Merge tag 'for_linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mst/vhost
authorLinus Torvalds <torvalds@linux-foundation.org>
Wed, 17 Jul 2019 18:26:09 +0000 (11:26 -0700)
committerLinus Torvalds <torvalds@linux-foundation.org>
Wed, 17 Jul 2019 18:26:09 +0000 (11:26 -0700)
Pull virtio, vhost updates from Michael Tsirkin:
 "Fixes, features, performance:

   - new iommu device

   - vhost guest memory access using vmap (just meta-data for now)

   - minor fixes"

* tag 'for_linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mst/vhost:
  virtio-mmio: add error check for platform_get_irq
  scsi: virtio_scsi: Use struct_size() helper
  iommu/virtio: Add event queue
  iommu/virtio: Add probe request
  iommu: Add virtio-iommu driver
  PCI: OF: Initialize dev->fwnode appropriately
  of: Allow the iommu-map property to omit untranslated devices
  dt-bindings: virtio: Add virtio-pci-iommu node
  dt-bindings: virtio-mmio: Add IOMMU description
  vhost: fix clang build warning
  vhost: access vq metadata through kernel virtual address
  vhost: factor out setting vring addr and num
  vhost: introduce helpers to get the size of metadata area
  vhost: rename vq_iotlb_prefetch() to vq_meta_prefetch()
  vhost: fine grain userspace memory accessors
  vhost: generalize adding used elem

1  2 
MAINTAINERS
drivers/iommu/virtio-iommu.c
drivers/scsi/virtio_scsi.c
drivers/vhost/net.c
drivers/vhost/vhost.c

diff --cc MAINTAINERS
Simple merge
index 0000000000000000000000000000000000000000,4620dd221ffd3db22ef223b311f6d20de4c8de4b..433f4d2ee9565be632c1e780e7cdedcdf4254e1d
mode 000000,100644..100644
--- /dev/null
@@@ -1,0 -1,1158 +1,1158 @@@
 -static int viommu_match_node(struct device *dev, void *data)
+ // SPDX-License-Identifier: GPL-2.0
+ /*
+  * Virtio driver for the paravirtualized IOMMU
+  *
+  * Copyright (C) 2018 Arm Limited
+  */
+ #define pr_fmt(fmt) KBUILD_MODNAME ": " fmt
+ #include <linux/amba/bus.h>
+ #include <linux/delay.h>
+ #include <linux/dma-iommu.h>
+ #include <linux/freezer.h>
+ #include <linux/interval_tree.h>
+ #include <linux/iommu.h>
+ #include <linux/module.h>
+ #include <linux/of_iommu.h>
+ #include <linux/of_platform.h>
+ #include <linux/pci.h>
+ #include <linux/platform_device.h>
+ #include <linux/virtio.h>
+ #include <linux/virtio_config.h>
+ #include <linux/virtio_ids.h>
+ #include <linux/wait.h>
+ #include <uapi/linux/virtio_iommu.h>
+ #define MSI_IOVA_BASE                 0x8000000
+ #define MSI_IOVA_LENGTH                       0x100000
+ #define VIOMMU_REQUEST_VQ             0
+ #define VIOMMU_EVENT_VQ                       1
+ #define VIOMMU_NR_VQS                 2
+ struct viommu_dev {
+       struct iommu_device             iommu;
+       struct device                   *dev;
+       struct virtio_device            *vdev;
+       struct ida                      domain_ids;
+       struct virtqueue                *vqs[VIOMMU_NR_VQS];
+       spinlock_t                      request_lock;
+       struct list_head                requests;
+       void                            *evts;
+       /* Device configuration */
+       struct iommu_domain_geometry    geometry;
+       u64                             pgsize_bitmap;
+       u8                              domain_bits;
+       u32                             probe_size;
+ };
+ struct viommu_mapping {
+       phys_addr_t                     paddr;
+       struct interval_tree_node       iova;
+       u32                             flags;
+ };
+ struct viommu_domain {
+       struct iommu_domain             domain;
+       struct viommu_dev               *viommu;
+       struct mutex                    mutex; /* protects viommu pointer */
+       unsigned int                    id;
+       spinlock_t                      mappings_lock;
+       struct rb_root_cached           mappings;
+       unsigned long                   nr_endpoints;
+ };
+ struct viommu_endpoint {
+       struct device                   *dev;
+       struct viommu_dev               *viommu;
+       struct viommu_domain            *vdomain;
+       struct list_head                resv_regions;
+ };
+ struct viommu_request {
+       struct list_head                list;
+       void                            *writeback;
+       unsigned int                    write_offset;
+       unsigned int                    len;
+       char                            buf[];
+ };
+ #define VIOMMU_FAULT_RESV_MASK                0xffffff00
+ struct viommu_event {
+       union {
+               u32                     head;
+               struct virtio_iommu_fault fault;
+       };
+ };
+ #define to_viommu_domain(domain)      \
+       container_of(domain, struct viommu_domain, domain)
+ static int viommu_get_req_errno(void *buf, size_t len)
+ {
+       struct virtio_iommu_req_tail *tail = buf + len - sizeof(*tail);
+       switch (tail->status) {
+       case VIRTIO_IOMMU_S_OK:
+               return 0;
+       case VIRTIO_IOMMU_S_UNSUPP:
+               return -ENOSYS;
+       case VIRTIO_IOMMU_S_INVAL:
+               return -EINVAL;
+       case VIRTIO_IOMMU_S_RANGE:
+               return -ERANGE;
+       case VIRTIO_IOMMU_S_NOENT:
+               return -ENOENT;
+       case VIRTIO_IOMMU_S_FAULT:
+               return -EFAULT;
+       case VIRTIO_IOMMU_S_IOERR:
+       case VIRTIO_IOMMU_S_DEVERR:
+       default:
+               return -EIO;
+       }
+ }
+ static void viommu_set_req_status(void *buf, size_t len, int status)
+ {
+       struct virtio_iommu_req_tail *tail = buf + len - sizeof(*tail);
+       tail->status = status;
+ }
+ static off_t viommu_get_write_desc_offset(struct viommu_dev *viommu,
+                                         struct virtio_iommu_req_head *req,
+                                         size_t len)
+ {
+       size_t tail_size = sizeof(struct virtio_iommu_req_tail);
+       if (req->type == VIRTIO_IOMMU_T_PROBE)
+               return len - viommu->probe_size - tail_size;
+       return len - tail_size;
+ }
+ /*
+  * __viommu_sync_req - Complete all in-flight requests
+  *
+  * Wait for all added requests to complete. When this function returns, all
+  * requests that were in-flight at the time of the call have completed.
+  */
+ static int __viommu_sync_req(struct viommu_dev *viommu)
+ {
+       int ret = 0;
+       unsigned int len;
+       size_t write_len;
+       struct viommu_request *req;
+       struct virtqueue *vq = viommu->vqs[VIOMMU_REQUEST_VQ];
+       assert_spin_locked(&viommu->request_lock);
+       virtqueue_kick(vq);
+       while (!list_empty(&viommu->requests)) {
+               len = 0;
+               req = virtqueue_get_buf(vq, &len);
+               if (!req)
+                       continue;
+               if (!len)
+                       viommu_set_req_status(req->buf, req->len,
+                                             VIRTIO_IOMMU_S_IOERR);
+               write_len = req->len - req->write_offset;
+               if (req->writeback && len == write_len)
+                       memcpy(req->writeback, req->buf + req->write_offset,
+                              write_len);
+               list_del(&req->list);
+               kfree(req);
+       }
+       return ret;
+ }
+ static int viommu_sync_req(struct viommu_dev *viommu)
+ {
+       int ret;
+       unsigned long flags;
+       spin_lock_irqsave(&viommu->request_lock, flags);
+       ret = __viommu_sync_req(viommu);
+       if (ret)
+               dev_dbg(viommu->dev, "could not sync requests (%d)\n", ret);
+       spin_unlock_irqrestore(&viommu->request_lock, flags);
+       return ret;
+ }
+ /*
+  * __viommu_add_request - Add one request to the queue
+  * @buf: pointer to the request buffer
+  * @len: length of the request buffer
+  * @writeback: copy data back to the buffer when the request completes.
+  *
+  * Add a request to the queue. Only synchronize the queue if it's already full.
+  * Otherwise don't kick the queue nor wait for requests to complete.
+  *
+  * When @writeback is true, data written by the device, including the request
+  * status, is copied into @buf after the request completes. This is unsafe if
+  * the caller allocates @buf on stack and drops the lock between add_req() and
+  * sync_req().
+  *
+  * Return 0 if the request was successfully added to the queue.
+  */
+ static int __viommu_add_req(struct viommu_dev *viommu, void *buf, size_t len,
+                           bool writeback)
+ {
+       int ret;
+       off_t write_offset;
+       struct viommu_request *req;
+       struct scatterlist top_sg, bottom_sg;
+       struct scatterlist *sg[2] = { &top_sg, &bottom_sg };
+       struct virtqueue *vq = viommu->vqs[VIOMMU_REQUEST_VQ];
+       assert_spin_locked(&viommu->request_lock);
+       write_offset = viommu_get_write_desc_offset(viommu, buf, len);
+       if (write_offset <= 0)
+               return -EINVAL;
+       req = kzalloc(sizeof(*req) + len, GFP_ATOMIC);
+       if (!req)
+               return -ENOMEM;
+       req->len = len;
+       if (writeback) {
+               req->writeback = buf + write_offset;
+               req->write_offset = write_offset;
+       }
+       memcpy(&req->buf, buf, write_offset);
+       sg_init_one(&top_sg, req->buf, write_offset);
+       sg_init_one(&bottom_sg, req->buf + write_offset, len - write_offset);
+       ret = virtqueue_add_sgs(vq, sg, 1, 1, req, GFP_ATOMIC);
+       if (ret == -ENOSPC) {
+               /* If the queue is full, sync and retry */
+               if (!__viommu_sync_req(viommu))
+                       ret = virtqueue_add_sgs(vq, sg, 1, 1, req, GFP_ATOMIC);
+       }
+       if (ret)
+               goto err_free;
+       list_add_tail(&req->list, &viommu->requests);
+       return 0;
+ err_free:
+       kfree(req);
+       return ret;
+ }
+ static int viommu_add_req(struct viommu_dev *viommu, void *buf, size_t len)
+ {
+       int ret;
+       unsigned long flags;
+       spin_lock_irqsave(&viommu->request_lock, flags);
+       ret = __viommu_add_req(viommu, buf, len, false);
+       if (ret)
+               dev_dbg(viommu->dev, "could not add request: %d\n", ret);
+       spin_unlock_irqrestore(&viommu->request_lock, flags);
+       return ret;
+ }
+ /*
+  * Send a request and wait for it to complete. Return the request status (as an
+  * errno)
+  */
+ static int viommu_send_req_sync(struct viommu_dev *viommu, void *buf,
+                               size_t len)
+ {
+       int ret;
+       unsigned long flags;
+       spin_lock_irqsave(&viommu->request_lock, flags);
+       ret = __viommu_add_req(viommu, buf, len, true);
+       if (ret) {
+               dev_dbg(viommu->dev, "could not add request (%d)\n", ret);
+               goto out_unlock;
+       }
+       ret = __viommu_sync_req(viommu);
+       if (ret) {
+               dev_dbg(viommu->dev, "could not sync requests (%d)\n", ret);
+               /* Fall-through (get the actual request status) */
+       }
+       ret = viommu_get_req_errno(buf, len);
+ out_unlock:
+       spin_unlock_irqrestore(&viommu->request_lock, flags);
+       return ret;
+ }
+ /*
+  * viommu_add_mapping - add a mapping to the internal tree
+  *
+  * On success, return the new mapping. Otherwise return NULL.
+  */
+ static int viommu_add_mapping(struct viommu_domain *vdomain, unsigned long iova,
+                             phys_addr_t paddr, size_t size, u32 flags)
+ {
+       unsigned long irqflags;
+       struct viommu_mapping *mapping;
+       mapping = kzalloc(sizeof(*mapping), GFP_ATOMIC);
+       if (!mapping)
+               return -ENOMEM;
+       mapping->paddr          = paddr;
+       mapping->iova.start     = iova;
+       mapping->iova.last      = iova + size - 1;
+       mapping->flags          = flags;
+       spin_lock_irqsave(&vdomain->mappings_lock, irqflags);
+       interval_tree_insert(&mapping->iova, &vdomain->mappings);
+       spin_unlock_irqrestore(&vdomain->mappings_lock, irqflags);
+       return 0;
+ }
+ /*
+  * viommu_del_mappings - remove mappings from the internal tree
+  *
+  * @vdomain: the domain
+  * @iova: start of the range
+  * @size: size of the range. A size of 0 corresponds to the entire address
+  *    space.
+  *
+  * On success, returns the number of unmapped bytes (>= size)
+  */
+ static size_t viommu_del_mappings(struct viommu_domain *vdomain,
+                                 unsigned long iova, size_t size)
+ {
+       size_t unmapped = 0;
+       unsigned long flags;
+       unsigned long last = iova + size - 1;
+       struct viommu_mapping *mapping = NULL;
+       struct interval_tree_node *node, *next;
+       spin_lock_irqsave(&vdomain->mappings_lock, flags);
+       next = interval_tree_iter_first(&vdomain->mappings, iova, last);
+       while (next) {
+               node = next;
+               mapping = container_of(node, struct viommu_mapping, iova);
+               next = interval_tree_iter_next(node, iova, last);
+               /* Trying to split a mapping? */
+               if (mapping->iova.start < iova)
+                       break;
+               /*
+                * Virtio-iommu doesn't allow UNMAP to split a mapping created
+                * with a single MAP request, so remove the full mapping.
+                */
+               unmapped += mapping->iova.last - mapping->iova.start + 1;
+               interval_tree_remove(node, &vdomain->mappings);
+               kfree(mapping);
+       }
+       spin_unlock_irqrestore(&vdomain->mappings_lock, flags);
+       return unmapped;
+ }
+ /*
+  * viommu_replay_mappings - re-send MAP requests
+  *
+  * When reattaching a domain that was previously detached from all endpoints,
+  * mappings were deleted from the device. Re-create the mappings available in
+  * the internal tree.
+  */
+ static int viommu_replay_mappings(struct viommu_domain *vdomain)
+ {
+       int ret = 0;
+       unsigned long flags;
+       struct viommu_mapping *mapping;
+       struct interval_tree_node *node;
+       struct virtio_iommu_req_map map;
+       spin_lock_irqsave(&vdomain->mappings_lock, flags);
+       node = interval_tree_iter_first(&vdomain->mappings, 0, -1UL);
+       while (node) {
+               mapping = container_of(node, struct viommu_mapping, iova);
+               map = (struct virtio_iommu_req_map) {
+                       .head.type      = VIRTIO_IOMMU_T_MAP,
+                       .domain         = cpu_to_le32(vdomain->id),
+                       .virt_start     = cpu_to_le64(mapping->iova.start),
+                       .virt_end       = cpu_to_le64(mapping->iova.last),
+                       .phys_start     = cpu_to_le64(mapping->paddr),
+                       .flags          = cpu_to_le32(mapping->flags),
+               };
+               ret = viommu_send_req_sync(vdomain->viommu, &map, sizeof(map));
+               if (ret)
+                       break;
+               node = interval_tree_iter_next(node, 0, -1UL);
+       }
+       spin_unlock_irqrestore(&vdomain->mappings_lock, flags);
+       return ret;
+ }
+ static int viommu_add_resv_mem(struct viommu_endpoint *vdev,
+                              struct virtio_iommu_probe_resv_mem *mem,
+                              size_t len)
+ {
+       size_t size;
+       u64 start64, end64;
+       phys_addr_t start, end;
+       struct iommu_resv_region *region = NULL;
+       unsigned long prot = IOMMU_WRITE | IOMMU_NOEXEC | IOMMU_MMIO;
+       start = start64 = le64_to_cpu(mem->start);
+       end = end64 = le64_to_cpu(mem->end);
+       size = end64 - start64 + 1;
+       /* Catch any overflow, including the unlikely end64 - start64 + 1 = 0 */
+       if (start != start64 || end != end64 || size < end64 - start64)
+               return -EOVERFLOW;
+       if (len < sizeof(*mem))
+               return -EINVAL;
+       switch (mem->subtype) {
+       default:
+               dev_warn(vdev->dev, "unknown resv mem subtype 0x%x\n",
+                        mem->subtype);
+               /* Fall-through */
+       case VIRTIO_IOMMU_RESV_MEM_T_RESERVED:
+               region = iommu_alloc_resv_region(start, size, 0,
+                                                IOMMU_RESV_RESERVED);
+               break;
+       case VIRTIO_IOMMU_RESV_MEM_T_MSI:
+               region = iommu_alloc_resv_region(start, size, prot,
+                                                IOMMU_RESV_MSI);
+               break;
+       }
+       if (!region)
+               return -ENOMEM;
+       list_add(&vdev->resv_regions, &region->list);
+       return 0;
+ }
+ static int viommu_probe_endpoint(struct viommu_dev *viommu, struct device *dev)
+ {
+       int ret;
+       u16 type, len;
+       size_t cur = 0;
+       size_t probe_len;
+       struct virtio_iommu_req_probe *probe;
+       struct virtio_iommu_probe_property *prop;
+       struct iommu_fwspec *fwspec = dev_iommu_fwspec_get(dev);
+       struct viommu_endpoint *vdev = fwspec->iommu_priv;
+       if (!fwspec->num_ids)
+               return -EINVAL;
+       probe_len = sizeof(*probe) + viommu->probe_size +
+                   sizeof(struct virtio_iommu_req_tail);
+       probe = kzalloc(probe_len, GFP_KERNEL);
+       if (!probe)
+               return -ENOMEM;
+       probe->head.type = VIRTIO_IOMMU_T_PROBE;
+       /*
+        * For now, assume that properties of an endpoint that outputs multiple
+        * IDs are consistent. Only probe the first one.
+        */
+       probe->endpoint = cpu_to_le32(fwspec->ids[0]);
+       ret = viommu_send_req_sync(viommu, probe, probe_len);
+       if (ret)
+               goto out_free;
+       prop = (void *)probe->properties;
+       type = le16_to_cpu(prop->type) & VIRTIO_IOMMU_PROBE_T_MASK;
+       while (type != VIRTIO_IOMMU_PROBE_T_NONE &&
+              cur < viommu->probe_size) {
+               len = le16_to_cpu(prop->length) + sizeof(*prop);
+               switch (type) {
+               case VIRTIO_IOMMU_PROBE_T_RESV_MEM:
+                       ret = viommu_add_resv_mem(vdev, (void *)prop, len);
+                       break;
+               default:
+                       dev_err(dev, "unknown viommu prop 0x%x\n", type);
+               }
+               if (ret)
+                       dev_err(dev, "failed to parse viommu prop 0x%x\n", type);
+               cur += len;
+               if (cur >= viommu->probe_size)
+                       break;
+               prop = (void *)probe->properties + cur;
+               type = le16_to_cpu(prop->type) & VIRTIO_IOMMU_PROBE_T_MASK;
+       }
+ out_free:
+       kfree(probe);
+       return ret;
+ }
+ static int viommu_fault_handler(struct viommu_dev *viommu,
+                               struct virtio_iommu_fault *fault)
+ {
+       char *reason_str;
+       u8 reason       = fault->reason;
+       u32 flags       = le32_to_cpu(fault->flags);
+       u32 endpoint    = le32_to_cpu(fault->endpoint);
+       u64 address     = le64_to_cpu(fault->address);
+       switch (reason) {
+       case VIRTIO_IOMMU_FAULT_R_DOMAIN:
+               reason_str = "domain";
+               break;
+       case VIRTIO_IOMMU_FAULT_R_MAPPING:
+               reason_str = "page";
+               break;
+       case VIRTIO_IOMMU_FAULT_R_UNKNOWN:
+       default:
+               reason_str = "unknown";
+               break;
+       }
+       /* TODO: find EP by ID and report_iommu_fault */
+       if (flags & VIRTIO_IOMMU_FAULT_F_ADDRESS)
+               dev_err_ratelimited(viommu->dev, "%s fault from EP %u at %#llx [%s%s%s]\n",
+                                   reason_str, endpoint, address,
+                                   flags & VIRTIO_IOMMU_FAULT_F_READ ? "R" : "",
+                                   flags & VIRTIO_IOMMU_FAULT_F_WRITE ? "W" : "",
+                                   flags & VIRTIO_IOMMU_FAULT_F_EXEC ? "X" : "");
+       else
+               dev_err_ratelimited(viommu->dev, "%s fault from EP %u\n",
+                                   reason_str, endpoint);
+       return 0;
+ }
+ static void viommu_event_handler(struct virtqueue *vq)
+ {
+       int ret;
+       unsigned int len;
+       struct scatterlist sg[1];
+       struct viommu_event *evt;
+       struct viommu_dev *viommu = vq->vdev->priv;
+       while ((evt = virtqueue_get_buf(vq, &len)) != NULL) {
+               if (len > sizeof(*evt)) {
+                       dev_err(viommu->dev,
+                               "invalid event buffer (len %u != %zu)\n",
+                               len, sizeof(*evt));
+               } else if (!(evt->head & VIOMMU_FAULT_RESV_MASK)) {
+                       viommu_fault_handler(viommu, &evt->fault);
+               }
+               sg_init_one(sg, evt, sizeof(*evt));
+               ret = virtqueue_add_inbuf(vq, sg, 1, evt, GFP_ATOMIC);
+               if (ret)
+                       dev_err(viommu->dev, "could not add event buffer\n");
+       }
+       virtqueue_kick(vq);
+ }
+ /* IOMMU API */
+ static struct iommu_domain *viommu_domain_alloc(unsigned type)
+ {
+       struct viommu_domain *vdomain;
+       if (type != IOMMU_DOMAIN_UNMANAGED && type != IOMMU_DOMAIN_DMA)
+               return NULL;
+       vdomain = kzalloc(sizeof(*vdomain), GFP_KERNEL);
+       if (!vdomain)
+               return NULL;
+       mutex_init(&vdomain->mutex);
+       spin_lock_init(&vdomain->mappings_lock);
+       vdomain->mappings = RB_ROOT_CACHED;
+       if (type == IOMMU_DOMAIN_DMA &&
+           iommu_get_dma_cookie(&vdomain->domain)) {
+               kfree(vdomain);
+               return NULL;
+       }
+       return &vdomain->domain;
+ }
+ static int viommu_domain_finalise(struct viommu_dev *viommu,
+                                 struct iommu_domain *domain)
+ {
+       int ret;
+       struct viommu_domain *vdomain = to_viommu_domain(domain);
+       unsigned int max_domain = viommu->domain_bits > 31 ? ~0 :
+                                 (1U << viommu->domain_bits) - 1;
+       vdomain->viommu         = viommu;
+       domain->pgsize_bitmap   = viommu->pgsize_bitmap;
+       domain->geometry        = viommu->geometry;
+       ret = ida_alloc_max(&viommu->domain_ids, max_domain, GFP_KERNEL);
+       if (ret >= 0)
+               vdomain->id = (unsigned int)ret;
+       return ret > 0 ? 0 : ret;
+ }
+ static void viommu_domain_free(struct iommu_domain *domain)
+ {
+       struct viommu_domain *vdomain = to_viommu_domain(domain);
+       iommu_put_dma_cookie(domain);
+       /* Free all remaining mappings (size 2^64) */
+       viommu_del_mappings(vdomain, 0, 0);
+       if (vdomain->viommu)
+               ida_free(&vdomain->viommu->domain_ids, vdomain->id);
+       kfree(vdomain);
+ }
+ static int viommu_attach_dev(struct iommu_domain *domain, struct device *dev)
+ {
+       int i;
+       int ret = 0;
+       struct virtio_iommu_req_attach req;
+       struct iommu_fwspec *fwspec = dev_iommu_fwspec_get(dev);
+       struct viommu_endpoint *vdev = fwspec->iommu_priv;
+       struct viommu_domain *vdomain = to_viommu_domain(domain);
+       mutex_lock(&vdomain->mutex);
+       if (!vdomain->viommu) {
+               /*
+                * Properly initialize the domain now that we know which viommu
+                * owns it.
+                */
+               ret = viommu_domain_finalise(vdev->viommu, domain);
+       } else if (vdomain->viommu != vdev->viommu) {
+               dev_err(dev, "cannot attach to foreign vIOMMU\n");
+               ret = -EXDEV;
+       }
+       mutex_unlock(&vdomain->mutex);
+       if (ret)
+               return ret;
+       /*
+        * In the virtio-iommu device, when attaching the endpoint to a new
+        * domain, it is detached from the old one and, if as as a result the
+        * old domain isn't attached to any endpoint, all mappings are removed
+        * from the old domain and it is freed.
+        *
+        * In the driver the old domain still exists, and its mappings will be
+        * recreated if it gets reattached to an endpoint. Otherwise it will be
+        * freed explicitly.
+        *
+        * vdev->vdomain is protected by group->mutex
+        */
+       if (vdev->vdomain)
+               vdev->vdomain->nr_endpoints--;
+       req = (struct virtio_iommu_req_attach) {
+               .head.type      = VIRTIO_IOMMU_T_ATTACH,
+               .domain         = cpu_to_le32(vdomain->id),
+       };
+       for (i = 0; i < fwspec->num_ids; i++) {
+               req.endpoint = cpu_to_le32(fwspec->ids[i]);
+               ret = viommu_send_req_sync(vdomain->viommu, &req, sizeof(req));
+               if (ret)
+                       return ret;
+       }
+       if (!vdomain->nr_endpoints) {
+               /*
+                * This endpoint is the first to be attached to the domain.
+                * Replay existing mappings (e.g. SW MSI).
+                */
+               ret = viommu_replay_mappings(vdomain);
+               if (ret)
+                       return ret;
+       }
+       vdomain->nr_endpoints++;
+       vdev->vdomain = vdomain;
+       return 0;
+ }
+ static int viommu_map(struct iommu_domain *domain, unsigned long iova,
+                     phys_addr_t paddr, size_t size, int prot)
+ {
+       int ret;
+       int flags;
+       struct virtio_iommu_req_map map;
+       struct viommu_domain *vdomain = to_viommu_domain(domain);
+       flags = (prot & IOMMU_READ ? VIRTIO_IOMMU_MAP_F_READ : 0) |
+               (prot & IOMMU_WRITE ? VIRTIO_IOMMU_MAP_F_WRITE : 0) |
+               (prot & IOMMU_MMIO ? VIRTIO_IOMMU_MAP_F_MMIO : 0);
+       ret = viommu_add_mapping(vdomain, iova, paddr, size, flags);
+       if (ret)
+               return ret;
+       map = (struct virtio_iommu_req_map) {
+               .head.type      = VIRTIO_IOMMU_T_MAP,
+               .domain         = cpu_to_le32(vdomain->id),
+               .virt_start     = cpu_to_le64(iova),
+               .phys_start     = cpu_to_le64(paddr),
+               .virt_end       = cpu_to_le64(iova + size - 1),
+               .flags          = cpu_to_le32(flags),
+       };
+       if (!vdomain->nr_endpoints)
+               return 0;
+       ret = viommu_send_req_sync(vdomain->viommu, &map, sizeof(map));
+       if (ret)
+               viommu_del_mappings(vdomain, iova, size);
+       return ret;
+ }
+ static size_t viommu_unmap(struct iommu_domain *domain, unsigned long iova,
+                          size_t size)
+ {
+       int ret = 0;
+       size_t unmapped;
+       struct virtio_iommu_req_unmap unmap;
+       struct viommu_domain *vdomain = to_viommu_domain(domain);
+       unmapped = viommu_del_mappings(vdomain, iova, size);
+       if (unmapped < size)
+               return 0;
+       /* Device already removed all mappings after detach. */
+       if (!vdomain->nr_endpoints)
+               return unmapped;
+       unmap = (struct virtio_iommu_req_unmap) {
+               .head.type      = VIRTIO_IOMMU_T_UNMAP,
+               .domain         = cpu_to_le32(vdomain->id),
+               .virt_start     = cpu_to_le64(iova),
+               .virt_end       = cpu_to_le64(iova + unmapped - 1),
+       };
+       ret = viommu_add_req(vdomain->viommu, &unmap, sizeof(unmap));
+       return ret ? 0 : unmapped;
+ }
+ static phys_addr_t viommu_iova_to_phys(struct iommu_domain *domain,
+                                      dma_addr_t iova)
+ {
+       u64 paddr = 0;
+       unsigned long flags;
+       struct viommu_mapping *mapping;
+       struct interval_tree_node *node;
+       struct viommu_domain *vdomain = to_viommu_domain(domain);
+       spin_lock_irqsave(&vdomain->mappings_lock, flags);
+       node = interval_tree_iter_first(&vdomain->mappings, iova, iova);
+       if (node) {
+               mapping = container_of(node, struct viommu_mapping, iova);
+               paddr = mapping->paddr + (iova - mapping->iova.start);
+       }
+       spin_unlock_irqrestore(&vdomain->mappings_lock, flags);
+       return paddr;
+ }
+ static void viommu_iotlb_sync(struct iommu_domain *domain)
+ {
+       struct viommu_domain *vdomain = to_viommu_domain(domain);
+       viommu_sync_req(vdomain->viommu);
+ }
+ static void viommu_get_resv_regions(struct device *dev, struct list_head *head)
+ {
+       struct iommu_resv_region *entry, *new_entry, *msi = NULL;
+       struct iommu_fwspec *fwspec = dev_iommu_fwspec_get(dev);
+       struct viommu_endpoint *vdev = fwspec->iommu_priv;
+       int prot = IOMMU_WRITE | IOMMU_NOEXEC | IOMMU_MMIO;
+       list_for_each_entry(entry, &vdev->resv_regions, list) {
+               if (entry->type == IOMMU_RESV_MSI)
+                       msi = entry;
+               new_entry = kmemdup(entry, sizeof(*entry), GFP_KERNEL);
+               if (!new_entry)
+                       return;
+               list_add_tail(&new_entry->list, head);
+       }
+       /*
+        * If the device didn't register any bypass MSI window, add a
+        * software-mapped region.
+        */
+       if (!msi) {
+               msi = iommu_alloc_resv_region(MSI_IOVA_BASE, MSI_IOVA_LENGTH,
+                                             prot, IOMMU_RESV_SW_MSI);
+               if (!msi)
+                       return;
+               list_add_tail(&msi->list, head);
+       }
+       iommu_dma_get_resv_regions(dev, head);
+ }
+ static void viommu_put_resv_regions(struct device *dev, struct list_head *head)
+ {
+       struct iommu_resv_region *entry, *next;
+       list_for_each_entry_safe(entry, next, head, list)
+               kfree(entry);
+ }
+ static struct iommu_ops viommu_ops;
+ static struct virtio_driver virtio_iommu_drv;
++static int viommu_match_node(struct device *dev, const void *data)
+ {
+       return dev->parent->fwnode == data;
+ }
+ static struct viommu_dev *viommu_get_by_fwnode(struct fwnode_handle *fwnode)
+ {
+       struct device *dev = driver_find_device(&virtio_iommu_drv.driver, NULL,
+                                               fwnode, viommu_match_node);
+       put_device(dev);
+       return dev ? dev_to_virtio(dev)->priv : NULL;
+ }
+ static int viommu_add_device(struct device *dev)
+ {
+       int ret;
+       struct iommu_group *group;
+       struct viommu_endpoint *vdev;
+       struct viommu_dev *viommu = NULL;
+       struct iommu_fwspec *fwspec = dev_iommu_fwspec_get(dev);
+       if (!fwspec || fwspec->ops != &viommu_ops)
+               return -ENODEV;
+       viommu = viommu_get_by_fwnode(fwspec->iommu_fwnode);
+       if (!viommu)
+               return -ENODEV;
+       vdev = kzalloc(sizeof(*vdev), GFP_KERNEL);
+       if (!vdev)
+               return -ENOMEM;
+       vdev->dev = dev;
+       vdev->viommu = viommu;
+       INIT_LIST_HEAD(&vdev->resv_regions);
+       fwspec->iommu_priv = vdev;
+       if (viommu->probe_size) {
+               /* Get additional information for this endpoint */
+               ret = viommu_probe_endpoint(viommu, dev);
+               if (ret)
+                       goto err_free_dev;
+       }
+       ret = iommu_device_link(&viommu->iommu, dev);
+       if (ret)
+               goto err_free_dev;
+       /*
+        * Last step creates a default domain and attaches to it. Everything
+        * must be ready.
+        */
+       group = iommu_group_get_for_dev(dev);
+       if (IS_ERR(group)) {
+               ret = PTR_ERR(group);
+               goto err_unlink_dev;
+       }
+       iommu_group_put(group);
+       return PTR_ERR_OR_ZERO(group);
+ err_unlink_dev:
+       iommu_device_unlink(&viommu->iommu, dev);
+ err_free_dev:
+       viommu_put_resv_regions(dev, &vdev->resv_regions);
+       kfree(vdev);
+       return ret;
+ }
+ static void viommu_remove_device(struct device *dev)
+ {
+       struct viommu_endpoint *vdev;
+       struct iommu_fwspec *fwspec = dev_iommu_fwspec_get(dev);
+       if (!fwspec || fwspec->ops != &viommu_ops)
+               return;
+       vdev = fwspec->iommu_priv;
+       iommu_group_remove_device(dev);
+       iommu_device_unlink(&vdev->viommu->iommu, dev);
+       viommu_put_resv_regions(dev, &vdev->resv_regions);
+       kfree(vdev);
+ }
+ static struct iommu_group *viommu_device_group(struct device *dev)
+ {
+       if (dev_is_pci(dev))
+               return pci_device_group(dev);
+       else
+               return generic_device_group(dev);
+ }
+ static int viommu_of_xlate(struct device *dev, struct of_phandle_args *args)
+ {
+       return iommu_fwspec_add_ids(dev, args->args, 1);
+ }
+ static struct iommu_ops viommu_ops = {
+       .domain_alloc           = viommu_domain_alloc,
+       .domain_free            = viommu_domain_free,
+       .attach_dev             = viommu_attach_dev,
+       .map                    = viommu_map,
+       .unmap                  = viommu_unmap,
+       .iova_to_phys           = viommu_iova_to_phys,
+       .iotlb_sync             = viommu_iotlb_sync,
+       .add_device             = viommu_add_device,
+       .remove_device          = viommu_remove_device,
+       .device_group           = viommu_device_group,
+       .get_resv_regions       = viommu_get_resv_regions,
+       .put_resv_regions       = viommu_put_resv_regions,
+       .of_xlate               = viommu_of_xlate,
+ };
+ static int viommu_init_vqs(struct viommu_dev *viommu)
+ {
+       struct virtio_device *vdev = dev_to_virtio(viommu->dev);
+       const char *names[] = { "request", "event" };
+       vq_callback_t *callbacks[] = {
+               NULL, /* No async requests */
+               viommu_event_handler,
+       };
+       return virtio_find_vqs(vdev, VIOMMU_NR_VQS, viommu->vqs, callbacks,
+                              names, NULL);
+ }
+ static int viommu_fill_evtq(struct viommu_dev *viommu)
+ {
+       int i, ret;
+       struct scatterlist sg[1];
+       struct viommu_event *evts;
+       struct virtqueue *vq = viommu->vqs[VIOMMU_EVENT_VQ];
+       size_t nr_evts = vq->num_free;
+       viommu->evts = evts = devm_kmalloc_array(viommu->dev, nr_evts,
+                                                sizeof(*evts), GFP_KERNEL);
+       if (!evts)
+               return -ENOMEM;
+       for (i = 0; i < nr_evts; i++) {
+               sg_init_one(sg, &evts[i], sizeof(*evts));
+               ret = virtqueue_add_inbuf(vq, sg, 1, &evts[i], GFP_KERNEL);
+               if (ret)
+                       return ret;
+       }
+       return 0;
+ }
+ static int viommu_probe(struct virtio_device *vdev)
+ {
+       struct device *parent_dev = vdev->dev.parent;
+       struct viommu_dev *viommu = NULL;
+       struct device *dev = &vdev->dev;
+       u64 input_start = 0;
+       u64 input_end = -1UL;
+       int ret;
+       if (!virtio_has_feature(vdev, VIRTIO_F_VERSION_1) ||
+           !virtio_has_feature(vdev, VIRTIO_IOMMU_F_MAP_UNMAP))
+               return -ENODEV;
+       viommu = devm_kzalloc(dev, sizeof(*viommu), GFP_KERNEL);
+       if (!viommu)
+               return -ENOMEM;
+       spin_lock_init(&viommu->request_lock);
+       ida_init(&viommu->domain_ids);
+       viommu->dev = dev;
+       viommu->vdev = vdev;
+       INIT_LIST_HEAD(&viommu->requests);
+       ret = viommu_init_vqs(viommu);
+       if (ret)
+               return ret;
+       virtio_cread(vdev, struct virtio_iommu_config, page_size_mask,
+                    &viommu->pgsize_bitmap);
+       if (!viommu->pgsize_bitmap) {
+               ret = -EINVAL;
+               goto err_free_vqs;
+       }
+       viommu->domain_bits = 32;
+       /* Optional features */
+       virtio_cread_feature(vdev, VIRTIO_IOMMU_F_INPUT_RANGE,
+                            struct virtio_iommu_config, input_range.start,
+                            &input_start);
+       virtio_cread_feature(vdev, VIRTIO_IOMMU_F_INPUT_RANGE,
+                            struct virtio_iommu_config, input_range.end,
+                            &input_end);
+       virtio_cread_feature(vdev, VIRTIO_IOMMU_F_DOMAIN_BITS,
+                            struct virtio_iommu_config, domain_bits,
+                            &viommu->domain_bits);
+       virtio_cread_feature(vdev, VIRTIO_IOMMU_F_PROBE,
+                            struct virtio_iommu_config, probe_size,
+                            &viommu->probe_size);
+       viommu->geometry = (struct iommu_domain_geometry) {
+               .aperture_start = input_start,
+               .aperture_end   = input_end,
+               .force_aperture = true,
+       };
+       viommu_ops.pgsize_bitmap = viommu->pgsize_bitmap;
+       virtio_device_ready(vdev);
+       /* Populate the event queue with buffers */
+       ret = viommu_fill_evtq(viommu);
+       if (ret)
+               goto err_free_vqs;
+       ret = iommu_device_sysfs_add(&viommu->iommu, dev, NULL, "%s",
+                                    virtio_bus_name(vdev));
+       if (ret)
+               goto err_free_vqs;
+       iommu_device_set_ops(&viommu->iommu, &viommu_ops);
+       iommu_device_set_fwnode(&viommu->iommu, parent_dev->fwnode);
+       iommu_device_register(&viommu->iommu);
+ #ifdef CONFIG_PCI
+       if (pci_bus_type.iommu_ops != &viommu_ops) {
+               pci_request_acs();
+               ret = bus_set_iommu(&pci_bus_type, &viommu_ops);
+               if (ret)
+                       goto err_unregister;
+       }
+ #endif
+ #ifdef CONFIG_ARM_AMBA
+       if (amba_bustype.iommu_ops != &viommu_ops) {
+               ret = bus_set_iommu(&amba_bustype, &viommu_ops);
+               if (ret)
+                       goto err_unregister;
+       }
+ #endif
+       if (platform_bus_type.iommu_ops != &viommu_ops) {
+               ret = bus_set_iommu(&platform_bus_type, &viommu_ops);
+               if (ret)
+                       goto err_unregister;
+       }
+       vdev->priv = viommu;
+       dev_info(dev, "input address: %u bits\n",
+                order_base_2(viommu->geometry.aperture_end));
+       dev_info(dev, "page mask: %#llx\n", viommu->pgsize_bitmap);
+       return 0;
+ err_unregister:
+       iommu_device_sysfs_remove(&viommu->iommu);
+       iommu_device_unregister(&viommu->iommu);
+ err_free_vqs:
+       vdev->config->del_vqs(vdev);
+       return ret;
+ }
+ static void viommu_remove(struct virtio_device *vdev)
+ {
+       struct viommu_dev *viommu = vdev->priv;
+       iommu_device_sysfs_remove(&viommu->iommu);
+       iommu_device_unregister(&viommu->iommu);
+       /* Stop all virtqueues */
+       vdev->config->reset(vdev);
+       vdev->config->del_vqs(vdev);
+       dev_info(&vdev->dev, "device removed\n");
+ }
+ static void viommu_config_changed(struct virtio_device *vdev)
+ {
+       dev_warn(&vdev->dev, "config changed\n");
+ }
+ static unsigned int features[] = {
+       VIRTIO_IOMMU_F_MAP_UNMAP,
+       VIRTIO_IOMMU_F_DOMAIN_BITS,
+       VIRTIO_IOMMU_F_INPUT_RANGE,
+       VIRTIO_IOMMU_F_PROBE,
+ };
+ static struct virtio_device_id id_table[] = {
+       { VIRTIO_ID_IOMMU, VIRTIO_DEV_ANY_ID },
+       { 0 },
+ };
+ static struct virtio_driver virtio_iommu_drv = {
+       .driver.name            = KBUILD_MODNAME,
+       .driver.owner           = THIS_MODULE,
+       .id_table               = id_table,
+       .feature_table          = features,
+       .feature_table_size     = ARRAY_SIZE(features),
+       .probe                  = viommu_probe,
+       .remove                 = viommu_remove,
+       .config_changed         = viommu_config_changed,
+ };
+ module_virtio_driver(virtio_iommu_drv);
+ MODULE_DESCRIPTION("Virtio IOMMU driver");
+ MODULE_AUTHOR("Jean-Philippe Brucker <jean-philippe.brucker@arm.com>");
+ MODULE_LICENSE("GPL v2");
Simple merge
Simple merge
Simple merge