linuxdcpp-team team mailing list archive

[poy <735512@xxxxxxxxxxxxxxxxxx>]

good job, this stack trace is spot on!

it is related to an atomic variable (BufferedSocket::sockets). atomics
are only implemented by GCC (MinGW); MSVC builds reproduce this
differently, using Interlocked APIs, through boost.atomic. this explains
why this bug only manifests itself in MinGW builds.

i'm not sure however what causes this freeze.

feel free to apply the previous procedure to other bugged threads and
send the results here. the last command (thread apply [index] bt) may be
replaced by "thread apply [index] bt full" to get more information.

** Changed in: dcplusplus
   Importance: Undecided => High

** Changed in: dcplusplus
       Status: Incomplete => Confirmed

-- 
You received this bug notification because you are a member of
Dcplusplus-team, which is subscribed to DC++.
https://bugs.launchpad.net/bugs/735512

Title:
  (0.782)high single threaded cpu usage and no traffic after a period of
  time

Status in DC++:
  Confirmed

Bug description:
  the release 0.782 version of dc++ seems to stop all traffic, as well
  as take up ~15% cpu constantly on my core i7 after a few hours of
  running it, the UI still works and it is reporting my upload traffic
  at 0kb/sec, the upload progress bars are there and they just sit at
  the same percentage when the traffic stop happened...  by this point
  my downloads have either completed or have entered the 'connection
  timed out' status, so there is also no download traffic

  this problem does not occur on the r2445 0.785 build

To manage notifications about this bug go to:
https://bugs.launchpad.net/dcplusplus/+bug/735512/+subscriptions