Re: [ANNOUNCE][PATCH 0 of 4] zedtrace, a general-purpose binary tracer

[Tom Zanussi <tzanussi@gmail.com>]

Hi,

On Fri, 2009-02-27 at 17:46 +0100, Frederic Weisbecker wrote:
> On Fri, Feb 27, 2009 at 03:00:38AM -0600, Tom Zanussi wrote:
> > 
> > Available tracepoints:
> > 
> > sched
> >   sched:kthread_stop
> >   sched:kthread_stop_ret
> >   sched:wait_task
> >   sched:wakeup
> >   sched:wakeup_new
> >   sched:switch
> >   sched:migrate_task
> >   sched:process_free
> >   sched:process_exit
> >   sched:process_wait
> >   sched:process_fork
> >   sched:signal_send
> > block
> >   block:rq_abort
> >   block:rq_insert
> >   block:rq_issue
> >   block:rq_requeue
> >   block:rq_complete
> >   block:bio_bounce
> >   block:bio_complete
> >   block:bio_backmerge
> >   block:bio_frontmerge
> >   block:bio_queue
> >   block:getrq
> >   block:sleeprq
> >   block:plug
> >   block:unplug_io
> >   block:unplug_timer
> >   block:split
> >   block:remap
> > workqueue
> >   workqueue:insertion
> >   workqueue:execution
> >   workqueue:creation
> >   workqueue:destruction
> 
> 
> You already wrote the subsystem, so I guess the subsystem: prefix
> is a bit too much. Anyway, sorting the tracepoints by subsystem is a good
> idea too.
> 

I add the subsystem prefix so you can just scrape the full event name
off the listing shown on the screen and feed it directly to the
command-line.

Good point about the sorting though, will put it on my todo list.

> > the infrastructure is there for. ;-)
> > 
> > 
> 
> 
> Great work.
>  

Thanks!