qemu-devel
[Top][All Lists]
Advanced

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

Re: [Qemu-devel] [PATCH 11/24] integratorcp: convert to memory API (RAM/


From: Peter Maydell
Subject: Re: [Qemu-devel] [PATCH 11/24] integratorcp: convert to memory API (RAM/flash only)
Date: Wed, 24 Aug 2011 12:22:18 +0100

On 24 August 2011 11:11, Avi Kivity <address@hidden> wrote:
> @@ -108,9 +111,15 @@ static uint32_t integratorcm_read(void *opaque, 
> target_phys_addr_t offset)
>  static void integratorcm_do_remap(integratorcm_state *s, int flash)
>  {
>     if (flash) {
> -        cpu_register_physical_memory(0, 0x100000, IO_MEM_RAM);
> +        if (s->flash_mapped) {
> +            sysbus_del_memory(&s->busdev, &s->flash);
> +            s->flash_mapped = false;
> +        }
>     } else {
> -        cpu_register_physical_memory(0, 0x100000, s->flash_offset | 
> IO_MEM_RAM);
> +        if (!s->flash_mapped) {
> +            sysbus_add_memory_overlap(&s->busdev, 0, &s->flash, 1);
> +            s->flash_mapped = true;
> +        }
>     }

This is introducing a new field in the device state which is actually
just tracking s->cm_init bit 2. At least it would be, except that
in integratorcm_set_ctrl this line:
    s->cm_init = (s->cm_init & ~ 5) | (value ^ 5);

appears to be using ^ when it meant & ...

This isn't a nak -- I'm happy for this to get committed and I'll fix
things up later, since the device doesn't have a VMStateDescription
that would be broken by the extra state field. (Or if I get round to
it before the series gets committed I'll send you a fix to squash
into this patch.)

-- PMM



reply via email to

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