[ltt-dev] [PATCH] Fix usage of trace_clock_freq_scale() as a multiplier

[Mathieu Desnoyers]

* Benjamin Poirier (benjamin.poirier at polymtl.ca) wrote:
> According to http://www.lttng.org/svn/trunk/lttv, svn revision 3150 commit
> log: freq scale is now a divisor
> ---
>  kernel/trace/trace-clock-32-to-64.c |    8 ++++----
>  1 files changed, 4 insertions(+), 4 deletions(-)
> 
> diff --git a/kernel/trace/trace-clock-32-to-64.c b/kernel/trace/trace-clock-32-to-64.c
> index b1fdceb..701aeb1 100644
> --- a/kernel/trace/trace-clock-32-to-64.c
> +++ b/kernel/trace/trace-clock-32-to-64.c
> @@ -158,10 +158,10 @@ static int __init precalc_stsc_interval(void)
>  	u64 rem_freq, rem_interval;
>  
>  	precalc_expire =
> -		__iter_div_u64_rem(HW_BITMASK,
> -		 ((__iter_div_u64_rem(trace_clock_frequency(), HZ, &rem_freq)
> -		     * trace_clock_freq_scale())
> -		    << 1)
> +		__iter_div_u64_rem(HW_BITMASK, (
> +		  __iter_div_u64_rem(trace_clock_frequency(),
> +		  HZ * trace_clock_freq_scale(), &rem_freq) << 1
> +		 )

Ah, right. Although all architectures currently use a freq_scale of 1 ;)

It will be useful if we ever want to express fraction of cycles/s to
improve clock frequency precision...

Will merge, thanks !

Mathieu

>  		 - 1
>  		 - (TC_EXPECTED_INTERRUPT_LATENCY * HZ / 1000), &rem_interval)
>  		>> 1;
> -- 
> 1.6.3.3
> 
> 
> _______________________________________________
> ltt-dev mailing list
> ltt-dev at lists.casi.polymtl.ca
> http://lists.casi.polymtl.ca/cgi-bin/mailman/listinfo/ltt-dev
> 

-- 
Mathieu Desnoyers
OpenPGP key fingerprint: 8CD5 52C3 8E3C 4140 715F  BA06 3F25 A8FE 3BAE 9A68