Re: [patch 1/3] sched_clock: document 4Mhz vs 1Mhz decision

[Ingo Molnar <mingo@kernel.org>]

* akpm@linux-foundation.org <akpm@linux-foundation.org> wrote:

> From: Stephen Boyd <sboyd@codeaurora.org>
> Subject: sched_clock: document 4Mhz vs 1Mhz decision
> 
> Bo Shen sent a patch to change this to 1Mhz instead of 4Mhz but according
> to Russell King the use of 4Mhz was intentional.  Add a comment to this
> effect so that others don't try to change the code as well.
> 
> Signed-off-by: Stephen Boyd <sboyd@codeaurora.org>
> Cc: Bo Shen <voice.shen@atmel.com>
> Cc: Russell King <rmk@arm.linux.org.uk>
> Cc: Thomas Gleixner <tglx@linutronix.de>
> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
> ---
> 
>  kernel/time/sched_clock.c |    4 ++++
>  1 file changed, 4 insertions(+)
> 
> diff -puN kernel/time/sched_clock.c~sched_clock-document-4mhz-vs-1mhz-decision kernel/time/sched_clock.c
> --- a/kernel/time/sched_clock.c~sched_clock-document-4mhz-vs-1mhz-decision
> +++ a/kernel/time/sched_clock.c
> @@ -132,6 +132,10 @@ void __init sched_clock_register(u64 (*r
>  	clocks_calc_mult_shift(&cd.mult, &cd.shift, rate, NSEC_PER_SEC, 3600);
>  
>  	r = rate;
> +	/*
> +	 * Use 4MHz instead of 1MHz so that things like 1.832Mhz show as
> +	 * 1832Khz
> +	 */
>  	if (r >= 4000000) {
>  		r /= 1000000;
>  		r_unit = 'M';

Instead of magic precision cutoff limits the better solution would be 
to simply display not just decimal part but also the first 3 digits of 
the fractional part:

          1.123  Hz
	990.134 kHz
	  1.832 MHz
	 12.314 GHz

etc. - then every number could be displayed in its natural unit 
without magic precision limits.

It would also be mostly self-documenting.

Thanks,

	Ingo