kernel/linux.git/samples, branch v4.2-rc2

bpf: BPF based latency tracing

2015-06-23T13:09:58+00:00

BPF offers another way to generate latency histograms. We attach kprobes at trace_preempt_off and trace_preempt_on and calculate the time it takes to from seeing the off/on transition. The first array is used to store the start time stamp. The key is the CPU id. The second array stores the log2(time diff). We need to use static allocation here (array and not hash tables). The kprobes hooking into trace_preempt_on|off should not calling any dynamic memory allocation or free path. We need to avoid recursivly getting called. Besides that, it reduces jitter in the measurement. CPU 0 latency : count distribution 1 -> 1 : 0 | | 2 -> 3 : 0 | | 4 -> 7 : 0 | | 8 -> 15 : 0 | | 16 -> 31 : 0 | | 32 -> 63 : 0 | | 64 -> 127 : 0 | | 128 -> 255 : 0 | | 256 -> 511 : 0 | | 512 -> 1023 : 0 | | 1024 -> 2047 : 0 | | 2048 -> 4095 : 166723 |*************************************** | 4096 -> 8191 : 19870 |*** | 8192 -> 16383 : 6324 | | 16384 -> 32767 : 1098 | | 32768 -> 65535 : 190 | | 65536 -> 131071 : 179 | | 131072 -> 262143 : 18 | | 262144 -> 524287 : 4 | | 524288 -> 1048575 : 1363 | | CPU 1 latency : count distribution 1 -> 1 : 0 | | 2 -> 3 : 0 | | 4 -> 7 : 0 | | 8 -> 15 : 0 | | 16 -> 31 : 0 | | 32 -> 63 : 0 | | 64 -> 127 : 0 | | 128 -> 255 : 0 | | 256 -> 511 : 0 | | 512 -> 1023 : 0 | | 1024 -> 2047 : 0 | | 2048 -> 4095 : 114042 |*************************************** | 4096 -> 8191 : 9587 |** | 8192 -> 16383 : 4140 | | 16384 -> 32767 : 673 | | 32768 -> 65535 : 179 | | 65536 -> 131071 : 29 | | 131072 -> 262143 : 4 | | 262144 -> 524287 : 1 | | 524288 -> 1048575 : 364 | | CPU 2 latency : count distribution 1 -> 1 : 0 | | 2 -> 3 : 0 | | 4 -> 7 : 0 | | 8 -> 15 : 0 | | 16 -> 31 : 0 | | 32 -> 63 : 0 | | 64 -> 127 : 0 | | 128 -> 255 : 0 | | 256 -> 511 : 0 | | 512 -> 1023 : 0 | | 1024 -> 2047 : 0 | | 2048 -> 4095 : 40147 |*************************************** | 4096 -> 8191 : 2300 |* | 8192 -> 16383 : 828 | | 16384 -> 32767 : 178 | | 32768 -> 65535 : 59 | | 65536 -> 131071 : 2 | | 131072 -> 262143 : 0 | | 262144 -> 524287 : 1 | | 524288 -> 1048575 : 174 | | CPU 3 latency : count distribution 1 -> 1 : 0 | | 2 -> 3 : 0 | | 4 -> 7 : 0 | | 8 -> 15 : 0 | | 16 -> 31 : 0 | | 32 -> 63 : 0 | | 64 -> 127 : 0 | | 128 -> 255 : 0 | | 256 -> 511 : 0 | | 512 -> 1023 : 0 | | 1024 -> 2047 : 0 | | 2048 -> 4095 : 29626 |*************************************** | 4096 -> 8191 : 2704 |** | 8192 -> 16383 : 1090 | | 16384 -> 32767 : 160 | | 32768 -> 65535 : 72 | | 65536 -> 131071 : 32 | | 131072 -> 262143 : 26 | | 262144 -> 524287 : 12 | | 524288 -> 1048575 : 298 | | All this is based on the trace3 examples written by Alexei Starovoitov . Signed-off-by: Daniel Wagner Cc: Alexei Starovoitov Cc: Alexei Starovoitov Cc: "David S. Miller" Cc: Daniel Borkmann Cc: Ingo Molnar Cc: linux-kernel@vger.kernel.org Cc: netdev@vger.kernel.org Acked-by: Alexei Starovoitov Acked-by: Daniel Borkmann Signed-off-by: David S. Miller

bpf: introduce current->pid, tgid, uid, gid, comm accessors

2015-06-15T22:53:50+00:00

eBPF programs attached to kprobes need to filter based on current->pid, uid and other fields, so introduce helper functions: u64 bpf_get_current_pid_tgid(void) Return: current->tgid << 32 | current->pid u64 bpf_get_current_uid_gid(void) Return: current_gid << 32 | current_uid bpf_get_current_comm(char *buf, int size_of_buf) stores current->comm into buf They can be used from the programs attached to TC as well to classify packets based on current task fields. Update tracex2 example to print histogram of write syscalls for each process instead of aggregated for all. Signed-off-by: Alexei Starovoitov Signed-off-by: David S. Miller

bpf: allow programs to write to certain skb fields

2015-06-07T09:01:33+00:00

allow programs read/write skb->mark, tc_index fields and ((struct qdisc_skb_cb *)cb)->data. mark and tc_index are generically useful in TC. cb[0]-cb[4] are primarily used to pass arguments from one program to another called via bpf_tail_call() which can be seen in sockex3_kern.c example. All fields of 'struct __sk_buff' are readable to socket and tc_cls_act progs. mark, tc_index are writeable from tc_cls_act only. cb[0]-cb[4] are writeable by both sockets and tc_cls_act. Add verifier tests and improve sample code. Signed-off-by: Alexei Starovoitov Signed-off-by: David S. Miller

bpf: make programs see skb->data == L2 for ingress and egress

2015-06-07T09:01:33+00:00

eBPF programs attached to ingress and egress qdiscs see inconsistent skb->data. For ingress L2 header is already pulled, whereas for egress it's present. This is known to program writers which are currently forced to use BPF_LL_OFF workaround. Since programs don't change skb internal pointers it is safe to do pull/push right around invocation of the program and earlier taps and later pt->func() will not be affected. Multiple taps via packet_rcv(), tpacket_rcv() are doing the same trick around run_filter/BPF_PROG_RUN even if skb_shared. This fix finally allows programs to use optimized LD_ABS/IND instructions without BPF_LL_OFF for higher performance. tc ingress + cls_bpf + samples/bpf/tcbpf1_kern.o w/o JIT w/JIT before 20.5 23.6 Mpps after 21.8 26.6 Mpps Old programs with BPF_LL_OFF will still work as-is. We can now undo most of the earlier workaround commit: a166151cbe33 ("bpf: fix bpf helpers to use skb->mac_header relative offsets") Signed-off-by: Alexei Starovoitov Acked-by: Jamal Hadi Salim Signed-off-by: David S. Miller

pktgen: add benchmark script pktgen_bench_xmit_mode_netif_receive.sh

2015-05-23T03:59:17+00:00

This script pktgen_bench_xmit_mode_netif_receive.sh is a benchmark script, which can be used for benchmarking part of the network stack. This can be used for performance improving or catching regression in that area. The script is developed for benchmarking ingress qdisc path, original idea by Alexei Starovoitov. This script don't really need any hardware. This is achieved via the recently introduced stack inject feature "xmit_mode netif_receive". See commit 62f64aed622b6 ("pktgen: introduce xmit_mode ''"). Signed-off-by: Jesper Dangaard Brouer Signed-off-by: David S. Miller

pktgen: add sample script pktgen_sample03_burst_single_flow.sh

2015-05-23T03:59:17+00:00

Add the pktgen samples script pktgen_sample03_burst_single_flow.sh that demonstrates how to acheive maximum performance. If correctly tuned[1] single CPU 10Gbit/s wirespeed small pkts is possible[2] which is 14.88Mpps. The trick is to take advantage of the "burst" feature introduced in commit 38b2cf2982dc73 ("net: pktgen: packet bursting via skb->xmit_more"). [1] http://netoptimizer.blogspot.dk/2014/06/pktgen-for-network-overload-testing.html [2] http://netoptimizer.blogspot.dk/2014/10/unlocked-10gbps-tx-wirespeed-smallest.html Signed-off-by: Jesper Dangaard Brouer Signed-off-by: David S. Miller

pktgen: add sample script pktgen_sample02_multiqueue.sh

2015-05-23T03:59:17+00:00

Add the pktgen samples script pktgen_sample02_multiqueue.sh that demonstrates generating packets on multiqueue NICs. Specifically notice the options "-t" that specifies how many kernel threads to activate. Also notice the flag QUEUE_MAP_CPU, which cause the SKB TX queue to be mapped to the CPU running the kernel thread. For best scalability people are also encourage to map NIC IRQ /proc/irq/*/smp_affinity to CPU number. Usage example with "-t" 4 threads and help: ./pktgen_sample02_multiqueue.sh -i eth4 -m 00:1B:21:3C:9D:F8 -t 4 Usage: ./pktgen_sample02_multiqueue.sh [-vx] -i ethX -i : ($DEV) output interface/device (required) -s : ($PKT_SIZE) packet size -d : ($DEST_IP) destination IP -m : ($DST_MAC) destination MAC-addr -t : ($THREADS) threads to start -c : ($SKB_CLONE) SKB clones send before alloc new SKB -b : ($BURST) HW level bursting of SKBs -v : ($VERBOSE) verbose -x : ($DEBUG) debug Removing pktgen.conf-2-1 and pktgen.conf-2-2 as these examples should be covered now. Signed-off-by: Jesper Dangaard Brouer Signed-off-by: David S. Miller

pktgen: add sample script pktgen_sample01_simple.sh

2015-05-23T03:59:17+00:00

Add the first basic pktgen samples script pktgen_sample01_simple.sh, which demonstrates the a simple use of the helper functions. Removing pktgen.conf-1-1 as that example should be covered now. The naming scheme pktgen_sampleNN, where NN is a number, should encourage reading the samples in a specific order. Script cause pktgen sending with a single thread and single interface, and introduce flow variation via random UDP source port. Usage example and help: ./pktgen_sample01_simple.sh -i eth4 -m 00:1B:21:3C:9D:F8 -d 192.168.8.2 Usage: ./pktgen_sample01_simple.sh [-vx] -i ethX -i : ($DEV) output interface/device (required) -s : ($PKT_SIZE) packet size -d : ($DEST_IP) destination IP -m : ($DST_MAC) destination MAC-addr -c : ($SKB_CLONE) SKB clones send before alloc new SKB -v : ($VERBOSE) verbose -x : ($DEBUG) debug Signed-off-by: Jesper Dangaard Brouer Signed-off-by: David S. Miller

pktgen: new pktgen helper functions for samples scripts

2015-05-23T03:59:16+00:00

Preparing for removing existing samples/pktgen/ scripts, and replacing these with easier to use samples. This commit provides two helper shell files, that can be "included" by shell source'ing. Namely "functions.sh" and "parameters.sh". The parameters.sh file support easy and consistant parameter parsing across the sample scripts. Usage example is printed on errors. The functions.sh file provides, three new shell functions for configuring the different components of pktgen: pg_ctrl(), pg_thread() and pg_set(). A slightly improved version of the old pgset() function is also provided for backwards compat. The new functions correspond to pktgens different components. * pg_ctrl() control "pgctrl" (/proc/net/pktgen/pgctrl) * pg_thread() control the kernel threads and binding to devices * pg_set() control setup of individual devices These changes are borrowed from: https://github.com/netoptimizer/network-testing/tree/master/pktgen Signed-off-by: Jesper Dangaard Brouer Signed-off-by: David S. Miller

samples/bpf: bpf_tail_call example for networking

2015-05-21T21:07:59+00:00

Usage: $ sudo ./sockex3 IP src.port -> dst.port bytes packets 127.0.0.1.42010 -> 127.0.0.1.12865 1568 8 127.0.0.1.59526 -> 127.0.0.1.33778 11422636 173070 127.0.0.1.33778 -> 127.0.0.1.59526 11260224828 341974 127.0.0.1.12865 -> 127.0.0.1.42010 1832 12 IP src.port -> dst.port bytes packets 127.0.0.1.42010 -> 127.0.0.1.12865 1568 8 127.0.0.1.59526 -> 127.0.0.1.33778 23198092 351486 127.0.0.1.33778 -> 127.0.0.1.59526 22972698518 698616 127.0.0.1.12865 -> 127.0.0.1.42010 1832 12 this example is similar to sockex2 in a way that it accumulates per-flow statistics, but it does packet parsing differently. sockex2 inlines full packet parser routine into single bpf program. This sockex3 example have 4 independent programs that parse vlan, mpls, ip, ipv6 and one main program that starts the process. bpf_tail_call() mechanism allows each program to be small and be called on demand potentially multiple times, so that many vlan, mpls, ip in ip, gre encapsulations can be parsed. These and other protocol parsers can be added or removed at runtime. TLVs can be parsed in similar manner. Note, tail_call_cnt dynamic check limits the number of tail calls to 32. Signed-off-by: Alexei Starovoitov Signed-off-by: David S. Miller