Fan Control fails after resume from pm-suspend

nblock · 2011-02-03 08:59:23

Hi,
i recently upgraded to 2.6.37 (on a hp nx6325 laptop) and noticed the following behaviour after a resume from pm-suspend:
The fan control stops working until cpu reaches a temperature of 60°. As soon as the temperatur passes this threshold the fan starts rotating at max. speed. Below 60°, the fan is turned off immediately.

pm-suspend.log:

disabled, not active

/usr/lib/pm-utils/sleep.d/01laptop-mode resume suspend: success.
Running hook /usr/lib/pm-utils/sleep.d/01grub resume suspend:

/usr/lib/pm-utils/sleep.d/01grub resume suspend: success.
Running hook /usr/lib/pm-utils/sleep.d/00powersave resume suspend:

/usr/lib/pm-utils/sleep.d/00powersave resume suspend: success.
Running hook /usr/lib/pm-utils/sleep.d/00logging resume suspend:

/usr/lib/pm-utils/sleep.d/00logging resume suspend: success.
Don Feb  3 09:16:21 CET 2011: Finished.

Any ideas where to look for the problem?

lagagnon · 2011-02-03 15:14:31

I had the same problem with a Toshiba Satellite L300 that was only solved by upgrading the BIOS, but that might not necessarily be your problem. However, usually it is a faulty DSDT that is sometimes alleviated by a BIOS upgrade - here is an old thread:
http://forum.notebookreview.com/linux-c … x6325.html

nblock · 2011-02-03 15:35:00

thx for your reply, lagagnon.
I took a quick look at the hp driver site and noticed, that the latest available bios-version is already installed.

na12 · 2011-02-06 08:23:27

I have same problem with HP Compaq 6715b laptop,but it reach 70° C before it turn on fan.

Edit:
It seems this bug is not solved yet.

http://us.generation-nt.com/answer/regr … tml?page=2

https://bugzilla.kernel.org/show_bug.cgi?id=19452

Last edited by na12 (2011-02-06 08:51:50)

daroczig · 2011-02-16 00:55:33

I have the exact same problem on a HP 6715b laptop. Do anyone have a solution to this annoying bug?

Edit: this bug only occurred after upgrading to the new kernel (2.6.37), before I had no problems like that. I am not sure if the bug reports linked above are really connected to my problem. Any thoughts?

Last edited by daroczig (2011-02-17 22:17:44)

na12 · 2011-02-20 19:43:46

daroczig wrote:

I have the exact same problem on a HP 6715b laptop. Do anyone have a solution to this annoying bug?
Edit: this bug only occurred after upgrading to the new kernel (2.6.37), before I had no problems like that. I am not sure if the bug reports linked above are really connected to my problem. Any thoughts?

New update to kernel 2.6.37-6 solved problem for me.

daroczig · 2011-02-20 21:43:42

Unfortunately not for me
I'll try today with 2.6.37.1-1, I hope with better luck.

Arpione · 2011-02-20 22:14:34

I had the same problem with the fan but 2.6.37.1-1 "solved" it for me. Now when I run pm-suspend the screen goes blank as it normally should, then it returns to X. So now I can't even suspend my machine (HP nx6325).

daroczig · 2011-02-21 14:53:22

I have the same problem with 2.6.37.1-1: suspend do not work.
The sceen goes blank then returns to X. The log shows the followings:

Feb 21 15:50:33 localhost kernel: Suspending console(s) (use no_console_suspend to debug)
Feb 21 15:50:33 localhost kernel: sd 2:0:0:0: [sda] Synchronizing SCSI cache
Feb 21 15:50:33 localhost kernel: sd 2:0:0:0: [sda] Stopping disk
Feb 21 15:50:33 localhost kernel: tpm_tis 00:03: Operation Timed out
Feb 21 15:50:33 localhost kernel: legacy_suspend(): pnp_bus_suspend+0x0/0xa0 returns -62
Feb 21 15:50:33 localhost kernel: PM: Device 00:03 failed to suspend: error -62
Feb 21 15:50:33 localhost kernel: PM: Some devices failed to suspend
Feb 21 15:50:33 localhost kernel: sd 2:0:0:0: [sda] Starting disk
Feb 21 15:50:33 localhost kernel: b43-phy0: Loading firmware version 478.104 (2008-07-01 00:50:23)
Feb 21 15:50:33 localhost kernel: ------------[ cut here ]------------
Feb 21 15:50:33 localhost kernel: WARNING: at drivers/base/power/main.c:106 device_pm_add+0xb6/0xc0()
Feb 21 15:50:33 localhost kernel: Hardware name: HP Compaq 6715b (GB835EA#AKC)
Feb 21 15:50:33 localhost kernel: Device: misc
Feb 21 15:50:33 localhost kernel: Parentless device registered during a PM transaction
Feb 21 15:50:33 localhost kernel: Modules linked in: fuse ipv6 vboxnetflt vboxdrv snd_seq_dummy snd_seq_oss snd_seq_midi_event loop snd_seq arc4 ecb radeon b43 ttm drm_kms_helper mac80211 snd_pcm_oss drm snd_mixer_oss joydev snd_hda_codec_analog ssb tpm_infineon i2c_algo_bit cfg80211 hp_wmi mmc_core sparse_keymap tg3 ppdev rfkill snd_hda_intel parport_pc libphy pcmcia snd_hda_codec snd_usb_audio snd_pcm ohci_hcd yenta_socket pcmcia_rsrc pcmcia_core firewire_ohci snd_hwdep tpm_tis snd_usbmidi_lib hp_accel lis3lv02d i2c_piix4 video tpm snd_timer firewire_core evdev ehci_hcd snd_rawmidi crc_itu_t serio_raw snd_page_alloc tpm_bios output wmi input_polldev thermal battery lp sg ac container usbcore i2c_core button fan snd_seq_device shpchp parport edac_core edac_mce_amd snd pci_hotplug psmouse soundcore k8temp cpufreq_ondemand powernow_k8 freq_table processor mperf ext4 mbcache jbd2 crc16 sd_mod sr_mod cdrom ahci libahci pata_atiixp pata_acpi libata scsi_mod
Feb 21 15:50:33 localhost kernel: Pid: 12986, comm: pm-suspend Not tainted 2.6.37-ARCH #1
Feb 21 15:50:33 localhost kernel: Call Trace:
Feb 21 15:50:33 localhost kernel: [<ffffffff81056a1a>] warn_slowpath_common+0x7a/0xb0
Feb 21 15:50:33 localhost kernel: [<ffffffff81056af1>] warn_slowpath_fmt+0x41/0x50
Feb 21 15:50:33 localhost kernel: [<ffffffff812b8596>] device_pm_add+0xb6/0xc0
Feb 21 15:50:33 localhost kernel: [<ffffffff812aea8e>] device_add+0x4ae/0x5c0
Feb 21 15:50:33 localhost kernel: [<ffffffff812aebb9>] device_register+0x19/0x20
Feb 21 15:50:33 localhost kernel: [<ffffffff812aed53>] device_create_vargs+0x193/0x1b0
Feb 21 15:50:33 localhost kernel: [<ffffffff812aed9c>] device_create+0x2c/0x30
Feb 21 15:50:33 localhost kernel: [<ffffffff813a5fc1>] ? mutex_lock+0x11/0x30
Feb 21 15:50:33 localhost kernel: [<ffffffff8129a3fa>] misc_register+0x8a/0x140
Feb 21 15:50:33 localhost kernel: [<ffffffff8129d36f>] hwrng_register+0xdf/0x170
Feb 21 15:50:33 localhost kernel: [<ffffffffa0426ab0>] b43_wireless_core_init+0xdb0/0x11a0 [b43]
Feb 21 15:50:33 localhost kernel: [<ffffffffa02d89e0>] ? wiphy_resume+0x0/0x90 [cfg80211]
Feb 21 15:50:33 localhost kernel: [<ffffffffa0427068>] b43_op_start+0x1c8/0x1e0 [b43]
Feb 21 15:50:33 localhost kernel: [<ffffffffa02d89e0>] ? wiphy_resume+0x0/0x90 [cfg80211]
Feb 21 15:50:33 localhost kernel: [<ffffffffa0477a15>] ieee80211_reconfig+0x335/0x400 [mac80211]
Feb 21 15:50:33 localhost kernel: [<ffffffffa02d89e0>] ? wiphy_resume+0x0/0x90 [cfg80211]
Feb 21 15:50:33 localhost kernel: [<ffffffffa046d8b8>] ieee80211_resume+0x28/0x70 [mac80211]
Feb 21 15:50:33 localhost kernel: [<ffffffffa02d8a55>] wiphy_resume+0x75/0x90 [cfg80211]
Feb 21 15:50:33 localhost kernel: [<ffffffff812b7ebe>] legacy_resume+0x3e/0x80
Feb 21 15:50:33 localhost kernel: [<ffffffff812b8025>] device_resume+0x125/0x130
Feb 21 15:50:33 localhost kernel: [<ffffffff812b8192>] dpm_resume_end+0x102/0x3a0
Feb 21 15:50:33 localhost kernel: [<ffffffff81096350>] suspend_devices_and_enter+0x90/0x1f0
Feb 21 15:50:33 localhost kernel: [<ffffffff810965d0>] enter_state+0x120/0x150
Feb 21 15:50:33 localhost kernel: [<ffffffff81095b16>] state_store+0xc6/0x100
Feb 21 15:50:33 localhost kernel: [<ffffffff811f3c27>] kobj_attr_store+0x17/0x20
Feb 21 15:50:33 localhost kernel: [<ffffffff8119f37f>] sysfs_write_file+0xcf/0x150
Feb 21 15:50:33 localhost kernel: [<ffffffff81135ac6>] vfs_write+0xc6/0x190
Feb 21 15:50:33 localhost kernel: [<ffffffff81135ddc>] sys_write+0x4c/0x80
Feb 21 15:50:33 localhost kernel: [<ffffffff8100bed2>] system_call_fastpath+0x16/0x1b
Feb 21 15:50:33 localhost kernel: ---[ end trace 19e4cbd37086936d ]---
Feb 21 15:50:33 localhost kernel: PM: resume of devices complete after 1300.325 msecs
Feb 21 15:50:33 localhost kernel: PM: Finishing wakeup.
Feb 21 15:50:33 localhost kernel: Restarting tasks ... done.
Feb 21 15:50:33 localhost kernel: video LNXVIDEO:00: Restoring backlight state

And the solution on linux.kernel: http://groups.google.com/group/linux.ke … 616cfdb4a9

Revert or wait!

Or rmmod tpm_tis.

Last edited by daroczig (2011-02-21 21:55:10)

na12 · 2011-02-23 14:12:34

daroczig wrote:

I have the same problem with 2.6.37.1-1: suspend do not work.

yep,same problem here with 2.6.37.1-1

daroczig · 2011-03-02 23:46:04

Got it working with kernel 2.6.37.2-1.
Well, only for suspend: after resuming the fan control still fails

daroczig · 2011-03-13 14:14:36

Same issue with kernel 2.6.37.3-1.

daroczig · 2011-03-17 19:06:18

And also with kernel 2.6.37.4-1

holstein · 2011-03-17 20:50:29

Hello, this is a long time (at least 6 months) that I have similar problem. Computer works correctly. But when I suspend and resume the fan don't start causing the CPU to overheat if he has to work hardly. This causes the computer to shut down a few minutes after boot up.

I use kernel 2.6.37.4-1 on "Linux 2.6.37-ARCH #1 SMP PREEMPT Tue Mar 15 09:21:17 CET 2011 x86_64 Intel(R) Core(TM)2 Quad CPU Q9000 @ 2.00GHz GenuineIntel GNU/Linux" from Gateway.
Video card : Nvidia GeForce GTX 260M

Edit : On my computer the temperature where fan starts is around 70 °C. It should be at full speed around 90°C. Once the temperature rised to 100°C whitout the fan starting. Comparing to the others presented I think this is very hot. On the Internet I saw that max T for my processor is around 105°C. But I think that the heat from the CPU travels to harddrive wich makes coming them to 50 to 60 °C which is high.

I use labtop-mode-tools.

Last edited by holstein (2011-03-21 17:48:04)

smartass · 2011-03-20 14:57:54

I have this problem too, or at least something similar.
First I though the reason was that I was using "$ sudo pm-suspend", and not as real root "# pm-suspend", but it turns out that it's not that. It worked in KDE4 a while ago, but I recently started using E17 an CLI, so can't be sure if it's that or the kernel upgrade to 2.6.37.4-1.

This is my experimental analysis so far:
If I suspend "# pm-supend" and then resume, fan will run for a second and then stop working.
If I let the temperature rise above a certain threshold (t>44 for me) and then suspend and immediately resume, the fan will work, because the temperature will still be above the threshold, but it seems to work differently then usual.
This happens regardless of the power state (ac/battery), but if I plug in AC after resume, something resets and the fan starts working again. I'm using laptop-mode-tools, I wonder if that could be connected.
Anyways, the fan starts working if t>50 (seems to be a trip-point, is there a file for that somewhere?) or so, but I find this behavior unsafe and concerning.

My assumption is that the thermal trip-points aren't read correctly upon resume, I think I read about a similar bug.

Does anyone more experienced have any idea how to debug this? My pm-suspend.log has just one line:
Activating Runtime PM for device type i2c

lagagnon · 2011-03-20 22:02:12

I don't believe these sort of fan problems after suspend/resume have anything to do with thermal trip points. IMHO they are the result of badly written BIOS code that does not play well with Linux or ACPI DSDT (https://wiki.archlinux.org/index.php/DSDT) that does not work well with your specific BIOS.

The only way I solved my problem was to flash updated BIOS onto my laptop (see the Toshiba Satellite L300 wiki page).

daroczig · 2011-03-21 00:19:51

@lagagnon: it might be true, but the problem occured only after upgrade to kernel 2.6.37.1 for me. Fans have been working correctly till 2.6.36 for ages with each kernel release.
Unfortunatelly there is no new BIOS upgrade for my laptop (HP 6715b) to try
And the fan does not start to operate even at 80 Celsius degrees after suspend, which sounds really dangerous to me. Well, using AMD, yet not normal heat

jlindgren · 2011-03-21 02:33:11

nblock wrote:

The fan control stops working until cpu reaches a temperature of 60°. As soon as the temperatur passes this threshold the fan starts rotating at max. speed. Below 60°, the fan is turned off immediately.

I've had this same problem on a Toshiba Satellite A305 since I bought the computer. As far I can tell it's an incompatibility between Linux and the BIOS. ("Insyde" BIOS. The problem never shows up in Windows.) The only thing I know of that seems to help (which I found through Google) is to use acpi_osi=Linux on the kernel command line. This isn't a perfect fix; but it seems to make the problem show up only after 5-10 suspend and resume cycles rather than after the first cycle.

EDIT: For me the problem doesn't seem tied to a particular kernel version. I think the earliest I had on this computer was 2.6.32 and I have 2.6.37.4 now.

Last edited by jlindgren (2011-03-21 02:34:50)

smartass · 2011-03-21 16:42:19

I'm also having this problem only recently, seems to be after the latest upgrade to 2.6.37.4-1
It used to work perfectly before.

Is there a way to determine if it's a BIOS problem? or which log files I should track?

EDIT: Btw, I also experienced this odd behavior upon reboot, repeatedly. Something is definitely not right!

Last edited by smartass (2011-03-21 20:18:17)

crispin · 2011-03-21 18:04:12

Same issue with HP Compaq 6715s.

daroczig · 2011-04-15 13:09:25

Seems to work at last again with kernel 2.6.38.2-1

smartass · 2011-04-15 14:26:04

yep, I just love these kind of issues that solve themselves eventually
Many thanks to the people that fixed this, though I don't know whom to give credit
hope this won't happen again

EDIT: nope, it's still there, rebooted my notebook while the temperature was still above 40 which I think is the first thermal trip point, so the fan turned on after the second trip point 50. Weird, but not really life threatening

I'm planning on making a clean, separate install (want to try out 64 bit arch and ext4 with e4rat) during the easter holidays, so I might use it afterwards to experiment with the kernel.

Last edited by smartass (2011-04-18 15:26:42)

Arch Linux

#1 2011-02-03 08:59:23

Fan Control fails after resume from pm-suspend

#2 2011-02-03 15:14:31

Re: Fan Control fails after resume from pm-suspend

#3 2011-02-03 15:35:00

Re: Fan Control fails after resume from pm-suspend

#4 2011-02-06 08:23:27

Re: Fan Control fails after resume from pm-suspend

#5 2011-02-16 00:55:33

Re: Fan Control fails after resume from pm-suspend

#6 2011-02-20 19:43:46

Re: Fan Control fails after resume from pm-suspend

#7 2011-02-20 21:43:42

Re: Fan Control fails after resume from pm-suspend

#8 2011-02-20 22:14:34

Re: Fan Control fails after resume from pm-suspend

#9 2011-02-21 14:53:22

Re: Fan Control fails after resume from pm-suspend

#10 2011-02-23 14:12:34

Re: Fan Control fails after resume from pm-suspend

#11 2011-03-02 23:46:04

Re: Fan Control fails after resume from pm-suspend

#12 2011-03-13 14:14:36

Re: Fan Control fails after resume from pm-suspend

#13 2011-03-17 19:06:18

Re: Fan Control fails after resume from pm-suspend

#14 2011-03-17 20:50:29

Re: Fan Control fails after resume from pm-suspend

#15 2011-03-20 14:57:54

Re: Fan Control fails after resume from pm-suspend

#16 2011-03-20 22:02:12

Re: Fan Control fails after resume from pm-suspend

#17 2011-03-21 00:19:51

Re: Fan Control fails after resume from pm-suspend

#18 2011-03-21 02:33:11

Re: Fan Control fails after resume from pm-suspend

#19 2011-03-21 16:42:19

Re: Fan Control fails after resume from pm-suspend

#20 2011-03-21 18:04:12

Re: Fan Control fails after resume from pm-suspend

#21 2011-04-15 13:09:25

Re: Fan Control fails after resume from pm-suspend

#22 2011-04-15 14:26:04

Re: Fan Control fails after resume from pm-suspend

Board footer