Results 1 to 7 of 7
  1.    #1  
    DO NOT RUN THIS KERNEL UNLESS YOU KNOW WHAT YOU ARE DOING. YOU ARE RESPONSIBLE FOR MAINTENANCE AND REPAIR OF YOUR OWN PHONE.

    I STILL HAVE NO PLANS TO BE ANYONE'S MOTHER.


    I have a new kernel based off the 1.4.1-6 tree, In addition to the previous kernel, this has:

    * Experimental temperature sensor enabled, available in /sys/devices/platform/omap34xx_temp/ - the clocking may be incorrect, please check cooked and raw values.
    * Experiments with SmartReflex
    * Force display off during suspend
    * Trimming of debug code from power handling
    * Further increase of delay for resume from suspend
    * Options enabled for iotop

    The power saving results are interesting; after charging the phone up fully then leaving it off charge overnight, I have the following assessment from Battery Monitor:

    Initial battery %: 100
    Current battery %: 83
    Runtime: 9h21m50s
    Drain per hour %: 1.82
    Remaining life (hrs): 45.72

    This is with WiFi and WAN on, Bluetooth off, GPS off. This is not a test of real world use, it is a test of the phone in standby waiting to receive a call. The battery will quite definitely not last 45 hours if the phone is actually active for any length of time, but I would be interested in results from anyone else regarding positive or negative impacts on battery life.

    I have committed my changes to the git tree for those who with to poke around; kernel is available at http://www.fop.ns.ca/pre/uImage.1.4.1-6.9 for those who wish to try it.

    Cheers, Steve
    Last edited by sbromwich; 04/17/2010 at 09:51 AM. Reason: speling
  2. #2  
    The temperature values, I get -40 and 0 (cooked and raw respectively), what should I be getting? Battery temp is at 35C.

    Any ideas?
  3.    #3  
    It seems like cooked is either -40 or 39, and raw is 0; ka6sox is digging through it, he said something about the clock timing IIRC. Hopefully he can fix it up

    Cheers, Steve
  4. #4  
    Quote Originally Posted by sbromwich View Post
    It seems like cooked is either -40 or 39, and raw is 0; ka6sox is digging through it, he said something about the clock timing IIRC. Hopefully he can fix it up

    Cheers, Steve
    I hope ya can fix the temperature, that'd be sweet


    Though, it's work linking this post Rod sent me over IRC.
    "There is no support for a temperature sensor in the OMAP35x. I would not suggest relying on accessing these registers in the OMAP35x device as there is no guarantee that these will be reliable values. You most likely want to use another temperature sensor."
    If you've liked my software, please consider to towards future development.

    Developer of many apps such as: WebOS Quick Install, WebOS Theme Builder, Ipk Packager, Unified Diff Creator, Internalz Pro, ComicShelf HD, LED Torch, over 70 patches and more.

    @JayCanuck @CanuckCoding Facebook
  5.    #5  
    [QUOTE=Jason Robitaille;2388299]I hope ya can fix the temperature, that'd be sweet

    I'm not the one working on it; ka6sox took it from a later kernel and backported it, I am just standing on his shoulders.

    Cheers, Steve
  6. #6  
    But the post talks about OMAP 35xx, and Pre has 3430, right? Could it be that that register is on Pre's Processor but not on later ones?
  7.    #7  
    Quote Originally Posted by NuttyBunny View Post
    But the post talks about OMAP 35xx, and Pre has 3430, right? Could it be that that register is on Pre's Processor but not on later ones?
    Certainly a possibility but ka6sox seemed to think he had a fix. Wait and see I guess

    Cheers, Steve

Posting Permissions