Date: Tue, 13 Jul 2004 23:41:43 +0900 From: Taku YAMAMOTO <taku@tackymt.homeip.net> To: freebsd-acpi@freebsd.org Subject: Re: [FOUND}: CPU Cx states not working with recent kernel on Thinkpad Message-ID: <20040713234143.051ddd3e.taku@tackymt.homeip.net> In-Reply-To: <20040713122312.GA47428@webcom.it> References: <20040706084203.GB19813@webcom.it> <20040713122312.GA47428@webcom.it>
next in thread | previous in thread | raw e-mail | index | archive | help
On Tue, 13 Jul 2004 14:23:13 +0200
Andrea Campi <andrea@webcom.it> wrote:
> Hi,
>
> On Tue, Jul 06, 2004 at 10:42:05AM +0200, Andrea Campi wrote:
> > I upgraded my Thinkpad 570E from a beginning of June kernel to new
> > sources cvsup'd after the recent changes (ACPICA import etc) a couple
> > of days ago. I now get this:
> >
> > ...
> > cpu0: <ACPI CPU (3 Cx states)> port 0x530-0x537 on acpi0
> > cpu0: Failed to attach throttling P_CNT
>
> I was able to determine this issue appeared with acpi.c rev 1.159 and
> related ("Add support to ACPI to manage its own resources."). Looking
> at devinfo -vr before and after the commit (see below) seems to
> confirm that something changed, although I can't understand how that
> caused the breakage. In fact, all the involved I/O ports are not
> reserved by any acpi_sysresource*, so I have no idea where to go
> from here.
>
> Nate (or anybody else), any idea?
I suspect it's because acpi_timer0 gets attached and sits on
not-yet-managed resources before acpi_sysresource0 reserves
resources for later use.
We can confirm that by temporarily disabling acpi_timer by
debug.acpi.disabled="timer" to see the working acpi_cpu.
Can we defer the attachment of acpi_timer0 after acpi_sysresource?
--
-|-__ YAMAMOTO, Taku
| __ < <taku@tackymt.homeip.net>
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20040713234143.051ddd3e.taku>