Re: dmidecode of HP Pavilion tx 1000

[Len Brown <lenb@kernel.org>]

On Tuesday 22 January 2008 16:24, Jakub Ambrozewicz wrote:
> 2008/1/19, Len Brown <lenb@kernel.org>:
> >
> > Thanks for the dmidecode output.
> 
> No problem. It's rather me who should thank for your work. So I do:
> Big Thanks. :)
> 
> > Do you notice any functional difference when booting with
> >
> > "acpi_osi=Linux"        (default for 2.6.22 and earlier)
> > vs.
> > "acpi_osi=!Linux"       (default for 2.6.23 and later)
> >
> > Also, please send me the output from acpidump.
> 
> I am really sorry that it took so long, but here they are.
> Hand made 2.6.23.14 (config attached). I have not noticed any
> differences  between using Linux and !Linux
> I still need to use nolapic because systems hangs on bootup while
> setting the system clock. I suspect it has something to do with
> interrupts, and tried irqpoll irqdebug acpi_irq_balance but it did not
> help a lot. Since then I haven't done much testing.
> One thing has changed since I last checked: the suspend to ram now
> works, but works both on "Linux" and "!Linux".
> 
> I am also attaching dmidecode again, as I have upgraded the bios to
> F.1C and I do not know if it changes anything.
> 

Thanks for the acpidump.
It shows that OSI(Linux) will have absolutely no effect
on the tx 1000 -- because the OSYS value that it effects
gets immediately over-written by the values associated
with Windows before it is accessed.

thanks,
-Len


                If (CondRefOf (\_OSI, Local0))
                {
                    If (\_OSI ("Linux"))
                    {
                        Store (0x03E8, OSYS)
                    }

                    If (\_OSI ("Windows 2001"))
                    {
                        Store (0x07D1, OSYS)
                    }