On Mon, Mar 07, 2016 at 07:52:22AM +, Xie, Huawei wrote:
> On 2/18/2016 9:48 PM, Yuanhan Liu wrote:
> > Current virtio_dev_merge_rx() implementation just looks like the
> > old rte_vhost_dequeue_burst(), full of twisted logic, that you
> > can see same code block in quite many different places.
On Mon, Mar 07, 2016 at 07:16:39AM +, Xie, Huawei wrote:
> On 3/7/2016 3:04 PM, Xie, Huawei wrote:
> > On 3/7/2016 2:49 PM, Yuanhan Liu wrote:
> >> On Mon, Mar 07, 2016 at 06:38:42AM +, Xie, Huawei wrote:
> >>> On 3/7/2016 2:35 PM, Yuanhan Liu wrote:
> On Mon, Mar 07, 2016 at 06:22:25A
On Mon, Mar 07, 2016 at 06:38:42AM +, Xie, Huawei wrote:
> On 3/7/2016 2:35 PM, Yuanhan Liu wrote:
> > On Mon, Mar 07, 2016 at 06:22:25AM +, Xie, Huawei wrote:
> >> On 2/18/2016 9:48 PM, Yuanhan Liu wrote:
> >>> + uint16_t idx = vq->avail->ring[avail_idx & (vq->size - 1)];
> >>> + uint32_t
On Mon, Mar 07, 2016 at 06:22:25AM +, Xie, Huawei wrote:
> On 2/18/2016 9:48 PM, Yuanhan Liu wrote:
> > + uint16_t idx = vq->avail->ring[avail_idx & (vq->size - 1)];
> > + uint32_t vec_id = *vec_idx;
> > + uint32_t len= *allocated;
> >
> There is bug not using volatile to retrieve t
On 3/7/2016 4:36 PM, Yuanhan Liu wrote:
> On Mon, Mar 07, 2016 at 07:52:22AM +, Xie, Huawei wrote:
>> On 2/18/2016 9:48 PM, Yuanhan Liu wrote:
>>> Current virtio_dev_merge_rx() implementation just looks like the
>>> old rte_vhost_dequeue_burst(), full of twisted logic, that you
>>> can see same
On 2/18/2016 9:48 PM, Yuanhan Liu wrote:
> Current virtio_dev_merge_rx() implementation just looks like the
> old rte_vhost_dequeue_burst(), full of twisted logic, that you
> can see same code block in quite many different places.
>
> However, the logic of virtio_dev_merge_rx() is quite similar to
On 3/7/2016 3:04 PM, Xie, Huawei wrote:
> On 3/7/2016 2:49 PM, Yuanhan Liu wrote:
>> On Mon, Mar 07, 2016 at 06:38:42AM +, Xie, Huawei wrote:
>>> On 3/7/2016 2:35 PM, Yuanhan Liu wrote:
On Mon, Mar 07, 2016 at 06:22:25AM +, Xie, Huawei wrote:
> On 2/18/2016 9:48 PM, Yuanhan Liu wro
On 3/7/2016 2:49 PM, Yuanhan Liu wrote:
> On Mon, Mar 07, 2016 at 06:38:42AM +, Xie, Huawei wrote:
>> On 3/7/2016 2:35 PM, Yuanhan Liu wrote:
>>> On Mon, Mar 07, 2016 at 06:22:25AM +, Xie, Huawei wrote:
On 2/18/2016 9:48 PM, Yuanhan Liu wrote:
> + uint16_t idx = vq->avail->ring[ava
On 3/7/2016 2:35 PM, Yuanhan Liu wrote:
> On Mon, Mar 07, 2016 at 06:22:25AM +, Xie, Huawei wrote:
>> On 2/18/2016 9:48 PM, Yuanhan Liu wrote:
>>> + uint16_t idx = vq->avail->ring[avail_idx & (vq->size - 1)];
>>> + uint32_t vec_id = *vec_idx;
>>> + uint32_t len= *allocated;
>>>
>> T
On 2/18/2016 9:48 PM, Yuanhan Liu wrote:
> + uint16_t idx = vq->avail->ring[avail_idx & (vq->size - 1)];
> + uint32_t vec_id = *vec_idx;
> + uint32_t len= *allocated;
>
There is bug not using volatile to retrieve the avail idx.
Current virtio_dev_merge_rx() implementation just looks like the
old rte_vhost_dequeue_burst(), full of twisted logic, that you
can see same code block in quite many different places.
However, the logic of virtio_dev_merge_rx() is quite similar to
virtio_dev_rx(). The big difference is that the m
11 matches
Mail list logo
