To follow-up, I returned the Dm1-1040us to HP and they replaced it with another (identical) machine. I no longer have any issues with the keyboard/touchpad freezing. I've been using the machine pretty heavily for two weeks and have had only one crash (under heavy load) but it looked to be completely unrealted.
Same thing. They replaced only the keyboard and since (twos weeks, heavy work) no keyboard freeze. Some hardlock when I'm plugin or unplugin the computer, but this is a totally different issue.
Send your computer back to HP, they do the job.
]]>The command acpi -V returns the following before the freeze:
Thermal 0: ok, 59.0 degrees C
Thermal 0: trip point 0 switches to mode hot at temperature 100.0 degrees C
Thermal 0: trip point 1 switches to mode passive at temperature 64.0 degrees C
Until here, now keyboard and touchpad are frozen:
Thermal 0: passive, 64.0 degrees C
Thermal 0: trip point 0 switches to mode hot at temperature 100.0 degrees C
Thermal 0: trip point 1 switches to mode passive at temperature 64.0 degrees C
At which point I can suspend the machine, let it cool down a little (just a couple seconds really) wake it back up and everything is fine until it hits 64 degrees again.
I haven't tested turning off thermal in the kernel yet -- I wanted to first play with the BIOS setting that forces the fan to run. While I get the error in both cases, the system seems to run much cooler if the BIOS setting is *disabled* that is if the fan is not set to run all the time the system stays cooler.
Would be nice if we could keep the Linux thermal settings enabled and turn-off the BIOS fan setting -- that seems like it would produce the best results.
[update 12-7-17]
Unfortunately, the temperature was a red herring. I get the freezes at different points and the trip point is not part of the equation. In fact, now when I run acpi -V the first trip point is at 97 degrees and not 64 like I posted above.
I tried multiple combinations of changing the fan BIOS setting and the thermal.off=1 setting with little to no effect. Once the machine has frozen, it seems like it will do so again almost immediately after a reboot and login. Just typing into the terminal will eventually cause a freeze of the keyboard.
I can reboot into Windows and everything seems to run fine. I'm starting to think this isn't heat related. And it doesn't seem dependent on cpu, disk, or network traffic.
]]>I can usually suspend the machine and restore will re-activate the keyboard/touchpad for a period time.
From the HP forum it appears that one of the thermal registers may be triggering a freeze -- which may may explain the randomness. The suggestion is to disable thermal events in kernel (pass boot parameter: thermal.off=1). Has anyone here tried this yet to see if these problems are all related? Example provided for updating GRUB:
example /etc/default/grub:
Edit:
GRUB_CMDLINE_LINUX_DEFAULT="quiet splash"
To:
GRUB_CMDLINE_LINUX_DEFAULT="quiet splash thermal.off=1"
Then
update-grub
This appears to "fix" the problem and the BIOS has a setting (on by default) to run the fan always.
I have not seen the freeze in Windows yet. I guess the real question here is whether this is a defect in the hardware or just a bug to be fixed in kernel and/or drivers for Linux and Windows.
]]>I noticed the freeze occurs more frequently when I move the laptop or when I put it on surface that is not very stable (bed, sofa etc) so maybe....
Or perhaps a blocked cooling air vent or intake?
]]>I experience the same problem on both Linux and Windows7.
I noticed the freeze occurs more frequently when I move the laptop or when I put it on surface that is not very stable (bed, sofa etc) so maybe the built-in accelerometer (HP CoolSense) has something to do with it?
BTW When it happens on Windows7, all I need to get it working again is to log out. No need to restart/suspend. Just log out using USB mouse and just before the windows login screen appears the touchpad and keyboard are working again.
]]>- it also happen on Window 7, but I was only able to get this issue two times since I have the laptop (even if I'm playing diablo III three hours a day on it + testing and compiling all my work on it). So perhaps the twos freezes I got are another issue (except that they behaved the same way)
- it may happen without loading any wireless or network module
- if I have brcmsmac/wl driver loaded and doing an heavy work, it freezes in ten seconds when it can takes minutes without the driver loaded
- unpluging the battery and pressing the power button for a few seconds (which may remove the residual power) helps a bit (ie, the issue takes more time to appears)
Something off topic, but did you get working wireless with the brcmsmac driver ? The wl one does not detect all networks and suffer from connection lost, but the brcmsmac see all the network with only 10% of signal strength)
]]>Though usually, even when not doing anything data-heavy, I still get the lockups. It's quite frustrating, and sometimes happens regularly.
Try to blacklist btusb module (bluetooth). It could reduce number of freezes - workaround but better than nothing.
I hope my brand new lapop (only 1.5 months old) doesn't need a motherboard repair already!
Probably it isn't hardware problem because everything works in windows 7 - see my comment above. I suppose that HP hw is strange but windows drivers somehow work with it. I'll write to kernel mailing list and I hope that they will be able to solve our problem.
I think there is an issue with heat as well.
I don't think so. I have temperature still around 60 C (hdd has 40 C) and IMHO it is ok. I didn't notice higher fan activity in windows. I think that problem is really in strange HP usb (or bus).
]]>You people seem to have correlated the issue with a possible conflict with either the Linux kernel or the Broadcom WL driver (or both). I've just now uninstalled the proprietary Broadcom wireless driver, and I'm using the open source driver. I'll see if I keep getting the lockups. Though usually, even when not doing anything data-heavy, I still get the lockups. It's quite frustrating, and sometimes happens regularly. I hope my brand new lapop (only 1.5 months old) doesn't need a motherboard repair already!
EDIT: I am still getting the lockups. Noticing another pattern - the first lockup takes quite a while to occur, whereas the following lockups are much more frequent, sometimes soon after a laptop restart. I think there is an issue with heat as well. In fact I don't think the fans are being controlled well - they should be working more then they are, but even when the left side of the DM1 is quite hot when I touch it, the cooling fan is barely audible (I have to lower my ear to the vent to hear something), which is unusual for a laptop that has been criticised for having quite a loud cooling fan! I think the Linux kernel is just not handling this laptop well....
]]>- boot from HP recovery dvd
- start windows 7 recovery
- soft freeze during copy of dvd number 3
- boot fresh installation of 64 bit windows 7
- copy 5 hp recovery dvds (20 GB) on hdd and no freeze
It seems that I get soft freeze always when some amount of data is sent thru usb/bus. But everything works in windows.
]]>- connected external mouse + dvd, kernel 3.4.4-1, blacklist: wl, snd_hda_intel, btusb, brcmsmac
- run for 3 hours without work, no freeze
- restart
- copy dvd to hdd, soft freeze when 3 GB were copied
- modprobe -r psmouse; modprobe psmouse - didn't work
- copy finished (4 GB)
- modprobe -r psmouse; modprobe psmouse - solved the problem, keyboard + touchpad worked again
- restart
- play openttd for 2 hours without sound (because of blacklisted snd_hda_intel) - no freeze
- remove snd_hda_intel from blacklist
- restart
- play openttd with sound and soft freeze after 1 hour
- restart
- connected external mouse + dvd
- boot live Ubuntu 12.04 from flash disc
- copy dvd to /tmp (memory disc), soft freeze when 500 MB were copied
I'll install windows 7 and test if I get soft freeze also in windows.
]]>Also I had a totally different freeze, the mouse was not frozen at all, only the keyboard, but modprobe -r psmouse did not bring the keyboard back... I had to sleep the computer to get it back.
]]> CPU0 CPU1
0: 52 0 IO-APIC-edge timer
1: 24 3264 IO-APIC-edge i8042
8: 0 1 IO-APIC-edge rtc0
9: 1438 187822 IO-APIC-fasteoi acpi
12: 81 4478 IO-APIC-edge i8042
16: 2 413 IO-APIC-fasteoi snd_hda_intel
17: 3 101 IO-APIC-fasteoi ehci_hcd:usb1, ehci_hcd:usb2, ehci_hcd:usb3
18: 1 26 IO-APIC-fasteoi ohci_hcd:usb4, ohci_hcd:usb5, ohci_hcd:usb6
19: 120 52539 IO-APIC-fasteoi ahci
41: 5536 5047134 PCI-MSI-edge fglrx[0]@PCI:0:1:0
42: 2 97 PCI-MSI-edge snd_hda_intel
43: 826 1014193 PCI-MSI-edge eth0
NMI: 3756 3761 Non-maskable interrupts
LOC: 13869735 13879818 Local timer interrupts
SPU: 0 0 Spurious interrupts
PMI: 3756 3761 Performance monitoring interrupts
IWI: 0 0 IRQ work interrupts
RTR: 0 0 APIC ICR read retries
RES: 1421768 1364352 Rescheduling interrupts
CAL: 331 23 Function call interrupts
TLB: 1717 2771 TLB shootdowns
TRM: 0 0 Thermal event interrupts
THR: 0 0 Threshold APIC interrupts
MCE: 0 0 Machine check exceptions
MCP: 129 129 Machine check polls
ERR: 0
MIS: 0
So eth0 is on its own line, so there may be no issue. Can you confirm I'm right ?
And with the broadcom wl driver, it is interesting:
CPU0 CPU1
0: 54 0 IO-APIC-edge timer
1: 54 2642 IO-APIC-edge i8042
8: 0 1 IO-APIC-edge rtc0
9: 4 632 IO-APIC-fasteoi acpi
12: 35 1665 IO-APIC-edge i8042
16: 306 55309 IO-APIC-fasteoi snd_hda_intel, eth0
17: 0 104 IO-APIC-fasteoi ehci_hcd:usb1, ehci_hcd:usb2, ehci_hcd:usb3
18: 1 26 IO-APIC-fasteoi ohci_hcd:usb4, ohci_hcd:usb5, ohci_hcd:usb6
19: 59 21582 IO-APIC-fasteoi ahci
41: 0 2 PCI-MSI-edge fglrx[0]@PCI:0:1:0
42: 2 98 PCI-MSI-edge snd_hda_intel
NMI: 4 7 Non-maskable interrupts
LOC: 28796 44487 Local timer interrupts
SPU: 0 0 Spurious interrupts
PMI: 4 7 Performance monitoring interrupts
IWI: 0 0 IRQ work interrupts
RTR: 0 0 APIC ICR read retries
RES: 29199 25170 Rescheduling interrupts
CAL: 58 19 Function call interrupts
TLB: 871 1006 TLB shootdowns
TRM: 0 0 Thermal event interrupts
THR: 0 0 Threshold APIC interrupts
MCE: 0 0 Machine check exceptions
MCP: 3 3 Machine check polls
ERR: 0
MIS: 0
eth0 is on the same line as snd_hda_intel.
By the way, my previous test runs all night and this morning when I touched the keyboard, it lags for half a second and get me a key input.... Then I modprobe r8169 and dhcpcd eth0 and after 2 minutes, it keyboard freezes. So there is definitly something here.
]]>This stuff is Weird.
...So now I'm wondering why the network can generate the mouse freeze. Is the ethernet AND wifi card broken, is it the driver ? Is it something on the kernel which is not used because of some optimization of the local loopback ?
Interrupt conflicts? See what is sharing interrupts with what.
...I'll have to sleep. More tomorrow.
Sleep is overrated.
]]>