Re: [Qemu-devel] [PATCH 3/3] Get system state configuration from QEMU and patch DSDT with it.

[Gleb Natapov]

On Sun, May 20, 2012 at 05:34:56PM +0300, Avi Kivity wrote:
> On 05/20/2012 04:57 PM, Gleb Natapov wrote:
> > On Sun, May 20, 2012 at 04:39:01PM +0300, Avi Kivity wrote:
> > > On 05/20/2012 03:59 PM, Gleb Natapov wrote:
> > > > > > > > > 
> > > > > > > > > Do we actually have to patch the DSDT?  Or can _S3 etc be 
> > > > > > > > > made into
> > > > > > > > > functions instead? (and talk to the bios, or even to fwcfg 
> > > > > > > > > directly?)
> > > > > > > > > 
> > > > > > > > We better not talk to fwcfg after OSPM is started since this is 
> > > > > > > > firmware
> > > > > > > > confing interface.
> > > > > > > 
> > > > > > > Why not?  The OS isn't going to talk to it, so we can have a 
> > > > > > > driver in ACPI.
> > > > > > > 
> > > > > > The OS is going to talk to it since the OS is the one who interprets
> > > > > > AML. 
> > > > > 
> > > > > I meant, not directly.  So the driver in ACPI has exclusive access.
> > > > > 
> > > > What's the difference?
> > > 
> > > ACPI is firmware, not OS.
> > AML is a data provided by firmware. AML's runtime is different from 
> > firmware's.
> 
> It's still firmware.
> 
We have to agree to disagree here :) It's just a data for OS to use as
far as I am concern.

> > > > > 
> > > > > It's an alternative to patching AML.  Sure it takes some effort to 
> > > > > write
> > > > > the driver, but afterwards we can modify the guest behaviour more
> > > > > easily.  One possible client is -M old, so you can revert to previous
> > > > > behaviour depending on fwcfg data.
> > > > -M old is easy to support with the current patch. You just set new
> > > > properties to compatibility values. The code is written with this in
> > > > mind. And this is not an alternative to patching AML as I am trying to
> > > > explain to you below. You can eliminate patching of s4 value, but that's
> > > > it, you still need to patch out _S3/_S4 names.
> > > 
> > > What about
> > > 
> > >   If (Fcfg(...)) {
> > >         Method()...
> > >   }
> > > 
> > > ?
> > syntax error, unexpected PARSEOP_IF
> 
> Unfortunately the ACPI spec forbids this construct, so either patching
> or double complication is necessary.
> 
It's not double if we will take all possible combinations into account.

> > > 
> > > (i.e.. define the method conditionally at runtime)
> > > 
> > > >  
> > > > > 
> > > > > (we don't need a driver in AML to avoid patching, we can have AML talk
> > > > > to the bios and the bios drive fwcfg; but I think we'll find uses for 
> > > > > a
> > > > > driver).
> > > > I am not sure what you mean. AML can't talk to the bios. It can read
> > > > values that bios put somewhere. 
> > > 
> > > That's what I meant - communicate through memory.
> > > 
> > What's the benefit? The patching is still needed. You need to pass
> > address of OperationRegion() to AML. You can do it either by patching or
> > by creating OperationRegion() code dynamically.
> 
> Or it can be a fixed address in low memory, or a scratch register in
> hardware.
> 
Both will work (fixed addresses are better be avoided and who needs
another PV device), but I do not see how either of them is better then
patching. What is your concern?

> >
> > > > I do not see advantage of this method
> > > > and it requires patching still.
> > > 
> > > For the existence of the names?  Yes, if we can't avoid it it's a
> > > problem.  But if we can avoid patching, we should.
> > > 
> > If we can, we should, but we can't as far as I see. The patching was here 
> > long before
> > this patch.
> 
> I agree we probably can't.
> 
> -- 
> error compiling committee.c: too many arguments to function

--
                        Gleb.