On 17/09/2021 14:10, Jan Beulich wrote:
> On 17.09.2021 10:45, Andrew Cooper wrote:
>> @@ -1336,13 +1338,16 @@ update_runq_load(const struct scheduler *ops,
>>      {
>>          struct {
>>              uint64_t rq_avgload, b_avgload;
>> -            unsigned rq_load:16, rq_id:8, shift:8;
>> -        } d;
>> -        d.rq_id = rqd->id;
>> -        d.rq_load = rqd->load;
>> -        d.rq_avgload = rqd->avgload;
>> -        d.b_avgload = rqd->b_avgload;
>> -        d.shift = P;
>> +            uint16_t rq_load; uint8_t rq_id, shift;
> Split into two lines? Preferably with this adjustment
> Reviewed-by: Jan Beulich <[email protected]>

Thanks.

> I'd like to note that the remaining uses of "unsigned int" or plain
> "int" in some of the instances you don't touch assume
> sizeof(int) == 32, while generally we assume only sizeof(int) >= 32.
> This is one of the cases where fixed width types are imo mandatory
> to use.

See patch 5.  There was far too much cleanup to merge with this patch.

~Andrew

Reply via email to