Hi Ferruh,
On 9/16/21 13:15, Ferruh Yigit wrote:
On 8/20/2021 1:44 PM, Jiayu Hu wrote:
Copy threshold is introduced in async vhost data path to select
the appropriate copy engine to do copies for higher efficiency.
However, it may cause packets out-of-order, and it also causes
data path performance unpredictable.
Just checking, if the patch is fixing data performance issue, should it be
backported?
This is experimental code that is not used in real product for now, and
will likely heavily change given OVS rejected this design.
So I would personally not bother backporting patches for it.
Jiayu, Chenbo, do you think otherwise?
Regards,
Maxime
Therefore, this patch removes copy threshold support in async vhost
data path.
Signed-off-by: Jiayu Hu <jiayu...@intel.com>
Signed-off-by: Cheng Jiang <cheng1.ji...@intel.com>
<...>
