views:

39856

answers:

11

Is there any difference between binary semaphore and mutex or they are essentialy same?

+31  A: 

Mutex can be released only by thread that had acquired it, while you can signal semaphore from any other thread (or process), so semaphores are more suitable for some synchronization problems like producer-consumer.

One Windows binary semaphores are more like event objects then mutants.

Mladen Jankovic
+1  A: 

The answer may depend on the target OS. For example, at least one RTOS implementation I'm familiar with will allow multiple sequential "get" operations against a single OS mutex, so long as they're all from within the same thread context. The multiple gets must be replaced by an equal number of puts before another thread will be allowed to get the mutex. This differs from binary semaphores, for which only a single get is allowed at a time, regardless of thread contexts.

The idea behind this type of mutex is that you protect an object by only allowing a single context to modify the data at a time. Even if the thread gets the mutex and then calls a function that further modifies the object (and gets/puts the protector mutex around its own operations), the operations should still be safe because they're all happening under a single thread.

{
    mutexGet();  // Other threads can no longer get the mutex.

    // Make changes to the protected object.
    // ...

    objectModify();  // Also gets/puts the mutex.  Only allowed from this thread context.

    // Make more changes to the protected object.
    // ...

    mutexPut();  // Finally allows other threads to get the mutex.
}

Of course, when using this feature, you must be certain that all accesses within a single thread really are safe!

I'm not sure how common this approach is, or whether it applies outside of the systems with which I'm familiar. For an example of this kind of mutex, see the ThreadX RTOS.

Casey Barker
The kind of mutex you are talking about are called "recursive mutex" and should be avoided since they are slow and tend to promote bad design: (see David Butenhof: http://www.zaval.org/resources/library/butenhof1.html)
Gaspard Bucher
Agreed. On this particular OS, I use the mutex service where I want to make it clear that the code is for "mutual exclusion" and not reference counting, but I don't use the recursive feature for fear of ugly unwinding. Still, in the context of the question, this is an important difference between a "mutex" and "binary semaphore."
Casey Barker
+14  A: 
Benoit
+4  A: 

There synchronization semantics are very different:

  • mutexes allows serialization of access to a given resource i.e. multiple threads wait for a lock, one at a time and as previously said, the thread owns the lock until it is done, most importantly only this particular thread can unlocks it.
  • a binary semaphore is a counter with value 0 and 1, a task blocking on it until any task does a sem_post. The semaphore advertise that a resource is available, and it provides the mechanism to wait until it is signaled as being available.

As such one can see a mutexe as a token passed from task to tasks and a semaphore as traffic red-light (it signals someone that it can proceed).

ppi
+2  A: 

On Windows, there are two differences between mutexes and binary semaphores:

  1. A mutex can only be released by the thread which has ownership, i.e. the thread which previously called the Wait function, (or which took ownership when creating it). A semaphore can be released by any thread.

  2. A thread can call a wait function repeatedly on a mutex without blocking. However, if you call a wait function twice on a binary semaphore without releasing the semaphore in between, the thread will block.

Good answer. In #2 you are describing a recursive mutex -- not all mutexes are necessarily recursive. E.g., http://www.cs.wustl.edu/~schmidt/ACE.FAQ.html#Q14
Dan
+1  A: 

Mutex work on blocking critical region, But Semaphore work on count.

+8  A: 

At a theoretical level, they are no different semantically. You can implement a mutex using semaphores or vice versa (see here for an example). In practice, the implementation is different and they offer slightly different services.

The practical difference (in terms of the system services surrounding them) is that the implementation of a mutex is aimed at being a more lightweight synchronisation mechanism. In oracle-speak, mutexes are known as latches and semaphores are known as waits.

At the lowest level, they use some sort of atomic test and set mechanism. This reads the current value of a memory location, computes some sort of conditional and writes out a value at that location in a single instruction that cannot be interrupted. This means that you can acquire a mutex and test to see if anyone else had it before you.

A typical mutex implementation has a process or thread executing the test-and-set instruction and evaluating whether anything else had set the mutex. A key point here is that there is no interaction with the scheduler, so we have no idea (and don't care) who has set the lock. Then we either give up our time slice and attempt it again when the task is re-scheduled or execute a spin-lock. A spin lock is an algorithm like:

Count down from 5000:
     i. Execute the test-and-set instruction
    ii. If the mutex is clear, we have acquired it in the previous instruction 
        so we can exit the loop
   iii. When we get to zero, give up our time slice.

When we have finished executing our protected code (known as a critical section) we just set the mutex value to zero or whatever means 'clear.' If multiple tasks are attempting to acquire the mutex they the next task that happens to be scheduled after the mutex is released will get access to the resource. Typically you would use mutexes to control a synchronised resource where exclusive access is only needed for very short periods of time, normally to make an update to a shared data structure.

A semaphore is a synchronised data structure (typically using a mutex) that has a count and some system call wrappers that interact with the scheduler in a bit more depth than the mutex libraries would. Semaphores are incremented and decremented and used to block tasks until something else is ready. See Producer/Consumer Problem for a simple example of this. Semaphores are initialised to some value - a binary semaphore is just a special case where the semaphore is initialised to 1. Posting to a semaphore has the effect of waking up a waiting process.

A basic semaphore algorithm looks like:

(somewhere in the program startup)
Initialise the semaphore to its start-up value.

Acquiring a semaphore
   i. (synchronised) Attempt to decrement the semaphore value
  ii. If the value would be less than zero, put the task on the tail of the list of tasks waiting on the semaphore and give up the time slice.

Posting a semaphore
   i. (synchronised) Increment the semaphore value
  ii. If the value is greater or equal to the amount requested in the post at the front of the queue, take that task off the queue and make it runnable.  
 iii. Repeat (ii) for all tasks until the posted value is exhausted or there are no more tasks waiting.

In the case of a binary semaphore the main practical difference between the two is the nature of the system services surrounding the actual data structure.

EDIT: As evan has rightly pointed out, spinlocks will slow down a single processor machine. You would only use a spinlock on a multi-processor box because on a single processor the process holding the mutex will never reset it while another task is running. Spinlocks are only useful on multi-processor architectures.

ConcernedOfTunbridgeWells
I don't think it is common practice for a mutex to be implemented with spinlocks. On a Uni-proc machine this would be absolutely terrible for performance.
Evan Teran
Normally you would only use spinlocks on multi-processor systems.
ConcernedOfTunbridgeWells
+28  A: 

The Toilet example is an enjoyable analogy:

Mutex:

Is a key to a toilet. One person can have the key - occupy the toilet - at the time. When finished, the person gives (frees) the key to the next person in the queue.

Officially: "Mutexes are typically used to serialise access to a section of re-entrant code that cannot be executed concurrently by more than one thread. A mutex object only allows one thread into a controlled section, forcing other threads which attempt to gain access to that section to wait until the first thread has exited from that section." Ref: Symbian Developer Library

(A mutex is really a semaphore with value 1.)

Semaphore:

Is the number of free identical toilet keys. Example, say we have four toilets with identical locks and keys. The semaphore count - the count of keys - is set to 4 at beginning (all four toilets are free), then the count value is decremented as people are coming in. If all toilets are full, ie. there are no free keys left, the semaphore count is 0. Now, when eq. one person leaves the toilet, semaphore is increased to 1 (one free key), and given to the next person in the queue.

Officially: "A semaphore restricts the number of simultaneous users of a shared resource up to a maximum number. Threads can request access to the resource (decrementing the semaphore), and can signal that they have finished using the resource (incrementing the semaphore)." Ref: Symbian Developer Library

david
Funny and simple example.
Thomman
... but this is regarding mutex vs counting semaphore. The question was asked about binary.
Roman Nikitchenko
+13  A: 

Nice articles on the topic:

From part 2:

The mutex is similar to the principles of the binary semaphore with one significant difference: the principle of ownership. Ownership is the simple concept that when a task locks (acquires) a mutex only it can unlock (release) it. If a task tries to unlock a mutex it hasn’t locked (thus doesn’t own) then an error condition is encountered and, most importantly, the mutex is not unlocked. If the mutual exclusion object doesn't have ownership then, irrelevant of what it is called, it is not a mutex.

teki
Thanks for the link, the explanations there are excellent. The link has changed: http://www.feabhas.com/blog/2009/09/mutex-vs-semaphores-%E2%80%93-part-1-semaphores/ (Use < Prev and Next > to navigate to the other two articles.
Aaron H.
@teki @Aaron I fixed the links
Judge Maygarden
Great links. Thanks
tommyk
A: 

Modified question is - What's the difference between A mutex and a "binary" semaphore in "Linux"?

Ans: Following are the differences – i) Scope – The scope of mutex is within a process address space which has created it and is used for synchronization of threads. Whereas semaphore can be used across process space and hence it can be used for interprocess synchronization.

ii) Mutex is lightweight and faster than semaphore. Futex is even faster.

iii) Mutex can be acquired by same thread successfully multiple times with condition that it should release it same number of times. Other thread trying to acquire will block. Whereas in case of semaphore if same process tries to acquire it again it blocks as it can be acquired only once.

Mickey
A: 

Apart from the fact that mutexes have an owner, the two objects may be optimized for different usage. Mutexes are designed to be held only for a short time; violating this can cause poor performance and unfair scheduling. For example, a running thread may be permitted to acquire a mutex, even though another thread is already blocked on it. Semaphores may provide more fairness, or fairness can be forced using several condition variables.

jilles