.. _glossary-p: ============================= Memory Management Glossary: P ============================= .. include:: alphabet.txt .. glossary:: padding Padding is redundant :term:`memory (2)` within the memory :term:`allocated` to an :term:`object`. It is usually inserted because of :term:`alignment` restrictions on the fields of the object or on the object itself. Padding is a form of :term:`internal fragmentation`. padding method .. mps:specific:: A :term:`format method` that is called by a :term:`moving ` :term:`pool ` to create a :term:`padding object`. See :c:type:`mps_fmt_pad_t`. padding object .. mps:specific:: A :term:`formatted object` that consists of :term:`padding`. One of three types of formatted objects, the other two being :term:`client objects` and :term:`forwarding objects`. page A :term:`virtual memory` system usually deals with :term:`memory (1)` :term:`blocks` of fixed size as units for :term:`paging`. These are known as *pages*. Pages are often 4 :term:`kB ` or 8 kB in size. This size is determined by the addressing hardware of the machine. page fault An exception when accessing :term:`virtual memory`, usually resulting in a :term:`page` being fetched from disk. A page fault is an exception occurring during the translation of :term:`virtual addresses` to :term:`physical addresses`. "Page fault" usually means an access to a page that has been :term:`paged out` and hence requires fetching from disk, but it is sometimes also used to mean :term:`invalid page fault` or :term:`protection fault`. .. seealso:: :term:`paging`, :term:`paged in`, :term:`paged out`, :term:`read fault`, :term:`write fault`. page marking Page marking is a form of :term:`card-marking ` where the :term:`card` is the same size as a :term:`page` page protection .. see:: :term:`protection`. Many operating systems support protection of :term:`memory (2)` :term:`pages`. Individual pages may be protected against a combination of read, write or execute accesses by a process. page table In a :term:`virtual memory` system, it is common to map between :term:`virtual addresses` and :term:`physical addresses` by means of a data structure called a *page table*. The :term:`page` number of an address is usually found from the most significant bits of the address; the remaining bits yield the offset of the :term:`memory location` within the page. The page table is normally indexed by page number and contains information on whether the page is currently in :term:`main memory`, and where it is in main memory or on disk. Conventional page tables are sized to the virtual :term:`address space` and store the entire virtual address space description of each process. Because of the need to keep the virtual-to-physical translation time low, a conventional page table is structured as a fixed, multi-level hierarchy, and can be very inefficient at representing a sparse virtual address space, unless the allocated pages are carefully aligned to the page table hierarchy. .. seealso:: :term:`inverted page table`. paged in In a :term:`virtual memory` system, :term:`memory (2)` is described as *paged in* if it is available in :term:`physical memory (1)`. .. similar:: :term:`swapped in`. .. opposite:: :term:`paged out`. .. seealso:: :term:`paging`. paged out In a :term:`virtual memory` system, :term:`memory (2)` is described as *paged out* if it is not available in :term:`physical memory (1)`. .. similar:: :term:`swapped out`. .. opposite:: :term:`paged in`. .. seealso:: :term:`paging`. paging In a :term:`virtual memory` system, *paging* is the act of transferring :term:`pages` between :term:`physical memory (1)` and :term:`backing store` (usually disk). When pages need to be paged out, a heuristic is used to select ones that will not be needed soon; "least recently used" is a popular one. .. similar:: :term:`swapping`. .. seealso:: :term:`paged in`, :term:`paged out`. palimpsest A :term:`block` of :term:`memory (2)` that has been :term:`allocated`, :term:`freed (1)` (or :term:`reclaimed`), and then allocated again. Such memory may contain data from the previous use if portions of it remain uninitialised. This commonly occurs on the :term:`stack`, especially if the compiler allocates large :term:`stack frames` in anticipation of allocating data structures on the stack. If the palimpsest is being :term:`scanned ` :term:`conservatively `, such left-over data may cause :term:`unreachable` :term:`objects` to appear :term:`reachable` and thus become :term:`floating garbage`. If it is scanned :term:`precisely `, such left-over data, if treated as :term:`pointers`, is a bug. parallel garbage collection .. aka:: *concurrent garbage collection*. A parallel or concurrent :term:`collector (2)` executes simultaneously with the :term:`mutator`, usually on a multi-processor machine. Concurrent :term:`garbage collection` must cope with the mutator changing :term:`objects` while collection occurs. The problem is similar to that of :term:`incremental GC `, but harder. The solution typically involves :term:`barrier (1)`. .. similar:: :term:`incremental `. .. seealso:: :term:`replicating garbage collector`. .. bibref:: :ref:`Doligez & Leroy (1993) `, :ref:`Doligez & Gonthier (1994) `. parked state .. mps:specific:: One of the three states an :term:`arena` can be in (the others being the :term:`clamped state` and the :term:`unclamped state`). In the parked state, no :term:`garbage collection` is in progress, no object motion occurs and the staleness of :term:`location dependencies` does not change. Call :c:func:`mps_arena_park` or :c:func:`mps_arena_collect` to put an arena into the parked state. perfect fit If an :term:`allocation ` request is satisfied exactly from a :term:`free block` with no :term:`fragmentation`, this is said to be a :term:`perfect fit`. .. seealso:: :term:`free block`, :term:`allocation mechanism`, :term:`best fit`. phantom reachable phantomly reachable In :term:`Java`, an object is *phantom reachable* if it is neither :term:`strongly ` nor :term:`softly ` nor :term:`weakly reachable` and has been :term:`finalized ` and there is a path from the :term:`roots` to it that contains at least one :term:`phantom reference`. When the Java :term:`collector (1)` determines that an object is phantom reachable, the :term:`reference objects` containing the phantom references are enqueued. The Java specification says that the phantom reference is not cleared when the reference object is enqueued, but actually, there's no way in the language to tell whether that has been done or not. In some implementations, JNI weak global references are weaker than phantom references, and provide a way to access phantom reachable objects. .. seealso:: :term:`reachability `. .. link:: `Class java.lang.ref.PhantomReference `_, `Reference Objects and Garbage Collection `_. phantom reference In :term:`Java` terminology, *phantom reference* is used to mean a :term:`reference` encapsulated in a :term:`reference object` of class ``PhantomReference``. Phantom references form one of three kinds of :term:`weak reference (1)` in Java. They are handy for performing clean-ups after an object has :term:`died ` and been :term:`finalized `. .. seealso:: :term:`phantom reachable`. .. link:: `Class java.lang.ref.PhantomReference `_, `Reference Objects and Garbage Collection `_. physical address .. aka:: *absolute address*. Physical :term:`addresses` are used to index into :term:`physical memory (1)`. On some systems, they are called *absolute addresses*. In a :term:`virtual memory` system the application program handles :term:`virtual addresses` and these are translated to physical addresses by the :term:`MMU`. .. opposite:: :term:`virtual address`. physical address space The physical :term:`address space` is the space of :term:`physical addresses`. .. opposite:: :term:`virtual address space`. physical memory (1) .. aka:: *real memory*. Physical memory is :term:`memory (1)` that is wired to directly to the processor, addressable by :term:`physical address`. This term is basically synonymous to :term:`main memory`, but is used in contrast to :term:`virtual memory` and :term:`backing store`. While modern computers usually have lots of :term:`virtual memory`, performance is still closely related to the quantity of physical memory available. If a system has insufficient physical memory, it may :term:`thrash`. .. similar:: :term:`main memory`. physical memory (2) .. aka:: *physical storage*. Physical memory is :term:`memory (1)` on physical storage devices, such as :term:`RAM` or disks. This term is often contrasted to :term:`virtual address space` that might not be mapped to any actual storage. .. similar:: :term:`memory (1)`. physical storage .. see:: :term:`physical memory (2)`. pig in the python .. aka:: *pig in the snake*. In a :term:`generational ` collector, when a large and long-lived :term:`object` is :term:`allocated` in :term:`nursery space`, collection effort will be wasted as that object survives and is :term:`promoted ` from :term:`generation` to generation. This is especially noticeable in a :term:`copying collector `, where the large object will be copied many times. This difficulty is similar to that of a python which swallows its prey whole and is somewhat immobilized as it digests it. Modern collectors permit objects to be allocated directly into appropriate generations or pools to avoid this problem. Long-lived objects can be allocated directly into long-term generations. Large objects can be allocated directly into pools with special support for large objects (such as copying by remapping, incremental copying, or not copying at all). .. seealso:: :term:`generational garbage collection`. pig in the snake .. see:: :term:`pig in the python`. pinning .. aka:: *nailing*. In :term:`copying garbage collection`, an object may not be movable because it is the target of an :term:`ambiguous reference` or because it is referenced by :term:`foreign code` that does not co-operate with the collector. Such an object is said to be *pinned*. placement policy .. see:: :term:`allocation policy`. platform .. mps:specific:: The term *platform* is used to refer to the combination of operating system, processor architecture, and compiler. See :ref:`topic-platform`. plinth .. mps:specific:: The plinth is a program module providing the MPS with all the support functions it needs from the execution environment. The plinth removes the need for external libraries, by getting the support from the :term:`client program`. See :ref:`topic-plinth`. pointer *Pointer* data types represent a reference to an :term:`object` or a :term:`location `. Pointers may be specialized by the type of the object referred to. Typically, pointers are represented by an :term:`address`, but they can be more complicated when they need to carry more information. For example, when the referent is smaller than a :term:`word`, an offset within the word might be needed. .. similar:: :term:`reference`, :term:`address`. .. seealso:: :term:`tag`. pool .. mps:specific:: A pool is responsible for requesting memory from the :term:`arena` and making it available to the :term:`client program` via :c:func:`mps_alloc` or via an :term:`allocation point`. Multiple pools can coexist in one arena. Pools belong to the type :c:type:`mps_pool_t`. See :ref:`topic-pool` and the :ref:`pool`. pool class .. mps:specific:: A value of type :c:type:`mps_class_t` describing a class of :term:`pools` that manage memory according to particular policy. See :ref:`pool`. precise garbage collection .. see:: :term:`exact garbage collection`. precise reference .. see:: :term:`exact reference`. precise root .. see:: :term:`exact root`. premature free .. aka:: *use after free*. A *premature free* or *use after free* occurs when :term:`memory (2)` is :term:`deallocated `, but is later accessed. Under :term:`manual memory management`, this usually occurs when one part of a program decides it has finished using a memory :term:`block`, and is unaware that another part of the program is still using it. This is rare under :term:`automatic memory management`. .. seealso:: :term:`double free`. premature promotion .. see:: :term:`premature tenuring`. premature tenuring .. aka:: *premature promotion*. When a short-lived :term:`object` :term:`allocated` in a :term:`generational garbage collector ` is :term:`promoted ` (due to poor timing) into a less-frequently collected :term:`generation`. This *prematurely tenured* object may become :term:`garbage` very soon after promotion, but will not be :term:`reclaimed` for some time because it is now in a less frequently collected generation. This problem is essentially due to quantization error: all objects in a generation are treated as if they have the same age, even though they range from as old as the previous promotion cycle to new-born. Modern :term:`collectors (1) ` offer several remedies for premature tenuring. If the client program knows that it is entering a phase that will create many short-lived objects, it can forestall all promotion until it knows it is done with those objects. Thus no objects will be prematurely promoted: they will all be seen as garbage. Another solution is to create :term:`buckets` within generations to more accurately classify objects by age and only promote those which have reached a certain minimum. primary storage .. see:: :term:`main memory`. promotion .. aka:: *tenuring*. Promotion or tenuring is the act of moving an :term:`object` from its current :term:`generation` to an *older* one (one that contains objects that are expected to survive longer). "Tenuring" is used particularly about promotion to the oldest generation. .. seealso:: :term:`generational garbage collection`. protectable root .. mps:specific:: A :term:`root` which the MPS may :term:`protect ` with a :term:`write barrier`. A protectable root is created by specifying the :term:`root mode` :c:macro:`MPS_RM_PROT` when calling a registration function such as :c:func:`mps_root_create`. protection .. aka:: *memory protection*, *page protection*. Many operating systems support protection of :term:`memory (2)` :term:`pages`. Individual pages may be protected against a combination of read, write or execute accesses by a process. A process which attempts a protected access will trigger a :term:`protection fault`. Protection is typically implemented in hardware by the :term:`MMU` as part of the support for :term:`virtual memory` . Pages can be protected for a number of reasons: a :term:`generational ` or :term:`incremental ` :term:`garbage collector` may want to place :term:`barriers (1)` on pages; an operating system may want to protect pages for security, or to implement "copy-on-write" or "demand-zero-filled" pages. .. seealso:: :term:`read fault`, :term:`write fault`. .. bibref:: :ref:`Appel et al. (1988) `, :ref:`Singhal et al. (1992) `, :ref:`Hosking & Moss (1993) `. protection exception .. see:: :term:`protection fault`. protection fault .. aka:: *barrier hit*, *protection exception*, *protection violation*. A protection fault is an exception or trap which occurs when a process attempts to access :term:`memory (2)` which has been :term:`protected `. .. relevance:: Some :term:`garbage collectors` use handlers for protection faults to provide :term:`barriers (1)`. .. seealso:: :term:`segmentation violation`, :term:`General Protection Fault`. protection violation .. see:: :term:`protection fault`.