mirror of https://lore.kernel.org/lkml/
 help / color / mirror / Atom feed
From: Steven Rostedt <rostedt@goodmis.org>
To: Ingo Molnar <mingo@elte.hu>
Cc: Thomas Gleixner <tglx@linutronix.de>,
	john stultz <johnstul@us.ibm.com>,
	linux-kernel@vger.kernel.org,
	david singleton <dsingleton@mvista.com>
Subject: Re: 2.6.14-rc3-rt2
Date: Fri, 7 Oct 2005 07:05:22 -0400 (EDT)	[thread overview]
Message-ID: <Pine.LNX.4.58.0510070538170.6608@localhost.localdomain> (raw)
In-Reply-To: <Pine.LNX.4.58.0510051204170.23350@localhost.localdomain>



On Wed, 5 Oct 2005, Steven Rostedt wrote:

>
> It seems that the problem comes down to the call to getnstimeofday in
> do_gettimeofday.
>

OK, not sure if anyone looked into this more or not.  But this patch seems
to at least fix the symptom if not the cure.  I changed
set_normalize_timespec to take a nsec_t type as its last parameter.
Since I don't see a problem with overflowing a 64 bit number, this works
for now.  But I still don't know the full extent of xtime_last_update not
updating during something like hackbench starving the timer softirq.

-- Steve

Index: linux-rt-quilt/include/linux/time.h
===================================================================
--- linux-rt-quilt.orig/include/linux/time.h	2005-10-07 05:11:00.000000000 -0400
+++ linux-rt-quilt/include/linux/time.h	2005-10-07 05:12:17.000000000 -0400
@@ -107,7 +107,7 @@
 extern struct timespec timespec_trunc(struct timespec t, unsigned gran);

 static inline void
-set_normalized_timespec (struct timespec *ts, time_t sec, long nsec)
+set_normalized_timespec (struct timespec *ts, time_t sec, nsec_t nsec)
 {
 	while (nsec > NSEC_PER_SEC) {
 		nsec -= NSEC_PER_SEC;
Index: linux-rt-quilt/kernel/timeofday.c
===================================================================
--- linux-rt-quilt.orig/kernel/timeofday.c	2005-10-06 08:04:56.000000000 -0400
+++ linux-rt-quilt/kernel/timeofday.c	2005-10-07 05:14:58.000000000 -0400
@@ -174,12 +174,12 @@
 			goto full;
 		set_normalized_timespec(&mono_last_update_ts,
 			mono_last_update_ts.tv_sec,
-			mono_last_update_ts.tv_nsec + (long) delta);
+			mono_last_update_ts.tv_nsec + delta);
 		if (unlikely(leapupdate))
 			set_normalized_timespec(&mono_wall_offset_ts,
 						mono_wall_offset_ts.tv_sec,
 						mono_wall_offset_ts.tv_nsec +
-						(long) leapupdate);
+						leapupdate);
 	}
 	xtime_last_update = system_time;
 	set_normalized_timespec(&xtime,
@@ -266,7 +266,7 @@

 	set_normalized_timespec(ts,
 				mono_last_update_ts.tv_sec,
-				mono_last_update_ts.tv_nsec + (long) delta);
+				mono_last_update_ts.tv_nsec + delta);
 }

 /**
@@ -359,7 +359,7 @@

 	set_normalized_timespec(ts,
 				ts->tv_sec,
-				ts->tv_nsec + (long) delta);
+				ts->tv_nsec + delta);

 }
 EXPORT_SYMBOL(getnstimeofday);

  reply	other threads:[~2005-10-07 11:06 UTC|newest]

Thread overview: 78+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2005-10-04  8:44 2.6.14-rc3-rt2 Ingo Molnar
2005-10-04 12:51 ` 2.6.14-rc3-rt2 K.R. Foley
2005-10-04 13:00   ` 2.6.14-rc3-rt2 Ingo Molnar
2005-10-04 13:45     ` 2.6.14-rc3-rt2 Steven Rostedt
2005-10-04 14:27       ` 2.6.14-rc3-rt2 Ingo Molnar
2005-10-04 15:16         ` 2.6.14-rc3-rt2 Dinakar Guniguntala
2005-10-04 15:32           ` 2.6.14-rc3-rt2 Ingo Molnar
2005-10-04 15:50             ` 2.6.14-rc3-rt2 Dinakar Guniguntala
2005-10-04 15:46               ` 2.6.14-rc3-rt2 Thomas Gleixner
2005-10-04 16:09                 ` 2.6.14-rc3-rt2 Dinakar Guniguntala
2005-10-04 15:34           ` 2.6.14-rc3-rt2 Thomas Gleixner
2005-10-04 16:09           ` 2.6.14-rc3-rt2 Daniel Walker
2005-10-04 16:32             ` 2.6.14-rc3-rt2 Steven Rostedt
2005-10-04 16:46               ` 2.6.14-rc3-rt2 Daniel Walker
2005-10-04 17:58             ` 2.6.14-rc3-rt2 Dinakar Guniguntala
2005-10-04 17:54               ` 2.6.14-rc3-rt2 Daniel Walker
2005-10-04 18:11                 ` 2.6.14-rc3-rt2 Dinakar Guniguntala
2005-10-04 18:27                   ` 2.6.14-rc3-rt2 Daniel Walker
2005-10-04 20:03                     ` 2.6.14-rc3-rt2 Dinakar Guniguntala
2005-10-05  7:37                       ` 2.6.14-rc3-rt2 Ingo Molnar
2005-10-04 14:25     ` 2.6.14-rc3-rt2 Steven Rostedt
2005-10-05  7:39       ` 2.6.14-rc3-rt2 Ingo Molnar
2005-10-04 16:44     ` 2.6.14-rc3-rt2 Mark Knecht
2005-10-04 18:11       ` 2.6.14-rc3-rt2 Mark Knecht
2005-10-04 18:20         ` 2.6.14-rc3-rt2 Thomas Gleixner
2005-10-04 18:58           ` 2.6.14-rc3-rt2 Mark Knecht
2005-10-04 20:45             ` 2.6.14-rc3-rt2 Thomas Gleixner
2005-10-04 20:49               ` 2.6.14-rc3-rt2 Mark Knecht
2005-10-04 20:53                 ` 2.6.14-rc3-rt2 Thomas Gleixner
2005-10-04 21:21                   ` 2.6.14-rc3-rt2 Mark Knecht
2005-10-05 10:56                 ` 2.6.14-rc3-rt2 Ingo Molnar
2005-10-05 17:14                   ` 2.6.14-rc3-rt2 Mark Knecht
2005-10-06  7:54                     ` 2.6.14-rc3-rt2 Steven Rostedt
2005-10-06  8:06                     ` 2.6.14-rc3-rt2 Steven Rostedt
2005-10-06  8:10                       ` 2.6.14-rc3-rt2 Ingo Molnar
2005-10-06  8:29                         ` 2.6.14-rc3-rt2 Steven Rostedt
2005-10-06  8:33                           ` 2.6.14-rc3-rt2 Ingo Molnar
2005-10-06  8:44                             ` 2.6.14-rc3-rt2 Steven Rostedt
2005-10-06 16:04                             ` 2.6.14-rc3-rt2 Mark Knecht
2005-10-06  8:37                           ` 2.6.14-rc3-rt2 Ingo Molnar
2005-10-06  8:37                         ` 2.6.14-rc3-rt2 Steven Rostedt
2005-10-06  8:49                           ` 2.6.14-rc3-rt2 Ingo Molnar
2005-10-06  9:48                             ` 2.6.14-rc3-rt2 Steven Rostedt
2005-10-06 10:04                               ` 2.6.14-rc3-rt2 Andi Kleen
2005-10-06 11:08                                 ` [PATCH] cleanup u32 flags in acpi spin_lock calls Steven Rostedt
2005-10-06 16:25                                 ` 2.6.14-rc3-rt2 Luck, Tony
2005-10-06 15:26                             ` 2.6.14-rc3-rt2 Steven Rostedt
2005-10-07 11:08                               ` 2.6.14-rc3-rt2 Steven Rostedt
2005-10-07 11:15                                 ` 2.6.14-rc3-rt2 Ingo Molnar
2005-10-07 12:38                                   ` 2.6.14-rc3-rt2 Steven Rostedt
2005-10-07 13:26                                     ` 2.6.14-rc3-rt2 Ingo Molnar
2005-10-11  8:01                               ` 2.6.14-rc3-rt2 Ingo Molnar
2005-10-11  9:09                                 ` 2.6.14-rc3-rt12 Steven Rostedt
2005-10-11  9:12                                   ` 2.6.14-rc3-rt12 Ingo Molnar
2005-10-06  9:48                           ` 2.6.14-rc3-rt2 Andi Kleen
2005-10-06  9:36                         ` 2.6.14-rc3-rt2 Andi Kleen
2005-10-04 20:51               ` 2.6.14-rc3-rt2 Thomas Gleixner
2005-10-05  7:12               ` 2.6.14-rc3-rt2 Steven Rostedt
2005-10-05 14:50                 ` 2.6.14-rc3-rt2 Thomas Gleixner
2005-10-05  7:41               ` 2.6.14-rc3-rt2 Ingo Molnar
2005-10-04 20:38           ` 2.6.14-rc3-rt2 Mark Knecht
2005-10-06 14:31           ` 2.6.14-rc3-rt2 Lee Revell
2005-10-06 15:00             ` 2.6.14-rc3-rt2 Steven Rostedt
2005-10-05 13:36 ` 2.6.14-rc3-rt2 Steven Rostedt
2005-10-05 14:29   ` 2.6.14-rc3-rt2 Steven Rostedt
2005-10-05 15:41     ` 2.6.14-rc3-rt2 Daniel Walker
2005-10-05 15:48     ` 2.6.14-rc3-rt2 Thomas Gleixner
2005-10-05 15:58       ` 2.6.14-rc3-rt2 david singleton
2005-10-05 16:08         ` 2.6.14-rc3-rt2 Thomas Gleixner
2005-10-05 16:13           ` 2.6.14-rc3-rt2 david singleton
2005-10-05 15:58       ` 2.6.14-rc3-rt2 Ingo Molnar
2005-10-05 16:07         ` 2.6.14-rc3-rt2 Steven Rostedt
2005-10-07 11:05           ` Steven Rostedt [this message]
2005-10-07 11:14             ` 2.6.14-rc3-rt2 Ingo Molnar
2005-10-05 14:41   ` 2.6.14-rc3-rt2 Ingo Molnar
2005-10-05 20:24 ` 2.6.14-rc3-rt2 Esben Nielsen
2005-10-06 16:47 ` 2.6.14-rc3-rt2 Todd Kneisel
2005-10-06 17:18   ` 2.6.14-rc3-rt2 david singleton

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=Pine.LNX.4.58.0510070538170.6608@localhost.localdomain \
    --to=rostedt@goodmis.org \
    --cc=dsingleton@mvista.com \
    --cc=johnstul@us.ibm.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=mingo@elte.hu \
    --cc=tglx@linutronix.de \
    /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

all inboxes | Powered by JetHome®