I'm in the process of writing a template library for data-caching in C++ where concurrent read can be done and concurrent write too, but not for the same key. The pattern can be explained with the following environment:
A mutex for the cache write.
A mutex for each key in the cache.
This way if a thread requests a key from the cache and is not present can start a locked calculation for that unique key. In the meantime other threads can retrieve or calculate data for other keys but a thread that tries to access the first key get locked-wait.
The main constraints are:
Never calculate the value for a key at the same time.
Calculating the value for 2 different keys can be done concurrently.
Data-retrieval must not lock other threads from retrieve data from other keys.
My other constraints but already resolved are:
fixed (known at compile time) maximum cache size with MRU-based ( most recently used ) thrashing.
retrieval by reference ( implicate mutexed shared counting )
I'm not sure using 1 mutex for each key is the right way to implement this but i didn't find any other substantially different way.
Do you know of other patterns to implements this or do you find this a suitable solution? I don't like the idea of having about 100 mutexs. ( the cache size is around 100 keys )
You want to lock and you want to wait. Thus there shall be "conditions" somewhere (as pthread_cond_t on Unix-like systems).
I suggest the following:
There is a global mutex which is used only to add or remove keys in the map.
The map maps keys to values, where values are wrappers. Each wrapper contains a condition and potentially a value. The condition is signaled when the value is set.
When a thread wishes to obtain a value from the cache, it first acquires the global mutex. It then looks in the map:
If there is a wrapper for that key, and that wrapper contains a value, then the thread has its value and may release the global mutex.
If there is a wrapper for that key but no value yet, then this means that some other thread is currently busy computing the value. The thread then blocks on the condition, to be awaken by the other thread when it has finished.
If there is no wrapper, then the thread registers a new wrapper in the map, and then proceeds to computing the value. When the value is computed, it sets the value and signals the condition.
In pseudo code this looks like this:
mutex_t global_mutex
hashmap_t map
lock(global_mutex)
w = map.get(key)
if (w == NULL) {
w = new Wrapper
map.put(key, w)
unlock(global_mutex)
v = compute_value()
lock(global_mutex)
w.set(v)
signal(w.cond)
unlock(global_mutex)
return v
} else {
v = w.get()
while (v == NULL) {
unlock-and-wait(global_mutex, w.cond)
v = w.get()
}
unlock(global_mutex)
return v
}
In pthreads terms, lock is pthread_mutex_lock(), unlock is pthread_mutex_unlock(), unlock-and-wait is pthread_cond_wait() and signal is pthread_cond_signal(). unlock-and-wait atomically releases the mutex and marks the thread as waiting on the condition; when the thread is awaken, the mutex is automatically reacquired.
This means that each wrapper will have to contain a condition. This embodies your various requirements:
No threads holds a mutex for a long period of time (either blocking or computing a value).
When a value is to be computed, only one thread does it, the other threads which wish to access the value just wait for it to be available.
Note that when a thread wishes to get a value and finds out that some other thread is already busy computing it, the threads ends up locking the global mutex twice: once in the beginning, and once when the value is available. A more complex solution, with one mutex per wrapper, may avoid the second locking, but unless contention is very high, I doubt that it is worth the effort.
About having many mutexes: mutexes are cheap. A mutex is basically an int, it costs nothing more than the four-or-so bytes of RAM used to store it. Beware of Windows terminology: in Win32, what I call here a mutex is deemed an "interlocked region"; what Win32 creates when CreateMutex() is called is something quite different, which is accessible from several distinct processes, and is much more expensive since it involves roundtrips to the kernel. Note that in Java, every single object instance contains a mutex, and Java developers do not seem to be overly grumpy on that subject.
You could use a mutex pool instead of allocating one mutex per resource. As reads are requested, first check the slot in question. If it already has a mutex tagged to it, block on that mutex. If not, assign a mutex to that slot and signal it, taking the mutex out of the pool. Once the mutex is unsignaled, clear the slot and return the mutex to the pool.
One possibility that would be a much simpler solution would be to use a single reader/writer lock on the entire cache. Given that you know there is a maximum number of entries (and it is relatively small), it sounds like adding new keys to the cache is a "rare" event. The general logic would be:
acquire read lock
search for key
if found
use the key
else
release read lock
acquire write lock
add key
release write lock
// acquire the read lock again and use it (probably encapsulate in a method)
endif
Not knowing more about the usage patterns, I can't say for sure if this is a good solution. It is very simple, though, and if the usage is predominantly reads, then it is very inexpensive in terms of locking.
Related
I used google benchmark to run following 3 tests and the result surprised me since the RW lock is ~4x slower than simple mutex in release mode. (and ~10x slower than simple mutex in debug mode)
void raw_access() {
(void) (gp->a + gp->b);
}
void mutex_access() {
std::lock_guard<std::mutex> guard(g_mutex);
(void) (gp->a + gp->b);
}
void rw_mutex_access() {
boost::shared_lock<boost::shared_mutex> l(g_rw_mutex);
(void) (gp->a + gp->b);
}
the result is:
2019-06-26 08:30:45
Running ./perf
Run on (4 X 2500 MHz CPU s)
CPU Caches:
L1 Data 32K (x2)
L1 Instruction 32K (x2)
L2 Unified 262K (x2)
L3 Unified 4194K (x1)
Load Average: 5.35, 3.22, 2.57
-----------------------------------------------------------
Benchmark Time CPU Iterations
-----------------------------------------------------------
BM_RawAccess 1.01 ns 1.01 ns 681922241
BM_MutexAccess 18.2 ns 18.2 ns 38479510
BM_RWMutexAccess 92.8 ns 92.8 ns 7561437
I didn't get the enough information via google, so hope some help here.
Thanks
I don't know the particulars of how the standard library/boost/etc. implementations differ, although it seems like the standard library version is faster (congrats, whoever wrote it).
So instead I'll try to explain the speed differences between various mutex types on a theoretical level, which will explain why a shared mutex (should) be slower.
Atomic Spin Lock
More-so as an academic exercise, consider the simplest thread-safety "mutex-like" implementation: a simple atomic spin lock.
In essence, this is nothing more than an std::atomic<bool> or an std::atomic_flag. It is initialized to false. To "lock" the mutex, you simply do an atomic compare-and-exchange operation in a loop until you get a false value (i.e. the previous value was false prior to atomically setting it to true).
std::atomic_flag flag = ATOMIC_FLAG_INIT;
// lock it by looping until we observe a false value
while (flag.test_and_set()) ;
// do stuff under "mutex" lock
// unlock by setting it back to false state
flag.clear();
However, due to the nature of this construct, it's what we call an "unfair" mutex because the order of threads that acquire the lock is not necessarily the order in which they began their attempts to lock it. That is, under high contention, it's possible a thread tries to lock and simply never succeed because other threads are luckier. It's very timing-sensitive. Imagine musical chairs.
Because of this, although it functions like a mutex, it's not what we consider a "mutex".
Mutex
A mutex can be thought of as building on top of an atomic spin lock (although it's not typically implemented as such, since they typically are implemented with support of the operating system and/or hardware).
In essence, a mutex is a step above atomic spin locks because it has a queue of waiting threads. This allows it to be "fair" because the order of lock acquisition is (more or less) the same as the order of locking attempts.
If you've noticed, if you run sizeof(std::mutex) it might be a bit larger than you might expect. On my platform it's 40 bytes. That extra space is used to hold state information, notably including some way of accessing a lock queue for each individual mutex.
When you try to lock a mutex, it performs some low-level thread-safety operation to have thread-safe access to the mutex's status information (e.g. atomic spin lock), checks the state of the mutex, adds your thread to the lock queue, and (typically) puts your thread to sleep while you wait so you don't burn precious CPU time. The low-level thread-safety operation (e.g. the atomic spin lock) is atomically released at the same time the thread goes to sleep (this is typically where OS or hardware support is necessary to be efficient).
Unlocking is performed by performing a low-level thread-safe operation (e.g. atomic spin lock), popping the next waiting thread from the queue, and waking it up. The thread that has been awoken now "owns" the lock. Rinse wash and repeat.
Shared Mutex
A shared mutex takes this concept a step further. It can be owned by a single thread for read/write permissions (like a normal mutex), or by multiple threads for read-only permissions (semantically, anyway - it's up to the programmer to ensure it's safe).
Thus, in addition to the unique ownership queue (like a normal mutex) it also has a shared ownership state. The shared ownership state could be simply a count of the number of threads that currently have shared ownership. If you inspect sizeof(std::shared_mutex) you'll find it's typically even larger than std::mutex. On my system, for instance, it's 56 bytes.
So when you go to lock a shared mutex, it has to do everything a normal mutex does, but additionally has to verify some other stuff. For instance, if you're trying to lock uniquely it must verify that there are no shared owners. And when you're trying to lock sharingly it must verify that there are no unique owners.
Because we typically want mutexes to be "fair", once a unique locker is in the queue, future sharing lock attempts must queue instead of acquiring the lock, even though it might currently be under sharing (i.e. non-unique) lock by several threads. This is to ensure shared owners don't "bully" a thread that wants unique ownership.
But this also goes the other way: the queuing logic must ensure a shared locker is never placed into an empty queue during shared ownership (because it should immediately succeed and be another shared owner).
Additionally, if there's a unique locker, followed by a shared locker, followed by a unique locker, it must (roughly) guarantee that acquisition order. So each entry in the lock queue also needs a flag denoting its purpose (i.e. shared vs. unique).
And then we think of the wake-up logic. When you unlock a shared mutex, the logic differs depending on the current ownership type of the mutex. If the unlocking thread has unique ownership or is the last shared owner it might have to wake up some thread(s) from the queue. It will either wake up all threads at the front of the queue who are requesting shared ownership, or a single thread at the front of the queue requesting unique ownership.
As you can imagine, all of this extra logic for who's locking for what reasons and how it changes depending not only on the current owners but also on the contents of the queue makes this potentially quite a bit slower. The hope is that you read significantly more frequent than you write, and thus you can have many sharing owners running concurrently, which mitigates the performance hit of coordinating all of that.
I am trying to develop an application with one producer and several consumers.
The producers is one process and each consumer is one process. The shared resource is some kind of buffer in the shared memory.
The producer should work completely independent from the consumers. It should not be blocked in any case. Therefor the consumers are responsible to check if the data they read from the shared memory is valid and handle it if the producer has already overwritten the data. (They do this using some kind of hashing. Not important.)
The consumers should be informed when new data is available in the buffer. I think boost interprocess conditions are suitable for this usecase. (More suitable would be boost signals2, but this library is not working in an interprocess way).
Conditionas always need a mutex. But I do not need the mutex in my producer. In the consumers I only need the mutex for condition#wait.
Is it ok to only use the codnition#notify_all in the producer and do not use the mutex? Or is this an abuse of the library?
Thanks in advance
It's okay to signal without holding the mutex, but it could lead to worst-case behaviour in rare cases (thread starvation).
Signaling under the mutex guarantees fair scheduling of the waiters under POSIX as far as I am aware ¹
That said, I think the commenters are right when they smell overcomplication in the design. I'd simplify. Optimize when you need it.
¹ See e.g. here: http://linux.die.net/man/3/pthread_cond_signal
The pthread_cond_broadcast() or pthread_cond_signal() functions may be called by a thread whether or not it currently owns the mutex that threads calling pthread_cond_wait() or pthread_cond_timedwait() have associated with the condition variable during their waits; however, if predictable scheduling behavior is required, then that mutex shall be locked by the thread calling pthread_cond_broadcast() or pthread_cond_signal().
The producer should work completely independent from the consumers. It
should not be blocked in any case.
Why not? This should not affect the performance if you do not lock too frequently. You can have a data counter in shared memory and you would lock access to that counter only. Data can be stored in circular buffer in shared memory and access to it does not need to be locked because consumers check how much data is available to read using counter. Of course consumers need to read data fast enough. If the data is overwritten then the internal consumer counter can be reset to the value of interprocess counter.
Also producer can store data using many threads. Each thread can calculate future position of the data at the beginning of the thread and then update the counter after the data is stored at the end of the thread. Then additional locking is needed for future position calculations so that this value can be passed between threads.
In details, the non-multithreaded scenario could work like this:
Producer loop:
receive X samples of data
lock access to interprocess counter, increment the counter, unlock the access
Then each consumer has it's own internal counter so that it can compare with interprocess counter if and how much data is available to read (simply polling for data):
Consumer loop:
lock access to interprocess counter, read the counter value, unlock the access
compare the read value with internal counter
if values equal // no data available
sleep, then continue to the beginning of the loop
else if data overwritten // no need for hashing here, counter can be use to figure that out although doing it this way is probably a bit risky
set internal counter to the value of the interprocess counter
then continue to the beginning of the loop
else
read available data
increment internal counter
I have a question regarding threads. It is known that basically when we call for mutex(lock) that means that thread keeps on executing the part of code uninterrupted by other threads until it meets mutex(unlock). (At least that's what they say in the book) So my question is if it is actually possible to have several scoped WriteLocks which do not interfere with each other. For example something like this:
If I have a buffer with N elements without any new elements coming, however with high frequency updates (like change value of Kth element) is it possible to set a different lock on each element so that the only time threads would stall and wait is if actually 2 or more threads are trying to update the same element?
To answer your question about N mutexes: yes, that is indeed possible. What resources are protected by a mutex depends entirely on you as the user of that mutex.
This leads to the first (statement) part of your question. A mutex by itself does not guarantee that a thread will work uninterrupted. All it guarantees is MUTual EXclusion - if thread B attempts to lock a mutex which thread A has locked, thread B will block (execute no code) until thread A unlocks the mutex.
This means mutexes can be used to guarantee that a thread executes a block of code uninterrupted; but this works only if all threads follow the same mutex-locking protocol around that block of code. Which means it is your responsibility to assign semantics (or meaning) to each individual mutex, and correctly adhere to those semantics in your code.
If you decide for the semantics to be "I have an array a of N data elements and an array m of N mutexes, and accessing a[i] can only be done when m[i] is locked," then that's how it will work.
The need to consistently stick to the same protocol is why you should generally encapsulate the mutex and the code/data protected by it in a class in some way or another, so that outside code doesn't need to know the details of the protocol. It just knows "call this member function, and the synchronisation will happen automagically." This "automagic" will be the class correcrtly implementing the protocol.
A crucial consideration when deciding between a mutex per array and a mutex per element is whether there are operations - like tracking the number of "in-use" array elements, the "active" element, or moving a pointer-to-array to a larger buffer - that can only be done safely by one thread while all the others are blocked.
A lesser but sometimes important consideration is the amount of extra memory more mutexes use.
If you genuinely need to do this kind of update as quickly as possible in a highly contested multi-threaded program, you may also want to learn about lock-free atomic types and their compare-and-swap / exchange operations, but I'd recommend against considering that unless profiling the existing locking is significant in your overall program performance.
A mutex does not stop other threads from running completely, it only stops other threads from locking the same mutex. I.e. while one thread is keeping the mutex locked, the operating system continues to do context switches letting other threads run also, but if any other thread is trying to lock the same mutex its execution will be halted until the mutex is unlocked.
So yes, you can indeed have several different mutexes and lock/unlock them independently. Just beware of deadlocks, i.e. if one thread can lock more than one mutex at a time you can run into a situation where thread 1 has locked mutex A and is trying to lock mutex B but blocks because thread 2 already has mutex B locked and it is trying to lock mutex A..
Its not completely clear that your use case is:
the threads gets a buffer assigned on that they have to work
the threads have some results and request a special buffer to update.
On the first variant you need some assignment logic that assigns a buffer to a thread.
This logic has to be exectued in an atomic way. so the best is to use a mutex to protect the assignment logic.
On the other variant it may be the best to have a vector of mutexes, one for each buffer element.
In Both cases the buffer does not need a protection because it (or better each field of it) is only accessed from one thread at a time.
You also may inform yourself about 'semaphores'. These contain a counter that allows to manage ressources that have a limited amount but more than one. Mutexes are a special case of semaphores with n=1.
You can have mutex per entry, C++11 mutex can be easily converted into an adaptive-spinlock, so you can achieve good CPU/Latency tradeoff.
Or, if you need very low latency yet have enough CPUs you can use an atomic "busy" flag per entry and spin in a tight compare-exchange loop on contention.
From experience, though, the best performance and scalability are achieved when concurrent writes are serialized via a command queue (or a queue of smaller immutable buffers to be concatenated at destination) and a single thread processing the queue.
I am wokring on pthread pool and there will be five separate thread and one queue. All the five threads are competing to get a job from the queue and I know the basic idea that I need to do lock/unlock and wait/signal.
But I am not sure how many mutex and cond variable I should have. Right now I have only one mutex and cond variable and all five thread will use it.
One mutex and at least one condition variable.
One mutex because there is one 'thing' (i.e. piece of memory) to synchronize access to: the shared state between all workers and the thread pushing the work.
One condition variable per, well, condition that one or more threads need to wait on. At the very least you need one condition variable for waiting on new jobs, the condition here being: "is there more stuff to do?" (or the converse: "is the work queue empty?").
A somewhat more substantive answer would be that there is a one-to-many relationship between a mutex and associated condition variables, and a one-to-one relationship between shared states and mutexes. From what you've told us and since you're learning, I recommend using only one shared state for your design. When or if you need more that one state I'd recommend looking for some higher level concepts (e.g. channels, futures/promises) to build up on abstraction.
In any case, don't use the same condition variable with different mutexes.
To elaborate on #Ivan's solution...
Instead of a mutex + condition variable, you can use a counting semaphore + atomic operations to create a very efficient queue.
semaphore dequeue_sem = 0;
semaphore enqueue_sem = 37; // or however large you want to bound your queue
Enqueue operation is just:
wait_for(enqueue_sem)
atomic_add_to_queue(element)
signal(dequeue_sem)
Dequeue operation is:
wait_for(dequeue_sem)
element = atomic_remove_from_queue()
signal(enqueue_sem)
The "atomic_add_to_queue" and "atomic_remove_from_queue" are typicaly implemented using atomic compare&exchange in a tight loop.
In addition to its symmetry, this formulation bounds the maximum size of the queue; if a thread calls enqueue() on a full queue, it will block. This is almost certainly what you want for any queue in a multi-threaded environment. (Your computer has finite memory; consuming it without bound should be avoided when possible.)
If you do stick with a mutex and condition variables, you want two conditions, one for enqueue to wait on (and deque to signal) and one for the other way around. The conditions mean "queue not full" and "queue not empty", respectively, and the enqueue/dequeue code is similarly symmetric.
I think that you could do stealing work from queue without locking at all via Interlocked operations if you organize it as stack/linked list (it will require semaphore instead of condition variable to prevent problem described in comments to this answer).
Pseudo-code is like that:
candidate = head.
if (candidate == null) wait_for_semaphore;
if (candidate == InterlockedCompareExchange(head, candidate->next, candidate)) perform_work(candidate->data);
else goto 1;
Of course, adding work to queue should be also via InterlockedCompareExchange in this case and signaling the semaphore.
I have a class instances which is being used in multiple threads. I am updating multiple member variables from one thread and reading the same member variables from one thread. What is the correct way to maintain the thread safety?
eg:
phthread_mutex_lock(&mutex1)
obj1.memberV1 = 1;
//unlock here?
Should I unlock the mutex over here? ( if another thread access the obj1 member variables 1 and 2 now, the accessed data might not be correct because memberV2 has not yet be updated. However, if I does not release the lock, the other thread might block because there is time consuming operation below.
//perform some time consuming operation which must be done before the assignment to memberV2 and after the assignment to memberV1
obj1.memberV2 = update field 2 from some calculation
pthread_mutex_unlock(&mutex1) //should I only unlock here?
Thanks
Your locking is correct. You should not release the lock early just to allow another thread to proceed (because that would allow the other thread to see the object in an inconsistent state.)
Perhaps it would be better to do something like:
//perform time consuming calculation
pthread_mutex_lock(&mutex1)
obj1.memberV1 = 1;
obj1.memberV2 = result;
pthread_mutex_unlock(&mutex1)
This of course assumes that the values used in the calculation won't be modified on any other thread.
Its hard to tell what you are doing that is causing problems. The mutex pattern is pretty simple. You Lock the mutex, access the shared data, unlock the mutex. This protects data, becuase the mutex will only let one thread get the lock at a time. Any thread that fails to get the lock has to wait till the mutex is unlocked. Unlocking wakes the waiters up. They will then fight to attain the lock. Losers go back to sleep. The time it takes to wake up might be multiple ms or more from the time the lock is released. Make sure you always unlock the mutex eventually.
Make sure you don't to keep locks locked for a long period of time. Most of the time, a long period of time is like a micro second. I prefer to keep it down around "a few lines of code." Thats why people have suggested that you do the long running calculation outside the lock. The reason for not keeping locks a long time is you increase the number of times other threads will hit the lock and have to spin or sleep, which decreases performance. You also increase the probability that your thread might be pre-empted while owning the lock, which means the lock is enabled while that thread sleeps. Thats even worse performance.
Threads that fail a lock dont have to sleep. Spinning means a thread encountering a locked mutex doesn't sleep, but loops repeatedly testing the lock for a predefine period before giving up and sleeping. This is a good idea if you have multiple cores or cores capable of multiple simultaneous threads. Multiple active threads means two threads can be executing the code at the same time. If the lock is around a small amount of code, then the thread that got the lock is going to be done real soon. the other thread need only wait a couple nano secs before it will get the lock. Remember, sleeping your thread is a context switch and some code to attach your thread to the waiters on the mutex, all have costs. Plus, once your thread sleeps, you have to wait for a period of time before the scheduler wakes it up. that could be multiple ms. Lookup spinlocks.
If you only have one core, then if a thread encounters a lock it means another sleeping thread owns the lock and no matter how long you spin it aint gonna unlock. So you would use a lock that sleeps a waiter immediately in hopes that the thread owning the lock will wake up and finish.
You should assume that a thread can be preempted at any machine code instruction. Also you should assume that each line of c code is probably many machine code instructions. The classic example is i++. This is one statement in c, but a read, an increment, and a store in machine code land.
If you really care about performance, try to use atomic operations first. Look to mutexes as a last resort. Most concurrency problems are easily solved with atomic operations (google gcc atomic operations to start learning) and very few problems really need mutexes. Mutexes are way way way slower.
Protect your shared data wherever it is written and wherever it is read. else...prepare for failure. You don't have to protect shared data during periods of time when only a single thread is active.
Its often useful to be able to run your app with 1 thread as well as N threads. This way you can debug race conditions easier.
Minimize the shared data that you protect with locks. Try to organize data into structures such that a single thread can gain exclusive access to the entire structure (perhaps by setting a single locked flag or version number or both) and not have to worry about anything after that. Then most of the code isnt cluttered with locks and race conditions.
Functions that ultimately write to shared variables should use temp variables until the last moment and then copy the results. Not only will the compiler generate better code, but accesses to shared variables especially changing them cause cache line updates between L2 and main ram and all sorts of other performance issues. Again if you don't care about performance disregard this. However i recommend you google the document "everything a programmer should know about memory" if you want to know more.
If you are reading a single variable from the shared data you probably don't need to lock as long as the variable is an integer type and not a member of a bitfield (bitfield members are read/written with multiple instructions). Read up on atomic operations. When you need to deal with multiple values, then you need a lock to make sure you didn't read version A of one value, get preempted, and then read version B of the next value. Same holds true for writing.
You will find that copies of data, even copies of entire structures come in handy. You can be working on building a new copy of the data and then swap it by changing a pointer in with one atomic operation. You can make a copy of the data and then do calculations on it without worrying if it changes.
So maybe what you want to do is:
lock the mutex
Make a copy of the input data to the long running calculation.
unlock the mutex
L1: Do the calculation
Lock the mutex
if the input data has changed and this matters
read the input data, unlock the mutex and go to L1
updata data
unlock mutex
Maybe, in the example above, you still store the result if the input changed, but go back and recalc. It depends if other threads can use a slightly out of date answer. Maybe other threads when they see that a thread is already doing the calculation simply change the input data and leave it to the busy thread to notice that and redo the calculation (there will be a race condition you need to handle if you do that, and easy one). That way the other threads can do other work rather than just sleep.
cheers.
Probably the best thing to do is:
temp = //perform some time consuming operation which must be done before the assignment to memberV2
pthread_mutex_lock(&mutex1)
obj1.memberV1 = 1;
obj1.memberV2 = temp; //result from previous calculation
pthread_mutex_unlock(&mutex1)
What I would do is separate the calculation from the update:
temp = some calculation
pthread_mutex_lock(&mutex1);
obj.memberV1 = 1;
obj.memberV2 = temp;
pthread_mutex_unlock(&mutex1);