Merge branch 'xfrm: IPsec hardware offload performance improvements'

Jianbo Liu says:

====================
This patch series optimizes IPsec crypto offload performance by
addressing a lock contention bottleneck using RSS.

The first patch refactors the xfrm_input to avoid a costly
unlock/relock cycle.

The second patch builds on this by removing a redundant replay check,
which is unnecessary for the synchronous hardware path.
====================

Signed-off-by: Steffen Klassert <steffen.klassert@secunet.com>
This commit is contained in:
Steffen Klassert 2025-10-28 07:46:28 +01:00
commit 7197e080de
1 changed files with 15 additions and 15 deletions

View File

@ -505,6 +505,7 @@ int xfrm_input(struct sk_buff *skb, int nexthdr, __be32 spi, int encap_type)
async = 1; async = 1;
dev_put(skb->dev); dev_put(skb->dev);
seq = XFRM_SKB_CB(skb)->seq.input.low; seq = XFRM_SKB_CB(skb)->seq.input.low;
spin_lock(&x->lock);
goto resume; goto resume;
} }
/* GRO call */ /* GRO call */
@ -541,9 +542,11 @@ int xfrm_input(struct sk_buff *skb, int nexthdr, __be32 spi, int encap_type)
XFRM_INC_STATS(net, LINUX_MIB_XFRMINHDRERROR); XFRM_INC_STATS(net, LINUX_MIB_XFRMINHDRERROR);
goto drop; goto drop;
} }
nexthdr = x->type_offload->input_tail(x, skb);
} }
goto lock; goto process;
} }
family = XFRM_SPI_SKB_CB(skb)->family; family = XFRM_SPI_SKB_CB(skb)->family;
@ -611,7 +614,12 @@ int xfrm_input(struct sk_buff *skb, int nexthdr, __be32 spi, int encap_type)
goto drop; goto drop;
} }
lock: process:
seq_hi = htonl(xfrm_replay_seqhi(x, seq));
XFRM_SKB_CB(skb)->seq.input.low = seq;
XFRM_SKB_CB(skb)->seq.input.hi = seq_hi;
spin_lock(&x->lock); spin_lock(&x->lock);
if (unlikely(x->km.state != XFRM_STATE_VALID)) { if (unlikely(x->km.state != XFRM_STATE_VALID)) {
@ -638,21 +646,13 @@ int xfrm_input(struct sk_buff *skb, int nexthdr, __be32 spi, int encap_type)
goto drop_unlock; goto drop_unlock;
} }
spin_unlock(&x->lock);
if (xfrm_tunnel_check(skb, x, family)) { if (xfrm_tunnel_check(skb, x, family)) {
XFRM_INC_STATS(net, LINUX_MIB_XFRMINSTATEMODEERROR); XFRM_INC_STATS(net, LINUX_MIB_XFRMINSTATEMODEERROR);
goto drop; goto drop_unlock;
} }
seq_hi = htonl(xfrm_replay_seqhi(x, seq)); if (!crypto_done) {
spin_unlock(&x->lock);
XFRM_SKB_CB(skb)->seq.input.low = seq;
XFRM_SKB_CB(skb)->seq.input.hi = seq_hi;
if (crypto_done) {
nexthdr = x->type_offload->input_tail(x, skb);
} else {
dev_hold(skb->dev); dev_hold(skb->dev);
nexthdr = x->type->input(x, skb); nexthdr = x->type->input(x, skb);
@ -660,9 +660,9 @@ int xfrm_input(struct sk_buff *skb, int nexthdr, __be32 spi, int encap_type)
return 0; return 0;
dev_put(skb->dev); dev_put(skb->dev);
spin_lock(&x->lock);
} }
resume: resume:
spin_lock(&x->lock);
if (nexthdr < 0) { if (nexthdr < 0) {
if (nexthdr == -EBADMSG) { if (nexthdr == -EBADMSG) {
xfrm_audit_state_icvfail(x, skb, xfrm_audit_state_icvfail(x, skb,
@ -676,7 +676,7 @@ int xfrm_input(struct sk_buff *skb, int nexthdr, __be32 spi, int encap_type)
/* only the first xfrm gets the encap type */ /* only the first xfrm gets the encap type */
encap_type = 0; encap_type = 0;
if (xfrm_replay_recheck(x, skb, seq)) { if (!crypto_done && xfrm_replay_recheck(x, skb, seq)) {
XFRM_INC_STATS(net, LINUX_MIB_XFRMINSTATESEQERROR); XFRM_INC_STATS(net, LINUX_MIB_XFRMINSTATESEQERROR);
goto drop_unlock; goto drop_unlock;
} }