public inbox for linux-acpi@vger.kernel.org
 help / color / mirror / Atom feed
* acpi change between 2.6.8.1 and 2.6.17 breaks Tyan S2466N power button
@ 2007-02-07 17:55 David Mathog
  2007-02-07 18:26 ` Alexey Starikovskiy
  0 siblings, 1 reply; 4+ messages in thread
From: David Mathog @ 2007-02-07 17:55 UTC (permalink / raw)
  To: linux-acpi

I'm trying to upgrade a cluster from Mandrake 10.1 (with 2.6.8.1
vanilla kernel) to Mandriva 2007 (2.6.17-8mdv kernel).  These machines
have Tyan S2466N-4M motherboards, single AMD Athlon MP 2200+ processors,
and the latest (and no doubt last) BIOS = 4.06.  ACPI is enabled in
the BIOS.

One test node was upgraded and worked correctly except for one
annoying problem:  afterwards that node would
no longer respond to the front power button following a
"poweroff".  To reboot that system it is now necessary to pull
the power cord, count to 20, plug it back in again, and then push
the power button.

There were a variety of changes to the acpi between those kernels
versions, but the one that I think maybe indicates the problem
is the new 

  /proc/acpi/wakeup

file, which on the problem machine contains only:

Device    Sleep state     Status
PCIO         5            disabled
OP2P         5            disabled
USB0         1            disabled

So I tried: echo "OP2P" > /proc/acpi/wakeup
which strangely enabled both of the first two
entries, neither of which was "* enabled", just "enabled".  This
was not enough to make the power button work following
a "poweroff".  

Those 3 wakeup devices were all that was listed in /var/log/dmesg,
so it isn't that the OS dropped the power button after the BIOS
reported its presence, it seems that the BIOS never told the OS
that the power button exists.  (How this worked with the older
OS I have no idea).  Sleep states listed in /var/log/dmesg are:

ACPI:  (supports S0 S1 S4 S5)

Here are all the ACPI lines in /var/log/dmesg:

 BIOS-e820: 000000003fef0000 - 000000003feff000 (ACPI data)
 BIOS-e820: 000000003feff000 - 000000003ff00000 (ACPI NVS)
ACPI: RSDP (v000 PTLTD                                 ) @ 0x000f7100
ACPI: RSDT (v001 PTLTD    RSDT   0x06040000  LTP 0x00000000) @ 0x3fefcf36
ACPI: FADT (v001 AMD    TECATE   0x06040000 PTL  0x000f4240) @ 0x3fefef3c
ACPI: MADT (v001 PTLTD           APIC   0x06040000  LTP 0x00000000) @
0x3fefefb0
ACPI: DSDT (v001    AMD  AMDACPI 0x06040000 MSFT 0x0100000d) @ 0x00000000
ACPI: PM-Timer IO Port: 0x8008
ACPI: Local APIC address 0xfee00000
ACPI: LAPIC (acpi_id[0x00] lapic_id[0x01] enabled)
ACPI: LAPIC_NMI (acpi_id[0x00] high edge lint[0x1])
ACPI: IOAPIC (id[0x00] address[0xfec00000] gsi_base[0])
ACPI: INT_SRC_OVR (bus 0 bus_irq 0 global_irq 2 high edge)
ACPI: IRQ0 used by override.
ACPI: IRQ2 used by override.
ACPI: IRQ9 used by override.
Using ACPI (MADT) for SMP configuration information
ACPI: Looking for DSDT in initramfs... error, file /DSDT.aml not found.
ACPI: bus type pci registered
ACPI: Subsystem revision 20060127
ACPI: Interpreter enabled
ACPI: Using IOAPIC for interrupt routing
ACPI: PCI Root Bridge [PCI0] (0000:00)
ACPI: Assume root bridge [\_SB_.PCI0] bus is 0
ACPI: PCI Interrupt Routing Table [\_SB_.PCI0._PRT]
ACPI: PCI Interrupt Routing Table [\_SB_.PCI0.AGP_._PRT]
ACPI: PCI Interrupt Routing Table [\_SB_.PCI0.OP2P._PRT]
ACPI: PCI Interrupt Link [LNKA] (IRQs 3 5 10 *11)
ACPI: PCI Interrupt Link [LNKB] (IRQs 3 5 10 11) *0, disabled.
ACPI: PCI Interrupt Link [LNKC] (IRQs 3 5 10 11) *0, disabled.
ACPI: PCI Interrupt Link [LNKD] (IRQs 3 5 *10 11)
pnp: PnP ACPI init
pnp: PnP ACPI: found 13 devices
PCI: Using ACPI for IRQ routing
apm: overridden by ACPI.
ACPI wakeup devices:
ACPI: (supports S0 S1 S4 S5)

On other systems (different motherboards and processors) with
recent kernels this /proc/acpi/wakeup contains either

PWRB   3   * enabled     (Mandriva 2006)

or

SLPB   4   * enabled     (Mandriva 2007)

and those machines do boot from a power button push following
"poweroff".

Boot parameters for both the working and nonworking Mandriva 2007
machines were the same (other than different partition numbers):

image=/boot/vmlinuz
        label="linux"
        root=/dev/hda5
        initrd=/boot/initrd.img
        append="resume=/dev/hda2"

The ACPI entries in config are:

CONFIG_ACPI=y
CONFIG_ACPI_SLEEP=y
CONFIG_ACPI_SLEEP_PROC_FS=y
CONFIG_ACPI_AC=m
CONFIG_ACPI_BATTERY=m
CONFIG_ACPI_BUTTON=m
CONFIG_ACPI_VIDEO=m
CONFIG_ACPI_HOTKEY=m
CONFIG_ACPI_FAN=m
CONFIG_ACPI_PROCESSOR=m
CONFIG_ACPI_HOTPLUG_CPU=y
CONFIG_ACPI_THERMAL=m
CONFIG_ACPI_ASUS=m
CONFIG_ACPI_IBM=m
CONFIG_ACPI_IBM_DOCK=y
CONFIG_ACPI_TOSHIBA=m
CONFIG_ACPI_CUSTOM_DSDT_INITRD=y
CONFIG_ACPI_BLACKLIST_YEAR=0
CONFIG_ACPI_EC=y
CONFIG_ACPI_POWER=y
CONFIG_ACPI_SYSTEM=y
CONFIG_ACPI_CONTAINER=m

What steps do I have to take to get a power button 
device into /proc/acpi/wakeup???

Any idea why the older kernel's ACPI variant was happy start up
from the front panel power switch following "poweroff", but the
newer one isn't? (There were no specific acpi kernel switches).

Thank you,

David Mathog
mathog@caltech.edu
Manager, Sequence Analysis Facility, Biology Division, Caltech

^ permalink raw reply	[flat|nested] 4+ messages in thread
* Re: acpi change between 2.6.8.1 and 2.6.17 breaks Tyan S2466N power button
@ 2007-02-07 18:37 David Mathog
  2007-02-07 18:56 ` Alexey Starikovskiy
  0 siblings, 1 reply; 4+ messages in thread
From: David Mathog @ 2007-02-07 18:37 UTC (permalink / raw)
  To: linux-acpi; +Cc: Alexey Starikovskiy



> Could it be that you don't load button module?

Good guess but "button" is loaded, that was the first
thing I checked.  Here is the output of lsmod:

Module                  Size  Used by
ipv6                  237120  12
w83627hf               24848  0
nfs                   215916  3
lockd                  56360  2 nfs
nfs_acl                 3776  1 nfs
sunrpc                137532  4 nfs,lockd,nfs_acl
eeprom                  6320  0
w83781d                31368  0
hwmon_vid               3392  2 w83627hf,w83781d
i2c_isa                 4352  2 w83627hf,w83781d
i2c_amd756              6436  0
i2c_core               17536  5 w83627hf,eeprom,w83781d,i2c_isa,i2c_amd756
3c59x                  40616  0
mii                     5664  1 3c59x
af_packet              17960  2
iptable_filter          3232  0
iptable_mangle          3104  0
iptable_nat             7652  0
ip_nat                 15692  1 iptable_nat
ip_conntrack           48320  2 iptable_nat,ip_nat
nfnetlink               5912  2 ip_nat,ip_conntrack
iptable_raw             2496  0
ip_tables              12740  4
iptable_filter,iptable_mangle,iptable_nat,iptable_raw
x_tables               12132  2 iptable_nat,ip_tables
video                  13924  0
thermal                11240  0
fan                     4036  0
container               3968  0
button                  5808  0
battery                 8260  0
ac                      4292  0
ide_cd                 35776  0
binfmt_misc             9800  1
loop                   15400  0
dm_mod                 49944  0
cpufreq_ondemand        6956  0
cpufreq_conservative     6952  0
cpufreq_powersave       2400  0
freq_table              4800  0
processor              21448  1 thermal
amd_k7_agp              7660  1
agpgart                27240  1 amd_k7_agp
hw_random               5688  0
evdev                   8864  1
tsdev                   7040  0
ext3                  124936  3
jbd                    51028  1 ext3


Next guess?

Thanks,

David Mathog
mathog@caltech.edu
Manager, Sequence Analysis Facility, Biology Division, Caltech

^ permalink raw reply	[flat|nested] 4+ messages in thread
end of thread, other threads:[~2007-02-07 18:57 UTC | newest]

Thread overview: 4+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2007-02-07 17:55 acpi change between 2.6.8.1 and 2.6.17 breaks Tyan S2466N power button David Mathog
2007-02-07 18:26 ` Alexey Starikovskiy
  -- strict thread matches above, loose matches on Subject: below --
2007-02-07 18:37 David Mathog
2007-02-07 18:56 ` Alexey Starikovskiy

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox