Re: [PATCH v3 5/5] timekeeping: utilize the suspend-nonstop clocksourceto count suspended time

[Thomas Gleixner]

On Wed, 6 Mar 2013, Feng Tang wrote:
> On Wed, Mar 06, 2013 at 03:15:49PM +0100, Thomas Gleixner wrote:
> > If the suspend time measured by the nonstop clocksource is 0.999 sec
> > then we throw it away and then let the RTC code inject inaccurate
> > sleep time? Brilliant design, really.
> 
> Emm, I wrote the code with an assumption that the sleep itself and the
> enter/exit processes will be longer than 1 second.
> 
> I can initialize the ts_delta to (0, 0} and change the check condition
> to 
> 	if (ts_delta.tv_sec || ts_delta.tv_nsec)
> 

To be honest, I hate this abuse of ts_delta as a condition. Use a
separate condition variable to make clear what the hell that code is
doing. This is not a fast path and I really prefer understandable code
over magic microoptimizations for no value.

Thanks,

	tglx
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/