Port-i386 archive

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Old Index]

ACPI quirk for ASUS CUR-DLS



                Hi,

I wrote a while ago about an issue with the CUR-DLS motherboard from
ASUS:
http://mail-index.netbsd.org/port-i386/2008/02/14/msg000199.html

I also realized I am not the first one to have observed this:
http://mail-index.netbsd.org/port-i386/2005/09/09/0011.html

I had the curiosity to check what Linux does in this regard. It turns
out they have a specific ACPI quirk, that disables the use of ACPI
altogether on this board:

arch/i386/kernel/dmi_scan.c, from Linux 2.4.36.2:
 902         { force_acpi_ht, "ASUS CUR-DLS", {
 903                         MATCH(DMI_BOARD_VENDOR, "ASUSTeK Computer INC."),
 904                         MATCH(DMI_BOARD_NAME, "CUR-DLS"),
 905                         NO_MATCH, NO_MATCH }},

where the force_acpi_ht does:

 501         if (!acpi_force) {
 502                 printk(KERN_NOTICE "%s detected: force use of acpi=ht\n", 
d->ident);
 503                 disable_acpi();

And in fact, booting the 4.0 GENERIC.MP kernel with "disable acpi" works
perfectly so far. Therefore, I figured I would try to add a
quirk for NetBSD. This seems to be done in dev/acpi/acpi_quirks.c, where
I added this:

 65         { ACPI_TABLE_DSDT, "ASUS  ", 0x00000000, AQ_GTE, "CUR-DLS ",
 66           ACPI_QUIRK_BROKEN },

But this does not fix the problem: the early AML breakpoint is still
hit, and then ACPI still seems to be used. How can I determine if this
entry is correct?

Cheers,
-- 
khorben



Home | Main Index | Thread Index | Old Index