What is the purpose of the countdownlatch? and a crash of my receiver

[jeffrey]

I have a receiver class, but I never used the countdownlatch objects that are
listed in this example program. http://activemq.apache.org/cms/example.html
Im wondering what the latch actually does and how will not using it affect
my program. Since upgrading to 3.5.0 my receiver now core dumps after the
return line at the end of my program.

#0 0x60000000d341e6a0:1 in apr_pvsprintf () at memory/unix/apr_pools.c:1176
#1 0x60000000e2906ed0:0 in decaf::lang::Exception::buildMessage (
this=0x2000000040089370,
format=0x60000000e1005530 "Failed to Lock OS Mutex", vargs=@0x7fffef00)
at decaf/lang/Exception.cpp:124
#2 0x60000000e299b5b0:0 in
decaf::lang::exceptions::RuntimeException::RuntimeException
(this=0x2000000040089370,
file=0x60000000e10054c0
"decaf/internal/util/concurrent/unix/PlatformThread.cpp", lineNumber=77,
msg=0x60000000e1005530 "Failed to Lock OS Mutex")
at decaf/lang/exceptions/RuntimeException.cpp:51
#3 0x60000000e2878160:0 in
decaf::internal::util::concurrent::PlatformThread::lockMutex (mutex=0x0)
at decaf/internal/util/concurrent/unix/PlatformThread.cpp:76
#4 0x60000000e286a5b0:0 in
decaf::internal::util::concurrent::Threading::attachToCurrentThread () at
decaf/internal/util/concurrent/Threading.cpp:927
#5 0x60000000e2863350:0 in
decaf::internal::util::concurrent::Threading::getCurrentThreadHandle () at
decaf/internal/util/concurrent/Threading.cpp:1273
#6 0x60000000e2867a40:0 in
decaf::internal::util::concurrent::Threading::join
(thread=0x2000000040092110, mills=0, nanos=0)
at decaf/internal/util/concurrent/Threading.cpp:1069
#7 0x60000000e2869af0:0 in
decaf::internal::util::concurrent::Threading::destroyThread
(thread=0x2000000040092110)
at decaf/internal/util/concurrent/Threading.cpp:893
#8 0x60000000e2985300:0 in decaf::lang::Thread::~Thread (
this=0x200000007fffeff0, _noname=0) at decaf/lang/Thread.cpp:130
#9 0x400bc90:0 in main () at cleanamqsms.cpp:373

It goes to delete the thread started for the listener and crashes along the
way, but this does not happen on my sender class that invokes a thread in
the exact same manor. Im wondering if that latch has anything to do with it
and also why it only started in 3.5.0 and did not happen in 3.4.5. I
already have this fix in my library
https://issues.apache.org/jira/browse/AMQCPP-449.

Thanks in advance.



--
View this message in context: http://activemq.2283324.n4.nabble.com/What-is-the-purpose-of-the-countdownlatch-and-a-crash-of-my-receiver-tp4662651.html
Sent from the ActiveMQ - User mailing list archive at Nabble.com.

[Timothy Bish]

Post by jeffrey
I have a receiver class, but I never used the countdownlatch objects that are
listed in this example program. http://activemq.apache.org/cms/example.html
Im wondering what the latch actually does and how will not using it affect
my program. Since upgrading to 3.5.0 my receiver now core dumps after the
return line at the end of my program.

It does what its named for, its provides a latch that will wait until
its count hits zero.

Not sure what's going on, recommend that you provide the code that
causes the error.

Post by jeffrey
#0 0x60000000d341e6a0:1 in apr_pvsprintf () at memory/unix/apr_pools.c:1176
#1 0x60000000e2906ed0:0 in decaf::lang::Exception::buildMessage (
this=0x2000000040089370,
at decaf/lang/Exception.cpp:124
#2 0x60000000e299b5b0:0 in
decaf::lang::exceptions::RuntimeException::RuntimeException
(this=0x2000000040089370,
file=0x60000000e10054c0
"decaf/internal/util/concurrent/unix/PlatformThread.cpp", lineNumber=77,
msg=0x60000000e1005530 "Failed to Lock OS Mutex")
at decaf/lang/exceptions/RuntimeException.cpp:51
#3 0x60000000e2878160:0 in
decaf::internal::util::concurrent::PlatformThread::lockMutex (mutex=0x0)
at decaf/internal/util/concurrent/unix/PlatformThread.cpp:76
#4 0x60000000e286a5b0:0 in
decaf::internal::util::concurrent::Threading::attachToCurrentThread () at
decaf/internal/util/concurrent/Threading.cpp:927
#5 0x60000000e2863350:0 in
decaf::internal::util::concurrent::Threading::getCurrentThreadHandle () at
decaf/internal/util/concurrent/Threading.cpp:1273
#6 0x60000000e2867a40:0 in
decaf::internal::util::concurrent::Threading::join
(thread=0x2000000040092110, mills=0, nanos=0)
at decaf/internal/util/concurrent/Threading.cpp:1069
#7 0x60000000e2869af0:0 in
decaf::internal::util::concurrent::Threading::destroyThread
(thread=0x2000000040092110)
at decaf/internal/util/concurrent/Threading.cpp:893
#8 0x60000000e2985300:0 in decaf::lang::Thread::~Thread (
this=0x200000007fffeff0, _noname=0) at decaf/lang/Thread.cpp:130
#9 0x400bc90:0 in main () at cleanamqsms.cpp:373
It goes to delete the thread started for the listener and crashes along the
way, but this does not happen on my sender class that invokes a thread in
the exact same manor. Im wondering if that latch has anything to do with it
and also why it only started in 3.5.0 and did not happen in 3.4.5. I
already have this fix in my library
https://issues.apache.org/jira/browse/AMQCPP-449.
Thanks in advance.
--
View this message in context: http://activemq.2283324.n4.nabble.com/What-is-the-purpose-of-the-countdownlatch-and-a-crash-of-my-receiver-tp4662651.html
Sent from the ActiveMQ - User mailing list archive at Nabble.com.

[jeffrey]

The program is broken down to simply

activemq::library::ActiveMQCPP::initializeLibrary();
HBconsumer *hbc = new HBconsumer(brokerURI, destURI, 3); //This is my
receiver class it implements public
ExceptionListener, public MessageListener, public Runnable
Thread HBthread(hbc); //
HBthread.start(); // If I take out these three lines and do not create
a thread, my program does nothing, but does not crash
when exiting
HBthread.join(); //
sleep(3); //
delete hbc;
activemq::library::ActiveMQCPP::shutdownLibrary(); // If I take out this
line and leave in the others, then it works fine
return 0;

I am still on HPUX11.31 btw, Ill get more information when I have time



--
View this message in context: http://activemq.2283324.n4.nabble.com/What-is-the-purpose-of-the-countdownlatch-and-a-crash-of-my-receiver-tp4662651p4662660.html
Sent from the ActiveMQ - User mailing list archive at Nabble.com.

[Timothy Bish]

Post by jeffrey
The program is broken down to simply
activemq::library::ActiveMQCPP::initializeLibrary();
HBconsumer *hbc = new HBconsumer(brokerURI, destURI, 3); //This is my
receiver class it implements public
ExceptionListener, public MessageListener, public Runnable
Thread HBthread(hbc); //
HBthread.start(); // If I take out these three lines and do not create
a thread, my program does nothing, but does not crash
when exiting
HBthread.join(); //
sleep(3); //
delete hbc;
activemq::library::ActiveMQCPP::shutdownLibrary(); // If I take out this
line and leave in the others, then it works fine
return 0;
I am still on HPUX11.31 btw, Ill get more information when I have time

Please create a new Jira and attach a complete source sample that
demonstrates the issue otherwise I can't test it.

Post by jeffrey
--
View this message in context: http://activemq.2283324.n4.nabble.com/What-is-the-purpose-of-the-countdownlatch-and-a-crash-of-my-receiver-tp4662651p4662660.html
Sent from the ActiveMQ - User mailing list archive at Nabble.com.

[jeffrey]

jire created
https://issues.apache.org/jira/browse/AMQCPP-452



--
View this message in context: http://activemq.2283324.n4.nabble.com/What-is-the-purpose-of-the-countdownlatch-and-a-crash-of-my-receiver-tp4662651p4662733.html
Sent from the ActiveMQ - User mailing list archive at Nabble.com.