Re: 2.6.32-rc5-mmotm1101 - lockdep whinge during early boot

[Thomas Gleixner <tglx@linutronix.de>]

On Thu, 5 Nov 2009, Thomas Gleixner wrote:
> On Thu, 5 Nov 2009, Martin Schwidefsky wrote:
> > >
> > > Which basically reverts commit 7285dd7fd375763bfb8ab1ac9cf3f1206f503c16
> > >
> > 
> > It is this commit in linux-next:
> > 
> > commit 66b1a72ceaa7da811db025855f19482f0a38e815
> > Author: Ingo Molnar <mingo@elte.hu>
> > Date:   Thu Sep 10 19:39:52 2009 +0200
> > 
> >     Revert "clocksource: Resolve cpu hotplug dead lock with TSC unstable"
> > 
> >     This reverts commit 7285dd7fd375763bfb8ab1ac9cf3f1206f503c16.
> > 
> >     causes crashes.
> > 
> >     Signed-off-by: Ingo Molnar <mingo@elte.hu>
> 
> Eeek. We fixed this in mainline via:
> 
> commit f79e0258ea1f04d63db499479b5fb855dff6dbc5
> Author: Martin Schwidefsky <schwidefsky@de.ibm.com>
> Date:   Fri Sep 11 15:33:05 2009 +0200
> 
>     clocksource: Resolve cpu hotplug dead lock with TSC unstable, fix crash
> 
> and
> 
> commit 54a6bc0b071c50150bc6d1da16c2cd9a963e288c
> Author: Thomas Gleixner <tglx@linutronix.de>
> Date:   Mon Sep 14 19:49:02 2009 +0200
> 
>     clocksource: Delay clocksource down rating to late boot
> 
> 
> These patches came via the tip tree right at the beginning of the
> merge window several days _before_ I asked Linus to pull the timer
> code for 32.
> 
> So there is something weird going on.

Ok, figured out what happened.

The revert has been in next via the tip-tree for some days when
testing turned up the problem and the fixes were not ready.

So it looks like Andrew picked exactly such a linux-next snapshot to
create his baseline patch.

linux-next, tip, mainline of today all have the correct and functional
code.

Thanks,

	tglx