← Back to team overview

kernel-packages team mailing list archive

[Bug 1231585] Re: ubuntuappmanager has a thread that is polling at 2Hz

 

[Expired for linux (Ubuntu) because there has been no activity for 60
days.]

** Changed in: linux (Ubuntu)
       Status: Incomplete => Expired

-- 
You received this bug notification because you are a member of Kernel
Packages, which is subscribed to linux in Ubuntu.
https://bugs.launchpad.net/bugs/1231585

Title:
  ubuntuappmanager has a thread that is polling at 2Hz

Status in “linux” package in Ubuntu:
  Expired

Bug description:
  Using today's phablet image on a Samsung Galaxy Nexus I observed that
  one of the threads of ubuntuappmanager is polling every 500ms.  I am
  filing this bug in a vain attempt to try and reduce system wakeups and
  hence reduce battery consumption.

  Tracing this with strace, I see:

  sudo strace -f -p $(pidof ubuntuappmanager)
  Process 477 attached with 8 threads
  [pid  1238] ioctl(13, BT819_FIFO_RESET_HIGH <unfinished ...>
  [pid   602] futex(0x412b4cac, FUTEX_WAKE_PRIVATE, 2147483647 <unfinished ...>
  [pid   601] epoll_wait(20,  <unfinished ...>
  [pid   602] <... futex resumed> )       = 0
  [pid   477] ioctl(13, BT819_FIFO_RESET_HIGH <unfinished ...>
  [pid   602] clock_gettime(CLOCK_MONOTONIC,  <unfinished ...>
  [pid  1239] ioctl(13, BT819_FIFO_RESET_HIGH <unfinished ...>
  [pid   606] ioctl(13, BT819_FIFO_RESET_HIGH <unfinished ...>
  [pid   604] clock_gettime(CLOCK_MONOTONIC,  <unfinished ...>
  [pid   603] write(6, "KeyEvents", 9 <unfinished ...>
  [pid   604] <... clock_gettime resumed> {17918, 745990561}) = 0
  [pid   603] <... write resumed> )       = 9
  [pid   604] epoll_wait(5,  <unfinished ...>
  [pid   603] clock_gettime(CLOCK_MONOTONIC,  <unfinished ...>
  [pid   602] <... clock_gettime resumed> {17918, 745624350}) = 0
  [pid   603] <... clock_gettime resumed> {17918, 746326255}) = 0
  [pid   603] write(7, "KeyEvents", 9 <unfinished ...>
  [pid   602] clock_gettime(CLOCK_MONOTONIC,  <unfinished ...>
  [pid   603] <... write resumed> )       = 9
  [pid   602] <... clock_gettime resumed> {17918, 746600913}) = 0
  [pid   603] epoll_wait(9,  <unfinished ...>
  [pid   602] epoll_wait(16,  <unfinished ...>
  [pid   604] <... epoll_wait resumed> {}, 16, 500) = 0
  [pid   604] clock_gettime(CLOCK_MONOTONIC, {17919, 248248861}) = 0
  [pid   604] epoll_wait(5, {}, 16, 500)  = 0
  [pid   604] clock_gettime(CLOCK_MONOTONIC, {17919, 749866293}) = 0
  [pid   604] epoll_wait(5, {}, 16, 500)  = 0
  [pid   604] clock_gettime(CLOCK_MONOTONIC, {17920, 409625813}) = 0
  [pid   604] epoll_wait(5, {}, 16, 500)  = 0
  [pid   604] clock_gettime(CLOCK_MONOTONIC, {17920, 911945151}) = 0

  where pid 604 epoll_waits on fd 5, which is an event poll fd and is
  timing out every 500ms.  Perhaps this blocking epoll_wait could wait
  for longer rather than continuously waking up the system every 2
  seconds?

To manage notifications about this bug go to:
https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1231585/+subscriptions