We Keep Coding

Monitor.Pulse in C# appears suboptimal : must be in lock scope

spoiler note: the question is the last phrase.
In C#, the classical pattern to use a condition variable is like this:
lock (answersQueue)
{
answersQueue.Enqueue(c);
Monitor.Pulse(answersQueue); // condition variable "notify one".
}
and some other thread:
lock (answersQueue)
{
while (answersQueue.Count == 0)
{
// unlock answer queue and sleeps here until notified.
Monitor.Wait(answersQueue);
}
...
}
that's an example taken from my code.
if I place the Pulse outside of the lock scope, it doesn't compile.
however, it is the correct way:
c.f:
http://msdn.microsoft.com/en-us/library/windows/desktop/ms686903(v=vs.85).aspx
and:
http://www.installsetupconfig.com/win32programming/threadprocesssynchronizationapis11_7.html
(search for "inside")
And indeed it is idiotic to signal the sleeping thread when you still are in your critical section. Because the sleeping thread CAN'T wake up (not immediately), BECAUSE it is INSIDE a criticial section as well !
Therefore, I hope that .NET or C# Pulse call is actually just flagging the lock object, so that when it goes out of scope it actually "pulses" the condition variable at this moment. Because otherwise, it would have an optimality issue.
So how come the design of the Monitor object was chosen to be that way ?
Edit:
I found the answer in this paper:
http://research.microsoft.com/pubs/64242/implementingcvs.pdf
section "Optimising Signal and Broadcast" and the previous section about NT kernel and how to make Condition Variable on top of Semaphores, which is the reason for introducing the "darned queues".
NOW that makes me a better engineer.

And indeed it is idiotic to signal the sleeping thread when you still are in your critical section. Because the sleeping thread CAN'T wake up
Pulse doesn't expect to get a thread running; it only expects to move a thread between the 2 queues (waiting and ready). The "not go do something" is part of releasing the lock via Exit (or the end of a lock). In reality, it isn't an issue because Monitor.Pulse typically happens right before a Wait or an Exit.
Therefore, I hope that .NET or C# Pulse call is actually just flagging the lock object, so that when it goes out of scope it actually "pulses" the condition variable at this moment. Because otherwise, it would have an optimality issue.
Again; these are different issues: moving between waiting and ready is one thing; exiting a lock already has all the code to actually activate the next ready thread.

You did not understood the basic problem of synchronization. What is a 'monitor', what does it mean that a thread sleeps and what does it mean that it is about to be woken up?
A monitor is a mid-level synchronization structure. This is not a low-level petty volatile boolean flag with bus-halting XCHG operation, and this is not high-level thread pool handler that requires dozens of other special mechanisms..
On a monitor, MANY threads may sleep. There are logical queues out there that i.e. preserver order of being put to sleep/woken up, or mechanisms that guarantee proper time scheduling and fairnees. I will not get into details, all of it is out there on the web, even on wiki.
Add to that that the operation is PULSE. Pulse is instantenous. It does not "stick". Pulse will wake those now sleeping. If after the pulse another one check the monitor, it will go to sleep.
Now imagine: you have a queue of 5 sleeping threads. One thread (6th) wants now to pulse them, and yet another (7th) wants to check the monitor.
6th and 7th are running in parallel, truly simultaneously, since you have quad-core CPU.
So, tell me, what would happen to the queue's implementtion if the 6th starts pulsing andwaking and removing woken threads from the queue, and in the same time the 7th one starts adding itself there?
To solve that, the internal queues would have to be internally synchronized and locked, so only one thread at time modifies them.
Um wait. We just stumbled upon a case where we wanted to SYNCHRONIZE something, and to do it properly we need to SYNCHRONIZE on another thing? Not good.
Therefore, the actual LOCK is done EXTERNALLY before you talk to the monitor itself. This is to achieve SINGLE LOCKING, instead of introduce several layers of hierarchical locks.
That way it is simplier, faster, and more resource-friendly.

Terminate loopless thread instantly without Abort or Suspend

I am implementing a protocol library. Here a simplified description.
The main thread within the main function will always check, whether some data is available on the the networkstream (within a tcpclient). Let us say response is the received message and thread is a running thread.
thread = new Thread(new ThreadStart(function));
thread.IsBackground = true;
thread.Start();
while(true){
response = receiveMessage();
if (response != null)
{
thread.Suspend();
//I am searching for an alternative for the line above and not thread.Abort().
thread2 = new Thread(new ThreadStart(function2));
thread2.IsBackground = true;
thread2.Start();
}
}
So far so good, there are actually more messages to come within the while loop and there is also a statemachine for handling different sort of incoming messages, but this should be enough.
(There are also more than just the functions "function" and "function2").
So anyways how the functions look inside is not clear in this application, since the protocol is hidden from the programmer and meant to be a library. This means the protocol will start some programmer-defined functions as a thread depending on at what state in the protocol the program is.
So if then a special response is received (e.g. a callAnotherFunction message), I want to terminate
a thread (here named "thread") abruptly, lets say within 100 ms. But I do not know whether it executes within a loop or without and how much processing is needed until it terminates.
How to stop these threads without deprecated Suspend or Exceptionthrowing Abort function?
(Note that I cannot force the programmer of the functions to catch the ThreadAbortException.)
Or do I need a different programme architecture?
(Btw I have decided to put the loop within receiveMessage for polling the network stream into the main function, since anytime a message can appear).

Starting a thread without having a reliable way to terminate it is a bad practice. Suspend/Abort are one of those unreliable ways to terminate a thread because you may terminate a thread in a state that corrupts your entire program and you have no way to avoid it from happening.
You can see how to kill a thread safely here: Killing a .NET thread
If the "user" is giving you a method to run in a thread, then the user should also give you a method to stop the code from running. Think of it as a contract: you promise the user that you will call the stop method and they promise that the stop method will actually stop the thread. If your user violates that contract then they will be responsible for the issues that arise, which is good because you don't want to be responsible for your user's errors :).
Note that I cannot force the programmer of the functions to catch the ThreadAbortException.
Since Suspend/Abort are bad practice, the programmer doesn't need to catch the ThreadAbortException, however they should catch the ThreadInterruptedException as part of their "contract."
Remember that there are two situations you need to worry about:
The thread is executing some code.
The thread is in a blocking state.
In the case that the thread is executing some code, all you can do is notify the thread that it can exit and wait until it processes the notification. You may also skip the waiting and assume that you've leaked a resource, in which case it's the user's fault again because they didn't design their stop method to terminate their thread in a timely fashion.
In the case where the thread is in a blocking state and it's not blocking on a notification construct (i.e. semaphore, manual reset event, etc) then you should call Thread.Interrupt() to get it out of the blocking state- the user must handle the ThreadInterruptedException.

Suspend is really evil especially in a way you are trying to use it - to stop thread execution forever. It will leave all locks that thread had and also will not release resources.
Thread Abort is slightly better since it will at least try to terminate thread cleaner and locks will have chance to be released.
To properly do that you really need your thread's code to cooperate in termination. Events, semaphores or even simple bool value checked by the thread may be enough.
It may be better to re-architect your solution to have queue of messages and process them on separate thread. Special message may simply empty the queue.

You need some sort of cancellation protocol between your application and wherever function comes from. Then you can share some sort of cancellation token between function and your message loop. If message loop recognizes that function needs to be stopped you signal that by setting that token which must be tested by function on proper occasions. The simplest way would be to share a condition variable which can be atomically set from within your message loop and atomically read from function.
I'd however consider using the proper Asynchronous IO patterns combined with Tasks provided by the .NET framework out-of-the box along with proper cancellation mechanisms.

So function refers to code which you have little control over? This is pretty typical of 3rd party libraries. Most of the time they do not have builtin abilities to gracefully terminate long running operations. Since you have no idea how these functions are implemented you have very few options. In fact, your only guaranteed safe option is to spin these operations up in their own process and communicate with them via WCF. That way if you need to terminate the operation abruptly you would just kill the process. Killing another process will not corrupt the state of the current process like what would happen if you called Thread.Abort on thread within the current process.

Thread: How to re-start thread once completed?

I have a method void DoWork(object input) that takes roughly 5 seconds to complete. I have read that Thread is better suited than ThreadPool for these longer operations but I have encountered a problem.
I click a button which calls threadRun.Start(input) which runs and completes fine. I click the button again and receive the following exception:
Thread is running or terminated; it cannot restart.
Can you not "reuse" a Thread? Should I use ThreadPool? Why is Thread "better suited for longer operations" compared to ThreadPool? If you can't reuse a thread, why use it at all (i.e. what advantages does it offer)?

Can you not "reuse" a Thread?
You can. But you have to code the thread not to terminate but to instead wait for more work. That's what a thread pool does.
Should I use ThreadPool?
If you want to re-use a thread, yes.
Why is Thread "better suited for longer operations" compared to ThreadPool?
Imagine a thread pool that is serving a large number of quick operations. You don't want to have too many threads, because the computer can only do so many things at a time. Each long operation you make the thread pool do ties up a thread from the pool. So the pool either has to have lots of extra threads or may run short of threads. Neither leads to an efficient thread pool design.
For longer operations, the overhead of creating and destroying a thread is very small in comparison to the cost of the operation. So the normal downside of using a thread just for the operation doesn't apply.
If you can't reuse a thread, why use it at all (i.e. what advantages does it offer)?
I'm assuming you mean using a thread dedicated to a job that then terminates over using a thread pool. The advantage is that the number of threads will always equal the number of jobs this way. This means you have to create a thread every time you start a job and destroy a thread every time you finish one, but you never have extra threads nor do you ever run short on threads. (This can be a good thing with I/O bound threads but can be a bad thing if most threads are CPU bound most of the time.)

Thread.Start documentation says:
Once the thread terminates, it cannot be restarted with another call
to Start.
Threads are not reusable. I have already faced this problem a while ago, the solution was to create a new Thread instance whenever needed.

It looks like this by by design.
I encountered the same problem and the only solution I could find was to recreate the thread. In my case I wasn't restarting the thread very often so I didn't look any further.
A search now has turned up this thread on social.msdn where the accepted answer states:
a stopped or aborted thread cannot be stated again.
The MSDN repeat this as well:
trying to restart an aborted thread by calling Start on a thread that has terminated throws a ThreadStateException.

As the message states, you cannot restart the thread. You can simply create a new thread for your next operation. Or, you might consider a design where the background thread keeps working until it completes all of your tasks, rather than launch a new thread for each one.

for(;;){} or while(true){} are useful constructs to 'reuse' a thread. Typically, the thread waits on some synchronization object at the top of these loops. In your example, you could wait on an event or semaphore and signal it from your button OnClick() handler.

It's just in background mode. It sounds like you need to use the ThreadPool because re-starting and re-creating Thread objects are very expensive operations. If you have a long running job that may last longer than your main process, then consider the use of a Windows Service.

Finding blocking calls in VS2010 solution

I've been tasked with removing blocking calls from a C# app. Turns out this is a requirement of the environment it'll be running on. I understand the concept of a blocking call, however, I'm not sure where to begin finding existing blocking calls.
So a few questions:
For any given function, how can I tell whether or not it is blocking? Is there any way besides looking up the documentation?
Is there any way to search for blocking in a project or solution? Eg. some plug-in that could tell me?

There's no automatic way I know of to find blocking calls. Most blocking code is used for thread or process synchronization such as lock, Monitor.Enter, Mutex and Semaphore/SemaphoreSlim waits, CountdownEvent and Barrier class use. There's also SpinLock and ReaderWriterLock/ReaderWriterLockSlim locks which block.
There are several Thread calls that are blocking. Thread.Sleep can technically be considered a blocking call, though it lasts a finite amount of time. Thread.Join waits for other threads to finish and is thus blocking.
For and While loops can be considered blocking as they will run until they are done, but usually they will use one of the calls above (especially lock) if they are waiting on a specific variable updated in another thread.
Keep in mind that removing any of these is likely to have a serious negative impact on thread safety.

Resource usage of ThreadPool RegisterWaitForSingleObject

I am writing a server application which processes request from multiple clients. For the processing of requests I am using the threadpool.
Some of these requests modify a database record, and I want to restrict the access to that specific record to one threadpool thread at a time. For this I am using named semaphores (other processes are also accessing these records).
For each new request that wants to modify a record, the thread should wait in line for its turn.
And this is where the question comes in:
As I don't want the threadpool to fill up with threads waiting for access to a record, I found the RegisterWaitForSingleObject method in the threadpool.
But when I read the documentation (MSDN) under the section Remarks:
New wait threads are created automatically when required. ...
Does this mean that the threadpool will fill up with wait-threads? And how does this affect the performance of the threadpool?
Any other suggestions to boost performance is more than welcome!
Thanks!

Your solution is a viable option. In the absence of more specific details I do not think I can offer other tangible options. However, let me try to illustrate why I think your current solution is, at the very least, based on sound theory.
Lets say you have 64 requests that came in simultaneously. It is reasonable to assume that the thread pool could dispatch each one of those requests to a thread immediately. So you might have 64 threads that immediately begin processing. Now lets assume that the mutex has already been acquired by another thread and it is held for a really long time. That means those 64 threads will be blocked for a long time waiting for the thread that currently owns the mutex to release it. That means those 64 threads are wasted on doing nothing.
On the other hand, if you choose to use RegisterWaitForSingleObject as opposed to using a blocking call to wait for the mutex to be released then you can immediately release those 64 waiting threads (work items) and allow them to be put back into the pool. If I were to implement my own version of RegisterWaitForSingleObject then I would use the WaitHandle.WaitAny method which allows me to specify up to 64 handles (I did not randomly choose 64 for the number of requests afterall) in a single blocking method call. I am not saying it would be easy, but I could replace my 64 waiting threads for only a single thread from the pool. I do not know how Microsoft implemented the RegisterWaitForSingleObject method, but I am guessing they did it in a manner that is at least as efficient as my strategy. To put this another way, you should be able to reduce the number of pending work items in the thread pool by at least a factor of 64 by using RegisterWaitForSingleObject.
So you see, your solution is based on sound theory. I am not saying that your solution is optimal, but I do believe your concern is unwarranted in regards to the specific question asked.

IMHO you should let the database do its own synchronization. All you need to do is to ensure that you're sync'ed within your process.
Interlocked class might be a premature optimization that is too complex to implement. I would recommend using higher-level sync objects, such as ReaderWriterLockSlim. Or better yet, a Monitor.

An approach to this problem that I've used before is to have the first thread that gets one of these work items be responsible for any other ones that occur while it's processing the work item(s), This is done by queueing the work items then dropping into a critical section to process the queue. Only the 'first' thread will drop into the critical section. If a thread can't get the critical section, it'll leave and let the thread already operating in the critical section handle the queued object.
It's really not very complicated - the only thing that might not be obvious is that when leaving the critical section, the processing thread has to do it in a way that doesn't potentially leave a late-arriving workitem on the queue. Basically, the 'processing' critical section lock has to be released while holding the queue lock. If not for this one requirement, a synchronized queue would be sufficient, and the code would really be simple!
Pseudo code:
// `workitem` is an object that contains the database modification request
//
// `queue` is a Queue<T> that can hold these workitem requests
//
// `processing_lock` is an object use to provide a lock
// to indicate a thread is processing the queue
// any number of threads can call this function, but only one
// will end up processing all the workitems.
//
// The other threads will simply drop the workitem in the queue
// and leave
void threadpoolHandleDatabaseUpdateRequest(workitem)
{
// put the workitem on a queue
Monitor.Enter(queue.SyncRoot);
queue.Enqueue(workitem);
Monitor.Exit(queue.SyncRoot);
bool doProcessing;
Monitor.TryEnter(processing_queue, doProcessing);
if (!doProcessing) {
// another thread has the processing lock, it'll
// handle the workitem
return;
}
for (;;) {
Monitor.Enter(queue.SyncRoot);
if (queue.Count() == 0) {
// done processing the queue
// release locks in an order that ensures
// a workitem won't get stranded on the queue
Monitor.Exit(processing_queue);
Monitor.Exit(queue.SyncRoot);
break;
}
workitem = queue.Dequeue();
Monitor.Exit(queue.SyncRoot);
// this will get the database mutex, do the update and release
// the database mutex
doDatabaseModification(workitem);
}
}

ThreadPool creates a wait thread for ~64 waitable objects.
Good comments are here: Thread.sleep vs Monitor.Wait vs RegisteredWaitHandle?