tty_init_dev: 24 callbacks suppressed

PATCH In review Borislav Petkov 6 participants 20 messages Oct 4, 2012 – Oct 5, 2012 TTY layer and serial drivers

lore

Share & tools

lore

Borislav Petkov

author

2012-10-04 09:21 UTC #0 lore raw

Hi,

I'm seeing this on today's Linus tree:

[   24.048278] tty_init_dev: 24 callbacks suppressed
[   45.630349] tty_init_dev: 3 callbacks suppressed

It is either from that WARN_RATELIMIT thing or the printk_ratelimited further below in tty_init_dev but I don't know for sure because the actual text message from both printk's doesn't come out in dmesg - only that something got suppressed.

And it quiets down later, after the machine has finished booting. Still, this doesn't tell me anything about any issue. So what's up?

Thanks.

Signature · 8 lines

Regards/Gruss,
Boris.

Advanced Micro Devices GmbH
Einsteinring 24, 85609 Dornach
GM: Alberto Bozzo
Reg: Dornach, Landkreis Muenchen
HRB Nr. 43632 WEEE Registernr: 129 19551

Markus Trippelsdorf

2012-10-04 11:23 UTC #1 lore raw

Quoted 7 lines · Borislav Petkov

Hi,

I'm seeing this on today's Linus tree:

[   24.048278] tty_init_dev: 24 callbacks suppressed
[   45.630349] tty_init_dev: 3 callbacks suppressed

Quoted 7 lines

It is either from that WARN_RATELIMIT thing or the printk_ratelimited further below in tty_init_dev but I don't know for sure because the actual text message from both printk's doesn't come out in dmesg - only that something got suppressed.
And it quiets down later, after the machine has finished booting. Still, this doesn't tell me anything about any issue. So what's up?

I'm seeing the same thing.
This can be fixed by a slightly modified version of:
https://patchwork.kernel.org/patch/1339221/

include/linux/ratelimit.h +9 ~~-18~~

index e11ccb4..2d02461 100644
@@ -46,20 +46,17 @@ extern int ___ratelimit(struct ratelimit_state *rs, const char *func);
 #define WARN_ON_RATELIMIT(condition, state)			\
 		WARN_ON((condition) && __ratelimit(state))
 
-#define __WARN_RATELIMIT(condition, state, format...)		\
-({								\
-	int rtn = 0;						\
-	if (unlikely(__ratelimit(state)))			\
-		rtn = WARN(condition, format);			\
-	rtn;							\
-})
-
-#define WARN_RATELIMIT(condition, format...)			\
+#define WARN_RATELIMIT(condition, state, fmt, ...)			\
 ({								\
 	static DEFINE_RATELIMIT_STATE(_rs,			\
 				      DEFAULT_RATELIMIT_INTERVAL,	\
 				      DEFAULT_RATELIMIT_BURST);	\
-	__WARN_RATELIMIT(condition, &_rs, format);		\
+	int rtn = !!(condition);				\
+								\
+	if (unlikely(rtn && __ratelimit(state)))		\
+		WARN(rtn, fmt, ##__VA_ARGS__);			\
+								\
+	rtn;							\
 })
 
 #else

@@ -67,15 +64,9 @@ extern int ___ratelimit(struct ratelimit_state *rs, const char *func);
 #define WARN_ON_RATELIMIT(condition, state)			\
 	WARN_ON(condition)
 
-#define __WARN_RATELIMIT(condition, state, format...)		\
-({								\
-	int rtn = WARN(condition, format);			\
-	rtn;							\
-})
-
-#define WARN_RATELIMIT(condition, format...)			\
+#define WARN_RATELIMIT(condition, fmt, ...)			\
 ({								\
-	int rtn = WARN(condition, format);			\
+	int rtn = WARN(condition, fmt, ##__VA_ARGS__);		\
 	rtn;							\
 })

Signature · 1 line

Markus

Markus Trippelsdorf

2012-10-04 11:52 UTC #2 lore raw

Quoted 13 lines · Markus Trippelsdorf

Quoted 12 lines · Borislav Petkov
Hi,
I'm seeing this on today's Linus tree:
[   24.048278] tty_init_dev: 24 callbacks suppressed
[   45.630349] tty_init_dev: 3 callbacks suppressed
It is either from that WARN_RATELIMIT thing or the printk_ratelimited further below in tty_init_dev but I don't know for sure because the actual text message from both printk's doesn't come out in dmesg - only that something got suppressed.

Quoted 6 lines

Quoted 2 lines
And it quiets down later, after the machine has finished booting. Still, this doesn't tell me anything about any issue. So what's up?
I'm seeing the same thing.
This can be fixed by a slightly modified version of:
https://patchwork.kernel.org/patch/1339221/

My first patch was wrong. This one should be correct:

include/linux/ratelimit.h +9 ~~-18~~

index e11ccb4..d8de255 100644
@@ -46,20 +46,17 @@ extern int ___ratelimit(struct ratelimit_state *rs, const char *func);
 #define WARN_ON_RATELIMIT(condition, state)			\
 		WARN_ON((condition) && __ratelimit(state))
 
-#define __WARN_RATELIMIT(condition, state, format...)		\
-({								\
-	int rtn = 0;						\
-	if (unlikely(__ratelimit(state)))			\
-		rtn = WARN(condition, format);			\
-	rtn;							\
-})
-
-#define WARN_RATELIMIT(condition, format...)			\
+#define WARN_RATELIMIT(condition, fmt, ...)			\
 ({								\
 	static DEFINE_RATELIMIT_STATE(_rs,			\
 				      DEFAULT_RATELIMIT_INTERVAL,	\
 				      DEFAULT_RATELIMIT_BURST);	\
-	__WARN_RATELIMIT(condition, &_rs, format);		\
+	int rtn = !!(condition);				\
+								\
+	if (unlikely(rtn && __ratelimit(&_rs)))		\
+		WARN(rtn, fmt, ##__VA_ARGS__);			\
+								\
+	rtn;							\
 })
 
 #else

@@ -67,15 +64,9 @@ extern int ___ratelimit(struct ratelimit_state *rs, const char *func);
 #define WARN_ON_RATELIMIT(condition, state)			\
 	WARN_ON(condition)
 
-#define __WARN_RATELIMIT(condition, state, format...)		\
-({								\
-	int rtn = WARN(condition, format);			\
-	rtn;							\
-})
-
-#define WARN_RATELIMIT(condition, format...)			\
+#define WARN_RATELIMIT(condition, fmt, ...)			\
 ({								\
-	int rtn = WARN(condition, format);			\
+	int rtn = WARN(condition, fmt, ##__VA_ARGS__);		\
 	rtn;							\
 })

Signature · 1 line

Markus

Borislav Petkov

author

2012-10-04 12:40 UTC #3 lore raw

Quoted 20 lines · Markus Trippelsdorf

include/linux/ratelimit.h +1 -9

index e11ccb4..d8de255 100644
@@ -46,20 +46,17 @@ extern int ___ratelimit(struct ratelimit_state *rs, const char *func);
 #define WARN_ON_RATELIMIT(condition, state)			\
 		WARN_ON((condition) && __ratelimit(state))
 
-#define __WARN_RATELIMIT(condition, state, format...)		\
-({								\
-	int rtn = 0;						\
-	if (unlikely(__ratelimit(state)))			\
-		rtn = WARN(condition, format);			\
-	rtn;							\
-})
-
-#define WARN_RATELIMIT(condition, format...)			\
+#define WARN_RATELIMIT(condition, fmt, ...)			\
 ({								\
 	static DEFINE_RATELIMIT_STATE(_rs,			\

include/linux/ratelimit.h extern int ___ratelimit(struct ratelimit_state *rs, const char *func); · line 60

Quoted 10 lines

include/linux/ratelimit.h +6 -1

 				      DEFAULT_RATELIMIT_INTERVAL,	\
 				      DEFAULT_RATELIMIT_BURST);	\
-	__WARN_RATELIMIT(condition, &_rs, format);		\
+	int rtn = !!(condition);				\
+								\
+	if (unlikely(rtn && __ratelimit(&_rs)))		\
+		WARN(rtn, fmt, ##__VA_ARGS__);			\
+								\
+	rtn;							\
 })

Aha, I see it. We need to look at the condition before the __ratelimit, otherwise we WARN unnecessarily, good catch.

include/linux/ratelimit.h extern int ___ratelimit(struct ratelimit_state *rs, const char *func); · line 67

Quoted 13 lines

include/linux/ratelimit.h +1 -7

 #else

@@ -67,15 +64,9 @@ extern int ___ratelimit(struct ratelimit_state *rs, const char *func);
 #define WARN_ON_RATELIMIT(condition, state)			\
 	WARN_ON(condition)
 
-#define __WARN_RATELIMIT(condition, state, format...)		\
-({								\
-	int rtn = WARN(condition, format);			\
-	rtn;							\
-})
-
-#define WARN_RATELIMIT(condition, format...)			\
+#define WARN_RATELIMIT(condition, fmt, ...)			\

... except this change is unrelated and unneeded - there's enough room in 80 cols to leave it as "format" instead of shortening it.

Other than that:

Acked-and-tested-by Borislav Petkov <[email protected]>

Thanks.

Signature · 8 lines

Regards/Gruss,
Boris.

Advanced Micro Devices GmbH
Einsteinring 24, 85609 Dornach
GM: Alberto Bozzo
Reg: Dornach, Landkreis Muenchen
HRB Nr. 43632 WEEE Registernr: 129 19551

Markus Trippelsdorf

2012-10-04 13:11 UTC #4 lore raw

Quoted 52 lines · Borislav Petkov

Quoted 20 lines · Markus Trippelsdorf

include/linux/ratelimit.h +1 -9

index e11ccb4..d8de255 100644
@@ -46,20 +46,17 @@ extern int ___ratelimit(struct ratelimit_state *rs, const char *func);
 #define WARN_ON_RATELIMIT(condition, state)			\
 		WARN_ON((condition) && __ratelimit(state))
 
-#define __WARN_RATELIMIT(condition, state, format...)		\
-({								\
-	int rtn = 0;						\
-	if (unlikely(__ratelimit(state)))			\
-		rtn = WARN(condition, format);			\
-	rtn;							\
-})
-
-#define WARN_RATELIMIT(condition, format...)			\
+#define WARN_RATELIMIT(condition, fmt, ...)			\
 ({								\
 	static DEFINE_RATELIMIT_STATE(_rs,			\

include/linux/ratelimit.h extern int ___ratelimit(struct ratelimit_state *rs, const char *func); · line 60

Quoted 10 lines

include/linux/ratelimit.h +6 -1

 				      DEFAULT_RATELIMIT_INTERVAL,	\
 				      DEFAULT_RATELIMIT_BURST);	\
-	__WARN_RATELIMIT(condition, &_rs, format);		\
+	int rtn = !!(condition);				\
+								\
+	if (unlikely(rtn && __ratelimit(&_rs)))		\
+		WARN(rtn, fmt, ##__VA_ARGS__);			\
+								\
+	rtn;							\
 })

Aha, I see it. We need to look at the condition before the __ratelimit, otherwise we WARN unnecessarily, good catch.

include/linux/ratelimit.h extern int ___ratelimit(struct ratelimit_state *rs, const char *func); · line 67

Quoted 13 lines

include/linux/ratelimit.h +1 -7

 #else

@@ -67,15 +64,9 @@ extern int ___ratelimit(struct ratelimit_state *rs, const char *func);
 #define WARN_ON_RATELIMIT(condition, state)			\
 	WARN_ON(condition)
 
-#define __WARN_RATELIMIT(condition, state, format...)		\
-({								\
-	int rtn = WARN(condition, format);			\
-	rtn;							\
-})
-
-#define WARN_RATELIMIT(condition, format...)			\
+#define WARN_RATELIMIT(condition, fmt, ...)			\

... except this change is unrelated and unneeded - there's enough room in 80 cols to leave it as "format" instead of shortening it.

Quoted 3 lines

Other than that:
Acked-and-tested-by Borislav Petkov <[email protected]>

I'll let Jiri handle this :). It's his patch anyway.

Signature · 1 line

Markus

Jiri Slaby

2012-10-05 11:17 UTC #5 lore raw

CCing Joe.

Quoted 53 lines · Markus Trippelsdorf

Quoted 52 lines · Borislav Petkov
Quoted 20 lines · Markus Trippelsdorf
include/linux/ratelimit.h +1 -9
index e11ccb4..d8de255 100644
@@ -46,20 +46,17 @@ extern int ___ratelimit(struct ratelimit_state *rs, const char *func);
 #define WARN_ON_RATELIMIT(condition, state)			\
 		WARN_ON((condition) && __ratelimit(state))
 
-#define __WARN_RATELIMIT(condition, state, format...)		\
-({								\
-	int rtn = 0;						\
-	if (unlikely(__ratelimit(state)))			\
-		rtn = WARN(condition, format);			\
-	rtn;							\
-})
-
-#define WARN_RATELIMIT(condition, format...)			\
+#define WARN_RATELIMIT(condition, fmt, ...)			\
 ({								\
 	static DEFINE_RATELIMIT_STATE(_rs,			\
include/linux/ratelimit.h extern int ___ratelimit(struct ratelimit_state *rs, const char *func); · line 60
Quoted 10 lines
include/linux/ratelimit.h +6 -1
 				      DEFAULT_RATELIMIT_INTERVAL,	\
 				      DEFAULT_RATELIMIT_BURST);	\
-	__WARN_RATELIMIT(condition, &_rs, format);		\
+	int rtn = !!(condition);				\
+								\
+	if (unlikely(rtn && __ratelimit(&_rs)))		\
+		WARN(rtn, fmt, ##__VA_ARGS__);			\
+								\
+	rtn;							\
 })
Aha, I see it. We need to look at the condition before the __ratelimit, otherwise we WARN unnecessarily, good catch.
include/linux/ratelimit.h extern int ___ratelimit(struct ratelimit_state *rs, const char *func); · line 67
Quoted 13 lines
include/linux/ratelimit.h +1 -7
 #else
@@ -67,15 +64,9 @@ extern int ___ratelimit(struct ratelimit_state *rs, const char *func);
 #define WARN_ON_RATELIMIT(condition, state)			\
 	WARN_ON(condition)
 
-#define __WARN_RATELIMIT(condition, state, format...)		\
-({								\
-	int rtn = WARN(condition, format);			\
-	rtn;							\
-})
-
-#define WARN_RATELIMIT(condition, format...)			\
+#define WARN_RATELIMIT(condition, fmt, ...)			\
... except this change is unrelated and unneeded - there's enough room in 80 cols to leave it as "format" instead of shortening it.

Quoted 5 lines

Quoted 3 lines
Other than that:
Acked-and-tested-by Borislav Petkov <[email protected]>
I'll let Jiri handle this :). It's his patch anyway.

Actually this is Joe's version of the patch. Joe, people started hitting the bug [1]. Could you resend your patch?

[1] https://patchwork.kernel.org/patch/1339221/

BTW what scares me that nobody noticed that bug until this is in the Linus's tree. Do people use -next at all or am I the only one user? (I didn't hit it as I have the patch in my local queue.)

thanks,

Signature · 2 lines

js
suse labs

Alan Cox

2012-10-05 11:21 UTC #6 lore raw

Quoted 8 lines

Actually this is Joe's version of the patch. Joe, people started hitting the bug [1]. Could you resend your patch?
[1] https://patchwork.kernel.org/patch/1339221/
BTW what scares me that nobody noticed that bug until this is in the Linus's tree. Do people use -next at all or am I the only one user? (I didn't hit it as I have the patch in my local queue.)

I run -next for various things. I'd noticed that funny but you'd started investigating it before I got to investigating.

Alan

Borislav Petkov

author

2012-10-05 12:27 UTC #7 lore raw

In reply to Jiri Slaby

Quoted 3 lines · Jiri Slaby

BTW what scares me that nobody noticed that bug until this is in the Linus's tree. Do people use -next at all or am I the only one user? (I didn't hit it as I have the patch in my local queue.)

Noticing this before -rc1 and fixing it around that time is still fine in my book. So can *someone* *please* send a patch already so that we can forget about this.

Thanks.

Signature · 8 lines

Regards/Gruss,
Boris.

Advanced Micro Devices GmbH
Einsteinring 24, 85609 Dornach
GM: Alberto Bozzo
Reg: Dornach, Landkreis Muenchen
HRB Nr. 43632 WEEE Registernr: 129 19551

Markus Trippelsdorf

2012-10-05 12:57 UTC #8 lore raw

PATCH 1/1 Fix bogus "callbacks suppressed" messages

On the current git tree one sees messages such as:
tty_init_dev: 24 callbacks suppressed
tty_init_dev: 3 callbacks suppressed

To fix this we need to look at condition before calling __ratelimit in the WARN_RATELIMIT macro. While at it remove the superfluous __WARN_RATELIMIT macros.

Original patch is from Joe Perches and Jiri Slaby.

Signed-off-by Markus Trippelsdorf <[email protected]>

Acked-and-tested-by Borislav Petkov <[email protected]>

---
 include/linux/ratelimit.h | 27 +++++++++------------------
 1 file changed, 9 insertions(+), 18 deletions(-)

include/linux/ratelimit.h +9 ~~-18~~

index e11ccb4..0a260d8 100644
@@ -46,20 +46,17 @@ extern int ___ratelimit(struct ratelimit_state *rs, const char *func);
 #define WARN_ON_RATELIMIT(condition, state)			\
 		WARN_ON((condition) && __ratelimit(state))
 
-#define __WARN_RATELIMIT(condition, state, format...)		\
-({								\
-	int rtn = 0;						\
-	if (unlikely(__ratelimit(state)))			\
-		rtn = WARN(condition, format);			\
-	rtn;							\
-})
-
-#define WARN_RATELIMIT(condition, format...)			\
+#define WARN_RATELIMIT(condition, format, ...)			\
 ({								\
 	static DEFINE_RATELIMIT_STATE(_rs,			\
 				      DEFAULT_RATELIMIT_INTERVAL,	\
 				      DEFAULT_RATELIMIT_BURST);	\
-	__WARN_RATELIMIT(condition, &_rs, format);		\
+	int rtn = !!(condition);				\
+								\
+	if (unlikely(rtn && __ratelimit(&_rs)))			\
+		WARN(rtn, format, ##__VA_ARGS__);		\
+								\
+	rtn;							\
 })
 
 #else

@@ -67,15 +64,9 @@ extern int ___ratelimit(struct ratelimit_state *rs, const char *func);
 #define WARN_ON_RATELIMIT(condition, state)			\
 	WARN_ON(condition)
 
-#define __WARN_RATELIMIT(condition, state, format...)		\
-({								\
-	int rtn = WARN(condition, format);			\
-	rtn;							\
-})
-
-#define WARN_RATELIMIT(condition, format...)			\
+#define WARN_RATELIMIT(condition, format, ...)			\
 ({								\
-	int rtn = WARN(condition, format);			\
+	int rtn = WARN(condition, format, ##__VA_ARGS__);	\
 	rtn;							\
 })

Signature · 1 line

Markus

Greg Kroah-Hartman

maintainer

2012-10-05 14:26 UTC #9 lore raw

Re: [PATCH] Fix bogus "callbacks suppressed" messages

Quoted 8 lines · Markus Trippelsdorf

On the current git tree one sees messages such as:
tty_init_dev: 24 callbacks suppressed
tty_init_dev: 3 callbacks suppressed
To fix this we need to look at condition before calling __ratelimit in the WARN_RATELIMIT macro. While at it remove the superfluous __WARN_RATELIMIT macros.

Quoted 7 lines

Original patch is from Joe Perches and Jiri Slaby.
Signed-off-by Markus Trippelsdorf <[email protected]>
Acked-and-tested-by Borislav Petkov <[email protected]>
---
 include/linux/ratelimit.h | 27 +++++++++------------------
 1 file changed, 9 insertions(+), 18 deletions(-)

I don't have a problem with this patch, but I don't understand why it's now showing up. There haven't been any changes in the ratelimit.h area recently that I can see, so why is this change needed now? What is in the tty layer that is causing this, just the fact that it's actually being used now?

thanks,

greg k-h

Markus Trippelsdorf

2012-10-05 15:28 UTC #10 lore raw

Re: [PATCH] Fix bogus "callbacks suppressed" messages

Quoted 17 lines · Greg Kroah-Hartman

Quoted 8 lines · Markus Trippelsdorf
On the current git tree one sees messages such as:
tty_init_dev: 24 callbacks suppressed
tty_init_dev: 3 callbacks suppressed
To fix this we need to look at condition before calling __ratelimit in the WARN_RATELIMIT macro. While at it remove the superfluous __WARN_RATELIMIT macros.
Quoted 7 lines
Original patch is from Joe Perches and Jiri Slaby.
Signed-off-by Markus Trippelsdorf <[email protected]>
Acked-and-tested-by Borislav Petkov <[email protected]>
---
 include/linux/ratelimit.h | 27 +++++++++------------------
 1 file changed, 9 insertions(+), 18 deletions(-)

Quoted 5 lines

I don't have a problem with this patch, but I don't understand why it's now showing up. There haven't been any changes in the ratelimit.h area recently that I can see, so why is this change needed now? What is in the tty layer that is causing this, just the fact that it's actually being used now?

See Jiri's recent commit:

commit 5d4121c04b3577e37e389b3553d442f44bb346d7

Author: Jiri Slaby <[email protected]>
Date:   Fri Aug 17 14:27:52 2012 +0200

TTY: check if tty->port is assigned

And if not, complain loudly. None in-kernel module should trigger that, but let us find out for sure. On the other hand, all the out-of-tree modules will hit that. Give them some time (maybe one release) to catch up.

Signed-off-by Jiri Slaby <[email protected]>

Signed-off-by Greg Kroah-Hartman <[email protected]>

drivers/tty/tty_io.c +4 -0

index 28c3e86..41e42f1 100644
@@ -1415,6 +1415,10 @@ struct tty_struct *tty_init_dev(struct tty_driver *driver, int idx)
 	if (!tty->port)
 		tty->port = driver->ports[idx];
 
+	WARN_RATELIMIT(!tty->port,
+			"%s: %s driver does not set tty->port. This will crash the kernel later. Fix the driver!\n",
+			__func__, tty->driver->name);
+
 	/*
 	 * Structures all installed ... call the ldisc open routines.
 	 * If we fail here just call release_tty to clean up.  No need

Signature · 1 line

Markus

Borislav Petkov

author

2012-10-05 15:30 UTC #11 lore raw

Re: [PATCH] Fix bogus "callbacks suppressed" messages

In reply to Greg Kroah-Hartman

Quoted 7 lines · Greg Kroah-Hartman

I don't have a problem with this patch, but I don't understand why it's now showing up. There haven't been any changes in the ratelimit.h area recently that I can see, so why is this change needed now? What is in the tty layer that is causing this, just the fact that it's actually being used now?
From my quick semi-skilled git history browsing, I'd say it's

5d4121c04b357 which added the WARN_RATELIMIT to tty_init_dev during the current merge window.

Signature · 8 lines

Regards/Gruss,
Boris.

Advanced Micro Devices GmbH
Einsteinring 24, 85609 Dornach
GM: Alberto Bozzo
Reg: Dornach, Landkreis Muenchen
HRB Nr. 43632 WEEE Registernr: 129 19551

Greg Kroah-Hartman

maintainer

2012-10-05 15:37 UTC #12 lore raw

Re: [PATCH] Fix bogus "callbacks suppressed" messages

Quoted 10 lines · Borislav Petkov

Quoted 7 lines · Greg Kroah-Hartman
I don't have a problem with this patch, but I don't understand why it's now showing up. There haven't been any changes in the ratelimit.h area recently that I can see, so why is this change needed now? What is in the tty layer that is causing this, just the fact that it's actually being used now?
From my quick semi-skilled git history browsing, I'd say it's
5d4121c04b357 which added the WARN_RATELIMIT to tty_init_dev during the current merge window.

So WARN_RATELIMIT was never working properly? If so, how far back does it go in kernel releases that this should be fixed?

thanks,

greg k-h

Markus Trippelsdorf

2012-10-05 15:41 UTC #13 lore raw

Re: [PATCH] Fix bogus "callbacks suppressed" messages

Quoted 8 lines · Greg Kroah-Hartman

Quoted 7 lines · Borislav Petkov
Quoted 5 lines · Greg Kroah-Hartman
I don't have a problem with this patch, but I don't understand why it's now showing up. There haven't been any changes in the ratelimit.h area recently that I can see, so why is this change needed now? What is in the tty layer that is causing this, just the fact that it's actually being used now?

Quoted 6 lines

Quoted 3 lines
Quoted 1 line
From my quick semi-skilled git history browsing, I'd say it's
5d4121c04b357 which added the WARN_RATELIMIT to tty_init_dev during the current merge window.
So WARN_RATELIMIT was never working properly? If so, how far back does it go in kernel releases that this should be fixed?

The only user until this merge window was net/core/filter.c. The WARN_RATELIMIT is used there since v3.0.

Signature · 1 line

Markus

Borislav Petkov

author

2012-10-05 15:43 UTC #14 lore raw

Re: [PATCH] Fix bogus "callbacks suppressed" messages

In reply to Greg Kroah-Hartman

Quoted 2 lines · Greg Kroah-Hartman

So WARN_RATELIMIT was never working properly? If so, how far back does it go in kernel releases that this should be fixed?

Since b3eec79b0776e which added it in May 2011. But the only one other user is net/core/filter.c.

Signature · 8 lines

Regards/Gruss,
Boris.

Advanced Micro Devices GmbH
Einsteinring 24, 85609 Dornach
GM: Alberto Bozzo
Reg: Dornach, Landkreis Muenchen
HRB Nr. 43632 WEEE Registernr: 129 19551

Markus Trippelsdorf

2012-10-05 15:48 UTC #15 lore raw

Re: [PATCH] Fix bogus "callbacks suppressed" messages

Quoted 6 lines · Borislav Petkov

Quoted 2 lines · Greg Kroah-Hartman
So WARN_RATELIMIT was never working properly? If so, how far back does it go in kernel releases that this should be fixed?
Since b3eec79b0776e which added it in May 2011. But the only one othe r user is net/core/filter.c.

But it doesn't matter, because the WARN_RATELIMIT in net/core/filter.c is guarded by a switch statement and uses WARN_RATELIMIT(1,...). So it could never trigger the bug.

Signature · 1 line

Markus

Borislav Petkov

author

2012-10-05 16:03 UTC #16 lore raw

Re: [PATCH] Fix bogus "callbacks suppressed" messages

Quoted 11 lines · Markus Trippelsdorf

Quoted 6 lines · Borislav Petkov
Quoted 2 lines · Greg Kroah-Hartman
So WARN_RATELIMIT was never working properly? If so, how far back does it go in kernel releases that this should be fixed?
Since b3eec79b0776e which added it in May 2011. But the only one othe r user is net/core/filter.c.
But it doesn't matter, because the WARN_RATELIMIT in net/core/filter.c is guarded by a switch statement and uses WARN_RATELIMIT(1,...). So it could never trigger the bug.

Yes, but you should say "it covers the bug with a brown paper bag because the condition is always true."

:-)

Signature · 8 lines

Regards/Gruss,
Boris.

Advanced Micro Devices GmbH
Einsteinring 24, 85609 Dornach
GM: Alberto Bozzo
Reg: Dornach, Landkreis Muenchen
HRB Nr. 43632 WEEE Registernr: 129 19551

Greg Kroah-Hartman

maintainer

2012-10-05 16:06 UTC #17 lore raw

Re: [PATCH] Fix bogus "callbacks suppressed" messages

In reply to Markus Trippelsdorf

Quoted 11 lines · Markus Trippelsdorf

Quoted 6 lines · Borislav Petkov
Quoted 2 lines · Greg Kroah-Hartman
So WARN_RATELIMIT was never working properly? If so, how far back does it go in kernel releases that this should be fixed?
Since b3eec79b0776e which added it in May 2011. But the only one othe r user is net/core/filter.c.
But it doesn't matter, because the WARN_RATELIMIT in net/core/filter.c is guarded by a switch statement and uses WARN_RATELIMIT(1,...). So it could never trigger the bug.

Ok, thanks for digging that all up, I'll just merge this in for 3.7.

greg k-h

Jiri Slaby

Acked-by on v1 1/1

2012-10-05 18:06 UTC #18 lore raw

Re: [PATCH] Fix bogus "callbacks suppressed" messages

In reply to Markus Trippelsdorf

Quoted 4 lines · Markus Trippelsdorf

On the current git tree one sees messages such as:
tty_init_dev: 24 callbacks suppressed
tty_init_dev: 3 callbacks suppressed

Quoted 8 lines

To fix this we need to look at condition before calling __ratelimit in the WARN_RATELIMIT macro. While at it remove the superfluous __WARN_RATELIMIT macros.
Original patch is from Joe Perches and Jiri Slaby.
Signed-off-by Markus Trippelsdorf <[email protected]>
Acked-and-tested-by Borislav Petkov <[email protected]>

Acked-by Jiri Slaby <[email protected]>

Thanks.

Signature · 2 lines

js
suse labs

Joe Perches

2012-10-05 15:33 UTC #19 lore raw

In reply to Jiri Slaby

Quoted 1 line · Jiri Slaby

CCing Joe.

[]

Quoted 6 lines

Quoted 1 line
I'll let Jiri handle this :). It's his patch anyway.
Actually this is Joe's version of the patch. Joe, people started hitting the bug [1]. Could you resend your patch?
[1] https://patchwork.kernel.org/patch/1339221/

Markus already did that.

I think it'd be fine if someone picked it up.

Quoted 3 lines

BTW what scares me that nobody noticed that bug until this is in the Linus's tree. Do people use -next at all or am I the only one user? (I didn't hit it as I have the patch in my local queue.)

I think you're the only actual user.

Does anyone else really use it as more than a tree integration compilation testbed?