// SPDX-License-Identifier: GPL-2.0 /* * event probes * * Part of this code was copied from kernel/trace/trace_kprobe.c written by * Masami Hiramatsu * * Copyright (C) 2021, VMware Inc, Steven Rostedt * Copyright (C) 2021, VMware Inc, Tzvetomir Stoyanov tz.stoyanov@gmail.com> * */ #include #include #include #include "trace_dynevent.h" #include "trace_probe.h" #include "trace_probe_tmpl.h" #define EPROBE_EVENT_SYSTEM "eprobes" struct trace_eprobe { /* tracepoint system */ const char *event_system; /* tracepoint event */ const char *event_name; struct trace_event_call *event; struct dyn_event devent; struct trace_probe tp; }; struct eprobe_data { struct trace_event_file *file; struct trace_eprobe *ep; }; static int __trace_eprobe_create(int argc, const char *argv[]); static void trace_event_probe_cleanup(struct trace_eprobe *ep) { if (!ep) return; trace_probe_cleanup(&ep->tp); kfree(ep->event_name); kfree(ep->event_system); if (ep->event) trace_event_put_ref(ep->event); kfree(ep); } static struct trace_eprobe *to_trace_eprobe(struct dyn_event *ev) { return container_of(ev, struct trace_eprobe, devent); } static int eprobe_dyn_event_create(const char *raw_command) { return trace_probe_create(raw_command, __trace_eprobe_create); } static int eprobe_dyn_event_show(struct seq_file *m, struct dyn_event *ev) { struct trace_eprobe *ep = to_trace_eprobe(ev); int i; seq_printf(m, "e:%s/%s", trace_probe_group_name(&ep->tp), trace_probe_name(&ep->tp)); seq_printf(m, " %s.%s", ep->event_system, ep->event_name); for (i = 0; i < ep->tp.nr_args; i++) seq_printf(m, " %s=%s", ep->tp.args[i].name, ep->tp.args[i].comm); seq_putc(m, '\n'); return 0; } static int unregister_trace_eprobe(struct trace_eprobe *ep) { /* If other probes are on the event, just unregister eprobe */ if (trace_probe_has_sibling(&ep->tp)) goto unreg; /* Enabled event can not be unregistered */ if (trace_probe_is_enabled(&ep->tp)) return -EBUSY; /* Will fail if probe is being used by ftrace or perf */ if (trace_probe_unregister_event_call(&ep->tp)) return -EBUSY; unreg: dyn_event_remove(&ep->devent); trace_probe_unlink(&ep->tp); return 0; } static int eprobe_dyn_event_release(struct dyn_event *ev) { struct trace_eprobe *ep = to_trace_eprobe(ev); int ret = unregister_trace_eprobe(ep); if (!ret) trace_event_probe_cleanup(ep); return ret; } static bool eprobe_dyn_event_is_busy(struct dyn_event *ev) { struct trace_eprobe *ep = to_trace_eprobe(ev); return trace_probe_is_enabled(&ep->tp); } static bool eprobe_dyn_event_match(const char *system, const char *event, int argc, const char **argv, struct dyn_event *ev) { struct trace_eprobe *ep = to_trace_eprobe(ev); const char *slash; /* * We match the following: * event only - match all eprobes with event name * system and event only - match all system/event probes * * The below has the above satisfied with more arguments: * * attached system/event - If the arg has the system and event * the probe is attached to, match * probes with the attachment. * * If any more args are given, then it requires a full match. */ /* * If system exists, but this probe is not part of that system * do not match. */ if (system && strcmp(trace_probe_group_name(&ep->tp), system) != 0) return false; /* Must match the event name */ if (strcmp(trace_probe_name(&ep->tp), event) != 0) return false; /* No arguments match all */ if (argc < 1) return true; /* First argument is the system/event the probe is attached to */ slash = strchr(argv[0], '/'); if (!slash) slash = strchr(argv[0], '.'); if (!slash) return false; if (strncmp(ep->event_system, argv[0], slash - argv[0])) return false; if (strcmp(ep->event_name, slash + 1)) return false; argc--; argv++; /* If there are no other args, then match */ if (argc < 1) return true; return trace_probe_match_command_args(&ep->tp, argc, argv); } static struct dyn_event_operations eprobe_dyn_event_ops = { .create = eprobe_dyn_event_create, .show = eprobe_dyn_event_show, .is_busy = eprobe_dyn_event_is_busy, .free = eprobe_dyn_event_release, .match = eprobe_dyn_event_match, }; static struct trace_eprobe *alloc_event_probe(const char *group, const char *this_event, struct trace_event_call *event, int nargs) { struct trace_eprobe *ep; const char *event_name; const char *sys_name; int ret = -ENOMEM; if (!event) return ERR_PTR(-ENODEV); sys_name = event->class->system; event_name = trace_event_name(event); ep = kzalloc(struct_size(ep, tp.args, nargs), GFP_KERNEL); if (!ep) { trace_event_put_ref(event); goto error; } ep->event = event; ep->event_name = kstrdup(event_name, GFP_KERNEL); if (!ep->event_name) goto error; ep->event_system = kstrdup(sys_name, GFP_KERNEL); if (!ep->event_system) goto error; ret = trace_probe_init(&ep->tp, this_event, group, false); if (ret < 0) goto error; dyn_event_init(&ep->devent, &eprobe_dyn_event_ops); return ep; error: trace_event_probe_cleanup(ep); return ERR_PTR(ret); } static int trace_eprobe_tp_arg_update(struct trace_eprobe *ep, int i) { struct probe_arg *parg = &ep->tp.args[i]; struct ftrace_event_field *field; struct list_head *head; head = trace_get_fields(ep->event); list_for_each_entry(field, head, link) { if (!strcmp(parg->code->data, field->name)) { kfree(parg->code->data); parg->code->data = field; return 0; } } kfree(parg->code->data); parg->code->data = NULL; return -ENOENT; } static int eprobe_event_define_fields(struct trace_event_call *event_call) { int ret; struct eprobe_trace_entry_head field; struct trace_probe *tp; tp = trace_probe_primary_from_call(event_call); if (WARN_ON_ONCE(!tp)) return -ENOENT; DEFINE_FIELD(unsigned int, type, FIELD_STRING_TYPE, 0); return traceprobe_define_arg_fields(event_call, sizeof(field), tp); } static struct trace_event_fields eprobe_fields_array[] = { { .type = TRACE_FUNCTION_TYPE, .define_fields = eprobe_event_define_fields }, {} }; /* Event entry printers */ static enum print_line_t print_eprobe_event(struct trace_iterator *iter, int flags, struct trace_event *event) { struct eprobe_trace_entry_head *field; struct trace_event_call *pevent; struct trace_event *probed_event; struct trace_seq *s = &iter->seq; struct trace_probe *tp; field = (struct eprobe_trace_entry_head *)iter->ent; tp = trace_probe_primary_from_call( container_of(event, struct trace_event_call, event)); if (WARN_ON_ONCE(!tp)) goto out; trace_seq_printf(s, "%s: (", trace_probe_name(tp)); probed_event = ftrace_find_event(field->type); if (probed_event) { pevent = container_of(probed_event, struct trace_event_call, event); trace_seq_printf(s, "%s.%s", pevent->class->system, trace_event_name(pevent)); } else { trace_seq_printf(s, "%u", field->type); } trace_seq_putc(s, ')'); if (print_probe_args(s, tp->args, tp->nr_args, (u8 *)&field[1], field) < 0) goto out; trace_seq_putc(s, '\n'); out: return trace_handle_return(s); } static unsigned long get_event_field(struct fetch_insn *code, void *rec) { struct ftrace_event_field *field = code->data; unsigned long val; void *addr; addr = rec + field->offset; switch (field->size) { case 1: if (field->is_signed) val = *(char *)addr; else val = *(unsigned char *)addr; break; case 2: if (field->is_signed) val = *(short *)addr; else val = *(unsigned short *)addr; break; case 4: if (field->is_signed) val = *(int *)addr; else val = *(unsigned int *)addr; break; default: if (field->is_signed) val = *(long *)addr; else val = *(unsigned long *)addr; break; } return val; } static int get_eprobe_size(struct trace_probe *tp, void *rec) { struct probe_arg *arg; int i, len, ret = 0; for (i = 0; i < tp->nr_args; i++) { arg = tp->args + i; if (unlikely(arg->dynamic)) { unsigned long val; val = get_event_field(arg->code, rec); len = process_fetch_insn_bottom(arg->code + 1, val, NULL, NULL); if (len > 0) ret += len; } } return ret; } /* Kprobe specific fetch functions */ /* Note that we don't verify it, since the code does not come from user space */ static int process_fetch_insn(struct fetch_insn *code, void *rec, void *dest, void *base) { unsigned long val; val = get_event_field(code, rec); return process_fetch_insn_bottom(code + 1, val, dest, base); } NOKPROBE_SYMBOL(process_fetch_insn) /* Return the length of string -- including null terminal byte */ static nokprobe_inline int fetch_store_strlen_user(unsigned long addr) { const void __user *uaddr = (__force const void __user *)addr; return strnlen_user_nofault(uaddr, MAX_STRING_SIZE); } /* Return the length of string -- including null terminal byte */ static nokprobe_inline int fetch_store_strlen(unsigned long addr) { int ret, len = 0; u8 c; #ifdef CONFIG_ARCH_HAS_NON_OVERLAPPING_ADDRESS_SPACE if (addr < TASK_SIZE) return fetch_store_strlen_user(addr); #endif do { ret = copy_from_kernel_nofault(&c, (u8 *)addr + len, 1); len++; } while (c && ret == 0 && len < MAX_STRING_SIZE); return (ret < 0) ? ret : len; } /* * Fetch a null-terminated string from user. Caller MUST set *(u32 *)buf * with max length and relative data location. */ static nokprobe_inline int fetch_store_string_user(unsigned long addr, void *dest, void *base) { const void __user *uaddr = (__force const void __user *)addr; int maxlen = get_loc_len(*(u32 *)dest); void *__dest; long ret; if (unlikely(!maxlen)) return -ENOMEM; __dest = get_loc_data(dest, base); ret = strncpy_from_user_nofault(__dest, uaddr, maxlen); if (ret >= 0) *(u32 *)dest = make_data_loc(ret, __dest - base); return ret; } /* * Fetch a null-terminated string. Caller MUST set *(u32 *)buf with max * length and relative data location. */ static nokprobe_inline int fetch_store_string(unsigned long addr, void *dest, void *base) { int maxlen = get_loc_len(*(u32 *)dest); void *__dest; long ret; #ifdef CONFIG_ARCH_HAS_NON_OVERLAPPING_ADDRESS_SPACE if ((unsigned long)addr < TASK_SIZE) return fetch_store_string_user(addr, dest, base); #endif if (unlikely(!maxlen)) return -ENOMEM; __dest = get_loc_data(dest, base); /* * Try to get string again, since the string can be changed while * probing. */ ret = strncpy_from_kernel_nofault(__dest, (void *)addr, maxlen); if (ret >= 0) *(u32 *)dest = make_data_loc(ret, __dest - base); return ret; } static nokprobe_inline int probe_mem_read_user(void *dest, void *src, size_t size) { const void __user *uaddr = (__force const void __user *)src; return copy_from_user_nofault(dest, uaddr, size); } static nokprobe_inline int probe_mem_read(void *dest, void *src, size_t size) { #ifdef CONFIG_ARCH_HAS_NON_OVERLAPPING_ADDRESS_SPACE if ((unsigned long)src < TASK_SIZE) return probe_mem_read_user(dest, src, size); #endif return copy_from_kernel_nofault(dest, src, size); } /* eprobe handler */ static inline void __eprobe_trace_func(struct eprobe_data *edata, void *rec) { struct eprobe_trace_entry_head *entry; struct trace_event_call *call = trace_probe_event_call(&edata->ep->tp); struct trace_event_buffer fbuffer; int dsize; if (WARN_ON_ONCE(call != edata->file->event_call)) return; if (trace_trigger_soft_disabled(edata->file)) return; dsize = get_eprobe_size(&edata->ep->tp, rec); entry = trace_event_buffer_reserve(&fbuffer, edata->file, sizeof(*entry) + edata->ep->tp.size + dsize); if (!entry) return; entry = fbuffer.entry = ring_buffer_event_data(fbuffer.event); if (edata->ep->event) entry->type = edata->ep->event->event.type; else entry->type = 0; store_trace_args(&entry[1], &edata->ep->tp, rec, sizeof(*entry), dsize); trace_event_buffer_commit(&fbuffer); } /* * The event probe implementation uses event triggers to get access to * the event it is attached to, but is not an actual trigger. The below * functions are just stubs to fulfill what is needed to use the trigger * infrastructure. */ static int eprobe_trigger_init(struct event_trigger_ops *ops, struct event_trigger_data *data) { return 0; } static void eprobe_trigger_free(struct event_trigger_ops *ops, struct event_trigger_data *data) { } static int eprobe_trigger_print(struct seq_file *m, struct event_trigger_ops *ops, struct event_trigger_data *data) { /* Do not print eprobe event triggers */ return 0; } static void eprobe_trigger_func(struct event_trigger_data *data, struct trace_buffer *buffer, void *rec, struct ring_buffer_event *rbe) { struct eprobe_data *edata = data->private_data; __eprobe_trace_func(edata, rec); } static struct event_trigger_ops eprobe_trigger_ops = { .func = eprobe_trigger_func, .print = eprobe_trigger_print, .init = eprobe_trigger_init, .free = eprobe_trigger_free, }; static int eprobe_trigger_cmd_func(struct event_command *cmd_ops, struct trace_event_file *file, char *glob, char *cmd, char *param) { return -1; } static int eprobe_trigger_reg_func(char *glob, struct event_trigger_ops *ops, struct event_trigger_data *data, struct trace_event_file *file) { return -1; } static void eprobe_trigger_unreg_func(char *glob, struct event_trigger_ops *ops, struct event_trigger_data *data, struct trace_event_file *file) { } static struct event_trigger_ops *eprobe_trigger_get_ops(char *cmd, char *param) { return &eprobe_trigger_ops; } static struct event_command event_trigger_cmd = { .name = "eprobe", .trigger_type = ETT_EVENT_EPROBE, .flags = EVENT_CMD_FL_NEEDS_REC, .func = eprobe_trigger_cmd_func, .reg = eprobe_trigger_reg_func, .unreg = eprobe_trigger_unreg_func, .unreg_all = NULL, .get_trigger_ops = eprobe_trigger_get_ops, .set_filter = NULL, }; static struct event_trigger_data * new_eprobe_trigger(struct trace_eprobe *ep, struct trace_event_file *file) { struct event_trigger_data *trigger; struct eprobe_data *edata; edata = kzalloc(sizeof(*edata), GFP_KERNEL); trigger = kzalloc(sizeof(*trigger), GFP_KERNEL); if (!trigger || !edata) { kfree(edata); kfree(trigger); return ERR_PTR(-ENOMEM); } trigger->flags = EVENT_TRIGGER_FL_PROBE; trigger->count = -1; trigger->ops = &eprobe_trigger_ops; /* * EVENT PROBE triggers are not registered as commands with * register_event_command(), as they are not controlled by the user * from the trigger file */ trigger->cmd_ops = &event_trigger_cmd; INIT_LIST_HEAD(&trigger->list); RCU_INIT_POINTER(trigger->filter, NULL); edata->file = file; edata->ep = ep; trigger->private_data = edata; return trigger; } static int enable_eprobe(struct trace_eprobe *ep, struct trace_event_file *eprobe_file) { struct event_trigger_data *trigger; struct trace_event_file *file; struct trace_array *tr = eprobe_file->tr; file = find_event_file(tr, ep->event_system, ep->event_name); if (!file) return -ENOENT; trigger = new_eprobe_trigger(ep, eprobe_file); if (IS_ERR(trigger)) return PTR_ERR(trigger); list_add_tail_rcu(&trigger->list, &file->triggers); trace_event_trigger_enable_disable(file, 1); update_cond_flag(file); return 0; } static struct trace_event_functions eprobe_funcs = { .trace = print_eprobe_event }; static int disable_eprobe(struct trace_eprobe *ep, struct trace_array *tr) { struct event_trigger_data *trigger; struct trace_event_file *file; struct eprobe_data *edata; file = find_event_file(tr, ep->event_system, ep->event_name); if (!file) return -ENOENT; list_for_each_entry(trigger, &file->triggers, list) { if (!(trigger->flags & EVENT_TRIGGER_FL_PROBE)) continue; edata = trigger->private_data; if (edata->ep == ep) break; } if (list_entry_is_head(trigger, &file->triggers, list)) return -ENODEV; list_del_rcu(&trigger->list); trace_event_trigger_enable_disable(file, 0); update_cond_flag(file); /* Make sure nothing is using the edata or trigger */ tracepoint_synchronize_unregister(); kfree(edata); kfree(trigger); return 0; } static int enable_trace_eprobe(struct trace_event_call *call, struct trace_event_file *file) { struct trace_probe *pos, *tp; struct trace_eprobe *ep; bool enabled; int ret = 0; tp = trace_probe_primary_from_call(call); if (WARN_ON_ONCE(!tp)) return -ENODEV; enabled = trace_probe_is_enabled(tp); /* This also changes "enabled" state */ if (file) { ret = trace_probe_add_file(tp, file); if (ret) return ret; } else trace_probe_set_flag(tp, TP_FLAG_PROFILE); if (enabled) return 0; list_for_each_entry(pos, trace_probe_probe_list(tp), list) { ep = container_of(pos, struct trace_eprobe, tp); ret = enable_eprobe(ep, file); if (ret) break; enabled = true; } if (ret) { /* Failed to enable one of them. Roll back all */ if (enabled) disable_eprobe(ep, file->tr); if (file) trace_probe_remove_file(tp, file); else trace_probe_clear_flag(tp, TP_FLAG_PROFILE); } return ret; } static int disable_trace_eprobe(struct trace_event_call *call, struct trace_event_file *file) { struct trace_probe *pos, *tp; struct trace_eprobe *ep; tp = trace_probe_primary_from_call(call); if (WARN_ON_ONCE(!tp)) return -ENODEV; if (file) { if (!trace_probe_get_file_link(tp, file)) return -ENOENT; if (!trace_probe_has_single_file(tp)) goto out; trace_probe_clear_flag(tp, TP_FLAG_TRACE); } else trace_probe_clear_flag(tp, TP_FLAG_PROFILE); if (!trace_probe_is_enabled(tp)) { list_for_each_entry(pos, trace_probe_probe_list(tp), list) { ep = container_of(pos, struct trace_eprobe, tp); disable_eprobe(ep, file->tr); } } out: if (file) /* * Synchronization is done in below function. For perf event, * file == NULL and perf_trace_event_unreg() calls * tracepoint_synchronize_unregister() to ensure synchronize * event. We don't need to care about it. */ trace_probe_remove_file(tp, file); return 0; } static int eprobe_register(struct trace_event_call *event, enum trace_reg type, void *data) { struct trace_event_file *file = data; switch (type) { case TRACE_REG_REGISTER: return enable_trace_eprobe(event, file); case TRACE_REG_UNREGISTER: return disable_trace_eprobe(event, file); #ifdef CONFIG_PERF_EVENTS case TRACE_REG_PERF_REGISTER: case TRACE_REG_PERF_UNREGISTER: case TRACE_REG_PERF_OPEN: case TRACE_REG_PERF_CLOSE: case TRACE_REG_PERF_ADD: case TRACE_REG_PERF_DEL: return 0; #endif } return 0; } static inline void init_trace_eprobe_call(struct trace_eprobe *ep) { struct trace_event_call *call = trace_probe_event_call(&ep->tp); call->flags = TRACE_EVENT_FL_EPROBE; call->event.funcs = &eprobe_funcs; call->class->fields_array = eprobe_fields_array; call->class->reg = eprobe_register; } static struct trace_event_call * find_and_get_event(const char *system, const char *event_name) { struct trace_event_call *tp_event; const char *name; list_for_each_entry(tp_event, &ftrace_events, list) { /* Skip other probes and ftrace events */ if (tp_event->flags & (TRACE_EVENT_FL_IGNORE_ENABLE | TRACE_EVENT_FL_KPROBE | TRACE_EVENT_FL_UPROBE | TRACE_EVENT_FL_EPROBE)) continue; if (!tp_event->class->system || strcmp(system, tp_event->class->system)) continue; name = trace_event_name(tp_event); if (!name || strcmp(event_name, name)) continue; if (!trace_event_try_get_ref(tp_event)) { return NULL; break; } return tp_event; break; } return NULL; } static int trace_eprobe_tp_update_arg(struct trace_eprobe *ep, const char *argv[], int i) { unsigned int flags = TPARG_FL_KERNEL | TPARG_FL_TPOINT; int ret; ret = traceprobe_parse_probe_arg(&ep->tp, i, argv[i], flags); if (ret) return ret; if (ep->tp.args[i].code->op == FETCH_OP_TP_ARG) ret = trace_eprobe_tp_arg_update(ep, i); return ret; } static int __trace_eprobe_create(int argc, const char *argv[]) { /* * Argument syntax: * e[:[GRP/]ENAME] SYSTEM.EVENT [FETCHARGS] * Fetch args: * =$[:TYPE] */ const char *event = NULL, *group = EPROBE_EVENT_SYSTEM; const char *sys_event = NULL, *sys_name = NULL; struct trace_event_call *event_call; struct trace_eprobe *ep = NULL; char buf1[MAX_EVENT_NAME_LEN]; char buf2[MAX_EVENT_NAME_LEN]; int ret = 0; int i; if (argc < 2 || argv[0][0] != 'e') return -ECANCELED; trace_probe_log_init("event_probe", argc, argv); event = strchr(&argv[0][1], ':'); if (event) { event++; ret = traceprobe_parse_event_name(&event, &group, buf1, event - argv[0]); if (ret) goto parse_error; } else { strscpy(buf1, argv[1], MAX_EVENT_NAME_LEN); sanitize_event_name(buf1); event = buf1; } if (!is_good_name(event) || !is_good_name(group)) goto parse_error; sys_event = argv[1]; ret = traceprobe_parse_event_name(&sys_event, &sys_name, buf2, sys_event - argv[1]); if (ret || !sys_name) goto parse_error; if (!is_good_name(sys_event) || !is_good_name(sys_name)) goto parse_error; mutex_lock(&event_mutex); event_call = find_and_get_event(sys_name, sys_event); ep = alloc_event_probe(group, event, event_call, argc - 2); mutex_unlock(&event_mutex); if (IS_ERR(ep)) { ret = PTR_ERR(ep); /* This must return -ENOMEM or missing event, else there is a bug */ WARN_ON_ONCE(ret != -ENOMEM && ret != -ENODEV); ep = NULL; goto error; } argc -= 2; argv += 2; /* parse arguments */ for (i = 0; i < argc && i < MAX_TRACE_ARGS; i++) { trace_probe_log_set_index(i + 2); ret = trace_eprobe_tp_update_arg(ep, argv, i); if (ret) goto error; } ret = traceprobe_set_print_fmt(&ep->tp, PROBE_PRINT_EVENT); if (ret < 0) goto error; init_trace_eprobe_call(ep); mutex_lock(&event_mutex); ret = trace_probe_register_event_call(&ep->tp); if (ret) { if (ret == -EEXIST) { trace_probe_log_set_index(0); trace_probe_log_err(0, EVENT_EXIST); } mutex_unlock(&event_mutex); goto error; } ret = dyn_event_add(&ep->devent, &ep->tp.event->call); mutex_unlock(&event_mutex); return ret; parse_error: ret = -EINVAL; error: trace_event_probe_cleanup(ep); return ret; } /* * Register dynevent at core_initcall. This allows kernel to setup eprobe * events in postcore_initcall without tracefs. */ static __init int trace_events_eprobe_init_early(void) { int err = 0; err = dyn_event_register(&eprobe_dyn_event_ops); if (err) pr_warn("Could not register eprobe_dyn_event_ops\n"); return err; } core_initcall(trace_events_eprobe_init_early);