Skip to contentSkip to conversation

kprobes: fix trace_probe flags in enable_trace_kprobe

PATCH In review Artem Savkov 4 participants 5 messages Jul 25, 2018 Tracing

lore
Share & tools
lore
#0 lore raw

PATCH 1/1 kprobes: fix trace_probe flags in enable_trace_kprobe

If enable_trace_kprobe fails to enable the probe in enable_k(ret)probe it returns an error, but does not unset the tp flags it set previosly. This results in a probe being considered enabled and failures like being unable to remove the probe through kprobe_events file since probes_open() expects every probe to be disabled.

---
 kernel/trace/trace_kprobe.c | 13 +++++++++++--
 1 file changed, 11 insertions(+), 2 deletions(-)
kernel/trace/trace_kprobe.c +11 -2
index 21f718472942..fb887ced5056 100644
@@ -400,11 +400,10 @@ static struct trace_kprobe *find_trace_kprobe(const char *event,
 static int
 enable_trace_kprobe(struct trace_kprobe *tk, struct trace_event_file *file)
 {
+	struct event_file_link *link;
 	int ret = 0;
 
 	if (file) {
-		struct event_file_link *link;
-
 		link = kmalloc(sizeof(*link), GFP_KERNEL);
 		if (!link) {
 			ret = -ENOMEM;
@@ -424,6 +423,16 @@ enable_trace_kprobe(struct trace_kprobe *tk, struct trace_event_file *file)
 		else
 			ret = enable_kprobe(&tk->rp.kp);
 	}
+
+	if (ret) {
+		if (file) {
+			list_del(&link->list);
+			kfree(link);
+			tk->tp.flags &= ~TP_FLAG_TRACE;
+		} else {
+			tk->tp.flags &= ~TP_FLAG_PROFILE;
+		}
+	}
  out:
 	return ret;
 }
Signature · 1 line
2.13.6
maintainer
#1 lore raw

[ Adding Masami to Cc ]

Quoted 6 lines · On Wed, 25 Jul 2018 12:28:26 +0200 Artem Savkov <[email protected]> wrote:

If enable_trace_kprobe fails to enable the probe in enable_k(ret)probe it returns an error, but does not unset the tp flags it set previosly. This results in a probe being considered enabled and failures like being unable to remove the probe through kprobe_events file since probes_open() expects every probe to be disabled.

Looks good to me. Masami, want to ack it?

-- Steve

Quoted 39 lines
---
 kernel/trace/trace_kprobe.c | 13 +++++++++++--
 1 file changed, 11 insertions(+), 2 deletions(-)
kernel/trace/trace_kprobe.c +11 -2
index 21f718472942..fb887ced5056 100644
@@ -400,11 +400,10 @@ static struct trace_kprobe *find_trace_kprobe(const char *event,
 static int
 enable_trace_kprobe(struct trace_kprobe *tk, struct trace_event_file *file)
 {
+	struct event_file_link *link;
 	int ret = 0;
 
 	if (file) {
-		struct event_file_link *link;
-
 		link = kmalloc(sizeof(*link), GFP_KERNEL);
 		if (!link) {
 			ret = -ENOMEM;
@@ -424,6 +423,16 @@ enable_trace_kprobe(struct trace_kprobe *tk, struct trace_event_file *file)
 		else
 			ret = enable_kprobe(&tk->rp.kp);
 	}
+
+	if (ret) {
+		if (file) {
+			list_del(&link->list);
+			kfree(link);
+			tk->tp.flags &= ~TP_FLAG_TRACE;
+		} else {
+			tk->tp.flags &= ~TP_FLAG_PROFILE;
+		}
+	}
  out:
 	return ret;
 }
maintainer Acked-by
#2 lore raw
Quoted 6 lines · On Wed, 25 Jul 2018 09:21:51 -0400 Steven Rostedt <[email protected]> wrote:

[ Adding Masami to Cc ]

On Wed, 25 Jul 2018 12:28:26 +0200
Artem Savkov <[email protected]> wrote:

Quoted 8 lines
Quoted 6 lines

If enable_trace_kprobe fails to enable the probe in enable_k(ret)probe it returns an error, but does not unset the tp flags it set previosly. This results in a probe being considered enabled and failures like being unable to remove the probe through kprobe_events file since probes_open() expects every probe to be disabled.

Looks good to me. Masami, want to ack it?

Yes, looks good to me too.

Thanks Artem!

Quoted 43 lines

-- Steve

Quoted 39 lines
---
 kernel/trace/trace_kprobe.c | 13 +++++++++++--
 1 file changed, 11 insertions(+), 2 deletions(-)
kernel/trace/trace_kprobe.c +11 -2
index 21f718472942..fb887ced5056 100644
@@ -400,11 +400,10 @@ static struct trace_kprobe *find_trace_kprobe(const char *event,
 static int
 enable_trace_kprobe(struct trace_kprobe *tk, struct trace_event_file *file)
 {
+	struct event_file_link *link;
 	int ret = 0;
 
 	if (file) {
-		struct event_file_link *link;
-
 		link = kmalloc(sizeof(*link), GFP_KERNEL);
 		if (!link) {
 			ret = -ENOMEM;
@@ -424,6 +423,16 @@ enable_trace_kprobe(struct trace_kprobe *tk, struct trace_event_file *file)
 		else
 			ret = enable_kprobe(&tk->rp.kp);
 	}
+
+	if (ret) {
+		if (file) {
+			list_del(&link->list);
+			kfree(link);
+			tk->tp.flags &= ~TP_FLAG_TRACE;
+		} else {
+			tk->tp.flags &= ~TP_FLAG_PROFILE;
+		}
+	}
  out:
 	return ret;
 }
Signature · 1 line
Masami Hiramatsu <[email protected]>
#3 lore raw
In reply to Artem Savkov
Quoted 2 lines · Artem Savkov

If enable_trace_kprobe fails to enable the probe in enable_k(ret)probe it returns an error, but does not unset the tp flags it set previosly.

"previously"

kernel/trace/trace_kprobe.c enable_trace_kprobe(struct trace_kprobe *tk, struct trace_event_file *file) · line 429

Quoted 8 lines
kernel/trace/trace_kprobe.c +4 -0
@@ -424,6 +423,16 @@ enable_trace_kprobe(struct trace_kprobe *tk, struct trace_event_file *file)
 		else
 			ret = enable_kprobe(&tk->rp.kp);
 	}
+
+	if (ret) {
+		if (file) {
+			list_del(&link->list);

Should this be list_del_rcu(), since it was added to the list with list_add_tail_rcu()?

Signature · 1 line
Josh
maintainer
#4 lore raw
Quoted 5 lines · On Wed, 25 Jul 2018 08:56:32 -0500 Josh Poimboeuf <[email protected]> wrote:
Quoted 2 lines · Artem Savkov

If enable_trace_kprobe fails to enable the probe in enable_k(ret)probe it returns an error, but does not unset the tp flags it set previosly.

"previously"

I tentatively added this to my queue. I updated the subject as:

tracing/kprobes: Fix trace_probe flags on enable_trace_kprobe() failure

But yeah, that needs to be fixed.

Quoted 13 lines

kernel/trace/trace_kprobe.c enable_trace_kprobe(struct trace_kprobe *tk, struct trace_event_file *file) · line 429

Quoted 9 lines
kernel/trace/trace_kprobe.c +4 -0
@@ -424,6 +423,16 @@ enable_trace_kprobe(struct trace_kprobe *tk, struct trace_event_file *file)
 		else
 			ret = enable_kprobe(&tk->rp.kp);
 	}
+
+	if (ret) {
+		if (file) {
+			list_del(&link->list);  

Should this be list_del_rcu(), since it was added to the list with list_add_tail_rcu()?

Good catch. Yes, that should be list_del_rcu().

Artem,

Can you send a v2 with the changes.

Thanks,

-- Steve