drivers/vhost/vdpa.c

   1 // SPDX-License-Identifier: GPL-2.0
   2 /*
   3  * Copyright (C) 2018-2020 Intel Corporation.
   4  * Copyright (C) 2020 Red Hat, Inc.
   5  *
   6  * Author: Tiwei Bie <tiwei.bie@intel.com>
   7  *         Jason Wang <jasowang@redhat.com>
   8  *
   9  * Thanks Michael S. Tsirkin for the valuable comments and
  10  * suggestions.  And thanks to Cunming Liang and Zhihong Wang for all
  11  * their supports.
  12  */
  13
  14 #include <linux/kernel.h>
  15 #include <linux/module.h>
  16 #include <linux/cdev.h>
  17 #include <linux/device.h>
  18 #include <linux/mm.h>
  19 #include <linux/slab.h>
  20 #include <linux/iommu.h>
  21 #include <linux/uuid.h>
  22 #include <linux/vdpa.h>
  23 #include <linux/nospec.h>
  24 #include <linux/vhost.h>
  25
  26 #include "vhost.h"
  27
  28 enum {
  29         VHOST_VDPA_BACKEND_FEATURES =
  30         (1ULL << VHOST_BACKEND_F_IOTLB_MSG_V2) |
  31         (1ULL << VHOST_BACKEND_F_IOTLB_BATCH),
  32 };
  33
  34 #define VHOST_VDPA_DEV_MAX (1U << MINORBITS)
  35
  36 struct vhost_vdpa {
  37         struct vhost_dev vdev;
  38         struct iommu_domain *domain;
  39         struct vhost_virtqueue *vqs;
  40         struct completion completion;
  41         struct vdpa_device *vdpa;
  42         struct device dev;
  43         struct cdev cdev;
  44         atomic_t opened;
  45         int nvqs;
  46         int virtio_id;
  47         int minor;
  48         struct eventfd_ctx *config_ctx;
  49         int in_batch;
  50         struct vdpa_iova_range range;
  51 };
  52
  53 static DEFINE_IDA(vhost_vdpa_ida);
  54
  55 static dev_t vhost_vdpa_major;
  56
  57 static void handle_vq_kick(struct vhost_work *work)
  58 {
  59         struct vhost_virtqueue *vq = container_of(work, struct vhost_virtqueue,
  60                                                   poll.work);
  61         struct vhost_vdpa *v = container_of(vq->dev, struct vhost_vdpa, vdev);
  62         const struct vdpa_config_ops *ops = v->vdpa->config;
  63
  64         ops->kick_vq(v->vdpa, vq - v->vqs);
  65 }
  66
  67 static irqreturn_t vhost_vdpa_virtqueue_cb(void *private)
  68 {
  69         struct vhost_virtqueue *vq = private;
  70         struct eventfd_ctx *call_ctx = vq->call_ctx.ctx;
  71
  72         if (call_ctx)
  73                 eventfd_signal(call_ctx, 1);
  74
  75         return IRQ_HANDLED;
  76 }
  77
  78 static irqreturn_t vhost_vdpa_config_cb(void *private)
  79 {
  80         struct vhost_vdpa *v = private;
  81         struct eventfd_ctx *config_ctx = v->config_ctx;
  82
  83         if (config_ctx)
  84                 eventfd_signal(config_ctx, 1);
  85
  86         return IRQ_HANDLED;
  87 }
  88
  89 static void vhost_vdpa_setup_vq_irq(struct vhost_vdpa *v, u16 qid)
  90 {
  91         struct vhost_virtqueue *vq = &v->vqs[qid];
  92         const struct vdpa_config_ops *ops = v->vdpa->config;
  93         struct vdpa_device *vdpa = v->vdpa;
  94         int ret, irq;
  95
  96         if (!ops->get_vq_irq)
  97                 return;
  98
  99         irq = ops->get_vq_irq(vdpa, qid);
 100         irq_bypass_unregister_producer(&vq->call_ctx.producer);
 101         if (!vq->call_ctx.ctx || irq < 0)
 102                 return;
 103
 104         vq->call_ctx.producer.token = vq->call_ctx.ctx;
 105         vq->call_ctx.producer.irq = irq;
 106         ret = irq_bypass_register_producer(&vq->call_ctx.producer);
 107         if (unlikely(ret))
 108                 dev_info(&v->dev, "vq %u, irq bypass producer (token %p) registration fails, ret =  %d\n",
 109                          qid, vq->call_ctx.producer.token, ret);
 110 }
 111
 112 static void vhost_vdpa_unsetup_vq_irq(struct vhost_vdpa *v, u16 qid)
 113 {
 114         struct vhost_virtqueue *vq = &v->vqs[qid];
 115
 116         irq_bypass_unregister_producer(&vq->call_ctx.producer);
 117 }
 118
 119 static int vhost_vdpa_reset(struct vhost_vdpa *v)
 120 {
 121         struct vdpa_device *vdpa = v->vdpa;
 122
 123         v->in_batch = 0;
 124
 125         return vdpa_reset(vdpa);
 126 }
 127
 128 static long vhost_vdpa_get_device_id(struct vhost_vdpa *v, u8 __user *argp)
 129 {
 130         struct vdpa_device *vdpa = v->vdpa;
 131         const struct vdpa_config_ops *ops = vdpa->config;
 132         u32 device_id;
 133
 134         device_id = ops->get_device_id(vdpa);
 135
 136         if (copy_to_user(argp, &device_id, sizeof(device_id)))
 137                 return -EFAULT;
 138
 139         return 0;
 140 }
 141
 142 static long vhost_vdpa_get_status(struct vhost_vdpa *v, u8 __user *statusp)
 143 {
 144         struct vdpa_device *vdpa = v->vdpa;
 145         const struct vdpa_config_ops *ops = vdpa->config;
 146         u8 status;
 147
 148         status = ops->get_status(vdpa);
 149
 150         if (copy_to_user(statusp, &status, sizeof(status)))
 151                 return -EFAULT;
 152
 153         return 0;
 154 }
 155
 156 static long vhost_vdpa_set_status(struct vhost_vdpa *v, u8 __user *statusp)
 157 {
 158         struct vdpa_device *vdpa = v->vdpa;
 159         const struct vdpa_config_ops *ops = vdpa->config;
 160         u8 status, status_old;
 161         int ret, nvqs = v->nvqs;
 162         u16 i;
 163
 164         if (copy_from_user(&status, statusp, sizeof(status)))
 165                 return -EFAULT;
 166
 167         status_old = ops->get_status(vdpa);
 168
 169         /*
 170          * Userspace shouldn't remove status bits unless reset the
 171          * status to 0.
 172          */
 173         if (status != 0 && (status_old & ~status) != 0)
 174                 return -EINVAL;
 175
 176         if ((status_old & VIRTIO_CONFIG_S_DRIVER_OK) && !(status & VIRTIO_CONFIG_S_DRIVER_OK))
 177                 for (i = 0; i < nvqs; i++)
 178                         vhost_vdpa_unsetup_vq_irq(v, i);
 179
 180         if (status == 0) {
 181                 ret = vdpa_reset(vdpa);
 182                 if (ret)
 183                         return ret;
 184         } else
 185                 vdpa_set_status(vdpa, status);
 186
 187         if ((status & VIRTIO_CONFIG_S_DRIVER_OK) && !(status_old & VIRTIO_CONFIG_S_DRIVER_OK))
 188                 for (i = 0; i < nvqs; i++)
 189                         vhost_vdpa_setup_vq_irq(v, i);
 190
 191         return 0;
 192 }
 193
 194 static int vhost_vdpa_config_validate(struct vhost_vdpa *v,
 195                                       struct vhost_vdpa_config *c)
 196 {
 197         struct vdpa_device *vdpa = v->vdpa;
 198         size_t size = vdpa->config->get_config_size(vdpa);
 199
 200         if (c->len == 0 || c->off > size)
 201                 return -EINVAL;
 202
 203         if (c->len > size - c->off)
 204                 return -E2BIG;
 205
 206         return 0;
 207 }
 208
 209 static long vhost_vdpa_get_config(struct vhost_vdpa *v,
 210                                   struct vhost_vdpa_config __user *c)
 211 {
 212         struct vdpa_device *vdpa = v->vdpa;
 213         struct vhost_vdpa_config config;
 214         unsigned long size = offsetof(struct vhost_vdpa_config, buf);
 215         u8 *buf;
 216
 217         if (copy_from_user(&config, c, size))
 218                 return -EFAULT;
 219         if (vhost_vdpa_config_validate(v, &config))
 220                 return -EINVAL;
 221         buf = kvzalloc(config.len, GFP_KERNEL);
 222         if (!buf)
 223                 return -ENOMEM;
 224
 225         vdpa_get_config(vdpa, config.off, buf, config.len);
 226
 227         if (copy_to_user(c->buf, buf, config.len)) {
 228                 kvfree(buf);
 229                 return -EFAULT;
 230         }
 231
 232         kvfree(buf);
 233         return 0;
 234 }
 235
 236 static long vhost_vdpa_set_config(struct vhost_vdpa *v,
 237                                   struct vhost_vdpa_config __user *c)
 238 {
 239         struct vdpa_device *vdpa = v->vdpa;
 240         struct vhost_vdpa_config config;
 241         unsigned long size = offsetof(struct vhost_vdpa_config, buf);
 242         u8 *buf;
 243
 244         if (copy_from_user(&config, c, size))
 245                 return -EFAULT;
 246         if (vhost_vdpa_config_validate(v, &config))
 247                 return -EINVAL;
 248
 249         buf = vmemdup_user(c->buf, config.len);
 250         if (IS_ERR(buf))
 251                 return PTR_ERR(buf);
 252
 253         vdpa_set_config(vdpa, config.off, buf, config.len);
 254
 255         kvfree(buf);
 256         return 0;
 257 }
 258
 259 static long vhost_vdpa_get_features(struct vhost_vdpa *v, u64 __user *featurep)
 260 {
 261         struct vdpa_device *vdpa = v->vdpa;
 262         const struct vdpa_config_ops *ops = vdpa->config;
 263         u64 features;
 264
 265         features = ops->get_device_features(vdpa);
 266
 267         if (copy_to_user(featurep, &features, sizeof(features)))
 268                 return -EFAULT;
 269
 270         return 0;
 271 }
 272
 273 static long vhost_vdpa_set_features(struct vhost_vdpa *v, u64 __user *featurep)
 274 {
 275         struct vdpa_device *vdpa = v->vdpa;
 276         const struct vdpa_config_ops *ops = vdpa->config;
 277         u64 features;
 278
 279         /*
 280          * It's not allowed to change the features after they have
 281          * been negotiated.
 282          */
 283         if (ops->get_status(vdpa) & VIRTIO_CONFIG_S_FEATURES_OK)
 284                 return -EBUSY;
 285
 286         if (copy_from_user(&features, featurep, sizeof(features)))
 287                 return -EFAULT;
 288
 289         if (vdpa_set_features(vdpa, features, false))
 290                 return -EINVAL;
 291
 292         return 0;
 293 }
 294
 295 static long vhost_vdpa_get_vring_num(struct vhost_vdpa *v, u16 __user *argp)
 296 {
 297         struct vdpa_device *vdpa = v->vdpa;
 298         const struct vdpa_config_ops *ops = vdpa->config;
 299         u16 num;
 300
 301         num = ops->get_vq_num_max(vdpa);
 302
 303         if (copy_to_user(argp, &num, sizeof(num)))
 304                 return -EFAULT;
 305
 306         return 0;
 307 }
 308
 309 static void vhost_vdpa_config_put(struct vhost_vdpa *v)
 310 {
 311         if (v->config_ctx) {
 312                 eventfd_ctx_put(v->config_ctx);
 313                 v->config_ctx = NULL;
 314         }
 315 }
 316
 317 static long vhost_vdpa_set_config_call(struct vhost_vdpa *v, u32 __user *argp)
 318 {
 319         struct vdpa_callback cb;
 320         int fd;
 321         struct eventfd_ctx *ctx;
 322
 323         cb.callback = vhost_vdpa_config_cb;
 324         cb.private = v;
 325         if (copy_from_user(&fd, argp, sizeof(fd)))
 326                 return  -EFAULT;
 327
 328         ctx = fd == VHOST_FILE_UNBIND ? NULL : eventfd_ctx_fdget(fd);
 329         swap(ctx, v->config_ctx);
 330
 331         if (!IS_ERR_OR_NULL(ctx))
 332                 eventfd_ctx_put(ctx);
 333
 334         if (IS_ERR(v->config_ctx)) {
 335                 long ret = PTR_ERR(v->config_ctx);
 336
 337                 v->config_ctx = NULL;
 338                 return ret;
 339         }
 340
 341         v->vdpa->config->set_config_cb(v->vdpa, &cb);
 342
 343         return 0;
 344 }
 345
 346 static long vhost_vdpa_get_iova_range(struct vhost_vdpa *v, u32 __user *argp)
 347 {
 348         struct vhost_vdpa_iova_range range = {
 349                 .first = v->range.first,
 350                 .last = v->range.last,
 351         };
 352
 353         if (copy_to_user(argp, &range, sizeof(range)))
 354                 return -EFAULT;
 355         return 0;
 356 }
 357
 358 static long vhost_vdpa_vring_ioctl(struct vhost_vdpa *v, unsigned int cmd,
 359                                    void __user *argp)
 360 {
 361         struct vdpa_device *vdpa = v->vdpa;
 362         const struct vdpa_config_ops *ops = vdpa->config;
 363         struct vdpa_vq_state vq_state;
 364         struct vdpa_callback cb;
 365         struct vhost_virtqueue *vq;
 366         struct vhost_vring_state s;
 367         u32 idx;
 368         long r;
 369
 370         r = get_user(idx, (u32 __user *)argp);
 371         if (r < 0)
 372                 return r;
 373
 374         if (idx >= v->nvqs)
 375                 return -ENOBUFS;
 376
 377         idx = array_index_nospec(idx, v->nvqs);
 378         vq = &v->vqs[idx];
 379
 380         switch (cmd) {
 381         case VHOST_VDPA_SET_VRING_ENABLE:
 382                 if (copy_from_user(&s, argp, sizeof(s)))
 383                         return -EFAULT;
 384                 ops->set_vq_ready(vdpa, idx, s.num);
 385                 return 0;
 386         case VHOST_GET_VRING_BASE:
 387                 r = ops->get_vq_state(v->vdpa, idx, &vq_state);
 388                 if (r)
 389                         return r;
 390
 391                 vq->last_avail_idx = vq_state.split.avail_index;
 392                 break;
 393         }
 394
 395         r = vhost_vring_ioctl(&v->vdev, cmd, argp);
 396         if (r)
 397                 return r;
 398
 399         switch (cmd) {
 400         case VHOST_SET_VRING_ADDR:
 401                 if (ops->set_vq_address(vdpa, idx,
 402                                         (u64)(uintptr_t)vq->desc,
 403                                         (u64)(uintptr_t)vq->avail,
 404                                         (u64)(uintptr_t)vq->used))
 405                         r = -EINVAL;
 406                 break;
 407
 408         case VHOST_SET_VRING_BASE:
 409                 vq_state.split.avail_index = vq->last_avail_idx;
 410                 if (ops->set_vq_state(vdpa, idx, &vq_state))
 411                         r = -EINVAL;
 412                 break;
 413
 414         case VHOST_SET_VRING_CALL:
 415                 if (vq->call_ctx.ctx) {
 416                         cb.callback = vhost_vdpa_virtqueue_cb;
 417                         cb.private = vq;
 418                 } else {
 419                         cb.callback = NULL;
 420                         cb.private = NULL;
 421                 }
 422                 ops->set_vq_cb(vdpa, idx, &cb);
 423                 vhost_vdpa_setup_vq_irq(v, idx);
 424                 break;
 425
 426         case VHOST_SET_VRING_NUM:
 427                 ops->set_vq_num(vdpa, idx, vq->num);
 428                 break;
 429         }
 430
 431         return r;
 432 }
 433
 434 static long vhost_vdpa_unlocked_ioctl(struct file *filep,
 435                                       unsigned int cmd, unsigned long arg)
 436 {
 437         struct vhost_vdpa *v = filep->private_data;
 438         struct vhost_dev *d = &v->vdev;
 439         void __user *argp = (void __user *)arg;
 440         u64 __user *featurep = argp;
 441         u64 features;
 442         long r = 0;
 443
 444         if (cmd == VHOST_SET_BACKEND_FEATURES) {
 445                 if (copy_from_user(&features, featurep, sizeof(features)))
 446                         return -EFAULT;
 447                 if (features & ~VHOST_VDPA_BACKEND_FEATURES)
 448                         return -EOPNOTSUPP;
 449                 vhost_set_backend_features(&v->vdev, features);
 450                 return 0;
 451         }
 452
 453         mutex_lock(&d->mutex);
 454
 455         switch (cmd) {
 456         case VHOST_VDPA_GET_DEVICE_ID:
 457                 r = vhost_vdpa_get_device_id(v, argp);
 458                 break;
 459         case VHOST_VDPA_GET_STATUS:
 460                 r = vhost_vdpa_get_status(v, argp);
 461                 break;
 462         case VHOST_VDPA_SET_STATUS:
 463                 r = vhost_vdpa_set_status(v, argp);
 464                 break;
 465         case VHOST_VDPA_GET_CONFIG:
 466                 r = vhost_vdpa_get_config(v, argp);
 467                 break;
 468         case VHOST_VDPA_SET_CONFIG:
 469                 r = vhost_vdpa_set_config(v, argp);
 470                 break;
 471         case VHOST_GET_FEATURES:
 472                 r = vhost_vdpa_get_features(v, argp);
 473                 break;
 474         case VHOST_SET_FEATURES:
 475                 r = vhost_vdpa_set_features(v, argp);
 476                 break;
 477         case VHOST_VDPA_GET_VRING_NUM:
 478                 r = vhost_vdpa_get_vring_num(v, argp);
 479                 break;
 480         case VHOST_SET_LOG_BASE:
 481         case VHOST_SET_LOG_FD:
 482                 r = -ENOIOCTLCMD;
 483                 break;
 484         case VHOST_VDPA_SET_CONFIG_CALL:
 485                 r = vhost_vdpa_set_config_call(v, argp);
 486                 break;
 487         case VHOST_GET_BACKEND_FEATURES:
 488                 features = VHOST_VDPA_BACKEND_FEATURES;
 489                 if (copy_to_user(featurep, &features, sizeof(features)))
 490                         r = -EFAULT;
 491                 break;
 492         case VHOST_VDPA_GET_IOVA_RANGE:
 493                 r = vhost_vdpa_get_iova_range(v, argp);
 494                 break;
 495         default:
 496                 r = vhost_dev_ioctl(&v->vdev, cmd, argp);
 497                 if (r == -ENOIOCTLCMD)
 498                         r = vhost_vdpa_vring_ioctl(v, cmd, argp);
 499                 break;
 500         }
 501
 502         mutex_unlock(&d->mutex);
 503         return r;
 504 }
 505
 506 static void vhost_vdpa_pa_unmap(struct vhost_vdpa *v, u64 start, u64 last)
 507 {
 508         struct vhost_dev *dev = &v->vdev;
 509         struct vhost_iotlb *iotlb = dev->iotlb;
 510         struct vhost_iotlb_map *map;
 511         struct page *page;
 512         unsigned long pfn, pinned;
 513
 514         while ((map = vhost_iotlb_itree_first(iotlb, start, last)) != NULL) {
 515                 pinned = PFN_DOWN(map->size);
 516                 for (pfn = PFN_DOWN(map->addr);
 517                      pinned > 0; pfn++, pinned--) {
 518                         page = pfn_to_page(pfn);
 519                         if (map->perm & VHOST_ACCESS_WO)
 520                                 set_page_dirty_lock(page);
 521                         unpin_user_page(page);
 522                 }
 523                 atomic64_sub(PFN_DOWN(map->size), &dev->mm->pinned_vm);
 524                 vhost_iotlb_map_free(iotlb, map);
 525         }
 526 }
 527
 528 static void vhost_vdpa_va_unmap(struct vhost_vdpa *v, u64 start, u64 last)
 529 {
 530         struct vhost_dev *dev = &v->vdev;
 531         struct vhost_iotlb *iotlb = dev->iotlb;
 532         struct vhost_iotlb_map *map;
 533         struct vdpa_map_file *map_file;
 534
 535         while ((map = vhost_iotlb_itree_first(iotlb, start, last)) != NULL) {
 536                 map_file = (struct vdpa_map_file *)map->opaque;
 537                 fput(map_file->file);
 538                 kfree(map_file);
 539                 vhost_iotlb_map_free(iotlb, map);
 540         }
 541 }
 542
 543 static void vhost_vdpa_iotlb_unmap(struct vhost_vdpa *v, u64 start, u64 last)
 544 {
 545         struct vdpa_device *vdpa = v->vdpa;
 546
 547         if (vdpa->use_va)
 548                 return vhost_vdpa_va_unmap(v, start, last);
 549
 550         return vhost_vdpa_pa_unmap(v, start, last);
 551 }
 552
 553 static void vhost_vdpa_iotlb_free(struct vhost_vdpa *v)
 554 {
 555         struct vhost_dev *dev = &v->vdev;
 556
 557         vhost_vdpa_iotlb_unmap(v, 0ULL, 0ULL - 1);
 558         kfree(dev->iotlb);
 559         dev->iotlb = NULL;
 560 }
 561
 562 static int perm_to_iommu_flags(u32 perm)
 563 {
 564         int flags = 0;
 565
 566         switch (perm) {
 567         case VHOST_ACCESS_WO:
 568                 flags |= IOMMU_WRITE;
 569                 break;
 570         case VHOST_ACCESS_RO:
 571                 flags |= IOMMU_READ;
 572                 break;
 573         case VHOST_ACCESS_RW:
 574                 flags |= (IOMMU_WRITE | IOMMU_READ);
 575                 break;
 576         default:
 577                 WARN(1, "invalidate vhost IOTLB permission\n");
 578                 break;
 579         }
 580
 581         return flags | IOMMU_CACHE;
 582 }
 583
 584 static int vhost_vdpa_map(struct vhost_vdpa *v, u64 iova,
 585                           u64 size, u64 pa, u32 perm, void *opaque)
 586 {
 587         struct vhost_dev *dev = &v->vdev;
 588         struct vdpa_device *vdpa = v->vdpa;
 589         const struct vdpa_config_ops *ops = vdpa->config;
 590         int r = 0;
 591
 592         r = vhost_iotlb_add_range_ctx(dev->iotlb, iova, iova + size - 1,
 593                                       pa, perm, opaque);
 594         if (r)
 595                 return r;
 596
 597         if (ops->dma_map) {
 598                 r = ops->dma_map(vdpa, iova, size, pa, perm, opaque);
 599         } else if (ops->set_map) {
 600                 if (!v->in_batch)
 601                         r = ops->set_map(vdpa, dev->iotlb);
 602         } else {
 603                 r = iommu_map(v->domain, iova, pa, size,
 604                               perm_to_iommu_flags(perm));
 605         }
 606         if (r) {
 607                 vhost_iotlb_del_range(dev->iotlb, iova, iova + size - 1);
 608                 return r;
 609         }
 610
 611         if (!vdpa->use_va)
 612                 atomic64_add(PFN_DOWN(size), &dev->mm->pinned_vm);
 613
 614         return 0;
 615 }
 616
 617 static void vhost_vdpa_unmap(struct vhost_vdpa *v, u64 iova, u64 size)
 618 {
 619         struct vhost_dev *dev = &v->vdev;
 620         struct vdpa_device *vdpa = v->vdpa;
 621         const struct vdpa_config_ops *ops = vdpa->config;
 622
 623         vhost_vdpa_iotlb_unmap(v, iova, iova + size - 1);
 624
 625         if (ops->dma_map) {
 626                 ops->dma_unmap(vdpa, iova, size);
 627         } else if (ops->set_map) {
 628                 if (!v->in_batch)
 629                         ops->set_map(vdpa, dev->iotlb);
 630         } else {
 631                 iommu_unmap(v->domain, iova, size);
 632         }
 633 }
 634
 635 static int vhost_vdpa_va_map(struct vhost_vdpa *v,
 636                              u64 iova, u64 size, u64 uaddr, u32 perm)
 637 {
 638         struct vhost_dev *dev = &v->vdev;
 639         u64 offset, map_size, map_iova = iova;
 640         struct vdpa_map_file *map_file;
 641         struct vm_area_struct *vma;
 642         int ret = 0;
 643
 644         mmap_read_lock(dev->mm);
 645
 646         while (size) {
 647                 vma = find_vma(dev->mm, uaddr);
 648                 if (!vma) {
 649                         ret = -EINVAL;
 650                         break;
 651                 }
 652                 map_size = min(size, vma->vm_end - uaddr);
 653                 if (!(vma->vm_file && (vma->vm_flags & VM_SHARED) &&
 654                         !(vma->vm_flags & (VM_IO | VM_PFNMAP))))
 655                         goto next;
 656
 657                 map_file = kzalloc(sizeof(*map_file), GFP_KERNEL);
 658                 if (!map_file) {
 659                         ret = -ENOMEM;
 660                         break;
 661                 }
 662                 offset = (vma->vm_pgoff << PAGE_SHIFT) + uaddr - vma->vm_start;
 663                 map_file->offset = offset;
 664                 map_file->file = get_file(vma->vm_file);
 665                 ret = vhost_vdpa_map(v, map_iova, map_size, uaddr,
 666                                      perm, map_file);
 667                 if (ret) {
 668                         fput(map_file->file);
 669                         kfree(map_file);
 670                         break;
 671                 }
 672 next:
 673                 size -= map_size;
 674                 uaddr += map_size;
 675                 map_iova += map_size;
 676         }
 677         if (ret)
 678                 vhost_vdpa_unmap(v, iova, map_iova - iova);
 679
 680         mmap_read_unlock(dev->mm);
 681
 682         return ret;
 683 }
 684
 685 static int vhost_vdpa_pa_map(struct vhost_vdpa *v,
 686                              u64 iova, u64 size, u64 uaddr, u32 perm)
 687 {
 688         struct vhost_dev *dev = &v->vdev;
 689         struct page **page_list;
 690         unsigned long list_size = PAGE_SIZE / sizeof(struct page *);
 691         unsigned int gup_flags = FOLL_LONGTERM;
 692         unsigned long npages, cur_base, map_pfn, last_pfn = 0;
 693         unsigned long lock_limit, sz2pin, nchunks, i;
 694         u64 start = iova;
 695         long pinned;
 696         int ret = 0;
 697
 698         /* Limit the use of memory for bookkeeping */
 699         page_list = (struct page **) __get_free_page(GFP_KERNEL);
 700         if (!page_list)
 701                 return -ENOMEM;
 702
 703         if (perm & VHOST_ACCESS_WO)
 704                 gup_flags |= FOLL_WRITE;
 705
 706         npages = PFN_UP(size + (iova & ~PAGE_MASK));
 707         if (!npages) {
 708                 ret = -EINVAL;
 709                 goto free;
 710         }
 711
 712         mmap_read_lock(dev->mm);
 713
 714         lock_limit = PFN_DOWN(rlimit(RLIMIT_MEMLOCK));
 715         if (npages + atomic64_read(&dev->mm->pinned_vm) > lock_limit) {
 716                 ret = -ENOMEM;
 717                 goto unlock;
 718         }
 719
 720         cur_base = uaddr & PAGE_MASK;
 721         iova &= PAGE_MASK;
 722         nchunks = 0;
 723
 724         while (npages) {
 725                 sz2pin = min_t(unsigned long, npages, list_size);
 726                 pinned = pin_user_pages(cur_base, sz2pin,
 727                                         gup_flags, page_list, NULL);
 728                 if (sz2pin != pinned) {
 729                         if (pinned < 0) {
 730                                 ret = pinned;
 731                         } else {
 732                                 unpin_user_pages(page_list, pinned);
 733                                 ret = -ENOMEM;
 734                         }
 735                         goto out;
 736                 }
 737                 nchunks++;
 738
 739                 if (!last_pfn)
 740                         map_pfn = page_to_pfn(page_list[0]);
 741
 742                 for (i = 0; i < pinned; i++) {
 743                         unsigned long this_pfn = page_to_pfn(page_list[i]);
 744                         u64 csize;
 745
 746                         if (last_pfn && (this_pfn != last_pfn + 1)) {
 747                                 /* Pin a contiguous chunk of memory */
 748                                 csize = PFN_PHYS(last_pfn - map_pfn + 1);
 749                                 ret = vhost_vdpa_map(v, iova, csize,
 750                                                      PFN_PHYS(map_pfn),
 751                                                      perm, NULL);
 752                                 if (ret) {
 753                                         /*
 754                                          * Unpin the pages that are left unmapped
 755                                          * from this point on in the current
 756                                          * page_list. The remaining outstanding
 757                                          * ones which may stride across several
 758                                          * chunks will be covered in the common
 759                                          * error path subsequently.
 760                                          */
 761                                         unpin_user_pages(&page_list[i],
 762                                                          pinned - i);
 763                                         goto out;
 764                                 }
 765
 766                                 map_pfn = this_pfn;
 767                                 iova += csize;
 768                                 nchunks = 0;
 769                         }
 770
 771                         last_pfn = this_pfn;
 772                 }
 773
 774                 cur_base += PFN_PHYS(pinned);
 775                 npages -= pinned;
 776         }
 777
 778         /* Pin the rest chunk */
 779         ret = vhost_vdpa_map(v, iova, PFN_PHYS(last_pfn - map_pfn + 1),
 780                              PFN_PHYS(map_pfn), perm, NULL);
 781 out:
 782         if (ret) {
 783                 if (nchunks) {
 784                         unsigned long pfn;
 785
 786                         /*
 787                          * Unpin the outstanding pages which are yet to be
 788                          * mapped but haven't due to vdpa_map() or
 789                          * pin_user_pages() failure.
 790                          *
 791                          * Mapped pages are accounted in vdpa_map(), hence
 792                          * the corresponding unpinning will be handled by
 793                          * vdpa_unmap().
 794                          */
 795                         WARN_ON(!last_pfn);
 796                         for (pfn = map_pfn; pfn <= last_pfn; pfn++)
 797                                 unpin_user_page(pfn_to_page(pfn));
 798                 }
 799                 vhost_vdpa_unmap(v, start, size);
 800         }
 801 unlock:
 802         mmap_read_unlock(dev->mm);
 803 free:
 804         free_page((unsigned long)page_list);
 805         return ret;
 806
 807 }
 808
 809 static int vhost_vdpa_process_iotlb_update(struct vhost_vdpa *v,
 810                                            struct vhost_iotlb_msg *msg)
 811 {
 812         struct vhost_dev *dev = &v->vdev;
 813         struct vdpa_device *vdpa = v->vdpa;
 814         struct vhost_iotlb *iotlb = dev->iotlb;
 815
 816         if (msg->iova < v->range.first || !msg->size ||
 817             msg->iova > U64_MAX - msg->size + 1 ||
 818             msg->iova + msg->size - 1 > v->range.last)
 819                 return -EINVAL;
 820
 821         if (vhost_iotlb_itree_first(iotlb, msg->iova,
 822                                     msg->iova + msg->size - 1))
 823                 return -EEXIST;
 824
 825         if (vdpa->use_va)
 826                 return vhost_vdpa_va_map(v, msg->iova, msg->size,
 827                                          msg->uaddr, msg->perm);
 828
 829         return vhost_vdpa_pa_map(v, msg->iova, msg->size, msg->uaddr,
 830                                  msg->perm);
 831 }
 832
 833 static int vhost_vdpa_process_iotlb_msg(struct vhost_dev *dev,
 834                                         struct vhost_iotlb_msg *msg)
 835 {
 836         struct vhost_vdpa *v = container_of(dev, struct vhost_vdpa, vdev);
 837         struct vdpa_device *vdpa = v->vdpa;
 838         const struct vdpa_config_ops *ops = vdpa->config;
 839         int r = 0;
 840
 841         mutex_lock(&dev->mutex);
 842
 843         r = vhost_dev_check_owner(dev);
 844         if (r)
 845                 goto unlock;
 846
 847         switch (msg->type) {
 848         case VHOST_IOTLB_UPDATE:
 849                 r = vhost_vdpa_process_iotlb_update(v, msg);
 850                 break;
 851         case VHOST_IOTLB_INVALIDATE:
 852                 vhost_vdpa_unmap(v, msg->iova, msg->size);
 853                 break;
 854         case VHOST_IOTLB_BATCH_BEGIN:
 855                 v->in_batch = true;
 856                 break;
 857         case VHOST_IOTLB_BATCH_END:
 858                 if (v->in_batch && ops->set_map)
 859                         ops->set_map(vdpa, dev->iotlb);
 860                 v->in_batch = false;
 861                 break;
 862         default:
 863                 r = -EINVAL;
 864                 break;
 865         }
 866 unlock:
 867         mutex_unlock(&dev->mutex);
 868
 869         return r;
 870 }
 871
 872 static ssize_t vhost_vdpa_chr_write_iter(struct kiocb *iocb,
 873                                          struct iov_iter *from)
 874 {
 875         struct file *file = iocb->ki_filp;
 876         struct vhost_vdpa *v = file->private_data;
 877         struct vhost_dev *dev = &v->vdev;
 878
 879         return vhost_chr_write_iter(dev, from);
 880 }
 881
 882 static int vhost_vdpa_alloc_domain(struct vhost_vdpa *v)
 883 {
 884         struct vdpa_device *vdpa = v->vdpa;
 885         const struct vdpa_config_ops *ops = vdpa->config;
 886         struct device *dma_dev = vdpa_get_dma_dev(vdpa);
 887         struct bus_type *bus;
 888         int ret;
 889
 890         /* Device want to do DMA by itself */
 891         if (ops->set_map || ops->dma_map)
 892                 return 0;
 893
 894         bus = dma_dev->bus;
 895         if (!bus)
 896                 return -EFAULT;
 897
 898         if (!iommu_capable(bus, IOMMU_CAP_CACHE_COHERENCY))
 899                 return -ENOTSUPP;
 900
 901         v->domain = iommu_domain_alloc(bus);
 902         if (!v->domain)
 903                 return -EIO;
 904
 905         ret = iommu_attach_device(v->domain, dma_dev);
 906         if (ret)
 907                 goto err_attach;
 908
 909         return 0;
 910
 911 err_attach:
 912         iommu_domain_free(v->domain);
 913         return ret;
 914 }
 915
 916 static void vhost_vdpa_free_domain(struct vhost_vdpa *v)
 917 {
 918         struct vdpa_device *vdpa = v->vdpa;
 919         struct device *dma_dev = vdpa_get_dma_dev(vdpa);
 920
 921         if (v->domain) {
 922                 iommu_detach_device(v->domain, dma_dev);
 923                 iommu_domain_free(v->domain);
 924         }
 925
 926         v->domain = NULL;
 927 }
 928
 929 static void vhost_vdpa_set_iova_range(struct vhost_vdpa *v)
 930 {
 931         struct vdpa_iova_range *range = &v->range;
 932         struct vdpa_device *vdpa = v->vdpa;
 933         const struct vdpa_config_ops *ops = vdpa->config;
 934
 935         if (ops->get_iova_range) {
 936                 *range = ops->get_iova_range(vdpa);
 937         } else if (v->domain && v->domain->geometry.force_aperture) {
 938                 range->first = v->domain->geometry.aperture_start;
 939                 range->last = v->domain->geometry.aperture_end;
 940         } else {
 941                 range->first = 0;
 942                 range->last = ULLONG_MAX;
 943         }
 944 }
 945
 946 static int vhost_vdpa_open(struct inode *inode, struct file *filep)
 947 {
 948         struct vhost_vdpa *v;
 949         struct vhost_dev *dev;
 950         struct vhost_virtqueue **vqs;
 951         int nvqs, i, r, opened;
 952
 953         v = container_of(inode->i_cdev, struct vhost_vdpa, cdev);
 954
 955         opened = atomic_cmpxchg(&v->opened, 0, 1);
 956         if (opened)
 957                 return -EBUSY;
 958
 959         nvqs = v->nvqs;
 960         r = vhost_vdpa_reset(v);
 961         if (r)
 962                 goto err;
 963
 964         vqs = kmalloc_array(nvqs, sizeof(*vqs), GFP_KERNEL);
 965         if (!vqs) {
 966                 r = -ENOMEM;
 967                 goto err;
 968         }
 969
 970         dev = &v->vdev;
 971         for (i = 0; i < nvqs; i++) {
 972                 vqs[i] = &v->vqs[i];
 973                 vqs[i]->handle_kick = handle_vq_kick;
 974         }
 975         vhost_dev_init(dev, vqs, nvqs, 0, 0, 0, false,
 976                        vhost_vdpa_process_iotlb_msg);
 977
 978         dev->iotlb = vhost_iotlb_alloc(0, 0);
 979         if (!dev->iotlb) {
 980                 r = -ENOMEM;
 981                 goto err_init_iotlb;
 982         }
 983
 984         r = vhost_vdpa_alloc_domain(v);
 985         if (r)
 986                 goto err_init_iotlb;
 987
 988         vhost_vdpa_set_iova_range(v);
 989
 990         filep->private_data = v;
 991
 992         return 0;
 993
 994 err_init_iotlb:
 995         vhost_dev_cleanup(&v->vdev);
 996         kfree(vqs);
 997 err:
 998         atomic_dec(&v->opened);
 999         return r;
1000 }
1001
1002 static void vhost_vdpa_clean_irq(struct vhost_vdpa *v)
1003 {
1004         int i;
1005
1006         for (i = 0; i < v->nvqs; i++)
1007                 vhost_vdpa_unsetup_vq_irq(v, i);
1008 }
1009
1010 static int vhost_vdpa_release(struct inode *inode, struct file *filep)
1011 {
1012         struct vhost_vdpa *v = filep->private_data;
1013         struct vhost_dev *d = &v->vdev;
1014
1015         mutex_lock(&d->mutex);
1016         filep->private_data = NULL;
1017         vhost_vdpa_clean_irq(v);
1018         vhost_vdpa_reset(v);
1019         vhost_dev_stop(&v->vdev);
1020         vhost_vdpa_iotlb_free(v);
1021         vhost_vdpa_free_domain(v);
1022         vhost_vdpa_config_put(v);
1023         vhost_dev_cleanup(&v->vdev);
1024         kfree(v->vdev.vqs);
1025         mutex_unlock(&d->mutex);
1026
1027         atomic_dec(&v->opened);
1028         complete(&v->completion);
1029
1030         return 0;
1031 }
1032
1033 #ifdef CONFIG_MMU
1034 static vm_fault_t vhost_vdpa_fault(struct vm_fault *vmf)
1035 {
1036         struct vhost_vdpa *v = vmf->vma->vm_file->private_data;
1037         struct vdpa_device *vdpa = v->vdpa;
1038         const struct vdpa_config_ops *ops = vdpa->config;
1039         struct vdpa_notification_area notify;
1040         struct vm_area_struct *vma = vmf->vma;
1041         u16 index = vma->vm_pgoff;
1042
1043         notify = ops->get_vq_notification(vdpa, index);
1044
1045         vma->vm_page_prot = pgprot_noncached(vma->vm_page_prot);
1046         if (remap_pfn_range(vma, vmf->address & PAGE_MASK,
1047                             PFN_DOWN(notify.addr), PAGE_SIZE,
1048                             vma->vm_page_prot))
1049                 return VM_FAULT_SIGBUS;
1050
1051         return VM_FAULT_NOPAGE;
1052 }
1053
1054 static const struct vm_operations_struct vhost_vdpa_vm_ops = {
1055         .fault = vhost_vdpa_fault,
1056 };
1057
1058 static int vhost_vdpa_mmap(struct file *file, struct vm_area_struct *vma)
1059 {
1060         struct vhost_vdpa *v = vma->vm_file->private_data;
1061         struct vdpa_device *vdpa = v->vdpa;
1062         const struct vdpa_config_ops *ops = vdpa->config;
1063         struct vdpa_notification_area notify;
1064         unsigned long index = vma->vm_pgoff;
1065
1066         if (vma->vm_end - vma->vm_start != PAGE_SIZE)
1067                 return -EINVAL;
1068         if ((vma->vm_flags & VM_SHARED) == 0)
1069                 return -EINVAL;
1070         if (vma->vm_flags & VM_READ)
1071                 return -EINVAL;
1072         if (index > 65535)
1073                 return -EINVAL;
1074         if (!ops->get_vq_notification)
1075                 return -ENOTSUPP;
1076
1077         /* To be safe and easily modelled by userspace, We only
1078          * support the doorbell which sits on the page boundary and
1079          * does not share the page with other registers.
1080          */
1081         notify = ops->get_vq_notification(vdpa, index);
1082         if (notify.addr & (PAGE_SIZE - 1))
1083                 return -EINVAL;
1084         if (vma->vm_end - vma->vm_start != notify.size)
1085                 return -ENOTSUPP;
1086
1087         vma->vm_flags |= VM_IO | VM_PFNMAP | VM_DONTEXPAND | VM_DONTDUMP;
1088         vma->vm_ops = &vhost_vdpa_vm_ops;
1089         return 0;
1090 }
1091 #endif /* CONFIG_MMU */
1092
1093 static const struct file_operations vhost_vdpa_fops = {
1094         .owner          = THIS_MODULE,
1095         .open           = vhost_vdpa_open,
1096         .release        = vhost_vdpa_release,
1097         .write_iter     = vhost_vdpa_chr_write_iter,
1098         .unlocked_ioctl = vhost_vdpa_unlocked_ioctl,
1099 #ifdef CONFIG_MMU
1100         .mmap           = vhost_vdpa_mmap,
1101 #endif /* CONFIG_MMU */
1102         .compat_ioctl   = compat_ptr_ioctl,
1103 };
1104
1105 static void vhost_vdpa_release_dev(struct device *device)
1106 {
1107         struct vhost_vdpa *v =
1108                container_of(device, struct vhost_vdpa, dev);
1109
1110         ida_simple_remove(&vhost_vdpa_ida, v->minor);
1111         kfree(v->vqs);
1112         kfree(v);
1113 }
1114
1115 static int vhost_vdpa_probe(struct vdpa_device *vdpa)
1116 {
1117         const struct vdpa_config_ops *ops = vdpa->config;
1118         struct vhost_vdpa *v;
1119         int minor;
1120         int r;
1121
1122         v = kzalloc(sizeof(*v), GFP_KERNEL | __GFP_RETRY_MAYFAIL);
1123         if (!v)
1124                 return -ENOMEM;
1125
1126         minor = ida_simple_get(&vhost_vdpa_ida, 0,
1127                                VHOST_VDPA_DEV_MAX, GFP_KERNEL);
1128         if (minor < 0) {
1129                 kfree(v);
1130                 return minor;
1131         }
1132
1133         atomic_set(&v->opened, 0);
1134         v->minor = minor;
1135         v->vdpa = vdpa;
1136         v->nvqs = vdpa->nvqs;
1137         v->virtio_id = ops->get_device_id(vdpa);
1138
1139         device_initialize(&v->dev);
1140         v->dev.release = vhost_vdpa_release_dev;
1141         v->dev.parent = &vdpa->dev;
1142         v->dev.devt = MKDEV(MAJOR(vhost_vdpa_major), minor);
1143         v->vqs = kmalloc_array(v->nvqs, sizeof(struct vhost_virtqueue),
1144                                GFP_KERNEL);
1145         if (!v->vqs) {
1146                 r = -ENOMEM;
1147                 goto err;
1148         }
1149
1150         r = dev_set_name(&v->dev, "vhost-vdpa-%u", minor);
1151         if (r)
1152                 goto err;
1153
1154         cdev_init(&v->cdev, &vhost_vdpa_fops);
1155         v->cdev.owner = THIS_MODULE;
1156
1157         r = cdev_device_add(&v->cdev, &v->dev);
1158         if (r)
1159                 goto err;
1160
1161         init_completion(&v->completion);
1162         vdpa_set_drvdata(vdpa, v);
1163
1164         return 0;
1165
1166 err:
1167         put_device(&v->dev);
1168         return r;
1169 }
1170
1171 static void vhost_vdpa_remove(struct vdpa_device *vdpa)
1172 {
1173         struct vhost_vdpa *v = vdpa_get_drvdata(vdpa);
1174         int opened;
1175
1176         cdev_device_del(&v->cdev, &v->dev);
1177
1178         do {
1179                 opened = atomic_cmpxchg(&v->opened, 0, 1);
1180                 if (!opened)
1181                         break;
1182                 wait_for_completion(&v->completion);
1183         } while (1);
1184
1185         put_device(&v->dev);
1186 }
1187
1188 static struct vdpa_driver vhost_vdpa_driver = {
1189         .driver = {
1190                 .name   = "vhost_vdpa",
1191         },
1192         .probe  = vhost_vdpa_probe,
1193         .remove = vhost_vdpa_remove,
1194 };
1195
1196 static int __init vhost_vdpa_init(void)
1197 {
1198         int r;
1199
1200         r = alloc_chrdev_region(&vhost_vdpa_major, 0, VHOST_VDPA_DEV_MAX,
1201                                 "vhost-vdpa");
1202         if (r)
1203                 goto err_alloc_chrdev;
1204
1205         r = vdpa_register_driver(&vhost_vdpa_driver);
1206         if (r)
1207                 goto err_vdpa_register_driver;
1208
1209         return 0;
1210
1211 err_vdpa_register_driver:
1212         unregister_chrdev_region(vhost_vdpa_major, VHOST_VDPA_DEV_MAX);
1213 err_alloc_chrdev:
1214         return r;
1215 }
1216 module_init(vhost_vdpa_init);
1217
1218 static void __exit vhost_vdpa_exit(void)
1219 {
1220         vdpa_unregister_driver(&vhost_vdpa_driver);
1221         unregister_chrdev_region(vhost_vdpa_major, VHOST_VDPA_DEV_MAX);
1222 }
1223 module_exit(vhost_vdpa_exit);
1224
1225 MODULE_VERSION("0.0.1");
1226 MODULE_LICENSE("GPL v2");
1227 MODULE_AUTHOR("Intel Corporation");
1228 MODULE_DESCRIPTION("vDPA-based vhost backend for virtio");