Re: [tip:tracing/urgent] tracing: Add pipe_close interface

[Ingo Molnar <mingo@elte.hu>]

* Steven Rostedt <rostedt@goodmis.org> wrote:

> On Wed, 2009-12-09 at 15:35 +0900, KOSAKI Motohiro wrote:
> 
> > > diff --git a/kernel/trace/trace.c b/kernel/trace/trace.c
> > > index 874f289..f804b40 100644
> > > --- a/kernel/trace/trace.c
> > > +++ b/kernel/trace/trace.c
> > > @@ -2898,6 +2898,10 @@ static int tracing_release_pipe(struct inode *inode, struct file *file)
> > >  	else
> > >  		cpumask_clear_cpu(iter->cpu_file, tracing_reader_cpumask);
> > >  
> > > +
> > > +	if (iter->trace->pipe_open)
> > > +		iter->trace->pipe_close(iter);
> > > +
> > 
> > What's happen if pipe_close is NULL? Wny following straightforward check
> > is wrong?
> > I mean the above description explain pipe_close is only useful if plugin
> > allocate something at pipe_open. then allowing NULL seems natural.
> > 
> > 	if (iter->trace->pipe_close)
> > 		iter->trace->pipe_close(iter);
> 
> Ug, good point (stupid cut & paste should be illegal).
> 
> Ingo, I'll fix this up too in the next patch set. Want me to rebase it 
> or just start with this fix?

I'd suggest to start with a fix. Thanks,

	Ingo