Re: [PATCH 2/2] x86/tsc: set LAPIC timer frequency to crystal clock frequency
From: Ingo Molnar
Date: Tue Apr 23 2019 - 06:30:42 EST
* Daniel Drake <drake@xxxxxxxxxxxx> wrote:
> On Mon, Apr 22, 2019 at 8:04 PM Ingo Molnar <mingo@xxxxxxxxxx> wrote:
> > Minor style nit: the parentheses are unnecessary, integer expressions
> > like this are evaluated left to right and multiplication and division has
> > the same precedence.
>
> Fair point, although the same could be said for cpu_khz_from_msr().
Yes, this standardization on mhz, if it makes sense, might have to be
propagated a bit to function names and any other variables.
Another naming quirk: what unit is 'lapic_timer_frequency' in? AFAICS
it's a "period" unit (number of clock cycles per jiffy), not a frequency
(which is number of cycles per second) - so the better name might be
lapic_timer_period?
> > But it might also make sense to actually store crystal_mhz instead of
> > crystal_khz, because both CPUID 15H and 16H provides MHz values.
> >
> > That way the above expression would simplify to:
> >
> > lapic_timer_frequency = crystal_mhz / HZ;
>
> Wouldn't it be
> lapic_timer_frequency = crystal_mhz * 1000000 / HZ;
> ?
Sorry, what I wanted to suggest is crystal_hz, not crystal_mhz.
I.e. store the raw unit that comes out of the CPUID, which appears to be
HZ, right?
That would change the calculation to:
lapic_timer_period = crystal_hz / HZ;
Note how clearly it's visible in that calculation that what we calculate
there is a 'period', not a frequency.
Thanks,
Ingo