[Openchrome-users] OpenChrome driver and power saving
Andrey Melentyev
andrey.melentyev
Mon May 14 10:08:50 PDT 2007
A few days ago Intel released PowerTOP software: "PowerTOP is a Linux tool
that finds the software component(s) that make your laptop use more power
than necessary while it is idle." http://www.linuxpowertop.org/
I'm running a laptop with Unichrome Pro IGP on VIA VN800 and I've made some
simple tests with PowerTOP. The results are not optimistic mainly because of
X server. When X is running with OpenChrome driver and nothing happens on the
screen, "<interrupt> via at pci:0000:01:00.0" causes 60-70% of wakeups (absolute
value is 60 per second). We can see a very similar behaivor of i915 graphics
driver: http://www.linuxpowertop.org/powertop.php - Intel guys say that there
is a bug (See section "Top causes for wakeups" on this page and
http://www.linuxpowertop.org/known.php#intelgfx for more info).
When X server is not running, powertop doesn't report any wakeups and 'libata'
takes the top position in the list. So I suggested that there can be a
similar bug in OpenChrome driver. Futhermore it is very suspicious that the
number of caused wakeups is equally 60. It reminds me of my refresh rate
which is 60Hz. = 60 1 / second. And PowerTOP's site says: "The Intel graphics
driver used to have a bug where it set up the hardware to generate an
interrupt (called VBLANK) each time the screen finished refreshing (typically
at 60Hz or 72Hz)."
PS: a screenshot with powertop running on my system is attached
Best regards,
Andrew.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: via_powertop.png
Type: image/png
Size: 18294 bytes
Desc: not available
Url : http://wiki.openchrome.org/pipermail/openchrome-users/attachments/20070514/0a5dcbed/via_powertop.png
More information about the Openchrome-users
mailing list