Re: [Qemu-devel] [PATCH v2 1/3] prep: qdev'ify Raven host bridge

[Anthony Liguori <anthony@codemonkey.ws>]

On 01/06/2012 06:06 PM, Andreas Färber wrote:
> Move initialization of vendor ID, etc. to PCIDeviceInfo.
> Add VMState.
>
> Signed-off-by: Andreas Färber<andreas.faerber@web.de>
> Cc: Hervé Poussineau<hpoussin@reactos.org>
> Cc: Michael S. Tsirkin<mst@redhat.com>
> Cc: Anthony Liguori<aliguori@us.ibm.com>
> ---
>   hw/prep_pci.c |   55 ++++++++++++++++++++++++++++++++++++++++++++++---------
>   1 files changed, 46 insertions(+), 9 deletions(-)
>
> diff --git a/hw/prep_pci.c b/hw/prep_pci.c
> index ea9fb69..741b273 100644
> --- a/hw/prep_pci.c
> +++ b/hw/prep_pci.c
> @@ -29,6 +29,10 @@
>
>   typedef PCIHostState PREPPCIState;
>
> +typedef struct RavenPCIState {
> +    PCIDevice dev;
> +} RavenPCIState;
> +
>   static inline uint32_t PPC_PCIIO_config(target_phys_addr_t addr)
>   {
>       int i;
> @@ -111,7 +115,6 @@ PCIBus *pci_prep_init(qemu_irq *pic,
>                         MemoryRegion *address_space_io)
>   {
>       PREPPCIState *s;
> -    PCIDevice *d;
>
>       s = g_malloc0(sizeof(PREPPCIState));
>       s->bus = pci_register_bus(NULL, "pci",
> @@ -133,16 +136,50 @@ PCIBus *pci_prep_init(qemu_irq *pic,
>       memory_region_init_io(&s->mmcfg,&PPC_PCIIO_ops, s, "pciio", 0x00400000);
>       memory_region_add_subregion(address_space_mem, 0x80800000,&s->mmcfg);
>
> -    /* PCI host bridge */
> -    d = pci_register_device(s->bus, "PREP Host Bridge - Motorola Raven",
> -                            sizeof(PCIDevice), 0, NULL, NULL);
> -    pci_config_set_vendor_id(d->config, PCI_VENDOR_ID_MOTOROLA);
> -    pci_config_set_device_id(d->config, PCI_DEVICE_ID_MOTOROLA_RAVEN);
> -    d->config[0x08] = 0x00; // revision
> -    pci_config_set_class(d->config, PCI_CLASS_BRIDGE_HOST);
> +    pci_create_simple(s->bus, 0, "raven");
> +
> +    return s->bus;
> +}
> +
> +static int raven_init(PCIDevice *d)
> +{
>       d->config[0x0C] = 0x08; // cache_line_size
>       d->config[0x0D] = 0x10; // latency_timer
>       d->config[0x34] = 0x00; // capabilities_pointer
>
> -    return s->bus;
> +    return 0;
>   }
> +
> +static const VMStateDescription vmstate_raven = {
> +    .name = "raven",
> +    .version_id = 0,
> +    .minimum_version_id = 0,
> +    .fields = (VMStateField[]) {
> +        VMSTATE_PCI_DEVICE(dev, RavenPCIState),
> +        VMSTATE_END_OF_LIST()
> +    },
> +};
> +
> +static PCIDeviceInfo raven_info = {
> +    .qdev.name = "raven",
> +    .qdev.desc = "PReP Host Bridge - Motorola Raven",
> +    .qdev.size = sizeof(RavenPCIState),
> +    .qdev.vmsd =&vmstate_raven,
> +    .qdev.no_user = 1,
> +    .no_hotplug = 1,
> +    .init = raven_init,
> +    .vendor_id = PCI_VENDOR_ID_MOTOROLA,
> +    .device_id = PCI_DEVICE_ID_MOTOROLA_RAVEN,
> +    .revision = 0x00,
> +    .class_id = PCI_CLASS_BRIDGE_HOST,
> +    .qdev.props = (Property[]) {
> +        DEFINE_PROP_END_OF_LIST()
> +    },
> +};
> +
> +static void raven_register_devices(void)
> +{
> +    pci_qdev_register(&raven_info);
> +}
> +
> +device_init(raven_register_devices)


How can the host bridge itself be a PCI device?  What we do for the i440fx is 
have a i440fx-pcihost device which is a SysBusDevice and then a create a 
separate PCIDevice (i440fx) which acts as the host device config space.

Regards,

Anthony Liguori