From: Charles Lepple <clepple@ghz.cc>
To: Tony Lindgren <tony@atomide.com>
Cc: psavo@iki.fi, lkml <linux-kernel@vger.kernel.org>,
john stultz <johnstul@us.ibm.com>
Subject: Re: [PATCH] amd76x_pm on 2.6.0-test9 cleanup
Date: Tue, 4 Nov 2003 17:38:33 -0500 [thread overview]
Message-ID: <9F0055D6-0F17-11D8-A943-003065DC6B50@ghz.cc> (raw)
In-Reply-To: <20031104200517.GD1042@atomide.com>
On Tuesday, November 4, 2003, at 03:05 PM, Tony Lindgren wrote:
> * Charles Lepple <clepple@ghz.cc> [031104 11:45]:
>> On Tuesday 04 November 2003 02:15 pm, Tony Lindgren wrote:
>>> I've heard of timing problems if it's compiled in, but supposedly
>>> they
>>> don't happen when loaded as module.
>>
>> In some of the earlier testX versions of the kernel, I did not see any
>> difference between compiling as a module, and compiling into the
>> kernel. (It
>> is currently a module on my system.)
>>
>> I did, however, manage to keep ntpd happy by reducing HZ to 100. Even
>> raising
>> HZ to 200 is enough to throw off its PLL. The machine is idle for 90%
>> of the
>> day, though, so I don't know if the PLL is adapting to the fact that
>> the
>> system is idling, but the values for tick look reasonable.
>
> Interesting, sounds like the idling causes missed timer interrupts?
> Can you
> briefly describe what's the easiest way to reproduce the timer
> problem, just
> change HZ to 200 and look at the system time?
Weird. On -test9-bk at HZ=1000, with amd76x_pm loaded as a module
(lazy_idle=800, the default), the system clock is running fast.
With ntpd running, the clock was stepped back 2.5 seconds twice in 20
minutes.
Here's what I get from adjtimexconfig (after stopping ntpd, of course):
# adjtimexconfig
Comparing clocks (this will take 70 sec)... adjusting system time by
-126.211 sec/day
Done
Now tick is 9985. I distinctly remember it being somewhat over 10,000
the last time I ran with HZ=1000 and amd_76x_pm active. With HZ=100,
adjtimexconfig sets tick=10002.
I'm not entirely sure what the "acpi" interrupt is doing-- it
increments about once every two seconds when the system is idle, and
various types of system activity make it happen more frequently. At
least I'm not getting any "irq 9: nobody cared!" messages anymore (the
button module is loaded, so I guess it is handling it). If I don't have
amd76x_pm loaded, the acpi interrupt is triggered a couple of times
after button is loaded, but then it doesn't happen again until I
actually press a button.
--
Charles Lepple
ghz.cc! clepple
next prev parent reply other threads:[~2003-11-04 22:39 UTC|newest]
Thread overview: 17+ messages / expand[flat|nested] mbox.gz Atom feed top
2003-11-04 0:22 [PATCH] amd76x_pm on 2.6.0-test9 cleanup Tony Lindgren
2003-11-04 18:41 ` john stultz
2003-11-04 19:15 ` Tony Lindgren
2003-11-04 19:44 ` Charles Lepple
2003-11-04 20:05 ` Tony Lindgren
2003-11-04 20:52 ` Charles Lepple
2003-11-04 22:38 ` Charles Lepple [this message]
2003-11-04 23:11 ` Tony Lindgren
2003-11-04 23:46 ` Pasi Savolainen
2003-11-04 20:21 ` Pasi Savolainen
2003-11-04 20:55 ` Tony Lindgren
2003-11-04 20:57 ` john stultz
2003-11-04 21:14 ` Tony Lindgren
2003-11-14 11:56 ` Pasi Savolainen
2003-11-04 21:01 ` Charles Lepple
2003-11-04 21:17 ` Tony Lindgren
2003-11-05 19:42 ` Felix Maibaum
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=9F0055D6-0F17-11D8-A943-003065DC6B50@ghz.cc \
--to=clepple@ghz.cc \
--cc=johnstul@us.ibm.com \
--cc=linux-kernel@vger.kernel.org \
--cc=psavo@iki.fi \
--cc=tony@atomide.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).