On Mon, May 30, 2016 at 05:59:33PM +0800, Jason Wang wrote:
OK and skb ring is 9% faster than the linked list, so together
On 2016å05æ23æ 18:43, Michael S. Tsirkin wrote:
This is in response to the proposal by Jason to make tunI change tun to use skb array, looks like it can give about 5% more faster
rx packet queue lockless using a circular buffer.
My testing seems to show that at least for the common usecase
in networking, which isn't lockless, circular buffer
with indices does not perform that well, because
each index access causes a cache line to bounce between
CPUs, and index access causes stalls due to the dependency.
than skb ring.
this is a 14% speedup?
And we usually don't need touch bhs during consume and produce (e.g for theMaybe I'll drop it in v6 then ...
case of tun).
Thanks
Could you post the full tun patchset please?