Re: [Bug #12162] Commit 7cd5b08be3 breaks startup on Toshiba Portege R500

From: Wim Van Sebroeck <wim@iguana.be>
To: Linus Torvalds <torvalds@linux-foundation.org>
Cc: "Rafael J. Wysocki" <rjw@sisk.pl>,
	Linux Kernel Mailing List <linux-kernel@vger.kernel.org>,
	Kernel Testers List <kernel-testers@vger.kernel.org>,
	Frans Pop <elendil@planet.nl>
Subject: Re: [Bug #12162] Commit 7cd5b08be3 breaks startup on Toshiba Portege R500
Date: Thu, 4 Dec 2008 23:45:47 +0100	[thread overview]
Message-ID: <20081204224547.GA2212@infomag.infomag.iguana.be> (raw)
In-Reply-To: <20081204080537.GI7017@infomag.infomag.iguana.be>

Hi All,

> I will test the watchdog part tonight on an ICH4 and ICH9 chipset.
> I indeed moved the code over from the supermicro_old_pre_stop code.
> And I'm sure this fix indeed needs to go in.

Test on ICH4 is OK.
Test on ICH9 -> the watchdog timer runs but doesn't reboot the system anymore.
The following bits are active in SMI_EN and thus need to be investigated:
	13 - TCO_EN (Not relevant since we set/unset this allready)
	 5 - APMC_EN
	 4 - SLP_SMI_EN
	 3 - LEGACY_USB_EN
	 1 - End Of SMI
	 0 - GBL_SMI_EN

A first test indicated that if bit 0 is forced to 0, the system reboots again.
The second test reveals that setting bits 5, 4, 3 and 1 to 0 doesn't help the rebooting of the ICH9 chipset.
So the real issue with ICH9 and rebooting has to do with the GBL_SMI_EN bit.

So next step will be to create an additional patch and then test it with Rafael and Frans...

Kind regards,
Wim.