From mboxrd@z Thu Jan 1 00:00:00 1970 From: "Ian Dobson" Date: Wed, 21 Dec 2011 17:55:31 +0000 Subject: Re: [lm-sensors] NCT6776 global registers 0x1c=0 and 0x24=0x5c Message-Id: List-Id: References: <20110410160433.GA22075@localhost> In-Reply-To: <20110410160433.GA22075@localhost> MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit To: lm-sensors@vger.kernel.org Hi, -------------------------------------------------- From: "Guenter Roeck" Sent: Wednesday, December 21, 2011 6:01 PM To: "Harald Judt" Cc: Subject: Re: [lm-sensors] NCT6776 global registers 0x1c=0 and 0x24=0x5c onASRock Extreme4 > Hi, > On Wed, Dec 21, 2011 at 09:50:31AM -0500, Harald Judt wrote: >> Hi, >> >> > On Tue, 12 Apr 2011 19:02:19 +0200, Ian Dobson wrote: >> >> Hi, >> >> >> >> -------------------------------------------------- >> >> From: "Mike Campin" >> >> Sent: Sunday, April 10, 2011 6:04 PM >> >> To: >> >> Subject: [lm-sensors] NCT6776 global registers 0x1c=0 and 0x24=0x5c >> >> onASRock >> >> Extreme4 >> >> >> >> > Hi, >> >> > >> >> > Only 2 of the 5 fans are showing up on my ASRock P67 Extreme4 >> >> > motherboard. >> >> > It due to the NCT6776 global configuration register 0x1c and 0x24 >> >> > values. >> >> > The fans work if I modify these registers prior to loading the >> >> > w83627ehf >> >> > driver. >> >> > >> >> > My question is where are these registers initialized? Should I ask >> >> > ASRock >> >> > to fix the BIOS? >> >> > >> >> > isadump -y -k 0x87,0x87 0x2e 0x2f >> >> > >> >> > 0 1 2 3 4 5 6 7 8 9 a b c d e f >> >> > 00: ff ff 00 ff ff ff ff 02 ff ff ff ff ff ff ff ff >> >> > 10: ff ff ff ff ff ff ff ff ff ff f8 0e 00 00 ff ff >> >> > 20: c3 33 ff 00 5c 00 00 98 00 ff 20 00 80 00 00 01 >> >> > 30: 00 ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff >> >> > >> >> > isaset -y -f 0x2e 0x87 >> >> > isaset -y -f 0x2e 0x87 >> >> > isaset -y 0x2e 0x2f 0x1c 0x3 >> >> > isaset -y 0x2e 0x2f 0x24 0x1c >> >> > >> >> > Thanks, Mike >> >> > >> >> > -- >> >> > Mike Campin >> >> > lm_sensors at ootsa.homelinux.net >> >> > >> >> > _______________________________________________ >> >> > lm-sensors mailing list >> >> > lm-sensors at lm-sensors.org >> >> > http://lists.lm-sensors.org/mailman/listinfo/lm-sensors >> >> >> >> ASRock designed the board, so they should setup the hardware >> >> monitoring chip >> >> correctly. Allowing users to change the chip configuration is a really >> >> bad >> >> idea, as many of the pins can have different functionality/electrical >> >> spcifications depending on the config. Setting up a pin incorrectly >> >> could >> >> blow the chip (Configuring a pin so that it's a source that's >> >> connected to >> >> an unprotected/limited sink for example). >> >> >> >> So ASRock should fix the BIOS. >> > >> > Seconded. But before flaming Asrock, two things worth checking: >> > * Availability of a BIOS update fixing the issue. >> > * Options in the BIOS to enable/disable the monitoring of specific >> > fans. I can imagine that the BIOS skips the configuration steps for >> > fans for which monitoring was disabled (although that would probably >> > mean more code than just doing it unconditionally...) >> > >> > -- >> > Jean Delvare >> >> I own an Asrock P67 Extreme4 too and have the same problem. It can be >> solved by following Mike Campin's instructions, which I successfully >> incorporated into /etc/init.d/lm_sensors. There is no BIOS available >> that fixes this issue, but I noticed that SpeedFan for Windows >> (www.almico.com/speedfan.php) seems to do something similar, otherwise >> it would show only 2 fans too. So maybe we could get that into >> lm_sensors, even if the BIOS is at fault here? >> > See above ... worst case this can result in defective hardware if > misconfigured. > > We could possibly introduce some kind of module parameter, but I don't > feel > too comfortable with that - wrong values may still blow up the hardware. > Any opinions, anyone ? > > Guenter > > _______________________________________________ > lm-sensors mailing list > lm-sensors@lm-sensors.org > http://lists.lm-sensors.org/mailman/listinfo/lm-sensors I wouldn't recommend a module parameter, leave the driver as it is. If there is a hack (using isaset) just use that. It's too easy for a user to set a module parameter hoping that it'll fix some random problem, and if the module kills the hardware, then the module is too blame. Asrock should fix the BIOS, that's the correct option. Regards Ian Dobson _______________________________________________ lm-sensors mailing list lm-sensors@lm-sensors.org http://lists.lm-sensors.org/mailman/listinfo/lm-sensors