Commit e2d0d7b2 authored by Masami Hiramatsu (Google)'s avatar Masami Hiramatsu (Google)
Browse files

tracing/probes: Add tracepoint support on fprobe_events 

Allow fprobe_events to trace raw tracepoints so that user can trace
tracepoints which don't have traceevent wrappers. This new event is
always available if the fprobe_events is enabled (thus no kconfig),
because the fprobe_events depends on the trace-event and traceporint.

e.g.
 # echo 't sched_overutilized_tp' >> dynamic_events
 # echo 't 9p_client_req' >> dynamic_events
 # cat dynamic_events
t:tracepoints/sched_overutilized_tp sched_overutilized_tp
t:tracepoints/_9p_client_req 9p_client_req

The event name is based on the tracepoint name, but if it is started
with digit character, an underscore '_' will be added.

NOTE: to avoid further confusion, this renames TPARG_FL_TPOINT to
TPARG_FL_TEVENT because this flag is used for eprobe (trace-event probe).
And reuse TPARG_FL_TPOINT for this raw tracepoint probe.

Link: https://lore.kernel.org/all/168507471874.913472.17214624519622959593.stgit@mhiramat.roam.corp.google.com/



Reported-by: default avatarkernel test robot <lkp@intel.com>
Link: https://lore.kernel.org/oe-kbuild-all/202305020453.afTJ3VVp-lkp@intel.com/


Signed-off-by: default avatarMasami Hiramatsu (Google) <mhiramat@kernel.org>
parent 4d42a765

include/linux/tracepoint-defs.h

+1 −0
Original line number Diff line number Diff line
@@ -35,6 +35,7 @@ struct tracepoint { 
	struct static_call_key *static_call_key;

	void *static_call_tramp;

	void *iterator;

	void *probestub;

	int (*regfunc)(void);

	void (*unregfunc)(void);

	struct tracepoint_func __rcu *funcs;

include/linux/tracepoint.h

+5 −0
Original line number Diff line number Diff line
@@ -303,6 +303,7 @@ static inline struct tracepoint *tracepoint_ptr_deref(tracepoint_ptr_t *p) 
	__section("__tracepoints_strings") = #_name;			\

	extern struct static_call_key STATIC_CALL_KEY(tp_func_##_name);	\

	int __traceiter_##_name(void *__data, proto);			\

	void __probestub_##_name(void *__data, proto);			\

	struct tracepoint __tracepoint_##_name	__used			\

	__section("__tracepoints") = {					\

		.name = __tpstrtab_##_name,				\
@@ -310,6 +311,7 @@ static inline struct tracepoint *tracepoint_ptr_deref(tracepoint_ptr_t *p) 
		.static_call_key = &STATIC_CALL_KEY(tp_func_##_name),	\

		.static_call_tramp = STATIC_CALL_TRAMP_ADDR(tp_func_##_name), \

		.iterator = &__traceiter_##_name,			\

		.probestub = &__probestub_##_name,			\

		.regfunc = _reg,					\

		.unregfunc = _unreg,					\

		.funcs = NULL };					\
@@ -330,6 +332,9 @@ static inline struct tracepoint *tracepoint_ptr_deref(tracepoint_ptr_t *p) 
		}							\

		return 0;						\

	}								\

	void __probestub_##_name(void *__data, proto)			\

	{								\

	}								\

	DEFINE_STATIC_CALL(tp_func_##_name, __traceiter_##_name);



#define DEFINE_TRACE(name, proto, args)		\

kernel/trace/trace.c

+1 −0
Original line number Diff line number Diff line
@@ -5681,6 +5681,7 @@ static const char readme_msg[] = 
#endif

#ifdef CONFIG_FPROBE_EVENTS

	"\t           f[:[<group>/][<event>]] <func-name>[%return] [<args>]\n"

	"\t           t[:[<group>/][<event>]] <tracepoint> [<args>]\n"

#endif

#ifdef CONFIG_HIST_TRIGGERS

	"\t           s:[synthetic/]<event> <field> [<field>]\n"

kernel/trace/trace_eprobe.c

+1 −1
Original line number Diff line number Diff line
@@ -817,7 +817,7 @@ find_and_get_event(const char *system, const char *event_name) 


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;

	unsigned int flags = TPARG_FL_KERNEL | TPARG_FL_TEVENT;

	int ret;



	ret = traceprobe_parse_probe_arg(&ep->tp, i, argv[i], flags);

kernel/trace/trace_fprobe.c

+127 −7
Original line number Diff line number Diff line
@@ -9,6 +9,7 @@ 
#include <linux/module.h>

#include <linux/rculist.h>

#include <linux/security.h>

#include <linux/tracepoint.h>

#include <linux/uaccess.h>



#include "trace_dynevent.h"
@@ -17,6 +18,7 @@ 
#include "trace_probe_tmpl.h"



#define FPROBE_EVENT_SYSTEM "fprobes"

#define TRACEPOINT_EVENT_SYSTEM "tracepoints"

#define RETHOOK_MAXACTIVE_MAX 4096



static int trace_fprobe_create(const char *raw_command);
@@ -41,6 +43,8 @@ struct trace_fprobe { 
	struct dyn_event	devent;

	struct fprobe		fp;

	const char		*symbol;

	struct tracepoint	*tpoint;

	struct module		*mod;

	struct trace_probe	tp;

};
@@ -68,6 +72,11 @@ static bool trace_fprobe_is_return(struct trace_fprobe *tf) 
	return tf->fp.exit_handler != NULL;

}



static bool trace_fprobe_is_tracepoint(struct trace_fprobe *tf)

{

	return tf->tpoint != NULL;

}



static const char *trace_fprobe_symbol(struct trace_fprobe *tf)

{

	return tf->symbol ? tf->symbol : "unknown";
@@ -668,6 +677,21 @@ static int __register_trace_fprobe(struct trace_fprobe *tf) 
	else

		tf->fp.flags |= FPROBE_FL_DISABLED;



	if (trace_fprobe_is_tracepoint(tf)) {

		struct tracepoint *tpoint = tf->tpoint;

		unsigned long ip = (unsigned long)tpoint->probestub;

		/*

		 * Here, we do 2 steps to enable fprobe on a tracepoint.

		 * At first, put __probestub_##TP function on the tracepoint

		 * and put a fprobe on the stub function.

		 */

		ret = tracepoint_probe_register_prio_may_exist(tpoint,

					tpoint->probestub, NULL, 0);

		if (ret < 0)

			return ret;

		return register_fprobe_ips(&tf->fp, &ip, 1);

	}



	/* TODO: handle filter, nofilter or symbol list */

	return register_fprobe(&tf->fp, tf->symbol, NULL);

}
@@ -678,6 +702,12 @@ static void __unregister_trace_fprobe(struct trace_fprobe *tf) 
	if (trace_fprobe_is_registered(tf)) {

		unregister_fprobe(&tf->fp);

		memset(&tf->fp, 0, sizeof(tf->fp));

		if (trace_fprobe_is_tracepoint(tf)) {

			tracepoint_probe_unregister(tf->tpoint,

					tf->tpoint->probestub, NULL);

			tf->tpoint = NULL;

			tf->mod = NULL;

		}

	}

}
@@ -741,7 +771,8 @@ static int append_trace_fprobe(struct trace_fprobe *tf, struct trace_fprobe *to) 
{

	int ret;



	if (trace_fprobe_is_return(tf) != trace_fprobe_is_return(to)) {

	if (trace_fprobe_is_return(tf) != trace_fprobe_is_return(to) ||

	    trace_fprobe_is_tracepoint(tf) != trace_fprobe_is_tracepoint(to)) {

		trace_probe_log_set_index(0);

		trace_probe_log_err(0, DIFF_PROBE_TYPE);

		return -EEXIST;
@@ -811,6 +842,60 @@ static int register_trace_fprobe(struct trace_fprobe *tf) 
	return ret;

}



#ifdef CONFIG_MODULES

static int __tracepoint_probe_module_cb(struct notifier_block *self,

					unsigned long val, void *data)

{

	struct tp_module *tp_mod = data;

	struct trace_fprobe *tf;

	struct dyn_event *pos;



	if (val != MODULE_STATE_GOING)

		return NOTIFY_DONE;



	mutex_lock(&event_mutex);

	for_each_trace_fprobe(tf, pos) {

		if (tp_mod->mod == tf->mod) {

			tracepoint_probe_unregister(tf->tpoint,

					tf->tpoint->probestub, NULL);

			tf->tpoint = NULL;

			tf->mod = NULL;

		}

	}

	mutex_unlock(&event_mutex);



	return NOTIFY_DONE;

}



static struct notifier_block tracepoint_module_nb = {

	.notifier_call = __tracepoint_probe_module_cb,

};

#endif /* CONFIG_MODULES */



struct __find_tracepoint_cb_data {

	const char *tp_name;

	struct tracepoint *tpoint;

};



static void __find_tracepoint_cb(struct tracepoint *tp, void *priv)

{

	struct __find_tracepoint_cb_data *data = priv;



	if (!data->tpoint && !strcmp(data->tp_name, tp->name))

		data->tpoint = tp;

}



static struct tracepoint *find_tracepoint(const char *tp_name)

{

	struct __find_tracepoint_cb_data data = {

		.tp_name = tp_name,

	};



	for_each_kernel_tracepoint(__find_tracepoint_cb, &data);



	return data.tpoint;

}



static int __trace_fprobe_create(int argc, const char *argv[])

{

	/*
@@ -819,6 +904,8 @@ static int __trace_fprobe_create(int argc, const char *argv[]) 
	 *      f[:[GRP/][EVENT]] [MOD:]KSYM [FETCHARGS]

	 *  - Add fexit probe:

	 *      f[N][:[GRP/][EVENT]] [MOD:]KSYM%return [FETCHARGS]

	 *  - Add tracepoint probe:

	 *      t[:[GRP/][EVENT]] TRACEPOINT [FETCHARGS]

	 *

	 * Fetch args:

	 *  $retval	: fetch return value
@@ -844,10 +931,16 @@ static int __trace_fprobe_create(int argc, const char *argv[]) 
	char buf[MAX_EVENT_NAME_LEN];

	char gbuf[MAX_EVENT_NAME_LEN];

	unsigned int flags = TPARG_FL_KERNEL | TPARG_FL_FPROBE;

	bool is_tracepoint = false;



	if (argv[0][0] != 'f' || argc < 2)

	if ((argv[0][0] != 'f' && argv[0][0] != 't') || argc < 2)

		return -ECANCELED;



	if (argv[0][0] == 't') {

		is_tracepoint = true;

		group = TRACEPOINT_EVENT_SYSTEM;

	}



	trace_probe_log_init("trace_fprobe", argc, argv);



	event = strchr(&argv[0][1], ':');
@@ -881,14 +974,14 @@ static int __trace_fprobe_create(int argc, const char *argv[]) 


	trace_probe_log_set_index(1);



	/* a symbol specified */

	/* a symbol(or tracepoint) must be specified */

	symbol = kstrdup(argv[1], GFP_KERNEL);

	if (!symbol)

		return -ENOMEM;



	tmp = strchr(symbol, '%');

	if (tmp) {

		if (!strcmp(tmp, "%return")) {

		if (!is_tracepoint && !strcmp(tmp, "%return")) {

			*tmp = '\0';

			is_return = true;

		} else {
@@ -907,6 +1000,9 @@ static int __trace_fprobe_create(int argc, const char *argv[]) 
	else

		flags |= TPARG_FL_FENTRY;



	if (is_tracepoint)

		flags |= TPARG_FL_TPOINT;



	trace_probe_log_set_index(0);

	if (event) {

		ret = traceprobe_parse_event_name(&event, &group, gbuf,
@@ -917,6 +1013,9 @@ static int __trace_fprobe_create(int argc, const char *argv[]) 


	if (!event) {

		/* Make a new event name */

		if (is_tracepoint)

			strscpy(buf, symbol, MAX_EVENT_NAME_LEN);

		else

			snprintf(buf, MAX_EVENT_NAME_LEN, "%s__%s", symbol,

				 is_return ? "exit" : "entry");

		sanitize_event_name(buf);
@@ -932,6 +1031,18 @@ static int __trace_fprobe_create(int argc, const char *argv[]) 
		WARN_ON_ONCE(ret != -ENOMEM);

		goto out;	/* We know tf is not allocated */

	}



	if (is_tracepoint) {

		tf->tpoint = find_tracepoint(tf->symbol);

		if (!tf->tpoint) {

			trace_probe_log_set_index(1);

			trace_probe_log_err(0, NO_TRACEPOINT);

			goto parse_error;

		}

		tf->mod = __module_text_address(

				(unsigned long)tf->tpoint->probestub);

	}



	argc -= 2; argv += 2;



	/* parse arguments */
@@ -991,6 +1102,9 @@ static int trace_fprobe_show(struct seq_file *m, struct dyn_event *ev) 
	struct trace_fprobe *tf = to_trace_fprobe(ev);

	int i;



	if (trace_fprobe_is_tracepoint(tf))

		seq_putc(m, 't');

	else

		seq_putc(m, 'f');

	if (trace_fprobe_is_return(tf) && tf->fp.nr_maxactive)

		seq_printf(m, "%d", tf->fp.nr_maxactive);
@@ -1048,6 +1162,12 @@ static __init int init_fprobe_trace_early(void) 
	if (ret)

		return ret;



#ifdef CONFIG_MODULES

	ret = register_tracepoint_module_notifier(&tracepoint_module_nb);

	if (ret)

		return ret;

#endif



	return 0;

}

core_initcall(init_fprobe_trace_early);