allocPool: Add option to disable the false sharing mitigations.

README.md

@@ -6,16 +6,22 @@ constructor (with threads) then offers you two functions: `getPtr()` and `return
 
 Using C++ concepts, we can use templates and require the class given to have a
 default constructor and to have a `.reset()` function. It will be used to clean the
-objects before giving them to another caller.
+objects before giving them to another caller instead of reallocating.
 
 We avoid false sharing by keeping a high amount of work per thread. This should
-lead to cache lines not being shared between threads. While this pool uses a hashmap 
-and a pivot to make `returnPtr(ptr)` extremely fast, the construction's main bottleneck is
-in the locking and unlocking of the hashmap's mutex. We need to do this since we cannot
-write in a `std::unordered_map` at different hashes concurrently.
+lead to cache lines not being shared between threads. In the advent that cache
+lines would overlap, we use a smaller amount of threads with a bit more work
+per thread. This option is enabled with the `bool enableFalseSharingMitigations`
+parameter in the constructor. It is on by default and can be disabled in the advent
+that you very big objects where writing pointers in the `std::vector` will not be the 
+bottleneck.
 
-It will automatically grow when the max capacity is reached, though there will
-be a performance penalty.
+While this pool uses a hashmap and a pivot to make `returnPtr(ptr)` extremely fast, 
+when saving the pointers, the main bottleneck is in the locking and unlocking of the 
+hashmap's mutex. We need to do this since we cannot write in a `std::unordered_map`
+at different hashes concurrently.
+
+It will automatically grow to twice its size when the max capacity is reached.
 
 ## Performance
 With a simple stub class and a pool of 10000 objects, using the pool to take a pointer 

allocPool.hpp

@@ -20,8 +20,8 @@ template<class T>
     requires std::default_initializable<T> && resetable<T>
 class allocPool {
 public:
-    explicit allocPool(size_t allocNumbers)
-        : vec(allocNumbers), pivot{allocNumbers} {
+    explicit allocPool(size_t allocNumbers, bool enableFalseSharingMitigations = true)
+        : vec(allocNumbers), pivot{allocNumbers}, falseSharingMitigations{enableFalseSharingMitigations} {
         memset(&(vec[0]), 0, sizeof(vec[0]) * vec.size());
         initArray(allocNumbers);
     }
@@ -57,6 +57,7 @@ private:
     std::mutex positionMapMutex;
     std::unordered_map<T *, size_t> positionMap;
     size_t pivot;
+    bool falseSharingMitigations;
 
     void initArray(size_t arrSize) {
         size_t amountOfThreads{std::thread::hardware_concurrency()};
@@ -68,8 +69,10 @@ private:
         threads.reserve(amountOfThreads);
 
         // We try to avoid false sharing by defining a minimum size.
-        amountPerThread = minObjPerThread > amountPerThread ? minObjPerThread : amountPerThread;
-        amountOfThreads = arrSize / amountPerThread;
+        if (falseSharingMitigations) {
+            amountPerThread = minObjPerThread > amountPerThread ? minObjPerThread : amountPerThread;
+            amountOfThreads = arrSize / amountPerThread;
+        }
 
         for (size_t i{}; i < amountOfThreads; i++)
             threads.emplace_back(&allocPool::initObjects, this, i * amountPerThread, amountPerThread);