Re: [PATCH 3/3] qemu-timer: reuse MIN macro in qemu_timeout_ns_to_ms

[Laurent Vivier]

Le 27/09/2019 à 13:57, Frediano Ziglio a écrit :
>>
>> Le 27/09/2019 à 12:32, Frediano Ziglio a écrit :
>>> Signed-off-by: Frediano Ziglio <address@hidden>
>>> ---
>>>  util/qemu-timer.c | 6 +-----
>>>  1 file changed, 1 insertion(+), 5 deletions(-)
>>>
>>> diff --git a/util/qemu-timer.c b/util/qemu-timer.c
>>> index d428fec567..9bd173ecda 100644
>>> --- a/util/qemu-timer.c
>>> +++ b/util/qemu-timer.c
>>> @@ -322,11 +322,7 @@ int qemu_timeout_ns_to_ms(int64_t ns)
>>>      ms = DIV_ROUND_UP(ns, SCALE_MS);
>>>  
>>>      /* To avoid overflow problems, limit this to 2^31, i.e. approx 25 days
>>>      */
>>> -    if (ms > (int64_t) INT32_MAX) {
>>> -        ms = INT32_MAX;
>>> -    }
>>> -
>>> -    return (int) ms;
>>> +    return (int) MIN(ms, INT32_MAX);
>>>  }
>>>  
>>
>> Perhaps it would be cleaner to have always the same return type for
>> MIN() with:
>>
>>     MIN(ms, (int64_t) INT32_MAX)
>>
> 
> The (int64_t) cast here is useless, ms is signed, C compilers knows how to
> convert and compare properly.

I agree.

> The function returns int and it's used in some cases to fill 32bit integer,
> changing the return type to int64_t is not trivial.
> But I suppose that here you mean instead that the 2 sides of the ternary
> operator used by MIN macro are int64_t and (probably) int32_t. In this case

Yes, it's what I mean.

Thanks,
Laurent