On Sat, 12 May 2007 07:04:25 -0700 (PDT) Tear <[EMAIL PROTECTED]> wrote:
> Summary: If ACPI is not enabled but APIC is,
> then there is trouble on Dell Optiplex GX240.
> If both are enabled or if both are disabled, then
> everything is fine. The attached patch removes
> Dell Optiplex GX240 from the ACPI blacklist.
>
> ---
>
> Hello,
>
> I have a Dell Optiplex GX240 and when I boot Linux, ACPI
> gets set up by only acpi=ht. dmesg shows the following line:
>
> DELL GX240 detected: force use of acpi=ht
>
> Everything seemed to be fine. However, I discovered that
> everything is not fine. The USB controller works so slowly
> that copying a few (uncached) 1 megabyte large photos from
> a USB-enabled digital camera takes many minutes instead of
> a couple of seconds.
>
> I am using Linux 2.6.21.1 on a Debian 4.0 ("Etch") system.
>
> I thought that this might be related to ACPI. So I tried
> to boot with _only_ "acpi=force" appended to the kernel
> command line. Voila, the USB controller started to work
> at full speed and copying photos from my digital camera
> took only seconds.
>
> I tested the system with "acpi=force" and could not find
> anything which did not work. So, can we please remove
> Dell Optiplex GX240 from the blacklist in
>
> ..../arch/i386/kernel/acpi/boot.c
>
> ? The attached patch does just that: It removes Dell
> Optiplex GX240 from the ACPI blacklist.
>
> I thought that this might be related to interrupts and
> APIC as well. (Note that this is APIC, not ACPI.) I tried
> booting with _only_ "noapic" and "nolapic" appended to the
> command line. Again, the USB controller started to work
> at full speed.
>
> If removing Dell Optiplex GX240 from the ACPI blacklist
> is not wanted/possible, then is there a way to disable
> APIC and LAPIC (note that this is APIC not ACPI) by
> default on Dell GX240 machines? (I.e. Can one patch
> the kernel so that APIC and LAPIC isn't used on
> these machines? - I know that I can use the noapic
> and nolapic options on the kernel command line.)
>
> Thank you for your attention.
>
> - Tear
>
> Note: Please include me in CC of your replies.
>
>
> --- linux-2.6.21.1.orig/arch/i386/kernel/acpi/boot.c 2007-05-01
> 17:10:30.000000000 +0300
> +++ linux-2.6.21.1/arch/i386/kernel/acpi/boot.c 2007-05-01
> 20:31:53.000000000 +0300
> @@ -971,14 +971,6 @@
> },
> {
> .callback = force_acpi_ht,
> - .ident = "DELL GX240",
> - .matches = {
> - DMI_MATCH(DMI_BOARD_VENDOR, "Dell Computer Corporation"),
> - DMI_MATCH(DMI_BOARD_NAME, "OptiPlex GX240"),
> - },
> - },
> - {
> - .callback = force_acpi_ht,
> .ident = "HP VISUALIZE NT Workstation",
> .matches = {
> DMI_MATCH(DMI_BOARD_VENDOR, "Hewlett-Packard"),
>
Thanks. Let's cc the acpi list.
The origin of that blacklist entry appears to be lost in the mists of time.
git-blame got tripped up by an intervening Lindent run and my gittiness
is insufficient for tracking changes before that.
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to [EMAIL PROTECTED]
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
