1# SPDX-License-Identifier: GPL-2.0-only
2
3menu "Memory Management options"
4
5#
6# For some reason microblaze and nios2 hard code SWAP=n.  Hopefully we can
7# add proper SWAP support to them, in which case this can be remove.
8#
9config ARCH_NO_SWAP
10	bool
11
12config ZPOOL
13	bool
14
15menuconfig SWAP
16	bool "Support for paging of anonymous memory (swap)"
17	depends on MMU && BLOCK && !ARCH_NO_SWAP
18	default y
19	help
20	  This option allows you to choose whether you want to have support
21	  for so called swap devices or swap files in your kernel that are
22	  used to provide more virtual memory than the actual RAM present
23	  in your computer.  If unsure say Y.
24
25config ZSWAP
26	bool "Compressed cache for swap pages"
27	depends on SWAP
28	select FRONTSWAP
29	select CRYPTO
30	select ZPOOL
31	help
32	  A lightweight compressed cache for swap pages.  It takes
33	  pages that are in the process of being swapped out and attempts to
34	  compress them into a dynamically allocated RAM-based memory pool.
35	  This can result in a significant I/O reduction on swap device and,
36	  in the case where decompressing from RAM is faster than swap device
37	  reads, can also improve workload performance.
38
39config ZSWAP_DEFAULT_ON
40	bool "Enable the compressed cache for swap pages by default"
41	depends on ZSWAP
42	help
43	  If selected, the compressed cache for swap pages will be enabled
44	  at boot, otherwise it will be disabled.
45
46	  The selection made here can be overridden by using the kernel
47	  command line 'zswap.enabled=' option.
48
49choice
50	prompt "Default compressor"
51	depends on ZSWAP
52	default ZSWAP_COMPRESSOR_DEFAULT_LZO
53	help
54	  Selects the default compression algorithm for the compressed cache
55	  for swap pages.
56
57	  For an overview what kind of performance can be expected from
58	  a particular compression algorithm please refer to the benchmarks
59	  available at the following LWN page:
60	  https://lwn.net/Articles/751795/
61
62	  If in doubt, select 'LZO'.
63
64	  The selection made here can be overridden by using the kernel
65	  command line 'zswap.compressor=' option.
66
67config ZSWAP_COMPRESSOR_DEFAULT_DEFLATE
68	bool "Deflate"
69	select CRYPTO_DEFLATE
70	help
71	  Use the Deflate algorithm as the default compression algorithm.
72
73config ZSWAP_COMPRESSOR_DEFAULT_LZO
74	bool "LZO"
75	select CRYPTO_LZO
76	help
77	  Use the LZO algorithm as the default compression algorithm.
78
79config ZSWAP_COMPRESSOR_DEFAULT_842
80	bool "842"
81	select CRYPTO_842
82	help
83	  Use the 842 algorithm as the default compression algorithm.
84
85config ZSWAP_COMPRESSOR_DEFAULT_LZ4
86	bool "LZ4"
87	select CRYPTO_LZ4
88	help
89	  Use the LZ4 algorithm as the default compression algorithm.
90
91config ZSWAP_COMPRESSOR_DEFAULT_LZ4HC
92	bool "LZ4HC"
93	select CRYPTO_LZ4HC
94	help
95	  Use the LZ4HC algorithm as the default compression algorithm.
96
97config ZSWAP_COMPRESSOR_DEFAULT_ZSTD
98	bool "zstd"
99	select CRYPTO_ZSTD
100	help
101	  Use the zstd algorithm as the default compression algorithm.
102endchoice
103
104config ZSWAP_COMPRESSOR_DEFAULT
105       string
106       depends on ZSWAP
107       default "deflate" if ZSWAP_COMPRESSOR_DEFAULT_DEFLATE
108       default "lzo" if ZSWAP_COMPRESSOR_DEFAULT_LZO
109       default "842" if ZSWAP_COMPRESSOR_DEFAULT_842
110       default "lz4" if ZSWAP_COMPRESSOR_DEFAULT_LZ4
111       default "lz4hc" if ZSWAP_COMPRESSOR_DEFAULT_LZ4HC
112       default "zstd" if ZSWAP_COMPRESSOR_DEFAULT_ZSTD
113       default ""
114
115choice
116	prompt "Default allocator"
117	depends on ZSWAP
118	default ZSWAP_ZPOOL_DEFAULT_ZBUD
119	help
120	  Selects the default allocator for the compressed cache for
121	  swap pages.
122	  The default is 'zbud' for compatibility, however please do
123	  read the description of each of the allocators below before
124	  making a right choice.
125
126	  The selection made here can be overridden by using the kernel
127	  command line 'zswap.zpool=' option.
128
129config ZSWAP_ZPOOL_DEFAULT_ZBUD
130	bool "zbud"
131	select ZBUD
132	help
133	  Use the zbud allocator as the default allocator.
134
135config ZSWAP_ZPOOL_DEFAULT_Z3FOLD
136	bool "z3fold"
137	select Z3FOLD
138	help
139	  Use the z3fold allocator as the default allocator.
140
141config ZSWAP_ZPOOL_DEFAULT_ZSMALLOC
142	bool "zsmalloc"
143	select ZSMALLOC
144	help
145	  Use the zsmalloc allocator as the default allocator.
146endchoice
147
148config ZSWAP_ZPOOL_DEFAULT
149       string
150       depends on ZSWAP
151       default "zbud" if ZSWAP_ZPOOL_DEFAULT_ZBUD
152       default "z3fold" if ZSWAP_ZPOOL_DEFAULT_Z3FOLD
153       default "zsmalloc" if ZSWAP_ZPOOL_DEFAULT_ZSMALLOC
154       default ""
155
156config ZBUD
157	tristate "2:1 compression allocator (zbud)"
158	depends on ZSWAP
159	help
160	  A special purpose allocator for storing compressed pages.
161	  It is designed to store up to two compressed pages per physical
162	  page.  While this design limits storage density, it has simple and
163	  deterministic reclaim properties that make it preferable to a higher
164	  density approach when reclaim will be used.
165
166config Z3FOLD
167	tristate "3:1 compression allocator (z3fold)"
168	depends on ZSWAP
169	help
170	  A special purpose allocator for storing compressed pages.
171	  It is designed to store up to three compressed pages per physical
172	  page. It is a ZBUD derivative so the simplicity and determinism are
173	  still there.
174
175config ZSMALLOC
176	tristate
177	prompt "N:1 compression allocator (zsmalloc)" if ZSWAP
178	depends on MMU
179	help
180	  zsmalloc is a slab-based memory allocator designed to store
181	  pages of various compression levels efficiently. It achieves
182	  the highest storage density with the least amount of fragmentation.
183
184config ZSMALLOC_STAT
185	bool "Export zsmalloc statistics"
186	depends on ZSMALLOC
187	select DEBUG_FS
188	help
189	  This option enables code in the zsmalloc to collect various
190	  statistics about what's happening in zsmalloc and exports that
191	  information to userspace via debugfs.
192	  If unsure, say N.
193
194menu "SLAB allocator options"
195
196choice
197	prompt "Choose SLAB allocator"
198	default SLUB
199	help
200	   This option allows to select a slab allocator.
201
202config SLAB
203	bool "SLAB"
204	depends on !PREEMPT_RT
205	select HAVE_HARDENED_USERCOPY_ALLOCATOR
206	help
207	  The regular slab allocator that is established and known to work
208	  well in all environments. It organizes cache hot objects in
209	  per cpu and per node queues.
210
211config SLUB
212	bool "SLUB (Unqueued Allocator)"
213	select HAVE_HARDENED_USERCOPY_ALLOCATOR
214	help
215	   SLUB is a slab allocator that minimizes cache line usage
216	   instead of managing queues of cached objects (SLAB approach).
217	   Per cpu caching is realized using slabs of objects instead
218	   of queues of objects. SLUB can use memory efficiently
219	   and has enhanced diagnostics. SLUB is the default choice for
220	   a slab allocator.
221
222config SLOB
223	depends on EXPERT
224	bool "SLOB (Simple Allocator)"
225	depends on !PREEMPT_RT
226	help
227	   SLOB replaces the stock allocator with a drastically simpler
228	   allocator. SLOB is generally more space efficient but
229	   does not perform as well on large systems.
230
231endchoice
232
233config SLAB_MERGE_DEFAULT
234	bool "Allow slab caches to be merged"
235	default y
236	depends on SLAB || SLUB
237	help
238	  For reduced kernel memory fragmentation, slab caches can be
239	  merged when they share the same size and other characteristics.
240	  This carries a risk of kernel heap overflows being able to
241	  overwrite objects from merged caches (and more easily control
242	  cache layout), which makes such heap attacks easier to exploit
243	  by attackers. By keeping caches unmerged, these kinds of exploits
244	  can usually only damage objects in the same cache. To disable
245	  merging at runtime, "slab_nomerge" can be passed on the kernel
246	  command line.
247
248config SLAB_FREELIST_RANDOM
249	bool "Randomize slab freelist"
250	depends on SLAB || SLUB
251	help
252	  Randomizes the freelist order used on creating new pages. This
253	  security feature reduces the predictability of the kernel slab
254	  allocator against heap overflows.
255
256config SLAB_FREELIST_HARDENED
257	bool "Harden slab freelist metadata"
258	depends on SLAB || SLUB
259	help
260	  Many kernel heap attacks try to target slab cache metadata and
261	  other infrastructure. This options makes minor performance
262	  sacrifices to harden the kernel slab allocator against common
263	  freelist exploit methods. Some slab implementations have more
264	  sanity-checking than others. This option is most effective with
265	  CONFIG_SLUB.
266
267config SLUB_STATS
268	default n
269	bool "Enable SLUB performance statistics"
270	depends on SLUB && SYSFS
271	help
272	  SLUB statistics are useful to debug SLUBs allocation behavior in
273	  order find ways to optimize the allocator. This should never be
274	  enabled for production use since keeping statistics slows down
275	  the allocator by a few percentage points. The slabinfo command
276	  supports the determination of the most active slabs to figure
277	  out which slabs are relevant to a particular load.
278	  Try running: slabinfo -DA
279
280config SLUB_CPU_PARTIAL
281	default y
282	depends on SLUB && SMP
283	bool "SLUB per cpu partial cache"
284	help
285	  Per cpu partial caches accelerate objects allocation and freeing
286	  that is local to a processor at the price of more indeterminism
287	  in the latency of the free. On overflow these caches will be cleared
288	  which requires the taking of locks that may cause latency spikes.
289	  Typically one would choose no for a realtime system.
290
291endmenu # SLAB allocator options
292
293config SHUFFLE_PAGE_ALLOCATOR
294	bool "Page allocator randomization"
295	default SLAB_FREELIST_RANDOM && ACPI_NUMA
296	help
297	  Randomization of the page allocator improves the average
298	  utilization of a direct-mapped memory-side-cache. See section
299	  5.2.27 Heterogeneous Memory Attribute Table (HMAT) in the ACPI
300	  6.2a specification for an example of how a platform advertises
301	  the presence of a memory-side-cache. There are also incidental
302	  security benefits as it reduces the predictability of page
303	  allocations to compliment SLAB_FREELIST_RANDOM, but the
304	  default granularity of shuffling on the "MAX_ORDER - 1" i.e,
305	  10th order of pages is selected based on cache utilization
306	  benefits on x86.
307
308	  While the randomization improves cache utilization it may
309	  negatively impact workloads on platforms without a cache. For
310	  this reason, by default, the randomization is enabled only
311	  after runtime detection of a direct-mapped memory-side-cache.
312	  Otherwise, the randomization may be force enabled with the
313	  'page_alloc.shuffle' kernel command line parameter.
314
315	  Say Y if unsure.
316
317config COMPAT_BRK
318	bool "Disable heap randomization"
319	default y
320	help
321	  Randomizing heap placement makes heap exploits harder, but it
322	  also breaks ancient binaries (including anything libc5 based).
323	  This option changes the bootup default to heap randomization
324	  disabled, and can be overridden at runtime by setting
325	  /proc/sys/kernel/randomize_va_space to 2.
326
327	  On non-ancient distros (post-2000 ones) N is usually a safe choice.
328
329config MMAP_ALLOW_UNINITIALIZED
330	bool "Allow mmapped anonymous memory to be uninitialized"
331	depends on EXPERT && !MMU
332	default n
333	help
334	  Normally, and according to the Linux spec, anonymous memory obtained
335	  from mmap() has its contents cleared before it is passed to
336	  userspace.  Enabling this config option allows you to request that
337	  mmap() skip that if it is given an MAP_UNINITIALIZED flag, thus
338	  providing a huge performance boost.  If this option is not enabled,
339	  then the flag will be ignored.
340
341	  This is taken advantage of by uClibc's malloc(), and also by
342	  ELF-FDPIC binfmt's brk and stack allocator.
343
344	  Because of the obvious security issues, this option should only be
345	  enabled on embedded devices where you control what is run in
346	  userspace.  Since that isn't generally a problem on no-MMU systems,
347	  it is normally safe to say Y here.
348
349	  See Documentation/admin-guide/mm/nommu-mmap.rst for more information.
350
351config SELECT_MEMORY_MODEL
352	def_bool y
353	depends on ARCH_SELECT_MEMORY_MODEL
354
355choice
356	prompt "Memory model"
357	depends on SELECT_MEMORY_MODEL
358	default SPARSEMEM_MANUAL if ARCH_SPARSEMEM_DEFAULT
359	default FLATMEM_MANUAL
360	help
361	  This option allows you to change some of the ways that
362	  Linux manages its memory internally. Most users will
363	  only have one option here selected by the architecture
364	  configuration. This is normal.
365
366config FLATMEM_MANUAL
367	bool "Flat Memory"
368	depends on !ARCH_SPARSEMEM_ENABLE || ARCH_FLATMEM_ENABLE
369	help
370	  This option is best suited for non-NUMA systems with
371	  flat address space. The FLATMEM is the most efficient
372	  system in terms of performance and resource consumption
373	  and it is the best option for smaller systems.
374
375	  For systems that have holes in their physical address
376	  spaces and for features like NUMA and memory hotplug,
377	  choose "Sparse Memory".
378
379	  If unsure, choose this option (Flat Memory) over any other.
380
381config SPARSEMEM_MANUAL
382	bool "Sparse Memory"
383	depends on ARCH_SPARSEMEM_ENABLE
384	help
385	  This will be the only option for some systems, including
386	  memory hot-plug systems.  This is normal.
387
388	  This option provides efficient support for systems with
389	  holes is their physical address space and allows memory
390	  hot-plug and hot-remove.
391
392	  If unsure, choose "Flat Memory" over this option.
393
394endchoice
395
396config SPARSEMEM
397	def_bool y
398	depends on (!SELECT_MEMORY_MODEL && ARCH_SPARSEMEM_ENABLE) || SPARSEMEM_MANUAL
399
400config FLATMEM
401	def_bool y
402	depends on !SPARSEMEM || FLATMEM_MANUAL
403
404#
405# SPARSEMEM_EXTREME (which is the default) does some bootmem
406# allocations when sparse_init() is called.  If this cannot
407# be done on your architecture, select this option.  However,
408# statically allocating the mem_section[] array can potentially
409# consume vast quantities of .bss, so be careful.
410#
411# This option will also potentially produce smaller runtime code
412# with gcc 3.4 and later.
413#
414config SPARSEMEM_STATIC
415	bool
416
417#
418# Architecture platforms which require a two level mem_section in SPARSEMEM
419# must select this option. This is usually for architecture platforms with
420# an extremely sparse physical address space.
421#
422config SPARSEMEM_EXTREME
423	def_bool y
424	depends on SPARSEMEM && !SPARSEMEM_STATIC
425
426config SPARSEMEM_VMEMMAP_ENABLE
427	bool
428
429config SPARSEMEM_VMEMMAP
430	bool "Sparse Memory virtual memmap"
431	depends on SPARSEMEM && SPARSEMEM_VMEMMAP_ENABLE
432	default y
433	help
434	  SPARSEMEM_VMEMMAP uses a virtually mapped memmap to optimise
435	  pfn_to_page and page_to_pfn operations.  This is the most
436	  efficient option when sufficient kernel resources are available.
437
438config HAVE_MEMBLOCK_PHYS_MAP
439	bool
440
441config HAVE_FAST_GUP
442	depends on MMU
443	bool
444
445# Don't discard allocated memory used to track "memory" and "reserved" memblocks
446# after early boot, so it can still be used to test for validity of memory.
447# Also, memblocks are updated with memory hot(un)plug.
448config ARCH_KEEP_MEMBLOCK
449	bool
450
451# Keep arch NUMA mapping infrastructure post-init.
452config NUMA_KEEP_MEMINFO
453	bool
454
455config MEMORY_ISOLATION
456	bool
457
458# IORESOURCE_SYSTEM_RAM regions in the kernel resource tree that are marked
459# IORESOURCE_EXCLUSIVE cannot be mapped to user space, for example, via
460# /dev/mem.
461config EXCLUSIVE_SYSTEM_RAM
462	def_bool y
463	depends on !DEVMEM || STRICT_DEVMEM
464
465#
466# Only be set on architectures that have completely implemented memory hotplug
467# feature. If you are not sure, don't touch it.
468#
469config HAVE_BOOTMEM_INFO_NODE
470	def_bool n
471
472config ARCH_ENABLE_MEMORY_HOTPLUG
473	bool
474
475config ARCH_ENABLE_MEMORY_HOTREMOVE
476	bool
477
478# eventually, we can have this option just 'select SPARSEMEM'
479menuconfig MEMORY_HOTPLUG
480	bool "Memory hotplug"
481	select MEMORY_ISOLATION
482	depends on SPARSEMEM
483	depends on ARCH_ENABLE_MEMORY_HOTPLUG
484	depends on 64BIT
485	select NUMA_KEEP_MEMINFO if NUMA
486
487if MEMORY_HOTPLUG
488
489config MEMORY_HOTPLUG_DEFAULT_ONLINE
490	bool "Online the newly added memory blocks by default"
491	depends on MEMORY_HOTPLUG
492	help
493	  This option sets the default policy setting for memory hotplug
494	  onlining policy (/sys/devices/system/memory/auto_online_blocks) which
495	  determines what happens to newly added memory regions. Policy setting
496	  can always be changed at runtime.
497	  See Documentation/admin-guide/mm/memory-hotplug.rst for more information.
498
499	  Say Y here if you want all hot-plugged memory blocks to appear in
500	  'online' state by default.
501	  Say N here if you want the default policy to keep all hot-plugged
502	  memory blocks in 'offline' state.
503
504config MEMORY_HOTREMOVE
505	bool "Allow for memory hot remove"
506	select HAVE_BOOTMEM_INFO_NODE if (X86_64 || PPC64)
507	depends on MEMORY_HOTPLUG && ARCH_ENABLE_MEMORY_HOTREMOVE
508	depends on MIGRATION
509
510config MHP_MEMMAP_ON_MEMORY
511	def_bool y
512	depends on MEMORY_HOTPLUG && SPARSEMEM_VMEMMAP
513	depends on ARCH_MHP_MEMMAP_ON_MEMORY_ENABLE
514
515endif # MEMORY_HOTPLUG
516
517# Heavily threaded applications may benefit from splitting the mm-wide
518# page_table_lock, so that faults on different parts of the user address
519# space can be handled with less contention: split it at this NR_CPUS.
520# Default to 4 for wider testing, though 8 might be more appropriate.
521# ARM's adjust_pte (unused if VIPT) depends on mm-wide page_table_lock.
522# PA-RISC 7xxx's spinlock_t would enlarge struct page from 32 to 44 bytes.
523# SPARC32 allocates multiple pte tables within a single page, and therefore
524# a per-page lock leads to problems when multiple tables need to be locked
525# at the same time (e.g. copy_page_range()).
526# DEBUG_SPINLOCK and DEBUG_LOCK_ALLOC spinlock_t also enlarge struct page.
527#
528config SPLIT_PTLOCK_CPUS
529	int
530	default "999999" if !MMU
531	default "999999" if ARM && !CPU_CACHE_VIPT
532	default "999999" if PARISC && !PA20
533	default "999999" if SPARC32
534	default "4"
535
536config ARCH_ENABLE_SPLIT_PMD_PTLOCK
537	bool
538
539#
540# support for memory balloon
541config MEMORY_BALLOON
542	bool
543
544#
545# support for memory balloon compaction
546config BALLOON_COMPACTION
547	bool "Allow for balloon memory compaction/migration"
548	def_bool y
549	depends on COMPACTION && MEMORY_BALLOON
550	help
551	  Memory fragmentation introduced by ballooning might reduce
552	  significantly the number of 2MB contiguous memory blocks that can be
553	  used within a guest, thus imposing performance penalties associated
554	  with the reduced number of transparent huge pages that could be used
555	  by the guest workload. Allowing the compaction & migration for memory
556	  pages enlisted as being part of memory balloon devices avoids the
557	  scenario aforementioned and helps improving memory defragmentation.
558
559#
560# support for memory compaction
561config COMPACTION
562	bool "Allow for memory compaction"
563	def_bool y
564	select MIGRATION
565	depends on MMU
566	help
567	  Compaction is the only memory management component to form
568	  high order (larger physically contiguous) memory blocks
569	  reliably. The page allocator relies on compaction heavily and
570	  the lack of the feature can lead to unexpected OOM killer
571	  invocations for high order memory requests. You shouldn't
572	  disable this option unless there really is a strong reason for
573	  it and then we would be really interested to hear about that at
574	  linux-mm@kvack.org.
575
576config COMPACT_UNEVICTABLE_DEFAULT
577	int
578	depends on COMPACTION
579	default 0 if PREEMPT_RT
580	default 1
581
582#
583# support for free page reporting
584config PAGE_REPORTING
585	bool "Free page reporting"
586	def_bool n
587	help
588	  Free page reporting allows for the incremental acquisition of
589	  free pages from the buddy allocator for the purpose of reporting
590	  those pages to another entity, such as a hypervisor, so that the
591	  memory can be freed within the host for other uses.
592
593#
594# support for page migration
595#
596config MIGRATION
597	bool "Page migration"
598	def_bool y
599	depends on (NUMA || ARCH_ENABLE_MEMORY_HOTREMOVE || COMPACTION || CMA) && MMU
600	help
601	  Allows the migration of the physical location of pages of processes
602	  while the virtual addresses are not changed. This is useful in
603	  two situations. The first is on NUMA systems to put pages nearer
604	  to the processors accessing. The second is when allocating huge
605	  pages as migration can relocate pages to satisfy a huge page
606	  allocation instead of reclaiming.
607
608config DEVICE_MIGRATION
609	def_bool MIGRATION && ZONE_DEVICE
610
611config ARCH_ENABLE_HUGEPAGE_MIGRATION
612	bool
613
614config ARCH_ENABLE_THP_MIGRATION
615	bool
616
617config HUGETLB_PAGE_SIZE_VARIABLE
618	def_bool n
619	help
620	  Allows the pageblock_order value to be dynamic instead of just standard
621	  HUGETLB_PAGE_ORDER when there are multiple HugeTLB page sizes available
622	  on a platform.
623
624	  Note that the pageblock_order cannot exceed MAX_ORDER - 1 and will be
625	  clamped down to MAX_ORDER - 1.
626
627config CONTIG_ALLOC
628	def_bool (MEMORY_ISOLATION && COMPACTION) || CMA
629
630config PHYS_ADDR_T_64BIT
631	def_bool 64BIT
632
633config BOUNCE
634	bool "Enable bounce buffers"
635	default y
636	depends on BLOCK && MMU && HIGHMEM
637	help
638	  Enable bounce buffers for devices that cannot access the full range of
639	  memory available to the CPU. Enabled by default when HIGHMEM is
640	  selected, but you may say n to override this.
641
642config MMU_NOTIFIER
643	bool
644	select SRCU
645	select INTERVAL_TREE
646
647config KSM
648	bool "Enable KSM for page merging"
649	depends on MMU
650	select XXHASH
651	help
652	  Enable Kernel Samepage Merging: KSM periodically scans those areas
653	  of an application's address space that an app has advised may be
654	  mergeable.  When it finds pages of identical content, it replaces
655	  the many instances by a single page with that content, so
656	  saving memory until one or another app needs to modify the content.
657	  Recommended for use with KVM, or with other duplicative applications.
658	  See Documentation/mm/ksm.rst for more information: KSM is inactive
659	  until a program has madvised that an area is MADV_MERGEABLE, and
660	  root has set /sys/kernel/mm/ksm/run to 1 (if CONFIG_SYSFS is set).
661
662config DEFAULT_MMAP_MIN_ADDR
663	int "Low address space to protect from user allocation"
664	depends on MMU
665	default 4096
666	help
667	  This is the portion of low virtual memory which should be protected
668	  from userspace allocation.  Keeping a user from writing to low pages
669	  can help reduce the impact of kernel NULL pointer bugs.
670
671	  For most ia64, ppc64 and x86 users with lots of address space
672	  a value of 65536 is reasonable and should cause no problems.
673	  On arm and other archs it should not be higher than 32768.
674	  Programs which use vm86 functionality or have some need to map
675	  this low address space will need CAP_SYS_RAWIO or disable this
676	  protection by setting the value to 0.
677
678	  This value can be changed after boot using the
679	  /proc/sys/vm/mmap_min_addr tunable.
680
681config ARCH_SUPPORTS_MEMORY_FAILURE
682	bool
683
684config MEMORY_FAILURE
685	depends on MMU
686	depends on ARCH_SUPPORTS_MEMORY_FAILURE
687	bool "Enable recovery from hardware memory errors"
688	select MEMORY_ISOLATION
689	select RAS
690	help
691	  Enables code to recover from some memory failures on systems
692	  with MCA recovery. This allows a system to continue running
693	  even when some of its memory has uncorrected errors. This requires
694	  special hardware support and typically ECC memory.
695
696config HWPOISON_INJECT
697	tristate "HWPoison pages injector"
698	depends on MEMORY_FAILURE && DEBUG_KERNEL && PROC_FS
699	select PROC_PAGE_MONITOR
700
701config NOMMU_INITIAL_TRIM_EXCESS
702	int "Turn on mmap() excess space trimming before booting"
703	depends on !MMU
704	default 1
705	help
706	  The NOMMU mmap() frequently needs to allocate large contiguous chunks
707	  of memory on which to store mappings, but it can only ask the system
708	  allocator for chunks in 2^N*PAGE_SIZE amounts - which is frequently
709	  more than it requires.  To deal with this, mmap() is able to trim off
710	  the excess and return it to the allocator.
711
712	  If trimming is enabled, the excess is trimmed off and returned to the
713	  system allocator, which can cause extra fragmentation, particularly
714	  if there are a lot of transient processes.
715
716	  If trimming is disabled, the excess is kept, but not used, which for
717	  long-term mappings means that the space is wasted.
718
719	  Trimming can be dynamically controlled through a sysctl option
720	  (/proc/sys/vm/nr_trim_pages) which specifies the minimum number of
721	  excess pages there must be before trimming should occur, or zero if
722	  no trimming is to occur.
723
724	  This option specifies the initial value of this option.  The default
725	  of 1 says that all excess pages should be trimmed.
726
727	  See Documentation/admin-guide/mm/nommu-mmap.rst for more information.
728
729config ARCH_WANT_GENERAL_HUGETLB
730	bool
731
732config ARCH_WANTS_THP_SWAP
733	def_bool n
734
735menuconfig TRANSPARENT_HUGEPAGE
736	bool "Transparent Hugepage Support"
737	depends on HAVE_ARCH_TRANSPARENT_HUGEPAGE && !PREEMPT_RT
738	select COMPACTION
739	select XARRAY_MULTI
740	help
741	  Transparent Hugepages allows the kernel to use huge pages and
742	  huge tlb transparently to the applications whenever possible.
743	  This feature can improve computing performance to certain
744	  applications by speeding up page faults during memory
745	  allocation, by reducing the number of tlb misses and by speeding
746	  up the pagetable walking.
747
748	  If memory constrained on embedded, you may want to say N.
749
750if TRANSPARENT_HUGEPAGE
751
752choice
753	prompt "Transparent Hugepage Support sysfs defaults"
754	depends on TRANSPARENT_HUGEPAGE
755	default TRANSPARENT_HUGEPAGE_ALWAYS
756	help
757	  Selects the sysfs defaults for Transparent Hugepage Support.
758
759	config TRANSPARENT_HUGEPAGE_ALWAYS
760		bool "always"
761	help
762	  Enabling Transparent Hugepage always, can increase the
763	  memory footprint of applications without a guaranteed
764	  benefit but it will work automatically for all applications.
765
766	config TRANSPARENT_HUGEPAGE_MADVISE
767		bool "madvise"
768	help
769	  Enabling Transparent Hugepage madvise, will only provide a
770	  performance improvement benefit to the applications using
771	  madvise(MADV_HUGEPAGE) but it won't risk to increase the
772	  memory footprint of applications without a guaranteed
773	  benefit.
774endchoice
775
776config THP_SWAP
777	def_bool y
778	depends on TRANSPARENT_HUGEPAGE && ARCH_WANTS_THP_SWAP && SWAP
779	help
780	  Swap transparent huge pages in one piece, without splitting.
781	  XXX: For now, swap cluster backing transparent huge page
782	  will be split after swapout.
783
784	  For selection by architectures with reasonable THP sizes.
785
786config READ_ONLY_THP_FOR_FS
787	bool "Read-only THP for filesystems (EXPERIMENTAL)"
788	depends on TRANSPARENT_HUGEPAGE && SHMEM
789
790	help
791	  Allow khugepaged to put read-only file-backed pages in THP.
792
793	  This is marked experimental because it is a new feature. Write
794	  support of file THPs will be developed in the next few release
795	  cycles.
796
797endif # TRANSPARENT_HUGEPAGE
798
799#
800# UP and nommu archs use km based percpu allocator
801#
802config NEED_PER_CPU_KM
803	depends on !SMP || !MMU
804	bool
805	default y
806
807config NEED_PER_CPU_EMBED_FIRST_CHUNK
808	bool
809
810config NEED_PER_CPU_PAGE_FIRST_CHUNK
811	bool
812
813config USE_PERCPU_NUMA_NODE_ID
814	bool
815
816config HAVE_SETUP_PER_CPU_AREA
817	bool
818
819config FRONTSWAP
820	bool
821
822config CMA
823	bool "Contiguous Memory Allocator"
824	depends on MMU
825	select MIGRATION
826	select MEMORY_ISOLATION
827	help
828	  This enables the Contiguous Memory Allocator which allows other
829	  subsystems to allocate big physically-contiguous blocks of memory.
830	  CMA reserves a region of memory and allows only movable pages to
831	  be allocated from it. This way, the kernel can use the memory for
832	  pagecache and when a subsystem requests for contiguous area, the
833	  allocated pages are migrated away to serve the contiguous request.
834
835	  If unsure, say "n".
836
837config CMA_DEBUG
838	bool "CMA debug messages (DEVELOPMENT)"
839	depends on DEBUG_KERNEL && CMA
840	help
841	  Turns on debug messages in CMA.  This produces KERN_DEBUG
842	  messages for every CMA call as well as various messages while
843	  processing calls such as dma_alloc_from_contiguous().
844	  This option does not affect warning and error messages.
845
846config CMA_DEBUGFS
847	bool "CMA debugfs interface"
848	depends on CMA && DEBUG_FS
849	help
850	  Turns on the DebugFS interface for CMA.
851
852config CMA_SYSFS
853	bool "CMA information through sysfs interface"
854	depends on CMA && SYSFS
855	help
856	  This option exposes some sysfs attributes to get information
857	  from CMA.
858
859config CMA_AREAS
860	int "Maximum count of the CMA areas"
861	depends on CMA
862	default 19 if NUMA
863	default 7
864	help
865	  CMA allows to create CMA areas for particular purpose, mainly,
866	  used as device private area. This parameter sets the maximum
867	  number of CMA area in the system.
868
869	  If unsure, leave the default value "7" in UMA and "19" in NUMA.
870
871config MEM_SOFT_DIRTY
872	bool "Track memory changes"
873	depends on CHECKPOINT_RESTORE && HAVE_ARCH_SOFT_DIRTY && PROC_FS
874	select PROC_PAGE_MONITOR
875	help
876	  This option enables memory changes tracking by introducing a
877	  soft-dirty bit on pte-s. This bit it set when someone writes
878	  into a page just as regular dirty bit, but unlike the latter
879	  it can be cleared by hands.
880
881	  See Documentation/admin-guide/mm/soft-dirty.rst for more details.
882
883config GENERIC_EARLY_IOREMAP
884	bool
885
886config STACK_MAX_DEFAULT_SIZE_MB
887	int "Default maximum user stack size for 32-bit processes (MB)"
888	default 100
889	range 8 2048
890	depends on STACK_GROWSUP && (!64BIT || COMPAT)
891	help
892	  This is the maximum stack size in Megabytes in the VM layout of 32-bit
893	  user processes when the stack grows upwards (currently only on parisc
894	  arch) when the RLIMIT_STACK hard limit is unlimited.
895
896	  A sane initial value is 100 MB.
897
898config DEFERRED_STRUCT_PAGE_INIT
899	bool "Defer initialisation of struct pages to kthreads"
900	depends on SPARSEMEM
901	depends on !NEED_PER_CPU_KM
902	depends on 64BIT
903	select PADATA
904	help
905	  Ordinarily all struct pages are initialised during early boot in a
906	  single thread. On very large machines this can take a considerable
907	  amount of time. If this option is set, large machines will bring up
908	  a subset of memmap at boot and then initialise the rest in parallel.
909	  This has a potential performance impact on tasks running early in the
910	  lifetime of the system until these kthreads finish the
911	  initialisation.
912
913config PAGE_IDLE_FLAG
914	bool
915	select PAGE_EXTENSION if !64BIT
916	help
917	  This adds PG_idle and PG_young flags to 'struct page'.  PTE Accessed
918	  bit writers can set the state of the bit in the flags so that PTE
919	  Accessed bit readers may avoid disturbance.
920
921config IDLE_PAGE_TRACKING
922	bool "Enable idle page tracking"
923	depends on SYSFS && MMU
924	select PAGE_IDLE_FLAG
925	help
926	  This feature allows to estimate the amount of user pages that have
927	  not been touched during a given period of time. This information can
928	  be useful to tune memory cgroup limits and/or for job placement
929	  within a compute cluster.
930
931	  See Documentation/admin-guide/mm/idle_page_tracking.rst for
932	  more details.
933
934config ARCH_HAS_CACHE_LINE_SIZE
935	bool
936
937config ARCH_HAS_CURRENT_STACK_POINTER
938	bool
939	help
940	  In support of HARDENED_USERCOPY performing stack variable lifetime
941	  checking, an architecture-agnostic way to find the stack pointer
942	  is needed. Once an architecture defines an unsigned long global
943	  register alias named "current_stack_pointer", this config can be
944	  selected.
945
946config ARCH_HAS_PTE_DEVMAP
947	bool
948
949config ARCH_HAS_ZONE_DMA_SET
950	bool
951
952config ZONE_DMA
953	bool "Support DMA zone" if ARCH_HAS_ZONE_DMA_SET
954	default y if ARM64 || X86
955
956config ZONE_DMA32
957	bool "Support DMA32 zone" if ARCH_HAS_ZONE_DMA_SET
958	depends on !X86_32
959	default y if ARM64
960
961config ZONE_DEVICE
962	bool "Device memory (pmem, HMM, etc...) hotplug support"
963	depends on MEMORY_HOTPLUG
964	depends on MEMORY_HOTREMOVE
965	depends on SPARSEMEM_VMEMMAP
966	depends on ARCH_HAS_PTE_DEVMAP
967	select XARRAY_MULTI
968
969	help
970	  Device memory hotplug support allows for establishing pmem,
971	  or other device driver discovered memory regions, in the
972	  memmap. This allows pfn_to_page() lookups of otherwise
973	  "device-physical" addresses which is needed for using a DAX
974	  mapping in an O_DIRECT operation, among other things.
975
976	  If FS_DAX is enabled, then say Y.
977
978#
979# Helpers to mirror range of the CPU page tables of a process into device page
980# tables.
981#
982config HMM_MIRROR
983	bool
984	depends on MMU
985
986config GET_FREE_REGION
987	depends on SPARSEMEM
988	bool
989
990config DEVICE_PRIVATE
991	bool "Unaddressable device memory (GPU memory, ...)"
992	depends on ZONE_DEVICE
993	select GET_FREE_REGION
994
995	help
996	  Allows creation of struct pages to represent unaddressable device
997	  memory; i.e., memory that is only accessible from the device (or
998	  group of devices). You likely also want to select HMM_MIRROR.
999
1000config VMAP_PFN
1001	bool
1002
1003config ARCH_USES_HIGH_VMA_FLAGS
1004	bool
1005config ARCH_HAS_PKEYS
1006	bool
1007
1008config VM_EVENT_COUNTERS
1009	default y
1010	bool "Enable VM event counters for /proc/vmstat" if EXPERT
1011	help
1012	  VM event counters are needed for event counts to be shown.
1013	  This option allows the disabling of the VM event counters
1014	  on EXPERT systems.  /proc/vmstat will only show page counts
1015	  if VM event counters are disabled.
1016
1017config PERCPU_STATS
1018	bool "Collect percpu memory statistics"
1019	help
1020	  This feature collects and exposes statistics via debugfs. The
1021	  information includes global and per chunk statistics, which can
1022	  be used to help understand percpu memory usage.
1023
1024config GUP_TEST
1025	bool "Enable infrastructure for get_user_pages()-related unit tests"
1026	depends on DEBUG_FS
1027	help
1028	  Provides /sys/kernel/debug/gup_test, which in turn provides a way
1029	  to make ioctl calls that can launch kernel-based unit tests for
1030	  the get_user_pages*() and pin_user_pages*() family of API calls.
1031
1032	  These tests include benchmark testing of the _fast variants of
1033	  get_user_pages*() and pin_user_pages*(), as well as smoke tests of
1034	  the non-_fast variants.
1035
1036	  There is also a sub-test that allows running dump_page() on any
1037	  of up to eight pages (selected by command line args) within the
1038	  range of user-space addresses. These pages are either pinned via
1039	  pin_user_pages*(), or pinned via get_user_pages*(), as specified
1040	  by other command line arguments.
1041
1042	  See tools/testing/selftests/vm/gup_test.c
1043
1044comment "GUP_TEST needs to have DEBUG_FS enabled"
1045	depends on !GUP_TEST && !DEBUG_FS
1046
1047config GUP_GET_PTE_LOW_HIGH
1048	bool
1049
1050config ARCH_HAS_PTE_SPECIAL
1051	bool
1052
1053#
1054# Some architectures require a special hugepage directory format that is
1055# required to support multiple hugepage sizes. For example a4fe3ce76
1056# "powerpc/mm: Allow more flexible layouts for hugepage pagetables"
1057# introduced it on powerpc.  This allows for a more flexible hugepage
1058# pagetable layouts.
1059#
1060config ARCH_HAS_HUGEPD
1061	bool
1062
1063config MAPPING_DIRTY_HELPERS
1064        bool
1065
1066config KMAP_LOCAL
1067	bool
1068
1069config KMAP_LOCAL_NON_LINEAR_PTE_ARRAY
1070	bool
1071
1072# struct io_mapping based helper.  Selected by drivers that need them
1073config IO_MAPPING
1074	bool
1075
1076config SECRETMEM
1077	def_bool ARCH_HAS_SET_DIRECT_MAP && !EMBEDDED
1078
1079config ANON_VMA_NAME
1080	bool "Anonymous VMA name support"
1081	depends on PROC_FS && ADVISE_SYSCALLS && MMU
1082
1083	help
1084	  Allow naming anonymous virtual memory areas.
1085
1086	  This feature allows assigning names to virtual memory areas. Assigned
1087	  names can be later retrieved from /proc/pid/maps and /proc/pid/smaps
1088	  and help identifying individual anonymous memory areas.
1089	  Assigning a name to anonymous virtual memory area might prevent that
1090	  area from being merged with adjacent virtual memory areas due to the
1091	  difference in their name.
1092
1093config USERFAULTFD
1094	bool "Enable userfaultfd() system call"
1095	depends on MMU
1096	help
1097	  Enable the userfaultfd() system call that allows to intercept and
1098	  handle page faults in userland.
1099
1100config HAVE_ARCH_USERFAULTFD_WP
1101	bool
1102	help
1103	  Arch has userfaultfd write protection support
1104
1105config HAVE_ARCH_USERFAULTFD_MINOR
1106	bool
1107	help
1108	  Arch has userfaultfd minor fault support
1109
1110config PTE_MARKER
1111	bool
1112
1113	help
1114	  Allows to create marker PTEs for file-backed memory.
1115
1116config PTE_MARKER_UFFD_WP
1117	bool "Userfaultfd write protection support for shmem/hugetlbfs"
1118	default y
1119	depends on HAVE_ARCH_USERFAULTFD_WP
1120	select PTE_MARKER
1121
1122	help
1123	  Allows to create marker PTEs for userfaultfd write protection
1124	  purposes.  It is required to enable userfaultfd write protection on
1125	  file-backed memory types like shmem and hugetlbfs.
1126
1127# multi-gen LRU {
1128config LRU_GEN
1129	bool "Multi-Gen LRU"
1130	depends on MMU
1131	# make sure folio->flags has enough spare bits
1132	depends on 64BIT || !SPARSEMEM || SPARSEMEM_VMEMMAP
1133	help
1134	  A high performance LRU implementation to overcommit memory. See
1135	  Documentation/admin-guide/mm/multigen_lru.rst for details.
1136
1137config LRU_GEN_ENABLED
1138	bool "Enable by default"
1139	depends on LRU_GEN
1140	help
1141	  This option enables the multi-gen LRU by default.
1142
1143config LRU_GEN_STATS
1144	bool "Full stats for debugging"
1145	depends on LRU_GEN
1146	help
1147	  Do not enable this option unless you plan to look at historical stats
1148	  from evicted generations for debugging purpose.
1149
1150	  This option has a per-memcg and per-node memory overhead.
1151# }
1152
1153source "mm/damon/Kconfig"
1154
1155endmenu
1156