Re: [dpdk-dev] [PATCH] vhost: flush used->idx update before reading avail->flags

["Xie, Huawei" <huawei.xie@intel.com>]

On 6/9/2015 4:47 PM, Michael S. Tsirkin wrote:
> On Tue, Jun 09, 2015 at 03:04:02PM +0800, Linhaifeng wrote:
>>
>> On 2015/4/24 15:27, Luke Gorrie wrote:
>>> On 24 April 2015 at 03:01, Linhaifeng <haifeng.lin@huawei.com> wrote:
>>>
>>>> If not add memory fence what would happen? Packets loss or interrupt
>>>> loss?How to test it ?
>>>>
>>> You should be able to test it like this:
>>>
>>> 1. Boot two Linux kernel (e.g. 3.13) guests.
>>> 2. Connect them via vhost switch.
>>> 3. Run continuous traffic between them (e.g. iperf).
>>>
>>> I would expect that within a reasonable timeframe (< 1 hour) one of the
>>> guests' network interfaces will hang indefinitely due to a missed interrupt.
>>>
>>> You won't be able to reproduce this using DPDK guests because they are not
>>> using the same interrupt suppression method.
>>>
>>> This is a serious real-world problem. I wouldn't deploy the vhost
>>> implementation without this fix.
>>>
>>> Cheers,
>>> -Luke
>>>
>> I think this patch can't resole this problem. On the other hand we still would miss interrupt.
>>
>> After add rte_mb() function the we want the case is :
>> 1.write used->idx. ring is full or empty.
>> 2.virtio_net open interrupt.
>> 3.read avail->flags.
>>
>> but this case(miss interrupt) would happen too:
>> 1.write used->idx. ring is full or empty.
>> 2.read avail->flags.
>> 3.virtio_net open interrupt.
>>
> That's why a correct guest, after detecting an empty used ring, must always
> re-check used idx at least once after writing avail->flags.
>
> By the way, similarly, host side must re-check avail idx after writing
> used flags. I don't see where snabbswitch does it - is that a bug
> in snabbswitch?
>
yes, both host and guest should recheck if there is more work added
after they toggle the flag.
For DPDK vHost, as it runs in polling mode, we will recheck avail idx
soon, so we don't need recheck.