From f0a56c17e64bb5e7cdb9295df2b5fc21e4949005 Mon Sep 17 00:00:00 2001
From: Kuniyuki Iwashima <kuniyu@amazon.com>
Date: Thu, 15 May 2025 19:27:18 -0700
Subject: [PATCH 01/16] inet: Remove rtnl_is_held arg of
 lwtunnel_valid_encap_type(_attr)?().

Commit f130a0cc1b4f ("inet: fix lwtunnel_valid_encap_type() lock
imbalance") added the rtnl_is_held argument as a temporary fix while
I'm converting nexthop and IPv6 routing table to per-netns RTNL or RCU.

Now all callers of lwtunnel_valid_encap_type() do not hold RTNL.

Let's remove the argument.

Signed-off-by: Kuniyuki Iwashima <kuniyu@amazon.com>
Link: https://patch.msgid.link/20250516022759.44392-3-kuniyu@amazon.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
---
 include/net/lwtunnel.h  | 13 +++++--------
 net/core/lwtunnel.c     | 15 +++------------
 net/ipv4/fib_frontend.c |  4 ++--
 net/ipv4/nexthop.c      |  3 +--
 net/ipv6/route.c        |  6 ++----
 5 files changed, 13 insertions(+), 28 deletions(-)

diff --git a/include/net/lwtunnel.h b/include/net/lwtunnel.h
index 39cd50300a18..c306ebe379a0 100644
--- a/include/net/lwtunnel.h
+++ b/include/net/lwtunnel.h
@@ -116,11 +116,9 @@ int lwtunnel_encap_add_ops(const struct lwtunnel_encap_ops *op,
 int lwtunnel_encap_del_ops(const struct lwtunnel_encap_ops *op,
 			   unsigned int num);
 int lwtunnel_valid_encap_type(u16 encap_type,
-			      struct netlink_ext_ack *extack,
-			      bool rtnl_is_held);
+			      struct netlink_ext_ack *extack);
 int lwtunnel_valid_encap_type_attr(struct nlattr *attr, int len,
-				   struct netlink_ext_ack *extack,
-				   bool rtnl_is_held);
+				   struct netlink_ext_ack *extack);
 int lwtunnel_build_state(struct net *net, u16 encap_type,
 			 struct nlattr *encap,
 			 unsigned int family, const void *cfg,
@@ -203,15 +201,14 @@ static inline int lwtunnel_encap_del_ops(const struct lwtunnel_encap_ops *op,
 }
 
 static inline int lwtunnel_valid_encap_type(u16 encap_type,
-					    struct netlink_ext_ack *extack,
-					    bool rtnl_is_held)
+					    struct netlink_ext_ack *extack)
 {
 	NL_SET_ERR_MSG(extack, "CONFIG_LWTUNNEL is not enabled in this kernel");
 	return -EOPNOTSUPP;
 }
+
 static inline int lwtunnel_valid_encap_type_attr(struct nlattr *attr, int len,
-						 struct netlink_ext_ack *extack,
-						 bool rtnl_is_held)
+						 struct netlink_ext_ack *extack)
 {
 	/* return 0 since we are not walking attr looking for
 	 * RTA_ENCAP_TYPE attribute on nexthops.
diff --git a/net/core/lwtunnel.c b/net/core/lwtunnel.c
index 60f27cb4e54f..f9d76d85d04f 100644
--- a/net/core/lwtunnel.c
+++ b/net/core/lwtunnel.c
@@ -149,8 +149,7 @@ int lwtunnel_build_state(struct net *net, u16 encap_type,
 }
 EXPORT_SYMBOL_GPL(lwtunnel_build_state);
 
-int lwtunnel_valid_encap_type(u16 encap_type, struct netlink_ext_ack *extack,
-			      bool rtnl_is_held)
+int lwtunnel_valid_encap_type(u16 encap_type, struct netlink_ext_ack *extack)
 {
 	const struct lwtunnel_encap_ops *ops;
 	int ret = -EINVAL;
@@ -167,12 +166,7 @@ int lwtunnel_valid_encap_type(u16 encap_type, struct netlink_ext_ack *extack,
 		const char *encap_type_str = lwtunnel_encap_str(encap_type);
 
 		if (encap_type_str) {
-			if (rtnl_is_held)
-				__rtnl_unlock();
 			request_module("rtnl-lwt-%s", encap_type_str);
-			if (rtnl_is_held)
-				rtnl_lock();
-
 			ops = rcu_access_pointer(lwtun_encaps[encap_type]);
 		}
 	}
@@ -186,8 +180,7 @@ int lwtunnel_valid_encap_type(u16 encap_type, struct netlink_ext_ack *extack,
 EXPORT_SYMBOL_GPL(lwtunnel_valid_encap_type);
 
 int lwtunnel_valid_encap_type_attr(struct nlattr *attr, int remaining,
-				   struct netlink_ext_ack *extack,
-				   bool rtnl_is_held)
+				   struct netlink_ext_ack *extack)
 {
 	struct rtnexthop *rtnh = (struct rtnexthop *)attr;
 	struct nlattr *nla_entype;
@@ -208,9 +201,7 @@ int lwtunnel_valid_encap_type_attr(struct nlattr *attr, int remaining,
 				}
 				encap_type = nla_get_u16(nla_entype);
 
-				if (lwtunnel_valid_encap_type(encap_type,
-							      extack,
-							      rtnl_is_held) != 0)
+				if (lwtunnel_valid_encap_type(encap_type, extack))
 					return -EOPNOTSUPP;
 			}
 		}
diff --git a/net/ipv4/fib_frontend.c b/net/ipv4/fib_frontend.c
index 57f088e5540e..fd1e1507a224 100644
--- a/net/ipv4/fib_frontend.c
+++ b/net/ipv4/fib_frontend.c
@@ -807,7 +807,7 @@ static int rtm_to_fib_config(struct net *net, struct sk_buff *skb,
 		case RTA_MULTIPATH:
 			err = lwtunnel_valid_encap_type_attr(nla_data(attr),
 							     nla_len(attr),
-							     extack, false);
+							     extack);
 			if (err < 0)
 				goto errout;
 			cfg->fc_mp = nla_data(attr);
@@ -825,7 +825,7 @@ static int rtm_to_fib_config(struct net *net, struct sk_buff *skb,
 		case RTA_ENCAP_TYPE:
 			cfg->fc_encap_type = nla_get_u16(attr);
 			err = lwtunnel_valid_encap_type(cfg->fc_encap_type,
-							extack, false);
+							extack);
 			if (err < 0)
 				goto errout;
 			break;
diff --git a/net/ipv4/nexthop.c b/net/ipv4/nexthop.c
index 823e4a783d2b..4397e89d3123 100644
--- a/net/ipv4/nexthop.c
+++ b/net/ipv4/nexthop.c
@@ -3180,8 +3180,7 @@ static int rtm_to_nh_config(struct net *net, struct sk_buff *skb,
 		}
 
 		cfg->nh_encap_type = nla_get_u16(tb[NHA_ENCAP_TYPE]);
-		err = lwtunnel_valid_encap_type(cfg->nh_encap_type,
-						extack, false);
+		err = lwtunnel_valid_encap_type(cfg->nh_encap_type, extack);
 		if (err < 0)
 			goto out;
 
diff --git a/net/ipv6/route.c b/net/ipv6/route.c
index 44300962230b..6baf177c529b 100644
--- a/net/ipv6/route.c
+++ b/net/ipv6/route.c
@@ -5172,8 +5172,7 @@ static int rtm_to_fib6_multipath_config(struct fib6_config *cfg,
 		rtnh = rtnh_next(rtnh, &remaining);
 	} while (rtnh_ok(rtnh, remaining));
 
-	return lwtunnel_valid_encap_type_attr(cfg->fc_mp, cfg->fc_mp_len,
-					      extack, false);
+	return lwtunnel_valid_encap_type_attr(cfg->fc_mp, cfg->fc_mp_len, extack);
 }
 
 static int rtm_to_fib6_config(struct sk_buff *skb, struct nlmsghdr *nlh,
@@ -5310,8 +5309,7 @@ static int rtm_to_fib6_config(struct sk_buff *skb, struct nlmsghdr *nlh,
 	if (tb[RTA_ENCAP_TYPE]) {
 		cfg->fc_encap_type = nla_get_u16(tb[RTA_ENCAP_TYPE]);
 
-		err = lwtunnel_valid_encap_type(cfg->fc_encap_type,
-						extack, false);
+		err = lwtunnel_valid_encap_type(cfg->fc_encap_type, extack);
 		if (err < 0)
 			goto errout;
 	}
-- 
2.51.0


From 8e5f1bb812741821e2a8ac221fba45cab6c73e43 Mon Sep 17 00:00:00 2001
From: Kuniyuki Iwashima <kuniyu@amazon.com>
Date: Thu, 15 May 2025 19:27:19 -0700
Subject: [PATCH 02/16] ipv6: Narrow down RCU critical section in
 inet6_rtm_newroute().

Commit 169fd62799e8 ("ipv6: Get rid of RTNL for SIOCADDRT and
RTM_NEWROUTE.") added rcu_read_lock() covering
ip6_route_info_create_nh() and __ip6_ins_rt() to guarantee that
nexthop and netdev will not go away.

However, as reported by syzkaller [0], ip_tun_build_state() calls
dst_cache_init() with GFP_KERNEL during the RCU critical section.

ip6_route_info_create_nh() fetches nexthop or netdev depending on
whether RTA_NH_ID is set, and struct fib6_info holds a refcount
of either of them by nexthop_get() or netdev_get_by_index().

netdev_get_by_index() looks up a dev and calls dev_hold() under RCU.

So, we need RCU only around nexthop_find_by_id() and nexthop_get()
( and a few more nexthop code).

Let's add rcu_read_lock() there and remove rcu_read_lock() in
ip6_route_add() and ip6_route_multipath_add().

Now these functions called from fib6_add() need RCU:

  - inet6_rt_notify()
  - fib6_drop_pcpu_from() (via fib6_purge_rt())
  - rt6_flush_exceptions() (via fib6_purge_rt())
  - ip6_ignore_linkdown() (via rt6_multipath_rebalance())

All callers of inet6_rt_notify() need RCU, so rcu_read_lock() is
added there.

[0]:
[ BUG: Invalid wait context ]
6.15.0-rc4-syzkaller-00746-g836b313a14a3 #0 Tainted: G W
 ----------------------------
syz-executor234/5832 is trying to lock:
ffffffff8e021688 (pcpu_alloc_mutex){+.+.}-{4:4}, at:
pcpu_alloc_noprof+0x284/0x16b0 mm/percpu.c:1782
other info that might help us debug this:
context-{5:5}
1 lock held by syz-executor234/5832:
 0: ffffffff8df3b860 (rcu_read_lock){....}-{1:3}, at: rcu_lock_acquire
include/linux/rcupdate.h:331 [inline]
 0: ffffffff8df3b860 (rcu_read_lock){....}-{1:3}, at: rcu_read_lock
include/linux/rcupdate.h:841 [inline]
 0: ffffffff8df3b860 (rcu_read_lock){....}-{1:3}, at:
ip6_route_add+0x4d/0x2f0 net/ipv6/route.c:3913
stack backtrace:
CPU: 0 UID: 0 PID: 5832 Comm: syz-executor234 Tainted: G W
6.15.0-rc4-syzkaller-00746-g836b313a14a3 #0 PREEMPT(full)
Tainted: [W]=WARN
Hardware name: Google Google Compute Engine/Google Compute Engine,
BIOS Google 04/29/2025
Call Trace:
 <TASK>
 dump_stack_lvl+0x189/0x250 lib/dump_stack.c:120
 print_lock_invalid_wait_context kernel/locking/lockdep.c:4831 [inline]
 check_wait_context kernel/locking/lockdep.c:4903 [inline]
 __lock_acquire+0xbcf/0xd20 kernel/locking/lockdep.c:5185
 lock_acquire+0x120/0x360 kernel/locking/lockdep.c:5866
 __mutex_lock_common kernel/locking/mutex.c:601 [inline]
 __mutex_lock+0x182/0xe80 kernel/locking/mutex.c:746
 pcpu_alloc_noprof+0x284/0x16b0 mm/percpu.c:1782
 dst_cache_init+0x37/0xc0 net/core/dst_cache.c:145
 ip_tun_build_state+0x193/0x6b0 net/ipv4/ip_tunnel_core.c:687
 lwtunnel_build_state+0x381/0x4c0 net/core/lwtunnel.c:137
 fib_nh_common_init+0x129/0x460 net/ipv4/fib_semantics.c:635
 fib6_nh_init+0x15e4/0x2030 net/ipv6/route.c:3669
 ip6_route_info_create_nh+0x139/0x870 net/ipv6/route.c:3866
 ip6_route_add+0xf6/0x2f0 net/ipv6/route.c:3915
 inet6_rtm_newroute+0x284/0x1c50 net/ipv6/route.c:5732
 rtnetlink_rcv_msg+0x7cc/0xb70 net/core/rtnetlink.c:6955
 netlink_rcv_skb+0x219/0x490 net/netlink/af_netlink.c:2534
 netlink_unicast_kernel net/netlink/af_netlink.c:1313 [inline]
 netlink_unicast+0x758/0x8d0 net/netlink/af_netlink.c:1339
 netlink_sendmsg+0x805/0xb30 net/netlink/af_netlink.c:1883
 sock_sendmsg_nosec net/socket.c:712 [inline]
 __sock_sendmsg+0x219/0x270 net/socket.c:727
 ____sys_sendmsg+0x505/0x830 net/socket.c:2566
 ___sys_sendmsg+0x21f/0x2a0 net/socket.c:2620
 __sys_sendmsg net/socket.c:2652 [inline]
 __do_sys_sendmsg net/socket.c:2657 [inline]
 __se_sys_sendmsg net/socket.c:2655 [inline]
 __x64_sys_sendmsg+0x19b/0x260 net/socket.c:2655
 do_syscall_x64 arch/x86/entry/syscall_64.c:63 [inline]
 do_syscall_64+0xf6/0x210 arch/x86/entry/syscall_64.c:94

Fixes: 169fd62799e8 ("ipv6: Get rid of RTNL for SIOCADDRT and RTM_NEWROUTE.")
Reported-by: syzbot+bcc12d6799364500fbec@syzkaller.appspotmail.com
Closes: https://syzkaller.appspot.com/bug?extid=bcc12d6799364500fbec
Reported-by: Eric Dumazet <edumazet@google.com>
Closes: https://lore.kernel.org/netdev/CANn89i+r1cGacVC_6n3-A-WSkAa_Nr+pmxJ7Gt+oP-P9by2aGw@mail.gmail.com/
Signed-off-by: Kuniyuki Iwashima <kuniyu@amazon.com>
Link: https://patch.msgid.link/20250516022759.44392-4-kuniyu@amazon.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
---
 net/ipv6/ip6_fib.c |  9 +++++++--
 net/ipv6/route.c   | 31 ++++++++++++++++++-------------
 2 files changed, 25 insertions(+), 15 deletions(-)

diff --git a/net/ipv6/ip6_fib.c b/net/ipv6/ip6_fib.c
index 88770ecd2da1..7094d7708686 100644
--- a/net/ipv6/ip6_fib.c
+++ b/net/ipv6/ip6_fib.c
@@ -1027,8 +1027,9 @@ static void fib6_drop_pcpu_from(struct fib6_info *f6i,
 			.table = table
 		};
 
-		nexthop_for_each_fib6_nh(f6i->nh, fib6_nh_drop_pcpu_from,
-					 &arg);
+		rcu_read_lock();
+		nexthop_for_each_fib6_nh(f6i->nh, fib6_nh_drop_pcpu_from, &arg);
+		rcu_read_unlock();
 	} else {
 		struct fib6_nh *fib6_nh;
 
@@ -1221,7 +1222,9 @@ next_iter:
 			fib6_nsiblings++;
 		}
 		BUG_ON(fib6_nsiblings != rt->fib6_nsiblings);
+		rcu_read_lock();
 		rt6_multipath_rebalance(temp_sibling);
+		rcu_read_unlock();
 	}
 
 	/*
@@ -1264,7 +1267,9 @@ add:
 					sibling->fib6_nsiblings--;
 				rt->fib6_nsiblings = 0;
 				list_del_rcu(&rt->fib6_siblings);
+				rcu_read_lock();
 				rt6_multipath_rebalance(next_sibling);
+				rcu_read_unlock();
 				return err;
 			}
 		}
diff --git a/net/ipv6/route.c b/net/ipv6/route.c
index 6baf177c529b..a87091dd06b1 100644
--- a/net/ipv6/route.c
+++ b/net/ipv6/route.c
@@ -1820,11 +1820,13 @@ static int rt6_nh_flush_exceptions(struct fib6_nh *nh, void *arg)
 
 void rt6_flush_exceptions(struct fib6_info *f6i)
 {
-	if (f6i->nh)
-		nexthop_for_each_fib6_nh(f6i->nh, rt6_nh_flush_exceptions,
-					 f6i);
-	else
+	if (f6i->nh) {
+		rcu_read_lock();
+		nexthop_for_each_fib6_nh(f6i->nh, rt6_nh_flush_exceptions, f6i);
+		rcu_read_unlock();
+	} else {
 		fib6_nh_flush_exceptions(f6i->fib6_nh, f6i);
+	}
 }
 
 /* Find cached rt in the hash table inside passed in rt
@@ -3841,6 +3843,8 @@ static int ip6_route_info_create_nh(struct fib6_info *rt,
 	if (cfg->fc_nh_id) {
 		struct nexthop *nh;
 
+		rcu_read_lock();
+
 		nh = nexthop_find_by_id(net, cfg->fc_nh_id);
 		if (!nh) {
 			err = -EINVAL;
@@ -3860,6 +3864,8 @@ static int ip6_route_info_create_nh(struct fib6_info *rt,
 
 		rt->nh = nh;
 		fib6_nh = nexthop_fib6_nh(rt->nh);
+
+		rcu_read_unlock();
 	} else {
 		int addr_type;
 
@@ -3895,6 +3901,7 @@ out_release:
 	fib6_info_release(rt);
 	return err;
 out_free:
+	rcu_read_unlock();
 	ip_fib_metrics_put(rt->fib6_metrics);
 	kfree(rt);
 	return err;
@@ -3910,16 +3917,12 @@ int ip6_route_add(struct fib6_config *cfg, gfp_t gfp_flags,
 	if (IS_ERR(rt))
 		return PTR_ERR(rt);
 
-	rcu_read_lock();
-
 	err = ip6_route_info_create_nh(rt, cfg, extack);
 	if (err)
-		goto unlock;
+		return err;
 
 	err = __ip6_ins_rt(rt, &cfg->fc_nlinfo, extack);
 	fib6_info_release(rt);
-unlock:
-	rcu_read_unlock();
 
 	return err;
 }
@@ -5534,8 +5537,6 @@ static int ip6_route_multipath_add(struct fib6_config *cfg,
 	if (err)
 		return err;
 
-	rcu_read_lock();
-
 	err = ip6_route_mpath_info_create_nh(&rt6_nh_list, extack);
 	if (err)
 		goto cleanup;
@@ -5627,8 +5628,6 @@ add_errout:
 	}
 
 cleanup:
-	rcu_read_unlock();
-
 	list_for_each_entry_safe(nh, nh_safe, &rt6_nh_list, list) {
 		fib6_info_release(nh->fib6_info);
 		list_del(&nh->list);
@@ -6410,6 +6409,8 @@ void inet6_rt_notify(int event, struct fib6_info *rt, struct nl_info *info,
 	err = -ENOBUFS;
 	seq = info->nlh ? info->nlh->nlmsg_seq : 0;
 
+	rcu_read_lock();
+
 	skb = nlmsg_new(rt6_nlmsg_size(rt), GFP_ATOMIC);
 	if (!skb)
 		goto errout;
@@ -6422,10 +6423,14 @@ void inet6_rt_notify(int event, struct fib6_info *rt, struct nl_info *info,
 		kfree_skb(skb);
 		goto errout;
 	}
+
+	rcu_read_unlock();
+
 	rtnl_notify(skb, net, info->portid, RTNLGRP_IPV6_ROUTE,
 		    info->nlh, GFP_ATOMIC);
 	return;
 errout:
+	rcu_read_unlock();
 	rtnl_set_sk_err(net, RTNLGRP_IPV6_ROUTE, err);
 }
 
-- 
2.51.0


From cefe6e131cc4f032110efe1687295e133f3d5964 Mon Sep 17 00:00:00 2001
From: Kuniyuki Iwashima <kuniyu@amazon.com>
Date: Thu, 15 May 2025 19:27:20 -0700
Subject: [PATCH 03/16] Revert "ipv6: sr: switch to GFP_ATOMIC flag to allocate
 memory during seg6local LWT setup"

The previous patch fixed the same issue mentioned in
commit 14a0087e7236 ("ipv6: sr: switch to GFP_ATOMIC
flag to allocate memory during seg6local LWT setup").

Let's revert it.

Signed-off-by: Kuniyuki Iwashima <kuniyu@amazon.com>
Reviewed-by: Andrea Mayer <andrea.mayer@uniroma2.it>
Link: https://patch.msgid.link/20250516022759.44392-5-kuniyu@amazon.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
---
 net/ipv6/seg6_local.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/net/ipv6/seg6_local.c b/net/ipv6/seg6_local.c
index ee5e448cc7a8..ac1dbd492c22 100644
--- a/net/ipv6/seg6_local.c
+++ b/net/ipv6/seg6_local.c
@@ -1671,7 +1671,7 @@ static int parse_nla_srh(struct nlattr **attrs, struct seg6_local_lwt *slwt,
 	if (!seg6_validate_srh(srh, len, false))
 		return -EINVAL;
 
-	slwt->srh = kmemdup(srh, len, GFP_ATOMIC);
+	slwt->srh = kmemdup(srh, len, GFP_KERNEL);
 	if (!slwt->srh)
 		return -ENOMEM;
 
@@ -1911,7 +1911,7 @@ static int parse_nla_bpf(struct nlattr **attrs, struct seg6_local_lwt *slwt,
 	if (!tb[SEG6_LOCAL_BPF_PROG] || !tb[SEG6_LOCAL_BPF_PROG_NAME])
 		return -EINVAL;
 
-	slwt->bpf.name = nla_memdup(tb[SEG6_LOCAL_BPF_PROG_NAME], GFP_ATOMIC);
+	slwt->bpf.name = nla_memdup(tb[SEG6_LOCAL_BPF_PROG_NAME], GFP_KERNEL);
 	if (!slwt->bpf.name)
 		return -ENOMEM;
 
@@ -1994,7 +1994,7 @@ static int parse_nla_counters(struct nlattr **attrs,
 		return -EINVAL;
 
 	/* counters are always zero initialized */
-	pcounters = seg6_local_alloc_pcpu_counters(GFP_ATOMIC);
+	pcounters = seg6_local_alloc_pcpu_counters(GFP_KERNEL);
 	if (!pcounters)
 		return -ENOMEM;
 
-- 
2.51.0


From 5e4a8cc7beb8567293e6d4230b14e95167759214 Mon Sep 17 00:00:00 2001
From: Kuniyuki Iwashima <kuniyu@amazon.com>
Date: Thu, 15 May 2025 19:27:21 -0700
Subject: [PATCH 04/16] Revert "ipv6: Factorise ip6_route_multipath_add()."

Commit 71c0efb6d12f ("ipv6: Factorise ip6_route_multipath_add().") split
a loop in ip6_route_multipath_add() so that we can put rcu_read_lock()
between ip6_route_info_create() and ip6_route_info_create_nh().

We no longer need to do so as ip6_route_info_create_nh() does not require
RCU now.

Let's revert the commit to simplify the code.

Signed-off-by: Kuniyuki Iwashima <kuniyu@amazon.com>
Link: https://patch.msgid.link/20250516022759.44392-6-kuniyu@amazon.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
---
 net/ipv6/route.c | 193 +++++++++++++++++------------------------------
 1 file changed, 70 insertions(+), 123 deletions(-)

diff --git a/net/ipv6/route.c b/net/ipv6/route.c
index a87091dd06b1..96ae21da9961 100644
--- a/net/ipv6/route.c
+++ b/net/ipv6/route.c
@@ -5335,131 +5335,29 @@ struct rt6_nh {
 	struct fib6_info *fib6_info;
 	struct fib6_config r_cfg;
 	struct list_head list;
-	int weight;
 };
 
-static void ip6_route_mpath_info_cleanup(struct list_head *rt6_nh_list)
+static int ip6_route_info_append(struct list_head *rt6_nh_list,
+				 struct fib6_info *rt,
+				 struct fib6_config *r_cfg)
 {
-	struct rt6_nh *nh, *nh_next;
+	struct rt6_nh *nh;
 
-	list_for_each_entry_safe(nh, nh_next, rt6_nh_list, list) {
-		struct fib6_info *rt = nh->fib6_info;
-
-		if (rt) {
-			free_percpu(rt->fib6_nh->nh_common.nhc_pcpu_rth_output);
-			free_percpu(rt->fib6_nh->rt6i_pcpu);
-			ip_fib_metrics_put(rt->fib6_metrics);
-			kfree(rt);
-		}
-
-		list_del(&nh->list);
-		kfree(nh);
+	list_for_each_entry(nh, rt6_nh_list, list) {
+		/* check if fib6_info already exists */
+		if (rt6_duplicate_nexthop(nh->fib6_info, rt))
+			return -EEXIST;
 	}
-}
-
-static int ip6_route_mpath_info_create(struct list_head *rt6_nh_list,
-				       struct fib6_config *cfg,
-				       struct netlink_ext_ack *extack)
-{
-	struct rtnexthop *rtnh;
-	int remaining;
-	int err;
-
-	remaining = cfg->fc_mp_len;
-	rtnh = (struct rtnexthop *)cfg->fc_mp;
-
-	/* Parse a Multipath Entry and build a list (rt6_nh_list) of
-	 * fib6_info structs per nexthop
-	 */
-	while (rtnh_ok(rtnh, remaining)) {
-		struct fib6_config r_cfg;
-		struct fib6_info *rt;
-		struct rt6_nh *nh;
-		int attrlen;
-
-		nh = kzalloc(sizeof(*nh), GFP_KERNEL);
-		if (!nh) {
-			err = -ENOMEM;
-			goto err;
-		}
 
-		list_add_tail(&nh->list, rt6_nh_list);
-
-		memcpy(&r_cfg, cfg, sizeof(*cfg));
-		if (rtnh->rtnh_ifindex)
-			r_cfg.fc_ifindex = rtnh->rtnh_ifindex;
-
-		attrlen = rtnh_attrlen(rtnh);
-		if (attrlen > 0) {
-			struct nlattr *nla, *attrs = rtnh_attrs(rtnh);
-
-			nla = nla_find(attrs, attrlen, RTA_GATEWAY);
-			if (nla) {
-				r_cfg.fc_gateway = nla_get_in6_addr(nla);
-				r_cfg.fc_flags |= RTF_GATEWAY;
-			}
-
-			r_cfg.fc_encap = nla_find(attrs, attrlen, RTA_ENCAP);
-			nla = nla_find(attrs, attrlen, RTA_ENCAP_TYPE);
-			if (nla)
-				r_cfg.fc_encap_type = nla_get_u16(nla);
-		}
-
-		r_cfg.fc_flags |= (rtnh->rtnh_flags & RTNH_F_ONLINK);
-
-		rt = ip6_route_info_create(&r_cfg, GFP_KERNEL, extack);
-		if (IS_ERR(rt)) {
-			err = PTR_ERR(rt);
-			goto err;
-		}
-
-		nh->fib6_info = rt;
-		nh->weight = rtnh->rtnh_hops + 1;
-		memcpy(&nh->r_cfg, &r_cfg, sizeof(r_cfg));
+	nh = kzalloc(sizeof(*nh), GFP_KERNEL);
+	if (!nh)
+		return -ENOMEM;
 
-		rtnh = rtnh_next(rtnh, &remaining);
-	}
+	nh->fib6_info = rt;
+	memcpy(&nh->r_cfg, r_cfg, sizeof(*r_cfg));
+	list_add_tail(&nh->list, rt6_nh_list);
 
 	return 0;
-err:
-	ip6_route_mpath_info_cleanup(rt6_nh_list);
-	return err;
-}
-
-static int ip6_route_mpath_info_create_nh(struct list_head *rt6_nh_list,
-					  struct netlink_ext_ack *extack)
-{
-	struct rt6_nh *nh, *nh_next, *nh_tmp;
-	LIST_HEAD(tmp);
-	int err;
-
-	list_for_each_entry_safe(nh, nh_next, rt6_nh_list, list) {
-		struct fib6_info *rt = nh->fib6_info;
-
-		err = ip6_route_info_create_nh(rt, &nh->r_cfg, extack);
-		if (err) {
-			nh->fib6_info = NULL;
-			goto err;
-		}
-
-		rt->fib6_nh->fib_nh_weight = nh->weight;
-
-		list_move_tail(&nh->list, &tmp);
-
-		list_for_each_entry(nh_tmp, rt6_nh_list, list) {
-			/* check if fib6_info already exists */
-			if (rt6_duplicate_nexthop(nh_tmp->fib6_info, rt)) {
-				err = -EEXIST;
-				goto err;
-			}
-		}
-	}
-out:
-	list_splice(&tmp, rt6_nh_list);
-	return err;
-err:
-	ip6_route_mpath_info_cleanup(rt6_nh_list);
-	goto out;
 }
 
 static void ip6_route_mpath_notify(struct fib6_info *rt,
@@ -5519,11 +5417,16 @@ static int ip6_route_multipath_add(struct fib6_config *cfg,
 	struct fib6_info *rt_notif = NULL, *rt_last = NULL;
 	struct nl_info *info = &cfg->fc_nlinfo;
 	struct rt6_nh *nh, *nh_safe;
+	struct fib6_config r_cfg;
+	struct rtnexthop *rtnh;
 	LIST_HEAD(rt6_nh_list);
 	struct rt6_nh *err_nh;
+	struct fib6_info *rt;
 	__u16 nlflags;
-	int nhn = 0;
+	int remaining;
+	int attrlen;
 	int replace;
+	int nhn = 0;
 	int err;
 
 	replace = (cfg->fc_nlinfo.nlh &&
@@ -5533,13 +5436,57 @@ static int ip6_route_multipath_add(struct fib6_config *cfg,
 	if (info->nlh && info->nlh->nlmsg_flags & NLM_F_APPEND)
 		nlflags |= NLM_F_APPEND;
 
-	err = ip6_route_mpath_info_create(&rt6_nh_list, cfg, extack);
-	if (err)
-		return err;
+	remaining = cfg->fc_mp_len;
+	rtnh = (struct rtnexthop *)cfg->fc_mp;
 
-	err = ip6_route_mpath_info_create_nh(&rt6_nh_list, extack);
-	if (err)
-		goto cleanup;
+	/* Parse a Multipath Entry and build a list (rt6_nh_list) of
+	 * fib6_info structs per nexthop
+	 */
+	while (rtnh_ok(rtnh, remaining)) {
+		memcpy(&r_cfg, cfg, sizeof(*cfg));
+		if (rtnh->rtnh_ifindex)
+			r_cfg.fc_ifindex = rtnh->rtnh_ifindex;
+
+		attrlen = rtnh_attrlen(rtnh);
+		if (attrlen > 0) {
+			struct nlattr *nla, *attrs = rtnh_attrs(rtnh);
+
+			nla = nla_find(attrs, attrlen, RTA_GATEWAY);
+			if (nla) {
+				r_cfg.fc_gateway = nla_get_in6_addr(nla);
+				r_cfg.fc_flags |= RTF_GATEWAY;
+			}
+
+			r_cfg.fc_encap = nla_find(attrs, attrlen, RTA_ENCAP);
+			nla = nla_find(attrs, attrlen, RTA_ENCAP_TYPE);
+			if (nla)
+				r_cfg.fc_encap_type = nla_get_u16(nla);
+		}
+
+		r_cfg.fc_flags |= (rtnh->rtnh_flags & RTNH_F_ONLINK);
+		rt = ip6_route_info_create(&r_cfg, GFP_KERNEL, extack);
+		if (IS_ERR(rt)) {
+			err = PTR_ERR(rt);
+			rt = NULL;
+			goto cleanup;
+		}
+
+		err = ip6_route_info_create_nh(rt, &r_cfg, extack);
+		if (err) {
+			rt = NULL;
+			goto cleanup;
+		}
+
+		rt->fib6_nh->fib_nh_weight = rtnh->rtnh_hops + 1;
+
+		err = ip6_route_info_append(&rt6_nh_list, rt, &r_cfg);
+		if (err) {
+			fib6_info_release(rt);
+			goto cleanup;
+		}
+
+		rtnh = rtnh_next(rtnh, &remaining);
+	}
 
 	/* for add and replace send one notification with all nexthops.
 	 * Skip the notification in fib6_add_rt2node and send one with
-- 
2.51.0


From d465bd07d16e37cd3aa25539ab187b372853808d Mon Sep 17 00:00:00 2001
From: Kuniyuki Iwashima <kuniyu@amazon.com>
Date: Thu, 15 May 2025 19:27:22 -0700
Subject: [PATCH 05/16] ipv6: Pass gfp_flags down to
 ip6_route_info_create_nh().

Since commit c4837b9853e5 ("ipv6: Split ip6_route_info_create()."),
ip6_route_info_create_nh() uses GFP_ATOMIC as it was expected to be
called under RCU.

Now, we can call it without RCU and use GFP_KERNEL.

Let's pass gfp_flags to ip6_route_info_create_nh().

Signed-off-by: Kuniyuki Iwashima <kuniyu@amazon.com>
Link: https://patch.msgid.link/20250516022759.44392-7-kuniyu@amazon.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
---
 net/ipv6/route.c | 9 +++++----
 1 file changed, 5 insertions(+), 4 deletions(-)

diff --git a/net/ipv6/route.c b/net/ipv6/route.c
index 96ae21da9961..dda913ebd2d3 100644
--- a/net/ipv6/route.c
+++ b/net/ipv6/route.c
@@ -3834,6 +3834,7 @@ err:
 
 static int ip6_route_info_create_nh(struct fib6_info *rt,
 				    struct fib6_config *cfg,
+				    gfp_t gfp_flags,
 				    struct netlink_ext_ack *extack)
 {
 	struct net *net = cfg->fc_nlinfo.nl_net;
@@ -3869,7 +3870,7 @@ static int ip6_route_info_create_nh(struct fib6_info *rt,
 	} else {
 		int addr_type;
 
-		err = fib6_nh_init(net, rt->fib6_nh, cfg, GFP_ATOMIC, extack);
+		err = fib6_nh_init(net, rt->fib6_nh, cfg, gfp_flags, extack);
 		if (err)
 			goto out_release;
 
@@ -3917,7 +3918,7 @@ int ip6_route_add(struct fib6_config *cfg, gfp_t gfp_flags,
 	if (IS_ERR(rt))
 		return PTR_ERR(rt);
 
-	err = ip6_route_info_create_nh(rt, cfg, extack);
+	err = ip6_route_info_create_nh(rt, cfg, gfp_flags, extack);
 	if (err)
 		return err;
 
@@ -4707,7 +4708,7 @@ struct fib6_info *addrconf_f6i_alloc(struct net *net,
 	if (IS_ERR(f6i))
 		return f6i;
 
-	err = ip6_route_info_create_nh(f6i, &cfg, extack);
+	err = ip6_route_info_create_nh(f6i, &cfg, gfp_flags, extack);
 	if (err)
 		return ERR_PTR(err);
 
@@ -5471,7 +5472,7 @@ static int ip6_route_multipath_add(struct fib6_config *cfg,
 			goto cleanup;
 		}
 
-		err = ip6_route_info_create_nh(rt, &r_cfg, extack);
+		err = ip6_route_info_create_nh(rt, &r_cfg, GFP_KERNEL, extack);
 		if (err) {
 			rt = NULL;
 			goto cleanup;
-- 
2.51.0


From 002dba13c824f1cf86f618f0d23d1f0ad3c93bbb Mon Sep 17 00:00:00 2001
From: Kuniyuki Iwashima <kuniyu@amazon.com>
Date: Thu, 15 May 2025 19:27:23 -0700
Subject: [PATCH 06/16] ipv6: Revert two per-cpu var allocation for
 RTM_NEWROUTE.

These two commits preallocated two per-cpu variables in
ip6_route_info_create() as fib_nh_common_init() and fib6_nh_init()
were expected to be called under RCU.

  * commit d27b9c40dbd6 ("ipv6: Preallocate nhc_pcpu_rth_output in
    ip6_route_info_create().")
  * commit 5720a328c3e9 ("ipv6: Preallocate rt->fib6_nh->rt6i_pcpu in
    ip6_route_info_create().")

Now these functions can be called without RCU and can use GFP_KERNEL.

Let's revert the commits.

Signed-off-by: Kuniyuki Iwashima <kuniyu@amazon.com>
Link: https://patch.msgid.link/20250516022759.44392-8-kuniyu@amazon.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
---
 net/ipv4/fib_semantics.c | 10 ++++------
 net/ipv6/route.c         | 34 +++-------------------------------
 2 files changed, 7 insertions(+), 37 deletions(-)

diff --git a/net/ipv4/fib_semantics.c b/net/ipv4/fib_semantics.c
index dabe2b7044ab..d643bd1a0d9d 100644
--- a/net/ipv4/fib_semantics.c
+++ b/net/ipv4/fib_semantics.c
@@ -617,12 +617,10 @@ int fib_nh_common_init(struct net *net, struct fib_nh_common *nhc,
 {
 	int err;
 
-	if (!nhc->nhc_pcpu_rth_output) {
-		nhc->nhc_pcpu_rth_output = alloc_percpu_gfp(struct rtable __rcu *,
-							    gfp_flags);
-		if (!nhc->nhc_pcpu_rth_output)
-			return -ENOMEM;
-	}
+	nhc->nhc_pcpu_rth_output = alloc_percpu_gfp(struct rtable __rcu *,
+						    gfp_flags);
+	if (!nhc->nhc_pcpu_rth_output)
+		return -ENOMEM;
 
 	if (encap) {
 		struct lwtunnel_state *lwtstate;
diff --git a/net/ipv6/route.c b/net/ipv6/route.c
index dda913ebd2d3..0143262094b0 100644
--- a/net/ipv6/route.c
+++ b/net/ipv6/route.c
@@ -3674,12 +3674,10 @@ int fib6_nh_init(struct net *net, struct fib6_nh *fib6_nh,
 		goto out;
 
 pcpu_alloc:
+	fib6_nh->rt6i_pcpu = alloc_percpu_gfp(struct rt6_info *, gfp_flags);
 	if (!fib6_nh->rt6i_pcpu) {
-		fib6_nh->rt6i_pcpu = alloc_percpu_gfp(struct rt6_info *, gfp_flags);
-		if (!fib6_nh->rt6i_pcpu) {
-			err = -ENOMEM;
-			goto out;
-		}
+		err = -ENOMEM;
+		goto out;
 	}
 
 	fib6_nh->fib_nh_dev = dev;
@@ -3739,24 +3737,6 @@ void fib6_nh_release_dsts(struct fib6_nh *fib6_nh)
 	}
 }
 
-static int fib6_nh_prealloc_percpu(struct fib6_nh *fib6_nh, gfp_t gfp_flags)
-{
-	struct fib_nh_common *nhc = &fib6_nh->nh_common;
-
-	fib6_nh->rt6i_pcpu = alloc_percpu_gfp(struct rt6_info *, gfp_flags);
-	if (!fib6_nh->rt6i_pcpu)
-		return -ENOMEM;
-
-	nhc->nhc_pcpu_rth_output = alloc_percpu_gfp(struct rtable __rcu *,
-						    gfp_flags);
-	if (!nhc->nhc_pcpu_rth_output) {
-		free_percpu(fib6_nh->rt6i_pcpu);
-		return -ENOMEM;
-	}
-
-	return 0;
-}
-
 static struct fib6_info *ip6_route_info_create(struct fib6_config *cfg,
 					       gfp_t gfp_flags,
 					       struct netlink_ext_ack *extack)
@@ -3794,12 +3774,6 @@ static struct fib6_info *ip6_route_info_create(struct fib6_config *cfg,
 		goto free;
 	}
 
-	if (!cfg->fc_nh_id) {
-		err = fib6_nh_prealloc_percpu(&rt->fib6_nh[0], gfp_flags);
-		if (err)
-			goto free_metrics;
-	}
-
 	if (cfg->fc_flags & RTF_ADDRCONF)
 		rt->dst_nocount = true;
 
@@ -3824,8 +3798,6 @@ static struct fib6_info *ip6_route_info_create(struct fib6_config *cfg,
 	rt->fib6_src.plen = cfg->fc_src_len;
 #endif
 	return rt;
-free_metrics:
-	ip_fib_metrics_put(rt->fib6_metrics);
 free:
 	kfree(rt);
 err:
-- 
2.51.0


From c52918744ee1e49cea86622a2633b9782446428f Mon Sep 17 00:00:00 2001
From: Lorenzo Bianconi <lorenzo@kernel.org>
Date: Fri, 16 May 2025 09:59:59 +0200
Subject: [PATCH 07/16] net: airoha: npu: Move memory allocation in
 airoha_npu_send_msg() caller

Move ppe_mbox_data struct memory allocation from airoha_npu_send_msg
routine to the caller one. This is a preliminary patch to enable wlan NPU
offloading and flow counter stats support.

Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org>
Reviewed-by: Simon Horman <horms@kernel.org>
Link: https://patch.msgid.link/20250516-airoha-en7581-flowstats-v2-1-06d5fbf28984@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
---
 drivers/net/ethernet/airoha/airoha_npu.c | 126 +++++++++++++----------
 1 file changed, 72 insertions(+), 54 deletions(-)

diff --git a/drivers/net/ethernet/airoha/airoha_npu.c b/drivers/net/ethernet/airoha/airoha_npu.c
index ead0625e781f..51c6fbd7b7b7 100644
--- a/drivers/net/ethernet/airoha/airoha_npu.c
+++ b/drivers/net/ethernet/airoha/airoha_npu.c
@@ -124,17 +124,12 @@ static int airoha_npu_send_msg(struct airoha_npu *npu, int func_id,
 	u16 core = 0; /* FIXME */
 	u32 val, offset = core << 4;
 	dma_addr_t dma_addr;
-	void *addr;
 	int ret;
 
-	addr = kmemdup(p, size, GFP_ATOMIC);
-	if (!addr)
-		return -ENOMEM;
-
-	dma_addr = dma_map_single(npu->dev, addr, size, DMA_TO_DEVICE);
+	dma_addr = dma_map_single(npu->dev, p, size, DMA_TO_DEVICE);
 	ret = dma_mapping_error(npu->dev, dma_addr);
 	if (ret)
-		goto out;
+		return ret;
 
 	spin_lock_bh(&npu->cores[core].lock);
 
@@ -155,8 +150,6 @@ static int airoha_npu_send_msg(struct airoha_npu *npu, int func_id,
 	spin_unlock_bh(&npu->cores[core].lock);
 
 	dma_unmap_single(npu->dev, dma_addr, size, DMA_TO_DEVICE);
-out:
-	kfree(addr);
 
 	return ret;
 }
@@ -261,76 +254,101 @@ static irqreturn_t airoha_npu_wdt_handler(int irq, void *core_instance)
 
 static int airoha_npu_ppe_init(struct airoha_npu *npu)
 {
-	struct ppe_mbox_data ppe_data = {
-		.func_type = NPU_OP_SET,
-		.func_id = PPE_FUNC_SET_WAIT_HWNAT_INIT,
-		.init_info = {
-			.ppe_type = PPE_TYPE_L2B_IPV4_IPV6,
-			.wan_mode = QDMA_WAN_ETHER,
-		},
-	};
+	struct ppe_mbox_data *ppe_data;
+	int err;
+
+	ppe_data = kzalloc(sizeof(*ppe_data), GFP_KERNEL);
+	if (!ppe_data)
+		return -ENOMEM;
 
-	return airoha_npu_send_msg(npu, NPU_FUNC_PPE, &ppe_data,
-				   sizeof(struct ppe_mbox_data));
+	ppe_data->func_type = NPU_OP_SET;
+	ppe_data->func_id = PPE_FUNC_SET_WAIT_HWNAT_INIT;
+	ppe_data->init_info.ppe_type = PPE_TYPE_L2B_IPV4_IPV6;
+	ppe_data->init_info.wan_mode = QDMA_WAN_ETHER;
+
+	err = airoha_npu_send_msg(npu, NPU_FUNC_PPE, ppe_data,
+				  sizeof(*ppe_data));
+	kfree(ppe_data);
+
+	return err;
 }
 
 static int airoha_npu_ppe_deinit(struct airoha_npu *npu)
 {
-	struct ppe_mbox_data ppe_data = {
-		.func_type = NPU_OP_SET,
-		.func_id = PPE_FUNC_SET_WAIT_HWNAT_DEINIT,
-	};
+	struct ppe_mbox_data *ppe_data;
+	int err;
+
+	ppe_data = kzalloc(sizeof(*ppe_data), GFP_KERNEL);
+	if (!ppe_data)
+		return -ENOMEM;
+
+	ppe_data->func_type = NPU_OP_SET;
+	ppe_data->func_id = PPE_FUNC_SET_WAIT_HWNAT_DEINIT;
 
-	return airoha_npu_send_msg(npu, NPU_FUNC_PPE, &ppe_data,
-				   sizeof(struct ppe_mbox_data));
+	err = airoha_npu_send_msg(npu, NPU_FUNC_PPE, ppe_data,
+				  sizeof(*ppe_data));
+	kfree(ppe_data);
+
+	return err;
 }
 
 static int airoha_npu_ppe_flush_sram_entries(struct airoha_npu *npu,
 					     dma_addr_t foe_addr,
 					     int sram_num_entries)
 {
-	struct ppe_mbox_data ppe_data = {
-		.func_type = NPU_OP_SET,
-		.func_id = PPE_FUNC_SET_WAIT_API,
-		.set_info = {
-			.func_id = PPE_SRAM_RESET_VAL,
-			.data = foe_addr,
-			.size = sram_num_entries,
-		},
-	};
+	struct ppe_mbox_data *ppe_data;
+	int err;
+
+	ppe_data = kzalloc(sizeof(*ppe_data), GFP_KERNEL);
+	if (!ppe_data)
+		return -ENOMEM;
+
+	ppe_data->func_type = NPU_OP_SET;
+	ppe_data->func_id = PPE_FUNC_SET_WAIT_API;
+	ppe_data->set_info.func_id = PPE_SRAM_RESET_VAL;
+	ppe_data->set_info.data = foe_addr;
+	ppe_data->set_info.size = sram_num_entries;
+
+	err = airoha_npu_send_msg(npu, NPU_FUNC_PPE, ppe_data,
+				  sizeof(*ppe_data));
+	kfree(ppe_data);
 
-	return airoha_npu_send_msg(npu, NPU_FUNC_PPE, &ppe_data,
-				   sizeof(struct ppe_mbox_data));
+	return err;
 }
 
 static int airoha_npu_foe_commit_entry(struct airoha_npu *npu,
 				       dma_addr_t foe_addr,
 				       u32 entry_size, u32 hash, bool ppe2)
 {
-	struct ppe_mbox_data ppe_data = {
-		.func_type = NPU_OP_SET,
-		.func_id = PPE_FUNC_SET_WAIT_API,
-		.set_info = {
-			.data = foe_addr,
-			.size = entry_size,
-		},
-	};
+	struct ppe_mbox_data *ppe_data;
 	int err;
 
-	ppe_data.set_info.func_id = ppe2 ? PPE2_SRAM_SET_ENTRY
-					 : PPE_SRAM_SET_ENTRY;
+	ppe_data = kzalloc(sizeof(*ppe_data), GFP_ATOMIC);
+	if (!ppe_data)
+		return -ENOMEM;
+
+	ppe_data->func_type = NPU_OP_SET;
+	ppe_data->func_id = PPE_FUNC_SET_WAIT_API;
+	ppe_data->set_info.data = foe_addr;
+	ppe_data->set_info.size = entry_size;
+	ppe_data->set_info.func_id = ppe2 ? PPE2_SRAM_SET_ENTRY
+					  : PPE_SRAM_SET_ENTRY;
 
-	err = airoha_npu_send_msg(npu, NPU_FUNC_PPE, &ppe_data,
-				  sizeof(struct ppe_mbox_data));
+	err = airoha_npu_send_msg(npu, NPU_FUNC_PPE, ppe_data,
+				  sizeof(*ppe_data));
 	if (err)
-		return err;
+		goto out;
+
+	ppe_data->set_info.func_id = PPE_SRAM_SET_VAL;
+	ppe_data->set_info.data = hash;
+	ppe_data->set_info.size = sizeof(u32);
 
-	ppe_data.set_info.func_id = PPE_SRAM_SET_VAL;
-	ppe_data.set_info.data = hash;
-	ppe_data.set_info.size = sizeof(u32);
+	err = airoha_npu_send_msg(npu, NPU_FUNC_PPE, ppe_data,
+				  sizeof(*ppe_data));
+out:
+	kfree(ppe_data);
 
-	return airoha_npu_send_msg(npu, NPU_FUNC_PPE, &ppe_data,
-				   sizeof(struct ppe_mbox_data));
+	return err;
 }
 
 struct airoha_npu *airoha_npu_get(struct device *dev)
-- 
2.51.0


From b81e0f2b58be37628b2e12f8dffdd63c84573e75 Mon Sep 17 00:00:00 2001
From: Lorenzo Bianconi <lorenzo@kernel.org>
Date: Fri, 16 May 2025 10:00:00 +0200
Subject: [PATCH 08/16] net: airoha: Add FLOW_CLS_STATS callback support

Introduce per-flow stats accounting to the flowtable hw offload in
the airoha_eth driver. Flow stats are split in the PPE and NPU modules:
- PPE: accounts for high 32bit of per-flow stats
- NPU: accounts for low 32bit of per-flow stats

FLOW_CLS_STATS can be enabled or disabled at compile time.

Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org>
Reviewed-by: Simon Horman <horms@kernel.org>
Link: https://patch.msgid.link/20250516-airoha-en7581-flowstats-v2-2-06d5fbf28984@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
---
 drivers/net/ethernet/airoha/Kconfig           |   7 +
 drivers/net/ethernet/airoha/airoha_eth.h      |  33 +++
 drivers/net/ethernet/airoha/airoha_npu.c      |  52 +++-
 drivers/net/ethernet/airoha/airoha_npu.h      |   4 +-
 drivers/net/ethernet/airoha/airoha_ppe.c      | 269 ++++++++++++++++--
 .../net/ethernet/airoha/airoha_ppe_debugfs.c  |   9 +-
 6 files changed, 354 insertions(+), 20 deletions(-)

diff --git a/drivers/net/ethernet/airoha/Kconfig b/drivers/net/ethernet/airoha/Kconfig
index 1a4cf6a259f6..ad3ce501e7a5 100644
--- a/drivers/net/ethernet/airoha/Kconfig
+++ b/drivers/net/ethernet/airoha/Kconfig
@@ -24,4 +24,11 @@ config NET_AIROHA
 	  This driver supports the gigabit ethernet MACs in the
 	  Airoha SoC family.
 
+config NET_AIROHA_FLOW_STATS
+	default y
+	bool "Airoha flow stats"
+	depends on NET_AIROHA && NET_AIROHA_NPU
+	help
+	  Enable Aiorha flowtable statistic counters.
+
 endif #NET_VENDOR_AIROHA
diff --git a/drivers/net/ethernet/airoha/airoha_eth.h b/drivers/net/ethernet/airoha/airoha_eth.h
index 53f39083a8b0..531a3c49c156 100644
--- a/drivers/net/ethernet/airoha/airoha_eth.h
+++ b/drivers/net/ethernet/airoha/airoha_eth.h
@@ -50,6 +50,14 @@
 #define PPE_NUM				2
 #define PPE1_SRAM_NUM_ENTRIES		(8 * 1024)
 #define PPE_SRAM_NUM_ENTRIES		(2 * PPE1_SRAM_NUM_ENTRIES)
+#ifdef CONFIG_NET_AIROHA_FLOW_STATS
+#define PPE1_STATS_NUM_ENTRIES		(4 * 1024)
+#else
+#define PPE1_STATS_NUM_ENTRIES		0
+#endif /* CONFIG_NET_AIROHA_FLOW_STATS */
+#define PPE_STATS_NUM_ENTRIES		(2 * PPE1_STATS_NUM_ENTRIES)
+#define PPE1_SRAM_NUM_DATA_ENTRIES	(PPE1_SRAM_NUM_ENTRIES - PPE1_STATS_NUM_ENTRIES)
+#define PPE_SRAM_NUM_DATA_ENTRIES	(2 * PPE1_SRAM_NUM_DATA_ENTRIES)
 #define PPE_DRAM_NUM_ENTRIES		(16 * 1024)
 #define PPE_NUM_ENTRIES			(PPE_SRAM_NUM_ENTRIES + PPE_DRAM_NUM_ENTRIES)
 #define PPE_HASH_MASK			(PPE_NUM_ENTRIES - 1)
@@ -261,6 +269,8 @@ struct airoha_foe_mac_info {
 
 	u16 pppoe_id;
 	u16 src_mac_lo;
+
+	u32 meter;
 };
 
 #define AIROHA_FOE_IB1_UNBIND_PREBIND		BIT(24)
@@ -296,6 +306,11 @@ struct airoha_foe_mac_info {
 #define AIROHA_FOE_TUNNEL			BIT(6)
 #define AIROHA_FOE_TUNNEL_ID			GENMASK(5, 0)
 
+#define AIROHA_FOE_TUNNEL_MTU			GENMASK(31, 16)
+#define AIROHA_FOE_ACNT_GRP3			GENMASK(15, 9)
+#define AIROHA_FOE_METER_GRP3			GENMASK(8, 5)
+#define AIROHA_FOE_METER_GRP2			GENMASK(4, 0)
+
 struct airoha_foe_bridge {
 	u32 dest_mac_hi;
 
@@ -379,6 +394,8 @@ struct airoha_foe_ipv6 {
 	u32 ib2;
 
 	struct airoha_foe_mac_info_common l2;
+
+	u32 meter;
 };
 
 struct airoha_foe_entry {
@@ -397,6 +414,16 @@ struct airoha_foe_entry {
 	};
 };
 
+struct airoha_foe_stats {
+	u32 bytes;
+	u32 packets;
+};
+
+struct airoha_foe_stats64 {
+	u64 bytes;
+	u64 packets;
+};
+
 struct airoha_flow_data {
 	struct ethhdr eth;
 
@@ -447,6 +474,7 @@ struct airoha_flow_table_entry {
 	struct hlist_node l2_subflow_node; /* PPE L2 subflow entry */
 	u32 hash;
 
+	struct airoha_foe_stats64 stats;
 	enum airoha_flow_entry_type type;
 
 	struct rhash_head node;
@@ -523,6 +551,9 @@ struct airoha_ppe {
 	struct hlist_head *foe_flow;
 	u16 foe_check_time[PPE_NUM_ENTRIES];
 
+	struct airoha_foe_stats *foe_stats;
+	dma_addr_t foe_stats_dma;
+
 	struct dentry *debugfs_dir;
 };
 
@@ -582,6 +613,8 @@ int airoha_ppe_init(struct airoha_eth *eth);
 void airoha_ppe_deinit(struct airoha_eth *eth);
 struct airoha_foe_entry *airoha_ppe_foe_get_entry(struct airoha_ppe *ppe,
 						  u32 hash);
+void airoha_ppe_foe_entry_get_stats(struct airoha_ppe *ppe, u32 hash,
+				    struct airoha_foe_stats64 *stats);
 
 #ifdef CONFIG_DEBUG_FS
 int airoha_ppe_debugfs_init(struct airoha_ppe *ppe);
diff --git a/drivers/net/ethernet/airoha/airoha_npu.c b/drivers/net/ethernet/airoha/airoha_npu.c
index 51c6fbd7b7b7..0e5b8c21b9aa 100644
--- a/drivers/net/ethernet/airoha/airoha_npu.c
+++ b/drivers/net/ethernet/airoha/airoha_npu.c
@@ -12,6 +12,7 @@
 #include <linux/of_reserved_mem.h>
 #include <linux/regmap.h>
 
+#include "airoha_eth.h"
 #include "airoha_npu.h"
 
 #define NPU_EN7581_FIRMWARE_DATA		"airoha/en7581_npu_data.bin"
@@ -72,6 +73,7 @@ enum {
 	PPE_FUNC_SET_WAIT_HWNAT_INIT,
 	PPE_FUNC_SET_WAIT_HWNAT_DEINIT,
 	PPE_FUNC_SET_WAIT_API,
+	PPE_FUNC_SET_WAIT_FLOW_STATS_SETUP,
 };
 
 enum {
@@ -115,6 +117,10 @@ struct ppe_mbox_data {
 			u32 size;
 			u32 data;
 		} set_info;
+		struct {
+			u32 npu_stats_addr;
+			u32 foe_stats_addr;
+		} stats_info;
 	};
 };
 
@@ -351,7 +357,40 @@ out:
 	return err;
 }
 
-struct airoha_npu *airoha_npu_get(struct device *dev)
+static int airoha_npu_stats_setup(struct airoha_npu *npu,
+				  dma_addr_t foe_stats_addr)
+{
+	int err, size = PPE_STATS_NUM_ENTRIES * sizeof(*npu->stats);
+	struct ppe_mbox_data *ppe_data;
+
+	if (!size) /* flow stats are disabled */
+		return 0;
+
+	ppe_data = kzalloc(sizeof(*ppe_data), GFP_ATOMIC);
+	if (!ppe_data)
+		return -ENOMEM;
+
+	ppe_data->func_type = NPU_OP_SET;
+	ppe_data->func_id = PPE_FUNC_SET_WAIT_FLOW_STATS_SETUP;
+	ppe_data->stats_info.foe_stats_addr = foe_stats_addr;
+
+	err = airoha_npu_send_msg(npu, NPU_FUNC_PPE, ppe_data,
+				  sizeof(*ppe_data));
+	if (err)
+		goto out;
+
+	npu->stats = devm_ioremap(npu->dev,
+				  ppe_data->stats_info.npu_stats_addr,
+				  size);
+	if (!npu->stats)
+		err = -ENOMEM;
+out:
+	kfree(ppe_data);
+
+	return err;
+}
+
+struct airoha_npu *airoha_npu_get(struct device *dev, dma_addr_t *stats_addr)
 {
 	struct platform_device *pdev;
 	struct device_node *np;
@@ -389,6 +428,17 @@ struct airoha_npu *airoha_npu_get(struct device *dev)
 		goto error_module_put;
 	}
 
+	if (stats_addr) {
+		int err;
+
+		err = airoha_npu_stats_setup(npu, *stats_addr);
+		if (err) {
+			dev_err(dev, "failed to allocate npu stats buffer\n");
+			npu = ERR_PTR(err);
+			goto error_module_put;
+		}
+	}
+
 	return npu;
 
 error_module_put:
diff --git a/drivers/net/ethernet/airoha/airoha_npu.h b/drivers/net/ethernet/airoha/airoha_npu.h
index a2b8ae4d9473..98ec3be74ce4 100644
--- a/drivers/net/ethernet/airoha/airoha_npu.h
+++ b/drivers/net/ethernet/airoha/airoha_npu.h
@@ -17,6 +17,8 @@ struct airoha_npu {
 		struct work_struct wdt_work;
 	} cores[NPU_NUM_CORES];
 
+	struct airoha_foe_stats __iomem *stats;
+
 	struct {
 		int (*ppe_init)(struct airoha_npu *npu);
 		int (*ppe_deinit)(struct airoha_npu *npu);
@@ -30,5 +32,5 @@ struct airoha_npu {
 	} ops;
 };
 
-struct airoha_npu *airoha_npu_get(struct device *dev);
+struct airoha_npu *airoha_npu_get(struct device *dev, dma_addr_t *stats_addr);
 void airoha_npu_put(struct airoha_npu *npu);
diff --git a/drivers/net/ethernet/airoha/airoha_ppe.c b/drivers/net/ethernet/airoha/airoha_ppe.c
index 6e9787c2843b..70598ba9f5ad 100644
--- a/drivers/net/ethernet/airoha/airoha_ppe.c
+++ b/drivers/net/ethernet/airoha/airoha_ppe.c
@@ -102,7 +102,7 @@ static void airoha_ppe_hw_init(struct airoha_ppe *ppe)
 
 	if (airoha_ppe2_is_enabled(eth)) {
 		sram_num_entries =
-			PPE_RAM_NUM_ENTRIES_SHIFT(PPE1_SRAM_NUM_ENTRIES);
+			PPE_RAM_NUM_ENTRIES_SHIFT(PPE1_SRAM_NUM_DATA_ENTRIES);
 		airoha_fe_rmw(eth, REG_PPE_TB_CFG(0),
 			      PPE_SRAM_TB_NUM_ENTRY_MASK |
 			      PPE_DRAM_TB_NUM_ENTRY_MASK,
@@ -119,7 +119,7 @@ static void airoha_ppe_hw_init(struct airoha_ppe *ppe)
 					 dram_num_entries));
 	} else {
 		sram_num_entries =
-			PPE_RAM_NUM_ENTRIES_SHIFT(PPE_SRAM_NUM_ENTRIES);
+			PPE_RAM_NUM_ENTRIES_SHIFT(PPE_SRAM_NUM_DATA_ENTRIES);
 		airoha_fe_rmw(eth, REG_PPE_TB_CFG(0),
 			      PPE_SRAM_TB_NUM_ENTRY_MASK |
 			      PPE_DRAM_TB_NUM_ENTRY_MASK,
@@ -417,6 +417,77 @@ static u32 airoha_ppe_foe_get_entry_hash(struct airoha_foe_entry *hwe)
 	return hash;
 }
 
+static u32 airoha_ppe_foe_get_flow_stats_index(struct airoha_ppe *ppe, u32 hash)
+{
+	if (!airoha_ppe2_is_enabled(ppe->eth))
+		return hash;
+
+	return hash >= PPE_STATS_NUM_ENTRIES ? hash - PPE1_STATS_NUM_ENTRIES
+					     : hash;
+}
+
+static void airoha_ppe_foe_flow_stat_entry_reset(struct airoha_ppe *ppe,
+						 struct airoha_npu *npu,
+						 int index)
+{
+	memset_io(&npu->stats[index], 0, sizeof(*npu->stats));
+	memset(&ppe->foe_stats[index], 0, sizeof(*ppe->foe_stats));
+}
+
+static void airoha_ppe_foe_flow_stats_reset(struct airoha_ppe *ppe,
+					    struct airoha_npu *npu)
+{
+	int i;
+
+	for (i = 0; i < PPE_STATS_NUM_ENTRIES; i++)
+		airoha_ppe_foe_flow_stat_entry_reset(ppe, npu, i);
+}
+
+static void airoha_ppe_foe_flow_stats_update(struct airoha_ppe *ppe,
+					     struct airoha_npu *npu,
+					     struct airoha_foe_entry *hwe,
+					     u32 hash)
+{
+	int type = FIELD_GET(AIROHA_FOE_IB1_BIND_PACKET_TYPE, hwe->ib1);
+	u32 index, pse_port, val, *data, *ib2, *meter;
+	u8 nbq;
+
+	index = airoha_ppe_foe_get_flow_stats_index(ppe, hash);
+	if (index >= PPE_STATS_NUM_ENTRIES)
+		return;
+
+	if (type == PPE_PKT_TYPE_BRIDGE) {
+		data = &hwe->bridge.data;
+		ib2 = &hwe->bridge.ib2;
+		meter = &hwe->bridge.l2.meter;
+	} else if (type >= PPE_PKT_TYPE_IPV6_ROUTE_3T) {
+		data = &hwe->ipv6.data;
+		ib2 = &hwe->ipv6.ib2;
+		meter = &hwe->ipv6.meter;
+	} else {
+		data = &hwe->ipv4.data;
+		ib2 = &hwe->ipv4.ib2;
+		meter = &hwe->ipv4.l2.meter;
+	}
+
+	airoha_ppe_foe_flow_stat_entry_reset(ppe, npu, index);
+
+	val = FIELD_GET(AIROHA_FOE_CHANNEL | AIROHA_FOE_QID, *data);
+	*data = (*data & ~AIROHA_FOE_ACTDP) |
+		FIELD_PREP(AIROHA_FOE_ACTDP, val);
+
+	val = *ib2 & (AIROHA_FOE_IB2_NBQ | AIROHA_FOE_IB2_PSE_PORT |
+		      AIROHA_FOE_IB2_PSE_QOS | AIROHA_FOE_IB2_FAST_PATH);
+	*meter |= FIELD_PREP(AIROHA_FOE_TUNNEL_MTU, val);
+
+	pse_port = FIELD_GET(AIROHA_FOE_IB2_PSE_PORT, *ib2);
+	nbq = pse_port == 1 ? 6 : 5;
+	*ib2 &= ~(AIROHA_FOE_IB2_NBQ | AIROHA_FOE_IB2_PSE_PORT |
+		  AIROHA_FOE_IB2_PSE_QOS);
+	*ib2 |= FIELD_PREP(AIROHA_FOE_IB2_PSE_PORT, 6) |
+		FIELD_PREP(AIROHA_FOE_IB2_NBQ, nbq);
+}
+
 struct airoha_foe_entry *airoha_ppe_foe_get_entry(struct airoha_ppe *ppe,
 						  u32 hash)
 {
@@ -470,6 +541,8 @@ static int airoha_ppe_foe_commit_entry(struct airoha_ppe *ppe,
 	struct airoha_foe_entry *hwe = ppe->foe + hash * sizeof(*hwe);
 	u32 ts = airoha_ppe_get_timestamp(ppe);
 	struct airoha_eth *eth = ppe->eth;
+	struct airoha_npu *npu;
+	int err = 0;
 
 	memcpy(&hwe->d, &e->d, sizeof(*hwe) - sizeof(hwe->ib1));
 	wmb();
@@ -478,25 +551,28 @@ static int airoha_ppe_foe_commit_entry(struct airoha_ppe *ppe,
 	e->ib1 |= FIELD_PREP(AIROHA_FOE_IB1_BIND_TIMESTAMP, ts);
 	hwe->ib1 = e->ib1;
 
+	rcu_read_lock();
+
+	npu = rcu_dereference(eth->npu);
+	if (!npu) {
+		err = -ENODEV;
+		goto unlock;
+	}
+
+	airoha_ppe_foe_flow_stats_update(ppe, npu, hwe, hash);
+
 	if (hash < PPE_SRAM_NUM_ENTRIES) {
 		dma_addr_t addr = ppe->foe_dma + hash * sizeof(*hwe);
 		bool ppe2 = airoha_ppe2_is_enabled(eth) &&
 			    hash >= PPE1_SRAM_NUM_ENTRIES;
-		struct airoha_npu *npu;
-		int err = -ENODEV;
 
-		rcu_read_lock();
-		npu = rcu_dereference(eth->npu);
-		if (npu)
-			err = npu->ops.ppe_foe_commit_entry(npu, addr,
-							    sizeof(*hwe), hash,
-							    ppe2);
-		rcu_read_unlock();
-
-		return err;
+		err = npu->ops.ppe_foe_commit_entry(npu, addr, sizeof(*hwe),
+						    hash, ppe2);
 	}
+unlock:
+	rcu_read_unlock();
 
-	return 0;
+	return err;
 }
 
 static void airoha_ppe_foe_remove_flow(struct airoha_ppe *ppe,
@@ -582,6 +658,7 @@ airoha_ppe_foe_commit_subflow_entry(struct airoha_ppe *ppe,
 		l2->common.etype = ETH_P_IPV6;
 
 	hwe.bridge.ib2 = e->data.bridge.ib2;
+	hwe.bridge.data = e->data.bridge.data;
 	airoha_ppe_foe_commit_entry(ppe, &hwe, hash);
 
 	return 0;
@@ -681,6 +758,98 @@ static int airoha_ppe_foe_flow_commit_entry(struct airoha_ppe *ppe,
 	return 0;
 }
 
+static int airoha_ppe_get_entry_idle_time(struct airoha_ppe *ppe, u32 ib1)
+{
+	u32 state = FIELD_GET(AIROHA_FOE_IB1_BIND_STATE, ib1);
+	u32 ts, ts_mask, now = airoha_ppe_get_timestamp(ppe);
+	int idle;
+
+	if (state == AIROHA_FOE_STATE_BIND) {
+		ts = FIELD_GET(AIROHA_FOE_IB1_BIND_TIMESTAMP, ib1);
+		ts_mask = AIROHA_FOE_IB1_BIND_TIMESTAMP;
+	} else {
+		ts = FIELD_GET(AIROHA_FOE_IB1_UNBIND_TIMESTAMP, ib1);
+		now = FIELD_GET(AIROHA_FOE_IB1_UNBIND_TIMESTAMP, now);
+		ts_mask = AIROHA_FOE_IB1_UNBIND_TIMESTAMP;
+	}
+	idle = now - ts;
+
+	return idle < 0 ? idle + ts_mask + 1 : idle;
+}
+
+static void
+airoha_ppe_foe_flow_l2_entry_update(struct airoha_ppe *ppe,
+				    struct airoha_flow_table_entry *e)
+{
+	int min_idle = airoha_ppe_get_entry_idle_time(ppe, e->data.ib1);
+	struct airoha_flow_table_entry *iter;
+	struct hlist_node *n;
+
+	lockdep_assert_held(&ppe_lock);
+
+	hlist_for_each_entry_safe(iter, n, &e->l2_flows, l2_subflow_node) {
+		struct airoha_foe_entry *hwe;
+		u32 ib1, state;
+		int idle;
+
+		hwe = airoha_ppe_foe_get_entry(ppe, iter->hash);
+		ib1 = READ_ONCE(hwe->ib1);
+
+		state = FIELD_GET(AIROHA_FOE_IB1_BIND_STATE, ib1);
+		if (state != AIROHA_FOE_STATE_BIND) {
+			iter->hash = 0xffff;
+			airoha_ppe_foe_remove_flow(ppe, iter);
+			continue;
+		}
+
+		idle = airoha_ppe_get_entry_idle_time(ppe, ib1);
+		if (idle >= min_idle)
+			continue;
+
+		min_idle = idle;
+		e->data.ib1 &= ~AIROHA_FOE_IB1_BIND_TIMESTAMP;
+		e->data.ib1 |= ib1 & AIROHA_FOE_IB1_BIND_TIMESTAMP;
+	}
+}
+
+static void airoha_ppe_foe_flow_entry_update(struct airoha_ppe *ppe,
+					     struct airoha_flow_table_entry *e)
+{
+	struct airoha_foe_entry *hwe_p, hwe = {};
+
+	spin_lock_bh(&ppe_lock);
+
+	if (e->type == FLOW_TYPE_L2) {
+		airoha_ppe_foe_flow_l2_entry_update(ppe, e);
+		goto unlock;
+	}
+
+	if (e->hash == 0xffff)
+		goto unlock;
+
+	hwe_p = airoha_ppe_foe_get_entry(ppe, e->hash);
+	if (!hwe_p)
+		goto unlock;
+
+	memcpy(&hwe, hwe_p, sizeof(*hwe_p));
+	if (!airoha_ppe_foe_compare_entry(e, &hwe)) {
+		e->hash = 0xffff;
+		goto unlock;
+	}
+
+	e->data.ib1 = hwe.ib1;
+unlock:
+	spin_unlock_bh(&ppe_lock);
+}
+
+static int airoha_ppe_entry_idle_time(struct airoha_ppe *ppe,
+				      struct airoha_flow_table_entry *e)
+{
+	airoha_ppe_foe_flow_entry_update(ppe, e);
+
+	return airoha_ppe_get_entry_idle_time(ppe, e->data.ib1);
+}
+
 static int airoha_ppe_flow_offload_replace(struct airoha_gdm_port *port,
 					   struct flow_cls_offload *f)
 {
@@ -896,6 +1065,60 @@ static int airoha_ppe_flow_offload_destroy(struct airoha_gdm_port *port,
 	return 0;
 }
 
+void airoha_ppe_foe_entry_get_stats(struct airoha_ppe *ppe, u32 hash,
+				    struct airoha_foe_stats64 *stats)
+{
+	u32 index = airoha_ppe_foe_get_flow_stats_index(ppe, hash);
+	struct airoha_eth *eth = ppe->eth;
+	struct airoha_npu *npu;
+
+	if (index >= PPE_STATS_NUM_ENTRIES)
+		return;
+
+	rcu_read_lock();
+
+	npu = rcu_dereference(eth->npu);
+	if (npu) {
+		u64 packets = ppe->foe_stats[index].packets;
+		u64 bytes = ppe->foe_stats[index].bytes;
+		struct airoha_foe_stats npu_stats;
+
+		memcpy_fromio(&npu_stats, &npu->stats[index],
+			      sizeof(*npu->stats));
+		stats->packets = packets << 32 | npu_stats.packets;
+		stats->bytes = bytes << 32 | npu_stats.bytes;
+	}
+
+	rcu_read_unlock();
+}
+
+static int airoha_ppe_flow_offload_stats(struct airoha_gdm_port *port,
+					 struct flow_cls_offload *f)
+{
+	struct airoha_eth *eth = port->qdma->eth;
+	struct airoha_flow_table_entry *e;
+	u32 idle;
+
+	e = rhashtable_lookup(&eth->flow_table, &f->cookie,
+			      airoha_flow_table_params);
+	if (!e)
+		return -ENOENT;
+
+	idle = airoha_ppe_entry_idle_time(eth->ppe, e);
+	f->stats.lastused = jiffies - idle * HZ;
+
+	if (e->hash != 0xffff) {
+		struct airoha_foe_stats64 stats = {};
+
+		airoha_ppe_foe_entry_get_stats(eth->ppe, e->hash, &stats);
+		f->stats.pkts += (stats.packets - e->stats.packets);
+		f->stats.bytes += (stats.bytes - e->stats.bytes);
+		e->stats = stats;
+	}
+
+	return 0;
+}
+
 static int airoha_ppe_flow_offload_cmd(struct airoha_gdm_port *port,
 				       struct flow_cls_offload *f)
 {
@@ -904,6 +1127,8 @@ static int airoha_ppe_flow_offload_cmd(struct airoha_gdm_port *port,
 		return airoha_ppe_flow_offload_replace(port, f);
 	case FLOW_CLS_DESTROY:
 		return airoha_ppe_flow_offload_destroy(port, f);
+	case FLOW_CLS_STATS:
+		return airoha_ppe_flow_offload_stats(port, f);
 	default:
 		break;
 	}
@@ -929,11 +1154,12 @@ static int airoha_ppe_flush_sram_entries(struct airoha_ppe *ppe,
 
 static struct airoha_npu *airoha_ppe_npu_get(struct airoha_eth *eth)
 {
-	struct airoha_npu *npu = airoha_npu_get(eth->dev);
+	struct airoha_npu *npu = airoha_npu_get(eth->dev,
+						&eth->ppe->foe_stats_dma);
 
 	if (IS_ERR(npu)) {
 		request_module("airoha-npu");
-		npu = airoha_npu_get(eth->dev);
+		npu = airoha_npu_get(eth->dev, &eth->ppe->foe_stats_dma);
 	}
 
 	return npu;
@@ -956,6 +1182,8 @@ static int airoha_ppe_offload_setup(struct airoha_eth *eth)
 	if (err)
 		goto error_npu_put;
 
+	airoha_ppe_foe_flow_stats_reset(eth->ppe, npu);
+
 	rcu_assign_pointer(eth->npu, npu);
 	synchronize_rcu();
 
@@ -1027,6 +1255,15 @@ int airoha_ppe_init(struct airoha_eth *eth)
 	if (!ppe->foe_flow)
 		return -ENOMEM;
 
+	foe_size = PPE_STATS_NUM_ENTRIES * sizeof(*ppe->foe_stats);
+	if (foe_size) {
+		ppe->foe_stats = dmam_alloc_coherent(eth->dev, foe_size,
+						     &ppe->foe_stats_dma,
+						     GFP_KERNEL);
+		if (!ppe->foe_stats)
+			return -ENOMEM;
+	}
+
 	err = rhashtable_init(&eth->flow_table, &airoha_flow_table_params);
 	if (err)
 		return err;
diff --git a/drivers/net/ethernet/airoha/airoha_ppe_debugfs.c b/drivers/net/ethernet/airoha/airoha_ppe_debugfs.c
index 3cdc6fd53fc7..05a756233f6a 100644
--- a/drivers/net/ethernet/airoha/airoha_ppe_debugfs.c
+++ b/drivers/net/ethernet/airoha/airoha_ppe_debugfs.c
@@ -61,6 +61,7 @@ static int airoha_ppe_debugfs_foe_show(struct seq_file *m, void *private,
 		u16 *src_port = NULL, *dest_port = NULL;
 		struct airoha_foe_mac_info_common *l2;
 		unsigned char h_source[ETH_ALEN] = {};
+		struct airoha_foe_stats64 stats = {};
 		unsigned char h_dest[ETH_ALEN];
 		struct airoha_foe_entry *hwe;
 		u32 type, state, ib2, data;
@@ -144,14 +145,18 @@ static int airoha_ppe_debugfs_foe_show(struct seq_file *m, void *private,
 				cpu_to_be16(hwe->ipv4.l2.src_mac_lo);
 		}
 
+		airoha_ppe_foe_entry_get_stats(ppe, i, &stats);
+
 		*((__be32 *)h_dest) = cpu_to_be32(l2->dest_mac_hi);
 		*((__be16 *)&h_dest[4]) = cpu_to_be16(l2->dest_mac_lo);
 		*((__be32 *)h_source) = cpu_to_be32(l2->src_mac_hi);
 
 		seq_printf(m, " eth=%pM->%pM etype=%04x data=%08x"
-			      " vlan=%d,%d ib1=%08x ib2=%08x\n",
+			      " vlan=%d,%d ib1=%08x ib2=%08x"
+			      " packets=%llu bytes=%llu\n",
 			   h_source, h_dest, l2->etype, data,
-			   l2->vlan1, l2->vlan2, hwe->ib1, ib2);
+			   l2->vlan1, l2->vlan2, hwe->ib1, ib2,
+			   stats.packets, stats.bytes);
 	}
 
 	return 0;
-- 
2.51.0


From a98326c151ea3d92e9496858cc2dacccd0870941 Mon Sep 17 00:00:00 2001
From: Lorenzo Bianconi <lorenzo@kernel.org>
Date: Fri, 16 May 2025 10:00:01 +0200
Subject: [PATCH 09/16] net: airoha: ppe: Disable packet keepalive

Since netfilter flowtable entries are now refreshed by flow-stats
polling, we can disable hw packet keepalive used to periodically send
packets belonging to offloaded flows to the kernel in order to refresh
flowtable entries.

Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org>
Reviewed-by: Simon Horman <horms@kernel.org>
Link: https://patch.msgid.link/20250516-airoha-en7581-flowstats-v2-3-06d5fbf28984@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
---
 drivers/net/ethernet/airoha/airoha_ppe.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/drivers/net/ethernet/airoha/airoha_ppe.c b/drivers/net/ethernet/airoha/airoha_ppe.c
index 70598ba9f5ad..2d273937f19c 100644
--- a/drivers/net/ethernet/airoha/airoha_ppe.c
+++ b/drivers/net/ethernet/airoha/airoha_ppe.c
@@ -84,6 +84,7 @@ static void airoha_ppe_hw_init(struct airoha_ppe *ppe)
 
 		airoha_fe_rmw(eth, REG_PPE_TB_CFG(i),
 			      PPE_TB_CFG_SEARCH_MISS_MASK |
+			      PPE_TB_CFG_KEEPALIVE_MASK |
 			      PPE_TB_ENTRY_SIZE_MASK,
 			      FIELD_PREP(PPE_TB_CFG_SEARCH_MISS_MASK, 3) |
 			      FIELD_PREP(PPE_TB_ENTRY_SIZE_MASK, 0));
-- 
2.51.0


From 8b8762eeec59b959fbca60afffe21265bce67168 Mon Sep 17 00:00:00 2001
From: Jakub Kicinski <kuba@kernel.org>
Date: Tue, 20 May 2025 09:19:05 -0700
Subject: [PATCH 10/16] tools: ynl-gen: add makefile deps for neigh

Kory is reporting build issues after recent additions to YNL
if the system headers are old.

Link: https://lore.kernel.org/20250519164949.597d6e92@kmaincent-XPS-13-7390
Reported-by: Kory Maincent <kory.maincent@bootlin.com>
Fixes: 0939a418b3b0 ("tools: ynl: submsg: reverse parse / error reporting")
Tested-by: Kory Maincent <kory.maincent@bootlin.com>
Link: https://patch.msgid.link/20250520161916.413298-2-kuba@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
---
 include/uapi/linux/neighbour.h | 4 ++--
 tools/net/ynl/Makefile.deps    | 3 ++-
 2 files changed, 4 insertions(+), 3 deletions(-)

diff --git a/include/uapi/linux/neighbour.h b/include/uapi/linux/neighbour.h
index 5e67a7eaf4a7..b851c36ad25d 100644
--- a/include/uapi/linux/neighbour.h
+++ b/include/uapi/linux/neighbour.h
@@ -1,6 +1,6 @@
 /* SPDX-License-Identifier: GPL-2.0 WITH Linux-syscall-note */
-#ifndef __LINUX_NEIGHBOUR_H
-#define __LINUX_NEIGHBOUR_H
+#ifndef _UAPI__LINUX_NEIGHBOUR_H
+#define _UAPI__LINUX_NEIGHBOUR_H
 
 #include <linux/types.h>
 #include <linux/netlink.h>
diff --git a/tools/net/ynl/Makefile.deps b/tools/net/ynl/Makefile.deps
index e5a5cb1b2cff..8c378356fc87 100644
--- a/tools/net/ynl/Makefile.deps
+++ b/tools/net/ynl/Makefile.deps
@@ -35,7 +35,8 @@ CFLAGS_rt-addr:=$(call get_hdr_inc,__LINUX_RTNETLINK_H,rtnetlink.h) \
 	$(call get_hdr_inc,__LINUX_IF_ADDR_H,if_addr.h)
 CFLAGS_rt-link:=$(call get_hdr_inc,__LINUX_RTNETLINK_H,rtnetlink.h) \
 	$(call get_hdr_inc,_LINUX_IF_LINK_H,if_link.h)
-CFLAGS_rt-neigh:=$(call get_hdr_inc,__LINUX_RTNETLINK_H,rtnetlink.h)
+CFLAGS_rt-neigh:=$(call get_hdr_inc,__LINUX_RTNETLINK_H,rtnetlink.h) \
+	$(call get_hdr_inc,__LINUX_NEIGHBOUR_H,neighbour.h)
 CFLAGS_rt-route:=$(call get_hdr_inc,__LINUX_RTNETLINK_H,rtnetlink.h)
 CFLAGS_rt-rule:=$(call get_hdr_inc,__LINUX_FIB_RULES_H,fib_rules.h)
 CFLAGS_tcp_metrics:=$(call get_hdr_inc,_LINUX_TCP_METRICS_H,tcp_metrics.h)
-- 
2.51.0


From e9033a846eb9a8ca21dee880b9ac2f5988ceb5f0 Mon Sep 17 00:00:00 2001
From: Jakub Kicinski <kuba@kernel.org>
Date: Tue, 20 May 2025 09:19:06 -0700
Subject: [PATCH 11/16] netlink: specs: tc: remove duplicate nests

tc-act-stats-attrs and tca-stats-attrs are almost identical.
The only difference is that the latter has sub-message decoding
for app, rather than declaring it as a binary attr.

tc-act-police-attrs and tc-police-attrs are identical but for
the TODO annotations.

Reviewed-by: Donald Hunter <donald.hunter@gmail.com>
Link: https://patch.msgid.link/20250520161916.413298-3-kuba@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
---
 Documentation/netlink/specs/tc.yaml | 78 ++---------------------------
 1 file changed, 4 insertions(+), 74 deletions(-)

diff --git a/Documentation/netlink/specs/tc.yaml b/Documentation/netlink/specs/tc.yaml
index 953aa837958b..c7e6a734cd12 100644
--- a/Documentation/netlink/specs/tc.yaml
+++ b/Documentation/netlink/specs/tc.yaml
@@ -1452,7 +1452,7 @@ attribute-sets:
       -
         name: stats
         type: nest
-        nested-attributes: tc-act-stats-attrs
+        nested-attributes: tca-stats-attrs
       -
         name: pad
         type: pad
@@ -1471,38 +1471,6 @@ attribute-sets:
       -
         name: in-hw-count
         type: u32
-  -
-    name: tc-act-stats-attrs
-    attributes:
-      -
-        name: basic
-        type: binary
-        struct: gnet-stats-basic
-      -
-        name: rate-est
-        type: binary
-        struct: gnet-stats-rate-est
-      -
-        name: queue
-        type: binary
-        struct: gnet-stats-queue
-      -
-        name: app
-        type: binary
-      -
-        name: rate-est64
-        type: binary
-        struct: gnet-stats-rate-est64
-      -
-        name: pad
-        type: pad
-      -
-        name: basic-hw
-        type: binary
-        struct: gnet-stats-basic
-      -
-        name: pkt64
-        type: u64
   -
     name: tc-act-bpf-attrs
     attributes:
@@ -1797,44 +1765,6 @@ attribute-sets:
       -
         name: key-ex
         type: binary
-  -
-    name: tc-act-police-attrs
-    attributes:
-      -
-        name: tbf
-        type: binary
-        struct: tc-police
-      -
-        name: rate
-        type: binary # TODO
-      -
-        name: peakrate
-        type: binary # TODO
-      -
-        name: avrate
-        type: u32
-      -
-        name: result
-        type: u32
-      -
-        name: tm
-        type: binary
-        struct: tcf-t
-      -
-        name: pad
-        type: pad
-      -
-        name: rate64
-        type: u64
-      -
-        name: peakrate64
-        type: u64
-      -
-        name: pktrate64
-        type: u64
-      -
-        name: pktburst64
-        type: u64
   -
     name: tc-act-simple-attrs
     attributes:
@@ -3327,10 +3257,10 @@ attribute-sets:
         struct: tc-police
       -
         name: rate
-        type: binary
+        type: binary # TODO
       -
         name: peakrate
-        type: binary
+        type: binary # TODO
       -
         name: avrate
         type: u32
@@ -3817,7 +3747,7 @@ sub-messages:
         attribute-set: tc-act-pedit-attrs
       -
         value: police
-        attribute-set: tc-act-police-attrs
+        attribute-set: tc-police-attrs
       -
         value: sample
         attribute-set: tc-act-sample-attrs
-- 
2.51.0


From eb1f803f9851e1bf40924a06c1db349540290bac Mon Sep 17 00:00:00 2001
From: Jakub Kicinski <kuba@kernel.org>
Date: Tue, 20 May 2025 09:19:07 -0700
Subject: [PATCH 12/16] netlink: specs: tc: use tc-gact instead of tc-gen as
 struct name

There is a define in the uAPI header called tc_gen which expands
to the "generic" TC action fields. This helps other actions include
the base fields without having to deal with nested structs.

A couple of actions (sample, gact) do not define extra fields,
so the spec used a common tc-gen struct for both of them.
Unfortunately this struct does not exist in C. Let's use gact's
(generic act's) struct for basic actions.

Reviewed-by: Donald Hunter <donald.hunter@gmail.com>
Link: https://patch.msgid.link/20250520161916.413298-4-kuba@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
---
 Documentation/netlink/specs/tc.yaml | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/Documentation/netlink/specs/tc.yaml b/Documentation/netlink/specs/tc.yaml
index c7e6a734cd12..697fdd1219d5 100644
--- a/Documentation/netlink/specs/tc.yaml
+++ b/Documentation/netlink/specs/tc.yaml
@@ -1186,7 +1186,7 @@ definitions:
         name: firstuse
         type: u64
   -
-    name: tc-gen
+    name: tc-gact
     type: struct
     members:
       -
@@ -3457,7 +3457,7 @@ attribute-sets:
       -
         name: parms
         type: binary
-        struct: tc-gen
+        struct: tc-gact
       -
         name: rate
         type: u32
@@ -3480,7 +3480,7 @@ attribute-sets:
       -
         name: parms
         type: binary
-        struct: tc-gen
+        struct: tc-gact
       -
         name: prob
         type: binary
-- 
2.51.0


From f9aec8025ab5fc440fcc56f31298750d9b039acc Mon Sep 17 00:00:00 2001
From: Jakub Kicinski <kuba@kernel.org>
Date: Tue, 20 May 2025 09:19:08 -0700
Subject: [PATCH 13/16] netlink: specs: tc: add C naming info

Add naming info needed by C code gen.

Reviewed-by: Donald Hunter <donald.hunter@gmail.com>
Link: https://patch.msgid.link/20250520161916.413298-5-kuba@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
---
 Documentation/netlink/specs/tc.yaml | 95 +++++++++++++++++++++++++++++
 1 file changed, 95 insertions(+)

diff --git a/Documentation/netlink/specs/tc.yaml b/Documentation/netlink/specs/tc.yaml
index 697fdd1219d5..8d5e5cb439e4 100644
--- a/Documentation/netlink/specs/tc.yaml
+++ b/Documentation/netlink/specs/tc.yaml
@@ -2,6 +2,7 @@
 
 name: tc
 protocol: netlink-raw
+uapi-header: linux/pkt_cls.h
 protonum: 0
 
 doc:
@@ -12,6 +13,7 @@ definitions:
   -
     name: tcmsg
     type: struct
+    header: linux/rtnetlink.h
     members:
       -
         name: family
@@ -34,6 +36,7 @@ definitions:
         type: u32
   -
     name: tc-cls-flags
+    enum-name:
     type: flags
     entries:
       - skip-hw
@@ -43,6 +46,8 @@ definitions:
       - verbose
   -
     name: tc-flower-key-ctrl-flags
+    name-prefix: tca-flower-key-flags-
+    enum-name:
     type: flags
     entries:
       - frag
@@ -630,6 +635,7 @@ definitions:
   -
     name: tc-ratespec
     type: struct
+    header: linux/pkt_sched.h
     members:
       -
         name: cell-log
@@ -1378,6 +1384,7 @@ definitions:
 attribute-sets:
   -
     name: tc-attrs
+    name-prefix: tca-
     attributes:
       -
         name: kind
@@ -1437,6 +1444,7 @@ attribute-sets:
         type: string
   -
     name: tc-act-attrs
+    name-prefix: tca-act-
     attributes:
       -
         name: kind
@@ -1473,6 +1481,8 @@ attribute-sets:
         type: u32
   -
     name: tc-act-bpf-attrs
+    name-prefix: tca-act-bpf-
+    header: linux/tc_act/tc_bpf.h
     attributes:
       -
         name: tm
@@ -1504,6 +1514,8 @@ attribute-sets:
         type: binary
   -
     name: tc-act-connmark-attrs
+    name-prefix: tca-connmark-
+    header: linux/tc_act/tc_connmark.h
     attributes:
       -
         name: parms
@@ -1517,6 +1529,8 @@ attribute-sets:
         type: pad
   -
     name: tc-act-csum-attrs
+    name-prefix: tca-csum-
+    header: linux/tc_act/tc_csum.h
     attributes:
       -
         name: parms
@@ -1530,6 +1544,8 @@ attribute-sets:
         type: pad
   -
     name: tc-act-ct-attrs
+    name-prefix: tca-ct-
+    header: linux/tc_act/tc_ct.h
     attributes:
       -
         name: parms
@@ -1592,6 +1608,8 @@ attribute-sets:
         type: u8
   -
     name: tc-act-ctinfo-attrs
+    name-prefix: tca-ctinfo-
+    header: linux/tc_act/tc_ctinfo.h
     attributes:
       -
         name: pad
@@ -1626,6 +1644,8 @@ attribute-sets:
         type: u64
   -
     name: tc-act-gate-attrs
+    name-prefix: tca-gate-
+    header: linux/tc_act/tc_gate.h
     attributes:
       -
         name: tm
@@ -1660,6 +1680,8 @@ attribute-sets:
         type: s32
   -
     name: tc-act-ife-attrs
+    name-prefix: tca-ife-
+    header: linux/tc_act/tc_ife.h
     attributes:
       -
         name: parms
@@ -1685,6 +1707,8 @@ attribute-sets:
         type: pad
   -
     name: tc-act-mirred-attrs
+    name-prefix: tca-mirred-
+    header: linux/tc_act/tc_mirred.h
     attributes:
       -
         name: tm
@@ -1701,6 +1725,8 @@ attribute-sets:
         type: binary
   -
     name: tc-act-mpls-attrs
+    name-prefix: tca-mpls-
+    header: linux/tc_act/tc_mpls.h
     attributes:
       -
         name: tm
@@ -1731,6 +1757,8 @@ attribute-sets:
         type: u8
   -
     name: tc-act-nat-attrs
+    name-prefix: tca-nat-
+    header: linux/tc_act/tc_nat.h
     attributes:
       -
         name: parms
@@ -1744,6 +1772,8 @@ attribute-sets:
         type: pad
   -
     name: tc-act-pedit-attrs
+    name-prefix: tca-pedit-
+    header: linux/tc_act/tc_pedit.h
     attributes:
       -
         name: tm
@@ -1767,6 +1797,8 @@ attribute-sets:
         type: binary
   -
     name: tc-act-simple-attrs
+    name-prefix: tca-def-
+    header: linux/tc_act/tc_defact.h
     attributes:
       -
         name: tm
@@ -1783,6 +1815,8 @@ attribute-sets:
         type: pad
   -
     name: tc-act-skbedit-attrs
+    name-prefix: tca-skbedit-
+    header: linux/tc_act/tc_skbedit.h
     attributes:
       -
         name: tm
@@ -1817,6 +1851,8 @@ attribute-sets:
         type: u16
   -
     name: tc-act-skbmod-attrs
+    name-prefix: tca-skbmod-
+    header: linux/tc_act/tc_skbmod.h
     attributes:
       -
         name: tm
@@ -1839,6 +1875,8 @@ attribute-sets:
         type: pad
   -
     name: tc-act-tunnel-key-attrs
+    name-prefix: tca-tunnel-key-
+    header: linux/tc_act/tc_tunnel_key.h
     attributes:
       -
         name: tm
@@ -1889,6 +1927,8 @@ attribute-sets:
         type: flag
   -
     name: tc-act-vlan-attrs
+    name-prefix: tca-vlan-
+    header: linux/tc_act/tc_vlan.h
     attributes:
       -
         name: tm
@@ -1918,6 +1958,7 @@ attribute-sets:
         type: binary
   -
     name: tc-basic-attrs
+    name-prefix: tca-basic-
     attributes:
       -
         name: classid
@@ -1944,6 +1985,7 @@ attribute-sets:
         type: pad
   -
     name: tc-bpf-attrs
+    name-prefix: tca-bpf-
     attributes:
       -
         name: act
@@ -1983,6 +2025,7 @@ attribute-sets:
         type: u32
   -
     name: tc-cake-attrs
+    name-prefix: tca-cake-
     attributes:
       -
         name: pad
@@ -2040,6 +2083,7 @@ attribute-sets:
         type: u32
   -
     name: tc-cake-stats-attrs
+    name-prefix: tca-cake-stats-
     attributes:
       -
         name: pad
@@ -2093,6 +2137,7 @@ attribute-sets:
         type: s32
   -
     name: tc-cake-tin-stats-attrs
+    name-prefix: tca-cake-tin-stats-
     attributes:
       -
         name: pad
@@ -2171,6 +2216,7 @@ attribute-sets:
         type: u32
   -
     name: tc-cbs-attrs
+    name-prefix: tca-cbs-
     attributes:
       -
         name: parms
@@ -2178,6 +2224,7 @@ attribute-sets:
         struct: tc-cbs-qopt
   -
     name: tc-cgroup-attrs
+    name-prefix: tca-cgroup-
     attributes:
       -
         name: act
@@ -2193,6 +2240,7 @@ attribute-sets:
         type: binary
   -
     name: tc-choke-attrs
+    name-prefix: tca-choke-
     attributes:
       -
         name: parms
@@ -2209,6 +2257,7 @@ attribute-sets:
         type: u32
   -
     name: tc-codel-attrs
+    name-prefix: tca-codel-
     attributes:
       -
         name: target
@@ -2227,12 +2276,15 @@ attribute-sets:
         type: u32
   -
     name: tc-drr-attrs
+    name-prefix: tca-drr-
     attributes:
       -
         name: quantum
         type: u32
   -
     name: tc-ematch-attrs
+    name-prefix: tca-ematch-
+    attr-max-name: tca-ematch-tree-max
     attributes:
       -
         name: tree-hdr
@@ -2243,6 +2295,7 @@ attribute-sets:
         type: binary
   -
     name: tc-flow-attrs
+    name-prefix: tca-flow-
     attributes:
       -
         name: keys
@@ -2283,6 +2336,7 @@ attribute-sets:
         type: u32
   -
     name: tc-flower-attrs
+    name-prefix: tca-flower-
     attributes:
       -
         name: classid
@@ -2709,6 +2763,7 @@ attribute-sets:
         enum-as-flags: true
   -
     name: tc-flower-key-enc-opts-attrs
+    name-prefix: tca-flower-key-enc-opts-
     attributes:
       -
         name: geneve
@@ -2728,6 +2783,7 @@ attribute-sets:
         nested-attributes: tc-flower-key-enc-opt-gtp-attrs
   -
     name: tc-flower-key-enc-opt-geneve-attrs
+    name-prefix: tca-flower-key-enc-opt-geneve-
     attributes:
       -
         name: class
@@ -2740,12 +2796,14 @@ attribute-sets:
         type: binary
   -
     name: tc-flower-key-enc-opt-vxlan-attrs
+    name-prefix: tca-flower-key-enc-opt-vxlan-
     attributes:
       -
         name: gbp
         type: u32
   -
     name: tc-flower-key-enc-opt-erspan-attrs
+    name-prefix: tca-flower-key-enc-opt-erspan-
     attributes:
       -
         name: ver
@@ -2761,6 +2819,7 @@ attribute-sets:
         type: u8
   -
     name: tc-flower-key-enc-opt-gtp-attrs
+    name-prefix: tca-flower-key-enc-opt-gtp-
     attributes:
       -
         name: pdu-type
@@ -2770,6 +2829,8 @@ attribute-sets:
         type: u8
   -
     name: tc-flower-key-mpls-opt-attrs
+    name-prefix: tca-flower-key-mpls-opt-
+    attr-max-name: tca-flower-key-mpls-opt-lse-max
     attributes:
       -
         name: lse-depth
@@ -2788,6 +2849,7 @@ attribute-sets:
         type: u32
   -
     name: tc-flower-key-cfm-attrs
+    name-prefix: tca-flower-key-cfm-
     attributes:
       -
         name: md-level
@@ -2797,6 +2859,7 @@ attribute-sets:
         type: u8
   -
     name: tc-fw-attrs
+    name-prefix: tca-fw-
     attributes:
       -
         name: classid
@@ -2818,6 +2881,7 @@ attribute-sets:
         type: u32
   -
     name: tc-gred-attrs
+    name-prefix: tca-gred-
     attributes:
       -
         name: parms
@@ -2843,6 +2907,7 @@ attribute-sets:
         nested-attributes: tca-gred-vq-list-attrs
   -
     name: tca-gred-vq-list-attrs
+    name-prefix: tca-gred-vq-
     attributes:
       -
         name: entry
@@ -2851,6 +2916,7 @@ attribute-sets:
         multi-attr: true
   -
     name: tca-gred-vq-entry-attrs
+    name-prefix: tca-gred-vq-
     attributes:
       -
         name: pad
@@ -2902,6 +2968,7 @@ attribute-sets:
         type: binary
   -
     name: tc-hhf-attrs
+    name-prefix: tca-hhf-
     attributes:
       -
         name: backlog-limit
@@ -2926,6 +2993,7 @@ attribute-sets:
         type: u32
   -
     name: tc-htb-attrs
+    name-prefix: tca-htb-
     attributes:
       -
         name: parms
@@ -2958,6 +3026,7 @@ attribute-sets:
         type: flag
   -
     name: tc-matchall-attrs
+    name-prefix: tca-matchall-
     attributes:
       -
         name: classid
@@ -2979,6 +3048,7 @@ attribute-sets:
         type: pad
   -
     name: tc-etf-attrs
+    name-prefix: tca-etf-
     attributes:
       -
         name: parms
@@ -2986,6 +3056,7 @@ attribute-sets:
         struct: tc-etf-qopt
   -
     name: tc-ets-attrs
+    name-prefix: tca-ets-
     attributes:
       -
         name: nbands
@@ -3011,6 +3082,7 @@ attribute-sets:
         multi-attr: true
   -
     name: tc-fq-attrs
+    name-prefix: tca-fq-
     attributes:
       -
         name: plimit
@@ -3082,6 +3154,7 @@ attribute-sets:
         doc: Weights for each band
   -
     name: tc-fq-codel-attrs
+    name-prefix: tca-fq-codel-
     attributes:
       -
         name: target
@@ -3118,6 +3191,7 @@ attribute-sets:
         type: u8
   -
     name: tc-fq-pie-attrs
+    name-prefix: tca-fq-pie-
     attributes:
       -
         name: limit
@@ -3157,6 +3231,7 @@ attribute-sets:
         type: u32
   -
     name: tc-netem-attrs
+    name-prefix: tca-netem-
     attributes:
       -
         name: corr
@@ -3210,6 +3285,7 @@ attribute-sets:
         type: u64
   -
     name: tc-netem-loss-attrs
+    name-prefix: netem-loss-
     attributes:
       -
         name: gi
@@ -3223,6 +3299,7 @@ attribute-sets:
         struct: tc-netem-gemodel
   -
     name: tc-pie-attrs
+    name-prefix: tca-pie-
     attributes:
       -
         name: target
@@ -3250,6 +3327,7 @@ attribute-sets:
         type: u32
   -
     name: tc-police-attrs
+    name-prefix: tca-police-
     attributes:
       -
         name: tbf
@@ -3288,6 +3366,7 @@ attribute-sets:
         type: u64
   -
     name: tc-qfq-attrs
+    name-prefix: tca-qfq-
     attributes:
       -
         name: weight
@@ -3297,6 +3376,7 @@ attribute-sets:
         type: u32
   -
     name: tc-red-attrs
+    name-prefix: tca-red-
     attributes:
       -
         name: parms
@@ -3319,6 +3399,7 @@ attribute-sets:
         type: u32
   -
     name: tc-route-attrs
+    name-prefix: tca-route4-
     attributes:
       -
         name: classid
@@ -3343,6 +3424,7 @@ attribute-sets:
         nested-attributes: tc-act-attrs
   -
     name: tc-taprio-attrs
+    name-prefix: tca-taprio-attr-
     attributes:
       -
         name: priomap
@@ -3386,6 +3468,7 @@ attribute-sets:
         nested-attributes: tc-taprio-tc-entry-attrs
   -
     name: tc-taprio-sched-entry-list
+    name-prefix: tca-taprio-sched-
     attributes:
       -
         name: entry
@@ -3394,6 +3477,7 @@ attribute-sets:
         multi-attr: true
   -
     name: tc-taprio-sched-entry
+    name-prefix: tca-taprio-sched-entry-
     attributes:
       -
         name: index
@@ -3409,6 +3493,7 @@ attribute-sets:
         type: u32
   -
     name: tc-taprio-tc-entry-attrs
+    name-prefix: tca-taprio-tc-entry-
     attributes:
       -
         name: index
@@ -3421,6 +3506,7 @@ attribute-sets:
         type: u32
   -
     name: tc-tbf-attrs
+    name-prefix: tca-tbf-
     attributes:
       -
         name: parms
@@ -3449,6 +3535,8 @@ attribute-sets:
         type: pad
   -
     name: tc-act-sample-attrs
+    name-prefix: tca-sample-
+    header: linux/tc_act/tc_sample.h
     attributes:
       -
         name: tm
@@ -3472,6 +3560,8 @@ attribute-sets:
         type: pad
   -
     name: tc-act-gact-attrs
+    name-prefix: tca-gact-
+    header: linux/tc_act/tc_gact.h
     attributes:
       -
         name: tm
@@ -3490,6 +3580,7 @@ attribute-sets:
         type: pad
   -
     name: tca-stab-attrs
+    name-prefix: tca-stab-
     attributes:
       -
         name: base
@@ -3500,6 +3591,8 @@ attribute-sets:
         type: binary
   -
     name: tca-stats-attrs
+    name-prefix: tca-stats-
+    header: linux/gen_stats.h
     attributes:
       -
         name: basic
@@ -3534,6 +3627,7 @@ attribute-sets:
         type: u64
   -
     name: tc-u32-attrs
+    name-prefix: tca-u32-
     attributes:
       -
         name: classid
@@ -3805,6 +3899,7 @@ sub-messages:
 
 operations:
   enum-model: directional
+  name-prefix: rtm-
   list:
     -
       name: newqdisc
-- 
2.51.0


From ba5a199b2401de3d5df1fba4d5dcd92908461a07 Mon Sep 17 00:00:00 2001
From: Jakub Kicinski <kuba@kernel.org>
Date: Tue, 20 May 2025 09:19:09 -0700
Subject: [PATCH 14/16] netlink: specs: tc: drop the family name prefix from
 attrs

All attribute sets and messages are prefixed with tc-.
The C codegen also adds the family name to all structs.
We end up with names like struct tc_tc_act_attrs.
Remove the tc- prefixes to shorten the names.
This should not impact Python as the attr set names
are never exposed to user, they are only used to refer
to things internally, in the encoder / decoder.

Reviewed-by: Donald Hunter <donald.hunter@gmail.com>
Link: https://patch.msgid.link/20250520161916.413298-6-kuba@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
---
 Documentation/netlink/specs/tc.yaml | 334 ++++++++++++++--------------
 1 file changed, 167 insertions(+), 167 deletions(-)

diff --git a/Documentation/netlink/specs/tc.yaml b/Documentation/netlink/specs/tc.yaml
index 8d5e5cb439e4..6e8db7adde3c 100644
--- a/Documentation/netlink/specs/tc.yaml
+++ b/Documentation/netlink/specs/tc.yaml
@@ -35,7 +35,7 @@ definitions:
         name: info
         type: u32
   -
-    name: tc-cls-flags
+    name: cls-flags
     enum-name:
     type: flags
     entries:
@@ -45,7 +45,7 @@ definitions:
       - not-in-nw
       - verbose
   -
-    name: tc-flower-key-ctrl-flags
+    name: flower-key-ctrl-flags
     name-prefix: tca-flower-key-flags-
     enum-name:
     type: flags
@@ -1383,7 +1383,7 @@ definitions:
         type: s32
 attribute-sets:
   -
-    name: tc-attrs
+    name: attrs
     name-prefix: tca-
     attributes:
       -
@@ -1392,7 +1392,7 @@ attribute-sets:
       -
         name: options
         type: sub-message
-        sub-message: tc-options-msg
+        sub-message: options-msg
         selector: kind
       -
         name: stats
@@ -1443,7 +1443,7 @@ attribute-sets:
         name: ext-warn-msg
         type: string
   -
-    name: tc-act-attrs
+    name: act-attrs
     name-prefix: tca-act-
     attributes:
       -
@@ -1452,7 +1452,7 @@ attribute-sets:
       -
         name: options
         type: sub-message
-        sub-message: tc-act-options-msg
+        sub-message: act-options-msg
         selector: kind
       -
         name: index
@@ -1480,7 +1480,7 @@ attribute-sets:
         name: in-hw-count
         type: u32
   -
-    name: tc-act-bpf-attrs
+    name: act-bpf-attrs
     name-prefix: tca-act-bpf-
     header: linux/tc_act/tc_bpf.h
     attributes:
@@ -1513,7 +1513,7 @@ attribute-sets:
         name: id
         type: binary
   -
-    name: tc-act-connmark-attrs
+    name: act-connmark-attrs
     name-prefix: tca-connmark-
     header: linux/tc_act/tc_connmark.h
     attributes:
@@ -1528,7 +1528,7 @@ attribute-sets:
         name: pad
         type: pad
   -
-    name: tc-act-csum-attrs
+    name: act-csum-attrs
     name-prefix: tca-csum-
     header: linux/tc_act/tc_csum.h
     attributes:
@@ -1543,7 +1543,7 @@ attribute-sets:
         name: pad
         type: pad
   -
-    name: tc-act-ct-attrs
+    name: act-ct-attrs
     name-prefix: tca-ct-
     header: linux/tc_act/tc_ct.h
     attributes:
@@ -1607,7 +1607,7 @@ attribute-sets:
         name: helper-proto
         type: u8
   -
-    name: tc-act-ctinfo-attrs
+    name: act-ctinfo-attrs
     name-prefix: tca-ctinfo-
     header: linux/tc_act/tc_ctinfo.h
     attributes:
@@ -1643,7 +1643,7 @@ attribute-sets:
         name: stats-cpmark-set
         type: u64
   -
-    name: tc-act-gate-attrs
+    name: act-gate-attrs
     name-prefix: tca-gate-
     header: linux/tc_act/tc_gate.h
     attributes:
@@ -1679,7 +1679,7 @@ attribute-sets:
         name: clockid
         type: s32
   -
-    name: tc-act-ife-attrs
+    name: act-ife-attrs
     name-prefix: tca-ife-
     header: linux/tc_act/tc_ife.h
     attributes:
@@ -1706,7 +1706,7 @@ attribute-sets:
         name: pad
         type: pad
   -
-    name: tc-act-mirred-attrs
+    name: act-mirred-attrs
     name-prefix: tca-mirred-
     header: linux/tc_act/tc_mirred.h
     attributes:
@@ -1724,7 +1724,7 @@ attribute-sets:
         name: blockid
         type: binary
   -
-    name: tc-act-mpls-attrs
+    name: act-mpls-attrs
     name-prefix: tca-mpls-
     header: linux/tc_act/tc_mpls.h
     attributes:
@@ -1756,7 +1756,7 @@ attribute-sets:
         name: bos
         type: u8
   -
-    name: tc-act-nat-attrs
+    name: act-nat-attrs
     name-prefix: tca-nat-
     header: linux/tc_act/tc_nat.h
     attributes:
@@ -1771,7 +1771,7 @@ attribute-sets:
         name: pad
         type: pad
   -
-    name: tc-act-pedit-attrs
+    name: act-pedit-attrs
     name-prefix: tca-pedit-
     header: linux/tc_act/tc_pedit.h
     attributes:
@@ -1796,7 +1796,7 @@ attribute-sets:
         name: key-ex
         type: binary
   -
-    name: tc-act-simple-attrs
+    name: act-simple-attrs
     name-prefix: tca-def-
     header: linux/tc_act/tc_defact.h
     attributes:
@@ -1814,7 +1814,7 @@ attribute-sets:
         name: pad
         type: pad
   -
-    name: tc-act-skbedit-attrs
+    name: act-skbedit-attrs
     name-prefix: tca-skbedit-
     header: linux/tc_act/tc_skbedit.h
     attributes:
@@ -1850,7 +1850,7 @@ attribute-sets:
         name: queue-mapping-max
         type: u16
   -
-    name: tc-act-skbmod-attrs
+    name: act-skbmod-attrs
     name-prefix: tca-skbmod-
     header: linux/tc_act/tc_skbmod.h
     attributes:
@@ -1874,7 +1874,7 @@ attribute-sets:
         name: pad
         type: pad
   -
-    name: tc-act-tunnel-key-attrs
+    name: act-tunnel-key-attrs
     name-prefix: tca-tunnel-key-
     header: linux/tc_act/tc_tunnel_key.h
     attributes:
@@ -1926,7 +1926,7 @@ attribute-sets:
         name: no-frag
         type: flag
   -
-    name: tc-act-vlan-attrs
+    name: act-vlan-attrs
     name-prefix: tca-vlan-
     header: linux/tc_act/tc_vlan.h
     attributes:
@@ -1957,7 +1957,7 @@ attribute-sets:
         name: push-eth-src
         type: binary
   -
-    name: tc-basic-attrs
+    name: basic-attrs
     name-prefix: tca-basic-
     attributes:
       -
@@ -1966,16 +1966,16 @@ attribute-sets:
       -
         name: ematches
         type: nest
-        nested-attributes: tc-ematch-attrs
+        nested-attributes: ematch-attrs
       -
         name: act
         type: indexed-array
         sub-type: nest
-        nested-attributes: tc-act-attrs
+        nested-attributes: act-attrs
       -
         name: police
         type: nest
-        nested-attributes: tc-police-attrs
+        nested-attributes: police-attrs
       -
         name: pcnt
         type: binary
@@ -1984,18 +1984,18 @@ attribute-sets:
         name: pad
         type: pad
   -
-    name: tc-bpf-attrs
+    name: bpf-attrs
     name-prefix: tca-bpf-
     attributes:
       -
         name: act
         type: indexed-array
         sub-type: nest
-        nested-attributes: tc-act-attrs
+        nested-attributes: act-attrs
       -
         name: police
         type: nest
-        nested-attributes: tc-police-attrs
+        nested-attributes: police-attrs
       -
         name: classid
         type: u32
@@ -2024,7 +2024,7 @@ attribute-sets:
         name: id
         type: u32
   -
-    name: tc-cake-attrs
+    name: cake-attrs
     name-prefix: tca-cake-
     attributes:
       -
@@ -2082,7 +2082,7 @@ attribute-sets:
         name: fwmark
         type: u32
   -
-    name: tc-cake-stats-attrs
+    name: cake-stats-attrs
     name-prefix: tca-cake-stats-
     attributes:
       -
@@ -2116,7 +2116,7 @@ attribute-sets:
         name: tin-stats
         type: indexed-array
         sub-type: nest
-        nested-attributes: tc-cake-tin-stats-attrs
+        nested-attributes: cake-tin-stats-attrs
       -
         name: deficit
         type: s32
@@ -2136,7 +2136,7 @@ attribute-sets:
         name: blue-timer-us
         type: s32
   -
-    name: tc-cake-tin-stats-attrs
+    name: cake-tin-stats-attrs
     name-prefix: tca-cake-tin-stats-
     attributes:
       -
@@ -2215,7 +2215,7 @@ attribute-sets:
         name: flow-quantum
         type: u32
   -
-    name: tc-cbs-attrs
+    name: cbs-attrs
     name-prefix: tca-cbs-
     attributes:
       -
@@ -2223,23 +2223,23 @@ attribute-sets:
         type: binary
         struct: tc-cbs-qopt
   -
-    name: tc-cgroup-attrs
+    name: cgroup-attrs
     name-prefix: tca-cgroup-
     attributes:
       -
         name: act
         type: indexed-array
         sub-type: nest
-        nested-attributes: tc-act-attrs
+        nested-attributes: act-attrs
       -
         name: police
         type: nest
-        nested-attributes: tc-police-attrs
+        nested-attributes: police-attrs
       -
         name: ematches
         type: binary
   -
-    name: tc-choke-attrs
+    name: choke-attrs
     name-prefix: tca-choke-
     attributes:
       -
@@ -2256,7 +2256,7 @@ attribute-sets:
         name: max-p
         type: u32
   -
-    name: tc-codel-attrs
+    name: codel-attrs
     name-prefix: tca-codel-
     attributes:
       -
@@ -2275,14 +2275,14 @@ attribute-sets:
         name: ce-threshold
         type: u32
   -
-    name: tc-drr-attrs
+    name: drr-attrs
     name-prefix: tca-drr-
     attributes:
       -
         name: quantum
         type: u32
   -
-    name: tc-ematch-attrs
+    name: ematch-attrs
     name-prefix: tca-ematch-
     attr-max-name: tca-ematch-tree-max
     attributes:
@@ -2294,7 +2294,7 @@ attribute-sets:
         name: tree-list
         type: binary
   -
-    name: tc-flow-attrs
+    name: flow-attrs
     name-prefix: tca-flow-
     attributes:
       -
@@ -2327,7 +2327,7 @@ attribute-sets:
       -
         name: police
         type: nest
-        nested-attributes: tc-police-attrs
+        nested-attributes: police-attrs
       -
         name: ematches
         type: binary
@@ -2335,7 +2335,7 @@ attribute-sets:
         name: perturb
         type: u32
   -
-    name: tc-flower-attrs
+    name: flower-attrs
     name-prefix: tca-flower-
     attributes:
       -
@@ -2348,7 +2348,7 @@ attribute-sets:
         name: act
         type: indexed-array
         sub-type: nest
-        nested-attributes: tc-act-attrs
+        nested-attributes: act-attrs
       -
         name: key-eth-dst
         type: binary
@@ -2427,7 +2427,7 @@ attribute-sets:
       -
         name: flags
         type: u32
-        enum: tc-cls-flags
+        enum: cls-flags
         enum-as-flags: true
       -
         name: key-vlan-id
@@ -2532,13 +2532,13 @@ attribute-sets:
         name: key-flags
         type: u32
         byte-order: big-endian
-        enum: tc-flower-key-ctrl-flags
+        enum: flower-key-ctrl-flags
         enum-as-flags: true
       -
         name: key-flags-mask
         type: u32
         byte-order: big-endian
-        enum: tc-flower-key-ctrl-flags
+        enum: flower-key-ctrl-flags
         enum-as-flags: true
       -
         name: key-icmpv4-code
@@ -2661,11 +2661,11 @@ attribute-sets:
       -
         name: key-enc-opts
         type: nest
-        nested-attributes: tc-flower-key-enc-opts-attrs
+        nested-attributes: flower-key-enc-opts-attrs
       -
         name: key-enc-opts-mask
         type: nest
-        nested-attributes: tc-flower-key-enc-opts-attrs
+        nested-attributes: flower-key-enc-opts-attrs
       -
         name: in-hw-count
         type: u32
@@ -2712,7 +2712,7 @@ attribute-sets:
       -
         name: key-mpls-opts
         type: nest
-        nested-attributes: tc-flower-key-mpls-opt-attrs
+        nested-attributes: flower-key-mpls-opt-attrs
       -
         name: key-hash
         type: u32
@@ -2740,7 +2740,7 @@ attribute-sets:
       -
         name: key-cfm
         type: nest
-        nested-attributes: tc-flower-key-cfm-attrs
+        nested-attributes: flower-key-cfm-attrs
       -
         name: key-spi
         type: u32
@@ -2753,36 +2753,36 @@ attribute-sets:
         name: key-enc-flags
         type: u32
         byte-order: big-endian
-        enum: tc-flower-key-ctrl-flags
+        enum: flower-key-ctrl-flags
         enum-as-flags: true
       -
         name: key-enc-flags-mask
         type: u32
         byte-order: big-endian
-        enum: tc-flower-key-ctrl-flags
+        enum: flower-key-ctrl-flags
         enum-as-flags: true
   -
-    name: tc-flower-key-enc-opts-attrs
+    name: flower-key-enc-opts-attrs
     name-prefix: tca-flower-key-enc-opts-
     attributes:
       -
         name: geneve
         type: nest
-        nested-attributes: tc-flower-key-enc-opt-geneve-attrs
+        nested-attributes: flower-key-enc-opt-geneve-attrs
       -
         name: vxlan
         type: nest
-        nested-attributes: tc-flower-key-enc-opt-vxlan-attrs
+        nested-attributes: flower-key-enc-opt-vxlan-attrs
       -
         name: erspan
         type: nest
-        nested-attributes: tc-flower-key-enc-opt-erspan-attrs
+        nested-attributes: flower-key-enc-opt-erspan-attrs
       -
         name: gtp
         type: nest
-        nested-attributes: tc-flower-key-enc-opt-gtp-attrs
+        nested-attributes: flower-key-enc-opt-gtp-attrs
   -
-    name: tc-flower-key-enc-opt-geneve-attrs
+    name: flower-key-enc-opt-geneve-attrs
     name-prefix: tca-flower-key-enc-opt-geneve-
     attributes:
       -
@@ -2795,14 +2795,14 @@ attribute-sets:
         name: data
         type: binary
   -
-    name: tc-flower-key-enc-opt-vxlan-attrs
+    name: flower-key-enc-opt-vxlan-attrs
     name-prefix: tca-flower-key-enc-opt-vxlan-
     attributes:
       -
         name: gbp
         type: u32
   -
-    name: tc-flower-key-enc-opt-erspan-attrs
+    name: flower-key-enc-opt-erspan-attrs
     name-prefix: tca-flower-key-enc-opt-erspan-
     attributes:
       -
@@ -2818,7 +2818,7 @@ attribute-sets:
         name: hwid
         type: u8
   -
-    name: tc-flower-key-enc-opt-gtp-attrs
+    name: flower-key-enc-opt-gtp-attrs
     name-prefix: tca-flower-key-enc-opt-gtp-
     attributes:
       -
@@ -2828,7 +2828,7 @@ attribute-sets:
         name: qfi
         type: u8
   -
-    name: tc-flower-key-mpls-opt-attrs
+    name: flower-key-mpls-opt-attrs
     name-prefix: tca-flower-key-mpls-opt-
     attr-max-name: tca-flower-key-mpls-opt-lse-max
     attributes:
@@ -2848,7 +2848,7 @@ attribute-sets:
         name: lse-label
         type: u32
   -
-    name: tc-flower-key-cfm-attrs
+    name: flower-key-cfm-attrs
     name-prefix: tca-flower-key-cfm-
     attributes:
       -
@@ -2858,7 +2858,7 @@ attribute-sets:
         name: opcode
         type: u8
   -
-    name: tc-fw-attrs
+    name: fw-attrs
     name-prefix: tca-fw-
     attributes:
       -
@@ -2867,7 +2867,7 @@ attribute-sets:
       -
         name: police
         type: nest
-        nested-attributes: tc-police-attrs
+        nested-attributes: police-attrs
       -
         name: indev
         type: string
@@ -2875,12 +2875,12 @@ attribute-sets:
         name: act
         type: indexed-array
         sub-type: nest
-        nested-attributes: tc-act-attrs
+        nested-attributes: act-attrs
       -
         name: mask
         type: u32
   -
-    name: tc-gred-attrs
+    name: gred-attrs
     name-prefix: tca-gred-
     attributes:
       -
@@ -2955,7 +2955,7 @@ attribute-sets:
         name: flags
         type: u32
   -
-    name: tc-hfsc-attrs
+    name: hfsc-attrs
     attributes:
       -
         name: rsc
@@ -2967,7 +2967,7 @@ attribute-sets:
         name: usc
         type: binary
   -
-    name: tc-hhf-attrs
+    name: hhf-attrs
     name-prefix: tca-hhf-
     attributes:
       -
@@ -2992,7 +2992,7 @@ attribute-sets:
         name: non-hh-weight
         type: u32
   -
-    name: tc-htb-attrs
+    name: htb-attrs
     name-prefix: tca-htb-
     attributes:
       -
@@ -3025,7 +3025,7 @@ attribute-sets:
         name: offload
         type: flag
   -
-    name: tc-matchall-attrs
+    name: matchall-attrs
     name-prefix: tca-matchall-
     attributes:
       -
@@ -3035,7 +3035,7 @@ attribute-sets:
         name: act
         type: indexed-array
         sub-type: nest
-        nested-attributes: tc-act-attrs
+        nested-attributes: act-attrs
       -
         name: flags
         type: u32
@@ -3047,7 +3047,7 @@ attribute-sets:
         name: pad
         type: pad
   -
-    name: tc-etf-attrs
+    name: etf-attrs
     name-prefix: tca-etf-
     attributes:
       -
@@ -3055,7 +3055,7 @@ attribute-sets:
         type: binary
         struct: tc-etf-qopt
   -
-    name: tc-ets-attrs
+    name: ets-attrs
     name-prefix: tca-ets-
     attributes:
       -
@@ -3067,7 +3067,7 @@ attribute-sets:
       -
         name: quanta
         type: nest
-        nested-attributes: tc-ets-attrs
+        nested-attributes: ets-attrs
       -
         name: quanta-band
         type: u32
@@ -3075,13 +3075,13 @@ attribute-sets:
       -
         name: priomap
         type: nest
-        nested-attributes: tc-ets-attrs
+        nested-attributes: ets-attrs
       -
         name: priomap-band
         type: u8
         multi-attr: true
   -
-    name: tc-fq-attrs
+    name: fq-attrs
     name-prefix: tca-fq-
     attributes:
       -
@@ -3153,7 +3153,7 @@ attribute-sets:
         sub-type: s32
         doc: Weights for each band
   -
-    name: tc-fq-codel-attrs
+    name: fq-codel-attrs
     name-prefix: tca-fq-codel-
     attributes:
       -
@@ -3190,7 +3190,7 @@ attribute-sets:
         name: ce-threshold-mask
         type: u8
   -
-    name: tc-fq-pie-attrs
+    name: fq-pie-attrs
     name-prefix: tca-fq-pie-
     attributes:
       -
@@ -3230,7 +3230,7 @@ attribute-sets:
         name: dq-rate-estimator
         type: u32
   -
-    name: tc-netem-attrs
+    name: netem-attrs
     name-prefix: tca-netem-
     attributes:
       -
@@ -3252,7 +3252,7 @@ attribute-sets:
       -
         name: loss
         type: nest
-        nested-attributes: tc-netem-loss-attrs
+        nested-attributes: netem-loss-attrs
       -
         name: rate
         type: binary
@@ -3284,7 +3284,7 @@ attribute-sets:
         name: prng-seed
         type: u64
   -
-    name: tc-netem-loss-attrs
+    name: netem-loss-attrs
     name-prefix: netem-loss-
     attributes:
       -
@@ -3298,7 +3298,7 @@ attribute-sets:
         doc: Gilbert Elliot models
         struct: tc-netem-gemodel
   -
-    name: tc-pie-attrs
+    name: pie-attrs
     name-prefix: tca-pie-
     attributes:
       -
@@ -3326,7 +3326,7 @@ attribute-sets:
         name: dq-rate-estimator
         type: u32
   -
-    name: tc-police-attrs
+    name: police-attrs
     name-prefix: tca-police-
     attributes:
       -
@@ -3365,7 +3365,7 @@ attribute-sets:
         name: pktburst64
         type: u64
   -
-    name: tc-qfq-attrs
+    name: qfq-attrs
     name-prefix: tca-qfq-
     attributes:
       -
@@ -3375,7 +3375,7 @@ attribute-sets:
         name: lmax
         type: u32
   -
-    name: tc-red-attrs
+    name: red-attrs
     name-prefix: tca-red-
     attributes:
       -
@@ -3398,7 +3398,7 @@ attribute-sets:
         name: mark-block
         type: u32
   -
-    name: tc-route-attrs
+    name: route-attrs
     name-prefix: tca-route4-
     attributes:
       -
@@ -3416,14 +3416,14 @@ attribute-sets:
       -
         name: police
         type: nest
-        nested-attributes: tc-police-attrs
+        nested-attributes: police-attrs
       -
         name: act
         type: indexed-array
         sub-type: nest
-        nested-attributes: tc-act-attrs
+        nested-attributes: act-attrs
   -
-    name: tc-taprio-attrs
+    name: taprio-attrs
     name-prefix: tca-taprio-attr-
     attributes:
       -
@@ -3433,14 +3433,14 @@ attribute-sets:
       -
         name: sched-entry-list
         type: nest
-        nested-attributes: tc-taprio-sched-entry-list
+        nested-attributes: taprio-sched-entry-list
       -
         name: sched-base-time
         type: s64
       -
         name: sched-single-entry
         type: nest
-        nested-attributes: tc-taprio-sched-entry
+        nested-attributes: taprio-sched-entry
       -
         name: sched-clockid
         type: s32
@@ -3465,18 +3465,18 @@ attribute-sets:
       -
         name: tc-entry
         type: nest
-        nested-attributes: tc-taprio-tc-entry-attrs
+        nested-attributes: taprio-tc-entry-attrs
   -
-    name: tc-taprio-sched-entry-list
+    name: taprio-sched-entry-list
     name-prefix: tca-taprio-sched-
     attributes:
       -
         name: entry
         type: nest
-        nested-attributes: tc-taprio-sched-entry
+        nested-attributes: taprio-sched-entry
         multi-attr: true
   -
-    name: tc-taprio-sched-entry
+    name: taprio-sched-entry
     name-prefix: tca-taprio-sched-entry-
     attributes:
       -
@@ -3492,7 +3492,7 @@ attribute-sets:
         name: interval
         type: u32
   -
-    name: tc-taprio-tc-entry-attrs
+    name: taprio-tc-entry-attrs
     name-prefix: tca-taprio-tc-entry-
     attributes:
       -
@@ -3505,7 +3505,7 @@ attribute-sets:
         name: fp
         type: u32
   -
-    name: tc-tbf-attrs
+    name: tbf-attrs
     name-prefix: tca-tbf-
     attributes:
       -
@@ -3534,7 +3534,7 @@ attribute-sets:
         name: pad
         type: pad
   -
-    name: tc-act-sample-attrs
+    name: act-sample-attrs
     name-prefix: tca-sample-
     header: linux/tc_act/tc_sample.h
     attributes:
@@ -3559,7 +3559,7 @@ attribute-sets:
         name: pad
         type: pad
   -
-    name: tc-act-gact-attrs
+    name: act-gact-attrs
     name-prefix: tca-gact-
     header: linux/tc_act/tc_gact.h
     attributes:
@@ -3626,7 +3626,7 @@ attribute-sets:
         name: pkt64
         type: u64
   -
-    name: tc-u32-attrs
+    name: u32-attrs
     name-prefix: tca-u32-
     attributes:
       -
@@ -3648,12 +3648,12 @@ attribute-sets:
       -
         name: police
         type: nest
-        nested-attributes: tc-police-attrs
+        nested-attributes: police-attrs
       -
         name: act
         type: indexed-array
         sub-type: nest
-        nested-attributes: tc-act-attrs
+        nested-attributes: act-attrs
       -
         name: indev
         type: string
@@ -3674,78 +3674,78 @@ attribute-sets:
 
 sub-messages:
   -
-    name: tc-options-msg
+    name: options-msg
     formats:
       -
         value: basic
-        attribute-set: tc-basic-attrs
+        attribute-set: basic-attrs
       -
         value: bpf
-        attribute-set: tc-bpf-attrs
+        attribute-set: bpf-attrs
       -
         value: bfifo
         fixed-header: tc-fifo-qopt
       -
         value: cake
-        attribute-set: tc-cake-attrs
+        attribute-set: cake-attrs
       -
         value: cbs
-        attribute-set: tc-cbs-attrs
+        attribute-set: cbs-attrs
       -
         value: cgroup
-        attribute-set: tc-cgroup-attrs
+        attribute-set: cgroup-attrs
       -
         value: choke
-        attribute-set: tc-choke-attrs
+        attribute-set: choke-attrs
       -
         value: clsact # no content
       -
         value: codel
-        attribute-set: tc-codel-attrs
+        attribute-set: codel-attrs
       -
         value: drr
-        attribute-set: tc-drr-attrs
+        attribute-set: drr-attrs
       -
         value: etf
-        attribute-set: tc-etf-attrs
+        attribute-set: etf-attrs
       -
         value: ets
-        attribute-set: tc-ets-attrs
+        attribute-set: ets-attrs
       -
         value: flow
-        attribute-set: tc-flow-attrs
+        attribute-set: flow-attrs
       -
         value: flower
-        attribute-set: tc-flower-attrs
+        attribute-set: flower-attrs
       -
         value: fq
-        attribute-set: tc-fq-attrs
+        attribute-set: fq-attrs
       -
         value: fq_codel
-        attribute-set: tc-fq-codel-attrs
+        attribute-set: fq-codel-attrs
       -
         value: fq_pie
-        attribute-set: tc-fq-pie-attrs
+        attribute-set: fq-pie-attrs
       -
         value: fw
-        attribute-set: tc-fw-attrs
+        attribute-set: fw-attrs
       -
         value: gred
-        attribute-set: tc-gred-attrs
+        attribute-set: gred-attrs
       -
         value: hfsc
         fixed-header: tc-hfsc-qopt
       -
         value: hhf
-        attribute-set: tc-hhf-attrs
+        attribute-set: hhf-attrs
       -
         value: htb
-        attribute-set: tc-htb-attrs
+        attribute-set: htb-attrs
       -
         value: ingress # no content
       -
         value: matchall
-        attribute-set: tc-matchall-attrs
+        attribute-set: matchall-attrs
       -
         value: mq # no content
       -
@@ -3757,7 +3757,7 @@ sub-messages:
       -
         value: netem
         fixed-header: tc-netem-qopt
-        attribute-set: tc-netem-attrs
+        attribute-set: netem-attrs
       -
         value: pfifo
         fixed-header: tc-fifo-qopt
@@ -3769,7 +3769,7 @@ sub-messages:
         fixed-header: tc-fifo-qopt
       -
         value: pie
-        attribute-set: tc-pie-attrs
+        attribute-set: pie-attrs
       -
         value: plug
         fixed-header: tc-plug-qopt
@@ -3778,13 +3778,13 @@ sub-messages:
         fixed-header: tc-prio-qopt
       -
         value: qfq
-        attribute-set: tc-qfq-attrs
+        attribute-set: qfq-attrs
       -
         value: red
-        attribute-set: tc-red-attrs
+        attribute-set: red-attrs
       -
         value: route
-        attribute-set: tc-route-attrs
+        attribute-set: route-attrs
       -
         value: sfb
         fixed-header: tc-sfb-qopt
@@ -3793,79 +3793,79 @@ sub-messages:
         fixed-header: tc-sfq-qopt-v1
       -
         value: taprio
-        attribute-set: tc-taprio-attrs
+        attribute-set: taprio-attrs
       -
         value: tbf
-        attribute-set: tc-tbf-attrs
+        attribute-set: tbf-attrs
       -
         value: u32
-        attribute-set: tc-u32-attrs
+        attribute-set: u32-attrs
   -
-    name: tc-act-options-msg
+    name: act-options-msg
     formats:
       -
         value: bpf
-        attribute-set: tc-act-bpf-attrs
+        attribute-set: act-bpf-attrs
       -
         value: connmark
-        attribute-set: tc-act-connmark-attrs
+        attribute-set: act-connmark-attrs
       -
         value: csum
-        attribute-set: tc-act-csum-attrs
+        attribute-set: act-csum-attrs
       -
         value: ct
-        attribute-set: tc-act-ct-attrs
+        attribute-set: act-ct-attrs
       -
         value: ctinfo
-        attribute-set: tc-act-ctinfo-attrs
+        attribute-set: act-ctinfo-attrs
       -
         value: gact
-        attribute-set: tc-act-gact-attrs
+        attribute-set: act-gact-attrs
       -
         value: gate
-        attribute-set: tc-act-gate-attrs
+        attribute-set: act-gate-attrs
       -
         value: ife
-        attribute-set: tc-act-ife-attrs
+        attribute-set: act-ife-attrs
       -
         value: mirred
-        attribute-set: tc-act-mirred-attrs
+        attribute-set: act-mirred-attrs
       -
         value: mpls
-        attribute-set: tc-act-mpls-attrs
+        attribute-set: act-mpls-attrs
       -
         value: nat
-        attribute-set: tc-act-nat-attrs
+        attribute-set: act-nat-attrs
       -
         value: pedit
-        attribute-set: tc-act-pedit-attrs
+        attribute-set: act-pedit-attrs
       -
         value: police
-        attribute-set: tc-police-attrs
+        attribute-set: police-attrs
       -
         value: sample
-        attribute-set: tc-act-sample-attrs
+        attribute-set: act-sample-attrs
       -
         value: simple
-        attribute-set: tc-act-simple-attrs
+        attribute-set: act-simple-attrs
       -
         value: skbedit
-        attribute-set: tc-act-skbedit-attrs
+        attribute-set: act-skbedit-attrs
       -
         value: skbmod
-        attribute-set: tc-act-skbmod-attrs
+        attribute-set: act-skbmod-attrs
       -
         value: tunnel_key
-        attribute-set: tc-act-tunnel-key-attrs
+        attribute-set: act-tunnel-key-attrs
       -
         value: vlan
-        attribute-set: tc-act-vlan-attrs
+        attribute-set: act-vlan-attrs
   -
     name: tca-stats-app-msg
     formats:
       -
         value: cake
-        attribute-set: tc-cake-stats-attrs
+        attribute-set: cake-stats-attrs
       -
         value: choke
         fixed-header: tc-choke-xstats
@@ -3904,7 +3904,7 @@ operations:
     -
       name: newqdisc
       doc: Create new tc qdisc.
-      attribute-set: tc-attrs
+      attribute-set: attrs
       fixed-header: tcmsg
       do:
         request:
@@ -3919,7 +3919,7 @@ operations:
     -
       name: delqdisc
       doc: Delete existing tc qdisc.
-      attribute-set: tc-attrs
+      attribute-set: attrs
       fixed-header: tcmsg
       do:
         request:
@@ -3927,7 +3927,7 @@ operations:
     -
       name: getqdisc
       doc: Get / dump tc qdisc information.
-      attribute-set: tc-attrs
+      attribute-set: attrs
       fixed-header: tcmsg
       do:
         request:
@@ -3951,7 +3951,7 @@ operations:
     -
       name: newtclass
       doc: Get / dump tc traffic class information.
-      attribute-set: tc-attrs
+      attribute-set: attrs
       fixed-header: tcmsg
       do:
         request:
@@ -3960,7 +3960,7 @@ operations:
     -
       name: deltclass
       doc: Get / dump tc traffic class information.
-      attribute-set: tc-attrs
+      attribute-set: attrs
       fixed-header: tcmsg
       do:
         request:
@@ -3968,7 +3968,7 @@ operations:
     -
       name: gettclass
       doc: Get / dump tc traffic class information.
-      attribute-set: tc-attrs
+      attribute-set: attrs
       fixed-header: tcmsg
       do:
         request:
@@ -3979,7 +3979,7 @@ operations:
     -
       name: newtfilter
       doc: Get / dump tc filter information.
-      attribute-set: tc-attrs
+      attribute-set: attrs
       fixed-header: tcmsg
       do:
         request:
@@ -3988,7 +3988,7 @@ operations:
     -
       name: deltfilter
       doc: Get / dump tc filter information.
-      attribute-set: tc-attrs
+      attribute-set: attrs
       fixed-header: tcmsg
       do:
         request:
@@ -3999,7 +3999,7 @@ operations:
     -
       name: gettfilter
       doc: Get / dump tc filter information.
-      attribute-set: tc-attrs
+      attribute-set: attrs
       fixed-header: tcmsg
       do:
         request:
@@ -4022,7 +4022,7 @@ operations:
     -
       name: newchain
       doc: Get / dump tc chain information.
-      attribute-set: tc-attrs
+      attribute-set: attrs
       fixed-header: tcmsg
       do:
         request:
@@ -4031,7 +4031,7 @@ operations:
     -
       name: delchain
       doc: Get / dump tc chain information.
-      attribute-set: tc-attrs
+      attribute-set: attrs
       fixed-header: tcmsg
       do:
         request:
@@ -4041,7 +4041,7 @@ operations:
     -
       name: getchain
       doc: Get / dump tc chain information.
-      attribute-set: tc-attrs
+      attribute-set: attrs
       fixed-header: tcmsg
       do:
         request:
-- 
2.51.0


From cb39645d9a6a8b84f2e820db7c2b49ebd4b18b2c Mon Sep 17 00:00:00 2001
From: Jakub Kicinski <kuba@kernel.org>
Date: Tue, 20 May 2025 09:19:10 -0700
Subject: [PATCH 15/16] tools: ynl-gen: support passing selector to a nest

In rtnetlink all submessages had the selector at the same level
of nesting as the submessage. We could refer to the relevant
attribute from the current struct. In TC, stats are one level
of nesting deeper than "kind". Teach the code-gen about structs
which need to be passed a selector by the caller for parsing.

Because structs are "topologically sorted" one pass of propagating
the selectors down is enough.

For generating netlink message we depend on the presence bits
so no selector passing needed there.

Reviewed-by: Donald Hunter <donald.hunter@gmail.com>
Link: https://patch.msgid.link/20250520161916.413298-7-kuba@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
---
 tools/net/ynl/pyynl/ynl_gen_c.py | 65 +++++++++++++++++++++++++++++---
 1 file changed, 60 insertions(+), 5 deletions(-)

diff --git a/tools/net/ynl/pyynl/ynl_gen_c.py b/tools/net/ynl/pyynl/ynl_gen_c.py
index 1f8cc34ab3f0..c1508d8c1e7a 100755
--- a/tools/net/ynl/pyynl/ynl_gen_c.py
+++ b/tools/net/ynl/pyynl/ynl_gen_c.py
@@ -685,7 +685,11 @@ class TypeNest(Type):
                             f"{self.enum_name}, {at}{var}->{self.c_name})")
 
     def _attr_get(self, ri, var):
-        get_lines = [f"if ({self.nested_render_name}_parse(&parg, attr))",
+        pns = self.family.pure_nested_structs[self.nested_attrs]
+        args = ["&parg", "attr"]
+        for sel in pns.external_selectors():
+            args.append(f'{var}->{sel.name}')
+        get_lines = [f"if ({self.nested_render_name}_parse({', '.join(args)}))",
                      "return YNL_PARSE_CB_ERROR;"]
         init_lines = [f"parg.rsp_policy = &{self.nested_render_name}_nest;",
                       f"parg.data = &{var}->{self.c_name};"]
@@ -890,15 +894,24 @@ class TypeSubMessage(TypeNest):
 
     def _attr_typol(self):
         typol = f'.type = YNL_PT_NEST, .nest = &{self.nested_render_name}_nest, '
-        typol += f'.is_submsg = 1, .selector_type = {self.attr_set[self["selector"]].value} '
+        typol += '.is_submsg = 1, '
+        # Reverse-parsing of the policy (ynl_err_walk() in ynl.c) does not
+        # support external selectors. No family uses sub-messages with external
+        # selector for requests so this is fine for now.
+        if not self.selector.is_external():
+            typol += f'.selector_type = {self.attr_set[self["selector"]].value} '
         return typol
 
     def _attr_get(self, ri, var):
         sel = c_lower(self['selector'])
-        get_lines = [f'if (!{var}->{sel})',
+        if self.selector.is_external():
+            sel_var = f"_sel_{sel}"
+        else:
+            sel_var = f"{var}->{sel}"
+        get_lines = [f'if (!{sel_var})',
                      f'return ynl_submsg_failed(yarg, "%s", "%s");' %
                         (self.name, self['selector']),
-                    f"if ({self.nested_render_name}_parse(&parg, {var}->{sel}, attr))",
+                    f"if ({self.nested_render_name}_parse(&parg, {sel_var}, attr))",
                      "return YNL_PARSE_CB_ERROR;"]
         init_lines = [f"parg.rsp_policy = &{self.nested_render_name}_nest;",
                       f"parg.data = &{var}->{self.c_name};"]
@@ -914,7 +927,15 @@ class Selector:
             self.attr.is_selector = True
             self._external = False
         else:
-            raise Exception("Passing selectors from external nests not supported")
+            # The selector will need to get passed down thru the structs
+            self.attr = None
+            self._external = True
+
+    def set_attr(self, attr):
+        self.attr = attr
+
+    def is_external(self):
+        return self._external
 
 
 class Struct:
@@ -976,6 +997,13 @@ class Struct:
             raise Exception("Inheriting different members not supported")
         self.inherited = [c_lower(x) for x in sorted(self._inherited)]
 
+    def external_selectors(self):
+        sels = []
+        for name, attr in self.attr_list:
+            if isinstance(attr, TypeSubMessage) and attr.selector.is_external():
+                sels.append(attr.selector)
+        return sels
+
     def free_needs_iter(self):
         for _, attr in self.attr_list:
             if attr.free_needs_iter():
@@ -1222,6 +1250,7 @@ class Family(SpecFamily):
         self._load_root_sets()
         self._load_nested_sets()
         self._load_attr_use()
+        self._load_selector_passing()
         self._load_hooks()
 
         self.kernel_policy = self.yaml.get('kernel-policy', 'split')
@@ -1436,6 +1465,30 @@ class Family(SpecFamily):
                 if attr in rs_members['reply']:
                     spec.set_reply()
 
+    def _load_selector_passing(self):
+        def all_structs():
+            for k, v in reversed(self.pure_nested_structs.items()):
+                yield k, v
+            for k, _ in self.root_sets.items():
+                yield k, None  # we don't have a struct, but it must be terminal
+
+        for attr_set, struct in all_structs():
+            for _, spec in self.attr_sets[attr_set].items():
+                if 'nested-attributes' in spec:
+                    child_name = spec['nested-attributes']
+                elif 'sub-message' in spec:
+                    child_name = spec.sub_message
+                else:
+                    continue
+
+                child = self.pure_nested_structs.get(child_name)
+                for selector in child.external_selectors():
+                    if selector.name in self.attr_sets[attr_set]:
+                        sel_attr = self.attr_sets[attr_set][selector.name]
+                        selector.set_attr(sel_attr)
+                    else:
+                        raise Exception("Passing selector thru more than one layer not supported")
+
     def _load_global_policy(self):
         global_set = set()
         attr_set_name = None
@@ -2183,6 +2236,8 @@ def parse_rsp_submsg(ri, struct):
 def parse_rsp_nested_prototype(ri, struct, suffix=';'):
     func_args = ['struct ynl_parse_arg *yarg',
                  'const struct nlattr *nested']
+    for sel in struct.external_selectors():
+        func_args.append('const char *_sel_' + sel.name)
     if struct.submsg:
         func_args.insert(1, 'const char *sel')
     for arg in struct.inherited:
-- 
2.51.0


From a66a170b68af0d588f4eadb9e5813c4edefe24f5 Mon Sep 17 00:00:00 2001
From: Jakub Kicinski <kuba@kernel.org>
Date: Tue, 20 May 2025 09:19:11 -0700
Subject: [PATCH 16/16] tools: ynl-gen: move fixed header info from RenderInfo
 to Struct

RenderInfo describes a request-response exchange. Struct describes
a parsed attribute set. For ease of parsing sub-messages with
fixed headers move fixed header info from RenderInfo to Struct.
No functional changes.

Reviewed-by: Donald Hunter <donald.hunter@gmail.com>
Link: https://patch.msgid.link/20250520161916.413298-8-kuba@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
---
 tools/net/ynl/pyynl/ynl_gen_c.py | 45 +++++++++++++++++++-------------
 1 file changed, 27 insertions(+), 18 deletions(-)

diff --git a/tools/net/ynl/pyynl/ynl_gen_c.py b/tools/net/ynl/pyynl/ynl_gen_c.py
index c1508d8c1e7a..bd1fadb2cf5a 100755
--- a/tools/net/ynl/pyynl/ynl_gen_c.py
+++ b/tools/net/ynl/pyynl/ynl_gen_c.py
@@ -939,7 +939,7 @@ class Selector:
 
 
 class Struct:
-    def __init__(self, family, space_name, type_list=None,
+    def __init__(self, family, space_name, type_list=None, fixed_header=None,
                  inherited=None, submsg=None):
         self.family = family
         self.space_name = space_name
@@ -947,6 +947,9 @@ class Struct:
         # Use list to catch comparisons with empty sets
         self._inherited = inherited if inherited is not None else []
         self.inherited = []
+        self.fixed_header = None
+        if fixed_header:
+            self.fixed_header = 'struct ' + c_lower(fixed_header)
         self.submsg = submsg
 
         self.nested = type_list is None
@@ -1345,7 +1348,9 @@ class Family(SpecFamily):
         nested = spec['nested-attributes']
         if nested not in self.root_sets:
             if nested not in self.pure_nested_structs:
-                self.pure_nested_structs[nested] = Struct(self, nested, inherited=inherit)
+                self.pure_nested_structs[nested] = \
+                    Struct(self, nested, inherited=inherit,
+                           fixed_header=spec.get('fixed-header'))
         else:
             raise Exception(f'Using attr set as root and nested not supported - {nested}')
 
@@ -1538,13 +1543,12 @@ class RenderInfo:
         self.op_mode = op_mode
         self.op = op
 
-        self.fixed_hdr = None
+        fixed_hdr = op.fixed_header if op else None
         self.fixed_hdr_len = 'ys->family->hdr_len'
         if op and op.fixed_header:
-            self.fixed_hdr = 'struct ' + c_lower(op.fixed_header)
             if op.fixed_header != family.fixed_header:
                 if family.is_classic():
-                    self.fixed_hdr_len = f"sizeof({self.fixed_hdr})"
+                    self.fixed_hdr_len = f"sizeof(struct {c_lower(fixed_hdr)})"
                 else:
                     raise Exception(f"Per-op fixed header not supported, yet")
 
@@ -1584,12 +1588,17 @@ class RenderInfo:
                 type_list = []
                 if op_dir in op[op_mode]:
                     type_list = op[op_mode][op_dir]['attributes']
-                self.struct[op_dir] = Struct(family, self.attr_set, type_list=type_list)
+                self.struct[op_dir] = Struct(family, self.attr_set,
+                                             fixed_header=fixed_hdr,
+                                             type_list=type_list)
         if op_mode == 'event':
-            self.struct['reply'] = Struct(family, self.attr_set, type_list=op['event']['attributes'])
+            self.struct['reply'] = Struct(family, self.attr_set,
+                                          fixed_header=fixed_hdr,
+                                          type_list=op['event']['attributes'])
 
     def type_empty(self, key):
-        return len(self.struct[key].attr_list) == 0 and self.fixed_hdr is None
+        return len(self.struct[key].attr_list) == 0 and \
+            self.struct['request'].fixed_header is None
 
     def needs_nlflags(self, direction):
         return self.op_mode == 'do' and direction == 'request' and self.family.is_classic()
@@ -2057,12 +2066,12 @@ def _multi_parse(ri, struct, init_lines, local_vars):
     if struct.nested:
         iter_line = "ynl_attr_for_each_nested(attr, nested)"
     else:
-        if ri.fixed_hdr:
+        if struct.fixed_header:
             local_vars += ['void *hdr;']
         iter_line = "ynl_attr_for_each(attr, nlh, yarg->ys->family->hdr_len)"
         if ri.op.fixed_header != ri.family.fixed_header:
             if ri.family.is_classic():
-                iter_line = f"ynl_attr_for_each(attr, nlh, sizeof({ri.fixed_hdr}))"
+                iter_line = f"ynl_attr_for_each(attr, nlh, sizeof({struct.fixed_header}))"
             else:
                 raise Exception(f"Per-op fixed header not supported, yet")
 
@@ -2104,12 +2113,12 @@ def _multi_parse(ri, struct, init_lines, local_vars):
     for arg in struct.inherited:
         ri.cw.p(f'dst->{arg} = {arg};')
 
-    if ri.fixed_hdr:
+    if struct.fixed_header:
         if ri.family.is_classic():
             ri.cw.p('hdr = ynl_nlmsg_data(nlh);')
         else:
             ri.cw.p('hdr = ynl_nlmsg_data_offset(nlh, sizeof(struct genlmsghdr));')
-        ri.cw.p(f"memcpy(&dst->_hdr, hdr, sizeof({ri.fixed_hdr}));")
+        ri.cw.p(f"memcpy(&dst->_hdr, hdr, sizeof({struct.fixed_header}));")
     for anest in sorted(all_multi):
         aspec = struct[anest]
         ri.cw.p(f"if (dst->{aspec.c_name})")
@@ -2303,7 +2312,7 @@ def print_req(ri):
         ret_err = 'NULL'
         local_vars += [f'{type_name(ri, rdir(direction))} *rsp;']
 
-    if ri.fixed_hdr:
+    if ri.struct["request"].fixed_header:
         local_vars += ['size_t hdr_len;',
                        'void *hdr;']
 
@@ -2327,7 +2336,7 @@ def print_req(ri):
         ri.cw.p(f"yrs.yarg.rsp_policy = &{ri.struct['reply'].render_name}_nest;")
     ri.cw.nl()
 
-    if ri.fixed_hdr:
+    if ri.struct['request'].fixed_header:
         ri.cw.p("hdr_len = sizeof(req->_hdr);")
         ri.cw.p("hdr = ynl_nlmsg_put_extra_header(nlh, hdr_len);")
         ri.cw.p("memcpy(hdr, &req->_hdr, hdr_len);")
@@ -2373,7 +2382,7 @@ def print_dump(ri):
                   'struct nlmsghdr *nlh;',
                   'int err;']
 
-    if ri.fixed_hdr:
+    if ri.struct['request'].fixed_header:
         local_vars += ['size_t hdr_len;',
                        'void *hdr;']
 
@@ -2394,7 +2403,7 @@ def print_dump(ri):
     else:
         ri.cw.p(f"nlh = ynl_gemsg_start_dump(ys, {ri.nl.get_family_id()}, {ri.op.enum_name}, 1);")
 
-    if ri.fixed_hdr:
+    if ri.struct['request'].fixed_header:
         ri.cw.p("hdr_len = sizeof(req->_hdr);")
         ri.cw.p("hdr = ynl_nlmsg_put_extra_header(nlh, hdr_len);")
         ri.cw.p("memcpy(hdr, &req->_hdr, hdr_len);")
@@ -2471,8 +2480,8 @@ def _print_type(ri, direction, struct):
     if ri.needs_nlflags(direction):
         ri.cw.p('__u16 _nlmsg_flags;')
         ri.cw.nl()
-    if ri.fixed_hdr:
-        ri.cw.p(ri.fixed_hdr + ' _hdr;')
+    if struct.fixed_header:
+        ri.cw.p(struct.fixed_header + ' _hdr;')
         ri.cw.nl()
 
     for type_filter in ['present', 'len', 'count']:
-- 
2.51.0