Hallo Konstantin, On Wed, 24 Jun 2009 16:23:50 +0200, Konstantin Kletschke wrote: > Okay, I even moved the w83627ehf.ko and asus_atk0110.ko away and > rebooted. > > Am 2009-06-24 14:53 +0200 schrieb Jean Delvare: > > > isaset 0x295 0x296 0x4e 0 # set bank to 0 > > isadump 0x295 0x296 > > >isaset 0x295 0x296 0x4e 0 > >isadump 0x295 0x296 > 0 1 2 3 4 5 6 7 8 9 a b c d e f > 00: 04 ff 04 3d 00 00 46 20 01 9b 01 01 3c 3c 04 04 > 10: 04 99 30 00 00 01 01 3c 43 17 00 00 32 ff ff c5 > 20: 80 dc cf cd d3 ff c3 2c ff 99 ff da 00 06 71 10 > 30: f7 94 c8 45 50 b2 8a 13 c1 90 85 01 27 01 01 ff > 40: 03 00 00 ff ff 00 3f 35 2d 21 20 40 10 95 00 a3 > 50: ff ff 00 ff ff ff 00 80 c1 6f ff ff 19 45 04 05 > 60: 04 3d 50 00 01 01 3c ff 0a ff 01 ff ff ff ff 28 > 70: 46 46 46 46 46 37 ff ff ff ff ff 4b 22 00 2c ff > 80: 04 ff 04 3d 00 00 46 20 01 9b 01 01 3c 3c 04 04 > 90: 04 99 30 00 00 01 01 3c 43 17 00 00 32 ff ff c5 > a0: 80 dc cf cd d3 ff c3 2c ff 99 ff da 00 06 71 10 > b0: f7 94 c8 45 50 b2 8a 13 c1 90 85 01 27 01 01 ff > c0: 03 00 00 ff ff 00 3f 35 2d 21 20 40 10 95 00 a3 > d0: ff ff 00 ff ff ff 00 80 c1 6f ff ff 19 45 04 05 > e0: 04 3d 50 00 01 01 3c ff 0a ff 01 ff ff ff ff 28 > f0: 46 46 46 46 46 37 ff ff ff ff ff 4b 22 00 2c ff > > > isaset 0x295 0x296 0x01 0xaa > > isadump 0x295 0x296 > > >isaset 0x295 0x296 0x01 0xaa > Data mismatch, wrote 0xaa, read 0xff back. > >isadump 0x295 0x296 > 0 1 2 3 4 5 6 7 8 9 a b c d e f > 00: 04 ff 04 3d 00 00 46 20 01 9b 01 01 3c 3c 04 04 > 10: 04 99 30 00 00 01 01 3c 43 17 00 00 32 ff ff cf > 20: 80 dc cf ce d3 ff c3 2c ff 99 ff da 00 06 71 10 > 30: f7 94 c8 45 50 b2 8a 13 c1 90 85 01 27 01 01 ff > 40: 03 00 00 ff ff 00 3f 35 2d 21 20 40 10 95 00 a3 > 50: ff ff 00 ff ff ff 00 80 c1 6f ff ff 19 45 04 05 > 60: 04 3d 50 00 01 01 3c ff 0a ff 01 ff ff ff ff 28 > 70: 46 46 46 46 46 37 ff ff ff ff ff 4b 22 00 2c ff > 80: 04 ff 04 3d 00 00 46 20 01 9b 01 01 3c 3c 04 04 > 90: 04 99 30 00 00 01 01 3c 43 17 00 00 32 ff ff cf > a0: 80 dc cf ce d3 ff c3 2c ff 99 ff da 00 06 71 10 > b0: f7 94 c8 45 50 b2 8a 13 c1 90 85 01 27 01 01 ff > c0: 03 00 00 ff ff 00 3f 35 2d 21 20 40 10 95 00 a3 > d0: ff ff 00 ff ff ff 00 80 c1 6f ff ff 19 45 04 05 > e0: 04 3d 50 00 01 01 3c ff 0a ff 01 ff ff ff ff 28 > f0: 46 46 46 46 46 37 ff ff ff ff ff 4b 22 00 2c ff > > 0x01 seems to be readonly. > > > isaset 0x295 0x296 0x03 0xbb > > isadump 0x295 0x296 > > >isaset 0x295 0x296 0x03 0xbb > Data mismatch, wrote 0xbb, read 0x37 back. > >isadump 0x295 0x296 > 0 1 2 3 4 5 6 7 8 9 a b c d e f > 00: 04 bb 04 37 00 00 46 20 01 9b 01 01 3c 3c 04 04 > 10: 04 99 30 00 00 01 01 3c 43 17 00 00 32 ff ff d1 > 20: 80 dc cf ce d3 ff c3 2b ff 9d ff da 00 06 71 10 > 30: f7 94 c8 45 50 b2 8a 13 c1 90 85 01 27 01 01 ff > 40: 03 00 00 ff ff 00 3f 35 2d 21 20 40 10 95 00 a3 > 50: ff ff 00 ff ff ff 00 80 c1 6f ff ff 19 45 04 05 > 60: 04 37 50 00 01 01 3c ff 0a ff 01 ff ff ff ff 28 > 70: 46 46 46 46 46 37 ff ff ff ff ff 4b 22 00 2b ff > 80: 04 bb 04 37 00 00 46 20 01 9b 01 01 3c 3c 04 04 > 90: 04 99 30 00 00 01 01 3c 43 17 00 00 32 ff ff d1 > a0: 80 dc cf ce d3 ff c3 2b ff 9d ff da 00 06 71 10 > b0: f7 94 c8 45 50 b2 8a 13 c1 90 85 01 27 01 01 ff > c0: 03 00 00 ff ff 00 3f 35 2d 21 20 40 10 95 00 a3 > d0: ff ff 00 ff ff ff 00 80 c1 6f ff ff 19 45 04 05 > e0: 04 37 50 00 01 01 3c ff 0a ff 01 ff ff ff ff 28 > f0: 46 46 46 46 46 37 ff ff ff ff ff 4b 22 00 2b ff > > 0x03 writes to 0x01... > > > isaset 0x295 0x296 0x11 0xcc > > isadump 0x295 0x296 > > >isaset 0x295 0x296 0x11 0xcc > >isadump 0x295 0x296 > 0 1 2 3 4 5 6 7 8 9 a b c d e f > 00: 04 bb 04 37 00 00 46 20 01 9b 01 01 3c 3c 04 04 > 10: 04 cc 30 00 00 01 01 3c 43 17 00 00 32 ff ff d3 > 20: 80 dc cf ce d3 ff c3 2b ff 9d ff da 00 06 71 10 > 30: f7 94 c8 45 50 b2 8a 13 c1 90 85 01 27 01 01 ff > 40: 03 00 00 ff ff 00 3f 35 2d 21 20 40 10 95 00 a3 > 50: ff ff 00 ff ff ff 00 80 c1 6f ff ff 19 45 04 05 > 60: 04 37 50 00 01 01 3c ff 0a ff 01 ff ff ff ff 28 > 70: 46 46 46 46 46 37 ff ff ff ff ff 4b 22 00 2b ff > 80: 04 bb 04 37 00 00 46 20 01 9b 01 01 3c 3c 04 04 > 90: 04 cc 30 00 00 01 01 3c 43 17 00 00 32 ff ff d3 > a0: 80 dc cf ce d3 ff c3 2b ff 9d ff da 00 06 71 10 > b0: f7 94 c8 45 50 b2 8a 13 c1 90 85 01 27 01 01 ff > c0: 03 00 00 ff ff 00 3f 35 2d 21 20 40 10 95 00 a3 > d0: ff ff 00 ff ff ff 00 80 c1 6f ff ff 19 45 04 05 > e0: 04 37 50 00 01 01 3c ff 0a ff 01 ff ff ff ff 28 > f0: 46 46 46 46 46 37 ff ff ff ff ff 4b 22 00 2b ff > > 0x11 indeed writes to 0x11. > > This reflects somehow how echo into the /sys filesystem reacts. Indeed. So this is not a bug in the driver, but rather the hardware misbehaving. I'll ask out contact at Nuvoton about this. > > If this works fine then that would be a bug in the driver. If it > > doesn't, that would be a hardware bug. > > What I wonder, is there any success report pwm controlling cpu fan from > userspace with this W83667HG-A chip? Support for the W83667HG is still relatively new. Officially this chip is only supported since kernel 2.6.30 which was released just 2 weeks ago. So we didn't have much feedback yet. > Meanwhile I read many reports not achieving this. There are two other > fan header (3-PIN non PWM) and many people connect their fans there and > doing DC regulation, even on windows platform (the other pin on some > p5q-xxx boards are able to do DC regulating there, I don't know if mine > is able to do yet). Could as well be that the chip has a silicon bug... -- Jean Delvare
