From: Christian Ehrig <cehrig@cloudflare.com>
To: bpf@vger.kernel.org
Cc: cehrig@cloudflare.com, kernel-team@cloudflare.com,
"David S. Miller" <davem@davemloft.net>,
Eric Dumazet <edumazet@google.com>,
Jakub Kicinski <kuba@kernel.org>, Paolo Abeni <pabeni@redhat.com>,
David Ahern <dsahern@kernel.org>,
linux-kernel@vger.kernel.org, netdev@vger.kernel.org
Subject: [PATCH bpf-next v3 2/3] bpf,fou: Add bpf_skb_{set,get}_fou_encap kfuncs
Date: Fri, 7 Apr 2023 15:38:54 +0200 [thread overview]
Message-ID: <e17c94a646b63e78ce0dbf3f04b2c33dc948a32d.1680874078.git.cehrig@cloudflare.com> (raw)
In-Reply-To: <cover.1680874078.git.cehrig@cloudflare.com>
Add two new kfuncs that allow a BPF tc-hook, installed on an ipip
device in collect-metadata mode, to control FOU encap parameters on a
per-packet level. The set of kfuncs is registered with the fou module.
The bpf_skb_set_fou_encap kfunc is supposed to be used in tandem and after
a successful call to the bpf_skb_set_tunnel_key bpf-helper. UDP source and
destination ports can be controlled by passing a struct bpf_fou_encap. A
source port of zero will auto-assign a source port. enum bpf_fou_encap_type
is used to specify if the egress path should FOU or GUE encap the packet.
On the ingress path bpf_skb_get_fou_encap can be used to read UDP source
and destination ports from the receiver's point of view and allows for
packet multiplexing across different destination ports within a single
BPF program and ipip device.
Signed-off-by: Christian Ehrig <cehrig@cloudflare.com>
---
include/net/fou.h | 2 +
net/ipv4/Makefile | 2 +-
net/ipv4/fou_bpf.c | 119 ++++++++++++++++++++++++++++++++++++++++++++
net/ipv4/fou_core.c | 5 ++
4 files changed, 127 insertions(+), 1 deletion(-)
create mode 100644 net/ipv4/fou_bpf.c
diff --git a/include/net/fou.h b/include/net/fou.h
index 80f56e275b08..824eb4b231fd 100644
--- a/include/net/fou.h
+++ b/include/net/fou.h
@@ -17,4 +17,6 @@ int __fou_build_header(struct sk_buff *skb, struct ip_tunnel_encap *e,
int __gue_build_header(struct sk_buff *skb, struct ip_tunnel_encap *e,
u8 *protocol, __be16 *sport, int type);
+int register_fou_bpf(void);
+
#endif
diff --git a/net/ipv4/Makefile b/net/ipv4/Makefile
index 880277c9fd07..b18ba8ef93ad 100644
--- a/net/ipv4/Makefile
+++ b/net/ipv4/Makefile
@@ -26,7 +26,7 @@ obj-$(CONFIG_IP_MROUTE) += ipmr.o
obj-$(CONFIG_IP_MROUTE_COMMON) += ipmr_base.o
obj-$(CONFIG_NET_IPIP) += ipip.o
gre-y := gre_demux.o
-fou-y := fou_core.o fou_nl.o
+fou-y := fou_core.o fou_nl.o fou_bpf.o
obj-$(CONFIG_NET_FOU) += fou.o
obj-$(CONFIG_NET_IPGRE_DEMUX) += gre.o
obj-$(CONFIG_NET_IPGRE) += ip_gre.o
diff --git a/net/ipv4/fou_bpf.c b/net/ipv4/fou_bpf.c
new file mode 100644
index 000000000000..3760a14b6b57
--- /dev/null
+++ b/net/ipv4/fou_bpf.c
@@ -0,0 +1,119 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/* Unstable Fou Helpers for TC-BPF hook
+ *
+ * These are called from SCHED_CLS BPF programs. Note that it is
+ * allowed to break compatibility for these functions since the interface they
+ * are exposed through to BPF programs is explicitly unstable.
+ */
+
+#include <linux/bpf.h>
+#include <linux/btf_ids.h>
+
+#include <net/dst_metadata.h>
+#include <net/fou.h>
+
+struct bpf_fou_encap {
+ __be16 sport;
+ __be16 dport;
+};
+
+enum bpf_fou_encap_type {
+ FOU_BPF_ENCAP_FOU,
+ FOU_BPF_ENCAP_GUE,
+};
+
+__diag_push();
+__diag_ignore_all("-Wmissing-prototypes",
+ "Global functions as their definitions will be in BTF");
+
+/* bpf_skb_set_fou_encap - Set FOU encap parameters
+ *
+ * This function allows for using GUE or FOU encapsulation together with an
+ * ipip device in collect-metadata mode.
+ *
+ * It is meant to be used in BPF tc-hooks and after a call to the
+ * bpf_skb_set_tunnel_key helper, responsible for setting IP addresses.
+ *
+ * Parameters:
+ * @skb_ctx Pointer to ctx (__sk_buff) in TC program. Cannot be NULL
+ * @encap Pointer to a `struct bpf_fou_encap` storing UDP src and
+ * dst ports. If sport is set to 0 the kernel will auto-assign a
+ * port. This is similar to using `encap-sport auto`.
+ * Cannot be NULL
+ * @type Encapsulation type for the packet. Their definitions are
+ * specified in `enum bpf_fou_encap_type`
+ */
+__bpf_kfunc int bpf_skb_set_fou_encap(struct __sk_buff *skb_ctx,
+ struct bpf_fou_encap *encap, int type)
+{
+ struct sk_buff *skb = (struct sk_buff *)skb_ctx;
+ struct ip_tunnel_info *info = skb_tunnel_info(skb);
+
+ if (unlikely(!encap))
+ return -EINVAL;
+
+ if (unlikely(!info || !(info->mode & IP_TUNNEL_INFO_TX)))
+ return -EINVAL;
+
+ switch (type) {
+ case FOU_BPF_ENCAP_FOU:
+ info->encap.type = TUNNEL_ENCAP_FOU;
+ break;
+ case FOU_BPF_ENCAP_GUE:
+ info->encap.type = TUNNEL_ENCAP_GUE;
+ break;
+ default:
+ info->encap.type = TUNNEL_ENCAP_NONE;
+ }
+
+ if (info->key.tun_flags & TUNNEL_CSUM)
+ info->encap.flags |= TUNNEL_ENCAP_FLAG_CSUM;
+
+ info->encap.sport = encap->sport;
+ info->encap.dport = encap->dport;
+
+ return 0;
+}
+
+/* bpf_skb_get_fou_encap - Get FOU encap parameters
+ *
+ * This function allows for reading encap metadata from a packet received
+ * on an ipip device in collect-metadata mode.
+ *
+ * Parameters:
+ * @skb_ctx Pointer to ctx (__sk_buff) in TC program. Cannot be NULL
+ * @encap Pointer to a struct bpf_fou_encap storing UDP source and
+ * destination port. Cannot be NULL
+ */
+__bpf_kfunc int bpf_skb_get_fou_encap(struct __sk_buff *skb_ctx,
+ struct bpf_fou_encap *encap)
+{
+ struct sk_buff *skb = (struct sk_buff *)skb_ctx;
+ struct ip_tunnel_info *info = skb_tunnel_info(skb);
+
+ if (unlikely(!info))
+ return -EINVAL;
+
+ encap->sport = info->encap.sport;
+ encap->dport = info->encap.dport;
+
+ return 0;
+}
+
+__diag_pop()
+
+BTF_SET8_START(fou_kfunc_set)
+BTF_ID_FLAGS(func, bpf_skb_set_fou_encap)
+BTF_ID_FLAGS(func, bpf_skb_get_fou_encap)
+BTF_SET8_END(fou_kfunc_set)
+
+static const struct btf_kfunc_id_set fou_bpf_kfunc_set = {
+ .owner = THIS_MODULE,
+ .set = &fou_kfunc_set,
+};
+
+int register_fou_bpf(void)
+{
+ return register_btf_kfunc_id_set(BPF_PROG_TYPE_SCHED_CLS,
+ &fou_bpf_kfunc_set);
+}
diff --git a/net/ipv4/fou_core.c b/net/ipv4/fou_core.c
index cafec9b4eee0..0c41076e31ed 100644
--- a/net/ipv4/fou_core.c
+++ b/net/ipv4/fou_core.c
@@ -1236,10 +1236,15 @@ static int __init fou_init(void)
if (ret < 0)
goto unregister;
+ ret = register_fou_bpf();
+ if (ret < 0)
+ goto kfunc_failed;
+
ret = ip_tunnel_encap_add_fou_ops();
if (ret == 0)
return 0;
+kfunc_failed:
genl_unregister_family(&fou_nl_family);
unregister:
unregister_pernet_device(&fou_net_ops);
--
2.39.2
next prev parent reply other threads:[~2023-04-07 13:40 UTC|newest]
Thread overview: 5+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-04-07 13:38 [PATCH bpf-next v3 0/3] Add FOU support for externally controlled ipip devices Christian Ehrig
2023-04-07 13:38 ` [PATCH bpf-next v3 1/3] ipip,ip_tunnel,sit: " Christian Ehrig
2023-04-07 13:38 ` Christian Ehrig [this message]
2023-04-07 13:38 ` [PATCH bpf-next v3 3/3] selftests/bpf: Test FOU kfuncs " Christian Ehrig
2023-04-12 23:50 ` [PATCH bpf-next v3 0/3] Add FOU support " patchwork-bot+netdevbpf
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=e17c94a646b63e78ce0dbf3f04b2c33dc948a32d.1680874078.git.cehrig@cloudflare.com \
--to=cehrig@cloudflare.com \
--cc=bpf@vger.kernel.org \
--cc=davem@davemloft.net \
--cc=dsahern@kernel.org \
--cc=edumazet@google.com \
--cc=kernel-team@cloudflare.com \
--cc=kuba@kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=netdev@vger.kernel.org \
--cc=pabeni@redhat.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox
all inboxes | Powered by JetHome®