Permalink
Commits on Jul 19, 2014
Commits on Jul 7, 2014
  1. Fix module name

    rickysarraf committed Jul 7, 2014
Commits on Jul 6, 2014
Commits on Jun 20, 2014
  1. Merge branch 'juan' into lmt-upstream

    rickysarraf committed Jun 20, 2014
    * juan:
      GUI refactor
  2. GUI refactor

    juan committed Jun 20, 2014
Commits on Jun 18, 2014
  1. Prepare for release 1.65

    rickysarraf committed Jun 18, 2014
Commits on Jun 12, 2014
  1. Merge pull request #33 from shuber2/missing-uevent

    rickysarraf committed Jun 12, 2014
    modules: fix grep error on missing $device/uevent
    
    
    Looks good. THank you.
  2. modules: fix grep error on missing $device/uevent

    shuber2 committed Jun 10, 2014
    Issue #33: A race at boot time may cause the removal of an USB device by
    the kernel between enumerating all devices and individually accessing
    sysfs uevent files. As a consequence, grep reports errors on missing
    files during boot. Fix this.
Commits on Jun 1, 2014
Commits on May 27, 2014
  1. Merge pull request #31 from shuber2/formatting

    rickysarraf committed May 27, 2014
    Formatting
    
    Thank you Stefan. I am merging it with the assumption that you already are running this formatted code on your box.
  2. LMT.py: fix indentation

    shuber2 committed May 27, 2014
  3. Merge pull request #30 from shuber2/ethernet-poweroff

    rickysarraf committed May 27, 2014
    Ethernet poweroff
Commits on May 25, 2014
  1. ethernet: replace sysfs/enabled by 'ip link down'

    shuber2 committed May 25, 2014
    Disabeling the ethernet pci device via "echo 0" to sysfs attr "enabled"
    does not necessarily power down the device but rather decrements the
    enabled counter. Even if the counter reaches 0 the device may still
    drain power.
    Shutting down the ethernet device via 'ip link set dev DEV down',
    however, reduces power consumptions to 0 mW as indicated by powertop.
  2. conf: fix typos

    shuber2 committed May 25, 2014
Commits on May 24, 2014
  1. Merge pull request #28 from shuber2/whitespace

    rickysarraf committed May 24, 2014
    Whitespace
    
    Thank you for this patch.
  2. modules: no expandtab indent

    shuber2 committed May 24, 2014
  3. Merge pull request #27 from alonbl/pci

    rickysarraf committed May 24, 2014
    pci: support both enable/enabled
  4. pci: support both enable/enabled

    alonbl committed May 24, 2014
    Starting with linux kernel 3.13, the sysfs attribute 'enable' for pci
    devices was renambed to 'enabled'.
    
    Support the two notations to allow seamless migration.
    
    Signed-off-by: Alon Bar-Lev <alon.barlev@gmail.com>
Commits on May 23, 2014
  1. Merge pull request #22 from shuber2/lmt-upstream

    rickysarraf committed May 23, 2014
    Add iwlwifi support, add support for linux kernel >=3.13
    
    @shuber2 Thank you. Looks good and merged.
Commits on May 22, 2014
  1. wireless-iwl-power: sysfs attr enbable -> enabled

    shuber2 committed May 22, 2014
    Starting with linux kernel 3.13, the sysfs attribute 'enable' for pci
    devices was renambed to 'enabled'.
Commits on Apr 17, 2014
  1. Fix path for enable / disable ethernet device

    rickysarraf committed Apr 17, 2014
    + [ x1 = x1 ]
    + [ -x /usr/bin/logger -a VERBOSE != STATUS ]
    + [ VERBOSE = MSG ]
    + [ VERBOSE = ERR ]
    + [ VERBOSE = VERBOSE ]
    + [ x0 = x1 ]
    + [ VERBOSE = VERBOSE ]
    + [ 1 = 0 ]
    + [ -z /sys/devices/pci0000:00/0000:00:19.0 ]
    + [ -f /sys/devices/pci0000:00/0000:00:19.0/enabled ]
    + [ x1 = x1 ]
    + [ -f /sys/devices/pci0000:00/0000:00:19.0/enabled ]
    + echo 0
    sh: echo: I/O error
    + log VERBOSE ethernet: Disabling ethernet device eth0
    + [ x1 = x1 ]
    + [ -x /usr/bin/logger -a VERBOSE != STATUS ]
    + [ VERBOSE = MSG ]
    + [ VERBOSE = ERR ]
    
    For some reason, I/O error is seen when writing to the file.
    This seems to be a cosmetic timing issue because irrespective of the
    error, the values are written to the file, and changes are effective.
    
    It could very well be a timing issue because the ethernet device is a
    PCI device too, which is also handle in Runtime PM
Commits on Mar 6, 2014
  1. Merge pull request #18 from sanya-m/lmt-upstream

    rickysarraf committed Mar 6, 2014
    Revert defaults for cpufreq, add module for intel_pstate
    
    Thank you for your contribution.
  2. Merge pull request #20 from schouleu/loop-hibernate

    rickysarraf committed Mar 6, 2014
    Fix multiple hibernate issue.
    
    Nice catch. Thank you for the analysis and the patch.
Commits on Mar 5, 2014
  1. Fix multiple hibernate issue

    schouleu committed with Sylvain Chouleur Mar 5, 2014
    Reload config when locks are taken
    
    Issue description:
    We are running out of battery, two laptop-mode
    instances are run concurrently, they call lmt_load_config() wich set
    ON_AC=0.
    The first one take locks and run lmt_main_function while the seconds
    waits for locks in lock_retry().
    The first one calls hibernate.
    The laptop is waken with AC plugged.
    The first instance exits and release locks
    The second one takes locks and run lmt_main_function but still have
    ON_AC=0 due to lmt_load_config() before the hibernation
    The second one calls hibernate whereas we are on AC and policy says do
    not hibernate on AC.
Commits on Jan 14, 2014
  1. When resuming, run LMT in force initialization mode

    rickysarraf committed Jan 14, 2014
    Because, during during, at times when LMT is invoked, the power
    subsystem might not be fully initialized. This leads to LMT getting
    invalid states.
    
    Scenario: User suspends on AC, and resumes on BATTERY. LMT gets the
    false state.
    
    Thanks: Andrey Vihrov
Commits on Dec 27, 2013
  1. Fix copy-paste mistake

    amezin committed Dec 27, 2013
Commits on Dec 26, 2013