net/ice: fix GTPU/PPPoE packets with no hash value
Checks
Commit Message
When RSS init, because of profile overlap, the GTPU_IPV4 packets
don't hit GTPU_INNER_IPV4 profile which causes no hash value. Because
of no PPPoE profile, the PPPoE packets also has no hash value. This
patch solves this issue by pulling GTPU_IPV4 profile into inner ipv4
group and creating related PPPoE profile at the same time.
Fixes: 8cadc41294c6 ("net/ice: initialize and update RSS based on user config")
Signed-off-by: Simei Su <simei.su@intel.com>
---
drivers/net/ice/ice_ethdev.c | 80 ++++++++++++++++++++++++++++++++++++++++++++
1 file changed, 80 insertions(+)
Comments
> -----Original Message-----
> From: Su, Simei <simei.su@intel.com>
> Sent: Tuesday, July 7, 2020 9:38 AM
> To: Zhang, Qi Z <qi.z.zhang@intel.com>
> Cc: dev@dpdk.org; Zhao1, Wei <wei.zhao1@intel.com>; Jiang, JunyuX
> <junyux.jiang@intel.com>; Lu, Nannan <nannan.lu@intel.com>; Su, Simei
> <simei.su@intel.com>
> Subject: [PATCH] net/ice: fix GTPU/PPPoE packets with no hash value
>
> When RSS init, because of profile overlap, the GTPU_IPV4 packets don't hit
> GTPU_INNER_IPV4 profile which causes no hash value. Because of no PPPoE
> profile, the PPPoE packets also has no hash value. This patch solves this issue
> by pulling GTPU_IPV4 profile into inner ipv4 group and creating related
> PPPoE profile at the same time.
>
> Fixes: 8cadc41294c6 ("net/ice: initialize and update RSS based on user
> config")
>
> Signed-off-by: Simei Su <simei.su@intel.com>
Acked-by: Qi Zhang <qi.z.zhang@intel.com>
Applied to dpdk-next-net-intel.
Thanks
Qi
> -----Original Message-----
> From: Zhang, Qi Z
> Sent: Tuesday, July 7, 2020 10:30 AM
> To: Su, Simei <simei.su@intel.com>
> Cc: dev@dpdk.org; Zhao1, Wei <wei.zhao1@intel.com>; Jiang, JunyuX
> <junyux.jiang@intel.com>; Lu, Nannan <nannan.lu@intel.com>
> Subject: RE: [PATCH] net/ice: fix GTPU/PPPoE packets with no hash value
>
>
>
> > -----Original Message-----
> > From: Su, Simei <simei.su@intel.com>
> > Sent: Tuesday, July 7, 2020 9:38 AM
> > To: Zhang, Qi Z <qi.z.zhang@intel.com>
> > Cc: dev@dpdk.org; Zhao1, Wei <wei.zhao1@intel.com>; Jiang, JunyuX
> > <junyux.jiang@intel.com>; Lu, Nannan <nannan.lu@intel.com>; Su, Simei
> > <simei.su@intel.com>
> > Subject: [PATCH] net/ice: fix GTPU/PPPoE packets with no hash value
> >
> > When RSS init, because of profile overlap, the GTPU_IPV4 packets don't
> > hit
> > GTPU_INNER_IPV4 profile which causes no hash value. Because of no
> > PPPoE profile, the PPPoE packets also has no hash value. This patch
> > solves this issue by pulling GTPU_IPV4 profile into inner ipv4 group
> > and creating related PPPoE profile at the same time.
> >
> > Fixes: 8cadc41294c6 ("net/ice: initialize and update RSS based on user
> > config")
> >
> > Signed-off-by: Simei Su <simei.su@intel.com>
>
> Acked-by: Qi Zhang <qi.z.zhang@intel.com>
>
> Applied to dpdk-next-net-intel.
There is a compile issue be fixed before be applied.
>
> Thanks
> Qi
@@ -2533,6 +2533,86 @@ static int ice_parse_devargs(struct rte_eth_dev *dev)
PMD_DRV_LOG(ERR, "%s SCTP_IPV6 rss flow fail %d",
__func__, ret);
}
+
+ if (rss_hf & ETH_RSS_IPV4) {
+ ret = ice_add_rss_cfg(hw, vsi->idx, ICE_FLOW_HASH_IPV4,
+ ICE_FLOW_SEG_HDR_GTPU_IP, 0);
+ if (ret)
+ PMD_DRV_LOG(ERR, "%s GTPU_IPV4 rss flow fail %d",
+ __func__, ret);
+
+ ret = ice_add_rss_cfg(hw, vsi->idx, ICE_FLOW_HASH_IPV4,
+ ICE_FLOW_SEG_HDR_PPPOE, 0);
+ if (ret)
+ PMD_DRV_LOG(ERR, "%s PPPoE_IPV4 rss flow fail %d",
+ __func__, ret);
+ }
+
+ if (rss_hf & ETH_RSS_IPV6) {
+ ret = ice_add_rss_cfg(hw, vsi->idx, ICE_FLOW_HASH_IPV6,
+ ICE_FLOW_SEG_HDR_GTPU_IP, 0);
+ if (ret)
+ PMD_DRV_LOG(ERR, "%s GTPU_IPV6 rss flow fail %d",
+ __func__, ret);
+
+ ret = ice_add_rss_cfg(hw, vsi->idx, ICE_FLOW_HASH_IPV6,
+ ICE_FLOW_SEG_HDR_PPPOE, 0);
+ if (ret)
+ PMD_DRV_LOG(ERR, "%s PPPoE_IPV6 rss flow fail %d",
+ __func__, ret);
+ }
+
+ if (rss_hf & ETH_RSS_NONFRAG_IPV4_UDP) {
+ ret = ice_add_rss_cfg(hw, vsi->idx, ICE_HASH_UDP_IPV4,
+ ICE_FLOW_SEG_HDR_PPPOE, 0);
+ if (ret)
+ PMD_DRV_LOG(ERR, "%s PPPoE_IPV4_UDP rss flow fail %d",
+ __func__, ret);
+ }
+
+ if (rss_hf & ETH_RSS_NONFRAG_IPV6_UDP) {
+ ret = ice_add_rss_cfg(hw, vsi->idx, ICE_HASH_UDP_IPV6,
+ ICE_FLOW_SEG_HDR_PPPOE, 0);
+ if (ret)
+ PMD_DRV_LOG(ERR, "%s PPPoE_IPV6_UDP rss flow fail %d",
+ __func__, ret);
+ }
+
+ if (rss_hf & ETH_RSS_NONFRAG_IPV4_TCP) {
+ ret = ice_add_rss_cfg(hw, vsi->idx, ICE_HASH_TCP_IPV4,
+ ICE_FLOW_SEG_HDR_PPPOE, 0);
+ if (ret)
+ PMD_DRV_LOG(ERR, "%s PPPoE_IPV4_TCP rss flow fail %d",
+ __func__, ret);
+ }
+
+ if (rss_hf & ETH_RSS_NONFRAG_IPV6_TCP) {
+ ret = ice_add_rss_cfg(hw, vsi->idx, ICE_HASH_TCP_IPV6,
+ ICE_FLOW_SEG_HDR_PPPOE, 0);
+ if (ret)
+ PMD_DRV_LOG(ERR, "%s PPPoE_IPV6_TCP rss flow fail %d",
+ __func__, ret);
+ }
+
+ if (rss_hf & ETH_RSS_NONFRAG_IPV4_SCTP) {
+ ret = ice_add_rss_cfg(hw, vsi->idx, ICE_HASH_SCTP_IPV4,
+ ICE_FLOW_SEG_HDR_PPPOE, 0);
+ if (ret)
+ PMD_DRV_LOG(ERR, "%s PPPoE_IPV4_SCTP rss flow fail %d",
+ __func__, ret);
+ }
+
+ if (rss_hf & ETH_RSS_NONFRAG_IPV6_SCTP) {
+ if (ret)
+ PMD_DRV_LOG(ERR, "%s GTPU_IPV6_SCTP rss flow fail %d",
+ __func__, ret);
+
+ ret = ice_add_rss_cfg(hw, vsi->idx, ICE_HASH_SCTP_IPV6,
+ ICE_FLOW_SEG_HDR_PPPOE, 0);
+ if (ret)
+ PMD_DRV_LOG(ERR, "%s PPPoE_IPV6_SCTP rss flow fail %d",
+ __func__, ret);
+ }
}
static int ice_init_rss(struct ice_pf *pf)