linux-kernel - Re: Enable histograms for events/ftrace/timerlat and osnoise

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite for Android: free password hash cracker in your pocket

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Message-ID: <20230327112715.023a94fd@gandalf.local.home>
Date:   Mon, 27 Mar 2023 11:27:15 -0400
From:   Steven Rostedt <rostedt@...dmis.org>
To:     Daniel Bristot de Oliveira <bristot@...hat.com>
Cc:     Masami Hiramatsu <mhiramat@...nel.org>,
        Tom Zanussi <zanussi@...nel.org>,
        LKML <linux-kernel@...r.kernel.org>
Subject: Re: Enable histograms for events/ftrace/timerlat and osnoise

On Mon, 27 Mar 2023 14:37:11 +0200
Daniel Bristot de Oliveira <bristot@...hat.com> wrote:

> On 3/24/23 17:13, Daniel Bristot de Oliveira wrote:
> > Hi Steven
> > 
> > I would like to collect histograms from timerlat and osnoise tracers event,
> > but I cannot collect them because they are not tracepoints but... tracer
> > events. So...
> > 
> >   # ls /sys/kernel/tracing/events/ftrace/timerlat/
> >   format	hist
> > 
> > no hist, filter, trigger files...  
> 
> Ooop, the hist file exists, what is missing is the trigger file, which is
> used to enable the histogram...
> 
> it seems that the "limitation" is around TRACE_EVENT_FL_IGNORE_ENABLE flag, and
> the fact that ftrace events cannot be enabled/disabled... 
> 
> but wait, as the hist file exist, is there any other way to enable
> the histogram? (or should we disable hist files for !TRACE_EVENT_FL_IGNORE_ENABLE)?

No, I think we just missed not creating the "hist" file, which should not
actually exist.

> 
> am I missing something?
> 
> > How hard would it be to enable histograms to the ftrace/osnoise and
> > ftrace/timerlat events?

I'll try to get some time to look at this. No guarantees :-p

-- Steve

> > 
> > This will allow me to reduce the overhead of rtla timerlat and rtla osnoise
> > tools...
> > 
> > Thanks in advance, Stephen!
> > -- Denny