* pal/power_info for real?
@ 2003-11-25 9:09 Duraid Madina
2003-11-25 18:24 ` Alex Williamson
0 siblings, 1 reply; 2+ messages in thread
From: Duraid Madina @ 2003-11-25 9:09 UTC (permalink / raw)
To: linux-ia64
Hi all,
/proc/pal/cpu0/power_info on my system reads:
Power level 0:
entry_latency : 9000 cycles
exit_latency : 8700 cycles
power consumption : 52000 mW
Cache+TLB coherency : Yes
Power level 1: not implemented
Power level 2:
entry_latency : 14000 cycles
exit_latency : 8700 cycles
power consumption : 52000 mW
Cache+TLB coherency : Yes
Power level 3: not implemented
Power level 4: not implemented
Power level 5: not implemented
Power level 6: not implemented
Power level 7: not implemented
Is this correct? Two power levels, with different entry latencies, but
the same power consumption?
Bracing myself for a particularly warm summer,
Duraid
^ permalink raw reply [flat|nested] 2+ messages in thread
* Re: pal/power_info for real?
2003-11-25 9:09 pal/power_info for real? Duraid Madina
@ 2003-11-25 18:24 ` Alex Williamson
0 siblings, 0 replies; 2+ messages in thread
From: Alex Williamson @ 2003-11-25 18:24 UTC (permalink / raw)
To: linux-ia64
On Tue, 2003-11-25 at 02:09, Duraid Madina wrote:
> Is this correct? Two power levels, with different entry latencies, but
> the same power consumption?
Yep, power level 0 is light_halt and 2 is halt2. I can't really
comment on the power consumption, but presumably they're shutting down
different portions of the chip. I see the same thing on my boxes. Note
that these are all halt state power consumptions, not full running power
consumption.
Alex
--
Alex Williamson HP Linux & Open Source Lab
^ permalink raw reply [flat|nested] 2+ messages in thread
end of thread, other threads:[~2003-11-25 18:24 UTC | newest]
Thread overview: 2+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2003-11-25 9:09 pal/power_info for real? Duraid Madina
2003-11-25 18:24 ` Alex Williamson
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox