On Mon, Jan 29, 2024 at 7:40 PM wangyunjian <wangyunjian@xxxxxxxxxx> wrote: > > > -----Original Message----- > > From: Jason Wang [mailto:jasowang@xxxxxxxxxx] > > Sent: Monday, January 29, 2024 11:03 AM > > To: wangyunjian <wangyunjian@xxxxxxxxxx> > > Cc: mst@xxxxxxxxxx; willemdebruijn.kernel@xxxxxxxxx; kuba@xxxxxxxxxx; > > davem@xxxxxxxxxxxxx; magnus.karlsson@xxxxxxxxx; netdev@xxxxxxxxxxxxxxx; > > linux-kernel@xxxxxxxxxxxxxxx; kvm@xxxxxxxxxxxxxxx; > > virtualization@xxxxxxxxxxxxxxx; xudingke <xudingke@xxxxxxxxxx> > > Subject: Re: [PATCH net-next 2/2] tun: AF_XDP Rx zero-copy support > > > > On Thu, Jan 25, 2024 at 8:54 PM wangyunjian <wangyunjian@xxxxxxxxxx> > > wrote: > > > > > > > > > > > > > -----Original Message----- > > > > From: Jason Wang [mailto:jasowang@xxxxxxxxxx] > > > > Sent: Thursday, January 25, 2024 12:49 PM > > > > To: wangyunjian <wangyunjian@xxxxxxxxxx> > > > > Cc: mst@xxxxxxxxxx; willemdebruijn.kernel@xxxxxxxxx; > > > > kuba@xxxxxxxxxx; davem@xxxxxxxxxxxxx; magnus.karlsson@xxxxxxxxx; > > > > netdev@xxxxxxxxxxxxxxx; linux-kernel@xxxxxxxxxxxxxxx; > > > > kvm@xxxxxxxxxxxxxxx; virtualization@xxxxxxxxxxxxxxx; xudingke > > > > <xudingke@xxxxxxxxxx> > > > > Subject: Re: [PATCH net-next 2/2] tun: AF_XDP Rx zero-copy support > > > > > > > > On Wed, Jan 24, 2024 at 5:38 PM Yunjian Wang > > > > <wangyunjian@xxxxxxxxxx> > > > > wrote: > > > > > > > > > > Now the zero-copy feature of AF_XDP socket is supported by some > > > > > drivers, which can reduce CPU utilization on the xdp program. > > > > > This patch set allows tun to support AF_XDP Rx zero-copy feature. > > > > > > > > > > This patch tries to address this by: > > > > > - Use peek_len to consume a xsk->desc and get xsk->desc length. > > > > > - When the tun support AF_XDP Rx zero-copy, the vq's array maybe empty. > > > > > So add a check for empty vq's array in vhost_net_buf_produce(). > > > > > - add XDP_SETUP_XSK_POOL and ndo_xsk_wakeup callback support > > > > > - add tun_put_user_desc function to copy the Rx data to VM > > > > > > > > Code explains themselves, let's explain why you need to do this. > > > > > > > > 1) why you want to use peek_len > > > > 2) for "vq's array", what does it mean? > > > > 3) from the view of TUN/TAP tun_put_user_desc() is the TX path, so I > > > > guess you meant TX zerocopy instead of RX (as I don't see codes for > > > > RX?) > > > > > > OK, I agree and use TX zerocopy instead of RX zerocopy. I meant RX > > > zerocopy from the view of vhost-net. > > > > Ok. > > > > > > > > > > > > > A big question is how could you handle GSO packets from > > userspace/guests? > > > > > > Now by disabling VM's TSO and csum feature. > > > > Btw, how could you do that? > > By set network backend-specific options: > <driver name='vhost'> > <host csum='off' gso='off' tso4='off' tso6='off' ecn='off' ufo='off' mrg_rxbuf='off'/> > <guest csum='off' tso4='off' tso6='off' ecn='off' ufo='off'/> > </driver> This is the mgmt work, but the problem is what happens if GSO is not disabled in the guest, or is there a way to: 1) forcing the guest GSO to be off 2) a graceful fallback Thanks > > Thanks > > > > > Thanks > > >
