Am Dienstag, den 05.09.2006, 18:20 +0200 schrieb Jean Delvare: > David, Matthias, > > > Nice to get some feedback on w83627ehf. The datasheet documents only 4 > > RPM sensors, but 5 pwm outputs. That's if your motherboard uses the > > standard w83627ehf without any additional electronics. > > No, that's the other way around. 5 RPM sensors, 4 PWM output. > > > Can you disconnect one fan at a time while in the BIOS readings and > > verify that each is independent of all other fans? I wonder if the > > BIOS is actually getting 5 independent RPM readings. > > The number of readings in the BIOS matches the number of fan monitoring > inputs of the chip - looks clear enough to me. Indeed, here's what the BIOS looks like: http://hentges.net/tmp/photos/p5w%20dh%20deluxe/hwmon/ > > Also, is there some way (a different OS, or in the BIOS) that you can > > actually control each of the fans independently? I'm looking for more > > info on your motherboard, but there is the possibility that the two > > CHA fans and the two PWR fans are actually hard-wired together. > > This is my conclusion as well. The PWM outputs wiring is decided by the > motherboard maker, and they may use the same PWM output for several > fans. Given that there are only 4 PWM outputs they actually had little > choice for at least two fans. It makes sense to me to control related > fans with a single command too. Looks like you are right, SpeedFan on the "other" OS behaves exactly like lm-sensors. both CHA and both PWR fans can only be controlled together :\ > > Can you provide a listing of the sysfs files exported by the driver on > > your system? The listing is attached. > The w83627ehf pins for fans 4 and 5 can be programmed as > > general purpose I/O, and the driver detects whether you have 3, 4, or > > 5 fan pins configured. (The setting is read, but not changed.) So...do > > you have pwm4 and pwm5 in your sysfs? > > Matthias, can you please provide a dump of your chip (isadump 0x295 > 0x296) at init time, before loading the w83627ehf driver? The dump is attached. It was taken before loading any modules (init=/bin/sh). > I'd try forcing the driver to use all 5 fan inputs. Just run "isaset > 0x295 0x296 0x47 0x05 0x05" before loading the w83627ehf driver, and > see what happens. It's supposed to be the default, but... I tried it, no success :\ I still only get 4 fans. > Maybe our heuristics to detect the use of the pins are not correct, > although they are simple enough and look OK to me. Or maybe the Asus > BIOS isn't configuring the chip properly (that'd be odd as they need > proper setup for displaying the values themselves.) > -- Matthias 'CoreDump' Hentges Webmaster of hentges.net and OpenZaurus developer. You can reach me in #openzaurus on Freenode. My OS: Debian SID. Geek by Nature, Linux by Choice -------------- next part -------------- 0 1 2 3 4 5 6 7 8 9 a b c d e f 00: 04 ff 04 ff 00 2b 3f 33 c0 9b c0 9b 3c 3c 01 01 10: 01 5e 70 00 00 01 01 3c 43 17 00 00 57 00 00 d3 20: 8f ef d3 d3 d5 cb cc 27 ad 89 a9 da 00 f7 fe ff 30: 7f 57 ed 1f fd f9 fe bf df 3c 37 0f 78 0f 0a 55 40: 03 fa af de ff ff 00 f4 2d ff 00 c4 98 95 06 a3 50: e4 1f 59 96 3a 39 3d 3d ff ff ff ff ff ff ff ff 60: 04 ff 03 21 f0 f0 3c ff 01 ff 01 ff ff ff ff ff 70: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff 80: 04 ff 04 ff 00 2b 3f 33 c0 9b c0 9b 3c 3c 01 01 90: 01 5e 70 00 00 01 01 3c 43 17 00 00 57 00 00 d3 a0: 8f ef d3 d3 d5 cb cc 27 ad 89 a9 da 00 f7 fe ff b0: 7f 57 ed 1f fd f9 fe bf df 3c 37 0f 78 0f 0a 55 c0: 03 00 00 de ff ff 00 f4 2d ff 00 c4 98 95 06 a3 d0: e4 1f 59 96 3a 39 3d 3d ff ff ff ff ff ff ff ff e0: 04 ff 03 21 f0 f0 3c ff 01 ff 01 ff ff ff ff ff f0: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff -------------- next part -------------- bus driver fan1_alarm fan1_div fan1_input fan1_min fan2_alarm fan2_div fan2_input fan2_min fan3_alarm fan3_div fan3_input fan3_min fan4_alarm fan4_div fan4_input fan4_min hwmon:hwmon0 in0_alarm in0_input in0_max in0_min in1_alarm in1_input in1_max in1_min in2_alarm in2_input in2_max in2_min in3_alarm in3_input in3_max in3_min in4_alarm in4_input in4_max in4_min in5_alarm in5_input in5_max in5_min in6_alarm in6_input in6_max in6_min in7_alarm in7_input in7_max in7_min in8_alarm in8_input in8_max in8_min in9_alarm in9_input in9_max in9_min name power pwm1 pwm1_enable pwm1_min_output pwm1_mode pwm1_stop_time pwm1_target pwm1_tolerance pwm2 pwm2_enable pwm2_min_output pwm2_mode pwm2_stop_time pwm2_target pwm2_tolerance pwm3 pwm3_enable pwm3_min_output pwm3_mode pwm3_stop_time pwm3_target pwm3_tolerance pwm4 pwm4_enable pwm4_min_output pwm4_mode pwm4_stop_time pwm4_target pwm4_tolerance subsystem temp1_alarm temp1_input temp1_max temp1_max_hyst temp2_alarm temp2_input temp2_max temp2_max_hyst temp3_alarm temp3_input temp3_max temp3_max_hyst uevent -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 189 bytes Desc: Dies ist ein digital signierter Nachrichtenteil Url : http://lists.lm-sensors.org/pipermail/lm-sensors/attachments/20060906/33bd7775/attachment.bin
