RE: _OSI boot hang

RE: _OSI boot hang

[Moore, Robert]

Not according to the MS white paper on OSI.  

http://www.microsoft.com/whdc/hwdev/tech/onnow/_OSI-method.mspx


Yes, \_OS used strings like "Microsoft Windows NT", but apparently not
OSI

Bob


-----Original Message-----
From: Nate Lawson [mailto:nate-Y6VGUYTwhu0@public.gmane.org] 
Sent: Wednesday, March 17, 2004 2:36 PM
To: Brown, Len
Cc: Moore, Robert; ACPI Developers
Subject: Re: [ACPI] _OSI boot hang

On Sat, 13 Mar 2004, Len Brown wrote:
> _OSI is in the ACPI patch on kernel.org and bkbits right now.
>
> However, I find that on my Toshiba Satellite Pro, it causes boot to
hang
> at Executing all Device _STA and_INI methods.
>
> >From ACPICA, it comes this way:
> drivers/acpi/utilities/utglobal.c
> const char
> *acpi_gbl_valid_osi_strings[ACPI_NUM_OSI_STRINGS] = {
>                        "Linux",
>                        "Windows 2000",
>                        "Windows 2001",
>                        "Windows 2001.1"};

That's not the right string for Win2k.  It's:
  "Microsoft Windows NT"

Since this is the most popular string, not emulating it is probably
causing some people problems.

-Nate


-------------------------------------------------------
This SF.Net email is sponsored by: IBM Linux Tutorials
Free Linux tutorial presented by Daniel Robbins, President and CEO of
GenToo technologies. Learn everything from fundamentals to system
administration.http://ads.osdn.com/?ad_id\x1470&alloc_id638&op=click

RE: _OSI boot hang

[Nate Lawson]

You're correct.  On a second check, I only found Windows 2001 and above in
terms of ASL checking for OSI strings.

-Nate

On Wed, 17 Mar 2004, Moore, Robert wrote:
> Not according to the MS white paper on OSI.
>
> http://www.microsoft.com/whdc/hwdev/tech/onnow/_OSI-method.mspx
>
>
> Yes, \_OS used strings like "Microsoft Windows NT", but apparently not
> OSI
>
> Bob
>
>
> -----Original Message-----
> From: Nate Lawson [mailto:nate-Y6VGUYTwhu0@public.gmane.org]
> Sent: Wednesday, March 17, 2004 2:36 PM
> To: Brown, Len
> Cc: Moore, Robert; ACPI Developers
> Subject: Re: [ACPI] _OSI boot hang
>
> On Sat, 13 Mar 2004, Len Brown wrote:
> > _OSI is in the ACPI patch on kernel.org and bkbits right now.
> >
> > However, I find that on my Toshiba Satellite Pro, it causes boot to
> hang
> > at Executing all Device _STA and_INI methods.
> >
> > >From ACPICA, it comes this way:
> > drivers/acpi/utilities/utglobal.c
> > const char
> > *acpi_gbl_valid_osi_strings[ACPI_NUM_OSI_STRINGS] = {
> >                        "Linux",
> >                        "Windows 2000",
> >                        "Windows 2001",
> >                        "Windows 2001.1"};
>
> That's not the right string for Win2k.  It's:
>   "Microsoft Windows NT"
>
> Since this is the most popular string, not emulating it is probably
> causing some people problems.
>
> -Nate
>
>
> -------------------------------------------------------
> This SF.Net email is sponsored by: IBM Linux Tutorials
> Free Linux tutorial presented by Daniel Robbins, President and CEO of
> GenToo technologies. Learn everything from fundamentals to system
> _______________________________________________
> 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: IBM Linux Tutorials
Free Linux tutorial presented by Daniel Robbins, President and CEO of
GenToo technologies. Learn everything from fundamentals to system
administration.http://ads.osdn.com/?ad_id=1470&alloc_id=3638&op=click

RE: _OSI (Re: Re: cvs commit: src/sys/contrib/dev/acpica acfreebsd.h)

[Yu, Luming]

> 
> On Wed, 11 Mar 2004, Len Brown wrote:
> > On Wed, 2004-03-10 at 21:26, Nate Lawson wrote:
> > > Bob, any word on _OSI implementation?
> >
> > Bob sent me an initial implementation for review today;-)
> >
> > I'm uncomfortable, however, at claiming to be compatible 
> with Windows,
> > since I don't really have a definition of exactly what that means...
> >
> > -Len
> 
> Doesn't matter to me.  I plan on claiming to be everything from Win 2k
> onward and sort out the rest on the mailing list.  :)
> 
> We already fixed some problems by claiming to be Win 2k for _OS.
> 

Cool! what kind of problems?  :)

--Luming


-------------------------------------------------------
This SF.Net email is sponsored by: IBM Linux Tutorials
Free Linux tutorial presented by Daniel Robbins, President and CEO of
GenToo technologies. Learn everything from fundamentals to system
administration.http://ads.osdn.com/?ad_id\x1470&alloc_id638&op=click

RE: _OSI (Re: Re: cvs commit: src/sys/contrib/dev/acpica acfreebsd.h)

[Nate Lawson]

On Thu, 11 Mar 2004, Yu, Luming wrote:
> > On Wed, 11 Mar 2004, Len Brown wrote:
> > > On Wed, 2004-03-10 at 21:26, Nate Lawson wrote:
> > > > Bob, any word on _OSI implementation?
> > >
> > > Bob sent me an initial implementation for review today;-)
> > >
> > > I'm uncomfortable, however, at claiming to be compatible
> > with Windows,
> > > since I don't really have a definition of exactly what that means...
> > >
> > > -Len
> >
> > Doesn't matter to me.  I plan on claiming to be everything from Win 2k
> > onward and sort out the rest on the mailing list.  :)
> >
> > We already fixed some problems by claiming to be Win 2k for _OS.
> >
>
> Cool! what kind of problems?  :)

PCI routing.

-Nate


-------------------------------------------------------
This SF.Net email is sponsored by: IBM Linux Tutorials
Free Linux tutorial presented by Daniel Robbins, President and CEO of
GenToo technologies. Learn everything from fundamentals to system
administration.http://ads.osdn.com/?ad_id=1470&alloc_id=3638&op=click

RE: _OSI (Re: Re: cvs commit: src/sys/contrib/dev/acpica acfreebsd.h)

[Len Brown]

On Thu, 2004-03-11 at 23:00, Nate Lawson wrote:
> On Thu, 11 Mar 2004, Yu, Luming wrote:
> > > On Wed, 11 Mar 2004, Len Brown wrote:
> > > > On Wed, 2004-03-10 at 21:26, Nate Lawson wrote:
> > > > > Bob, any word on _OSI implementation?
> > > >
> > > > Bob sent me an initial implementation for review today;-)
> > > >
> > > > I'm uncomfortable, however, at claiming to be compatible
> > > with Windows,
> > > > since I don't really have a definition of exactly what that means...
> > > >
> > > > -Len
> > >
> > > Doesn't matter to me.  I plan on claiming to be everything from Win 2k
> > > onward and sort out the rest on the mailing list.  :)
> > >
> > > We already fixed some problems by claiming to be Win 2k for _OS.
> > >
> >
> > Cool! what kind of problems?  :)
> 
> PCI routing.
> 
> -Nate

You don't need to tempt me any more than that!;-)

While I don't like blindly claiming compatibility with something that is
neither specified nor visible, I *do* like the idea of causing the AML
to follow the *tested* path that it follows under Windows.

thanks,
-Len
 



-------------------------------------------------------
This SF.Net email is sponsored by: IBM Linux Tutorials
Free Linux tutorial presented by Daniel Robbins, President and CEO of
GenToo technologies. Learn everything from fundamentals to system
administration.http://ads.osdn.com/?ad_id=1470&alloc_id=3638&op=click

_OSI boot hang

[Len Brown]

_OSI is in the ACPI patch on kernel.org and bkbits right now.

However, I find that on my Toshiba Satellite Pro, it causes boot to hang
at Executing all Device _STA and_INI methods.

>From ACPICA, it comes this way:
drivers/acpi/utilities/utglobal.c
const char                         
*acpi_gbl_valid_osi_strings[ACPI_NUM_OSI_STRINGS] = {
                       "Linux",
                       "Windows 2000",
                       "Windows 2001",
                       "Windows 2001.1"};


I find that it is the match on "Windows 2001" that hangs my laptop...

clues?

-Len

ACPI: Subsystem revision 20040311
 tbxface-0117 [03] acpi_load_tables      : ACPI Tables successfully
acquired
Parsing all Control
Methods:...................................................................................................
Table [DSDT](id F005) - 682 Objects with 72 Devices 183 Methods 6
Regions
Parsing all Control Methods:.....................
Table [SSDT](id F003) - 46 Objects with 6 Devices 21 Methods 0 Regions
ACPI Namespace successfully loaded at root c04b717c
ACPI: IRQ9 SCI: Level Trigger.
evxfevnt-0093 [04] acpi_enable           : Transition to ACPI mode
successful
evgpeblk-0747 [06] ev_create_gpe_block   : GPE 00 to 31 [_GPE] 4 regs at
000000000000D828 on int 9
Completing Region/Field/Buffer/Package
initialization:......................................................
Initialized 6/6 Regions 0/0 Fields 18/18 Buffers 30/30 Packages (736
nodes)
Executing all Device _STA and_INI
methods:..................................................



-------------------------------------------------------
This SF.Net email is sponsored by: IBM Linux Tutorials
Free Linux tutorial presented by Daniel Robbins, President and CEO of
GenToo technologies. Learn everything from fundamentals to system
administration.http://ads.osdn.com/?ad_id=1470&alloc_id=3638&op=click

Re: _OSI boot hang

[Len Brown]

Hmmm, now my Toshiba is not hanging.  Maybe the hang wasn't related to
_OSI afterall?

In any case, if you experience issues with the latest
"acpi_osi=" or "acpi_serialize" may be helpful -- do speak up if you
find they're necessary.

thanks,
-Len
----
kernel-parameters.txt:

        acpi_osi=       [HW,ACPI] empty param disables _OSI

        acpi_serialize  [HW,ACPI] force serialization of AML methods




-------------------------------------------------------
This SF.Net email is sponsored by: IBM Linux Tutorials
Free Linux tutorial presented by Daniel Robbins, President and CEO of
GenToo technologies. Learn everything from fundamentals to system
administration.http://ads.osdn.com/?ad_id=1470&alloc_id=3638&op=click

Re: _OSI boot hang

[Nate Lawson]

On Sat, 13 Mar 2004, Len Brown wrote:
> _OSI is in the ACPI patch on kernel.org and bkbits right now.
>
> However, I find that on my Toshiba Satellite Pro, it causes boot to hang
> at Executing all Device _STA and_INI methods.
>
> >From ACPICA, it comes this way:
> drivers/acpi/utilities/utglobal.c
> const char
> *acpi_gbl_valid_osi_strings[ACPI_NUM_OSI_STRINGS] = {
>                        "Linux",
>                        "Windows 2000",
>                        "Windows 2001",
>                        "Windows 2001.1"};

That's not the right string for Win2k.  It's:
  "Microsoft Windows NT"

Since this is the most popular string, not emulating it is probably
causing some people problems.

-Nate


-------------------------------------------------------
This SF.Net email is sponsored by: IBM Linux Tutorials
Free Linux tutorial presented by Daniel Robbins, President and CEO of
GenToo technologies. Learn everything from fundamentals to system
administration.http://ads.osdn.com/?ad_id=1470&alloc_id=3638&op=click