From mboxrd@z Thu Jan 1 00:00:00 1970 From: Adam Nielsen Date: Tue, 20 Oct 2009 11:50:31 +0000 Subject: Re: [lm-sensors] Gigabyte GA-P35-DS3P - VID Zero value Message-Id: <4ADDA407.8030306@shikadi.net> List-Id: References: <20091019164516.701032uxlen5ai2k@mail.simonandkate.net> In-Reply-To: <20091019164516.701032uxlen5ai2k@mail.simonandkate.net> MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit To: lm-sensors@vger.kernel.org > The raw VID value is at 0xfc, value is 0x3f, which means all VID pins > are high. Which means they are not wired to the CPU, otherwise at least > one of them would be low. > > Now, looking at register 0x27, it appears that all VID input pins have > been configured for their alternative function (GPIO.) This is not the > default value for this register, which means that the motherboard > vendor decided to not use these pins for VID input but for another > function. > > BTW... the pins in question are in input mode. So who knows... maybe > they ARE used for VID monitoring, and GPIO was preferred over true VID > for obscure reasons (incompatible voltage levels?) You may want to run, > as root: > > isadump -f 0x800 16 > > and see if the 3rd value would make sense as a VID value for your CPU. Ah interesting. It doesn't seem to for me though: 0 1 2 3 4 5 6 7 8 9 a b c d e f 0800: ff 01 cf fa f7 07 ff ff ff ff ff ff ff ff ff ff VRM v11 (in hwmon-vid) says values can't be > 0xb2, although I don't know how many bits are wired up to the VID. So is 0x802 the GPIO input? I wonder what it's wired up to...? > One thing we could do is teach the it87 driver to not export the VID > value if any of VID pins 0-3 are configured for GPIO function. That > wouldn't be too difficult, but would you be able to test a kernel patch? I can test an it87 patch if you need a guinea pig - especially if it'll compile against 2.6.30 so I don't have to reboot ;-) Cheers, Adam. _______________________________________________ lm-sensors mailing list lm-sensors@lm-sensors.org http://lists.lm-sensors.org/mailman/listinfo/lm-sensors