Re: [dpdk-dev] [PATCH v4 04/19] vhost: fix payload size of reply

[Maxime Coquelin <maxime.coquelin@redhat.com>]

On 10/09/2018 12:30 PM, Maxime Coquelin wrote:
> 
> 
> On 10/09/2018 12:21 PM, Tiwei Bie wrote:
>> On Mon, Oct 08, 2018 at 05:25:42PM +0200, Maxime Coquelin wrote:
>>> QEMU doesn't expect any payload for the reply of
>>> VHOST_USER_SET_LOG_BASE request, so don't send any.
>>> Note that the Vhost-user specification isn't clear about
>>> it and would need to be fixed.
>>>
>>> Fixes: 54f9e32305d4 ("vhost: handle dirty pages logging request")
>>> Cc: stable@dpdk.org
>>>
>>> Reported-by: Ilya Maximets <i.maximets@samsung.com>
>>> Signed-off-by: Maxime Coquelin <maxime.coquelin@redhat.com>
>>> Acked-by: Ilya Maximets <i.maximets@samsung.com>
>>> ---
>>>   lib/librte_vhost/vhost_user.c | 6 +++++-
>>>   1 file changed, 5 insertions(+), 1 deletion(-)
>>>
>>> diff --git a/lib/librte_vhost/vhost_user.c 
>>> b/lib/librte_vhost/vhost_user.c
>>> index 7f3e86778..71a0e7dd7 100644
>>> --- a/lib/librte_vhost/vhost_user.c
>>> +++ b/lib/librte_vhost/vhost_user.c
>>> @@ -1296,7 +1296,11 @@ vhost_user_set_log_base(struct virtio_net 
>>> **pdev, struct VhostUserMsg *msg)
>>>       dev->log_base = dev->log_addr + off;
>>>       dev->log_size = size;
>>> -    msg->size = sizeof(msg->payload.u64);
>>> +    /*
>>> +     * The spec is not clear about it (yet), but QEMU doesn't expect
>>> +     * any payload in the reply.
>>> +     */
>>> +    msg->size = 0;
>>
>> I think the spec is clear about it that no payload is
>> expected in the reply:
>>
>> https://github.com/qemu/qemu/blob/7c69b7c84964/docs/interop/vhost-user.txt#L496 
>>
> It isn't really clear because there are two cases[0]:
> 1. VHOST_USER_PROTOCOL_F_LOG_SHMFD not negotiated: In this case no reply
>     has to be sent.
> 2. VHOST_USER_PROTOCOL_F_LOG_SHMFD negotiated: In this case reply with
>     empty payload hsa to be sent.
> 
> [0]: 
> https://github.com/qemu/qemu/blob/7c69b7c849641a39ba3defa40d384a2ba24cd7a2/docs/interop/vhost-user.txt#L177 
> 

And I would add that N/A isn't clear, because SET_VRING_ADDR specifies
N/A as slave payload, whereas the implementation does not implement a
reply.

>> But below line in the spec needs to be fixed:
>>
>> https://github.com/qemu/qemu/blob/7c69b7c84964/docs/interop/vhost-user.txt#L495 
>>
> 
> Indeed.
> 
>>
>>>       return VH_RESULT_REPLY;
>>>   }
>>> -- 
>>> 2.17.1
>>>