General ACPI/Linux question

General ACPI/Linux question

[Constantinos Antoniou]

Hello,

I have a laptop (Compaq Presarion 1700T) and would really really like to
have some ACPI support (battery status is very important, suspend to ram
or disk - anything faster than reboot- would be very very useful and
welcome). 

I have not tried every possible trick, and in any case I am hoping to
get some things working this year... either with the introduction of
ACPI in RH8.1 or otherwise (ok, this is not flame-bait. Let me move on
:)) This is not a system/problem specific post, however.

My question is: One of the most usual things I see in the various
newsgroup posts is that DSDT tables are buggy, and people need to
disassemble them, edit them, replace them etc etc. However, windows
power management on the same machines works. So what is the situation? 

The way I (don't) understand it, the ACPI implementation expects some
things to be defined correctly in the DSDT table. However, some OSes
manage to work with whatever is there.

And to go one step further, if I change this DSDT thingie, will the ACPI
functionality in the other OSes still work? So, would an option be to be
more lenient (I understand that there are standards and people should
respect them...)

Thanks in advance and please be considerate in the replies - IANA ACPI
expert.

Costas


-- 
Constantinos Antoniou
Ph.D. Candidate
Massachusetts Institute of Technology
Intelligent Transportation Systems Program
77 Massachusetts Ave., NE20-208, Cambridge, MA 02139
(T) 617-252-1113 * (F) 617-252-1130 * (email) costas-3s7WtUTddSA@public.gmane.org



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

RE: General ACPI/Linux question

[Adachi, Kenichi]

Hi Constantinos,

> My question is: One of the most usual things I see in the 
> various newsgroup posts is that DSDT tables are buggy, and 
> people need to disassemble them, edit them, replace them etc 
> etc. However, windows power management on the same machines 
> works. So what is the situation? 
> 
> The way I (don't) understand it, the ACPI implementation 
> expects some things to be defined correctly in the DSDT 
> table. However, some OSes manage to work with whatever is there.

If your DSDT is totally broken, Windows won't work perfectly either. If
the problem is well known like "PC Vendor X's Model Y" has buggy PCI
Routing Table etc., Windows tries to save the machine by applying hack
logic. See %SystemRoot%\Inf\BiosInfo.inf for more details on this.
However, of course this approach is not ideal, and they(Microsoft) keep
telling BIOS writers to create the flawless BIOS. Only few aspects are
Windows specific, but as a whole they also want to have ACPI BIOS
perfectly honoring ACPI spec.

> And to go one step further, if I change this DSDT thingie, 
> will the ACPI functionality in the other OSes still work? So, 
> would an option be to be more lenient (I understand that 
> there are standards and people should respect them...)

You are not changing the ACPI stuffs stored in ROM, and the other OSes
still honor the ones in ROM (though they could have their own way to
replace ACPI tables, at least they don't understand the way Linux does),
changing DSDT specifically for Linux won't do any harm to other OSes (in
theory).


Thanks,
- Adachi, Kenichi 




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

Re: General ACPI/Linux question

[Philippe Clérié]

For what it's worth, I have ACPI working fine on a 1700T with 
Gentoo Linux and kernel 2.4.20-acpi-r9. As far as I can tell 
this kernel is vanilla + acpi 20021212. No modifications were 
needed. 

This is the first kernel I have tried that gave me some info on 
the battery. klaptop is working and shows a battery icon when 
on battery or an AC plug when on AC. And it pops up a message 
when battery power is low. But it does not show whether 
capacity is at 90, 50 or 20%. 

I have not checked any of the sleep modes since I don't use them 
at all so I can't comment on that.

For the moment I'm happy.

Regards,
Philippe

On Saturday, 25 January 2003 20:54, Constantinos Antoniou wrote:
> Hello,
>
> I have a laptop (Compaq Presarion 1700T) and would really
> really like to have some ACPI support (battery status is very
> important, suspend to ram or disk - anything faster than
> reboot- would be very very useful and welcome).
>
> I have not tried every possible trick, and in any case I am
> hoping to get some things working this year... either with
> the introduction of ACPI in RH8.1 or otherwise (ok, this is
> not flame-bait. Let me move on
>
> :)) This is not a system/problem specific post, however.
>
> My question is: One of the most usual things I see in the
> various newsgroup posts is that DSDT tables are buggy, and
> people need to disassemble them, edit them, replace them etc
> etc. However, windows power management on the same machines
> works. So what is the situation?
>
> The way I (don't) understand it, the ACPI implementation
> expects some things to be defined correctly in the DSDT
> table. However, some OSes manage to work with whatever is
> there.
>
> And to go one step further, if I change this DSDT thingie,
> will the ACPI functionality in the other OSes still work? So,
> would an option be to be more lenient (I understand that
> there are standards and people should respect them...)
>
> Thanks in advance and please be considerate in the replies -
> IANA ACPI expert.
>
> Costas

-- 
Best regards,
Philippe Clérié


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

Re: General ACPI/Linux question

[Constantinos Antoniou]

On Sun, 2003-01-26 at 08:23, Philippe Clérié wrote:
> For what it's worth, I have ACPI working fine on a 1700T with 
> Gentoo Linux and kernel 2.4.20-acpi-r9. As far as I can tell 
> this kernel is vanilla + acpi 20021212. No modifications were 
> needed. 
> 

This is very nice to hear. 


> This is the first kernel I have tried that gave me some info on 
> the battery. klaptop is working and shows a battery icon when 
> on battery or an AC plug when on AC. And it pops up a message 
> when battery power is low. But it does not show whether 
> capacity is at 90, 50 or 20%. 
> 

Having the ability to (at least) gracefully shutdown is a "good thing"
(TM)

> I have not checked any of the sleep modes since I don't use them 
> at all so I can't comment on that.
> 

Could you or anybody else perhaps provide some feedback on the attached
image? I have compiled a vanilla 2.4.20 with acpi-20021212-2.4.20.diff.
When I try to boot it freezes at the point of the attached (camera
captured - ~280K image) point.

http://mit.edu/costas/www/img_0092.jpg

Please let me know if any other information would help.

Thanks,

Costas




> For the moment I'm happy.
> 
> Regards,
> Philippe
> 
> On Saturday, 25 January 2003 20:54, Constantinos Antoniou wrote:
> > Hello,
> >
> > I have a laptop (Compaq Presarion 1700T) and would really
> > really like to have some ACPI support (battery status is very
> > important, suspend to ram or disk - anything faster than
> > reboot- would be very very useful and welcome).
> >
> > I have not tried every possible trick, and in any case I am
> > hoping to get some things working this year... either with
> > the introduction of ACPI in RH8.1 or otherwise (ok, this is
> > not flame-bait. Let me move on
> >
> > :)) This is not a system/problem specific post, however.
> >
> > My question is: One of the most usual things I see in the
> > various newsgroup posts is that DSDT tables are buggy, and
> > people need to disassemble them, edit them, replace them etc
> > etc. However, windows power management on the same machines
> > works. So what is the situation?
> >
> > The way I (don't) understand it, the ACPI implementation
> > expects some things to be defined correctly in the DSDT
> > table. However, some OSes manage to work with whatever is
> > there.
> >
> > And to go one step further, if I change this DSDT thingie,
> > will the ACPI functionality in the other OSes still work? So,
> > would an option be to be more lenient (I understand that
> > there are standards and people should respect them...)
> >
> > Thanks in advance and please be considerate in the replies -
> > IANA ACPI expert.
> >
> > Costas
-- 
Constantinos Antoniou
Ph.D. Candidate
Massachusetts Institute of Technology
Intelligent Transportation Systems Program
77 Massachusetts Ave., NE20-208, Cambridge, MA 02139
(T) 617-252-1113 * (F) 617-252-1130 * (email) costas-3s7WtUTddSA@public.gmane.org



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

Re: General ACPI/Linux question

[Philippe Clérié]

>
> Could you or anybody else perhaps provide some feedback on
> the attached image? I have compiled a vanilla 2.4.20 with
> acpi-20021212-2.4.20.diff. When I try to boot it freezes at
> the point of the attached (camera captured - ~280K image)
> point.
>
> http://mit.edu/costas/www/img_0092.jpg
>
> Please let me know if any other information would help.
>

I should have mentionned earlier that the only way the thing 
would boot is to add pci=noacpi to the kernel boot params. 
Sorry for the omission.

Regards,
Philippe


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

Re: General ACPI/Linux question

[Sérgio Monteiro Basto]

Compaq presarios 7xx have a problem with APIC (Advance Programmer
interrupt control ) that still to be resolved, so if you have a stable
computer, and work with acpi  you have to disable APIC from kernel.
I think this same problem reflexes in compaq presario 17xx. 

from more information you can read 
http://dude.noc.clara.net/~faye/compaq2701ea/
http://www.pps.jussieu.fr/~jch/software/presario/
http://wwwhome.cs.utwente.nl/~berensch/compaq.html
http://www.geocities.com/adrianturcu/presario724EA.html
and my own
that gone be move in next days and it out of date
http://codecs.home.sapo.pt/

hope saying something new for you

Sérgio Basto 
On Sun, 2003-01-26 at 15:55, Constantinos Antoniou wrote:
> On Sun, 2003-01-26 at 08:23, Philippe Clérié wrote:
> > For what it's worth, I have ACPI working fine on a 1700T with 
> > Gentoo Linux and kernel 2.4.20-acpi-r9. As far as I can tell 
> > this kernel is vanilla + acpi 20021212. No modifications were 
> > needed. 
> > 
> 
> This is very nice to hear. 
> 
> 
> > This is the first kernel I have tried that gave me some info on 
> > the battery. klaptop is working and shows a battery icon when 
> > on battery or an AC plug when on AC. And it pops up a message 
> > when battery power is low. But it does not show whether 
> > capacity is at 90, 50 or 20%. 
> > 
> 
> Having the ability to (at least) gracefully shutdown is a "good thing"
> (TM)
> 
> > I have not checked any of the sleep modes since I don't use them 
> > at all so I can't comment on that.
> > 
> 
> Could you or anybody else perhaps provide some feedback on the attached
> image? I have compiled a vanilla 2.4.20 with acpi-20021212-2.4.20.diff.
> When I try to boot it freezes at the point of the attached (camera
> captured - ~280K image) point.
> 
> http://mit.edu/costas/www/img_0092.jpg
> 
> Please let me know if any other information would help.
> 
> Thanks,
> 
> Costas
> 
> 
> 
> 
> > For the moment I'm happy.
> > 
> > Regards,
> > Philippe
> > 
> > On Saturday, 25 January 2003 20:54, Constantinos Antoniou wrote:
> > > Hello,
> > >
> > > I have a laptop (Compaq Presarion 1700T) and would really
> > > really like to have some ACPI support (battery status is very
> > > important, suspend to ram or disk - anything faster than
> > > reboot- would be very very useful and welcome).
> > >
> > > I have not tried every possible trick, and in any case I am
> > > hoping to get some things working this year... either with
> > > the introduction of ACPI in RH8.1 or otherwise (ok, this is
> > > not flame-bait. Let me move on
> > >
> > > :)) This is not a system/problem specific post, however.
> > >
> > > My question is: One of the most usual things I see in the
> > > various newsgroup posts is that DSDT tables are buggy, and
> > > people need to disassemble them, edit them, replace them etc
> > > etc. However, windows power management on the same machines
> > > works. So what is the situation?
> > >
> > > The way I (don't) understand it, the ACPI implementation
> > > expects some things to be defined correctly in the DSDT
> > > table. However, some OSes manage to work with whatever is
> > > there.
> > >
> > > And to go one step further, if I change this DSDT thingie,
> > > will the ACPI functionality in the other OSes still work? So,
> > > would an option be to be more lenient (I understand that
> > > there are standards and people should respect them...)
> > >
> > > Thanks in advance and please be considerate in the replies -
> > > IANA ACPI expert.
> > >
> > > Costas
> -- 
> Constantinos Antoniou
> Ph.D. Candidate
> Massachusetts Institute of Technology
> Intelligent Transportation Systems Program
> 77 Massachusetts Ave., NE20-208, Cambridge, MA 02139
> (T) 617-252-1113 * (F) 617-252-1130 * (email) costas-3s7WtUTddSA@public.gmane.org
> 
> 
> 
> -------------------------------------------------------
> 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

Re: General ACPI/Linux question

[Sérgio Monteiro Basto]

On Sun, 2003-01-26 at 17:37, Sérgio Monteiro Basto wrote:
> Compaq presarios 7xx have a problem with APIC (Advance Programmer
> interrupt control ) that still to be resolved, so if you have a stable

so if you want have a stable

> computer, and work with acpi  you have to disable APIC from kernel.
> I think this same problem reflexes in compaq presario 17xx. 
> 
> from more information you can read 
> http://dude.noc.clara.net/~faye/compaq2701ea/
> http://www.pps.jussieu.fr/~jch/software/presario/
> http://wwwhome.cs.utwente.nl/~berensch/compaq.html
> http://www.geocities.com/adrianturcu/presario724EA.html
> and my own
> that gone be move in next days and it out of date
> http://codecs.home.sapo.pt/
> 
> hope saying something new for you
> 
> Sérgio Basto 
> On Sun, 2003-01-26 at 15:55, Constantinos Antoniou wrote:
> > On Sun, 2003-01-26 at 08:23, Philippe Clérié wrote:
> > > For what it's worth, I have ACPI working fine on a 1700T with 
> > > Gentoo Linux and kernel 2.4.20-acpi-r9. As far as I can tell 
> > > this kernel is vanilla + acpi 20021212. No modifications were 
> > > needed. 
> > > 
> > 
> > This is very nice to hear. 
> > 
> > 
> > > This is the first kernel I have tried that gave me some info on 
> > > the battery. klaptop is working and shows a battery icon when 
> > > on battery or an AC plug when on AC. And it pops up a message 
> > > when battery power is low. But it does not show whether 
> > > capacity is at 90, 50 or 20%. 
> > > 
> > 
> > Having the ability to (at least) gracefully shutdown is a "good thing"
> > (TM)
> > 
> > > I have not checked any of the sleep modes since I don't use them 
> > > at all so I can't comment on that.
> > > 
> > 
> > Could you or anybody else perhaps provide some feedback on the attached
> > image? I have compiled a vanilla 2.4.20 with acpi-20021212-2.4.20.diff.
> > When I try to boot it freezes at the point of the attached (camera
> > captured - ~280K image) point.
> > 
> > http://mit.edu/costas/www/img_0092.jpg
> > 
> > Please let me know if any other information would help.
> > 
> > Thanks,
> > 
> > Costas
> > 
> > 
> > 
> > 
> > > For the moment I'm happy.
> > > 
> > > Regards,
> > > Philippe
> > > 
> > > On Saturday, 25 January 2003 20:54, Constantinos Antoniou wrote:
> > > > Hello,
> > > >
> > > > I have a laptop (Compaq Presarion 1700T) and would really
> > > > really like to have some ACPI support (battery status is very
> > > > important, suspend to ram or disk - anything faster than
> > > > reboot- would be very very useful and welcome).
> > > >
> > > > I have not tried every possible trick, and in any case I am
> > > > hoping to get some things working this year... either with
> > > > the introduction of ACPI in RH8.1 or otherwise (ok, this is
> > > > not flame-bait. Let me move on
> > > >
> > > > :)) This is not a system/problem specific post, however.
> > > >
> > > > My question is: One of the most usual things I see in the
> > > > various newsgroup posts is that DSDT tables are buggy, and
> > > > people need to disassemble them, edit them, replace them etc
> > > > etc. However, windows power management on the same machines
> > > > works. So what is the situation?
> > > >
> > > > The way I (don't) understand it, the ACPI implementation
> > > > expects some things to be defined correctly in the DSDT
> > > > table. However, some OSes manage to work with whatever is
> > > > there.
> > > >
> > > > And to go one step further, if I change this DSDT thingie,
> > > > will the ACPI functionality in the other OSes still work? So,
> > > > would an option be to be more lenient (I understand that
> > > > there are standards and people should respect them...)
> > > >
> > > > Thanks in advance and please be considerate in the replies -
> > > > IANA ACPI expert.
> > > >
> > > > Costas
> > -- 
> > Constantinos Antoniou
> > Ph.D. Candidate
> > Massachusetts Institute of Technology
> > Intelligent Transportation Systems Program
> > 77 Massachusetts Ave., NE20-208, Cambridge, MA 02139
> > (T) 617-252-1113 * (F) 617-252-1130 * (email) costas-3s7WtUTddSA@public.gmane.org
> > 
> > 
> > 
> > -------------------------------------------------------
> > 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
> _______________________________________________
> 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

RE: General ACPI/Linux question

[Grover, Andrew]

> From: Constantinos Antoniou [mailto:costas-FrUbXkNCsVf2fBVCVOL8/A@public.gmane.org] 
> The way I (don't) understand it, the ACPI implementation expects some
> things to be defined correctly in the DSDT table. However, some OSes
> manage to work with whatever is there.
> 
> And to go one step further, if I change this DSDT thingie, 
> will the ACPI
> functionality in the other OSes still work? So, would an 
> option be to be
> more lenient (I understand that there are standards and people should
> respect them...)

You can override your DSDT by changing the Linux ACPI code. This doesn't
change the DSDT in your BIOS, however, so booting another OS will just
result in using the BIOS one (which is probably what you wanted.)

Regards -- Andy


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