ThreadLocal

folly/ThreadLocal.h

Improved thread local storage for non-trivial types.

• ~4x faster than boost::thread_specific_ptr.
• Similar speed as using pthread_getspecific directly, but only consumes a single pthread_key_t per Tag template param.
• Expands on the thread_specific_ptr API with accessAllThreads and extended custom deleter support.

Usage

---

The API of ThreadLocalPtr is very close to boost::thread_specific_ptr with the notable addition of the accessAllThreadsmethod. There is also a ThreadLocal class which is a thin wrapper around ThreadLocalPtr that manages allocation automatically (creates a new object the first time it is dereferenced from each thread).

ThreadLocalPtr simply gives you a place to put and access a pointer local to each thread such that it will be destroyed appropriately.

{
  folly::ThreadLocalPtr<Widget> w;
  w.reset(new Widget(0), Widget::customDeleterA);
  std::thread([&w]() {
    w.reset(new Widget(1), Widget::customDeleterB);
    w.get()->mangleWidget();
  } // Widget(1) is destroyed with customDeleterB
} // Widget(0) is destroyed with customDeleterA

Note that customDeleterB will get called with TLPDestructionMode::THIS_THREAD and customerDeleterA will get called withTLPDestructionMode::ALL_THREADS. This is to distinguish between thread exit vs. the entire ThreadLocalPtr getting destroyed, in which case there is cleanup work that may be avoided.

The accessAllThreads interface is provided to walk all the thread local child objects of a parent. accessAllThreads initializes an accessor which holds a global lock that blocks all creation and destruction of ThreadLocal objects with the same Tag and can be used as an iterable container. Typical use is for frequent write, infrequent read data access patterns such as counters. Note that you must specify a unique Tag type so you don't block other ThreadLocal object usage, and you should try to minimize the lifetime of the accessor so the lock is held for as short as possible).

The following example is a simplification of folly/ThreadCachedInt.h. It keeps track of a counter value and allows multiple threads to add to the count without synchronization. In order to get the total count, read() iterates through all the thread local values via accessAllThreads() and sums them up. class NewTag is used to break the global mutex so that this class won't block other ThreadLocal usage when read() is called.

Note that read() holds the global mutex which blocks construction, destruction, and read() for other SimpleThreadCachedInt's, but does not block add(). Also, since it uses the unique NewTag, SimpleThreadCachedInt does not affect other ThreadLocal usage.

class SimpleThreadCachedInt {

  class NewTag;  // Segments the global mutex
  ThreadLocal<int,NewTag> val_;

 public:
  void add(int val) {
    *val_ += val;  // operator*() gives a reference to the thread local instance
  }

  int read() {
    int ret = 0;
    // accessAllThreads acquires the global lock
    for (const auto& i : val_.accessAllThreads()) {
      ret += i;
    }  // Global lock is released on scope exit
    return ret;
  }
};

Implementation

---

We keep a __thread array of pointers to objects (ThreadEntry::elements) where each array has an index for each unique instance of the ThreadLocalPtr object. Each ThreadLocalPtr object has a unique id that is an index into these arrays so we can fetch the correct object from thread local storage very efficiently.

In order to prevent unbounded growth of the id space and thus huge ThreadEntry::elements arrays, for example due to continuous creation and destruction of ThreadLocalPtr objects, we keep track of all active instances by linking them together into a list. When an instance is destroyed we remove it from the chain and insert the id into freeIds_ for reuse. These operations require a global mutex, but only happen at construction and destruction time. accessAllThreads also acquires this global mutex.

We use a single global pthread_key_t per Tag to manage object destruction and memory cleanup upon thread exit because there is a finite number of pthread_key_t's available per machine.