Skip to contentSkip to conversation

version 0.37 of rt-tests available

PATCH Applied Clark Williams 4 participants 7 messages Apr 29, 2009 – Jul 2, 2009

lore
Share & tools
lore
#0 lore raw

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

I've pushed 0.37 of the rt-tests up to my git repo at:

git://git.kernel.org/pub/scm/linux/kernel/git/clrkwllms/rt-tests.git

tarballs available at:

http://www.kernel.org/pub/linux/kernel/people/clrkwllms/rt-tests/

This version has a couple of fixes from GeunSik Lim <[email protected]> (man page fixes and two new options for wakeup and wakeup_rt tracing).

After seeing that we have a never-ending stream of ftrace tracing functions and options available to us, I decided that we need a better way to specify the tracers. So I added the --tracer option (which accepts the name of a tracing function) and the --traceopt option (which accepts tracing options), e.g.:

$ cyclictest -b 500 --tracer=events --traceopts=noraw --traceopts=nohex

I guess I could make traceopts take a comma separated list, but for now it was just easier to allow multiple invocations on a command line.

I've left the other tracing options alone for now, since I'm sure people have scripts that use them. I wouldn't mind getting rid of them in the future though.

Clark
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.0.10 (GNU/Linux)

iEYEARECAAYFAkn4vYoACgkQHyuj/+TTEp0eRgCgyJy7jg4o7qEGxqlwCoZRqwBq Ct4An129sVkC7Md6fokUDpW8QOw03PPR
=HPJj
-----END PGP SIGNATURE-----

#1 lore raw

Thank you for thinking about needs of the frace options using cyclictest.

As you said, I think that most rt developers will need ftrace utility based on command line for tuning internal kernel latencies in their complicated realtime system.

Maybe, If cyclictest is written by contribution of many developers successfully, Most peoples will use this cyclictest utility generally.
In special, Interested developers will make gui based cyclictest(ex:gcyclictest) utility
additionally like gmplayer based on mplayer and tsclient based on rdesktop.

So, I agree with your opinion absoultely.

And, Clark Williams. Is a below remark right?
This year is 2009 as you know.
How about you modify date info from "(C) 2008-2008 " to "(C) 2008-2009". (Just my thinking...)

I am happy because you append your name in cyclictest source code. In gerneal, When some developers want to send their patches and opinion, They often
find author and maintainer from source code like me.

< cyclictest source file>
* High resolution timer test software *
* (C) 2008-2008 Clark Williams <[email protected]>
* (C) 2005-2007 Thomas Gleixner <[email protected]>

2009/4/30 Clark Williams <[email protected]>:

Quoted 40 lines

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

I've pushed 0.37 of the rt-tests up to my git repo at:

git://git.kernel.org/pub/scm/linux/kernel/git/clrkwllms/rt-tests.git

tarballs available at:

http://www.kernel.org/pub/linux/kernel/people/clrkwllms/rt-tests/

This version has a couple of fixes from GeunSik Lim <[email protected]> (man page fixes and two new options for wakeup and wakeup_rt tracing).

After seeing that we have a never-ending stream of ftrace tracing functions and options available to us, I decided that we need a better way to specify the tracers. So I added the --tracer option (which accepts the name of a tracing function) and the --traceopt option (which accepts tracing options), e.g.:

$ cyclictest -b 500 --tracer=events --traceopts=noraw --traceopts=nohex

I guess I could make traceopts take a comma separated list, but for now it was just easier to allow multiple invocations on a command line.

I've left the other tracing options alone for now, since I'm sure people have scripts that use them. I wouldn't mind getting rid of them in the future though.

Clark
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.0.10 (GNU/Linux)

iEYEARECAAYFAkn4vYoACgkQHyuj/+TTEp0eRgCgyJy7jg4o7qEGxqlwCoZRqwBq Ct4An129sVkC7Md6fokUDpW8QOw03PPR
=HPJj
-----END PGP SIGNATURE-----

Signature · 7 lines
Regards,
GeunSik Lim
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to [email protected]
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/
#2 lore raw

Dear Clark Willams.

It's strange. It seems that you did't run "#> git commit -a" command.

When I visited http://git.kernel.org/?p=linux/kernel/git/clrkwllms/rt-tests.git;a=summary
webpage, I can't find v0.37 related "diff" contents in the shortlog
table.

Please, Check "commit" command.
How about the others?

Thanks,
GeunSik Lim.

2009/4/30 GeunSik Lim <[email protected]>:

Quoted 84 lines

Thank you for thinking about needs of the frace options using cyclictest.

As you said, I think that most rt developers will need ftrace utility based on command line for tuning internal kernel latencies in their complicated realtime system.

Maybe, If cyclictest is written by contribution of many developers successfully, Most peoples will use this cyclictest utility generally.
In special, Interested developers will make gui based cyclictest(ex:gcyclictest) utility
additionally like gmplayer based on mplayer and tsclient based on rdesktop.

So, I agree with your opinion absoultely.

And, Clark Williams. Is a below remark right?
This year is 2009 as you know.
How about you modify  date info from "(C) 2008-2008 " to "(C) 2008-2009". (Just my thinking...)

I am happy because you append your name in cyclictest source code. In gerneal, When some developers want to send their patches and opinion, They often
find author and maintainer from source code like me.

< cyclictest source file>
 * High resolution timer test software *
 * (C) 2008-2008 Clark Williams <[email protected]>
 * (C) 2005-2007 Thomas Gleixner <[email protected]>

2009/4/30 Clark Williams <[email protected]>:

Quoted 31 lines

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

I've pushed 0.37 of the rt-tests up to my git repo at:

git://git.kernel.org/pub/scm/linux/kernel/git/clrkwllms/rt-tests.git

tarballs available at:

http://www.kernel.org/pub/linux/kernel/people/clrkwllms/rt-tests/

This version has a couple of fixes from GeunSik Lim <[email protected]> (man page fixes and two new options for wakeup and wakeup_rt tracing).

After seeing that we have a never-ending stream of ftrace tracing functions and options available to us, I decided that we need a better way to specify the tracers. So I added the --tracer option (which accepts the name of a tracing function) and the --traceopt option (which accepts tracing options), e.g.:

$ cyclictest -b 500 --tracer=events --traceopts=noraw --traceopts=nohex

I guess I could make traceopts take a comma separated list, but for now it was just easier to allow multiple invocations on a command line.

I've left the other tracing options alone for now, since I'm sure people have scripts that use them. I wouldn't mind getting rid of them in the future though.

Quoted 9 lines

Clark
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.0.10 (GNU/Linux)

iEYEARECAAYFAkn4vYoACgkQHyuj/+TTEp0eRgCgyJy7jg4o7qEGxqlwCoZRqwBq Ct4An129sVkC7Md6fokUDpW8QOw03PPR
=HPJj
-----END PGP SIGNATURE-----

--
Regards,
GeunSik Lim
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [email protected]
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

Signature · 11 lines
Regards,
GeunSik Lim
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to [email protected]
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/
--
To unsubscribe from this list: send the line "unsubscribe linux-rt-users" in
the body of a message to [email protected]
More majordomo info at  http://vger.kernel.org/majordomo-info.html
#3 lore raw
Quoted 7 lines · GeunSik Lim

Dear Clark Willams.

It's strange. It seems that you did't run "#> git commit -a" command.

When I visited http://git.kernel.org/?p=linux/kernel/git/clrkwllms/rt-tests.git;a=summary
webpage, I can't find v0.37 related "diff" contents in the shortlog
table.

The problem is, that only the tag was pushed, not the master branch. GeunSik Lim, what you can do is check out Clark's master branch (being at 0.36) and simply pull in v0.37.

Best regards
Uwe

Signature · 6 lines
Pengutronix e.K.                              | Uwe Kleine-König            |
Industrial Linux Solutions                    | http://www.pengutronix.de/  |
--
To unsubscribe from this list: send the line "unsubscribe linux-rt-users" in
the body of a message to [email protected]
More majordomo info at  http://vger.kernel.org/majordomo-info.html
#4 lore raw

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Quoted 12 lines · On Thu, 30 Apr 2009 08:52:59 +0200 Uwe Kleine-König…
Quoted 7 lines · GeunSik Lim

Dear Clark Willams.

It's strange. It seems that you did't run "#> git commit -a" command.

When I visited http://git.kernel.org/?p=linux/kernel/git/clrkwllms/rt-tests.git;a=summary
webpage, I can't find v0.37 related "diff" contents in the shortlog
table.

The problem is, that only the tag was pushed, not the master branch. GeunSik Lim, what you can do is check out Clark's master branch (being at 0.36) and simply pull in v0.37.

No the problem is that I'm an idiot :). I did all my work on a "clark" branch and forgot to merge to master before pushing. Sigh..

That's done now.

Clark
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.0.10 (GNU/Linux)

iEYEARECAAYFAkn5sWsACgkQHyuj/+TTEp1KXwCggnwJocR3gc1NFsT75b0+FJwz J4kAoIDdp+FYhdeMmEY1QxBS7cOmQhCY
=xAcu
-----END PGP SIGNATURE-----

#5 lore raw

PATCH 1/1 clock_nanosleep interrupt

In reply to Clark Williams

Subject: Avoid segfault of cyclictest if it gets immediately interrupted.

From: Daniel Gollub <[email protected]>

If clock_nanosleep() gets interrupted this could result in a negative time diff from calcdiff().

With the histogram patch this leads to a segfault, since the time diff is used as index for the histogram array:

Core was generated by `/usr/bin/cyclictest -n -q -p 99 -t 2 -i 500 -l 1000000 -h
20000'.
Program terminated with signal 11, Segmentation fault.
#0 0x0000000000402324 in timerthread (param=<value optimized out>)

     at src/cyclictest/cyclictest.c:339
     339                             stat->hist_array[diff] += 1;
     (gdb) p diff
     $1 = -751974
---
 src/cyclictest/cyclictest.c |   17 ++++++++++++-----
 1 file changed, 12 insertions(+), 5 deletions(-)
src/cyclictest/cyclictest.c +12 -5
Index: rt-tests/src/cyclictest/cyclictest.c
===================================================================
@@ -376,7 +376,7 @@ void *timerthread(void *param)
 	while (!shutdown) {
 
 		long diff;
-		int sigs;
+		int sigs, ret;
 
 		/* Wait for next period */
 		switch (par->mode) {
@@ -387,17 +387,24 @@ void *timerthread(void *param)
 			break;
 
 		case MODE_CLOCK_NANOSLEEP:
-			if (par->timermode == TIMER_ABSTIME)
-				clock_nanosleep(par->clock, TIMER_ABSTIME,
+			if (par->timermode == TIMER_ABSTIME) {
+				ret = clock_nanosleep(par->clock, TIMER_ABSTIME,
 						&next, NULL);
-			else {
+			} else {
 				clock_gettime(par->clock, &now);
-				clock_nanosleep(par->clock, TIMER_RELTIME,
+				ret = clock_nanosleep(par->clock, TIMER_RELTIME,
 						&interval, NULL);
 				next.tv_sec = now.tv_sec + interval.tv_sec;
 				next.tv_nsec = now.tv_nsec + interval.tv_nsec;
 				tsnorm(&next);
 			}
+
+			/* Avoid negative calcdiff result if clock_nanosleep() 
+			 * gets interrupted.
+			 */
+			if (ret == EINTR)
+				goto out;
+
 			break;
 
 		case MODE_SYS_NANOSLEEP:
author Applied
#6 lore raw

Re: [PATCH] clock_nanosleep interrupt

Quoted 5 lines · On Wed, 01 Jul 2009 18:57:43 -0700 Sven-Thorsten Dietrich…

Subject: Avoid segfault of cyclictest if it gets immediately interrupted.

From: Daniel Gollub <[email protected]>

Thanks, applied; will go out with 0.45 shortly.

Clark