Re: [PATCH v1] vhost-vdpa : add support for vIOMMU

[Eugenio Perez Martin]

On Thu, Oct 6, 2022 at 7:44 AM Cindy Lu <lulu@redhat.com> wrote:
>
> Add support for vIOMMU. Register a memory listener to dma_as in
> vhost_vdpa_dev_start
> - during region_add register a specific IOMMU notifier, and store all 
> notifiers in a list.
> - during region_del, compare and delete the IOMMU notifier from the list
> - also change the IOTLB batch flag to support IOMMU batch send
>
> Verified in vp_vdpa and vdpa_sim_net driver
>
> Signed-off-by: Cindy Lu <lulu@redhat.com>
> ---
>  hw/virtio/vhost-vdpa.c         | 253 +++++++++++++++++++++++++++++++--
>  include/hw/virtio/vhost-vdpa.h |  26 +++-
>  2 files changed, 267 insertions(+), 12 deletions(-)
>
> diff --git a/hw/virtio/vhost-vdpa.c b/hw/virtio/vhost-vdpa.c
> index 3ff9ce3501..d2ac40c261 100644
> --- a/hw/virtio/vhost-vdpa.c
> +++ b/hw/virtio/vhost-vdpa.c
> @@ -26,6 +26,7 @@
>  #include "cpu.h"
>  #include "trace.h"
>  #include "qapi/error.h"
> +#include "hw/virtio/virtio-access.h"
>
>  /*
>   * Return one past the end of the end of section. Be careful with uint64_t
> @@ -136,14 +137,15 @@ static void vhost_vdpa_listener_begin_batch(struct 
> vhost_vdpa *v)
>      }
>  }
>
> -static void vhost_vdpa_iotlb_batch_begin_once(struct vhost_vdpa *v)
> +static void vhost_vdpa_iotlb_batch_begin_once(struct vhost_vdpa *v,
> +                                              enum iotlb_batch_flag flag)
>  {
>      if (v->dev->backend_cap & (0x1ULL << VHOST_BACKEND_F_IOTLB_BATCH) &&
>          !v->iotlb_batch_begin_sent) {
>          vhost_vdpa_listener_begin_batch(v);
>      }
>
> -    v->iotlb_batch_begin_sent = true;
> +    v->iotlb_batch_begin_sent |= flag;
>  }
>
>  static void vhost_vdpa_listener_commit(MemoryListener *listener)
> @@ -157,7 +159,7 @@ static void vhost_vdpa_listener_commit(MemoryListener 
> *listener)
>          return;
>      }
>
> -    if (!v->iotlb_batch_begin_sent) {
> +    if (!(v->iotlb_batch_begin_sent & VDPA_IOTLB_BATCH_SEND)) {
>          return;
>      }
>
> @@ -169,8 +171,7 @@ static void vhost_vdpa_listener_commit(MemoryListener 
> *listener)
>          error_report("failed to write, fd=%d, errno=%d (%s)",
>                       fd, errno, strerror(errno));
>      }
> -
> -    v->iotlb_batch_begin_sent = false;
> +    v->iotlb_batch_begin_sent &= ~VDPA_IOTLB_BATCH_SEND;
>  }
>
>  static void vhost_vdpa_listener_region_add(MemoryListener *listener,
> @@ -186,6 +187,9 @@ static void vhost_vdpa_listener_region_add(MemoryListener 
> *listener,
>                                              v->iova_range.last)) {
>          return;
>      }
> +    if (memory_region_is_iommu(section->mr)) {
> +        return;
> +    }
>
>      if (unlikely((section->offset_within_address_space & ~TARGET_PAGE_MASK) 
> !=
>                   (section->offset_within_region & ~TARGET_PAGE_MASK))) {
> @@ -227,9 +231,9 @@ static void vhost_vdpa_listener_region_add(MemoryListener 
> *listener,
>          iova = mem_region.iova;
>      }
>
> -    vhost_vdpa_iotlb_batch_begin_once(v);
> -    ret = vhost_vdpa_dma_map(v, iova, int128_get64(llsize),
> -                             vaddr, section->readonly);
> +    vhost_vdpa_iotlb_batch_begin_once(v, VDPA_IOTLB_BATCH_SEND);
> +    ret = vhost_vdpa_dma_map(v, iova, int128_get64(llsize), vaddr,
> +                             section->readonly);
>      if (ret) {
>          error_report("vhost vdpa map fail!");
>          goto fail;
> @@ -260,6 +264,9 @@ static void vhost_vdpa_listener_region_del(MemoryListener 
> *listener,
>                                              v->iova_range.last)) {
>          return;
>      }
> +    if (memory_region_is_iommu(section->mr)) {
> +        return;
> +    }
>
>      if (unlikely((section->offset_within_address_space & ~TARGET_PAGE_MASK) 
> !=
>                   (section->offset_within_region & ~TARGET_PAGE_MASK))) {
> @@ -292,7 +299,7 @@ static void vhost_vdpa_listener_region_del(MemoryListener 
> *listener,
>          iova = result->iova;
>          vhost_iova_tree_remove(v->iova_tree, result);
>      }
> -    vhost_vdpa_iotlb_batch_begin_once(v);
> +    vhost_vdpa_iotlb_batch_begin_once(v, VDPA_IOTLB_BATCH_SEND);
>      ret = vhost_vdpa_dma_unmap(v, iova, int128_get64(llsize));
>      if (ret) {
>          error_report("vhost_vdpa dma unmap error!");
> @@ -312,6 +319,212 @@ static const MemoryListener vhost_vdpa_memory_listener 
> = {
>      .region_del = vhost_vdpa_listener_region_del,
>  };
>
> +static void vhost_vdpa_listener_iommu_commit(MemoryListener *listener)
> +{
> +    struct vhost_vdpa *v =
> +        container_of(listener, struct vhost_vdpa, iommu_listener);
> +    struct vhost_dev *dev = v->dev;
> +    struct vhost_msg_v2 msg = {};
> +    int fd = v->device_fd;
> +
> +    if (!(dev->backend_cap & (0x1ULL << VHOST_BACKEND_F_IOTLB_BATCH))) {
> +        return;
> +    }
> +
> +    if (!(v->iotlb_batch_begin_sent & VDPA_IOTLB_BATCH_IOMMU_SEND)) {
> +        return;
> +    }
> +
> +    msg.type = v->msg_type;
> +    msg.iotlb.type = VHOST_IOTLB_BATCH_END;
> +
> +    trace_vhost_vdpa_listener_commit(v, fd, msg.type, msg.iotlb.type);
> +    if (write(fd, &msg, sizeof(msg)) != sizeof(msg)) {
> +        error_report("failed to write, fd=%d, errno=%d (%s)", fd, errno,
> +                     strerror(errno));
> +    }
> +    v->iotlb_batch_begin_sent &= ~VDPA_IOTLB_BATCH_IOMMU_SEND;
> +}
> +
> +/* Called with rcu_read_lock held.  */
> +static bool vhost_vdpa_get_xlat_addr(IOMMUTLBEntry *iotlb, void **vaddr,
> +                                     ram_addr_t *ram_addr, bool *read_only)
> +{
> +    MemoryRegion *mr;
> +    hwaddr xlat;
> +    hwaddr len = iotlb->addr_mask + 1;
> +    bool writable = iotlb->perm & IOMMU_WO;
> +
> +    /*
> +     * The IOMMU TLB entry we have just covers translation through
> +     * this IOMMU to its immediate target.  We need to translate
> +     * it the rest of the way through to memory.
> +     */
> +    mr = address_space_translate(&address_space_memory, 
> iotlb->translated_addr,
> +                                 &xlat, &len, writable, 
> MEMTXATTRS_UNSPECIFIED);
> +    if (!memory_region_is_ram(mr)) {
> +        error_report("iommu map to non memory area %" HWADDR_PRIx "", xlat);
> +        return false;
> +    } else if (memory_region_has_ram_discard_manager(mr)) {
> +        RamDiscardManager *rdm = memory_region_get_ram_discard_manager(mr);
> +        MemoryRegionSection tmp = {
> +            .mr = mr,
> +            .offset_within_region = xlat,
> +            .size = int128_make64(len),
> +        };
> +
> +        if (!ram_discard_manager_is_populated(rdm, &tmp)) {
> +            error_report("iommu map to discarded memory (e.g., unplugged via"
> +                         " virtio-mem): %" HWADDR_PRIx "",
> +                         iotlb->translated_addr);
> +            return false;
> +        }
> +    }
> +
> +    /*
> +     * Translation truncates length to the IOMMU page size,
> +     * check that it did not truncate too much.
> +     */
> +    if (len & iotlb->addr_mask) {
> +        error_report("iommu has granularity incompatible with target AS");
> +        return false;
> +    }
> +
> +    if (vaddr) {
> +        *vaddr = memory_region_get_ram_ptr(mr) + xlat;
> +    }
> +
> +    if (ram_addr) {
> +        *ram_addr = memory_region_get_ram_addr(mr) + xlat;
> +    }
> +
> +    if (read_only) {
> +        *read_only = !writable || mr->readonly;
> +    }
> +
> +    return true;
> +}
> +
> +static void vhost_vdpa_iommu_map_notify(IOMMUNotifier *n, IOMMUTLBEntry 
> *iotlb)
> +{
> +    struct vdpa_iommu *iommu = container_of(n, struct vdpa_iommu, n);
> +
> +    hwaddr iova = iotlb->iova + iommu->iommu_offset;
> +    struct vhost_vdpa *v = iommu->dev;
> +    void *vaddr;
> +    int ret;
> +
> +

Extra space here.

> +    if (iotlb->target_as != &address_space_memory) {
> +        error_report("Wrong target AS \"%s\", only system memory is allowed",
> +                     iotlb->target_as->name ? iotlb->target_as->name : 
> "none");
> +        return;
> +    }
> +    rcu_read_lock();
> +    vhost_vdpa_iotlb_batch_begin_once(v, VDPA_IOTLB_BATCH_IOMMU_SEND);
> +
> +    if ((iotlb->perm & IOMMU_RW) != IOMMU_NONE) {
> +        bool read_only;
> +
> +        if (!vhost_vdpa_get_xlat_addr(iotlb, &vaddr, NULL, &read_only)) {
> +            goto out;
> +        }
> +        ret =
> +            vhost_vdpa_dma_map(v, iova, iotlb->addr_mask + 1, vaddr, 
> read_only);
> +        if (ret) {
> +            error_report("vhost_vdpa_dma_map(%p, 0x%" HWADDR_PRIx ", "
> +                         "0x%" HWADDR_PRIx ", %p) = %d (%m)",
> +                         v, iova, iotlb->addr_mask + 1, vaddr, ret);
> +        }
> +    } else {
> +        ret = vhost_vdpa_dma_unmap(v, iova, iotlb->addr_mask + 1);
> +        if (ret) {
> +            error_report("vhost_vdpa_dma_unmap(%p, 0x%" HWADDR_PRIx ", "
> +                         "0x%" HWADDR_PRIx ") = %d (%m)",
> +                         v, iova, iotlb->addr_mask + 1, ret);
> +        }
> +    }
> +out:

Maybe we can simplify this goto with RCU_READ_LOCK_GUARD macro?

> +    rcu_read_unlock();
> +}
> +
> +static void vhost_vdpa_iommu_region_add(MemoryListener *listener,
> +                                        MemoryRegionSection *section)
> +{
> +    struct vhost_vdpa *v =
> +        container_of(listener, struct vhost_vdpa, iommu_listener);
> +
> +    struct vdpa_iommu *iommu;
> +    Int128 end;
> +    int iommu_idx;
> +    IOMMUMemoryRegion *iommu_mr;
> +    int ret;
> +
> +    if (!memory_region_is_iommu(section->mr)) {
> +        return;
> +    }
> +
> +    iommu_mr = IOMMU_MEMORY_REGION(section->mr);
> +
> +    iommu = g_malloc0(sizeof(*iommu));
> +    end =
> +        int128_add(int128_make64(section->offset_within_region), 
> section->size);
> +    end = int128_sub(end, int128_one());
> +    iommu_idx =
> +        memory_region_iommu_attrs_to_index(iommu_mr, MEMTXATTRS_UNSPECIFIED);
> +    iommu->iommu_mr = iommu_mr;
> +
> +    iommu_notifier_init(
> +        &iommu->n, vhost_vdpa_iommu_map_notify, IOMMU_NOTIFIER_IOTLB_EVENTS,
> +        section->offset_within_region, int128_get64(end), iommu_idx);
> +    iommu->iommu_offset =
> +        section->offset_within_address_space - section->offset_within_region;
> +    iommu->dev = v;
> +
> +

Extra newline.

> +    ret = memory_region_register_iommu_notifier(section->mr, &iommu->n, 
> NULL);
> +    if (ret) {
> +        g_free(iommu);
> +        return;
> +    }
> +
> +    QLIST_INSERT_HEAD(&v->iommu_list, iommu, iommu_next);
> +    memory_region_iommu_replay(iommu->iommu_mr, &iommu->n);
> +
> +    return;
> +}
> +
> +static void vhost_vdpa_iommu_region_del(MemoryListener *listener,
> +                                        MemoryRegionSection *section)
> +{
> +    struct vhost_vdpa *v =
> +        container_of(listener, struct vhost_vdpa, iommu_listener);
> +
> +    struct vdpa_iommu *iommu;
> +
> +    if (!memory_region_is_iommu(section->mr)) {
> +        return;
> +    }
> +
> +    QLIST_FOREACH(iommu, &v->iommu_list, iommu_next)
> +    {
> +        if (MEMORY_REGION(iommu->iommu_mr) == section->mr &&
> +            iommu->n.start == section->offset_within_region) {
> +            memory_region_unregister_iommu_notifier(section->mr, &iommu->n);
> +            QLIST_REMOVE(iommu, iommu_next);
> +            g_free(iommu);
> +            break;
> +        }
> +    }
> +}
> +
> +static const MemoryListener vhost_vdpa_iommu_listener = {
> +    .name = "vhost-vdpa-iommu",
> +    .region_add = vhost_vdpa_iommu_region_add,
> +    .region_del = vhost_vdpa_iommu_region_del,
> +    .commit = vhost_vdpa_listener_iommu_commit,
> +};
> +
>  static int vhost_vdpa_call(struct vhost_dev *dev, unsigned long int request,
>                               void *arg)
>  {
> @@ -432,6 +645,17 @@ static int vhost_vdpa_init_svq(struct vhost_dev *hdev, 
> struct vhost_vdpa *v,
>      v->shadow_vqs = g_steal_pointer(&shadow_vqs);
>      return 0;
>  }
> +static bool vdpa_dev_has_iommu(struct vhost_dev *dev)
> +{
> +    VirtIODevice *vdev = dev->vdev;
> +
> +    if (vdev) {
> +        return virtio_bus_device_iommu_enabled(vdev) &&
> +               virtio_host_has_feature(vdev, VIRTIO_F_IOMMU_PLATFORM);
> +    } else {
> +        return false;
> +    }
> +}
>
>  static int vhost_vdpa_init(struct vhost_dev *dev, void *opaque, Error **errp)
>  {
> @@ -454,6 +678,7 @@ static int vhost_vdpa_init(struct vhost_dev *dev, void 
> *opaque, Error **errp)
>      v->dev = dev;
>      dev->opaque =  opaque ;
>      v->listener = vhost_vdpa_memory_listener;
> +    v->iommu_listener = vhost_vdpa_iommu_listener;
>      v->msg_type = VHOST_IOTLB_MSG_V2;
>      ret = vhost_vdpa_init_svq(dev, v, errp);
>      if (ret) {
> @@ -587,7 +812,6 @@ static int vhost_vdpa_cleanup(struct vhost_dev *dev)
>      v = dev->opaque;
>      trace_vhost_vdpa_cleanup(dev, v);
>      vhost_vdpa_host_notifiers_uninit(dev, dev->nvqs);
> -    memory_listener_unregister(&v->listener);
>      vhost_vdpa_svq_cleanup(dev);
>
>      dev->opaque = NULL;
> @@ -1128,11 +1352,18 @@ static int vhost_vdpa_dev_start(struct vhost_dev 
> *dev, bool started)
>
>      if (started) {
>          memory_listener_register(&v->listener, &address_space_memory);
> +        if (vdpa_dev_has_iommu(dev)) {
> +            memory_listener_register(&v->iommu_listener, dev->vdev->dma_as);
> +        }
> +
>          return vhost_vdpa_add_status(dev, VIRTIO_CONFIG_S_DRIVER_OK);
>      } else {
>          vhost_vdpa_reset_device(dev);
>          vhost_vdpa_add_status(dev, VIRTIO_CONFIG_S_ACKNOWLEDGE |
> -                                   VIRTIO_CONFIG_S_DRIVER);
> +                                       VIRTIO_CONFIG_S_DRIVER);

Why this extra indentation?

> +        if (vdpa_dev_has_iommu(dev)) {
> +            memory_listener_unregister(&v->iommu_listener);
> +        }
>          memory_listener_unregister(&v->listener);
>
>          return 0;
> diff --git a/include/hw/virtio/vhost-vdpa.h b/include/hw/virtio/vhost-vdpa.h
> index d10a89303e..4fcd144c06 100644
> --- a/include/hw/virtio/vhost-vdpa.h
> +++ b/include/hw/virtio/vhost-vdpa.h
> @@ -24,12 +24,25 @@ typedef struct VhostVDPAHostNotifier {
>      void *addr;
>  } VhostVDPAHostNotifier;
>
> +

Extra newline.

> +typedef enum iotlb_batch_flag {
> +    /* Notify IOTLB_BATCH start*/
> +    VDPA_IOTLB_BATCH_SEND = 0x1,
> +    /* Notify IOTLB_BATCH iommu start*/
> +    VDPA_IOTLB_BATCH_IOMMU_SEND = 0x2,
> +    /* Notify IOTLB_BATCH stop*/
> +    VDPA_IOTLB_BATCH_SEND_STOP = 0x4,
> +    /* Notify IOTLB_BATCH iommu stop*/
> +    VDPA_IOTLB_BATCH_IOMMU_SEND_STOP = 0x08,

It seems to me that VDPA_IOTLB_BATCH_SEND_STOP and
VDPA_IOTLB_BATCH_IOMMU_SEND_STOP are never used in the patch.

Also, I'm not sure about the interactions between them. We should have
either only a batch of GPA->QEMU VA or a batch of GVA->QEMU VA, but
never to interleave them. So I would either go back to a boolean or I
would check if they are interleaving, something I think must not
happen in a sane QEMU.

> +} IotlbBatchFlag;
> +
>  typedef struct vhost_vdpa {
>      int device_fd;
>      int index;
>      uint32_t msg_type;
> -    bool iotlb_batch_begin_sent;
> +    uint32_t iotlb_batch_begin_sent;
>      MemoryListener listener;
> +    MemoryListener iommu_listener;
>      struct vhost_vdpa_iova_range iova_range;
>      uint64_t acked_features;
>      bool shadow_vqs_enabled;
> @@ -41,8 +54,19 @@ typedef struct vhost_vdpa {
>      void *shadow_vq_ops_opaque;
>      struct vhost_dev *dev;
>      VhostVDPAHostNotifier notifier[VIRTIO_QUEUE_MAX];
> +    QLIST_HEAD(, vdpa_iommu) iommu_list;
> +    IOMMUNotifier n;
> +

Extra newline.

Thanks!

>  } VhostVDPA;
>
> +struct vdpa_iommu {
> +    struct vhost_vdpa *dev;
> +    IOMMUMemoryRegion *iommu_mr;
> +    hwaddr iommu_offset;
> +    IOMMUNotifier n;
> +    QLIST_ENTRY(vdpa_iommu) iommu_next;
> +};
> +
>  int vhost_vdpa_dma_map(struct vhost_vdpa *v, hwaddr iova, hwaddr size,
>                         void *vaddr, bool readonly);
>  int vhost_vdpa_dma_unmap(struct vhost_vdpa *v, hwaddr iova, hwaddr size);
> --
> 2.34.3
>
>