http://defect.opensolaris.org/bz/show_bug.cgi?id=3838
Eric Saxe <eric.saxe at sun.com> changed:
What |Removed |Added
----------------------------------------------------------------------------
CC| |eric.saxe at sun.com
--- Comment #2 from Eric Saxe <eric.saxe at sun.com> 2008-10-10 14:34:32 ---
(In reply to comment #1)
> Eric,
>
> It sounds good. Had couple of comments/questions to make myself clear.
>
> 1) event-mode will only exist for x86 right? If that is the case, by default
> "cpupm enable" for x86 will be event mode and for sparc it will be
> "polled-mode". Does that make sense?
Right, cpupm enable specifies an enabling of CPU power management using the
system's default mechanism...where that default may vary depending on what's
possible / supported. That makes sense.
> 2) If "polled-mode" has to be used, then I assume "cpupm enable" has to be
> replaced by "cpupm polled-mode".
Right, the idea is if you want to be specific about the mechanism used you can.
Otherwise, the system will go with the default.
> 3) Instead of having 4 behaviors, How about having three.
>
> a)enable-event-mode
> b) enable-polling-mode
> c)disable
>
> x86 will support all the three and sparc supports just 'b' and 'c'. Just my
> thoughts.
I had thought about that. Here's why I'm leaning towards preserving the enable
behavior:
- It's backwards compatible, existing power.conf files will work in a way
consistent with "the principle of least surprise", since "enable" will
still
be legal.
- It allows users not to have to care about the best mechanism to use,
which (given 1 above) may vary in certain cases. They can say "enable" as
they have done in the past.
- If we didn't support enable, and the default behavior varies in cases, then
I was thinking this would mean we would have to ship different default
power.conf files with the right defaults in place for the system...which
could be messy...and might be easier to deal with at run time.
Thanks,
-Eric
--
Configure bugmail: http://defect.opensolaris.org/bz/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.
