Re: [PATCH v3 1/3] hw/pci: Introduce pci_dma_memory_region() helper

[Mark Cave-Ayland]

On 26/08/2021 21:07, Philippe Mathieu-Daudé wrote:

> pci_get_address_space() returns an AddressSpace. In some
> cases we want a MemoryRegion. Add the pci_dma_memory_region()
> equivalent helper.
>
> Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>
> ---
>   include/hw/pci/pci.h | 5 +++++
>   1 file changed, 5 insertions(+)
>
> diff --git a/include/hw/pci/pci.h b/include/hw/pci/pci.h
> index d0f4266e372..5860f42e400 100644
> --- a/include/hw/pci/pci.h
> +++ b/include/hw/pci/pci.h
> @@ -786,6 +786,11 @@ static inline AddressSpace 
> *pci_get_address_space(PCIDevice *dev)
>       return &dev->bus_master_as;
>   }
>   
> +static inline MemoryRegion *pci_dma_memory_region(PCIDevice *dev)
> +{
> +    return &dev->bus_master_container_region;
> +}
> +
>   /**
>    * pci_dma_rw: Read from or write to an address space from PCI device.
>    *

This patch doesn't quite feel right: there are already a number of devices that have 
been using pci_get_address_space() for some time without requiring access to the 
underlying MemoryRegion. So then the first question I ask myself when I see this 
patch is why do we suddenly need it now?


ATB,

Mark.