[libvirt] 答复： Re: [PATCH] libxl: support assignment from a pool of SRIOV VFs

[Chun Yan Liu]

>>> Jim Fehlig <jfehlig at suse.com> 2016-2-20 上午 7:30 >>>
On 01/25/2016 07:24 PM, Chunyan Liu wrote:
> Add codes to support creating domain with network defition of assigning
> SRIOV VF from a pool. And fix hot plug and unplug SRIOV VF under this
> kind of network defition.
>
> Signed-off-by: Chunyan Liu <cyliu at suse.com>
> ---
>  src/libxl/libxl_conf.c   |  5 +++--
>  src/libxl/libxl_domain.c | 48 +++++++++++++++++++++++++++++++++++++++++++++++-
>  src/libxl/libxl_driver.c | 12 ++++++++++++
>  tests/Makefile.am        |  3 +++
>  4 files changed, 65 insertions(+), 3 deletions(-)
>
> diff --git a/src/libxl/libxl_conf.c b/src/libxl/libxl_conf.c
> index 6320421..f50c68a 100644
> --- a/src/libxl/libxl_conf.c
> +++ b/src/libxl/libxl_conf.c
> @@ -1260,7 +1260,8 @@ libxlMakeNicList(virDomainDefPtr def,  libxl_domain_config *d_config)
>          return -1;
>  
>      for (i = 0; i < nnics; i++) {
> -        if (l_nics[i]->type == VIR_DOMAIN_NET_TYPE_HOSTDEV)
> +        if (l_nics[i]->type == VIR_DOMAIN_NET_TYPE_HOSTDEV ||
> +            l_nics[i]->data.network.actual->type == VIR_DOMAIN_NET_TYPE_HOSTDEV)

 > Elsewhere we use the virDomainNetGetActualType() accessor function.
 Ah, yes, I'll update.

>              continue;
>  
>          if (libxlMakeNic(def, l_nics[i], &x_nics[nvnics]))
> @@ -1278,7 +1279,7 @@ libxlMakeNicList(virDomainDefPtr def,  libxl_domain_config *d_config)
>  
>      VIR_SHRINK_N(x_nics, nnics, nnics - nvnics);
>      d_config->nics = x_nics;
> -    d_config->num_nics = nnics;
> +    d_config->num_nics = nvnics;

 > This looks like a bug fix. If so, it should be in a separate patch.
 Yes. Will put in a separate patch.

>  
>      return 0;
>  
> diff --git a/src/libxl/libxl_domain.c b/src/libxl/libxl_domain.c
> index cf5c9f6..9bf7a5a 100644
> --- a/src/libxl/libxl_domain.c
> +++ b/src/libxl/libxl_domain.c
> @@ -35,6 +35,7 @@
>  #include "virstring.h"
>  #include "virtime.h"
>  #include "locking/domain_lock.h"
> +#include "network/bridge_driver.h"
>  
>  #define VIR_FROM_THIS VIR_FROM_LIBXL
>  
> @@ -314,7 +315,7 @@ libxlDomainDeviceDefPostParse(virDomainDeviceDefPtr dev,
>          virDomainHostdevSubsysPCIPtr pcisrc;
>  
>          if (dev->type == VIR_DOMAIN_DEVICE_NET)
> -            hostdev = &(dev->data.net)->data.hostdev.def;
> +            hostdev = &dev->data.net->data.hostdev.def;

 > This looks like a bug fix too.
 Yes. With '()' has no harm, but it doesn't need.

>          else
>              hostdev = dev->data.hostdev;
>          pcisrc = &hostdev->source.subsys.u.pci;
> @@ -916,6 +917,48 @@ libxlConsoleCallback(libxl_ctx *ctx, libxl_event *ev, void *for_callback)
>      libxl_event_free(ctx, ev);
>  }
>  
> +static int
> +libxlNetworkPrepareDevices(virDomainDefPtr def)
> +{
> +    int ret = -1;
> +    size_t i;
> +
> +    for (i = 0; i < def->nnets; i++) {
> +        virDomainNetDefPtr net = def->nets[i];
> +        int actualType;
> +
> +        /* If appropriate, grab a physical device from the configured
> +         * network's pool of devices, or resolve bridge device name
> +         * to the one defined in the network definition.
> +         */
> +        if (networkAllocateActualDevice(def, net) < 0)
> +            goto cleanup;
> +
> +        actualType = virDomainNetGetActualType(net);
> +        if (actualType == VIR_DOMAIN_NET_TYPE_HOSTDEV &&
> +            net->type == VIR_DOMAIN_NET_TYPE_NETWORK) {
> +            /* Each type='hostdev' network device must also have a
> +             * corresponding entry in the hostdevs array. For netdevs
> +             * that are hardcoded as type='hostdev', this is already
> +             * done by the parser, but for those allocated from a
> +             * network / determined at runtime, we need to do it
> +             * separately.
> +             */
> +            virDomainHostdevDefPtr hostdev = virDomainNetGetActualHostdev(net);
> +            virDomainHostdevSubsysPCIPtr pcisrc = &hostdev->source.subsys.u.pci;
> +

 > I n > see if the hostdev is already in use. Is that needed here too?
 Later when doing HostdevAttach, it will check. So I think it's not necessary.

> +            if (hostdev->mode == VIR_DOMAIN_HOSTDEV_MODE_SUBSYS &&
> +                hostdev->source.subsys.type == VIR_DOMAIN_HOSTDEV_SUBSYS_TYPE_PCI)

 > Does this need to include checking if backend is already set (pcisrc->backend ==
 > VIR_DOMAIN_HOSTDEV_PCI_BACKEND_DEFAULT) ? I suppose it should always be
 > VIR_DOMAIN_HOSTDEV_PCI_BACKEND_XEN.
 There are two cases, one is specified as <interface type='hostdev'>, for this case, it
 will be handled during DeviceDefParse phase, it's already set as BACKEND_XEN;
 another case is from a pool of SRIOV vfs, it won't be handled during DeviceDefParse
 as a hostdev, so the backend type is still BACKEND_DEFAULT, for this case, we
 need to set it to be BACKEND_XEN.
 

> +                pcisrc->backend = VIR_DOMAIN_HOSTDEV_PCI_BACKEND_XEN;
> +
> +            if (virDomainHostdevInsert(def, hostdev) < 0)
> +                goto cleanup;
> +        }
> +    }
> +    ret = 0;
> + cleanup:
> +    return ret;
> +}

 > Does anything done by this function need to be undone when the domain is shutdown?
 No, when domain is shutdown, all is handled by hostdev process.

 Chunyan

 > Regards,
 > Jim