Re: [PATCH V2 -tip 3/4] Tracing/ftrace: Adapt mmiotrace to the new type of print_line

[Steven Rostedt <rostedt@goodmis.org>]

On Tue, 30 Sep 2008, Fr?d?ric Weisbecker wrote:

> 2008/9/30 Steven Rostedt <rostedt@goodmis.org>:
> >
> > OK, I'll nuke the ring_buffer_lock :-/
> >
> > The trace_pipe calls that and then calls print_trace_line which will
> > eventually lock the buffer again. This is a bug on my part. I'll fix that
> > today.
> >
> > Thanks,
> 
> Strange, I can't see any case where print_trace_line could call the
> ring_buffer_lock.
> Hmm, I will see in your patch.

Opps, I mean the find_next_entry_inc will do the lock.

The ring_buffer_lock locks all CPU buffers, and the find_next_entry_inc
will read from the ring buffer which will also lock the buffer. Hence
a deadlock.

-- Steve

> 
> Ingo, I just saw one damage from the merging, trace_empty() returns
> TRACE_TYPE_HANDLED. The type is wrong, it should return 1.
> It's not urgent since the value is the same. Should I send a patch for
> such a small error?