I've got a class that's using ASIO to run a simple server listening for 
command messages from a remote interface.<div><br></div><div>It&#39;s got 
some private members to help with this (class abridged for shortness):</div>
<div><br></div><div>class ClassifierContext {</div><div>private:</div><div>    io_service
      ioService_;</div><div>    t_socket_shptr  socket_;</div><div>};</div><div><br></div><div>It
 runs it&#39;s event loop in it&#39;s own thread (started with boost::thread)</div>
<div><br></div><div><div>// Hosts thread to run IO event loop</div><div>void
 ClassifierContext::io_loop() {</div><div>  try {</div><div>   
 connect(); </div><div>  } catch (std::exception &amp;e) {</div><div>   
 RERROR(&quot;Error connecting to site server -- %s&quot;, e.what());</div>
<div>  }</div><div><br></div><div>  // Work object prevents run() from 
stopping when it runs out of work</div><div>  asio::io_service::work 
work(ioService_);</div><div>  ioService_.run();</div><div>}</div><div><br></div><div>
And it periodically sends data back along a single socket associated 
with the ioService and also has</div><div>an asynchronous read pending 
to listen for command messages (I can post this code if needed, it&#39;s 
just a couple async_sends and an async_read)</div>
<div><br></div><div>The problem I&#39;m seeing comes when the destructor for
 the class is run:</div><div><br></div><div><div>ClassifierContext::~ClassifierContext()
 {</div><div>  ioService_.stop();</div><div>}</div><div><br>
</div><div>Periodically (every 20 times or so) I get a segfault when the
 destructor for ioService_ is called.  A representative stack trace 
looks like this:</div><div><div><br></div><div>#0  0x0921bfd6 in ?? ()</div><div>
#1  0x08125877 in boost::asio::detail::reactor_op_queue&lt;int&gt;::destroy_operations
 (</div><div>    this=0x8ec5f40)</div><div>    at 
opt/linux/include/boost/asio/detail/reactor_op_queue.hpp:268</div><div>#2
  0x081274ad in boost::asio::detail::epoll_reactor&lt;false&gt;::shutdown_service
 (</div>
<div>    this=0x8ec5ed8)</div><div>    at opt/linux/include/boost/asio/detail/epoll_reactor.hpp:119</div><div>#3
  0x08127b3b in ~service_registry (this=0x8ec5958)</div><div>    at 
opt/linux/include/boost/asio/detail/service_registry.hpp:75</div>
<div>#4  0x08127c29 in ~io_service (this=0x8492360)</div><div>    at 
opt/linux/include/boost/asio/impl/io_service.ipp:62</div><div>#5 
 0x08117110 in ~ClassifierContext (this=0x8492360)</div><div>    at 
ClassifierContext.cpp:91</div>
<div>#6  0xb781cbb9 in exit () from /lib/tls/i686/cmov/libc.so.6</div><div>#7
  0xb780477d in __libc_start_main () from /lib/tls/i686/cmov/libc.so.6</div><div>#8
  0x081079c1 in _start () at ../sysdeps/i386/elf/start.S:119</div>
<div><br></div></div><div>Has anyone seen a segfault like this before? 
 It&#39;s intermittent so that tells me there&#39;s some kind of race condition.
  Any idea how I can work around it?</div></div></div>