Lines Matching +full:memory +full:- +full:to +full:- +full:memory
1 # SPDX-License-Identifier: GPL-2.0-only
3 menu "Memory Management options"
10 prompt "Memory model"
16 This option allows you to change some of the ways that
17 Linux manages its memory internally. Most users will
22 bool "Flat Memory"
25 This option is best suited for non-NUMA systems with
31 spaces and for features like NUMA and memory hotplug,
32 choose "Sparse Memory".
34 If unsure, choose this option (Flat Memory) over any other.
37 bool "Discontiguous Memory"
41 memory systems, over FLATMEM. These systems have holes
45 Although "Discontiguous Memory" is still used by several
47 "Sparse Memory".
49 If unsure, choose "Sparse Memory" over this option.
52 bool "Sparse Memory"
56 memory hot-plug systems. This is normal.
59 holes is their physical address space and allows memory
60 hot-plug and hot-remove.
62 If unsure, choose "Flat Memory" over this option.
72 File-LRU is a mechanism that put file page in global lru list,
80 Add more attributes in memory cgroup, these attribute is used
81 to show information, shrink memory, swapin page and so on.
84 bool "Enable zswapd thread to reclaim anon pages in background"
91 zram will exchanged to eswap by a certain percentage.
111 # to represent different areas of memory. This variable allows
112 # those dependencies to exist individually.
144 bool "Sparse Memory virtual memmap"
148 SPARSEMEM_VMEMMAP uses a virtually mapped memmap to optimise
162 # Don't discard allocated memory used to track "memory" and "reserved" memblocks
163 # after early boot, so it can still be used to test for validity of memory.
164 # Also, memblocks are updated with memory hot(un)plug.
168 # Keep arch NUMA mapping infrastructure post-init.
176 # Only be set on architectures that have completely implemented memory hotplug
184 bool "Allow for memory hot-add"
196 bool "Online the newly added memory blocks by default"
199 This option sets the default policy setting for memory hotplug
200 onlining policy (/sys/devices/system/memory/auto_online_blocks) which
201 determines what happens to newly added memory regions. Policy setting
203 See Documentation/admin-guide/mm/memory-hotplug.rst for more information.
205 Say Y here if you want all hot-plugged memory blocks to appear in
207 Say N here if you want the default policy to keep all hot-plugged
208 memory blocks in 'offline' state.
211 bool "Allow for memory hot remove"
216 # Heavily threaded applications may benefit from splitting the mm-wide
219 # Default to 4 for wider testing, though 8 might be more appropriate.
220 # ARM's adjust_pte (unused if VIPT) depends on mm-wide page_table_lock.
221 # PA-RISC 7xxx's spinlock_t would enlarge struct page from 32 to 44 bytes.
223 # a per-page lock leads to problems when multiple tables need to be locked
239 # support for memory balloon
244 # support for memory balloon compaction
246 bool "Allow for balloon memory compaction/migration"
250 Memory fragmentation introduced by ballooning might reduce
251 significantly the number of 2MB contiguous memory blocks that can be
254 by the guest workload. Allowing the compaction & migration for memory
255 pages enlisted as being part of memory balloon devices avoids the
256 scenario aforementioned and helps improving memory defragmentation.
259 # support for memory compaction
261 bool "Allow for memory compaction"
266 Compaction is the only memory management component to form
267 high order (larger physically contiguous) memory blocks
269 the lack of the feature can lead to unexpected OOM killer
270 invocations for high order memory requests. You shouldn't
272 it and then we would be really interested to hear about that at
273 linux-mm@kvack.org.
283 those pages to another entity, such as a hypervisor, so that the
284 memory can be freed within the host for other uses.
296 two situations. The first is on NUMA systems to put pages nearer
297 to the processors accessing. The second is when allocating huge
298 pages as migration can relocate pages to satisfy a huge page
319 the full range of memory available to the CPU. Enabled
321 may say n to override this.
345 saving memory until one or another app needs to modify the content.
349 root has set /sys/kernel/mm/ksm/run to 1 (if CONFIG_SYSFS is set).
352 int "Low address space to protect from user allocation"
356 This is the portion of low virtual memory which should be protected
357 from userspace allocation. Keeping a user from writing to low pages
363 Programs which use vm86 functionality or have some need to map
365 protection by setting the value to 0.
376 bool "Enable recovery from hardware memory errors"
380 Enables code to recover from some memory failures on systems
381 with MCA recovery. This allows a system to continue running
382 even when some of its memory has uncorrected errors. This requires
383 special hardware support and typically ECC memory.
395 The NOMMU mmap() frequently needs to allocate large contiguous chunks
396 of memory on which to store mappings, but it can only ask the system
397 allocator for chunks in 2^N*PAGE_SIZE amounts - which is frequently
398 more than it requires. To deal with this, mmap() is able to trim off
399 the excess and return it to the allocator.
401 If trimming is enabled, the excess is trimmed off and returned to the
406 long-term mappings means that the space is wasted.
411 no trimming is to occur.
416 See Documentation/admin-guide/mm/nommu-mmap.rst for more information.
424 Transparent Hugepages allows the kernel to use huge pages and
425 huge tlb transparently to the applications whenever possible.
426 This feature can improve computing performance to certain
427 applications by speeding up page faults during memory
431 If memory constrained on embedded, you may want to say N.
444 memory footprint of applications without a guaranteed
451 performance improvement benefit to the applications using
452 madvise(MADV_HUGEPAGE) but it won't risk to increase the
453 memory footprint of applications without a guaranteed
479 bool "Enable cleancache driver to cache clean pages if tmem is present"
481 Cleancache can be thought of as a page-granularity victim cache
483 (PFRA) would like to keep around, but can't since there isn't enough
484 memory. So when the PFRA "evicts" a page, it first attempts to use
485 cleancache code to put the data contained in that page into
486 "transcendent memory", memory that is not directly accessible or
488 time-varying size. And when a cleancache-enabled
489 filesystem wishes to access a page in a file on disk, it first
490 checks cleancache to see if it already contains it; if it does,
492 When a transcendent memory driver is available (such as zcache or
493 Xen transcendent memory), a significant I/O reduction
495 are reduced to a single pointer-compare-against-NULL resulting
498 If unsure, say Y to enable cleancache
501 bool "Enable frontswap to cache swap pages if tmem is present"
506 "transcendent memory", memory that is not directly accessible or
508 time-varying size. When space in transcendent memory is available,
510 available, all frontswap calls are reduced to a single pointer-
511 compare-against-NULL resulting in a negligible performance hit
514 If unsure, say Y to enable frontswap.
517 bool "Contiguous Memory Allocator"
522 This enables the Contiguous Memory Allocator which allows other
523 subsystems to allocate big physically-contiguous blocks of memory.
524 CMA reserves a region of memory and allows only movable pages to
525 be allocated from it. This way, the kernel can use the memory for
527 allocated pages are migrated away to serve the contiguous request.
552 CMA allows to create CMA areas for particular purpose, mainly,
562 If enabled, it will add MIGRATE_CMA to pcp lists and movable
563 allocations with __GFP_CMA flag will use cma areas prior to
569 bool "Track memory changes"
573 This option enables memory changes tracking by introducing a
574 soft-dirty bit on pte-s. This bit it set when someone writes
578 See Documentation/admin-guide/mm/soft-dirty.rst for more details.
586 pages that are in the process of being swapped out and attempts to
587 compress them into a dynamically allocated RAM-based memory pool.
593 v3.11) that interacts heavily with memory reclaim. While these
594 interactions don't cause any known issues on simple memory setups,
607 a particular compression algorithm please refer to the benchmarks
716 tristate "Common API for compressed memory storage"
718 Compressed memory storage API. This allows using either zbud or
722 tristate "Low (Up to 2x) density storage for compressed pages"
725 It is designed to store up to two compressed pages per physical
727 deterministic reclaim properties that make it preferable to a higher
731 tristate "Up to 3x density storage for compressed pages"
735 It is designed to store up to three compressed pages per physical
740 tristate "Memory allocator for compressed pages"
743 zsmalloc is a slab-based memory allocator designed to store
744 compressed RAM pages. zsmalloc uses virtual memory mapping
745 in order to reduce fragmentation. However, this results in a
746 non-standard allocator interface where a handle, not a pointer, is
747 returned by an alloc(). This handle must be mapped in order to
755 This option enables code in the zsmalloc to collect various
757 information to userspace via debugfs.
764 int "Maximum user stack size for 32-bit processes (MB)"
769 This is the maximum stack size in Megabytes in the VM layout of 32-bit
771 arch). The stack will be located at the highest memory address minus
772 the given value, unless the RLIMIT_STACK hard limit is changed to a
778 bool "Defer initialisation of struct pages to kthreads"
797 This feature allows to estimate the amount of user pages that have
799 be useful to tune memory cgroup limits and/or for job placement
802 See Documentation/admin-guide/mm/idle_page_tracking.rst for
809 bool "Device memory (pmem, HMM, etc...) hotplug support"
817 Device memory hotplug support allows for establishing pmem,
818 or other device driver discovered memory regions, in the
820 "device-physical" addresses which is needed for using a DAX
829 # Helpers to mirror range of the CPU page tables of a process into device page
837 bool "Unaddressable device memory (GPU memory, ...)"
842 Allows creation of struct pages to represent unaddressable device
843 memory; i.e., memory that is only accessible from the device (or
844 group of devices). You likely also want to select HMM_MIRROR.
858 bool "Collect percpu memory statistics"
862 be used to help understand percpu memory usage.
876 bool "Read-only THP for filesystems (EXPERIMENTAL)"
880 Allow khugepaged to put read-only file-backed pages in THP.
891 # required to support multiple hugepage sizes. For example a4fe3ce76
907 Allow naming anonymous virtual memory areas.
909 This feature allows assigning names to virtual memory areas. Assigned
911 and help identifying individual anonymous memory areas.
912 Assigning a name to anonymous virtual memory area might prevent that
913 area from being merged with adjacent virtual memory areas due to the