multithreading: which variables need mutex protection when communication via a condition variable? - c

I have a question on the interplay between condition variables and associated mutex locks (it arose from a simplified example I was presenting in a lecture, confusing myself in the process). Two threads are exchanging data (lets say an int n indicating an array size, and a double *d which points to the array) via shared variables in the memory of their process. I use an additional int flag (initially flag = 0) to indicate when the data (n and d) is ready (flag = 1), a pthread_mutex_t mtx, and a condition variable pthread_cond_t cnd.
This part is from the receiver thread which waits until flag becomes 1 under the protection of the mutex lock, but afterward processes n and d without protection:
while (1) {
pthread_mutex_lock(&mtx);
while (!flag) {
pthread_cond_wait(&cnd, &mtx);
}
pthread_mutex_unlock(&mtx);
// use n and d
}
This part is from the sender thread which sets n and d beforehand without protection by the mutex lock, but sets flag while the mutex is locked:
n = 10;
d = malloc(n * sizeof(float));
pthread_mutex_lock(&mtx);
flag = 1;
pthread_cond_signal(&cnd);
pthread_mutex_unlock(&mtx);
It is clear that you need the mutex in the sender since otherwise you have a "lost wakeup call" problem (see https://stackoverflow.com/a/4544494/3852630).
My question is different: I'm not sure which variables have to be set (in the sender thread) or read out (in the receiver thread) inside the region protected by the mutex lock, and which variables don't need to be protected by the mutex lock. Is it sufficient to protect flag on both sides, or do n and d also need protection?
Memory visiblity (see the rule below) between sender and receiver should be guaranteed by the call to pthread_cond_signal(), so the necessary pairwise memory barriers should be there (in combination with pthread_cond_wait()).
I'm aware that this is an unusual case. Typically my applications modify a task list in the sender and pop tasks from the list in the receiver, and the associated mutex lock protects of the list operations on both sides. However I'm not sure what would be necessary in the case above. Could the danger be that the compiler (which is not aware of the concurrent access to variables) optimizes away the write and/or read access to the variables? Are there other problems if n and d are not protected by the mutex?
Thanks for your help!
David R. Butenhof: Programming with POSIX Threads, p.89: "Whatever memory values a thread can see when is signals ... a condition variable can also be seen by any thread that is awakened by that signal ...".

At the low level of memory ordering, the rules aren't specified in terms of "the region protected by a mutex", but in terms of synchronization. Whenever two different threads access the same non-atomic object, then unless they are both reads, there must be a synchronization operation in between, to ensure that one of the accesses definitely happens before the other.
The way to achieve synchronization is to have one thread perform a release operation (such as unlocking a mutex) after accessing the shared variables, and have the other thread perform an acquire operation (such as locking a mutex) before accessing them, in such a way that program logic guarantees the acquire must have happened after the release.
And here, you have that. The sender thread does perform a mutex unlock after accessing n and d (last line of the sender code). And the receiver does perform a mutex lock before accessing them (inside pthread_cond_wait). The setting and testing of flag ensures that, when we exit the while (!flag) loop, the most recent lock of the mutex by the receiver did happen after the unlock by the sender. So synchronization is achieved.
The compiler and CPU must not perform any optimization that would defeat this, so in particular they can't optimize away the accesses to n and d, or reorder them around the synchronizing operations. This is usually ensured by treating the release/acquire operations as barriers. Any accesses to potentially shared objects that occur in program order before a release barrier must actually be performed and flushed to coherent memory/cache prior to anything that comes after the barrier (in this case, before any other thread may see the mutex as unlocked). If special CPU instructions are needed to ensure global visibility, the compiler must emit them. And vice versa for acquire barriers: any access that occurs in program order after an acquire barrier must not be reordered before it.
To say it another way, the compiler treats the release barrier as an operation that may potentially read all of memory; so all variables must be written out before that point, so that the actual contents of memory at that point will match what an abstract machine would have. Likewise, an acquire barrier is treated as an operation that may potentially write all of memory, and all variables must be reloaded from memory afterwards. The only exception would be local variables for which the compiler can prove that no other thread could legally know their address; those can be safely kept in registers or otherwise reordered.
It's true that, after the synchronizing lock operation, the receiver happened to unlock the mutex again, but that isn't relevant here. That unlock doesn't synchronize with anything in this particular program, and it has no impact on its execution. Likewise, for purposes of synchronizing access to n and d, it didn't matter whether the sender locked the mutex before or after accessing them. (Though it was important that the sender locked the mutex before writing to flag; that's how we ensure that any earlier reads of flag by the receiver really did happen before the write, instead of racing with it.)
The principle that "accesses to shared variables should be inside a critical region protected by a mutex" is just a higher-level abstraction that is one way to ensure that accesses by different threads always have a synchronizing unlock-lock pair in between them. And in cases where the variables could be accessed over and over again, you normally would want a lock before, and an unlock after, every such access, which is equivalent to the "critical section" principle. But this principle is not itself the fundamental rule.
That said, in real life you probably do want to follow this principle as much as possible, since it will make it easier to write correct code and avoid subtle bugs, and likewise make it easier for other programmers to verify that your code is correct. So even though it is not strictly necessary in this program for the accesses to n,d to be "protected" by the mutex, it would probably be wise to do so anyway, unless there is a significant and measurable benefit (performance or otherwise) to be gained by avoiding it.
The condition variable doesn't play a role in the race avoidance here, except insofar as the pthread_cond_wait locked the mutex. Functionally, it is equivalent to having the receiver simply do a tight spin loop of "lock mutex; test flag; unlock mutex", but without wasting all those CPU cycles.
And I think that the quote from Butenhof is mistaken, or at best misleading. My understanding is that pthread_cond_signal by itself is not guaranteed to be a barrier of any kind, and in fact has no memory ordering effect whatsoever. POSIX doesn't directly address memory ordering, but this is the case for the standard C equivalent cnd_signal. There would be no point in having pthread_cond_signal ensure global visibility unless you could make use of it by assuming that all those accesses were visible by the time the corresponding pthread_cond_wait returns. But since pthread_cond_wait can wake up spuriously, you cannot ever safely make such an assumption.
In this program, the necessary release barrier in the sender comes not from pthread_cond_signal, but rather from the subsequent pthread_mutex_unlock.

Is it sufficient to protect flag on both sides, or do n and d also need protection?
In principle, if you use your mutex, CV, and flag in such a way that the the writer does not modify n, d, or *d after setting flag under mutex protection, and the reader cannot access n, d, or *d until after it observes the modification of flag (under protection of the same mutex), then you can rely on the reader observing the writer's last-written values of n, d, and *d. This is more or less a hand-rolled semaphore.
In practice, you should use whichever system synchronization objects you have chosen (mutexes, semaphores, etc.) to protect all the shared data. Doing so is easier to reason about and less prone to spawn bugs. Often, it's simpler, too.

Related

Mutex on pointers to a shared variable

I am new to thread programming. I know that mutexes are used to protect access to shared data in a multi-threaded program.
Suppose I have one thread with variable a and a second one with the pointer variable p that holds the address of a. Is the code safe if, in the second thread, I lock a mutex before I modify the value of a using the pointer variable? From my understanding it is safe.
Can you confirm? And also can you provide the reason why it is true or why it is not true?
I am working with c and pthreads.
The general rule when doing multithreading is that shared variables among threads that are read and written need to be accessed serially, which means that you must use some sort of synchronization primitive. Mutexes are a popular choice, but no matter what you end up using, you just need to remember that before reading from or writing to a shared variable, you need to acquire a lock to ensure consistency.
So, as long as every thread in your code agrees to always use the same lock before accessing a given variable, you're all good.
Now, to answer your specific questions:
Is the code safe if, in the second thread, I lock a mutex before I
modify the value of a using the pointer variable?
It depends. How do you read a on the first thread? The first thread needs to lock the mutex too before accessing a in any way. If both threads lock the same mutex before reading or writing the value of a, then it is safe.
It's safe because the region of code between the mutex lock and unlock is exclusive (as long as every thread respects the rule that before doing Y, they need to acquire lock X), since only one thread at a time can have the lock.
As for this comment:
And if the mutex is locked before p is used, then both a and p are
protected? The conclusion being that every memory reference present in
a section where a mutex is locked is protected, even if the memory is
indirectly referenced?
Mutexes don't protect memory regions or references, they protect a region of code. Whatever you make between locking and unlocking is exclusive; that's it. So, if every thread accessing or modifying either of a or p locks the same mutex before and unlocks afterwards, then as a side-effect you have synchronized accesses.
TL;DR Mutexes allow you to write code that never executes in parallel, you get to choose what that code does - a remarkably popular pattern is to access and modify shared variables.

POSIX threads and global variables in C on Linux

If I have two threads and one global variable (one thread constantly loops to read the variable; the other constantly loops to write to it) would anything happen that shouldn't? (ex: exceptions, errors). If it, does what is a way to prevent this. I was reading about mutex locks and that they allow exclusive access to a variable to one thread. Does this mean that only that thread can read and write to it and no other?
Would anything happen that shouldn't?
It depends in part on the type of the variables. If the variable is, say, a string (long array of characters), then if the writer and the reader access it at the same time, it is completely undefined what the reader will see.
This is why mutexes and other coordinating mechanisms are provided by pthreads.
Does this mean that only that thread can read and write to it and no other?
Mutexes ensure that at most one thread that is using the mutex can have permission to proceed. All other threads using the same mutex will be held up until the first thread releases the mutex. Therefore, if the code is written properly, at any time, only one thread will be able to access the variable. If the code is not written properly, then:
one thread might access the variable without checking that it has permission to do so
one thread might acquire the mutex and never release it
one thread might destroy the mutex without notifying the other
None of these is desirable behaviour, but the mere existence of a mutex does not prevent any of these happening.
Nevertheless, your code could reasonably use a mutex carefully and then the access to the global variable would be properly controlled. While it has permission via the mutex, either thread could modify the variable, or just read the variable. Either will be safe from interference by the other thread.
Does this mean that only that thread can read and write to it and no other?
It means that only one thread can read or write to the global variable at a time.
The two threads will not race amongst themselves to access the global variable neither will they access it at the same time at any given point of time.
In short the access to the global variable is Synchronized.
First; In C/C++ unsynchronized read/write of variable does not generate any exceptions or system error, BUT it can generate application level errors -- mostly because you are unlikely to fully understand how the memory is accessed, and whether it is atomic unless you look at the generated assembler. A multi core CPU may likely create hard-to-debug race conditions when you access shared memory without synchronization.
Hence
Second; You should always use synchronization -- such as mutex locks -- when dealing with shared memory. A mutex lock is cheap; so it will not really impact performance if done right. Rule of thumb; keep the lcok for as short as possible, such as just for the duration of reading/incrementing/writing the shared memory.
However, from your description, it sounds like that one of your threads is doing nothing BUT waiting for the shared meory to change state before doing something -- that is a bad multi-threaded design which cost unnecessary CPU burn, so
Third; Look at using semaphores (sem_create/wait/post) for synchronization between your threads if you are trying to send a "message" from one thread to the other
As others already said, when communicating between threads through "normal" objects you have to take care of race conditions. Besides mutexes and other lock structures that are relatively heavy weight, the new C standard (C11) provides atomic types and operations that are guaranteed to be race-free. Most modern processors provide instructions for such types and many modern compilers (in particular gcc on linux) already provide their proper interfaces for such operations.
If the threads truly are only one producer and only one consumer, then (barring compiler bugs) then
1) marking the variable as volatile, and
2) making sure that it is correctly aligned, so as to avoid interleaved fetches and stores
will allow you to do this without locking.

Implementing mutex in a user level thread library

I am developing a user level thread library as part of a project. I came up with an approach to implement mutex. I would like to see ur views before going on with it. Basically, i need to implement just 3 functions in my library
mutex_init, mutex_lock and mutex_unlock
I thought my mutex_t structure would look something like
typedef struct
{
int available; //indicates whether the mutex is locked or unlocked
queue listofwaitingthreads;
gtthread_t owningthread;
}mutex_t;
In my mutex_lock function, i will first check if the mutex is available in a while loop. If it is not, i will yield the processor for the next thread to execute.
In my mutex_unlock function, i will check if the owner thread is the current thread. If it is, i will set available to 0.
Is this the way to go about it ? Also, what about deadlock? Should i take care of those conditions in my user level library or should i leave the application programmers to write code properly ?
This won't work, because you have a race condition. If 2 threads try to catch the lock at the same time, both will see available == 0, and both will think they succeeded with taking the mutex.
If you want to do this properly, and without using an already-existing lock, You must access hardware operations like TAS, CAS, etc.
There are algorithms that give you mutual exclusion without such hardware support, but they make some assumptions that are many times false. For more details about this, I highly recommend reading Herlihy and Shavit's The art of multiprocessor programming, chapter 7.
You shouldn't worry about deadlocks in this level - mutex locks should be simple enough, and there is some assumption that the programmer using them should use care not to cause deadlocks (advanced mutexes can check for self-deadlock, meaning a thread that calls lock twice without calling unlock in the middle).
Not only that you have to do atomic operations to read and modify the flag (as Eran pointed out) you also have to watch that your queue is capable to have concurrent accesses. This is not completely trivial, sort of hen and egg problem.
But if you'd really implement this by spinning, you wouldn't even need to have such a queue. The access order to the lock then would be mainly random, though.
Probably just yielding would also not be enough, this can be quite costly if you have threads holding the lock for more than some processor cycles. Consider using nanosleep with a low time value for the wait.
In general, a mutex implementation should look like:
Lock:
while (trylock()==failed) {
atomic_inc(waiter_cnt);
atomic_sleep_if_locked();
atomic_dec(waiter_cnt);
}
Trylock:
return atomic_swap(&lock, 1);
Unlock:
atomic_store(&lock, 0);
if (waiter_cnt) wakeup_sleepers();
Things get more complex if you want recursive mutexes, mutexes that can synchronize their own destruction (i.e. freeing the mutex is safe as soon as you get the lock), etc.
Note that atomic_sleep_if_locked and wakeup_sleepers correspond to FUTEX_WAIT and FUTEX_WAKE ops on Linux. The other atomics are probably CPU instructions, but could be system calls or kernel-assisted userspace function code, as in the case of Linux/ARM and the 0xffff0fc0 atomic compare-and-swap call.
You do not need atomic instructions for a user level thread library, because all the threads are going to be user level threads of the same process. So actually when your process is given the time slice to execute, you are running multiple threads during that time slice but on the same processor. So, no two threads are going to be in the library function at the same time. Considering that the functions for mutex are already in the library, mutual exclusion is guaranteed.

Is this usage of condition variables ALWAYS subject to a lost-signal race?

Suppose a condition variable is used in a situation where the signaling thread modifies the state affecting the truth value of the predicate and calls pthread_cond_signal without holding the mutex associated with the condition variable? Is it true that this type of usage is always subject to race conditions where the signal may be missed?
To me, there seems to always be an obvious race:
Waiter evaluates the predicate as false, but before it can begin waiting...
Another thread changes state in a way that makes the predicate true.
That other thread calls pthread_cond_signal, which does nothing because there are no waiters yet.
The waiter thread enters pthread_cond_wait, unaware that the predicate is now true, and waits indefinitely.
But does this same kind of race condition always exist if the situation is changed so that either (A) the mutex is held while calling pthread_cond_signal, just not while changing the state, or (B) so that the mutex is held while changing the state, just not while calling pthread_cond_signal?
I'm asking from a standpoint of wanting to know if there are any valid uses of the above not-best-practices usages, i.e. whether a correct condition-variable implementation needs to account for such usages in avoiding race conditions itself, or whether it can ignore them because they're already inherently racy.
The fundamental race here looks like this:
THREAD A THREAD B
Mutex lock
Check state
Change state
Signal
cvar wait
(never awakens)
If we take a lock EITHER on the state change OR the signal, OR both, then we avoid this; it's not possible for both the state-change and the signal to occur while thread A is in its critical section and holding the lock.
If we consider the reverse case, where thread A interleaves into thread B, there's no problem:
THREAD A THREAD B
Change state
Mutex lock
Check state
( no need to wait )
Mutex unlock
Signal (nobody cares)
So there's no particular need for thread B to hold a mutex over the entire operation; it just need to hold the mutex for some, possible infinitesimally small interval, between the state change and signal. Of course, if the state itself requires locking for safe manipulation, then the lock must be held over the state change as well.
Finally, note that dropping the mutex early is unlikely to be a performance improvement in most cases. Requiring the mutex to be held reduces contention over the internal locks in the condition variable, and in modern pthreads implementations, the system can 'move' the waiting thread from waiting on the cvar to waiting on the mutex without waking it up (thus avoiding it waking up only to immediately block on the mutex).
As pointed out in the comments, dropping the mutex may improve performance in some cases, by reducing the number of syscalls needed. Then again it could also lead to extra contention on the condition variable's internal mutex. Hard to say. It's probably not worth worrying about in any case.
Note that the applicable standards require that pthread_cond_signal be safely callable without holding the mutex:
The pthread_cond_signal() or pthread_cond_broadcast() 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 [...]
This usually means that condition variables have an internal lock over their internal data structures, or otherwise use some very careful lock-free algorithm.
The state must be modified inside a mutex, if for no other reason than the possibility of spurious wake-ups, which would lead to the reader reading the state while the writer is in the middle of writing it.
You can call pthread_cond_signal anytime after the state is changed. It doesn't have to be inside the mutex. POSIX guarantees that at least one waiter will awaken to check the new state. More to the point:
Calling pthread_cond_signal doesn't guarantee that a reader will acquire the mutex first. Another writer might get in before a reader gets a chance to check the new status. Condition variables don't guarantee that readers immediately follow writers (After all, what if there are no readers?)
Calling it after releasing the lock is actually better, since you don't risk having the just-awoken reader immediately going back to sleep trying to acquire the lock that the writer is still holding.
EDIT: #DietrichEpp makes a good point in the comments. The writer must change the state in such a way that the reader can never access an inconsistent state. It can do so either by acquiring the mutex used in the condition-variable, as I indicate above, or by ensuring that all state-changes are atomic.
The answer is, there is a race, and to eliminate that race, you must do this:
/* atomic op outside of mutex, and then: */
pthread_mutex_lock(&m);
pthread_mutex_unlock(&m);
pthread_cond_signal(&c);
The protection of the data doesn't matter, because you don't hold the mutex when calling pthread_cond_signal anyway.
See, by locking and unlocking the mutex, you have created a barrier. During that brief moment when the signaler has the mutex, there is a certainty: no other thread has the mutex. This means no other thread is executing any critical regions.
This means that all threads are either about to get the mutex to discover the change you have posted, or else they have already found that change and ran off with it (releasing the mutex), or else have not found they are looking for and have atomically given up the mutex to gone to sleep (and are guaranteed to be waiting nicely on the condition).
Without the mutex lock/unlock, you have no synchronization. The signal will sometimes fire as threads which didn't see the changed atomic value are transitioning to their atomic sleep to wait for it.
So this is what the mutex does from the point of view of a thread which is signaling. You can get the atomicity of access from something else, but not the synchronization.
P.S. I have implemented this logic before. The situation was in the Linux kernel (using my own mutexes and condition variables).
In my situation, it was impossible for the signaler to hold the mutex for the atomic operation on shared data. Why? Because the signaler did the operation in user space, inside a buffer shared between the kernel and user, and then (in some situations) made a system call into the kernel to wake up a thread. User space simply made some modifications to the buffer, and then if some conditions were satisfied, it would perform an ioctl.
So in the ioctl call I did the mutex lock/unlock thing, and then hit the condition variable. This ensured that the thread would not miss the wake up related to that latest modification posted by user space.
At first I just had the condition variable signal, but it looked wrong without the involvement of the mutex, so I reasoned about the situation a little bit and realized that the mutex must simply be locked and unlocked to conform to the synchronization ritual which eliminates the lost wakeup.

Deferred bcast wakeup for condition variables - is it valid?

I'm implementing pthread condition variables (based on Linux futexes) and I have an idea for avoiding the "stampede effect" on pthread_cond_broadcast with process-shared condition variables. For non-process-shared cond vars, futex requeue operations are traditionally (i.e. by NPTL) used to requeue waiters from the cond var's futex to the mutex's futex without waking them up, but this is in general impossible for process-shared cond vars, because pthread_cond_broadcast might not have a valid pointer to the associated mutex. In a worst case scenario, the mutex might not even be mapped in its memory space.
My idea for overcoming this issue is to have pthread_cond_broadcast only directly wake one waiter, and have that waiter perform the requeue operation when it wakes up, since it does have the needed pointer to the mutex.
Naturally there are a lot of ugly race conditions to consider if I pursue this approach, but if they can be overcome, are there any other reasons such an implementation would be invalid or undesirable? One potential issue I can think of that might not be able to be overcome is the race where the waiter (a separate process) responsible for the requeue gets killed before it can act, but it might be possible to overcome even this by putting the condvar futex in the robust mutex list so that the kernel performs a wake on it when the process dies.
There may be waiters belonging to multiple address spaces, each of which has mapped the mutex associated with the futex at a different address in memory. I'm not sure if FUTEX_REQUEUE is safe to use when the requeue point may not be mapped at the same address in all waiters; if it does then this isn't a problem.
There are other problems that won't be detected by robust futexes; for example, if your chosen waiter is busy in a signal handler, you could be kept waiting an arbitrarily long time. [As discussed in the comments, these are not an issue]
Note that with robust futexes, you must set the value of the futex & 0x3FFFFFFF to be the TID of the thread to be woken up; you must also set bit FUTEX_WAITERS on if you want a wakeup. This means that you must choose which thread to awaken from the broadcasting thread, or you will be unable to deal with thread death immediately after the FUTEX_WAKE. You'll also need to deal with the possibility of the thread dying immediately before the waker thread writes its TID into the state variable - perhaps having a 'pending master' field that is also registered in the robust mutex system would be a good idea.
I see no reason why this can't work, then, as long as you make sure to deal with the thread exit issues carefully. That said, it may be best to simply define in the kernel an extension to FUTEX_WAIT that takes a requeue point and comparison value as an argument, and let the kernel handle this in a simple, race-free manner.
I just don't see why you assume that the corresponding mutex might not be known. It is clearly stated
The effect of using more than one mutex for concurrent
pthread_cond_timedwait() or pthread_cond_wait() operations on the same
condition variable
is undefined; that is, a condition variable becomes bound to
a unique mutex when a thread waits on the condition variable, and this
(dynamic)
binding shall end when the wait returns.
So even for process shared mutexes and conditions this must hold, and any user space process must always have mapped the same and unique mutex that is associated to the condition.
Allowing users to associate different mutexes to a condition at the same time is nothing that I would support.

Resources