Re: [PATCH 4/4] tracing: add per-subsystem filtering

[Tom Zanussi <tzanussi@gmail.com>]

Hi,

On Mon, 2009-03-23 at 14:24 -0400, Steven Rostedt wrote:
> On Sun, 22 Mar 2009, Tom Zanussi wrote:
> >  
> > +struct event_subsystem {
> > +	struct list_head	list;
> > +	const char		*name;
> > +	struct dentry		*entry;
> > +	struct filter_pred	**preds;
> > +};
> > +
> > +#define events_for_each(event)						\
> > +	for (event = __start_ftrace_events;				\
> > +	     (unsigned long)event < (unsigned long)__stop_ftrace_events; \
> > +	     event++)
> > +
> >  #define MAX_FILTER_PRED 8
> >  
> >  struct filter_pred;
> > @@ -832,6 +844,9 @@ extern int filter_add_pred(struct ftrace_event_call *call,
> >  			   struct filter_pred *pred);
> >  extern void filter_free_preds(struct ftrace_event_call *call);
> >  extern int filter_match_preds(struct ftrace_event_call *call, void *rec);
> > +extern void filter_free_subsystem_preds(struct event_subsystem *system);
> > +extern int filter_add_subsystem_pred(struct event_subsystem *system,
> > +				     struct filter_pred *pred);
> >  
> >  void event_trace_printk(unsigned long ip, const char *fmt, ...);
> >  extern struct ftrace_event_call __start_ftrace_events[];
> > diff --git a/kernel/trace/trace_events.c b/kernel/trace/trace_events.c
> > index 97470c4..97d4daa 100644
> > --- a/kernel/trace/trace_events.c
> > +++ b/kernel/trace/trace_events.c
> > @@ -524,6 +524,71 @@ event_filter_write(struct file *filp, const char __user *ubuf, size_t cnt,
> >  	return cnt;
> >  }
> >  
> > +static ssize_t
> > +subsystem_filter_read(struct file *filp, char __user *ubuf, size_t cnt,
> > +		      loff_t *ppos)
> > +{
> > +	struct event_subsystem *system = filp->private_data;
> > +	struct trace_seq *s;
> 
> Again, trace_seq is not used, might as well use your own buffer.
> 
> > +	int r;
> > +
> > +	if (*ppos)
> > +		return 0;
> > +
> > +	s = kmalloc(sizeof(*s), GFP_KERNEL);
> > +	if (!s)
> > +		return -ENOMEM;
> > +
> > +	trace_seq_init(s);
> > +
> > +	r = filter_print_preds(system->preds, s->buffer);
> > +	r = simple_read_from_buffer(ubuf, cnt, ppos, s->buffer, r);
> > +
> > +	kfree(s);
> > +
> > +	return r;
> > +}
> > +
> > +static ssize_t
> > +subsystem_filter_write(struct file *filp, const char __user *ubuf, size_t cnt,
> > +		       loff_t *ppos)
> > +{
> > +	struct event_subsystem *system = filp->private_data;
> > +	char buf[64], *pbuf = buf;
> > +	struct filter_pred *pred;
> > +	int err;
> > +
> > +	if (cnt >= sizeof(buf))
> > +		return -EINVAL;
> > +
> > +	if (copy_from_user(&buf, ubuf, cnt))
> > +		return -EFAULT;
> > +
> > +	pred = kzalloc(sizeof(*pred), GFP_KERNEL);
> > +	if (!pred)
> > +		return -ENOMEM;
> > +
> > +	err = filter_parse(&pbuf, pred);
> > +	if (err < 0) {
> > +		filter_free_pred(pred);
> > +		return err;
> > +	}
> > +
> > +	if (pred->clear) {
> > +		filter_free_subsystem_preds(system);
> 
> is "system" correct here?

Do you mean the naming i.e. would be better as "subsystem"?  Otherwise,
I think it's correct.

> 
> > +		return cnt;
> > +	}
> > +
> > +	if (filter_add_subsystem_pred(system, pred)) {
> > +		filter_free_pred(pred);
> > +		return -EINVAL;
> > +	}
> > +
> > +	*ppos += cnt;
> > +
> > +	return cnt;
> > +}
> > +
> >  static const struct seq_operations show_event_seq_ops = {
> >  	.start = t_start,
> >  	.next = t_next,
> > @@ -575,6 +640,12 @@ static const struct file_operations ftrace_event_filter_fops = {
> >  	.write = event_filter_write,
> >  };
> >  
> > +static const struct file_operations ftrace_subsystem_filter_fops = {
> > +	.open = tracing_open_generic,
> > +	.read = subsystem_filter_read,
> > +	.write = subsystem_filter_write,
> > +};
> > +
> >  static struct dentry *event_trace_events_dir(void)
> >  {
> >  	static struct dentry *d_tracer;
> > @@ -595,18 +666,13 @@ static struct dentry *event_trace_events_dir(void)
> >  	return d_events;
> >  }
> >  
> > -struct event_subsystem {
> > -	struct list_head	list;
> > -	const char		*name;
> > -	struct dentry		*entry;
> > -};
> > -
> >  static LIST_HEAD(event_subsystems);
> >  
> >  static struct dentry *
> >  event_subsystem_dir(const char *name, struct dentry *d_events)
> >  {
> >  	struct event_subsystem *system;
> > +	struct dentry *entry;
> >  
> >  	/* First see if we did not already create this dir */
> >  	list_for_each_entry(system, &event_subsystems, list) {
> > @@ -633,6 +699,14 @@ event_subsystem_dir(const char *name, struct dentry *d_events)
> >  	system->name = name;
> >  	list_add(&system->list, &event_subsystems);
> >  
> > +	system->preds = NULL;
> > +
> > +	entry = debugfs_create_file("filter", 0444, system->entry, system,
> > +				    &ftrace_subsystem_filter_fops);
> > +	if (!entry)
> > +		pr_warning("Could not create debugfs "
> > +			   "'%s/filter' entry\n", name);
> > +
> >  	return system->entry;
> >  }
> >  
> > diff --git a/kernel/trace/trace_events_filter.c b/kernel/trace/trace_events_filter.c
> > index 8e8c5fa..1ab20ce 100644
> > --- a/kernel/trace/trace_events_filter.c
> > +++ b/kernel/trace/trace_events_filter.c
> > @@ -181,6 +181,27 @@ void filter_free_preds(struct ftrace_event_call *call)
> >  	}
> >  }
> >  
> > +void filter_free_subsystem_preds(struct event_subsystem *system)
> > +{
> > +	struct ftrace_event_call *call = __start_ftrace_events;
> > +	int i;
> > +
> > +	if (system->preds) {
> > +		for (i = 0; i < MAX_FILTER_PRED; i++)
> > +			filter_free_pred(system->preds[i]);
> > +		kfree(system->preds);
> > +		system->preds = NULL;
> > +	}
> > +
> > +	events_for_each(call) {
> > +		if (!call->name || !call->regfunc)
> > +			continue;
> > +
> > +		if (!strcmp(call->system, system->name))
> > +			filter_free_preds(call);
> > +	}
> > +}
> > +
> >  static int __filter_add_pred(struct ftrace_event_call *call,
> >  			     struct filter_pred *pred)
> >  {
> > @@ -250,6 +271,65 @@ int filter_add_pred(struct ftrace_event_call *call, struct filter_pred *pred)
> >  	return __filter_add_pred(call, pred);
> >  }
> >  
> > +static struct filter_pred *copy_pred(struct filter_pred *pred)
> > +{
> > +	struct filter_pred *new_pred = kmalloc(sizeof(*pred), GFP_KERNEL);
> > +	if (!new_pred)
> > +		return NULL;
> > +
> > +	memcpy(new_pred, pred, sizeof(*pred));
> > +	if (pred->str_val) {
> > +		new_pred->str_val = kstrdup(pred->str_val, GFP_KERNEL);
> > +		new_pred->field_name = kstrdup(pred->field_name, GFP_KERNEL);
> > +		if (!new_pred->str_val) {
> > +			kfree(new_pred);
> 
> Shouldn't there be a check for field_name too?
> 

Yes - I posted a patch to copy_pred() yesterday to fix that.

Tom

> -- Steve
> 
> > +			return NULL;
> > +		}
> > +	}
> > +
> > +	return new_pred;
> > +}
> > +