[PATCH] ACPI update (20030122)

[PATCH] ACPI update (20030122)

[Grover, Andrew]

Hi all,

The latest ACPI patch is now available at http://sf.net/projects/acpi .
Non-Linux packages will be available within 24 hours from
http://developer.intel.com/technology/iapc/acpi/downloads.htm .

The patches are big primarily because we finally got with the program
and moved our headers from drivers/acpi/include to include/acpi.

Regards -- Andy

22 January 2003.  Summary of changes for version 20030122.


1) ACPI CA Core Subsystem:

Added a check for constructs of the form:  Store (Local0,
Local0) where Local0 is not initialized.  Apparently, some
BIOS programmers believe that this is a NOOP.  Since this
store doesn't do anything anyway, the new prototype behavior
will ignore this error.  This is a case where we can relax the
strict checking in the interpreter in the name of
compatibility.


2) Linux

The AcpiSrc Source Conversion Utility has been released with
the Unix package for the first time.  This is the utility
that is used to convert the ACPI CA base source code to the
Linux version.

(Both) Handle P_BLK lengths shorter than 6 more gracefully

(Both) Move more headers to include/acpi, and delete an unused
header.

(Both) Move drivers/acpi/include directory to include/acpi

(Both) Boot functions don't use cmdline, so don't pass it
around

(Both) Remove include of unused header (Adrian Bunk)

(Both) acpiphp.h includes both linux/acpi.h and acpi_bus.h.
Since the former now also includes the latter, acpiphp.h only needs the
one, now.

(2.5) Make it possible to select method of bios restoring
after S3 resume. [=> no more ugly ifdefs] (Pavel Machek)

(2.5) Make proc write interfaces work (Pavel Machek)

(2.5) Properly init/clean up in cpufreq/acpi (Dominik
Brodowski)

(2.5) Break out ACPI Perf code into its own module, under
cpufreq (Dominik Brodowski)

(2.4) S4BIOS support (Ducrot Bruno)

(2.4) Fix acpiphp_glue.c for latest ACPI struct changes
(Sergio Visinoni)


3) iASL Compiler:

Added support to disassemble SSDT and PSDTs.

Implemented support to obtain SSDTs from the Windows registry
if available.

-----------------------------
Andrew Grover
Intel Labs / Mobile Architecture
andrew.grover-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org

Re: [PATCH] ACPI update (20030122)

[Dave Jones]

On Thu, Jan 23, 2003 at 02:44:55PM -0800, Grover, Andrew wrote:

 > The latest ACPI patch is now available at http://sf.net/projects/acpi .
 > Non-Linux packages will be available within 24 hours from
 > http://developer.intel.com/technology/iapc/acpi/downloads.htm .

I've noticed that with .59 some of my boxes no longer have
functioning NICs unless I boot with acpi=off. Packets get
transmitted, but nothing ever gets received.
Seen this with a 3c509, an 8139, and an e100.

Known bug? Fixed in this patch ?

		Dave

-- 
| Dave Jones.        http://www.codemonkey.org.uk
| SuSE Labs

Re: [PATCH] ACPI update (20030122)

[Marco Walther]

A quick question,

is there a reason why iasl (iasl-linux-20021205) would need glibc-2.3?
I can play with different kernels on my box but if I've to upgrade my
libc as well, it's getting harder to maintain a stable stand:-(

Thanks,
-- Marco

-- 

-- 


-------------------------------------------------------
This SF.NET email is sponsored by:
SourceForge Enterprise Edition + IBM + LinuxWorld = Something 2 See!
http://www.vasoftware.com

Re: [PATCH] ACPI update (20030122)

[Sérgio Monteiro Basto]

you can donwnload acpica-unix-(...)
and compile it with:
cd /acpica-unix-(...)/compiler
make
On Fri, 2003-01-24 at 00:43, Marco Walther wrote:
> A quick question,
> 
> is there a reason why iasl (iasl-linux-20021205) would need glibc-2.3?
> I can play with different kernels on my box but if I've to upgrade my
> libc as well, it's getting harder to maintain a stable stand:-(
> 
> Thanks,
> -- Marco
> 
> -- 
> 
> -- 
> 
> 
> -------------------------------------------------------
> This SF.NET email is sponsored by:
> SourceForge Enterprise Edition + IBM + LinuxWorld = Something 2 See!
> http://www.vasoftware.com
> _______________________________________________
> Acpi-devel mailing list
> Acpi-devel-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.org
> https://lists.sourceforge.net/lists/listinfo/acpi-devel




-------------------------------------------------------
This SF.NET email is sponsored by:
SourceForge Enterprise Edition + IBM + LinuxWorld = Something 2 See!
http://www.vasoftware.com

[TRIVIAL] Re: [PATCH] ACPI update (20030122)

[Sergio Visinoni]

[-- Attachment #1: Type: text/plain, Size: 349 bytes --]

* Grover, Andrew (andrew.grover-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org) wrote:
> (2.4) S4BIOS support (Ducrot Bruno)

Attached is the missing s4bios portion (the
acpi_enter_sleep_state_s4bios function ) + fixes to make it
build correctly on a 2.4.20 tree.

It should apply correctly on 2.4.21-pre3 as well (not tested).

Greetings,
Sergio Visinoni

[-- Attachment #2: acpi-20030122-missing-s4bios.patch --]
[-- Type: text/plain, Size: 2682 bytes --]

--- linux-2.4.20/include/acpi/acpixf.h.s4bios~	2003-01-24 02:34:04.000000000 +0100
+++ linux-2.4.20/include/acpi/acpixf.h	2003-01-24 02:37:01.000000000 +0100
@@ -380,6 +380,10 @@
 	u8                              sleep_state);
 
 acpi_status
+acpi_enter_sleep_state_s4bios (
+	void);
+
+acpi_status
 acpi_leave_sleep_state (
 	u8                              sleep_state);
 
--- linux-2.4.20/drivers/acpi/hardware/hwsleep.c.s4bios~	2003-01-24 02:35:41.000000000 +0100
+++ linux-2.4.20/drivers/acpi/hardware/hwsleep.c	2003-01-24 02:35:53.000000000 +0100
@@ -316,6 +316,54 @@
 	return_ACPI_STATUS (AE_OK);
 }
 
+
+/******************************************************************************
+ *
+ * FUNCTION:    acpi_enter_sleep_state_s4bios
+ *
+ * PARAMETERS:  None
+ *
+ * RETURN:      Status
+ *
+ * DESCRIPTION: Perform a s4 bios request.
+ *              THIS FUNCTION MUST BE CALLED WITH INTERRUPTS DISABLED
+ *
+ ******************************************************************************/
+
+acpi_status
+acpi_enter_sleep_state_s4bios (
+	void)
+{
+	u32                     in_value;
+	acpi_status             status;
+
+
+	ACPI_FUNCTION_TRACE ("Acpi_enter_sleep_state_s4bios");
+
+
+	acpi_set_register (ACPI_BITREG_WAKE_STATUS, 1, ACPI_MTX_LOCK);
+	acpi_hw_clear_acpi_status();
+
+	acpi_hw_disable_non_wakeup_gpes();
+
+	ACPI_FLUSH_CPU_CACHE();
+
+	status = acpi_os_write_port (acpi_gbl_FADT->smi_cmd, (acpi_integer) acpi_gbl_FADT->S4bios_req, 8);
+
+	do {
+		acpi_os_stall(1000);
+		status = acpi_get_register (ACPI_BITREG_WAKE_STATUS, &in_value, ACPI_MTX_LOCK);
+		if (ACPI_FAILURE (status)) {
+			return_ACPI_STATUS (status);
+		}
+	} while (!in_value);
+
+	printk(KERN_DEBUG "acpi: s4bios coming to live\n");
+
+	return_ACPI_STATUS (AE_OK);
+}
+
+
 /******************************************************************************
  *
  * FUNCTION:    acpi_leave_sleep_state
@@ -339,6 +387,8 @@
 
 	ACPI_FUNCTION_TRACE ("acpi_leave_sleep_state");
 
+	/* Be sure to have BM arbitration */
+	status = acpi_set_register (ACPI_BITREG_ARB_DISABLE, 0, ACPI_MTX_LOCK);
 
 	/* Ensure enter_sleep_state_prep -> enter_sleep_state ordering */
 
@@ -371,8 +421,5 @@
 		return_ACPI_STATUS (status);
 	}
 
-	/* Disable BM arbitration */
-	status = acpi_set_register (ACPI_BITREG_ARB_DISABLE, 0, ACPI_MTX_LOCK);
-
 	return_ACPI_STATUS (status);
 }
--- linux-2.4.20/drivers/acpi/include/acpixf.h.s4bios~	2001-10-24 23:06:22.000000000 +0200
+++ linux-2.4.20/drivers/acpi/include/acpixf.h	2003-01-24 02:38:17.000000000 +0100
@@ -320,6 +320,10 @@
 	u8 sleep_state);
 
 acpi_status
+acpi_enter_sleep_state_s4bios (
+	void);
+
+acpi_status
 acpi_leave_sleep_state (
 	u8 sleep_state);
 

Re: [ACPI] [TRIVIAL] Re: [PATCH] ACPI update (20030122)

[Ducrot Bruno]

On Fri, Jan 24, 2003 at 03:22:22AM +0100, Sergio Visinoni wrote:
> * Grover, Andrew (andrew.grover-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org) wrote:
> > (2.4) S4BIOS support (Ducrot Bruno)
> 
> Attached is the missing s4bios portion (the
> acpi_enter_sleep_state_s4bios function ) + fixes to make it
> build correctly on a 2.4.20 tree.
> 
> It should apply correctly on 2.4.21-pre3 as well (not tested).
> 

It should be OK (I think..)

Appart the following chunk:

--- linux-2.4.20/drivers/acpi/hardware/hwsleep.c.s4bios~	2003-01-24 02:35:41.000000000 +0100
+++ linux-2.4.20/drivers/acpi/hardware/hwsleep.c	2003-01-24 02:35:53.000000000 +0100
@@ -339,6 +387,8 @@
 
 	ACPI_FUNCTION_TRACE ("acpi_leave_sleep_state");
 
+	/* Be sure to have BM arbitration */
+	status = acpi_set_register (ACPI_BITREG_ARB_DISABLE, 0, ACPI_MTX_LOCK);
 
 	/* Ensure enter_sleep_state_prep -> enter_sleep_state ordering */
 
@@ -371,8 +421,5 @@
 		return_ACPI_STATUS (status);
 	}
 
-	/* Disable BM arbitration */
-	status = acpi_set_register (ACPI_BITREG_ARB_DISABLE, 0, ACPI_MTX_LOCK);
-
 	return_ACPI_STATUS (status);
 }


which is a bug fix for a different purpose.

We can not trust the bios when we have to enter the _WAK method.
Those we re-enable the BM arbitration before (the comment Disable
is actually a mistake).

Cheers,


-- 
Ducrot Bruno

--  Which is worse:  ignorance or apathy?
--  Don't know.  Don't care.

Re: [TRIVIAL] Re: [PATCH] ACPI update (20030122)

[Jochen Reinwand]

On Friday 24 January 2003 16:51, Ducrot Bruno wrote:
> On Fri, Jan 24, 2003 at 03:22:22AM +0100, Sergio Visinoni wrote:
> > * Grover, Andrew (andrew.grover-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org) wrote:
> > > (2.4) S4BIOS support (Ducrot Bruno)
> >
> > Attached is the missing s4bios portion (the
> > acpi_enter_sleep_state_s4bios function ) + fixes to make it
> > build correctly on a 2.4.20 tree.
> >
> > It should apply correctly on 2.4.21-pre3 as well (not tested).
>
> It should be OK (I think..)

I would really like to test it on a 2.4.20, but the new acpi patch does not 
apply cleanly on a vanilla 2.4.20. Is there a special backport patch 
available or do I have to figure it out on my own?

regards,
Jochen


-------------------------------------------------------
This SF.NET email is sponsored by:
SourceForge Enterprise Edition + IBM + LinuxWorld = Something 2 See!
http://www.vasoftware.com

Re: [TRIVIAL] Re: [PATCH] ACPI update (20030122)

[Ducrot Bruno]

On Fri, Jan 24, 2003 at 07:21:35PM +0100, Jochen Reinwand wrote:
> On Friday 24 January 2003 16:51, Ducrot Bruno wrote:
> > On Fri, Jan 24, 2003 at 03:22:22AM +0100, Sergio Visinoni wrote:
> > > * Grover, Andrew (andrew.grover-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org) wrote:
> > > > (2.4) S4BIOS support (Ducrot Bruno)
> > >
> > > Attached is the missing s4bios portion (the
> > > acpi_enter_sleep_state_s4bios function ) + fixes to make it
> > > build correctly on a 2.4.20 tree.
> > >
> > > It should apply correctly on 2.4.21-pre3 as well (not tested).
> >
> > It should be OK (I think..)
> 
> I would really like to test it on a 2.4.20, but the new acpi patch does not 
> apply cleanly on a vanilla 2.4.20. Is there a special backport patch 
> available or do I have to figure it out on my own?
> 

I am sorry, but I can not provide a patch for newest acpi patch for
vanilla 2.4.20 due to lack of time.

Cheers,

-- 
Ducrot Bruno

--  Which is worse:  ignorance or apathy?
--  Don't know.  Don't care.


-------------------------------------------------------
This SF.NET email is sponsored by:
SourceForge Enterprise Edition + IBM + LinuxWorld = Something 2 See!
http://www.vasoftware.com

RE: [PATCH] ACPI update (20030122)

[Grover, Andrew]

> From: Dave Jones [mailto:davej-rdkfGonbjUTCLXcRTR1eJlpr/1R2p/CL@public.gmane.org] 
> I've noticed that with .59 some of my boxes no longer have
> functioning NICs unless I boot with acpi=off. Packets get
> transmitted, but nothing ever gets received.
> Seen this with a 3c509, an 8139, and an e100.
> 
> Known bug? Fixed in this patch ?

Known bug. Not fixed in this patch. Are you sure this is new in recent
kernels?

Regards -- Andy

Re: [PATCH] ACPI update (20030122)

[Dave Jones]

On Thu, Jan 23, 2003 at 04:09:07PM -0800, Grover, Andrew wrote:

 > > Known bug? Fixed in this patch ?
 > Known bug. Not fixed in this patch. Are you sure this is new in recent
 > kernels?

Only seems to started biting me this last week for some reason.

		Dave

-- 
| Dave Jones.        http://www.codemonkey.org.uk
| SuSE Labs

RE: [PATCH] ACPI update (20030122)

[Grover, Andrew]

> From: Marco Walther [mailto:Marco.Walther-xsfywfwIY+M@public.gmane.org] 
> A quick question,
> 
> is there a reason why iasl (iasl-linux-20021205) would need glibc-2.3?
> I can play with different kernels on my box but if I've to upgrade my
> libc as well, it's getting harder to maintain a stable stand:-(

That just happens to be what my system uses (RH 8). If it's not too hard
to link against something else, just tell me how, and I'll do it. :-)

Failing that, you should be able to D/L the source (it's in the
acpica-unix package via the intel.com link) and compile it yourself.

Regards -- Andy


-------------------------------------------------------
This SF.NET email is sponsored by:
SourceForge Enterprise Edition + IBM + LinuxWorld = Something 2 See!
http://www.vasoftware.com

RE: [PATCH] ACPI update (20030122)

[Marco Walther]

>>>>> "AG" == Andrew Grover <Grover> writes:

>> From: Marco Walther [mailto:Marco.Walther-xsfywfwIY+M@public.gmane.org] 
>> A quick question,
>> 
>> is there a reason why iasl (iasl-linux-20021205) would need glibc-2.3?
>> I can play with different kernels on my box but if I've to upgrade my
>> libc as well, it's getting harder to maintain a stable stand:-(

AG> That just happens to be what my system uses (RH 8). If it's not too hard
AG> to link against something else, just tell me how, and I'll do it. :-)

AG> Failing that, you should be able to D/L the source (it's in the
AG> acpica-unix package via the intel.com link) and compile it yourself.

Ok, this worked for me;-)

Thanks,
-- Marco


AG> Regards -- Andy

-- 


-------------------------------------------------------
This SF.NET email is sponsored by:
SourceForge Enterprise Edition + IBM + LinuxWorld = Something 2 See!
http://www.vasoftware.com

Re: [PATCH] ACPI update (20030122)

[Marc-Christian Petersen]

Hi Dave,

>> I've noticed that with .59 some of my boxes no longer have
>> functioning NICs unless I boot with acpi=off. Packets get
>> transmitted, but nothing ever gets received.
>> Seen this with a 3c509, an 8139, and an e100.
>> Known bug? Fixed in this patch ?

>> Known bug. Not fixed in this patch. Are you sure this is new in recent
>> kernels?
> Only seems to started biting me this last week for some reason.

hehe, I had exactly the same issues starting with 2.4.18 + acpi stuff.

I've noticed that almost every 2nd release broke network cards.
Already reported sometimes.

ciao, Marc


-------------------------------------------------------
This SF.NET email is sponsored by:
SourceForge Enterprise Edition + IBM + LinuxWorld = Something 2 See!
http://www.vasoftware.com