[ovs-dev] [PATCHv5] netdev-afxdp: Add need_wakeup supprt.
ciara.loftus at intel.com
Wed Oct 16 16:13:14 UTC 2019
> The patch adds support for using need_wakeup flag in AF_XDP rings.
> A new option, use_need_wakeup, is added. When this option is used,
> it means that OVS has to explicitly wake up the kernel RX, using poll()
> syscall and wake up TX, using sendto() syscall. This feature improves
> the performance by avoiding unnecessary sendto syscalls for TX.
> For RX, instead of kernel always busy-spinning on fille queue, OVS wakes
> up the kernel RX processing when fill queue is replenished.
> The need_wakeup feature is merged into Linux kernel bpf-next tee with
> 77cd0d7b3f25 ("xsk: add support for need_wakeup flag in AF_XDP rings")
> OVS enables it by default. Running the feature before this version causes
> xsk bind fails, please use options:use_need_wakeup=false to disable it.
> If users enable it but runs in an older version of libbpf, then the
> need_wakeup feature has no effect, and a warning message is logged.
> For virtual interface, it's better set use_need_wakeup=false, since
> the virtual device's AF_XDP xmit is synchronous: the sendto syscall
> enters kernel and process the TX packet on tx queue directly.
> On Intel Xeon E5-2620 v3 2.4GHz system, performance of physical port
> to physical port improves from 6.1Mpps to 7.3Mpps.
Thanks for the patch.
I'm wondering if/how you've pinned the IRQs for the physical ports in the test above.
Are the IRQs pinned to standalone cores or does the PMD thread share the same core?
I'd be interested in seeing performance results for both, if you have them.
> Suggested-by: Ilya Maximets <i.maximets at ovn.org>
> Signed-off-by: William Tu <u9012063 at gmail.com>
More information about the dev