ACPI device for ASUS EEEPC 1101HA not added
Daniel Mack
because the asus-laptop driver fails to match the ACPI device named
"ASUS010". However, acpidump tells me a device with that name does
actually exist - the (simplified) disassembly of the DSDT is below.
It looks like the gigantic parser state machine fails to add that
device for whatever reason. Any ideas how to fix that?
Thanks,
Daniel
/*
* Intel ACPI Component Architecture
* AML Disassembler version 20090521
*/
DefinitionBlock ("out/DSDT-0x3f660430-0.aml", "DSDT", 2, "A1359", "A1359000", 0x00000000)
{
[...]
Scope (_SB)
{
[...]
Device (PCI0)
{
Name (_HID, EisaId ("PNP0A08"))
Name (_ADR, Zero)
[...]
Device (SBRG)
{
[...]
Scope (\_SB)
{
Name (ATKP, Zero)
Device (ATKD)
{
Name (_HID, "ASUS010")
Name (_UID, 0x01010100)
[...]
The full version of the dump is here:
http://caiaq.de/download/tmp/DSDT-0x3f660430-0.dsl
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majo...@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
andrej.g...@udo.edu
you need to whitelist your eee pc for OSI(Linux) in drivers/acpi/blacklist.c
like this:
+ /*
+ * On newer Eeepc, the interface used by eeepc-laptop (ASUS010)
+ * is disabled without _OSI(Linux)
+ */
+ {
+ .callback = dmi_enable_osi_linux,
+ .ident = "Asus Eeepc-1101HA",
+ .matches = {
+ DMI_MATCH(DMI_SYS_VENDOR, "ASUSTeK Computer INC."),
+ DMI_MATCH(DMI_PRODUCT_NAME, "1101HA"),
+ },
+ },
Regards,
Andrej
Daniel Mack writes:
> ------------------------------------------------------------------------------
> The Planet: dedicated and managed hosting, cloud storage, colocation
> Stay online with enterprise data centers and the best network in the business
> Choose flexible plans and management services without long-term contracts
> Personal 24x7 support from experience hosting pros just a phone call away.
> http://p.sf.net/sfu/theplanet-com
> _______________________________________________
> Acpi4asus-user mailing list
> Acpi4as...@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/acpi4asus-user
Daniel Mack
> you need to whitelist your eee pc for OSI(Linux) in drivers/acpi/blacklist.c
> like this:
>
> + /*
> + * On newer Eeepc, the interface used by eeepc-laptop (ASUS010)
> + * is disabled without _OSI(Linux)
> + */
> + {
> + .callback = dmi_enable_osi_linux,
> + .ident = "Asus Eeepc-1101HA",
> + .matches = {
> + DMI_MATCH(DMI_SYS_VENDOR, "ASUSTeK Computer INC."),
> + DMI_MATCH(DMI_PRODUCT_NAME, "1101HA"),
> + },
> + },
Great, thanks. That worked. Is that already queued anywhere or do you
want me to send a proper patch?
Daniel
> Daniel Mack writes:
>
> >On my new 1101HA Eeepc, the special function keys do not work. This is
> >because the asus-laptop driver fails to match the ACPI device named
> >"ASUS010". However, acpidump tells me a device with that name does
> >actually exist - the (simplified) disassembly of the DSDT is below.
> >
> >It looks like the gigantic parser state machine fails to add that
> >device for whatever reason. Any ideas how to fix that?
> >
> >
> >Thanks,
> >Daniel
> >
andrej.g...@udo.edu
there was a patch for 1005ha and 1000h (or so), but i can't find it
in the repository anymore. I miss it in 2.6.33-rc* too.
acpi/blacklist: enable OSI(Linux) on newer eeepc
patch from git://git.iksaif.net/acpi4asus.git.
Regards,
Andrej
Zhang Rui
> On my new 1101HA Eeepc, the special function keys do not work. This is
> because the asus-laptop driver fails to match the ACPI device named
> "ASUS010". However, acpidump tells me a device with that name does
> actually exist - the (simplified) disassembly of the DSDT is below.
>
> It looks like the gigantic parser state machine fails to add that
> device for whatever reason. Any ideas how to fix that?
>
~/src/linux-2.6$ grep ASUS010 drivers/platform/*/*
drivers/platform/x86/eeepc-laptop.c:#define EEEPC_ACPI_HID "ASUS010"
thanks,
rui
> To unsubscribe from this list: send the line "unsubscribe linux-acpi" in
Daniel Mack
> On Mon, 2010-02-01 at 11:25 +0800, Daniel Mack wrote:
> > On my new 1101HA Eeepc, the special function keys do not work. This is
> > because the asus-laptop driver fails to match the ACPI device named
> > "ASUS010". However, acpidump tells me a device with that name does
> > actually exist - the (simplified) disassembly of the DSDT is below.
> >
> > It looks like the gigantic parser state machine fails to add that
> > device for whatever reason. Any ideas how to fix that?
> >
> eeepc is handled in eeepci-laptop driver.
>
> ~/src/linux-2.6$ grep ASUS010 drivers/platform/*/*
> drivers/platform/x86/eeepc-laptop.c:#define EEEPC_ACPI_HID "ASUS010"
Yes, that I know. The driver was just unable to match the device because
it was not detected by the ACPI core. Works now with the patch Andrej
sent. I would still like to make sure it's getting merged mainline
sooner or later.
Thanks,
Daniel
Len Brown
> Hi,
>
> you need to whitelist your eee pc for OSI(Linux) in drivers/acpi/blacklist.c
> like this:
>
> + /*
> + * On newer Eeepc, the interface used by eeepc-laptop (ASUS010)
> + * is disabled without _OSI(Linux)
> + */
> + {
> + .callback = dmi_enable_osi_linux,
> + .ident = "Asus Eeepc-1101HA",
> + .matches = {
> + DMI_MATCH(DMI_SYS_VENDOR, "ASUSTeK Computer INC."),
> + DMI_MATCH(DMI_PRODUCT_NAME, "1101HA"),
> + },
> + },
Not necessarily the right fix. We have gone to a lot of trouble
to discourage BIOS vendors from depending on the ill-defined OSI(Linux),
so I hesitate to invoke it -- even for a workaround.
The problem at hand is that ASUS010 is not enabled for an OS
that claims compatibility with Win7 (MSOS() == MSW7) below.
Scope (\_SB)
{
Name (ATKP, Zero)
Device (ATKD)
{
Name (_HID, "ASUS010")
Name (_UID, 0x01010100)
Method (_STA, 0, NotSerialized)
{
If (LEqual (MSOS (), MSW7))
{
Return (Zero)
}
Else
{
Return (0x0F)
}
Return (Zero)
Return (0x0F)
}
(heh, see any indication of lack of quality in this code?:-)
MSOS() does this:
Scope (\)
{
Name (OSLX, 0x10)
Name (OSMS, 0x20)
Name (MS98, 0x21)
Name (MSME, 0x22)
Name (MS2K, 0x23)
Name (MSXP, 0x24)
Name (MSVT, 0x25)
Name (MSW7, 0x26)
Name (OSFG, Ones)
Method (MSOS, 0, NotSerialized)
{
If (LNotEqual (OSFG, Ones))
{
Return (OSFG)
}
Store (Zero, OSFG)
If (CondRefOf (_OSI, Local0))
{
If (_OSI ("Windows 2001"))
{
Store (MSXP, OSFG)
}
If (_OSI ("Windows 2001 SP1"))
{
Store (MSXP, OSFG)
}
If (_OSI ("Windows 2001 SP2"))
{
Store (MSXP, OSFG)
}
If (_OSI ("Windows 2006"))
{
Store (MSVT, OSFG)
}
If (_OSI ("Windows 2009"))
{
Store (MSW7, OSFG)
}
If (_OSI ("Linux"))
{
Store (OSLX, OSFG)
}
Return (OSFG)
}
Else
So I expect if you apply no patch, but boot with 'acpi_osi="!Windows 2009"'
then that would also work properly, as OSFG above will be set
to "MSVT" instead of "OSLX".
Looking through the DSDT, there are no references to OSLX or MSVT,
or MSXP, for that matter. However, there is an additional reference
to MSV7 in the temperature reading part of thermal zone TZ00,
that looks like some sort of OS-specific initialization, perhaps
a workaround. So also check that /proc/acpi/thermal_zone/*/temperature
still work before and after.
thanks,
-Len Brown, Intel Open Source Technology Center
Corentin Chary
This kind of things is present on all newer eeepc.
I tried to contact asus about that, without real success and it is
likely that the xandros distribribution
shipped with some eeepc include an acpi blacklisiting patch.
The thing is that on win7 ASUS010 is disabled, but an equivalent wmi
interface is enable, and it could
be possible to write a driver for it. Now, someone need to write it,
or send me a free 1005HA so I can do it :).
If nothing is done for 2.6.34, we may need to envisage some blacklisting :/.
Thanks
--
Corentin Chary
http://xf.iksaif.net
Matthew Garrett
looks pretty much identical to the existing one, it's just indirected
throguh WMI.
--
Matthew Garrett | mj...@srcf.ucam.org