netdev - Re: [PATCH bpf-next] xsk: include XDP meta data in AF

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <CAJ+HfNhmtLrrALC_4fQNofh+Y3E5_QxXrqDsqin2jiPLwcMoVw@mail.gmail.com>
Date:   Thu, 30 Aug 2018 14:51:05 +0200
From:   Björn Töpel <bjorn.topel@...il.com>
To:     Daniel Borkmann <daniel@...earbox.net>
Cc:     "Karlsson, Magnus" <magnus.karlsson@...el.com>,
        Magnus Karlsson <magnus.karlsson@...il.com>, ast@...nel.org,
        Netdev <netdev@...r.kernel.org>,
        Björn Töpel <bjorn.topel@...el.com>,
        Jesper Dangaard Brouer <brouer@...hat.com>
Subject: Re: [PATCH bpf-next] xsk: include XDP meta data in AF_XDP frames

Den tors 30 aug. 2018 kl 14:37 skrev Daniel Borkmann <daniel@...earbox.net>:
>
> On 08/30/2018 10:09 AM, Björn Töpel wrote:
> > From: Björn Töpel <bjorn.topel@...el.com>
> >
> > Previously, the AF_XDP (XDP_DRV/XDP_SKB copy-mode) ingress logic did
> > not include XDP meta data in the data buffers copied out to the user
> > application.
> >
> > In this commit, we check if meta data is available, and if so, it is
> > prepended to the frame.
> >
> > Signed-off-by: Björn Töpel <bjorn.topel@...el.com>
> > ---
> >  net/xdp/xsk.c | 37 ++++++++++++++++++++++++++++---------
> >  1 file changed, 28 insertions(+), 9 deletions(-)
> >
> > diff --git a/net/xdp/xsk.c b/net/xdp/xsk.c
> > index 4e937cd7c17d..817e4cee1540 100644
> > --- a/net/xdp/xsk.c
> > +++ b/net/xdp/xsk.c
> > @@ -55,20 +55,30 @@ EXPORT_SYMBOL(xsk_umem_discard_addr);
> >
> >  static int __xsk_rcv(struct xdp_sock *xs, struct xdp_buff *xdp, u32 len)
> >  {
> > -     void *buffer;
> > +     void *to_buf, *from_buf;
> > +     u32 metalen;
> >       u64 addr;
> >       int err;
> >
> >       if (!xskq_peek_addr(xs->umem->fq, &addr) ||
> > -         len > xs->umem->chunk_size_nohr) {
> > +         len > xs->umem->chunk_size_nohr - XDP_PACKET_HEADROOM) {
> >               xs->rx_dropped++;
> >               return -ENOSPC;
> >       }
> >
> >       addr += xs->umem->headroom;
> >
> > -     buffer = xdp_umem_get_data(xs->umem, addr);
> > -     memcpy(buffer, xdp->data, len);
> > +     if (xdp_data_meta_unsupported(xdp)) {
>
> Nit: Probably makes sense to wrap with unlikely() since we expect all
> drivers to implement this.
>

I'll send a v2 fixing this...

> > +             from_buf = xdp->data;
> > +             metalen = 0;
> > +     } else {
> > +             from_buf = xdp->data_meta;
> > +             metalen = xdp->data - xdp->data_meta;
> > +     }
> > +
> > +     to_buf = xdp_umem_get_data(xs->umem, addr);
> > +     memcpy(to_buf, from_buf, len + metalen);
> > +     addr += metalen;
> >       err = xskq_produce_batch_desc(xs->rx, addr, len);
> >       if (!err) {
> >               xskq_discard_addr(xs->umem->fq);
> > @@ -111,8 +121,8 @@ void xsk_flush(struct xdp_sock *xs)
> >
> >  int xsk_generic_rcv(struct xdp_sock *xs, struct xdp_buff *xdp)
> >  {
> > -     u32 len = xdp->data_end - xdp->data;
> > -     void *buffer;
> > +     u32 metalen, len = xdp->data_end - xdp->data;
> > +     void *to_buf, *from_buf;
> >       u64 addr;
> >       int err;
> >
> > @@ -120,15 +130,24 @@ int xsk_generic_rcv(struct xdp_sock *xs, struct xdp_buff *xdp)
> >               return -EINVAL;
> >
> >       if (!xskq_peek_addr(xs->umem->fq, &addr) ||
> > -         len > xs->umem->chunk_size_nohr) {
> > +         len > xs->umem->chunk_size_nohr - XDP_PACKET_HEADROOM) {
> >               xs->rx_dropped++;
> >               return -ENOSPC;
> >       }
> >
> >       addr += xs->umem->headroom;
> >
> > -     buffer = xdp_umem_get_data(xs->umem, addr);
> > -     memcpy(buffer, xdp->data, len);
> > +     if (xdp_data_meta_unsupported(xdp)) {
> > +             from_buf = xdp->data;
> > +             metalen = 0;
>
> Note that this condition should be dead code. netif_receive_generic_xdp()
> sets xdp->data_meta to xdp->data, so all good here and above is never hit.
>

...and this! Thanks for pointing this out!


Björn

> > +     } else {
> > +             from_buf = xdp->data_meta;
> > +             metalen = xdp->data - xdp->data_meta;
> > +     }
> > +
> > +     to_buf = xdp_umem_get_data(xs->umem, addr);
> > +     memcpy(to_buf, from_buf, len + metalen);
> > +     addr += metalen;
> >       err = xskq_produce_batch_desc(xs->rx, addr, len);
> >       if (!err) {
> >               xskq_discard_addr(xs->umem->fq);
> >
>