> i2c-isa loads fine. w83781d fails at step 2. > > ==> /var/log/syslog <== > Sep 10 19:15:10 defiant kernel: i2c_adapter i2c-0: registered as > adapter > #0 > > <defiant> [2004-09-10 at 19:15:17] ~ -> modprobe w83781d > > ==> /var/log/syslog <== > Sep 10 19:15:25 defiant kernel: i2c-core: driver w83781d registered. > Sep 10 19:15:25 defiant kernel: i2c_adapter i2c-0: found normal isa > entry for adapter 9191, addr 0290 > Sep 10 19:15:25 defiant kernel: i2c_adapter i2c-0: Detection of > w83781d chip failed at step 2 OK, at least we now know where it breaks. The strange thing is that the very same detection trick is used in sensors-detect, where it seems to work fine for you. I admit I don't quite understand what's going on. Attached is a modified patch that will make the failure more verbose. Please revert the previous patch, apply this one, and test again. The patch will also ignore the step 2 error, so that we can see whether or not this is the only step failing. Alternatively, you can try to load the w83781d module with the "force=9191,0x290" parameter and see how it goes. This should skip all the detection steps. > > I would also appreciate the output of "isadump 0x295 0x296". > > Here's the output before I loaded any modules. > > 0 1 2 3 4 5 6 7 8 9 a b c d e f > 00: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff > 10: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff > 20: 7e 7d de b5 c7 d7 ca 1c ff 5c ff ff 00 ff 00 ed > 30: b0 80 24 00 c2 08 24 00 00 c0 a1 00 20 06 00 00 > 40: 01 d8 0f ff 7f 00 40 a0 2d 01 01 40 01 95 80 5c > 50: 00 7f 80 ff ff ff 00 80 11 ff ff ff ff ff ff ff > 60: 7e 7d de b5 c7 d7 ca 1c ff 5c ff ff 00 ff 00 ed > 70: b0 80 24 00 c2 08 24 00 00 c0 a1 00 20 06 00 00 > 80: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff > 90: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff > a0: 7e 7d de b5 c7 d7 ca 1c ff 5c ff ff 00 ff 00 ed > b0: b0 80 24 00 c2 08 24 00 00 c0 a1 00 20 06 00 00 > c0: 01 00 00 ff 7f 00 40 a0 2d 01 01 40 01 95 80 5c > d0: 01 75 80 ff ff ff 00 80 11 ff ff ff ff ff ff ff > e0: 7e 7d de b5 c7 d7 ca 1c ff 5c ff ff 00 ff 00 ed > f0: b0 80 24 00 c2 08 24 00 00 c0 a1 00 20 06 00 00 > > Here's the output after I loaded i2c-isa and w83781d. > > 0 1 2 3 4 5 6 7 8 9 a b c d e f > 00: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff > 10: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff > 20: 7d 7d df b5 c7 d7 ca 1c ff 5c ff ff 00 ff 00 ed > 30: b0 80 24 00 c2 08 24 00 00 c0 a1 00 20 06 00 00 > 40: 01 c8 0f ff 7f 00 40 a0 2d 01 01 40 01 95 80 5c > 50: 02 56 80 ff ff ff 00 80 11 ff ff ff ff ff ff ff > 60: 7d 7d df b5 c7 d7 ca 1c ff 5c ff ff 00 ff 00 ed > 70: b0 80 24 00 c2 08 24 00 00 c0 a1 00 20 06 00 00 > 80: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff > 90: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff > a0: 7d 7d df b5 c7 d7 ca 1c ff 5c ff ff 00 ff 00 ed > b0: b0 80 24 00 c2 08 24 00 00 c0 a1 00 20 06 00 00 > c0: 01 00 00 ff 7f 00 40 a0 2d 01 01 40 01 95 80 5c > d0: 03 45 80 ff ff ff 00 80 11 ff ff ff ff ff ff ff > e0: 7d 7d df b5 c7 d7 ca 1c ff 5c ff ff 00 ff 00 ed > f0: b0 80 24 00 c2 08 24 00 00 c0 a1 00 20 06 00 00 > > I tried another isadump later and diff'ed the results and some lines > were different. I guess it's constantly changing? This is the sensor > data? True. Registers containing measured values are likely to change, not by much but very often. Nothing to be afraid of. You can also notice that the chip registers are addressed over 7 bits, not 8. Thus 0x00-0x7F and 0x80-0xFF are "the same" (modulo the changes in measured values). -- Jean "Khali" Delvare http://khali.linux-fr.org/ -------------- next part -------------- An embedded and charset-unspecified text was scrubbed... Name: w83781d-more-debug-2.diff Url: http://lists.lm-sensors.org/pipermail/lm-sensors/attachments/20040911/88bb69f1/attachment.pl
