在 2021/9/24 15:35, wangjie (L) 写道:
On 2021/9/22 22:56, Jian Shen wrote:
Use netdev_feature_xxx helpers to replace the logical operation for netdev features.
Signed-off-by: Jian Shen shenjian15@huawei.com
include/net/udp.h | 8 ++++++-- include/net/udp_tunnel.h | 12 ++++++++---- net/ipv4/af_inet.c | 3 ++- net/ipv4/esp4_offload.c | 28 +++++++++++++++++----------- net/ipv4/gre_offload.c | 8 +++++--- net/ipv4/ip_gre.c | 36 ++++++++++++++++++++++-------------- net/ipv4/ip_output.c | 23 +++++++++++++++-------- net/ipv4/ip_tunnel.c | 4 +++- net/ipv4/ip_vti.c | 2 +- net/ipv4/ipip.c | 6 +++--- net/ipv4/ipmr.c | 2 +- net/ipv4/tcp_offload.c | 5 ++++- net/ipv4/udp_offload.c | 28 +++++++++++++++++----------- 13 files changed, 104 insertions(+), 61 deletions(-)
diff --git a/include/net/udp.h b/include/net/udp.h index 360df454356c..3855b9775cbf 100644 --- a/include/net/udp.h +++ b/include/net/udp.h @@ -481,14 +481,18 @@ void udpv6_encap_enable(void); static inline struct sk_buff *udp_rcv_segment(struct sock *sk, struct sk_buff *skb, bool ipv4) { - netdev_features_t features = NETIF_F_SG; + netdev_features_t features; struct sk_buff *segs;
+ netdev_feature_zero(&features); + netdev_feature_set_bit(NETIF_F_SG_BIT, &features);
/* Avoid csum recalculation by skb_segment unless userspace explicitly * asks for the final checksum values */ if (!inet_get_convert_csum(sk)) - features |= NETIF_F_IP_CSUM | NETIF_F_IPV6_CSUM; + netdev_feature_set_bits(NETIF_F_IP_CSUM | NETIF_F_IPV6_CSUM, + &features);
/* UDP segmentation expects packets of type CHECKSUM_PARTIAL or * CHECKSUM_NONE in __udp_gso_segment. UDP GRO indeed builds partial diff --git a/include/net/udp_tunnel.h b/include/net/udp_tunnel.h index afc7ce713657..68c1a540a733 100644 --- a/include/net/udp_tunnel.h +++ b/include/net/udp_tunnel.h @@ -129,7 +129,8 @@ void udp_tunnel_notify_del_rx_port(struct socket *sock, unsigned short type); static inline void udp_tunnel_get_rx_info(struct net_device *dev) { ASSERT_RTNL(); - if (!(dev->features & NETIF_F_RX_UDP_TUNNEL_PORT)) + if (!netdev_feature_test_bit(NETIF_F_RX_UDP_TUNNEL_PORT_BIT, + dev->features)) return; call_netdevice_notifiers(NETDEV_UDP_TUNNEL_PUSH_INFO, dev); } @@ -137,7 +138,8 @@ static inline void udp_tunnel_get_rx_info(struct net_device *dev) static inline void udp_tunnel_drop_rx_info(struct net_device *dev) { ASSERT_RTNL(); - if (!(dev->features & NETIF_F_RX_UDP_TUNNEL_PORT)) + if (!netdev_feature_test_bit(NETIF_F_RX_UDP_TUNNEL_PORT_BIT, + dev->features)) return; call_netdevice_notifiers(NETDEV_UDP_TUNNEL_DROP_INFO, dev); } @@ -326,7 +328,8 @@ udp_tunnel_nic_set_port_priv(struct net_device *dev, unsigned int table, static inline void udp_tunnel_nic_add_port(struct net_device *dev, struct udp_tunnel_info *ti) { - if (!(dev->features & NETIF_F_RX_UDP_TUNNEL_PORT)) + if (!netdev_feature_test_bit(NETIF_F_RX_UDP_TUNNEL_PORT_BIT, + dev->features)) return; if (udp_tunnel_nic_ops) udp_tunnel_nic_ops->add_port(dev, ti); @@ -335,7 +338,8 @@ udp_tunnel_nic_add_port(struct net_device *dev, struct udp_tunnel_info *ti) static inline void udp_tunnel_nic_del_port(struct net_device *dev, struct udp_tunnel_info *ti) { - if (!(dev->features & NETIF_F_RX_UDP_TUNNEL_PORT)) + if (!netdev_feature_test_bit(NETIF_F_RX_UDP_TUNNEL_PORT_BIT, + dev->features)) return; if (udp_tunnel_nic_ops) udp_tunnel_nic_ops->del_port(dev, ti); diff --git a/net/ipv4/af_inet.c b/net/ipv4/af_inet.c index 1d816a5fd3eb..3b5c16041990 100644 --- a/net/ipv4/af_inet.c +++ b/net/ipv4/af_inet.c @@ -1361,7 +1361,8 @@ struct sk_buff *inet_gso_segment(struct sk_buff *skb,
encap = SKB_GSO_CB(skb)->encap_level > 0; if (encap) - features &= skb->dev->hw_enc_features; + netdev_feature_and(&features, features, + skb->dev->hw_enc_features); SKB_GSO_CB(skb)->encap_level += ihl;
skb_reset_transport_header(skb); diff --git a/net/ipv4/esp4_offload.c b/net/ipv4/esp4_offload.c index 8e4e9aa12130..f7e77a9176d6 100644 --- a/net/ipv4/esp4_offload.c +++ b/net/ipv4/esp4_offload.c @@ -189,8 +189,8 @@ static struct sk_buff *esp4_gso_segment(struct sk_buff *skb, struct xfrm_state *x; struct ip_esp_hdr *esph; struct crypto_aead *aead; - netdev_features_t esp_features = features; struct xfrm_offload *xo = xfrm_offload(skb); + netdev_features_t esp_features; struct sec_path *sp;
if (!xo) @@ -214,14 +214,19 @@ static struct sk_buff *esp4_gso_segment(struct sk_buff *skb,
skb->encap_hdr_csum = 1;
- if ((!(skb->dev->gso_partial_features & NETIF_F_HW_ESP) && - !(features & NETIF_F_HW_ESP)) || x->xso.dev != skb->dev) - esp_features = features & ~(NETIF_F_SG | NETIF_F_CSUM_MASK | - NETIF_F_SCTP_CRC); - else if (!(features & NETIF_F_HW_ESP_TX_CSUM) && - !(skb->dev->gso_partial_features & NETIF_F_HW_ESP_TX_CSUM)) - esp_features = features & ~(NETIF_F_CSUM_MASK | - NETIF_F_SCTP_CRC); + netdev_feature_copy(&esp_features, features); + if ((!netdev_feature_test_bit(NETIF_F_HW_ESP_BIT, + skb->dev->gso_partial_features) && + !netdev_feature_test_bit(NETIF_F_HW_ESP_BIT, features)) || + x->xso.dev != skb->dev) + netdev_feature_clear_bits(NETIF_F_SG | NETIF_F_CSUM_MASK | + NETIF_F_SCTP_CRC, &esp_features);
the origin line use "features & ~(NETIF_F_CSUM_MASK | NETIF_F_SCTP_CRC)", the replaced line use esp_features directly?
I have call netdev_feature_copy before this, so esp_features is the same as features.
+ else if (!netdev_feature_test_bit(NETIF_F_HW_ESP_TX_CSUM_BIT, + features) && + !netdev_feature_test_bit(NETIF_F_HW_ESP_TX_CSUM_BIT, + skb->dev->gso_partial_features)) + netdev_feature_clear_bits(NETIF_F_CSUM_MASK | NETIF_F_SCTP_CRC, + &esp_features);
the origin line use "features", the replaced line use "esp_features" directly?
same as above.
xo->flags |= XFRM_GSO_SEGMENT;
@@ -261,8 +266,9 @@ static int esp_xmit(struct xfrm_state *x, struct sk_buff *skb, netdev_features_ if (!xo) return -EINVAL;
- if ((!(features & NETIF_F_HW_ESP) && - !(skb->dev->gso_partial_features & NETIF_F_HW_ESP)) || + if ((!netdev_feature_test_bit(NETIF_F_HW_ESP_BIT, features) && + !netdev_feature_test_bit(NETIF_F_HW_ESP_BIT, + skb->dev->gso_partial_features)) || x->xso.dev != skb->dev) { xo->flags |= CRYPTO_FALLBACK; hw_offload = false; diff --git a/net/ipv4/gre_offload.c b/net/ipv4/gre_offload.c index 1121a9d5fed9..65f22fa749be 100644 --- a/net/ipv4/gre_offload.c +++ b/net/ipv4/gre_offload.c @@ -43,14 +43,16 @@ static struct sk_buff *gre_gso_segment(struct sk_buff *skb, need_csum = !!(skb_shinfo(skb)->gso_type & SKB_GSO_GRE_CSUM); skb->encap_hdr_csum = need_csum;
- features &= skb->dev->hw_enc_features; + netdev_feature_and(&features, features, skb->dev->hw_enc_features);
if (need_csum) - features &= ~NETIF_F_SCTP_CRC; + netdev_feature_clear_bit(NETIF_F_SCTP_CRC_BIT, &features);
need_ipsec = skb_dst(skb) && dst_xfrm(skb_dst(skb)); /* Try to offload checksum if possible */ offload_csum = !!(need_csum && !need_ipsec && - (skb->dev->features & NETIF_F_HW_CSUM)); + netdev_feature_test_bit(NETIF_F_HW_CSUM_BIT, + skb->dev->features));
/* segment inner packet. */ segs = skb_mac_gso_segment(skb, features); diff --git a/net/ipv4/ip_gre.c b/net/ipv4/ip_gre.c index 0fe6c936dc54..e083ef6a8454 100644 --- a/net/ipv4/ip_gre.c +++ b/net/ipv4/ip_gre.c @@ -767,16 +767,22 @@ static void ipgre_link_update(struct net_device *dev, bool set_mtu) if (!(tunnel->parms.o_flags & TUNNEL_SEQ)) { if (!(tunnel->parms.o_flags & TUNNEL_CSUM) || tunnel->encap.type == TUNNEL_ENCAP_NONE) { - dev->features |= NETIF_F_GSO_SOFTWARE; - dev->hw_features |= NETIF_F_GSO_SOFTWARE; + netdev_feature_set_bits(NETIF_F_GSO_SOFTWARE, + &dev->features); + netdev_feature_set_bits(NETIF_F_GSO_SOFTWARE, + &dev->hw_features); } else { - dev->features &= ~NETIF_F_GSO_SOFTWARE; - dev->hw_features &= ~NETIF_F_GSO_SOFTWARE; + netdev_feature_clear_bits(NETIF_F_GSO_SOFTWARE, + &dev->features); + netdev_feature_clear_bits(NETIF_F_GSO_SOFTWARE, + &dev->hw_features); } - dev->features |= NETIF_F_LLTX; + netdev_feature_set_bit(NETIF_F_LLTX_BIT, &dev->features); } else { - dev->hw_features &= ~NETIF_F_GSO_SOFTWARE; - dev->features &= ~(NETIF_F_LLTX | NETIF_F_GSO_SOFTWARE); + netdev_feature_clear_bits(NETIF_F_GSO_SOFTWARE, + &dev->hw_features); + netdev_feature_clear_bits(NETIF_F_LLTX | NETIF_F_GSO_SOFTWARE, + &dev->features); } }
@@ -958,8 +964,8 @@ static void __gre_tunnel_init(struct net_device *dev) tunnel->hlen = tunnel->tun_hlen + tunnel->encap_hlen; dev->needed_headroom = tunnel->hlen + sizeof(tunnel->parms.iph);
- dev->features |= GRE_FEATURES; - dev->hw_features |= GRE_FEATURES; + netdev_feature_set_bits(GRE_FEATURES, &dev->features); + netdev_feature_set_bits(GRE_FEATURES, &dev->hw_features);
if (!(tunnel->parms.o_flags & TUNNEL_SEQ)) { /* TCP offload with GRE SEQ is not supported, nor @@ -968,14 +974,16 @@ static void __gre_tunnel_init(struct net_device *dev) */ if (!(tunnel->parms.o_flags & TUNNEL_CSUM) || (tunnel->encap.type == TUNNEL_ENCAP_NONE)) { - dev->features |= NETIF_F_GSO_SOFTWARE; - dev->hw_features |= NETIF_F_GSO_SOFTWARE; + netdev_feature_set_bits(NETIF_F_GSO_SOFTWARE, + &dev->features); + netdev_feature_set_bits(NETIF_F_GSO_SOFTWARE, + &dev->hw_features); }
/* Can use a lockless transmit, unless we generate * output sequences */ - dev->features |= NETIF_F_LLTX; + netdev_feature_set_bit(NETIF_F_LLTX_BIT, &dev->features); } }
@@ -1303,8 +1311,8 @@ static int erspan_tunnel_init(struct net_device *dev) tunnel->hlen = tunnel->tun_hlen + tunnel->encap_hlen + erspan_hdr_len(tunnel->erspan_ver);
- dev->features |= GRE_FEATURES; - dev->hw_features |= GRE_FEATURES; + netdev_feature_set_bits(GRE_FEATURES, &dev->features); + netdev_feature_set_bits(GRE_FEATURES, &dev->hw_features); dev->priv_flags |= IFF_LIVE_ADDR_CHANGE; netif_keep_dst(dev);
diff --git a/net/ipv4/ip_output.c b/net/ipv4/ip_output.c index 8d552f1b7f62..9f7cba9c9498 100644 --- a/net/ipv4/ip_output.c +++ b/net/ipv4/ip_output.c @@ -257,7 +257,8 @@ static int ip_finish_output_gso(struct net *net, struct sock *sk, */ netif_skb_features(skb, &features); BUILD_BUG_ON(sizeof(*IPCB(skb)) > SKB_GSO_CB_OFFSET); - segs = skb_gso_segment(skb, features & ~NETIF_F_GSO_MASK); + netdev_feature_clear_bits(NETIF_F_GSO_MASK, &features); + segs = skb_gso_segment(skb, features); if (IS_ERR_OR_NULL(segs)) { kfree_skb(skb); return -ENOMEM; @@ -996,9 +997,11 @@ static int __ip_append_data(struct sock *sk, */ if (transhdrlen && length + fragheaderlen <= mtu && - rt->dst.dev->features & (NETIF_F_HW_CSUM | NETIF_F_IP_CSUM) && + netdev_feature_test_bits(NETIF_F_HW_CSUM | NETIF_F_IP_CSUM, + rt->dst.dev->features) && (!(flags & MSG_MORE) || cork->gso_size) && - (!exthdrlen || (rt->dst.dev->features & NETIF_F_HW_ESP_TX_CSUM))) + (!exthdrlen || netdev_feature_test_bits(NETIF_F_HW_ESP_TX_CSUM_BIT, + rt->dst.dev->features)))
netdev_feature_test_bits -> netdev_feature_test_bit ?
yes, will fix it, thanks
csummode = CHECKSUM_PARTIAL;
if (flags & MSG_ZEROCOPY && length && sock_flag(sk, SOCK_ZEROCOPY)) { @@ -1006,7 +1009,8 @@ static int __ip_append_data(struct sock *sk, if (!uarg) return -ENOBUFS; extra_uref = !skb_zcopy(skb); /* only ref on new uarg */ - if (rt->dst.dev->features & NETIF_F_SG && + if (netdev_feature_test_bits(NETIF_F_SG_BIT, + rt->dst.dev->features) &&
netdev_feature_test_bits -> netdev_feature_test_bit?
yes, will fix it, thanks
csummode == CHECKSUM_PARTIAL) { paged = true; } else { @@ -1069,11 +1073,13 @@ static int __ip_append_data(struct sock *sk, alloc_extra += rt->dst.trailer_len;
if ((flags & MSG_MORE) && - !(rt->dst.dev->features&NETIF_F_SG)) + !netdev_feature_test_bits(NETIF_F_SG_BIT, + rt->dst.dev->features))
netdev_feature_test_bits -> netdev_feature_test_bit?
yes, will fix it, thanks
alloclen = mtu; else if (!paged && (fraglen + alloc_extra < SKB_MAX_ALLOC || - !(rt->dst.dev->features & NETIF_F_SG))) + !netdev_feature_test_bits(NETIF_F_SG_BIT, + rt->dst.dev->features)))
netdev_feature_test_bits -> netdev_feature_test_bit?
yes, will fix it, thanks
alloclen = fraglen; else { alloclen = min_t(int, fraglen, MAX_HEADER); @@ -1161,7 +1167,8 @@ static int __ip_append_data(struct sock *sk, if (copy > length) copy = length;
- if (!(rt->dst.dev->features&NETIF_F_SG) && + if (!netdev_feature_test_bits(NETIF_F_SG_BIT, + rt->dst.dev->features) &&
netdev_feature_test_bits -> netdev_feature_test_bit?
yes, will fix it, thanks
skb_tailroom(skb) >= copy) { unsigned int off;
@@ -1340,7 +1347,7 @@ ssize_t ip_append_page(struct sock *sk, struct flowi4 *fl4, struct page *page, if (cork->flags & IPCORK_OPT) opt = cork->opt;
- if (!(rt->dst.dev->features & NETIF_F_SG)) + if (!netdev_feature_test_bits(NETIF_F_SG_BIT, rt->dst.dev->features))
netdev_feature_test_bits -> netdev_feature_test_bit?
yes, will fix it, thanks
return -EOPNOTSUPP;
hh_len = LL_RESERVED_SPACE(rt->dst.dev); diff --git a/net/ipv4/ip_tunnel.c b/net/ipv4/ip_tunnel.c index fe9101d3d69e..ac5671e5ea28 100644 --- a/net/ipv4/ip_tunnel.c +++ b/net/ipv4/ip_tunnel.c @@ -1072,7 +1072,9 @@ int ip_tunnel_init_net(struct net *net, unsigned int ip_tnl_net_id, * Allowing to move it to another netns is clearly unsafe. */ if (!IS_ERR(itn->fb_tunnel_dev)) { - itn->fb_tunnel_dev->features |= NETIF_F_NETNS_LOCAL; + netdev_feature_set_bit(NETIF_F_NETNS_LOCAL_BIT,
- &itn->fb_tunnel_dev->features);
itn->fb_tunnel_dev->mtu = ip_tunnel_bind_dev(itn->fb_tunnel_dev); ip_tunnel_add(itn, netdev_priv(itn->fb_tunnel_dev)); itn->type = itn->fb_tunnel_dev->type; diff --git a/net/ipv4/ip_vti.c b/net/ipv4/ip_vti.c index efe25a0172e6..bf94ef6a1f2c 100644 --- a/net/ipv4/ip_vti.c +++ b/net/ipv4/ip_vti.c @@ -430,7 +430,7 @@ static int vti_tunnel_init(struct net_device *dev)
dev->flags = IFF_NOARP; dev->addr_len = 4; - dev->features |= NETIF_F_LLTX; + netdev_feature_set_bit(NETIF_F_LLTX_BIT, &dev->features); netif_keep_dst(dev);
return ip_tunnel_init(dev); diff --git a/net/ipv4/ipip.c b/net/ipv4/ipip.c index 3aa78ccbec3e..b587d472fad5 100644 --- a/net/ipv4/ipip.c +++ b/net/ipv4/ipip.c @@ -368,11 +368,11 @@ static void ipip_tunnel_setup(struct net_device *dev) dev->type = ARPHRD_TUNNEL; dev->flags = IFF_NOARP; dev->addr_len = 4; - dev->features |= NETIF_F_LLTX; + netdev_feature_set_bit(NETIF_F_LLTX_BIT, &dev->features); netif_keep_dst(dev);
- dev->features |= IPIP_FEATURES; - dev->hw_features |= IPIP_FEATURES; + netdev_feature_set_bits(IPIP_FEATURES, &dev->features); + netdev_feature_set_bits(IPIP_FEATURES, &dev->hw_features); ip_tunnel_setup(dev, ipip_net_id); }
diff --git a/net/ipv4/ipmr.c b/net/ipv4/ipmr.c index 2dda856ca260..750a4a102014 100644 --- a/net/ipv4/ipmr.c +++ b/net/ipv4/ipmr.c @@ -532,7 +532,7 @@ static void reg_vif_setup(struct net_device *dev) dev->flags = IFF_NOARP; dev->netdev_ops = ®_vif_netdev_ops; dev->needs_free_netdev = true; - dev->features |= NETIF_F_NETNS_LOCAL; + netdev_feature_set_bit(NETIF_F_NETNS_LOCAL_BIT, &dev->features); }
static struct net_device *ipmr_reg_vif(struct net *net, struct mr_table *mrt) diff --git a/net/ipv4/tcp_offload.c b/net/ipv4/tcp_offload.c index fc61cd3fea65..4bf0a0d382f6 100644 --- a/net/ipv4/tcp_offload.c +++ b/net/ipv4/tcp_offload.c @@ -65,6 +65,7 @@ struct sk_buff *tcp_gso_segment(struct sk_buff *skb, struct sk_buff *gso_skb = skb; __sum16 newcheck; bool ooo_okay, copy_destructor; + netdev_features_t tmp;
th = tcp_hdr(skb); thlen = th->doff * 4; @@ -81,7 +82,9 @@ struct sk_buff *tcp_gso_segment(struct sk_buff *skb, if (unlikely(skb->len <= mss)) goto out;
- if (skb_gso_ok(skb, features | NETIF_F_GSO_ROBUST)) { + netdev_feature_copy(&tmp, features); + netdev_feature_set_bit(NETIF_F_GSO_ROBUST_BIT, &tmp); + if (skb_gso_ok(skb, tmp)) { /* Packet is from an untrusted source, reset gso_segs. */
skb_shinfo(skb)->gso_segs = DIV_ROUND_UP(skb->len, mss); diff --git a/net/ipv4/udp_offload.c b/net/ipv4/udp_offload.c index 86d32a1e62ac..e4c14e142bde 100644 --- a/net/ipv4/udp_offload.c +++ b/net/ipv4/udp_offload.c @@ -63,22 +63,26 @@ static struct sk_buff *__skb_udp_tunnel_segment(struct sk_buff *skb, /* Try to offload checksum if possible */ offload_csum = !!(need_csum && !need_ipsec && - (skb->dev->features & - (is_ipv6 ? (NETIF_F_HW_CSUM | NETIF_F_IPV6_CSUM) : - (NETIF_F_HW_CSUM | NETIF_F_IP_CSUM))));
- features &= skb->dev->hw_enc_features; + is_ipv6 ? + netdev_feature_test_bits(NETIF_F_HW_CSUM | + NETIF_F_IPV6_CSUM, + skb->dev->features) : + netdev_feature_test_bits(NETIF_F_HW_CSUM | + NETIF_F_IP_CSUM, + skb->dev->features));
+ netdev_feature_and(&features, features, skb->dev->hw_enc_features); if (need_csum) - features &= ~NETIF_F_SCTP_CRC; + netdev_feature_clear_bit(NETIF_F_SCTP_CRC_BIT, &features);
/* The only checksum offload we care about from here on out is the * outer one so strip the existing checksum feature flags and * instead set the flag based on our outer checksum offload value. */ if (remcsum) { - features &= ~NETIF_F_CSUM_MASK; + netdev_feature_clear_bits(NETIF_F_CSUM_MASK, &features); if (!need_csum || offload_csum) - features |= NETIF_F_HW_CSUM; + netdev_feature_set_bit(NETIF_F_HW_CSUM_BIT, &features); }
/* segment inner packet. */ @@ -414,7 +418,7 @@ static struct sk_buff *udp4_ufo_fragment(struct sk_buff *skb, * software prior to segmenting the frame. */ if (!skb->encap_hdr_csum) - features |= NETIF_F_HW_CSUM; + netdev_feature_set_bit(NETIF_F_HW_CSUM_BIT, &features);
/* Fragment the skb. IP headers of the fragments are updated in * inet_gso_segment() @@ -520,10 +524,12 @@ struct sk_buff *udp_gro_receive(struct list_head *head, struct sk_buff *skb, */ NAPI_GRO_CB(skb)->is_flist = 0; if (!sk || !udp_sk(sk)->gro_receive) { - if (skb->dev->features & NETIF_F_GRO_FRAGLIST) + if (netdev_feature_test_bit(NETIF_F_GRO_FRAGLIST_BIT, + skb->dev->features)) NAPI_GRO_CB(skb)->is_flist = sk ? !udp_sk(sk)->gro_enabled : 1;
- if ((!sk && (skb->dev->features & NETIF_F_GRO_UDP_FWD)) || + if ((!sk && netdev_feature_test_bit(NETIF_F_GRO_UDP_FWD_BIT, + skb->dev->features)) || (sk && udp_sk(sk)->gro_enabled) || NAPI_GRO_CB(skb)->is_flist) return call_gro_receive(udp_gro_receive_segment, head, skb);
.