[PATCH net-next] xfrm: reduce struct sec_path size

[PATCH net-next] xfrm: reduce struct sec_path size

[Paolo Abeni]

The mentioned struct has an hole and uses unnecessary wide type to
store MAC length and indexes of very small arrays.

It's also embedded into the skb_extensions, and the latter, due
to recent CAN changes, may exceeds the 192 bytes mark (3 cachelines
on x86_64 arch) on some reasonable configurations.

Reordering and the sec_path fields, shrinking xfrm_offload.orig_mac_len
to 16 bits and xfrm_offload.{len,olen,verified_cnt} to u8, we can save
16 bytes and keep skb_extensions size under control.

Before:

struct sec_path {
	int                        len;
	int                        olen;
	int                        verified_cnt;

	/* XXX 4 bytes hole, try to pack */$
	struct xfrm_state *        xvec[6];
	struct xfrm_offload ovec[1];

	/* size: 88, cachelines: 2, members: 5 */
	/* sum members: 84, holes: 1, sum holes: 4 */
	/* last cacheline: 24 bytes */
};

After:

struct sec_path {
	struct xfrm_state *        xvec[6];
	struct xfrm_offload        ovec[1];
	/* typedef u8 -> __u8 */ unsigned char              len;
	/* typedef u8 -> __u8 */ unsigned char              olen;
	/* typedef u8 -> __u8 */ unsigned char              verified_cnt;

	/* size: 72, cachelines: 2, members: 5 */
	/* padding: 1 */
	/* last cacheline: 8 bytes */
};

Signed-off-by: Paolo Abeni <pabeni@redhat.com>
---
RFC -> v1:
  - shrink lens fields - thanks Florian
- RFC: https://lore.kernel.org/netdev/a2552992ca2fa29754340eaa05b1a185fca7a381.1770309837.git.pabeni@redhat.com/
---
 include/net/xfrm.h | 10 +++++-----
 1 file changed, 5 insertions(+), 5 deletions(-)

diff --git a/include/net/xfrm.h b/include/net/xfrm.h
index 0a14daaa5dd4..10d3edde6b2f 100644
--- a/include/net/xfrm.h
+++ b/include/net/xfrm.h
@@ -1156,19 +1156,19 @@ struct xfrm_offload {
 #define CRYPTO_INVALID_PROTOCOL			128
 
 	/* Used to keep whole l2 header for transport mode GRO */
-	__u32			orig_mac_len;
+	__u16			orig_mac_len;
 
 	__u8			proto;
 	__u8			inner_ipproto;
 };
 
 struct sec_path {
-	int			len;
-	int			olen;
-	int			verified_cnt;
-
 	struct xfrm_state	*xvec[XFRM_MAX_DEPTH];
 	struct xfrm_offload	ovec[XFRM_MAX_OFFLOAD_DEPTH];
+
+	u8			len;
+	u8			olen;
+	u8			verified_cnt;
 };
 
 struct sec_path *secpath_set(struct sk_buff *skb);
-- 
2.52.0

Re: [PATCH net-next] xfrm: reduce struct sec_path size

[Florian Westphal]

Paolo Abeni <pabeni@redhat.com> wrote:
> The mentioned struct has an hole and uses unnecessary wide type to
> store MAC length and indexes of very small arrays.
> 
> It's also embedded into the skb_extensions, and the latter, due
> to recent CAN changes, may exceeds the 192 bytes mark (3 cachelines
> on x86_64 arch) on some reasonable configurations.
> 
> Reordering and the sec_path fields, shrinking xfrm_offload.orig_mac_len
> to 16 bits and xfrm_offload.{len,olen,verified_cnt} to u8, we can save
> 16 bytes and keep skb_extensions size under control.

Reviewed-by: Florian Westphal <fw@strlen.de>

Re: [PATCH net-next] xfrm: reduce struct sec_path size

[Steffen Klassert]

On Fri, Feb 06, 2026 at 06:14:36PM +0100, Paolo Abeni wrote:
> The mentioned struct has an hole and uses unnecessary wide type to
> store MAC length and indexes of very small arrays.
> 
> It's also embedded into the skb_extensions, and the latter, due
> to recent CAN changes, may exceeds the 192 bytes mark (3 cachelines
> on x86_64 arch) on some reasonable configurations.
> 
> Reordering and the sec_path fields, shrinking xfrm_offload.orig_mac_len
> to 16 bits and xfrm_offload.{len,olen,verified_cnt} to u8, we can save
> 16 bytes and keep skb_extensions size under control.
> 
> Before:
> 
> struct sec_path {
> 	int                        len;
> 	int                        olen;
> 	int                        verified_cnt;
> 
> 	/* XXX 4 bytes hole, try to pack */$
> 	struct xfrm_state *        xvec[6];
> 	struct xfrm_offload ovec[1];
> 
> 	/* size: 88, cachelines: 2, members: 5 */
> 	/* sum members: 84, holes: 1, sum holes: 4 */
> 	/* last cacheline: 24 bytes */
> };
> 
> After:
> 
> struct sec_path {
> 	struct xfrm_state *        xvec[6];
> 	struct xfrm_offload        ovec[1];
> 	/* typedef u8 -> __u8 */ unsigned char              len;
> 	/* typedef u8 -> __u8 */ unsigned char              olen;
> 	/* typedef u8 -> __u8 */ unsigned char              verified_cnt;
> 
> 	/* size: 72, cachelines: 2, members: 5 */
> 	/* padding: 1 */
> 	/* last cacheline: 8 bytes */
> };
> 
> Signed-off-by: Paolo Abeni <pabeni@redhat.com>

Reviewed-by: Steffen Klassert <steffen.klassert@secunet.com>

Re: [PATCH net-next] xfrm: reduce struct sec_path size

[patchwork-bot+netdevbpf]

Hello:

This patch was applied to netdev/net-next.git (main)
by Jakub Kicinski <kuba@kernel.org>:

On Fri,  6 Feb 2026 18:14:36 +0100 you wrote:
> The mentioned struct has an hole and uses unnecessary wide type to
> store MAC length and indexes of very small arrays.
> 
> It's also embedded into the skb_extensions, and the latter, due
> to recent CAN changes, may exceeds the 192 bytes mark (3 cachelines
> on x86_64 arch) on some reasonable configurations.
> 
> [...]

Here is the summary with links:
  - [net-next] xfrm: reduce struct sec_path size
    https://git.kernel.org/netdev/net-next/c/dc010e1b4bb6

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