From ac5e379da2d700d0b0fe6ee826d78a178afe136d Mon Sep 17 00:00:00 2001 From: Alban Crequy Date: Thu, 6 Apr 2017 11:06:38 +0200 Subject: [PATCH] vendor: update gobpf and tcptracer-bpf It will help to get missing kretprobe events during the ebpf tracker initialization and missing connect & accept events, specially visible on single-cores: - https://github.com/weaveworks/tcptracer-bpf/pull/36 - https://github.com/iovisor/gobpf/pull/39 It makes use of a Linux patch that should be in Linux >= 4.12 (unreleased yet): https://github.com/iovisor/bcc/issues/1072#issuecomment-289777787 It falls back gracefully on current/older Linux kernels. The new gobpf vendoring also includes other small fixes such as error handling. --- vendor/github.com/iovisor/gobpf/elf/elf.go | 95 ++- .../iovisor/gobpf/elf/include/bpf.h | 608 ++++++++++++++++++ vendor/github.com/iovisor/gobpf/elf/module.go | 155 ++++- .../iovisor/gobpf/elf/module_unsupported.go | 18 +- .../tcptracer-bpf/pkg/tracer/tracer.go | 8 +- vendor/manifest | 4 +- 6 files changed, 838 insertions(+), 50 deletions(-) create mode 100644 vendor/github.com/iovisor/gobpf/elf/include/bpf.h diff --git a/vendor/github.com/iovisor/gobpf/elf/elf.go b/vendor/github.com/iovisor/gobpf/elf/elf.go index c348b7d711..4047136ca2 100644 --- a/vendor/github.com/iovisor/gobpf/elf/elf.go +++ b/vendor/github.com/iovisor/gobpf/elf/elf.go @@ -43,7 +43,7 @@ import ( #include #include #include -#include +#include "include/bpf.h" #include #include #include @@ -257,9 +257,9 @@ func elfReadMaps(file *elf.File) (map[string]*Map, error) { mapCount := len(data) / C.sizeof_struct_bpf_map_def for i := 0; i < mapCount; i++ { pos := i * C.sizeof_struct_bpf_map_def - cm := C.bpf_load_map((*C.bpf_map_def)(unsafe.Pointer(&data[pos]))) + cm, err := C.bpf_load_map((*C.bpf_map_def)(unsafe.Pointer(&data[pos]))) if cm == nil { - return nil, fmt.Errorf("error while loading map %s", section.Name) + return nil, fmt.Errorf("error while loading map %q: %v", section.Name, err) } m := &Map{ @@ -411,10 +411,25 @@ func (b *Module) Load() error { processed[section.Info] = true secName := rsection.Name + isKprobe := strings.HasPrefix(secName, "kprobe/") isKretprobe := strings.HasPrefix(secName, "kretprobe/") + isCgroupSkb := strings.HasPrefix(secName, "cgroup/skb") + isCgroupSock := strings.HasPrefix(secName, "cgroup/sock") + + var progType uint32 + switch { + case isKprobe: + fallthrough + case isKretprobe: + progType = uint32(C.BPF_PROG_TYPE_KPROBE) + case isCgroupSkb: + progType = uint32(C.BPF_PROG_TYPE_CGROUP_SKB) + case isCgroupSock: + progType = uint32(C.BPF_PROG_TYPE_CGROUP_SOCK) + } - if isKprobe || isKretprobe { + if isKprobe || isKretprobe || isCgroupSkb || isCgroupSock { rdata, err := rsection.Data() if err != nil { return err @@ -431,17 +446,31 @@ func (b *Module) Load() error { insns := (*C.struct_bpf_insn)(unsafe.Pointer(&rdata[0])) - progFd := C.bpf_prog_load(C.BPF_PROG_TYPE_KPROBE, + progFd := C.bpf_prog_load(progType, insns, C.int(rsection.Size), (*C.char)(lp), C.int(version), (*C.char)(unsafe.Pointer(&b.log[0])), C.int(len(b.log))) if progFd < 0 { return fmt.Errorf("error while loading %q:\n%s", secName, b.log) } - b.probes[secName] = &Kprobe{ - Name: secName, - insns: insns, - fd: int(progFd), + + switch { + case isKprobe: + fallthrough + case isKretprobe: + b.probes[secName] = &Kprobe{ + Name: secName, + insns: insns, + fd: int(progFd), + } + case isCgroupSkb: + fallthrough + case isCgroupSock: + b.cgroupPrograms[secName] = &CgroupProgram{ + Name: secName, + insns: insns, + fd: int(progFd), + } } } } @@ -452,7 +481,26 @@ func (b *Module) Load() error { continue } - if strings.HasPrefix(section.Name, "kprobe/") || strings.HasPrefix(section.Name, "kretprobe/") { + secName := section.Name + + isKprobe := strings.HasPrefix(secName, "kprobe/") + isKretprobe := strings.HasPrefix(secName, "kretprobe/") + isCgroupSkb := strings.HasPrefix(secName, "cgroup/skb") + isCgroupSock := strings.HasPrefix(secName, "cgroup/sock") + + var progType uint32 + switch { + case isKprobe: + fallthrough + case isKretprobe: + progType = uint32(C.BPF_PROG_TYPE_KPROBE) + case isCgroupSkb: + progType = uint32(C.BPF_PROG_TYPE_CGROUP_SKB) + case isCgroupSock: + progType = uint32(C.BPF_PROG_TYPE_CGROUP_SOCK) + } + + if isKprobe || isKretprobe || isCgroupSkb || isCgroupSock { data, err := section.Data() if err != nil { return err @@ -464,16 +512,31 @@ func (b *Module) Load() error { insns := (*C.struct_bpf_insn)(unsafe.Pointer(&data[0])) - fd := C.bpf_prog_load(C.BPF_PROG_TYPE_KPROBE, + progFd := C.bpf_prog_load(progType, insns, C.int(section.Size), (*C.char)(lp), C.int(version), (*C.char)(unsafe.Pointer(&b.log[0])), C.int(len(b.log))) - if fd < 0 { + if progFd < 0 { return fmt.Errorf("error while loading %q:\n%s", section.Name, b.log) } - b.probes[section.Name] = &Kprobe{ - Name: section.Name, - fd: int(fd), + + switch { + case isKprobe: + fallthrough + case isKretprobe: + b.probes[secName] = &Kprobe{ + Name: secName, + insns: insns, + fd: int(progFd), + } + case isCgroupSkb: + fallthrough + case isCgroupSock: + b.cgroupPrograms[secName] = &CgroupProgram{ + Name: secName, + insns: insns, + fd: int(progFd), + } } } } @@ -517,7 +580,7 @@ func (b *Module) initializePerfMaps() error { // assign perf fd tp map ret, err := C.bpf_update_element(C.int(b.maps[name].m.fd), unsafe.Pointer(&cpu), unsafe.Pointer(&pmuFD), C.BPF_ANY) if ret != 0 { - return fmt.Errorf("cannot assign perf fd to map %q: %s (cpu %d)", name, err, cpu) + return fmt.Errorf("cannot assign perf fd to map %q: %v (cpu %d)", name, err, cpu) } b.maps[name].pmuFDs = append(b.maps[name].pmuFDs, pmuFD) diff --git a/vendor/github.com/iovisor/gobpf/elf/include/bpf.h b/vendor/github.com/iovisor/gobpf/elf/include/bpf.h new file mode 100644 index 0000000000..0eb0e87dbe --- /dev/null +++ b/vendor/github.com/iovisor/gobpf/elf/include/bpf.h @@ -0,0 +1,608 @@ +/* Copyright (c) 2011-2014 PLUMgrid, http://plumgrid.com + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of version 2 of the GNU General Public + * License as published by the Free Software Foundation. + */ +#ifndef _UAPI__LINUX_BPF_H__ +#define _UAPI__LINUX_BPF_H__ + +#include +#include + +/* Extended instruction set based on top of classic BPF */ + +/* instruction classes */ +#define BPF_ALU64 0x07 /* alu mode in double word width */ + +/* ld/ldx fields */ +#define BPF_DW 0x18 /* double word */ +#define BPF_XADD 0xc0 /* exclusive add */ + +/* alu/jmp fields */ +#define BPF_MOV 0xb0 /* mov reg to reg */ +#define BPF_ARSH 0xc0 /* sign extending arithmetic shift right */ + +/* change endianness of a register */ +#define BPF_END 0xd0 /* flags for endianness conversion: */ +#define BPF_TO_LE 0x00 /* convert to little-endian */ +#define BPF_TO_BE 0x08 /* convert to big-endian */ +#define BPF_FROM_LE BPF_TO_LE +#define BPF_FROM_BE BPF_TO_BE + +#define BPF_JNE 0x50 /* jump != */ +#define BPF_JSGT 0x60 /* SGT is signed '>', GT in x86 */ +#define BPF_JSGE 0x70 /* SGE is signed '>=', GE in x86 */ +#define BPF_CALL 0x80 /* function call */ +#define BPF_EXIT 0x90 /* function return */ + +/* Register numbers */ +enum { + BPF_REG_0 = 0, + BPF_REG_1, + BPF_REG_2, + BPF_REG_3, + BPF_REG_4, + BPF_REG_5, + BPF_REG_6, + BPF_REG_7, + BPF_REG_8, + BPF_REG_9, + BPF_REG_10, + __MAX_BPF_REG, +}; + +/* BPF has 10 general purpose 64-bit registers and stack frame. */ +#define MAX_BPF_REG __MAX_BPF_REG + +struct bpf_insn { + __u8 code; /* opcode */ + __u8 dst_reg:4; /* dest register */ + __u8 src_reg:4; /* source register */ + __s16 off; /* signed offset */ + __s32 imm; /* signed immediate constant */ +}; + +/* BPF syscall commands, see bpf(2) man-page for details. */ +enum bpf_cmd { + BPF_MAP_CREATE, + BPF_MAP_LOOKUP_ELEM, + BPF_MAP_UPDATE_ELEM, + BPF_MAP_DELETE_ELEM, + BPF_MAP_GET_NEXT_KEY, + BPF_PROG_LOAD, + BPF_OBJ_PIN, + BPF_OBJ_GET, + BPF_PROG_ATTACH, + BPF_PROG_DETACH, +}; + +enum bpf_map_type { + BPF_MAP_TYPE_UNSPEC, + BPF_MAP_TYPE_HASH, + BPF_MAP_TYPE_ARRAY, + BPF_MAP_TYPE_PROG_ARRAY, + BPF_MAP_TYPE_PERF_EVENT_ARRAY, + BPF_MAP_TYPE_PERCPU_HASH, + BPF_MAP_TYPE_PERCPU_ARRAY, + BPF_MAP_TYPE_STACK_TRACE, + BPF_MAP_TYPE_CGROUP_ARRAY, + BPF_MAP_TYPE_LRU_HASH, + BPF_MAP_TYPE_LRU_PERCPU_HASH, +}; + +enum bpf_prog_type { + BPF_PROG_TYPE_UNSPEC, + BPF_PROG_TYPE_SOCKET_FILTER, + BPF_PROG_TYPE_KPROBE, + BPF_PROG_TYPE_SCHED_CLS, + BPF_PROG_TYPE_SCHED_ACT, + BPF_PROG_TYPE_TRACEPOINT, + BPF_PROG_TYPE_XDP, + BPF_PROG_TYPE_PERF_EVENT, + BPF_PROG_TYPE_CGROUP_SKB, + BPF_PROG_TYPE_CGROUP_SOCK, + BPF_PROG_TYPE_LWT_IN, + BPF_PROG_TYPE_LWT_OUT, + BPF_PROG_TYPE_LWT_XMIT, +}; + +enum bpf_attach_type { + BPF_CGROUP_INET_INGRESS, + BPF_CGROUP_INET_EGRESS, + BPF_CGROUP_INET_SOCK_CREATE, + __MAX_BPF_ATTACH_TYPE +}; + +#define MAX_BPF_ATTACH_TYPE __MAX_BPF_ATTACH_TYPE + +#define BPF_PSEUDO_MAP_FD 1 + +/* flags for BPF_MAP_UPDATE_ELEM command */ +#define BPF_ANY 0 /* create new element or update existing */ +#define BPF_NOEXIST 1 /* create new element if it didn't exist */ +#define BPF_EXIST 2 /* update existing element */ + +#define BPF_F_NO_PREALLOC (1U << 0) +/* Instead of having one common LRU list in the + * BPF_MAP_TYPE_LRU_[PERCPU_]HASH map, use a percpu LRU list + * which can scale and perform better. + * Note, the LRU nodes (including free nodes) cannot be moved + * across different LRU lists. + */ +#define BPF_F_NO_COMMON_LRU (1U << 1) + +union bpf_attr { + struct { /* anonymous struct used by BPF_MAP_CREATE command */ + __u32 map_type; /* one of enum bpf_map_type */ + __u32 key_size; /* size of key in bytes */ + __u32 value_size; /* size of value in bytes */ + __u32 max_entries; /* max number of entries in a map */ + __u32 map_flags; /* prealloc or not */ + }; + + struct { /* anonymous struct used by BPF_MAP_*_ELEM commands */ + __u32 map_fd; + __aligned_u64 key; + union { + __aligned_u64 value; + __aligned_u64 next_key; + }; + __u64 flags; + }; + + struct { /* anonymous struct used by BPF_PROG_LOAD command */ + __u32 prog_type; /* one of enum bpf_prog_type */ + __u32 insn_cnt; + __aligned_u64 insns; + __aligned_u64 license; + __u32 log_level; /* verbosity level of verifier */ + __u32 log_size; /* size of user buffer */ + __aligned_u64 log_buf; /* user supplied buffer */ + __u32 kern_version; /* checked when prog_type=kprobe */ + }; + + struct { /* anonymous struct used by BPF_OBJ_* commands */ + __aligned_u64 pathname; + __u32 bpf_fd; + }; + + struct { /* anonymous struct used by BPF_PROG_ATTACH/DETACH commands */ + __u32 target_fd; /* container object to attach to */ + __u32 attach_bpf_fd; /* eBPF program to attach */ + __u32 attach_type; + }; +} __attribute__((aligned(8))); + +/* BPF helper function descriptions: + * + * void *bpf_map_lookup_elem(&map, &key) + * Return: Map value or NULL + * + * int bpf_map_update_elem(&map, &key, &value, flags) + * Return: 0 on success or negative error + * + * int bpf_map_delete_elem(&map, &key) + * Return: 0 on success or negative error + * + * int bpf_probe_read(void *dst, int size, void *src) + * Return: 0 on success or negative error + * + * u64 bpf_ktime_get_ns(void) + * Return: current ktime + * + * int bpf_trace_printk(const char *fmt, int fmt_size, ...) + * Return: length of buffer written or negative error + * + * u32 bpf_prandom_u32(void) + * Return: random value + * + * u32 bpf_raw_smp_processor_id(void) + * Return: SMP processor ID + * + * int bpf_skb_store_bytes(skb, offset, from, len, flags) + * store bytes into packet + * @skb: pointer to skb + * @offset: offset within packet from skb->mac_header + * @from: pointer where to copy bytes from + * @len: number of bytes to store into packet + * @flags: bit 0 - if true, recompute skb->csum + * other bits - reserved + * Return: 0 on success or negative error + * + * int bpf_l3_csum_replace(skb, offset, from, to, flags) + * recompute IP checksum + * @skb: pointer to skb + * @offset: offset within packet where IP checksum is located + * @from: old value of header field + * @to: new value of header field + * @flags: bits 0-3 - size of header field + * other bits - reserved + * Return: 0 on success or negative error + * + * int bpf_l4_csum_replace(skb, offset, from, to, flags) + * recompute TCP/UDP checksum + * @skb: pointer to skb + * @offset: offset within packet where TCP/UDP checksum is located + * @from: old value of header field + * @to: new value of header field + * @flags: bits 0-3 - size of header field + * bit 4 - is pseudo header + * other bits - reserved + * Return: 0 on success or negative error + * + * int bpf_tail_call(ctx, prog_array_map, index) + * jump into another BPF program + * @ctx: context pointer passed to next program + * @prog_array_map: pointer to map which type is BPF_MAP_TYPE_PROG_ARRAY + * @index: index inside array that selects specific program to run + * Return: 0 on success or negative error + * + * int bpf_clone_redirect(skb, ifindex, flags) + * redirect to another netdev + * @skb: pointer to skb + * @ifindex: ifindex of the net device + * @flags: bit 0 - if set, redirect to ingress instead of egress + * other bits - reserved + * Return: 0 on success or negative error + * + * 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 + * + * int bpf_get_current_comm(char *buf, int size_of_buf) + * stores current->comm into buf + * Return: 0 on success or negative error + * + * u32 bpf_get_cgroup_classid(skb) + * retrieve a proc's classid + * @skb: pointer to skb + * Return: classid if != 0 + * + * int bpf_skb_vlan_push(skb, vlan_proto, vlan_tci) + * Return: 0 on success or negative error + * + * int bpf_skb_vlan_pop(skb) + * Return: 0 on success or negative error + * + * int bpf_skb_get_tunnel_key(skb, key, size, flags) + * int bpf_skb_set_tunnel_key(skb, key, size, flags) + * retrieve or populate tunnel metadata + * @skb: pointer to skb + * @key: pointer to 'struct bpf_tunnel_key' + * @size: size of 'struct bpf_tunnel_key' + * @flags: room for future extensions + * Return: 0 on success or negative error + * + * u64 bpf_perf_event_read(&map, index) + * Return: Number events read or error code + * + * int bpf_redirect(ifindex, flags) + * redirect to another netdev + * @ifindex: ifindex of the net device + * @flags: bit 0 - if set, redirect to ingress instead of egress + * other bits - reserved + * Return: TC_ACT_REDIRECT + * + * u32 bpf_get_route_realm(skb) + * retrieve a dst's tclassid + * @skb: pointer to skb + * Return: realm if != 0 + * + * int bpf_perf_event_output(ctx, map, index, data, size) + * output perf raw sample + * @ctx: struct pt_regs* + * @map: pointer to perf_event_array map + * @index: index of event in the map + * @data: data on stack to be output as raw data + * @size: size of data + * Return: 0 on success or negative error + * + * int bpf_get_stackid(ctx, map, flags) + * walk user or kernel stack and return id + * @ctx: struct pt_regs* + * @map: pointer to stack_trace map + * @flags: bits 0-7 - numer of stack frames to skip + * bit 8 - collect user stack instead of kernel + * bit 9 - compare stacks by hash only + * bit 10 - if two different stacks hash into the same stackid + * discard old + * other bits - reserved + * Return: >= 0 stackid on success or negative error + * + * s64 bpf_csum_diff(from, from_size, to, to_size, seed) + * calculate csum diff + * @from: raw from buffer + * @from_size: length of from buffer + * @to: raw to buffer + * @to_size: length of to buffer + * @seed: optional seed + * Return: csum result or negative error code + * + * int bpf_skb_get_tunnel_opt(skb, opt, size) + * retrieve tunnel options metadata + * @skb: pointer to skb + * @opt: pointer to raw tunnel option data + * @size: size of @opt + * Return: option size + * + * int bpf_skb_set_tunnel_opt(skb, opt, size) + * populate tunnel options metadata + * @skb: pointer to skb + * @opt: pointer to raw tunnel option data + * @size: size of @opt + * Return: 0 on success or negative error + * + * int bpf_skb_change_proto(skb, proto, flags) + * Change protocol of the skb. Currently supported is v4 -> v6, + * v6 -> v4 transitions. The helper will also resize the skb. eBPF + * program is expected to fill the new headers via skb_store_bytes + * and lX_csum_replace. + * @skb: pointer to skb + * @proto: new skb->protocol type + * @flags: reserved + * Return: 0 on success or negative error + * + * int bpf_skb_change_type(skb, type) + * Change packet type of skb. + * @skb: pointer to skb + * @type: new skb->pkt_type type + * Return: 0 on success or negative error + * + * int bpf_skb_under_cgroup(skb, map, index) + * Check cgroup2 membership of skb + * @skb: pointer to skb + * @map: pointer to bpf_map in BPF_MAP_TYPE_CGROUP_ARRAY type + * @index: index of the cgroup in the bpf_map + * Return: + * == 0 skb failed the cgroup2 descendant test + * == 1 skb succeeded the cgroup2 descendant test + * < 0 error + * + * u32 bpf_get_hash_recalc(skb) + * Retrieve and possibly recalculate skb->hash. + * @skb: pointer to skb + * Return: hash + * + * u64 bpf_get_current_task(void) + * Returns current task_struct + * Return: current + * + * int bpf_probe_write_user(void *dst, void *src, int len) + * safely attempt to write to a location + * @dst: destination address in userspace + * @src: source address on stack + * @len: number of bytes to copy + * Return: 0 on success or negative error + * + * int bpf_current_task_under_cgroup(map, index) + * Check cgroup2 membership of current task + * @map: pointer to bpf_map in BPF_MAP_TYPE_CGROUP_ARRAY type + * @index: index of the cgroup in the bpf_map + * Return: + * == 0 current failed the cgroup2 descendant test + * == 1 current succeeded the cgroup2 descendant test + * < 0 error + * + * int bpf_skb_change_tail(skb, len, flags) + * The helper will resize the skb to the given new size, to be used f.e. + * with control messages. + * @skb: pointer to skb + * @len: new skb length + * @flags: reserved + * Return: 0 on success or negative error + * + * int bpf_skb_pull_data(skb, len) + * The helper will pull in non-linear data in case the skb is non-linear + * and not all of len are part of the linear section. Only needed for + * read/write with direct packet access. + * @skb: pointer to skb + * @len: len to make read/writeable + * Return: 0 on success or negative error + * + * s64 bpf_csum_update(skb, csum) + * Adds csum into skb->csum in case of CHECKSUM_COMPLETE. + * @skb: pointer to skb + * @csum: csum to add + * Return: csum on success or negative error + * + * void bpf_set_hash_invalid(skb) + * Invalidate current skb->hash. + * @skb: pointer to skb + * + * int bpf_get_numa_node_id() + * Return: Id of current NUMA node. + * + * int bpf_skb_change_head() + * Grows headroom of skb and adjusts MAC header offset accordingly. + * Will extends/reallocae as required automatically. + * May change skb data pointer and will thus invalidate any check + * performed for direct packet access. + * @skb: pointer to skb + * @len: length of header to be pushed in front + * @flags: Flags (unused for now) + * Return: 0 on success or negative error + * + * int bpf_xdp_adjust_head(xdp_md, delta) + * Adjust the xdp_md.data by delta + * @xdp_md: pointer to xdp_md + * @delta: An positive/negative integer to be added to xdp_md.data + * Return: 0 on success or negative on error + */ +#define __BPF_FUNC_MAPPER(FN) \ + FN(unspec), \ + FN(map_lookup_elem), \ + FN(map_update_elem), \ + FN(map_delete_elem), \ + FN(probe_read), \ + FN(ktime_get_ns), \ + FN(trace_printk), \ + FN(get_prandom_u32), \ + FN(get_smp_processor_id), \ + FN(skb_store_bytes), \ + FN(l3_csum_replace), \ + FN(l4_csum_replace), \ + FN(tail_call), \ + FN(clone_redirect), \ + FN(get_current_pid_tgid), \ + FN(get_current_uid_gid), \ + FN(get_current_comm), \ + FN(get_cgroup_classid), \ + FN(skb_vlan_push), \ + FN(skb_vlan_pop), \ + FN(skb_get_tunnel_key), \ + FN(skb_set_tunnel_key), \ + FN(perf_event_read), \ + FN(redirect), \ + FN(get_route_realm), \ + FN(perf_event_output), \ + FN(skb_load_bytes), \ + FN(get_stackid), \ + FN(csum_diff), \ + FN(skb_get_tunnel_opt), \ + FN(skb_set_tunnel_opt), \ + FN(skb_change_proto), \ + FN(skb_change_type), \ + FN(skb_under_cgroup), \ + FN(get_hash_recalc), \ + FN(get_current_task), \ + FN(probe_write_user), \ + FN(current_task_under_cgroup), \ + FN(skb_change_tail), \ + FN(skb_pull_data), \ + FN(csum_update), \ + FN(set_hash_invalid), \ + FN(get_numa_node_id), \ + FN(skb_change_head), \ + FN(xdp_adjust_head), + +/* integer value in 'imm' field of BPF_CALL instruction selects which helper + * function eBPF program intends to call + */ +#define __BPF_ENUM_FN(x) BPF_FUNC_ ## x +enum bpf_func_id { + __BPF_FUNC_MAPPER(__BPF_ENUM_FN) + __BPF_FUNC_MAX_ID, +}; +#undef __BPF_ENUM_FN + +/* All flags used by eBPF helper functions, placed here. */ + +/* BPF_FUNC_skb_store_bytes flags. */ +#define BPF_F_RECOMPUTE_CSUM (1ULL << 0) +#define BPF_F_INVALIDATE_HASH (1ULL << 1) + +/* BPF_FUNC_l3_csum_replace and BPF_FUNC_l4_csum_replace flags. + * First 4 bits are for passing the header field size. + */ +#define BPF_F_HDR_FIELD_MASK 0xfULL + +/* BPF_FUNC_l4_csum_replace flags. */ +#define BPF_F_PSEUDO_HDR (1ULL << 4) +#define BPF_F_MARK_MANGLED_0 (1ULL << 5) + +/* BPF_FUNC_clone_redirect and BPF_FUNC_redirect flags. */ +#define BPF_F_INGRESS (1ULL << 0) + +/* BPF_FUNC_skb_set_tunnel_key and BPF_FUNC_skb_get_tunnel_key flags. */ +#define BPF_F_TUNINFO_IPV6 (1ULL << 0) + +/* BPF_FUNC_get_stackid flags. */ +#define BPF_F_SKIP_FIELD_MASK 0xffULL +#define BPF_F_USER_STACK (1ULL << 8) +#define BPF_F_FAST_STACK_CMP (1ULL << 9) +#define BPF_F_REUSE_STACKID (1ULL << 10) + +/* BPF_FUNC_skb_set_tunnel_key flags. */ +#define BPF_F_ZERO_CSUM_TX (1ULL << 1) +#define BPF_F_DONT_FRAGMENT (1ULL << 2) + +/* BPF_FUNC_perf_event_output and BPF_FUNC_perf_event_read flags. */ +#define BPF_F_INDEX_MASK 0xffffffffULL +#define BPF_F_CURRENT_CPU BPF_F_INDEX_MASK +/* BPF_FUNC_perf_event_output for sk_buff input context. */ +#define BPF_F_CTXLEN_MASK (0xfffffULL << 32) + +/* user accessible mirror of in-kernel sk_buff. + * new fields can only be added to the end of this structure + */ +struct __sk_buff { + __u32 len; + __u32 pkt_type; + __u32 mark; + __u32 queue_mapping; + __u32 protocol; + __u32 vlan_present; + __u32 vlan_tci; + __u32 vlan_proto; + __u32 priority; + __u32 ingress_ifindex; + __u32 ifindex; + __u32 tc_index; + __u32 cb[5]; + __u32 hash; + __u32 tc_classid; + __u32 data; + __u32 data_end; +}; + +struct bpf_tunnel_key { + __u32 tunnel_id; + union { + __u32 remote_ipv4; + __u32 remote_ipv6[4]; + }; + __u8 tunnel_tos; + __u8 tunnel_ttl; + __u16 tunnel_ext; + __u32 tunnel_label; +}; + +/* Generic BPF return codes which all BPF program types may support. + * The values are binary compatible with their TC_ACT_* counter-part to + * provide backwards compatibility with existing SCHED_CLS and SCHED_ACT + * programs. + * + * XDP is handled seprately, see XDP_*. + */ +enum bpf_ret_code { + BPF_OK = 0, + /* 1 reserved */ + BPF_DROP = 2, + /* 3-6 reserved */ + BPF_REDIRECT = 7, + /* >127 are reserved for prog type specific return codes */ +}; + +struct bpf_sock { + __u32 bound_dev_if; + __u32 family; + __u32 type; + __u32 protocol; +}; + +#define XDP_PACKET_HEADROOM 256 + +/* User return codes for XDP prog type. + * A valid XDP program must return one of these defined values. All other + * return codes are reserved for future use. Unknown return codes will result + * in packet drop. + */ +enum xdp_action { + XDP_ABORTED = 0, + XDP_DROP, + XDP_PASS, + XDP_TX, +}; + +/* user accessible metadata for XDP packet hook + * new fields must be added to the end of this structure + */ +struct xdp_md { + __u32 data; + __u32 data_end; +}; + +#endif /* _UAPI__LINUX_BPF_H__ */ diff --git a/vendor/github.com/iovisor/gobpf/elf/module.go b/vendor/github.com/iovisor/gobpf/elf/module.go index d5cef78908..1bd269255b 100644 --- a/vendor/github.com/iovisor/gobpf/elf/module.go +++ b/vendor/github.com/iovisor/gobpf/elf/module.go @@ -20,6 +20,7 @@ package elf import ( "debug/elf" + "errors" "fmt" "io" "io/ioutil" @@ -31,7 +32,8 @@ import ( /* #include -#include +#include +#include "include/bpf.h" #include #include @@ -48,6 +50,19 @@ static int perf_event_open_tracepoint(int tracepoint_id, int pid, int cpu, return syscall(__NR_perf_event_open, &attr, pid, cpu, group_fd, flags); } + +int bpf_prog_attach(int prog_fd, int target_fd, enum bpf_attach_type type) +{ + union bpf_attr attr; + + bzero(&attr, sizeof(attr)); + attr.target_fd = target_fd; + attr.attach_bpf_fd = prog_fd; + attr.attach_type = type; + + // BPF_PROG_ATTACH = 8 + return syscall(__NR_bpf, 8, &attr, sizeof(attr)); +} */ import "C" @@ -56,9 +71,10 @@ type Module struct { fileReader io.ReaderAt file *elf.File - log []byte - maps map[string]*Map - probes map[string]*Kprobe + log []byte + maps map[string]*Map + probes map[string]*Kprobe + cgroupPrograms map[string]*CgroupProgram } // Kprobe represents a kprobe or kretprobe and has to be declared @@ -70,11 +86,27 @@ type Kprobe struct { efd int } +type AttachType int + +const ( + IngressType AttachType = iota + EgressType + SockCreateType +) + +// CgroupProgram represents a cgroup skb/sock program +type CgroupProgram struct { + Name string + insns *C.struct_bpf_insn + fd int +} + func NewModule(fileName string) *Module { return &Module{ - fileName: fileName, - probes: make(map[string]*Kprobe), - log: make([]byte, 65536), + fileName: fileName, + probes: make(map[string]*Kprobe), + cgroupPrograms: make(map[string]*CgroupProgram), + log: make([]byte, 65536), } } @@ -86,7 +118,46 @@ func NewModuleFromReader(fileReader io.ReaderAt) *Module { } } -func (b *Module) EnableKprobe(secName string) error { +var kprobeIDNotExist error = errors.New("kprobe id file doesn't exist") + +func writeKprobeEvent(probeType, eventName, funcName, maxactiveStr string) (int, error) { + kprobeEventsFileName := "/sys/kernel/debug/tracing/kprobe_events" + f, err := os.OpenFile(kprobeEventsFileName, os.O_APPEND|os.O_WRONLY, 0666) + if err != nil { + return -1, fmt.Errorf("cannot open kprobe_events: %v\n", err) + } + defer f.Close() + + cmd := fmt.Sprintf("%s%s:%s %s\n", probeType, maxactiveStr, eventName, funcName) + _, err = f.WriteString(cmd) + if err != nil { + return -1, fmt.Errorf("cannot write %q to kprobe_events: %v\n", cmd, err) + } + + kprobeIdFile := fmt.Sprintf("/sys/kernel/debug/tracing/events/kprobes/%s/id", eventName) + kprobeIdBytes, err := ioutil.ReadFile(kprobeIdFile) + if err != nil { + if os.IsNotExist(err) { + return -1, kprobeIDNotExist + } + return -1, fmt.Errorf("cannot read kprobe id: %v\n", err) + } + + kprobeId, err := strconv.Atoi(strings.TrimSpace(string(kprobeIdBytes))) + if err != nil { + return -1, fmt.Errorf("invalid kprobe id: %v\n", err) + } + + return kprobeId, nil +} + +// EnableKprobe enables a kprobe/kretprobe identified by secName. +// For kretprobes, you can configure the maximum number of instances +// of the function that can be probed simultaneously with maxactive. +// If maxactive is 0 it will be set to the default value: if CONFIG_PREEMPT is +// enabled, this is max(10, 2*NR_CPUS); otherwise, it is NR_CPUS. +// For kprobes, maxactive is ignored. +func (b *Module) EnableKprobe(secName string, maxactive int) error { var probeType, funcName string isKretprobe := strings.HasPrefix(secName, "kretprobe/") probe, ok := b.probes[secName] @@ -94,36 +165,26 @@ func (b *Module) EnableKprobe(secName string) error { return fmt.Errorf("no such kprobe %q", secName) } progFd := probe.fd + var maxactiveStr string if isKretprobe { probeType = "r" funcName = strings.TrimPrefix(secName, "kretprobe/") + if maxactive > 0 { + maxactiveStr = fmt.Sprintf("%d", maxactive) + } } else { probeType = "p" funcName = strings.TrimPrefix(secName, "kprobe/") } eventName := probeType + funcName - kprobeEventsFileName := "/sys/kernel/debug/tracing/kprobe_events" - f, err := os.OpenFile(kprobeEventsFileName, os.O_APPEND|os.O_WRONLY, 0666) - if err != nil { - return fmt.Errorf("cannot open kprobe_events: %v\n", err) + kprobeId, err := writeKprobeEvent(probeType, eventName, funcName, maxactiveStr) + // fallback without maxactive + if err == kprobeIDNotExist { + kprobeId, err = writeKprobeEvent(probeType, eventName, funcName, "") } - defer f.Close() - - cmd := fmt.Sprintf("%s:%s %s\n", probeType, eventName, funcName) - _, err = f.WriteString(cmd) if err != nil { - return fmt.Errorf("cannot write %q to kprobe_events: %v\n", cmd, err) - } - - kprobeIdFile := fmt.Sprintf("/sys/kernel/debug/tracing/events/kprobes/%s/id", eventName) - kprobeIdBytes, err := ioutil.ReadFile(kprobeIdFile) - if err != nil { - return fmt.Errorf("cannot read kprobe id: %v\n", err) - } - kprobeId, err := strconv.Atoi(strings.TrimSpace(string(kprobeIdBytes))) - if err != nil { - return fmt.Errorf("invalid kprobe id): %v\n", err) + return err } efd := C.perf_event_open_tracepoint(C.int(kprobeId), -1 /* pid */, 0 /* cpu */, -1 /* group_fd */, C.PERF_FLAG_FD_CLOEXEC) @@ -144,6 +205,8 @@ func (b *Module) EnableKprobe(secName string) error { return nil } +// IterKprobes returns a channel that emits the kprobes that included in the +// module. func (b *Module) IterKprobes() <-chan *Kprobe { ch := make(chan *Kprobe) go func() { @@ -155,13 +218,47 @@ func (b *Module) IterKprobes() <-chan *Kprobe { return ch } -func (b *Module) EnableKprobes() error { +// EnableKprobes enables all kprobes/kretprobes included in the module. The +// value in maxactive will be applied to all the kretprobes. +func (b *Module) EnableKprobes(maxactive int) error { var err error for _, kprobe := range b.probes { - err = b.EnableKprobe(kprobe.Name) + err = b.EnableKprobe(kprobe.Name, maxactive) if err != nil { return err } } return nil } + +func (b *Module) IterCgroupProgram() <-chan *CgroupProgram { + ch := make(chan *CgroupProgram) + go func() { + for name := range b.cgroupPrograms { + ch <- b.cgroupPrograms[name] + } + close(ch) + }() + return ch +} + +func (b *Module) CgroupProgram(name string) *CgroupProgram { + return b.cgroupPrograms[name] +} + +func (b *Module) AttachCgroupProgram(cgroupProg *CgroupProgram, cgroupPath string, attachType AttachType) error { + f, err := os.Open(cgroupPath) + if err != nil { + return fmt.Errorf("error opening cgroup %q: %v", cgroupPath, err) + } + defer f.Close() + + progFd := C.int(cgroupProg.fd) + cgroupFd := C.int(f.Fd()) + ret, err := C.bpf_prog_attach(progFd, cgroupFd, uint32(attachType)) + if ret < 0 { + return fmt.Errorf("failed to attach prog to cgroup %q: %v\n", cgroupPath, err) + } + + return nil +} diff --git a/vendor/github.com/iovisor/gobpf/elf/module_unsupported.go b/vendor/github.com/iovisor/gobpf/elf/module_unsupported.go index d978ee3a24..74765ec7e9 100644 --- a/vendor/github.com/iovisor/gobpf/elf/module_unsupported.go +++ b/vendor/github.com/iovisor/gobpf/elf/module_unsupported.go @@ -9,6 +9,8 @@ import ( type Module struct{} type Kprobe struct{} +type CgroupProgram struct{} +type AttachType struct{} func NewModule(fileName string) *Module { return nil @@ -18,7 +20,7 @@ func NewModuleFromReader(fileReader io.ReaderAt) *Module { return nil } -func (b *Module) EnableKprobe(secName string) error { +func (b *Module) EnableKprobe(secName string, maxactive int) error { return fmt.Errorf("not supported") } @@ -26,6 +28,18 @@ func (b *Module) IterKprobes() <-chan *Kprobe { return nil } -func (b *Module) EnableKprobes() error { +func (b *Module) EnableKprobes(maxactive int) error { + return fmt.Errorf("not supported") +} + +func (b *Module) IterCgroupProgram() <-chan *CgroupProgram { + return nil +} + +func (b *Module) CgroupProgram(name string) *CgroupProgram { + return nil +} + +func (b *Module) AttachProgram(cgroupProg *CgroupProgram, cgroupPath string, attachType AttachType) error { return fmt.Errorf("not supported") } diff --git a/vendor/github.com/weaveworks/tcptracer-bpf/pkg/tracer/tracer.go b/vendor/github.com/weaveworks/tcptracer-bpf/pkg/tracer/tracer.go index 766cbb0a53..3e16f643a6 100644 --- a/vendor/github.com/weaveworks/tcptracer-bpf/pkg/tracer/tracer.go +++ b/vendor/github.com/weaveworks/tcptracer-bpf/pkg/tracer/tracer.go @@ -16,6 +16,12 @@ type Tracer struct { stopChan chan struct{} } +// maxActive configures the maximum number of instances of the probed functions +// that can be handled simultaneously. +// This value should be enough to handle typical workloads (for example, some +// amount of processes blocked on the accept syscall). +const maxActive = 128 + func TracerAsset() ([]byte, error) { buf, err := Asset("tcptracer-ebpf.o") if err != nil { @@ -41,7 +47,7 @@ func NewTracer(tcpEventCbV4 func(TcpV4), tcpEventCbV6 func(TcpV6)) (*Tracer, err return nil, err } - err = m.EnableKprobes() + err = m.EnableKprobes(maxActive) if err != nil { return nil, err } diff --git a/vendor/manifest b/vendor/manifest index 30ec28ec74..16c7345aed 100644 --- a/vendor/manifest +++ b/vendor/manifest @@ -1020,7 +1020,7 @@ "importpath": "github.com/iovisor/gobpf/elf", "repository": "https://github.com/iovisor/gobpf", "vcs": "git", - "revision": "21a9e281bf73650b9245d4ebcf111716338ae652", + "revision": "65e4048660d6c4339ebae113ac55b1af6f01305d", "branch": "master", "path": "/elf", "notests": true @@ -1462,7 +1462,7 @@ "importpath": "github.com/weaveworks/tcptracer-bpf", "repository": "https://github.com/weaveworks/tcptracer-bpf", "vcs": "git", - "revision": "5b0b56d81d81cf9739739d347798a04ba1754b2e", + "revision": "b715a3b635b8d9c4a096bbd6009826b57fe64c38", "branch": "master", "notests": true },