search for: mlx_features

Clang warns several times when building for 32-bit ARM along the lines of: drivers/vdpa/mlx5/net/mlx5_vnet.c:1462:31: warning: shift count >= width of type [-Wshift-count-overflow] ndev->mvdev.mlx_features |= BIT(VIRTIO_F_VERSION_1); ^~~~~~~~~~~~~~~~~~~~~~~ This is related to the BIT macro, which uses an unsigned long literal, which is 32-bit on ARM so having a shift equal to or larger than 32 will cause this warning, such as the above, where VIRTIO_F_VERS...

Clang warns several times when building for 32-bit ARM along the lines of: drivers/vdpa/mlx5/net/mlx5_vnet.c:1462:31: warning: shift count >= width of type [-Wshift-count-overflow] ndev->mvdev.mlx_features |= BIT(VIRTIO_F_VERSION_1); ^~~~~~~~~~~~~~~~~~~~~~~ This is related to the BIT macro, which uses an unsigned long literal, which is 32-bit on ARM so having a shift equal to or larger than 32 will cause this warning, such as the above, where VIRTIO_F_VERS...

On 8/21/20 3:50 PM, Nathan Chancellor wrote: > Clang warns several times when building for 32-bit ARM along the lines > of: > > drivers/vdpa/mlx5/net/mlx5_vnet.c:1462:31: warning: shift count >= width > of type [-Wshift-count-overflow] > ndev->mvdev.mlx_features |= BIT(VIRTIO_F_VERSION_1); > ^~~~~~~~~~~~~~~~~~~~~~~ > > This is related to the BIT macro, which uses an unsigned long literal, > which is 32-bit on ARM so having a shift equal to or larger than 32 will > cause this warning, such as the a...

...; + struct mlx5_vdpa_dev *mvdev = to_mvdev(vdev); >>> + struct mlx5_vdpa_net *ndev = to_mlx5_vdpa_ndev(mvdev); >>> + u16 dev_features; >>> + >>> + dev_features = MLX5_CAP_DEV_VDPA_EMULATION(mvdev->mdev, device_features_bits_mask); >>> + ndev->mvdev.mlx_features = mlx_to_vritio_features(dev_features); >>> + if (MLX5_CAP_DEV_VDPA_EMULATION(mvdev->mdev, virtio_version_1_0)) >>> + ndev->mvdev.mlx_features |= BIT(VIRTIO_F_VERSION_1); >> >> This is interesting. This suggests !VIRTIO_F_VERSION_1 && >> VIRTIO_F_I...

...st struct mlx5_vdpa_dev *mvdev) > @@ -797,6 +813,7 @@ static int create_virtqueue(struct mlx5_vdpa_net *ndev, struct mlx5_vdpa_virtque > int inlen = MLX5_ST_SZ_BYTES(create_virtio_net_q_in); > u32 out[MLX5_ST_SZ_DW(create_virtio_net_q_out)] = {}; > void *obj_context; > + u16 mlx_features; > void *cmd_hdr; > void *vq_ctx; > void *in; > @@ -812,6 +829,7 @@ static int create_virtqueue(struct mlx5_vdpa_net *ndev, struct mlx5_vdpa_virtque > goto err_alloc; > } > > + mlx_features = get_features(ndev->mvdev.actual_features); > cmd_hdr =...

...ev *mvdev) > @@ -797,6 +817,7 @@ static int create_virtqueue(struct mlx5_vdpa_net *ndev, struct mlx5_vdpa_virtque > int inlen = MLX5_ST_SZ_BYTES(create_virtio_net_q_in); > u32 out[MLX5_ST_SZ_DW(create_virtio_net_q_out)] = {}; > void *obj_context; > + u16 mlx_features; > void *cmd_hdr; > void *vq_ctx; > void *in; > @@ -812,6 +833,7 @@ static int create_virtqueue(struct mlx5_vdpa_net *ndev, struct mlx5_vdpa_virtque > goto err_alloc; > } > > + mlx_features = get_features(ndev->mvdev.a...

This patchset is pre-requisite to export and provision device config attributes and features for vdpa live migration, in a way backward and forward compatibility can be retained. The follow up work [1] will need to be built around the new feature provisioning uAPI, with which it's easier to formalize migration compatibility support at the driver level. Thanks, -Siwei [1] [PATCH v3 0/4] vDPA:

This patchset is pre-requisite to export and provision device config attributes and features for vdpa live migration, in a way backward and forward compatibility can be retained. The follow up work [1] will need to be built around the new feature provisioning uAPI, with which it's easier to formalize migration compatibility support at the driver level. Thanks, -Siwei [1] [PATCH v3 0/4] vDPA:

...es(struct vdpa_device *vdev) > +{ > + struct mlx5_vdpa_dev *mvdev = to_mvdev(vdev); > + struct mlx5_vdpa_net *ndev = to_mlx5_vdpa_ndev(mvdev); > + u16 dev_features; > + > + dev_features = MLX5_CAP_DEV_VDPA_EMULATION(mvdev->mdev, device_features_bits_mask); > + ndev->mvdev.mlx_features = mlx_to_vritio_features(dev_features); > + if (MLX5_CAP_DEV_VDPA_EMULATION(mvdev->mdev, virtio_version_1_0)) > + ndev->mvdev.mlx_features |= BIT(VIRTIO_F_VERSION_1); This is interesting. This suggests !VIRTIO_F_VERSION_1 && VIRTIO_F_IOMMU_PLATFORM is valid. But virito spec...

This patchset is pre-requisite to export and provision device config attributes and features for vdpa live migration, in a way backward and forward compatibility can be retained. The follow up work [1] will need to be built around the new feature provisioning uAPI, with which it's easier to formalize migration compatibility support at the driver level. Thanks, -Siwei [1] [PATCH v3 0/4] vDPA:

This patchset is pre-requisite to export and provision device config attributes and features for vdpa live migration, in a way backward and forward compatibility can be retained. The follow up work [1] will need to be built around the new feature provisioning uAPI, with which it's easier to formalize migration compatibility support at the driver level. Thanks, -Siwei [1] [PATCH v3 0/4] vDPA:

...tic int mlx5_vdpa_set_features(struct vdpa_device *vdev, u64 features) { struct mlx5_vdpa_dev *mvdev = to_mvdev(vdev); @@ -1519,6 +1527,8 @@ static int mlx5_vdpa_set_features(struct vdpa_device *vdev, u64 features) return err; ndev->mvdev.actual_features = features & ndev->mvdev.mlx_features; + ndev->config.mtu = __cpu_to_virtio16(mlx5_vdpa_is_little_endian(mvdev), + ndev->mtu); return err; } @@ -1925,7 +1935,7 @@ void *mlx5_vdpa_add_dev(struct mlx5_core_dev *mdev) init_mvqs(ndev); mutex_init(&ndev->reslock); config = &ndev->config; - err = mlx5_...

...mvdev, features, true); > @@ -2310,7 +2362,11 @@ static int mlx5_vdpa_set_driver_features(struct vdpa_device *vdev, u64 features) > if (err) > return err; > > + cur_features = ndev->mvdev.actual_features; > ndev->mvdev.actual_features = features & ndev->mvdev.mlx_features; > + if (f_status_was_set(cur_features, ndev->mvdev.actual_features)) > + update_link_state(mvdev); > + > if (ndev->mvdev.actual_features & BIT_ULL(VIRTIO_NET_F_MQ)) > ndev->rqt_size = mlx5vdpa16_to_cpu(mvdev, ndev->config.max_virtqueue_pairs); > else >...

...vdpa_set_driver_features(struct vdpa_device *vdev, u64 features) > >> if (err) > >> return err; > >> > >> + old = ndev->mvdev.actual_features; > >> ndev->mvdev.actual_features = features & ndev->mvdev.mlx_features; > >> + if (~old & ndev->mvdev.actual_features & BIT_ULL(VIRTIO_NET_F_STATUS)) > >> + register_link_notifier(ndev); > >> + > > I still think it's better to move this check during set_status. > > Otherwise, we may have a small...