Merge branch 'AF_XDP-initial-support'

Björn Töpel says:

====================
This patch set introduces a new address family called AF_XDP that is
optimized for high performance packet processing and, in upcoming
patch sets, zero-copy semantics. In this patch set, we have removed
all zero-copy related code in order to make it smaller, simpler and
hopefully more review friendly. This patch set only supports copy-mode
for the generic XDP path (XDP_SKB) for both RX and TX and copy-mode
for RX using the XDP_DRV path. Zero-copy support requires XDP and
driver changes that Jesper Dangaard Brouer is working on. Some of his
work has already been accepted. We will publish our zero-copy support
for RX and TX on top of his patch sets at a later point in time.

An AF_XDP socket (XSK) is created with the normal socket()
syscall. Associated with each XSK are two queues: the RX queue and the
TX queue. A socket can receive packets on the RX queue and it can send
packets on the TX queue. These queues are registered and sized with
the setsockopts XDP_RX_RING and XDP_TX_RING, respectively. It is
mandatory to have at least one of these queues for each socket. In
contrast to AF_PACKET V2/V3 these descriptor queues are separated from
packet buffers. An RX or TX descriptor points to a data buffer in a
memory area called a UMEM. RX and TX can share the same UMEM so that a
packet does not have to be copied between RX and TX. Moreover, if a
packet needs to be kept for a while due to a possible retransmit, the
descriptor that points to that packet can be changed to point to
another and reused right away. This again avoids copying data.

This new dedicated packet buffer area is call a UMEM. It consists of a
number of equally size frames and each frame has a unique frame id. A
descriptor in one of the queues references a frame by referencing its
frame id. The user space allocates memory for this UMEM using whatever
means it feels is most appropriate (malloc, mmap, huge pages,
etc). This memory area is then registered with the kernel using the new
setsockopt XDP_UMEM_REG. The UMEM also has two queues: the FILL queue
and the COMPLETION queue. The fill queue is used by the application to
send down frame ids for the kernel to fill in with RX packet
data. References to these frames will then appear in the RX queue of
the XSK once they have been received. The completion queue, on the
other hand, contains frame ids that the kernel has transmitted
completely and can now be used again by user space, for either TX or
RX. Thus, the frame ids appearing in the completion queue are ids that
were previously transmitted using the TX queue. In summary, the RX and
FILL queues are used for the RX path and the TX and COMPLETION queues
are used for the TX path.

The socket is then finally bound with a bind() call to a device and a
specific queue id on that device, and it is not until bind is
completed that traffic starts to flow. Note that in this patch set,
all packet data is copied out to user-space.

A new feature in this patch set is that the UMEM can be shared between
processes, if desired. If a process wants to do this, it simply skips
the registration of the UMEM and its corresponding two queues, sets a
flag in the bind call and submits the XSK of the process it would like
to share UMEM with as well as its own newly created XSK socket. The
new process will then receive frame id references in its own RX queue
that point to this shared UMEM. Note that since the queue structures
are single-consumer / single-producer (for performance reasons), the
new process has to create its own socket with associated RX and TX
queues, since it cannot share this with the other process. This is
also the reason that there is only one set of FILL and COMPLETION
queues per UMEM. It is the responsibility of a single process to
handle the UMEM. If multiple-producer / multiple-consumer queues are
implemented in the future, this requirement could be relaxed.

How is then packets distributed between these two XSK? We have
introduced a new BPF map called XSKMAP (or BPF_MAP_TYPE_XSKMAP in
full). The user-space application can place an XSK at an arbitrary
place in this map. The XDP program can then redirect a packet to a
specific index in this map and at this point XDP validates that the
XSK in that map was indeed bound to that device and queue number. If
not, the packet is dropped. If the map is empty at that index, the
packet is also dropped. This also means that it is currently mandatory
to have an XDP program loaded (and one XSK in the XSKMAP) to be able
to get any traffic to user space through the XSK.

AF_XDP can operate in two different modes: XDP_SKB and XDP_DRV. If the
driver does not have support for XDP, or XDP_SKB is explicitly chosen
when loading the XDP program, XDP_SKB mode is employed that uses SKBs
together with the generic XDP support and copies out the data to user
space. A fallback mode that works for any network device. On the other
hand, if the driver has support for XDP, it will be used by the AF_XDP
code to provide better performance, but there is still a copy of the
data into user space.

There is a xdpsock benchmarking/test application included that
demonstrates how to use AF_XDP sockets with both private and shared
UMEMs. Say that you would like your UDP traffic from port 4242 to end
up in queue 16, that we will enable AF_XDP on. Here, we use ethtool
for this:

      ethtool -N p3p2 rx-flow-hash udp4 fn
      ethtool -N p3p2 flow-type udp4 src-port 4242 dst-port 4242 \
          action 16

Running the rxdrop benchmark in XDP_DRV mode can then be done
using:

      samples/bpf/xdpsock -i p3p2 -q 16 -r -N

For XDP_SKB mode, use the switch "-S" instead of "-N" and all options
can be displayed with "-h", as usual.

We have run some benchmarks on a dual socket system with two Broadwell
E5 2660 @ 2.0 GHz with hyperthreading turned off. Each socket has 14
cores which gives a total of 28, but only two cores are used in these
experiments. One for TR/RX and one for the user space application. The
memory is DDR4 @ 2133 MT/s (1067 MHz) and the size of each DIMM is
8192MB and with 8 of those DIMMs in the system we have 64 GB of total
memory. The compiler used is gcc (Ubuntu 7.3.0-16ubuntu3) 7.3.0. The
NIC is Intel I40E 40Gbit/s using the i40e driver.

Below are the results in Mpps of the I40E NIC benchmark runs for 64
and 1500 byte packets, generated by a commercial packet generator HW
outputing packets at full 40 Gbit/s line rate. The results are without
retpoline so that we can compare against previous numbers. With
retpoline, the AF_XDP numbers drop with between 10 - 15 percent.

AF_XDP performance 64 byte packets. Results from V2 in parenthesis.
Benchmark   XDP_SKB   XDP_DRV
rxdrop       2.9(3.0)   9.6(9.5)
txpush       2.6(2.5)   NA*
l2fwd        1.9(1.9)   2.5(2.5) (TX using XDP_SKB in both cases)

AF_XDP performance 1500 byte packets:
Benchmark   XDP_SKB   XDP_DRV
rxdrop       2.1(2.2)   3.3(3.3)
l2fwd        1.4(1.4)   1.8(1.8) (TX using XDP_SKB in both cases)

* NA since we have no support for TX using the XDP_DRV infrastructure
  in this patch set. This is for a future patch set since it involves
  changes to the XDP NDOs. Some of this has been upstreamed by Jesper
  Dangaard Brouer.

XDP performance on our system as a base line:

64 byte packets:
XDP stats       CPU     pps         issue-pps
XDP-RX CPU      16      32.3(32.9)M  0

1500 byte packets:
XDP stats       CPU     pps         issue-pps
XDP-RX CPU      16      3.3(3.3)M    0

Changes from V2:

* Fixed a race in XSKMAP map found by Will. The code has been
  completely rearchitected and is now simpler, faster, and hopefully
  also not racy. Please review and check if it holds.

If you would like to diff V2 against V3, you can find them here:
https://github.com/bjoto/linux/tree/af-xdp-v2-on-bpf-next
https://github.com/bjoto/linux/tree/af-xdp-v3-on-bpf-next

The structure of the patch set is as follows:

Patches 1-3: Basic socket and umem plumbing
Patches 4-9: RX support together with the new XSKMAP
Patches 10-13: TX support
Patch 14: Statistics support with getsockopt()
Patch 15: Sample application

We based this patch set on bpf-next commit a3fe1f6 ("tools:
bpftool: change time format for program 'loaded at:' information")

To do for this patch set:

* Syzkaller torture session being worked on

Post-series plan:

* Optimize performance

* Kernel selftest

* Kernel load module support of AF_XDP would be nice. Unclear how to
  achieve this though since our XDP code depends on net/core.

* Support for AF_XDP sockets without an XPD program loaded. In this
  case all the traffic on a queue should go up to the user space socket.

* Daniel Borkmann's suggestion for a "copy to XDP socket, and return
  XDP_PASS" for a tcpdump-like functionality.

* And of course getting to zero-copy support in small increments,
  starting with TX then adding RX.

Thanks: Björn and Magnus
====================

Acked-by: Willem de Bruijn <[email protected]>
Acked-by: David S. Miller <[email protected]>
Acked-by: Daniel Borkmann <[email protected]>
Signed-off-by: Alexei Starovoitov <[email protected]>

Author: Alexei Starovoitov

Documentation/networking/af_xdp.rst

@@ -0,0 +1,297 @@
+.. SPDX-License-Identifier: GPL-2.0
+
+======
+AF_XDP
+======
+
+Overview
+========
+
+AF_XDP is an address family that is optimized for high performance
+packet processing.
+
+This document assumes that the reader is familiar with BPF and XDP. If
+not, the Cilium project has an excellent reference guide at
+http://cilium.readthedocs.io/en/doc-1.0/bpf/.
+
+Using the XDP_REDIRECT action from an XDP program, the program can
+redirect ingress frames to other XDP enabled netdevs, using the
+bpf_redirect_map() function. AF_XDP sockets enable the possibility for
+XDP programs to redirect frames to a memory buffer in a user-space
+application.
+
+An AF_XDP socket (XSK) is created with the normal socket()
+syscall. Associated with each XSK are two rings: the RX ring and the
+TX ring. A socket can receive packets on the RX ring and it can send
+packets on the TX ring. These rings are registered and sized with the
+setsockopts XDP_RX_RING and XDP_TX_RING, respectively. It is mandatory
+to have at least one of these rings for each socket. An RX or TX
+descriptor ring points to a data buffer in a memory area called a
+UMEM. RX and TX can share the same UMEM so that a packet does not have
+to be copied between RX and TX. Moreover, if a packet needs to be kept
+for a while due to a possible retransmit, the descriptor that points
+to that packet can be changed to point to another and reused right
+away. This again avoids copying data.
+
+The UMEM consists of a number of equally size frames and each frame
+has a unique frame id. A descriptor in one of the rings references a
+frame by referencing its frame id. The user space allocates memory for
+this UMEM using whatever means it feels is most appropriate (malloc,
+mmap, huge pages, etc). This memory area is then registered with the
+kernel using the new setsockopt XDP_UMEM_REG. The UMEM also has two
+rings: the FILL ring and the COMPLETION ring. The fill ring is used by
+the application to send down frame ids for the kernel to fill in with
+RX packet data. References to these frames will then appear in the RX
+ring once each packet has been received. The completion ring, on the
+other hand, contains frame ids that the kernel has transmitted
+completely and can now be used again by user space, for either TX or
+RX. Thus, the frame ids appearing in the completion ring are ids that
+were previously transmitted using the TX ring. In summary, the RX and
+FILL rings are used for the RX path and the TX and COMPLETION rings
+are used for the TX path.
+
+The socket is then finally bound with a bind() call to a device and a
+specific queue id on that device, and it is not until bind is
+completed that traffic starts to flow.
+
+The UMEM can be shared between processes, if desired. If a process
+wants to do this, it simply skips the registration of the UMEM and its
+corresponding two rings, sets the XDP_SHARED_UMEM flag in the bind
+call and submits the XSK of the process it would like to share UMEM
+with as well as its own newly created XSK socket. The new process will
+then receive frame id references in its own RX ring that point to this
+shared UMEM. Note that since the ring structures are single-consumer /
+single-producer (for performance reasons), the new process has to
+create its own socket with associated RX and TX rings, since it cannot
+share this with the other process. This is also the reason that there
+is only one set of FILL and COMPLETION rings per UMEM. It is the
+responsibility of a single process to handle the UMEM.
+
+How is then packets distributed from an XDP program to the XSKs? There
+is a BPF map called XSKMAP (or BPF_MAP_TYPE_XSKMAP in full). The
+user-space application can place an XSK at an arbitrary place in this
+map. The XDP program can then redirect a packet to a specific index in
+this map and at this point XDP validates that the XSK in that map was
+indeed bound to that device and ring number. If not, the packet is
+dropped. If the map is empty at that index, the packet is also
+dropped. This also means that it is currently mandatory to have an XDP
+program loaded (and one XSK in the XSKMAP) to be able to get any
+traffic to user space through the XSK.
+
+AF_XDP can operate in two different modes: XDP_SKB and XDP_DRV. If the
+driver does not have support for XDP, or XDP_SKB is explicitly chosen
+when loading the XDP program, XDP_SKB mode is employed that uses SKBs
+together with the generic XDP support and copies out the data to user
+space. A fallback mode that works for any network device. On the other
+hand, if the driver has support for XDP, it will be used by the AF_XDP
+code to provide better performance, but there is still a copy of the
+data into user space.
+
+Concepts
+========
+
+In order to use an AF_XDP socket, a number of associated objects need
+to be setup.
+
+Jonathan Corbet has also written an excellent article on LWN,
+"Accelerating networking with AF_XDP". It can be found at
+https://lwn.net/Articles/750845/.
+
+UMEM
+----
+
+UMEM is a region of virtual contiguous memory, divided into
+equal-sized frames. An UMEM is associated to a netdev and a specific
+queue id of that netdev. It is created and configured (frame size,
+frame headroom, start address and size) by using the XDP_UMEM_REG
+setsockopt system call. A UMEM is bound to a netdev and queue id, via
+the bind() system call.
+
+An AF_XDP is socket linked to a single UMEM, but one UMEM can have
+multiple AF_XDP sockets. To share an UMEM created via one socket A,
+the next socket B can do this by setting the XDP_SHARED_UMEM flag in
+struct sockaddr_xdp member sxdp_flags, and passing the file descriptor
+of A to struct sockaddr_xdp member sxdp_shared_umem_fd.
+
+The UMEM has two single-producer/single-consumer rings, that are used
+to transfer ownership of UMEM frames between the kernel and the
+user-space application.
+
+Rings
+-----
+
+There are a four different kind of rings: Fill, Completion, RX and
+TX. All rings are single-producer/single-consumer, so the user-space
+application need explicit synchronization of multiple
+processes/threads are reading/writing to them.
+
+The UMEM uses two rings: Fill and Completion. Each socket associated
+with the UMEM must have an RX queue, TX queue or both. Say, that there
+is a setup with four sockets (all doing TX and RX). Then there will be
+one Fill ring, one Completion ring, four TX rings and four RX rings.
+
+The rings are head(producer)/tail(consumer) based rings. A producer
+writes the data ring at the index pointed out by struct xdp_ring
+producer member, and increasing the producer index. A consumer reads
+the data ring at the index pointed out by struct xdp_ring consumer
+member, and increasing the consumer index.
+
+The rings are configured and created via the _RING setsockopt system
+calls and mmapped to user-space using the appropriate offset to mmap()
+(XDP_PGOFF_RX_RING, XDP_PGOFF_TX_RING, XDP_UMEM_PGOFF_FILL_RING and
+XDP_UMEM_PGOFF_COMPLETION_RING).
+
+The size of the rings need to be of size power of two.
+
+UMEM Fill Ring
+~~~~~~~~~~~~~~
+
+The Fill ring is used to transfer ownership of UMEM frames from
+user-space to kernel-space. The UMEM indicies are passed in the
+ring. As an example, if the UMEM is 64k and each frame is 4k, then the
+UMEM has 16 frames and can pass indicies between 0 and 15.
+
+Frames passed to the kernel are used for the ingress path (RX rings).
+
+The user application produces UMEM indicies to this ring.
+
+UMEM Completetion Ring
+~~~~~~~~~~~~~~~~~~~~~~
+
+The Completion Ring is used transfer ownership of UMEM frames from
+kernel-space to user-space. Just like the Fill ring, UMEM indicies are
+used.
+
+Frames passed from the kernel to user-space are frames that has been
+sent (TX ring) and can be used by user-space again.
+
+The user application consumes UMEM indicies from this ring.
+
+
+RX Ring
+~~~~~~~
+
+The RX ring is the receiving side of a socket. Each entry in the ring
+is a struct xdp_desc descriptor. The descriptor contains UMEM index
+(idx), the length of the data (len), the offset into the frame
+(offset).
+
+If no frames have been passed to kernel via the Fill ring, no
+descriptors will (or can) appear on the RX ring.
+
+The user application consumes struct xdp_desc descriptors from this
+ring.
+
+TX Ring
+~~~~~~~
+
+The TX ring is used to send frames. The struct xdp_desc descriptor is
+filled (index, length and offset) and passed into the ring.
+
+To start the transfer a sendmsg() system call is required. This might
+be relaxed in the future.
+
+The user application produces struct xdp_desc descriptors to this
+ring.
+
+XSKMAP / BPF_MAP_TYPE_XSKMAP
+----------------------------
+
+On XDP side there is a BPF map type BPF_MAP_TYPE_XSKMAP (XSKMAP) that
+is used in conjunction with bpf_redirect_map() to pass the ingress
+frame to a socket.
+
+The user application inserts the socket into the map, via the bpf()
+system call.
+
+Note that if an XDP program tries to redirect to a socket that does
+not match the queue configuration and netdev, the frame will be
+dropped. E.g. an AF_XDP socket is bound to netdev eth0 and
+queue 17. Only the XDP program executing for eth0 and queue 17 will
+successfully pass data to the socket. Please refer to the sample
+application (samples/bpf/) in for an example.
+
+Usage
+=====
+
+In order to use AF_XDP sockets there are two parts needed. The
+user-space application and the XDP program. For a complete setup and
+usage example, please refer to the sample application. The user-space
+side is xdpsock_user.c and the XDP side xdpsock_kern.c.
+
+Naive ring dequeue and enqueue could look like this::
+
+    // typedef struct xdp_rxtx_ring RING;
+    // typedef struct xdp_umem_ring RING;
+
+    // typedef struct xdp_desc RING_TYPE;
+    // typedef __u32 RING_TYPE;
+
+    int dequeue_one(RING *ring, RING_TYPE *item)
+    {
+        __u32 entries = ring->ptrs.producer - ring->ptrs.consumer;
+
+        if (entries == 0)
+            return -1;
+
+        // read-barrier!
+
+        *item = ring->desc[ring->ptrs.consumer & (RING_SIZE - 1)];
+        ring->ptrs.consumer++;
+        return 0;
+    }
+
+    int enqueue_one(RING *ring, const RING_TYPE *item)
+    {
+        u32 free_entries = RING_SIZE - (ring->ptrs.producer - ring->ptrs.consumer);
+
+        if (free_entries == 0)
+            return -1;
+
+        ring->desc[ring->ptrs.producer & (RING_SIZE - 1)] = *item;
+
+        // write-barrier!
+
+        ring->ptrs.producer++;
+        return 0;
+    }
+
+
+For a more optimized version, please refer to the sample application.
+
+Sample application
+==================
+
+There is a xdpsock benchmarking/test application included that
+demonstrates how to use AF_XDP sockets with both private and shared
+UMEMs. Say that you would like your UDP traffic from port 4242 to end
+up in queue 16, that we will enable AF_XDP on. Here, we use ethtool
+for this::
+
+      ethtool -N p3p2 rx-flow-hash udp4 fn
+      ethtool -N p3p2 flow-type udp4 src-port 4242 dst-port 4242 \
+          action 16
+
+Running the rxdrop benchmark in XDP_DRV mode can then be done
+using::
+
+      samples/bpf/xdpsock -i p3p2 -q 16 -r -N
+
+For XDP_SKB mode, use the switch "-S" instead of "-N" and all options
+can be displayed with "-h", as usual.
+
+Credits
+=======
+
+- Björn Töpel (AF_XDP core)
+- Magnus Karlsson (AF_XDP core)
+- Alexander Duyck
+- Alexei Starovoitov
+- Daniel Borkmann
+- Jesper Dangaard Brouer
+- John Fastabend
+- Jonathan Corbet (LWN coverage)
+- Michael S. Tsirkin
+- Qi Z Zhang
+- Willem de Bruijn
+

Documentation/networking/index.rst

@@ -6,6 +6,7 @@ Contents:
 .. toctree::
    :maxdepth: 2
 
+   af_xdp
    batman-adv
    can
    dpaa2/index

MAINTAINERS

@@ -15424,6 +15424,14 @@ T:	git git://linuxtv.org/media_tree.git
 S:	Maintained
 F:	drivers/media/tuners/tuner-xc2028.*
 
+XDP SOCKETS (AF_XDP)
+M:	Björn Töpel <[email protected]>
+M:	Magnus Karlsson <[email protected]>
+L:	[email protected]
+S:	Maintained
+F:	kernel/bpf/xskmap.c
+F:	net/xdp/
+
 XEN BLOCK SUBSYSTEM
 M:	Konrad Rzeszutek Wilk <[email protected]>
 M:	Roger Pau Monné <[email protected]>

include/linux/bpf.h

@@ -676,6 +676,31 @@ static inline int sock_map_prog(struct bpf_map *map,
 }
 #endif
 
+#if defined(CONFIG_XDP_SOCKETS)
+struct xdp_sock;
+struct xdp_sock *__xsk_map_lookup_elem(struct bpf_map *map, u32 key);
+int __xsk_map_redirect(struct bpf_map *map, struct xdp_buff *xdp,
+		       struct xdp_sock *xs);
+void __xsk_map_flush(struct bpf_map *map);
+#else
+struct xdp_sock;
+static inline struct xdp_sock *__xsk_map_lookup_elem(struct bpf_map *map,
+						     u32 key)
+{
+	return NULL;
+}
+
+static inline int __xsk_map_redirect(struct bpf_map *map, struct xdp_buff *xdp,
+				     struct xdp_sock *xs)
+{
+	return -EOPNOTSUPP;
+}
+
+static inline void __xsk_map_flush(struct bpf_map *map)
+{
+}
+#endif
+
 /* verifier prototypes for helper functions called from eBPF programs */
 extern const struct bpf_func_proto bpf_map_lookup_elem_proto;
 extern const struct bpf_func_proto bpf_map_update_elem_proto;

include/linux/bpf_types.h

@@ -49,4 +49,7 @@ BPF_MAP_TYPE(BPF_MAP_TYPE_DEVMAP, dev_map_ops)
 BPF_MAP_TYPE(BPF_MAP_TYPE_SOCKMAP, sock_map_ops)
 #endif
 BPF_MAP_TYPE(BPF_MAP_TYPE_CPUMAP, cpu_map_ops)
+#if defined(CONFIG_XDP_SOCKETS)
+BPF_MAP_TYPE(BPF_MAP_TYPE_XSKMAP, xsk_map_ops)
+#endif
 #endif

include/linux/filter.h

@@ -760,7 +760,7 @@ struct bpf_prog *bpf_patch_insn_single(struct bpf_prog *prog, u32 off,
  * This does not appear to be a real limitation for existing software.
  */
 int xdp_do_generic_redirect(struct net_device *dev, struct sk_buff *skb,
-			    struct bpf_prog *prog);
+			    struct xdp_buff *xdp, struct bpf_prog *prog);
 int xdp_do_redirect(struct net_device *dev,
 		    struct xdp_buff *xdp,
 		    struct bpf_prog *prog);

include/linux/netdevice.h

@@ -2486,6 +2486,7 @@ void dev_disable_lro(struct net_device *dev);
 int dev_loopback_xmit(struct net *net, struct sock *sk, struct sk_buff *newskb);
 int dev_queue_xmit(struct sk_buff *skb);
 int dev_queue_xmit_accel(struct sk_buff *skb, void *accel_priv);
+int dev_direct_xmit(struct sk_buff *skb, u16 queue_id);
 int register_netdevice(struct net_device *dev);
 void unregister_netdevice_queue(struct net_device *dev, struct list_head *head);
 void unregister_netdevice_many(struct list_head *head);