Re: [PATCH] trace: Add a free on close control mechanism for buffer_size_kb

[Steven Rostedt <rostedt@goodmis.org>]

On Fri, 2011-04-29 at 16:16 -0700, Vaibhav Nagarnaik wrote:

> +static ssize_t tracing_resize_ring_buffer(unsigned long size)
> +{
> +	int cpu, ret = 0;
> +
> +	mutex_lock(&trace_types_lock);
> +
> +	tracing_stop();
> +
> +	/* disable all cpu buffers */
> +	for_each_tracing_cpu(cpu) {
> +		if (global_trace.data[cpu])
> +			atomic_inc(&global_trace.data[cpu]->disabled);
> +		if (max_tr.data[cpu])
> +			atomic_inc(&max_tr.data[cpu]->disabled);
> +	}
> +
> +	if (size != global_trace.entries)
> +		ret = __tracing_resize_ring_buffer(size);
> +
> +	/* If check pages failed, return ENOMEM */
> +	if (tracing_disabled)
> +		ret = -ENOMEM;

This is incorrect. tracing_disabled is set from
__tracing_resize_ring_buffer() if the max buffer can't become the same
size as the global buffer.

What you should do is just check if ret is less than zero here.

	if (ret < 0)
		ret = -ENOMEM;

-- Steve


> +
> +	for_each_tracing_cpu(cpu) {
> +		if (global_trace.data[cpu])
> +			atomic_dec(&global_trace.data[cpu]->disabled);
> +		if (max_tr.data[cpu])
> +			atomic_dec(&max_tr.data[cpu]->disabled);
> +	}
> +
> +	tracing_start();
> +	mutex_unlock(&trace_types_lock);
> +
> +	return ret;
> +}
> +
>