[Fwd: Re: ASL help with latest acpi and compaq presario 2800t]

[Fwd: Re: ASL help with latest acpi and compaq presario 2800t]

[Darren Benham]

oops... for got to include the list...

-------- Original Message --------
Subject: Re: [ACPI] ASL help with latest acpi and compaq presario 2800t
From: "Darren Benham" <dbenham-FG1iuTdj8bisTnJN9+BGXg@public.gmane.org>
Date: Mon, February 10, 2003 12:01 pm
To: <ducrot-kk6yZipjEM5g9hUCZPvPmw@public.gmane.org>


Ducrot Bruno said:
> On Sun, Feb 09, 2003 at 05:44:48PM -0700, Kevin Fenzi wrote:
>
> Try:
>
> --- dsdt.asl	2003/02/10 13:49:25	1.1
> +++ dsdt.asl	2003/02/10 13:54:53
> @@ -1916,7 +1915,7 @@
>                         }
>                         Method(_OFF) {
>                             If(LOr(\_SB_.C020, LEqual(0x0, 0x1))) {
> -                                Return(0x0)
> +                                /* Return(0x0) */
>                             }
>                             C132(0x3, 0x0)
>                         }
> @@ -2202,7 +2201,7 @@
>                         }
>                         Method(_OFF) {
>                             If(LOr(\_SB_.C023, LEqual(0x0, 0x1))) {
> -                                Return(0x0)
> +                                /* Return(0x0) */
>                             }
>                             C132(0x1, 0x0)
>                         }

I question these two changes.  It really does seem as if the "return"
statements are to prevent the rest of the code from being executed in
certain cases.  Wouldn't it be wiser to leave them in and either 1)
ignore the warning or 2) put "return statements at the end of the
block?--  Darren


-- 
Darren




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

Re: [Fwd: Re: ASL help with latest acpi and compaq presario 2800t]

[Ducrot Bruno]

On Mon, Feb 10, 2003 at 12:04:56PM -0800, Darren Benham wrote:
> oops... for got to include the list...
> 
> -------- Original Message --------
> Subject: Re: [ACPI] ASL help with latest acpi and compaq presario 2800t
> From: "Darren Benham" <dbenham-FG1iuTdj8bisTnJN9+BGXg@public.gmane.org>
> Date: Mon, February 10, 2003 12:01 pm
> To: <ducrot-kk6yZipjEM5g9hUCZPvPmw@public.gmane.org>
> 
> 
> Ducrot Bruno said:
> > On Sun, Feb 09, 2003 at 05:44:48PM -0700, Kevin Fenzi wrote:
> >
> > Try:
> >
> > --- dsdt.asl	2003/02/10 13:49:25	1.1
> > +++ dsdt.asl	2003/02/10 13:54:53
> > @@ -1916,7 +1915,7 @@
> >                         }
> >                         Method(_OFF) {
> >                             If(LOr(\_SB_.C020, LEqual(0x0, 0x1))) {
> > -                                Return(0x0)
> > +                                /* Return(0x0) */
> >                             }
> >                             C132(0x3, 0x0)
> >                         }
> > @@ -2202,7 +2201,7 @@
> >                         }
> >                         Method(_OFF) {
> >                             If(LOr(\_SB_.C023, LEqual(0x0, 0x1))) {
> > -                                Return(0x0)
> > +                                /* Return(0x0) */
> >                             }
> >                             C132(0x1, 0x0)
> >                         }
> 
> I question these two changes.  It really does seem as if the "return"
> statements are to prevent the rest of the code from being executed in
> certain cases.

Exact.  I am completly wrong here.

You should consider to do something like that instead:


                         Method(_OFF) {
                             If(LOr(\_SB_.C020, LEqual(0x0, 0x1))) {
                                Store ("_OFF: already off ?", Debug)
                            } Else {
                                C132(0x3, 0x0)
                            }
                        }


> Wouldn't it be wiser to leave them in and either 1)
> ignore the warning or 2) put "return statements at the end of the
> block?--  Darren
> 

point 2) can not work here.
An _OFF method must return nothing (see acpi specs 2.0b, 7.1.2)
OSPM will call _OFF, then call _STA in order to check if
the power ressource is really off.

Note also that the debug statement is only here to check if
"LOr(\_SB_.C020, LEqual(0x0, 0x1)" is really for that.  I am
really not sure here.  Hard to follow this ASL since
the names are not, well, the best that can be choosen..


-- 
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: [Fwd: Re: ASL help with latest acpi and compaq presario 2800t]

[dbenham-FG1iuTdj8bisTnJN9+BGXg]

On Tue, Feb 11, 2003 at 02:37:11PM +0100, Ducrot Bruno wrote:
> Exact.  I am completly wrong here.
> 
> You should consider to do something like that instead:
> 
> 
>                          Method(_OFF) {
>                              If(LOr(\_SB_.C020, LEqual(0x0, 0x1))) {
>                                 Store ("_OFF: already off ?", Debug)
>                             } Else {
>                                 C132(0x3, 0x0)
>                             }
>                         }
> 
> 

What about using NOOP instead of the Store?  Would that work also?

> point 2) can not work here.
> An _OFF method must return nothing (see acpi specs 2.0b, 7.1.2)
> OSPM will call _OFF, then call _STA in order to check if
> the power ressource is really off.
> 
> Note also that the debug statement is only here to check if
> "LOr(\_SB_.C020, LEqual(0x0, 0x1)" is really for that.  I am
> really not sure here.  Hard to follow this ASL since
> the names are not, well, the best that can be choosen..

I'm just starting to read the *gack* specs.  What did 1.0b say about
_OFF returning a value?  I'd lay odds the ASM is coded to v1 and not
v2...


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

Re: [Fwd: Re: ASL help with latest acpi and compaq presario 2800t]

[Ducrot Bruno]

On Tue, Feb 11, 2003 at 09:10:05AM -0800, dbenham-FG1iuTdj8bisTnJN9+BGXg@public.gmane.org wrote:
> On Tue, Feb 11, 2003 at 02:37:11PM +0100, Ducrot Bruno wrote:
> > Exact.  I am completly wrong here.
> > 
> > You should consider to do something like that instead:
> > 
> > 
> >                          Method(_OFF) {
> >                              If(LOr(\_SB_.C020, LEqual(0x0, 0x1))) {
> >                                 Store ("_OFF: already off ?", Debug)
> >                             } Else {
> >                                 C132(0x3, 0x0)
> >                             }
> >                         }
> > 
> > 
> 
> What about using NOOP instead of the Store?  Would that work also?

Store is here only to debug things.  You can kill it after testing.
You can also reverse the logic in the If().

> 
> > point 2) can not work here.
> > An _OFF method must return nothing (see acpi specs 2.0b, 7.1.2)
> > OSPM will call _OFF, then call _STA in order to check if
> > the power ressource is really off.
> > 
> > Note also that the debug statement is only here to check if
> > "LOr(\_SB_.C020, LEqual(0x0, 0x1)" is really for that.  I am
> > really not sure here.  Hard to follow this ASL since
> > the names are not, well, the best that can be choosen..
> 
> I'm just starting to read the *gack* specs.  What did 1.0b say about
> _OFF returning a value?  I'd lay odds the ASM is coded to v1 and not
> v2...

1.0b say the same thing (1.0b 7.4.3).

-- 
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