diff mbox

[v3,1/2] console: Replace #if 0 with atomic var 'ignore_console_lock_warning'

Message ID 20180719101601.25076-2-tzimmermann@suse.de (mailing list archive)
State New, archived
Headers show

Commit Message

Thomas Zimmermann July 19, 2018, 10:16 a.m. UTC
The macro WARN_CONSOLE_UNLOCKED prints a warning when a thread enters
the console's critical section without having acquired the console
lock. The console lock can be ignored when debugging the console using
printk, but this makes WARN_CONSOLE_UNLOCKED generate unnecessary
warnings.

The variable ignore_console_lock_warning temporarily disables
WARN_CONSOLE_UNLOCKED. Developers interested in debugging the console's
critical sections should increment it before entering the CS and
decrement it after leaving the CS. Setting ignore_console_lock_warning
is only for debugging. Regular operation should not manipulate it.

Acknoledgements: This patch is based on an earlier version by Steven
Rostedt. The use of atomic increment/decrement was suggested by Petr
Mladek.

Link: http://lkml.kernel.org/r/717e6337-e7a6-7a92-1c1b-8929a25696b5@suse.de
Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de>
---
 include/linux/console.h | 14 +++++++++-----
 kernel/printk/printk.c  |  3 +++
 2 files changed, 12 insertions(+), 5 deletions(-)

Comments

Petr Mladek July 20, 2018, 9:07 a.m. UTC | #1
On Thu 2018-07-19 12:16:00, Thomas Zimmermann wrote:
> The macro WARN_CONSOLE_UNLOCKED prints a warning when a thread enters
> the console's critical section without having acquired the console
> lock. The console lock can be ignored when debugging the console using
> printk, but this makes WARN_CONSOLE_UNLOCKED generate unnecessary
> warnings.
> 
> The variable ignore_console_lock_warning temporarily disables
> WARN_CONSOLE_UNLOCKED. Developers interested in debugging the console's
> critical sections should increment it before entering the CS and
> decrement it after leaving the CS. Setting ignore_console_lock_warning
> is only for debugging. Regular operation should not manipulate it.
> 
> Acknoledgements: This patch is based on an earlier version by Steven
> Rostedt. The use of atomic increment/decrement was suggested by Petr
> Mladek.
> 
> Link: http://lkml.kernel.org/r/717e6337-e7a6-7a92-1c1b-8929a25696b5@suse.de
> Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de>

Acked-by: Petr Mladek <pmladek@suse.com>

I have vacation the following two weeks. I prefer this version over
the previous one but I could live with both.

Best Regards,
Petr
Bartlomiej Zolnierkiewicz July 31, 2018, 10:45 a.m. UTC | #2
On Thursday, July 19, 2018 12:16:00 PM Thomas Zimmermann wrote:
> The macro WARN_CONSOLE_UNLOCKED prints a warning when a thread enters
> the console's critical section without having acquired the console
> lock. The console lock can be ignored when debugging the console using
> printk, but this makes WARN_CONSOLE_UNLOCKED generate unnecessary
> warnings.
> 
> The variable ignore_console_lock_warning temporarily disables
> WARN_CONSOLE_UNLOCKED. Developers interested in debugging the console's
> critical sections should increment it before entering the CS and
> decrement it after leaving the CS. Setting ignore_console_lock_warning
> is only for debugging. Regular operation should not manipulate it.
> 
> Acknoledgements: This patch is based on an earlier version by Steven
> Rostedt. The use of atomic increment/decrement was suggested by Petr
> Mladek.
> 
> Link: http://lkml.kernel.org/r/717e6337-e7a6-7a92-1c1b-8929a25696b5@suse.de
> Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de>

Patch queued for 4.19 (w/ <asm/atomic.h> -> <linux/atomic.h> fixup), thanks.

Best regards,
--
Bartlomiej Zolnierkiewicz
Samsung R&D Institute Poland
Samsung Electronics
Steven Rostedt July 31, 2018, 3:38 p.m. UTC | #3
On Thu, 19 Jul 2018 12:16:00 +0200
Thomas Zimmermann <tzimmermann@suse.de> wrote:

> The macro WARN_CONSOLE_UNLOCKED prints a warning when a thread enters
> the console's critical section without having acquired the console
> lock. The console lock can be ignored when debugging the console using
> printk, but this makes WARN_CONSOLE_UNLOCKED generate unnecessary
> warnings.
> 
> The variable ignore_console_lock_warning temporarily disables
> WARN_CONSOLE_UNLOCKED. Developers interested in debugging the console's
> critical sections should increment it before entering the CS and
> decrement it after leaving the CS. Setting ignore_console_lock_warning
> is only for debugging. Regular operation should not manipulate it.
> 
> Acknoledgements: This patch is based on an earlier version by Steven
> Rostedt. The use of atomic increment/decrement was suggested by Petr
> Mladek.

You can also add a reference to that patch as well:

Link: http://lkml.kernel.org/r/20180712092938.22e33641@gandalf.local.home

-- Steve

> 
> Link: http://lkml.kernel.org/r/717e6337-e7a6-7a92-1c1b-8929a25696b5@suse.de
> Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de>
> ---
>  include/linux/console.h | 14 +++++++++-----
>  kernel/printk/printk.c  |  3 +++
>  2 files changed, 12 insertions(+), 5 deletions(-)
> 
> diff --git a/include/linux/console.h b/include/linux/console.h
> index f59f3dbca65c..a3307bcab947 100644
> --- a/include/linux/console.h
> +++ b/include/linux/console.h
> @@ -14,6 +14,7 @@
>  #ifndef _LINUX_CONSOLE_H_
>  #define _LINUX_CONSOLE_H_ 1
>  
> +#include <asm/atomic.h>
>  #include <linux/types.h>
>  
>  struct vc_data;
> @@ -201,11 +202,14 @@ void vcs_make_sysfs(int index);
>  void vcs_remove_sysfs(int index);
>  
>  /* Some debug stub to catch some of the obvious races in the VT code */
> -#if 1
> -#define WARN_CONSOLE_UNLOCKED()	WARN_ON(!is_console_locked() && !oops_in_progress)
> -#else
> -#define WARN_CONSOLE_UNLOCKED()
> -#endif
> +#define WARN_CONSOLE_UNLOCKED()						\
> +	WARN_ON(!atomic_read(&ignore_console_lock_warning) &&		\
> +		!is_console_locked() && !oops_in_progress)
> +/*
> + * Increment ignore_console_lock_warning if you need to quiet
> + * WARN_CONSOLE_UNLOCKED() for debugging purposes.
> + */
> +extern atomic_t ignore_console_lock_warning;
>  
>  /* VESA Blanking Levels */
>  #define VESA_NO_BLANKING        0
> diff --git a/kernel/printk/printk.c b/kernel/printk/printk.c
> index 3f041e7cbfc9..7d32a86758cd 100644
> --- a/kernel/printk/printk.c
> +++ b/kernel/printk/printk.c
> @@ -66,6 +66,9 @@ int console_printk[4] = {
>  	CONSOLE_LOGLEVEL_DEFAULT,	/* default_console_loglevel */
>  };
>  
> +atomic_t ignore_console_lock_warning __read_mostly = ATOMIC_INIT(0);
> +EXPORT_SYMBOL(ignore_console_lock_warning);
> +
>  /*
>   * Low level drivers may need that to know if they can schedule in
>   * their unblank() callback or not. So let's export it.
diff mbox

Patch

diff --git a/include/linux/console.h b/include/linux/console.h
index f59f3dbca65c..a3307bcab947 100644
--- a/include/linux/console.h
+++ b/include/linux/console.h
@@ -14,6 +14,7 @@ 
 #ifndef _LINUX_CONSOLE_H_
 #define _LINUX_CONSOLE_H_ 1
 
+#include <asm/atomic.h>
 #include <linux/types.h>
 
 struct vc_data;
@@ -201,11 +202,14 @@  void vcs_make_sysfs(int index);
 void vcs_remove_sysfs(int index);
 
 /* Some debug stub to catch some of the obvious races in the VT code */
-#if 1
-#define WARN_CONSOLE_UNLOCKED()	WARN_ON(!is_console_locked() && !oops_in_progress)
-#else
-#define WARN_CONSOLE_UNLOCKED()
-#endif
+#define WARN_CONSOLE_UNLOCKED()						\
+	WARN_ON(!atomic_read(&ignore_console_lock_warning) &&		\
+		!is_console_locked() && !oops_in_progress)
+/*
+ * Increment ignore_console_lock_warning if you need to quiet
+ * WARN_CONSOLE_UNLOCKED() for debugging purposes.
+ */
+extern atomic_t ignore_console_lock_warning;
 
 /* VESA Blanking Levels */
 #define VESA_NO_BLANKING        0
diff --git a/kernel/printk/printk.c b/kernel/printk/printk.c
index 3f041e7cbfc9..7d32a86758cd 100644
--- a/kernel/printk/printk.c
+++ b/kernel/printk/printk.c
@@ -66,6 +66,9 @@  int console_printk[4] = {
 	CONSOLE_LOGLEVEL_DEFAULT,	/* default_console_loglevel */
 };
 
+atomic_t ignore_console_lock_warning __read_mostly = ATOMIC_INIT(0);
+EXPORT_SYMBOL(ignore_console_lock_warning);
+
 /*
  * Low level drivers may need that to know if they can schedule in
  * their unblank() callback or not. So let's export it.