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

Bug#495541: linux-image-2.6.26-1-686: Please reactivate CONFIG_ACPI_PROCFS_POWER



clone 495541 -1
reassign -1 powertop
retitle -1 fix assumption on /sys exported power files
stop

On Mon, Aug 18, 2008 at 07:52:21PM +0200, Christian Perrier wrote:
> Quoting maximilian attems (max@stro.at):
> 
> > > So, the support for it was reinstated by Maks in 2.6.24-3.
> > 
> > yes it was to give userspace some time to react.
> > we allways stated that we *don't* guarantee to have it set for lenny.
> 
> What I regret is that there should have been more communication with
> developers.
> 
> Being the source of the change, the consequences should have been
> tracked by the kernel team and the change made at the moment all
> userspace utilities have been updated. Which is what I call a
> transition.

no the kernel is not a lib
if an app doesn't work with default kernel it is plain buggy.

 
> Well:
> 
> root@mykerinos:/etc/acpi/events> dpkg -s powertop
> Package: powertop
> .../...
> Version: 1.10-1
> 
> Is there something I'm missing here ?
> 
> "my" powertop is still reporting that "no ACPI power usage estimate
> available"

right we debugged that with jcristau on irc and powertop is buggy.
if current_now or energy_now or voltage_now is not present in
for example /sys/class/power_supply/BAT0/ no value will be shown.
 
> See also #491396 for acpi-support: that one lead me to think that
> there are still utilities that are broken by dropping /proc/acpi
> support and it would better be reinstated.
> 
> (indeed all this plus following Sulamita Garcia's talk at DC8 and the
> discussions we had later brings me to the idea of a "Power Management
> team", dealing with these issues and who could be a good reference
> point in such situations, but that's mostly another story)

deprecated ACPI_PROCFS_POWER will be set for lenny as it is show now
that too much userspace is still too buggy to interpret the sys files.

kind regards

-- 
maks



Reply to: