RE: [PATCH/RFT 1/1] OMAP2+: cpufreq: scale voltage along with frequency

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



Hi Kevin,

On Thu, Feb 23, 2012 at 00:12:06, Hilman, Kevin wrote:
> > And in my patch plus - minus was not used as regulator framework will
> > try to set voltage for the least voltage which sometimes corresponds
> > to exact OPP required value.
> 
> sometimes?

I was not clear in my previous statement, let me explain it differently.

Regulator framework will always try to set lowest voltage in the range.
Upon, regulator_set_voltage(reg, OPPVOLT, (OPPVOLT+RESOLUTION-1)), if
OPPVOLT is a value that can be set by the regulator, that will be set by
the regulator, this is what I meant by 'sometimes'. Otherwise it will set
the next possible step.

> Using your example above, what if the closest value was 1.259V?
> Wouldn't you then need +/- range?

In that case, it will set to next step after 1.259V.

If +/- is used, it may happen that SoC will work for a particular frequency
at a voltage lower than it has been characterized (if you ask me to define
characterized, I don't know, but what I know is that SoC can work at a
lower frequency for the voltage of an OPP, but not vice versa). Still as
voltage will be only very less than that specified by OPP, it may not be
a problem to use +/-

> Please have a look at my v2.
> 
> The drivers will get pre and post notifiers, with the caveat that upon
> DVFS failure, the freq in the post notifier might be the same as the one
> in the pre notifier, indicating that no change was made.

Yes, I overlooked your freq.new update for error condition.

Earlier while adding support for DVFS on AM335X, it was noticed that, lpj
was going wrong after dvfs failure, then those error notifiers were added
to recover properly. Later it was realized that it was due to a bug in
cpufreq framework,

"d08de0c1  [CPUFREQ] update lpj only if frequency has changed"

fixed it, probably those extra handling in  my patch is not required now,
if drivers can understand your pre & post notifiers properly (if drivers
can't, then maybe it is driver problem)

Regards
Afzal
--
To unsubscribe from this list: send the line "unsubscribe linux-omap" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html


[Index of Archives]     [Linux Arm (vger)]     [ARM Kernel]     [ARM MSM]     [Linux Tegra]     [Linux WPAN Networking]     [Linux Wireless Networking]     [Maemo Users]     [Linux USB Devel]     [Video for Linux]     [Linux Audio Users]     [Yosemite Trails]     [Linux Kernel]     [Linux SCSI]

  Powered by Linux