qemu-devel
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Qemu-devel] Memory API: handling unassigned physical memory


From: Peter Maydell
Subject: Re: [Qemu-devel] Memory API: handling unassigned physical memory
Date: Mon, 30 Apr 2012 15:03:46 +0100

On 30 April 2012 14:52, Mark Cave-Ayland <address@hidden> wrote:
> The key problem is that this doesn't worked with shared peripherals, such as
> the ESP device which is also used on various PPC Mac models as well as
> SPARC. That's because its init function looks like this:
>
>
>
> void esp_init(target_phys_addr_t espaddr, int it_shift,
>              ESPDMAMemoryReadWriteFunc dma_memory_read,
>              ESPDMAMemoryReadWriteFunc dma_memory_write,
>              void *dma_opaque, qemu_irq irq, qemu_irq *reset,
>              qemu_irq *dma_enable)
> {
>    ...
>    ...
>    sysbus_mmio_map(s, 0, espaddr);
>    ...
> }
>
>
> Therefore I can't change it to my (modified) sbus_mmio_map() function
> because it would break other non-SPARC platforms, and AIUI there is nothing
> in the memory API that allows me to move a subregion to a different
> MemoryRegion parent, even if I can get a reference to it with
> sysbus_mmio_get_region() after the sysbus_mmio_map() call - or have I
> misunderstood something?

Init functions like esp_init should be purely convenience functions
which create, set properties on, and map the sysbus/qdev device. If
they make use of private knowledge about the internals of the device
then this is wrong and should be fixed. For esp_init() it looks like
the handling of dma_memory_read, dma_memory_write, dma_opaque, it_shift
and dma_enabled are wrong.

If you fix that then you can just ignore the convenience function,
and create, configure and map the device as appropriate for you.

-- PMM



reply via email to

[Prev in Thread] Current Thread [Next in Thread]