Hi: I'm still working towards getting sensor information provided correctly for this GBT 2CEWH mobo. I found an ap, named moodss, that can record values and chart them once lm_sensors can be used. It runs so I have hopes it will work & I do not have to code something here from scratch. Yesterday '' i2cdetect -l '' would not work. Today, currently, it does. So here is the output that was requested : #> i2cdetect -l i2c-4 i2c NVIDIA I2C Device Algorithm unavailable i2c-3 i2c NVIDIA I2C Device Algorithm unavailable i2c-2 i2c NVIDIA I2C Device Algorithm unavailable i2c-1 smbus SMBus nForce2 adapter at 5040 Non-I2C SMBus adapter i2c-0 smbus SMBus nForce2 adapter at 5000 Non-I2C SMBus adapter It's 11:26:02 CDT (UTC-0500) on Wed Jun 14, week 24 in 2006. That IS the result that has been previously listed when using i2cdetect -l. Any ideas as to why it works sometimes but not others? (I suspect openSUSE config but do not know the smart place to start looking to confirm/fix that. ...) Also, oddly, I was unable to use i2cdump after first boot this AM. After a few minutes it started working just as it did earlier this AM before shutdown. A brief look at logs only provides a clue as to why. During boot, before login: ... Jun 14 10:45:10 name kernel: i2c_adapter i2c-2: SMBus Quick command not supported, can't probe for chips Jun 14 10:45:10 name kernel: i2c_adapter i2c-3: SMBus Quick command not supported, can't probe for chips Jun 14 10:45:10 name kernel: i2c_adapter i2c-4: SMBus Quick command not supported, can't probe for chips ... _BUT later, after login was done @ ~10:46AM, an additional entry: ... Jun 14 10:51:30 name kernel: i2c /dev entries driver ... I was able to get the i2cdump at 10:53 so suspect that the above entry is significant as to why I could not before. Is that a "SUSE bug/freature" or something to do with the way the devices are setup by the kernel? IOW, I want monitoring to begin ASAP so I can record & chart it, even w/o login if that is possible/sensible, so do I need some mod to SUSE, kernel or sensors?(or all of the above?) On the subject of temperatures: I ran a lot of dumps this AM(my yesterday) and have determined that on this mobo, for i2c-1, 2E: Register 26, CPU0 temperature! Register 27, ?? - this is always the 2nd coolest temp; ~23*c to 29*c. Register 28, CPU1 temperature! Register 29, ?? ~27*c to 32*c all the time Register 2a, Front temperature!( always the lowest temp ) Register 2b, Rear temperature! ( the BIOS & GBT have named it "Rear" temp ) I don't think I am going to be able to pin-down what registers 27 and 29 monitor until we can get lm_sensors working correctly so values can be recorded & charted. Both fluctuate, a little, as the CPUs temps go up and down but I cannot see any correlation, if there is any, with the small dataset that I have. ( I have hopes that those two are, respectively, the NV 2050 and NV 2200 chipsets ... ) Thanks for listening as well as any assistance/hints that can be provided. Philip's tip to run a dump using the suspected chip device address really helped nail where things are located! [I don't why I had not done that specific command before ... suppose my ''sensors'' knowledge base was just too low. (still is..., :) ] I have written a sensors.conf section for this device which has everything needed except some of PWM control parameters(and of course usability with lm_sensors...). I also have part of a chip doc similar to what one finds in the /usr/share/doc/packages/sensors/chips/ for a description to be added after compatibility is established. I am adding to it as I gather more *correct* info. I am investigating other ways/chips to get compatibility but have found none so far. If there is something specifically to look at, please inform me. Regards, Ric Johnson ================== --- Philip Pokorny <ppokorny at penguincomputing.com> wrote: > Ric Johnson wrote: > > >There is no URL as GBT sent the Datasheet to me. It is PDF. I can send it > but > >do not have a web server at this time so cannot post it. > > > > > It would *really* help if we could see the datasheet. How big is it? > > >The DS says the "default slave adddress is 0101110b" and GBT told me they > >"followed SMSC recommendations" ... > >There is a table showing > > "SMBUS Address[7:1] " as 0101 110b which is suppose is the same as 101110 > >binary or 2E hex. > > > > > Yes, that's the default address. > > >i2cdetect -l returns nothing. I am not sure why since lsmod has quite a > few > >"i2c*" items in it while sensors is running. > > > > > Perhaps that's a bug with i2cdetect... We can follow up on that problem > later. > > >#> i2cdetect 0 > >WARNING! This program can confuse your I2C bus, cause data loss and worse! > >I will probe file /dev/i2c-0. > >I will probe address range 0x03-0x77. > >Continue? [Y/n] > > 0 1 2 3 4 5 6 7 8 9 a b c d e f > >00: XX XX XX XX XX 08 XX XX XX XX XX XX XX > >10: XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX > >20: XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX > >30: XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX > >40: XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX > >50: XX XX UU UU XX XX UU UU XX XX XX XX XX XX XX XX > >60: XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX > >70: XX XX XX XX XX XX XX XX > > > > > > > No, that's probably just your memory DIMM's (at 52, 53, 56, 67) (do you > have 4 DIMM's installed?) Not sure what might be at address 0x08... > > >I have run i2cdump 0 0x2e but the results are only "XX". Perhaps that is > that > >wrong command parameter/address ... > > > > > That's because there is no device at address 0x2e on bus 0. Note the XX > in square 20 x e in the output above. > > >Since it *looks* like i2c-1 is where all the easter eggs are hidden, I did > a > >dump and got all kinds of stuff: > > > > > Perfect, the output from 'i2cdetect -l' would have listed what busses > you have and we would have gotten around to asking you to 'i2cdetect 1' > and so on, but you've figured that out for yourself. Excellent... > > >#> i2cdump 1 0x2e > >No size specified (using byte-data access) > >WARNING! This program can confuse your I2C bus, cause data loss and worse! > >I will probe file /dev/i2c-1, address 0x2e, mode byte > >Continue? [Y/n] > > 0 1 2 3 4 5 6 7 8 9 a b c d e f 0123456789abcdef > >00: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ................ > >10: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ................ > >20: c8 b0 bb 64 c0 b3 2d 17 31 1b 16 22 c4 0c 37 0e ???d??-?1??"??7? > >30: 52 0e 40 0d 1a 07 80 80 80 00 00 00 00 00 5c b1 R?@??????.....\? > >40: 05 00 00 00 00 00 00 00 3e 00 00 ff 00 ff 00 ff ?.......>....... > >50: 00 ff 00 ff 00 ff 81 7f 81 7f 81 7f 81 7f 81 7f ......?????????? > >60: 81 7f ff ff ff ff ff ff ff ff ff ff 00 00 00 ff ??.............. > >70: 00 ff 00 ff 00 ff 00 ff 00 ff fc 3e 7e 00 00 00 ..........?>~... > >80: e2 e2 e2 99 90 cc cc cc 00 00 80 80 80 00 00 00 ????????..???... > >90: 23 23 23 23 23 23 64 64 64 64 64 64 44 44 44 94 ######ddddddDDD? > >a0: 00 40 80 00 04 04 04 04 2c 2c 2c 00 0a 00 2d 2d .@?.????,,,.?.-- > >b0: 2d 2d 2d 2d 00 00 00 00 a0 00 05 25 12 ff ff 00 ----....?.?%?... > >c0: 00 00 00 00 7f 7f 7f 7f 7f 7f c0 00 00 00 00 00 ....???????..... > >d0: 00 ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ................ > >e0: 00 00 c0 8a 68 00 20 00 1c 00 00 00 00 00 00 00 ..??h. .?....... > >f0: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ................ > >It's 22:24:24 CDT (UTC-0500) on Tue Jun 13, week 24 in 2006. > > > >OF course, that might be an eeprom or something. > > > No, that's probably your chip. Go ahead and try i2cdump on one of your > eeproms 'i2cdump 0 0x52' and you'll see that the output is very different. > > >Any ideas? > > > > > > > Yes, but without the data sheet it's just speculation... > > I would say that you appear to have 6 voltage sensors (0x20-25), and 6 > temperature sensors (as you said previously, 0x26-2b). I'm going to > guess that you have five (5) 16-bit fan speed sensors (0x2c-35). > > The device ID signature registers are probably 0x3e (value 0x5c) and > 0x3f (value 0xb1) > > The min/max limits for the voltages probably start at 0x4a through > 0x55. Then the limits for the temperatures (0x56-0x61) and fan speeds > (0x62-0x6b). > > I would guess that the rest of the registers are perhaps fan speed > control or some other features on the chip. Hopefully they are > described in the datasheet. > > If that's correct, then we just need to know what the scaling values are > for the voltages and the fan speed equation. And your temperatures are: > > 2d = 45 degC > 17 = 23 degC > 31 = 49 degC > 1b = 27 degC > 16 = 22 degC > 22 = 34 degC > > If that's right, the temperatures appear to be rather low. Do you have this > machine in a cooled computer room? > > > Thanks, > :v) >
