Re: [Patch 3/6] statistics infrastructure - prerequisite: timestamp

[Martin Peschke <mp3@de.ibm.com>]

Balbir Singh wrote:
>> +static inline int nsec_to_timestamp(char *s, unsigned long long t)
>> +{
>> +       unsigned long nsec_rem = do_div(t, 1000000000);
> 
> Could we please use NSEC_PER_SEC. I cannot count the number of zeros
> after the 1.

Sure. I tried to keep my changes as small as possible ;-)

>> +       return sprintf(s, "[%5lu.%06lu]", (unsigned long)t, 
>> nsec_rem/1000);
>> +}
> 
> Something symbolic for the 1000 would be better.  NSECS_PER_USEC probably?

Makes sense.

Thanks,
Martin