[Qt-interest] QMutex deadlock without using QMutex

[Gabriel M. Beddingfield]

On Tue, 27 Apr 2010, Andre Somers wrote:

>> Perhaps switch from Qt::QueuedConnection to
>> Qt::BlockingQueuedConnection (or the reverse).  You might get lucky.
> Ehh? Swithing to a blocking connection would defeat the purpose of using
[snip]
> think... Or do I totally misunderstand the documentation here? If I try
> it, I get up to about 25% CPU usage on my quad core system. That's
> single threaded performance.

But it didn't deadlock, right?

Ok, scratch that one.  :-)

You said that the worker threads signal the coordinator 
thread.  In response to this signal, the coordinator thread 
signals the workers.  Right?

What happens when the workers respond faster than the 
coordinator thread expects?  I.e. while it's still 
processing the signal from the worker thread (and is 
dispatching the next workloa)d... it gets another signal 
from the same worker thread.

hth,
Gabriel