From mboxrd@z Thu Jan 1 00:00:00 1970 From: Shirley Ma Subject: Re: [PATCH V4 2/8] netdevice.h: Add a new zerocopy device flag Date: Thu, 05 May 2011 14:02:17 -0700 Message-ID: <1304629337.7076.62.camel@localhost.localdomain> References: <1304495717.20660.67.camel@localhost.localdomain> <4DC1DD9D.6070708@us.ibm.com> Mime-Version: 1.0 Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: 7bit Cc: David Miller , mst@redhat.com, Eric Dumazet , Avi Kivity , Arnd Bergmann , netdev@vger.kernel.org, kvm@vger.kernel.org, linux-kernel@vger.kernel.org To: Sridhar Samudrala Return-path: In-Reply-To: <4DC1DD9D.6070708@us.ibm.com> Sender: linux-kernel-owner@vger.kernel.org List-Id: netdev.vger.kernel.org On Wed, 2011-05-04 at 16:13 -0700, Sridhar Samudrala wrote: > On Wed, 2011-05-04 at 00:55 -0700, Shirley Ma wrote: > > Signed-off-by: Shirley Ma > > --- > > > > include/linux/netdevice.h | 10 ++++++++++ > > 1 files changed, 10 insertions(+), 0 deletions(-) > > > > diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h > > index 0249fe7..0808f1e 100644 > > --- a/include/linux/netdevice.h > > +++ b/include/linux/netdevice.h > > @@ -1067,6 +1067,16 @@ struct net_device { > > #define NETIF_F_RXHASH (1 << 28) /* Receive hashing offload */ > > #define NETIF_F_RXCSUM (1 << 29) /* Receive checksumming offload */ > > > > +/* > > + * Bit 31 is for device to map userspace buffers -- zerocopy > > + * Device can set this flag when it supports HIGHDMA. > > + * Device can't recycle this kind of skb buffers. > > + * There are 256 bytes copied, the rest of buffers are mapped. > > + * The userspace callback should only be called when last reference to this skb > > + * is gone. > > + */ > > +#define NETIF_F_ZEROCOPY (1 << 31) > > + > Should we make this TX_ZEROCOPY as we may need a seperate flag for > RX_ZEROCOPY in future? > > Same with socket flag. We might be able to use the same flag for both TX and RX. If RX zero-copy requires only HIGHDMA feature only. If not, then we might need different feature bit. Same for sock flag. > Thanks > Sridhar > >