Re: [Xen-devel] [PATCH V3 33/41] xen/arm: Add versatile express platform

[Julien Grall <julien.grall@xxxxxxxxxx>]

On 05/10/2013 03:06 PM, Ian Campbell wrote:

> On Fri, 2013-05-10 at 15:00 +0100, Julien Grall wrote:
>> On 05/10/2013 10:44 AM, Ian Campbell wrote:
>>
>>>> @@ -91,6 +93,47 @@ out:
>>>>  }
>>>>  
>>>>  /*
>>>> + * TODO: Get base address from the device tree
>>>> + * See arm,vexpress-reset node
>>>> + */
>>>> +static void vexpress_reset(void)
>>>> +{
>>>> +    void __iomem *base;
>>>> +    void __iomem *sp810;
>>>> +
>>>> +    /* Use the SP810 system controller to force a reset */
>>>> +    base = ioremap_attr(SP810_ADDRESS & PAGE_MASK, PAGE_SIZE,
>>>> +                        PAGE_HYPERVISOR_NOCACHE);
>>>> +    if ( !base )
>>>> +    {
>>>> +        dprintk(XENLOG_ERR, "Unable to map SP810\n");
>>>> +        return;
>>>> +    }
>>>> +
>>>> +    sp810 = base + (SP810_ADDRESS & ~PAGE_MASK);
>>>
>>> Didn't I see you making vunmap, which iounmap is based on, take care of
>>> the page offsets itself in an earlier patch? Or is that not the reason
>>> you are going through the base + offset dance?
>>
>>
>> I wasn't not sure if you will accept ioremap(..., 8); I will fix the patch.
> 
> I'm confused, what does (..., 8) mean?

Because we only need to the first 8 bytes.

> What I was asking about was why the code couldn't be:
>       sp810 = ioremap_..(SP810_ADDRESS, PAGE_SIZE,..)
>       iowritel(sp810,...)
>       iounmap(sp810)


Indeed. I will use this solution.

-- 
Julien


_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxx
http://lists.xen.org/xen-devel