[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [Qemu-devel] [PATCH V2 07/11] virtio-pci: address space translation
From: |
Michael S. Tsirkin |
Subject: |
Re: [Qemu-devel] [PATCH V2 07/11] virtio-pci: address space translation service (ATS) support |
Date: |
Thu, 3 Nov 2016 21:49:10 +0200 |
On Thu, Nov 03, 2016 at 05:27:19PM +0800, Jason Wang wrote:
> This patches enable the Address Translation Service support for virtio
> pci devices. This is needed for a guest visible Device IOTLB
> implementation and will be required by vhost device IOTLB API
> implementation for intel IOMMU.
>
> Cc: Michael S. Tsirkin <address@hidden>
> Signed-off-by: Jason Wang <address@hidden>
I'd like to understand why do you think this is strictly required.
Won't setting CM bit in the IOMMU do the trick.
Also, could you remind me pls - can guests just disable ATS?
What happens then?
> ---
> hw/pci/pcie.c | 16 ++++++++++++++++
> hw/virtio/virtio-pci.c | 7 +++++++
> hw/virtio/virtio-pci.h | 4 ++++
> include/hw/pci/pcie.h | 4 ++++
> include/standard-headers/linux/pci_regs.h | 1 +
> 5 files changed, 32 insertions(+)
>
> diff --git a/hw/pci/pcie.c b/hw/pci/pcie.c
> index 99cfb45..02195d9 100644
> --- a/hw/pci/pcie.c
> +++ b/hw/pci/pcie.c
> @@ -717,3 +717,19 @@ void pcie_dev_ser_num_init(PCIDevice *dev, uint16_t
> offset, uint64_t ser_num)
> PCI_EXT_CAP_DSN_SIZEOF);
> pci_set_quad(dev->config + offset + pci_dsn_cap, ser_num);
> }
> +
> +void pcie_ats_init(PCIDevice *dev, uint16_t offset)
> +{
> + pcie_add_capability(dev, PCI_EXT_CAP_ID_ATS, 0x1,
> + offset, PCI_EXT_CAP_ATS_SIZEOF);
> +
> + dev->exp.ats_cap = offset;
> +
> + /* Invalidate Queue Depth 0, Page Aligned Request 0 */
> + pci_set_word(dev->config + offset + PCI_ATS_CAP, 0);
> + /* STU 0, Disabled by default */
> + pci_set_word(dev->config + offset + PCI_ATS_CTRL, 0);
> +
> + pci_set_word(dev->wmask + dev->exp.ats_cap + PCI_ATS_CTRL, 0x800f);
> +}
> +
> diff --git a/hw/virtio/virtio-pci.c b/hw/virtio/virtio-pci.c
> index 6ceb43e..e357bdf 100644
> --- a/hw/virtio/virtio-pci.c
> +++ b/hw/virtio/virtio-pci.c
> @@ -1838,6 +1838,11 @@ static void virtio_pci_realize(PCIDevice *pci_dev,
> Error **errp)
> * PCI Power Management Interface Specification.
> */
> pci_set_word(pci_dev->config + pos + PCI_PM_PMC, 0x3);
> +
> + if (proxy->flags & VIRTIO_PCI_FLAG_ATS) {
> + pcie_ats_init(pci_dev, 256);
> + }
> +
> } else {
> /*
> * make future invocations of pci_is_express() return false
> @@ -1889,6 +1894,8 @@ static Property virtio_pci_properties[] = {
> VIRTIO_PCI_FLAG_DISABLE_PCIE_BIT, false),
> DEFINE_PROP_BIT("page-per-vq", VirtIOPCIProxy, flags,
> VIRTIO_PCI_FLAG_PAGE_PER_VQ_BIT, false),
> + DEFINE_PROP_BIT("ats", VirtIOPCIProxy, flags,
> + VIRTIO_PCI_FLAG_ATS_BIT, false),
> DEFINE_PROP_END_OF_LIST(),
> };
>
> diff --git a/hw/virtio/virtio-pci.h b/hw/virtio/virtio-pci.h
> index b4edea6..057d49d 100644
> --- a/hw/virtio/virtio-pci.h
> +++ b/hw/virtio/virtio-pci.h
> @@ -69,6 +69,7 @@ enum {
> VIRTIO_PCI_FLAG_MODERN_PIO_NOTIFY_BIT,
> VIRTIO_PCI_FLAG_DISABLE_PCIE_BIT,
> VIRTIO_PCI_FLAG_PAGE_PER_VQ_BIT,
> + VIRTIO_PCI_FLAG_ATS_BIT,
> };
>
> /* Need to activate work-arounds for buggy guests at vmstate load. */
> @@ -93,6 +94,9 @@ enum {
> #define VIRTIO_PCI_FLAG_PAGE_PER_VQ \
> (1 << VIRTIO_PCI_FLAG_PAGE_PER_VQ_BIT)
>
> +/* address space translation service */
> +#define VIRTIO_PCI_FLAG_ATS (1 << VIRTIO_PCI_FLAG_ATS_BIT)
> +
> typedef struct {
> MSIMessage msg;
> int virq;
> diff --git a/include/hw/pci/pcie.h b/include/hw/pci/pcie.h
> index 056d25e..b08451d 100644
> --- a/include/hw/pci/pcie.h
> +++ b/include/hw/pci/pcie.h
> @@ -74,6 +74,9 @@ struct PCIExpressDevice {
> /* AER */
> uint16_t aer_cap;
> PCIEAERLog aer_log;
> +
> + /* Offset of ATS capability in config space */
> + uint16_t ats_cap;
> };
>
> #define COMPAT_PROP_PCP "power_controller_present"
> @@ -120,6 +123,7 @@ void pcie_add_capability(PCIDevice *dev,
>
> void pcie_ari_init(PCIDevice *dev, uint16_t offset, uint16_t nextfn);
> void pcie_dev_ser_num_init(PCIDevice *dev, uint16_t offset, uint64_t
> ser_num);
> +void pcie_ats_init(PCIDevice *dev, uint16_t offset);
>
> extern const VMStateDescription vmstate_pcie_device;
>
> diff --git a/include/standard-headers/linux/pci_regs.h
> b/include/standard-headers/linux/pci_regs.h
> index 4040951..ac426a0 100644
> --- a/include/standard-headers/linux/pci_regs.h
> +++ b/include/standard-headers/linux/pci_regs.h
> @@ -674,6 +674,7 @@
> #define PCI_EXT_CAP_ID_MAX PCI_EXT_CAP_ID_DPC
>
> #define PCI_EXT_CAP_DSN_SIZEOF 12
> +#define PCI_EXT_CAP_ATS_SIZEOF 8
> #define PCI_EXT_CAP_MCAST_ENDPOINT_SIZEOF 40
>
> /* Advanced Error Reporting */
> --
> 2.7.4
- [Qemu-devel] [PATCH V2 03/11] intel_iommu: name vtd address space with devfn, (continued)
- [Qemu-devel] [PATCH V2 03/11] intel_iommu: name vtd address space with devfn, Jason Wang, 2016/11/03
- [Qemu-devel] [PATCH V2 02/11] virtio: convert to use DMA api, Jason Wang, 2016/11/03
- [Qemu-devel] [PATCH V2 04/11] intel_iommu: allocate new key when creating new address space, Jason Wang, 2016/11/03
- [Qemu-devel] [PATCH V2 05/11] exec: introduce address_space_get_iotlb_entry(), Jason Wang, 2016/11/03
- [Qemu-devel] [PATCH V2 06/11] intel_iommu: support device iotlb descriptor, Jason Wang, 2016/11/03
- [Qemu-devel] [PATCH V2 07/11] virtio-pci: address space translation service (ATS) support, Jason Wang, 2016/11/03
- Re: [Qemu-devel] [PATCH V2 07/11] virtio-pci: address space translation service (ATS) support,
Michael S. Tsirkin <=
- Re: [Qemu-devel] [PATCH V2 07/11] virtio-pci: address space translation service (ATS) support, Jason Wang, 2016/11/04
- Re: [Qemu-devel] [PATCH V2 07/11] virtio-pci: address space translation service (ATS) support, Michael S. Tsirkin, 2016/11/10
- Re: [Qemu-devel] [PATCH V2 07/11] virtio-pci: address space translation service (ATS) support, Jason Wang, 2016/11/10
- Re: [Qemu-devel] [PATCH V2 07/11] virtio-pci: address space translation service (ATS) support, Michael S. Tsirkin, 2016/11/10
- Re: [Qemu-devel] [PATCH V2 07/11] virtio-pci: address space translation service (ATS) support, Jason Wang, 2016/11/10
[Qemu-devel] [PATCH V2 08/11] acpi: add ATSR for q35, Jason Wang, 2016/11/03
[Qemu-devel] [PATCH V2 09/11] memory: handle alias for iommu notifier, Jason Wang, 2016/11/03
[Qemu-devel] [PATCH V2 10/11] memory: handle alias in memory_region_is_iommu(), Jason Wang, 2016/11/03