ACPI patches updated (20021111)

ACPI patches updated (20021111)

[Grover, Andrew]

Hi all,

New patches are up on sf.net/projects/acpi. Non-Linux-specific releases will
be available by tomorrow evening at
http://developer.intel.com/technology/iapc/acpi/downloads.htm .

If you've been having problems with reading battery or other information
being very slow, please try this release and report if problems persist.

Regards -- Andy

----------------------------------------
11 November 2002.  Summary of changes for version 20021111.


0) ACPI Specification 2.0B is released and is now available
at:  http://www.acpi.info/index.html

1) Linux

Module loading/unloading fixes (John Cagle)

Other minor fixes

2) ACPI CA Core Subsystem:

Implemented support for the ACPI 2.0 SMBus Operation Regions.
This includes the early detection and handoff of the request
to the SMBus region handler (avoiding all of the complex field
support code), and support for the bidirectional return packet
from an SMBus write operation.  This paves the way for the
development of SMBus drivers in each host operating system.

Fixed a problem where the semaphore WAIT_FOREVER constant was
defined as 32 bits, but must be 16 bits according to the ACPI
specification.  This had the side effect of causing ASL
Mutex/Event timeouts even though the ASL code requested a wait
forever.  Changed all internal references to the ACPI timeout
parameter to 16 bits to prevent future problems.  Changed the
name of WAIT_FOREVER to ACPI_WAIT_FOREVER.



-----------------------------
Andrew Grover
Intel Labs / Mobile Architecture
andrew.grover@intel.com

Re: ACPI patches updated (20021111)

[Stephen Hemminger]

Will this fix problems with IRQ routing.
On our SMP test machines, ACPI has to be disabled otherwise the SCSI
disk controllers don't work.

This is a major pain, and ACPI should be default off until it gets
fixed.

Re: ACPI patches updated (20021111)

[Zed Pobre]

[-- Attachment #1: Type: text/plain, Size: 772 bytes --]

On Wed, Nov 13, 2002 at 01:33:03PM -0800, Stephen Hemminger wrote:
> Will this fix problems with IRQ routing.
> On our SMP test machines, ACPI has to be disabled otherwise the SCSI
> disk controllers don't work.

    As a further data point, if ACPI is enabled on my non-SMP test
machine, USB stops working.


> This is a major pain, and ACPI should be default off until it gets
> fixed.

    Well, this kind of goes both ways, unfortunately.  If ACPI is
disabled on that same test machine, the network stops working,
complaining about possible IRQ conflicts (I sent mail earlier about
this but got not response).

-- 
Zed Pobre <zed@debian.org> a.k.a. Zed Pobre <zed@resonant.org>
PGP key and fingerprint available on finger; encrypted mail welcomed.

[-- Attachment #2: Type: application/pgp-signature, Size: 481 bytes --]

Re: ACPI patches updated (20021111)

[Zed Pobre]

[-- Attachment #1: Type: text/plain, Size: 758 bytes --]

On Fri, Nov 15, 2002 at 09:01:13AM -0600, Zed Pobre wrote:
> On Wed, Nov 13, 2002 at 01:33:03PM -0800, Stephen Hemminger wrote:
> > Will this fix problems with IRQ routing.
> > On our SMP test machines, ACPI has to be disabled otherwise the SCSI
> > disk controllers don't work.
> 
>     As a further data point, if ACPI is enabled on my non-SMP test
> machine, USB stops working.
[... and if ACPI is disabled, the net stops working...]

    I take this back.  After seeing the thread about buggy IO-APIC on
uniprocessor machines, I tested disabling IO-APIC, and all of my
problems magically went away.

-- 
Zed Pobre <zed@debian.org> a.k.a. Zed Pobre <zed@resonant.org>
PGP key and fingerprint available on finger; encrypted mail welcomed.

[-- Attachment #2: Type: application/pgp-signature, Size: 481 bytes --]

Re: ACPI patches updated (20021111)

[Stian Jordet]

fre, 2002-11-15 kl. 16:01 skrev Zed Pobre:
> On Wed, Nov 13, 2002 at 01:33:03PM -0800, Stephen Hemminger wrote:
> > Will this fix problems with IRQ routing.
> > On our SMP test machines, ACPI has to be disabled otherwise the SCSI
> > disk controllers don't work.
> 
>     As a further data point, if ACPI is enabled on my non-SMP test
> machine, USB stops working.
Very good to hear I'm not the only one with that problem, though, I have
a smp-machine, but it utterly refuses to work with acpi.

Regards,
Stian Jordet

Re: ACPI patches updated (20021111)

[Zed Pobre]

[-- Attachment #1: Type: text/plain, Size: 772 bytes --]

On Fri, Nov 15, 2002 at 11:29:00PM +0100, Stian Jordet wrote:
> fre, 2002-11-15 kl. 16:01 skrev Zed Pobre:
> > 
> >     As a further data point, if ACPI is enabled on my non-SMP test
> > machine, USB stops working.
>
> Very good to hear I'm not the only one with that problem, though, I have
> a smp-machine, but it utterly refuses to work with acpi.

    Interesting.  As I pointed out in my other reply, my problem was
being caused by IO-APIC being enabled, but the thread that lead me to
discovering this implied that it was only buggy for the uniprocessor
case.  This implies that it might also be buggy for SMP.

-- 
Zed Pobre <zed@debian.org> a.k.a. Zed Pobre <zed@resonant.org>
PGP key and fingerprint available on finger; encrypted mail welcomed.

[-- Attachment #2: Type: application/pgp-signature, Size: 481 bytes --]

RE: ACPI patches updated (20021111)

[Grover, Andrew]

> From: Stephen Hemminger [mailto:shemminger@osdl.org] 
> Will this fix problems with IRQ routing.
> On our SMP test machines, ACPI has to be disabled otherwise the SCSI
> disk controllers don't work.
> 
> This is a major pain, and ACPI should be default off until it gets
> fixed.

ACPI has not yet been adequately tested on machines with multiple IO-APICs.
More assistance in this area would be gratefully accepted. In the meantime,
"acpi=off" works pretty well to disable ACPI-related configuration problems.

Regards -- Andy