Re: Regression from 2.6.28-rc7: Commit 7cd5b08be3 breaks startup on Toshiba Portege R500

[Linus Torvalds <torvalds@linux-foundation.org>]

On Wed, 3 Dec 2008, Rafael J. Wysocki wrote:
> 
> commit 7cd5b08be3c489df11b559fef210b81133764ad4
> Author: Wim Van Sebroeck <wim@iguana.be>
> Date:   Wed Nov 19 19:39:58 2008 +0000
> 
>     [WATCHDOG] iTCO_wdt : problem with rebooting on new ICH9 based motherboards
> 
>     Bugzilla #9868: On Intel motherboards with the ICH9 based I/O controllers
>     (Like DP35DP and DG33FB) the iTCO timer counts but it doesn't reboot the
>     system after the counter expires.
> 
>     This patch fixes this by moving the enabling & disabling of the TCO_EN bit
>     in the SMI_EN register into the start and stop code.
> 
>     Signed-off-by: Wim Van Sebroeck <wim@iguana.be>
> 
> makes CPU fan spin with full speed and causes hald-addon-acpi to take 100% of
> CPU time on Toshiba Portege R500 with openSUSE 11.0 user space.  This starts as
> soon as hal starts or thereabouts and never ends, which makes the current
> mainline unusable on that box.

Hmm. I wonder if the bug is this one-liner (it really looks like an 
idiotic typo).

Can you test?

		Linus
---
 drivers/watchdog/iTCO_wdt.c |    2 +-
 1 files changed, 1 insertions(+), 1 deletions(-)

diff --git a/drivers/watchdog/iTCO_wdt.c b/drivers/watchdog/iTCO_wdt.c
index 26173a2..5b395a4 100644
--- a/drivers/watchdog/iTCO_wdt.c
+++ b/drivers/watchdog/iTCO_wdt.c
@@ -392,7 +392,7 @@ static int iTCO_wdt_stop(void)
 
 	/* Bit 13: TCO_EN -> 1 = Enables the TCO logic to generate SMI# */
 	val32 = inl(SMI_EN);
-	val32 &= 0x00002000;
+	val32 |= 0x00002000;
 	outl(val32, SMI_EN);
 
 	/* Set the NO_REBOOT bit to prevent later reboots, just for sure */