diff options
Diffstat (limited to 'drivers/vdpa/vdpa_user/vduse_dev.c')
| -rw-r--r-- | drivers/vdpa/vdpa_user/vduse_dev.c | 180 |
1 files changed, 180 insertions, 0 deletions
diff --git a/drivers/vdpa/vdpa_user/vduse_dev.c b/drivers/vdpa/vdpa_user/vduse_dev.c index 3bc27de58f46..41c0b29739f1 100644 --- a/drivers/vdpa/vdpa_user/vduse_dev.c +++ b/drivers/vdpa/vdpa_user/vduse_dev.c @@ -21,6 +21,8 @@ #include <linux/uio.h> #include <linux/vdpa.h> #include <linux/nospec.h> +#include <linux/vmalloc.h> +#include <linux/sched/mm.h> #include <uapi/linux/vduse.h> #include <uapi/linux/vdpa.h> #include <uapi/linux/virtio_config.h> @@ -64,6 +66,13 @@ struct vduse_vdpa { struct vduse_dev *dev; }; +struct vduse_umem { + unsigned long iova; + unsigned long npages; + struct page **pages; + struct mm_struct *mm; +}; + struct vduse_dev { struct vduse_vdpa *vdev; struct device *dev; @@ -95,6 +104,8 @@ struct vduse_dev { u8 status; u32 vq_num; u32 vq_align; + struct vduse_umem *umem; + struct mutex mem_lock; }; struct vduse_dev_msg { @@ -917,6 +928,102 @@ unlock: return ret; } +static int vduse_dev_dereg_umem(struct vduse_dev *dev, + u64 iova, u64 size) +{ + int ret; + + mutex_lock(&dev->mem_lock); + ret = -ENOENT; + if (!dev->umem) + goto unlock; + + ret = -EINVAL; + if (dev->umem->iova != iova || size != dev->domain->bounce_size) + goto unlock; + + vduse_domain_remove_user_bounce_pages(dev->domain); + unpin_user_pages_dirty_lock(dev->umem->pages, + dev->umem->npages, true); + atomic64_sub(dev->umem->npages, &dev->umem->mm->pinned_vm); + mmdrop(dev->umem->mm); + vfree(dev->umem->pages); + kfree(dev->umem); + dev->umem = NULL; + ret = 0; +unlock: + mutex_unlock(&dev->mem_lock); + return ret; +} + +static int vduse_dev_reg_umem(struct vduse_dev *dev, + u64 iova, u64 uaddr, u64 size) +{ + struct page **page_list = NULL; + struct vduse_umem *umem = NULL; + long pinned = 0; + unsigned long npages, lock_limit; + int ret; + + if (!dev->domain->bounce_map || + size != dev->domain->bounce_size || + iova != 0 || uaddr & ~PAGE_MASK) + return -EINVAL; + + mutex_lock(&dev->mem_lock); + ret = -EEXIST; + if (dev->umem) + goto unlock; + + ret = -ENOMEM; + npages = size >> PAGE_SHIFT; + page_list = __vmalloc(array_size(npages, sizeof(struct page *)), + GFP_KERNEL_ACCOUNT); + umem = kzalloc(sizeof(*umem), GFP_KERNEL); + if (!page_list || !umem) + goto unlock; + + mmap_read_lock(current->mm); + + lock_limit = PFN_DOWN(rlimit(RLIMIT_MEMLOCK)); + if (npages + atomic64_read(¤t->mm->pinned_vm) > lock_limit) + goto out; + + pinned = pin_user_pages(uaddr, npages, FOLL_LONGTERM | FOLL_WRITE, + page_list, NULL); + if (pinned != npages) { + ret = pinned < 0 ? pinned : -ENOMEM; + goto out; + } + + ret = vduse_domain_add_user_bounce_pages(dev->domain, + page_list, pinned); + if (ret) + goto out; + + atomic64_add(npages, ¤t->mm->pinned_vm); + + umem->pages = page_list; + umem->npages = pinned; + umem->iova = iova; + umem->mm = current->mm; + mmgrab(current->mm); + + dev->umem = umem; +out: + if (ret && pinned > 0) + unpin_user_pages(page_list, pinned); + + mmap_read_unlock(current->mm); +unlock: + if (ret) { + vfree(page_list); + kfree(umem); + } + mutex_unlock(&dev->mem_lock); + return ret; +} + static long vduse_dev_ioctl(struct file *file, unsigned int cmd, unsigned long arg) { @@ -1089,6 +1196,77 @@ static long vduse_dev_ioctl(struct file *file, unsigned int cmd, ret = vduse_dev_queue_irq_work(dev, &dev->vqs[index].inject); break; } + case VDUSE_IOTLB_REG_UMEM: { + struct vduse_iova_umem umem; + + ret = -EFAULT; + if (copy_from_user(&umem, argp, sizeof(umem))) + break; + + ret = -EINVAL; + if (!is_mem_zero((const char *)umem.reserved, + sizeof(umem.reserved))) + break; + + ret = vduse_dev_reg_umem(dev, umem.iova, + umem.uaddr, umem.size); + break; + } + case VDUSE_IOTLB_DEREG_UMEM: { + struct vduse_iova_umem umem; + + ret = -EFAULT; + if (copy_from_user(&umem, argp, sizeof(umem))) + break; + + ret = -EINVAL; + if (!is_mem_zero((const char *)umem.reserved, + sizeof(umem.reserved))) + break; + + ret = vduse_dev_dereg_umem(dev, umem.iova, + umem.size); + break; + } + case VDUSE_IOTLB_GET_INFO: { + struct vduse_iova_info info; + struct vhost_iotlb_map *map; + struct vduse_iova_domain *domain = dev->domain; + + ret = -EFAULT; + if (copy_from_user(&info, argp, sizeof(info))) + break; + + ret = -EINVAL; + if (info.start > info.last) + break; + + if (!is_mem_zero((const char *)info.reserved, + sizeof(info.reserved))) + break; + + spin_lock(&domain->iotlb_lock); + map = vhost_iotlb_itree_first(domain->iotlb, + info.start, info.last); + if (map) { + info.start = map->start; + info.last = map->last; + info.capability = 0; + if (domain->bounce_map && map->start == 0 && + map->last == domain->bounce_size - 1) + info.capability |= VDUSE_IOVA_CAP_UMEM; + } + spin_unlock(&domain->iotlb_lock); + if (!map) + break; + + ret = -EFAULT; + if (copy_to_user(argp, &info, sizeof(info))) + break; + + ret = 0; + break; + } default: ret = -ENOIOCTLCMD; break; @@ -1101,6 +1279,7 @@ static int vduse_dev_release(struct inode *inode, struct file *file) { struct vduse_dev *dev = file->private_data; + vduse_dev_dereg_umem(dev, 0, dev->domain->bounce_size); spin_lock(&dev->msg_lock); /* Make sure the inflight messages can processed after reconncection */ list_splice_init(&dev->recv_list, &dev->send_list); @@ -1163,6 +1342,7 @@ static struct vduse_dev *vduse_dev_create(void) return NULL; mutex_init(&dev->lock); + mutex_init(&dev->mem_lock); spin_lock_init(&dev->msg_lock); INIT_LIST_HEAD(&dev->send_list); INIT_LIST_HEAD(&dev->recv_list); |