Looks like it's because you're not calling ros::shutdown() before you unload
the plugin.

Josh

On Sun, Sep 12, 2010 at 5:59 AM, Ruben Smits
<ruben.smits@mech.kuleuven.be>wrote:

> On Saturday 11 September 2010 20:03:48 Josh Faust wrote:
> > init() shouldn't be blocking -- if it does it's a bug.  It's designed
> this
> > way so you can call ros::master::check().  Creation of the first
> > NodeHandle or calling start() should be what's blocking.
>
> Ok, sorry for the mistake, it was indeed ros::start() that blocked my code.
> I
> changed it into the following:
>
> bool loadRTTPlugin(RTT::TaskContext* c){
>    log(Info)<<"Initializing ROS node"<<endlog();
>    if(!ros::isInitialized()){
>      int argc=0;
>      char* argv[0];
>      ros::init(argc,argv,"rtt",ros::init_options::AnonymousName);
>      if(ros::master::check())
>          ros::start();
>      else{
>          log(Error)<<"No ros::master available"<<endlog();
>          return false;
>      }
>    }
>    static ros::AsyncSpinner spinner(1); // Use 1 threads
>    spinner.start();
>    log(Info)<<"ROS node spinner started"<<endlog();
>    return true;
>  }
>
> But now I get the the following assertion:
>
> 0.719 [ ERROR  ][DeploymentComponent::import] No ros::master available
> 0.719 [ ERROR  ][DeploymentComponent::import] Failed to load RTT Plugin
> 'ros_integration': plugin refused to load into this process. Unloading.
> deployer-gnulinux: /usr/include/boost/thread/pthread/mutex.hpp:50: void
> boost::mutex::lock(): Assertion `!pthread_mutex_lock(&m)' failed.
>
> Program received signal SIGABRT, Aborted.
> 0x00007ffff4caca75 in *__GI_raise (sig=<value optimized out>)
>    at ../nptl/sysdeps/unix/sysv/linux/raise.c:64
> 64      ../nptl/sysdeps/unix/sysv/linux/raise.c: No such file or directory.
>        in ../nptl/sysdeps/unix/sysv/linux/raise.c
> (gdb) bt
> #0  0x00007ffff4caca75 in *__GI_raise (sig=<value optimized out>)
>    at ../nptl/sysdeps/unix/sysv/linux/raise.c:64
> #1  0x00007ffff4cb05c0 in *__GI_abort () at abort.c:92
> #2  0x00007ffff4ca5941 in *__GI___assert_fail (
>    assertion=0x7ffff124f7e5 "!pthread_mutex_lock(&m)",
>    file=<value optimized out>, line=50,
>    function=0x7ffff124f9a0 "void boost::mutex::lock()") at assert.c:81
> #3  0x00007ffff12453db in boost::mutex::lock (loc=0x7ffff1beea20,
>    name=..., level=ros::console::levels::Debug)
>    at /usr/include/boost/thread/pthread/mutex.hpp:50
> #4  boost::unique_lock<boost::mutex>::lock (loc=0x7ffff1beea20, name=...,
>    level=ros::console::levels::Debug)
>    at /usr/include/boost/thread/locks.hpp:338
> #5  unique_lock (loc=0x7ffff1beea20, name=...,
>    level=ros::console::levels::Debug)
>    at /usr/include/boost/thread/locks.hpp:224
> #6  ros::console::initializeLogLocation (loc=0x7ffff1beea20, name=...,
>    level=ros::console::levels::Debug)
>    at
> /opt/ros/cturtle/ros/core/rosconsole/src/rosconsole/rosconsole.cpp:6---
> Type <return> to continue, or q <return> to quit---
> 10
> #7  0x00007ffff1923e4a in ros::atexitCallback ()
>    at /opt/ros/cturtle/ros/core/roscpp/src/libros/init.cpp:149
> #8  0x00007ffff4cb2630 in __cxa_finalize (d=0x7ffff1bee000)
>    at cxa_finalize.c:56
> #9  0x00007ffff18f23d6 in __do_global_dtors_aux ()
>   from /opt/ros/cturtle/ros/core/roscpp/lib/libros.so
> #10 0x0000000000000015 in ?? ()
> #11 0x00007fffffffc8d0 in ?? ()
> #12 0x00007ffff19abc81 in _fini ()
>   from /opt/ros/cturtle/ros/core/roscpp/lib/libros.so
> #13 0x0000000000000032 in ?? ()
> #14 0x00007ffff7df0972 in _dl_close_worker (map=<value optimized out>)
>    at dl-close.c:271
> #15 0x0000000000680f80 in ?? ()
>
> Another question:
>
> Would it be possible to start my own master if none is available? Can it be
> done using ros::master::init()?
>
> Ruben
> > Josh
> >
> > On Sat, Sep 11, 2010 at 7:53 AM, Ruben Smits
> > <ruben.smits@mech.kuleuven.be<mailto:ruben.smits@mech.kuleuven.be>>
> wrote:
> > Hi,
> >
> > I would like to know if there is a way to call ros::init and not block
> > untill a ros::master is available but just return an error or the like if
> > a ros::master is not available so that my code can continue.
> >
> > Ruben
> > _______________________________________________
> > ros-users mailing list
> > ros-users@code.ros.org<mailto:ros-users@code.ros.org>
> > https://code.ros.org/mailman/listinfo/ros-users
> _______________________________________________
> ros-users mailing list
> ros-users@code.ros.org
> https://code.ros.org/mailman/listinfo/ros-users
>