Hi, Shashank, I also met such a defect when I run CTS for android-4.0.3_r1 on some devices.
A test engineer report me that randomly some case will fail for adb timeout. After some investigate, I found that the child process forked by adbd is not return from fork(). I checked the child's call stack, it is suspend/blocked/sleeping on the futex_wait_queue_me(). Also I found the child process kept this file as open: /acct/uid/0/tasks. Someone other also found this issue and reported at here: https://groups.google.com/forum/?fromgroups#!topic/android-platform/Gvw5X7g3dLg 在 2011年12月30日星期五UTC+8上午2时54分07秒，Shashank写道： > > We are facing some random issues with futex while porting ICS on our > devices. > > Issue description: > --------------------- > > We are seeing random futex deadlock issues on ICS with different > processes on Android. > > One of the example is, we run a perl script that install/uninstall > multiple apks on device using adb. As install process adb daemon > (adbd) on the device forks another child process to handle this new > install request. In the failure case I saw that this new adbd child > process was sleeping in the futex queue, because of which adb request > on host was just waiting for reply back. This scenario is very random. > > In another case we saw dexpot getting stuck randomly during boot up > while running a malloc call. From stack trace it again turned out it > was sleeping in the futex queue (malloc -> dlmalloc -> futex). > > > Questions: > -------------- > > 1) Has anyone seen such issue? > 2) How folks at Google generally debug pthead/futex code in libc > library under bionic? > > Thanks! > --Shashank > -- unsubscribe: [email protected] website: http://groups.google.com/group/android-porting
