admin-guide/mm/concepts.rst

41 The physical system memory is divided into page frames, or pages. The
48 pages. These mappings are described by page tables that allow
53 addresses of actual pages used by the software. The tables at higher
54 levels contain physical addresses of the pages belonging to the lower
64 Huge Pages
75 Many modern CPU architectures allow mapping of the memory pages
77 it is possible to map 2M and even 1G pages using entries in the second
78 and the third level page tables. In Linux such pages are called
79 `huge`. Usage of huge pages significantly reduces pressure on TLB,
83 memory with the huge pages. The first one is `HugeTLB filesystem`, or
86 the memory and mapped using huge pages. The hugetlbfs is described at
89 Another, more recent, mechanism that enables use of the huge pages is
92 the system memory should and can be mapped by the huge pages, THP
105 groups memory pages into `zones` according to their possible
109 contain normally addressed pages.
123 own set of zones, lists of free and used pages and various statistics
136 storage device. The written pages are marked as `dirty` and when Linux
163 memory management. The pages that can be freed at any time, either
167 reclaimable pages are page cache and anonymous memory.
169 In most cases, the pages holding internal kernel data and used as DMA
171 their user. Such pages are called `unreclaimable`. However, in certain
172 circumstances, even pages occupied with kernel data structures can be
178 The process of freeing the reclaimable physical memory pages and
180 pages either asynchronously or synchronously, depending on the state
183 pages supply. As the load increases, the amount of the free pages goes
186 asynchronously scan memory pages and either just free them if the data
188 device (remember those dirty pages?). As memory usage increases even
191 until enough memory pages are reclaimed to satisfy the request.
198 scattered physical pages as virtually contiguous range, sometimes it is
202 addresses the fragmentation issue. This mechanism moves occupied pages
203 from the lower part of a memory zone to free pages in the upper part
204 of the zone. When a compaction scan is finished free pages are grouped