Received: by 2002:ad5:474a:0:0:0:0:0 with SMTP id i10csp312568imu; Mon, 5 Nov 2018 01:03:15 -0800 (PST) X-Google-Smtp-Source: AJdET5exNUHEWsKAv00gw0KcFHT+k384hNJMJyHV2K8Lo0ettJBtgu6HK8C49J8IO4LBGQDXs67J X-Received: by 2002:a17:902:622:: with SMTP id 31-v6mr21295676plg.310.1541408595548; Mon, 05 Nov 2018 01:03:15 -0800 (PST) ARC-Seal: i=1; a=rsa-sha256; t=1541408595; cv=none; d=google.com; s=arc-20160816; b=Zl7710+vLMYYt3isd3nXT89L5U5k0yIL6wzX+Zet16/m9U9GlCktDMCanTRH9eUnhS xuZfSj7/q3EQubV06RdzY4BB3H4WYIYyWiyyvpquuvkDEwLOx0guTtLd7Q6/p2/XpUlE St4lhIrUiKomnUp6zxERUkQ+c/rlKto2HPrKrV7gILpeHvs08TSf2xgHLfU6L5XdQQLN r6KUWykQzKszOihJYUMsIPrEZBK1XmR4rdUCTAcjk1+B5iM9SYvcb7J/2aCyUDBTnvWF 2Rej5ZAPxebHgbF7XGKrZWrLmJvieJHF8B0hBaXOSUVLPQ4FxDFlDE+iKhd11gZUvXEl iMOQ== ARC-Message-Signature: i=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=arc-20160816; h=list-id:precedence:sender:content-transfer-encoding:mime-version :user-agent:references:in-reply-to:message-id:date:subject:cc:to :from:dkim-signature; bh=GlWRgzKJnGddHIUDo6TQcCEqxqzpw6e6pb7SluZnmEI=; b=RTd5k+QoRnxGTSdCYipeCZhEm1pnKVTqXhAEYeMj7q3Bs62Zj7yBmBoo1OIAVJ0NPq hw3/cnQAMPk/S16kJvyluBJeI20mJLP7ICmDZ1Kbd288BvWAbyLWE/yKo3doQoAD5cob Ok/zo2x+qAHQJGH4g//Vuv2XVwEH5AJ3ZgjI9c85vwUOgDO3U/y5O5pO7N6kTcbLpKJN dEI3g8RYEVxcfZmUzzLQbXkji37C188j3NLxhgVvGxCm8bMlyfYJ90fvsbek2fhQIqKO xYL5m4Ei+p2TioyfWeblZPSaIfcj7pUAARtkssF0tVxigns9ieh2TMd0fkULtYgZ+P2i qHYw== ARC-Authentication-Results: i=1; mx.google.com; dkim=pass header.i=@kernel.org header.s=default header.b="S4Du6hQ/"; spf=pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) smtp.mailfrom=linux-kernel-owner@vger.kernel.org; dmarc=pass (p=NONE sp=NONE dis=NONE) header.from=kernel.org Return-Path: Received: from vger.kernel.org (vger.kernel.org. [209.132.180.67]) by mx.google.com with ESMTP id 1-v6si33314135pgo.354.2018.11.05.01.02.59; Mon, 05 Nov 2018 01:03:15 -0800 (PST) Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67; Authentication-Results: mx.google.com; dkim=pass header.i=@kernel.org header.s=default header.b="S4Du6hQ/"; spf=pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) smtp.mailfrom=linux-kernel-owner@vger.kernel.org; dmarc=pass (p=NONE sp=NONE dis=NONE) header.from=kernel.org Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1728890AbeKESVM (ORCPT + 99 others); Mon, 5 Nov 2018 13:21:12 -0500 Received: from mail.kernel.org ([198.145.29.99]:50466 "EHLO mail.kernel.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1726186AbeKESVM (ORCPT ); Mon, 5 Nov 2018 13:21:12 -0500 Received: from localhost.localdomain (NE2965lan1.rev.em-net.ne.jp [210.141.244.193]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by mail.kernel.org (Postfix) with ESMTPSA id 3480D20862; Mon, 5 Nov 2018 09:02:30 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=kernel.org; s=default; t=1541408551; bh=T2CI+eIVzdrZvRAYmKmaisUpWOh1IBdBkzRmnSjgzVU=; h=From:To:Cc:Subject:Date:In-Reply-To:References:From; b=S4Du6hQ/2Os/kmwPOKiAApkVX78fI/xkaHCaBbcqkV5F8YGtog6uop+OtRKrZ4EBT vsge0l2rImL1+UTNyucpqHYfb/JhNFINcrt1rdyFaniBux5jJyEHTSzc03DQMKqqzV KvQC7PC2Y0KyHI9S/pso4eEz8uinnN/7CK1hBDMU= From: Masami Hiramatsu To: Steven Rostedt Cc: linux-kernel@vger.kernel.org, Ingo Molnar , Tom Zanussi , Ravi Bangoria Subject: [PATCH v2 05/12] tracing: Add unified dynamic event framework Date: Mon, 5 Nov 2018 18:02:08 +0900 Message-Id: <154140852824.17322.12250362185969352095.stgit@devbox> X-Mailer: git-send-email 2.13.6 In-Reply-To: <154140838606.17322.15294184388075458777.stgit@devbox> References: <154140838606.17322.15294184388075458777.stgit@devbox> User-Agent: StGit/0.17.1-dirty MIME-Version: 1.0 Content-Type: text/plain; charset="utf-8" Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Add unified dynamic event framework for ftrace kprobes, uprobes and synthetic events. Those dynamic events can be co-exist on same file because those syntax doesn't over-wrapped. This introduces a framework part which provides a unified tracefs interface and operations. Signed-off-by: Masami Hiramatsu --- Changes in v2: - Fix to lock dyn_event_mutex when freeing all events. - Generalize and export dyn_event_release_all(). - To fix dependency issue, lock event_mutex before dyn_event_mutex. - Separate create and delete operation. - If create operation returns -ECANCELED, try next operation. - Pass const argv to create operation for safety. - Add .match operation and generic delete API. - Add precise description for data structures. - Remove dyn_event_mutex, use event_mutex instead. --- kernel/trace/Kconfig | 3 + kernel/trace/Makefile | 1 kernel/trace/trace.c | 4 + kernel/trace/trace_dynevent.c | 210 +++++++++++++++++++++++++++++++++++++++++ kernel/trace/trace_dynevent.h | 119 +++++++++++++++++++++++ 5 files changed, 337 insertions(+) create mode 100644 kernel/trace/trace_dynevent.c create mode 100644 kernel/trace/trace_dynevent.h diff --git a/kernel/trace/Kconfig b/kernel/trace/Kconfig index 5e3de28c7677..bf2e8a5a91f1 100644 --- a/kernel/trace/Kconfig +++ b/kernel/trace/Kconfig @@ -518,6 +518,9 @@ config BPF_EVENTS help This allows the user to attach BPF programs to kprobe events. +config DYNAMIC_EVENTS + def_bool n + config PROBE_EVENTS def_bool n diff --git a/kernel/trace/Makefile b/kernel/trace/Makefile index f81dadbc7c4a..9ff3c4fa91b6 100644 --- a/kernel/trace/Makefile +++ b/kernel/trace/Makefile @@ -78,6 +78,7 @@ endif ifeq ($(CONFIG_TRACING),y) obj-$(CONFIG_KGDB_KDB) += trace_kdb.o endif +obj-$(CONFIG_DYNAMIC_EVENTS) += trace_dynevent.o obj-$(CONFIG_PROBE_EVENTS) += trace_probe.o obj-$(CONFIG_UPROBE_EVENTS) += trace_uprobe.o diff --git a/kernel/trace/trace.c b/kernel/trace/trace.c index ff1c4b20cd0a..2886e92e8eab 100644 --- a/kernel/trace/trace.c +++ b/kernel/trace/trace.c @@ -4604,6 +4604,10 @@ static const char readme_msg[] = "\t\t\t traces\n" #endif #endif /* CONFIG_STACK_TRACER */ +#ifdef CONFIG_DYNAMIC_EVENTS + " dynamic_events\t\t- Add/remove/show the generic dynamic events\n" + "\t\t\t Write into this file to define/undefine new trace events.\n" +#endif #ifdef CONFIG_KPROBE_EVENTS " kprobe_events\t\t- Add/remove/show the kernel dynamic events\n" "\t\t\t Write into this file to define/undefine new trace events.\n" diff --git a/kernel/trace/trace_dynevent.c b/kernel/trace/trace_dynevent.c new file mode 100644 index 000000000000..f17a887abb66 --- /dev/null +++ b/kernel/trace/trace_dynevent.c @@ -0,0 +1,210 @@ +// SPDX-License-Identifier: GPL-2.0 +/* + * Generic dynamic event control interface + * + * Copyright (C) 2018 Masami Hiramatsu + */ + +#include +#include +#include +#include +#include +#include + +#include "trace.h" +#include "trace_dynevent.h" + +static DEFINE_MUTEX(dyn_event_ops_mutex); +static LIST_HEAD(dyn_event_ops_list); + +int dyn_event_register(struct dyn_event_operations *ops) +{ + if (!ops || !ops->create || !ops->show || !ops->is_busy || + !ops->free || !ops->match) + return -EINVAL; + + INIT_LIST_HEAD(&ops->list); + mutex_lock(&dyn_event_ops_mutex); + list_add_tail(&ops->list, &dyn_event_ops_list); + mutex_unlock(&dyn_event_ops_mutex); + return 0; +} + +int dyn_event_release(int argc, char **argv, struct dyn_event_operations *type) +{ + struct dyn_event *pos, *n; + char *system = NULL, *event, *p; + int ret = -ENOENT; + + if (argv[0][1] != ':') + return -EINVAL; + + event = &argv[0][2]; + p = strchr(event, '/'); + if (p) { + system = event; + event = p + 1; + *p = '\0'; + } + if (event[0] == '\0') + return -EINVAL; + + mutex_lock(&event_mutex); + for_each_dyn_event_safe(pos, n) { + if (type && type != pos->ops) + continue; + if (pos->ops->match(system, event, pos)) { + ret = pos->ops->free(pos); + break; + } + } + mutex_unlock(&event_mutex); + + return ret; +} + +static int create_dyn_event(int argc, char **argv) +{ + struct dyn_event_operations *ops; + int ret; + + if (argv[0][0] == '-') + return dyn_event_release(argc, argv, NULL); + + mutex_lock(&dyn_event_ops_mutex); + list_for_each_entry(ops, &dyn_event_ops_list, list) { + ret = ops->create(argc, (const char **)argv); + if (!ret || ret != -ECANCELED) + break; + } + mutex_unlock(&dyn_event_ops_mutex); + if (ret == -ECANCELED) + ret = -EINVAL; + + return ret; +} + +/* Protected by event_mutex */ +LIST_HEAD(dyn_event_list); + +void *dyn_event_seq_start(struct seq_file *m, loff_t *pos) +{ + mutex_lock(&event_mutex); + return seq_list_start(&dyn_event_list, *pos); +} + +void *dyn_event_seq_next(struct seq_file *m, void *v, loff_t *pos) +{ + return seq_list_next(v, &dyn_event_list, pos); +} + +void dyn_event_seq_stop(struct seq_file *m, void *v) +{ + mutex_unlock(&event_mutex); +} + +static int dyn_event_seq_show(struct seq_file *m, void *v) +{ + struct dyn_event *ev = v; + + if (ev && ev->ops) + return ev->ops->show(m, ev); + + return 0; +} + +static const struct seq_operations dyn_event_seq_op = { + .start = dyn_event_seq_start, + .next = dyn_event_seq_next, + .stop = dyn_event_seq_stop, + .show = dyn_event_seq_show +}; + +/* + * dyn_events_release_all - Release all specific events + * @type: the dyn_event_operations * which filters releasing events + * + * This releases all events which ->ops matches @type. If @type is NULL, + * all events are released. + * Return -EBUSY if any of them are in use, and return other errors when + * it failed to free the given event. Except for -EBUSY, event releasing + * process will be aborted at that point and there may be some other + * releasable events on the list. + */ +int dyn_events_release_all(struct dyn_event_operations *type) +{ + struct dyn_event *ev, *tmp; + int ret = 0; + + mutex_lock(&event_mutex); + for_each_dyn_event(ev) { + if (type && ev->ops != type) + continue; + if (ev->ops->is_busy(ev)) { + ret = -EBUSY; + goto out; + } + } + for_each_dyn_event_safe(ev, tmp) { + if (type && ev->ops != type) + continue; + ret = ev->ops->free(ev); + if (ret) + break; + } +out: + mutex_unlock(&event_mutex); + + return ret; +} + +static int dyn_event_open(struct inode *inode, struct file *file) +{ + int ret; + + if ((file->f_mode & FMODE_WRITE) && (file->f_flags & O_TRUNC)) { + ret = dyn_events_release_all(NULL); + if (ret < 0) + return ret; + } + + return seq_open(file, &dyn_event_seq_op); +} + +static ssize_t dyn_event_write(struct file *file, const char __user *buffer, + size_t count, loff_t *ppos) +{ + return trace_parse_run_command(file, buffer, count, ppos, + create_dyn_event); +} + +static const struct file_operations dynamic_events_ops = { + .owner = THIS_MODULE, + .open = dyn_event_open, + .read = seq_read, + .llseek = seq_lseek, + .release = seq_release, + .write = dyn_event_write, +}; + +/* Make a tracefs interface for controlling dynamic events */ +static __init int init_dynamic_event(void) +{ + struct dentry *d_tracer; + struct dentry *entry; + + d_tracer = tracing_init_dentry(); + if (IS_ERR(d_tracer)) + return 0; + + entry = tracefs_create_file("dynamic_events", 0644, d_tracer, + NULL, &dynamic_events_ops); + + /* Event list interface */ + if (!entry) + pr_warn("Could not create tracefs 'dynamic_events' entry\n"); + + return 0; +} +fs_initcall(init_dynamic_event); diff --git a/kernel/trace/trace_dynevent.h b/kernel/trace/trace_dynevent.h new file mode 100644 index 000000000000..8c334064e4d6 --- /dev/null +++ b/kernel/trace/trace_dynevent.h @@ -0,0 +1,119 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +/* + * Common header file for generic dynamic events. + */ + +#ifndef _TRACE_DYNEVENT_H +#define _TRACE_DYNEVENT_H + +#include +#include +#include +#include + +#include "trace.h" + +struct dyn_event; + +/** + * struct dyn_event_operations - Methods for each type of dynamic events + * + * These methods must be set for each type, since there is no default method. + * Before using this for dyn_event_init(), it must be registered by + * dyn_event_register(). + * + * @create: Parse and create event method. This is invoked when user passes + * a event definition to dynamic_events interface. This must not destruct + * the arguments and return -ECANCELED if given arguments doesn't match its + * command prefix. + * @show: Showing method. This is invoked when user reads the event definitions + * via dynamic_events interface. + * @is_busy: Check whether given event is busy so that it can not be deleted. + * Return true if it is busy, otherwides false. + * @free: Delete the given event. Return 0 if success, otherwides error. + * @match: Check whether given event and system name match this event. + * Return true if it matches, otherwides false. + * + * Except for @create, these methods are called under holding event_mutex. + */ +struct dyn_event_operations { + struct list_head list; + int (*create)(int argc, const char *argv[]); + int (*show)(struct seq_file *m, struct dyn_event *ev); + bool (*is_busy)(struct dyn_event *ev); + int (*free)(struct dyn_event *ev); + bool (*match)(const char *system, const char *event, + struct dyn_event *ev); +}; + +/* Register new dyn_event type -- must be called at first */ +int dyn_event_register(struct dyn_event_operations *ops); + +/** + * struct dyn_event - Dynamic event list header + * + * The dyn_event structure encapsulates a list and a pointer to the operators + * for making a global list of dynamic events. + * User must includes this in each event structure, so that those events can + * be added/removed via dynamic_events interface. + */ +struct dyn_event { + struct list_head list; + struct dyn_event_operations *ops; +}; + +extern struct list_head dyn_event_list; + +static inline +int dyn_event_init(struct dyn_event *ev, struct dyn_event_operations *ops) +{ + if (!ev || !ops) + return -EINVAL; + + INIT_LIST_HEAD(&ev->list); + ev->ops = ops; + return 0; +} + +static inline int dyn_event_add(struct dyn_event *ev) +{ + lockdep_assert_held(&event_mutex); + + if (!ev || !ev->ops) + return -EINVAL; + + list_add_tail(&ev->list, &dyn_event_list); + return 0; +} + +static inline void dyn_event_remove(struct dyn_event *ev) +{ + lockdep_assert_held(&event_mutex); + list_del_init(&ev->list); +} + +void *dyn_event_seq_start(struct seq_file *m, loff_t *pos); +void *dyn_event_seq_next(struct seq_file *m, void *v, loff_t *pos); +void dyn_event_seq_stop(struct seq_file *m, void *v); +int dyn_events_release_all(struct dyn_event_operations *type); +int dyn_event_release(int argc, char **argv, struct dyn_event_operations *type); + +/* + * for_each_dyn_event - iterate over the dyn_event list + * @pos: the struct dyn_event * to use as a loop cursor + * + * This is just a basement of for_each macro. Wrap this for + * each actual event structure with ops filtering. + */ +#define for_each_dyn_event(pos) \ + list_for_each_entry(pos, &dyn_event_list, list) + +/* + * for_each_dyn_event - iterate over the dyn_event list safely + * @pos: the struct dyn_event * to use as a loop cursor + * @n: the struct dyn_event * to use as temporary storage + */ +#define for_each_dyn_event_safe(pos, n) \ + list_for_each_entry_safe(pos, n, &dyn_event_list, list) + +#endif