqueues is actually a chain of descriptors. This
1087 * function returns the next descriptor in the chain,
1088 * or -1U if we're at the end. */
1089static unsigned next_desc(struct vring_desc *desc)
1090{
1091 unsigned int next;
1092
1093 /* If this descriptor says it doesn't chain, we're done. */
1094 if (!(desc->flags & VRING_DESC_F_NEXT))
1095 return -1U;
1096
1097 /* Check they're not leading us off end of descriptors. */
1098 next = desc->next;
1099 /* Make sure compiler knows to grab that: we don't want it changing! */
1100 /* We will use the result as an index in an array, so most
1101 * architectures only need a compiler barrier here. */
1102 read_barrier_depends();
1103
1104 return next;
1105}
1106
1107static int get_indirect(struct vhost_dev *dev, struct vhost_virtqueue *vq,
1108 struct iovec iov[], unsigned int iov_size,
1109 unsigned int *out_num, unsigned int *in_num,
1110 struct vhost_log *log, unsigned int *log_num,
1111 struct vring_desc *indirect)
1112{
1113 struct vring_desc desc;
1114 unsigned int i = 0, count, found = 0;
1115 int ret;
1116
1117 /* Sanity check */
1118 if (unlikely(indirect->len % sizeof desc)) {
1119 vq_err(vq, "Invalid length in indirect descriptor: "
1120 "len 0x%llx not multiple of 0x%zx\n",
1121 (unsigned long long)indirect->len,
1122 sizeof desc);
1123 return -EINVAL;
1124 }
1125
1126 ret = translate_desc(dev, indirect->addr, indirect->len, vq->indirect,
1127 UIO_MAXIOV);
1128 if (unlikely(ret < 0)) {
1129 vq_err(vq, "Translation failure %d in indirect.\n", ret);
1130 return ret;
1131 }
1132
1133 /* We will use the result as an address to read from, so most
1134 * architectures only need a compiler barrier here. */
1135 read_barrier_depends();
1136
1137 count = indirect->len / sizeof desc;
1138 /* Buffers are chained via a 16 bit next field, so
1139 * we can have at most 2^16 of these. */
1140 if (unlikely(count > USHRT_MAX + 1)) {
1141 vq_err(vq, "Indirect buffer length too big: %d\n",
1142 indirect->len);
1143 return -E2BIG;
1144 }
1145
1146 do {
1147 unsigned iov_count = *in_num + *out_num;
1148 if (unlikely(++found > count)) {
1149 vq_err(vq, "Loop detected: last one at %u "
1150 "indirect size %u\n",
1151 i, count);
1152 return -EINVAL;
1153 }
1154 if (unlikely(memcpy_fromiovec((unsigned char *)&desc,
1155 vq->indirect, sizeof desc))) {
1156 vq_err(vq, "Failed indirect descriptor: idx %d, %zx\n",
1157 i, (size_t)indirect->addr + i * sizeof desc);
1158 return -EINVAL;
1159 }
1160 if (unlikely(desc.flags & VRING_DESC_F_INDIRECT)) {
1161 vq_err(vq, "Nested indirect descriptor: idx %d, %zx\n",
1162 i, (size_t)indirect->addr + i * sizeof desc);
1163 return -EINVAL;
1164 }
1165
1166 ret = translate_desc(dev, desc.addr, desc.len, iov + iov_count,
1167 iov_size - iov_count);
1168 if (unlikely(ret < 0)) {
1169 vq_err(vq, "Translation failure %d indirect idx %d\n",
1170 ret, i);
1171 return ret;
1172 }
1173 /* If this is an input descriptor, increment that count. */
1174 if (desc.flags & VRING_DESC_F_WRITE) {
1175 *in_num += ret;
1176 if (unlikely(log)) {
1177 log[*log_num].addr = desc.addr;
1178 log[*log_num].len = desc.len;
1179 ++*log_num;
1180 }
1181 } else {
1182 /* If it's an output descriptor, they're all supposed
1183 * to come before any input descriptors. */
1184 if (unlikely(*in_num)) {
1185 vq_err(vq, "Indirect descriptor "
1186 "has out after in: idx %d\n", i);
1187 return -EINVAL;
1188 }
1189 *out_num += ret;
1190 }
1191 } while ((i = next_desc(&desc)) != -1);
1192 return 0;
1193}
1194
1195/* This looks in the virtqueue and for the first available buffer, and converts
1196 * it to an iovec for convenient access. Since descriptors consist of some
1197 * number of output then some number of input descriptors, it's actually two
1198 * iovecs, but we pack them into one and note how many of each there were.
1199 *
1200 * This function returns the descriptor number found, or vq->num (which is
1201 * never a valid descriptor number) if none was found. A negative code is
1202 * returned on error. */
1203int vhost_get_vq_desc(struct vhost_dev *dev, struct vhost_virtqueue *vq,
1204 struct iovec iov[], unsigned int iov_size,
1205 unsigned int *out_num, unsigned int *in_num,
1206 struct vhost_log *log, unsigned int *log_num)
1207{
1208 struct vring_desc desc;
1209 unsigned int i, head, found = 0;
1210 u16 last_avail_idx;
1211 int ret;
1212
1213 /* Check it isn't doing very strange things with descriptor numbers. */
1214 last_avail_idx = vq->last_avail_idx;
1215 if (unlikely(__get_user(vq->avail_idx, &vq->avail->idx))) {
1216 vq_err(vq, "Failed to access avail idx at %p\n",
1217 &vq->avail->idx);
1218 return -EFAULT;
1219 }
1220
1221 if (unlikely((u16)(vq->avail_idx - last_avail_idx) > vq->num)) {
1222 vq_err(vq, "Guest moved used index from %u to %u",
1223 last_avail_idx, vq->avail_idx);
1224 return -EFAULT;
1225 }
1226
1227 /* If there's nothing new since last we looked, return invalid. */
1228 if (vq->avail_idx == last_avail_idx)
1229 return vq->num;
1230
1231 /* Only get avail ring entries after they have been exposed by guest. */
1232 smp_rmb();
1233
1234 /* Grab the next descriptor number they're advertising, and increment
1235 * the index we've seen. */
1236 if (unlikely(__get_user(head,
1237 &vq->avail->ring[last_avail_idx % vq->num]))) {
1238 vq_err(vq, "Failed to read head: idx %d address %p\n",
1239 last_avail_idx,
1240 &vq->avail->ring[last_avail_idx % vq->num]);
1241 return -EFAULT;
1242 }
1243
1244 /* If their number is silly, that's an error. */
1245 if (unlikely(head >= vq->num)) {
1246 vq_err(vq, "Guest says index %u > %u is available",
1247 head, vq->num);
1248 return -EINVAL;
1249 }
1250
1251 /* When we start there are none of either input nor output. */
1252 *out_num = *in_num = 0;
1253 if (unlikely(log))
1254 *log_num = 0;
1255
1256 i = head;
1257 do {
1258 unsigned iov_count = *in_num + *out_num;
1259 if (unlikely(i >= vq->num)) {
1260 vq_err(vq, "Desc index is %u > %u, head = %u",
1261 i, vq->num, head);
1262 return -EINVAL;
1263 }
1264 if (unlikely(++found > vq->num)) {
1265 vq_err(vq, "Loop detected: last one at %u "
1266 "vq size %u head %u\n",
1267 i, vq->num, head);
1268 return -EINVAL;
1269 }
1270 ret = __copy_from_user(&desc, vq->desc + i, sizeof desc);
1271 if (unlikely(ret)) {
1272 vq_err(vq, "Failed to get descriptor: idx %d addr %p\n",
1273 i, vq->desc + i);
1274 return -EFAULT;
1275 }
1276 if (desc.flags & VRING_DESC_F_INDIRECT) {
1277 ret = get_indirect(dev, vq, iov, iov_size,
1278 out_num, in_num,
1279 log, log_num, &desc);
1280 if (unlikely(ret < 0)) {
1281 vq_err(vq, "Failure detected "
1282 "in indirect descriptor at idx %d\n", i);
1283 return ret;
1284 }
1285 continue;
1286 }
1287
1288 ret = translate_desc(dev, desc.addr, desc.len, iov + iov_count,
1289 iov_size - iov_count);
1290 if (unlikely(ret < 0)) {
1291 vq_err(vq, "Translation failure %d descriptor idx %d\n",
1292 ret, i);
1293 return ret;
1294 }
1295 if (desc.flags & VRING_DESC_F_WRITE) {
1296 /* If this is an input descriptor,
1297 * increment that count. */
1298 *in_num += ret;
1299 if (unlikely(log)) {
1300 log[*log_num].addr = desc.addr;
1301 log[*log_num].len = desc.len;
1302 ++*log_num;
1303 }
1304 } else {
1305 /* If it's an output descriptor, they're all supposed
1306 * to come before any input descriptors. */
1307 if (unlikely(*in_num)) {
1308 vq_err(vq, "Descriptor has out after in: "
1309 "idx %d\n", i);
1310 return -EINVAL;
1311 }
1312 *out_num += ret;
1313 }
1314 } while ((i = next_desc(&desc)) != -1);
1315
1316 /* On success, increment avail index. */
1317 vq->last_avail_idx++;
1318
1319 /* Assume notifications from guest are disabled at this point,
1320 * if they aren't we would need to update avail_event index. */
1321 BUG_ON(!(vq->used_flags & VRING_USED_F_NO_NOTIFY));
1322 return head;
1323}
1324
1325/* Reverse the effect of vhost_get_vq_desc. Useful for error handling. */
1326void vhost_discard_vq_desc(struct vhost_virtqueue *vq, int n)
1327{
1328 vq->last_avail_idx -= n;
1329}
1330
1331/* After we've used one of their buffers, we tell them about it. We'll then
1332 * want to notify the guest, using eventfd. */
1333int vhost_add_used(struct vhost_virtqueue *vq, unsigned int head, int len)
1334{
1335 struct vring_used_elem __user *used;
1336
1337 /* The virtqueue contains a ring of used buffers. Get a pointer to the
1338 * next entry in that used ring. */
1339 used = &vq->used->ring[vq->last_used_idx % vq->num];
1340 if (__put_user(head, &used->id)) {
1341 vq_err(vq, "Failed to write used id");
1342 return -EFAULT;
1343 }
1344 if (__put_user(len, &used->len)) {
1345 vq_err(vq, "Failed to write used len");
1346 return -EFAULT;
1347 }
1348 /* Make sure buffer is written before we update index. */
1349 smp_wmb();
1350 if (__put_user(vq->last_used_idx + 1, &vq->used->idx)) {
1351 vq_err(vq, "Failed to increment used idx");
1352 return -EFAULT;
1353 }
1354 if (unlikely(vq->log_used)) {
1355 /* Make sure data is seen before log. */
1356 smp_wmb();
1357 /* Log used ring entry write. */
1358 log_write(vq->log_base,
1359 vq->log_addr +
1360 ((void __user *)used - (void __user *)vq->used),
1361 sizeof *used);
1362 /* Log used index update. */
1363 log_write(vq->log_base,
1364 vq->log_addr + offsetof(struct vring_used, idx),
1365 sizeof vq->used->idx);
1366 if (vq->log_ctx)
1367 eventfd_signal(vq->log_ctx, 1);
1368 }
1369 vq->last_used_idx++;
1370 /* If the driver never bothers to signal in a very long while,
1371 * used index might wrap around. If that happens, invalidate
1372 * signalled_used index we stored. TODO: make sure driver
1373 * signals at least once in 2^16 and remove this. */
1374 if (unlikely(vq->last_used_idx == vq->signalled_used))
1375 vq->signalled_used_valid = false;
1376 return 0;
1377}
1378
1379static int __vhost_add_used_n(struct vhost_virtqueue *vq,
1380 struct vring_used_elem *heads,
1381 unsigned count)
1382{
1383 struct vring_used_elem __user *used;
1384 u16 old, new;
1385 int start;
1386
1387 start = vq->last_used_idx % vq->num;
1388 used = vq->used->ring + start;
1389 if (__copy_to_user(used, heads, count * sizeof *used)) {
1390 vq_err(vq, "Failed to write used");
1391 return -EFAULT;
1392 }
1393 if (unlikely(vq->log_used)) {
1394 /* Make sure data is seen before log. */
1395 smp_wmb();
1396 /* Log used ring entry write. */
1397 log_write(vq->log_base,
1398 vq->log_addr +
1399 ((void __user *)used - (void __user *)vq->used),
1400 count * sizeof *used);
1401 }
1402 old = vq->last_used_idx;
1403 new = (vq->last_used_idx += count);
1404 /* If the driver never bothers to signal in a very long while,
1405 * used index might wrap around. If that happens, invalidate
1406 * signalled_used index we stored. TODO: make sure driver
1407 * signals at least once in 2^16 and remove this. */
1408 if (unlikely((u16)(new - vq->signalled_used) < (u16)(new - old)))
1409 vq->signalled_used_valid = false;
1410 return 0;
1411}
1412
1413/* After we've used one of their buffers, we tell them about it. We'll then
1414 * want to notify the guest, using eventfd. */
1415int vhost_add_used_n(struct vhost_virtqueue *vq, struct vring_used_elem *heads,
1416 unsigned count)
1417{
1418 int start, n, r;
1419
1420 start = vq->last_used_idx % vq->num;
1421 n = vq->num - start;
1422 if (n < count) {
1423 r = __vhost_add_used_n(vq, heads, n);
1424 if (r < 0)
1425 return r;
1426 heads += n;
1427 count -= n;
1428 }
1429 r = __vhost_add_used_n(vq, heads, count);
1430
1431 /* Make sure buffer is written before we update index. */
1432 smp_wmb();
1433 if (put_user(vq->last_used_idx, &vq->used->idx)) {
1434 vq_err(vq, "Failed to increment used idx");
1435 return -EFAULT;
1436 }
1437 if (unlikely(vq->log_used)) {
1438 /* Log used index update. */
1439 log_write(vq->log_base,
1440 vq->log_addr + offsetof(struct vring_used, idx),
1441 sizeof vq->used->idx);
1442 if (vq->log_ctx)
1443 eventfd_signal(vq->log_ctx, 1);
1444 }
1445 return r;
1446}
1447
1448static bool vhost_notify(struct vhost_dev *dev, struct vhost_virtqueue *vq)
1449{
1450 __u16 old, new, event;
1451 bool v;
1452 /* Flush out used index updates. This is paired
1453 * with the barrier that the Guest executes when enabling
1454 * interrupts. */
1455 smp_mb();
1456
1457 if (vhost_has_feature(dev, VIRTIO_F_NOTIFY_ON_EMPTY) &&
1458 unlikely(vq->avail_idx == vq->last_avail_idx))
1459 return true;
1460
1461 if (!vhost_has_feature(dev, VIRTIO_RING_F_EVENT_IDX)) {
1462 __u16 flags;
1463 if (__get_user(flags, &vq->avail->flags)) {
1464 vq_err(vq, "Failed to get flags");
1465 return true;
1466 }
1467 return !(flags & VRING_AVAIL_F_NO_INTERRUPT);
1468 }
1469 old = vq->signalled_used;
1470 v = vq->signalled_used_valid;
1471 new = vq->signalled_used = vq->last_used_idx;
1472 vq->signalled_used_valid = true;
1473
1474 if (unlikely(!v))
1475 return true;
1476
1477 if (get_user(event, vhost_used_event(vq))) {
1478 vq_err(vq, "Failed to get used event idx");
1479 return true;
1480 }
1481 return vring_need_event(event, new, old);
1482}
1483
1484/* This actually signals the guest, using eventfd. */
1485void vhost_signal(struct vhost_dev *dev, struct vhost_virtqueue *vq)
1486{
1487 /* Signal the Guest tell them we used something up. */
1488 if (vq->call_ctx && vhost_notify(dev, vq))
1489 eventfd_signal(vq->call_ctx, 1);
1490}
1491
1492/* And here's the combo meal deal. Supersize me! */
1493void vhost_add_used_and_signal(struct vhost_dev *dev,
1494 struct vhost_virtqueue *vq,
1495 unsigned int head, int len)
1496{
1497 vhost_add_used(vq, head, len);
1498 vhost_signal(dev, vq);
1499}
1500
1501/* multi-buffer version of vhost_add_used_and_signal */
1502void vhost_add_used_and_signal_n(struct vhost_dev *dev,
1503 struct vhost_virtqueue *vq,
1504 struct vring_used_elem *heads, unsigned count)
1505{
1506 vhost_add_used_n(vq, heads, count);
1507 vhost_signal(dev, vq);
1508}
1509
1510/* OK, now we need to know about added descriptors. */
1511bool vhost_enable_notify(struct vhost_dev *dev, struct vhost_virtqueue *vq)
1512{
1513 u16 avail_idx;
1514 int r;
1515
1516 if (!(vq->used_flags & VRING_USED_F_NO_NOTIFY))
1517 return false;
1518 vq->used_flags &= ~VRING_USED_F_NO_NOTIFY;
1519 if (!vhost_has_feature(dev, VIRTIO_RING_F_EVENT_IDX)) {
1520 r = vhost_update_used_flags(vq);
1521 if (r) {
1522 vq_err(vq, "Failed to enable notification at %p: %d\n",
1523 &vq->used->flags, r);
1524 return false;
1525 }
1526 } else {
1527 r = vhost_update_avail_event(vq, vq->avail_idx);
1528 if (r) {
1529 vq_err(vq, "Failed to update avail event index at %p: %d\n",
1530 vhost_avail_event(vq), r);
1531 return false;
1532 }
1533 }
1534 /* They could have slipped one in as we were doing that: make
1535 * sure it's written, then check again. */
1536 smp_mb();
1537 r = __get_user(avail_idx, &vq->avail->idx);
1538 if (r) {
1539 vq_err(vq, "Failed to check avail idx at %p: %d\n",
1540 &vq->avail->idx, r);
1541 return false;
1542 }
1543
1544 return avail_idx != vq->avail_idx;
1545}
1546
1547/* We don't need to be notified again. */
1548void vhost_disable_notify(struct vhost_dev *dev, struct vhost_virtqueue *vq)
1549{
1550 int r;
1551
1552 if (vq->used_flags & VRING_USED_F_NO_NOTIFY)
1553 return;
1554 vq->used_flags |= VRING_USED_F_NO_NOTIFY;
1555 if (!vhost_has_feature(dev, VIRTIO_RING_F_EVENT_IDX)) {
1556 r = vhost_update_used_flags(vq);
1557 if (r)
1558 vq_err(vq, "Failed to enable notification at %p: %d\n",
1559 &vq->used->flags, r);
1560 }
1561}
1562
1563static void vhost_zerocopy_done_signal(struct kref *kref)
1564{
1565 struct vhost_ubuf_ref *ubufs = container_of(kref, struct vhost_ubuf_ref,
1566 kref);
1567 wake_up(&ubufs->wait);
1568}
1569
1570struct vhost_ubuf_ref *vhost_ubuf_alloc(struct vhost_virtqueue *vq,
1571 bool zcopy)
1572{
1573 struct vhost_ubuf_ref *ubufs;
1574 /* No zero copy backend? Nothing to count. */
1575 if (!zcopy)
1576 return NULL;
1577 ubufs = kmalloc(sizeof *ubufs, GFP_KERNEL);
1578 if (!ubufs)
1579 return ERR_PTR(-ENOMEM);
1580 kref_init(&ubufs->kref);
1581 init_waitqueue_head(&ubufs->wait);
1582 ubufs->vq = vq;
1583 return ubufs;
1584}
1585
1586void vhost_ubuf_put(struct vhost_ubuf_ref *ubufs)
1587{
1588 kref_put(&ubufs->kref, vhost_zerocopy_done_signal);
1589}
1590
1591void vhost_ubuf_put_and_wait(struct vhost_ubuf_ref *ubufs)
1592{
1593 kref_put(&ubufs->kref, vhost_zerocopy_done_signal);
1594 wait_event(ubufs->wait, !atomic_read(&ubufs->kref.refcount));
1595 kfree(ubufs);
1596}
1597
1598void vhost_zerocopy_callback(void *arg)
1599{
1600 struct ubuf_info *ubuf = arg;
1601 struct vhost_ubuf_ref *ubufs = ubuf->arg;
1602 struct vhost_virtqueue *vq = ubufs->vq;
1603
1604 /* set len = 1 to mark this desc buffers done DMA */
1605 vq->heads[ubuf->desc].len = VHOST_DMA_DONE_LEN;
1606 kref_put(&ubufs->kref, vhost_zerocopy_done_signal);
1607}
1// SPDX-License-Identifier: GPL-2.0-only
2/* Copyright (C) 2009 Red Hat, Inc.
3 * Copyright (C) 2006 Rusty Russell IBM Corporation
4 *
5 * Author: Michael S. Tsirkin <mst@redhat.com>
6 *
7 * Inspiration, some code, and most witty comments come from
8 * Documentation/virtual/lguest/lguest.c, by Rusty Russell
9 *
10 * Generic code for virtio server in host kernel.
11 */
12
13#include <linux/eventfd.h>
14#include <linux/vhost.h>
15#include <linux/uio.h>
16#include <linux/mm.h>
17#include <linux/miscdevice.h>
18#include <linux/mutex.h>
19#include <linux/poll.h>
20#include <linux/file.h>
21#include <linux/highmem.h>
22#include <linux/slab.h>
23#include <linux/vmalloc.h>
24#include <linux/kthread.h>
25#include <linux/module.h>
26#include <linux/sort.h>
27#include <linux/sched/mm.h>
28#include <linux/sched/signal.h>
29#include <linux/sched/vhost_task.h>
30#include <linux/interval_tree_generic.h>
31#include <linux/nospec.h>
32#include <linux/kcov.h>
33
34#include "vhost.h"
35
36static ushort max_mem_regions = 64;
37module_param(max_mem_regions, ushort, 0444);
38MODULE_PARM_DESC(max_mem_regions,
39 "Maximum number of memory regions in memory map. (default: 64)");
40static int max_iotlb_entries = 2048;
41module_param(max_iotlb_entries, int, 0444);
42MODULE_PARM_DESC(max_iotlb_entries,
43 "Maximum number of iotlb entries. (default: 2048)");
44
45enum {
46 VHOST_MEMORY_F_LOG = 0x1,
47};
48
49#define vhost_used_event(vq) ((__virtio16 __user *)&vq->avail->ring[vq->num])
50#define vhost_avail_event(vq) ((__virtio16 __user *)&vq->used->ring[vq->num])
51
52#ifdef CONFIG_VHOST_CROSS_ENDIAN_LEGACY
53static void vhost_disable_cross_endian(struct vhost_virtqueue *vq)
54{
55 vq->user_be = !virtio_legacy_is_little_endian();
56}
57
58static void vhost_enable_cross_endian_big(struct vhost_virtqueue *vq)
59{
60 vq->user_be = true;
61}
62
63static void vhost_enable_cross_endian_little(struct vhost_virtqueue *vq)
64{
65 vq->user_be = false;
66}
67
68static long vhost_set_vring_endian(struct vhost_virtqueue *vq, int __user *argp)
69{
70 struct vhost_vring_state s;
71
72 if (vq->private_data)
73 return -EBUSY;
74
75 if (copy_from_user(&s, argp, sizeof(s)))
76 return -EFAULT;
77
78 if (s.num != VHOST_VRING_LITTLE_ENDIAN &&
79 s.num != VHOST_VRING_BIG_ENDIAN)
80 return -EINVAL;
81
82 if (s.num == VHOST_VRING_BIG_ENDIAN)
83 vhost_enable_cross_endian_big(vq);
84 else
85 vhost_enable_cross_endian_little(vq);
86
87 return 0;
88}
89
90static long vhost_get_vring_endian(struct vhost_virtqueue *vq, u32 idx,
91 int __user *argp)
92{
93 struct vhost_vring_state s = {
94 .index = idx,
95 .num = vq->user_be
96 };
97
98 if (copy_to_user(argp, &s, sizeof(s)))
99 return -EFAULT;
100
101 return 0;
102}
103
104static void vhost_init_is_le(struct vhost_virtqueue *vq)
105{
106 /* Note for legacy virtio: user_be is initialized at reset time
107 * according to the host endianness. If userspace does not set an
108 * explicit endianness, the default behavior is native endian, as
109 * expected by legacy virtio.
110 */
111 vq->is_le = vhost_has_feature(vq, VIRTIO_F_VERSION_1) || !vq->user_be;
112}
113#else
114static void vhost_disable_cross_endian(struct vhost_virtqueue *vq)
115{
116}
117
118static long vhost_set_vring_endian(struct vhost_virtqueue *vq, int __user *argp)
119{
120 return -ENOIOCTLCMD;
121}
122
123static long vhost_get_vring_endian(struct vhost_virtqueue *vq, u32 idx,
124 int __user *argp)
125{
126 return -ENOIOCTLCMD;
127}
128
129static void vhost_init_is_le(struct vhost_virtqueue *vq)
130{
131 vq->is_le = vhost_has_feature(vq, VIRTIO_F_VERSION_1)
132 || virtio_legacy_is_little_endian();
133}
134#endif /* CONFIG_VHOST_CROSS_ENDIAN_LEGACY */
135
136static void vhost_reset_is_le(struct vhost_virtqueue *vq)
137{
138 vhost_init_is_le(vq);
139}
140
141struct vhost_flush_struct {
142 struct vhost_work work;
143 struct completion wait_event;
144};
145
146static void vhost_flush_work(struct vhost_work *work)
147{
148 struct vhost_flush_struct *s;
149
150 s = container_of(work, struct vhost_flush_struct, work);
151 complete(&s->wait_event);
152}
153
154static void vhost_poll_func(struct file *file, wait_queue_head_t *wqh,
155 poll_table *pt)
156{
157 struct vhost_poll *poll;
158
159 poll = container_of(pt, struct vhost_poll, table);
160 poll->wqh = wqh;
161 add_wait_queue(wqh, &poll->wait);
162}
163
164static int vhost_poll_wakeup(wait_queue_entry_t *wait, unsigned mode, int sync,
165 void *key)
166{
167 struct vhost_poll *poll = container_of(wait, struct vhost_poll, wait);
168 struct vhost_work *work = &poll->work;
169
170 if (!(key_to_poll(key) & poll->mask))
171 return 0;
172
173 if (!poll->dev->use_worker)
174 work->fn(work);
175 else
176 vhost_poll_queue(poll);
177
178 return 0;
179}
180
181void vhost_work_init(struct vhost_work *work, vhost_work_fn_t fn)
182{
183 clear_bit(VHOST_WORK_QUEUED, &work->flags);
184 work->fn = fn;
185}
186EXPORT_SYMBOL_GPL(vhost_work_init);
187
188/* Init poll structure */
189void vhost_poll_init(struct vhost_poll *poll, vhost_work_fn_t fn,
190 __poll_t mask, struct vhost_dev *dev,
191 struct vhost_virtqueue *vq)
192{
193 init_waitqueue_func_entry(&poll->wait, vhost_poll_wakeup);
194 init_poll_funcptr(&poll->table, vhost_poll_func);
195 poll->mask = mask;
196 poll->dev = dev;
197 poll->wqh = NULL;
198 poll->vq = vq;
199
200 vhost_work_init(&poll->work, fn);
201}
202EXPORT_SYMBOL_GPL(vhost_poll_init);
203
204/* Start polling a file. We add ourselves to file's wait queue. The caller must
205 * keep a reference to a file until after vhost_poll_stop is called. */
206int vhost_poll_start(struct vhost_poll *poll, struct file *file)
207{
208 __poll_t mask;
209
210 if (poll->wqh)
211 return 0;
212
213 mask = vfs_poll(file, &poll->table);
214 if (mask)
215 vhost_poll_wakeup(&poll->wait, 0, 0, poll_to_key(mask));
216 if (mask & EPOLLERR) {
217 vhost_poll_stop(poll);
218 return -EINVAL;
219 }
220
221 return 0;
222}
223EXPORT_SYMBOL_GPL(vhost_poll_start);
224
225/* Stop polling a file. After this function returns, it becomes safe to drop the
226 * file reference. You must also flush afterwards. */
227void vhost_poll_stop(struct vhost_poll *poll)
228{
229 if (poll->wqh) {
230 remove_wait_queue(poll->wqh, &poll->wait);
231 poll->wqh = NULL;
232 }
233}
234EXPORT_SYMBOL_GPL(vhost_poll_stop);
235
236static void vhost_worker_queue(struct vhost_worker *worker,
237 struct vhost_work *work)
238{
239 if (!test_and_set_bit(VHOST_WORK_QUEUED, &work->flags)) {
240 /* We can only add the work to the list after we're
241 * sure it was not in the list.
242 * test_and_set_bit() implies a memory barrier.
243 */
244 llist_add(&work->node, &worker->work_list);
245 vhost_task_wake(worker->vtsk);
246 }
247}
248
249bool vhost_vq_work_queue(struct vhost_virtqueue *vq, struct vhost_work *work)
250{
251 struct vhost_worker *worker;
252 bool queued = false;
253
254 rcu_read_lock();
255 worker = rcu_dereference(vq->worker);
256 if (worker) {
257 queued = true;
258 vhost_worker_queue(worker, work);
259 }
260 rcu_read_unlock();
261
262 return queued;
263}
264EXPORT_SYMBOL_GPL(vhost_vq_work_queue);
265
266void vhost_vq_flush(struct vhost_virtqueue *vq)
267{
268 struct vhost_flush_struct flush;
269
270 init_completion(&flush.wait_event);
271 vhost_work_init(&flush.work, vhost_flush_work);
272
273 if (vhost_vq_work_queue(vq, &flush.work))
274 wait_for_completion(&flush.wait_event);
275}
276EXPORT_SYMBOL_GPL(vhost_vq_flush);
277
278/**
279 * vhost_worker_flush - flush a worker
280 * @worker: worker to flush
281 *
282 * This does not use RCU to protect the worker, so the device or worker
283 * mutex must be held.
284 */
285static void vhost_worker_flush(struct vhost_worker *worker)
286{
287 struct vhost_flush_struct flush;
288
289 init_completion(&flush.wait_event);
290 vhost_work_init(&flush.work, vhost_flush_work);
291
292 vhost_worker_queue(worker, &flush.work);
293 wait_for_completion(&flush.wait_event);
294}
295
296void vhost_dev_flush(struct vhost_dev *dev)
297{
298 struct vhost_worker *worker;
299 unsigned long i;
300
301 xa_for_each(&dev->worker_xa, i, worker) {
302 mutex_lock(&worker->mutex);
303 if (!worker->attachment_cnt) {
304 mutex_unlock(&worker->mutex);
305 continue;
306 }
307 vhost_worker_flush(worker);
308 mutex_unlock(&worker->mutex);
309 }
310}
311EXPORT_SYMBOL_GPL(vhost_dev_flush);
312
313/* A lockless hint for busy polling code to exit the loop */
314bool vhost_vq_has_work(struct vhost_virtqueue *vq)
315{
316 struct vhost_worker *worker;
317 bool has_work = false;
318
319 rcu_read_lock();
320 worker = rcu_dereference(vq->worker);
321 if (worker && !llist_empty(&worker->work_list))
322 has_work = true;
323 rcu_read_unlock();
324
325 return has_work;
326}
327EXPORT_SYMBOL_GPL(vhost_vq_has_work);
328
329void vhost_poll_queue(struct vhost_poll *poll)
330{
331 vhost_vq_work_queue(poll->vq, &poll->work);
332}
333EXPORT_SYMBOL_GPL(vhost_poll_queue);
334
335static void __vhost_vq_meta_reset(struct vhost_virtqueue *vq)
336{
337 int j;
338
339 for (j = 0; j < VHOST_NUM_ADDRS; j++)
340 vq->meta_iotlb[j] = NULL;
341}
342
343static void vhost_vq_meta_reset(struct vhost_dev *d)
344{
345 int i;
346
347 for (i = 0; i < d->nvqs; ++i)
348 __vhost_vq_meta_reset(d->vqs[i]);
349}
350
351static void vhost_vring_call_reset(struct vhost_vring_call *call_ctx)
352{
353 call_ctx->ctx = NULL;
354 memset(&call_ctx->producer, 0x0, sizeof(struct irq_bypass_producer));
355}
356
357bool vhost_vq_is_setup(struct vhost_virtqueue *vq)
358{
359 return vq->avail && vq->desc && vq->used && vhost_vq_access_ok(vq);
360}
361EXPORT_SYMBOL_GPL(vhost_vq_is_setup);
362
363static void vhost_vq_reset(struct vhost_dev *dev,
364 struct vhost_virtqueue *vq)
365{
366 vq->num = 1;
367 vq->desc = NULL;
368 vq->avail = NULL;
369 vq->used = NULL;
370 vq->last_avail_idx = 0;
371 vq->avail_idx = 0;
372 vq->last_used_idx = 0;
373 vq->signalled_used = 0;
374 vq->signalled_used_valid = false;
375 vq->used_flags = 0;
376 vq->log_used = false;
377 vq->log_addr = -1ull;
378 vq->private_data = NULL;
379 vq->acked_features = 0;
380 vq->acked_backend_features = 0;
381 vq->log_base = NULL;
382 vq->error_ctx = NULL;
383 vq->kick = NULL;
384 vq->log_ctx = NULL;
385 vhost_disable_cross_endian(vq);
386 vhost_reset_is_le(vq);
387 vq->busyloop_timeout = 0;
388 vq->umem = NULL;
389 vq->iotlb = NULL;
390 rcu_assign_pointer(vq->worker, NULL);
391 vhost_vring_call_reset(&vq->call_ctx);
392 __vhost_vq_meta_reset(vq);
393}
394
395static bool vhost_worker(void *data)
396{
397 struct vhost_worker *worker = data;
398 struct vhost_work *work, *work_next;
399 struct llist_node *node;
400
401 node = llist_del_all(&worker->work_list);
402 if (node) {
403 __set_current_state(TASK_RUNNING);
404
405 node = llist_reverse_order(node);
406 /* make sure flag is seen after deletion */
407 smp_wmb();
408 llist_for_each_entry_safe(work, work_next, node, node) {
409 clear_bit(VHOST_WORK_QUEUED, &work->flags);
410 kcov_remote_start_common(worker->kcov_handle);
411 work->fn(work);
412 kcov_remote_stop();
413 cond_resched();
414 }
415 }
416
417 return !!node;
418}
419
420static void vhost_vq_free_iovecs(struct vhost_virtqueue *vq)
421{
422 kfree(vq->indirect);
423 vq->indirect = NULL;
424 kfree(vq->log);
425 vq->log = NULL;
426 kfree(vq->heads);
427 vq->heads = NULL;
428}
429
430/* Helper to allocate iovec buffers for all vqs. */
431static long vhost_dev_alloc_iovecs(struct vhost_dev *dev)
432{
433 struct vhost_virtqueue *vq;
434 int i;
435
436 for (i = 0; i < dev->nvqs; ++i) {
437 vq = dev->vqs[i];
438 vq->indirect = kmalloc_array(UIO_MAXIOV,
439 sizeof(*vq->indirect),
440 GFP_KERNEL);
441 vq->log = kmalloc_array(dev->iov_limit, sizeof(*vq->log),
442 GFP_KERNEL);
443 vq->heads = kmalloc_array(dev->iov_limit, sizeof(*vq->heads),
444 GFP_KERNEL);
445 if (!vq->indirect || !vq->log || !vq->heads)
446 goto err_nomem;
447 }
448 return 0;
449
450err_nomem:
451 for (; i >= 0; --i)
452 vhost_vq_free_iovecs(dev->vqs[i]);
453 return -ENOMEM;
454}
455
456static void vhost_dev_free_iovecs(struct vhost_dev *dev)
457{
458 int i;
459
460 for (i = 0; i < dev->nvqs; ++i)
461 vhost_vq_free_iovecs(dev->vqs[i]);
462}
463
464bool vhost_exceeds_weight(struct vhost_virtqueue *vq,
465 int pkts, int total_len)
466{
467 struct vhost_dev *dev = vq->dev;
468
469 if ((dev->byte_weight && total_len >= dev->byte_weight) ||
470 pkts >= dev->weight) {
471 vhost_poll_queue(&vq->poll);
472 return true;
473 }
474
475 return false;
476}
477EXPORT_SYMBOL_GPL(vhost_exceeds_weight);
478
479static size_t vhost_get_avail_size(struct vhost_virtqueue *vq,
480 unsigned int num)
481{
482 size_t event __maybe_unused =
483 vhost_has_feature(vq, VIRTIO_RING_F_EVENT_IDX) ? 2 : 0;
484
485 return size_add(struct_size(vq->avail, ring, num), event);
486}
487
488static size_t vhost_get_used_size(struct vhost_virtqueue *vq,
489 unsigned int num)
490{
491 size_t event __maybe_unused =
492 vhost_has_feature(vq, VIRTIO_RING_F_EVENT_IDX) ? 2 : 0;
493
494 return size_add(struct_size(vq->used, ring, num), event);
495}
496
497static size_t vhost_get_desc_size(struct vhost_virtqueue *vq,
498 unsigned int num)
499{
500 return sizeof(*vq->desc) * num;
501}
502
503void vhost_dev_init(struct vhost_dev *dev,
504 struct vhost_virtqueue **vqs, int nvqs,
505 int iov_limit, int weight, int byte_weight,
506 bool use_worker,
507 int (*msg_handler)(struct vhost_dev *dev, u32 asid,
508 struct vhost_iotlb_msg *msg))
509{
510 struct vhost_virtqueue *vq;
511 int i;
512
513 dev->vqs = vqs;
514 dev->nvqs = nvqs;
515 mutex_init(&dev->mutex);
516 dev->log_ctx = NULL;
517 dev->umem = NULL;
518 dev->iotlb = NULL;
519 dev->mm = NULL;
520 dev->iov_limit = iov_limit;
521 dev->weight = weight;
522 dev->byte_weight = byte_weight;
523 dev->use_worker = use_worker;
524 dev->msg_handler = msg_handler;
525 init_waitqueue_head(&dev->wait);
526 INIT_LIST_HEAD(&dev->read_list);
527 INIT_LIST_HEAD(&dev->pending_list);
528 spin_lock_init(&dev->iotlb_lock);
529 xa_init_flags(&dev->worker_xa, XA_FLAGS_ALLOC);
530
531 for (i = 0; i < dev->nvqs; ++i) {
532 vq = dev->vqs[i];
533 vq->log = NULL;
534 vq->indirect = NULL;
535 vq->heads = NULL;
536 vq->dev = dev;
537 mutex_init(&vq->mutex);
538 vhost_vq_reset(dev, vq);
539 if (vq->handle_kick)
540 vhost_poll_init(&vq->poll, vq->handle_kick,
541 EPOLLIN, dev, vq);
542 }
543}
544EXPORT_SYMBOL_GPL(vhost_dev_init);
545
546/* Caller should have device mutex */
547long vhost_dev_check_owner(struct vhost_dev *dev)
548{
549 /* Are you the owner? If not, I don't think you mean to do that */
550 return dev->mm == current->mm ? 0 : -EPERM;
551}
552EXPORT_SYMBOL_GPL(vhost_dev_check_owner);
553
554/* Caller should have device mutex */
555bool vhost_dev_has_owner(struct vhost_dev *dev)
556{
557 return dev->mm;
558}
559EXPORT_SYMBOL_GPL(vhost_dev_has_owner);
560
561static void vhost_attach_mm(struct vhost_dev *dev)
562{
563 /* No owner, become one */
564 if (dev->use_worker) {
565 dev->mm = get_task_mm(current);
566 } else {
567 /* vDPA device does not use worker thead, so there's
568 * no need to hold the address space for mm. This help
569 * to avoid deadlock in the case of mmap() which may
570 * held the refcnt of the file and depends on release
571 * method to remove vma.
572 */
573 dev->mm = current->mm;
574 mmgrab(dev->mm);
575 }
576}
577
578static void vhost_detach_mm(struct vhost_dev *dev)
579{
580 if (!dev->mm)
581 return;
582
583 if (dev->use_worker)
584 mmput(dev->mm);
585 else
586 mmdrop(dev->mm);
587
588 dev->mm = NULL;
589}
590
591static void vhost_worker_destroy(struct vhost_dev *dev,
592 struct vhost_worker *worker)
593{
594 if (!worker)
595 return;
596
597 WARN_ON(!llist_empty(&worker->work_list));
598 xa_erase(&dev->worker_xa, worker->id);
599 vhost_task_stop(worker->vtsk);
600 kfree(worker);
601}
602
603static void vhost_workers_free(struct vhost_dev *dev)
604{
605 struct vhost_worker *worker;
606 unsigned long i;
607
608 if (!dev->use_worker)
609 return;
610
611 for (i = 0; i < dev->nvqs; i++)
612 rcu_assign_pointer(dev->vqs[i]->worker, NULL);
613 /*
614 * Free the default worker we created and cleanup workers userspace
615 * created but couldn't clean up (it forgot or crashed).
616 */
617 xa_for_each(&dev->worker_xa, i, worker)
618 vhost_worker_destroy(dev, worker);
619 xa_destroy(&dev->worker_xa);
620}
621
622static struct vhost_worker *vhost_worker_create(struct vhost_dev *dev)
623{
624 struct vhost_worker *worker;
625 struct vhost_task *vtsk;
626 char name[TASK_COMM_LEN];
627 int ret;
628 u32 id;
629
630 worker = kzalloc(sizeof(*worker), GFP_KERNEL_ACCOUNT);
631 if (!worker)
632 return NULL;
633
634 snprintf(name, sizeof(name), "vhost-%d", current->pid);
635
636 vtsk = vhost_task_create(vhost_worker, worker, name);
637 if (!vtsk)
638 goto free_worker;
639
640 mutex_init(&worker->mutex);
641 init_llist_head(&worker->work_list);
642 worker->kcov_handle = kcov_common_handle();
643 worker->vtsk = vtsk;
644
645 vhost_task_start(vtsk);
646
647 ret = xa_alloc(&dev->worker_xa, &id, worker, xa_limit_32b, GFP_KERNEL);
648 if (ret < 0)
649 goto stop_worker;
650 worker->id = id;
651
652 return worker;
653
654stop_worker:
655 vhost_task_stop(vtsk);
656free_worker:
657 kfree(worker);
658 return NULL;
659}
660
661/* Caller must have device mutex */
662static void __vhost_vq_attach_worker(struct vhost_virtqueue *vq,
663 struct vhost_worker *worker)
664{
665 struct vhost_worker *old_worker;
666
667 old_worker = rcu_dereference_check(vq->worker,
668 lockdep_is_held(&vq->dev->mutex));
669
670 mutex_lock(&worker->mutex);
671 worker->attachment_cnt++;
672 mutex_unlock(&worker->mutex);
673 rcu_assign_pointer(vq->worker, worker);
674
675 if (!old_worker)
676 return;
677 /*
678 * Take the worker mutex to make sure we see the work queued from
679 * device wide flushes which doesn't use RCU for execution.
680 */
681 mutex_lock(&old_worker->mutex);
682 old_worker->attachment_cnt--;
683 /*
684 * We don't want to call synchronize_rcu for every vq during setup
685 * because it will slow down VM startup. If we haven't done
686 * VHOST_SET_VRING_KICK and not done the driver specific
687 * SET_ENDPOINT/RUNNUNG then we can skip the sync since there will
688 * not be any works queued for scsi and net.
689 */
690 mutex_lock(&vq->mutex);
691 if (!vhost_vq_get_backend(vq) && !vq->kick) {
692 mutex_unlock(&vq->mutex);
693 mutex_unlock(&old_worker->mutex);
694 /*
695 * vsock can queue anytime after VHOST_VSOCK_SET_GUEST_CID.
696 * Warn if it adds support for multiple workers but forgets to
697 * handle the early queueing case.
698 */
699 WARN_ON(!old_worker->attachment_cnt &&
700 !llist_empty(&old_worker->work_list));
701 return;
702 }
703 mutex_unlock(&vq->mutex);
704
705 /* Make sure new vq queue/flush/poll calls see the new worker */
706 synchronize_rcu();
707 /* Make sure whatever was queued gets run */
708 vhost_worker_flush(old_worker);
709 mutex_unlock(&old_worker->mutex);
710}
711
712 /* Caller must have device mutex */
713static int vhost_vq_attach_worker(struct vhost_virtqueue *vq,
714 struct vhost_vring_worker *info)
715{
716 unsigned long index = info->worker_id;
717 struct vhost_dev *dev = vq->dev;
718 struct vhost_worker *worker;
719
720 if (!dev->use_worker)
721 return -EINVAL;
722
723 worker = xa_find(&dev->worker_xa, &index, UINT_MAX, XA_PRESENT);
724 if (!worker || worker->id != info->worker_id)
725 return -ENODEV;
726
727 __vhost_vq_attach_worker(vq, worker);
728 return 0;
729}
730
731/* Caller must have device mutex */
732static int vhost_new_worker(struct vhost_dev *dev,
733 struct vhost_worker_state *info)
734{
735 struct vhost_worker *worker;
736
737 worker = vhost_worker_create(dev);
738 if (!worker)
739 return -ENOMEM;
740
741 info->worker_id = worker->id;
742 return 0;
743}
744
745/* Caller must have device mutex */
746static int vhost_free_worker(struct vhost_dev *dev,
747 struct vhost_worker_state *info)
748{
749 unsigned long index = info->worker_id;
750 struct vhost_worker *worker;
751
752 worker = xa_find(&dev->worker_xa, &index, UINT_MAX, XA_PRESENT);
753 if (!worker || worker->id != info->worker_id)
754 return -ENODEV;
755
756 mutex_lock(&worker->mutex);
757 if (worker->attachment_cnt) {
758 mutex_unlock(&worker->mutex);
759 return -EBUSY;
760 }
761 mutex_unlock(&worker->mutex);
762
763 vhost_worker_destroy(dev, worker);
764 return 0;
765}
766
767static int vhost_get_vq_from_user(struct vhost_dev *dev, void __user *argp,
768 struct vhost_virtqueue **vq, u32 *id)
769{
770 u32 __user *idxp = argp;
771 u32 idx;
772 long r;
773
774 r = get_user(idx, idxp);
775 if (r < 0)
776 return r;
777
778 if (idx >= dev->nvqs)
779 return -ENOBUFS;
780
781 idx = array_index_nospec(idx, dev->nvqs);
782
783 *vq = dev->vqs[idx];
784 *id = idx;
785 return 0;
786}
787
788/* Caller must have device mutex */
789long vhost_worker_ioctl(struct vhost_dev *dev, unsigned int ioctl,
790 void __user *argp)
791{
792 struct vhost_vring_worker ring_worker;
793 struct vhost_worker_state state;
794 struct vhost_worker *worker;
795 struct vhost_virtqueue *vq;
796 long ret;
797 u32 idx;
798
799 if (!dev->use_worker)
800 return -EINVAL;
801
802 if (!vhost_dev_has_owner(dev))
803 return -EINVAL;
804
805 ret = vhost_dev_check_owner(dev);
806 if (ret)
807 return ret;
808
809 switch (ioctl) {
810 /* dev worker ioctls */
811 case VHOST_NEW_WORKER:
812 ret = vhost_new_worker(dev, &state);
813 if (!ret && copy_to_user(argp, &state, sizeof(state)))
814 ret = -EFAULT;
815 return ret;
816 case VHOST_FREE_WORKER:
817 if (copy_from_user(&state, argp, sizeof(state)))
818 return -EFAULT;
819 return vhost_free_worker(dev, &state);
820 /* vring worker ioctls */
821 case VHOST_ATTACH_VRING_WORKER:
822 case VHOST_GET_VRING_WORKER:
823 break;
824 default:
825 return -ENOIOCTLCMD;
826 }
827
828 ret = vhost_get_vq_from_user(dev, argp, &vq, &idx);
829 if (ret)
830 return ret;
831
832 switch (ioctl) {
833 case VHOST_ATTACH_VRING_WORKER:
834 if (copy_from_user(&ring_worker, argp, sizeof(ring_worker))) {
835 ret = -EFAULT;
836 break;
837 }
838
839 ret = vhost_vq_attach_worker(vq, &ring_worker);
840 break;
841 case VHOST_GET_VRING_WORKER:
842 worker = rcu_dereference_check(vq->worker,
843 lockdep_is_held(&dev->mutex));
844 if (!worker) {
845 ret = -EINVAL;
846 break;
847 }
848
849 ring_worker.index = idx;
850 ring_worker.worker_id = worker->id;
851
852 if (copy_to_user(argp, &ring_worker, sizeof(ring_worker)))
853 ret = -EFAULT;
854 break;
855 default:
856 ret = -ENOIOCTLCMD;
857 break;
858 }
859
860 return ret;
861}
862EXPORT_SYMBOL_GPL(vhost_worker_ioctl);
863
864/* Caller should have device mutex */
865long vhost_dev_set_owner(struct vhost_dev *dev)
866{
867 struct vhost_worker *worker;
868 int err, i;
869
870 /* Is there an owner already? */
871 if (vhost_dev_has_owner(dev)) {
872 err = -EBUSY;
873 goto err_mm;
874 }
875
876 vhost_attach_mm(dev);
877
878 err = vhost_dev_alloc_iovecs(dev);
879 if (err)
880 goto err_iovecs;
881
882 if (dev->use_worker) {
883 /*
884 * This should be done last, because vsock can queue work
885 * before VHOST_SET_OWNER so it simplifies the failure path
886 * below since we don't have to worry about vsock queueing
887 * while we free the worker.
888 */
889 worker = vhost_worker_create(dev);
890 if (!worker) {
891 err = -ENOMEM;
892 goto err_worker;
893 }
894
895 for (i = 0; i < dev->nvqs; i++)
896 __vhost_vq_attach_worker(dev->vqs[i], worker);
897 }
898
899 return 0;
900
901err_worker:
902 vhost_dev_free_iovecs(dev);
903err_iovecs:
904 vhost_detach_mm(dev);
905err_mm:
906 return err;
907}
908EXPORT_SYMBOL_GPL(vhost_dev_set_owner);
909
910static struct vhost_iotlb *iotlb_alloc(void)
911{
912 return vhost_iotlb_alloc(max_iotlb_entries,
913 VHOST_IOTLB_FLAG_RETIRE);
914}
915
916struct vhost_iotlb *vhost_dev_reset_owner_prepare(void)
917{
918 return iotlb_alloc();
919}
920EXPORT_SYMBOL_GPL(vhost_dev_reset_owner_prepare);
921
922/* Caller should have device mutex */
923void vhost_dev_reset_owner(struct vhost_dev *dev, struct vhost_iotlb *umem)
924{
925 int i;
926
927 vhost_dev_cleanup(dev);
928
929 dev->umem = umem;
930 /* We don't need VQ locks below since vhost_dev_cleanup makes sure
931 * VQs aren't running.
932 */
933 for (i = 0; i < dev->nvqs; ++i)
934 dev->vqs[i]->umem = umem;
935}
936EXPORT_SYMBOL_GPL(vhost_dev_reset_owner);
937
938void vhost_dev_stop(struct vhost_dev *dev)
939{
940 int i;
941
942 for (i = 0; i < dev->nvqs; ++i) {
943 if (dev->vqs[i]->kick && dev->vqs[i]->handle_kick)
944 vhost_poll_stop(&dev->vqs[i]->poll);
945 }
946
947 vhost_dev_flush(dev);
948}
949EXPORT_SYMBOL_GPL(vhost_dev_stop);
950
951void vhost_clear_msg(struct vhost_dev *dev)
952{
953 struct vhost_msg_node *node, *n;
954
955 spin_lock(&dev->iotlb_lock);
956
957 list_for_each_entry_safe(node, n, &dev->read_list, node) {
958 list_del(&node->node);
959 kfree(node);
960 }
961
962 list_for_each_entry_safe(node, n, &dev->pending_list, node) {
963 list_del(&node->node);
964 kfree(node);
965 }
966
967 spin_unlock(&dev->iotlb_lock);
968}
969EXPORT_SYMBOL_GPL(vhost_clear_msg);
970
971void vhost_dev_cleanup(struct vhost_dev *dev)
972{
973 int i;
974
975 for (i = 0; i < dev->nvqs; ++i) {
976 if (dev->vqs[i]->error_ctx)
977 eventfd_ctx_put(dev->vqs[i]->error_ctx);
978 if (dev->vqs[i]->kick)
979 fput(dev->vqs[i]->kick);
980 if (dev->vqs[i]->call_ctx.ctx)
981 eventfd_ctx_put(dev->vqs[i]->call_ctx.ctx);
982 vhost_vq_reset(dev, dev->vqs[i]);
983 }
984 vhost_dev_free_iovecs(dev);
985 if (dev->log_ctx)
986 eventfd_ctx_put(dev->log_ctx);
987 dev->log_ctx = NULL;
988 /* No one will access memory at this point */
989 vhost_iotlb_free(dev->umem);
990 dev->umem = NULL;
991 vhost_iotlb_free(dev->iotlb);
992 dev->iotlb = NULL;
993 vhost_clear_msg(dev);
994 wake_up_interruptible_poll(&dev->wait, EPOLLIN | EPOLLRDNORM);
995 vhost_workers_free(dev);
996 vhost_detach_mm(dev);
997}
998EXPORT_SYMBOL_GPL(vhost_dev_cleanup);
999
1000static bool log_access_ok(void __user *log_base, u64 addr, unsigned long sz)
1001{
1002 u64 a = addr / VHOST_PAGE_SIZE / 8;
1003
1004 /* Make sure 64 bit math will not overflow. */
1005 if (a > ULONG_MAX - (unsigned long)log_base ||
1006 a + (unsigned long)log_base > ULONG_MAX)
1007 return false;
1008
1009 return access_ok(log_base + a,
1010 (sz + VHOST_PAGE_SIZE * 8 - 1) / VHOST_PAGE_SIZE / 8);
1011}
1012
1013/* Make sure 64 bit math will not overflow. */
1014static bool vhost_overflow(u64 uaddr, u64 size)
1015{
1016 if (uaddr > ULONG_MAX || size > ULONG_MAX)
1017 return true;
1018
1019 if (!size)
1020 return false;
1021
1022 return uaddr > ULONG_MAX - size + 1;
1023}
1024
1025/* Caller should have vq mutex and device mutex. */
1026static bool vq_memory_access_ok(void __user *log_base, struct vhost_iotlb *umem,
1027 int log_all)
1028{
1029 struct vhost_iotlb_map *map;
1030
1031 if (!umem)
1032 return false;
1033
1034 list_for_each_entry(map, &umem->list, link) {
1035 unsigned long a = map->addr;
1036
1037 if (vhost_overflow(map->addr, map->size))
1038 return false;
1039
1040
1041 if (!access_ok((void __user *)a, map->size))
1042 return false;
1043 else if (log_all && !log_access_ok(log_base,
1044 map->start,
1045 map->size))
1046 return false;
1047 }
1048 return true;
1049}
1050
1051static inline void __user *vhost_vq_meta_fetch(struct vhost_virtqueue *vq,
1052 u64 addr, unsigned int size,
1053 int type)
1054{
1055 const struct vhost_iotlb_map *map = vq->meta_iotlb[type];
1056
1057 if (!map)
1058 return NULL;
1059
1060 return (void __user *)(uintptr_t)(map->addr + addr - map->start);
1061}
1062
1063/* Can we switch to this memory table? */
1064/* Caller should have device mutex but not vq mutex */
1065static bool memory_access_ok(struct vhost_dev *d, struct vhost_iotlb *umem,
1066 int log_all)
1067{
1068 int i;
1069
1070 for (i = 0; i < d->nvqs; ++i) {
1071 bool ok;
1072 bool log;
1073
1074 mutex_lock(&d->vqs[i]->mutex);
1075 log = log_all || vhost_has_feature(d->vqs[i], VHOST_F_LOG_ALL);
1076 /* If ring is inactive, will check when it's enabled. */
1077 if (d->vqs[i]->private_data)
1078 ok = vq_memory_access_ok(d->vqs[i]->log_base,
1079 umem, log);
1080 else
1081 ok = true;
1082 mutex_unlock(&d->vqs[i]->mutex);
1083 if (!ok)
1084 return false;
1085 }
1086 return true;
1087}
1088
1089static int translate_desc(struct vhost_virtqueue *vq, u64 addr, u32 len,
1090 struct iovec iov[], int iov_size, int access);
1091
1092static int vhost_copy_to_user(struct vhost_virtqueue *vq, void __user *to,
1093 const void *from, unsigned size)
1094{
1095 int ret;
1096
1097 if (!vq->iotlb)
1098 return __copy_to_user(to, from, size);
1099 else {
1100 /* This function should be called after iotlb
1101 * prefetch, which means we're sure that all vq
1102 * could be access through iotlb. So -EAGAIN should
1103 * not happen in this case.
1104 */
1105 struct iov_iter t;
1106 void __user *uaddr = vhost_vq_meta_fetch(vq,
1107 (u64)(uintptr_t)to, size,
1108 VHOST_ADDR_USED);
1109
1110 if (uaddr)
1111 return __copy_to_user(uaddr, from, size);
1112
1113 ret = translate_desc(vq, (u64)(uintptr_t)to, size, vq->iotlb_iov,
1114 ARRAY_SIZE(vq->iotlb_iov),
1115 VHOST_ACCESS_WO);
1116 if (ret < 0)
1117 goto out;
1118 iov_iter_init(&t, ITER_DEST, vq->iotlb_iov, ret, size);
1119 ret = copy_to_iter(from, size, &t);
1120 if (ret == size)
1121 ret = 0;
1122 }
1123out:
1124 return ret;
1125}
1126
1127static int vhost_copy_from_user(struct vhost_virtqueue *vq, void *to,
1128 void __user *from, unsigned size)
1129{
1130 int ret;
1131
1132 if (!vq->iotlb)
1133 return __copy_from_user(to, from, size);
1134 else {
1135 /* This function should be called after iotlb
1136 * prefetch, which means we're sure that vq
1137 * could be access through iotlb. So -EAGAIN should
1138 * not happen in this case.
1139 */
1140 void __user *uaddr = vhost_vq_meta_fetch(vq,
1141 (u64)(uintptr_t)from, size,
1142 VHOST_ADDR_DESC);
1143 struct iov_iter f;
1144
1145 if (uaddr)
1146 return __copy_from_user(to, uaddr, size);
1147
1148 ret = translate_desc(vq, (u64)(uintptr_t)from, size, vq->iotlb_iov,
1149 ARRAY_SIZE(vq->iotlb_iov),
1150 VHOST_ACCESS_RO);
1151 if (ret < 0) {
1152 vq_err(vq, "IOTLB translation failure: uaddr "
1153 "%p size 0x%llx\n", from,
1154 (unsigned long long) size);
1155 goto out;
1156 }
1157 iov_iter_init(&f, ITER_SOURCE, vq->iotlb_iov, ret, size);
1158 ret = copy_from_iter(to, size, &f);
1159 if (ret == size)
1160 ret = 0;
1161 }
1162
1163out:
1164 return ret;
1165}
1166
1167static void __user *__vhost_get_user_slow(struct vhost_virtqueue *vq,
1168 void __user *addr, unsigned int size,
1169 int type)
1170{
1171 int ret;
1172
1173 ret = translate_desc(vq, (u64)(uintptr_t)addr, size, vq->iotlb_iov,
1174 ARRAY_SIZE(vq->iotlb_iov),
1175 VHOST_ACCESS_RO);
1176 if (ret < 0) {
1177 vq_err(vq, "IOTLB translation failure: uaddr "
1178 "%p size 0x%llx\n", addr,
1179 (unsigned long long) size);
1180 return NULL;
1181 }
1182
1183 if (ret != 1 || vq->iotlb_iov[0].iov_len != size) {
1184 vq_err(vq, "Non atomic userspace memory access: uaddr "
1185 "%p size 0x%llx\n", addr,
1186 (unsigned long long) size);
1187 return NULL;
1188 }
1189
1190 return vq->iotlb_iov[0].iov_base;
1191}
1192
1193/* This function should be called after iotlb
1194 * prefetch, which means we're sure that vq
1195 * could be access through iotlb. So -EAGAIN should
1196 * not happen in this case.
1197 */
1198static inline void __user *__vhost_get_user(struct vhost_virtqueue *vq,
1199 void __user *addr, unsigned int size,
1200 int type)
1201{
1202 void __user *uaddr = vhost_vq_meta_fetch(vq,
1203 (u64)(uintptr_t)addr, size, type);
1204 if (uaddr)
1205 return uaddr;
1206
1207 return __vhost_get_user_slow(vq, addr, size, type);
1208}
1209
1210#define vhost_put_user(vq, x, ptr) \
1211({ \
1212 int ret; \
1213 if (!vq->iotlb) { \
1214 ret = __put_user(x, ptr); \
1215 } else { \
1216 __typeof__(ptr) to = \
1217 (__typeof__(ptr)) __vhost_get_user(vq, ptr, \
1218 sizeof(*ptr), VHOST_ADDR_USED); \
1219 if (to != NULL) \
1220 ret = __put_user(x, to); \
1221 else \
1222 ret = -EFAULT; \
1223 } \
1224 ret; \
1225})
1226
1227static inline int vhost_put_avail_event(struct vhost_virtqueue *vq)
1228{
1229 return vhost_put_user(vq, cpu_to_vhost16(vq, vq->avail_idx),
1230 vhost_avail_event(vq));
1231}
1232
1233static inline int vhost_put_used(struct vhost_virtqueue *vq,
1234 struct vring_used_elem *head, int idx,
1235 int count)
1236{
1237 return vhost_copy_to_user(vq, vq->used->ring + idx, head,
1238 count * sizeof(*head));
1239}
1240
1241static inline int vhost_put_used_flags(struct vhost_virtqueue *vq)
1242
1243{
1244 return vhost_put_user(vq, cpu_to_vhost16(vq, vq->used_flags),
1245 &vq->used->flags);
1246}
1247
1248static inline int vhost_put_used_idx(struct vhost_virtqueue *vq)
1249
1250{
1251 return vhost_put_user(vq, cpu_to_vhost16(vq, vq->last_used_idx),
1252 &vq->used->idx);
1253}
1254
1255#define vhost_get_user(vq, x, ptr, type) \
1256({ \
1257 int ret; \
1258 if (!vq->iotlb) { \
1259 ret = __get_user(x, ptr); \
1260 } else { \
1261 __typeof__(ptr) from = \
1262 (__typeof__(ptr)) __vhost_get_user(vq, ptr, \
1263 sizeof(*ptr), \
1264 type); \
1265 if (from != NULL) \
1266 ret = __get_user(x, from); \
1267 else \
1268 ret = -EFAULT; \
1269 } \
1270 ret; \
1271})
1272
1273#define vhost_get_avail(vq, x, ptr) \
1274 vhost_get_user(vq, x, ptr, VHOST_ADDR_AVAIL)
1275
1276#define vhost_get_used(vq, x, ptr) \
1277 vhost_get_user(vq, x, ptr, VHOST_ADDR_USED)
1278
1279static void vhost_dev_lock_vqs(struct vhost_dev *d)
1280{
1281 int i = 0;
1282 for (i = 0; i < d->nvqs; ++i)
1283 mutex_lock_nested(&d->vqs[i]->mutex, i);
1284}
1285
1286static void vhost_dev_unlock_vqs(struct vhost_dev *d)
1287{
1288 int i = 0;
1289 for (i = 0; i < d->nvqs; ++i)
1290 mutex_unlock(&d->vqs[i]->mutex);
1291}
1292
1293static inline int vhost_get_avail_idx(struct vhost_virtqueue *vq,
1294 __virtio16 *idx)
1295{
1296 return vhost_get_avail(vq, *idx, &vq->avail->idx);
1297}
1298
1299static inline int vhost_get_avail_head(struct vhost_virtqueue *vq,
1300 __virtio16 *head, int idx)
1301{
1302 return vhost_get_avail(vq, *head,
1303 &vq->avail->ring[idx & (vq->num - 1)]);
1304}
1305
1306static inline int vhost_get_avail_flags(struct vhost_virtqueue *vq,
1307 __virtio16 *flags)
1308{
1309 return vhost_get_avail(vq, *flags, &vq->avail->flags);
1310}
1311
1312static inline int vhost_get_used_event(struct vhost_virtqueue *vq,
1313 __virtio16 *event)
1314{
1315 return vhost_get_avail(vq, *event, vhost_used_event(vq));
1316}
1317
1318static inline int vhost_get_used_idx(struct vhost_virtqueue *vq,
1319 __virtio16 *idx)
1320{
1321 return vhost_get_used(vq, *idx, &vq->used->idx);
1322}
1323
1324static inline int vhost_get_desc(struct vhost_virtqueue *vq,
1325 struct vring_desc *desc, int idx)
1326{
1327 return vhost_copy_from_user(vq, desc, vq->desc + idx, sizeof(*desc));
1328}
1329
1330static void vhost_iotlb_notify_vq(struct vhost_dev *d,
1331 struct vhost_iotlb_msg *msg)
1332{
1333 struct vhost_msg_node *node, *n;
1334
1335 spin_lock(&d->iotlb_lock);
1336
1337 list_for_each_entry_safe(node, n, &d->pending_list, node) {
1338 struct vhost_iotlb_msg *vq_msg = &node->msg.iotlb;
1339 if (msg->iova <= vq_msg->iova &&
1340 msg->iova + msg->size - 1 >= vq_msg->iova &&
1341 vq_msg->type == VHOST_IOTLB_MISS) {
1342 vhost_poll_queue(&node->vq->poll);
1343 list_del(&node->node);
1344 kfree(node);
1345 }
1346 }
1347
1348 spin_unlock(&d->iotlb_lock);
1349}
1350
1351static bool umem_access_ok(u64 uaddr, u64 size, int access)
1352{
1353 unsigned long a = uaddr;
1354
1355 /* Make sure 64 bit math will not overflow. */
1356 if (vhost_overflow(uaddr, size))
1357 return false;
1358
1359 if ((access & VHOST_ACCESS_RO) &&
1360 !access_ok((void __user *)a, size))
1361 return false;
1362 if ((access & VHOST_ACCESS_WO) &&
1363 !access_ok((void __user *)a, size))
1364 return false;
1365 return true;
1366}
1367
1368static int vhost_process_iotlb_msg(struct vhost_dev *dev, u32 asid,
1369 struct vhost_iotlb_msg *msg)
1370{
1371 int ret = 0;
1372
1373 if (asid != 0)
1374 return -EINVAL;
1375
1376 mutex_lock(&dev->mutex);
1377 vhost_dev_lock_vqs(dev);
1378 switch (msg->type) {
1379 case VHOST_IOTLB_UPDATE:
1380 if (!dev->iotlb) {
1381 ret = -EFAULT;
1382 break;
1383 }
1384 if (!umem_access_ok(msg->uaddr, msg->size, msg->perm)) {
1385 ret = -EFAULT;
1386 break;
1387 }
1388 vhost_vq_meta_reset(dev);
1389 if (vhost_iotlb_add_range(dev->iotlb, msg->iova,
1390 msg->iova + msg->size - 1,
1391 msg->uaddr, msg->perm)) {
1392 ret = -ENOMEM;
1393 break;
1394 }
1395 vhost_iotlb_notify_vq(dev, msg);
1396 break;
1397 case VHOST_IOTLB_INVALIDATE:
1398 if (!dev->iotlb) {
1399 ret = -EFAULT;
1400 break;
1401 }
1402 vhost_vq_meta_reset(dev);
1403 vhost_iotlb_del_range(dev->iotlb, msg->iova,
1404 msg->iova + msg->size - 1);
1405 break;
1406 default:
1407 ret = -EINVAL;
1408 break;
1409 }
1410
1411 vhost_dev_unlock_vqs(dev);
1412 mutex_unlock(&dev->mutex);
1413
1414 return ret;
1415}
1416ssize_t vhost_chr_write_iter(struct vhost_dev *dev,
1417 struct iov_iter *from)
1418{
1419 struct vhost_iotlb_msg msg;
1420 size_t offset;
1421 int type, ret;
1422 u32 asid = 0;
1423
1424 ret = copy_from_iter(&type, sizeof(type), from);
1425 if (ret != sizeof(type)) {
1426 ret = -EINVAL;
1427 goto done;
1428 }
1429
1430 switch (type) {
1431 case VHOST_IOTLB_MSG:
1432 /* There maybe a hole after type for V1 message type,
1433 * so skip it here.
1434 */
1435 offset = offsetof(struct vhost_msg, iotlb) - sizeof(int);
1436 break;
1437 case VHOST_IOTLB_MSG_V2:
1438 if (vhost_backend_has_feature(dev->vqs[0],
1439 VHOST_BACKEND_F_IOTLB_ASID)) {
1440 ret = copy_from_iter(&asid, sizeof(asid), from);
1441 if (ret != sizeof(asid)) {
1442 ret = -EINVAL;
1443 goto done;
1444 }
1445 offset = 0;
1446 } else
1447 offset = sizeof(__u32);
1448 break;
1449 default:
1450 ret = -EINVAL;
1451 goto done;
1452 }
1453
1454 iov_iter_advance(from, offset);
1455 ret = copy_from_iter(&msg, sizeof(msg), from);
1456 if (ret != sizeof(msg)) {
1457 ret = -EINVAL;
1458 goto done;
1459 }
1460
1461 if (msg.type == VHOST_IOTLB_UPDATE && msg.size == 0) {
1462 ret = -EINVAL;
1463 goto done;
1464 }
1465
1466 if (dev->msg_handler)
1467 ret = dev->msg_handler(dev, asid, &msg);
1468 else
1469 ret = vhost_process_iotlb_msg(dev, asid, &msg);
1470 if (ret) {
1471 ret = -EFAULT;
1472 goto done;
1473 }
1474
1475 ret = (type == VHOST_IOTLB_MSG) ? sizeof(struct vhost_msg) :
1476 sizeof(struct vhost_msg_v2);
1477done:
1478 return ret;
1479}
1480EXPORT_SYMBOL(vhost_chr_write_iter);
1481
1482__poll_t vhost_chr_poll(struct file *file, struct vhost_dev *dev,
1483 poll_table *wait)
1484{
1485 __poll_t mask = 0;
1486
1487 poll_wait(file, &dev->wait, wait);
1488
1489 if (!list_empty(&dev->read_list))
1490 mask |= EPOLLIN | EPOLLRDNORM;
1491
1492 return mask;
1493}
1494EXPORT_SYMBOL(vhost_chr_poll);
1495
1496ssize_t vhost_chr_read_iter(struct vhost_dev *dev, struct iov_iter *to,
1497 int noblock)
1498{
1499 DEFINE_WAIT(wait);
1500 struct vhost_msg_node *node;
1501 ssize_t ret = 0;
1502 unsigned size = sizeof(struct vhost_msg);
1503
1504 if (iov_iter_count(to) < size)
1505 return 0;
1506
1507 while (1) {
1508 if (!noblock)
1509 prepare_to_wait(&dev->wait, &wait,
1510 TASK_INTERRUPTIBLE);
1511
1512 node = vhost_dequeue_msg(dev, &dev->read_list);
1513 if (node)
1514 break;
1515 if (noblock) {
1516 ret = -EAGAIN;
1517 break;
1518 }
1519 if (signal_pending(current)) {
1520 ret = -ERESTARTSYS;
1521 break;
1522 }
1523 if (!dev->iotlb) {
1524 ret = -EBADFD;
1525 break;
1526 }
1527
1528 schedule();
1529 }
1530
1531 if (!noblock)
1532 finish_wait(&dev->wait, &wait);
1533
1534 if (node) {
1535 struct vhost_iotlb_msg *msg;
1536 void *start = &node->msg;
1537
1538 switch (node->msg.type) {
1539 case VHOST_IOTLB_MSG:
1540 size = sizeof(node->msg);
1541 msg = &node->msg.iotlb;
1542 break;
1543 case VHOST_IOTLB_MSG_V2:
1544 size = sizeof(node->msg_v2);
1545 msg = &node->msg_v2.iotlb;
1546 break;
1547 default:
1548 BUG();
1549 break;
1550 }
1551
1552 ret = copy_to_iter(start, size, to);
1553 if (ret != size || msg->type != VHOST_IOTLB_MISS) {
1554 kfree(node);
1555 return ret;
1556 }
1557 vhost_enqueue_msg(dev, &dev->pending_list, node);
1558 }
1559
1560 return ret;
1561}
1562EXPORT_SYMBOL_GPL(vhost_chr_read_iter);
1563
1564static int vhost_iotlb_miss(struct vhost_virtqueue *vq, u64 iova, int access)
1565{
1566 struct vhost_dev *dev = vq->dev;
1567 struct vhost_msg_node *node;
1568 struct vhost_iotlb_msg *msg;
1569 bool v2 = vhost_backend_has_feature(vq, VHOST_BACKEND_F_IOTLB_MSG_V2);
1570
1571 node = vhost_new_msg(vq, v2 ? VHOST_IOTLB_MSG_V2 : VHOST_IOTLB_MSG);
1572 if (!node)
1573 return -ENOMEM;
1574
1575 if (v2) {
1576 node->msg_v2.type = VHOST_IOTLB_MSG_V2;
1577 msg = &node->msg_v2.iotlb;
1578 } else {
1579 msg = &node->msg.iotlb;
1580 }
1581
1582 msg->type = VHOST_IOTLB_MISS;
1583 msg->iova = iova;
1584 msg->perm = access;
1585
1586 vhost_enqueue_msg(dev, &dev->read_list, node);
1587
1588 return 0;
1589}
1590
1591static bool vq_access_ok(struct vhost_virtqueue *vq, unsigned int num,
1592 vring_desc_t __user *desc,
1593 vring_avail_t __user *avail,
1594 vring_used_t __user *used)
1595
1596{
1597 /* If an IOTLB device is present, the vring addresses are
1598 * GIOVAs. Access validation occurs at prefetch time. */
1599 if (vq->iotlb)
1600 return true;
1601
1602 return access_ok(desc, vhost_get_desc_size(vq, num)) &&
1603 access_ok(avail, vhost_get_avail_size(vq, num)) &&
1604 access_ok(used, vhost_get_used_size(vq, num));
1605}
1606
1607static void vhost_vq_meta_update(struct vhost_virtqueue *vq,
1608 const struct vhost_iotlb_map *map,
1609 int type)
1610{
1611 int access = (type == VHOST_ADDR_USED) ?
1612 VHOST_ACCESS_WO : VHOST_ACCESS_RO;
1613
1614 if (likely(map->perm & access))
1615 vq->meta_iotlb[type] = map;
1616}
1617
1618static bool iotlb_access_ok(struct vhost_virtqueue *vq,
1619 int access, u64 addr, u64 len, int type)
1620{
1621 const struct vhost_iotlb_map *map;
1622 struct vhost_iotlb *umem = vq->iotlb;
1623 u64 s = 0, size, orig_addr = addr, last = addr + len - 1;
1624
1625 if (vhost_vq_meta_fetch(vq, addr, len, type))
1626 return true;
1627
1628 while (len > s) {
1629 map = vhost_iotlb_itree_first(umem, addr, last);
1630 if (map == NULL || map->start > addr) {
1631 vhost_iotlb_miss(vq, addr, access);
1632 return false;
1633 } else if (!(map->perm & access)) {
1634 /* Report the possible access violation by
1635 * request another translation from userspace.
1636 */
1637 return false;
1638 }
1639
1640 size = map->size - addr + map->start;
1641
1642 if (orig_addr == addr && size >= len)
1643 vhost_vq_meta_update(vq, map, type);
1644
1645 s += size;
1646 addr += size;
1647 }
1648
1649 return true;
1650}
1651
1652int vq_meta_prefetch(struct vhost_virtqueue *vq)
1653{
1654 unsigned int num = vq->num;
1655
1656 if (!vq->iotlb)
1657 return 1;
1658
1659 return iotlb_access_ok(vq, VHOST_MAP_RO, (u64)(uintptr_t)vq->desc,
1660 vhost_get_desc_size(vq, num), VHOST_ADDR_DESC) &&
1661 iotlb_access_ok(vq, VHOST_MAP_RO, (u64)(uintptr_t)vq->avail,
1662 vhost_get_avail_size(vq, num),
1663 VHOST_ADDR_AVAIL) &&
1664 iotlb_access_ok(vq, VHOST_MAP_WO, (u64)(uintptr_t)vq->used,
1665 vhost_get_used_size(vq, num), VHOST_ADDR_USED);
1666}
1667EXPORT_SYMBOL_GPL(vq_meta_prefetch);
1668
1669/* Can we log writes? */
1670/* Caller should have device mutex but not vq mutex */
1671bool vhost_log_access_ok(struct vhost_dev *dev)
1672{
1673 return memory_access_ok(dev, dev->umem, 1);
1674}
1675EXPORT_SYMBOL_GPL(vhost_log_access_ok);
1676
1677static bool vq_log_used_access_ok(struct vhost_virtqueue *vq,
1678 void __user *log_base,
1679 bool log_used,
1680 u64 log_addr)
1681{
1682 /* If an IOTLB device is present, log_addr is a GIOVA that
1683 * will never be logged by log_used(). */
1684 if (vq->iotlb)
1685 return true;
1686
1687 return !log_used || log_access_ok(log_base, log_addr,
1688 vhost_get_used_size(vq, vq->num));
1689}
1690
1691/* Verify access for write logging. */
1692/* Caller should have vq mutex and device mutex */
1693static bool vq_log_access_ok(struct vhost_virtqueue *vq,
1694 void __user *log_base)
1695{
1696 return vq_memory_access_ok(log_base, vq->umem,
1697 vhost_has_feature(vq, VHOST_F_LOG_ALL)) &&
1698 vq_log_used_access_ok(vq, log_base, vq->log_used, vq->log_addr);
1699}
1700
1701/* Can we start vq? */
1702/* Caller should have vq mutex and device mutex */
1703bool vhost_vq_access_ok(struct vhost_virtqueue *vq)
1704{
1705 if (!vq_log_access_ok(vq, vq->log_base))
1706 return false;
1707
1708 return vq_access_ok(vq, vq->num, vq->desc, vq->avail, vq->used);
1709}
1710EXPORT_SYMBOL_GPL(vhost_vq_access_ok);
1711
1712static long vhost_set_memory(struct vhost_dev *d, struct vhost_memory __user *m)
1713{
1714 struct vhost_memory mem, *newmem;
1715 struct vhost_memory_region *region;
1716 struct vhost_iotlb *newumem, *oldumem;
1717 unsigned long size = offsetof(struct vhost_memory, regions);
1718 int i;
1719
1720 if (copy_from_user(&mem, m, size))
1721 return -EFAULT;
1722 if (mem.padding)
1723 return -EOPNOTSUPP;
1724 if (mem.nregions > max_mem_regions)
1725 return -E2BIG;
1726 newmem = kvzalloc(struct_size(newmem, regions, mem.nregions),
1727 GFP_KERNEL);
1728 if (!newmem)
1729 return -ENOMEM;
1730
1731 memcpy(newmem, &mem, size);
1732 if (copy_from_user(newmem->regions, m->regions,
1733 flex_array_size(newmem, regions, mem.nregions))) {
1734 kvfree(newmem);
1735 return -EFAULT;
1736 }
1737
1738 newumem = iotlb_alloc();
1739 if (!newumem) {
1740 kvfree(newmem);
1741 return -ENOMEM;
1742 }
1743
1744 for (region = newmem->regions;
1745 region < newmem->regions + mem.nregions;
1746 region++) {
1747 if (vhost_iotlb_add_range(newumem,
1748 region->guest_phys_addr,
1749 region->guest_phys_addr +
1750 region->memory_size - 1,
1751 region->userspace_addr,
1752 VHOST_MAP_RW))
1753 goto err;
1754 }
1755
1756 if (!memory_access_ok(d, newumem, 0))
1757 goto err;
1758
1759 oldumem = d->umem;
1760 d->umem = newumem;
1761
1762 /* All memory accesses are done under some VQ mutex. */
1763 for (i = 0; i < d->nvqs; ++i) {
1764 mutex_lock(&d->vqs[i]->mutex);
1765 d->vqs[i]->umem = newumem;
1766 mutex_unlock(&d->vqs[i]->mutex);
1767 }
1768
1769 kvfree(newmem);
1770 vhost_iotlb_free(oldumem);
1771 return 0;
1772
1773err:
1774 vhost_iotlb_free(newumem);
1775 kvfree(newmem);
1776 return -EFAULT;
1777}
1778
1779static long vhost_vring_set_num(struct vhost_dev *d,
1780 struct vhost_virtqueue *vq,
1781 void __user *argp)
1782{
1783 struct vhost_vring_state s;
1784
1785 /* Resizing ring with an active backend?
1786 * You don't want to do that. */
1787 if (vq->private_data)
1788 return -EBUSY;
1789
1790 if (copy_from_user(&s, argp, sizeof s))
1791 return -EFAULT;
1792
1793 if (!s.num || s.num > 0xffff || (s.num & (s.num - 1)))
1794 return -EINVAL;
1795 vq->num = s.num;
1796
1797 return 0;
1798}
1799
1800static long vhost_vring_set_addr(struct vhost_dev *d,
1801 struct vhost_virtqueue *vq,
1802 void __user *argp)
1803{
1804 struct vhost_vring_addr a;
1805
1806 if (copy_from_user(&a, argp, sizeof a))
1807 return -EFAULT;
1808 if (a.flags & ~(0x1 << VHOST_VRING_F_LOG))
1809 return -EOPNOTSUPP;
1810
1811 /* For 32bit, verify that the top 32bits of the user
1812 data are set to zero. */
1813 if ((u64)(unsigned long)a.desc_user_addr != a.desc_user_addr ||
1814 (u64)(unsigned long)a.used_user_addr != a.used_user_addr ||
1815 (u64)(unsigned long)a.avail_user_addr != a.avail_user_addr)
1816 return -EFAULT;
1817
1818 /* Make sure it's safe to cast pointers to vring types. */
1819 BUILD_BUG_ON(__alignof__ *vq->avail > VRING_AVAIL_ALIGN_SIZE);
1820 BUILD_BUG_ON(__alignof__ *vq->used > VRING_USED_ALIGN_SIZE);
1821 if ((a.avail_user_addr & (VRING_AVAIL_ALIGN_SIZE - 1)) ||
1822 (a.used_user_addr & (VRING_USED_ALIGN_SIZE - 1)) ||
1823 (a.log_guest_addr & (VRING_USED_ALIGN_SIZE - 1)))
1824 return -EINVAL;
1825
1826 /* We only verify access here if backend is configured.
1827 * If it is not, we don't as size might not have been setup.
1828 * We will verify when backend is configured. */
1829 if (vq->private_data) {
1830 if (!vq_access_ok(vq, vq->num,
1831 (void __user *)(unsigned long)a.desc_user_addr,
1832 (void __user *)(unsigned long)a.avail_user_addr,
1833 (void __user *)(unsigned long)a.used_user_addr))
1834 return -EINVAL;
1835
1836 /* Also validate log access for used ring if enabled. */
1837 if (!vq_log_used_access_ok(vq, vq->log_base,
1838 a.flags & (0x1 << VHOST_VRING_F_LOG),
1839 a.log_guest_addr))
1840 return -EINVAL;
1841 }
1842
1843 vq->log_used = !!(a.flags & (0x1 << VHOST_VRING_F_LOG));
1844 vq->desc = (void __user *)(unsigned long)a.desc_user_addr;
1845 vq->avail = (void __user *)(unsigned long)a.avail_user_addr;
1846 vq->log_addr = a.log_guest_addr;
1847 vq->used = (void __user *)(unsigned long)a.used_user_addr;
1848
1849 return 0;
1850}
1851
1852static long vhost_vring_set_num_addr(struct vhost_dev *d,
1853 struct vhost_virtqueue *vq,
1854 unsigned int ioctl,
1855 void __user *argp)
1856{
1857 long r;
1858
1859 mutex_lock(&vq->mutex);
1860
1861 switch (ioctl) {
1862 case VHOST_SET_VRING_NUM:
1863 r = vhost_vring_set_num(d, vq, argp);
1864 break;
1865 case VHOST_SET_VRING_ADDR:
1866 r = vhost_vring_set_addr(d, vq, argp);
1867 break;
1868 default:
1869 BUG();
1870 }
1871
1872 mutex_unlock(&vq->mutex);
1873
1874 return r;
1875}
1876long vhost_vring_ioctl(struct vhost_dev *d, unsigned int ioctl, void __user *argp)
1877{
1878 struct file *eventfp, *filep = NULL;
1879 bool pollstart = false, pollstop = false;
1880 struct eventfd_ctx *ctx = NULL;
1881 struct vhost_virtqueue *vq;
1882 struct vhost_vring_state s;
1883 struct vhost_vring_file f;
1884 u32 idx;
1885 long r;
1886
1887 r = vhost_get_vq_from_user(d, argp, &vq, &idx);
1888 if (r < 0)
1889 return r;
1890
1891 if (ioctl == VHOST_SET_VRING_NUM ||
1892 ioctl == VHOST_SET_VRING_ADDR) {
1893 return vhost_vring_set_num_addr(d, vq, ioctl, argp);
1894 }
1895
1896 mutex_lock(&vq->mutex);
1897
1898 switch (ioctl) {
1899 case VHOST_SET_VRING_BASE:
1900 /* Moving base with an active backend?
1901 * You don't want to do that. */
1902 if (vq->private_data) {
1903 r = -EBUSY;
1904 break;
1905 }
1906 if (copy_from_user(&s, argp, sizeof s)) {
1907 r = -EFAULT;
1908 break;
1909 }
1910 if (vhost_has_feature(vq, VIRTIO_F_RING_PACKED)) {
1911 vq->last_avail_idx = s.num & 0xffff;
1912 vq->last_used_idx = (s.num >> 16) & 0xffff;
1913 } else {
1914 if (s.num > 0xffff) {
1915 r = -EINVAL;
1916 break;
1917 }
1918 vq->last_avail_idx = s.num;
1919 }
1920 /* Forget the cached index value. */
1921 vq->avail_idx = vq->last_avail_idx;
1922 break;
1923 case VHOST_GET_VRING_BASE:
1924 s.index = idx;
1925 if (vhost_has_feature(vq, VIRTIO_F_RING_PACKED))
1926 s.num = (u32)vq->last_avail_idx | ((u32)vq->last_used_idx << 16);
1927 else
1928 s.num = vq->last_avail_idx;
1929 if (copy_to_user(argp, &s, sizeof s))
1930 r = -EFAULT;
1931 break;
1932 case VHOST_SET_VRING_KICK:
1933 if (copy_from_user(&f, argp, sizeof f)) {
1934 r = -EFAULT;
1935 break;
1936 }
1937 eventfp = f.fd == VHOST_FILE_UNBIND ? NULL : eventfd_fget(f.fd);
1938 if (IS_ERR(eventfp)) {
1939 r = PTR_ERR(eventfp);
1940 break;
1941 }
1942 if (eventfp != vq->kick) {
1943 pollstop = (filep = vq->kick) != NULL;
1944 pollstart = (vq->kick = eventfp) != NULL;
1945 } else
1946 filep = eventfp;
1947 break;
1948 case VHOST_SET_VRING_CALL:
1949 if (copy_from_user(&f, argp, sizeof f)) {
1950 r = -EFAULT;
1951 break;
1952 }
1953 ctx = f.fd == VHOST_FILE_UNBIND ? NULL : eventfd_ctx_fdget(f.fd);
1954 if (IS_ERR(ctx)) {
1955 r = PTR_ERR(ctx);
1956 break;
1957 }
1958
1959 swap(ctx, vq->call_ctx.ctx);
1960 break;
1961 case VHOST_SET_VRING_ERR:
1962 if (copy_from_user(&f, argp, sizeof f)) {
1963 r = -EFAULT;
1964 break;
1965 }
1966 ctx = f.fd == VHOST_FILE_UNBIND ? NULL : eventfd_ctx_fdget(f.fd);
1967 if (IS_ERR(ctx)) {
1968 r = PTR_ERR(ctx);
1969 break;
1970 }
1971 swap(ctx, vq->error_ctx);
1972 break;
1973 case VHOST_SET_VRING_ENDIAN:
1974 r = vhost_set_vring_endian(vq, argp);
1975 break;
1976 case VHOST_GET_VRING_ENDIAN:
1977 r = vhost_get_vring_endian(vq, idx, argp);
1978 break;
1979 case VHOST_SET_VRING_BUSYLOOP_TIMEOUT:
1980 if (copy_from_user(&s, argp, sizeof(s))) {
1981 r = -EFAULT;
1982 break;
1983 }
1984 vq->busyloop_timeout = s.num;
1985 break;
1986 case VHOST_GET_VRING_BUSYLOOP_TIMEOUT:
1987 s.index = idx;
1988 s.num = vq->busyloop_timeout;
1989 if (copy_to_user(argp, &s, sizeof(s)))
1990 r = -EFAULT;
1991 break;
1992 default:
1993 r = -ENOIOCTLCMD;
1994 }
1995
1996 if (pollstop && vq->handle_kick)
1997 vhost_poll_stop(&vq->poll);
1998
1999 if (!IS_ERR_OR_NULL(ctx))
2000 eventfd_ctx_put(ctx);
2001 if (filep)
2002 fput(filep);
2003
2004 if (pollstart && vq->handle_kick)
2005 r = vhost_poll_start(&vq->poll, vq->kick);
2006
2007 mutex_unlock(&vq->mutex);
2008
2009 if (pollstop && vq->handle_kick)
2010 vhost_dev_flush(vq->poll.dev);
2011 return r;
2012}
2013EXPORT_SYMBOL_GPL(vhost_vring_ioctl);
2014
2015int vhost_init_device_iotlb(struct vhost_dev *d)
2016{
2017 struct vhost_iotlb *niotlb, *oiotlb;
2018 int i;
2019
2020 niotlb = iotlb_alloc();
2021 if (!niotlb)
2022 return -ENOMEM;
2023
2024 oiotlb = d->iotlb;
2025 d->iotlb = niotlb;
2026
2027 for (i = 0; i < d->nvqs; ++i) {
2028 struct vhost_virtqueue *vq = d->vqs[i];
2029
2030 mutex_lock(&vq->mutex);
2031 vq->iotlb = niotlb;
2032 __vhost_vq_meta_reset(vq);
2033 mutex_unlock(&vq->mutex);
2034 }
2035
2036 vhost_iotlb_free(oiotlb);
2037
2038 return 0;
2039}
2040EXPORT_SYMBOL_GPL(vhost_init_device_iotlb);
2041
2042/* Caller must have device mutex */
2043long vhost_dev_ioctl(struct vhost_dev *d, unsigned int ioctl, void __user *argp)
2044{
2045 struct eventfd_ctx *ctx;
2046 u64 p;
2047 long r;
2048 int i, fd;
2049
2050 /* If you are not the owner, you can become one */
2051 if (ioctl == VHOST_SET_OWNER) {
2052 r = vhost_dev_set_owner(d);
2053 goto done;
2054 }
2055
2056 /* You must be the owner to do anything else */
2057 r = vhost_dev_check_owner(d);
2058 if (r)
2059 goto done;
2060
2061 switch (ioctl) {
2062 case VHOST_SET_MEM_TABLE:
2063 r = vhost_set_memory(d, argp);
2064 break;
2065 case VHOST_SET_LOG_BASE:
2066 if (copy_from_user(&p, argp, sizeof p)) {
2067 r = -EFAULT;
2068 break;
2069 }
2070 if ((u64)(unsigned long)p != p) {
2071 r = -EFAULT;
2072 break;
2073 }
2074 for (i = 0; i < d->nvqs; ++i) {
2075 struct vhost_virtqueue *vq;
2076 void __user *base = (void __user *)(unsigned long)p;
2077 vq = d->vqs[i];
2078 mutex_lock(&vq->mutex);
2079 /* If ring is inactive, will check when it's enabled. */
2080 if (vq->private_data && !vq_log_access_ok(vq, base))
2081 r = -EFAULT;
2082 else
2083 vq->log_base = base;
2084 mutex_unlock(&vq->mutex);
2085 }
2086 break;
2087 case VHOST_SET_LOG_FD:
2088 r = get_user(fd, (int __user *)argp);
2089 if (r < 0)
2090 break;
2091 ctx = fd == VHOST_FILE_UNBIND ? NULL : eventfd_ctx_fdget(fd);
2092 if (IS_ERR(ctx)) {
2093 r = PTR_ERR(ctx);
2094 break;
2095 }
2096 swap(ctx, d->log_ctx);
2097 for (i = 0; i < d->nvqs; ++i) {
2098 mutex_lock(&d->vqs[i]->mutex);
2099 d->vqs[i]->log_ctx = d->log_ctx;
2100 mutex_unlock(&d->vqs[i]->mutex);
2101 }
2102 if (ctx)
2103 eventfd_ctx_put(ctx);
2104 break;
2105 default:
2106 r = -ENOIOCTLCMD;
2107 break;
2108 }
2109done:
2110 return r;
2111}
2112EXPORT_SYMBOL_GPL(vhost_dev_ioctl);
2113
2114/* TODO: This is really inefficient. We need something like get_user()
2115 * (instruction directly accesses the data, with an exception table entry
2116 * returning -EFAULT). See Documentation/arch/x86/exception-tables.rst.
2117 */
2118static int set_bit_to_user(int nr, void __user *addr)
2119{
2120 unsigned long log = (unsigned long)addr;
2121 struct page *page;
2122 void *base;
2123 int bit = nr + (log % PAGE_SIZE) * 8;
2124 int r;
2125
2126 r = pin_user_pages_fast(log, 1, FOLL_WRITE, &page);
2127 if (r < 0)
2128 return r;
2129 BUG_ON(r != 1);
2130 base = kmap_atomic(page);
2131 set_bit(bit, base);
2132 kunmap_atomic(base);
2133 unpin_user_pages_dirty_lock(&page, 1, true);
2134 return 0;
2135}
2136
2137static int log_write(void __user *log_base,
2138 u64 write_address, u64 write_length)
2139{
2140 u64 write_page = write_address / VHOST_PAGE_SIZE;
2141 int r;
2142
2143 if (!write_length)
2144 return 0;
2145 write_length += write_address % VHOST_PAGE_SIZE;
2146 for (;;) {
2147 u64 base = (u64)(unsigned long)log_base;
2148 u64 log = base + write_page / 8;
2149 int bit = write_page % 8;
2150 if ((u64)(unsigned long)log != log)
2151 return -EFAULT;
2152 r = set_bit_to_user(bit, (void __user *)(unsigned long)log);
2153 if (r < 0)
2154 return r;
2155 if (write_length <= VHOST_PAGE_SIZE)
2156 break;
2157 write_length -= VHOST_PAGE_SIZE;
2158 write_page += 1;
2159 }
2160 return r;
2161}
2162
2163static int log_write_hva(struct vhost_virtqueue *vq, u64 hva, u64 len)
2164{
2165 struct vhost_iotlb *umem = vq->umem;
2166 struct vhost_iotlb_map *u;
2167 u64 start, end, l, min;
2168 int r;
2169 bool hit = false;
2170
2171 while (len) {
2172 min = len;
2173 /* More than one GPAs can be mapped into a single HVA. So
2174 * iterate all possible umems here to be safe.
2175 */
2176 list_for_each_entry(u, &umem->list, link) {
2177 if (u->addr > hva - 1 + len ||
2178 u->addr - 1 + u->size < hva)
2179 continue;
2180 start = max(u->addr, hva);
2181 end = min(u->addr - 1 + u->size, hva - 1 + len);
2182 l = end - start + 1;
2183 r = log_write(vq->log_base,
2184 u->start + start - u->addr,
2185 l);
2186 if (r < 0)
2187 return r;
2188 hit = true;
2189 min = min(l, min);
2190 }
2191
2192 if (!hit)
2193 return -EFAULT;
2194
2195 len -= min;
2196 hva += min;
2197 }
2198
2199 return 0;
2200}
2201
2202static int log_used(struct vhost_virtqueue *vq, u64 used_offset, u64 len)
2203{
2204 struct iovec *iov = vq->log_iov;
2205 int i, ret;
2206
2207 if (!vq->iotlb)
2208 return log_write(vq->log_base, vq->log_addr + used_offset, len);
2209
2210 ret = translate_desc(vq, (uintptr_t)vq->used + used_offset,
2211 len, iov, 64, VHOST_ACCESS_WO);
2212 if (ret < 0)
2213 return ret;
2214
2215 for (i = 0; i < ret; i++) {
2216 ret = log_write_hva(vq, (uintptr_t)iov[i].iov_base,
2217 iov[i].iov_len);
2218 if (ret)
2219 return ret;
2220 }
2221
2222 return 0;
2223}
2224
2225int vhost_log_write(struct vhost_virtqueue *vq, struct vhost_log *log,
2226 unsigned int log_num, u64 len, struct iovec *iov, int count)
2227{
2228 int i, r;
2229
2230 /* Make sure data written is seen before log. */
2231 smp_wmb();
2232
2233 if (vq->iotlb) {
2234 for (i = 0; i < count; i++) {
2235 r = log_write_hva(vq, (uintptr_t)iov[i].iov_base,
2236 iov[i].iov_len);
2237 if (r < 0)
2238 return r;
2239 }
2240 return 0;
2241 }
2242
2243 for (i = 0; i < log_num; ++i) {
2244 u64 l = min(log[i].len, len);
2245 r = log_write(vq->log_base, log[i].addr, l);
2246 if (r < 0)
2247 return r;
2248 len -= l;
2249 if (!len) {
2250 if (vq->log_ctx)
2251 eventfd_signal(vq->log_ctx);
2252 return 0;
2253 }
2254 }
2255 /* Length written exceeds what we have stored. This is a bug. */
2256 BUG();
2257 return 0;
2258}
2259EXPORT_SYMBOL_GPL(vhost_log_write);
2260
2261static int vhost_update_used_flags(struct vhost_virtqueue *vq)
2262{
2263 void __user *used;
2264 if (vhost_put_used_flags(vq))
2265 return -EFAULT;
2266 if (unlikely(vq->log_used)) {
2267 /* Make sure the flag is seen before log. */
2268 smp_wmb();
2269 /* Log used flag write. */
2270 used = &vq->used->flags;
2271 log_used(vq, (used - (void __user *)vq->used),
2272 sizeof vq->used->flags);
2273 if (vq->log_ctx)
2274 eventfd_signal(vq->log_ctx);
2275 }
2276 return 0;
2277}
2278
2279static int vhost_update_avail_event(struct vhost_virtqueue *vq)
2280{
2281 if (vhost_put_avail_event(vq))
2282 return -EFAULT;
2283 if (unlikely(vq->log_used)) {
2284 void __user *used;
2285 /* Make sure the event is seen before log. */
2286 smp_wmb();
2287 /* Log avail event write */
2288 used = vhost_avail_event(vq);
2289 log_used(vq, (used - (void __user *)vq->used),
2290 sizeof *vhost_avail_event(vq));
2291 if (vq->log_ctx)
2292 eventfd_signal(vq->log_ctx);
2293 }
2294 return 0;
2295}
2296
2297int vhost_vq_init_access(struct vhost_virtqueue *vq)
2298{
2299 __virtio16 last_used_idx;
2300 int r;
2301 bool is_le = vq->is_le;
2302
2303 if (!vq->private_data)
2304 return 0;
2305
2306 vhost_init_is_le(vq);
2307
2308 r = vhost_update_used_flags(vq);
2309 if (r)
2310 goto err;
2311 vq->signalled_used_valid = false;
2312 if (!vq->iotlb &&
2313 !access_ok(&vq->used->idx, sizeof vq->used->idx)) {
2314 r = -EFAULT;
2315 goto err;
2316 }
2317 r = vhost_get_used_idx(vq, &last_used_idx);
2318 if (r) {
2319 vq_err(vq, "Can't access used idx at %p\n",
2320 &vq->used->idx);
2321 goto err;
2322 }
2323 vq->last_used_idx = vhost16_to_cpu(vq, last_used_idx);
2324 return 0;
2325
2326err:
2327 vq->is_le = is_le;
2328 return r;
2329}
2330EXPORT_SYMBOL_GPL(vhost_vq_init_access);
2331
2332static int translate_desc(struct vhost_virtqueue *vq, u64 addr, u32 len,
2333 struct iovec iov[], int iov_size, int access)
2334{
2335 const struct vhost_iotlb_map *map;
2336 struct vhost_dev *dev = vq->dev;
2337 struct vhost_iotlb *umem = dev->iotlb ? dev->iotlb : dev->umem;
2338 struct iovec *_iov;
2339 u64 s = 0, last = addr + len - 1;
2340 int ret = 0;
2341
2342 while ((u64)len > s) {
2343 u64 size;
2344 if (unlikely(ret >= iov_size)) {
2345 ret = -ENOBUFS;
2346 break;
2347 }
2348
2349 map = vhost_iotlb_itree_first(umem, addr, last);
2350 if (map == NULL || map->start > addr) {
2351 if (umem != dev->iotlb) {
2352 ret = -EFAULT;
2353 break;
2354 }
2355 ret = -EAGAIN;
2356 break;
2357 } else if (!(map->perm & access)) {
2358 ret = -EPERM;
2359 break;
2360 }
2361
2362 _iov = iov + ret;
2363 size = map->size - addr + map->start;
2364 _iov->iov_len = min((u64)len - s, size);
2365 _iov->iov_base = (void __user *)(unsigned long)
2366 (map->addr + addr - map->start);
2367 s += size;
2368 addr += size;
2369 ++ret;
2370 }
2371
2372 if (ret == -EAGAIN)
2373 vhost_iotlb_miss(vq, addr, access);
2374 return ret;
2375}
2376
2377/* Each buffer in the virtqueues is actually a chain of descriptors. This
2378 * function returns the next descriptor in the chain,
2379 * or -1U if we're at the end. */
2380static unsigned next_desc(struct vhost_virtqueue *vq, struct vring_desc *desc)
2381{
2382 unsigned int next;
2383
2384 /* If this descriptor says it doesn't chain, we're done. */
2385 if (!(desc->flags & cpu_to_vhost16(vq, VRING_DESC_F_NEXT)))
2386 return -1U;
2387
2388 /* Check they're not leading us off end of descriptors. */
2389 next = vhost16_to_cpu(vq, READ_ONCE(desc->next));
2390 return next;
2391}
2392
2393static int get_indirect(struct vhost_virtqueue *vq,
2394 struct iovec iov[], unsigned int iov_size,
2395 unsigned int *out_num, unsigned int *in_num,
2396 struct vhost_log *log, unsigned int *log_num,
2397 struct vring_desc *indirect)
2398{
2399 struct vring_desc desc;
2400 unsigned int i = 0, count, found = 0;
2401 u32 len = vhost32_to_cpu(vq, indirect->len);
2402 struct iov_iter from;
2403 int ret, access;
2404
2405 /* Sanity check */
2406 if (unlikely(len % sizeof desc)) {
2407 vq_err(vq, "Invalid length in indirect descriptor: "
2408 "len 0x%llx not multiple of 0x%zx\n",
2409 (unsigned long long)len,
2410 sizeof desc);
2411 return -EINVAL;
2412 }
2413
2414 ret = translate_desc(vq, vhost64_to_cpu(vq, indirect->addr), len, vq->indirect,
2415 UIO_MAXIOV, VHOST_ACCESS_RO);
2416 if (unlikely(ret < 0)) {
2417 if (ret != -EAGAIN)
2418 vq_err(vq, "Translation failure %d in indirect.\n", ret);
2419 return ret;
2420 }
2421 iov_iter_init(&from, ITER_SOURCE, vq->indirect, ret, len);
2422 count = len / sizeof desc;
2423 /* Buffers are chained via a 16 bit next field, so
2424 * we can have at most 2^16 of these. */
2425 if (unlikely(count > USHRT_MAX + 1)) {
2426 vq_err(vq, "Indirect buffer length too big: %d\n",
2427 indirect->len);
2428 return -E2BIG;
2429 }
2430
2431 do {
2432 unsigned iov_count = *in_num + *out_num;
2433 if (unlikely(++found > count)) {
2434 vq_err(vq, "Loop detected: last one at %u "
2435 "indirect size %u\n",
2436 i, count);
2437 return -EINVAL;
2438 }
2439 if (unlikely(!copy_from_iter_full(&desc, sizeof(desc), &from))) {
2440 vq_err(vq, "Failed indirect descriptor: idx %d, %zx\n",
2441 i, (size_t)vhost64_to_cpu(vq, indirect->addr) + i * sizeof desc);
2442 return -EINVAL;
2443 }
2444 if (unlikely(desc.flags & cpu_to_vhost16(vq, VRING_DESC_F_INDIRECT))) {
2445 vq_err(vq, "Nested indirect descriptor: idx %d, %zx\n",
2446 i, (size_t)vhost64_to_cpu(vq, indirect->addr) + i * sizeof desc);
2447 return -EINVAL;
2448 }
2449
2450 if (desc.flags & cpu_to_vhost16(vq, VRING_DESC_F_WRITE))
2451 access = VHOST_ACCESS_WO;
2452 else
2453 access = VHOST_ACCESS_RO;
2454
2455 ret = translate_desc(vq, vhost64_to_cpu(vq, desc.addr),
2456 vhost32_to_cpu(vq, desc.len), iov + iov_count,
2457 iov_size - iov_count, access);
2458 if (unlikely(ret < 0)) {
2459 if (ret != -EAGAIN)
2460 vq_err(vq, "Translation failure %d indirect idx %d\n",
2461 ret, i);
2462 return ret;
2463 }
2464 /* If this is an input descriptor, increment that count. */
2465 if (access == VHOST_ACCESS_WO) {
2466 *in_num += ret;
2467 if (unlikely(log && ret)) {
2468 log[*log_num].addr = vhost64_to_cpu(vq, desc.addr);
2469 log[*log_num].len = vhost32_to_cpu(vq, desc.len);
2470 ++*log_num;
2471 }
2472 } else {
2473 /* If it's an output descriptor, they're all supposed
2474 * to come before any input descriptors. */
2475 if (unlikely(*in_num)) {
2476 vq_err(vq, "Indirect descriptor "
2477 "has out after in: idx %d\n", i);
2478 return -EINVAL;
2479 }
2480 *out_num += ret;
2481 }
2482 } while ((i = next_desc(vq, &desc)) != -1);
2483 return 0;
2484}
2485
2486/* This looks in the virtqueue and for the first available buffer, and converts
2487 * it to an iovec for convenient access. Since descriptors consist of some
2488 * number of output then some number of input descriptors, it's actually two
2489 * iovecs, but we pack them into one and note how many of each there were.
2490 *
2491 * This function returns the descriptor number found, or vq->num (which is
2492 * never a valid descriptor number) if none was found. A negative code is
2493 * returned on error. */
2494int vhost_get_vq_desc(struct vhost_virtqueue *vq,
2495 struct iovec iov[], unsigned int iov_size,
2496 unsigned int *out_num, unsigned int *in_num,
2497 struct vhost_log *log, unsigned int *log_num)
2498{
2499 struct vring_desc desc;
2500 unsigned int i, head, found = 0;
2501 u16 last_avail_idx;
2502 __virtio16 avail_idx;
2503 __virtio16 ring_head;
2504 int ret, access;
2505
2506 /* Check it isn't doing very strange things with descriptor numbers. */
2507 last_avail_idx = vq->last_avail_idx;
2508
2509 if (vq->avail_idx == vq->last_avail_idx) {
2510 if (unlikely(vhost_get_avail_idx(vq, &avail_idx))) {
2511 vq_err(vq, "Failed to access avail idx at %p\n",
2512 &vq->avail->idx);
2513 return -EFAULT;
2514 }
2515 vq->avail_idx = vhost16_to_cpu(vq, avail_idx);
2516
2517 if (unlikely((u16)(vq->avail_idx - last_avail_idx) > vq->num)) {
2518 vq_err(vq, "Guest moved used index from %u to %u",
2519 last_avail_idx, vq->avail_idx);
2520 return -EFAULT;
2521 }
2522
2523 /* If there's nothing new since last we looked, return
2524 * invalid.
2525 */
2526 if (vq->avail_idx == last_avail_idx)
2527 return vq->num;
2528
2529 /* Only get avail ring entries after they have been
2530 * exposed by guest.
2531 */
2532 smp_rmb();
2533 }
2534
2535 /* Grab the next descriptor number they're advertising, and increment
2536 * the index we've seen. */
2537 if (unlikely(vhost_get_avail_head(vq, &ring_head, last_avail_idx))) {
2538 vq_err(vq, "Failed to read head: idx %d address %p\n",
2539 last_avail_idx,
2540 &vq->avail->ring[last_avail_idx % vq->num]);
2541 return -EFAULT;
2542 }
2543
2544 head = vhost16_to_cpu(vq, ring_head);
2545
2546 /* If their number is silly, that's an error. */
2547 if (unlikely(head >= vq->num)) {
2548 vq_err(vq, "Guest says index %u > %u is available",
2549 head, vq->num);
2550 return -EINVAL;
2551 }
2552
2553 /* When we start there are none of either input nor output. */
2554 *out_num = *in_num = 0;
2555 if (unlikely(log))
2556 *log_num = 0;
2557
2558 i = head;
2559 do {
2560 unsigned iov_count = *in_num + *out_num;
2561 if (unlikely(i >= vq->num)) {
2562 vq_err(vq, "Desc index is %u > %u, head = %u",
2563 i, vq->num, head);
2564 return -EINVAL;
2565 }
2566 if (unlikely(++found > vq->num)) {
2567 vq_err(vq, "Loop detected: last one at %u "
2568 "vq size %u head %u\n",
2569 i, vq->num, head);
2570 return -EINVAL;
2571 }
2572 ret = vhost_get_desc(vq, &desc, i);
2573 if (unlikely(ret)) {
2574 vq_err(vq, "Failed to get descriptor: idx %d addr %p\n",
2575 i, vq->desc + i);
2576 return -EFAULT;
2577 }
2578 if (desc.flags & cpu_to_vhost16(vq, VRING_DESC_F_INDIRECT)) {
2579 ret = get_indirect(vq, iov, iov_size,
2580 out_num, in_num,
2581 log, log_num, &desc);
2582 if (unlikely(ret < 0)) {
2583 if (ret != -EAGAIN)
2584 vq_err(vq, "Failure detected "
2585 "in indirect descriptor at idx %d\n", i);
2586 return ret;
2587 }
2588 continue;
2589 }
2590
2591 if (desc.flags & cpu_to_vhost16(vq, VRING_DESC_F_WRITE))
2592 access = VHOST_ACCESS_WO;
2593 else
2594 access = VHOST_ACCESS_RO;
2595 ret = translate_desc(vq, vhost64_to_cpu(vq, desc.addr),
2596 vhost32_to_cpu(vq, desc.len), iov + iov_count,
2597 iov_size - iov_count, access);
2598 if (unlikely(ret < 0)) {
2599 if (ret != -EAGAIN)
2600 vq_err(vq, "Translation failure %d descriptor idx %d\n",
2601 ret, i);
2602 return ret;
2603 }
2604 if (access == VHOST_ACCESS_WO) {
2605 /* If this is an input descriptor,
2606 * increment that count. */
2607 *in_num += ret;
2608 if (unlikely(log && ret)) {
2609 log[*log_num].addr = vhost64_to_cpu(vq, desc.addr);
2610 log[*log_num].len = vhost32_to_cpu(vq, desc.len);
2611 ++*log_num;
2612 }
2613 } else {
2614 /* If it's an output descriptor, they're all supposed
2615 * to come before any input descriptors. */
2616 if (unlikely(*in_num)) {
2617 vq_err(vq, "Descriptor has out after in: "
2618 "idx %d\n", i);
2619 return -EINVAL;
2620 }
2621 *out_num += ret;
2622 }
2623 } while ((i = next_desc(vq, &desc)) != -1);
2624
2625 /* On success, increment avail index. */
2626 vq->last_avail_idx++;
2627
2628 /* Assume notifications from guest are disabled at this point,
2629 * if they aren't we would need to update avail_event index. */
2630 BUG_ON(!(vq->used_flags & VRING_USED_F_NO_NOTIFY));
2631 return head;
2632}
2633EXPORT_SYMBOL_GPL(vhost_get_vq_desc);
2634
2635/* Reverse the effect of vhost_get_vq_desc. Useful for error handling. */
2636void vhost_discard_vq_desc(struct vhost_virtqueue *vq, int n)
2637{
2638 vq->last_avail_idx -= n;
2639}
2640EXPORT_SYMBOL_GPL(vhost_discard_vq_desc);
2641
2642/* After we've used one of their buffers, we tell them about it. We'll then
2643 * want to notify the guest, using eventfd. */
2644int vhost_add_used(struct vhost_virtqueue *vq, unsigned int head, int len)
2645{
2646 struct vring_used_elem heads = {
2647 cpu_to_vhost32(vq, head),
2648 cpu_to_vhost32(vq, len)
2649 };
2650
2651 return vhost_add_used_n(vq, &heads, 1);
2652}
2653EXPORT_SYMBOL_GPL(vhost_add_used);
2654
2655static int __vhost_add_used_n(struct vhost_virtqueue *vq,
2656 struct vring_used_elem *heads,
2657 unsigned count)
2658{
2659 vring_used_elem_t __user *used;
2660 u16 old, new;
2661 int start;
2662
2663 start = vq->last_used_idx & (vq->num - 1);
2664 used = vq->used->ring + start;
2665 if (vhost_put_used(vq, heads, start, count)) {
2666 vq_err(vq, "Failed to write used");
2667 return -EFAULT;
2668 }
2669 if (unlikely(vq->log_used)) {
2670 /* Make sure data is seen before log. */
2671 smp_wmb();
2672 /* Log used ring entry write. */
2673 log_used(vq, ((void __user *)used - (void __user *)vq->used),
2674 count * sizeof *used);
2675 }
2676 old = vq->last_used_idx;
2677 new = (vq->last_used_idx += count);
2678 /* If the driver never bothers to signal in a very long while,
2679 * used index might wrap around. If that happens, invalidate
2680 * signalled_used index we stored. TODO: make sure driver
2681 * signals at least once in 2^16 and remove this. */
2682 if (unlikely((u16)(new - vq->signalled_used) < (u16)(new - old)))
2683 vq->signalled_used_valid = false;
2684 return 0;
2685}
2686
2687/* After we've used one of their buffers, we tell them about it. We'll then
2688 * want to notify the guest, using eventfd. */
2689int vhost_add_used_n(struct vhost_virtqueue *vq, struct vring_used_elem *heads,
2690 unsigned count)
2691{
2692 int start, n, r;
2693
2694 start = vq->last_used_idx & (vq->num - 1);
2695 n = vq->num - start;
2696 if (n < count) {
2697 r = __vhost_add_used_n(vq, heads, n);
2698 if (r < 0)
2699 return r;
2700 heads += n;
2701 count -= n;
2702 }
2703 r = __vhost_add_used_n(vq, heads, count);
2704
2705 /* Make sure buffer is written before we update index. */
2706 smp_wmb();
2707 if (vhost_put_used_idx(vq)) {
2708 vq_err(vq, "Failed to increment used idx");
2709 return -EFAULT;
2710 }
2711 if (unlikely(vq->log_used)) {
2712 /* Make sure used idx is seen before log. */
2713 smp_wmb();
2714 /* Log used index update. */
2715 log_used(vq, offsetof(struct vring_used, idx),
2716 sizeof vq->used->idx);
2717 if (vq->log_ctx)
2718 eventfd_signal(vq->log_ctx);
2719 }
2720 return r;
2721}
2722EXPORT_SYMBOL_GPL(vhost_add_used_n);
2723
2724static bool vhost_notify(struct vhost_dev *dev, struct vhost_virtqueue *vq)
2725{
2726 __u16 old, new;
2727 __virtio16 event;
2728 bool v;
2729 /* Flush out used index updates. This is paired
2730 * with the barrier that the Guest executes when enabling
2731 * interrupts. */
2732 smp_mb();
2733
2734 if (vhost_has_feature(vq, VIRTIO_F_NOTIFY_ON_EMPTY) &&
2735 unlikely(vq->avail_idx == vq->last_avail_idx))
2736 return true;
2737
2738 if (!vhost_has_feature(vq, VIRTIO_RING_F_EVENT_IDX)) {
2739 __virtio16 flags;
2740 if (vhost_get_avail_flags(vq, &flags)) {
2741 vq_err(vq, "Failed to get flags");
2742 return true;
2743 }
2744 return !(flags & cpu_to_vhost16(vq, VRING_AVAIL_F_NO_INTERRUPT));
2745 }
2746 old = vq->signalled_used;
2747 v = vq->signalled_used_valid;
2748 new = vq->signalled_used = vq->last_used_idx;
2749 vq->signalled_used_valid = true;
2750
2751 if (unlikely(!v))
2752 return true;
2753
2754 if (vhost_get_used_event(vq, &event)) {
2755 vq_err(vq, "Failed to get used event idx");
2756 return true;
2757 }
2758 return vring_need_event(vhost16_to_cpu(vq, event), new, old);
2759}
2760
2761/* This actually signals the guest, using eventfd. */
2762void vhost_signal(struct vhost_dev *dev, struct vhost_virtqueue *vq)
2763{
2764 /* Signal the Guest tell them we used something up. */
2765 if (vq->call_ctx.ctx && vhost_notify(dev, vq))
2766 eventfd_signal(vq->call_ctx.ctx);
2767}
2768EXPORT_SYMBOL_GPL(vhost_signal);
2769
2770/* And here's the combo meal deal. Supersize me! */
2771void vhost_add_used_and_signal(struct vhost_dev *dev,
2772 struct vhost_virtqueue *vq,
2773 unsigned int head, int len)
2774{
2775 vhost_add_used(vq, head, len);
2776 vhost_signal(dev, vq);
2777}
2778EXPORT_SYMBOL_GPL(vhost_add_used_and_signal);
2779
2780/* multi-buffer version of vhost_add_used_and_signal */
2781void vhost_add_used_and_signal_n(struct vhost_dev *dev,
2782 struct vhost_virtqueue *vq,
2783 struct vring_used_elem *heads, unsigned count)
2784{
2785 vhost_add_used_n(vq, heads, count);
2786 vhost_signal(dev, vq);
2787}
2788EXPORT_SYMBOL_GPL(vhost_add_used_and_signal_n);
2789
2790/* return true if we're sure that avaiable ring is empty */
2791bool vhost_vq_avail_empty(struct vhost_dev *dev, struct vhost_virtqueue *vq)
2792{
2793 __virtio16 avail_idx;
2794 int r;
2795
2796 if (vq->avail_idx != vq->last_avail_idx)
2797 return false;
2798
2799 r = vhost_get_avail_idx(vq, &avail_idx);
2800 if (unlikely(r))
2801 return false;
2802 vq->avail_idx = vhost16_to_cpu(vq, avail_idx);
2803
2804 return vq->avail_idx == vq->last_avail_idx;
2805}
2806EXPORT_SYMBOL_GPL(vhost_vq_avail_empty);
2807
2808/* OK, now we need to know about added descriptors. */
2809bool vhost_enable_notify(struct vhost_dev *dev, struct vhost_virtqueue *vq)
2810{
2811 __virtio16 avail_idx;
2812 int r;
2813
2814 if (!(vq->used_flags & VRING_USED_F_NO_NOTIFY))
2815 return false;
2816 vq->used_flags &= ~VRING_USED_F_NO_NOTIFY;
2817 if (!vhost_has_feature(vq, VIRTIO_RING_F_EVENT_IDX)) {
2818 r = vhost_update_used_flags(vq);
2819 if (r) {
2820 vq_err(vq, "Failed to enable notification at %p: %d\n",
2821 &vq->used->flags, r);
2822 return false;
2823 }
2824 } else {
2825 r = vhost_update_avail_event(vq);
2826 if (r) {
2827 vq_err(vq, "Failed to update avail event index at %p: %d\n",
2828 vhost_avail_event(vq), r);
2829 return false;
2830 }
2831 }
2832 /* They could have slipped one in as we were doing that: make
2833 * sure it's written, then check again. */
2834 smp_mb();
2835 r = vhost_get_avail_idx(vq, &avail_idx);
2836 if (r) {
2837 vq_err(vq, "Failed to check avail idx at %p: %d\n",
2838 &vq->avail->idx, r);
2839 return false;
2840 }
2841 vq->avail_idx = vhost16_to_cpu(vq, avail_idx);
2842
2843 return vq->avail_idx != vq->last_avail_idx;
2844}
2845EXPORT_SYMBOL_GPL(vhost_enable_notify);
2846
2847/* We don't need to be notified again. */
2848void vhost_disable_notify(struct vhost_dev *dev, struct vhost_virtqueue *vq)
2849{
2850 int r;
2851
2852 if (vq->used_flags & VRING_USED_F_NO_NOTIFY)
2853 return;
2854 vq->used_flags |= VRING_USED_F_NO_NOTIFY;
2855 if (!vhost_has_feature(vq, VIRTIO_RING_F_EVENT_IDX)) {
2856 r = vhost_update_used_flags(vq);
2857 if (r)
2858 vq_err(vq, "Failed to disable notification at %p: %d\n",
2859 &vq->used->flags, r);
2860 }
2861}
2862EXPORT_SYMBOL_GPL(vhost_disable_notify);
2863
2864/* Create a new message. */
2865struct vhost_msg_node *vhost_new_msg(struct vhost_virtqueue *vq, int type)
2866{
2867 /* Make sure all padding within the structure is initialized. */
2868 struct vhost_msg_node *node = kzalloc(sizeof(*node), GFP_KERNEL);
2869 if (!node)
2870 return NULL;
2871
2872 node->vq = vq;
2873 node->msg.type = type;
2874 return node;
2875}
2876EXPORT_SYMBOL_GPL(vhost_new_msg);
2877
2878void vhost_enqueue_msg(struct vhost_dev *dev, struct list_head *head,
2879 struct vhost_msg_node *node)
2880{
2881 spin_lock(&dev->iotlb_lock);
2882 list_add_tail(&node->node, head);
2883 spin_unlock(&dev->iotlb_lock);
2884
2885 wake_up_interruptible_poll(&dev->wait, EPOLLIN | EPOLLRDNORM);
2886}
2887EXPORT_SYMBOL_GPL(vhost_enqueue_msg);
2888
2889struct vhost_msg_node *vhost_dequeue_msg(struct vhost_dev *dev,
2890 struct list_head *head)
2891{
2892 struct vhost_msg_node *node = NULL;
2893
2894 spin_lock(&dev->iotlb_lock);
2895 if (!list_empty(head)) {
2896 node = list_first_entry(head, struct vhost_msg_node,
2897 node);
2898 list_del(&node->node);
2899 }
2900 spin_unlock(&dev->iotlb_lock);
2901
2902 return node;
2903}
2904EXPORT_SYMBOL_GPL(vhost_dequeue_msg);
2905
2906void vhost_set_backend_features(struct vhost_dev *dev, u64 features)
2907{
2908 struct vhost_virtqueue *vq;
2909 int i;
2910
2911 mutex_lock(&dev->mutex);
2912 for (i = 0; i < dev->nvqs; ++i) {
2913 vq = dev->vqs[i];
2914 mutex_lock(&vq->mutex);
2915 vq->acked_backend_features = features;
2916 mutex_unlock(&vq->mutex);
2917 }
2918 mutex_unlock(&dev->mutex);
2919}
2920EXPORT_SYMBOL_GPL(vhost_set_backend_features);
2921
2922static int __init vhost_init(void)
2923{
2924 return 0;
2925}
2926
2927static void __exit vhost_exit(void)
2928{
2929}
2930
2931module_init(vhost_init);
2932module_exit(vhost_exit);
2933
2934MODULE_VERSION("0.0.1");
2935MODULE_LICENSE("GPL v2");
2936MODULE_AUTHOR("Michael S. Tsirkin");
2937MODULE_DESCRIPTION("Host kernel accelerator for virtio");