From: john stultz <johnstul@us.ibm.com>
To: Roman Zippel <zippel@linux-m68k.org>
Cc: Ingo Molnar <mingo@elte.hu>, Thomas Gleixner <tglx@linutronix.de>,
Andrew Morton <akpm@osdl.org>,
linux-kernel@vger.kernel.org
Subject: Re: [RFC] HZ free ntp
Date: Tue, 19 Dec 2006 17:32:36 -0800 [thread overview]
Message-ID: <1166578357.5594.3.camel@localhost> (raw)
In-Reply-To: <Pine.LNX.4.64.0612132125450.1867@scrub.home>
On Wed, 2006-12-13 at 21:40 +0100, Roman Zippel wrote:
> On Wed, 13 Dec 2006, john stultz wrote:
> > > You cannot choose arbitrary intervals otherwise you get other problems,
> > > e.g. with your patch time_offset handling is broken.
> >
> > I'm not seeing this yet. Any more details?
>
> time_offset is scaled to HZ in do_adjtimex, which needs to be changed as
> well.
Ah, thanks! Fixed.
> > > You don't have to introduce anything new, it's tick_length that changes
> > > and HZ that becomes a variable in this function.
> >
> > So, forgive me for rehashing this, but it seems we're cross talking
> > again. The context here is the dynticks code. Where HZ doesn't change,
> > but we get interrupts at much reduced rates.
>
> I know and all you have to change in the ntp and some related code is to
> replace HZ there with a variable, thus make it changable, so you can
> increase the update interval (i.e. it becomes 1s/hz instead of 1s/HZ).
Untested patch below. Does this vibe better with you are suggesting?
Any other suggestions or feedback?
thanks
-john
diff --git a/include/linux/timex.h b/include/linux/timex.h
index db501dc..8241e6e 100644
--- a/include/linux/timex.h
+++ b/include/linux/timex.h
@@ -286,6 +286,9 @@ #endif /* !CONFIG_TIME_INTERPOLATION */
#define TICK_LENGTH_SHIFT 32
+#define NTP_INTERVAL_FREQ (HZ)
+#define NTP_INTERVAL_LENGTH (NSEC_PER_SEC/NTP_INTERVAL_FREQ)
+
/* Returns how long ticks are at present, in ns / 2^(SHIFT_SCALE-10). */
extern u64 current_tick_length(void);
diff --git a/kernel/time/ntp.c b/kernel/time/ntp.c
index 3afeaa3..eb12509 100644
--- a/kernel/time/ntp.c
+++ b/kernel/time/ntp.c
@@ -24,7 +24,7 @@ static u64 tick_length, tick_length_base
#define MAX_TICKADJ 500 /* microsecs */
#define MAX_TICKADJ_SCALED (((u64)(MAX_TICKADJ * NSEC_PER_USEC) << \
- TICK_LENGTH_SHIFT) / HZ)
+ TICK_LENGTH_SHIFT) / NTP_INTERVAL_FREQ)
/*
* phase-lock loop variables
@@ -46,13 +46,17 @@ #define CLOCK_TICK_ADJUST (((s64)CLOCK_T
static void ntp_update_frequency(void)
{
- tick_length_base = (u64)(tick_usec * NSEC_PER_USEC * USER_HZ) << TICK_LENGTH_SHIFT;
- tick_length_base += (s64)CLOCK_TICK_ADJUST << TICK_LENGTH_SHIFT;
- tick_length_base += (s64)time_freq << (TICK_LENGTH_SHIFT - SHIFT_NSEC);
+ u64 second_length = (u64)(tick_usec * NSEC_PER_USEC * USER_HZ)
+ << TICK_LENGTH_SHIFT;
+ second_length += (s64)CLOCK_TICK_ADJUST << TICK_LENGTH_SHIFT;
+ second_length += (s64)time_freq << (TICK_LENGTH_SHIFT - SHIFT_NSEC);
- do_div(tick_length_base, HZ);
+ tick_length_base = second_length;
- tick_nsec = tick_length_base >> TICK_LENGTH_SHIFT;
+ do_div(second_length, HZ);
+ tick_nsec = second_length >> TICK_LENGTH_SHIFT;
+
+ do_div(tick_length_base, NTP_INTERVAL_FREQ);
}
/**
@@ -162,7 +166,7 @@ void second_overflow(void)
tick_length -= MAX_TICKADJ_SCALED;
} else {
tick_length += (s64)(time_adjust * NSEC_PER_USEC /
- HZ) << TICK_LENGTH_SHIFT;
+ NTP_INTERVAL_FREQ) << TICK_LENGTH_SHIFT;
time_adjust = 0;
}
}
@@ -239,7 +243,8 @@ #endif
result = -EINVAL;
goto leave;
}
- time_freq = ((s64)txc->freq * NSEC_PER_USEC) >> (SHIFT_USEC - SHIFT_NSEC);
+ time_freq = ((s64)txc->freq * NSEC_PER_USEC)
+ >> (SHIFT_USEC - SHIFT_NSEC);
}
if (txc->modes & ADJ_MAXERROR) {
@@ -309,7 +314,8 @@ #endif
freq_adj += time_freq;
freq_adj = min(freq_adj, (s64)MAXFREQ_NSEC);
time_freq = max(freq_adj, (s64)-MAXFREQ_NSEC);
- time_offset = (time_offset / HZ) << SHIFT_UPDATE;
+ time_offset = (time_offset / NTP_INTERVAL_FREQ)
+ << SHIFT_UPDATE;
} /* STA_PLL */
} /* txc->modes & ADJ_OFFSET */
if (txc->modes & ADJ_TICK)
@@ -324,8 +330,10 @@ leave: if ((time_status & (STA_UNSYNC|ST
if ((txc->modes & ADJ_OFFSET_SINGLESHOT) == ADJ_OFFSET_SINGLESHOT)
txc->offset = save_adjust;
else
- txc->offset = shift_right(time_offset, SHIFT_UPDATE) * HZ / 1000;
- txc->freq = (time_freq / NSEC_PER_USEC) << (SHIFT_USEC - SHIFT_NSEC);
+ txc->offset = shift_right(time_offset, SHIFT_UPDATE)
+ * NTP_INTERVAL_FREQ / 1000;
+ txc->freq = (time_freq / NSEC_PER_USEC)
+ << (SHIFT_USEC - SHIFT_NSEC);
txc->maxerror = time_maxerror;
txc->esterror = time_esterror;
txc->status = time_status;
diff --git a/kernel/timer.c b/kernel/timer.c
index 0256ab4..616b7fd 100644
--- a/kernel/timer.c
+++ b/kernel/timer.c
@@ -890,7 +890,7 @@ void __init timekeeping_init(void)
ntp_clear();
clock = clocksource_get_next();
- clocksource_calculate_interval(clock, tick_nsec);
+ clocksource_calculate_interval(clock, NTP_INTERVAL_LENGTH);
clock->cycle_last = clocksource_read(clock);
write_sequnlock_irqrestore(&xtime_lock, flags);
@@ -1092,7 +1092,7 @@ #endif
if (change_clocksource()) {
clock->error = 0;
clock->xtime_nsec = 0;
- clocksource_calculate_interval(clock, tick_nsec);
+ clocksource_calculate_interval(clock, NTP_INTERVAL_LENGTH);
}
}
next prev parent reply other threads:[~2006-12-20 1:36 UTC|newest]
Thread overview: 85+ messages / expand[flat|nested] mbox.gz Atom feed top
2006-12-05 4:40 -mm merge plans for 2.6.20 Andrew Morton
2006-12-05 4:56 ` Jeff Garzik
2006-12-05 5:41 ` Andrew Morton
2006-12-05 7:04 ` Jens Axboe
2006-12-05 15:00 ` Mark Lord
2006-12-06 19:19 ` Conke Hu
2006-12-06 19:26 ` Randy Dunlap
2006-12-06 19:40 ` Jeff Garzik
2006-12-06 22:36 ` Andrew Morton
2006-12-05 5:14 ` Paul Mackerras
2006-12-05 5:42 ` Andrew Morton
2006-12-05 5:53 ` Nick Piggin
2006-12-05 5:49 ` Nick Piggin
2006-12-05 8:36 ` Gautham R Shenoy
2006-12-05 8:47 ` Peter Zijlstra
2006-12-05 11:06 ` ext2 future [was Re: -mm merge plans for 2.6.20] Pavel Machek
2006-12-05 13:23 ` -mm merge plans for 2.6.20 John W. Linville
2006-12-05 14:27 ` Roman Zippel
2006-12-06 3:46 ` Horst Schirmeier
2006-12-05 16:02 ` Dave Jones
2006-12-12 17:49 ` Dave Jones
2006-12-19 5:20 ` Nick Piggin
2006-12-19 6:44 ` Dave Jones
2006-12-19 7:02 ` Nick Piggin
2007-01-07 17:36 ` page_mapcount(page) went negative Dave Jones
2007-01-10 23:53 ` Nick Piggin
2006-12-05 17:35 ` -mm merge plans for 2.6.20 James Simmons
2006-12-05 18:01 ` Andrew Morton
2006-12-05 18:25 ` James Simmons
2006-12-05 18:37 ` [PATCH] backlight sysfs change to the fbdev drivers James Simmons
2006-12-05 19:43 ` -mm merge plans for 2.6.20 Andrew Morton
2006-12-05 19:59 ` James Simmons
2006-12-05 20:20 ` Andrew Morton
2006-12-05 21:34 ` James Simmons
2006-12-06 23:40 ` Andrew Morton
2006-12-07 14:31 ` James Simmons
2006-12-05 20:40 ` Miguel Ojeda
2006-12-06 14:42 ` James Simmons
2006-12-05 19:18 ` Josef Sipek
2006-12-05 19:21 ` [PATCH 1/2] fsstack: Make fsstack_copy_attr_all copy inode size Josef Sipek
2006-12-05 19:22 ` [PATCH 2/2] fsstack: Fix up ecryptfs's fsstack usage Josef Sipek
2006-12-05 22:28 ` Andrew Morton
2006-12-05 22:38 ` Josef Sipek
2006-12-05 22:49 ` Andrew Morton
2006-12-05 23:16 ` Josef Sipek
2006-12-05 21:00 ` -mm merge plans for 2.6.20 Ingo Molnar
2006-12-05 21:17 ` -mm merge plans for 2.6.20, scheduler bits Ingo Molnar
2006-12-05 20:59 ` Siddha, Suresh B
2006-12-05 21:47 ` Ingo Molnar
2006-12-05 21:29 ` Miguel Ojeda
2006-12-06 2:59 ` -mm merge plans for 2.6.20 Roman Zippel
2006-12-06 4:30 ` Andrew Morton
2006-12-06 8:32 ` Thomas Gleixner
2006-12-06 12:54 ` Roman Zippel
2006-12-06 13:11 ` Ingo Molnar
2006-12-06 14:33 ` Roman Zippel
2006-12-06 15:22 ` Ingo Molnar
2006-12-06 16:42 ` Roman Zippel
2006-12-06 16:58 ` Ingo Molnar
2006-12-06 16:59 ` Ingo Molnar
2006-12-12 20:40 ` [RFC] HZ free ntp john stultz
2006-12-13 9:51 ` Ingo Molnar
2006-12-13 18:48 ` john stultz
2006-12-13 13:47 ` Roman Zippel
2006-12-13 19:19 ` john stultz
2006-12-13 20:40 ` Roman Zippel
2006-12-20 1:32 ` john stultz [this message]
2006-12-20 1:54 ` john stultz
2006-12-21 4:26 ` Andrew Morton
2007-01-01 18:29 ` Roman Zippel
2007-01-02 19:46 ` john stultz
2007-01-02 20:50 ` john stultz
2007-01-06 16:56 ` Roman Zippel
2007-01-22 19:27 ` [patch] HZ-free NTP Ingo Molnar
2007-01-22 19:39 ` Ingo Molnar
2007-01-01 16:27 ` [RFC] HZ free ntp Roman Zippel
2007-01-02 19:42 ` john stultz
2007-01-06 16:46 ` Roman Zippel
2006-12-06 12:33 ` -mm merge plans for 2.6.20 Roman Zippel
2006-12-08 14:09 ` Stephen Smalley
2006-12-08 20:58 ` Andrew Morton
2006-12-10 15:07 ` Mimi Zohar
2006-12-09 9:30 ` Randy Dunlap
2006-12-09 9:44 ` Andrew Morton
2006-12-10 20:12 ` Randy Dunlap
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=1166578357.5594.3.camel@localhost \
--to=johnstul@us.ibm.com \
--cc=akpm@osdl.org \
--cc=linux-kernel@vger.kernel.org \
--cc=mingo@elte.hu \
--cc=tglx@linutronix.de \
--cc=zippel@linux-m68k.org \
/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