Re: [PATCH v2 1/2] vhost: Add smp_rmb() in vhost_vq_avail_empty()

[Gavin Shan <gshan@redhat.com>]

On 3/27/24 14:08, Gavin Shan wrote:
> On 3/27/24 12:44, Jason Wang wrote:
>> On Wed, Mar 27, 2024 at 10:34 AM Jason Wang <jasowang@redhat.com> wrote:
>>> On Wed, Mar 27, 2024 at 7:39 AM Gavin Shan <gshan@redhat.com> wrote:
>>>>
>>>> A smp_rmb() has been missed in vhost_vq_avail_empty(), spotted by
>>>> Will Deacon <will@kernel.org>. Otherwise, it's not ensured the
>>>> available ring entries pushed by guest can be observed by vhost
>>>> in time, leading to stale available ring entries fetched by vhost
>>>> in vhost_get_vq_desc(), as reported by Yihuang Yu on NVidia's
>>>> grace-hopper (ARM64) platform.
>>>>
>>>>    /home/gavin/sandbox/qemu.main/build/qemu-system-aarch64      \
>>>>    -accel kvm -machine virt,gic-version=host -cpu host          \
>>>>    -smp maxcpus=1,cpus=1,sockets=1,clusters=1,cores=1,threads=1 \
>>>>    -m 4096M,slots=16,maxmem=64G                                 \
>>>>    -object memory-backend-ram,id=mem0,size=4096M                \
>>>>     :                                                           \
>>>>    -netdev tap,id=vnet0,vhost=true                              \
>>>>    -device virtio-net-pci,bus=pcie.8,netdev=vnet0,mac=52:54:00:f1:26:b0
>>>>     :
>>>>    guest# netperf -H 10.26.1.81 -l 60 -C -c -t UDP_STREAM
>>>>    virtio_net virtio0: output.0:id 100 is not a head!
>>>>
>>>> Add the missed smp_rmb() in vhost_vq_avail_empty(). Note that it
>>>> should be safe until vq->avail_idx is changed by commit 275bf960ac697
>>>> ("vhost: better detection of available buffers").
>>>>
>>>> Fixes: 275bf960ac697 ("vhost: better detection of available buffers")
>>>> Cc: <stable@kernel.org> # v4.11+
>>>> Reported-by: Yihuang Yu <yihyu@redhat.com>
>>>> Signed-off-by: Gavin Shan <gshan@redhat.com>
>>>> ---
>>>>   drivers/vhost/vhost.c | 11 ++++++++++-
>>>>   1 file changed, 10 insertions(+), 1 deletion(-)
>>>>
>>>> diff --git a/drivers/vhost/vhost.c b/drivers/vhost/vhost.c
>>>> index 045f666b4f12..00445ab172b3 100644
>>>> --- a/drivers/vhost/vhost.c
>>>> +++ b/drivers/vhost/vhost.c
>>>> @@ -2799,9 +2799,18 @@ bool vhost_vq_avail_empty(struct vhost_dev *dev, struct vhost_virtqueue *vq)
>>>>          r = vhost_get_avail_idx(vq, &avail_idx);
>>>>          if (unlikely(r))
>>>>                  return false;
>>>> +
>>>>          vq->avail_idx = vhost16_to_cpu(vq, avail_idx);
>>>> +       if (vq->avail_idx != vq->last_avail_idx) {
>>>> +               /* Similar to what's done in vhost_get_vq_desc(), we need
>>>> +                * to ensure the available ring entries have been exposed
>>>> +                * by guest.
>>>> +                */
>>>
>>> We need to be more verbose here. For example, which load needs to be
>>> ordered with which load.
>>>
>>> The rmb in vhost_get_vq_desc() is used to order the load of avail idx
>>> and the load of head. It is paired with e.g virtio_wmb() in
>>> virtqueue_add_split().
>>>
>>> vhost_vq_avail_empty() are mostly used as a hint in
>>> vhost_net_busy_poll() which is under the protection of the vq mutex.
>>>
>>> An exception is the tx_can_batch(), but in that case it doesn't even
>>> want to read the head.
>>
>> Ok, if it is needed only in that path, maybe we can move the barriers there.
>>
> 
> [cc Will Deacon]
> 
> Jason, appreciate for your review and comments. I think PATCH[1/2] is
> the fix for the hypothesis, meaning PATCH[2/2] is the real fix. However,
> it would be nice to fix all of them in one shoot. I will try with PATCH[2/2]
> only to see if our issue will disappear or not. However, the issue still
> exists if PATCH[2/2] is missed.
> 

Jason, PATCH[2/2] is sufficient to fix our current issue. I tried with PATCH[2/2]
only and unable to hit the issue. However, PATCH[1/2] may be needed by other scenarios.
So it would be nice to fix them in one shoot.


> Firstly, We were failing on the transmit queue and {tvq, rvq}->busyloop_timeout
> == false if I remember correctly. So the added smp_rmb() in vhost_vq_avail_empty()
> is only a concern to tx_can_batch(). A mutex isn't enough to ensure the order
> for the available index and available ring entry (head). For example, vhost_vq_avail_empty()
> called by tx_can_batch() can see next available index, but its corresponding
> available ring entry (head) may not be seen by vhost yet if smp_rmb() is missed.
> The next call to get_tx_bufs(), where the available ring entry (head) doesn't
> arrived yet, leading to stale available ring entry (head) being fetched.
> 
>    handle_tx_copy
>      get_tx_bufs                 // smp_rmb() won't be executed when vq->avail_idx != vq->last_avail_idx
>      tx_can_batch
>        vhost_vq_avail_empty      // vq->avail_idx is updated from vq->avail->idx
> 
> The reason why I added smp_rmb() to vhost_vq_avail_empty() is because the function
> is a exposed API, even it's only used by drivers/vhost/net.c at present. It means
> the API has been broken internally. So it seems more appropriate to fix it up in
> vhost_vq_avail_empty() so that the API's users needn't worry about the memory access
> order.
> 
>>>
>>>
>>>> +               smp_rmb();
>>>> +               return false;
>>>> +       }
>>>>
>>>> -       return vq->avail_idx == vq->last_avail_idx;
>>>> +       return true;
>>>>   }
>>>>   EXPORT_SYMBOL_GPL(vhost_vq_avail_empty);

Thanks,
Gavin