Databases
StackOverflow2013
Postsdbo
POboost.thread dead-lock and self-deletion
 

Note that there are some explanatory texts on larger screens.

plurals
  1. POboost.thread dead-lock and self-deletion
    primarykey
    data
    text
    <p>I am using <code>boost::thread_group</code> to create(using <code>thread_group::create_thread()</code>) and dispatch threads. In order to limit the max thread numbers, at the end of each thread, I remove the thread from the <code>thread_group</code> and delete the thread itself(so that I could decide whether new threads need to be created). However it hangs somewhere between the creation and deletion of the last thread (say the 999th one of 999 in total). </p> <p>My questions are: </p> <ul> <li>is it OK to delete the thread from within itself like what I do? if not, what the best way to achieve this</li> <li>why does my code hangs?</li> </ul> <p>Below are the related code:</p> <p>//1- code to create and dispatch thread </p> <pre><code> { //mutex for map&lt;thread_id, thread*&gt; operations boost::mutex::scoped_lock lk(m_mutex_for_ptr); // create a thread for this-&gt;f(duplicate_hashes) boost::thread* p = m_thread_group.create_thread(boost::bind( &amp;detectiveT&lt;equal_predicate&gt;::f, this, duplicate_hashes )); // save the &lt;thread_id,thread pointer&gt; map for later lookup &amp; deletion m_thread_ptrs.insert(make_pair(p-&gt;get_id(), p)); // log to console for debug cout &lt;&lt; "thread created: " &lt;&lt; p-&gt;get_id() &lt;&lt; ", " &lt;&lt; m_thread_group.size() &lt;&lt; ", " m_thread_ptrs.size() &lt;&lt; "\n"; } </code></pre> <p>//2- code of the thread execution </p> <pre><code>void f(list&lt;map_iterator_type&gt;&amp; l) { Do_something(l); boost::this_thread::at_thread_exit(boost::bind( &amp;detectiveT&lt;equal_predicate&gt;::remove_this_thread, this )); } </code></pre> <p>//3- code to delete the thread itself </p> <pre><code>void remove_this_thread() { { //mutex for map&lt;thread_id, thread*&gt; operations boost::mutex::scoped_lock lk(m_mutex_for_ptr); boost::thread::id this_id(boost::this_thread::get_id()); map&lt;boost::thread::id, boost::thread*&gt;::iterator itr; itr = (m_thread_ptrs.find(this_id)); if(m_thread_ptrs.end() != itr) { // remove it from the control of thread_group m_thread_group.remove_thread(itr-&gt;second); // delete it delete itr-&gt;second; // remove from the map m_thread_ptrs.erase(this_id); // log to console for debug cout &lt;&lt; "thread erased: " &lt;&lt; this_id &lt;&lt; ", " &lt;&lt; m_thread_group.size() &lt;&lt; ", " &lt;&lt; m_thread_ptrs.size() &lt;&lt; "\n"; } } } </code></pre>
    singulars
    1. PO
      singulars
      1. PTAnswer
    1. This table or related slice is empty.
    1. PTQuestion
    1. USLightness Races in Orbit
    1. USt.g.
    plurals
    1. This table or related slice is empty.
    1. This table or related slice is empty.
    1. This table or related slice is empty.
    1. PO
      singulars
      1. PTAnswer
    1. VO
      singulars
      1. POboost.thread dead-lock and self-deletion
      1. This table or related slice is empty.
      1. VTUpMod
    2. VO
      singulars
      1. POboost.thread dead-lock and self-deletion
      1. USLightness Races in Orbit
      1. VTBountyStart
    3. VO
      singulars
      1. POboost.thread dead-lock and self-deletion
      1. USmirt
      1. VTFavorite
    1. COIf you run the program in the debugger, what is happening in the various threads when it hangs?
      singulars
      1. POboost.thread dead-lock and self-deletion
      1. USDavid Norman
    2. COfrom procexp.exe, I can see all threads are in "Wait:UserRequest" state. I suspect some recursive locks happens between the thread_group internal lock and the outer lock I introduce(i.e. lk(m_mutex_for_ptr)), but not sure how. I was once lucky by commentting out this line " << m_thread_group.size() << ", " " in void remove_this_thread(), but could not reproduce that...
      singulars
      1. POboost.thread dead-lock and self-deletion
      1. USt.g.
    3. CO@Lightness Races in Orbit: Just call `detach()` on the thread and delete its object. This will make a thread release its resources automatically upon exit (see http://www.kernel.org/doc/man-pages/online/pages/man3/pthread_create.3.html). Otherwise it will either deadlock joining itself or become a zombie because nobody will join it when it exits.
      singulars
      1. POboost.thread dead-lock and self-deletion
      1. This table or related slice is empty.
 

Querying!

 
Guidance
A row detail

Detail views are divided into sections. All the information in the data section comes from columns in the selected row. The other sections display data from other, related rows.

Related data can be related in a to-one or a to-many fashion. Captions of data related in a to-many fashion link to a list view showing a filtered view of the table.

Try moving around until you find a non-empty to-many entry and click on the label to get to one. You can move back to the root by clicking on the database name in the header.

SQuiL has stopped working due to an internal error.

If you are curious you may find further information in the browser console, which is accessible through the devtools (F12).

Reload