acpi problem - Hardware

This is a discussion on acpi problem - Hardware ; My desktop computer uses a DFI AM33-EL motherboard and Slackware Linux 10.2, which provides kernel versions 2.4.31 and 2.6.13. I want the machine to power down nicely when the power button is pressed, but this is not working. The kernels ...

+ Reply to Thread
Results 1 to 3 of 3

Thread: acpi problem

  1. acpi problem

    My desktop computer uses a DFI AM33-EL motherboard and Slackware Linux
    10.2, which provides kernel versions 2.4.31 and 2.6.13. I want the
    machine to power down nicely when the power button is pressed, but this
    is not working.

    The kernels I am using both include ACPI support, (as does the
    motherboard). The evidence is that the 'halt' command shuts down and
    turns off the power using either version of the kernel (although not
    the 2.4.31 kernel instaled 'out of the box'.

    As I understand it, pressing the power button generates an interrupt
    (9) and the kernel then puts a message into /proc/acpi/events. The
    acpid daemon is woken by the interrupt and reads the file, logs a
    message to /var/log/acpid and takes action prescribed by files in
    /etc/acpi. None of this appears to occur on my machine.

    I have used cat /proc/interrupts to check what is happening. The line
    for acpi does not change after the power button is pressed. The line
    for keyboard, etc all change as appropriate. I have checked the BIOS
    setup and ACPI is enabled.

    Any help or constructive comments would be appreciated

    PeterM


  2. Re: acpi problem

    On 18 Jan 2007 05:16:04 -0800, mitchell.peter.j@edumail.vic.gov.au
    staggered into the Black Sun and said:
    > kernel versions 2.4.31 and 2.6.13. I want the machine to power down
    > nicely when the power button is pressed, but this is not working.
    >
    > The kernels I am using both include ACPI support, (as does the
    > motherboard). The evidence is that the 'halt' command shuts down and
    > turns off the power


    That's not evidence for ACPI working--it could be using APM, after all.
    You need to make sure ACPI support has been loaded, and ACPI support for
    the devices you're using has been loaded as well. Make sure that
    /proc/acpi contains button/ . If not, modprobe button.

    > As I understand it, pressing the power button generates an interrupt
    > (9) and the kernel then puts a message into /proc/acpi/events. acpid
    > is woken by the interrupt and reads the file, logs a message to
    > /var/log/acpid and takes action prescribed by files in /etc/acpi. None
    > of this appears to occur on my machine.


    Is acpid running? "ps auxw | grep acpi" should show kacpid and acpid.
    Also note that pressing the button may or may not generate the events
    you expect. On my laptop, closing the lid generates an event of type
    button/lid , opening it generates an event of type processor/processor .

    > I have used cat /proc/interrupts to check what is happening. The line
    > for acpi does not change after the power button is pressed.


    Funky. I'm getting an ACPI interrupt about every second or so without
    anything happening, but laptops are weird that way.

    --
    If you go hit yourself with a plank, it would make a freely available
    Unix-like operating system to everyone.
    --MegaHAL, trained on RMS and the BOFH
    Matt G|There is no Darkness in Eternity/But only Light too dim for us to see

  3. Re: acpi problem


    Dances With Crows wrote:
    > That's not evidence for ACPI working--it could be using APM, after all.
    > You need to make sure ACPI support has been loaded, and ACPI support for
    > the devices you're using has been loaded as well. Make sure that
    > /proc/acpi contains button/ . If not, modprobe button.


    AHA!

    Yes ACPI was running (kernel had it in, APM out, various ACPI lines in
    dmesg).
    But I had not loaded the button module.

    Thanks

    PeterM


+ Reply to Thread