1config SELECT_MEMORY_MODEL 2 def_bool y 3 depends on ARCH_SELECT_MEMORY_MODEL 4 5choice 6 prompt "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 11 12config FLATMEM_MANUAL 13 bool "Flat Memory" 14 depends on !(ARCH_DISCONTIGMEM_ENABLE || ARCH_SPARSEMEM_ENABLE) || ARCH_FLATMEM_ENABLE 15 help 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 19 and a correct option. 20 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". 28 29 If unsure, choose this option (Flat Memory) over any other. 30 31config DISCONTIGMEM_MANUAL 32 bool "Discontiguous Memory" 33 depends on ARCH_DISCONTIGMEM_ENABLE 34 help 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 41 this option imposes. 42 43 Many NUMA configurations will have this as the only option. 44 45 If unsure, choose "Flat Memory" over this option. 46 47config SPARSEMEM_MANUAL 48 bool "Sparse Memory" 49 depends on ARCH_SPARSEMEM_ENABLE 50 help 51 This will be the only option for some systems, including 52 memory hotplug systems. This is normal. 53 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. 58 59 If unsure, choose "Discontiguous Memory" or "Flat Memory" 60 over this option. 61 62endchoice 63 64config DISCONTIGMEM 65 def_bool y 66 depends on (!SELECT_MEMORY_MODEL && ARCH_DISCONTIGMEM_ENABLE) || DISCONTIGMEM_MANUAL 67 68config SPARSEMEM 69 def_bool y 70 depends on (!SELECT_MEMORY_MODEL && ARCH_SPARSEMEM_ENABLE) || SPARSEMEM_MANUAL 71 72config FLATMEM 73 def_bool y 74 depends on (!DISCONTIGMEM && !SPARSEMEM) || FLATMEM_MANUAL 75 76config FLAT_NODE_MEM_MAP 77 def_bool y 78 depends on !SPARSEMEM 79 80# 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. 84# 85config NEED_MULTIPLE_NODES 86 def_bool y 87 depends on DISCONTIGMEM || NUMA 88 89config HAVE_MEMORY_PRESENT 90 def_bool y 91 depends on ARCH_HAVE_MEMORY_PRESENT || SPARSEMEM 92 93# 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. 99# 100# This option will also potentially produce smaller runtime code 101# with gcc 3.4 and later. 102# 103config SPARSEMEM_STATIC 104 bool 105 106# 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. 110# 111config SPARSEMEM_EXTREME 112 def_bool y 113 depends on SPARSEMEM && !SPARSEMEM_STATIC 114 115config SPARSEMEM_VMEMMAP_ENABLE 116 bool 117 118config SPARSEMEM_ALLOC_MEM_MAP_TOGETHER 119 def_bool y 120 depends on SPARSEMEM && X86_64 121 122config SPARSEMEM_VMEMMAP 123 bool "Sparse Memory virtual memmap" 124 depends on SPARSEMEM && SPARSEMEM_VMEMMAP_ENABLE 125 default y 126 help 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. 130 131config HAVE_MEMBLOCK 132 bool 133 134config HAVE_MEMBLOCK_NODE_MAP 135 bool 136 137config HAVE_MEMBLOCK_PHYS_MAP 138 bool 139 140config HAVE_GENERIC_RCU_GUP 141 bool 142 143config ARCH_DISCARD_MEMBLOCK 144 bool 145 146config NO_BOOTMEM 147 bool 148 149config MEMORY_ISOLATION 150 bool 151 152config MOVABLE_NODE 153 bool "Enable to assign a node which has only movable memory" 154 depends on HAVE_MEMBLOCK 155 depends on NO_BOOTMEM 156 depends on X86_64 157 depends on NUMA 158 default n 159 help 160 Allow a node to have only movable memory. Pages used by the kernel, 161 such as direct mapping pages cannot be migrated. So the corresponding 162 memory device cannot be hotplugged. This option allows the following 163 two things: 164 - When the system is booting, node full of hotpluggable memory can 165 be arranged to have only movable memory so that the whole node can 166 be hot-removed. (need movable_node boot option specified). 167 - After the system is up, the option allows users to online all the 168 memory of a node as movable memory so that the whole node can be 169 hot-removed. 170 171 Users who don't use the memory hotplug feature are fine with this 172 option on since they don't specify movable_node boot option or they 173 don't online memory as movable. 174 175 Say Y here if you want to hotplug a whole node. 176 Say N here if you want kernel to use memory on all nodes evenly. 177 178# 179# Only be set on architectures that have completely implemented memory hotplug 180# feature. If you are not sure, don't touch it. 181# 182config HAVE_BOOTMEM_INFO_NODE 183 def_bool n 184 185# eventually, we can have this option just 'select SPARSEMEM' 186config MEMORY_HOTPLUG 187 bool "Allow for memory hot-add" 188 depends on SPARSEMEM || X86_64_ACPI_NUMA 189 depends on ARCH_ENABLE_MEMORY_HOTPLUG 190 191config MEMORY_HOTPLUG_SPARSE 192 def_bool y 193 depends on SPARSEMEM && MEMORY_HOTPLUG 194 195config MEMORY_HOTREMOVE 196 bool "Allow for memory hot remove" 197 select MEMORY_ISOLATION 198 select HAVE_BOOTMEM_INFO_NODE if (X86_64 || PPC64) 199 depends on MEMORY_HOTPLUG && ARCH_ENABLE_MEMORY_HOTREMOVE 200 depends on MIGRATION 201 202# Heavily threaded applications may benefit from splitting the mm-wide 203# page_table_lock, so that faults on different parts of the user address 204# space can be handled with less contention: split it at this NR_CPUS. 205# Default to 4 for wider testing, though 8 might be more appropriate. 206# ARM's adjust_pte (unused if VIPT) depends on mm-wide page_table_lock. 207# PA-RISC 7xxx's spinlock_t would enlarge struct page from 32 to 44 bytes. 208# DEBUG_SPINLOCK and DEBUG_LOCK_ALLOC spinlock_t also enlarge struct page. 209# 210config SPLIT_PTLOCK_CPUS 211 int 212 default "999999" if !MMU 213 default "999999" if ARM && !CPU_CACHE_VIPT 214 default "999999" if PARISC && !PA20 215 default "4" 216 217config ARCH_ENABLE_SPLIT_PMD_PTLOCK 218 bool 219 220# 221# support for memory balloon 222config MEMORY_BALLOON 223 bool 224 225# 226# support for memory balloon compaction 227config BALLOON_COMPACTION 228 bool "Allow for balloon memory compaction/migration" 229 def_bool y 230 depends on COMPACTION && MEMORY_BALLOON 231 help 232 Memory fragmentation introduced by ballooning might reduce 233 significantly the number of 2MB contiguous memory blocks that can be 234 used within a guest, thus imposing performance penalties associated 235 with the reduced number of transparent huge pages that could be used 236 by the guest workload. Allowing the compaction & migration for memory 237 pages enlisted as being part of memory balloon devices avoids the 238 scenario aforementioned and helps improving memory defragmentation. 239 240# 241# support for memory compaction 242config COMPACTION 243 bool "Allow for memory compaction" 244 def_bool y 245 select MIGRATION 246 depends on MMU 247 help 248 Allows the compaction of memory for the allocation of huge pages. 249 250# 251# support for page migration 252# 253config MIGRATION 254 bool "Page migration" 255 def_bool y 256 depends on (NUMA || ARCH_ENABLE_MEMORY_HOTREMOVE || COMPACTION || CMA) && MMU 257 help 258 Allows the migration of the physical location of pages of processes 259 while the virtual addresses are not changed. This is useful in 260 two situations. The first is on NUMA systems to put pages nearer 261 to the processors accessing. The second is when allocating huge 262 pages as migration can relocate pages to satisfy a huge page 263 allocation instead of reclaiming. 264 265config ARCH_ENABLE_HUGEPAGE_MIGRATION 266 bool 267 268config PHYS_ADDR_T_64BIT 269 def_bool 64BIT || ARCH_PHYS_ADDR_T_64BIT 270 271config ZONE_DMA_FLAG 272 int 273 default "0" if !ZONE_DMA 274 default "1" 275 276config BOUNCE 277 bool "Enable bounce buffers" 278 default y 279 depends on BLOCK && MMU && (ZONE_DMA || HIGHMEM) 280 help 281 Enable bounce buffers for devices that cannot access 282 the full range of memory available to the CPU. Enabled 283 by default when ZONE_DMA or HIGHMEM is selected, but you 284 may say n to override this. 285 286# On the 'tile' arch, USB OHCI needs the bounce pool since tilegx will often 287# have more than 4GB of memory, but we don't currently use the IOTLB to present 288# a 32-bit address to OHCI. So we need to use a bounce pool instead. 289config NEED_BOUNCE_POOL 290 bool 291 default y if TILE && USB_OHCI_HCD 292 293config NR_QUICK 294 int 295 depends on QUICKLIST 296 default "2" if AVR32 297 default "1" 298 299config VIRT_TO_BUS 300 bool 301 help 302 An architecture should select this if it implements the 303 deprecated interface virt_to_bus(). All new architectures 304 should probably not select this. 305 306 307config MMU_NOTIFIER 308 bool 309 select SRCU 310 311config KSM 312 bool "Enable KSM for page merging" 313 depends on MMU 314 help 315 Enable Kernel Samepage Merging: KSM periodically scans those areas 316 of an application's address space that an app has advised may be 317 mergeable. When it finds pages of identical content, it replaces 318 the many instances by a single page with that content, so 319 saving memory until one or another app needs to modify the content. 320 Recommended for use with KVM, or with other duplicative applications. 321 See Documentation/vm/ksm.txt for more information: KSM is inactive 322 until a program has madvised that an area is MADV_MERGEABLE, and 323 root has set /sys/kernel/mm/ksm/run to 1 (if CONFIG_SYSFS is set). 324 325config DEFAULT_MMAP_MIN_ADDR 326 int "Low address space to protect from user allocation" 327 depends on MMU 328 default 4096 329 help 330 This is the portion of low virtual memory which should be protected 331 from userspace allocation. Keeping a user from writing to low pages 332 can help reduce the impact of kernel NULL pointer bugs. 333 334 For most ia64, ppc64 and x86 users with lots of address space 335 a value of 65536 is reasonable and should cause no problems. 336 On arm and other archs it should not be higher than 32768. 337 Programs which use vm86 functionality or have some need to map 338 this low address space will need CAP_SYS_RAWIO or disable this 339 protection by setting the value to 0. 340 341 This value can be changed after boot using the 342 /proc/sys/vm/mmap_min_addr tunable. 343 344config ARCH_SUPPORTS_MEMORY_FAILURE 345 bool 346 347config MEMORY_FAILURE 348 depends on MMU 349 depends on ARCH_SUPPORTS_MEMORY_FAILURE 350 bool "Enable recovery from hardware memory errors" 351 select MEMORY_ISOLATION 352 select RAS 353 help 354 Enables code to recover from some memory failures on systems 355 with MCA recovery. This allows a system to continue running 356 even when some of its memory has uncorrected errors. This requires 357 special hardware support and typically ECC memory. 358 359config HWPOISON_INJECT 360 tristate "HWPoison pages injector" 361 depends on MEMORY_FAILURE && DEBUG_KERNEL && PROC_FS 362 select PROC_PAGE_MONITOR 363 364config NOMMU_INITIAL_TRIM_EXCESS 365 int "Turn on mmap() excess space trimming before booting" 366 depends on !MMU 367 default 1 368 help 369 The NOMMU mmap() frequently needs to allocate large contiguous chunks 370 of memory on which to store mappings, but it can only ask the system 371 allocator for chunks in 2^N*PAGE_SIZE amounts - which is frequently 372 more than it requires. To deal with this, mmap() is able to trim off 373 the excess and return it to the allocator. 374 375 If trimming is enabled, the excess is trimmed off and returned to the 376 system allocator, which can cause extra fragmentation, particularly 377 if there are a lot of transient processes. 378 379 If trimming is disabled, the excess is kept, but not used, which for 380 long-term mappings means that the space is wasted. 381 382 Trimming can be dynamically controlled through a sysctl option 383 (/proc/sys/vm/nr_trim_pages) which specifies the minimum number of 384 excess pages there must be before trimming should occur, or zero if 385 no trimming is to occur. 386 387 This option specifies the initial value of this option. The default 388 of 1 says that all excess pages should be trimmed. 389 390 See Documentation/nommu-mmap.txt for more information. 391 392config TRANSPARENT_HUGEPAGE 393 bool "Transparent Hugepage Support" 394 depends on HAVE_ARCH_TRANSPARENT_HUGEPAGE 395 select COMPACTION 396 help 397 Transparent Hugepages allows the kernel to use huge pages and 398 huge tlb transparently to the applications whenever possible. 399 This feature can improve computing performance to certain 400 applications by speeding up page faults during memory 401 allocation, by reducing the number of tlb misses and by speeding 402 up the pagetable walking. 403 404 If memory constrained on embedded, you may want to say N. 405 406choice 407 prompt "Transparent Hugepage Support sysfs defaults" 408 depends on TRANSPARENT_HUGEPAGE 409 default TRANSPARENT_HUGEPAGE_ALWAYS 410 help 411 Selects the sysfs defaults for Transparent Hugepage Support. 412 413 config TRANSPARENT_HUGEPAGE_ALWAYS 414 bool "always" 415 help 416 Enabling Transparent Hugepage always, can increase the 417 memory footprint of applications without a guaranteed 418 benefit but it will work automatically for all applications. 419 420 config TRANSPARENT_HUGEPAGE_MADVISE 421 bool "madvise" 422 help 423 Enabling Transparent Hugepage madvise, will only provide a 424 performance improvement benefit to the applications using 425 madvise(MADV_HUGEPAGE) but it won't risk to increase the 426 memory footprint of applications without a guaranteed 427 benefit. 428endchoice 429 430# 431# UP and nommu archs use km based percpu allocator 432# 433config NEED_PER_CPU_KM 434 depends on !SMP 435 bool 436 default y 437 438config CLEANCACHE 439 bool "Enable cleancache driver to cache clean pages if tmem is present" 440 default n 441 help 442 Cleancache can be thought of as a page-granularity victim cache 443 for clean pages that the kernel's pageframe replacement algorithm 444 (PFRA) would like to keep around, but can't since there isn't enough 445 memory. So when the PFRA "evicts" a page, it first attempts to use 446 cleancache code to put the data contained in that page into 447 "transcendent memory", memory that is not directly accessible or 448 addressable by the kernel and is of unknown and possibly 449 time-varying size. And when a cleancache-enabled 450 filesystem wishes to access a page in a file on disk, it first 451 checks cleancache to see if it already contains it; if it does, 452 the page is copied into the kernel and a disk access is avoided. 453 When a transcendent memory driver is available (such as zcache or 454 Xen transcendent memory), a significant I/O reduction 455 may be achieved. When none is available, all cleancache calls 456 are reduced to a single pointer-compare-against-NULL resulting 457 in a negligible performance hit. 458 459 If unsure, say Y to enable cleancache 460 461config FRONTSWAP 462 bool "Enable frontswap to cache swap pages if tmem is present" 463 depends on SWAP 464 default n 465 help 466 Frontswap is so named because it can be thought of as the opposite 467 of a "backing" store for a swap device. The data is stored into 468 "transcendent memory", memory that is not directly accessible or 469 addressable by the kernel and is of unknown and possibly 470 time-varying size. When space in transcendent memory is available, 471 a significant swap I/O reduction may be achieved. When none is 472 available, all frontswap calls are reduced to a single pointer- 473 compare-against-NULL resulting in a negligible performance hit 474 and swap data is stored as normal on the matching swap device. 475 476 If unsure, say Y to enable frontswap. 477 478config CMA 479 bool "Contiguous Memory Allocator" 480 depends on HAVE_MEMBLOCK && MMU 481 select MIGRATION 482 select MEMORY_ISOLATION 483 help 484 This enables the Contiguous Memory Allocator which allows other 485 subsystems to allocate big physically-contiguous blocks of memory. 486 CMA reserves a region of memory and allows only movable pages to 487 be allocated from it. This way, the kernel can use the memory for 488 pagecache and when a subsystem requests for contiguous area, the 489 allocated pages are migrated away to serve the contiguous request. 490 491 If unsure, say "n". 492 493config CMA_DEBUG 494 bool "CMA debug messages (DEVELOPMENT)" 495 depends on DEBUG_KERNEL && CMA 496 help 497 Turns on debug messages in CMA. This produces KERN_DEBUG 498 messages for every CMA call as well as various messages while 499 processing calls such as dma_alloc_from_contiguous(). 500 This option does not affect warning and error messages. 501 502config CMA_DEBUGFS 503 bool "CMA debugfs interface" 504 depends on CMA && DEBUG_FS 505 help 506 Turns on the DebugFS interface for CMA. 507 508config CMA_AREAS 509 int "Maximum count of the CMA areas" 510 depends on CMA 511 default 7 512 help 513 CMA allows to create CMA areas for particular purpose, mainly, 514 used as device private area. This parameter sets the maximum 515 number of CMA area in the system. 516 517 If unsure, leave the default value "7". 518 519config MEM_SOFT_DIRTY 520 bool "Track memory changes" 521 depends on CHECKPOINT_RESTORE && HAVE_ARCH_SOFT_DIRTY && PROC_FS 522 select PROC_PAGE_MONITOR 523 help 524 This option enables memory changes tracking by introducing a 525 soft-dirty bit on pte-s. This bit it set when someone writes 526 into a page just as regular dirty bit, but unlike the latter 527 it can be cleared by hands. 528 529 See Documentation/vm/soft-dirty.txt for more details. 530 531config ZSWAP 532 bool "Compressed cache for swap pages (EXPERIMENTAL)" 533 depends on FRONTSWAP && CRYPTO=y 534 select CRYPTO_LZO 535 select ZPOOL 536 default n 537 help 538 A lightweight compressed cache for swap pages. It takes 539 pages that are in the process of being swapped out and attempts to 540 compress them into a dynamically allocated RAM-based memory pool. 541 This can result in a significant I/O reduction on swap device and, 542 in the case where decompressing from RAM is faster that swap device 543 reads, can also improve workload performance. 544 545 This is marked experimental because it is a new feature (as of 546 v3.11) that interacts heavily with memory reclaim. While these 547 interactions don't cause any known issues on simple memory setups, 548 they have not be fully explored on the large set of potential 549 configurations and workloads that exist. 550 551config ZPOOL 552 tristate "Common API for compressed memory storage" 553 default n 554 help 555 Compressed memory storage API. This allows using either zbud or 556 zsmalloc. 557 558config ZBUD 559 tristate "Low density storage for compressed pages" 560 default n 561 help 562 A special purpose allocator for storing compressed pages. 563 It is designed to store up to two compressed pages per physical 564 page. While this design limits storage density, it has simple and 565 deterministic reclaim properties that make it preferable to a higher 566 density approach when reclaim will be used. 567 568config ZSMALLOC 569 tristate "Memory allocator for compressed pages" 570 depends on MMU 571 default n 572 help 573 zsmalloc is a slab-based memory allocator designed to store 574 compressed RAM pages. zsmalloc uses virtual memory mapping 575 in order to reduce fragmentation. However, this results in a 576 non-standard allocator interface where a handle, not a pointer, is 577 returned by an alloc(). This handle must be mapped in order to 578 access the allocated space. 579 580config PGTABLE_MAPPING 581 bool "Use page table mapping to access object in zsmalloc" 582 depends on ZSMALLOC 583 help 584 By default, zsmalloc uses a copy-based object mapping method to 585 access allocations that span two pages. However, if a particular 586 architecture (ex, ARM) performs VM mapping faster than copying, 587 then you should select this. This causes zsmalloc to use page table 588 mapping rather than copying for object mapping. 589 590 You can check speed with zsmalloc benchmark: 591 https://github.com/spartacus06/zsmapbench 592 593config ZSMALLOC_STAT 594 bool "Export zsmalloc statistics" 595 depends on ZSMALLOC 596 select DEBUG_FS 597 help 598 This option enables code in the zsmalloc to collect various 599 statistics about whats happening in zsmalloc and exports that 600 information to userspace via debugfs. 601 If unsure, say N. 602 603config GENERIC_EARLY_IOREMAP 604 bool 605 606config MAX_STACK_SIZE_MB 607 int "Maximum user stack size for 32-bit processes (MB)" 608 default 80 609 range 8 256 if METAG 610 range 8 2048 611 depends on STACK_GROWSUP && (!64BIT || COMPAT) 612 help 613 This is the maximum stack size in Megabytes in the VM layout of 32-bit 614 user processes when the stack grows upwards (currently only on parisc 615 and metag arch). The stack will be located at the highest memory 616 address minus the given value, unless the RLIMIT_STACK hard limit is 617 changed to a smaller value in which case that is used. 618 619 A sane initial value is 80 MB. 620 621# For architectures that support deferred memory initialisation 622config ARCH_SUPPORTS_DEFERRED_STRUCT_PAGE_INIT 623 bool 624 625config DEFERRED_STRUCT_PAGE_INIT 626 bool "Defer initialisation of struct pages to kthreads" 627 default n 628 depends on ARCH_SUPPORTS_DEFERRED_STRUCT_PAGE_INIT 629 depends on MEMORY_HOTPLUG 630 help 631 Ordinarily all struct pages are initialised during early boot in a 632 single thread. On very large machines this can take a considerable 633 amount of time. If this option is set, large machines will bring up 634 a subset of memmap at boot and then initialise the rest in parallel 635 by starting one-off "pgdatinitX" kernel thread for each node X. This 636 has a potential performance impact on processes running early in the 637 lifetime of the system until these kthreads finish the 638 initialisation. 639 640config IDLE_PAGE_TRACKING 641 bool "Enable idle page tracking" 642 depends on SYSFS && MMU 643 select PAGE_EXTENSION if !64BIT 644 help 645 This feature allows to estimate the amount of user pages that have 646 not been touched during a given period of time. This information can 647 be useful to tune memory cgroup limits and/or for job placement 648 within a compute cluster. 649 650 See Documentation/vm/idle_page_tracking.txt for more details. 651 652config ZONE_DEVICE 653 bool "Device memory (pmem, etc...) hotplug support" if EXPERT 654 depends on MEMORY_HOTPLUG 655 depends on MEMORY_HOTREMOVE 656 depends on SPARSEMEM_VMEMMAP 657 depends on X86_64 #arch_add_memory() comprehends device memory 658 659 help 660 Device memory hotplug support allows for establishing pmem, 661 or other device driver discovered memory regions, in the 662 memmap. This allows pfn_to_page() lookups of otherwise 663 "device-physical" addresses which is needed for using a DAX 664 mapping in an O_DIRECT operation, among other things. 665 666 If FS_DAX is enabled, then say Y. 667 668config FRAME_VECTOR 669 bool 670 671config ARCH_USES_HIGH_VMA_FLAGS 672 bool 673config ARCH_HAS_PKEYS 674 bool 675