PM WG weekly status meeting

Wednesday, 19-10-2011

Attendees

Name

Email

IRC Nick

Amit Kucheria

amit.kucheria@linaro.org

amitk

Mike Turquette

mike.turquette@linaro.org

mturquette

Vincent Guittot

vincent.guittot@linaro.org

vingu

Amit Kachhap

amit.kachhap@linaro.org

amitdk

Steven Jahnke

steve.jahnke@linaro.org

frautha?

Daniel Lezcano

daniel.lezcano@linaro.org

dlezcano

Mounir Bsaibes

mounir.bsaibes@linaro.org

mounir

Robin Randhawa

robin.randhawa@arm.com

?

Lorenzo Pieralisi

lorenzo.pieralisi@arm.com

?

Agenda

New Action items

  • AmitD get patches into linaro kernel, post to linaro-dev and handle today.
  • Mounir to follow up with Andy green on the thermal framework whether it is included in the Ti kernel. If so, highlight it to Marketing team.

Old Action Items

  • AmitK to create two new Connect sessions: 1) cpu c groups 2) Common save & restore code - DONE

  • Rob to send a note to Russel asking whether he accepts Freescale cpuidle driver, or Russell only accepts a consolidated driver.
  • AmitD to host Lorenzo's patches (branch) - DONE
  • AmitK to figure out what powersupply PNWG needs for use for measurement
  • Amitk to contact Kevin Hillman about which NIDAK to use - Trying to get Dave to talk about his H/W dungle.
  • Mike to send a schematic how to solder a resistor across VBAT, power measurement.
  • Rob to test sched_mc patch on i.MX6 and give feedback
  • AmitD to send content of /proc/interrupt on Samsung board to Daniel - DONE
  • Rob to send an email to the group to solicit feedback on a generic cpuidle driver

Minutes

  • Success Stories
    • ondemand governor of cpufreq driver and the sampling_down factor. The original patch has not been done by PMWG but we have tested it on ARM platform and correct bug. This is currently used by ST-E platform and have improved performance
    • For the cpu hotplug, not sure that it's a success story ... but it's used by ARM platform
    • cpu hotplug offline experimental, useful experimental tool, got Peter to talk to us.
  • Feedback on the upstream mapping document
    • When in research mode and discussion with upstream, it is hard to plan ahead of time, we don't know what is the outcome maybe from the discussion. a blueprint may start with 6 work items and end up with 30 Work items.
    • you can get stuck ona WI for a whole month. If you have not defined the WI correctly you could get stuck on the item for long time.
  • Connect attendance
    • Vincent will attend remotely
    • Robin may come
  • Review of blueprints
    • Thermal Framework is done. Need to check with TI landing team whether it is integrated in their kernel. If so, highlight it to Marketing
    • Common clock some items will be moved to November, the git tree not published yet, the second version of the core common clock is being discussed on LKML.
    • Sched_mc will integrate at the end of the week will publish the Vincent git tree at the end of this week, but will not integrate in this linaro-kernel release
    • Adapt Samsung sensor driver to in-kernel thermal framework. most work is done. not published yet. The wiki WI will not be done this week.
  • releases
    • powerdebug and qa-power will be released tomorrow
    • powertop - no release tomorrow, the fix is not done, just found the cause.

WorkingGroups/PowerManagement/Meetings/2011-10-19 (last modified 2011-10-19 17:29:02)