linux-rt-users.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: John Kacur <jkacur@redhat.com>
To: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Cc: linux-rt-users@vger.kernel.org,
	Clark Williams <williams@redhat.com>,
	 Steven Rostedt <rostedt@goodmis.org>
Subject: Re: [PATCH 07/10] rt-utils: Decouple writting trace-marks with trace stopping
Date: Fri, 19 Sep 2025 14:49:16 -0400 (EDT)	[thread overview]
Message-ID: <c2fa419a-75bd-ca4c-db95-eb09035d0287@redhat.com> (raw)
In-Reply-To: <20250903111717.1391196-8-bigeasy@linutronix.de>



On Wed, 3 Sep 2025, Sebastian Andrzej Siewior wrote:

> Most users write a trace mark and expect that the trace stops afterwads.
> Decoupling writting and stopping the trace so it can be used in context
> where trace stopping is not needed.
> 
> Make tracemark() writting just to the tracemark file.
> Add tracing_stop() which stops tracing and let current users of
> tracemark() use it.
> 
> Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
> ---
>  src/cyclictest/cyclictest.c         | 1 +
>  src/include/rt-utils.h              | 1 +
>  src/lib/rt-utils.c                  | 8 ++++++--
>  src/oslat/oslat.c                   | 1 +
>  src/sched_deadline/cyclicdeadline.c | 1 +
>  5 files changed, 10 insertions(+), 2 deletions(-)
> 
> diff --git a/src/cyclictest/cyclictest.c b/src/cyclictest/cyclictest.c
> index 890da5d869c92..b4bce8915b432 100644
> --- a/src/cyclictest/cyclictest.c
> +++ b/src/cyclictest/cyclictest.c
> @@ -852,6 +852,7 @@ static void *timerthread(void *param)
>  				break_thread_id = stat->tid;
>  				tracemark("hit latency threshold (%llu > %d)",
>  					  (unsigned long long) diff, tracelimit);
> +				tracing_stop();
>  				break_thread_value = diff;
>  			}
>  			pthread_mutex_unlock(&break_thread_id_lock);
> diff --git a/src/include/rt-utils.h b/src/include/rt-utils.h
> index 569023c5d3051..e148dd77fd7d8 100644
> --- a/src/include/rt-utils.h
> +++ b/src/include/rt-utils.h
> @@ -31,6 +31,7 @@ int parse_mem_string(char *str, uint64_t *val);
>  
>  void enable_trace_mark(void);
>  void tracemark(char *fmt, ...) __attribute__((format(printf, 1, 2)));
> +void tracing_stop(void);
>  void disable_trace_mark(void);
>  
>  #define MSEC_PER_SEC		1000
> diff --git a/src/lib/rt-utils.c b/src/lib/rt-utils.c
> index 294fc9683a122..e504b17738411 100644
> --- a/src/lib/rt-utils.c
> +++ b/src/lib/rt-utils.c
> @@ -427,7 +427,7 @@ void tracemark(char *fmt, ...)
>  
>  	/* bail out if we're not tracing */
>  	/* or if the kernel doesn't support trace_mark */
> -	if (tracemark_fd < 0 || trace_fd < 0)
> +	if (tracemark_fd < 0)
>  		return;
>  
>  	va_start(ap, fmt);
> @@ -436,8 +436,12 @@ void tracemark(char *fmt, ...)
>  
>  	/* write the tracemark message */
>  	write(tracemark_fd, tracebuf, len);
> +}
>  
> -	/* now stop any trace */
> +void tracing_stop(void)
> +{
> +	if (trace_fd < 0)
> +		return;
>  	write(trace_fd, "0\n", 2);
>  }
>  
> diff --git a/src/oslat/oslat.c b/src/oslat/oslat.c
> index 9e5a08a2fbd92..4291221dc6521 100644
> --- a/src/oslat/oslat.c
> +++ b/src/oslat/oslat.c
> @@ -340,6 +340,7 @@ static void insert_bucket(struct thread *t, stamp_t value)
>  		char *line = "%s: Trace threshold (%d us) triggered on cpu %d with %.*f us!\n";
>  		tracemark(line, g.app_name, g.trace_threshold, t->core_i,
>  			  g.precision, us);
> +		tracing_stop();
>  		err_quit(line, g.app_name, g.trace_threshold, t->core_i,
>  			 g.precision, us);
>  	}
> diff --git a/src/sched_deadline/cyclicdeadline.c b/src/sched_deadline/cyclicdeadline.c
> index 04e61edc68d83..e2f827deb275a 100644
> --- a/src/sched_deadline/cyclicdeadline.c
> +++ b/src/sched_deadline/cyclicdeadline.c
> @@ -885,6 +885,7 @@ void *run_deadline(void *data)
>  				break_thread_value = stat->max;
>  				tracemark("hit latency threshold (%lld > %d)",
>  						 (unsigned long long) stat->max, tracelimit);
> +				tracing_stop();
>  			}
>  			pthread_mutex_unlock(&break_thread_id_lock);
>  			break;
> -- 
> 2.51.0
> 
> 
    - Fixed spelling of "writing" and "afterwards"
    Signed-off-by: John Kacur <jkacur@redhat.com>


  reply	other threads:[~2025-09-19 18:49 UTC|newest]

Thread overview: 21+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2025-09-03 11:17 [PATCH 00/10] rt-tests: Remove debugfs usage Sebastian Andrzej Siewior
2025-09-03 11:17 ` [PATCH 01/10] README: Drop "debug" from the tracing path Sebastian Andrzej Siewior
2025-09-19 18:18   ` John Kacur
2025-09-03 11:17 ` [PATCH 02/10] rt-utils: Remove mount_debugfs() Sebastian Andrzej Siewior
2025-09-19 18:20   ` John Kacur
2025-09-03 11:17 ` [PATCH 03/10] rt-utils: Rename get_debugfileprefix() and let it look for tracefs Sebastian Andrzej Siewior
2025-09-19 18:29   ` John Kacur
2025-09-03 11:17 ` [PATCH 04/10] queuelat, rt-migrate-test: Use tracefs for tracing Sebastian Andrzej Siewior
2025-09-19 18:39   ` John Kacur
2025-09-03 11:17 ` [PATCH 05/10] rt-utils: Remove fileprefix Sebastian Andrzej Siewior
2025-09-19 18:41   ` John Kacur
2025-09-03 11:17 ` [PATCH 06/10] rt-utils: Simplify trace_file_exists() Sebastian Andrzej Siewior
2025-09-19 18:42   ` John Kacur
2025-09-03 11:17 ` [PATCH 07/10] rt-utils: Decouple writting trace-marks with trace stopping Sebastian Andrzej Siewior
2025-09-19 18:49   ` John Kacur [this message]
2025-09-03 11:17 ` [PATCH 08/10] sched_deadline: Use tracemark() from rt-utils instead cusstom implementation Sebastian Andrzej Siewior
2025-09-19 18:52   ` John Kacur
2025-09-03 11:17 ` [PATCH 09/10] sched_deadline: Check the resulting size returned from snprintf() Sebastian Andrzej Siewior
2025-09-19 18:54   ` John Kacur
2025-09-03 11:17 ` [PATCH 10/10] cyclicdeadline: Don't test for /sys/kernel/debug/sched_features Sebastian Andrzej Siewior
2025-09-19 18:55   ` John Kacur

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=c2fa419a-75bd-ca4c-db95-eb09035d0287@redhat.com \
    --to=jkacur@redhat.com \
    --cc=bigeasy@linutronix.de \
    --cc=linux-rt-users@vger.kernel.org \
    --cc=rostedt@goodmis.org \
    --cc=williams@redhat.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).