[PATCH net-next 00/14] mlxsw: Add support for IP-in-IP with IPv6 underlay

[PATCH net-next 00/14] mlxsw: Add support for IP-in-IP with IPv6 underlay

[Ido Schimmel]

From: Ido Schimmel <idosch@nvidia.com>

Currently, mlxsw only supports IP-in-IP with IPv4 underlay. Traffic
routed through 'gre' netdevs is encapsulated with IPv4 and GRE headers.
Similarly, incoming IPv4 GRE packets are decapsulated and routed in the
overlay VRF (which can be the same as the underlay VRF).

This patchset adds support for IPv6 underlay using the 'ip6gre' netdev.
Due to architectural differences between Spectrum-1 and later ASICs,
this functionality is only supported on Spectrum-2 onwards (the software
data path is used for Spectrum-1).

Patchset overview:

Patches #1-#5 are preparations.

Patches #6-#9 add and extend required device registers.

Patches #10-#14 gradually add IPv6 underlay support.

A follow-up patchset will add net/forwarding/ selftests.

Amit Cohen (14):
  mlxsw: spectrum_router: Create common function for
    fib_entry_type_unset() code
  mlxsw: spectrum_ipip: Pass IP tunnel parameters by reference and as
    'const'
  mlxsw: spectrum_router: Fix arguments alignment
  mlxsw: spectrum_ipip: Create common function for
    mlxsw_sp_ipip_ol_netdev_change_gre()
  mlxsw: Take tunnel's type into account when searching underlay device
  mlxsw: reg: Add Router IP version Six Register
  mlxsw: reg: Add support for rtdp_ipip6_pack()
  mlxsw: reg: Add support for ratr_ipip6_entry_pack()
  mlxsw: reg: Add support for ritr_loopback_ipip6_pack()
  mlxsw: Create separate ipip_ops_arr for different ASICs
  mlxsw: spectrum_ipip: Add mlxsw_sp_ipip_gre6_ops
  mlxsw: Add IPV6_ADDRESS kvdl entry type
  mlxsw: spectrum_router: Increase parsing depth for IPv6 decapsulation
  mlxsw: Add support for IP-in-IP with IPv6 underlay for Spectrum-2 and
    above

 drivers/net/ethernet/mellanox/mlxsw/reg.h     |  86 +++-
 .../net/ethernet/mellanox/mlxsw/spectrum.h    |   2 +
 .../ethernet/mellanox/mlxsw/spectrum2_kvdl.c  |   1 +
 .../ethernet/mellanox/mlxsw/spectrum_ipip.c   | 432 ++++++++++++++++--
 .../ethernet/mellanox/mlxsw/spectrum_ipip.h   |  27 +-
 .../ethernet/mellanox/mlxsw/spectrum_router.c | 186 ++++++--
 .../ethernet/mellanox/mlxsw/spectrum_router.h |   2 +
 7 files changed, 641 insertions(+), 95 deletions(-)

-- 
2.31.1

[PATCH net-next 01/14] mlxsw: spectrum_router: Create common function for fib_entry_type_unset() code

[Ido Schimmel]

From: Amit Cohen <amcohen@nvidia.com>

mlxsw_sp_fib4_entry_type_unset() is not specific for IPv4 FIB entry,
move the code to mlxsw_sp_fib_entry_type_unset(), and call this function
from mlxsw_sp_fib4_entry_type_unset() so then it will be used for IPv6
also.

Signed-off-by: Amit Cohen <amcohen@nvidia.com>
Reviewed-by: Petr Machata <petrm@nvidia.com>
Signed-off-by: Ido Schimmel <idosch@nvidia.com>
---
 .../net/ethernet/mellanox/mlxsw/spectrum_router.c   | 13 ++++++++++---
 1 file changed, 10 insertions(+), 3 deletions(-)

diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c
index 331d26c1181e..0454f3bc27d3 100644
--- a/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c
+++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c
@@ -6069,8 +6069,8 @@ mlxsw_sp_fib4_entry_type_set(struct mlxsw_sp *mlxsw_sp,
 }
 
 static void
-mlxsw_sp_fib4_entry_type_unset(struct mlxsw_sp *mlxsw_sp,
-			       struct mlxsw_sp_fib_entry *fib_entry)
+mlxsw_sp_fib_entry_type_unset(struct mlxsw_sp *mlxsw_sp,
+			      struct mlxsw_sp_fib_entry *fib_entry)
 {
 	switch (fib_entry->type) {
 	case MLXSW_SP_FIB_ENTRY_TYPE_IPIP_DECAP:
@@ -6081,6 +6081,13 @@ mlxsw_sp_fib4_entry_type_unset(struct mlxsw_sp *mlxsw_sp,
 	}
 }
 
+static void
+mlxsw_sp_fib4_entry_type_unset(struct mlxsw_sp *mlxsw_sp,
+			       struct mlxsw_sp_fib4_entry *fib4_entry)
+{
+	mlxsw_sp_fib_entry_type_unset(mlxsw_sp, &fib4_entry->common);
+}
+
 static struct mlxsw_sp_fib4_entry *
 mlxsw_sp_fib4_entry_create(struct mlxsw_sp *mlxsw_sp,
 			   struct mlxsw_sp_fib_node *fib_node,
@@ -6141,7 +6148,7 @@ static void mlxsw_sp_fib4_entry_destroy(struct mlxsw_sp *mlxsw_sp,
 	struct mlxsw_sp_fib_node *fib_node = fib4_entry->common.fib_node;
 
 	fib_info_put(fib4_entry->fi);
-	mlxsw_sp_fib4_entry_type_unset(mlxsw_sp, &fib4_entry->common);
+	mlxsw_sp_fib4_entry_type_unset(mlxsw_sp, fib4_entry);
 	mlxsw_sp_nexthop_group_vr_unlink(fib4_entry->common.nh_group,
 					 fib_node->fib);
 	mlxsw_sp_nexthop4_group_put(mlxsw_sp, &fib4_entry->common);
-- 
2.31.1

[PATCH net-next 02/14] mlxsw: spectrum_ipip: Pass IP tunnel parameters by reference and as 'const'

[Ido Schimmel]

From: Amit Cohen <amcohen@nvidia.com>

Currently, there are several functions that handle 'struct ip_tunnel_parm'
and 'struct __ip6_tnl_parm'.

Change the mentioned functions to get IP tunnel parameters by reference
and as 'const'.

Signed-off-by: Amit Cohen <amcohen@nvidia.com>
Reviewed-by: Petr Machata <petrm@nvidia.com>
Signed-off-by: Ido Schimmel <idosch@nvidia.com>
---
 .../ethernet/mellanox/mlxsw/spectrum_ipip.c   | 66 +++++++++----------
 1 file changed, 33 insertions(+), 33 deletions(-)

diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_ipip.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_ipip.c
index 5facabd86882..9aeb6fe76c06 100644
--- a/drivers/net/ethernet/mellanox/mlxsw/spectrum_ipip.c
+++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_ipip.c
@@ -24,50 +24,50 @@ mlxsw_sp_ipip_netdev_parms6(const struct net_device *ol_dev)
 	return tun->parms;
 }
 
-static bool mlxsw_sp_ipip_parms4_has_ikey(struct ip_tunnel_parm parms)
+static bool mlxsw_sp_ipip_parms4_has_ikey(const struct ip_tunnel_parm *parms)
 {
-	return !!(parms.i_flags & TUNNEL_KEY);
+	return !!(parms->i_flags & TUNNEL_KEY);
 }
 
-static bool mlxsw_sp_ipip_parms4_has_okey(struct ip_tunnel_parm parms)
+static bool mlxsw_sp_ipip_parms4_has_okey(const struct ip_tunnel_parm *parms)
 {
-	return !!(parms.o_flags & TUNNEL_KEY);
+	return !!(parms->o_flags & TUNNEL_KEY);
 }
 
-static u32 mlxsw_sp_ipip_parms4_ikey(struct ip_tunnel_parm parms)
+static u32 mlxsw_sp_ipip_parms4_ikey(const struct ip_tunnel_parm *parms)
 {
 	return mlxsw_sp_ipip_parms4_has_ikey(parms) ?
-		be32_to_cpu(parms.i_key) : 0;
+		be32_to_cpu(parms->i_key) : 0;
 }
 
-static u32 mlxsw_sp_ipip_parms4_okey(struct ip_tunnel_parm parms)
+static u32 mlxsw_sp_ipip_parms4_okey(const struct ip_tunnel_parm *parms)
 {
 	return mlxsw_sp_ipip_parms4_has_okey(parms) ?
-		be32_to_cpu(parms.o_key) : 0;
+		be32_to_cpu(parms->o_key) : 0;
 }
 
 static union mlxsw_sp_l3addr
-mlxsw_sp_ipip_parms4_saddr(struct ip_tunnel_parm parms)
+mlxsw_sp_ipip_parms4_saddr(const struct ip_tunnel_parm *parms)
 {
-	return (union mlxsw_sp_l3addr) { .addr4 = parms.iph.saddr };
+	return (union mlxsw_sp_l3addr) { .addr4 = parms->iph.saddr };
 }
 
 static union mlxsw_sp_l3addr
-mlxsw_sp_ipip_parms6_saddr(struct __ip6_tnl_parm parms)
+mlxsw_sp_ipip_parms6_saddr(const struct __ip6_tnl_parm *parms)
 {
-	return (union mlxsw_sp_l3addr) { .addr6 = parms.laddr };
+	return (union mlxsw_sp_l3addr) { .addr6 = parms->laddr };
 }
 
 static union mlxsw_sp_l3addr
-mlxsw_sp_ipip_parms4_daddr(struct ip_tunnel_parm parms)
+mlxsw_sp_ipip_parms4_daddr(const struct ip_tunnel_parm *parms)
 {
-	return (union mlxsw_sp_l3addr) { .addr4 = parms.iph.daddr };
+	return (union mlxsw_sp_l3addr) { .addr4 = parms->iph.daddr };
 }
 
 static union mlxsw_sp_l3addr
-mlxsw_sp_ipip_parms6_daddr(struct __ip6_tnl_parm parms)
+mlxsw_sp_ipip_parms6_daddr(const struct __ip6_tnl_parm *parms)
 {
-	return (union mlxsw_sp_l3addr) { .addr6 = parms.raddr };
+	return (union mlxsw_sp_l3addr) { .addr6 = parms->raddr };
 }
 
 union mlxsw_sp_l3addr
@@ -80,10 +80,10 @@ mlxsw_sp_ipip_netdev_saddr(enum mlxsw_sp_l3proto proto,
 	switch (proto) {
 	case MLXSW_SP_L3_PROTO_IPV4:
 		parms4 = mlxsw_sp_ipip_netdev_parms4(ol_dev);
-		return mlxsw_sp_ipip_parms4_saddr(parms4);
+		return mlxsw_sp_ipip_parms4_saddr(&parms4);
 	case MLXSW_SP_L3_PROTO_IPV6:
 		parms6 = mlxsw_sp_ipip_netdev_parms6(ol_dev);
-		return mlxsw_sp_ipip_parms6_saddr(parms6);
+		return mlxsw_sp_ipip_parms6_saddr(&parms6);
 	}
 
 	WARN_ON(1);
@@ -95,7 +95,7 @@ static __be32 mlxsw_sp_ipip_netdev_daddr4(const struct net_device *ol_dev)
 
 	struct ip_tunnel_parm parms4 = mlxsw_sp_ipip_netdev_parms4(ol_dev);
 
-	return mlxsw_sp_ipip_parms4_daddr(parms4).addr4;
+	return mlxsw_sp_ipip_parms4_daddr(&parms4).addr4;
 }
 
 static union mlxsw_sp_l3addr
@@ -108,10 +108,10 @@ mlxsw_sp_ipip_netdev_daddr(enum mlxsw_sp_l3proto proto,
 	switch (proto) {
 	case MLXSW_SP_L3_PROTO_IPV4:
 		parms4 = mlxsw_sp_ipip_netdev_parms4(ol_dev);
-		return mlxsw_sp_ipip_parms4_daddr(parms4);
+		return mlxsw_sp_ipip_parms4_daddr(&parms4);
 	case MLXSW_SP_L3_PROTO_IPV6:
 		parms6 = mlxsw_sp_ipip_netdev_parms6(ol_dev);
-		return mlxsw_sp_ipip_parms6_daddr(parms6);
+		return mlxsw_sp_ipip_parms6_daddr(&parms6);
 	}
 
 	WARN_ON(1);
@@ -158,8 +158,8 @@ mlxsw_sp_ipip_decap_config_gre4(struct mlxsw_sp *mlxsw_sp,
 	u32 ikey;
 
 	parms = mlxsw_sp_ipip_netdev_parms4(ipip_entry->ol_dev);
-	has_ikey = mlxsw_sp_ipip_parms4_has_ikey(parms);
-	ikey = mlxsw_sp_ipip_parms4_ikey(parms);
+	has_ikey = mlxsw_sp_ipip_parms4_has_ikey(&parms);
+	ikey = mlxsw_sp_ipip_parms4_ikey(&parms);
 
 	mlxsw_reg_rtdp_pack(rtdp_pl, MLXSW_REG_RTDP_TYPE_IPIP, tunnel_index);
 	mlxsw_reg_rtdp_egress_router_interface_set(rtdp_pl, ul_rif_id);
@@ -218,12 +218,12 @@ mlxsw_sp_ipip_ol_loopback_config_gre4(struct mlxsw_sp *mlxsw_sp,
 	struct ip_tunnel_parm parms = mlxsw_sp_ipip_netdev_parms4(ol_dev);
 	enum mlxsw_reg_ritr_loopback_ipip_type lb_ipipt;
 
-	lb_ipipt = mlxsw_sp_ipip_parms4_has_okey(parms) ?
+	lb_ipipt = mlxsw_sp_ipip_parms4_has_okey(&parms) ?
 		MLXSW_REG_RITR_LOOPBACK_IPIP_TYPE_IP_IN_GRE_KEY_IN_IP :
 		MLXSW_REG_RITR_LOOPBACK_IPIP_TYPE_IP_IN_GRE_IN_IP;
 	return (struct mlxsw_sp_rif_ipip_lb_config){
 		.lb_ipipt = lb_ipipt,
-		.okey = mlxsw_sp_ipip_parms4_okey(parms),
+		.okey = mlxsw_sp_ipip_parms4_okey(&parms),
 		.ul_protocol = MLXSW_SP_L3_PROTO_IPV4,
 		.saddr = mlxsw_sp_ipip_netdev_saddr(MLXSW_SP_L3_PROTO_IPV4,
 						    ol_dev),
@@ -245,10 +245,10 @@ mlxsw_sp_ipip_ol_netdev_change_gre4(struct mlxsw_sp *mlxsw_sp,
 
 	new_parms = mlxsw_sp_ipip_netdev_parms4(ipip_entry->ol_dev);
 
-	new_saddr = mlxsw_sp_ipip_parms4_saddr(new_parms);
-	old_saddr = mlxsw_sp_ipip_parms4_saddr(ipip_entry->parms4);
-	new_daddr = mlxsw_sp_ipip_parms4_daddr(new_parms);
-	old_daddr = mlxsw_sp_ipip_parms4_daddr(ipip_entry->parms4);
+	new_saddr = mlxsw_sp_ipip_parms4_saddr(&new_parms);
+	old_saddr = mlxsw_sp_ipip_parms4_saddr(&ipip_entry->parms4);
+	new_daddr = mlxsw_sp_ipip_parms4_daddr(&new_parms);
+	old_daddr = mlxsw_sp_ipip_parms4_daddr(&ipip_entry->parms4);
 
 	if (!mlxsw_sp_l3addr_eq(&new_saddr, &old_saddr)) {
 		u16 ul_tb_id = mlxsw_sp_ipip_dev_ul_tb_id(ipip_entry->ol_dev);
@@ -265,14 +265,14 @@ mlxsw_sp_ipip_ol_netdev_change_gre4(struct mlxsw_sp *mlxsw_sp,
 		}
 
 		update_tunnel = true;
-	} else if ((mlxsw_sp_ipip_parms4_okey(ipip_entry->parms4) !=
-		    mlxsw_sp_ipip_parms4_okey(new_parms)) ||
+	} else if ((mlxsw_sp_ipip_parms4_okey(&ipip_entry->parms4) !=
+		    mlxsw_sp_ipip_parms4_okey(&new_parms)) ||
 		   ipip_entry->parms4.link != new_parms.link) {
 		update_tunnel = true;
 	} else if (!mlxsw_sp_l3addr_eq(&new_daddr, &old_daddr)) {
 		update_nhs = true;
-	} else if (mlxsw_sp_ipip_parms4_ikey(ipip_entry->parms4) !=
-		   mlxsw_sp_ipip_parms4_ikey(new_parms)) {
+	} else if (mlxsw_sp_ipip_parms4_ikey(&ipip_entry->parms4) !=
+		   mlxsw_sp_ipip_parms4_ikey(&new_parms)) {
 		update_decap = true;
 	}
 
-- 
2.31.1

[PATCH net-next 03/14] mlxsw: spectrum_router: Fix arguments alignment

[Ido Schimmel]

From: Amit Cohen <amcohen@nvidia.com>

Suppress the following checkpatch.pl check [1] by adding a variable to
store the IP-in-IP options. Noticed while adding equivalent IPv6 code in
subsequent patches.

[1]
CHECK: Alignment should match open parenthesis
+               mlxsw_reg_ritr_loopback_ipip4_pack(ritr_pl, lb_cf.lb_ipipt,
+
+ MLXSW_REG_RITR_LOOPBACK_IPIP_OPTIONS_GRE_KEY_PRESET,

Signed-off-by: Amit Cohen <amcohen@nvidia.com>
Reviewed-by: Petr Machata <petrm@nvidia.com>
Signed-off-by: Ido Schimmel <idosch@nvidia.com>
---
 drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c | 7 +++++--
 1 file changed, 5 insertions(+), 2 deletions(-)

diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c
index 0454f3bc27d3..61f1e7d58128 100644
--- a/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c
+++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c
@@ -1536,19 +1536,22 @@ mlxsw_sp_rif_ipip_lb_op(struct mlxsw_sp_rif_ipip_lb *lb_rif, u16 ul_vr_id,
 			u16 ul_rif_id, bool enable)
 {
 	struct mlxsw_sp_rif_ipip_lb_config lb_cf = lb_rif->lb_config;
+	enum mlxsw_reg_ritr_loopback_ipip_options ipip_options;
 	struct mlxsw_sp_rif *rif = &lb_rif->common;
 	struct mlxsw_sp *mlxsw_sp = rif->mlxsw_sp;
 	char ritr_pl[MLXSW_REG_RITR_LEN];
 	u32 saddr4;
 
+	ipip_options = MLXSW_REG_RITR_LOOPBACK_IPIP_OPTIONS_GRE_KEY_PRESET;
 	switch (lb_cf.ul_protocol) {
 	case MLXSW_SP_L3_PROTO_IPV4:
 		saddr4 = be32_to_cpu(lb_cf.saddr.addr4);
 		mlxsw_reg_ritr_pack(ritr_pl, enable, MLXSW_REG_RITR_LOOPBACK_IF,
 				    rif->rif_index, rif->vr_id, rif->dev->mtu);
 		mlxsw_reg_ritr_loopback_ipip4_pack(ritr_pl, lb_cf.lb_ipipt,
-			    MLXSW_REG_RITR_LOOPBACK_IPIP_OPTIONS_GRE_KEY_PRESET,
-			    ul_vr_id, ul_rif_id, saddr4, lb_cf.okey);
+						   ipip_options, ul_vr_id,
+						   ul_rif_id, saddr4,
+						   lb_cf.okey);
 		break;
 
 	case MLXSW_SP_L3_PROTO_IPV6:
-- 
2.31.1

[PATCH net-next 04/14] mlxsw: spectrum_ipip: Create common function for mlxsw_sp_ipip_ol_netdev_change_gre()

[Ido Schimmel]

From: Amit Cohen <amcohen@nvidia.com>

The function mlxsw_sp_ipip_ol_netdev_change_gre4() contains code that
can be shared between IPv4 and IPv6.
The only difference is the way that arguments are taken from tunnel
parameters, which are different between IPv4 and IPv6.

For that, add structure 'mlxsw_sp_ipip_parms' to hold all the required
parameters for the function and save it as part of
'struct mlxsw_sp_ipip_entry' instead of the existing structure that is
not shared between IPv4 and IPv6. Add new operation as part of
'mlxsw_sp_ipip_ops' to initialize the new structure.

Then mlxsw_sp_ipip_ol_netdev_change_gre{4,6}() will prepare the new
structure and both will call the same function.

Signed-off-by: Amit Cohen <amcohen@nvidia.com>
Reviewed-by: Petr Machata <petrm@nvidia.com>
Signed-off-by: Ido Schimmel <idosch@nvidia.com>
---
 .../ethernet/mellanox/mlxsw/spectrum_ipip.c   | 69 ++++++++++++-------
 .../ethernet/mellanox/mlxsw/spectrum_ipip.h   | 16 ++++-
 .../ethernet/mellanox/mlxsw/spectrum_router.c | 11 +--
 3 files changed, 59 insertions(+), 37 deletions(-)

diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_ipip.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_ipip.c
index 9aeb6fe76c06..2164e940abba 100644
--- a/drivers/net/ethernet/mellanox/mlxsw/spectrum_ipip.c
+++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_ipip.c
@@ -125,6 +125,21 @@ bool mlxsw_sp_l3addr_is_zero(union mlxsw_sp_l3addr addr)
 	return !memcmp(&addr, &naddr, sizeof(naddr));
 }
 
+static struct mlxsw_sp_ipip_parms
+mlxsw_sp_ipip_netdev_parms_init_gre4(const struct net_device *ol_dev)
+{
+	struct ip_tunnel_parm parms = mlxsw_sp_ipip_netdev_parms4(ol_dev);
+
+	return (struct mlxsw_sp_ipip_parms) {
+		.proto = MLXSW_SP_L3_PROTO_IPV4,
+		.saddr = mlxsw_sp_ipip_parms4_saddr(&parms),
+		.daddr = mlxsw_sp_ipip_parms4_daddr(&parms),
+		.link = parms.link,
+		.ikey = mlxsw_sp_ipip_parms4_ikey(&parms),
+		.okey = mlxsw_sp_ipip_parms4_okey(&parms),
+	};
+}
+
 static int
 mlxsw_sp_ipip_nexthop_update_gre4(struct mlxsw_sp *mlxsw_sp, u32 adj_index,
 				  struct mlxsw_sp_ipip_entry *ipip_entry,
@@ -231,48 +246,39 @@ mlxsw_sp_ipip_ol_loopback_config_gre4(struct mlxsw_sp *mlxsw_sp,
 }
 
 static int
-mlxsw_sp_ipip_ol_netdev_change_gre4(struct mlxsw_sp *mlxsw_sp,
-				    struct mlxsw_sp_ipip_entry *ipip_entry,
-				    struct netlink_ext_ack *extack)
+mlxsw_sp_ipip_ol_netdev_change_gre(struct mlxsw_sp *mlxsw_sp,
+				   struct mlxsw_sp_ipip_entry *ipip_entry,
+				   const struct mlxsw_sp_ipip_parms *new_parms,
+				   struct netlink_ext_ack *extack)
 {
-	union mlxsw_sp_l3addr old_saddr, new_saddr;
-	union mlxsw_sp_l3addr old_daddr, new_daddr;
-	struct ip_tunnel_parm new_parms;
+	const struct mlxsw_sp_ipip_parms *old_parms = &ipip_entry->parms;
 	bool update_tunnel = false;
 	bool update_decap = false;
 	bool update_nhs = false;
 	int err = 0;
 
-	new_parms = mlxsw_sp_ipip_netdev_parms4(ipip_entry->ol_dev);
-
-	new_saddr = mlxsw_sp_ipip_parms4_saddr(&new_parms);
-	old_saddr = mlxsw_sp_ipip_parms4_saddr(&ipip_entry->parms4);
-	new_daddr = mlxsw_sp_ipip_parms4_daddr(&new_parms);
-	old_daddr = mlxsw_sp_ipip_parms4_daddr(&ipip_entry->parms4);
-
-	if (!mlxsw_sp_l3addr_eq(&new_saddr, &old_saddr)) {
+	if (!mlxsw_sp_l3addr_eq(&new_parms->saddr, &old_parms->saddr)) {
 		u16 ul_tb_id = mlxsw_sp_ipip_dev_ul_tb_id(ipip_entry->ol_dev);
 
 		/* Since the local address has changed, if there is another
 		 * tunnel with a matching saddr, both need to be demoted.
 		 */
 		if (mlxsw_sp_ipip_demote_tunnel_by_saddr(mlxsw_sp,
-							 MLXSW_SP_L3_PROTO_IPV4,
-							 new_saddr, ul_tb_id,
+							 new_parms->proto,
+							 new_parms->saddr,
+							 ul_tb_id,
 							 ipip_entry)) {
 			mlxsw_sp_ipip_entry_demote_tunnel(mlxsw_sp, ipip_entry);
 			return 0;
 		}
 
 		update_tunnel = true;
-	} else if ((mlxsw_sp_ipip_parms4_okey(&ipip_entry->parms4) !=
-		    mlxsw_sp_ipip_parms4_okey(&new_parms)) ||
-		   ipip_entry->parms4.link != new_parms.link) {
+	} else if (old_parms->okey != new_parms->okey ||
+		   old_parms->link != new_parms->link) {
 		update_tunnel = true;
-	} else if (!mlxsw_sp_l3addr_eq(&new_daddr, &old_daddr)) {
+	} else if (!mlxsw_sp_l3addr_eq(&new_parms->daddr, &old_parms->daddr)) {
 		update_nhs = true;
-	} else if (mlxsw_sp_ipip_parms4_ikey(&ipip_entry->parms4) !=
-		   mlxsw_sp_ipip_parms4_ikey(&new_parms)) {
+	} else if (old_parms->ikey != new_parms->ikey) {
 		update_decap = true;
 	}
 
@@ -288,14 +294,29 @@ mlxsw_sp_ipip_ol_netdev_change_gre4(struct mlxsw_sp *mlxsw_sp,
 		err = __mlxsw_sp_ipip_entry_update_tunnel(mlxsw_sp, ipip_entry,
 							  false, false, false,
 							  extack);
+	if (err)
+		return err;
+
+	ipip_entry->parms = *new_parms;
+	return 0;
+}
+
+static int
+mlxsw_sp_ipip_ol_netdev_change_gre4(struct mlxsw_sp *mlxsw_sp,
+				    struct mlxsw_sp_ipip_entry *ipip_entry,
+				    struct netlink_ext_ack *extack)
+{
+	struct mlxsw_sp_ipip_parms new_parms;
 
-	ipip_entry->parms4 = new_parms;
-	return err;
+	new_parms = mlxsw_sp_ipip_netdev_parms_init_gre4(ipip_entry->ol_dev);
+	return mlxsw_sp_ipip_ol_netdev_change_gre(mlxsw_sp, ipip_entry,
+						  &new_parms, extack);
 }
 
 static const struct mlxsw_sp_ipip_ops mlxsw_sp_ipip_gre4_ops = {
 	.dev_type = ARPHRD_IPGRE,
 	.ul_proto = MLXSW_SP_L3_PROTO_IPV4,
+	.parms_init = mlxsw_sp_ipip_netdev_parms_init_gre4,
 	.nexthop_update = mlxsw_sp_ipip_nexthop_update_gre4,
 	.decap_config = mlxsw_sp_ipip_decap_config_gre4,
 	.can_offload = mlxsw_sp_ipip_can_offload_gre4,
diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_ipip.h b/drivers/net/ethernet/mellanox/mlxsw/spectrum_ipip.h
index f0837b42d1d6..bc6866d1c070 100644
--- a/drivers/net/ethernet/mellanox/mlxsw/spectrum_ipip.h
+++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_ipip.h
@@ -24,21 +24,31 @@ enum mlxsw_sp_ipip_type {
 	MLXSW_SP_IPIP_TYPE_MAX,
 };
 
+struct mlxsw_sp_ipip_parms {
+	enum mlxsw_sp_l3proto proto;
+	union mlxsw_sp_l3addr saddr;
+	union mlxsw_sp_l3addr daddr;
+	int link;
+	u32 ikey;
+	u32 okey;
+};
+
 struct mlxsw_sp_ipip_entry {
 	enum mlxsw_sp_ipip_type ipipt;
 	struct net_device *ol_dev; /* Overlay. */
 	struct mlxsw_sp_rif_ipip_lb *ol_lb;
 	struct mlxsw_sp_fib_entry *decap_fib_entry;
 	struct list_head ipip_list_node;
-	union {
-		struct ip_tunnel_parm parms4;
-	};
+	struct mlxsw_sp_ipip_parms parms;
 };
 
 struct mlxsw_sp_ipip_ops {
 	int dev_type;
 	enum mlxsw_sp_l3proto ul_proto; /* Underlay. */
 
+	struct mlxsw_sp_ipip_parms
+	(*parms_init)(const struct net_device *ol_dev);
+
 	int (*nexthop_update)(struct mlxsw_sp *mlxsw_sp, u32 adj_index,
 			      struct mlxsw_sp_ipip_entry *ipip_entry,
 			      bool force, char *ratr_pl);
diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c
index 61f1e7d58128..b79662048ef7 100644
--- a/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c
+++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c
@@ -1131,16 +1131,7 @@ mlxsw_sp_ipip_entry_alloc(struct mlxsw_sp *mlxsw_sp,
 
 	ipip_entry->ipipt = ipipt;
 	ipip_entry->ol_dev = ol_dev;
-
-	switch (ipip_ops->ul_proto) {
-	case MLXSW_SP_L3_PROTO_IPV4:
-		ipip_entry->parms4 = mlxsw_sp_ipip_netdev_parms4(ol_dev);
-		break;
-	case MLXSW_SP_L3_PROTO_IPV6:
-		WARN_ON(1);
-		break;
-	}
-
+	ipip_entry->parms = ipip_ops->parms_init(ol_dev);
 	return ipip_entry;
 
 err_ol_ipip_lb_create:
-- 
2.31.1

[PATCH net-next 05/14] mlxsw: Take tunnel's type into account when searching underlay device

[Ido Schimmel]

From: Amit Cohen <amcohen@nvidia.com>

The function __mlxsw_sp_ipip_netdev_ul_dev_get() returns the underlay
device that corresponds to the overlay device that it gets.
Currently, this function assumes that the tunnel is IPv4 GRE, because it
is the only one that is supported by mlxsw driver.

This assumption will no longer be correct when IPv6 GRE support is added,
resulting in wrong underlay device being returned.
Instead, check 'ol_dev->type' and return the underlay device accordingly.

Move the function to spectrum_ipip.c because spectrum_router.c should not
be aware to tunnel type.

Signed-off-by: Amit Cohen <amcohen@nvidia.com>
Reviewed-by: Petr Machata <petrm@nvidia.com>
Signed-off-by: Ido Schimmel <idosch@nvidia.com>
---
 .../net/ethernet/mellanox/mlxsw/spectrum_ipip.c | 15 +++++++++++++++
 .../ethernet/mellanox/mlxsw/spectrum_router.c   | 17 ++++-------------
 .../ethernet/mellanox/mlxsw/spectrum_router.h   |  2 ++
 3 files changed, 21 insertions(+), 13 deletions(-)

diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_ipip.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_ipip.c
index 2164e940abba..3c07e3a70fb6 100644
--- a/drivers/net/ethernet/mellanox/mlxsw/spectrum_ipip.c
+++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_ipip.c
@@ -384,3 +384,18 @@ int mlxsw_sp_ipip_ecn_decap_init(struct mlxsw_sp *mlxsw_sp)
 
 	return 0;
 }
+
+struct net_device *
+mlxsw_sp_ipip_netdev_ul_dev_get(const struct net_device *ol_dev)
+{
+	struct net *net = dev_net(ol_dev);
+	struct ip_tunnel *tun4;
+
+	switch (ol_dev->type) {
+	case ARPHRD_IPGRE:
+		tun4 = netdev_priv(ol_dev);
+		return dev_get_by_index_rcu(net, tun4->parms.link);
+	default:
+		return NULL;
+	}
+}
diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c
index b79662048ef7..2582404043af 100644
--- a/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c
+++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c
@@ -1055,22 +1055,13 @@ static void mlxsw_sp_vrs_fini(struct mlxsw_sp *mlxsw_sp)
 	kfree(mlxsw_sp->router->vrs);
 }
 
-static struct net_device *
-__mlxsw_sp_ipip_netdev_ul_dev_get(const struct net_device *ol_dev)
-{
-	struct ip_tunnel *tun = netdev_priv(ol_dev);
-	struct net *net = dev_net(ol_dev);
-
-	return dev_get_by_index_rcu(net, tun->parms.link);
-}
-
 u32 mlxsw_sp_ipip_dev_ul_tb_id(const struct net_device *ol_dev)
 {
 	struct net_device *d;
 	u32 tb_id;
 
 	rcu_read_lock();
-	d = __mlxsw_sp_ipip_netdev_ul_dev_get(ol_dev);
+	d = mlxsw_sp_ipip_netdev_ul_dev_get(ol_dev);
 	if (d)
 		tb_id = l3mdev_fib_table(d) ? : RT_TABLE_MAIN;
 	else
@@ -1441,7 +1432,7 @@ mlxsw_sp_ipip_entry_find_by_ul_dev(const struct mlxsw_sp *mlxsw_sp,
 		struct net_device *ipip_ul_dev;
 
 		rcu_read_lock();
-		ipip_ul_dev = __mlxsw_sp_ipip_netdev_ul_dev_get(ol_dev);
+		ipip_ul_dev = mlxsw_sp_ipip_netdev_ul_dev_get(ol_dev);
 		rcu_read_unlock();
 
 		if (ipip_ul_dev == ul_dev)
@@ -1821,7 +1812,7 @@ static void mlxsw_sp_ipip_demote_tunnel_by_ul_netdev(struct mlxsw_sp *mlxsw_sp,
 		struct net_device *ipip_ul_dev;
 
 		rcu_read_lock();
-		ipip_ul_dev = __mlxsw_sp_ipip_netdev_ul_dev_get(ol_dev);
+		ipip_ul_dev = mlxsw_sp_ipip_netdev_ul_dev_get(ol_dev);
 		rcu_read_unlock();
 		if (ipip_ul_dev == ul_dev)
 			mlxsw_sp_ipip_entry_demote_tunnel(mlxsw_sp, ipip_entry);
@@ -4146,7 +4137,7 @@ static bool mlxsw_sp_ipip_netdev_ul_up(struct net_device *ol_dev)
 	bool is_up;
 
 	rcu_read_lock();
-	ul_dev = __mlxsw_sp_ipip_netdev_ul_dev_get(ol_dev);
+	ul_dev = mlxsw_sp_ipip_netdev_ul_dev_get(ol_dev);
 	is_up = ul_dev ? (ul_dev->flags & IFF_UP) : true;
 	rcu_read_unlock();
 
diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.h b/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.h
index cc32d25c3bb2..1d0d28f8ff05 100644
--- a/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.h
+++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.h
@@ -226,6 +226,8 @@ static inline bool mlxsw_sp_l3addr_eq(const union mlxsw_sp_l3addr *addr1,
 
 int mlxsw_sp_ipip_ecn_encap_init(struct mlxsw_sp *mlxsw_sp);
 int mlxsw_sp_ipip_ecn_decap_init(struct mlxsw_sp *mlxsw_sp);
+struct net_device *
+mlxsw_sp_ipip_netdev_ul_dev_get(const struct net_device *ol_dev);
 
 extern const struct mlxsw_sp_router_ll_ops mlxsw_sp_router_ll_xm_ops;
 
-- 
2.31.1

[PATCH net-next 06/14] mlxsw: reg: Add Router IP version Six Register

[Ido Schimmel]

From: Amit Cohen <amcohen@nvidia.com>

The RIPS register is used to store IPv6 addresses for use by the NVE and
IP-in-IP.

For IPv6 underlay support, RATR register needs to hold a pointer to the
remote IPv6 address for encapsulation and RTDP register needs to hold a
pointer to the local IPv6 address for decapsulation check.

Add the required register for saving IPv6 addresses.

Signed-off-by: Amit Cohen <amcohen@nvidia.com>
Reviewed-by: Petr Machata <petrm@nvidia.com>
Signed-off-by: Ido Schimmel <idosch@nvidia.com>
---
 drivers/net/ethernet/mellanox/mlxsw/reg.h | 32 +++++++++++++++++++++++
 1 file changed, 32 insertions(+)

diff --git a/drivers/net/ethernet/mellanox/mlxsw/reg.h b/drivers/net/ethernet/mellanox/mlxsw/reg.h
index 6c96e124e4c8..22f5c26fc0e7 100644
--- a/drivers/net/ethernet/mellanox/mlxsw/reg.h
+++ b/drivers/net/ethernet/mellanox/mlxsw/reg.h
@@ -8200,6 +8200,37 @@ mlxsw_reg_rtdp_ipip4_pack(char *payload, u16 irif,
 	mlxsw_reg_rtdp_ipip_expected_gre_key_set(payload, expected_gre_key);
 }
 
+/* RIPS - Router IP version Six Register
+ * -------------------------------------
+ * The RIPS register is used to store IPv6 addresses for use by the NVE and
+ * IPinIP
+ */
+#define MLXSW_REG_RIPS_ID 0x8021
+#define MLXSW_REG_RIPS_LEN 0x14
+
+MLXSW_REG_DEFINE(rips, MLXSW_REG_RIPS_ID, MLXSW_REG_RIPS_LEN);
+
+/* reg_rips_index
+ * Index to IPv6 address.
+ * For Spectrum, the index is to the KVD linear.
+ * Access: Index
+ */
+MLXSW_ITEM32(reg, rips, index, 0x00, 0, 24);
+
+/* reg_rips_ipv6
+ * IPv6 address
+ * Access: RW
+ */
+MLXSW_ITEM_BUF(reg, rips, ipv6, 0x04, 16);
+
+static inline void mlxsw_reg_rips_pack(char *payload, u32 index,
+				       const struct in6_addr *ipv6)
+{
+	MLXSW_REG_ZERO(rips, payload);
+	mlxsw_reg_rips_index_set(payload, index);
+	mlxsw_reg_rips_ipv6_memcpy_to(payload, (const char *)ipv6);
+}
+
 /* RATRAD - Router Adjacency Table Activity Dump Register
  * ------------------------------------------------------
  * The RATRAD register is used to dump and optionally clear activity bits of
@@ -12281,6 +12312,7 @@ static const struct mlxsw_reg_info *mlxsw_reg_infos[] = {
 	MLXSW_REG(rtar),
 	MLXSW_REG(ratr),
 	MLXSW_REG(rtdp),
+	MLXSW_REG(rips),
 	MLXSW_REG(ratrad),
 	MLXSW_REG(rdpm),
 	MLXSW_REG(ricnt),
-- 
2.31.1

[PATCH net-next 07/14] mlxsw: reg: Add support for rtdp_ipip6_pack()

[Ido Schimmel]

From: Amit Cohen <amcohen@nvidia.com>

The RTDP register is used for configuring the tunnel decapsulation
properties of NVE and IP-in-IP.

Linux tunnels verify packets before decapsulation based on the packet's
source IP, which must match tunnel remote IP.
RTDP is used to configure decapsulation so that it filters out packets that
are not IPv6 or have the wrong source IP or wrong GRE key.

For IP-in-IP entry, source IPv4 is saved as part of this register and
source IPv6 is saved by RIPS register and RTDP saves pointer to it.

Create common function for configuring both IPv4 and IPv6 and add
dedicated functions for each protocol.

Signed-off-by: Amit Cohen <amcohen@nvidia.com>
Reviewed-by: Petr Machata <petrm@nvidia.com>
Signed-off-by: Ido Schimmel <idosch@nvidia.com>
---
 drivers/net/ethernet/mellanox/mlxsw/reg.h | 31 +++++++++++++++++++----
 1 file changed, 26 insertions(+), 5 deletions(-)

diff --git a/drivers/net/ethernet/mellanox/mlxsw/reg.h b/drivers/net/ethernet/mellanox/mlxsw/reg.h
index 22f5c26fc0e7..f9f614e3c57c 100644
--- a/drivers/net/ethernet/mellanox/mlxsw/reg.h
+++ b/drivers/net/ethernet/mellanox/mlxsw/reg.h
@@ -8187,19 +8187,40 @@ static inline void mlxsw_reg_rtdp_pack(char *payload,
 }
 
 static inline void
-mlxsw_reg_rtdp_ipip4_pack(char *payload, u16 irif,
-			  enum mlxsw_reg_rtdp_ipip_sip_check sip_check,
-			  unsigned int type_check, bool gre_key_check,
-			  u32 ipv4_usip, u32 expected_gre_key)
+mlxsw_reg_rtdp_ipip_pack(char *payload, u16 irif,
+			 enum mlxsw_reg_rtdp_ipip_sip_check sip_check,
+			 unsigned int type_check, bool gre_key_check,
+			 u32 expected_gre_key)
 {
 	mlxsw_reg_rtdp_ipip_irif_set(payload, irif);
 	mlxsw_reg_rtdp_ipip_sip_check_set(payload, sip_check);
 	mlxsw_reg_rtdp_ipip_type_check_set(payload, type_check);
 	mlxsw_reg_rtdp_ipip_gre_key_check_set(payload, gre_key_check);
-	mlxsw_reg_rtdp_ipip_ipv4_usip_set(payload, ipv4_usip);
 	mlxsw_reg_rtdp_ipip_expected_gre_key_set(payload, expected_gre_key);
 }
 
+static inline void
+mlxsw_reg_rtdp_ipip4_pack(char *payload, u16 irif,
+			  enum mlxsw_reg_rtdp_ipip_sip_check sip_check,
+			  unsigned int type_check, bool gre_key_check,
+			  u32 ipv4_usip, u32 expected_gre_key)
+{
+	mlxsw_reg_rtdp_ipip_pack(payload, irif, sip_check, type_check,
+				 gre_key_check, expected_gre_key);
+	mlxsw_reg_rtdp_ipip_ipv4_usip_set(payload, ipv4_usip);
+}
+
+static inline void
+mlxsw_reg_rtdp_ipip6_pack(char *payload, u16 irif,
+			  enum mlxsw_reg_rtdp_ipip_sip_check sip_check,
+			  unsigned int type_check, bool gre_key_check,
+			  u32 ipv6_usip_ptr, u32 expected_gre_key)
+{
+	mlxsw_reg_rtdp_ipip_pack(payload, irif, sip_check, type_check,
+				 gre_key_check, expected_gre_key);
+	mlxsw_reg_rtdp_ipip_ipv6_usip_ptr_set(payload, ipv6_usip_ptr);
+}
+
 /* RIPS - Router IP version Six Register
  * -------------------------------------
  * The RIPS register is used to store IPv6 addresses for use by the NVE and
-- 
2.31.1

[PATCH net-next 08/14] mlxsw: reg: Add support for ratr_ipip6_entry_pack()

[Ido Schimmel]

From: Amit Cohen <amcohen@nvidia.com>

The RATR register is used to configure the Router Adjacency (next-hop)
Table.

For IP-in-IP entry, underlay destination IPv4 is saved as part of this
register and underlay destination IPv6 is saved by RIPS register and RATR
saves pointer to it.

Add function for setting IPv6 IP-in-IP configuration.

Signed-off-by: Amit Cohen <amcohen@nvidia.com>
Reviewed-by: Petr Machata <petrm@nvidia.com>
Signed-off-by: Ido Schimmel <idosch@nvidia.com>
---
 drivers/net/ethernet/mellanox/mlxsw/reg.h | 6 ++++++
 1 file changed, 6 insertions(+)

diff --git a/drivers/net/ethernet/mellanox/mlxsw/reg.h b/drivers/net/ethernet/mellanox/mlxsw/reg.h
index f9f614e3c57c..5cc4b1ee7e7b 100644
--- a/drivers/net/ethernet/mellanox/mlxsw/reg.h
+++ b/drivers/net/ethernet/mellanox/mlxsw/reg.h
@@ -7002,6 +7002,12 @@ static inline void mlxsw_reg_ratr_ipip4_entry_pack(char *payload, u32 ipv4_udip)
 	mlxsw_reg_ratr_ipip_ipv4_udip_set(payload, ipv4_udip);
 }
 
+static inline void mlxsw_reg_ratr_ipip6_entry_pack(char *payload, u32 ipv6_ptr)
+{
+	mlxsw_reg_ratr_ipip_type_set(payload, MLXSW_REG_RATR_IPIP_TYPE_IPV6);
+	mlxsw_reg_ratr_ipip_ipv6_ptr_set(payload, ipv6_ptr);
+}
+
 static inline void mlxsw_reg_ratr_counter_pack(char *payload, u64 counter_index,
 					       bool counter_enable)
 {
-- 
2.31.1

[PATCH net-next 09/14] mlxsw: reg: Add support for ritr_loopback_ipip6_pack()

[Ido Schimmel]

From: Amit Cohen <amcohen@nvidia.com>

The RITR register is used to configure the router interface table.

For IP-in-IP, it stores the underlay source IP address for encapsulation
and also the ingress RIF for the underlay lookup.

Add support for IPv6 IP-in-IP configuration.

Signed-off-by: Amit Cohen <amcohen@nvidia.com>
Reviewed-by: Petr Machata <petrm@nvidia.com>
Signed-off-by: Ido Schimmel <idosch@nvidia.com>
---
 drivers/net/ethernet/mellanox/mlxsw/reg.h | 17 +++++++++++++++++
 1 file changed, 17 insertions(+)

diff --git a/drivers/net/ethernet/mellanox/mlxsw/reg.h b/drivers/net/ethernet/mellanox/mlxsw/reg.h
index 5cc4b1ee7e7b..c5fad3c94fac 100644
--- a/drivers/net/ethernet/mellanox/mlxsw/reg.h
+++ b/drivers/net/ethernet/mellanox/mlxsw/reg.h
@@ -6734,6 +6734,23 @@ mlxsw_reg_ritr_loopback_ipip4_pack(char *payload,
 	mlxsw_reg_ritr_loopback_ipip_usip4_set(payload, usip);
 }
 
+static inline void
+mlxsw_reg_ritr_loopback_ipip6_pack(char *payload,
+				   enum mlxsw_reg_ritr_loopback_ipip_type ipip_type,
+				   enum mlxsw_reg_ritr_loopback_ipip_options options,
+				   u16 uvr_id, u16 underlay_rif,
+				   const struct in6_addr *usip, u32 gre_key)
+{
+	enum mlxsw_reg_ritr_loopback_protocol protocol =
+		MLXSW_REG_RITR_LOOPBACK_PROTOCOL_IPIP_IPV6;
+
+	mlxsw_reg_ritr_loopback_protocol_set(payload, protocol);
+	mlxsw_reg_ritr_loopback_ipip_common_pack(payload, ipip_type, options,
+						 uvr_id, underlay_rif, gre_key);
+	mlxsw_reg_ritr_loopback_ipip_usip6_memcpy_to(payload,
+						     (const char *)usip);
+}
+
 /* RTAR - Router TCAM Allocation Register
  * --------------------------------------
  * This register is used for allocation of regions in the TCAM table.
-- 
2.31.1

[PATCH net-next 10/14] mlxsw: Create separate ipip_ops_arr for different ASICs

[Ido Schimmel]

From: Amit Cohen <amcohen@nvidia.com>

Currently, there is support for IP-in-IP only with IPv4 underlay for all
supported Spectrum ASICs.
The next patches will add support for IPv6 underlay only for Spectrum-2
and above.

Add infrastructure for splitting IP-in-IP support between different
ASICs - create separate ipip_ops_arr and add ipips_init function to set the
right ops.

Signed-off-by: Amit Cohen <amcohen@nvidia.com>
Reviewed-by: Petr Machata <petrm@nvidia.com>
Signed-off-by: Ido Schimmel <idosch@nvidia.com>
---
 .../ethernet/mellanox/mlxsw/spectrum_ipip.c    |  6 +++++-
 .../ethernet/mellanox/mlxsw/spectrum_ipip.h    |  3 ++-
 .../ethernet/mellanox/mlxsw/spectrum_router.c  | 18 ++++++++++++++++--
 3 files changed, 23 insertions(+), 4 deletions(-)

diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_ipip.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_ipip.c
index 3c07e3a70fb6..93d183d5100d 100644
--- a/drivers/net/ethernet/mellanox/mlxsw/spectrum_ipip.c
+++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_ipip.c
@@ -324,7 +324,11 @@ static const struct mlxsw_sp_ipip_ops mlxsw_sp_ipip_gre4_ops = {
 	.ol_netdev_change = mlxsw_sp_ipip_ol_netdev_change_gre4,
 };
 
-const struct mlxsw_sp_ipip_ops *mlxsw_sp_ipip_ops_arr[] = {
+const struct mlxsw_sp_ipip_ops *mlxsw_sp1_ipip_ops_arr[] = {
+	[MLXSW_SP_IPIP_TYPE_GRE4] = &mlxsw_sp_ipip_gre4_ops,
+};
+
+const struct mlxsw_sp_ipip_ops *mlxsw_sp2_ipip_ops_arr[] = {
 	[MLXSW_SP_IPIP_TYPE_GRE4] = &mlxsw_sp_ipip_gre4_ops,
 };
 
diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_ipip.h b/drivers/net/ethernet/mellanox/mlxsw/spectrum_ipip.h
index bc6866d1c070..4426a44d546f 100644
--- a/drivers/net/ethernet/mellanox/mlxsw/spectrum_ipip.h
+++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_ipip.h
@@ -70,6 +70,7 @@ struct mlxsw_sp_ipip_ops {
 				struct netlink_ext_ack *extack);
 };
 
-extern const struct mlxsw_sp_ipip_ops *mlxsw_sp_ipip_ops_arr[];
+extern const struct mlxsw_sp_ipip_ops *mlxsw_sp1_ipip_ops_arr[];
+extern const struct mlxsw_sp_ipip_ops *mlxsw_sp2_ipip_ops_arr[];
 
 #endif /* _MLXSW_IPIP_H_*/
diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c
index 2582404043af..321f19f21d18 100644
--- a/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c
+++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c
@@ -115,6 +115,7 @@ struct mlxsw_sp_rif_ops {
 
 struct mlxsw_sp_router_ops {
 	int (*init)(struct mlxsw_sp *mlxsw_sp);
+	int (*ipips_init)(struct mlxsw_sp *mlxsw_sp);
 };
 
 static struct mlxsw_sp_rif *
@@ -9468,7 +9469,6 @@ static int mlxsw_sp_ipips_init(struct mlxsw_sp *mlxsw_sp)
 {
 	int err;
 
-	mlxsw_sp->router->ipip_ops_arr = mlxsw_sp_ipip_ops_arr;
 	INIT_LIST_HEAD(&mlxsw_sp->router->ipip_list);
 
 	err = mlxsw_sp_ipip_ecn_encap_init(mlxsw_sp);
@@ -9481,6 +9481,18 @@ static int mlxsw_sp_ipips_init(struct mlxsw_sp *mlxsw_sp)
 	return mlxsw_sp_ipip_config_tigcr(mlxsw_sp);
 }
 
+static int mlxsw_sp1_ipips_init(struct mlxsw_sp *mlxsw_sp)
+{
+	mlxsw_sp->router->ipip_ops_arr = mlxsw_sp1_ipip_ops_arr;
+	return mlxsw_sp_ipips_init(mlxsw_sp);
+}
+
+static int mlxsw_sp2_ipips_init(struct mlxsw_sp *mlxsw_sp)
+{
+	mlxsw_sp->router->ipip_ops_arr = mlxsw_sp2_ipip_ops_arr;
+	return mlxsw_sp_ipips_init(mlxsw_sp);
+}
+
 static void mlxsw_sp_ipips_fini(struct mlxsw_sp *mlxsw_sp)
 {
 	WARN_ON(!list_empty(&mlxsw_sp->router->ipip_list));
@@ -9895,6 +9907,7 @@ static int mlxsw_sp1_router_init(struct mlxsw_sp *mlxsw_sp)
 
 const struct mlxsw_sp_router_ops mlxsw_sp1_router_ops = {
 	.init = mlxsw_sp1_router_init,
+	.ipips_init = mlxsw_sp1_ipips_init,
 };
 
 static int mlxsw_sp2_router_init(struct mlxsw_sp *mlxsw_sp)
@@ -9910,6 +9923,7 @@ static int mlxsw_sp2_router_init(struct mlxsw_sp *mlxsw_sp)
 
 const struct mlxsw_sp_router_ops mlxsw_sp2_router_ops = {
 	.init = mlxsw_sp2_router_init,
+	.ipips_init = mlxsw_sp2_ipips_init,
 };
 
 int mlxsw_sp_router_init(struct mlxsw_sp *mlxsw_sp,
@@ -9955,7 +9969,7 @@ int mlxsw_sp_router_init(struct mlxsw_sp *mlxsw_sp,
 	if (err)
 		goto err_rifs_init;
 
-	err = mlxsw_sp_ipips_init(mlxsw_sp);
+	err = mlxsw_sp->router_ops->ipips_init(mlxsw_sp);
 	if (err)
 		goto err_ipips_init;
 
-- 
2.31.1

[PATCH net-next 11/14] mlxsw: spectrum_ipip: Add mlxsw_sp_ipip_gre6_ops

[Ido Schimmel]

From: Amit Cohen <amcohen@nvidia.com>

Add operations for IP-in-IP GRE6. For now the function can_offload()
returns false and the other functions warn as they should never be called.

A later patch will add dedicated operations for Spectrum-2 which will
support IPv6 underlay, so use 'mlxsw_sp1_' prefix for the new operations.

Signed-off-by: Amit Cohen <amcohen@nvidia.com>
Reviewed-by: Petr Machata <petrm@nvidia.com>
Signed-off-by: Ido Schimmel <idosch@nvidia.com>
---
 .../ethernet/mellanox/mlxsw/spectrum_ipip.c   | 65 +++++++++++++++++++
 .../ethernet/mellanox/mlxsw/spectrum_ipip.h   |  1 +
 2 files changed, 66 insertions(+)

diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_ipip.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_ipip.c
index 93d183d5100d..4bb4a3e3a2aa 100644
--- a/drivers/net/ethernet/mellanox/mlxsw/spectrum_ipip.c
+++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_ipip.c
@@ -324,12 +324,77 @@ static const struct mlxsw_sp_ipip_ops mlxsw_sp_ipip_gre4_ops = {
 	.ol_netdev_change = mlxsw_sp_ipip_ol_netdev_change_gre4,
 };
 
+static struct mlxsw_sp_ipip_parms
+mlxsw_sp1_ipip_netdev_parms_init_gre6(const struct net_device *ol_dev)
+{
+	struct mlxsw_sp_ipip_parms parms = {0};
+
+	WARN_ON_ONCE(1);
+	return parms;
+}
+
+static int
+mlxsw_sp1_ipip_nexthop_update_gre6(struct mlxsw_sp *mlxsw_sp, u32 adj_index,
+				   struct mlxsw_sp_ipip_entry *ipip_entry,
+				   bool force, char *ratr_pl)
+{
+	WARN_ON_ONCE(1);
+	return -EINVAL;
+}
+
+static int
+mlxsw_sp1_ipip_decap_config_gre6(struct mlxsw_sp *mlxsw_sp,
+				 struct mlxsw_sp_ipip_entry *ipip_entry,
+				 u32 tunnel_index)
+{
+	WARN_ON_ONCE(1);
+	return -EINVAL;
+}
+
+static bool mlxsw_sp1_ipip_can_offload_gre6(const struct mlxsw_sp *mlxsw_sp,
+					    const struct net_device *ol_dev)
+{
+	return false;
+}
+
+static struct mlxsw_sp_rif_ipip_lb_config
+mlxsw_sp1_ipip_ol_loopback_config_gre6(struct mlxsw_sp *mlxsw_sp,
+				       const struct net_device *ol_dev)
+{
+	struct mlxsw_sp_rif_ipip_lb_config config = {0};
+
+	WARN_ON_ONCE(1);
+	return config;
+}
+
+static int
+mlxsw_sp1_ipip_ol_netdev_change_gre6(struct mlxsw_sp *mlxsw_sp,
+				     struct mlxsw_sp_ipip_entry *ipip_entry,
+				     struct netlink_ext_ack *extack)
+{
+	WARN_ON_ONCE(1);
+	return -EINVAL;
+}
+
+static const struct mlxsw_sp_ipip_ops mlxsw_sp1_ipip_gre6_ops = {
+	.dev_type = ARPHRD_IP6GRE,
+	.ul_proto = MLXSW_SP_L3_PROTO_IPV6,
+	.parms_init = mlxsw_sp1_ipip_netdev_parms_init_gre6,
+	.nexthop_update = mlxsw_sp1_ipip_nexthop_update_gre6,
+	.decap_config = mlxsw_sp1_ipip_decap_config_gre6,
+	.can_offload = mlxsw_sp1_ipip_can_offload_gre6,
+	.ol_loopback_config = mlxsw_sp1_ipip_ol_loopback_config_gre6,
+	.ol_netdev_change = mlxsw_sp1_ipip_ol_netdev_change_gre6,
+};
+
 const struct mlxsw_sp_ipip_ops *mlxsw_sp1_ipip_ops_arr[] = {
 	[MLXSW_SP_IPIP_TYPE_GRE4] = &mlxsw_sp_ipip_gre4_ops,
+	[MLXSW_SP_IPIP_TYPE_GRE6] = &mlxsw_sp1_ipip_gre6_ops,
 };
 
 const struct mlxsw_sp_ipip_ops *mlxsw_sp2_ipip_ops_arr[] = {
 	[MLXSW_SP_IPIP_TYPE_GRE4] = &mlxsw_sp_ipip_gre4_ops,
+	[MLXSW_SP_IPIP_TYPE_GRE6] = &mlxsw_sp1_ipip_gre6_ops,
 };
 
 static int mlxsw_sp_ipip_ecn_encap_init_one(struct mlxsw_sp *mlxsw_sp,
diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_ipip.h b/drivers/net/ethernet/mellanox/mlxsw/spectrum_ipip.h
index 4426a44d546f..5d93337b9a2d 100644
--- a/drivers/net/ethernet/mellanox/mlxsw/spectrum_ipip.h
+++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_ipip.h
@@ -21,6 +21,7 @@ bool mlxsw_sp_l3addr_is_zero(union mlxsw_sp_l3addr addr);
 
 enum mlxsw_sp_ipip_type {
 	MLXSW_SP_IPIP_TYPE_GRE4,
+	MLXSW_SP_IPIP_TYPE_GRE6,
 	MLXSW_SP_IPIP_TYPE_MAX,
 };
 
-- 
2.31.1

[PATCH net-next 12/14] mlxsw: Add IPV6_ADDRESS kvdl entry type

[Ido Schimmel]

From: Amit Cohen <amcohen@nvidia.com>

Add support for allocating and freeing KVD entries for IPv6 addresses.

These addresses are programmed by the RIPS register and referenced by
the RATR and RTDP registers for IPv6 underlay encapsulation and
decapsulation, respectively.

Signed-off-by: Amit Cohen <amcohen@nvidia.com>
Reviewed-by: Petr Machata <petrm@nvidia.com>
Signed-off-by: Ido Schimmel <idosch@nvidia.com>
---
 drivers/net/ethernet/mellanox/mlxsw/spectrum.h       | 2 ++
 drivers/net/ethernet/mellanox/mlxsw/spectrum2_kvdl.c | 1 +
 2 files changed, 3 insertions(+)

diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum.h b/drivers/net/ethernet/mellanox/mlxsw/spectrum.h
index 83ab1ea92d31..0ebbd9b04b89 100644
--- a/drivers/net/ethernet/mellanox/mlxsw/spectrum.h
+++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum.h
@@ -747,6 +747,7 @@ enum mlxsw_sp_kvdl_entry_type {
 	MLXSW_SP_KVDL_ENTRY_TYPE_ACTSET,
 	MLXSW_SP_KVDL_ENTRY_TYPE_PBS,
 	MLXSW_SP_KVDL_ENTRY_TYPE_MCRIGR,
+	MLXSW_SP_KVDL_ENTRY_TYPE_IPV6_ADDRESS,
 	MLXSW_SP_KVDL_ENTRY_TYPE_TNUMT,
 };
 
@@ -758,6 +759,7 @@ mlxsw_sp_kvdl_entry_size(enum mlxsw_sp_kvdl_entry_type type)
 	case MLXSW_SP_KVDL_ENTRY_TYPE_ACTSET:
 	case MLXSW_SP_KVDL_ENTRY_TYPE_PBS:
 	case MLXSW_SP_KVDL_ENTRY_TYPE_MCRIGR:
+	case MLXSW_SP_KVDL_ENTRY_TYPE_IPV6_ADDRESS:
 	case MLXSW_SP_KVDL_ENTRY_TYPE_TNUMT:
 	default:
 		return 1;
diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum2_kvdl.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum2_kvdl.c
index 3a73d654017f..10ae1115de6c 100644
--- a/drivers/net/ethernet/mellanox/mlxsw/spectrum2_kvdl.c
+++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum2_kvdl.c
@@ -35,6 +35,7 @@ static const struct mlxsw_sp2_kvdl_part_info mlxsw_sp2_kvdl_parts_info[] = {
 				 MAX_KVD_ACTION_SETS),
 	MLXSW_SP2_KVDL_PART_INFO(PBS, 0x24, KVD_SIZE, KVD_SIZE),
 	MLXSW_SP2_KVDL_PART_INFO(MCRIGR, 0x26, KVD_SIZE, KVD_SIZE),
+	MLXSW_SP2_KVDL_PART_INFO(IPV6_ADDRESS, 0x28, KVD_SIZE, KVD_SIZE),
 	MLXSW_SP2_KVDL_PART_INFO(TNUMT, 0x29, KVD_SIZE, KVD_SIZE),
 };
 
-- 
2.31.1

[PATCH net-next 13/14] mlxsw: spectrum_router: Increase parsing depth for IPv6 decapsulation

[Ido Schimmel]

From: Amit Cohen <amcohen@nvidia.com>

The Spectrum ASIC has a configurable limit on how deep into the packet
it parses. By default, the limit is 96 bytes.

For IP-in-IP packets, with IPv6 outer and inner headers, the default
parsing depth is not enough and without increasing it such packets cannot
be properly decapsulated.

Use the existing API to set parsing depth, call it once for each
decapsulation entry when it is created/destroyed.
There is no need to protect the code with new mutex because 'router->lock'
is already taken in these code paths.

Signed-off-by: Amit Cohen <amcohen@nvidia.com>
Reviewed-by: Petr Machata <petrm@nvidia.com>
Signed-off-by: Ido Schimmel <idosch@nvidia.com>
---
 .../ethernet/mellanox/mlxsw/spectrum_ipip.c   |  2 +
 .../ethernet/mellanox/mlxsw/spectrum_ipip.h   |  1 +
 .../ethernet/mellanox/mlxsw/spectrum_router.c | 40 +++++++++++++++++++
 3 files changed, 43 insertions(+)

diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_ipip.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_ipip.c
index 4bb4a3e3a2aa..37a1ad92ac91 100644
--- a/drivers/net/ethernet/mellanox/mlxsw/spectrum_ipip.c
+++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_ipip.c
@@ -316,6 +316,7 @@ mlxsw_sp_ipip_ol_netdev_change_gre4(struct mlxsw_sp *mlxsw_sp,
 static const struct mlxsw_sp_ipip_ops mlxsw_sp_ipip_gre4_ops = {
 	.dev_type = ARPHRD_IPGRE,
 	.ul_proto = MLXSW_SP_L3_PROTO_IPV4,
+	.inc_parsing_depth = false,
 	.parms_init = mlxsw_sp_ipip_netdev_parms_init_gre4,
 	.nexthop_update = mlxsw_sp_ipip_nexthop_update_gre4,
 	.decap_config = mlxsw_sp_ipip_decap_config_gre4,
@@ -379,6 +380,7 @@ mlxsw_sp1_ipip_ol_netdev_change_gre6(struct mlxsw_sp *mlxsw_sp,
 static const struct mlxsw_sp_ipip_ops mlxsw_sp1_ipip_gre6_ops = {
 	.dev_type = ARPHRD_IP6GRE,
 	.ul_proto = MLXSW_SP_L3_PROTO_IPV6,
+	.inc_parsing_depth = true,
 	.parms_init = mlxsw_sp1_ipip_netdev_parms_init_gre6,
 	.nexthop_update = mlxsw_sp1_ipip_nexthop_update_gre6,
 	.decap_config = mlxsw_sp1_ipip_decap_config_gre6,
diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_ipip.h b/drivers/net/ethernet/mellanox/mlxsw/spectrum_ipip.h
index 5d93337b9a2d..2444f09d3fb1 100644
--- a/drivers/net/ethernet/mellanox/mlxsw/spectrum_ipip.h
+++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_ipip.h
@@ -46,6 +46,7 @@ struct mlxsw_sp_ipip_entry {
 struct mlxsw_sp_ipip_ops {
 	int dev_type;
 	enum mlxsw_sp_l3proto ul_proto; /* Underlay. */
+	bool inc_parsing_depth;
 
 	struct mlxsw_sp_ipip_parms
 	(*parms_init)(const struct net_device *ol_dev);
diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c
index 321f19f21d18..ed3f308d69cd 100644
--- a/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c
+++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c
@@ -1157,6 +1157,32 @@ mlxsw_sp_ipip_entry_saddr_matches(struct mlxsw_sp *mlxsw_sp,
 	       mlxsw_sp_l3addr_eq(&tun_saddr, &saddr);
 }
 
+static int mlxsw_sp_ipip_decap_parsing_depth_inc(struct mlxsw_sp *mlxsw_sp,
+						 enum mlxsw_sp_ipip_type ipipt)
+{
+	const struct mlxsw_sp_ipip_ops *ipip_ops;
+
+	ipip_ops = mlxsw_sp->router->ipip_ops_arr[ipipt];
+
+	/* Not all tunnels require to increase the default pasing depth
+	 * (96 bytes).
+	 */
+	if (ipip_ops->inc_parsing_depth)
+		return mlxsw_sp_parsing_depth_inc(mlxsw_sp);
+
+	return 0;
+}
+
+static void mlxsw_sp_ipip_decap_parsing_depth_dec(struct mlxsw_sp *mlxsw_sp,
+						  enum mlxsw_sp_ipip_type ipipt)
+{
+	const struct mlxsw_sp_ipip_ops *ipip_ops =
+		mlxsw_sp->router->ipip_ops_arr[ipipt];
+
+	if (ipip_ops->inc_parsing_depth)
+		mlxsw_sp_parsing_depth_dec(mlxsw_sp);
+}
+
 static int
 mlxsw_sp_fib_entry_decap_init(struct mlxsw_sp *mlxsw_sp,
 			      struct mlxsw_sp_fib_entry *fib_entry,
@@ -1170,18 +1196,32 @@ mlxsw_sp_fib_entry_decap_init(struct mlxsw_sp *mlxsw_sp,
 	if (err)
 		return err;
 
+	err = mlxsw_sp_ipip_decap_parsing_depth_inc(mlxsw_sp,
+						    ipip_entry->ipipt);
+	if (err)
+		goto err_parsing_depth_inc;
+
 	ipip_entry->decap_fib_entry = fib_entry;
 	fib_entry->decap.ipip_entry = ipip_entry;
 	fib_entry->decap.tunnel_index = tunnel_index;
+
 	return 0;
+
+err_parsing_depth_inc:
+	mlxsw_sp_kvdl_free(mlxsw_sp, MLXSW_SP_KVDL_ENTRY_TYPE_ADJ, 1,
+			   fib_entry->decap.tunnel_index);
+	return err;
 }
 
 static void mlxsw_sp_fib_entry_decap_fini(struct mlxsw_sp *mlxsw_sp,
 					  struct mlxsw_sp_fib_entry *fib_entry)
 {
+	enum mlxsw_sp_ipip_type ipipt = fib_entry->decap.ipip_entry->ipipt;
+
 	/* Unlink this node from the IPIP entry that it's the decap entry of. */
 	fib_entry->decap.ipip_entry->decap_fib_entry = NULL;
 	fib_entry->decap.ipip_entry = NULL;
+	mlxsw_sp_ipip_decap_parsing_depth_dec(mlxsw_sp, ipipt);
 	mlxsw_sp_kvdl_free(mlxsw_sp, MLXSW_SP_KVDL_ENTRY_TYPE_ADJ,
 			   1, fib_entry->decap.tunnel_index);
 }
-- 
2.31.1

[PATCH net-next 14/14] mlxsw: Add support for IP-in-IP with IPv6 underlay for Spectrum-2 and above

[Ido Schimmel]

From: Amit Cohen <amcohen@nvidia.com>

Currently, mlxsw driver supports IP-in-IP only with IPv4 underlay.
Add support for IPv6 underlay for Spectrum-2 and above.

Most of the configurations are same to IPv4, the main difference between
IPv4 and IPv6 is related to saving IP addresses.
IPv6 addresses are saved as part of KVD and the relevant registers hold
pointer to them.
Add API for that as part of ipip_ops, so then only Spectrum-2 and above
will save IPv6 addresses in this way.

Signed-off-by: Amit Cohen <amcohen@nvidia.com>
Reviewed-by: Petr Machata <petrm@nvidia.com>
Signed-off-by: Ido Schimmel <idosch@nvidia.com>
---
 .../ethernet/mellanox/mlxsw/spectrum_ipip.c   | 227 +++++++++++++++++-
 .../ethernet/mellanox/mlxsw/spectrum_ipip.h   |   6 +
 .../ethernet/mellanox/mlxsw/spectrum_router.c |  86 ++++++-
 3 files changed, 309 insertions(+), 10 deletions(-)

diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_ipip.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_ipip.c
index 37a1ad92ac91..ad3926de88f2 100644
--- a/drivers/net/ethernet/mellanox/mlxsw/spectrum_ipip.c
+++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_ipip.c
@@ -29,23 +29,45 @@ static bool mlxsw_sp_ipip_parms4_has_ikey(const struct ip_tunnel_parm *parms)
 	return !!(parms->i_flags & TUNNEL_KEY);
 }
 
+static bool mlxsw_sp_ipip_parms6_has_ikey(const struct __ip6_tnl_parm *parms)
+{
+	return !!(parms->i_flags & TUNNEL_KEY);
+}
+
 static bool mlxsw_sp_ipip_parms4_has_okey(const struct ip_tunnel_parm *parms)
 {
 	return !!(parms->o_flags & TUNNEL_KEY);
 }
 
+static bool mlxsw_sp_ipip_parms6_has_okey(const struct __ip6_tnl_parm *parms)
+{
+	return !!(parms->o_flags & TUNNEL_KEY);
+}
+
 static u32 mlxsw_sp_ipip_parms4_ikey(const struct ip_tunnel_parm *parms)
 {
 	return mlxsw_sp_ipip_parms4_has_ikey(parms) ?
 		be32_to_cpu(parms->i_key) : 0;
 }
 
+static u32 mlxsw_sp_ipip_parms6_ikey(const struct __ip6_tnl_parm *parms)
+{
+	return mlxsw_sp_ipip_parms6_has_ikey(parms) ?
+		be32_to_cpu(parms->i_key) : 0;
+}
+
 static u32 mlxsw_sp_ipip_parms4_okey(const struct ip_tunnel_parm *parms)
 {
 	return mlxsw_sp_ipip_parms4_has_okey(parms) ?
 		be32_to_cpu(parms->o_key) : 0;
 }
 
+static u32 mlxsw_sp_ipip_parms6_okey(const struct __ip6_tnl_parm *parms)
+{
+	return mlxsw_sp_ipip_parms6_has_okey(parms) ?
+		be32_to_cpu(parms->o_key) : 0;
+}
+
 static union mlxsw_sp_l3addr
 mlxsw_sp_ipip_parms4_saddr(const struct ip_tunnel_parm *parms)
 {
@@ -313,6 +335,19 @@ mlxsw_sp_ipip_ol_netdev_change_gre4(struct mlxsw_sp *mlxsw_sp,
 						  &new_parms, extack);
 }
 
+static int
+mlxsw_sp_ipip_rem_addr_set_gre4(struct mlxsw_sp *mlxsw_sp,
+				struct mlxsw_sp_ipip_entry *ipip_entry)
+{
+	return 0;
+}
+
+static void
+mlxsw_sp_ipip_rem_addr_unset_gre4(struct mlxsw_sp *mlxsw_sp,
+				  const struct mlxsw_sp_ipip_entry *ipip_entry)
+{
+}
+
 static const struct mlxsw_sp_ipip_ops mlxsw_sp_ipip_gre4_ops = {
 	.dev_type = ARPHRD_IPGRE,
 	.ul_proto = MLXSW_SP_L3_PROTO_IPV4,
@@ -323,6 +358,8 @@ static const struct mlxsw_sp_ipip_ops mlxsw_sp_ipip_gre4_ops = {
 	.can_offload = mlxsw_sp_ipip_can_offload_gre4,
 	.ol_loopback_config = mlxsw_sp_ipip_ol_loopback_config_gre4,
 	.ol_netdev_change = mlxsw_sp_ipip_ol_netdev_change_gre4,
+	.rem_ip_addr_set = mlxsw_sp_ipip_rem_addr_set_gre4,
+	.rem_ip_addr_unset = mlxsw_sp_ipip_rem_addr_unset_gre4,
 };
 
 static struct mlxsw_sp_ipip_parms
@@ -377,6 +414,21 @@ mlxsw_sp1_ipip_ol_netdev_change_gre6(struct mlxsw_sp *mlxsw_sp,
 	return -EINVAL;
 }
 
+static int
+mlxsw_sp1_ipip_rem_addr_set_gre6(struct mlxsw_sp *mlxsw_sp,
+				 struct mlxsw_sp_ipip_entry *ipip_entry)
+{
+	WARN_ON_ONCE(1);
+	return -EINVAL;
+}
+
+static void
+mlxsw_sp1_ipip_rem_addr_unset_gre6(struct mlxsw_sp *mlxsw_sp,
+				   const struct mlxsw_sp_ipip_entry *ipip_entry)
+{
+	WARN_ON_ONCE(1);
+}
+
 static const struct mlxsw_sp_ipip_ops mlxsw_sp1_ipip_gre6_ops = {
 	.dev_type = ARPHRD_IP6GRE,
 	.ul_proto = MLXSW_SP_L3_PROTO_IPV6,
@@ -387,6 +439,8 @@ static const struct mlxsw_sp_ipip_ops mlxsw_sp1_ipip_gre6_ops = {
 	.can_offload = mlxsw_sp1_ipip_can_offload_gre6,
 	.ol_loopback_config = mlxsw_sp1_ipip_ol_loopback_config_gre6,
 	.ol_netdev_change = mlxsw_sp1_ipip_ol_netdev_change_gre6,
+	.rem_ip_addr_set = mlxsw_sp1_ipip_rem_addr_set_gre6,
+	.rem_ip_addr_unset = mlxsw_sp1_ipip_rem_addr_unset_gre6,
 };
 
 const struct mlxsw_sp_ipip_ops *mlxsw_sp1_ipip_ops_arr[] = {
@@ -394,9 +448,176 @@ const struct mlxsw_sp_ipip_ops *mlxsw_sp1_ipip_ops_arr[] = {
 	[MLXSW_SP_IPIP_TYPE_GRE6] = &mlxsw_sp1_ipip_gre6_ops,
 };
 
+static struct mlxsw_sp_ipip_parms
+mlxsw_sp2_ipip_netdev_parms_init_gre6(const struct net_device *ol_dev)
+{
+	struct __ip6_tnl_parm parms = mlxsw_sp_ipip_netdev_parms6(ol_dev);
+
+	return (struct mlxsw_sp_ipip_parms) {
+		.proto = MLXSW_SP_L3_PROTO_IPV6,
+		.saddr = mlxsw_sp_ipip_parms6_saddr(&parms),
+		.daddr = mlxsw_sp_ipip_parms6_daddr(&parms),
+		.link = parms.link,
+		.ikey = mlxsw_sp_ipip_parms6_ikey(&parms),
+		.okey = mlxsw_sp_ipip_parms6_okey(&parms),
+	};
+}
+
+static int
+mlxsw_sp2_ipip_nexthop_update_gre6(struct mlxsw_sp *mlxsw_sp, u32 adj_index,
+				   struct mlxsw_sp_ipip_entry *ipip_entry,
+				   bool force, char *ratr_pl)
+{
+	u16 rif_index = mlxsw_sp_ipip_lb_rif_index(ipip_entry->ol_lb);
+	enum mlxsw_reg_ratr_op op;
+
+	op = force ? MLXSW_REG_RATR_OP_WRITE_WRITE_ENTRY :
+		     MLXSW_REG_RATR_OP_WRITE_WRITE_ENTRY_ON_ACTIVITY;
+	mlxsw_reg_ratr_pack(ratr_pl, op, true, MLXSW_REG_RATR_TYPE_IPIP,
+			    adj_index, rif_index);
+	mlxsw_reg_ratr_ipip6_entry_pack(ratr_pl,
+					ipip_entry->dip_kvdl_index);
+
+	return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(ratr), ratr_pl);
+}
+
+static int
+mlxsw_sp2_ipip_decap_config_gre6(struct mlxsw_sp *mlxsw_sp,
+				 struct mlxsw_sp_ipip_entry *ipip_entry,
+				 u32 tunnel_index)
+{
+	u16 rif_index = mlxsw_sp_ipip_lb_rif_index(ipip_entry->ol_lb);
+	u16 ul_rif_id = mlxsw_sp_ipip_lb_ul_rif_id(ipip_entry->ol_lb);
+	char rtdp_pl[MLXSW_REG_RTDP_LEN];
+	struct __ip6_tnl_parm parms;
+	unsigned int type_check;
+	bool has_ikey;
+	u32 ikey;
+
+	parms = mlxsw_sp_ipip_netdev_parms6(ipip_entry->ol_dev);
+	has_ikey = mlxsw_sp_ipip_parms6_has_ikey(&parms);
+	ikey = mlxsw_sp_ipip_parms6_ikey(&parms);
+
+	mlxsw_reg_rtdp_pack(rtdp_pl, MLXSW_REG_RTDP_TYPE_IPIP, tunnel_index);
+	mlxsw_reg_rtdp_egress_router_interface_set(rtdp_pl, ul_rif_id);
+
+	type_check = has_ikey ?
+		MLXSW_REG_RTDP_IPIP_TYPE_CHECK_ALLOW_GRE_KEY :
+		MLXSW_REG_RTDP_IPIP_TYPE_CHECK_ALLOW_GRE;
+
+	/* Linux demuxes tunnels based on packet SIP (which must match tunnel
+	 * remote IP). Thus configure decap so that it filters out packets that
+	 * are not IPv6 or have the wrong SIP. IPIP_DECAP_ERROR trap is
+	 * generated for packets that fail this criterion. Linux then handles
+	 * such packets in slow path and generates ICMP destination unreachable.
+	 */
+	mlxsw_reg_rtdp_ipip6_pack(rtdp_pl, rif_index,
+				  MLXSW_REG_RTDP_IPIP_SIP_CHECK_FILTER_IPV6,
+				  type_check, has_ikey,
+				  ipip_entry->dip_kvdl_index, ikey);
+
+	return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(rtdp), rtdp_pl);
+}
+
+static bool mlxsw_sp2_ipip_can_offload_gre6(const struct mlxsw_sp *mlxsw_sp,
+					    const struct net_device *ol_dev)
+{
+	struct __ip6_tnl_parm tparm = mlxsw_sp_ipip_netdev_parms6(ol_dev);
+	bool inherit_tos = tparm.flags & IP6_TNL_F_USE_ORIG_TCLASS;
+	bool inherit_ttl = tparm.hop_limit == 0;
+	__be16 okflags = TUNNEL_KEY; /* We can't offload any other features. */
+
+	return (tparm.i_flags & ~okflags) == 0 &&
+	       (tparm.o_flags & ~okflags) == 0 &&
+	       inherit_ttl && inherit_tos &&
+	       mlxsw_sp_ipip_tunnel_complete(MLXSW_SP_L3_PROTO_IPV6, ol_dev);
+}
+
+static struct mlxsw_sp_rif_ipip_lb_config
+mlxsw_sp2_ipip_ol_loopback_config_gre6(struct mlxsw_sp *mlxsw_sp,
+				       const struct net_device *ol_dev)
+{
+	struct __ip6_tnl_parm parms = mlxsw_sp_ipip_netdev_parms6(ol_dev);
+	enum mlxsw_reg_ritr_loopback_ipip_type lb_ipipt;
+
+	lb_ipipt = mlxsw_sp_ipip_parms6_has_okey(&parms) ?
+		MLXSW_REG_RITR_LOOPBACK_IPIP_TYPE_IP_IN_GRE_KEY_IN_IP :
+		MLXSW_REG_RITR_LOOPBACK_IPIP_TYPE_IP_IN_GRE_IN_IP;
+	return (struct mlxsw_sp_rif_ipip_lb_config){
+		.lb_ipipt = lb_ipipt,
+		.okey = mlxsw_sp_ipip_parms6_okey(&parms),
+		.ul_protocol = MLXSW_SP_L3_PROTO_IPV6,
+		.saddr = mlxsw_sp_ipip_netdev_saddr(MLXSW_SP_L3_PROTO_IPV6,
+						    ol_dev),
+	};
+}
+
+static int
+mlxsw_sp2_ipip_ol_netdev_change_gre6(struct mlxsw_sp *mlxsw_sp,
+				     struct mlxsw_sp_ipip_entry *ipip_entry,
+				     struct netlink_ext_ack *extack)
+{
+	struct mlxsw_sp_ipip_parms new_parms;
+
+	new_parms = mlxsw_sp2_ipip_netdev_parms_init_gre6(ipip_entry->ol_dev);
+	return mlxsw_sp_ipip_ol_netdev_change_gre(mlxsw_sp, ipip_entry,
+						  &new_parms, extack);
+}
+
+static int
+mlxsw_sp2_ipip_rem_addr_set_gre6(struct mlxsw_sp *mlxsw_sp,
+				 struct mlxsw_sp_ipip_entry *ipip_entry)
+{
+	char rips_pl[MLXSW_REG_RIPS_LEN];
+	struct __ip6_tnl_parm parms6;
+	int err;
+
+	err = mlxsw_sp_kvdl_alloc(mlxsw_sp,
+				  MLXSW_SP_KVDL_ENTRY_TYPE_IPV6_ADDRESS, 1,
+				  &ipip_entry->dip_kvdl_index);
+	if (err)
+		return err;
+
+	parms6 = mlxsw_sp_ipip_netdev_parms6(ipip_entry->ol_dev);
+	mlxsw_reg_rips_pack(rips_pl, ipip_entry->dip_kvdl_index,
+			    &parms6.raddr);
+	err = mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(rips), rips_pl);
+	if (err)
+		goto err_rips_write;
+
+	return 0;
+
+err_rips_write:
+	mlxsw_sp_kvdl_free(mlxsw_sp, MLXSW_SP_KVDL_ENTRY_TYPE_IPV6_ADDRESS, 1,
+			   ipip_entry->dip_kvdl_index);
+	return err;
+}
+
+static void
+mlxsw_sp2_ipip_rem_addr_unset_gre6(struct mlxsw_sp *mlxsw_sp,
+				   const struct mlxsw_sp_ipip_entry *ipip_entry)
+{
+	mlxsw_sp_kvdl_free(mlxsw_sp, MLXSW_SP_KVDL_ENTRY_TYPE_IPV6_ADDRESS, 1,
+			   ipip_entry->dip_kvdl_index);
+}
+
+static const struct mlxsw_sp_ipip_ops mlxsw_sp2_ipip_gre6_ops = {
+	.dev_type = ARPHRD_IP6GRE,
+	.ul_proto = MLXSW_SP_L3_PROTO_IPV6,
+	.inc_parsing_depth = true,
+	.parms_init = mlxsw_sp2_ipip_netdev_parms_init_gre6,
+	.nexthop_update = mlxsw_sp2_ipip_nexthop_update_gre6,
+	.decap_config = mlxsw_sp2_ipip_decap_config_gre6,
+	.can_offload = mlxsw_sp2_ipip_can_offload_gre6,
+	.ol_loopback_config = mlxsw_sp2_ipip_ol_loopback_config_gre6,
+	.ol_netdev_change = mlxsw_sp2_ipip_ol_netdev_change_gre6,
+	.rem_ip_addr_set = mlxsw_sp2_ipip_rem_addr_set_gre6,
+	.rem_ip_addr_unset = mlxsw_sp2_ipip_rem_addr_unset_gre6,
+};
+
 const struct mlxsw_sp_ipip_ops *mlxsw_sp2_ipip_ops_arr[] = {
 	[MLXSW_SP_IPIP_TYPE_GRE4] = &mlxsw_sp_ipip_gre4_ops,
-	[MLXSW_SP_IPIP_TYPE_GRE6] = &mlxsw_sp1_ipip_gre6_ops,
+	[MLXSW_SP_IPIP_TYPE_GRE6] = &mlxsw_sp2_ipip_gre6_ops,
 };
 
 static int mlxsw_sp_ipip_ecn_encap_init_one(struct mlxsw_sp *mlxsw_sp,
@@ -461,11 +682,15 @@ mlxsw_sp_ipip_netdev_ul_dev_get(const struct net_device *ol_dev)
 {
 	struct net *net = dev_net(ol_dev);
 	struct ip_tunnel *tun4;
+	struct ip6_tnl *tun6;
 
 	switch (ol_dev->type) {
 	case ARPHRD_IPGRE:
 		tun4 = netdev_priv(ol_dev);
 		return dev_get_by_index_rcu(net, tun4->parms.link);
+	case ARPHRD_IP6GRE:
+		tun6 = netdev_priv(ol_dev);
+		return dev_get_by_index_rcu(net, tun6->parms.link);
 	default:
 		return NULL;
 	}
diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_ipip.h b/drivers/net/ethernet/mellanox/mlxsw/spectrum_ipip.h
index 2444f09d3fb1..8cc259dcc8d0 100644
--- a/drivers/net/ethernet/mellanox/mlxsw/spectrum_ipip.h
+++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_ipip.h
@@ -7,6 +7,7 @@
 #include "spectrum_router.h"
 #include <net/ip_fib.h>
 #include <linux/if_tunnel.h>
+#include <net/ip6_tunnel.h>
 
 struct ip_tunnel_parm
 mlxsw_sp_ipip_netdev_parms4(const struct net_device *ol_dev);
@@ -41,6 +42,7 @@ struct mlxsw_sp_ipip_entry {
 	struct mlxsw_sp_fib_entry *decap_fib_entry;
 	struct list_head ipip_list_node;
 	struct mlxsw_sp_ipip_parms parms;
+	u32 dip_kvdl_index;
 };
 
 struct mlxsw_sp_ipip_ops {
@@ -70,6 +72,10 @@ struct mlxsw_sp_ipip_ops {
 	int (*ol_netdev_change)(struct mlxsw_sp *mlxsw_sp,
 				struct mlxsw_sp_ipip_entry *ipip_entry,
 				struct netlink_ext_ack *extack);
+	int (*rem_ip_addr_set)(struct mlxsw_sp *mlxsw_sp,
+			       struct mlxsw_sp_ipip_entry *ipip_entry);
+	void (*rem_ip_addr_unset)(struct mlxsw_sp *mlxsw_sp,
+				  const struct mlxsw_sp_ipip_entry *ipip_entry);
 };
 
 extern const struct mlxsw_sp_ipip_ops *mlxsw_sp1_ipip_ops_arr[];
diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c
index ed3f308d69cd..1e141b5944cd 100644
--- a/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c
+++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c
@@ -1108,6 +1108,7 @@ mlxsw_sp_ipip_entry_alloc(struct mlxsw_sp *mlxsw_sp,
 	const struct mlxsw_sp_ipip_ops *ipip_ops;
 	struct mlxsw_sp_ipip_entry *ipip_entry;
 	struct mlxsw_sp_ipip_entry *ret = NULL;
+	int err;
 
 	ipip_ops = mlxsw_sp->router->ipip_ops_arr[ipipt];
 	ipip_entry = kzalloc(sizeof(*ipip_entry), GFP_KERNEL);
@@ -1124,16 +1125,29 @@ mlxsw_sp_ipip_entry_alloc(struct mlxsw_sp *mlxsw_sp,
 	ipip_entry->ipipt = ipipt;
 	ipip_entry->ol_dev = ol_dev;
 	ipip_entry->parms = ipip_ops->parms_init(ol_dev);
+
+	err = ipip_ops->rem_ip_addr_set(mlxsw_sp, ipip_entry);
+	if (err) {
+		ret = ERR_PTR(err);
+		goto err_rem_ip_addr_set;
+	}
+
 	return ipip_entry;
 
+err_rem_ip_addr_set:
+	mlxsw_sp_rif_destroy(&ipip_entry->ol_lb->common);
 err_ol_ipip_lb_create:
 	kfree(ipip_entry);
 	return ret;
 }
 
-static void
-mlxsw_sp_ipip_entry_dealloc(struct mlxsw_sp_ipip_entry *ipip_entry)
+static void mlxsw_sp_ipip_entry_dealloc(struct mlxsw_sp *mlxsw_sp,
+					struct mlxsw_sp_ipip_entry *ipip_entry)
 {
+	const struct mlxsw_sp_ipip_ops *ipip_ops =
+		mlxsw_sp->router->ipip_ops_arr[ipip_entry->ipipt];
+
+	ipip_ops->rem_ip_addr_unset(mlxsw_sp, ipip_entry);
 	mlxsw_sp_rif_destroy(&ipip_entry->ol_lb->common);
 	kfree(ipip_entry);
 }
@@ -1332,6 +1346,11 @@ mlxsw_sp_ipip_entry_find_decap(struct mlxsw_sp *mlxsw_sp,
 		saddr_len = 4;
 		saddr_prefix_len = 32;
 		break;
+	case MLXSW_SP_L3_PROTO_IPV6:
+		saddrp = &saddr.addr6;
+		saddr_len = 16;
+		saddr_prefix_len = 128;
+		break;
 	default:
 		WARN_ON(1);
 		return NULL;
@@ -1368,7 +1387,7 @@ mlxsw_sp_ipip_entry_destroy(struct mlxsw_sp *mlxsw_sp,
 			    struct mlxsw_sp_ipip_entry *ipip_entry)
 {
 	list_del(&ipip_entry->ipip_list_node);
-	mlxsw_sp_ipip_entry_dealloc(ipip_entry);
+	mlxsw_sp_ipip_entry_dealloc(mlxsw_sp, ipip_entry);
 }
 
 static bool
@@ -1563,6 +1582,7 @@ mlxsw_sp_rif_ipip_lb_op(struct mlxsw_sp_rif_ipip_lb *lb_rif, u16 ul_vr_id,
 	struct mlxsw_sp_rif *rif = &lb_rif->common;
 	struct mlxsw_sp *mlxsw_sp = rif->mlxsw_sp;
 	char ritr_pl[MLXSW_REG_RITR_LEN];
+	struct in6_addr *saddr6;
 	u32 saddr4;
 
 	ipip_options = MLXSW_REG_RITR_LOOPBACK_IPIP_OPTIONS_GRE_KEY_PRESET;
@@ -1578,7 +1598,14 @@ mlxsw_sp_rif_ipip_lb_op(struct mlxsw_sp_rif_ipip_lb *lb_rif, u16 ul_vr_id,
 		break;
 
 	case MLXSW_SP_L3_PROTO_IPV6:
-		return -EAFNOSUPPORT;
+		saddr6 = &lb_cf.saddr.addr6;
+		mlxsw_reg_ritr_pack(ritr_pl, enable, MLXSW_REG_RITR_LOOPBACK_IF,
+				    rif->rif_index, rif->vr_id, rif->dev->mtu);
+		mlxsw_reg_ritr_loopback_ipip6_pack(ritr_pl, lb_cf.lb_ipipt,
+						   ipip_options, ul_vr_id,
+						   ul_rif_id, saddr6,
+						   lb_cf.okey);
+		break;
 	}
 
 	return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(ritr), ritr_pl);
@@ -6960,11 +6987,38 @@ mlxsw_sp_fib6_entry_nexthop_del(struct mlxsw_sp *mlxsw_sp,
 	mlxsw_sp_nexthop6_group_update(mlxsw_sp, op_ctx, fib6_entry);
 }
 
-static void mlxsw_sp_fib6_entry_type_set(struct mlxsw_sp *mlxsw_sp,
-					 struct mlxsw_sp_fib_entry *fib_entry,
-					 const struct fib6_info *rt)
+static int
+mlxsw_sp_fib6_entry_type_set_local(struct mlxsw_sp *mlxsw_sp,
+				   struct mlxsw_sp_fib_entry *fib_entry,
+				   const struct fib6_info *rt)
+{
+	struct mlxsw_sp_nexthop_group_info *nhgi = fib_entry->nh_group->nhgi;
+	union mlxsw_sp_l3addr dip = { .addr6 = rt->fib6_dst.addr };
+	int ifindex = nhgi->nexthops[0].ifindex;
+	struct mlxsw_sp_ipip_entry *ipip_entry;
+
+	fib_entry->type = MLXSW_SP_FIB_ENTRY_TYPE_TRAP;
+	ipip_entry = mlxsw_sp_ipip_entry_find_by_decap(mlxsw_sp, ifindex,
+						       MLXSW_SP_L3_PROTO_IPV6,
+						       dip);
+
+	if (ipip_entry && ipip_entry->ol_dev->flags & IFF_UP) {
+		fib_entry->type = MLXSW_SP_FIB_ENTRY_TYPE_IPIP_DECAP;
+		return mlxsw_sp_fib_entry_decap_init(mlxsw_sp, fib_entry,
+						     ipip_entry);
+	}
+
+	return 0;
+}
+
+static int mlxsw_sp_fib6_entry_type_set(struct mlxsw_sp *mlxsw_sp,
+					struct mlxsw_sp_fib_entry *fib_entry,
+					const struct fib6_info *rt)
 {
-	if (rt->fib6_flags & (RTF_LOCAL | RTF_ANYCAST))
+	if (rt->fib6_flags & RTF_LOCAL)
+		return mlxsw_sp_fib6_entry_type_set_local(mlxsw_sp, fib_entry,
+							  rt);
+	if (rt->fib6_flags & RTF_ANYCAST)
 		fib_entry->type = MLXSW_SP_FIB_ENTRY_TYPE_TRAP;
 	else if (rt->fib6_type == RTN_BLACKHOLE)
 		fib_entry->type = MLXSW_SP_FIB_ENTRY_TYPE_BLACKHOLE;
@@ -6974,6 +7028,8 @@ static void mlxsw_sp_fib6_entry_type_set(struct mlxsw_sp *mlxsw_sp,
 		fib_entry->type = MLXSW_SP_FIB_ENTRY_TYPE_REMOTE;
 	else
 		fib_entry->type = MLXSW_SP_FIB_ENTRY_TYPE_LOCAL;
+
+	return 0;
 }
 
 static void
@@ -7031,12 +7087,16 @@ mlxsw_sp_fib6_entry_create(struct mlxsw_sp *mlxsw_sp,
 	if (err)
 		goto err_nexthop_group_vr_link;
 
-	mlxsw_sp_fib6_entry_type_set(mlxsw_sp, fib_entry, rt_arr[0]);
+	err = mlxsw_sp_fib6_entry_type_set(mlxsw_sp, fib_entry, rt_arr[0]);
+	if (err)
+		goto err_fib6_entry_type_set;
 
 	fib_entry->fib_node = fib_node;
 
 	return fib6_entry;
 
+err_fib6_entry_type_set:
+	mlxsw_sp_nexthop_group_vr_unlink(fib_entry->nh_group, fib_node->fib);
 err_nexthop_group_vr_link:
 	mlxsw_sp_nexthop6_group_put(mlxsw_sp, fib_entry);
 err_nexthop6_group_get:
@@ -7055,11 +7115,19 @@ mlxsw_sp_fib6_entry_create(struct mlxsw_sp *mlxsw_sp,
 	return ERR_PTR(err);
 }
 
+static void
+mlxsw_sp_fib6_entry_type_unset(struct mlxsw_sp *mlxsw_sp,
+			       struct mlxsw_sp_fib6_entry *fib6_entry)
+{
+	mlxsw_sp_fib_entry_type_unset(mlxsw_sp, &fib6_entry->common);
+}
+
 static void mlxsw_sp_fib6_entry_destroy(struct mlxsw_sp *mlxsw_sp,
 					struct mlxsw_sp_fib6_entry *fib6_entry)
 {
 	struct mlxsw_sp_fib_node *fib_node = fib6_entry->common.fib_node;
 
+	mlxsw_sp_fib6_entry_type_unset(mlxsw_sp, fib6_entry);
 	mlxsw_sp_nexthop_group_vr_unlink(fib6_entry->common.nh_group,
 					 fib_node->fib);
 	mlxsw_sp_nexthop6_group_put(mlxsw_sp, &fib6_entry->common);
-- 
2.31.1

Re: [PATCH net-next 00/14] mlxsw: Add support for IP-in-IP with IPv6 underlay

[patchwork-bot+netdevbpf]

Hello:

This series was applied to netdev/net-next.git (refs/heads/master):

On Thu, 23 Sep 2021 15:36:46 +0300 you wrote:
> From: Ido Schimmel <idosch@nvidia.com>
> 
> Currently, mlxsw only supports IP-in-IP with IPv4 underlay. Traffic
> routed through 'gre' netdevs is encapsulated with IPv4 and GRE headers.
> Similarly, incoming IPv4 GRE packets are decapsulated and routed in the
> overlay VRF (which can be the same as the underlay VRF).
> 
> [...]

Here is the summary with links:
  - [net-next,01/14] mlxsw: spectrum_router: Create common function for fib_entry_type_unset() code
    https://git.kernel.org/netdev/net-next/c/45bce5c99d46
  - [net-next,02/14] mlxsw: spectrum_ipip: Pass IP tunnel parameters by reference and as 'const'
    https://git.kernel.org/netdev/net-next/c/aa6fd8f177d6
  - [net-next,03/14] mlxsw: spectrum_router: Fix arguments alignment
    https://git.kernel.org/netdev/net-next/c/8aba32cea3f3
  - [net-next,04/14] mlxsw: spectrum_ipip: Create common function for mlxsw_sp_ipip_ol_netdev_change_gre()
    https://git.kernel.org/netdev/net-next/c/80ef2abcddbc
  - [net-next,05/14] mlxsw: Take tunnel's type into account when searching underlay device
    https://git.kernel.org/netdev/net-next/c/59bf980dd90f
  - [net-next,06/14] mlxsw: reg: Add Router IP version Six Register
    https://git.kernel.org/netdev/net-next/c/dd8a9552d484
  - [net-next,07/14] mlxsw: reg: Add support for rtdp_ipip6_pack()
    https://git.kernel.org/netdev/net-next/c/a917bb271d16
  - [net-next,08/14] mlxsw: reg: Add support for ratr_ipip6_entry_pack()
    https://git.kernel.org/netdev/net-next/c/c729ae8d6cbc
  - [net-next,09/14] mlxsw: reg: Add support for ritr_loopback_ipip6_pack()
    https://git.kernel.org/netdev/net-next/c/36c2ab890b8f
  - [net-next,10/14] mlxsw: Create separate ipip_ops_arr for different ASICs
    https://git.kernel.org/netdev/net-next/c/a82feba686e8
  - [net-next,11/14] mlxsw: spectrum_ipip: Add mlxsw_sp_ipip_gre6_ops
    https://git.kernel.org/netdev/net-next/c/713e8502fd3e
  - [net-next,12/14] mlxsw: Add IPV6_ADDRESS kvdl entry type
    https://git.kernel.org/netdev/net-next/c/53eedd61dea9
  - [net-next,13/14] mlxsw: spectrum_router: Increase parsing depth for IPv6 decapsulation
    https://git.kernel.org/netdev/net-next/c/8d4f10463cd6
  - [net-next,14/14] mlxsw: Add support for IP-in-IP with IPv6 underlay for Spectrum-2 and above
    https://git.kernel.org/netdev/net-next/c/ba1c71324bc2

You are awesome, thank you!
--
Deet-doot-dot, I am a bot.
https://korg.docs.kernel.org/patchwork/pwbot.html