From patchwork Thu Jan 12 22:35:07 2017 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: "Samudrala, Sridhar" X-Patchwork-Id: 714692 X-Patchwork-Delegate: jeffrey.t.kirsher@intel.com Return-Path: X-Original-To: incoming@patchwork.ozlabs.org Delivered-To: patchwork-incoming@bilbo.ozlabs.org Received: from silver.osuosl.org (smtp3.osuosl.org [140.211.166.136]) (using TLSv1.2 with cipher AECDH-AES256-SHA (256/256 bits)) (No client certificate requested) by ozlabs.org (Postfix) with ESMTPS id 3v00vt3Yvgz9srY for ; Fri, 13 Jan 2017 09:35:38 +1100 (AEDT) Received: from localhost (localhost [127.0.0.1]) by silver.osuosl.org (Postfix) with ESMTP id 90DB42DC91; Thu, 12 Jan 2017 22:35:36 +0000 (UTC) X-Virus-Scanned: amavisd-new at osuosl.org Received: from silver.osuosl.org ([127.0.0.1]) by localhost (.osuosl.org [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id c6+82J7GaLqe; Thu, 12 Jan 2017 22:35:31 +0000 (UTC) Received: from ash.osuosl.org (ash.osuosl.org [140.211.166.34]) by silver.osuosl.org (Postfix) with ESMTP id 2A49F308C1; Thu, 12 Jan 2017 22:35:28 +0000 (UTC) X-Original-To: intel-wired-lan@lists.osuosl.org Delivered-To: intel-wired-lan@lists.osuosl.org Received: from silver.osuosl.org (smtp3.osuosl.org [140.211.166.136]) by ash.osuosl.org (Postfix) with ESMTP id 3014D1C001E for ; Thu, 12 Jan 2017 22:35:20 +0000 (UTC) Received: from localhost (localhost [127.0.0.1]) by silver.osuosl.org (Postfix) with ESMTP id 2973C2E134 for ; Thu, 12 Jan 2017 22:35:20 +0000 (UTC) X-Virus-Scanned: amavisd-new at osuosl.org Received: from silver.osuosl.org ([127.0.0.1]) by localhost (.osuosl.org [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id TclWNX8+NboH for ; Thu, 12 Jan 2017 22:35:16 +0000 (UTC) X-Greylist: domain auto-whitelisted by SQLgrey-1.7.6 Received: from mga01.intel.com (mga01.intel.com [192.55.52.88]) by silver.osuosl.org (Postfix) with ESMTPS id 6B0D02DC8B for ; Thu, 12 Jan 2017 22:35:16 +0000 (UTC) Received: from fmsmga003.fm.intel.com ([10.253.24.29]) by fmsmga101.fm.intel.com with ESMTP; 12 Jan 2017 14:35:16 -0800 X-ExtLoop1: 1 X-IronPort-AV: E=Sophos;i="5.33,220,1477983600"; d="scan'208";a="808217575" Received: from sri-mi-02.jf.intel.com ([10.166.188.51]) by FMSMGA003.fm.intel.com with ESMTP; 12 Jan 2017 14:35:16 -0800 From: Sridhar Samudrala To: alexander.h.duyck@intel.com, john.r.fastabend@intel.com, anjali.singhai@intel.com, jakub.kicinski@netronome.com, davem@davemloft.net, scott.d.peterson@intel.com, gerlitz.or@gmail.com, jiri@resnulli.us, intel-wired-lan@lists.osuosl.org, netdev@vger.kernel.org Date: Thu, 12 Jan 2017 14:35:07 -0800 Message-Id: <1484260510-9162-5-git-send-email-sridhar.samudrala@intel.com> X-Mailer: git-send-email 2.5.5 In-Reply-To: <1484260510-9162-1-git-send-email-sridhar.samudrala@intel.com> References: <1484260510-9162-1-git-send-email-sridhar.samudrala@intel.com> Subject: [Intel-wired-lan] [next-queue v4 PATCH 4/7] net: store port/representator id in metadata_dst X-BeenThere: intel-wired-lan@lists.osuosl.org X-Mailman-Version: 2.1.18-1 Precedence: list List-Id: Intel Wired Ethernet Linux Kernel Driver Development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , MIME-Version: 1.0 Errors-To: intel-wired-lan-bounces@lists.osuosl.org Sender: "Intel-wired-lan" From: Jakub Kicinski Switches and modern SR-IOV enabled NICs may multiplex traffic from Port representators and control messages over single set of hardware queues. Control messages and muxed traffic may need ordered delivery. Those requirements make it hard to comfortably use TC infrastructure today unless we have a way of attaching metadata to skbs at the upper device. Because single set of queues is used for many netdevs stopping TC/sched queues of all of them reliably is impossible and lower device has to retreat to returning NETDEV_TX_BUSY and usually has to take extra locks on the fastpath. This patch attempts to enable port/representative devs to attach metadata to skbs which carry port id. This way representatives can be queueless and all queuing can be performed at the lower netdev in the usual way. Traffic arriving on the port/representative interfaces will be have metadata attached and will subsequently be queued to the lower device for transmission. The lower device should recognize the metadata and translate it to HW specific format which is most likely either a special header inserted before the network headers or descriptor/metadata fields. Metadata is associated with the lower device by storing the netdev pointer along with port id so that if TC decides to redirect or mirror the new netdev will not try to interpret it. This is mostly for SR-IOV devices since switches don't have lower netdevs today. Signed-off-by: Jakub Kicinski Signed-off-by: Sridhar Samudrala Tested-by: Andrew Bowers --- include/net/dst_metadata.h | 41 ++++++++++++++++++++++++++++++++--------- net/core/dst.c | 15 ++++++++++----- net/core/filter.c | 1 + net/ipv4/ip_tunnel_core.c | 6 ++++-- net/openvswitch/flow_netlink.c | 4 +++- 5 files changed, 50 insertions(+), 17 deletions(-) diff --git a/include/net/dst_metadata.h b/include/net/dst_metadata.h index 701fc81..a803129 100644 --- a/include/net/dst_metadata.h +++ b/include/net/dst_metadata.h @@ -5,10 +5,22 @@ #include #include +enum metadata_type { + METADATA_IP_TUNNEL, + METADATA_HW_PORT_MUX, +}; + +struct hw_port_info { + struct net_device *lower_dev; + u32 port_id; +}; + struct metadata_dst { struct dst_entry dst; + enum metadata_type type; union { struct ip_tunnel_info tun_info; + struct hw_port_info port_info; } u; }; @@ -27,7 +39,7 @@ static inline struct ip_tunnel_info *skb_tunnel_info(struct sk_buff *skb) struct metadata_dst *md_dst = skb_metadata_dst(skb); struct dst_entry *dst; - if (md_dst) + if (md_dst && md_dst->type == METADATA_IP_TUNNEL) return &md_dst->u.tun_info; dst = skb_dst(skb); @@ -55,22 +67,33 @@ static inline int skb_metadata_dst_cmp(const struct sk_buff *skb_a, a = (const struct metadata_dst *) skb_dst(skb_a); b = (const struct metadata_dst *) skb_dst(skb_b); - if (!a != !b || a->u.tun_info.options_len != b->u.tun_info.options_len) + if (!a != !b || a->type != b->type) return 1; - return memcmp(&a->u.tun_info, &b->u.tun_info, - sizeof(a->u.tun_info) + a->u.tun_info.options_len); + switch (a->type) { + case METADATA_HW_PORT_MUX: + return memcmp(&a->u.port_info, &b->u.port_info, + sizeof(a->u.port_info)); + case METADATA_IP_TUNNEL: + return memcmp(&a->u.tun_info, &b->u.tun_info, + sizeof(a->u.tun_info) + + a->u.tun_info.options_len); + default: + return 1; + } } void metadata_dst_free(struct metadata_dst *); -struct metadata_dst *metadata_dst_alloc(u8 optslen, gfp_t flags); -struct metadata_dst __percpu *metadata_dst_alloc_percpu(u8 optslen, gfp_t flags); +struct metadata_dst *metadata_dst_alloc(u8 optslen, enum metadata_type type, + gfp_t flags); +struct metadata_dst __percpu * +metadata_dst_alloc_percpu(u8 optslen, enum metadata_type type, gfp_t flags); static inline struct metadata_dst *tun_rx_dst(int md_size) { struct metadata_dst *tun_dst; - tun_dst = metadata_dst_alloc(md_size, GFP_ATOMIC); + tun_dst = metadata_dst_alloc(md_size, METADATA_IP_TUNNEL, GFP_ATOMIC); if (!tun_dst) return NULL; @@ -85,11 +108,11 @@ static inline struct metadata_dst *tun_dst_unclone(struct sk_buff *skb) int md_size; struct metadata_dst *new_md; - if (!md_dst) + if (!md_dst || md_dst->type != METADATA_IP_TUNNEL) return ERR_PTR(-EINVAL); md_size = md_dst->u.tun_info.options_len; - new_md = metadata_dst_alloc(md_size, GFP_ATOMIC); + new_md = metadata_dst_alloc(md_size, METADATA_IP_TUNNEL, GFP_ATOMIC); if (!new_md) return ERR_PTR(-ENOMEM); diff --git a/net/core/dst.c b/net/core/dst.c index b5cbbe0..62dd4e4 100644 --- a/net/core/dst.c +++ b/net/core/dst.c @@ -367,7 +367,9 @@ static int dst_md_discard(struct sk_buff *skb) return 0; } -static void __metadata_dst_init(struct metadata_dst *md_dst, u8 optslen) +static void __metadata_dst_init(struct metadata_dst *md_dst, + enum metadata_type type, u8 optslen) + { struct dst_entry *dst; @@ -379,9 +381,11 @@ static void __metadata_dst_init(struct metadata_dst *md_dst, u8 optslen) dst->output = dst_md_discard_out; memset(dst + 1, 0, sizeof(*md_dst) + optslen - sizeof(*dst)); + md_dst->type = type; } -struct metadata_dst *metadata_dst_alloc(u8 optslen, gfp_t flags) +struct metadata_dst *metadata_dst_alloc(u8 optslen, enum metadata_type type, + gfp_t flags) { struct metadata_dst *md_dst; @@ -389,7 +393,7 @@ struct metadata_dst *metadata_dst_alloc(u8 optslen, gfp_t flags) if (!md_dst) return NULL; - __metadata_dst_init(md_dst, optslen); + __metadata_dst_init(md_dst, type, optslen); return md_dst; } @@ -403,7 +407,8 @@ void metadata_dst_free(struct metadata_dst *md_dst) kfree(md_dst); } -struct metadata_dst __percpu *metadata_dst_alloc_percpu(u8 optslen, gfp_t flags) +struct metadata_dst __percpu * +metadata_dst_alloc_percpu(u8 optslen, enum metadata_type type, gfp_t flags) { int cpu; struct metadata_dst __percpu *md_dst; @@ -414,7 +419,7 @@ struct metadata_dst __percpu *metadata_dst_alloc_percpu(u8 optslen, gfp_t flags) return NULL; for_each_possible_cpu(cpu) - __metadata_dst_init(per_cpu_ptr(md_dst, cpu), optslen); + __metadata_dst_init(per_cpu_ptr(md_dst, cpu), type, optslen); return md_dst; } diff --git a/net/core/filter.c b/net/core/filter.c index 1969b3f..617ca0c 100644 --- a/net/core/filter.c +++ b/net/core/filter.c @@ -2521,6 +2521,7 @@ bpf_get_skb_set_tunnel_proto(enum bpf_func_id which) * that is holding verifier mutex. */ md_dst = metadata_dst_alloc_percpu(IP_TUNNEL_OPTS_MAX, + METADATA_IP_TUNNEL, GFP_KERNEL); if (!md_dst) return NULL; diff --git a/net/ipv4/ip_tunnel_core.c b/net/ipv4/ip_tunnel_core.c index 5476110..4544b35 100644 --- a/net/ipv4/ip_tunnel_core.c +++ b/net/ipv4/ip_tunnel_core.c @@ -134,10 +134,12 @@ struct metadata_dst *iptunnel_metadata_reply(struct metadata_dst *md, struct metadata_dst *res; struct ip_tunnel_info *dst, *src; - if (!md || md->u.tun_info.mode & IP_TUNNEL_INFO_TX) + if (!md || md->type != METADATA_IP_TUNNEL || + md->u.tun_info.mode & IP_TUNNEL_INFO_TX) + return NULL; - res = metadata_dst_alloc(0, flags); + res = metadata_dst_alloc(0, METADATA_IP_TUNNEL, flags); if (!res) return NULL; diff --git a/net/openvswitch/flow_netlink.c b/net/openvswitch/flow_netlink.c index c87d359..164b4f1 100644 --- a/net/openvswitch/flow_netlink.c +++ b/net/openvswitch/flow_netlink.c @@ -2105,7 +2105,9 @@ static int validate_and_copy_set_tun(const struct nlattr *attr, if (start < 0) return start; - tun_dst = metadata_dst_alloc(key.tun_opts_len, GFP_KERNEL); + tun_dst = metadata_dst_alloc(key.tun_opts_len, METADATA_IP_TUNNEL, + GFP_KERNEL); + if (!tun_dst) return -ENOMEM;