1 config SELECT_MEMORY_MODEL
3 depends on ARCH_SELECT_MEMORY_MODEL
7 depends on SELECT_MEMORY_MODEL
8 default DISCONTIGMEM_MANUAL if ARCH_DISCONTIGMEM_DEFAULT
9 default SPARSEMEM_MANUAL if ARCH_SPARSEMEM_DEFAULT
10 default FLATMEM_MANUAL
14 depends on !(ARCH_DISCONTIGMEM_ENABLE || ARCH_SPARSEMEM_ENABLE) || ARCH_FLATMEM_ENABLE
16 This option allows you to change some of the ways that
17 Linux manages its memory internally. Most users will
18 only have one option here: FLATMEM. This is normal
21 Some users of more advanced features like NUMA and
22 memory hotplug may have different options here.
23 DISCONTIGMEM is a more mature, better tested system,
24 but is incompatible with memory hotplug and may suffer
25 decreased performance over SPARSEMEM. If unsure between
26 "Sparse Memory" and "Discontiguous Memory", choose
27 "Discontiguous Memory".
29 If unsure, choose this option (Flat Memory) over any other.
31 config DISCONTIGMEM_MANUAL
32 bool "Discontiguous Memory"
33 depends on ARCH_DISCONTIGMEM_ENABLE
35 This option provides enhanced support for discontiguous
36 memory systems, over FLATMEM. These systems have holes
37 in their physical address spaces, and this option provides
38 more efficient handling of these holes. However, the vast
39 majority of hardware has quite flat address spaces, and
40 can have degraded performance from the extra overhead that
43 Many NUMA configurations will have this as the only option.
45 If unsure, choose "Flat Memory" over this option.
47 config SPARSEMEM_MANUAL
49 depends on ARCH_SPARSEMEM_ENABLE
51 This will be the only option for some systems, including
52 memory hotplug systems. This is normal.
54 For many other systems, this will be an alternative to
55 "Discontiguous Memory". This option provides some potential
56 performance benefits, along with decreased code complexity,
57 but it is newer, and more experimental.
59 If unsure, choose "Discontiguous Memory" or "Flat Memory"
66 depends on (!SELECT_MEMORY_MODEL && ARCH_DISCONTIGMEM_ENABLE) || DISCONTIGMEM_MANUAL
70 depends on (!SELECT_MEMORY_MODEL && ARCH_SPARSEMEM_ENABLE) || SPARSEMEM_MANUAL
74 depends on (!DISCONTIGMEM && !SPARSEMEM) || FLATMEM_MANUAL
76 config FLAT_NODE_MEM_MAP
81 # Both the NUMA code and DISCONTIGMEM use arrays of pg_data_t's
82 # to represent different areas of memory. This variable allows
83 # those dependencies to exist individually.
85 config NEED_MULTIPLE_NODES
87 depends on DISCONTIGMEM || NUMA
89 config HAVE_MEMORY_PRESENT
91 depends on ARCH_HAVE_MEMORY_PRESENT || SPARSEMEM
94 # SPARSEMEM_EXTREME (which is the default) does some bootmem
95 # allocations when memory_present() is called. If this cannot
96 # be done on your architecture, select this option. However,
97 # statically allocating the mem_section[] array can potentially
98 # consume vast quantities of .bss, so be careful.
100 # This option will also potentially produce smaller runtime code
101 # with gcc 3.4 and later.
103 config SPARSEMEM_STATIC
107 # Architecture platforms which require a two level mem_section in SPARSEMEM
108 # must select this option. This is usually for architecture platforms with
109 # an extremely sparse physical address space.
111 config SPARSEMEM_EXTREME
113 depends on SPARSEMEM && !SPARSEMEM_STATIC
115 config SPARSEMEM_VMEMMAP_ENABLE
118 config SPARSEMEM_ALLOC_MEM_MAP_TOGETHER
120 depends on SPARSEMEM && X86_64
122 config SPARSEMEM_VMEMMAP
123 bool "Sparse Memory virtual memmap"
124 depends on SPARSEMEM && SPARSEMEM_VMEMMAP_ENABLE
127 SPARSEMEM_VMEMMAP uses a virtually mapped memmap to optimise
128 pfn_to_page and page_to_pfn operations. This is the most
129 efficient option when sufficient kernel resources are available.
134 config HAVE_MEMBLOCK_NODE_MAP
137 config HAVE_MEMBLOCK_PHYS_MAP
140 config HAVE_GENERIC_GUP
143 config ARCH_DISCARD_MEMBLOCK
149 config MEMORY_ISOLATION
153 # Only be set on architectures that have completely implemented memory hotplug
154 # feature. If you are not sure, don't touch it.
156 config HAVE_BOOTMEM_INFO_NODE
159 # eventually, we can have this option just 'select SPARSEMEM'
160 config MEMORY_HOTPLUG
161 bool "Allow for memory hot-add"
162 depends on SPARSEMEM || X86_64_ACPI_NUMA
163 depends on ARCH_ENABLE_MEMORY_HOTPLUG
165 config MEMORY_HOTPLUG_SPARSE
167 depends on SPARSEMEM && MEMORY_HOTPLUG
169 config MEMORY_HOTPLUG_DEFAULT_ONLINE
170 bool "Online the newly added memory blocks by default"
172 depends on MEMORY_HOTPLUG
174 This option sets the default policy setting for memory hotplug
175 onlining policy (/sys/devices/system/memory/auto_online_blocks) which
176 determines what happens to newly added memory regions. Policy setting
177 can always be changed at runtime.
178 See Documentation/memory-hotplug.txt for more information.
180 Say Y here if you want all hot-plugged memory blocks to appear in
181 'online' state by default.
182 Say N here if you want the default policy to keep all hot-plugged
183 memory blocks in 'offline' state.
185 config MEMORY_HOTREMOVE
186 bool "Allow for memory hot remove"
187 select MEMORY_ISOLATION
188 select HAVE_BOOTMEM_INFO_NODE if (X86_64 || PPC64)
189 depends on MEMORY_HOTPLUG && ARCH_ENABLE_MEMORY_HOTREMOVE
192 # Heavily threaded applications may benefit from splitting the mm-wide
193 # page_table_lock, so that faults on different parts of the user address
194 # space can be handled with less contention: split it at this NR_CPUS.
195 # Default to 4 for wider testing, though 8 might be more appropriate.
196 # ARM's adjust_pte (unused if VIPT) depends on mm-wide page_table_lock.
197 # PA-RISC 7xxx's spinlock_t would enlarge struct page from 32 to 44 bytes.
198 # DEBUG_SPINLOCK and DEBUG_LOCK_ALLOC spinlock_t also enlarge struct page.
200 config SPLIT_PTLOCK_CPUS
202 default "999999" if !MMU
203 default "999999" if ARM && !CPU_CACHE_VIPT
204 default "999999" if PARISC && !PA20
207 config ARCH_ENABLE_SPLIT_PMD_PTLOCK
211 # support for memory balloon
212 config MEMORY_BALLOON
216 # support for memory balloon compaction
217 config BALLOON_COMPACTION
218 bool "Allow for balloon memory compaction/migration"
220 depends on COMPACTION && MEMORY_BALLOON
222 Memory fragmentation introduced by ballooning might reduce
223 significantly the number of 2MB contiguous memory blocks that can be
224 used within a guest, thus imposing performance penalties associated
225 with the reduced number of transparent huge pages that could be used
226 by the guest workload. Allowing the compaction & migration for memory
227 pages enlisted as being part of memory balloon devices avoids the
228 scenario aforementioned and helps improving memory defragmentation.
231 # support for memory compaction
233 bool "Allow for memory compaction"
238 Compaction is the only memory management component to form
239 high order (larger physically contiguous) memory blocks
240 reliably. The page allocator relies on compaction heavily and
241 the lack of the feature can lead to unexpected OOM killer
242 invocations for high order memory requests. You shouldn't
243 disable this option unless there really is a strong reason for
244 it and then we would be really interested to hear about that at
248 # support for page migration
251 bool "Page migration"
253 depends on (NUMA || ARCH_ENABLE_MEMORY_HOTREMOVE || COMPACTION || CMA) && MMU
255 Allows the migration of the physical location of pages of processes
256 while the virtual addresses are not changed. This is useful in
257 two situations. The first is on NUMA systems to put pages nearer
258 to the processors accessing. The second is when allocating huge
259 pages as migration can relocate pages to satisfy a huge page
260 allocation instead of reclaiming.
262 config ARCH_ENABLE_HUGEPAGE_MIGRATION
265 config ARCH_ENABLE_THP_MIGRATION
268 config PHYS_ADDR_T_64BIT
269 def_bool 64BIT || ARCH_PHYS_ADDR_T_64BIT
272 bool "Enable bounce buffers"
274 depends on BLOCK && MMU && (ZONE_DMA || HIGHMEM)
276 Enable bounce buffers for devices that cannot access
277 the full range of memory available to the CPU. Enabled
278 by default when ZONE_DMA or HIGHMEM is selected, but you
279 may say n to override this.
289 An architecture should select this if it implements the
290 deprecated interface virt_to_bus(). All new architectures
291 should probably not select this.
299 bool "Enable KSM for page merging"
302 Enable Kernel Samepage Merging: KSM periodically scans those areas
303 of an application's address space that an app has advised may be
304 mergeable. When it finds pages of identical content, it replaces
305 the many instances by a single page with that content, so
306 saving memory until one or another app needs to modify the content.
307 Recommended for use with KVM, or with other duplicative applications.
308 See Documentation/vm/ksm.txt for more information: KSM is inactive
309 until a program has madvised that an area is MADV_MERGEABLE, and
310 root has set /sys/kernel/mm/ksm/run to 1 (if CONFIG_SYSFS is set).
312 config DEFAULT_MMAP_MIN_ADDR
313 int "Low address space to protect from user allocation"
317 This is the portion of low virtual memory which should be protected
318 from userspace allocation. Keeping a user from writing to low pages
319 can help reduce the impact of kernel NULL pointer bugs.
321 For most ia64, ppc64 and x86 users with lots of address space
322 a value of 65536 is reasonable and should cause no problems.
323 On arm and other archs it should not be higher than 32768.
324 Programs which use vm86 functionality or have some need to map
325 this low address space will need CAP_SYS_RAWIO or disable this
326 protection by setting the value to 0.
328 This value can be changed after boot using the
329 /proc/sys/vm/mmap_min_addr tunable.
331 config ARCH_SUPPORTS_MEMORY_FAILURE
334 config MEMORY_FAILURE
336 depends on ARCH_SUPPORTS_MEMORY_FAILURE
337 bool "Enable recovery from hardware memory errors"
338 select MEMORY_ISOLATION
341 Enables code to recover from some memory failures on systems
342 with MCA recovery. This allows a system to continue running
343 even when some of its memory has uncorrected errors. This requires
344 special hardware support and typically ECC memory.
346 config HWPOISON_INJECT
347 tristate "HWPoison pages injector"
348 depends on MEMORY_FAILURE && DEBUG_KERNEL && PROC_FS
349 select PROC_PAGE_MONITOR
351 config NOMMU_INITIAL_TRIM_EXCESS
352 int "Turn on mmap() excess space trimming before booting"
356 The NOMMU mmap() frequently needs to allocate large contiguous chunks
357 of memory on which to store mappings, but it can only ask the system
358 allocator for chunks in 2^N*PAGE_SIZE amounts - which is frequently
359 more than it requires. To deal with this, mmap() is able to trim off
360 the excess and return it to the allocator.
362 If trimming is enabled, the excess is trimmed off and returned to the
363 system allocator, which can cause extra fragmentation, particularly
364 if there are a lot of transient processes.
366 If trimming is disabled, the excess is kept, but not used, which for
367 long-term mappings means that the space is wasted.
369 Trimming can be dynamically controlled through a sysctl option
370 (/proc/sys/vm/nr_trim_pages) which specifies the minimum number of
371 excess pages there must be before trimming should occur, or zero if
372 no trimming is to occur.
374 This option specifies the initial value of this option. The default
375 of 1 says that all excess pages should be trimmed.
377 See Documentation/nommu-mmap.txt for more information.
379 config TRANSPARENT_HUGEPAGE
380 bool "Transparent Hugepage Support"
381 depends on HAVE_ARCH_TRANSPARENT_HUGEPAGE
383 select RADIX_TREE_MULTIORDER
385 Transparent Hugepages allows the kernel to use huge pages and
386 huge tlb transparently to the applications whenever possible.
387 This feature can improve computing performance to certain
388 applications by speeding up page faults during memory
389 allocation, by reducing the number of tlb misses and by speeding
390 up the pagetable walking.
392 If memory constrained on embedded, you may want to say N.
395 prompt "Transparent Hugepage Support sysfs defaults"
396 depends on TRANSPARENT_HUGEPAGE
397 default TRANSPARENT_HUGEPAGE_ALWAYS
399 Selects the sysfs defaults for Transparent Hugepage Support.
401 config TRANSPARENT_HUGEPAGE_ALWAYS
404 Enabling Transparent Hugepage always, can increase the
405 memory footprint of applications without a guaranteed
406 benefit but it will work automatically for all applications.
408 config TRANSPARENT_HUGEPAGE_MADVISE
411 Enabling Transparent Hugepage madvise, will only provide a
412 performance improvement benefit to the applications using
413 madvise(MADV_HUGEPAGE) but it won't risk to increase the
414 memory footprint of applications without a guaranteed
418 config ARCH_WANTS_THP_SWAP
423 depends on TRANSPARENT_HUGEPAGE && ARCH_WANTS_THP_SWAP
425 Swap transparent huge pages in one piece, without splitting.
426 XXX: For now this only does clustered swap space allocation.
428 For selection by architectures with reasonable THP sizes.
430 config TRANSPARENT_HUGE_PAGECACHE
432 depends on TRANSPARENT_HUGEPAGE
435 # UP and nommu archs use km based percpu allocator
437 config NEED_PER_CPU_KM
443 bool "Enable cleancache driver to cache clean pages if tmem is present"
446 Cleancache can be thought of as a page-granularity victim cache
447 for clean pages that the kernel's pageframe replacement algorithm
448 (PFRA) would like to keep around, but can't since there isn't enough
449 memory. So when the PFRA "evicts" a page, it first attempts to use
450 cleancache code to put the data contained in that page into
451 "transcendent memory", memory that is not directly accessible or
452 addressable by the kernel and is of unknown and possibly
453 time-varying size. And when a cleancache-enabled
454 filesystem wishes to access a page in a file on disk, it first
455 checks cleancache to see if it already contains it; if it does,
456 the page is copied into the kernel and a disk access is avoided.
457 When a transcendent memory driver is available (such as zcache or
458 Xen transcendent memory), a significant I/O reduction
459 may be achieved. When none is available, all cleancache calls
460 are reduced to a single pointer-compare-against-NULL resulting
461 in a negligible performance hit.
463 If unsure, say Y to enable cleancache
466 bool "Enable frontswap to cache swap pages if tmem is present"
470 Frontswap is so named because it can be thought of as the opposite
471 of a "backing" store for a swap device. The data is stored into
472 "transcendent memory", memory that is not directly accessible or
473 addressable by the kernel and is of unknown and possibly
474 time-varying size. When space in transcendent memory is available,
475 a significant swap I/O reduction may be achieved. When none is
476 available, all frontswap calls are reduced to a single pointer-
477 compare-against-NULL resulting in a negligible performance hit
478 and swap data is stored as normal on the matching swap device.
480 If unsure, say Y to enable frontswap.
483 bool "Contiguous Memory Allocator"
484 depends on HAVE_MEMBLOCK && MMU
486 select MEMORY_ISOLATION
488 This enables the Contiguous Memory Allocator which allows other
489 subsystems to allocate big physically-contiguous blocks of memory.
490 CMA reserves a region of memory and allows only movable pages to
491 be allocated from it. This way, the kernel can use the memory for
492 pagecache and when a subsystem requests for contiguous area, the
493 allocated pages are migrated away to serve the contiguous request.
498 bool "CMA debug messages (DEVELOPMENT)"
499 depends on DEBUG_KERNEL && CMA
501 Turns on debug messages in CMA. This produces KERN_DEBUG
502 messages for every CMA call as well as various messages while
503 processing calls such as dma_alloc_from_contiguous().
504 This option does not affect warning and error messages.
507 bool "CMA debugfs interface"
508 depends on CMA && DEBUG_FS
510 Turns on the DebugFS interface for CMA.
513 int "Maximum count of the CMA areas"
517 CMA allows to create CMA areas for particular purpose, mainly,
518 used as device private area. This parameter sets the maximum
519 number of CMA area in the system.
521 If unsure, leave the default value "7".
523 config MEM_SOFT_DIRTY
524 bool "Track memory changes"
525 depends on CHECKPOINT_RESTORE && HAVE_ARCH_SOFT_DIRTY && PROC_FS
526 select PROC_PAGE_MONITOR
528 This option enables memory changes tracking by introducing a
529 soft-dirty bit on pte-s. This bit it set when someone writes
530 into a page just as regular dirty bit, but unlike the latter
531 it can be cleared by hands.
533 See Documentation/vm/soft-dirty.txt for more details.
536 bool "Compressed cache for swap pages (EXPERIMENTAL)"
537 depends on FRONTSWAP && CRYPTO=y
542 A lightweight compressed cache for swap pages. It takes
543 pages that are in the process of being swapped out and attempts to
544 compress them into a dynamically allocated RAM-based memory pool.
545 This can result in a significant I/O reduction on swap device and,
546 in the case where decompressing from RAM is faster that swap device
547 reads, can also improve workload performance.
549 This is marked experimental because it is a new feature (as of
550 v3.11) that interacts heavily with memory reclaim. While these
551 interactions don't cause any known issues on simple memory setups,
552 they have not be fully explored on the large set of potential
553 configurations and workloads that exist.
556 tristate "Common API for compressed memory storage"
559 Compressed memory storage API. This allows using either zbud or
563 tristate "Low (Up to 2x) density storage for compressed pages"
566 A special purpose allocator for storing compressed pages.
567 It is designed to store up to two compressed pages per physical
568 page. While this design limits storage density, it has simple and
569 deterministic reclaim properties that make it preferable to a higher
570 density approach when reclaim will be used.
573 tristate "Up to 3x density storage for compressed pages"
577 A special purpose allocator for storing compressed pages.
578 It is designed to store up to three compressed pages per physical
579 page. It is a ZBUD derivative so the simplicity and determinism are
583 tristate "Memory allocator for compressed pages"
587 zsmalloc is a slab-based memory allocator designed to store
588 compressed RAM pages. zsmalloc uses virtual memory mapping
589 in order to reduce fragmentation. However, this results in a
590 non-standard allocator interface where a handle, not a pointer, is
591 returned by an alloc(). This handle must be mapped in order to
592 access the allocated space.
594 config PGTABLE_MAPPING
595 bool "Use page table mapping to access object in zsmalloc"
598 By default, zsmalloc uses a copy-based object mapping method to
599 access allocations that span two pages. However, if a particular
600 architecture (ex, ARM) performs VM mapping faster than copying,
601 then you should select this. This causes zsmalloc to use page table
602 mapping rather than copying for object mapping.
604 You can check speed with zsmalloc benchmark:
605 https://github.com/spartacus06/zsmapbench
608 bool "Export zsmalloc statistics"
612 This option enables code in the zsmalloc to collect various
613 statistics about whats happening in zsmalloc and exports that
614 information to userspace via debugfs.
617 config GENERIC_EARLY_IOREMAP
620 config MAX_STACK_SIZE_MB
621 int "Maximum user stack size for 32-bit processes (MB)"
624 depends on STACK_GROWSUP && (!64BIT || COMPAT)
626 This is the maximum stack size in Megabytes in the VM layout of 32-bit
627 user processes when the stack grows upwards (currently only on parisc
628 arch). The stack will be located at the highest memory address minus
629 the given value, unless the RLIMIT_STACK hard limit is changed to a
630 smaller value in which case that is used.
632 A sane initial value is 80 MB.
634 config DEFERRED_STRUCT_PAGE_INIT
635 bool "Defer initialisation of struct pages to kthreads"
637 depends on NO_BOOTMEM
639 depends on !NEED_PER_CPU_KM
641 Ordinarily all struct pages are initialised during early boot in a
642 single thread. On very large machines this can take a considerable
643 amount of time. If this option is set, large machines will bring up
644 a subset of memmap at boot and then initialise the rest in parallel
645 by starting one-off "pgdatinitX" kernel thread for each node X. This
646 has a potential performance impact on processes running early in the
647 lifetime of the system until these kthreads finish the
650 config IDLE_PAGE_TRACKING
651 bool "Enable idle page tracking"
652 depends on SYSFS && MMU
653 select PAGE_EXTENSION if !64BIT
655 This feature allows to estimate the amount of user pages that have
656 not been touched during a given period of time. This information can
657 be useful to tune memory cgroup limits and/or for job placement
658 within a compute cluster.
660 See Documentation/vm/idle_page_tracking.txt for more details.
662 # arch_add_memory() comprehends device memory
663 config ARCH_HAS_ZONE_DEVICE
667 bool "Device memory (pmem, HMM, etc...) hotplug support"
668 depends on MEMORY_HOTPLUG
669 depends on MEMORY_HOTREMOVE
670 depends on SPARSEMEM_VMEMMAP
671 depends on ARCH_HAS_ZONE_DEVICE
672 select RADIX_TREE_MULTIORDER
675 Device memory hotplug support allows for establishing pmem,
676 or other device driver discovered memory regions, in the
677 memmap. This allows pfn_to_page() lookups of otherwise
678 "device-physical" addresses which is needed for using a DAX
679 mapping in an O_DIRECT operation, among other things.
681 If FS_DAX is enabled, then say Y.
686 depends on (X86_64 || PPC64)
687 depends on ZONE_DEVICE
688 depends on MMU && 64BIT
689 depends on MEMORY_HOTPLUG
690 depends on MEMORY_HOTREMOVE
691 depends on SPARSEMEM_VMEMMAP
693 config MIGRATE_VMA_HELPER
698 select MIGRATE_VMA_HELPER
701 bool "HMM mirror CPU page table into a device page table"
702 depends on ARCH_HAS_HMM
706 Select HMM_MIRROR if you want to mirror range of the CPU page table of a
707 process into a device page table. Here, mirror means "keep synchronized".
708 Prerequisites: the device must provide the ability to write-protect its
709 page tables (at PAGE_SIZE granularity), and must be able to recover from
710 the resulting potential page faults.
712 config DEVICE_PRIVATE
713 bool "Unaddressable device memory (GPU memory, ...)"
714 depends on ARCH_HAS_HMM
718 Allows creation of struct pages to represent unaddressable device
719 memory; i.e., memory that is only accessible from the device (or
720 group of devices). You likely also want to select HMM_MIRROR.
723 bool "Addressable device memory (like GPU memory)"
724 depends on ARCH_HAS_HMM
728 Allows creation of struct pages to represent addressable device
729 memory; i.e., memory that is accessible from both the device and
735 config ARCH_USES_HIGH_VMA_FLAGS
737 config ARCH_HAS_PKEYS
741 bool "Collect percpu memory statistics"
744 This feature collects and exposes statistics via debugfs. The
745 information includes global and per chunk statistics, which can
746 be used to help understand percpu memory usage.
749 bool "Enable infrastructure for get_user_pages_fast() benchmarking"
752 Provides /sys/kernel/debug/gup_benchmark that helps with testing
753 performance of get_user_pages_fast().
755 See tools/testing/selftests/vm/gup_benchmark.c