tracing: Restore system filter behavior

PATCH In review Li Zefan 2 participants 8 messages Nov 1, 2011 – Dec 6, 2011

Subsystems (2)

Tracing ftrace

lore

Share & tools

lore

Li Zefan

author

2011-11-01 01:07 UTC #0 lore raw

PATCH 1/1 tracing: Restore system filter behavior

Though not all events have field 'prev_pid', it was allowed to do this:

  # echo 'prev_pid == 100' > events/sched/filter

but commit 75b8e98263fdb0bfbdeba60d4db463259f1fe8a2 (tracing/filter: Swap entire filter of events) broke it without any reason.

Signed-off-by Li Zefan <[email protected]>

---
 include/linux/ftrace_event.h       |    2 ++
 kernel/trace/trace_events_filter.c |    7 ++++++-
 2 files changed, 8 insertions(+), 1 deletions(-)

include/linux/ftrace_event.h +2 -0

index 96efa67..c3da42d 100644
@@ -172,6 +172,7 @@ enum {
 	TRACE_EVENT_FL_FILTERED_BIT,
 	TRACE_EVENT_FL_RECORDED_CMD_BIT,
 	TRACE_EVENT_FL_CAP_ANY_BIT,
+	TRACE_EVENT_FL_NO_SET_FILTER_BIT,
 };
 
 enum {

@@ -179,6 +180,7 @@ enum {
 	TRACE_EVENT_FL_FILTERED		= (1 << TRACE_EVENT_FL_FILTERED_BIT),
 	TRACE_EVENT_FL_RECORDED_CMD	= (1 << TRACE_EVENT_FL_RECORDED_CMD_BIT),
 	TRACE_EVENT_FL_CAP_ANY		= (1 << TRACE_EVENT_FL_CAP_ANY_BIT),
+	TRACE_EVENT_FL_NO_SET_FILTER	= (1 << TRACE_EVENT_FL_NO_SET_FILTER_BIT),
 };
 
 struct ftrace_event_call {

kernel/trace/trace_events_filter.c +6 -1

index 816d3d0..86040d9 100644
@@ -1649,7 +1649,9 @@ static int replace_system_preds(struct event_subsystem *system,
 		 */
 		err = replace_preds(call, NULL, ps, filter_string, true);
 		if (err)
-			goto fail;
+			call->flags |= TRACE_EVENT_FL_NO_SET_FILTER;
+		else
+			call->flags &= ~TRACE_EVENT_FL_NO_SET_FILTER;
 	}
 
 	list_for_each_entry(call, &ftrace_events, list) {

@@ -1658,6 +1660,9 @@ static int replace_system_preds(struct event_subsystem *system,
 		if (strcmp(call->class->system, system->name) != 0)
 			continue;
 
+		if (call->flags & TRACE_EVENT_FL_NO_SET_FILTER)
+			continue;
+
 		filter_item = kzalloc(sizeof(*filter_item), GFP_KERNEL);
 		if (!filter_item)
 			goto fail_mem;

Signature · 1 line

1.7.3.1

Steven Rostedt

maintainer

2011-11-02 18:22 UTC #1 lore raw

Quoted 6 lines · Li Zefan

Though not all events have field 'prev_pid', it was allowed to do this:
  # echo 'prev_pid == 100' > events/sched/filter
but commit 75b8e98263fdb0bfbdeba60d4db463259f1fe8a2 (tracing/filter: Swap entire filter of events) broke it without any reason.

I wouldn't say "without any reason" as there was a reason.

But I'm not pleased with the way this all works still.

# echo 'prev_pid == 100' > /debug/tracing/events/sched/filter
# cat /debug/tracing/events/sched/filter
prev_pid == 100

# cat /debug/tracing/events/sched/sched_migrate_task/filter
none

# echo 'orig_cpu == 1' > /debug/tracing/events/sched/sched_migrate_task/filter
# cat /debug/tracing/events/sched/sched_migrate_task/filter
orig_cpu == 1

# cat /debug/tracing/events/sched/filter
prev_pid == 100

The above shows how things are just ambiguous. I have no problem in using the top filter to set multiple events. But the top filter should not keep the state of what was set. Perhaps just have system event filters always show default text. Like:

# cat /debug/tracing/events/sched/filter
### global filter ###
# Use this to set multiple event filters
# Only affects events that have the event fields specified in the filter

I'll add your patch, but are you OK with the above always printing for system event filters? Just to remove the ambiguous state.

-- Steve

Steven Rostedt

maintainer

2011-11-02 18:51 UTC #2 lore raw

Quoted 5 lines · Steven Rostedt

The above shows how things are just ambiguous. I have no problem in using the top filter to set multiple events. But the top filter should not keep the state of what was set. Perhaps just have system event filters always show default text. Like:

Quoted 8 lines

# cat /debug/tracing/events/sched/filter
### global filter ###
# Use this to set multiple event filters
# Only affects events that have the event fields specified in the filter
I'll add your patch, but are you OK with the above always printing for system event filters? Just to remove the ambiguous state.

As the filter is also used to show errors, I'll only have it print the "message" if the filter worked. If there's an error, then the error message will display instead.

-- Steve

Steven Rostedt

maintainer

2011-11-02 20:45 UTC #3 lore raw

Quoted 13 lines · Steven Rostedt

Quoted 11 lines · Steven Rostedt
The above shows how things are just ambiguous. I have no problem in using the top filter to set multiple events. But the top filter should not keep the state of what was set. Perhaps just have system event filters always show default text. Like:
# cat /debug/tracing/events/sched/filter
### global filter ###
# Use this to set multiple event filters
# Only affects events that have the event fields specified in the filter

Quoted 8 lines

Quoted 2 lines
I'll add your patch, but are you OK with the above always printing for system event filters? Just to remove the ambiguous state.
As the filter is also used to show errors, I'll only have it print the "message" if the filter worked. If there's an error, then the error message will display instead.

What's your thought on the below patch?

-- Steve

kernel/trace/trace_events_filter.c +17 -7

index 86040d9..6dee2b5 100644
@@ -27,6 +27,12 @@
 #include "trace.h"
 #include "trace_output.h"
 
+#define DEFAULT_SYS_FILTER_MESSAGE					\
+	"### global filter ###\n"					\
+	"# Use this to set filters for multiple events.\n"		\
+	"# Only events with the given fields will be affected.\n"	\
+	"# If no events are modified, an error message will be displayed here"
+
 enum filter_op_ids
 {
 	OP_OR,

@@ -646,7 +652,7 @@ void print_subsystem_event_filter(struct event_subsystem *system,
 	if (filter && filter->filter_string)
 		trace_seq_printf(s, "%s\n", filter->filter_string);
 	else
-		trace_seq_printf(s, "none\n");
+		trace_seq_printf(s, DEFAULT_SYS_FILTER_MESSAGE "\n");
 	mutex_unlock(&event_mutex);
 }

@@ -1838,7 +1844,8 @@ int apply_subsystem_event_filter(struct event_subsystem *system,
 	if (!filter)
 		goto out;
 
-	replace_filter_string(filter, filter_string);
+	/* System filters just show a default message */
+	replace_filter_string(filter, DEFAULT_SYS_FILTER_MESSAGE);
 	/*
 	 * No event actually uses the system filter
 	 * we can free it without synchronize_sched().

@@ -1848,14 +1855,12 @@ int apply_subsystem_event_filter(struct event_subsystem *system,
 
 	parse_init(ps, filter_ops, filter_string);
 	err = filter_parse(ps);
-	if (err) {
-		append_filter_err(ps, system->filter);
-		goto out;
-	}
+	if (err)
+		goto err_filter;
 
 	err = replace_system_preds(system, ps, filter_string);
 	if (err)
-		append_filter_err(ps, system->filter);
+		goto err_filter;
 
 out:
 	filter_opstack_clear(ps);

@@ -1865,6 +1870,11 @@ out_unlock:
 	mutex_unlock(&event_mutex);
 
 	return err;
+
+err_filter:
+	replace_filter_string(filter, filter_string);
+	append_filter_err(ps, system->filter);
+	goto out;
 }
 
 #ifdef CONFIG_PERF_EVENTS

Li Zefan

author

2011-11-03 01:24 UTC #4 lore raw

Quoted 18 lines · Steven Rostedt

Quoted 17 lines · Steven Rostedt
Quoted 13 lines · Steven Rostedt
The above shows how things are just ambiguous. I have no problem in using the top filter to set multiple events. But the top filter should not keep the state of what was set. Perhaps just have system event filters always show default text. Like:
# cat /debug/tracing/events/sched/filter
### global filter ###
# Use this to set multiple event filters
# Only affects events that have the event fields specified in the filter
I'll add your patch, but are you OK with the above always printing for system event filters? Just to remove the ambiguous state.

Quoted 7 lines

Quoted 4 lines
As the filter is also used to show errors, I'll only have it print the "message" if the filter worked. If there's an error, then the error message will display instead.
What's your thought on the below patch?

I'm fine with this idea.

a comment below..

Quoted 29 lines

-- Steve

kernel/trace/trace_events_filter.c +7 -1

index 86040d9..6dee2b5 100644
@@ -27,6 +27,12 @@
 #include "trace.h"
 #include "trace_output.h"
 
+#define DEFAULT_SYS_FILTER_MESSAGE					\
+	"### global filter ###\n"					\
+	"# Use this to set filters for multiple events.\n"		\
+	"# Only events with the given fields will be affected.\n"	\
+	"# If no events are modified, an error message will be displayed here"
+
 enum filter_op_ids
 {
 	OP_OR,

@@ -646,7 +652,7 @@ void print_subsystem_event_filter(struct event_subsystem *system,
 	if (filter && filter->filter_string)
 		trace_seq_printf(s, "%s\n", filter->filter_string);
 	else
-		trace_seq_printf(s, "none\n");
+		trace_seq_printf(s, DEFAULT_SYS_FILTER_MESSAGE "\n");
 	mutex_unlock(&event_mutex);
 }

kernel/trace/trace_events_filter.c int apply_subsystem_event_filter(struct event_subsystem *system, · line 1848

Quoted 6 lines

kernel/trace/trace_events_filter.c +2 -1

@@ -1838,7 +1844,8 @@ int apply_subsystem_event_filter(struct event_subsystem *system,
 	if (!filter)
 		goto out;
 
-	replace_filter_string(filter, filter_string);
+	/* System filters just show a default message */
+	replace_filter_string(filter, DEFAULT_SYS_FILTER_MESSAGE);

Seems we can call remove_filter_string() instead to save memory.

Quoted 36 lines

kernel/trace/trace_events_filter.c +8 -5

 	/*
 	 * No event actually uses the system filter
 	 * we can free it without synchronize_sched().

@@ -1848,14 +1855,12 @@ int apply_subsystem_event_filter(struct event_subsystem *system,
 
 	parse_init(ps, filter_ops, filter_string);
 	err = filter_parse(ps);
-	if (err) {
-		append_filter_err(ps, system->filter);
-		goto out;
-	}
+	if (err)
+		goto err_filter;
 
 	err = replace_system_preds(system, ps, filter_string);
 	if (err)
-		append_filter_err(ps, system->filter);
+		goto err_filter;
 
 out:
 	filter_opstack_clear(ps);

@@ -1865,6 +1870,11 @@ out_unlock:
 	mutex_unlock(&event_mutex);
 
 	return err;
+
+err_filter:
+	replace_filter_string(filter, filter_string);
+	append_filter_err(ps, system->filter);
+	goto out;
 }
 
 #ifdef CONFIG_PERF_EVENTS

Steven Rostedt

maintainer

2011-11-03 01:35 UTC #5 lore raw

Quoted 6 lines · Li Zefan

kernel/trace/trace_events_filter.c int apply_subsystem_event_filter(struct event_subsystem *system, · line 1848
Quoted 4 lines
kernel/trace/trace_events_filter.c +2 -1
-	replace_filter_string(filter, filter_string);
+	/* System filters just show a default message */
+	replace_filter_string(filter, DEFAULT_SYS_FILTER_MESSAGE);
Seems we can call remove_filter_string() instead to save memory.

I originally had it that way, but for some reason I had to change it. Not sure why. I'll put it back and see if that reason pops up again ;)

Thanks,

-- Steve

tip-bot for Li Zefan

author

2011-11-18 23:05 UTC #6 lore raw

In reply to Li Zefan

Commit-ID:  ed0449af5373abd766c79fbf83254bebc996bd23
Gitweb:     http://git.kernel.org/tip/ed0449af5373abd766c79fbf83254bebc996bd23
Author:     Li Zefan <[email protected]>
AuthorDate: Tue, 1 Nov 2011 09:09:35 +0800
Committer:  Steven Rostedt <[email protected]>
CommitDate: Wed, 2 Nov 2011 13:56:25 -0400

tracing: Restore system filter behavior

Though not all events have field 'prev_pid', it was allowed to do this:

  # echo 'prev_pid == 100' > events/sched/filter

but commit 75b8e98263fdb0bfbdeba60d4db463259f1fe8a2 (tracing/filter: Swap entire filter of events) broke it without any reason.

Link http://lkml.kernel.org/r/[email protected] open here

Signed-off-by Li Zefan <[email protected]>

Signed-off-by Steven Rostedt <[email protected]>

---
 include/linux/ftrace_event.h       |    2 ++
 kernel/trace/trace_events_filter.c |    7 ++++++-
 2 files changed, 8 insertions(+), 1 deletions(-)

include/linux/ftrace_event.h +2 -0

index 96efa67..c3da42d 100644
@@ -172,6 +172,7 @@ enum {
 	TRACE_EVENT_FL_FILTERED_BIT,
 	TRACE_EVENT_FL_RECORDED_CMD_BIT,
 	TRACE_EVENT_FL_CAP_ANY_BIT,
+	TRACE_EVENT_FL_NO_SET_FILTER_BIT,
 };
 
 enum {

@@ -179,6 +180,7 @@ enum {
 	TRACE_EVENT_FL_FILTERED		= (1 << TRACE_EVENT_FL_FILTERED_BIT),
 	TRACE_EVENT_FL_RECORDED_CMD	= (1 << TRACE_EVENT_FL_RECORDED_CMD_BIT),
 	TRACE_EVENT_FL_CAP_ANY		= (1 << TRACE_EVENT_FL_CAP_ANY_BIT),
+	TRACE_EVENT_FL_NO_SET_FILTER	= (1 << TRACE_EVENT_FL_NO_SET_FILTER_BIT),
 };
 
 struct ftrace_event_call {

kernel/trace/trace_events_filter.c +6 -1

index 816d3d0..86040d9 100644
@@ -1649,7 +1649,9 @@ static int replace_system_preds(struct event_subsystem *system,
 		 */
 		err = replace_preds(call, NULL, ps, filter_string, true);
 		if (err)
-			goto fail;
+			call->flags |= TRACE_EVENT_FL_NO_SET_FILTER;
+		else
+			call->flags &= ~TRACE_EVENT_FL_NO_SET_FILTER;
 	}
 
 	list_for_each_entry(call, &ftrace_events, list) {

@@ -1658,6 +1660,9 @@ static int replace_system_preds(struct event_subsystem *system,
 		if (strcmp(call->class->system, system->name) != 0)
 			continue;
 
+		if (call->flags & TRACE_EVENT_FL_NO_SET_FILTER)
+			continue;
+
 		filter_item = kzalloc(sizeof(*filter_item), GFP_KERNEL);
 		if (!filter_item)
 			goto fail_mem;

tip-bot for Li Zefan

author

2011-12-06 06:26 UTC #7 lore raw

In reply to Li Zefan

Commit-ID:  27b14b56af081ec7edeefb3a38b2c9577cc5ef48
Gitweb:     http://git.kernel.org/tip/27b14b56af081ec7edeefb3a38b2c9577cc5ef48
Author:     Li Zefan <[email protected]>
AuthorDate: Tue, 1 Nov 2011 09:09:35 +0800
Committer:  Steven Rostedt <[email protected]>
CommitDate: Mon, 5 Dec 2011 13:28:45 -0500

tracing: Restore system filter behavior

Though not all events have field 'prev_pid', it was allowed to do this:

  # echo 'prev_pid == 100' > events/sched/filter

but commit 75b8e98263fdb0bfbdeba60d4db463259f1fe8a2 (tracing/filter: Swap entire filter of events) broke it without any reason.

Link http://lkml.kernel.org/r/[email protected] open here

Signed-off-by Li Zefan <[email protected]>

Signed-off-by Steven Rostedt <[email protected]>

---
 include/linux/ftrace_event.h       |    2 ++
 kernel/trace/trace_events_filter.c |    7 ++++++-
 2 files changed, 8 insertions(+), 1 deletions(-)

include/linux/ftrace_event.h +2 -0

index 96efa67..c3da42d 100644
@@ -172,6 +172,7 @@ enum {
 	TRACE_EVENT_FL_FILTERED_BIT,
 	TRACE_EVENT_FL_RECORDED_CMD_BIT,
 	TRACE_EVENT_FL_CAP_ANY_BIT,
+	TRACE_EVENT_FL_NO_SET_FILTER_BIT,
 };
 
 enum {

@@ -179,6 +180,7 @@ enum {
 	TRACE_EVENT_FL_FILTERED		= (1 << TRACE_EVENT_FL_FILTERED_BIT),
 	TRACE_EVENT_FL_RECORDED_CMD	= (1 << TRACE_EVENT_FL_RECORDED_CMD_BIT),
 	TRACE_EVENT_FL_CAP_ANY		= (1 << TRACE_EVENT_FL_CAP_ANY_BIT),
+	TRACE_EVENT_FL_NO_SET_FILTER	= (1 << TRACE_EVENT_FL_NO_SET_FILTER_BIT),
 };
 
 struct ftrace_event_call {

kernel/trace/trace_events_filter.c +6 -1

index d6e7926..95dc31e 100644
@@ -1649,7 +1649,9 @@ static int replace_system_preds(struct event_subsystem *system,
 		 */
 		err = replace_preds(call, NULL, ps, filter_string, true);
 		if (err)
-			goto fail;
+			call->flags |= TRACE_EVENT_FL_NO_SET_FILTER;
+		else
+			call->flags &= ~TRACE_EVENT_FL_NO_SET_FILTER;
 	}
 
 	list_for_each_entry(call, &ftrace_events, list) {

@@ -1658,6 +1660,9 @@ static int replace_system_preds(struct event_subsystem *system,
 		if (strcmp(call->class->system, system->name) != 0)
 			continue;
 
+		if (call->flags & TRACE_EVENT_FL_NO_SET_FILTER)
+			continue;
+
 		filter_item = kzalloc(sizeof(*filter_item), GFP_KERNEL);
 		if (!filter_item)
 			goto fail_mem;