Re: [Qemu-devel] [PATCH 2/3] acpi_piix4: Add stub functions for CPU eject callback

[Avi Kivity]

On 01/16/2012 01:32 PM, Vasilis Liaskovitis wrote:
> On Sun, Jan 15, 2012 at 02:38:52PM +0200, Avi Kivity wrote:
> > On 01/13/2012 01:11 PM, Vasilis Liaskovitis wrote:
> > > Signed-off-by: Vasilis Liaskovitis <address@hidden>
> > > ---
> > >  hw/acpi_piix4.c |   15 +++++++++++++++
> > >  1 files changed, 15 insertions(+), 0 deletions(-)
> > >
> > > diff --git a/hw/acpi_piix4.c b/hw/acpi_piix4.c
> > > index d5743b6..8bf30dd 100644
> > > --- a/hw/acpi_piix4.c
> > > +++ b/hw/acpi_piix4.c
> > > @@ -37,6 +37,7 @@
> > >  
> > >  #define GPE_BASE 0xafe0
> > >  #define PROC_BASE 0xaf00
> > > +#define PROC_EJ_BASE 0xaf20
> > >
> > 
> > We're adding stuff to piix4 which was never there.  At a minimum this
> > needs to be documented.  Also needs to be -M pc-1.1 and later only.
>
> Where should this be documented? PCI/ACPI hotplug addresses are documented in
> docs/specs/acpi_pci_hotplug.txt 

A pleasant surprise

> but for CPU hotplug documentation (i.e.
> for the existing PROC_BASE) I don't see relevant documentation. I will
> create a docs/specs/acpi_cpu_hotplug.txt if that sounds reasonable.

I suggest renaming it to acpi_hotplug.txt, so it covers both cases.

> For pc-1.1, a new QEMUmachine type will be needed I assume. Should a check be
> made against the machine version in the piix4 code? any relevant examples? 
>

The standard practice is to set a property.  See for example
pc_machine_v0_14 in hw/pc_piix.c, it autosets properties for devices
(erroneously called "drivers" in the code).

btw, I notice the I/O ports are write only and don't remember their
state.  I can't think offhand if there's anything bad about it (in fact
not having state makes live migration more robust), but perhaps someone
else will.


-- 
error compiling committee.c: too many arguments to function