Memory Management

Python uses a combination of reference counting and a generational garbage collector to manage memory automatically. Understanding these mechanisms helps you write efficient code and debug memory leaks.

Reference Counting

Every Python object maintains a reference count. When the count drops to zero, the object is immediately deallocated. This is the primary memory management mechanism in CPython.

Garbage Collector (GC)

Reference counting alone cannot handle cyclic references (A references B, B references A). Python's cyclic GC runs periodically and uses three generations. Objects surviving collection are promoted to older generations, which are scanned less frequently.

Memory Pools

CPython uses a private heap with a pymalloc allocator for objects under 512 bytes. This reduces overhead from calling the OS allocator for every small object. Larger allocations go directly to the system malloc.

Common Pitfalls

• Mutable default arguments: Default arguments are created once and shared across calls.
• Forgotten references in caches: Storing objects in a global dict prevents deallocation forever.
• Circular references with __del__: Prior to Python 3.4, objects in reference cycles with __del__ methods were never collected.