[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [Qemu-arm] [PATCH for-4.2 v10 11/15] virtio-iommu: Expose the IOAPIC
From: |
Michael S. Tsirkin |
Subject: |
Re: [Qemu-arm] [PATCH for-4.2 v10 11/15] virtio-iommu: Expose the IOAPIC MSI reserved region when relevant |
Date: |
Tue, 30 Jul 2019 15:38:08 -0400 |
On Tue, Jul 30, 2019 at 07:21:33PM +0200, Eric Auger wrote:
> We introduce a new msi_bypass field which indicates whether
> the IOAPIC MSI window [0xFEE00000 - 0xFEEFFFFF] must be exposed
> as a reserved region. By default the field is set to true at
> instantiation time. Later on we will introduce a property at
> virtio pci proxy level to turn it off.
>
> Signed-off-by: Eric Auger <address@hidden>
>
> ---
>
> v8 -> v9:
> - pass IOAPIC_RANGE_END to virtio_iommu_register_resv_region
> - take into account the change in the struct virtio_iommu_probe_resv_mem
> definition
> - We just introduce the field here. A property will be introduced later on
> at pci proxy level.
> ---
> hw/virtio/virtio-iommu.c | 36 ++++++++++++++++++++++++++++++++
> include/hw/virtio/virtio-iommu.h | 1 +
> 2 files changed, 37 insertions(+)
>
> diff --git a/hw/virtio/virtio-iommu.c b/hw/virtio/virtio-iommu.c
> index 66be9a4627..74038288b0 100644
> --- a/hw/virtio/virtio-iommu.c
> +++ b/hw/virtio/virtio-iommu.c
> @@ -39,6 +39,9 @@
> #define VIOMMU_DEFAULT_QUEUE_SIZE 256
> #define VIOMMU_PROBE_SIZE 512
>
> +#define IOAPIC_RANGE_START (0xfee00000)
> +#define IOAPIC_RANGE_END (0xfeefffff)
> +
> #define SUPPORTED_PROBE_PROPERTIES (\
> 1 << VIRTIO_IOMMU_PROBE_T_RESV_MEM)
>
Sorry where are these numbers coming from?
Does this really work on all platforms?
With all guests?
> @@ -100,6 +103,30 @@ static void
> virtio_iommu_detach_endpoint_from_domain(viommu_endpoint *ep)
> ep->domain = NULL;
> }
>
> +static void virtio_iommu_register_resv_region(viommu_endpoint *ep,
> + uint8_t subtype,
> + uint64_t start, uint64_t end)
> +{
> + viommu_interval *interval;
> + struct virtio_iommu_probe_resv_mem *resv_reg_prop;
> + size_t prop_size = sizeof(struct virtio_iommu_probe_resv_mem);
> + size_t value_size = prop_size -
> + sizeof(struct virtio_iommu_probe_property);
> +
> + interval = g_malloc0(sizeof(*interval));
> + interval->low = start;
> + interval->high = end;
> +
> + resv_reg_prop = g_malloc0(prop_size);
> + resv_reg_prop->head.type = VIRTIO_IOMMU_PROBE_T_RESV_MEM;
> + resv_reg_prop->head.length = cpu_to_le64(value_size);
> + resv_reg_prop->subtype = cpu_to_le64(subtype);
> + resv_reg_prop->start = cpu_to_le64(start);
> + resv_reg_prop->end = cpu_to_le64(end);
> +
> + g_tree_insert(ep->reserved_regions, interval, resv_reg_prop);
> +}
> +
> static viommu_endpoint *virtio_iommu_get_endpoint(VirtIOIOMMU *s,
> uint32_t ep_id)
> {
> @@ -117,6 +144,12 @@ static viommu_endpoint
> *virtio_iommu_get_endpoint(VirtIOIOMMU *s,
> ep->reserved_regions = g_tree_new_full((GCompareDataFunc)interval_cmp,
> NULL, (GDestroyNotify)g_free,
> (GDestroyNotify)g_free);
> + if (s->msi_bypass) {
> + virtio_iommu_register_resv_region(ep, VIRTIO_IOMMU_RESV_MEM_T_MSI,
> + IOAPIC_RANGE_START,
> + IOAPIC_RANGE_END);
> + }
> +
> return ep;
> }
>
> @@ -822,6 +855,9 @@ static void virtio_iommu_set_status(VirtIODevice *vdev,
> uint8_t status)
>
> static void virtio_iommu_instance_init(Object *obj)
> {
> + VirtIOIOMMU *s = VIRTIO_IOMMU(obj);
> +
> + s->msi_bypass = true;
> }
>
> static const VMStateDescription vmstate_virtio_iommu = {
> diff --git a/include/hw/virtio/virtio-iommu.h
> b/include/hw/virtio/virtio-iommu.h
> index f55f48d304..56c8b4e57f 100644
> --- a/include/hw/virtio/virtio-iommu.h
> +++ b/include/hw/virtio/virtio-iommu.h
> @@ -59,6 +59,7 @@ typedef struct VirtIOIOMMU {
> GTree *domains;
> QemuMutex mutex;
> GTree *endpoints;
> + bool msi_bypass;
> } VirtIOIOMMU;
>
> #endif
> --
> 2.20.1
- [Qemu-arm] [PATCH for-4.2 v10 02/15] linux-headers: update against 5.3-rc2, (continued)
- [Qemu-arm] [PATCH for-4.2 v10 02/15] linux-headers: update against 5.3-rc2, Eric Auger, 2019/07/30
- [Qemu-arm] [PATCH for-4.2 v10 03/15] virtio-iommu: Add skeleton, Eric Auger, 2019/07/30
- [Qemu-arm] [PATCH for-4.2 v10 04/15] virtio-iommu: Decode the command payload, Eric Auger, 2019/07/30
- [Qemu-arm] [PATCH for-4.2 v10 05/15] virtio-iommu: Add the iommu regions, Eric Auger, 2019/07/30
- [Qemu-arm] [PATCH for-4.2 v10 06/15] virtio-iommu: Endpoint and domains structs and helpers, Eric Auger, 2019/07/30
- [Qemu-arm] [PATCH for-4.2 v10 07/15] virtio-iommu: Implement attach/detach command, Eric Auger, 2019/07/30
- [Qemu-arm] [PATCH for-4.2 v10 08/15] virtio-iommu: Implement map/unmap, Eric Auger, 2019/07/30
- [Qemu-arm] [PATCH for-4.2 v10 09/15] virtio-iommu: Implement translate, Eric Auger, 2019/07/30
- [Qemu-arm] [PATCH for-4.2 v10 10/15] virtio-iommu: Implement probe request, Eric Auger, 2019/07/30
- [Qemu-arm] [PATCH for-4.2 v10 11/15] virtio-iommu: Expose the IOAPIC MSI reserved region when relevant, Eric Auger, 2019/07/30
- Re: [Qemu-arm] [PATCH for-4.2 v10 11/15] virtio-iommu: Expose the IOAPIC MSI reserved region when relevant,
Michael S. Tsirkin <=
- Re: [Qemu-arm] [PATCH for-4.2 v10 11/15] virtio-iommu: Expose the IOAPIC MSI reserved region when relevant, Tian, Kevin, 2019/07/30
- Re: [Qemu-arm] [PATCH for-4.2 v10 11/15] virtio-iommu: Expose the IOAPIC MSI reserved region when relevant, Auger Eric, 2019/07/31
- Re: [Qemu-arm] [PATCH for-4.2 v10 11/15] virtio-iommu: Expose the IOAPIC MSI reserved region when relevant, Michael S. Tsirkin, 2019/07/31
- Re: [Qemu-arm] [PATCH for-4.2 v10 11/15] virtio-iommu: Expose the IOAPIC MSI reserved region when relevant, Auger Eric, 2019/07/31
- Re: [Qemu-arm] [PATCH for-4.2 v10 11/15] virtio-iommu: Expose the IOAPIC MSI reserved region when relevant, Tian, Kevin, 2019/07/31
[Qemu-arm] [PATCH for-4.2 v10 12/15] virtio-iommu: Implement fault reporting, Eric Auger, 2019/07/30
[Qemu-arm] [PATCH for-4.2 v10 13/15] virtio_iommu: Handle reserved regions in translation process, Eric Auger, 2019/07/30
[Qemu-arm] [PATCH for-4.2 v10 14/15] virtio-iommu-pci: Add virtio iommu pci support, Eric Auger, 2019/07/30
[Qemu-arm] [PATCH for-4.2 v10 15/15] hw/arm/virt: Add the virtio-iommu device tree mappings, Eric Auger, 2019/07/30