Re: [Qemu-devel] [RFC v9 11/18] vfio: add check host bus reset is support or not

[Alex Williamson]

On Tue, 2015-06-09 at 11:37 +0800, Chen Fan wrote:
> we introduce a has_bus_reset capability to sign the vfio
> devices if support host bus reset.

Old description, no sign of has_bus_reset here.

> 
> Signed-off-by: Chen Fan <address@hidden>
> ---
>  hw/vfio/pci.c | 169 
> ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
>  1 file changed, 169 insertions(+)
> 
> diff --git a/hw/vfio/pci.c b/hw/vfio/pci.c
> index 06006ce..97af143 100644
> --- a/hw/vfio/pci.c
> +++ b/hw/vfio/pci.c
> @@ -32,6 +32,7 @@
>  #include "hw/pci/msi.h"
>  #include "hw/pci/msix.h"
>  #include "hw/pci/pci.h"
> +#include "hw/pci/pci_bus.h"
>  #include "hw/pci/pci_bridge.h"
>  #include "qemu-common.h"
>  #include "qemu/error-report.h"
> @@ -2921,6 +2922,169 @@ out:
>      return ret;
>  }
>  
> +struct VFIODeviceFind {
> +    PCIDevice *pdev;
> +    bool found;
> +};
> +
> +static void find_devices(PCIBus *bus, void *opaque)
> +{
> +    struct VFIODeviceFind *find = opaque;
> +    int i;
> +
> +    if (find->found) {
> +        return;
> +    }
> +
> +    for (i = 0; i < ARRAY_SIZE(bus->devices); i++) {
> +        if (!bus->devices[i]) {
> +            continue;
> +        }
> +
> +        if (bus->devices[i] == find->pdev) {
> +            find->found = true;
> +            break;
> +        }
> +    }
> +}
> +
> +static void vfio_check_device_reset(PCIBus *bus, void *opaque)
> +{
> +    int i;
> +    PCIDevice *dev;
> +    VFIOPCIDevice *vdev;
> +    struct VFIODeviceFind *find = opaque;
> +
> +    if (find->found) {
> +        return;
> +    }
> +
> +    for (i = 0; i < ARRAY_SIZE(bus->devices); i++) {
> +        if (!bus->devices[i]) {
> +            continue;
> +        }
> +        dev = bus->devices[i];
> +        if (!object_dynamic_cast(OBJECT(dev), "vfio-pci")) {
> +            continue;
> +        }
> +        vdev = DO_UPCAST(VFIOPCIDevice, pdev, dev);
> +        if (!(vdev->features & VFIO_FEATURE_ENABLE_AER) &&
> +            !vdev->vbasedev.reset_works &&
> +            !vdev->has_flr &&
> +            !vdev->has_pm_reset) {

I don't think these last two are necessary, these are internal
differentiation of the reset mechanism used, not separate reset
mechanisms.  We don't require FLR and PM reset.

> +            find->pdev = dev;
> +            find->found = true;
> +            break;
> +        }
> +    }
> +}
> +
> +static int vfio_check_host_bus_reset(VFIOPCIDevice *vdev)
> +{
> +    PCIBus *bus = vdev->pdev.bus;
> +    struct vfio_pci_hot_reset_info *info = NULL;
> +    struct vfio_pci_dependent_device *devices;
> +    VFIOGroup *group;
> +    int ret, i;
> +    struct VFIODeviceFind find;
> +
> +    ret = vfio_get_hot_reset_info(vdev, &info);
> +    if (ret) {
> +        goto out;
> +    }
> +
> +    /* List all affected devices by bus reset */
> +    devices = &info->devices[0];
> +
> +    /* Verify that we have all the groups required */
> +    for (i = 0; i < info->count; i++) {
> +        PCIHostDeviceAddress host;
> +        VFIOPCIDevice *tmp;
> +        VFIODevice *vbasedev_iter;
> +
> +        host.domain = devices[i].segment;
> +        host.bus = devices[i].bus;
> +        host.slot = PCI_SLOT(devices[i].devfn);
> +        host.function = PCI_FUNC(devices[i].devfn);
> +
> +        /* Skip the current device */
> +        if (vfio_pci_host_match(&host, &vdev->host)) {
> +            continue;
> +        }
> +
> +        /* Ensure we own the group of the affected device */
> +        QLIST_FOREACH(group, &vfio_group_list, next) {
> +            if (group->groupid == devices[i].group_id) {
> +                break;
> +            }
> +        }
> +
> +        if (!group) {
> +            ret = -1;
> +            goto out;
> +        }
> +
> +        /* Ensure affected devices for reset under the same bus */
> +        QLIST_FOREACH(vbasedev_iter, &group->device_list, next) {
> +            if (vbasedev_iter->type != VFIO_DEVICE_TYPE_PCI) {
> +                continue;
> +            }
> +            tmp = container_of(vbasedev_iter, VFIOPCIDevice, vbasedev);
> +            if (vfio_pci_host_match(&host, &tmp->host)) {
> +                find.pdev = &tmp->pdev;
> +                find.found = false;
> +
> +                pci_for_each_bus(bus, find_devices, &find);
> +                if (!find.found) {
> +                    ret = -1;
> +                    goto out;
> +                }
> +                break;
> +            }
> +        }
> +    }
> +
> +    /*
> +     * Check the all vfio pci devices on or below the target bus
> +     * have a reset mechanism at least.
> +     */
> +    find.pdev = NULL;
> +    find.found = false;
> +    pci_for_each_bus(bus, vfio_check_device_reset, &find);
> +    if (find.found) {
> +        ret = -1;
> +        goto out;
> +    }
> +
> +    ret = 0;
> +out:
> +    g_free(info);
> +    return ret;
> +}
> +
> +static int vfio_check_devices_host_bus_reset(void)
> +{
> +    VFIOGroup *group;
> +    VFIODevice *vbasedev;
> +    VFIOPCIDevice *vdev;
> +
> +    /* Check All vfio-pci devices if have bus reset capability */
> +    QLIST_FOREACH(group, &vfio_group_list, next) {
> +        QLIST_FOREACH(vbasedev, &group->device_list, next) {
> +            vdev = container_of(vbasedev, VFIOPCIDevice, vbasedev);
> +            if ((vdev->features & VFIO_FEATURE_ENABLE_AER) &&
> +                vfio_check_host_bus_reset(vdev)) {
> +                error_report("vfio: Cannot enable AER for device %s, "
> +                             "which does not support host bus reset.",
> +                              vdev->vbasedev.name);
> +                 return -1;
> +            }
> +        }
> +    }
> +
> +    return 0;
> +}
> +
>  static int vfio_setup_aer(VFIOPCIDevice *vdev, uint8_t cap_ver,
>                            int pos, uint16_t size)
>  {
> @@ -2963,6 +3127,11 @@ static int vfio_setup_aer(VFIOPCIDevice *vdev, uint8_t 
> cap_ver,
>          goto error;
>      }
>  
> +    /* Make sure this devices does conflict the existing aer topology */
> +    if (vfio_check_devices_host_bus_reset()) {
> +        goto error;
> +    }
> +
>      errcap = vfio_pci_read_config(pdev, pdev->exp.aer_cap + PCI_ERR_CAP, 4);
>      /*
>       * The ability to record multiple headers is depending on