[for-linus][PATCH 00/13] tracing: Updates for 5.17-rc

[for-linus][PATCH 00/13] tracing: Updates for 5.17-rc

[Steven Rostedt]

Tracing fixes for 5.17:

 - rtla (Real-Time Linux Analysis tool): fix typo in man page

 - rtla: Update API -e to -E before it is released

 - rlla: Error message fix and memory leak fix

 - Partially unline trace event soft disable to shrink text

 - Fix function graph start up test

 - Have triggers affect the trace instance they are in and not top level

 - Have osnoise sleep in the units it says it uses

 - Remove unused ftrace stub function

 - Remove event probe reduntant info from event in the buffer

 - Fix group ownership setting in tracefs

 - Ensure trace buffer is minimum size to prevent crashes

  git://git.kernel.org/pub/scm/linux/kernel/git/rostedt/linux-trace.git
ftrace/urgent

Head SHA1: f4b8715634f04fa03e730f1493ee919b56468de8


Arnaldo Carvalho de Melo (1):
      rtla: Fix systme -> system typo on man page

Christophe Leroy (2):
      tracing: Uninline trace_trigger_soft_disabled() partly
      tracing: Fix selftest config check for function graph start up test

Daniel Bristot de Oliveira (5):
      tracing: Dump stacktrace trigger to the corresponding instance
      tracing/osnoise: Make osnoise_main to sleep for microseconds
      rtla/hist: Make -E the short version of --entries
      rtla/osnoise: Free params at the exit
      rtla/osnoise: Fix error message when failing to enable trace instance

Nathan Chancellor (1):
      ftrace: Remove unused ftrace_startup_enable() stub

Steven Rostedt (Google) (3):
      tracing: Have traceon and traceoff trigger honor the instance
      eprobes: Remove redundant event type information
      tracefs: Set the group ownership in apply_options() not parse_options()

Sven Schnelle (1):
      tracing: Ensure trace buffer is at least 4096 bytes large

----
 Documentation/tools/rtla/common_hist_options.rst   |  2 +-
 .../tools/rtla/common_osnoise_description.rst      |  2 +-
 Documentation/tools/rtla/rtla-osnoise-hist.rst     |  2 +-
 fs/tracefs/inode.c                                 |  4 +-
 include/linux/trace_events.h                       | 22 ++++---
 kernel/trace/ftrace.c                              |  1 -
 kernel/trace/trace.c                               | 10 +--
 kernel/trace/trace.h                               |  1 -
 kernel/trace/trace_eprobe.c                        | 16 +++--
 kernel/trace/trace_events_trigger.c                | 73 +++++++++++++++++++---
 kernel/trace/trace_osnoise.c                       | 53 +++++++++-------
 kernel/trace/trace_probe.c                         | 10 +--
 kernel/trace/trace_probe.h                         |  1 -
 kernel/trace/trace_selftest.c                      |  6 +-
 tools/tracing/rtla/src/osnoise.c                   |  2 +-
 tools/tracing/rtla/src/osnoise_hist.c              | 10 +--
 tools/tracing/rtla/src/osnoise_top.c               |  1 +
 tools/tracing/rtla/src/timerlat_hist.c             | 10 +--
 18 files changed, 148 insertions(+), 78 deletions(-)

[for-linus][PATCH 01/13] rtla: Fix systme -> system typo on man page

[Steven Rostedt]

From: Arnaldo Carvalho de Melo <acme@kernel.org>

Link: https://lkml.kernel.org/r/YhZsZxqk+IaFxorj@kernel.org

Fixes: 496082df01bb08a4 ("rtla: Add rtla osnoise man page")
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Acked-by: Daniel Bristot de Oliveira <bristot@kernel.org>
Signed-off-by: Steven Rostedt (Google) <rostedt@goodmis.org>
---
 Documentation/tools/rtla/common_osnoise_description.rst | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/Documentation/tools/rtla/common_osnoise_description.rst b/Documentation/tools/rtla/common_osnoise_description.rst
index 8973c5df888f..d5d61615b967 100644
--- a/Documentation/tools/rtla/common_osnoise_description.rst
+++ b/Documentation/tools/rtla/common_osnoise_description.rst
@@ -1,7 +1,7 @@
 The **rtla osnoise** tool is an interface for the *osnoise* tracer. The
 *osnoise* tracer dispatches a kernel thread per-cpu. These threads read the
 time in a loop while with preemption, softirq and IRQs enabled, thus
-allowing all the sources of operating systme noise during its execution.
+allowing all the sources of operating system noise during its execution.
 The *osnoise*'s tracer threads take note of the delta between each time
 read, along with an interference counter of all sources of interference.
 At the end of each period, the *osnoise* tracer displays a summary of
-- 
2.34.1

[for-linus][PATCH 02/13] tracing: Dump stacktrace trigger to the corresponding instance

[Steven Rostedt]

From: Daniel Bristot de Oliveira <bristot@kernel.org>

The stacktrace event trigger is not dumping the stacktrace to the instance
where it was enabled, but to the global "instance."

Use the private_data, pointing to the trigger file, to figure out the
corresponding trace instance, and use it in the trigger action, like
snapshot_trigger does.

Link: https://lkml.kernel.org/r/afbb0b4f18ba92c276865bc97204d438473f4ebc.1645396236.git.bristot@kernel.org

Cc: stable@vger.kernel.org
Fixes: ae63b31e4d0e2 ("tracing: Separate out trace events from global variables")
Reviewed-by: Tom Zanussi <zanussi@kernel.org>
Tested-by: Tom Zanussi <zanussi@kernel.org>
Signed-off-by: Daniel Bristot de Oliveira <bristot@kernel.org>
Signed-off-by: Steven Rostedt (Google) <rostedt@goodmis.org>
---
 kernel/trace/trace_events_trigger.c | 7 ++++++-
 1 file changed, 6 insertions(+), 1 deletion(-)

diff --git a/kernel/trace/trace_events_trigger.c b/kernel/trace/trace_events_trigger.c
index d00fee705f9c..e0d50c9577f3 100644
--- a/kernel/trace/trace_events_trigger.c
+++ b/kernel/trace/trace_events_trigger.c
@@ -1540,7 +1540,12 @@ stacktrace_trigger(struct event_trigger_data *data,
 		   struct trace_buffer *buffer,  void *rec,
 		   struct ring_buffer_event *event)
 {
-	trace_dump_stack(STACK_SKIP);
+	struct trace_event_file *file = data->private_data;
+
+	if (file)
+		__trace_stack(file->tr, tracing_gen_ctx(), STACK_SKIP);
+	else
+		trace_dump_stack(STACK_SKIP);
 }
 
 static void
-- 
2.34.1

[for-linus][PATCH 03/13] tracing: Have traceon and traceoff trigger honor the instance

[Steven Rostedt]

From: "Steven Rostedt (Google)" <rostedt@goodmis.org>

If a trigger is set on an event to disable or enable tracing within an
instance, then tracing should be disabled or enabled in the instance and
not at the top level, which is confusing to users.

Link: https://lkml.kernel.org/r/20220223223837.14f94ec3@rorschach.local.home

Cc: Tom Zanussi <zanussi@kernel.org>
Cc: stable@vger.kernel.org
Fixes: ae63b31e4d0e2 ("tracing: Separate out trace events from global variables")
Tested-by: Daniel Bristot de Oliveira <bristot@kernel.org>
Signed-off-by: Steven Rostedt (Google) <rostedt@goodmis.org>
---
 kernel/trace/trace_events_trigger.c | 52 +++++++++++++++++++++++++----
 1 file changed, 46 insertions(+), 6 deletions(-)

diff --git a/kernel/trace/trace_events_trigger.c b/kernel/trace/trace_events_trigger.c
index e0d50c9577f3..efe563140f27 100644
--- a/kernel/trace/trace_events_trigger.c
+++ b/kernel/trace/trace_events_trigger.c
@@ -1295,6 +1295,16 @@ traceon_trigger(struct event_trigger_data *data,
 		struct trace_buffer *buffer, void *rec,
 		struct ring_buffer_event *event)
 {
+	struct trace_event_file *file = data->private_data;
+
+	if (file) {
+		if (tracer_tracing_is_on(file->tr))
+			return;
+
+		tracer_tracing_on(file->tr);
+		return;
+	}
+
 	if (tracing_is_on())
 		return;
 
@@ -1306,8 +1316,15 @@ traceon_count_trigger(struct event_trigger_data *data,
 		      struct trace_buffer *buffer, void *rec,
 		      struct ring_buffer_event *event)
 {
-	if (tracing_is_on())
-		return;
+	struct trace_event_file *file = data->private_data;
+
+	if (file) {
+		if (tracer_tracing_is_on(file->tr))
+			return;
+	} else {
+		if (tracing_is_on())
+			return;
+	}
 
 	if (!data->count)
 		return;
@@ -1315,7 +1332,10 @@ traceon_count_trigger(struct event_trigger_data *data,
 	if (data->count != -1)
 		(data->count)--;
 
-	tracing_on();
+	if (file)
+		tracer_tracing_on(file->tr);
+	else
+		tracing_on();
 }
 
 static void
@@ -1323,6 +1343,16 @@ traceoff_trigger(struct event_trigger_data *data,
 		 struct trace_buffer *buffer, void *rec,
 		 struct ring_buffer_event *event)
 {
+	struct trace_event_file *file = data->private_data;
+
+	if (file) {
+		if (!tracer_tracing_is_on(file->tr))
+			return;
+
+		tracer_tracing_off(file->tr);
+		return;
+	}
+
 	if (!tracing_is_on())
 		return;
 
@@ -1334,8 +1364,15 @@ traceoff_count_trigger(struct event_trigger_data *data,
 		       struct trace_buffer *buffer, void *rec,
 		       struct ring_buffer_event *event)
 {
-	if (!tracing_is_on())
-		return;
+	struct trace_event_file *file = data->private_data;
+
+	if (file) {
+		if (!tracer_tracing_is_on(file->tr))
+			return;
+	} else {
+		if (!tracing_is_on())
+			return;
+	}
 
 	if (!data->count)
 		return;
@@ -1343,7 +1380,10 @@ traceoff_count_trigger(struct event_trigger_data *data,
 	if (data->count != -1)
 		(data->count)--;
 
-	tracing_off();
+	if (file)
+		tracer_tracing_off(file->tr);
+	else
+		tracing_off();
 }
 
 static int
-- 
2.34.1

[for-linus][PATCH 04/13] eprobes: Remove redundant event type information

[Steven Rostedt]

From: "Steven Rostedt (Google)" <rostedt@goodmis.org>

Currently, the event probes save the type of the event they are attached
to when recording the event. For example:

  # echo 'e:switch sched/sched_switch prev_state=$prev_state prev_prio=$prev_prio next_pid=$next_pid next_prio=$next_prio' > dynamic_events
  # cat events/eprobes/switch/format

 name: switch
 ID: 1717
 format:
        field:unsigned short common_type;       offset:0;       size:2; signed:0;
        field:unsigned char common_flags;       offset:2;       size:1; signed:0;
        field:unsigned char common_preempt_count;       offset:3;       size:1; signed:0;
        field:int common_pid;   offset:4;       size:4; signed:1;

        field:unsigned int __probe_type;        offset:8;       size:4; signed:0;
        field:u64 prev_state;   offset:12;      size:8; signed:0;
        field:u64 prev_prio;    offset:20;      size:8; signed:0;
        field:u64 next_pid;     offset:28;      size:8; signed:0;
        field:u64 next_prio;    offset:36;      size:8; signed:0;

 print fmt: "(%u) prev_state=0x%Lx prev_prio=0x%Lx next_pid=0x%Lx next_prio=0x%Lx", REC->__probe_type, REC->prev_state, REC->prev_prio, REC->next_pid, REC->next_prio

The __probe_type adds 4 bytes to every event.

One of the reasons for creating eprobes is to limit what is traced in an
event to be able to limit what is written into the ring buffer. Having
this redundant 4 bytes to every event takes away from this.

The event that is recorded can be retrieved from the event probe itself,
that is available when the trace is happening. For user space tools, it
could simply read the dynamic_event file to find the event they are for.
So there is really no reason to write this information into the ring
buffer for every event.

Link: https://lkml.kernel.org/r/20220218190057.2f5a19a8@gandalf.local.home

Acked-by: Masami Hiramatsu <mhiramat@kernel.org>
Reviewed-by: Joel Fernandes <joel@joelfernandes.org>
Signed-off-by: Steven Rostedt (Google) <rostedt@goodmis.org>
---
 kernel/trace/trace.h        |  1 -
 kernel/trace/trace_eprobe.c | 16 +++++++---------
 kernel/trace/trace_probe.c  | 10 +++++-----
 kernel/trace/trace_probe.h  |  1 -
 4 files changed, 12 insertions(+), 16 deletions(-)

diff --git a/kernel/trace/trace.h b/kernel/trace/trace.h
index d038ddbf1bea..c5b09c31e077 100644
--- a/kernel/trace/trace.h
+++ b/kernel/trace/trace.h
@@ -136,7 +136,6 @@ struct kprobe_trace_entry_head {
 
 struct eprobe_trace_entry_head {
 	struct trace_entry	ent;
-	unsigned int		type;
 };
 
 struct kretprobe_trace_entry_head {
diff --git a/kernel/trace/trace_eprobe.c b/kernel/trace/trace_eprobe.c
index 191db32dec46..541aa13581b9 100644
--- a/kernel/trace/trace_eprobe.c
+++ b/kernel/trace/trace_eprobe.c
@@ -242,7 +242,6 @@ static int trace_eprobe_tp_arg_update(struct trace_eprobe *ep, int i)
 
 static int eprobe_event_define_fields(struct trace_event_call *event_call)
 {
-	int ret;
 	struct eprobe_trace_entry_head field;
 	struct trace_probe *tp;
 
@@ -250,8 +249,6 @@ static int eprobe_event_define_fields(struct trace_event_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);
 }
 
@@ -270,7 +267,9 @@ print_eprobe_event(struct trace_iterator *iter, int flags,
 	struct trace_event_call *pevent;
 	struct trace_event *probed_event;
 	struct trace_seq *s = &iter->seq;
+	struct trace_eprobe *ep;
 	struct trace_probe *tp;
+	unsigned int type;
 
 	field = (struct eprobe_trace_entry_head *)iter->ent;
 	tp = trace_probe_primary_from_call(
@@ -278,15 +277,18 @@ print_eprobe_event(struct trace_iterator *iter, int flags,
 	if (WARN_ON_ONCE(!tp))
 		goto out;
 
+	ep = container_of(tp, struct trace_eprobe, tp);
+	type = ep->event->event.type;
+
 	trace_seq_printf(s, "%s: (", trace_probe_name(tp));
 
-	probed_event = ftrace_find_event(field->type);
+	probed_event = ftrace_find_event(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_printf(s, "%u", type);
 	}
 
 	trace_seq_putc(s, ')');
@@ -498,10 +500,6 @@ __eprobe_trace_func(struct eprobe_data *edata, void *rec)
 		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);
diff --git a/kernel/trace/trace_probe.c b/kernel/trace/trace_probe.c
index 73d90179b51b..80863c6508e5 100644
--- a/kernel/trace/trace_probe.c
+++ b/kernel/trace/trace_probe.c
@@ -871,15 +871,15 @@ static int __set_print_fmt(struct trace_probe *tp, char *buf, int len,
 	switch (ptype) {
 	case PROBE_PRINT_NORMAL:
 		fmt = "(%lx)";
-		arg = "REC->" FIELD_STRING_IP;
+		arg = ", REC->" FIELD_STRING_IP;
 		break;
 	case PROBE_PRINT_RETURN:
 		fmt = "(%lx <- %lx)";
-		arg = "REC->" FIELD_STRING_FUNC ", REC->" FIELD_STRING_RETIP;
+		arg = ", REC->" FIELD_STRING_FUNC ", REC->" FIELD_STRING_RETIP;
 		break;
 	case PROBE_PRINT_EVENT:
-		fmt = "(%u)";
-		arg = "REC->" FIELD_STRING_TYPE;
+		fmt = "";
+		arg = "";
 		break;
 	default:
 		WARN_ON_ONCE(1);
@@ -903,7 +903,7 @@ static int __set_print_fmt(struct trace_probe *tp, char *buf, int len,
 					parg->type->fmt);
 	}
 
-	pos += snprintf(buf + pos, LEN_OR_ZERO, "\", %s", arg);
+	pos += snprintf(buf + pos, LEN_OR_ZERO, "\"%s", arg);
 
 	for (i = 0; i < tp->nr_args; i++) {
 		parg = tp->args + i;
diff --git a/kernel/trace/trace_probe.h b/kernel/trace/trace_probe.h
index 99e7a5df025e..92cc149af0fd 100644
--- a/kernel/trace/trace_probe.h
+++ b/kernel/trace/trace_probe.h
@@ -38,7 +38,6 @@
 #define FIELD_STRING_IP		"__probe_ip"
 #define FIELD_STRING_RETIP	"__probe_ret_ip"
 #define FIELD_STRING_FUNC	"__probe_func"
-#define FIELD_STRING_TYPE	"__probe_type"
 
 #undef DEFINE_FIELD
 #define DEFINE_FIELD(type, item, name, is_signed)			\
-- 
2.34.1

[for-linus][PATCH 05/13] tracing: Uninline trace_trigger_soft_disabled() partly

[Steven Rostedt]

From: Christophe Leroy <christophe.leroy@csgroup.eu>

On a powerpc32 build with CONFIG_CC_OPTIMISE_FOR_SIZE, the inline
keyword is not honored and trace_trigger_soft_disabled() appears
approx 50 times in vmlinux.

Adding -Winline to the build, the following message appears:

	./include/linux/trace_events.h:712:1: error: inlining failed in call to 'trace_trigger_soft_disabled': call is unlikely and code size would grow [-Werror=inline]

That function is rather big for an inlined function:

	c003df60 <trace_trigger_soft_disabled>:
	c003df60:	94 21 ff f0 	stwu    r1,-16(r1)
	c003df64:	7c 08 02 a6 	mflr    r0
	c003df68:	90 01 00 14 	stw     r0,20(r1)
	c003df6c:	bf c1 00 08 	stmw    r30,8(r1)
	c003df70:	83 e3 00 24 	lwz     r31,36(r3)
	c003df74:	73 e9 01 00 	andi.   r9,r31,256
	c003df78:	41 82 00 10 	beq     c003df88 <trace_trigger_soft_disabled+0x28>
	c003df7c:	38 60 00 00 	li      r3,0
	c003df80:	39 61 00 10 	addi    r11,r1,16
	c003df84:	4b fd 60 ac 	b       c0014030 <_rest32gpr_30_x>
	c003df88:	73 e9 00 80 	andi.   r9,r31,128
	c003df8c:	7c 7e 1b 78 	mr      r30,r3
	c003df90:	41 a2 00 14 	beq     c003dfa4 <trace_trigger_soft_disabled+0x44>
	c003df94:	38 c0 00 00 	li      r6,0
	c003df98:	38 a0 00 00 	li      r5,0
	c003df9c:	38 80 00 00 	li      r4,0
	c003dfa0:	48 05 c5 f1 	bl      c009a590 <event_triggers_call>
	c003dfa4:	73 e9 00 40 	andi.   r9,r31,64
	c003dfa8:	40 82 00 28 	bne     c003dfd0 <trace_trigger_soft_disabled+0x70>
	c003dfac:	73 ff 02 00 	andi.   r31,r31,512
	c003dfb0:	41 82 ff cc 	beq     c003df7c <trace_trigger_soft_disabled+0x1c>
	c003dfb4:	80 01 00 14 	lwz     r0,20(r1)
	c003dfb8:	83 e1 00 0c 	lwz     r31,12(r1)
	c003dfbc:	7f c3 f3 78 	mr      r3,r30
	c003dfc0:	83 c1 00 08 	lwz     r30,8(r1)
	c003dfc4:	7c 08 03 a6 	mtlr    r0
	c003dfc8:	38 21 00 10 	addi    r1,r1,16
	c003dfcc:	48 05 6f 6c 	b       c0094f38 <trace_event_ignore_this_pid>
	c003dfd0:	38 60 00 01 	li      r3,1
	c003dfd4:	4b ff ff ac 	b       c003df80 <trace_trigger_soft_disabled+0x20>

However it is located in a hot path so inlining it is important.
But forcing inlining of the entire function by using __always_inline
leads to increasing the text size by approx 20 kbytes.

Instead, split the fonction in two parts, one part with the likely
fast path, flagged __always_inline, and a second part out of line.

With this change, on a powerpc32 with CONFIG_CC_OPTIMISE_FOR_SIZE
vmlinux text increases by only 1,4 kbytes, which is partly
compensated by a decrease of vmlinux data by 7 kbytes.

On ppc64_defconfig which has CONFIG_CC_OPTIMISE_FOR_SPEED, this
change reduces vmlinux text by more than 30 kbytes.

Link: https://lkml.kernel.org/r/69ce0986a52d026d381d612801d978aa4f977460.1644563295.git.christophe.leroy@csgroup.eu

Signed-off-by: Christophe Leroy <christophe.leroy@csgroup.eu>
Signed-off-by: Steven Rostedt (Google) <rostedt@goodmis.org>
---
 include/linux/trace_events.h        | 22 ++++++++++++----------
 kernel/trace/trace_events_trigger.c | 14 ++++++++++++++
 2 files changed, 26 insertions(+), 10 deletions(-)

diff --git a/include/linux/trace_events.h b/include/linux/trace_events.h
index 70c069aef02c..dcea51fb60e2 100644
--- a/include/linux/trace_events.h
+++ b/include/linux/trace_events.h
@@ -699,6 +699,8 @@ event_triggers_post_call(struct trace_event_file *file,
 
 bool trace_event_ignore_this_pid(struct trace_event_file *trace_file);
 
+bool __trace_trigger_soft_disabled(struct trace_event_file *file);
+
 /**
  * trace_trigger_soft_disabled - do triggers and test if soft disabled
  * @file: The file pointer of the event to test
@@ -708,20 +710,20 @@ bool trace_event_ignore_this_pid(struct trace_event_file *trace_file);
  * triggers that require testing the fields, it will return true,
  * otherwise false.
  */
-static inline bool
+static __always_inline bool
 trace_trigger_soft_disabled(struct trace_event_file *file)
 {
 	unsigned long eflags = file->flags;
 
-	if (!(eflags & EVENT_FILE_FL_TRIGGER_COND)) {
-		if (eflags & EVENT_FILE_FL_TRIGGER_MODE)
-			event_triggers_call(file, NULL, NULL, NULL);
-		if (eflags & EVENT_FILE_FL_SOFT_DISABLED)
-			return true;
-		if (eflags & EVENT_FILE_FL_PID_FILTER)
-			return trace_event_ignore_this_pid(file);
-	}
-	return false;
+	if (likely(!(eflags & (EVENT_FILE_FL_TRIGGER_MODE |
+			       EVENT_FILE_FL_SOFT_DISABLED |
+			       EVENT_FILE_FL_PID_FILTER))))
+		return false;
+
+	if (likely(eflags & EVENT_FILE_FL_TRIGGER_COND))
+		return false;
+
+	return __trace_trigger_soft_disabled(file);
 }
 
 #ifdef CONFIG_BPF_EVENTS
diff --git a/kernel/trace/trace_events_trigger.c b/kernel/trace/trace_events_trigger.c
index efe563140f27..7eb9d04f1c2e 100644
--- a/kernel/trace/trace_events_trigger.c
+++ b/kernel/trace/trace_events_trigger.c
@@ -84,6 +84,20 @@ event_triggers_call(struct trace_event_file *file,
 }
 EXPORT_SYMBOL_GPL(event_triggers_call);
 
+bool __trace_trigger_soft_disabled(struct trace_event_file *file)
+{
+	unsigned long eflags = file->flags;
+
+	if (eflags & EVENT_FILE_FL_TRIGGER_MODE)
+		event_triggers_call(file, NULL, NULL, NULL);
+	if (eflags & EVENT_FILE_FL_SOFT_DISABLED)
+		return true;
+	if (eflags & EVENT_FILE_FL_PID_FILTER)
+		return trace_event_ignore_this_pid(file);
+	return false;
+}
+EXPORT_SYMBOL_GPL(__trace_trigger_soft_disabled);
+
 /**
  * event_triggers_post_call - Call 'post_triggers' for a trace event
  * @file: The trace_event_file associated with the event
-- 
2.34.1

[for-linus][PATCH 06/13] tracing: Ensure trace buffer is at least 4096 bytes large

[Steven Rostedt]

From: Sven Schnelle <svens@linux.ibm.com>

Booting the kernel with 'trace_buf_size=1' give a warning at
boot during the ftrace selftests:

[    0.892809] Running postponed tracer tests:
[    0.892893] Testing tracer function:
[    0.901899] Callback from call_rcu_tasks_trace() invoked.
[    0.983829] Callback from call_rcu_tasks_rude() invoked.
[    1.072003] .. bad ring buffer .. corrupted trace buffer ..
[    1.091944] Callback from call_rcu_tasks() invoked.
[    1.097695] PASSED
[    1.097701] Testing dynamic ftrace: .. filter failed count=0 ..FAILED!
[    1.353474] ------------[ cut here ]------------
[    1.353478] WARNING: CPU: 0 PID: 1 at kernel/trace/trace.c:1951 run_tracer_selftest+0x13c/0x1b0

Therefore enforce a minimum of 4096 bytes to make the selftest pass.

Link: https://lkml.kernel.org/r/20220214134456.1751749-1-svens@linux.ibm.com

Signed-off-by: Sven Schnelle <svens@linux.ibm.com>
Signed-off-by: Steven Rostedt (Google) <rostedt@goodmis.org>
---
 kernel/trace/trace.c | 10 ++++++----
 1 file changed, 6 insertions(+), 4 deletions(-)

diff --git a/kernel/trace/trace.c b/kernel/trace/trace.c
index 7c2578efde26..3050892d1812 100644
--- a/kernel/trace/trace.c
+++ b/kernel/trace/trace.c
@@ -1474,10 +1474,12 @@ static int __init set_buf_size(char *str)
 	if (!str)
 		return 0;
 	buf_size = memparse(str, &str);
-	/* nr_entries can not be zero */
-	if (buf_size == 0)
-		return 0;
-	trace_buf_size = buf_size;
+	/*
+	 * nr_entries can not be zero and the startup
+	 * tests require some buffer space. Therefore
+	 * ensure we have at least 4096 bytes of buffer.
+	 */
+	trace_buf_size = max(4096UL, buf_size);
 	return 1;
 }
 __setup("trace_buf_size=", set_buf_size);
-- 
2.34.1

[for-linus][PATCH 07/13] ftrace: Remove unused ftrace_startup_enable() stub

[Steven Rostedt]

From: Nathan Chancellor <nathan@kernel.org>

When building with clang + CONFIG_DYNAMIC_FTRACE=n + W=1, there is a
warning:

  kernel/trace/ftrace.c:7194:20: error: unused function 'ftrace_startup_enable' [-Werror,-Wunused-function]
  static inline void ftrace_startup_enable(int command) { }
                     ^
  1 error generated.

Clang warns on instances of static inline functions in .c files with W=1
after commit 6863f5643dd7 ("kbuild: allow Clang to find unused static
inline functions for W=1 build").

The ftrace_startup_enable() stub has been unused since
commit e1effa0144a1 ("ftrace: Annotate the ops operation on update"),
where its use outside of the CONFIG_DYNAMIC_TRACE section was replaced
by ftrace_startup_all().  Remove it to resolve the warning.

Link: https://lkml.kernel.org/r/20220214192847.488166-1-nathan@kernel.org

Reported-by: kernel test robot <lkp@intel.com>
Signed-off-by: Nathan Chancellor <nathan@kernel.org>
Signed-off-by: Steven Rostedt (Google) <rostedt@goodmis.org>
---
 kernel/trace/ftrace.c | 1 -
 1 file changed, 1 deletion(-)

diff --git a/kernel/trace/ftrace.c b/kernel/trace/ftrace.c
index f9feb197b2da..a4b462b6f944 100644
--- a/kernel/trace/ftrace.c
+++ b/kernel/trace/ftrace.c
@@ -7191,7 +7191,6 @@ static int __init ftrace_nodyn_init(void)
 core_initcall(ftrace_nodyn_init);
 
 static inline int ftrace_init_dyn_tracefs(struct dentry *d_tracer) { return 0; }
-static inline void ftrace_startup_enable(int command) { }
 static inline void ftrace_startup_all(int command) { }
 
 # define ftrace_startup_sysctl()	do { } while (0)
-- 
2.34.1

[for-linus][PATCH 08/13] tracing/osnoise: Make osnoise_main to sleep for microseconds

[Steven Rostedt]

From: Daniel Bristot de Oliveira <bristot@kernel.org>

osnoise's runtime and period are in the microseconds scale, but it is
currently sleeping in the millisecond's scale. This behavior roots in the
usage of hwlat as the skeleton for osnoise.

Make osnoise to sleep in the microseconds scale. Also, move the sleep to
a specialized function.

Link: https://lkml.kernel.org/r/302aa6c7bdf2d131719b22901905e9da122a11b2.1645197336.git.bristot@kernel.org

Cc: Ingo Molnar <mingo@redhat.com>
Signed-off-by: Daniel Bristot de Oliveira <bristot@kernel.org>
Signed-off-by: Steven Rostedt (Google) <rostedt@goodmis.org>
---
 kernel/trace/trace_osnoise.c | 53 ++++++++++++++++++++++--------------
 1 file changed, 32 insertions(+), 21 deletions(-)

diff --git a/kernel/trace/trace_osnoise.c b/kernel/trace/trace_osnoise.c
index 870a08da5b48..cfddb30e65ab 100644
--- a/kernel/trace/trace_osnoise.c
+++ b/kernel/trace/trace_osnoise.c
@@ -1436,6 +1436,37 @@ static int run_osnoise(void)
 static struct cpumask osnoise_cpumask;
 static struct cpumask save_cpumask;
 
+/*
+ * osnoise_sleep - sleep until the next period
+ */
+static void osnoise_sleep(void)
+{
+	u64 interval;
+	ktime_t wake_time;
+
+	mutex_lock(&interface_lock);
+	interval = osnoise_data.sample_period - osnoise_data.sample_runtime;
+	mutex_unlock(&interface_lock);
+
+	/*
+	 * differently from hwlat_detector, the osnoise tracer can run
+	 * without a pause because preemption is on.
+	 */
+	if (!interval) {
+		/* Let synchronize_rcu_tasks() make progress */
+		cond_resched_tasks_rcu_qs();
+		return;
+	}
+
+	wake_time = ktime_add_us(ktime_get(), interval);
+	__set_current_state(TASK_INTERRUPTIBLE);
+
+	while (schedule_hrtimeout_range(&wake_time, 0, HRTIMER_MODE_ABS)) {
+		if (kthread_should_stop())
+			break;
+	}
+}
+
 /*
  * osnoise_main - The osnoise detection kernel thread
  *
@@ -1444,30 +1475,10 @@ static struct cpumask save_cpumask;
  */
 static int osnoise_main(void *data)
 {
-	u64 interval;
 
 	while (!kthread_should_stop()) {
-
 		run_osnoise();
-
-		mutex_lock(&interface_lock);
-		interval = osnoise_data.sample_period - osnoise_data.sample_runtime;
-		mutex_unlock(&interface_lock);
-
-		do_div(interval, USEC_PER_MSEC);
-
-		/*
-		 * differently from hwlat_detector, the osnoise tracer can run
-		 * without a pause because preemption is on.
-		 */
-		if (interval < 1) {
-			/* Let synchronize_rcu_tasks() make progress */
-			cond_resched_tasks_rcu_qs();
-			continue;
-		}
-
-		if (msleep_interruptible(interval))
-			break;
+		osnoise_sleep();
 	}
 
 	return 0;
-- 
2.34.1

[for-linus][PATCH 09/13] tracefs: Set the group ownership in apply_options() not parse_options()

[Steven Rostedt]

From: "Steven Rostedt (Google)" <rostedt@goodmis.org>

Al Viro brought it to my attention that the dentries may not be filled
when the parse_options() is called, causing the call to set_gid() to
possibly crash. It should only be called if parse_options() succeeds
totally anyway.

He suggested the logical place to do the update is in apply_options().

Cc: stable@vger.kernel.org
Reported-by: Al Viro <viro@zeniv.linux.org.uk>
Fixes: 48b27b6b5191 ("tracefs: Set all files to the same group ownership as the mount option")
Signed-off-by: Steven Rostedt (Google) <rostedt@goodmis.org>
---
 fs/tracefs/inode.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

diff --git a/fs/tracefs/inode.c b/fs/tracefs/inode.c
index bafc02bf8220..3638d330ff5a 100644
--- a/fs/tracefs/inode.c
+++ b/fs/tracefs/inode.c
@@ -264,7 +264,6 @@ static int tracefs_parse_options(char *data, struct tracefs_mount_opts *opts)
 			if (!gid_valid(gid))
 				return -EINVAL;
 			opts->gid = gid;
-			set_gid(tracefs_mount->mnt_root, gid);
 			break;
 		case Opt_mode:
 			if (match_octal(&args[0], &option))
@@ -293,6 +292,9 @@ static int tracefs_apply_options(struct super_block *sb)
 	inode->i_uid = opts->uid;
 	inode->i_gid = opts->gid;
 
+	if (tracefs_mount && tracefs_mount->mnt_root)
+		set_gid(tracefs_mount->mnt_root, opts->gid);
+
 	return 0;
 }
 
-- 
2.34.1

[for-linus][PATCH 10/13] tracing: Fix selftest config check for function graph start up test

[Steven Rostedt]

From: Christophe Leroy <christophe.leroy@csgroup.eu>

CONFIG_DYNAMIC_FTRACE_WITH_DIRECT_CALLS is required to test
direct tramp.

Link: https://lkml.kernel.org/r/bdc7e594e13b0891c1d61bc8d56c94b1890eaed7.1640017960.git.christophe.leroy@csgroup.eu

Signed-off-by: Christophe Leroy <christophe.leroy@csgroup.eu>
Signed-off-by: Steven Rostedt (Google) <rostedt@goodmis.org>
---
 kernel/trace/trace_selftest.c | 6 ++----
 1 file changed, 2 insertions(+), 4 deletions(-)

diff --git a/kernel/trace/trace_selftest.c b/kernel/trace/trace_selftest.c
index afd937a46496..abcadbe933bb 100644
--- a/kernel/trace/trace_selftest.c
+++ b/kernel/trace/trace_selftest.c
@@ -784,9 +784,7 @@ static struct fgraph_ops fgraph_ops __initdata  = {
 	.retfunc		= &trace_graph_return,
 };
 
-#if defined(CONFIG_DYNAMIC_FTRACE) && \
-    defined(CONFIG_HAVE_DYNAMIC_FTRACE_WITH_ARGS)
-#define TEST_DIRECT_TRAMP
+#ifdef CONFIG_DYNAMIC_FTRACE_WITH_DIRECT_CALLS
 noinline __noclone static void trace_direct_tramp(void) { }
 #endif
 
@@ -849,7 +847,7 @@ trace_selftest_startup_function_graph(struct tracer *trace,
 		goto out;
 	}
 
-#ifdef TEST_DIRECT_TRAMP
+#ifdef CONFIG_DYNAMIC_FTRACE_WITH_DIRECT_CALLS
 	tracing_reset_online_cpus(&tr->array_buffer);
 	set_graph_array(tr);
 
-- 
2.34.1

[for-linus][PATCH 11/13] rtla/hist: Make -E the short version of --entries

[Steven Rostedt]

From: Daniel Bristot de Oliveira <bristot@kernel.org>

Currently, --entries uses -e as the short version in the hist mode of
timerlat and osnoise tools. But as -e is already used to enable events
on trace sessions by other tools, thus let's keep it available for the
same usage for all rtla tools.

Make -E the short version of --entries for hist mode on all tools.

Note: rtla was merged in this merge window, so rtla was not released yet.

Link: https://lkml.kernel.org/r/5dbf0cbe7364d3a05e708926b41a097c59a02b1e.1645206561.git.bristot@kernel.org

Cc: Jonathan Corbet <corbet@lwn.net>
Signed-off-by: Daniel Bristot de Oliveira <bristot@kernel.org>
Signed-off-by: Steven Rostedt (Google) <rostedt@goodmis.org>
---
 Documentation/tools/rtla/common_hist_options.rst |  2 +-
 Documentation/tools/rtla/rtla-osnoise-hist.rst   |  2 +-
 tools/tracing/rtla/src/osnoise_hist.c            | 10 +++++-----
 tools/tracing/rtla/src/timerlat_hist.c           | 10 +++++-----
 4 files changed, 12 insertions(+), 12 deletions(-)

diff --git a/Documentation/tools/rtla/common_hist_options.rst b/Documentation/tools/rtla/common_hist_options.rst
index 0266cd08a6c9..df53ff835bfb 100644
--- a/Documentation/tools/rtla/common_hist_options.rst
+++ b/Documentation/tools/rtla/common_hist_options.rst
@@ -2,7 +2,7 @@
 
         Set the histogram bucket size (default *1*).
 
-**-e**, **--entries** *N*
+**-E**, **--entries** *N*
 
         Set the number of entries of the histogram (default 256).
 
diff --git a/Documentation/tools/rtla/rtla-osnoise-hist.rst b/Documentation/tools/rtla/rtla-osnoise-hist.rst
index 52298ddd8701..f2e79d22c4c4 100644
--- a/Documentation/tools/rtla/rtla-osnoise-hist.rst
+++ b/Documentation/tools/rtla/rtla-osnoise-hist.rst
@@ -36,7 +36,7 @@ default). The reason for reducing the runtime is to avoid starving the
 **rtla** tool. The tool is also set to run for *one minute*. The output
 histogram is set to group outputs in buckets of *10us* and *25* entries::
 
-  [root@f34 ~/]# rtla osnoise hist -P F:1 -c 0-11 -r 900000 -d 1M -b 10 -e 25
+  [root@f34 ~/]# rtla osnoise hist -P F:1 -c 0-11 -r 900000 -d 1M -b 10 -E 25
   # RTLA osnoise histogram
   # Time unit is microseconds (us)
   # Duration:   0 00:01:00
diff --git a/tools/tracing/rtla/src/osnoise_hist.c b/tools/tracing/rtla/src/osnoise_hist.c
index 1f0b7fce55cf..52c053cc1789 100644
--- a/tools/tracing/rtla/src/osnoise_hist.c
+++ b/tools/tracing/rtla/src/osnoise_hist.c
@@ -426,7 +426,7 @@ static void osnoise_hist_usage(char *usage)
 	static const char * const msg[] = {
 		"",
 		"  usage: rtla osnoise hist [-h] [-D] [-d s] [-p us] [-r us] [-s us] [-S us] [-t[=file]] \\",
-		"	  [-c cpu-list] [-P priority] [-b N] [-e N] [--no-header] [--no-summary] \\",
+		"	  [-c cpu-list] [-P priority] [-b N] [-E N] [--no-header] [--no-summary] \\",
 		"	  [--no-index] [--with-zeros]",
 		"",
 		"	  -h/--help: print this menu",
@@ -439,7 +439,7 @@ static void osnoise_hist_usage(char *usage)
 		"	  -D/--debug: print debug info",
 		"	  -t/--trace[=file]: save the stopped trace to [file|osnoise_trace.txt]",
 		"	  -b/--bucket-size N: set the histogram bucket size (default 1)",
-		"	  -e/--entries N: set the number of entries of the histogram (default 256)",
+		"	  -E/--entries N: set the number of entries of the histogram (default 256)",
 		"	     --no-header: do not print header",
 		"	     --no-summary: do not print summary",
 		"	     --no-index: do not print index",
@@ -486,7 +486,7 @@ static struct osnoise_hist_params
 	while (1) {
 		static struct option long_options[] = {
 			{"bucket-size",		required_argument,	0, 'b'},
-			{"entries",		required_argument,	0, 'e'},
+			{"entries",		required_argument,	0, 'E'},
 			{"cpus",		required_argument,	0, 'c'},
 			{"debug",		no_argument,		0, 'D'},
 			{"duration",		required_argument,	0, 'd'},
@@ -507,7 +507,7 @@ static struct osnoise_hist_params
 		/* getopt_long stores the option index here. */
 		int option_index = 0;
 
-		c = getopt_long(argc, argv, "c:b:d:e:Dhp:P:r:s:S:t::0123",
+		c = getopt_long(argc, argv, "c:b:d:E:Dhp:P:r:s:S:t::0123",
 				 long_options, &option_index);
 
 		/* detect the end of the options. */
@@ -534,7 +534,7 @@ static struct osnoise_hist_params
 			if (!params->duration)
 				osnoise_hist_usage("Invalid -D duration\n");
 			break;
-		case 'e':
+		case 'E':
 			params->entries = get_llong_from_str(optarg);
 			if ((params->entries < 10) || (params->entries > 9999999))
 				osnoise_hist_usage("Entries must be > 10 and < 9999999\n");
diff --git a/tools/tracing/rtla/src/timerlat_hist.c b/tools/tracing/rtla/src/timerlat_hist.c
index 436a799f9adf..237e1735afa7 100644
--- a/tools/tracing/rtla/src/timerlat_hist.c
+++ b/tools/tracing/rtla/src/timerlat_hist.c
@@ -429,7 +429,7 @@ static void timerlat_hist_usage(char *usage)
 	char *msg[] = {
 		"",
 		"  usage: [rtla] timerlat hist [-h] [-q] [-d s] [-D] [-n] [-p us] [-i us] [-T us] [-s us] [-t[=file]] \\",
-		"         [-c cpu-list] [-P priority] [-e N] [-b N]  [--no-irq] [--no-thread] [--no-header] [--no-summary] \\",
+		"         [-c cpu-list] [-P priority] [-E N] [-b N]  [--no-irq] [--no-thread] [--no-header] [--no-summary] \\",
 		"         [--no-index] [--with-zeros]",
 		"",
 		"	  -h/--help: print this menu",
@@ -443,7 +443,7 @@ static void timerlat_hist_usage(char *usage)
 		"	  -T/--trace[=file]: save the stopped trace to [file|timerlat_trace.txt]",
 		"	  -n/--nano: display data in nanoseconds",
 		"	  -b/--bucket-size N: set the histogram bucket size (default 1)",
-		"	  -e/--entries N: set the number of entries of the histogram (default 256)",
+		"	  -E/--entries N: set the number of entries of the histogram (default 256)",
 		"	     --no-irq: ignore IRQ latencies",
 		"	     --no-thread: ignore thread latencies",
 		"	     --no-header: do not print header",
@@ -494,7 +494,7 @@ static struct timerlat_hist_params
 			{"cpus",		required_argument,	0, 'c'},
 			{"bucket-size",		required_argument,	0, 'b'},
 			{"debug",		no_argument,		0, 'D'},
-			{"entries",		required_argument,	0, 'e'},
+			{"entries",		required_argument,	0, 'E'},
 			{"duration",		required_argument,	0, 'd'},
 			{"help",		no_argument,		0, 'h'},
 			{"irq",			required_argument,	0, 'i'},
@@ -516,7 +516,7 @@ static struct timerlat_hist_params
 		/* getopt_long stores the option index here. */
 		int option_index = 0;
 
-		c = getopt_long(argc, argv, "c:b:d:e:Dhi:np:P:s:t::T:012345",
+		c = getopt_long(argc, argv, "c:b:d:E:Dhi:np:P:s:t::T:012345",
 				 long_options, &option_index);
 
 		/* detect the end of the options. */
@@ -543,7 +543,7 @@ static struct timerlat_hist_params
 			if (!params->duration)
 				timerlat_hist_usage("Invalid -D duration\n");
 			break;
-		case 'e':
+		case 'E':
 			params->entries = get_llong_from_str(optarg);
 			if ((params->entries < 10) || (params->entries > 9999999))
 					timerlat_hist_usage("Entries must be > 10 and < 9999999\n");
-- 
2.34.1

[for-linus][PATCH 12/13] rtla/osnoise: Free params at the exit

[Steven Rostedt]

From: Daniel Bristot de Oliveira <bristot@kernel.org>

The variable that stores the parsed command line arguments are not
being free()d at the rtla osnoise top exit path.

Free params variable before exiting.

Link: https://lkml.kernel.org/r/0be31d8259c7c53b98a39769d60cfeecd8421785.1645206561.git.bristot@kernel.org

Fixes: 1eceb2fc2ca5 ("rtla/osnoise: Add osnoise top mode")
Cc: Jonathan Corbet <corbet@lwn.net>
Signed-off-by: Daniel Bristot de Oliveira <bristot@kernel.org>
Signed-off-by: Steven Rostedt (Google) <rostedt@goodmis.org>
---
 tools/tracing/rtla/src/osnoise_top.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/tools/tracing/rtla/src/osnoise_top.c b/tools/tracing/rtla/src/osnoise_top.c
index c67dc28ef716..7af769b9c0de 100644
--- a/tools/tracing/rtla/src/osnoise_top.c
+++ b/tools/tracing/rtla/src/osnoise_top.c
@@ -573,6 +573,7 @@ int osnoise_top_main(int argc, char **argv)
 	osnoise_free_top(tool->data);
 	osnoise_destroy_tool(record);
 	osnoise_destroy_tool(tool);
+	free(params);
 out_exit:
 	exit(return_value);
 }
-- 
2.34.1

[for-linus][PATCH 13/13] rtla/osnoise: Fix error message when failing to enable trace instance

[Steven Rostedt]

From: Daniel Bristot de Oliveira <bristot@kernel.org>

When a trace instance creation fails, tools are printing:

	Could not enable -> osnoiser <- tracer for tracing

Print the actual (and correct) name of the tracer it fails to enable.

Link: https://lkml.kernel.org/r/53ef0582605af91eca14b19dba9fc9febb95d4f9.1645206561.git.bristot@kernel.org

Fixes: b1696371d865 ("rtla: Helper functions for rtla")
Cc: Jonathan Corbet <corbet@lwn.net>
Signed-off-by: Daniel Bristot de Oliveira <bristot@kernel.org>
Signed-off-by: Steven Rostedt (Google) <rostedt@goodmis.org>
---
 tools/tracing/rtla/src/osnoise.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/tools/tracing/rtla/src/osnoise.c b/tools/tracing/rtla/src/osnoise.c
index 5648f9252e58..e60f1862bad0 100644
--- a/tools/tracing/rtla/src/osnoise.c
+++ b/tools/tracing/rtla/src/osnoise.c
@@ -810,7 +810,7 @@ struct osnoise_tool *osnoise_init_trace_tool(char *tracer)
 
 	retval = enable_tracer_by_name(trace->trace.inst, tracer);
 	if (retval) {
-		err_msg("Could not enable osnoiser tracer for tracing\n");
+		err_msg("Could not enable %s tracer for tracing\n", tracer);
 		goto out_err;
 	}
 
-- 
2.34.1

Re: [for-linus][PATCH 09/13] tracefs: Set the group ownership in apply_options() not parse_options()

[Al Viro]

On Fri, Feb 25, 2022 at 11:52:00AM -0500, Steven Rostedt wrote:

> @@ -293,6 +292,9 @@ static int tracefs_apply_options(struct super_block *sb)
>  	inode->i_uid = opts->uid;
>  	inode->i_gid = opts->gid;
>  
> +	if (tracefs_mount && tracefs_mount->mnt_root)
> +		set_gid(tracefs_mount->mnt_root, opts->gid);
> +

Umm...  Why bother with tracefs_mount here in the first place?
You have the superblock the operation is applied for - right in the
arguments.

Just make that
	set_gid(sb->s_root, opts->gid);
and be done with that.  Same place in tracefs_apply_options()...

Incidentally, I'm not sure you need to keep ->i_gid assignment - set_gid()
won't miss that inode, so...

Re: [for-linus][PATCH 09/13] tracefs: Set the group ownership in apply_options() not parse_options()

[Steven Rostedt]

On Fri, 25 Feb 2022 17:11:00 +0000
Al Viro <viro@zeniv.linux.org.uk> wrote:

> On Fri, Feb 25, 2022 at 11:52:00AM -0500, Steven Rostedt wrote:
> 
> > @@ -293,6 +292,9 @@ static int tracefs_apply_options(struct super_block *sb)
> >  	inode->i_uid = opts->uid;
> >  	inode->i_gid = opts->gid;
> >  
> > +	if (tracefs_mount && tracefs_mount->mnt_root)
> > +		set_gid(tracefs_mount->mnt_root, opts->gid);
> > +  
> 
> Umm...  Why bother with tracefs_mount here in the first place?
> You have the superblock the operation is applied for - right in the
> arguments.

Ah, because I didn't realize they were one and the same ;-)

> 
> Just make that
> 	set_gid(sb->s_root, opts->gid);
> and be done with that.  Same place in tracefs_apply_options()...
> 
> Incidentally, I'm not sure you need to keep ->i_gid assignment - set_gid()
> won't miss that inode, so...

Makes sense. Thanks.

I guess I wont be sending my pull request to Linus now :-/

-- Steve