netdev - Re: [bpf-next RFC 1/3] xsk: add xsk_umem_consume_tx

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: <CALDO+SYWu92Wi+s8SAM2fXUekn0yMnxoqbwr5aTMKbockdMvgg@mail.gmail.com>
Date:   Tue, 18 Dec 2018 10:39:51 -0800
From:   William Tu <u9012063@...il.com>
To:     Björn Töpel <bjorn.topel@...il.com>
Cc:     Magnus Karlsson <magnus.karlsson@...il.com>,
        Alexei Starovoitov <ast@...nel.org>,
        Daniel Borkmann <daniel@...earbox.net>,
        Netdev <netdev@...r.kernel.org>, makita.toshiaki@....ntt.co.jp,
        "Karlsson, Magnus" <magnus.karlsson@...el.com>
Subject: Re: [bpf-next RFC 1/3] xsk: add xsk_umem_consume_tx_virtual.

On Tue, Dec 18, 2018 at 6:34 AM Björn Töpel <bjorn.topel@...il.com> wrote:
>
> Den mån 17 dec. 2018 kl 20:40 skrev William Tu <u9012063@...il.com>:
> >
> > Currently the xsk_umem_consume_tx expects only the physical NICs so
> > the api returns a dma address.  This patch introduce the new function
> > to return the virtual address, when XSK is used by a virtual device.
> >
> > Signed-off-by: William Tu <u9012063@...il.com>
> > ---
> >  include/net/xdp_sock.h |  7 +++++++
> >  net/xdp/xsk.c          | 24 ++++++++++++++++++++++++
> >  2 files changed, 31 insertions(+)
> >
> > diff --git a/include/net/xdp_sock.h b/include/net/xdp_sock.h
> > index 13acb9803a6d..8de6b8456945 100644
> > --- a/include/net/xdp_sock.h
> > +++ b/include/net/xdp_sock.h
> > @@ -81,6 +81,7 @@ u64 *xsk_umem_peek_addr(struct xdp_umem *umem, u64 *addr);
> >  void xsk_umem_discard_addr(struct xdp_umem *umem);
> >  void xsk_umem_complete_tx(struct xdp_umem *umem, u32 nb_entries);
> >  bool xsk_umem_consume_tx(struct xdp_umem *umem, dma_addr_t *dma, u32 *len);
> > +bool xsk_umem_consume_tx_virtual(struct xdp_umem *umem, char **addr, u32 *len);
> >  void xsk_umem_consume_tx_done(struct xdp_umem *umem);
> >  struct xdp_umem_fq_reuse *xsk_reuseq_prepare(u32 nentries);
> >  struct xdp_umem_fq_reuse *xsk_reuseq_swap(struct xdp_umem *umem,
> > @@ -165,6 +166,12 @@ static inline bool xsk_umem_consume_tx(struct xdp_umem *umem, dma_addr_t *dma,
> >         return false;
> >  }
> >
> > +static inline bool xsk_umem_consume_tx_virtual(struct xdp_umem *umem,
> > +                                              char **dma, u32 *len)
>
> Nit: dma->addr
>
> > +{
> > +       return false;
> > +}
> > +
> >  static inline void xsk_umem_consume_tx_done(struct xdp_umem *umem)
> >  {
> >  }
> > diff --git a/net/xdp/xsk.c b/net/xdp/xsk.c
> > index 07156f43d295..379f5e9d0c81 100644
> > --- a/net/xdp/xsk.c
> > +++ b/net/xdp/xsk.c
> > @@ -197,6 +197,30 @@ bool xsk_umem_consume_tx(struct xdp_umem *umem, dma_addr_t *dma, u32 *len)
> >  }
> >  EXPORT_SYMBOL(xsk_umem_consume_tx);
> >
> > +bool xsk_umem_consume_tx_virtual(struct xdp_umem *umem, char **addr, u32 *len)
>
> Prefer void ** in favor of char**?
>
> > +{
> > +       struct xdp_desc desc;
> > +       struct xdp_sock *xs;
> > +
> > +       rcu_read_lock();
> > +       list_for_each_entry_rcu(xs, &umem->xsk_list, list) {
> > +               if (!xskq_peek_desc(xs->tx, &desc))
> > +                        continue;
> > +               if (xskq_produce_addr_lazy(umem->cq, desc.addr))
> > +                       goto out;
> > +
> > +               *addr = xdp_umem_get_data(umem, desc.addr);
> > +               *len = desc.len;
> > +               xskq_discard_desc(xs->tx);
> > +               rcu_read_unlock();
> > +               return true;
> > +       }
> > +out:
> > +       rcu_read_unlock();
> > +       return false;
> > +}
> > +EXPORT_SYMBOL(xsk_umem_consume_tx_virtual);
>
> Is possible to make xsk_umem_consume_tx_virtual and
> xsk_umem_consume_tx share a common implementation. They are very
> similar.

Sure, will do it in next version.
Thanks
William
>
> > +
> >  static int xsk_zc_xmit(struct sock *sk)
> >  {
> >         struct xdp_sock *xs = xdp_sk(sk);
> > --
> > 2.7.4
> >