Date:	Wed, 17 Dec 2008 08:25:37 -0500 (EST)
From:	Steven Rostedt <rostedt@...dmis.org>
To:	Andrew Morton <akpm@...ux-foundation.org>
cc:	LKML <linux-kernel@...r.kernel.org>, Ingo Molnar <mingo@...e.hu>,
	Frédéric Weisbecker <fweisbec@...il.com>
Subject: Re: [PATCH] trace: add a way to enable or disable the stack tracer


On Tue, 16 Dec 2008, Andrew Morton wrote:

> >     
> >     The stack tracer currently is either on when configured in or
> >     off when it is not. It can not be disabled when it is configured on.
> >     (besides disabling the function tracer that it uses)
> >     
> >     This patch adds a way to enable or disable the stack tracer at
> >     run time. It defaults off on bootup, but a kernel parameter 'stacktrace'
> >     has been added to enable it on bootup.
> >     
> >     A new sysctl has been added "kernel.stack_tracer_enabled" to let
> >     the user enable or disable the stack tracer at run time.
> >     
> 
> Why do we need the boot parameter if there's a runtime control?

Because it would be a regression without it. The current stack tracer lets 
you know find the biggest stack size since initcall. There's some pretty 
big stack sizes in the boot up process, and I want to make it easy to 
record them. My test kernels now have this defaulted on.

> 
> > 
> > diff --git a/Documentation/kernel-parameters.txt b/Documentation/kernel-parameters.txt
> > index 16aaa84..95fa674 100644
> > --- a/Documentation/kernel-parameters.txt
> > +++ b/Documentation/kernel-parameters.txt
> > @@ -89,6 +89,7 @@ parameter is applicable:
> >  	SPARC	Sparc architecture is enabled.
> >  	SWSUSP	Software suspend (hibernation) is enabled.
> >  	SUSPEND	System suspend states are enabled.
> > +	FTRACE	Function tracing enabled.
> >  	TS	Appropriate touchscreen support is enabled.
> >  	USB	USB support is enabled.
> >  	USBHID	USB Human Interface Device support is enabled.
> > @@ -2197,6 +2198,9 @@ and is between 256 and 4096 characters. It is defined in the file
> >  	st=		[HW,SCSI] SCSI tape parameters (buffers, etc.)
> >  			See Documentation/scsi/st.txt.
> >  
> > +	stacktrace	[FTRACE]
> > +			Enabled the stack tracer on boot up.
> > +
> >  	sti=		[PARISC,HW]
> >  			Format: <num>
> >  			Set the STI (builtin display/keyboard on the HP-PARISC
> > diff --git a/include/linux/ftrace.h b/include/linux/ftrace.h
> > index 286af82..04b52e6 100644
> > --- a/include/linux/ftrace.h
> > +++ b/include/linux/ftrace.h
> > @@ -87,6 +87,14 @@ static inline void ftrace_stop(void) { }
> >  static inline void ftrace_start(void) { }
> >  #endif /* CONFIG_FUNCTION_TRACER */
> >  
> > +#ifdef CONFIG_STACK_TRACER
> > +extern int stack_tracer_enabled;
> > +int
> > +stack_trace_sysctl(struct ctl_table *table, int write,
> > +		   struct file *file, void __user *buffer, size_t *lenp,
> > +		   loff_t *ppos);
> > +#endif
> 
> The ifdef could be omitted here.

Yeah I thought about that too. But it is in a header file, and I might be 
adding more stuff as well. The ifdef here is more to annotate what the 
prototype is for than to protect it. This is one of the few cases that 
adding a #ifdef is cleaner than to omit it (IMHO).

> 
> >  #ifdef CONFIG_DYNAMIC_FTRACE
> >  /* asm/ftrace.h must be defined for archs supporting dynamic ftrace */
> >  #include <asm/ftrace.h>
> > diff --git a/kernel/sysctl.c b/kernel/sysctl.c
> > index c7bf3dd..2414fa8 100644
> > --- a/kernel/sysctl.c
> > +++ b/kernel/sysctl.c
> > @@ -488,6 +488,16 @@ static struct ctl_table kern_table[] = {
> >  		.proc_handler	= &ftrace_enable_sysctl,
> >  	},
> >  #endif
> > +#ifdef CONFIG_STACK_TRACER
> > +	{
> > +		.ctl_name	= CTL_UNNUMBERED,
> > +		.procname	= "stack_tracer_enabled",
> > +		.data		= &stack_tracer_enabled,
> > +		.maxlen		= sizeof(int),
> > +		.mode		= 0644,
> > +		.proc_handler	= &stack_trace_sysctl,
> > +	},
> > +#endif
> >  #ifdef CONFIG_TRACING
> >  	{
> >  		.ctl_name	= CTL_UNNUMBERED,
> > diff --git a/kernel/trace/Kconfig b/kernel/trace/Kconfig
> > index d8bae6f..e2a4ff6 100644
> > --- a/kernel/trace/Kconfig
> > +++ b/kernel/trace/Kconfig
> > @@ -244,10 +244,15 @@ config STACK_TRACER
> >  
> >  	  This tracer works by hooking into every function call that the
> >  	  kernel executes, and keeping a maximum stack depth value and
> > -	  stack-trace saved. Because this logic has to execute in every
> > -	  kernel function, all the time, this option can slow down the
> > -	  kernel measurably and is generally intended for kernel
> > -	  developers only.
> > +	  stack-trace saved.  If this is configured with DYNAMIC_FTRACE
> > +	  then it will not have any overhead while the stack tracer
> > +	  is disabled.
> > +
> > +	  To enable the stack tracer on bootup, pass in 'stacktrace'
> > +	  on the kernel command line.
> > +
> > +	  The stack tracer can also be enabled or disabled via the
> > +	  sysctl kernel.stack_tracer_enabled
> >  
> >  	  Say N if unsure.
> >  
> > diff --git a/kernel/trace/trace_stack.c b/kernel/trace/trace_stack.c
> > index 0b863f2..4842c96 100644
> > --- a/kernel/trace/trace_stack.c
> > +++ b/kernel/trace/trace_stack.c
> > @@ -10,6 +10,7 @@
> >  #include <linux/debugfs.h>
> >  #include <linux/ftrace.h>
> >  #include <linux/module.h>
> > +#include <linux/sysctl.h>
> >  #include <linux/init.h>
> >  #include <linux/fs.h>
> >  #include "trace.h"
> > @@ -31,6 +32,10 @@ static raw_spinlock_t max_stack_lock =
> >  
> >  static int stack_trace_disabled __read_mostly;
> >  static DEFINE_PER_CPU(int, trace_active);
> > +static DEFINE_MUTEX(stack_sysctl_mutex);
> > +
> > +int stack_tracer_enabled;
> > +static int last_stack_tracer_enabled;
> >  
> >  static inline void check_stack(void)
> >  {
> > @@ -174,7 +179,7 @@ stack_max_size_write(struct file *filp, const char __user *ubuf,
> >  	return count;
> >  }
> >  
> > -static struct file_operations stack_max_size_fops = {
> > +static const struct file_operations stack_max_size_fops = {
> >  	.open		= tracing_open_generic,
> >  	.read		= stack_max_size_read,
> >  	.write		= stack_max_size_write,
> > @@ -272,7 +277,7 @@ static int t_show(struct seq_file *m, void *v)
> >  	return 0;
> >  }
> >  
> > -static struct seq_operations stack_trace_seq_ops = {
> > +static const struct seq_operations stack_trace_seq_ops = {
> >  	.start		= t_start,
> >  	.next		= t_next,
> >  	.stop		= t_stop,
> > @@ -288,12 +293,48 @@ static int stack_trace_open(struct inode *inode, struct file *file)
> >  	return ret;
> >  }
> >  
> > -static struct file_operations stack_trace_fops = {
> > +static const struct file_operations stack_trace_fops = {
> >  	.open		= stack_trace_open,
> >  	.read		= seq_read,
> >  	.llseek		= seq_lseek,
> >  };
> >  
> > +int
> > +stack_trace_sysctl(struct ctl_table *table, int write,
> > +		   struct file *file, void __user *buffer, size_t *lenp,
> > +		   loff_t *ppos)
> > +{
> > +	int ret;
> > +
> > +	mutex_lock(&stack_sysctl_mutex);
> > +
> > +	ret  = proc_dointvec(table, write, file, buffer, lenp, ppos);
> 
> funny whitespace

Hmm, I cut and pasted this function from ftrace.c. It must be there too.
(goes look, and sure enough it is)

> 
> > +	if (ret || !write ||
> > +	    (last_stack_tracer_enabled == stack_tracer_enabled))
> > +		goto out;
> > +
> > +	last_stack_tracer_enabled = stack_tracer_enabled;
> > +
> > +	if (stack_tracer_enabled)
> > +		register_ftrace_function(&trace_ops);
> > +	else
> > +		unregister_ftrace_function(&trace_ops);
> > +
> > + out:
> > +	mutex_unlock(&stack_sysctl_mutex);
> > +	return ret;
> > +}
> > +
> > +static int start_stack_trace __initdata;
> > +
> > +static __init int enable_stacktrace(char *str)
> > +{
> > +	start_stack_trace = 1;
> > +	return 1;
> > +}
> > +__setup("stacktrace", enable_stacktrace);
> > +
> >  static __init int stack_trace_init(void)
> >  {
> >  	struct dentry *d_tracer;
> > @@ -311,7 +352,10 @@ static __init int stack_trace_init(void)
> >  	if (!entry)
> >  		pr_warning("Could not create debugfs 'stack_trace' entry\n");
> >  
> > -	register_ftrace_function(&trace_ops);
> > +	if (start_stack_trace) {
> > +		register_ftrace_function(&trace_ops);
> > +		stack_tracer_enabled = 1;
> > +	}
> 
> I think we could do away with start_stack_trace and just set
> stack_tracer_enabled=1 in enable_stacktrace()?

Ah yes. I'll write a follow up patch to fix that.

-- Steve

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

Powered by blists - more mailing lists