Page 6 of 14 FirstFirst 1234567891011 ... LastLast
Results 101 to 120 of 274
  1. angiest's Avatar
    Posts
    933 Posts
    Global Posts
    952 Global Posts
    #101  
    Quote Originally Posted by IlyaK View Post
    So if I installed SR600 through the IPK, after a reboot that patch is no longer working?

    If so, I'm happy with my phones performance without the patch. The 1.3.5 update is working good for me. Do I need to do anything to remove it?
    I don't know if anyone has verified SmartReflex is being deactivated on Luna restart, but it does seem the CPU is being throttled back to 500mHz on Luna restart.
  2. #102  
    Okay...then I should be good. I've done a reboot of my phone a few times since applying the patch.
    Palm Pre | WebOS 1.4.1.1 | Sprint
    My CarDomain - 5/30/09
  3. a2k
    a2k is offline
    a2k's Avatar
    Posts
    103 Posts
    Global Posts
    111 Global Posts
    #103  
    Quote Originally Posted by rayln View Post
    SmartReflex is a voltage monitoring process supported by TI's OMAP line.

    the op assumed that it was SmartReflex that was... incompatible (for lack of a better word) with 1.3.5 and simply removed all lines of code beside pertaining to it.

    the only things left in his executable file are notes from the original patch and the following line of code repeated twice:
    echo 600000 > /sys/devices/system/cpu/cpu0/cpufreq/scaling_setspeed

    this causes the cpu to clock at 600mhz but the same effect can be achieved by simply entering the above line once in a terminal session.

    what we've uncovered is that *something* is causing the cpu to revert upon a luna restart or phone reboot.

    my hunch is that luna itself is the cause of this due to the above issue I've encountered.

    upon entering and closing a second terminal session, luna restarts and the cpu reverts to 500mhz.

    luna is the java sys mgr layer sitting atop the underlying linux and restarts upon a phone reboot as well. hence why I believe luna is at the root of this.

    my hunch involves luna running checks or possibly running an 'initialize' settings upon start (set cpu to 500mhz etc.).

    as to why the second session closing would cause luna to restart, I have no idea.

    I haven't looked through, but I assume the terminal "app" runs an open console command or similar.

    I haven't delved deeply at all regarding luna's interactions with linux but I have a few scenarios I'd like to explore.

    presumably, ANY tampering with the cpu settings (scaling back etc.) would be reverted upon a luna restart.

    except for the facts, please take my musings with a grain of salt. I'm no linux/webOS guru and I do not read java.
    Thank you rayln.

    So by having the OP's script in the event.d directory, does it execute every time the phone or Luna is restarted, thus clocking back to 600?
  4. angiest's Avatar
    Posts
    933 Posts
    Global Posts
    952 Global Posts
    #104  
    Quote Originally Posted by a2k View Post
    Thank you rayln.

    So by having the OP's script in the event.d directory, does it execute every time the phone or Luna is restarted, thus clocking back to 600?
    The /etc/event.d directory is used by the upstart init system used by WebOS. The init system is how UNIX-like systems run startup services (it's a tad more complicated than that). Therefore, the scripts in event.d are only run when WebOS is booted or restarted and not when Luna is restarted. (Again, it is actually a little more complicated.)
  5. #105  
    Does that mean that restarting JavaScript after normal or Luna reboot would set it back at 600?
  6. a2k
    a2k is offline
    a2k's Avatar
    Posts
    103 Posts
    Global Posts
    111 Global Posts
    #106  
    Quote Originally Posted by angiest View Post
    The /etc/event.d directory is used by the upstart init system used by WebOS. The init system is how UNIX-like systems run startup services (it's a tad more complicated than that). Therefore, the scripts in event.d are only run when WebOS is booted or restarted and not when Luna is restarted. (Again, it is actually a little more complicated.)

    Thanks for explaining. But a good question was just brought up. Does restarting JavaScript (from the button in the LunaManager application ) run startup services from the event.d directory once again?
  7. #107  
    Wow!

    I go to sleep and wake up and finally people believe me! woohoo!

    Now i'm going to see if SR is still activated even tho the CPU is clocked lower after Luna start/restart.
    Live free or DIE!
  8. rayln's Avatar
    Posts
    215 Posts
    Global Posts
    239 Global Posts
    #108  
    Quote Originally Posted by angiest View Post
    I don't know if anyone has verified SmartReflex is being deactivated on Luna restart, but it does seem the CPU is being throttled back to 500mHz on Luna restart.
    that's true, the SmartReflex process itself hasn't been checked for deactivation.

    Quote Originally Posted by a2k View Post
    Thanks for explaining. But a good question was just brought up. Does restarting JavaScript (from the button in the LunaManager application ) run startup services from the event.d directory once again?
    as far as I understand, no.
    java doesn't have access to such low level facilities.

    of course, I haven't checked.
    but I'm pretty sure.

    feel free to try however.
  9. #109  
    looks like SR is still activated after Luna restart. so running SR500 is fine.

    root@palm-webos-device:/# echo 600000 > /sys/devices/system/cpu/cpu0/cpufreq/scaling_setspeed
    root@palm-webos-device:/# echo -n 1 > /sys/power/sr_vdd1_autocomp
    root@palm-webos-device:/# echo -n 1 > /sys/power/sr_vdd2_autocomp
    root@palm-webos-device:/# cat /sys/devices/system/cpu/cpu0/cpufreq/scaling_cur_freq
    600000
    root@palm-webos-device:/#
    root@palm-webos-device:/# cat /sys/power/sr_vdd1_autocomp
    1
    root@palm-webos-device:/# cat /sys/power/sr_vdd2_autocomp
    1
    root@palm-webos-device:/# pkill LunaSysMgr
    root@palm-webos-device:/#
    root@palm-webos-device:/# cat /sys/devices/system/cpu/cpu0/cpufreq/scaling_cur_freq
    500000
    root@palm-webos-device:/# cat /sys/power/sr_vdd1_autocomp
    1
    root@palm-webos-device:/# cat /sys/power/sr_vdd2_autocomp
    1
    Live free or DIE!
  10. #110  
    smartreflex is working despite some people reporting instability.

    I've checked and smartreflex gets turned on by my event.d script, and it stays on even if luna restarts. It seems as if only cpu clock speed is resetting.

    I read somewhere on here someone said it might be powerd.. If it's not hardcoded into luna then it is most likely a power saving daemon (such as powerd)
  11. #111  
    Quote Originally Posted by xstaticxgpx View Post
    smartreflex is working despite some people reporting instability.

    I've checked and smartreflex gets turned on by my event.d script, and it stays on even if luna restarts. It seems as if only cpu clock speed is resetting.

    I read somewhere on here someone said it might be powerd.. If it's not hardcoded into luna then it is most likely a power saving daemon (such as powerd)
    nope, not powerd.

    root@palm-webos-device:/# echo 600000 > /sys/devices/system/cpu/cpu0cpufreq/scaling_setspeed
    root@palmwebos-device:/# cat /sys/devices/system/cpu/cpu0/cpufreq/scaling_cur_f
    req
    600000
    root@palm-webos-device:/# pkill powerd
    root@palm-webos-device:/# cat /sys/devices/system/cpu/cpu0/cpufreqscaling_cur_f
    req
    600000
    Live free or DIE!
  12. #112  
    Quote Originally Posted by unixpsycho View Post
    nope, not powerd.

    root@palm-webos-device:/# echo 600000 > /sys/devices/system/cpu/cpu0cpufreq/scaling_setspeed
    root@palmwebos-device:/# cat /sys/devices/system/cpu/cpu0/cpufreq/scaling_cur_f
    req
    600000
    root@palm-webos-device:/# pkill powerd
    root@palm-webos-device:/# cat /sys/devices/system/cpu/cpu0/cpufreqscaling_cur_f
    req
    600000
    what exactly are those commands supposed to show? You set the cpu to 600mhz, you make sure it's running at 600, then you kill powerd and recheck to make sure it's 600???

    honestly that is the most redundant series of commands I've seen in awhile, how about setting to to 600, killing powerd (using upstart), restarting luna AND THEN checking if it's still 600


    but I digress from the thread.
  13. #113  
    Quote Originally Posted by xstaticxgpx View Post
    what exactly are those commands supposed to show? You set the cpu to 600mhz, you make sure it's running at 600, then you kill powerd and recheck to make sure it's 600???

    honestly that is the most redundant series of commands I've seen in awhile, how about setting to to 600, killing powerd (using upstart), restarting luna AND THEN checking if it's still 600


    but I digress from the thread.
    um, i know. I do UNIX support for a living.

    I did that so other people know what is going on. If you read further back in the thread you will know why i dumbed it down.

    Since I am being picked on i'll logoff this thread.
    Live free or DIE!
  14. #114  
    Quote Originally Posted by unixpsycho View Post
    um, i know. I do UNIX support for a living.

    I did that so other people know what is going on. If you read further back in the thread you will know why i dumbed it down.

    Since I am being picked on i'll logoff this thread.
    unix support... For a living!? Damn that's a triple negative =P
  15. #115  
    Quote Originally Posted by xstaticxgpx View Post
    unix support... For a living!? Damn that's a triple negative =P
    yeah, i dont get out much.
    Live free or DIE!
  16. angiest's Avatar
    Posts
    933 Posts
    Global Posts
    952 Global Posts
    #116  
    Quote Originally Posted by a2k View Post
    Thanks for explaining. But a good question was just brought up. Does restarting JavaScript (from the button in the LunaManager application ) run startup services from the event.d directory once again?
    I'm not terribly familiar with how upstart (the init system introduced by Ubuntu, IIRC) works exactly, but I would think not. The Restart Java option simply restarts an individual process. event.d is read at boot and possibly at other times subject to the inner workings of upstart.
  17. #117  
    root@palm-webos-device: cat /sys/devices/system/cpu/cpu0/cpufreq/stats/time_in_state
    600000 0
    550000 1717
    500000 1003767
    250000 0
    125000 0
    root@palm-webos-device:

    this is what i get. How can i tell if i am running at 600mhz?

    Edit: and how is it suppos to looke like if i am running sf600
    Last edited by LittleCZR; 12/30/2009 at 11:58 AM. Reason: forgot
  18. rayln's Avatar
    Posts
    215 Posts
    Global Posts
    239 Global Posts
    #118  
    Quote Originally Posted by LittleCZR View Post
    root@palm-webos-device: cat /sys/devices/system/cpu/cpu0/cpufreq/stats/time_in_state
    600000 0
    550000 1717
    500000 1003767
    250000 0
    125000 0
    root@palm-webos-device:

    this is what i get. How can i tell if i am running at 600mhz?

    Edit: and how is it suppos to looke like if i am running sf600
    enter this in a terminal session
    Code:
    cat /sys/devices/system/cpu/cpu0/cpufreq/stats/time_in_state
    ...after any commands you enter that affect frequency.

    600000 0 <-- this one should.
    550000 1717
    500000 1003767 <-- this number (whatever it was BEFORE you affected frequency) should not increase.
    250000 0
    125000 0
  19. angiest's Avatar
    Posts
    933 Posts
    Global Posts
    952 Global Posts
    #119  
    Quote Originally Posted by LittleCZR View Post
    root@palm-webos-device: cat /sys/devices/system/cpu/cpu0/cpufreq/stats/time_in_state
    600000 0
    550000 1717
    500000 1003767
    250000 0
    125000 0
    root@palm-webos-device:

    this is what i get. How can i tell if i am running at 600mhz?

    Edit: and how is it suppos to looke like if i am running sf600
    You aren't. I assume the number next to the frequency is ticks of the CPU. You spent 1717 ticks at 550mHz and 1003767 ticks at 500mHz, but no time at 600, 250, and 125.
  20. #120  
    Quote Originally Posted by angiest View Post
    You aren't. I assume the number next to the frequency is ticks of the CPU. You spent 1717 ticks at 550mHz and 1003767 ticks at 500mHz, but no time at 600, 250, and 125.
    Ok, should i uninstall the patch now? hopefully some one fixes it.
Page 6 of 14 FirstFirst 1234567891011 ... LastLast

Posting Permissions