On Fri, 5 Mar 2021 12:07:28 -0800
William Tu wrote:
> Thanks Björn and Jesper for feedbacks.
>
> On Fri, Mar 5, 2021 at 7:34 AM Björn Töpel wrote:
> >
> > On 2021-03-05 16:07, Jesper Dangaard Brouer wrote:
> > > On Fri, 5 Mar 2021 14:56:02 +0100
> > > Björn Töpel wrote:
> > >
> > >> On 202
Thanks Björn and Jesper for feedbacks.
On Fri, Mar 5, 2021 at 7:34 AM Björn Töpel wrote:
>
> On 2021-03-05 16:07, Jesper Dangaard Brouer wrote:
> > On Fri, 5 Mar 2021 14:56:02 +0100
> > Björn Töpel wrote:
> >
> >> On 2021-03-05 11:13, Jesper Dangaard Brouer wrote:
>
>
> [...]
> >>
> >> You'd lik
On 2021-03-05 16:07, Jesper Dangaard Brouer wrote:
On Fri, 5 Mar 2021 14:56:02 +0100
Björn Töpel wrote:
On 2021-03-05 11:13, Jesper Dangaard Brouer wrote:
[...]
We also want this sizeof-offset to be dynamic.
And I hope that AF_XDP could provide this info, on the size of
data_meta area. I
On Fri, 5 Mar 2021 14:56:02 +0100
Björn Töpel wrote:
> On 2021-03-05 11:13, Jesper Dangaard Brouer wrote:
> >
> > Bjørn and Magnus please take a look at my questions inlined below.
> >
> > On Thu, 4 Mar 2021 10:27:05 -0800
> > William Tu wrote:
> >
> >> One big problem of netdev-afxdp is t
On 2021-03-05 11:13, Jesper Dangaard Brouer wrote:
Bjørn and Magnus please take a look at my questions inlined below.
On Thu, 4 Mar 2021 10:27:05 -0800
William Tu wrote:
One big problem of netdev-afxdp is that there is no metadata support
from the hardware at all. For example, OVS netdev-a
Bjørn and Magnus please take a look at my questions inlined below.
On Thu, 4 Mar 2021 10:27:05 -0800
William Tu wrote:
> One big problem of netdev-afxdp is that there is no metadata support
> from the hardware at all. For example, OVS netdev-afxdp has to do rxhash,
> or TCP checksum in softwa
One big problem of netdev-afxdp is that there is no metadata support
from the hardware at all. For example, OVS netdev-afxdp has to do rxhash,
or TCP checksum in software, resulting in high performance overhead.
A generic meta data type for XDP frame using BTF is proposed[1] and
there is sample i