SpeichermanagementSubsystemen

Memory management, or Speichermanagement in German, is the set of techniques used by a computer system to allocate, track, and reclaim memory. Its primary goals are to provide each process with a usable address space, protect memory from unauthorized access, and use physical memory efficiently to sustain performance.

Allocation and deallocation handle dynamic requests. The stack supports automatic, LIFO allocation for function calls, while the heap provides a general-purpose area for dynamically allocated objects via allocators. Common strategies include first-fit or best-fit placement, buddy systems, slab allocators, and arena allocators. Deallocation returns memory to the pool, which may trigger compaction or reuse.

Virtual memory is a central concept in modern systems. The operating system translates virtual addresses to physical addresses, typically using paging or segmentation. A page table records mappings, and a translation lookaside buffer accelerates repeated translations. Memory protection bits enforce isolation between processes and prevent interventions such as buffer overflows. Paging also enables overcommitment and the illusion of larger address spaces.

Fragmentation is a persistent challenge. External fragmentation arises when free memory becomes divided into small blocks; internal fragmentation occurs when allocated blocks are larger than requested. Techniques such as paging, compaction, or slab allocation mitigate fragmentation and improve locality.

Garbage collection is used in managed runtimes to reclaim memory automatically by identifying unreachable objects. Various algorithms exist, including mark-sweep, copying collectors, and generational collectors, each balancing pause times, throughput, and memory footprint.

In practice, memory management blends hardware support, such as the MMU and TLB, with software strategies across operating systems, runtimes, and applications. Efficient memory management is critical for performance, reliability, and scalability.