http://bugzilla.novell.com/show_bug.cgi?id=529018 Marcus Meissner changed: What |Removed |Added ---------------------------------------------------------------------------- AssignedTo|bnc-team-screening@forge.pr |trenn@novell.com |ovo.novell.com | Summary|/usr/bin/sensors hängs on |/usr/bin/sensors hangs on |ASUS P5W DH Delux |ASUS P5W DH Delux -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

http://bugzilla.novell.com/show_bug.cgi?id=529018 User Joachim.Reichelt@helmholtz-hzi.de added comment http://bugzilla.novell.com/show_bug.cgi?id=529018#c2 Joachim Reichelt changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEEDINFO |NEW Info Provider|jdelvare@novell.com | --- Comment #2 from Joachim Reichelt 2009-08-08 11:26:06 MDT --- Created an attachment (id=311460) --> (http://bugzilla.novell.com/attachment.cgi?id=311460) tar cvf attachments.tar dmesg-08-08.09 acpidump.txt boot.* As I'm not sure whether this is a pure kernel bug or related to the upgrade from older openSUSE versions (11.0 -> 11.1 (alpha+beta) -> 11.2 (M1, M2, M3, M4, M5) I'll stay here at now. acpidump seems not to be installed at 11.2 (by default), so I used an older one. I got 2 errors: # /home/old/usr/sbin/acpidump > acpidump.txt Wrong checksum for OEMB Wrong checksum for OEMB! I attached the output from the /var/log/boot* file too Add: The system loaded the "old" driver for sensors too: /etc/sysconfig/lm_sensors MODULE_0=w83627ehf HWMON_MODULES="w83627ehf" -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

http://bugzilla.novell.com/show_bug.cgi?id=529018 User Joachim.Reichelt@helmholtz-hzi.de added comment http://bugzilla.novell.com/show_bug.cgi?id=529018#c4 --- Comment #4 from Joachim Reichelt 2009-08-09 13:38:12 MDT --- I did some more tests: in terminal1 I start sensors. It hangs before readin the first fan in terminal2 I enter: # killall -STOP sensors;killall -CONT sensors in terminal1 I got the whole output, BUT: The value for FAN1 is garbage! All other values are the same for atk0110 and w83627ehf. blacklisting would help me, but not fix the "bug". So can I help here? debug?... Is there a way to use atk0110 for fancontrol? It seems to stop my cpu fan. -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

http://bugzilla.novell.com/show_bug.cgi?id=529018 User Joachim.Reichelt@helmholtz-hzi.de added comment http://bugzilla.novell.com/show_bug.cgi?id=529018#c6 --- Comment #6 from Joachim Reichelt 2009-08-10 06:44:37 MDT --- The w83627ehf reports "good" fan speeds. The problem is only in FAN1. Each time I got different values. If the speed whould be as high as reported by the atk0110, the fan whould fly away :-; values are > 10.000rpm In reality the value is ~1000rpm, same value ist seen in the bios. PS.: Why not break the while-loop in two independen ones. This will stop oscillations raised by noise in fan_cnt, as e.g. by noninitialized variables. I know, there are many funny things on this asus board. -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

http://bugzilla.novell.com/show_bug.cgi?id=529018 User Joachim.Reichelt@helmholtz-hzi.de added comment http://bugzilla.novell.com/show_bug.cgi?id=529018#c8 Joachim Reichelt changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEEDINFO |ASSIGNED Info Provider|Joachim.Reichelt@helmholtz- | |hzi.de | --- Comment #8 from Joachim Reichelt 2009-08-10 15:10:59 MDT --- O.K. I removed all w83627ehf related stuff from /etc/... and now sensors gives good results. But now my CPU goes straight to 55 deg without any load. And adding load it goes up. But the fan stays at ~1100rmp. So how to setup fancontrol? (But that shoud be another bugreport) So how to make shure, not to load the ACPI module and the nativ HW driver? Couse that was the reason for the failure. Now I have: # lsmod Module Size Used by nls_utf8 1888 0 xt_tcpudp 3168 1 xt_pkttype 1696 3 ipt_LOG 6308 8 xt_limit 2660 8 snd_pcm_oss 51456 0 snd_mixer_oss 18464 1 snd_pcm_oss snd_seq 64656 0 af_packet 23648 0 microcode 17548 0 xt_NOTRACK 1664 2 ipt_REJECT 3200 3 xt_state 2240 4 iptable_raw 2880 1 iptable_filter 3648 1 nf_conntrack_netbios_ns 2464 0 nf_conntrack_ipv4 11720 4 nf_conntrack 84476 4 xt_NOTRACK,xt_state,nf_conntrack_netbios_ns,nf_conntrack_ipv4 nf_defrag_ipv4 2112 1 nf_conntrack_ipv4 ip_tables 13520 2 iptable_raw,iptable_filter ip6_tables 14960 0 x_tables 19140 9 xt_tcpudp,xt_pkttype,ipt_LOG,xt_limit,xt_NOTRACK,ipt_REJECT,xt_state,ip_tables,ip6_tables nls_iso8859_1 4032 2 nls_cp437 5696 2 vfat 12576 2 fat 59072 1 vfat fuse 75356 11 loop 17164 0 dm_mod 84036 0 arc4 1984 2 ecb 3296 2 snd_usb_audio 95552 2 snd_usb_lib 21504 1 snd_usb_audio snd_rawmidi 26400 1 snd_usb_lib snd_seq_device 8588 2 snd_seq,snd_rawmidi snd_hda_codec_realtek 231172 1 snd_hda_intel 31136 4 snd_hda_codec 94496 2 snd_hda_codec_realtek,snd_hda_intel snd_hwdep 8612 2 snd_usb_audio,snd_hda_codec snd_pcm 97444 4 snd_pcm_oss,snd_usb_audio,snd_hda_intel,snd_hda_codec rtl8187 54468 0 snd_timer 26760 2 snd_seq,snd_pcm snd 74820 25 snd_pcm_oss,snd_mixer_oss,snd_seq,snd_usb_audio,snd_usb_lib,snd_rawmidi,snd_seq_device,snd_hda_codec_realtek,snd_hda_intel,snd_hda_codec,snd_hwdep,snd_pcm,snd_timer mac80211 255720 1 rtl8187 led_class 4836 1 rtl8187 eeprom_93cx6 2176 1 rtl8187 ohci1394 32784 0 iTCO_wdt 12484 0 mag_quickcam_messenger 104808 0 iTCO_vendor_support 3908 1 iTCO_wdt soundcore 8480 1 snd cfg80211 103488 2 rtl8187,mac80211 rfkill 22580 1 cfg80211 gspca_stv06xx 29000 0 usbhid 51520 0 hid 46816 1 usbhid usb_storage 57056 1 gspca_main 26304 1 gspca_stv06xx videodev 39104 2 mag_quickcam_messenger,gspca_main v4l1_compat 15876 1 videodev asus_atk0110 13760 0 ieee1394 101956 1 ohci1394 snd_page_alloc 10248 2 snd_hda_intel,snd_pcm pcspkr 2816 0 i2c_i801 12820 0 sr_mod 17348 0 cdrom 40480 1 sr_mod sg 32500 0 floppy 61252 0 button 6576 0 nvidia 8952932 26 sky2 52452 0 i2c_core 32212 3 videodev,i2c_i801,nvidia i82975x_edac 4740 0 edac_core 49236 1 i82975x_edac rtc_cmos 14860 0 rtc_core 22684 1 rtc_cmos rtc_lib 3200 1 rtc_core uhci_hcd 29104 0 ehci_hcd 56272 0 pata_acpi 4672 0 edd 10472 0 fan 5060 0 processor 49584 0 ide_pci_generic 4068 0 piix 6792 0 ide_core 122412 2 ide_pci_generic,piix sym53c8xx 83928 0 scsi_transport_spi 27456 1 sym53c8xx ata_generic 4900 0 pata_jmicron 3232 3 ata_piix 25188 5 ahci 37520 0 libata 198412 5 pata_acpi,ata_generic,pata_jmicron,ata_piix,ahci thermal 21052 0 thermal_sys 18056 3 fan,processor,thermal hwmon 3708 2 asus_atk0110,thermal_sys Now I can: # modprobe w83627ehf # sensors with success and "good" results! And here is the reason for the lock: Now I started fancontrol (using w83627ehf!!!) Now I have to wait till fancontrol gives me access to the fan etc. -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

http://bugzilla.novell.com/show_bug.cgi?id=529018 User kronos.it@gmail.com added comment http://bugzilla.novell.com/show_bug.cgi?id=529018#c10 --- Comment #10 from Luca Tettamanti 2009-08-11 04:17:53 MDT --- Are you overriding acpi_enforce_resources? You shouldn't be able to load w83627ehf driver like that... The monitoring chip uses index+data register scheme (plus banks...), having two drivers loaded at the same time is guaranteed to case troubles. Please attach the content of /proc/ioports -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

http://bugzilla.novell.com/show_bug.cgi?id=529018 User trenn@novell.com added comment http://bugzilla.novell.com/show_bug.cgi?id=529018#c12 --- Comment #12 from Thomas Renninger 2009-08-11 08:09:51 MDT ---

Are you overriding acpi_enforce_resources? The acpi_enforce_resources mechanism does not work anymore. Hopefully Len will revive the functionality for 2.6.31 soon and post a patches for stable@kernel.org (got removed for 2.6.30 already AFAIK and sooner kernels need the memleak fix from lin ming). Also see mail thread: http://markmail.org/message/d4at27mzaivtpbz3#query:+page:1+mid:46xdznqq53zhg...

Subject: Re: [PATCH] ACPICA: Revert "ACPICA: Remove obsolete acpi_os_validate_address interface" on the acpi mailing list. -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

http://bugzilla.novell.com/show_bug.cgi?id=529018 User kronos.it@gmail.com added comment http://bugzilla.novell.com/show_bug.cgi?id=529018#c14 --- Comment #14 from Luca Tettamanti 2009-08-11 08:21:58 MDT --- (In reply to comment #13)

Joachim is using 2.6.31-rc5 though, which contains the patch that you mentioned.

Ops, I was looking at my own log... he's running 2.6.31-rc4, so the conflict is not detected. -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

http://bugzilla.novell.com/show_bug.cgi?id=529018 User trenn@novell.com added comment http://bugzilla.novell.com/show_bug.cgi?id=529018#c16 --- Comment #16 from Thomas Renninger 2009-09-07 03:08:46 MDT ---

Thomas, please do not suggest blacklisting the asus_atk0110 driver in the future Did I? If, this was more for testing or as a workaround, not for a final solution.

-- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

http://bugzilla.novell.com/show_bug.cgi?id=529018 User trenn@novell.com added comment http://bugzilla.novell.com/show_bug.cgi?id=529018#c18 --- Comment #18 from Thomas Renninger 2009-09-07 06:09:06 MDT ---

In this case there's might be a bug in the ACPI code, the driver hangs looping there. Hmm, but it works if only the atk0110 driver is loaded? I've seen the loop, but it (endless loop) may only happen if another hwmon driver is also accessing the HW and the driver gets unexpected values? Not sure, the loop is very unfortunate and bad code style (if the sensor values are not sanity checked before).

-- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

http://bugzilla.novell.com/show_bug.cgi?id=529018 User Joachim.Reichelt@helmholtz-hzi.de added comment http://bugzilla.novell.com/show_bug.cgi?id=529018#c20 --- Comment #20 from Joachim Reichelt 2009-09-07 14:36:32 MDT --- By the way: joachim:~ # rmmod asus_atk0110 joachim:~ # modprobe w83627ehf FATAL: Error inserting w83627ehf (/lib/modules/2.6.31-rc8-29-default/kernel/drivers/hwmon/w83627ehf.ko): Device or resource busy joachim:~ # So I cannot load w83627ehf.ko at all(?) Is there any possibility to change the fan speed? -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

http://bugzilla.novell.com/show_bug.cgi?id=529018 User kronos.it@gmail.com added comment http://bugzilla.novell.com/show_bug.cgi?id=529018#c22 --- Comment #22 from Luca Tettamanti 2009-09-08 05:00:58 MDT --- (In reply to comment #21)

As to changing the fan speeds, that would require pwm attributes to be created by the asus_atk0110 driver. For now, this driver has only been reading from the chip. It's up to Luca to decide whether it is possible, and safe, to add more features to the driver.

It's not possible to control the PWM directly; it should be possible to change the "profile" (as defined in the BIOS), but I haven't been able to make it work yet. -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

http://bugzilla.novell.com/show_bug.cgi?id=529018 http://bugzilla.novell.com/show_bug.cgi?id=529018#c24 Jean Delvare changed: What |Removed |Added ---------------------------------------------------------------------------- Status|ASSIGNED |NEEDINFO Info Provider| |Joachim.Reichelt@helmholtz- | |hzi.de --- Comment #24 from Jean Delvare 2010-02-15 17:53:13 UTC --- The fan speed garbage could be caused by the fan speed control. Fan speed control is achieved by not providing the fan its full continuous power supply and this can cause noise in the speed signal especially at very low speeds. Some monitoring chips have special circuitry to compensate for this, some do not. Try disabling the automatic fan speed control and see if it helps. It could also be that your fan is simply defective and never properly reporting its speed. In that case, the problem will survive disabling the automatic fan speed control. In both cases, I don't think this is a problem in the asus_atk0110 driver, this is a hardware problem and I presume you see the same problems in the BIOS. -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

http://bugzilla.novell.com/show_bug.cgi?id=529018 http://bugzilla.novell.com/show_bug.cgi?id=529018#c26 --- Comment #26 from Joachim Reichelt 2010-02-28 16:48:02 UTC --- I just did a test: as /proc/acpi/fan is empty here, I did a # rmmod fan and this seems to fix the trouble. now I always get good results without waiting: CPU FAN Speed: 602 RPM (min = 0 RPM) CHASSIS1 FAN Speed:1506 RPM (min = 800 RPM) CHASSIS2 FAN Speed:1442 RPM (min = 800 RPM) POWER1 FAN Speed: 0 RPM (min = 800 RPM) So it seems to me a problem between these two modules on this board. # uname -a Linux Joachim-PC 2.6.33-rc7-3-desktop #1 SMP PREEMPT 2010-02-12 17:37:20 +0100 x86_64 x86_64 x86_64 GNU/Linux -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

http://bugzilla.novell.com/show_bug.cgi?id=529018 http://bugzilla.novell.com/show_bug.cgi?id=529018#c28 Joachim Reichelt changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEEDINFO |ASSIGNED Info Provider|Joachim.Reichelt@helmholtz- | |hzi.de | --- Comment #28 from Joachim Reichelt 2010-03-01 22:35:24 UTC --- I just did an upgrade to M3. During this I checked /etc/sysconfig/kernel and found: INITRD_MODULES="thermal pata_jmicron ata_generic sym53c8xx processor fan" and so I removed the "fan" now I get good results. Maybe this "fan" was leftover during upgrades. But the first install was 11.2 goldmaster. -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

http://bugzilla.novell.com/show_bug.cgi?id=529018 http://bugzilla.novell.com/show_bug.cgi?id=529018#c30 --- Comment #30 from Joachim Reichelt 2010-03-02 19:54:00 UTC --- It seem to be bios: acpidump --addr 0xFFFF0000 --length 0xFFFF| od -c | less 0000000 224 324 \0 \0 \0 \0 \0 \0 \0 \0 \0 \0 \0 \0 \0 \0 0000020 \0 \0 \0 \0 \0 \0 \0 \0 \0 \0 \0 \0 \0 \0 \0 \0 * 0000100 P 5 W D H R O M \0 \0 \0 \0 \0 0000120 \0 \0 \0 \0 \0 035 G \0 \0 022 P \0 Q I \0 \0 0000140 001 \0 \0 \0 001 \0 \0 \0 \0 \0 \0 \0 \0 \0 \0 \0 0000160 351 \r 002 351 344 f 351 177 321 351 G 002 351 ; 035 351 0000200 277 351 345 f 351 210 \t 351 w 037 351 J 351 3 0000220 215 351 220 b 351 003 } 351 ^ 216 351 347 321 351 016 k 0000240 351 332 a 351 360 203 351 251 i 351 304 ~ 351 246 020 351 0000260 ( y 351 e u 351 \0 k 351 347 001 351 211 017 351 = I cannot find any PNP0C0B: find /sys/devices/ -print0 | xargs -0 ls -ld | grep -i 0C0B is empty. So it seems to be all o.k. -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.