Re: [PATCH] x86/vrtc: Use a real type for rtc_ioport_write()

On 28/07/2026 1:36 pm, Jan Beulich wrote:
> On 28.07.2026 14:31, Andrew Cooper wrote:
>> Signed-off-by: Andrew Cooper <[email protected]>
> Reviewed-by: Jan Beulich <[email protected]>

Thanks.

>
>> --- a/xen/arch/x86/hvm/rtc.c
>> +++ b/xen/arch/x86/hvm/rtc.c
>> @@ -469,9 +469,8 @@ static void cf_check rtc_alarm_cb(void *opaque)
>>      spin_unlock(&s->lock);
>>  }
>>  
>> -static int rtc_ioport_write(void *opaque, uint32_t addr, uint32_t data)
>> +static int rtc_ioport_write(RTCState *s, uint32_t addr, uint32_t data)
>>  {
>> -    RTCState *s = opaque;
>>      struct domain *d = vrtc_domain(s);
>>      uint32_t orig;
> It's entirely unclear why it was ever done like this. Even in 3.2 it could
> already have been done the type-correct way.

There are a whole bunch of callbacks which use this pattern.  They must
continue to use void * for the function types to work.

I wonder if at any point in the past rtc_ioport_write() was used
directly as a pointee, but I haven't gone looking.

~Andrew