bpf: add initial bpf tracepoints

This work adds a number of tracepoints to paths that are either
considered slow-path or exception-like states, where monitoring or
inspecting them would be desirable.

For bpf(2) syscall, tracepoints have been placed for main commands
when they succeed. In XDP case, tracepoint is for exceptions, that
is, f.e. on abnormal BPF program exit such as unknown or XDP_ABORTED
return code, or when error occurs during XDP_TX action and the packet
could not be forwarded.

Both have been split into separate event headers, and can be further
extended. Worst case, if they unexpectedly should get into our way in
future, they can also removed [1]. Of course, these tracepoints (like
any other) can be analyzed by eBPF itself, etc. Example output:

  # ./perf record -a -e bpf:* sleep 10
  # ./perf script
  sock_example  6197 [005]   283.980322:      bpf:bpf_map_create: map type=ARRAY ufd=4 key=4 val=8 max=256 flags=0
  sock_example  6197 [005]   283.980721:       bpf:bpf_prog_load: prog=a5ea8fa30ea6849c type=SOCKET_FILTER ufd=5
  sock_example  6197 [005]   283.988423:   bpf:bpf_prog_get_type: prog=a5ea8fa30ea6849c type=SOCKET_FILTER
  sock_example  6197 [005]   283.988443: bpf:bpf_map_lookup_elem: map type=ARRAY ufd=4 key=[06 00 00 00] val=[00 00 00 00 00 00 00 00]
  [...]
  sock_example  6197 [005]   288.990868: bpf:bpf_map_lookup_elem: map type=ARRAY ufd=4 key=[01 00 00 00] val=[14 00 00 00 00 00 00 00]
       swapper     0 [005]   289.338243:    bpf:bpf_prog_put_rcu: prog=a5ea8fa30ea6849c type=SOCKET_FILTER

  [1] https://lwn.net/Articles/705270/

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

Author: borkmann

drivers/net/ethernet/mellanox/mlx4/en_rx.c

@@ -33,6 +33,7 @@
 
 #include <net/busy_poll.h>
 #include <linux/bpf.h>
+#include <linux/bpf_trace.h>
 #include <linux/mlx4/cq.h>
 #include <linux/slab.h>
 #include <linux/mlx4/qp.h>
@@ -926,10 +927,12 @@ int mlx4_en_process_rx_cq(struct net_device *dev, struct mlx4_en_cq *cq, int bud
 							length, cq->ring,
 							&doorbell_pending)))
 					goto consumed;
+				trace_xdp_exception(dev, xdp_prog, act);
 				goto xdp_drop_no_cnt; /* Drop on xmit failure */
 			default:
 				bpf_warn_invalid_xdp_action(act);
 			case XDP_ABORTED:
+				trace_xdp_exception(dev, xdp_prog, act);
 			case XDP_DROP:
 				ring->xdp_drop++;
 xdp_drop_no_cnt:

drivers/net/ethernet/mellanox/mlx5/core/en_rx.c

@@ -33,6 +33,7 @@
 #include <linux/ip.h>
 #include <linux/ipv6.h>
 #include <linux/tcp.h>
+#include <linux/bpf_trace.h>
 #include <net/busy_poll.h>
 #include "en.h"
 #include "en_tc.h"
@@ -640,7 +641,7 @@ static inline void mlx5e_xmit_xdp_doorbell(struct mlx5e_sq *sq)
 	mlx5e_tx_notify_hw(sq, &wqe->ctrl, 0);
 }
 
-static inline void mlx5e_xmit_xdp_frame(struct mlx5e_rq *rq,
+static inline bool mlx5e_xmit_xdp_frame(struct mlx5e_rq *rq,
 					struct mlx5e_dma_info *di,
 					const struct xdp_buff *xdp)
 {
@@ -662,7 +663,7 @@ static inline void mlx5e_xmit_xdp_frame(struct mlx5e_rq *rq,
 		     MLX5E_SW2HW_MTU(rq->netdev->mtu) < dma_len)) {
 		rq->stats.xdp_drop++;
 		mlx5e_page_release(rq, di, true);
-		return;
+		return false;
 	}
 
 	if (unlikely(!mlx5e_sq_has_room_for(sq, MLX5E_XDP_TX_WQEBBS))) {
@@ -673,7 +674,7 @@ static inline void mlx5e_xmit_xdp_frame(struct mlx5e_rq *rq,
 		}
 		rq->stats.xdp_tx_full++;
 		mlx5e_page_release(rq, di, true);
-		return;
+		return false;
 	}
 
 	dma_len -= MLX5E_XDP_MIN_INLINE;
@@ -703,6 +704,7 @@ static inline void mlx5e_xmit_xdp_frame(struct mlx5e_rq *rq,
 
 	sq->db.xdp.doorbell = true;
 	rq->stats.xdp_tx++;
+	return true;
 }
 
 /* returns true if packet was consumed by xdp */
@@ -728,11 +730,13 @@ static inline int mlx5e_xdp_handle(struct mlx5e_rq *rq,
 		*len = xdp.data_end - xdp.data;
 		return false;
 	case XDP_TX:
-		mlx5e_xmit_xdp_frame(rq, di, &xdp);
+		if (unlikely(!mlx5e_xmit_xdp_frame(rq, di, &xdp)))
+			trace_xdp_exception(rq->netdev, prog, act);
 		return true;
 	default:
 		bpf_warn_invalid_xdp_action(act);
 	case XDP_ABORTED:
+		trace_xdp_exception(rq->netdev, prog, act);
 	case XDP_DROP:
 		rq->stats.xdp_drop++;
 		mlx5e_page_release(rq, di, true);

drivers/net/ethernet/netronome/nfp/nfp_net_common.c

@@ -42,6 +42,7 @@
  */
 
 #include <linux/bpf.h>
+#include <linux/bpf_trace.h>
 #include <linux/module.h>
 #include <linux/kernel.h>
 #include <linux/init.h>
@@ -1459,7 +1460,7 @@ nfp_net_rx_drop(struct nfp_net_r_vector *r_vec, struct nfp_net_rx_ring *rx_ring,
 		dev_kfree_skb_any(skb);
 }
 
-static void
+static bool
 nfp_net_tx_xdp_buf(struct nfp_net *nn, struct nfp_net_rx_ring *rx_ring,
 		   struct nfp_net_tx_ring *tx_ring,
 		   struct nfp_net_rx_buf *rxbuf, unsigned int pkt_off,
@@ -1473,13 +1474,13 @@ nfp_net_tx_xdp_buf(struct nfp_net *nn, struct nfp_net_rx_ring *rx_ring,
 
 	if (unlikely(nfp_net_tx_full(tx_ring, 1))) {
 		nfp_net_rx_drop(rx_ring->r_vec, rx_ring, rxbuf, NULL);
-		return;
+		return false;
 	}
 
 	new_frag = nfp_net_napi_alloc_one(nn, DMA_BIDIRECTIONAL, &new_dma_addr);
 	if (unlikely(!new_frag)) {
 		nfp_net_rx_drop(rx_ring->r_vec, rx_ring, rxbuf, NULL);
-		return;
+		return false;
 	}
 	nfp_net_rx_give_one(rx_ring, new_frag, new_dma_addr);
 
@@ -1509,6 +1510,7 @@ nfp_net_tx_xdp_buf(struct nfp_net *nn, struct nfp_net_rx_ring *rx_ring,
 
 	tx_ring->wr_p++;
 	tx_ring->wr_ptr_add++;
+	return true;
 }
 
 static int nfp_net_run_xdp(struct bpf_prog *prog, void *data, unsigned int len)
@@ -1613,12 +1615,15 @@ static int nfp_net_rx(struct nfp_net_rx_ring *rx_ring, int budget)
 			case XDP_PASS:
 				break;
 			case XDP_TX:
-				nfp_net_tx_xdp_buf(nn, rx_ring, tx_ring, rxbuf,
-						   pkt_off, pkt_len);
+				if (unlikely(!nfp_net_tx_xdp_buf(nn, rx_ring,
+								 tx_ring, rxbuf,
+								 pkt_off, pkt_len)))
+					trace_xdp_exception(nn->netdev, xdp_prog, act);
 				continue;
 			default:
 				bpf_warn_invalid_xdp_action(act);
 			case XDP_ABORTED:
+				trace_xdp_exception(nn->netdev, xdp_prog, act);
 			case XDP_DROP:
 				nfp_net_rx_give_one(rx_ring, rxbuf->frag,
 						    rxbuf->dma_addr);

drivers/net/ethernet/qlogic/qede/qede_fp.c

@@ -32,6 +32,7 @@
 #include <linux/netdevice.h>
 #include <linux/etherdevice.h>
 #include <linux/skbuff.h>
+#include <linux/bpf_trace.h>
 #include <net/udp_tunnel.h>
 #include <linux/ip.h>
 #include <net/ipv6.h>
@@ -1016,6 +1017,7 @@ static bool qede_rx_xdp(struct qede_dev *edev,
 		/* We need the replacement buffer before transmit. */
 		if (qede_alloc_rx_buffer(rxq, true)) {
 			qede_recycle_rx_bd_ring(rxq, 1);
+			trace_xdp_exception(edev->ndev, prog, act);
 			return false;
 		}
 
@@ -1026,6 +1028,7 @@ static bool qede_rx_xdp(struct qede_dev *edev,
 			dma_unmap_page(rxq->dev, bd->mapping,
 				       PAGE_SIZE, DMA_BIDIRECTIONAL);
 			__free_page(bd->data);
+			trace_xdp_exception(edev->ndev, prog, act);
 		}
 
 		/* Regardless, we've consumed an Rx BD */
@@ -1035,6 +1038,7 @@ static bool qede_rx_xdp(struct qede_dev *edev,
 	default:
 		bpf_warn_invalid_xdp_action(act);
 	case XDP_ABORTED:
+		trace_xdp_exception(edev->ndev, prog, act);
 	case XDP_DROP:
 		qede_recycle_rx_bd_ring(rxq, cqe->bd_num);
 	}

drivers/net/virtio_net.c

@@ -23,6 +23,7 @@
 #include <linux/virtio.h>
 #include <linux/virtio_net.h>
 #include <linux/bpf.h>
+#include <linux/bpf_trace.h>
 #include <linux/scatterlist.h>
 #include <linux/if_vlan.h>
 #include <linux/slab.h>
@@ -330,7 +331,7 @@ static struct sk_buff *page_to_skb(struct virtnet_info *vi,
 	return skb;
 }
 
-static void virtnet_xdp_xmit(struct virtnet_info *vi,
+static bool virtnet_xdp_xmit(struct virtnet_info *vi,
 			     struct receive_queue *rq,
 			     struct send_queue *sq,
 			     struct xdp_buff *xdp,
@@ -382,10 +383,12 @@ static void virtnet_xdp_xmit(struct virtnet_info *vi,
 			put_page(page);
 		} else /* small buffer */
 			kfree_skb(data);
-		return; // On error abort to avoid unnecessary kick
+		/* On error abort to avoid unnecessary kick */
+		return false;
 	}
 
 	virtqueue_kick(sq->vq);
+	return true;
 }
 
 static u32 do_xdp_prog(struct virtnet_info *vi,
@@ -421,11 +424,14 @@ static u32 do_xdp_prog(struct virtnet_info *vi,
 			vi->xdp_queue_pairs +
 			smp_processor_id();
 		xdp.data = buf;
-		virtnet_xdp_xmit(vi, rq, &vi->sq[qp], &xdp, data);
+		if (unlikely(!virtnet_xdp_xmit(vi, rq, &vi->sq[qp], &xdp,
+					       data)))
+			trace_xdp_exception(vi->dev, xdp_prog, act);
 		return XDP_TX;
 	default:
 		bpf_warn_invalid_xdp_action(act);
 	case XDP_ABORTED:
+		trace_xdp_exception(vi->dev, xdp_prog, act);
 	case XDP_DROP:
 		return XDP_DROP;
 	}

include/linux/bpf_trace.h

@@ -0,0 +1,7 @@
+#ifndef __LINUX_BPF_TRACE_H__
+#define __LINUX_BPF_TRACE_H__
+
+#include <trace/events/bpf.h>
+#include <trace/events/xdp.h>
+
+#endif /* __LINUX_BPF_TRACE_H__ */