Am 26.09.2011 08:15, schrieb Zhi Yong Wu:
> On Fri, Sep 23, 2011 at 11:54 PM, Kevin Wolf <kw...@redhat.com> wrote:
>>> +}
>>> +
>>> +static void bdrv_block_timer(void *opaque)
>>> +{
>>> + BlockDriverState *bs = opaque;
>>> + BlockQueue *queue = bs->block_queue;
>>> +
>>> + qemu_block_queue_flush(queue);
>>
>> Hm, didn't really notice it while reading patch 1, but
>> qemu_block_queue_flush() is misleading. It's really something like
> Why do you say this is misleading?
>> qemu_block_queue_submit().
> Right. It will resubmit all enqueued I/O requests.
For me, flush sounds as if it waits for completion of all requests.
Kevin
