public inbox for linux-kernel@vger.kernel.org
 help / color / mirror / Atom feed
From: Martin Schwidefsky <schwidefsky@de.ibm.com>
To: Thomas Gleixner <tglx@linutronix.de>
Cc: Ingo Molnar <mingo@elte.hu>,
	Peter Zijlstra <a.p.zijlstra@chello.nl>,
	john stultz <johnstul@us.ibm.com>,
	linux-kernel@vger.kernel.org
Subject: Re: [circular locking bug] Re: [patch 00/15] clocksource / timekeeping rework V4 (resend V3 + bug fix)
Date: Thu, 20 Aug 2009 18:53:44 +0200	[thread overview]
Message-ID: <20090820185344.1c0e48f8@skybase> (raw)
In-Reply-To: <alpine.LFD.2.00.0908201809190.3361@localhost.localdomain>

On Thu, 20 Aug 2009 18:14:07 +0200 (CEST)
Thomas Gleixner <tglx@linutronix.de> wrote:

> On Thu, 20 Aug 2009, Martin Schwidefsky wrote:
> > On Thu, 20 Aug 2009 11:58:21 +0200
> > > There should be a time management kernel thread instead 
> > > (or workqueue), which does a proper state machine of all 
> > > these properties - without having to call this stuff from 
> > > within a timer handler.
> > 
> > We could use that time managment kernel thread for the watchdog
> > downgrade as well. Dunno if it is worth to create another kernel thread
> > that just sits there doing nothing for 99.9% of the time.
> > 
> > As for the fix: my brains starts to hurt looking at the pit clocksource
> > code. Why does it set CLOCK_EVT_FEAT_ONESHOT but then unregisters the
> > clocksource when the mode is set to CLOCK_EVT_MODE_ONESHOT?? That does
> > not make any sense to me. I would have expected that the pit does not
> > set CLOCK_EVT_MODE_ONESHOT. The timekeeping code wouldn't try use the
> > clock for one-shot if the bit is not set. And to unregister the clock
> > only because the mode is set to shutdown or unused doesn't seem to be
> > necessary either. My fix would be to remove the CLOCK_EVT_MODE_ONESHOT
> > bit from the features mask and to remove the clocksource_unregister
> > from the set_mode callback.
> 
> No. On UP machines we can run in oneshot mode with the PIT. 
> 
> The disable PIT clocksource hack was done in commit
> 1a0c009ac53de4a7664a1239936f0bc258133156. Yes, in hindsight we should
> have done 3f68535adad8dd89499505a65fb25d0e02d118cc in the first place.
> 
> So now the simple and correct fix is to remove the unregister call.

Sorry if I don't see the obvious but how can that work? If the pit as
clocksource is switched to CLOCK_EVT_MODE_ONESHOT it simply vanishes.
How is it possible that a UP machine runs with the PIT in oneshot mode?

And yes, hindsight is easier than foresight.

-- 
blue skies,
   Martin.

"Reality continues to ruin my life." - Calvin.


  reply	other threads:[~2009-08-20 16:53 UTC|newest]

Thread overview: 78+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2009-08-14 13:47 [patch 00/15] clocksource / timekeeping rework V4 (resend V3 + bug fix) Martin Schwidefsky
2009-08-14 13:47 ` [patch 01/15] introduce timekeeping_leap_insert Martin Schwidefsky
2009-08-15  9:01   ` [tip:timers/core] timekeeping: Introduce timekeeping_leap_insert tip-bot for John Stultz
2009-08-14 13:47 ` [patch 02/15] remove clocksource inline functions Martin Schwidefsky
2009-08-15  9:01   ` [tip:timers/core] timekeeping: Remove " tip-bot for Martin Schwidefsky
2009-08-14 13:47 ` [patch 03/15] reset of cycle_last for tsc clocksource Martin Schwidefsky
2009-08-15  9:01   ` [tip:timers/core] timekeeping: Move reset of cycle_last for tsc clocksource to tsc tip-bot for Martin Schwidefsky
2009-08-14 13:47 ` [patch 04/15] cleanup clocksource selection Martin Schwidefsky
2009-08-15  1:42   ` john stultz
2009-08-15  1:43     ` john stultz
2009-08-17  7:34     ` Martin Schwidefsky
2009-08-15  9:02   ` [tip:timers/core] clocksource: Cleanup " tip-bot for Martin Schwidefsky
2009-08-14 13:47 ` [patch 05/15] clocksource watchdog highres enablement Martin Schwidefsky
2009-08-15  9:02   ` [tip:timers/core] clocksource: Delay " tip-bot for Martin Schwidefsky
2009-08-14 13:47 ` [patch 06/15] clocksource watchdog resume logic Martin Schwidefsky
2009-08-15  9:02   ` [tip:timers/core] clocksource: Simplify " tip-bot for Martin Schwidefsky
2009-08-14 13:47 ` [patch 07/15] clocksource watchdog refactoring Martin Schwidefsky
2009-08-15  9:02   ` [tip:timers/core] clocksource: Refactor clocksource watchdog tip-bot for Martin Schwidefsky
2009-08-14 13:47 ` [patch 08/15] clocksource watchdog work Martin Schwidefsky
2009-08-15  9:03   ` [tip:timers/core] clocksource: Move watchdog downgrade to a work queue thread tip-bot for Martin Schwidefsky
2009-08-14 13:47 ` [patch 09/15] introduce struct timekeeper Martin Schwidefsky
2009-08-15  9:03   ` [tip:timers/core] timekeeping: Introduce " tip-bot for Martin Schwidefsky
2009-08-14 13:47 ` [patch 10/15] add xtime_shift and ntp_error_shift to " Martin Schwidefsky
2009-08-15  9:03   ` [tip:timers/core] timekeeping: Add " tip-bot for Martin Schwidefsky
2009-08-15  9:04   ` [patch 10/15] add " Thomas Gleixner
2009-08-14 13:47 ` [patch 11/15] move NTP adjusted clock multiplier " Martin Schwidefsky
2009-08-15  9:03   ` [tip:timers/core] timekeeping: Move " tip-bot for Martin Schwidefsky
2009-08-14 13:47 ` [patch 12/15] timekeeper read clock helper functions Martin Schwidefsky
2009-08-15  9:03   ` [tip:timers/core] timekeeping: Add timekeeper read_clock " tip-bot for Martin Schwidefsky
2009-08-14 13:47 ` [patch 13/15] update clocksource with stop_machine Martin Schwidefsky
2009-08-15  9:04   ` [tip:timers/core] timekeeping: Update " tip-bot for Martin Schwidefsky
2009-08-14 13:47 ` [patch 14/15] read_persistent_clock should return a timespec Martin Schwidefsky
2009-08-15  9:04   ` [tip:timers/core] timekeeping: Increase granularity of read_persistent_clock() tip-bot for Martin Schwidefsky
2009-08-22 10:32     ` Ingo Molnar
2009-08-22 15:15       ` Martin Schwidefsky
2009-08-22 15:33         ` Ingo Molnar
2009-08-22 20:23           ` Martin Schwidefsky
2009-08-23  8:53             ` Ingo Molnar
2009-08-23  9:03             ` [tip:timers/core] timekeeping: Increase granularity of read_persistent_clock(), build fix tip-bot for Martin Schwidefsky
2009-08-23  3:33           ` [tip:timers/core] timekeeping: Increase granularity of read_persistent_clock() Paul Mackerras
2009-08-23  8:47             ` Ingo Molnar
2009-08-24  3:20               ` Paul Mackerras
2009-08-24  8:23                 ` Ingo Molnar
2009-08-25  3:49                   ` Paul Mackerras
2009-08-25  8:26                     ` Ingo Molnar
2009-08-25  9:57                       ` Paul Mackerras
2009-08-25 10:17                         ` Ingo Molnar
2009-08-25 11:33                           ` Paul Mackerras
2009-08-25 13:50                             ` Ingo Molnar
2009-08-25 21:33                               ` Theodore Tso
2009-08-25 22:03                                 ` Ingo Molnar
2009-08-26  0:26                                   ` Paul Mackerras
2009-08-26  0:22                                 ` Paul Mackerras
2009-08-25 23:48                               ` Paul Mackerras
2009-08-26  9:44                               ` Benjamin Herrenschmidt
2009-08-14 13:47 ` [patch 15/15] introduce read_boot_clock Martin Schwidefsky
2009-08-15  9:04   ` [tip:timers/core] timekeeping: Introduce read_boot_clock tip-bot for Martin Schwidefsky
2009-08-14 14:08 ` [patch 00/15] clocksource / timekeeping rework V4 (resend V3 + bug fix) Thomas Gleixner
2009-08-14 14:22   ` Martin Schwidefsky
2009-08-14 22:56 ` john stultz
2009-08-15  1:46 ` john stultz
2009-08-15  9:01   ` Thomas Gleixner
2009-08-15  9:52     ` Ingo Molnar
2009-08-15 10:08       ` Thomas Gleixner
2009-08-17  7:40         ` Martin Schwidefsky
2009-08-17  8:45           ` Thomas Gleixner
2009-08-17  9:28             ` [circular locking bug] " Ingo Molnar
2009-08-17 17:53               ` Martin Schwidefsky
2009-08-18 15:09               ` Martin Schwidefsky
2009-08-19 10:06                 ` [tip:timers/core] clocksource: Avoid clocksource watchdog circular locking dependency tip-bot for Martin Schwidefsky
2009-08-19 20:25                 ` [circular locking bug] Re: [patch 00/15] clocksource / timekeeping rework V4 (resend V3 + bug fix) Ingo Molnar
2009-08-20  9:28                   ` Martin Schwidefsky
2009-08-20  9:58                     ` Ingo Molnar
2009-08-20 10:35                       ` Martin Schwidefsky
2009-08-20 16:14                         ` Thomas Gleixner
2009-08-20 16:53                           ` Martin Schwidefsky [this message]
2009-08-20 19:08                             ` Thomas Gleixner
2009-08-19  9:46             ` [tip:timers/core] clocksource: Protect the watchdog rating changes with clocksource_mutex tip-bot for Thomas Gleixner

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=20090820185344.1c0e48f8@skybase \
    --to=schwidefsky@de.ibm.com \
    --cc=a.p.zijlstra@chello.nl \
    --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