Re: [PATCH] vdpa/mlx5: Use random MAC for the vdpa net instance

[Jason Wang <jasowang@redhat.com>]

On 2020/12/2 下午9:04, Michael S. Tsirkin wrote:
> On Wed, Dec 02, 2020 at 08:56:37PM +0800, Jason Wang wrote:
>> On 2020/12/2 下午5:30, Michael S. Tsirkin wrote:
>>> On Wed, Dec 02, 2020 at 12:18:36PM +0800, Jason Wang wrote:
>>>> On 2020/12/1 下午5:23, Cindy Lu wrote:
>>>>> On Mon, Nov 30, 2020 at 11:33 PM Michael S. Tsirkin <mst@redhat.com> wrote:
>>>>>> On Mon, Nov 30, 2020 at 06:41:45PM +0800, Cindy Lu wrote:
>>>>>>> On Mon, Nov 30, 2020 at 5:33 PM Michael S. Tsirkin <mst@redhat.com> wrote:
>>>>>>>> On Mon, Nov 30, 2020 at 11:27:59AM +0200, Eli Cohen wrote:
>>>>>>>>> On Mon, Nov 30, 2020 at 04:00:51AM -0500, Michael S. Tsirkin wrote:
>>>>>>>>>> On Mon, Nov 30, 2020 at 08:27:46AM +0200, Eli Cohen wrote:
>>>>>>>>>>> On Sun, Nov 29, 2020 at 03:08:22PM -0500, Michael S. Tsirkin wrote:
>>>>>>>>>>>> On Sun, Nov 29, 2020 at 08:43:51AM +0200, Eli Cohen wrote:
>>>>>>>>>>>>> We should not try to use the VF MAC address as that is used by the
>>>>>>>>>>>>> regular (e.g. mlx5_core) NIC implementation. Instead, use a random
>>>>>>>>>>>>> generated MAC address.
>>>>>>>>>>>>>
>>>>>>>>>>>>> Suggested by: Cindy Lu <lulu@redhat.com>
>>>>>>>>>>>>> Fixes: 1a86b377aa21 ("vdpa/mlx5: Add VDPA driver for supported mlx5 devices")
>>>>>>>>>>>>> Signed-off-by: Eli Cohen <elic@nvidia.com>
>>>>>>>>>>>> I didn't realise it's possible to use VF in two ways
>>>>>>>>>>>> with and without vdpa.
>>>>>>>>>>> Using a VF you can create quite a few resources, e.g. send queues
>>>>>>>>>>> recieve queues, virtio_net queues etc. So you can possibly create
>>>>>>>>>>> several instances of vdpa net devices and nic net devices.
>>>>>>>>>>>
>>>>>>>>>>>> Could you include a bit more description on the failure
>>>>>>>>>>>> mode?
>>>>>>>>>>> Well, using the MAC address of the nic vport is wrong since that is the
>>>>>>>>>>> MAC of the regular NIC implementation of mlx5_core.
>>>>>>>>>> Right but ATM it doesn't coexist with vdpa so what's the problem?
>>>>>>>>>>
>>>>>>>>> This call is wrong:  mlx5_query_nic_vport_mac_address()
>>>>>>>>>
>>>>>>>>>>>> Is switching to a random mac for such an unusual
>>>>>>>>>>>> configuration really justified?
>>>>>>>>>>> Since I can't use the NIC's MAC address, I have two options:
>>>>>>>>>>> 1. To get the MAC address as was chosen by the user administering the
>>>>>>>>>>>       NIC. This should invoke the set_config callback. Unfortunately this
>>>>>>>>>>>       is not implemented yet.
>>>>>>>>>>>
>>>>>>>>>>> 2. Use a random MAC address. This is OK since if (1) is implemented it
>>>>>>>>>>>       can always override this random configuration.
>>>>>>>>>>>
>>>>>>>>>>>> It looks like changing a MAC could break some guests,
>>>>>>>>>>>> can it not?
>>>>>>>>>>>>
>>>>>>>>>>> No, it will not. The current version of mlx5 VDPA does not allow regular
>>>>>>>>>>> NIC driver and VDPA to co-exist. I have patches ready that enable that
>>>>>>>>>>> from steering point of view. I will post them here once other patches on
>>>>>>>>>>> which they depend will be merged.
>>>>>>>>>>>
>>>>>>>>>>> https://patchwork.ozlabs.org/project/netdev/patch/20201120230339.651609-12-saeedm@nvidia.com/
>>>>>>>>>> Could you be more explicit on the following points:
>>>>>>>>>> - which configuration is broken ATM (as in, two device have identical
>>>>>>>>>>      macs? any other issues)?
>>>>>>>>> The only wrong thing is the call to  mlx5_query_nic_vport_mac_address().
>>>>>>>>> It's not breaking anything yet is wrong. The random MAC address setting
>>>>>>>>> is required for the steering patches.
>>>>>>>> Okay so I'm not sure the Fixes tag at least is appropriate if it's a
>>>>>>>> dependency of a new feature.
>>>>>>>>
>>>>>>>>>> - why won't device MAC change from guest point of view?
>>>>>>>>>>
>>>>>>>>> It's lack of implementation in qemu as far as I know.
>>>>>>>> Sorry not sure I understand. What's not implemented in QEMU?
>>>>>>>>
>>>>>>> HI Michael, there are some bug in qemu to set_config, this will fix in future,
>>>>>>> But this patch is still needed, because without this patch the mlx
>>>>>>> driver will give an 0 mac address to qemu
>>>>>>> and qemu will overwrite the default mac address.  This will cause traffic down.
>>>>>> Hmm the patch description says VF mac address, not 0 address. Confused.
>>>>>> If there's no mac we can clear VIRTIO_NET_F_MAC and have guest
>>>>>> use a random value ...
>>>> I'm not sure this can work for all types of vDPA (e.g it could not be a
>>>> learning bridge in the swtich).
>>>>
>>>>
>>>>> hi Michael，
>>>>> I have tried as your suggestion, seems even remove the
>>>>> VIRTIO_NET_F_MAC the qemu will still call get_cinfig and overwrite the
>>>>> default address in  VM,
>>>> This looks a bug in qemu, in guest driver we had:
>>>>
>>>>       /* Configuration may specify what MAC to use.  Otherwise random. */
>>>>       if (virtio_has_feature(vdev, VIRTIO_NET_F_MAC))
>>>>           virtio_cread_bytes(vdev,
>>>>                      offsetof(struct virtio_net_config, mac),
>>>>                      dev->dev_addr, dev->addr_len);
>>>>       else
>>>>           eth_hw_addr_random(dev);
>>>>
>>>>
>>>>> this process is like
>>>>> vdpa _init -->qemu call get_config ->mlx driver will give  an mac
>>>>> address with all 0-->
>>>>> qemu will not check this mac address and use it --> overwrite the mac
>>>>> address in qemu
>>>>>
>>>>> So for my understanding there are several method to fix this problem
>>>>>
>>>>> 1, qemu check the mac address, if the mac address is all 0, qemu will
>>>>> ignore it and set the random mac address to mlx driver.
>>>> So my understanding is that, if mac address is all 0, vDPA parent should not
>>>> advertise VIRTIO_NET_F_MAC. And qemu should emulate this feature as you did:
>>>>
>>>> 1) get a random mac
>>> To me this looks like a spec violation.
>>>
>>> If the driver negotiates the VIRTIO_NET_F_MAC feature, the driver MUST set
>>> the physical address of the NIC to \field{mac}.  Otherwise, it SHOULD
>>> use a locally-administered MAC address (see \hyperref[intro:IEEE 802]{IEEE 802},
>>> ``9.2 48-bit universal LAN MAC addresses'').
>>
>> One question here, what did "set" mean here consider the mac is given by the
>> device itself?
>>
>
> That is my understanding, and this seems to be what linux guests do.
>
>>> While not said explicitly, the assumption I think is that the local
>>> MAC is not a local one.
>>>
>>>
>>>> 2) advertise VIRTIO_NET_F_MAC
>>>> 3) set the random mac to vDPA through set_config
>>> that part looks wrong to me. Setting mac through set_config was
>>> a pre-virtio-1.0 way to send mac to device. In 1.0 we have
>>> VIRTIO_NET_CTRL_MAC_ADDR_SET for that:
>>>
>>>
>>> 	When using the legacy interface, \field{mac} is driver-writable
>>> 	which provided a way for drivers to update the MAC without
>>> 	negotiating VIRTIO_NET_F_CTRL_MAC_ADDR.
>>
>> Looks like it doesn't prevent us from doing so.
>  From writing into mac?
> Yes it does:
> 	Device configuration fields are listed below, they are read-only for a driver. The \field{mac} address field
> 	always exists (though is only valid if VIRTIO_NET_F_MAC is set), and
> 	\field{status} only exists if VIRTIO_NET_F_STATUS is set.


Okay, so this is a hint that the vDPA parent needs to check and ignore 
the write if version 1 is negotiated.


>
>
>> Otherwise this brings an
>> implicit dependency for control virtqueue if we want to support 1.0?
>>
>> Thanks
> With 1.0 you either need VIRTIO_NET_F_CTRL_MAC_ADDR or VIRTIO_NET_F_MAC.
>
>

Ok, so it looks to me VIRTIO_NET_F_MAC is the only choice for mlx5 
consider control vq is not supported.

Thanks


>>>
>>>
>>>> 4) advertise the random mac to emulated config to guest
>>>>
>>>>
>>>>> 2. mlx driver checks the mac address and if this mac is 0, return fail
>>>>> to qemu, but this need to change the UAPI.
>>>> uAPI is probably fine since ioctl can fail.  We can change the to allow the
>>>> set_config to fail but virito spec doesn't have a way to advertise the error
>>>> in this case. Anyway, the driver only risk itself for setting a wrong value,
>>>> so we're probably fine.
>>>>
>>>> Thanks
>>>>
>>>>
>>>>> 3. mlx driver it shelf should get an correct mac address while it init.
>>>>> 4. add check in qemu get_config function  , if there is not F_MAC Then
>>>>> ignore the mac address from mlx driver
>>>>>
>>>>> not sure which method is more suitable ?
>>>>>
>>>>> Thanks
>>>>> Cindy
>>>>>
>>>>>
>>>>>
>>>>>>>>>>>>> ---
>>>>>>>>>>>>>     drivers/vdpa/mlx5/net/mlx5_vnet.c | 5 +----
>>>>>>>>>>>>>     1 file changed, 1 insertion(+), 4 deletions(-)
>>>>>>>>>>>>>
>>>>>>>>>>>>> diff --git a/drivers/vdpa/mlx5/net/mlx5_vnet.c b/drivers/vdpa/mlx5/net/mlx5_vnet.c
>>>>>>>>>>>>> index 1fa6fcac8299..80d06d958b8b 100644
>>>>>>>>>>>>> --- a/drivers/vdpa/mlx5/net/mlx5_vnet.c
>>>>>>>>>>>>> +++ b/drivers/vdpa/mlx5/net/mlx5_vnet.c
>>>>>>>>>>>>> @@ -1955,10 +1955,7 @@ void *mlx5_vdpa_add_dev(struct mlx5_core_dev *mdev)
>>>>>>>>>>>>>          if (err)
>>>>>>>>>>>>>                  goto err_mtu;
>>>>>>>>>>>>>
>>>>>>>>>>>>> -     err = mlx5_query_nic_vport_mac_address(mdev, 0, 0, config->mac);
>>>>>>>>>>>>> -     if (err)
>>>>>>>>>>>>> -             goto err_mtu;
>>>>>>>>>>>>> -
>>>>>>>>>>>>> +     eth_random_addr(config->mac);
>>>>>>>>>>>>>          mvdev->vdev.dma_dev = mdev->device;
>>>>>>>>>>>>>          err = mlx5_vdpa_alloc_resources(&ndev->mvdev);
>>>>>>>>>>>>>          if (err)
>>>>>>>>>>>>> --
>>>>>>>>>>>>> 2.26.2