Problem using threads and user interface with signal/slot

ROOT
1

dear all,

I’m working with root v5.23/02.
I have an application with two threads which more or less look like this:

void *
monitorFunction(void *ptr)
{
  AliRawReaderRoot reader("/home/preghenella/ALICE.2008/data/67577/raw/09000067577018.10.root");
  while(1) {
    reader.RewindEvents();
    Int_t eventCounter = 0;
    while (reader.NextEvent()) printf("event %d\n", eventCounter++);
  }
}

Int_t
main(int argc, char *argv[])
{

  TApplication application("spyke", &argc, argv);
  AliTOFMonitor monitor;
  monitor.CreateFullMonitorScheme();
  new spykeGUI(&monitor, NULL);
  TThread *monitorThread = new TThread("monitorThread", monitorFunction);
  monitorThread->Run();
  application.Run();

  return 0;
  
}

the call spykeGUI(…) creates a user interface with buttons and canvases. the interactive behaviour of the GUI is obtained using signal/slot mechanism. pressing a button causes a set of calls to be executed and eventually a histogram is drawn on a canvas.

in summary the main thread contains the user interface, while the other thread (monitorThread) does some operations reading from a file.

what happens is that if I push one of the GUI buttons in a certain moment both threads hang.

this is the stack trace of the two threads when they are hanging:

#0  0x005db7a2 in _dl_sysinfo_int80 () from /lib/ld-linux.so.2
#1  0x013313de in __lll_mutex_lock_wait () from /lib/tls/libpthread.so.0
#2  0x0132e00b in _L_mutex_lock_35 () from /lib/tls/libpthread.so.0
#3  0x00000000 in ?? ()
#0  0x005db7a2 in _dl_sysinfo_int80 () from /lib/ld-linux.so.2
#1  0x013313de in __lll_mutex_lock_wait () from /lib/tls/libpthread.so.0
#2  0x0132e00b in _L_mutex_lock_35 () from /lib/tls/libpthread.so.0
#3  0x00000001 in ?? ()
#4  0x0738057c in ?? () from /home/alisoft/root/trunk/lib/libThread.so.5.23
#5  0xb7f2aba0 in ?? ()
#6  0x107d0ca0 in ?? ()
#7  0xb7f29fc8 in ?? ()
#8  0x07359303 in TPosixMutex::Lock ()
   from /home/alisoft/root/trunk/lib/libThread.so.5.23
#9  0x07359303 in TPosixMutex::Lock ()
   from /home/alisoft/root/trunk/lib/libThread.so.5.23
#10 0x0735465b in TMutex::Lock ()
   from /home/alisoft/root/trunk/lib/libThread.so.5.23
#11 0x007d1973 in TCint::RecursiveRemove ()
   from /home/alisoft/root/trunk/lib/libCore.so.5.23
#12 0x0078987e in THashList::RecursiveRemove ()
   from /home/alisoft/root/trunk/lib/libCore.so.5.23
#13 0x007286e7 in TObject::~TObject$base ()
   from /home/alisoft/root/trunk/lib/libCore.so.5.23
#14 0x039436d3 in ~AliRawData (this=0xb6fa88d8)
    at /home/alisoft/aliroot/pro/RAW/AliRawData.h:24
#15 0x0391b14a in ~AliRawEquipment (this=0xb6f7fe90)
    at RAW/AliRawEquipment.cxx:106
#16 0x00787718 in TCollection::GarbageCollect ()
   from /home/alisoft/root/trunk/lib/libCore.so.5.23
#17 0x0078ef73 in TObjArray::Delete ()
   from /home/alisoft/root/trunk/lib/libCore.so.5.23
#18 0x0391b9c1 in ~AliRawEvent (this=0xb6f7f9c0) at RAW/AliRawEvent.cxx:190
#19 0x00787718 in TCollection::GarbageCollect ()
   from /home/alisoft/root/trunk/lib/libCore.so.5.23
#20 0x0078ef73 in TObjArray::Delete ()
   from /home/alisoft/root/trunk/lib/libCore.so.5.23
#21 0x0391b9f4 in ~AliRawEvent (this=0xb718c480) at RAW/AliRawEvent.cxx:193
#22 0x01609e2c in AliRawReaderRoot::NextEvent (this=0xb7f2a2c0)
    at RAW/AliRawReaderRoot.cxx:528
#23 0x080495de in monitorFunction (ptr=0x0) at spyke.cxx:23
#24 0x07356fc5 in TThread::Function ()
   from /home/alisoft/root/trunk/lib/libThread.so.5.23
#25 0x0132c3cc in start_thread () from /lib/tls/libpthread.so.0
#26 0x04774b3e in clone () from /lib/tls/libc.so.6

both are in the same portion of code… and it seems that both threads are trying to lock a mutex.

does anybody have an idea of what’s going on and how can I deal with it (i.e. avoid hangs)?

thank you very much,
roberto

2

Hi Roberto,

Could you post a short example reproducing the problem, please?

Cheers,
Bertrand.