Re: [RFC net-next PATCH 3/5] net: introduce XDP driver features interface

2017-05-30 Thread Jesper Dangaard Brouer
On Mon, 22 May 2017 19:07:44 +0200 Daniel Borkmann wrote: > On 05/22/2017 04:49 PM, Jesper Dangaard Brouer wrote: > > > > How do we move forward from here? > > If we introduce such feature bits one day, one possibility I see > that more or less could work is to propagate this into tail call >

Re: [RFC net-next PATCH 3/5] net: introduce XDP driver features interface

2017-05-22 Thread Daniel Borkmann
On 05/22/2017 04:49 PM, Jesper Dangaard Brouer wrote: On Sun, 21 May 2017 02:58:19 +0200 Daniel Borkmann wrote: On 05/20/2017 09:53 AM, Jesper Dangaard Brouer wrote: On Fri, 19 May 2017 19:13:29 +0200 Daniel Borkmann wrote: On 05/18/2017 05:41 PM, Jesper Dangaard Brouer wrote: There is a fu

Re: [RFC net-next PATCH 3/5] net: introduce XDP driver features interface

2017-05-22 Thread Jesper Dangaard Brouer
On Sun, 21 May 2017 02:58:19 +0200 Daniel Borkmann wrote: > On 05/20/2017 09:53 AM, Jesper Dangaard Brouer wrote: > > On Fri, 19 May 2017 19:13:29 +0200 > > Daniel Borkmann wrote: > > > >> On 05/18/2017 05:41 PM, Jesper Dangaard Brouer wrote: > >>> There is a fundamental difference between n

Re: [RFC net-next PATCH 3/5] net: introduce XDP driver features interface

2017-05-20 Thread Daniel Borkmann
On 05/20/2017 09:53 AM, Jesper Dangaard Brouer wrote: On Fri, 19 May 2017 19:13:29 +0200 Daniel Borkmann wrote: On 05/18/2017 05:41 PM, Jesper Dangaard Brouer wrote: There is a fundamental difference between normal eBPF programs and (XDP) eBPF programs getting attached in a driver. For normal

Re: [RFC net-next PATCH 3/5] net: introduce XDP driver features interface

2017-05-20 Thread Jesper Dangaard Brouer
On Fri, 19 May 2017 19:13:29 +0200 Daniel Borkmann wrote: > On 05/18/2017 05:41 PM, Jesper Dangaard Brouer wrote: > > There is a fundamental difference between normal eBPF programs > > and (XDP) eBPF programs getting attached in a driver. For normal > > eBPF programs it is easy to add a new bpf f

Re: [RFC net-next PATCH 3/5] net: introduce XDP driver features interface

2017-05-19 Thread David Miller
From: Daniel Borkmann Date: Fri, 19 May 2017 19:13:29 +0200 > The problem is that once you add bits markers to bpf_prog like we > used to do in the past, then as you do in patch 4/5 with the > xdp_rxhash_needed bit, they will need to be turned /on/ > unconditionally when a prog has tail calls. Y

Re: [RFC net-next PATCH 3/5] net: introduce XDP driver features interface

2017-05-19 Thread Daniel Borkmann
On 05/18/2017 05:41 PM, Jesper Dangaard Brouer wrote: There is a fundamental difference between normal eBPF programs and (XDP) eBPF programs getting attached in a driver. For normal eBPF programs it is easy to add a new bpf feature, like a bpf helper, because is it strongly tied to the feature be

[RFC net-next PATCH 3/5] net: introduce XDP driver features interface

2017-05-18 Thread Jesper Dangaard Brouer
There is a fundamental difference between normal eBPF programs and (XDP) eBPF programs getting attached in a driver. For normal eBPF programs it is easy to add a new bpf feature, like a bpf helper, because is it strongly tied to the feature being available in the current core kernel code. When dri