Blame view
drivers/vhost/vhost.c
40.5 KB
3a4d5c94e vhost_net: a kern... |
1 2 3 4 5 6 |
/* Copyright (C) 2009 Red Hat, Inc. * Copyright (C) 2006 Rusty Russell IBM Corporation * * Author: Michael S. Tsirkin <mst@redhat.com> * * Inspiration, some code, and most witty comments come from |
615165875 Correct occurrenc... |
7 |
* Documentation/virtual/lguest/lguest.c, by Rusty Russell |
3a4d5c94e vhost_net: a kern... |
8 9 10 11 12 13 14 15 16 17 |
* * This work is licensed under the terms of the GNU GPL, version 2. * * Generic code for virtio server in host kernel. */ #include <linux/eventfd.h> #include <linux/vhost.h> #include <linux/virtio_net.h> #include <linux/mm.h> |
64e1c8074 vhost-net: batch ... |
18 |
#include <linux/mmu_context.h> |
3a4d5c94e vhost_net: a kern... |
19 20 |
#include <linux/miscdevice.h> #include <linux/mutex.h> |
3a4d5c94e vhost_net: a kern... |
21 22 23 24 |
#include <linux/rcupdate.h> #include <linux/poll.h> #include <linux/file.h> #include <linux/highmem.h> |
5a0e3ad6a include cleanup: ... |
25 |
#include <linux/slab.h> |
c23f3445e vhost: replace vh... |
26 |
#include <linux/kthread.h> |
9e3d19572 vhost: apply cgro... |
27 |
#include <linux/cgroup.h> |
3a4d5c94e vhost_net: a kern... |
28 29 30 31 |
#include <linux/net.h> #include <linux/if_packet.h> #include <linux/if_arp.h> |
3a4d5c94e vhost_net: a kern... |
32 33 34 35 36 37 |
#include "vhost.h" enum { VHOST_MEMORY_MAX_NREGIONS = 64, VHOST_MEMORY_F_LOG = 0x1, }; |
bab632d69 vhost: vhost TX z... |
38 |
static unsigned vhost_zcopy_mask __read_mostly; |
8ea8cf89e vhost: support ev... |
39 40 |
#define vhost_used_event(vq) ((u16 __user *)&vq->avail->ring[vq->num]) #define vhost_avail_event(vq) ((u16 __user *)&vq->used->ring[vq->num]) |
3a4d5c94e vhost_net: a kern... |
41 42 43 44 |
static void vhost_poll_func(struct file *file, wait_queue_head_t *wqh, poll_table *pt) { struct vhost_poll *poll; |
3a4d5c94e vhost_net: a kern... |
45 |
|
d47effe1b vhost: Cleanup vh... |
46 |
poll = container_of(pt, struct vhost_poll, table); |
3a4d5c94e vhost_net: a kern... |
47 48 49 50 51 52 53 |
poll->wqh = wqh; add_wait_queue(wqh, &poll->wait); } static int vhost_poll_wakeup(wait_queue_t *wait, unsigned mode, int sync, void *key) { |
c23f3445e vhost: replace vh... |
54 |
struct vhost_poll *poll = container_of(wait, struct vhost_poll, wait); |
3a4d5c94e vhost_net: a kern... |
55 56 |
if (!((unsigned long)key & poll->mask)) return 0; |
c23f3445e vhost: replace vh... |
57 |
vhost_poll_queue(poll); |
3a4d5c94e vhost_net: a kern... |
58 59 |
return 0; } |
87d6a412b vhost: fix attach... |
60 61 62 63 64 65 66 67 |
static void vhost_work_init(struct vhost_work *work, vhost_work_fn_t fn) { INIT_LIST_HEAD(&work->node); work->fn = fn; init_waitqueue_head(&work->done); work->flushing = 0; work->queue_seq = work->done_seq = 0; } |
3a4d5c94e vhost_net: a kern... |
68 |
/* Init poll structure */ |
c23f3445e vhost: replace vh... |
69 70 |
void vhost_poll_init(struct vhost_poll *poll, vhost_work_fn_t fn, unsigned long mask, struct vhost_dev *dev) |
3a4d5c94e vhost_net: a kern... |
71 |
{ |
3a4d5c94e vhost_net: a kern... |
72 73 74 |
init_waitqueue_func_entry(&poll->wait, vhost_poll_wakeup); init_poll_funcptr(&poll->table, vhost_poll_func); poll->mask = mask; |
c23f3445e vhost: replace vh... |
75 |
poll->dev = dev; |
87d6a412b vhost: fix attach... |
76 |
vhost_work_init(&poll->work, fn); |
3a4d5c94e vhost_net: a kern... |
77 78 79 80 81 82 83 |
} /* Start polling a file. We add ourselves to file's wait queue. The caller must * keep a reference to a file until after vhost_poll_stop is called. */ void vhost_poll_start(struct vhost_poll *poll, struct file *file) { unsigned long mask; |
d47effe1b vhost: Cleanup vh... |
84 |
|
3a4d5c94e vhost_net: a kern... |
85 86 87 88 89 90 91 92 93 94 95 |
mask = file->f_op->poll(file, &poll->table); if (mask) vhost_poll_wakeup(&poll->wait, 0, 0, (void *)mask); } /* Stop polling a file. After this function returns, it becomes safe to drop the * file reference. You must also flush afterwards. */ void vhost_poll_stop(struct vhost_poll *poll) { remove_wait_queue(poll->wqh, &poll->wait); } |
0174b0c30 vhost: fix signed... |
96 97 98 99 |
static bool vhost_work_seq_done(struct vhost_dev *dev, struct vhost_work *work, unsigned seq) { int left; |
d47effe1b vhost: Cleanup vh... |
100 |
|
0174b0c30 vhost: fix signed... |
101 102 103 104 105 |
spin_lock_irq(&dev->work_lock); left = seq - work->done_seq; spin_unlock_irq(&dev->work_lock); return left <= 0; } |
87d6a412b vhost: fix attach... |
106 |
static void vhost_work_flush(struct vhost_dev *dev, struct vhost_work *work) |
3a4d5c94e vhost_net: a kern... |
107 |
{ |
c23f3445e vhost: replace vh... |
108 |
unsigned seq; |
c23f3445e vhost: replace vh... |
109 |
int flushing; |
87d6a412b vhost: fix attach... |
110 |
spin_lock_irq(&dev->work_lock); |
c23f3445e vhost: replace vh... |
111 112 |
seq = work->queue_seq; work->flushing++; |
87d6a412b vhost: fix attach... |
113 |
spin_unlock_irq(&dev->work_lock); |
0174b0c30 vhost: fix signed... |
114 |
wait_event(work->done, vhost_work_seq_done(dev, work, seq)); |
87d6a412b vhost: fix attach... |
115 |
spin_lock_irq(&dev->work_lock); |
c23f3445e vhost: replace vh... |
116 |
flushing = --work->flushing; |
87d6a412b vhost: fix attach... |
117 |
spin_unlock_irq(&dev->work_lock); |
c23f3445e vhost: replace vh... |
118 |
BUG_ON(flushing < 0); |
3a4d5c94e vhost_net: a kern... |
119 |
} |
87d6a412b vhost: fix attach... |
120 121 122 123 124 125 126 127 128 |
/* Flush any work that has been scheduled. When calling this, don't hold any * locks that are also used by the callback. */ void vhost_poll_flush(struct vhost_poll *poll) { vhost_work_flush(poll->dev, &poll->work); } static inline void vhost_work_queue(struct vhost_dev *dev, struct vhost_work *work) |
3a4d5c94e vhost_net: a kern... |
129 |
{ |
c23f3445e vhost: replace vh... |
130 131 132 133 134 135 136 137 138 |
unsigned long flags; spin_lock_irqsave(&dev->work_lock, flags); if (list_empty(&work->node)) { list_add_tail(&work->node, &dev->work_list); work->queue_seq++; wake_up_process(dev->worker); } spin_unlock_irqrestore(&dev->work_lock, flags); |
3a4d5c94e vhost_net: a kern... |
139 |
} |
87d6a412b vhost: fix attach... |
140 141 142 143 |
void vhost_poll_queue(struct vhost_poll *poll) { vhost_work_queue(poll->dev, &poll->work); } |
3a4d5c94e vhost_net: a kern... |
144 145 146 147 148 149 150 151 152 153 |
static void vhost_vq_reset(struct vhost_dev *dev, struct vhost_virtqueue *vq) { vq->num = 1; vq->desc = NULL; vq->avail = NULL; vq->used = NULL; vq->last_avail_idx = 0; vq->avail_idx = 0; vq->last_used_idx = 0; |
8ea8cf89e vhost: support ev... |
154 155 |
vq->signalled_used = 0; vq->signalled_used_valid = false; |
3a4d5c94e vhost_net: a kern... |
156 |
vq->used_flags = 0; |
3a4d5c94e vhost_net: a kern... |
157 158 |
vq->log_used = false; vq->log_addr = -1ull; |
8dd014adf vhost-net: mergea... |
159 160 |
vq->vhost_hlen = 0; vq->sock_hlen = 0; |
3a4d5c94e vhost_net: a kern... |
161 162 163 164 165 166 167 |
vq->private_data = NULL; vq->log_base = NULL; vq->error_ctx = NULL; vq->error = NULL; vq->kick = NULL; vq->call_ctx = NULL; vq->call = NULL; |
73a99f083 vhost: initialize... |
168 |
vq->log_ctx = NULL; |
bab632d69 vhost: vhost TX z... |
169 170 171 |
vq->upend_idx = 0; vq->done_idx = 0; vq->ubufs = NULL; |
3a4d5c94e vhost_net: a kern... |
172 |
} |
c23f3445e vhost: replace vh... |
173 174 175 176 177 |
static int vhost_worker(void *data) { struct vhost_dev *dev = data; struct vhost_work *work = NULL; unsigned uninitialized_var(seq); |
64e1c8074 vhost-net: batch ... |
178 |
use_mm(dev->mm); |
c23f3445e vhost: replace vh... |
179 180 181 182 183 184 185 186 187 188 189 190 191 192 |
for (;;) { /* mb paired w/ kthread_stop */ set_current_state(TASK_INTERRUPTIBLE); spin_lock_irq(&dev->work_lock); if (work) { work->done_seq = seq; if (work->flushing) wake_up_all(&work->done); } if (kthread_should_stop()) { spin_unlock_irq(&dev->work_lock); __set_current_state(TASK_RUNNING); |
64e1c8074 vhost-net: batch ... |
193 |
break; |
c23f3445e vhost: replace vh... |
194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 |
} if (!list_empty(&dev->work_list)) { work = list_first_entry(&dev->work_list, struct vhost_work, node); list_del_init(&work->node); seq = work->queue_seq; } else work = NULL; spin_unlock_irq(&dev->work_lock); if (work) { __set_current_state(TASK_RUNNING); work->fn(work); } else schedule(); } |
64e1c8074 vhost-net: batch ... |
211 212 |
unuse_mm(dev->mm); return 0; |
c23f3445e vhost: replace vh... |
213 |
} |
bab632d69 vhost: vhost TX z... |
214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 |
static void vhost_vq_free_iovecs(struct vhost_virtqueue *vq) { kfree(vq->indirect); vq->indirect = NULL; kfree(vq->log); vq->log = NULL; kfree(vq->heads); vq->heads = NULL; kfree(vq->ubuf_info); vq->ubuf_info = NULL; } void vhost_enable_zcopy(int vq) { vhost_zcopy_mask |= 0x1 << vq; } |
e0e9b4064 vhost: max s/g to... |
230 231 232 233 |
/* Helper to allocate iovec buffers for all vqs. */ static long vhost_dev_alloc_iovecs(struct vhost_dev *dev) { int i; |
bab632d69 vhost: vhost TX z... |
234 |
bool zcopy; |
d47effe1b vhost: Cleanup vh... |
235 |
|
e0e9b4064 vhost: max s/g to... |
236 237 238 239 240 241 242 |
for (i = 0; i < dev->nvqs; ++i) { dev->vqs[i].indirect = kmalloc(sizeof *dev->vqs[i].indirect * UIO_MAXIOV, GFP_KERNEL); dev->vqs[i].log = kmalloc(sizeof *dev->vqs[i].log * UIO_MAXIOV, GFP_KERNEL); dev->vqs[i].heads = kmalloc(sizeof *dev->vqs[i].heads * UIO_MAXIOV, GFP_KERNEL); |
bab632d69 vhost: vhost TX z... |
243 244 245 246 247 |
zcopy = vhost_zcopy_mask & (0x1 << i); if (zcopy) dev->vqs[i].ubuf_info = kmalloc(sizeof *dev->vqs[i].ubuf_info * UIO_MAXIOV, GFP_KERNEL); |
e0e9b4064 vhost: max s/g to... |
248 |
if (!dev->vqs[i].indirect || !dev->vqs[i].log || |
bab632d69 vhost: vhost TX z... |
249 250 |
!dev->vqs[i].heads || (zcopy && !dev->vqs[i].ubuf_info)) |
e0e9b4064 vhost: max s/g to... |
251 252 253 |
goto err_nomem; } return 0; |
d47effe1b vhost: Cleanup vh... |
254 |
|
e0e9b4064 vhost: max s/g to... |
255 |
err_nomem: |
bab632d69 vhost: vhost TX z... |
256 257 |
for (; i >= 0; --i) vhost_vq_free_iovecs(&dev->vqs[i]); |
e0e9b4064 vhost: max s/g to... |
258 259 260 261 262 263 |
return -ENOMEM; } static void vhost_dev_free_iovecs(struct vhost_dev *dev) { int i; |
d47effe1b vhost: Cleanup vh... |
264 |
|
bab632d69 vhost: vhost TX z... |
265 266 |
for (i = 0; i < dev->nvqs; ++i) vhost_vq_free_iovecs(&dev->vqs[i]); |
e0e9b4064 vhost: max s/g to... |
267 |
} |
3a4d5c94e vhost_net: a kern... |
268 269 270 271 |
long vhost_dev_init(struct vhost_dev *dev, struct vhost_virtqueue *vqs, int nvqs) { int i; |
c23f3445e vhost: replace vh... |
272 |
|
3a4d5c94e vhost_net: a kern... |
273 274 275 276 277 278 279 |
dev->vqs = vqs; dev->nvqs = nvqs; mutex_init(&dev->mutex); dev->log_ctx = NULL; dev->log_file = NULL; dev->memory = NULL; dev->mm = NULL; |
c23f3445e vhost: replace vh... |
280 281 282 |
spin_lock_init(&dev->work_lock); INIT_LIST_HEAD(&dev->work_list); dev->worker = NULL; |
3a4d5c94e vhost_net: a kern... |
283 284 |
for (i = 0; i < dev->nvqs; ++i) { |
e0e9b4064 vhost: max s/g to... |
285 286 287 |
dev->vqs[i].log = NULL; dev->vqs[i].indirect = NULL; dev->vqs[i].heads = NULL; |
bab632d69 vhost: vhost TX z... |
288 |
dev->vqs[i].ubuf_info = NULL; |
3a4d5c94e vhost_net: a kern... |
289 290 291 292 293 |
dev->vqs[i].dev = dev; mutex_init(&dev->vqs[i].mutex); vhost_vq_reset(dev, dev->vqs + i); if (dev->vqs[i].handle_kick) vhost_poll_init(&dev->vqs[i].poll, |
c23f3445e vhost: replace vh... |
294 |
dev->vqs[i].handle_kick, POLLIN, dev); |
3a4d5c94e vhost_net: a kern... |
295 |
} |
c23f3445e vhost: replace vh... |
296 |
|
3a4d5c94e vhost_net: a kern... |
297 298 299 300 301 302 303 304 305 |
return 0; } /* Caller should have device mutex */ long vhost_dev_check_owner(struct vhost_dev *dev) { /* Are you the owner? If not, I don't think you mean to do that */ return dev->mm == current->mm ? 0 : -EPERM; } |
87d6a412b vhost: fix attach... |
306 |
struct vhost_attach_cgroups_struct { |
d47effe1b vhost: Cleanup vh... |
307 308 309 |
struct vhost_work work; struct task_struct *owner; int ret; |
87d6a412b vhost: fix attach... |
310 311 312 313 |
}; static void vhost_attach_cgroups_work(struct vhost_work *work) { |
d47effe1b vhost: Cleanup vh... |
314 315 316 317 |
struct vhost_attach_cgroups_struct *s; s = container_of(work, struct vhost_attach_cgroups_struct, work); s->ret = cgroup_attach_task_all(s->owner, current); |
87d6a412b vhost: fix attach... |
318 319 320 321 |
} static int vhost_attach_cgroups(struct vhost_dev *dev) { |
d47effe1b vhost: Cleanup vh... |
322 323 324 325 326 327 328 |
struct vhost_attach_cgroups_struct attach; attach.owner = current; vhost_work_init(&attach.work, vhost_attach_cgroups_work); vhost_work_queue(dev, &attach.work); vhost_work_flush(dev, &attach.work); return attach.ret; |
87d6a412b vhost: fix attach... |
329 |
} |
3a4d5c94e vhost_net: a kern... |
330 331 332 |
/* Caller should have device mutex */ static long vhost_dev_set_owner(struct vhost_dev *dev) { |
c23f3445e vhost: replace vh... |
333 334 |
struct task_struct *worker; int err; |
d47effe1b vhost: Cleanup vh... |
335 |
|
3a4d5c94e vhost_net: a kern... |
336 |
/* Is there an owner already? */ |
c23f3445e vhost: replace vh... |
337 338 339 340 |
if (dev->mm) { err = -EBUSY; goto err_mm; } |
d47effe1b vhost: Cleanup vh... |
341 |
|
3a4d5c94e vhost_net: a kern... |
342 343 |
/* No owner, become one */ dev->mm = get_task_mm(current); |
c23f3445e vhost: replace vh... |
344 345 346 347 348 349 350 |
worker = kthread_create(vhost_worker, dev, "vhost-%d", current->pid); if (IS_ERR(worker)) { err = PTR_ERR(worker); goto err_worker; } dev->worker = worker; |
87d6a412b vhost: fix attach... |
351 352 353 |
wake_up_process(worker); /* avoid contributing to loadavg */ err = vhost_attach_cgroups(dev); |
9e3d19572 vhost: apply cgro... |
354 355 |
if (err) goto err_cgroup; |
c23f3445e vhost: replace vh... |
356 |
|
e0e9b4064 vhost: max s/g to... |
357 358 359 |
err = vhost_dev_alloc_iovecs(dev); if (err) goto err_cgroup; |
3a4d5c94e vhost_net: a kern... |
360 |
return 0; |
9e3d19572 vhost: apply cgro... |
361 362 |
err_cgroup: kthread_stop(worker); |
615cc2211 vhost: error hand... |
363 |
dev->worker = NULL; |
c23f3445e vhost: replace vh... |
364 365 366 367 368 369 |
err_worker: if (dev->mm) mmput(dev->mm); dev->mm = NULL; err_mm: return err; |
3a4d5c94e vhost_net: a kern... |
370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 |
} /* Caller should have device mutex */ long vhost_dev_reset_owner(struct vhost_dev *dev) { struct vhost_memory *memory; /* Restore memory to default empty mapping. */ memory = kmalloc(offsetof(struct vhost_memory, regions), GFP_KERNEL); if (!memory) return -ENOMEM; vhost_dev_cleanup(dev); memory->nregions = 0; |
28457ee69 vhost: add __rcu ... |
385 |
RCU_INIT_POINTER(dev->memory, memory); |
3a4d5c94e vhost_net: a kern... |
386 387 |
return 0; } |
bab632d69 vhost: vhost TX z... |
388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 |
/* In case of DMA done not in order in lower device driver for some reason. * upend_idx is used to track end of used idx, done_idx is used to track head * of used idx. Once lower device DMA done contiguously, we will signal KVM * guest used idx. */ int vhost_zerocopy_signal_used(struct vhost_virtqueue *vq) { int i; int j = 0; for (i = vq->done_idx; i != vq->upend_idx; i = (i + 1) % UIO_MAXIOV) { if ((vq->heads[i].len == VHOST_DMA_DONE_LEN)) { vq->heads[i].len = VHOST_DMA_CLEAR_LEN; vhost_add_used_and_signal(vq->dev, vq, vq->heads[i].id, 0); ++j; } else break; } if (j) vq->done_idx = i; return j; } |
3a4d5c94e vhost_net: a kern... |
411 412 413 414 |
/* Caller should have device mutex */ void vhost_dev_cleanup(struct vhost_dev *dev) { int i; |
d47effe1b vhost: Cleanup vh... |
415 |
|
3a4d5c94e vhost_net: a kern... |
416 417 418 419 420 |
for (i = 0; i < dev->nvqs; ++i) { if (dev->vqs[i].kick && dev->vqs[i].handle_kick) { vhost_poll_stop(&dev->vqs[i].poll); vhost_poll_flush(&dev->vqs[i].poll); } |
bab632d69 vhost: vhost TX z... |
421 422 423 424 425 426 |
/* Wait for all lower device DMAs done. */ if (dev->vqs[i].ubufs) vhost_ubuf_put_and_wait(dev->vqs[i].ubufs); /* Signal guest as appropriate. */ vhost_zerocopy_signal_used(&dev->vqs[i]); |
3a4d5c94e vhost_net: a kern... |
427 428 429 430 431 432 433 434 435 436 437 438 |
if (dev->vqs[i].error_ctx) eventfd_ctx_put(dev->vqs[i].error_ctx); if (dev->vqs[i].error) fput(dev->vqs[i].error); if (dev->vqs[i].kick) fput(dev->vqs[i].kick); if (dev->vqs[i].call_ctx) eventfd_ctx_put(dev->vqs[i].call_ctx); if (dev->vqs[i].call) fput(dev->vqs[i].call); vhost_vq_reset(dev, dev->vqs + i); } |
e0e9b4064 vhost: max s/g to... |
439 |
vhost_dev_free_iovecs(dev); |
3a4d5c94e vhost_net: a kern... |
440 441 442 443 444 445 446 |
if (dev->log_ctx) eventfd_ctx_put(dev->log_ctx); dev->log_ctx = NULL; if (dev->log_file) fput(dev->log_file); dev->log_file = NULL; /* No one will access memory at this point */ |
28457ee69 vhost: add __rcu ... |
447 448 449 |
kfree(rcu_dereference_protected(dev->memory, lockdep_is_held(&dev->mutex))); RCU_INIT_POINTER(dev->memory, NULL); |
c23f3445e vhost: replace vh... |
450 |
WARN_ON(!list_empty(&dev->work_list)); |
78b620ce9 vhost: stop worke... |
451 452 453 454 |
if (dev->worker) { kthread_stop(dev->worker); dev->worker = NULL; } |
533a19b4b vhost: put mm aft... |
455 456 457 |
if (dev->mm) mmput(dev->mm); dev->mm = NULL; |
3a4d5c94e vhost_net: a kern... |
458 459 460 461 462 |
} static int log_access_ok(void __user *log_base, u64 addr, unsigned long sz) { u64 a = addr / VHOST_PAGE_SIZE / 8; |
d47effe1b vhost: Cleanup vh... |
463 |
|
3a4d5c94e vhost_net: a kern... |
464 465 466 |
/* Make sure 64 bit math will not overflow. */ if (a > ULONG_MAX - (unsigned long)log_base || a + (unsigned long)log_base > ULONG_MAX) |
6d97e55f7 vhost: fix return... |
467 |
return 0; |
3a4d5c94e vhost_net: a kern... |
468 469 470 471 472 473 474 475 476 477 |
return access_ok(VERIFY_WRITE, log_base + a, (sz + VHOST_PAGE_SIZE * 8 - 1) / VHOST_PAGE_SIZE / 8); } /* Caller should have vq mutex and device mutex. */ static int vq_memory_access_ok(void __user *log_base, struct vhost_memory *mem, int log_all) { int i; |
179b284e2 vhost-net: fix vq... |
478 |
|
f8322fbe0 vhost: whitespace... |
479 480 |
if (!mem) return 0; |
179b284e2 vhost-net: fix vq... |
481 |
|
3a4d5c94e vhost_net: a kern... |
482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 |
for (i = 0; i < mem->nregions; ++i) { struct vhost_memory_region *m = mem->regions + i; unsigned long a = m->userspace_addr; if (m->memory_size > ULONG_MAX) return 0; else if (!access_ok(VERIFY_WRITE, (void __user *)a, m->memory_size)) return 0; else if (log_all && !log_access_ok(log_base, m->guest_phys_addr, m->memory_size)) return 0; } return 1; } /* Can we switch to this memory table? */ /* Caller should have device mutex but not vq mutex */ static int memory_access_ok(struct vhost_dev *d, struct vhost_memory *mem, int log_all) { int i; |
d47effe1b vhost: Cleanup vh... |
504 |
|
3a4d5c94e vhost_net: a kern... |
505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 |
for (i = 0; i < d->nvqs; ++i) { int ok; mutex_lock(&d->vqs[i].mutex); /* If ring is inactive, will check when it's enabled. */ if (d->vqs[i].private_data) ok = vq_memory_access_ok(d->vqs[i].log_base, mem, log_all); else ok = 1; mutex_unlock(&d->vqs[i].mutex); if (!ok) return 0; } return 1; } |
8ea8cf89e vhost: support ev... |
520 |
static int vq_access_ok(struct vhost_dev *d, unsigned int num, |
3a4d5c94e vhost_net: a kern... |
521 522 523 524 |
struct vring_desc __user *desc, struct vring_avail __user *avail, struct vring_used __user *used) { |
8ea8cf89e vhost: support ev... |
525 |
size_t s = vhost_has_feature(d, VIRTIO_RING_F_EVENT_IDX) ? 2 : 0; |
3a4d5c94e vhost_net: a kern... |
526 527 |
return access_ok(VERIFY_READ, desc, num * sizeof *desc) && access_ok(VERIFY_READ, avail, |
8ea8cf89e vhost: support ev... |
528 |
sizeof *avail + num * sizeof *avail->ring + s) && |
3a4d5c94e vhost_net: a kern... |
529 |
access_ok(VERIFY_WRITE, used, |
8ea8cf89e vhost: support ev... |
530 |
sizeof *used + num * sizeof *used->ring + s); |
3a4d5c94e vhost_net: a kern... |
531 532 533 534 535 536 |
} /* Can we log writes? */ /* Caller should have device mutex but not vq mutex */ int vhost_log_access_ok(struct vhost_dev *dev) { |
28457ee69 vhost: add __rcu ... |
537 538 539 540 541 |
struct vhost_memory *mp; mp = rcu_dereference_protected(dev->memory, lockdep_is_held(&dev->mutex)); return memory_access_ok(dev, mp, 1); |
3a4d5c94e vhost_net: a kern... |
542 543 544 545 |
} /* Verify access for write logging. */ /* Caller should have vq mutex and device mutex */ |
8ea8cf89e vhost: support ev... |
546 547 |
static int vq_log_access_ok(struct vhost_dev *d, struct vhost_virtqueue *vq, void __user *log_base) |
3a4d5c94e vhost_net: a kern... |
548 |
{ |
28457ee69 vhost: add __rcu ... |
549 |
struct vhost_memory *mp; |
8ea8cf89e vhost: support ev... |
550 |
size_t s = vhost_has_feature(d, VIRTIO_RING_F_EVENT_IDX) ? 2 : 0; |
28457ee69 vhost: add __rcu ... |
551 552 553 554 |
mp = rcu_dereference_protected(vq->dev->memory, lockdep_is_held(&vq->mutex)); return vq_memory_access_ok(log_base, mp, |
3a4d5c94e vhost_net: a kern... |
555 556 557 |
vhost_has_feature(vq->dev, VHOST_F_LOG_ALL)) && (!vq->log_used || log_access_ok(log_base, vq->log_addr, sizeof *vq->used + |
8ea8cf89e vhost: support ev... |
558 |
vq->num * sizeof *vq->used->ring + s)); |
3a4d5c94e vhost_net: a kern... |
559 560 561 562 563 564 |
} /* Can we start vq? */ /* Caller should have vq mutex and device mutex */ int vhost_vq_access_ok(struct vhost_virtqueue *vq) { |
8ea8cf89e vhost: support ev... |
565 566 |
return vq_access_ok(vq->dev, vq->num, vq->desc, vq->avail, vq->used) && vq_log_access_ok(vq->dev, vq, vq->log_base); |
3a4d5c94e vhost_net: a kern... |
567 568 569 570 571 572 |
} static long vhost_set_memory(struct vhost_dev *d, struct vhost_memory __user *m) { struct vhost_memory mem, *newmem, *oldmem; unsigned long size = offsetof(struct vhost_memory, regions); |
d47effe1b vhost: Cleanup vh... |
573 |
|
7ad9c9d27 vhost: fix to che... |
574 575 |
if (copy_from_user(&mem, m, size)) return -EFAULT; |
3a4d5c94e vhost_net: a kern... |
576 577 578 579 580 581 582 583 584 |
if (mem.padding) return -EOPNOTSUPP; if (mem.nregions > VHOST_MEMORY_MAX_NREGIONS) return -E2BIG; newmem = kmalloc(size + mem.nregions * sizeof *m->regions, GFP_KERNEL); if (!newmem) return -ENOMEM; memcpy(newmem, &mem, size); |
7ad9c9d27 vhost: fix to che... |
585 586 |
if (copy_from_user(newmem->regions, m->regions, mem.nregions * sizeof *m->regions)) { |
3a4d5c94e vhost_net: a kern... |
587 |
kfree(newmem); |
7ad9c9d27 vhost: fix to che... |
588 |
return -EFAULT; |
3a4d5c94e vhost_net: a kern... |
589 |
} |
d47effe1b vhost: Cleanup vh... |
590 591 |
if (!memory_access_ok(d, newmem, vhost_has_feature(d, VHOST_F_LOG_ALL))) { |
a02c37891 vhost: fix the me... |
592 |
kfree(newmem); |
3a4d5c94e vhost_net: a kern... |
593 |
return -EFAULT; |
a02c37891 vhost: fix the me... |
594 |
} |
28457ee69 vhost: add __rcu ... |
595 596 |
oldmem = rcu_dereference_protected(d->memory, lockdep_is_held(&d->mutex)); |
3a4d5c94e vhost_net: a kern... |
597 598 599 600 601 |
rcu_assign_pointer(d->memory, newmem); synchronize_rcu(); kfree(oldmem); return 0; } |
3a4d5c94e vhost_net: a kern... |
602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 |
static long vhost_set_vring(struct vhost_dev *d, int ioctl, void __user *argp) { struct file *eventfp, *filep = NULL, *pollstart = NULL, *pollstop = NULL; struct eventfd_ctx *ctx = NULL; u32 __user *idxp = argp; struct vhost_virtqueue *vq; struct vhost_vring_state s; struct vhost_vring_file f; struct vhost_vring_addr a; u32 idx; long r; r = get_user(idx, idxp); if (r < 0) return r; |
0f3d9a174 vhost: Fix host p... |
618 |
if (idx >= d->nvqs) |
3a4d5c94e vhost_net: a kern... |
619 620 621 622 623 624 625 626 627 628 629 630 631 632 |
return -ENOBUFS; vq = d->vqs + idx; mutex_lock(&vq->mutex); switch (ioctl) { case VHOST_SET_VRING_NUM: /* Resizing ring with an active backend? * You don't want to do that. */ if (vq->private_data) { r = -EBUSY; break; } |
7ad9c9d27 vhost: fix to che... |
633 634 |
if (copy_from_user(&s, argp, sizeof s)) { r = -EFAULT; |
3a4d5c94e vhost_net: a kern... |
635 |
break; |
7ad9c9d27 vhost: fix to che... |
636 |
} |
3a4d5c94e vhost_net: a kern... |
637 638 639 640 641 642 643 644 645 646 647 648 649 |
if (!s.num || s.num > 0xffff || (s.num & (s.num - 1))) { r = -EINVAL; break; } vq->num = s.num; break; case VHOST_SET_VRING_BASE: /* Moving base with an active backend? * You don't want to do that. */ if (vq->private_data) { r = -EBUSY; break; } |
7ad9c9d27 vhost: fix to che... |
650 651 |
if (copy_from_user(&s, argp, sizeof s)) { r = -EFAULT; |
3a4d5c94e vhost_net: a kern... |
652 |
break; |
7ad9c9d27 vhost: fix to che... |
653 |
} |
3a4d5c94e vhost_net: a kern... |
654 655 656 657 658 659 660 661 662 663 664 |
if (s.num > 0xffff) { r = -EINVAL; break; } vq->last_avail_idx = s.num; /* Forget the cached index value. */ vq->avail_idx = vq->last_avail_idx; break; case VHOST_GET_VRING_BASE: s.index = idx; s.num = vq->last_avail_idx; |
7ad9c9d27 vhost: fix to che... |
665 666 |
if (copy_to_user(argp, &s, sizeof s)) r = -EFAULT; |
3a4d5c94e vhost_net: a kern... |
667 668 |
break; case VHOST_SET_VRING_ADDR: |
7ad9c9d27 vhost: fix to che... |
669 670 |
if (copy_from_user(&a, argp, sizeof a)) { r = -EFAULT; |
3a4d5c94e vhost_net: a kern... |
671 |
break; |
7ad9c9d27 vhost: fix to che... |
672 |
} |
3a4d5c94e vhost_net: a kern... |
673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 |
if (a.flags & ~(0x1 << VHOST_VRING_F_LOG)) { r = -EOPNOTSUPP; break; } /* For 32bit, verify that the top 32bits of the user data are set to zero. */ if ((u64)(unsigned long)a.desc_user_addr != a.desc_user_addr || (u64)(unsigned long)a.used_user_addr != a.used_user_addr || (u64)(unsigned long)a.avail_user_addr != a.avail_user_addr) { r = -EFAULT; break; } if ((a.avail_user_addr & (sizeof *vq->avail->ring - 1)) || (a.used_user_addr & (sizeof *vq->used->ring - 1)) || (a.log_guest_addr & (sizeof *vq->used->ring - 1))) { r = -EINVAL; break; } /* We only verify access here if backend is configured. * If it is not, we don't as size might not have been setup. * We will verify when backend is configured. */ if (vq->private_data) { |
8ea8cf89e vhost: support ev... |
696 |
if (!vq_access_ok(d, vq->num, |
3a4d5c94e vhost_net: a kern... |
697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 |
(void __user *)(unsigned long)a.desc_user_addr, (void __user *)(unsigned long)a.avail_user_addr, (void __user *)(unsigned long)a.used_user_addr)) { r = -EINVAL; break; } /* Also validate log access for used ring if enabled. */ if ((a.flags & (0x1 << VHOST_VRING_F_LOG)) && !log_access_ok(vq->log_base, a.log_guest_addr, sizeof *vq->used + vq->num * sizeof *vq->used->ring)) { r = -EINVAL; break; } } |
3a4d5c94e vhost_net: a kern... |
713 714 715 716 717 718 719 |
vq->log_used = !!(a.flags & (0x1 << VHOST_VRING_F_LOG)); vq->desc = (void __user *)(unsigned long)a.desc_user_addr; vq->avail = (void __user *)(unsigned long)a.avail_user_addr; vq->log_addr = a.log_guest_addr; vq->used = (void __user *)(unsigned long)a.used_user_addr; break; case VHOST_SET_VRING_KICK: |
7ad9c9d27 vhost: fix to che... |
720 721 |
if (copy_from_user(&f, argp, sizeof f)) { r = -EFAULT; |
3a4d5c94e vhost_net: a kern... |
722 |
break; |
7ad9c9d27 vhost: fix to che... |
723 |
} |
3a4d5c94e vhost_net: a kern... |
724 |
eventfp = f.fd == -1 ? NULL : eventfd_fget(f.fd); |
535297a6a vhost: fix error ... |
725 726 727 728 |
if (IS_ERR(eventfp)) { r = PTR_ERR(eventfp); break; } |
3a4d5c94e vhost_net: a kern... |
729 730 731 732 733 734 735 |
if (eventfp != vq->kick) { pollstop = filep = vq->kick; pollstart = vq->kick = eventfp; } else filep = eventfp; break; case VHOST_SET_VRING_CALL: |
7ad9c9d27 vhost: fix to che... |
736 737 |
if (copy_from_user(&f, argp, sizeof f)) { r = -EFAULT; |
3a4d5c94e vhost_net: a kern... |
738 |
break; |
7ad9c9d27 vhost: fix to che... |
739 |
} |
3a4d5c94e vhost_net: a kern... |
740 |
eventfp = f.fd == -1 ? NULL : eventfd_fget(f.fd); |
535297a6a vhost: fix error ... |
741 742 743 744 |
if (IS_ERR(eventfp)) { r = PTR_ERR(eventfp); break; } |
3a4d5c94e vhost_net: a kern... |
745 746 747 748 749 750 751 752 753 754 |
if (eventfp != vq->call) { filep = vq->call; ctx = vq->call_ctx; vq->call = eventfp; vq->call_ctx = eventfp ? eventfd_ctx_fileget(eventfp) : NULL; } else filep = eventfp; break; case VHOST_SET_VRING_ERR: |
7ad9c9d27 vhost: fix to che... |
755 756 |
if (copy_from_user(&f, argp, sizeof f)) { r = -EFAULT; |
3a4d5c94e vhost_net: a kern... |
757 |
break; |
7ad9c9d27 vhost: fix to che... |
758 |
} |
3a4d5c94e vhost_net: a kern... |
759 |
eventfp = f.fd == -1 ? NULL : eventfd_fget(f.fd); |
535297a6a vhost: fix error ... |
760 761 762 763 |
if (IS_ERR(eventfp)) { r = PTR_ERR(eventfp); break; } |
3a4d5c94e vhost_net: a kern... |
764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 |
if (eventfp != vq->error) { filep = vq->error; vq->error = eventfp; ctx = vq->error_ctx; vq->error_ctx = eventfp ? eventfd_ctx_fileget(eventfp) : NULL; } else filep = eventfp; break; default: r = -ENOIOCTLCMD; } if (pollstop && vq->handle_kick) vhost_poll_stop(&vq->poll); if (ctx) eventfd_ctx_put(ctx); if (filep) fput(filep); if (pollstart && vq->handle_kick) vhost_poll_start(&vq->poll, vq->kick); mutex_unlock(&vq->mutex); if (pollstop && vq->handle_kick) vhost_poll_flush(&vq->poll); return r; } /* Caller must have device mutex */ long vhost_dev_ioctl(struct vhost_dev *d, unsigned int ioctl, unsigned long arg) { void __user *argp = (void __user *)arg; struct file *eventfp, *filep = NULL; struct eventfd_ctx *ctx = NULL; u64 p; long r; int i, fd; /* If you are not the owner, you can become one */ if (ioctl == VHOST_SET_OWNER) { r = vhost_dev_set_owner(d); goto done; } /* You must be the owner to do anything else */ r = vhost_dev_check_owner(d); if (r) goto done; switch (ioctl) { case VHOST_SET_MEM_TABLE: r = vhost_set_memory(d, argp); break; case VHOST_SET_LOG_BASE: |
7ad9c9d27 vhost: fix to che... |
821 822 |
if (copy_from_user(&p, argp, sizeof p)) { r = -EFAULT; |
3a4d5c94e vhost_net: a kern... |
823 |
break; |
7ad9c9d27 vhost: fix to che... |
824 |
} |
3a4d5c94e vhost_net: a kern... |
825 826 827 828 829 830 831 832 833 834 |
if ((u64)(unsigned long)p != p) { r = -EFAULT; break; } for (i = 0; i < d->nvqs; ++i) { struct vhost_virtqueue *vq; void __user *base = (void __user *)(unsigned long)p; vq = d->vqs + i; mutex_lock(&vq->mutex); /* If ring is inactive, will check when it's enabled. */ |
8ea8cf89e vhost: support ev... |
835 |
if (vq->private_data && !vq_log_access_ok(d, vq, base)) |
3a4d5c94e vhost_net: a kern... |
836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 |
r = -EFAULT; else vq->log_base = base; mutex_unlock(&vq->mutex); } break; case VHOST_SET_LOG_FD: r = get_user(fd, (int __user *)argp); if (r < 0) break; eventfp = fd == -1 ? NULL : eventfd_fget(fd); if (IS_ERR(eventfp)) { r = PTR_ERR(eventfp); break; } if (eventfp != d->log_file) { filep = d->log_file; ctx = d->log_ctx; d->log_ctx = eventfp ? eventfd_ctx_fileget(eventfp) : NULL; } else filep = eventfp; for (i = 0; i < d->nvqs; ++i) { mutex_lock(&d->vqs[i].mutex); d->vqs[i].log_ctx = d->log_ctx; mutex_unlock(&d->vqs[i].mutex); } if (ctx) eventfd_ctx_put(ctx); if (filep) fput(filep); break; default: r = vhost_set_vring(d, ioctl, argp); break; } done: return r; } static const struct vhost_memory_region *find_region(struct vhost_memory *mem, __u64 addr, __u32 len) { struct vhost_memory_region *reg; int i; |
d47effe1b vhost: Cleanup vh... |
881 |
|
3a4d5c94e vhost_net: a kern... |
882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 |
/* linear search is not brilliant, but we really have on the order of 6 * regions in practice */ for (i = 0; i < mem->nregions; ++i) { reg = mem->regions + i; if (reg->guest_phys_addr <= addr && reg->guest_phys_addr + reg->memory_size - 1 >= addr) return reg; } return NULL; } /* TODO: This is really inefficient. We need something like get_user() * (instruction directly accesses the data, with an exception table entry * returning -EFAULT). See Documentation/x86/exception-tables.txt. */ static int set_bit_to_user(int nr, void __user *addr) { unsigned long log = (unsigned long)addr; struct page *page; void *base; int bit = nr + (log % PAGE_SIZE) * 8; int r; |
d47effe1b vhost: Cleanup vh... |
904 |
|
3a4d5c94e vhost_net: a kern... |
905 |
r = get_user_pages_fast(log, 1, 1, &page); |
d6db3f5c1 vhost: fix get_us... |
906 |
if (r < 0) |
3a4d5c94e vhost_net: a kern... |
907 |
return r; |
d6db3f5c1 vhost: fix get_us... |
908 |
BUG_ON(r != 1); |
3a4d5c94e vhost_net: a kern... |
909 910 911 912 913 914 915 916 917 918 919 |
base = kmap_atomic(page, KM_USER0); set_bit(bit, base); kunmap_atomic(base, KM_USER0); set_page_dirty_lock(page); put_page(page); return 0; } static int log_write(void __user *log_base, u64 write_address, u64 write_length) { |
28831ee60 vhost: better var... |
920 |
u64 write_page = write_address / VHOST_PAGE_SIZE; |
3a4d5c94e vhost_net: a kern... |
921 |
int r; |
d47effe1b vhost: Cleanup vh... |
922 |
|
3a4d5c94e vhost_net: a kern... |
923 924 |
if (!write_length) return 0; |
3bf9be40f vhost: correctly ... |
925 |
write_length += write_address % VHOST_PAGE_SIZE; |
3a4d5c94e vhost_net: a kern... |
926 927 |
for (;;) { u64 base = (u64)(unsigned long)log_base; |
28831ee60 vhost: better var... |
928 929 |
u64 log = base + write_page / 8; int bit = write_page % 8; |
3a4d5c94e vhost_net: a kern... |
930 931 932 933 934 935 936 937 |
if ((u64)(unsigned long)log != log) return -EFAULT; r = set_bit_to_user(bit, (void __user *)(unsigned long)log); if (r < 0) return r; if (write_length <= VHOST_PAGE_SIZE) break; write_length -= VHOST_PAGE_SIZE; |
28831ee60 vhost: better var... |
938 |
write_page += 1; |
3a4d5c94e vhost_net: a kern... |
939 940 941 942 943 944 945 946 947 948 |
} return r; } int vhost_log_write(struct vhost_virtqueue *vq, struct vhost_log *log, unsigned int log_num, u64 len) { int i, r; /* Make sure data written is seen before log. */ |
5659338c8 vhost-net: switch... |
949 |
smp_wmb(); |
3a4d5c94e vhost_net: a kern... |
950 951 952 953 954 955 |
for (i = 0; i < log_num; ++i) { u64 l = min(log[i].len, len); r = log_write(vq->log_base, log[i].addr, l); if (r < 0) return r; len -= l; |
5786aee8b vhost: fix log ct... |
956 957 958 |
if (!len) { if (vq->log_ctx) eventfd_signal(vq->log_ctx, 1); |
3a4d5c94e vhost_net: a kern... |
959 |
return 0; |
5786aee8b vhost: fix log ct... |
960 |
} |
3a4d5c94e vhost_net: a kern... |
961 |
} |
3a4d5c94e vhost_net: a kern... |
962 963 964 965 |
/* Length written exceeds what we have stored. This is a bug. */ BUG(); return 0; } |
2723feaa8 vhost: set log wh... |
966 967 968 |
static int vhost_update_used_flags(struct vhost_virtqueue *vq) { void __user *used; |
b834226b0 vhost: optimize i... |
969 |
if (__put_user(vq->used_flags, &vq->used->flags) < 0) |
2723feaa8 vhost: set log wh... |
970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 |
return -EFAULT; if (unlikely(vq->log_used)) { /* Make sure the flag is seen before log. */ smp_wmb(); /* Log used flag write. */ used = &vq->used->flags; log_write(vq->log_base, vq->log_addr + (used - (void __user *)vq->used), sizeof vq->used->flags); if (vq->log_ctx) eventfd_signal(vq->log_ctx, 1); } return 0; } static int vhost_update_avail_event(struct vhost_virtqueue *vq, u16 avail_event) { |
b834226b0 vhost: optimize i... |
987 |
if (__put_user(vq->avail_idx, vhost_avail_event(vq))) |
2723feaa8 vhost: set log wh... |
988 989 990 991 992 993 994 995 996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 |
return -EFAULT; if (unlikely(vq->log_used)) { void __user *used; /* Make sure the event is seen before log. */ smp_wmb(); /* Log avail event write */ used = vhost_avail_event(vq); log_write(vq->log_base, vq->log_addr + (used - (void __user *)vq->used), sizeof *vhost_avail_event(vq)); if (vq->log_ctx) eventfd_signal(vq->log_ctx, 1); } return 0; } int vhost_init_used(struct vhost_virtqueue *vq) { int r; if (!vq->private_data) return 0; r = vhost_update_used_flags(vq); if (r) return r; vq->signalled_used_valid = false; return get_user(vq->last_used_idx, &vq->used->idx); } |
a8d3782f9 vhost: fix sparse... |
1016 1017 |
static int translate_desc(struct vhost_dev *dev, u64 addr, u32 len, struct iovec iov[], int iov_size) |
3a4d5c94e vhost_net: a kern... |
1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 |
{ const struct vhost_memory_region *reg; struct vhost_memory *mem; struct iovec *_iov; u64 s = 0; int ret = 0; rcu_read_lock(); mem = rcu_dereference(dev->memory); while ((u64)len > s) { u64 size; |
7b3384fc3 vhost: add unlike... |
1030 |
if (unlikely(ret >= iov_size)) { |
3a4d5c94e vhost_net: a kern... |
1031 1032 1033 1034 |
ret = -ENOBUFS; break; } reg = find_region(mem, addr, len); |
7b3384fc3 vhost: add unlike... |
1035 |
if (unlikely(!reg)) { |
3a4d5c94e vhost_net: a kern... |
1036 1037 1038 1039 1040 1041 |
ret = -EFAULT; break; } _iov = iov + ret; size = reg->memory_size - addr + reg->guest_phys_addr; _iov->iov_len = min((u64)len, size); |
a8d3782f9 vhost: fix sparse... |
1042 |
_iov->iov_base = (void __user *)(unsigned long) |
3a4d5c94e vhost_net: a kern... |
1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 1072 |
(reg->userspace_addr + addr - reg->guest_phys_addr); s += size; addr += size; ++ret; } rcu_read_unlock(); return ret; } /* Each buffer in the virtqueues is actually a chain of descriptors. This * function returns the next descriptor in the chain, * or -1U if we're at the end. */ static unsigned next_desc(struct vring_desc *desc) { unsigned int next; /* If this descriptor says it doesn't chain, we're done. */ if (!(desc->flags & VRING_DESC_F_NEXT)) return -1U; /* Check they're not leading us off end of descriptors. */ next = desc->next; /* Make sure compiler knows to grab that: we don't want it changing! */ /* We will use the result as an index in an array, so most * architectures only need a compiler barrier here. */ read_barrier_depends(); return next; } |
7b3384fc3 vhost: add unlike... |
1073 1074 1075 1076 1077 |
static int get_indirect(struct vhost_dev *dev, struct vhost_virtqueue *vq, struct iovec iov[], unsigned int iov_size, unsigned int *out_num, unsigned int *in_num, struct vhost_log *log, unsigned int *log_num, struct vring_desc *indirect) |
3a4d5c94e vhost_net: a kern... |
1078 1079 1080 1081 1082 1083 |
{ struct vring_desc desc; unsigned int i = 0, count, found = 0; int ret; /* Sanity check */ |
7b3384fc3 vhost: add unlike... |
1084 |
if (unlikely(indirect->len % sizeof desc)) { |
3a4d5c94e vhost_net: a kern... |
1085 1086 1087 1088 1089 1090 1091 1092 1093 |
vq_err(vq, "Invalid length in indirect descriptor: " "len 0x%llx not multiple of 0x%zx ", (unsigned long long)indirect->len, sizeof desc); return -EINVAL; } ret = translate_desc(dev, indirect->addr, indirect->len, vq->indirect, |
e0e9b4064 vhost: max s/g to... |
1094 |
UIO_MAXIOV); |
7b3384fc3 vhost: add unlike... |
1095 |
if (unlikely(ret < 0)) { |
3a4d5c94e vhost_net: a kern... |
1096 1097 1098 1099 1100 1101 1102 1103 1104 1105 1106 1107 |
vq_err(vq, "Translation failure %d in indirect. ", ret); return ret; } /* We will use the result as an address to read from, so most * architectures only need a compiler barrier here. */ read_barrier_depends(); count = indirect->len / sizeof desc; /* Buffers are chained via a 16 bit next field, so * we can have at most 2^16 of these. */ |
7b3384fc3 vhost: add unlike... |
1108 |
if (unlikely(count > USHRT_MAX + 1)) { |
3a4d5c94e vhost_net: a kern... |
1109 1110 1111 1112 1113 1114 1115 1116 |
vq_err(vq, "Indirect buffer length too big: %d ", indirect->len); return -E2BIG; } do { unsigned iov_count = *in_num + *out_num; |
7b3384fc3 vhost: add unlike... |
1117 |
if (unlikely(++found > count)) { |
3a4d5c94e vhost_net: a kern... |
1118 1119 1120 1121 1122 1123 |
vq_err(vq, "Loop detected: last one at %u " "indirect size %u ", i, count); return -EINVAL; } |
d47effe1b vhost: Cleanup vh... |
1124 1125 |
if (unlikely(memcpy_fromiovec((unsigned char *)&desc, vq->indirect, sizeof desc))) { |
3a4d5c94e vhost_net: a kern... |
1126 1127 1128 1129 1130 |
vq_err(vq, "Failed indirect descriptor: idx %d, %zx ", i, (size_t)indirect->addr + i * sizeof desc); return -EINVAL; } |
7b3384fc3 vhost: add unlike... |
1131 |
if (unlikely(desc.flags & VRING_DESC_F_INDIRECT)) { |
3a4d5c94e vhost_net: a kern... |
1132 1133 1134 1135 1136 1137 1138 1139 |
vq_err(vq, "Nested indirect descriptor: idx %d, %zx ", i, (size_t)indirect->addr + i * sizeof desc); return -EINVAL; } ret = translate_desc(dev, desc.addr, desc.len, iov + iov_count, iov_size - iov_count); |
7b3384fc3 vhost: add unlike... |
1140 |
if (unlikely(ret < 0)) { |
3a4d5c94e vhost_net: a kern... |
1141 1142 1143 1144 1145 1146 1147 1148 1149 1150 1151 1152 1153 1154 1155 1156 |
vq_err(vq, "Translation failure %d indirect idx %d ", ret, i); return ret; } /* If this is an input descriptor, increment that count. */ if (desc.flags & VRING_DESC_F_WRITE) { *in_num += ret; if (unlikely(log)) { log[*log_num].addr = desc.addr; log[*log_num].len = desc.len; ++*log_num; } } else { /* If it's an output descriptor, they're all supposed * to come before any input descriptors. */ |
7b3384fc3 vhost: add unlike... |
1157 |
if (unlikely(*in_num)) { |
3a4d5c94e vhost_net: a kern... |
1158 1159 1160 1161 1162 1163 1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 |
vq_err(vq, "Indirect descriptor " "has out after in: idx %d ", i); return -EINVAL; } *out_num += ret; } } while ((i = next_desc(&desc)) != -1); return 0; } /* This looks in the virtqueue and for the first available buffer, and converts * it to an iovec for convenient access. Since descriptors consist of some * number of output then some number of input descriptors, it's actually two * iovecs, but we pack them into one and note how many of each there were. * |
d5675bd20 vhost: break out ... |
1174 1175 1176 1177 1178 1179 1180 |
* This function returns the descriptor number found, or vq->num (which is * never a valid descriptor number) if none was found. A negative code is * returned on error. */ int vhost_get_vq_desc(struct vhost_dev *dev, struct vhost_virtqueue *vq, struct iovec iov[], unsigned int iov_size, unsigned int *out_num, unsigned int *in_num, struct vhost_log *log, unsigned int *log_num) |
3a4d5c94e vhost_net: a kern... |
1181 1182 1183 1184 1185 1186 1187 1188 |
{ struct vring_desc desc; unsigned int i, head, found = 0; u16 last_avail_idx; int ret; /* Check it isn't doing very strange things with descriptor numbers. */ last_avail_idx = vq->last_avail_idx; |
8b7347aab vhost: get/put_us... |
1189 |
if (unlikely(__get_user(vq->avail_idx, &vq->avail->idx))) { |
3a4d5c94e vhost_net: a kern... |
1190 1191 1192 |
vq_err(vq, "Failed to access avail idx at %p ", &vq->avail->idx); |
d5675bd20 vhost: break out ... |
1193 |
return -EFAULT; |
3a4d5c94e vhost_net: a kern... |
1194 |
} |
7b3384fc3 vhost: add unlike... |
1195 |
if (unlikely((u16)(vq->avail_idx - last_avail_idx) > vq->num)) { |
3a4d5c94e vhost_net: a kern... |
1196 1197 |
vq_err(vq, "Guest moved used index from %u to %u", last_avail_idx, vq->avail_idx); |
d5675bd20 vhost: break out ... |
1198 |
return -EFAULT; |
3a4d5c94e vhost_net: a kern... |
1199 1200 1201 1202 1203 1204 1205 |
} /* If there's nothing new since last we looked, return invalid. */ if (vq->avail_idx == last_avail_idx) return vq->num; /* Only get avail ring entries after they have been exposed by guest. */ |
5659338c8 vhost-net: switch... |
1206 |
smp_rmb(); |
3a4d5c94e vhost_net: a kern... |
1207 1208 1209 |
/* Grab the next descriptor number they're advertising, and increment * the index we've seen. */ |
8b7347aab vhost: get/put_us... |
1210 1211 |
if (unlikely(__get_user(head, &vq->avail->ring[last_avail_idx % vq->num]))) { |
3a4d5c94e vhost_net: a kern... |
1212 1213 1214 1215 |
vq_err(vq, "Failed to read head: idx %d address %p ", last_avail_idx, &vq->avail->ring[last_avail_idx % vq->num]); |
d5675bd20 vhost: break out ... |
1216 |
return -EFAULT; |
3a4d5c94e vhost_net: a kern... |
1217 1218 1219 |
} /* If their number is silly, that's an error. */ |
7b3384fc3 vhost: add unlike... |
1220 |
if (unlikely(head >= vq->num)) { |
3a4d5c94e vhost_net: a kern... |
1221 1222 |
vq_err(vq, "Guest says index %u > %u is available", head, vq->num); |
d5675bd20 vhost: break out ... |
1223 |
return -EINVAL; |
3a4d5c94e vhost_net: a kern... |
1224 1225 1226 1227 1228 1229 1230 1231 1232 1233 |
} /* When we start there are none of either input nor output. */ *out_num = *in_num = 0; if (unlikely(log)) *log_num = 0; i = head; do { unsigned iov_count = *in_num + *out_num; |
7b3384fc3 vhost: add unlike... |
1234 |
if (unlikely(i >= vq->num)) { |
3a4d5c94e vhost_net: a kern... |
1235 1236 |
vq_err(vq, "Desc index is %u > %u, head = %u", i, vq->num, head); |
d5675bd20 vhost: break out ... |
1237 |
return -EINVAL; |
3a4d5c94e vhost_net: a kern... |
1238 |
} |
7b3384fc3 vhost: add unlike... |
1239 |
if (unlikely(++found > vq->num)) { |
3a4d5c94e vhost_net: a kern... |
1240 1241 1242 1243 |
vq_err(vq, "Loop detected: last one at %u " "vq size %u head %u ", i, vq->num, head); |
d5675bd20 vhost: break out ... |
1244 |
return -EINVAL; |
3a4d5c94e vhost_net: a kern... |
1245 |
} |
fcc042a28 vhost: copy_from_... |
1246 |
ret = __copy_from_user(&desc, vq->desc + i, sizeof desc); |
7b3384fc3 vhost: add unlike... |
1247 |
if (unlikely(ret)) { |
3a4d5c94e vhost_net: a kern... |
1248 1249 1250 |
vq_err(vq, "Failed to get descriptor: idx %d addr %p ", i, vq->desc + i); |
d5675bd20 vhost: break out ... |
1251 |
return -EFAULT; |
3a4d5c94e vhost_net: a kern... |
1252 1253 1254 1255 1256 |
} if (desc.flags & VRING_DESC_F_INDIRECT) { ret = get_indirect(dev, vq, iov, iov_size, out_num, in_num, log, log_num, &desc); |
7b3384fc3 vhost: add unlike... |
1257 |
if (unlikely(ret < 0)) { |
3a4d5c94e vhost_net: a kern... |
1258 1259 1260 |
vq_err(vq, "Failure detected " "in indirect descriptor at idx %d ", i); |
d5675bd20 vhost: break out ... |
1261 |
return ret; |
3a4d5c94e vhost_net: a kern... |
1262 1263 1264 1265 1266 1267 |
} continue; } ret = translate_desc(dev, desc.addr, desc.len, iov + iov_count, iov_size - iov_count); |
7b3384fc3 vhost: add unlike... |
1268 |
if (unlikely(ret < 0)) { |
3a4d5c94e vhost_net: a kern... |
1269 1270 1271 |
vq_err(vq, "Translation failure %d descriptor idx %d ", ret, i); |
d5675bd20 vhost: break out ... |
1272 |
return ret; |
3a4d5c94e vhost_net: a kern... |
1273 1274 1275 1276 1277 1278 1279 1280 1281 1282 1283 1284 1285 |
} if (desc.flags & VRING_DESC_F_WRITE) { /* If this is an input descriptor, * increment that count. */ *in_num += ret; if (unlikely(log)) { log[*log_num].addr = desc.addr; log[*log_num].len = desc.len; ++*log_num; } } else { /* If it's an output descriptor, they're all supposed * to come before any input descriptors. */ |
7b3384fc3 vhost: add unlike... |
1286 |
if (unlikely(*in_num)) { |
3a4d5c94e vhost_net: a kern... |
1287 1288 1289 |
vq_err(vq, "Descriptor has out after in: " "idx %d ", i); |
d5675bd20 vhost: break out ... |
1290 |
return -EINVAL; |
3a4d5c94e vhost_net: a kern... |
1291 1292 1293 1294 1295 1296 1297 |
} *out_num += ret; } } while ((i = next_desc(&desc)) != -1); /* On success, increment avail index. */ vq->last_avail_idx++; |
8ea8cf89e vhost: support ev... |
1298 1299 1300 1301 |
/* Assume notifications from guest are disabled at this point, * if they aren't we would need to update avail_event index. */ BUG_ON(!(vq->used_flags & VRING_USED_F_NO_NOTIFY)); |
3a4d5c94e vhost_net: a kern... |
1302 1303 1304 1305 |
return head; } /* Reverse the effect of vhost_get_vq_desc. Useful for error handling. */ |
8dd014adf vhost-net: mergea... |
1306 |
void vhost_discard_vq_desc(struct vhost_virtqueue *vq, int n) |
3a4d5c94e vhost_net: a kern... |
1307 |
{ |
8dd014adf vhost-net: mergea... |
1308 |
vq->last_avail_idx -= n; |
3a4d5c94e vhost_net: a kern... |
1309 1310 1311 1312 1313 1314 |
} /* After we've used one of their buffers, we tell them about it. We'll then * want to notify the guest, using eventfd. */ int vhost_add_used(struct vhost_virtqueue *vq, unsigned int head, int len) { |
a8d3782f9 vhost: fix sparse... |
1315 |
struct vring_used_elem __user *used; |
3a4d5c94e vhost_net: a kern... |
1316 1317 1318 1319 |
/* The virtqueue contains a ring of used buffers. Get a pointer to the * next entry in that used ring. */ used = &vq->used->ring[vq->last_used_idx % vq->num]; |
8b7347aab vhost: get/put_us... |
1320 |
if (__put_user(head, &used->id)) { |
3a4d5c94e vhost_net: a kern... |
1321 1322 1323 |
vq_err(vq, "Failed to write used id"); return -EFAULT; } |
8b7347aab vhost: get/put_us... |
1324 |
if (__put_user(len, &used->len)) { |
3a4d5c94e vhost_net: a kern... |
1325 1326 1327 1328 |
vq_err(vq, "Failed to write used len"); return -EFAULT; } /* Make sure buffer is written before we update index. */ |
5659338c8 vhost-net: switch... |
1329 |
smp_wmb(); |
8b7347aab vhost: get/put_us... |
1330 |
if (__put_user(vq->last_used_idx + 1, &vq->used->idx)) { |
3a4d5c94e vhost_net: a kern... |
1331 1332 1333 1334 1335 |
vq_err(vq, "Failed to increment used idx"); return -EFAULT; } if (unlikely(vq->log_used)) { /* Make sure data is seen before log. */ |
5659338c8 vhost-net: switch... |
1336 |
smp_wmb(); |
86e9424d7 vhost: logging th... |
1337 1338 |
/* Log used ring entry write. */ log_write(vq->log_base, |
a8d3782f9 vhost: fix sparse... |
1339 1340 |
vq->log_addr + ((void __user *)used - (void __user *)vq->used), |
86e9424d7 vhost: logging th... |
1341 1342 1343 1344 1345 |
sizeof *used); /* Log used index update. */ log_write(vq->log_base, vq->log_addr + offsetof(struct vring_used, idx), sizeof vq->used->idx); |
3a4d5c94e vhost_net: a kern... |
1346 1347 1348 1349 |
if (vq->log_ctx) eventfd_signal(vq->log_ctx, 1); } vq->last_used_idx++; |
8ea8cf89e vhost: support ev... |
1350 1351 1352 1353 1354 1355 |
/* If the driver never bothers to signal in a very long while, * used index might wrap around. If that happens, invalidate * signalled_used index we stored. TODO: make sure driver * signals at least once in 2^16 and remove this. */ if (unlikely(vq->last_used_idx == vq->signalled_used)) vq->signalled_used_valid = false; |
3a4d5c94e vhost_net: a kern... |
1356 1357 |
return 0; } |
8dd014adf vhost-net: mergea... |
1358 1359 1360 1361 1362 |
static int __vhost_add_used_n(struct vhost_virtqueue *vq, struct vring_used_elem *heads, unsigned count) { struct vring_used_elem __user *used; |
8ea8cf89e vhost: support ev... |
1363 |
u16 old, new; |
8dd014adf vhost-net: mergea... |
1364 1365 1366 1367 |
int start; start = vq->last_used_idx % vq->num; used = vq->used->ring + start; |
dfe5ac5b1 vhost: copy_to_us... |
1368 |
if (__copy_to_user(used, heads, count * sizeof *used)) { |
8dd014adf vhost-net: mergea... |
1369 1370 1371 1372 1373 1374 1375 1376 1377 1378 1379 1380 |
vq_err(vq, "Failed to write used"); return -EFAULT; } if (unlikely(vq->log_used)) { /* Make sure data is seen before log. */ smp_wmb(); /* Log used ring entry write. */ log_write(vq->log_base, vq->log_addr + ((void __user *)used - (void __user *)vq->used), count * sizeof *used); } |
8ea8cf89e vhost: support ev... |
1381 1382 1383 1384 1385 1386 1387 1388 |
old = vq->last_used_idx; new = (vq->last_used_idx += count); /* If the driver never bothers to signal in a very long while, * used index might wrap around. If that happens, invalidate * signalled_used index we stored. TODO: make sure driver * signals at least once in 2^16 and remove this. */ if (unlikely((u16)(new - vq->signalled_used) < (u16)(new - old))) vq->signalled_used_valid = false; |
8dd014adf vhost-net: mergea... |
1389 1390 1391 1392 1393 1394 1395 1396 1397 1398 1399 1400 1401 1402 1403 1404 1405 1406 1407 1408 1409 1410 1411 1412 1413 1414 1415 1416 1417 1418 1419 1420 1421 1422 1423 1424 1425 |
return 0; } /* After we've used one of their buffers, we tell them about it. We'll then * want to notify the guest, using eventfd. */ int vhost_add_used_n(struct vhost_virtqueue *vq, struct vring_used_elem *heads, unsigned count) { int start, n, r; start = vq->last_used_idx % vq->num; n = vq->num - start; if (n < count) { r = __vhost_add_used_n(vq, heads, n); if (r < 0) return r; heads += n; count -= n; } r = __vhost_add_used_n(vq, heads, count); /* Make sure buffer is written before we update index. */ smp_wmb(); if (put_user(vq->last_used_idx, &vq->used->idx)) { vq_err(vq, "Failed to increment used idx"); return -EFAULT; } if (unlikely(vq->log_used)) { /* Log used index update. */ log_write(vq->log_base, vq->log_addr + offsetof(struct vring_used, idx), sizeof vq->used->idx); if (vq->log_ctx) eventfd_signal(vq->log_ctx, 1); } return r; } |
8ea8cf89e vhost: support ev... |
1426 |
static bool vhost_notify(struct vhost_dev *dev, struct vhost_virtqueue *vq) |
3a4d5c94e vhost_net: a kern... |
1427 |
{ |
8ea8cf89e vhost: support ev... |
1428 1429 |
__u16 old, new, event; bool v; |
0d4993563 vhost: fix barrie... |
1430 1431 1432 1433 |
/* Flush out used index updates. This is paired * with the barrier that the Guest executes when enabling * interrupts. */ smp_mb(); |
8ea8cf89e vhost: support ev... |
1434 1435 1436 1437 1438 1439 1440 1441 1442 1443 1444 |
if (vhost_has_feature(dev, VIRTIO_F_NOTIFY_ON_EMPTY) && unlikely(vq->avail_idx == vq->last_avail_idx)) return true; if (!vhost_has_feature(dev, VIRTIO_RING_F_EVENT_IDX)) { __u16 flags; if (__get_user(flags, &vq->avail->flags)) { vq_err(vq, "Failed to get flags"); return true; } return !(flags & VRING_AVAIL_F_NO_INTERRUPT); |
3a4d5c94e vhost_net: a kern... |
1445 |
} |
8ea8cf89e vhost: support ev... |
1446 1447 1448 1449 |
old = vq->signalled_used; v = vq->signalled_used_valid; new = vq->signalled_used = vq->last_used_idx; vq->signalled_used_valid = true; |
3a4d5c94e vhost_net: a kern... |
1450 |
|
8ea8cf89e vhost: support ev... |
1451 1452 |
if (unlikely(!v)) return true; |
3a4d5c94e vhost_net: a kern... |
1453 |
|
8ea8cf89e vhost: support ev... |
1454 1455 1456 1457 1458 1459 1460 1461 1462 1463 |
if (get_user(event, vhost_used_event(vq))) { vq_err(vq, "Failed to get used event idx"); return true; } return vring_need_event(event, new, old); } /* This actually signals the guest, using eventfd. */ void vhost_signal(struct vhost_dev *dev, struct vhost_virtqueue *vq) { |
3a4d5c94e vhost_net: a kern... |
1464 |
/* Signal the Guest tell them we used something up. */ |
8ea8cf89e vhost: support ev... |
1465 |
if (vq->call_ctx && vhost_notify(dev, vq)) |
3a4d5c94e vhost_net: a kern... |
1466 1467 1468 1469 1470 1471 1472 1473 1474 1475 1476 |
eventfd_signal(vq->call_ctx, 1); } /* And here's the combo meal deal. Supersize me! */ void vhost_add_used_and_signal(struct vhost_dev *dev, struct vhost_virtqueue *vq, unsigned int head, int len) { vhost_add_used(vq, head, len); vhost_signal(dev, vq); } |
8dd014adf vhost-net: mergea... |
1477 1478 1479 1480 1481 1482 1483 1484 |
/* multi-buffer version of vhost_add_used_and_signal */ void vhost_add_used_and_signal_n(struct vhost_dev *dev, struct vhost_virtqueue *vq, struct vring_used_elem *heads, unsigned count) { vhost_add_used_n(vq, heads, count); vhost_signal(dev, vq); } |
3a4d5c94e vhost_net: a kern... |
1485 |
/* OK, now we need to know about added descriptors. */ |
8ea8cf89e vhost: support ev... |
1486 |
bool vhost_enable_notify(struct vhost_dev *dev, struct vhost_virtqueue *vq) |
3a4d5c94e vhost_net: a kern... |
1487 1488 1489 |
{ u16 avail_idx; int r; |
d47effe1b vhost: Cleanup vh... |
1490 |
|
3a4d5c94e vhost_net: a kern... |
1491 1492 1493 |
if (!(vq->used_flags & VRING_USED_F_NO_NOTIFY)) return false; vq->used_flags &= ~VRING_USED_F_NO_NOTIFY; |
8ea8cf89e vhost: support ev... |
1494 |
if (!vhost_has_feature(dev, VIRTIO_RING_F_EVENT_IDX)) { |
2723feaa8 vhost: set log wh... |
1495 |
r = vhost_update_used_flags(vq); |
8ea8cf89e vhost: support ev... |
1496 1497 1498 1499 1500 1501 1502 |
if (r) { vq_err(vq, "Failed to enable notification at %p: %d ", &vq->used->flags, r); return false; } } else { |
2723feaa8 vhost: set log wh... |
1503 |
r = vhost_update_avail_event(vq, vq->avail_idx); |
8ea8cf89e vhost: support ev... |
1504 1505 1506 1507 1508 1509 1510 |
if (r) { vq_err(vq, "Failed to update avail event index at %p: %d ", vhost_avail_event(vq), r); return false; } } |
3a4d5c94e vhost_net: a kern... |
1511 1512 |
/* They could have slipped one in as we were doing that: make * sure it's written, then check again. */ |
5659338c8 vhost-net: switch... |
1513 |
smp_mb(); |
8b7347aab vhost: get/put_us... |
1514 |
r = __get_user(avail_idx, &vq->avail->idx); |
3a4d5c94e vhost_net: a kern... |
1515 1516 1517 1518 1519 1520 |
if (r) { vq_err(vq, "Failed to check avail idx at %p: %d ", &vq->avail->idx, r); return false; } |
8dd014adf vhost-net: mergea... |
1521 |
return avail_idx != vq->avail_idx; |
3a4d5c94e vhost_net: a kern... |
1522 1523 1524 |
} /* We don't need to be notified again. */ |
8ea8cf89e vhost: support ev... |
1525 |
void vhost_disable_notify(struct vhost_dev *dev, struct vhost_virtqueue *vq) |
3a4d5c94e vhost_net: a kern... |
1526 1527 |
{ int r; |
d47effe1b vhost: Cleanup vh... |
1528 |
|
3a4d5c94e vhost_net: a kern... |
1529 1530 1531 |
if (vq->used_flags & VRING_USED_F_NO_NOTIFY) return; vq->used_flags |= VRING_USED_F_NO_NOTIFY; |
8ea8cf89e vhost: support ev... |
1532 |
if (!vhost_has_feature(dev, VIRTIO_RING_F_EVENT_IDX)) { |
2723feaa8 vhost: set log wh... |
1533 |
r = vhost_update_used_flags(vq); |
8ea8cf89e vhost: support ev... |
1534 1535 1536 1537 1538 |
if (r) vq_err(vq, "Failed to enable notification at %p: %d ", &vq->used->flags, r); } |
3a4d5c94e vhost_net: a kern... |
1539 |
} |
bab632d69 vhost: vhost TX z... |
1540 1541 1542 1543 1544 1545 1546 1547 1548 1549 1550 1551 1552 1553 1554 1555 1556 1557 1558 |
static void vhost_zerocopy_done_signal(struct kref *kref) { struct vhost_ubuf_ref *ubufs = container_of(kref, struct vhost_ubuf_ref, kref); wake_up(&ubufs->wait); } struct vhost_ubuf_ref *vhost_ubuf_alloc(struct vhost_virtqueue *vq, bool zcopy) { struct vhost_ubuf_ref *ubufs; /* No zero copy backend? Nothing to count. */ if (!zcopy) return NULL; ubufs = kmalloc(sizeof *ubufs, GFP_KERNEL); if (!ubufs) return ERR_PTR(-ENOMEM); kref_init(&ubufs->kref); |
bab632d69 vhost: vhost TX z... |
1559 1560 1561 1562 1563 1564 1565 1566 1567 1568 1569 1570 1571 1572 1573 1574 1575 1576 1577 1578 1579 1580 1581 1582 1583 1584 1585 |
init_waitqueue_head(&ubufs->wait); ubufs->vq = vq; return ubufs; } void vhost_ubuf_put(struct vhost_ubuf_ref *ubufs) { kref_put(&ubufs->kref, vhost_zerocopy_done_signal); } void vhost_ubuf_put_and_wait(struct vhost_ubuf_ref *ubufs) { kref_put(&ubufs->kref, vhost_zerocopy_done_signal); wait_event(ubufs->wait, !atomic_read(&ubufs->kref.refcount)); kfree(ubufs); } void vhost_zerocopy_callback(void *arg) { struct ubuf_info *ubuf = arg; struct vhost_ubuf_ref *ubufs = ubuf->arg; struct vhost_virtqueue *vq = ubufs->vq; /* set len = 1 to mark this desc buffers done DMA */ vq->heads[ubuf->desc].len = VHOST_DMA_DONE_LEN; kref_put(&ubufs->kref, vhost_zerocopy_done_signal); } |