Re: [PATCH 1/2] x86/vPIT: re-order functions

[Jan Beulich <jbeulich@xxxxxxxx>]

On 01.06.2023 11:17, Roger Pau Monné wrote:
> On Tue, May 30, 2023 at 05:30:02PM +0200, Jan Beulich wrote:
>> To avoid the need for a forward declaration of pit_load_count() in a
>> subsequent change, move it earlier in the file (along with its helper
>> callback).
>>
>> Signed-off-by: Jan Beulich <jbeulich@xxxxxxxx>
> 
> Reviewed-by: Roger Pau Monné <roger.pau@xxxxxxxxxx>

Thanks.

> Just a couple of nits, which you might also noticed but decided to not
> fix given this is just code movement.

Indeed, I meant this to be pure code movement. Nevertheless I'd be happy
to take care of style issues, if that's deemed okay in a "pure code
movement" patch. However, ...

>> --- a/xen/arch/x86/emul-i8254.c
>> +++ b/xen/arch/x86/emul-i8254.c
>> @@ -87,6 +87,57 @@ static int pit_get_count(PITState *pit,
>>      return counter;
>>  }
>>  
>> +static void cf_check pit_time_fired(struct vcpu *v, void *priv)
> 
> Seems like v could be constified?

... the function being used as a callback, I doubt adding const would
be possible. Otoh ...

>> +{
>> +    uint64_t *count_load_time = priv;

... there's a blank line missing here, if I was to go for style
adjustments.

Jan

>> +    TRACE_0D(TRC_HVM_EMUL_PIT_TIMER_CB);
>> +    *count_load_time = get_guest_time(v);
>> +}
>> +
>> +static void pit_load_count(PITState *pit, int channel, int val)
>> +{
>> +    u32 period;
> 
> uint32_t
> 
> Thanks, Roger.