> -----Original Message-----
> From: Zhang, Qi Z <qi.z.zh...@intel.com>
> Sent: 2022年1月2日 16:46
> To: Liu, KevinX <kevinx....@intel.com>; dev@dpdk.org
> Cc: Zhang, RobinX <robinx.zh...@intel.com>; Wang, Jie1X
> <jie1x.w...@intel.com>; Liu, KevinX <kevinx....@intel.com>
> Subject: RE: [PATCH] net/ice: fix error forwarding IPv6 VXLAN packet
>
>
>
> > -----Original Message-----
> > From: Kevin Liu <kevinx....@intel.com>
> > Sent: Wednesday, December 8, 2021 5:56 PM
> > To: dev@dpdk.org
> > Cc: Zhang, RobinX <robinx.zh...@intel.com>; Wang, Jie1X
> > <jie1x.w...@intel.com>; Liu, KevinX <kevinx....@intel.com>
> > Subject: [PATCH] net/ice: fix error forwarding IPv6 VXLAN packet
> >
> > In ice_txd_enable_offload(), when set tunnel packet Tx checksum
> > offload enable, td_offset should be set with outer l2/l3 len instead of
> > inner
> l2/l3 len.
> >
> > This patch fix the bug that the checksum engine can forward tunnle packets.
>
> s/tunnle/tunnel
>
> >
> > Fixes: 28f9002ab67f ("net/ice: add Tx AVX512 offload path")
> >
> > Signed-off-by: Kevin Liu <kevinx....@intel.com>
> > ---
> > drivers/net/ice/ice_rxtx_vec_common.h | 52
> > +++++++++++++++++++--------
> > 1 file changed, 37 insertions(+), 15 deletions(-)
> >
> > diff --git a/drivers/net/ice/ice_rxtx_vec_common.h
> > b/drivers/net/ice/ice_rxtx_vec_common.h
> > index dfe60c81d9..8ff01046e1 100644
> > --- a/drivers/net/ice/ice_rxtx_vec_common.h
> > +++ b/drivers/net/ice/ice_rxtx_vec_common.h
> > @@ -364,23 +364,45 @@ ice_txd_enable_offload(struct rte_mbuf *tx_pkt,
> > uint32_t td_offset = 0;
> >
> > /* Tx Checksum Offload */
> > -/* SET MACLEN */
> > -td_offset |= (tx_pkt->l2_len >> 1) <<
> > +/*Tunnel package usage outer len enable L2/L3 checksum offload*/ if
> > +(ol_flags & RTE_MBUF_F_TX_TUNNEL_MASK) {
> > +/* SET MACLEN */
> > +td_offset |= (tx_pkt->outer_l2_len >> 1) <<
> > ICE_TX_DESC_LEN_MACLEN_S;
> >
> > -/* Enable L3 checksum offload */
> > -if (ol_flags & RTE_MBUF_F_TX_IP_CKSUM) { -td_cmd |=
> > ICE_TX_DESC_CMD_IIPT_IPV4_CSUM; -td_offset |= (tx_pkt->l3_len >> 2)
> <<
> > -ICE_TX_DESC_LEN_IPLEN_S; -} else if (ol_flags & RTE_MBUF_F_TX_IPV4) {
> > -td_cmd |= ICE_TX_DESC_CMD_IIPT_IPV4; -td_offset |= (tx_pkt->l3_len >>
> > 2) << -ICE_TX_DESC_LEN_IPLEN_S; -} else if (ol_flags &
> > RTE_MBUF_F_TX_IPV6) { -td_cmd |= ICE_TX_DESC_CMD_IIPT_IPV6; -
> td_offset
> > |= (tx_pkt->l3_len >> 2) << -ICE_TX_DESC_LEN_IPLEN_S;
> > +/* Enable L3 checksum offload */
> > +if (ol_flags & RTE_MBUF_F_TX_IP_CKSUM) { td_cmd |=
> > +ICE_TX_DESC_CMD_IIPT_IPV4_CSUM; td_offset |= (tx_pkt-
> >outer_l3_len >>
> > +2) << ICE_TX_DESC_LEN_IPLEN_S;
>
> Is this fix also cover IPv4 Tx checksum offload?
Yes, cover it.
> Please refine the title to reflect what exactly the patch does if necessary.
>
> Btw, could you also check if any fix needed in ice_txd_enable_checksum?
> I saw inconsistent offset configure on l3_len between scaler path and vector
> path.
> Its better to always keep them identical.
>
>
I checked code in ice_txd_enable_checksum, and it should also be fixed here. I
will send a new patch of v2.
Thanks
Kevin
