[v2] Dinamycally switch to vhost shadow virtqueues at vdpa net migration

[RFC v2 01/13] vdpa: fix VHOST_BACKEND_F_IOTLB_ASID flag check

Posted by Eugenio Pérez 3 years ago

VHOST_BACKEND_F_IOTLB_ASID is the feature bit, not the bitmask. Since
the device under test also provided VHOST_BACKEND_F_IOTLB_MSG_V2 and
VHOST_BACKEND_F_IOTLB_BATCH, this went unnoticed.

Fixes: c1a1008685 ("vdpa: always start CVQ in SVQ mode if possible")
Signed-off-by: Eugenio Pérez <eperezma@redhat.com>
---
 net/vhost-vdpa.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/net/vhost-vdpa.c b/net/vhost-vdpa.c
index 1a13a34d35..de5ed8ff22 100644
--- a/net/vhost-vdpa.c
+++ b/net/vhost-vdpa.c
@@ -384,7 +384,7 @@ static int vhost_vdpa_net_cvq_start(NetClientState *nc)
             g_strerror(errno), errno);
         return -1;
     }
-    if (!(backend_features & VHOST_BACKEND_F_IOTLB_ASID) ||
+    if (!(backend_features & BIT_ULL(VHOST_BACKEND_F_IOTLB_ASID)) ||
         !vhost_vdpa_net_valid_svq_features(v->dev->features, NULL)) {
         return 0;
     }
-- 
2.31.1

Re: [RFC v2 01/13] vdpa: fix VHOST_BACKEND_F_IOTLB_ASID flag check

Posted by Jason Wang 3 years ago

On Fri, Jan 13, 2023 at 1:24 AM Eugenio Pérez <eperezma@redhat.com> wrote:
>
> VHOST_BACKEND_F_IOTLB_ASID is the feature bit, not the bitmask. Since
> the device under test also provided VHOST_BACKEND_F_IOTLB_MSG_V2 and
> VHOST_BACKEND_F_IOTLB_BATCH, this went unnoticed.
>
> Fixes: c1a1008685 ("vdpa: always start CVQ in SVQ mode if possible")
> Signed-off-by: Eugenio Pérez <eperezma@redhat.com>

Acked-by: Jason Wang <jasowang@redhat.com>

Do we need this for -stable?

Thanks

> ---
>  net/vhost-vdpa.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/net/vhost-vdpa.c b/net/vhost-vdpa.c
> index 1a13a34d35..de5ed8ff22 100644
> --- a/net/vhost-vdpa.c
> +++ b/net/vhost-vdpa.c
> @@ -384,7 +384,7 @@ static int vhost_vdpa_net_cvq_start(NetClientState *nc)
>              g_strerror(errno), errno);
>          return -1;
>      }
> -    if (!(backend_features & VHOST_BACKEND_F_IOTLB_ASID) ||
> +    if (!(backend_features & BIT_ULL(VHOST_BACKEND_F_IOTLB_ASID)) ||
>          !vhost_vdpa_net_valid_svq_features(v->dev->features, NULL)) {
>          return 0;
>      }
> --
> 2.31.1
>

Re: [RFC v2 01/13] vdpa: fix VHOST_BACKEND_F_IOTLB_ASID flag check

Posted by Eugenio Perez Martin 3 years ago

On Fri, Jan 13, 2023 at 4:12 AM Jason Wang <jasowang@redhat.com> wrote:
>
> On Fri, Jan 13, 2023 at 1:24 AM Eugenio Pérez <eperezma@redhat.com> wrote:
> >
> > VHOST_BACKEND_F_IOTLB_ASID is the feature bit, not the bitmask. Since
> > the device under test also provided VHOST_BACKEND_F_IOTLB_MSG_V2 and
> > VHOST_BACKEND_F_IOTLB_BATCH, this went unnoticed.
> >
> > Fixes: c1a1008685 ("vdpa: always start CVQ in SVQ mode if possible")
> > Signed-off-by: Eugenio Pérez <eperezma@redhat.com>
>
> Acked-by: Jason Wang <jasowang@redhat.com>
>
> Do we need this for -stable?
>

The commit c1a1008685 was introduced in this development window so
there is no stable version of qemu with that patch. But I'm ok to CC
stable just in case for sure.

Thanks!

> Thanks
>
> > ---
> >  net/vhost-vdpa.c | 2 +-
> >  1 file changed, 1 insertion(+), 1 deletion(-)
> >
> > diff --git a/net/vhost-vdpa.c b/net/vhost-vdpa.c
> > index 1a13a34d35..de5ed8ff22 100644
> > --- a/net/vhost-vdpa.c
> > +++ b/net/vhost-vdpa.c
> > @@ -384,7 +384,7 @@ static int vhost_vdpa_net_cvq_start(NetClientState *nc)
> >              g_strerror(errno), errno);
> >          return -1;
> >      }
> > -    if (!(backend_features & VHOST_BACKEND_F_IOTLB_ASID) ||
> > +    if (!(backend_features & BIT_ULL(VHOST_BACKEND_F_IOTLB_ASID)) ||
> >          !vhost_vdpa_net_valid_svq_features(v->dev->features, NULL)) {
> >          return 0;
> >      }
> > --
> > 2.31.1
> >
>

Re: [RFC v2 01/13] vdpa: fix VHOST_BACKEND_F_IOTLB_ASID flag check

Posted by Jason Wang 3 years ago

在 2023/1/13 14:42, Eugenio Perez Martin 写道:
> On Fri, Jan 13, 2023 at 4:12 AM Jason Wang <jasowang@redhat.com> wrote:
>> On Fri, Jan 13, 2023 at 1:24 AM Eugenio Pérez <eperezma@redhat.com> wrote:
>>> VHOST_BACKEND_F_IOTLB_ASID is the feature bit, not the bitmask. Since
>>> the device under test also provided VHOST_BACKEND_F_IOTLB_MSG_V2 and
>>> VHOST_BACKEND_F_IOTLB_BATCH, this went unnoticed.
>>>
>>> Fixes: c1a1008685 ("vdpa: always start CVQ in SVQ mode if possible")
>>> Signed-off-by: Eugenio Pérez <eperezma@redhat.com>
>> Acked-by: Jason Wang <jasowang@redhat.com>
>>
>> Do we need this for -stable?
>>
> The commit c1a1008685 was introduced in this development window so
> there is no stable version of qemu with that patch. But I'm ok to CC
> stable just in case for sure.


Right, just have a check and it doesn't there for 7.2 so there should be 
no need for that.

Thanks


>
> Thanks!
>
>> Thanks
>>
>>> ---
>>>   net/vhost-vdpa.c | 2 +-
>>>   1 file changed, 1 insertion(+), 1 deletion(-)
>>>
>>> diff --git a/net/vhost-vdpa.c b/net/vhost-vdpa.c
>>> index 1a13a34d35..de5ed8ff22 100644
>>> --- a/net/vhost-vdpa.c
>>> +++ b/net/vhost-vdpa.c
>>> @@ -384,7 +384,7 @@ static int vhost_vdpa_net_cvq_start(NetClientState *nc)
>>>               g_strerror(errno), errno);
>>>           return -1;
>>>       }
>>> -    if (!(backend_features & VHOST_BACKEND_F_IOTLB_ASID) ||
>>> +    if (!(backend_features & BIT_ULL(VHOST_BACKEND_F_IOTLB_ASID)) ||
>>>           !vhost_vdpa_net_valid_svq_features(v->dev->features, NULL)) {
>>>           return 0;
>>>       }
>>> --
>>> 2.31.1
>>>