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 CRYPTO 29 select ZPOOL 30 help 31 A lightweight compressed cache for swap pages. It takes 32 pages that are in the process of being swapped out and attempts to 33 compress them into a dynamically allocated RAM-based memory pool. 34 This can result in a significant I/O reduction on swap device and, 35 in the case where decompressing from RAM is faster than swap device 36 reads, can also improve workload performance. 37 38config ZSWAP_DEFAULT_ON 39 bool "Enable the compressed cache for swap pages by default" 40 depends on ZSWAP 41 help 42 If selected, the compressed cache for swap pages will be enabled 43 at boot, otherwise it will be disabled. 44 45 The selection made here can be overridden by using the kernel 46 command line 'zswap.enabled=' option. 47 48config ZSWAP_SHRINKER_DEFAULT_ON 49 bool "Shrink the zswap pool on memory pressure" 50 depends on ZSWAP 51 default n 52 help 53 If selected, the zswap shrinker will be enabled, and the pages 54 stored in the zswap pool will become available for reclaim (i.e 55 written back to the backing swap device) on memory pressure. 56 57 This means that zswap writeback could happen even if the pool is 58 not yet full, or the cgroup zswap limit has not been reached, 59 reducing the chance that cold pages will reside in the zswap pool 60 and consume memory indefinitely. 61 62choice 63 prompt "Default compressor" 64 depends on ZSWAP 65 default ZSWAP_COMPRESSOR_DEFAULT_LZO 66 help 67 Selects the default compression algorithm for the compressed cache 68 for swap pages. 69 70 For an overview what kind of performance can be expected from 71 a particular compression algorithm please refer to the benchmarks 72 available at the following LWN page: 73 https://lwn.net/Articles/751795/ 74 75 If in doubt, select 'LZO'. 76 77 The selection made here can be overridden by using the kernel 78 command line 'zswap.compressor=' option. 79 80config ZSWAP_COMPRESSOR_DEFAULT_DEFLATE 81 bool "Deflate" 82 select CRYPTO_DEFLATE 83 help 84 Use the Deflate algorithm as the default compression algorithm. 85 86config ZSWAP_COMPRESSOR_DEFAULT_LZO 87 bool "LZO" 88 select CRYPTO_LZO 89 help 90 Use the LZO algorithm as the default compression algorithm. 91 92config ZSWAP_COMPRESSOR_DEFAULT_842 93 bool "842" 94 select CRYPTO_842 95 help 96 Use the 842 algorithm as the default compression algorithm. 97 98config ZSWAP_COMPRESSOR_DEFAULT_LZ4 99 bool "LZ4" 100 select CRYPTO_LZ4 101 help 102 Use the LZ4 algorithm as the default compression algorithm. 103 104config ZSWAP_COMPRESSOR_DEFAULT_LZ4HC 105 bool "LZ4HC" 106 select CRYPTO_LZ4HC 107 help 108 Use the LZ4HC algorithm as the default compression algorithm. 109 110config ZSWAP_COMPRESSOR_DEFAULT_ZSTD 111 bool "zstd" 112 select CRYPTO_ZSTD 113 help 114 Use the zstd algorithm as the default compression algorithm. 115endchoice 116 117config ZSWAP_COMPRESSOR_DEFAULT 118 string 119 depends on ZSWAP 120 default "deflate" if ZSWAP_COMPRESSOR_DEFAULT_DEFLATE 121 default "lzo" if ZSWAP_COMPRESSOR_DEFAULT_LZO 122 default "842" if ZSWAP_COMPRESSOR_DEFAULT_842 123 default "lz4" if ZSWAP_COMPRESSOR_DEFAULT_LZ4 124 default "lz4hc" if ZSWAP_COMPRESSOR_DEFAULT_LZ4HC 125 default "zstd" if ZSWAP_COMPRESSOR_DEFAULT_ZSTD 126 default "" 127 128choice 129 prompt "Default allocator" 130 depends on ZSWAP 131 default ZSWAP_ZPOOL_DEFAULT_ZSMALLOC if MMU 132 default ZSWAP_ZPOOL_DEFAULT_ZBUD 133 help 134 Selects the default allocator for the compressed cache for 135 swap pages. 136 The default is 'zbud' for compatibility, however please do 137 read the description of each of the allocators below before 138 making a right choice. 139 140 The selection made here can be overridden by using the kernel 141 command line 'zswap.zpool=' option. 142 143config ZSWAP_ZPOOL_DEFAULT_ZBUD 144 bool "zbud" 145 select ZBUD 146 help 147 Use the zbud allocator as the default allocator. 148 149config ZSWAP_ZPOOL_DEFAULT_Z3FOLD 150 bool "z3fold" 151 select Z3FOLD 152 help 153 Use the z3fold allocator as the default allocator. 154 155config ZSWAP_ZPOOL_DEFAULT_ZSMALLOC 156 bool "zsmalloc" 157 select ZSMALLOC 158 help 159 Use the zsmalloc allocator as the default allocator. 160endchoice 161 162config ZSWAP_ZPOOL_DEFAULT 163 string 164 depends on ZSWAP 165 default "zbud" if ZSWAP_ZPOOL_DEFAULT_ZBUD 166 default "z3fold" if ZSWAP_ZPOOL_DEFAULT_Z3FOLD 167 default "zsmalloc" if ZSWAP_ZPOOL_DEFAULT_ZSMALLOC 168 default "" 169 170config ZBUD 171 tristate "2:1 compression allocator (zbud)" 172 depends on ZSWAP 173 help 174 A special purpose allocator for storing compressed pages. 175 It is designed to store up to two compressed pages per physical 176 page. While this design limits storage density, it has simple and 177 deterministic reclaim properties that make it preferable to a higher 178 density approach when reclaim will be used. 179 180config Z3FOLD 181 tristate "3:1 compression allocator (z3fold)" 182 depends on ZSWAP 183 help 184 A special purpose allocator for storing compressed pages. 185 It is designed to store up to three compressed pages per physical 186 page. It is a ZBUD derivative so the simplicity and determinism are 187 still there. 188 189config ZSMALLOC 190 tristate 191 prompt "N:1 compression allocator (zsmalloc)" if ZSWAP 192 depends on MMU 193 help 194 zsmalloc is a slab-based memory allocator designed to store 195 pages of various compression levels efficiently. It achieves 196 the highest storage density with the least amount of fragmentation. 197 198config ZSMALLOC_STAT 199 bool "Export zsmalloc statistics" 200 depends on ZSMALLOC 201 select DEBUG_FS 202 help 203 This option enables code in the zsmalloc to collect various 204 statistics about what's happening in zsmalloc and exports that 205 information to userspace via debugfs. 206 If unsure, say N. 207 208config ZSMALLOC_CHAIN_SIZE 209 int "Maximum number of physical pages per-zspage" 210 default 8 211 range 4 16 212 depends on ZSMALLOC 213 help 214 This option sets the upper limit on the number of physical pages 215 that a zmalloc page (zspage) can consist of. The optimal zspage 216 chain size is calculated for each size class during the 217 initialization of the pool. 218 219 Changing this option can alter the characteristics of size classes, 220 such as the number of pages per zspage and the number of objects 221 per zspage. This can also result in different configurations of 222 the pool, as zsmalloc merges size classes with similar 223 characteristics. 224 225 For more information, see zsmalloc documentation. 226 227menu "Slab allocator options" 228 229config SLUB 230 def_bool y 231 232config SLUB_TINY 233 bool "Configure for minimal memory footprint" 234 depends on EXPERT 235 select SLAB_MERGE_DEFAULT 236 help 237 Configures the slab allocator in a way to achieve minimal memory 238 footprint, sacrificing scalability, debugging and other features. 239 This is intended only for the smallest system that had used the 240 SLOB allocator and is not recommended for systems with more than 241 16MB RAM. 242 243 If unsure, say N. 244 245config SLAB_MERGE_DEFAULT 246 bool "Allow slab caches to be merged" 247 default y 248 help 249 For reduced kernel memory fragmentation, slab caches can be 250 merged when they share the same size and other characteristics. 251 This carries a risk of kernel heap overflows being able to 252 overwrite objects from merged caches (and more easily control 253 cache layout), which makes such heap attacks easier to exploit 254 by attackers. By keeping caches unmerged, these kinds of exploits 255 can usually only damage objects in the same cache. To disable 256 merging at runtime, "slab_nomerge" can be passed on the kernel 257 command line. 258 259config SLAB_FREELIST_RANDOM 260 bool "Randomize slab freelist" 261 depends on !SLUB_TINY 262 help 263 Randomizes the freelist order used on creating new pages. This 264 security feature reduces the predictability of the kernel slab 265 allocator against heap overflows. 266 267config SLAB_FREELIST_HARDENED 268 bool "Harden slab freelist metadata" 269 depends on !SLUB_TINY 270 help 271 Many kernel heap attacks try to target slab cache metadata and 272 other infrastructure. This options makes minor performance 273 sacrifices to harden the kernel slab allocator against common 274 freelist exploit methods. 275 276config SLUB_STATS 277 default n 278 bool "Enable performance statistics" 279 depends on SYSFS && !SLUB_TINY 280 help 281 The statistics are useful to debug slab allocation behavior in 282 order find ways to optimize the allocator. This should never be 283 enabled for production use since keeping statistics slows down 284 the allocator by a few percentage points. The slabinfo command 285 supports the determination of the most active slabs to figure 286 out which slabs are relevant to a particular load. 287 Try running: slabinfo -DA 288 289config SLUB_CPU_PARTIAL 290 default y 291 depends on SMP && !SLUB_TINY 292 bool "Enable per cpu partial caches" 293 help 294 Per cpu partial caches accelerate objects allocation and freeing 295 that is local to a processor at the price of more indeterminism 296 in the latency of the free. On overflow these caches will be cleared 297 which requires the taking of locks that may cause latency spikes. 298 Typically one would choose no for a realtime system. 299 300config RANDOM_KMALLOC_CACHES 301 default n 302 depends on !SLUB_TINY 303 bool "Randomize slab caches for normal kmalloc" 304 help 305 A hardening feature that creates multiple copies of slab caches for 306 normal kmalloc allocation and makes kmalloc randomly pick one based 307 on code address, which makes the attackers more difficult to spray 308 vulnerable memory objects on the heap for the purpose of exploiting 309 memory vulnerabilities. 310 311 Currently the number of copies is set to 16, a reasonably large value 312 that effectively diverges the memory objects allocated for different 313 subsystems or modules into different caches, at the expense of a 314 limited degree of memory and CPU overhead that relates to hardware and 315 system workload. 316 317endmenu # Slab allocator options 318 319config SHUFFLE_PAGE_ALLOCATOR 320 bool "Page allocator randomization" 321 default SLAB_FREELIST_RANDOM && ACPI_NUMA 322 help 323 Randomization of the page allocator improves the average 324 utilization of a direct-mapped memory-side-cache. See section 325 5.2.27 Heterogeneous Memory Attribute Table (HMAT) in the ACPI 326 6.2a specification for an example of how a platform advertises 327 the presence of a memory-side-cache. There are also incidental 328 security benefits as it reduces the predictability of page 329 allocations to compliment SLAB_FREELIST_RANDOM, but the 330 default granularity of shuffling on the MAX_PAGE_ORDER i.e, 10th 331 order of pages is selected based on cache utilization benefits 332 on x86. 333 334 While the randomization improves cache utilization it may 335 negatively impact workloads on platforms without a cache. For 336 this reason, by default, the randomization is not enabled even 337 if SHUFFLE_PAGE_ALLOCATOR=y. The randomization may be force enabled 338 with the 'page_alloc.shuffle' kernel command line parameter. 339 340 Say Y if unsure. 341 342config COMPAT_BRK 343 bool "Disable heap randomization" 344 default y 345 help 346 Randomizing heap placement makes heap exploits harder, but it 347 also breaks ancient binaries (including anything libc5 based). 348 This option changes the bootup default to heap randomization 349 disabled, and can be overridden at runtime by setting 350 /proc/sys/kernel/randomize_va_space to 2. 351 352 On non-ancient distros (post-2000 ones) N is usually a safe choice. 353 354config MMAP_ALLOW_UNINITIALIZED 355 bool "Allow mmapped anonymous memory to be uninitialized" 356 depends on EXPERT && !MMU 357 default n 358 help 359 Normally, and according to the Linux spec, anonymous memory obtained 360 from mmap() has its contents cleared before it is passed to 361 userspace. Enabling this config option allows you to request that 362 mmap() skip that if it is given an MAP_UNINITIALIZED flag, thus 363 providing a huge performance boost. If this option is not enabled, 364 then the flag will be ignored. 365 366 This is taken advantage of by uClibc's malloc(), and also by 367 ELF-FDPIC binfmt's brk and stack allocator. 368 369 Because of the obvious security issues, this option should only be 370 enabled on embedded devices where you control what is run in 371 userspace. Since that isn't generally a problem on no-MMU systems, 372 it is normally safe to say Y here. 373 374 See Documentation/admin-guide/mm/nommu-mmap.rst for more information. 375 376config SELECT_MEMORY_MODEL 377 def_bool y 378 depends on ARCH_SELECT_MEMORY_MODEL 379 380choice 381 prompt "Memory model" 382 depends on SELECT_MEMORY_MODEL 383 default SPARSEMEM_MANUAL if ARCH_SPARSEMEM_DEFAULT 384 default FLATMEM_MANUAL 385 help 386 This option allows you to change some of the ways that 387 Linux manages its memory internally. Most users will 388 only have one option here selected by the architecture 389 configuration. This is normal. 390 391config FLATMEM_MANUAL 392 bool "Flat Memory" 393 depends on !ARCH_SPARSEMEM_ENABLE || ARCH_FLATMEM_ENABLE 394 help 395 This option is best suited for non-NUMA systems with 396 flat address space. The FLATMEM is the most efficient 397 system in terms of performance and resource consumption 398 and it is the best option for smaller systems. 399 400 For systems that have holes in their physical address 401 spaces and for features like NUMA and memory hotplug, 402 choose "Sparse Memory". 403 404 If unsure, choose this option (Flat Memory) over any other. 405 406config SPARSEMEM_MANUAL 407 bool "Sparse Memory" 408 depends on ARCH_SPARSEMEM_ENABLE 409 help 410 This will be the only option for some systems, including 411 memory hot-plug systems. This is normal. 412 413 This option provides efficient support for systems with 414 holes is their physical address space and allows memory 415 hot-plug and hot-remove. 416 417 If unsure, choose "Flat Memory" over this option. 418 419endchoice 420 421config SPARSEMEM 422 def_bool y 423 depends on (!SELECT_MEMORY_MODEL && ARCH_SPARSEMEM_ENABLE) || SPARSEMEM_MANUAL 424 425config FLATMEM 426 def_bool y 427 depends on !SPARSEMEM || FLATMEM_MANUAL 428 429# 430# SPARSEMEM_EXTREME (which is the default) does some bootmem 431# allocations when sparse_init() is called. If this cannot 432# be done on your architecture, select this option. However, 433# statically allocating the mem_section[] array can potentially 434# consume vast quantities of .bss, so be careful. 435# 436# This option will also potentially produce smaller runtime code 437# with gcc 3.4 and later. 438# 439config SPARSEMEM_STATIC 440 bool 441 442# 443# Architecture platforms which require a two level mem_section in SPARSEMEM 444# must select this option. This is usually for architecture platforms with 445# an extremely sparse physical address space. 446# 447config SPARSEMEM_EXTREME 448 def_bool y 449 depends on SPARSEMEM && !SPARSEMEM_STATIC 450 451config SPARSEMEM_VMEMMAP_ENABLE 452 bool 453 454config SPARSEMEM_VMEMMAP 455 bool "Sparse Memory virtual memmap" 456 depends on SPARSEMEM && SPARSEMEM_VMEMMAP_ENABLE 457 default y 458 help 459 SPARSEMEM_VMEMMAP uses a virtually mapped memmap to optimise 460 pfn_to_page and page_to_pfn operations. This is the most 461 efficient option when sufficient kernel resources are available. 462# 463# Select this config option from the architecture Kconfig, if it is preferred 464# to enable the feature of HugeTLB/dev_dax vmemmap optimization. 465# 466config ARCH_WANT_OPTIMIZE_DAX_VMEMMAP 467 bool 468 469config ARCH_WANT_OPTIMIZE_HUGETLB_VMEMMAP 470 bool 471 472config HAVE_MEMBLOCK_PHYS_MAP 473 bool 474 475config HAVE_GUP_FAST 476 depends on MMU 477 bool 478 479# Don't discard allocated memory used to track "memory" and "reserved" memblocks 480# after early boot, so it can still be used to test for validity of memory. 481# Also, memblocks are updated with memory hot(un)plug. 482config ARCH_KEEP_MEMBLOCK 483 bool 484 485# Keep arch NUMA mapping infrastructure post-init. 486config NUMA_KEEP_MEMINFO 487 bool 488 489config MEMORY_ISOLATION 490 bool 491 492# IORESOURCE_SYSTEM_RAM regions in the kernel resource tree that are marked 493# IORESOURCE_EXCLUSIVE cannot be mapped to user space, for example, via 494# /dev/mem. 495config EXCLUSIVE_SYSTEM_RAM 496 def_bool y 497 depends on !DEVMEM || STRICT_DEVMEM 498 499# 500# Only be set on architectures that have completely implemented memory hotplug 501# feature. If you are not sure, don't touch it. 502# 503config HAVE_BOOTMEM_INFO_NODE 504 def_bool n 505 506config ARCH_ENABLE_MEMORY_HOTPLUG 507 bool 508 509config ARCH_ENABLE_MEMORY_HOTREMOVE 510 bool 511 512# eventually, we can have this option just 'select SPARSEMEM' 513menuconfig MEMORY_HOTPLUG 514 bool "Memory hotplug" 515 select MEMORY_ISOLATION 516 depends on SPARSEMEM 517 depends on ARCH_ENABLE_MEMORY_HOTPLUG 518 depends on 64BIT 519 select NUMA_KEEP_MEMINFO if NUMA 520 521if MEMORY_HOTPLUG 522 523config MEMORY_HOTPLUG_DEFAULT_ONLINE 524 bool "Online the newly added memory blocks by default" 525 depends on MEMORY_HOTPLUG 526 help 527 This option sets the default policy setting for memory hotplug 528 onlining policy (/sys/devices/system/memory/auto_online_blocks) which 529 determines what happens to newly added memory regions. Policy setting 530 can always be changed at runtime. 531 See Documentation/admin-guide/mm/memory-hotplug.rst for more information. 532 533 Say Y here if you want all hot-plugged memory blocks to appear in 534 'online' state by default. 535 Say N here if you want the default policy to keep all hot-plugged 536 memory blocks in 'offline' state. 537 538config MEMORY_HOTREMOVE 539 bool "Allow for memory hot remove" 540 select HAVE_BOOTMEM_INFO_NODE if (X86_64 || PPC64) 541 depends on MEMORY_HOTPLUG && ARCH_ENABLE_MEMORY_HOTREMOVE 542 depends on MIGRATION 543 544config MHP_MEMMAP_ON_MEMORY 545 def_bool y 546 depends on MEMORY_HOTPLUG && SPARSEMEM_VMEMMAP 547 depends on ARCH_MHP_MEMMAP_ON_MEMORY_ENABLE 548 549endif # MEMORY_HOTPLUG 550 551config ARCH_MHP_MEMMAP_ON_MEMORY_ENABLE 552 bool 553 554# Heavily threaded applications may benefit from splitting the mm-wide 555# page_table_lock, so that faults on different parts of the user address 556# space can be handled with less contention: split it at this NR_CPUS. 557# Default to 4 for wider testing, though 8 might be more appropriate. 558# ARM's adjust_pte (unused if VIPT) depends on mm-wide page_table_lock. 559# PA-RISC 7xxx's spinlock_t would enlarge struct page from 32 to 44 bytes. 560# SPARC32 allocates multiple pte tables within a single page, and therefore 561# a per-page lock leads to problems when multiple tables need to be locked 562# at the same time (e.g. copy_page_range()). 563# DEBUG_SPINLOCK and DEBUG_LOCK_ALLOC spinlock_t also enlarge struct page. 564# 565config SPLIT_PTLOCK_CPUS 566 int 567 default "999999" if !MMU 568 default "999999" if ARM && !CPU_CACHE_VIPT 569 default "999999" if PARISC && !PA20 570 default "999999" if SPARC32 571 default "4" 572 573config ARCH_ENABLE_SPLIT_PMD_PTLOCK 574 bool 575 576# 577# support for memory balloon 578config MEMORY_BALLOON 579 bool 580 581# 582# support for memory balloon compaction 583config BALLOON_COMPACTION 584 bool "Allow for balloon memory compaction/migration" 585 default y 586 depends on COMPACTION && MEMORY_BALLOON 587 help 588 Memory fragmentation introduced by ballooning might reduce 589 significantly the number of 2MB contiguous memory blocks that can be 590 used within a guest, thus imposing performance penalties associated 591 with the reduced number of transparent huge pages that could be used 592 by the guest workload. Allowing the compaction & migration for memory 593 pages enlisted as being part of memory balloon devices avoids the 594 scenario aforementioned and helps improving memory defragmentation. 595 596# 597# support for memory compaction 598config COMPACTION 599 bool "Allow for memory compaction" 600 default y 601 select MIGRATION 602 depends on MMU 603 help 604 Compaction is the only memory management component to form 605 high order (larger physically contiguous) memory blocks 606 reliably. The page allocator relies on compaction heavily and 607 the lack of the feature can lead to unexpected OOM killer 608 invocations for high order memory requests. You shouldn't 609 disable this option unless there really is a strong reason for 610 it and then we would be really interested to hear about that at 611 linux-mm@kvack.org. 612 613config COMPACT_UNEVICTABLE_DEFAULT 614 int 615 depends on COMPACTION 616 default 0 if PREEMPT_RT 617 default 1 618 619# 620# support for free page reporting 621config PAGE_REPORTING 622 bool "Free page reporting" 623 help 624 Free page reporting allows for the incremental acquisition of 625 free pages from the buddy allocator for the purpose of reporting 626 those pages to another entity, such as a hypervisor, so that the 627 memory can be freed within the host for other uses. 628 629# 630# support for page migration 631# 632config MIGRATION 633 bool "Page migration" 634 default y 635 depends on (NUMA || ARCH_ENABLE_MEMORY_HOTREMOVE || COMPACTION || CMA) && MMU 636 help 637 Allows the migration of the physical location of pages of processes 638 while the virtual addresses are not changed. This is useful in 639 two situations. The first is on NUMA systems to put pages nearer 640 to the processors accessing. The second is when allocating huge 641 pages as migration can relocate pages to satisfy a huge page 642 allocation instead of reclaiming. 643 644config DEVICE_MIGRATION 645 def_bool MIGRATION && ZONE_DEVICE 646 647config ARCH_ENABLE_HUGEPAGE_MIGRATION 648 bool 649 650config ARCH_ENABLE_THP_MIGRATION 651 bool 652 653config HUGETLB_PAGE_SIZE_VARIABLE 654 def_bool n 655 help 656 Allows the pageblock_order value to be dynamic instead of just standard 657 HUGETLB_PAGE_ORDER when there are multiple HugeTLB page sizes available 658 on a platform. 659 660 Note that the pageblock_order cannot exceed MAX_PAGE_ORDER and will be 661 clamped down to MAX_PAGE_ORDER. 662 663config CONTIG_ALLOC 664 def_bool (MEMORY_ISOLATION && COMPACTION) || CMA 665 666config PCP_BATCH_SCALE_MAX 667 int "Maximum scale factor of PCP (Per-CPU pageset) batch allocate/free" 668 default 5 669 range 0 6 670 help 671 In page allocator, PCP (Per-CPU pageset) is refilled and drained in 672 batches. The batch number is scaled automatically to improve page 673 allocation/free throughput. But too large scale factor may hurt 674 latency. This option sets the upper limit of scale factor to limit 675 the maximum latency. 676 677config PHYS_ADDR_T_64BIT 678 def_bool 64BIT 679 680config BOUNCE 681 bool "Enable bounce buffers" 682 default y 683 depends on BLOCK && MMU && HIGHMEM 684 help 685 Enable bounce buffers for devices that cannot access the full range of 686 memory available to the CPU. Enabled by default when HIGHMEM is 687 selected, but you may say n to override this. 688 689config MMU_NOTIFIER 690 bool 691 select INTERVAL_TREE 692 693config KSM 694 bool "Enable KSM for page merging" 695 depends on MMU 696 select XXHASH 697 help 698 Enable Kernel Samepage Merging: KSM periodically scans those areas 699 of an application's address space that an app has advised may be 700 mergeable. When it finds pages of identical content, it replaces 701 the many instances by a single page with that content, so 702 saving memory until one or another app needs to modify the content. 703 Recommended for use with KVM, or with other duplicative applications. 704 See Documentation/mm/ksm.rst for more information: KSM is inactive 705 until a program has madvised that an area is MADV_MERGEABLE, and 706 root has set /sys/kernel/mm/ksm/run to 1 (if CONFIG_SYSFS is set). 707 708config DEFAULT_MMAP_MIN_ADDR 709 int "Low address space to protect from user allocation" 710 depends on MMU 711 default 4096 712 help 713 This is the portion of low virtual memory which should be protected 714 from userspace allocation. Keeping a user from writing to low pages 715 can help reduce the impact of kernel NULL pointer bugs. 716 717 For most ppc64 and x86 users with lots of address space 718 a value of 65536 is reasonable and should cause no problems. 719 On arm and other archs it should not be higher than 32768. 720 Programs which use vm86 functionality or have some need to map 721 this low address space will need CAP_SYS_RAWIO or disable this 722 protection by setting the value to 0. 723 724 This value can be changed after boot using the 725 /proc/sys/vm/mmap_min_addr tunable. 726 727config ARCH_SUPPORTS_MEMORY_FAILURE 728 bool 729 730config MEMORY_FAILURE 731 depends on MMU 732 depends on ARCH_SUPPORTS_MEMORY_FAILURE 733 bool "Enable recovery from hardware memory errors" 734 select MEMORY_ISOLATION 735 select RAS 736 help 737 Enables code to recover from some memory failures on systems 738 with MCA recovery. This allows a system to continue running 739 even when some of its memory has uncorrected errors. This requires 740 special hardware support and typically ECC memory. 741 742config HWPOISON_INJECT 743 tristate "HWPoison pages injector" 744 depends on MEMORY_FAILURE && DEBUG_KERNEL && PROC_FS 745 select PROC_PAGE_MONITOR 746 747config NOMMU_INITIAL_TRIM_EXCESS 748 int "Turn on mmap() excess space trimming before booting" 749 depends on !MMU 750 default 1 751 help 752 The NOMMU mmap() frequently needs to allocate large contiguous chunks 753 of memory on which to store mappings, but it can only ask the system 754 allocator for chunks in 2^N*PAGE_SIZE amounts - which is frequently 755 more than it requires. To deal with this, mmap() is able to trim off 756 the excess and return it to the allocator. 757 758 If trimming is enabled, the excess is trimmed off and returned to the 759 system allocator, which can cause extra fragmentation, particularly 760 if there are a lot of transient processes. 761 762 If trimming is disabled, the excess is kept, but not used, which for 763 long-term mappings means that the space is wasted. 764 765 Trimming can be dynamically controlled through a sysctl option 766 (/proc/sys/vm/nr_trim_pages) which specifies the minimum number of 767 excess pages there must be before trimming should occur, or zero if 768 no trimming is to occur. 769 770 This option specifies the initial value of this option. The default 771 of 1 says that all excess pages should be trimmed. 772 773 See Documentation/admin-guide/mm/nommu-mmap.rst for more information. 774 775config ARCH_WANT_GENERAL_HUGETLB 776 bool 777 778config ARCH_WANTS_THP_SWAP 779 def_bool n 780 781menuconfig TRANSPARENT_HUGEPAGE 782 bool "Transparent Hugepage Support" 783 depends on HAVE_ARCH_TRANSPARENT_HUGEPAGE && !PREEMPT_RT 784 select COMPACTION 785 select XARRAY_MULTI 786 help 787 Transparent Hugepages allows the kernel to use huge pages and 788 huge tlb transparently to the applications whenever possible. 789 This feature can improve computing performance to certain 790 applications by speeding up page faults during memory 791 allocation, by reducing the number of tlb misses and by speeding 792 up the pagetable walking. 793 794 If memory constrained on embedded, you may want to say N. 795 796if TRANSPARENT_HUGEPAGE 797 798choice 799 prompt "Transparent Hugepage Support sysfs defaults" 800 depends on TRANSPARENT_HUGEPAGE 801 default TRANSPARENT_HUGEPAGE_ALWAYS 802 help 803 Selects the sysfs defaults for Transparent Hugepage Support. 804 805 config TRANSPARENT_HUGEPAGE_ALWAYS 806 bool "always" 807 help 808 Enabling Transparent Hugepage always, can increase the 809 memory footprint of applications without a guaranteed 810 benefit but it will work automatically for all applications. 811 812 config TRANSPARENT_HUGEPAGE_MADVISE 813 bool "madvise" 814 help 815 Enabling Transparent Hugepage madvise, will only provide a 816 performance improvement benefit to the applications using 817 madvise(MADV_HUGEPAGE) but it won't risk to increase the 818 memory footprint of applications without a guaranteed 819 benefit. 820 821 config TRANSPARENT_HUGEPAGE_NEVER 822 bool "never" 823 help 824 Disable Transparent Hugepage by default. It can still be 825 enabled at runtime via sysfs. 826endchoice 827 828config THP_SWAP 829 def_bool y 830 depends on TRANSPARENT_HUGEPAGE && ARCH_WANTS_THP_SWAP && SWAP && 64BIT 831 help 832 Swap transparent huge pages in one piece, without splitting. 833 XXX: For now, swap cluster backing transparent huge page 834 will be split after swapout. 835 836 For selection by architectures with reasonable THP sizes. 837 838config READ_ONLY_THP_FOR_FS 839 bool "Read-only THP for filesystems (EXPERIMENTAL)" 840 depends on TRANSPARENT_HUGEPAGE && SHMEM 841 842 help 843 Allow khugepaged to put read-only file-backed pages in THP. 844 845 This is marked experimental because it is a new feature. Write 846 support of file THPs will be developed in the next few release 847 cycles. 848 849endif # TRANSPARENT_HUGEPAGE 850 851# 852# The architecture supports pgtable leaves that is larger than PAGE_SIZE 853# 854config PGTABLE_HAS_HUGE_LEAVES 855 def_bool TRANSPARENT_HUGEPAGE || HUGETLB_PAGE 856 857# 858# UP and nommu archs use km based percpu allocator 859# 860config NEED_PER_CPU_KM 861 depends on !SMP || !MMU 862 bool 863 default y 864 865config NEED_PER_CPU_EMBED_FIRST_CHUNK 866 bool 867 868config NEED_PER_CPU_PAGE_FIRST_CHUNK 869 bool 870 871config USE_PERCPU_NUMA_NODE_ID 872 bool 873 874config HAVE_SETUP_PER_CPU_AREA 875 bool 876 877config CMA 878 bool "Contiguous Memory Allocator" 879 depends on MMU 880 select MIGRATION 881 select MEMORY_ISOLATION 882 help 883 This enables the Contiguous Memory Allocator which allows other 884 subsystems to allocate big physically-contiguous blocks of memory. 885 CMA reserves a region of memory and allows only movable pages to 886 be allocated from it. This way, the kernel can use the memory for 887 pagecache and when a subsystem requests for contiguous area, the 888 allocated pages are migrated away to serve the contiguous request. 889 890 If unsure, say "n". 891 892config CMA_DEBUGFS 893 bool "CMA debugfs interface" 894 depends on CMA && DEBUG_FS 895 help 896 Turns on the DebugFS interface for CMA. 897 898config CMA_SYSFS 899 bool "CMA information through sysfs interface" 900 depends on CMA && SYSFS 901 help 902 This option exposes some sysfs attributes to get information 903 from CMA. 904 905config CMA_AREAS 906 int "Maximum count of the CMA areas" 907 depends on CMA 908 default 20 if NUMA 909 default 8 910 help 911 CMA allows to create CMA areas for particular purpose, mainly, 912 used as device private area. This parameter sets the maximum 913 number of CMA area in the system. 914 915 If unsure, leave the default value "8" in UMA and "20" in NUMA. 916 917config MEM_SOFT_DIRTY 918 bool "Track memory changes" 919 depends on CHECKPOINT_RESTORE && HAVE_ARCH_SOFT_DIRTY && PROC_FS 920 select PROC_PAGE_MONITOR 921 help 922 This option enables memory changes tracking by introducing a 923 soft-dirty bit on pte-s. This bit it set when someone writes 924 into a page just as regular dirty bit, but unlike the latter 925 it can be cleared by hands. 926 927 See Documentation/admin-guide/mm/soft-dirty.rst for more details. 928 929config GENERIC_EARLY_IOREMAP 930 bool 931 932config STACK_MAX_DEFAULT_SIZE_MB 933 int "Default maximum user stack size for 32-bit processes (MB)" 934 default 100 935 range 8 2048 936 depends on STACK_GROWSUP && (!64BIT || COMPAT) 937 help 938 This is the maximum stack size in Megabytes in the VM layout of 32-bit 939 user processes when the stack grows upwards (currently only on parisc 940 arch) when the RLIMIT_STACK hard limit is unlimited. 941 942 A sane initial value is 100 MB. 943 944config DEFERRED_STRUCT_PAGE_INIT 945 bool "Defer initialisation of struct pages to kthreads" 946 depends on SPARSEMEM 947 depends on !NEED_PER_CPU_KM 948 depends on 64BIT 949 select PADATA 950 help 951 Ordinarily all struct pages are initialised during early boot in a 952 single thread. On very large machines this can take a considerable 953 amount of time. If this option is set, large machines will bring up 954 a subset of memmap at boot and then initialise the rest in parallel. 955 This has a potential performance impact on tasks running early in the 956 lifetime of the system until these kthreads finish the 957 initialisation. 958 959config PAGE_IDLE_FLAG 960 bool 961 select PAGE_EXTENSION if !64BIT 962 help 963 This adds PG_idle and PG_young flags to 'struct page'. PTE Accessed 964 bit writers can set the state of the bit in the flags so that PTE 965 Accessed bit readers may avoid disturbance. 966 967config IDLE_PAGE_TRACKING 968 bool "Enable idle page tracking" 969 depends on SYSFS && MMU 970 select PAGE_IDLE_FLAG 971 help 972 This feature allows to estimate the amount of user pages that have 973 not been touched during a given period of time. This information can 974 be useful to tune memory cgroup limits and/or for job placement 975 within a compute cluster. 976 977 See Documentation/admin-guide/mm/idle_page_tracking.rst for 978 more details. 979 980# Architectures which implement cpu_dcache_is_aliasing() to query 981# whether the data caches are aliased (VIVT or VIPT with dcache 982# aliasing) need to select this. 983config ARCH_HAS_CPU_CACHE_ALIASING 984 bool 985 986config ARCH_HAS_CACHE_LINE_SIZE 987 bool 988 989config ARCH_HAS_CURRENT_STACK_POINTER 990 bool 991 help 992 In support of HARDENED_USERCOPY performing stack variable lifetime 993 checking, an architecture-agnostic way to find the stack pointer 994 is needed. Once an architecture defines an unsigned long global 995 register alias named "current_stack_pointer", this config can be 996 selected. 997 998config ARCH_HAS_PTE_DEVMAP 999 bool 1000 1001config ARCH_HAS_ZONE_DMA_SET 1002 bool 1003 1004config ZONE_DMA 1005 bool "Support DMA zone" if ARCH_HAS_ZONE_DMA_SET 1006 default y if ARM64 || X86 1007 1008config ZONE_DMA32 1009 bool "Support DMA32 zone" if ARCH_HAS_ZONE_DMA_SET 1010 depends on !X86_32 1011 default y if ARM64 1012 1013config ZONE_DEVICE 1014 bool "Device memory (pmem, HMM, etc...) hotplug support" 1015 depends on MEMORY_HOTPLUG 1016 depends on MEMORY_HOTREMOVE 1017 depends on SPARSEMEM_VMEMMAP 1018 depends on ARCH_HAS_PTE_DEVMAP 1019 select XARRAY_MULTI 1020 1021 help 1022 Device memory hotplug support allows for establishing pmem, 1023 or other device driver discovered memory regions, in the 1024 memmap. This allows pfn_to_page() lookups of otherwise 1025 "device-physical" addresses which is needed for using a DAX 1026 mapping in an O_DIRECT operation, among other things. 1027 1028 If FS_DAX is enabled, then say Y. 1029 1030# 1031# Helpers to mirror range of the CPU page tables of a process into device page 1032# tables. 1033# 1034config HMM_MIRROR 1035 bool 1036 depends on MMU 1037 1038config GET_FREE_REGION 1039 depends on SPARSEMEM 1040 bool 1041 1042config DEVICE_PRIVATE 1043 bool "Unaddressable device memory (GPU memory, ...)" 1044 depends on ZONE_DEVICE 1045 select GET_FREE_REGION 1046 1047 help 1048 Allows creation of struct pages to represent unaddressable device 1049 memory; i.e., memory that is only accessible from the device (or 1050 group of devices). You likely also want to select HMM_MIRROR. 1051 1052config VMAP_PFN 1053 bool 1054 1055config ARCH_USES_HIGH_VMA_FLAGS 1056 bool 1057config ARCH_HAS_PKEYS 1058 bool 1059 1060config ARCH_USES_PG_ARCH_X 1061 bool 1062 help 1063 Enable the definition of PG_arch_x page flags with x > 1. Only 1064 suitable for 64-bit architectures with CONFIG_FLATMEM or 1065 CONFIG_SPARSEMEM_VMEMMAP enabled, otherwise there may not be 1066 enough room for additional bits in page->flags. 1067 1068config VM_EVENT_COUNTERS 1069 default y 1070 bool "Enable VM event counters for /proc/vmstat" if EXPERT 1071 help 1072 VM event counters are needed for event counts to be shown. 1073 This option allows the disabling of the VM event counters 1074 on EXPERT systems. /proc/vmstat will only show page counts 1075 if VM event counters are disabled. 1076 1077config PERCPU_STATS 1078 bool "Collect percpu memory statistics" 1079 help 1080 This feature collects and exposes statistics via debugfs. The 1081 information includes global and per chunk statistics, which can 1082 be used to help understand percpu memory usage. 1083 1084config GUP_TEST 1085 bool "Enable infrastructure for get_user_pages()-related unit tests" 1086 depends on DEBUG_FS 1087 help 1088 Provides /sys/kernel/debug/gup_test, which in turn provides a way 1089 to make ioctl calls that can launch kernel-based unit tests for 1090 the get_user_pages*() and pin_user_pages*() family of API calls. 1091 1092 These tests include benchmark testing of the _fast variants of 1093 get_user_pages*() and pin_user_pages*(), as well as smoke tests of 1094 the non-_fast variants. 1095 1096 There is also a sub-test that allows running dump_page() on any 1097 of up to eight pages (selected by command line args) within the 1098 range of user-space addresses. These pages are either pinned via 1099 pin_user_pages*(), or pinned via get_user_pages*(), as specified 1100 by other command line arguments. 1101 1102 See tools/testing/selftests/mm/gup_test.c 1103 1104comment "GUP_TEST needs to have DEBUG_FS enabled" 1105 depends on !GUP_TEST && !DEBUG_FS 1106 1107config GUP_GET_PXX_LOW_HIGH 1108 bool 1109 1110config DMAPOOL_TEST 1111 tristate "Enable a module to run time tests on dma_pool" 1112 depends on HAS_DMA 1113 help 1114 Provides a test module that will allocate and free many blocks of 1115 various sizes and report how long it takes. This is intended to 1116 provide a consistent way to measure how changes to the 1117 dma_pool_alloc/free routines affect performance. 1118 1119config ARCH_HAS_PTE_SPECIAL 1120 bool 1121 1122# 1123# Some architectures require a special hugepage directory format that is 1124# required to support multiple hugepage sizes. For example a4fe3ce76 1125# "powerpc/mm: Allow more flexible layouts for hugepage pagetables" 1126# introduced it on powerpc. This allows for a more flexible hugepage 1127# pagetable layouts. 1128# 1129config ARCH_HAS_HUGEPD 1130 bool 1131 1132config MAPPING_DIRTY_HELPERS 1133 bool 1134 1135config KMAP_LOCAL 1136 bool 1137 1138config KMAP_LOCAL_NON_LINEAR_PTE_ARRAY 1139 bool 1140 1141# struct io_mapping based helper. Selected by drivers that need them 1142config IO_MAPPING 1143 bool 1144 1145config MEMFD_CREATE 1146 bool "Enable memfd_create() system call" if EXPERT 1147 1148config SECRETMEM 1149 default y 1150 bool "Enable memfd_secret() system call" if EXPERT 1151 depends on ARCH_HAS_SET_DIRECT_MAP 1152 help 1153 Enable the memfd_secret() system call with the ability to create 1154 memory areas visible only in the context of the owning process and 1155 not mapped to other processes and other kernel page tables. 1156 1157config ANON_VMA_NAME 1158 bool "Anonymous VMA name support" 1159 depends on PROC_FS && ADVISE_SYSCALLS && MMU 1160 1161 help 1162 Allow naming anonymous virtual memory areas. 1163 1164 This feature allows assigning names to virtual memory areas. Assigned 1165 names can be later retrieved from /proc/pid/maps and /proc/pid/smaps 1166 and help identifying individual anonymous memory areas. 1167 Assigning a name to anonymous virtual memory area might prevent that 1168 area from being merged with adjacent virtual memory areas due to the 1169 difference in their name. 1170 1171config HAVE_ARCH_USERFAULTFD_WP 1172 bool 1173 help 1174 Arch has userfaultfd write protection support 1175 1176config HAVE_ARCH_USERFAULTFD_MINOR 1177 bool 1178 help 1179 Arch has userfaultfd minor fault support 1180 1181menuconfig USERFAULTFD 1182 bool "Enable userfaultfd() system call" 1183 depends on MMU 1184 help 1185 Enable the userfaultfd() system call that allows to intercept and 1186 handle page faults in userland. 1187 1188if USERFAULTFD 1189config PTE_MARKER_UFFD_WP 1190 bool "Userfaultfd write protection support for shmem/hugetlbfs" 1191 default y 1192 depends on HAVE_ARCH_USERFAULTFD_WP 1193 1194 help 1195 Allows to create marker PTEs for userfaultfd write protection 1196 purposes. It is required to enable userfaultfd write protection on 1197 file-backed memory types like shmem and hugetlbfs. 1198endif # USERFAULTFD 1199 1200# multi-gen LRU { 1201config LRU_GEN 1202 bool "Multi-Gen LRU" 1203 depends on MMU 1204 # make sure folio->flags has enough spare bits 1205 depends on 64BIT || !SPARSEMEM || SPARSEMEM_VMEMMAP 1206 help 1207 A high performance LRU implementation to overcommit memory. See 1208 Documentation/admin-guide/mm/multigen_lru.rst for details. 1209 1210config LRU_GEN_ENABLED 1211 bool "Enable by default" 1212 depends on LRU_GEN 1213 help 1214 This option enables the multi-gen LRU by default. 1215 1216config LRU_GEN_STATS 1217 bool "Full stats for debugging" 1218 depends on LRU_GEN 1219 help 1220 Do not enable this option unless you plan to look at historical stats 1221 from evicted generations for debugging purpose. 1222 1223 This option has a per-memcg and per-node memory overhead. 1224 1225config LRU_GEN_WALKS_MMU 1226 def_bool y 1227 depends on LRU_GEN && ARCH_HAS_HW_PTE_YOUNG 1228# } 1229 1230config ARCH_SUPPORTS_PER_VMA_LOCK 1231 def_bool n 1232 1233config PER_VMA_LOCK 1234 def_bool y 1235 depends on ARCH_SUPPORTS_PER_VMA_LOCK && MMU && SMP 1236 help 1237 Allow per-vma locking during page fault handling. 1238 1239 This feature allows locking each virtual memory area separately when 1240 handling page faults instead of taking mmap_lock. 1241 1242config LOCK_MM_AND_FIND_VMA 1243 bool 1244 depends on !STACK_GROWSUP 1245 1246config IOMMU_MM_DATA 1247 bool 1248 1249config EXECMEM 1250 bool 1251 1252source "mm/damon/Kconfig" 1253 1254endmenu 1255