Re: [PATCH 1/4] tracing: add run-time field descriptions for eventfiltering

[Steven Rostedt]

On Sun, 22 Mar 2009, Tom Zanussi wrote:
> @@ -19,6 +19,34 @@
>  
>  static DEFINE_MUTEX(event_mutex);
>  
> +int trace_define_field(struct ftrace_event_call *call, char *type,
> +		       char *name, int offset, int size)
> +{
> +	struct ftrace_event_field *field;
> +
> +	field = kmalloc(sizeof(*field), GFP_KERNEL);
> +	if (!field)
> +		goto err;
> +	field->name = kstrdup(name, GFP_KERNEL);
> +	if (!field->name)
> +		goto err;
> +	field->type = kstrdup(type, GFP_KERNEL);
> +	if (!field->type)
> +		goto err;
> +	field->offset = offset;
> +	field->size = size;
> +	list_add(&field->link, &call->fields);
> +
> +	return 0;
> +err:
> +	if (field) {
> +		kfree(field->name);
> +		kfree(field->type);

Field was not allocated with kzalloc, if we failed to allocate name, then 
type is unknown, and kfree(field->type) may corrupt the system.

What I would do is:

	field = kzalloc(...);
	if (!field)
		return -ENOMEM;

And then we could also get rid of the if (field) check in err.


> +	}
> +	kfree(field);
> +	return -ENOMEM;
> +}
> +

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