On Mon, 1 Oct 2018 17:37:06 +0300
Ilias Apalodimas wrote:
> On Mon, Oct 01, 2018 at 03:48:45PM +0200, Jesper Dangaard Brouer wrote:
> > On Mon, 1 Oct 2018 14:20:21 +0300
> > Ilias Apalodimas wrote:
> >
> > > On Mon, Oct 01, 2018 at 01:03:13PM +0200, Jesper Dangaard Brouer wrote:
> > > > On
On Mon, Oct 01, 2018 at 03:48:45PM +0200, Jesper Dangaard Brouer wrote:
> On Mon, 1 Oct 2018 14:20:21 +0300
> Ilias Apalodimas wrote:
>
> > On Mon, Oct 01, 2018 at 01:03:13PM +0200, Jesper Dangaard Brouer wrote:
> > > On Mon, 1 Oct 2018 12:56:58 +0300
> > > Ilias Apalodimas wrote:
> > >
> >
On Mon, 1 Oct 2018 14:20:21 +0300
Ilias Apalodimas wrote:
> On Mon, Oct 01, 2018 at 01:03:13PM +0200, Jesper Dangaard Brouer wrote:
> > On Mon, 1 Oct 2018 12:56:58 +0300
> > Ilias Apalodimas wrote:
> >
> > > > > #2: You have allocations on the XDP fast-path.
> > > > >
> > > > > The REAL
On Mon, Oct 01, 2018 at 01:03:13PM +0200, Jesper Dangaard Brouer wrote:
> On Mon, 1 Oct 2018 12:56:58 +0300
> Ilias Apalodimas wrote:
>
> > > > #2: You have allocations on the XDP fast-path.
> > > >
> > > > The REAL secret behind the XDP performance is to avoid allocations on
> > > > the
On Mon, 1 Oct 2018 12:56:58 +0300
Ilias Apalodimas wrote:
> > > #2: You have allocations on the XDP fast-path.
> > >
> > > The REAL secret behind the XDP performance is to avoid allocations on
> > > the fast-path. While I just told you to use the page-allocator and
> > > order-0 pages, this
> > #2: You have allocations on the XDP fast-path.
> >
> > The REAL secret behind the XDP performance is to avoid allocations on
> > the fast-path. While I just told you to use the page-allocator and
> > order-0 pages, this will actually kill performance. Thus, to make this
> > fast, you need a
On Mon, Oct 01, 2018 at 11:26:31AM +0200, Jesper Dangaard Brouer wrote:
>
> On Sat, 29 Sep 2018 14:28:01 +0300 Ilias Apalodimas
> wrote:
>
> > +static void *netsec_alloc_rx_data(struct netsec_priv *priv,
> > + dma_addr_t *dma_handle, u16 *desc_len)
> > +{
> > +
On Sat, 29 Sep 2018 14:28:01 +0300 Ilias Apalodimas
wrote:
> +static void *netsec_alloc_rx_data(struct netsec_priv *priv,
> + dma_addr_t *dma_handle, u16 *desc_len)
> +{
> + size_t len = priv->ndev->mtu + ETH_HLEN + 2 * VLAN_HLEN + NET_SKB_PAD +
> +
Current driver dynamically allocates an skb and maps it as DMA rx buffer.
A following patch introduces XDP functionality, so we need a
different allocation scheme. Buffers are allocated dynamically and
mapped into hardware. During the Rx operation the driver uses
build_skb() to produce the