Skip to contentSkip to conversation

n800 high idle power consumption

DISCUSSION Kalle Valo 2 participants 4 messages Apr 24, 2009

lore
Share & tools
lore

Kalle Valo

author
#0 lore raw

Hello all,

I was testing 2.6.30-rc3-omap1 on n800 and noticed that idle power consumption is very high, ten times compared to a 2.6.29 I tested few weeks back. I'm guessing that CPU doesn't sleep at all.

Is this a regression? How should I debug this further?

Signature · 1 line
Kalle Valo
author
#1 lore raw
Quoted 1 line · Kalle Valo

Hello all,

Hello again,

Quoted 5 lines

I was testing 2.6.30-rc3-omap1 on n800 and noticed that idle power consumption is very high, ten times compared to a 2.6.29 I tested few weeks back. I'm guessing that CPU doesn't sleep at all.

Is this a regression? How should I debug this further?

Oh yeah, I forgot to mention that I have included Jarkko's patch "ARM:
OMAP: Fix tusb6010 init error and compilation warning" in my builds. Otherwise usb won't work.

Signature · 1 line
Kalle Valo
#2 lore raw
In reply to Kalle Valo
Quoted 7 lines · Kalle Valo

Hello all,

I was testing 2.6.30-rc3-omap1 on n800 and noticed that idle power consumption is very high, ten times compared to a 2.6.29 I tested few weeks back. I'm guessing that CPU doesn't sleep at all.

Is this a regression? How should I debug this further?

Between .29 and .30, the PM code in l-o was dropped and re-sync'd to kernel.org code. This is so that PM can be separated out into a separate branch and submitted upstream independently.

The current PM branch is in the process of being re-based onto 2.6.30. When it's ready, I'd appreciate any testing on n800.

Kevin

author
#3 lore raw
Quoted 8 lines · Kevin Hilman
Quoted 6 lines · Kalle Valo

Hello all,

I was testing 2.6.30-rc3-omap1 on n800 and noticed that idle power consumption is very high, ten times compared to a 2.6.29 I tested few weeks back. I'm guessing that CPU doesn't sleep at all.

Quoted 5 lines
Quoted 1 line

Is this a regression? How should I debug this further?

Between .29 and .30, the PM code in l-o was dropped and re-sync'd to kernel.org code. This is so that PM can be separated out into a separate branch and submitted upstream independently.

Ah, of course. I read about this but didn't understand the implications :) Thanks for pointing this out to me.

Quoted 2 lines

The current PM branch is in the process of being re-based onto 2.6.30. When it's ready, I'd appreciate any testing on n800.

Ok. Please let me know when it's ready.

Signature · 1 line
Kalle Valo