[PATCH bpf-next v2 0/7] samples/bpf: xdpsock app enhancements

[PATCH bpf-next v2 0/7] samples/bpf: xdpsock app enhancements

[Ong Boon Leong]

First of all, sorry for taking more time to get back to this series and
thanks to all valuble feedback in series-1 at [1] from Jesper and Song
Liu.

Since then I have looked into what Jesper suggested in [2] and worked on
revising the patch series into several patches for ease of review:

v1->v2:
1/7: [No change]. Add VLAN tag (ID & Priority) to the generated Tx-Only
     frames.

2/7: [No change]. Add DMAC and SMAC setting to the generated Tx-Only
     frames. If parameters are not set, previous DMAC and SMAC are used.

3/7: [New]. Add support for selecting different CLOCK for clock_gettime()
     used in get_nsecs.

4/7: [New]. This is a total rework from series-1 3/4-patch [3]. It uses
     clock_nanosleep() suggested by Jesper. In addition, added statistic
     for Tx schedule variance under application stat (-a|--app-stats).
     Make the cyclic Tx operation and --poll mode to be mutually-
     exclusive. Still, the ability to specify TX cycle time and used
     together with batch size and packet count remain the same.

5/7: [New]. Add the support for TX process schedule policy and priority
     setting. By default, SCHED_OTHER policy is used. This too is matching
     the schedule policy setting in [2].

6/7: [Change]. This is update from series-1 4/4-patch [4]. Added TX clean
     process time-out in 1s granularity with configurable retries count
     (-O|--retries).

7/7: [New]. Added timestamp for TX packet following pktgen_hdr format
     matching the implementation in [2]. However, the sequence ID remains
     the same as it is instead of process schedule diff in [2].

To summarize on what program options have been added with v2 series
using an example below:-

 DMAC (-G)                 = fa:8d:f1:e2:0b:e8
 SMAC (-H)                 = ce:17:07:17:3e:3a

 VLAN tagged (-V)
 VLAN ID (-J)              = 12
 VLAN Pri (-K)             = 3

 Tx Queue (-q)             = 3
 Cycle Time in us (-T)     = 1000
 Batch (-b)                = 2
 Packet Count              = 6
 Tx schedule policy (-W)   = FIFO
 Tx schedule priority (-U) = 50
 Clock selection (-w)      = REALTIME

 Tx timeout retries(-O)    = 5
 Tx timestamp (-y)
 Cyclic Tx schedule stat (-a)

Note: xdpsock sets UDP dest-port and src-port to 0x1000 as default.

 Sending Board
 =============
 $ xdpsock -i eth0 -t -N -z -H ce:17:07:17:3e:3a -G fa:8d:f1:e2:0b:e8 \
   -V -J 12 -K 3 -q 3 \
   -T 1000 -b 2 -C 6 -W FIFO -U 50 -w REALTIME \
   -O 5 -y -a

  sock0@eth0:3 txonly xdp-drv
                    pps            pkts           0.00
 rx                 0              0
 tx                 0              6

                    calls/s        count
 rx empty polls     0              0
 fill fail polls    0              0
 copy tx sendtos    0              0
 tx wakeup sendtos  0              5
 opt polls          0              0

                    period     min        ave        max        cycle
 Cyclic TX          1000000    31033      32009      33397      3

 Receiving Board
 ===============
 $ tcpdump -nei eth0 udp port 0x1000 -vv -Q in -X \
    --time-stamp-precision nano
tcpdump: listening on eth0, link-type EN10MB (Ethernet), capture size 262144 bytes
03:46:40.520111580 ce:17:07:17:3e:3a > fa:8d:f1:e2:0b:e8, ethertype 802.1Q (0x8100), length 62: vlan 12, p 3, ethertype IPv4, (tos 0x0, ttl 64, id 0, offset 0, flags [none], proto UDP (17), length 44)
    10.10.10.16.4096 > 10.10.10.32.4096: [udp sum ok] UDP, length 16
        0x0000:  4500 002c 0000 0000 4011 527e 0a0a 0a10  E..,....@.R~....
        0x0010:  0a0a 0a20 1000 1000 0018 e997 be9b e955  ...............U
        0x0020:  0000 0000 61cd 2ba1 0006 987c            ....a.+....|
03:46:40.520112163 ce:17:07:17:3e:3a > fa:8d:f1:e2:0b:e8, ethertype 802.1Q (0x8100), length 62: vlan 12, p 3, ethertype IPv4, (tos 0x0, ttl 64, id 0, offset 0, flags [none], proto UDP (17), length 44)
    10.10.10.16.4096 > 10.10.10.32.4096: [udp sum ok] UDP, length 16
        0x0000:  4500 002c 0000 0000 4011 527e 0a0a 0a10  E..,....@.R~....
        0x0010:  0a0a 0a20 1000 1000 0018 e996 be9b e955  ...............U
        0x0020:  0000 0001 61cd 2ba1 0006 987c            ....a.+....|
03:46:40.521066860 ce:17:07:17:3e:3a > fa:8d:f1:e2:0b:e8, ethertype 802.1Q (0x8100), length 62: vlan 12, p 3, ethertype IPv4, (tos 0x0, ttl 64, id 0, offset 0, flags [none], proto UDP (17), length 44)
    10.10.10.16.4096 > 10.10.10.32.4096: [udp sum ok] UDP, length 16
        0x0000:  4500 002c 0000 0000 4011 527e 0a0a 0a10  E..,....@.R~....
        0x0010:  0a0a 0a20 1000 1000 0018 e5af be9b e955  ...............U
        0x0020:  0000 0002 61cd 2ba1 0006 9c62            ....a.+....b
03:46:40.521067012 ce:17:07:17:3e:3a > fa:8d:f1:e2:0b:e8, ethertype 802.1Q (0x8100), length 62: vlan 12, p 3, ethertype IPv4, (tos 0x0, ttl 64, id 0, offset 0, flags [none], proto UDP (17), length 44)
    10.10.10.16.4096 > 10.10.10.32.4096: [udp sum ok] UDP, length 16
        0x0000:  4500 002c 0000 0000 4011 527e 0a0a 0a10  E..,....@.R~....
        0x0010:  0a0a 0a20 1000 1000 0018 e5ae be9b e955  ...............U
        0x0020:  0000 0003 61cd 2ba1 0006 9c62            ....a.+....b
03:46:40.522061935 ce:17:07:17:3e:3a > fa:8d:f1:e2:0b:e8, ethertype 802.1Q (0x8100), length 62: vlan 12, p 3, ethertype IPv4, (tos 0x0, ttl 64, id 0, offset 0, flags [none], proto UDP (17), length 44)
    10.10.10.16.4096 > 10.10.10.32.4096: [udp sum ok] UDP, length 16
        0x0000:  4500 002c 0000 0000 4011 527e 0a0a 0a10  E..,....@.R~....
        0x0010:  0a0a 0a20 1000 1000 0018 e1c5 be9b e955  ...............U
        0x0020:  0000 0004 61cd 2ba1 0006 a04a            ....a.+....J
03:46:40.522062173 ce:17:07:17:3e:3a > fa:8d:f1:e2:0b:e8, ethertype 802.1Q (0x8100), length 62: vlan 12, p 3, ethertype IPv4, (tos 0x0, ttl 64, id 0, offset 0, flags [none], proto UDP (17), length 44)
    10.10.10.16.4096 > 10.10.10.32.4096: [udp sum ok] UDP, length 16
        0x0000:  4500 002c 0000 0000 4011 527e 0a0a 0a10  E..,....@.R~....
        0x0010:  0a0a 0a20 1000 1000 0018 e1c4 be9b e955  ...............U
        0x0020:  0000 0005 61cd 2ba1 0006 a04a            ....a.+....J

I have tested the above with both tagged and untagged packet format and
based on the timestamp in tcpdump found that the timing of the batch
cyclic transmission is correct.

Appreciate if community can give the patch series v2 a try and point out
any gap.

Thanks
Boon Leong

[1] https://patchwork.kernel.org/project/netdevbpf/cover/20211124091821.3916046-1-boon.leong.ong@intel.com/
[2] https://github.com/netoptimizer/network-testing/blob/master/src/udp_pacer.c
[3] https://patchwork.kernel.org/project/netdevbpf/patch/20211124091821.3916046-4-boon.leong.ong@intel.com/
[4] https://patchwork.kernel.org/project/netdevbpf/patch/20211124091821.3916046-5-boon.leong.ong@intel.com/

Ong Boon Leong (7):
  samples/bpf: xdpsock: add VLAN support for Tx-only operation
  samples/bpf: xdpsock: add Dest and Src MAC setting for Tx-only
    operation
  samples/bpf: xdpsock: add clockid selection support
  samples/bpf: xdpsock: add cyclic TX operation capability
  samples/bpf: xdpsock: add sched policy and priority support
  samples/bpf: xdpsock: add time-out for cleaning Tx
  samples/bpf: xdpsock: add timestamp for Tx-only operation

 samples/bpf/xdpsock_user.c | 363 ++++++++++++++++++++++++++++++++++---
 1 file changed, 341 insertions(+), 22 deletions(-)

-- 
2.25.1

[PATCH bpf-next v2 1/7] samples/bpf: xdpsock: add VLAN support for Tx-only operation

[Ong Boon Leong]

In multi-queue environment testing, the support for VLAN-tag based
steering is useful. So, this patch adds the capability to add
VLAN tag (VLAN ID and Priority) to the generated Tx frame.

To set the VLAN ID=10 and Priority=2 for Tx only through TxQ=3:
 $ xdpsock -i eth0 -t -N -z -q 3 -V -J 10 -K 2

If VLAN ID (-J) and Priority (-K) is set, it default to
  VLAN ID = 1
  VLAN Priority = 0.

For example, VLAN-tagged Tx only, xdp copy mode through TxQ=1:
 $ xdpsock -i eth0 -t -N -c -q 1 -V

Acked-by: Song Liu <songliubraving@fb.com>
Signed-off-by: Ong Boon Leong <boon.leong.ong@intel.com>
---
 samples/bpf/xdpsock_user.c | 90 +++++++++++++++++++++++++++++++-------
 1 file changed, 75 insertions(+), 15 deletions(-)

diff --git a/samples/bpf/xdpsock_user.c b/samples/bpf/xdpsock_user.c
index 616d663d55a..d5e298ccf49 100644
--- a/samples/bpf/xdpsock_user.c
+++ b/samples/bpf/xdpsock_user.c
@@ -56,6 +56,12 @@
 
 #define DEBUG_HEXDUMP 0
 
+#define VLAN_PRIO_MASK		0xe000 /* Priority Code Point */
+#define VLAN_PRIO_SHIFT		13
+#define VLAN_VID_MASK		0x0fff /* VLAN Identifier */
+#define VLAN_VID__DEFAULT	1
+#define VLAN_PRI__DEFAULT	0
+
 typedef __u64 u64;
 typedef __u32 u32;
 typedef __u16 u16;
@@ -81,6 +87,9 @@ static u32 opt_batch_size = 64;
 static int opt_pkt_count;
 static u16 opt_pkt_size = MIN_PKT_SIZE;
 static u32 opt_pkt_fill_pattern = 0x12345678;
+static bool opt_vlan_tag;
+static u16 opt_pkt_vlan_id = VLAN_VID__DEFAULT;
+static u16 opt_pkt_vlan_pri = VLAN_PRI__DEFAULT;
 static bool opt_extra_stats;
 static bool opt_quiet;
 static bool opt_app_stats;
@@ -101,6 +110,14 @@ static u32 prog_id;
 static bool opt_busy_poll;
 static bool opt_reduced_cap;
 
+struct vlan_ethhdr {
+	unsigned char h_dest[6];
+	unsigned char h_source[6];
+	__be16 h_vlan_proto;
+	__be16 h_vlan_TCI;
+	__be16 h_vlan_encapsulated_proto;
+};
+
 struct xsk_ring_stats {
 	unsigned long rx_npkts;
 	unsigned long tx_npkts;
@@ -740,11 +757,13 @@ static inline u16 udp_csum(u32 saddr, u32 daddr, u32 len,
 
 #define ETH_FCS_SIZE 4
 
-#define PKT_HDR_SIZE (sizeof(struct ethhdr) + sizeof(struct iphdr) + \
+#define ETH_HDR_SIZE (opt_vlan_tag ? sizeof(struct vlan_ethhdr) : \
+		      sizeof(struct ethhdr))
+#define PKT_HDR_SIZE (ETH_HDR_SIZE + sizeof(struct iphdr) + \
 		      sizeof(struct udphdr))
 
 #define PKT_SIZE		(opt_pkt_size - ETH_FCS_SIZE)
-#define IP_PKT_SIZE		(PKT_SIZE - sizeof(struct ethhdr))
+#define IP_PKT_SIZE		(PKT_SIZE - ETH_HDR_SIZE)
 #define UDP_PKT_SIZE		(IP_PKT_SIZE - sizeof(struct iphdr))
 #define UDP_PKT_DATA_SIZE	(UDP_PKT_SIZE - sizeof(struct udphdr))
 
@@ -752,17 +771,42 @@ static u8 pkt_data[XSK_UMEM__DEFAULT_FRAME_SIZE];
 
 static void gen_eth_hdr_data(void)
 {
-	struct udphdr *udp_hdr = (struct udphdr *)(pkt_data +
-						   sizeof(struct ethhdr) +
-						   sizeof(struct iphdr));
-	struct iphdr *ip_hdr = (struct iphdr *)(pkt_data +
-						sizeof(struct ethhdr));
-	struct ethhdr *eth_hdr = (struct ethhdr *)pkt_data;
-
-	/* ethernet header */
-	memcpy(eth_hdr->h_dest, "\x3c\xfd\xfe\x9e\x7f\x71", ETH_ALEN);
-	memcpy(eth_hdr->h_source, "\xec\xb1\xd7\x98\x3a\xc0", ETH_ALEN);
-	eth_hdr->h_proto = htons(ETH_P_IP);
+	struct udphdr *udp_hdr;
+	struct iphdr *ip_hdr;
+
+	if (opt_vlan_tag) {
+		struct vlan_ethhdr *veth_hdr = (struct vlan_ethhdr *)pkt_data;
+		u16 vlan_tci = 0;
+
+		udp_hdr = (struct udphdr *)(pkt_data +
+					    sizeof(struct vlan_ethhdr) +
+					    sizeof(struct iphdr));
+		ip_hdr = (struct iphdr *)(pkt_data +
+					  sizeof(struct vlan_ethhdr));
+
+		/* ethernet & VLAN header */
+		memcpy(veth_hdr->h_dest, "\x3c\xfd\xfe\x9e\x7f\x71", ETH_ALEN);
+		memcpy(veth_hdr->h_source, "\xec\xb1\xd7\x98\x3a\xc0", ETH_ALEN);
+		veth_hdr->h_vlan_proto = htons(ETH_P_8021Q);
+		vlan_tci = opt_pkt_vlan_id & VLAN_VID_MASK;
+		vlan_tci |= (opt_pkt_vlan_pri << VLAN_PRIO_SHIFT) & VLAN_PRIO_MASK;
+		veth_hdr->h_vlan_TCI = htons(vlan_tci);
+		veth_hdr->h_vlan_encapsulated_proto = htons(ETH_P_IP);
+	} else {
+		struct ethhdr *eth_hdr = (struct ethhdr *)pkt_data;
+
+		udp_hdr = (struct udphdr *)(pkt_data +
+					    sizeof(struct ethhdr) +
+					    sizeof(struct iphdr));
+		ip_hdr = (struct iphdr *)(pkt_data +
+					  sizeof(struct ethhdr));
+
+		/* ethernet header */
+		memcpy(eth_hdr->h_dest, "\x3c\xfd\xfe\x9e\x7f\x71", ETH_ALEN);
+		memcpy(eth_hdr->h_source, "\xec\xb1\xd7\x98\x3a\xc0", ETH_ALEN);
+		eth_hdr->h_proto = htons(ETH_P_IP);
+	}
+
 
 	/* IP header */
 	ip_hdr->version = IPVERSION;
@@ -920,6 +964,9 @@ static struct option long_options[] = {
 	{"tx-pkt-count", required_argument, 0, 'C'},
 	{"tx-pkt-size", required_argument, 0, 's'},
 	{"tx-pkt-pattern", required_argument, 0, 'P'},
+	{"tx-vlan", no_argument, 0, 'V'},
+	{"tx-vlan-id", required_argument, 0, 'J'},
+	{"tx-vlan-pri", required_argument, 0, 'K'},
 	{"extra-stats", no_argument, 0, 'x'},
 	{"quiet", no_argument, 0, 'Q'},
 	{"app-stats", no_argument, 0, 'a'},
@@ -960,6 +1007,9 @@ static void usage(const char *prog)
 		"			(Default: %d bytes)\n"
 		"			Min size: %d, Max size %d.\n"
 		"  -P, --tx-pkt-pattern=nPacket fill pattern. Default: 0x%x\n"
+		"  -V, --tx-vlan        Send VLAN tagged  packets (For -t|--txonly)\n"
+		"  -J, --tx-vlan-id=n   Tx VLAN ID [1-4095]. Default: %d (For -V|--tx-vlan)\n"
+		"  -K, --tx-vlan-pri=n  Tx VLAN Priority [0-7]. Default: %d (For -V|--tx-vlan)\n"
 		"  -x, --extra-stats	Display extra statistics.\n"
 		"  -Q, --quiet          Do not display any stats.\n"
 		"  -a, --app-stats	Display application (syscall) statistics.\n"
@@ -969,7 +1019,8 @@ static void usage(const char *prog)
 		"\n";
 	fprintf(stderr, str, prog, XSK_UMEM__DEFAULT_FRAME_SIZE,
 		opt_batch_size, MIN_PKT_SIZE, MIN_PKT_SIZE,
-		XSK_UMEM__DEFAULT_FRAME_SIZE, opt_pkt_fill_pattern);
+		XSK_UMEM__DEFAULT_FRAME_SIZE, opt_pkt_fill_pattern,
+		VLAN_VID__DEFAULT, VLAN_PRI__DEFAULT);
 
 	exit(EXIT_FAILURE);
 }
@@ -981,7 +1032,7 @@ static void parse_command_line(int argc, char **argv)
 	opterr = 0;
 
 	for (;;) {
-		c = getopt_long(argc, argv, "Frtli:q:pSNn:czf:muMd:b:C:s:P:xQaI:BR",
+		c = getopt_long(argc, argv, "Frtli:q:pSNn:czf:muMd:b:C:s:P:VJ:K:xQaI:BR",
 				long_options, &option_index);
 		if (c == -1)
 			break;
@@ -1062,6 +1113,15 @@ static void parse_command_line(int argc, char **argv)
 		case 'P':
 			opt_pkt_fill_pattern = strtol(optarg, NULL, 16);
 			break;
+		case 'V':
+			opt_vlan_tag = true;
+			break;
+		case 'J':
+			opt_pkt_vlan_id = atoi(optarg);
+			break;
+		case 'K':
+			opt_pkt_vlan_pri = atoi(optarg);
+			break;
 		case 'x':
 			opt_extra_stats = 1;
 			break;
-- 
2.25.1

[PATCH bpf-next v2 2/7] samples/bpf: xdpsock: add Dest and Src MAC setting for Tx-only operation

[Ong Boon Leong]

To set Dest MAC address (-G|--tx-dmac) only:
 $ xdpsock -i eth0 -t -N -z -G aa:bb:cc:dd:ee:ff

To set Source MAC address (-H|--tx-smac) only:
 $ xdpsock -i eth0 -t -N -z -H 11:22:33:44:55:66

To set both Dest and Source MAC address:
 $ xdpsock -i eth0 -t -N -z -G aa:bb:cc:dd:ee:ff \
   -H 11:22:33:44:55:66

The default Dest and Source MAC address remain the same as before.

Acked-by: Song Liu <songliubraving@fb.com>
Acked-by: Jesper Dangaard Brouer <brouer@redhat.com>
Signed-off-by: Ong Boon Leong <boon.leong.ong@intel.com>
---
 samples/bpf/xdpsock_user.c | 35 ++++++++++++++++++++++++++++++-----
 1 file changed, 30 insertions(+), 5 deletions(-)

diff --git a/samples/bpf/xdpsock_user.c b/samples/bpf/xdpsock_user.c
index d5e298ccf49..c9a8748a460 100644
--- a/samples/bpf/xdpsock_user.c
+++ b/samples/bpf/xdpsock_user.c
@@ -14,6 +14,7 @@
 #include <arpa/inet.h>
 #include <locale.h>
 #include <net/ethernet.h>
+#include <netinet/ether.h>
 #include <net/if.h>
 #include <poll.h>
 #include <pthread.h>
@@ -90,6 +91,10 @@ static u32 opt_pkt_fill_pattern = 0x12345678;
 static bool opt_vlan_tag;
 static u16 opt_pkt_vlan_id = VLAN_VID__DEFAULT;
 static u16 opt_pkt_vlan_pri = VLAN_PRI__DEFAULT;
+static struct ether_addr opt_txdmac = {{ 0x3c, 0xfd, 0xfe,
+					 0x9e, 0x7f, 0x71 }};
+static struct ether_addr opt_txsmac = {{ 0xec, 0xb1, 0xd7,
+					 0x98, 0x3a, 0xc0 }};
 static bool opt_extra_stats;
 static bool opt_quiet;
 static bool opt_app_stats;
@@ -785,8 +790,8 @@ static void gen_eth_hdr_data(void)
 					  sizeof(struct vlan_ethhdr));
 
 		/* ethernet & VLAN header */
-		memcpy(veth_hdr->h_dest, "\x3c\xfd\xfe\x9e\x7f\x71", ETH_ALEN);
-		memcpy(veth_hdr->h_source, "\xec\xb1\xd7\x98\x3a\xc0", ETH_ALEN);
+		memcpy(veth_hdr->h_dest, &opt_txdmac, ETH_ALEN);
+		memcpy(veth_hdr->h_source, &opt_txsmac, ETH_ALEN);
 		veth_hdr->h_vlan_proto = htons(ETH_P_8021Q);
 		vlan_tci = opt_pkt_vlan_id & VLAN_VID_MASK;
 		vlan_tci |= (opt_pkt_vlan_pri << VLAN_PRIO_SHIFT) & VLAN_PRIO_MASK;
@@ -802,8 +807,8 @@ static void gen_eth_hdr_data(void)
 					  sizeof(struct ethhdr));
 
 		/* ethernet header */
-		memcpy(eth_hdr->h_dest, "\x3c\xfd\xfe\x9e\x7f\x71", ETH_ALEN);
-		memcpy(eth_hdr->h_source, "\xec\xb1\xd7\x98\x3a\xc0", ETH_ALEN);
+		memcpy(eth_hdr->h_dest, &opt_txdmac, ETH_ALEN);
+		memcpy(eth_hdr->h_source, &opt_txsmac, ETH_ALEN);
 		eth_hdr->h_proto = htons(ETH_P_IP);
 	}
 
@@ -967,6 +972,8 @@ static struct option long_options[] = {
 	{"tx-vlan", no_argument, 0, 'V'},
 	{"tx-vlan-id", required_argument, 0, 'J'},
 	{"tx-vlan-pri", required_argument, 0, 'K'},
+	{"tx-dmac", required_argument, 0, 'G'},
+	{"tx-smac", required_argument, 0, 'H'},
 	{"extra-stats", no_argument, 0, 'x'},
 	{"quiet", no_argument, 0, 'Q'},
 	{"app-stats", no_argument, 0, 'a'},
@@ -1010,6 +1017,8 @@ static void usage(const char *prog)
 		"  -V, --tx-vlan        Send VLAN tagged  packets (For -t|--txonly)\n"
 		"  -J, --tx-vlan-id=n   Tx VLAN ID [1-4095]. Default: %d (For -V|--tx-vlan)\n"
 		"  -K, --tx-vlan-pri=n  Tx VLAN Priority [0-7]. Default: %d (For -V|--tx-vlan)\n"
+		"  -G, --tx-dmac=<MAC>  Dest MAC addr of TX frame in aa:bb:cc:dd:ee:ff format (For -V|--tx-vlan)\n"
+		"  -H, --tx-smac=<MAC>  Src MAC addr of TX frame in aa:bb:cc:dd:ee:ff format (For -V|--tx-vlan)\n"
 		"  -x, --extra-stats	Display extra statistics.\n"
 		"  -Q, --quiet          Do not display any stats.\n"
 		"  -a, --app-stats	Display application (syscall) statistics.\n"
@@ -1032,7 +1041,7 @@ static void parse_command_line(int argc, char **argv)
 	opterr = 0;
 
 	for (;;) {
-		c = getopt_long(argc, argv, "Frtli:q:pSNn:czf:muMd:b:C:s:P:VJ:K:xQaI:BR",
+		c = getopt_long(argc, argv, "Frtli:q:pSNn:czf:muMd:b:C:s:P:VJ:K:G:H:xQaI:BR",
 				long_options, &option_index);
 		if (c == -1)
 			break;
@@ -1122,6 +1131,22 @@ static void parse_command_line(int argc, char **argv)
 		case 'K':
 			opt_pkt_vlan_pri = atoi(optarg);
 			break;
+		case 'G':
+			if (!ether_aton_r(optarg,
+					  (struct ether_addr *)&opt_txdmac)) {
+				fprintf(stderr, "Invalid dmac address:%s\n",
+					optarg);
+				usage(basename(argv[0]));
+			}
+			break;
+		case 'H':
+			if (!ether_aton_r(optarg,
+					  (struct ether_addr *)&opt_txsmac)) {
+				fprintf(stderr, "Invalid smac address:%s\n",
+					optarg);
+				usage(basename(argv[0]));
+			}
+			break;
 		case 'x':
 			opt_extra_stats = 1;
 			break;
-- 
2.25.1

[PATCH bpf-next v2 3/7] samples/bpf: xdpsock: add clockid selection support

[Ong Boon Leong]

User specifies the clock selection by using -w CLOCK or --clock=CLOCK
where CLOCK=[REALTIME, TAI, BOOTTIME, MONOTONIC].

The default CLOCK selection is MONOTONIC.

The implementation of clock selection parsing is borrowed from
iproute2/tc/q_taprio.c

Signed-off-by: Ong Boon Leong <boon.leong.ong@intel.com>
---
 samples/bpf/xdpsock_user.c | 40 ++++++++++++++++++++++++++++++++++++--
 1 file changed, 38 insertions(+), 2 deletions(-)

diff --git a/samples/bpf/xdpsock_user.c b/samples/bpf/xdpsock_user.c
index c9a8748a460..e6e9a20375c 100644
--- a/samples/bpf/xdpsock_user.c
+++ b/samples/bpf/xdpsock_user.c
@@ -114,6 +114,7 @@ static u32 opt_num_xsks = 1;
 static u32 prog_id;
 static bool opt_busy_poll;
 static bool opt_reduced_cap;
+static clockid_t opt_clock = CLOCK_MONOTONIC;
 
 struct vlan_ethhdr {
 	unsigned char h_dest[6];
@@ -178,15 +179,40 @@ struct xsk_socket_info {
 	u32 outstanding_tx;
 };
 
+static const struct clockid_map {
+	const char *name;
+	clockid_t clockid;
+} clockids_map[] = {
+	{ "REALTIME", CLOCK_REALTIME },
+	{ "TAI", CLOCK_TAI },
+	{ "BOOTTIME", CLOCK_BOOTTIME },
+	{ "MONOTONIC", CLOCK_MONOTONIC },
+	{ NULL }
+};
+
 static int num_socks;
 struct xsk_socket_info *xsks[MAX_SOCKS];
 int sock;
 
+static int get_clockid(clockid_t *id, const char *name)
+{
+	const struct clockid_map *clk;
+
+	for (clk = clockids_map; clk->name; clk++) {
+		if (strcasecmp(clk->name, name) == 0) {
+			*id = clk->clockid;
+			return 0;
+		}
+	}
+
+	return -1;
+}
+
 static unsigned long get_nsecs(void)
 {
 	struct timespec ts;
 
-	clock_gettime(CLOCK_MONOTONIC, &ts);
+	clock_gettime(opt_clock, &ts);
 	return ts.tv_sec * 1000000000UL + ts.tv_nsec;
 }
 
@@ -965,6 +991,7 @@ static struct option long_options[] = {
 	{"shared-umem", no_argument, 0, 'M'},
 	{"force", no_argument, 0, 'F'},
 	{"duration", required_argument, 0, 'd'},
+	{"clock", required_argument, 0, 'w'},
 	{"batch-size", required_argument, 0, 'b'},
 	{"tx-pkt-count", required_argument, 0, 'C'},
 	{"tx-pkt-size", required_argument, 0, 's'},
@@ -1006,6 +1033,7 @@ static void usage(const char *prog)
 		"  -F, --force		Force loading the XDP prog\n"
 		"  -d, --duration=n	Duration in secs to run command.\n"
 		"			Default: forever.\n"
+		"  -w, --clock=CLOCK	Clock NAME (default MONOTONIC).\n"
 		"  -b, --batch-size=n	Batch size for sending or receiving\n"
 		"			packets. Default: %d\n"
 		"  -C, --tx-pkt-count=n	Number of packets to send.\n"
@@ -1041,7 +1069,7 @@ static void parse_command_line(int argc, char **argv)
 	opterr = 0;
 
 	for (;;) {
-		c = getopt_long(argc, argv, "Frtli:q:pSNn:czf:muMd:b:C:s:P:VJ:K:G:H:xQaI:BR",
+		c = getopt_long(argc, argv, "Frtli:q:pSNn:w:czf:muMd:b:C:s:P:VJ:K:G:H:xQaI:BR",
 				long_options, &option_index);
 		if (c == -1)
 			break;
@@ -1075,6 +1103,14 @@ static void parse_command_line(int argc, char **argv)
 		case 'n':
 			opt_interval = atoi(optarg);
 			break;
+		case 'w':
+			if (get_clockid(&opt_clock, optarg)) {
+				fprintf(stderr,
+					"ERROR: Invalid clock %s. Default to CLOCK_MONOTONIC.\n",
+					optarg);
+				opt_clock = CLOCK_MONOTONIC;
+			}
+			break;
 		case 'z':
 			opt_xdp_bind_flags |= XDP_ZEROCOPY;
 			break;
-- 
2.25.1

[PATCH bpf-next v2 4/7] samples/bpf: xdpsock: add cyclic TX operation capability

[Ong Boon Leong]

Tx cycle time is in micro-seconds unit. By combining the batch size (-b M)
and Tx cycle time (-T|--tx-cycle N), xdpsock now can transmit batch-size of
packets every N-us periodically. Cyclic TX operation is not applicable if
--poll mode is used.

To transmit 16 packets every 1ms cycle time for total of 100000 packets
silently:
 $ xdpsock -i eth0 -T -N -z -T 1000 -b 16 -C 100000

To print cyclic TX schedule variance stats, use --app-stats|-a:
 $ xdpsock -i eth0 -T -N -z -T 1000 -b 16 -C 100000 -a

 sock0@eth0:0 txonly xdp-drv
                   pps            pkts           0.00
rx                 0              0
tx                 0              100000

                   calls/s        count
rx empty polls     0              0
fill fail polls    0              0
copy tx sendtos    0              0
tx wakeup sendtos  0              6254
opt polls          0              0

                   period     min        ave        max        cycle
Cyclic TX          1000000    53507      75334      712642     6250

Signed-off-by: Ong Boon Leong <boon.leong.ong@intel.com>
---
 samples/bpf/xdpsock_user.c | 85 +++++++++++++++++++++++++++++++++++---
 1 file changed, 80 insertions(+), 5 deletions(-)

diff --git a/samples/bpf/xdpsock_user.c b/samples/bpf/xdpsock_user.c
index e6e9a20375c..a2a42ec4b0e 100644
--- a/samples/bpf/xdpsock_user.c
+++ b/samples/bpf/xdpsock_user.c
@@ -63,12 +63,19 @@
 #define VLAN_VID__DEFAULT	1
 #define VLAN_PRI__DEFAULT	0
 
+#define NSEC_PER_SEC		1000000000UL
+#define NSEC_PER_USEC		1000
+
 typedef __u64 u64;
 typedef __u32 u32;
 typedef __u16 u16;
 typedef __u8  u8;
 
 static unsigned long prev_time;
+static long tx_cycle_diff_min;
+static long tx_cycle_diff_max;
+static double tx_cycle_diff_ave;
+static long tx_cycle_cnt;
 
 enum benchmark_type {
 	BENCH_RXDROP = 0,
@@ -115,6 +122,7 @@ static u32 prog_id;
 static bool opt_busy_poll;
 static bool opt_reduced_cap;
 static clockid_t opt_clock = CLOCK_MONOTONIC;
+static unsigned long opt_tx_cycle_ns;
 
 struct vlan_ethhdr {
 	unsigned char h_dest[6];
@@ -305,6 +313,15 @@ static void dump_app_stats(long dt)
 		xsks[i]->app_stats.prev_tx_wakeup_sendtos = xsks[i]->app_stats.tx_wakeup_sendtos;
 		xsks[i]->app_stats.prev_opt_polls = xsks[i]->app_stats.opt_polls;
 	}
+
+	if (opt_tx_cycle_ns) {
+		printf("\n%-18s %-10s %-10s %-10s %-10s %-10s\n",
+		       "", "period", "min", "ave", "max", "cycle");
+		printf("%-18s %-10lu %-10lu %-10lu %-10lu %-10lu\n",
+		       "Cyclic TX", opt_tx_cycle_ns, tx_cycle_diff_min,
+		       (long)(tx_cycle_diff_ave / tx_cycle_cnt),
+		       tx_cycle_diff_max, tx_cycle_cnt);
+	}
 }
 
 static bool get_interrupt_number(void)
@@ -1001,6 +1018,7 @@ static struct option long_options[] = {
 	{"tx-vlan-pri", required_argument, 0, 'K'},
 	{"tx-dmac", required_argument, 0, 'G'},
 	{"tx-smac", required_argument, 0, 'H'},
+	{"tx-cycle", required_argument, 0, 'T'},
 	{"extra-stats", no_argument, 0, 'x'},
 	{"quiet", no_argument, 0, 'Q'},
 	{"app-stats", no_argument, 0, 'a'},
@@ -1047,6 +1065,7 @@ static void usage(const char *prog)
 		"  -K, --tx-vlan-pri=n  Tx VLAN Priority [0-7]. Default: %d (For -V|--tx-vlan)\n"
 		"  -G, --tx-dmac=<MAC>  Dest MAC addr of TX frame in aa:bb:cc:dd:ee:ff format (For -V|--tx-vlan)\n"
 		"  -H, --tx-smac=<MAC>  Src MAC addr of TX frame in aa:bb:cc:dd:ee:ff format (For -V|--tx-vlan)\n"
+		"  -T, --tx-cycle=n     Tx cycle time in micro-seconds (For -t|--txonly).\n"
 		"  -x, --extra-stats	Display extra statistics.\n"
 		"  -Q, --quiet          Do not display any stats.\n"
 		"  -a, --app-stats	Display application (syscall) statistics.\n"
@@ -1069,7 +1088,7 @@ static void parse_command_line(int argc, char **argv)
 	opterr = 0;
 
 	for (;;) {
-		c = getopt_long(argc, argv, "Frtli:q:pSNn:w:czf:muMd:b:C:s:P:VJ:K:G:H:xQaI:BR",
+		c = getopt_long(argc, argv, "Frtli:q:pSNn:w:czf:muMd:b:C:s:P:VJ:K:G:H:T:xQaI:BR",
 				long_options, &option_index);
 		if (c == -1)
 			break;
@@ -1183,6 +1202,10 @@ static void parse_command_line(int argc, char **argv)
 				usage(basename(argv[0]));
 			}
 			break;
+		case 'T':
+			opt_tx_cycle_ns = atoi(optarg);
+			opt_tx_cycle_ns *= NSEC_PER_USEC;
+			break;
 		case 'x':
 			opt_extra_stats = 1;
 			break;
@@ -1388,7 +1411,7 @@ static void rx_drop_all(void)
 	}
 }
 
-static void tx_only(struct xsk_socket_info *xsk, u32 *frame_nb, int batch_size)
+static int tx_only(struct xsk_socket_info *xsk, u32 *frame_nb, int batch_size)
 {
 	u32 idx;
 	unsigned int i;
@@ -1397,7 +1420,7 @@ static void tx_only(struct xsk_socket_info *xsk, u32 *frame_nb, int batch_size)
 				      batch_size) {
 		complete_tx_only(xsk, batch_size);
 		if (benchmark_done)
-			return;
+			return 0;
 	}
 
 	for (i = 0; i < batch_size; i++) {
@@ -1413,6 +1436,8 @@ static void tx_only(struct xsk_socket_info *xsk, u32 *frame_nb, int batch_size)
 	*frame_nb += batch_size;
 	*frame_nb %= NUM_FRAMES;
 	complete_tx_only(xsk, batch_size);
+
+	return batch_size;
 }
 
 static inline int get_batch_size(int pkt_cnt)
@@ -1446,16 +1471,39 @@ static void tx_only_all(void)
 {
 	struct pollfd fds[MAX_SOCKS] = {};
 	u32 frame_nb[MAX_SOCKS] = {};
+	unsigned long next_tx_ns = 0;
 	int pkt_cnt = 0;
 	int i, ret;
 
+	if (opt_poll && opt_tx_cycle_ns) {
+		fprintf(stderr,
+			"Error: --poll and --tx-cycles are both set\n");
+		return;
+	}
+
 	for (i = 0; i < num_socks; i++) {
 		fds[0].fd = xsk_socket__fd(xsks[i]->xsk);
 		fds[0].events = POLLOUT;
 	}
 
+	if (opt_tx_cycle_ns) {
+		/* Align Tx time to micro-second boundary */
+		next_tx_ns = (get_nsecs() / NSEC_PER_USEC + 1) *
+			     NSEC_PER_USEC;
+		next_tx_ns += opt_tx_cycle_ns;
+
+		/* Initialize periodic Tx scheduling variance */
+		tx_cycle_diff_min = 1000000000;
+		tx_cycle_diff_max = 0;
+		tx_cycle_diff_ave = 0.0;
+	}
+
 	while ((opt_pkt_count && pkt_cnt < opt_pkt_count) || !opt_pkt_count) {
 		int batch_size = get_batch_size(pkt_cnt);
+		struct timespec next;
+		int tx_cnt = 0;
+		long diff;
+		int err;
 
 		if (opt_poll) {
 			for (i = 0; i < num_socks; i++)
@@ -1468,13 +1516,40 @@ static void tx_only_all(void)
 				continue;
 		}
 
+		if (opt_tx_cycle_ns) {
+			next.tv_sec = next_tx_ns / NSEC_PER_SEC;
+			next.tv_nsec = next_tx_ns % NSEC_PER_SEC;
+			err = clock_nanosleep(opt_clock, TIMER_ABSTIME, &next, NULL);
+			if (err) {
+				if (err != EINTR)
+					fprintf(stderr,
+						"clock_nanosleep failed. Err:%d errno:%d\n",
+						err, errno);
+				break;
+			}
+
+			/* Measure periodic Tx scheduling variance */
+			diff = get_nsecs() - next_tx_ns;
+			if (diff < tx_cycle_diff_min)
+				tx_cycle_diff_min = diff;
+
+			if (diff > tx_cycle_diff_max)
+				tx_cycle_diff_max = diff;
+
+			tx_cycle_diff_ave += (double)diff;
+			tx_cycle_cnt++;
+		}
+
 		for (i = 0; i < num_socks; i++)
-			tx_only(xsks[i], &frame_nb[i], batch_size);
+			tx_cnt += tx_only(xsks[i], &frame_nb[i], batch_size);
 
-		pkt_cnt += batch_size;
+		pkt_cnt += tx_cnt;
 
 		if (benchmark_done)
 			break;
+
+		if (opt_tx_cycle_ns)
+			next_tx_ns += opt_tx_cycle_ns;
 	}
 
 	if (opt_pkt_count)
-- 
2.25.1

[PATCH bpf-next v2 5/7] samples/bpf: xdpsock: add sched policy and priority support

[Ong Boon Leong]

By default, TX schedule policy is SCHED_OTHER (round-robin time-sharing).
To improve TX cyclic scheduling, we add SCHED_FIFO policy and its priority
by using -W FIFO or --policy=FIFO and -U <PRIO> or --schpri=<PRIO>.

A) From xdpsock --app-stats, for SCHED_OTHER policy:
   $ xdpsock -i eth0 -t -N -z -T 1000 -b 16 -C 100000 -a

                      period     min        ave        max        cycle
   Cyclic TX          1000000    53507      75334      712642     6250

B) For SCHED_FIFO policy and schpri=50:
   $ xdpsock -i eth0 -t -N -z -T 1000 -b 16 -C 100000 -a -W FIFO -U 50

                      period     min        ave        max        cycle
   Cyclic TX          1000000    3699       24859      54397      6250

Signed-off-by: Ong Boon Leong <boon.leong.ong@intel.com>
---
 samples/bpf/xdpsock_user.c | 61 ++++++++++++++++++++++++++++++++++++--
 1 file changed, 59 insertions(+), 2 deletions(-)

diff --git a/samples/bpf/xdpsock_user.c b/samples/bpf/xdpsock_user.c
index a2a42ec4b0e..b7d0f536f97 100644
--- a/samples/bpf/xdpsock_user.c
+++ b/samples/bpf/xdpsock_user.c
@@ -31,6 +31,7 @@
 #include <sys/un.h>
 #include <time.h>
 #include <unistd.h>
+#include <sched.h>
 
 #include <bpf/libbpf.h>
 #include <bpf/xsk.h>
@@ -66,6 +67,8 @@
 #define NSEC_PER_SEC		1000000000UL
 #define NSEC_PER_USEC		1000
 
+#define SCHED_PRI__DEFAULT	0
+
 typedef __u64 u64;
 typedef __u32 u32;
 typedef __u16 u16;
@@ -123,6 +126,8 @@ static bool opt_busy_poll;
 static bool opt_reduced_cap;
 static clockid_t opt_clock = CLOCK_MONOTONIC;
 static unsigned long opt_tx_cycle_ns;
+static int opt_schpolicy = SCHED_OTHER;
+static int opt_schprio = SCHED_PRI__DEFAULT;
 
 struct vlan_ethhdr {
 	unsigned char h_dest[6];
@@ -198,6 +203,15 @@ static const struct clockid_map {
 	{ NULL }
 };
 
+static const struct sched_map {
+	const char *name;
+	int policy;
+} schmap[] = {
+	{ "OTHER", SCHED_OTHER },
+	{ "FIFO", SCHED_FIFO },
+	{ NULL }
+};
+
 static int num_socks;
 struct xsk_socket_info *xsks[MAX_SOCKS];
 int sock;
@@ -216,6 +230,20 @@ static int get_clockid(clockid_t *id, const char *name)
 	return -1;
 }
 
+static int get_schpolicy(int *policy, const char *name)
+{
+	const struct sched_map *sch;
+
+	for (sch = schmap; sch->name; sch++) {
+		if (strcasecmp(sch->name, name) == 0) {
+			*policy = sch->policy;
+			return 0;
+		}
+	}
+
+	return -1;
+}
+
 static unsigned long get_nsecs(void)
 {
 	struct timespec ts;
@@ -1019,6 +1047,8 @@ static struct option long_options[] = {
 	{"tx-dmac", required_argument, 0, 'G'},
 	{"tx-smac", required_argument, 0, 'H'},
 	{"tx-cycle", required_argument, 0, 'T'},
+	{"policy", required_argument, 0, 'W'},
+	{"schpri", required_argument, 0, 'U'},
 	{"extra-stats", no_argument, 0, 'x'},
 	{"quiet", no_argument, 0, 'Q'},
 	{"app-stats", no_argument, 0, 'a'},
@@ -1066,6 +1096,8 @@ static void usage(const char *prog)
 		"  -G, --tx-dmac=<MAC>  Dest MAC addr of TX frame in aa:bb:cc:dd:ee:ff format (For -V|--tx-vlan)\n"
 		"  -H, --tx-smac=<MAC>  Src MAC addr of TX frame in aa:bb:cc:dd:ee:ff format (For -V|--tx-vlan)\n"
 		"  -T, --tx-cycle=n     Tx cycle time in micro-seconds (For -t|--txonly).\n"
+		"  -W, --policy=POLICY  Schedule policy. Default: SCHED_OTHER\n"
+		"  -U, --schpri=n       Schedule priority. Default: %d\n"
 		"  -x, --extra-stats	Display extra statistics.\n"
 		"  -Q, --quiet          Do not display any stats.\n"
 		"  -a, --app-stats	Display application (syscall) statistics.\n"
@@ -1076,7 +1108,8 @@ static void usage(const char *prog)
 	fprintf(stderr, str, prog, XSK_UMEM__DEFAULT_FRAME_SIZE,
 		opt_batch_size, MIN_PKT_SIZE, MIN_PKT_SIZE,
 		XSK_UMEM__DEFAULT_FRAME_SIZE, opt_pkt_fill_pattern,
-		VLAN_VID__DEFAULT, VLAN_PRI__DEFAULT);
+		VLAN_VID__DEFAULT, VLAN_PRI__DEFAULT,
+		SCHED_PRI__DEFAULT);
 
 	exit(EXIT_FAILURE);
 }
@@ -1088,7 +1121,8 @@ static void parse_command_line(int argc, char **argv)
 	opterr = 0;
 
 	for (;;) {
-		c = getopt_long(argc, argv, "Frtli:q:pSNn:w:czf:muMd:b:C:s:P:VJ:K:G:H:T:xQaI:BR",
+		c = getopt_long(argc, argv,
+				"Frtli:q:pSNn:w:czf:muMd:b:C:s:P:VJ:K:G:H:T:W:U:xQaI:BR",
 				long_options, &option_index);
 		if (c == -1)
 			break;
@@ -1206,6 +1240,17 @@ static void parse_command_line(int argc, char **argv)
 			opt_tx_cycle_ns = atoi(optarg);
 			opt_tx_cycle_ns *= NSEC_PER_USEC;
 			break;
+		case 'W':
+			if (get_schpolicy(&opt_schpolicy, optarg)) {
+				fprintf(stderr,
+					"ERROR: Invalid policy %s. Default to SCHED_OTHER.\n",
+					optarg);
+				opt_schpolicy = SCHED_OTHER;
+			}
+			break;
+		case 'U':
+			opt_schprio = atoi(optarg);
+			break;
 		case 'x':
 			opt_extra_stats = 1;
 			break;
@@ -1780,6 +1825,7 @@ int main(int argc, char **argv)
 	struct __user_cap_data_struct data[2] = { { 0 } };
 	struct rlimit r = {RLIM_INFINITY, RLIM_INFINITY};
 	bool rx = false, tx = false;
+	struct sched_param schparam;
 	struct xsk_umem_info *umem;
 	struct bpf_object *obj;
 	int xsks_map_fd = 0;
@@ -1881,6 +1927,16 @@ int main(int argc, char **argv)
 	prev_time = get_nsecs();
 	start_time = prev_time;
 
+	/* Configure sched priority for better wake-up accuracy */
+	memset(&schparam, 0, sizeof(schparam));
+	schparam.sched_priority = opt_schprio;
+	ret = sched_setscheduler(0, opt_schpolicy, &schparam);
+	if (ret) {
+		fprintf(stderr, "Error(%d) in setting priority(%d): %s\n",
+			errno, opt_schprio, strerror(errno));
+		goto out;
+	}
+
 	if (opt_bench == BENCH_RXDROP)
 		rx_drop_all();
 	else if (opt_bench == BENCH_TXONLY)
@@ -1888,6 +1944,7 @@ int main(int argc, char **argv)
 	else
 		l2fwd_all();
 
+out:
 	benchmark_done = true;
 
 	if (!opt_quiet)
-- 
2.25.1

[PATCH bpf-next v2 6/7] samples/bpf: xdpsock: add time-out for cleaning Tx

[Ong Boon Leong]

When user sets tx-pkt-count and in case where there are invalid Tx frame,
the complete_tx_only_all() process polls indefinitely. So, this patch
adds a time-out mechanism into the process so that the application
can terminate automatically after it retries 3*polling interval duration.

v1->v2:
 Thanks to Jesper's and Song Liu's suggestion.
 - clean-up git message to remove polling log
 - make the Tx time-out retries configurable with 1s granularity

Signed-off-by: Ong Boon Leong <boon.leong.ong@intel.com>
---
 samples/bpf/xdpsock_user.c | 11 +++++++++--
 1 file changed, 9 insertions(+), 2 deletions(-)

diff --git a/samples/bpf/xdpsock_user.c b/samples/bpf/xdpsock_user.c
index b7d0f536f97..319cb3cdb22 100644
--- a/samples/bpf/xdpsock_user.c
+++ b/samples/bpf/xdpsock_user.c
@@ -113,6 +113,7 @@ static u32 irq_no;
 static int irqs_at_init = -1;
 static int opt_poll;
 static int opt_interval = 1;
+static int opt_retries = 3;
 static u32 opt_xdp_bind_flags = XDP_USE_NEED_WAKEUP;
 static u32 opt_umem_flags;
 static int opt_unaligned_chunks;
@@ -1028,6 +1029,7 @@ static struct option long_options[] = {
 	{"xdp-skb", no_argument, 0, 'S'},
 	{"xdp-native", no_argument, 0, 'N'},
 	{"interval", required_argument, 0, 'n'},
+	{"retries", required_argument, 0, 'O'},
 	{"zero-copy", no_argument, 0, 'z'},
 	{"copy", no_argument, 0, 'c'},
 	{"frame-size", required_argument, 0, 'f'},
@@ -1072,6 +1074,7 @@ static void usage(const char *prog)
 		"  -S, --xdp-skb=n	Use XDP skb-mod\n"
 		"  -N, --xdp-native=n	Enforce XDP native mode\n"
 		"  -n, --interval=n	Specify statistics update interval (default 1 sec).\n"
+		"  -O, --retries=n	Specify time-out retries (1s interval) attempt (default 3).\n"
 		"  -z, --zero-copy      Force zero-copy mode.\n"
 		"  -c, --copy           Force copy mode.\n"
 		"  -m, --no-need-wakeup Turn off use of driver need wakeup flag.\n"
@@ -1122,7 +1125,7 @@ static void parse_command_line(int argc, char **argv)
 
 	for (;;) {
 		c = getopt_long(argc, argv,
-				"Frtli:q:pSNn:w:czf:muMd:b:C:s:P:VJ:K:G:H:T:W:U:xQaI:BR",
+				"Frtli:q:pSNn:w:O:czf:muMd:b:C:s:P:VJ:K:G:H:T:W:U:xQaI:BR",
 				long_options, &option_index);
 		if (c == -1)
 			break;
@@ -1164,6 +1167,9 @@ static void parse_command_line(int argc, char **argv)
 				opt_clock = CLOCK_MONOTONIC;
 			}
 			break;
+		case 'O':
+			opt_retries = atoi(optarg);
+			break;
 		case 'z':
 			opt_xdp_bind_flags |= XDP_ZEROCOPY;
 			break;
@@ -1509,7 +1515,8 @@ static void complete_tx_only_all(void)
 				pending = !!xsks[i]->outstanding_tx;
 			}
 		}
-	} while (pending);
+		sleep(1);
+	} while (pending && opt_retries-- > 0);
 }
 
 static void tx_only_all(void)
-- 
2.25.1

[PATCH bpf-next v2 7/7] samples/bpf: xdpsock: add timestamp for Tx-only operation

[Ong Boon Leong]

It may be useful to add timestamp for Tx packets for continuous or cyclic
transmit operation. The timestamp and sequence ID of a Tx packet are
stored according to pktgen header format. To enable per-packet timestamp,
use -y|--tstamp option. If timestamp is off, pktgen header is not
included in the UDP payload. This means receiving side can use the magic
number for pktgen for differentiation.

The implementation supports both VLAN tagged and untagged option. By
default, the minimum packet size is set at 64B. However, if VLAN tagged
is on (-V), the minimum packet size is increased to 66B just so to fit
the pktgen_hdr size.

Added hex_dump() into the code path just for future cross-checking.
As before, simply change to "#define DEBUG_HEXDUMP 1" to inspect the
accuracy of TX packet.

Signed-off-by: Ong Boon Leong <boon.leong.ong@intel.com>
---
 samples/bpf/xdpsock_user.c | 77 +++++++++++++++++++++++++++++++++-----
 1 file changed, 68 insertions(+), 9 deletions(-)

diff --git a/samples/bpf/xdpsock_user.c b/samples/bpf/xdpsock_user.c
index 319cb3cdb22..aa50864e441 100644
--- a/samples/bpf/xdpsock_user.c
+++ b/samples/bpf/xdpsock_user.c
@@ -111,6 +111,7 @@ static bool opt_app_stats;
 static const char *opt_irq_str = "";
 static u32 irq_no;
 static int irqs_at_init = -1;
+static u32 sequence;
 static int opt_poll;
 static int opt_interval = 1;
 static int opt_retries = 3;
@@ -129,6 +130,7 @@ static clockid_t opt_clock = CLOCK_MONOTONIC;
 static unsigned long opt_tx_cycle_ns;
 static int opt_schpolicy = SCHED_OTHER;
 static int opt_schprio = SCHED_PRI__DEFAULT;
+static bool opt_tstamp;
 
 struct vlan_ethhdr {
 	unsigned char h_dest[6];
@@ -138,6 +140,14 @@ struct vlan_ethhdr {
 	__be16 h_vlan_encapsulated_proto;
 };
 
+#define PKTGEN_MAGIC 0xbe9be955
+struct pktgen_hdr {
+	__be32 pgh_magic;
+	__be32 seq_num;
+	__be32 tv_sec;
+	__be32 tv_usec;
+};
+
 struct xsk_ring_stats {
 	unsigned long rx_npkts;
 	unsigned long tx_npkts;
@@ -836,18 +846,25 @@ static inline u16 udp_csum(u32 saddr, u32 daddr, u32 len,
 
 #define ETH_HDR_SIZE (opt_vlan_tag ? sizeof(struct vlan_ethhdr) : \
 		      sizeof(struct ethhdr))
+#define PKTGEN_HDR_SIZE (opt_tstamp ? sizeof(struct pktgen_hdr) : 0)
 #define PKT_HDR_SIZE (ETH_HDR_SIZE + sizeof(struct iphdr) + \
-		      sizeof(struct udphdr))
+		      sizeof(struct udphdr) + PKTGEN_HDR_SIZE)
+#define PKTGEN_HDR_OFFSET (ETH_HDR_SIZE + sizeof(struct iphdr) + \
+			   sizeof(struct udphdr))
+#define PKTGEN_SIZE_MIN (PKTGEN_HDR_OFFSET + sizeof(struct pktgen_hdr) + \
+			 ETH_FCS_SIZE)
 
 #define PKT_SIZE		(opt_pkt_size - ETH_FCS_SIZE)
 #define IP_PKT_SIZE		(PKT_SIZE - ETH_HDR_SIZE)
 #define UDP_PKT_SIZE		(IP_PKT_SIZE - sizeof(struct iphdr))
-#define UDP_PKT_DATA_SIZE	(UDP_PKT_SIZE - sizeof(struct udphdr))
+#define UDP_PKT_DATA_SIZE	(UDP_PKT_SIZE - \
+				 (sizeof(struct udphdr) + PKTGEN_HDR_SIZE))
 
 static u8 pkt_data[XSK_UMEM__DEFAULT_FRAME_SIZE];
 
 static void gen_eth_hdr_data(void)
 {
+	struct pktgen_hdr *pktgen_hdr;
 	struct udphdr *udp_hdr;
 	struct iphdr *ip_hdr;
 
@@ -860,7 +877,10 @@ static void gen_eth_hdr_data(void)
 					    sizeof(struct iphdr));
 		ip_hdr = (struct iphdr *)(pkt_data +
 					  sizeof(struct vlan_ethhdr));
-
+		pktgen_hdr = (struct pktgen_hdr *)(pkt_data +
+						   sizeof(struct vlan_ethhdr) +
+						   sizeof(struct iphdr) +
+						   sizeof(struct udphdr));
 		/* ethernet & VLAN header */
 		memcpy(veth_hdr->h_dest, &opt_txdmac, ETH_ALEN);
 		memcpy(veth_hdr->h_source, &opt_txsmac, ETH_ALEN);
@@ -877,7 +897,10 @@ static void gen_eth_hdr_data(void)
 					    sizeof(struct iphdr));
 		ip_hdr = (struct iphdr *)(pkt_data +
 					  sizeof(struct ethhdr));
-
+		pktgen_hdr = (struct pktgen_hdr *)(pkt_data +
+						   sizeof(struct ethhdr) +
+						   sizeof(struct iphdr) +
+						   sizeof(struct udphdr));
 		/* ethernet header */
 		memcpy(eth_hdr->h_dest, &opt_txdmac, ETH_ALEN);
 		memcpy(eth_hdr->h_source, &opt_txsmac, ETH_ALEN);
@@ -906,6 +929,9 @@ static void gen_eth_hdr_data(void)
 	udp_hdr->dest = htons(0x1000);
 	udp_hdr->len = htons(UDP_PKT_SIZE);
 
+	if (opt_tstamp)
+		pktgen_hdr->pgh_magic = htonl(PKTGEN_MAGIC);
+
 	/* UDP data */
 	memset32_htonl(pkt_data + PKT_HDR_SIZE, opt_pkt_fill_pattern,
 		       UDP_PKT_DATA_SIZE);
@@ -1049,6 +1075,7 @@ static struct option long_options[] = {
 	{"tx-dmac", required_argument, 0, 'G'},
 	{"tx-smac", required_argument, 0, 'H'},
 	{"tx-cycle", required_argument, 0, 'T'},
+	{"tstamp", no_argument, 0, 'y'},
 	{"policy", required_argument, 0, 'W'},
 	{"schpri", required_argument, 0, 'U'},
 	{"extra-stats", no_argument, 0, 'x'},
@@ -1099,6 +1126,7 @@ static void usage(const char *prog)
 		"  -G, --tx-dmac=<MAC>  Dest MAC addr of TX frame in aa:bb:cc:dd:ee:ff format (For -V|--tx-vlan)\n"
 		"  -H, --tx-smac=<MAC>  Src MAC addr of TX frame in aa:bb:cc:dd:ee:ff format (For -V|--tx-vlan)\n"
 		"  -T, --tx-cycle=n     Tx cycle time in micro-seconds (For -t|--txonly).\n"
+		"  -y, --tstamp         Add time-stamp to packet (For -t|--txonly).\n"
 		"  -W, --policy=POLICY  Schedule policy. Default: SCHED_OTHER\n"
 		"  -U, --schpri=n       Schedule priority. Default: %d\n"
 		"  -x, --extra-stats	Display extra statistics.\n"
@@ -1125,7 +1153,7 @@ static void parse_command_line(int argc, char **argv)
 
 	for (;;) {
 		c = getopt_long(argc, argv,
-				"Frtli:q:pSNn:w:O:czf:muMd:b:C:s:P:VJ:K:G:H:T:W:U:xQaI:BR",
+				"Frtli:q:pSNn:w:O:czf:muMd:b:C:s:P:VJ:K:G:H:T:yW:U:xQaI:BR",
 				long_options, &option_index);
 		if (c == -1)
 			break;
@@ -1246,6 +1274,9 @@ static void parse_command_line(int argc, char **argv)
 			opt_tx_cycle_ns = atoi(optarg);
 			opt_tx_cycle_ns *= NSEC_PER_USEC;
 			break;
+		case 'y':
+			opt_tstamp = 1;
+			break;
 		case 'W':
 			if (get_schpolicy(&opt_schpolicy, optarg)) {
 				fprintf(stderr,
@@ -1462,9 +1493,10 @@ static void rx_drop_all(void)
 	}
 }
 
-static int tx_only(struct xsk_socket_info *xsk, u32 *frame_nb, int batch_size)
+static int tx_only(struct xsk_socket_info *xsk, u32 *frame_nb,
+		   int batch_size, unsigned long tx_ns)
 {
-	u32 idx;
+	u32 idx, tv_sec, tv_usec;
 	unsigned int i;
 
 	while (xsk_ring_prod__reserve(&xsk->tx, batch_size, &idx) <
@@ -1474,11 +1506,31 @@ static int tx_only(struct xsk_socket_info *xsk, u32 *frame_nb, int batch_size)
 			return 0;
 	}
 
+	if (opt_tstamp) {
+		tv_sec = (u32)(tx_ns / NSEC_PER_SEC);
+		tv_usec = (u32)((tx_ns % NSEC_PER_SEC) / 1000);
+	}
+
 	for (i = 0; i < batch_size; i++) {
 		struct xdp_desc *tx_desc = xsk_ring_prod__tx_desc(&xsk->tx,
 								  idx + i);
 		tx_desc->addr = (*frame_nb + i) * opt_xsk_frame_size;
 		tx_desc->len = PKT_SIZE;
+
+		if (opt_tstamp) {
+			struct pktgen_hdr *pktgen_hdr;
+			u64 addr = tx_desc->addr;
+			char *pkt;
+
+			pkt = xsk_umem__get_data(xsk->umem->buffer, addr);
+			pktgen_hdr = (struct pktgen_hdr *)(pkt + PKTGEN_HDR_OFFSET);
+
+			pktgen_hdr->seq_num = htonl(sequence++);
+			pktgen_hdr->tv_sec = htonl(tv_sec);
+			pktgen_hdr->tv_usec = htonl(tv_usec);
+
+			hex_dump(pkt, PKT_SIZE, addr);
+		}
 	}
 
 	xsk_ring_prod__submit(&xsk->tx, batch_size);
@@ -1552,6 +1604,7 @@ static void tx_only_all(void)
 
 	while ((opt_pkt_count && pkt_cnt < opt_pkt_count) || !opt_pkt_count) {
 		int batch_size = get_batch_size(pkt_cnt);
+		unsigned long tx_ns = 0;
 		struct timespec next;
 		int tx_cnt = 0;
 		long diff;
@@ -1581,7 +1634,8 @@ static void tx_only_all(void)
 			}
 
 			/* Measure periodic Tx scheduling variance */
-			diff = get_nsecs() - next_tx_ns;
+			tx_ns = get_nsecs();
+			diff = tx_ns - next_tx_ns;
 			if (diff < tx_cycle_diff_min)
 				tx_cycle_diff_min = diff;
 
@@ -1590,10 +1644,12 @@ static void tx_only_all(void)
 
 			tx_cycle_diff_ave += (double)diff;
 			tx_cycle_cnt++;
+		} else if (opt_tstamp) {
+			tx_ns = get_nsecs();
 		}
 
 		for (i = 0; i < num_socks; i++)
-			tx_cnt += tx_only(xsks[i], &frame_nb[i], batch_size);
+			tx_cnt += tx_only(xsks[i], &frame_nb[i], batch_size, tx_ns);
 
 		pkt_cnt += tx_cnt;
 
@@ -1895,6 +1951,9 @@ int main(int argc, char **argv)
 		apply_setsockopt(xsks[i]);
 
 	if (opt_bench == BENCH_TXONLY) {
+		if (opt_tstamp && opt_pkt_size < PKTGEN_SIZE_MIN)
+			opt_pkt_size = PKTGEN_SIZE_MIN;
+
 		gen_eth_hdr_data();
 
 		for (i = 0; i < NUM_FRAMES; i++)
-- 
2.25.1