On 14/10/2016 11:43, Fam Zheng wrote:
> On Thu, 10/13 19:34, Paolo Bonzini wrote:
>> Ensure that there are no changes between the last check to
>> bdrv_get_dirty_count and the switch to the target.
>>
>> There is already a bdrv_drained_end call, we only need to ensure
>> that bdrv_drained_begin is not called twice.
>>
>> Cc: qemu-sta...@nongnu.org
>
> Cc stable? I don't see an existing bug here, can you explain?
Hmm, I was not sure that mirror was safe for dataplane devices without
the drained section, but it looks like there is no "hole". So no need
to Cc stable.
>> @@ -802,9 +812,10 @@ immediate_exit:
>>
>> data = g_malloc(sizeof(*data));
>> data->ret = ret;
>> - /* Before we switch to target in mirror_exit, make sure data doesn't
>> - * change. */
>> - bdrv_drained_begin(bs);
>> +
>> + if (need_drain) {
>
> Not sure whether this if block is necessary (i.e. when !(cnt == 0 &&
> should_complete)), but it certainly doesn't hurt.
Yes, the alternative is to have something similar, to skip the
bdrv_drained_end, in mirror_exit.
Paolo
>> + bdrv_drained_begin(bs);
>> + }
>> block_job_defer_to_main_loop(&s->common, mirror_exit, data);
>> }
>>
>> --
>> 2.7.4
>>
>>
>
> Fam
>
