[PATCH] x86: Use preempt_disable_notrace() in cycles_2_ns()

[PATCH] x86: Use preempt_disable_notrace() in cycles_2_ns()

[Steven Rostedt]

When debug preempt is enabled, preempt_disable() can be traced by
function and function graph tracing.

There's a place in the function graph tracer that calls trace_clock()
which eventually calls cycles_2_ns() outside of the recursion
protection. When cycles_2_ns() calls preempt_disable() it gets traced
and the graph tracer will go into a recursive loop causing a crash or
worse, a triple fault.

Simple fix is to use preempt_disable_notrace() in cycles_2_ns, which
makes sense because the preempt_disable() tracing may use that code
too, and it tracing it, even with recursion protection is rather
pointless.

Signed-off-by: Steven Rostedt <rostedt@goodmis.org>

diff --git a/arch/x86/kernel/tsc.c b/arch/x86/kernel/tsc.c
index 19e5adb..acb3b60 100644
--- a/arch/x86/kernel/tsc.c
+++ b/arch/x86/kernel/tsc.c
@@ -209,7 +209,7 @@ static inline unsigned long long cycles_2_ns(unsigned long long cyc)
 	 * dance when its actually needed.
 	 */
 
-	preempt_disable();
+	preempt_disable_notrace();
 	data = this_cpu_read(cyc2ns.head);
 	tail = this_cpu_read(cyc2ns.tail);
 
@@ -229,7 +229,7 @@ static inline unsigned long long cycles_2_ns(unsigned long long cyc)
 		if (!--data->__count)
 			this_cpu_write(cyc2ns.tail, data);
 	}
-	preempt_enable();
+	preempt_enable_notrace();
 
 	return ns;
 }

Re: [PATCH] x86: Use preempt_disable_notrace() in cycles_2_ns()

[Peter Zijlstra]

On Tue, Feb 04, 2014 at 02:13:15PM -0500, Steven Rostedt wrote:
> When debug preempt is enabled, preempt_disable() can be traced by
> function and function graph tracing.
> 
> There's a place in the function graph tracer that calls trace_clock()
> which eventually calls cycles_2_ns() outside of the recursion
> protection. When cycles_2_ns() calls preempt_disable() it gets traced
> and the graph tracer will go into a recursive loop causing a crash or
> worse, a triple fault.
> 
> Simple fix is to use preempt_disable_notrace() in cycles_2_ns, which
> makes sense because the preempt_disable() tracing may use that code
> too, and it tracing it, even with recursion protection is rather
> pointless.
> 
> Signed-off-by: Steven Rostedt <rostedt@goodmis.org>

Indeed, thanks!

Re: [PATCH] x86: Use preempt_disable_notrace() in cycles_2_ns()

[Thomas Gleixner]

On Tue, 4 Feb 2014, Steven Rostedt wrote:

> When debug preempt is enabled, preempt_disable() can be traced by
> function and function graph tracing.
> 
> There's a place in the function graph tracer that calls trace_clock()
> which eventually calls cycles_2_ns() outside of the recursion
> protection. When cycles_2_ns() calls preempt_disable() it gets traced
> and the graph tracer will go into a recursive loop causing a crash or
> worse, a triple fault.
> 
> Simple fix is to use preempt_disable_notrace() in cycles_2_ns, which
> makes sense because the preempt_disable() tracing may use that code
> too, and it tracing it, even with recursion protection is rather
> pointless.
> 
> Signed-off-by: Steven Rostedt <rostedt@goodmis.org>

Acked-by-me

> diff --git a/arch/x86/kernel/tsc.c b/arch/x86/kernel/tsc.c
> index 19e5adb..acb3b60 100644
> --- a/arch/x86/kernel/tsc.c
> +++ b/arch/x86/kernel/tsc.c
> @@ -209,7 +209,7 @@ static inline unsigned long long cycles_2_ns(unsigned long long cyc)
>  	 * dance when its actually needed.
>  	 */
>  
> -	preempt_disable();
> +	preempt_disable_notrace();
>  	data = this_cpu_read(cyc2ns.head);
>  	tail = this_cpu_read(cyc2ns.tail);
>  
> @@ -229,7 +229,7 @@ static inline unsigned long long cycles_2_ns(unsigned long long cyc)
>  		if (!--data->__count)
>  			this_cpu_write(cyc2ns.tail, data);
>  	}
> -	preempt_enable();
> +	preempt_enable_notrace();
>  
>  	return ns;
>  }
> 

[tip:x86/urgent] x86: Use preempt_disable_notrace() in cycles_2_ns()

[tip-bot for Steven Rostedt]

Commit-ID:  569d6557ab957d6ae7e97a46ae669174be4189e6
Gitweb:     http://git.kernel.org/tip/569d6557ab957d6ae7e97a46ae669174be4189e6
Author:     Steven Rostedt <rostedt@goodmis.org>
AuthorDate: Tue, 4 Feb 2014 14:13:15 -0500
Committer:  Ingo Molnar <mingo@kernel.org>
CommitDate: Sun, 9 Feb 2014 13:09:08 +0100

x86: Use preempt_disable_notrace() in cycles_2_ns()

When debug preempt is enabled, preempt_disable() can be traced by
function and function graph tracing.

There's a place in the function graph tracer that calls trace_clock()
which eventually calls cycles_2_ns() outside of the recursion
protection. When cycles_2_ns() calls preempt_disable() it gets traced
and the graph tracer will go into a recursive loop causing a crash or
worse, a triple fault.

Simple fix is to use preempt_disable_notrace() in cycles_2_ns, which
makes sense because the preempt_disable() tracing may use that code
too, and it tracing it, even with recursion protection is rather
pointless.

Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Acked-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Peter Zijlstra <peterz@infradead.org>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Link: http://lkml.kernel.org/r/20140204141315.2a968a72@gandalf.local.home
Signed-off-by: Ingo Molnar <mingo@kernel.org>
---
 arch/x86/kernel/tsc.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/arch/x86/kernel/tsc.c b/arch/x86/kernel/tsc.c
index 19e5adb..acb3b60 100644
--- a/arch/x86/kernel/tsc.c
+++ b/arch/x86/kernel/tsc.c
@@ -209,7 +209,7 @@ static inline unsigned long long cycles_2_ns(unsigned long long cyc)
 	 * dance when its actually needed.
 	 */
 
-	preempt_disable();
+	preempt_disable_notrace();
 	data = this_cpu_read(cyc2ns.head);
 	tail = this_cpu_read(cyc2ns.tail);
 
@@ -229,7 +229,7 @@ static inline unsigned long long cycles_2_ns(unsigned long long cyc)
 		if (!--data->__count)
 			this_cpu_write(cyc2ns.tail, data);
 	}
-	preempt_enable();
+	preempt_enable_notrace();
 
 	return ns;
 }